paperID
stringlengths 36
36
| pwc_id
stringlengths 8
47
| arxiv_id
stringlengths 6
16
⌀ | nips_id
float64 | url_abs
stringlengths 18
329
| url_pdf
stringlengths 18
742
| title
stringlengths 8
325
| abstract
stringlengths 1
7.27k
⌀ | authors
stringlengths 2
7.06k
| published
stringlengths 10
10
⌀ | conference
stringlengths 12
47
⌀ | conference_url_abs
stringlengths 16
198
⌀ | conference_url_pdf
stringlengths 27
199
⌀ | proceeding
stringlengths 6
47
⌀ | taskID
stringlengths 7
1.44k
| areaID
stringclasses 688
values | embedding
stringlengths 9.26k
12.5k
| umap_embedding
stringlengths 29
44
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
8c8ffc82-eb59-4ee6-9cba-f488b5fed1bb | lung-nodule-segmentation-and-low-confidence | 2303.08416 | null | https://arxiv.org/abs/2303.08416v4 | https://arxiv.org/pdf/2303.08416v4.pdf | Lung Nodule Segmentation and Low-Confidence Region Prediction with Uncertainty-Aware Attention Mechanism | Radiologists have different training and clinical experiences, which may result in various segmentation annotations for lung nodules, causing segmentation uncertainty. Conventional methods usually select a single annotation as the learning target or try to learn a latent space of various annotations, but these approaches waste the valuable information of consensus or disagreements ingrained in the multiple annotations. In this paper, we propose an Uncertainty-Aware Attention Mechanism (UAAM) that utilizes consensus and disagreements among multiple annotations to facilitate better segmentation. To achieve this, we introduce the Multi-Confidence Mask (MCM), which is a combination of a Low-Confidence (LC) Mask and a High-Confidence (HC) Mask. The LC mask indicates regions with a low segmentation confidence, which may cause different segmentation options among radiologists. Following UAAM, we further design an Uncertainty-Guide Segmentation Network (UGS-Net), which contains three modules: a Feature Extracting Module that captures a general feature of a lung nodule, an Uncertainty-Aware Module that produces three features for the annotations' union, intersection, and annotation set, and an Intersection-Union Constraining Module that uses distances between the three features to balance the predictions of final segmentation, LC mask, and HC mask. To fully demonstrate the performance of our method, we propose a Complex Nodule Validation on LIDC-IDRI, which tests UGS-Net's segmentation performance on lung nodules that are difficult to segment using U-Net. Experimental results demonstrate that our method can significantly improve the segmentation performance on nodules with poor segmentation by U-Net. | ['S. Kevin Zhou', 'Xiaohong Zhang', 'Chen Liu', 'Zhulin An', 'Yue Zhang', 'Qiuli Wang', 'Han Yang'] | 2023-03-15 | null | null | null | null | ['lung-nodule-segmentation'] | ['medical'] | [ 1.03918649e-01 6.09717429e-01 -4.23396617e-01 -4.70238328e-01
-1.07945168e+00 -3.86451066e-01 1.39049575e-01 -1.23723589e-01
-1.70330271e-01 5.35371661e-01 2.93807238e-01 -3.43346596e-01
-9.12262946e-02 -5.99245727e-01 -2.89461613e-01 -9.33870196e-01
4.48024511e-01 5.70808828e-01 8.44516098e-01 4.63143080e-01
-2.26414263e-01 1.54137835e-01 -8.87383878e-01 5.39882064e-01
1.22659600e+00 1.14570367e+00 4.13878113e-01 2.30720758e-01
-2.25891307e-01 5.41721523e-01 -3.74131918e-01 -7.62312934e-02
1.90296113e-01 -3.93643200e-01 -9.01256263e-01 2.46280402e-01
-2.32766807e-01 -2.43389249e-01 2.66169067e-02 1.25591958e+00
2.94139266e-01 -2.46289223e-01 8.91197860e-01 -1.02620673e+00
-3.37479234e-01 1.02528584e+00 -5.13377666e-01 -9.99512449e-02
-1.85891181e-01 3.17769498e-01 8.91222060e-01 -6.29194558e-01
2.80604035e-01 1.05649233e+00 7.53138542e-01 4.16735619e-01
-8.65645111e-01 -6.34158313e-01 1.97697416e-01 -2.49466479e-01
-1.49010980e+00 1.33653088e-02 4.60893959e-01 -6.92149460e-01
3.28849047e-01 3.42778265e-01 6.95636749e-01 6.39696658e-01
5.36235154e-01 9.97794449e-01 8.55857849e-01 -1.01966940e-01
2.49074832e-01 3.87157381e-01 6.46253750e-02 8.94973636e-01
3.79782468e-01 -2.05740586e-01 2.99519271e-01 -1.91783413e-01
9.40881848e-01 3.56243439e-02 -5.25476992e-01 -3.44058812e-01
-1.38701510e+00 7.33436167e-01 7.60509491e-01 2.62281358e-01
-2.19590291e-01 -1.12058491e-01 2.21400753e-01 -3.32594097e-01
2.59558529e-01 3.78243208e-01 -4.28805083e-01 4.75885153e-01
-9.80180085e-01 -3.11436981e-01 8.69435608e-01 9.83602941e-01
6.59037352e-01 -4.45157081e-01 -1.03925145e+00 6.87169075e-01
7.51251817e-01 9.27962065e-02 9.15242910e-01 -7.68641591e-01
8.69007930e-02 8.19792211e-01 -2.56087501e-02 -7.17618227e-01
-3.26260835e-01 -4.73580778e-01 -1.00182581e+00 -4.37833667e-02
1.30740017e-01 -2.47359589e-01 -1.35275173e+00 1.47586846e+00
2.73570418e-01 2.79717386e-01 -1.61142319e-01 9.94604468e-01
9.96047199e-01 1.94699094e-01 3.34387720e-02 -4.78171825e-01
1.34253001e+00 -1.35605729e+00 -7.18213022e-01 -1.05270289e-01
6.81468666e-01 -6.17469490e-01 8.29058707e-01 -1.30110877e-02
-8.07111859e-01 -4.76875693e-01 -9.02354896e-01 2.76239693e-01
6.31429330e-02 3.86325300e-01 3.29890043e-01 4.71736431e-01
-8.88765991e-01 3.67904037e-01 -1.00358403e+00 1.42898792e-02
5.87892234e-01 3.69027495e-01 -1.75780095e-02 1.14316233e-01
-1.18841445e+00 7.50630558e-01 7.19599307e-01 2.78721482e-01
-8.32509935e-01 -5.99826097e-01 -7.55702615e-01 1.25627741e-01
9.22462285e-01 -5.90636551e-01 1.41433167e+00 -8.01821768e-01
-1.35552883e+00 6.64079607e-01 4.11389619e-02 -1.42584488e-01
8.11308920e-01 2.26264372e-01 -1.76039040e-01 -1.18983880e-01
3.56210113e-01 9.26941574e-01 5.98093271e-01 -1.24353278e+00
-7.48907626e-01 3.91744636e-02 -3.99334431e-01 3.13435942e-01
1.60012275e-01 -4.69368577e-01 -9.91133988e-01 -5.39386332e-01
5.91537297e-01 -1.02090645e+00 -7.24406660e-01 -1.71760228e-02
-9.85022902e-01 -2.73628712e-01 8.14423919e-01 -3.54681402e-01
1.44030857e+00 -2.08799219e+00 -8.44163150e-02 5.90850532e-01
4.24116969e-01 1.98697537e-01 2.33176470e-01 -5.21612048e-01
7.23661408e-02 6.05579555e-01 -5.03497660e-01 1.02776084e-02
-1.49680749e-01 4.89023149e-01 2.62045741e-01 1.25560537e-01
3.08808982e-01 1.03219688e+00 -9.23461139e-01 -1.18410480e+00
1.84884220e-01 3.25779952e-02 -5.73600352e-01 3.61625999e-01
-4.58392352e-01 7.02619910e-01 -8.99588346e-01 8.06198239e-01
4.83932406e-01 -6.60744369e-01 1.77953497e-01 -5.34459412e-01
-9.05772597e-02 -9.96567905e-02 -1.23108363e+00 1.40456533e+00
6.06970675e-02 8.00718740e-02 1.47794485e-02 -4.26202714e-01
7.17466891e-01 4.80459750e-01 6.66590214e-01 1.44067660e-01
3.77260119e-01 3.50624681e-01 2.94379383e-01 -6.06182516e-01
-3.22901718e-02 5.99955469e-02 -5.57845123e-02 3.46661508e-01
-1.23816736e-01 -4.52302933e-01 -2.99534332e-02 5.87196834e-02
1.10753715e+00 -1.74898729e-01 5.31171858e-01 -4.32303309e-01
6.58374965e-01 -1.46703035e-01 1.06237364e+00 8.25003803e-01
-6.94089949e-01 8.73499036e-01 9.41430330e-01 -6.29442185e-02
-6.61465108e-01 -9.67216671e-01 -4.42108750e-01 3.90227109e-01
3.46061498e-01 -1.90599471e-01 -7.85704970e-01 -1.36288857e+00
-1.41289979e-01 5.09982169e-01 -8.02485168e-01 -1.57845646e-01
-2.38667682e-01 -6.07919395e-01 2.99519330e-01 6.73044741e-01
6.52310967e-01 -1.06969857e+00 -3.71111929e-01 6.95636570e-02
-3.23164731e-01 -7.54639506e-01 -9.25682068e-01 3.46168160e-01
-6.90073133e-01 -1.23924065e+00 -6.21868491e-01 -7.35517681e-01
9.47559953e-01 3.02335601e-02 1.01705730e+00 1.47942513e-01
-1.97268263e-01 1.42563939e-01 -1.71479866e-01 -4.15990084e-01
-4.25624162e-01 2.17473403e-01 -2.87399024e-01 -1.24352664e-01
1.61693990e-01 -5.64131737e-02 -6.08672738e-01 8.16994011e-01
-9.67560410e-01 2.99073428e-01 1.05543816e+00 9.76317883e-01
1.16250801e+00 1.16129190e-01 2.84013420e-01 -1.21856570e+00
4.38575655e-01 -7.19940364e-01 -5.36530316e-01 5.94697177e-01
-6.37353957e-01 6.88387826e-02 1.97074309e-01 -4.09502029e-01
-1.08140481e+00 4.01112080e-01 1.31802065e-02 -5.95904052e-01
-6.58694282e-02 6.67886853e-01 -3.00105840e-01 1.33407041e-01
4.52267468e-01 -1.15947872e-01 1.04506567e-01 -5.51047735e-02
1.73045814e-01 7.85799682e-01 4.55747068e-01 -3.21924537e-01
4.85019207e-01 2.07220316e-01 -3.57116789e-01 -7.06062615e-02
-1.34551013e+00 -6.44676626e-01 -8.44635129e-01 -1.63592651e-01
1.13873529e+00 -7.11677969e-01 -2.39787176e-01 1.59881279e-01
-8.58534098e-01 -2.19924480e-01 -4.94817406e-01 7.21481323e-01
-3.82978827e-01 2.25911453e-01 -4.74601448e-01 -4.37503964e-01
-2.59562850e-01 -1.82205725e+00 1.08809400e+00 7.34835923e-01
-2.16544524e-01 -8.74386311e-01 -1.78616717e-01 2.28107288e-01
2.66683042e-01 9.71132368e-02 8.64984572e-01 -1.06416285e+00
-8.27897310e-01 -6.32622540e-02 -5.02507508e-01 4.34475869e-01
5.03895104e-01 1.57170445e-01 -8.08560252e-01 -1.62778050e-01
4.37727235e-02 -2.77311623e-01 1.00280356e+00 8.01597476e-01
1.62773466e+00 -9.84004587e-02 -9.05399144e-01 5.82244456e-01
1.23796237e+00 2.53296226e-01 4.86079544e-01 -6.42557293e-02
7.26814151e-01 2.95878619e-01 7.17749357e-01 2.36459956e-01
9.51087549e-02 2.09648028e-01 6.52390420e-01 -2.82625407e-01
-8.18430185e-02 -3.00308503e-02 -8.32770392e-02 8.60505521e-01
1.47926539e-01 -3.22284639e-01 -1.03923678e+00 5.46951711e-01
-2.04695511e+00 -4.34307188e-01 -4.78934310e-02 1.80403805e+00
1.18760049e+00 3.45536947e-01 -4.89115894e-01 -3.79629999e-01
9.09051180e-01 9.54762697e-02 -7.39209712e-01 1.47750184e-01
4.11586910e-01 -2.66280770e-01 4.47589964e-01 3.63587230e-01
-1.31915081e+00 7.14978218e-01 6.11283541e+00 1.07510734e+00
-9.99752343e-01 -2.89264880e-02 9.85512853e-01 3.46223235e-01
-5.88499904e-01 -1.40322834e-01 -8.34443927e-01 6.05838895e-01
2.66649246e-01 4.30259630e-02 -3.33652198e-01 1.04091644e+00
4.97671105e-02 -2.62252271e-01 -1.17652929e+00 5.40491521e-01
-1.53894499e-01 -1.32400024e+00 -9.68184415e-03 -6.53540641e-02
1.06453717e+00 7.71734789e-02 2.11159941e-02 3.86413664e-01
5.21999896e-01 -1.09184921e+00 3.07720572e-01 7.14799166e-01
9.06670570e-01 -3.42410952e-01 1.28968716e+00 4.34597701e-01
-1.24242103e+00 1.33250207e-01 -3.91196281e-01 8.33455980e-01
-1.14531284e-02 8.32165360e-01 -1.31291771e+00 7.15724587e-01
5.75647175e-01 4.82962549e-01 -5.09026468e-01 1.31005192e+00
-3.25824678e-01 6.36489987e-01 -3.59554082e-01 9.46585312e-02
4.07463759e-01 -9.14793983e-02 4.66325819e-01 1.15701401e+00
3.50539148e-01 1.82559252e-01 7.29863763e-01 1.17992282e+00
-3.61388847e-02 3.65281031e-02 -1.62751880e-02 2.42951170e-01
5.88704944e-01 1.50237477e+00 -1.04740286e+00 -4.08864945e-01
-3.09924513e-01 6.43284798e-01 -2.62365770e-02 7.70716816e-02
-1.07360125e+00 -2.21002707e-03 9.98625755e-02 -6.11854382e-02
1.81713909e-01 4.10350204e-01 -3.99341345e-01 -9.21625793e-01
-2.89551646e-01 -4.72837001e-01 5.60247481e-01 -7.90433943e-01
-1.40849912e+00 6.65116727e-01 2.58650985e-02 -1.38958979e+00
6.95657134e-02 -4.12882864e-01 -8.54308963e-01 9.40366328e-01
-1.26590490e+00 -1.00372088e+00 -6.13663375e-01 3.20801824e-01
6.69125617e-01 -8.94843340e-02 5.52524686e-01 -4.21112999e-02
-7.48645306e-01 4.22920227e-01 -1.05768397e-01 2.74976254e-01
8.45522463e-01 -1.39053869e+00 -3.30562145e-01 5.27547419e-01
-2.87981510e-01 3.00477177e-01 1.42979994e-01 -1.01771879e+00
-4.27012026e-01 -1.59874094e+00 3.77261668e-01 -3.66106123e-01
4.19600308e-01 3.39639217e-01 -1.06794930e+00 9.01748180e-01
-1.61667511e-01 3.71584326e-01 7.89196134e-01 -2.59236842e-01
1.90529883e-01 1.51842490e-01 -1.19596887e+00 6.59810364e-01
5.79321146e-01 -7.84166157e-02 -6.16880357e-01 4.16567475e-01
1.11451650e+00 -7.12581336e-01 -9.91173685e-01 9.97778296e-01
4.29013759e-01 -8.33641052e-01 6.64165914e-01 -1.37182862e-01
3.53557199e-01 -5.67083120e-01 1.87022507e-01 -1.30246139e+00
-6.93928838e-01 -3.11770802e-03 2.34344944e-01 1.12402749e+00
7.41435170e-01 -3.85913193e-01 8.91470075e-01 8.70599508e-01
-5.87037921e-01 -1.30915451e+00 -7.24253356e-01 -2.79941499e-01
-8.83402228e-02 -3.57434958e-01 6.55310392e-01 9.60688293e-01
-2.70705998e-01 -1.27238348e-01 1.55469894e-01 3.47526789e-01
1.68351859e-01 9.13009886e-03 1.68209985e-01 -1.25297070e+00
-2.52903253e-01 -6.14394069e-01 -4.44081798e-03 -8.53012264e-01
-5.63899428e-02 -9.72124338e-01 5.62638521e-01 -1.77169156e+00
5.44741750e-01 -7.48432040e-01 -5.60425699e-01 7.31939256e-01
-3.85070801e-01 -1.22171842e-01 -2.75784791e-01 5.59236944e-01
-8.31197023e-01 2.98844665e-01 1.86038566e+00 -1.99280903e-01
-3.66244912e-01 4.28627014e-01 -7.32742190e-01 1.17174649e+00
6.24286652e-01 -4.97934192e-01 -4.64804381e-01 -7.35617727e-02
-2.77533263e-01 1.79235429e-01 4.07701582e-02 -8.91881108e-01
3.99632186e-01 -2.93469727e-01 7.01904416e-01 -1.00836861e+00
-3.06082755e-01 -8.68474305e-01 1.49320185e-01 6.02782249e-01
-4.10644323e-01 -6.85915351e-01 -2.93009523e-02 6.76334798e-01
-2.79345095e-01 -4.87467051e-01 8.63076508e-01 -4.69229817e-01
-4.28805828e-01 6.60903454e-01 -3.24351937e-01 -9.02332738e-02
1.25264060e+00 -2.45577171e-01 -1.13084808e-01 6.38655424e-02
-1.01864505e+00 8.91732037e-01 2.30145425e-01 9.92811248e-02
4.52263355e-01 -1.40372920e+00 -7.32586324e-01 3.52262706e-01
9.06172395e-02 8.84159207e-01 2.10909873e-01 1.11695433e+00
-4.64346826e-01 4.58785534e-01 1.45521611e-01 -1.12791193e+00
-1.01365495e+00 2.65740246e-01 7.30013609e-01 -7.82173157e-01
-3.01293612e-01 1.18002999e+00 5.30737579e-01 -5.00671268e-01
2.35674024e-01 -7.28877842e-01 -4.12567466e-01 -4.64399308e-02
3.17236856e-02 4.93167974e-02 -1.30430028e-01 -3.19941014e-01
-2.67823577e-01 3.28748554e-01 -2.76107192e-01 1.51615396e-01
5.25738120e-01 -1.85653642e-02 -3.00045103e-01 3.86066526e-01
6.83197975e-01 -7.76460534e-03 -1.50052285e+00 -4.13958341e-01
-1.19815534e-02 -2.27609247e-01 1.42722160e-01 -9.27339613e-01
-1.14043450e+00 4.12438810e-01 5.10252178e-01 1.74444705e-01
9.99502122e-01 3.65123957e-01 6.01974130e-01 4.17744368e-02
6.46988675e-02 -8.55505049e-01 1.63820878e-01 3.76869440e-01
8.17949653e-01 -1.50028384e+00 -6.99498132e-02 -7.39777684e-01
-9.09744859e-01 9.51134920e-01 1.13894022e+00 3.33738178e-01
9.43363845e-01 3.81133318e-01 1.54800579e-01 -1.18146606e-01
-6.77022278e-01 -2.67504275e-01 7.27471769e-01 2.45043039e-01
3.71768177e-01 2.83476114e-01 -3.16762537e-01 1.14853919e+00
1.28390923e-01 -4.33490193e-03 2.43857950e-01 6.70603991e-01
-7.13744283e-01 -9.22745824e-01 -3.07565063e-01 9.89847302e-01
-4.51276392e-01 8.30052048e-02 -2.80074596e-01 7.75528729e-01
6.36038780e-01 5.69779813e-01 1.04460865e-01 -4.05437022e-01
3.10694743e-02 8.15360982e-05 -2.72323955e-02 -1.05725026e+00
-5.11916816e-01 5.95986903e-01 -2.03017563e-01 -4.58214968e-01
-2.16839522e-01 -4.72870171e-01 -1.53941441e+00 4.44122523e-01
-9.91585433e-01 3.21545660e-01 8.14304873e-02 1.02424216e+00
-4.86356653e-02 1.02319920e+00 5.16709983e-01 -4.67386872e-01
-5.76282263e-01 -1.00794256e+00 -5.39494574e-01 6.67717084e-02
5.62217310e-02 -5.81408679e-01 -3.32263023e-01 7.27735553e-03] | [14.722505569458008, -2.124925374984741] |
fb5960bf-6885-4e6c-a922-ace6f2d348d0 | prediction-based-one-shot-dynamic-parking | 2208.14231 | null | https://arxiv.org/abs/2208.14231v1 | https://arxiv.org/pdf/2208.14231v1.pdf | Prediction-based One-shot Dynamic Parking Pricing | Many U.S. metropolitan cities are notorious for their severe shortage of parking spots. To this end, we present a proactive prediction-driven optimization framework to dynamically adjust parking prices. We use state-of-the-art deep learning technologies such as neural ordinary differential equations (NODEs) to design our future parking occupancy rate prediction model given historical occupancy rates and price information. Owing to the continuous and bijective characteristics of NODEs, in addition, we design a one-shot price optimization method given a pre-trained prediction model, which requires only one iteration to find the optimal solution. In other words, we optimize the price input to the pre-trained prediction model to achieve targeted occupancy rates in the parking blocks. We conduct experiments with the data collected in San Francisco and Seattle for years. Our prediction model shows the best accuracy in comparison with various temporal or spatio-temporal forecasting models. Our one-shot optimization method greatly outperforms other black-box and white-box search methods in terms of the search time and always returns the optimal price solution. | ['Noseong Park', 'Jeongwhan Choi', 'Heejoo Shin', 'Seoyoung Hong'] | 2022-08-30 | null | null | null | null | ['spatio-temporal-forecasting'] | ['time-series'] | [-7.40671337e-01 -1.24861397e-01 -5.90300918e-01 -4.44200367e-01
-4.70656425e-01 4.08186764e-01 2.85860419e-01 -3.09823871e-01
-6.25192106e-01 1.11005390e+00 8.11869577e-02 -8.43033433e-01
-2.23245293e-01 -1.18653584e+00 -6.70486152e-01 -5.51758289e-01
-1.09727025e-01 6.01990283e-01 -2.01396309e-02 -4.68901157e-01
2.29028419e-01 3.33707392e-01 -1.52191603e+00 -3.37458283e-01
1.34109378e+00 1.06169856e+00 5.90053737e-01 3.30474705e-01
-3.19884837e-01 4.67554569e-01 1.19782910e-02 -6.13321438e-02
6.14176571e-01 2.94388980e-01 -3.24402213e-01 -3.05204391e-01
-2.30745316e-01 -5.78988075e-01 -7.02793479e-01 5.34908772e-01
3.87465358e-01 2.38442868e-01 3.94000649e-01 -1.36469483e+00
-7.42174447e-01 6.45225108e-01 -4.85454798e-01 3.81208539e-01
-5.49667478e-01 4.31266934e-01 1.07531512e+00 -1.10677028e+00
-7.76570141e-02 8.45442116e-01 7.76584148e-01 4.58375901e-01
-1.12343895e+00 -6.07051194e-01 3.35768431e-01 2.83825248e-01
-1.71160197e+00 -4.63287979e-01 7.35799909e-01 -2.66935617e-01
1.34794641e+00 2.69049078e-01 9.13245678e-01 2.75951326e-01
4.64639425e-01 8.21328223e-01 5.17576993e-01 -4.74299304e-02
4.62150514e-01 8.66017416e-02 -4.86156255e-01 5.24661839e-01
-9.56938043e-02 3.67829472e-01 -2.38152191e-01 -1.69409469e-01
6.31997406e-01 4.57409412e-01 3.88082206e-01 -2.51517713e-01
-7.27541864e-01 9.64817524e-01 7.18643606e-01 6.75550327e-02
-6.99512601e-01 6.01852536e-01 -7.17519969e-02 -2.53344595e-01
5.82864702e-01 2.48681590e-01 -6.21882915e-01 -2.58715540e-01
-1.38230002e+00 4.62180436e-01 6.09063208e-01 6.26451373e-01
9.72546995e-01 4.83936101e-01 -5.87712340e-02 7.56154954e-01
5.24789095e-01 7.68100679e-01 4.64302063e-01 -1.09147549e+00
5.00497580e-01 3.91251802e-01 6.60501361e-01 -8.97944272e-01
-4.34717685e-01 -2.82327235e-01 -7.42490888e-01 -1.01960123e-01
9.08028409e-02 -4.10374492e-01 -8.17294955e-01 1.59660423e+00
7.53751472e-02 3.09193671e-01 -5.20818420e-02 8.43465865e-01
2.88667470e-01 9.78832185e-01 2.01958373e-01 -1.38804764e-01
8.59378695e-01 -1.09444010e+00 -4.30475384e-01 -5.53291380e-01
5.87529242e-01 -2.25374565e-01 1.01514387e+00 -3.35216671e-01
-1.15961385e+00 -3.79448414e-01 -7.87791252e-01 2.75078982e-01
-6.99075282e-01 7.60560483e-02 6.62739456e-01 2.79141694e-01
-1.19967365e+00 7.75929332e-01 -8.05971324e-01 -1.25754520e-01
5.15720487e-01 6.04228675e-01 5.92152476e-01 3.95192921e-01
-1.54667759e+00 9.86489475e-01 2.45675608e-03 1.44164413e-01
-5.00937998e-01 -1.07004511e+00 -7.68056929e-01 4.94175881e-01
5.20471819e-02 -3.87018442e-01 1.53636670e+00 -3.53302568e-01
-1.24987221e+00 3.92391622e-01 -4.34092045e-01 -8.20010364e-01
3.53129178e-01 2.38608181e-01 -5.05611479e-01 -6.14531755e-01
4.55259949e-01 1.06427026e+00 4.37072188e-01 -8.51764917e-01
-9.15533960e-01 -3.79510410e-02 -3.02455097e-01 2.25564718e-01
-4.55639541e-01 -6.16678894e-01 -4.13256943e-01 -2.32176006e-01
-2.79339939e-01 -7.31134176e-01 -9.43766057e-01 1.54795107e-02
-7.48174116e-02 -3.19073945e-01 7.29499042e-01 -6.23227835e-01
1.84425962e+00 -1.90342283e+00 -5.90613186e-01 2.28896827e-01
2.89541371e-02 3.30879726e-02 1.78373739e-01 2.61897564e-01
2.45626956e-01 4.17875089e-02 -7.57630393e-02 -6.15061104e-01
3.42480212e-01 6.45624101e-01 -3.39945912e-01 1.91828892e-01
2.02951953e-01 1.40950203e+00 -7.13615775e-01 -6.31161749e-01
5.82250476e-01 1.98288143e-01 -5.57240665e-01 -1.64040744e-01
-3.85973722e-01 -4.86355871e-02 -7.25370169e-01 7.56508708e-01
7.56454885e-01 -4.96930063e-01 -8.87177736e-02 3.75880182e-01
-7.03960776e-01 4.15688664e-01 -8.72667968e-01 1.15761864e+00
-8.69090915e-01 6.48018539e-01 8.14887602e-03 -1.20232856e+00
1.09725511e+00 -1.58891797e-01 1.00707269e+00 -1.39421773e+00
-1.36925504e-01 4.67436850e-01 -3.45986754e-01 -1.80759922e-01
9.76287901e-01 1.52454227e-01 -3.29838514e-01 4.06359822e-01
-8.51201534e-01 2.48986140e-01 4.00825366e-02 -1.37846231e-01
7.03460336e-01 -2.15454713e-01 -2.51734108e-02 -4.45180416e-01
3.73980671e-01 3.34842145e-01 7.83360481e-01 6.89762175e-01
-5.60099423e-01 9.21678171e-02 5.60942777e-02 -9.98995960e-01
-1.51272309e+00 -9.87684131e-01 -1.64664075e-01 8.89017284e-01
9.57670063e-02 4.93745357e-02 -4.24741030e-01 7.79190566e-03
4.92443085e-01 8.34333479e-01 -4.40747827e-01 7.28819594e-02
-6.94411874e-01 -7.78638005e-01 9.00075808e-02 6.02650762e-01
7.53416538e-01 -8.60729277e-01 -8.38995934e-01 5.77386200e-01
-3.13371681e-02 -7.83821404e-01 -5.57119668e-01 1.40838683e-01
-8.00301373e-01 -3.70890081e-01 -7.99534500e-01 -6.42540395e-01
5.88896990e-01 2.77748704e-01 1.04511631e+00 8.02579224e-02
-6.42714798e-02 -5.34179918e-02 4.45883512e-01 -3.49071354e-01
1.66365862e-01 2.61495590e-01 1.19218402e-01 -2.44602665e-01
8.08349490e-01 -6.11264229e-01 -9.77650762e-01 4.62768674e-01
-2.47004852e-01 1.22576535e-01 5.37376881e-01 7.08264768e-01
9.16546285e-01 3.67711306e-01 6.50949180e-01 -1.18437767e-01
7.48397052e-01 -7.57528961e-01 -1.14466918e+00 6.84583113e-02
-1.24309373e+00 2.48037845e-01 8.16222012e-01 -2.46999830e-01
-7.67484128e-01 3.82949412e-01 -1.40258055e-02 -5.81554890e-01
2.92295277e-01 6.02149248e-01 2.60788947e-01 1.67608321e-01
3.02285492e-01 4.50153589e-01 -1.65646933e-02 -3.11882526e-01
2.47643456e-01 8.30136716e-01 4.41490024e-01 -2.91688651e-01
7.07880080e-01 4.76403862e-01 -1.04334399e-01 -5.97254992e-01
-3.80521208e-01 -4.52507943e-01 -4.40614581e-01 -3.90934199e-01
6.15621984e-01 -1.00984657e+00 -9.10479724e-01 4.03078735e-01
-9.25219715e-01 -6.64523244e-01 -2.96601951e-01 2.13002525e-02
-6.16979659e-01 -1.02372438e-01 -2.03503102e-01 -1.17360353e+00
-5.78710437e-01 -9.64627445e-01 8.33477318e-01 6.26594782e-01
1.12441137e-01 -1.07794201e+00 7.71758631e-02 7.97989145e-02
1.07097006e+00 -2.76723117e-01 5.09088457e-01 -4.25175838e-02
-7.08164811e-01 -1.14277206e-01 -2.90411085e-01 -1.46710217e-01
-6.78353477e-03 -1.42351449e-01 -6.13993883e-01 9.38201472e-02
-5.41721106e-01 4.03475165e-01 7.80572414e-01 1.00797844e+00
1.43816006e+00 -4.97612178e-01 -6.92674518e-01 3.79263908e-01
1.51077783e+00 4.30267304e-01 6.95579112e-01 7.14921415e-01
2.09953308e-01 1.59096122e-01 6.98963225e-01 7.95591414e-01
9.97064114e-01 6.67853475e-01 6.05663657e-01 -2.46367663e-01
6.10310972e-01 -5.36523461e-01 9.03566703e-02 3.24831575e-01
3.12939972e-01 1.53189853e-01 -1.28247714e+00 1.03362060e+00
-2.33666921e+00 -1.07354772e+00 8.18776935e-02 2.07544684e+00
6.00250244e-01 4.08442706e-01 2.72620440e-01 -3.73665094e-01
5.46780884e-01 2.73688436e-01 -1.11185288e+00 -7.59631395e-01
6.73750937e-02 8.25624838e-02 1.36233187e+00 5.28726041e-01
-9.89156604e-01 9.99305964e-01 6.83199549e+00 7.21778154e-01
-1.22834659e+00 -2.72399802e-02 1.12351286e+00 -3.82235557e-01
-4.40579027e-01 -1.02451235e-01 -9.91013706e-01 9.79959190e-01
1.43892622e+00 -5.94878733e-01 8.07703137e-01 1.26725578e+00
1.01015496e+00 -1.38055861e-01 -5.32069445e-01 9.49405909e-01
-5.90737045e-01 -2.06552529e+00 -5.56034148e-01 5.43333709e-01
7.55194902e-01 5.55044651e-01 2.89810032e-01 5.95523715e-01
6.04532003e-01 -1.12420559e+00 3.67555261e-01 8.16953719e-01
4.87973779e-01 -8.24626625e-01 4.77845460e-01 5.71073771e-01
-1.51192224e+00 -4.28557366e-01 -5.70001960e-01 -5.50960898e-01
6.37964487e-01 7.16494143e-01 -6.85258806e-01 -1.40082091e-01
8.83334994e-01 6.17776096e-01 -1.24903047e-03 1.08865380e+00
3.32607508e-01 1.69703215e-01 -7.52976835e-01 -4.78445888e-01
6.73504889e-01 -1.46021008e-01 -7.03226868e-03 6.83072567e-01
5.30644715e-01 -2.32703350e-02 2.42063314e-01 1.42400098e+00
1.07200876e-01 -1.06278114e-01 -5.55102646e-01 1.28626823e-01
7.86301970e-01 9.88262236e-01 -4.45973240e-02 -2.21919119e-01
-5.43247461e-01 4.29118544e-01 1.83470607e-01 2.56517380e-01
-1.16187227e+00 -2.51912683e-01 1.12846446e+00 3.52739096e-01
6.61172986e-01 -4.05542374e-01 -6.69595838e-01 -7.14759409e-01
-1.53027177e-01 1.14010975e-01 -8.03950578e-02 -8.01807284e-01
-1.08530152e+00 1.64682735e-02 -1.15366466e-01 -1.03221071e+00
-3.23965609e-01 -4.75201190e-01 -1.13549685e+00 1.11040998e+00
-2.01828432e+00 -5.26708126e-01 4.50716447e-03 5.29056430e-01
6.70643568e-01 -2.10684836e-01 5.01441419e-01 6.53329253e-01
-8.09663594e-01 4.51159000e-01 3.55122328e-01 3.26678827e-02
-7.58588165e-02 -9.07724798e-01 8.68479729e-01 6.10092103e-01
-5.33935070e-01 -3.07421237e-02 6.75613165e-01 -5.68452060e-01
-1.34434283e+00 -1.12302947e+00 1.39877093e+00 1.10552646e-01
9.67320800e-01 -9.71080586e-02 -5.85214555e-01 2.63786703e-01
-4.34332974e-02 -5.23537062e-02 2.48317301e-01 -1.36153236e-01
4.94801491e-01 -4.23136413e-01 -1.04210293e+00 8.72952640e-01
7.26755977e-01 -2.01687917e-01 -6.97129369e-02 3.92055273e-01
6.78081214e-01 -2.84821033e-01 -6.10878348e-01 2.18300760e-01
4.86322373e-01 -6.24945581e-01 9.50000823e-01 -4.56124067e-01
3.85757297e-01 2.39048097e-02 -2.72092074e-01 -1.10028267e+00
-6.00797951e-01 -4.41374630e-01 -1.67356044e-01 6.75632000e-01
1.00869679e+00 -8.04475248e-01 1.44332290e+00 1.56167543e+00
-1.63446411e-01 -1.37541628e+00 -1.40093064e+00 -7.02279270e-01
2.04296008e-01 -5.20638287e-01 1.16453362e+00 5.38025498e-01
-1.01603419e-01 -1.86494395e-01 -6.95415199e-01 1.40253350e-01
5.72102308e-01 3.02806854e-01 6.69330597e-01 -1.01504052e+00
6.90838993e-02 -6.34693384e-01 -4.32806741e-03 -1.29107475e+00
2.91470945e-01 -5.19803762e-01 6.88100383e-02 -1.83964193e+00
-1.08290292e-01 -1.01372790e+00 -4.87632245e-01 5.80843449e-01
2.23010167e-01 -4.00705755e-01 -6.49412498e-02 3.64853561e-01
-4.23086673e-01 9.07749474e-01 9.57237363e-01 -3.70140314e-01
-6.21914148e-01 2.34727517e-01 -6.11832738e-01 2.41621047e-01
1.47086751e+00 -1.98415920e-01 -2.72677451e-01 -5.69581985e-01
2.17323273e-01 3.70592922e-02 2.57300466e-01 -8.63389850e-01
4.52133238e-01 -8.77547622e-01 2.64028072e-01 -1.17217875e+00
5.46243250e-01 -7.53084421e-01 1.92255095e-01 6.46362901e-01
1.66717991e-01 3.15863043e-01 4.22615290e-01 3.45006406e-01
1.52545094e-01 2.05628395e-01 6.53308272e-01 -1.93873718e-01
-1.22874951e+00 7.26833999e-01 -5.81902206e-01 -1.90438345e-01
1.12051141e+00 -5.11565089e-01 -3.72586459e-01 -4.53275144e-01
-4.85676467e-01 1.09540772e+00 2.59076029e-01 3.18613082e-01
5.96853912e-01 -1.57877946e+00 -2.74860412e-01 1.74523532e-01
-2.97138602e-01 -6.18330985e-02 2.64860958e-01 8.95995438e-01
-3.63243401e-01 9.59072948e-01 -1.05456032e-01 -5.21600246e-01
-3.95477772e-01 5.79339325e-01 8.50787818e-01 -3.85118544e-01
-4.49605048e-01 4.06116843e-01 -4.44245785e-01 -6.25940859e-01
1.04092076e-01 -2.41532147e-01 1.05860896e-01 -2.13904530e-01
1.68470457e-01 2.79926419e-01 -2.35314041e-01 -5.38465381e-01
-4.41302389e-01 2.38592029e-01 2.74479181e-01 -1.23572554e-02
1.54072452e+00 -3.25880557e-01 3.03217828e-01 2.28058264e-01
1.05078197e+00 -5.70058763e-01 -1.53348255e+00 -2.88065374e-01
-7.72043914e-02 -4.51774895e-01 4.76008147e-01 -4.39342707e-01
-1.35925806e+00 5.50964296e-01 6.76106572e-01 2.88005888e-01
8.09299529e-01 -2.26462737e-01 1.36867917e+00 3.99487555e-01
1.74900070e-01 -1.91955590e+00 -3.96459341e-01 6.09736621e-01
2.75413334e-01 -1.35981429e+00 -1.30413502e-01 2.99192935e-01
-5.26541710e-01 6.72329187e-01 6.59103692e-01 -1.53430328e-01
1.26675820e+00 2.16160506e-01 -1.99989647e-01 -1.09911186e-03
-1.06887376e+00 -2.31437400e-01 -2.60196358e-01 2.86029339e-01
-1.39716387e-01 5.90054691e-01 -3.42804968e-01 4.36084241e-01
-3.27341914e-01 1.92647070e-01 2.46032864e-01 8.75328481e-01
-8.75134766e-01 -8.62377048e-01 -2.04357117e-01 8.26482832e-01
6.33985773e-02 -3.07081610e-01 2.16293484e-01 7.19527066e-01
-2.19352439e-01 7.13677764e-01 5.87856472e-01 -3.65490377e-01
2.79411107e-01 6.20424524e-02 -4.69308555e-01 -7.96010569e-02
2.91995727e-03 -1.96079314e-01 -8.09944049e-02 -6.82358384e-01
6.22488596e-02 -7.22549915e-01 -1.52034628e+00 -9.78130579e-01
-1.26343757e-01 1.54341847e-01 9.17525649e-01 8.52800965e-01
7.72322774e-01 1.81990266e-01 1.00118411e+00 -1.01323307e+00
-6.32467866e-01 -3.55501860e-01 -6.11091495e-01 -2.82226115e-01
1.90708414e-01 -8.44401658e-01 -3.15837234e-01 -6.73242390e-01] | [6.00666618347168, 1.6874703168869019] |
902d6c43-1bc7-4e1c-b251-28769e5f4b54 | grammatical-error-detection-using-error-and | null | null | https://aclanthology.org/I17-1005 | https://aclanthology.org/I17-1005.pdf | Grammatical Error Detection Using Error- and Grammaticality-Specific Word Embeddings | In this study, we improve grammatical error detection by learning word embeddings that consider grammaticality and error patterns. Most existing algorithms for learning word embeddings usually model only the syntactic context of words so that classifiers treat erroneous and correct words as similar inputs. We address the problem of contextual information by considering learner errors. Specifically, we propose two models: one model that employs grammatical error patterns and another model that considers grammaticality of the target word. We determine grammaticality of n-gram sequence from the annotated error tags and extract grammatical error patterns for word embeddings from large-scale learner corpora. Experimental results show that a bidirectional long-short term memory model initialized by our word embeddings achieved the state-of-the-art accuracy by a large margin in an English grammatical error detection task on the First Certificate in English dataset. | ['Yuya Sakaizawa', 'Masahiro Kaneko', 'Mamoru Komachi'] | 2017-11-01 | grammatical-error-detection-using-error-and-1 | https://aclanthology.org/I17-1005 | https://aclanthology.org/I17-1005.pdf | ijcnlp-2017-11 | ['learning-word-embeddings', 'grammatical-error-detection'] | ['methodology', 'natural-language-processing'] | [-1.76478669e-01 5.45576913e-03 -1.39394194e-01 -6.28643692e-01
-7.15061426e-01 -2.39159748e-01 -4.18933816e-02 1.06533158e+00
-1.08800125e+00 4.84423786e-01 3.26169401e-01 -7.42500544e-01
7.58062601e-02 -1.00562811e+00 -8.99864912e-01 1.78558175e-02
9.83201340e-02 2.04149872e-01 -1.33342281e-01 -3.42191666e-01
5.18929482e-01 -5.23469935e-04 -1.30954921e+00 2.10760966e-01
1.40502882e+00 5.79704881e-01 3.71576160e-01 8.55470836e-01
-7.88777053e-01 6.76634610e-01 -7.04435945e-01 -8.21538627e-01
-3.39919329e-01 -3.32362682e-01 -1.07053363e+00 -4.67190981e-01
6.02710307e-01 4.21906188e-02 -2.42741823e-01 1.38648272e+00
4.62542385e-01 1.29238665e-01 3.01086694e-01 -7.16847241e-01
-1.21284592e+00 6.36809528e-01 2.27321580e-01 4.57680166e-01
5.30812562e-01 -2.36248702e-01 1.28271067e+00 -1.21006012e+00
5.09601593e-01 1.18043447e+00 9.77899253e-01 8.27271461e-01
-1.13262975e+00 -7.35644579e-01 3.27856421e-01 3.85902166e-01
-1.26698947e+00 1.22251004e-01 3.62029880e-01 -4.16465342e-01
1.61411250e+00 -5.58847608e-03 5.83272099e-01 1.11815000e+00
8.85912359e-01 5.95990717e-01 7.85428822e-01 -1.11414063e+00
-7.73938075e-02 5.89349903e-02 9.77958918e-01 1.18751085e+00
4.41385090e-01 2.53891528e-01 -6.12944126e-01 -1.78249478e-01
-1.09224662e-01 1.75374169e-02 -2.63376474e-01 2.61195511e-01
-7.81935036e-01 1.05157769e+00 1.44131929e-01 4.46706474e-01
1.48633718e-01 5.89200556e-01 4.95972574e-01 7.83873558e-01
6.51694357e-01 3.91213924e-01 -6.56478345e-01 -4.06000704e-01
-4.27905113e-01 5.62853664e-02 4.37635839e-01 9.66454923e-01
7.43593395e-01 -1.80138811e-01 1.98299617e-01 1.06355572e+00
7.99296200e-01 3.46142322e-01 1.05953455e+00 -1.14553876e-01
4.85715806e-01 6.55153155e-01 -3.81972611e-01 -8.20657551e-01
-3.45136076e-01 -3.41464937e-01 -1.59723908e-01 -4.44717593e-02
3.29146177e-01 7.87913650e-02 -1.05479670e+00 1.76588225e+00
1.56052420e-02 2.80502170e-01 9.76383835e-02 4.16419744e-01
1.04367316e+00 4.67404038e-01 6.21001422e-01 1.06427737e-01
1.40539908e+00 -8.20674241e-01 -1.05892062e+00 -5.68424463e-01
1.67138767e+00 -9.14374053e-01 1.25854957e+00 1.81828171e-01
-7.65091896e-01 -6.44592822e-01 -1.03891647e+00 -3.40547472e-01
-7.22428024e-01 -8.77647698e-02 5.08148447e-02 7.58654177e-01
-6.73378766e-01 7.51296520e-01 -6.91039503e-01 -4.57366467e-01
-8.18194747e-02 6.83947280e-02 -4.82791841e-01 -3.48925531e-01
-1.31606507e+00 1.13611209e+00 3.53797346e-01 -2.34390065e-01
-4.32378560e-01 -8.44827473e-01 -1.40256166e+00 -8.42068195e-02
-4.31864887e-01 -3.68060350e-01 1.13042283e+00 -8.28110576e-01
-8.46725941e-01 1.08565593e+00 -4.22433019e-01 -2.49840200e-01
-1.78234875e-01 -4.23498005e-01 -6.86081886e-01 -6.19461179e-01
-6.25994056e-02 2.77884841e-01 4.53298181e-01 -8.51404488e-01
-7.89084613e-01 -5.56990385e-01 -3.25122982e-01 -1.55174673e-01
-7.09228158e-01 1.52170107e-01 1.41594067e-01 -9.32277679e-01
-1.93140376e-02 -7.05782533e-01 -6.48856303e-03 -1.67969704e-01
1.98509142e-01 -9.74248946e-01 3.87055904e-01 -9.82033670e-01
1.98971319e+00 -2.20801687e+00 1.31324247e-01 1.22180961e-01
-5.04224561e-02 3.24415565e-01 -5.85387051e-01 3.20982069e-01
-4.09625977e-01 5.60933769e-01 -9.26169604e-02 -5.48738182e-01
2.77683660e-02 6.48908257e-01 -2.98859358e-01 2.58179188e-01
3.10723990e-01 7.88822114e-01 -1.25752568e+00 -4.58492815e-01
-1.07811674e-01 1.56219885e-01 -7.40699828e-01 4.17140812e-01
1.99588329e-01 -2.83042252e-01 -9.07493606e-02 4.78768975e-01
4.14301366e-01 3.45472693e-01 2.83714175e-01 2.98952550e-01
1.41623870e-01 9.62572694e-01 -1.04700005e+00 2.02033377e+00
-1.03332818e+00 4.16810960e-01 -6.03749096e-01 -7.41471708e-01
1.09865677e+00 2.71828532e-01 -1.37251660e-01 -7.73816884e-01
8.70799832e-03 6.72918975e-01 1.00977518e-01 -7.23246574e-01
8.32810640e-01 -4.45548445e-01 -3.89823526e-01 3.62460941e-01
7.65419543e-01 1.54182568e-01 3.88798676e-02 5.45989126e-02
1.32515371e+00 1.40208140e-01 2.30313569e-01 -3.92103851e-01
4.39778239e-01 -2.17564076e-01 6.80850744e-01 5.10963380e-01
-2.54683524e-01 3.19292724e-01 1.00635320e-01 -6.57165825e-01
-8.47402096e-01 -9.69228268e-01 -5.51380098e-01 1.34748399e+00
4.44334634e-02 -1.07833409e+00 -4.96107548e-01 -1.49886274e+00
3.84330392e-01 1.18629813e+00 -6.04236543e-01 -6.02260828e-01
-8.01930070e-01 -4.92498666e-01 6.80556178e-01 6.75629139e-01
-2.93438673e-01 -1.01082814e+00 -1.20588832e-01 4.44298267e-01
-6.55771568e-02 -7.72965252e-01 -6.29494607e-01 5.22877097e-01
-1.02449226e+00 -1.18625081e+00 1.89803943e-01 -1.23693681e+00
8.67846131e-01 -3.68683249e-01 1.45500231e+00 7.48343170e-01
-3.76436412e-01 3.50293458e-01 -6.75113022e-01 -4.76152062e-01
-6.03288054e-01 -8.49452093e-02 1.20475560e-01 -6.46396399e-01
1.15026140e+00 -8.13080072e-02 5.49841411e-02 -1.80616409e-01
-8.37014139e-01 -5.93862355e-01 -1.21880062e-02 1.22074437e+00
3.63026261e-01 -3.75865251e-01 5.83725512e-01 -9.02724206e-01
8.06780875e-01 -4.71511930e-01 -1.27387375e-01 4.04155761e-01
-1.14610207e+00 5.00096858e-01 5.47079146e-01 -1.84649363e-01
-6.93238318e-01 -3.57529670e-01 -9.22198176e-01 1.22074358e-01
-1.85479850e-01 5.57548106e-01 2.00756326e-01 -5.76212071e-02
5.75356662e-01 1.71539962e-01 -4.22555238e-01 -5.99345922e-01
1.71041921e-01 9.07472193e-01 8.87062699e-02 -9.11216140e-01
2.03067154e-01 -3.72903377e-01 -4.40872610e-01 -5.18753588e-01
-9.07565057e-01 -4.27383274e-01 -6.89497411e-01 -1.26154423e-01
8.70778084e-01 -7.99691319e-01 -9.90820825e-02 1.98649421e-01
-1.57172716e+00 1.79640919e-01 -4.15944189e-01 7.15399563e-01
-1.01011239e-01 4.91599798e-01 -8.36719453e-01 -5.69336057e-01
-1.70339063e-01 -1.10358763e+00 9.34894562e-01 -1.45007461e-01
-5.54572642e-01 -1.43642628e+00 5.16548395e-01 4.86346930e-02
1.62130713e-01 -4.83381361e-01 1.42413759e+00 -8.92075241e-01
-2.37541832e-03 -1.39501497e-01 1.66957781e-01 6.02766037e-01
-8.92861262e-02 -2.98746258e-01 -7.57770717e-01 -2.44995520e-01
-3.59937847e-01 -3.39479029e-01 1.08024955e+00 -1.37351260e-01
1.21281278e+00 -3.83039527e-02 -1.63179204e-01 6.37561560e-01
1.73544574e+00 3.15566082e-03 6.37499392e-01 4.54672694e-01
6.68372929e-01 3.41620296e-01 6.95762217e-01 1.44231319e-01
3.45423609e-01 3.73890787e-01 3.92722368e-01 5.67038715e-01
-2.53819227e-01 -6.97271407e-01 5.47210515e-01 1.47528136e+00
4.56784964e-01 -4.05928016e-01 -1.18287981e+00 1.15144551e+00
-1.48127925e+00 -4.85725731e-01 -4.06394362e-01 2.18128753e+00
1.12726259e+00 -9.49362945e-03 -6.44827724e-01 7.42959306e-02
5.11501551e-01 -1.56420469e-02 3.31819773e-01 -1.26785398e+00
3.57656419e-01 1.00053847e+00 3.48347694e-01 8.92289340e-01
-7.66405761e-01 1.06707847e+00 6.18456411e+00 7.66862333e-01
-5.95883608e-01 5.01140118e-01 1.56831935e-01 1.33626878e-01
-7.69331932e-01 -8.24947432e-02 -1.28942025e+00 7.63749778e-01
1.47190022e+00 5.09587750e-02 -1.53657079e-01 6.44735396e-01
-2.29743913e-01 -2.38709152e-02 -1.20411837e+00 8.13182831e-01
3.36631745e-01 -9.23029125e-01 1.03896357e-01 -9.13404897e-02
4.57688779e-01 -1.02204330e-01 -2.42767707e-01 7.50865281e-01
5.96127570e-01 -1.36146188e+00 5.78432560e-01 6.34814799e-01
6.01270378e-01 -9.99384761e-01 1.10034323e+00 2.62306690e-01
-9.49169874e-01 -1.44552290e-01 -7.69033790e-01 -2.68939584e-01
1.05955033e-02 7.35625207e-01 -3.44588667e-01 2.78945655e-01
6.88519359e-01 8.53897214e-01 -7.78458655e-01 8.32831800e-01
-6.46141827e-01 8.64228487e-01 2.99748145e-02 -1.86888114e-01
1.59711346e-01 -7.56329596e-02 3.39013368e-01 1.64035392e+00
7.03830063e-01 -1.13861807e-01 8.91820192e-02 5.69172502e-01
-5.60258746e-01 4.40506369e-01 -7.73636699e-01 5.12402542e-02
4.93871391e-01 8.23646963e-01 1.55486777e-01 -8.00640881e-02
-6.54411674e-01 1.05949223e+00 9.70392525e-01 -1.87646806e-01
-7.01869667e-01 -7.17610955e-01 1.30725777e+00 -8.52750540e-02
1.93437636e-01 -3.47151548e-01 -2.74365336e-01 -1.06788957e+00
1.80499911e-01 -7.16062844e-01 5.87810993e-01 -3.93946022e-01
-1.41647494e+00 4.70712125e-01 -6.72874808e-01 -7.79883146e-01
1.23350434e-01 -1.06795061e+00 -5.92852473e-01 9.53940809e-01
-1.87241220e+00 -8.04574490e-01 -6.90765306e-03 6.91624656e-02
8.57068181e-01 -1.20744377e-01 1.33709586e+00 5.99397600e-01
-5.83075047e-01 1.02780592e+00 1.36308521e-01 4.76989746e-01
1.06859517e+00 -1.51061654e+00 3.79675269e-01 8.69988084e-01
5.13780177e-01 7.48932719e-01 4.41469520e-01 -7.71237016e-01
-1.16479254e+00 -1.19683027e+00 2.29549527e+00 -6.59707844e-01
8.07241797e-01 -2.19227567e-01 -1.17199481e+00 9.64688540e-01
2.93106496e-01 4.84755069e-01 1.12917399e+00 5.36151826e-01
-5.05508304e-01 2.84851879e-01 -9.41483736e-01 6.87720627e-02
1.14305258e+00 -7.68466353e-01 -1.25023651e+00 2.45584786e-01
7.18680143e-01 -4.37244445e-01 -1.03567481e+00 2.41802752e-01
2.22385690e-01 -3.91820699e-01 5.13160765e-01 -1.40439391e+00
7.40600109e-01 -3.31340246e-02 -3.00127894e-01 -1.87525988e+00
-2.96101838e-01 1.70153499e-01 -7.17902705e-02 1.19270182e+00
5.27928948e-01 -4.45530623e-01 2.74051517e-01 3.89189273e-01
-5.63862741e-01 -9.02962267e-01 -1.06353557e+00 -9.18161571e-01
7.56002486e-01 -9.08883750e-01 4.16286916e-01 1.10467362e+00
4.26009089e-01 5.97331859e-02 1.23664975e-01 1.99442387e-01
2.40052059e-01 -2.59382755e-01 1.07891314e-01 -1.12830234e+00
1.48436561e-01 -8.15966576e-02 -8.71912181e-01 -1.07304561e+00
9.33397233e-01 -1.50284600e+00 1.93712756e-01 -1.25419295e+00
1.29035056e-01 -6.48435891e-01 -5.39424181e-01 3.44014257e-01
-7.42316663e-01 1.92650054e-02 -1.26440406e-01 -5.65684617e-01
-4.77684736e-01 6.69679999e-01 4.56887096e-01 -1.82991743e-01
3.93951684e-01 -6.63975000e-01 -3.50388139e-01 7.62841225e-01
5.98716676e-01 -1.05068386e+00 1.50978535e-01 -1.12039423e+00
7.02133000e-01 -5.94492555e-01 1.73739210e-01 -7.85708964e-01
9.86223295e-02 7.46258199e-02 1.26576722e-01 1.96740520e-03
-4.62856352e-01 -9.12806571e-01 -5.69734693e-01 5.59031963e-01
-6.47634506e-01 7.41995394e-01 3.42331082e-01 8.03671420e-01
-4.42344576e-01 -1.10089040e+00 6.79903805e-01 7.79013336e-02
-8.99415910e-01 -4.66662496e-02 -5.91187835e-01 3.90607476e-01
8.41785014e-01 1.60350978e-01 -3.13853621e-01 2.47881204e-01
-8.38159263e-01 2.33703196e-01 2.75582433e-01 8.73808146e-01
9.44367588e-01 -1.77314055e+00 -7.08957434e-01 5.85484207e-01
6.21048272e-01 -4.83352274e-01 -1.60443977e-01 4.54491556e-01
-4.60552245e-01 2.25251690e-01 2.08780393e-01 -3.72589022e-01
-1.81585491e+00 4.30398405e-01 5.34155071e-01 -4.72223282e-01
-3.33482325e-01 1.29349732e+00 -4.41269398e-01 -1.12640440e+00
3.43165398e-01 -7.31820762e-01 -8.54159743e-02 -1.36701643e-01
5.00696480e-01 1.78281233e-01 4.96780783e-01 -5.79776347e-01
-3.44744295e-01 6.36014163e-01 -5.04006818e-02 5.05530059e-01
1.23813522e+00 -2.68236399e-02 -1.57499075e-01 6.04385674e-01
1.36353040e+00 1.83459654e-01 -1.23733915e-01 -3.62747461e-01
5.85139811e-01 -8.01600635e-01 3.81584257e-01 -7.15786040e-01
-7.17956901e-01 1.14283085e+00 8.88801455e-01 -2.48942152e-01
6.68774962e-01 -6.58073202e-02 1.18173444e+00 2.82654226e-01
3.04024220e-01 -1.30853212e+00 -2.99892109e-02 1.19425535e+00
3.74373645e-01 -1.29474115e+00 -5.40729761e-01 -2.68903196e-01
3.06612300e-03 1.55775166e+00 8.88088644e-01 -2.15739667e-01
8.14680338e-01 2.36218393e-01 3.27829644e-02 -2.13661805e-01
-7.91136742e-01 -1.66918308e-01 3.68232548e-01 2.58288264e-01
1.19085026e+00 3.07243653e-02 -1.01962006e+00 7.17228591e-01
-6.70631826e-02 -1.72311395e-01 4.28039461e-01 1.11796927e+00
-7.35950053e-01 -1.81548619e+00 -1.84341639e-01 4.31924284e-01
-6.55523360e-01 -5.81453562e-01 -1.11283801e-01 4.34585422e-01
5.10744810e-01 7.72823989e-01 4.80802089e-01 -4.97201383e-01
6.20481074e-01 8.91551971e-01 7.41459548e-01 -1.05128920e+00
-1.06856227e+00 -8.13365638e-01 2.38098606e-01 -6.34261429e-01
1.51524305e-01 -4.70530421e-01 -1.58611405e+00 -3.17256868e-01
-6.52020931e-01 3.22327882e-01 7.96624660e-01 1.17441833e+00
3.41789693e-01 7.59417772e-01 2.45984271e-01 1.97273344e-01
-7.24113166e-01 -1.15096653e+00 -4.98996943e-01 8.07062089e-01
2.59626269e-01 -4.29992199e-01 -3.38192701e-01 -1.05778351e-01] | [11.00619125366211, 10.707874298095703] |
89754d1b-9c0c-42d9-bd71-ab1229c1029c | zero-shot-object-detection | 1804.04340 | null | http://arxiv.org/abs/1804.04340v2 | http://arxiv.org/pdf/1804.04340v2.pdf | Zero-Shot Object Detection | We introduce and tackle the problem of zero-shot object detection (ZSD),
which aims to detect object classes which are not observed during training. We
work with a challenging set of object classes, not restricting ourselves to
similar and/or fine-grained categories as in prior works on zero-shot
classification. We present a principled approach by first adapting
visual-semantic embeddings for ZSD. We then discuss the problems associated
with selecting a background class and motivate two background-aware approaches
for learning robust detectors. One of these models uses a fixed background
class and the other is based on iterative latent assignments. We also outline
the challenge associated with using a limited number of training classes and
propose a solution based on dense sampling of the semantic label space using
auxiliary data with a large number of categories. We propose novel splits of
two standard detection datasets - MSCOCO and VisualGenome, and present
extensive empirical results in both the traditional and generalized zero-shot
settings to highlight the benefits of the proposed methods. We provide useful
insights into the algorithm and conclude by posing some open questions to
encourage further research. | ['Gaurav Sharma', 'Ankan Bansal', 'Ajay Divakaran', 'Karan Sikka', 'Rama Chellappa'] | 2018-04-12 | zero-shot-object-detection-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Ankan_Bansal_Zero-Shot_Object_Detection_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Ankan_Bansal_Zero-Shot_Object_Detection_ECCV_2018_paper.pdf | eccv-2018-9 | ['zero-shot-object-detection'] | ['computer-vision'] | [ 5.56941748e-01 2.33017839e-02 -2.22627833e-01 -2.39817828e-01
-6.81739271e-01 -3.64260852e-01 7.94497252e-01 4.99394313e-02
-4.25583184e-01 2.98972666e-01 -4.17071814e-03 -7.10265292e-03
-1.38263941e-01 -6.37552917e-01 -3.89404923e-01 -7.77817786e-01
1.32953346e-01 4.13275957e-01 9.34056342e-01 1.87665567e-01
2.20363468e-01 2.64026105e-01 -2.09970808e+00 8.20314735e-02
3.59478921e-01 9.55878556e-01 3.60490799e-01 6.98673010e-01
-2.11868480e-01 8.41787219e-01 -6.12266421e-01 -3.96428436e-01
5.64437151e-01 -4.40253109e-01 -5.72108984e-01 8.28439951e-01
6.64514720e-01 -3.34244728e-01 -1.67254701e-01 1.25111878e+00
5.86341918e-01 5.29923201e-01 1.01079404e+00 -1.59154332e+00
-4.88987982e-01 9.28306952e-02 -6.66815579e-01 6.32515490e-01
-1.79991573e-01 2.18361430e-02 1.06938553e+00 -1.19249761e+00
7.21845806e-01 1.32528400e+00 7.56293237e-01 8.00323546e-01
-1.34819019e+00 -2.84029335e-01 4.18210208e-01 3.49639386e-01
-1.55635190e+00 -5.83701849e-01 5.70680320e-01 -9.75329757e-01
6.41299009e-01 5.12745976e-02 4.43360448e-01 1.22193956e+00
-4.60097909e-01 8.06726575e-01 9.57865715e-01 -8.39357972e-01
5.93909383e-01 4.87337142e-01 7.97213316e-01 7.26267338e-01
5.47946334e-01 -2.94867475e-02 -3.91827762e-01 -4.17799115e-01
4.52458441e-01 1.66256920e-01 -3.48922201e-02 -1.17871952e+00
-9.66955662e-01 1.13369167e+00 9.36179981e-02 9.66880322e-02
-8.90507102e-02 -3.96931283e-02 3.88811737e-01 -1.03557065e-01
5.45004368e-01 2.92119104e-02 -1.66475147e-01 2.92388678e-01
-8.90050709e-01 1.57310590e-01 6.77181482e-01 1.16383052e+00
7.53068030e-01 -3.56383398e-02 -5.63904643e-01 1.09267318e+00
2.14235187e-01 2.55332977e-01 2.72932410e-01 -9.06384885e-01
3.56696993e-02 2.57402480e-01 4.07617271e-01 -8.06207359e-01
-8.84814709e-02 -4.25908625e-01 -3.47701132e-01 3.98386210e-01
3.93901289e-01 1.06039107e-01 -1.29387712e+00 1.67383790e+00
5.99391937e-01 6.41399384e-01 -7.51339942e-02 8.03214252e-01
7.67386794e-01 3.29979211e-01 7.98115432e-02 -2.32361630e-01
1.44166458e+00 -1.21297514e+00 -6.26178980e-01 -3.79600048e-01
3.87429744e-01 -4.69735146e-01 8.45792770e-01 5.40802069e-02
-6.01991475e-01 -5.98657131e-01 -1.14197898e+00 7.89497048e-03
-6.05609596e-01 1.40064344e-01 5.25026321e-01 8.79052162e-01
-8.57536554e-01 4.13906932e-01 -8.51671815e-01 -9.11122739e-01
5.78210711e-01 -7.68365413e-02 2.39180829e-02 -9.29265544e-02
-7.84467280e-01 8.62523139e-01 4.51343387e-01 -4.71094191e-01
-1.30055463e+00 -4.71004158e-01 -8.50489199e-01 -1.91944058e-03
6.80060208e-01 -4.29877967e-01 1.38627720e+00 -5.90511978e-01
-1.08903253e+00 1.28358233e+00 -3.02289605e-01 -3.63391042e-01
4.62254256e-01 -1.19796216e-01 -1.61132336e-01 2.99286306e-01
4.92805660e-01 5.19413352e-01 9.82937098e-01 -1.27908206e+00
-9.88379419e-01 -1.95587724e-01 1.47427082e-01 1.04272313e-01
-4.99956042e-01 2.28771389e-01 -7.22864509e-01 -5.35379171e-01
7.68209621e-02 -8.53944123e-01 -3.64498317e-01 4.07677531e-01
-3.30561578e-01 -3.31214339e-01 8.21794868e-01 -1.28123283e-01
9.79115367e-01 -2.23428226e+00 3.33555974e-02 -1.99459761e-01
3.97691846e-01 2.55518198e-01 -7.56016448e-02 2.07901657e-01
7.61614516e-02 -2.31697947e-01 -2.84792960e-01 -6.08307004e-01
2.36728102e-01 2.85427213e-01 -4.24802005e-01 6.23377025e-01
2.97555923e-01 6.22178316e-01 -1.14380491e+00 -6.79001629e-01
4.63037729e-01 2.00579494e-01 -3.64379227e-01 2.02114359e-01
1.06798150e-02 -1.39366925e-01 -3.56606662e-01 8.37723076e-01
6.51636660e-01 -3.78038436e-01 -7.02860355e-02 3.37609686e-02
-1.11020125e-01 -6.85776919e-02 -1.53803432e+00 1.34698462e+00
1.06036745e-01 5.67943811e-01 1.01657473e-02 -1.25363719e+00
6.52872205e-01 5.81488898e-03 3.13850194e-01 -1.21941030e-01
2.31202915e-01 -1.61274463e-01 -3.47527146e-01 -3.58751774e-01
4.21933442e-01 -4.32313263e-01 -2.50349902e-02 3.99282426e-01
5.86445451e-01 1.20607451e-01 2.61066854e-01 3.95874530e-01
1.07192981e+00 -2.40545869e-02 6.53528750e-01 -3.45922142e-01
1.26245916e-01 -4.08855267e-02 7.69584000e-01 1.49838221e+00
-7.92298198e-01 6.05346859e-01 3.23650092e-01 -2.48720542e-01
-8.26575041e-01 -1.20005083e+00 -3.54864508e-01 1.49629724e+00
4.78848904e-01 -3.46768975e-01 -6.16183758e-01 -7.94346750e-01
5.09925187e-02 7.02311635e-01 -9.46433663e-01 -1.54378206e-01
4.62618954e-02 -1.10388052e+00 1.14566423e-01 6.91980004e-01
9.35430378e-02 -6.53266191e-01 -8.48287880e-01 -2.84118354e-02
1.68804657e-02 -1.06830299e+00 -8.16024840e-02 6.59277856e-01
-6.02697313e-01 -1.21424091e+00 -6.45271122e-01 -8.28613997e-01
5.61131775e-01 1.06646323e+00 9.87899244e-01 -9.69435200e-02
-7.29434073e-01 7.73704708e-01 -5.09618402e-01 -5.80251753e-01
-2.14483991e-01 -2.47812584e-01 2.51321971e-01 2.52411425e-01
8.84282887e-01 -7.62171224e-02 -3.24090391e-01 4.29423541e-01
-6.84124708e-01 -2.73465700e-02 3.64719689e-01 7.66922772e-01
4.91843522e-01 -1.00551754e-01 3.53129178e-01 -9.44527030e-01
1.95278466e-01 -7.09493160e-01 -5.67619383e-01 2.59630322e-01
-3.12992692e-01 -2.04331622e-01 -1.21421684e-02 -6.08777463e-01
-1.02703595e+00 1.43521205e-01 3.26653004e-01 -8.68579686e-01
-3.06775868e-01 -3.61852944e-01 -9.53813046e-02 -1.36309639e-01
8.05722117e-01 -1.07616916e-01 -2.52839476e-01 -5.45150101e-01
6.03378236e-01 6.67843223e-01 4.30452675e-01 -4.11905617e-01
9.88050759e-01 8.51336181e-01 -1.66606262e-01 -1.08797669e+00
-1.34674883e+00 -1.27025211e+00 -9.62844253e-01 -2.07482040e-01
9.88092661e-01 -9.93711770e-01 3.98288928e-02 2.71408170e-01
-9.32917655e-01 -3.49104702e-01 -7.03978539e-01 3.71631563e-01
-6.77365541e-01 5.68945348e-01 -4.61230934e-01 -1.14785349e+00
1.00926220e-01 -7.75895119e-01 1.38305366e+00 1.16279379e-01
-1.53668197e-02 -8.41417253e-01 1.76348895e-01 9.14743450e-03
6.62969500e-02 1.42512038e-01 5.33466518e-01 -8.06478143e-01
-5.88459015e-01 -2.62815982e-01 -3.92262578e-01 2.56943852e-01
1.94918513e-01 -2.40798220e-01 -1.30919158e+00 -4.25608426e-01
9.58341509e-02 -5.03706872e-01 1.35661614e+00 5.06274700e-01
8.74460459e-01 2.81630456e-01 -7.41745889e-01 4.81319934e-01
1.53373384e+00 1.33605143e-02 1.29143059e-01 3.28723133e-01
5.09855688e-01 6.28521323e-01 9.51867938e-01 5.53596020e-01
7.49534667e-02 7.70292580e-01 4.31006372e-01 -8.41376558e-03
-2.63176590e-01 -1.85038149e-01 -1.62769239e-02 3.68175805e-01
1.32345244e-01 -2.63267785e-01 -6.66113973e-01 8.93262267e-01
-2.07912803e+00 -1.10413623e+00 6.30025491e-02 2.25860643e+00
3.51365179e-01 2.05637723e-01 3.94479752e-01 1.18183307e-01
1.14318311e+00 1.65465266e-01 -4.34515983e-01 2.35159650e-01
-6.77231252e-02 8.65091383e-02 4.83786881e-01 4.66055453e-01
-1.70570052e+00 9.98892903e-01 6.84054422e+00 7.79948175e-01
-5.87822318e-01 5.45077443e-01 1.72966138e-01 -1.44231766e-01
3.38400453e-01 1.75686687e-01 -1.39164746e+00 3.22627485e-01
6.17639124e-01 -6.42141998e-02 -5.62833808e-02 1.27858162e+00
-1.75538048e-01 -1.23827733e-01 -1.06223786e+00 9.21804547e-01
4.96065110e-01 -1.20105004e+00 -2.02856734e-01 -1.21384658e-01
7.35022902e-01 1.15840897e-01 -2.60887653e-01 5.45112312e-01
5.46288610e-01 -4.17288065e-01 9.48451161e-01 3.29195023e-01
6.09118402e-01 -1.84644401e-01 3.06552261e-01 4.16653454e-01
-1.26940358e+00 -4.58629489e-01 -9.07070160e-01 -1.32616714e-01
-3.63251865e-02 2.98745096e-01 -6.31214499e-01 1.82181522e-01
8.58187139e-01 8.09764445e-01 -8.22019696e-01 1.40467083e+00
-4.22278456e-02 5.78502417e-01 -2.80073106e-01 -2.93842852e-02
8.70415419e-02 -9.32311546e-03 6.18447840e-01 1.40855634e+00
1.85288608e-01 7.31318295e-02 6.00470781e-01 7.91659951e-01
2.28371367e-01 -1.64860219e-01 -6.07962966e-01 2.03055024e-01
5.12534678e-01 1.42032194e+00 -1.26850331e+00 -7.09615052e-01
-8.08497846e-01 1.01463771e+00 3.65624815e-01 3.51285845e-01
-8.30341578e-01 -4.43877339e-01 5.96849203e-01 1.15945451e-01
7.17286110e-01 -2.55988091e-02 2.04818413e-01 -1.49767184e+00
-2.27797061e-01 -2.72706360e-01 7.63272703e-01 -5.35062253e-01
-1.44881070e+00 2.30755940e-01 3.83143246e-01 -1.30259299e+00
1.83782861e-01 -8.52914512e-01 -6.26791239e-01 4.57681775e-01
-1.57666898e+00 -1.11494327e+00 -5.78119814e-01 4.12298769e-01
9.26131308e-01 -1.32531047e-01 7.17420340e-01 2.21618652e-01
-5.19632101e-01 2.60011584e-01 3.11787814e-01 9.23437029e-02
6.54010773e-01 -1.30362189e+00 2.25567222e-01 1.15099108e+00
2.30349869e-01 3.83020192e-01 8.64012837e-01 -4.34110612e-01
-8.71670663e-01 -1.48171437e+00 5.92859030e-01 -7.20975459e-01
7.02529550e-01 -8.12962174e-01 -8.08549285e-01 7.75364518e-01
-1.92730755e-01 4.67649817e-01 7.49713182e-01 1.29546106e-01
-2.86723435e-01 2.24771380e-01 -1.08007395e+00 3.35093826e-01
1.41673970e+00 -4.12434638e-01 -9.91788983e-01 5.20463288e-01
6.31241143e-01 6.97220713e-02 -4.05335724e-02 2.43162632e-01
1.79868907e-01 -7.86263466e-01 1.11252868e+00 -7.74755716e-01
-2.14935854e-01 -5.11340737e-01 -5.40562510e-01 -9.75275218e-01
-7.09620595e-01 -2.14909047e-01 -2.86463737e-01 1.21958041e+00
-2.82345712e-01 -3.80748153e-01 8.31495047e-01 2.14167133e-01
-1.02940530e-01 -2.11112574e-01 -9.01960135e-01 -1.11695230e+00
-2.01273903e-01 -3.65250260e-01 -1.32117122e-01 9.51474011e-01
-3.57840210e-01 4.68765467e-01 -4.96661842e-01 3.80842328e-01
1.13341296e+00 2.37564176e-01 9.03125405e-01 -1.59670091e+00
-4.20397282e-01 -1.68441564e-01 -8.15970838e-01 -9.36749518e-01
4.57794480e-02 -8.33835840e-01 3.10516357e-01 -1.30316389e+00
7.93124139e-01 -1.56405896e-01 -5.73237836e-01 5.27793288e-01
-3.05674762e-01 5.37830293e-01 1.99072599e-01 2.34858304e-01
-1.11950004e+00 5.38176298e-01 3.91408503e-01 -2.21846640e-01
6.43086284e-02 1.46873835e-02 -6.92012191e-01 7.99182713e-01
3.77782255e-01 -5.61172307e-01 -4.70045656e-01 -5.43437712e-02
-4.08479124e-01 -4.28373784e-01 7.12853909e-01 -1.02062654e+00
1.36885792e-01 -2.15839669e-01 3.58032644e-01 -6.01597846e-01
5.37047982e-01 -5.90893924e-01 -4.30128485e-01 3.82820278e-01
-2.59348571e-01 -6.89002395e-01 -4.44895662e-02 1.13135183e+00
1.35512382e-01 -6.37031496e-01 1.33240986e+00 -2.64524579e-01
-1.29222906e+00 2.07390442e-01 -4.83332098e-01 1.09570377e-01
1.50767422e+00 -4.97435004e-01 -1.25359431e-01 1.27736479e-02
-1.04840529e+00 1.91840574e-01 5.97975850e-01 6.41259015e-01
3.67335528e-01 -1.30666661e+00 -3.90321136e-01 1.36189550e-01
6.48534358e-01 -4.08699661e-01 2.93050315e-02 5.83015561e-01
-2.81643625e-02 3.29089195e-01 -6.88313544e-02 -8.78004193e-01
-1.45394218e+00 1.07709777e+00 3.12215537e-01 1.86903730e-01
-7.60563076e-01 9.84016716e-01 7.42390931e-01 -1.93829387e-01
5.99002063e-01 7.49487728e-02 -2.66079962e-01 2.90552855e-01
6.92308128e-01 5.36496341e-01 -1.44467726e-01 -6.56858087e-01
-5.31350613e-01 4.93630558e-01 -4.63791229e-02 9.14813429e-02
1.28704917e+00 -2.40005121e-01 2.84507692e-01 7.90325880e-01
8.31936240e-01 -2.55369782e-01 -1.45481467e+00 -6.41925156e-01
2.73907244e-01 -7.61090338e-01 1.14360549e-01 -2.67316967e-01
-5.62316835e-01 1.06411052e+00 9.81991172e-01 2.58798957e-01
8.04248095e-01 4.30487871e-01 -2.69273221e-02 3.45978439e-01
3.96881938e-01 -1.23614049e+00 3.97248209e-01 3.77251357e-01
2.83515364e-01 -1.40099549e+00 -8.80463868e-02 -5.38357794e-01
-5.38512588e-01 9.70577598e-01 6.84748888e-01 -1.31196961e-01
6.58514738e-01 -1.82202570e-02 -6.22477829e-02 -2.93783695e-01
-6.11028194e-01 -9.75535274e-01 9.08891857e-02 8.90890777e-01
-1.23541847e-01 -1.51389316e-01 7.70590678e-02 3.91176581e-01
5.26757121e-01 -1.30384624e-01 5.26447296e-01 1.28525150e+00
-1.04512882e+00 -7.28881717e-01 -4.50265467e-01 5.26203990e-01
-1.09706081e-01 5.09188995e-02 -3.74879241e-01 6.59086823e-01
4.87557679e-01 1.06899083e+00 1.61603123e-01 3.51890712e-03
1.61918223e-01 4.05607194e-01 4.39858586e-01 -1.19497252e+00
4.57225621e-01 1.38368130e-01 -1.02808110e-01 -3.28510404e-01
-6.10460639e-01 -8.30463767e-01 -5.51807582e-01 2.90866137e-01
-7.55759835e-01 -2.54836269e-02 2.28460789e-01 7.49433279e-01
1.15997933e-01 3.63127410e-01 4.88363892e-01 -1.17805314e+00
-7.42841363e-01 -8.84094656e-01 -8.48575950e-01 5.71053386e-01
4.83787984e-01 -1.43409836e+00 -7.29312718e-01 1.48416758e-01] | [9.436176300048828, 1.6066113710403442] |
f8f6d1b3-e748-4614-a34c-6629bc4189f8 | a-massive-scale-semantic-similarity-dataset | 2306.17810 | null | https://arxiv.org/abs/2306.17810v1 | https://arxiv.org/pdf/2306.17810v1.pdf | A Massive Scale Semantic Similarity Dataset of Historical English | A diversity of tasks use language models trained on semantic similarity data. While there are a variety of datasets that capture semantic similarity, they are either constructed from modern web data or are relatively small datasets created in the past decade by human annotators. This study utilizes a novel source, newly digitized articles from off-copyright, local U.S. newspapers, to assemble a massive-scale semantic similarity dataset spanning 70 years from 1920 to 1989 and containing nearly 400M positive semantic similarity pairs. Historically, around half of articles in U.S. local newspapers came from newswires like the Associated Press. While local papers reproduced articles from the newswire, they wrote their own headlines, which form abstractive summaries of the associated articles. We associate articles and their headlines by exploiting document layouts and language understanding. We then use deep neural methods to detect which articles are from the same underlying source, in the presence of substantial noise and abridgement. The headlines of reproduced articles form positive semantic similarity pairs. The resulting publicly available HEADLINES dataset is significantly larger than most existing semantic similarity datasets and covers a much longer span of time. It will facilitate the application of contrastively trained semantic similarity models to a variety of tasks, including the study of semantic change across space and time. | ['Melissa Dell', 'Emily Silcock'] | 2023-06-30 | null | null | null | null | ['semantic-textual-similarity', 'semantic-similarity'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.97829972e-02 -3.28404635e-01 -4.10968155e-01 -3.99730295e-01
-1.02414775e+00 -1.10616648e+00 1.22843993e+00 8.99008274e-01
-6.04207635e-01 4.00412232e-01 9.80944693e-01 -2.00218245e-01
-2.51987934e-01 -8.22783589e-01 -6.88236356e-01 1.50341734e-01
3.76738161e-01 4.69688773e-01 2.25358218e-01 -5.32503605e-01
1.04149199e+00 1.52050480e-01 -1.22772515e+00 4.39787716e-01
8.50704491e-01 9.89771366e-01 5.37542939e-01 3.04669738e-01
-5.75473785e-01 4.24700230e-01 -6.93963468e-01 -5.22132516e-01
2.09987178e-01 -3.26620907e-01 -9.70484912e-01 -1.64440960e-01
1.03018260e+00 2.06446230e-01 -5.38730800e-01 1.26134849e+00
3.13189328e-01 3.07004511e-01 6.05276585e-01 -1.06129944e+00
-1.44371963e+00 8.42108727e-01 -3.54091167e-01 7.29021370e-01
5.91103852e-01 -2.69769222e-01 1.57400918e+00 -7.06298113e-01
1.10887182e+00 1.31949115e+00 9.57356572e-01 -7.41252452e-02
-9.65386569e-01 -5.96606433e-01 -2.33883709e-02 2.80160338e-01
-9.87060189e-01 -3.51280868e-01 8.13073814e-01 -3.92982841e-01
8.88351798e-01 8.52766111e-02 5.00821114e-01 1.60580301e+00
1.31424993e-01 4.96824652e-01 1.01073384e+00 -2.35467702e-01
1.16923720e-01 -2.58009303e-02 4.41055030e-01 3.51690769e-01
1.17955841e-01 -5.02212584e-01 -5.88670075e-01 -3.61599177e-01
4.01178151e-01 2.45203152e-01 -2.92352796e-01 -2.97868345e-02
-1.53608119e+00 9.68115509e-01 6.26370966e-01 6.96163177e-01
-1.98224504e-02 -1.94752961e-01 8.37923944e-01 5.36602437e-01
6.54126227e-01 1.07362401e+00 -2.87990898e-01 -6.35379478e-02
-9.82089341e-01 5.63111067e-01 1.02295136e+00 1.22639155e+00
6.10808551e-01 -5.76729655e-01 1.08732894e-01 1.14971805e+00
1.21733442e-01 6.40141487e-01 9.87438619e-01 -8.83143842e-01
8.41889679e-01 6.31682992e-01 9.96984635e-03 -1.79590583e+00
-3.82017553e-01 -4.47702169e-01 -3.74665797e-01 -6.57138526e-01
2.61219800e-01 2.89356560e-01 -5.74737191e-01 1.48556244e+00
-2.91613042e-01 -1.05733603e-01 4.16240320e-02 8.56888473e-01
1.06583345e+00 5.96495986e-01 1.63143352e-02 3.34996134e-01
1.51058519e+00 -9.66997683e-01 -5.89484155e-01 -4.77114141e-01
6.67204440e-01 -1.21633708e+00 1.36693776e+00 -1.27953859e-02
-1.03915286e+00 -5.61610103e-01 -1.16509271e+00 -4.23773378e-01
-9.60417092e-01 -3.13891172e-01 2.64115661e-01 -1.10957049e-01
-9.93951797e-01 9.30457771e-01 -1.31280512e-01 -1.13617039e+00
4.76266831e-01 -5.07296622e-01 -3.41520578e-01 -5.44782765e-02
-1.46816850e+00 1.08011699e+00 1.61779150e-01 -6.53389513e-01
-2.65393615e-01 -9.67485607e-01 -8.99099886e-01 -9.69337896e-02
2.29267091e-01 -6.75864697e-01 1.38554120e+00 -1.09263194e+00
-7.15321064e-01 1.30299485e+00 -4.85136956e-02 -4.46339339e-01
3.06013495e-01 -3.35055113e-01 -8.32784295e-01 1.58159479e-01
7.92730093e-01 2.21341103e-01 4.69898134e-01 -8.25263560e-01
-6.43804073e-01 -4.27534640e-01 -2.39956472e-02 1.74435705e-01
-6.11721218e-01 3.68742794e-01 -4.01411444e-01 -1.33521545e+00
1.00239471e-01 -7.74072409e-01 1.94249287e-01 2.75905896e-02
-3.03699672e-01 -3.82160127e-01 7.40403950e-01 -9.20590103e-01
1.21258223e+00 -2.14361310e+00 -1.04702286e-01 5.24550378e-02
2.89087415e-01 -1.61408454e-01 -5.36021709e-01 8.00543427e-01
9.33827683e-02 3.53654236e-01 -2.37280399e-01 -1.24885879e-01
1.59526452e-01 -1.25754327e-01 -7.83317864e-01 3.75918090e-01
-2.15620488e-01 1.05076706e+00 -1.28035152e+00 -2.98886091e-01
-1.91533297e-01 -1.37971282e-01 -2.13325933e-01 -1.82294831e-01
-1.45330876e-01 -1.82420209e-01 -4.34960037e-01 4.75700736e-01
2.46762231e-01 -3.97598803e-01 -1.00458637e-02 -4.19692755e-01
-5.98858371e-02 9.31586742e-01 -3.93814385e-01 2.16678905e+00
-6.04975104e-01 1.24375200e+00 -3.92951638e-01 -1.13576269e+00
1.08546770e+00 -8.58146846e-02 1.90014184e-01 -1.13356400e+00
7.01480880e-02 3.27308714e-01 -2.65117288e-01 -3.36663157e-01
1.06543899e+00 -1.36483267e-01 -5.96458673e-01 6.83961332e-01
-5.48937842e-02 -4.50060546e-01 2.36451045e-01 4.64867532e-01
1.29336596e+00 -3.29223514e-01 4.06418890e-01 -5.05364895e-01
1.12491496e-01 3.47294569e-01 2.18764067e-01 8.19401205e-01
-1.95664912e-01 7.62892306e-01 2.17403457e-01 -4.32805836e-01
-1.61951649e+00 -1.22550786e+00 -3.74977171e-01 1.11890662e+00
3.96786004e-01 -6.42410815e-01 -3.22654963e-01 -4.66410577e-01
2.84773439e-01 7.96526074e-01 -5.78933835e-01 -2.33093724e-01
-4.41234678e-01 -2.52532452e-01 5.34111083e-01 5.73033810e-01
5.36010742e-01 -8.39069486e-01 4.71174996e-03 2.86566406e-01
-4.29942459e-01 -1.18692827e+00 -7.89793193e-01 -1.24455199e-01
-4.95245069e-01 -1.15022969e+00 -8.02077889e-01 -1.00896919e+00
2.74032384e-01 5.74489117e-01 1.43682671e+00 -2.62042224e-01
-9.42814276e-02 3.80641848e-01 -5.01130998e-01 -3.01017404e-01
-6.30178332e-01 2.28998840e-01 7.07001835e-02 -4.07915115e-01
7.27374077e-01 -2.90458262e-01 -2.60972887e-01 9.93062928e-03
-8.58981848e-01 -2.97737479e-01 1.13314264e-01 7.23530829e-01
2.85549134e-01 -1.95629865e-01 6.81874156e-01 -9.02756810e-01
1.01230180e+00 -1.11059475e+00 -5.45589700e-02 2.99891382e-01
-6.23763084e-01 -2.27713436e-02 8.48301053e-01 -2.98742920e-01
-8.68783951e-01 -8.89996588e-01 2.60355562e-01 -1.20793819e-01
-1.18943617e-01 7.83905268e-01 3.56746465e-01 5.18954277e-01
8.49088728e-01 2.09456477e-02 -1.90523699e-01 -7.53777385e-01
3.53372425e-01 9.50580597e-01 9.35612202e-01 -4.59371090e-01
6.70752943e-01 3.78556669e-01 -5.52138329e-01 -7.22845137e-01
-1.25060964e+00 -7.08138764e-01 -2.63306350e-01 1.42564103e-01
6.41304433e-01 -9.09457624e-01 -7.48955458e-03 5.37757874e-01
-1.18771613e+00 8.82272050e-02 -2.46955901e-01 2.58133233e-01
-3.22178006e-01 4.72062886e-01 -5.22650421e-01 3.13703716e-01
-2.58740664e-01 -4.36512738e-01 8.18814516e-01 3.97862084e-02
-9.59677875e-01 -1.37663984e+00 2.24961311e-01 4.61934268e-01
4.01296258e-01 -5.81886843e-02 1.12439406e+00 -1.29194868e+00
1.32655546e-01 -4.15059209e-01 -3.83558601e-01 1.30056813e-01
3.96118224e-01 1.32070228e-01 -3.62405509e-01 -1.81772262e-01
-1.09982304e-01 -2.91909367e-01 7.82748997e-01 1.04235277e-01
1.14507079e+00 -5.35156131e-01 -4.82053399e-01 2.84820676e-01
1.23289824e+00 7.18685538e-02 2.97871053e-01 1.01355326e+00
5.65787256e-01 4.73445326e-01 3.15518260e-01 4.34500009e-01
5.36585629e-01 6.06644928e-01 -6.84561431e-02 4.07007560e-02
-2.44011238e-01 -5.42281985e-01 1.19809672e-01 1.19381642e+00
4.56831872e-01 -4.24327523e-01 -1.14325869e+00 8.70655894e-01
-1.74512482e+00 -1.16126084e+00 7.50954077e-02 2.02948833e+00
1.01622069e+00 2.75286317e-01 -1.27245516e-01 -2.72372812e-01
9.92166042e-01 4.17494565e-01 -5.46815693e-01 -3.39981377e-01
-5.61343849e-01 6.28303736e-02 4.81035322e-01 2.50055611e-01
-1.04097414e+00 7.86100686e-01 6.38970757e+00 7.90169001e-01
-9.70500112e-01 -3.95041192e-03 2.94543386e-01 -2.63329417e-01
-7.80093789e-01 5.52673712e-02 -6.14003599e-01 9.36813116e-01
1.02430451e+00 -7.17446923e-01 3.13132107e-01 7.47242928e-01
5.14788441e-02 8.77594948e-02 -1.06706107e+00 9.32468534e-01
6.35484874e-01 -1.66394389e+00 1.99221969e-01 -5.21712840e-01
9.47573483e-01 3.87125075e-01 4.25457582e-02 9.46602374e-02
4.16494131e-01 -7.77400970e-01 1.10104024e+00 6.10940158e-01
5.83147824e-01 -5.70198834e-01 6.02606356e-01 -1.18286282e-01
-9.23714399e-01 -1.98254418e-02 -4.05319124e-01 -9.28459987e-02
2.46027991e-01 7.25259840e-01 -4.05885547e-01 2.05178469e-01
9.85133886e-01 1.64459383e+00 -9.32539403e-01 8.00450683e-01
1.06366454e-02 3.15854162e-01 5.89812808e-02 -3.18915516e-01
4.98567015e-01 -1.25825897e-01 6.95852757e-01 1.33305931e+00
4.61340070e-01 -2.42056057e-01 3.60865965e-02 9.10007477e-01
-5.46750247e-01 1.48008376e-01 -8.54687631e-01 -5.43997705e-01
9.94407833e-01 1.04511344e+00 -6.61585808e-01 -5.84072053e-01
-8.07000160e-01 1.05353713e+00 6.30519927e-01 2.98722237e-01
-6.51362419e-01 -8.87400329e-01 6.73966527e-01 2.18439609e-01
-1.12432845e-01 -1.70171082e-01 -4.00337934e-01 -1.12989211e+00
1.02337070e-01 -8.13781619e-01 3.87010366e-01 -1.08041143e+00
-2.31112504e+00 5.06528258e-01 -1.97133288e-01 -1.02007496e+00
5.00510857e-02 -4.47529465e-01 -6.39822125e-01 7.71062970e-01
-1.12586796e+00 -8.06276381e-01 -3.34413707e-01 2.86841094e-01
8.93615246e-01 -5.60337722e-01 5.92843831e-01 1.20655872e-01
-1.65572450e-01 4.27881002e-01 6.60799444e-01 2.70469725e-01
1.23380494e+00 -1.17249250e+00 9.67959225e-01 4.94505495e-01
2.16125056e-01 8.04610193e-01 6.83083951e-01 -8.19176495e-01
-8.86783481e-01 -9.41754520e-01 1.36155009e+00 -7.11257219e-01
1.53157127e+00 -1.35313317e-01 -9.87160265e-01 5.62420309e-01
4.41265613e-01 -4.58584726e-01 7.49737084e-01 9.29953307e-02
-7.90169001e-01 1.39818117e-01 -8.56018364e-01 8.58922005e-01
1.29361379e+00 -1.18276787e+00 -1.51142704e+00 5.80322802e-01
9.07839060e-01 -3.03756800e-02 -9.16225135e-01 -2.85535678e-02
4.74479139e-01 -3.93637866e-01 1.02206039e+00 -9.90562379e-01
1.12774467e+00 1.95862949e-02 -3.77393991e-01 -1.67004848e+00
-5.52755117e-01 -3.39662790e-01 3.32752854e-01 1.34346986e+00
4.52201873e-01 -6.79255724e-01 8.69190097e-02 5.90177059e-01
-5.60092926e-01 -3.02087754e-01 -5.97969830e-01 -9.36366916e-01
3.78076732e-01 -2.09530920e-01 5.03459454e-01 1.55437326e+00
3.10047239e-01 5.53636253e-01 2.97370553e-01 -5.75570703e-01
2.22204179e-01 3.55554551e-01 2.65059203e-01 -1.32347775e+00
-3.25886421e-02 -9.18547392e-01 -4.49495852e-01 -9.08430159e-01
5.48186898e-01 -1.49270976e+00 -1.73710406e-01 -1.55404174e+00
5.63523710e-01 -4.80861008e-01 -2.08468407e-01 3.41258824e-01
-1.35306567e-01 1.52145535e-01 -1.49098877e-03 4.99993116e-01
-5.18923998e-01 3.71306747e-01 1.04647553e+00 -4.51986343e-01
1.30724609e-01 -5.78110337e-01 -1.11448848e+00 8.31873000e-01
7.92898655e-01 -2.81928599e-01 -2.85398692e-01 -6.65389776e-01
6.46031260e-01 -6.34502172e-01 5.80618143e-01 -8.64776492e-01
7.66802952e-02 -1.78889006e-01 3.53126466e-01 -3.34930778e-01
-2.71254480e-01 -6.06995106e-01 -8.23999196e-02 1.45867586e-01
-1.09272361e+00 6.69544637e-01 6.47988915e-02 8.32884312e-01
-4.13630456e-01 -3.70104671e-01 6.83726132e-01 -3.45951498e-01
-9.49388921e-01 -8.33977833e-02 -3.55300725e-01 8.61221552e-01
7.33363152e-01 -1.41389892e-01 -1.00110626e+00 -3.24415356e-01
-2.91486830e-01 -1.47598758e-01 8.02066803e-01 1.15809107e+00
3.30312550e-01 -1.48125470e+00 -7.08244860e-01 -2.98207581e-01
4.75860596e-01 -4.38790828e-01 -1.30610630e-01 4.21371937e-01
-3.65122557e-01 4.46141452e-01 -2.38118976e-01 -7.19470829e-02
-7.36786723e-01 6.25914574e-01 -1.76795289e-01 2.20869258e-01
-8.70501459e-01 6.08541191e-01 3.96952704e-02 -4.17089343e-01
-1.23824008e-01 -3.27084720e-01 -9.44003761e-02 4.03522074e-01
5.94370723e-01 4.51994777e-01 7.82809872e-03 -7.01590598e-01
-2.25811616e-01 5.70788443e-01 -1.63982823e-01 -1.17268965e-01
1.48917937e+00 -4.82709438e-01 -1.81874633e-01 7.18303263e-01
1.83724630e+00 -7.41524696e-02 -5.08136749e-01 -6.32908463e-01
3.53622288e-01 -4.73366678e-01 -2.13609487e-02 -7.88962066e-01
-7.35476792e-01 5.17635465e-01 3.70727474e-04 3.19325507e-01
6.25033438e-01 5.08309007e-01 1.35105395e+00 5.86602092e-01
1.71684936e-01 -1.47511530e+00 2.23571077e-01 8.14162970e-01
1.25929976e+00 -1.30393934e+00 1.37344208e-02 4.93508615e-02
-5.78902125e-01 9.72892344e-01 3.58291000e-01 -2.27620214e-01
5.40608287e-01 -1.18024945e-02 -1.35095164e-01 -3.56874675e-01
-4.05422598e-01 2.66996443e-01 5.61359346e-01 4.61493820e-01
5.18356562e-01 -1.13982789e-01 -5.05769789e-01 1.00525582e+00
-6.27848446e-01 -4.64161366e-01 5.16151726e-01 8.48991692e-01
-5.90948761e-01 -6.93898976e-01 -1.34469301e-01 8.72792244e-01
-2.72235245e-01 -5.10579586e-01 -7.66101658e-01 4.78244901e-01
-2.60747850e-01 8.03226888e-01 5.90687871e-01 -2.02169821e-01
1.71801448e-01 3.73481512e-02 5.71170896e-02 -6.42061532e-01
-6.06185794e-01 -5.69558859e-01 2.57147551e-01 -4.21180099e-01
-4.49166328e-01 -7.32015729e-01 -1.23528659e+00 -6.17622435e-01
1.80332929e-01 -3.82506810e-02 7.15432823e-01 1.02509177e+00
4.71857697e-01 2.11874932e-01 5.42261362e-01 -5.67819953e-01
-4.79560792e-01 -9.32001233e-01 -6.51429415e-01 1.12414622e+00
1.38278663e-01 -4.12011713e-01 -4.88823801e-01 2.03947052e-01] | [10.979652404785156, 8.976202964782715] |
2326461c-adf2-4c28-83fa-25007641d559 | adaptive-multi-source-predictor-for-zero-shot | 2303.10383 | null | https://arxiv.org/abs/2303.10383v1 | https://arxiv.org/pdf/2303.10383v1.pdf | Adaptive Multi-source Predictor for Zero-shot Video Object Segmentation | Both static and moving objects usually exist in real-life videos. Most video object segmentation methods only focus on exacting and exploiting motion cues to perceive moving objects. Once faced with static objects frames, moving object predictors may predict failed results caused by uncertain motion information, such as low-quality optical flow maps. Besides, many sources such as RGB, depth, optical flow and static saliency can provide useful information about the objects. However, existing approaches only utilize the RGB or RGB and optical flow. In this paper, we propose a novel adaptive multi-source predictor for zero-shot video object segmentation. In the static object predictor, the RGB source is converted to depth and static saliency sources, simultaneously. In the moving object predictor, we propose the multi-source fusion structure. First, the spatial importance of each source is highlighted with the help of the interoceptive spatial attention module (ISAM). Second, the motion-enhanced module (MEM) is designed to generate pure foreground motion attention for improving both static and moving features used in the decoder. Furthermore, we design a feature purification module (FPM) to filter the inter-source incompatible features. By the ISAM, MEM and FPM, the multi-source features are effectively fused. In addition, we put forward an adaptive predictor fusion network (APF) to evaluate the quality of optical flow and fuse the predictions from the static object predictor and the moving object predictor in order to prevent over-reliance on the failed results caused by low-quality optical flow maps. Experiments show that the proposed model outperforms the state-of-the-art methods on three challenging ZVOS benchmarks. And, the static object predictor can precisely predicts a high-quality depth map and static saliency map at the same time. | ['Huchuan Lu', 'Lihe Zhang', 'Jiaxing Yang', 'Youwei Pang', 'Shijie Chang', 'Xiaoqi Zhao'] | 2023-03-18 | null | null | null | null | ['video-object-segmentation', 'video-semantic-segmentation'] | ['computer-vision', 'computer-vision'] | [ 1.47794709e-01 -3.85360360e-01 -2.42932737e-01 -2.00382471e-01
-4.01756644e-01 -1.13961466e-01 2.08500683e-01 -2.05120519e-01
-4.40414160e-01 6.09908402e-01 1.31178305e-01 1.49217799e-01
5.09637184e-02 -6.31903470e-01 -6.41846538e-01 -8.22447717e-01
3.41095269e-01 -1.14020124e-01 1.14834082e+00 2.58196555e-02
4.82182801e-01 9.74750295e-02 -1.81383169e+00 3.78384501e-01
1.02960801e+00 1.25421798e+00 8.92938852e-01 6.22116268e-01
-3.59170407e-01 1.06319809e+00 -4.99598116e-01 3.50281633e-02
1.56422138e-01 -5.35242975e-01 -5.73001266e-01 7.44721666e-02
4.59858894e-01 -8.62792909e-01 -4.46049154e-01 1.29643452e+00
3.64511222e-01 2.70221412e-01 1.16098277e-01 -1.44709074e+00
-4.72729027e-01 2.55357563e-01 -6.76562786e-01 7.63830721e-01
3.01310033e-01 5.33176184e-01 5.77256083e-01 -8.33480299e-01
6.96300328e-01 1.27388144e+00 5.01807146e-02 5.38857818e-01
-6.52768493e-01 -6.26572669e-01 5.84399819e-01 7.62191236e-01
-1.11904132e+00 -4.97042358e-01 9.99493122e-01 -4.09698904e-01
5.09596229e-01 2.48780504e-01 7.88309395e-01 6.78142786e-01
2.78137237e-01 1.22829795e+00 7.77744234e-01 2.20176458e-01
1.05334677e-01 1.32379502e-01 8.93163159e-02 7.31716037e-01
1.13016598e-01 9.48392376e-02 -6.64822280e-01 3.28521460e-01
8.79317522e-01 2.15430290e-01 -8.03140044e-01 -1.34291023e-01
-1.35843527e+00 2.00798735e-01 5.98822713e-01 3.56587529e-01
-3.55002254e-01 -3.66893150e-02 6.58130124e-02 -2.53085732e-01
1.35637492e-01 2.21218746e-02 -4.83696699e-01 -2.13652402e-01
-1.06620502e+00 -2.40024310e-02 3.53349179e-01 8.42779160e-01
1.05728889e+00 1.95208877e-01 -5.55987477e-01 4.24310803e-01
5.55080354e-01 6.33741200e-01 7.83056200e-01 -1.14663768e+00
4.89801645e-01 5.87299943e-01 3.28354836e-01 -1.38167226e+00
-3.12612951e-01 -2.96656013e-01 -6.20708942e-01 4.90013175e-02
3.22265267e-01 2.07039908e-01 -1.12088156e+00 1.59988666e+00
5.24976254e-01 8.62161875e-01 1.26771601e-02 1.71601045e+00
1.11508799e+00 8.03957641e-01 7.34954551e-02 -4.84947383e-01
9.70123112e-01 -1.25975358e+00 -8.34869027e-01 -4.00550425e-01
2.05178857e-01 -7.20825553e-01 8.64315093e-01 6.76963925e-02
-1.08897316e+00 -8.43555808e-01 -1.06427693e+00 -2.20191911e-01
2.10701022e-02 7.49515966e-02 3.96980286e-01 2.61456221e-01
-8.16813946e-01 6.27871931e-01 -9.17427659e-01 9.80618075e-02
4.24341261e-01 3.80993664e-01 -4.53863740e-02 -2.48424232e-01
-1.20059359e+00 6.90832078e-01 5.31080067e-01 3.84488136e-01
-1.00293326e+00 -5.84907651e-01 -6.78896189e-01 2.37243045e-02
5.22317708e-01 -6.93314314e-01 1.05327845e+00 -1.38984394e+00
-1.42125261e+00 2.04045549e-01 -5.27828693e-01 -4.47028317e-02
3.40293974e-01 -1.47698537e-01 -3.33649457e-01 5.60417652e-01
2.23731264e-01 9.38257635e-01 8.94975662e-01 -1.26573658e+00
-1.22045588e+00 -2.27502465e-01 4.33314145e-02 4.76801485e-01
-8.47164765e-02 -2.37401173e-01 -7.15602160e-01 -3.80595565e-01
4.26494807e-01 -4.82702315e-01 3.83237302e-02 -2.33397516e-03
-3.31389397e-01 4.34371531e-02 1.04959226e+00 -8.37865949e-01
1.33830643e+00 -2.13719273e+00 3.54732007e-01 -2.02051654e-01
3.28941941e-01 4.66856629e-01 -9.67540517e-02 -6.29764259e-01
3.17484647e-01 -4.81691808e-02 -2.38541201e-01 -4.57972586e-02
-5.56973517e-01 2.96291351e-01 -1.52473748e-01 3.90214473e-01
3.12708765e-01 9.81081009e-01 -1.29227340e+00 -8.83484900e-01
6.53110802e-01 4.43172038e-01 -4.75904047e-01 2.08193958e-01
-2.33641997e-01 7.13518798e-01 -7.54635334e-01 8.03328037e-01
8.16355765e-01 -1.69958472e-01 -3.40092808e-01 -3.95664126e-01
-3.10651213e-01 1.23797558e-01 -1.34210956e+00 1.78958368e+00
-1.46582825e-02 4.78008360e-01 1.11449659e-01 -5.11232853e-01
6.25205159e-01 1.70491815e-01 6.68364644e-01 -8.43925714e-01
2.79227436e-01 3.81247371e-01 9.38523337e-02 -6.63273096e-01
5.25354207e-01 2.73676723e-01 4.83437389e-01 -2.29722280e-02
1.21870525e-01 2.82854974e-01 6.35001361e-02 1.75357625e-01
8.09486032e-01 3.15373391e-01 -1.64274231e-01 3.11635099e-02
9.64623511e-01 -8.96466896e-02 1.26709914e+00 4.05661851e-01
-7.55572975e-01 8.99488211e-01 2.91913450e-01 -3.18981767e-01
-7.67067552e-01 -9.48451877e-01 1.12443820e-01 7.71836638e-01
1.28097057e+00 -1.93581834e-01 -5.87977231e-01 -6.08518898e-01
-1.95207387e-01 4.61194485e-01 -4.17008042e-01 -2.88144827e-01
-5.95123827e-01 -5.51608086e-01 -1.10209852e-01 4.92385685e-01
8.82952809e-01 -1.09977102e+00 -9.43928301e-01 4.77334023e-01
-5.59517622e-01 -1.21903157e+00 -6.80157065e-01 -3.58271450e-01
-8.09648693e-01 -1.04885709e+00 -8.44641447e-01 -6.27923369e-01
3.88898343e-01 8.13083947e-01 6.37512505e-01 2.51965791e-01
7.06231818e-02 -7.85417797e-04 -3.46060842e-01 -5.87991090e-04
2.16305479e-02 -1.95200756e-01 -1.82109594e-01 3.74970764e-01
2.08659530e-01 -3.47719908e-01 -1.05033052e+00 4.82155204e-01
-8.98404062e-01 4.63382006e-01 5.55687189e-01 5.54945230e-01
6.23524010e-01 -1.88040987e-01 3.41861963e-01 -3.19401205e-01
-1.41888365e-01 -5.61245620e-01 -4.49525654e-01 1.26255780e-01
-2.94478983e-01 -9.97048616e-02 4.32119429e-01 -5.81312001e-01
-1.17840958e+00 1.35795131e-01 2.36624591e-02 -9.13203359e-01
-3.37021761e-02 1.23708531e-01 -4.61156726e-01 2.12968439e-02
2.13549301e-01 4.45920378e-01 -2.21026853e-01 -3.26575786e-01
8.47648680e-02 5.53387523e-01 7.53036499e-01 -5.07638790e-02
5.05214810e-01 5.97618461e-01 -1.33779675e-01 -5.01541138e-01
-8.54338884e-01 -6.41996861e-01 -4.96117413e-01 -6.19932353e-01
1.11053813e+00 -9.27680016e-01 -7.10037708e-01 7.61941612e-01
-1.34966600e+00 -5.83214983e-02 6.05592616e-02 6.20987535e-01
-4.24560755e-01 3.39323312e-01 -5.83563566e-01 -8.12264860e-01
-1.91571265e-01 -1.57212794e+00 1.25988603e+00 9.02704895e-01
3.54007483e-01 -5.33313632e-01 -4.29381281e-01 3.35405707e-01
3.26338559e-01 1.87228303e-02 3.00261587e-01 -1.78331137e-01
-1.32012498e+00 3.81743342e-01 -5.48751831e-01 1.33738160e-01
2.03514144e-01 5.29737361e-02 -1.04053915e+00 5.24286628e-02
2.26899058e-01 2.64103055e-01 1.05759716e+00 6.18332505e-01
9.83001709e-01 -1.45206243e-01 -3.51926118e-01 8.98110628e-01
1.31531584e+00 5.33761621e-01 6.21062756e-01 3.65514249e-01
1.20924795e+00 5.63208818e-01 9.43477690e-01 2.10515246e-01
5.51962197e-01 5.85862517e-01 6.25848234e-01 2.38606613e-02
-3.05494189e-01 -1.59136787e-01 4.82968956e-01 7.26165593e-01
-6.19967803e-02 -1.12452917e-01 -6.25025570e-01 6.87594831e-01
-2.11000800e+00 -1.03385341e+00 -3.47488672e-01 2.12808895e+00
6.57918274e-01 2.76292473e-01 -1.76098272e-01 5.56197912e-02
8.12314510e-01 2.41992176e-01 -7.99111128e-01 1.77777767e-01
-4.31321532e-01 -2.67365068e-01 4.44043964e-01 4.88351852e-01
-1.00933123e+00 9.96875823e-01 4.21825790e+00 8.10129941e-01
-1.39245319e+00 4.01773781e-01 6.46994233e-01 -2.58191019e-01
-4.00442064e-01 1.12214454e-01 -8.19142401e-01 9.58135545e-01
3.84382874e-01 -7.43017271e-02 4.07291293e-01 7.55556285e-01
3.72553617e-01 -6.20034039e-01 -7.36407340e-01 1.02542269e+00
7.53307641e-02 -1.16564929e+00 -1.15171902e-01 -2.23807648e-01
7.81315207e-01 -3.44421752e-02 2.35110912e-02 -4.02633920e-02
-4.65186656e-01 -5.86470366e-01 9.84147072e-01 8.97359788e-01
3.37660402e-01 -5.91227829e-01 8.33493888e-01 4.07602876e-01
-1.38810527e+00 -2.93410748e-01 -3.77826393e-01 1.15138561e-01
4.64101911e-01 6.68849349e-01 -2.12598652e-01 6.86258554e-01
8.31671536e-01 1.06388831e+00 -5.79180419e-01 1.30289626e+00
-2.02392086e-01 2.66010255e-01 -3.27349544e-01 1.34426892e-01
3.45662743e-01 -1.40359312e-01 8.25121582e-01 7.48376369e-01
2.37029493e-01 3.54818493e-01 3.12938273e-01 9.43642795e-01
4.67006117e-01 -1.80567075e-02 1.91821400e-02 2.79294431e-01
3.18058610e-01 1.18257678e+00 -7.96012700e-01 -5.91682196e-01
-5.16561151e-01 1.14106858e+00 -2.11807750e-02 5.28298140e-01
-9.59679008e-01 -1.15729854e-01 7.00173140e-01 -9.85836796e-03
3.93743694e-01 -5.94482087e-02 -2.15272859e-01 -1.47479486e+00
1.39496714e-01 -3.44222069e-01 1.21402226e-01 -1.12951767e+00
-7.66609371e-01 5.77856719e-01 -2.85451770e-01 -1.38035464e+00
1.32928025e-02 -3.60591084e-01 -4.83331919e-01 9.88488793e-01
-1.84066856e+00 -7.91246891e-01 -6.46866202e-01 6.43927217e-01
8.17777932e-01 1.63009480e-01 -2.12752409e-02 5.04220843e-01
-7.63247967e-01 1.34110019e-01 -2.76681483e-01 -8.91393423e-02
6.72609925e-01 -8.97597313e-01 -8.99548382e-02 1.17320883e+00
-1.14871204e-01 2.62767851e-01 5.13583124e-01 -9.27036941e-01
-1.25494432e+00 -1.19372523e+00 4.85864371e-01 -2.65996218e-01
2.01211840e-01 9.92025286e-02 -1.30431497e+00 2.54515201e-01
-2.78324723e-01 3.80846083e-01 -1.10004276e-01 -7.57532001e-01
2.80314267e-01 -2.30272293e-01 -9.19292092e-01 4.04994965e-01
9.91262436e-01 -2.82179207e-01 -4.54173476e-01 -1.95575505e-02
1.02188754e+00 -6.93587244e-01 -4.44860488e-01 6.37302995e-01
4.65771973e-01 -1.24953020e+00 8.85467708e-01 -2.20064178e-01
6.11923218e-01 -9.34758723e-01 -2.52398644e-02 -8.76757860e-01
-2.58883893e-01 -2.93383300e-01 -4.83614475e-01 1.23051250e+00
-7.86633641e-02 -2.68104017e-01 7.69232988e-01 5.53969085e-01
-4.43564415e-01 -7.53260493e-01 -9.42407012e-01 -3.33196282e-01
-6.67870164e-01 -4.44530278e-01 4.89136517e-01 7.19915152e-01
-4.79211271e-01 1.54976413e-01 -4.02475089e-01 3.01748067e-01
5.18150508e-01 3.00402552e-01 4.89970684e-01 -9.82239783e-01
-1.74188092e-01 -5.53323746e-01 -6.05974436e-01 -1.35303056e+00
1.88206527e-02 -5.87359190e-01 3.10576439e-01 -1.49636102e+00
1.94186464e-01 -2.71266907e-01 -6.09388471e-01 2.15590954e-01
-7.95613825e-01 1.56634927e-01 5.78360915e-01 4.16431427e-01
-7.67993808e-01 6.25727296e-01 1.71937203e+00 -1.51680276e-01
-5.21804035e-01 -6.12800270e-02 -3.47822905e-01 7.72328734e-01
6.02226317e-01 -3.19935203e-01 -5.02419829e-01 -5.48004925e-01
-3.28214139e-01 3.62212092e-01 5.94772398e-01 -1.20868790e+00
5.53235114e-01 -3.88249338e-01 6.86505675e-01 -7.90954530e-01
3.08619171e-01 -6.68591499e-01 -4.47671078e-02 4.09954876e-01
1.62208602e-01 -2.92318851e-01 6.70573339e-02 7.14803040e-01
-3.70728105e-01 -7.56388456e-02 7.51854897e-01 -1.77977711e-01
-1.34650326e+00 5.98278105e-01 -6.56354427e-02 -4.08886001e-02
1.01884699e+00 -5.08099496e-01 -4.29395348e-01 -1.64361879e-01
-5.16077816e-01 6.42484784e-01 4.37240422e-01 7.18561113e-01
1.00841570e+00 -1.11442029e+00 -3.58446836e-01 3.30425620e-01
-1.27758920e-01 2.82154262e-01 7.12301731e-01 1.28670800e+00
-4.44705486e-01 1.89074934e-01 -4.91871685e-01 -9.12437975e-01
-9.35096502e-01 7.65050650e-01 4.76306558e-01 3.09751630e-01
-3.57458740e-01 8.89817059e-01 5.40069222e-01 4.15242821e-01
2.65320744e-02 -5.40907800e-01 -4.41007644e-01 -3.27401757e-02
7.72784293e-01 4.16735649e-01 -3.57759953e-01 -1.00756156e+00
-4.51678365e-01 7.36250520e-01 2.04399243e-01 -7.67809376e-02
8.71905029e-01 -6.29010677e-01 -5.79193346e-02 6.03331327e-01
1.08958721e+00 -2.16342881e-01 -1.70154464e+00 -1.71150073e-01
-3.59480977e-01 -8.41359854e-01 3.70110244e-01 -5.58489382e-01
-1.45844376e+00 1.04592156e+00 7.43455350e-01 -3.03862780e-01
1.41319859e+00 -1.98111475e-01 1.15614986e+00 -2.99974650e-01
3.59385669e-01 -8.98362756e-01 6.86306283e-02 3.59815657e-01
4.20958370e-01 -1.34458053e+00 -1.80345878e-01 -6.44558907e-01
-8.90161037e-01 9.65961695e-01 1.16886628e+00 3.86428498e-02
4.42598403e-01 -1.37534307e-03 -7.78776454e-03 2.24758148e-01
-6.80371404e-01 -5.00908434e-01 5.20040512e-01 4.69465226e-01
-8.51797760e-02 -2.87806123e-01 -2.26509646e-01 7.65708685e-01
1.95053235e-01 7.54177868e-02 5.67535460e-01 7.61209786e-01
-7.25585580e-01 -6.23348713e-01 -3.21994811e-01 4.10010189e-01
-3.76309305e-01 -3.97013798e-02 2.33470835e-02 3.43529940e-01
6.04231477e-01 9.89452302e-01 2.65987605e-01 -6.57245994e-01
6.38666376e-02 -2.68080026e-01 2.55142123e-01 -3.23133647e-01
-3.84958416e-01 3.50362003e-01 -3.60694885e-01 -9.54062462e-01
-7.07624614e-01 -5.31740367e-01 -1.62953973e+00 -1.18844688e-01
-3.83028358e-01 -7.55229890e-02 3.77439559e-01 1.02688587e+00
2.48624355e-01 7.33325899e-01 4.96161342e-01 -1.07023680e+00
1.46350771e-01 -7.64545739e-01 -5.40663838e-01 3.88527691e-01
6.52580976e-01 -8.80786836e-01 -3.80616695e-01 7.15573132e-02] | [9.39942455291748, -0.3724464774131775] |
e1f72492-3571-4f4d-a15f-dda60075ab85 | citecaselaw-citation-worthiness-detection-in | 2305.03508 | null | https://arxiv.org/abs/2305.03508v1 | https://arxiv.org/pdf/2305.03508v1.pdf | CiteCaseLAW: Citation Worthiness Detection in Caselaw for Legal Assistive Writing | In legal document writing, one of the key elements is properly citing the case laws and other sources to substantiate claims and arguments. Understanding the legal domain and identifying appropriate citation context or cite-worthy sentences are challenging tasks that demand expensive manual annotation. The presence of jargon, language semantics, and high domain specificity makes legal language complex, making any associated legal task hard for automation. The current work focuses on the problem of citation-worthiness identification. It is designed as the initial step in today's citation recommendation systems to lighten the burden of extracting an adequate set of citation contexts. To accomplish this, we introduce a labeled dataset of 178M sentences for citation-worthiness detection in the legal domain from the Caselaw Access Project (CAP). The performance of various deep learning models was examined on this novel dataset. The domain-specific pre-trained model tends to outperform other models, with an 88% F1-score for the citation-worthiness detection task. | ['Ponnurangam Kumaraguru', 'Rajiv Ratn Shah', 'Yaman Kumar', 'Reshma Sheik', 'Gitansh Satija', 'Pritish Wadhwa', 'Mann Khatri'] | 2023-05-03 | null | null | null | null | ['specificity'] | ['natural-language-processing'] | [ 2.25798979e-01 -5.44128940e-02 -6.48416638e-01 -1.93603292e-01
-1.40317273e+00 -7.43448973e-01 8.70711744e-01 2.87420571e-01
-4.08054173e-01 7.86415160e-01 4.56422925e-01 -1.04229796e+00
-6.84466124e-01 -3.29971403e-01 -3.82494062e-01 -1.23594508e-01
5.43237448e-01 4.68075901e-01 -1.10674329e-01 -2.51909882e-01
1.10644960e+00 4.22701001e-01 -1.07315314e+00 4.13040340e-01
1.13885105e+00 8.14659774e-01 2.16268823e-02 3.21612477e-01
-5.76050639e-01 1.06487250e+00 -1.07294118e+00 -8.43093812e-01
7.86688738e-03 -2.63342649e-01 -8.92072737e-01 -3.29725116e-01
9.39673960e-01 -3.38640481e-01 -3.00101757e-01 1.00610530e+00
4.92097259e-01 -2.63828844e-01 9.85994101e-01 -9.29424167e-01
-1.01888192e+00 8.45075071e-01 -5.85905194e-01 9.63774264e-01
2.83541858e-01 -1.51854977e-01 1.28489912e+00 -5.67784786e-01
9.78317618e-01 9.10377085e-01 6.25873625e-01 3.87990355e-01
-6.67294383e-01 -6.84742928e-01 -6.65521994e-02 3.74114066e-01
-8.42442393e-01 -5.84424794e-01 9.90715325e-01 -9.78975713e-01
1.05609834e+00 -5.53926490e-02 1.94563255e-01 1.55528319e+00
3.16485316e-01 3.98750991e-01 9.60198462e-01 -6.49375200e-01
1.19425267e-01 -1.86018601e-01 6.65502250e-01 1.22581884e-01
7.56049454e-01 -2.17405155e-01 -2.18228400e-01 -5.71920097e-01
7.94046074e-02 -2.81036049e-02 -2.26764157e-02 5.48968077e-01
-6.03900731e-01 7.93678641e-01 -1.45754725e-01 7.92408109e-01
-4.09846485e-01 7.04109743e-02 5.29854178e-01 2.54048675e-01
4.63642985e-01 8.05259526e-01 -1.69945776e-01 -3.78970742e-01
-1.13929498e+00 6.69327080e-01 7.17698216e-01 8.49945426e-01
-2.79680751e-02 -3.84699762e-01 -5.90061784e-01 1.01138318e+00
2.49983206e-01 4.77613360e-01 1.58086345e-01 -1.07991385e+00
1.16069043e+00 7.60912240e-01 1.67977676e-01 -1.03566265e+00
-1.87566608e-01 -7.66851962e-01 -3.39024454e-01 1.52273148e-01
5.08132100e-01 -1.56877592e-01 -8.19381654e-01 1.29522336e+00
-2.94766396e-01 -3.34498912e-01 5.00216186e-02 3.91995609e-01
6.87784195e-01 4.15700316e-01 4.42388594e-01 -1.02794625e-01
1.31048870e+00 -2.64593959e-01 -9.25611615e-01 -3.43596190e-01
6.55085266e-01 -1.07445073e+00 9.57660675e-01 -1.12219572e-01
-8.91207457e-01 -1.94599815e-02 -1.10489058e+00 -4.01841521e-01
-4.18036908e-01 3.15191746e-01 4.90361661e-01 2.64732003e-01
-3.11325103e-01 5.18666863e-01 5.53089790e-02 -2.06557468e-01
9.90225494e-01 -4.12545890e-01 -7.52335712e-02 -2.42148429e-01
-1.30958009e+00 1.33371842e+00 7.46532285e-04 -1.27691790e-01
-1.78079262e-01 -8.41311276e-01 -4.40076619e-01 3.02183151e-01
5.15353858e-01 -3.54594409e-01 1.28952897e+00 -2.72774279e-01
-4.70999181e-01 1.14700675e+00 -1.50020450e-01 -4.63800251e-01
6.97103500e-01 -3.04398954e-01 -7.83944905e-01 4.41008154e-03
9.48300600e-01 -1.98046759e-01 7.68821836e-01 -8.01051319e-01
-7.39492655e-01 -2.25527585e-01 2.28402525e-01 -4.49174345e-01
-2.69809842e-01 7.57154405e-01 -8.35115463e-02 -8.22782278e-01
-3.93789023e-01 -6.74475253e-01 1.14849918e-01 -4.31773812e-01
-4.01444376e-01 -7.85047650e-01 8.38023722e-01 -9.86300766e-01
1.55051661e+00 -1.91476059e+00 -5.15113235e-01 8.37013200e-02
4.53705907e-01 3.78476083e-01 8.17146897e-02 5.85292876e-01
-4.51285318e-02 5.41771531e-01 -2.88351268e-01 1.50307280e-03
2.84610301e-01 -1.84549429e-02 -6.93311512e-01 2.86241472e-01
3.17732424e-01 8.90774965e-01 -8.33836138e-01 -7.04658151e-01
-2.99387366e-01 1.51004583e-01 -2.56425977e-01 -1.98358968e-01
-2.44165435e-01 8.30601156e-02 -7.00159431e-01 8.61860812e-01
4.80535984e-01 -5.47514498e-01 3.22966352e-02 3.86146791e-02
-1.86197668e-01 1.10002851e+00 -6.42012894e-01 1.16871440e+00
-4.17712480e-01 1.00654280e+00 -1.23781919e-01 -8.10279906e-01
7.31761515e-01 2.90655851e-01 3.61323327e-01 -1.21996176e+00
2.21231297e-01 6.70074761e-01 1.77613571e-01 -9.62493598e-01
2.50659496e-01 1.45229623e-02 -3.06146771e-01 7.63337851e-01
-4.20061588e-01 2.33653352e-01 7.59954810e-01 4.62506086e-01
1.20182121e+00 -2.31865689e-01 1.15199432e-01 -1.97083980e-01
3.76588374e-01 2.00259492e-01 6.35716140e-01 9.05457914e-01
4.51417714e-02 3.20736319e-01 8.05260599e-01 -4.64307457e-01
-9.07898664e-01 -5.19608974e-01 -4.87211764e-01 6.91374481e-01
-4.03336406e-01 -2.40198284e-01 -6.23868883e-01 -9.38230574e-01
2.66986549e-01 1.08237910e+00 -4.96755630e-01 4.75536324e-02
-7.83882618e-01 -4.16502148e-01 5.74474096e-01 5.25389612e-01
3.01289856e-01 -1.05871439e+00 -7.57799208e-01 3.16311210e-01
-3.72488499e-01 -1.21346223e+00 -3.60028207e-01 -2.18372062e-01
-3.00151587e-01 -1.57241488e+00 -6.17725074e-01 -4.36905175e-01
3.00117671e-01 -1.23992629e-01 9.57750320e-01 4.07237113e-01
-2.20644578e-01 -9.08710808e-02 -1.77815914e-01 -7.77890444e-01
-6.00621462e-01 2.48401731e-01 -5.16983271e-01 -4.49668586e-01
9.77094650e-01 -3.56147856e-01 -3.06560509e-02 -4.03428704e-01
-8.31783414e-01 -4.55227584e-01 5.56693554e-01 6.90970123e-01
4.82605509e-02 -3.31192553e-01 1.10268056e+00 -1.21797013e+00
1.51659691e+00 -5.24939001e-01 -7.63493240e-01 3.93337965e-01
-1.00543141e+00 -1.51338279e-01 3.65064234e-01 -8.72787759e-02
-1.03048766e+00 -8.36180389e-01 -2.62614042e-01 5.75242452e-02
-2.67715380e-02 8.61090362e-01 1.05989680e-01 3.99926573e-01
8.33225489e-01 -3.53737533e-01 -1.65634602e-01 -6.06919527e-01
1.64773434e-01 1.15775132e+00 5.25066614e-01 -6.85248911e-01
6.67454779e-01 6.06164113e-02 -1.25693738e-01 -3.84870976e-01
-1.45296264e+00 -4.20742244e-01 -5.56612432e-01 -2.49941051e-01
5.82122087e-01 -3.28546733e-01 -3.53210747e-01 -8.97832960e-02
-1.64142096e+00 3.62002820e-01 4.91379388e-03 3.50863010e-01
1.20347470e-01 4.49033111e-01 -3.94560456e-01 -7.37626433e-01
-5.06614208e-01 -1.04773712e+00 8.39676261e-01 -8.48033652e-02
-7.64352143e-01 -7.16229141e-01 4.56219502e-02 1.06345296e+00
4.52251792e-01 3.17736894e-01 1.47245324e+00 -9.04274642e-01
-1.29631832e-01 -6.61808729e-01 -4.82289791e-01 2.68895745e-01
-8.69486202e-03 1.49675518e-01 -8.66118133e-01 3.26625347e-01
-5.06370291e-02 3.24396696e-03 9.46583271e-01 2.56637216e-01
9.10116851e-01 -5.55354416e-01 -3.10855657e-01 4.75087762e-02
1.17448592e+00 4.84676450e-01 5.02933204e-01 8.13198864e-01
4.74144250e-01 5.62863231e-01 3.70828450e-01 2.36927122e-01
2.62721300e-01 4.81244713e-01 1.83075424e-02 2.44388968e-01
-2.44348377e-01 -2.03912985e-02 -6.72974735e-02 2.04524577e-01
-2.35508963e-01 -3.02529216e-01 -1.39271045e+00 8.22860479e-01
-1.71292555e+00 -1.52348578e+00 -2.41674840e-01 1.46341276e+00
7.54926503e-01 5.28209329e-01 -1.21301852e-01 3.16756606e-01
7.07472622e-01 1.33583695e-01 -3.97687018e-01 -6.43187106e-01
-1.63381994e-01 1.36859104e-01 3.26075375e-01 3.18799883e-01
-1.10340500e+00 6.01161659e-01 6.00736094e+00 7.17256606e-01
-9.38827813e-01 2.15528905e-01 4.35046852e-01 -6.63893446e-02
-5.28505385e-01 7.51934946e-02 -9.11073685e-01 9.88938510e-01
8.53456378e-01 -3.08482081e-01 -9.47590172e-02 9.80113685e-01
3.04774821e-01 2.04217464e-01 -9.69664752e-01 6.62302434e-01
2.46518433e-01 -2.07385373e+00 1.89180952e-04 4.20408815e-01
5.86450219e-01 7.93255940e-02 1.97258522e-03 3.79107118e-01
4.15820256e-02 -9.66246486e-01 8.34980130e-01 3.76837969e-01
7.37672150e-01 -4.07892644e-01 1.12157726e+00 2.24939510e-01
-1.39111727e-01 -4.27249759e-01 -2.92020857e-01 -1.26218140e-01
4.37213838e-01 8.08628321e-01 -6.43532813e-01 1.66421399e-01
5.56587398e-01 8.47315192e-01 -3.72128278e-01 9.88306224e-01
-3.51942331e-01 7.25232065e-01 8.27081576e-02 -1.41017795e-01
4.67243552e-01 -1.74719933e-02 5.73142648e-01 1.41199219e+00
3.22811097e-01 1.27148122e-01 -1.34788901e-01 1.13227642e+00
-7.70856678e-01 -4.44047637e-02 -7.37351060e-01 -4.43401307e-01
7.82486498e-01 9.23713744e-01 -3.06547552e-01 -4.02316064e-01
-5.17607808e-01 1.56705886e-01 2.62644768e-01 2.45960847e-01
-6.68351352e-01 -5.98622978e-01 4.50267375e-01 4.64226574e-01
2.33326629e-01 8.63571465e-02 -6.83713377e-01 -8.75735283e-01
4.74674821e-01 -8.03585291e-01 5.79575837e-01 -6.95901990e-01
-1.64262152e+00 4.36479270e-01 -1.88146043e-03 -1.24612927e+00
-2.16138154e-01 -5.67776620e-01 -8.82044733e-01 1.04941356e+00
-1.88304484e+00 -1.13560152e+00 1.41661599e-01 -7.86533728e-02
4.69858825e-01 -4.73755032e-01 3.63115579e-01 7.17084587e-01
-5.69542825e-01 4.75730777e-01 1.49533480e-01 4.62707758e-01
7.96588242e-01 -9.56837893e-01 2.83453524e-01 1.08826971e+00
-1.18503785e-02 1.08519208e+00 5.35343170e-01 -9.61907864e-01
-8.84004533e-01 -7.69809604e-01 1.81300056e+00 -8.72159541e-01
1.01362431e+00 1.11651361e-01 -9.22017336e-01 4.13621426e-01
3.98859501e-01 -6.93861604e-01 7.90552497e-01 4.19349313e-01
-7.49785900e-01 -3.20726149e-02 -1.08187485e+00 4.69270527e-01
1.01576769e+00 -8.45725715e-01 -1.41888070e+00 6.66953385e-01
2.32973278e-01 -1.85141519e-01 -7.14695215e-01 5.13131991e-02
5.36228597e-01 -3.71727884e-01 7.93855846e-01 -8.43916595e-01
9.61443186e-01 -5.37619218e-02 1.15213253e-01 -6.90206230e-01
-3.57524306e-01 -4.12151515e-01 1.61415490e-03 1.59088922e+00
8.03988636e-01 -4.90850776e-01 1.31756380e-01 1.00442517e+00
-2.87365139e-01 -5.68637133e-01 -1.14722502e+00 -7.47583210e-01
3.01384211e-01 -7.20372319e-01 7.25886524e-01 1.35053718e+00
2.91088015e-01 8.68463039e-01 6.88061193e-02 -2.16510922e-01
5.76223552e-01 5.65785229e-01 3.09307724e-01 -1.58263826e+00
1.42708883e-01 -9.24439907e-01 -3.93598601e-02 -4.80546117e-01
5.72277367e-01 -9.70658064e-01 -5.36569595e-01 -1.98581052e+00
2.59161055e-01 -4.85188782e-01 -2.00484961e-01 5.74703693e-01
-1.82633132e-01 -3.54699790e-01 6.38831556e-02 5.27164340e-01
-3.61537218e-01 -1.85166672e-01 9.71973598e-01 -6.78589463e-01
3.24243367e-01 -2.19137874e-02 -1.39803147e+00 7.00573921e-01
7.36030936e-01 -7.36942470e-01 -2.16775373e-01 -8.03088784e-01
4.90596682e-01 -1.64309397e-01 2.23692372e-01 -5.81052423e-01
3.80515277e-01 -4.29419249e-01 1.03807978e-01 -7.41860151e-01
-1.72446415e-01 -5.26523709e-01 -3.99551153e-01 9.80699658e-02
-8.44438791e-01 1.47929892e-01 4.51751128e-02 4.51181412e-01
-2.04059422e-01 -7.68814802e-01 4.20578301e-01 -9.53106135e-02
-4.80046362e-01 -1.09047830e-01 -5.16481638e-01 5.18850446e-01
6.73228204e-01 -1.85065284e-01 -9.39858377e-01 -9.72112939e-02
-1.93182155e-01 -5.94081394e-02 2.41425291e-01 6.06411159e-01
4.13489729e-01 -8.81215990e-01 -1.00504506e+00 -2.97856629e-01
3.51828188e-01 -3.33307534e-01 -1.25211418e-01 7.25731015e-01
-3.15073699e-01 7.11545110e-01 2.80408002e-02 6.46126196e-02
-1.03567040e+00 3.32312584e-01 -2.53385939e-02 -1.76571414e-01
-7.71987140e-01 5.11223614e-01 -5.21371484e-01 2.19933584e-01
1.14504792e-01 -4.16487038e-01 -7.52955079e-01 2.18120843e-01
7.53786147e-01 5.87922692e-01 2.89639741e-01 -5.95026970e-01
-3.89095277e-01 5.10731280e-01 -4.20473069e-01 1.25334635e-01
1.54550493e+00 2.72001505e-01 -4.09002185e-01 -3.61774489e-03
1.06177175e+00 2.82775462e-01 -2.75867313e-01 -4.35841493e-02
8.21646750e-01 -5.87235808e-01 2.69491732e-01 -1.15198648e+00
-6.79246366e-01 8.73392701e-01 5.62710979e-04 3.30838352e-01
4.18930590e-01 2.76688367e-01 8.32133055e-01 4.55668807e-01
-4.31311391e-02 -1.51854122e+00 -4.79765534e-01 7.19681144e-01
1.23074710e+00 -1.18251491e+00 4.08859067e-02 -7.77864382e-02
-5.03628612e-01 1.24886382e+00 3.72592688e-01 2.08076015e-01
5.37294149e-01 2.41059616e-01 2.32245713e-01 -8.95515800e-01
-4.65748787e-01 1.08319595e-01 4.90189433e-01 2.84672141e-01
7.13739157e-01 -2.84695476e-01 -9.90625441e-01 7.03351736e-01
-1.57581568e-02 9.32668746e-02 6.28379762e-01 9.54139352e-01
-3.36979866e-01 -1.03712523e+00 -2.61453331e-01 1.10367596e+00
-1.18652141e+00 -2.85450578e-01 -7.69922018e-01 6.89911902e-01
8.79654214e-02 1.14571083e+00 -1.50096506e-01 1.91380978e-01
3.88551652e-01 1.86322927e-01 -4.98519875e-02 -6.15962029e-01
-6.06974840e-01 -1.24209756e-02 7.34935820e-01 -2.36987844e-02
-4.32048470e-01 -7.29115725e-01 -1.09655321e+00 -2.66476005e-01
-2.47069389e-01 3.22714567e-01 6.55053377e-01 1.54749727e+00
5.99578500e-01 5.04260778e-01 -8.75141397e-02 5.51177524e-02
-7.56679893e-01 -1.05245769e+00 -2.15216517e-01 7.01733172e-01
2.68000185e-01 -8.21708500e-01 -3.37886900e-01 4.64726686e-02] | [9.941329956054688, 9.28042221069336] |
c01ae2e7-c04c-4cd0-bbec-a501429ad327 | sustainability-using-renewable-electricity | 2202.13101 | null | https://arxiv.org/abs/2202.13101v1 | https://arxiv.org/pdf/2202.13101v1.pdf | Sustainability using Renewable Electricity (SuRE) towards NetZero Emissions | Demand for energy has increased significantly across the globe due to increase in population and economic growth. Growth in energy demand poses serious threat to the environment since majority of the energy sources are non-renewable and based on fossil fuels, which leads to emission of harmful greenhouse gases. Organizations across the world are facing challenges in transitioning from fossil fuels-based sources to greener sources to reduce their carbon footprint. As a step towards achieving Net-Zero emission target, we present a scalable AI based solution that can be used by organizations to increase their overall renewable electricity share in total energy consumption. Our solution provides facilities with accurate energy demand forecast, recommendation for procurement of renewable electricity to optimize cost and carbon offset recommendations to compensate for Greenhouse Gas (GHG) emissions. This solution has been used in production for more than a year for four facilities and has increased their renewable electricity share significantly. | ['Rengaraj Ramasubbu', 'Tamal Bhattacharyya', 'Sandeep Vaity', 'Harshit Sampgaon', 'Rajesh kumar Palani', 'Gopali Contractor', 'Sreedhar Seetharam', 'Bhushan Jagyasi', 'Pallavi Gawade', 'Saurabh Pashine', 'Jinu Jayan'] | 2022-02-26 | null | null | null | null | ['total-energy'] | ['miscellaneous'] | [-5.27582988e-02 2.10094512e-01 -1.76492810e-01 2.44819090e-01
-2.89478987e-01 -7.62149811e-01 4.82426316e-01 4.49687801e-02
1.72926217e-01 1.17271233e+00 5.11368692e-01 -2.30655864e-01
1.16795816e-01 -1.37604821e+00 8.64112750e-03 -6.40457869e-01
5.47100544e-01 2.70614833e-01 -3.83842468e-01 -2.19645336e-01
7.79950798e-01 5.97269475e-01 -1.09266782e+00 -5.86358488e-01
1.28932416e+00 5.78786552e-01 3.42125624e-01 4.65033650e-01
-2.03709498e-01 7.31556937e-02 -2.99312949e-01 4.95005101e-01
7.54235029e-01 -4.92900223e-01 -7.19497740e-01 -7.01775610e-01
-3.35917979e-01 -3.54395658e-01 4.52768087e-01 1.08347499e+00
4.24250662e-01 1.93354845e-01 6.40996635e-01 -1.33394229e+00
-7.25880980e-01 4.52846587e-01 -6.55283272e-01 -3.65656056e-02
-2.39137009e-01 3.49202037e-01 5.02143562e-01 -5.12378931e-01
2.14457050e-01 1.02733481e+00 1.79405794e-01 4.27681625e-01
-9.97077465e-01 -1.04899049e+00 -5.05774677e-01 5.13016060e-02
-1.24551761e+00 -2.10960656e-01 5.73314667e-01 -1.90709963e-01
1.80809748e+00 5.70522964e-01 1.22261012e+00 1.12916045e-01
6.59404695e-01 -5.73386073e-01 1.37523007e+00 -7.00183690e-01
4.33841377e-01 -7.29904184e-03 -3.90737444e-01 -6.40675575e-02
1.07611549e+00 -1.11818522e-01 9.56274644e-02 6.09832443e-02
4.96398538e-01 -8.29594433e-02 4.45921496e-02 4.33703452e-01
-1.00389802e+00 3.56020719e-01 4.49171484e-01 7.20552564e-01
-5.13678312e-01 7.41640031e-01 -2.67494977e-01 -2.43546411e-01
1.66378602e-01 6.51593506e-01 -3.76854241e-01 -4.63180453e-01
-4.91919935e-01 -1.10716835e-01 6.73136830e-01 6.37896836e-01
8.46596479e-01 4.84970540e-01 1.35611594e-01 3.66974920e-01
4.65463310e-01 1.20858979e+00 2.33035564e-01 -1.52214253e+00
1.10642398e-02 9.09077764e-01 7.14017570e-01 -8.75061631e-01
-3.72741997e-01 -4.45729017e-01 -6.45188272e-01 3.05232465e-01
-3.16487849e-01 -4.08544868e-01 -1.02757740e+00 1.21054876e+00
2.29514658e-01 -2.46567726e-01 -1.92243978e-01 3.65031660e-01
-4.98259068e-02 1.23131335e+00 3.48176122e-01 -5.11610746e-01
1.01785195e+00 -7.97232330e-01 -8.90941799e-01 2.60656834e-01
2.46677071e-01 -3.50375056e-01 5.23771048e-01 2.64227867e-01
-1.10535836e+00 -1.28234684e-01 -7.79253960e-01 2.00383514e-01
-8.93871009e-01 -5.52701414e-01 1.46288946e-01 9.57798719e-01
-9.37308371e-01 3.66454393e-01 -7.16791570e-01 -7.03388155e-01
2.35496506e-01 3.08411628e-01 2.25879043e-01 2.72608459e-01
-7.80902088e-01 1.43144548e+00 7.78917432e-01 -1.63088843e-01
-5.57569504e-01 -8.89270484e-01 -1.45675927e-01 4.98731136e-01
1.33356929e-01 -6.42145216e-01 8.74320924e-01 -3.26316476e-01
-1.19429636e+00 7.99454302e-02 1.43434480e-01 -1.96374223e-01
2.20250636e-01 -8.26751664e-02 -7.74853349e-01 -1.88539863e-01
-8.92656147e-02 6.26470506e-01 8.80542845e-02 -1.19580615e+00
-9.34933841e-01 -3.52322072e-01 -4.21386540e-01 3.20956886e-01
-5.69745421e-01 -2.57825285e-01 7.81073272e-01 -4.32072654e-02
-3.85891318e-01 -1.37882268e+00 -2.92659253e-01 -1.04178464e+00
-2.34023049e-01 -5.16604006e-01 1.25407052e+00 -6.56271636e-01
1.16479564e+00 -1.29616976e+00 -3.64712387e-01 2.89278030e-01
-3.26311529e-01 8.96358937e-02 3.26291889e-01 1.04635835e+00
5.14091067e-02 1.05620241e+00 -1.64848026e-02 7.51649857e-01
1.47112504e-01 1.94729373e-01 -4.18758355e-02 -4.93090935e-02
-6.63088914e-03 6.54176474e-01 -1.19860053e+00 7.28697106e-02
5.14427602e-01 4.57042158e-01 7.22776875e-02 3.55781801e-02
-1.30548954e-01 6.69845760e-01 -3.64261001e-01 8.90600264e-01
8.89858842e-01 8.82108435e-02 3.13073754e-01 -1.10917903e-01
-9.96039152e-01 1.03688486e-01 -8.78293991e-01 1.25363874e+00
-7.93064952e-01 3.03850353e-01 -1.11825824e-01 -5.50578356e-01
7.95279026e-01 2.71473289e-01 7.84131467e-01 -9.37933028e-01
-1.92153051e-01 8.21835577e-01 2.13705570e-01 -2.96500593e-01
6.12734377e-01 -3.54826033e-01 2.31365860e-01 3.99295837e-01
-4.56281513e-01 -4.23825473e-01 5.25754809e-01 -2.28237569e-01
9.04692233e-01 3.75334173e-01 4.16503698e-01 -1.00677562e+00
4.68794912e-01 1.08943179e-01 6.86464250e-01 -2.33505294e-01
-1.12383731e-01 -2.84340769e-01 -4.11730170e-01 -2.32182339e-01
-1.27181709e+00 -9.38923538e-01 4.29957872e-03 8.13212991e-01
2.07836524e-01 1.55846670e-01 -6.06833279e-01 -7.07695633e-02
-6.81652427e-02 1.64852190e+00 4.20175456e-02 6.44943789e-02
-6.08562648e-01 -7.12740958e-01 -1.53919175e-01 3.48211199e-01
1.01283813e+00 -5.81834853e-01 -1.24517381e+00 3.44308764e-01
-3.59112710e-01 -5.28052688e-01 -3.21450740e-01 1.65599525e-01
-6.62178755e-01 -7.84781337e-01 -6.21136189e-01 3.24984640e-02
8.02604556e-01 4.94853437e-01 1.00174284e+00 4.43357183e-03
-6.21927559e-01 3.66268963e-01 -1.31149199e-02 -1.33083212e+00
-2.92636812e-01 4.73550614e-03 3.06163579e-01 -1.02955198e+00
2.77588427e-01 -6.38076067e-01 -1.39716697e+00 -3.46723548e-03
-5.04650652e-01 2.32681885e-01 4.31569010e-01 -2.15356573e-01
4.95134056e-01 7.34622717e-01 1.34936571e+00 -4.97583538e-01
6.71344936e-01 -8.98199141e-01 -7.24722743e-01 1.80151924e-01
-1.54018486e+00 -1.33184075e-01 5.05759358e-01 5.14551163e-01
-1.38888264e+00 -2.89618522e-01 5.13626158e-01 3.32463503e-01
3.08497678e-02 8.42292756e-02 -7.62555897e-02 1.58201009e-01
4.07037744e-03 -2.60129124e-01 -5.54002643e-01 -3.33315879e-01
2.90477723e-01 6.72387242e-01 5.50161839e-01 -3.21490884e-01
1.40063393e+00 1.94606021e-01 5.96124470e-01 -6.00002587e-01
-7.54303485e-02 -3.00028652e-01 -1.15781970e-01 -5.15342355e-01
9.71606255e-01 -7.31079698e-01 -7.34094083e-01 -3.66713293e-02
-6.64862514e-01 -1.14225507e-01 -1.19927652e-01 3.89188588e-01
-9.43674613e-03 -3.41837913e-01 4.39710617e-01 -1.54618955e+00
-9.17558312e-01 -4.15867090e-01 1.09198883e-01 1.11614692e+00
-1.51003778e-01 -7.55048752e-01 4.37709600e-01 1.76112607e-01
1.18811572e+00 9.95108366e-01 1.16087162e+00 -4.16722707e-02
-9.22677934e-01 2.34942451e-01 -3.39464843e-01 4.24926579e-01
1.05417991e+00 2.61172950e-01 -5.85424125e-01 -5.02035737e-01
-1.59348682e-01 2.44815961e-01 4.26003903e-01 2.77081788e-01
9.68842924e-01 -6.94717646e-01 -5.67087471e-01 -8.10722560e-02
2.46130157e+00 9.68987942e-01 6.52585685e-01 4.74467784e-01
3.86077583e-01 3.67727429e-01 6.04386330e-01 4.75832522e-01
3.42000037e-01 -6.06003143e-02 8.80723357e-01 2.34142855e-01
8.40958729e-02 1.20425401e-02 8.68328614e-04 9.86184776e-01
-8.38867366e-01 -2.17887640e-01 -1.18332863e+00 9.49885488e-01
-1.53125310e+00 -1.02570677e+00 -4.40035224e-01 2.42046642e+00
6.09644234e-01 -3.91928285e-01 6.44631609e-02 -2.18485340e-01
5.83776176e-01 -3.59284222e-01 -6.88378453e-01 -9.30487573e-01
5.44098437e-01 1.63567886e-01 1.22517860e+00 2.58018579e-02
-8.81038979e-02 2.67558932e-01 6.43236589e+00 1.45712599e-01
-1.15601993e+00 7.69402012e-02 6.80388689e-01 -2.79602408e-01
-9.44129705e-01 2.86959559e-01 -6.06523335e-01 7.96832800e-01
1.68370295e+00 -1.48968232e+00 8.68422687e-01 5.22813559e-01
9.46135104e-01 -4.99548703e-01 -6.54184163e-01 5.74727416e-01
-6.34052932e-01 -1.53754640e+00 -2.21223891e-01 7.02730417e-01
1.51975119e+00 9.97103155e-02 -3.28422993e-01 6.08805157e-02
8.51581514e-01 -1.11550200e+00 1.76659346e-01 6.91488981e-01
7.25132108e-01 -1.55707777e+00 5.02097905e-01 4.10748303e-01
-1.47145164e+00 -4.67241347e-01 -2.87282318e-01 -8.27397779e-02
1.16847076e-01 6.12346947e-01 -9.49482620e-01 8.22588325e-01
9.10242558e-01 -3.48769486e-01 4.61323857e-02 6.21923745e-01
1.45407975e-01 5.78731596e-01 -6.11306489e-01 1.40764385e-01
1.31808162e-01 -7.73053527e-01 2.98817843e-01 9.95802879e-01
9.68525112e-01 4.10183758e-01 -4.55458388e-02 8.80967319e-01
-3.58351082e-01 6.87081665e-02 -7.15865910e-01 -4.78733182e-01
1.16269302e+00 1.67504740e+00 -8.97725821e-01 -7.59847909e-02
-2.13555336e-01 2.09133565e-01 -6.04854405e-01 8.84625241e-02
-8.38797927e-01 -7.21415818e-01 4.44451600e-01 2.49502629e-01
-2.67358512e-01 -4.56090093e-01 -2.82145202e-01 -2.63494253e-01
-4.44773048e-01 -1.98417366e-01 -2.04325452e-01 -5.35185695e-01
-7.31808186e-01 -1.78185448e-01 2.09058017e-01 -7.91716576e-01
-3.88932139e-01 -9.55201164e-02 -1.00009608e+00 1.26410627e+00
-1.91360211e+00 -1.00139773e+00 -6.92870975e-01 -4.43993974e-03
6.87776446e-01 9.13052857e-02 8.86435211e-01 -2.10199952e-02
-4.57194835e-01 -5.70049942e-01 4.73460048e-01 -6.13817215e-01
5.14788866e-01 -1.63885379e+00 -1.87877547e-02 8.72927427e-01
-7.52667308e-01 6.33850396e-01 6.51894271e-01 -8.70913267e-01
-1.85163069e+00 -1.30295861e+00 3.79491180e-01 3.57079618e-02
5.06007791e-01 2.27806076e-01 -2.75016218e-01 4.04900163e-01
1.25693738e+00 -6.61875725e-01 1.01145387e+00 -5.31447768e-01
2.47211784e-01 -3.03373486e-01 -1.78151417e+00 4.33305323e-01
7.14557588e-01 9.05168056e-02 3.81513983e-02 2.94249803e-01
6.27359748e-01 2.86257029e-01 -1.19909859e+00 2.64416516e-01
6.26718223e-01 -6.40111268e-01 2.46473938e-01 -5.18819280e-02
3.48786354e-01 -4.51614290e-01 -1.32265091e-01 -1.70686209e+00
-7.20448196e-01 -9.25707281e-01 -1.99617878e-01 1.64294255e+00
1.93050936e-01 -1.05966413e+00 5.48786968e-02 1.32286215e+00
-3.85803312e-01 -4.23758298e-01 -8.53433132e-01 -9.37016845e-01
2.82114059e-01 3.55083257e-01 1.00159848e+00 9.64704692e-01
-1.23213499e-03 1.18128113e-01 -1.05814293e-01 -1.05716139e-01
1.04841816e+00 5.53970002e-02 5.10185540e-01 -1.38557935e+00
7.00816691e-01 -4.12916422e-01 3.20145309e-01 2.14268431e-01
-4.33363259e-01 -8.14599931e-01 -1.86487734e-01 -2.58583975e+00
4.83370095e-01 -5.54363467e-02 -5.50229788e-01 7.60662496e-01
3.26846629e-01 4.00727838e-01 4.21049505e-01 1.64168701e-01
4.35771525e-01 2.84976095e-01 1.00398767e+00 3.75087000e-02
-1.82816729e-01 -7.40765452e-01 -1.13289511e+00 6.80342436e-01
1.26897442e+00 -3.25370818e-01 -5.99999964e-01 -2.35803574e-01
5.75137138e-01 -4.60350156e-01 -1.87118471e-01 -1.16080797e+00
-1.84110537e-01 -1.28517890e+00 3.09930384e-01 -7.81974554e-01
-4.11289155e-01 -1.49719965e+00 1.21354842e+00 9.96548355e-01
5.05279958e-01 -8.55611712e-02 -2.88237687e-02 2.78219998e-01
4.91737902e-01 -1.23070687e-01 8.82301569e-01 -4.36079532e-01
-5.10772645e-01 -1.45357817e-01 -4.22057182e-01 -6.64090812e-01
1.64328873e+00 -5.05548894e-01 -1.11203003e+00 7.15832487e-02
-1.75702363e-01 6.49413288e-01 7.97570348e-01 4.33887124e-01
-1.69268399e-01 -1.41689730e+00 -7.82797694e-01 -6.10591054e-01
-4.70516980e-01 -2.42744267e-01 2.28744615e-02 5.05841792e-01
-8.44769657e-01 8.70445371e-01 -8.56880426e-01 1.98219404e-01
-8.94978046e-01 4.33174968e-01 4.21129882e-01 -3.23254764e-01
-2.42334470e-01 1.01334065e-01 -2.49963328e-01 2.23884493e-01
-5.89824200e-01 -4.18334275e-01 2.12862045e-01 -3.58718038e-02
3.86460215e-01 1.41046417e+00 1.23897102e-02 -2.30765164e-01
-5.56217313e-01 5.04193485e-01 5.36025047e-01 5.19428737e-02
1.56334209e+00 -4.59396631e-01 -5.44565320e-01 6.64836884e-01
7.74388790e-01 3.40655625e-01 -7.04775214e-01 9.73162055e-01
-3.15183919e-04 -6.29247785e-01 3.70706171e-01 -1.47652173e+00
-9.39039648e-01 1.63170516e-01 9.52338338e-01 7.35151172e-01
1.51742589e+00 -6.03265584e-01 1.00375748e+00 4.31385964e-01
6.52192652e-01 -1.88215387e+00 -1.92139670e-01 -5.65085150e-02
8.53072166e-01 -8.49437594e-01 3.29689473e-01 -9.84942541e-02
9.51746181e-02 1.19603622e+00 6.09070361e-01 -1.06718741e-01
3.57073218e-01 1.71391696e-01 -4.65741128e-01 1.71512246e-01
-6.92742407e-01 -1.14372447e-01 -5.64186931e-01 7.91072309e-01
7.92158902e-01 4.71645594e-01 -8.67248774e-01 -3.11857074e-01
1.53348938e-01 2.13245407e-01 6.99914396e-01 8.86371911e-01
-8.96844685e-01 -1.23520601e+00 -5.03482997e-01 7.52830088e-01
-5.97865582e-01 1.43352086e-02 -2.79909492e-01 5.80466986e-01
5.19802272e-01 1.30657578e+00 8.20747688e-02 2.31003195e-01
2.08990082e-01 1.79737017e-01 2.16722220e-01 -3.51495177e-01
-3.77021044e-01 6.27082214e-02 3.19752336e-01 -2.27811426e-01
-5.42935371e-01 -3.39902312e-01 -2.08189440e+00 -7.82151937e-01
-2.98672885e-01 3.18360418e-01 1.52101529e+00 6.10329866e-01
5.75438499e-01 8.28212738e-01 1.08196342e+00 -8.41113627e-01
-3.29542786e-01 -1.03734934e+00 -5.89603603e-01 -1.95746467e-01
-1.84406266e-01 -4.17985976e-01 -1.79012775e-01 3.11031044e-02] | [5.750117301940918, 2.5645477771759033] |
469a40de-fd38-4cde-96af-3d35bc1c535f | diff-nst-diffusion-interleaving-for | 2307.04157 | null | https://arxiv.org/abs/2307.04157v2 | https://arxiv.org/pdf/2307.04157v2.pdf | DIFF-NST: Diffusion Interleaving For deFormable Neural Style Transfer | Neural Style Transfer (NST) is the field of study applying neural techniques to modify the artistic appearance of a content image to match the style of a reference style image. Traditionally, NST methods have focused on texture-based image edits, affecting mostly low level information and keeping most image structures the same. However, style-based deformation of the content is desirable for some styles, especially in cases where the style is abstract or the primary concept of the style is in its deformed rendition of some content. With the recent introduction of diffusion models, such as Stable Diffusion, we can access far more powerful image generation techniques, enabling new possibilities. In our work, we propose using this new class of models to perform style transfer while enabling deformable style transfer, an elusive capability in previous models. We show how leveraging the priors of these models can expose new artistic controls at inference time, and we document our findings in exploring this new direction for the field of style transfer. | ['John Collomosse', 'Nicholas Kolkin', 'Eli Shechtman', 'Andrew Gilbert', 'Gemma Canet Tarrés', 'Dan Ruta'] | 2023-07-09 | null | null | null | null | ['style-transfer', 'image-generation'] | ['computer-vision', 'computer-vision'] | [ 6.75882518e-01 3.55505466e-01 2.07623243e-01 -1.42125651e-01
7.23979101e-02 -8.70811522e-01 8.23753715e-01 -2.46951342e-01
-3.26992631e-01 6.75467730e-01 1.48428753e-01 1.06474772e-01
8.54239613e-02 -1.11250770e+00 -9.00208116e-01 -7.39196956e-01
1.78488657e-01 2.52091438e-01 4.22193527e-01 -5.80019474e-01
4.07384098e-01 7.26870596e-01 -1.37741888e+00 4.08711046e-01
5.61195195e-01 7.61509418e-01 2.51782924e-01 7.07060218e-01
-2.53794283e-01 6.01754427e-01 -7.12206960e-01 -5.69686711e-01
3.73429030e-01 -7.27577388e-01 -6.64072692e-01 -5.96223585e-02
6.74843788e-01 -1.08302973e-01 -3.34449232e-01 1.10558987e+00
2.77016282e-01 1.36267975e-01 8.32472146e-01 -9.68822718e-01
-1.16406512e+00 4.37143117e-01 -4.99150723e-01 -1.07417934e-01
3.79800081e-01 1.94974616e-02 5.91718078e-01 -5.44654727e-01
1.23503947e+00 1.43520486e+00 5.75601101e-01 9.91169691e-01
-1.70503795e+00 -3.72506082e-01 4.44325283e-02 1.26448683e-02
-1.10158539e+00 -3.23066473e-01 1.10255468e+00 -3.38695586e-01
2.26380378e-01 5.36229610e-01 9.18624341e-01 1.41185999e+00
4.24771458e-01 7.69001126e-01 1.57493424e+00 -6.15915358e-01
2.63770819e-01 4.06392932e-01 -4.81251389e-01 4.14906114e-01
-7.00621232e-02 2.72178084e-01 -4.53410327e-01 1.50458768e-01
1.25829148e+00 -1.54077366e-01 -3.96732122e-01 -5.19865513e-01
-1.13420630e+00 6.74240410e-01 4.57722157e-01 5.33237219e-01
-2.21925974e-01 1.71677172e-01 6.51503131e-02 5.29373109e-01
6.71235919e-01 7.94781268e-01 -1.69669747e-01 -6.77218810e-02
-1.11909509e+00 4.04308826e-01 7.73387730e-01 8.16973507e-01
7.65596449e-01 1.54263556e-01 -5.52508652e-01 6.11791909e-01
-6.05366491e-02 3.45838130e-01 2.33198792e-01 -1.10593212e+00
-1.23527147e-01 5.00533879e-01 1.28246527e-02 -9.97937977e-01
1.37483120e-01 -2.32680619e-01 -7.54681468e-01 9.16657269e-01
5.34329951e-01 4.81638797e-02 -1.01450062e+00 1.79885387e+00
1.88009650e-01 -9.96569023e-02 -2.62285739e-01 6.74720943e-01
2.67065167e-01 5.29059529e-01 7.94183314e-02 7.97967166e-02
1.01181841e+00 -5.69410801e-01 -7.03417778e-01 1.64038539e-01
1.36919126e-01 -9.97252822e-01 1.20980883e+00 3.55568707e-01
-1.35525727e+00 -5.22288084e-01 -8.85696173e-01 -2.42432520e-01
-5.19331634e-01 -4.33405995e-01 4.54913884e-01 6.36528194e-01
-1.40471709e+00 9.19520676e-01 -5.87756753e-01 -5.40226936e-01
4.91150588e-01 1.59302771e-01 -2.76240170e-01 2.85989821e-01
-1.06355155e+00 9.62171018e-01 1.38731629e-01 -1.35157064e-01
-6.77028477e-01 -9.15330052e-01 -5.76501608e-01 -5.16518578e-02
2.21422568e-01 -9.02385175e-01 8.30764413e-01 -1.61488223e+00
-2.00761938e+00 1.10596597e+00 8.54780674e-02 -1.20688185e-01
1.07665730e+00 -6.35896623e-02 -1.48218557e-01 -2.72322726e-02
-2.58504212e-01 1.14620626e+00 1.35241902e+00 -1.51576161e+00
-1.38767734e-01 -1.52230665e-01 1.39130533e-01 -4.87241447e-02
-4.56381410e-01 -1.41957805e-01 -2.69635558e-01 -1.30656052e+00
-2.37414286e-01 -1.08692777e+00 -8.26352760e-02 6.71860337e-01
-2.44944096e-01 1.53367713e-01 1.26701558e+00 -4.94078547e-01
8.26996148e-01 -2.23063350e+00 6.84501290e-01 2.79848963e-01
2.80629158e-01 3.23967934e-01 -3.42140436e-01 5.29071331e-01
-8.77998397e-02 2.83337593e-01 -4.93657380e-01 -3.90737832e-01
-1.02058582e-01 3.02685022e-01 -4.30639207e-01 1.17212318e-01
3.99056196e-01 1.09063387e+00 -7.01341927e-01 -3.93779337e-01
5.94032519e-02 7.95866787e-01 -7.86061823e-01 2.64806509e-01
-3.29916477e-01 1.01302528e+00 -3.40270370e-01 1.56631783e-01
5.90177298e-01 2.11837769e-01 -1.42720222e-01 -2.30519399e-01
-1.73553854e-01 -3.53720903e-01 -9.27861214e-01 1.94874978e+00
-4.35625494e-01 7.83718109e-01 1.65626913e-01 -6.56279385e-01
8.24247897e-01 5.33400550e-02 2.19186619e-01 -6.43536806e-01
1.05997756e-01 -1.25344202e-01 -4.82453778e-02 -3.22437733e-01
6.52895451e-01 -3.97229135e-01 2.60762781e-01 5.16795099e-01
-1.87744409e-01 -7.25353777e-01 1.05257101e-01 1.32719204e-02
7.83861935e-01 6.31914437e-01 -1.94917232e-01 -4.66738969e-01
2.45736063e-01 -2.75588810e-01 1.67095974e-01 7.09017634e-01
1.79371372e-01 1.05302954e+00 4.93741632e-01 -4.70901370e-01
-1.43105662e+00 -1.21906221e+00 -9.21181887e-02 8.42385709e-01
-1.40878469e-01 5.81958215e-04 -1.05595636e+00 -5.97326458e-01
-1.79769307e-01 7.73055434e-01 -1.02175438e+00 -3.71917605e-01
-7.06460297e-01 -3.00769895e-01 4.89236444e-01 2.75478691e-01
4.81264442e-01 -1.46699822e+00 -5.95514655e-01 1.95410758e-01
2.69269675e-01 -7.80524254e-01 -6.76291704e-01 -6.50329143e-02
-9.96503353e-01 -6.11978054e-01 -1.44866550e+00 -6.81581795e-01
8.63471389e-01 -3.16280186e-01 1.17750645e+00 1.71574667e-01
-3.32461476e-01 4.33597058e-01 -2.61365354e-01 -3.77644897e-01
-8.69817674e-01 8.63098577e-02 -1.32636219e-01 3.96081120e-01
-4.23395216e-01 -9.61484313e-01 -5.59575856e-01 1.50344446e-01
-1.57390761e+00 2.24238813e-01 5.15001595e-01 6.38057113e-01
3.05595249e-01 -1.61755502e-01 2.33410463e-01 -1.07871377e+00
7.14588106e-01 -9.13559645e-03 -3.51565599e-01 1.57077387e-01
-4.41568166e-01 3.57972115e-01 5.03796279e-01 -9.32534277e-01
-1.29591751e+00 -2.08996981e-01 -5.28074205e-02 -6.40004694e-01
-2.86043853e-01 1.42569557e-01 -9.87835228e-02 -4.42424744e-01
7.14568615e-01 2.98629940e-01 2.40279794e-01 -5.28528631e-01
6.70164168e-01 6.79316595e-02 3.54179919e-01 -7.47718871e-01
1.04450929e+00 7.76667297e-01 1.85949951e-01 -8.21223736e-01
-4.08961743e-01 3.80187333e-01 -8.66916358e-01 -2.63867110e-01
9.10525858e-01 -2.17143729e-01 -4.40716147e-01 5.92678845e-01
-1.18235123e+00 -6.86633766e-01 -9.80950594e-01 7.34413555e-03
-7.26339757e-01 4.17429477e-01 -5.61252594e-01 -4.65968519e-01
1.78384669e-02 -1.12919521e+00 9.95458245e-01 2.14532927e-01
-5.77455819e-01 -1.27145135e+00 2.27460787e-01 -1.33993223e-01
7.71124423e-01 4.40343052e-01 1.12179840e+00 -5.01117297e-03
-5.87940931e-01 -4.47293147e-02 -1.21569648e-01 5.00938535e-01
3.45208138e-01 6.32186309e-02 -9.36872184e-01 -3.01627427e-01
8.42855126e-02 -3.27275731e-02 7.31215715e-01 2.06932187e-01
1.27319968e+00 -2.98768699e-01 -1.67714357e-01 7.31821716e-01
1.21393347e+00 2.35481262e-01 1.06668687e+00 3.22175533e-01
7.67691612e-01 8.30180705e-01 1.09111555e-02 2.82735564e-02
-8.26977640e-02 9.32385862e-01 5.97622544e-02 -4.89535302e-01
-6.51381314e-01 -2.44432688e-01 2.77799129e-01 4.62780356e-01
-4.99709696e-01 -2.22662881e-01 -3.52138102e-01 2.47418806e-01
-1.53267622e+00 -1.04872012e+00 1.54964194e-01 2.21841717e+00
1.06759751e+00 2.24946260e-01 -1.09076157e-01 -5.11509888e-02
6.22489631e-01 9.33111683e-02 -5.50647497e-01 -5.72948039e-01
-2.37515703e-01 4.67362285e-01 2.17386097e-01 5.02073407e-01
-7.31512606e-01 1.05147231e+00 7.07523441e+00 8.99990678e-01
-1.43283653e+00 -9.54153165e-02 6.10935867e-01 -7.97071029e-03
-7.13083565e-01 -5.69993891e-02 -3.40755612e-01 4.66302395e-01
5.06233335e-01 2.02627946e-02 6.85905159e-01 4.95653212e-01
2.14147598e-01 -1.02750465e-01 -1.18001187e+00 6.48616493e-01
1.41598538e-01 -1.45227158e+00 5.78162193e-01 2.44005859e-01
8.21731508e-01 -7.40926564e-01 4.72688526e-01 6.43931702e-03
9.57370549e-02 -8.35118890e-01 1.01276886e+00 1.01500571e+00
1.01863873e+00 -4.84988660e-01 7.51226619e-02 -5.51334172e-02
-7.00775623e-01 2.51700312e-01 -8.21724385e-02 -1.88933779e-02
1.57454774e-01 4.37015712e-01 -3.20582747e-01 2.90205687e-01
6.92745864e-01 5.85089386e-01 -7.42417097e-01 7.28125513e-01
-3.18102688e-02 3.41210127e-01 -9.54530314e-02 1.52383327e-01
-1.29400855e-02 -4.80378538e-01 9.40554142e-01 1.01709461e+00
3.87547225e-01 -9.46951285e-02 -2.66720116e-01 1.49175990e+00
-1.94956064e-02 -3.77453938e-02 -9.50548708e-01 4.67870794e-02
-1.53578758e-01 1.16141880e+00 -9.64064300e-01 -2.88839936e-01
-5.41211702e-02 1.52238691e+00 1.51118353e-01 4.01753336e-01
-6.43551767e-01 -3.77906084e-01 5.68755448e-01 4.07545000e-01
3.00662577e-01 -3.72157335e-01 -3.58449578e-01 -1.09482408e+00
-8.63513872e-02 -8.24462891e-01 -2.68694520e-01 -1.06161582e+00
-1.21265244e+00 7.10031331e-01 1.15405060e-01 -1.02121937e+00
5.54882735e-02 -5.35699069e-01 -7.44865000e-01 8.86721492e-01
-1.18776965e+00 -1.37669206e+00 -4.53848131e-02 5.96860588e-01
5.37392378e-01 -1.61570637e-03 7.06628680e-01 1.24875247e-01
2.27699690e-02 4.88902062e-01 1.05510443e-01 -1.76448077e-01
9.80187356e-01 -1.32270825e+00 3.93857181e-01 6.78982079e-01
1.02645010e-01 7.25375712e-01 1.03071451e+00 -6.90129817e-01
-1.26459157e+00 -7.60574281e-01 4.87879395e-01 -7.47646928e-01
4.57474142e-01 -5.27274728e-01 -1.13335872e+00 4.08343583e-01
5.32645404e-01 -4.41661656e-01 2.37656981e-01 -1.95886001e-01
-1.45560920e-01 -1.13195293e-02 -1.11258459e+00 1.02356923e+00
1.13531029e+00 -5.04570246e-01 -4.84983653e-01 -4.36096378e-02
5.96743107e-01 -3.30451578e-01 -8.23701322e-01 2.09595799e-01
7.36136556e-01 -1.03830743e+00 9.30125773e-01 -4.46625143e-01
4.47620064e-01 -3.73888940e-01 1.90179586e-01 -1.60926306e+00
-3.85972589e-01 -6.91668928e-01 1.84716418e-01 1.40031707e+00
2.88459450e-01 -3.49937290e-01 7.66410947e-01 5.82623005e-01
-5.46131469e-02 -2.08817557e-01 -8.04758489e-01 -7.60307252e-01
3.96678358e-01 -1.11724362e-01 5.50649047e-01 9.56339180e-01
-5.24319589e-01 -1.48398310e-01 -3.64641935e-01 -4.25782055e-01
5.06234884e-01 -1.63655177e-01 6.73845470e-01 -1.17845809e+00
-3.27251196e-01 -7.36626148e-01 -3.30230743e-01 -6.70732021e-01
1.47275820e-01 -8.29064667e-01 -1.94921240e-01 -1.11430395e+00
-3.93554680e-02 -3.81039351e-01 -5.52604459e-02 2.60282516e-01
5.77890500e-02 5.55101693e-01 5.16601682e-01 2.32096702e-01
4.53198738e-02 6.91300333e-01 2.03004098e+00 -2.62107491e-01
-2.10176781e-01 -9.32012796e-02 -5.69892287e-01 6.48510337e-01
5.72749972e-01 -4.70819086e-01 -3.30999643e-01 -4.04793561e-01
3.31057519e-01 -3.06251645e-01 4.23951268e-01 -9.64144230e-01
-1.32083267e-01 -1.83037609e-01 5.81427395e-01 1.13284759e-01
3.27433825e-01 -8.99288476e-01 4.12590295e-01 4.67040718e-01
-5.16799867e-01 1.22895904e-01 2.79274642e-01 5.22065997e-01
-8.37691799e-02 -9.58273411e-02 9.89876926e-01 -2.88612723e-01
-5.98835170e-01 1.31859168e-01 -5.02468765e-01 -9.43653435e-02
8.69617581e-01 -4.70271766e-01 9.60023850e-02 -4.41639692e-01
-9.51645732e-01 -5.73067009e-01 9.28269863e-01 7.27389932e-01
3.69955629e-01 -1.34811473e+00 -5.27409554e-01 3.35981548e-01
-1.29899293e-01 -1.69590488e-01 2.07561746e-01 5.15763879e-01
-5.11663914e-01 -1.45589799e-01 -7.69756913e-01 -4.87829328e-01
-9.45803404e-01 7.07805395e-01 3.28065187e-01 -2.16319915e-02
-7.24930644e-01 5.39610386e-01 6.97141707e-01 -2.68099517e-01
-9.32690129e-02 -2.64696270e-01 -1.38230249e-01 -6.47690818e-02
3.94976854e-01 1.95363343e-01 -1.15720145e-01 -3.47486019e-01
1.92043573e-01 8.38393331e-01 6.48522473e-05 -3.90488923e-01
1.20415974e+00 -1.13952346e-01 -3.42326909e-01 4.65830475e-01
9.01200116e-01 2.20237121e-01 -1.63010192e+00 7.08935633e-02
-2.73306608e-01 -4.36447531e-01 1.31348753e-03 -9.07361805e-01
-1.10892689e+00 8.81192207e-01 8.31199288e-01 3.01146537e-01
1.09870970e+00 -1.12873860e-01 5.47332048e-01 1.62921678e-02
4.12140876e-01 -1.02149260e+00 2.30502725e-01 3.20598513e-01
1.39759648e+00 -8.82164419e-01 -2.76007205e-01 -1.65976092e-01
-5.21164834e-01 1.17475605e+00 4.44192857e-01 -2.60059983e-01
6.70626819e-01 2.74779230e-01 1.81649476e-01 -1.26042679e-01
-2.05595374e-01 1.05835021e-01 5.04407823e-01 7.69086540e-01
4.07119185e-01 -1.62311837e-01 -1.88558191e-01 -1.60242662e-01
-2.56179363e-01 8.05485770e-02 6.84788406e-01 9.63705182e-01
3.53830941e-02 -1.53892052e+00 -4.52994645e-01 1.22700781e-02
-3.98543209e-01 -1.93937853e-01 -5.87929904e-01 8.11857760e-01
9.88033265e-02 4.78345484e-01 8.86896253e-02 -1.27577320e-01
3.39793563e-01 1.18913457e-01 8.97837102e-01 -4.77642834e-01
-5.75525582e-01 7.16905072e-02 -3.58269721e-01 -4.48902726e-01
-5.05035698e-01 -6.71216190e-01 -5.30786395e-01 -3.67100239e-01
1.51765691e-02 -1.94670886e-01 6.71245813e-01 8.30591321e-01
3.74162912e-01 6.09802723e-01 5.17245054e-01 -1.31639671e+00
-5.31553105e-02 -8.15881491e-01 -6.57078564e-01 7.25639641e-01
1.56593099e-01 -6.20823741e-01 -1.05428219e-01 6.39050364e-01] | [11.583467483520508, -0.42698201537132263] |
b6036d68-89ed-4277-8fda-88916fa1e62c | criticality-as-it-could-be-organizational | 1704.05255 | null | http://arxiv.org/abs/1704.05255v3 | http://arxiv.org/pdf/1704.05255v3.pdf | Criticality as It Could Be: organizational invariance as self-organized criticality in embodied agents | This paper outlines a methodological approach for designing adaptive agents
driving themselves near points of criticality. Using a synthetic approach we
construct a conceptual model that, instead of specifying mechanistic
requirements to generate criticality, exploits the maintenance of an
organizational structure capable of reproducing critical behavior. Our approach
exploits the well-known principle of universality, which classifies critical
phenomena inside a few universality classes of systems independently of their
specific mechanisms or topologies. In particular, we implement an artificial
embodied agent controlled by a neural network maintaining a correlation
structure randomly sampled from a lattice Ising model at a critical point. We
evaluate the agent in two classical reinforcement learning scenarios: the
Mountain Car benchmark and the Acrobot double pendulum, finding that in both
cases the neural controller reaches a point of criticality, which coincides
with a transition point between two regimes of the agent's behaviour,
maximizing the mutual information between neurons and sensorimotor patterns.
Finally, we discuss the possible applications of this synthetic approach to the
comprehension of deeper principles connected to the pervasive presence of
criticality in biological and cognitive systems. | ['Manuel G. Bedia', 'Miguel Aguilera'] | 2017-04-18 | null | null | null | null | ['acrobot'] | ['playing-games'] | [ 2.14726895e-01 4.64905828e-01 9.37164649e-02 2.06437990e-01
1.16685897e-01 -4.30781454e-01 1.11737013e+00 2.50426918e-01
-4.95429754e-01 9.39260781e-01 -1.76339433e-01 -9.78418738e-02
-5.28981268e-01 -6.62694573e-01 -7.41998315e-01 -1.27154422e+00
-4.47369426e-01 4.81864125e-01 4.91471499e-01 -8.77456427e-01
4.80811328e-01 1.38146430e-01 -2.01645827e+00 -7.95265734e-02
1.11808419e+00 5.80882072e-01 3.82295549e-01 8.33038032e-01
4.92654413e-01 5.32922089e-01 -3.47196966e-01 1.92413732e-01
1.49991170e-01 -1.00005400e+00 -6.47492290e-01 2.29745880e-02
-7.60837570e-02 5.57800412e-01 -8.50059539e-02 1.07404816e+00
9.98776779e-02 2.34383017e-01 1.19931042e+00 -1.20126951e+00
-6.89070165e-01 5.37072659e-01 -1.00114547e-01 3.25861007e-01
1.70941979e-01 5.24301052e-01 1.02256644e+00 -9.65229869e-02
4.79861081e-01 1.04362988e+00 5.91057241e-01 9.23836350e-01
-1.48789370e+00 -7.70129561e-02 -1.65503353e-01 -2.22843718e-02
-1.11933506e+00 -3.74764442e-01 5.08478343e-01 -6.11073911e-01
8.28576684e-01 -9.70371366e-02 1.02960718e+00 1.00004864e+00
1.17287052e+00 -9.64718638e-04 1.42438424e+00 -7.79857278e-01
9.79551256e-01 -2.38261893e-01 -8.94006458e-04 8.20816934e-01
5.02420485e-01 6.42514229e-01 -5.21287858e-01 -1.99821293e-01
7.01969504e-01 -5.15246511e-01 -1.96026608e-01 -6.36905432e-01
-1.36661947e+00 5.29538512e-01 2.33564645e-01 6.62322164e-01
-5.12359321e-01 3.55646044e-01 4.33686264e-02 6.17000163e-01
-3.03719312e-01 9.74579215e-01 -2.80179024e-01 1.72389925e-01
-2.24138230e-01 3.36357713e-01 7.30748296e-01 4.33988363e-01
4.39405352e-01 -4.15968783e-02 3.57222736e-01 1.91065043e-01
3.10482740e-01 2.82415658e-01 6.53393507e-01 -1.19947398e+00
-4.37080979e-01 7.18975008e-01 1.35048926e-01 -4.67429310e-01
-6.51242733e-01 -4.32351619e-01 -6.99680150e-01 6.72034800e-01
3.36775810e-01 -1.64401740e-01 -4.57852602e-01 2.13664150e+00
5.50923645e-02 -3.19458127e-01 6.41726255e-01 7.50337780e-01
2.34371778e-02 4.55796748e-01 1.47319630e-01 -5.58273196e-01
1.27455938e+00 -4.30439144e-01 -3.58546048e-01 -6.01493008e-02
1.54861704e-01 -4.53962460e-02 1.18651319e+00 2.63455421e-01
-1.15240848e+00 -2.09328264e-01 -1.54028666e+00 6.24343455e-01
-4.47205305e-01 -7.56585121e-01 5.44105589e-01 3.22877169e-01
-1.23799443e+00 9.03686225e-01 -7.77832150e-01 -8.89297009e-01
-3.37967217e-01 3.55395466e-01 -1.21714324e-01 7.97909141e-01
-1.38337970e+00 1.06981432e+00 6.26330733e-01 -3.78242671e-01
-1.06632900e+00 -3.70147347e-01 -2.03418747e-01 3.52273248e-02
8.23441800e-03 -1.03425443e+00 1.14477038e+00 -1.35928500e+00
-1.64551127e+00 9.68326151e-01 3.45232666e-01 -3.89997900e-01
2.78745621e-01 5.61426997e-01 -1.47758320e-01 2.01834947e-01
-1.85460430e-02 6.38967454e-01 6.37144089e-01 -1.46614325e+00
-1.97646797e-01 -2.72219211e-01 2.53296971e-01 2.01979950e-01
-1.21285126e-01 -4.94726062e-01 5.89711666e-01 -2.64470607e-01
1.61675066e-01 -1.17125893e+00 -5.74323475e-01 -4.27972436e-01
-9.26053226e-02 -4.04682547e-01 1.73696965e-01 2.57758260e-01
8.13706100e-01 -1.84168577e+00 6.86648011e-01 3.09096575e-01
3.22972327e-01 -1.43090382e-01 6.31052256e-02 1.09832573e+00
-8.65267217e-02 -7.29376748e-02 -4.95599777e-01 4.96696770e-01
1.71749055e-01 2.50760347e-01 -1.35368958e-01 4.31470573e-01
-7.52518326e-02 7.71499515e-01 -8.48868370e-01 -2.78597146e-01
-2.20839888e-01 2.97859728e-01 -7.63059437e-01 -7.69142807e-02
-6.23758852e-01 7.46984601e-01 -4.43672508e-01 1.36399686e-01
-9.52736381e-03 -8.48255306e-02 4.00644839e-01 5.43819368e-01
-5.95193863e-01 8.52274522e-03 -7.75352657e-01 1.16611242e+00
-1.01216890e-01 4.31972086e-01 -4.16092910e-02 -1.08205450e+00
7.32376575e-01 3.28844100e-01 4.20098156e-01 -8.25430393e-01
3.86608243e-01 3.50182384e-01 5.99611759e-01 -4.02548373e-01
1.54704496e-01 -4.03144449e-01 -3.27782243e-01 7.48941839e-01
2.03594595e-01 -2.37612307e-01 2.85129458e-01 -1.04868393e-02
1.36433172e+00 -3.92978266e-02 4.27138418e-01 -1.28305042e+00
5.66330254e-01 3.08827031e-03 1.99708551e-01 1.06215596e+00
-4.92484868e-01 2.88603842e-01 9.15458858e-01 -4.93958235e-01
-1.43029344e+00 -1.21129119e+00 -3.18326980e-01 7.55142450e-01
5.00392735e-01 -1.81906506e-01 -1.39855957e+00 2.84133673e-01
-2.73489833e-01 6.06399655e-01 -1.14624560e+00 -8.40585530e-01
-4.82742906e-01 -9.25548315e-01 3.15676570e-01 -3.04218799e-01
5.54631531e-01 -1.48774958e+00 -1.58557546e+00 -7.36449892e-03
3.09803206e-02 -4.90460992e-01 2.43245680e-02 6.51082456e-01
-7.38256097e-01 -1.17849970e+00 -2.32073054e-01 -7.07997620e-01
6.21527016e-01 -2.16102730e-02 1.02768254e+00 4.71440226e-01
-3.01308036e-01 4.81736362e-01 -7.10558146e-02 -4.22724709e-02
-1.11388350e+00 2.24244565e-01 7.01259613e-01 -2.62579709e-01
-3.65475118e-02 -9.88079011e-01 -7.43414462e-01 2.43391424e-01
-1.14602077e+00 3.95808257e-02 2.60669947e-01 6.62871242e-01
1.17884792e-01 5.82353696e-02 7.88897395e-01 9.02579874e-02
9.99584556e-01 -3.68169874e-01 -4.59426552e-01 3.14393014e-01
-7.84051001e-01 5.81552386e-01 7.12535083e-01 -4.62121516e-01
-7.71543145e-01 -6.36048764e-02 3.17413896e-01 5.31320751e-01
-2.12385003e-02 1.57262847e-01 -5.82286529e-02 -6.35958388e-02
9.46215510e-01 4.91706043e-01 4.96377081e-01 1.85373887e-01
1.93295956e-01 4.17739868e-01 3.73922259e-01 -8.40946198e-01
5.05600750e-01 3.71003479e-01 3.16791773e-01 -1.01948845e+00
-8.76353607e-02 4.30944413e-01 -7.48484552e-01 -5.81150532e-01
8.27923000e-01 -1.98892087e-01 -1.18354440e+00 6.69615626e-01
-9.48262155e-01 -5.41176915e-01 -7.61694491e-01 1.66753054e-01
-1.52127504e+00 -7.85462633e-02 -4.74268883e-01 -7.61900127e-01
1.03440069e-01 -7.97802567e-01 5.69972098e-01 2.89877385e-01
-2.11768344e-01 -9.95909512e-01 9.17093933e-01 -3.68337899e-01
4.70205754e-01 4.35065627e-01 1.46363556e+00 -2.44463101e-01
-3.98205012e-01 1.10374503e-01 5.09553194e-01 -3.40839773e-01
-2.46612355e-01 2.67429262e-01 -6.80242658e-01 -2.11209893e-01
4.40309122e-02 -1.49240643e-01 8.46780419e-01 3.27292651e-01
-4.05598357e-02 -2.01632798e-01 -2.89331943e-01 8.60567465e-02
1.48579240e+00 6.63590252e-01 5.34815490e-01 5.47174335e-01
-2.45814934e-01 9.55314398e-01 -8.56985375e-02 2.27899745e-01
1.52006283e-01 4.54246283e-01 5.90465009e-01 4.88223255e-01
3.50726604e-01 -1.32033061e-02 4.90931094e-01 8.12588155e-01
-4.11648691e-01 1.87011078e-01 -1.02800083e+00 3.16364378e-01
-1.94474232e+00 -1.16953123e+00 2.28647426e-01 2.40770984e+00
8.19524169e-01 2.10790709e-01 2.86075026e-01 -1.52879402e-01
8.39576781e-01 -9.79390889e-02 -6.33946240e-01 -5.80059826e-01
-4.14191097e-01 -3.67187262e-01 2.43096396e-01 6.85316145e-01
-6.34225845e-01 8.83154809e-01 7.30755758e+00 3.22236568e-01
-8.33123088e-01 2.05755293e-01 4.38350171e-01 1.34097904e-01
-1.48383826e-01 2.38676861e-01 -3.28331023e-01 4.00107235e-01
1.36500859e+00 -4.25612360e-01 7.09880054e-01 9.40498710e-02
1.83611497e-01 -5.16592741e-01 -9.00608003e-01 7.85001293e-02
-2.58401901e-01 -1.31220627e+00 -1.39901519e-01 3.68926257e-01
7.73037791e-01 -1.32442474e-01 1.21338539e-01 1.25819566e-02
6.81824863e-01 -9.24773335e-01 9.16893601e-01 8.79748821e-01
3.64098579e-01 -7.82410920e-01 1.12811528e-01 6.94848061e-01
-7.25354016e-01 -3.94576311e-01 -3.01153392e-01 -4.10684884e-01
-2.06049740e-01 5.07533215e-02 -2.51059793e-02 -2.90533364e-01
5.23380518e-01 -5.87235717e-03 -5.59597790e-01 7.78295875e-01
1.32850185e-01 -3.05108540e-02 -3.23162973e-01 -4.59236741e-01
3.24869096e-01 -4.60260779e-01 7.90411651e-01 6.02901042e-01
9.46012661e-02 1.77431434e-01 -3.96515101e-01 9.19114530e-01
5.58184862e-01 -2.33547136e-01 -1.03251493e+00 1.46450266e-01
1.59052685e-01 9.14806545e-01 -1.37220144e+00 -1.57243479e-02
2.42392749e-01 5.39044142e-01 3.98002982e-01 1.70394227e-01
-8.89784038e-01 -3.10966849e-01 5.72142363e-01 1.71069354e-01
9.15164351e-02 -4.86893058e-01 -1.47541687e-01 -1.12538254e+00
-2.77369916e-01 -7.22622454e-01 -1.76286623e-01 -4.99290586e-01
-7.95789242e-01 8.16161513e-01 5.13579734e-02 -8.03261995e-01
-3.36221784e-01 -3.12254220e-01 -6.94370568e-01 4.46448848e-02
-7.40844309e-01 -6.20205045e-01 1.09861121e-02 6.12314641e-01
-4.24854457e-02 -3.50528061e-01 8.82344425e-01 -4.13056910e-01
-6.04558587e-01 7.46398568e-02 5.58843911e-01 -6.38509631e-01
6.98521584e-02 -1.37504005e+00 3.48337710e-01 5.25882185e-01
-3.22303742e-01 8.95878732e-01 1.33961666e+00 -5.65987408e-01
-1.21819603e+00 -4.49672908e-01 6.95758343e-01 -3.88859987e-01
8.38052750e-01 -5.11359334e-01 -6.21654630e-01 2.44732454e-01
8.38619649e-01 -4.86773044e-01 3.03276181e-01 -1.43261507e-01
-1.15338825e-01 9.18050334e-02 -9.91792738e-01 1.21126425e+00
1.21843088e+00 -2.70333648e-01 -6.95545435e-01 2.35291839e-01
5.85199714e-01 5.29594362e-01 -4.58620161e-01 2.21792638e-01
7.91269958e-01 -1.36978889e+00 5.32929301e-01 -6.95123315e-01
5.31746328e-01 -1.08916342e-01 -1.97234079e-01 -1.55174839e+00
-4.79378909e-01 -1.07032490e+00 1.65435955e-01 6.49193883e-01
2.94572413e-01 -1.06595516e+00 3.09286922e-01 2.09553778e-01
1.44697160e-01 -6.05815649e-01 -1.44139147e+00 -7.44088531e-01
7.40697861e-01 3.44991714e-01 2.91030705e-01 6.28998101e-01
6.98830128e-01 4.05801594e-01 2.19626531e-01 -2.93607324e-01
7.20661581e-01 -1.83315769e-01 2.12692946e-01 -1.36511624e+00
-1.88062906e-01 -7.36187875e-01 -4.59960431e-01 -3.58189076e-01
2.32917115e-01 -6.35372341e-01 2.63179988e-01 -9.91897941e-01
2.29844883e-01 -3.22777241e-01 -4.86845404e-01 -2.22816244e-01
4.99242008e-01 7.62012750e-02 1.00472972e-01 5.96097708e-01
-6.62215292e-01 5.39517164e-01 1.04258192e+00 3.41946900e-01
-3.72398585e-01 -3.37731719e-01 -9.70938027e-01 8.69003356e-01
1.06490552e+00 -4.21623826e-01 -4.60373312e-01 3.84589195e-01
8.03843915e-01 1.66092724e-01 4.64356512e-01 -1.39084435e+00
4.86837119e-01 -4.44616288e-01 -2.09200889e-01 3.39036465e-01
3.94691639e-02 -7.45617330e-01 1.71833262e-01 1.24648547e+00
-6.69808507e-01 5.26636183e-01 -2.37745777e-01 6.58428431e-01
2.91470647e-01 -4.18309987e-01 1.29232657e+00 -3.06468934e-01
-2.40931824e-01 -3.86827976e-01 -1.25645602e+00 1.45015912e-02
1.63977301e+00 -2.51047105e-01 -6.06338918e-01 -8.63029361e-02
-7.35835373e-01 1.30416855e-01 8.70900273e-01 2.83252478e-01
1.73760056e-01 -1.00007761e+00 -4.02947932e-01 4.29037333e-01
1.00906705e-02 -8.14588904e-01 1.78811364e-02 8.59472752e-01
-7.28879690e-01 4.93307590e-01 -9.25549269e-01 -5.07888556e-01
-6.12296045e-01 7.58619308e-01 1.02292049e+00 -7.07224980e-02
-4.93520349e-01 1.78751618e-01 4.08355922e-01 -4.16823268e-01
-3.02781075e-01 -1.39369398e-01 -2.76794195e-01 3.25440951e-02
2.01623857e-01 6.62157685e-02 -1.96313441e-01 -6.99168801e-01
-2.38497823e-01 7.13060558e-01 3.17849904e-01 -5.50920069e-01
1.09883499e+00 -1.18317984e-01 -3.97042423e-01 9.00173426e-01
5.33744931e-01 -3.29270184e-01 -1.41999471e+00 1.84085399e-01
2.76308149e-01 3.28567207e-01 -4.22215164e-01 -4.12309855e-01
-4.80757415e-01 5.49723804e-01 5.10064960e-01 9.22469854e-01
9.00179684e-01 3.39100838e-01 1.79028615e-01 6.99007094e-01
6.13841772e-01 -1.19468141e+00 3.25634211e-01 6.36380136e-01
1.02857745e+00 -7.56788194e-01 -2.86985368e-01 3.40268642e-01
-5.03748298e-01 1.08003724e+00 5.12958109e-01 -9.26402509e-01
7.97464788e-01 4.42672372e-01 -2.91111529e-01 -3.38473946e-01
-1.55551744e+00 -1.92140982e-01 -2.60202706e-01 8.08878362e-01
1.36482745e-01 1.29472956e-01 -6.53920293e-01 1.24854468e-01
-3.51688504e-01 -4.15063471e-01 9.14024949e-01 8.82078946e-01
-1.14328384e+00 -7.62654722e-01 -1.13037065e-01 -1.55841455e-01
-3.34523767e-01 4.57814075e-02 -6.74582839e-01 9.54565227e-01
5.32736368e-02 7.77013719e-01 2.42803410e-01 -2.40117639e-01
9.16673318e-02 2.17079371e-01 7.23127127e-01 -1.89431429e-01
-6.31054819e-01 -3.20967644e-01 -4.14136410e-01 -1.75021321e-01
-5.39702296e-01 -1.01075995e+00 -1.34936893e+00 -4.47891623e-01
4.31293584e-02 5.56714296e-01 4.33079302e-01 9.16024625e-01
2.27944821e-01 5.74320197e-01 3.96166384e-01 -8.11523318e-01
-4.74810034e-01 -6.55942261e-01 -6.42539918e-01 -2.83912290e-02
6.21231437e-01 -8.15350115e-01 -4.28525358e-01 1.10141344e-01] | [5.563549041748047, 4.131024360656738] |
41545c02-57d0-467c-9135-3b1aa81bba62 | data-free-quantization-via-mixed-precision | 2307.00498 | null | https://arxiv.org/abs/2307.00498v1 | https://arxiv.org/pdf/2307.00498v1.pdf | Data-Free Quantization via Mixed-Precision Compensation without Fine-Tuning | Neural network quantization is a very promising solution in the field of model compression, but its resulting accuracy highly depends on a training/fine-tuning process and requires the original data. This not only brings heavy computation and time costs but also is not conducive to privacy and sensitive information protection. Therefore, a few recent works are starting to focus on data-free quantization. However, data-free quantization does not perform well while dealing with ultra-low precision quantization. Although researchers utilize generative methods of synthetic data to address this problem partially, data synthesis needs to take a lot of computation and time. In this paper, we propose a data-free mixed-precision compensation (DF-MPC) method to recover the performance of an ultra-low precision quantized model without any data and fine-tuning process. By assuming the quantized error caused by a low-precision quantized layer can be restored via the reconstruction of a high-precision quantized layer, we mathematically formulate the reconstruction loss between the pre-trained full-precision model and its layer-wise mixed-precision quantized model. Based on our formulation, we theoretically deduce the closed-form solution by minimizing the reconstruction loss of the feature maps. Since DF-MPC does not require any original/synthetic data, it is a more efficient method to approximate the full-precision model. Experimentally, our DF-MPC is able to achieve higher accuracy for an ultra-low precision quantized model compared to the recent methods without any data and fine-tuning process. | ['Yong liu', 'Guanzhong Tian', 'Mengmeng Wang', 'Tianxin Huang', 'Shipeng Bai', 'Jun Chen'] | 2023-07-02 | null | null | null | null | ['data-free-quantization', 'quantization', 'data-free-quantization', 'model-compression'] | ['computer-vision', 'methodology', 'methodology', 'methodology'] | [ 3.44766766e-01 -1.72436163e-01 -3.18980545e-01 -3.50892127e-01
-9.20861959e-01 -2.20516007e-02 4.66626585e-01 3.66085500e-01
-4.32539701e-01 7.45766342e-01 -1.12246215e-01 -1.28797710e-01
6.86762109e-02 -1.20452535e+00 -1.14738882e+00 -6.86369896e-01
2.57808000e-01 7.91897625e-02 2.80242950e-01 -1.48547247e-01
2.68767655e-01 3.16138089e-01 -1.75801933e+00 2.01659888e-01
1.00649953e+00 1.39745021e+00 2.46639043e-01 2.80343711e-01
5.94168156e-02 3.53897303e-01 -7.13413835e-01 -5.18544257e-01
5.80239296e-01 -3.50606024e-01 -3.79146069e-01 -2.20562771e-01
3.88202161e-01 -6.55714512e-01 -3.79035741e-01 1.57175314e+00
5.03524899e-01 -1.82280600e-01 5.80943823e-01 -1.10918140e+00
-6.64650261e-01 3.76016706e-01 -3.44732463e-01 -3.64316583e-01
-3.65876332e-02 -1.32592455e-01 6.22228146e-01 -6.28622949e-01
4.61056709e-01 9.63577807e-01 6.68007791e-01 4.15015876e-01
-1.30481660e+00 -9.24552381e-01 -5.46136558e-01 1.11111350e-01
-1.93770397e+00 -5.35450459e-01 7.97076404e-01 -1.91679448e-01
6.44850552e-01 3.83518398e-01 5.67393184e-01 6.03809655e-01
4.27613050e-01 4.28413332e-01 1.04791796e+00 -5.15331268e-01
5.95322132e-01 4.11360830e-01 -1.05302118e-01 5.25272131e-01
3.97430897e-01 2.18953788e-01 -3.98383081e-01 -8.98902267e-02
8.30356836e-01 1.00295536e-01 -6.43678784e-01 -3.00595462e-01
-6.97405219e-01 6.23218536e-01 6.03173912e-01 4.06777203e-01
-2.58665413e-01 2.21447989e-01 3.99654299e-01 4.18338984e-01
3.26957434e-01 4.44383800e-01 -2.28079319e-01 3.92605811e-02
-1.55320263e+00 2.45438486e-01 5.56373060e-01 1.00015092e+00
1.01231313e+00 1.69370517e-01 -2.19123274e-01 5.95349312e-01
2.89454728e-01 3.50897640e-01 6.61560655e-01 -9.14961159e-01
5.84366858e-01 5.57429671e-01 1.11283578e-01 -1.21444297e+00
2.84488983e-02 -3.15066904e-01 -1.36738694e+00 3.78977895e-01
1.62613243e-01 2.46960536e-01 -8.10199201e-01 1.48138535e+00
1.79442227e-01 1.71353027e-01 1.87598273e-01 8.12502146e-01
3.74586821e-01 9.83362198e-01 -3.00083369e-01 -5.19020438e-01
1.07624900e+00 -7.17692554e-01 -9.81661499e-01 3.76541406e-01
5.50073922e-01 -6.42893136e-01 1.17051363e+00 4.33854878e-01
-1.05955291e+00 -6.10799253e-01 -1.68891430e+00 -4.05947834e-01
-3.04687768e-01 1.56916291e-01 1.67092159e-01 1.01381648e+00
-9.50460672e-01 9.17167008e-01 -6.63519025e-01 2.54009277e-01
4.40706551e-01 4.65237200e-01 -3.25235814e-01 -1.54207826e-01
-1.45151997e+00 7.31543124e-01 5.32434583e-01 -4.61425781e-02
-5.34067988e-01 -7.81944156e-01 -6.77022517e-01 2.65828371e-01
1.75465450e-01 -4.64627922e-01 8.88230681e-01 -4.94966894e-01
-1.55823100e+00 2.69537151e-01 -1.14082173e-01 -7.58875191e-01
5.98174155e-01 -5.12369759e-02 -3.53084683e-01 1.21963784e-01
-2.16578081e-01 4.37789291e-01 1.18639314e+00 -1.19331217e+00
-3.69070172e-01 -3.48143727e-01 -3.06158751e-01 -1.91608384e-01
-7.78792918e-01 -5.15966952e-01 -3.72272074e-01 -7.55579531e-01
4.84448597e-02 -4.82300818e-01 -2.71035396e-02 4.37047452e-01
-3.12925041e-01 2.92601496e-01 1.03617287e+00 -6.26657486e-01
1.47235942e+00 -2.42305636e+00 -4.87272441e-01 2.31318146e-01
3.15786451e-02 7.96205342e-01 2.05097318e-01 2.27399871e-01
2.61449486e-01 3.84392560e-01 -4.22460616e-01 -5.92564940e-01
6.65722936e-02 3.35708052e-01 -4.70782906e-01 4.73530710e-01
8.30106065e-02 6.14932060e-01 -6.66804790e-01 -4.58647877e-01
4.17054236e-01 8.89162898e-01 -8.12006474e-01 4.54907306e-02
-1.98541239e-01 2.06434608e-01 -2.93457448e-01 6.06953740e-01
1.02848113e+00 -4.05887477e-02 -1.23165078e-01 -4.28247422e-01
-1.31872594e-01 4.51191813e-02 -1.14139163e+00 1.59516490e+00
-4.61279750e-01 2.89649755e-01 2.56761789e-01 -8.36543739e-01
1.29806769e+00 3.72873247e-01 1.14316285e-01 -1.01627839e+00
3.57305944e-01 6.96815312e-01 -5.96272469e-01 -2.23984271e-01
8.57419848e-01 -2.13287070e-01 4.18940112e-02 -5.91833256e-02
-1.37172773e-01 -4.08377796e-01 -1.90096751e-01 -1.74722508e-01
7.71924913e-01 -2.04580933e-01 2.34764561e-01 -1.57572374e-01
6.10463977e-01 -3.93880725e-01 7.00661004e-01 3.90752017e-01
-4.49606702e-02 7.70262957e-01 2.72917330e-01 -8.35368484e-02
-1.38283539e+00 -6.96963310e-01 -4.31160539e-01 4.00616020e-01
3.03583592e-01 -6.91320300e-01 -9.47167933e-01 -2.05118865e-01
-1.39180541e-01 7.90915132e-01 -2.18896091e-01 -4.84754622e-01
-2.15464696e-01 -5.12267470e-01 7.68658042e-01 1.81689307e-01
1.08888102e+00 -4.52457964e-01 -4.52305019e-01 8.87859687e-02
-1.84433125e-02 -9.13276851e-01 -4.88768905e-01 2.92042732e-01
-8.59620392e-01 -5.50545990e-01 -7.10565686e-01 -6.14026606e-01
5.78383803e-01 -5.98947667e-02 5.57372928e-01 2.36746281e-01
2.61685736e-02 -4.24972147e-01 -2.67743856e-01 -2.79494286e-01
-5.03615797e-01 5.15097603e-02 6.43552989e-02 1.90826982e-01
1.14495963e-01 -6.85207844e-01 -4.51316267e-01 1.53891847e-01
-1.19738925e+00 -5.19706048e-02 5.26096046e-01 9.85597253e-01
1.09521377e+00 5.50851166e-01 5.32791555e-01 -5.75645089e-01
5.65489292e-01 -4.75359969e-02 -9.61410940e-01 1.71154797e-01
-1.04588068e+00 3.43995869e-01 1.09508228e+00 -1.72710195e-01
-5.70118427e-01 2.92722106e-01 -3.31929356e-01 -9.61578250e-01
2.20754117e-01 9.26205143e-02 -5.30870557e-01 -3.56293261e-01
6.99645400e-01 3.66031051e-01 -6.86364546e-02 -5.96577585e-01
2.07737833e-01 9.08759415e-01 7.25796700e-01 -3.70792121e-01
7.57125199e-01 2.83346027e-01 2.60188252e-01 -7.91568160e-01
-2.87361026e-01 -8.20580944e-02 -4.30742353e-01 2.46305525e-01
7.32523441e-01 -7.32035875e-01 -6.27558351e-01 4.63487059e-01
-1.05726767e+00 -6.23888224e-02 -5.48455298e-01 4.06898797e-01
-5.56200564e-01 4.84092444e-01 -3.54484826e-01 -7.89097726e-01
-5.53390682e-01 -1.45990586e+00 1.06605256e+00 -5.37310727e-02
1.99197963e-01 -5.14770091e-01 -3.52371007e-01 1.06944881e-01
5.33336461e-01 2.96867609e-01 7.95587718e-01 -2.17148572e-01
-7.86237955e-01 -4.49651927e-01 -2.45432854e-01 8.51710260e-01
5.69929779e-02 -1.97226584e-01 -1.02859151e+00 -3.11480850e-01
4.96359557e-01 -3.16965461e-01 8.02947521e-01 8.76856074e-02
1.45542037e+00 -7.47378230e-01 -7.67340139e-03 1.02597523e+00
1.56546926e+00 1.83699988e-02 9.36690569e-01 9.05738771e-02
5.34683764e-01 1.59898162e-01 6.96113646e-01 5.33669770e-01
1.25633553e-01 5.94065428e-01 5.84459424e-01 2.71380842e-01
-3.01963780e-02 -5.69903970e-01 1.59703135e-01 6.91630304e-01
1.60540491e-01 -1.90973088e-01 -5.18464863e-01 3.57535869e-01
-1.51410449e+00 -8.13504100e-01 -1.66305408e-01 2.78782320e+00
1.08421946e+00 1.50753886e-01 -3.94044757e-01 8.77393603e-01
6.56129777e-01 1.02592491e-01 -3.74607325e-01 -5.41517437e-01
1.29612610e-01 3.27651739e-01 7.99393117e-01 4.45737600e-01
-1.11686921e+00 4.54918683e-01 5.91247702e+00 1.33968532e+00
-1.57980084e+00 1.05344661e-01 6.60716951e-01 -5.46007864e-02
-4.72534597e-01 -2.80775905e-01 -9.04265344e-01 8.09194088e-01
1.21512270e+00 -3.09142113e-01 4.28986877e-01 1.07077205e+00
-4.52464186e-02 -8.21913034e-02 -1.09749675e+00 1.12800813e+00
-1.56679690e-01 -1.46990681e+00 2.04418227e-01 3.21582943e-01
6.62733257e-01 -4.66192573e-01 2.39610448e-01 5.20946644e-02
-5.21908760e-01 -9.82663751e-01 6.94444656e-01 4.20239449e-01
1.31397963e+00 -9.95336056e-01 7.92514443e-01 7.21180499e-01
-1.06069422e+00 -1.11217014e-02 -8.31253827e-01 -2.94401441e-02
-3.96620743e-02 1.00334013e+00 -7.01688826e-01 5.70776165e-01
6.77182496e-01 2.07038835e-01 -3.81466478e-01 9.08337474e-01
7.50685716e-03 3.46165210e-01 -4.57099855e-01 -5.18546440e-02
-3.15811820e-02 -1.62525088e-01 2.63790518e-01 9.12131488e-01
7.30009437e-01 -2.82113831e-02 -1.50414750e-01 9.16461527e-01
-2.76878417e-01 7.35632330e-02 -5.63422561e-01 -1.10223398e-01
6.95355892e-01 8.05169880e-01 -2.04018027e-01 -3.43446881e-01
-1.10792547e-01 1.07611585e+00 -8.95781629e-03 -4.89934608e-02
-6.51502371e-01 -5.79750001e-01 4.72413331e-01 5.56399226e-01
2.89106637e-01 -1.38040170e-01 -5.80824614e-01 -9.80461597e-01
2.97360033e-01 -5.98899662e-01 -8.71022567e-02 -6.29415274e-01
-9.32603776e-01 5.78110516e-01 -1.99049652e-01 -1.57978415e+00
-3.70975405e-01 -1.66023940e-01 -2.04143417e-03 1.09056330e+00
-1.60069835e+00 -9.53545094e-01 -1.21320903e-01 7.89484501e-01
4.62740064e-02 1.50986649e-02 1.14511108e+00 6.27882302e-01
-1.88789770e-01 1.18846309e+00 5.28989971e-01 -2.04603583e-01
6.33688927e-01 -7.79861331e-01 4.74046133e-02 7.85947025e-01
-1.77566752e-01 7.01203465e-01 6.62835062e-01 -5.36534786e-01
-1.46468258e+00 -1.44676936e+00 1.00562131e+00 1.30132839e-01
7.21626133e-02 -3.76071036e-01 -1.24236286e+00 2.29586601e-01
-1.62427232e-01 2.76464283e-01 3.31169724e-01 -7.01967120e-01
-2.92593569e-01 -5.36241293e-01 -1.85890102e+00 2.42301971e-01
4.07003403e-01 -7.02210307e-01 -3.32626998e-01 5.96412569e-02
9.72725868e-01 -4.22192067e-01 -1.09169626e+00 4.88104641e-01
2.92822778e-01 -1.06413603e+00 9.45286095e-01 4.53547955e-01
4.78979945e-01 -4.15111661e-01 -5.86059451e-01 -8.79557073e-01
5.59144374e-03 -5.70062816e-01 -2.02320322e-01 1.33946359e+00
1.79820865e-01 -6.62532449e-01 7.94276714e-01 6.91977978e-01
-1.99650332e-01 -9.31703091e-01 -1.18505228e+00 -9.00814056e-01
4.53924350e-02 -3.56780052e-01 8.28413069e-01 8.36301327e-01
-2.09102347e-01 -1.51795387e-01 -7.05573916e-01 2.89774686e-01
6.82971656e-01 -1.81082651e-01 6.86613023e-01 -1.01871121e+00
-2.45511413e-01 -1.23199485e-01 -6.34616077e-01 -1.04987979e+00
-2.57741302e-01 -8.39131713e-01 2.04877853e-02 -1.19152999e+00
-4.28234458e-01 -5.40217042e-01 -1.06838837e-01 3.15455019e-01
3.30839634e-01 5.03352702e-01 1.35727704e-01 3.14152896e-01
-6.04845732e-02 8.37591469e-01 1.42581689e+00 -1.74291924e-01
-2.20363334e-01 -5.85438386e-02 -5.30152202e-01 5.41343212e-01
7.63725817e-01 -4.94575232e-01 -5.91473281e-01 -2.31638089e-01
-3.48254517e-02 1.61325052e-01 2.81656086e-01 -1.42823672e+00
4.68084663e-01 3.19765843e-02 3.43394220e-01 -6.52793288e-01
7.04352260e-01 -1.18550611e+00 4.09325600e-01 5.39814055e-01
-1.76051959e-01 -4.47714746e-01 -4.35266234e-02 5.18977046e-01
-4.98102695e-01 -5.01396358e-01 1.05290294e+00 -1.32665280e-02
-4.40622717e-01 2.85518527e-01 -9.44163427e-02 -4.62018728e-01
9.26245153e-01 -3.13497156e-01 -1.73435524e-01 -2.28981167e-01
-3.74417424e-01 -1.19036131e-01 8.03417921e-01 9.65813398e-02
7.22670197e-01 -1.48516190e+00 -2.52890199e-01 7.98028529e-01
-1.48246750e-01 2.99083114e-01 1.58443972e-01 3.25756699e-01
-6.60019159e-01 4.06375021e-01 -1.10849857e-01 -5.04877210e-01
-9.78094935e-01 9.01617348e-01 1.85250327e-01 -1.93070903e-01
-5.20690322e-01 6.70140088e-01 -1.52368009e-01 -2.03361675e-01
3.96847308e-01 -6.23277307e-01 2.69942015e-01 -2.16059685e-01
8.41131508e-01 4.64534312e-01 3.37707281e-01 -5.48965096e-01
-8.00960064e-02 6.37631893e-01 1.80680186e-01 -7.76582658e-02
1.09231746e+00 -2.60963757e-02 -8.77172425e-02 3.01469266e-01
1.73700631e+00 -1.87605709e-01 -1.03635406e+00 -1.09610550e-01
-5.67367971e-01 -8.16693187e-01 2.88690388e-01 -3.44086766e-01
-1.18536556e+00 1.19393349e+00 7.33866215e-01 1.30227745e-01
1.50351095e+00 -5.78962922e-01 1.00682938e+00 3.26567382e-01
8.27059627e-01 -1.12010300e+00 -1.84135482e-01 2.22570121e-01
9.28271592e-01 -8.16599071e-01 1.50325060e-01 -5.33604741e-01
-1.57521129e-01 8.63482714e-01 3.81422967e-01 -2.27247864e-01
9.11288857e-01 4.93364781e-01 -3.03501368e-01 3.73367459e-01
-4.19919163e-01 4.63440567e-01 1.34718627e-01 6.95452392e-01
6.29047304e-02 2.85632908e-02 -4.68789876e-01 6.63081288e-01
-5.68899751e-01 4.42341238e-01 4.04404104e-01 8.54605019e-01
-4.26104486e-01 -1.21353543e+00 -5.42856216e-01 5.81021428e-01
-4.45825368e-01 -4.06898335e-02 -2.79323058e-03 3.43037128e-01
4.85025495e-01 7.87119210e-01 9.78619605e-02 -8.00659716e-01
2.92522162e-01 2.86256764e-02 2.15264291e-01 -2.51021117e-01
-5.53679824e-01 5.50311804e-02 -3.99029553e-01 -6.85860097e-01
-1.04453042e-01 -2.36296073e-01 -1.32450140e+00 -7.25832939e-01
-4.65879500e-01 5.68484254e-02 8.96401227e-01 4.35036063e-01
5.05379260e-01 4.65474993e-01 6.07829332e-01 -8.19412768e-01
-8.07138205e-01 -5.77115297e-01 -7.84803689e-01 1.34888083e-01
5.22716880e-01 -3.50249529e-01 -2.92608261e-01 -1.96325988e-01] | [8.729589462280273, 2.9730257987976074] |
5974d1ba-4698-44a5-8bbc-f6a7d21752d7 | an-examination-of-wearable-sensors-and-video | 2307.04516 | null | https://arxiv.org/abs/2307.04516v1 | https://arxiv.org/pdf/2307.04516v1.pdf | An Examination of Wearable Sensors and Video Data Capture for Human Exercise Classification | Wearable sensors such as Inertial Measurement Units (IMUs) are often used to assess the performance of human exercise. Common approaches use handcrafted features based on domain expertise or automatically extracted features using time series analysis. Multiple sensors are required to achieve high classification accuracy, which is not very practical. These sensors require calibration and synchronization and may lead to discomfort over longer time periods. Recent work utilizing computer vision techniques has shown similar performance using video, without the need for manual feature engineering, and avoiding some pitfalls such as sensor calibration and placement on the body. In this paper, we compare the performance of IMUs to a video-based approach for human exercise classification on two real-world datasets consisting of Military Press and Rowing exercises. We compare the performance using a single camera that captures video in the frontal view versus using 5 IMUs placed on different parts of the body. We observe that an approach based on a single camera can outperform a single IMU by 10 percentage points on average. Additionally, a minimum of 3 IMUs are required to outperform a single camera. We observe that working with the raw data using multivariate time series classifiers outperforms traditional approaches based on handcrafted or automatically extracted features. Finally, we show that an ensemble model combining the data from a single camera with a single IMU outperforms either data modality. Our work opens up new and more realistic avenues for this application, where a video captured using a readily available smartphone camera, combined with a single sensor, can be used for effective human exercise classification. | ['Georgiana Ifrim', 'Brian Caulfield', "Martin O'Reilly", 'Darragh Whelan', 'Thach Le Nguyen', 'Timilehin B. Aderinola', 'Antonio Bevilacqua', 'Ashish Singh'] | 2023-07-10 | null | null | null | null | ['feature-engineering', 'time-series'] | ['methodology', 'time-series'] | [ 4.12077725e-01 -5.43721735e-01 -2.35868514e-01 -9.76492371e-03
-6.57914221e-01 -6.59734964e-01 6.04674257e-02 3.74970585e-01
-7.30587423e-01 4.70171362e-01 4.19951342e-02 -6.48874464e-03
-3.26162241e-02 -5.61229885e-01 -6.57029867e-01 -4.36567813e-01
8.36055726e-02 -6.53744042e-02 2.13423833e-01 -6.20589033e-02
9.77290645e-02 3.52525771e-01 -1.66992903e+00 -2.17143789e-01
3.77478093e-01 1.24365926e+00 1.57616764e-01 9.78545368e-01
4.06226397e-01 2.30364308e-01 -8.72187972e-01 3.34813669e-02
5.36005259e-01 -3.87866884e-01 -7.22196028e-02 4.31141913e-01
6.89535677e-01 -3.76374602e-01 3.56305689e-02 5.04919052e-01
8.50465655e-01 3.59259605e-01 8.01098570e-02 -1.04145777e+00
3.08240712e-01 -1.08105905e-01 -3.51464540e-01 3.25098515e-01
1.09366584e+00 2.92133600e-01 5.47056913e-01 -2.86969572e-01
4.59488004e-01 4.68335509e-01 1.22276568e+00 9.46921781e-02
-1.36549926e+00 -3.92452151e-01 -3.07586461e-01 2.24905282e-01
-1.39897275e+00 -2.64532685e-01 8.96407008e-01 -6.24058127e-01
7.92395949e-01 4.48294848e-01 1.12002277e+00 1.04853439e+00
4.04387891e-01 3.65130693e-01 1.22808111e+00 -4.24742460e-01
3.16183239e-01 3.38407903e-04 8.83683562e-02 4.82429951e-01
4.75062639e-01 -1.77203435e-02 -6.08107746e-01 -1.44843400e-01
7.19061494e-01 6.30258560e-01 -3.82879615e-01 -3.58783126e-01
-1.32604861e+00 7.14867115e-01 1.16309874e-01 4.45446938e-01
-6.55928254e-01 9.55318958e-02 4.38918710e-01 3.54854763e-01
3.22035253e-01 6.48566663e-01 -5.04227877e-01 -8.93782079e-01
-1.22807956e+00 8.00838619e-02 6.63377464e-01 6.32522404e-01
5.41992843e-01 -1.11492753e-01 2.19005510e-01 5.47305644e-01
-1.23890921e-01 4.46311414e-01 7.54843116e-01 -1.02475715e+00
2.54776746e-01 8.10181022e-01 2.25158095e-01 -1.25881624e+00
-7.14258850e-01 -2.41593450e-01 -2.81261325e-01 3.10310274e-01
6.64862692e-01 -5.47050655e-01 -4.24137384e-01 1.25635874e+00
5.22221446e-01 4.65298623e-01 -2.67643094e-01 1.00295746e+00
3.30050111e-01 1.09917097e-01 -2.00705513e-01 -2.56404996e-01
1.51757371e+00 -4.94431227e-01 -5.83179176e-01 -1.51558593e-01
8.14890921e-01 -8.61024559e-01 1.04758942e+00 7.38950253e-01
-8.51616144e-01 -7.08734334e-01 -1.53002429e+00 3.00516635e-01
-3.57926309e-01 3.39065701e-01 2.87589103e-01 1.10589969e+00
-6.16401434e-01 1.03206682e+00 -1.14518678e+00 -6.64219797e-01
-1.53745681e-01 5.41660666e-01 -7.26291299e-01 8.86058956e-02
-8.38973880e-01 8.22121739e-01 1.01783164e-01 -1.32609578e-02
-2.34577954e-01 -6.35742307e-01 -1.12992036e+00 -3.46002966e-01
4.89390045e-01 -6.32381439e-01 1.08110118e+00 -9.69547272e-01
-1.71722686e+00 5.65151453e-01 9.14532542e-02 -5.93174934e-01
5.99814594e-01 -8.13082159e-01 -3.52514684e-01 5.28067827e-01
-1.22629412e-01 -5.47296554e-02 8.01623940e-01 -5.00247657e-01
-3.24541897e-01 -4.24901575e-01 2.02923015e-01 1.63253590e-01
-5.96608281e-01 -1.22569866e-01 -9.35019627e-02 -4.85716164e-01
-9.95270237e-02 -1.54953158e+00 -3.07709631e-02 -1.43551454e-01
-5.94249442e-02 2.55860031e-01 7.34093249e-01 -8.13836277e-01
1.36118805e+00 -2.11988926e+00 1.08113259e-01 2.20718488e-01
3.40120606e-02 3.01095605e-01 5.44530928e-01 3.08692485e-01
-1.26620919e-01 -1.58334091e-01 2.09869787e-01 -1.17577948e-02
-4.26628470e-01 2.25571528e-01 3.91321361e-01 7.32122481e-01
-4.10355836e-01 4.40109640e-01 -8.34076941e-01 -4.12405312e-01
8.04523468e-01 4.25378382e-01 -3.84747177e-01 1.32525504e-01
4.17374909e-01 4.69461381e-01 -4.06981632e-02 6.92723811e-01
1.13831699e-01 -2.30705109e-03 3.05336744e-01 -3.43726456e-01
-1.26322238e-02 -9.30673853e-02 -1.67353570e+00 1.85644078e+00
-6.00149632e-01 7.59235263e-01 -1.58767775e-01 -1.10681605e+00
7.88471997e-01 5.93451440e-01 9.79125857e-01 -3.25039595e-01
1.90280169e-01 1.56450868e-02 -1.89574301e-01 -7.52221823e-01
3.34772348e-01 -2.87779383e-02 -2.24838600e-01 3.64355385e-01
6.15260005e-02 -8.38718377e-03 2.66157150e-01 -2.35441417e-01
1.40812814e+00 2.96701938e-01 6.48533762e-01 -5.94586842e-02
3.33173752e-01 1.23797255e-02 4.84643579e-01 4.57255781e-01
-2.55361825e-01 7.84421563e-01 -2.02622171e-02 -5.46195030e-01
-8.02732050e-01 -7.79680729e-01 1.53043956e-01 7.96200514e-01
2.02674251e-02 -8.90057504e-01 -7.52670109e-01 -3.65773976e-01
7.99623206e-02 6.29593879e-02 -4.96087015e-01 -1.19378708e-01
-5.68295479e-01 -4.17272270e-01 4.80961561e-01 9.20239091e-01
3.54389995e-01 -3.86976719e-01 -1.41775274e+00 4.10461128e-01
-1.34281799e-01 -1.29224527e+00 -4.13243473e-01 -1.15450464e-01
-1.18214202e+00 -1.24765635e+00 -8.43199492e-01 -1.42674074e-02
4.62674499e-01 4.37036455e-01 8.77715170e-01 1.03133880e-01
-3.56418639e-01 1.15396941e+00 -5.03449559e-01 -5.20649612e-01
3.02152127e-01 -7.03029260e-02 4.79140043e-01 2.76698470e-02
3.29727709e-01 -5.69733500e-01 -8.34613621e-01 4.39522624e-01
-5.61648786e-01 -1.95849121e-01 2.81773686e-01 6.57133102e-01
5.15784025e-01 -1.22515559e-01 2.49218903e-02 -3.37520510e-01
4.50951368e-01 -3.58798116e-01 -2.45888427e-01 -1.54359534e-01
-5.01094878e-01 -3.66144270e-01 4.36674416e-01 -7.89699852e-01
-4.56961900e-01 5.32657325e-01 1.90329254e-01 -5.82800388e-01
-2.34090120e-01 5.06958008e-01 2.71096915e-01 -2.16545552e-01
9.36538994e-01 -1.72355086e-01 3.90616119e-01 -5.20871580e-01
-2.21776798e-01 8.19001377e-01 4.93655026e-01 -4.99849647e-01
4.33952391e-01 4.39963549e-01 2.28480548e-01 -1.16865849e+00
-5.51201642e-01 -8.08382154e-01 -8.76580358e-01 -7.67308593e-01
8.14936042e-01 -1.08503842e+00 -8.74745309e-01 3.05679560e-01
-5.55017114e-01 -9.01178643e-02 -1.89920411e-01 1.03404903e+00
-6.30188882e-01 4.20914382e-01 -3.92545611e-01 -9.04439211e-01
-2.11549342e-01 -8.86279166e-01 1.37124181e+00 5.37313700e-01
-8.03699493e-01 -9.46339905e-01 2.10268553e-02 7.69650638e-01
2.11297154e-01 8.27500880e-01 -1.73465282e-01 -3.94083411e-01
1.31280437e-01 -8.85807812e-01 6.21318460e-01 4.21941429e-01
2.76216745e-01 -1.25892416e-01 -6.70450926e-01 -4.21907872e-01
1.95776999e-01 -6.37940690e-02 1.82666004e-01 3.34287107e-01
8.35246682e-01 1.25016555e-01 -2.00609148e-01 2.59904087e-01
1.40774810e+00 1.42674237e-01 5.46252668e-01 5.05204320e-01
6.18607461e-01 2.16145828e-01 8.07695270e-01 5.38777173e-01
1.58895105e-01 9.76958036e-01 1.03699178e-01 -8.17439705e-02
2.35429898e-01 4.15803641e-02 7.28279650e-01 5.22150457e-01
-8.43604088e-01 2.78374255e-01 -8.01005423e-01 3.14557135e-01
-1.69034493e+00 -1.01147521e+00 -1.18019342e-01 2.73741579e+00
4.10931259e-01 1.60186470e-01 7.06686258e-01 7.43272901e-01
6.55700862e-01 -2.65049487e-01 -2.42750049e-01 -3.06932837e-01
3.93493146e-01 2.91083276e-01 7.62127340e-01 1.84153885e-01
-1.05417037e+00 -7.12045329e-03 6.55510855e+00 2.16155633e-01
-1.38252902e+00 1.24022044e-01 5.97680770e-02 -6.24741375e-01
5.38400412e-01 -1.03885103e-02 -5.04330158e-01 7.97538102e-01
1.26388001e+00 -1.73592418e-02 1.08569853e-01 9.81003106e-01
4.01903510e-01 -7.87235260e-01 -1.02360439e+00 1.34891272e+00
2.87297457e-01 -8.72481287e-01 -7.40040600e-01 3.41720760e-01
3.67960751e-01 -2.01711297e-01 -4.57601756e-01 -1.51724694e-02
-5.71994781e-01 -5.70923030e-01 3.53475839e-01 7.47358143e-01
4.91433948e-01 -4.77945268e-01 8.29109430e-01 3.17193538e-01
-1.29901886e+00 -1.28668472e-02 7.87646696e-02 -7.58567274e-01
4.21947390e-02 6.54472411e-01 -6.78070426e-01 6.21244073e-01
7.26368248e-01 6.99021399e-01 -7.17979968e-01 1.01633108e+00
1.42593130e-01 6.19381845e-01 -7.79388726e-01 -6.29371703e-02
-2.22781315e-01 -1.77962705e-01 5.70734322e-01 9.10617232e-01
7.76826203e-01 -6.45396113e-02 5.09681165e-01 -1.08298279e-01
4.89642084e-01 2.17102274e-01 -9.01843488e-01 2.07896233e-02
1.09346829e-01 1.34820247e+00 -6.23524666e-01 -4.59555387e-01
-6.56625509e-01 8.87402296e-01 -2.44889081e-01 -8.10938627e-02
-7.91082919e-01 -3.96540582e-01 6.77631259e-01 6.15497053e-01
8.58193338e-02 -7.25752175e-01 -1.87610343e-01 -1.24017274e+00
2.54551649e-01 -7.72801936e-01 5.52855611e-01 -7.40678012e-01
-6.93033040e-01 1.50017381e-01 9.27137956e-02 -1.95686924e+00
-6.69796407e-01 -6.72591269e-01 -5.64566493e-01 4.25706744e-01
-6.71795726e-01 -6.25600636e-01 -7.20875740e-01 6.90034032e-01
5.04790425e-01 1.65013403e-01 8.91118348e-01 4.66295928e-01
-6.51314020e-01 2.91197687e-01 -2.69427091e-01 1.01103686e-01
7.71941066e-01 -1.17153907e+00 -3.13891977e-01 7.29649365e-01
1.30268842e-01 5.97827196e-01 8.42171252e-01 -5.94978094e-01
-1.77479303e+00 -4.80955482e-01 4.75453049e-01 -7.38029003e-01
3.80069971e-01 -4.01149020e-02 -3.05763185e-01 6.90844953e-01
-1.31871104e-02 1.26200214e-01 1.30724585e+00 2.26915091e-01
3.72748584e-01 -3.73340756e-01 -1.06805754e+00 3.13565016e-01
7.51878977e-01 -4.83265311e-01 -6.39805198e-01 3.17756265e-01
-5.10004200e-02 -7.13189065e-01 -1.41085565e+00 3.53353679e-01
1.09887052e+00 -8.53112698e-01 1.02300227e+00 -2.31728047e-01
9.76883098e-02 -3.82684827e-01 -5.87495342e-02 -1.33934009e+00
5.17250150e-02 -6.57103121e-01 -3.68472822e-02 8.82458270e-01
-4.96649854e-02 -5.51451921e-01 7.86709607e-01 8.27123702e-01
1.66943505e-01 -5.47709525e-01 -8.72424185e-01 -9.84005213e-01
-6.39289618e-01 -6.37618244e-01 1.69788003e-01 7.06814647e-01
3.53031218e-01 2.73871958e-01 -5.94660223e-01 -1.76373020e-01
3.64391953e-01 3.21667343e-02 1.24447191e+00 -1.44471550e+00
-3.61388475e-01 1.05375923e-01 -1.21043789e+00 -5.78995466e-01
-4.04670596e-01 -3.38266373e-01 -3.84369195e-01 -1.08322644e+00
-1.61879689e-01 2.27213591e-01 -1.22924879e-01 2.02285305e-01
-6.04379512e-02 6.71452999e-01 3.34582508e-01 -7.15355128e-02
-4.30428296e-01 -1.62324324e-01 8.36191833e-01 1.06894448e-01
-3.38040650e-01 1.74614519e-01 -4.75732982e-01 1.04627037e+00
7.45664537e-01 -3.76659006e-01 -2.58953214e-01 -2.70469021e-02
3.12167853e-01 2.04319805e-01 4.28911805e-01 -1.69169569e+00
2.65418947e-01 7.58119747e-02 7.84916818e-01 -1.66458804e-02
6.25650227e-01 -1.13535845e+00 5.04237771e-01 6.26202643e-01
2.26671785e-01 4.17148143e-01 1.09962240e-01 5.16663671e-01
-1.85222730e-01 -1.49988338e-01 3.89180899e-01 -3.25929046e-01
-6.62927091e-01 -3.06256950e-01 -5.22401631e-01 -3.67648393e-01
1.39725983e+00 -7.70366609e-01 2.34181225e-01 -5.10583341e-01
-1.03956258e+00 -5.74424714e-02 5.70311606e-01 3.70635450e-01
2.85934299e-01 -1.42016602e+00 -1.82919547e-01 2.40238756e-01
9.14753675e-02 -5.91916442e-01 1.98154896e-01 1.22013152e+00
-5.43673575e-01 2.22544551e-01 -4.67036515e-01 -1.03966534e+00
-1.72134781e+00 3.17991585e-01 2.80023396e-01 1.91420410e-02
-6.45381927e-01 1.08066294e-02 -7.44451702e-01 -4.81088422e-02
-1.08599506e-01 -4.77688849e-01 -2.11310342e-01 4.33597744e-01
5.98252118e-01 6.87116027e-01 4.12283510e-01 -6.02073729e-01
-4.37777191e-01 9.49910700e-01 4.68172371e-01 -1.41205430e-01
1.02522004e+00 -1.99414641e-01 6.05759025e-01 8.32445920e-01
1.03604770e+00 2.01323867e-01 -1.09120250e+00 -4.79319654e-02
-1.54523224e-01 -6.87060177e-01 2.51615755e-02 -3.00942153e-01
-8.20378006e-01 6.62599802e-01 1.00837803e+00 2.51519352e-01
1.28143835e+00 -6.28815234e-01 8.85507405e-01 2.85920292e-01
5.90516806e-01 -1.46392167e+00 1.14837319e-01 -4.59865890e-02
5.24884582e-01 -1.21018493e+00 4.35977221e-01 -2.33433187e-01
-6.46061599e-01 1.13469899e+00 4.00656372e-01 -3.47515017e-01
7.01796770e-01 1.93865463e-01 2.07933001e-02 3.36429700e-02
-2.55596697e-01 -4.45288420e-01 3.14034462e-01 4.79035884e-01
5.97524524e-01 2.43465304e-01 -6.79613829e-01 4.55809414e-01
-2.85996288e-01 3.49309444e-01 6.33955777e-01 1.61618125e+00
-2.06614718e-01 -1.07142568e+00 -8.31198394e-01 6.74023569e-01
-7.64415920e-01 4.77250248e-01 -2.49215961e-01 7.57588387e-01
2.49519467e-01 1.18172681e+00 1.44369574e-02 -7.59400427e-01
7.00658143e-01 3.65825921e-01 6.52857780e-01 -5.51420212e-01
-8.68215919e-01 1.36759982e-01 2.09298596e-01 -9.24385846e-01
-7.36933053e-01 -1.04172695e+00 -7.09241509e-01 -2.64925450e-01
-4.06123936e-01 -9.96070653e-02 8.60707819e-01 9.67232704e-01
5.71646035e-01 3.71091753e-01 4.64363158e-01 -1.25601161e+00
-2.42160320e-01 -1.00322545e+00 -6.60031915e-01 5.36846459e-01
2.28200868e-01 -9.56067443e-01 -2.16495857e-01 4.65426028e-01] | [7.432644844055176, 0.4443418085575104] |
583c3eef-882d-48f4-a0cf-eac4685e2b57 | a-score-level-fusion-method-for-eye-movement | 1601.03333 | null | http://arxiv.org/abs/1601.03333v1 | http://arxiv.org/pdf/1601.03333v1.pdf | A Score-level Fusion Method for Eye Movement Biometrics | This paper proposes a novel framework for the use of eye movement patterns
for biometric applications. Eye movements contain abundant information about
cognitive brain functions, neural pathways, etc. In the proposed method, eye
movement data is classified into fixations and saccades. Features extracted
from fixations and saccades are used by a Gaussian Radial Basis Function
Network (GRBFN) based method for biometric authentication. A score fusion
approach is adopted to classify the data in the output layer. In the evaluation
stage, the algorithm has been tested using two types of stimuli: random dot
following on a screen and text reading. The results indicate the strength of
eye movement pattern as a biometric modality. The algorithm has been evaluated
on BioEye 2015 database and found to outperform all the other methods. Eye
movements are generated by a complex oculomotor plant which is very hard to
spoof by mechanical replicas. Use of eye movement dynamics along with iris
recognition technology may lead to a robust counterfeit-resistant person
identification system. | ['Aurobinda Routray', 'Anjith George'] | 2016-01-13 | null | null | null | null | ['person-identification'] | ['computer-vision'] | [ 1.78078711e-01 -3.06888968e-01 1.40670445e-02 3.10234027e-03
6.84023619e-01 -4.21242803e-01 5.68300903e-01 -9.53355134e-02
-5.85807383e-01 7.63650417e-01 -6.03562631e-02 -4.06628698e-01
-3.40999335e-01 -2.28026763e-01 -1.42244488e-01 -6.60783589e-01
3.90868038e-01 -2.90448993e-01 1.30947009e-01 -1.56717762e-01
8.98761332e-01 8.56866658e-01 -2.01486254e+00 5.36432862e-02
5.25972843e-01 8.66161764e-01 -1.25472918e-01 8.35957587e-01
1.29104868e-01 3.75996500e-01 -6.81731403e-01 -2.59204004e-02
2.76220232e-01 -3.91431391e-01 -5.07503688e-01 1.43139372e-02
1.20212056e-01 5.69106415e-02 1.11066721e-01 9.81544673e-01
6.60494208e-01 2.04438031e-01 9.32173073e-01 -1.06747651e+00
-9.02071714e-01 -2.36289784e-01 -7.34575450e-01 6.64456129e-01
6.82699263e-01 2.72082776e-01 3.03436667e-02 -8.49278927e-01
3.92167002e-01 9.75136757e-01 4.57275540e-01 8.37536812e-01
-1.17091691e+00 -6.65936708e-01 -7.79172957e-01 3.56118947e-01
-1.30742347e+00 -5.85390925e-01 5.32838762e-01 -6.70981407e-01
1.12447190e+00 3.67381245e-01 5.83271027e-01 5.33588111e-01
6.11432254e-01 3.00657660e-01 1.54243302e+00 -8.29656899e-01
-1.58602372e-01 5.51337361e-01 8.86682928e-01 5.90425789e-01
5.99062562e-01 5.02790213e-01 -6.58431530e-01 6.13637865e-02
4.23628360e-01 -3.90785038e-02 -2.59338677e-01 2.47692183e-01
-7.34929681e-01 3.10463578e-01 4.32961136e-02 5.29426217e-01
-6.06852293e-01 -4.46019083e-01 -1.50252834e-01 3.29870433e-01
-4.35950817e-04 2.28328213e-01 -2.48633012e-01 -1.85809717e-01
-8.78804207e-01 -2.71201078e-02 5.11348248e-01 2.30337515e-01
2.32013941e-01 -9.00055319e-02 -3.05534542e-01 5.13004005e-01
8.86507332e-01 8.44995975e-01 6.78092003e-01 -1.30984157e-01
-5.65050021e-02 9.49298620e-01 1.02659635e-01 -1.00812447e+00
-6.51914716e-01 2.06762314e-01 -4.44486588e-01 7.28800893e-01
7.90319800e-01 -1.41826078e-01 -1.07092834e+00 1.08091295e+00
4.30545062e-01 1.14148416e-01 -7.10186586e-02 6.07962549e-01
8.21339726e-01 2.92993784e-01 -1.09822378e-02 -2.75422633e-01
1.63281906e+00 -1.91945568e-01 -1.00702631e+00 6.75535381e-01
1.07095815e-01 -8.62926424e-01 6.55263484e-01 6.44130647e-01
-1.01966393e+00 -7.45241582e-01 -1.06529772e+00 2.21371457e-01
-6.85947716e-01 4.01778907e-01 3.70810837e-01 1.54545963e+00
-8.39371026e-01 2.52639920e-01 -8.14167082e-01 -4.87235785e-01
3.16436023e-01 8.88043642e-01 -4.39938873e-01 5.47547877e-01
-6.01895571e-01 1.15252852e+00 3.64427231e-02 2.25811511e-01
1.01987541e-01 -1.54341102e-01 -5.52514195e-01 -2.45240033e-01
-5.14003038e-01 -3.27614933e-01 5.19849896e-01 -6.60459697e-01
-1.79663646e+00 9.47926402e-01 -7.37177789e-01 -3.76255572e-01
-3.31929736e-02 1.62254438e-01 -7.83658683e-01 1.62994206e-01
-5.12530446e-01 3.33922327e-01 1.06810987e+00 -3.88549656e-01
-6.20967686e-01 -7.90935040e-01 -5.67299902e-01 -1.13363810e-01
-2.86422312e-01 6.16017163e-01 1.86617389e-01 -3.32955718e-01
-3.04863006e-02 -9.01009381e-01 7.45429218e-01 -4.14996386e-01
-3.56894553e-01 -5.97510457e-01 8.11022103e-01 -6.76200151e-01
1.57152200e+00 -2.09207296e+00 -2.28178903e-01 6.16096199e-01
1.53023973e-01 8.28812242e-01 2.12163538e-01 1.97379008e-01
-2.12810725e-01 1.22707583e-01 4.29539323e-01 1.38045013e-01
-1.44456849e-01 -5.59485972e-01 1.73356771e-01 7.34664440e-01
1.43945873e-01 7.03502595e-01 -9.76377577e-02 -4.65334672e-03
2.09040314e-01 6.41079485e-01 5.63527942e-02 -1.88541993e-01
6.90174580e-01 2.28383392e-01 -2.41683573e-02 8.68053019e-01
6.25277281e-01 4.22993377e-02 -6.97914541e-01 -2.24921852e-01
-3.33052427e-01 -2.70282645e-02 -1.07341552e+00 9.95419979e-01
2.86273688e-01 1.10269010e+00 -5.72921813e-01 -4.88467455e-01
1.11081982e+00 4.70206350e-01 6.36859983e-02 -6.82059824e-01
5.62277019e-01 3.19019482e-02 5.74085236e-01 -9.62406218e-01
2.19060391e-01 1.42338216e-01 8.80883038e-01 7.84076750e-01
-1.57879740e-02 5.22363365e-01 1.45764545e-01 -4.47104007e-01
5.48280239e-01 2.43566647e-01 4.03166801e-01 -3.06510150e-01
1.03210545e+00 -3.42584699e-01 1.58490584e-05 3.53967786e-01
-5.13222039e-01 -6.07158393e-02 9.28455964e-02 -5.08464336e-01
-7.05258965e-01 -5.46070635e-01 -5.93728423e-01 5.04782438e-01
-9.79953539e-03 2.32041821e-01 -1.08518767e+00 -1.71097636e-01
1.25369346e-02 2.87972659e-01 -7.56416321e-01 -8.66097659e-02
7.18865171e-02 -8.19414437e-01 6.74589872e-01 -1.62483100e-02
5.05597472e-01 -1.29796124e+00 -1.12953615e+00 -4.98077534e-02
6.23889923e-01 -5.44434309e-01 -1.07322223e-01 -4.36153412e-01
-7.83575416e-01 -1.49510276e+00 -6.73262358e-01 -4.27708924e-01
7.82046735e-01 2.05018982e-01 2.23450705e-01 2.16901332e-01
-7.25218475e-01 2.97795326e-01 -1.00594454e-01 -1.09387088e+00
-1.49305195e-01 -3.55764747e-01 5.28576195e-01 5.00375092e-01
1.62426150e+00 -8.49603396e-03 -6.77191734e-01 2.22949430e-01
-5.30374467e-01 -3.83329660e-01 2.62315452e-01 6.38608277e-01
-5.11763096e-02 -2.96617160e-03 2.55991280e-01 -4.33077574e-01
1.10649800e+00 -2.84068316e-01 -6.98305726e-01 3.28148007e-01
-8.01798165e-01 -5.39139286e-02 -1.46016568e-01 -5.60800374e-01
-7.04098940e-01 -3.09664518e-01 4.59468216e-01 1.80935234e-01
-5.23287654e-01 2.24516124e-01 3.50737274e-01 -6.12259448e-01
1.03726077e+00 3.63436341e-01 6.23255074e-01 -2.26788655e-01
-5.48762202e-01 1.18832743e+00 3.24081123e-01 1.92295894e-01
5.95667303e-01 3.50969911e-01 3.40468824e-01 -1.35881603e+00
4.60321605e-02 -4.87887949e-01 -4.95040894e-01 -4.86078531e-01
9.33822930e-01 -2.85163611e-01 -1.60620117e+00 1.13064730e+00
-1.00149047e+00 3.99801463e-01 4.92581129e-01 8.68592739e-01
-6.37753587e-03 1.89592987e-01 -3.03128332e-01 -1.60147738e+00
-5.79740107e-01 -9.54919934e-01 3.88489068e-01 9.82791841e-01
-2.27071762e-01 -7.69970655e-01 -4.61412258e-02 3.03162247e-01
4.85636413e-01 1.16872162e-01 4.69538391e-01 -8.30400944e-01
-2.89820880e-01 -5.71882844e-01 -1.58170074e-01 1.86953738e-01
4.83017892e-01 4.97441113e-01 -1.14745319e+00 -1.77948875e-03
1.76694125e-01 1.00366123e-01 3.90911967e-01 6.91435039e-01
3.82073611e-01 1.70134902e-01 -2.15147778e-01 4.89038616e-01
1.25088871e+00 8.87304783e-01 9.44113910e-01 4.47557747e-01
2.01011553e-01 6.89526856e-01 1.82110682e-01 1.34559095e-01
3.13344449e-02 4.25978810e-01 -7.68531188e-02 1.52697518e-01
-9.34674367e-02 4.23884869e-01 2.82591343e-01 1.01531208e-01
-7.27982998e-01 -1.83415785e-01 -9.95173633e-01 5.39392345e-02
-1.13488352e+00 -1.19917846e+00 -4.88035917e-01 2.34208107e+00
5.33661366e-01 2.37770565e-02 4.58279520e-01 7.49457359e-01
8.63859653e-01 -7.67414808e-01 -3.70355904e-01 -5.70480645e-01
-9.73819122e-02 6.39312267e-01 3.46838355e-01 3.86786491e-01
-8.04269850e-01 6.32346153e-01 6.15980434e+00 1.86966136e-01
-1.51797795e+00 -2.70868272e-01 1.96289882e-01 -3.38774808e-02
3.84145439e-01 -4.90721762e-01 -1.24668586e+00 9.04496491e-01
1.12037754e+00 -1.01814875e-02 5.94147921e-01 -1.19510427e-01
5.23719490e-01 -7.33276069e-01 -5.31919241e-01 1.37345421e+00
2.76066840e-01 -1.08067560e+00 -2.59097487e-01 4.12190974e-01
2.83094674e-01 -2.77970463e-01 3.31089884e-01 -3.95842582e-01
-5.16130209e-01 -1.25043595e+00 2.23345701e-02 1.28436863e+00
7.88073421e-01 -6.04847550e-01 6.41691506e-01 7.53400922e-02
-6.69524193e-01 -1.77846834e-01 -2.87671238e-01 -1.87856212e-01
-2.93001205e-01 -2.37599507e-01 -8.12047839e-01 8.16275366e-03
4.25846636e-01 3.89825821e-01 -8.16033244e-01 1.48382819e+00
1.64665103e-01 6.21592641e-01 -5.29607654e-01 -5.40536523e-01
-2.24984989e-01 -1.89601019e-01 6.60407603e-01 8.46813619e-01
4.80857819e-01 4.33238782e-02 -9.57392573e-01 8.54295552e-01
4.91736054e-01 3.04713041e-01 -8.35627794e-01 -3.55629832e-01
4.62415427e-01 9.31465209e-01 -7.92496800e-01 1.19847255e-02
-4.06838953e-01 6.11871302e-01 -4.67667729e-01 4.47293520e-01
-2.38234773e-01 -7.88430691e-01 4.72683579e-01 2.08542883e-01
-9.37001780e-02 3.17164063e-02 -4.73028690e-01 -8.06151509e-01
-2.30680898e-01 -7.68376589e-01 -5.29968217e-02 -8.23688567e-01
-6.50974572e-01 6.38711631e-01 -2.12867945e-01 -1.01730978e+00
-2.11904988e-01 -1.16515517e+00 -4.57069159e-01 1.69027817e+00
-1.04690957e+00 -9.43263710e-01 -3.10102105e-01 8.60889971e-01
1.65590018e-01 -1.13202143e+00 8.84570241e-01 -8.50614011e-02
-7.35750496e-01 7.33114421e-01 -2.37198338e-01 1.20508172e-01
6.20871246e-01 -1.06463277e+00 -5.48618706e-03 1.13989544e+00
1.13480628e-01 1.11149693e+00 5.30567467e-01 -6.28695488e-01
-1.18439174e+00 -1.58183724e-01 1.07006264e+00 -8.15877259e-01
2.51020938e-01 2.10510060e-01 -6.30310476e-01 8.39738771e-02
4.57759470e-01 -2.69948304e-01 1.21140456e+00 -2.01876536e-01
8.77655596e-02 4.40909527e-02 -1.38849962e+00 3.72498989e-01
2.55142510e-01 -7.08510756e-01 -8.73711705e-01 -1.88204646e-01
-3.39408785e-01 -8.78155977e-02 -5.70822597e-01 6.46907091e-02
1.06378615e+00 -1.03989148e+00 5.64642012e-01 -9.44901943e-01
-1.34892836e-01 -6.08699739e-01 3.07570159e-01 -7.68233418e-01
-1.60957038e-01 -9.94083405e-01 -2.93592572e-01 1.00850534e+00
4.03097749e-01 -9.56926167e-01 5.75905323e-01 8.04315627e-01
6.11869872e-01 -3.79998803e-01 -5.12832880e-01 -4.95423526e-01
-5.26792467e-01 -3.61088887e-02 4.96886820e-01 6.13706529e-01
1.86728686e-01 2.67214745e-01 -1.11428551e-01 1.38318181e-01
7.35272944e-01 -5.24307668e-01 6.54460847e-01 -1.76189590e+00
4.32230905e-02 -5.48897505e-01 -1.03255880e+00 -1.77652940e-01
-3.69883180e-01 -4.64685410e-01 -6.99759841e-01 -9.12844300e-01
-1.39964953e-01 1.56988651e-01 -6.32469118e-01 2.75374204e-01
-1.10930942e-01 5.55297792e-01 8.00415576e-02 7.33476281e-02
4.59543824e-01 -3.64924997e-01 9.62723374e-01 2.64567912e-01
-5.73001146e-01 4.50913578e-01 -4.56323028e-01 3.84975821e-01
7.84125984e-01 -1.51709363e-01 -5.14953792e-01 1.24554284e-01
3.04239124e-01 -1.92601264e-01 3.23113918e-01 -1.21844339e+00
6.38344109e-01 1.47851422e-01 7.29479551e-01 -6.73162758e-01
4.50058170e-02 -6.54798985e-01 1.49271756e-01 5.20121038e-01
9.70459729e-02 2.70034492e-01 4.37126309e-01 3.53063047e-01
2.62166280e-03 -3.48243564e-01 8.27820659e-01 3.80177587e-01
-3.96787912e-01 -2.74587609e-02 -5.96412957e-01 -6.67368829e-01
1.34301198e+00 -1.25218415e+00 -4.36701864e-01 1.52350023e-01
-9.50757623e-01 -2.69975543e-01 2.74455816e-01 3.33711773e-01
5.85586250e-01 -9.70911682e-01 -4.80941862e-01 9.72072184e-01
-8.11742768e-02 -9.54384089e-01 2.50464510e-02 1.25085092e+00
-6.72480762e-01 8.20332766e-01 -9.25258100e-01 -5.18629968e-01
-2.02296734e+00 4.94151950e-01 4.17914301e-01 5.57163656e-01
-8.07122812e-02 8.86002362e-01 -6.67025089e-01 5.47922790e-01
2.90299594e-01 -1.12341762e-01 -1.21768582e+00 2.64957666e-01
1.06471264e+00 6.06650651e-01 3.02838236e-02 -7.17635155e-01
-3.39475423e-01 6.48533583e-01 -7.57381842e-02 -2.04150781e-01
1.00498295e+00 -1.74502164e-01 -2.68193036e-01 5.09433150e-01
5.84330857e-01 2.11263925e-01 -6.97691858e-01 6.47796467e-02
2.36525014e-01 -4.06669706e-01 5.80616621e-03 -1.06105947e+00
-5.57406366e-01 9.55624461e-01 1.22298014e+00 5.50906420e-01
1.16984212e+00 -8.93605292e-01 1.00916788e-01 3.80162537e-01
1.30606517e-01 -1.13930416e+00 -4.59216267e-01 1.05772533e-01
5.00592887e-01 -1.10393703e+00 -1.19400829e-01 2.20659330e-01
-3.93611372e-01 1.53996134e+00 3.71291816e-01 -6.19643815e-02
1.06140339e+00 1.36283636e-01 1.34697661e-01 -1.96178511e-01
-5.90898633e-01 -3.51100415e-01 8.90613139e-01 9.36491191e-01
7.17550159e-01 -1.03091866e-01 -8.54040444e-01 2.52537102e-01
-6.14372082e-02 5.58026373e-01 5.14629662e-01 6.68492854e-01
-4.17064130e-01 -1.02244151e+00 -7.43143022e-01 5.14147997e-01
-8.79376471e-01 -5.66767566e-02 -4.55893308e-01 7.19058931e-01
1.32937789e-01 1.20416498e+00 5.44776767e-02 -4.62888181e-01
7.78502747e-02 4.38028306e-01 6.53213799e-01 -3.29345345e-01
-6.74715996e-01 -9.15125236e-02 -4.64321882e-01 -3.47895950e-01
-8.00503373e-01 -8.37000310e-01 -8.70236993e-01 -3.67601633e-01
-3.90913278e-01 -3.23381387e-02 1.19237280e+00 1.00349116e+00
3.63485962e-01 2.20611587e-01 3.34409356e-01 -3.48411024e-01
-2.04288721e-01 -1.39304245e+00 -5.40722907e-01 2.55822420e-01
5.85547388e-01 -9.38812256e-01 -1.70274734e-01 2.95559198e-01] | [13.770920753479004, 0.49560460448265076] |
a610886a-9744-448d-8c61-91185be3f56e | best-sources-forward-domain-generalization | 1806.05810 | null | http://arxiv.org/abs/1806.05810v1 | http://arxiv.org/pdf/1806.05810v1.pdf | Best sources forward: domain generalization through source-specific nets | A long standing problem in visual object categorization is the ability of
algorithms to generalize across different testing conditions. The problem has
been formalized as a covariate shift among the probability distributions
generating the training data (source) and the test data (target) and several
domain adaptation methods have been proposed to address this issue. While these
approaches have considered the single source-single target scenario, it is
plausible to have multiple sources and require adaptation to any possible
target domain. This last scenario, named Domain Generalization (DG), is the
focus of our work. Differently from previous DG methods which learn domain
invariant representations from source data, we design a deep network with
multiple domain-specific classifiers, each associated to a source domain. At
test time we estimate the probabilities that a target sample belongs to each
source domain and exploit them to optimally fuse the classifiers predictions.
To further improve the generalization ability of our model, we also introduced
a domain agnostic component supporting the final classifier. Experiments on two
public benchmarks demonstrate the power of our approach. | ['Samuel Rota Bulò', 'Massimiliano Mancini', 'Barbara Caputo', 'Elisa Ricci'] | 2018-06-15 | null | null | null | null | ['object-categorization'] | ['computer-vision'] | [ 4.98503953e-01 6.29916862e-02 -2.14440599e-01 -7.31984556e-01
-4.07441825e-01 -7.21987545e-01 7.48876691e-01 2.34140962e-01
-1.97353840e-01 8.31950843e-01 -3.61924060e-02 -8.16207156e-02
-1.24071039e-01 -8.52566779e-01 -8.26601624e-01 -7.98724949e-01
7.41280988e-02 5.45444250e-01 6.17335856e-01 5.32341376e-02
1.61179692e-01 4.92928714e-01 -1.61067021e+00 6.35976553e-01
9.73731816e-01 1.11600423e+00 1.78497076e-01 4.10906792e-01
-1.77187592e-01 2.88737416e-01 -9.77385521e-01 -2.87684709e-01
3.96597266e-01 -4.39568281e-01 -7.63742685e-01 2.17363909e-01
4.86912221e-01 3.18690948e-02 1.27487183e-01 1.09317625e+00
3.36824834e-01 4.15739790e-02 1.06719816e+00 -1.77359235e+00
-7.79570282e-01 5.46743631e-01 -4.50736672e-01 1.64640561e-01
1.01850368e-01 1.58952083e-02 7.45471239e-01 -6.38874888e-01
5.84564984e-01 1.14686489e+00 4.34963584e-01 7.50943482e-01
-1.47285819e+00 -7.17280507e-01 4.26049888e-01 2.09099948e-01
-1.14987493e+00 -9.09218788e-02 8.78254473e-01 -6.96381271e-01
4.43948001e-01 -1.40102804e-01 2.74459988e-01 1.38419783e+00
9.62106511e-02 7.93505728e-01 1.43032193e+00 -4.73438382e-01
5.91635108e-01 6.70189440e-01 3.51470202e-01 7.31576830e-02
3.00784618e-01 -4.91494350e-02 -3.02793562e-01 -1.46865025e-01
3.68356705e-01 -1.31357715e-01 -3.43568295e-01 -9.94829595e-01
-9.30594921e-01 8.28727067e-01 6.46244347e-01 2.85642147e-01
-1.32646725e-01 -3.37138891e-01 4.75780308e-01 3.97862315e-01
4.44862962e-01 3.03359866e-01 -4.76261228e-01 3.98501873e-01
-6.61149263e-01 4.16320920e-01 8.03533018e-01 9.72566962e-01
8.76574874e-01 -9.60697904e-02 -3.40112060e-01 8.85881543e-01
3.34206894e-02 2.68049687e-01 6.61816955e-01 -3.45648974e-01
4.65651512e-01 9.52879786e-01 2.37916745e-02 -7.00302958e-01
-3.19147497e-01 -7.93482423e-01 -6.92471087e-01 5.43156385e-01
6.78027213e-01 -3.96500938e-02 -1.16003859e+00 2.09724975e+00
3.61167520e-01 2.87762195e-01 2.77167320e-01 7.36997426e-01
5.13374269e-01 4.22354400e-01 4.42213833e-01 1.49330050e-01
8.95856857e-01 -7.71082520e-01 -1.70741379e-01 -5.34924567e-01
4.49309468e-01 -3.93563271e-01 9.71623600e-01 3.74190807e-01
-6.09256029e-01 -6.67215586e-01 -1.24403477e+00 2.94035882e-01
-7.35559702e-01 3.31170261e-02 3.55806768e-01 7.17105985e-01
-7.26423383e-01 4.07933682e-01 -3.46870601e-01 -5.64210713e-01
6.26121342e-01 2.94222921e-01 -5.10052323e-01 -1.42719001e-01
-1.04940391e+00 8.69173110e-01 8.41829658e-01 -5.09866834e-01
-9.12772417e-01 -7.08360195e-01 -4.85855192e-01 2.24022940e-01
3.49515498e-01 -6.44241154e-01 1.21916556e+00 -1.55845034e+00
-1.12326455e+00 9.34660733e-01 -1.75882634e-02 -6.82781816e-01
5.24931908e-01 9.55412388e-02 -4.14489985e-01 -9.14000273e-02
1.49761766e-01 7.61323273e-01 1.09346390e+00 -1.56830275e+00
-9.09640253e-01 -3.78077060e-01 1.53736115e-01 5.01200482e-02
-3.60801131e-01 -2.32274249e-01 -2.23855134e-02 -7.13853180e-01
-1.56174172e-02 -9.08522487e-01 1.21711060e-01 9.73026454e-02
-2.88934290e-01 -3.87282193e-01 9.40079927e-01 -2.62709230e-01
8.06016922e-01 -2.27573752e+00 3.32371533e-01 1.49828821e-01
-1.40413968e-02 4.12382901e-01 -1.81722105e-01 1.15347728e-01
-4.51634586e-01 -4.57152948e-02 -5.07930815e-01 -1.28286287e-01
-1.93453545e-03 2.52419770e-01 -6.82923555e-01 2.54483789e-01
5.27458549e-01 6.03737652e-01 -6.59585595e-01 -1.62051484e-01
7.25881476e-03 2.89063334e-01 -5.11161149e-01 2.46867180e-01
-4.57355976e-01 4.35111880e-01 -3.00521463e-01 3.92560363e-01
9.42134917e-01 -2.13550016e-01 1.52164370e-01 -5.50624309e-03
3.70971680e-01 5.58432490e-02 -1.36560273e+00 1.29927886e+00
-3.43605638e-01 2.55107641e-01 -1.92149386e-01 -1.36064303e+00
1.18288767e+00 9.89823341e-02 2.24463627e-01 -3.87114435e-01
-3.75108942e-02 2.24308386e-01 1.99694932e-01 -7.17328638e-02
2.11034462e-01 -4.27940339e-01 -2.06293061e-01 2.69982159e-01
3.68396848e-01 3.13126296e-01 8.71046484e-02 1.99577231e-02
8.13311636e-01 3.34079802e-01 4.85744596e-01 -3.06930780e-01
5.79224288e-01 1.63682953e-01 7.13145137e-01 8.12146783e-01
-3.11412543e-01 7.52541304e-01 7.78463602e-01 -3.04071575e-01
-9.36539471e-01 -1.33625293e+00 -2.48143986e-01 1.12920737e+00
2.59833962e-01 -3.88240255e-02 -5.88261724e-01 -1.22828627e+00
2.45130301e-01 8.41540158e-01 -8.39111030e-01 -4.99613911e-01
-2.63954163e-01 -6.34665489e-01 2.21441537e-01 7.83406496e-01
4.53319997e-01 -9.11689520e-01 -6.95835829e-01 1.64554089e-01
9.00002271e-02 -9.96640086e-01 -5.18710352e-02 5.19851506e-01
-6.83721542e-01 -1.12963474e+00 -7.47692645e-01 -6.16032660e-01
5.88072181e-01 8.84914771e-02 1.06977510e+00 -3.87460798e-01
-1.45622298e-01 3.86787951e-01 -4.98528093e-01 -7.72799611e-01
-6.15795672e-01 3.94236483e-02 1.22087076e-01 3.98576945e-01
4.99643773e-01 -5.39006174e-01 -3.74379575e-01 4.52504575e-01
-1.09839594e+00 -1.59605473e-01 6.10774279e-01 7.76429415e-01
4.34497714e-01 1.36079609e-01 8.82606685e-01 -1.05527127e+00
4.83801812e-01 -8.44748139e-01 -5.35372555e-01 4.24268156e-01
-4.90630150e-01 2.31462300e-01 7.96176732e-01 -9.57877636e-01
-1.06869817e+00 2.81286299e-01 3.06632817e-01 -4.92555827e-01
-6.52055085e-01 3.68180811e-01 -5.68181157e-01 2.15500250e-01
8.85492027e-01 1.85870931e-01 -3.75510268e-02 -6.08511567e-01
2.39391834e-01 6.14560008e-01 5.20934522e-01 -6.13020480e-01
8.82302582e-01 4.14300978e-01 9.70062974e-04 -4.59441096e-01
-8.21563900e-01 -3.20537746e-01 -9.08985317e-01 3.06826904e-02
7.55582035e-01 -7.72778094e-01 -2.15083525e-01 4.22924995e-01
-1.01401508e+00 -3.28326643e-01 -3.91992569e-01 1.50984481e-01
-5.15116334e-01 1.71594381e-01 2.64256716e-01 -5.43805242e-01
2.20468298e-01 -1.17691398e+00 8.92309904e-01 3.28874975e-01
-2.31918126e-01 -1.12895644e+00 -6.51007369e-02 -1.32354945e-01
4.13294852e-01 3.44988972e-01 1.20218647e+00 -1.30894947e+00
-3.65078926e-01 -2.07020983e-01 -3.21211755e-01 6.14848971e-01
3.14980090e-01 -3.62671018e-01 -1.15906787e+00 -3.63478333e-01
-9.01006386e-02 -2.73276478e-01 1.00202274e+00 2.58732527e-01
1.25172341e+00 -1.68534845e-01 -6.36101782e-01 4.64333296e-01
1.44773757e+00 1.37179017e-01 4.62702781e-01 3.78355891e-01
3.21338028e-01 7.18836427e-01 6.36123180e-01 2.59541243e-01
7.59254619e-02 8.95057440e-01 4.15728390e-01 -2.00836137e-02
-2.55289644e-01 -1.50474474e-01 3.53611976e-01 -2.53717244e-01
3.71047109e-01 -5.32612622e-01 -9.70542133e-01 6.61236703e-01
-1.68183279e+00 -8.13635767e-01 1.32382169e-01 2.32226658e+00
5.54023266e-01 2.41494879e-01 4.03001010e-01 1.87875211e-01
7.73609459e-01 -1.48500219e-01 -8.48149478e-01 -3.05927396e-01
-1.80207327e-01 1.35873392e-01 2.70435601e-01 9.67429206e-02
-1.29443955e+00 6.71554267e-01 6.07131529e+00 6.41382754e-01
-1.39810801e+00 -5.49504682e-02 4.76730585e-01 1.42666131e-01
-2.41710292e-03 -2.41518375e-02 -1.01286733e+00 4.52263266e-01
7.06847370e-01 -4.80781406e-01 -8.68408680e-02 1.25386202e+00
-4.41570342e-01 -4.23437841e-02 -1.42472005e+00 5.34769237e-01
1.35809794e-01 -9.37138975e-01 2.69421160e-01 1.11511126e-01
5.96995413e-01 -2.39905253e-01 2.57723033e-01 4.77010816e-01
3.57060432e-01 -8.61165106e-01 6.77628219e-01 3.51936489e-01
7.73991764e-01 -5.99078953e-01 4.77253795e-01 4.33977574e-01
-9.48844671e-01 -4.74631697e-01 -4.51367021e-01 7.06642419e-02
-2.61927634e-01 3.40225846e-01 -1.21072912e+00 6.15315497e-01
7.25308001e-01 7.04054296e-01 -1.04888165e+00 1.25248623e+00
-1.65693417e-01 4.71837014e-01 -1.83397636e-01 2.95961291e-01
-6.75840154e-02 1.84279516e-01 6.95562601e-01 1.00809026e+00
2.89496213e-01 -4.20396239e-01 2.87621677e-01 9.75384712e-01
6.35915548e-02 -1.08038716e-01 -7.29495943e-01 2.24112660e-01
3.15872788e-01 1.10434055e+00 -7.15148032e-01 -4.28629160e-01
-3.83853078e-01 1.00276482e+00 4.06108320e-01 4.32737529e-01
-8.14021349e-01 -2.37011194e-01 6.75895333e-01 2.58519769e-01
5.52507877e-01 8.72886926e-02 -4.15125906e-01 -1.16346776e+00
1.64100766e-01 -8.52210224e-01 7.16237247e-01 -5.75977206e-01
-1.80433786e+00 6.20033920e-01 4.10251260e-01 -1.54438734e+00
-3.08641881e-01 -9.67190921e-01 -5.38896859e-01 1.07190180e+00
-1.57682312e+00 -1.25394666e+00 -3.47932756e-01 6.97575152e-01
4.66617703e-01 -3.63711804e-01 7.45553136e-01 5.90773532e-03
-2.39373341e-01 6.03012621e-01 7.40408525e-02 -2.03626920e-02
1.04935169e+00 -1.47745633e+00 1.04202591e-01 8.68909955e-01
-1.20113678e-02 5.76492846e-01 8.02879751e-01 -6.15884840e-01
-8.20741236e-01 -1.41211092e+00 5.01700997e-01 -5.09465635e-01
5.25568247e-01 -5.05457640e-01 -1.38206685e+00 7.45435715e-01
4.67114300e-02 1.51550338e-01 7.02082574e-01 6.48830011e-02
-7.75948822e-01 -2.27238998e-01 -1.25099421e+00 3.57941210e-01
9.77091312e-01 -1.89964861e-01 -7.29325414e-01 1.42308280e-01
4.12058324e-01 -2.20411286e-01 -5.89132845e-01 4.60871607e-01
2.49321908e-01 -1.10749829e+00 8.73984396e-01 -7.62311220e-01
3.97166520e-01 -4.41200376e-01 -1.91868037e-01 -1.68331635e+00
-3.21136713e-01 3.09634984e-01 4.18042913e-02 1.45793176e+00
3.63397747e-01 -7.90388644e-01 6.46757007e-01 4.88419414e-01
-1.05883613e-01 -3.75348747e-01 -1.01936638e+00 -1.09377849e+00
5.42956531e-01 -2.68793494e-01 8.67236495e-01 8.23253930e-01
-3.07305157e-01 3.73387009e-01 -1.27942637e-01 4.48867917e-01
5.03382742e-01 4.21149254e-01 8.01823854e-01 -1.69150174e+00
-3.88722986e-01 -5.10114312e-01 -6.15685225e-01 -7.68394530e-01
3.78667831e-01 -1.19504368e+00 -6.54235249e-04 -1.40498507e+00
2.36547351e-01 -5.53042412e-01 -4.21316028e-01 6.92230403e-01
-1.34422943e-01 7.88786933e-02 2.06360653e-01 3.11199129e-02
-3.50917161e-01 4.05122876e-01 8.32901776e-01 -1.93279400e-01
-1.96337566e-01 2.49343991e-01 -1.01891053e+00 5.66615462e-01
7.57797599e-01 -5.15442073e-01 -6.23995006e-01 -2.81221628e-01
-2.63956338e-01 -3.08130682e-01 7.13643849e-01 -1.30415893e+00
8.57306123e-02 -2.46122673e-01 6.45542979e-01 -2.93647975e-01
2.05260485e-01 -1.08028209e+00 1.84395835e-02 2.26184830e-01
-4.74748462e-01 -3.78818572e-01 3.18561941e-01 7.77909696e-01
-2.46217430e-01 -2.14549825e-01 1.21598637e+00 1.46355197e-01
-9.94656801e-01 1.46969318e-01 -2.40606777e-02 -3.78969163e-02
1.45875931e+00 -3.55085641e-01 -4.44048464e-01 -1.00650825e-01
-7.53326893e-01 5.14256135e-02 5.93270123e-01 7.47356772e-01
4.38929707e-01 -1.23030329e+00 -6.23677731e-01 3.68029445e-01
5.58109701e-01 -9.67356712e-02 1.15868367e-01 3.31299573e-01
2.91754931e-01 2.43371159e-01 -5.02484858e-01 -8.42661917e-01
-1.14368129e+00 9.37891006e-01 5.09769738e-01 -1.22823253e-01
-3.70802015e-01 6.81718051e-01 6.42549694e-01 -3.63577843e-01
1.59858182e-01 -1.82822973e-01 -3.17910701e-01 1.20657861e-01
4.68967617e-01 6.59038946e-02 3.23064588e-02 -5.92518866e-01
-4.06370938e-01 3.67431641e-01 -2.34724805e-01 7.55072534e-02
1.27992225e+00 8.33418742e-02 1.88826591e-01 6.41027272e-01
1.08408749e+00 -2.50116050e-01 -1.39565003e+00 -4.25009727e-01
2.94331640e-01 -4.47373748e-01 -4.65315878e-01 -1.14755690e+00
-7.82401323e-01 1.01209342e+00 7.97597826e-01 2.97765106e-01
1.38452888e+00 1.86768666e-01 1.18987240e-01 6.93571344e-02
2.10119948e-01 -9.06437159e-01 1.12578727e-01 3.82794172e-01
9.42535996e-01 -1.33938432e+00 -2.53449231e-01 -4.73545045e-01
-8.78925622e-01 1.09412467e+00 9.72434938e-01 -2.03531310e-01
4.79673594e-01 1.20692946e-01 -3.28196748e-03 2.61786997e-01
-7.35345900e-01 -1.92332074e-01 5.20435333e-01 1.03556979e+00
1.96278766e-01 -6.87714070e-02 -4.38429751e-02 7.55274594e-01
1.08660713e-01 1.54887855e-01 5.16350746e-01 8.82843196e-01
-3.29532862e-01 -1.46808684e+00 -3.93388748e-01 3.70194912e-01
-1.28207862e-01 2.73194104e-01 -5.09114563e-01 7.85958886e-01
3.31202149e-01 7.41593361e-01 1.35509729e-01 -2.44074956e-01
4.55363244e-01 5.43207765e-01 2.79041052e-01 -8.93972635e-01
-4.72030103e-01 -2.92838633e-01 -3.21343184e-01 -2.33261421e-01
-4.20919865e-01 -5.56144297e-01 -9.28629160e-01 3.39625925e-01
1.46232899e-02 -4.09286208e-02 5.79930723e-01 8.36233735e-01
4.56119567e-01 5.08117437e-01 5.38961530e-01 -5.32033861e-01
-7.35758364e-01 -8.09663534e-01 -5.90492666e-01 7.25705922e-01
4.31887805e-01 -1.10639250e+00 -3.04483443e-01 1.51406765e-01] | [9.946327209472656, 2.8199901580810547] |
8b2c6eba-26ff-4273-a88a-c7bd6544eba7 | a-real-time-and-high-precision-method-for | null | null | https://link.springer.com/article/10.1007/s00521-021-06526-1 | https://link.springer.com/content/pdf/10.1007/s00521-021-06526-1.pdf | A real-time and high-precision method for small traffic-signs recognition | As a fundamental element of the traffic system, traffic signs reduce the risk of accidents by providing essential information about the road condition to drivers, pedestrians, etc. With the rapid progress of computer vision and artificial intelligence, traffic-signs recognition systems have been applied for the advanced driver assistance system and auto driving system, to help drivers and self-driving vehicles capture the important road information precisely. However, in real applications, small traffic-signs recognition is still challenging. In this article, we propose an efficient method for small-size traffic-signs recognition, named traffic-signs recognition small-aware, with the inspiration of the state-of-the-art object detection framework YOLOv4 and YOLOv5. In general, there are four contributions in our work: (1) for the Backbone of the model, we introduce high-level features to construct a better detector head; (2) for the Neck of the model, receptive field blockcross is utilized for capturing the contextual information of feature map; (3) for the Head of the model, we refine the detector head grid to achieve more accurate detection of small traffic signs; (4) for the input, we propose a data augmentation method named Random Erasing-Attention, which can increase difficult samples and enhance the robustness of the model. Real experiments on the challenging dataset TT100K demonstrate that our method can achieve significant performance improvement compared with the state of the art. Moreover, it is a real-time method and shows huge potential applications in advanced driver assistance system and auto driving system. | ['Ronghui Zhang', 'Zhihan Lv', 'Wenquan Chen', 'Kunkun Jia', 'Junzhou Chen'] | 2021-09-25 | null | null | null | neural-computing-and-applications-2021-9 | ['traffic-sign-recognition', 'traffic-sign-detection'] | ['computer-vision', 'computer-vision'] | [-7.09093660e-02 -2.36979365e-01 -2.90727854e-01 -2.94368178e-01
-2.95397252e-01 1.16832338e-01 4.30419087e-01 -4.72612441e-01
-3.44493538e-01 3.49427968e-01 -5.84416278e-02 -3.37851137e-01
-5.18509671e-02 -6.89538896e-01 -3.50142062e-01 -9.40233111e-01
5.03419816e-01 5.76473735e-02 8.95291567e-01 -4.27554309e-01
3.86290550e-01 6.60939991e-01 -2.07075524e+00 4.02431237e-03
1.08048809e+00 1.11561465e+00 2.96269000e-01 2.37472534e-01
-6.04624487e-02 6.64515495e-01 -3.11509043e-01 -1.25337988e-01
2.42412865e-01 -2.86336273e-01 -5.98752089e-02 3.63236619e-03
5.32227099e-01 -2.33949214e-01 -7.86111712e-01 1.08151209e+00
5.19712329e-01 -1.98613126e-02 5.55800915e-01 -1.46591711e+00
-2.76365072e-01 -1.69928133e-01 -5.18361509e-01 4.92956609e-01
-4.26421344e-01 6.36854708e-01 5.00411510e-01 -9.83392477e-01
3.44083995e-01 1.13274324e+00 4.41638857e-01 5.42274833e-01
-5.46536148e-01 -9.60881531e-01 1.42300814e-01 1.18193138e+00
-1.41048241e+00 -6.16661727e-01 8.53434265e-01 -4.05656129e-01
5.34753442e-01 2.88079381e-01 7.29034543e-01 5.04595757e-01
-5.73281497e-02 1.10936809e+00 1.05942023e+00 -2.16183275e-01
-4.02547233e-02 1.90954104e-01 5.44815481e-01 7.30806410e-01
1.90295860e-01 5.02081871e-01 -1.28962681e-01 4.98963088e-01
4.80387509e-01 1.11700825e-01 -7.20120687e-03 -2.47060582e-01
-1.04112923e+00 6.48251593e-01 6.45722330e-01 3.23775291e-01
-5.12940288e-01 5.53610511e-02 2.86608726e-01 5.05607277e-02
-3.99398059e-03 -1.67995289e-01 -6.00975305e-02 -2.96507418e-01
-4.75611866e-01 2.64300138e-01 2.72281945e-01 7.82263517e-01
8.16556513e-01 2.65172422e-01 -4.84325945e-01 9.27009165e-01
1.21141665e-01 1.05001199e+00 4.93972570e-01 -5.86759150e-01
5.96844018e-01 8.05907071e-01 -1.24303736e-01 -9.05636907e-01
-4.76005763e-01 -5.40225565e-01 -8.76627028e-01 3.38008553e-01
3.26317161e-01 1.51710406e-01 -1.11167228e+00 1.26861620e+00
3.23542953e-01 5.69264591e-01 -1.46465302e-01 9.75724339e-01
8.90717626e-01 5.77095985e-01 6.14099875e-02 8.71206596e-02
1.53213859e+00 -1.15026343e+00 -8.29018831e-01 -3.64437997e-01
6.83534563e-01 -4.87516463e-01 7.40013957e-01 2.29365811e-01
-5.55187106e-01 -1.08579063e+00 -1.01685250e+00 1.62658177e-03
-6.10869706e-01 5.72281659e-01 5.53597808e-01 7.55134404e-01
-6.51979864e-01 -1.02211110e-01 -4.50718015e-01 -2.94874132e-01
6.26128256e-01 2.18287885e-01 -2.69235373e-01 -3.67469013e-01
-1.17339444e+00 1.30244780e+00 2.52324730e-01 5.57272375e-01
-6.79081380e-01 -4.43502098e-01 -8.00384343e-01 1.34996753e-02
5.91448545e-01 -2.80447453e-01 8.73504221e-01 -4.16991115e-01
-1.25924766e+00 7.08539546e-01 -5.37248135e-01 -4.26763386e-01
4.63808805e-01 -6.85196519e-02 -7.31388271e-01 -1.13937803e-01
2.82302275e-02 7.12970614e-01 9.52696085e-01 -9.55406308e-01
-1.17845464e+00 -2.24969760e-01 -3.93895417e-01 7.49021396e-02
-7.10197613e-02 -6.40310049e-02 -6.99651599e-01 -2.10537866e-01
-5.56861684e-02 -9.36549664e-01 -1.53301567e-01 -6.46886081e-02
-2.02493295e-01 -5.91104984e-01 1.41584444e+00 -5.04145265e-01
1.25886238e+00 -2.64320326e+00 -3.11842978e-01 3.66156787e-01
3.39321852e-01 1.07463109e+00 -1.98134571e-01 -1.13734037e-01
1.07756712e-01 -3.56561720e-01 -5.09422235e-02 -1.94093678e-02
-3.82541940e-02 3.72295916e-01 -3.06560814e-01 3.75627846e-01
3.80296081e-01 1.18792403e+00 -6.16003692e-01 -5.74601471e-01
9.79542851e-01 3.35076898e-01 -1.01026461e-01 -1.14424499e-02
4.20212865e-01 2.85321295e-01 -6.80745244e-01 5.81682444e-01
8.87009561e-01 3.51838857e-01 -7.33352184e-01 -2.76016146e-01
-5.68742335e-01 8.55276734e-02 -1.18545020e+00 7.95965850e-01
-3.68350089e-01 9.66213822e-01 -1.24755967e-02 -1.21303999e+00
1.17970443e+00 -7.99117312e-02 1.24025956e-01 -1.33894718e+00
2.05962986e-01 2.83444196e-01 3.73514324e-01 -7.92932153e-01
3.71916026e-01 2.49098316e-01 9.80415195e-02 -1.86000034e-01
-4.22489464e-01 7.13888407e-02 4.28332001e-01 -1.04788207e-01
8.30367923e-01 -3.22420329e-01 -1.36783225e-02 1.45515740e-01
1.12344682e+00 -1.48518339e-01 6.52519703e-01 5.64148068e-01
-6.31430328e-01 2.15905651e-01 1.70397237e-01 -5.31163514e-01
-7.62403786e-01 -7.74162412e-01 -3.96626443e-01 6.59239352e-01
4.10877705e-01 1.42425876e-02 -6.48689508e-01 -6.36046648e-01
2.75144249e-01 7.84890234e-01 -5.28423667e-01 -5.04823625e-01
-9.62851107e-01 -5.85254669e-01 4.93837684e-01 8.59457076e-01
1.10687160e+00 -1.20989680e+00 -4.65982497e-01 5.29160500e-02
-8.96138325e-02 -1.38361382e+00 -3.87974828e-01 -1.58602014e-01
-4.86811429e-01 -1.16207099e+00 -7.56380022e-01 -9.86244023e-01
6.96330011e-01 7.90868104e-01 2.92416632e-01 3.47264498e-01
-4.65946972e-01 -9.13683921e-02 -1.07869111e-01 -6.58260942e-01
-1.84003457e-01 -1.13754407e-01 -9.25050154e-02 4.48954850e-01
7.30179608e-01 -9.91515294e-02 -6.69982493e-01 8.35777700e-01
-3.94739240e-01 -9.51424167e-02 1.01592040e+00 7.47537196e-01
3.32616746e-01 1.65751539e-02 7.27400124e-01 -4.44793105e-01
3.74583036e-01 -2.01721396e-02 -7.93203771e-01 5.02398610e-02
-5.98909140e-01 -7.16828182e-02 5.33691287e-01 -1.75055653e-01
-9.51412082e-01 -4.59601693e-02 -3.84530693e-01 -2.71630645e-01
-4.40083683e-01 -7.63879418e-02 -4.39943314e-01 -3.66593987e-01
2.66176403e-01 5.20940006e-01 1.48741454e-01 -4.28649604e-01
4.03649598e-01 1.01020968e+00 7.23997831e-01 1.35388762e-01
9.75037992e-01 4.36925352e-01 3.98662180e-01 -1.09404254e+00
-4.22790438e-01 -8.62254798e-01 -7.39783347e-01 -4.51687187e-01
7.37192571e-01 -6.62373662e-01 -1.02373981e+00 8.42061460e-01
-9.22066331e-01 -4.13995609e-02 -2.73498774e-01 6.26034200e-01
-2.87198186e-01 3.69496375e-01 -2.34439559e-02 -8.73694003e-01
-8.59713182e-02 -1.28257847e+00 9.35410500e-01 5.37686646e-01
4.69535828e-01 -5.31601608e-01 -2.84363121e-01 5.76333046e-01
5.84432423e-01 -2.52228916e-01 7.28336453e-01 -4.31010187e-01
-9.38923776e-01 -5.42263210e-01 -7.67966926e-01 6.16327941e-01
3.70899364e-02 -1.94382906e-01 -1.08106208e+00 3.06036472e-01
-2.75975943e-01 2.62979828e-02 1.24259615e+00 3.92066628e-01
1.06670690e+00 1.18363231e-01 -7.03039706e-01 5.64303696e-01
8.00789356e-01 5.38105667e-01 1.06473756e+00 3.08045596e-01
7.73032010e-01 7.16591537e-01 8.98984492e-01 -2.39924365e-03
5.64269543e-01 8.45800161e-01 2.91161418e-01 -4.65446085e-01
-6.32357001e-01 -2.74837166e-01 2.63757974e-01 7.08765149e-01
-1.10206798e-01 1.51273802e-01 -5.73014498e-01 7.17537105e-01
-1.95287621e+00 -1.27431822e+00 -5.98871112e-01 2.18420053e+00
1.50657281e-01 2.72694916e-01 1.93268135e-01 4.75480884e-01
7.58153498e-01 9.74851623e-02 -6.67449951e-01 -1.63484126e-01
-1.46870375e-01 -1.21117443e-01 6.43031836e-01 3.25420588e-01
-1.03257775e+00 1.16607404e+00 5.27198696e+00 1.25086915e+00
-1.21061659e+00 3.21094203e-03 2.73558497e-01 3.33130419e-01
2.57183760e-01 -8.61855373e-02 -1.40322757e+00 6.99898243e-01
6.21482015e-01 3.58462892e-02 2.37495318e-01 9.43751633e-01
4.64141846e-01 -2.22888798e-01 -6.63116872e-01 1.16975152e+00
2.12975994e-01 -1.11368561e+00 -3.01205695e-01 1.22088797e-01
3.24484169e-01 6.94526881e-02 1.27407014e-01 6.35089457e-01
-2.40642354e-01 -7.53568411e-01 4.14862990e-01 5.58466792e-01
7.23325014e-01 -7.38954902e-01 9.42629874e-01 5.15862584e-01
-1.39116764e+00 -2.72437513e-01 -4.38806981e-01 2.20898405e-01
4.38896716e-01 4.59809721e-01 -6.32460952e-01 3.34428102e-01
3.89013410e-01 9.25418556e-01 -8.86508286e-01 1.55766904e+00
-4.78764415e-01 7.31999278e-01 -3.35216284e-01 -2.71890163e-01
3.56564730e-01 -2.49049142e-01 5.92162430e-01 1.16701424e+00
7.06198663e-02 -1.61516406e-02 3.29858231e-05 6.57880425e-01
2.91144937e-01 -7.60846958e-02 -7.66165078e-01 4.10455555e-01
3.07221919e-01 1.26128376e+00 -4.43472922e-01 -5.24754822e-01
-3.71419311e-01 4.52852309e-01 -8.85268897e-02 3.36426288e-01
-8.54239047e-01 -7.96545982e-01 6.53532386e-01 2.03875467e-01
5.28692365e-01 -1.54355571e-01 -3.86379153e-01 -7.89110780e-01
2.62803733e-01 -4.91394937e-01 2.18759980e-02 -6.90394223e-01
-6.35186076e-01 3.25759947e-01 -4.29635383e-02 -1.42101419e+00
6.59197867e-02 -8.30903053e-01 -7.55621016e-01 8.55691195e-01
-2.04923344e+00 -1.08855867e+00 -7.12405682e-01 5.39974391e-01
6.39794350e-01 -4.29787576e-01 1.53192610e-01 6.76097095e-01
-8.00121009e-01 6.49202704e-01 1.46222666e-01 2.57591486e-01
5.09750247e-01 -5.61187863e-01 5.01871824e-01 1.02580500e+00
-9.63623673e-02 2.53056437e-01 2.12504461e-01 -5.11251509e-01
-1.25067568e+00 -1.16414726e+00 9.13540184e-01 -2.02087551e-01
5.04013360e-01 -2.31456906e-01 -8.77011001e-01 2.06424639e-01
-1.49715126e-01 1.98011100e-01 6.44951239e-02 -1.97899103e-01
-1.07072480e-02 -7.81875193e-01 -1.02191246e+00 5.24240375e-01
9.59979534e-01 -2.28405237e-01 -6.80043101e-01 1.57656133e-01
2.45318726e-01 -1.55038908e-01 -1.09616622e-01 4.31602687e-01
5.90744019e-01 -7.45645046e-01 9.49104607e-01 -1.68412372e-01
-3.06328833e-01 -7.81589627e-01 3.09855670e-01 -1.05412769e+00
-4.87675637e-01 -2.60762930e-01 -6.14027632e-03 9.75094438e-01
5.14681712e-02 -1.02181923e+00 6.78872585e-01 3.89872611e-01
-6.16689920e-01 -7.16984749e-01 -1.21243048e+00 -7.66218126e-01
-3.83060217e-01 -7.82530487e-01 5.63037753e-01 2.52716154e-01
-3.70574474e-01 3.06370556e-01 -3.50746900e-01 7.43483528e-02
5.59474647e-01 -1.84399530e-01 1.09660017e+00 -1.22555351e+00
3.35453659e-01 -7.90372670e-01 -9.39261794e-01 -1.55420494e+00
-3.17197144e-02 -6.37537599e-01 2.49509931e-01 -1.54284573e+00
-8.40495378e-02 -5.61688960e-01 -3.60739678e-01 4.59820777e-01
-2.99668729e-01 2.27714464e-01 2.45209545e-01 1.25628501e-01
-5.80585301e-01 7.43262649e-01 1.47779572e+00 -2.46005714e-01
-1.13827683e-01 3.99994373e-01 -4.46188897e-01 6.81638002e-01
4.24591750e-01 -2.33048812e-01 -1.59381688e-01 1.87338293e-02
-3.45735788e-01 -3.35905015e-01 7.54044831e-01 -1.13623679e+00
4.38035905e-01 -4.95627038e-02 6.47475421e-02 -1.15507567e+00
4.39774930e-01 -8.42573524e-01 -5.33927262e-01 7.84320951e-01
1.30863652e-01 -3.23836118e-01 2.37851039e-01 4.55059350e-01
-3.75650883e-01 -2.20323533e-01 9.32300091e-01 3.85013580e-01
-1.36198270e+00 2.81069130e-01 -4.95073378e-01 -1.08476102e-01
1.27219665e+00 -5.14247179e-01 -5.90345562e-01 -1.67673990e-01
-1.33294106e-01 6.46980107e-01 -7.33416229e-02 7.32521653e-01
8.54453087e-01 -1.33494115e+00 -7.22644627e-01 7.42285728e-01
3.13324362e-01 -1.12476692e-01 4.20969218e-01 1.24101007e+00
-3.82293671e-01 9.71695364e-01 -1.59460947e-01 -7.23288715e-01
-1.29535735e+00 5.20575225e-01 2.65396059e-01 8.24379176e-02
-6.85166299e-01 4.34506744e-01 4.56095725e-01 -9.09335352e-03
2.72363573e-01 -5.56425154e-01 -6.13227665e-01 -6.19283691e-02
8.38967443e-01 7.60402560e-01 1.08539961e-01 -1.12889433e+00
-5.07644773e-01 1.09338200e+00 -1.67913273e-01 3.09656233e-01
9.74613845e-01 -1.43811062e-01 3.24612826e-01 1.47447372e-02
9.29949522e-01 -1.25157550e-01 -1.17187774e+00 -2.13399440e-01
-1.86620221e-01 -6.04732692e-01 2.32913390e-01 -6.05666697e-01
-1.15532017e+00 1.28254235e+00 9.75358009e-01 -6.99773431e-02
9.72089648e-01 -8.94252211e-02 1.10719419e+00 5.37039280e-01
2.36061826e-01 -1.11679626e+00 -3.15903038e-01 5.01723170e-01
6.88589633e-01 -1.45903313e+00 -2.72256762e-01 -4.20713097e-01
-8.57252061e-01 1.01052988e+00 7.96342850e-01 -3.83699010e-03
4.71838295e-01 6.31504655e-02 1.51665539e-01 -8.01947638e-02
-4.40706879e-01 -8.62830639e-01 6.56344235e-01 8.54546070e-01
-2.97789931e-01 -7.64613152e-02 -4.03397381e-01 3.62975508e-01
6.42898902e-02 1.74913988e-01 3.84671427e-02 4.94541168e-01
-8.58973980e-01 -8.51537466e-01 -3.68474424e-01 4.45248127e-01
2.24571720e-01 -3.69304866e-02 -1.31353751e-01 9.24702644e-01
5.03965259e-01 1.10205388e+00 1.86852012e-02 -4.56959754e-01
8.53410363e-01 5.96421920e-02 1.20068431e-01 -1.94472790e-01
-1.70092404e-01 -2.19221950e-01 5.06087281e-02 -4.97621000e-01
-1.29067153e-01 -7.35274255e-01 -1.28759611e+00 -9.27975327e-02
-5.19503474e-01 4.91263531e-02 7.75067329e-01 1.25764716e+00
4.42579806e-01 5.93699217e-01 8.12022686e-01 -6.62904084e-01
-3.73791575e-01 -1.01082861e+00 -3.76664430e-01 3.66662413e-01
4.26539779e-01 -1.07962930e+00 -2.47395292e-01 -1.73146278e-01] | [7.99085807800293, -0.8021739721298218] |
981995f4-815d-449c-bcea-82bb0252313c | layoutlmv2-multi-modal-pre-training-for | 2012.14740 | null | https://arxiv.org/abs/2012.14740v4 | https://arxiv.org/pdf/2012.14740v4.pdf | LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding | Pre-training of text and layout has proved effective in a variety of visually-rich document understanding tasks due to its effective model architecture and the advantage of large-scale unlabeled scanned/digital-born documents. We propose LayoutLMv2 architecture with new pre-training tasks to model the interaction among text, layout, and image in a single multi-modal framework. Specifically, with a two-stream multi-modal Transformer encoder, LayoutLMv2 uses not only the existing masked visual-language modeling task but also the new text-image alignment and text-image matching tasks, which make it better capture the cross-modality interaction in the pre-training stage. Meanwhile, it also integrates a spatial-aware self-attention mechanism into the Transformer architecture so that the model can fully understand the relative positional relationship among different text blocks. Experiment results show that LayoutLMv2 outperforms LayoutLM by a large margin and achieves new state-of-the-art results on a wide variety of downstream visually-rich document understanding tasks, including FUNSD (0.7895 $\to$ 0.8420), CORD (0.9493 $\to$ 0.9601), SROIE (0.9524 $\to$ 0.9781), Kleister-NDA (0.8340 $\to$ 0.8520), RVL-CDIP (0.9443 $\to$ 0.9564), and DocVQA (0.7295 $\to$ 0.8672). We made our model and code publicly available at \url{https://aka.ms/layoutlmv2}. | ['Lidong Zhou', 'Min Zhang', 'Wanxiang Che', 'Cha Zhang', 'Dinei Florencio', 'Yijuan Lu', 'Guoxin Wang', 'Furu Wei', 'Lei Cui', 'Tengchao Lv', 'Yiheng Xu', 'Yang Xu'] | 2020-12-29 | null | https://aclanthology.org/2021.acl-long.201 | https://aclanthology.org/2021.acl-long.201.pdf | acl-2021-5 | ['document-image-classification', 'document-layout-analysis', 'semantic-entity-labeling', 'key-information-extraction'] | ['computer-vision', 'computer-vision', 'natural-language-processing', 'natural-language-processing'] | [-1.00082465e-01 -2.47249305e-02 3.39252427e-02 -3.56404632e-01
-9.71712053e-01 -7.62663722e-01 6.37669146e-01 -4.50322405e-02
-1.58714026e-01 2.70891100e-01 2.93666452e-01 -6.28078938e-01
2.97882929e-02 -7.27765441e-01 -1.12306714e+00 -3.82550091e-01
2.74620503e-01 5.71104527e-01 1.75230309e-01 -3.04739654e-01
1.72634751e-01 1.15326889e-01 -1.12301481e+00 6.23618364e-01
1.03275192e+00 9.12631989e-01 7.95872509e-01 7.17856109e-01
-5.06333292e-01 6.70462608e-01 -4.42570239e-01 -5.77487826e-01
-1.37368634e-01 -2.23304302e-01 -8.25851500e-01 5.50670028e-02
7.60018408e-01 -5.38844526e-01 -5.82190931e-01 8.42440844e-01
4.79384989e-01 -4.37648259e-02 7.45822430e-01 -9.64277983e-01
-1.03604841e+00 5.58610380e-01 -1.02240884e+00 -3.56395207e-02
1.25497520e-01 1.67839393e-01 1.01467025e+00 -9.78082359e-01
6.88450396e-01 1.36865366e+00 3.67739856e-01 4.38530356e-01
-1.01627088e+00 -7.45922148e-01 4.18922752e-01 3.10541958e-01
-1.26138723e+00 -2.60146022e-01 5.31373501e-01 -3.93043905e-01
9.49814975e-01 1.34662390e-01 2.95651823e-01 1.03935933e+00
1.76783856e-02 1.30673850e+00 8.13038111e-01 -4.59976196e-01
-3.67265463e-01 4.08331538e-03 7.17760921e-02 1.03565085e+00
-1.14588007e-01 -3.05744261e-01 -3.77550960e-01 5.22173285e-01
8.92833650e-01 -1.04689755e-01 -2.84704089e-01 -1.92716390e-01
-9.48246360e-01 5.97484827e-01 7.67788589e-01 3.73772651e-01
8.78986865e-02 2.81223387e-01 4.29018766e-01 -6.61791582e-03
1.78831324e-01 7.76698813e-02 -2.19145268e-01 5.72351441e-02
-7.89225638e-01 -6.47697076e-02 4.59418967e-02 1.15888441e+00
7.60340929e-01 1.61384538e-01 -3.40618104e-01 1.07749486e+00
6.63758814e-01 8.76132309e-01 1.50781274e-01 -6.80755198e-01
1.09560096e+00 6.89701080e-01 -1.93425715e-01 -8.52491915e-01
-3.32802415e-01 -3.90638441e-01 -1.18448150e+00 -1.14978459e-02
3.51003230e-01 3.77157092e-01 -1.14496958e+00 1.71979427e+00
-4.08310853e-02 -3.32511753e-01 1.91089641e-02 7.32598066e-01
9.82064009e-01 9.96397436e-01 1.15592524e-01 2.01726228e-01
1.43989229e+00 -1.27902973e+00 -5.21569192e-01 -6.60208523e-01
6.72669172e-01 -1.02685463e+00 1.57538140e+00 2.51330912e-01
-1.30319095e+00 -8.68857205e-01 -1.01321769e+00 -6.08440816e-01
-3.27388734e-01 6.01852238e-01 4.73403811e-01 3.09805155e-01
-1.04731858e+00 -3.22199836e-02 -6.37012541e-01 -3.26623440e-01
7.08765209e-01 1.12511523e-01 -2.37611100e-01 -7.77000248e-01
-8.76167655e-01 4.72658515e-01 2.55436033e-01 1.58336014e-01
-9.03917730e-01 -6.39614463e-01 -8.81827712e-01 1.32946983e-01
2.64732122e-01 -5.87259650e-01 1.05915761e+00 -7.06944048e-01
-1.00147963e+00 9.43768144e-01 -5.11354625e-01 -1.13295913e-01
6.16308749e-01 -2.72499144e-01 -4.47856516e-01 2.99107432e-01
3.24269027e-01 1.01959443e+00 5.08374453e-01 -1.50905943e+00
-4.15850133e-01 -5.65534174e-01 1.03389509e-01 3.49298149e-01
-3.72669250e-01 -1.27710000e-01 -1.30954957e+00 -7.07120299e-01
-9.33224335e-02 -6.14008605e-01 2.68157542e-01 1.65441215e-01
-5.76789856e-01 -1.12829879e-01 9.53919053e-01 -9.68323171e-01
1.06034005e+00 -2.03111053e+00 1.57590345e-01 8.69416539e-03
1.34768248e-01 3.29883933e-01 -5.81665039e-01 5.40000319e-01
2.02731237e-01 1.84946761e-01 -1.17462181e-01 -6.15241945e-01
7.58961141e-02 -9.21398103e-02 -3.31142038e-01 1.76803499e-01
-6.49126247e-02 1.61540425e+00 -3.99160028e-01 -5.87625265e-01
4.83594894e-01 5.31441271e-01 -2.32631654e-01 2.65278250e-01
-4.86444175e-01 2.19454050e-01 -4.48688626e-01 7.84112453e-01
9.72474992e-01 -6.96659625e-01 1.81864351e-01 -6.11108601e-01
-6.12775050e-02 -7.66918808e-02 -9.35250700e-01 2.16319990e+00
-6.51941538e-01 9.67976868e-01 1.66402116e-01 -8.80214691e-01
7.47291744e-01 1.81393046e-02 6.86562359e-02 -1.44717336e+00
1.22085482e-01 -2.52906904e-02 -1.23240717e-01 -3.91079932e-01
4.48962808e-01 1.83053434e-01 1.69375667e-03 5.82527041e-01
1.61697999e-01 1.94561314e-02 3.72343153e-01 7.55040467e-01
7.78021038e-01 1.08608477e-01 -1.34847432e-01 -2.03594610e-01
5.80600798e-01 -1.91085041e-01 -8.15914869e-02 6.42511666e-01
9.86641943e-02 6.14037275e-01 2.81158835e-01 6.10537864e-02
-1.23669386e+00 -1.12831461e+00 1.73256360e-03 1.15572286e+00
5.01909137e-01 -4.21307981e-01 -5.84612966e-01 -4.28852022e-01
-7.80732483e-02 7.27395713e-01 -5.42233109e-01 -6.06755801e-02
-5.33707321e-01 -2.80565381e-01 6.00192547e-01 9.84035075e-01
8.58938217e-01 -9.76329446e-01 5.66206351e-02 -1.32912382e-01
-4.00944620e-01 -1.13443887e+00 -7.63344765e-01 -9.92278382e-02
-5.85312009e-01 -9.12823915e-01 -9.68991160e-01 -9.72086132e-01
7.30682611e-01 5.92279673e-01 1.04847074e+00 2.75684863e-01
-3.66026640e-01 3.85754228e-01 -3.30630958e-01 -2.38134921e-01
-2.98122242e-02 -4.04849574e-02 -4.09490794e-01 -2.32104361e-01
1.03326105e-01 -9.84853357e-02 -5.11397660e-01 4.54498261e-01
-8.88209939e-01 4.39429253e-01 7.86882639e-01 1.02817607e+00
5.30972421e-01 -1.29552469e-01 2.56998926e-01 -5.98200440e-01
3.38503152e-01 -1.83596954e-01 -5.67944586e-01 7.28770196e-01
-4.31389332e-01 -6.80067837e-02 6.93959713e-01 -2.92148858e-01
-1.38301170e+00 -2.81238139e-01 -1.34817421e-01 -5.64531922e-01
-1.19060151e-01 4.05040264e-01 -7.04238474e-01 2.95029044e-01
1.78707018e-01 5.30446589e-01 -2.10438907e-01 -5.81591368e-01
6.69113755e-01 6.59091294e-01 8.38710248e-01 -5.37792325e-01
8.63070548e-01 5.37554860e-01 -3.96598667e-01 -8.10158730e-01
-8.22949946e-01 -3.30886185e-01 -5.78590512e-01 -1.03811137e-01
9.23931420e-01 -1.04681957e+00 -8.89094174e-01 6.53425992e-01
-1.09452105e+00 -6.56956613e-01 2.29612797e-01 1.12890899e-01
-3.06735009e-01 5.52767456e-01 -6.56516671e-01 -6.23824775e-01
-5.05835652e-01 -1.04939973e+00 1.27464199e+00 3.44613552e-01
1.64895788e-01 -8.95448387e-01 -4.82092649e-01 8.96346927e-01
2.65804440e-01 -2.47074962e-01 1.18790317e+00 -1.88608378e-01
-1.01630557e+00 1.05147511e-01 -9.92507160e-01 2.92809546e-01
-1.10814691e-01 -7.47140869e-02 -8.58520925e-01 -5.10801911e-01
-6.21211469e-01 -5.39747179e-01 1.07240438e+00 5.30899644e-01
1.42437589e+00 3.71542736e-03 -5.13114154e-01 8.01995754e-01
1.47254038e+00 4.71937060e-01 9.13464010e-01 3.64166707e-01
1.21045148e+00 4.71109509e-01 6.08651519e-01 9.90609899e-02
6.66431367e-01 5.47939301e-01 4.76555258e-01 -3.94291759e-01
-4.64547008e-01 -6.38502598e-01 1.22494861e-01 6.53593600e-01
2.56630629e-01 -8.21987987e-01 -1.04547906e+00 5.51382661e-01
-1.71535182e+00 -8.40173721e-01 -2.79889435e-01 1.75272167e+00
5.24047375e-01 1.86245024e-01 -1.97404429e-01 -1.67823941e-01
6.98579848e-01 2.75430948e-01 -5.84010005e-01 -3.01705062e-01
-5.42021036e-01 8.84959772e-02 2.12260619e-01 4.49708611e-01
-1.05784059e+00 1.18878591e+00 4.29711056e+00 1.10934246e+00
-9.19640779e-01 -7.06679597e-02 7.77431846e-01 2.54138894e-02
-3.73612732e-01 -2.40033805e-01 -7.25194037e-01 5.37647784e-01
4.01125669e-01 -8.41031596e-03 2.77302533e-01 4.33563679e-01
4.73909341e-02 -1.76139325e-01 -9.48563814e-01 1.10051847e+00
2.86611974e-01 -1.61708140e+00 3.16116244e-01 4.90539894e-02
6.76206291e-01 2.25567315e-02 3.04162472e-01 2.31073931e-01
1.81497097e-01 -1.38927281e+00 9.62559104e-01 2.93938935e-01
1.36472344e+00 -7.27781951e-01 5.10710597e-01 2.18186125e-01
-1.65440845e+00 -2.12100133e-01 -2.94658840e-01 3.56284410e-01
3.94986391e-01 3.87026221e-01 -4.85061258e-01 7.56637096e-01
8.90039265e-01 8.95899475e-01 -5.11414707e-01 5.99288166e-01
-2.87592143e-01 3.35966259e-01 -7.14873523e-02 3.63437198e-02
4.52559948e-01 -5.83071001e-02 1.23163685e-01 1.26339567e+00
2.25690648e-01 3.43702696e-02 -8.76493379e-02 8.90332222e-01
-2.80961931e-01 1.47740424e-01 -3.32346380e-01 -1.00869097e-01
4.38110173e-01 1.05248797e+00 -5.05178809e-01 -3.37660789e-01
-5.38679898e-01 1.11479354e+00 3.44057620e-01 5.69039583e-01
-8.60872149e-01 -4.86108482e-01 3.36963415e-01 1.47813618e-01
6.03307903e-01 -2.54956186e-01 -2.96340317e-01 -1.11672318e+00
1.81203544e-01 -8.27048123e-01 3.82383764e-01 -1.44598472e+00
-1.11841285e+00 5.98368049e-01 -2.40898833e-01 -9.85338449e-01
1.67062089e-01 -8.07037294e-01 -5.94350278e-01 1.14380503e+00
-1.28859794e+00 -1.78575671e+00 -5.14112592e-01 5.55846989e-01
7.97114432e-01 -2.52651721e-01 5.53110838e-01 3.66425633e-01
-5.85680425e-01 8.05328965e-01 3.81021351e-01 2.75297791e-01
8.42539728e-01 -1.03645360e+00 5.56926310e-01 8.16545904e-01
2.00226933e-01 5.20793200e-01 1.00570537e-01 -6.45057917e-01
-1.48161006e+00 -1.08564007e+00 5.63848853e-01 -3.83242279e-01
4.82212573e-01 -6.49419844e-01 -1.07976496e+00 7.51540184e-01
4.58838552e-01 -3.40247273e-01 2.62286067e-01 -5.35897724e-02
-7.11910784e-01 -1.19833373e-01 -6.94345653e-01 6.02719188e-01
1.16975939e+00 -7.47964442e-01 -3.01382661e-01 1.41273543e-01
5.93130529e-01 -5.44125319e-01 -6.65598214e-01 1.80816978e-01
6.50910556e-01 -7.63058722e-01 1.18108249e+00 -3.03420812e-01
7.81119168e-01 -2.37539783e-01 -2.51748353e-01 -8.79511893e-01
-4.07777041e-01 -1.22668803e-01 -9.65985954e-02 1.54379880e+00
3.41243744e-01 -2.32527509e-01 5.99611759e-01 2.16325849e-01
-2.82143056e-01 -6.74479067e-01 -6.83506012e-01 -5.37593663e-01
3.15659523e-01 -4.47285652e-01 3.00119251e-01 7.65385985e-01
-2.62567759e-01 5.58287978e-01 -3.78637284e-01 1.18174985e-01
6.13998115e-01 3.37836057e-01 7.84928024e-01 -8.32739055e-01
-2.60144889e-01 -5.68338037e-01 3.64660360e-02 -1.76491272e+00
5.46956286e-02 -9.67595279e-01 -2.89903641e-01 -2.12793064e+00
6.41524792e-01 -3.44920099e-01 -1.73851714e-01 6.27883673e-01
-1.04955956e-01 2.12750226e-01 3.24892849e-01 3.83320600e-02
-7.57300436e-01 6.21041179e-01 1.56757331e+00 -6.35405064e-01
2.32075915e-01 -4.69528854e-01 -8.41111600e-01 4.47991222e-01
6.63473487e-01 1.89693466e-01 -5.92445910e-01 -1.10351753e+00
1.43264076e-02 -1.28896255e-02 5.92589200e-01 -5.66996813e-01
1.98732063e-01 1.32293552e-02 7.98555255e-01 -1.16847718e+00
4.23055887e-01 -4.00863945e-01 -1.89463735e-01 2.44976938e-01
-2.20440060e-01 1.04768328e-01 5.66945374e-01 5.14265180e-01
-2.91502625e-01 -4.81229536e-02 6.99096501e-01 -1.01730572e-02
-1.13584864e+00 2.55437046e-01 -2.35651881e-02 1.44888744e-01
8.50311279e-01 -2.80843794e-01 -8.75371635e-01 -4.79073912e-01
-3.23584110e-01 6.41127646e-01 4.85280544e-01 6.91892147e-01
6.63898289e-01 -1.10273671e+00 -6.08393610e-01 1.56455070e-01
2.72541702e-01 1.80364713e-01 9.79029536e-01 5.06053627e-01
-3.66889089e-01 7.32236266e-01 -1.28492653e-01 -7.35767066e-01
-1.47531044e+00 4.73788857e-01 2.41333872e-01 -1.74036995e-01
-6.47445261e-01 1.17886102e+00 7.56861031e-01 -3.82576585e-01
3.83897066e-01 -1.70509234e-01 2.12229803e-01 -1.28593743e-01
5.75308859e-01 2.61414289e-01 -1.46860540e-01 -7.65668273e-01
-5.04889011e-01 9.68313217e-01 -3.20417047e-01 -5.06035089e-02
1.05343473e+00 -3.56447577e-01 -1.20726265e-02 1.08003281e-01
1.56939709e+00 -1.39595523e-01 -1.21151543e+00 -2.36017868e-01
-5.29131353e-01 -3.62184346e-01 -6.80251271e-02 -1.22907877e+00
-1.10414028e+00 1.33958125e+00 5.16188025e-01 -2.96648622e-01
1.26614404e+00 3.04619491e-01 6.87980950e-01 1.74612701e-01
1.28886312e-01 -1.06784153e+00 5.16530871e-01 5.93263507e-01
1.00314617e+00 -1.35467815e+00 3.57164145e-02 -2.93863058e-01
-8.47855449e-01 8.90717506e-01 9.51361001e-01 4.15170252e-01
2.97864139e-01 1.89818606e-01 1.57568663e-01 -1.04810357e-01
-6.74924076e-01 -1.28583431e-01 5.68195701e-01 5.57597280e-01
5.24446547e-01 -6.56920299e-02 -3.99357341e-02 4.92039740e-01
1.74187738e-02 -3.71091485e-01 2.33640015e-01 7.54693270e-01
-1.59690276e-01 -1.05373037e+00 -1.70136288e-01 3.63530695e-01
-2.24706065e-02 -2.78927207e-01 -2.81350851e-01 9.29277837e-01
-6.75154626e-02 7.90733159e-01 2.96053082e-01 -1.95562318e-01
3.07157576e-01 -3.05939056e-02 5.11543691e-01 -2.29442209e-01
-1.38010651e-01 4.57374066e-01 -7.71931186e-02 -4.76586908e-01
-1.43065900e-01 -3.33813667e-01 -1.59533131e+00 -2.84980297e-01
-2.59010911e-01 -2.11921468e-01 6.14877641e-01 7.60846555e-01
4.61032152e-01 8.71696830e-01 2.34644115e-01 -6.36732221e-01
5.38256485e-03 -9.16560531e-01 -5.38589835e-01 2.89119989e-01
6.18406199e-02 -5.92868030e-01 8.99922326e-02 8.63013417e-02] | [11.425856590270996, 2.1499826908111572] |
f33c50fe-3c62-47a7-99e2-04e074554692 | a-flexible-extensible-software-framework-for | 2005.07786 | null | https://arxiv.org/abs/2005.07786v1 | https://arxiv.org/pdf/2005.07786v1.pdf | A flexible, extensible software framework for model compression based on the LC algorithm | We propose a software framework based on the ideas of the Learning-Compression (LC) algorithm, that allows a user to compress a neural network or other machine learning model using different compression schemes with minimal effort. Currently, the supported compressions include pruning, quantization, low-rank methods (including automatically learning the layer ranks), and combinations of those, and the user can choose different compression types for different parts of a neural network. The LC algorithm alternates two types of steps until convergence: a learning (L) step, which trains a model on a dataset (using an algorithm such as SGD); and a compression (C) step, which compresses the model parameters (using a compression scheme such as low-rank or quantization). This decoupling of the "machine learning" aspect from the "signal compression" aspect means that changing the model or the compression type amounts to calling the corresponding subroutine in the L or C step, respectively. The library fully supports this by design, which makes it flexible and extensible. This does not come at the expense of performance: the runtime needed to compress a model is comparable to that of training the model in the first place; and the compressed model is competitive in terms of prediction accuracy and compression ratio with other algorithms (which are often specialized for specific models or compression schemes). The library is written in Python and PyTorch and available in Github. | ['Miguel Á. Carreira-Perpiñán', 'Yerlan Idelbayev'] | 2020-05-15 | null | null | null | null | ['low-rank-compression'] | ['computer-code'] | [ 3.80123287e-01 1.10139124e-01 -2.52670974e-01 -3.35030496e-01
-4.55737084e-01 -3.32559764e-01 4.91349131e-01 3.02104294e-01
-6.05266809e-01 4.18734580e-01 -2.97505081e-01 -5.63159466e-01
-2.00992107e-01 -9.66517627e-01 -7.80133486e-01 -7.11574674e-01
-5.25326952e-02 6.44194603e-01 5.27527511e-01 1.78091228e-01
3.41927409e-01 6.43233120e-01 -1.96204269e+00 3.92160296e-01
5.70962727e-01 1.17447221e+00 4.96411890e-01 7.93078601e-01
-4.26529795e-01 6.29772484e-01 -2.22720325e-01 -2.87228614e-01
3.83145213e-01 -4.36739981e-01 -8.21828723e-01 -2.48489857e-01
1.89336210e-01 -3.43544185e-01 -2.28389189e-01 1.04594791e+00
3.14044327e-01 -1.52988851e-01 4.41841424e-01 -9.47583139e-01
2.56599724e-01 7.30906010e-01 -2.26365045e-01 -8.64898637e-02
-1.13246322e-01 -1.81733876e-01 7.83923626e-01 -6.28808081e-01
3.51331741e-01 9.84086633e-01 7.29328871e-01 4.53296900e-01
-1.43652487e+00 -6.68435693e-01 -2.15768352e-01 2.17044711e-01
-1.35298419e+00 -5.93046486e-01 4.99509066e-01 -4.84275460e-01
8.62955034e-01 5.39829254e-01 7.90832937e-01 6.88217819e-01
3.93719934e-02 7.17964709e-01 7.36973166e-01 -6.07312202e-01
6.63158298e-01 1.58729419e-01 3.23160648e-01 6.10284567e-01
2.81165689e-01 1.46813467e-01 -3.31409454e-01 -2.00596780e-01
6.64902091e-01 -4.96238917e-02 -2.60703653e-01 -3.55090916e-01
-8.48912120e-01 8.21586132e-01 1.11412384e-01 3.70457172e-01
-3.15913498e-01 1.34724736e-01 7.28281379e-01 5.52004278e-01
2.07263619e-01 3.20527673e-01 -6.23573422e-01 -2.87140369e-01
-1.61948848e+00 3.12469065e-01 1.12469459e+00 7.40891755e-01
9.28910553e-01 -4.12920713e-02 1.02017298e-01 1.00271559e+00
2.20860571e-01 9.11247730e-03 8.36183488e-01 -1.04359102e+00
2.68155545e-01 5.18493950e-01 -3.28948855e-01 -8.17404628e-01
-4.51545209e-01 -5.81461489e-01 -1.00274098e+00 3.32949698e-01
3.13819021e-01 -2.33376604e-02 -6.96380794e-01 1.59704423e+00
1.69403136e-01 2.71733433e-01 -4.67178673e-02 5.89215934e-01
7.15032279e-01 6.82300627e-01 -1.31821007e-01 -3.42045039e-01
1.15491092e+00 -8.68233919e-01 -4.58900750e-01 -1.86396495e-01
8.64289820e-01 -5.85317910e-01 1.10968828e+00 7.27979064e-01
-1.29555142e+00 -6.09908342e-01 -1.24909043e+00 -1.87899787e-02
-5.15442610e-01 1.67038471e-01 4.46064055e-01 4.36986357e-01
-1.12889576e+00 1.20801091e+00 -9.90225852e-01 -1.11670867e-01
1.82497486e-01 5.59328079e-01 -1.06706187e-01 1.81647539e-02
-1.14962697e+00 6.93465292e-01 7.76695609e-01 -2.44248495e-01
-6.50770426e-01 -6.50249004e-01 -7.35980570e-01 4.99201596e-01
1.93632200e-01 -5.50016284e-01 1.13822079e+00 -1.11968243e+00
-1.53268039e+00 6.91219509e-01 -1.11733032e-02 -8.42800319e-01
5.47026813e-01 -6.68819994e-02 1.53778968e-02 1.84660882e-01
-5.26296496e-01 7.02021182e-01 8.84044707e-01 -9.16151822e-01
-4.55825001e-01 -2.94390142e-01 -5.16873337e-02 -4.75048572e-02
-5.07484674e-01 -3.83274779e-02 -8.19026411e-01 -6.36672258e-01
2.15028167e-01 -8.12674940e-01 -1.84260160e-01 8.78731981e-02
-1.70954809e-01 -1.05296001e-01 7.78148413e-01 -7.49928355e-01
1.56742108e+00 -2.29432511e+00 2.27492228e-01 4.43599164e-01
-3.13190110e-02 5.30934334e-01 -6.82467669e-02 3.72313291e-01
-3.15819710e-01 1.77813433e-02 -4.56899136e-01 -5.91414094e-01
-1.89217284e-01 4.52602893e-01 -1.14432387e-01 1.81581244e-01
-1.14684045e-01 3.07376981e-01 -6.44690633e-01 -5.09924710e-01
1.25630125e-01 5.42545855e-01 -7.39389241e-01 1.22253820e-01
-2.30043158e-01 -6.21115938e-02 -1.74801439e-01 1.66609138e-01
6.54429555e-01 -2.51635432e-01 3.79782379e-01 -2.84554094e-01
-2.53882051e-01 7.20714748e-01 -1.61090922e+00 1.59258604e+00
-4.27361399e-01 4.58225161e-01 1.18126884e-01 -1.13544154e+00
9.16584373e-01 3.89755368e-01 2.27434248e-01 -1.45687431e-01
-2.71926448e-03 2.71581650e-01 -1.48375094e-01 -3.32229942e-01
2.50742257e-01 2.78549790e-01 4.49944377e-01 5.30162513e-01
2.66916245e-01 -1.45700976e-01 6.01860106e-01 2.38204794e-03
1.16814947e+00 2.72765398e-01 2.91549146e-01 -1.86289877e-01
4.94367331e-01 -1.49761021e-01 3.76944989e-01 6.09449446e-01
4.49244916e-01 5.06298959e-01 6.81167543e-01 -3.04813772e-01
-1.12242401e+00 -6.66246533e-01 -3.61753285e-01 1.07071686e+00
-4.33696151e-01 -9.10474420e-01 -1.03619897e+00 -3.29198480e-01
-1.90525025e-01 6.51219368e-01 -3.42255384e-01 -2.05736235e-01
-4.53112245e-01 -5.59117317e-01 3.99389118e-01 2.90810496e-01
5.78143597e-01 -1.04175150e+00 -7.66788125e-01 7.03247041e-02
4.20160927e-02 -7.11645544e-01 -1.10025316e-01 7.93722332e-01
-1.46156740e+00 -9.10516381e-01 -2.79479653e-01 -6.40589476e-01
6.53369308e-01 -4.05559018e-02 8.60584140e-01 2.98308134e-01
3.89645025e-02 -1.03509314e-01 -3.36419642e-01 -2.13849068e-01
-4.61519361e-01 2.93736428e-01 -1.67742074e-01 -1.23332262e-01
2.14641109e-01 -9.64339733e-01 -3.07156503e-01 -1.07589476e-01
-1.25409102e+00 3.38763326e-01 6.18645668e-01 6.41822636e-01
8.43426526e-01 3.75330061e-01 6.98075891e-02 -1.02266383e+00
5.01225710e-01 -3.68917793e-01 -9.56805408e-01 4.39187773e-02
-9.12010252e-01 3.42622489e-01 9.69118536e-01 -3.74956399e-01
-3.21320981e-01 4.80294317e-01 -3.52088898e-01 -8.10882330e-01
-1.91199139e-01 7.90901005e-01 -1.58650488e-01 -5.81426173e-02
5.01782954e-01 4.01931167e-01 2.44307429e-01 -9.41863477e-01
-2.36399546e-02 6.22857690e-01 3.93986404e-01 -3.07838887e-01
6.53859854e-01 -1.29538393e-02 3.06863580e-02 -7.55173266e-01
-5.96068203e-01 -2.14206412e-01 -6.43441975e-01 2.62606554e-02
5.24696350e-01 -5.60317397e-01 -5.42411625e-01 3.59652728e-01
-1.02366412e+00 -4.48519498e-01 -5.30219138e-01 5.14002442e-01
-5.21484494e-01 3.46837282e-01 -8.02286685e-01 -4.68081802e-01
-3.07147741e-01 -1.18585134e+00 6.43684506e-01 2.36332156e-02
-1.25667140e-01 -6.99571669e-01 -5.53486273e-02 4.39885771e-03
4.01375502e-01 -9.35875103e-02 1.12972844e+00 -8.42733383e-01
-3.99172962e-01 -4.08330500e-01 -2.08522044e-02 7.84526467e-01
-3.05666775e-01 4.99648787e-02 -9.64569509e-01 -3.58001173e-01
2.21107349e-01 -2.37230301e-01 9.68307674e-01 3.76515567e-01
1.79155874e+00 -6.20537281e-01 -1.61289811e-01 8.51209998e-01
1.37598443e+00 6.05612583e-02 8.58001947e-01 3.03861231e-01
2.89537132e-01 5.03551126e-01 1.19320996e-01 3.44724536e-01
-1.46411555e-02 6.21600509e-01 4.44262624e-01 -1.21415325e-01
4.59553115e-02 -2.31766090e-01 3.91065329e-01 1.05155575e+00
-5.48073091e-02 3.31424475e-01 -7.87229598e-01 6.49268255e-02
-1.73274553e+00 -1.02988613e+00 -2.47641169e-02 2.72136235e+00
1.12366533e+00 3.87017488e-01 1.60481691e-01 9.51635361e-01
4.08108234e-01 -5.66154756e-02 -3.61038089e-01 -6.55105770e-01
2.31188834e-01 3.30017686e-01 6.08888090e-01 6.60480440e-01
-8.78581166e-01 6.09948397e-01 6.39747715e+00 1.09371948e+00
-1.34787738e+00 -6.36871234e-02 5.12200117e-01 -1.74708292e-01
-4.96922471e-02 4.33831513e-02 -9.59590673e-01 4.85769033e-01
1.32776940e+00 4.73907702e-02 8.89234066e-01 1.02397311e+00
2.32942373e-01 -1.00462779e-01 -1.29576600e+00 8.18624020e-01
-1.92801952e-01 -1.42292368e+00 1.37057528e-01 1.38062239e-01
5.50553910e-02 1.29419997e-01 -2.48423979e-01 2.33463883e-01
-9.77295116e-02 -7.26509035e-01 6.94849074e-01 5.36090076e-01
7.00768054e-01 -8.83194327e-01 6.13897562e-01 7.72480369e-01
-9.86372173e-01 -7.47621879e-02 -5.87354064e-01 1.12333715e-01
-1.69155613e-01 8.40954065e-01 -5.73987186e-01 3.38711888e-01
8.27047527e-01 5.88432789e-01 -6.47952735e-01 1.16981578e+00
-6.38707653e-02 9.95105386e-01 -4.32966322e-01 2.82488614e-01
4.33827303e-02 -4.56954539e-01 4.63882387e-01 1.47021067e+00
2.75683403e-01 -2.02045798e-01 -2.43427791e-02 6.78872228e-01
9.71642584e-02 2.24512398e-01 -3.44815612e-01 1.04222037e-01
5.18212557e-01 1.26904798e+00 -6.98642612e-01 -4.51252252e-01
-3.16194445e-01 7.18431652e-01 2.75400966e-01 1.72607288e-01
-5.71241200e-01 -5.74795604e-01 1.46172523e-01 4.38837886e-01
5.33584654e-01 -1.08344227e-01 -3.02544594e-01 -8.28235030e-01
3.15168276e-02 -1.05674314e+00 4.27070379e-01 -5.69723725e-01
-7.51161158e-01 5.69819152e-01 2.37218827e-01 -1.19734728e+00
-3.40053350e-01 -4.91430193e-01 -2.86084831e-01 7.00000644e-01
-1.46128619e+00 -5.36969304e-01 -1.46512344e-01 5.65883934e-01
3.13956857e-01 -7.14943111e-02 9.93153632e-01 4.91124690e-01
-4.39166516e-01 5.92235804e-01 1.27604544e-01 -2.73596626e-02
3.96373510e-01 -1.03907144e+00 -9.76179987e-02 5.82058847e-01
8.55758190e-02 6.86708331e-01 6.65841162e-01 -2.59498179e-01
-1.19467616e+00 -1.04143941e+00 1.27072620e+00 3.30050319e-01
3.37002486e-01 -4.58463967e-01 -1.18215573e+00 4.85388786e-01
-2.30325416e-01 -1.58337548e-01 7.42512643e-01 9.10204351e-02
-2.72469074e-01 -4.79025394e-01 -1.08964944e+00 4.26397175e-01
6.57105207e-01 -3.57060432e-01 -3.28999847e-01 4.31169689e-01
6.85614049e-01 -3.49449039e-01 -8.95292044e-01 3.47471833e-01
5.30484438e-01 -1.18802619e+00 8.71303558e-01 -3.57554048e-01
5.34560978e-01 -3.41067225e-01 -3.09009314e-01 -8.40023220e-01
-2.94229835e-01 -4.92891103e-01 -6.75731063e-01 1.11942065e+00
5.92487574e-01 -5.61444223e-01 6.84227824e-01 3.38725805e-01
-1.71876699e-01 -9.48937297e-01 -9.51575518e-01 -6.97506905e-01
-1.46242112e-01 -6.01261973e-01 6.13069415e-01 7.67300606e-01
-6.03353605e-02 2.82458216e-01 -2.27019548e-01 -1.65888052e-02
3.35698098e-01 5.39018586e-02 6.44965351e-01 -1.44053447e+00
-8.71710241e-01 -5.82425535e-01 -3.07535142e-01 -1.15532935e+00
-6.30006567e-02 -1.07836187e+00 -2.69382209e-01 -1.27256787e+00
2.90372465e-02 -5.81301928e-01 -2.99173176e-01 8.46994936e-01
2.87855804e-01 1.72067787e-02 1.83189824e-01 4.74455297e-01
-1.80501238e-01 1.92600027e-01 7.10822701e-01 7.60225058e-02
-3.56252342e-01 3.62798750e-01 -3.81387681e-01 8.87875319e-01
7.24947333e-01 -6.51193440e-01 -3.85268539e-01 -3.32134664e-01
2.04886481e-01 1.25936046e-01 2.90179133e-01 -1.34753489e+00
3.41117978e-01 1.73391297e-01 3.31820816e-01 -5.25922716e-01
2.80388862e-01 -8.89148593e-01 3.70009571e-01 7.95845568e-01
-5.24737656e-01 -2.51631290e-02 2.08851784e-01 1.51226491e-01
-1.84067175e-01 -8.16030085e-01 1.07735896e+00 -6.68443739e-02
-1.38900116e-01 1.66406319e-01 -2.87555933e-01 -3.25304300e-01
7.10274875e-01 -3.85565370e-01 9.31147411e-02 -3.06418389e-01
-8.49396408e-01 -1.37109548e-01 4.89326715e-01 1.45239696e-01
4.08470511e-01 -1.08198321e+00 -4.69495595e-01 5.13042748e-01
-3.05553675e-01 1.36826202e-01 -2.49095157e-01 9.03091252e-01
-5.24352610e-01 2.43750647e-01 1.23841584e-01 -5.34313321e-01
-1.25800884e+00 5.63862920e-01 4.24585760e-01 -5.60200453e-01
-6.72649205e-01 5.27626395e-01 -3.13657165e-01 -3.20655197e-01
6.01074636e-01 -2.80598074e-01 -2.51802176e-01 4.92926799e-02
7.31729567e-01 3.77160996e-01 4.07157630e-01 -2.25025266e-01
-9.46352072e-03 4.10354048e-01 -4.56885323e-02 -3.01139913e-02
1.51979065e+00 2.89035499e-01 -2.67429769e-01 5.86486518e-01
1.23743510e+00 -2.70497501e-01 -1.12947917e+00 -1.70177579e-01
3.34117636e-02 -1.94140598e-01 3.63108248e-01 -5.55861175e-01
-1.20564246e+00 8.61160457e-01 6.29528165e-01 4.00494128e-01
1.52429795e+00 -2.57921159e-01 5.34530759e-01 4.05271769e-01
3.13670397e-01 -1.12944877e+00 -4.16729182e-01 7.00960398e-01
7.66329229e-01 -5.22444367e-01 7.55429938e-02 -1.85897857e-01
-2.51446933e-01 1.34859776e+00 1.72807142e-01 -2.05586612e-01
8.39512825e-01 6.61167562e-01 -2.41889134e-01 1.99661218e-02
-9.16063368e-01 8.31618607e-02 2.75580972e-01 2.24720761e-01
4.80161130e-01 -2.31366783e-01 -6.89703226e-01 5.58071375e-01
-4.81707692e-01 3.35600525e-01 2.14010864e-01 8.86013329e-01
-6.54987395e-01 -1.37532282e+00 -2.19479233e-01 8.06050062e-01
-2.96537757e-01 -3.04273129e-01 -6.43333942e-02 5.39712131e-01
4.24619496e-01 6.52319729e-01 1.94154859e-01 -5.97190380e-01
1.52061030e-01 2.99388826e-01 4.10354942e-01 -5.62316120e-01
-6.69340432e-01 1.84354171e-01 -3.00853811e-02 -7.97794580e-01
-2.57618278e-01 -7.20105946e-01 -1.23031223e+00 -3.99531335e-01
-2.68258005e-01 3.14348340e-01 8.86500478e-01 8.96521270e-01
2.54523128e-01 1.81284681e-01 4.83843654e-01 -1.02986097e+00
-6.48743391e-01 -8.79911005e-01 -5.77903152e-01 -3.22584808e-02
8.60011056e-02 -3.33042532e-01 -5.48024237e-01 2.87566990e-01] | [8.54240894317627, 3.32901930809021] |
c9361820-564f-406c-a5a2-5b8745d64417 | early-recognition-of-human-activities-from | 1406.5309 | null | http://arxiv.org/abs/1406.5309v2 | http://arxiv.org/pdf/1406.5309v2.pdf | Early Recognition of Human Activities from First-Person Videos Using Onset Representations | In this paper, we propose a methodology for early recognition of human
activities from videos taken with a first-person viewpoint. Early recognition,
which is also known as activity prediction, is an ability to infer an ongoing
activity at its early stage. We present an algorithm to perform recognition of
activities targeted at the camera from streaming videos, making the system to
predict intended activities of the interacting person and avoid harmful events
before they actually happen. We introduce the novel concept of 'onset' that
efficiently summarizes pre-activity observations, and design an approach to
consider event history in addition to ongoing video observation for early
first-person recognition of activities. We propose to represent onset using
cascade histograms of time series gradients, and we describe a novel
algorithmic setup to take advantage of onset for early recognition of
activities. The experimental results clearly illustrate that the proposed
concept of onset enables better/earlier recognition of human activities from
first-person videos. | ['Thomas J. Fuchs', 'M. S. Ryoo', 'J. K. Aggarwal', 'Lu Xia', 'Larry Matthies'] | 2014-06-20 | null | null | null | null | ['person-recognition', 'activity-prediction', 'activity-prediction'] | ['computer-vision', 'computer-vision', 'time-series'] | [ 7.26187944e-01 -2.27142438e-01 -4.34673689e-02 -2.35214740e-01
-2.15032488e-01 -4.02415991e-01 8.69832516e-01 1.01822875e-01
-4.83302057e-01 4.54316348e-01 5.05495369e-01 3.17043126e-01
-1.98021501e-01 -3.19623768e-01 -5.40990531e-01 -7.88372874e-01
-6.18659377e-01 7.49625713e-02 4.69802618e-01 3.09638143e-01
2.95864463e-01 9.40082967e-01 -2.06163859e+00 7.02155769e-01
2.26835236e-01 6.81475878e-01 6.62411600e-02 1.37473226e+00
1.97877735e-01 1.40602469e+00 -8.59337807e-01 1.37937322e-01
8.81801322e-02 -8.11236441e-01 -6.43430471e-01 7.42262363e-01
1.15832806e-01 -5.57224929e-01 -2.94707239e-01 4.86710966e-01
1.66431785e-01 3.12034786e-01 7.06583798e-01 -1.39054799e+00
2.02271432e-01 -3.30963917e-02 -2.28729963e-01 7.95652449e-01
1.34695065e+00 2.26528734e-01 5.69769919e-01 -8.76694083e-01
8.21253538e-01 8.24782729e-01 5.37226379e-01 3.63035321e-01
-8.18382859e-01 -6.27995431e-02 2.77185559e-01 7.04922974e-01
-1.42179966e+00 -6.09565020e-01 7.40476489e-01 -7.45223105e-01
9.40322280e-01 4.36594993e-01 1.09635448e+00 1.40100944e+00
2.15406954e-01 1.22600281e+00 9.01803911e-01 -5.92203021e-01
2.83689916e-01 -1.57867208e-01 2.40023330e-01 5.26530027e-01
4.74401377e-02 2.51825958e-01 -8.99828732e-01 -2.19779819e-01
7.55545378e-01 4.95323718e-01 -3.56166542e-01 -2.35794693e-01
-1.32081914e+00 1.65056542e-01 -3.26190472e-01 4.94109899e-01
-7.91740537e-01 1.59657635e-02 3.06112409e-01 2.42258117e-01
3.40538174e-01 -1.02290295e-01 1.52981102e-01 -6.13371611e-01
-1.01897812e+00 3.47241282e-01 9.88835931e-01 7.02403188e-01
4.66294259e-01 -3.03588957e-01 -4.35114503e-01 2.70413578e-01
5.94061203e-02 2.90086836e-01 3.25769514e-01 -8.57285440e-01
9.92369503e-02 6.36771262e-01 5.14425218e-01 -9.69772398e-01
-1.16225548e-01 -1.67627260e-01 -3.68247539e-01 2.72436380e-01
6.66818202e-01 4.85262796e-02 -4.90316749e-01 1.05123198e+00
3.72834831e-01 7.20514834e-01 1.45877227e-01 8.08866918e-01
2.27974355e-01 9.27335322e-01 -2.28825714e-02 -8.77652347e-01
1.31292689e+00 -9.20895457e-01 -8.81488681e-01 -3.85327376e-02
3.87684524e-01 -3.97382557e-01 4.87178862e-01 8.40949059e-01
-9.48731363e-01 -8.37933838e-01 -6.86083794e-01 4.28187966e-01
-3.18359137e-01 2.80451238e-01 3.58463466e-01 4.56184477e-01
-7.24448264e-01 8.24116707e-01 -1.13335919e+00 -7.74585187e-01
1.35084344e-02 3.33966523e-01 -4.55781281e-01 2.64938205e-01
-5.95159829e-01 6.23465478e-01 4.07762796e-01 1.96647450e-01
-1.36202574e+00 -1.91836953e-01 -6.28752828e-01 -6.38892222e-03
5.51582158e-01 -2.06948116e-01 1.26610124e+00 -1.48426676e+00
-1.37369311e+00 7.96224475e-01 -5.73679447e-01 -7.94551075e-01
6.62337363e-01 -6.88740194e-01 -5.97746074e-01 7.86560118e-01
-3.31209660e-01 7.00521469e-02 1.25222862e+00 -8.93151641e-01
-9.38449204e-01 -2.08881110e-01 -1.14784926e-01 4.02534455e-01
-3.67483348e-01 4.22845781e-01 -2.76227683e-01 -4.72768933e-01
-3.13981891e-01 -7.00603187e-01 -2.95748785e-02 -6.53576851e-02
2.17257112e-01 -2.38850996e-01 9.23671365e-01 -8.59017015e-01
1.33327281e+00 -2.12471080e+00 3.22915353e-02 1.68914363e-01
1.23258308e-01 2.91960388e-01 2.82956272e-01 8.63824904e-01
-2.80599177e-01 -5.92063844e-01 1.62290297e-02 -1.95175260e-01
-4.96808499e-01 1.79435953e-01 -3.29953581e-01 5.80796897e-01
3.02627236e-01 5.40682077e-01 -1.01800215e+00 -5.59382856e-01
5.64397454e-01 5.14641225e-01 -8.71716514e-02 7.68162489e-01
4.36013453e-02 5.89731038e-01 -4.07957822e-01 6.42380357e-01
2.68615723e-01 -1.50771867e-02 1.99094057e-01 2.04859123e-01
-4.80716228e-01 -1.57232106e-01 -1.38809061e+00 1.26634037e+00
1.74715608e-01 6.93973958e-01 -2.21584260e-01 -9.79238808e-01
6.56271160e-01 5.75833499e-01 8.44397724e-01 -2.52053469e-01
1.71735635e-04 -1.26205489e-01 -5.02042472e-01 -9.33972299e-01
3.17965537e-01 3.56038868e-01 1.21056437e-01 2.82830298e-01
2.11408764e-01 8.81034315e-01 5.60431004e-01 -1.73679423e-02
1.42390323e+00 5.81341088e-01 8.77682567e-01 1.95481867e-01
9.77432311e-01 -8.76142904e-02 3.55602831e-01 9.20960128e-01
-4.83842939e-01 4.25796092e-01 4.25674945e-01 -7.61737287e-01
-6.67365253e-01 -9.39837337e-01 5.70056617e-01 1.20578134e+00
6.18304014e-02 -3.92014772e-01 -7.75298715e-01 -8.01744401e-01
-4.23187912e-01 2.28131115e-01 -5.57754695e-01 6.47110641e-02
-8.70606482e-01 -4.02178913e-01 3.64067495e-01 6.38767779e-01
4.15920824e-01 -1.11159992e+00 -1.11949444e+00 3.70737851e-01
-1.69279605e-01 -1.19225836e+00 -3.78286570e-01 2.18537226e-02
-9.84554768e-01 -1.25730467e+00 -9.98126864e-01 -5.64538360e-01
6.73862219e-01 2.98015773e-01 8.23131442e-01 8.06662664e-02
-2.78295755e-01 1.21600664e+00 -7.33467162e-01 -8.37516859e-02
-3.65010083e-01 -4.34123129e-01 2.01582029e-01 9.24626112e-01
5.39004624e-01 -6.24889076e-01 -6.96565211e-01 3.93023372e-01
-7.52934337e-01 -1.11516245e-01 6.35315120e-01 4.20933217e-01
3.82460624e-01 1.19052671e-01 -1.23322584e-01 -4.37451541e-01
1.69188261e-01 -3.02664161e-01 -3.98973197e-01 4.89715755e-01
-1.96910709e-01 -2.42892995e-01 5.56227863e-01 -5.63261747e-01
-1.31129646e+00 5.56048453e-01 3.14949341e-02 -4.80323583e-01
-8.92070055e-01 5.65726049e-02 -2.04963565e-01 8.77040625e-02
5.22498369e-01 6.42860234e-01 -4.11851943e-01 -5.92746258e-01
-5.46266623e-02 5.63085675e-01 6.84096754e-01 -4.15507585e-01
2.71552384e-01 7.93661296e-01 -9.24183801e-02 -1.29872310e+00
-3.76653194e-01 -1.15888608e+00 -9.81748641e-01 -1.19085562e+00
8.56991291e-01 -9.69534874e-01 -9.12932694e-01 7.81308115e-01
-1.27202952e+00 -1.32403925e-01 -4.18430358e-01 6.28976047e-01
-5.99241793e-01 7.68396080e-01 -4.90222543e-01 -1.49763203e+00
2.03984547e-02 -4.01130497e-01 1.06778574e+00 4.15090084e-01
-3.46224099e-01 -1.03810859e+00 2.90016562e-01 3.09036523e-01
-2.17758909e-01 4.25194591e-01 -1.70190275e-01 -8.40145528e-01
-6.59435749e-01 -4.47396368e-01 3.07691067e-01 2.87421703e-01
1.22604528e-02 1.88844606e-01 -1.09643638e+00 -3.05281252e-01
1.79729268e-01 3.48055035e-01 5.98420084e-01 3.18356782e-01
7.24036038e-01 -2.82425284e-01 -6.36129558e-01 2.57743657e-01
1.15166223e+00 6.06813550e-01 8.63026440e-01 7.32898638e-02
4.45644826e-01 7.24003613e-01 1.00050807e+00 7.55886614e-01
-1.30033821e-01 5.99405587e-01 9.42738652e-02 1.64413810e-01
4.12328690e-02 -1.92450598e-01 1.14036417e+00 4.70236361e-01
-8.66710305e-01 -3.98975462e-01 -6.47388041e-01 5.62721491e-01
-2.02074051e+00 -1.61712217e+00 -1.98612735e-01 2.50730824e+00
4.55603898e-02 3.76904190e-01 7.20295072e-01 4.44318146e-01
9.43121552e-01 2.19657809e-01 -2.48016976e-02 -1.37624025e-01
1.54772058e-01 -1.25820726e-01 2.21515387e-01 4.12543863e-01
-1.34550273e+00 5.39026499e-01 6.89388037e+00 5.53237975e-01
-8.39717507e-01 2.09215321e-02 1.93499163e-01 -1.35738879e-01
5.74925303e-01 9.72996354e-02 -1.01230741e+00 5.08630991e-01
1.13717353e+00 -1.66883767e-02 1.41169861e-01 8.85653973e-01
6.75133288e-01 -5.28858304e-01 -1.30499184e+00 9.82299864e-01
4.50438708e-01 -8.33739042e-01 1.10309228e-01 6.80233240e-02
2.54531652e-01 -7.47533560e-01 -6.27352238e-01 -8.23012963e-02
-4.86137360e-01 -4.87895370e-01 6.93681002e-01 1.09085321e+00
2.20857382e-01 -6.69144928e-01 3.66590559e-01 6.97161078e-01
-1.54357588e+00 -3.54602784e-01 2.03398913e-01 -6.57262921e-01
3.97303641e-01 5.20472765e-01 -9.85887885e-01 4.41968441e-01
5.62366426e-01 1.13177562e+00 -5.96660674e-01 1.12223983e+00
-1.75075218e-01 6.83449686e-01 -3.31945360e-01 -1.14942186e-01
-1.23302661e-01 -1.31403968e-01 8.70835781e-01 1.54812205e+00
5.25895596e-01 3.24582696e-01 4.15085584e-01 3.87665182e-01
7.44446218e-01 -1.04450278e-01 -9.85702932e-01 -3.05720598e-01
-5.26852310e-02 9.32059705e-01 -1.02800906e+00 -7.17201531e-01
-4.20076311e-01 1.62456441e+00 -2.36337334e-01 1.75264701e-01
-7.55839288e-01 -2.74941087e-01 2.03267336e-01 3.03320557e-01
3.52595925e-01 -4.22041178e-01 4.97525334e-01 -1.33142233e+00
2.64914811e-01 -6.19723082e-01 7.24515736e-01 -1.02673662e+00
-6.20774031e-01 3.90947312e-01 4.26794261e-01 -1.63178813e+00
-5.83023906e-01 -7.46090233e-01 -7.71919847e-01 2.04524264e-01
-8.25505853e-01 -1.03974628e+00 -4.24471110e-01 8.06901276e-01
9.97302890e-01 1.94158219e-02 5.49940348e-01 6.29632175e-02
-4.84046221e-01 -8.52446109e-02 -2.52268374e-01 1.25825763e-01
3.93900752e-01 -1.20056224e+00 -9.12035350e-03 1.39731705e+00
4.77706909e-01 3.24383914e-01 9.97284412e-01 -8.75273347e-01
-1.20503271e+00 -4.02022898e-01 1.13922203e+00 -5.81052184e-01
2.91292578e-01 -3.52895439e-01 -7.08282292e-01 8.13045323e-01
2.73253888e-01 -2.55065739e-01 7.54930019e-01 -3.07408035e-01
2.77428687e-01 -3.02584827e-01 -7.32178628e-01 4.97579694e-01
1.28516555e+00 -5.45620203e-01 -1.13977861e+00 2.91206896e-01
1.71416830e-02 -8.08824822e-02 -5.81571460e-01 2.28436932e-01
7.13685751e-01 -1.30435824e+00 1.01819634e+00 -3.19072545e-01
-2.04014406e-02 -3.68276596e-01 3.88037503e-01 -7.15194762e-01
-2.37649769e-01 -1.02489281e+00 -8.36190462e-01 1.09401417e+00
-3.64269406e-01 -1.28105983e-01 9.72420096e-01 3.64422023e-01
1.06325485e-01 -8.08455870e-02 -6.95090771e-01 -9.02847767e-01
-1.28664017e+00 -4.23080683e-01 2.05125492e-02 4.23992217e-01
2.30250433e-01 -1.57118380e-01 -8.33626747e-01 3.17619026e-01
6.28729224e-01 -5.96347451e-02 9.25186098e-01 -1.08956075e+00
-4.32896435e-01 2.85815876e-02 -8.55291545e-01 -1.25887132e+00
-2.48777777e-01 -7.03765005e-02 1.20952129e-01 -1.04123557e+00
2.45793328e-01 8.33401561e-01 -4.50633079e-01 1.60827890e-01
4.43291552e-02 1.60056911e-02 -5.97337075e-02 2.86530793e-01
-9.68722045e-01 1.28481865e-01 5.69331408e-01 2.64542967e-01
-3.08278978e-01 3.63364637e-01 1.79676324e-01 8.56787622e-01
2.62249857e-01 -3.95145774e-01 -3.34403545e-01 1.82998568e-01
-6.14334680e-02 3.73450369e-01 9.45845127e-01 -1.64020777e+00
4.66383219e-01 -2.09533691e-01 6.05075181e-01 -6.83068931e-01
5.05066693e-01 -1.03533590e+00 3.94576848e-01 7.28969872e-01
-3.35157007e-01 7.73425549e-02 -1.99928448e-01 9.92391109e-01
-3.13400716e-01 -2.19302148e-01 4.01467443e-01 -2.41698757e-01
-1.36552501e+00 -1.69900388e-01 -1.20832741e+00 -4.92626637e-01
1.48405802e+00 -6.95463777e-01 2.60300338e-01 -5.22799730e-01
-1.38341641e+00 -1.71508580e-01 2.68577963e-01 2.07228899e-01
7.72493303e-01 -1.10275817e+00 -4.22392726e-01 3.61070991e-01
1.77406222e-01 -8.33630860e-01 4.65925634e-01 1.05163753e+00
-7.82812059e-01 1.60161778e-01 -3.20710182e-01 -7.26411939e-01
-1.89421940e+00 8.33868146e-01 1.48082301e-01 -3.58229935e-01
-8.30412328e-01 5.56694627e-01 -7.02551752e-02 6.21675253e-01
3.78762931e-01 -3.18122864e-01 -6.70401990e-01 2.13144228e-01
1.01912129e+00 7.92931914e-01 -2.23497435e-01 -7.84606755e-01
-5.71562350e-01 4.52188760e-01 6.62286440e-03 -3.14780653e-01
1.07216465e+00 -2.17427745e-01 1.88317299e-01 6.96059585e-01
8.14226806e-01 -8.38855558e-05 -1.75462449e+00 1.11909941e-01
4.95347083e-01 -7.28437066e-01 -4.55444962e-01 -5.62699139e-01
-4.96439874e-01 8.19229901e-01 7.68967569e-01 4.10692066e-01
1.43152654e+00 -1.70820788e-01 6.04588628e-01 5.16737223e-01
4.03194904e-01 -1.18117332e+00 3.69629562e-01 1.06515370e-01
7.25841224e-01 -8.22611988e-01 2.60324421e-04 -4.53665406e-01
-7.25834668e-01 1.42831802e+00 3.48073691e-01 -2.62476772e-01
4.15412903e-01 1.81812331e-01 -2.25389987e-01 -1.59507111e-01
-7.98994243e-01 -6.46676660e-01 3.08567494e-01 6.19019568e-01
-1.26294866e-02 -2.62051553e-01 -3.05973649e-01 3.35497320e-01
7.43782103e-01 5.19207299e-01 5.46029985e-01 1.49545252e+00
-6.40304029e-01 -7.55131185e-01 -8.77185166e-01 1.21016828e-02
-4.73585188e-01 5.14194250e-01 -7.46647716e-01 5.88304937e-01
4.12912160e-01 8.81106079e-01 1.89587072e-01 -4.29280937e-01
5.38570344e-01 4.97677654e-01 6.28952324e-01 -4.39347893e-01
-4.04102474e-01 3.48032176e-01 2.26877674e-01 -8.95574510e-01
-9.42422807e-01 -1.14702499e+00 -7.53587127e-01 1.03912637e-01
3.90000381e-02 2.67666101e-01 1.52159855e-01 1.21776938e+00
2.79342622e-01 1.93695083e-01 7.56183624e-01 -1.17056096e+00
2.02851128e-02 -5.98906040e-01 -6.75647318e-01 6.03780210e-01
4.28042650e-01 -5.24528742e-01 -5.02697110e-01 8.18477690e-01] | [8.191540718078613, 0.35192158818244934] |
70006695-0241-4887-b2d1-c24b6f8726e1 | a-sentiment-and-emotion-aware-multimodal | null | null | https://aclanthology.org/2022.coling-1.587 | https://aclanthology.org/2022.coling-1.587.pdf | A Sentiment and Emotion Aware Multimodal Multiparty Humor Recognition in Multilingual Conversational Setting | In this paper, we hypothesize that humor is closely related to sentiment and emotions. Also, due to the tremendous growth in multilingual content, there is a great demand for building models and systems that support multilingual information access. To end this, we first extend the recently released Multimodal Multiparty Hindi Humor (M2H2) dataset by adding parallel English utterances corresponding to Hindi utterances and then annotating each utterance with sentiment and emotion classes. We name it Sentiment, Humor, and Emotion aware Multilingual Multimodal Multiparty Dataset (SHEMuD). Therefore, we propose a multitask framework wherein the primary task is humor detection, and the auxiliary tasks are sentiment and emotion identification. We design a multitasking framework wherein we first propose a Context Transformer to capture the deep contextual relationships with the input utterances. We then propose a Sentiment and Emotion aware Embedding (SE-Embedding) to get the overall representation of a particular emotion and sentiment w.r.t. the specific humor situation. Experimental results on the SHEMuD show the efficacy of our approach and shows that multitask learning offers an improvement over the single-task framework for both monolingual (4.86 points in Hindi and 5.9 points in English in F1-score) and multilingual (5.17 points in F1-score) setting. | ['Pushpak Bhattacharyya', 'Asif Ekbal', 'Aseem Arora', 'Gopendra Vikram Singh', 'Dushyant Singh Chauhan'] | null | null | null | null | coling-2022-10 | ['humor-detection'] | ['natural-language-processing'] | [-4.32566077e-01 -2.28129998e-01 2.45816261e-01 -5.69649339e-01
-9.25189078e-01 -5.47204614e-01 6.17021382e-01 -2.04550717e-02
-4.85385686e-01 6.03012264e-01 7.31311023e-01 1.57618746e-01
4.08285260e-01 -3.35851818e-01 -4.28407490e-01 -4.90214229e-01
3.38698149e-01 2.01347932e-01 -3.43280703e-01 -7.22186625e-01
1.32365331e-01 -2.36974269e-01 -1.16140509e+00 6.43832147e-01
7.17709243e-01 8.98313642e-01 1.30900785e-01 6.57282770e-01
1.04961395e-01 1.40691710e+00 -3.71854872e-01 -8.30800056e-01
-2.13568538e-01 -4.65503305e-01 -9.69651759e-01 9.08737704e-02
1.41718552e-01 -1.44086242e-01 -2.62456387e-01 8.22051108e-01
8.83675337e-01 4.01309311e-01 4.54335123e-01 -1.30271208e+00
-7.65178382e-01 8.12743485e-01 -6.05004132e-01 -2.74326384e-01
3.53994787e-01 -2.00515196e-01 1.22790694e+00 -1.48697329e+00
4.92526203e-01 1.30933750e+00 5.75455666e-01 5.11385918e-01
-7.02522576e-01 -5.00983655e-01 -2.78767884e-01 5.08263350e-01
-1.10642397e+00 -4.34605777e-01 1.13444209e+00 -3.96415353e-01
6.68083072e-01 1.31689057e-01 9.20992941e-02 1.34973133e+00
-8.19339044e-03 1.33578324e+00 1.27731955e+00 -3.68038565e-01
-1.25390574e-01 6.50913417e-01 2.49685392e-01 5.22942901e-01
-7.41066277e-01 -6.18331492e-01 -9.29278791e-01 -2.85176374e-03
6.12339936e-02 -1.27018020e-01 -1.88701987e-01 2.15535581e-01
-1.15281343e+00 1.24258113e+00 2.24259272e-01 2.45686978e-01
-2.70086139e-01 -1.33999363e-01 9.31074083e-01 4.71683890e-01
3.54170710e-01 1.58511654e-01 -3.57623130e-01 -2.39803806e-01
-6.86588764e-01 2.85642803e-01 8.20914328e-01 8.82721364e-01
7.51087844e-01 -7.14927241e-02 -7.70434588e-02 1.54183638e+00
3.05282116e-01 6.87730670e-01 6.87134564e-01 -7.17516482e-01
4.46365029e-01 4.88511682e-01 -5.74753620e-02 -1.21033597e+00
-6.68921113e-01 -1.41317531e-01 -7.07897186e-01 -2.85709232e-01
3.34828235e-02 -3.63131344e-01 -2.02395827e-01 1.90195870e+00
2.36377671e-01 -5.00855148e-01 4.09742773e-01 1.08727765e+00
1.09233749e+00 8.78965914e-01 -1.53587051e-02 -1.71589389e-01
1.51883364e+00 -1.37247646e+00 -9.26426411e-01 -2.71466076e-01
6.14085793e-01 -1.28318608e+00 1.49974060e+00 2.92599976e-01
-8.86748374e-01 -3.61418277e-01 -8.73792470e-01 -6.11273766e-01
-4.53453690e-01 2.01759145e-01 2.40798816e-01 4.13620919e-01
-5.57890296e-01 -1.65742159e-01 -1.25002608e-01 -2.64273107e-01
-1.91826805e-01 2.88825110e-03 -3.52872163e-01 -4.25701551e-02
-1.67565107e+00 1.00231695e+00 1.09247260e-01 2.43181922e-02
-8.59887540e-01 -2.62474567e-01 -1.10665131e+00 -2.40436345e-01
2.19714846e-02 -1.60831720e-01 1.17761278e+00 -1.04161537e+00
-1.44043195e+00 8.74920666e-01 -2.06004858e-01 -1.85453713e-01
1.36508361e-01 -2.69796193e-01 -4.42424476e-01 -1.08740442e-02
1.25694513e-01 5.39182484e-01 6.96223617e-01 -1.31116748e+00
-6.56493247e-01 -3.98993582e-01 1.10370740e-01 4.82505113e-01
-9.14985120e-01 4.18467671e-01 -3.21593404e-01 -3.84438694e-01
-3.25154811e-01 -1.08834434e+00 2.93636143e-01 -9.27790999e-01
-3.98346603e-01 -2.28802040e-01 1.00063503e+00 -9.75462973e-01
1.31802845e+00 -2.15109205e+00 4.43034679e-01 -1.94709629e-01
1.55061245e-01 -2.22242489e-01 -1.78873017e-01 5.80147684e-01
1.99922934e-01 -2.83928335e-01 -1.49872109e-01 -6.71632051e-01
4.58115637e-01 1.54881403e-01 -3.05310398e-01 3.26029599e-01
-1.18958615e-01 1.03449941e+00 -7.17003942e-01 -5.41613758e-01
2.41125803e-02 6.20544851e-01 -5.61690450e-01 4.03001279e-01
2.27888107e-01 5.48071444e-01 -1.19854681e-01 7.90658176e-01
3.57869804e-01 -1.05763219e-01 2.72670507e-01 -3.08822781e-01
-1.84778407e-01 3.36680040e-02 -6.61907911e-01 1.60058439e+00
-1.07371044e+00 7.26755500e-01 5.88341318e-02 -4.93441373e-01
1.07169545e+00 5.98246515e-01 4.16892797e-01 -7.51654804e-01
4.61059421e-01 2.92093277e-01 -1.54810101e-01 -7.30218709e-01
8.97228897e-01 -6.32479548e-01 -9.01876748e-01 4.29348499e-01
3.26939851e-01 -6.68067709e-02 -7.08468109e-02 2.89180964e-01
5.79859436e-01 -3.03038239e-01 3.44035059e-01 -1.58313334e-01
8.69425297e-01 -3.63154650e-01 4.95561749e-01 1.81992680e-01
-4.81131494e-01 4.23580945e-01 5.64259768e-01 -3.49359751e-01
-9.58775759e-01 -4.43225265e-01 -1.22111194e-01 1.73947227e+00
-1.23344630e-01 -3.49088550e-01 -5.62965333e-01 -6.17561877e-01
-2.23747358e-01 7.08238304e-01 -6.69052541e-01 1.25916272e-01
-4.04054552e-01 -8.73800755e-01 5.27027726e-01 2.75269717e-01
5.18467546e-01 -1.25121224e+00 -5.31651974e-01 1.22636676e-01
-9.80108082e-01 -1.37527740e+00 -6.35408938e-01 2.73946524e-01
-1.97618734e-02 -6.44506514e-01 -6.18369341e-01 -9.36170340e-01
1.67288110e-02 6.02948964e-02 1.06583691e+00 -3.35575014e-01
2.45568186e-01 3.59706879e-01 -7.16675639e-01 -2.40939170e-01
-1.88325614e-01 1.49324968e-01 7.07758963e-02 2.47519255e-01
5.52020133e-01 -3.95204633e-01 -3.15506816e-01 1.26239479e-01
-8.58091950e-01 8.13889578e-02 3.95097315e-01 1.06422257e+00
1.75579637e-01 -2.28299096e-01 8.60893846e-01 -8.07123482e-01
7.94426084e-01 -9.39306080e-01 1.05817162e-01 1.19147897e-01
-1.63084492e-01 -3.18184912e-01 7.64202476e-01 -2.81003207e-01
-1.18893838e+00 3.25085409e-02 -1.26984671e-01 -2.06501797e-01
-5.60764261e-02 8.61594677e-01 -2.35298827e-01 1.41378880e-01
3.88080269e-01 3.30793150e-02 -3.71151626e-01 -3.30692917e-01
5.98350763e-01 1.23126650e+00 6.18534684e-01 -7.23670304e-01
4.06941473e-01 2.25728065e-01 -3.17101985e-01 -9.05964673e-01
-1.18743229e+00 -6.22870088e-01 -5.26928127e-01 -5.85375965e-01
1.08548462e+00 -1.12076950e+00 -9.65542674e-01 5.67116261e-01
-1.36257458e+00 -2.27554157e-01 4.01675284e-01 2.85588503e-01
-6.59773171e-01 3.47777963e-01 -8.95273805e-01 -1.05911756e+00
-5.34614265e-01 -1.19061029e+00 9.29201245e-01 1.02391625e-02
-3.68567795e-01 -1.22620821e+00 3.49636674e-01 9.33727384e-01
2.61864632e-01 2.11529389e-01 1.09060860e+00 -8.46028388e-01
2.57984906e-01 -1.56074241e-01 -3.18587571e-01 4.87142950e-01
-1.69272646e-01 -5.20635307e-01 -1.07410061e+00 -1.06855989e-01
2.81749219e-01 -1.31660175e+00 6.21552408e-01 -5.92436232e-02
5.35268128e-01 -5.25597811e-01 5.57680428e-01 1.28966346e-01
1.37184680e+00 -4.53712121e-02 4.88364369e-01 4.54336166e-01
8.54632258e-01 1.04032850e+00 6.02101803e-01 8.04046094e-01
1.16440475e+00 6.80836380e-01 1.87443763e-01 -7.36482665e-02
1.33188665e-01 -3.26223522e-01 8.49139869e-01 1.59068155e+00
3.12199384e-01 6.36574253e-02 -1.04890418e+00 7.86422014e-01
-2.21550989e+00 -9.51537073e-01 -4.12168533e-01 1.81439054e+00
1.13644683e+00 -5.47607005e-01 2.94204205e-01 -9.68375579e-02
4.73000169e-01 1.50921330e-01 -2.67289877e-01 -6.61017954e-01
-4.93063301e-01 -2.96192318e-01 -1.53076937e-02 9.24542785e-01
-1.15417135e+00 1.22288597e+00 5.16291142e+00 6.46295547e-01
-1.18249202e+00 5.64913690e-01 5.85281610e-01 -2.10091472e-01
-3.41414809e-01 -2.54631191e-01 -7.24803686e-01 3.58611226e-01
1.01353776e+00 -1.75355136e-01 5.84412813e-01 8.37869942e-01
2.56705374e-01 1.57354981e-01 -7.94499815e-01 1.17418277e+00
7.23830044e-01 -5.94789624e-01 -1.43843353e-01 -2.07621664e-01
9.66606140e-01 9.18372255e-03 2.08095655e-01 7.64710665e-01
-2.48786267e-02 -9.24709022e-01 9.84991550e-01 2.47600287e-01
4.90185440e-01 -1.13317716e+00 1.11620736e+00 3.39416265e-01
-8.58240843e-01 -1.69481263e-01 -2.22682551e-01 -4.71259393e-02
2.98367172e-01 2.29498461e-01 -5.95537126e-01 2.90920377e-01
6.54767931e-01 6.55486584e-01 -4.55967396e-01 2.98394173e-01
-3.57641309e-01 4.02399480e-01 1.65357053e-01 -1.23301782e-01
4.09687907e-01 -2.34874040e-02 3.72246265e-01 1.65500486e+00
5.57798408e-02 -1.64279565e-02 3.60240102e-01 3.60993147e-01
-3.40661317e-01 7.89412618e-01 -4.99787033e-01 1.00742415e-01
2.01660067e-01 1.74447823e+00 -5.75998798e-02 -1.87798515e-01
-7.52628863e-01 1.39144552e+00 4.53950286e-01 2.85514295e-01
-9.72929358e-01 -5.13160408e-01 2.64153838e-01 -5.97235024e-01
3.36236507e-02 -1.43761098e-01 -3.32073689e-01 -1.37957299e+00
-2.40053937e-01 -1.19032252e+00 4.27712947e-01 -6.55012071e-01
-1.42582715e+00 7.82715201e-01 -4.27255571e-01 -7.73119628e-01
-2.08612785e-01 -5.77731073e-01 -3.76324803e-01 7.21291304e-01
-1.63424540e+00 -1.70868015e+00 -1.55261427e-01 9.15977836e-01
8.39490473e-01 -2.28492215e-01 7.15450644e-01 5.73085845e-01
-5.47591984e-01 6.60296023e-01 -1.08492151e-01 1.61249742e-01
1.13712335e+00 -1.32483959e+00 -4.77414817e-01 4.55800533e-01
-1.56576950e-02 5.09296775e-01 8.79395425e-01 -2.83771008e-01
-1.38343215e+00 -9.07480180e-01 1.53746641e+00 -6.50866687e-01
1.09372866e+00 -4.11053956e-01 -7.83354938e-01 6.67049587e-01
7.58683741e-01 -6.37746155e-01 1.14431703e+00 4.73493218e-01
-5.80196321e-01 1.23093233e-01 -1.02123618e+00 5.04939318e-01
1.88401848e-01 -9.08100903e-01 -4.48420972e-01 3.96263897e-01
6.45814598e-01 -1.18086770e-01 -9.54550326e-01 3.84130567e-01
7.33467162e-01 -8.68189514e-01 4.13464636e-01 -6.11796856e-01
1.15429485e+00 -8.59986469e-02 -9.63000834e-01 -1.19534302e+00
-1.52696863e-01 -4.83157068e-01 3.65573317e-02 1.41244805e+00
4.80018705e-01 -3.88161354e-02 2.63820231e-01 5.30757606e-01
-1.29223466e-01 -6.32857680e-01 -7.30906188e-01 -2.63664842e-01
4.42380548e-01 -5.10547101e-01 1.57527730e-01 1.41631377e+00
7.44968355e-01 1.11452925e+00 -1.23371470e+00 2.79626045e-02
5.32067120e-01 2.85385728e-01 8.37268829e-01 -6.00432098e-01
-2.24732652e-01 -3.09817880e-01 -8.05882439e-02 -7.54952192e-01
6.29248381e-01 -9.40324605e-01 1.59534782e-01 -1.01197326e+00
7.59549916e-01 1.67821884e-01 -4.35075700e-01 6.84901297e-01
-3.14163774e-01 4.98469353e-01 4.16274488e-01 2.17095062e-01
-9.19375718e-01 9.44257140e-01 8.16052556e-01 -3.14547382e-02
-1.11486189e-01 -5.77141464e-01 -7.44750559e-01 8.01809669e-01
7.67341137e-01 -1.52034506e-01 -2.06068218e-01 -3.88580918e-01
5.25004506e-01 2.71060616e-01 2.42856443e-01 -6.19956315e-01
1.45562917e-01 -1.88118331e-02 1.59421302e-02 -6.18705809e-01
4.99965042e-01 -7.02765107e-01 -4.75567400e-01 -7.92769045e-02
-5.35334527e-01 3.66429575e-02 8.94010440e-02 1.33185670e-01
-6.42990172e-01 -2.04921067e-01 8.67561519e-01 5.21549322e-02
-7.89847970e-01 -8.85428786e-02 -4.30434167e-01 6.31943112e-03
6.93469584e-01 6.02010906e-01 -2.99364299e-01 -8.92537594e-01
-6.59705460e-01 4.55512077e-01 3.20215732e-01 7.03988910e-01
6.83796942e-01 -1.65088618e+00 -9.57078218e-01 -2.27223948e-01
4.69976187e-01 -8.61547291e-01 5.62658906e-01 1.09342420e+00
9.93748903e-02 4.59054768e-01 -2.42829874e-01 -1.49377033e-01
-1.42817450e+00 2.93345720e-01 1.04759201e-01 -1.65069178e-01
4.62178420e-03 6.90363824e-01 3.33536208e-01 -1.07695186e+00
1.23833857e-01 5.50765991e-01 -4.04662430e-01 5.18008471e-01
4.98398602e-01 4.32135552e-01 -1.56702131e-01 -1.38687885e+00
-3.06748539e-01 2.99991161e-01 -7.72200152e-02 -5.60970664e-01
1.27922809e+00 -5.53131521e-01 -5.23648679e-01 1.13899660e+00
1.58467793e+00 3.24416131e-01 -5.78070045e-01 -2.94727206e-01
3.28792334e-02 -1.85375720e-01 8.05583149e-02 -1.04197729e+00
-7.99703121e-01 1.10376704e+00 2.77659327e-01 4.50898521e-02
1.10240531e+00 3.36520635e-02 1.25304949e+00 4.33496714e-01
1.99838236e-01 -1.41615009e+00 3.88530880e-01 1.10632968e+00
1.16051924e+00 -1.42334211e+00 -4.62615341e-01 1.08843975e-01
-1.54657030e+00 1.11008477e+00 6.36566341e-01 2.89427668e-01
3.41235697e-01 -5.59919737e-02 5.55399179e-01 -3.52466315e-01
-8.50790501e-01 -6.32568225e-02 2.75130719e-01 6.86770752e-02
1.00568402e+00 2.03237340e-01 -4.75320727e-01 1.10450685e+00
-4.40236032e-01 -3.75214636e-01 6.06153727e-01 4.85934168e-01
-6.27838850e-01 -9.09671128e-01 -3.36630583e-01 -2.69406229e-01
-5.56493819e-01 -2.86953688e-01 -6.55103207e-01 4.66443360e-01
4.35450561e-02 1.27064991e+00 -5.67504823e-01 -7.84621060e-01
3.41257036e-01 4.61225599e-01 1.04935162e-01 -3.00128162e-01
-7.75094092e-01 2.33458757e-01 2.53183663e-01 -2.47833475e-01
-3.58885795e-01 -5.40653229e-01 -1.15185869e+00 -4.25823092e-01
-3.45235243e-02 1.42010942e-01 8.15234840e-01 9.30110097e-01
-2.47503794e-03 1.25053689e-01 1.13513052e+00 -6.31404042e-01
-4.18823242e-01 -1.12398708e+00 -5.84354579e-01 7.01915145e-01
2.99671799e-01 -2.83669204e-01 -3.74084622e-01 1.20891616e-01] | [13.064263343811035, 5.36147403717041] |
22867581-4bf8-402d-abc7-2e28ebc7f8ed | evaluating-multi-focus-natural-language | null | null | https://aclanthology.org/L12-1468 | https://aclanthology.org/L12-1468.pdf | Evaluating Multi-focus Natural Language Queries over Data Services | Natural language interfaces to data services will be a key technology to guarantee access to huge data repositories in an effortless way. This involves solving the complex problem of recognizing a relevant service or service composition given an ambiguous, potentially ungrammatical natural language question. As a first step toward this goal, we study methods for identifying the salient terms (or foci) in natural language questions, classifying the latter according to a taxonomy of services and extracting additional relevant information in order to route them to suitable data services. While current approaches deal with single-focus (and therefore single-domain) questions, we investigate multi-focus questions in the aim of supporting conjunctive queries over the data services they refer to. Since such complex queries have seldom been studied in the literature, we have collected an ad-hoc dataset, SeCo-600, containing 600 multi-domain queries annotated with a number of linguistic and pragmatic features. Our experiments with the dataset have allowed us to reach very high accuracy in different phases of query analysis, especially when adopting machine learning methods. | ['Pietro La Torre', 'Vincenzo Guerrisi', 'Silvia Quarteroni'] | 2012-05-01 | null | null | null | lrec-2012-5 | ['service-composition'] | ['miscellaneous'] | [ 2.52988815e-01 3.16940665e-01 -4.92736511e-02 -5.80596864e-01
-9.12659347e-01 -8.00279915e-01 1.05291498e+00 4.35253501e-01
-4.84664857e-01 4.06983286e-01 5.87798715e-01 -6.50664687e-01
-4.57510263e-01 -6.26981020e-01 -1.50222585e-01 -1.35338411e-01
2.46508762e-01 9.40848887e-01 7.64160395e-01 -6.09955788e-01
5.20782530e-01 5.41158795e-01 -1.88168728e+00 5.10858595e-01
7.02003419e-01 8.28888357e-01 4.14482087e-01 2.18499318e-01
-9.06253457e-01 4.32501733e-01 -3.88208002e-01 -6.07110798e-01
-1.08039007e-01 -3.25252473e-01 -1.45556462e+00 1.15944691e-01
-5.06361276e-02 2.18993813e-01 3.26033413e-01 1.18837810e+00
1.28826529e-01 -1.18386030e-01 4.07286286e-01 -1.25748312e+00
-2.21812144e-01 6.39887393e-01 1.44503072e-01 2.00078279e-01
1.07388973e+00 -1.03627905e-01 1.48034680e+00 -6.74191535e-01
8.17835808e-01 1.28295076e+00 -4.67300601e-02 6.99980438e-01
-1.01180279e+00 -7.49749243e-02 1.70098394e-01 2.46564448e-01
-1.20493472e+00 -5.75881600e-01 6.82999194e-01 -3.04318249e-01
1.05261505e+00 7.18805611e-01 1.34527802e-01 7.83253670e-01
-3.79495233e-01 5.69152594e-01 9.01472747e-01 -8.35071743e-01
1.29387617e-01 6.95004106e-01 3.37691784e-01 2.93627262e-01
3.58730495e-01 -4.00527626e-01 -3.74132067e-01 -4.90088940e-01
1.23903938e-01 -3.26081932e-01 -2.40183711e-01 -1.95638523e-01
-1.07298625e+00 6.57406867e-01 2.59582209e-03 1.05533028e+00
-4.52993333e-01 -4.22928810e-01 3.87991786e-01 6.12138510e-01
4.34908345e-02 8.43165815e-01 -8.63283455e-01 -2.70003498e-01
-4.11130637e-01 6.10519886e-01 1.52889800e+00 1.26791465e+00
8.08899939e-01 -5.77556610e-01 1.98284656e-01 6.89554036e-01
3.52885932e-01 1.83936238e-01 4.01868016e-01 -8.79438579e-01
5.14280319e-01 1.24603343e+00 6.27797604e-01 -9.23086166e-01
-3.09020877e-01 5.83567284e-02 1.25127696e-02 -2.89813340e-01
6.70238137e-01 3.03366899e-01 -2.59499609e-01 1.43090379e+00
5.29085934e-01 -8.36099863e-01 2.05616653e-01 9.57624793e-01
7.91340053e-01 4.59923506e-01 1.75116528e-02 -4.83117819e-01
1.85680509e+00 -3.26132089e-01 -7.04105437e-01 -2.52380610e-01
5.81071854e-01 -8.85641754e-01 1.63107181e+00 3.61485243e-01
-8.43810618e-01 7.00663254e-02 -3.98274392e-01 -8.69729593e-02
-7.64699399e-01 -3.64103079e-01 5.15102208e-01 4.27430600e-01
-8.20784867e-01 -3.01308185e-02 -2.39242703e-01 -9.14659739e-01
-2.71290213e-01 1.21033698e-01 -3.18516344e-01 -1.41484380e-01
-1.27615857e+00 8.75501037e-01 2.14487612e-01 -2.15833157e-01
-3.44035864e-01 -1.27314985e-01 -6.17850006e-01 1.40210360e-01
9.79188025e-01 -3.46089125e-01 1.52709115e+00 -8.57610464e-01
-9.15853560e-01 1.15813231e+00 -5.30577958e-01 -2.48432666e-01
1.07038014e-01 1.23668641e-01 -8.40470731e-01 1.66701764e-01
3.66762996e-01 2.61263520e-01 5.03033340e-01 -1.28960741e+00
-1.04703903e+00 -6.74173892e-01 4.90251303e-01 4.16423306e-02
-3.21900994e-01 9.47389960e-01 -4.10178810e-01 -2.42190197e-01
2.57895023e-01 -4.71712291e-01 -5.38591556e-02 -3.53209734e-01
-1.23167120e-01 -7.27781236e-01 4.85123068e-01 -5.17994463e-01
1.38596904e+00 -1.73452437e+00 1.34466901e-01 4.49967720e-02
2.89413240e-03 6.00776672e-02 1.51849329e-01 7.84324229e-01
1.33320853e-01 3.07564646e-01 -2.28815049e-01 1.96962461e-01
4.68790889e-01 4.86720026e-01 -4.61378396e-01 -9.61457267e-02
2.41409197e-01 7.29811013e-01 -7.94435203e-01 -7.79327393e-01
-1.63805932e-01 3.08514317e-03 -6.19705141e-01 3.01037461e-01
-8.63063693e-01 2.02062711e-01 -7.73022652e-01 9.13211942e-01
4.75749969e-02 -1.38981521e-01 3.07340264e-01 -7.75742680e-02
-1.82839349e-01 8.22395444e-01 -1.30174232e+00 1.32091391e+00
-5.93885005e-01 1.59840658e-01 2.20369086e-01 -9.76237237e-01
8.22721541e-01 4.91726607e-01 6.48724735e-01 -8.97081077e-01
-1.20026603e-01 7.32480586e-01 -1.37539759e-01 -1.03400123e+00
5.57998776e-01 -2.52252340e-01 -5.29176652e-01 5.49833775e-01
-1.32629588e-01 -6.84821904e-02 4.75829929e-01 7.14461878e-02
9.84593868e-01 4.40783873e-02 4.64109868e-01 -4.26321000e-01
9.67409611e-01 3.98798168e-01 2.36534297e-01 5.48473835e-01
-2.13969294e-02 1.97016582e-01 6.47946060e-01 -4.75349814e-01
-6.65484250e-01 -5.49612224e-01 2.75488999e-02 1.33676398e+00
2.37817883e-01 -6.83313191e-01 -5.80856740e-01 -7.40277231e-01
-1.65727898e-01 8.89613152e-01 -1.20743312e-01 1.27530292e-01
-6.62929535e-01 -2.86295980e-01 2.86354661e-01 -6.34939447e-02
-1.51351362e-01 -1.40843356e+00 -9.44389164e-01 3.91611129e-01
-3.73796999e-01 -1.39415944e+00 -2.31159076e-01 8.49070698e-02
-4.86428738e-01 -1.38587224e+00 -1.49427190e-01 -6.78144693e-01
5.23470521e-01 3.66989709e-02 1.53905141e+00 3.33869398e-01
-5.72392680e-02 4.88341808e-01 -6.80080891e-01 -3.29052150e-01
-5.87916553e-01 2.10503504e-01 -3.08659971e-01 -3.32418308e-02
1.02716660e+00 -4.50887650e-01 -2.46621892e-01 5.60780227e-01
-1.39577031e+00 -4.32579219e-01 3.39860827e-01 3.61106396e-01
2.36874953e-01 -1.52041599e-01 5.71147382e-01 -8.01779509e-01
1.10324264e+00 -5.43958902e-01 -8.26953292e-01 4.85990822e-01
-8.09626222e-01 3.63274693e-01 5.57321906e-01 -2.75304347e-01
-9.90100503e-01 -1.82097107e-01 -3.69807988e-01 5.38128197e-01
-6.10576451e-01 6.62623644e-01 -6.59050763e-01 2.08765611e-01
7.43337989e-01 2.74033517e-01 -2.85321295e-01 -9.07726765e-01
4.14195925e-01 9.83798206e-01 2.87535697e-01 -9.88296509e-01
6.51315510e-01 2.41533101e-01 -2.06724524e-01 -7.92062581e-01
-5.27054250e-01 -9.72680867e-01 -4.31246161e-01 -1.49440750e-01
5.36179841e-01 -9.00109410e-02 -8.90863001e-01 -1.98719040e-01
-1.11760736e+00 2.58383542e-01 -3.41013700e-01 8.56086961e-04
-5.71868122e-01 3.69542181e-01 -9.74962637e-02 -9.65722620e-01
-1.93650380e-01 -1.18300760e+00 1.25590479e+00 -1.20244645e-01
-5.07447004e-01 -6.46928549e-01 -1.74821183e-01 4.87974912e-01
6.69882894e-01 -2.47802883e-01 1.32432854e+00 -1.23533428e+00
-4.74181920e-01 -3.45942587e-01 -1.66429982e-01 -1.56459570e-01
3.77425343e-01 -3.66678596e-01 -5.44870615e-01 1.08270250e-01
2.40100071e-01 -1.22845985e-01 1.60937533e-01 -4.71732110e-01
7.54824281e-01 -6.45173132e-01 -1.39916658e-01 -1.77681550e-01
1.50732136e+00 2.30791152e-01 3.97773474e-01 4.93254006e-01
5.32616302e-02 1.10308719e+00 8.31915557e-01 2.55791396e-01
6.02398038e-01 9.11247432e-01 4.04232770e-01 5.39425731e-01
1.75813779e-01 -1.74887180e-01 -5.19793741e-02 4.54968244e-01
4.37960476e-01 -6.08273558e-02 -1.27367508e+00 7.46878088e-01
-1.60978925e+00 -6.93558812e-01 -1.82463363e-01 2.00387311e+00
9.87134993e-01 1.70704082e-01 3.21733892e-01 2.01122612e-01
3.61842990e-01 -5.36180884e-02 -1.60494998e-01 -2.94610411e-01
1.35715961e-01 -1.55806124e-01 -1.09064803e-01 5.70966244e-01
-5.27039111e-01 9.36690569e-01 5.52117300e+00 3.41295660e-01
-9.59346294e-01 -1.59501575e-03 -1.05290852e-01 3.81362557e-01
-9.11017776e-01 3.41998756e-01 -9.72924650e-01 3.52965236e-01
9.65452909e-01 -3.10702741e-01 6.05247676e-01 5.45429707e-01
2.73011267e-01 -2.52459973e-01 -1.22726774e+00 6.50152326e-01
-1.09702572e-02 -1.09429169e+00 3.41400743e-01 5.79652842e-03
-8.79699364e-02 -2.20803857e-01 -6.20013595e-01 1.61741689e-01
-6.73364699e-02 -6.67731643e-01 7.10630774e-01 6.26435399e-01
3.90519381e-01 -3.38896602e-01 5.58366120e-01 6.97182238e-01
-8.78578722e-01 -2.55116910e-01 7.41083873e-03 8.78712721e-03
2.11835802e-01 2.28415176e-01 -7.56536186e-01 6.78138256e-01
6.95422590e-01 -5.57543822e-02 -3.70210350e-01 9.37153041e-01
8.48090947e-02 2.65691578e-01 -4.56089079e-01 -6.02249503e-01
1.43727720e-01 -2.71690160e-01 7.27557600e-01 1.10847425e+00
9.08174440e-02 1.15213729e-01 4.28490341e-02 8.89599979e-01
1.48780316e-01 4.79306668e-01 -5.40104032e-01 -2.73286164e-01
2.68303871e-01 1.04446685e+00 -6.98235154e-01 -2.69145787e-01
-7.17259943e-01 5.06825626e-01 -5.72528616e-02 3.98463935e-01
-2.15496302e-01 -4.52818364e-01 6.08933210e-01 5.32841265e-01
-1.45319164e-01 -2.25433201e-01 1.15148515e-01 -1.27427554e+00
7.74772465e-01 -1.29908597e+00 7.54356563e-01 -7.26345062e-01
-1.17209303e+00 8.08515072e-01 3.40141833e-01 -1.00493336e+00
-6.61549389e-01 -4.85948175e-01 1.97482910e-02 9.10980880e-01
-1.71202302e+00 -1.05918670e+00 -9.32487994e-02 6.25177979e-01
4.80674922e-01 1.96103416e-02 9.83819127e-01 5.74568748e-01
-3.01600643e-03 -6.54762685e-02 -6.45723581e-01 -2.36918524e-01
5.49520612e-01 -1.12346387e+00 2.03668237e-01 8.02701175e-01
3.26141000e-01 9.48197961e-01 1.11246169e+00 -4.11886126e-01
-1.74180186e+00 -4.27541316e-01 1.87981915e+00 -3.46363485e-01
9.08276737e-01 -3.92572790e-01 -1.15617907e+00 5.07906377e-01
1.04776151e-01 -2.16912925e-01 5.43748498e-01 2.55812239e-02
-2.81730920e-01 -3.35012704e-01 -1.27902210e+00 5.41057944e-01
1.11337268e+00 -6.67858481e-01 -1.13582742e+00 4.59514052e-01
9.70086634e-01 1.11271068e-01 -6.79283619e-01 3.92025858e-01
1.52492866e-01 -6.71604514e-01 5.72399795e-01 -1.18365836e+00
-1.47894084e-01 -4.65979695e-01 -5.83261609e-01 -6.90200448e-01
4.08166826e-01 -7.95755267e-01 1.38519391e-01 1.42404056e+00
7.45395839e-01 -6.04247630e-01 5.56049109e-01 1.07525778e+00
-2.58618444e-02 -5.77715695e-01 -9.74710941e-01 -5.09348810e-01
-2.49582723e-01 -6.61990821e-01 1.03268766e+00 7.99469590e-01
4.13637578e-01 2.21776769e-01 2.83685952e-01 8.24187100e-02
2.31139868e-01 5.59244812e-01 6.01282597e-01 -1.50153244e+00
-5.78445941e-02 -5.79454958e-01 -2.02492714e-01 -9.64173317e-01
2.10241735e-01 -7.80299067e-01 7.89861158e-02 -1.69205606e+00
-2.44325757e-01 -6.14478171e-01 1.48796782e-01 4.53375489e-01
1.45704538e-01 -3.87167394e-01 -1.54534474e-01 2.11943820e-01
-6.62573218e-01 1.61004722e-01 8.27087998e-01 9.71964076e-02
-1.37532443e-01 4.02341783e-01 -1.19431841e+00 5.62899590e-01
5.23132920e-01 -5.21299660e-01 -2.90218145e-01 -3.94808978e-01
6.19268119e-01 1.28183782e-01 1.70716733e-01 -4.48600799e-01
4.50695723e-01 -6.06382251e-01 -7.16669261e-01 -7.49076530e-02
5.12024090e-02 -1.51054943e+00 2.94146966e-02 3.01952869e-01
-5.47255754e-01 1.69021577e-01 -2.63155073e-01 1.18302919e-01
-4.75231409e-01 -6.79360330e-01 3.65969896e-01 -4.33766782e-01
-8.54706228e-01 7.22274259e-02 -4.41859663e-01 4.57719713e-01
8.12109888e-01 -4.03688522e-03 -8.17650277e-03 -3.41271043e-01
-7.31460333e-01 2.45378852e-01 5.78782499e-01 9.17838812e-01
5.03513634e-01 -9.31275845e-01 -3.54549289e-01 1.64151311e-01
6.93236113e-01 -2.23519459e-01 -6.62658691e-01 4.52255189e-01
-3.52543443e-01 8.55109632e-01 5.27114198e-02 -3.48604053e-01
-1.04081571e+00 7.48181343e-01 4.52394098e-01 -1.51482644e-02
-3.61251324e-01 3.15342367e-01 -4.75017339e-01 -6.50474489e-01
5.34517407e-01 -5.29751599e-01 -5.14690638e-01 2.82380462e-01
5.25663853e-01 -1.37040779e-01 4.30030793e-01 -8.17440212e-01
-6.02416098e-01 2.67879188e-01 2.75300592e-01 -3.81833315e-01
1.30019104e+00 -4.17706370e-01 -5.99478543e-01 3.79838884e-01
1.03278732e+00 6.18663691e-02 -3.15417349e-01 -6.47565484e-01
1.24962795e+00 -5.77672303e-01 -5.22748232e-01 -7.53279984e-01
-5.16544342e-01 4.40899909e-01 2.67638490e-02 1.13971555e+00
1.25418723e+00 6.35954499e-01 6.19670033e-01 5.44934452e-01
8.21436167e-01 -8.20704937e-01 -2.15859771e-01 4.70007360e-01
1.06382525e+00 -1.11023104e+00 -4.84482020e-01 -4.45442587e-01
-3.79107028e-01 1.02931917e+00 4.20082033e-01 5.92008471e-01
5.08012652e-01 1.48883387e-01 1.50463700e-01 -6.78826511e-01
-8.87526929e-01 -5.98174512e-01 1.13150053e-01 3.23684692e-01
3.86728466e-01 -2.72106200e-01 -9.05232191e-01 5.36574602e-01
-6.78711981e-02 3.28820460e-02 3.12809289e-01 1.16707206e+00
-6.19962692e-01 -1.62175083e+00 -1.98981076e-01 4.47642565e-01
-7.68324614e-01 2.63510048e-02 -7.44396031e-01 6.74380481e-01
-2.34275311e-01 1.30058992e+00 -2.63800949e-01 1.56853590e-02
8.27828586e-01 2.91086197e-01 1.76124826e-01 -7.22874463e-01
-6.77871168e-01 -1.27073914e-01 5.97952962e-01 -5.60521245e-01
-5.39693415e-01 -7.49691010e-01 -1.20290172e+00 2.92227596e-01
-3.19155902e-01 7.47719824e-01 9.52502549e-01 1.29133976e+00
2.70884424e-01 -3.92268710e-02 4.78151530e-01 -1.89655617e-01
-7.46866465e-01 -6.81331456e-01 -2.06411868e-01 8.02616835e-01
2.49395654e-01 -3.57249379e-01 -3.41056049e-01 -1.76614877e-02] | [9.883530616760254, 7.964869499206543] |
26e237d2-2e8a-4eca-993e-c5438bf65a64 | a-method-of-generating-measurable-panoramic | 2010.14270 | null | https://arxiv.org/abs/2010.14270v1 | https://arxiv.org/pdf/2010.14270v1.pdf | A Method of Generating Measurable Panoramic Image for Indoor Mobile Measurement System | This paper designs a technique route to generate high-quality panoramic image with depth information, which involves two critical research hotspots: fusion of LiDAR and image data and image stitching. For the fusion of 3D points and image data, since a sparse depth map can be firstly generated by projecting LiDAR point onto the RGB image plane based on our reliable calibrated and synchronized sensors, we adopt a parameter self-adaptive framework to produce 2D dense depth map. For image stitching, optimal seamline for the overlapping area is searched using a graph-cuts-based method to alleviate the geometric influence and image blending based on the pyramid multi-band is utilized to eliminate the photometric effects near the stitching line. Since each pixel is associated with a depth value, we design this depth value as a radius in the spherical projection which can further project the panoramic image to the world coordinate and consequently produces a high-quality measurable panoramic image. The purposed method is tested on the data from our data collection platform and presents a satisfactory application prospects. | ['Sheng Yang', 'Xiaodong Gong', 'Zemin Wang', 'Dong Xu', 'Hongyu Qiu', 'Zhirong Hu', 'Jingbin Liu', 'Hao Ma'] | 2020-10-27 | null | null | null | null | ['image-stitching'] | ['computer-vision'] | [ 4.92993832e-01 -3.21356714e-01 4.35306542e-02 -2.92763203e-01
-3.03858697e-01 -2.10846767e-01 3.15667003e-01 -5.12469828e-01
-4.03834671e-01 2.66535878e-01 -1.63054559e-02 2.59750690e-02
-2.32769102e-01 -1.14770925e+00 -5.37831128e-01 -6.11031651e-01
5.80905080e-01 2.83359647e-01 3.93863380e-01 -1.84654355e-01
5.00358522e-01 5.83232045e-01 -1.70182920e+00 -2.60624975e-01
1.06551170e+00 8.65279257e-01 4.83381540e-01 1.86955884e-01
-2.54594773e-01 -1.70016348e-01 -3.50767136e-01 -3.75840217e-01
7.47582555e-01 -3.20607543e-01 -8.99629891e-02 5.76607764e-01
5.02451837e-01 -5.66615045e-01 -1.19143501e-01 1.32969570e+00
2.78654993e-01 -3.04908991e-01 3.63925606e-01 -1.17731285e+00
-2.13795274e-01 -3.68716493e-02 -1.32180965e+00 -4.44029152e-01
2.63290852e-01 -3.99550386e-02 5.10269701e-01 -8.12949479e-01
6.85356200e-01 1.27213132e+00 4.23856944e-01 -3.06933969e-02
-9.27454770e-01 -8.06374669e-01 -5.18654168e-01 -1.01380259e-01
-1.44428909e+00 -6.29165247e-02 1.27563012e+00 -1.13426439e-01
5.08589335e-02 2.89423108e-01 9.65143323e-01 2.37451002e-01
5.53784192e-01 5.81687875e-02 1.17445803e+00 -6.14514828e-01
-5.66171706e-02 -4.22836356e-02 -1.77000940e-01 6.21718943e-01
3.55410099e-01 3.38518500e-01 -4.12509263e-01 4.87366877e-02
1.38560557e+00 6.45341873e-01 -5.30182898e-01 -5.85459352e-01
-1.33200932e+00 5.43183684e-01 5.59570372e-01 1.99239522e-01
-3.36019427e-01 -1.12753145e-01 -2.03202173e-01 2.25011725e-02
1.64755180e-01 1.13027975e-01 6.67790622e-02 1.70319021e-01
-9.54499543e-01 -5.82398176e-02 1.94094121e-01 8.87364864e-01
1.42277408e+00 -1.11263782e-01 5.35932124e-01 7.92894542e-01
5.63886821e-01 8.85006130e-01 3.62864196e-01 -1.17339432e+00
5.12698889e-01 1.09030652e+00 -6.14938438e-02 -1.17259324e+00
-2.23798662e-01 1.57564238e-01 -8.33014965e-01 4.68312204e-01
4.21591029e-02 -6.81350827e-02 -7.86130726e-01 9.90275323e-01
6.26499712e-01 -1.54271517e-02 3.74691226e-02 1.13645899e+00
5.35266876e-01 8.49468410e-01 -6.52750254e-01 -5.10715127e-01
1.48232174e+00 -5.27173042e-01 -6.91995621e-01 -1.47930533e-01
-1.57927293e-02 -1.12226605e+00 1.14042485e+00 5.03183961e-01
-9.18675840e-01 -5.29619038e-01 -1.29814470e+00 -1.74634516e-01
1.20506868e-01 7.04649314e-02 3.25235218e-01 5.40562391e-01
-7.04761505e-01 1.03926249e-01 -3.49693328e-01 -1.94469884e-01
-1.94852710e-01 -2.09469162e-02 -4.60158795e-01 -3.91772985e-01
-8.29239309e-01 7.62913167e-01 3.79835635e-01 4.57272343e-02
-1.99011758e-01 -5.15122950e-01 -7.10644960e-01 -3.06448370e-01
2.54003376e-01 -6.06631219e-01 5.91709673e-01 -6.68485880e-01
-1.56056416e+00 9.48362589e-01 -2.60530636e-02 2.84058731e-02
3.46903801e-01 -1.15698978e-01 -4.37117487e-01 3.62720549e-01
1.25480935e-01 6.13045394e-01 9.70102489e-01 -1.34829950e+00
-6.34627759e-01 -6.82623982e-01 -5.76080978e-01 7.90876210e-01
-2.01754317e-01 -1.75018668e-01 -7.77909935e-01 -2.95197457e-01
9.09809351e-01 -6.73118949e-01 -2.81175941e-01 6.54252619e-02
-2.88454771e-01 3.48095447e-01 1.27415192e+00 -4.40322250e-01
8.48766387e-01 -2.23464918e+00 9.58580244e-03 3.86898607e-01
-6.98089302e-02 -1.07061692e-01 2.94038028e-01 2.55598724e-01
1.61200672e-01 -2.86396921e-01 -5.60928404e-01 -1.85200527e-01
-6.10833943e-01 8.47977772e-02 -1.77880615e-01 5.33906758e-01
-4.27901804e-01 4.90458012e-01 -5.48585832e-01 -9.01822448e-01
7.48986006e-01 4.20076072e-01 -2.44663596e-01 2.62149245e-01
-8.10331628e-02 1.73214078e-01 -4.53308553e-01 7.56770670e-01
1.41062713e+00 3.94913077e-01 -3.22417289e-01 -4.74933743e-01
-6.61214948e-01 -4.84848678e-01 -1.40933299e+00 2.02245975e+00
-3.53499323e-01 2.57728040e-01 2.76258647e-01 -1.53548092e-01
1.72472286e+00 -7.66948909e-02 6.49511933e-01 -7.77195811e-01
1.70072734e-01 2.04654112e-01 -5.47998786e-01 -3.50781560e-01
8.83742630e-01 -1.77422345e-01 1.71652697e-02 4.92433429e-01
-4.62213933e-01 -1.10013986e+00 -3.32701504e-01 -8.05576220e-02
3.31863016e-01 1.03087783e-01 4.29163128e-02 -1.49760798e-01
6.40830398e-01 2.74112791e-01 6.32832408e-01 5.38025983e-02
3.04045230e-01 1.01036394e+00 -4.01499085e-02 -3.01912934e-01
-1.39888477e+00 -1.03337598e+00 -3.84011686e-01 -4.62450739e-03
9.54852343e-01 -1.84902266e-01 -7.39327431e-01 -4.25991490e-02
-1.97480947e-01 3.73140037e-01 -2.01719567e-01 -6.75358921e-02
-4.89002973e-01 -5.67562878e-01 -1.36539251e-01 -1.57123506e-01
1.14659846e+00 -9.28404152e-01 -8.65157783e-01 1.14843324e-01
-5.28379157e-02 -7.14311421e-01 -5.05167603e-01 -2.32822433e-01
-1.14842927e+00 -1.33739352e+00 -7.63556361e-01 -6.38375998e-01
6.50282562e-01 9.12965953e-01 5.95613658e-01 -5.31113110e-02
-3.01610291e-01 2.44298190e-01 -1.41597345e-01 -6.91018403e-02
-4.97379377e-02 -4.34197217e-01 -2.24060312e-01 5.52224703e-02
1.98776856e-01 -6.06499434e-01 -9.54793155e-01 6.46915555e-01
-1.09803557e+00 6.27088130e-01 7.20186055e-01 3.24731797e-01
7.10097075e-01 3.19623262e-01 -1.33040562e-01 -1.91311985e-01
3.36815923e-01 -2.16377955e-02 -1.17963648e+00 9.98977125e-02
-6.12167716e-01 -2.65566736e-01 7.83089474e-02 -6.24536797e-02
-1.27257204e+00 3.57965142e-01 1.62861347e-01 -5.05393267e-01
4.47993837e-02 4.54688594e-02 -3.89754981e-01 -2.15996251e-01
4.68911380e-01 4.43026930e-01 4.08091158e-01 -3.35233867e-01
6.72050774e-01 8.66551578e-01 9.70325708e-01 -1.92827597e-01
1.05004871e+00 9.35341239e-01 1.60790190e-01 -9.74682629e-01
-1.75978765e-01 -4.53985155e-01 -8.10643077e-01 -4.25741166e-01
1.15486562e+00 -9.01568055e-01 -5.10280550e-01 6.97019517e-01
-1.19958270e+00 3.02393079e-01 -7.50293434e-02 7.22028911e-01
-4.55015540e-01 7.12459147e-01 -3.18487495e-01 -5.99602401e-01
-5.31145394e-01 -1.24571097e+00 1.10807180e+00 5.55892050e-01
3.88062149e-01 -5.18638253e-01 2.72677749e-01 3.31384927e-01
-1.15917578e-01 2.66380578e-01 5.10306478e-01 5.70404172e-01
-1.07608080e+00 -9.00161862e-02 -2.72420615e-01 2.14647010e-01
2.90209949e-01 4.78302419e-01 -8.13356161e-01 1.81951106e-01
4.95636195e-01 5.86809069e-02 5.53994298e-01 3.75297368e-01
7.46543944e-01 1.38615459e-01 -2.98175931e-01 9.84430015e-01
1.78263533e+00 3.16863298e-01 9.52309191e-01 6.35523558e-01
8.17285776e-01 5.68929017e-01 1.02934802e+00 1.98672041e-01
3.55603606e-01 7.26563334e-01 5.85478663e-01 -2.78937876e-01
6.85165226e-02 -4.94541347e-01 1.38031140e-01 6.29496098e-01
4.00680490e-02 1.57923400e-01 -6.19379997e-01 2.12463453e-01
-1.51378977e+00 -5.95044136e-01 -4.32405144e-01 2.58130312e+00
6.92168057e-01 3.16712074e-02 -2.51762658e-01 2.32428402e-01
9.77131128e-01 1.81244314e-01 -2.64738321e-01 -1.71397224e-01
-2.03138471e-01 -6.97195604e-02 7.59215117e-01 6.65951431e-01
-5.19692957e-01 8.76356840e-01 5.63120604e+00 8.01190197e-01
-1.25803518e+00 -2.99577296e-01 1.14053912e-01 2.82829791e-01
-6.55393422e-01 3.33541930e-01 -6.83077335e-01 5.40102541e-01
2.34612271e-01 -1.61539063e-01 2.07638264e-01 7.79912710e-01
2.42004290e-01 -7.32786477e-01 -3.51992697e-01 1.44214714e+00
1.13762915e-01 -1.10692656e+00 7.10028782e-02 3.40418726e-01
8.62781227e-01 -1.60314396e-01 -4.15307991e-02 -7.54754126e-01
1.29823074e-01 -5.28325021e-01 4.54090834e-01 5.54781258e-01
9.08941269e-01 -8.73385370e-01 3.66831839e-01 4.19351369e-01
-1.27514899e+00 1.26211777e-01 -7.10568249e-01 1.10098600e-01
5.84389985e-01 1.00710058e+00 -6.57032192e-01 8.42849076e-01
7.20247746e-01 7.68111825e-01 -4.75024492e-01 1.18889201e+00
-3.00926358e-01 -1.87805206e-01 -6.69606209e-01 4.29449826e-01
-1.39385657e-02 -1.33157182e+00 4.06287283e-01 3.93125176e-01
8.98526609e-01 1.86507687e-01 -1.44568190e-01 8.93061817e-01
2.77550548e-01 1.63144410e-01 -7.80859530e-01 4.88039136e-01
8.15876782e-01 1.45077515e+00 -7.72970855e-01 -1.56168208e-01
-3.21863055e-01 9.19271469e-01 -2.22348243e-01 7.33705759e-02
-6.11659050e-01 -3.26350480e-01 1.36204392e-01 1.65660650e-01
-5.40618263e-02 -4.27591503e-01 -6.73326492e-01 -9.37003911e-01
5.78891970e-02 -5.25998473e-01 2.28607342e-01 -1.32953918e+00
-4.94795978e-01 3.46767157e-01 -1.97213516e-02 -1.55292237e+00
1.20287597e-01 -1.47247687e-01 -8.86851490e-01 9.26105678e-01
-1.28092825e+00 -1.03222990e+00 -8.31708491e-01 9.02088702e-01
1.37639090e-01 2.12955207e-01 5.51721394e-01 5.01547791e-02
-2.78708458e-01 -2.14387640e-01 8.28169957e-02 -3.22209805e-01
7.74176657e-01 -7.18380392e-01 5.44416793e-02 1.02211773e+00
1.84585135e-02 3.35430175e-01 5.38400948e-01 -7.94119835e-01
-1.48690534e+00 -6.36274338e-01 3.93859088e-01 -2.11009905e-01
1.14231080e-01 -8.53234455e-02 -7.84505486e-01 4.02773410e-01
9.11075398e-02 -4.17235821e-01 -6.29636496e-02 -5.96441746e-01
3.49116288e-02 -4.77040976e-01 -1.29530585e+00 6.36916339e-01
8.43584538e-01 -3.23433578e-01 -8.44521821e-01 4.29663770e-02
7.46693492e-01 -5.30018985e-01 -7.40945041e-01 5.67048788e-01
4.68989879e-01 -1.51877379e+00 9.88423288e-01 7.78966248e-01
4.65943724e-01 -7.87897110e-01 4.35673147e-02 -1.06074870e+00
2.68536419e-01 -5.68985045e-01 7.34377384e-01 1.27119637e+00
-4.81074303e-02 -6.01704836e-01 7.64728129e-01 4.25309837e-01
-1.38930425e-01 -2.80461431e-01 -8.21315169e-01 -2.34597623e-01
-5.15128970e-01 -1.63838863e-01 8.34440827e-01 7.30720997e-01
-1.34534091e-01 2.48228684e-01 -4.63698626e-01 2.99047083e-01
1.05531561e+00 6.07977152e-01 1.22230971e+00 -1.15906703e+00
2.97277957e-01 -2.23897561e-01 -3.11944246e-01 -1.09593606e+00
-4.03511792e-01 -3.86966527e-01 -1.15566300e-02 -1.47802317e+00
-9.85687450e-02 -6.98124647e-01 4.31720078e-01 -6.48942441e-02
9.97188762e-02 3.69637579e-01 6.42514154e-02 6.24175608e-01
1.19629927e-01 6.68968558e-01 1.51559603e+00 2.96766937e-01
-4.96086717e-01 2.68312134e-02 -3.64162564e-01 6.91952825e-01
3.87198269e-01 -2.19040677e-01 -4.32397515e-01 -4.17394340e-01
1.36429057e-01 4.56189215e-01 9.14570540e-02 -1.04684937e+00
3.82109821e-01 -4.08982903e-01 3.75168562e-01 -1.19809604e+00
6.74485743e-01 -1.34192276e+00 5.79841673e-01 5.33034325e-01
4.05121118e-01 3.05547267e-02 -2.47738674e-01 4.29666132e-01
-2.19843432e-01 -2.11919263e-01 8.62367272e-01 -1.21761486e-01
-7.64631569e-01 4.27045912e-01 2.92717725e-01 -5.68603992e-01
1.25947368e+00 -7.38605440e-01 -2.54749835e-01 -2.97080845e-01
7.91076943e-02 2.33406559e-01 1.23335683e+00 2.34207362e-01
1.05218828e+00 -1.55704665e+00 -3.65363717e-01 7.61499584e-01
1.48652211e-01 5.05106747e-01 3.64984959e-01 4.36952949e-01
-1.08109057e+00 5.35210408e-02 -4.72264439e-01 -9.98253703e-01
-1.29374421e+00 3.82106602e-01 2.25811169e-01 4.47152048e-01
-9.22089040e-01 3.85327607e-01 1.29577085e-01 -3.73896092e-01
-2.27095306e-01 -2.22892657e-01 -1.62665367e-01 -1.49356753e-01
4.10917252e-01 4.59154963e-01 -1.08167224e-01 -7.58164346e-01
-1.84107572e-01 1.50320482e+00 3.79144639e-01 -6.01857483e-01
1.09236860e+00 -5.16345561e-01 -3.57150227e-01 1.30798057e-01
1.09207797e+00 4.49402630e-01 -1.43222654e+00 -1.91427097e-01
-4.20956880e-01 -1.06241393e+00 1.48697391e-01 -1.30475178e-01
-1.14082432e+00 8.24257553e-01 6.14260256e-01 -7.14783892e-02
1.20326042e+00 -1.67954117e-01 7.65734911e-01 -1.03229195e-01
4.59754795e-01 -1.00688303e+00 1.60089899e-02 1.10304216e-02
7.58564770e-01 -9.96315956e-01 4.52565551e-01 -5.98865449e-01
-5.70048034e-01 1.44185317e+00 6.90194488e-01 -2.44129777e-01
4.93740737e-01 8.99378657e-02 2.61778563e-01 -3.70964020e-01
1.07280798e-01 8.36562291e-02 8.77297744e-02 5.16239524e-01
-1.98298261e-01 -1.55983046e-01 -4.72582668e-01 -1.51556984e-01
-6.28755569e-01 -9.42166895e-03 6.76121294e-01 6.92571640e-01
-9.56006229e-01 -1.00792933e+00 -9.73643303e-01 -3.01155094e-02
2.26105154e-01 1.02713101e-01 -6.98200241e-02 7.28114963e-01
1.43067196e-01 7.57953107e-01 4.48982120e-01 -3.63204986e-01
3.55555952e-01 -3.09772313e-01 4.13866937e-01 -3.79205436e-01
1.63645044e-01 3.51304501e-01 -4.28053886e-01 -6.29418373e-01
-4.85910773e-01 -3.35790545e-01 -1.32495964e+00 -2.43096739e-01
-4.20850247e-01 1.53231621e-01 1.04914963e+00 5.81907094e-01
-1.65442973e-02 -2.32784361e-01 1.08110940e+00 -8.69232059e-01
8.24909136e-02 -6.07726574e-01 -8.65474582e-01 2.32167184e-01
9.41159129e-02 -5.10509849e-01 -6.29292905e-01 -1.41822189e-01] | [9.044339179992676, -2.463674306869507] |
5229ee30-a780-4b3f-bbef-b02433d5d236 | learning-scene-geometry-for-visual | null | null | https://hal.archives-ouvertes.fr/hal-02057378/document | https://hal.archives-ouvertes.fr/hal-02057378/document | Learning Scene Geometry for Visual Localization in Challenging Conditions | We propose a new approach for outdoor large scale image based localization that can deal with challenging scenarios like cross-season, cross-weather, day/night and longterm localization. The key component of our method is a new learned global image descriptor, that can effectively benefit from scene geometry information during training. At test time, our system is capable of inferring the depth map related to the query image and use it to increase localization accuracy.We are able to increase recall@1 performances by 2.15% on cross-weather and long-term localization scenario and by 4.24% points on a challenging winter/summer localization sequence versus state-of-the-art methods. Our method can also use weakly annotated data to localize night images across a reference dataset of daytime images. | ['Valerie Gouet-Brunet，Cedric Demonceaux1', 'Nathan Piasco', 'Desire Sidibe'] | 2019-05-01 | null | null | null | international-conference-on-robotics-and-2 | ['image-based-localization'] | ['computer-vision'] | [-1.73136726e-01 -6.22239113e-01 2.53707051e-01 -7.00940549e-01
-1.33922601e+00 -1.06556547e+00 5.04841745e-01 2.98126161e-01
-8.87576580e-01 6.97088420e-01 -1.25877231e-01 1.57788441e-01
-9.73774195e-02 -8.01078677e-01 -8.37229609e-01 -6.27385259e-01
-3.01438957e-01 2.30437964e-01 4.69714969e-01 -2.29239091e-01
2.59417802e-01 7.51641572e-01 -1.61987460e+00 1.13283060e-01
5.07601142e-01 1.26792502e+00 4.69643116e-01 9.07986879e-01
2.37211913e-01 3.53112161e-01 -8.70740056e-01 5.43834753e-02
4.23712105e-01 1.57073840e-01 -4.97050315e-01 3.16687189e-02
1.05290353e+00 -1.26982316e-01 -2.56655086e-03 7.23229527e-01
9.08134460e-01 3.05793434e-01 1.85920209e-01 -9.53122437e-01
-3.19598839e-02 -3.21118146e-01 -3.89904112e-01 3.13896120e-01
6.27212048e-01 -4.18074206e-02 6.46623492e-01 -8.39741647e-01
4.41202492e-01 7.13150024e-01 1.04650390e+00 -1.38371974e-01
-9.40951228e-01 -4.78641033e-01 1.39433548e-01 1.04181558e-01
-1.94380057e+00 -3.61653268e-01 2.90534347e-01 -1.63995877e-01
1.05664146e+00 4.33674067e-01 3.75446171e-01 7.13565588e-01
5.32225780e-02 2.73768723e-01 1.39265084e+00 -2.82186031e-01
3.01066220e-01 1.38369352e-01 -2.15037122e-01 7.90636897e-01
-1.60222813e-01 6.02654666e-02 -5.87639153e-01 -6.29114583e-02
3.03232521e-01 1.80333084e-03 -3.76589119e-01 -3.86515141e-01
-1.19989192e+00 5.37735760e-01 9.11044598e-01 1.68664798e-01
-1.94870487e-01 2.24444464e-01 1.86348796e-01 2.38229036e-01
4.75584954e-01 7.03163564e-01 -7.42798746e-01 -2.14820892e-01
-1.21934319e+00 3.43073785e-01 8.35510850e-01 9.87407863e-01
1.21168017e+00 -4.34378892e-01 -6.06983490e-02 8.27926636e-01
-4.28193584e-02 1.22448480e+00 4.08169061e-01 -8.14937055e-01
3.52906108e-01 2.39032239e-01 4.25139487e-01 -1.00978494e+00
-5.94065607e-01 -6.10874653e-01 -4.66411889e-01 -9.24035087e-02
2.36705929e-01 1.82552412e-01 -8.35504591e-01 1.39622116e+00
2.29350924e-01 2.88466632e-01 7.57455826e-02 9.61805582e-01
4.56985265e-01 5.93864620e-01 -2.78472126e-01 1.82351604e-01
1.25584233e+00 -1.21085787e+00 -2.62879044e-01 -5.90242147e-01
5.98643303e-01 -9.11376536e-01 9.02383089e-01 1.81687862e-01
-4.59967196e-01 -8.40769470e-01 -8.81093323e-01 1.00490518e-01
-8.97476375e-01 6.09114230e-01 4.06483829e-01 4.33214337e-01
-1.39930761e+00 2.04532817e-01 -7.05463409e-01 -1.04814422e+00
-1.45018041e-01 4.04194564e-01 -7.16718793e-01 -4.74785447e-01
-9.15015221e-01 8.31824303e-01 2.15638041e-01 2.19828233e-01
-1.17177939e+00 -6.09825373e-01 -9.17606175e-01 -2.88175672e-01
1.33911759e-01 -4.80958998e-01 8.85608017e-01 -5.36109507e-01
-1.04175556e+00 8.43786359e-01 -4.00562257e-01 -6.06508374e-01
3.77335489e-01 -4.71826613e-01 -5.77716708e-01 1.96069002e-01
5.47733903e-01 8.02281082e-01 5.54761171e-01 -1.26296210e+00
-7.44982481e-01 -3.28027368e-01 2.27187961e-01 2.91646570e-01
-5.52532300e-02 -3.54229689e-01 -7.47085690e-01 -2.59390652e-01
1.21642895e-01 -1.31054962e+00 -2.01735005e-01 1.01514421e-01
-3.45049500e-02 2.84354270e-01 8.69891644e-01 -3.58334124e-01
6.94813371e-01 -2.11718869e+00 -3.44679713e-01 2.63055991e-02
-4.47323918e-01 4.56028692e-02 -2.28836164e-01 7.16801643e-01
3.95170152e-01 -1.94656134e-01 -1.39684781e-01 -6.26508355e-01
-2.49164283e-01 4.01890635e-01 -3.26888561e-01 7.39168644e-01
-2.53860168e-02 7.15989351e-01 -1.01040578e+00 -1.95625499e-01
5.73219776e-01 4.63443547e-01 -1.73182145e-01 4.30544108e-01
1.99193403e-01 5.71374059e-01 -1.00143306e-01 8.28840792e-01
9.59573627e-01 2.72426177e-02 -2.11130440e-01 -3.46109450e-01
-5.20991981e-01 -7.18095601e-02 -1.08335292e+00 2.28228569e+00
-9.48962390e-01 9.13601875e-01 -9.87054035e-02 -6.68629885e-01
9.95960414e-01 -1.84632346e-01 3.91195789e-02 -9.93527114e-01
-3.23598087e-01 2.37415358e-01 -7.01398790e-01 -5.13791203e-01
6.46323383e-01 2.33199283e-01 -3.00285488e-01 -1.03018172e-01
7.60109276e-02 -3.80601734e-01 -1.84084494e-02 -2.19574884e-01
1.18261886e+00 2.19330207e-01 1.60822153e-01 -5.11315942e-01
7.49353468e-01 -1.65655464e-02 1.68296099e-01 1.08409572e+00
-1.23662263e-01 9.17958677e-01 -1.15256459e-01 -8.61684680e-01
-7.57028461e-01 -1.08761585e+00 -1.70156091e-01 1.15633881e+00
3.84167314e-01 -4.74380583e-01 -2.76352465e-01 -6.28603756e-01
-5.30305505e-02 2.41338238e-01 -6.77568793e-01 1.69033691e-01
-2.12924913e-01 -6.66764200e-01 7.68850803e-01 4.55841810e-01
9.63763714e-01 -5.55806160e-01 -7.11324811e-01 5.05684391e-02
-3.51621121e-01 -1.64513922e+00 -2.98709035e-01 2.25720972e-01
-4.35657114e-01 -8.76749754e-01 -6.49099767e-01 -5.96518219e-01
6.42841697e-01 5.13043582e-01 1.15085030e+00 -2.21890286e-01
-5.25308192e-01 6.09233558e-01 -3.42518747e-01 -4.46146615e-02
3.88443887e-01 2.81570137e-01 1.67840093e-01 8.08856413e-02
-8.75956342e-02 -4.50943291e-01 -8.92179072e-01 7.41634965e-01
-6.24198079e-01 -4.39627230e-01 5.29841959e-01 6.81796074e-01
8.86072814e-01 -8.67963359e-02 -1.80101804e-02 -4.09047842e-01
1.18943360e-02 -3.52278203e-01 -1.00238550e+00 3.40791047e-01
-4.75239247e-01 -4.63826917e-02 5.00835240e-01 -6.26119003e-02
-5.53699613e-01 4.66027617e-01 -1.43030286e-01 -3.00465167e-01
-4.03016567e-01 2.79957503e-01 8.90810192e-02 -8.14487517e-01
7.47391701e-01 4.96443629e-01 -5.19138336e-01 -4.38607395e-01
2.77527690e-01 6.33544505e-01 6.79155290e-01 -5.95810771e-01
9.02313828e-01 8.81792903e-01 1.92525700e-01 -1.02936304e+00
-1.20453274e+00 -1.06280684e+00 -9.33816373e-01 -6.30869046e-02
8.56243610e-01 -1.60131955e+00 -6.00525796e-01 3.82787198e-01
-8.64316881e-01 -4.63365257e-01 4.47355770e-02 4.76573348e-01
-3.55732769e-01 5.33855259e-02 -1.57012306e-02 -6.45042539e-01
-1.45230964e-01 -9.73635256e-01 1.80210912e+00 3.49912345e-01
3.28754216e-01 -1.05986154e+00 2.46409640e-01 2.33839616e-01
7.33255923e-01 4.49194968e-01 -1.39117628e-01 -2.11233243e-01
-8.60271871e-01 -6.31184578e-01 -2.85003066e-01 3.62661064e-01
2.00359225e-01 -4.31565791e-01 -1.18881571e+00 -6.35799766e-01
-2.93552130e-01 -4.76995707e-01 8.28713417e-01 3.91434804e-02
1.00758541e+00 -6.87643066e-02 -3.57570350e-01 1.09353280e+00
1.86888552e+00 -2.94478923e-01 6.40378118e-01 6.19386435e-01
5.81308722e-01 3.30434740e-01 1.07832992e+00 4.10697937e-01
8.22469473e-01 9.59489942e-01 5.86090386e-01 -3.06753129e-01
1.38315156e-01 -4.57333252e-02 1.74857512e-01 2.92755604e-01
1.51369289e-01 -2.57214874e-01 -9.73683298e-01 6.92525804e-01
-1.85918617e+00 -6.38062596e-01 3.93043421e-02 2.45231390e+00
1.57049015e-01 2.69479323e-02 -2.62679607e-01 -4.57965761e-01
2.66816318e-01 3.30219209e-01 -1.56694040e-01 -8.22764635e-02
-2.60964811e-01 1.44224778e-01 1.37684691e+00 5.55154264e-01
-1.51709068e+00 1.09016967e+00 6.40080070e+00 6.68004632e-01
-1.30842757e+00 1.63484097e-01 3.48052442e-01 -6.66292664e-03
4.01161194e-01 1.09759785e-01 -9.77040410e-01 2.44629353e-01
1.17011023e+00 3.33218634e-01 4.26146686e-01 1.01741409e+00
1.37826383e-01 -6.56298399e-01 -7.67293155e-01 1.14812279e+00
4.91192490e-01 -1.18739748e+00 -5.35681188e-01 1.71365701e-02
8.96642923e-01 7.30103254e-01 -2.49272913e-01 3.10826182e-01
-2.36180469e-01 -8.82913947e-01 5.81973076e-01 6.24059796e-01
9.73118305e-01 -6.25687540e-01 9.58563387e-01 4.72231537e-01
-1.66808105e+00 4.06490220e-03 -4.46899503e-01 -1.02648981e-01
-4.16518599e-02 4.22119826e-01 -9.29509461e-01 6.89115405e-01
1.11933625e+00 8.53275955e-01 -1.24894226e+00 1.34526801e+00
-3.49219948e-01 2.30576783e-01 -8.15141439e-01 1.54385149e-01
4.64763165e-01 1.49614364e-01 1.45698324e-01 1.41092932e+00
5.40122747e-01 -1.60233319e-01 6.13859713e-01 3.23231876e-01
2.29704127e-01 -1.22423440e-01 -9.05328870e-01 4.90090400e-01
3.30423921e-01 1.62797189e+00 -6.57549739e-01 -2.10287973e-01
-1.24716088e-01 1.45901942e+00 2.30974779e-01 2.71430522e-01
-9.82907534e-01 -4.74807948e-01 7.19913185e-01 5.99065870e-02
5.36936045e-01 -6.86857462e-01 5.30739665e-01 -1.14878309e+00
2.01195374e-01 -4.43580538e-01 7.89843723e-02 -1.03131711e+00
-9.23097253e-01 8.67080271e-01 -1.04824968e-01 -1.27172482e+00
-2.54004207e-02 -5.52409351e-01 -3.81048590e-01 9.42143857e-01
-1.80207455e+00 -1.65755248e+00 -1.03177559e+00 6.32999718e-01
3.88550580e-01 1.30353317e-01 1.24139690e+00 6.13556921e-01
1.42745484e-04 4.30589139e-01 4.34343010e-01 -4.75375447e-03
9.98300672e-01 -1.32717097e+00 3.56626153e-01 8.51658344e-01
4.51490849e-01 5.10275960e-01 5.43230772e-01 -1.75383195e-01
-1.44231105e+00 -1.46085596e+00 8.09191704e-01 -7.35532820e-01
5.26212692e-01 -8.37779880e-01 -4.43583459e-01 3.88924718e-01
-6.88330978e-02 7.15980709e-01 5.13398468e-01 7.33153224e-02
-4.58533645e-01 -7.71027386e-01 -1.08646643e+00 -2.52209119e-02
9.14790809e-01 -9.29901302e-01 2.65698135e-02 8.25027645e-01
5.85445821e-01 -7.95362115e-01 -9.30601180e-01 4.17149276e-01
4.88501906e-01 -1.10227609e+00 1.19181299e+00 2.28354827e-01
-1.86022028e-01 -7.05561817e-01 -6.27214670e-01 -1.16920805e+00
3.74882296e-02 -3.41398716e-01 3.88978273e-01 1.14885163e+00
1.21551834e-01 -6.14089191e-01 3.82589996e-01 -5.71953580e-02
-1.06882110e-01 -4.30967540e-01 -1.07033420e+00 -1.13982558e+00
-6.03139758e-01 -4.91660088e-01 3.96092623e-01 5.81651807e-01
-8.88281584e-01 -2.66966764e-02 -5.15969276e-01 8.88475358e-01
5.26696146e-01 3.24078053e-01 1.06198299e+00 -8.28358471e-01
-7.29322955e-02 2.53104717e-01 -8.60576332e-01 -1.05205214e+00
8.51950794e-02 -5.13478637e-01 3.61677647e-01 -1.70102751e+00
-3.24571669e-01 -5.32653987e-01 -4.80639517e-01 5.03850162e-01
2.50258923e-01 9.69267845e-01 -2.20313352e-02 2.39808694e-01
-1.14752328e+00 4.13159847e-01 5.05249977e-01 5.08222431e-02
1.22461125e-01 -4.79680374e-02 -1.28183752e-01 3.25980604e-01
6.30025625e-01 -5.04482090e-01 -2.49036595e-01 -5.82871974e-01
1.82320476e-01 -6.93741143e-02 6.57777071e-01 -1.67238522e+00
3.67610097e-01 8.47875625e-02 5.06783545e-01 -6.07828021e-01
6.74390078e-01 -9.52212214e-01 3.72381397e-02 3.13755631e-01
6.71914890e-02 3.45736027e-01 4.40938801e-01 7.07926154e-01
-4.93048251e-01 1.54271692e-01 6.12276912e-01 -2.00502425e-01
-1.24880576e+00 2.48270750e-01 9.87112522e-02 -1.05863705e-01
1.10599673e+00 8.64016488e-02 -3.99978161e-01 -3.22353303e-01
-5.02405524e-01 3.51077259e-01 6.70489967e-01 6.16294384e-01
3.37583005e-01 -1.17974579e+00 -3.81883472e-01 2.48877332e-01
7.92895675e-01 -1.08980007e-01 8.50981846e-02 8.10839653e-01
-9.95901227e-01 6.15547717e-01 -7.79233724e-02 -1.05024683e+00
-1.20728147e+00 2.98557431e-01 5.96183360e-01 -9.13861170e-02
-2.91320860e-01 9.68542457e-01 7.04364553e-02 -6.25554085e-01
4.24229987e-02 -4.05021578e-01 2.37935141e-01 3.51131745e-02
5.25061905e-01 2.32722741e-02 4.38144058e-01 -8.11692178e-01
-9.60171819e-01 1.09299994e+00 3.69739234e-01 -1.75561145e-01
1.18700385e+00 -4.58443522e-01 -1.59794360e-01 5.88145792e-01
1.63073075e+00 2.74934769e-01 -1.25547147e+00 -1.08154662e-01
-1.26623720e-01 -7.80320942e-01 1.06912762e-01 -1.00192451e+00
-8.59808087e-01 7.17376232e-01 1.42491055e+00 3.20534222e-02
1.30503929e+00 6.05512932e-02 4.73039001e-01 7.65216172e-01
7.60291576e-01 -8.03907990e-01 -2.06159979e-01 6.07073486e-01
6.71894848e-01 -1.69692779e+00 8.65827948e-02 9.57420766e-02
-4.59504932e-01 1.12080669e+00 4.13194656e-01 -3.17932129e-01
5.83173573e-01 2.49566421e-01 3.56117874e-01 -1.84554085e-01
-3.62601697e-01 -5.12611628e-01 4.62739199e-01 5.85615814e-01
1.82173669e-01 8.32681060e-02 2.33679429e-01 -2.38983765e-01
-8.49184617e-02 -3.43665302e-01 7.22472519e-02 9.91461515e-01
-3.61131996e-01 -9.47911263e-01 -3.74262601e-01 -1.50925457e-01
-1.97583035e-01 -2.12406993e-01 -1.81881636e-01 8.58060122e-01
5.83926976e-01 8.07859600e-01 1.35832205e-01 -3.72876555e-01
4.15672630e-01 -7.08808377e-02 4.12318498e-01 -4.55117196e-01
-5.37107825e-01 -1.01210373e-02 -1.01727247e-01 -1.06012964e+00
-5.87236166e-01 -5.06739259e-01 -9.02129650e-01 1.04556195e-01
-2.09837765e-01 2.53375292e-01 1.27869880e+00 7.64842689e-01
6.09366834e-01 2.78342128e-01 7.59696364e-01 -1.02419448e+00
4.85375673e-02 -7.37649918e-01 -4.81537461e-01 1.93405449e-01
7.87601292e-01 -4.42901999e-01 -4.01537657e-01 -1.45888403e-02] | [7.621814727783203, -2.016674280166626] |
a15ca712-a097-4c82-825f-e608062e00bc | a-study-on-the-integration-of-pre-trained-ssl | 2211.05869 | null | https://arxiv.org/abs/2211.05869v1 | https://arxiv.org/pdf/2211.05869v1.pdf | A Study on the Integration of Pre-trained SSL, ASR, LM and SLU Models for Spoken Language Understanding | Collecting sufficient labeled data for spoken language understanding (SLU) is expensive and time-consuming. Recent studies achieved promising results by using pre-trained models in low-resource scenarios. Inspired by this, we aim to ask: which (if any) pre-training strategies can improve performance across SLU benchmarks? To answer this question, we employ four types of pre-trained models and their combinations for SLU. We leverage self-supervised speech and language models (LM) pre-trained on large quantities of unpaired data to extract strong speech and text representations. We also explore using supervised models pre-trained on larger external automatic speech recognition (ASR) or SLU corpora. We conduct extensive experiments on the SLU Evaluation (SLUE) benchmark and observe self-supervised pre-trained models to be more powerful, with pre-trained LM and speech models being most beneficial for the Sentiment Analysis and Named Entity Recognition task, respectively. | ['Shinji Watanabe', 'Xuankai Chang', 'Siddharth Dalmia', 'Karthik Ganesan', 'Sujay Kumar', 'Yushi Ueda', 'Yosuke Higuchi', 'Siddhant Arora', 'Yifan Peng'] | 2022-11-10 | null | null | null | null | ['spoken-language-understanding', 'spoken-language-understanding'] | ['natural-language-processing', 'speech'] | [ 4.06872869e-01 2.68922389e-01 -1.67891532e-01 -7.21396327e-01
-1.31394386e+00 -5.87287366e-01 6.74719334e-01 6.06902912e-02
-6.41593993e-01 6.45967662e-01 7.09842265e-01 -5.77038825e-01
6.23170614e-01 -4.04502600e-01 -7.22338855e-01 -3.19964252e-02
8.88918936e-02 3.77249569e-01 -1.16724648e-01 -3.19167674e-01
1.39814923e-02 -7.56211430e-02 -1.50337696e+00 5.35567522e-01
9.47556436e-01 8.42238426e-01 3.00462544e-01 8.20528388e-01
-4.68212843e-01 1.13749337e+00 -5.46679318e-01 -1.20446786e-01
-7.60543197e-02 -6.91968739e-01 -1.11218977e+00 1.52606413e-01
7.24580213e-02 -2.31485903e-01 -3.31177860e-02 6.24380350e-01
6.94247961e-01 3.15352499e-01 5.97767055e-01 -7.27048218e-01
-6.73908830e-01 1.10044849e+00 -1.05095774e-01 3.10068071e-01
6.57683849e-01 -1.59393754e-02 1.25779331e+00 -1.02674270e+00
6.56644762e-01 1.21441734e+00 4.99630749e-01 9.50814068e-01
-1.02591050e+00 -5.60259044e-01 1.11532353e-01 -1.23356171e-02
-1.20455265e+00 -1.16964626e+00 5.12659669e-01 7.54846931e-02
1.47322071e+00 2.67544448e-01 -2.31342420e-01 1.55106103e+00
-5.37609279e-01 1.49042487e+00 1.30039787e+00 -8.63923728e-01
3.27428699e-01 5.23253322e-01 4.45429415e-01 3.76830399e-01
-2.90069968e-01 -1.91135406e-01 -8.00478160e-01 2.75140926e-02
7.97073543e-02 -4.01306152e-01 -5.07616162e-01 1.04696080e-01
-1.17020190e+00 9.76915121e-01 -5.23068160e-02 4.86528516e-01
-2.02628762e-01 -3.57768625e-01 6.26697540e-01 5.76393485e-01
8.34475875e-01 6.61535025e-01 -1.07328773e+00 -5.61518729e-01
-9.70481575e-01 -6.02207959e-01 1.02955687e+00 9.47122097e-01
7.70682871e-01 1.23725884e-01 -6.85157627e-02 1.38596797e+00
2.74980485e-01 5.83676457e-01 8.32859576e-01 -4.75254208e-01
6.91720307e-01 4.13597882e-01 -2.87596315e-01 -1.06220283e-01
-3.90111446e-01 -3.13724726e-01 -4.90700752e-01 -3.54054093e-01
3.39494824e-01 -5.43423593e-01 -9.54977751e-01 1.72737658e+00
-2.35825032e-02 3.90122920e-01 8.63418519e-01 5.02547979e-01
1.18755996e+00 1.05121791e+00 2.93320358e-01 -2.03757674e-01
1.12612271e+00 -1.20403433e+00 -6.89909518e-01 -6.29988611e-01
1.42807508e+00 -7.65119493e-01 1.44434917e+00 -3.97592364e-03
-7.97350109e-01 -4.26044405e-01 -8.03696752e-01 6.08041845e-02
-5.29558241e-01 4.39847916e-01 2.77169526e-01 8.22273314e-01
-1.17279637e+00 2.51233637e-01 -7.32450783e-01 -5.57146251e-01
1.91043586e-01 1.18320949e-01 -4.06849802e-01 -1.54841408e-01
-1.48764265e+00 8.69297266e-01 1.54813722e-01 5.06954081e-02
-8.92066360e-01 -4.78122652e-01 -1.29841959e+00 1.07723236e-01
5.51193237e-01 -2.12022766e-01 1.50402153e+00 -1.14138162e+00
-1.81517696e+00 1.03552616e+00 -4.61471349e-01 -6.30092382e-01
1.62718073e-01 -1.69823766e-01 -2.93372512e-01 -5.37764318e-02
-1.81705087e-01 5.93398452e-01 5.55109382e-01 -1.18520224e+00
-3.91099006e-01 -2.66170233e-01 -2.62456596e-01 2.57428020e-01
-6.37996376e-01 3.72850418e-01 -1.12147339e-01 -4.97571498e-01
-2.34389290e-01 -7.56228268e-01 -6.95293099e-02 -1.04209459e+00
-3.90132248e-01 -5.72200179e-01 7.20017016e-01 -9.07303274e-01
1.29277432e+00 -1.96638751e+00 -4.10058275e-02 -9.03738439e-02
-3.46914858e-01 6.05096877e-01 -5.34506500e-01 4.72712487e-01
-6.30201176e-02 3.34235847e-01 -2.07866490e-01 -8.74680459e-01
-7.32354224e-02 4.85513717e-01 -4.62765396e-01 -1.27113178e-01
5.94321668e-01 9.61161315e-01 -8.89606774e-01 -2.46677905e-01
2.27000624e-01 3.74759495e-01 -3.84890676e-01 5.83226502e-01
-2.04939827e-01 5.81753790e-01 -5.08456230e-01 6.52105451e-01
3.09522688e-01 -1.54638588e-01 1.73827916e-01 6.82803392e-02
8.59028921e-02 9.46537733e-01 -9.88032401e-01 1.58379352e+00
-1.15156710e+00 5.84368408e-01 1.06063634e-01 -1.11396360e+00
9.92244840e-01 5.23129404e-01 3.71584855e-02 -7.79198885e-01
4.47818696e-01 3.98236066e-01 -2.38821104e-01 -3.91904801e-01
3.54404181e-01 -8.52862448e-02 -1.18789621e-01 7.45738506e-01
5.59336841e-01 -1.01504110e-01 -5.13978936e-02 2.84330130e-01
1.35215902e+00 -1.71393663e-01 4.40505654e-01 -1.99163735e-01
7.99294055e-01 -2.46103555e-01 1.69222683e-01 7.97886372e-01
-4.34504747e-01 7.53126383e-01 1.94516748e-01 1.03996314e-01
-6.91290796e-01 -6.30490780e-01 -6.58380464e-02 1.60009158e+00
-4.71347153e-01 -5.07268846e-01 -9.00919735e-01 -1.08275986e+00
-3.78463656e-01 1.12235737e+00 -3.63374829e-01 -7.18499124e-02
-4.07571346e-01 -5.94519973e-01 5.87937295e-01 7.82263160e-01
1.53407991e-01 -1.41454792e+00 -5.67289209e-03 1.76039979e-01
-2.53368795e-01 -1.53058636e+00 -4.60489154e-01 4.44149852e-01
-5.71763873e-01 -6.09182477e-01 -8.29607069e-01 -8.03473592e-01
5.84299326e-01 8.44001397e-02 1.31521940e+00 -9.91841778e-02
3.18512112e-01 6.50867224e-01 -1.04476643e+00 -5.17875314e-01
-9.52314436e-01 5.49539626e-01 1.57897368e-01 9.93861482e-02
5.80934405e-01 -1.34056389e-01 7.93721899e-02 3.63416404e-01
-7.38956571e-01 -9.92317721e-02 6.39584363e-01 8.98450017e-01
3.19060236e-01 -5.38729250e-01 1.06943786e+00 -1.12542260e+00
5.65791309e-01 -4.26347405e-01 -2.31325924e-01 5.43511271e-01
-1.94310561e-01 2.06583470e-01 7.08531439e-01 -4.02013183e-01
-1.26103258e+00 4.40288633e-02 -7.44038761e-01 -2.53789693e-01
-3.76905411e-01 6.69740975e-01 -3.47430110e-01 2.14256898e-01
5.77500105e-01 3.86391252e-01 -1.61702976e-01 -5.89953959e-01
4.82846856e-01 1.36491132e+00 1.96334600e-01 -5.35441458e-01
2.83928931e-01 1.37692049e-01 -9.43763673e-01 -1.46167076e+00
-1.44215190e+00 -7.87974179e-01 -6.04816318e-01 7.12039769e-02
8.61809850e-01 -1.08824313e+00 -1.10394657e-01 3.48615766e-01
-1.14484680e+00 -7.46711791e-01 -1.30720884e-01 4.88901377e-01
-3.84570867e-01 3.13646466e-01 -5.52518070e-01 -1.14306295e+00
-5.18221974e-01 -9.91390646e-01 1.18996775e+00 1.95761949e-01
-4.21851337e-01 -1.14671183e+00 2.48708412e-01 7.49771953e-01
5.19480705e-01 -5.69046080e-01 4.59368765e-01 -1.50283349e+00
-1.00372687e-01 -1.13485187e-01 -1.66339487e-01 9.81304049e-01
-5.66135626e-04 -3.63038719e-01 -1.50201607e+00 -1.62372306e-01
-5.77751605e-04 -1.06761789e+00 8.75580311e-01 9.25207511e-02
9.01542544e-01 -1.62456810e-01 -3.57576162e-02 2.34167010e-01
7.30802476e-01 -4.47633192e-02 2.56731570e-01 7.71299675e-02
4.85866249e-01 8.89149010e-01 4.24123734e-01 1.45939127e-01
6.77446127e-01 2.45697290e-01 -1.38449460e-01 -7.27939680e-02
-8.29978511e-02 -3.48009795e-01 8.55855286e-01 1.32643187e+00
3.23060572e-01 -4.61223751e-01 -1.11113524e+00 7.31318533e-01
-1.63556623e+00 -4.10823852e-01 1.41543046e-01 2.00451446e+00
1.01075590e+00 1.07822441e-01 -1.71112984e-01 -1.44134656e-01
4.09439892e-01 3.56375158e-01 -1.83257207e-01 -3.35548073e-01
-3.62117290e-01 4.33490366e-01 2.53192127e-01 5.02994418e-01
-1.14614475e+00 1.29617918e+00 6.35646152e+00 8.89622450e-01
-1.22562456e+00 4.34241414e-01 1.01878107e+00 1.56233698e-01
-4.02640939e-01 -7.47569054e-02 -1.10691822e+00 1.69608995e-01
1.66609931e+00 1.07749932e-01 2.36619696e-01 8.56313705e-01
7.35064819e-02 -7.73258582e-02 -1.11107433e+00 8.19747210e-01
3.96351576e-01 -1.23316193e+00 -6.80701574e-03 -3.84658039e-01
7.67819583e-01 7.30888247e-01 -2.63400137e-01 8.93860698e-01
5.67445457e-01 -9.35608625e-01 3.38307202e-01 -6.41769767e-02
6.34105086e-01 -5.38304090e-01 1.14103818e+00 5.45252979e-01
-9.58218038e-01 2.11664349e-01 -2.67372161e-01 3.82999592e-02
2.99324304e-01 2.43103594e-01 -1.06928122e+00 5.67637384e-01
5.07930279e-01 7.08410203e-01 -5.26376247e-01 2.20247075e-01
-4.63431746e-01 1.29567754e+00 -5.21172047e-01 -3.27171475e-01
5.88489771e-01 -5.97739890e-02 3.92100334e-01 1.44818628e+00
6.38823658e-02 1.08856998e-01 1.45569205e-01 5.18730462e-01
-4.25680101e-01 5.98306715e-01 -6.79877460e-01 -5.36732137e-01
3.40855926e-01 1.23426509e+00 -6.99118733e-01 -6.08919561e-01
-9.20106113e-01 9.79624093e-01 6.27470255e-01 2.83133417e-01
-2.66434669e-01 -3.18633504e-02 4.11943674e-01 -1.84045866e-01
1.15562700e-01 -2.96165496e-01 -2.38922819e-01 -1.63309300e+00
-2.30853423e-01 -1.03073561e+00 5.53605676e-01 -6.65537655e-01
-1.27044618e+00 1.09817564e+00 -5.32368243e-01 -8.41480732e-01
-6.70634329e-01 -6.52919233e-01 -7.09814310e-01 6.93883836e-01
-2.00888848e+00 -1.30835521e+00 2.01070935e-01 3.95067781e-01
1.09409249e+00 -2.63080746e-01 9.86308455e-01 2.18347907e-01
-8.49806845e-01 6.39273643e-01 1.48529112e-01 4.65333283e-01
8.43409538e-01 -1.10015309e+00 3.57857317e-01 7.67952263e-01
6.97243989e-01 5.43176353e-01 5.96630573e-01 -4.84906673e-01
-1.31673586e+00 -1.00224543e+00 1.22397637e+00 -1.09525347e+00
9.15877104e-01 -6.90354943e-01 -1.08561373e+00 9.15534437e-01
3.75159621e-01 -9.01802704e-02 1.05805933e+00 5.10064244e-01
-1.15595236e-01 3.14440876e-01 -6.55955017e-01 2.76580781e-01
8.48311841e-01 -9.18125629e-01 -8.94841731e-01 2.54824251e-01
8.03278029e-01 -1.59496143e-01 -5.73608279e-01 4.19981003e-01
8.96981731e-02 -6.37180924e-01 6.92172110e-01 -8.89583528e-01
3.39370012e-01 2.61913121e-01 -3.37610990e-01 -1.57116961e+00
6.46719456e-01 -5.86821318e-01 1.54012546e-01 1.63965583e+00
1.03937626e+00 -6.45608187e-01 6.07064545e-01 4.64761645e-01
-2.15349302e-01 -7.47847199e-01 -8.12530339e-01 -7.43594885e-01
-2.27636248e-02 -7.48542428e-01 8.60186964e-02 1.11374962e+00
2.18042731e-01 9.29058194e-01 -3.11707675e-01 3.69798280e-02
1.54598549e-01 -1.91330820e-01 8.47826421e-01 -7.44300127e-01
-1.37624182e-02 -1.96819812e-01 4.64811772e-02 -1.15093935e+00
9.81260419e-01 -9.69471931e-01 3.54321122e-01 -1.32859194e+00
-4.58809473e-02 -3.58657807e-01 -2.72805512e-01 7.79025912e-01
-4.58625823e-01 4.78964485e-03 5.13438657e-02 -3.83456387e-02
-8.39138269e-01 1.02016878e+00 6.68963671e-01 3.29358876e-02
-3.51443589e-01 1.55261070e-01 -6.21665895e-01 6.88878119e-01
5.08278728e-01 -2.95296878e-01 -3.32772523e-01 -3.92606050e-01
-1.06916770e-01 5.53883836e-02 -3.32721531e-01 -6.15941703e-01
8.17501992e-02 2.46414751e-01 -2.35862687e-01 -4.95134145e-01
1.97837383e-01 -4.40991610e-01 -1.00263035e+00 2.72688852e-03
-5.10958135e-01 -3.55675191e-01 2.59878904e-01 1.75626934e-01
-6.61718786e-01 -5.27840257e-01 7.28680491e-01 -1.63852677e-01
-8.86154413e-01 -6.98187202e-02 -6.11233890e-01 3.58276486e-01
6.83977067e-01 2.63657510e-01 -3.12318299e-02 -9.78427470e-01
-7.85179496e-01 4.47991759e-01 -1.95945110e-02 6.35314882e-01
4.28366929e-01 -7.70265639e-01 -7.99572885e-01 3.18009555e-01
3.38837594e-01 -1.29804164e-01 1.76294744e-01 7.48624921e-01
-1.11809365e-01 7.28946030e-01 3.47890973e-01 -4.07168716e-01
-1.19432580e+00 2.40197986e-01 1.53917000e-01 -4.80600238e-01
-1.37851954e-01 1.19089484e+00 6.23864383e-02 -1.30642962e+00
3.24320197e-01 -4.26004201e-01 -3.64417821e-01 2.21453831e-01
7.27676332e-01 1.56832159e-01 2.25670815e-01 -7.60712206e-01
-4.85220253e-01 1.32427514e-01 -1.08277872e-01 -1.40155628e-01
1.41883969e+00 -4.93915349e-01 3.15037876e-01 5.71358025e-01
1.43274808e+00 1.25280932e-01 -8.47019076e-01 -5.38995624e-01
4.31012511e-01 2.12581411e-01 2.51788944e-01 -6.94722652e-01
-6.46146894e-01 1.04298317e+00 6.09898120e-02 2.22533673e-01
8.54347527e-01 3.25332016e-01 9.78205204e-01 7.91544259e-01
3.19143116e-01 -1.19414508e+00 1.03785761e-01 1.00925267e+00
7.29827642e-01 -1.79634833e+00 -6.63435340e-01 -4.22514737e-01
-1.06863666e+00 8.39550495e-01 6.59194171e-01 1.01158708e-01
7.58395255e-01 2.90551573e-01 6.50269866e-01 1.40085235e-01
-8.32701802e-01 -5.24536133e-01 3.66699338e-01 2.73880363e-01
6.76618040e-01 -7.46499524e-02 -6.09113798e-02 8.89310777e-01
-2.80755550e-01 -1.35691136e-01 5.04437387e-01 9.87042606e-01
-3.75763357e-01 -1.20121372e+00 -1.90617874e-01 4.98500288e-01
-3.68961155e-01 -4.38800871e-01 -6.50889993e-01 4.81313139e-01
-5.74448764e-01 1.36190987e+00 -1.18488193e-01 -2.19543159e-01
3.18247020e-01 5.74175477e-01 4.53510843e-02 -1.08886325e+00
-5.19242942e-01 1.29834324e-01 6.46753490e-01 -6.10291719e-01
-4.33885485e-01 -5.95199704e-01 -1.16564023e+00 3.55003566e-01
-6.39710784e-01 2.91419566e-01 6.26818061e-01 1.40790963e+00
2.66650736e-01 4.08441544e-01 8.61608863e-01 -5.79889059e-01
-7.40584254e-01 -1.41116834e+00 -2.61820704e-01 3.26823711e-01
2.04730421e-01 -2.34093800e-01 -6.76347613e-01 4.47744550e-03] | [13.990569114685059, 6.955099105834961] |
84232544-dbca-423d-a60c-194b9b60f114 | dudonet-dual-domain-network-for-ct-metal-1 | 1907.00273 | null | https://arxiv.org/abs/1907.00273v1 | https://arxiv.org/pdf/1907.00273v1.pdf | DuDoNet: Dual Domain Network for CT Metal Artifact Reduction | Computed tomography (CT) is an imaging modality widely used for medical diagnosis and treatment. CT images are often corrupted by undesirable artifacts when metallic implants are carried by patients, which creates the problem of metal artifact reduction (MAR). Existing methods for reducing the artifacts due to metallic implants are inadequate for two main reasons. First, metal artifacts are structured and non-local so that simple image domain enhancement approaches would not suffice. Second, the MAR approaches which attempt to reduce metal artifacts in the X-ray projection (sinogram) domain inevitably lead to severe secondary artifact due to sinogram inconsistency. To overcome these difficulties, we propose an end-to-end trainable Dual Domain Network (DuDoNet) to simultaneously restore sinogram consistency and enhance CT images. The linkage between the sigogram and image domains is a novel Radon inversion layer that allows the gradients to back-propagate from the image domain to the sinogram domain during training. Extensive experiments show that our method achieves significant improvements over other single domain MAR approaches. To the best of our knowledge, it is the first end-to-end dual-domain network for MAR. | ['Shaohua Kevin Zhou', 'Wei-An Lin', 'Jiebo Luo', 'Jingdan Zhang', 'Rama Chellappa', 'Xiaohang Sun', 'Haofu Liao', 'Cheng Peng'] | 2019-06-29 | dudonet-dual-domain-network-for-ct-metal | http://openaccess.thecvf.com/content_CVPR_2019/html/Lin_DuDoNet_Dual_Domain_Network_for_CT_Metal_Artifact_Reduction_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Lin_DuDoNet_Dual_Domain_Network_for_CT_Metal_Artifact_Reduction_CVPR_2019_paper.pdf | cvpr-2019-6 | ['metal-artifact-reduction'] | ['medical'] | [ 3.77739400e-01 -3.58459353e-02 1.76549405e-01 -3.30938160e-01
-1.01473987e+00 -1.33386068e-02 1.96003750e-01 -2.19710365e-01
-3.24151963e-01 7.16846883e-01 2.24943325e-01 -2.02503294e-01
-1.77875906e-01 -7.51240909e-01 -6.68986619e-01 -7.15061247e-01
2.17692226e-01 4.93590146e-01 5.09106100e-01 -4.71291542e-02
-9.86226350e-02 4.40571010e-01 -6.71457529e-01 5.57059407e-01
9.49728906e-01 8.24081242e-01 4.09904778e-01 2.47765183e-01
1.20509841e-01 1.01499856e+00 -2.14912996e-01 1.20671406e-01
3.13879728e-01 -7.32752204e-01 -7.01959372e-01 3.09210978e-02
1.56121165e-01 -5.65017045e-01 -4.81759220e-01 1.13608432e+00
7.68908441e-01 -8.21604058e-02 6.61892414e-01 -7.09402144e-01
-7.61893749e-01 5.88085353e-01 -8.13026488e-01 2.18341500e-01
-4.71170954e-02 -2.71760430e-02 7.14112446e-02 -1.06588137e+00
7.04946160e-01 7.45383799e-01 1.16984046e+00 7.67729759e-01
-9.02473450e-01 -5.46955645e-01 -2.83981770e-01 4.95104212e-03
-9.15667832e-01 -1.37728021e-01 1.01032758e+00 -3.45827729e-01
5.97883523e-01 2.34574974e-01 5.26179552e-01 7.82306910e-01
5.25334954e-01 6.21822417e-01 1.21064496e+00 -3.86681080e-01
1.92771778e-01 2.55705602e-03 -2.14905456e-01 6.36550546e-01
2.76930690e-01 1.55582592e-01 -1.34984940e-01 -1.71383306e-01
1.15114689e+00 2.58171886e-01 -6.05513036e-01 -4.07391757e-01
-9.15276527e-01 5.26385009e-01 7.10549891e-01 6.93968773e-01
-7.15797007e-01 2.38383114e-01 3.44311357e-01 2.80912966e-02
5.36621511e-01 3.98013771e-01 -2.29795218e-01 1.79745018e-01
-1.05350220e+00 3.57167944e-02 3.18529665e-01 4.93415713e-01
9.39422250e-02 1.44486070e-01 -9.18959975e-02 1.03099489e+00
3.85976881e-01 5.29939234e-01 8.84331882e-01 -6.57845199e-01
3.43142003e-01 2.10237771e-01 -6.60806298e-02 -6.84930563e-01
-5.03128946e-01 -5.51889658e-01 -1.01061547e+00 5.01577139e-01
4.16437119e-01 -7.91986510e-02 -1.33451736e+00 1.41170967e+00
3.93689364e-01 2.95627505e-01 -1.36022255e-01 1.25646305e+00
1.04018617e+00 3.79947335e-01 1.45573756e-02 -2.53828526e-01
1.36768961e+00 -8.37278545e-01 -9.60778058e-01 -4.11781520e-01
5.12848318e-01 -9.33822632e-01 9.91431832e-01 3.67655694e-01
-1.65200162e+00 -3.22102159e-01 -1.19824731e+00 -1.63775738e-02
1.67443901e-01 -1.10871583e-01 3.09701651e-01 4.96246725e-01
-8.71788263e-01 8.89830470e-01 -1.18882644e+00 2.70911515e-01
7.01433659e-01 3.37585658e-01 -1.14918076e-01 -2.28496104e-01
-1.15518105e+00 1.10291588e+00 -1.61546335e-01 1.86809838e-01
-6.23192608e-01 -1.14620769e+00 -7.87149847e-01 -4.14000452e-01
2.54677743e-01 -6.16566956e-01 1.55069435e+00 -9.46625769e-01
-1.42408907e+00 7.04887927e-01 2.06027236e-02 -3.60488147e-01
8.15361559e-01 -2.80255407e-01 -3.27163160e-01 3.42883736e-01
2.26556718e-01 9.49100181e-02 8.57898235e-01 -1.45302331e+00
-1.53089792e-01 -5.01006007e-01 -7.12344646e-01 2.13229477e-01
-5.13038784e-02 6.13776483e-02 -4.56373274e-01 -1.01469898e+00
7.36334562e-01 -8.72696579e-01 -3.94365698e-01 2.84485459e-01
-1.43600896e-01 2.92794883e-01 8.06254864e-01 -1.07904732e+00
9.47149336e-01 -2.00593042e+00 -2.13615805e-01 1.27403781e-01
3.60432744e-01 1.82849482e-01 8.34540799e-02 -2.16369182e-01
-5.30938268e-01 -2.25577369e-01 -7.11995423e-01 -1.70388460e-01
-6.34947777e-01 -1.44799463e-02 2.24701613e-02 7.53301263e-01
-2.53818512e-01 8.53492200e-01 -8.98866951e-01 -5.40520668e-01
3.33431989e-01 7.39586592e-01 -4.35195297e-01 5.17108664e-02
6.04881160e-02 9.72833276e-01 -5.37156343e-01 3.36043179e-01
1.18353450e+00 -5.40006459e-01 1.12522796e-01 -4.29041147e-01
-5.97520806e-02 3.38737577e-01 -9.15462852e-01 1.80518663e+00
-6.51452422e-01 2.51163006e-01 1.33740574e-01 -7.02533782e-01
3.94398302e-01 7.01181650e-01 1.14211094e+00 -1.22276103e+00
1.71320736e-01 6.77186012e-01 -4.30467725e-02 -5.65630794e-01
-2.50618923e-02 -8.70534420e-01 3.56084108e-01 4.40630466e-01
-4.55150545e-01 -4.89729971e-01 -4.43687648e-01 -1.13752544e-01
1.06327629e+00 1.02645457e-01 1.15561010e-02 -1.78682745e-01
3.28347772e-01 9.15133208e-02 6.14228606e-01 4.71062779e-01
-1.66716188e-01 1.18794012e+00 -1.29831642e-01 -5.51550150e-01
-1.17309856e+00 -1.39817333e+00 -4.29874897e-01 2.16209278e-01
3.18894446e-01 2.61978358e-01 -7.07850695e-01 -6.40689671e-01
-4.17247444e-01 4.41940695e-01 -4.51846510e-01 -2.77211577e-01
-1.15108967e+00 -7.85056949e-01 4.02507901e-01 9.69469965e-01
7.77893722e-01 -1.04384077e+00 -6.20156944e-01 5.62430143e-01
-4.70755547e-01 -1.11725485e+00 -7.19978333e-01 2.49303952e-01
-1.47065306e+00 -9.92381692e-01 -1.25569153e+00 -1.03079128e+00
9.16017234e-01 2.22955108e-01 1.12452459e+00 1.40743017e-01
-4.20701146e-01 -3.14454362e-02 -9.37283859e-02 -1.71930045e-01
-6.16834521e-01 -6.46071494e-01 -2.07764879e-01 -4.53001589e-01
-6.49493560e-02 -5.98349571e-01 -9.74483013e-01 3.22474241e-01
-1.12859333e+00 2.46709570e-01 4.98562306e-01 1.01230323e+00
7.42122710e-01 1.85729355e-01 3.86137873e-01 -9.84524667e-01
4.88662660e-01 -1.47455618e-01 -5.09208083e-01 1.57604516e-01
-4.79651481e-01 1.16889484e-01 5.67181528e-01 -2.94415563e-01
-1.48670077e+00 2.22676173e-01 -4.95256603e-01 -3.25913072e-01
6.98162289e-03 3.75738412e-01 2.37375483e-01 -2.54148871e-01
9.84256148e-01 2.25449339e-01 -3.84412818e-02 -3.85911852e-01
-9.08554122e-02 3.38177800e-01 6.49969757e-01 -2.26431757e-01
7.02171564e-01 8.11062753e-01 1.18555076e-01 -5.98157406e-01
-7.97263980e-01 -3.50974053e-01 -2.85817087e-01 -1.41111180e-01
1.05812144e+00 -8.13533127e-01 -2.05634803e-01 5.81615686e-01
-1.20860207e+00 -3.32384288e-01 -3.91180217e-01 8.05437744e-01
-4.71180797e-01 7.12215781e-01 -1.06793463e+00 -4.15614516e-01
-6.08501971e-01 -1.43230748e+00 8.12361240e-01 1.14654541e-01
-5.00630848e-02 -8.62460434e-01 -3.35655361e-02 2.34669328e-01
5.10461330e-01 2.50663161e-01 7.70953774e-01 -3.36237997e-02
-3.49710941e-01 1.36998489e-01 -3.27996910e-01 4.03767288e-01
3.85318369e-01 -7.58050501e-01 -8.64785194e-01 -2.12117255e-01
9.14956093e-01 -1.61263689e-01 6.50228620e-01 1.09318864e+00
1.39768791e+00 7.63040856e-02 -3.14078450e-01 8.03454578e-01
1.48852432e+00 6.00564361e-01 8.84251654e-01 4.44103301e-01
5.32722056e-01 1.02365471e-01 4.35068101e-01 3.16263735e-01
-7.15439096e-02 6.62476718e-01 2.48305365e-01 -6.24721050e-01
-6.99513018e-01 -3.09806094e-02 -1.29988238e-01 5.64189851e-01
-7.74278492e-02 -1.96868484e-03 -9.46627021e-01 5.58683217e-01
-1.46987510e+00 -6.91447139e-01 -5.67100108e-01 2.06704211e+00
1.02837348e+00 -6.65691644e-02 -3.35263312e-01 1.29851431e-01
5.24918675e-01 -1.66644916e-01 -4.38797802e-01 1.52487919e-01
1.29950345e-01 5.27140439e-01 6.92408800e-01 4.85110968e-01
-9.77755070e-01 3.01744729e-01 6.25582933e+00 5.44246733e-01
-1.25539660e+00 5.74498177e-01 5.05782425e-01 2.11081326e-01
-4.17615056e-01 -3.84416193e-01 -1.27228200e-01 5.43084621e-01
4.11132604e-01 3.91036093e-01 7.14456588e-02 6.35723054e-01
1.99142888e-01 -2.16633260e-01 -8.87945592e-01 1.05768836e+00
-1.62733123e-02 -1.26119137e+00 -5.16277671e-01 -1.58721402e-01
8.37948799e-01 2.01434389e-01 2.60041386e-01 -1.22941792e-01
2.30265170e-01 -8.36310387e-01 5.65280378e-01 2.11684316e-01
9.01856959e-01 -4.94206309e-01 6.81501150e-01 6.11558408e-02
-8.61866057e-01 2.47878835e-01 -1.18940569e-01 5.12005389e-01
5.65927863e-01 1.03059411e+00 -1.01973045e+00 5.40671349e-01
7.41738677e-01 4.78491366e-01 -5.59125617e-02 1.19490695e+00
-2.66041964e-01 3.37640226e-01 -2.69167989e-01 5.26544809e-01
1.71082214e-01 -6.54758811e-02 4.52933609e-01 8.98225546e-01
5.59688449e-01 2.61779398e-01 -1.25830337e-01 9.27328527e-01
-1.42685071e-01 -2.72793710e-01 -5.62121630e-01 5.27860999e-01
-3.99821550e-02 7.54759133e-01 -8.63868594e-01 -3.60952377e-01
-5.26510835e-01 1.24662006e+00 -3.60591829e-01 2.95944780e-01
-1.02688038e+00 2.72024870e-02 -1.17895892e-02 4.93858784e-01
-3.06260511e-02 -1.40752733e-01 -1.02735376e+00 -9.40902054e-01
3.26378733e-01 -6.94617569e-01 3.78745675e-01 -8.44715714e-01
-1.37400305e+00 8.08610201e-01 -2.29749486e-01 -1.55986810e+00
-2.30721384e-02 -4.20210391e-01 -4.54235286e-01 8.07108223e-01
-1.70996678e+00 -8.85800719e-01 -4.98528063e-01 6.30298376e-01
5.61173379e-01 2.43428752e-01 4.70227540e-01 9.11573172e-01
2.09502757e-01 3.65012497e-01 2.08712548e-01 1.04771644e-01
7.61700690e-01 -1.12358010e+00 1.60927445e-01 7.50387251e-01
-3.32864791e-01 3.09927195e-01 6.08623028e-01 -1.06157112e+00
-1.17526615e+00 -9.97788727e-01 5.93162954e-01 -2.24480644e-01
3.11842680e-01 1.13855660e-01 -1.00009799e+00 5.08513272e-01
1.14873454e-01 4.72079992e-01 4.26560044e-01 -5.02628505e-01
1.10448956e-01 4.27836701e-02 -1.46600521e+00 5.52089214e-01
6.80992365e-01 -2.48213202e-01 -6.54512703e-01 4.78138834e-01
4.85796750e-01 -8.78823400e-01 -7.70136654e-01 6.80447638e-01
3.29600751e-01 -8.54031265e-01 1.19775760e+00 -8.30448791e-02
8.33194733e-01 -3.36604238e-01 2.28983164e-01 -1.47285962e+00
-3.35567951e-01 -1.70049429e-01 2.45619521e-01 5.10450006e-01
3.63749981e-01 -5.04918098e-01 9.47586417e-01 5.58769464e-01
-6.26604795e-01 -4.01019454e-01 -1.02162278e+00 -7.24368453e-01
2.26064712e-01 -4.79472250e-01 3.71815085e-01 1.08077312e+00
-2.09079638e-01 -1.13120735e-01 -3.61165166e-01 3.38763028e-01
7.92762339e-01 -1.10221811e-01 -5.24740480e-02 -9.25422490e-01
-4.88822728e-01 -2.67292321e-01 1.25359178e-01 -9.93915677e-01
-5.62158167e-01 -8.89663696e-01 5.71196854e-01 -1.82546544e+00
2.43241429e-01 -7.32424438e-01 -3.18433374e-01 2.70153791e-01
-8.97577628e-02 4.36755002e-01 -1.04926161e-01 2.32074007e-01
-8.63929000e-03 4.83030826e-01 1.98447716e+00 -8.74154121e-02
-2.99095035e-01 7.22363591e-02 -4.43673790e-01 9.73982215e-01
8.49310338e-01 -7.38474131e-01 -3.66650403e-01 -8.50910008e-01
7.98398852e-02 5.15399814e-01 4.44172233e-01 -1.10318661e+00
2.36337662e-01 4.94437888e-02 6.70624256e-01 -7.22841978e-01
3.17482799e-01 -1.16394174e+00 2.45505363e-01 8.54987025e-01
2.19133566e-03 -4.19205353e-02 2.99926788e-01 2.13132128e-01
-3.73192549e-01 -2.57875919e-01 1.46224964e+00 -4.37971950e-01
-1.88758880e-01 3.00352305e-01 -4.36788529e-01 -4.74238768e-02
8.82047296e-01 -2.18475759e-01 4.67801988e-02 -2.25122452e-01
-7.87575424e-01 -2.20894501e-01 2.58111775e-01 4.25456539e-02
1.09523904e+00 -1.40770090e+00 -6.36820138e-01 1.90527245e-01
-3.02578062e-01 3.71704429e-01 6.54381752e-01 1.21613181e+00
-9.21198130e-01 1.42292410e-01 -1.17289096e-01 -6.09819174e-01
-1.12528348e+00 3.79093707e-01 7.96271205e-01 -6.24356985e-01
-1.23166108e+00 9.89637256e-01 4.16994423e-01 -2.96782464e-01
-2.07014550e-02 -4.75535542e-01 3.78730893e-01 -6.94478333e-01
6.44672215e-01 2.01342568e-01 6.01208985e-01 -2.47492194e-01
-3.93738002e-01 6.70651913e-01 -1.78130403e-01 -2.37050742e-01
1.68120289e+00 7.80433118e-02 -5.37956879e-02 -2.05035344e-01
1.01991403e+00 -1.78274676e-01 -1.16107965e+00 -3.11210990e-01
-2.29085520e-01 -4.74358469e-01 4.04322565e-01 -1.06737220e+00
-1.47311902e+00 8.73577714e-01 1.05074942e+00 -1.97582752e-01
1.41480267e+00 -9.09428298e-02 1.28476083e+00 -1.65930361e-01
2.07297221e-01 -1.15063655e+00 4.15402114e-01 2.21134618e-01
9.86811459e-01 -1.11417723e+00 1.57072708e-01 -6.96695983e-01
-7.11374104e-01 1.03808928e+00 6.15871608e-01 -8.03545490e-02
7.76880920e-01 7.64622986e-01 2.38635987e-01 -3.89834285e-01
4.04076464e-02 3.39216858e-01 2.77445257e-01 6.33751869e-01
7.44110763e-01 -1.89785913e-01 -5.34583569e-01 4.02484179e-01
3.28708261e-01 4.30847168e-01 4.16302651e-01 1.26838529e+00
-3.06360275e-01 -1.13880444e+00 -8.29789162e-01 2.76926249e-01
-7.01229095e-01 -9.88685712e-02 1.51265591e-01 5.65152407e-01
3.55602987e-02 6.66779518e-01 -3.39661181e-01 1.81569651e-01
4.97241974e-01 -3.54752600e-01 7.34978497e-01 -4.57073689e-01
-6.90478921e-01 4.64412421e-01 -2.76864946e-01 -3.91006708e-01
-3.05231273e-01 -4.46626753e-01 -1.84644651e+00 9.73403677e-02
-1.96952790e-01 -8.48543122e-02 6.82299674e-01 7.95012474e-01
-3.20649147e-02 1.10073864e+00 5.33253193e-01 -5.64472675e-01
-4.73651111e-01 -8.10177982e-01 -6.10069931e-01 6.45290434e-01
3.52466106e-01 -5.16426325e-01 -5.54668456e-02 1.05424672e-01] | [13.483664512634277, -2.5426745414733887] |
b9fb0732-56d3-45a5-97eb-91a325a27e77 | hypergraph-convolutional-network-based-weakly | 2211.01174 | null | https://arxiv.org/abs/2211.01174v1 | https://arxiv.org/pdf/2211.01174v1.pdf | Hypergraph Convolutional Network based Weakly Supervised Point Cloud Semantic Segmentation with Scene-Level Annotations | Point cloud segmentation with scene-level annotations is a promising but challenging task. Currently, the most popular way is to employ the class activation map (CAM) to locate discriminative regions and then generate point-level pseudo labels from scene-level annotations. However, these methods always suffer from the point imbalance among categories, as well as the sparse and incomplete supervision from CAM. In this paper, we propose a novel weighted hypergraph convolutional network-based method, called WHCN, to confront the challenges of learning point-wise labels from scene-level annotations. Firstly, in order to simultaneously overcome the point imbalance among different categories and reduce the model complexity, superpoints of a training point cloud are generated by exploiting the geometrically homogeneous partition. Then, a hypergraph is constructed based on the high-confidence superpoint-level seeds which are converted from scene-level annotations. Secondly, the WHCN takes the hypergraph as input and learns to predict high-precision point-level pseudo labels by label propagation. Besides the backbone network consisting of spectral hypergraph convolution blocks, a hyperedge attention module is learned to adjust the weights of hyperedges in the WHCN. Finally, a segmentation network is trained by these pseudo point cloud labels. We comprehensively conduct experiments on the ScanNet and S3DIS segmentation datasets. Experimental results demonstrate that the proposed WHCN is effective to predict the point labels with scene annotations, and yields state-of-the-art results in the community. The source code is available at http://zhiyongsu.github.io/Project/WHCN.html. | ['Zhiyong Su', 'Weiqing Li', 'Yuewei Dai', 'Peng Zhang', 'Zhuheng Lu'] | 2022-11-02 | null | null | null | null | ['point-cloud-segmentation'] | ['computer-vision'] | [ 2.31563553e-01 2.86506772e-01 -2.48573467e-01 -5.03039002e-01
-8.30402434e-01 -1.86334148e-01 1.71310544e-01 -5.36105707e-02
-2.36973122e-01 2.37379104e-01 -4.17315781e-01 2.04147156e-02
5.01246043e-02 -1.04887521e+00 -9.61418808e-01 -5.87278187e-01
1.90447688e-01 7.50104070e-01 7.90787101e-01 -1.19542060e-02
3.17228347e-01 5.98110616e-01 -1.50676334e+00 5.39854430e-02
1.12027383e+00 1.10381305e+00 5.73064506e-01 1.43404350e-01
-5.77662826e-01 4.16033238e-01 -6.22512288e-02 -1.92490429e-01
3.78463656e-01 -1.36579275e-02 -7.90486574e-01 3.45708430e-01
4.63681251e-01 -1.10610105e-01 -2.52962440e-01 1.29519546e+00
1.51251838e-01 3.33414555e-01 5.47691762e-01 -1.26706827e+00
-3.26442927e-01 3.97726774e-01 -9.93511677e-01 -1.82655126e-01
-1.64539963e-01 2.33505011e-01 9.42443848e-01 -1.12321568e+00
5.21434724e-01 1.14981544e+00 6.01902068e-01 3.43051255e-01
-1.07667077e+00 -1.02744627e+00 3.96356851e-01 2.41214886e-01
-1.69607687e+00 9.92515683e-02 1.21976745e+00 -5.01037836e-01
5.71027458e-01 -3.46166454e-02 8.13002765e-01 5.76268971e-01
-3.43252897e-01 7.97230721e-01 6.46457255e-01 -1.86275601e-01
1.52768210e-01 -1.44143909e-01 1.79875284e-01 8.91683996e-01
2.90054083e-02 -2.26787150e-01 -2.10094035e-01 -4.33070883e-02
1.08021462e+00 3.63340139e-01 -4.11565185e-01 -5.64696014e-01
-1.20887136e+00 8.22651386e-01 1.23726380e+00 1.34575993e-01
-3.98662359e-01 1.59475520e-01 1.20121069e-01 -4.94429648e-01
6.18813932e-01 2.66760796e-01 -3.75014305e-01 5.28723061e-01
-1.04046965e+00 -3.85421030e-02 4.12346840e-01 1.30345821e+00
1.27503061e+00 -1.52054086e-01 -8.77462029e-02 9.51730609e-01
5.72035074e-01 4.37654406e-01 4.81000394e-02 -9.81171429e-01
5.71636498e-01 1.10967755e+00 -2.38494828e-01 -1.07272935e+00
-4.48743910e-01 -6.01418972e-01 -8.20529699e-01 1.88709021e-01
1.04372561e-01 1.36150777e-01 -1.52565300e+00 1.23259771e+00
6.09492064e-01 7.84950912e-01 -4.01377618e-01 1.22995329e+00
1.02896714e+00 8.49279583e-01 2.48285949e-01 1.78958207e-01
1.16215742e+00 -1.17684078e+00 -1.70323491e-01 -3.04934919e-01
3.45611393e-01 -4.73315090e-01 1.01433575e+00 1.57649107e-02
-8.88221145e-01 -7.15372801e-01 -8.38861346e-01 -1.48184508e-01
-2.16241732e-01 9.42770019e-02 5.75978041e-01 6.15637749e-02
-9.31721747e-01 3.52807820e-01 -1.00185633e+00 -7.43924677e-02
8.32220912e-01 3.88236821e-01 -1.42687075e-02 -2.44829267e-01
-9.26896691e-01 2.70671278e-01 8.02328169e-01 2.42000729e-01
-6.88751936e-01 -9.70608294e-01 -1.01427317e+00 2.06240252e-01
4.54225779e-01 -6.13520145e-01 1.09713364e+00 -8.90597761e-01
-1.15624380e+00 8.82114470e-01 9.59957540e-02 7.22387880e-02
3.81295770e-01 2.31601089e-01 -1.05347902e-01 2.81882197e-01
4.60886478e-01 1.20379782e+00 5.70573330e-01 -1.73078012e+00
-1.00258827e+00 -3.43500286e-01 -1.27880961e-01 2.63868779e-01
3.65677476e-02 -4.49763834e-01 -1.15523648e+00 -3.72208118e-01
6.79536998e-01 -9.15352523e-01 -5.95800042e-01 -7.49732628e-02
-7.77986467e-01 -4.35552835e-01 8.28057408e-01 -3.77349257e-01
8.43755662e-01 -2.19355798e+00 4.14019935e-02 6.40233397e-01
2.94446379e-01 6.01880513e-02 -4.59754542e-02 -6.18069805e-02
-7.38465041e-02 1.53298508e-02 -8.60767365e-01 -2.82701880e-01
-1.38409972e-01 1.85409009e-01 -1.38490379e-01 5.09947956e-01
3.53388906e-01 9.05788124e-01 -9.96054709e-01 -7.68330336e-01
6.59339011e-01 5.98836005e-01 -5.22818923e-01 1.60294726e-01
-5.69216490e-01 5.93948126e-01 -6.85108244e-01 8.19015563e-01
1.04917955e+00 -5.52527428e-01 -3.27869296e-01 -3.58030289e-01
-2.05795199e-01 -1.87826067e-01 -1.12769282e+00 1.95166123e+00
-1.96260542e-01 3.51185054e-01 -7.85599723e-02 -8.08169305e-01
9.66279864e-01 -1.25259891e-01 6.58368409e-01 -4.21219468e-01
1.69795513e-01 2.92711943e-01 -2.31987283e-01 -2.75640905e-01
3.64867777e-01 -1.57738943e-02 6.67306632e-02 -2.20851868e-01
1.25834092e-01 -4.67497557e-01 -8.28973297e-03 1.57337740e-01
8.03326726e-01 1.20465145e-01 -2.43869811e-01 -7.94562697e-02
6.68171465e-01 4.90278602e-01 8.43292713e-01 3.98954481e-01
1.54800832e-01 1.07235789e+00 2.39749551e-01 -4.13573742e-01
-1.06408715e+00 -1.01594365e+00 -3.49568516e-01 6.76217139e-01
8.31672549e-01 -2.53657550e-01 -8.18521440e-01 -7.07934976e-01
1.95355583e-02 6.14169896e-01 -4.10834670e-01 -1.17462039e-01
-6.07429266e-01 -5.33538461e-01 1.56824306e-01 6.17420375e-01
7.60899067e-01 -1.21855414e+00 -3.34072679e-01 1.51417866e-01
-2.34361097e-01 -1.11643100e+00 -4.17403758e-01 9.26464610e-03
-8.75447273e-01 -1.13476861e+00 -7.62784660e-01 -1.06967151e+00
9.61761296e-01 4.01031315e-01 1.10609436e+00 3.47782075e-01
-1.18375517e-01 4.33216169e-02 -3.73954922e-01 -1.89321294e-01
2.08533660e-01 3.94768268e-01 -3.96591038e-01 1.12517945e-01
3.24917704e-01 -5.10510981e-01 -8.62798631e-01 4.92244631e-01
-8.65952075e-01 3.27510118e-01 7.45732188e-01 6.03889704e-01
1.28874660e+00 1.13180578e-01 2.76211560e-01 -1.20387530e+00
-2.52378106e-01 -6.23341143e-01 -9.41063702e-01 5.00754565e-02
-3.57619882e-01 -2.34836146e-01 3.64782572e-01 -1.20671973e-01
-1.08688033e+00 5.56440711e-01 -4.51323241e-01 -8.37226868e-01
-3.42081785e-01 4.73782480e-01 -3.05149287e-01 -3.10229391e-01
3.29174876e-01 4.48914133e-02 -5.16576111e-01 -4.02370185e-01
4.60565388e-01 4.91400629e-01 6.73436046e-01 -4.84882116e-01
1.00835407e+00 6.06563389e-01 -7.32082874e-02 -7.63802886e-01
-1.03502154e+00 -9.02834475e-01 -6.91979110e-01 -4.79196250e-01
1.30098057e+00 -1.02960896e+00 -3.06090713e-01 3.15460265e-01
-1.21232295e+00 -4.86415178e-01 -3.06011736e-01 4.71848905e-01
-5.22743881e-01 2.04116300e-01 -5.65863788e-01 -4.29632664e-01
-1.73689350e-01 -1.21848238e+00 1.39842629e+00 6.42656982e-01
4.23289686e-01 -6.55059278e-01 -1.79585278e-01 4.05089736e-01
-2.52382368e-01 4.24882948e-01 8.11076641e-01 -2.88284123e-01
-1.17992175e+00 -1.29270837e-01 -6.89864814e-01 1.63149580e-01
-1.45867497e-01 1.45091340e-02 -9.05349314e-01 -7.99524188e-02
-2.35254541e-01 -1.13546461e-01 8.95219266e-01 5.48296392e-01
1.78099012e+00 1.72918648e-01 -5.98819971e-01 1.10160959e+00
1.56711257e+00 -3.77383968e-03 5.60321867e-01 2.58590281e-01
1.40015364e+00 4.87421304e-01 6.73413992e-01 1.78036451e-01
5.04105330e-01 5.38956106e-01 7.32465148e-01 -4.25460577e-01
-2.22046420e-01 -4.97039735e-01 -2.86700308e-01 8.06324840e-01
-1.25219047e-01 -1.28157765e-01 -1.21296251e+00 5.09842694e-01
-1.87319636e+00 -5.90574384e-01 -6.62443459e-01 1.95326960e+00
5.56579292e-01 2.17308074e-01 -1.47010908e-01 -2.37709627e-01
1.09405625e+00 1.56726539e-01 -8.08952093e-01 3.44487727e-01
6.41715154e-02 9.80464742e-02 7.83797443e-01 3.82687151e-01
-1.24743187e+00 1.27649522e+00 4.07804537e+00 1.07148385e+00
-9.04908657e-01 2.38519058e-01 7.37372041e-01 2.38202050e-01
-2.61613399e-01 5.42027764e-02 -7.54517376e-01 6.54010653e-01
3.02198797e-01 2.29375750e-01 7.56724700e-02 1.12326252e+00
-6.73695579e-02 1.30501658e-01 -7.13503063e-01 1.14390683e+00
4.36971299e-02 -1.35697007e+00 -7.77650774e-02 -2.63619609e-02
1.07223916e+00 5.75890183e-01 -2.20241562e-01 3.04383487e-01
3.21792543e-01 -7.39223957e-01 6.50725603e-01 4.92795885e-01
1.00752270e+00 -9.70766783e-01 7.11260140e-01 3.68194878e-01
-1.51023662e+00 9.51434374e-02 -6.81358516e-01 4.27066475e-01
3.09146971e-01 7.24141836e-01 -5.89445591e-01 6.29413068e-01
8.55586886e-01 8.37630451e-01 -5.22528470e-01 1.49942625e+00
-4.13860530e-01 5.46615839e-01 -4.12238151e-01 3.91729444e-01
4.94957089e-01 -3.79851818e-01 2.70568430e-01 1.20199430e+00
3.50452602e-01 3.92548323e-01 6.18980587e-01 1.20523322e+00
-2.17791602e-01 1.45553937e-02 -2.93720275e-01 4.82313007e-01
3.30990076e-01 1.69157720e+00 -1.23979485e+00 -2.78878152e-01
-3.59910667e-01 8.05076897e-01 5.27202368e-01 5.21223068e-01
-8.24463546e-01 -3.78784835e-01 4.14041936e-01 2.15619549e-01
1.75823510e-01 -2.36503720e-01 -4.70336646e-01 -1.04259288e+00
-3.50172147e-02 -1.80864513e-01 3.18113744e-01 -9.28943455e-01
-1.32814419e+00 6.63412631e-01 9.00026411e-02 -1.44980705e+00
3.19983661e-01 -3.42289567e-01 -7.75487721e-01 9.53536093e-01
-1.64422011e+00 -1.14749432e+00 -8.58615696e-01 4.17676240e-01
8.34405303e-01 3.07533115e-01 1.70868725e-01 3.98157358e-01
-6.87500298e-01 2.12336898e-01 -2.56862313e-01 2.64976591e-01
2.44010359e-01 -1.28082919e+00 3.39102536e-01 6.57249749e-01
3.44882370e-03 1.89226255e-01 1.45848528e-01 -9.84684944e-01
-7.33708382e-01 -1.61583042e+00 4.76052523e-01 -3.21332902e-01
4.34416771e-01 -4.72981721e-01 -1.30794001e+00 4.52550471e-01
-1.78752914e-01 3.43194902e-01 1.88007414e-01 -2.73705125e-01
-8.55357498e-02 4.84007075e-02 -9.90273535e-01 4.09491897e-01
1.24274957e+00 -3.60967278e-01 -3.62554371e-01 6.78385556e-01
1.09066296e+00 -7.80981779e-01 -7.11334705e-01 7.51457632e-01
-1.05676882e-01 -6.23918235e-01 1.06548083e+00 -1.16165936e-01
6.00981832e-01 -7.18961120e-01 1.35872468e-01 -1.19921124e+00
-5.11624873e-01 2.19160691e-01 2.01542407e-01 1.28258491e+00
4.76532251e-01 -3.65011871e-01 1.29505372e+00 6.99984789e-01
-7.70638287e-01 -8.02585661e-01 -7.95760095e-01 -5.43222785e-01
-1.25813991e-01 -4.86680716e-01 6.64480865e-01 1.05509913e+00
-6.45406604e-01 4.02530551e-01 1.54403672e-01 6.58479214e-01
8.43276858e-01 2.54811436e-01 5.94430327e-01 -1.43970454e+00
1.07560590e-01 -4.74684119e-01 -5.17117381e-01 -1.10611606e+00
4.11763966e-01 -1.30256522e+00 4.54092681e-01 -1.79265368e+00
9.95884836e-02 -9.49139714e-01 -3.20550859e-01 4.02764380e-01
-3.69150341e-01 5.16492903e-01 1.35321170e-01 4.68503892e-01
-7.12279260e-01 7.14800715e-01 1.36071539e+00 -3.62751156e-01
-3.18195105e-01 5.13406657e-02 -2.26092830e-01 9.16464269e-01
9.02982354e-01 -5.57173729e-01 -4.88335997e-01 -5.44534981e-01
3.40632349e-02 -1.63647652e-01 5.72445869e-01 -1.21321023e+00
4.95633870e-01 -1.98266879e-01 6.04170024e-01 -1.13304222e+00
3.81584048e-01 -1.02205169e+00 1.24229886e-01 1.48116767e-01
-5.15815131e-02 -2.34768108e-01 -2.83263866e-02 7.36027360e-01
-1.57711461e-01 -2.13207677e-01 8.17527533e-01 -2.40504250e-01
-1.08450222e+00 9.27023470e-01 3.23724121e-01 -3.32341231e-02
1.09141588e+00 -2.96148390e-01 -1.47805512e-01 1.36746138e-01
-5.42028248e-01 8.50130916e-01 5.75730383e-01 3.09451610e-01
8.69705737e-01 -1.31956255e+00 -5.03296673e-01 3.59254539e-01
3.98843795e-01 1.12006199e+00 6.33632183e-01 7.65748501e-01
-7.73070455e-01 9.56147537e-02 -2.75498838e-03 -1.09857416e+00
-7.57798433e-01 5.27554750e-01 3.33432794e-01 2.60708660e-01
-9.93618250e-01 1.10578620e+00 6.09429061e-01 -6.72855914e-01
2.23019972e-01 -3.43952864e-01 -2.42238507e-01 -2.21472964e-01
8.66990685e-02 9.65598002e-02 9.05435532e-02 -9.05547261e-01
-3.63130569e-01 9.70270514e-01 3.33189145e-02 1.24014914e-01
1.18945909e+00 1.24501154e-01 -9.60599482e-02 2.91096061e-01
1.11005521e+00 -4.19158369e-01 -1.53275001e+00 -2.75190294e-01
-1.19468562e-01 -6.15939856e-01 2.35729769e-01 -4.66263890e-01
-1.56142807e+00 8.42293024e-01 5.53565383e-01 -6.37121499e-02
1.00069511e+00 3.77082258e-01 8.97137344e-01 8.74734297e-02
4.63789284e-01 -9.23076928e-01 -1.92654818e-01 3.18194687e-01
6.84164941e-01 -1.33939445e+00 -2.01659173e-01 -1.00629377e+00
-4.76745605e-01 8.59909296e-01 1.14839482e+00 -4.01269048e-01
7.23934650e-01 -2.55866081e-01 -5.66445803e-03 -5.69190085e-01
-2.16066744e-02 -4.35463876e-01 3.50301594e-01 5.71639836e-01
-1.89650469e-02 1.41199484e-01 -1.36095667e-02 4.02921140e-01
-6.11744672e-02 -1.43040016e-01 1.66933298e-01 5.31827271e-01
-5.32636344e-01 -8.66918921e-01 -3.32779646e-01 7.19093084e-01
1.37190804e-01 -5.67640327e-02 -2.64211476e-01 6.44775033e-01
5.08655071e-01 7.01424301e-01 3.40034753e-01 -6.23323321e-01
4.64186966e-01 -2.65424907e-01 -2.20255479e-02 -9.52265799e-01
-3.32132161e-01 1.18786126e-01 -3.98588717e-01 -6.30484402e-01
-3.87402058e-01 -3.93220693e-01 -1.82619143e+00 -3.73998508e-02
-5.31241715e-01 2.61058688e-01 5.74608743e-01 6.64084911e-01
2.34376758e-01 7.02080190e-01 6.79241419e-01 -1.30237830e+00
1.97382629e-01 -7.04435349e-01 -4.66296375e-01 3.55908841e-01
7.81991258e-02 -8.37025046e-01 -2.59952217e-01 -1.74458966e-01] | [8.063189506530762, -3.1577274799346924] |
ff8e5365-d229-4d32-ad81-48b96ca5db16 | graph-matching-with-bi-level-noisy | 2212.04085 | null | https://arxiv.org/abs/2212.04085v2 | https://arxiv.org/pdf/2212.04085v2.pdf | Graph Matching with Bi-level Noisy Correspondence | In this paper, we study a novel and widely existing problem in graph matching (GM), namely, Bi-level Noisy Correspondence (BNC), which refers to node-level noisy correspondence (NNC) and edge-level noisy correspondence (ENC). In brief, on the one hand, due to the poor recognizability and viewpoint differences between images, it is inevitable to inaccurately annotate some keypoints with offset and confusion, leading to the mismatch between two associated nodes, i.e., NNC. On the other hand, the noisy node-to-node correspondence will further contaminate the edge-to-edge correspondence, thus leading to ENC. For the BNC challenge, we propose a novel method termed Contrastive Matching with Momentum Distillation. Specifically, the proposed method is with a robust quadratic contrastive loss which enjoys the following merits: i) better exploring the node-to-node and edge-to-edge correlations through a GM customized quadratic contrastive learning paradigm; ii) adaptively penalizing the noisy assignments based on the confidence estimated by the momentum teacher. Extensive experiments on three real-world datasets show the robustness of our model compared with 12 competitive baselines. | ['Xi Peng', 'Changqing Zhang', 'Peng Hu', 'Jun Yu', 'Mouxing Yang', 'Yijie Lin'] | 2022-12-08 | null | null | null | null | ['graph-matching'] | ['graphs'] | [ 9.61165354e-02 2.93526542e-03 1.07692992e-02 -2.59955168e-01
-9.37966883e-01 -2.10607156e-01 4.06738341e-01 9.82659906e-02
-1.44650057e-01 4.24700141e-01 8.64117816e-02 3.71467099e-02
-2.34775037e-01 -5.79875112e-01 -6.74782872e-01 -8.80905271e-01
6.33373559e-02 2.58599669e-01 2.15054378e-01 -1.11435495e-01
1.36937112e-01 3.58506888e-01 -1.10571909e+00 -5.67292534e-02
1.00721753e+00 1.21151054e+00 2.44971886e-01 1.31770581e-01
-1.40652940e-01 4.35259163e-01 -3.29895347e-01 -6.46465063e-01
3.66175890e-01 -3.20842832e-01 -4.28741276e-01 2.52832137e-02
7.01855421e-01 -1.72190249e-01 -5.07217586e-01 1.58680987e+00
4.51857179e-01 9.79609862e-02 3.73058528e-01 -1.66251671e+00
-5.89747310e-01 3.53683531e-01 -1.06238675e+00 7.16004670e-02
2.30322734e-01 2.07311392e-01 1.07532716e+00 -1.07117403e+00
6.35815680e-01 1.39969409e+00 8.90987039e-01 2.67180651e-01
-1.15024376e+00 -7.61969984e-01 3.88183117e-01 3.02213341e-01
-1.77077115e+00 -1.74267858e-01 1.21156788e+00 -2.49700502e-01
3.16195339e-01 1.74489260e-01 5.45234501e-01 8.24400842e-01
1.17126167e-01 5.52978396e-01 1.06930697e+00 -8.95522162e-02
-1.21007495e-01 6.65159076e-02 -2.93361157e-01 7.82898009e-01
2.12136805e-01 4.04944539e-01 -2.50325620e-01 -2.30546311e-01
6.54937983e-01 -7.34135555e-03 -4.79662955e-01 -4.43814903e-01
-1.23881686e+00 5.33963382e-01 1.01588798e+00 1.91426516e-01
-2.86237866e-01 2.13311464e-01 2.82862127e-01 2.03387767e-01
5.47946870e-01 3.52545381e-02 -9.71086249e-02 1.36357591e-01
-9.12246943e-01 2.32084408e-01 5.75628161e-01 1.19939613e+00
9.67840433e-01 -2.72258967e-01 -2.87018538e-01 7.26486683e-01
3.69087934e-01 2.89493710e-01 1.68792665e-01 -7.58408546e-01
8.82681727e-01 6.68491483e-01 -4.36275937e-02 -1.88183439e+00
-2.98087418e-01 -7.47496188e-01 -1.33199310e+00 -2.44646259e-02
2.63755530e-01 1.85417712e-01 -6.21781945e-01 1.92520154e+00
4.52866405e-01 6.96765423e-01 -3.17535013e-01 1.03521061e+00
1.16587234e+00 4.51232165e-01 5.71998507e-02 -1.00871712e-01
1.07956815e+00 -1.19225824e+00 -6.76160514e-01 -1.08201653e-01
3.21218073e-01 -8.96993458e-01 7.96579897e-01 -9.12236422e-02
-1.03903830e+00 -6.60167694e-01 -1.02636838e+00 5.25295958e-02
-3.17269772e-01 1.14276007e-01 3.25922370e-01 2.29926571e-01
-1.03701222e+00 6.57254755e-01 -5.33898950e-01 -8.87297615e-02
2.88948834e-01 9.98503268e-02 -3.60281944e-01 -2.61704236e-01
-1.16618478e+00 6.78456843e-01 3.64408344e-01 5.22523046e-01
-3.58819008e-01 -7.09561586e-01 -7.93098927e-01 1.27024606e-01
6.43577635e-01 -5.87120593e-01 6.88049972e-01 -7.62670875e-01
-1.01503575e+00 7.88580179e-01 -1.86018035e-01 1.09213561e-01
9.94151592e-01 3.38502415e-02 -4.79823977e-01 1.87469255e-02
3.19073558e-01 8.24232101e-01 6.38803184e-01 -1.50376046e+00
-5.83311975e-01 -3.67345124e-01 -1.30854800e-01 3.52154344e-01
-1.69253811e-01 -2.43567333e-01 -9.66386855e-01 -1.04986870e+00
7.28793859e-01 -8.86223614e-01 -1.10238537e-01 5.17907858e-01
-5.42243123e-01 -3.94869298e-01 9.58610058e-01 -6.56976044e-01
1.12093961e+00 -2.27967715e+00 1.30529493e-01 4.44855720e-01
5.58897018e-01 3.13732289e-02 -4.18924600e-01 3.24208975e-01
-1.59550875e-01 9.97056141e-02 -1.45444021e-01 -5.62843084e-01
-1.74973577e-01 1.49483353e-01 7.50558078e-02 6.19287431e-01
2.75733918e-01 1.11150801e+00 -1.32984662e+00 -8.44247520e-01
6.35554567e-02 3.86094153e-01 -2.78026342e-01 1.63542524e-01
1.94297254e-01 4.59359676e-01 -3.61584306e-01 8.07721496e-01
1.22896886e+00 -3.15367401e-01 -2.10117728e-01 -7.40044832e-01
2.11719811e-01 -1.41029045e-01 -1.45530951e+00 1.50903141e+00
-1.27715364e-01 3.68890345e-01 3.97433043e-01 -1.06248510e+00
9.54952955e-01 1.59671903e-01 4.11623836e-01 -6.31566048e-01
-8.01514601e-04 4.35876220e-01 -2.26360783e-01 -1.61779284e-01
6.63557529e-01 9.36985835e-02 5.89620061e-02 2.89672650e-02
-1.39459759e-01 1.11911394e-01 -1.66829482e-01 4.49049294e-01
8.36632490e-01 2.92589795e-02 1.31372750e-01 -3.64679992e-01
6.49304748e-01 -2.84640461e-01 9.38456237e-01 6.37701273e-01
-4.87901896e-01 8.71671677e-01 4.55018044e-01 -1.05221085e-01
-7.56829202e-01 -9.21403289e-01 1.68113206e-02 4.95343953e-01
8.93804073e-01 -2.92537451e-01 -5.36464274e-01 -7.32510090e-01
7.34937713e-02 1.14131682e-01 -5.19002438e-01 -3.61804008e-01
-7.69446075e-01 -3.93052787e-01 3.24688107e-01 6.13283753e-01
7.06489980e-01 -7.42725492e-01 3.88912976e-01 2.67678827e-01
-4.16007578e-01 -1.20180368e+00 -1.22678816e+00 -1.84453845e-01
-7.12565005e-01 -1.22513580e+00 -9.49255884e-01 -9.42335486e-01
1.07153368e+00 7.15475798e-01 1.19945717e+00 5.50883949e-01
-2.06096813e-01 1.08294614e-01 -2.89629012e-01 1.39613017e-01
-1.17008232e-01 -1.92138195e-01 -1.16374262e-01 1.46538066e-02
5.58929145e-02 -5.04782498e-01 -7.32321203e-01 6.51038408e-01
-5.95050633e-01 7.61768445e-02 7.01756060e-01 1.04940557e+00
8.53750646e-01 7.25322589e-02 3.25323999e-01 -7.56209433e-01
5.46796679e-01 -5.23480356e-01 -6.43363118e-01 5.12578309e-01
-8.07534873e-01 -2.23520130e-01 5.72471023e-01 -5.05997837e-01
-7.46482968e-01 -3.48158255e-02 -1.42870117e-02 -7.31870294e-01
2.76839495e-01 4.39302891e-01 -3.97549629e-01 -6.45383537e-01
1.02577627e-01 2.29829505e-01 -1.58372432e-01 -3.25480610e-01
4.08527792e-01 3.82610649e-01 8.66264582e-01 -6.48319185e-01
1.25096488e+00 3.84102643e-01 2.45898560e-01 -2.93097615e-01
-6.05771959e-01 -6.45519376e-01 -4.26473498e-01 -2.90842056e-01
5.17313004e-01 -1.02605057e+00 -6.08413756e-01 7.17175305e-01
-1.21506047e+00 1.75162733e-01 1.12240808e-02 3.93165380e-01
-2.25852236e-01 7.22886860e-01 -6.27567053e-01 -5.45384407e-01
-4.41652447e-01 -1.20433247e+00 1.10978603e+00 4.36661899e-01
1.96041316e-01 -9.96346712e-01 -2.52747864e-01 1.46658480e-01
4.20368791e-01 4.63538796e-01 7.11812317e-01 -3.33906442e-01
-7.52597868e-01 -3.26365113e-01 -8.66457701e-01 2.46775195e-01
5.00081442e-02 2.02926137e-02 -6.10229969e-01 -5.57863295e-01
-2.81378299e-01 -1.36663914e-01 6.94692910e-01 1.75932765e-01
1.12766588e+00 -1.08681418e-01 -5.52755952e-01 8.41712892e-01
1.42507267e+00 -8.71443301e-02 5.11400044e-01 8.40050057e-02
9.86369133e-01 5.59954226e-01 7.83855081e-01 8.75248536e-02
5.68892121e-01 9.75779653e-01 4.29937482e-01 -3.11262816e-01
-4.31926757e-01 -7.10483849e-01 -2.33849347e-01 1.00204790e+00
2.37320706e-01 -2.27032259e-01 -3.75410199e-01 4.88898456e-01
-2.05478859e+00 -5.70235610e-01 -3.14507544e-01 2.13334489e+00
5.47803760e-01 1.23934560e-01 -1.88119039e-01 -2.15499341e-01
1.23195362e+00 3.63078445e-01 -4.48274523e-01 2.69478798e-01
-2.87303448e-01 -2.55502194e-01 4.61043239e-01 3.31770360e-01
-9.38601315e-01 6.16097808e-01 4.93111610e+00 1.14395678e+00
-9.43429291e-01 4.24103327e-02 8.37038457e-01 3.29769760e-01
-4.18695688e-01 1.09236300e-01 -7.03698635e-01 9.65186179e-01
-8.53057429e-02 -8.73665288e-02 5.15275359e-01 7.19604790e-01
6.64380132e-05 1.67147577e-01 -9.88261461e-01 1.23322964e+00
2.83228587e-02 -1.16271460e+00 -7.81238079e-02 -6.00381568e-02
9.70855355e-01 -2.11796701e-01 -4.80826944e-02 3.87544632e-01
5.58062457e-02 -7.06408560e-01 7.37660646e-01 4.93912190e-01
8.69292557e-01 -6.23944938e-01 8.74700904e-01 2.18889505e-01
-1.73078120e+00 4.14824843e-01 -5.34428239e-01 5.34918904e-01
1.42497823e-01 6.51891232e-01 -1.22565351e-01 9.53041553e-01
7.68959165e-01 8.07574153e-01 -5.18437028e-01 1.31134403e+00
-2.07720175e-01 1.64961666e-01 -3.67854804e-01 1.11298814e-01
3.63773942e-01 -4.55263942e-01 8.17660451e-01 9.47507441e-01
4.42041397e-01 6.58442751e-02 3.64513814e-01 1.11383152e+00
-4.69075322e-01 -1.23528112e-02 -1.03372268e-01 4.56296504e-01
8.66382539e-01 1.54402280e+00 -6.63355649e-01 -1.68728039e-01
-3.90798002e-01 1.12372494e+00 4.64785457e-01 4.08994824e-01
-7.39315927e-01 -4.86596078e-01 2.66863704e-01 -1.15963139e-01
1.86062396e-01 2.02584222e-01 -2.29015246e-01 -1.06215990e+00
5.95858097e-01 -6.03183031e-01 3.96203250e-01 -5.72654545e-01
-1.81624615e+00 5.98007143e-01 -1.90095633e-01 -1.50239682e+00
2.85061479e-01 -6.98782951e-02 -1.03841209e+00 1.17831039e+00
-1.67428386e+00 -1.15181959e+00 -9.10616159e-01 3.77336353e-01
6.45136684e-02 1.45312816e-01 2.97748893e-01 6.51918054e-01
-5.43996453e-01 1.02731478e+00 2.08947778e-01 1.54250443e-01
8.08678389e-01 -1.17697680e+00 5.99115610e-01 8.07877362e-01
-7.33729750e-02 4.05725420e-01 4.31107670e-01 -7.57054448e-01
-1.14329183e+00 -1.40857029e+00 9.46393549e-01 7.74759576e-02
5.98244965e-01 -2.82509744e-01 -1.15558791e+00 2.80474305e-01
-1.83436185e-01 6.13883615e-01 4.91162613e-02 -1.89351261e-01
-4.40669358e-01 -3.67572397e-01 -1.29889727e+00 6.19848132e-01
1.18035865e+00 -5.97920775e-01 -1.27336532e-01 3.24439168e-01
6.88263416e-01 -6.89913571e-01 -8.06225836e-01 6.43123269e-01
3.99986804e-01 -8.19241643e-01 9.56359684e-01 -2.38576353e-01
2.76046455e-01 -4.80524600e-01 -1.26118178e-03 -1.32936263e+00
-5.60988069e-01 -6.78497851e-01 -5.13856299e-02 1.61750114e+00
1.11865856e-01 -6.83307230e-01 9.79605556e-01 4.35624540e-01
-1.11137152e-01 -1.04370713e+00 -1.17317367e+00 -1.02227104e+00
-1.43190399e-01 -5.23386337e-02 7.22185671e-01 1.05478787e+00
-3.57367575e-01 2.81004030e-02 -5.60346246e-01 3.47604990e-01
8.47346604e-01 1.03938699e-01 6.79493725e-01 -1.03832757e+00
-4.07691926e-01 -6.21152222e-01 -5.33457816e-01 -1.51525104e+00
-1.63832530e-02 -8.24685812e-01 3.15566272e-01 -1.36214352e+00
3.70336771e-01 -8.76946151e-01 -3.30699503e-01 4.75877374e-02
-7.47536004e-01 3.31778884e-01 1.56649902e-01 4.36237931e-01
-6.93158984e-01 7.07725346e-01 1.44032896e+00 -3.01011622e-01
8.69710892e-02 2.29728892e-02 -6.50714576e-01 4.77507353e-01
3.44458878e-01 -6.38033807e-01 -1.11890526e-03 -2.28370160e-01
1.08266219e-01 1.57280505e-01 5.43738723e-01 -7.19631672e-01
4.79899764e-01 5.48587218e-02 1.59989953e-01 -8.27539563e-01
2.94818282e-01 -8.77221584e-01 1.42143130e-01 4.18734550e-01
-1.08221158e-01 1.49051249e-01 -6.12175986e-02 1.06388032e+00
-4.68413591e-01 -1.79589778e-01 6.81348264e-01 -2.79979315e-04
-7.49445677e-01 6.04840100e-01 4.09220666e-01 3.30177605e-01
1.04726410e+00 -2.59893954e-01 -4.94127721e-01 -5.26708484e-01
-2.69707650e-01 6.97638810e-01 4.52456862e-01 4.32697594e-01
5.63587546e-01 -1.81975472e+00 -9.01964307e-01 9.76726115e-02
5.05874336e-01 1.24597244e-01 3.76150012e-01 1.14300418e+00
-2.08159640e-01 7.40935579e-02 2.62014896e-01 -6.33285105e-01
-1.20716846e+00 4.80991453e-01 3.53227645e-01 -4.04306144e-01
-5.46830654e-01 1.12624037e+00 1.90907300e-01 -5.50862312e-01
5.14376044e-01 -3.69405448e-02 1.72308281e-01 -2.22061232e-01
1.56078696e-01 5.61462164e-01 1.03595935e-01 -8.44751298e-01
-5.35227776e-01 7.87020326e-01 -1.61584705e-01 4.49020088e-01
8.21386814e-01 -1.88175067e-01 -1.99074388e-01 -8.22725073e-02
1.57328033e+00 -1.40768930e-01 -1.33109629e+00 -6.89064920e-01
-7.54328147e-02 -7.82373667e-01 1.21374324e-01 -6.03724360e-01
-1.44707847e+00 6.36578619e-01 5.99832714e-01 -2.02808063e-02
9.80779648e-01 -1.04677618e-01 1.07372177e+00 8.32042843e-02
3.22224319e-01 -9.39078093e-01 4.14893627e-02 7.56430924e-02
7.51455486e-01 -1.40184474e+00 1.10938631e-01 -7.79143929e-01
-4.22249913e-01 9.38234031e-01 7.06512988e-01 -2.09858388e-01
7.07606733e-01 7.85634145e-02 9.28970799e-02 -1.35586023e-01
-3.73015434e-01 4.31370400e-02 6.19226575e-01 4.66449291e-01
1.05018303e-01 6.09388202e-02 -4.02422607e-01 3.62407386e-01
1.57834426e-01 -3.57912511e-01 5.42068109e-02 5.91236234e-01
1.29151735e-02 -9.88430977e-01 -3.41871053e-01 4.48123425e-01
-2.06779018e-01 -2.16057777e-01 -3.85349631e-01 7.51102448e-01
2.37337843e-01 9.82146382e-01 4.24253456e-02 -5.31338871e-01
3.64149570e-01 -6.78445220e-01 1.91128418e-01 -1.55240297e-01
-5.25102615e-01 1.19185828e-01 -2.35241979e-01 -6.11279190e-01
-2.76054084e-01 -5.16824126e-01 -9.76946294e-01 -3.84815931e-01
-8.25372577e-01 1.45328730e-01 3.99678886e-01 7.19443798e-01
4.50230747e-01 3.82344812e-01 8.07483554e-01 -7.67965734e-01
-7.34504282e-01 -9.05997634e-01 -5.51125169e-01 7.25655496e-01
1.48934960e-01 -7.33746588e-01 -6.46041453e-01 -4.85334963e-01] | [7.238724708557129, 6.429582595825195] |
63b297fb-d546-4ec4-9a08-5141f0ea5093 | candidate-scoring-using-web-based-measure-for | null | null | https://aclanthology.org/W13-4420 | https://aclanthology.org/W13-4420.pdf | Candidate Scoring Using Web-Based Measure for Chinese Spelling Error Correction | null | ['Chung-Hsien Wu', 'Chao-Hong Liu', 'Liang-Chih Yu'] | 2013-10-01 | null | null | null | ws-2013-10 | ['csc'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.430234432220459, 3.865859031677246] |
853bad7a-072b-4e8b-9dfd-30b02b365b6d | certified-reasoning-with-language-models | 2306.04031 | null | https://arxiv.org/abs/2306.04031v1 | https://arxiv.org/pdf/2306.04031v1.pdf | Certified Reasoning with Language Models | Language models often achieve higher accuracy when reasoning step-by-step in complex tasks. However, their reasoning can be unsound, inconsistent, or rely on undesirable prior assumptions. To tackle these issues, we introduce a class of tools for language models called guides that use state and incremental constraints to guide generation. A guide can be invoked by the model to constrain its own generation to a set of valid statements given by the tool. In turn, the model's choices can change the guide's state. We show how a general system for logical reasoning can be used as a guide, which we call LogicGuide. Given a reasoning problem in natural language, a model can formalize its assumptions for LogicGuide and then guarantee that its reasoning steps are sound. In experiments with the PrOntoQA and ProofWriter reasoning datasets, LogicGuide significantly improves the performance of GPT-3, GPT-3.5 Turbo and LLaMA (accuracy gains up to 35%). LogicGuide also drastically reduces content effects: the interference of prior and current assumptions that both humans and language models have been shown to suffer from. Finally, we explore bootstrapping LLaMA 13B from its own reasoning and find that LogicGuide is critical: by training only on certified self-generated reasoning, LLaMA can self-improve, avoiding learning from its own hallucinations. | ['Noah D. Goodman', 'Eric Zelikman', 'Kanishk Gandhi', 'Gabriel Poesia'] | 2023-06-06 | null | null | null | null | ['logical-reasoning'] | ['reasoning'] | [ 1.83905568e-02 9.08464968e-01 -2.23229527e-01 -3.90310436e-01
-7.68601596e-01 -7.65722930e-01 7.71288455e-01 -1.96323812e-01
1.57787725e-01 8.14968586e-01 1.35490939e-01 -1.01088715e+00
8.75925422e-02 -1.04490590e+00 -7.44543374e-01 3.13696824e-02
1.51860848e-01 8.88179421e-01 3.17641348e-01 -4.24044311e-01
1.89682722e-01 2.94173837e-01 -1.26467299e+00 7.36990690e-01
1.14066780e+00 4.10336524e-01 -2.00442493e-01 9.17349398e-01
-4.70789969e-01 1.83456409e+00 -5.75500667e-01 -5.95486045e-01
2.88950633e-02 -5.10449827e-01 -1.26359046e+00 -2.75975406e-01
1.88479111e-01 -6.14004731e-01 7.94761404e-02 1.07375586e+00
-5.73735200e-02 -4.51740362e-02 4.65338409e-01 -1.68383372e+00
-6.10045731e-01 1.45405841e+00 1.17087409e-01 -3.73103678e-01
7.15060234e-01 7.67147958e-01 8.52972865e-01 -3.42873335e-01
6.94286227e-01 1.77732217e+00 7.25413561e-01 1.08470070e+00
-1.50444376e+00 -7.10478961e-01 2.85758048e-01 9.95180458e-02
-1.15206790e+00 -5.67145228e-01 5.35113394e-01 -3.77182543e-01
1.50047219e+00 4.52342600e-01 4.37648237e-01 1.09749246e+00
4.03935224e-01 8.13034713e-01 1.38450825e+00 -5.99378049e-01
3.80870193e-01 3.80005181e-01 2.94346660e-01 1.06395853e+00
1.75801173e-01 8.77653882e-02 -5.72315335e-01 -5.14903069e-01
3.86807710e-01 -7.05142558e-01 5.21806628e-02 -1.92304865e-01
-9.12949026e-01 7.40214825e-01 -4.45564240e-02 1.12688635e-02
-1.10956781e-01 3.72853398e-01 1.99249536e-01 4.87410396e-01
-2.46174604e-01 9.52047527e-01 -3.35332185e-01 -2.34033763e-01
-7.45765924e-01 8.11419249e-01 1.38443661e+00 9.94787872e-01
3.96251589e-01 5.47625124e-02 -2.62569249e-01 3.83358240e-01
6.74271286e-01 8.09900403e-01 2.75266618e-01 -1.61064351e+00
2.61787295e-01 6.22074902e-01 4.20943528e-01 -8.04985404e-01
-4.07265961e-01 -1.16784461e-01 -3.04336309e-01 4.66815442e-01
4.51058745e-01 -1.88818544e-01 -6.96126401e-01 1.92622471e+00
9.58477110e-02 -3.66753846e-01 5.93099236e-01 6.43148959e-01
4.92686987e-01 7.30100811e-01 1.35820016e-01 -1.47871360e-01
1.03064227e+00 -8.65437984e-01 -6.38667583e-01 -6.52570665e-01
8.45771432e-01 -3.47085148e-01 1.27293062e+00 9.39906359e-01
-1.60995305e+00 -2.37751037e-01 -1.09764850e+00 -1.41387001e-01
-3.95341553e-02 -3.38523597e-01 9.86270249e-01 4.30160850e-01
-1.21701002e+00 3.62473905e-01 -9.12295401e-01 -1.70437783e-01
1.12428419e-01 2.13209808e-01 4.06571105e-02 -1.99773028e-01
-1.34621620e+00 1.25362122e+00 5.20355940e-01 -2.25759000e-01
-1.14047372e+00 -4.70800877e-01 -8.56431901e-01 2.49083843e-02
7.12717295e-01 -8.56804371e-01 1.98991823e+00 -7.84711003e-01
-1.70594728e+00 7.23507702e-01 -1.67755097e-01 -9.07083750e-01
8.70236933e-01 -1.67387813e-01 -4.78496641e-01 1.97888240e-02
1.96779951e-01 6.21960580e-01 4.91337180e-01 -1.12338758e+00
-3.65505815e-01 5.09405732e-02 6.76541567e-01 -7.87193626e-02
2.86459476e-01 1.15098715e-04 -1.81194603e-01 4.71171103e-02
4.24292088e-02 -9.65309858e-01 -9.57775414e-02 6.30403757e-02
-6.47535503e-01 -3.61894518e-01 3.01004410e-01 -4.79045123e-01
1.15191150e+00 -1.78508222e+00 9.32213478e-03 2.93736339e-01
1.01343870e-01 2.11305708e-01 -8.92295316e-02 4.82076734e-01
1.41038239e-01 4.95476365e-01 -1.32441685e-01 2.16252953e-01
5.19303918e-01 4.36803401e-01 -7.78160989e-01 -1.41173705e-01
1.99539781e-01 9.18934584e-01 -9.10913348e-01 -6.19207799e-01
6.53317943e-02 -3.48669151e-03 -1.05832088e+00 2.93576121e-01
-1.04601836e+00 -9.11883041e-02 -3.51698875e-01 4.42018896e-01
4.75113690e-01 -3.60771239e-01 6.23114049e-01 3.27816188e-01
1.94319576e-01 7.62099564e-01 -1.06379294e+00 1.37521446e+00
-7.53000617e-01 2.39908949e-01 -3.22607420e-02 -4.27945614e-01
7.15677083e-01 2.84253985e-01 -4.73553896e-01 -6.10222340e-01
-1.30158544e-01 2.29383796e-01 3.04510385e-01 -6.02070212e-01
1.41681403e-01 -5.25619328e-01 -1.62307486e-01 9.53148067e-01
-2.95705169e-01 -8.19193006e-01 2.87695974e-01 5.82320631e-01
1.22526193e+00 2.60767192e-01 3.67882639e-01 -1.10691182e-01
5.09000480e-01 3.98086786e-01 4.80036944e-01 1.37440479e+00
1.53137371e-01 2.74484046e-02 8.00565779e-01 -4.25041258e-01
-8.21436584e-01 -1.15471709e+00 2.45123446e-01 9.63953257e-01
-2.09747121e-01 -7.79030859e-01 -6.74928427e-01 -8.18787396e-01
-3.27354968e-02 1.82346356e+00 -1.92293897e-01 -4.61312443e-01
-2.99903452e-01 -1.18173204e-01 1.12428641e+00 5.72139025e-01
6.00871742e-01 -1.25376964e+00 -8.47524166e-01 2.70789236e-01
-4.36707586e-01 -9.38024282e-01 7.52257481e-02 8.35835263e-02
-7.57249594e-01 -1.11917162e+00 3.79538417e-01 -1.44411534e-01
6.96702361e-01 -2.54962027e-01 1.25390720e+00 6.25771523e-01
3.93335760e-01 3.60605896e-01 -1.02075443e-01 -3.70467037e-01
-1.36574924e+00 -1.11118272e-01 -6.07570522e-02 -8.79281163e-01
3.43491554e-01 -3.91200989e-01 2.69106776e-01 1.49406731e-01
-7.72441566e-01 7.21342742e-01 3.02176595e-01 6.90925419e-01
1.77793019e-02 3.20093840e-01 2.50200957e-01 -1.33722520e+00
9.66819286e-01 -2.23142743e-01 -7.69537628e-01 4.87670839e-01
-9.58900511e-01 6.54905021e-01 8.60360801e-01 -2.24949360e-01
-1.32591915e+00 -3.46093088e-01 8.62568244e-03 -2.86441036e-02
1.49344094e-02 7.26013958e-01 -7.74775818e-02 3.64767075e-01
1.04847026e+00 2.33038291e-01 5.24647199e-02 -1.15529262e-02
4.56790417e-01 4.67577279e-01 6.93197906e-01 -1.48085010e+00
9.12328780e-01 1.37071982e-01 -2.47799352e-01 -1.79864436e-01
-1.05183399e+00 4.87170547e-01 8.11220631e-02 5.91407940e-02
3.97718370e-01 -7.16794074e-01 -9.53703284e-01 3.11881036e-01
-1.34060049e+00 -1.06755555e+00 -1.42599016e-01 -3.58091807e-03
-7.73400426e-01 1.45780325e-01 -6.81720078e-01 -1.26180565e+00
-2.95922965e-01 -1.08172584e+00 5.14975369e-01 1.10225365e-01
-1.00831151e+00 -8.38116348e-01 -1.59891441e-01 4.49690908e-01
4.33413506e-01 -1.52891472e-01 1.48222363e+00 -8.76041651e-01
-5.91690779e-01 7.67213479e-02 8.47224239e-03 4.20249015e-01
-2.53097355e-01 2.35849172e-01 -8.71573985e-01 1.72413036e-01
-9.03545544e-02 -7.85882831e-01 2.40182862e-01 -2.76762515e-01
8.89980853e-01 -9.35776949e-01 -1.53525174e-01 2.16073111e-01
8.42117667e-01 3.12971085e-01 6.30074918e-01 3.59782040e-01
8.52207243e-02 4.03574407e-01 5.18536806e-01 1.25907108e-01
5.50958812e-01 3.46958399e-01 3.61459106e-02 4.76688743e-01
1.78273022e-02 -7.10490704e-01 6.54793620e-01 1.53313309e-01
1.15471654e-01 -1.09996922e-01 -1.38426006e+00 2.58138567e-01
-1.94821239e+00 -1.17954421e+00 7.62702376e-02 1.84103942e+00
1.47141004e+00 7.06945181e-01 -3.09253991e-01 7.97978267e-02
1.15782961e-01 -2.00403139e-01 -6.62022352e-01 -1.08917856e+00
1.17944501e-01 1.42241195e-01 -1.44359484e-01 1.14064264e+00
-2.87981987e-01 1.29736352e+00 7.27812910e+00 3.97076994e-01
-9.04960334e-01 -1.35056406e-01 1.86448440e-01 -7.56267831e-03
-8.56049657e-01 3.58914614e-01 -7.01601565e-01 1.06071375e-01
1.11214519e+00 -5.77227473e-01 8.83337200e-01 9.23423707e-01
6.34465665e-02 -3.23814720e-01 -1.61176944e+00 3.76558989e-01
1.20470174e-01 -1.32923985e+00 2.48834565e-01 -3.64690661e-01
4.91876185e-01 -3.83000106e-01 -1.07822731e-01 8.66257906e-01
1.22387278e+00 -1.25915611e+00 1.22762024e+00 5.20028114e-01
4.22811002e-01 -7.85467267e-01 5.56735933e-01 8.87053430e-01
-2.30442241e-01 -1.78990349e-01 -1.85209680e-02 -6.01965427e-01
-1.01709887e-01 3.15057546e-01 -1.42438340e+00 1.04119278e-01
2.35736936e-01 5.23731783e-02 -5.20388365e-01 3.34222078e-01
-9.81138825e-01 8.02379906e-01 -3.92627209e-01 -2.36485839e-01
9.15457532e-02 1.81892335e-01 3.95677626e-01 1.09274220e+00
5.38847223e-02 3.15369040e-01 2.91513562e-01 1.52948666e+00
2.42311165e-01 -5.39817631e-01 -5.55348694e-01 -2.62325376e-01
5.64853370e-01 7.69247174e-01 -2.96896905e-01 -7.76170671e-01
-3.18435542e-02 5.91082036e-01 3.43789637e-01 3.98828745e-01
-9.40981209e-01 -4.96468917e-02 4.50847954e-01 1.64679691e-01
-1.99430257e-01 -1.22831717e-01 -1.56995162e-01 -1.10030603e+00
-1.33914858e-01 -1.56710744e+00 3.07335347e-01 -1.63890278e+00
-9.92060184e-01 4.92897630e-01 3.82414848e-01 -4.51355428e-01
-8.24077487e-01 -5.06453693e-01 -4.56022829e-01 8.52611005e-01
-1.17949939e+00 -9.68226373e-01 3.22904475e-02 3.80274296e-01
3.47281903e-01 -1.51896235e-02 1.05171120e+00 -5.04675210e-01
-1.05913460e-01 3.55305135e-01 -9.22237396e-01 1.16612196e-01
4.69553679e-01 -1.28140414e+00 4.29885328e-01 9.28425968e-01
-9.13416818e-02 1.21802807e+00 1.13147092e+00 -7.86543429e-01
-1.51071787e+00 -8.13088894e-01 1.06320333e+00 -7.58242071e-01
9.23234463e-01 -1.90722629e-01 -9.44905996e-01 1.35790265e+00
1.31150946e-01 -4.03901368e-01 3.41890454e-01 3.07388157e-01
-8.58036160e-01 1.94274247e-01 -1.22528994e+00 1.09551585e+00
1.14112186e+00 -8.37212622e-01 -1.27929771e+00 4.93186772e-01
6.43366992e-01 -7.86119699e-01 -3.58167082e-01 1.16175180e-02
6.12414598e-01 -1.19640744e+00 6.50176346e-01 -1.08213663e+00
6.09753847e-01 -5.15959084e-01 -1.47243440e-01 -1.20448923e+00
-2.90242106e-01 -8.98156226e-01 -3.91240329e-01 1.02704394e+00
6.90755963e-01 -7.71850705e-01 2.95806140e-01 1.31584513e+00
7.27977697e-03 -4.50790614e-01 -4.12459522e-01 -6.43860340e-01
3.19946796e-01 -9.74460721e-01 7.80846179e-01 6.25162542e-01
6.36186957e-01 4.67768818e-01 -9.96091142e-02 2.91080296e-01
3.35681856e-01 2.79990107e-01 9.48663294e-01 -9.76715744e-01
-7.44263053e-01 -3.02134663e-01 3.16852331e-01 -7.69403934e-01
5.92667639e-01 -1.17922211e+00 4.17144120e-01 -1.68879211e+00
1.35633722e-01 -4.11313951e-01 2.56154329e-01 1.18361664e+00
1.00676529e-01 -3.59769315e-01 3.52417618e-01 -5.59337772e-02
-6.66644633e-01 -7.51690045e-02 1.14858961e+00 -3.98571253e-01
-1.53186023e-01 -1.96416512e-01 -1.13667607e+00 1.03373110e+00
7.12801158e-01 -3.06749791e-01 -7.91961133e-01 -3.75030816e-01
1.00107360e+00 3.40338409e-01 5.52152336e-01 -1.08877051e+00
3.81551027e-01 -6.76875353e-01 7.38169327e-02 -1.04244053e-01
8.09885655e-03 -5.38904846e-01 3.85439515e-01 6.53827786e-01
-8.98042500e-01 -2.32968237e-02 4.58122402e-01 -1.23492643e-01
2.32406229e-01 -2.75225401e-01 5.73495150e-01 -3.41008306e-01
-5.23686051e-01 -5.30064940e-01 -8.26413035e-01 3.56436700e-01
6.58479095e-01 2.49602750e-01 -5.89126587e-01 -5.73069930e-01
-6.86634064e-01 5.65906286e-01 5.48483908e-01 2.05969781e-01
5.22264659e-01 -9.64365482e-01 -5.41948736e-01 2.75646716e-01
1.02486826e-01 1.43924460e-01 -8.32171813e-02 7.09078133e-01
-5.73148429e-01 4.14250106e-01 -8.13776627e-02 -3.64313900e-01
-1.02535748e+00 5.38778067e-01 6.70143485e-01 -4.87338662e-01
-4.64919925e-01 7.30246127e-01 -6.41920716e-02 -7.64732599e-01
-1.79770011e-02 -8.30430865e-01 3.14112067e-01 -5.14034152e-01
7.14604318e-01 -1.88047186e-01 -1.10630773e-01 1.41868010e-01
-5.17852068e-01 2.44058836e-02 -1.73946679e-01 -5.36484540e-01
9.51906800e-01 2.30012655e-01 -3.03734213e-01 5.73716521e-01
1.70204476e-01 2.40723863e-01 -9.34757829e-01 -9.51433256e-02
-8.41786191e-02 -7.55580664e-02 -2.31721893e-01 -1.64803636e+00
-3.44920665e-01 6.38823807e-01 -3.03617537e-01 2.44199842e-01
6.71876192e-01 -6.12394884e-02 4.66082841e-01 1.09622979e+00
7.81118572e-01 -9.04632986e-01 -1.39483795e-01 6.59076154e-01
1.15357912e+00 -8.98712873e-01 3.09555815e-03 -1.43473342e-01
-8.04744840e-01 1.02331519e+00 8.69398654e-01 1.13011658e-01
4.77936678e-02 7.25088120e-01 2.60551274e-01 -1.23113701e-02
-1.48474431e+00 3.62560034e-01 -1.62987009e-01 6.76156044e-01
2.55157232e-01 2.21520320e-01 4.26455354e-03 9.40360963e-01
-7.41170287e-01 5.79372287e-01 7.93312788e-01 1.05880833e+00
-4.19363320e-01 -1.06804299e+00 -7.00504839e-01 1.47474796e-01
-9.22414511e-02 -1.90495551e-01 -6.74951196e-01 9.16939139e-01
3.40797901e-02 1.19318950e+00 -2.92726576e-01 -1.71988234e-01
7.85413682e-02 5.41994631e-01 7.84843743e-01 -8.87776732e-01
-4.86010820e-01 -3.93184811e-01 7.76914537e-01 -8.07924867e-01
-1.92507505e-02 -5.31697214e-01 -1.95835865e+00 -6.41341984e-01
1.37005150e-01 4.25671786e-01 1.11983344e-01 1.16737437e+00
1.05360761e-01 4.47611719e-01 -4.21941988e-02 -4.66091633e-02
-1.01193964e+00 -5.73917389e-01 -1.06227636e-01 1.64206401e-01
1.51356697e-01 -2.82063216e-01 -4.56988722e-01 1.32610425e-01] | [9.245381355285645, 7.2210845947265625] |
988b48c4-ac62-41c8-acdf-61c3c517d466 | a-conditional-random-field-model-for-context | 1906.07383 | null | https://arxiv.org/abs/1906.07383v1 | https://arxiv.org/pdf/1906.07383v1.pdf | A Conditional Random Field Model for Context Aware Cloud Detection in Sky Images | A conditional random field (CRF) model for cloud detection in ground based sky images is presented. We show that very high cloud detection accuracy can be achieved by combining a discriminative classifier and a higher order clique potential in a CRF framework. The image is first divided into homogeneous regions using a mean shift clustering algorithm and then a CRF model is defined over these regions. The various parameters involved are estimated using training data and the inference is performed using Iterated Conditional Modes (ICM) algorithm. We demonstrate how taking spatial context into account can boost the accuracy. We present qualitative and quantitative results to prove the superior performance of this framework in comparison with other state of the art methods applied for cloud detection. | ['Jeffrey J. Rodriguez', 'Alexander D. Cronin', 'Vijai T. Jayadevan'] | 2019-06-18 | null | null | null | null | ['cloud-detection'] | ['computer-vision'] | [ 2.98987478e-01 -3.52115899e-01 6.38230816e-02 -5.45862138e-01
-8.51982117e-01 -6.43464565e-01 7.90771246e-01 1.35908708e-01
-2.85602480e-01 6.82161152e-01 -4.93574888e-01 -1.88852519e-01
-1.41427055e-01 -7.84921050e-01 -4.43625450e-01 -1.03569388e+00
-1.89451367e-01 8.94050717e-01 6.76304996e-01 3.23912919e-01
4.00769323e-01 7.58558333e-01 -1.56819153e+00 1.14033446e-01
1.24465561e+00 6.36788309e-01 7.52950013e-01 8.21697414e-01
-1.24366067e-01 8.14577579e-01 -4.42085624e-01 2.86248118e-01
2.20778376e-01 -1.97730854e-01 -9.32096481e-01 4.34325010e-01
5.09341061e-01 1.50716066e-01 3.21267128e-01 1.21528244e+00
-2.14677267e-02 1.78019226e-01 1.01190603e+00 -9.11568701e-01
6.02573045e-02 -1.02856643e-02 -5.11291921e-01 1.32057190e-01
2.36673445e-01 -1.42386347e-01 8.99528444e-01 -1.12856519e+00
7.29109824e-01 8.46010149e-01 3.85807306e-01 -8.00537616e-02
-1.30162704e+00 -5.11415958e-01 1.17807724e-01 9.81807411e-02
-1.72295856e+00 1.86685205e-01 4.45998549e-01 -7.55852878e-01
9.50951815e-01 2.72890717e-01 8.09819639e-01 4.88136895e-02
1.46509288e-02 6.07775033e-01 1.63618851e+00 -1.02259827e+00
5.38484335e-01 2.89388686e-01 1.16760962e-01 4.08484071e-01
2.60512203e-01 8.06419924e-02 -3.30244362e-01 -2.52408832e-01
6.36009216e-01 -2.95170490e-02 6.94909543e-02 -3.24570030e-01
-7.83836544e-01 1.13302290e+00 7.10955858e-01 3.13686430e-01
-3.84011000e-01 8.75747129e-02 -2.07422420e-01 -3.44569892e-01
5.65142870e-01 1.50825769e-01 -1.23438574e-01 5.08044302e-01
-1.56142795e+00 4.46498811e-01 6.30205333e-01 1.02613175e+00
1.08212519e+00 -2.61784941e-01 -2.09622495e-02 3.54898036e-01
6.78466976e-01 9.92649674e-01 -3.62695485e-01 -9.20107603e-01
-5.10402583e-02 3.87234896e-01 2.90582478e-01 -7.48666227e-01
-1.52619334e-03 -4.46922004e-01 -5.93126595e-01 5.54843068e-01
1.52726457e-01 1.89402953e-01 -1.18416345e+00 8.09744239e-01
5.13057649e-01 5.03753841e-01 -1.44717231e-01 8.26981187e-01
5.11483967e-01 4.12207931e-01 1.44156277e-01 -3.90964568e-01
1.20747375e+00 -7.82090783e-01 -5.93548417e-01 -5.80363497e-02
2.28901669e-01 -9.69670177e-01 3.63967717e-01 5.63766837e-01
-5.35762548e-01 -4.47716981e-01 -7.69079924e-01 4.16653514e-01
-5.42804778e-01 2.28463128e-01 9.38003778e-01 7.79203534e-01
-9.13103640e-01 4.13420916e-01 -9.17881548e-01 -2.14796379e-01
2.65643507e-01 4.42709446e-01 5.57872579e-02 -2.15730265e-01
-7.13117838e-01 8.38714242e-01 2.59082973e-01 1.61579981e-01
-7.94309616e-01 -1.19710118e-01 -5.07740080e-01 -1.55256286e-01
-5.74999303e-02 -6.94731593e-01 1.00230753e+00 -8.46491694e-01
-1.12325609e+00 1.12648237e+00 -6.34536207e-01 -4.89821881e-01
3.43854189e-01 -9.61837545e-02 -2.51373887e-01 6.97473526e-01
1.14805318e-01 5.50100923e-01 9.27876353e-01 -1.75639129e+00
-7.31106162e-01 -2.14696065e-01 -1.90673515e-01 1.00741953e-01
5.97398639e-01 4.53625858e-01 -4.79521990e-01 -1.74499348e-01
5.95858395e-01 -1.24701166e+00 -6.57864749e-01 -5.48987210e-01
-4.54288781e-01 -5.55003434e-02 6.95168197e-01 -4.66362655e-01
9.68154252e-01 -1.58572924e+00 -6.31008968e-02 8.21806371e-01
2.80982077e-01 -1.47797298e-02 4.78149533e-01 5.24858832e-01
8.95116329e-02 9.87562984e-02 -6.69920564e-01 -3.33925188e-01
-3.20800990e-01 3.67284745e-01 -1.98199973e-01 7.54162729e-01
2.14551881e-01 5.39425790e-01 -8.32228303e-01 -8.72491121e-01
6.34245813e-01 5.95527589e-01 -3.41678023e-01 1.99415684e-01
-5.08039653e-01 7.83590436e-01 -4.77885485e-01 7.21046329e-01
1.49854219e+00 -3.37411880e-01 3.07652742e-01 4.15023595e-01
-4.53562468e-01 1.70809433e-01 -1.24112773e+00 1.26529241e+00
-1.81666002e-01 6.10624194e-01 1.14242792e-01 -6.72945321e-01
9.53010559e-01 1.52432755e-01 3.03708941e-01 -2.62246072e-01
-2.19598338e-01 1.99550465e-01 -3.58842701e-01 -3.90083045e-01
5.76360703e-01 -4.30309415e-01 3.95893931e-01 -1.69978052e-01
-7.26322234e-02 -7.60294616e-01 -1.83143392e-01 4.91617233e-01
6.34474933e-01 2.51662403e-01 2.02306390e-01 -6.13685012e-01
4.70073879e-01 5.94171882e-01 2.98225939e-01 8.72487009e-01
-6.51088506e-02 7.56991267e-01 2.09874045e-02 -1.59676358e-01
-6.23617411e-01 -9.61658001e-01 -5.73138177e-01 5.93840003e-01
1.93145856e-01 -3.05619270e-01 -8.09064090e-01 -4.43721563e-01
-3.08973752e-02 5.01620889e-01 -5.04340351e-01 6.43761754e-01
-3.19855005e-01 -1.16236305e+00 -4.15999256e-02 3.12664956e-01
3.86692315e-01 -8.37412715e-01 -5.11784315e-01 -6.58096001e-02
-2.01079234e-01 -1.37846422e+00 2.93865472e-01 4.81361210e-01
-1.08940208e+00 -8.90017211e-01 -1.79208592e-01 -6.87388420e-01
8.18848848e-01 5.86216271e-01 1.50001681e+00 4.01559412e-01
-4.67477262e-01 3.93108696e-01 -4.97863889e-01 -5.19290447e-01
-2.29211956e-01 -2.36889407e-01 -1.78805813e-01 -8.46509635e-02
6.01108015e-01 -4.47403520e-01 -5.75819492e-01 5.72965294e-02
-8.25701535e-01 -1.26122788e-01 5.11596799e-01 5.46210051e-01
9.99632239e-01 2.30723396e-01 -1.67607382e-01 -1.20751941e+00
-2.84363866e-01 -3.80985707e-01 -1.31815457e+00 1.05780341e-01
-6.75660431e-01 -2.82427013e-01 -6.71700463e-02 1.76865250e-01
-1.15702665e+00 9.06924546e-01 2.35784888e-01 -5.33824086e-01
-5.76304317e-01 4.27834511e-01 9.17919204e-02 -7.24036336e-01
5.89635670e-01 2.91693658e-01 -5.58220863e-01 -3.67419153e-01
3.10961664e-01 6.65048599e-01 4.61348146e-01 -4.57381606e-01
1.09656787e+00 9.63803947e-01 4.72992837e-01 -1.03814936e+00
-7.05493689e-01 -1.32375646e+00 -1.50011146e+00 -6.24167979e-01
1.18431139e+00 -1.15085971e+00 -4.40385222e-01 3.34880769e-01
-1.06921434e+00 -1.87614158e-01 1.00698143e-01 7.33227730e-01
-3.84397805e-01 4.73097712e-01 -2.22830519e-01 -1.68015516e+00
2.56442185e-02 -7.64610350e-01 1.32775986e+00 3.27596366e-01
6.18426740e-01 -1.08649933e+00 3.52084398e-01 5.62214732e-01
2.94986337e-01 3.94505203e-01 4.51074421e-01 -3.25495332e-01
-1.33542812e+00 -1.57716751e-01 -3.02794069e-01 1.61897436e-01
-2.51172423e-01 5.11946321e-01 -1.43943286e+00 -1.13994263e-01
-8.32527280e-02 2.42859740e-02 1.25760388e+00 7.58953571e-01
7.65363038e-01 1.70514137e-01 -6.34413242e-01 4.89528060e-01
2.21187854e+00 -2.54424542e-01 8.85918140e-01 1.62665084e-01
7.34243631e-01 4.19443846e-01 9.02946353e-01 2.57925957e-01
2.57182032e-01 5.64232230e-01 8.16370130e-01 -2.46172786e-01
6.93107694e-02 1.03620350e-01 -2.23668665e-01 6.82646394e-01
-5.14381766e-01 -1.35849295e-02 -1.09014881e+00 6.80236578e-01
-1.73969543e+00 -8.94152462e-01 -1.06636548e+00 2.42895317e+00
4.16271150e-01 5.92808090e-02 1.06712840e-02 -9.62786749e-03
7.75835812e-01 -2.05090135e-01 1.29795879e-01 -1.52423969e-02
-1.87632903e-01 4.58360553e-01 9.26492274e-01 8.62993896e-01
-1.47826922e+00 1.23815215e+00 7.57982969e+00 8.66694450e-01
-9.06051278e-01 2.58115411e-01 2.39091232e-01 4.46002871e-01
-1.18193708e-01 7.15799749e-01 -8.99037719e-01 2.83841908e-01
6.78080916e-01 5.06092131e-01 2.27418035e-01 8.07608426e-01
3.22064906e-01 -9.60325956e-01 -1.93287507e-01 6.59013212e-01
-1.76169276e-01 -1.23000503e+00 -3.32763016e-01 2.67436802e-01
1.01862299e+00 6.77187741e-01 -4.29642290e-01 -1.09651268e-01
5.98499298e-01 -1.10020423e+00 5.16038954e-01 9.39095080e-01
6.80297971e-01 -5.46249509e-01 5.86576045e-01 4.34220105e-01
-1.36827064e+00 2.80652344e-01 -6.15769386e-01 -1.98468328e-01
-9.53297690e-02 1.09333861e+00 -1.12232196e+00 8.30545843e-01
6.72969639e-01 2.82989055e-01 -6.22406185e-01 1.47834480e+00
-2.61850387e-01 9.84180629e-01 -6.20589256e-01 6.31064996e-02
1.59500793e-01 -8.00926387e-01 5.36470473e-01 1.71925306e+00
-1.23110078e-01 1.45529076e-01 4.27471042e-01 8.51965129e-01
4.32706743e-01 1.29759401e-01 -5.56994200e-01 4.48197454e-01
3.35924715e-01 1.52551448e+00 -1.23807991e+00 -5.07064998e-01
-3.59428972e-01 7.80359745e-01 1.21039003e-01 3.04935634e-01
-5.02472043e-01 1.49569571e-01 1.37894541e-01 2.19793394e-01
7.75824964e-01 -5.05271971e-01 -3.00840318e-01 -1.21870935e+00
-1.36135623e-01 -2.62928128e-01 2.34936833e-01 -1.06554425e+00
-1.22414899e+00 5.35824597e-01 2.50750959e-01 -1.40519083e+00
-2.06349880e-01 -6.49338543e-01 -7.86395252e-01 1.07147169e+00
-1.74701011e+00 -1.56850457e+00 -5.91540933e-01 4.42784399e-01
1.36486664e-01 3.30893368e-01 1.03241038e+00 -6.12183250e-02
-8.62926543e-02 -3.34385902e-01 3.93502831e-01 -7.48172849e-02
2.79496372e-01 -1.78385842e+00 -9.00656804e-02 1.15025818e+00
2.48408049e-01 4.85669464e-01 8.56268227e-01 -7.79159248e-01
-8.51000845e-01 -1.09916937e+00 1.13961124e+00 -6.93368971e-01
2.67635942e-01 -5.26362002e-01 -9.05746818e-01 3.95601839e-01
3.45482469e-01 4.56439883e-01 5.27541995e-01 1.60569668e-01
-3.06034297e-01 3.15603554e-01 -1.13190579e+00 -3.35437328e-01
5.14966965e-01 -6.12377644e-01 -2.41710275e-01 9.83542204e-01
3.36317688e-01 -2.88850397e-01 -6.94931090e-01 4.35499638e-01
1.87451154e-01 -1.25277936e+00 7.14302003e-01 -3.02379787e-01
-1.79692712e-02 -7.80622900e-01 -3.87530893e-01 -8.42073500e-01
-3.82971883e-01 -3.06712151e-01 3.40333968e-01 9.35423255e-01
2.80998558e-01 -3.32803041e-01 7.60666728e-01 2.81338841e-01
2.00499728e-01 -2.93366760e-01 -1.00416625e+00 -7.30537772e-01
2.43398994e-01 -4.90172446e-01 1.72498539e-01 1.07055652e+00
-2.91950613e-01 1.57961160e-01 -1.79201096e-01 9.12336290e-01
9.07610655e-01 8.23620200e-01 6.84092224e-01 -1.60525513e+00
-2.38211751e-01 1.03397287e-01 -4.19980705e-01 -7.92007208e-01
-8.41167271e-02 -7.44558871e-01 2.04800844e-01 -1.51463389e+00
5.15288174e-01 -7.62574434e-01 4.90673259e-02 -6.10889904e-02
-1.99169025e-01 2.74083495e-01 1.49809822e-01 6.37112260e-01
-8.81229281e-01 9.50947180e-02 8.66104007e-01 2.57300943e-01
1.06460698e-01 3.03274989e-01 1.88319996e-01 6.66749060e-01
7.04387963e-01 -8.50862384e-01 2.66446084e-01 -3.57678533e-02
2.80518740e-01 1.22470692e-01 6.87629998e-01 -1.13995337e+00
4.65797812e-01 -3.37954700e-01 5.32234192e-01 -1.02122271e+00
3.65270048e-01 -7.29986548e-01 -6.47431007e-03 2.35058159e-01
3.18718970e-01 -3.20404500e-01 -6.38244813e-03 8.69819283e-01
-4.79987830e-01 -5.60454011e-01 1.04535997e+00 -3.19693297e-01
-5.30763149e-01 5.56621887e-02 -5.63366413e-01 -5.40267766e-01
7.94574261e-01 -4.58304808e-02 1.91085264e-01 -4.04829651e-01
-9.56717432e-01 8.27246010e-02 6.41422391e-01 -4.30962324e-01
3.93738747e-01 -8.44983757e-01 -7.77780414e-01 2.31388770e-02
1.91511825e-01 2.71726727e-01 1.42802149e-01 1.00592899e+00
-7.88509965e-01 4.66020554e-01 2.77785689e-01 -1.21141362e+00
-1.44351232e+00 4.95542347e-01 5.67100644e-01 -4.61880893e-01
-4.27285463e-01 6.54944420e-01 -3.96366930e-03 -4.52002645e-01
-1.73607096e-01 -2.97570944e-01 -2.69751340e-01 -2.65200198e-01
3.99912000e-02 2.35246688e-01 2.55991161e-01 -8.79824162e-01
-6.84535325e-01 7.83532023e-01 3.55533272e-01 -2.69226521e-01
1.13993323e+00 -1.27831727e-01 -3.96761000e-01 1.23668060e-01
5.77960134e-01 5.49101889e-01 -1.09093511e+00 -2.31173828e-01
1.07988343e-03 -8.20254385e-01 3.71452808e-01 -7.55137026e-01
-8.80092442e-01 8.30738664e-01 8.06204379e-01 3.73033315e-01
1.18723130e+00 1.65826753e-01 -1.31028771e-01 1.86997145e-01
4.74623919e-01 -9.31992888e-01 -4.59365249e-01 3.77128780e-01
3.60246688e-01 -1.52897096e+00 3.91967088e-01 -1.21897495e+00
-6.80653632e-01 9.34931099e-01 7.10333064e-02 -6.54650569e-01
9.70597863e-01 2.98128724e-01 9.67651680e-02 -6.25590205e-01
-4.64518607e-01 -8.64062786e-01 2.43633181e-01 7.01625466e-01
1.91058114e-01 5.88655233e-01 -2.79789399e-02 -1.86927635e-02
8.21154490e-02 -8.99518877e-02 4.25110817e-01 8.82982552e-01
-7.38478065e-01 -1.11280179e+00 -8.84233475e-01 3.79667610e-01
-4.73517656e-01 -3.14808071e-01 -4.25862134e-01 7.16193736e-01
5.15316486e-01 1.22216332e+00 8.76480192e-02 -1.94437336e-02
-3.08847070e-01 1.03522450e-01 5.76114476e-01 -8.08665276e-01
-4.48448002e-01 5.15370667e-01 -2.51597404e-01 -2.13410094e-01
-1.16111100e+00 -9.22903240e-01 -1.28444898e+00 2.00601354e-01
-1.10727465e+00 3.00854057e-01 9.70149636e-01 9.65011597e-01
1.23177864e-01 1.69661596e-01 7.71025896e-01 -9.97198105e-01
8.23293254e-02 -1.01506734e+00 -9.26485360e-01 -1.09692156e-01
2.96004474e-01 -7.06739068e-01 -7.31189072e-01 3.14815849e-01] | [9.73434829711914, -1.7646949291229248] |
08765102-930d-426b-aa92-c645d5aeea4d | do-different-tracking-tasks-require-different | 2107.02156 | null | https://arxiv.org/abs/2107.02156v2 | https://arxiv.org/pdf/2107.02156v2.pdf | Do Different Tracking Tasks Require Different Appearance Models? | Tracking objects of interest in a video is one of the most popular and widely applicable problems in computer vision. However, with the years, a Cambrian explosion of use cases and benchmarks has fragmented the problem in a multitude of different experimental setups. As a consequence, the literature has fragmented too, and now novel approaches proposed by the community are usually specialised to fit only one specific setup. To understand to what extent this specialisation is necessary, in this work we present UniTrack, a solution to address five different tasks within the same framework. UniTrack consists of a single and task-agnostic appearance model, which can be learned in a supervised or self-supervised fashion, and multiple ``heads'' that address individual tasks and do not require training. We show how most tracking tasks can be solved within this framework, and that the same appearance model can be successfully used to obtain results that are competitive against specialised methods for most of the tasks considered. The framework also allows us to analyse appearance models obtained with the most recent self-supervised methods, thus extending their evaluation and comparison to a larger variety of important problems. | ['Luca Bertinetto', 'Philip H. S. Torr', 'Shengjin Wang', 'Ya-Li Li', 'Hengshuang Zhao', 'Zhongdao Wang'] | 2021-07-05 | null | http://proceedings.neurips.cc/paper/2021/hash/06997f04a7db92466a2baa6ebc8b872d-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/06997f04a7db92466a2baa6ebc8b872d-Paper.pdf | neurips-2021-12 | ['multiple-people-tracking', 'video-instance-segmentation', 'online-multi-object-tracking', 'video-object-tracking', 'multi-object-tracking-and-segmentation'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 2.43848115e-01 -2.00014755e-01 -1.69798825e-02 -1.75690055e-01
-2.78758913e-01 -6.15939856e-01 8.58446836e-01 -3.30066681e-03
-5.06623447e-01 5.00135362e-01 -4.69490409e-01 6.07622676e-02
-2.35158786e-01 -2.88906783e-01 -5.95072687e-01 -9.39093769e-01
2.00021043e-02 7.37082362e-01 9.04095113e-01 -1.85044214e-01
-7.35481083e-03 6.47233307e-01 -1.97420752e+00 9.95311514e-02
4.32381302e-01 9.18254137e-01 3.44248444e-01 4.90986198e-01
4.29034382e-02 3.67243409e-01 -4.92821574e-01 -6.20662630e-01
4.58942980e-01 -2.88512528e-01 -8.74068856e-01 4.92000550e-01
1.05844033e+00 2.61027753e-01 5.55614606e-02 1.02166057e+00
3.97156566e-01 -6.85852543e-02 6.07847452e-01 -1.36578619e+00
-1.07657634e-01 1.69100352e-02 -5.44511199e-01 1.79916173e-01
1.94486246e-01 -1.91239893e-01 8.26074481e-01 -4.32349503e-01
8.57737064e-01 1.06546092e+00 8.77886593e-01 5.19261956e-01
-1.56475747e+00 -2.71208972e-01 2.69092947e-01 2.75608510e-01
-1.03838980e+00 -4.17251259e-01 6.79408312e-01 -5.26479125e-01
5.09583354e-01 3.36690873e-01 6.24881685e-01 1.11259925e+00
2.96656266e-02 8.02906752e-01 1.45761228e+00 -6.27888441e-01
-1.06172003e-02 4.66842324e-01 1.06779121e-01 6.29129112e-01
4.87330973e-01 3.58252004e-02 -2.64755506e-02 -7.16114938e-02
7.46258616e-01 -6.71103820e-02 -1.89412639e-01 -1.15884256e+00
-1.28359818e+00 6.81510746e-01 2.45038286e-01 7.49999881e-01
2.39388198e-02 -2.40768865e-02 6.42907083e-01 4.62544501e-01
4.85104918e-01 3.12489241e-01 -4.81380582e-01 3.38523500e-02
-9.91947711e-01 3.09356123e-01 8.79541636e-01 7.17954695e-01
6.71135306e-01 -3.98397334e-02 8.37240368e-02 9.23914731e-01
2.99896210e-01 5.09358287e-01 4.15923506e-01 -8.03325057e-01
5.21942461e-03 3.57716650e-01 9.79348570e-02 -1.01152492e+00
-5.46647608e-01 -5.43622673e-01 -6.43546879e-01 6.38367236e-01
8.64713788e-01 1.21774666e-01 -7.61355817e-01 1.77621090e+00
5.42063236e-01 1.77003995e-01 -3.21309745e-01 7.78795660e-01
5.60993016e-01 1.96035758e-01 -1.14423349e-01 -2.29606137e-01
1.37236702e+00 -1.13883626e+00 -4.42782402e-01 -1.83291256e-01
5.46401739e-01 -1.10332501e+00 5.01474559e-01 7.19363928e-01
-8.47800195e-01 -6.71274424e-01 -9.13965940e-01 2.71081239e-01
-5.54721296e-01 2.28355408e-01 5.49946725e-01 7.93005228e-01
-1.41017294e+00 7.11299539e-01 -5.96862316e-01 -8.66201699e-01
1.29326150e-01 4.06629413e-01 -5.65191507e-01 1.28239527e-01
-8.46861124e-01 1.32801485e+00 3.46177727e-01 1.00932717e-02
-4.36808497e-01 -1.83340386e-01 -6.77087843e-01 -4.29737300e-01
6.25340700e-01 -7.62971401e-01 1.18481767e+00 -1.45984566e+00
-1.38525891e+00 1.44804931e+00 2.62078084e-02 -4.86917138e-01
9.00996625e-01 -2.66640331e-03 -3.04431856e-01 1.28547996e-01
-8.42193067e-02 5.98338068e-01 1.28769732e+00 -1.41545522e+00
-7.14462519e-01 -3.29537004e-01 8.77707377e-02 -3.65220606e-02
-2.54363060e-01 3.59469593e-01 -7.39000082e-01 -8.87758076e-01
-3.07799786e-01 -1.20344269e+00 -1.25610024e-01 1.42164052e-01
1.60297379e-02 -2.88077474e-01 9.25909579e-01 -3.73992264e-01
9.83455598e-01 -1.96966577e+00 6.75724983e-01 -8.18919688e-02
1.81139559e-01 6.60562873e-01 -9.45891887e-02 3.34822178e-01
-2.43143857e-01 -3.98067534e-01 -4.18442607e-01 -6.92503214e-01
1.05299443e-01 4.11269635e-01 6.67485595e-02 7.84499109e-01
-3.78610939e-02 7.34516740e-01 -8.89211416e-01 -7.18724370e-01
4.96847421e-01 4.40330416e-01 -3.71401161e-02 4.51604165e-02
-3.15991104e-01 4.60401654e-01 -2.89904833e-01 3.71445507e-01
6.11994267e-01 -1.48260042e-01 2.87773814e-02 5.70856407e-03
-1.39337614e-01 -3.61572683e-01 -1.38728738e+00 1.62503612e+00
-3.29943329e-01 6.92300498e-01 3.50639760e-01 -1.39642310e+00
8.81901860e-01 3.06602120e-01 7.34790921e-01 -4.63011235e-01
2.12754339e-01 2.95238256e-01 6.15279116e-02 -4.08790350e-01
1.67463467e-01 -2.63631701e-01 2.05689207e-01 3.51398885e-01
2.56099820e-01 8.46807137e-02 3.99134398e-01 -1.32692963e-01
8.90092373e-01 5.37386417e-01 3.87577266e-01 -2.87000924e-01
8.76048684e-01 3.21941860e-02 2.55174369e-01 6.22175634e-01
-3.19144249e-01 7.31514275e-01 2.00979650e-01 -5.70465207e-01
-1.04575038e+00 -7.86601663e-01 -4.44200516e-01 1.10465825e+00
1.86420098e-01 -2.03922004e-01 -7.11855292e-01 -8.47466648e-01
9.13514644e-02 3.81435491e-02 -9.21728373e-01 2.62074947e-01
-5.58025420e-01 -7.77927935e-01 2.44455978e-01 2.89738894e-01
1.86618000e-01 -1.35823596e+00 -7.63393521e-01 2.24675536e-01
2.17489198e-01 -1.27102816e+00 -6.86582550e-02 1.46414086e-01
-9.29325104e-01 -1.34072018e+00 -1.12442946e+00 -6.87004328e-01
6.02939904e-01 5.29158831e-01 1.32384396e+00 2.66876340e-01
-5.18390954e-01 8.30172956e-01 -5.97792327e-01 -3.83230835e-01
-5.40031314e-01 9.61356163e-02 -4.88609225e-02 3.85729134e-01
2.33673871e-01 -5.07035315e-01 -2.69999295e-01 6.66598022e-01
-1.06423175e+00 -1.52190641e-01 7.05608964e-01 6.15838110e-01
1.69605806e-01 -2.59624392e-01 3.31880838e-01 -9.34730589e-01
5.69065399e-02 -1.09158643e-01 -7.77567804e-01 4.10554886e-01
-4.40209538e-01 -2.20389925e-02 6.80109441e-01 -4.48268086e-01
-8.84736180e-01 3.67752552e-01 -1.10409118e-01 -4.17362481e-01
-6.23454809e-01 -3.30483392e-02 -1.30031571e-01 -5.67064345e-01
5.15103698e-01 3.00836235e-01 3.60579848e-01 -6.88980639e-01
3.69606853e-01 3.66511852e-01 3.71030331e-01 -4.81012344e-01
1.15197122e+00 6.67744517e-01 2.76323467e-01 -9.30856287e-01
-6.98519886e-01 -9.31251049e-01 -9.96140778e-01 -4.85003799e-01
7.34583080e-01 -5.08083940e-01 -3.41664463e-01 6.61639214e-01
-9.84878719e-01 -3.65348905e-01 -2.26101398e-01 2.64760375e-01
-7.79165149e-01 9.16526318e-01 -2.00577050e-01 -7.41667509e-01
2.03010179e-02 -1.16898263e+00 1.13104832e+00 1.05216749e-01
-5.12095615e-02 -1.33970690e+00 4.01103765e-01 1.93847820e-01
4.20893878e-01 3.80924851e-01 4.04568762e-01 -5.88196158e-01
-3.08583945e-01 -3.71124893e-01 -2.29240581e-01 5.42225242e-01
2.63753444e-01 1.28923818e-01 -9.54956830e-01 -5.53325295e-01
-2.51443963e-03 -1.29241541e-01 9.61135268e-01 2.27185577e-01
7.51584411e-01 2.58110732e-01 -6.11661553e-01 4.47618484e-01
1.52368140e+00 -1.21929519e-01 5.45669496e-01 7.06199586e-01
5.08876503e-01 8.13322246e-01 5.34589887e-01 -1.73237398e-01
9.45670083e-02 1.49422717e+00 5.19611180e-01 -2.95560211e-01
-2.03205124e-01 5.01906633e-01 3.52815300e-01 5.40073872e-01
-5.63701272e-01 1.07543588e-01 -6.66249275e-01 5.20726323e-01
-2.15974545e+00 -1.08734584e+00 -4.49860007e-01 2.51471758e+00
5.03233790e-01 8.87526721e-02 8.14695895e-01 1.41865939e-01
8.49908948e-01 1.58018246e-01 -6.03707582e-02 -2.05814734e-01
-3.74842957e-02 1.44101590e-01 3.85162592e-01 2.38990083e-01
-1.43151808e+00 6.66474581e-01 6.24881077e+00 8.08547258e-01
-1.17219520e+00 1.28387630e-01 1.16530336e-01 3.68100315e-01
3.73594284e-01 -1.01717353e-01 -8.46899629e-01 4.98263150e-01
6.02078199e-01 1.20935693e-01 8.92386436e-02 9.48142707e-01
-1.18573986e-01 -9.88986865e-02 -1.12302911e+00 1.00199389e+00
5.25964141e-01 -8.39727104e-01 -3.39992315e-01 9.89446267e-02
5.24377823e-01 -1.79664284e-01 7.55185708e-02 2.38697037e-01
-8.71977769e-03 -6.91842794e-01 5.80567777e-01 3.78757149e-01
3.98530513e-01 -7.23233745e-02 6.59310281e-01 2.82337695e-01
-1.35730743e+00 5.84072992e-02 -3.72167021e-01 1.18011117e-01
2.13184088e-01 2.71098435e-01 -2.80875206e-01 1.01668990e+00
7.67608225e-01 9.26366925e-01 -9.60833728e-01 1.79313684e+00
-1.69931743e-02 2.77937084e-01 -2.97590047e-01 9.68642384e-02
2.77655393e-01 -3.60920578e-01 7.12515831e-01 1.41410971e+00
4.29095067e-02 -6.49463475e-01 8.45323354e-02 2.79857278e-01
4.32493538e-01 3.57608616e-01 -6.06642306e-01 5.09557843e-01
-2.84942746e-01 1.62195110e+00 -1.14033830e+00 -3.41876626e-01
-6.50705576e-01 1.05430090e+00 3.87163222e-01 3.66452113e-02
-1.00841570e+00 -6.83518033e-03 3.24405491e-01 2.40546584e-01
7.08807528e-01 -4.71698381e-02 3.02785724e-01 -1.30305302e+00
1.46079719e-01 -8.77622545e-01 4.39439595e-01 -5.94356894e-01
-1.38210917e+00 7.85717607e-01 4.12820756e-01 -1.50422120e+00
-1.44911021e-01 -9.98110294e-01 -4.95575279e-01 5.18734038e-01
-1.63366461e+00 -1.24682271e+00 -3.80848557e-01 4.99822199e-01
5.62217057e-01 -1.59859419e-01 8.45894277e-01 4.60887879e-01
-6.12245619e-01 2.92730093e-01 2.76614606e-01 2.89366413e-02
1.13530409e+00 -1.61167741e+00 1.69103652e-01 7.70597279e-01
4.72864389e-01 3.90691817e-01 9.59833741e-01 -2.05611393e-01
-1.20036280e+00 -9.47755516e-01 7.49561429e-01 -7.05862403e-01
7.88885117e-01 -4.07153398e-01 -8.77351940e-01 5.55909812e-01
2.48291165e-01 3.82381111e-01 3.26371133e-01 1.29069388e-01
-3.69915336e-01 -1.91408291e-01 -8.67719114e-01 3.10609818e-01
1.01277387e+00 2.41300445e-02 -8.05731714e-01 2.72513181e-01
6.22038767e-02 -2.49538273e-01 -6.53566539e-01 3.78714353e-01
5.45714915e-01 -1.20099044e+00 1.09826541e+00 -5.48078179e-01
-7.02355802e-02 -4.65854585e-01 2.09983885e-01 -1.01389945e+00
-1.21662490e-01 -6.59082830e-01 -1.99869290e-01 1.15794158e+00
6.43247142e-02 -7.54030645e-01 7.19872415e-01 1.85955629e-01
5.48149608e-02 -5.11718869e-01 -1.05247760e+00 -1.12622941e+00
1.06212096e-02 -9.18007866e-02 1.10868923e-02 8.09185326e-01
-4.60903555e-01 2.32896283e-01 -5.72539449e-01 -2.40007088e-01
8.66067588e-01 3.24947268e-01 1.14176059e+00 -1.72808444e+00
-4.66022253e-01 -8.62882912e-01 -8.87684882e-01 -9.98901606e-01
9.21166912e-02 -7.81963229e-01 -6.23634085e-02 -1.27096641e+00
2.52966106e-01 -5.60269952e-01 -2.49478891e-01 3.86370629e-01
-1.97891399e-01 6.28210306e-01 3.91665578e-01 2.80181557e-01
-9.15485680e-01 1.55429497e-01 9.71601486e-01 -6.50876164e-02
1.75572813e-01 4.55965668e-01 -2.57329583e-01 8.47067535e-01
5.30510902e-01 -4.68833089e-01 -2.33820528e-01 -1.28909439e-01
1.23095974e-01 -4.24785674e-01 5.91929138e-01 -1.14956415e+00
1.55876681e-01 1.78293169e-01 2.46919870e-01 -1.89994872e-01
3.79154742e-01 -1.20774758e+00 2.90902436e-01 3.42992544e-01
1.21820234e-01 7.17244297e-02 2.66317993e-01 5.82943320e-01
-1.39601141e-01 -5.57359159e-01 9.51599002e-01 -2.13085160e-01
-8.43383968e-01 2.30822027e-01 -2.30392948e-01 -1.01678632e-01
1.44641376e+00 -6.09567106e-01 1.38645619e-01 -2.13637173e-01
-1.02439761e+00 -6.31933883e-02 8.70509267e-01 5.22875905e-01
2.65718065e-02 -1.12356377e+00 -6.71137452e-01 -2.39722148e-01
2.49647528e-01 -2.85832435e-01 1.65631101e-01 1.19288194e+00
-5.15377879e-01 5.13629198e-01 -5.24859369e-01 -9.50170755e-01
-1.74929738e+00 8.88073564e-01 4.02078569e-01 -4.12253380e-01
-8.07366848e-01 4.20342922e-01 1.98259890e-01 -2.18752444e-01
3.63609582e-01 4.01046053e-02 -3.19767028e-01 1.96444511e-01
3.55648667e-01 1.70351401e-01 1.43146113e-01 -1.06714451e+00
-3.90630215e-01 9.99975801e-01 5.02410997e-03 1.87757090e-01
1.39217138e+00 -3.01454812e-02 -6.17177151e-02 4.18220699e-01
9.76920664e-01 -1.02499433e-01 -1.42255616e+00 -2.29502201e-01
2.36808792e-01 -4.58953917e-01 -2.23191887e-01 -5.45178294e-01
-1.17174315e+00 7.36042500e-01 6.17569685e-01 7.22220004e-01
1.18700945e+00 6.27838373e-02 2.11542383e-01 8.83662105e-02
5.73430657e-01 -7.77638674e-01 -2.95013543e-02 1.60199314e-01
7.55816162e-01 -1.35100126e+00 1.25740707e-01 -5.62881947e-01
-4.80832100e-01 1.25406826e+00 4.34391558e-01 -1.90881312e-01
4.53342617e-01 1.55136451e-01 1.50429502e-01 -1.05384856e-01
-3.84247959e-01 -6.66873634e-01 4.97450799e-01 8.29290092e-01
4.21181738e-01 -3.17573190e-01 -4.70860869e-01 -4.05160636e-02
3.91136676e-01 -1.79580674e-01 2.40633577e-01 9.38223422e-01
-3.94127756e-01 -1.84107697e+00 -6.31922185e-01 2.45937347e-01
-4.13933814e-01 4.28146243e-01 -5.68604469e-01 1.25027704e+00
3.04061353e-01 5.85152745e-01 -3.02510798e-01 5.79359978e-02
3.98933798e-01 1.86615154e-01 1.02156055e+00 -5.82728922e-01
-8.16059172e-01 1.95649102e-01 -5.15620336e-02 -4.47475016e-01
-1.21573436e+00 -9.31499779e-01 -4.31529641e-01 4.87816632e-02
-5.08994043e-01 1.44531310e-01 5.30417144e-01 1.02969778e+00
-1.16509460e-01 2.33066902e-01 4.51659113e-01 -1.21437144e+00
-3.82193863e-01 -8.24912965e-01 -5.30402005e-01 7.18645632e-01
1.70426235e-01 -1.18489134e+00 -2.26822138e-01 2.30101958e-01] | [8.66832447052002, -0.7928475737571716] |
9a16504a-161c-46ca-9c22-7148ef92f95f | north-sami-dialect-identification-with-self | 2305.11864 | null | https://arxiv.org/abs/2305.11864v1 | https://arxiv.org/pdf/2305.11864v1.pdf | North Sámi Dialect Identification with Self-supervised Speech Models | The North S\'{a}mi (NS) language encapsulates four primary dialectal variants that are related but that also have differences in their phonology, morphology, and vocabulary. The unique geopolitical location of NS speakers means that in many cases they are bilingual in S\'{a}mi as well as in the dominant state language: Norwegian, Swedish, or Finnish. This enables us to study the NS variants both with respect to the spoken state language and their acoustic characteristics. In this paper, we investigate an extensive set of acoustic features, including MFCCs and prosodic features, as well as state-of-the-art self-supervised representations, namely, XLS-R, WavLM, and HuBERT, for the automatic detection of the four NS variants. In addition, we examine how the majority state language is reflected in the dialects. Our results show that NS dialects are influenced by the state language and that the four dialects are separable, reaching high classification accuracy, especially with the XLS-R model. | ['Katri Hiovain-Asikainen', 'Sofoklis Kakouros'] | 2023-05-19 | null | null | null | null | ['dialect-identification'] | ['natural-language-processing'] | [-4.07539815e-01 -2.12453768e-01 -3.09560120e-01 -3.52577865e-01
-4.12443042e-01 -8.49137664e-01 7.12389946e-01 -8.82975459e-02
-4.41791713e-01 4.37428594e-01 5.03979266e-01 -5.73152065e-01
-1.21732615e-01 -6.45387650e-01 -1.11060657e-01 -6.20019734e-01
4.75082397e-02 4.29917544e-01 1.58492744e-01 -9.42936301e-01
-2.61057559e-02 4.15682852e-01 -1.74767983e+00 -6.81845024e-02
9.82055366e-01 5.14695704e-01 4.31886792e-01 4.44628417e-01
-3.40020239e-01 1.79319933e-01 -7.74212122e-01 -2.00962260e-01
-4.86104488e-02 -5.35173535e-01 -6.39123738e-01 -3.06235552e-01
2.46846214e-01 1.89152136e-01 -2.98124611e-01 1.08938384e+00
2.94759631e-01 -1.17192250e-02 9.21968639e-01 -6.19929254e-01
-5.38985848e-01 1.16121733e+00 -2.11422369e-01 2.91490078e-01
4.27278042e-01 -5.03314324e-02 1.11293447e+00 -6.08572423e-01
5.76321006e-01 1.39886642e+00 5.08923292e-01 4.43693697e-01
-1.26364625e+00 -7.08934963e-01 3.30575615e-01 2.25212947e-01
-1.61424935e+00 -7.59541392e-01 7.64612734e-01 -6.04254544e-01
6.81842744e-01 3.33362252e-01 6.74332440e-01 9.64832485e-01
4.08755504e-02 6.28505647e-01 1.56611252e+00 -6.28608406e-01
2.27430508e-01 4.01829094e-01 4.59252983e-01 4.74843293e-01
-1.11133894e-02 1.31367996e-01 -5.15437126e-01 -1.05079755e-01
4.11280751e-01 -5.84633350e-01 -5.57138979e-01 1.87972769e-01
-1.03696644e+00 7.74326265e-01 -2.31884941e-02 1.00839257e+00
-1.44378796e-01 -3.88236851e-01 2.85604000e-01 4.79259431e-01
2.05874637e-01 1.73914492e-01 -7.30379999e-01 -2.06349045e-01
-6.47118151e-01 -8.68489519e-02 9.67006445e-01 5.48829198e-01
7.01380491e-01 5.16828179e-01 4.85989511e-01 1.34415901e+00
5.04408300e-01 8.57366621e-01 1.01399660e+00 -4.51249868e-01
3.03345144e-01 1.81739286e-01 -4.27990884e-01 -5.07184267e-01
-4.64259773e-01 -2.67365366e-01 -6.31295919e-01 8.82676840e-02
4.10416692e-01 -9.29485634e-02 -7.39671111e-01 2.06772256e+00
6.30597919e-02 -3.07980120e-01 4.44634736e-01 5.46506941e-01
9.17076826e-01 8.77828896e-01 2.09325999e-02 -3.33084732e-01
1.35507417e+00 -4.58478719e-01 -6.62257552e-01 -3.15203875e-01
3.95693034e-01 -7.78388500e-01 1.24315310e+00 2.74470866e-01
-9.40110564e-01 -6.93686545e-01 -1.05539966e+00 3.58758748e-01
-5.37871957e-01 3.05605352e-01 2.44877383e-01 1.23484576e+00
-1.08802485e+00 4.67417151e-01 -7.31567025e-01 -4.93933797e-01
-5.67915320e-01 2.61378318e-01 -4.96255547e-01 5.69803417e-01
-1.59547853e+00 9.13305044e-01 1.78786486e-01 2.01340705e-01
-4.03311044e-01 -3.27106327e-01 -1.08781064e+00 -1.10010587e-01
-2.97931790e-01 1.89243719e-01 1.12331653e+00 -8.86373341e-01
-1.87385499e+00 9.53387141e-01 -1.58190891e-01 1.19386487e-01
1.61469981e-01 1.21812053e-01 -1.10626864e+00 -5.02188206e-02
1.20022632e-01 2.10395426e-01 3.95623386e-01 -1.08301079e+00
-7.54956663e-01 -5.73138773e-01 -4.00721967e-01 2.73411751e-01
-2.41754174e-01 3.49132687e-01 -2.06839200e-02 -7.01587617e-01
4.15719777e-01 -9.72834647e-01 6.71754777e-02 -7.74727046e-01
-3.15522492e-01 -2.55665928e-01 3.96287590e-01 -8.80392611e-01
1.26161420e+00 -2.45908570e+00 1.99549690e-01 5.87289870e-01
-2.54663378e-01 2.82040745e-01 3.85186449e-02 5.22316515e-01
-1.45084942e-02 1.63844898e-01 -1.61571518e-01 -1.38878614e-01
2.17318356e-01 6.62385285e-01 -8.92974287e-02 5.42637467e-01
-1.15499564e-01 4.00793612e-01 -6.08805001e-01 -1.38347119e-01
1.96829021e-01 4.15524006e-01 -8.50369558e-02 -9.97952074e-02
2.34866604e-01 5.03015816e-01 -1.55351898e-02 3.97389412e-01
5.80181301e-01 8.54222953e-01 5.63485324e-01 9.38543230e-02
-7.81604707e-01 8.53694797e-01 -1.20738828e+00 1.06107676e+00
-6.61080062e-01 5.82021475e-01 5.19186497e-01 -8.30102265e-01
1.04037988e+00 3.93080235e-01 -4.07603756e-02 -6.13043129e-01
7.22326562e-02 7.49024451e-01 5.37674904e-01 -1.92657396e-01
6.69852376e-01 -2.14167997e-01 -4.47456926e-01 1.97092757e-01
1.07595742e-01 -1.34177983e-01 3.07433933e-01 -2.73720115e-01
4.65162486e-01 -1.29760548e-01 5.84940076e-01 -9.59242523e-01
5.94776928e-01 -4.24107462e-01 8.56059134e-01 4.52789724e-01
-2.15797797e-01 3.83855850e-01 4.74509180e-01 2.36510918e-01
-3.93449068e-01 -1.37774551e+00 -6.66461408e-01 1.23659182e+00
1.69135228e-01 -2.07963929e-01 -5.63422859e-01 -8.01233724e-02
-9.89772230e-02 8.84110808e-01 -3.22357178e-01 -3.17618161e-01
-8.88953865e-01 -6.17443323e-01 8.04704249e-01 4.60036874e-01
1.80233359e-01 -1.08476019e+00 -2.05517992e-01 1.44334123e-01
-2.33019918e-01 -9.03653622e-01 -4.11825180e-01 5.11170685e-01
-4.30602759e-01 -7.29781210e-01 -4.93261456e-01 -1.13956594e+00
7.04887658e-02 -2.04151407e-01 9.01908636e-01 -4.19425219e-01
2.84428328e-01 2.37884343e-01 -4.16447401e-01 -3.25762480e-01
-1.07025003e+00 2.94106215e-01 3.96386236e-01 3.44141051e-02
3.06642562e-01 -5.71941316e-01 -6.40338659e-02 4.03455615e-01
-5.76928139e-01 -5.31890213e-01 3.99854600e-01 2.59834021e-01
3.41984928e-01 -3.14766653e-02 5.42090535e-01 -7.64921546e-01
2.82168627e-01 -4.77484137e-01 -3.54188234e-01 1.94209907e-02
-3.05892617e-01 1.70825268e-04 6.62192404e-01 -3.89391840e-01
-1.05240214e+00 -7.25168511e-02 -8.38970184e-01 1.46609440e-01
-5.32829523e-01 5.19564509e-01 -6.10966861e-01 2.66419649e-01
5.15855134e-01 4.93030399e-01 -8.42843130e-02 -6.99275136e-01
2.43288636e-01 1.10171211e+00 7.59985983e-01 -4.21593130e-01
7.14687526e-01 2.00787097e-01 -6.08176172e-01 -1.68341017e+00
-1.83630317e-01 -5.68699360e-01 -8.06696892e-01 -6.23813681e-02
8.78405869e-01 -7.76148617e-01 -3.36090237e-01 7.34291196e-01
-9.37246859e-01 -2.54167050e-01 -2.42334142e-01 8.64474237e-01
-3.57497334e-01 3.54183823e-01 -8.16796780e-01 -8.82703364e-01
-2.96657179e-02 -1.28450263e+00 8.20506454e-01 7.26052672e-02
-5.14263153e-01 -1.02858567e+00 2.52112538e-01 -9.01169702e-02
1.63952962e-01 -3.12855989e-01 1.32652259e+00 -8.44887018e-01
3.00551713e-01 2.76959777e-01 4.14979219e-01 5.75136781e-01
5.15405416e-01 1.94302157e-01 -1.04714489e+00 -1.07825145e-01
1.72998488e-01 2.38477051e-01 8.89314055e-01 4.36037481e-01
-5.73009392e-03 -6.86144307e-02 4.35099714e-02 4.94238019e-01
1.07284689e+00 5.78761280e-01 3.20682377e-01 -3.48860398e-02
4.19463903e-01 8.18307757e-01 2.81834960e-01 -4.33076769e-02
5.00977516e-01 5.99307179e-01 -6.88287392e-02 1.31784603e-01
-1.79177180e-01 -4.29690443e-02 1.14833546e+00 1.73591292e+00
7.08545297e-02 5.49871475e-02 -9.10241783e-01 7.38270342e-01
-1.32785809e+00 -6.78541362e-01 -3.78934234e-01 2.23155499e+00
8.85192454e-01 2.69839559e-02 1.89973682e-01 5.08046329e-01
9.30606067e-01 3.32199454e-01 1.17519960e-01 -7.31808126e-01
-5.31782985e-01 3.50377142e-01 3.09851974e-01 6.36775732e-01
-8.84127557e-01 1.07161665e+00 6.68915176e+00 7.73621917e-01
-1.44531560e+00 -1.14574239e-01 1.32724583e-01 3.29194933e-01
-4.89600033e-01 -8.62776488e-02 -1.01430857e+00 4.75409925e-01
1.24528074e+00 1.03505842e-01 2.61037022e-01 5.66876531e-01
1.21728525e-01 -8.46614316e-02 -8.40011060e-01 5.95527351e-01
1.33084252e-01 -6.68192506e-01 -2.36808792e-01 8.34524408e-02
3.71177793e-01 3.17505509e-01 1.28075406e-01 3.57643098e-01
1.74139693e-01 -6.38038695e-01 1.32253230e+00 5.04438579e-03
8.57416868e-01 -6.99962437e-01 6.07660115e-01 2.47537255e-01
-1.52631521e+00 3.40679549e-02 -2.96846241e-01 -6.99045211e-02
3.07080775e-01 3.82264227e-01 -3.21716934e-01 5.76731682e-01
6.16173923e-01 2.80171841e-01 -3.68267983e-01 4.65433210e-01
-3.26260358e-01 1.21464467e+00 -3.76209110e-01 -1.31530449e-01
1.38354108e-01 -6.15533710e-01 8.09501946e-01 1.34251857e+00
3.93825889e-01 -1.86193317e-01 3.18589002e-01 5.46881378e-01
4.97460067e-01 5.03323674e-01 -4.41960812e-01 -2.72392165e-02
4.29154575e-01 8.57988894e-01 -7.80956626e-01 -1.47115037e-01
-4.70359653e-01 6.60172701e-01 -5.86607940e-02 2.97701955e-01
-3.34884912e-01 -5.37045598e-01 1.10745442e+00 1.75240159e-01
3.42437953e-01 -4.71344501e-01 2.96872761e-03 -1.00230813e+00
-9.43848416e-02 -7.25881577e-01 1.35634854e-01 -1.20736420e-01
-1.35756838e+00 1.12483740e+00 5.12784719e-02 -8.54098439e-01
-5.98133564e-01 -8.12507987e-01 -5.00135899e-01 1.00784540e+00
-1.50695539e+00 -1.06518459e+00 2.56907403e-01 5.81841767e-01
3.14023554e-01 -5.42261899e-01 9.84213412e-01 3.85916471e-01
-6.12955689e-01 5.47513485e-01 4.87514436e-01 2.91879267e-01
5.34196019e-01 -1.45982623e+00 4.64937210e-01 5.33613682e-01
5.05698681e-01 4.49721664e-01 5.33685684e-01 -4.15370047e-01
-8.90954494e-01 -7.91381419e-01 1.13582730e+00 6.96669742e-02
8.44613075e-01 -5.91411054e-01 -6.94713414e-01 4.89654928e-01
1.79603934e-01 -4.95837033e-01 8.46552074e-01 4.05724794e-01
-4.21735793e-01 -1.70894295e-01 -8.37271810e-01 7.50703037e-01
1.00523043e+00 -8.47910225e-01 -8.46224904e-01 -1.10903025e-01
6.36097848e-01 1.57478124e-01 -7.18868971e-01 1.51239216e-01
5.94815373e-01 -1.20437336e+00 4.50529963e-01 -1.63340420e-02
-1.76899254e-01 -1.48791507e-01 -4.67181921e-01 -1.67000687e+00
-3.02284122e-01 -6.01141274e-01 7.36349106e-01 1.71775484e+00
5.47605753e-01 -1.09588110e+00 3.05736482e-01 -1.23766445e-01
-3.82462114e-01 -9.07472968e-02 -1.31556451e+00 -9.95799303e-01
4.84145343e-01 -5.06594300e-01 5.94524384e-01 9.28634584e-01
9.87640098e-02 4.67021972e-01 1.11199044e-01 9.90214571e-02
6.47181571e-02 1.60204202e-01 1.42620146e-01 -1.46820307e+00
-3.51954490e-01 -8.31967711e-01 -6.33393109e-01 -7.80881703e-01
6.65378571e-01 -9.97792900e-01 1.92416325e-01 -1.07498312e+00
-6.34341598e-01 -5.72211087e-01 -3.04260015e-01 1.89929098e-01
8.70048702e-02 1.34693772e-01 5.79571277e-02 7.17998743e-02
3.41476649e-01 5.03179729e-01 6.19538367e-01 1.48115074e-03
-6.03656113e-01 3.63190800e-01 -5.47946692e-01 9.88512158e-01
9.25455987e-01 -9.19479355e-02 -2.80697197e-01 -2.57301867e-01
-5.49932532e-02 8.94977972e-02 -2.38045409e-01 -9.62805688e-01
-1.86184123e-01 4.70352136e-02 -1.54401749e-01 -3.44967335e-01
3.60581815e-01 -3.67636263e-01 9.85226780e-03 5.17353654e-01
-8.11340064e-02 2.15270475e-01 1.50552958e-01 -2.73734033e-02
-5.47219336e-01 -4.18486774e-01 1.01897728e+00 -1.87539477e-02
-6.45595968e-01 -2.51943022e-01 -1.13344133e+00 6.28754124e-02
6.81812525e-01 -3.66855174e-01 -1.40919704e-02 -4.15754825e-01
-6.19311512e-01 -2.37567410e-01 2.70033896e-01 4.18639719e-01
3.36642563e-02 -1.04181397e+00 -8.99906874e-01 6.64600015e-01
2.74968650e-02 -6.93143427e-01 1.83821589e-01 6.83423698e-01
-4.70740288e-01 3.02257776e-01 -3.48451912e-01 -4.10051048e-01
-1.06808054e+00 6.74664527e-02 2.65039802e-01 4.78225537e-02
-2.50581622e-01 6.45360649e-01 3.94040018e-01 -1.08774686e+00
-1.50836492e-02 -3.65213186e-01 -6.20322943e-01 4.67172086e-01
7.95563087e-02 3.75011176e-01 3.54081206e-02 -1.48414469e+00
-5.37664533e-01 6.79960310e-01 1.63414255e-01 -5.18131435e-01
8.85667622e-01 -2.76974648e-01 -4.65582609e-01 1.15497816e+00
1.09946549e+00 8.42688382e-01 -2.78549135e-01 -9.83925164e-03
3.92231755e-02 7.80428350e-02 -1.91314042e-01 -5.28307438e-01
-8.73987973e-01 6.96665943e-01 5.29309273e-01 3.43924463e-01
9.03249085e-01 2.86218107e-01 6.41608536e-01 4.45618927e-02
1.99548885e-01 -1.13505936e+00 -8.32904518e-01 8.39405596e-01
6.83184505e-01 -6.35763288e-01 -6.28138363e-01 -5.80373466e-01
-7.40197539e-01 8.42633247e-01 2.84188241e-01 -6.05737269e-02
1.04095459e+00 3.18890154e-01 6.84796810e-01 1.63571090e-01
-4.38593805e-01 -5.84727645e-01 1.75887212e-01 8.10361505e-01
7.21566200e-01 4.82800573e-01 -6.38777971e-01 7.30061591e-01
-1.07073402e+00 -8.87035310e-01 5.69007754e-01 5.75191855e-01
-3.64033073e-01 -1.33333552e+00 -5.19264579e-01 3.18650275e-01
-3.33257675e-01 -7.39048123e-02 -3.76034439e-01 8.97957802e-01
2.53137797e-01 1.13406777e+00 3.49488348e-01 -5.21395326e-01
4.68558520e-01 3.32831770e-01 1.71590582e-01 -6.15198493e-01
-8.39046478e-01 4.23818558e-01 3.10769558e-01 5.84771819e-02
-2.34226912e-01 -9.45037723e-01 -1.35531950e+00 -1.97601631e-01
-3.12649488e-01 5.00982225e-01 7.34136820e-01 9.72372949e-01
-2.77222902e-01 2.89411634e-01 8.56739938e-01 -7.26541281e-01
-6.27775013e-01 -1.08870852e+00 -1.31790733e+00 1.10903166e-01
2.37779453e-01 -4.07351017e-01 -5.98003924e-01 -3.77255619e-01] | [14.321491241455078, 6.746435642242432] |
b921cc74-7387-4c0c-8fa4-6af46135cba3 | a-machine-learning-approach-to-doa-estimation | 2009.12858 | null | https://arxiv.org/abs/2009.12858v2 | https://arxiv.org/pdf/2009.12858v2.pdf | A Machine Learning Approach to DoA Estimation and Model Order Selection for Antenna Arrays with Subarray Sampling | In this paper, we study the problem of direction of arrival estimation and model order selection for systems employing subarray sampling. Thereby, we focus on scenarios, where the number of active sources is not smaller than the number of simultaneously sampled antenna elements. For this purpose, we propose new schemes based on neural networks and estimators that combine neural networks with gradient steps on the likelihood function. These methods are able to outperform existing estimators in terms of mean squared error and model selection accuracy, especially in the low snapshot domain, at a drastically lower computational complexity. | ['Wolfgang Utschick', 'Andreas Barthelme'] | 2020-09-27 | null | null | null | null | ['direction-of-arrival-estimation'] | ['audio'] | [ 2.91787028e-01 -4.09370631e-01 6.62005246e-02 -2.99604505e-01
-7.99706638e-01 -5.62510490e-01 2.50034124e-01 9.25408974e-02
-4.37478364e-01 1.02509952e+00 -2.59160876e-01 -4.63548034e-01
-7.25521147e-01 -7.27751732e-01 -2.38939315e-01 -9.46368814e-01
-4.50282633e-01 2.18495414e-01 -1.68779925e-01 6.61285594e-02
1.84986845e-01 9.17352319e-01 -1.09889865e+00 -2.92768300e-01
6.52148664e-01 1.51875722e+00 -2.18342498e-01 7.21298218e-01
4.86854792e-01 5.72461367e-01 -7.42379487e-01 4.35277540e-03
2.05567032e-01 -5.81287563e-01 1.10972226e-01 -1.38369739e-01
2.11999133e-01 -1.91607729e-01 -1.53015465e-01 9.03607607e-01
1.01566863e+00 1.61890343e-01 8.75386715e-01 -8.97850811e-01
1.50198758e-01 6.39507651e-01 -3.50674361e-01 3.87469769e-01
-8.67932104e-03 -5.66003859e-01 3.37091744e-01 -1.16446400e+00
2.69475859e-02 7.75285184e-01 9.72170234e-01 -5.38939014e-02
-1.03684461e+00 -5.91630459e-01 -1.83672398e-01 1.87708899e-01
-1.72273791e+00 -9.36051369e-01 9.64814782e-01 -1.82628199e-01
4.36949044e-01 4.04301316e-01 3.59583825e-01 6.05699718e-01
4.95921746e-02 4.22749609e-01 7.32476413e-01 -9.84098971e-01
6.25102818e-01 1.32973596e-01 6.26350641e-02 4.16722119e-01
7.64414430e-01 -1.11660339e-01 -2.60162652e-01 -5.32246053e-01
8.20337176e-01 -4.12903637e-01 -3.21018904e-01 -5.99508047e-01
-9.60425079e-01 1.01816905e+00 6.81592338e-03 6.61946774e-01
-7.40907907e-01 2.24671826e-01 -8.43679681e-02 2.35757172e-01
5.97532630e-01 8.65614951e-01 -2.76470274e-01 2.78507382e-01
-1.20181298e+00 -2.46033762e-02 1.08897066e+00 6.77051127e-01
3.00529331e-01 6.60848022e-01 -1.14599086e-01 9.09392178e-01
3.38933557e-01 7.33833730e-01 3.82683502e-04 -7.88024485e-01
2.21201539e-01 4.52959836e-02 3.90915811e-01 -1.25246465e+00
-7.81628430e-01 -1.34573340e+00 -1.10594201e+00 -1.12510942e-01
5.54827929e-01 -1.02737784e+00 -4.89008009e-01 1.49552500e+00
4.17697966e-01 4.23656292e-02 2.73232460e-01 7.00954378e-01
2.35001996e-01 6.94401264e-01 -5.06300151e-01 -9.17682469e-01
9.68798697e-01 -5.83661735e-01 -9.53673840e-01 -1.66582212e-01
3.02763969e-01 -8.37836504e-01 -6.67986646e-02 7.73404062e-01
-1.38087380e+00 -4.33870137e-01 -1.23533547e+00 9.04040694e-01
-3.82549055e-02 9.31988955e-01 5.17805219e-01 1.04744542e+00
-1.00332224e+00 2.06325144e-01 -5.80510199e-01 -1.01969495e-01
8.05661827e-02 3.42220098e-01 2.72236675e-01 1.83263466e-01
-1.07628512e+00 5.05584657e-01 3.05046737e-01 6.38926029e-01
-3.79678935e-01 -4.59764272e-01 -6.31788552e-01 3.03101450e-01
3.45746756e-01 -3.92495304e-01 1.15827549e+00 -1.03618217e+00
-1.47708869e+00 -1.06035538e-01 -2.02592015e-01 -5.17340958e-01
3.21566492e-01 -7.01611489e-02 -7.46993959e-01 1.16213821e-01
-3.48684222e-01 -2.88781434e-01 1.02347541e+00 -1.05460978e+00
-7.99857676e-01 -2.84531832e-01 -1.75322250e-01 -4.98363227e-02
-4.02821094e-01 -1.73628375e-01 -6.26526102e-02 -5.89659393e-01
5.34434259e-01 -6.68886662e-01 -6.54488027e-01 -1.29482195e-01
-2.38871142e-01 7.32021779e-02 2.48329341e-01 -5.16765416e-01
1.50703371e+00 -2.05509973e+00 -2.26455092e-01 9.01540637e-01
-1.04863644e-01 3.49415034e-01 1.30197495e-01 7.02545524e-01
1.90814897e-01 -4.73834753e-01 -4.32047322e-02 1.71744227e-01
-3.50886732e-01 -3.09786558e-01 2.91868281e-02 6.81059122e-01
-3.17282253e-03 1.25057101e-01 -6.14950001e-01 -8.03647190e-02
5.41046560e-02 2.35248685e-01 -3.60819548e-01 3.02577466e-01
3.59216034e-01 3.28771383e-01 -6.40688121e-01 5.09069026e-01
9.31721151e-01 -2.71701872e-01 4.39967334e-01 -3.99334162e-01
-4.12229538e-01 1.59672543e-01 -1.82564986e+00 9.94292736e-01
-9.12000597e-01 8.40493202e-01 4.37874466e-01 -1.54127705e+00
1.07651234e+00 4.68162179e-01 6.78542256e-01 -4.39953417e-01
2.87635833e-01 6.75614536e-01 1.09407552e-01 -3.09688628e-01
1.36070818e-01 6.68975115e-02 1.12253822e-01 2.90123463e-01
-4.74640206e-02 2.71392494e-01 4.04854834e-01 -4.10510182e-01
7.87699521e-01 -6.05097473e-01 8.50551128e-01 -2.86531866e-01
7.66507030e-01 -4.63533252e-01 3.49849910e-01 1.16598427e+00
1.89742714e-01 6.42197505e-02 4.04507399e-01 -3.15246642e-01
-7.37915576e-01 -7.93814361e-01 -3.67347926e-01 7.64700234e-01
1.58595920e-01 2.84219086e-01 -5.58146954e-01 -1.52301371e-01
-1.19752198e-01 1.00601637e+00 -3.08029830e-01 1.14238374e-02
-7.79766440e-01 -1.02412438e+00 4.98540074e-01 2.62185603e-01
2.89064139e-01 -4.11808163e-01 -8.77480149e-01 5.40201962e-01
2.19635338e-01 -1.00138628e+00 1.31194919e-01 6.62922442e-01
-8.98147106e-01 -7.53184259e-01 -1.05046582e+00 -6.64344788e-01
7.18558550e-01 1.42861724e-01 9.48777497e-01 -4.45362926e-01
1.80420563e-01 3.39601725e-01 -3.47502351e-01 -4.88808274e-01
-2.87448596e-02 2.85825253e-01 6.01762347e-02 4.82225806e-01
-1.36632562e-01 -6.68217897e-01 -4.31424230e-01 4.24394995e-01
-6.05163157e-01 -4.31666225e-01 7.49246478e-01 9.45192993e-01
1.78442940e-01 1.89836100e-01 8.70925128e-01 -5.32567203e-01
7.79177010e-01 -6.86419547e-01 -1.04549670e+00 3.45510870e-01
-6.06847465e-01 -4.25115153e-02 7.68145680e-01 -3.76408041e-01
-1.14894223e+00 4.00889553e-02 -3.35562408e-01 2.30891481e-01
-7.87255540e-02 7.06606448e-01 6.05915487e-02 -5.20720959e-01
7.45697439e-01 3.20801347e-01 -4.50360656e-01 -4.86224264e-01
-1.07583173e-01 7.43237793e-01 6.55113086e-02 7.63585344e-02
5.09958148e-01 2.50613272e-01 5.04181445e-01 -1.08473396e+00
-5.64233124e-01 -4.43925858e-01 -3.83975178e-01 -2.24097297e-01
-3.90877165e-02 -6.39487505e-01 -4.15792435e-01 3.92212600e-01
-1.13975036e+00 1.95025980e-01 6.94321468e-02 1.22190273e+00
-3.83033991e-01 6.64007440e-02 -1.95036694e-01 -1.42536604e+00
-2.11101562e-01 -6.84119165e-01 6.11394227e-01 2.92015135e-01
6.25295565e-02 -1.05101132e+00 -5.00017218e-02 -5.39376676e-01
7.85272896e-01 1.21828072e-01 6.25819564e-01 -8.90590608e-01
-3.30370963e-01 -6.85773790e-01 -2.34777525e-01 1.95742249e-01
1.28292575e-01 -3.92933935e-01 -6.44881666e-01 -3.83358777e-01
1.97563320e-01 2.06388056e-01 3.88956517e-01 1.14461529e+00
9.05990303e-01 -4.24546927e-01 -5.13322532e-01 7.03883708e-01
1.67159414e+00 8.34022939e-01 3.60450387e-01 1.54963702e-01
1.01219580e-01 3.93916905e-01 4.41902727e-01 8.17305624e-01
-2.78846800e-01 7.21388936e-01 1.71021268e-01 -1.11870110e-01
2.64169812e-01 4.65151757e-01 -3.45088601e-01 6.84518874e-01
1.59236088e-01 -9.96396124e-01 -6.98104084e-01 5.25106609e-01
-1.64405596e+00 -7.63780892e-01 -2.18376368e-01 2.32708192e+00
2.42903292e-01 1.37600228e-01 8.76836628e-02 2.56178111e-01
7.09311128e-01 1.50200874e-01 -3.92766923e-01 -3.22021216e-01
-3.89756151e-02 -3.87573205e-02 7.38865256e-01 4.22698557e-01
-1.12720215e+00 -6.02301769e-02 6.83400917e+00 1.03221118e+00
-1.22220623e+00 -3.95439491e-02 5.67622483e-01 -9.91945900e-03
3.50619592e-02 -3.52140337e-01 -9.01287615e-01 4.27091002e-01
1.24204266e+00 1.81093216e-01 1.12083316e-01 6.52823091e-01
1.89220861e-01 -3.17278087e-01 -7.47159898e-01 9.16530669e-01
2.49180689e-01 -1.17741621e+00 -4.55591589e-01 -2.30874985e-01
8.25897396e-01 -4.00041074e-01 -9.60361306e-03 -7.06776083e-02
-5.34776211e-01 -6.71609402e-01 4.56891298e-01 8.36250842e-01
2.54020751e-01 -9.93247688e-01 1.09681702e+00 4.96945322e-01
-8.45673442e-01 -4.51120138e-01 -1.96497560e-01 -2.00824916e-01
4.72650617e-01 1.23498809e+00 -9.39337313e-01 6.93677008e-01
2.29825631e-01 3.60964909e-02 -2.67897211e-02 1.84493268e+00
1.43426701e-01 9.78885293e-01 -7.92692542e-01 -7.23961473e-01
3.59671980e-01 -9.27417874e-02 6.53976440e-01 1.20203161e+00
1.02198231e+00 5.18850759e-02 1.50735870e-01 3.43508631e-01
2.47955322e-01 1.07724153e-01 -3.89980525e-01 3.42722237e-01
7.14623868e-01 1.06053472e+00 -8.50146234e-01 -2.14520961e-01
-2.23524690e-01 2.87580460e-01 -2.59254366e-01 6.31440580e-01
-4.88732934e-01 -8.73183370e-01 4.90321890e-02 -3.79834063e-02
5.94124377e-01 -5.04867494e-01 -3.29847038e-01 -6.67681694e-01
5.27547719e-03 -6.32642508e-01 2.63899475e-01 -2.46509627e-01
-1.09390247e+00 6.35489166e-01 1.65475532e-01 -1.45985520e+00
-7.68211305e-01 -6.48456395e-01 -4.29063380e-01 7.77994454e-01
-1.36033213e+00 -5.92404842e-01 1.44734770e-01 -2.46191267e-02
4.08843428e-01 -4.59832519e-01 7.06978261e-01 6.06672645e-01
-4.02321637e-01 6.67414367e-01 1.01983142e+00 -1.39225945e-01
3.19755286e-01 -8.65338981e-01 -6.89130202e-02 9.10299242e-01
1.43679768e-01 4.70906317e-01 1.02776551e+00 -1.25576407e-01
-1.14727223e+00 -7.71363556e-01 8.65104198e-01 1.63762793e-01
2.89468020e-01 -2.98485488e-01 -3.86864275e-01 3.47645879e-02
1.23398043e-01 3.60997990e-02 7.56941915e-01 1.31275415e-01
5.11196494e-01 -5.09052396e-01 -9.62273479e-01 4.89173055e-01
4.81226623e-01 2.06903681e-01 7.82012418e-02 3.94808263e-01
-1.62738279e-01 -3.07583153e-01 -7.89346993e-01 6.67570353e-01
8.48645389e-01 -8.37022424e-01 1.15099859e+00 -5.70375361e-02
-4.40474957e-01 -1.71977982e-01 -4.86542732e-01 -1.37406158e+00
-5.62908292e-01 -4.85059232e-01 -3.84063989e-01 9.68345702e-01
6.96545422e-01 -7.56375730e-01 7.43430674e-01 -2.15423897e-01
2.12108716e-01 -9.16355610e-01 -1.23429966e+00 -9.35097456e-01
-4.69536543e-01 -3.29454213e-01 3.06073785e-01 5.27245998e-01
-4.99890924e-01 3.09085488e-01 -7.50943661e-01 5.36526680e-01
8.51001561e-01 8.64892527e-02 2.78817594e-01 -1.18003762e+00
-5.99334180e-01 -3.90726507e-01 -8.19186941e-02 -1.10808325e+00
-2.25307047e-01 -1.16559431e-01 2.30270147e-01 -1.34834182e+00
-5.37553668e-01 -5.82683742e-01 -4.34101999e-01 -1.28358185e-01
7.23881349e-02 3.43215734e-01 -2.59582251e-01 -2.30587229e-01
-5.29303551e-01 2.82029003e-01 6.35944784e-01 -1.19775524e-02
-3.50501239e-01 8.77006471e-01 -3.61665696e-01 9.51618791e-01
9.27911103e-01 -5.02933741e-01 -3.12506825e-01 -4.13953692e-01
3.38221192e-01 4.61611807e-01 8.22972655e-02 -1.49533415e+00
3.24499220e-01 7.09324628e-02 7.85548985e-01 -8.00191164e-01
4.01001662e-01 -9.14884448e-01 1.90503240e-01 5.72129130e-01
-5.00054777e-01 -2.12505415e-01 1.33640647e-01 4.91357297e-01
-1.53569937e-01 -7.99241066e-01 6.86269939e-01 3.48217636e-01
-3.91280770e-01 -1.37719706e-01 -9.89645541e-01 -4.38817263e-01
8.00603330e-01 -2.91675299e-01 3.26623321e-01 -8.11581373e-01
-4.81878906e-01 -1.68551192e-01 -4.08096075e-01 -1.75162196e-01
3.75182420e-01 -1.39267755e+00 -7.18253136e-01 2.71647841e-01
-1.20711260e-01 -6.29454434e-01 3.22216094e-01 1.08842456e+00
-3.13351661e-01 8.74864042e-01 4.37329374e-02 -4.94485617e-01
-9.89374459e-01 1.20192803e-01 5.59331894e-01 -2.68622011e-01
3.26928735e-01 9.07613575e-01 -2.27110207e-01 -2.38183841e-01
2.83101916e-01 -1.24670006e-02 -5.65517902e-01 2.42983595e-01
4.88469303e-01 7.22487569e-01 2.98751324e-01 -2.41522074e-01
-5.00478745e-01 6.11696005e-01 3.89565557e-01 -8.70399997e-02
1.17911685e+00 -2.47925505e-01 -1.28141537e-01 4.51769531e-01
1.22379565e+00 4.15562928e-01 -9.24722135e-01 -2.65460491e-01
-5.58450865e-03 -3.92702758e-01 3.63989294e-01 -8.45862627e-01
-7.98144341e-01 6.71977460e-01 1.25442171e+00 7.24616468e-01
1.48931277e+00 -3.09682131e-01 2.88180918e-01 8.23186576e-01
3.29370052e-01 -9.24011290e-01 -3.10408711e-01 3.23199332e-01
6.44821763e-01 -8.36152077e-01 1.41869456e-01 -3.05231184e-01
1.40901282e-01 1.46993661e+00 2.70451009e-01 -3.22642505e-01
8.42489362e-01 2.76282996e-01 5.65581284e-02 1.82817265e-01
-4.96511936e-01 -6.46709651e-02 5.76504648e-01 6.10939026e-01
5.99716902e-01 -5.02178222e-02 -7.70558178e-01 4.07976419e-01
1.83826268e-01 -1.51999071e-01 4.42533731e-01 7.24712431e-01
-7.47862756e-01 -9.60407615e-01 -8.83610547e-01 7.42911994e-01
-4.99368221e-01 -5.12223952e-02 9.02593061e-02 5.33172607e-01
5.13695665e-02 1.17173767e+00 4.23880927e-02 1.83575377e-01
4.74251807e-01 -3.83247823e-01 4.84077245e-01 -8.59569833e-02
-5.45507260e-02 5.09430826e-01 6.00388110e-01 -1.83443241e-02
-3.55332196e-01 -7.56943703e-01 -3.13832372e-01 2.24216342e-01
-8.66665781e-01 6.42133951e-01 6.79600120e-01 8.59333217e-01
2.14873075e-01 7.48699784e-01 1.23416746e+00 -8.66154253e-01
-7.38799453e-01 -1.00249195e+00 -8.92406404e-01 -6.56597733e-01
5.70286751e-01 -5.36273181e-01 -2.17653140e-01 -3.07807237e-01] | [6.538086414337158, 1.308972716331482] |
74730f9d-88af-4ba4-9892-81873f9c5473 | towards-explainable-in-the-wild-video-quality | 2305.12726 | null | https://arxiv.org/abs/2305.12726v1 | https://arxiv.org/pdf/2305.12726v1.pdf | Towards Explainable In-the-Wild Video Quality Assessment: a Database and a Language-Prompted Approach | The proliferation of in-the-wild videos has greatly expanded the Video Quality Assessment (VQA) problem. Unlike early definitions that usually focus on limited distortion types, VQA on in-the-wild videos is especially challenging as it could be affected by complicated factors, including various distortions and diverse contents. Though subjective studies have collected overall quality scores for these videos, how the abstract quality scores relate with specific factors is still obscure, hindering VQA methods from more concrete quality evaluations (e.g. sharpness of a video). To solve this problem, we collect over two million opinions on 4,543 in-the-wild videos on 13 dimensions of quality-related factors, including in-capture authentic distortions (e.g. motion blur, noise, flicker), errors introduced by compression and transmission, and higher-level experiences on semantic contents and aesthetic issues (e.g. composition, camera trajectory), to establish the multi-dimensional Maxwell database. Specifically, we ask the subjects to label among a positive, a negative, and a neural choice for each dimension. These explanation-level opinions allow us to measure the relationships between specific quality factors and abstract subjective quality ratings, and to benchmark different categories of VQA algorithms on each dimension, so as to more comprehensively analyze their strengths and weaknesses. Furthermore, we propose the MaxVQA, a language-prompted VQA approach that modifies vision-language foundation model CLIP to better capture important quality issues as observed in our analyses. The MaxVQA can jointly evaluate various specific quality factors and final quality scores with state-of-the-art accuracy on all dimensions, and superb generalization ability on existing datasets. Code and data available at \url{https://github.com/VQAssessment/MaxVQA}. | ['Weisi Lin', 'Qiong Yan', 'Wenxiu Sun', 'Annan Wang', 'Jingwen Hou', 'Chaofeng Chen', 'Liang Liao', 'Erli Zhang', 'HaoNing Wu'] | 2023-05-22 | null | null | null | null | ['video-quality-assessment', 'video-quality-assessment'] | ['computer-vision', 'time-series'] | [-3.11983705e-01 -6.56914353e-01 -1.74598113e-01 -5.35215974e-01
-9.06103194e-01 -9.35520470e-01 2.83631772e-01 -1.22324452e-02
-6.35386109e-02 4.27712023e-01 6.57562256e-01 1.69218909e-02
-2.12950334e-01 -6.11682951e-01 -6.36487782e-01 -4.63531494e-01
-1.43675208e-02 -2.17859551e-01 6.61078915e-02 -1.73851714e-01
2.79191881e-01 1.99542359e-01 -1.54089224e+00 4.67862099e-01
1.11704493e+00 1.31612897e+00 6.59219176e-02 5.63552380e-01
1.18543953e-01 6.73672736e-01 -9.46479440e-01 -1.00929928e+00
1.07687913e-01 -3.10313255e-01 -3.46032351e-01 3.56639892e-01
5.95633686e-01 -6.46824360e-01 -3.94735128e-01 1.33462024e+00
5.58706462e-01 -9.87932533e-02 6.56537414e-01 -1.43143272e+00
-1.09853148e+00 2.26770267e-01 -3.93994689e-01 2.50601530e-01
6.88549876e-01 6.82631075e-01 1.33879852e+00 -9.76444125e-01
5.84141076e-01 1.40723431e+00 5.45617402e-01 3.87627274e-01
-9.77759123e-01 -6.72450662e-01 1.11919887e-01 7.22342193e-01
-1.22967470e+00 -3.31959069e-01 8.26359034e-01 -6.54956579e-01
4.75526214e-01 2.85611570e-01 7.18931317e-01 1.42513001e+00
2.55700052e-01 8.78006697e-01 8.98026407e-01 2.71633983e-01
3.60732079e-01 7.07573593e-02 -1.25369817e-01 3.12154353e-01
1.01485156e-01 4.45255190e-02 -4.46244270e-01 8.01093280e-02
5.58261454e-01 -1.03706874e-01 -6.30412400e-01 -2.80190408e-01
-1.10675955e+00 6.27903521e-01 4.05453086e-01 1.12915143e-01
-4.04492408e-01 -5.92458360e-02 3.87368768e-01 3.54244083e-01
2.29594424e-01 6.15112662e-01 -5.41684270e-01 -5.15908122e-01
-9.04256940e-01 3.24939132e-01 4.37795758e-01 9.66454923e-01
5.89986086e-01 1.41422451e-01 -5.57373703e-01 9.36216354e-01
3.60920638e-01 8.19134295e-01 1.77311063e-01 -1.35686588e+00
5.56834400e-01 3.97592247e-01 2.25218207e-01 -1.34800041e+00
-1.64588168e-01 -3.78647208e-01 -8.47643554e-01 1.56925783e-01
2.51322180e-01 -2.82115154e-02 -6.83803141e-01 1.70325696e+00
-8.84811506e-02 -1.65595785e-01 -1.57713100e-01 1.44968283e+00
1.05964255e+00 8.78204286e-01 9.43149179e-02 -2.97406435e-01
1.41148651e+00 -8.94138634e-01 -9.87873793e-01 -2.45417599e-02
2.51273125e-01 -8.53926718e-01 1.44875526e+00 8.30269456e-01
-1.26460838e+00 -8.73308837e-01 -1.04154408e+00 -8.11998770e-02
-1.27378181e-01 8.06553960e-02 3.02801043e-01 6.43868685e-01
-1.14444232e+00 3.90149295e-01 -3.32572728e-01 -5.34199066e-02
4.96592134e-01 -7.33519867e-02 -1.93191752e-01 -3.98658454e-01
-1.41832602e+00 7.65612960e-01 -1.08740866e-01 -5.13979383e-02
-1.18412542e+00 -8.86038065e-01 -8.35556090e-01 -8.28546137e-02
5.71000278e-01 -6.92290008e-01 1.19778788e+00 -1.27546406e+00
-1.15506923e+00 4.86712575e-01 -3.07621695e-02 5.76710403e-02
4.14063454e-01 -3.58776659e-01 -8.70493650e-01 3.04608762e-01
1.06102616e-01 5.32100856e-01 9.38793957e-01 -1.41851878e+00
-5.08661330e-01 -2.90166795e-01 5.73108554e-01 1.30687192e-01
-3.58560830e-01 2.46746391e-02 -8.99318695e-01 -1.01877499e+00
-2.28476450e-01 -6.50064766e-01 2.79832393e-01 4.25599396e-01
-2.10621469e-02 -5.34824282e-02 5.19523501e-01 -9.37348425e-01
1.52555048e+00 -2.29028988e+00 3.82156998e-01 -1.23344317e-01
3.51924628e-01 2.31409103e-01 -4.73596156e-01 1.24876417e-01
1.80287704e-01 4.19769794e-01 -1.10974638e-02 -8.71349201e-02
1.73118308e-01 1.00319721e-01 -2.63064075e-02 2.49250978e-01
4.24552143e-01 9.01960492e-01 -9.27642465e-01 -4.44442421e-01
2.13162065e-01 5.63934684e-01 -7.76288748e-01 3.40264380e-01
-1.45455927e-01 1.72226399e-01 -3.71851176e-01 1.02281761e+00
7.94274509e-01 -2.23671779e-01 -2.33024687e-01 -8.98269653e-01
1.33405358e-01 -5.40392250e-02 -1.13834989e+00 1.72592604e+00
-4.69662875e-01 8.23046803e-01 2.26963945e-02 -4.42872077e-01
6.21437311e-01 3.75975162e-01 4.41666633e-01 -1.00021374e+00
1.14385925e-01 8.01331922e-02 -1.00501820e-01 -8.96378994e-01
5.68898976e-01 4.37324122e-02 -7.78728863e-03 9.22055508e-04
3.01718622e-01 -2.28671238e-01 2.50444859e-01 2.51513183e-01
9.37484264e-01 5.89411706e-03 5.59529364e-02 6.22189902e-02
3.57242733e-01 -2.80666113e-01 6.83729529e-01 4.44181502e-01
-5.68852305e-01 9.79395032e-01 6.25978351e-01 -1.48743659e-01
-1.05290091e+00 -1.30291224e+00 -8.83218944e-02 8.51780891e-01
5.16836405e-01 -7.03104079e-01 -7.01755583e-01 -6.12550676e-01
-1.50824443e-01 6.84606433e-01 -4.99537706e-01 -2.96146601e-01
-2.41536554e-02 -3.34663689e-01 4.63303566e-01 3.22904587e-01
5.82325876e-01 -8.97670090e-01 -2.68885046e-01 1.09239640e-02
-6.53917789e-01 -1.16847694e+00 -6.76909089e-01 -5.46230972e-01
-5.45104682e-01 -1.07604945e+00 -9.16316211e-01 -1.35085702e-01
2.09982440e-01 3.68268162e-01 1.56401253e+00 9.91427619e-03
8.70390534e-02 5.41705906e-01 -7.00768530e-01 -3.98038253e-02
-1.68392614e-01 -5.59085786e-01 1.39895499e-01 8.16462040e-02
2.46019065e-01 -3.61720473e-01 -8.98759842e-01 6.39568448e-01
-1.13971388e+00 -1.91948533e-01 5.26715636e-01 5.05279779e-01
4.37090129e-01 2.73868352e-01 2.33454600e-01 -2.47091964e-01
7.60885715e-01 -6.28078997e-01 -2.68353403e-01 2.81404972e-01
-5.55069566e-01 -2.80324996e-01 4.73462731e-01 -5.67051232e-01
-9.16301012e-01 -7.86405683e-01 -2.20658585e-01 -7.21828640e-01
-1.27842128e-01 4.81625199e-01 -7.25712776e-01 2.37675801e-01
6.30738676e-01 1.69234462e-02 -3.22236925e-01 -3.02487344e-01
3.54370356e-01 5.70833564e-01 3.85172337e-01 -5.72295845e-01
7.93300331e-01 1.28005803e-01 -4.65490073e-01 -6.80158854e-01
-8.53969455e-01 -2.41533235e-01 -9.28512886e-02 -6.27491772e-01
9.57037270e-01 -1.19503486e+00 -5.56432128e-01 5.94184279e-01
-1.27392673e+00 2.83818878e-03 -9.40320939e-02 5.06213188e-01
-4.26652700e-01 4.86631185e-01 -5.78660309e-01 -6.19099498e-01
-1.63897555e-02 -1.58444655e+00 9.39160466e-01 2.03587711e-01
-1.93351388e-01 -7.41803527e-01 -2.04741150e-01 7.02197790e-01
3.85867298e-01 -1.01647370e-01 8.30193937e-01 8.90450999e-02
-5.59102118e-01 6.07476123e-02 -4.11676317e-01 7.70981550e-01
9.24165025e-02 2.48734012e-01 -8.56580436e-01 -2.17895299e-01
9.56627280e-02 -3.31670940e-01 5.74490011e-01 5.76297164e-01
1.35012615e+00 -4.52040017e-01 4.44869101e-01 7.87442684e-01
1.35029531e+00 2.84874260e-01 1.07539427e+00 2.36813143e-01
6.66247785e-01 5.14994800e-01 8.76598299e-01 6.31054342e-01
4.41705316e-01 8.20584774e-01 6.61699474e-01 3.17684337e-02
-2.32844785e-01 -8.86405706e-02 7.51644313e-01 8.31687868e-01
-2.30739325e-01 -7.71878302e-01 -6.44616187e-01 4.43017572e-01
-1.38823676e+00 -9.40042675e-01 5.54948952e-03 1.86590374e+00
7.55772293e-01 9.06990692e-02 1.41777083e-01 1.39229804e-01
5.66545486e-01 4.27425325e-01 -6.46753609e-01 -2.57936716e-01
-4.90268350e-01 -1.80314720e-01 8.69582817e-02 3.17509204e-01
-8.36608708e-01 5.45048475e-01 5.58172846e+00 1.06620800e+00
-1.04453194e+00 2.17576288e-02 7.41188824e-01 -3.29119623e-01
-7.94239104e-01 -2.83775181e-01 -3.54471415e-01 6.85137331e-01
6.25119150e-01 5.31820357e-02 6.80580258e-01 6.57863557e-01
5.73490441e-01 6.46113977e-02 -1.03307724e+00 1.50678170e+00
2.34300032e-01 -1.11948287e+00 4.67630833e-01 -3.30175571e-02
7.58491457e-01 -2.58914381e-01 4.05298889e-01 3.06958675e-01
-2.62990892e-01 -9.87589598e-01 1.10567176e+00 6.38831615e-01
9.46872354e-01 -4.67037171e-01 8.29083085e-01 -9.33402479e-02
-1.11166644e+00 -1.29034609e-01 -2.50806570e-01 3.17688212e-02
2.55856544e-01 7.55487859e-01 1.44506767e-01 5.33980310e-01
1.04343736e+00 9.34627652e-01 -8.00380886e-01 1.00187135e+00
-2.68310696e-01 6.05179548e-01 1.48949444e-01 4.70228866e-02
6.78104758e-02 -1.79120243e-01 6.08627796e-01 1.09435225e+00
6.80018425e-01 4.00579572e-01 -3.24384809e-01 8.94325078e-01
-1.05079509e-01 1.78724546e-02 -3.35066885e-01 -1.77099541e-01
3.61270636e-01 1.11635780e+00 -6.11320958e-02 -2.02324241e-01
-6.70823634e-01 1.01047599e+00 -1.94793165e-01 7.01927662e-01
-1.18890584e+00 -1.47752270e-01 1.19520426e+00 1.91854700e-01
3.34095865e-01 -8.11405554e-02 -1.37237489e-01 -1.51830971e+00
3.92627567e-01 -1.45012510e+00 2.42808506e-01 -1.32473934e+00
-1.49269688e+00 6.20599210e-01 -1.28043517e-01 -1.69706964e+00
7.97695369e-02 -6.19131505e-01 -2.61650413e-01 5.31262279e-01
-1.36784267e+00 -7.68981457e-01 -5.61262667e-01 7.58348227e-01
7.37754703e-01 -1.14183374e-01 3.66507292e-01 7.23404527e-01
-3.62496674e-01 6.36363328e-01 -1.79763541e-01 -1.88457575e-02
9.70618069e-01 -9.73730385e-01 -2.74323728e-02 7.74152994e-01
3.98541279e-02 2.96057314e-01 8.53845119e-01 -4.07064378e-01
-1.55000877e+00 -8.41999054e-01 4.92781848e-01 -5.09116650e-01
6.46052122e-01 -1.23118069e-02 -8.74500692e-01 6.22704476e-02
8.51811767e-02 -4.54013161e-02 6.06239796e-01 -1.07410945e-01
-6.39459312e-01 -3.22524905e-01 -1.05624592e+00 5.90072811e-01
1.16519797e+00 -6.98503792e-01 -3.20486903e-01 -2.19923586e-01
7.90058255e-01 -7.63985589e-02 -1.09351599e+00 4.96013910e-01
6.45997941e-01 -1.36966276e+00 1.03382897e+00 -3.02111983e-01
9.93225276e-01 -4.25088227e-01 -6.02617860e-01 -1.46165907e+00
-5.95700920e-01 -3.45446914e-01 -1.98236957e-01 1.42349768e+00
3.63456100e-01 -9.05126110e-02 1.79802194e-01 6.54358745e-01
-1.42802075e-01 -7.26442516e-01 -6.30491853e-01 -5.01439631e-01
-1.57264665e-01 -9.90832210e-01 8.14346075e-01 9.14303064e-01
-3.67199033e-01 2.29575247e-01 -7.31712401e-01 2.52920896e-01
4.56906527e-01 -1.72471344e-01 5.80297709e-01 -8.52107525e-01
-3.59907091e-01 -6.86131120e-01 -6.15076840e-01 -1.22845161e+00
-2.72544235e-01 -3.24061096e-01 -2.11680681e-01 -1.56629705e+00
2.43859410e-01 -8.64699930e-02 -3.27482343e-01 7.75114913e-03
-2.14949176e-01 3.10594797e-01 4.51699346e-01 1.39839917e-01
-9.15898740e-01 8.63764584e-01 1.67793453e+00 -4.50500906e-01
7.48229921e-02 -1.87272400e-01 -7.22312808e-01 6.33097112e-01
5.09991229e-01 -9.68209356e-02 -5.65782964e-01 -9.10332501e-01
6.63048327e-01 3.85279059e-01 4.58146214e-01 -1.11551821e+00
-1.68700367e-01 -3.30460787e-01 4.18038845e-01 -4.04193252e-01
4.38454747e-01 -9.35939491e-01 2.79019713e-01 3.90202217e-02
-2.16728494e-01 1.59943581e-01 1.61885664e-01 4.60964710e-01
-6.02999330e-01 5.12592979e-02 5.45450032e-01 3.12756710e-02
-9.12580907e-01 4.71782058e-01 -1.89689264e-01 2.74727166e-01
7.29810238e-01 -3.57361406e-01 -4.96378630e-01 -9.11844134e-01
-5.84886193e-01 2.56649792e-01 7.29853213e-01 8.16543937e-01
9.05767679e-01 -1.53560328e+00 -8.68619919e-01 -5.25532663e-02
4.35180306e-01 -4.34528977e-01 7.60591507e-01 6.05192780e-01
-3.65231484e-01 3.70563567e-02 -3.76829445e-01 -5.80762923e-01
-1.06701267e+00 6.44576788e-01 2.79647946e-01 6.36164919e-02
3.82963382e-03 6.94995105e-01 5.07775724e-01 1.39357466e-02
3.20727617e-01 -4.39860493e-01 -3.40833336e-01 3.50493222e-01
6.56317115e-01 4.44526553e-01 -9.26179439e-02 -8.26818347e-01
-3.55466038e-01 8.85042906e-01 4.01243329e-01 2.92933686e-03
1.12789309e+00 -3.98873001e-01 1.94874793e-01 4.66516435e-01
1.30193424e+00 -3.65281099e-04 -1.45840549e+00 -2.16114640e-01
-5.33501148e-01 -7.30465114e-01 1.20147258e-01 -1.08876348e+00
-1.46795404e+00 9.94146466e-01 8.28899324e-01 1.34762898e-01
1.58706057e+00 1.26429717e-03 7.76461542e-01 -1.37900993e-01
3.22917223e-01 -1.03883159e+00 5.43397427e-01 2.54828066e-01
1.30152106e+00 -1.38391984e+00 7.00735077e-02 -2.21146658e-01
-1.11526883e+00 9.02300954e-01 5.67457378e-01 3.20167720e-01
6.14849091e-01 -1.25594229e-01 2.65402764e-01 -9.29043740e-02
-7.48835385e-01 3.26528959e-02 8.73717785e-01 6.15532458e-01
3.42063546e-01 1.44863620e-01 -1.71916008e-01 9.34290290e-01
-1.85083300e-01 -6.48069456e-02 5.23828566e-01 3.59430999e-01
-9.83678326e-02 -7.17532456e-01 -2.91632682e-01 3.67476851e-01
-4.83203381e-01 -2.31629740e-02 -1.89048603e-01 5.06006539e-01
4.52731043e-01 1.34417522e+00 -3.09621803e-02 -8.15052688e-01
6.53213859e-01 -5.43838203e-01 3.53734285e-01 -1.20503403e-01
-4.66875643e-01 6.39263727e-03 7.19857514e-02 -9.46303427e-01
-4.81603265e-01 -5.57610035e-01 -7.40780115e-01 -5.54475546e-01
-1.01125792e-01 -1.48395762e-01 6.33675873e-01 7.08879054e-01
2.60682017e-01 5.16067863e-01 5.79980135e-01 -7.70376921e-01
-3.08750063e-01 -8.21904004e-01 -5.56648195e-01 9.16549444e-01
5.08626521e-01 -7.33250976e-01 -6.08359575e-01 1.26816854e-01] | [11.776507377624512, -1.7992511987686157] |
26c9d415-2186-4b7d-83f7-4e1fefb4742b | multi-modal-forest-optimization-algorithm | null | null | https://doi.org/10.1007/s00521-019-04113-z | https://link.springer.com/article/10.1007%2Fs00521-019-04113-z | Multi-Modal Forest Optimization Algorithm | Multi-modal optimization algorithms are one of the most challenging issues in the field of optimization. Most real-world problems have more than one solution; therefore, the potential role of multi-modal optimization algorithms is rather significant. Multi-modal problems consider several global and local optima. Therefore, during the search process, most of the points should be detected by the algorithm. The forest optimization algorithm has been recently introduced as a new evolutionary algorithm with the capability of solving unimodal problems. This paper presents the multi-modal forest optimization algorithm (MMFOA), which is constructed by applying a clustering technique, based on niching methods, to the unimodal forest optimization algorithm. The MMFOA operates by dividing the population of the forest into subpopulations to locate existing local and global optima. Subpopulations are generated by the Basic Sequential Algorithmic Scheme with a radius neighborhood. As population size is self-adaptive in MMFOA, population size can be increased in functions with too many local and global optima. The proposed algorithm is evaluated by a set of multi-modal benchmark functions. The experiment results show that not only is the population size low, but also that the convergence speed is high, and that the algorithm is efficient in solving multi-modal problems. | ['Taymaz Rahkar-Farshi', 'Mohammad-Reza Feizi-Derakhshi', 'Mohanna Orujpour'] | 2019-03-05 | null | null | null | null | ['metaheuristic-optimization'] | ['methodology'] | [ 2.34815236e-02 -5.12367666e-01 -1.36797741e-01 1.77233204e-01
-2.65521884e-01 -2.24784002e-01 1.98306352e-01 1.13073364e-01
-3.75245363e-01 7.95625508e-01 1.08406600e-02 1.09992675e-01
-6.25340402e-01 -1.23979676e+00 -1.45717055e-01 -1.40909290e+00
1.47396728e-01 7.95754075e-01 2.20020965e-01 -2.77676731e-01
6.65797293e-01 2.34045297e-01 -2.04681158e+00 2.40782574e-02
1.42507100e+00 8.59911263e-01 4.00432467e-01 3.10144663e-01
-3.45861226e-01 7.79790729e-02 -6.78555846e-01 1.59414753e-01
1.06227867e-01 -4.66374636e-01 -7.56660700e-01 1.03468485e-01
-5.70448518e-01 3.49057376e-01 6.01234555e-01 1.09726143e+00
7.50232399e-01 4.37045515e-01 7.72932768e-01 -1.37820363e+00
-4.49676096e-01 4.52034622e-01 -9.55139041e-01 1.22058168e-01
1.62416808e-02 5.59627563e-02 7.14943171e-01 -7.06162512e-01
2.71736681e-01 1.34622526e+00 4.41336989e-01 3.58327508e-01
-1.06686711e+00 -3.34333807e-01 -4.63804835e-03 4.85033691e-01
-1.70507717e+00 -4.76342924e-02 6.35226905e-01 -1.57311603e-01
6.20861232e-01 5.77908218e-01 8.58908892e-01 3.23924571e-02
4.48771030e-01 5.73042750e-01 1.06317890e+00 -6.26259208e-01
4.99441952e-01 2.19943896e-02 1.11954182e-01 6.18302703e-01
4.20548320e-01 8.91851559e-02 -3.38205367e-01 -3.11394423e-01
3.39544803e-01 -1.92852169e-01 -2.72445977e-01 -4.59773004e-01
-1.09472096e+00 1.10360742e+00 4.10716593e-01 4.82362568e-01
-5.26423752e-01 -2.84744382e-01 7.77123645e-02 -5.16209267e-02
8.04437548e-02 3.89597028e-01 4.29344736e-03 -7.44116157e-02
-8.80437970e-01 2.71071762e-01 4.53292787e-01 2.55780697e-01
7.54171491e-01 -6.80969805e-02 4.15842272e-02 1.01846445e+00
4.30114686e-01 5.32488644e-01 9.36346829e-01 -5.69725931e-01
2.17344955e-01 1.12958121e+00 5.05346693e-02 -1.32373047e+00
-5.00600159e-01 -3.23366106e-01 -9.70989048e-01 3.88405681e-01
2.94225484e-01 -1.50654987e-01 -7.59492934e-01 1.32445407e+00
6.17691100e-01 -3.90922219e-01 1.84583634e-01 8.76206577e-01
6.24415219e-01 1.02123225e+00 -1.99518442e-01 -8.07210565e-01
9.99538898e-01 -1.16809595e+00 -7.37135291e-01 -2.51709316e-02
7.93884695e-02 -7.56011248e-01 9.62465048e-01 2.42051855e-01
-8.23961139e-01 -3.70441377e-01 -9.21179056e-01 7.89037824e-01
-5.22792876e-01 2.95257550e-02 4.75712419e-01 8.43487680e-01
-7.60210454e-01 2.51969635e-01 -5.15933335e-01 -4.63034242e-01
2.64637619e-02 4.90562409e-01 1.56108424e-01 -1.39080090e-02
-8.98373306e-01 6.61606967e-01 9.16441917e-01 3.85522187e-01
-3.78996372e-01 -2.69234069e-02 -3.89566511e-01 3.50221880e-02
2.97938466e-01 -7.16760695e-01 4.01068538e-01 -1.08216703e+00
-1.45376587e+00 2.42446780e-01 -5.66415250e-01 1.46950096e-01
2.60158658e-01 4.01923418e-01 -4.10692453e-01 -2.04092320e-02
2.45242730e-01 3.14065903e-01 7.10344374e-01 -1.42491758e+00
-8.39402318e-01 -4.44638342e-01 -4.32228774e-01 6.22611523e-01
-3.94819558e-01 1.18076921e-01 1.54881952e-02 -4.46202457e-01
3.42106313e-01 -7.40880787e-01 -4.11453962e-01 -7.15236723e-01
-2.46816903e-01 -2.86136389e-01 9.54481065e-01 -2.11910069e-01
1.62720692e+00 -1.99814367e+00 6.09956205e-01 8.29355896e-01
-2.44248301e-01 -4.41031270e-02 8.24034587e-02 3.41031462e-01
1.53441861e-01 1.63223416e-01 -5.34356952e-01 4.17314470e-01
-2.49313235e-01 3.07695985e-01 3.84252936e-01 2.83513516e-01
-2.55542725e-01 5.19685566e-01 -7.10565984e-01 -8.37257862e-01
-1.12693915e-02 6.08370034e-03 -4.71326590e-01 -7.41110593e-02
-1.46851838e-01 2.43074715e-01 -5.43646574e-01 1.11271107e+00
8.29738498e-01 -5.34416288e-02 -3.39912847e-02 -6.86747804e-02
-5.44712603e-01 -6.96325660e-01 -1.73515999e+00 1.07147992e+00
-9.30330455e-02 -4.18277318e-03 2.36892521e-01 -9.81449544e-01
9.79246259e-01 1.10044234e-01 7.28287876e-01 -4.03229296e-01
4.52520996e-01 4.09522176e-01 5.32425120e-02 -5.62439382e-01
6.40824080e-01 -2.69969463e-01 1.77751601e-01 6.59731805e-01
-3.75383168e-01 -3.40090110e-03 6.00356996e-01 -3.61763895e-01
4.32143331e-01 -3.43752414e-01 2.22416103e-01 -5.36729097e-01
8.96307647e-01 4.75291163e-01 8.35910022e-01 2.63468713e-01
-1.14885651e-01 3.48464400e-01 -1.33170545e-01 -3.15958202e-01
-5.77828765e-01 -8.20286453e-01 -2.46564597e-01 9.89537537e-01
7.00688541e-01 -1.92443803e-01 -7.08940387e-01 -2.31054798e-01
-9.47830528e-02 5.59210181e-01 -4.83403653e-01 -2.03307733e-01
-5.16127586e-01 -1.80261827e+00 1.28064901e-01 3.24909128e-02
9.33014870e-01 -1.24123883e+00 -7.14394629e-01 2.28725329e-01
-5.07573664e-01 -6.83833137e-02 -2.52076447e-01 -1.97973624e-02
-1.09860444e+00 -9.39952672e-01 -8.84271026e-01 -1.09729803e+00
9.18441534e-01 4.23291653e-01 6.67670906e-01 3.38841081e-01
-8.84148777e-02 -2.85890652e-03 -4.44013029e-01 -2.68859506e-01
-2.51500815e-01 1.80601910e-01 -5.23445234e-02 2.97282338e-01
3.19273621e-01 -5.13326883e-01 -3.13811630e-01 7.83806086e-01
-6.77029014e-01 -1.41327351e-01 7.10048497e-01 1.15419590e+00
8.97271037e-01 1.21286607e+00 6.93190336e-01 -2.11462855e-01
9.74474549e-01 -3.50785166e-01 -4.83070076e-01 4.59959239e-01
-8.09488118e-01 8.49621370e-02 5.95065117e-01 -4.22387242e-01
-1.09771860e+00 2.44831964e-02 1.41610861e-01 -7.96678439e-02
-1.91566739e-02 7.81250358e-01 -5.23380876e-01 -3.85548353e-01
4.12792206e-01 6.20675266e-01 2.90016502e-01 -2.35910729e-01
2.31544003e-02 8.40650380e-01 2.06020728e-01 -5.87524176e-01
7.26492882e-01 2.09194764e-01 1.85915247e-01 -8.86582732e-01
2.45492123e-02 -4.61515874e-01 -3.80168259e-01 -7.72888184e-01
8.20722818e-01 -1.17239609e-01 -8.72343719e-01 8.32852244e-01
-6.86750889e-01 1.37510270e-01 -7.61755705e-02 3.20301533e-01
-2.95149058e-01 3.27598453e-01 3.81996296e-02 -1.05921972e+00
-5.06558299e-01 -1.24428403e+00 5.97928941e-01 8.15722108e-01
4.90609743e-02 -8.09430182e-01 1.31272510e-01 4.46095735e-01
4.46610659e-01 3.39988530e-01 1.24074924e+00 -8.82032216e-02
-4.24405962e-01 3.01092733e-02 3.01068038e-01 -2.50306964e-01
2.29497075e-01 3.33946347e-01 -3.02312642e-01 -3.11516374e-01
4.47194874e-02 3.85315083e-02 5.23493648e-01 7.21703410e-01
8.72547030e-01 -2.46050388e-01 -6.57633960e-01 4.97584432e-01
1.66600180e+00 7.42273092e-01 4.71412033e-01 1.06760633e+00
3.19417924e-01 6.36074483e-01 1.11283684e+00 4.02596742e-01
3.17377120e-01 3.83763462e-01 6.99412465e-01 9.98971388e-02
4.54191417e-01 1.02652699e-01 1.79878205e-01 8.59369755e-01
-5.06942630e-01 -2.87703067e-01 -8.46569240e-01 6.36731684e-01
-2.01813531e+00 -1.19921470e+00 -2.68165410e-01 2.16950631e+00
5.21350622e-01 -1.93588257e-01 4.94748741e-01 5.07357240e-01
1.33548391e+00 -5.15076667e-02 -5.79906881e-01 -5.51770389e-01
-3.86041373e-01 -1.62231058e-01 4.17155802e-01 3.88100713e-01
-9.23730791e-01 5.65853417e-01 5.99302149e+00 9.65411544e-01
-1.16740751e+00 -2.47881830e-01 5.02991080e-01 -1.28517032e-01
-3.02099824e-01 -9.52682495e-02 -6.45554423e-01 7.48083770e-01
4.03769433e-01 -5.05039632e-01 8.36113513e-01 6.39077127e-01
3.33828151e-01 -6.32012844e-01 -1.30261615e-01 8.67740214e-01
-3.17097977e-02 -1.01819599e+00 1.69704244e-01 1.35624101e-02
9.24511015e-01 -5.95944464e-01 7.99101740e-02 -2.52961904e-01
-9.04576667e-03 -1.10455692e+00 5.99588573e-01 3.00938845e-01
2.18391299e-01 -1.36755705e+00 9.47048187e-01 6.38389826e-01
-1.54733789e+00 -5.17132401e-01 -3.70629132e-01 1.79247677e-01
4.19583410e-01 5.69290757e-01 -1.40812889e-01 9.34720099e-01
1.06117833e+00 2.98456043e-01 -6.23192489e-01 1.52759588e+00
2.06886679e-01 2.30896637e-01 -6.09103560e-01 -5.91753542e-01
1.97487235e-01 -8.71033728e-01 8.13739955e-01 4.68994707e-01
5.68992257e-01 -4.14809883e-02 1.24502510e-01 6.73376203e-01
5.93412757e-01 6.81862175e-01 -2.09333837e-01 4.66753878e-02
6.12825632e-01 1.08935499e+00 -1.01646411e+00 4.05277461e-02
1.08261727e-01 4.92265165e-01 -1.80282831e-01 1.37307331e-01
-8.14762890e-01 -6.29506648e-01 5.28512411e-02 -1.50974244e-01
3.93573530e-02 -4.35535200e-02 -6.15875244e-01 -6.48201883e-01
-1.50159955e-01 -9.13588047e-01 7.71350443e-01 -5.54467857e-01
-9.33227777e-01 5.44033587e-01 -1.02453962e-01 -9.44187105e-01
-7.50899091e-02 -3.76705170e-01 -9.60257590e-01 8.68080318e-01
-1.06395543e+00 -9.46670830e-01 -5.33006310e-01 7.80080020e-01
5.01291156e-01 -4.02502567e-01 6.47790730e-01 2.22459719e-01
-8.29020977e-01 3.45124274e-01 5.21637201e-01 -5.07915139e-01
3.63103837e-01 -9.78996456e-01 -7.29027987e-01 8.82888496e-01
-4.33357954e-01 4.81614172e-01 7.33357251e-01 -6.60843909e-01
-1.25770748e+00 -7.43895411e-01 7.02342510e-01 2.73516506e-01
1.74594983e-01 2.20480129e-01 -6.33690894e-01 5.31317182e-02
2.59862214e-01 -7.03145027e-01 6.81597054e-01 2.46230457e-02
8.54652882e-01 -8.52024108e-02 -1.46474707e+00 4.98413801e-01
5.86657345e-01 3.51270765e-01 -3.78825426e-01 2.31087372e-01
4.25469726e-02 -1.57836735e-01 -8.04951906e-01 5.33795655e-01
5.25670111e-01 -9.15977180e-01 9.39537585e-01 4.94010709e-02
1.54183924e-01 -8.77643943e-01 5.92464861e-03 -1.54653788e+00
-5.21704972e-01 -2.54063517e-01 1.46513239e-01 1.41388547e+00
3.74594122e-01 -9.34543192e-01 7.58828640e-01 2.11865664e-01
1.67949319e-01 -8.50823104e-01 -9.77435648e-01 -8.39783430e-01
1.51931345e-01 4.49103594e-01 8.99878144e-01 7.33923793e-01
-8.35809559e-02 3.17057908e-01 -1.20962493e-01 1.14542149e-01
6.43661916e-01 7.46406674e-01 4.73555714e-01 -1.37139630e+00
-1.24667741e-01 -7.88806498e-01 -4.22508009e-02 -4.86391634e-01
-1.36324689e-01 -6.24801815e-01 7.20405132e-02 -1.60503078e+00
2.98557490e-01 -6.18626654e-01 -6.47650510e-02 2.30046019e-01
-3.47321987e-01 -1.27136067e-01 5.13040982e-02 4.76182729e-01
-1.74594611e-01 8.16183627e-01 1.27963495e+00 -2.95637488e-01
-7.39214361e-01 3.62106562e-01 -4.44352657e-01 5.35236776e-01
9.42938566e-01 -3.84833574e-01 -2.70779490e-01 -3.44899036e-02
-7.20910728e-02 -1.56808108e-01 -2.35865012e-01 -1.05102229e+00
3.47390652e-01 -6.28914714e-01 1.19480707e-01 -7.74995148e-01
1.47230729e-01 -9.03965890e-01 4.99689847e-01 7.85467863e-01
2.56846249e-01 3.94216090e-01 8.02325159e-02 4.27307874e-01
-4.78467554e-01 -5.68841219e-01 9.67414618e-01 -4.84751612e-02
-8.36402535e-01 -1.73604205e-01 -7.94471264e-01 -2.09938005e-01
1.47905850e+00 -7.92115808e-01 -1.94518045e-01 -1.23965368e-01
-6.15891159e-01 6.42601490e-01 6.00921094e-01 1.08810455e-01
5.37008405e-01 -1.36170602e+00 -4.05792773e-01 7.02716112e-02
-1.97662786e-01 1.50093421e-01 2.20021904e-01 9.29047763e-01
-7.81453729e-01 1.43010095e-01 -5.24107933e-01 -6.27874672e-01
-1.65031576e+00 5.86803675e-01 5.23092210e-01 -1.17721617e-01
1.96292829e-02 6.30917490e-01 -3.61297667e-01 -5.59456229e-01
2.71791723e-02 3.08274090e-01 -7.38058269e-01 3.25819671e-01
2.48068005e-01 9.37929094e-01 -2.85295814e-01 -9.37123120e-01
-5.97370267e-01 1.10358226e+00 4.67381626e-01 -1.52207136e-01
1.42858422e+00 -1.32570386e-01 -7.78203011e-01 1.66394681e-01
6.68947875e-01 1.08161807e-01 -2.91991115e-01 3.94593507e-01
-3.24582368e-01 -5.83690584e-01 1.38845846e-01 -8.86071205e-01
-1.15635633e+00 4.59028959e-01 7.36641467e-01 4.40350622e-01
1.63665199e+00 -4.21800941e-01 6.16100609e-01 2.85925865e-02
4.54903334e-01 -1.45816624e+00 1.87591906e-03 3.20892036e-01
6.79535925e-01 -9.58281755e-01 -4.21143249e-02 -4.15792197e-01
-4.50282931e-01 1.21283054e+00 7.71497548e-01 2.44626969e-01
5.42395592e-01 3.67660597e-02 -8.27952772e-02 -7.96402991e-02
-3.77389103e-01 -1.78426445e-01 -6.00305460e-02 4.80408698e-01
-1.20244294e-01 3.34409773e-02 -1.02477181e+00 5.40386796e-01
-2.09489495e-01 -2.30904222e-01 2.11035296e-01 1.12871718e+00
-1.00858188e+00 -1.14399672e+00 -1.19371629e+00 3.79813224e-01
-1.30490199e-01 1.38265163e-01 -4.10553157e-01 5.65175593e-01
3.78846914e-01 1.38178658e+00 -7.03516081e-02 -4.28297102e-01
5.36988042e-02 -1.79614514e-01 2.08091095e-01 5.27863130e-02
-5.93995690e-01 4.96573374e-02 -1.97396338e-01 -1.50466710e-01
-7.15607166e-01 -7.43599057e-01 -1.42418635e+00 -5.76400638e-01
-7.67918289e-01 8.42487276e-01 5.06771624e-01 7.86136568e-01
1.24688342e-01 4.56030458e-01 7.42424905e-01 -7.24015653e-01
-1.16095364e-01 -5.40969193e-01 -6.19167089e-01 -1.44398466e-01
-2.86543399e-01 -1.02661502e+00 -3.85591209e-01 -2.90988684e-01] | [5.754266262054443, 3.529407501220703] |
80df35ee-a3be-4892-95d0-5479c8f5e542 | guaranteed-bounds-for-posterior-inference-in | 2204.02948 | null | https://arxiv.org/abs/2204.02948v2 | https://arxiv.org/pdf/2204.02948v2.pdf | Guaranteed Bounds for Posterior Inference in Universal Probabilistic Programming | We propose a new method to approximate the posterior distribution of probabilistic programs by means of computing guaranteed bounds. The starting point of our work is an interval-based trace semantics for a recursive, higher-order probabilistic programming language with continuous distributions. Taking the form of (super-/subadditive) measures, these lower/upper bounds are non-stochastic and provably correct: using the semantics, we prove that the actual posterior of a given program is sandwiched between the lower and upper bounds (soundness); moreover the bounds converge to the posterior (completeness). As a practical and sound approximation, we introduce a weight-aware interval type system, which automatically infers interval bounds on not just the return value but also weight of program executions, simultaneously. We have built a tool implementation, called GuBPI, which automatically computes these posterior lower/upper bounds. Our evaluation on examples from the literature shows that the bounds are useful, and can even be used to recognise wrong outputs from stochastic posterior inference procedures. | ['Fabian Zaiser', 'Luke Ong', 'Raven Beutner'] | 2022-04-06 | null | null | null | null | ['probabilistic-programming'] | ['methodology'] | [ 1.25115499e-01 2.48056531e-01 -2.06956297e-01 -3.62844259e-01
-1.00531018e+00 -8.06263089e-01 5.38227379e-01 3.91500086e-01
-8.54298323e-02 7.30470598e-01 -1.27896473e-01 -7.16850698e-01
-2.40626335e-01 -1.07038379e+00 -9.28858042e-01 -5.45927465e-01
-3.14073354e-01 9.48991060e-01 8.81981850e-01 2.73274988e-01
3.89504552e-01 3.60900432e-01 -1.91175807e+00 2.21819207e-01
6.33666277e-01 9.29310501e-01 -3.49991053e-01 1.18197429e+00
-2.16516525e-01 5.70016861e-01 -6.62681758e-01 -3.65206689e-01
-2.36322373e-01 -2.66577929e-01 -7.03559816e-01 -5.86489618e-01
-6.52060285e-02 -5.42707503e-01 3.55484873e-01 1.50768983e+00
-2.23490760e-01 -1.94342867e-01 9.15238857e-01 -1.66837907e+00
-1.72374636e-01 1.33904409e+00 -5.12348771e-01 -1.59846276e-01
6.57187581e-01 -1.96116045e-01 1.06267941e+00 -1.49074942e-01
3.10885489e-01 1.21969259e+00 9.38913107e-01 2.39645869e-01
-1.83402753e+00 -1.12900302e-01 -1.95839003e-01 -3.74752283e-01
-1.55229080e+00 -4.11759615e-02 2.78216332e-01 -6.52064621e-01
8.31590593e-01 6.37379289e-01 2.65190184e-01 8.85073483e-01
3.14085811e-01 7.59917021e-01 1.17352366e+00 -7.22922146e-01
6.15439117e-01 4.58194017e-01 5.51392436e-01 6.90223277e-01
5.70544302e-01 4.96828109e-01 -1.08440250e-01 -1.10261679e+00
4.57429796e-01 7.41328523e-02 -1.43959224e-01 -2.76379019e-01
-1.02413702e+00 6.31102920e-01 -6.94842219e-01 1.58809602e-01
-2.22576819e-02 6.15660906e-01 5.55555284e-01 -3.08931922e-03
2.60380954e-01 -2.10508302e-01 -4.64487553e-01 -5.76697886e-01
-1.13839900e+00 5.55687606e-01 1.59961474e+00 1.21896362e+00
6.29042566e-01 -4.72592384e-01 -5.12943625e-01 2.52058268e-01
7.02874541e-01 7.62931108e-01 -1.73775762e-01 -9.77266014e-01
3.58788297e-02 2.16613919e-01 6.73456907e-01 -8.29714656e-01
1.36803076e-01 2.08474949e-01 -1.01153843e-01 5.01600027e-01
7.69417644e-01 1.30931526e-01 -4.51447129e-01 1.97392356e+00
1.18533939e-01 -3.10271662e-02 -2.35572666e-01 2.60155976e-01
-2.71877497e-01 7.68507421e-01 1.68411016e-01 -3.54666740e-01
1.35300589e+00 2.17889436e-02 -5.59951544e-01 3.70675087e-01
6.20571256e-01 -5.34076095e-01 1.05606377e+00 6.37329519e-01
-9.78125453e-01 -9.49879643e-03 -1.28553855e+00 3.67975384e-01
-1.35483995e-01 -2.49760766e-02 6.22501671e-01 1.25817192e+00
-9.32413518e-01 7.83874452e-01 -1.29200292e+00 2.55641162e-01
-7.59226233e-02 1.15647860e-01 1.27102286e-01 3.20347220e-01
-7.71077096e-01 6.53543711e-01 7.53992081e-01 -7.45996311e-02
-9.60980952e-01 -7.17927873e-01 -6.58232450e-01 1.87654182e-01
3.07410628e-01 -1.89281896e-01 1.59251356e+00 -5.26990950e-01
-1.63895023e+00 8.63713086e-01 -1.23293847e-01 -5.55008650e-01
6.89001679e-01 6.76754257e-03 -1.94166124e-01 -2.35606000e-01
-6.45910725e-02 -2.77965426e-01 5.78215837e-01 -1.18491316e+00
-7.87993431e-01 -3.73223275e-01 2.54500300e-01 -6.29732430e-01
8.27119425e-02 2.01740935e-01 3.07687442e-03 2.36928859e-03
-9.53281894e-02 -8.59034479e-01 -1.46044731e-01 -3.14317346e-01
-6.92490935e-01 -4.65702176e-01 2.21345931e-01 -3.79835635e-01
1.57957673e+00 -2.11602354e+00 -1.23233825e-01 6.35112703e-01
3.71402642e-03 -3.63502353e-01 7.28318155e-01 3.99132997e-01
1.33797556e-01 3.22780579e-01 -6.98473096e-01 -1.12235304e-02
8.53232265e-01 4.15190905e-01 -8.28281224e-01 8.03889334e-01
-5.60198203e-02 3.06526452e-01 -8.97178233e-01 -6.06487989e-01
1.06643289e-01 -3.70693989e-02 -5.92744052e-01 3.49289954e-01
-8.73706818e-01 -4.63755608e-01 -3.24815661e-01 2.79237688e-01
8.50349486e-01 1.08209275e-01 4.16750312e-01 2.89202631e-01
-2.03366429e-01 2.68352151e-01 -1.59568441e+00 1.29972410e+00
-6.13529801e-01 1.44660488e-01 -1.39627308e-01 -4.75326747e-01
7.52345920e-01 2.07250178e-01 1.91531442e-02 2.01521635e-01
1.97643787e-01 2.79967010e-01 -4.60601211e-01 -1.74196452e-01
6.73397005e-01 -1.54350653e-01 -5.73382318e-01 9.37981725e-01
-6.43724650e-02 -2.82538742e-01 3.66770089e-01 9.37172584e-03
9.47145581e-01 6.59292459e-01 2.99211383e-01 -6.78412616e-01
6.52474999e-01 5.71754016e-03 4.73063469e-01 9.66501594e-01
9.59075466e-02 3.80009353e-01 1.54396343e+00 -6.89959005e-02
-9.77859735e-01 -1.51620531e+00 -2.02261060e-01 1.33350158e+00
-3.14434290e-01 -5.24360716e-01 -1.05250692e+00 -5.28769076e-01
6.00637384e-02 1.13561296e+00 -6.38780415e-01 -1.21795692e-01
-1.25934392e-01 -6.56478941e-01 8.13749731e-01 7.35752821e-01
-2.86118209e-01 -4.01394099e-01 -8.60800922e-01 2.58438826e-01
1.32344946e-01 -6.31381750e-01 -7.06855595e-01 3.27968836e-01
-7.69608498e-01 -1.10178339e+00 -3.00448239e-01 -8.86404663e-02
4.82999593e-01 -7.16012061e-01 1.27534151e+00 -1.56774715e-01
4.41135578e-02 4.19888407e-01 1.32159829e-01 -5.48707187e-01
-1.00457406e+00 -4.30503726e-01 -2.22073048e-01 -1.96404010e-01
3.17641616e-01 -5.87058008e-01 6.77720979e-02 2.48708755e-01
-1.02299583e+00 -4.86039132e-01 1.05730124e-01 6.81918085e-01
7.45367885e-01 3.06446314e-01 -2.97291968e-02 -1.27033210e+00
8.27434421e-01 -3.78968775e-01 -1.61453331e+00 4.18412000e-01
-5.40139854e-01 7.49420404e-01 7.20275939e-01 -4.49736029e-01
-1.08010638e+00 9.53400135e-02 -1.10954553e-01 -1.67991191e-01
-2.29367495e-01 5.39672911e-01 -2.48949334e-01 4.46996182e-01
7.39932120e-01 1.96847856e-01 -5.63124977e-02 -9.06561166e-02
3.35465372e-01 6.23715103e-01 8.56300473e-01 -1.54807663e+00
2.88157851e-01 4.96171981e-01 2.29165763e-01 -2.59260952e-01
-5.49463212e-01 -2.95987874e-01 -1.30588636e-01 1.63800009e-02
1.63414985e-01 -2.07082957e-01 -1.46786940e+00 9.97311845e-02
-1.25901175e+00 -3.65975797e-01 -2.83429921e-01 3.12453419e-01
-9.21193242e-01 5.51936269e-01 -5.74121892e-01 -1.71897423e+00
-2.54682396e-02 -1.24919260e+00 1.15123153e+00 -3.36082280e-02
-4.71993148e-01 -1.00131047e+00 4.17960554e-01 -5.79212606e-01
1.54743344e-01 5.11406541e-01 1.18740952e+00 -9.52943981e-01
-3.36572856e-01 -5.86725175e-01 -1.68550253e-01 4.74681616e-01
-7.39312530e-01 9.08040643e-01 -8.71288836e-01 3.83810580e-01
5.81094325e-02 2.49011740e-01 3.75469416e-01 2.92656749e-01
1.08792281e+00 -5.71427703e-01 -4.04812783e-01 5.66393286e-02
1.64687753e+00 -2.15703487e-01 6.70396924e-01 -1.90889791e-01
-2.98138894e-03 3.57952744e-01 5.00075698e-01 7.45175004e-01
1.65446475e-01 3.82869840e-01 2.39085227e-01 1.02849960e+00
3.86634409e-01 -3.91307056e-01 6.73871577e-01 2.54799902e-01
-1.17478974e-01 2.19894841e-01 -1.01528883e+00 6.80960298e-01
-1.78501093e+00 -1.21666169e+00 -4.54579055e-01 2.92180300e+00
1.33807874e+00 5.36842942e-01 5.10838330e-01 2.88531482e-01
8.25266778e-01 -3.91202837e-01 -7.70952180e-02 -9.52284575e-01
5.68696022e-01 3.78593922e-01 7.53032088e-01 7.78970242e-01
-7.09455550e-01 7.96950981e-02 7.02368736e+00 9.46536601e-01
-5.23062646e-01 1.30177468e-01 3.65134627e-01 5.60785115e-01
-7.27824986e-01 3.11348379e-01 -8.45734835e-01 6.78130388e-01
1.75837278e+00 -6.16839468e-01 3.54622781e-01 1.54403794e+00
-4.92150068e-01 -5.69786429e-01 -1.75244880e+00 5.02106965e-01
-4.60741937e-01 -1.04181612e+00 -6.28580511e-01 6.54703677e-02
4.61488008e-01 -3.59723002e-01 -1.47894487e-01 4.29902613e-01
8.44902098e-01 -8.26963544e-01 1.31260872e+00 7.31943190e-01
7.92615473e-01 -1.24298108e+00 1.05513310e+00 5.70432961e-01
-9.65215147e-01 1.83298677e-01 -2.09447056e-01 1.17527908e-02
-1.94002360e-01 1.13127708e+00 -8.54189634e-01 2.98248768e-01
2.85824239e-01 -7.10804909e-02 6.73434734e-02 8.98737371e-01
-4.18058306e-01 7.11872995e-01 -9.75320756e-01 -3.75703365e-01
-2.98187464e-01 -1.64433822e-01 3.71415198e-01 1.52778065e+00
2.52005935e-01 -3.82803649e-01 1.52910128e-01 1.46028757e+00
2.69595921e-01 -4.94657069e-01 -4.91935253e-01 -7.34879747e-02
5.10013044e-01 8.64526272e-01 -8.12192738e-01 -5.38969874e-01
-5.07238731e-02 3.02844524e-01 -2.59226505e-02 1.72788963e-01
-1.17503309e+00 -7.24967837e-01 4.48588610e-01 -4.44155127e-01
1.66237131e-01 -9.46336165e-02 -7.23742962e-01 -7.00844824e-01
2.10838214e-01 -5.81085145e-01 5.80718517e-01 -1.58056617e-01
-1.12071931e+00 3.43930423e-01 5.71282089e-01 -7.06520200e-01
-7.60557771e-01 -7.61472523e-01 -5.56630254e-01 1.11354744e+00
-7.85661697e-01 -4.24380034e-01 3.48653197e-01 1.10029489e-01
-4.66940820e-01 6.47803187e-01 1.21444106e+00 -2.56501466e-01
-3.00839692e-01 6.46585464e-01 -2.27438528e-02 -2.25173727e-01
9.84973609e-02 -1.68288267e+00 -6.98001832e-02 9.86726761e-01
-1.65438652e-01 7.49888897e-01 1.50080895e+00 -5.50718069e-01
-1.71191943e+00 -7.41306186e-01 6.92315817e-01 -7.94863760e-01
1.05564177e+00 -3.42004329e-01 -8.47261429e-01 9.06025648e-01
-1.71869978e-01 -2.43666489e-02 5.61843514e-01 4.28317338e-01
-9.24653590e-01 -7.72118047e-02 -1.43371212e+00 4.01796371e-01
4.19212759e-01 -8.09784532e-01 -8.22389603e-01 2.65090048e-01
5.21622717e-01 -5.21924376e-01 -1.24482620e+00 1.06305674e-01
9.30832863e-01 -1.33130658e+00 6.57559454e-01 -1.91424146e-01
2.09622487e-01 -6.57256961e-01 -5.53217053e-01 -6.83069825e-01
3.53210241e-01 -9.02148724e-01 -5.27707577e-01 1.24866569e+00
5.85439861e-01 -9.54295695e-01 3.81178766e-01 1.02594841e+00
9.56970230e-02 -5.06239831e-01 -9.63434160e-01 -9.25745547e-01
-1.52080879e-01 -1.19516218e+00 8.35092306e-01 2.74684340e-01
5.41728318e-01 -4.13362712e-01 7.64772967e-02 5.78676343e-01
8.56953144e-01 4.79324043e-01 4.50637013e-01 -1.07168376e+00
-8.92192721e-01 -7.12073803e-01 -4.76316303e-01 -4.37465042e-01
2.84086972e-01 -5.09184301e-01 6.67968512e-01 -7.73504555e-01
3.01711887e-01 -3.99728239e-01 -6.34397268e-02 4.09900784e-01
2.31857389e-01 -2.61431545e-01 -5.17111063e-01 -2.51982272e-01
-5.44707477e-01 8.73809606e-02 2.50908703e-01 1.89358428e-01
-5.44818677e-03 5.18601179e-01 -3.75788122e-01 9.23871160e-01
4.38061595e-01 -8.47995520e-01 -1.78002208e-01 2.51513362e-01
8.43217373e-01 5.73529899e-01 4.00622934e-01 -8.46604228e-01
2.17763335e-01 -2.34755129e-01 -2.08266959e-01 -7.82791972e-01
-3.70132148e-01 -8.02800894e-01 5.44407487e-01 5.18881619e-01
-5.80302000e-01 -1.64780259e-01 1.95766687e-01 8.06520164e-01
8.16527605e-02 -8.16653967e-01 5.08434832e-01 8.82156938e-02
-1.50408611e-01 -8.12229216e-02 -3.51960182e-01 1.51030729e-02
9.04284656e-01 2.81260461e-01 -2.68911589e-02 1.58576488e-01
-6.03349447e-01 -1.22647129e-01 6.11932099e-01 -4.31075245e-01
2.61110216e-01 -9.56610620e-01 -5.17858505e-01 2.32601445e-02
2.67480284e-01 -2.03730777e-01 -6.49573728e-02 8.36663544e-01
-6.20927811e-01 2.53150046e-01 1.59594014e-01 -8.93039286e-01
-8.37026954e-01 7.43102491e-01 3.47287096e-02 -3.22427958e-01
-2.46747941e-01 4.79890406e-01 -2.60251552e-01 -4.48397875e-01
3.71338814e-01 -8.83503675e-01 5.45748949e-01 -3.75947416e-01
9.94197488e-01 5.36969125e-01 -4.07809741e-04 2.93755680e-01
-4.40462947e-01 -7.30154440e-02 1.40132904e-01 -6.35851622e-01
1.05718434e+00 1.65568754e-01 -7.35213578e-01 9.82312620e-01
9.57398772e-01 2.37225235e-01 -1.01345718e+00 4.96857911e-02
5.12621641e-01 -3.64496589e-01 -2.40291283e-01 -6.25138938e-01
-2.06598490e-01 8.02903473e-01 1.67442828e-01 7.95830250e-01
9.00842130e-01 1.06152214e-01 2.09584862e-01 2.49669448e-01
8.24794412e-01 -9.13943946e-01 -9.87766027e-01 4.40872043e-01
5.74851632e-01 -4.25684571e-01 -3.98644879e-02 -2.47200891e-01
-1.46844730e-01 1.31420100e+00 -4.24147211e-02 -2.51588136e-01
5.29063404e-01 9.86429214e-01 -8.32053900e-01 3.60154539e-01
-9.02824044e-01 7.38052055e-02 -7.25681335e-02 6.73571825e-01
3.82649124e-01 5.88018656e-01 -2.97304392e-01 9.87354755e-01
-2.00832546e-01 2.18959644e-01 8.81117821e-01 9.70768154e-01
-5.10247648e-01 -8.96746874e-01 -7.12655544e-01 3.54538769e-01
-7.01823771e-01 -1.51084112e-02 3.43456239e-01 7.13098586e-01
-4.30421263e-01 7.24316120e-01 -1.58913106e-01 -1.73129424e-01
8.68932754e-02 2.20872700e-01 7.78241515e-01 -6.00877881e-01
2.63604280e-02 -2.29778081e-01 2.88327873e-01 -6.29338562e-01
4.03356031e-02 -5.48265457e-01 -1.18336737e+00 -5.82557738e-01
-3.22503328e-01 4.67629254e-01 8.26378047e-01 8.92396927e-01
1.90821551e-02 2.70213664e-01 4.43737566e-01 -5.18009007e-01
-1.31909966e+00 -7.89111197e-01 -8.48639607e-01 -1.87996671e-01
8.68899673e-02 -3.58365625e-01 -6.85595095e-01 2.15251222e-01] | [8.379842758178711, 6.444100856781006] |
45619806-acc6-492c-84b2-86680f763763 | group-emotion-recognition-using-machine | 1905.01118 | null | https://arxiv.org/abs/1905.01118v1 | https://arxiv.org/pdf/1905.01118v1.pdf | Group Emotion Recognition Using Machine Learning | Automatic facial emotion recognition is a challenging task that has gained significant scientific interest over the past few years, but the problem of emotion recognition for a group of people has been less extensively studied. However, it is slowly gaining popularity due to the massive amount of data available on social networking sites containing images of groups of people participating in various social events. Group emotion recognition is a challenging problem due to obstructions like head and body pose variations, occlusions, variable lighting conditions, variance of actors, varied indoor and outdoor settings and image quality. The objective of this task is to classify a group's perceived emotion as Positive, Neutral or Negative. In this report, we describe our solution which is a hybrid machine learning system that incorporates deep neural networks and Bayesian classifiers. Deep Convolutional Neural Networks (CNNs) work from bottom to top, analysing facial expressions expressed by individual faces extracted from the image. The Bayesian network works from top to bottom, inferring the global emotion for the image, by integrating the visual features of the contents of the image obtained through a scene descriptor. In the final pipeline, the group emotion category predicted by an ensemble of CNNs in the bottom-up module is passed as input to the Bayesian Network in the top-down module and an overall prediction for the image is obtained. Experimental results show that the stated system achieves 65.27% accuracy on the validation set which is in line with state-of-the-art results. As an outcome of this project, a Progressive Web Application and an accompanying Android app with a simple and intuitive user interface are presented, allowing users to test out the system with their own pictures. | ['Samanyou Garg'] | 2019-05-03 | null | null | null | null | ['facial-emotion-recognition'] | ['computer-vision'] | [-9.94410664e-02 7.95156881e-02 1.94223374e-01 -9.79479432e-01
-1.07868031e-01 -1.69116899e-01 5.10363042e-01 2.08720993e-02
-4.60040778e-01 3.57211083e-01 1.57994125e-02 4.37544465e-01
5.37722185e-02 -5.15890360e-01 -3.08208942e-01 -6.78401649e-01
-1.62310690e-01 1.93865746e-01 -2.81707831e-02 -1.01978965e-01
4.94040400e-02 7.27942705e-01 -2.09248781e+00 6.13954663e-01
8.05163160e-02 1.60511351e+00 -1.67622522e-01 5.57475209e-01
-1.44302230e-02 6.47438228e-01 -5.44873893e-01 -3.32351089e-01
-4.56103384e-02 -2.23792151e-01 -3.75662029e-01 2.11293429e-01
4.17293131e-01 -2.32140198e-01 2.56879658e-01 8.67374599e-01
7.27823973e-01 1.08756170e-01 5.01954556e-01 -1.44433534e+00
1.16535395e-01 -5.76410741e-02 -2.44579822e-01 -2.06799507e-01
6.40280545e-01 4.76965569e-02 5.81008911e-01 -9.36632097e-01
5.98788857e-01 1.43945551e+00 5.49185693e-01 5.24981976e-01
-1.20876253e+00 -8.63035321e-01 8.90292972e-02 4.43857253e-01
-1.40855312e+00 -5.54575741e-01 8.41042101e-01 -5.01682818e-01
9.78975356e-01 4.72390354e-02 9.17921782e-01 1.27121627e+00
1.55819759e-01 4.93965715e-01 1.23071384e+00 -3.23318332e-01
5.47608376e-01 6.84421539e-01 -1.39359087e-01 6.62478864e-01
-3.67817521e-01 -1.99964076e-01 -7.48941422e-01 -5.69660589e-02
1.10979140e-01 2.30135489e-02 9.19251367e-02 -4.56236482e-01
-3.71757656e-01 8.06103230e-01 4.26056594e-01 4.01518971e-01
-7.27332413e-01 -6.38941005e-02 4.93271112e-01 1.79711789e-01
7.22321212e-01 -1.98556110e-01 -5.43056130e-01 -2.03667924e-01
-9.40906048e-01 3.04695845e-01 1.02799118e+00 2.22102746e-01
8.71985078e-01 -2.60863364e-01 1.27437651e-01 9.72117364e-01
7.91950941e-01 3.25732768e-01 2.93950975e-01 -6.68434322e-01
-2.70109236e-01 7.26177394e-01 -6.57327399e-02 -1.48171473e+00
-6.72476649e-01 -7.72898346e-02 -6.48390591e-01 6.13843560e-01
2.85900921e-01 -3.39171618e-01 -6.27675831e-01 1.61365342e+00
5.93914568e-01 8.03733431e-03 -1.19957715e-01 9.08599675e-01
9.69488680e-01 4.88858759e-01 3.42648804e-01 -2.10788757e-01
1.70206892e+00 -4.82045621e-01 -7.95317113e-01 -1.88686088e-01
2.93368697e-01 -6.60216510e-01 5.08217514e-01 8.03682864e-01
-7.00948060e-01 -7.17357159e-01 -8.37448657e-01 2.51816452e-01
-6.88776433e-01 4.08387244e-01 4.92130041e-01 9.39583361e-01
-1.17319512e+00 5.02644002e-01 -5.64482212e-01 -8.43429208e-01
6.00892007e-01 6.30547881e-01 -7.29591787e-01 8.77512246e-02
-1.05691767e+00 9.22418177e-01 1.07408032e-01 4.90396172e-01
-6.16268873e-01 -3.78367811e-01 -9.15081561e-01 -3.11630964e-02
5.18679731e-02 -2.72663981e-01 9.32215810e-01 -1.73465598e+00
-1.91757607e+00 1.10415542e+00 -2.11266175e-01 -1.20424582e-02
5.12672007e-01 -2.15682775e-01 -5.98007321e-01 1.52162462e-01
-2.32292414e-01 8.69483411e-01 9.91196871e-01 -1.05032456e+00
-5.53349376e-01 -7.80399084e-01 -2.34586388e-01 8.05775225e-02
-2.68667191e-01 6.05594575e-01 -4.78999525e-01 2.57291440e-02
-1.59299880e-01 -9.24598217e-01 -1.63932927e-02 7.22888410e-02
1.17182301e-03 -4.95316207e-01 1.07040811e+00 -6.81381941e-01
8.48300099e-01 -2.47657657e+00 2.18951441e-02 4.86793160e-01
-2.02248037e-01 1.76375791e-01 1.55921161e-01 1.85321122e-01
-4.15991068e-01 -9.61566716e-02 2.08923101e-01 -6.33280456e-01
2.07422256e-01 1.80647466e-02 1.76993251e-01 5.29236555e-01
3.30572933e-01 3.24870586e-01 -6.26870632e-01 -4.19574082e-01
4.99363959e-01 8.86813819e-01 -3.90972048e-01 3.73550653e-01
-3.21207121e-02 4.06779587e-01 -1.50958836e-01 6.14677310e-01
8.13180506e-01 2.91729957e-01 3.76277834e-01 -3.06577146e-01
-6.04181364e-02 -2.21681088e-01 -1.41500163e+00 1.46660042e+00
-3.88111353e-01 7.55794764e-01 3.24786901e-01 -9.53702211e-01
1.29626763e+00 5.22505522e-01 3.38624209e-01 -4.95553970e-01
5.82704425e-01 1.38887111e-02 -1.35611787e-01 -9.34352219e-01
2.94271767e-01 -3.02118480e-01 -2.51228120e-02 2.50749618e-01
3.44425917e-01 9.75968093e-02 -1.22838698e-01 -2.32179090e-01
7.51451373e-01 3.52833599e-01 2.82933921e-01 3.08883823e-02
7.08686829e-01 -5.91619194e-01 3.73295099e-01 3.09733748e-01
-4.85782713e-01 3.69672835e-01 8.59428942e-01 -7.54735529e-01
-5.82813501e-01 -6.75558984e-01 -3.59982029e-02 1.20534670e+00
-3.68933320e-01 -3.59570771e-01 -8.24977338e-01 -5.34983575e-01
-2.12993696e-01 3.36937875e-01 -9.94146943e-01 -6.53005391e-02
-8.66111889e-02 -6.18439317e-01 2.33305275e-01 1.17102422e-01
5.24678946e-01 -1.52179992e+00 -8.47774208e-01 -3.70944999e-02
2.67745666e-02 -1.10099006e+00 5.53199887e-01 4.28301655e-02
-4.41612989e-01 -7.80020356e-01 -4.20271099e-01 -4.29318547e-01
5.66940248e-01 -3.35054129e-01 1.00024438e+00 -9.68807042e-02
-5.56211710e-01 5.37865758e-01 -2.80634463e-01 -8.10989201e-01
-6.19331226e-02 -1.89371809e-01 -2.01068092e-02 8.08918655e-01
7.21039534e-01 -4.93624628e-01 -6.61981761e-01 2.49573633e-01
-7.71456480e-01 -2.11852789e-01 2.62928843e-01 3.34701777e-01
2.14303270e-01 1.10948598e-02 2.72373497e-01 -4.63085532e-01
4.16435391e-01 -5.48656583e-01 -5.00021636e-01 -6.60236552e-02
-6.90741390e-02 -4.61888671e-01 3.11669648e-01 -3.01833034e-01
-1.11115503e+00 3.81638318e-01 -3.80564630e-01 -2.61740953e-01
-9.14674044e-01 3.38839233e-01 -5.28599620e-01 8.41790512e-02
5.19288182e-01 -2.61767477e-01 3.21686566e-01 -1.65486693e-01
5.74613959e-02 9.28042889e-01 6.31029755e-02 -1.65448159e-01
1.26631647e-01 5.81305444e-01 3.85704115e-02 -1.04285061e+00
-7.27120042e-01 -3.48109543e-01 -7.50006497e-01 -1.12748659e+00
1.11596513e+00 -8.12038600e-01 -1.12669611e+00 7.26268947e-01
-1.01852322e+00 -2.47773498e-01 2.86897749e-01 4.53801781e-01
-4.02637005e-01 -1.32172748e-01 -9.20965672e-02 -1.16734850e+00
-1.56673655e-01 -1.10928988e+00 1.14282501e+00 5.02117813e-01
-5.12629986e-01 -9.30684805e-01 -5.51898405e-02 4.32892680e-01
3.34065735e-01 4.95838821e-01 4.47149962e-01 -6.81376457e-01
-2.78969370e-02 -5.64229429e-01 -1.51615173e-01 6.96802855e-01
-6.31319210e-02 3.15932304e-01 -1.48512232e+00 7.52099231e-02
-1.19617067e-01 -5.19590318e-01 4.73843694e-01 3.46978664e-01
1.14133775e+00 2.00164258e-01 -7.36412331e-02 3.11453909e-01
1.31138670e+00 5.71619458e-02 8.02183390e-01 2.03356341e-01
4.00714040e-01 1.24854124e+00 5.42593896e-01 7.22201645e-01
2.16728419e-01 8.88832092e-01 8.20758343e-01 -7.96828419e-02
2.14944288e-01 3.53130013e-01 5.97296715e-01 -1.62840560e-02
-8.56616348e-02 3.18333171e-02 -7.21621871e-01 1.50459602e-01
-1.79852855e+00 -1.04070950e+00 -1.54604837e-01 2.09575605e+00
2.05393702e-01 -1.88457593e-01 2.27640986e-01 1.38027877e-01
6.20556295e-01 1.13744467e-01 -1.95549801e-01 -9.75774646e-01
1.33012027e-01 2.38196865e-01 -1.79879129e-01 2.68782586e-01
-1.22799802e+00 8.35216403e-01 5.46422338e+00 5.85629642e-01
-1.62462056e+00 -9.99122486e-02 9.49606538e-01 -2.26405859e-01
4.62281853e-01 -2.93323934e-01 -7.90390372e-01 3.45768511e-01
1.18155110e+00 4.72093225e-01 3.05246919e-01 9.79556739e-01
5.79487145e-01 -5.68930626e-01 -9.89704609e-01 1.14359951e+00
5.97295463e-01 -6.42761469e-01 -5.30791044e-01 1.13095053e-01
2.30271086e-01 -1.09186545e-01 7.63028935e-02 3.85997504e-01
-1.69019401e-01 -1.14083099e+00 7.68558025e-01 8.91954362e-01
3.27934265e-01 -9.46181595e-01 9.69554067e-01 1.63469881e-01
-7.19511211e-01 -1.38654619e-01 -1.32385179e-01 -2.62887895e-01
1.35495573e-01 5.81129968e-01 -9.17518914e-01 2.80908644e-01
1.14057302e+00 5.49060643e-01 -5.80016911e-01 6.51780307e-01
-2.97664255e-01 3.76274914e-01 -3.48966748e-01 -2.39700273e-01
2.11860146e-02 -1.14039712e-01 1.92379922e-01 1.41250598e+00
2.93006212e-01 3.55886742e-02 -8.62640291e-02 5.53873420e-01
1.77093282e-01 4.09938902e-01 -5.49522102e-01 3.23382974e-01
-2.21811384e-01 1.96302998e+00 -8.13492537e-01 -2.44937241e-01
-2.24198163e-01 9.43836331e-01 3.98991436e-01 1.29571378e-01
-7.24892378e-01 -1.34314656e-01 7.11220741e-01 -1.32351041e-01
3.18281472e-01 2.06093743e-01 1.35141835e-01 -7.73394167e-01
-7.62723908e-02 -8.51743579e-01 3.05912405e-01 -1.09341133e+00
-1.06749463e+00 6.58480644e-01 -8.60040113e-02 -7.84184992e-01
-3.30743134e-01 -9.33783650e-01 -4.99115318e-01 8.09977949e-01
-1.02618766e+00 -1.10314655e+00 -7.00422168e-01 5.73392510e-01
1.97258577e-01 -1.85613841e-01 1.20582759e+00 2.69034863e-01
-6.57676280e-01 2.67543912e-01 -3.43643010e-01 1.03728943e-01
8.48527193e-01 -9.58366752e-01 -5.91899514e-01 4.23558444e-01
-1.67310581e-01 3.10410351e-01 8.43391240e-01 -2.53928453e-01
-1.00859034e+00 -7.78150916e-01 1.12728226e+00 -2.16252044e-01
4.92612243e-01 -5.76036751e-01 -5.73187530e-01 2.39558801e-01
3.00837129e-01 2.96909839e-01 1.11783695e+00 1.88283801e-01
-4.11371171e-01 -4.78805900e-01 -1.47580266e+00 3.89407068e-01
3.79516423e-01 -4.22120482e-01 -1.34893671e-01 -1.34652415e-02
-2.56949276e-01 -1.25172466e-01 -8.51288378e-01 3.71608973e-01
1.10277987e+00 -1.63428330e+00 5.65387726e-01 -3.37962061e-01
4.81343061e-01 -1.59478217e-01 -2.28640601e-01 -1.20769608e+00
6.60615563e-02 -1.31071404e-01 3.26826692e-01 1.38072991e+00
1.09905489e-01 -4.41844434e-01 7.34501004e-01 8.33301067e-01
3.57372701e-01 -6.93824232e-01 -8.04155469e-01 -8.69664103e-02
-6.28830314e-01 -7.45377421e-01 2.18175590e-01 5.30490696e-01
-3.69190872e-02 2.12996230e-01 -3.36797148e-01 1.15919635e-01
2.79200345e-01 -2.96603888e-01 1.04189348e+00 -1.40124238e+00
-2.96287413e-04 -4.06550795e-01 -9.16591465e-01 -1.91354677e-01
4.21613127e-01 -4.95678931e-01 -3.07103023e-02 -1.34140289e+00
8.34389552e-02 4.90431376e-02 -1.63438976e-01 4.87480521e-01
2.21954107e-01 7.21751213e-01 3.08787733e-01 -3.64325881e-01
-7.91607738e-01 3.64181906e-01 5.05193174e-01 5.63280657e-02
-1.20519638e-01 1.28672663e-02 -4.41114008e-01 9.87914562e-01
8.05460572e-01 -3.69477421e-01 -2.14757286e-02 1.13752298e-01
5.03814697e-01 -1.94686428e-01 5.35308301e-01 -1.11882722e+00
1.55479852e-02 2.96656072e-01 7.97260582e-01 -4.07170087e-01
7.34100699e-01 -1.13048220e+00 3.55505347e-01 2.85959393e-01
-1.13807805e-01 -2.77812600e-01 3.23372215e-01 2.69307941e-01
-3.79124671e-01 -2.01907694e-01 9.00509596e-01 -9.51321796e-03
-6.60318494e-01 1.34244323e-01 -6.18024945e-01 -7.21465051e-01
1.14036369e+00 -3.78571719e-01 2.54745632e-01 -6.18683457e-01
-1.24546242e+00 -1.79206133e-01 1.53542086e-01 6.05964124e-01
4.36417103e-01 -1.17147112e+00 -5.82284331e-01 3.77665162e-01
1.35600477e-01 -3.96862626e-01 6.90262794e-01 9.23863947e-01
-2.99007118e-01 4.15385934e-03 -5.11148214e-01 -8.28661263e-01
-2.00375223e+00 1.76176250e-01 4.60308164e-01 9.11819655e-03
-5.01935184e-02 8.05546939e-01 -1.40331730e-01 -3.82655501e-01
3.28312337e-01 1.01083174e-01 -8.26531589e-01 7.20591366e-01
7.97858179e-01 1.64142266e-01 2.43140936e-01 -1.10531104e+00
-4.60607797e-01 4.70436096e-01 1.21456146e-01 -1.64177552e-01
1.64125264e+00 -6.33596033e-02 -5.37845075e-01 5.73289573e-01
1.47405303e+00 -2.28927791e-01 -1.02669740e+00 1.79318637e-01
-1.60761252e-01 -3.46414149e-01 2.38035023e-02 -8.30183089e-01
-1.01266968e+00 1.01528513e+00 1.23230553e+00 2.63572544e-01
1.34648812e+00 -1.96937062e-02 -1.08897932e-01 2.02100918e-01
1.18193813e-01 -1.25584483e+00 1.44950166e-01 4.57338333e-01
9.64772642e-01 -1.25172460e+00 -4.57359385e-03 -2.01233461e-01
-6.50133252e-01 1.33088410e+00 4.47324485e-01 -1.31570682e-01
9.75212455e-01 -3.44516225e-02 3.53263944e-01 -3.61229926e-01
-8.26993823e-01 -1.62929401e-01 2.40890503e-01 5.80285311e-01
5.79951704e-01 3.63745131e-02 -7.35048205e-02 6.99254036e-01
-1.99516758e-01 2.28161559e-01 1.11103930e-01 5.30388713e-01
-3.77171218e-01 -1.04021418e+00 -6.18596613e-01 1.86230958e-01
-7.07795203e-01 4.00494307e-01 -4.91921693e-01 5.93069196e-01
7.50631928e-01 1.17138720e+00 2.46941864e-01 -3.00943017e-01
2.79868931e-01 4.08389479e-01 3.79050016e-01 -4.20318007e-01
-9.53995705e-01 3.41699980e-02 2.11270958e-01 -8.78188133e-01
-7.08159745e-01 -8.99947822e-01 -8.71403873e-01 2.06049886e-02
1.63856640e-01 -1.23181760e-01 1.30922186e+00 9.75791276e-01
4.08650339e-01 3.13583404e-01 6.85807467e-01 -1.35804510e+00
2.27165177e-01 -1.17033195e+00 -5.95920086e-01 4.69046056e-01
-4.09074016e-02 -7.01985478e-01 -4.18528587e-01 1.98177427e-01] | [13.509827613830566, 2.1185038089752197] |
f8eef892-9932-4752-aa0a-0dd593b4057f | contournet-taking-a-further-step-toward | 2004.04940 | null | https://arxiv.org/abs/2004.04940v1 | https://arxiv.org/pdf/2004.04940v1.pdf | ContourNet: Taking a Further Step toward Accurate Arbitrary-shaped Scene Text Detection | Scene text detection has witnessed rapid development in recent years. However, there still exists two main challenges: 1) many methods suffer from false positives in their text representations; 2) the large scale variance of scene texts makes it hard for network to learn samples. In this paper, we propose the ContourNet, which effectively handles these two problems taking a further step toward accurate arbitrary-shaped text detection. At first, a scale-insensitive Adaptive Region Proposal Network (Adaptive-RPN) is proposed to generate text proposals by only focusing on the Intersection over Union (IoU) values between predicted and ground-truth bounding boxes. Then a novel Local Orthogonal Texture-aware Module (LOTM) models the local texture information of proposal features in two orthogonal directions and represents text region with a set of contour points. Considering that the strong unidirectional or weakly orthogonal activation is usually caused by the monotonous texture characteristic of false-positive patterns (e.g. streaks.), our method effectively suppresses these false positives by only outputting predictions with high response value in both orthogonal directions. This gives more accurate description of text regions. Extensive experiments on three challenging datasets (Total-Text, CTW1500 and ICDAR2015) verify that our method achieves the state-of-the-art performance. Code is available at https://github.com/wangyuxin87/ContourNet. | ['Zheng-Jun Zha', 'Mengting Xing', 'Hongtao Xie', 'Yuxin Wang', 'Yongdong Zhang', 'Zilong Fu'] | 2020-04-10 | contournet-taking-a-further-step-toward-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Wang_ContourNet_Taking_a_Further_Step_Toward_Accurate_Arbitrary-Shaped_Scene_Text_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Wang_ContourNet_Taking_a_Further_Step_Toward_Accurate_Arbitrary-Shaped_Scene_Text_CVPR_2020_paper.pdf | cvpr-2020-6 | ['scene-text-detection'] | ['computer-vision'] | [ 2.79111952e-01 -1.70843363e-01 -9.58373621e-02 -3.70976448e-01
-6.83993042e-01 -3.73071551e-01 7.29124248e-01 -1.62935495e-01
-1.16871566e-01 4.18493599e-01 1.70554847e-01 3.40280756e-02
3.42983544e-01 -8.85697544e-01 -6.27477407e-01 -8.42247009e-01
3.38297665e-01 5.95933855e-01 8.57335210e-01 -1.81279808e-01
4.40053076e-01 2.87186682e-01 -1.30159342e+00 8.70629847e-01
7.52897799e-01 1.15369034e+00 1.01956882e-01 4.79365617e-01
-5.87568164e-01 5.57585299e-01 -5.81487954e-01 -3.96325827e-01
3.23482513e-01 -3.05606663e-01 -3.74219388e-01 2.43387282e-01
6.80135787e-01 -4.86647338e-01 -2.40781754e-01 1.30191755e+00
5.99988937e-01 1.08226515e-01 8.69628489e-01 -9.58018780e-01
-6.27877355e-01 7.25800931e-01 -1.29063296e+00 1.96556836e-01
1.56970099e-01 5.93878329e-02 1.01804042e+00 -1.47189605e+00
5.38300097e-01 1.39886773e+00 7.79401302e-01 3.67465317e-01
-1.06189156e+00 -7.30239868e-01 4.72788215e-01 -5.03157936e-02
-1.71282446e+00 -1.12693824e-01 9.46700811e-01 -1.91111594e-01
5.86487055e-01 4.03618962e-01 5.00558555e-01 1.37192559e+00
3.01062763e-01 1.43962920e+00 9.03826654e-01 -2.50810236e-01
7.71677122e-02 1.65115267e-01 9.69951414e-03 6.35703266e-01
2.67400861e-01 -1.59443542e-01 -5.90235233e-01 -1.76819321e-02
9.50150609e-01 5.10104895e-02 -2.97154665e-01 -3.70750964e-01
-1.27596998e+00 7.57999420e-01 5.39322555e-01 3.25921088e-01
-8.00131038e-02 -1.27405033e-01 3.63256872e-01 -1.52414531e-01
6.54819608e-01 -2.10622311e-01 -2.49487564e-01 3.44804734e-01
-1.03316057e+00 1.62166685e-01 3.82002026e-01 9.39179301e-01
4.25268352e-01 2.72253990e-01 -3.61371160e-01 1.20145249e+00
2.24820092e-01 7.85407364e-01 6.99335456e-01 5.47930896e-02
7.68556893e-01 7.92776704e-01 -5.44727407e-02 -1.43696439e+00
-4.54431176e-01 -4.68288362e-01 -1.20585203e+00 -6.51444942e-02
4.54596937e-01 2.99156215e-02 -9.85459447e-01 1.08130801e+00
4.86503482e-01 3.88844833e-02 -3.73576790e-01 1.20818937e+00
9.85672116e-01 8.55801880e-01 -4.92576696e-02 9.77313221e-02
1.42484558e+00 -9.17690992e-01 -6.56968653e-01 -3.94591689e-01
3.66310596e-01 -1.01082301e+00 1.18528020e+00 4.20645624e-01
-7.53705800e-01 -5.32234073e-01 -8.69452834e-01 1.00367785e-01
-4.32371497e-01 6.42504811e-01 2.49190375e-01 4.23085511e-01
-8.06203902e-01 2.24708661e-01 -5.57591677e-01 -3.20256144e-01
5.88482678e-01 2.01073319e-01 8.09464753e-02 -5.62193512e-04
-9.33165133e-01 3.57192665e-01 4.75711167e-01 3.58436942e-01
-3.22970271e-01 -1.40380323e-01 -6.71757042e-01 -8.81010145e-02
6.07770383e-01 -2.10175008e-01 8.79540443e-01 -1.16323709e+00
-1.29560804e+00 6.93250597e-01 -3.17438990e-02 -2.25313291e-01
9.83861566e-01 -2.57230997e-01 -3.85258526e-01 2.29905009e-01
3.17063183e-01 8.66462171e-01 1.16852903e+00 -1.24469459e+00
-6.98585391e-01 -3.34243059e-01 -7.09300458e-01 4.00046647e-01
-3.88952345e-01 4.12657745e-02 -6.87810361e-01 -1.32902980e+00
6.52276576e-01 -6.15487874e-01 2.13884730e-02 4.21936572e-01
-7.83684552e-01 -4.45938796e-01 1.05638444e+00 -4.59303737e-01
1.26884282e+00 -1.94835162e+00 -4.54719424e-01 2.42782727e-01
4.91768792e-02 1.11044265e-01 -7.23070800e-02 9.20901373e-02
4.47722301e-02 1.91014647e-01 -2.95662314e-01 -2.27452993e-01
-1.16055325e-01 -1.77046448e-01 -7.79405832e-01 5.73885500e-01
3.74288678e-01 8.45101714e-01 -5.84426463e-01 -9.76352096e-01
5.17185807e-01 3.84697646e-01 -2.13532314e-01 -1.15726084e-01
-3.54235232e-01 -2.39575412e-02 -8.10749769e-01 8.91137123e-01
1.06381631e+00 -3.84944797e-01 1.13814205e-01 -2.90968329e-01
-9.99524593e-02 -8.23922381e-02 -1.50598168e+00 1.18253183e+00
1.78448230e-01 6.49325728e-01 -2.37722829e-01 -7.74894476e-01
1.10867810e+00 7.87889063e-02 1.74737424e-01 -7.17481494e-01
4.45172489e-01 6.22529862e-03 -3.70950192e-01 -3.51802677e-01
6.41665518e-01 3.47361565e-02 4.53892387e-02 3.31402987e-01
-4.31776345e-01 2.36871000e-02 -2.52036713e-02 2.20520590e-02
5.71663320e-01 3.24682653e-01 2.70753473e-01 -4.11906123e-01
6.15552545e-01 4.87163812e-02 6.42993271e-01 9.37375546e-01
-4.01924372e-01 1.06436419e+00 4.84043121e-01 -8.39238048e-01
-9.25882339e-01 -1.14415860e+00 -3.01892549e-01 1.19611168e+00
3.95238876e-01 -2.03785494e-01 -6.09240353e-01 -8.77013326e-01
-1.38980523e-01 5.11887670e-01 -8.95089149e-01 1.48090109e-01
-8.55395973e-01 -9.53279436e-01 5.11441171e-01 6.85594141e-01
9.05115366e-01 -1.20822465e+00 -3.84812355e-01 2.03912687e-02
-3.37603301e-01 -1.07618868e+00 -8.16971123e-01 -2.56325193e-02
-9.58202720e-01 -6.87717319e-01 -9.30583775e-01 -8.86202097e-01
8.98141861e-01 5.21042109e-01 7.71638274e-01 3.22284177e-02
-4.43362236e-01 -1.66924298e-01 -3.93911332e-01 -2.59440392e-01
-1.46684349e-02 -2.07429081e-01 -2.35702485e-01 3.24285179e-01
4.19443816e-01 2.27242127e-01 -6.96598053e-01 6.62670195e-01
-8.15872312e-01 4.00740862e-01 7.11666882e-01 1.03089559e+00
8.58647406e-01 2.62734592e-01 3.55507761e-01 -7.83157051e-01
4.64745790e-01 -2.28350416e-01 -5.23576260e-01 2.47048363e-01
-2.83220112e-01 -2.60179758e-01 7.83082306e-01 -7.29868591e-01
-1.18285060e+00 1.17740050e-01 1.70059130e-01 -4.39323753e-01
-3.24196815e-01 4.91656130e-03 7.25643411e-02 2.12729707e-01
7.14328945e-01 6.49761200e-01 -3.73535514e-01 -3.14081758e-01
6.21079057e-02 6.49745166e-01 2.68936336e-01 -5.30909777e-01
8.56116593e-01 8.71236324e-01 -2.48966292e-01 -1.07926786e+00
-8.71188521e-01 -5.79125464e-01 -6.12666011e-01 -3.45186144e-01
7.04939485e-01 -7.64106452e-01 -4.70069468e-01 5.27169466e-01
-9.44538176e-01 -3.97989124e-01 3.23492065e-02 1.46970645e-01
-1.88480332e-01 6.70400620e-01 -5.74584544e-01 -9.61401284e-01
-7.40321577e-01 -9.02359009e-01 1.46728897e+00 3.14150244e-01
-4.80077416e-02 -8.09063971e-01 -2.57405609e-01 1.57285661e-01
2.16601759e-01 7.30743557e-02 5.62976658e-01 -5.99229276e-01
-4.81828719e-01 -2.46404648e-01 -6.70998991e-01 -1.93797443e-02
-1.35026395e-01 1.63913563e-01 -9.77119982e-01 -1.81991100e-01
-1.74724057e-01 -3.16158175e-01 1.00954795e+00 5.48365295e-01
1.39211607e+00 -3.00516546e-01 -4.01695311e-01 5.00974238e-01
1.27410758e+00 -4.63091582e-02 3.45661312e-01 2.59091437e-01
6.73382401e-01 5.82282186e-01 8.79171729e-01 5.26950836e-01
7.81207206e-03 6.51316166e-01 2.00349435e-01 -3.20781916e-01
-2.22245798e-01 -3.12947959e-01 1.97029188e-01 4.05927807e-01
3.12431306e-01 -4.79356855e-01 -9.66452718e-01 4.18515444e-01
-1.95220947e+00 -8.57414007e-01 -5.20845413e-01 2.07547832e+00
6.60507739e-01 5.11222363e-01 3.84814218e-02 -2.36348365e-04
1.14405990e+00 2.42584124e-01 -5.77063322e-01 7.37369955e-02
-5.38460612e-01 -6.77290112e-02 3.92817765e-01 1.80293292e-01
-1.56720555e+00 1.34996057e+00 5.48755550e+00 1.31917441e+00
-1.27199376e+00 -2.06820264e-01 1.04738510e+00 5.64474463e-02
1.01026669e-01 -4.58012491e-01 -9.41772223e-01 4.95020807e-01
1.08159363e-01 4.08267677e-01 7.49935657e-02 9.68021035e-01
2.69887060e-01 -5.46729006e-02 -4.36288118e-01 9.41451490e-01
2.97393590e-01 -1.32082784e+00 4.74693179e-01 -3.35705101e-01
8.16812634e-01 1.02638535e-01 1.96957469e-01 2.23872796e-01
1.73336357e-01 -7.41811395e-01 1.00039506e+00 2.38801181e-01
9.95684147e-01 -5.69791079e-01 7.73320377e-01 3.92566085e-01
-1.57391822e+00 1.15675874e-01 -7.25679517e-01 3.69840324e-01
-3.03302348e-01 6.14681184e-01 -1.14536011e+00 2.28066921e-01
9.28310454e-01 7.58148432e-01 -7.19105422e-01 8.74005616e-01
-2.93822914e-01 7.47264564e-01 -5.72569132e-01 -4.74882215e-01
2.95181811e-01 -1.53092042e-01 5.79480588e-01 1.53285968e+00
7.67255276e-02 -5.16305007e-02 3.75850677e-01 1.11382890e+00
-1.62034836e-02 5.41294396e-01 -4.73426700e-01 3.60383958e-01
2.96593577e-01 1.47703969e+00 -1.59942758e+00 -5.17580986e-01
-2.36037999e-01 9.16765153e-01 1.08138159e-01 3.31937075e-01
-8.35734189e-01 -2.15884104e-01 -7.54524767e-02 7.26043805e-02
4.27972496e-01 1.79271117e-01 -7.42961407e-01 -1.29997849e+00
1.90499231e-01 -7.55739212e-01 5.18445194e-01 -9.61859882e-01
-1.41137481e+00 5.01336455e-01 -3.24527323e-01 -1.53237569e+00
4.01269406e-01 -7.10123062e-01 -8.61943662e-01 7.24002421e-01
-1.17841446e+00 -1.37724626e+00 -4.66067135e-01 5.72251976e-01
1.19366634e+00 1.15654662e-01 3.52428764e-01 -1.03389420e-01
-8.19014013e-01 6.97001696e-01 1.01175070e-01 5.01426458e-01
9.52738583e-01 -1.19159663e+00 6.57830715e-01 8.50196421e-01
4.67257053e-02 4.57788616e-01 5.22235870e-01 -1.07693136e+00
-1.06636977e+00 -1.27826524e+00 4.09529328e-01 -3.74896079e-01
5.73796749e-01 -6.31189823e-01 -1.07321680e+00 3.66123199e-01
-1.40056118e-01 1.87998727e-01 2.00244039e-01 -2.83406615e-01
-3.62744838e-01 9.05782580e-02 -9.68092382e-01 9.51687455e-01
7.65937686e-01 -2.09558923e-02 -4.26042885e-01 4.93743688e-01
3.69424105e-01 -4.49136287e-01 -2.07108572e-01 4.22394812e-01
5.53104639e-01 -1.10169029e+00 8.24187636e-01 -1.14135846e-01
3.17532182e-01 -4.57031131e-01 -1.19954310e-01 -6.09856009e-01
-3.00908804e-01 -2.96125829e-01 6.68846518e-02 1.00998878e+00
2.30021596e-01 -6.23281479e-01 1.13140965e+00 -8.54174420e-02
-2.59181764e-02 -1.09476173e+00 -8.13029349e-01 -4.27013665e-01
5.61658107e-02 -3.83133918e-01 3.80383343e-01 1.04666162e+00
-2.21000969e-01 3.44966203e-01 -3.07890713e-01 1.52092889e-01
5.36194503e-01 2.61941403e-01 7.22217917e-01 -1.10088444e+00
-3.79136167e-02 -7.78824449e-01 -1.48705840e-01 -1.47754526e+00
-3.06242704e-01 -7.11273611e-01 3.10708672e-01 -1.30029321e+00
4.20357406e-01 -4.88168746e-01 -4.10770848e-02 4.37831908e-01
-4.56882268e-01 4.29850131e-01 -9.12858769e-02 4.33368027e-01
-7.79269695e-01 6.65862203e-01 1.42430162e+00 -1.65504292e-01
-9.77295786e-02 2.36808155e-02 -3.81772369e-01 1.03995669e+00
9.22621727e-01 -3.78107250e-01 -6.88047186e-02 -2.20227987e-01
2.21637562e-01 -1.31471664e-01 3.36595297e-01 -8.50313723e-01
2.84068465e-01 -2.88867474e-01 1.04698765e+00 -1.37932122e+00
1.80269778e-01 -6.60261333e-01 -4.57033396e-01 4.23560023e-01
-2.80673563e-01 -1.64353803e-01 2.73822397e-01 7.31933475e-01
-3.26178782e-03 -6.00106157e-02 9.51197743e-01 -3.90649075e-03
-5.73104858e-01 2.57186025e-01 -3.47533762e-01 5.80383092e-03
7.83239365e-01 -4.89253193e-01 -5.71751595e-01 -1.37291327e-01
-2.19531819e-01 2.10348099e-01 3.39726776e-01 5.02920687e-01
8.67419183e-01 -1.12526000e+00 -7.76923358e-01 4.02386248e-01
1.51111364e-01 2.66779035e-01 3.84461462e-01 8.11688721e-01
-6.43705010e-01 3.65912110e-01 1.43763676e-01 -1.03097045e+00
-1.17839873e+00 3.04371804e-01 4.15806890e-01 -2.67999411e-01
-1.05986071e+00 8.93072546e-01 8.79854441e-01 -4.23696280e-01
2.76575744e-01 -2.66851693e-01 -2.32893661e-01 -1.54044136e-01
5.37428916e-01 1.66810662e-01 -6.17159605e-02 -7.24250257e-01
-2.14756519e-01 9.20579314e-01 -4.42614973e-01 1.12530813e-01
8.86539102e-01 -1.93367293e-03 1.15930662e-01 4.02661651e-01
8.16125810e-01 7.47429878e-02 -1.37394154e+00 -3.78734261e-01
-1.18920423e-01 -5.33023238e-01 3.08951624e-02 -8.09277654e-01
-9.45258141e-01 9.80753839e-01 6.63448691e-01 1.51484743e-01
9.48555768e-01 -2.13024378e-01 6.72627509e-01 4.31036234e-01
4.95435409e-02 -1.35293782e+00 4.95288104e-01 4.92667615e-01
9.56372559e-01 -1.38785613e+00 1.38637066e-01 -6.33768737e-01
-7.70471990e-01 1.38188112e+00 1.00319493e+00 -2.46967316e-01
3.41669798e-01 2.75500238e-01 1.81588873e-01 -1.58235252e-01
-5.89125872e-01 9.13842320e-02 3.81867826e-01 3.28285277e-01
4.32566255e-01 2.01542914e-01 -9.44255292e-02 4.64590698e-01
-2.70621870e-02 -6.44981384e-01 3.91695261e-01 8.10444832e-01
-7.48165727e-01 -4.96682703e-01 -9.07101691e-01 7.70587742e-01
-5.50527215e-01 -1.86591908e-01 -6.70424879e-01 9.07591701e-01
-8.87807980e-02 6.39609158e-01 2.64541835e-01 -3.01220089e-01
1.16714932e-01 -6.85703456e-02 -3.09302262e-03 -4.12514776e-01
-3.93990248e-01 7.25841463e-01 -3.11129719e-01 -3.90283525e-01
1.30180925e-01 -6.73409283e-01 -1.58013320e+00 -6.05217703e-02
-7.71896124e-01 -3.09896380e-01 4.57208782e-01 5.22580564e-01
9.39593092e-02 5.12232244e-01 4.90796685e-01 -7.66323805e-01
-4.94382948e-01 -1.11832643e+00 -4.96236891e-01 4.32917744e-01
3.26592512e-02 -6.40066087e-01 -4.05208439e-01 1.59584776e-01] | [12.07292366027832, 2.26259708404541] |
61a3a4d2-e0d1-496f-815d-e138430ff6a0 | pointwise-paraphrase-appraisal-is-potentially | 2005.11996 | null | https://arxiv.org/abs/2005.11996v2 | https://arxiv.org/pdf/2005.11996v2.pdf | Pointwise Paraphrase Appraisal is Potentially Problematic | The prevailing approach for training and evaluating paraphrase identification models is constructed as a binary classification problem: the model is given a pair of sentences, and is judged by how accurately it classifies pairs as either paraphrases or non-paraphrases. This pointwise-based evaluation method does not match well the objective of most real world applications, so the goal of our work is to understand how models which perform well under pointwise evaluation may fail in practice and find better methods for evaluating paraphrase identification models. As a first step towards that goal, we show that although the standard way of fine-tuning BERT for paraphrase identification by pairing two sentences as one sequence results in a model with state-of-the-art performance, that model may perform poorly on simple tasks like identifying pairs with two identical sentences. Moreover, we show that these models may even predict a pair of randomly-selected sentences with higher paraphrase score than a pair of identical ones. | ['Yangfeng Ji', 'David Evans', 'Hannah Chen'] | 2020-05-25 | pointwise-paraphrase-appraisal-is-potentially-1 | https://aclanthology.org/2020.acl-srw.20 | https://aclanthology.org/2020.acl-srw.20.pdf | acl-2020-6 | ['paraphrase-identification'] | ['natural-language-processing'] | [ 3.74298602e-01 -9.35008749e-02 -1.97337210e-01 -6.65825188e-01
-8.91838372e-01 -8.60785007e-01 7.33650684e-01 5.12791932e-01
-3.39815795e-01 6.81127012e-01 2.29223534e-01 -5.21396458e-01
7.35927699e-03 -6.72450840e-01 -6.19529188e-01 -4.22536522e-01
3.39342654e-01 7.36568153e-01 1.62245110e-01 -3.82281303e-01
5.98470926e-01 2.60315567e-01 -1.48524857e+00 7.64372706e-01
8.71660829e-01 7.44144440e-01 8.02965984e-02 8.39077055e-01
-6.85780263e-03 7.78274596e-01 -6.27974749e-01 -8.47397387e-01
2.84327894e-01 -8.44300628e-01 -1.46723223e+00 -1.28549322e-01
6.77000701e-01 4.11386462e-03 -1.20192155e-01 1.11135149e+00
2.30823532e-01 -2.53744572e-01 8.70547414e-01 -1.04412985e+00
-6.46502495e-01 6.66295171e-01 -4.35414873e-02 2.73280948e-01
1.09084749e+00 -1.28508851e-01 1.22949445e+00 -7.43061900e-01
3.11687440e-01 9.47026551e-01 9.11648095e-01 3.67668927e-01
-1.42137539e+00 -4.60509956e-01 -3.80475372e-01 4.47847068e-01
-1.05369055e+00 -6.00083232e-01 5.87373376e-01 -3.44873220e-01
9.64665413e-01 6.52265072e-01 5.47475517e-01 1.21857858e+00
4.53400053e-02 6.28905416e-01 1.35112357e+00 -7.33899891e-01
2.55932249e-02 4.40171361e-01 6.49502695e-01 3.36157262e-01
1.76835164e-01 -3.70268263e-02 -3.67742121e-01 -2.98074335e-01
2.14828283e-01 -1.28908798e-01 -2.71495104e-01 -3.97682339e-01
-1.16119659e+00 8.86502624e-01 2.48635799e-01 6.82890296e-01
-1.17435120e-01 -4.12088513e-01 4.78263497e-01 1.04045951e+00
2.68637896e-01 1.10290730e+00 -1.82586536e-01 -3.06828052e-01
-1.18455195e+00 4.69975740e-01 1.07294858e+00 6.64333224e-01
5.30977488e-01 -6.84933007e-01 -2.42998019e-01 1.25174844e+00
-1.39313266e-01 1.04503654e-01 9.89971519e-01 -8.40808630e-01
5.33852935e-01 5.82472444e-01 2.22771510e-01 -7.91217446e-01
-1.65028080e-01 -2.36786619e-01 -6.05935514e-01 1.22388966e-01
6.31261408e-01 3.54571939e-01 -3.69174868e-01 1.61491907e+00
-4.84834760e-01 -1.55541211e-01 1.25597194e-01 6.56228065e-01
5.44203043e-01 5.14177680e-01 -1.60904780e-01 -3.40998352e-01
1.19225371e+00 -9.83662844e-01 -2.12762654e-01 -4.96087700e-01
8.32697332e-01 -1.06090295e+00 1.26983440e+00 2.23783985e-01
-1.50834095e+00 -8.76817703e-01 -1.23798752e+00 6.79882094e-02
-1.35566086e-01 5.19750193e-02 2.48894885e-01 7.92217910e-01
-1.05421567e+00 1.00295269e+00 -2.49711290e-01 -7.13334560e-01
-5.21264486e-02 2.12886810e-01 -5.78675210e-01 6.35870844e-02
-1.32019162e+00 1.45337951e+00 3.58821839e-01 -2.25774288e-01
-4.23766196e-01 -4.51364219e-01 -6.36058807e-01 3.42431754e-01
-1.98027745e-01 -7.90808141e-01 1.53015423e+00 -1.26883101e+00
-1.21409357e+00 1.62878704e+00 -5.18802702e-01 -7.44365156e-01
6.90663099e-01 1.23431318e-01 -3.72182310e-01 2.28933301e-02
2.08966568e-01 1.87553823e-01 7.15339720e-01 -1.08095956e+00
-4.98248756e-01 -5.02017021e-01 2.53441304e-01 3.29656273e-01
-3.93935382e-01 3.73967588e-01 1.77295059e-01 -4.80199605e-01
2.74656564e-01 -8.39794397e-01 1.26097932e-01 -3.07629734e-01
-4.20751780e-01 -2.26236165e-01 2.58436620e-01 -5.66134214e-01
1.41186666e+00 -1.71199226e+00 8.99912119e-02 6.41167685e-02
7.44047239e-02 5.30781567e-01 -8.98229480e-02 7.75039732e-01
-6.76825762e-01 2.67070919e-01 -1.98384449e-01 -6.02767348e-01
4.86746952e-02 -1.77017033e-01 -4.56332266e-01 2.45205790e-01
1.30626783e-01 7.99594879e-01 -1.10024714e+00 -3.96589041e-01
5.37396260e-02 -3.40439767e-01 -2.65236557e-01 4.10411507e-01
3.08140278e-01 -7.96083584e-02 -2.63988692e-02 2.11035594e-01
4.53364044e-01 -2.97861308e-01 3.48384231e-01 4.14759852e-02
1.97416484e-01 6.69266522e-01 -7.61374474e-01 1.16617656e+00
-5.96007109e-01 8.14468205e-01 -2.77702719e-01 -1.47417021e+00
9.00154293e-01 3.06961894e-01 -4.25296463e-02 -6.58224821e-01
-1.53224126e-01 4.88545418e-01 1.16808131e-01 -7.01602817e-01
4.90692675e-01 -6.12029850e-01 -2.39275992e-01 7.35214174e-01
-1.03402697e-01 -3.38540554e-01 2.14990199e-01 1.22943819e-01
1.17350996e+00 -4.56856817e-01 4.62724537e-01 -1.51361644e-01
9.04098332e-01 -4.36501205e-02 7.75590613e-02 1.27836204e+00
-3.70296836e-01 9.22383308e-01 7.81064272e-01 -2.87078232e-01
-1.44509900e+00 -1.26715040e+00 -2.01265231e-01 9.20043051e-01
1.05920821e-01 -4.37564552e-01 -7.75017738e-01 -7.70641863e-01
-2.35899631e-02 9.22048748e-01 -6.20047927e-01 -4.13466096e-01
-5.33378184e-01 -5.26794076e-01 5.80475867e-01 4.14554238e-01
2.48729795e-01 -1.06819844e+00 -1.07361168e-01 -2.61569954e-02
-4.91949558e-01 -7.34962821e-01 -4.07218099e-01 2.19721258e-01
-7.84859896e-01 -1.17943048e+00 -8.62354457e-01 -1.21649981e+00
4.79908407e-01 4.09006506e-01 1.55231941e+00 2.43671775e-01
2.97592700e-01 -2.15661168e-01 -5.27360618e-01 -1.39962006e-02
-1.29051375e+00 1.00512765e-01 1.20935507e-01 -1.21041581e-01
6.13735020e-01 -5.51643550e-01 -4.87493396e-01 5.13803899e-01
-7.23478675e-01 -2.22430564e-02 4.47891772e-01 1.18044972e+00
-1.59848124e-01 -3.19130778e-01 7.01992869e-01 -1.02685380e+00
1.22264349e+00 -4.62795466e-01 7.35567212e-02 6.83787882e-01
-7.68133640e-01 6.64447546e-02 1.00689554e+00 -2.50045538e-01
-5.39941549e-01 -1.21848904e-01 -3.03325802e-01 -1.56313390e-01
-3.36889595e-01 3.24017465e-01 3.88651758e-01 -4.04119641e-02
9.66599524e-01 6.06864035e-01 2.68891037e-01 -3.64576757e-01
-3.09704337e-02 1.12767279e+00 5.57893276e-01 -3.06579024e-01
7.96044052e-01 -9.69154313e-02 -3.86729360e-01 -6.46213710e-01
-8.45464110e-01 -8.39272439e-01 -7.20940351e-01 3.65923531e-02
4.52222019e-01 -7.09419966e-01 -4.96295571e-01 5.00879824e-01
-1.23040569e+00 -3.10455635e-02 -1.55945644e-01 8.37361813e-02
-8.05016935e-01 1.02069163e+00 -4.45843041e-01 -5.89842200e-01
-2.25537464e-01 -1.10993981e+00 9.46093380e-01 -9.43839327e-02
-9.23622489e-01 -9.56158638e-01 3.48572850e-01 8.43371391e-01
2.23104551e-01 -5.44615626e-01 1.23074865e+00 -1.18969023e+00
7.36936331e-02 -6.32288516e-01 -9.51685756e-02 7.78904915e-01
1.09791428e-01 -1.44513443e-01 -8.51646483e-01 -3.14088672e-01
3.80120367e-01 -6.38066351e-01 6.85926139e-01 5.28282262e-02
9.86712873e-01 -1.83932394e-01 -2.20927477e-01 3.05496603e-01
1.13545215e+00 2.34270915e-02 7.00513363e-01 3.33562493e-01
1.29381597e-01 7.24843621e-01 5.10027647e-01 -6.67600408e-02
4.81695272e-02 1.06805921e+00 3.45573425e-02 1.85858458e-01
5.66790029e-02 -4.48447555e-01 2.28186205e-01 7.81637549e-01
3.89656514e-01 -2.05099940e-01 -8.09358776e-01 5.56958973e-01
-1.87100542e+00 -1.52683735e+00 -2.62877345e-01 2.61657214e+00
8.44350994e-01 4.02143210e-01 5.97352684e-01 6.00162804e-01
8.89876604e-01 1.11504197e-01 -1.25125960e-01 -7.46200025e-01
-1.31822124e-01 2.52432823e-01 -2.55003711e-03 4.73590940e-01
-9.38743412e-01 6.83080137e-01 7.49331760e+00 7.82969892e-01
-7.88101792e-01 -1.83359846e-01 7.77050316e-01 1.71985045e-01
-2.85272002e-01 1.16616040e-01 -4.24486101e-01 8.61299276e-01
1.07388031e+00 -4.40595657e-01 3.26893479e-01 7.06055582e-01
2.60087818e-01 -6.96802512e-02 -1.66276062e+00 1.01034892e+00
3.10193121e-01 -1.21231973e+00 1.09104954e-01 -4.14497882e-01
5.64064980e-01 -2.59229273e-01 -1.19418256e-01 5.48792422e-01
1.20777972e-01 -1.17451954e+00 5.93866646e-01 2.40932107e-01
3.35295051e-01 -2.96473920e-01 9.33025181e-01 6.79456770e-01
-5.91078579e-01 -2.08113223e-01 -5.20445824e-01 -4.30471838e-01
2.62806714e-02 2.02489421e-01 -9.39855456e-01 4.30608004e-01
3.04077923e-01 5.51209390e-01 -1.04904246e+00 1.37525475e+00
-1.49381533e-01 5.85399389e-01 1.57903023e-02 -3.73972833e-01
1.37831504e-02 -2.82775849e-01 4.29325879e-01 1.31432068e+00
1.26554802e-01 -4.00577843e-01 -2.38544956e-01 5.85566878e-01
1.20994411e-01 1.96127728e-01 -8.48015249e-01 1.78887337e-01
5.51731110e-01 7.57314622e-01 -3.40046644e-01 -6.41523421e-01
-2.72703469e-01 1.32348728e+00 6.99450254e-01 1.38879120e-01
-6.67841613e-01 -5.32306790e-01 5.00566125e-01 1.51430398e-01
-7.97784925e-02 2.58210331e-01 -6.72753096e-01 -1.39779592e+00
3.21430236e-01 -1.05143321e+00 2.58709311e-01 -1.10278976e+00
-1.71600735e+00 6.75583661e-01 -9.49907824e-02 -1.39833534e+00
-6.20807886e-01 -5.61959684e-01 -1.04943609e+00 1.15486431e+00
-9.68112350e-01 -6.81860745e-01 -2.85042077e-01 1.36101648e-01
7.60814548e-01 -1.39309853e-01 9.18576062e-01 5.18226959e-02
-2.47989669e-01 8.29177678e-01 3.39289993e-01 1.96498469e-01
8.26337278e-01 -1.44398963e+00 7.05592155e-01 7.38888562e-01
4.29161191e-01 7.70686805e-01 1.16441941e+00 -2.27620110e-01
-6.44574165e-01 -5.01029074e-01 1.69165349e+00 -6.97400451e-01
8.30771744e-01 -2.13477954e-01 -1.06160891e+00 3.96588057e-01
9.24522355e-02 -5.72113812e-01 5.35040379e-01 3.09979647e-01
-4.78243768e-01 2.77727377e-02 -9.82949436e-01 6.79258108e-01
1.11609066e+00 -1.07692063e+00 -1.33578205e+00 8.38705242e-01
2.08123311e-01 -5.80237014e-03 -6.44171953e-01 2.44155824e-01
7.54861057e-01 -1.59583008e+00 1.19818068e+00 -1.13466477e+00
9.59677458e-01 1.49920464e-01 -1.67402402e-02 -1.48525095e+00
-4.97193336e-01 -2.81719804e-01 3.45994383e-01 9.65652525e-01
5.00751197e-01 -6.05795920e-01 1.01967680e+00 4.42873418e-01
4.79215980e-02 -7.97162533e-01 -7.81680524e-01 -1.17150593e+00
4.01662618e-01 5.45585342e-03 2.45428801e-01 9.41876888e-01
6.97945297e-01 7.28882015e-01 -1.72130510e-01 -3.07779104e-01
3.23671430e-01 4.42858070e-01 6.57106340e-01 -1.07546043e+00
-6.16482317e-01 -8.99988532e-01 -5.68891227e-01 -1.08804798e+00
5.54438055e-01 -1.08859575e+00 2.11956892e-02 -1.21594739e+00
6.41433537e-01 -2.54910499e-01 -8.40338692e-02 2.48650778e-02
-5.08159518e-01 3.75982285e-01 2.66142070e-01 5.09393156e-01
-3.76064062e-01 1.61855727e-01 6.33750856e-01 -2.09720686e-01
1.71034157e-01 6.62988722e-01 -6.93716705e-01 4.60092634e-01
7.24445581e-01 -4.68171686e-01 -4.43576992e-01 -1.78943038e-01
1.70663342e-01 3.42086405e-01 3.52550328e-01 -9.85446036e-01
5.15173525e-02 4.03145291e-02 1.29957289e-01 -1.28739923e-01
2.32833400e-01 -5.40533543e-01 2.39163831e-01 5.60113490e-01
-9.43185627e-01 5.32551348e-01 -3.58518481e-01 3.94791216e-01
-4.69675690e-01 -1.19166291e+00 9.04678047e-01 -2.51270473e-01
-3.29506636e-01 -3.55363458e-01 -3.27136248e-01 2.23996025e-02
1.00196970e+00 -7.06155002e-01 -4.14150447e-01 -6.59358144e-01
-7.57904947e-01 3.80249210e-02 9.39152062e-01 4.22065109e-01
4.38947439e-01 -1.04804254e+00 -8.54548693e-01 2.44975328e-01
4.42239285e-01 -8.49275351e-01 5.27017303e-02 4.99493778e-01
-6.26958251e-01 6.51925504e-01 -2.14525446e-01 -4.85370815e-01
-1.80773342e+00 4.78465259e-01 5.10518372e-01 -6.01536095e-01
7.25118117e-03 8.51585984e-01 3.95349488e-02 -5.64210594e-01
-1.07265403e-02 -1.43025205e-01 -1.88135073e-01 -8.04479942e-02
3.65289807e-01 1.40459642e-01 3.65384012e-01 -7.12852538e-01
-1.93293512e-01 4.40676272e-01 -3.48056942e-01 -3.68791558e-02
8.54344249e-01 -2.81994212e-02 -6.53901547e-02 5.82120359e-01
1.61325026e+00 -2.17830241e-01 -4.09982115e-01 -3.19394208e-02
6.67948872e-02 -5.45099258e-01 -5.11208475e-01 -7.63560236e-01
-1.07553646e-01 8.17780912e-01 2.37094253e-01 8.65049064e-01
8.50640714e-01 -7.12748021e-02 5.77133179e-01 5.59558332e-01
3.84700149e-01 -7.33745635e-01 1.88076913e-01 5.48572123e-01
9.27369535e-01 -1.35974765e+00 -1.07477903e-01 -3.83829176e-01
-6.84951425e-01 1.23056793e+00 3.63283694e-01 -2.90369362e-01
2.68383652e-01 -4.27773446e-01 -6.09088549e-03 3.92122753e-02
-7.67416179e-01 1.69650748e-01 2.12040469e-01 4.18923140e-01
5.65341294e-01 -1.18669137e-01 -6.42039299e-01 1.95800319e-01
-6.19647205e-01 -7.59712681e-02 6.11693203e-01 5.63337088e-01
-4.36354935e-01 -1.35809720e+00 -1.40245691e-01 8.61042559e-01
-3.00444067e-01 -1.74667567e-01 -7.58112907e-01 3.77671272e-01
-4.98386085e-01 1.07272780e+00 1.49834692e-01 -6.47991300e-01
3.66799057e-01 4.18745488e-01 5.38019598e-01 -7.82291114e-01
-1.05581522e+00 -8.67531300e-01 4.36556011e-01 -1.34329185e-01
-1.90640122e-01 -7.35599101e-01 -3.44246000e-01 -4.93499190e-01
-2.20407918e-01 4.81418818e-01 3.95227253e-01 1.09288955e+00
4.88064438e-02 -1.61373392e-01 1.00901973e+00 -5.79350293e-01
-1.37245429e+00 -1.03204954e+00 -4.93299901e-01 1.24121833e+00
2.33067840e-01 -1.68428078e-01 -7.01021254e-01 -6.51696771e-02] | [11.323047637939453, 9.16833209991455] |
1e96424a-aaa8-4fed-9e42-66896468c28c | add-2022-the-first-audio-deep-synthesis | 2202.08433 | null | https://arxiv.org/abs/2202.08433v2 | https://arxiv.org/pdf/2202.08433v2.pdf | ADD 2022: the First Audio Deep Synthesis Detection Challenge | Audio deepfake detection is an emerging topic, which was included in the ASVspoof 2021. However, the recent shared tasks have not covered many real-life and challenging scenarios. The first Audio Deep synthesis Detection challenge (ADD) was motivated to fill in the gap. The ADD 2022 includes three tracks: low-quality fake audio detection (LF), partially fake audio detection (PF) and audio fake game (FG). The LF track focuses on dealing with bona fide and fully fake utterances with various real-world noises etc. The PF track aims to distinguish the partially fake audio from the real. The FG track is a rivalry game, which includes two tasks: an audio generation task and an audio fake detection task. In this paper, we describe the datasets, evaluation metrics, and protocols. We also report major findings that reflect the recent advances in audio deepfake detection tasks. | ['Bin Liu', 'Zheng Lian', 'Haizhou Li', 'Zhengqi Wen', 'Le Xu', 'Xinrui Yan', 'Shuai Zhang', 'Shiming Wang', 'Shan Liang', 'Cunhang Fan', 'Ye Bai', 'Zhengkun Tian', 'Tao Wang', 'Chenglong Wang', 'Haoxin Ma', 'Shuai Nie', 'JianHua Tao', 'Ruibo Fu', 'Jiangyan Yi'] | 2022-02-17 | null | null | null | null | ['audio-generation'] | ['audio'] | [-3.99603844e-02 -1.27570421e-01 1.94348976e-01 1.54916555e-01
-1.62395549e+00 -4.99621928e-01 3.88651162e-01 -2.00126216e-01
1.80565044e-01 5.73996603e-01 6.04712665e-01 1.59708455e-01
3.74064118e-01 -1.95948347e-01 -6.09999418e-01 -5.64640403e-01
-2.02421039e-01 1.25073373e-01 5.07019937e-01 -3.43138784e-01
1.09284021e-01 4.69162762e-02 -1.71339691e+00 9.96821761e-01
1.77691385e-01 1.11327505e+00 -2.84942836e-01 1.00595391e+00
6.30673409e-01 8.69112611e-01 -1.29928148e+00 -4.71853107e-01
8.26353282e-02 -4.65418458e-01 -7.96481252e-01 -2.68326849e-01
5.32433689e-01 -7.53714263e-01 -5.57729185e-01 1.12560558e+00
1.38498127e+00 -2.41484925e-01 3.94600362e-01 -2.16480637e+00
-2.10561439e-01 9.25564706e-01 -4.82512087e-01 3.78961593e-01
7.82346129e-01 3.67144644e-01 9.15219486e-01 -8.46571922e-01
3.11199099e-01 1.50030112e+00 8.98951411e-01 5.16593814e-01
-5.62111557e-01 -1.42622101e+00 -3.74342620e-01 6.31271899e-01
-1.51565945e+00 -8.75414908e-01 9.07665849e-01 -6.13418400e-01
6.42461538e-01 3.94284695e-01 5.51089823e-01 1.79670262e+00
-1.48284927e-01 1.11750853e+00 8.12513053e-01 -1.14787139e-01
-7.32586309e-02 -6.81780353e-02 -2.31483653e-01 1.07304245e-01
-1.02044299e-01 4.89056319e-01 -1.07076705e+00 -5.78400373e-01
2.21683428e-01 -9.26104188e-01 -5.19804180e-01 3.69068414e-01
-1.11383450e+00 6.76038802e-01 -1.42668381e-01 4.12522703e-01
-1.49263918e-01 6.91737652e-01 1.17321765e+00 7.68725872e-01
4.84196782e-01 4.41900939e-01 -7.75223691e-03 -6.68091774e-01
-1.01849043e+00 5.64111352e-01 4.39321756e-01 7.07130969e-01
-9.48477536e-03 5.47069013e-01 -3.08649033e-01 8.49063396e-01
1.60147637e-01 3.92322600e-01 7.29945123e-01 -6.89933419e-01
6.88365817e-01 -6.05098128e-01 3.38012844e-01 -1.34101593e+00
-1.42257944e-01 -4.32458103e-01 -5.84338665e-01 -2.19593614e-01
1.79397479e-01 -1.59817114e-01 -4.01798695e-01 1.59204054e+00
2.15195373e-01 8.50964189e-01 -3.09864372e-01 1.21804893e+00
1.17731631e+00 6.09144270e-01 -3.95505279e-01 -5.25805317e-02
1.34819961e+00 -8.82208228e-01 -1.20051503e+00 1.29584610e-01
8.32094550e-01 -1.33941174e+00 1.07838416e+00 1.02757967e+00
-1.01137817e+00 -4.81730878e-01 -1.12614739e+00 1.47918966e-02
-1.13458745e-01 -1.74424294e-02 3.22644204e-01 1.13062143e+00
-9.14554954e-01 3.04882109e-01 -2.58592755e-01 1.83768943e-01
4.51313972e-01 -1.88528486e-02 -4.86635089e-01 2.77757287e-01
-1.86321962e+00 2.98794687e-01 1.43679947e-01 -5.05693555e-02
-1.62821782e+00 -5.33850014e-01 -4.20387268e-01 -1.54861376e-01
3.89045030e-01 -2.61044353e-01 1.66785443e+00 -1.03049397e+00
-1.62639618e+00 1.07561672e+00 4.58249271e-01 -6.90740705e-01
8.99592102e-01 -4.92115557e-01 -1.07347333e+00 1.65701002e-01
2.16587365e-01 5.44757724e-01 1.41934347e+00 -1.09230793e+00
-6.01653039e-01 8.01204741e-02 -3.06076765e-01 -5.80195971e-02
-6.94831461e-02 5.77781022e-01 8.56376067e-03 -1.20769382e+00
-2.91656107e-01 -8.19302022e-01 7.49087453e-01 -2.60517687e-01
-8.74780834e-01 -3.18895350e-03 1.25015509e+00 -7.54802346e-01
1.46792781e+00 -2.67465162e+00 -2.28947490e-01 -3.77623081e-01
2.32689217e-01 3.35944533e-01 -2.20248491e-01 6.82333350e-01
-3.24779361e-01 1.30437836e-01 3.54721546e-01 -4.52398270e-01
3.68228495e-01 -4.82439429e-01 -1.04725099e+00 6.41173840e-01
-1.72639921e-01 5.93660057e-01 -1.01471436e+00 -2.03512594e-01
-2.24101961e-01 6.71225563e-02 -7.42327511e-01 1.93079188e-01
-7.42279971e-03 2.10465491e-01 1.60717294e-01 8.10942411e-01
8.29666138e-01 5.75759113e-01 -3.73462945e-01 -4.23021913e-01
7.41706043e-02 6.59794807e-01 -1.31884980e+00 1.43994415e+00
1.77543797e-02 9.84774232e-01 5.13748109e-01 -7.70755768e-01
6.27808332e-01 9.26077902e-01 3.79358232e-01 -3.59016299e-01
3.38150650e-01 6.28330231e-01 1.54424952e-02 -7.14813054e-01
7.83452392e-01 -2.20597446e-01 -4.37150151e-01 6.35962129e-01
2.07345620e-01 -4.57059741e-01 -5.61551571e-01 3.37586164e-01
1.44106948e+00 -3.74199182e-01 -1.60913691e-02 1.93057761e-01
2.30936587e-01 -5.12196839e-01 3.96802127e-01 9.03829753e-01
-1.00814831e+00 8.72339487e-01 5.98182976e-01 -1.07707977e-01
-5.50449669e-01 -1.13673484e+00 3.12192082e-01 9.02309299e-01
3.83773558e-02 -7.20094800e-01 -7.10562527e-01 -6.73841119e-01
1.98117383e-02 2.90860474e-01 -2.58510888e-01 -5.61887145e-01
-3.52744311e-01 -3.13436240e-01 1.82901883e+00 -1.09818086e-01
6.73688591e-01 -1.16590893e+00 -2.40683034e-01 2.79241204e-01
-1.06045735e+00 -1.46522725e+00 -7.40890920e-01 -2.20041990e-01
4.28195633e-02 -8.05982769e-01 -3.95201772e-01 -6.98688507e-01
-6.70973599e-01 4.38817233e-01 8.37322950e-01 -1.36493221e-01
-4.93489981e-01 1.24337748e-01 -6.84859216e-01 -6.68171942e-01
-6.30521178e-01 -3.89607877e-01 4.41553265e-01 1.54187873e-01
1.60531476e-01 -5.58247685e-01 -2.72665143e-01 5.64950168e-01
-7.67114341e-01 -3.47869784e-01 9.42527801e-02 8.42831016e-01
2.22592771e-01 2.42458627e-01 9.82617617e-01 -2.70793915e-01
1.04329824e+00 -5.87600946e-01 1.79928076e-03 -4.62850749e-01
3.36423397e-01 -6.68152094e-01 2.60174453e-01 -8.28712881e-01
-4.08848614e-01 -3.76563102e-01 -5.30920804e-01 -7.26947665e-01
-2.16191307e-01 9.08944011e-02 -3.28426808e-01 1.20252796e-01
6.77321851e-01 6.90809339e-02 -3.96381408e-01 -4.99785632e-01
2.04066887e-01 1.47572112e+00 8.16736698e-01 -3.27122837e-01
6.73538923e-01 4.27661240e-01 -5.00525653e-01 -9.64573205e-01
-6.37324512e-01 -5.15111983e-01 -6.06528595e-02 -5.01246214e-01
3.64657968e-01 -1.14942110e+00 -6.82324171e-01 1.25597048e+00
-1.44624758e+00 -1.12249896e-01 -2.48587638e-01 5.36610186e-01
-7.56209910e-01 7.24042118e-01 -8.50124300e-01 -9.20460761e-01
-3.03247929e-01 -1.35766411e+00 1.54488707e+00 -9.36085820e-01
-4.36656505e-01 -6.38097748e-02 2.61634499e-01 5.98734140e-01
1.23797379e-01 3.81437957e-01 4.97248828e-01 -6.39882326e-01
-2.20258966e-01 -3.74173492e-01 -5.55570871e-02 6.10031068e-01
9.64054465e-03 -1.12930655e-01 -1.65712500e+00 -2.70967543e-01
6.39572665e-02 -8.33418787e-01 6.66910052e-01 2.09774196e-01
1.03485215e+00 -3.84792000e-01 3.46113369e-02 2.40258768e-01
6.18513167e-01 1.27311930e-01 9.38538849e-01 -4.57791686e-02
4.44017231e-01 4.07979459e-01 8.60256672e-01 7.43399501e-01
4.16818187e-02 1.25351250e+00 8.13439906e-01 3.42582315e-01
-4.98950213e-01 -3.91606063e-01 9.26358640e-01 1.15241337e+00
4.95448679e-01 -6.53789997e-01 -7.91355729e-01 6.52255714e-01
-1.52396309e+00 -1.25887978e+00 -6.00473940e-01 1.95276701e+00
6.56789541e-01 1.51729569e-01 7.47832537e-01 1.11734700e+00
9.89775002e-01 1.87433586e-01 -6.62782649e-03 -5.68405330e-01
-3.00490528e-01 1.27119079e-01 -8.82791132e-02 3.64235371e-01
-1.44288063e+00 1.15712774e+00 6.47812176e+00 1.57030809e+00
-1.14570689e+00 7.51118898e-01 2.15954185e-01 -2.49515355e-01
1.35312721e-01 -3.83912086e-01 -4.05413538e-01 8.37390780e-01
1.01223624e+00 1.01643754e-02 1.79575503e-01 4.91939843e-01
5.48154891e-01 1.60386220e-01 -8.34009886e-01 1.38406575e+00
1.50713012e-01 -9.78926718e-01 6.43682852e-02 -1.98309924e-02
2.02442899e-01 9.71097723e-02 3.66239876e-01 4.71256644e-01
-1.92928240e-02 -8.73183966e-01 1.33689570e+00 3.50227095e-02
9.67418432e-01 -8.51069331e-01 7.60708511e-01 2.02937901e-01
-1.13976431e+00 -8.80651996e-02 -2.45970748e-02 -7.48677999e-02
3.59447896e-01 7.26215959e-01 -8.31573188e-01 3.60830665e-01
9.92499769e-01 5.32550991e-01 -6.37009833e-03 1.06698430e+00
-1.11229062e-01 8.20327282e-01 -1.14795320e-01 1.40689388e-01
1.23518296e-01 6.29554689e-01 1.12458205e+00 1.49423647e+00
3.95410597e-01 -4.26148802e-01 -6.86972514e-02 5.68593144e-01
-2.56958038e-01 1.94397301e-03 -6.83877289e-01 -2.12891579e-01
8.26586604e-01 8.14304650e-01 -1.90235212e-01 -1.83615610e-01
2.49978647e-01 1.24435949e+00 -3.96155179e-01 -3.20796698e-01
-1.32194614e+00 -6.19177520e-01 1.03883278e+00 1.14171170e-01
8.56062025e-02 2.04972789e-01 2.87187040e-01 -1.06105709e+00
-8.95557627e-02 -1.48605192e+00 2.90412694e-01 -1.13539064e+00
-1.20841682e+00 4.79020149e-01 -2.12677434e-01 -1.70587969e+00
-1.28438294e-01 -1.05907507e-01 -3.47949803e-01 3.40573967e-01
-1.00909054e+00 -9.29696023e-01 -2.34953031e-01 7.67627478e-01
4.79493201e-01 -2.33640328e-01 7.43098438e-01 9.85569298e-01
-2.94155508e-01 1.10908937e+00 -1.15640603e-01 1.32244259e-01
1.15572250e+00 -7.05685318e-01 4.57488805e-01 4.60064977e-01
2.22712263e-01 -1.80851221e-01 1.09982800e+00 -5.21893501e-01
-8.27426255e-01 -1.07751513e+00 8.28120232e-01 -3.26886684e-01
1.08093619e+00 -7.68481851e-01 -5.73295295e-01 2.94649392e-01
-6.59215003e-02 8.53822753e-02 5.80848873e-01 -5.05205214e-01
-8.26309741e-01 5.48240989e-02 -1.31952202e+00 2.58391619e-01
1.04582667e+00 -9.87521946e-01 -1.49716228e-01 4.86479759e-01
1.03648031e+00 -4.81516421e-01 -3.72167349e-01 1.99866414e-01
6.94107294e-01 -1.17181170e+00 9.71748769e-01 -5.38311243e-01
3.18721205e-01 -2.44980931e-01 -4.27055627e-01 -1.37458014e+00
6.63751960e-02 -1.28449106e+00 -2.21596733e-01 1.28363824e+00
-2.86770552e-01 -2.76743799e-01 5.11991680e-01 -8.85372162e-01
-4.56471235e-01 -1.42059997e-01 -1.52478564e+00 -1.26665890e+00
-2.21392453e-01 -1.07953620e+00 7.42265403e-01 1.26474297e+00
8.37760195e-02 1.90064311e-01 -1.16998184e+00 2.47303620e-01
4.02244955e-01 -3.30699623e-01 1.06968391e+00 -9.86840308e-01
-5.88241518e-01 -2.83840507e-01 -1.14967465e+00 -1.00262058e+00
-2.96903327e-02 -4.78430033e-01 7.52453282e-02 -5.25649309e-01
-2.56361127e-01 9.21888575e-02 -4.14720597e-03 2.36422211e-01
2.80696541e-01 7.74562240e-01 2.65698045e-01 1.71034113e-01
-5.26783824e-01 5.63879311e-01 8.34240794e-01 -3.53499353e-01
1.59418866e-01 1.61669061e-01 -4.76474196e-01 6.78927422e-01
7.70327866e-01 -9.19198155e-01 -1.98531076e-01 -2.95373976e-01
2.01727733e-01 3.16132814e-01 7.55204618e-01 -1.17835188e+00
-1.83563471e-01 2.71709472e-01 -6.05846047e-01 -8.05927634e-01
8.87121499e-01 -4.23733622e-01 8.57031792e-02 5.91883659e-01
-2.15408802e-01 -3.58704850e-02 2.36733630e-01 6.53629065e-01
-5.99294066e-01 9.50747505e-02 9.54061806e-01 1.88405171e-01
-1.96809247e-01 -2.89134961e-02 -7.68770337e-01 3.50786835e-01
7.87480116e-01 -1.14907533e-01 -6.06576562e-01 -1.15047014e+00
-8.79008055e-01 -3.58362377e-01 -2.95371085e-01 7.50099540e-01
8.56401503e-01 -1.65754914e+00 -9.16825831e-01 -8.23575333e-02
1.77758902e-01 -6.62771404e-01 4.95916724e-01 9.22498882e-01
-4.82975066e-01 2.83829689e-01 -8.59711990e-02 -4.31929082e-01
-1.73996854e+00 4.33004200e-01 3.70129824e-01 -8.24107155e-02
-3.75089765e-01 1.21645641e+00 1.73970200e-02 -1.50053382e-01
6.33743405e-01 -2.03470975e-01 1.22057468e-01 1.03623949e-01
6.22641385e-01 7.39984989e-01 2.85373092e-01 -1.03472662e+00
-3.77443731e-01 -1.61386803e-01 2.23095283e-01 -4.73074049e-01
9.43093181e-01 -3.36487666e-02 1.13498196e-01 6.43734336e-01
1.35134518e+00 3.12871873e-01 -4.51441199e-01 -1.34011641e-01
-1.80871189e-01 -6.23124063e-01 3.08292538e-01 -9.56612945e-01
-9.57334697e-01 1.20912015e+00 7.22482502e-01 5.08725047e-01
1.09087849e+00 -2.01962143e-01 1.28563857e+00 -2.00650215e-01
5.34538150e-01 -1.07587290e+00 7.60128260e-01 5.65947175e-01
1.32696497e+00 -7.15502977e-01 -4.03000265e-01 -5.96143425e-01
-5.61585188e-01 7.39280879e-01 3.90425771e-01 2.33569909e-02
7.31878340e-01 4.41382915e-01 8.95457417e-02 -1.58332929e-01
-7.71712601e-01 4.92467768e-02 1.05669901e-01 7.89811432e-01
2.20571682e-01 1.45609468e-01 -2.05615144e-02 1.02570832e+00
-7.48972058e-01 -2.03541234e-01 7.70081043e-01 7.02087641e-01
-4.16119277e-01 -7.72176862e-01 -6.84856296e-01 9.26893875e-02
-7.74238646e-01 -1.58675462e-01 -1.14898455e+00 5.47742903e-01
3.94459456e-01 1.49504066e+00 -2.48267263e-01 -1.18162417e+00
2.21010312e-01 -2.11510256e-01 3.90574962e-01 -4.16748494e-01
-9.48831379e-01 3.51677448e-01 3.44050497e-01 -6.79555953e-01
4.80634607e-02 -3.69335443e-01 -5.55982411e-01 -5.32082438e-01
-7.75995612e-01 1.93797767e-01 5.29504180e-01 7.07441509e-01
3.57984066e-01 4.48370755e-01 6.63407862e-01 -8.81468117e-01
-5.70259690e-01 -1.11607623e+00 -8.54585171e-01 2.70373166e-01
7.62827754e-01 -7.64394641e-01 -8.87303710e-01 -2.96856195e-01] | [14.153407096862793, 5.741184234619141] |
49675a62-331f-4d12-9a3e-baf25debaf70 | food-recognition-using-fusion-of-classifiers | 1709.04864 | null | http://arxiv.org/abs/1709.04864v1 | http://arxiv.org/pdf/1709.04864v1.pdf | Food Recognition using Fusion of Classifiers based on CNNs | With the arrival of convolutional neural networks, the complex problem of
food recognition has experienced an important improvement in recent years. The
best results have been obtained using methods based on very deep convolutional
neural networks, which show that the deeper the model,the better the
classification accuracy will be obtain. However, very deep neural networks may
suffer from the overfitting problem. In this paper, we propose a combination of
multiple classifiers based on different convolutional models that complement
each other and thus, achieve an improvement in performance. The evaluation of
our approach is done on two public datasets: Food-101 as a dataset with a wide
variety of fine-grained dishes, and Food-11 as a dataset of high-level food
categories, where our approach outperforms the independent CNN models. | ['Marc Bolaños', 'Petia Radeva', 'Eduardo Aguilar'] | 2017-09-14 | null | null | null | null | ['food-recognition'] | ['computer-vision'] | [-1.55200273e-01 -2.94410974e-01 -2.83083439e-01 -4.69928205e-01
-1.60787582e-01 -3.24736118e-01 3.09637219e-01 6.90362692e-01
-3.18701655e-01 6.45809889e-01 3.24429244e-01 2.47477423e-02
-4.20162156e-02 -1.27283680e+00 -9.53878880e-01 -6.26628101e-01
-2.48295978e-01 3.15348879e-02 8.46939906e-03 -4.06584978e-01
-1.50160894e-01 9.43589211e-02 -1.67132306e+00 4.82209861e-01
7.50096142e-01 1.24532509e+00 -5.28787859e-02 5.44629157e-01
-1.73743144e-01 1.02855706e+00 -4.09815758e-01 -3.23573232e-01
3.67714286e-01 -1.01109438e-01 -5.85814118e-01 -1.11332916e-01
2.66025186e-01 -3.02317113e-01 -3.86199094e-02 1.17211688e+00
2.75349319e-01 -1.81043968e-01 5.32611310e-01 -9.54183757e-01
-1.12446272e+00 1.00136757e+00 -3.28987867e-01 3.61135826e-02
3.08128297e-01 -3.42467986e-02 7.07986355e-01 -5.73075056e-01
1.71136614e-02 1.25102758e+00 1.16420650e+00 1.96862683e-01
-1.13140428e+00 -7.53249407e-01 6.80424720e-02 1.20093442e-01
-1.26089263e+00 -1.52126357e-01 6.11181378e-01 -3.89450133e-01
9.81288910e-01 -5.68010062e-02 8.33205998e-01 1.05864143e+00
9.52747315e-02 7.66128957e-01 9.96148288e-01 -5.91234863e-01
3.61692086e-02 8.46604556e-02 7.28138685e-01 7.14453697e-01
6.83495939e-01 3.33253533e-01 -9.59274992e-02 8.88968483e-02
5.50137579e-01 3.91691208e-01 -8.57071877e-02 -1.47965804e-01
-9.56492007e-01 1.35207748e+00 8.89018536e-01 6.78888440e-01
-5.50425708e-01 -8.24998617e-02 4.16257113e-01 2.86664307e-01
6.14114523e-01 2.19458193e-01 -6.44574940e-01 4.90554959e-01
-7.11465597e-01 4.18097615e-01 1.06437325e+00 6.00040197e-01
4.22706604e-01 -8.54339078e-02 -1.08856084e-02 8.15876365e-01
4.47624147e-01 4.52877373e-01 4.90287393e-01 -2.81110823e-01
2.05858663e-01 7.98121572e-01 1.80428118e-01 -1.27943945e+00
-8.97145569e-01 -6.54868245e-01 -1.25544345e+00 1.49371594e-01
7.40994513e-01 -4.60354388e-02 -1.15339136e+00 1.67527425e+00
1.50722384e-01 -1.53246477e-01 3.48240286e-01 8.36088300e-01
1.46659982e+00 6.23748600e-01 4.96396393e-01 3.95969033e-01
1.41698861e+00 -1.22188210e+00 -7.21138656e-01 -1.11227736e-01
5.29051960e-01 -7.30578303e-01 7.02077925e-01 7.17186868e-01
-7.90293813e-01 -1.11765957e+00 -1.48676336e+00 -8.57408997e-03
-9.11486030e-01 1.06157824e-01 1.07356656e+00 8.14445555e-01
-9.30110157e-01 7.54372478e-01 -7.04276979e-01 -5.02266169e-01
4.59631234e-01 4.06972468e-01 -3.62365901e-01 -2.75067508e-01
-1.29063749e+00 7.94329524e-01 9.25080538e-01 1.08201072e-01
-5.41631341e-01 -6.69551551e-01 -1.00182867e+00 2.92590618e-01
-5.65185444e-03 -6.24032199e-01 1.11378002e+00 -1.38438189e+00
-1.19696569e+00 7.39652574e-01 6.71418905e-01 -7.12349236e-01
1.90569237e-01 -2.20373392e-01 -6.21300995e-01 -2.17792183e-01
-1.36142865e-01 8.34753275e-01 3.36914450e-01 -1.02386296e+00
-6.50858879e-01 -2.92212605e-01 4.04037714e-01 -3.34483534e-01
-5.47672689e-01 1.05466969e-01 1.65232509e-01 -5.56801319e-01
-2.70126551e-01 -8.07001770e-01 -3.06718379e-01 -8.89818519e-02
-2.85611331e-01 -3.90839517e-01 1.95687532e-01 -4.85084087e-01
1.01758480e+00 -2.24112511e+00 -1.44892514e-01 -2.18888134e-01
1.27045557e-01 4.68816012e-01 -1.88331053e-01 1.43435210e-01
-2.92492479e-01 7.62859657e-02 1.18825801e-01 6.73920363e-02
-8.96686912e-02 2.98457563e-01 3.12289536e-01 2.27158844e-01
3.06107163e-01 8.59753191e-01 -1.00612497e+00 -2.24020362e-01
3.39120448e-01 5.81701159e-01 -4.83631819e-01 3.90805267e-02
-3.45265120e-01 3.10844965e-02 -3.58145833e-01 7.75445819e-01
9.50952888e-01 -5.66414356e-01 3.55732106e-02 -3.91619086e-01
-2.37923920e-01 -2.84284115e-01 -1.01152432e+00 1.85538733e+00
-1.26696169e-01 1.27262414e-01 -1.83117136e-01 -1.20777965e+00
9.91185188e-01 1.91726908e-01 5.29192090e-01 -7.06772327e-01
4.97383356e-01 1.38957858e-01 3.61348033e-01 -4.72443312e-01
3.33463639e-01 -1.53250741e-02 -1.23798572e-01 -9.06772818e-03
2.58012056e-01 5.33236563e-01 3.98162007e-01 -4.12809879e-01
8.54832292e-01 1.92045793e-01 5.04651666e-01 -6.41042054e-01
3.85236531e-01 3.61057192e-01 3.21818978e-01 6.73312604e-01
-3.85517746e-01 4.13431674e-01 -5.41719310e-02 -1.19593537e+00
-9.09663200e-01 -8.13273728e-01 -2.20006675e-01 8.32876682e-01
-1.90730706e-01 -9.88160148e-02 -6.60449445e-01 -6.17843807e-01
2.91667819e-01 3.06902856e-01 -1.01458526e+00 -2.45941654e-01
-4.43552464e-01 -1.22866881e+00 4.14738357e-01 8.13176095e-01
9.41395879e-01 -1.11129034e+00 -6.48726344e-01 5.32252789e-01
-1.56114146e-01 -9.52710092e-01 5.00876131e-03 5.62000692e-01
-7.20307410e-01 -1.30535483e+00 -8.06013346e-01 -8.28416169e-01
1.46284372e-01 3.93831253e-01 1.60375500e+00 4.67512101e-01
-2.57715911e-01 -2.81983852e-01 -8.09545279e-01 -9.49042022e-01
-6.23795271e-01 3.32468152e-01 -3.08791816e-01 -6.68697283e-02
1.00718093e+00 -2.72204816e-01 -7.50028551e-01 -1.06181622e-01
-8.95322323e-01 4.94267903e-02 4.93592948e-01 8.54843855e-01
3.53420496e-01 2.02755451e-01 6.69576406e-01 -6.22508466e-01
1.24728642e-01 -8.44772935e-01 -3.21190417e-01 1.23520143e-01
-4.59127873e-01 -1.07444143e-02 8.73228788e-01 -4.98075694e-01
-6.86457217e-01 2.70410717e-01 -6.38565481e-01 -1.01413481e-01
-6.21215582e-01 6.57159448e-01 4.89761941e-02 -1.52005460e-02
7.41587222e-01 -1.18914098e-01 -1.65587500e-01 -7.82036662e-01
2.63412863e-01 5.17536223e-01 3.87982994e-01 -2.46152192e-01
3.09057385e-01 9.34357643e-02 -1.04786545e-01 -5.46222448e-01
-1.29169285e+00 -3.92273098e-01 -7.19376624e-01 -1.14197224e-01
1.13142896e+00 -1.01097882e+00 -6.08969688e-01 8.22902083e-01
-8.17036271e-01 -2.35973462e-01 -9.89774168e-02 7.12865114e-01
-8.42893720e-02 1.04646444e-01 -7.21690536e-01 -3.22538346e-01
-5.78136444e-01 -9.87097740e-01 6.26045585e-01 3.57300729e-01
-7.22091794e-02 -1.07489157e+00 5.26724420e-02 3.21006961e-02
6.08592510e-01 8.02217245e-01 1.02183664e+00 -8.08347881e-01
-1.44410089e-01 -2.25521132e-01 -4.26036060e-01 4.51402247e-01
3.96200031e-01 4.69309241e-02 -9.97599304e-01 -2.02915922e-01
-1.40207425e-01 -4.49689865e-01 1.25521576e+00 6.93040788e-01
9.89061356e-01 6.42332882e-02 -1.65363982e-01 5.82490921e-01
1.86819839e+00 2.97750473e-01 5.83068371e-01 6.61410809e-01
5.22643268e-01 3.03765953e-01 3.96781385e-01 1.53409898e-01
5.51967859e-01 3.83114338e-01 7.36996412e-01 -6.48581147e-01
-4.38329756e-01 -1.77670226e-01 1.21958911e-01 8.66063535e-01
-2.00327069e-01 -3.39785874e-01 -6.55715644e-01 5.41072965e-01
-1.80474377e+00 -6.94625556e-01 -4.54511464e-01 1.74689960e+00
5.70875943e-01 3.92557196e-02 5.19116342e-01 6.44742966e-01
4.64445710e-01 -1.74982801e-01 -4.02907193e-01 -4.58984017e-01
-1.37242928e-01 6.02721214e-01 4.35076445e-01 -2.98235714e-02
-1.78500283e+00 5.70556343e-01 7.16621494e+00 4.60155606e-01
-1.00933814e+00 1.69799864e-01 6.50641859e-01 2.44717419e-01
4.08160269e-01 -7.97730386e-01 -7.55708337e-01 4.38534111e-01
1.18261576e+00 3.33318055e-01 1.39212221e-01 8.49475622e-01
-3.41887116e-01 -1.82143748e-02 -1.12603915e+00 7.52897918e-01
5.92047237e-02 -1.10808933e+00 -4.15463299e-02 -1.27025753e-01
6.84203923e-01 1.14757210e-01 -6.04587607e-03 5.35964966e-01
7.27452993e-01 -1.15029681e+00 8.28981161e-01 5.10156929e-01
2.49540374e-01 -9.44088995e-01 1.05209005e+00 3.69174838e-01
-1.30444121e+00 -3.50981414e-01 -6.59498751e-01 -4.76800114e-01
-4.41856325e-01 7.70821154e-01 -3.18581581e-01 8.97623479e-01
1.16410422e+00 9.28353310e-01 -6.23346567e-01 1.13587093e+00
-8.71673133e-03 6.02446437e-01 -3.63439709e-01 -2.83300251e-01
4.45170969e-01 5.15846163e-02 -2.75375456e-01 1.45933723e+00
3.47427368e-01 1.04955949e-01 5.24066389e-01 7.06542492e-01
8.74922797e-02 4.79210287e-01 -5.40218294e-01 -6.62043095e-02
-2.12961182e-01 1.38228595e+00 -9.75197017e-01 -4.86788779e-01
-8.86634469e-01 4.95231539e-01 4.29611117e-01 -1.95979446e-01
-8.15579474e-01 -7.69154951e-02 5.54714739e-01 -3.76871824e-01
3.50377917e-01 1.37552572e-02 -1.31671593e-01 -1.01315105e+00
-3.47558290e-01 -8.45018566e-01 5.27106881e-01 -4.19006437e-01
-1.53466296e+00 5.67987800e-01 -1.75625563e-01 -9.69747126e-01
5.39470129e-02 -8.37311208e-01 -1.51170179e-01 5.94436646e-01
-1.98465168e+00 -1.30384552e+00 -4.77755785e-01 5.50933659e-01
5.17555058e-01 8.68949145e-02 1.42505991e+00 9.09198046e-01
-4.16939288e-01 5.43702483e-01 1.57295898e-01 5.64309359e-01
4.94921356e-01 -1.17805851e+00 1.95716456e-01 4.11279887e-01
9.37967077e-02 1.62380591e-01 5.25857031e-01 -3.63045931e-01
-1.29650462e+00 -1.33806002e+00 8.24716628e-01 -2.00440407e-01
3.40172410e-01 -2.40108192e-01 -8.21697414e-01 4.92445767e-01
3.09381813e-01 1.54270063e-04 9.91539657e-01 2.37377867e-01
-5.04415452e-01 -1.68841973e-01 -1.42675018e+00 -1.30716905e-01
8.72975051e-01 2.24078402e-01 -5.41539609e-01 4.02410239e-01
7.53345251e-01 -1.24547400e-01 -1.23549128e+00 6.84470832e-01
7.79184461e-01 -1.09856081e+00 1.18046284e+00 -6.39730573e-01
7.65670300e-01 1.34593956e-02 -3.41123134e-01 -1.51675618e+00
-9.28058207e-01 6.51005387e-01 -1.51848421e-02 1.06216967e+00
2.86873668e-01 -5.03624797e-01 5.31178296e-01 5.97707257e-02
-1.36409208e-01 -4.93157893e-01 -3.10273170e-01 -7.77298272e-01
4.59737629e-01 -6.58059120e-03 1.03126729e+00 1.14999831e+00
-3.16856384e-01 2.70044506e-01 -3.82917255e-01 -6.27926812e-02
4.53323513e-01 3.08326811e-01 5.51254332e-01 -1.71890593e+00
-2.64969528e-01 -4.25131559e-01 -5.23610890e-01 -6.28248155e-01
-4.75015827e-02 -1.05017865e+00 1.42633647e-01 -1.59076416e+00
5.52219689e-01 -3.07646543e-01 -9.39668655e-01 9.07303751e-01
-1.95931479e-01 5.15211344e-01 1.91135630e-01 -2.17629552e-01
-2.89842308e-01 7.89145567e-03 1.09088349e+00 -5.45127094e-01
1.29544521e-02 -1.64942518e-01 -9.22916293e-01 9.38564420e-01
9.65853930e-01 -4.44869787e-01 -1.13579758e-01 -5.12845516e-01
2.28342742e-01 -2.38030747e-01 2.80872792e-01 -1.46415508e+00
-2.84773260e-01 1.95797145e-01 1.02610087e+00 -5.56575239e-01
1.81536768e-02 -9.01124418e-01 4.28089380e-01 1.05990338e+00
-3.28548193e-01 -1.01438969e-01 3.61125916e-01 3.12742710e-01
-1.99846461e-01 -1.94454238e-01 7.36220717e-01 -5.15341997e-01
-5.53179801e-01 3.24891627e-01 -1.99813649e-01 -6.49743736e-01
7.12091267e-01 3.64654735e-02 -1.91213056e-01 2.57865906e-01
-7.88314581e-01 -7.16494620e-02 1.23185828e-01 7.41784096e-01
2.33414739e-01 -1.57623434e+00 -9.32214856e-01 3.27121913e-01
2.28571326e-01 -3.41407388e-01 3.97728151e-03 4.11181659e-01
-5.24276376e-01 4.48465973e-01 -5.48331439e-01 -5.85629523e-01
-1.00859165e+00 1.04019022e+00 3.78185064e-01 -5.69220424e-01
-7.47173250e-01 6.55834913e-01 5.83663397e-02 -4.06554788e-01
2.77811319e-01 -9.06083941e-01 -8.37855041e-01 3.24559435e-02
8.91254485e-01 1.04856282e-01 2.27086380e-01 -5.14009953e-01
-1.28086746e-01 5.04577518e-01 -2.95218993e-02 8.87164235e-01
1.71103001e+00 2.37145275e-01 8.90121385e-02 3.81001800e-01
1.01346564e+00 -6.12627387e-01 -8.81191850e-01 -1.68960974e-01
-1.12226263e-01 -2.90634215e-01 2.34806180e-01 -1.17702413e+00
-1.46965194e+00 8.43479514e-01 1.07236481e+00 7.76629210e-01
1.19876122e+00 -2.92861789e-01 8.73033106e-01 2.74690181e-01
4.84887272e-01 -6.21731341e-01 -1.37716994e-01 4.63339299e-01
5.69025457e-01 -1.56861317e+00 -1.54461786e-01 -4.58118409e-01
1.86914042e-01 1.49564064e+00 4.06238049e-01 -4.27872241e-01
1.12825513e+00 3.64601821e-01 1.41527578e-01 -2.39232853e-01
-2.98480093e-01 -4.19336557e-01 1.13403767e-01 6.53335452e-01
6.40909910e-01 4.25708294e-01 -3.13251108e-01 9.84502971e-01
-1.89927712e-01 7.64275432e-01 6.58450797e-02 7.44567513e-01
-6.19417191e-01 -1.07581234e+00 -4.42087024e-01 3.92646641e-01
-7.68844128e-01 -5.71840070e-02 -1.94193870e-01 9.53893244e-01
5.87522089e-01 1.34331417e+00 9.34929177e-02 -2.98883408e-01
4.48056847e-01 1.37732290e-02 8.11656773e-01 -4.35396522e-01
-1.08410704e+00 -8.55726078e-02 2.02481806e-01 -3.78259480e-01
-1.04477453e+00 -4.52391177e-01 -1.10032737e+00 -5.50749958e-01
-3.14925760e-01 -6.16678856e-02 6.26457870e-01 7.76307464e-01
-1.53847769e-01 7.39926577e-01 4.67708617e-01 -8.50536883e-01
-5.81772447e-01 -1.18363440e+00 -6.54728174e-01 6.55388772e-01
3.13467324e-01 -7.08623171e-01 -1.28202513e-01 -6.74929544e-02] | [11.563477516174316, 4.38442850112915] |
34d444ee-7c65-4bf1-b563-0089657d2e5b | ds-1000-a-natural-and-reliable-benchmark-for | 2211.11501 | null | https://arxiv.org/abs/2211.11501v1 | https://arxiv.org/pdf/2211.11501v1.pdf | DS-1000: A Natural and Reliable Benchmark for Data Science Code Generation | We introduce DS-1000, a code generation benchmark with a thousand data science problems spanning seven Python libraries, such as NumPy and Pandas. Compared to prior works, DS-1000 incorporates three core features. First, our problems reflect diverse, realistic, and practical use cases since we collected them from StackOverflow. Second, our automatic evaluation is highly specific (reliable) -- across all Codex-002-predicted solutions that our evaluation accept, only 1.8% of them are incorrect; we achieve this with multi-criteria metrics, checking both functional correctness by running test cases and surface-form constraints by restricting API usages or keywords. Finally, we proactively defend against memorization by slightly modifying our problems to be different from the original StackOverflow source; consequently, models cannot answer them correctly by memorizing the solutions from pre-training. The current best public system (Codex-002) achieves 43.3% accuracy, leaving ample room for improvement. We release our benchmark at https://ds1000-code-gen.github.io. | ['Tao Yu', 'Sida Wang', 'Daniel Fried', 'Scott Wen-tau Yih', 'Luke Zettlemoyer', 'Ruiqi Zhong', 'Tianyi Zhang', 'Yiming Wang', 'Chengxi Li', 'Yuhang Lai'] | 2022-11-18 | null | null | null | null | ['memorization'] | ['natural-language-processing'] | [-1.50468796e-01 -9.30276886e-02 -2.55253911e-01 -4.45475340e-01
-1.13721430e+00 -1.02729404e+00 1.66427448e-01 5.66093884e-02
-6.45187348e-02 8.04464579e-01 1.10646822e-01 -8.66954088e-01
1.60687089e-01 -8.25679779e-01 -9.67224538e-01 -1.30608037e-01
-1.56599477e-01 6.25060648e-02 4.81379509e-01 -6.55146986e-02
6.17460787e-01 -1.25140309e-01 -1.65530419e+00 6.43932760e-01
1.24348390e+00 4.22514439e-01 -2.53102761e-02 8.91848505e-01
-1.21041521e-01 8.44783485e-01 -9.26842749e-01 -6.77076936e-01
2.80574918e-01 1.69073623e-02 -1.20174396e+00 -5.27459383e-01
6.22167468e-01 -2.61416644e-01 7.10642338e-02 1.25475633e+00
3.99841785e-01 -3.66346031e-01 6.53033257e-02 -1.48192537e+00
-6.32064760e-01 9.78467166e-01 -5.10339797e-01 2.04974070e-01
6.36050224e-01 8.02200139e-01 1.35583282e+00 -6.13111258e-01
7.44172573e-01 9.00109649e-01 9.90149558e-01 7.26068318e-01
-1.45845175e+00 -6.22402847e-01 -6.44035041e-02 5.20465374e-02
-1.18531358e+00 -4.76272553e-01 7.04988763e-02 -7.31595755e-01
1.45406437e+00 9.51966643e-01 2.07384944e-01 1.16796243e+00
7.66657442e-02 5.73815286e-01 1.00786698e+00 -8.53907913e-02
2.05163345e-01 -3.41682471e-02 6.61412001e-01 6.80863619e-01
4.42451686e-01 -2.22653523e-01 -8.38601142e-02 -9.49007571e-01
-6.90366551e-02 -3.07080328e-01 -2.88990527e-01 7.85883702e-03
-1.28271174e+00 4.52717245e-01 7.33800456e-02 2.77593844e-02
1.59029692e-01 2.71068782e-01 5.93465090e-01 3.43605757e-01
-1.08837456e-01 1.00540292e+00 -8.98589134e-01 -6.00332558e-01
-8.62583578e-01 8.49091411e-01 1.39693260e+00 1.24717593e+00
9.24899340e-01 -8.51050690e-02 -2.24766403e-01 5.54135442e-01
1.07659519e-01 2.67106444e-01 7.58915126e-01 -1.18841267e+00
8.17972481e-01 8.18551123e-01 6.04493357e-02 -5.75357199e-01
-3.24625582e-01 -3.24151248e-01 4.66969050e-02 1.74315214e-01
4.78605330e-01 -3.23006928e-01 -4.52293903e-01 1.58635914e+00
2.82500181e-02 -1.15829436e-02 -1.06443301e-01 6.58052742e-01
8.53031218e-01 4.88464296e-01 5.15954569e-02 2.68048674e-01
1.07869685e+00 -9.85063195e-01 -3.03137247e-02 -4.34778243e-01
1.07524264e+00 -9.72135723e-01 1.49530876e+00 5.35777032e-01
-1.22871482e+00 -1.48071408e-01 -9.78803337e-01 1.06511444e-01
-1.15710512e-01 -1.49812981e-01 9.09034073e-01 6.61966503e-01
-1.25852561e+00 5.75739145e-01 -7.62916744e-01 -3.14954549e-01
3.37867849e-02 1.58230603e-01 -1.64604872e-01 8.54849294e-02
-7.27976561e-01 5.78805029e-01 3.47140402e-01 -4.55183625e-01
-8.07308614e-01 -1.10077620e+00 -6.07289553e-01 2.47597769e-01
3.80168378e-01 -5.72915554e-01 1.68703234e+00 -5.49407005e-01
-9.59104002e-01 1.01975906e+00 -2.55639344e-01 -1.72321945e-01
7.44044065e-01 -2.10611701e-01 -2.65905827e-01 -4.94746804e-01
3.56334567e-01 1.76459178e-01 2.46952370e-01 -1.09635305e+00
-5.41385531e-01 9.99661759e-02 4.59499270e-01 -5.26589513e-01
-1.60384342e-01 3.90492767e-01 -5.23164928e-01 -3.68530422e-01
-2.54294038e-01 -9.86992002e-01 -1.48669764e-01 -2.45353147e-01
-7.89421022e-01 -2.30215281e-01 2.27129743e-01 -7.02643573e-01
1.53447008e+00 -2.13501525e+00 -9.61581394e-02 2.13415205e-01
4.65793431e-01 1.99443445e-01 -4.35952216e-01 2.76664495e-01
-2.45640025e-01 6.13043904e-01 -4.98009741e-01 -1.61962345e-01
3.12184751e-01 -4.70370464e-02 -5.54232419e-01 2.46434540e-01
2.99816638e-01 7.26137102e-01 -8.65019560e-01 -2.81545848e-01
-2.81178296e-01 -1.45998105e-01 -1.34639263e+00 3.61420840e-01
-8.37202489e-01 -1.73762500e-01 -4.84923095e-01 9.01916564e-01
8.35484087e-01 -5.15900433e-01 2.95120716e-01 3.68108511e-01
-3.49291503e-01 8.69831443e-01 -1.10583782e+00 1.47011220e+00
-5.12575090e-01 4.61327851e-01 -1.14776149e-01 -3.38017613e-01
6.47995293e-01 -1.72123626e-01 1.75741047e-01 -7.06640124e-01
-4.13550675e-01 3.81139606e-01 1.26125515e-01 -6.85924053e-01
6.11698568e-01 7.34962881e-01 -3.54376316e-01 7.11886287e-01
-2.30952531e-01 -7.79113965e-04 5.40163934e-01 2.74597138e-01
1.96611893e+00 2.74145603e-01 1.22815482e-02 -3.57702285e-01
4.55016702e-01 1.40734151e-01 9.73621547e-01 1.04534209e+00
-1.39742251e-02 8.33988369e-01 1.14465833e+00 -3.13060611e-01
-1.09467590e+00 -1.04964411e+00 -2.62568414e-01 1.42159796e+00
-2.84568250e-01 -1.21331239e+00 -1.00927925e+00 -8.11503589e-01
2.08204404e-01 1.08786273e+00 -5.11498570e-01 1.47445649e-01
-8.50065947e-01 -9.18035686e-01 9.52033937e-01 3.62673938e-01
1.36578918e-01 -9.93734181e-01 -3.42799366e-01 3.20347339e-01
-2.44922206e-01 -8.22447240e-01 -4.11558002e-01 -3.97261009e-02
-4.71564949e-01 -1.61090052e+00 -1.66394934e-01 -3.00711244e-01
4.91105139e-01 2.04672441e-02 1.83050001e+00 8.89273167e-01
-5.73207259e-01 1.74709097e-01 -2.04424009e-01 1.82731733e-01
-7.84916759e-01 3.98941785e-01 -4.79964763e-01 -8.07678342e-01
4.58824396e-01 -6.38375938e-01 -4.20206696e-01 4.93760377e-01
-6.59149468e-01 -1.83373883e-01 1.39333054e-01 5.38316488e-01
1.70609236e-01 -4.90312099e-01 2.52962232e-01 -1.25464296e+00
5.72363913e-01 -1.13023245e+00 -1.01060688e+00 3.92528832e-01
-4.86245990e-01 -7.48973386e-03 8.54923308e-01 -1.06867172e-01
-8.06072235e-01 -4.18276280e-01 -4.07372117e-01 1.16452374e-01
-4.33324929e-03 2.95923889e-01 -4.87622283e-02 4.38367203e-02
1.25948811e+00 1.42875075e-01 -2.28776708e-01 -6.67536795e-01
9.72154066e-02 4.21503097e-01 5.23636818e-01 -1.47938752e+00
8.05004418e-01 -1.49983689e-01 -6.24952674e-01 -2.58186013e-01
-5.53492963e-01 -1.40358090e-01 1.04329966e-01 3.33079308e-01
2.47683048e-01 -6.64011478e-01 -1.24760354e+00 5.40287435e-01
-1.27432621e+00 -5.36565959e-01 1.94838598e-01 -8.87465775e-02
-2.64312357e-01 5.10352194e-01 -7.90921569e-01 -3.82010907e-01
-5.38700283e-01 -1.62635148e+00 8.80244792e-01 -6.82090875e-03
-6.30150914e-01 -6.21599436e-01 1.71426669e-01 3.32828134e-01
7.69992769e-01 2.41685182e-01 1.10791552e+00 -7.39705205e-01
-7.63561487e-01 1.09190747e-01 -3.11515629e-01 2.64976114e-01
-2.36091301e-01 8.30087185e-01 -9.60983455e-01 -2.91877449e-01
-2.88311362e-01 -4.15248036e-01 5.55972993e-01 -3.30109328e-01
1.70963061e+00 -6.07885957e-01 -2.60319859e-01 9.39379752e-01
1.42092395e+00 -4.03631687e-01 6.12816751e-01 8.53392184e-01
5.35772502e-01 2.22079560e-01 2.21366912e-01 6.98165834e-01
6.89095318e-01 3.51411819e-01 6.85767531e-01 4.68077302e-01
1.39215171e-01 3.77114490e-02 5.89404762e-01 6.36842072e-01
2.18414113e-01 -1.12282567e-01 -1.44502699e+00 5.70604801e-01
-1.81160641e+00 -7.46752679e-01 -6.08017564e-01 2.29849672e+00
1.41198623e+00 3.71428549e-01 2.53333330e-01 -1.90575778e-01
5.09073138e-01 -1.39939025e-01 -5.86895406e-01 -4.22742993e-01
2.03919429e-02 2.40365237e-01 3.84832978e-01 3.77152890e-01
-8.53402197e-01 7.08569109e-01 6.40303898e+00 3.73835474e-01
-1.18847394e+00 1.10997660e-02 3.89863968e-01 -2.29112908e-01
-7.41661489e-01 1.45368516e-01 -1.12654984e+00 8.87091100e-01
9.93053377e-01 -5.78408897e-01 6.62214041e-01 1.26079893e+00
-3.71982872e-01 -5.97651768e-03 -1.45465696e+00 2.47507542e-01
-2.48947009e-01 -1.37737167e+00 -3.92194211e-01 -1.41265705e-01
7.90912330e-01 5.49587071e-01 5.21881394e-02 7.91361451e-01
7.72534430e-01 -9.86065686e-01 1.00133073e+00 3.89675260e-01
6.13748789e-01 -2.17540130e-01 4.75745231e-01 2.30487928e-01
-7.16306567e-01 -5.65542653e-02 -3.09157580e-01 -1.43156007e-01
-3.59723479e-01 6.66368842e-01 -8.53756666e-01 2.40754515e-01
9.85597491e-01 4.45772201e-01 -1.22507179e+00 1.46505952e+00
-1.56207994e-01 1.06652379e+00 -5.06058753e-01 -3.45752239e-02
1.04991375e-02 4.12377983e-01 4.74385202e-01 1.55634284e+00
2.57266194e-01 -4.03348804e-01 1.97322499e-02 1.40105081e+00
-2.01554403e-01 -5.89811653e-02 -2.25162759e-01 1.77751482e-01
7.44485140e-01 1.06544030e+00 -8.93120021e-02 -3.40353310e-01
-5.52481294e-01 3.87308836e-01 6.23162925e-01 3.26457918e-01
-1.07724428e+00 -4.37395751e-01 1.17373395e+00 1.83024585e-01
5.89528009e-02 -3.10801826e-02 -6.36185586e-01 -1.50432730e+00
6.43433511e-01 -1.37355936e+00 5.11620581e-01 -5.87003350e-01
-1.44436061e+00 5.76622367e-01 -2.24457249e-01 -8.87197137e-01
-2.58669287e-01 -5.92553377e-01 -1.02213883e+00 9.56411242e-01
-1.46693897e+00 -6.68630004e-01 -3.45757931e-01 1.02870472e-01
8.68154913e-02 9.07284692e-02 9.21796679e-01 6.08911514e-01
-7.63561845e-01 1.11319780e+00 1.09322652e-01 5.34571968e-02
8.60501230e-01 -1.48735380e+00 1.19757855e+00 8.40661526e-01
-7.11641192e-01 1.29131556e+00 7.47788906e-01 -5.97165346e-01
-1.63410997e+00 -1.14346182e+00 8.09843898e-01 -8.21933806e-01
1.05888188e+00 -4.12476003e-01 -1.42030561e+00 1.01279783e+00
-9.49288439e-03 1.87414154e-01 5.89825451e-01 2.91026860e-01
-8.89634013e-01 9.20326933e-02 -1.10133934e+00 6.31048381e-01
1.07503223e+00 -2.77625144e-01 -5.12300789e-01 4.93137240e-01
7.44597375e-01 -7.31700957e-01 -9.03478801e-01 1.65587500e-01
4.12617028e-01 -1.25978768e+00 8.36848795e-01 -7.99823821e-01
8.16422939e-01 -3.92621636e-01 -3.18205625e-01 -1.04251921e+00
-1.55673981e-01 -8.21494460e-01 -2.03982949e-01 1.46566486e+00
7.14011133e-01 -8.28404784e-01 6.61551416e-01 1.07620871e+00
-2.56568611e-01 -5.88582814e-01 -5.15730917e-01 -6.89317346e-01
3.18283290e-01 -5.10976255e-01 1.13807964e+00 1.29201162e+00
1.38548046e-01 -2.22043574e-01 1.08806025e-02 1.07166767e-01
6.97835326e-01 9.08862576e-02 1.08524227e+00 -8.41469467e-01
-8.74062836e-01 -7.02883780e-01 -8.04672092e-02 -7.63312399e-01
3.40231389e-01 -1.22125351e+00 -1.49872378e-01 -1.05887008e+00
3.53905380e-01 -7.26893067e-01 7.93121904e-02 9.73925650e-01
-5.37284017e-01 -2.25897804e-02 2.26133525e-01 2.35213682e-01
-8.23995948e-01 9.23905671e-02 6.18333757e-01 4.59892415e-02
1.58432171e-01 1.08341798e-01 -8.96622956e-01 6.62428439e-01
8.33405197e-01 -5.59806764e-01 2.96347976e-01 -7.94183850e-01
8.10460687e-01 1.34204790e-01 4.54050601e-01 -1.02032495e+00
2.06686538e-02 -2.77527124e-01 -1.77482530e-01 -2.23195776e-01
-2.93436885e-01 -3.24422002e-01 2.84816355e-01 4.16665465e-01
-4.36965227e-01 4.00830954e-01 4.00632888e-01 -5.84264025e-02
1.83448657e-01 -5.89519799e-01 5.62152803e-01 -3.33755285e-01
-7.91086912e-01 -1.41305730e-01 -3.08872283e-01 6.59811139e-01
8.53136063e-01 1.68964475e-01 -1.21131325e+00 2.89259076e-01
-2.91242182e-01 5.62227249e-01 9.64428067e-01 6.10499084e-01
4.07185070e-02 -8.54501486e-01 -8.50374460e-01 1.43834040e-01
5.00545323e-01 -4.71250936e-02 -5.76949902e-02 4.77105588e-01
-8.90791655e-01 9.09385309e-02 -5.20062819e-03 -5.76985121e-01
-1.10809147e+00 3.07641327e-01 3.57473642e-01 -7.28904158e-02
-5.15315711e-01 7.77621746e-01 -1.10952131e-01 -1.07150888e+00
9.70206037e-02 -5.97586811e-01 4.22010422e-01 -6.47477269e-01
6.76079869e-01 5.34883618e-01 3.59926909e-01 1.11377314e-01
-5.45298219e-01 7.94067308e-02 -1.54088244e-01 4.98088270e-01
1.48557973e+00 5.76387167e-01 -6.04294181e-01 1.26628309e-01
1.24944031e+00 4.22417074e-01 -9.41925466e-01 -1.66744664e-01
3.00267965e-01 -7.21489608e-01 -6.19119942e-01 -1.14072824e+00
-8.84465694e-01 5.02777696e-01 -1.49130449e-02 3.59363407e-01
6.09624088e-01 -1.40472323e-01 6.09230757e-01 6.88767552e-01
4.30412114e-01 -5.57263315e-01 -2.67758250e-01 7.58397102e-01
6.99495316e-01 -1.07215703e+00 4.45623361e-02 -2.67176449e-01
-1.18194677e-01 1.08316481e+00 1.33746552e+00 -1.45284692e-02
2.17310444e-01 5.61010301e-01 -2.18896315e-01 7.10428283e-02
-1.28112662e+00 4.36310709e-01 -6.11468591e-02 2.96436191e-01
8.53448570e-01 1.15122236e-01 -3.56636941e-01 8.96510422e-01
-5.49851835e-01 3.78150120e-02 9.44220483e-01 1.16601396e+00
-1.68357611e-01 -1.15345001e+00 -5.35451651e-01 6.66786134e-01
-7.73785651e-01 -3.39334846e-01 -2.14544728e-01 6.12464964e-01
-2.30662271e-01 5.19464135e-01 -8.05613920e-02 -3.61054331e-01
3.66159886e-01 -3.88853662e-02 2.35933617e-01 -8.08896899e-01
-1.19104934e+00 -7.07762420e-01 4.63957131e-01 -9.01772976e-01
5.25265753e-01 -6.33541107e-01 -1.15863168e+00 -1.03757453e+00
1.77535781e-04 2.28767589e-01 6.67310119e-01 4.29777503e-01
6.94828331e-01 5.33060551e-01 4.32102025e-01 -5.35519123e-01
-1.01640284e+00 -6.33046329e-01 1.66112185e-01 4.50668603e-01
1.65229321e-01 -2.33102724e-01 -5.65234482e-01 -6.96211681e-02] | [7.798279285430908, 7.718754291534424] |
8af0cdd3-7400-46dc-a50b-164cebf49851 | algorithms-for-weighted-pushdown-automata | 2210.06884 | null | https://arxiv.org/abs/2210.06884v3 | https://arxiv.org/pdf/2210.06884v3.pdf | Algorithms for Weighted Pushdown Automata | Weighted pushdown automata (WPDAs) are at the core of many natural language processing tasks, like syntax-based statistical machine translation and transition-based dependency parsing. As most existing dynamic programming algorithms are designed for context-free grammars (CFGs), algorithms for PDAs often resort to a PDA-to-CFG conversion. In this paper, we develop novel algorithms that operate directly on WPDAs. Our algorithms are inspired by Lang's algorithm, but use a more general definition of pushdown automaton and either reduce the space requirements by a factor of $|\Gamma|$ (the size of the stack alphabet) or reduce the runtime by a factor of more than $|Q|$ (the number of states). When run on the same class of PDAs as Lang's algorithm, our algorithm is both more space-efficient by a factor of $|\Gamma|$ and more time-efficient by a factor of $|Q| \cdot |\Gamma|$. | ['David Chiang', 'Ryan Cotterell', 'Tim Vieira', 'Brian DuSell', 'Alexandra Butoi'] | 2022-10-13 | null | null | null | null | ['dependency-parsing', 'transition-based-dependency-parsing'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.98631334e-01 1.34542659e-01 -6.77844062e-02 -1.49845704e-01
-7.67450690e-01 -7.31527209e-01 5.78015983e-01 7.08015501e-01
-6.85222924e-01 2.71851927e-01 -2.10102364e-01 -1.17706001e+00
6.54861480e-02 -1.37648737e+00 -5.51287830e-01 -3.59658748e-01
-3.44230831e-01 6.70715630e-01 9.59824860e-01 -4.67589378e-01
6.09663278e-02 3.83253545e-01 -1.66364300e+00 -9.66572836e-02
5.25671065e-01 4.90509093e-01 1.35396495e-01 1.06824434e+00
-1.02106297e+00 2.93404847e-01 -1.18110046e-01 -3.34330559e-01
2.71459401e-01 -6.08505726e-01 -9.75315690e-01 -2.93037236e-01
-1.93251707e-02 6.31026253e-02 -1.91034675e-01 1.20308793e+00
1.72255859e-02 2.56563187e-01 -6.75979406e-02 -1.13039422e+00
3.59772518e-02 9.86892164e-01 -2.24544689e-01 2.67788440e-01
3.74643415e-01 -1.81980655e-01 1.26114726e+00 -4.40278053e-01
6.75458431e-01 1.29327953e+00 1.61828205e-01 5.63999712e-01
-1.33844995e+00 -2.53960907e-01 5.81639349e-01 -9.99773368e-02
-1.06290972e+00 -1.87224343e-01 4.89726216e-01 7.27178121e-04
1.26817548e+00 5.84432542e-01 4.52427447e-01 4.30265486e-01
1.53130507e-02 5.26627004e-01 1.22080564e+00 -1.01556277e+00
3.69711757e-01 -4.53817070e-01 5.25342107e-01 1.01804805e+00
3.38244915e-01 -7.81781748e-02 -1.68506861e-01 -4.87515628e-01
5.13736010e-01 -1.41188651e-01 4.03730452e-01 -3.91361862e-01
-1.12944221e+00 1.06352770e+00 -5.19593596e-01 5.06859899e-01
1.86134636e-01 2.57725835e-01 4.78847444e-01 3.87754649e-01
-2.38641687e-02 1.73116699e-01 -4.59840864e-01 -5.00586748e-01
-5.76463521e-01 3.59457165e-01 1.16430223e+00 1.02622962e+00
7.70300806e-01 -2.51902729e-01 1.86700344e-01 4.69381332e-01
1.45522937e-01 9.41767931e-01 1.29578829e-01 -7.30253339e-01
6.87449932e-01 7.77455986e-01 1.65390909e-01 -3.73453736e-01
-6.38698518e-01 1.59059882e-01 -4.03936177e-01 1.21143073e-01
8.52438867e-01 -3.41254957e-02 -8.59367549e-01 2.05008936e+00
4.77152258e-01 -5.25067389e-01 -1.56565487e-01 4.06782717e-01
-1.32874176e-01 1.08245122e+00 6.10039979e-02 -5.22113979e-01
1.72577524e+00 -8.43536675e-01 -3.92582744e-01 -5.32032609e-01
1.11361241e+00 -5.30065835e-01 1.35383415e+00 2.14012310e-01
-1.32608020e+00 -6.22498505e-02 -9.94335949e-01 6.27928239e-04
-2.36054271e-01 -6.17092431e-01 8.84036839e-01 1.05079997e+00
-1.01693439e+00 3.64160001e-01 -1.28010035e+00 -2.18716025e-01
-3.25537890e-01 5.61860383e-01 -2.96941757e-01 -1.12227149e-01
-8.42251241e-01 7.20963776e-01 2.87906706e-01 -3.27887028e-01
-4.98406380e-01 -2.17619359e-01 -9.02759433e-01 2.20810905e-01
6.70166910e-01 -3.31720799e-01 1.47898281e+00 -5.48263967e-01
-1.89497399e+00 7.55132020e-01 -3.43797326e-01 -4.83240753e-01
2.08983988e-01 2.59194285e-01 -3.28533828e-01 -6.13991171e-02
-1.55966341e-01 -1.68524254e-02 3.95189226e-01 -4.18514311e-01
-7.79883504e-01 -5.81427217e-01 4.24964964e-01 -1.15320802e-01
1.33898810e-01 7.39220440e-01 -3.93251866e-01 -2.27640659e-01
3.95231903e-01 -1.51091433e+00 -6.78591073e-01 -4.31524396e-01
-5.40762097e-02 -3.81314486e-01 1.85939029e-01 -2.38496676e-01
1.48824239e+00 -2.06809998e+00 2.07001477e-01 3.64678741e-01
2.98759311e-01 3.80640835e-01 -2.26468757e-01 7.28474140e-01
2.41971910e-01 1.03595346e-01 -7.12754786e-01 2.74521597e-02
2.06103727e-01 4.21580911e-01 -1.64193928e-01 2.74838120e-01
-2.06070971e-02 5.63605487e-01 -9.39131856e-01 -3.89336884e-01
4.57209582e-03 -2.77058154e-01 -6.46019161e-01 9.30409133e-02
-6.42966509e-01 -1.42934442e-01 -4.72307742e-01 2.62466282e-01
6.14409089e-01 7.01931491e-02 9.17961538e-01 7.83171892e-01
-6.15888119e-01 6.36739612e-01 -1.41294301e+00 1.82793713e+00
-6.82145596e-01 1.27145603e-01 1.86371133e-01 -7.43342400e-01
9.60488737e-01 -2.29062196e-02 1.91764012e-01 -7.38228381e-01
2.55331323e-02 5.04415751e-01 1.48554251e-01 5.16648181e-02
4.57775682e-01 -3.43055010e-01 -7.67220557e-01 1.12352848e+00
-3.12342495e-01 4.63934522e-03 6.32728696e-01 1.45382434e-01
1.63714099e+00 1.91360041e-02 2.87367821e-01 -3.61981124e-01
7.17523813e-01 4.26245891e-02 7.57215858e-01 7.83552945e-01
-1.05019003e-01 -4.72737104e-02 1.03254151e+00 -5.51850975e-01
-1.09591758e+00 -1.00849044e+00 2.14843124e-01 1.73148727e+00
-3.78854363e-03 -8.40478897e-01 -9.88983154e-01 -6.35450125e-01
-4.70399350e-01 6.83438361e-01 -3.10900331e-01 4.90000136e-02
-1.16018283e+00 -5.74037135e-01 7.90900946e-01 6.67661488e-01
2.57712871e-01 -9.02982891e-01 -9.89090204e-01 5.73291659e-01
2.04282656e-01 -1.07666230e+00 -3.59318733e-01 4.11442220e-01
-1.12661183e+00 -8.33950281e-01 -2.56762922e-01 -7.84103751e-01
6.84839070e-01 -8.30970779e-02 8.74083757e-01 1.21642061e-01
1.44748524e-01 -2.04319209e-01 -3.98902178e-01 -4.36321825e-01
-8.40066433e-01 2.71727622e-01 1.05603419e-01 -4.48874652e-01
4.21696365e-01 -5.14131069e-01 -9.71203148e-02 6.52217194e-02
-1.01765728e+00 1.27167642e-01 3.13509077e-01 6.51488543e-01
5.46731651e-01 -2.13867262e-01 -1.95499927e-01 -1.37621415e+00
4.29214358e-01 1.36638865e-01 -1.31770432e+00 1.70130402e-01
-6.80384874e-01 7.81979680e-01 1.01513004e+00 -3.79469275e-01
-8.50022078e-01 3.03590477e-01 -4.00114119e-01 1.28625691e-01
7.18792528e-02 4.98679489e-01 -2.45446756e-01 1.90464526e-01
4.44950908e-01 2.68601209e-01 -5.55208363e-02 -4.46525097e-01
5.43889165e-01 2.77428925e-01 2.85744935e-01 -8.24571192e-01
6.90976620e-01 3.47080171e-01 4.89373356e-01 -3.81126285e-01
-6.70860708e-02 -7.71486685e-02 -3.25579047e-01 4.69725639e-01
4.59906310e-01 -2.03249633e-01 -6.11321926e-01 4.53867406e-01
-8.75589788e-01 -4.88930076e-01 -1.44484967e-01 3.13927621e-01
-4.46114361e-01 5.63407183e-01 -5.63581765e-01 -8.82495165e-01
-3.88868302e-01 -1.15292215e+00 7.21123695e-01 -3.08019444e-02
-8.72516930e-02 -6.01819098e-01 5.27813733e-01 -9.43624750e-02
2.68165380e-01 1.16172852e-02 1.61730278e+00 -7.76058435e-01
-2.20233977e-01 -3.14758420e-01 -6.29680604e-02 8.06728303e-02
-3.30441773e-01 4.97447960e-02 -2.85537511e-01 -1.62699401e-01
-2.77007461e-01 4.27172929e-01 3.97886515e-01 1.88566223e-02
7.05764472e-01 -2.32676804e-01 -2.99208522e-01 3.95399556e-02
1.42964756e+00 5.04801154e-01 5.91638267e-01 2.43993551e-01
4.20555413e-01 4.67526942e-01 3.50117892e-01 1.50964901e-01
4.52935427e-01 6.32772624e-01 2.16371939e-01 3.33589882e-01
1.05630279e-01 -3.73457670e-01 6.24871373e-01 1.07575011e+00
-2.41729859e-02 1.80643108e-02 -1.38115537e+00 6.91581607e-01
-1.73984015e+00 -3.14228594e-01 -3.19001436e-01 2.35877132e+00
7.73435712e-01 4.75776672e-01 4.57348078e-01 3.84663880e-01
8.46013665e-01 1.33386970e-01 2.44467109e-02 -1.57807434e+00
2.83787489e-01 9.71763968e-01 9.02393341e-01 6.52551353e-01
-6.20345771e-01 1.16710508e+00 5.94150162e+00 4.47894782e-01
-9.09338355e-01 2.62579829e-01 -3.82498913e-02 -1.24775000e-01
-3.89535904e-01 6.65707767e-01 -6.95689082e-01 5.04152656e-01
1.54551589e+00 -2.48418152e-01 5.34871042e-01 8.81430447e-01
-2.93973029e-01 -3.55296731e-01 -1.17198646e+00 6.91416383e-01
-3.63091379e-01 -1.02382624e+00 -3.57708991e-01 2.51848161e-01
3.17609787e-01 -6.50711507e-02 -4.28761542e-01 4.01851088e-01
6.30525529e-01 -4.54579353e-01 7.74423599e-01 -3.24294627e-01
7.80909121e-01 -8.58658969e-01 3.97020578e-01 4.64443594e-01
-1.32254970e+00 -2.13264987e-01 -2.33339921e-01 -3.82370353e-01
4.09989268e-01 5.10587692e-01 -2.99742728e-01 2.63184607e-01
4.16485906e-01 -6.14287198e-01 -1.58338204e-01 4.43867445e-01
-4.10614312e-01 6.96960330e-01 -7.60807335e-01 -4.88934338e-01
3.97207201e-01 -4.26318198e-01 6.11786902e-01 1.36896837e+00
2.17229411e-01 3.89743328e-01 2.36022979e-01 2.12254107e-01
-1.12857111e-01 1.75284714e-01 -2.48277053e-01 -3.43710631e-01
5.06264865e-01 8.84294629e-01 -1.02640724e+00 -4.56786931e-01
-5.74712098e-01 5.02385139e-01 2.37246394e-01 -3.79568011e-01
-7.61211514e-01 -9.09284055e-01 5.06226718e-01 3.27320129e-01
4.49811548e-01 -8.47658992e-01 -1.93685088e-02 -8.90356898e-01
1.58366725e-01 -7.83890307e-01 5.17087996e-01 -6.34278432e-02
-2.97596455e-01 6.64853573e-01 2.02978194e-01 -5.36916137e-01
-3.55138332e-01 -5.99991858e-01 -3.98894668e-01 7.00019896e-01
-1.15648711e+00 -7.63969898e-01 3.22114587e-01 4.61545259e-01
2.40353763e-01 2.60804176e-01 1.30506575e+00 3.75796445e-02
-5.98274171e-01 4.62156266e-01 1.23017542e-02 3.79198678e-02
5.12936004e-02 -1.13577712e+00 9.53014910e-01 1.41827106e+00
2.26500034e-01 7.98536777e-01 7.40512073e-01 -2.68789113e-01
-2.00422835e+00 -7.08749771e-01 1.31680763e+00 -1.22236952e-01
7.90792763e-01 -7.18467474e-01 -7.38051534e-01 6.21151626e-01
-8.25069994e-02 8.65250379e-02 5.00259340e-01 5.17194331e-01
-6.89205468e-01 -3.18459630e-01 -9.41780448e-01 7.41604388e-01
1.35499549e+00 -5.75929761e-01 -6.57575369e-01 -4.94146869e-02
7.48253047e-01 -6.26878023e-01 -8.00399721e-01 -5.85046262e-02
5.99585652e-01 -6.28968298e-01 1.97417989e-01 -6.00861430e-01
-1.25126392e-01 -4.64581966e-01 -2.13153154e-01 -8.98775876e-01
-3.48258376e-01 -9.96377051e-01 -2.70038307e-01 8.67752790e-01
6.48588002e-01 -8.40806603e-01 5.07756770e-01 6.30337000e-01
-1.78281218e-01 -6.08664811e-01 -1.14223945e+00 -9.10555422e-01
3.32210273e-01 -9.36941028e-01 8.37075114e-01 3.96900415e-01
5.34929514e-01 2.85045117e-01 1.11515827e-01 3.61523293e-02
3.45989615e-01 4.63272601e-01 5.42421281e-01 -1.20520973e+00
-6.15780354e-01 -4.54603404e-01 -2.42178097e-01 -8.35780740e-01
-8.03616568e-02 -8.59249234e-01 1.77814886e-01 -1.30570900e+00
1.42545864e-01 -7.86541164e-01 -2.82428265e-01 7.52096951e-01
2.19608828e-01 -1.61704615e-01 4.04703051e-01 -8.18436965e-02
-5.27051628e-01 -3.32822323e-01 7.07725108e-01 1.86931610e-01
-3.45212430e-01 5.69542497e-02 -5.78424692e-01 5.22497118e-01
6.41537189e-01 -8.53662133e-01 -1.99509963e-01 -6.74645424e-01
9.96783316e-01 2.35451952e-01 -2.65579164e-01 -7.19911456e-01
6.32553101e-02 -5.61609089e-01 -8.42894793e-01 -1.57484770e-01
-2.21483745e-02 -3.41636389e-01 1.93286955e-01 6.87426448e-01
-2.28250712e-01 5.54495513e-01 3.07256699e-01 5.01378655e-01
1.51816010e-01 -2.74725646e-01 8.03626299e-01 -2.17856064e-01
-3.63858700e-01 2.23510772e-01 -8.87814283e-01 -8.31264630e-03
9.68387783e-01 -1.00584075e-01 -2.09009349e-01 9.34653580e-02
-6.71183944e-01 -1.22394457e-01 5.46186447e-01 3.59737337e-01
1.84065253e-01 -8.33222151e-01 -2.29229033e-01 3.11799794e-01
1.16469353e-01 1.06955439e-01 8.96365440e-04 8.91867161e-01
-8.49475265e-01 7.20041394e-01 -1.18634723e-01 -1.98221639e-01
-1.43210399e+00 7.05867589e-01 -2.10219964e-01 -7.89221168e-01
-7.45235741e-01 7.94408560e-01 -1.40562281e-01 -2.80709535e-01
-2.49821067e-01 -8.39075506e-01 4.71811295e-01 -3.12971413e-01
6.78375542e-01 4.56351042e-01 2.35411495e-01 -3.06294531e-01
-7.88010299e-01 3.41738820e-01 -1.20071754e-01 -5.37871718e-01
1.26889169e+00 1.08281389e-01 -4.02971119e-01 3.03360492e-01
6.83992863e-01 3.44096810e-01 -8.03952277e-01 -4.22965080e-01
4.11344349e-01 -1.06120266e-01 -8.28174427e-02 -3.90266448e-01
-4.56507534e-01 7.62174428e-01 4.34081167e-01 5.31185329e-01
1.06075084e+00 1.55484080e-01 1.05001700e+00 4.33698416e-01
9.12879586e-01 -1.23465598e+00 -6.01930976e-01 8.57695103e-01
2.21540891e-02 -4.85668421e-01 -2.46174738e-01 -5.03542840e-01
-2.95887262e-01 9.96939301e-01 7.67962560e-02 -1.35806859e-01
3.07605207e-01 7.07420588e-01 -2.91078687e-01 2.84041166e-01
-8.27681541e-01 -3.64106596e-01 -5.38759410e-01 3.23380768e-01
1.14312336e-01 4.67830986e-01 -9.33852315e-01 7.86290824e-01
-1.00716040e-01 -1.66040465e-01 6.81734025e-01 1.59868848e+00
-7.20378399e-01 -2.19648790e+00 -2.90434420e-01 2.44127825e-01
-5.70512414e-01 -2.38942966e-01 -1.04554594e-01 5.93633533e-01
-1.29111320e-01 8.58742476e-01 1.65018290e-01 -2.38698006e-01
3.91659409e-01 5.15761733e-01 8.82193863e-01 -9.30693984e-01
-5.08583188e-01 -1.81291938e-01 4.25888628e-01 -5.65787256e-01
-1.94770411e-01 -8.72339845e-01 -1.72963643e+00 -6.93291545e-01
-3.44683945e-01 3.12649757e-01 9.49723899e-01 1.01355684e+00
2.78286368e-01 7.24818185e-02 5.76496303e-01 6.02131523e-02
-6.07742250e-01 -6.01380646e-01 -5.10469139e-01 -1.28687024e-01
-2.56194055e-01 -3.14557791e-01 7.44762793e-02 -3.28266859e-01] | [10.360260963439941, 9.610509872436523] |
7334ab49-761b-46f1-ba52-d5ce6b9bce5d | transformer-unet-raw-image-processing-with | 2109.08417 | null | https://arxiv.org/abs/2109.08417v1 | https://arxiv.org/pdf/2109.08417v1.pdf | Transformer-Unet: Raw Image Processing with Unet | Medical image segmentation have drawn massive attention as it is important in biomedical image analysis. Good segmentation results can assist doctors with their judgement and further improve patients' experience. Among many available pipelines in medical image analysis, Unet is one of the most popular neural networks as it keeps raw features by adding concatenation between encoder and decoder, which makes it still widely used in industrial field. In the mean time, as a popular model which dominates natural language process tasks, transformer is now introduced to computer vision tasks and have seen promising results in object detection, image classification and semantic segmentation tasks. Therefore, the combination of transformer and Unet is supposed to be more efficient than both methods working individually. In this article, we propose Transformer-Unet by adding transformer modules in raw images instead of feature maps in Unet and test our network in CT82 datasets for Pancreas segmentation accordingly. We form an end-to-end network and gain segmentation results better than many previous Unet based algorithms in our experiment. We demonstrate our network and show our experimental results in this paper accordingly. | ['Lei Hu', 'Xuquan Ji', 'Yonghong Zhang', 'Youyang Sha'] | 2021-09-17 | null | null | null | null | ['pancreas-segmentation'] | ['medical'] | [ 5.38249910e-01 1.24336466e-01 -1.92219522e-02 -4.61545855e-01
-4.86111194e-01 -3.09788287e-01 2.92576730e-01 3.24897408e-01
-7.24427402e-01 3.91103894e-01 -5.15707098e-02 -3.18870217e-01
1.95538420e-02 -7.58915007e-01 -5.76379418e-01 -6.81980312e-01
1.19870171e-01 6.32728994e-01 6.31479442e-01 -2.83134338e-02
-5.03347293e-02 1.36059940e-01 -9.37478900e-01 4.31536883e-01
9.69919205e-01 9.21244979e-01 5.17806947e-01 4.07667220e-01
-3.45659912e-01 8.88041258e-01 -2.66462713e-01 -7.11039364e-01
2.01435909e-01 -5.19421935e-01 -9.12766635e-01 -5.37192263e-02
-3.84547189e-02 -9.85155627e-02 -1.64377153e-01 1.27297688e+00
7.47137785e-01 -3.15073490e-01 2.66921818e-01 -8.44823658e-01
-3.42884928e-01 1.13717568e+00 -6.66163087e-01 2.81643957e-01
3.54304933e-03 3.63469198e-02 5.53933382e-01 -4.08104479e-01
5.19856989e-01 1.10484934e+00 8.08331370e-01 4.30073291e-01
-7.08207548e-01 -6.38128757e-01 -7.63854086e-02 2.81836361e-01
-1.00253534e+00 -2.69861668e-01 5.52788377e-01 -4.83833030e-02
7.99270213e-01 3.10727209e-01 8.25740337e-01 7.24141300e-01
4.09736216e-01 1.27097082e+00 8.02875936e-01 -1.62314296e-01
-4.53374870e-02 -5.87720796e-02 8.48734677e-02 8.51770759e-01
7.16865733e-02 -4.21717018e-01 5.44857681e-02 4.74910259e-01
8.59158814e-01 1.77667454e-01 -2.62287468e-01 -2.42083490e-01
-1.68474472e+00 6.58498704e-01 8.86054754e-01 4.38703954e-01
-4.05102104e-01 1.39818802e-01 7.36887753e-01 6.58060366e-04
2.42577717e-01 2.02906191e-01 -4.55867201e-01 -3.50517668e-02
-7.45727539e-01 3.73888835e-02 4.02040720e-01 8.96209836e-01
1.23429962e-01 -2.35749647e-01 -4.29274768e-01 9.61286128e-01
1.21697031e-01 2.51508951e-01 9.10615027e-01 -6.29574239e-01
3.59199733e-01 9.56553817e-01 -5.67556679e-01 -7.15737402e-01
-6.99811578e-01 -9.18603778e-01 -1.27249300e+00 -2.54735082e-01
3.81915718e-01 9.33113042e-03 -1.16475701e+00 1.41062486e+00
2.63067812e-01 4.30859506e-01 1.51148647e-01 9.22176421e-01
1.09818017e+00 5.06385028e-01 2.01495871e-01 -9.48996469e-02
1.69274020e+00 -1.25358021e+00 -8.32715034e-01 -1.84695840e-01
6.36062145e-01 -9.72388804e-01 5.99752247e-01 5.70692480e-01
-9.75434899e-01 -3.82111311e-01 -8.70258033e-01 -1.81767330e-01
-3.45814258e-01 -1.69021878e-02 7.99052596e-01 4.95250314e-01
-9.39440012e-01 5.43035865e-01 -1.12290943e+00 -5.63209057e-01
9.80985284e-01 4.22877699e-01 -2.61855781e-01 -1.13695003e-01
-1.06880474e+00 9.06740725e-01 6.24487877e-01 4.64242905e-01
-5.69036603e-01 -2.97718704e-01 -7.83756733e-01 -4.63759191e-02
4.54476237e-01 -9.33736145e-01 1.57509935e+00 -7.13232219e-01
-1.30293715e+00 7.60197580e-01 9.10060555e-02 -7.96102583e-01
6.71849608e-01 -4.63247852e-04 -7.01971054e-02 1.99505001e-01
1.14321947e-01 9.13472831e-01 4.03797925e-01 -6.17422819e-01
-8.83382618e-01 -2.21896112e-01 -1.78365067e-01 9.29767564e-02
-1.46199793e-01 -1.04585933e-02 -8.20884347e-01 -5.52269816e-01
3.25352818e-01 -7.50808001e-01 -4.60085988e-01 -7.68966784e-05
-6.09412909e-01 -2.77443916e-01 8.38331699e-01 -6.90422773e-01
9.80052650e-01 -1.95793879e+00 7.36488104e-02 -6.89998642e-02
4.44666773e-01 5.05481660e-01 1.15398698e-01 -7.64483362e-02
-5.12069054e-02 6.65753707e-02 -5.94546020e-01 -1.62002772e-01
-1.79682150e-01 2.86010116e-01 3.65673900e-01 3.99565786e-01
1.47521392e-01 1.22170627e+00 -8.50528479e-01 -1.02770293e+00
5.15584230e-01 5.06978750e-01 -5.24834871e-01 9.95701551e-02
-1.87797457e-01 6.88756049e-01 -6.33786082e-01 5.95339239e-01
6.43677354e-01 -4.12660390e-01 -2.53130700e-02 -4.47312117e-01
-8.76434743e-02 -1.07788801e-01 -6.33917570e-01 2.11148667e+00
-3.01945210e-01 4.79870081e-01 2.01154537e-02 -1.31253946e+00
6.54743075e-01 4.23785836e-01 6.60931170e-01 -9.05740321e-01
5.65654933e-01 2.28535682e-01 3.07500601e-01 -8.82192850e-01
5.39439507e-02 -2.21579358e-01 -1.01499572e-01 -1.47431297e-02
-5.83308041e-02 -3.19559500e-02 4.06821668e-01 1.11490302e-01
1.06173813e+00 1.33457914e-01 3.71627748e-01 -1.36433318e-01
6.65331900e-01 1.94843784e-01 7.16530621e-01 3.18943560e-01
-3.53636056e-01 7.21430838e-01 5.60043037e-01 -4.95469272e-01
-9.63078141e-01 -7.49256968e-01 -3.01116168e-01 5.74474573e-01
1.91836014e-01 -2.77489692e-01 -9.82634723e-01 -7.56440639e-01
-3.44766378e-01 4.32896435e-01 -4.83788252e-01 8.38064179e-02
-8.57522488e-01 -9.64495838e-01 7.19133615e-01 5.71471810e-01
1.17955053e+00 -1.23104489e+00 -8.37917507e-01 3.47680300e-01
-3.10985029e-01 -1.19760466e+00 -3.88119698e-01 3.24960113e-01
-9.28748429e-01 -1.20660698e+00 -1.11164320e+00 -1.27724087e+00
6.23671889e-01 8.64627361e-02 8.80127251e-01 6.73978627e-02
-5.58152974e-01 -3.24488610e-01 -5.64930558e-01 -4.36912477e-01
-4.26953316e-01 4.27344203e-01 -6.86241090e-01 -2.61475295e-01
1.92344204e-01 -3.78365904e-01 -8.58345270e-01 8.22787583e-02
-1.10595608e+00 6.21647120e-01 9.88086879e-01 7.76718855e-01
4.62677568e-01 2.26989582e-01 3.43916714e-01 -1.23724055e+00
4.70177054e-01 -2.70431489e-01 -4.98455793e-01 2.34754264e-01
-4.12839085e-01 -6.34460244e-03 6.96655750e-01 4.06554602e-02
-9.56926405e-01 9.85201001e-02 -7.05678642e-01 -2.34919354e-01
-1.25236571e-01 6.50244832e-01 -1.02630645e-01 1.93988875e-01
1.05341204e-01 3.79803300e-01 1.97028756e-01 -6.57814264e-01
-5.49915470e-02 6.01966202e-01 6.39083147e-01 -8.88159722e-02
2.92413652e-01 1.75400764e-01 -2.25413404e-03 -5.67075670e-01
-7.48485506e-01 -4.15691763e-01 -5.61069310e-01 -4.21945639e-02
1.31198263e+00 -6.57164931e-01 -8.34852815e-01 6.96421981e-01
-1.09909904e+00 -5.17336167e-02 2.47484818e-02 3.76160175e-01
-1.62844166e-01 4.33721274e-01 -8.79867673e-01 -2.87946254e-01
-6.84032381e-01 -1.71794081e+00 1.10391259e+00 5.82100868e-01
1.36986762e-01 -1.01708162e+00 -3.17306012e-01 3.21259260e-01
6.21918023e-01 2.86087364e-01 9.96163547e-01 -7.41241157e-01
-5.91713369e-01 -1.87012509e-01 -5.29375434e-01 3.67839128e-01
1.60531819e-01 -4.07655448e-01 -7.74227083e-01 -1.57534733e-01
2.31747865e-03 7.43508935e-02 1.09010041e+00 5.71752429e-01
1.46785998e+00 -2.49514487e-02 -6.12577200e-01 8.40179145e-01
1.53481662e+00 3.30806315e-01 7.13983715e-01 5.62611699e-01
7.00500488e-01 4.34817433e-01 2.21299544e-01 2.62673311e-02
3.54831487e-01 4.20003563e-01 6.68353200e-01 -6.20980740e-01
-2.14302182e-01 -1.50219053e-02 3.80879417e-02 1.14832246e+00
2.16511697e-01 -3.98146659e-01 -9.72581923e-01 4.87956315e-01
-1.83041847e+00 -6.62709534e-01 -4.06823426e-01 1.80517089e+00
8.42041254e-01 2.97461689e-01 -8.41395631e-02 1.41210884e-01
7.57661462e-01 -2.36091495e-01 -5.66177309e-01 -1.68493897e-01
-6.61613233e-03 4.86140519e-01 6.24969065e-01 1.17828816e-01
-1.38076580e+00 8.83128524e-01 5.20701885e+00 9.14556861e-01
-1.16866910e+00 2.33801365e-01 7.71114588e-01 2.46185407e-01
7.26360381e-02 -3.27894777e-01 -4.05529618e-01 5.24932563e-01
5.81232250e-01 5.33728823e-02 -7.22887293e-02 5.50766051e-01
2.00080946e-01 -1.76793441e-01 -9.96850848e-01 9.62937415e-01
-2.04730093e-01 -1.24148273e+00 1.01357698e-01 -2.93326467e-01
2.50171334e-01 4.24121656e-02 -3.18025821e-03 3.17810595e-01
1.52175993e-01 -1.04373014e+00 4.78726715e-01 1.70598894e-01
4.50795621e-01 -6.54826999e-01 1.45428526e+00 4.39210266e-01
-1.14148486e+00 1.11222848e-01 -2.15638086e-01 2.88542837e-01
4.17356819e-01 7.64080226e-01 -1.12332535e+00 7.59807527e-01
6.35086179e-01 9.04330671e-01 -6.64000452e-01 1.71827137e+00
-1.14868037e-01 5.99624336e-01 -2.57779986e-01 1.15605490e-03
4.79858279e-01 -1.82668298e-01 3.79766226e-01 1.32162189e+00
2.74843603e-01 -4.96104248e-02 3.37215811e-01 5.73205054e-01
-2.09516615e-01 2.09240213e-01 -1.89736873e-01 1.13018021e-01
4.37948183e-04 1.46154881e+00 -1.47219276e+00 -4.96945649e-01
-1.52424857e-01 9.75688696e-01 -1.03962570e-01 -1.65285870e-01
-1.15367675e+00 -5.54675221e-01 2.37943038e-01 -1.16832808e-01
2.36271232e-01 2.13670179e-01 -9.28650051e-02 -9.75261152e-01
-1.77485675e-01 -8.05485070e-01 2.98442781e-01 -6.28098726e-01
-1.08200407e+00 8.94905686e-01 -1.91807806e-01 -1.22376716e+00
1.50091693e-01 -7.75545180e-01 -2.86043108e-01 6.07836485e-01
-1.69795346e+00 -1.26755619e+00 -3.94531965e-01 4.72860754e-01
9.17209506e-01 1.93245217e-01 6.52117848e-01 7.09932148e-01
-9.78421926e-01 3.54329318e-01 -6.78476272e-03 6.25943184e-01
6.49161816e-01 -1.36391866e+00 3.26375276e-01 8.28716099e-01
-2.03829184e-01 5.14812469e-01 5.46240985e-01 -6.04234934e-01
-1.34340322e+00 -1.28696871e+00 6.17864668e-01 1.06227286e-01
3.34259272e-01 -9.49648023e-02 -6.24808311e-01 7.22059608e-01
6.44559205e-01 -8.63744244e-02 4.02059406e-01 -5.17929137e-01
2.31889382e-01 -1.70562133e-01 -1.12074256e+00 4.18492258e-01
8.42049241e-01 2.64492989e-01 -4.97115433e-01 4.37770277e-01
8.67004693e-01 -8.36272061e-01 -8.79697084e-01 5.77738047e-01
4.03290212e-01 -1.00833392e+00 8.03652167e-01 -1.04823790e-01
4.82104093e-01 -4.24833953e-01 3.11955094e-01 -1.10204375e+00
-2.46932954e-01 -2.60616481e-01 5.60381353e-01 1.12157083e+00
3.97151262e-01 -7.31268466e-01 7.75913417e-01 7.13293403e-02
-5.98970771e-01 -8.65546525e-01 -8.20747137e-01 -3.22900921e-01
1.59456834e-01 -3.87960225e-01 5.52138269e-01 7.33175695e-01
-1.91644654e-01 4.35087681e-01 -2.26833355e-02 -2.18618423e-01
7.48777151e-01 8.22083279e-02 3.00491571e-01 -1.17981160e+00
2.58630943e-02 -7.71896958e-01 -3.92062902e-01 -1.06267178e+00
-3.63018215e-01 -1.34234130e+00 2.76953518e-01 -1.97125673e+00
5.84024310e-01 -4.79920655e-01 -2.59443134e-01 7.15428591e-01
-8.81883353e-02 5.12696743e-01 3.52238193e-02 5.14871553e-02
-8.33115458e-01 2.67986774e-01 1.79482424e+00 -3.23495299e-01
1.39977783e-01 8.11175928e-02 -6.46567345e-01 6.93978012e-01
7.93541253e-01 -3.94026071e-01 -2.09967464e-01 -8.27643692e-01
6.70634210e-02 -4.03823238e-03 2.48583138e-01 -1.08552837e+00
4.50813621e-01 3.99632543e-01 5.85942626e-01 -7.38061368e-01
-9.76858214e-02 -1.03208590e+00 3.25794928e-02 7.65194356e-01
-3.14953089e-01 2.14899048e-01 2.00057358e-01 2.27336556e-01
-4.50652659e-01 -1.09276451e-01 8.52975845e-01 -5.54015458e-01
-7.64383197e-01 3.77802283e-01 -3.55135530e-01 -2.25904077e-01
1.09900558e+00 -2.08324492e-01 -1.54145852e-01 7.32490867e-02
-6.82919919e-01 5.51371753e-01 1.30745217e-01 2.92982191e-01
5.70113122e-01 -9.13484156e-01 -8.50312352e-01 9.31656733e-02
-7.67719224e-02 5.64481497e-01 3.09213936e-01 1.38336158e+00
-1.09797251e+00 6.64491057e-01 -1.20467804e-01 -9.97687221e-01
-1.20747948e+00 4.99815196e-01 2.64884025e-01 -5.52505791e-01
-8.46889615e-01 1.00269699e+00 2.70194024e-01 -3.76513481e-01
2.78781921e-01 -8.16159248e-01 -4.66697752e-01 -1.68033138e-01
2.17832997e-01 4.71844450e-02 2.50779599e-01 -5.28477967e-01
-3.15136492e-01 4.75665182e-01 -3.67342442e-01 4.09510821e-01
1.43069375e+00 -8.33942648e-03 -5.04479945e-01 2.32952405e-02
1.18320048e+00 -4.28658307e-01 -8.09587836e-01 -1.07617728e-01
9.24944654e-02 -6.12355489e-03 2.01585427e-01 -1.03504574e+00
-1.64347088e+00 9.85415876e-01 7.98036397e-01 1.99470639e-01
1.45420647e+00 -8.75695646e-02 1.34874606e+00 8.12439546e-02
3.92898142e-01 -6.45559132e-01 -2.53703445e-01 2.73137391e-01
3.19959581e-01 -1.23249459e+00 -9.78837237e-02 -6.10609531e-01
-4.64659214e-01 1.15649390e+00 4.33137298e-01 -1.89578459e-02
3.94342571e-01 4.40880418e-01 1.46544859e-01 -9.75167826e-02
-4.74292636e-01 -2.21318170e-01 -2.15140376e-02 3.00342143e-01
7.62973547e-01 -7.73182660e-02 -4.78232682e-01 5.99279821e-01
-1.80934802e-01 3.29330742e-01 2.33160585e-01 9.21496809e-01
-4.32669669e-01 -1.34842443e+00 -1.37279332e-01 7.19713449e-01
-9.06745076e-01 -1.26995549e-01 1.00504436e-01 7.81979263e-01
4.35156077e-01 7.03747451e-01 -2.02560574e-01 -1.91296905e-01
3.69487792e-01 -1.18483715e-01 5.18281877e-01 -4.42532331e-01
-8.94978464e-01 3.24144810e-01 -1.32241562e-01 -2.98828840e-01
-4.83334422e-01 -6.09396040e-01 -1.51996279e+00 -1.63421053e-02
-3.05406988e-01 2.34317854e-01 6.71802938e-01 1.05311024e+00
3.09706503e-03 1.24185324e+00 1.85910240e-01 -5.69415808e-01
-2.08057433e-01 -1.13001275e+00 -1.17982075e-01 2.74188936e-01
1.24923490e-01 -3.34568411e-01 2.25037336e-01 2.26765603e-01] | [14.56950855255127, -2.6020796298980713] |
a5aacd11-ff94-436b-b65a-d0848b11bd68 | everybody-is-unique-towards-unbiased-human | 2107.06239 | null | https://arxiv.org/abs/2107.06239v1 | https://arxiv.org/pdf/2107.06239v1.pdf | Everybody Is Unique: Towards Unbiased Human Mesh Recovery | We consider the problem of obese human mesh recovery, i.e., fitting a parametric human mesh to images of obese people. Despite obese person mesh fitting being an important problem with numerous applications (e.g., healthcare), much recent progress in mesh recovery has been restricted to images of non-obese people. In this work, we identify this crucial gap in the current literature by presenting and discussing limitations of existing algorithms. Next, we present a simple baseline to address this problem that is scalable and can be easily used in conjunction with existing algorithms to improve their performance. Finally, we present a generalized human mesh optimization algorithm that substantially improves the performance of existing methods on both obese person images as well as community-standard benchmark datasets. A key innovation of this technique is that it does not rely on supervision from expensive-to-create mesh parameters. Instead, starting from widely and cheaply available 2D keypoints annotations, our method automatically generates mesh parameters that can in turn be used to re-train and fine-tune any existing mesh estimation algorithm. This way, we show our method acts as a drop-in to improve the performance of a wide variety of contemporary mesh estimation methods. We conduct extensive experiments on multiple datasets comprising both standard and obese person images and demonstrate the efficacy of our proposed techniques. | ['Ziyan Wu', 'Terrence Chen', 'Srikrishna Karanam', 'Meng Zheng', 'Ren Li'] | 2021-07-13 | null | null | null | null | ['human-mesh-recovery'] | ['computer-vision'] | [ 3.39024693e-01 1.79517761e-01 -1.09496824e-01 -1.98987409e-01
-8.06666315e-01 -1.57794997e-01 2.52739906e-01 1.07238255e-01
-2.60832936e-01 5.36109626e-01 6.31614998e-02 1.67617425e-01
3.40406559e-02 -7.08207190e-01 -7.15693831e-01 -2.66805410e-01
5.92126511e-02 9.85679030e-01 3.10027450e-01 -4.30577725e-01
-1.63945109e-02 4.17395353e-01 -1.45743918e+00 -1.69921666e-01
9.10729885e-01 8.91769707e-01 -2.63041347e-01 3.24127555e-01
2.06374168e-01 3.90901119e-02 -1.11906245e-01 -6.38536334e-01
6.09177411e-01 -2.60921866e-01 -9.65627491e-01 4.26060617e-01
8.54797602e-01 -2.18516305e-01 -7.81395510e-02 9.30805981e-01
7.22050428e-01 2.55528212e-01 6.22918785e-01 -1.03776777e+00
-2.12665170e-01 2.19863877e-01 -1.00019693e+00 -2.19673470e-01
5.66952944e-01 -2.31042318e-02 5.75781584e-01 -1.09234750e+00
8.80645633e-01 1.31546569e+00 1.23897660e+00 6.12340748e-01
-1.36024523e+00 -7.36043572e-01 4.16869447e-02 -1.21839009e-01
-1.49861479e+00 -6.11587346e-01 1.00938451e+00 -5.74335814e-01
4.88886088e-01 4.20916229e-01 1.22022021e+00 8.97219598e-01
-1.03011623e-01 5.92320323e-01 9.56269860e-01 -3.90938401e-01
-7.66891390e-02 -2.09481493e-01 -2.73797065e-01 1.09886658e+00
1.88980401e-01 -8.18754733e-02 -3.00919503e-01 -4.64297742e-01
1.15419662e+00 -1.11875936e-01 -2.41944402e-01 -8.69909823e-01
-1.32768798e+00 6.77051127e-01 3.45926493e-01 -2.65252918e-01
-4.51869994e-01 2.36592487e-01 4.39473957e-01 4.40280624e-02
7.52284586e-01 2.50164896e-01 -6.37908205e-02 8.52477923e-02
-1.13633931e+00 7.02598631e-01 6.76131308e-01 8.45744371e-01
6.56432033e-01 -3.33640277e-01 8.08320194e-02 1.06216562e+00
2.11219192e-01 3.54526848e-01 -2.48415228e-02 -1.08692241e+00
4.59546894e-01 6.55419827e-01 9.78418216e-02 -1.34139073e+00
-4.53600705e-01 -8.51092935e-02 -9.51016545e-01 2.19951898e-01
4.78107274e-01 -7.71663105e-03 -8.55720699e-01 1.46720445e+00
8.40885401e-01 4.62064087e-01 -5.76738954e-01 9.12403762e-01
8.29191804e-01 5.95949031e-02 6.71857148e-02 -6.97561204e-02
1.39621484e+00 -1.15352869e+00 -4.56931114e-01 -1.67923011e-02
5.92548400e-02 -8.49392951e-01 1.04445243e+00 4.29130882e-01
-1.52491391e+00 -4.77124453e-01 -7.65422165e-01 -2.67614245e-01
-3.25375162e-02 -2.40040883e-01 7.89175153e-01 3.10783297e-01
-9.68081534e-01 8.37613523e-01 -7.84706533e-01 -6.03748024e-01
5.82721114e-01 5.81459403e-01 -5.11185348e-01 6.36151899e-03
-9.68406618e-01 8.83032441e-01 1.88059822e-01 7.93176368e-02
-4.00569081e-01 -9.23615754e-01 -9.04506505e-01 -4.21993375e-01
5.65966666e-01 -1.47467244e+00 1.26088309e+00 -8.73448789e-01
-1.50147235e+00 1.33095825e+00 -1.52198419e-01 -1.85181528e-01
1.06917989e+00 -2.99129486e-01 -5.95642999e-02 4.25014794e-01
1.16958939e-01 9.43696260e-01 9.29227829e-01 -1.35499740e+00
-5.19638717e-01 -2.08056778e-01 -1.46159222e-02 1.57310635e-01
-3.58403958e-02 8.52709264e-02 -8.35824072e-01 -9.35494959e-01
2.13347569e-01 -1.20916831e+00 -4.87216949e-01 5.50148785e-01
-5.34319460e-01 -2.38996640e-01 4.96405423e-01 -7.03322291e-01
1.16830802e+00 -1.65978587e+00 5.16057670e-01 5.14629126e-01
5.35980165e-01 -4.96431217e-02 1.56161353e-01 3.51704121e-01
-6.70267716e-02 -4.54691611e-03 -5.85600197e-01 -6.75090253e-01
-1.05169695e-02 1.73566472e-02 1.86541125e-01 8.09016168e-01
8.05119798e-02 1.02006757e+00 -9.19465065e-01 -9.51529562e-01
4.21269655e-01 6.58184230e-01 -7.71178484e-01 -6.07486069e-02
-7.75149316e-02 9.10898626e-01 -2.65489995e-01 8.73179018e-01
6.48991168e-01 -2.93509901e-01 4.64599282e-02 -4.16574270e-01
1.07866935e-01 -1.98642567e-01 -1.53490639e+00 2.16149855e+00
-2.98255771e-01 2.32699495e-02 1.98091164e-01 -1.00930631e+00
5.14813125e-01 4.56945956e-01 1.08191872e+00 -3.26792389e-01
2.70570487e-01 3.14959377e-01 -4.75587666e-01 -4.77274269e-01
3.70316625e-01 -2.35476106e-01 -6.42081425e-02 3.64673078e-01
-3.16869885e-01 -1.24435045e-01 1.15074940e-01 -1.06195256e-01
7.45943666e-01 3.14234912e-01 5.84481895e-01 -2.89839298e-01
6.29736364e-01 2.86755990e-02 5.38951993e-01 3.66414636e-01
-9.92304981e-02 8.14904869e-01 1.08745452e-02 -6.56805694e-01
-1.17372811e+00 -9.51540291e-01 -3.68061662e-01 8.80749345e-01
3.47354025e-01 -6.74180269e-01 -8.56478453e-01 -6.43776119e-01
2.48802036e-01 -2.12808266e-01 -8.28052580e-01 1.99761212e-01
-1.00576985e+00 -5.11782527e-01 4.19638872e-01 5.91211081e-01
4.40063357e-01 -8.62685680e-01 -5.73203444e-01 1.59908772e-01
-4.93130863e-01 -9.48838174e-01 -6.93963587e-01 -6.27454996e-01
-9.59894359e-01 -1.31194389e+00 -1.02024913e+00 -8.16739142e-01
9.63165462e-01 1.80159792e-01 1.58103001e+00 6.06147349e-01
-4.69600141e-01 6.19936049e-01 -6.41481802e-02 -4.27587301e-01
-1.49742886e-01 1.90353587e-01 1.51221275e-01 -7.22626746e-02
-1.32428616e-01 -7.65466630e-01 -9.48543429e-01 3.95348936e-01
-6.31942928e-01 3.54010582e-01 3.44311804e-01 5.98296046e-01
1.08194792e+00 -8.56736451e-02 3.20659280e-01 -1.11989295e+00
4.07245129e-01 -3.39592367e-01 -3.12534958e-01 7.81785622e-02
-5.01787186e-01 -2.65515804e-01 1.85688078e-01 -5.30538917e-01
-5.83918333e-01 4.01507676e-01 -3.87117386e-01 -5.03816009e-01
3.11797950e-02 2.35171169e-01 5.06923022e-03 -5.61568022e-01
3.96788627e-01 -3.71260852e-01 1.23164013e-01 -7.14784682e-01
3.79894972e-01 3.39728177e-01 7.69149542e-01 -9.03339922e-01
1.12582183e+00 7.58465886e-01 2.56226778e-01 -7.17481256e-01
-7.13113546e-01 -5.71874976e-01 -8.65381539e-01 -2.67568499e-01
7.28766263e-01 -7.85746396e-01 -8.40583742e-01 2.68644810e-01
-8.35679531e-01 -3.63708079e-01 -2.44069144e-01 6.05954118e-02
-8.08173358e-01 5.99739313e-01 -5.42000234e-01 -4.64331180e-01
-6.21282578e-01 -1.29479158e+00 1.50555909e+00 -8.98303315e-02
-6.21827126e-01 -1.13540566e+00 2.02325508e-01 4.89316136e-01
2.61673331e-01 8.76736343e-01 7.23132908e-01 -2.78639458e-02
-2.77419895e-01 -1.49227530e-01 -4.19807397e-02 -1.10682689e-01
3.73122911e-03 -1.41102970e-01 -5.53198874e-01 -4.05798107e-01
-4.19148535e-01 -3.22476119e-01 6.75767779e-01 3.65301728e-01
1.34041715e+00 -1.83937132e-01 -5.96831203e-01 7.79038012e-01
1.30310869e+00 -6.51220500e-01 4.87569004e-01 2.73011357e-01
1.04468799e+00 6.46900237e-01 5.49899936e-01 2.89796799e-01
7.07431316e-01 1.16343355e+00 2.57554710e-01 -5.16129017e-01
-3.09351772e-01 -1.96612850e-01 -2.54626960e-01 6.52974129e-01
-7.50314713e-01 4.32391673e-01 -1.01594472e+00 3.40858370e-01
-1.95353115e+00 -8.18840206e-01 -1.46002769e-01 2.24368858e+00
1.09542668e+00 -2.00824469e-01 6.69261277e-01 1.45300150e-01
5.89740396e-01 -1.68587551e-01 -4.07617837e-01 1.76681861e-01
3.93633574e-01 5.84452391e-01 2.78664678e-01 4.78177458e-01
-1.42019725e+00 7.35050082e-01 6.59804773e+00 6.75949991e-01
-8.14543307e-01 8.82637724e-02 1.91233158e-01 -1.41427703e-02
-4.94541563e-02 -1.96140617e-01 -4.62181628e-01 3.16510171e-01
3.29169393e-01 -4.89762686e-02 4.68819618e-01 7.70347536e-01
1.84850216e-01 8.35325271e-02 -1.37150526e+00 1.35477936e+00
1.54726118e-01 -1.23241246e+00 -4.74255346e-02 2.41393849e-01
7.27082908e-01 -4.30777371e-01 -1.28287777e-01 1.23215675e-01
3.83301103e-03 -1.17683852e+00 7.88801134e-01 6.67848051e-01
1.10390663e+00 -7.88761973e-01 4.93193239e-01 2.57086873e-01
-1.60383427e+00 3.85390520e-01 -2.66540170e-01 1.34849492e-02
3.67164344e-01 6.30862117e-01 -4.36424404e-01 7.09320664e-01
6.64675415e-01 7.66858220e-01 -4.90224153e-01 1.22420907e+00
-1.06959417e-02 1.38645381e-01 -4.73393440e-01 7.41244912e-01
-2.47255310e-01 -2.51843929e-01 5.32630622e-01 1.11219263e+00
2.50352491e-02 2.13603720e-01 8.96473169e-01 6.49110556e-01
-2.91580141e-01 5.01105845e-01 -5.12926638e-01 4.92479563e-01
3.74497712e-01 1.26208842e+00 -8.34628582e-01 -4.31016028e-01
-3.64566475e-01 9.62464333e-01 4.80175763e-01 4.52884398e-02
-9.51056480e-01 -1.72289629e-02 4.67560410e-01 6.93386614e-01
-5.67876734e-02 -8.39445740e-02 -1.96208760e-01 -1.11353254e+00
9.47808996e-02 -1.04376984e+00 3.02570164e-01 -2.83143103e-01
-1.49255109e+00 2.08471611e-01 2.50138998e-01 -1.26567340e+00
-1.17868565e-01 -1.76360592e-01 -4.18846101e-01 6.98564649e-01
-1.39991355e+00 -1.59050608e+00 -6.58567309e-01 6.55735373e-01
6.47402346e-01 4.52860266e-01 8.79574716e-01 6.19834542e-01
-6.42304599e-01 7.16388106e-01 -4.41017985e-01 2.24331945e-01
9.13696170e-01 -1.22455394e+00 5.10216773e-01 5.51878512e-01
-1.87690541e-01 6.86069429e-01 6.51350617e-01 -1.01093698e+00
-1.26306295e+00 -1.18144166e+00 5.35046220e-01 -5.99362910e-01
2.71147549e-01 -3.12851310e-01 -7.34053135e-01 8.53370130e-01
-1.12769261e-01 2.94080585e-01 6.39475048e-01 1.15477934e-01
-3.50335985e-02 1.41738459e-01 -1.22886097e+00 7.05102265e-01
1.42651665e+00 6.00122400e-02 -5.47773659e-01 4.96835619e-01
2.93051451e-01 -1.07973754e+00 -1.18558133e+00 8.53893816e-01
8.41013968e-01 -8.74121726e-01 1.55300701e+00 -5.34695745e-01
3.58689457e-01 -1.03262112e-01 1.29197031e-01 -1.01605451e+00
-3.66895467e-01 -8.40602100e-01 -4.73912060e-01 8.58457565e-01
-1.10364661e-01 -4.08352345e-01 8.27570736e-01 7.35041678e-01
-1.03550225e-01 -1.24013972e+00 -8.76874328e-01 -4.77730036e-01
1.21575452e-01 -2.47608542e-01 6.00424170e-01 9.71836925e-01
-1.83808059e-01 -1.61038563e-02 -5.77222466e-01 -5.35908416e-02
9.93746579e-01 1.28687397e-01 1.17617691e+00 -1.74406874e+00
-1.01704039e-01 -3.78117412e-01 -3.52150053e-01 -9.61012244e-01
3.14695388e-01 -1.02530217e+00 5.14118485e-02 -1.64548230e+00
3.75796288e-01 -8.23886514e-01 9.57482904e-02 4.92183566e-01
-4.67295110e-01 9.10267591e-01 9.23006311e-02 3.88817966e-01
-5.99653184e-01 2.74395078e-01 1.44121325e+00 -6.73236176e-02
-1.71509519e-01 1.26535326e-01 -6.23773515e-01 1.25245643e+00
5.94265759e-01 -3.91088217e-01 -3.34160835e-01 -1.97033405e-01
2.96063989e-01 -1.02153257e-01 5.92570484e-01 -1.03160989e+00
-5.55783845e-02 -1.93468735e-01 4.80677843e-01 -4.82628763e-01
3.48789781e-01 -8.85230005e-01 4.04977590e-01 3.39737177e-01
1.22886449e-02 2.72202224e-01 2.95043159e-02 5.60454190e-01
3.12233686e-01 -8.48721713e-02 1.00302601e+00 -4.99588758e-01
-4.12456781e-01 6.45319819e-01 2.51828969e-01 3.48998159e-01
1.16262364e+00 -4.46610034e-01 2.55146265e-01 -1.44660339e-01
-7.90604293e-01 3.34859312e-01 1.01035845e+00 2.31026292e-01
6.74744606e-01 -1.53487206e+00 -6.84571683e-01 4.64950912e-02
3.66010107e-02 3.95678073e-01 1.00750029e-01 1.25991476e+00
-5.68772197e-01 -1.31750852e-02 -1.26759008e-01 -6.72753155e-01
-1.36285603e+00 4.86074984e-01 3.68439287e-01 -3.35013151e-01
-1.29358566e+00 6.79518461e-01 -3.19696032e-02 -4.55702156e-01
4.30188715e-01 -2.32321769e-01 -6.10516183e-02 -1.63658977e-01
4.23287451e-01 6.19465888e-01 3.56474482e-02 -8.98249388e-01
-3.72410923e-01 1.17666864e+00 2.66608506e-01 8.36436227e-02
1.22312760e+00 -1.96530297e-01 -1.87348813e-01 3.65356326e-01
9.04077232e-01 3.97656471e-01 -9.36751783e-01 -1.83722809e-01
-2.14417458e-01 -4.83475089e-01 -2.11750343e-01 -3.77223223e-01
-1.16908288e+00 4.78113621e-01 4.11208004e-01 -2.13598207e-01
1.00645888e+00 5.37368953e-02 1.34574842e+00 2.23176274e-02
5.65547705e-01 -9.65503871e-01 6.85192868e-02 -3.99490409e-02
9.82329965e-01 -1.27961779e+00 4.65459406e-01 -9.94412601e-01
-3.39891911e-01 9.58710074e-01 3.87668341e-01 -3.68712276e-01
7.10937798e-01 1.32805496e-01 -2.35771593e-02 -4.98173833e-01
-1.19332105e-01 -1.58292994e-01 7.21360683e-01 6.08312011e-01
4.09452230e-01 -3.75312380e-02 -4.22517836e-01 3.07467520e-01
-4.35773939e-01 2.09499508e-01 7.93592557e-02 9.09690917e-01
-2.43529901e-01 -1.38904715e+00 -4.97432798e-01 5.29292285e-01
-4.45557773e-01 1.59741506e-01 -1.98070824e-01 9.33318138e-01
4.57585841e-01 5.33424616e-01 -1.83283985e-01 -1.49633005e-01
7.00999737e-01 7.88385863e-04 8.09790790e-01 -7.43698895e-01
-5.54827213e-01 3.48790698e-02 -1.81958713e-02 -7.07830191e-01
-7.06130326e-01 -6.34743094e-01 -1.17137790e+00 -4.65454698e-01
-1.41522184e-01 -1.40271753e-01 2.93437123e-01 7.79402494e-01
1.84877723e-01 3.89365196e-01 1.23851776e-01 -1.44501400e+00
-3.22353095e-01 -6.46203935e-01 -2.89008021e-01 9.29261208e-01
1.39465153e-01 -1.35194361e+00 1.07521713e-01 2.19620481e-01] | [7.08579683303833, -1.1219227313995361] |
9a189336-cc40-415d-b7bc-6b9caf6189bf | all-about-structure-adapting-structural | 1903.12212 | null | http://arxiv.org/abs/1903.12212v1 | http://arxiv.org/pdf/1903.12212v1.pdf | All about Structure: Adapting Structural Information across Domains for Boosting Semantic Segmentation | In this paper we tackle the problem of unsupervised domain adaptation for the
task of semantic segmentation, where we attempt to transfer the knowledge
learned upon synthetic datasets with ground-truth labels to real-world images
without any annotation. With the hypothesis that the structural content of
images is the most informative and decisive factor to semantic segmentation and
can be readily shared across domains, we propose a Domain Invariant Structure
Extraction (DISE) framework to disentangle images into domain-invariant
structure and domain-specific texture representations, which can further
realize image-translation across domains and enable label transfer to improve
segmentation performance. Extensive experiments verify the effectiveness of our
proposed DISE model and demonstrate its superiority over several
state-of-the-art approaches. | ['Wei-Chen Chiu', 'Wen-Hsiao Peng', 'Hui-Po Wang', 'Wei-Lun Chang'] | 2019-03-26 | all-about-structure-adapting-structural-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Chang_All_About_Structure_Adapting_Structural_Information_Across_Domains_for_Boosting_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Chang_All_About_Structure_Adapting_Structural_Information_Across_Domains_for_Boosting_CVPR_2019_paper.pdf | cvpr-2019-6 | ['synthetic-to-real-translation'] | ['computer-vision'] | [ 7.38174677e-01 1.26615763e-01 -3.74845237e-01 -5.67609072e-01
-7.12746561e-01 -7.49917090e-01 5.84747195e-01 -1.91316009e-01
-3.77932131e-01 7.23745763e-01 -1.61766320e-01 -5.81668317e-02
4.78452221e-02 -7.71094501e-01 -8.75880241e-01 -7.40613401e-01
4.54541177e-01 7.44684577e-01 4.46730942e-01 -5.16802520e-02
-7.39144441e-03 4.09531325e-01 -1.31977212e+00 2.75842309e-01
1.19211841e+00 1.03556705e+00 4.36257869e-01 2.41583794e-01
-2.66487956e-01 5.48854470e-01 -3.50565553e-01 -2.02396616e-01
2.64655948e-01 -4.38978195e-01 -1.24491787e+00 6.53261900e-01
3.86548758e-01 -5.99857941e-02 -1.10330366e-01 1.38112187e+00
7.28863180e-02 -2.55776942e-02 1.03985775e+00 -1.04951632e+00
-7.76819587e-01 1.32538721e-01 -6.00532055e-01 -6.28858730e-02
5.07898256e-02 2.90458426e-02 9.23174858e-01 -5.95500231e-01
1.08789897e+00 1.07957017e+00 2.29785040e-01 5.34600198e-01
-1.61689174e+00 -4.68040198e-01 3.28969449e-01 -2.42402479e-02
-1.24535704e+00 -1.69108137e-01 1.04338551e+00 -6.96399927e-01
1.25149846e-01 -7.68332705e-02 3.92455190e-01 1.20550406e+00
-1.94090724e-01 9.80977714e-01 1.45387304e+00 -3.24738353e-01
2.35864624e-01 6.38745427e-02 -1.42531261e-01 7.77775228e-01
4.08946484e-01 -1.28392413e-01 -3.27753812e-01 2.53013819e-01
1.12543118e+00 -1.41672939e-01 -1.71555936e-01 -7.98624396e-01
-1.45089877e+00 6.86717987e-01 5.52575707e-01 3.00198972e-01
-2.83751160e-01 -1.45928696e-01 4.30165231e-01 6.49032518e-02
5.74962020e-01 5.27785301e-01 -6.43692017e-01 4.74039793e-01
-7.70925939e-01 9.68793333e-02 4.68358338e-01 1.12447441e+00
1.02976489e+00 1.15586976e-02 -1.43234134e-01 8.80437136e-01
1.19870566e-01 6.57951832e-01 4.17951196e-01 -9.77124214e-01
2.19695628e-01 8.49134743e-01 5.80067001e-02 -7.08058894e-01
-1.66546017e-01 -3.05227935e-01 -7.88523078e-01 1.90131627e-02
6.47232592e-01 1.30217895e-03 -1.21749222e+00 1.74900377e+00
4.29482132e-01 2.99678147e-01 8.98938850e-02 1.08981645e+00
8.33173752e-01 2.22019777e-01 3.40396583e-01 1.72528386e-01
1.37794375e+00 -8.98741603e-01 -3.03874940e-01 -5.25959015e-01
3.36647153e-01 -6.79537594e-01 1.20064950e+00 9.87549201e-02
-5.61806858e-01 -6.83210492e-01 -1.02210796e+00 -5.80673888e-02
-1.64912224e-01 1.34962574e-01 6.97533667e-01 3.52033615e-01
-5.55971086e-01 2.31338575e-01 -8.07768941e-01 -4.36270475e-01
8.34505677e-01 1.84816301e-01 -5.36403954e-01 -1.32478267e-01
-9.09134865e-01 4.45160896e-01 8.91707301e-01 -2.57893026e-01
-1.09837604e+00 -5.61011612e-01 -9.36978340e-01 -4.00186270e-01
5.91383636e-01 -8.76070380e-01 1.08973873e+00 -1.42358220e+00
-1.45519841e+00 1.50601649e+00 -9.23676938e-02 -2.39385054e-01
3.94893855e-01 1.86560676e-02 -1.16225168e-01 5.06175995e-01
4.23950672e-01 8.90584171e-01 8.85259151e-01 -1.62457395e+00
-4.99118924e-01 -4.17030632e-01 -2.28511612e-03 1.81242689e-01
-1.07398763e-01 -3.79924774e-01 -5.93175352e-01 -8.20591271e-01
2.64964670e-01 -1.03615010e+00 -3.51968467e-01 1.22673944e-01
-5.24274766e-01 1.43294066e-01 6.62312746e-01 -7.00559735e-01
5.41349053e-01 -2.04206848e+00 5.34153640e-01 1.34444445e-01
1.05485886e-01 2.25056469e-01 -2.76482463e-01 -8.56503844e-02
1.37453437e-01 -1.38023168e-01 -6.96505964e-01 5.25544100e-02
-1.55349374e-01 5.43901563e-01 -2.02476636e-01 3.68654788e-01
4.55411583e-01 1.29419351e+00 -9.14135993e-01 -8.12074065e-01
2.81097114e-01 1.40659541e-01 -3.73435706e-01 4.40656096e-01
-7.34388649e-01 1.19949782e+00 -9.97710109e-01 5.93018532e-01
7.50538468e-01 -5.14710724e-01 4.35825646e-01 -3.17980915e-01
3.26836318e-01 8.40170681e-02 -7.79828846e-01 2.07448339e+00
-4.39172179e-01 3.98365110e-01 6.28124699e-02 -1.41413951e+00
1.05772257e+00 -6.23872224e-03 4.59328592e-01 -9.27560747e-01
2.41233930e-01 1.52589142e-01 -3.45800996e-01 -5.04861593e-01
1.39558300e-01 -3.82922769e-01 -2.38136724e-01 3.12189370e-01
2.35405952e-01 -4.96932119e-01 -5.22714993e-03 5.07378206e-02
5.47046125e-01 3.63522530e-01 1.89398348e-01 -4.71866846e-01
7.51879096e-01 1.94321603e-01 6.99088573e-01 3.12293291e-01
-1.09261282e-01 7.47059345e-01 4.29045141e-01 -2.16409490e-01
-1.12968647e+00 -1.31748259e+00 -1.18296579e-01 1.09315073e+00
5.99600971e-01 4.29674268e-01 -1.08392894e+00 -1.03452861e+00
-1.67627353e-03 4.80869800e-01 -7.01271296e-01 -2.77210176e-02
-3.49065810e-01 -5.58894396e-01 3.98728341e-01 6.58676267e-01
8.26026559e-01 -9.79789793e-01 -2.52783209e-01 -4.52242531e-02
-4.64692563e-01 -1.76613784e+00 -3.94427687e-01 2.42434833e-02
-9.11756635e-01 -1.11943710e+00 -8.29708159e-01 -1.21848607e+00
1.02459729e+00 2.27002919e-01 1.18942261e+00 -1.77572012e-01
-2.40970030e-01 2.87479162e-01 -3.24938029e-01 -1.06815606e-01
-4.70251352e-01 1.90145031e-01 -3.67529571e-01 2.62021095e-01
4.29656431e-02 -4.99768764e-01 -7.32474327e-01 5.27512670e-01
-1.17068422e+00 3.30162495e-01 6.26269519e-01 8.76724184e-01
1.07149804e+00 -1.63752392e-01 5.43902814e-01 -1.40895772e+00
2.87251323e-01 -1.57210380e-01 -5.96014917e-01 4.54970360e-01
-3.34333569e-01 3.18658501e-01 5.46047866e-01 -3.39711487e-01
-1.37437069e+00 3.72280389e-01 1.48004442e-01 -2.81994939e-01
-5.31289697e-01 2.85733581e-01 -5.62823713e-01 -2.35379145e-01
5.92989445e-01 4.69032913e-01 3.04609220e-02 -4.16709602e-01
7.54267275e-01 3.24472755e-01 9.29616868e-01 -1.09368348e+00
7.29085207e-01 8.67410064e-01 3.54773588e-02 -7.59638131e-01
-1.13376439e+00 -5.39149940e-01 -1.20507109e+00 9.88224149e-02
1.30104995e+00 -1.12688529e+00 -1.41614467e-01 5.48760951e-01
-1.02708375e+00 -5.34601271e-01 -3.51901561e-01 1.11054942e-01
-9.84132230e-01 4.37782079e-01 -4.84556705e-01 -1.97256714e-01
7.32654482e-02 -1.06556082e+00 1.37199378e+00 2.40766719e-01
4.26101461e-02 -1.34492648e+00 -4.28137928e-02 6.62441790e-01
-1.21450759e-01 5.19199789e-01 9.71297443e-01 -5.09493232e-01
-7.57815182e-01 2.23238721e-01 -7.02010691e-01 5.96329153e-01
2.12664872e-01 -3.73684108e-01 -9.87863839e-01 -8.83726403e-02
-2.09652513e-01 -4.20043617e-01 8.82288516e-01 3.40179056e-01
1.23823667e+00 2.09079422e-02 -4.08920139e-01 7.73380399e-01
1.52119899e+00 -1.43407270e-01 4.29857850e-01 2.27785662e-01
9.07318950e-01 6.93864405e-01 7.83029497e-01 2.14318465e-02
4.10041869e-01 7.04034626e-01 1.99862093e-01 -4.67243105e-01
-3.78596127e-01 -4.53067601e-01 7.31314123e-02 6.72293305e-01
5.35874590e-02 -8.69684815e-02 -8.79239798e-01 6.93296313e-01
-1.70226443e+00 -3.71906936e-01 -3.69853191e-02 1.75509858e+00
8.81787539e-01 1.49436131e-01 7.68643767e-02 -2.57078648e-01
7.34196067e-01 1.14577457e-01 -7.37246871e-01 -1.20963119e-01
-2.34298438e-01 2.91441649e-01 7.52986670e-01 2.85760909e-01
-1.36260259e+00 1.39979291e+00 6.57831097e+00 8.67189765e-01
-1.12211668e+00 1.71970218e-01 7.58163333e-01 6.39655292e-01
-3.71971488e-01 -1.71177685e-01 -2.60865957e-01 2.99261332e-01
4.36482668e-01 -1.64504334e-01 2.74289966e-01 8.89221072e-01
-1.44536301e-01 -3.71676534e-02 -1.17711830e+00 6.90855205e-01
-7.31983110e-02 -1.24166143e+00 2.43329674e-01 -9.69629958e-02
1.27211070e+00 -1.11777455e-01 2.29962111e-01 -1.55594319e-01
5.25582254e-01 -8.85209739e-01 6.69362724e-01 3.68737578e-01
1.17213368e+00 -3.25571984e-01 4.96610850e-01 2.71761745e-01
-1.21659076e+00 3.46875280e-01 -3.36945593e-01 1.83018655e-01
1.20328620e-01 5.17907977e-01 -1.01267457e+00 8.24055791e-01
3.41335416e-01 1.00805688e+00 -4.47699815e-01 5.41174114e-01
-5.94074190e-01 5.11065722e-01 -2.00390331e-02 3.52276295e-01
2.51512498e-01 -4.04695690e-01 3.10812473e-01 1.16376758e+00
-6.70771971e-02 -2.34010555e-02 4.30234641e-01 1.11109221e+00
-1.47005543e-01 5.04816100e-02 -5.74314892e-01 -1.33809671e-01
2.89675951e-01 1.00741839e+00 -1.20654988e+00 -5.33044875e-01
-3.19192857e-01 1.33698559e+00 2.58540601e-01 6.48270965e-01
-6.29323840e-01 -3.73537838e-02 7.20420063e-01 1.25735030e-01
5.15905499e-01 -3.03528398e-01 -6.20212018e-01 -1.25645709e+00
-7.57706314e-02 -7.21185565e-01 2.11491972e-01 -7.64787197e-01
-1.61631358e+00 5.52076995e-01 7.67287193e-03 -1.19673300e+00
-1.16077736e-02 -8.71866822e-01 -2.42474452e-01 6.97136402e-01
-1.70394051e+00 -1.65933907e+00 -2.65444338e-01 7.19854891e-01
6.21562302e-01 -1.35258391e-01 7.15259552e-01 1.42268911e-01
-3.86651695e-01 3.41220945e-01 2.88044810e-01 3.46861094e-01
5.97561717e-01 -1.20644641e+00 4.66131240e-01 7.54079580e-01
2.34813720e-01 3.07110280e-01 4.97032553e-01 -6.20969415e-01
-1.03133154e+00 -1.40998936e+00 2.28797168e-01 -4.49142992e-01
6.06276393e-01 -4.30353612e-01 -9.52099979e-01 6.31652415e-01
-1.63233921e-01 1.66046381e-01 4.84806418e-01 -1.69383794e-01
-6.45023942e-01 3.01129539e-02 -1.21641588e+00 3.49107563e-01
1.21889603e+00 -7.32776999e-01 -7.48684108e-01 2.10463628e-01
8.37619305e-01 -3.60469848e-01 -7.79428780e-01 4.61446375e-01
3.41643870e-01 -8.16895485e-01 1.17586482e+00 -6.02723539e-01
5.52364767e-01 -3.40371072e-01 -2.07236603e-01 -1.14141595e+00
-7.64104649e-02 -7.64356777e-02 5.67932367e-01 1.19035149e+00
2.86232442e-01 -5.71388483e-01 8.99581075e-01 3.32935184e-01
1.45922437e-01 -2.15533882e-01 -8.10175300e-01 -8.47211003e-01
2.32065797e-01 -2.91876733e-01 5.14195621e-01 1.18462873e+00
-5.46070755e-01 3.18347752e-01 -1.17437519e-01 1.52073443e-01
7.22636402e-01 5.36672950e-01 7.33602345e-01 -1.41692233e+00
-2.28051186e-01 -2.47847766e-01 -5.38421273e-01 -1.30522740e+00
6.89066947e-01 -1.13048148e+00 2.25347236e-01 -1.52873790e+00
4.53370839e-01 -5.59577286e-01 -4.29373562e-01 3.02936435e-01
-1.21437952e-01 5.26302874e-01 8.12155288e-03 1.80900872e-01
-9.61913824e-01 6.43810570e-01 1.93805468e+00 -2.97210842e-01
1.67292908e-01 -2.64772445e-01 -8.08628023e-01 8.27864826e-01
4.57323134e-01 -4.43294048e-01 -6.33554280e-01 -4.70762342e-01
-2.27768645e-01 -7.55479857e-02 5.03322303e-01 -6.87619209e-01
-3.11948001e-01 -4.88762796e-01 2.36659259e-01 -1.47921890e-02
9.38285794e-03 -8.70282769e-01 -7.53815472e-02 2.70129740e-01
-3.93438011e-01 -6.20792747e-01 1.57925263e-01 7.84643829e-01
-5.13507485e-01 7.20442757e-02 9.31446195e-01 -1.19805172e-01
-1.06905127e+00 3.32277536e-01 4.95108403e-02 5.40632665e-01
1.15845454e+00 -2.24981830e-01 -1.73757404e-01 8.21686164e-03
-7.58346677e-01 1.14103518e-01 8.98420751e-01 3.70528340e-01
4.98399734e-01 -1.26602530e+00 -6.12093925e-01 3.78843039e-01
4.40008581e-01 3.90598536e-01 1.97930336e-01 4.32840645e-01
-6.58028364e-01 3.53079587e-01 -5.60075819e-01 -9.49808002e-01
-9.86545563e-01 4.18166727e-01 2.62252271e-01 -1.89904109e-01
-5.16610861e-01 8.19932818e-01 9.73395288e-01 -6.49497151e-01
-2.20369980e-01 -2.99107164e-01 1.02886014e-01 -2.31056720e-01
8.12579170e-02 -9.59796086e-02 -1.70720994e-01 -7.61380076e-01
-2.56422997e-01 9.78027582e-01 -1.72018576e-02 -4.04172465e-02
1.21350884e+00 -1.69110298e-01 -4.55906726e-02 3.36725771e-01
1.14282978e+00 -4.76453513e-01 -1.61972308e+00 -5.73619723e-01
1.02030888e-01 -5.60170293e-01 -1.33515745e-01 -7.83853590e-01
-1.16475928e+00 9.19741452e-01 5.40711761e-01 -2.57803261e-01
1.16040635e+00 4.22680318e-01 9.04268503e-01 1.23384289e-01
5.39621711e-01 -1.09881258e+00 3.33950937e-01 2.97233522e-01
6.03385866e-01 -1.54960716e+00 -1.81044996e-01 -9.03578460e-01
-1.03232276e+00 7.26453662e-01 5.76787472e-01 -2.33459577e-01
5.43197930e-01 -6.64362982e-02 1.01918317e-01 -1.35941342e-01
-2.84963995e-01 -5.45024514e-01 6.18221104e-01 1.08509779e+00
2.22405478e-01 3.26829702e-01 -1.39305830e-01 4.65624630e-01
1.98371857e-01 1.54049113e-01 3.35585400e-02 5.96588492e-01
-3.40673715e-01 -1.43135965e+00 -6.15994111e-02 8.16772282e-02
-2.12101787e-01 2.91131854e-01 -5.20252228e-01 7.11452961e-01
2.52985001e-01 4.85157907e-01 1.64178461e-02 -1.77745372e-01
4.11289573e-01 1.33629844e-01 7.46849954e-01 -8.59089613e-01
-7.44866505e-02 1.25775486e-01 -1.46477446e-01 -3.54441822e-01
-6.98827982e-01 -6.01892769e-01 -1.37640905e+00 1.80378288e-01
9.05034840e-02 -1.50742963e-01 5.12209833e-01 1.07824767e+00
3.39487255e-01 6.79163039e-01 4.27292258e-01 -6.48419023e-01
-1.79777667e-01 -6.03360951e-01 -6.95723832e-01 1.02362895e+00
2.13437006e-01 -9.02668118e-01 1.02704346e-01 7.03022540e-01] | [9.745046615600586, 1.1554534435272217] |
8c8409e0-8e75-4b6b-8905-65f4e0627965 | bof-ucb-a-bayesian-optimistic-frequentist | 2307.03587 | null | https://arxiv.org/abs/2307.03587v1 | https://arxiv.org/pdf/2307.03587v1.pdf | BOF-UCB: A Bayesian-Optimistic Frequentist Algorithm for Non-Stationary Contextual Bandits | We propose a novel Bayesian-Optimistic Frequentist Upper Confidence Bound (BOF-UCB) algorithm for stochastic contextual linear bandits in non-stationary environments. This unique combination of Bayesian and frequentist principles enhances adaptability and performance in dynamic settings. The BOF-UCB algorithm utilizes sequential Bayesian updates to infer the posterior distribution of the unknown regression parameter, and subsequently employs a frequentist approach to compute the Upper Confidence Bound (UCB) by maximizing the expected reward over the posterior distribution. We provide theoretical guarantees of BOF-UCB's performance and demonstrate its effectiveness in balancing exploration and exploitation on synthetic datasets and classical control tasks in a reinforcement learning setting. Our results show that BOF-UCB outperforms existing methods, making it a promising solution for sequential decision-making in non-stationary environments. | ['Melih Kandemir', 'Abdullah Akgül', 'Nicklas Werge'] | 2023-07-07 | null | null | null | null | ['multi-armed-bandits', 'decision-making'] | ['miscellaneous', 'reasoning'] | [-1.72266439e-02 1.47931064e-02 -1.00352848e+00 -1.53806776e-01
-9.42225575e-01 -3.60463679e-01 4.82427925e-01 1.13267034e-01
-5.35422325e-01 1.56557441e+00 -6.77736178e-02 -7.90823042e-01
-6.58898592e-01 -5.81167519e-01 -9.29648221e-01 -7.33124077e-01
-4.68627572e-01 7.46243417e-01 2.62468189e-01 1.90799534e-01
3.46041203e-01 2.06105024e-01 -1.28239584e+00 -2.26341128e-01
8.74082983e-01 1.18922687e+00 4.45030153e-01 6.34405434e-01
3.23775709e-01 4.03137147e-01 -4.35021609e-01 -2.49650553e-01
1.66060746e-01 -1.01889700e-01 -5.17965078e-01 -3.42209756e-01
-2.33404905e-01 -5.66114306e-01 3.16518068e-01 1.02404225e+00
1.98460668e-01 5.96816957e-01 7.22239971e-01 -1.13445985e+00
-2.06138894e-01 9.66137290e-01 -1.10542500e+00 6.50561154e-01
3.19006622e-01 5.67908101e-02 1.21414506e+00 -9.64301974e-02
3.64161134e-01 1.66029918e+00 6.17179692e-01 3.66265237e-01
-1.65843463e+00 -6.06699169e-01 8.74211729e-01 2.23841950e-01
-1.08950281e+00 -1.34748802e-01 2.05269605e-01 -1.67293638e-01
8.27061832e-01 3.90868396e-01 1.06302965e+00 1.23077250e+00
1.44429862e-01 1.26112831e+00 1.36158729e+00 -6.84001207e-01
9.35375333e-01 -2.06206776e-02 1.27620459e-01 4.03266758e-01
8.49812925e-01 7.46082604e-01 -7.86271274e-01 -5.48678160e-01
7.44111717e-01 -7.02153891e-02 -9.22780484e-02 -5.63555121e-01
-8.78166020e-01 1.03589487e+00 -8.00284520e-02 -4.30915982e-01
-6.92275524e-01 6.88522875e-01 9.65103954e-02 2.30416417e-01
4.86067414e-01 1.31175250e-01 -4.26934361e-01 -3.64307225e-01
-8.46170962e-01 7.53828108e-01 8.69752109e-01 9.75067139e-01
2.33497709e-01 2.48298347e-02 -6.07917786e-01 6.07051671e-01
6.25702977e-01 8.77271771e-01 3.66778582e-01 -1.14197230e+00
4.06337380e-01 -3.84181887e-01 1.15538824e+00 -5.11724234e-01
-2.75412630e-02 -5.43352008e-01 -1.03276134e-01 2.82735169e-01
4.58025873e-01 -3.66936952e-01 -6.63116992e-01 1.82214916e+00
6.87318385e-01 1.69955920e-02 -1.09293982e-01 7.73310244e-01
-3.87754202e-01 7.86371350e-01 5.73190115e-02 -7.13808298e-01
8.52378011e-01 -7.16118395e-01 -7.89099276e-01 -3.55628401e-01
8.55617076e-02 -1.90231457e-01 1.03061008e+00 6.92450941e-01
-1.00999212e+00 2.05452949e-01 -1.09961748e+00 8.31294298e-01
6.57966807e-02 -3.77498895e-01 8.24944079e-01 8.80293131e-01
-5.63736618e-01 5.63562930e-01 -1.13421071e+00 4.17525023e-02
4.63936776e-01 1.93308033e-02 5.44094324e-01 6.88201487e-02
-8.51688504e-01 7.49766767e-01 8.84878039e-01 2.04996914e-02
-1.28184378e+00 -5.18852711e-01 -4.35106456e-01 2.11199179e-01
1.17894363e+00 -6.20431960e-01 1.82159448e+00 -5.39407253e-01
-1.92775488e+00 1.09234750e-01 -8.85397345e-02 -1.11514366e+00
8.19651425e-01 -4.57962960e-01 1.77450310e-02 -2.01210469e-01
8.14951137e-02 6.92959353e-02 9.48393643e-01 -1.05145895e+00
-1.17399192e+00 -2.01947093e-01 3.35864052e-02 3.58951807e-01
1.53520526e-02 -4.28957671e-01 9.38111842e-02 -5.50017297e-01
-7.70255178e-02 -9.81759965e-01 -4.84139979e-01 -5.96486688e-01
-4.16435301e-01 -4.13011223e-01 3.20331246e-01 -9.83322859e-02
1.55662274e+00 -1.72734261e+00 -2.91261822e-01 6.31578684e-01
-4.53688085e-01 -2.73753136e-01 3.36357057e-01 2.23029047e-01
4.63562250e-01 3.14983726e-02 -1.12915471e-01 -3.41202617e-02
3.10915500e-01 3.85100484e-01 -6.31625414e-01 5.11472285e-01
-2.97559589e-01 5.16352475e-01 -1.15889812e+00 -2.97014326e-01
-7.02207163e-02 -3.82378131e-01 -6.73085451e-01 -4.04674234e-03
-7.10053682e-01 3.05289067e-02 -7.30607748e-01 5.97384214e-01
3.35507125e-01 -2.01314881e-01 6.88162446e-01 6.84837103e-01
-9.73965749e-02 1.33159235e-01 -1.52633524e+00 1.14866567e+00
-3.74813437e-01 1.13213830e-01 2.07829133e-01 -1.11983907e+00
5.53470016e-01 -1.78024992e-01 1.97414562e-01 -4.49843109e-01
-3.06903943e-02 2.96279013e-01 -3.15018266e-01 -2.05759466e-01
2.36870930e-01 -3.16970825e-01 -1.62051380e-01 7.85805583e-01
-1.99430093e-01 -1.76602125e-01 4.23107296e-01 3.10073923e-02
6.37717247e-01 5.72865665e-01 7.75970936e-01 -8.00326943e-01
-2.46682428e-02 -2.18711779e-01 7.58050799e-01 1.65602517e+00
-2.95436054e-01 -3.39651376e-01 6.67969942e-01 -3.71804416e-01
-6.12337470e-01 -1.24515998e+00 -3.19032013e-01 1.59918511e+00
2.29964182e-01 -4.46192771e-02 -4.49622452e-01 -5.49501956e-01
6.68875277e-01 1.33962381e+00 -7.44138837e-01 -4.08889614e-02
-1.06465511e-01 -1.03455317e+00 -3.36714476e-01 4.01755542e-01
1.79805294e-01 -5.62166750e-01 -1.20561957e+00 4.73982751e-01
-4.00332399e-02 -4.84951586e-01 -4.83387262e-01 4.86430019e-01
-8.67760360e-01 -8.46493065e-01 -5.70182979e-01 2.45689049e-01
1.87338591e-01 7.05849156e-02 8.96636844e-01 -5.67404211e-01
7.14399144e-02 3.85225326e-01 -2.08944380e-02 -9.82354403e-01
-1.07889093e-01 -1.28744528e-01 2.80834377e-01 -2.81950012e-02
1.11398675e-01 -4.68575627e-01 -7.26398468e-01 3.34432811e-01
-4.76453274e-01 -2.64477432e-01 2.47709274e-01 1.08964694e+00
8.26890528e-01 -8.08905661e-02 8.37386131e-01 -8.24039757e-01
1.01716626e+00 -6.57716632e-01 -1.49690843e+00 4.62172121e-01
-9.99101341e-01 3.39206755e-01 -5.77165261e-02 -8.26452494e-01
-1.54392767e+00 -1.57901928e-01 5.47877967e-01 -2.06611633e-01
4.69631881e-01 6.81153655e-01 3.54457438e-01 4.78805363e-01
6.47965848e-01 -4.47904319e-02 -1.94635674e-01 -3.85819376e-01
2.45651737e-01 5.34009993e-01 4.17877555e-01 -1.35289717e+00
-2.40256102e-03 4.23911244e-01 8.24989378e-02 -3.73308599e-01
-1.26837564e+00 -2.38203049e-01 1.90461606e-01 -2.62325138e-01
2.03704864e-01 -7.41569996e-01 -1.30930030e+00 -8.09628293e-02
-5.10313988e-01 -7.97361314e-01 -4.79513407e-01 8.21156681e-01
-1.26897871e+00 4.33479361e-02 -1.92228973e-01 -2.03016019e+00
-9.05223712e-02 -8.56952786e-01 7.39422560e-01 3.94470215e-01
-2.47772545e-01 -7.73436129e-01 1.34497806e-01 1.51672333e-01
7.31630102e-02 3.39315057e-01 7.55025566e-01 -4.83168989e-01
-5.30208111e-01 1.52841462e-02 1.39471561e-01 -1.52844444e-01
-2.81810939e-01 -2.38037780e-02 -4.26408201e-01 -5.68328023e-01
-2.53172010e-01 -6.13865018e-01 9.88883615e-01 1.09658718e+00
1.03554559e+00 -7.21843839e-01 -5.19414783e-01 3.42731297e-01
1.16516650e+00 4.41922337e-01 -6.07980452e-02 7.42105186e-01
-3.93408626e-01 2.53545702e-01 1.26140273e+00 1.24128795e+00
1.32206142e-01 4.86762315e-01 5.97302377e-01 7.68284619e-01
6.73684716e-01 -4.56802279e-01 3.72824401e-01 -2.84031272e-01
-9.37666297e-02 -1.67591944e-01 -6.55853629e-01 7.39175558e-01
-2.44449854e+00 -1.13097739e+00 4.28575009e-01 2.57796097e+00
1.21633410e+00 5.26018918e-01 6.48816466e-01 -2.25328207e-01
8.38852286e-01 -1.86963528e-01 -1.17594922e+00 -5.91624975e-01
7.45996088e-02 -1.47687286e-01 7.19037592e-01 6.92949891e-01
-8.41536999e-01 6.05718493e-01 7.37768269e+00 1.08033311e+00
-5.67548692e-01 1.00376755e-01 8.61332893e-01 -7.81063378e-01
-2.37991706e-01 -1.35700673e-01 -1.22461438e+00 6.34396791e-01
1.15609503e+00 -5.15592515e-01 6.70104980e-01 1.18841660e+00
2.23342761e-01 -7.78433859e-01 -1.03811753e+00 7.06528246e-01
-6.64858997e-01 -1.39643121e+00 -3.62896472e-01 8.85274038e-02
1.10648584e+00 5.48780411e-02 2.49598667e-01 3.82803023e-01
1.51766634e+00 -6.88725531e-01 1.28460157e+00 5.90964556e-01
6.74148798e-01 -1.13723195e+00 4.30880755e-01 6.63910627e-01
-4.47835922e-01 -7.68405437e-01 -3.16792965e-01 -5.83933778e-02
1.81182161e-01 6.56222880e-01 -9.50839698e-01 6.31833449e-02
9.64347303e-01 1.87881783e-01 3.62030059e-01 1.11997962e+00
-3.26515555e-01 7.73572862e-01 -9.42137837e-01 -6.48905754e-01
6.14678800e-01 -3.81835878e-01 5.83295524e-01 1.10412955e+00
2.81564027e-01 -1.83275551e-01 4.48757499e-01 8.55047107e-01
2.40609199e-01 -9.17847306e-02 -8.65573436e-02 4.98150215e-02
9.51266408e-01 5.63320875e-01 -7.48896182e-01 -4.30284441e-01
2.55802214e-01 1.44316956e-01 4.10095483e-01 5.53736269e-01
-8.27822268e-01 -9.59839225e-02 5.96237957e-01 -3.40557992e-01
7.53078163e-01 1.01204708e-01 -2.67074257e-01 -7.98158765e-01
-1.31366938e-01 -6.72725976e-01 1.01565695e+00 -4.11355376e-01
-9.76784706e-01 -9.53055546e-02 7.01926053e-01 -4.83948857e-01
-7.01971412e-01 -3.08381766e-01 -2.48616472e-01 6.74181759e-01
-1.46696198e+00 -5.76930523e-01 5.59001148e-01 4.57987815e-01
4.99100387e-01 1.12835556e-01 4.42160815e-01 -6.14328325e-01
-7.36361086e-01 3.08528543e-01 9.84055102e-01 -8.42359006e-01
3.04900378e-01 -1.43406582e+00 -4.34134677e-02 6.55847967e-01
-1.92697704e-01 5.66381574e-01 1.13194680e+00 -9.08062458e-01
-1.16977561e+00 -6.20521903e-01 -7.63725638e-02 -2.08210051e-02
9.41440046e-01 -1.25003755e-01 -2.49894351e-01 8.25625181e-01
-1.69621691e-01 -1.11837059e-01 6.03656650e-01 5.85016549e-01
-2.99159288e-01 -3.33320379e-01 -1.25842226e+00 6.94630802e-01
8.79296005e-01 6.50233775e-02 -4.20742840e-01 4.54220980e-01
6.73809946e-01 -3.18171382e-01 -6.58974409e-01 2.33076751e-01
9.96254325e-01 -7.66218066e-01 8.40216458e-01 -8.73002052e-01
-3.00091822e-02 8.32080096e-02 -3.88772845e-01 -1.51816952e+00
-2.26297349e-01 -1.36994529e+00 -7.55109191e-01 5.94008029e-01
5.39990842e-01 -8.63604367e-01 6.28334522e-01 3.35684657e-01
2.75697857e-01 -9.49877560e-01 -1.31628168e+00 -1.06834674e+00
1.89959742e-02 -5.44116259e-01 6.86953783e-01 4.16433513e-01
1.06591292e-01 -1.98717251e-01 -6.56339586e-01 -2.31377650e-02
1.02478766e+00 6.29908681e-01 4.93106931e-01 -1.07966757e+00
-8.24164808e-01 -5.37253499e-01 5.14304996e-01 -1.10092306e+00
6.09195046e-02 -1.96375355e-01 3.49627793e-01 -9.62559760e-01
2.68001080e-01 -4.52554345e-01 -5.45909524e-01 2.26794362e-01
-1.81036085e-01 -2.83742845e-01 3.37199047e-02 -6.27225917e-03
-9.36955333e-01 5.27499437e-01 1.02708960e+00 -5.89184985e-02
-3.82366240e-01 5.53201616e-01 -7.75184095e-01 6.96241796e-01
6.38284802e-01 -5.39131820e-01 -5.26027977e-01 1.23425320e-01
4.68317628e-01 3.90213311e-01 -1.13111243e-01 -3.54144543e-01
2.63945963e-02 -1.09763753e+00 3.07008058e-01 -8.09163511e-01
1.79524779e-01 -5.63408911e-01 2.55282037e-02 7.77715623e-01
-5.79142034e-01 -2.82428533e-01 6.68144375e-02 1.42408228e+00
4.74614501e-01 -4.86227781e-01 8.07148695e-01 -1.86190411e-01
-3.91185492e-01 -6.30431622e-02 -6.65094674e-01 -6.17858022e-03
1.13118458e+00 -5.93408607e-02 -1.28959745e-01 -6.06151044e-01
-8.25827599e-01 7.28108168e-01 -1.20116509e-01 -1.90803930e-01
3.74657661e-01 -1.07578933e+00 -5.98656058e-01 -5.05715683e-02
-9.06482637e-02 -2.14934591e-02 -7.68350437e-02 7.34399080e-01
-1.41564488e-01 5.98042727e-01 6.92033693e-02 -5.41197896e-01
-8.79761100e-01 5.89521468e-01 2.42156938e-01 -7.34580815e-01
-1.90361097e-01 9.25828338e-01 -1.11935258e-01 4.82538380e-02
5.31010807e-01 -3.04875016e-01 2.67990232e-01 2.31596511e-02
7.04351366e-01 7.50869453e-01 -2.88805157e-01 4.00711536e-01
-1.32405788e-01 -3.97510797e-01 -2.47770682e-01 -6.47189140e-01
1.43049419e+00 -5.29442012e-01 2.55218595e-01 7.27509558e-01
2.52323091e-01 -2.61213154e-01 -1.82699060e+00 -4.99087453e-01
5.02152503e-01 -7.04618573e-01 4.50793386e-01 -1.11124527e+00
-3.43055069e-01 1.90506473e-01 3.76938760e-01 3.76142353e-01
7.23473251e-01 -1.49569973e-01 1.93713650e-01 8.29944730e-01
6.93527579e-01 -1.53299212e+00 -1.66240454e-01 3.75580251e-01
7.83288419e-01 -9.37432468e-01 5.52343309e-01 1.36501715e-01
-6.69369161e-01 8.11539412e-01 2.87015319e-01 1.10043800e-02
6.07847810e-01 3.14066410e-01 -5.59275806e-01 4.37165111e-01
-1.23443627e+00 -2.08250970e-01 -2.15197895e-02 5.49349725e-01
-1.83765277e-01 4.87468898e-01 -4.45072502e-01 6.24438047e-01
-2.80862212e-01 5.33986511e-03 3.19662094e-01 1.15045822e+00
-7.37565517e-01 -7.13877439e-01 -7.35873997e-01 5.94059110e-01
-7.03235805e-01 1.26887292e-01 1.33357510e-01 6.63146436e-01
-5.75557232e-01 9.15531456e-01 5.35235479e-02 2.87456602e-01
-6.32316247e-02 -8.63591209e-02 6.65828168e-01 -2.70351261e-01
2.49931403e-02 6.76039159e-01 3.26622605e-01 -7.11738467e-01
-1.51799455e-01 -1.04023302e+00 -6.28998220e-01 -2.68730432e-01
-6.66121066e-01 5.60272694e-01 7.64875352e-01 7.00822413e-01
4.84614745e-02 2.28351727e-01 6.30746841e-01 -7.01178372e-01
-1.64060342e+00 -8.84784400e-01 -8.55216980e-01 -3.07387784e-02
4.76945728e-01 -1.06388092e+00 -2.40272522e-01 -3.49199355e-01] | [4.500918388366699, 3.196922540664673] |
2139081a-bf36-455b-85be-9e8784d52276 | converting-the-point-of-view-of-messages | 2010.02600 | null | https://arxiv.org/abs/2010.02600v2 | https://arxiv.org/pdf/2010.02600v2.pdf | Converting the Point of View of Messages Spoken to Virtual Assistants | Virtual Assistants can be quite literal at times. If the user says "tell Bob I love him," most virtual assistants will extract the message "I love him" and send it to the user's contact named Bob, rather than properly converting the message to "I love you." We designed a system to allow virtual assistants to take a voice message from one user, convert the point of view of the message, and then deliver the result to its target user. We developed a rule-based model, which integrates a linear text classification model, part-of-speech tagging, and constituency parsing with rule-based transformation methods. We also investigated Neural Machine Translation (NMT) approaches, including LSTMs, CopyNet, and T5. We explored 5 metrics to gauge both naturalness and faithfulness automatically, and we chose to use BLEU plus METEOR for faithfulness and relative perplexity using a separately trained language model (GPT) for naturalness. Transformer-Copynet and T5 performed similarly on faithfulness metrics, with T5 achieving slight edge, a BLEU score of 63.8 and a METEOR score of 83.0. CopyNet was the most natural, with a relative perplexity of 1.59. CopyNet also has 37 times fewer parameters than T5. We have publicly released our dataset, which is composed of 46,565 crowd-sourced samples. | ['Jack G. M. FitzGerald', 'Dennis Liang', 'Sai Srujana Buddi', 'Vera Zu', 'Isabelle G. Lee'] | 2020-10-06 | null | https://aclanthology.org/2020.findings-emnlp.15 | https://aclanthology.org/2020.findings-emnlp.15.pdf | findings-of-the-association-for-computational | ['constituency-parsing'] | ['natural-language-processing'] | [ 1.02167241e-01 4.25796181e-01 -9.67185199e-02 -6.86458349e-01
-1.05410397e+00 -8.16937149e-01 8.81558120e-01 -1.62254229e-01
-5.84639907e-01 7.90049434e-01 5.53241313e-01 -8.04318547e-01
4.92098600e-01 -6.94406748e-01 -5.13531804e-01 -1.68148503e-01
5.73495448e-01 8.24536204e-01 -2.66469926e-01 -5.00652015e-01
-2.37767160e-01 2.76504010e-01 -8.19047689e-01 3.95953447e-01
7.91200042e-01 6.30070448e-01 1.28092468e-01 8.87716591e-01
-2.82157570e-01 9.34965670e-01 -6.77189410e-01 -9.64081466e-01
9.62773431e-03 -3.59697759e-01 -1.14014435e+00 -1.27958298e-01
4.97661710e-01 -4.52290118e-01 -5.21194577e-01 8.05158317e-01
4.70413566e-01 2.88632721e-01 4.07616168e-01 -1.13441229e+00
-7.66961157e-01 8.36375356e-01 -2.57844292e-02 -8.97445977e-02
6.66370809e-01 2.51078039e-01 1.27447915e+00 -7.18417764e-01
7.84057200e-01 1.38972783e+00 7.77056038e-01 8.18575144e-01
-1.35044241e+00 -4.73735392e-01 -1.47978604e-01 -2.37251729e-01
-9.58401978e-01 -7.20410526e-01 3.59426975e-01 -3.33248496e-01
1.41694772e+00 4.66730654e-01 3.49875391e-01 1.61233497e+00
2.60838687e-01 6.08555496e-01 1.09920740e+00 -5.07716179e-01
2.28626207e-02 5.67637324e-01 1.23482861e-01 8.32441688e-01
-2.32858658e-01 -1.96446106e-01 -4.55516070e-01 -4.30112034e-01
5.29907227e-01 -2.94098649e-02 -1.62101954e-01 2.46271297e-01
-1.37466586e+00 8.71942282e-01 1.88974142e-01 3.17423671e-01
-3.92884612e-01 2.54047140e-02 3.67574960e-01 5.48771620e-01
3.73760372e-01 3.71216089e-01 -4.95091438e-01 -4.11751658e-01
-7.33949006e-01 2.08947703e-01 1.30743647e+00 1.25161171e+00
5.13256550e-01 -1.38651013e-01 -3.99132937e-01 1.18000174e+00
2.04829797e-01 7.98959970e-01 4.80906814e-01 -1.31072760e+00
6.78693891e-01 3.06808919e-01 2.26242170e-01 -6.32980883e-01
-2.76673704e-01 -1.25037163e-01 -6.30118608e-01 -3.33817542e-01
3.86837959e-01 -3.16644907e-01 -8.18587720e-01 1.83682501e+00
-1.61569148e-01 -7.59863734e-01 6.14832118e-02 6.55674458e-01
8.64465475e-01 8.24577868e-01 -8.31244141e-03 -2.42273778e-01
1.38253236e+00 -9.70870674e-01 -7.94342995e-01 -5.50590932e-01
6.31082833e-01 -1.05143523e+00 1.59807909e+00 1.02880679e-01
-1.20936441e+00 -3.56643945e-01 -7.72034168e-01 -3.29428256e-01
-1.82752043e-01 8.70275795e-02 4.72104073e-01 6.33432508e-01
-1.45363760e+00 6.91295266e-01 -7.48766899e-01 -9.86990690e-01
-5.57745434e-02 2.47242212e-01 -4.66086328e-01 6.74163252e-02
-1.20848238e+00 1.25339627e+00 -7.37577453e-02 -2.97495127e-01
-6.99410856e-01 -2.02842653e-01 -8.73260379e-01 -2.54747942e-02
-1.25470385e-01 -7.74882197e-01 1.86894655e+00 -8.37444603e-01
-1.73575997e+00 9.90592301e-01 -5.05709946e-01 -3.90752524e-01
5.59208095e-01 1.37844132e-02 -4.50019211e-01 -7.97539577e-02
2.46238157e-01 7.38493264e-01 5.70300102e-01 -1.09930766e+00
-5.35804749e-01 -3.55112851e-01 6.70053065e-02 -1.90003135e-03
-4.44626957e-01 3.79291415e-01 -3.31046283e-01 -3.26400638e-01
3.77407260e-02 -1.07278740e+00 2.21066207e-01 -2.37834871e-01
-3.72952938e-01 -2.37636566e-01 4.01625156e-01 -1.03241217e+00
1.05976188e+00 -1.89020753e+00 -1.52853847e-01 -1.42580599e-01
2.20847383e-01 2.36187831e-01 -3.55693758e-01 4.44710255e-01
2.31300533e-01 1.76904738e-01 -1.74496353e-01 -6.45188868e-01
1.63048699e-01 4.56684917e-01 -5.05734563e-01 -4.83964868e-05
7.96942189e-02 9.98533368e-01 -8.27958643e-01 -5.30822992e-01
-2.33524386e-03 4.96380389e-01 -6.56981707e-01 4.12958473e-01
-1.42992198e-01 -1.25618726e-01 -2.54284907e-02 4.72617418e-01
3.43744397e-01 -8.80973861e-02 3.34748358e-01 1.20610014e-01
3.75376968e-03 9.09726679e-01 -4.27366853e-01 1.38418353e+00
-9.25758779e-01 9.62791979e-01 2.93702334e-01 -2.76457727e-01
1.07171512e+00 4.78357524e-01 5.62203974e-02 -5.62486649e-01
3.01520407e-01 1.29185840e-01 -1.97660267e-01 -5.00279188e-01
8.49436641e-01 -3.86268824e-01 -2.18537629e-01 6.99285269e-01
1.65956572e-01 -2.85583496e-01 -9.33060646e-02 4.34473902e-01
1.29692256e+00 1.59994870e-01 8.87508020e-02 9.81623977e-02
2.50105038e-02 7.50029013e-02 3.43316853e-01 7.90076554e-01
-3.10030341e-01 5.40710032e-01 4.71603066e-01 -2.02512115e-01
-1.24058294e+00 -1.14734352e+00 2.35235631e-01 1.51287782e+00
-3.55292857e-01 -4.49394107e-01 -9.59619820e-01 -6.09213889e-01
-2.35964030e-01 1.44745922e+00 -2.35331371e-01 -6.62330240e-02
-5.96613050e-01 -3.82541329e-01 8.76170933e-01 3.38025093e-01
4.78062749e-01 -1.15269518e+00 -5.92632703e-02 3.23930472e-01
-1.09399974e+00 -1.32598388e+00 -7.70210683e-01 2.67855853e-01
-8.12015116e-01 -3.33316714e-01 -3.95685792e-01 -7.76853323e-01
3.31850290e-01 2.96849579e-01 1.19181013e+00 -2.44373322e-01
3.92042845e-01 2.05307931e-01 -3.12773615e-01 -2.05942050e-01
-9.38967943e-01 2.57353723e-01 2.48459920e-01 -3.23561013e-01
6.44224942e-01 -5.64872742e-01 3.39870751e-02 3.50536704e-01
-6.14599943e-01 2.36874633e-02 4.96459812e-01 8.41369689e-01
1.46445204e-02 -6.39743805e-01 1.86371431e-01 -7.84271955e-01
1.03085935e+00 -1.41823620e-01 5.00445105e-02 2.48439759e-01
-5.57825804e-01 5.70182800e-02 8.88183296e-01 -4.51075763e-01
-1.17395639e+00 -1.11404523e-01 -3.58444095e-01 -1.11882076e-01
-2.93820836e-02 5.72483018e-02 -2.34452069e-01 3.51317674e-01
9.00406539e-01 2.57264018e-01 1.21526308e-01 -4.64022964e-01
5.19447625e-01 1.40221858e+00 8.00155163e-01 -6.45389795e-01
4.68745351e-01 1.62558034e-01 -8.49493742e-01 -8.00975025e-01
-7.09537506e-01 -1.86422452e-01 -4.76108283e-01 8.19057375e-02
6.99630141e-01 -7.76201785e-01 -7.79921293e-01 3.07851195e-01
-1.70262051e+00 -4.94679391e-01 -2.94835307e-02 4.72809851e-01
-6.10905647e-01 3.62401545e-01 -9.98072743e-01 -6.15095556e-01
-7.31973112e-01 -9.15151596e-01 7.85389364e-01 -9.29698870e-02
-9.27385747e-01 -9.52063859e-01 6.93212226e-02 8.63559842e-01
6.00656509e-01 -2.50987649e-01 1.04374456e+00 -1.00459397e+00
5.79148196e-02 -1.71635255e-01 -3.21504772e-01 4.42835152e-01
1.59876645e-01 1.21758856e-01 -1.16140962e+00 -1.62918434e-01
6.92601353e-02 -2.74450153e-01 5.93754888e-01 9.42160040e-02
3.60990822e-01 -8.49162817e-01 -1.19302884e-01 3.98187757e-01
8.92177582e-01 2.15889245e-01 6.11819506e-01 2.94550180e-01
6.04049921e-01 5.34723938e-01 2.68005908e-01 2.78200030e-01
7.41528630e-01 7.00789332e-01 3.28458026e-02 1.33028209e-01
-4.54998761e-02 -7.68463671e-01 8.54822934e-01 1.13224077e+00
-3.12501602e-02 -3.46902162e-01 -8.64096045e-01 3.12380135e-01
-1.53899896e+00 -9.93379831e-01 -3.28718424e-02 1.94102144e+00
1.10177970e+00 3.37375224e-01 3.48859169e-02 -3.17134798e-01
7.83552825e-01 1.52382240e-01 -1.34514153e-01 -9.42036331e-01
1.30609006e-01 2.99489293e-02 5.09804606e-01 1.02862847e+00
-7.52344370e-01 1.41488314e+00 6.73795509e+00 4.31798130e-01
-1.15741789e+00 3.00626099e-01 4.68773246e-01 -1.85507342e-01
-5.36525011e-01 1.06144287e-01 -8.24338973e-01 4.39151853e-01
1.32426524e+00 6.85969368e-02 8.63828421e-01 8.23530495e-01
4.60302562e-01 1.76907137e-01 -1.28629029e+00 7.78969944e-01
9.26371664e-02 -8.83705854e-01 1.80473864e-01 2.45549604e-02
3.67389530e-01 3.48163247e-01 -1.98273763e-01 6.16202772e-01
7.07886577e-01 -9.58891332e-01 7.57002115e-01 4.72460628e-01
9.70523596e-01 -3.99955720e-01 6.25980139e-01 4.79222506e-01
-6.49151564e-01 1.74878091e-01 -3.63793969e-01 -3.17554563e-01
1.42132029e-01 2.62695104e-01 -1.34833324e+00 1.61196999e-02
5.28060913e-01 1.53072700e-01 -1.47557631e-01 9.56165642e-02
-3.80698472e-01 6.51416898e-01 -3.46069574e-01 -4.59463745e-01
2.27764666e-01 -2.24893600e-01 5.84710538e-01 1.46281576e+00
2.56495357e-01 1.90920889e-01 -1.64728582e-01 9.55310404e-01
-4.55876768e-01 9.19342507e-03 -6.54459715e-01 -2.07874566e-01
7.65282452e-01 1.37261200e+00 -3.16586375e-01 -5.45098960e-01
-3.07951391e-01 1.37481940e+00 3.06062102e-01 3.68261933e-01
-7.79182673e-01 -5.98676145e-01 7.20648050e-01 2.16697097e-01
-1.41940966e-01 -3.05916369e-01 -3.76783490e-01 -1.16905081e+00
1.95665017e-01 -1.12755764e+00 -3.68446894e-02 -1.01506412e+00
-1.07309508e+00 1.04589283e+00 -3.44636202e-01 -7.62341261e-01
-5.65148056e-01 -5.42735755e-01 -7.39600062e-01 9.44733620e-01
-9.21512485e-01 -1.22820950e+00 -1.98818088e-01 3.80107522e-01
5.94618917e-01 -3.24985385e-01 1.11793053e+00 1.70756236e-01
-4.52839911e-01 8.83217454e-01 7.90748186e-03 3.35922033e-01
9.45558786e-01 -1.06887388e+00 1.01482916e+00 6.27938569e-01
-4.47370559e-02 1.02628374e+00 7.84666240e-01 -5.97080231e-01
-1.30087531e+00 -1.08997118e+00 1.98113739e+00 -8.07686865e-01
7.56408036e-01 -5.20353317e-01 -7.71940827e-01 1.12545693e+00
2.46882975e-01 -6.09271049e-01 5.61866045e-01 1.29246980e-01
-4.68237817e-01 4.53250073e-02 -1.31228054e+00 7.70413816e-01
1.07615268e+00 -9.37617719e-01 -8.66876185e-01 4.57653999e-01
1.21748316e+00 -1.53225049e-01 -7.54561841e-01 -7.75733292e-02
6.27758801e-01 -8.08226705e-01 5.80655754e-01 -5.87242007e-01
4.25106287e-01 2.11624414e-01 -4.88332003e-01 -1.40428185e+00
-4.97736394e-01 -9.21362996e-01 2.67384946e-01 1.48839760e+00
7.59949148e-01 -8.92937362e-01 4.96020854e-01 1.07487690e+00
-6.39476776e-02 -3.30101401e-01 -7.39411592e-01 -6.40615880e-01
1.47433460e-01 -5.92414677e-01 5.40322423e-01 9.88884926e-01
4.17781949e-01 7.76347041e-01 -4.03774679e-01 -3.95858228e-01
2.34407082e-01 -3.07144552e-01 6.41918361e-01 -9.20737982e-01
-3.29434067e-01 -4.26852077e-01 2.50982344e-01 -1.20534801e+00
2.44568288e-01 -1.21015954e+00 1.55380324e-01 -1.75827920e+00
2.22797692e-01 -2.85332143e-01 3.11887950e-01 9.33735430e-01
6.62551746e-02 6.58811182e-02 1.57803714e-01 3.99415970e-01
-1.48159578e-01 4.57013547e-01 9.90610957e-01 -2.37505972e-01
-5.13873041e-01 1.30305201e-01 -9.70565498e-01 6.90484047e-01
8.64213347e-01 -4.40505147e-01 -3.16116288e-02 -5.42236686e-01
4.19496596e-02 3.12736839e-01 2.86975771e-01 -5.75788379e-01
1.19641408e-01 -1.75432950e-01 1.98887989e-01 -1.40343696e-01
5.20614922e-01 -4.78899270e-01 -7.82804191e-02 3.67104173e-01
-5.79536140e-01 5.68952709e-02 -1.01470547e-02 -7.94697180e-02
1.50664583e-01 -1.52640879e-01 7.18592763e-01 -1.34657383e-01
-8.07591453e-02 -5.25773950e-02 -6.45782709e-01 -3.88083514e-03
3.43088597e-01 5.04978932e-02 -4.37508434e-01 -8.52631807e-01
-4.68645304e-01 4.84437943e-02 5.74850619e-01 4.86026347e-01
5.32529652e-01 -1.34656501e+00 -7.57122755e-01 1.81156605e-01
-4.57936898e-04 -5.31240225e-01 -3.42674375e-01 6.63098156e-01
-4.95053411e-01 5.49268305e-01 -1.08571216e-01 -2.55474120e-01
-1.31413186e+00 1.86430827e-01 3.28256786e-01 -1.56313051e-02
-4.28106189e-01 7.27962911e-01 -3.43149304e-01 -1.20475245e+00
2.97354966e-01 -4.53800529e-01 1.43525064e-01 -3.72731872e-02
4.11821336e-01 4.45723653e-01 8.86091217e-02 -5.99124134e-01
-3.61445546e-01 5.24024740e-02 -1.93588674e-01 -7.01441526e-01
1.05546176e+00 -1.17683783e-01 -3.60807478e-01 3.48480999e-01
1.34464693e+00 3.29803675e-02 -7.25494504e-01 -4.68552858e-01
-1.75863206e-01 -2.87831664e-01 -1.81302547e-01 -8.33843529e-01
-4.80278641e-01 8.19306374e-01 9.16301608e-02 2.28564575e-01
6.36163175e-01 3.10944747e-02 1.28676772e+00 8.43214571e-01
3.96544427e-01 -1.31790972e+00 -9.81202796e-02 1.00122476e+00
8.31308067e-01 -1.18132043e+00 -3.66583467e-01 -1.04801785e-02
-9.60868955e-01 9.25669134e-01 3.48406196e-01 4.22585189e-01
1.68935984e-01 3.46611917e-01 3.83047462e-01 1.52004391e-01
-9.45105970e-01 1.91993460e-01 -1.39979839e-01 6.24690175e-01
7.95347214e-01 4.47949082e-01 -1.92663725e-02 7.18941271e-01
-9.87948537e-01 2.07222383e-02 5.53191304e-01 6.48763061e-01
-6.75449610e-01 -1.05997062e+00 -3.73358190e-01 3.62420350e-01
-3.03437978e-01 -3.95141602e-01 -9.18378949e-01 4.18469429e-01
-3.38978142e-01 1.41773474e+00 2.71622781e-02 -9.86300170e-01
4.44084674e-01 4.44032192e-01 2.36849144e-01 -5.00164449e-01
-8.02850068e-01 -2.01717824e-01 6.22272730e-01 -3.15823019e-01
1.06949307e-01 -4.86917228e-01 -1.41179907e+00 -1.07270002e+00
7.54240155e-03 1.72747195e-01 8.09786797e-01 9.04765844e-01
2.14994848e-01 1.68065410e-02 7.66654193e-01 -5.81343949e-01
-8.57403040e-01 -1.35101211e+00 -1.27133653e-01 2.24410906e-01
1.85171127e-01 1.65616006e-01 -3.39355439e-01 2.78828014e-02] | [14.374934196472168, 7.243810653686523] |
ee8c0be8-92ce-4516-8dc1-4d8d722884c7 | lexglue-a-benchmark-dataset-for-legal | 2110.00976 | null | https://arxiv.org/abs/2110.00976v4 | https://arxiv.org/pdf/2110.00976v4.pdf | LexGLUE: A Benchmark Dataset for Legal Language Understanding in English | Laws and their interpretations, legal arguments and agreements\ are typically expressed in writing, leading to the production of vast corpora of legal text. Their analysis, which is at the center of legal practice, becomes increasingly elaborate as these collections grow in size. Natural language understanding (NLU) technologies can be a valuable tool to support legal practitioners in these endeavors. Their usefulness, however, largely depends on whether current state-of-the-art models can generalize across various tasks in the legal domain. To answer this currently open question, we introduce the Legal General Language Understanding Evaluation (LexGLUE) benchmark, a collection of datasets for evaluating model performance across a diverse set of legal NLU tasks in a standardized way. We also provide an evaluation and analysis of several generic and legal-oriented models demonstrating that the latter consistently offer performance improvements across multiple tasks. | ['Nikolaos Aletras', 'Daniel Martin Katz', 'Ion Androutsopoulos', 'Michael Bommarito', 'Dirk Hartung', 'Abhik Jana', 'Ilias Chalkidis'] | 2021-10-03 | null | https://aclanthology.org/2022.acl-long.297 | https://aclanthology.org/2022.acl-long.297.pdf | acl-2022-5 | ['multiple-choice-qa'] | ['natural-language-processing'] | [ 1.54914588e-01 2.09068358e-01 -8.68510425e-01 -5.39669037e-01
-1.28733325e+00 -9.04532909e-01 8.76323164e-01 2.65449524e-01
-3.12624127e-01 9.00699079e-01 5.98386049e-01 -8.46730649e-01
-2.96637118e-01 -4.00263846e-01 -5.77057660e-01 3.84467803e-02
2.13619247e-01 7.34819829e-01 7.55284503e-02 -4.45744544e-01
3.25058639e-01 2.63271064e-01 -1.00103915e+00 5.99046826e-01
1.47128892e+00 5.27794123e-01 -6.29063666e-01 2.83263505e-01
-4.89082128e-01 1.26249290e+00 -4.29867327e-01 -1.10673237e+00
3.28654796e-01 -1.42675921e-01 -1.11573613e+00 -4.98416781e-01
9.42146599e-01 -4.56923604e-01 -1.24014355e-01 8.72165561e-01
1.74152225e-01 -2.63710201e-01 6.87722921e-01 -1.04636788e+00
-8.74737799e-01 7.39575863e-01 -4.42701340e-01 3.47191572e-01
6.96862042e-01 1.11580230e-02 1.60161555e+00 -1.91140935e-01
1.30464017e+00 1.30000186e+00 7.08663583e-01 7.02845931e-01
-1.24465823e+00 -4.46376026e-01 7.83430710e-02 3.17207456e-01
-7.73288727e-01 -4.78948176e-01 5.29030323e-01 -7.71221817e-01
1.26926386e+00 -1.47937164e-01 -2.30935048e-02 1.09406948e+00
3.18710297e-01 8.32444131e-01 1.06244981e+00 -4.31521982e-01
-2.65002996e-02 2.92058699e-02 7.86865413e-01 3.35475296e-01
6.56066000e-01 -9.31635499e-02 -3.04954976e-01 -4.45763111e-01
1.79277524e-01 -5.01402199e-01 -1.34091303e-01 -1.13700800e-01
-4.08055663e-01 1.01410246e+00 -9.57102142e-03 5.33230543e-01
-8.47212225e-02 1.65486544e-01 6.92823708e-01 6.03126943e-01
3.84957194e-01 4.81106400e-01 -3.40953678e-01 -6.87078536e-01
-5.72619140e-01 8.11463892e-01 1.11294806e+00 6.12616360e-01
4.42861557e-01 -5.44435978e-01 4.05980572e-02 9.26154077e-01
4.17296827e-01 3.48611653e-01 -1.76050633e-01 -1.13891482e+00
1.17057729e+00 9.74313915e-01 8.63995329e-02 -3.08331847e-01
-1.03828013e-01 -9.11818072e-02 -1.63064793e-01 2.68709749e-01
8.01452219e-01 -1.00477800e-01 -8.69110882e-01 1.49961901e+00
9.48346555e-02 -3.03833038e-01 3.75343293e-01 2.43948981e-01
5.47928393e-01 2.91107446e-01 5.61243713e-01 1.88760564e-01
1.11718249e+00 -3.84404361e-01 -6.99376166e-01 -5.29775798e-01
1.03173983e+00 -8.03013802e-01 1.19371343e+00 1.46673217e-01
-9.60340142e-01 -1.18924178e-01 -7.53132224e-01 -7.70459890e-01
-6.16114616e-01 -1.70700058e-01 8.57357860e-01 8.68018508e-01
-5.29294670e-01 4.12183642e-01 -2.70500153e-01 -4.06507641e-01
9.95225132e-01 -1.81588814e-01 -4.55101579e-01 -3.93243194e-01
-1.40703130e+00 1.24544418e+00 2.73518234e-01 -2.87847191e-01
-2.40160525e-01 -1.00209892e+00 -9.85575080e-01 1.75586164e-01
6.23495460e-01 -2.93711752e-01 1.41342556e+00 -1.81834221e-01
-5.51119030e-01 1.31540275e+00 -2.18881965e-01 -8.02341878e-01
8.18822801e-01 -5.20427048e-01 -4.48410660e-01 -1.63377851e-01
7.56093085e-01 1.91130206e-01 1.53785199e-01 -1.14661968e+00
-8.02703142e-01 -3.89975399e-01 5.28175950e-01 -3.82457584e-01
2.60778964e-01 6.57054067e-01 -1.81346312e-01 -2.98300326e-01
-3.60968322e-01 -6.32217824e-01 -9.93894115e-02 -7.70044327e-02
-2.76672423e-01 -6.45545006e-01 7.31309235e-01 -8.45247149e-01
1.41549230e+00 -1.78658938e+00 -2.64935613e-01 2.12323487e-01
4.31961231e-02 6.27448201e-01 1.15730360e-01 7.61095166e-01
-3.81292813e-02 7.03181803e-01 -3.36242795e-01 -1.12978883e-01
3.93145531e-01 6.58586860e-01 -7.08344638e-01 9.63150859e-02
9.19413269e-02 1.14748430e+00 -6.82438254e-01 -7.32678354e-01
9.54246968e-02 1.36576565e-02 -6.00459278e-01 -3.09920251e-01
-6.90450072e-01 4.91132468e-01 -8.92006576e-01 7.46958137e-01
3.40648592e-01 -1.85762703e-01 2.73329943e-01 2.82517105e-01
-3.83029468e-02 6.95326686e-01 -6.12174749e-01 1.76473486e+00
-3.92456174e-01 6.34025455e-01 -1.44433379e-01 -8.98182094e-01
4.84391987e-01 4.34153706e-01 2.18198285e-01 -9.74240363e-01
1.23269938e-01 3.91107142e-01 3.67622346e-01 -9.30180371e-01
2.86274888e-02 -2.83151925e-01 -3.25486720e-01 7.88424373e-01
-4.64392692e-01 1.00821234e-01 9.71928060e-01 4.19310123e-01
1.16998076e+00 1.57072127e-01 6.30595922e-01 -2.41465226e-01
6.36313021e-01 3.84416610e-01 5.01645923e-01 8.95591855e-01
-3.25485379e-01 3.33776735e-02 9.77893889e-01 -8.30132961e-01
-8.83939028e-01 -1.07725370e+00 -6.24801874e-01 7.30682135e-01
-2.75275379e-01 -3.66056442e-01 -7.10950792e-01 -1.04369283e+00
4.98879582e-01 9.91432786e-01 -6.82891071e-01 3.48103732e-01
-8.31706464e-01 -3.84292513e-01 7.71923900e-01 6.36242807e-01
4.41515625e-01 -1.11773229e+00 -5.51742315e-01 3.73530358e-01
-1.63416311e-01 -1.58810878e+00 9.57993418e-03 -4.95700210e-01
-6.02254450e-01 -1.82635534e+00 -1.23563558e-01 -2.20060691e-01
2.25022823e-01 -3.75829846e-01 1.17472231e+00 4.88315523e-01
-2.67884701e-01 3.61645579e-01 -4.41560417e-01 -8.20977330e-01
-1.07705390e+00 2.52859265e-01 -4.87937123e-01 -4.08755183e-01
8.65546107e-01 -4.85386133e-01 1.35366425e-01 -8.70756283e-02
-1.04860723e+00 -2.48590708e-01 2.50588268e-01 4.06774193e-01
2.23447140e-02 -5.89209497e-01 9.31724668e-01 -1.69479883e+00
1.33409226e+00 -3.65863919e-01 -6.76229060e-01 7.35471189e-01
-6.40284956e-01 6.82471469e-02 6.52857423e-01 3.40928942e-01
-1.42326510e+00 -7.77783334e-01 -3.74094099e-01 3.55348021e-01
-3.41469884e-01 7.68332481e-01 -2.41138265e-02 1.87859595e-01
9.81953442e-01 -1.77986637e-01 -8.92797038e-02 -6.25788450e-01
6.64706111e-01 8.68272841e-01 5.52587926e-01 -1.09427536e+00
9.06509936e-01 5.21856606e-01 -2.59256274e-01 -5.63129783e-01
-1.29098916e+00 -4.31874454e-01 -8.75122070e-01 -1.54317424e-01
9.86256182e-01 -4.91000265e-01 -6.14673674e-01 -2.53369436e-02
-1.46228015e+00 -2.91734099e-01 -1.85596526e-01 2.13981822e-01
-3.07766795e-01 5.83752871e-01 -6.28138959e-01 -7.86515236e-01
-2.02061996e-01 -1.13114142e+00 9.90725279e-01 3.10596097e-02
-6.16188526e-01 -1.20639324e+00 2.02859029e-01 1.36361945e+00
1.68336257e-01 4.53481406e-01 1.71289945e+00 -9.20032501e-01
-5.35853267e-01 -5.67591608e-01 -5.38939834e-01 4.46589470e-01
2.06820238e-02 -1.56751320e-01 -8.39937150e-01 2.12334692e-01
-3.47793221e-01 -5.98481894e-01 6.86795831e-01 1.13926396e-01
5.08504510e-01 -8.22642893e-02 -3.06237638e-01 1.13698486e-02
1.29244161e+00 5.13552189e-01 8.14063787e-01 4.70989615e-01
3.16084146e-01 6.55165195e-01 5.88487327e-01 -8.93472433e-02
6.30615056e-01 4.81000304e-01 -9.14038122e-02 1.65684521e-01
-4.77287993e-02 -2.17552930e-01 -5.70636541e-02 9.03388038e-02
-5.09548545e-01 -1.09417401e-01 -1.44893098e+00 5.55804491e-01
-2.10762525e+00 -1.28183067e+00 -1.11928158e-01 1.67014694e+00
6.25593603e-01 2.73738027e-01 -1.80266649e-01 -5.68568446e-02
1.99237451e-01 3.68679702e-01 -7.71015823e-01 -7.06046522e-01
-2.29152571e-02 1.92261472e-01 1.48780882e-01 4.60722029e-01
-9.05690789e-01 1.17327547e+00 6.79488659e+00 6.27284586e-01
-6.02548659e-01 2.30641499e-01 5.70205867e-01 3.23198348e-01
-5.04116058e-01 2.82877326e-01 -7.80743539e-01 3.49156797e-01
6.84603691e-01 -5.15787363e-01 2.55248547e-01 9.83341992e-01
4.50810790e-01 -1.78685278e-01 -1.36187541e+00 4.75609839e-01
-2.63776854e-02 -1.64076614e+00 2.10722327e-01 4.48062599e-01
7.53606915e-01 2.57124841e-01 -3.18614364e-01 4.84414637e-01
4.74553555e-01 -1.18564761e+00 4.43104863e-01 9.85928252e-02
5.57879150e-01 -5.43032959e-02 7.55432069e-01 3.82231355e-01
-8.12226593e-01 -2.99546689e-01 -2.42214859e-01 -1.30029038e-01
7.70463347e-01 6.73021227e-02 -5.21288157e-01 7.54479706e-01
4.68677759e-01 7.52738357e-01 -3.39441299e-01 6.70343161e-01
-4.76094991e-01 4.91042703e-01 -3.91851217e-02 2.59005189e-01
6.10383034e-01 -4.28950012e-01 3.26853991e-01 1.08347893e+00
-2.15448573e-01 2.91347802e-01 9.86868143e-02 9.53314722e-01
-4.65481490e-01 3.07616293e-01 -8.91646266e-01 -6.88272342e-02
2.89587617e-01 5.44033349e-01 -1.09261222e-01 -5.35621405e-01
-1.01268899e+00 8.03929493e-02 5.54669797e-01 3.32564682e-01
-8.41955483e-01 1.87241420e-01 8.80147755e-01 3.83916527e-01
-2.62148738e-01 -6.76963925e-02 -3.39101791e-01 -1.27250028e+00
4.57432926e-01 -1.11014521e+00 7.41977572e-01 -6.30805433e-01
-1.32358599e+00 3.37662846e-01 3.72930437e-01 -8.84060740e-01
-5.19306004e-01 -7.02118397e-01 -6.34964049e-01 6.09525919e-01
-1.97783530e+00 -1.26236951e+00 1.40974492e-01 2.39969239e-01
6.28985882e-01 -3.02418262e-01 5.40325224e-01 6.68574691e-01
-3.25130075e-01 2.91053593e-01 -1.05119064e-01 5.58620989e-01
7.33968496e-01 -1.15681660e+00 6.24731779e-01 9.11898315e-01
2.69484192e-01 1.09550893e+00 4.48430061e-01 -9.27840948e-01
-8.79499376e-01 -5.76999187e-01 1.35584593e+00 -8.00560832e-01
1.12595701e+00 -1.37094393e-01 -1.13670361e+00 1.16296101e+00
2.14141905e-01 -5.54393709e-01 8.15013826e-01 5.16261876e-01
-9.14328337e-01 6.52082413e-02 -1.05942225e+00 5.92669785e-01
1.33868206e+00 -8.11224639e-01 -1.28472459e+00 6.87936127e-01
1.62543327e-01 -3.63882959e-01 -1.17343581e+00 2.53948152e-01
7.94325471e-01 -1.04837811e+00 7.42978990e-01 -1.28540158e+00
6.62585855e-01 3.55752707e-02 1.43557027e-01 -5.65359831e-01
3.41815025e-01 -4.67559636e-01 3.94912273e-01 1.19613636e+00
1.04509175e+00 -1.04847515e+00 7.15737343e-01 1.42041934e+00
-1.61803998e-02 -6.95132911e-01 -1.15032589e+00 -8.41688931e-01
4.64630842e-01 -9.83793557e-01 4.56066251e-01 1.04246759e+00
4.14129883e-01 4.65798885e-01 -1.72450617e-01 -1.66248634e-01
6.57359481e-01 4.33904976e-01 1.03724217e+00 -1.57177484e+00
4.49203141e-02 -4.67252880e-01 -3.71223539e-01 -8.08031142e-01
5.59410214e-01 -9.34572279e-01 -6.22187555e-01 -2.00071311e+00
1.08595811e-01 -2.73696333e-01 2.80921549e-01 7.89577782e-01
-9.40435305e-02 -4.42488879e-01 2.86401451e-01 4.85249609e-01
-4.29487556e-01 2.88856514e-02 1.00095260e+00 -1.88904181e-01
1.14008360e-01 -2.99362037e-02 -9.05390918e-01 9.88577068e-01
5.96455276e-01 -4.10516202e-01 -5.61934710e-01 -7.28107035e-01
2.18874380e-01 -1.33068264e-01 -2.71877111e-03 -6.93360507e-01
2.52951443e-01 -1.74503818e-01 -3.73221546e-01 -1.62553653e-01
1.03782743e-01 -6.28953874e-01 -2.80088216e-01 3.26349914e-01
-4.48404998e-01 -1.13141850e-01 8.78195092e-02 4.42967355e-01
-5.04423440e-01 -5.23472011e-01 5.89041889e-01 -1.77761644e-01
-7.48407185e-01 4.46755141e-02 -1.32622629e-01 6.31541550e-01
1.05708659e+00 -3.80159229e-01 -9.78526533e-01 -3.99892300e-01
-5.47000408e-01 6.65651083e-01 5.01552522e-01 4.93400365e-01
1.29778266e-01 -6.40230834e-01 -9.19354141e-01 -2.31505737e-01
3.68686229e-01 -1.47676587e-01 1.96677595e-01 6.29436314e-01
-7.00722158e-01 8.52327406e-01 2.23435000e-01 -5.86104318e-02
-1.34967601e+00 2.40196250e-02 3.38916928e-01 -8.41765821e-01
-6.21257007e-01 2.24586442e-01 3.46440263e-02 -4.07269150e-01
1.60243679e-02 -4.37678158e-01 -3.24221551e-01 -5.37984706e-02
5.04487336e-01 2.11643860e-01 -2.55343169e-01 -7.20529258e-01
-2.37351254e-01 7.72212029e-01 -5.03428042e-01 8.37642699e-02
1.52462780e+00 3.48184675e-01 -2.45274276e-01 -5.53384572e-02
9.07484293e-01 -4.57899794e-02 -6.15201354e-01 -1.67154625e-01
8.61028135e-01 -6.01061165e-01 -6.08619750e-01 -1.07982802e+00
-5.55913210e-01 7.12403119e-01 -1.77338813e-02 3.51815969e-01
4.67995167e-01 4.20553923e-01 8.53765428e-01 5.91871023e-01
4.79745507e-01 -1.26275730e+00 -6.70361459e-01 4.10961688e-01
9.16898370e-01 -1.43067181e+00 1.03750110e-01 -7.79086173e-01
-6.58088624e-01 9.23747182e-01 5.12312114e-01 1.79009855e-01
4.63494003e-01 1.93810895e-01 7.43491530e-01 -6.77740455e-01
-4.64833885e-01 -2.67897576e-01 2.05384776e-01 4.61003065e-01
6.15928829e-01 -3.78095776e-01 -8.42061281e-01 2.98721999e-01
-2.56518694e-03 3.17940235e-01 5.85225105e-01 9.65131462e-01
-3.04758161e-01 -1.77579331e+00 -9.64132994e-02 6.83403432e-01
-8.96116912e-01 -1.20334096e-01 -7.09752321e-01 1.44591677e+00
1.27055153e-01 1.01081753e+00 -4.08266008e-01 4.49886799e-01
6.46068335e-01 3.33462894e-01 1.81099266e-01 -7.94439673e-01
-6.69036686e-01 -3.73276144e-01 9.09087598e-01 -5.62382102e-01
-6.21463656e-01 -6.71145856e-01 -1.27105308e+00 -8.77239257e-02
1.27531886e-01 3.26205671e-01 4.26068991e-01 1.51350355e+00
1.47998407e-01 -7.61441560e-03 -3.78693759e-01 2.05105618e-01
-5.79073489e-01 -7.39657521e-01 -4.48531657e-01 9.60063398e-01
-1.65656194e-01 -4.18377995e-01 -1.01821646e-01 6.62896782e-02] | [9.933937072753906, 9.232254981994629] |
41621940-5205-4835-a2cf-e9553de540bd | bi-stride-multi-scale-graph-neural-network | 2210.02573 | null | https://arxiv.org/abs/2210.02573v4 | https://arxiv.org/pdf/2210.02573v4.pdf | Efficient Learning of Mesh-Based Physical Simulation with BSMS-GNN | Learning the physical simulation on large-scale meshes with flat Graph Neural Networks (GNNs) and stacking Message Passings (MPs) is challenging due to the scaling complexity w.r.t. the number of nodes and over-smoothing. There has been growing interest in the community to introduce \textit{multi-scale} structures to GNNs for physical simulation. However, current state-of-the-art methods are limited by their reliance on the labor-intensive drawing of coarser meshes or building coarser levels based on spatial proximity, which can introduce wrong edges across geometry boundaries. Inspired by the bipartite graph determination, we propose a novel pooling strategy, \textit{bi-stride} to tackle the aforementioned limitations. Bi-stride pools nodes on every other frontier of the breadth-first search (BFS), without the need for the manual drawing of coarser meshes and avoiding the wrong edges by spatial proximity. Additionally, it enables a one-MP scheme per level and non-parametrized pooling and unpooling by interpolations, resembling U-Nets, which significantly reduces computational costs. Experiments show that the proposed framework, \textit{BSMS-GNN}, significantly outperforms existing methods in terms of both accuracy and computational efficiency in representative physical simulations. | ['Chenfanfu Jiang', 'Minchen Li', 'Menglei Chai', 'Yadi Cao'] | 2022-10-05 | null | null | null | null | ['physical-simulations'] | ['miscellaneous'] | [-4.75840531e-02 6.90067792e-03 2.55491287e-01 2.29973018e-01
-4.89652038e-01 -2.50975102e-01 2.47769982e-01 3.54538441e-01
-4.32703495e-01 1.07346249e+00 -3.52110744e-01 -3.29685867e-01
-5.30668020e-01 -1.32558715e+00 -1.07510328e+00 -6.84226692e-01
-5.00344753e-01 2.74954319e-01 7.19047546e-01 -2.52451181e-01
3.24023366e-01 9.12682891e-01 -1.23368239e+00 -7.36810593e-03
8.54589760e-01 1.17307937e+00 4.39522676e-02 4.90250558e-01
-2.56806463e-01 3.75407100e-01 -3.55684519e-01 -2.44264841e-01
2.58394241e-01 -1.44807577e-01 -5.74323475e-01 -3.47110897e-01
3.60858917e-01 -1.25305593e-01 -6.83495641e-01 1.03001750e+00
4.64665413e-01 2.96957135e-01 3.52961838e-01 -1.26195943e+00
-4.53351915e-01 6.26931965e-01 -9.22838271e-01 5.07414006e-02
-3.33219111e-01 1.84237540e-01 5.49450159e-01 -7.58221209e-01
6.05118155e-01 1.25520825e+00 1.27192414e+00 -7.02725071e-03
-1.40103853e+00 -8.34970355e-01 1.43684566e-01 -1.08564146e-01
-1.76247919e+00 -8.87860283e-02 8.41740072e-01 -2.85287589e-01
6.94542050e-01 2.62931675e-01 6.52421772e-01 6.31222427e-01
4.55952466e-01 3.83828506e-02 1.05626535e+00 6.14787405e-03
3.64833474e-01 -3.71766716e-01 6.30844682e-02 8.75834525e-01
4.14435625e-01 -1.08396210e-01 -4.79185194e-01 -3.96724045e-01
1.67967987e+00 -1.18140712e-01 -2.62225300e-01 -4.58422840e-01
-1.23983228e+00 5.70874631e-01 9.63780403e-01 2.53851444e-01
-4.22626466e-01 6.49588227e-01 5.45639098e-01 -8.78390670e-03
4.78065133e-01 4.66591060e-01 -2.97554880e-01 2.45210722e-01
-1.34702539e+00 3.12305659e-01 7.12813377e-01 9.05156910e-01
1.08124375e+00 1.20553039e-01 -6.79678470e-02 5.94939709e-01
-2.40081120e-02 4.29036856e-01 -2.47023016e-01 -8.87576282e-01
5.23374140e-01 6.77920997e-01 1.12062760e-01 -1.43833208e+00
-9.31563377e-01 -5.69799304e-01 -1.77493429e+00 4.03525084e-01
5.57214558e-01 -4.19787496e-01 -9.14220333e-01 1.58468926e+00
4.31585133e-01 5.56767583e-01 -3.40028554e-01 9.14048851e-01
7.17267632e-01 7.66330302e-01 4.75414768e-02 7.01523572e-02
1.17626822e+00 -7.41040409e-01 -2.42464721e-01 2.51121849e-01
4.61391658e-01 -5.38597047e-01 6.94237530e-01 1.74237669e-01
-1.24838006e+00 -5.97983003e-01 -1.12520170e+00 1.65676817e-01
-4.58174706e-01 -9.19871673e-04 6.42232060e-01 3.58234227e-01
-1.35260129e+00 1.28320456e+00 -8.93528759e-01 -1.22551555e-02
4.53589976e-01 6.91833496e-01 -1.25762209e-01 3.33968073e-01
-1.31595004e+00 3.57068807e-01 2.97018856e-01 5.33677518e-01
-5.74438632e-01 -1.02703738e+00 -7.77360737e-01 2.57023036e-01
4.62926418e-01 -5.64831674e-01 4.11853999e-01 -5.94044328e-01
-1.63701522e+00 4.71172109e-02 1.93581969e-01 -4.87360954e-01
6.00718796e-01 2.36894354e-01 -2.64421254e-01 1.05772682e-01
-2.52194941e-01 5.65779924e-01 7.33673751e-01 -1.26582909e+00
-2.84964532e-01 -1.04920536e-01 2.00070530e-01 -1.56041592e-01
-7.79772475e-02 -3.40179861e-01 -4.67829257e-01 -8.42393517e-01
2.26871043e-01 -8.33227038e-01 -6.55034661e-01 4.25450783e-03
-6.35645628e-01 -1.37812838e-01 7.23562181e-01 -5.52515864e-01
1.39974058e+00 -1.90838718e+00 1.60562158e-01 7.01400757e-01
7.17753291e-01 1.58256859e-01 -7.02850670e-02 5.40596187e-01
3.23060900e-01 1.55955344e-01 -2.94302613e-01 -1.64390713e-01
4.61653457e-04 -1.37496903e-03 5.72702885e-02 6.74790800e-01
3.10113519e-01 7.76011944e-01 -9.22402382e-01 -5.84469378e-01
2.53190249e-01 5.91343522e-01 -6.19410396e-01 -3.06405455e-01
-1.14162311e-01 5.15870869e-01 -5.94556451e-01 3.10028046e-01
1.11427295e+00 -6.13937199e-01 5.17844297e-02 -5.08238018e-01
-5.29764473e-01 1.07221536e-01 -1.69593751e+00 1.45784986e+00
-5.15906572e-01 2.72347987e-01 5.89755833e-01 -9.47057664e-01
9.32417154e-01 1.44883156e-01 4.17375207e-01 -5.19815803e-01
1.40803277e-01 3.14650714e-01 -2.43452206e-01 6.22410700e-03
6.16664231e-01 -2.69137919e-02 -2.81296730e-01 1.09597452e-01
-1.84033930e-01 -5.01224436e-02 1.38547763e-01 1.78776547e-01
1.40584838e+00 4.83401045e-02 -1.84264407e-01 -7.00648785e-01
6.08409643e-01 -2.32203007e-01 6.49250031e-01 8.98655415e-01
8.65478069e-02 4.73154455e-01 6.70074105e-01 -5.64209700e-01
-1.17019260e+00 -9.84561026e-01 -2.24806704e-02 7.77153432e-01
4.63981062e-01 -2.53545761e-01 -8.16969275e-01 -3.40303749e-01
2.59215623e-01 9.44018140e-02 -5.34525633e-01 3.70856136e-01
-9.71556127e-01 -5.45230925e-01 7.73806393e-01 4.77920443e-01
6.76510036e-01 -1.16836715e+00 -4.29278553e-01 6.24528587e-01
3.57658982e-01 -9.35246825e-01 -2.01161116e-01 1.21880218e-01
-8.51192832e-01 -8.04012835e-01 -8.54138315e-01 -6.79752052e-01
7.96425164e-01 -5.85182831e-02 1.16867959e+00 4.28514808e-01
-9.91632715e-02 -2.59292752e-01 1.31081343e-01 2.07691833e-01
-2.82303058e-02 4.45281893e-01 -8.64522606e-02 -3.37203518e-02
-3.23963791e-01 -1.03290546e+00 -8.32500994e-01 3.72592270e-01
-7.68302202e-01 4.04378414e-01 6.45872176e-01 5.83639920e-01
7.18390286e-01 1.93002939e-01 7.02967525e-01 -7.39895821e-01
4.77178514e-01 -5.86515486e-01 -9.44053948e-01 4.88339514e-02
7.79009536e-02 8.30434784e-02 1.01353705e+00 -3.80583793e-01
-7.88492322e-01 -1.90194398e-01 -1.05987228e-01 -5.09791017e-01
3.51762250e-02 2.99475431e-01 2.47734413e-01 -9.46962774e-01
4.04685795e-01 -4.54651192e-02 -3.01329702e-01 -3.27551693e-01
2.76588023e-01 1.16280057e-01 3.39517593e-01 -7.30563104e-01
7.28246033e-01 5.04791379e-01 7.12826312e-01 -1.05408275e+00
-1.93837509e-01 1.64695606e-02 -5.53532898e-01 -4.78992760e-01
7.25290477e-01 -4.73528922e-01 -1.10389328e+00 7.03213751e-01
-1.25673425e+00 -6.26420259e-01 4.92680147e-02 1.60008729e-01
-2.29863569e-01 4.41991985e-01 -1.03472674e+00 -6.86982453e-01
-3.41474354e-01 -1.27629948e+00 9.29183543e-01 3.40782017e-01
-1.63474649e-01 -8.37928474e-01 -3.74548793e-01 -3.37956011e-01
8.13921213e-01 5.76488376e-01 9.80707526e-01 -1.16266191e-01
-1.07284331e+00 7.71906897e-02 -9.06831682e-01 -2.63357698e-03
-1.57534391e-01 -7.12039769e-02 -4.25596744e-01 -3.96200001e-01
-3.29369307e-01 -1.59960657e-01 6.23906314e-01 5.12031436e-01
1.47194254e+00 -1.75024882e-01 -4.10331100e-01 7.04749465e-01
1.73902631e+00 -2.60456532e-01 6.24079168e-01 9.96759441e-03
1.10591841e+00 3.42549115e-01 -2.27190793e-01 4.67750311e-01
3.12150806e-01 4.15198296e-01 3.35552126e-01 -3.80028367e-01
-2.69267946e-01 4.44562621e-02 -2.48727053e-01 9.58960712e-01
-1.36166632e-01 -1.94626048e-01 -1.00842845e+00 5.86720943e-01
-1.85774434e+00 -4.29988533e-01 -4.19230103e-01 2.06900477e+00
5.60518324e-01 4.69090939e-01 -2.04883948e-01 -4.77896184e-02
1.00960207e+00 2.96016186e-01 -4.97095168e-01 -3.24371129e-01
2.29470022e-02 4.23082232e-01 1.04066813e+00 4.95346934e-01
-8.76685441e-01 6.07817531e-01 4.89581728e+00 1.40749598e+00
-1.21104419e+00 -4.92021348e-03 7.15311170e-01 5.24325594e-02
-1.09531999e-01 -1.79787785e-01 -8.46611559e-01 6.12970889e-01
7.16496944e-01 2.21314907e-01 6.38498604e-01 4.12511587e-01
2.93729097e-01 -1.77875698e-01 -6.99028671e-01 6.68632925e-01
-3.62122178e-01 -1.86800873e+00 5.27434759e-02 -2.45599858e-02
8.99498522e-01 1.90877065e-01 -3.24893326e-01 1.84955791e-01
4.92632687e-01 -8.35583746e-01 6.03293061e-01 5.82964838e-01
7.90788412e-01 -8.74469995e-01 5.82290530e-01 1.76172838e-01
-1.79424214e+00 3.49797547e-01 -2.56429702e-01 -8.00367370e-02
3.93261850e-01 9.61488783e-01 -1.57110542e-01 8.75650942e-01
7.30792105e-01 9.98750925e-02 -1.75004512e-01 1.11578691e+00
3.26204330e-01 3.15589279e-01 -8.66731405e-01 -3.17223281e-01
6.89385772e-01 -3.14493716e-01 4.77521122e-01 1.18735218e+00
3.61120701e-01 1.52370455e-02 4.45507735e-01 9.38004613e-01
-3.67573261e-01 5.25074489e-02 -3.38061899e-01 3.56542349e-01
5.70093274e-01 1.28155184e+00 -1.36417413e+00 -4.29794073e-01
-3.82662296e-01 4.98958737e-01 5.43836176e-01 5.09462297e-01
-8.68147194e-01 -7.09262133e-01 3.88197392e-01 4.90784287e-01
5.45519888e-01 -4.85705137e-01 -6.57252431e-01 -5.13697028e-01
3.05190198e-02 -4.07007962e-01 3.23433764e-02 -4.09671634e-01
-1.40259469e+00 6.93610966e-01 -1.41463980e-01 -9.54886973e-01
6.79288030e-01 -3.54546458e-01 -5.46035469e-01 1.03682721e+00
-1.43975484e+00 -1.07059133e+00 -3.95201832e-01 4.03166085e-01
1.79439053e-01 5.64511061e-01 3.49294484e-01 5.41875720e-01
-4.25684482e-01 5.04317343e-01 7.46012628e-02 8.65258053e-02
2.39973709e-01 -8.22233379e-01 6.88936293e-01 6.46211088e-01
-6.48681462e-01 5.57435811e-01 4.89250928e-01 -1.18144643e+00
-1.44648921e+00 -1.15615988e+00 5.99567533e-01 1.78970486e-01
1.05023730e+00 -5.10238588e-01 -1.20070195e+00 2.08078071e-01
1.05848694e-02 2.89493084e-01 -3.39196622e-02 -1.70478821e-01
1.09659739e-01 -1.81225285e-01 -1.19214380e+00 7.84148693e-01
1.08184826e+00 -2.23646909e-01 1.04936481e-01 9.87164453e-02
6.39911413e-01 -5.03349543e-01 -1.10802150e+00 5.95633149e-01
2.95545846e-01 -8.26642692e-01 1.06564164e+00 -1.35946795e-01
3.42554003e-01 -5.45551777e-01 2.86535919e-01 -1.05370915e+00
-4.73655999e-01 -8.11139405e-01 1.63854435e-01 1.11937571e+00
3.26339424e-01 -7.10164130e-01 9.79341447e-01 3.59731764e-01
-2.24922881e-01 -1.04708815e+00 -1.29600072e+00 -6.88887835e-01
8.62384215e-02 -3.66298765e-01 7.65926361e-01 8.80173564e-01
-1.97385117e-01 -9.18413401e-02 -2.98203111e-01 6.46625757e-01
9.23495948e-01 1.06302649e-03 7.74497569e-01 -1.25831831e+00
-1.89412847e-01 -5.94868243e-01 -3.30724716e-01 -1.13726997e+00
-2.32913271e-01 -6.44230008e-01 1.48257941e-01 -1.53890014e+00
-3.78382832e-01 -1.04851663e+00 -2.39846289e-01 1.94542840e-01
2.23785248e-02 5.78319371e-01 -9.35613960e-02 -5.32644242e-02
-6.89418077e-01 4.99106735e-01 1.52266204e+00 2.18223080e-01
-1.22317493e-01 -3.57010514e-01 1.48921506e-02 8.72735322e-01
7.75930345e-01 -3.76709759e-01 4.27442277e-03 -4.43469554e-01
5.19845128e-01 4.12943423e-01 6.26828730e-01 -1.41514647e+00
4.93866444e-01 7.35357851e-02 3.53200078e-01 -8.15959036e-01
2.61207879e-01 -6.63095653e-01 4.59758788e-01 4.18131053e-01
9.76682082e-02 3.34497690e-01 5.68676293e-01 5.57446957e-01
1.40295878e-01 -2.34261900e-02 8.32010508e-01 -2.17070490e-01
-3.88162166e-01 3.68196458e-01 -3.34080398e-01 1.63282119e-02
8.23008418e-01 -1.22543290e-01 -1.81390196e-01 1.20628655e-01
-5.28869629e-01 3.69379997e-01 4.88224447e-01 -6.32703826e-02
1.54978663e-01 -1.35862362e+00 -3.82485181e-01 2.13021860e-01
-5.42723656e-01 3.41037452e-01 8.56850207e-01 9.12188232e-01
-9.20988619e-01 9.63517725e-02 -2.62853384e-01 -3.94312918e-01
-6.76133394e-01 3.36444288e-01 2.79789984e-01 -7.67381728e-01
-8.86131942e-01 8.52143049e-01 1.61527038e-01 -3.14024031e-01
2.06253335e-01 -5.85291743e-01 1.56577617e-01 -1.81432888e-01
7.87796825e-02 9.07970369e-01 2.41355747e-01 -2.98090160e-01
-1.74885035e-01 6.16715193e-01 1.27882838e-01 2.91750491e-01
1.31471992e+00 -6.96850196e-02 -4.78000373e-01 9.18834060e-02
9.24950957e-01 -6.69414476e-02 -1.51300514e+00 -2.19277963e-01
-1.32140338e-01 -1.20770916e-01 1.67850807e-01 -2.83473760e-01
-1.22895145e+00 5.58406770e-01 1.71661258e-01 5.78399539e-01
7.40252256e-01 -2.88398623e-01 1.28001523e+00 3.02579671e-01
6.87628567e-01 -1.08774924e+00 -3.00780088e-01 6.37141943e-01
7.69469738e-01 -5.84402323e-01 6.49714321e-02 -7.25757241e-01
2.02831075e-01 1.13047171e+00 6.43912137e-01 -7.05440044e-01
9.87288177e-01 6.16357148e-01 -4.06794012e-01 -4.96854007e-01
-7.71549717e-02 2.79078722e-01 -6.67619482e-02 6.34979084e-02
3.32479775e-02 1.41336797e-02 -3.09148073e-01 3.70475113e-01
-3.28242481e-02 2.10355714e-01 2.32020155e-01 9.58239675e-01
-3.09673876e-01 -8.48221064e-01 -3.97871464e-01 7.68472612e-01
-1.55067265e-01 -2.11503282e-01 2.43239045e-01 8.51604044e-01
2.43379131e-01 3.49330992e-01 2.16893643e-01 -2.05410108e-01
3.94307762e-01 -2.70405561e-01 3.60655516e-01 -1.57383084e-01
-8.20663989e-01 -1.85844302e-01 4.06703018e-02 -5.88479221e-01
-6.52717054e-02 -2.88571775e-01 -1.45591605e+00 -6.70027435e-01
-2.99445868e-01 1.24336734e-01 4.70111191e-01 7.53083467e-01
5.89037538e-01 9.61015999e-01 3.34891617e-01 -1.70963812e+00
-1.89660400e-01 -8.13199341e-01 -5.42254627e-01 7.87126198e-02
2.66553611e-01 -9.48089719e-01 -3.13921213e-01 -4.99060720e-01] | [6.3353986740112305, 3.3200008869171143] |
1c9130e9-c6d3-4629-90d8-1111bcd57966 | celltrack-r-cnn-a-novel-end-to-end-deep | 2102.10377 | null | https://arxiv.org/abs/2102.10377v1 | https://arxiv.org/pdf/2102.10377v1.pdf | CellTrack R-CNN: A Novel End-To-End Deep Neural Network for Cell Segmentation and Tracking in Microscopy Images | Cell segmentation and tracking in microscopy images are of great significance to new discoveries in biology and medicine. In this study, we propose a novel approach to combine cell segmentation and cell tracking into a unified end-to-end deep learning based framework, where cell detection and segmentation are performed with a current instance segmentation pipeline and cell tracking is implemented by integrating Siamese Network with the pipeline. Besides, tracking performance is improved by incorporating spatial information into the network and fusing spatial and visual prediction. Our approach was evaluated on the DeepCell benchmark dataset. Despite being simple and efficient, our method outperforms state-of-the-art algorithms in terms of both cell segmentation and cell tracking accuracies. | ['Weidong Cai', 'Wojciech Chrzanowski', "Lauren O'Donnell", 'Fan Zhang', 'Chaoyi Zhang', 'Yang song', 'Yuqian Chen'] | 2021-02-20 | null | null | null | null | ['cell-detection'] | ['computer-vision'] | [-1.29582405e-01 -3.54392439e-01 1.43634006e-01 -3.14981528e-02
-5.68694711e-01 -6.60777271e-01 3.66374373e-01 5.19692957e-01
-1.09461796e+00 8.25884521e-01 -3.68030846e-01 -1.51161179e-01
3.73394668e-01 -5.13867736e-01 -6.05842054e-01 -9.37578738e-01
1.19930923e-01 6.39933288e-01 7.27964699e-01 2.42279306e-01
2.72256434e-01 9.52118158e-01 -8.13484013e-01 -1.00892112e-02
7.47178614e-01 8.48303795e-01 1.55024037e-01 1.13546979e+00
-9.79794562e-02 5.79314768e-01 -2.71566272e-01 -4.29633074e-02
8.01791772e-02 -2.54193425e-01 -7.95624554e-01 -7.36558735e-02
1.92232266e-01 -1.22280918e-01 4.88708727e-03 7.40970612e-01
7.53387153e-01 -9.50878859e-02 6.70994639e-01 -7.74396896e-01
-2.10988984e-01 2.70416085e-02 -6.83956981e-01 4.89816040e-01
-4.13130522e-02 4.54517394e-01 4.67066407e-01 -6.64360702e-01
1.09939981e+00 8.17679107e-01 6.78797126e-01 5.36770225e-01
-1.68975675e+00 -4.00609583e-01 6.05940707e-02 -2.04241112e-01
-1.35221827e+00 -3.45697373e-01 4.28974241e-01 -7.70644605e-01
8.90323222e-01 -1.39422074e-01 9.33871865e-01 4.39859539e-01
4.17538762e-01 9.75053310e-01 7.43362188e-01 -1.29671216e-01
4.68274593e-01 -2.13906392e-01 2.15185434e-01 7.76935875e-01
2.32156217e-01 -6.89329877e-02 -1.74678221e-01 7.93812051e-02
1.04872799e+00 1.73899963e-01 -1.32032499e-01 -4.51878965e-01
-1.68773663e+00 4.67329144e-01 4.76951718e-01 4.78842616e-01
-3.48566353e-01 4.24382299e-01 6.08972251e-01 -3.55696410e-01
4.00001317e-01 2.83023208e-01 -5.88027537e-01 1.73370633e-02
-1.25332594e+00 3.47358942e-01 6.53158545e-01 5.58778524e-01
5.14111578e-01 -2.70813555e-01 -4.88256633e-01 3.67078304e-01
4.21331167e-01 1.66158348e-01 3.55638862e-01 -1.14227808e+00
-3.66708189e-01 8.31652820e-01 9.46148261e-02 -6.23746872e-01
-8.88785839e-01 -5.85316956e-01 -6.79034114e-01 3.99543554e-01
9.98835862e-01 -2.55638897e-01 -1.14948308e+00 1.38589191e+00
6.90765619e-01 4.53699142e-01 -1.99278861e-01 9.65265632e-01
8.00132275e-01 3.82794499e-01 2.92164207e-01 -1.25679806e-01
1.27759790e+00 -1.04698193e+00 -7.78509915e-01 2.30397299e-01
1.11327231e+00 -5.29658437e-01 6.03428304e-01 2.25164369e-01
-1.07175612e+00 -3.75149965e-01 -6.59964740e-01 -4.58474219e-01
-6.20515585e-01 1.91213980e-01 5.80626309e-01 2.77002335e-01
-1.11388779e+00 5.59099197e-01 -1.39035201e+00 -6.12125397e-01
1.23903525e+00 6.22440338e-01 -3.21207702e-01 2.97460109e-01
-3.04038435e-01 5.67666888e-01 4.25893307e-01 -5.23163974e-02
-6.65103912e-01 -1.06660640e+00 -6.88155353e-01 -1.48204193e-01
-3.17425616e-02 -1.02059150e+00 1.09942031e+00 -2.40064412e-01
-1.54262650e+00 1.08298886e+00 -6.07653737e-01 -6.75859571e-01
9.20268476e-01 2.76519964e-03 1.63563624e-01 4.11401302e-01
6.84909448e-02 1.13469052e+00 -2.02104181e-01 -1.23157966e+00
-8.47007334e-01 -4.79119182e-01 -5.43978810e-01 -1.36935636e-01
1.46868050e-01 -1.19230635e-01 -7.87245810e-01 -3.41135144e-01
-3.73125672e-02 -8.03361416e-01 -5.61190546e-01 6.13631189e-01
-4.29225296e-01 1.12480417e-01 1.16414583e+00 -6.87212229e-01
8.95616531e-01 -1.88583803e+00 2.55612612e-01 8.59896839e-02
3.89886022e-01 4.17118162e-01 1.82181299e-01 -4.76572737e-02
4.37152952e-01 6.24329373e-02 -2.08144888e-01 -6.44190907e-01
-2.91711390e-01 -1.46596774e-01 2.24254653e-01 8.67112875e-01
-7.18201790e-03 1.43944621e+00 -9.35014606e-01 -9.85169709e-01
4.87214237e-01 8.06518912e-01 -5.26390851e-01 -3.60691398e-02
-3.16777021e-01 1.25288701e+00 -1.64872363e-01 1.07139277e+00
5.77833176e-01 -5.10362029e-01 1.14418194e-02 -9.07207951e-02
-2.84575135e-01 -3.99590433e-01 -7.42251158e-01 1.68230283e+00
7.92935491e-02 5.48738897e-01 2.67960101e-01 -7.23554969e-01
4.57668513e-01 2.20577046e-01 9.11781609e-01 -4.08848524e-01
3.93912345e-01 2.57853389e-01 -6.47432357e-02 -2.99915463e-01
7.05084279e-02 -1.34736791e-01 2.98695147e-01 8.29879194e-02
4.99678738e-02 -1.21046510e-03 5.98493576e-01 5.52742556e-02
9.15567279e-01 5.89311004e-01 6.08277209e-02 -3.76538187e-01
7.18933105e-01 3.87730688e-01 5.67039192e-01 3.53351116e-01
-6.92401290e-01 6.40598953e-01 6.50669158e-01 -4.77785110e-01
-9.81458664e-01 -8.60006511e-01 -1.92755654e-01 9.72957194e-01
2.05149263e-01 8.89280662e-02 -9.49901044e-01 -6.30644739e-01
6.44481257e-02 -6.90838555e-03 -7.23993897e-01 5.07517457e-01
-6.85084224e-01 -7.89052606e-01 6.33474171e-01 6.11216366e-01
3.19798052e-01 -1.00017965e+00 -8.07008028e-01 5.23641765e-01
1.04684778e-01 -1.39158738e+00 -4.92913455e-01 2.63318092e-01
-8.57386589e-01 -1.04680729e+00 -1.16615498e+00 -1.04854763e+00
7.37501502e-01 7.07095116e-02 7.93832660e-01 2.39314899e-01
-5.78381360e-01 -6.79757595e-02 9.83699858e-02 -3.74374002e-01
-1.20227255e-01 3.08563471e-01 -4.04229730e-01 -1.69854969e-01
2.02066138e-01 -3.43633890e-01 -1.01829159e+00 1.47152871e-01
-7.67914593e-01 6.82419464e-02 3.09984714e-01 8.10289741e-01
1.28028417e+00 -3.98772001e-01 2.53270715e-01 -1.11309922e+00
1.19225539e-01 -1.27424628e-01 -9.13192868e-01 3.16062197e-02
-2.23429009e-01 -3.71496648e-01 7.08691418e-01 -2.61681944e-01
-6.18136823e-01 7.70073056e-01 -3.59767318e-01 -3.05063665e-01
-3.86743963e-01 4.56528276e-01 2.15039924e-02 -5.28323650e-01
2.57149965e-01 2.50477523e-01 3.51395935e-01 -1.98018730e-01
1.48695558e-01 2.48424876e-02 8.05335701e-01 -7.69448876e-02
3.56144577e-01 1.12505674e+00 4.83777732e-01 -7.42716432e-01
-6.15421116e-01 -7.61403799e-01 -1.13219106e+00 -2.93766767e-01
1.31982553e+00 -7.77042449e-01 -1.09190989e+00 8.80651355e-01
-1.15738726e+00 -7.18261480e-01 -1.27457827e-01 3.54662567e-01
-6.19433939e-01 3.28104168e-01 -1.04056406e+00 -5.45230746e-01
-3.14188719e-01 -1.22844517e+00 1.35040057e+00 5.10025442e-01
-2.23216847e-01 -1.49881911e+00 3.33185673e-01 1.92797258e-01
3.64104658e-01 6.41852498e-01 6.31915808e-01 -7.40338743e-01
-8.33370805e-01 -3.05570662e-01 -3.45169067e-01 -2.89836347e-01
2.73951665e-02 3.28209132e-01 -9.52082813e-01 -3.31074804e-01
-7.25400865e-01 6.42073452e-02 1.13798237e+00 9.77011383e-01
1.18674827e+00 3.11465144e-01 -1.08693683e+00 1.10242701e+00
1.59728861e+00 1.77261367e-01 4.40697879e-01 3.36834162e-01
9.92953241e-01 4.58182454e-01 3.71143430e-01 1.59456730e-01
3.24272037e-01 5.39046526e-01 3.23380351e-01 -7.10491180e-01
-2.56780177e-01 1.50814369e-01 -2.07752407e-01 2.69166112e-01
1.31874472e-01 -2.35376462e-01 -1.24437046e+00 8.04797709e-01
-2.01409006e+00 -8.07349741e-01 -5.26789844e-01 1.83450246e+00
7.81959474e-01 3.32863741e-02 4.32553768e-01 -6.34602644e-03
5.93863368e-01 -3.61042708e-01 -7.81474888e-01 1.48180366e-01
-2.62909621e-01 1.12480998e-01 6.06579840e-01 6.91262126e-01
-1.39096785e+00 1.30458546e+00 6.61960602e+00 6.63383067e-01
-1.30588293e+00 1.20105240e-02 1.10535216e+00 -2.01316953e-01
3.18245769e-01 -1.83848485e-01 -9.37110484e-01 4.63644892e-01
4.72904056e-01 1.90577149e-01 -1.57256238e-02 2.42457032e-01
4.06548977e-01 -3.44604999e-01 -1.14432132e+00 7.68359005e-01
-4.59994823e-01 -1.87175047e+00 -1.49095580e-01 4.58088726e-01
6.80901825e-01 1.15397938e-01 -7.58090988e-02 -1.59479946e-01
2.87595809e-01 -1.01652408e+00 4.76085961e-01 7.61559308e-01
5.23170888e-01 -7.19180942e-01 1.02808583e+00 3.92069310e-01
-1.13570344e+00 3.14012468e-01 -5.19401878e-02 1.57301053e-01
5.01812339e-01 6.04127526e-01 -7.56011784e-01 1.73420370e-01
4.94419515e-01 8.69365513e-01 -5.22005737e-01 1.61446369e+00
3.60821515e-01 4.05405879e-01 -3.22430402e-01 8.26002806e-02
1.99180230e-01 -2.75440782e-01 3.34118605e-01 1.81984591e+00
1.53828561e-01 -1.42477512e-01 4.70228761e-01 1.01582253e+00
-1.04410626e-01 -6.52133673e-02 -3.78097028e-01 -3.20742466e-02
3.33965927e-01 1.58951032e+00 -1.62915707e+00 -4.40796524e-01
-1.84676737e-01 8.81410778e-01 2.75032312e-01 4.03209537e-01
-8.49822283e-01 -3.85704607e-01 4.87742573e-01 2.63857752e-01
5.65454125e-01 -2.62057334e-01 -6.94663048e-01 -7.24413455e-01
-6.06419683e-01 -8.68575796e-02 2.37459078e-01 -3.11038733e-01
-1.10599518e+00 1.19422801e-01 -5.05367279e-01 -8.56611133e-01
1.78336173e-01 -8.14238131e-01 -5.78824520e-01 8.07658434e-01
-1.79375291e+00 -1.32849991e+00 -2.60334492e-01 2.59034693e-01
2.51327306e-01 2.25140527e-01 5.71983457e-01 3.21240067e-01
-8.54312122e-01 2.68243849e-01 1.63251385e-01 3.97470027e-01
4.59326416e-01 -1.42490792e+00 3.49264681e-01 7.06026196e-01
-1.56415835e-01 5.24981141e-01 5.40195107e-01 -6.86441839e-01
-1.11211133e+00 -1.39400268e+00 7.51260459e-01 -4.41078514e-01
6.05831265e-01 -3.57963085e-01 -9.49157476e-01 4.03181046e-01
2.14895144e-01 6.75139129e-01 8.28752816e-01 -3.94318759e-01
3.61859620e-01 2.15978310e-01 -1.32166088e+00 6.87966645e-01
7.48545766e-01 -1.79606691e-01 2.38563865e-01 2.64480233e-01
4.58958924e-01 -7.63635874e-01 -1.01559019e+00 2.39583254e-01
5.90573013e-01 -7.36613929e-01 9.04152751e-01 -2.31660545e-01
1.86694995e-01 -7.55047023e-01 3.58473122e-01 -9.41851318e-01
-4.34603572e-01 -4.96496946e-01 -3.51913124e-01 9.89833832e-01
4.64071840e-01 -4.09118444e-01 1.24231768e+00 2.03285351e-01
-1.47659361e-01 -1.11932397e+00 -1.03475034e+00 -6.63104177e-01
4.88488317e-01 1.29314899e-01 2.91556597e-01 7.07137942e-01
-4.81832437e-02 2.94569761e-01 1.43305644e-01 -5.45816086e-02
6.42952383e-01 -1.56600606e-02 8.33098054e-01 -1.34855223e+00
9.60604846e-02 -8.98307085e-01 -4.76811707e-01 -1.11346459e+00
1.58265948e-01 -9.22123373e-01 4.46999557e-02 -1.83433497e+00
2.55223870e-01 2.93332636e-02 -3.80053997e-01 3.79641682e-01
-2.06002802e-01 6.65892661e-01 1.44619986e-01 1.87735096e-01
-9.59458172e-01 2.17914388e-01 1.56270087e+00 -1.54695690e-01
-3.89063984e-01 -3.62738371e-01 -3.15428674e-01 6.24140918e-01
6.75783396e-01 -2.14547291e-01 1.79646790e-01 -2.79690564e-01
-1.92218557e-01 -6.52322248e-02 6.26217365e-01 -1.25215721e+00
7.55913198e-01 2.52335250e-01 8.71282935e-01 -8.73293042e-01
1.78708345e-01 -6.59928858e-01 -2.85379272e-02 7.97528565e-01
-2.47634783e-01 -3.01962912e-01 4.69191790e-01 5.18491089e-01
-9.84779671e-02 3.10669541e-01 1.27533877e+00 -1.04390599e-01
-2.68473327e-01 4.27489012e-01 -6.16896212e-01 -8.11008364e-02
1.38831711e+00 -8.36568594e-01 -3.41740698e-01 2.54568815e-01
-9.13930416e-01 4.87315834e-01 9.05567825e-01 -3.72969925e-01
3.73923570e-01 -1.01747847e+00 -4.84531105e-01 -5.13147982e-03
3.15112025e-02 3.63171786e-01 1.44736931e-01 1.46156573e+00
-1.18782902e+00 5.82352579e-01 -1.80295035e-01 -1.07749844e+00
-1.19617569e+00 5.02005756e-01 7.18507111e-01 -4.95218098e-01
-4.09067452e-01 9.29299772e-01 2.69221634e-01 -4.77954775e-01
3.00182134e-01 -6.02093220e-01 -3.75611514e-01 -2.10526079e-01
2.60804683e-01 4.81326014e-01 -2.14468881e-01 -7.87934303e-01
-5.65704644e-01 8.47353339e-01 -2.44530886e-01 1.92177221e-01
1.17569661e+00 -1.37858182e-01 -2.75179446e-01 5.17193973e-01
1.10925794e+00 -1.11093670e-01 -1.69556308e+00 2.47794688e-02
1.66745231e-01 -7.55145354e-03 1.23651549e-01 -9.40674961e-01
-1.39740050e+00 8.35121930e-01 7.27157593e-01 -3.75573449e-02
8.15741718e-01 -2.09188629e-02 1.10145199e+00 1.95843160e-01
1.42738760e-01 -1.00554848e+00 -1.91077098e-01 6.95735395e-01
7.18222260e-02 -1.34722102e+00 4.84919250e-02 -5.40674269e-01
-4.04439867e-01 9.11692321e-01 6.47788227e-01 -1.57885090e-01
7.24019051e-01 5.33829093e-01 2.64170945e-01 -2.79423118e-01
-6.55405223e-01 -2.81065196e-01 5.83913326e-02 7.12171674e-01
8.17610681e-01 -3.22956234e-01 -2.72065043e-01 3.49650145e-01
3.77044380e-01 4.96772498e-01 9.57627967e-02 9.96993482e-01
-5.49852371e-01 -8.50087166e-01 -6.23485968e-02 2.26678580e-01
-9.32779789e-01 1.68062568e-01 -3.31599802e-01 6.91389382e-01
2.72262573e-01 4.42263246e-01 1.43616676e-01 1.65590331e-01
2.63102595e-02 -2.57283241e-01 2.74394542e-01 -4.68823969e-01
-7.54453242e-01 5.42616367e-01 -5.56241155e-01 -5.67326546e-01
-4.51647848e-01 -7.69777417e-01 -1.99783993e+00 -1.05842002e-01
-1.53433338e-01 -1.49593160e-01 3.94858152e-01 9.53216195e-01
5.50226390e-01 7.87949741e-01 -4.98539172e-02 -1.09887302e+00
3.37422162e-01 -5.69677770e-01 -5.23527622e-01 1.08485788e-01
4.31628764e-01 -1.81340545e-01 1.45498574e-01 4.95498776e-01] | [14.604710578918457, -3.2330551147460938] |
a869c912-c114-4c7b-9529-993aed1622be | spp-net-deep-absolute-pose-regression-with | 1712.03452 | null | http://arxiv.org/abs/1712.03452v1 | http://arxiv.org/pdf/1712.03452v1.pdf | SPP-Net: Deep Absolute Pose Regression with Synthetic Views | Image based localization is one of the important problems in computer vision
due to its wide applicability in robotics, augmented reality, and autonomous
systems. There is a rich set of methods described in the literature how to
geometrically register a 2D image w.r.t.\ a 3D model. Recently, methods based
on deep (and convolutional) feedforward networks (CNNs) became popular for pose
regression. However, these CNN-based methods are still less accurate than
geometry based methods despite being fast and memory efficient. In this work we
design a deep neural network architecture based on sparse feature descriptors
to estimate the absolute pose of an image. Our choice of using sparse feature
descriptors has two major advantages: first, our network is significantly
smaller than the CNNs proposed in the literature for this task---thereby making
our approach more efficient and scalable. Second---and more importantly---,
usage of sparse features allows to augment the training data with synthetic
viewpoints, which leads to substantial improvements in the generalization
performance to unseen poses. Thus, our proposed method aims to combine the best
of the two worlds---feature-based localization and CNN-based pose
regression--to achieve state-of-the-art performance in the absolute pose
estimation. A detailed analysis of the proposed architecture and a rigorous
evaluation on the existing datasets are provided to support our method. | ['Christopher Zach', 'Pulak Purkait', 'Cheng Zhao'] | 2017-12-09 | null | null | null | null | ['image-based-localization'] | ['computer-vision'] | [-1.28981739e-01 -1.28336787e-01 -1.95709616e-01 -4.16944206e-01
-6.00909352e-01 -1.49202764e-01 4.09635305e-01 -1.00808382e-01
-4.73068506e-01 4.88039583e-01 -2.73587015e-02 1.99890420e-01
-1.99273184e-01 -8.06433678e-01 -8.43515694e-01 -4.93374139e-01
-2.66268048e-02 4.74499524e-01 9.01225433e-02 -3.27421308e-01
3.93830448e-01 9.40265834e-01 -1.67426825e+00 -4.05115664e-01
5.23209155e-01 1.34893334e+00 2.75349110e-01 9.08463225e-02
1.98509693e-01 5.18333852e-01 -2.69581944e-01 -1.29602715e-01
4.76026773e-01 -1.93476111e-01 -3.54515076e-01 -5.94519116e-02
7.28282213e-01 -3.40224087e-01 -6.75859213e-01 9.36462164e-01
6.38989389e-01 1.75639689e-01 5.75561225e-01 -1.05338550e+00
-7.20091581e-01 7.34963417e-02 -6.12182021e-01 -1.40291899e-01
4.62788582e-01 -2.93596715e-01 8.19176316e-01 -1.05283964e+00
6.78731203e-01 1.05050945e+00 9.70526636e-01 4.50751871e-01
-8.90268207e-01 -6.87114358e-01 -2.37937048e-02 1.52826369e-01
-1.78096116e+00 -3.57756734e-01 1.09034932e+00 -2.37807125e-01
1.01607418e+00 -9.89368930e-02 8.28357935e-01 9.21946228e-01
3.76981795e-01 5.76289952e-01 8.12649906e-01 -4.45930898e-01
2.27039084e-01 -1.63420904e-02 -3.31901044e-01 1.04276526e+00
4.36390489e-01 1.18316993e-01 -6.47455513e-01 1.68878570e-01
1.16584647e+00 2.87475675e-01 -2.07803801e-01 -1.01003647e+00
-1.27361059e+00 8.52964342e-01 1.08863854e+00 2.80122578e-01
-3.75916868e-01 4.95298088e-01 2.60468751e-01 -7.13070035e-02
5.29024303e-01 6.43755794e-01 -3.63853455e-01 4.28526215e-02
-9.35077965e-01 4.18281794e-01 6.95497751e-01 1.00423467e+00
8.73029411e-01 1.84294388e-01 4.62386698e-01 7.00034559e-01
4.20086861e-01 6.01486802e-01 5.55518091e-01 -8.75423729e-01
5.02515495e-01 6.63744986e-01 7.11416006e-02 -1.57169557e+00
-7.72874475e-01 -5.61021805e-01 -1.00690889e+00 2.07345888e-01
2.43523687e-01 1.62802473e-01 -8.34205389e-01 1.75322378e+00
1.64623976e-01 1.96642011e-01 -1.10808320e-01 1.15275133e+00
7.85858810e-01 2.73048341e-01 -3.21046382e-01 1.41576484e-01
1.03829002e+00 -6.94100678e-01 -5.82814097e-01 -2.54811078e-01
4.73377287e-01 -6.42526269e-01 6.95553064e-01 3.05300504e-01
-7.86393940e-01 -6.30877495e-01 -1.36380994e+00 -1.40760764e-01
-3.85418624e-01 4.13899988e-01 7.45626390e-01 4.46037263e-01
-1.14848101e+00 5.06708980e-01 -9.40191507e-01 -7.11272418e-01
3.46228898e-01 7.57668555e-01 -7.49527335e-01 -7.00528100e-02
-8.41183603e-01 1.21065080e+00 3.09128970e-01 2.22412542e-01
-5.30726612e-01 -2.68754393e-01 -1.24933946e+00 -1.11549221e-01
1.10440321e-01 -6.35785043e-01 1.06256437e+00 -7.10283458e-01
-1.61251438e+00 8.05501640e-01 -1.11600704e-01 -5.50649166e-01
3.74458134e-01 -5.12992203e-01 -4.49177064e-02 7.14015812e-02
1.23306982e-01 9.09634173e-01 7.34956443e-01 -1.21969056e+00
-3.71764153e-01 -6.26798868e-01 6.10417202e-02 3.64261448e-01
-3.60163480e-01 -4.34116423e-01 -4.78549778e-01 -5.04359126e-01
8.03711474e-01 -1.24975789e+00 -4.33409631e-01 3.22686285e-01
-2.81473875e-01 -7.60086924e-02 8.33351254e-01 -2.96608865e-01
6.96209311e-01 -2.07746029e+00 3.19505632e-01 1.04040906e-01
2.87499279e-01 2.87787348e-01 -7.39576817e-02 4.44694251e-01
1.09016737e-02 -2.88061768e-01 1.41804442e-01 -5.05748212e-01
-3.05587709e-01 2.92939007e-01 -1.50558189e-01 1.02656794e+00
2.24581450e-01 9.92811322e-01 -8.00350666e-01 -2.76828319e-01
7.35279083e-01 7.66213655e-01 -4.74280268e-01 9.51806009e-02
2.01528594e-02 6.63359165e-01 -5.08482754e-01 5.29697120e-01
7.07634628e-01 -1.06337398e-01 -2.37291548e-02 -4.34535205e-01
-2.08861336e-01 8.50686356e-02 -1.27549326e+00 2.30015707e+00
-6.57581449e-01 6.48736954e-01 -2.23494023e-01 -1.11584473e+00
1.31681204e+00 9.36906263e-02 6.37432158e-01 -7.10944355e-01
4.45365340e-01 5.46635151e-01 -2.30898380e-01 -3.18774968e-01
6.31420672e-01 1.47743121e-01 -1.49742872e-01 -1.20057136e-01
3.01021516e-01 -4.37697411e-01 -8.95524770e-02 -3.30945373e-01
8.45316112e-01 4.62886214e-01 5.34486890e-01 -9.50884074e-02
6.40607417e-01 -8.58193357e-03 4.44339305e-01 4.27255988e-01
-3.87315415e-02 8.55459929e-01 -3.73611227e-02 -7.81917810e-01
-9.62722123e-01 -7.98740745e-01 -4.11635637e-02 4.54063058e-01
6.38558745e-01 -1.48404166e-01 -4.77939874e-01 -3.50371957e-01
1.77121118e-01 1.77939981e-01 -5.55371463e-01 -9.88858566e-02
-8.51005316e-01 -1.59665018e-01 3.63359541e-01 7.64008999e-01
7.51763046e-01 -8.48589361e-01 -7.60194421e-01 6.51536062e-02
-7.25270659e-02 -1.14247811e+00 -3.27944607e-02 1.97258681e-01
-1.04549313e+00 -1.13932300e+00 -8.68717551e-01 -1.00552917e+00
7.91955531e-01 5.61383903e-01 7.31157124e-01 -2.64380425e-01
-3.10013443e-01 3.33811015e-01 -3.19955349e-01 -3.47423911e-01
9.04223844e-02 3.45092744e-01 3.52439702e-01 -1.89311698e-01
2.46728763e-01 -5.95107794e-01 -7.16089606e-01 3.59713763e-01
-6.72361493e-01 -1.93670928e-01 8.60789537e-01 7.49402761e-01
7.44300485e-01 -3.19050133e-01 2.69668072e-01 -5.02740860e-01
2.29920149e-01 -2.52001405e-01 -7.56716371e-01 -9.42571238e-02
-4.34525043e-01 2.26652190e-01 7.18417287e-01 -2.89393932e-01
-5.98393321e-01 5.91421247e-01 -2.25142047e-01 -6.58783376e-01
-2.60095835e-01 5.85362196e-01 3.34858187e-02 -7.71469653e-01
6.98520422e-01 1.85805306e-01 1.59838453e-01 -5.52024484e-01
4.34693664e-01 5.19156098e-01 5.35234451e-01 -2.69931674e-01
9.20741022e-01 6.04665577e-01 5.81789911e-01 -8.04413199e-01
-8.04707527e-01 -6.47010744e-01 -8.72655332e-01 -1.19244367e-01
7.96168149e-01 -1.22456670e+00 -8.33790183e-01 4.52846408e-01
-1.27099848e+00 2.12010369e-01 -7.44849741e-02 8.02322149e-01
-8.32303464e-01 2.51194715e-01 -3.45811486e-01 -6.89417362e-01
-1.66800395e-01 -1.35428870e+00 1.20463467e+00 3.27249497e-01
-6.95082471e-02 -8.49269569e-01 2.11192325e-01 8.72058049e-02
5.13106525e-01 4.85544384e-01 3.69828582e-01 -5.44665217e-01
-5.76602459e-01 -7.04303384e-01 -3.21978003e-01 1.88334331e-01
9.25154313e-02 -3.78354400e-01 -8.87630522e-01 -4.71997976e-01
-1.71953458e-02 -4.66890365e-01 6.03912234e-01 3.95209014e-01
1.05306780e+00 1.12214364e-01 -4.12745029e-01 8.31280708e-01
1.60703909e+00 -1.48644716e-01 4.59207207e-01 4.94549304e-01
7.24894345e-01 2.57932842e-01 7.15851367e-01 4.10406590e-01
4.65850890e-01 9.93514597e-01 8.91399503e-01 -1.03792973e-01
-1.35482997e-01 -3.61091793e-01 1.05405627e-02 9.16712999e-01
-3.01055849e-01 9.64677706e-03 -8.00127745e-01 4.10498410e-01
-1.93886638e+00 -6.13882124e-01 -1.54034663e-02 2.33771825e+00
2.55684733e-01 -1.48855895e-01 -1.63789600e-01 9.65415835e-02
5.27850866e-01 2.76697040e-01 -3.75130028e-01 -1.09592900e-02
1.42337143e-01 3.72866809e-01 6.84099197e-01 1.31789356e-01
-1.39157951e+00 8.55580330e-01 5.88909626e+00 5.14130771e-01
-1.53999257e+00 -1.24427460e-01 1.45678237e-01 1.17774658e-01
3.36672902e-01 -3.31469476e-01 -8.11561704e-01 6.09913431e-02
6.55481517e-01 2.41385773e-01 1.52115554e-01 1.30301857e+00
7.22702313e-03 -2.41155595e-01 -1.19473052e+00 1.30764389e+00
5.83029270e-01 -1.45833492e+00 -1.85592449e-03 1.15770340e-01
6.79270566e-01 2.04322159e-01 1.82964459e-01 7.58918300e-02
-1.57327622e-01 -1.01448965e+00 9.75013614e-01 2.79076338e-01
9.12539840e-01 -8.50292683e-01 9.87117171e-01 4.50720668e-01
-1.31484497e+00 -1.33093685e-01 -5.81476212e-01 -1.89963758e-01
9.34653636e-03 4.39631969e-01 -6.36369765e-01 7.20644534e-01
6.08435869e-01 8.97183537e-01 -4.86165255e-01 1.29708934e+00
-2.32777521e-01 -1.32032588e-01 -3.79851460e-01 -2.19697893e-01
2.26050422e-01 -1.09151162e-01 2.71448553e-01 8.04233909e-01
5.51140964e-01 -2.32367426e-01 3.72646809e-01 6.44504488e-01
-1.01845108e-01 2.99709532e-02 -1.00067294e+00 3.13746810e-01
4.38093007e-01 1.26179194e+00 -7.14283049e-01 1.96712211e-01
-5.06878197e-01 8.76860917e-01 5.72184503e-01 4.56315763e-02
-6.93536162e-01 -6.00257516e-01 5.75330436e-01 1.52167290e-01
3.26404959e-01 -9.42267835e-01 -2.46176630e-01 -1.09854507e+00
5.50646298e-02 -4.84362960e-01 -1.89928964e-01 -5.97624302e-01
-1.05011213e+00 6.84208274e-01 -6.92236498e-02 -1.67406762e+00
-3.98211151e-01 -9.37102377e-01 -1.11806192e-01 7.87744939e-01
-1.63837492e+00 -1.48038626e+00 -7.02702820e-01 6.11559808e-01
5.77186227e-01 -1.41389623e-01 1.02411711e+00 3.07746142e-01
-2.38048166e-01 4.88672435e-01 6.51988834e-02 3.01235467e-01
5.72382510e-01 -9.03625846e-01 3.48659277e-01 6.10602975e-01
3.87555003e-01 7.01641917e-01 4.46125656e-01 -4.56626624e-01
-1.80527878e+00 -9.71066892e-01 7.48568118e-01 -3.80934507e-01
5.05252361e-01 -3.75780582e-01 -3.75252038e-01 5.27533710e-01
-2.06096992e-01 3.70177716e-01 3.37235838e-01 -4.61580157e-02
-3.72171670e-01 -3.30706000e-01 -1.03129029e+00 4.53734696e-01
9.64481533e-01 -4.71124738e-01 -5.18908501e-01 2.26133212e-01
4.31462318e-01 -8.00937414e-01 -7.76553750e-01 5.13824761e-01
6.39457941e-01 -1.05918825e+00 1.14551795e+00 -5.70964180e-02
2.83380032e-01 -3.75522852e-01 -4.46809590e-01 -1.25487828e+00
-2.91400969e-01 -3.22076321e-01 -1.49241686e-01 6.70183480e-01
3.75316925e-02 -6.58210337e-01 1.04536819e+00 1.97445422e-01
-2.32866541e-01 -8.12537014e-01 -1.21656513e+00 -8.35693359e-01
-1.49080664e-01 -2.74623752e-01 3.02576393e-01 5.82684875e-01
-2.96892464e-01 1.58062175e-01 -4.31749910e-01 3.80387723e-01
5.10304272e-01 5.19471988e-02 9.86682594e-01 -1.41290343e+00
1.37415096e-01 -1.65734753e-01 -1.10698509e+00 -1.26431751e+00
2.75031596e-01 -6.05474055e-01 2.67159730e-01 -1.44544041e+00
-8.86034817e-02 -5.23105979e-01 -1.53447226e-01 3.63888621e-01
3.55263233e-01 6.69747531e-01 1.86688796e-01 3.89237642e-01
-5.86807728e-01 7.04846501e-01 1.17873752e+00 6.46643564e-02
-1.60109326e-01 9.28308815e-03 -3.05732965e-01 7.55740166e-01
6.37117684e-01 -3.14146280e-01 -3.17317754e-01 -4.84218985e-01
1.96581692e-01 -6.38666190e-03 4.53227997e-01 -1.50126898e+00
3.34606677e-01 1.99072793e-01 6.84297323e-01 -5.70809305e-01
7.86176801e-01 -1.06064427e+00 1.18007483e-02 6.19972765e-01
-1.12527929e-01 1.73828274e-01 -1.42705655e-02 7.40374029e-01
-4.54513043e-01 -1.53036326e-01 7.59556592e-01 -3.64094764e-01
-9.45281982e-01 4.19712275e-01 -9.15529951e-03 -2.47244552e-01
9.30816293e-01 -3.45961303e-01 -3.95550020e-02 -4.97737616e-01
-3.60923201e-01 -2.59366184e-01 6.30686462e-01 5.44960082e-01
9.34059799e-01 -1.45671284e+00 -2.78067172e-01 3.70462775e-01
3.11164558e-01 2.91383862e-01 -3.92300375e-02 9.83293414e-01
-1.08987856e+00 7.58646131e-01 -3.88075829e-01 -8.05208921e-01
-9.03775156e-01 4.25370872e-01 2.70084560e-01 -7.75737083e-03
-5.39740801e-01 7.94348538e-01 4.71193530e-03 -7.21710086e-01
4.25875634e-01 -3.52855295e-01 -9.25725773e-02 -2.88101047e-01
1.16953999e-01 2.65727103e-01 2.54109770e-01 -1.00477469e+00
-6.25442684e-01 1.11271775e+00 1.95774823e-01 1.34444416e-01
1.51966548e+00 -6.09064428e-03 -1.05737723e-01 2.97704875e-01
1.41964102e+00 -2.07635820e-01 -1.08991647e+00 -3.00477624e-01
-4.37612794e-02 -6.39551878e-01 1.83520690e-01 -3.47640038e-01
-1.06943059e+00 9.82097685e-01 7.45833576e-01 -1.36271939e-01
8.74191165e-01 -1.16827093e-01 6.25903070e-01 6.71224356e-01
8.87226582e-01 -8.17277133e-01 2.59973377e-01 6.71301126e-01
9.89394486e-01 -1.41903377e+00 1.99451014e-01 -4.26448643e-01
-2.86354721e-01 1.40247571e+00 6.74490988e-01 -6.58274353e-01
7.09298909e-01 -5.57512492e-02 1.87684178e-01 -2.59741008e-01
8.32635611e-02 -2.31653601e-01 4.69410270e-01 7.06545293e-01
4.61049080e-01 -2.21832857e-01 -1.21054277e-01 1.56249896e-01
-1.74467295e-01 7.51285702e-02 2.01069266e-01 9.96716499e-01
-3.21608007e-01 -9.24798250e-01 -4.16594535e-01 1.56789362e-01
-1.58734564e-02 2.99053669e-01 -3.24506193e-01 1.07103384e+00
1.74845442e-01 7.18994737e-01 1.35261491e-02 -6.01267636e-01
4.80695248e-01 -2.91209608e-01 7.20046580e-01 -6.66611910e-01
-3.45081657e-01 -5.47105037e-02 -3.03094417e-01 -9.06894386e-01
-6.88417912e-01 -5.06632864e-01 -1.06058490e+00 -1.87820986e-01
-6.09362006e-01 -2.82710165e-01 1.11615682e+00 8.85432065e-01
3.73180807e-01 2.10472330e-01 6.49098575e-01 -1.65852952e+00
-6.19036674e-01 -7.90777147e-01 -6.37586832e-01 2.38377169e-01
2.38010406e-01 -1.10722911e+00 -8.44032019e-02 -4.25781429e-01] | [7.794301986694336, -2.0841474533081055] |
2b1c7e7f-f002-4f0c-ab7f-4253f4ba2049 | hospital-length-of-stay-prediction-based-on | 2303.09817 | null | https://arxiv.org/abs/2303.09817v1 | https://arxiv.org/pdf/2303.09817v1.pdf | Hospital Length of Stay Prediction Based on Multi-modal Data towards Trustworthy Human-AI Collaboration in Radiomics | To what extent can the patient's length of stay in a hospital be predicted using only an X-ray image? We answer this question by comparing the performance of machine learning survival models on a novel multi-modal dataset created from 1235 images with textual radiology reports annotated by humans. Although black-box models predict better on average than interpretable ones, like Cox proportional hazards, they are not inherently understandable. To overcome this trust issue, we introduce time-dependent model explanations into the human-AI decision making process. Explaining models built on both: human-annotated and algorithm-extracted radiomics features provides valuable insights for physicians working in a hospital. We believe the presented approach to be general and widely applicable to other time-to-event medical use cases. For reproducibility, we open-source code and the TLOS dataset at https://github.com/mi2datalab/xlungs-trustworthy-los-prediction. | ['Przemysław Biecek', 'Patryk Szatkowski', 'Przemysław Bombiński', 'Bartlomiej Sobieski', 'Hubert Baniecki'] | 2023-03-17 | null | null | null | null | ['length-of-stay-prediction'] | ['medical'] | [ 1.12073310e-02 7.78336763e-01 -4.87603813e-01 -7.59893417e-01
-1.12951994e+00 -4.26882237e-01 2.82488525e-01 8.10443878e-01
-3.41892719e-01 7.81474292e-01 5.05475581e-01 -9.76449490e-01
-4.21068192e-01 -5.07313013e-01 -4.88123626e-01 -5.36629558e-01
-1.91229954e-01 8.51496160e-01 -8.29684064e-02 1.78669408e-01
1.23026729e-01 5.21073461e-01 -8.35300207e-01 6.60978019e-01
3.70142698e-01 7.19972372e-01 -8.23554397e-02 7.81161726e-01
4.02186692e-01 1.45958364e+00 -7.74621442e-02 -3.53365004e-01
4.43478450e-02 -4.60544616e-01 -8.75894248e-01 -2.29094088e-01
-1.56325117e-01 -3.78166556e-01 -4.38935906e-01 3.39787632e-01
4.09553021e-01 -4.84674037e-01 9.20713425e-01 -1.12924135e+00
-5.14173210e-01 8.61193061e-01 -1.45651475e-01 4.16876853e-01
3.78277600e-01 2.61367410e-01 5.97913444e-01 -5.57855666e-01
6.95955634e-01 7.83461869e-01 7.61698246e-01 5.09335220e-01
-1.12146163e+00 -4.52270269e-01 -3.37499768e-01 1.44072473e-01
-1.10158968e+00 -4.27616723e-02 2.43161559e-01 -7.32318580e-01
5.80594599e-01 7.84928381e-01 5.93660176e-01 1.17026663e+00
9.01594043e-01 4.44235355e-01 1.19152844e+00 -4.42132920e-01
1.66328624e-02 2.52703011e-01 2.26281315e-01 1.01656818e+00
3.96158189e-01 4.31464404e-01 -3.34707648e-01 -6.73124254e-01
7.88137555e-01 5.25789917e-01 -4.31460887e-01 -2.25373521e-01
-1.67325997e+00 1.00286007e+00 4.80550677e-01 2.99431890e-01
-5.39482415e-01 2.26486385e-01 4.66507822e-01 1.73196882e-01
3.68088394e-01 5.97909212e-01 -7.45747030e-01 4.92470199e-03
-9.67558384e-01 -6.84185922e-02 7.35182762e-01 7.97148108e-01
8.55198577e-02 -3.97218823e-01 -1.44662663e-01 2.19631925e-01
3.60528171e-01 1.38392121e-01 5.73049664e-01 -7.77553618e-01
-3.08790743e-01 4.94570255e-01 -6.47517145e-02 -5.82341075e-01
-1.02739418e+00 -4.65653628e-01 -7.53428280e-01 1.42471984e-01
5.44136822e-01 -4.67253625e-02 -9.54620838e-01 1.18963540e+00
-3.32271904e-02 -2.41991207e-01 -3.90550159e-02 1.06972075e+00
9.79014993e-01 1.92425072e-01 5.60895860e-01 -2.46777222e-01
1.70216763e+00 -8.81268203e-01 -7.10283101e-01 -3.78400087e-02
1.22436988e+00 -6.48174584e-01 8.41120660e-01 3.90862793e-01
-1.10125124e+00 -1.07441723e-01 -7.60000050e-01 2.30875574e-02
-1.64785147e-01 2.37498984e-01 7.43657172e-01 2.26606712e-01
-7.07980812e-01 7.78379321e-01 -1.13934970e+00 -5.69301724e-01
5.78536749e-01 2.48398528e-01 -6.31423771e-01 8.14406667e-03
-9.82642710e-01 1.32766569e+00 2.85591662e-01 -1.45075575e-01
-1.01443493e+00 -8.77960742e-01 -6.00489557e-01 -9.70788598e-02
4.16894466e-01 -1.18630123e+00 1.44593787e+00 -7.25911140e-01
-6.74029827e-01 1.11390734e+00 -1.52852133e-01 -4.86142635e-01
9.29709256e-01 -1.32989094e-01 -3.62071335e-01 2.65572429e-01
6.26671538e-02 2.60055929e-01 3.45108509e-01 -1.49516881e+00
-3.93298030e-01 -4.73336756e-01 -1.59342110e-01 -3.12041909e-01
6.74405620e-02 3.32597084e-02 2.05184370e-02 -9.00441170e-01
1.24866776e-01 -1.11957467e+00 -7.12423325e-01 4.11262780e-01
-5.16142130e-01 1.59781724e-01 1.22694701e-01 -8.38660479e-01
1.32740974e+00 -1.94911635e+00 -2.67419010e-01 9.30718631e-02
6.81959987e-01 -4.03124928e-01 4.41271842e-01 4.63967860e-01
-5.96762717e-01 3.48784864e-01 -3.60252410e-01 -1.05188869e-01
-3.18822235e-01 1.57810390e-01 1.13891726e-02 6.65201902e-01
-3.50422263e-02 9.92701292e-01 -8.71104538e-01 -9.04871702e-01
3.23917329e-01 4.13637996e-01 -3.14681172e-01 2.13179350e-01
1.81881651e-01 9.17909384e-01 -5.99869907e-01 6.08342230e-01
4.93773818e-02 -7.31669009e-01 1.66619644e-01 -6.04389831e-02
8.08645710e-02 -4.81798500e-02 -3.54693681e-01 1.65510941e+00
-3.82492959e-01 4.92489845e-01 -5.44245064e-01 -5.94242811e-01
6.45827711e-01 6.98150635e-01 6.76499903e-01 -2.98599035e-01
4.23283696e-01 2.75128812e-01 8.12998638e-02 -6.73277378e-01
-8.41997787e-02 -7.69070327e-01 -8.17042142e-02 5.23567259e-01
-2.43116766e-01 -1.47022605e-01 -3.47385317e-01 3.13739210e-01
1.36827075e+00 -1.48633435e-01 9.56152320e-01 -2.75451124e-01
2.20537052e-01 5.38767517e-01 3.01735520e-01 9.43377972e-01
-2.43631721e-01 9.60923493e-01 5.89315653e-01 -9.92292345e-01
-1.06301999e+00 -1.02045870e+00 -7.31455982e-01 6.26000166e-01
-1.82604983e-01 -4.19505745e-01 -2.98216790e-01 -7.95024872e-01
-7.93473348e-02 1.24496067e+00 -1.18111372e+00 -1.42120838e-01
-2.41047591e-01 -7.19880164e-01 2.52531558e-01 7.25984573e-01
-4.96092975e-01 -8.96513820e-01 -1.17419457e+00 2.68063009e-01
-1.18797027e-01 -7.74789989e-01 -6.32468313e-02 5.03957570e-01
-1.00922298e+00 -1.27084851e+00 -7.33432710e-01 -1.03671618e-01
7.33356595e-01 -2.74251729e-01 1.41277015e+00 6.08212113e-01
-6.42689228e-01 5.48524857e-01 -5.46670794e-01 -8.94421101e-01
-9.03332531e-01 -1.25995651e-01 -2.46293291e-01 -6.45969987e-01
3.68360519e-01 -1.33948103e-01 -9.59311783e-01 3.40432823e-01
-9.16741312e-01 4.73915696e-01 7.21718013e-01 9.10561979e-01
6.68951035e-01 -3.84143889e-01 2.12368697e-01 -1.30276287e+00
2.15702221e-01 -7.46162295e-01 -8.24748874e-02 3.36534202e-01
-9.77780044e-01 5.39155416e-02 4.01879936e-01 1.88233424e-03
-6.93480372e-01 2.08931223e-01 1.08886383e-01 -9.96281132e-02
-4.02651042e-01 6.71481609e-01 7.05955327e-01 3.39681685e-01
8.42571557e-01 -2.24438861e-01 1.01259790e-01 -7.48285353e-02
-6.16563670e-03 4.56502736e-01 3.91602069e-01 -2.01952338e-01
5.88507056e-01 7.35949576e-01 2.44011804e-01 -2.90746540e-02
-1.04041243e+00 -3.96328300e-01 -7.03625262e-01 -3.03148210e-01
1.05550158e+00 -8.53359342e-01 -6.70418441e-01 -3.83477002e-01
-1.08856463e+00 -3.52343678e-01 -3.82696390e-01 8.93550694e-01
-8.64867091e-01 -1.09215438e-01 -5.48893094e-01 -6.11807644e-01
-2.85236716e-01 -1.15387976e+00 9.87558544e-01 -1.48601308e-01
-9.13505495e-01 -1.18673241e+00 5.46965487e-02 4.66598392e-01
2.54355669e-01 7.07790196e-01 1.15801919e+00 -1.05394459e+00
-4.81351733e-01 -6.04158938e-01 -1.15670398e-01 -3.19936574e-01
-3.15697305e-02 8.72621611e-02 -8.99610937e-01 -5.15500642e-02
6.66164011e-02 -1.30992964e-01 7.82588363e-01 6.66455209e-01
1.48615921e+00 -2.50793308e-01 -6.29619479e-01 2.98087120e-01
1.40112865e+00 1.36099398e-01 5.89505732e-01 5.68411648e-01
4.25669432e-01 6.41563952e-01 8.06501031e-01 7.28650689e-01
5.47242999e-01 3.02670628e-01 5.10981798e-01 -5.50384700e-01
2.38299340e-01 -1.37795471e-02 -2.88017303e-01 3.53596687e-01
-4.34731036e-01 -2.47251615e-01 -1.68754470e+00 4.63954449e-01
-1.91299355e+00 -6.88282669e-01 -6.10733032e-01 1.92681670e+00
5.73431134e-01 2.13425308e-01 -1.85401604e-01 3.66609953e-02
2.81342804e-01 -2.04768762e-01 -2.86803246e-01 -2.93089837e-01
1.96154863e-01 -1.43016115e-01 7.47843921e-01 3.59472275e-01
-8.13093305e-01 3.05254161e-01 6.38493824e+00 3.09267461e-01
-1.00013173e+00 4.60568666e-01 1.10161150e+00 -1.28506362e-01
-4.23466146e-01 2.31494457e-01 -4.31726202e-02 1.62065655e-01
1.27218413e+00 -2.25678667e-01 -1.31570682e-01 8.95357728e-01
6.47909522e-01 -2.96727806e-01 -1.47099388e+00 7.00520992e-01
-8.87657478e-02 -1.56762505e+00 -1.91070750e-01 -1.55917732e-02
2.05140486e-01 -1.50290117e-01 -1.12286650e-01 -8.88346732e-02
1.20948918e-01 -1.38906217e+00 8.41983795e-01 1.03379107e+00
9.93054509e-01 -2.19750240e-01 1.09231973e+00 1.76562458e-01
-5.37565649e-01 -7.63277188e-02 2.72118784e-02 2.27182597e-01
2.80901819e-01 4.64994997e-01 -1.36103189e+00 7.66709626e-01
7.48643637e-01 5.01339793e-01 -7.75968254e-01 8.81534994e-01
-1.17960513e-01 8.05636346e-01 6.68539032e-02 4.03898805e-01
-1.04985209e-02 3.76119047e-01 2.27767378e-01 1.16693676e+00
4.46674556e-01 6.24717414e-01 -2.72907823e-01 6.42364681e-01
4.69559580e-01 1.63394049e-01 -6.02894783e-01 8.70159045e-02
-5.65605052e-03 1.30397022e+00 -1.00369000e+00 -4.38037187e-01
-5.40400922e-01 7.75976121e-01 -1.76887944e-01 5.19356169e-02
-1.05880833e+00 3.31840754e-01 -1.21437937e-01 8.35483968e-01
-4.01866764e-01 5.57910129e-02 -8.13973367e-01 -9.83865798e-01
-4.50231999e-01 -6.52863681e-01 9.68617141e-01 -1.37713659e+00
-1.20956552e+00 9.89107728e-01 4.73960638e-02 -1.45663941e+00
-3.21174353e-01 -6.39429748e-01 -3.14621180e-01 5.73210895e-01
-1.40040052e+00 -1.30376804e+00 -4.95814860e-01 2.82100588e-01
3.41447949e-01 -8.39332417e-02 1.26962221e+00 4.49452065e-02
-1.49217427e-01 2.73918420e-01 -1.57166079e-01 2.76874173e-02
1.05577767e+00 -1.33773017e+00 -9.29955915e-02 7.34940916e-02
-1.15240656e-01 6.45211101e-01 9.69709814e-01 -5.50050914e-01
-8.63065898e-01 -9.00356948e-01 8.37715864e-01 -1.12097633e+00
7.69856870e-01 2.14260831e-01 -9.88023520e-01 1.06069458e+00
7.41889104e-02 9.00926366e-02 1.27367198e+00 -1.09782413e-01
1.21409930e-02 4.42754537e-01 -9.98409510e-01 2.91738123e-01
6.60842717e-01 -2.11710021e-01 -7.16780782e-01 6.20525241e-01
4.61267233e-01 -3.98400635e-01 -1.44395816e+00 6.10776126e-01
6.29328430e-01 -1.02429914e+00 7.80465782e-01 -1.02450001e+00
9.70000565e-01 -4.40180153e-02 -1.03265811e-02 -9.76880670e-01
-2.03016743e-01 -1.05707511e-01 3.28920722e-01 3.50899488e-01
9.10256565e-01 -6.44742548e-01 6.03280604e-01 9.91358042e-01
-1.13875479e-01 -1.13821709e+00 -8.34333360e-01 -2.88747996e-01
1.72375992e-01 -4.94270205e-01 3.33900809e-01 1.24677908e+00
1.97867244e-01 -1.33243307e-01 -1.02130048e-01 3.50201994e-01
4.97683138e-01 3.90216745e-02 4.20253962e-01 -1.35421062e+00
-2.47777730e-01 8.10052976e-02 -4.46931928e-01 5.57969883e-02
-9.13699418e-02 -9.43312824e-01 -2.25492820e-01 -1.78549850e+00
8.28274429e-01 -5.36879539e-01 -4.45569932e-01 7.70440698e-01
-2.21077651e-01 6.38783276e-02 9.83383209e-02 6.99882448e-01
-3.38381171e-01 1.66551456e-01 1.17246616e+00 2.18007445e-01
2.28604108e-01 -2.30553970e-02 -5.59098840e-01 1.01202297e+00
6.20820999e-01 -1.19367754e+00 -1.14789464e-01 -2.50812799e-01
1.17225781e-01 8.81299317e-01 8.65299940e-01 -8.82155955e-01
1.57064661e-01 -1.45617366e-01 5.30681372e-01 -4.04443085e-01
7.86776692e-02 -1.14480388e+00 6.21435642e-01 9.45853353e-01
-5.99794924e-01 1.88250884e-01 2.38006592e-01 5.19578755e-01
-1.44185856e-01 -1.15760706e-01 6.71440780e-01 -4.10346359e-01
-3.06191862e-01 2.54918247e-01 -5.65701306e-01 -3.30948532e-01
1.31505752e+00 -1.77800611e-01 -4.11145300e-01 -6.26330435e-01
-1.27953017e+00 5.75540252e-02 5.79602659e-01 1.39376298e-01
4.93696511e-01 -8.38104546e-01 -1.09651244e+00 -3.51801097e-01
4.96009290e-01 -1.44026265e-01 4.75832462e-01 1.34129250e+00
-9.94735003e-01 6.51833236e-01 -3.84372994e-02 -5.54454863e-01
-1.23207629e+00 7.69310117e-01 4.41754341e-01 -2.95785904e-01
-6.92901134e-01 4.83183026e-01 3.89571160e-01 -2.79476494e-01
-2.45892152e-01 -2.30351299e-01 1.13355756e-01 -2.15738341e-01
3.54200393e-01 1.26961619e-03 1.28355213e-02 -5.22950053e-01
-4.88531113e-01 2.15134278e-01 -1.79744929e-01 -1.28623545e-01
1.59669662e+00 -7.65653327e-02 3.46286483e-02 8.31199229e-01
9.15202439e-01 -1.95794716e-01 -8.05028737e-01 1.11833230e-01
1.05696745e-01 -3.54820907e-01 7.40381181e-02 -1.15870690e+00
-7.74995923e-01 7.78476834e-01 8.57983708e-01 1.87768221e-01
9.38050032e-01 4.99856085e-01 3.27290714e-01 8.00895318e-02
2.31468558e-01 -4.24991310e-01 -2.88584400e-02 -3.03884000e-01
1.16675949e+00 -1.66292691e+00 4.61366773e-01 -4.79603022e-01
-1.16884875e+00 1.35928285e+00 2.94244766e-01 1.94435686e-01
8.26297164e-01 2.14179352e-01 5.35487711e-01 -5.82452416e-01
-1.00130785e+00 3.00357342e-01 2.22389221e-01 1.87667489e-01
8.72801960e-01 3.81097347e-01 -4.00758415e-01 9.85786855e-01
-1.68498114e-01 4.52481002e-01 8.53763998e-01 9.81352568e-01
4.24348526e-02 -8.99704814e-01 -4.87322688e-01 7.33468413e-01
-6.93703532e-01 -2.73148060e-01 -1.46171182e-01 1.00134468e+00
-1.50701612e-01 6.89224482e-01 -2.68669367e-01 -7.88604915e-02
3.76473576e-01 4.52717952e-02 8.59148875e-02 -6.24063492e-01
-5.68838179e-01 -9.14399326e-03 1.29452303e-01 -6.21280491e-01
-3.87304842e-01 -6.56108856e-01 -1.62488151e+00 -1.96505070e-01
-1.24344788e-01 1.50484860e-01 6.91691995e-01 8.05942655e-01
1.14646152e-01 8.95881474e-01 2.48560712e-01 -2.77872205e-01
-3.65888238e-01 -7.46853352e-01 -4.71873671e-01 3.25999320e-01
3.38962466e-01 -5.27508259e-01 -5.14058530e-01 3.81330907e-01] | [8.460079193115234, 5.721365451812744] |
2d6e6b66-01ae-4b48-b3f4-69c35749eb9c | conversational-answer-generation-and | 2103.06500 | null | https://arxiv.org/abs/2103.06500v1 | https://arxiv.org/pdf/2103.06500v1.pdf | Conversational Answer Generation and Factuality for Reading Comprehension Question-Answering | Question answering (QA) is an important use case on voice assistants. A popular approach to QA is extractive reading comprehension (RC) which finds an answer span in a text passage. However, extractive answers are often unnatural in a conversational context which results in suboptimal user experience. In this work, we investigate conversational answer generation for QA. We propose AnswerBART, an end-to-end generative RC model which combines answer generation from multiple passages with passage ranking and answerability. Moreover, a hurdle in applying generative RC are hallucinations where the answer is factually inconsistent with the passage text. We leverage recent work from summarization to evaluate factuality. Experiments show that AnswerBART significantly improves over previous best published results on MS MARCO 2.1 NLGEN by 2.5 ROUGE-L and NarrativeQA by 9.4 ROUGE-L. | ['Vikas Bhardwaj', 'Hakan Inan', 'Debojeet Chatterjee', 'Barlas Oguz', 'Stan Peshterliev'] | 2021-03-11 | null | null | null | null | ['passage-ranking'] | ['natural-language-processing'] | [ 2.42058218e-01 7.14102745e-01 4.32853669e-01 -2.46897489e-01
-1.63547981e+00 -8.17931294e-01 7.78045416e-01 2.49978542e-01
-1.46526456e-01 1.16094959e+00 1.06058574e+00 -3.94255966e-01
-1.20200925e-01 -5.90012968e-01 -3.60737473e-01 4.06958647e-02
4.32624429e-01 9.51434314e-01 6.94270134e-02 -8.62702906e-01
5.14782965e-01 -2.33029023e-01 -1.29812539e+00 9.73545074e-01
1.46438551e+00 4.63339627e-01 3.35963517e-01 1.46479332e+00
-5.12944877e-01 1.47854602e+00 -1.23217463e+00 -6.53639972e-01
-3.97860527e-01 -1.24564540e+00 -1.78005970e+00 -2.54442453e-01
7.10171342e-01 -4.55364585e-01 -2.01030634e-02 7.80819118e-01
8.23344171e-01 2.46844485e-01 6.00828588e-01 -8.53453398e-01
-6.62774444e-01 7.63995290e-01 2.29948670e-01 3.51960033e-01
1.42772055e+00 1.63881227e-01 1.14519966e+00 -7.75576711e-01
8.02586675e-01 1.63914704e+00 4.38255191e-01 8.51363897e-01
-9.74451721e-01 2.97357678e-01 -4.50798154e-01 4.88487989e-01
-7.16184556e-01 -8.14357162e-01 4.03664678e-01 2.95004547e-02
1.31818974e+00 1.00103045e+00 3.77062708e-01 1.09193039e+00
2.70374626e-01 1.08165467e+00 9.51844692e-01 -5.92763126e-01
2.28241131e-01 -1.18053034e-01 2.95045078e-01 4.48025197e-01
-3.67431223e-01 -5.12998879e-01 -6.94321632e-01 -3.54542643e-01
-1.35254785e-02 -6.95317328e-01 -6.04421258e-01 4.37256753e-01
-1.07898080e+00 8.84159803e-01 -6.29541799e-02 -9.48150977e-02
-5.40494442e-01 -1.81829423e-01 4.29872036e-01 7.67945766e-01
2.09778190e-01 1.09537792e+00 -1.76518500e-01 -8.42646658e-01
-7.90387988e-01 1.00078487e+00 1.65467322e+00 9.49174583e-01
1.65449128e-01 -2.63007790e-01 -9.43332076e-01 1.10511160e+00
6.15763739e-02 7.36593843e-01 6.00580156e-01 -1.51842785e+00
7.63945520e-01 4.90959883e-01 4.52956140e-01 -7.73300648e-01
-2.17217758e-01 -3.38738412e-01 -3.64578485e-01 -3.40738863e-01
5.17194033e-01 -3.76153350e-01 -4.83005911e-01 1.34465337e+00
5.49409911e-02 -6.56990826e-01 6.42732441e-01 7.64066517e-01
1.60018194e+00 9.03914034e-01 -1.58948287e-01 -5.09583890e-01
1.45007932e+00 -1.21900344e+00 -1.24365509e+00 -4.07649845e-01
5.34287274e-01 -1.09345257e+00 1.46957672e+00 3.45704883e-01
-1.43794119e+00 -3.81023675e-01 -7.28280246e-01 -4.64144886e-01
2.06118673e-01 5.07022589e-02 1.21589534e-01 3.38129640e-01
-1.03786123e+00 1.90278068e-01 -1.13104485e-01 -4.15973693e-01
-5.59658781e-02 -5.33179715e-02 3.78833562e-02 -2.02267855e-01
-1.30755544e+00 1.08991730e+00 1.40265495e-01 -1.20268069e-01
-6.46002412e-01 -5.33997357e-01 -6.97430789e-01 -2.21929662e-02
5.83645582e-01 -1.17377698e+00 2.34141111e+00 -4.97497082e-01
-1.78222525e+00 5.06154418e-01 -6.94063723e-01 -6.66570187e-01
4.99366462e-01 -7.03788400e-01 -4.98706758e-01 5.43527067e-01
3.12827975e-01 6.03427529e-01 7.05042005e-01 -1.18619764e+00
-4.23780322e-01 -1.49469689e-01 4.04707104e-01 8.14074457e-01
3.81993324e-01 -3.33183520e-02 1.54607669e-02 -1.42225176e-01
-2.33338727e-03 -4.90399361e-01 -1.43929189e-02 -9.43278611e-01
-6.45762563e-01 -6.02801263e-01 5.20269156e-01 -1.33182418e+00
1.59862173e+00 -1.45487154e+00 1.39814749e-01 -3.52540106e-01
3.56442988e-01 3.00828069e-01 -2.84378648e-01 1.00526333e+00
4.49873447e-01 2.49214321e-02 -2.15982303e-01 -8.62778258e-03
1.18140109e-01 9.83867049e-02 -6.83717668e-01 -4.37137961e-01
2.38418877e-01 1.25644708e+00 -1.19494390e+00 -6.13863289e-01
-1.84425563e-01 -8.51372629e-02 -5.44344306e-01 6.32101715e-01
-6.80124164e-01 3.81819993e-01 -3.32104206e-01 4.39901620e-01
2.09249169e-01 -6.86170161e-02 -1.39835745e-01 3.39354843e-01
3.35920192e-02 1.07816696e+00 -5.33159673e-01 1.76069951e+00
-5.25938630e-01 6.45572543e-01 -2.01702356e-01 -1.36462346e-01
7.06070006e-01 5.57014108e-01 -3.32339764e-01 -7.72516847e-01
-1.30396917e-01 2.97761559e-01 1.20156586e-01 -1.11474609e+00
1.19591296e+00 -1.59885108e-01 -2.19610557e-01 5.66495299e-01
1.25924215e-01 -9.01898980e-01 5.36233902e-01 7.36976147e-01
1.31200898e+00 -2.21242327e-02 5.37264228e-01 4.71141451e-04
5.87816238e-01 4.25678909e-01 -7.89896324e-02 1.23808360e+00
-8.90383720e-02 8.40054512e-01 7.62817740e-01 5.63489869e-02
-6.14176273e-01 -1.08698404e+00 4.08067137e-01 8.57872307e-01
-2.59358138e-01 -8.86825621e-01 -1.21747422e+00 -7.91101635e-01
-5.46710432e-01 1.57071888e+00 -2.04058632e-01 -1.30440056e-01
-6.65320814e-01 -1.12723209e-01 7.66649961e-01 2.17034191e-01
5.80570340e-01 -1.45387900e+00 -7.01304197e-01 5.08455873e-01
-1.27582240e+00 -9.23487484e-01 -4.45874602e-01 -3.23582232e-01
-8.79112780e-01 -7.97805309e-01 -7.61163056e-01 -4.25180495e-01
1.17652684e-01 1.99069276e-01 1.81354642e+00 5.16402395e-03
2.48309970e-01 6.84976399e-01 -6.92341626e-01 -5.27015746e-01
-9.74939942e-01 3.38604450e-01 -3.52582276e-01 -5.63425064e-01
3.96435469e-01 -2.51724333e-01 -5.20275295e-01 -1.03833228e-02
-7.40742564e-01 1.55326754e-01 3.57879609e-01 8.77991557e-01
7.64664039e-02 -5.45995951e-01 1.18129146e+00 -8.94322038e-01
1.79832244e+00 -4.69592869e-01 2.42298484e-01 5.56681514e-01
-2.26063013e-01 2.02605486e-01 4.75405961e-01 6.50472846e-03
-1.45525587e+00 -5.35256505e-01 -5.92155695e-01 4.50531930e-01
-2.09225014e-01 5.37351131e-01 -6.04824051e-02 6.45190775e-01
1.19952714e+00 1.74355760e-01 1.88145548e-01 -1.74453214e-01
6.99409544e-01 9.38544333e-01 7.27936625e-01 -5.02304435e-01
3.94924939e-01 -3.80837061e-02 -6.28216743e-01 -1.04904401e+00
-1.14155555e+00 -5.44711471e-01 -8.08050856e-02 -5.17774582e-01
6.78694427e-01 -4.77271199e-01 -6.17865741e-01 -1.07590005e-01
-1.60900950e+00 -1.62405401e-01 -4.48363006e-01 7.77567104e-02
-9.16808605e-01 6.90589309e-01 -5.21750450e-01 -1.02094471e+00
-9.05343890e-01 -7.20605195e-01 8.97974432e-01 3.95185947e-01
-1.37686634e+00 -7.39633977e-01 2.88291693e-01 1.03703535e+00
4.23176348e-01 4.26506856e-03 8.89280140e-01 -6.97826266e-01
-1.82460219e-01 -1.22248076e-01 1.44069508e-01 1.60903007e-01
-4.68160994e-02 -3.76449496e-01 -9.29531813e-01 1.88750118e-01
3.55871558e-01 -7.34837651e-01 6.77299142e-01 4.73255143e-02
5.43685317e-01 -9.51040030e-01 2.61189699e-01 -4.62949812e-01
7.28413939e-01 1.82940722e-01 8.94883990e-01 6.00607283e-02
3.29116821e-01 9.81794000e-01 7.21583188e-01 2.63734579e-01
5.41493893e-01 4.14876103e-01 1.25996262e-01 4.68854785e-01
-4.00615424e-01 -5.87743759e-01 5.24492145e-01 1.14872313e+00
-3.14243957e-02 -5.81643581e-01 -8.41987371e-01 5.91760039e-01
-1.88833737e+00 -1.28135741e+00 -5.71561694e-01 1.79237473e+00
1.17764187e+00 1.55896610e-02 1.19486496e-01 8.41695741e-02
1.79107174e-01 4.56752740e-02 -1.44866779e-01 -8.22794437e-01
-1.43035233e-01 4.19343144e-01 -4.08907354e-01 9.62170482e-01
-3.84837210e-01 9.06210065e-01 6.14739895e+00 7.51863420e-01
-3.17861378e-01 1.81875408e-01 2.33475998e-01 -1.25666201e-01
-6.62411273e-01 3.38673592e-03 -5.51918685e-01 9.79193896e-02
1.06207132e+00 -3.56721014e-01 4.15725291e-01 4.74828809e-01
2.83860862e-01 -5.21274745e-01 -1.13342726e+00 7.59959340e-01
5.22956312e-01 -1.14066827e+00 3.72999549e-01 -4.36141789e-01
5.02688348e-01 -4.21170980e-01 -1.27991483e-01 7.32304335e-01
1.90706849e-01 -1.17801833e+00 5.76155782e-01 6.85184479e-01
4.55505133e-01 -8.19528401e-01 9.62924182e-01 6.31131411e-01
-2.93752223e-01 1.05857059e-01 -2.46347204e-01 -3.69284451e-01
6.73472106e-01 2.57919461e-01 -1.32474983e+00 5.50347388e-01
3.62064719e-01 -3.49347405e-02 -6.93636417e-01 9.63672280e-01
-7.40754604e-01 8.06379974e-01 -2.37828828e-02 -4.96356010e-01
2.91292846e-01 1.01759948e-03 1.11625540e+00 1.12879503e+00
3.20012003e-01 4.71232474e-01 -1.95423111e-01 6.71233654e-01
-7.73983002e-02 2.84963667e-01 -6.36510551e-01 -1.11788966e-01
3.46378446e-01 9.41279650e-01 -1.12538017e-01 -4.57435578e-01
4.70805950e-02 1.43871224e+00 3.33220631e-01 3.80219400e-01
-2.68644035e-01 -6.03811800e-01 -4.19853628e-03 -1.35089323e-01
-1.97798565e-01 -7.09376708e-02 -2.22060814e-01 -9.89462256e-01
2.34397039e-01 -1.56195235e+00 3.97004813e-01 -1.17474198e+00
-1.10451567e+00 8.27487528e-01 -5.47096282e-02 -9.66953576e-01
-1.03656185e+00 -1.61700889e-01 -7.02213109e-01 8.29329371e-01
-1.12854683e+00 -7.07128763e-01 -2.52872646e-01 1.49292409e-01
1.24504018e+00 1.51503664e-02 1.18109071e+00 -1.60990387e-01
3.24096419e-02 3.47580373e-01 -3.08981717e-01 -3.87708724e-01
6.83017612e-01 -1.61343133e+00 4.32413638e-01 6.38734877e-01
3.04390758e-01 8.49546850e-01 1.18532264e+00 -6.90706611e-01
-1.27811372e+00 -6.49963737e-01 1.59461963e+00 -1.03957975e+00
3.40161830e-01 1.18183270e-01 -9.95879292e-01 3.20406973e-01
1.06306505e+00 -1.10873663e+00 7.63107538e-01 1.84184656e-01
-9.62066650e-03 2.91041344e-01 -9.81291771e-01 1.03270221e+00
8.27776134e-01 -7.36172259e-01 -1.51834011e+00 5.40723443e-01
1.00662982e+00 -6.42627180e-01 -5.15807450e-01 1.25862792e-01
2.60885388e-01 -9.21346664e-01 6.56774879e-01 -6.94230914e-01
8.58862340e-01 -2.90139139e-01 -6.58531487e-02 -1.71834052e+00
1.78571120e-01 -1.18676007e+00 -6.07225657e-01 1.14165044e+00
7.19499111e-01 -1.60387918e-01 1.90789759e-01 6.55236959e-01
-3.68695855e-01 -5.02385557e-01 -8.28997016e-01 -5.13213336e-01
7.30449408e-02 -3.49892557e-01 5.71864367e-01 4.51941133e-01
5.18316448e-01 1.33317840e+00 -2.20644698e-01 -3.08516979e-01
2.18007371e-01 -1.20499833e-02 8.70835423e-01 -9.27538812e-01
-4.10362214e-01 -3.51471782e-01 2.72620678e-01 -1.49514210e+00
3.19127738e-02 -7.24551201e-01 3.36706579e-01 -2.13187814e+00
8.70663226e-02 4.97365445e-01 7.12071061e-01 -4.25093733e-02
-5.04725397e-01 -1.54705390e-01 2.03814954e-01 1.03817999e-01
-9.40583467e-01 7.64671028e-01 1.49322271e+00 -1.35014102e-01
-4.80893552e-01 2.20375657e-01 -9.55711722e-01 5.72459757e-01
1.09958279e+00 -1.98191434e-01 -6.34799719e-01 -4.69655067e-01
5.81735969e-01 6.28693700e-01 2.49830127e-01 -7.51528144e-01
3.01395059e-01 1.13415383e-01 -1.21503964e-01 -9.35085416e-01
3.18871379e-01 -9.20300111e-02 -3.10165137e-01 1.11899003e-01
-7.58039415e-01 3.11731458e-01 7.14498237e-02 3.65402222e-01
-4.03943151e-01 -7.90545821e-01 7.21782222e-02 -4.79185045e-01
-4.28162128e-01 -5.77044427e-01 -9.36409950e-01 6.81611538e-01
2.73735881e-01 -1.21179380e-01 -6.38301730e-01 -1.35156667e+00
-5.24205625e-01 4.49369311e-01 -1.92532450e-01 4.64487195e-01
1.05274808e+00 -1.10428357e+00 -1.23275685e+00 -4.02609169e-01
3.19213986e-01 -7.99209252e-02 2.82444060e-01 6.46733701e-01
-6.37484550e-01 6.56062663e-01 1.20251574e-01 -3.51311952e-01
-1.42472732e+00 -1.56790167e-01 2.70704806e-01 -2.38433033e-01
-5.99854350e-01 7.85354018e-01 -4.63684678e-01 -6.81156099e-01
4.58073840e-02 -2.01705590e-01 -6.72666490e-01 2.62695849e-01
9.78266358e-01 7.00944245e-01 2.75994956e-01 -2.08810017e-01
2.30464622e-01 -1.31660700e-01 -3.54490668e-01 -9.08027411e-01
6.88869476e-01 -4.09733981e-01 -3.65928113e-01 6.14575446e-01
7.80535936e-01 2.23347589e-01 -5.80637038e-01 -5.82632534e-02
3.37826610e-01 -2.27262914e-01 -3.75916421e-01 -1.37845147e+00
1.90516710e-01 6.70311868e-01 3.90715376e-02 5.59183002e-01
8.12453389e-01 1.14789382e-01 1.17558157e+00 1.10466623e+00
1.32371873e-01 -1.23560274e+00 3.44380707e-01 1.14688492e+00
1.66113079e+00 -1.11440825e+00 -3.23485255e-01 -1.73475668e-01
-1.01428950e+00 9.94587004e-01 6.12716019e-01 4.77538884e-01
-3.44311297e-01 -3.21273088e-01 4.30338979e-01 -4.96214896e-01
-9.87793803e-01 -2.51652747e-01 4.33119833e-01 5.28982580e-01
7.21975148e-01 1.78083152e-01 -6.84448838e-01 7.12454438e-01
-1.06163585e+00 -2.51079798e-01 8.13776195e-01 8.92909586e-01
-5.63420177e-01 -8.87016177e-01 -3.54920119e-01 5.03208816e-01
-5.20388305e-01 -3.93671602e-01 -1.09623408e+00 3.20085764e-01
-6.95567548e-01 1.86796224e+00 -3.76402706e-01 -5.43225184e-02
5.61689854e-01 5.23733318e-01 6.10179126e-01 -8.96957755e-01
-9.57769632e-01 -4.24693793e-01 1.09086907e+00 -3.25820386e-01
-1.96277067e-01 -5.82283378e-01 -1.29066813e+00 -1.88070968e-01
-2.80315369e-01 6.65684938e-01 2.92238355e-01 1.00074232e+00
4.46512073e-01 5.02309322e-01 2.01967120e-01 1.23437243e-02
-8.16586971e-01 -1.49764252e+00 1.79011285e-01 2.80832112e-01
4.60493028e-01 1.86202884e-01 -2.45203048e-01 1.04485378e-01] | [11.798077583312988, 8.09407901763916] |
e6e0c3ce-db19-4992-a472-bac3e250f4d4 | a-novel-disparity-transformation-algorithm | 1808.02837 | null | http://arxiv.org/abs/1808.02837v1 | http://arxiv.org/pdf/1808.02837v1.pdf | A Novel Disparity Transformation Algorithm for Road Segmentation | The disparity information provided by stereo cameras has enabled advanced
driver assistance systems to estimate road area more accurately and
effectively. In this paper, a novel disparity transformation algorithm is
proposed to extract road areas from dense disparity maps by making the
disparity value of the road pixels become similar. The transformation is
achieved using two parameters: roll angle and fitted disparity value with
respect to each row. To achieve a better processing efficiency, golden section
search and dynamic programming are utilised to estimate the roll angle and the
fitted disparity value, respectively. By performing a rotation around the
estimated roll angle, the disparity distribution of each row becomes very
compact. This further improves the accuracy of the road model estimation, as
demonstrated by the various experimental results in this paper. Finally, the
Otsu's thresholding method is applied to the transformed disparity map and the
roads can be accurately segmented at pixel level. | ['Mohammud Junaid Bocus', 'Naim Dahnoun', 'Rui Fan'] | 2018-08-08 | null | null | null | null | ['road-segementation'] | ['computer-vision'] | [ 3.12054157e-01 -4.73928563e-02 -2.89187372e-01 -5.08041441e-01
-5.03622890e-02 -6.25888780e-02 4.67632085e-01 -8.96714106e-02
-6.26928031e-01 8.91353607e-01 -3.92877050e-02 -4.95164573e-01
2.21260846e-01 -1.06064463e+00 -3.92976552e-01 -6.47508681e-01
6.77089453e-01 1.04100453e-02 5.26748240e-01 4.11003120e-02
8.65983844e-01 6.72845781e-01 -1.99662781e+00 -3.09229493e-01
1.24027658e+00 9.58290160e-01 5.25439978e-01 4.68464077e-01
-2.81445593e-01 3.89580488e-01 -1.50067285e-01 -3.68577033e-01
5.81791341e-01 -1.85654476e-01 -4.30898368e-01 2.75855839e-01
6.07441843e-01 -7.96250343e-01 -2.24297985e-01 1.30523968e+00
1.82168856e-01 4.25582416e-02 6.02095246e-01 -8.32491040e-01
2.76446402e-01 -1.41728997e-01 -1.36300898e+00 1.56745329e-01
1.31150201e-01 -8.44390690e-03 2.89047718e-01 -8.60718131e-01
4.84391809e-01 1.04645050e+00 4.14167285e-01 -1.18515812e-01
-8.46968710e-01 -7.45933712e-01 -2.75628209e-01 6.00513518e-01
-1.49109423e+00 -4.00866300e-01 7.85520852e-01 -2.77686417e-01
6.00619733e-01 6.60154372e-02 7.74643302e-01 -2.23467886e-01
4.26307023e-01 3.83356482e-01 1.41021883e+00 -7.32947707e-01
-2.08032019e-02 1.44056052e-01 7.67771015e-03 7.15741336e-01
5.68016112e-01 3.13018680e-01 -1.18182570e-01 4.65598971e-01
7.92621374e-01 -5.56309484e-02 -3.17869097e-01 -4.63337481e-01
-9.26543832e-01 6.01359069e-01 4.96279687e-01 -7.71751627e-02
-5.36325693e-01 -7.10112751e-02 2.58849084e-01 -1.85591519e-01
2.40856975e-01 -1.72025174e-01 2.14731395e-01 -1.22681685e-01
-8.73636544e-01 -2.93576214e-02 2.43970916e-01 8.73220563e-01
1.47599971e+00 1.89216547e-02 3.30116898e-01 8.23157549e-01
3.92455786e-01 7.98712909e-01 2.65678853e-01 -1.08113110e+00
7.72125244e-01 6.52702570e-01 2.70542763e-02 -1.13020694e+00
-2.22798198e-01 -1.74132772e-02 -7.36923575e-01 7.22914517e-01
5.93118072e-01 -2.18177617e-01 -1.11407459e+00 9.18382704e-01
5.74639261e-01 2.83747256e-01 1.01153284e-01 8.07948828e-01
4.32124436e-01 7.52312064e-01 -2.26027980e-01 -1.74552649e-01
1.30232275e+00 -7.52685905e-01 -8.07435334e-01 -5.87924242e-01
3.99011731e-01 -1.06630087e+00 6.12649202e-01 2.10607171e-01
-8.90053511e-01 -6.98662817e-01 -1.15054286e+00 -2.18521401e-01
-1.25359043e-01 3.99372846e-01 2.66481757e-01 5.38557351e-01
-8.81322265e-01 -1.36693150e-01 -5.70031881e-01 -1.33072942e-01
2.34908879e-01 1.59101576e-01 -2.40199372e-01 -3.66973758e-01
-9.51479912e-01 1.27096748e+00 4.42185432e-01 4.31122988e-01
1.53242961e-01 -3.22963417e-01 -1.13545275e+00 -3.31794709e-01
7.22418353e-02 -5.79955876e-01 9.33921099e-01 -5.33334136e-01
-1.49116015e+00 1.14088011e+00 -7.32986748e-01 -5.50958753e-01
5.09656966e-01 -5.94604015e-02 -1.99236587e-01 4.26995426e-01
3.61936241e-01 8.42635572e-01 7.69189000e-01 -1.15592337e+00
-1.42933440e+00 -5.84534943e-01 -3.77959937e-01 7.23846972e-01
3.45074356e-01 -3.53322327e-01 -6.67670667e-01 1.19355403e-01
5.37410378e-01 -7.25220978e-01 -3.66450518e-01 7.81594366e-02
-3.10308397e-01 2.66574234e-01 9.70836878e-01 -7.77221680e-01
1.28663301e+00 -2.10497832e+00 -4.46859747e-01 5.89799047e-01
2.17166677e-01 3.97702694e-01 4.37640876e-01 -1.33978069e-01
2.83119082e-01 -5.11513412e-01 -4.09198761e-01 6.27438575e-02
-6.23057067e-01 1.44263685e-01 -1.73327684e-01 5.07239044e-01
-7.66097456e-02 4.64659929e-01 -5.28186858e-01 -7.32956707e-01
8.62101972e-01 3.38753968e-01 -1.11682817e-01 -6.43371791e-02
6.17417395e-01 2.49638334e-01 -4.06160563e-01 4.19961452e-01
1.42331159e+00 7.14385450e-01 -1.32496521e-01 -3.70930791e-01
-7.70539582e-01 4.33161706e-02 -1.21453869e+00 9.12100136e-01
-5.92068076e-01 1.22373736e+00 5.75453117e-02 -6.59215271e-01
1.56374681e+00 -1.63131565e-01 2.63127536e-01 -1.04496992e+00
1.33936897e-01 3.48794371e-01 -4.79379743e-02 -2.68000811e-01
9.52843964e-01 -9.18801576e-02 2.96117127e-01 1.29673362e-01
-9.74790096e-01 -6.49747610e-01 2.81300247e-01 -2.52074540e-01
1.34527668e-01 -1.48190022e-01 5.52753270e-01 -1.66951492e-01
8.65721881e-01 2.87013799e-01 5.03507078e-01 2.91916937e-01
-2.44577870e-01 4.00358737e-01 1.88934565e-01 -4.44534183e-01
-1.31894803e+00 -8.58424544e-01 -5.73573589e-01 -1.75327063e-02
8.86532605e-01 2.74468869e-01 -8.18166077e-01 -5.56339789e-03
1.24298565e-01 6.17556036e-01 -2.45113924e-01 4.34204787e-02
-8.11936855e-01 -4.26352382e-01 -5.03125191e-02 4.53259170e-01
1.36462045e+00 -5.63700140e-01 -1.03094149e+00 1.48684829e-01
-3.39735508e-01 -1.17571831e+00 -2.78328240e-01 -2.50446767e-01
-1.01445806e+00 -1.01714158e+00 -6.47030115e-01 -9.97305393e-01
8.59477401e-01 9.33730185e-01 5.20565808e-01 -7.14475960e-02
-1.85513169e-01 -2.81064332e-01 9.41866264e-02 -4.50408638e-01
-1.22817636e-01 -2.48615131e-01 -3.87727350e-01 1.17161190e-02
6.10849619e-01 -2.12870657e-01 -9.19601083e-01 6.93978131e-01
-4.18219447e-01 5.69450319e-01 6.41133666e-01 4.73551691e-01
7.51796663e-01 2.95374304e-01 1.40067965e-01 -5.24520636e-01
3.20979595e-01 -2.23254375e-02 -1.13223779e+00 -2.74932683e-01
-7.46945918e-01 -1.29314706e-01 2.28676811e-01 1.38944834e-01
-1.46234190e+00 2.19023466e-01 -9.70570743e-02 1.63723044e-02
-1.68438569e-01 9.72077101e-02 -7.11814985e-02 -1.84349284e-01
4.39146012e-01 2.58492351e-01 5.16675353e-01 1.66823249e-02
3.45284611e-01 1.00456393e+00 9.74522531e-01 1.93571448e-01
7.81742334e-01 8.28110814e-01 2.64758050e-01 -9.56168056e-01
-3.54171842e-01 -7.82220542e-01 -7.37117171e-01 -5.96383631e-01
9.18637097e-01 -1.04767644e+00 -5.75879157e-01 8.82198274e-01
-8.74110997e-01 8.11198130e-02 6.55405596e-02 9.33708608e-01
-4.65953290e-01 6.44195139e-01 -2.71291077e-01 -7.31626451e-01
-2.47891232e-01 -1.19149184e+00 5.55036426e-01 7.70419180e-01
9.28059146e-02 -8.47822070e-01 -1.33839548e-01 4.28943008e-01
2.85009414e-01 7.34814629e-02 6.63220584e-01 6.50592685e-01
-8.42818975e-01 -2.73598433e-01 -7.22729445e-01 1.93619832e-01
1.39054105e-01 3.27405304e-01 -8.54974151e-01 3.34255010e-01
-6.14394918e-02 4.88402486e-01 8.91820788e-01 8.73981893e-01
6.09600663e-01 1.23940043e-01 -5.57823777e-01 7.94001043e-01
1.55871439e+00 5.72435975e-01 1.21508968e+00 8.69474530e-01
6.18962824e-01 7.73252964e-01 1.27977359e+00 2.27882341e-01
5.25587380e-01 6.40701652e-01 4.21823382e-01 -3.25303197e-01
-3.00021887e-01 -1.67287499e-01 -9.14356112e-03 4.22502697e-01
-9.88762900e-02 1.40393972e-01 -8.80069911e-01 5.36664367e-01
-1.48113585e+00 -9.68732238e-01 -7.25470364e-01 2.48221898e+00
6.12069070e-01 2.24677503e-01 -2.18871802e-01 3.21056575e-01
1.13038838e+00 8.79516974e-02 -4.95433182e-01 -8.75336587e-01
-9.49250832e-02 -1.14876635e-01 1.10101938e+00 9.73447025e-01
-7.92716444e-01 9.40680325e-01 6.09377146e+00 8.43591452e-01
-1.27250624e+00 -5.54257214e-01 5.52020609e-01 4.08569068e-01
-3.05964440e-01 1.02263004e-01 -1.15784168e+00 5.87090313e-01
3.97266865e-01 -3.71660203e-01 1.36782497e-01 5.95852077e-01
6.62007153e-01 -1.24471223e+00 -2.04478443e-01 1.14665949e+00
-2.90097564e-01 -1.01993775e+00 -1.64653242e-01 3.65326613e-01
8.27149868e-01 -1.56848505e-01 1.52574688e-01 -2.50250518e-01
-7.07773417e-02 -5.22898018e-01 4.06318188e-01 5.23356080e-01
9.82654512e-01 -1.04360783e+00 8.88089001e-01 4.07220036e-01
-1.45782268e+00 3.93619388e-02 -5.31193197e-01 -2.65954077e-01
4.86759394e-01 8.35105300e-01 -1.25562024e+00 3.62400264e-01
3.55854690e-01 6.76290095e-01 -4.00689453e-01 1.37404966e+00
-3.97878110e-01 5.74031547e-02 -3.62802595e-01 1.93918273e-01
2.66438246e-01 -1.01715803e+00 4.55460668e-01 8.60871971e-01
4.60304797e-01 -1.10961914e-01 -3.87576699e-01 4.81538981e-01
3.24619025e-01 2.02688277e-01 -7.60234416e-01 7.82451868e-01
6.51245356e-01 1.01959825e+00 -6.66962206e-01 -4.40847188e-01
-4.29331392e-01 5.18559277e-01 -6.64019883e-02 3.77418965e-01
-6.21858299e-01 -7.47805774e-01 6.26083255e-01 4.01405334e-01
1.87713623e-01 -2.48793140e-01 -8.43977928e-01 -6.12674773e-01
1.64246544e-01 -1.93159372e-01 -1.09204717e-01 -8.62708628e-01
-3.14267665e-01 4.04992312e-01 8.45587552e-02 -1.53867257e+00
-5.07656634e-01 -3.81989419e-01 -7.02660620e-01 1.28765678e+00
-2.07349944e+00 -6.22856379e-01 -6.46236062e-01 3.81246448e-01
4.30734426e-01 3.12827341e-02 9.73209813e-02 1.73853427e-01
-5.96713960e-01 2.21624374e-01 2.96557605e-01 -1.08706564e-01
4.85594988e-01 -8.20613444e-01 2.72205651e-01 1.22904158e+00
-6.31196022e-01 2.23575190e-01 7.28019953e-01 -6.28168762e-01
-6.38967454e-01 -7.92800665e-01 9.87277567e-01 2.04191193e-01
2.07623348e-01 2.56880224e-01 -7.88902104e-01 1.89036474e-01
1.39792070e-01 -3.73011112e-01 2.28150133e-02 -5.13723731e-01
2.34635025e-01 -5.41529894e-01 -1.22018647e+00 5.93188167e-01
5.40198505e-01 -5.60095131e-01 -3.89482707e-01 -4.23032105e-01
2.02686768e-02 -6.12535894e-01 -3.92321020e-01 4.55779046e-01
7.88262844e-01 -1.32073498e+00 8.80755603e-01 5.20363033e-01
2.38451362e-01 -7.58069158e-01 2.37742573e-01 -1.04805732e+00
1.34391636e-01 -1.80552170e-01 3.28326523e-01 8.86135876e-01
3.77551645e-01 -1.00850773e+00 9.00956690e-01 6.24452174e-01
-1.45358220e-01 -7.56997049e-01 -8.79417777e-01 -2.01645300e-01
-5.55274904e-01 -1.48649678e-01 4.18061286e-01 4.07748997e-01
-2.32307628e-01 9.54723805e-02 -1.06695101e-01 2.01109484e-01
7.58618474e-01 3.18188548e-01 1.00495410e+00 -1.00748050e+00
4.87737745e-01 -4.62981284e-01 -7.41963208e-01 -1.60355675e+00
-8.95908400e-02 -3.02196085e-01 9.67747420e-02 -1.65223873e+00
-3.56515720e-02 -5.16152024e-01 3.11576277e-01 -1.91703439e-01
-3.72336447e-01 4.45368648e-01 -2.09561720e-01 3.77022266e-01
3.34388196e-01 2.76762515e-01 1.58235776e+00 1.56473070e-01
-4.04750258e-01 4.64068741e-01 -4.28072602e-01 9.27790821e-01
1.07247710e+00 -3.78519073e-02 -5.43923736e-01 -4.25759614e-01
-5.26726097e-02 1.19977899e-01 9.28875580e-02 -9.63844240e-01
3.88127387e-01 -4.50253002e-02 3.18530291e-01 -1.33513606e+00
4.68418509e-01 -7.89503753e-01 -4.84372750e-02 5.31251132e-01
2.53901660e-01 -2.23769690e-03 2.61766791e-01 4.28359181e-01
-5.52014291e-01 -3.43710780e-01 1.13973439e+00 3.04675847e-01
-1.17988312e+00 3.33976187e-03 -5.54809809e-01 -4.56462592e-01
1.39117944e+00 -1.29356599e+00 -2.57322729e-01 -5.16288221e-01
-9.25097615e-02 2.93670863e-01 7.84921050e-01 -1.16078973e-01
1.00611627e+00 -1.20197356e+00 -6.21175349e-01 5.92435002e-01
3.54733802e-02 1.99617788e-01 4.12604123e-01 1.03932667e+00
-1.31620395e+00 4.94553685e-01 -4.79222596e-01 -7.89845824e-01
-1.56180096e+00 1.56213909e-01 3.13164830e-01 1.19961448e-01
-6.16635501e-01 3.31476152e-01 1.95672318e-01 1.16300382e-01
-1.49706304e-01 -3.75792205e-01 -5.41900635e-01 5.22389784e-02
5.49222708e-01 7.55077720e-01 -1.62657201e-01 -9.73214269e-01
-1.85945407e-01 1.47269452e+00 -9.25763398e-02 -2.63360441e-01
6.65540159e-01 -9.34803545e-01 1.19883409e-02 -6.28688857e-02
1.16950595e+00 3.05609673e-01 -1.61880195e+00 -1.47045910e-01
-3.84314924e-01 -1.05931044e+00 4.10122812e-01 -2.25878984e-01
-1.08944523e+00 9.59470689e-01 6.59536958e-01 -1.50267914e-01
1.33287871e+00 -4.65985239e-01 9.05022800e-01 -4.05639708e-02
2.42490187e-01 -1.32723343e+00 -6.30051374e-01 3.13793659e-01
2.64527172e-01 -1.14762950e+00 2.37671942e-01 -8.43876421e-01
-7.52046645e-01 1.35697460e+00 5.95355093e-01 1.46132693e-01
3.71409923e-01 2.58373350e-01 3.36006641e-01 1.66356415e-01
-7.34622777e-02 -5.69800854e-01 1.03371643e-01 6.86749816e-01
-3.37401591e-02 -1.73039418e-02 -5.56019604e-01 -7.84071565e-01
-3.32691818e-01 -1.51063338e-01 7.92817414e-01 6.60284281e-01
-1.17558515e+00 -8.03397894e-01 -6.47843182e-01 3.46781522e-01
1.55518390e-02 3.97355203e-03 1.46044746e-01 8.32652330e-01
1.02386355e-01 1.12017453e+00 5.69320917e-01 -1.46345228e-01
3.83089095e-01 -5.45685887e-01 3.57532561e-01 -1.91543579e-01
3.05756837e-01 -1.92979500e-02 2.29611546e-01 -4.07034576e-01
-4.15088743e-01 -4.49655324e-01 -1.52042234e+00 -5.86918354e-01
-3.72619987e-01 4.08310629e-02 9.92635846e-01 6.25760734e-01
1.37416422e-01 -1.27581969e-01 1.07984173e+00 -5.75039327e-01
2.25333478e-02 -4.87654477e-01 -6.24202669e-01 -4.49884832e-02
2.27580532e-01 -5.74413478e-01 -4.54163939e-01 1.01592429e-02] | [8.85372543334961, -2.3170382976531982] |
7916a54f-2ff7-4c20-afb5-333b53384beb | revisiting-self-supervised-visual | 1901.09005 | null | http://arxiv.org/abs/1901.09005v1 | http://arxiv.org/pdf/1901.09005v1.pdf | Revisiting Self-Supervised Visual Representation Learning | Unsupervised visual representation learning remains a largely unsolved
problem in computer vision research. Among a big body of recently proposed
approaches for unsupervised learning of visual representations, a class of
self-supervised techniques achieves superior performance on many challenging
benchmarks. A large number of the pretext tasks for self-supervised learning
have been studied, but other important aspects, such as the choice of
convolutional neural networks (CNN), has not received equal attention.
Therefore, we revisit numerous previously proposed self-supervised models,
conduct a thorough large scale study and, as a result, uncover multiple crucial
insights. We challenge a number of common practices in selfsupervised visual
representation learning and observe that standard recipes for CNN design do not
always translate to self-supervised representation learning. As part of our
study, we drastically boost the performance of previously proposed techniques
and outperform previously published state-of-the-art results by a large margin. | ['Alexander Kolesnikov', 'Lucas Beyer', 'Xiaohua Zhai'] | 2019-01-25 | revisiting-self-supervised-visual-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Kolesnikov_Revisiting_Self-Supervised_Visual_Representation_Learning_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Kolesnikov_Revisiting_Self-Supervised_Visual_Representation_Learning_CVPR_2019_paper.pdf | cvpr-2019-6 | ['self-supervised-image-classification'] | ['computer-vision'] | [ 5.81471145e-01 5.71093597e-02 -6.52225733e-01 -3.67248714e-01
-3.18800837e-01 -6.29342318e-01 8.59043002e-01 1.91570491e-01
-2.37407878e-01 5.18784523e-01 3.64479303e-01 -2.14705184e-01
-1.71803877e-01 -4.24952239e-01 -6.43988431e-01 -7.08208025e-01
1.86886583e-02 2.14061588e-01 9.87901911e-02 -2.51161039e-01
4.12188530e-01 4.17979151e-01 -1.88609493e+00 3.42431009e-01
6.01317465e-01 9.23290908e-01 -1.61872998e-01 4.96327341e-01
-1.25300735e-01 1.30284131e+00 -3.82367730e-01 -1.27054840e-01
1.38365865e-01 -4.78705913e-01 -1.10269284e+00 4.66328889e-01
6.68748677e-01 3.96027863e-02 -4.05533671e-01 1.00941157e+00
2.68765628e-01 2.10877225e-01 1.12610602e+00 -1.36412656e+00
-9.62073088e-01 5.46757936e-01 -7.35199988e-01 2.72789448e-01
-1.92085892e-01 8.39172602e-02 1.27098894e+00 -7.21733689e-01
6.89452887e-01 8.98225904e-01 5.66168249e-01 3.76453698e-01
-1.24268603e+00 -5.50119102e-01 1.32723540e-01 2.33921051e-01
-1.14516079e+00 -5.28331995e-01 9.83033061e-01 -6.32985473e-01
8.15992355e-01 7.56829605e-02 5.03307700e-01 1.19721806e+00
-2.56078064e-01 9.78596270e-01 1.36382818e+00 -7.39743173e-01
1.60879225e-01 3.45237941e-01 4.39506561e-01 8.39405179e-01
5.23937583e-01 2.05929592e-01 -4.65357989e-01 1.50159122e-02
8.30303967e-01 2.71215290e-01 3.26583348e-02 -9.12170649e-01
-1.12499058e+00 1.01321328e+00 6.89475179e-01 4.51193243e-01
6.27861544e-02 2.82343864e-01 6.37500107e-01 5.49231291e-01
5.00787318e-01 4.86156553e-01 -2.59341776e-01 7.52034113e-02
-8.70163560e-01 -1.06003113e-01 4.76632595e-01 1.03047740e+00
1.08806384e+00 4.49404746e-01 -1.89966559e-01 8.86434376e-01
6.60270825e-02 9.22346041e-02 6.12379432e-01 -5.75836122e-01
1.77317694e-01 8.00174654e-01 -3.44105691e-01 -1.07508516e+00
-2.83688873e-01 -7.58296311e-01 -1.26097763e+00 3.27698886e-01
4.16593701e-01 5.38987704e-02 -1.02163708e+00 1.35432923e+00
-1.64257869e-01 -4.33913805e-02 1.34210408e-01 7.37013698e-01
1.03807306e+00 3.62356633e-01 2.12138146e-01 -1.89828351e-01
1.05557001e+00 -1.34911811e+00 -4.96555567e-01 -3.95103663e-01
5.93381464e-01 -6.20025516e-01 7.57639647e-01 3.64594102e-01
-7.34652340e-01 -8.18820059e-01 -1.08588219e+00 -5.83589785e-02
-4.93682444e-01 2.92633176e-01 1.17509413e+00 5.19173443e-01
-9.88620996e-01 5.84187567e-01 -5.22192836e-01 -8.24633002e-01
8.43837023e-01 2.52387077e-01 -5.61607420e-01 -9.81648192e-02
-6.27846122e-01 7.16828167e-01 4.67854857e-01 -2.49249175e-01
-1.05771899e+00 -3.77348393e-01 -8.23589265e-01 -5.88613600e-02
3.22715729e-01 -5.01652181e-01 1.19644940e+00 -1.53494275e+00
-1.14428735e+00 1.28313637e+00 -1.20795384e-01 -7.52831399e-01
3.58409762e-01 -9.51042771e-02 -2.89738178e-01 2.96571434e-01
-7.16863796e-02 7.71426380e-01 1.22114325e+00 -1.56058848e+00
-5.04540384e-01 -8.98026451e-02 -1.45713724e-02 3.92423235e-02
-8.75724316e-01 -2.43272837e-02 -3.58190089e-01 -8.84649515e-01
-1.01384170e-01 -7.81503201e-01 -3.85951549e-01 1.58365294e-01
-4.80040610e-01 -2.97504395e-01 6.30595624e-01 1.95769891e-01
9.37291145e-01 -2.38752985e+00 1.00943465e-02 7.62050152e-02
6.15681171e-01 3.66749227e-01 -1.06126443e-01 4.53907818e-01
-4.85606313e-01 1.06444724e-01 -4.18566823e-01 -5.55797935e-01
-1.29409999e-01 3.17565471e-01 -6.01660430e-01 6.27696276e-01
3.45218688e-01 1.17743540e+00 -9.51663196e-01 -6.08746111e-01
2.85297006e-01 2.80156314e-01 -2.98970610e-01 3.54725689e-01
-3.42032947e-02 3.89631391e-01 -2.92734444e-01 5.54407656e-01
4.28089947e-01 -7.50263453e-01 1.89551219e-01 -2.16903806e-01
-1.34525061e-01 1.84235647e-01 -8.83485913e-01 1.65604174e+00
8.67374912e-02 8.82845044e-01 -4.62152660e-01 -1.69230235e+00
1.08755481e+00 3.18653174e-02 5.41065872e-01 -6.52124345e-01
1.81497931e-01 -1.22408293e-01 5.04552349e-02 -4.44172621e-01
6.28276050e-01 -4.19421196e-02 2.26863146e-01 6.33986235e-01
5.27832866e-01 3.84651601e-01 2.99971372e-01 2.41153255e-01
9.50208306e-01 5.06532416e-02 7.65372157e-01 -3.65282595e-01
7.11566359e-02 3.55324179e-01 1.22187771e-01 9.60613728e-01
-3.14316690e-01 9.48474586e-01 6.18547797e-01 -5.97700357e-01
-9.34239149e-01 -8.24536085e-01 -1.90085843e-02 1.40545070e+00
-7.52823055e-02 -4.86834526e-01 -4.08549458e-01 -8.92105937e-01
-8.24605860e-03 1.20920269e-03 -1.13041699e+00 -1.68231502e-01
-3.88333261e-01 -7.13685095e-01 4.15498585e-01 7.58104086e-01
2.37270519e-01 -1.34895337e+00 -5.05949378e-01 -1.17753498e-01
4.57175285e-01 -1.18495929e+00 -9.78744030e-02 6.41040862e-01
-1.12666607e+00 -1.25716090e+00 -8.43392193e-01 -1.21987224e+00
1.15114009e+00 9.86542523e-01 1.36692810e+00 5.10292590e-01
-3.92989099e-01 4.43514884e-01 -6.27570868e-01 -4.06186402e-01
-2.84460217e-01 3.29457581e-01 -1.67208612e-01 1.80309802e-01
2.70281911e-01 -6.06921256e-01 -5.14950335e-01 8.06474164e-02
-1.02720177e+00 -5.26168048e-02 8.06854725e-01 8.14170599e-01
5.99978447e-01 -1.32834464e-01 6.45221233e-01 -1.52358711e+00
3.20117354e-01 -4.67020929e-01 -3.48398894e-01 1.47339761e-01
-7.14023292e-01 1.08293153e-01 6.84896171e-01 -5.56278765e-01
-6.00653231e-01 2.61293530e-01 2.30626658e-01 -5.41114807e-01
-3.80892754e-01 4.74399686e-01 2.54559726e-01 -1.95759237e-01
8.75864387e-01 4.21333075e-01 1.35842144e-01 -3.84222716e-01
6.99018061e-01 5.12029946e-01 4.60400999e-01 -4.05072480e-01
1.13503695e+00 7.89754331e-01 -1.16483048e-01 -9.43305731e-01
-1.09071958e+00 -6.77414775e-01 -8.48130703e-01 -9.16980058e-02
5.76714277e-01 -9.32017922e-01 -2.98129946e-01 1.59047812e-01
-8.36522043e-01 -4.18215603e-01 -5.59895635e-01 1.51798695e-01
-5.77622294e-01 5.98194659e-01 -1.11346990e-01 -8.22440922e-01
-2.37610012e-01 -8.94488990e-01 7.29358792e-01 2.81127304e-01
-3.65725249e-01 -1.12734210e+00 1.81886107e-01 2.45493039e-01
5.83525956e-01 1.56935468e-01 8.49433839e-01 -6.47298813e-01
-5.19969463e-01 3.27088945e-02 -5.80521047e-01 4.19866502e-01
2.90849507e-01 4.95471694e-02 -1.18254387e+00 -5.08050263e-01
-4.62779492e-01 -8.16378653e-01 1.43067217e+00 2.64957041e-01
1.29476643e+00 -8.28436464e-02 -1.84384227e-01 8.05534601e-01
1.40640974e+00 -2.64269114e-01 7.39677489e-01 4.94632095e-01
8.65566909e-01 6.12376094e-01 3.38187277e-01 2.67605931e-01
2.85352677e-01 2.55440682e-01 5.90294719e-01 -5.08531690e-01
-2.13509738e-01 -2.62218088e-01 1.80370748e-01 8.68332326e-01
-3.67122531e-01 1.08633153e-02 -6.87645137e-01 8.42133403e-01
-1.92973447e+00 -9.34788823e-01 2.86959764e-02 2.13685822e+00
6.07211709e-01 1.67196169e-01 3.91635001e-01 3.49539459e-01
5.73218822e-01 5.30487835e-01 -4.76416796e-01 -5.93693219e-02
-3.14209670e-01 3.43768746e-01 5.16013920e-01 -1.82256848e-01
-1.41158593e+00 1.18947828e+00 7.26924467e+00 6.11948311e-01
-1.20307207e+00 -5.74381836e-02 5.99802136e-01 3.45274955e-01
-1.14032768e-01 1.03915080e-01 -5.40948808e-01 -4.74999547e-02
4.53170747e-01 4.34278361e-02 1.69926986e-01 1.17861617e+00
-2.78274655e-01 1.45949662e-01 -1.14334857e+00 1.26611471e+00
4.30081248e-01 -1.61631417e+00 1.35588557e-01 -1.40874967e-01
1.23927236e+00 1.17062256e-01 2.22728387e-01 4.22470897e-01
3.02940339e-01 -1.40747821e+00 4.33104545e-01 2.48425201e-01
8.39553237e-01 -6.65376723e-01 3.76059741e-01 -9.11743268e-02
-1.27295232e+00 -1.51164040e-01 -6.26816273e-01 -1.35447785e-01
-3.33903223e-01 4.34511065e-01 -2.66515702e-01 3.57425272e-01
6.52580082e-01 1.31833100e+00 -1.03055131e+00 1.23219585e+00
-3.75316799e-01 1.05465782e+00 1.85071081e-01 4.08024760e-03
4.23897088e-01 1.53166771e-01 1.86272115e-01 1.38302088e+00
-3.55190814e-01 -3.58923763e-01 4.70685996e-02 6.60295725e-01
-3.50618601e-01 2.28772759e-01 -9.34880495e-01 -4.00649190e-01
2.70531382e-02 1.37548602e+00 -9.57658529e-01 -3.66228491e-01
-4.91989940e-01 8.24807465e-01 8.13347042e-01 5.06804645e-01
-3.47218603e-01 -1.18418209e-01 4.28476274e-01 8.78867656e-02
4.84033108e-01 -3.30161035e-01 -3.20228398e-01 -1.21381247e+00
-1.23324990e-01 -9.93149042e-01 6.47517920e-01 -5.80883920e-01
-1.61120784e+00 5.82329154e-01 -1.26112968e-01 -1.62764156e+00
-1.86611995e-01 -7.48345435e-01 -7.47983515e-01 3.90753075e-02
-1.98990369e+00 -1.21167970e+00 -4.47130471e-01 6.66968644e-01
6.10930204e-01 -5.99960327e-01 8.77023458e-01 3.66197936e-02
-4.81251866e-01 7.06648707e-01 2.60261208e-01 4.56458092e-01
7.90385127e-01 -1.24307096e+00 3.86380196e-01 6.42796814e-01
7.47574270e-01 8.42097700e-01 4.27508891e-01 -1.93049297e-01
-1.53121054e+00 -1.24641490e+00 5.27212322e-01 -3.46948653e-01
7.94925392e-01 -3.85150075e-01 -9.71384346e-01 6.79018378e-01
3.87429535e-01 5.00303268e-01 7.23273396e-01 2.10778743e-01
-8.12710106e-01 -8.50623921e-02 -6.51372790e-01 4.62015927e-01
1.20388377e+00 -6.33477807e-01 -5.79357624e-01 3.79944175e-01
4.18524176e-01 6.44451156e-02 -4.20906961e-01 2.81387955e-01
4.81120616e-01 -9.57028329e-01 1.12098014e+00 -1.06185102e+00
7.13237703e-01 -1.45449132e-01 1.93975531e-02 -1.08011639e+00
-3.39741975e-01 -6.75836325e-01 -3.22463542e-01 1.14871991e+00
1.89798832e-01 -5.35339952e-01 9.61464345e-01 -1.87465519e-01
-2.96120383e-02 -8.01160872e-01 -4.53603268e-01 -8.00786257e-01
1.18356727e-01 -3.03108901e-01 -9.24953222e-02 1.23388851e+00
-1.68464687e-02 5.45829952e-01 -7.14263320e-01 -2.86208928e-01
9.89623427e-01 3.45641822e-01 9.80157554e-01 -1.36395431e+00
-1.20412387e-01 -6.14773571e-01 -6.71606600e-01 -1.19031477e+00
1.73693419e-01 -1.17286348e+00 -2.14763269e-01 -1.70192134e+00
5.93161523e-01 -2.30680048e-01 -5.66484153e-01 6.65008724e-01
-1.57454293e-02 7.55250096e-01 9.76772383e-02 5.06983280e-01
-1.06076503e+00 4.84965831e-01 1.07849216e+00 -4.35263038e-01
2.90559023e-03 -3.07112336e-01 -1.10732710e+00 7.08023727e-01
8.73141468e-01 -5.86227596e-01 -6.19649589e-01 -4.20185328e-01
1.53603449e-01 -6.08564615e-01 4.02931958e-01 -9.49595273e-01
2.72687048e-01 -2.54209429e-01 4.39913154e-01 -3.42006475e-01
-1.42671034e-01 -6.63491666e-01 -3.72468501e-01 2.04660431e-01
-5.85340500e-01 -2.53808480e-02 -3.83084230e-02 6.63200498e-01
-3.18856597e-01 -1.86679184e-01 9.78287220e-01 -2.17008799e-01
-1.00729370e+00 3.57448041e-01 -4.14434910e-01 2.24046230e-01
8.25595796e-01 -5.29042602e-01 -4.18031007e-01 -3.22768867e-01
-4.22634423e-01 -2.94151679e-02 3.64204735e-01 6.57629728e-01
6.40662491e-01 -1.23045802e+00 -6.17933273e-01 1.25852793e-01
6.69169188e-01 -5.02896272e-02 1.29331857e-01 6.83324814e-01
-2.58808315e-01 3.47159356e-01 -3.97111237e-01 -6.84259832e-01
-1.13973939e+00 5.73003113e-01 -8.41054171e-02 -2.65361577e-01
-8.04331541e-01 6.60073280e-01 2.93872297e-01 -1.65054023e-01
5.76766193e-01 -5.87653816e-02 -6.49109423e-01 2.00828210e-01
5.71188867e-01 1.15577571e-01 -1.06006898e-01 -8.20161760e-01
-1.19152531e-01 6.97858512e-01 -3.51448268e-01 4.82707292e-01
1.71967912e+00 1.08843461e-01 3.12627107e-02 5.45631766e-01
1.41057515e+00 -4.06542003e-01 -1.13765049e+00 -3.69967312e-01
4.86115851e-02 -1.83854073e-01 -1.38987124e-01 -2.97220379e-01
-1.38060164e+00 9.15875971e-01 5.72156131e-01 3.47888708e-01
1.02879238e+00 2.20433578e-01 1.35340855e-01 7.40518272e-01
1.99945673e-01 -9.96747255e-01 7.04018235e-01 6.50840282e-01
7.49702692e-01 -1.65132523e+00 3.43886584e-01 -3.68101597e-01
-7.82317936e-01 1.20395696e+00 5.91804087e-01 -5.78067899e-01
5.39389491e-01 1.02613188e-01 2.35370725e-01 -4.68728960e-01
-4.57699507e-01 -7.70649850e-01 4.92416918e-01 9.36911285e-01
6.32389426e-01 -2.89890051e-01 6.39271438e-02 2.47252285e-01
1.45021509e-02 -9.86244157e-02 2.59938031e-01 1.07622325e+00
-4.44895864e-01 -1.02464616e+00 -5.96902333e-02 5.61653256e-01
-1.71770722e-01 -1.03122845e-01 -4.61429566e-01 7.10653603e-01
-1.13725558e-01 7.57915437e-01 1.06340557e-01 -2.27892429e-01
1.30814627e-01 -1.38669193e-01 4.41573560e-01 -7.99874127e-01
-6.89925849e-01 -2.18182400e-01 -1.29396021e-01 -1.40968978e-01
-9.23538148e-01 -3.61089826e-01 -9.89578247e-01 -1.78616736e-02
-2.50614256e-01 -1.25978589e-01 2.84844756e-01 1.02273047e+00
1.83086589e-01 4.15399909e-01 7.64000177e-01 -9.03490186e-01
-4.12681699e-01 -9.29051161e-01 -5.96459270e-01 7.52510130e-01
3.26913834e-01 -8.00431907e-01 -3.03907722e-01 3.87462318e-01] | [9.50986385345459, 2.412519693374634] |
6c373075-6c2d-4fd8-8ffa-b633e58b7141 | latent-odes-for-irregularly-sampled-time | 1907.03907 | null | https://arxiv.org/abs/1907.03907v1 | https://arxiv.org/pdf/1907.03907v1.pdf | Latent ODEs for Irregularly-Sampled Time Series | Time series with non-uniform intervals occur in many applications, and are difficult to model using standard recurrent neural networks (RNNs). We generalize RNNs to have continuous-time hidden dynamics defined by ordinary differential equations (ODEs), a model we call ODE-RNNs. Furthermore, we use ODE-RNNs to replace the recognition network of the recently-proposed Latent ODE model. Both ODE-RNNs and Latent ODEs can naturally handle arbitrary time gaps between observations, and can explicitly model the probability of observation times using Poisson processes. We show experimentally that these ODE-based models outperform their RNN-based counterparts on irregularly-sampled data. | ['Yulia Rubanova', 'David Duvenaud', 'Ricky T. Q. Chen'] | 2019-07-08 | null | null | null | null | ['multivariate-time-series-imputation'] | ['time-series'] | [-1.11887738e-01 -5.23636006e-02 -1.36900619e-01 9.26598981e-02
-1.03937939e-01 -3.37998837e-01 7.52923548e-01 -4.41464692e-01
-2.10607916e-01 8.16275716e-01 1.59479275e-01 -5.05285442e-01
-6.13241270e-02 -7.67107427e-01 -4.00281847e-01 -8.15889597e-01
-3.59165549e-01 6.34566844e-01 -1.32920578e-01 -1.19037740e-01
-2.87170082e-01 6.13076031e-01 -1.21178162e+00 -3.49685222e-01
4.79070812e-01 7.87995636e-01 -1.82909057e-01 9.47569788e-01
-9.32374150e-02 1.25566995e+00 -7.65107691e-01 1.98262468e-01
3.93031724e-02 -3.54355633e-01 -1.36136472e-01 -1.89779133e-01
-6.86383367e-01 -4.00183797e-01 -1.15753496e+00 5.58936775e-01
1.31490663e-01 3.79819691e-01 1.07090497e+00 -1.39174509e+00
-1.34624815e+00 4.75140721e-01 -2.46176094e-01 5.65667629e-01
-2.02541556e-02 1.29360100e-02 6.71021461e-01 -5.76483786e-01
2.16546625e-01 1.33792531e+00 1.09821343e+00 7.21465826e-01
-1.47459698e+00 -4.50018704e-01 2.23488882e-01 -9.74753201e-02
-1.25538063e+00 -1.26678884e-01 5.46171069e-01 -4.43251312e-01
1.19653690e+00 2.00153384e-02 6.76690102e-01 2.04735756e+00
4.08780754e-01 9.01511073e-01 9.36604261e-01 -1.48102030e-01
3.56530249e-01 -6.08364642e-01 3.45986813e-01 1.17085546e-01
1.22997016e-01 2.60403842e-01 -7.09742829e-02 -4.57964808e-01
1.63177490e+00 9.56926584e-01 1.29429353e-02 1.48678944e-01
-1.11840630e+00 9.96911407e-01 -1.26398042e-01 2.83078164e-01
-8.06656361e-01 5.36015272e-01 3.07566822e-01 5.43496907e-01
6.44756317e-01 -3.77874337e-02 -3.21937919e-01 -3.70817661e-01
-7.44720697e-01 1.61861762e-01 1.29023254e+00 1.06123185e+00
2.79584587e-01 7.83754051e-01 -2.10859552e-01 7.08577275e-01
8.22125655e-03 5.87721825e-01 8.40278029e-01 -9.43776131e-01
-1.34524042e-02 1.21495478e-01 2.80713290e-01 -6.73951566e-01
-3.75206381e-01 -3.76406312e-01 -1.64078522e+00 -2.54950762e-01
3.13651115e-01 -3.46052557e-01 -1.12390292e+00 1.70807135e+00
-3.48415583e-01 7.29629517e-01 3.82397175e-01 2.62127191e-01
2.59657115e-01 1.23269391e+00 -3.12853694e-01 -7.92551637e-01
6.11313283e-01 -8.02557409e-01 -1.23855090e+00 3.21872503e-01
2.69054949e-01 -1.72273248e-01 4.90516275e-01 6.00649603e-02
-1.22512162e+00 -2.76387125e-01 -4.46616709e-01 5.59233055e-02
-4.07961577e-01 -1.41149700e-01 3.23725611e-01 -5.75547963e-02
-1.27474749e+00 7.55026996e-01 -1.43674541e+00 -9.82524455e-02
-1.83225647e-01 2.62986511e-01 3.32158506e-01 4.77641851e-01
-1.49979138e+00 7.77572453e-01 -1.42067835e-01 6.97364748e-01
-1.28133142e+00 -5.47808826e-01 -9.37173545e-01 -9.23515037e-02
2.42557496e-01 -4.82876748e-01 1.71531177e+00 -3.74611765e-01
-1.73765540e+00 2.98426211e-01 -5.93564749e-01 -9.09776151e-01
5.33932269e-01 -1.72616452e-01 -8.63041162e-01 -2.20795065e-01
-4.96699614e-03 -5.90580143e-02 9.50917244e-01 -7.16581881e-01
5.55245876e-02 1.87052622e-01 -3.23799103e-01 -3.40965956e-01
-1.40971830e-02 -1.29948407e-01 -1.49177745e-01 -1.00960207e+00
1.29682533e-02 -1.13533413e+00 -6.06273413e-01 -2.58636028e-01
-4.88987446e-01 -5.25161028e-01 1.14878809e+00 -5.30677497e-01
1.47258496e+00 -1.92532802e+00 1.90363616e-01 -1.46275721e-02
4.55952734e-01 2.71678060e-01 -1.00879848e-01 7.69964874e-01
-2.34163180e-01 -2.46157479e-02 -3.12409937e-01 -7.74079978e-01
2.87519023e-02 1.05953670e+00 -8.67449939e-01 4.84499395e-01
1.95104241e-01 1.05900419e+00 -7.67086446e-01 1.32287771e-01
1.02661863e-01 6.15046322e-01 1.09599955e-01 3.01216573e-01
-2.86234468e-01 3.63445878e-01 -2.12428644e-01 5.25785804e-01
1.21742629e-01 -6.38941348e-01 -2.18264878e-01 5.39593816e-01
-3.94189954e-01 2.26667240e-01 -9.08708274e-01 6.55818582e-01
-5.98800123e-01 1.01306510e+00 -4.31519926e-01 -9.27980304e-01
1.02794206e+00 7.54360616e-01 4.25600171e-01 -3.40789855e-01
6.14763983e-02 1.23662002e-01 -1.43931419e-01 -1.79737449e-01
4.59351838e-01 -1.56781539e-01 -1.15256868e-01 7.10905373e-01
-9.27729253e-03 5.29594839e-01 8.99188891e-02 -1.02709703e-01
1.29196644e+00 -2.60380268e-01 2.65496850e-01 1.40128940e-01
-6.64365478e-03 -5.08650422e-01 7.34740019e-01 1.11827254e+00
8.10532738e-03 5.06873131e-01 7.26426959e-01 -5.80927908e-01
-1.40605867e+00 -1.45642340e+00 -9.80318040e-02 5.85809410e-01
-3.80553365e-01 -1.38846368e-01 -2.24087015e-01 1.19677849e-01
-1.16687335e-01 5.99306822e-01 -8.73182118e-01 -7.46675953e-03
-5.39193034e-01 -7.52512753e-01 8.95245254e-01 1.01050639e+00
7.91974887e-02 -1.36644924e+00 -3.40561926e-01 9.22764719e-01
-7.12664947e-02 -1.02710986e+00 -4.37729806e-01 5.09943187e-01
-8.79844129e-01 -5.67046285e-01 -1.44242096e+00 -5.48884094e-01
4.00592774e-01 -3.64614800e-02 9.57280576e-01 -4.19177204e-01
-2.24096149e-01 5.41846812e-01 -6.92650974e-02 -5.27784050e-01
-6.23754680e-01 6.09907508e-02 5.28142571e-01 1.49094192e-02
3.79255444e-01 -1.03597069e+00 -3.15095812e-01 4.71877515e-01
-1.17982137e+00 -1.57643273e-01 2.47021765e-01 8.32788289e-01
6.36378169e-01 -1.27292946e-01 6.69477105e-01 -4.12790865e-01
9.65919316e-01 -7.32399464e-01 -8.10342729e-01 2.19169468e-01
-2.25888208e-01 2.00760394e-01 7.46012688e-01 -1.27422667e+00
-5.96217513e-01 -5.02859950e-01 1.30970761e-01 -1.14902306e+00
5.09907957e-03 5.69607615e-01 6.17519140e-01 6.61005676e-01
6.96692616e-02 6.01685643e-01 1.46908954e-01 -3.80820662e-01
8.38229209e-02 4.20970500e-01 6.45765305e-01 -3.15840840e-01
5.74623406e-01 6.35158241e-01 5.06442264e-02 -9.90990758e-01
-7.99418330e-01 -3.92639518e-01 -4.03966218e-01 6.83285967e-02
7.75154531e-01 -9.76044238e-01 -9.85182047e-01 9.73126233e-01
-1.54458523e+00 -8.38430107e-01 -7.13832617e-01 5.88330507e-01
-9.05007124e-01 -8.37994367e-03 -1.59674048e+00 -1.52398431e+00
1.97631940e-02 -5.61113834e-01 7.94109643e-01 9.19697508e-02
-4.32347625e-01 -1.85067451e+00 4.68257546e-01 -9.39606726e-01
6.42531574e-01 3.88182253e-01 6.95810795e-01 -6.70779407e-01
-3.63632202e-01 -4.35017735e-01 -1.48195550e-02 4.33283895e-01
2.89487839e-01 2.44612545e-01 -5.93559027e-01 1.68939874e-01
4.85394746e-01 1.84784636e-01 8.53808343e-01 9.27275121e-01
1.20395684e+00 -4.40748155e-01 -2.77343124e-01 5.14020920e-01
1.12796426e+00 4.17209238e-01 6.99074149e-01 -2.34209164e-03
7.00570464e-01 2.44324833e-01 -2.76689112e-01 6.62600398e-01
3.33239257e-01 1.93187043e-01 2.66348481e-01 -1.20372280e-01
6.27522349e-01 -2.18032971e-01 7.95083523e-01 1.47246897e+00
-3.67958188e-01 -3.76361430e-01 -9.27043974e-01 7.21972764e-01
-2.24610901e+00 -1.31526601e+00 -5.66821158e-01 1.78240371e+00
6.12382293e-01 1.07145518e-01 3.09643209e-01 -1.37800857e-01
9.89958465e-01 2.16801167e-01 -1.03904414e+00 -5.21199048e-01
-5.20261645e-01 1.12469032e-01 5.82148910e-01 3.78147781e-01
-9.44743991e-01 5.92883408e-01 8.38141251e+00 2.81573892e-01
-8.55632901e-01 1.27043545e-01 3.82050961e-01 -1.36776984e-01
1.54431481e-02 -2.97933012e-01 -1.05664074e+00 5.26117384e-01
1.61025226e+00 -3.77916306e-01 5.14914513e-01 5.63540339e-01
6.37364805e-01 4.28619564e-01 -1.08698726e+00 1.03211987e+00
-2.63824135e-01 -1.18328846e+00 1.13966204e-02 4.56012040e-01
8.41272712e-01 2.73746908e-01 3.47570121e-01 5.83355963e-01
1.03242970e+00 -1.23962200e+00 2.75684804e-01 1.14412892e+00
5.90534031e-01 -5.16111493e-01 6.40350878e-01 4.60289538e-01
-1.18548751e+00 -2.61874869e-02 -6.14620209e-01 -5.47337055e-01
6.71441078e-01 7.31015563e-01 -2.36749515e-01 1.04893625e-01
4.22759712e-01 1.20014989e+00 1.80256963e-01 8.79234910e-01
6.74662599e-03 8.67986023e-01 -5.48441172e-01 2.16342174e-02
5.09997725e-01 -2.90626347e-01 7.89801598e-01 9.62097287e-01
6.95697844e-01 -1.42002702e-01 -1.21600963e-02 1.17540526e+00
1.78891823e-01 -7.76605904e-01 -9.91361380e-01 -3.76587987e-01
3.60401839e-01 7.19560981e-01 -5.42148054e-01 -4.74307537e-01
-3.85337561e-01 1.06765187e+00 1.06798820e-01 1.01467907e+00
-1.05570018e+00 -1.41770735e-01 9.79266584e-01 -1.02646619e-01
4.97155607e-01 -7.57585108e-01 3.70220333e-01 -1.41315866e+00
-2.16236427e-01 -3.42020154e-01 1.57156676e-01 -9.95598793e-01
-1.97438157e+00 7.61311650e-01 8.10218081e-02 -1.31652606e+00
-8.73889744e-01 -6.18813157e-01 -7.68186331e-01 1.13268876e+00
-1.15308499e+00 -5.28988123e-01 2.02103958e-01 7.54823804e-01
5.93696058e-01 5.20481393e-02 9.95452642e-01 -1.17526710e-01
-8.37621987e-01 9.91254952e-03 6.83177590e-01 4.06738400e-01
5.02756909e-02 -1.33671021e+00 9.55301464e-01 5.62234700e-01
-6.67283311e-02 8.66336584e-01 7.80000269e-01 -7.99544036e-01
-1.10935056e+00 -1.43364084e+00 6.55242920e-01 -4.61469322e-01
1.15355492e+00 -4.20705736e-01 -1.34177363e+00 1.30442333e+00
1.18847668e-01 1.06655881e-01 6.62284672e-01 6.42556623e-02
-2.64547884e-01 6.89502537e-01 -3.91284704e-01 7.31438756e-01
1.00344574e+00 -9.06483293e-01 -7.58805752e-01 1.37548342e-01
9.02774930e-01 -2.89533794e-01 -7.99602568e-01 2.42264673e-01
3.75624239e-01 -4.38997060e-01 8.09105754e-01 -7.68987834e-01
3.40424657e-01 7.41465688e-02 -2.10626572e-02 -1.44897354e+00
-4.94602919e-01 -1.10562992e+00 -1.04939592e+00 9.25585628e-01
1.76774964e-01 -1.00417113e+00 3.50761145e-01 4.42630112e-01
1.44245178e-01 -6.22631073e-01 -8.21172416e-01 -1.44419849e+00
1.57556888e-02 -5.63793719e-01 4.99495000e-01 9.34526563e-01
-4.67162699e-01 -4.12216969e-02 -7.91750848e-01 1.55108660e-01
4.86328810e-01 -3.01303208e-01 2.91399539e-01 -1.54892373e+00
-3.25907707e-01 -5.10684192e-01 -2.11083233e-01 -1.45849764e+00
4.24637944e-01 -2.75946915e-01 1.78301528e-01 -1.41593444e+00
-1.67343453e-01 -1.17650069e-01 -3.66823256e-01 1.69281736e-01
1.84862062e-01 -9.90866795e-02 -2.91982979e-01 6.97414160e-01
-5.50716817e-01 8.91099870e-01 8.26214790e-01 3.65066767e-01
-5.01452744e-01 4.58034217e-01 -1.06711552e-01 8.66545975e-01
7.61941254e-01 -6.01241052e-01 -2.57235497e-01 -1.37571460e-02
2.38631785e-01 6.11451864e-01 7.95441926e-01 -8.56656671e-01
5.05421281e-01 -2.51247555e-01 2.70456791e-01 -1.19006371e+00
4.27204937e-01 -4.54828054e-01 3.71367961e-01 3.17356169e-01
-5.75043201e-01 7.97944248e-01 9.34099555e-02 1.02863181e+00
-3.22860539e-01 5.19430861e-02 1.51511729e-01 -6.28098920e-02
-1.19573548e-01 7.38607645e-01 -1.28595877e+00 1.24816177e-02
8.16231608e-01 -2.54252285e-01 -1.33341447e-01 -1.07094026e+00
-1.39300382e+00 9.72337574e-02 5.02154194e-02 2.30149135e-01
4.74691421e-01 -1.41730297e+00 -2.06438988e-01 1.10112943e-01
-2.62770891e-01 1.73184380e-01 2.99677819e-01 9.91020143e-01
-3.30658071e-02 5.45448542e-01 1.42237768e-01 -6.20790839e-01
-5.68594694e-01 6.12666190e-01 5.63228607e-01 -6.40704453e-01
-9.14835930e-01 2.68254101e-01 2.49848798e-01 -4.58885610e-01
4.22916442e-01 -8.07417929e-01 7.89009258e-02 8.90354440e-02
6.83420599e-01 6.05087161e-01 -5.68944633e-01 -3.62305701e-01
1.73201665e-01 1.22313015e-01 -3.60528789e-02 -4.55300897e-01
1.50379217e+00 -2.34001249e-01 -1.57608718e-01 1.83397937e+00
1.18868148e+00 -7.06175148e-01 -1.52506554e+00 -5.15311360e-01
-1.09862350e-01 5.36234975e-01 -4.06785995e-01 -1.86339706e-01
-6.24583125e-01 1.10184300e+00 2.41059661e-02 1.04686487e+00
6.24132752e-01 -7.88611695e-02 9.03322995e-01 3.39696079e-01
-6.81263627e-03 -7.95352995e-01 4.31403629e-02 1.33942485e+00
7.81198025e-01 -5.14989913e-01 -8.24230850e-01 4.73435931e-02
-3.26332480e-01 1.27149415e+00 2.76712894e-01 -5.78450859e-01
1.05204952e+00 6.10554636e-01 -1.47728562e-01 4.16822806e-02
-1.31986237e+00 -1.70946136e-01 -1.29096523e-01 6.71094656e-01
1.58199936e-01 6.74675852e-02 4.82397020e-01 5.79057693e-01
-7.26159588e-02 3.19193453e-01 9.21618640e-01 8.21724772e-01
1.65513173e-01 -6.85137451e-01 -4.23438609e-01 5.62122285e-01
-4.50921178e-01 -1.79150730e-01 1.48142606e-01 7.59668529e-01
-8.11259210e-01 6.75414503e-01 7.94565141e-01 1.03312451e-02
1.43167544e-02 2.39187762e-01 4.85884063e-02 -5.59478641e-01
-2.22273871e-01 1.44913346e-01 -4.06305492e-01 -1.61853194e-01
-3.03562164e-01 -8.10525298e-01 -1.07248449e+00 -5.93846440e-01
-1.38826728e-01 -9.72397402e-02 4.33833659e-01 1.19839907e+00
2.44573802e-01 8.96637082e-01 4.85887259e-01 -7.54334152e-01
-7.45222390e-01 -1.06934226e+00 -7.88056076e-01 -1.15321063e-01
9.09945130e-01 -5.49334943e-01 -8.14670503e-01 -6.96703345e-02] | [6.978858947753906, 3.3885955810546875] |
63f76444-a722-42fd-b7dc-8c085daa7a58 | global-counterfactual-explainer-for-graph | 2210.11695 | null | https://arxiv.org/abs/2210.11695v2 | https://arxiv.org/pdf/2210.11695v2.pdf | Global Counterfactual Explainer for Graph Neural Networks | Graph neural networks (GNNs) find applications in various domains such as computational biology, natural language processing, and computer security. Owing to their popularity, there is an increasing need to explain GNN predictions since GNNs are black-box machine learning models. One way to address this is counterfactual reasoning where the objective is to change the GNN prediction by minimal changes in the input graph. Existing methods for counterfactual explanation of GNNs are limited to instance-specific local reasoning. This approach has two major limitations of not being able to offer global recourse policies and overloading human cognitive ability with too much information. In this work, we study the global explainability of GNNs through global counterfactual reasoning. Specifically, we want to find a small set of representative counterfactual graphs that explains all input graphs. Towards this goal, we propose GCFExplainer, a novel algorithm powered by vertex-reinforced random walks on an edit map of graphs with a greedy summary. Extensive experiments on real graph datasets show that the global explanation from GCFExplainer provides important high-level insights of the model behavior and achieves a 46.9% gain in recourse coverage and a 9.5% reduction in recourse cost compared to the state-of-the-art local counterfactual explainers. | ['Ambuj Singh', 'Sayan Ranu', 'Sourav Medya', 'Zexi Huang', 'Mert Kosan'] | 2022-10-21 | null | null | null | null | ['counterfactual-explanation'] | ['miscellaneous'] | [ 4.13692981e-01 1.10179484e+00 -5.88705719e-01 -1.63601398e-01
-1.46849761e-02 -1.74519807e-01 6.41501844e-01 2.42951527e-01
9.57509503e-02 1.18684435e+00 1.97625056e-01 -9.17474091e-01
-7.26771057e-01 -1.25222909e+00 -1.08074105e+00 -3.77092153e-01
-1.51073679e-01 5.56871057e-01 -1.42084092e-01 -3.93759429e-01
1.70733184e-01 2.95339674e-01 -1.03045273e+00 2.05578282e-02
1.28912449e+00 3.52931947e-01 1.60439406e-02 2.63746798e-01
-1.01431377e-01 6.92703247e-01 -2.82713264e-01 -8.03245425e-01
3.73054534e-01 -7.89851010e-01 -8.56393754e-01 -2.05368429e-01
7.65214860e-02 3.18812989e-02 -5.93986392e-01 1.13438833e+00
2.21036196e-01 4.26896304e-01 5.23143530e-01 -1.62012756e+00
-8.49945009e-01 1.32891893e+00 -6.92098618e-01 2.58082092e-01
8.21513217e-03 1.45229593e-01 1.16812372e+00 -1.78046182e-01
6.39244318e-01 1.37066519e+00 5.98216653e-01 8.17169726e-01
-1.42808616e+00 -6.55041575e-01 4.82199818e-01 3.25501055e-01
-8.59849155e-01 1.31324768e-01 1.03873777e+00 1.66601893e-02
1.10356605e+00 6.04062498e-01 7.32410729e-01 9.47023213e-01
6.05462372e-01 3.74752343e-01 8.80443990e-01 -4.96611238e-01
3.85675222e-01 -2.92018205e-01 1.16557954e-02 7.59799600e-01
8.84620011e-01 4.27143157e-01 -5.13068438e-01 -2.27970123e-01
6.72773838e-01 2.72797137e-01 -4.64129537e-01 -6.91092551e-01
-9.29128587e-01 1.17788351e+00 9.87003624e-01 1.29067581e-02
-6.11043990e-01 6.67301238e-01 2.85110921e-01 2.33044460e-01
6.36403561e-01 7.58577466e-01 -4.92342979e-01 4.67746347e-01
-4.13408637e-01 2.65733331e-01 7.38214552e-01 5.19667387e-01
6.37386978e-01 2.31635854e-01 -5.70296012e-02 4.00198936e-01
2.48572111e-01 3.37072194e-01 4.56026733e-01 -6.43260300e-01
5.96168876e-01 8.20420086e-01 -2.68855810e-01 -1.39369953e+00
-4.97154921e-01 -8.00700963e-01 -1.12844491e+00 2.74043549e-02
2.32485950e-01 -2.50796616e-01 -8.30987871e-01 1.96251714e+00
3.66313875e-01 2.77637094e-01 -9.51144025e-02 9.08726931e-01
3.97739708e-01 4.32138175e-01 1.29356295e-01 -3.22805405e-01
9.24608111e-01 -1.03242695e+00 -6.02689445e-01 -5.85899711e-01
6.84972167e-01 2.55883727e-02 7.88986981e-01 -8.35971385e-02
-5.93552947e-01 -3.52207571e-02 -8.56411815e-01 4.56047297e-01
-2.89445728e-01 -5.06518960e-01 1.07208633e+00 6.90554142e-01
-9.85148072e-01 9.08459365e-01 -4.95272279e-01 -3.25444043e-01
5.29682577e-01 4.53679830e-01 -7.96087012e-02 -1.63159847e-01
-1.49626768e+00 8.64209652e-01 7.09820986e-01 8.62168819e-02
-6.78930998e-01 -8.40036333e-01 -8.83607805e-01 4.67720419e-01
1.08757269e+00 -1.12472320e+00 9.13989604e-01 -9.63656962e-01
-1.10076463e+00 3.16731542e-01 2.12570608e-01 -1.13920736e+00
7.33074725e-01 3.37736666e-01 -3.10085624e-01 -3.77778351e-01
1.72983035e-01 4.33019489e-01 5.54629564e-01 -1.03285432e+00
-3.43532234e-01 -3.95380855e-01 2.82057017e-01 2.71788955e-01
-6.75189719e-02 -6.81196511e-01 1.01497091e-01 -6.58258617e-01
1.45557269e-01 -1.05057323e+00 -7.61724114e-01 -3.89039248e-01
-9.25778627e-01 -2.01453567e-01 4.65218365e-01 -4.37800169e-01
1.05700469e+00 -1.39973569e+00 -1.06693193e-01 4.09859091e-01
6.09229445e-01 -1.38383862e-02 -1.59151047e-01 3.45832497e-01
-3.71418238e-01 5.58356702e-01 -3.02204221e-01 2.38806367e-01
1.27029985e-01 9.41961333e-02 -4.80371892e-01 3.01507771e-01
-3.50676507e-01 1.27557313e+00 -1.12062693e+00 -8.61562230e-03
1.14917748e-01 -3.27351056e-02 -6.56916261e-01 -2.92407066e-01
-4.87206876e-01 -1.20937690e-01 -4.69301194e-01 1.00762524e-01
4.92337644e-01 -4.28730816e-01 7.75166750e-01 3.48113090e-01
4.12361085e-01 4.15603966e-01 -9.23821151e-01 1.14520478e+00
-3.32232118e-01 6.38479054e-01 -6.73139393e-01 -1.22136152e+00
8.21881473e-01 7.86627159e-02 1.62357002e-01 -5.89795411e-01
9.02709514e-02 2.23106474e-01 3.36439908e-01 -7.08922930e-03
1.92363665e-01 -3.14005375e-01 -7.33630359e-02 8.82966638e-01
-4.15723592e-01 9.01006684e-02 -2.73117162e-02 3.53444725e-01
1.15525806e+00 -2.15713397e-01 9.94712055e-01 -4.20726568e-01
1.21685281e-01 1.22435741e-01 5.34263670e-01 1.12693357e+00
-1.46231735e-02 1.76484212e-01 7.47455537e-01 -7.81028807e-01
-8.73606145e-01 -8.66995156e-01 5.54264009e-01 7.89876997e-01
1.48057625e-01 -8.41030031e-02 -7.75513351e-01 -1.16398025e+00
1.66685343e-01 1.23719323e+00 -8.62729967e-01 -7.75200188e-01
-4.36806262e-01 -8.26350331e-01 2.27265716e-01 3.61647189e-01
6.85426712e-01 -1.18541765e+00 -5.62969387e-01 1.48188800e-01
-2.52367228e-01 -5.35419106e-01 -5.92073977e-01 -5.50106168e-02
-1.06230283e+00 -1.32526886e+00 -2.54756093e-01 -1.50291532e-01
8.37064743e-01 4.62475210e-01 1.01349068e+00 3.44932854e-01
-4.43430655e-02 7.62933195e-02 6.22549430e-02 -7.39397287e-01
-4.25604880e-01 1.40519381e-01 2.04713598e-01 -2.50392795e-01
2.10100546e-01 -6.41972244e-01 -5.89986801e-01 7.73591101e-02
-7.45207489e-01 4.59104747e-01 5.92488050e-01 9.55015719e-01
4.21306223e-01 1.89073369e-01 9.35188413e-01 -1.45714736e+00
1.01305246e+00 -6.65472627e-01 -4.97585863e-01 4.11850601e-01
-1.42487717e+00 3.51863056e-01 8.78806949e-01 -3.59969586e-01
-1.18136799e+00 -2.40461096e-01 4.25795555e-01 -7.98237622e-02
1.85547829e-01 9.25335765e-01 -1.05384439e-01 1.39545783e-01
8.97100151e-01 1.92445874e-01 -5.52320257e-02 -2.53213421e-02
6.90406859e-01 5.71199358e-02 2.82130748e-01 -2.17724219e-01
8.07491124e-01 3.26147825e-01 4.71780360e-01 -3.38379711e-01
-7.94721246e-01 2.17191219e-01 -1.78239390e-01 -3.16030473e-01
3.90168697e-01 -2.78173298e-01 -9.81139898e-01 -1.60172567e-01
-1.11849642e+00 -3.37265968e-01 -5.00039995e-01 2.96684712e-01
-7.38750756e-01 2.33800456e-01 2.16821581e-02 -7.76518464e-01
-4.36656833e-01 -5.21924019e-01 1.73185036e-01 2.23474756e-01
-2.25198209e-01 -1.19130468e+00 -9.16452706e-03 3.78873110e-01
3.51223677e-01 5.60710371e-01 1.18233263e+00 -9.46008384e-01
-8.23531449e-01 1.49958173e-03 -2.33951479e-01 -3.96825194e-01
2.03571707e-01 -5.13048112e-01 -4.89131749e-01 -1.23306513e-01
-2.13620976e-01 2.83445716e-01 9.90456402e-01 8.03133190e-01
1.12958586e+00 -1.03464246e+00 -7.66780019e-01 3.03141624e-01
1.41489017e+00 3.08578789e-01 3.97841424e-01 4.58762228e-01
6.52449310e-01 6.56748652e-01 4.31795657e-01 2.60639638e-01
3.58695865e-01 4.92564857e-01 9.43187177e-01 3.81766856e-02
7.87063912e-02 -7.11394191e-01 -6.44102395e-02 6.21110722e-02
-3.00108433e-01 -5.67413926e-01 -1.05310750e+00 6.67825401e-01
-2.44348502e+00 -1.11537027e+00 -3.14912647e-01 2.07998443e+00
3.81425321e-01 2.24783331e-01 -1.12598203e-01 -1.69055574e-02
1.07048404e+00 2.49631166e-01 -9.16719913e-01 -6.08963788e-01
1.14358246e-01 -3.41934532e-01 7.17702985e-01 4.79105890e-01
-6.35130703e-01 7.83713281e-01 5.30232954e+00 7.54693627e-01
-7.49560058e-01 -1.17545426e-02 9.05004680e-01 -1.67838275e-01
-8.95906270e-01 2.22275123e-01 -2.19522253e-01 3.23918879e-01
1.03686666e+00 -9.73666728e-01 8.16615403e-01 8.50429475e-01
4.12765801e-01 2.92341471e-01 -9.89579856e-01 5.19727170e-01
-1.29839599e-01 -1.88547039e+00 4.74769920e-01 2.45002806e-01
1.06983125e+00 -1.35456830e-01 -5.46673648e-02 2.98790157e-01
7.25693345e-01 -1.13558090e+00 4.77514058e-01 5.62016010e-01
5.33640563e-01 -1.11572111e+00 7.62941718e-01 5.97410381e-01
-8.09021473e-01 -2.71434754e-01 -5.52071512e-01 -3.89814734e-01
3.20564816e-03 7.23159254e-01 -1.22943246e+00 7.67544925e-01
2.45474860e-01 2.29050949e-01 -1.61197722e-01 8.33430171e-01
-5.85722268e-01 6.45226121e-01 -3.42896804e-02 -5.09805799e-01
1.79086983e-01 -5.31066842e-02 6.71203494e-01 6.79886103e-01
2.41986349e-01 2.62544572e-01 -1.84982806e-01 1.13417244e+00
-4.87810493e-01 3.02412044e-02 -1.12050378e+00 2.59424895e-02
4.28050011e-01 7.52142370e-01 -8.97659302e-01 -9.76861715e-02
3.36842202e-02 7.45991886e-01 4.85097170e-01 3.93967122e-01
-1.03036916e+00 -2.57999927e-01 4.74475920e-01 2.46503517e-01
-3.83459069e-02 3.96885335e-01 -5.35043895e-01 -9.46175277e-01
-2.47834533e-01 -7.91838586e-01 6.88918531e-01 -6.81108057e-01
-1.09435821e+00 4.39345747e-01 -2.42823269e-02 -6.15897954e-01
-5.23499787e-01 -2.50927418e-01 -8.97458673e-01 6.45036161e-01
-1.21848869e+00 -1.08166873e+00 -1.43787637e-01 2.42609397e-01
5.31974554e-01 -8.17425251e-02 4.89722967e-01 -4.52605218e-01
-4.07140374e-01 4.33069885e-01 1.10293321e-01 -2.50572503e-01
3.43335979e-02 -1.27614880e+00 7.30718076e-01 9.93965626e-01
4.14703101e-01 6.77162945e-01 1.11060047e+00 -9.81874704e-01
-9.77321088e-01 -1.25718427e+00 1.12620819e+00 -7.88921192e-02
6.21786177e-01 -2.27839258e-02 -7.03123927e-01 9.29194212e-01
1.23835139e-01 -8.90189558e-02 1.58929601e-01 3.80262941e-01
-5.16380854e-02 -9.14205685e-02 -1.19770849e+00 1.13626540e+00
1.48782480e+00 -1.52356595e-01 -4.65124130e-01 4.53913271e-01
1.07446432e+00 -2.13926986e-01 -7.79488012e-02 2.75441915e-01
2.06613630e-01 -8.96113217e-01 8.79036605e-01 -1.11163628e+00
6.62934840e-01 4.66865823e-02 3.05829406e-01 -1.86618614e+00
-3.70210409e-01 -8.56106460e-01 -2.10464969e-01 7.52617002e-01
6.57009304e-01 -1.25392187e+00 1.11911976e+00 9.11154807e-01
1.36772722e-01 -9.42290843e-01 -9.86479938e-01 -8.40328991e-01
-1.83832139e-01 -3.94869983e-01 1.09295940e+00 1.13522589e+00
2.60376304e-01 3.83527130e-01 -4.87533897e-01 2.64659207e-02
8.80840719e-01 4.76862758e-01 7.01483428e-01 -1.24542272e+00
-2.15958580e-01 -5.03607094e-01 -2.73683876e-01 -5.79861999e-01
4.61781234e-01 -1.16221344e+00 -2.29868099e-01 -1.93873823e+00
5.58332205e-01 -4.40361723e-03 -2.88770407e-01 6.32204890e-01
-2.48608321e-01 -2.05781281e-01 1.56283051e-01 -1.17032439e-01
-3.21868390e-01 6.63942814e-01 1.08761358e+00 -3.64345670e-01
-2.21040949e-01 2.58184075e-01 -1.17995691e+00 7.63615608e-01
1.08347642e+00 -8.24205697e-01 -8.35934877e-01 -3.46567295e-02
4.38044608e-01 2.37104744e-01 6.09692216e-01 -3.84247243e-01
2.63564318e-01 -6.51750922e-01 2.34386977e-02 -1.32414177e-01
-2.86682844e-01 -6.91705287e-01 5.27348518e-01 1.04048395e+00
-5.56408584e-01 4.47000470e-03 4.53152694e-02 1.25268686e+00
2.56483734e-01 -1.64557695e-01 4.22727853e-01 -2.00238198e-01
-5.59372127e-01 3.12765449e-01 -2.49249190e-01 -1.43745944e-01
1.14341366e+00 -1.65921882e-01 -7.50947893e-01 -8.15790892e-01
-5.36857247e-01 2.49309108e-01 1.18976101e-01 4.29305136e-01
6.53450072e-01 -1.25078273e+00 -6.18986905e-01 -2.31174245e-01
-6.25149384e-02 -2.50495225e-01 4.16700035e-01 5.61380386e-01
-2.37387538e-01 8.87924314e-01 -2.46372446e-01 1.02322288e-01
-9.46317077e-01 7.47985065e-01 6.05383456e-01 -7.84009218e-01
-6.48089707e-01 5.70729375e-01 3.84036630e-01 -8.02891791e-01
-4.63812977e-01 -1.17196299e-01 -1.18889146e-01 -4.55502033e-01
-4.84847687e-02 6.12724483e-01 -1.90182537e-01 9.76000167e-03
-2.10074708e-01 -2.95535356e-01 7.88713843e-02 2.33884096e-01
1.55861199e+00 -1.60976157e-01 -1.13756090e-01 -2.18102033e-03
6.04700029e-01 -1.84295967e-01 -1.00617850e+00 -6.28202111e-02
2.52368033e-01 -4.96812493e-01 -9.82564986e-02 -9.11918700e-01
-1.03778493e+00 4.21554953e-01 1.62456278e-02 7.25951076e-01
1.00367904e+00 4.66848025e-03 4.15614486e-01 5.83662570e-01
4.80686367e-01 -8.38596523e-01 -2.45866999e-01 1.31298497e-01
1.09022903e+00 -1.08112466e+00 2.00083405e-01 -4.74671841e-01
-4.66165066e-01 7.80455947e-01 6.99057817e-01 -9.38789472e-02
3.93538326e-01 -4.58732933e-01 -4.14150923e-01 -4.34374750e-01
-9.22765255e-01 1.47633344e-01 4.27052349e-01 5.97567677e-01
9.62122157e-02 4.92970645e-01 -6.17323935e-01 8.30581307e-01
-3.89315993e-01 -2.49231502e-01 9.13494408e-01 1.07954152e-01
-2.79179543e-01 -6.43796206e-01 -1.41314998e-01 9.47712123e-01
-2.47174844e-01 -1.94500133e-01 -5.45812011e-01 1.11251855e+00
-3.12685996e-01 8.83701086e-01 -2.51437008e-01 -2.04371214e-01
2.33841002e-01 -1.30695716e-01 1.08123854e-01 -4.91237521e-01
-3.00946653e-01 -5.74734449e-01 2.24519923e-01 -6.53228521e-01
-5.16594648e-02 -4.42921013e-01 -1.34092033e+00 -7.61406660e-01
-5.18397450e-01 2.45635733e-01 4.60926443e-01 1.12210131e+00
4.95238453e-01 9.19491470e-01 4.06918287e-01 -3.53128761e-01
-7.44141102e-01 -7.22370565e-01 -5.45471907e-01 1.82819933e-01
6.34319186e-02 -5.65547287e-01 -4.44921762e-01 -4.81054515e-01] | [8.430273056030273, 5.8806610107421875] |
7d11aad1-ff6e-4ad8-b3bd-7a1ee32ed74b | native-language-identification-on-text-and | 1707.07182 | null | http://arxiv.org/abs/1707.07182v1 | http://arxiv.org/pdf/1707.07182v1.pdf | Native Language Identification on Text and Speech | This paper presents an ensemble system combining the output of multiple SVM
classifiers to native language identification (NLI). The system was submitted
to the NLI Shared Task 2017 fusion track which featured students essays and
spoken responses in form of audio transcriptions and iVectors by non-native
English speakers of eleven native languages. Our system competed in the
challenge under the team name ZCD and was based on an ensemble of SVM
classifiers trained on character n-grams achieving 83.58% accuracy and ranking
3rd in the shared task. | ['Marcos Zampieri', 'Alina Maria Ciobanu', 'Liviu P. Dinu'] | 2017-07-22 | native-language-identification-on-text-and-1 | https://aclanthology.org/W17-5045 | https://aclanthology.org/W17-5045.pdf | ws-2017-9 | ['native-language-identification'] | ['natural-language-processing'] | [ 1.88530520e-01 -1.16048142e-01 -4.98284727e-01 -4.98006642e-01
-1.15565217e+00 -1.21355891e+00 6.44527316e-01 3.12242270e-01
-6.19688570e-01 7.16497660e-01 2.31038392e-01 -6.59077048e-01
9.50135663e-02 -1.88993052e-01 -3.64654392e-01 -4.68827076e-02
6.05361342e-01 5.74133575e-01 -2.24862099e-01 -1.25984743e-01
2.54948646e-01 2.38218337e-01 -1.23953938e+00 7.41464138e-01
1.01003456e+00 9.36039925e-01 -3.81576955e-01 1.22686374e+00
-2.84988821e-01 1.07032084e+00 -1.00869787e+00 -4.37424928e-01
-1.61907867e-01 -1.26656711e-01 -1.06265175e+00 -8.73931587e-01
1.20869672e+00 2.67937966e-02 -4.50390697e-01 9.23114955e-01
5.64858437e-01 3.43161762e-01 7.55363166e-01 -1.20133948e+00
-8.46215665e-01 1.07938910e+00 1.63345546e-01 4.76269752e-01
1.07699966e+00 -2.39340037e-01 1.25721133e+00 -1.27401459e+00
5.51876783e-01 1.17750883e+00 7.51938641e-01 5.45187235e-01
-1.39524317e+00 -1.18060946e+00 -1.95461243e-01 3.54790509e-01
-9.88413155e-01 -6.96945310e-01 3.30981106e-01 -8.52347434e-01
1.35122836e+00 4.21589792e-01 7.77885467e-02 1.72845066e+00
5.78669608e-02 1.41977406e+00 1.63126409e+00 -6.67880476e-01
-6.80673867e-02 8.50818932e-01 1.07209301e+00 4.18776661e-01
-2.94274777e-01 -1.00217782e-01 -1.23343837e+00 -3.31900775e-01
-1.95011288e-01 -6.44544125e-01 -2.08064809e-01 3.49666506e-01
-1.41109121e+00 1.02947628e+00 -2.78202623e-01 2.64690429e-01
2.35668257e-01 -4.47843492e-01 5.21034300e-01 7.33861387e-01
5.14857590e-01 7.01135933e-01 -8.04904759e-01 -4.00310218e-01
-8.82333398e-01 1.09512076e-01 1.12915301e+00 5.94590545e-01
1.16750389e-01 2.24350706e-01 -1.28869802e-01 1.07605660e+00
1.64246917e-01 5.71529031e-01 1.00380766e+00 -5.41554391e-01
6.13851666e-01 6.03807271e-01 -2.62195349e-01 -3.34416002e-01
-1.84239537e-01 -3.08799714e-01 -4.53305840e-01 2.80743986e-01
6.03751302e-01 -3.38845789e-01 -5.34193575e-01 1.24459279e+00
-2.31652409e-01 3.41290206e-01 4.85376298e-01 5.01200080e-01
1.74584103e+00 6.86784506e-01 2.87340194e-01 3.44807863e-01
9.76159394e-01 -1.14683926e+00 -7.06727743e-01 -1.16692849e-01
6.36023641e-01 -1.04195762e+00 5.70542932e-01 9.40457702e-01
-1.02916014e+00 -1.04138076e+00 -1.13999355e+00 9.35884118e-02
-5.87894320e-01 4.75805759e-01 1.09258533e-01 8.76668036e-01
-9.76251543e-01 3.27420324e-01 -1.01995483e-01 -1.66525096e-01
9.30989683e-02 4.81260091e-01 -6.39233947e-01 3.14482361e-01
-1.29142761e+00 1.11145365e+00 7.43486034e-03 -5.64431429e-01
-4.40943778e-01 -1.16553402e+00 -8.58755589e-01 -9.79911163e-02
-4.48813915e-01 1.70967743e-01 1.42391360e+00 -9.46192801e-01
-2.02606416e+00 1.33089697e+00 -2.15275027e-02 -4.96093661e-01
4.82126474e-01 -3.27844858e-01 -5.75438440e-01 -2.35502526e-01
5.05758487e-02 2.46553883e-01 1.63154572e-01 -3.11203927e-01
-6.60711586e-01 -2.45758235e-01 -5.09647846e-01 2.01465428e-01
-6.59474194e-01 6.61377490e-01 6.39833212e-01 -3.90142471e-01
3.13200206e-02 -7.30576038e-01 3.84362817e-01 -1.01329005e+00
-3.59847605e-01 -1.01857829e+00 7.68374383e-01 -1.02786827e+00
1.08546937e+00 -1.86529481e+00 4.03937697e-01 -1.49640320e-02
-5.40656783e-02 5.17117918e-01 -2.22494081e-02 2.58810699e-01
-4.36465591e-01 9.22712162e-02 5.36130965e-01 -5.40440679e-01
1.54270977e-01 -4.76349324e-01 -9.47528005e-01 2.93694168e-01
-3.12849618e-02 7.34323621e-01 -9.10752594e-01 -2.57662773e-01
1.25905037e-01 1.13504723e-01 1.50933653e-01 6.73184752e-01
4.25218135e-01 4.48111355e-01 1.54579863e-01 7.24794149e-01
9.13559049e-02 2.61409611e-01 -1.02140203e-01 2.66819209e-01
-3.26071084e-01 9.77613568e-01 -1.10683799e+00 1.61237240e+00
-4.37498599e-01 9.82849658e-01 3.14154476e-01 -1.39178073e+00
1.42055607e+00 7.11158395e-01 3.20952572e-02 -1.13315962e-01
1.25208008e-03 5.91574669e-01 2.35902481e-02 -3.16148661e-02
2.21670702e-01 4.03327465e-01 -6.70314729e-01 6.54389799e-01
9.90270376e-01 -2.04803601e-01 -3.28919470e-01 1.77510008e-01
9.95563567e-01 1.99427381e-01 5.98114610e-01 -4.60913122e-01
1.13108313e+00 9.06160548e-02 1.56652242e-01 1.17534041e+00
-7.09607780e-01 3.08161080e-01 2.14841459e-02 -5.52834988e-01
-6.01209521e-01 -1.02192783e+00 -3.63713562e-01 1.84350240e+00
-5.82015812e-01 -2.77306288e-01 -4.76228565e-01 -8.32656443e-01
6.07179627e-02 1.00717938e+00 -3.90273750e-01 -2.49556918e-02
-4.68334377e-01 -2.27950774e-02 1.09011602e+00 4.78158057e-01
-2.86717452e-02 -1.19048929e+00 -2.14953553e-02 3.10703963e-01
-1.06611043e-01 -1.12732887e+00 -6.02314770e-01 7.37406075e-01
-7.26673156e-02 -5.16931713e-01 -7.62436211e-01 -1.10324633e+00
-1.55397076e-02 -1.61374316e-01 8.23929787e-01 -3.89104366e-01
-2.99918711e-01 7.40120471e-01 -2.44655564e-01 -8.98311675e-01
-6.56940877e-01 6.90479338e-01 6.64096534e-01 -1.39159203e-01
9.06606495e-01 -1.63505048e-01 4.46622193e-01 -2.02516299e-02
1.50225997e-01 -8.65284503e-02 -5.68878651e-03 9.95980024e-01
-1.61694273e-01 -8.08641195e-01 9.00161028e-01 -4.79893833e-01
7.12224424e-01 -3.48190099e-01 -4.51449156e-01 7.25337386e-01
-3.54256600e-01 -2.33591214e-01 8.01124990e-01 -8.24052274e-01
-9.36402142e-01 1.20404869e-01 -1.46677822e-01 -1.81987971e-01
-5.89271426e-01 4.31517482e-01 -9.59564969e-02 -4.24663901e-01
6.38775289e-01 3.42529029e-01 -2.46767700e-01 -4.90893632e-01
2.17995886e-02 1.46511531e+00 9.16549504e-01 -7.19140649e-01
4.28295970e-01 -7.44555950e-01 -5.55547833e-01 -8.63223314e-01
-1.16021967e+00 -6.30511224e-01 -8.93504202e-01 -3.42156291e-01
7.69627392e-01 -1.28555691e+00 -1.06132579e+00 9.77940798e-01
-1.33680916e+00 -8.68089870e-02 2.13066235e-01 7.16646254e-01
-1.74644724e-01 -1.94694415e-01 -7.84897447e-01 -1.00741851e+00
-5.92691481e-01 -1.33835948e+00 5.35409212e-01 5.56115508e-01
-1.03242755e+00 -1.01910555e+00 4.52598363e-01 1.01396763e+00
6.01185381e-01 -4.78825867e-01 4.42741036e-01 -1.94852817e+00
1.55801058e-01 -4.75104928e-01 1.73803657e-01 4.37095910e-01
-4.16408569e-01 1.40873075e-01 -1.50397730e+00 -1.58699498e-01
-3.26504737e-01 -1.11800086e+00 7.84751415e-01 -1.12833977e-01
8.89795184e-01 -2.68521279e-01 -9.82201248e-02 5.50130963e-01
8.12580585e-01 1.63766459e-01 -2.56014585e-01 5.02815425e-01
7.23529994e-01 7.72692323e-01 -6.02116510e-02 -1.35972708e-01
2.62056112e-01 9.61429656e-01 -3.71164560e-01 5.61273038e-01
5.86396689e-03 -2.65382618e-01 7.75645196e-01 1.17601621e+00
4.25822973e-01 -1.28595442e-01 -1.53679228e+00 6.11722052e-01
-1.39428318e+00 -9.37179029e-01 -4.33337182e-01 1.94971025e+00
1.19650185e+00 5.21937152e-03 2.04437047e-01 3.54343317e-02
3.54641110e-01 -1.20970227e-01 -1.16338268e-01 -1.26123691e+00
-2.73047060e-01 7.72304118e-01 5.02424121e-01 1.02599776e+00
-1.51276052e+00 8.50903988e-01 6.94209099e+00 8.75861406e-01
-1.34378362e+00 2.97989964e-01 3.85989636e-01 1.03069004e-02
-5.15704714e-02 -4.54642653e-01 -1.75427997e+00 3.88187647e-01
1.77802396e+00 -7.29816496e-01 1.24660403e-01 9.28069115e-01
-8.05471182e-01 1.03256419e-01 -9.43008840e-01 6.07229054e-01
3.52895826e-01 -1.33953977e+00 -4.55973119e-01 -2.10232690e-01
9.44711506e-01 2.96728134e-01 4.27991420e-01 7.46376634e-01
8.00358593e-01 -1.64583361e+00 7.60843992e-01 5.00086546e-01
8.04065228e-01 -5.72562695e-01 7.95753062e-01 7.14633584e-01
-6.34336412e-01 -8.61524567e-02 1.69825971e-01 -3.19786549e-01
-5.42352974e-01 -1.14368506e-01 -1.22697091e+00 3.32766265e-01
6.78777099e-01 8.11718464e-01 -4.71730471e-01 5.39777577e-01
-2.08676293e-01 1.38659859e+00 -3.15329939e-01 -5.14582098e-01
1.22136801e-01 8.32265839e-02 5.63225925e-01 1.74351668e+00
1.45405725e-01 1.21286064e-01 6.12160265e-01 5.32679141e-01
1.79472249e-02 3.55324090e-01 -5.91395140e-01 -5.15907891e-02
6.33712649e-01 1.33063054e+00 -4.75474186e-02 -6.26396298e-01
-5.73611677e-01 1.03473127e+00 4.30400163e-01 1.05345502e-01
-3.63302648e-01 -6.54183686e-01 7.63748705e-01 -5.06788552e-01
-1.69426665e-01 4.96337749e-02 -9.58545685e-01 -1.20898104e+00
-6.22964621e-01 -1.18460190e+00 4.08793747e-01 -5.61049640e-01
-1.50227940e+00 8.30745220e-01 -3.00077945e-01 -1.00438631e+00
-4.05841619e-01 -1.03696251e+00 -9.38459277e-01 1.60672879e+00
-8.47175539e-01 -9.49639499e-01 -3.98741215e-02 2.10567549e-01
6.67952657e-01 -1.22170103e+00 1.47277939e+00 3.39376628e-02
-8.99922192e-01 1.32828426e+00 3.41972649e-01 3.92175585e-01
1.28935134e+00 -1.54473090e+00 4.44973648e-01 3.56647730e-01
6.47162259e-01 5.84706366e-01 5.48422158e-01 -4.00452256e-01
-9.82572913e-01 -7.16174304e-01 1.95192075e+00 -1.08130229e+00
9.04358447e-01 -4.96634960e-01 -9.46425378e-01 8.12484741e-01
7.00704813e-01 -3.32440555e-01 1.40217841e+00 6.78432733e-02
-5.27993977e-01 2.90995151e-01 -1.01400685e+00 -1.57818437e-01
7.53173828e-02 -9.55734789e-01 -1.04169214e+00 5.09879053e-01
5.56053638e-01 -7.53795981e-01 -1.08382237e+00 1.54720873e-01
7.06239402e-01 -4.84394640e-01 1.10691273e+00 -1.14952362e+00
6.99437797e-01 3.43504578e-01 -1.82077408e-01 -1.36294031e+00
1.65092908e-02 -2.18954131e-01 -2.12086178e-02 1.47142017e+00
5.08201301e-01 -4.57576901e-01 3.78626108e-01 5.32650471e-01
-6.95983246e-02 -5.82164407e-01 -1.11982620e+00 -7.90982187e-01
4.64463204e-01 -4.81466949e-01 2.10153237e-02 1.38043737e+00
6.30284548e-01 7.41645694e-01 -1.45564079e-01 -3.24726045e-01
6.77949309e-01 -7.27034956e-02 6.48983896e-01 -1.68051088e+00
-9.09288079e-02 -9.85525906e-01 -3.34980220e-01 -6.58202410e-01
1.04998517e+00 -1.68830776e+00 -2.78881639e-01 -7.91795373e-01
3.58955422e-03 1.03968240e-01 -3.64304781e-01 8.81447375e-01
-3.42864722e-01 3.54963064e-01 2.03066155e-01 1.29290044e-01
-5.11260808e-01 1.00581154e-01 1.52208462e-01 -5.18023312e-01
-1.40948787e-01 2.42296010e-01 -6.72764599e-01 6.77320838e-01
5.42761505e-01 -1.79390654e-01 1.87872931e-01 -5.06836921e-02
-2.42012680e-01 8.57698172e-03 9.89035517e-02 -1.08645439e+00
6.24274254e-01 -1.42563626e-01 5.84912539e-01 -6.27585888e-01
1.65853098e-01 -8.47819224e-02 -3.19490910e-01 2.85659373e-01
-1.07486200e+00 -2.34321147e-01 5.37668288e-01 -3.08735311e-01
-4.39530283e-01 -6.04123890e-01 6.76713467e-01 1.67852104e-01
-5.20650685e-01 -2.42220029e-01 -7.32448757e-01 7.47772753e-02
8.91240180e-01 -1.13238312e-01 -3.08638126e-01 -3.69542688e-01
-7.21085131e-01 1.66231543e-01 -7.30965361e-02 8.47474515e-01
3.44335318e-01 -1.25941014e+00 -1.28491104e+00 6.44119263e-01
-1.79187817e-04 -8.83781970e-01 -5.36203608e-02 6.34553790e-01
-2.12384790e-01 1.20313466e+00 -2.55336046e-01 -4.67544347e-01
-1.85307229e+00 -1.30937368e-01 9.82725993e-02 -4.27139342e-01
-3.94389592e-02 1.75484729e+00 -4.33273673e-01 -1.31429172e+00
6.55183554e-01 3.12289715e-01 -6.40901744e-01 4.52066928e-01
1.09322464e+00 2.93084532e-01 2.49318272e-01 -8.41454089e-01
-5.84854245e-01 1.63928568e-01 -2.14622676e-01 -3.47653210e-01
1.21232212e+00 4.22466308e-01 1.73850015e-01 9.71651375e-01
1.39365542e+00 1.93340540e-01 -8.49578604e-02 -4.80452359e-01
1.30332246e-01 1.10182837e-01 1.11139022e-01 -1.30726314e+00
-1.92077622e-01 1.06053305e+00 3.77171695e-01 -1.46664321e-01
1.44968972e-01 -2.94844568e-01 6.46403790e-01 4.96623755e-01
2.78237313e-02 -9.53354657e-01 -1.45348474e-01 1.01248336e+00
6.45268202e-01 -1.36381197e+00 -2.13544756e-01 6.36848882e-02
-6.79224730e-01 1.59977949e+00 9.64885116e-01 -1.30245939e-01
4.75642264e-01 3.25239420e-01 4.72159445e-01 5.94127059e-01
-1.00649381e+00 4.17115241e-01 1.10658741e+00 4.25779700e-01
1.10965192e+00 3.73921216e-01 -1.81035399e-01 1.37781870e+00
-5.16674101e-01 -2.95487344e-01 5.28111875e-01 3.54101330e-01
-3.56732190e-01 -1.05229509e+00 -7.07667828e-01 3.52527261e-01
-5.89084625e-01 -1.83639616e-01 -8.23691547e-01 1.60068810e-01
-1.45814270e-02 9.98873055e-01 -1.18329115e-01 -6.59991860e-01
1.37931958e-01 1.22890985e+00 2.70694286e-01 -7.70013094e-01
-1.38606989e+00 -7.12709486e-01 5.12386635e-02 -2.93623567e-01
1.80002764e-01 -1.11805761e+00 -6.32229447e-01 1.79839768e-02
-9.79480073e-02 2.52442449e-01 7.93865681e-01 1.10266149e+00
-2.35343367e-01 4.30424720e-01 4.52185690e-01 -5.24178088e-01
-1.31948757e+00 -1.26079583e+00 -2.65924543e-01 8.18243697e-02
3.97323608e-01 -3.23322862e-01 -8.24523032e-01 -1.36951432e-01] | [10.41236400604248, 10.512048721313477] |
e9a4d7b0-1e07-400c-b4ad-ea7f76cfa498 | on-the-surprising-effectiveness-of | 2209.07474 | null | https://arxiv.org/abs/2209.07474v3 | https://arxiv.org/pdf/2209.07474v3.pdf | On the Surprising Effectiveness of Transformers in Low-Labeled Video Recognition | Recently vision transformers have been shown to be competitive with convolution-based methods (CNNs) broadly across multiple vision tasks. The less restrictive inductive bias of transformers endows greater representational capacity in comparison with CNNs. However, in the image classification setting this flexibility comes with a trade-off with respect to sample efficiency, where transformers require ImageNet-scale training. This notion has carried over to video where transformers have not yet been explored for video classification in the low-labeled or semi-supervised settings. Our work empirically explores the low data regime for video classification and discovers that, surprisingly, transformers perform extremely well in the low-labeled video setting compared to CNNs. We specifically evaluate video vision transformers across two contrasting video datasets (Kinetics-400 and SomethingSomething-V2) and perform thorough analysis and ablation studies to explain this observation using the predominant features of video transformer architectures. We even show that using just the labeled data, transformers significantly outperform complex semi-supervised CNN methods that leverage large-scale unlabeled data as well. Our experiments inform our recommendation that semi-supervised learning video work should consider the use of video transformers in the future. | ['Zsolt Kira', 'Ömer Mubarek', 'Farrukh Rahman'] | 2022-09-15 | null | null | null | null | ['video-classification', 'video-recognition'] | ['computer-vision', 'computer-vision'] | [ 2.72571236e-01 1.21086404e-01 -2.56292433e-01 -3.42388988e-01
-5.18626511e-01 -7.57041216e-01 9.06679749e-01 -5.16422391e-01
-6.67690814e-01 4.33044165e-01 3.82832557e-01 -5.45414388e-01
-1.54115455e-02 -5.57475507e-01 -8.17688048e-01 -5.97898543e-01
-2.71623749e-02 1.93161577e-01 2.41623923e-01 -2.97690406e-02
-6.67997152e-02 1.08580723e-01 -1.61903524e+00 5.16682744e-01
2.20089152e-01 1.19616306e+00 4.12147529e-02 6.84591830e-01
1.18637830e-01 1.48502564e+00 -1.24664560e-01 -5.77454805e-01
5.64314067e-01 -1.48760110e-01 -9.38272476e-01 2.34075457e-01
1.17026651e+00 -5.48852503e-01 -7.35025764e-01 8.51886868e-01
3.04701775e-01 -1.00396827e-01 6.24173760e-01 -1.41212559e+00
-7.44520545e-01 5.95408738e-01 -2.93540359e-01 6.35974348e-01
5.64327724e-02 5.72622359e-01 1.23544812e+00 -9.27069306e-01
6.35233939e-01 1.02088618e+00 9.99444604e-01 5.66590667e-01
-1.23033726e+00 -4.83290970e-01 3.09871793e-01 1.47895873e-01
-1.02046776e+00 -7.98741460e-01 3.62606913e-01 -6.60412252e-01
1.17444396e+00 -1.51532263e-01 9.00229275e-01 1.57661438e+00
-6.44940659e-02 7.82995343e-01 1.10598886e+00 -1.25393525e-01
1.84668481e-01 2.82466244e-02 9.34927315e-02 1.12262619e+00
2.52700925e-01 1.25638098e-01 -7.58600712e-01 2.42186427e-01
7.54306257e-01 3.36804241e-02 -5.42431414e-01 -3.21931720e-01
-1.36708808e+00 7.87271500e-01 3.60344201e-01 3.35166991e-01
-1.70682788e-01 6.58590078e-01 7.52294123e-01 6.27441287e-01
5.23555398e-01 4.86099571e-01 -4.80480760e-01 -4.45772469e-01
-1.17270768e+00 -9.29802377e-03 7.06473231e-01 1.12728715e+00
5.68565726e-01 4.55071479e-01 -1.16424493e-01 4.61620599e-01
4.09792438e-02 1.63635164e-01 6.20031953e-01 -1.20362520e+00
5.79372406e-01 3.66913706e-01 -2.80136704e-01 -4.94796216e-01
-2.53883123e-01 -3.97654206e-01 -7.61693835e-01 3.52469474e-01
9.51973498e-01 -1.79711240e-03 -9.41543579e-01 1.79557037e+00
-4.21588957e-01 1.99976489e-01 8.99631232e-02 8.39457929e-01
1.07384920e+00 1.93345860e-01 2.14223757e-01 1.94985010e-02
1.37664688e+00 -1.08716691e+00 -2.29621157e-01 -1.94150627e-01
6.90087557e-01 -2.95295268e-01 1.33306408e+00 5.22395849e-01
-1.05347562e+00 -4.85875040e-01 -1.03282773e+00 -2.80709088e-01
-3.22069615e-01 1.32850572e-01 9.13701296e-01 7.12952256e-01
-1.57823658e+00 6.20321631e-01 -8.06228757e-01 -8.16029012e-01
1.08451200e+00 3.91298681e-01 -4.74042624e-01 -1.94322485e-02
-8.34387958e-01 8.22792470e-01 6.90755099e-02 1.50111122e-02
-1.49935400e+00 -8.36600423e-01 -8.24619532e-01 9.83676165e-02
3.30624163e-01 -8.26601386e-01 1.41741192e+00 -1.51870906e+00
-1.25113654e+00 1.18507564e+00 5.83037846e-02 -8.69283199e-01
6.92092240e-01 3.88265960e-02 1.30452984e-03 5.43738008e-01
-7.09462836e-02 1.07368648e+00 1.00861144e+00 -7.81769454e-01
-5.51065624e-01 -5.44660650e-02 8.36112380e-01 3.74755226e-02
-6.83247805e-01 -2.62066454e-01 -2.93790698e-01 -4.49319571e-01
-2.41449893e-01 -9.90931153e-01 6.14764653e-02 3.35985541e-01
-2.19838783e-01 -2.16742784e-01 7.65240967e-01 -2.08668411e-01
5.82737505e-01 -2.06605768e+00 8.85473378e-03 -1.86250687e-01
7.51676917e-01 2.59470433e-01 -2.77006149e-01 2.13945180e-01
-2.73047060e-01 8.56642127e-02 5.82989529e-02 -4.65264231e-01
-6.59998879e-02 7.55746439e-02 -3.28956962e-01 6.14952028e-01
1.93936765e-01 1.27574241e+00 -8.53200853e-01 -5.55538177e-01
3.05331796e-01 3.65843356e-01 -7.29611099e-01 -1.08482949e-01
-1.88562691e-01 4.78513166e-02 -1.11166790e-01 7.94295490e-01
1.24747932e-01 -6.48148358e-01 1.11063860e-01 -6.61705732e-01
-1.31604612e-01 1.40831307e-01 -2.68350393e-01 1.83899295e+00
-3.71574819e-01 1.24501801e+00 -2.29390543e-02 -1.30449021e+00
6.62190169e-02 4.83933628e-01 4.60755110e-01 -6.22354746e-01
1.34017274e-01 1.35863747e-03 7.48548955e-02 -6.37768686e-01
3.68715286e-01 -1.95011958e-01 4.88109708e-01 3.52963537e-01
6.32420480e-01 7.28744492e-02 1.67587087e-01 5.00721753e-01
1.39183652e+00 3.99562329e-01 -6.91781044e-02 -4.74550575e-01
1.06794350e-01 2.06972390e-01 1.37137741e-01 9.62162137e-01
-4.93534654e-01 6.57600224e-01 5.82300246e-01 -5.00252783e-01
-1.11047399e+00 -1.11761165e+00 -6.61733150e-02 1.13212633e+00
-9.86632183e-02 -7.06087768e-01 -8.09280276e-01 -8.72157872e-01
-2.20067531e-01 5.61959706e-02 -9.48018968e-01 -1.29023686e-01
-2.29932323e-01 -5.90125799e-01 9.77733076e-01 8.48481536e-01
6.89372897e-01 -8.48400354e-01 -8.27684045e-01 -1.26258850e-01
6.57478198e-02 -1.56941330e+00 -2.30055213e-01 4.42757428e-01
-9.41007257e-01 -1.25681806e+00 -7.57650852e-01 -7.00410068e-01
5.46324253e-01 4.50667471e-01 1.18441594e+00 1.97907034e-02
-2.19537601e-01 1.09963083e+00 -3.38272512e-01 -1.14289373e-01
-6.64698929e-02 1.21938713e-01 1.11038625e-01 -1.99068666e-01
4.02969182e-01 -8.19526553e-01 -7.63108432e-01 2.62678742e-01
-7.03053296e-01 1.40353218e-01 6.39928460e-01 9.10229087e-01
1.08001903e-01 -1.44135758e-01 3.52208316e-01 -8.91514063e-01
2.14777559e-01 -4.13603812e-01 -3.80091846e-01 1.27792671e-01
-4.97222900e-01 -3.62922321e-03 7.80411303e-01 -6.19232595e-01
-8.29001606e-01 5.03517278e-02 1.31758198e-01 -8.68309677e-01
3.19925509e-02 3.97286534e-01 1.75242081e-01 -3.74548227e-01
7.22939968e-01 2.73367316e-01 1.94579944e-01 -7.74917081e-02
3.78150612e-01 3.14961851e-01 5.40703416e-01 -5.35172939e-01
8.22941899e-01 8.96056473e-01 -8.71281847e-02 -1.02010953e+00
-9.41545963e-01 -3.07138115e-01 -5.37012696e-01 -3.77612352e-01
1.07687366e+00 -1.30416775e+00 -9.17932034e-01 3.42406690e-01
-8.29530239e-01 -6.32818282e-01 -5.28894007e-01 4.99369383e-01
-8.02881181e-01 3.07699919e-01 -8.57414246e-01 -4.87686813e-01
-1.98946655e-01 -1.31251490e+00 8.03859353e-01 -1.62008971e-01
-1.71314180e-01 -1.08548641e+00 -3.41600627e-01 4.56446379e-01
6.01646006e-01 -1.68457985e-01 6.83013558e-01 -6.20890796e-01
-7.78169632e-01 2.17887580e-01 -5.19748449e-01 5.89549839e-01
-4.20042723e-02 -6.10213261e-03 -1.55784309e+00 -3.94517690e-01
-1.61965325e-01 -9.50464427e-01 1.28760839e+00 3.51970792e-01
1.35311413e+00 -1.19215205e-01 -1.60060208e-02 9.42984343e-01
1.31926143e+00 -2.06976786e-01 6.75035655e-01 2.61568308e-01
9.18719053e-01 4.04675573e-01 -2.23816261e-02 4.59950194e-02
3.77321005e-01 4.09599096e-01 5.40565550e-01 -1.17709406e-01
-3.42828423e-01 -2.90026605e-01 6.76628888e-01 6.81621075e-01
-3.83703440e-01 -2.07923472e-01 -6.72367036e-01 4.68268961e-01
-1.74279571e+00 -1.28927279e+00 2.04917237e-01 1.91354454e+00
5.77239215e-01 5.77418864e-01 3.64093006e-01 1.75757766e-01
2.58715868e-01 3.45257938e-01 -4.88331109e-01 -7.45342299e-02
-2.22554073e-01 2.19640374e-01 7.00414062e-01 -4.17146012e-02
-1.25011468e+00 8.51539791e-01 7.14178896e+00 7.01691985e-01
-1.29313231e+00 3.43274683e-01 7.39340007e-01 -3.91820818e-01
-1.09470718e-01 4.50816602e-02 -3.78989607e-01 2.98557103e-01
1.09282279e+00 1.74782693e-01 4.82991248e-01 8.77709329e-01
2.84595741e-03 7.99630135e-02 -1.64717340e+00 1.25833762e+00
2.81227846e-02 -1.70528698e+00 1.28045231e-01 1.77128971e-01
3.91838849e-01 6.29691601e-01 3.45137775e-01 3.63120437e-01
2.87347734e-01 -1.24981606e+00 1.12660611e+00 1.69540718e-01
1.14824915e+00 -1.23980694e-01 2.83836275e-01 -9.97963473e-02
-1.19166958e+00 -2.44123071e-01 -3.48055154e-01 -2.38108993e-01
-2.28713468e-01 1.79061890e-01 -6.23966455e-01 1.46269307e-01
9.79581773e-01 1.40025973e+00 -8.65948260e-01 6.48699284e-01
1.85477927e-01 9.39739764e-01 -2.34236583e-01 1.83572277e-01
6.99243426e-01 -1.10728340e-02 1.71802744e-01 1.32269371e+00
1.77753139e-02 1.91746633e-02 -6.01726249e-02 7.72264838e-01
-2.93170482e-01 -3.44959199e-01 -1.03681409e+00 -4.25549418e-01
1.33036882e-01 1.18160987e+00 -8.11998963e-01 -5.04622757e-01
-8.56582880e-01 7.81112850e-01 4.67812032e-01 5.39373755e-01
-6.51409924e-01 4.48070131e-02 5.49707353e-01 3.27688187e-01
4.28338289e-01 -1.70302913e-01 1.88868009e-02 -1.60431659e+00
-3.92472968e-02 -8.02503109e-01 2.77774155e-01 -9.90065873e-01
-1.23012388e+00 5.87437212e-01 -5.85352890e-02 -1.35498393e+00
-3.99370529e-02 -1.12823761e+00 -5.47326624e-01 2.13457227e-01
-1.45846820e+00 -1.29749966e+00 -4.19523448e-01 9.21283662e-01
7.57125795e-01 -3.63271832e-01 6.26662552e-01 3.87533635e-01
-4.48768407e-01 6.96964800e-01 -2.18704373e-01 5.80148876e-01
4.58908200e-01 -1.18068707e+00 2.13992402e-01 7.91647673e-01
5.36197424e-01 6.59281194e-01 4.68956977e-01 2.72884611e-02
-1.68417251e+00 -1.00838387e+00 2.48894915e-01 -7.64686465e-01
8.38955820e-01 -5.77627540e-01 -3.91467929e-01 1.13646877e+00
5.11784792e-01 4.08636004e-01 4.90154058e-01 1.12885125e-01
-9.01231170e-01 -8.50515068e-02 -9.09446180e-01 4.93988156e-01
1.62111294e+00 -1.09486866e+00 -4.27506089e-01 3.13671708e-01
7.15024114e-01 5.88851124e-02 -7.26314962e-01 2.13459939e-01
7.25029647e-01 -1.30743372e+00 1.03893781e+00 -7.34124422e-01
7.52926230e-01 -4.66971733e-02 -3.28740358e-01 -1.09138989e+00
-2.09848493e-01 -4.98516768e-01 -6.47779033e-02 1.01436186e+00
3.41947943e-01 -5.91301084e-01 9.69234586e-01 3.88887018e-01
-2.09056318e-01 -6.83574796e-01 -8.72591734e-01 -8.26216877e-01
7.31946602e-02 -6.58705413e-01 2.59425081e-02 9.29947019e-01
9.52031016e-02 3.59453976e-01 -2.20408380e-01 -4.12975848e-01
7.00447857e-01 -2.05993935e-01 4.89508390e-01 -9.59642529e-01
-4.05831367e-01 -5.84005415e-01 -8.59652936e-01 -1.14930356e+00
4.00787294e-01 -9.70916748e-01 -2.27523625e-01 -1.02795649e+00
3.61491531e-01 -4.09183383e-01 -2.49589160e-01 6.81411862e-01
3.64065081e-01 7.18509436e-01 2.35833332e-01 3.49758387e-01
-9.86524940e-01 3.38801771e-01 1.06631541e+00 -4.42832887e-01
4.61721361e-01 -2.59021014e-01 -7.76984632e-01 8.78478765e-01
5.40289402e-01 -3.28884423e-02 -8.95814836e-01 -6.94284260e-01
1.34909615e-01 -2.40090758e-01 7.77200103e-01 -1.13743591e+00
2.38002867e-01 1.66990146e-01 4.23642576e-01 -2.80000512e-02
4.71371025e-01 -7.91533113e-01 -1.94685876e-01 3.71915698e-01
-5.84782898e-01 1.46895021e-01 1.60850748e-01 6.79816604e-01
-1.76863462e-01 1.12641871e-01 7.45533884e-01 -4.28244412e-01
-1.03795993e+00 6.02532744e-01 -7.50332296e-01 4.00797248e-01
7.70093203e-01 -5.95939636e-01 -6.01142645e-01 -5.97077072e-01
-7.38920271e-01 -7.93071538e-02 5.63149154e-01 3.36318374e-01
6.63266957e-01 -1.08031714e+00 -4.96266067e-01 -1.06589468e-02
3.53773117e-01 -2.46789828e-01 -1.05068982e-01 1.10667682e+00
-4.46345389e-01 4.79603648e-01 -3.40390652e-01 -8.81903887e-01
-9.69679534e-01 5.99063277e-01 5.67647755e-01 1.32445484e-01
-8.07940125e-01 9.30711269e-01 4.99350101e-01 1.91691350e-02
4.77663249e-01 -6.00814462e-01 5.20029962e-02 1.30091533e-01
3.93192291e-01 9.34471339e-02 5.78695759e-02 -5.43786824e-01
-3.51187944e-01 3.65550786e-01 -6.62747547e-02 -1.12193018e-01
1.42733085e+00 -2.22189464e-02 1.88470304e-01 4.95122641e-01
1.31670451e+00 -4.35731471e-01 -1.60024750e+00 -1.66882202e-01
-1.20607063e-01 -2.35829294e-01 1.47659272e-01 -4.73766714e-01
-1.48578858e+00 1.00165641e+00 4.42581296e-01 9.37930793e-02
9.65721250e-01 3.37044038e-02 3.26442093e-01 6.83168232e-01
5.17101884e-01 -7.46808946e-01 3.35110933e-01 3.93138558e-01
3.82306457e-01 -1.37970817e+00 -6.56561777e-02 -1.24987602e-01
-6.05578184e-01 1.08011901e+00 7.75191844e-01 -2.39338964e-01
7.44432926e-01 3.55127215e-01 -1.03618070e-01 -4.63302910e-01
-1.13912964e+00 -3.53173494e-01 -4.63161469e-02 7.25171804e-01
5.39792538e-01 -3.73156488e-01 4.13422227e-01 2.35510990e-01
-9.77502167e-02 2.66563326e-01 5.82614064e-01 8.17645371e-01
-1.20538943e-01 -6.54976547e-01 1.76458851e-01 8.36591661e-01
-5.19014299e-01 -4.06986207e-01 -3.21915925e-01 9.66674805e-01
2.15255022e-02 9.44897115e-01 2.74521500e-01 -6.39734685e-01
-7.23551661e-02 4.37672213e-02 9.59651232e-01 -4.90272075e-01
-7.11599410e-01 -3.10006082e-01 2.50227720e-01 -7.35724092e-01
-1.05075860e+00 -5.65916419e-01 -7.60213256e-01 -4.92506117e-01
-1.83239773e-01 -1.31409476e-02 3.04017425e-01 9.81836438e-01
1.15707181e-01 2.98498094e-01 1.27876446e-01 -1.03082383e+00
-4.66679543e-01 -7.80615509e-01 -4.58839446e-01 4.26679045e-01
4.60628450e-01 -7.18339860e-01 -4.91463155e-01 3.80830497e-01] | [9.342842102050781, 1.3416926860809326] |
65ee3759-b648-435c-9676-0476f8fe8c9e | toward-a-universal-model-for-shape-from | null | null | http://openaccess.thecvf.com/content_CVPR_2020/html/Verbin_Toward_a_Universal_Model_for_Shape_From_Texture_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Verbin_Toward_a_Universal_Model_for_Shape_From_Texture_CVPR_2020_paper.pdf | Toward a Universal Model for Shape From Texture | We consider the shape from texture problem, where the input is a single image of a curved, textured surface, and the texture and shape are both a priori unknown. We formulate this task as a three-player game between a shape process, a texture process, and a discriminator. The discriminator adapts a set of non-linear filters to try to distinguish image patches created by the texture process from those created by the shape process, while the shape and texture processes try to create image patches that are indistinguishable from those of the other. An equilibrium of this game yields two things: an estimate of the 2.5D surface from the shape process, and a stochastic texture synthesis model from the texture process. Experiments show that this approach is robust to common non-idealities such as shading, gloss, and clutter. We also find that it succeeds for a wide variety of texture types, including both periodic textures and those composed of isolated textons, which have previously required distinct and specialized processing.
| [' Todd Zickler', 'Dor Verbin'] | 2020-06-01 | null | null | null | cvpr-2020-6 | ['shape-from-texture'] | ['computer-vision'] | [ 7.52833366e-01 6.79533556e-02 6.26260221e-01 -1.16835594e-01
-6.94158673e-01 -8.21561635e-01 6.77930653e-01 -2.73890585e-01
-1.01395240e-02 3.53113741e-01 -2.69875705e-01 -9.39996168e-02
1.54266477e-01 -9.42283094e-01 -6.08615220e-01 -1.23724973e+00
2.26643175e-01 8.31127107e-01 5.39682508e-01 -2.23471880e-01
4.83462363e-01 3.98828208e-01 -1.81088161e+00 5.40460289e-01
5.33595324e-01 1.21772873e+00 2.47585446e-01 1.06551826e+00
-1.64744288e-01 3.01317722e-01 -3.83187324e-01 -2.48268738e-01
5.18845081e-01 -5.05410194e-01 -5.00245631e-01 6.25914514e-01
4.56209809e-01 4.01804484e-02 1.97249427e-01 1.15684903e+00
3.74251485e-01 3.13104615e-02 1.13474047e+00 -5.32944500e-01
-5.08575797e-01 -3.52809995e-01 -8.58244359e-01 -4.12865460e-01
3.18414509e-01 1.92922816e-01 5.73383629e-01 -8.16831529e-01
6.53423071e-01 1.47433138e+00 7.30805337e-01 4.10178006e-01
-1.70389926e+00 -7.03500360e-02 -2.02322230e-01 -5.92252076e-01
-1.28569567e+00 -5.45506299e-01 6.27783239e-01 -5.30412018e-01
4.34638888e-01 5.20961702e-01 4.28120404e-01 8.05537879e-01
6.58289909e-01 4.87043798e-01 1.51306844e+00 -5.99785507e-01
5.16683400e-01 3.10941041e-01 -5.25949776e-01 6.12939179e-01
9.09947008e-02 2.88339913e-01 -1.12690307e-01 -4.65210855e-01
1.15823185e+00 -4.49325740e-01 -2.44813919e-01 -6.47661865e-01
-6.31167829e-01 4.23102349e-01 -2.99951077e-01 -2.40670331e-02
-4.29055840e-01 -1.33792415e-01 -2.83393711e-01 5.41582942e-01
6.68212354e-01 4.12725449e-01 -1.82500318e-01 1.59922823e-01
-5.85291505e-01 2.34510124e-01 1.22712660e+00 7.70821214e-01
9.52171564e-01 -9.98118892e-02 2.44322062e-01 8.10664892e-01
2.57420063e-01 1.21120262e+00 -7.38360435e-02 -9.25741017e-01
-1.43950828e-03 2.40240678e-01 4.02657330e-01 -1.00278676e+00
-2.23938953e-02 3.02406311e-01 -5.21042407e-01 1.00187862e+00
9.17690277e-01 -1.87331006e-01 -1.19680154e+00 1.40142119e+00
5.04095137e-01 -1.20582961e-01 -1.84816048e-01 9.93834972e-01
5.60713887e-01 5.22950888e-01 -6.38572693e-01 9.20225575e-05
1.28354537e+00 -4.32472616e-01 -2.97683716e-01 -1.64102033e-01
9.68327746e-02 -1.39002478e+00 1.02308559e+00 5.28160870e-01
-1.38431978e+00 -3.60061646e-01 -7.79492080e-01 1.46580622e-01
-2.65201870e-02 6.23505674e-02 1.77005976e-01 8.06652427e-01
-1.08482409e+00 5.17887890e-01 -5.42848587e-01 -4.02958751e-01
1.75269395e-02 1.42619863e-01 -3.24865371e-01 -1.00914262e-01
-4.83558953e-01 6.66656137e-01 -2.59665251e-01 1.94181755e-01
-6.02470398e-01 3.74866240e-02 -5.06657183e-01 -1.56739533e-01
3.61638039e-01 -5.34622788e-01 1.23413730e+00 -1.74944758e+00
-1.81596673e+00 1.27154577e+00 -4.73881990e-01 3.03370804e-01
6.07446194e-01 9.10427868e-02 -2.09854499e-01 9.03235450e-02
-5.78495115e-02 -6.67767152e-02 1.39881289e+00 -1.74232173e+00
-5.05289555e-01 -4.80970412e-01 -4.64635879e-01 5.13731480e-01
7.27408409e-01 -2.02833220e-01 -5.38261235e-01 -3.95832032e-01
4.84596342e-01 -9.28468227e-01 -1.73526242e-01 1.39826015e-01
-4.44676340e-01 3.85932148e-01 6.97689891e-01 -3.72706980e-01
4.74610955e-01 -2.29364944e+00 -2.75836326e-02 6.95042014e-01
7.28338361e-02 -2.07409516e-01 -3.61961216e-01 4.19250399e-01
7.62350783e-02 1.61344796e-01 -1.29790500e-01 -2.65968591e-01
-1.55815110e-01 1.75091311e-01 -3.09624702e-01 7.26576686e-01
3.57126385e-01 5.29055357e-01 -6.05695784e-01 -1.36876464e-01
-2.06348583e-01 1.89183220e-01 -3.45594078e-01 2.39164397e-01
-4.55096483e-01 3.79166245e-01 -6.38134360e-01 6.32038593e-01
9.58879113e-01 -5.89394867e-02 3.29463065e-01 1.24543421e-01
-3.45953315e-01 -1.11389123e-01 -1.54919541e+00 8.59507561e-01
-1.22668900e-01 4.69971001e-01 5.66958547e-01 -6.06462121e-01
1.25151312e+00 2.52654523e-01 2.79116124e-01 -7.42380798e-01
3.71246576e-01 5.18366992e-01 3.87075990e-02 -6.34518743e-01
5.55247426e-01 -7.18657553e-01 1.61363289e-01 7.58751273e-01
-2.66575545e-01 -9.41872239e-01 -3.02315265e-01 -3.67889643e-01
1.01880610e+00 2.23206654e-01 -1.17325373e-01 -5.57537675e-01
1.30693600e-01 -9.78940129e-02 4.22802806e-01 1.11710417e+00
2.07104847e-01 1.15596294e+00 7.40704656e-01 -6.66909039e-01
-1.26086390e+00 -1.27871919e+00 -5.21626137e-02 6.64943457e-01
6.30906224e-01 3.38728666e-01 -7.02722907e-01 -1.96141765e-01
1.80376500e-01 2.97119766e-01 -9.21001792e-01 2.84996182e-02
-3.84936124e-01 -6.30288839e-01 2.79949252e-02 -4.30862904e-02
3.01169306e-01 -9.89712656e-01 -5.12448788e-01 1.36653304e-01
-6.89148903e-02 -7.05860078e-01 -6.54209077e-01 2.03945532e-01
-6.76558197e-01 -1.16716135e+00 -6.83096945e-01 -8.98908377e-01
9.14474487e-01 2.62699008e-01 1.37942863e+00 1.94502816e-01
-3.31989378e-01 5.40819585e-01 -2.44023234e-01 -4.94358957e-01
-5.26343465e-01 -7.28707433e-01 -2.51343429e-01 8.20105433e-01
3.73151246e-03 -2.80922771e-01 -3.27105850e-01 9.06883895e-01
-9.16336060e-01 1.30829319e-01 4.37246412e-01 8.28725755e-01
9.56426263e-01 5.50890088e-01 -1.22627221e-01 -9.08016562e-01
3.63182962e-01 -1.81177929e-01 -7.14221656e-01 2.63243616e-01
-5.38940616e-02 8.86746272e-02 3.36439133e-01 -6.97884500e-01
-1.56072164e+00 1.75778612e-01 2.45818943e-01 -4.44551073e-02
-4.36531216e-01 2.67336033e-02 -3.22398841e-01 -3.03199589e-01
9.46303427e-01 2.96985298e-01 3.40316474e-01 -5.70488930e-01
1.19369917e-01 5.01813412e-01 4.90516096e-01 -9.47844386e-01
9.47842598e-01 9.44052160e-01 -1.54712602e-01 -1.20757806e+00
-3.95971209e-01 -8.45407322e-02 -5.05949259e-01 -2.03671977e-01
7.91352689e-01 -4.11447644e-01 -4.25389677e-01 1.02889216e+00
-1.13069630e+00 -1.00299919e+00 -6.65821314e-01 1.29392892e-01
-8.25645149e-01 1.46514729e-01 -3.48938555e-01 -1.07528067e+00
2.55849004e-01 -1.05784488e+00 1.32965899e+00 4.84974474e-01
6.86936378e-02 -1.12339175e+00 2.66163349e-01 -3.01855560e-02
4.94615018e-01 2.58850694e-01 1.14428735e+00 -1.62986949e-01
-5.98829091e-01 -3.03085506e-01 -1.18121848e-01 2.28624076e-01
3.97477329e-01 2.91113228e-01 -1.08836484e+00 -2.14617878e-01
5.30795574e-01 -3.20706695e-01 7.55788028e-01 4.98639166e-01
5.14148176e-01 -1.70605227e-01 -1.19084336e-01 5.85095525e-01
1.46502447e+00 5.49194038e-01 9.17049944e-01 -2.32859105e-02
3.20605636e-01 9.02500689e-01 2.96103448e-01 1.26370996e-01
-3.42041850e-02 5.82308888e-01 2.65343249e-01 -6.18121266e-01
-3.87426242e-02 1.73607588e-01 5.54118641e-02 2.35624239e-01
-1.84062690e-01 -5.59442997e-01 -8.82915556e-01 2.95852095e-01
-1.55813193e+00 -9.90941048e-01 -1.45332381e-01 2.65555120e+00
5.55179954e-01 1.24157585e-01 -2.47323513e-03 -1.98838994e-01
8.57954443e-01 -8.76326263e-02 -5.42866886e-01 -6.51711047e-01
-6.15477324e-01 3.18988323e-01 4.39720005e-01 6.71718597e-01
-9.89834487e-01 9.38665509e-01 7.40864182e+00 6.89601600e-01
-1.34326994e+00 -5.42642117e-01 9.33692336e-01 3.31184298e-01
-6.68302357e-01 1.07350945e-01 -2.79659241e-01 2.78775305e-01
1.79197475e-01 1.01721391e-01 6.56707942e-01 2.25406066e-01
9.06603411e-02 -7.22790897e-01 -8.58108342e-01 7.13807285e-01
8.21179971e-02 -7.20417261e-01 -1.48639411e-01 4.77836914e-02
8.26904774e-01 -1.36858478e-01 3.33775997e-01 -4.26927567e-01
5.36324084e-01 -1.12351346e+00 8.17504823e-01 8.14363062e-01
9.25021470e-01 -1.36421800e-01 6.30702138e-01 2.37115934e-01
-1.10948670e+00 3.06559086e-01 -2.55683273e-01 -9.97688174e-02
-4.64902967e-02 6.70342982e-01 -3.66327852e-01 2.50006109e-01
5.43625653e-01 1.21920900e-02 -1.01723894e-01 1.10267770e+00
-1.03118150e-02 3.09600443e-01 -6.54278755e-01 1.75032899e-01
-6.00430742e-02 -6.38305604e-01 5.98573089e-01 9.09398377e-01
3.63673389e-01 4.11373138e-01 2.44004607e-01 9.64932561e-01
4.87327039e-01 9.50192288e-02 -7.70025194e-01 1.78174242e-01
1.47164509e-01 1.07033372e+00 -1.17437911e+00 -1.94803238e-01
-3.47867936e-01 1.04159319e+00 -1.03073865e-01 7.38123000e-01
-2.54007488e-01 -2.86580324e-01 5.04662395e-01 2.54235923e-01
4.92870092e-01 -1.47642180e-01 -5.81821322e-01 -1.10101855e+00
-6.75566792e-02 -9.79629219e-01 -1.43922269e-01 -1.07941723e+00
-1.43512201e+00 5.94835460e-01 -3.59239846e-01 -1.03579772e+00
7.28103295e-02 -7.49280572e-01 -8.10082853e-01 1.36339319e+00
-8.76858354e-01 -8.59579027e-01 -1.32802591e-01 4.38681930e-01
1.73625991e-01 1.46116555e-01 7.78258383e-01 -2.90174216e-01
-2.21750494e-02 -2.75996290e-02 4.56813425e-01 -2.53633261e-01
5.11273503e-01 -1.30401766e+00 5.72894394e-01 5.76038003e-01
-2.49538627e-02 9.19026583e-02 8.53070557e-01 -6.94631100e-01
-1.54055846e+00 -5.20179272e-01 6.28186941e-01 -6.56114042e-01
3.18143725e-01 -5.08629978e-01 -1.04867494e+00 1.25001088e-01
-2.16215402e-01 -1.02045037e-01 1.29624665e-01 -1.67187244e-01
-3.74311715e-01 2.03744844e-01 -1.15699875e+00 6.68865144e-01
5.76108992e-01 -5.15701115e-01 -3.72451633e-01 1.23433201e-02
-3.98055613e-01 -6.07164860e-01 -3.96253407e-01 1.55419350e-01
1.04604471e+00 -1.25413752e+00 5.66632330e-01 -4.10196483e-01
2.58621126e-01 -3.14087242e-01 -3.38798821e-01 -1.32753801e+00
-4.65665787e-01 -7.59708941e-01 6.97667003e-01 7.38100410e-01
4.97067928e-01 -8.75972807e-01 9.27555621e-01 6.33077979e-01
2.34912202e-01 -4.83527392e-01 -7.98575580e-01 -9.39406276e-01
1.06602088e-01 9.49092582e-03 2.54717231e-01 6.35726631e-01
-5.81727624e-01 2.62632042e-01 -2.62081981e-01 1.24813192e-01
7.05252707e-01 5.47095001e-01 8.73826206e-01 -1.46039283e+00
-6.60594285e-01 -4.83483553e-01 -1.03546433e-01 -1.16030991e+00
-3.99972498e-01 -3.19048554e-01 6.89787626e-01 -1.16065788e+00
8.67874399e-02 -7.58680105e-01 3.46876323e-01 2.89305657e-01
-1.20740794e-01 1.73022941e-01 -7.86629394e-02 2.02120394e-01
-5.91930486e-02 2.31249690e-01 1.45684183e+00 3.62354517e-02
-4.91947562e-01 2.18504563e-01 -6.29133344e-01 9.82653618e-01
5.93333364e-01 -4.77829278e-01 -3.61120343e-01 -5.61483979e-01
1.16913535e-01 2.95316547e-01 2.69861072e-01 -5.50220191e-01
-1.21258780e-01 -3.91972929e-01 4.65272009e-01 -4.39782627e-02
4.10625368e-01 -6.68438017e-01 4.99182642e-01 -1.20285712e-02
4.75721918e-02 -1.22052923e-01 1.82414621e-01 5.79820812e-01
1.10020682e-01 -6.30187988e-02 1.05054319e+00 -2.10313812e-01
-1.46275520e-01 1.32580996e-01 -6.45226657e-01 1.29677266e-01
6.21215999e-01 -7.14756548e-01 -3.33848476e-01 -6.98904514e-01
-8.26805711e-01 -1.74407616e-01 1.07798672e+00 7.64621645e-02
3.83804768e-01 -1.13873959e+00 -6.44850314e-01 7.79385865e-01
-3.70417684e-01 -1.03481095e-02 1.96600869e-01 7.60993779e-01
-5.47464848e-01 -3.71693105e-01 -1.20814957e-01 -6.73817039e-01
-1.32729268e+00 1.37079284e-01 6.24208212e-01 1.72862746e-02
-6.04997694e-01 8.87953043e-01 9.69869733e-01 -4.41365153e-01
-2.22924426e-01 -1.74270511e-01 3.35419983e-01 -3.35495442e-01
5.08081496e-01 5.78579493e-02 7.69672683e-03 -7.00511396e-01
7.17897043e-02 1.08039200e+00 2.73496777e-01 -4.49071348e-01
1.07631671e+00 -2.89740503e-01 -9.92407799e-02 6.69758737e-01
8.79118979e-01 3.07089537e-01 -1.61935329e+00 -2.80190229e-01
-2.11990952e-01 -6.57732368e-01 -2.17619106e-01 -7.84897685e-01
-9.17272806e-01 8.15247059e-01 4.21105504e-01 7.27001250e-01
1.18107843e+00 -4.14999425e-02 3.87481779e-01 2.05616370e-01
5.00279367e-01 -1.00526345e+00 1.48390099e-01 7.42304921e-01
9.80406046e-01 -8.06690693e-01 -2.02835232e-01 -5.03191352e-01
-5.96199870e-01 1.34573722e+00 2.81257927e-01 -3.84299695e-01
7.32469738e-01 6.95024073e-01 4.93077070e-01 -3.68906617e-01
-8.31788361e-01 -2.71107644e-01 3.38308901e-01 6.22168601e-01
-7.57145742e-03 1.11529501e-02 3.27205062e-01 1.02732204e-01
4.97237556e-02 -3.43902886e-01 4.64389205e-01 8.71930838e-01
-6.16752446e-01 -9.42210615e-01 -8.52553368e-01 4.87650692e-01
-2.45142490e-01 4.90366332e-02 -6.85493469e-01 6.87307179e-01
5.61952442e-02 6.96507335e-01 4.05373633e-01 -3.00134420e-01
4.96043742e-01 -2.70942505e-02 7.05837190e-01 -6.02473378e-01
-1.95573151e-01 7.29212523e-01 2.57499427e-01 -4.47739989e-01
-8.15256983e-02 -8.23235333e-01 -6.12294376e-01 -1.87528566e-01
-5.17850995e-01 -9.48088393e-02 5.81776798e-01 7.61050105e-01
8.60276893e-02 -1.46073028e-01 7.36081302e-01 -1.19527376e+00
-3.60378265e-01 -4.57731664e-01 -1.19940341e+00 5.37060380e-01
3.04641694e-01 -5.98685086e-01 -6.49826109e-01 2.97143966e-01] | [9.798440933227539, -2.960512161254883] |
65155cc1-df8f-467a-92cc-d61a98ce5a21 | a-cognition-based-attention-model-for | null | null | https://aclanthology.org/D17-1048 | https://aclanthology.org/D17-1048.pdf | A Cognition Based Attention Model for Sentiment Analysis | Attention models are proposed in sentiment analysis because some words are more important than others. However,most existing methods either use local context based text information or user preference information. In this work, we propose a novel attention model trained by cognition grounded eye-tracking data. A reading prediction model is first built using eye-tracking data as dependent data and other features in the context as independent data. The predicted reading time is then used to build a cognition based attention (CBA) layer for neural sentiment analysis. As a comprehensive model, We can capture attentions of words in sentences as well as sentences in documents. Different attention mechanisms can also be incorporated to capture other aspects of attentions. Evaluations show the CBA based method outperforms the state-of-the-art local context based attention methods significantly. This brings insight to how cognition grounded data can be brought into NLP tasks. | ['Chu-Ren Huang', 'Minglei Li', 'Rong Xiang', 'Yunfei Long', 'Qin Lu'] | 2017-09-01 | null | null | null | emnlp-2017-9 | ['product-recommendation'] | ['miscellaneous'] | [ 7.83711523e-02 -5.20476475e-02 -1.72042295e-01 -7.89230585e-01
-4.35327262e-01 -1.74606219e-01 6.35320902e-01 5.68700016e-01
-6.35448635e-01 3.99395198e-01 7.05485523e-01 -2.73524612e-01
1.13027066e-01 -6.16923332e-01 -4.91039485e-01 -4.10185575e-01
6.06853187e-01 8.99314582e-02 3.26555341e-01 -4.82548356e-01
9.49325383e-01 -7.21239299e-02 -1.44141448e+00 5.68524837e-01
9.91110444e-01 9.65269625e-01 7.69458175e-01 7.63853610e-01
-5.06431162e-01 8.90446723e-01 -6.45427406e-01 -3.00683051e-01
-3.69473219e-01 -4.83681977e-01 -1.02978563e+00 -2.77096063e-01
2.49582276e-01 -2.43718959e-02 2.56557256e-01 1.02404654e+00
5.22923291e-01 6.90725088e-01 3.78794104e-01 -9.34377015e-01
-1.80775237e+00 4.81709808e-01 -5.08716047e-01 7.23468184e-01
5.35523057e-01 -3.63006182e-02 1.19128621e+00 -1.20298254e+00
9.34278071e-02 1.37121105e+00 1.29905760e-01 6.76576257e-01
-6.57020092e-01 -3.43475938e-01 1.05611575e+00 6.53644562e-01
-7.65107572e-01 -1.96960434e-01 9.09622192e-01 -1.59893334e-01
1.24208617e+00 3.19968104e-01 7.21914709e-01 1.07727742e+00
5.67465901e-01 1.30283034e+00 8.25557292e-01 -8.25320065e-01
1.81202456e-01 1.76788941e-01 8.13023031e-01 2.96684116e-01
-1.05661727e-01 -2.95434892e-01 -8.48869443e-01 3.48735482e-01
1.69334970e-02 5.24553180e-01 -4.96217728e-01 4.05884743e-01
-1.04828644e+00 9.35196877e-01 7.17684209e-01 4.07185256e-01
-6.23982906e-01 -9.30246264e-02 8.12525526e-02 3.36360186e-02
8.73777032e-01 5.83377779e-01 -8.55909526e-01 1.14163328e-02
-6.23128712e-01 -1.05075590e-01 3.07872504e-01 8.29118431e-01
6.73053980e-01 -2.21808821e-01 -9.18846190e-01 6.97789907e-01
8.93714607e-01 4.02143061e-01 9.64294434e-01 -8.70148242e-02
5.18948317e-01 8.94058168e-01 1.08395204e-01 -1.01396978e+00
-6.42839909e-01 -3.44583601e-01 -3.18833828e-01 -2.56433249e-01
-1.75645113e-01 -2.34526426e-01 -8.74620676e-01 1.61538053e+00
1.50798246e-01 -5.79409152e-02 -9.74468567e-05 9.86115038e-01
9.17263031e-01 8.45165193e-01 3.77653748e-01 -1.42888844e-01
1.67360532e+00 -1.47437143e+00 -1.09778368e+00 -7.00957417e-01
6.29875898e-01 -7.86805511e-01 1.53165030e+00 2.27983296e-01
-1.01386011e+00 -8.02769780e-01 -8.95890832e-01 -5.46728730e-01
-7.36369312e-01 1.13688558e-01 4.01305377e-01 4.83036637e-01
-1.15657651e+00 1.79526165e-01 -5.77996492e-01 -5.98471165e-01
4.77713764e-01 1.72953144e-01 3.20631325e-01 5.73263429e-02
-1.34714448e+00 1.12735176e+00 5.01736142e-02 3.78660440e-01
-4.19427723e-01 -4.76523370e-01 -9.77912068e-01 4.69234496e-01
1.40537143e-01 -5.45576930e-01 1.38958323e+00 -1.22563529e+00
-1.55957294e+00 6.44450843e-01 -7.69540906e-01 -1.77204460e-01
-4.00132358e-01 -7.62157977e-01 -7.15918660e-01 -1.59244969e-01
7.95715973e-02 5.65626740e-01 7.53459513e-01 -8.53190601e-01
-7.05807686e-01 -4.84708786e-01 2.44439423e-01 3.87921274e-01
-8.49886596e-01 2.73711473e-01 -5.95584452e-01 -6.08574629e-01
-3.32014441e-01 -7.12036610e-01 -7.54114464e-02 -3.11936617e-01
-4.11762118e-01 -7.00947821e-01 8.76010776e-01 -5.48570693e-01
1.49456120e+00 -1.77354610e+00 5.20797037e-02 -2.36309484e-01
4.58515994e-02 4.25385177e-01 -4.50638503e-01 3.20898980e-01
-1.21196382e-01 2.33763739e-01 3.26300830e-01 -6.62781537e-01
-5.20443358e-02 -2.75988251e-01 -3.51873457e-01 -1.00650284e-02
4.27085757e-01 1.30874527e+00 -1.08150148e+00 -2.65486121e-01
1.77626491e-01 4.56644684e-01 -6.42953634e-01 1.38080359e-01
-3.24555695e-01 5.04430473e-01 -6.59696817e-01 5.04204571e-01
4.17849481e-01 -4.25001144e-01 -2.10251629e-01 -5.70328049e-02
-1.59764424e-01 5.30881524e-01 -2.92019248e-01 1.71248269e+00
-8.07166815e-01 9.96667624e-01 -4.08985347e-01 -7.42664993e-01
1.01213562e+00 1.44381747e-01 -2.42650196e-01 -1.12558842e+00
5.25243580e-01 -6.44069076e-01 1.61133543e-01 -7.61038065e-01
8.26669633e-01 2.77290028e-02 1.55262396e-01 6.49681926e-01
1.83272790e-02 3.83248925e-01 -2.77862679e-02 1.64737314e-01
4.80763316e-01 -1.39399758e-02 2.53230065e-01 -5.83890259e-01
8.29170346e-01 -2.10239619e-01 2.68152773e-01 7.16218293e-01
-3.25805008e-01 5.65871119e-01 3.50576222e-01 -3.24813724e-01
-3.81103784e-01 -1.48404315e-01 1.08026065e-01 1.89214981e+00
3.41267169e-01 -7.15597093e-01 -8.58356178e-01 -7.54032135e-01
-5.12749135e-01 1.20651817e+00 -1.18986773e+00 -5.72968245e-01
-4.47470844e-02 -6.17069781e-01 -4.03774709e-01 1.02375376e+00
4.69494224e-01 -1.55988479e+00 -5.65311491e-01 -3.88011038e-02
-1.70480177e-01 -6.79627240e-01 -7.18863726e-01 1.08184591e-01
-7.11142242e-01 -8.40946376e-01 -5.81892133e-01 -9.26992118e-01
8.26471090e-01 3.79087150e-01 1.13095164e+00 4.38692421e-01
2.43335053e-01 4.26626593e-01 -8.25191200e-01 -1.28715801e+00
3.17078114e-01 1.64122373e-01 -1.32402539e-01 1.85377389e-01
1.35200930e+00 2.40121946e-01 -9.18163061e-01 4.99409810e-02
-5.27060688e-01 -6.28958866e-02 4.22977507e-01 8.05855811e-01
4.66367096e-01 -3.89241189e-01 6.17478430e-01 -9.30526376e-01
1.27384067e+00 -6.18241847e-01 -3.61671269e-01 5.42933345e-01
-7.91680634e-01 1.19216383e-01 5.85997820e-01 -4.10188228e-01
-1.33849490e+00 -3.05050731e-01 -1.34992704e-01 -1.01805389e-01
-2.41380721e-01 8.38960290e-01 -1.86846957e-01 4.07206744e-01
4.60095257e-01 1.19472846e-01 -2.59349465e-01 -4.45697904e-01
1.14775449e-01 9.01721478e-01 -1.73626900e-01 -2.49462619e-01
-1.52985230e-01 -7.84601201e-04 -5.54445565e-01 -5.71516216e-01
-1.72041845e+00 -5.56570351e-01 -7.32477665e-01 -1.62958473e-01
1.20202518e+00 -6.60854101e-01 -8.35230708e-01 3.00062925e-01
-1.25790238e+00 -2.55802989e-01 1.02714738e-02 4.30397332e-01
-1.50758952e-01 -7.84507394e-02 -3.83860797e-01 -9.84502077e-01
-6.63018942e-01 -1.20639729e+00 1.24963295e+00 5.85507870e-01
-4.09559906e-01 -1.42294824e+00 8.80832151e-02 3.69790643e-01
5.11663139e-01 -5.13362765e-01 6.48051798e-01 -1.01677513e+00
-1.30415916e-01 -3.20473909e-01 -1.67993516e-01 8.82248953e-03
2.18573794e-01 -5.48395850e-02 -1.16105831e+00 2.16619838e-02
2.42730439e-01 -3.02470326e-01 9.75189030e-01 7.26690412e-01
1.58699894e+00 -1.54166549e-01 -4.51835752e-01 9.38722342e-02
1.16312957e+00 3.62435430e-01 7.39413321e-01 4.45858508e-01
6.93403542e-01 6.47612333e-01 9.25540924e-01 2.71401167e-01
6.14811540e-01 3.70523393e-01 4.19713944e-01 -4.22321409e-02
1.57888278e-01 -1.07145362e-01 3.35391015e-01 8.18978667e-01
1.10841757e-02 -7.71686435e-01 -1.00433886e+00 7.42836297e-01
-2.04958510e+00 -9.77835357e-01 -3.96342129e-01 1.92810941e+00
6.83015406e-01 1.64227292e-01 -2.94213057e-01 -1.99707612e-01
8.92091751e-01 1.27944425e-01 -5.94121754e-01 -8.88164282e-01
1.50634319e-01 1.36880279e-01 -2.29433313e-01 6.19328737e-01
-1.10043812e+00 1.09912312e+00 6.19020367e+00 4.13972229e-01
-1.20600986e+00 1.99105844e-01 7.01552570e-01 -3.17817479e-01
-3.73730123e-01 -1.10563397e-01 -1.09713578e+00 6.64107680e-01
1.07773936e+00 -1.43898502e-01 -5.48685901e-02 8.84440780e-01
3.24817508e-01 -2.64045477e-01 -1.06904888e+00 7.79735386e-01
6.14925861e-01 -1.09020817e+00 2.11330041e-01 -2.02358544e-01
6.89199567e-01 -6.44336417e-02 2.78862327e-01 5.53220272e-01
5.30336834e-02 -8.94724846e-01 6.00548089e-01 9.40400481e-01
2.35394359e-01 -7.35231400e-01 9.78220344e-01 4.40513223e-01
-9.33528721e-01 -2.13603914e-01 -6.24302387e-01 -4.25338477e-01
1.66450188e-01 1.80764318e-01 -3.93301249e-01 9.60914791e-02
7.43757665e-01 1.07671976e+00 -9.03102458e-01 8.69755626e-01
-6.97258115e-01 7.76768565e-01 2.24102661e-01 -8.06239784e-01
4.01573837e-01 9.85386297e-02 6.52928874e-02 1.16308272e+00
1.96004733e-01 3.53703827e-01 -1.50425285e-01 7.63798416e-01
-1.00612730e-01 4.08872187e-01 -2.33966112e-01 1.62528470e-01
1.40829489e-01 1.47686994e+00 -6.81811094e-01 -3.85317534e-01
-8.30680907e-01 1.05838883e+00 5.80105901e-01 5.15048623e-01
-7.63421297e-01 -4.45711821e-01 5.35245299e-01 -2.61191964e-01
5.42050362e-01 2.27991864e-01 -4.26179707e-01 -1.24802005e+00
-2.07326055e-01 -4.19323415e-01 3.75584155e-01 -1.48035514e+00
-1.20013320e+00 6.58177853e-01 -4.54791665e-01 -1.07857311e+00
1.21054739e-01 -8.83401036e-01 -1.12310207e+00 1.13457942e+00
-1.79683161e+00 -9.65864778e-01 -4.32478219e-01 6.55571461e-01
1.18179858e+00 -4.61612381e-02 8.57948244e-01 -2.09663242e-01
-8.49775136e-01 6.08411729e-01 -1.13560684e-01 9.28868875e-02
8.01289201e-01 -1.35357606e+00 2.95357108e-01 1.01531291e+00
2.45058045e-01 1.03587210e+00 4.38973546e-01 -6.53893054e-01
-1.08131433e+00 -9.89376426e-01 1.26698744e+00 -8.39144349e-01
5.21889925e-01 -2.41754934e-01 -1.01752150e+00 9.44087982e-01
8.99302185e-01 -1.34921178e-01 1.08405662e+00 6.24023855e-01
5.45144901e-02 1.58548892e-01 -6.77242994e-01 6.74607813e-01
5.47148705e-01 -5.45079529e-01 -1.02535689e+00 2.43627235e-01
1.02880180e+00 -1.48534387e-01 -3.77047598e-01 -1.33003294e-01
2.31337488e-01 -7.25343883e-01 6.20681345e-01 -9.41160798e-01
7.41539299e-01 -1.57508869e-02 1.47041842e-01 -1.58128309e+00
-7.10600853e-01 -1.47528052e-01 -3.21813822e-01 1.30417788e+00
7.01862335e-01 -4.16728467e-01 4.11106944e-01 1.05393696e+00
-3.94843668e-01 -7.81813204e-01 -1.93893880e-01 -2.33908981e-01
1.72301650e-01 -4.64718997e-01 6.86443865e-01 9.95503545e-01
5.11744857e-01 1.05783474e+00 -1.74636945e-01 5.68365417e-02
-1.42069831e-01 2.51340568e-01 2.08924398e-01 -1.35939014e+00
2.56330252e-01 -4.56377417e-01 2.32638363e-02 -1.29558480e+00
3.40742588e-01 -6.24952257e-01 1.10891223e-01 -1.83068883e+00
2.39477813e-01 2.14645788e-01 -9.08616066e-01 6.32243454e-01
-9.80472267e-01 -9.33524743e-02 2.53297716e-01 -9.55172181e-02
-1.09723437e+00 8.56463850e-01 1.45731163e+00 -3.29602920e-02
-1.96989521e-01 -3.07949120e-03 -1.26198995e+00 7.84373462e-01
8.80757332e-01 -2.60237515e-01 -5.79509556e-01 -7.99688935e-01
5.91200471e-01 -3.82867336e-01 1.00655325e-01 -6.89290285e-01
6.35363936e-01 -1.89324781e-01 6.60179019e-01 -9.06733990e-01
3.20699364e-01 -7.37951398e-01 -1.06648922e+00 5.98156862e-02
-7.42714465e-01 1.43100500e-01 3.69005710e-01 7.58112669e-01
-3.69789422e-01 -4.84194010e-01 3.09248924e-01 -3.45972180e-02
-9.18747544e-01 2.26566449e-01 -4.88866091e-01 2.16347091e-02
8.18826079e-01 -1.24071382e-01 -4.75677133e-01 -5.59818029e-01
-6.89997196e-01 5.43477654e-01 1.13563593e-02 9.65922236e-01
7.38732576e-01 -1.19917822e+00 -4.54627514e-01 1.46894082e-01
4.78964627e-01 -7.62013867e-02 3.43759775e-01 9.20380354e-01
1.00807451e-01 7.82275021e-01 3.54293399e-02 -4.53240812e-01
-1.21451342e+00 1.07274318e+00 7.78320804e-02 7.19297528e-02
-1.54496849e-01 1.17615700e+00 4.76236194e-01 -2.68650770e-01
7.89022148e-02 -5.82321107e-01 -1.09158254e+00 3.82027745e-01
1.17640233e+00 -7.49729574e-02 1.66168198e-01 -5.53779542e-01
-4.90653604e-01 7.36456037e-01 -4.11317974e-01 1.98288739e-01
1.28421104e+00 -6.12213373e-01 6.53394833e-02 6.09884083e-01
8.99235070e-01 -5.22353053e-02 -1.19656205e+00 -2.71461189e-01
4.99327853e-02 -3.17097366e-01 5.33387184e-01 -1.04891193e+00
-1.08600903e+00 1.40781248e+00 4.12984222e-01 3.89065385e-01
1.23100758e+00 7.91144818e-02 6.03276491e-01 4.38813627e-01
-1.56678498e-01 -1.32884979e+00 2.65507877e-01 8.10033739e-01
1.12291682e+00 -1.63082266e+00 -1.54697210e-01 4.04947363e-02
-1.08396649e+00 1.04319394e+00 1.12652588e+00 -6.14301860e-02
8.13393116e-01 -1.53809816e-01 1.39113575e-01 -5.23103416e-01
-1.02181399e+00 -3.69112790e-01 7.30997384e-01 4.81715918e-01
9.98885691e-01 -4.32559073e-01 -3.52978081e-01 1.27748823e+00
-5.41920774e-03 -6.53859526e-02 3.54767203e-01 9.13451433e-01
-4.94075984e-01 -7.88284540e-01 -3.46000463e-01 4.24564868e-01
-4.70540464e-01 -7.09628522e-01 -5.44049978e-01 3.38494092e-01
-1.61161795e-01 1.28930020e+00 4.33420181e-01 -2.75116652e-01
2.07302734e-01 2.85189658e-01 -7.67341629e-02 -1.00161481e+00
-8.92917633e-01 -4.80961055e-02 -3.07669789e-01 -5.04260063e-01
-5.82055449e-01 -5.15916646e-01 -1.17423618e+00 -2.76264641e-03
-9.46167767e-01 9.44772363e-02 6.19566143e-01 1.16493046e+00
7.72767603e-01 9.37594295e-01 3.50207031e-01 -6.63190663e-01
1.53714970e-01 -1.30240667e+00 -1.91291168e-01 3.34132731e-01
4.54337090e-01 -5.17186165e-01 -1.99396208e-01 9.91699621e-02] | [11.437372207641602, 6.650262355804443] |
8ab5e246-1597-4b88-854e-1bd6d2f2edf1 | near-optimal-multi-agent-learning-for-safe | 2210.06380 | null | https://arxiv.org/abs/2210.06380v1 | https://arxiv.org/pdf/2210.06380v1.pdf | Near-Optimal Multi-Agent Learning for Safe Coverage Control | In multi-agent coverage control problems, agents navigate their environment to reach locations that maximize the coverage of some density. In practice, the density is rarely known $\textit{a priori}$, further complicating the original NP-hard problem. Moreover, in many applications, agents cannot visit arbitrary locations due to $\textit{a priori}$ unknown safety constraints. In this paper, we aim to efficiently learn the density to approximately solve the coverage problem while preserving the agents' safety. We first propose a conditionally linear submodular coverage function that facilitates theoretical analysis. Utilizing this structure, we develop MacOpt, a novel algorithm that efficiently trades off the exploration-exploitation dilemma due to partial observability, and show that it achieves sublinear regret. Next, we extend results on single-agent safe exploration to our multi-agent setting and propose SafeMac for safe coverage and exploration. We analyze SafeMac and give first of its kind results: near optimal coverage in finite time while provably guaranteeing safety. We extensively evaluate our algorithms on synthetic and real problems, including a bio-diversity monitoring task under safety constraints, where SafeMac outperforms competing methods. | ['Andreas Krause', 'Melanie N. Zeilinger', 'Matteo Turchetta', 'Manish Prajapat'] | 2022-10-12 | null | null | null | null | ['safe-exploration'] | ['robots'] | [ 2.63480306e-01 7.17252493e-01 -5.82770944e-01 1.45807475e-01
-1.02541375e+00 -9.68342483e-01 5.37249818e-02 5.31487525e-01
-5.76041400e-01 1.42500103e+00 -1.73423082e-01 -2.66057789e-01
-5.04093826e-01 -1.06871951e+00 -1.12724555e+00 -1.04099560e+00
-6.20096147e-01 7.92524517e-01 -7.94464201e-02 -5.46611100e-03
1.47661595e-02 8.60761032e-02 -1.20879960e+00 -5.69149613e-01
1.05970061e+00 9.26965058e-01 1.49242878e-01 5.48261046e-01
6.23683155e-01 8.20232987e-01 -6.15315437e-01 6.01631738e-02
4.21085238e-01 -3.97436172e-01 -6.78689599e-01 2.06684411e-01
-4.56535131e-01 -4.59415913e-01 -4.13266495e-02 1.27988434e+00
1.56696036e-01 -9.28068766e-04 5.02972901e-01 -1.69198966e+00
-1.72622353e-01 1.01601529e+00 -9.63559330e-01 -1.50564983e-01
3.57958108e-01 1.37810066e-01 8.91079843e-01 3.66229594e-01
6.24330699e-01 1.00938761e+00 4.16294485e-01 6.12877607e-01
-1.21571815e+00 -3.96465600e-01 4.67925847e-01 -4.17272031e-01
-1.19553757e+00 -2.02465639e-01 1.89308554e-01 -2.76314497e-01
7.56863892e-01 5.48441887e-01 8.02186668e-01 7.34552026e-01
4.06365037e-01 7.81501889e-01 9.96468246e-01 -7.02018812e-02
7.06904054e-01 4.92363200e-02 -3.23922276e-01 6.35038853e-01
9.68700826e-01 4.64306086e-01 -2.95211345e-01 -4.05091733e-01
5.37318528e-01 1.63273010e-02 -5.60869396e-01 -5.55010319e-01
-1.13693607e+00 1.05804801e+00 2.54383236e-01 -1.39788389e-01
-3.89723748e-01 5.63205361e-01 2.67873872e-02 4.95481938e-01
1.43474758e-01 4.80710387e-01 -3.11720282e-01 -1.82866454e-02
-3.87968570e-01 4.03597713e-01 1.09165716e+00 1.32485616e+00
4.59206939e-01 -2.08165377e-01 1.00587681e-01 -3.17858309e-02
2.47771263e-01 7.60897100e-01 -4.31361198e-01 -1.20058835e+00
6.02606475e-01 5.28584421e-01 8.78055513e-01 -7.18030870e-01
-7.16715932e-01 -5.51460683e-01 -7.81859040e-01 3.71699989e-01
4.98491883e-01 -7.49542058e-01 -5.65296233e-01 2.12571192e+00
5.64876854e-01 -2.34708592e-01 2.83821136e-01 7.83687174e-01
-2.34972775e-01 6.66791022e-01 -2.52574831e-01 -1.07527614e+00
9.43684042e-01 -7.94343054e-01 -6.65528297e-01 -4.04953778e-01
8.65314484e-01 1.82856888e-01 6.43709660e-01 4.40613806e-01
-1.18950856e+00 6.75744116e-01 -1.18303776e+00 7.23438621e-01
-2.89943162e-02 -5.13193667e-01 6.43296242e-01 7.92652488e-01
-8.36150467e-01 1.37162894e-01 -1.03825605e+00 -3.10426950e-01
5.43849051e-01 4.03622240e-01 -1.49457023e-01 -9.90426168e-02
-6.72632039e-01 7.27174282e-01 5.01973450e-01 -3.13700020e-01
-1.54606724e+00 -6.64631844e-01 -1.05868244e+00 -4.67928536e-02
1.26648664e+00 -6.27240181e-01 1.19744921e+00 -2.26572230e-01
-1.14878559e+00 3.37459475e-01 1.40439972e-01 -7.93068111e-01
7.01664746e-01 2.81137973e-01 3.66407841e-01 1.09139830e-01
3.18210185e-01 4.27428544e-01 9.59025100e-02 -1.47178137e+00
-1.08424461e+00 -3.27294469e-01 7.32401371e-01 2.81864196e-01
-2.78154641e-01 -6.16035998e-01 2.09003985e-02 -1.62984326e-01
-2.85897523e-01 -9.33479667e-01 -8.88047218e-01 -2.37658601e-02
-5.77343285e-01 1.30590916e-01 1.90864697e-01 4.51350808e-02
9.36048508e-01 -1.77133191e+00 3.72148514e-01 3.27876061e-01
1.32878289e-01 -5.94832778e-01 -1.63199440e-01 6.48665071e-01
9.95070279e-01 -2.67295279e-02 -7.33937025e-01 -4.09337491e-01
1.12867475e-01 4.39238250e-01 -4.84731086e-02 1.00860167e+00
-4.27647024e-01 7.50509679e-01 -9.94190395e-01 -3.61750335e-01
-8.82379040e-02 -7.47819468e-02 -5.82364619e-01 -6.31612390e-02
-7.82863319e-01 4.06405300e-01 -7.37249911e-01 8.70063186e-01
6.94947243e-01 -2.00065672e-01 4.53500450e-01 6.58483684e-01
-1.46812141e-01 -4.02058929e-01 -9.68280077e-01 1.46886563e+00
-3.53448510e-01 -6.80548400e-02 7.75231123e-01 -9.37678993e-01
2.50832826e-01 2.14701500e-02 8.94054770e-01 -4.58079815e-01
4.29983497e-01 2.35477984e-01 -4.25497919e-01 -1.48684904e-01
1.61284700e-01 -1.69748276e-01 -5.49252450e-01 7.82490611e-01
-6.05547547e-01 -5.79356141e-02 1.04217716e-01 1.89698607e-01
1.44760561e+00 -2.74685264e-01 5.12440026e-01 -6.82807624e-01
-2.09464841e-02 3.96939784e-01 7.70884573e-01 1.05945659e+00
-2.46768802e-01 7.06799254e-02 9.26502883e-01 -1.31502256e-01
-6.81623399e-01 -7.48378098e-01 2.17223421e-01 7.24080324e-01
7.81390309e-01 4.68469150e-02 -9.28641438e-01 -6.73776209e-01
2.79480428e-01 8.64335418e-01 -1.14757872e+00 8.25964194e-03
-1.92535982e-01 -9.55526292e-01 2.76179343e-01 1.38556674e-01
2.20124155e-01 -5.82015872e-01 -1.34372473e+00 3.10180813e-01
-9.58757177e-02 -7.64093876e-01 -6.55989528e-01 4.01163340e-01
-4.30246979e-01 -1.26909530e+00 -4.98494565e-01 -4.50706750e-01
8.72005999e-01 3.56605917e-01 6.89887524e-01 1.22580379e-01
-1.75209671e-01 5.24405897e-01 -3.25777531e-01 -7.07260847e-01
-1.23347983e-01 1.22603670e-01 5.36307581e-02 -4.19456899e-01
-6.39232155e-03 -4.88152921e-01 -4.36461002e-01 1.86819613e-01
-9.66501236e-01 -1.39833614e-01 4.01090473e-01 5.01955688e-01
1.00327027e+00 5.05116463e-01 7.87395537e-01 -5.76079071e-01
5.26337445e-01 -8.03251147e-01 -1.40534294e+00 3.92666548e-01
-3.92667890e-01 9.94546637e-02 2.67630905e-01 -3.42334688e-01
-5.92937291e-01 3.70725930e-01 2.97327638e-01 -1.07361205e-01
3.34285349e-01 5.09072244e-01 -2.76160359e-01 -2.43693173e-01
3.57848555e-01 2.28935599e-01 3.04458499e-01 7.35477433e-02
6.81801513e-02 1.42261505e-01 2.08334818e-01 -6.89730108e-01
7.19682634e-01 8.53321612e-01 4.99934494e-01 -3.98495555e-01
-8.72833788e-01 4.03372273e-02 1.77297682e-01 -1.20471485e-01
4.75252479e-01 -8.63318384e-01 -1.67038822e+00 2.17053965e-02
-8.08140993e-01 -7.61188865e-01 -6.98745787e-01 2.52397627e-01
-1.00740910e+00 2.00732544e-01 -1.19192265e-01 -1.53657389e+00
-1.14446163e-01 -1.09027326e+00 8.73447061e-01 1.98700175e-01
2.63504803e-01 -9.58890021e-01 3.57599109e-01 5.47249708e-03
3.72291356e-01 8.81786704e-01 5.34586072e-01 -2.09388956e-01
-9.68627930e-01 1.64914466e-02 3.31177860e-01 -5.35801351e-01
3.10425758e-02 -6.99585080e-01 -2.90010810e-01 -8.69796455e-01
7.17559457e-02 -5.47490716e-01 7.32487082e-01 7.95135915e-01
9.46355224e-01 -1.07471704e+00 -9.04401302e-01 5.48688591e-01
1.53144622e+00 3.60803962e-01 2.57714897e-01 6.49482906e-01
-4.51041907e-02 6.86197817e-01 1.06281543e+00 1.10878372e+00
6.91416204e-01 4.29460377e-01 1.25489712e+00 3.58822882e-01
8.53468537e-01 -1.28168449e-01 2.92640716e-01 -3.01538646e-01
1.15780488e-01 -9.32423830e-01 -4.29102451e-01 7.35645354e-01
-2.06320024e+00 -7.04144955e-01 3.48202318e-01 2.42205644e+00
9.33144033e-01 -7.12020621e-02 5.47886908e-01 -1.28701344e-01
5.14379144e-01 -1.09807953e-01 -1.01812601e+00 -9.78164673e-02
-3.17741990e-01 -5.51383674e-01 1.18921232e+00 1.12338138e+00
-9.94897842e-01 4.72012699e-01 5.88740301e+00 8.05112600e-01
-3.48665863e-01 2.93185830e-01 7.87138462e-01 -7.27332950e-01
-6.41363800e-01 -1.89950123e-01 -6.28655791e-01 2.53629386e-01
5.50749183e-01 -5.43899536e-01 8.42083395e-01 8.69693756e-01
1.23851657e-01 -6.28750741e-01 -1.18896890e+00 4.94865060e-01
-2.33029157e-01 -1.18795562e+00 -4.65408385e-01 6.62624896e-01
1.05005372e+00 -9.10333544e-02 -2.07810430e-03 -1.63326338e-01
9.21104193e-01 -1.13120615e+00 9.01031315e-01 1.76732019e-01
7.65058696e-01 -1.30237842e+00 5.98980010e-01 7.96649218e-01
-1.14252222e+00 -4.40307915e-01 -4.13609028e-01 -1.45775646e-01
6.75492942e-01 5.23203135e-01 -4.57213968e-01 5.37778139e-01
5.95589817e-01 2.58149564e-01 3.16525280e-01 1.08981717e+00
-4.04314101e-02 -6.40651286e-02 -8.48717034e-01 -3.44672799e-01
3.47137123e-01 -1.62515104e-01 9.76776481e-01 5.72868586e-01
4.62102383e-01 4.80978042e-01 4.90250856e-01 8.39002669e-01
1.56073389e-03 -3.80330801e-01 -7.53167212e-01 -6.88543022e-02
8.82406175e-01 9.80077446e-01 -7.42685318e-01 1.53391689e-01
1.01664647e-01 5.64432323e-01 2.72293031e-01 2.55302250e-01
-1.04279625e+00 -2.35803649e-01 6.93564594e-01 -1.28950506e-01
1.30300611e-01 -4.99471389e-02 -5.09786665e-01 -6.47892177e-01
6.14192672e-02 -5.27929962e-01 6.37868047e-01 1.53217807e-01
-9.57427204e-01 3.55823785e-01 6.99526072e-02 -9.38832402e-01
-2.28026450e-01 -1.15383342e-01 -3.05332124e-01 1.73496902e-01
-1.42988634e+00 -9.24354672e-01 -6.66793957e-02 3.54783177e-01
1.81728095e-01 1.22226425e-01 4.39957559e-01 -1.42382577e-01
-5.56473196e-01 6.08224332e-01 2.11161345e-01 -7.31720328e-01
-7.14671090e-02 -1.06470335e+00 -1.42798096e-01 9.01480317e-01
-6.43691778e-01 2.08320484e-01 9.69052732e-01 -8.52441788e-01
-1.88478553e+00 -1.31904328e+00 1.13277562e-01 -2.38769010e-01
5.83392799e-01 -3.49043280e-01 -2.00984672e-01 8.83939326e-01
2.17958912e-01 -1.68981418e-01 2.97796160e-01 -4.70196903e-01
1.86291665e-01 1.78534742e-02 -1.77769935e+00 6.44635558e-01
1.28066826e+00 4.18543100e-01 1.95574299e-01 4.06328261e-01
1.18458056e+00 -3.51155281e-01 -7.10244417e-01 4.89453226e-01
3.58026356e-01 -6.33701622e-01 6.05886817e-01 -2.46666923e-01
-1.56166777e-01 -3.05412173e-01 -4.21950847e-01 -1.37149549e+00
-4.14596796e-02 -1.22219408e+00 -3.46639335e-01 7.93532729e-01
6.07590616e-01 -9.66007769e-01 1.02845919e+00 6.19203269e-01
-2.84927189e-02 -8.29234004e-01 -1.46048021e+00 -1.12068665e+00
2.12985829e-01 5.18960580e-02 7.93857753e-01 6.71368599e-01
5.13930678e-01 -3.51752937e-01 -6.18798912e-01 6.49027169e-01
1.21436012e+00 1.68984517e-01 4.55626547e-01 -7.94329405e-01
-4.37369496e-01 -1.39083892e-01 2.49526843e-01 -9.63830888e-01
5.77073283e-02 -3.05999249e-01 5.13952255e-01 -1.56228876e+00
6.40535414e-01 -9.16099846e-01 2.36865819e-01 6.16363883e-01
2.56422728e-01 -1.57325774e-01 3.96925732e-02 -2.10432798e-01
-1.17350078e+00 7.10605145e-01 1.29148400e+00 -4.56572652e-01
-4.56910640e-01 9.91808027e-02 -1.04301083e+00 4.48902816e-01
8.84488821e-01 -5.84126174e-01 -7.49099910e-01 -4.68355626e-01
5.01242220e-01 6.65638030e-01 1.41835615e-01 -8.49275649e-01
2.96273232e-01 -1.06393588e+00 -4.17493671e-01 -5.23540795e-01
4.26848412e-01 -1.12477219e+00 5.69972336e-01 1.14224625e+00
-2.73981601e-01 -3.80067706e-01 2.39218585e-02 1.15411031e+00
4.39587504e-01 -3.69517118e-01 7.78884053e-01 -2.52745207e-02
2.76598167e-02 6.12563908e-01 -5.27016282e-01 1.54096901e-01
1.94585896e+00 -1.93098020e-02 -7.05323935e-01 -6.86053276e-01
-3.35992843e-01 1.18994069e+00 8.55688453e-01 -2.03850523e-01
4.27675515e-01 -9.66587842e-01 -6.88553154e-01 -2.12129831e-01
1.72483042e-01 4.10481274e-01 3.37966472e-01 9.99239087e-01
-3.10035288e-01 4.50549215e-01 8.11617300e-02 -3.13861310e-01
-8.96695375e-01 9.95848835e-01 5.29668331e-01 -2.45497227e-01
-3.07068378e-01 7.68875957e-01 3.53372455e-01 -1.48006514e-01
5.43146372e-01 -2.54537195e-01 5.11978678e-02 -2.67981231e-01
6.63829625e-01 5.98470509e-01 -5.31381905e-01 -6.75334558e-02
-4.66355979e-01 4.08540875e-01 1.21413760e-01 -4.87214208e-01
1.41170537e+00 -4.98834580e-01 -1.30499512e-01 -2.08808765e-01
6.30905330e-01 -2.30041165e-02 -1.72405255e+00 7.89030343e-02
-1.99853957e-01 -4.92055029e-01 -2.34451890e-01 -8.53142262e-01
-1.17860925e+00 1.44736379e-01 -4.30212542e-03 4.78469282e-01
1.18463290e+00 3.30316871e-01 4.32900578e-01 5.71894288e-01
1.39118659e+00 -9.01737273e-01 -2.13706732e-01 5.93114235e-02
9.15659189e-01 -9.80140924e-01 -9.69688129e-03 -4.40385312e-01
-5.93940616e-01 4.20973212e-01 5.69267094e-01 1.01792328e-01
3.33021194e-01 8.85626674e-01 -6.82943404e-01 -9.76065472e-02
-1.00208712e+00 -2.28792116e-01 -7.48220503e-01 8.75053883e-01
-6.46949053e-01 4.02609080e-01 -2.55194217e-01 7.65600502e-01
8.76539201e-02 -2.67995924e-01 8.91866505e-01 1.25280118e+00
-1.05310810e+00 -8.53772521e-01 -5.51588714e-01 5.56121469e-01
-4.40025866e-01 4.48656529e-01 -3.08614224e-01 7.02453077e-01
-4.60021049e-02 1.43354857e+00 -4.74166460e-02 -7.12591708e-02
3.45558091e-03 -9.46263075e-01 6.01215482e-01 -2.18067944e-01
1.11598492e-01 7.50164539e-02 2.90988505e-01 -6.15525067e-01
-3.32813144e-01 -7.75937140e-01 -1.31771863e+00 -5.50110757e-01
-3.98801357e-01 5.22248685e-01 2.53813744e-01 8.11227858e-01
9.17229429e-02 3.31259429e-01 7.07684398e-01 -4.83105630e-01
-8.96968305e-01 -6.20466053e-01 -8.53554070e-01 -5.22931278e-01
7.26486385e-01 -7.20959187e-01 -5.14337718e-01 -5.28115630e-01] | [4.4004974365234375, 2.8118948936462402] |
cb6d902c-7765-42ed-be3d-56bacef34ef1 | learning-a-self-supervised-tone-mapping | 2110.09866 | null | https://arxiv.org/abs/2110.09866v1 | https://arxiv.org/pdf/2110.09866v1.pdf | Learning a self-supervised tone mapping operator via feature contrast masking loss | High Dynamic Range (HDR) content is becoming ubiquitous due to the rapid development of capture technologies. Nevertheless, the dynamic range of common display devices is still limited, therefore tone mapping (TM) remains a key challenge for image visualization. Recent work has demonstrated that neural networks can achieve remarkable performance in this task when compared to traditional methods, however, the quality of the results of these learning-based methods is limited by the training data. Most existing works use as training set a curated selection of best-performing results from existing traditional tone mapping operators (often guided by a quality metric), therefore, the quality of newly generated results is fundamentally limited by the performance of such operators. This quality might be even further limited by the pool of HDR content that is used for training. In this work we propose a learning-based self-supervised tone mapping operator that is trained at test time specifically for each HDR image and does not need any data labeling. The key novelty of our approach is a carefully designed loss function built upon fundamental knowledge on contrast perception that allows for directly comparing the content in the HDR and tone mapped images. We achieve this goal by reformulating classic VGG feature maps into feature contrast maps that normalize local feature differences by their average magnitude in a local neighborhood, allowing our loss to account for contrast masking effects. We perform extensive ablation studies and exploration of parameters and demonstrate that our solution outperforms existing approaches with a single set of fixed parameters, as confirmed by both objective and subjective metrics. | ['Ana Serrano', 'Karol Myszkowski', 'Hans-Peter Seidel', 'Bin Chen', 'Chao Wang'] | 2021-10-19 | null | null | null | null | ['tone-mapping'] | ['computer-vision'] | [ 3.93684298e-01 -4.32285428e-01 -1.11161023e-01 -2.83326477e-01
-5.17738402e-01 -4.60871398e-01 5.92050016e-01 1.04332112e-01
-5.38324594e-01 6.38944924e-01 -2.10909247e-01 -3.21884789e-02
-2.43380666e-01 -7.62585163e-01 -6.25904918e-01 -7.12195814e-01
-1.69556528e-01 -5.84612601e-03 5.75516880e-01 -4.77461040e-01
1.80042133e-01 4.54321742e-01 -1.92027450e+00 1.33482441e-01
1.26605260e+00 1.33670413e+00 5.18472552e-01 4.32317019e-01
-7.58475671e-03 5.49996912e-01 -6.91309214e-01 8.78104195e-03
5.18195629e-01 -6.82478249e-01 -4.32088077e-01 -1.72274634e-01
4.87738907e-01 -2.07108557e-01 -2.71231592e-01 1.10774767e+00
6.60223365e-01 2.32563496e-01 3.25476944e-01 -1.02177072e+00
-6.62234247e-01 2.82188565e-01 -4.89600897e-01 3.70344818e-01
1.50292546e-01 1.73045561e-01 7.84930289e-01 -7.02965617e-01
5.82802176e-01 7.20936298e-01 4.81583565e-01 3.01255286e-01
-1.45864260e+00 -6.73034191e-01 7.85852000e-02 3.14684391e-01
-1.52879655e+00 -1.91544697e-01 9.89694059e-01 -2.46010885e-01
6.84775651e-01 5.14841199e-01 7.79281020e-01 7.28853285e-01
3.58052105e-02 3.12925339e-01 1.69503045e+00 -4.77897197e-01
2.93022066e-01 2.54987210e-01 -2.30827987e-01 5.38519979e-01
-1.51466042e-01 2.20605806e-01 -7.00028658e-01 2.76951104e-01
9.71109152e-01 -3.02730799e-01 -6.52278364e-01 -5.89013994e-01
-1.04438412e+00 6.23084962e-01 6.70735717e-01 5.28561950e-01
-1.02078184e-01 -1.06596172e-01 1.34775952e-01 5.17433345e-01
4.63486642e-01 6.88048720e-01 -9.07220021e-02 -1.68932304e-01
-1.01755214e+00 5.98399192e-02 2.86978275e-01 3.95802677e-01
8.63281250e-01 7.78904855e-02 -3.77630621e-01 1.10828388e+00
-2.88554758e-01 3.79215717e-01 6.54097378e-01 -8.18693578e-01
1.31847352e-01 6.01048768e-01 1.48912771e-02 -1.16708410e+00
-3.60245019e-01 -6.39986813e-01 -8.89919877e-01 7.66225398e-01
5.25603354e-01 1.15289092e-01 -9.35755014e-01 1.95369315e+00
2.02454850e-01 4.46851291e-02 -3.04477304e-01 1.17285585e+00
5.76246858e-01 5.47149479e-01 -9.32695493e-02 -2.82909304e-01
1.01064706e+00 -5.93328357e-01 -7.49893486e-01 -3.05860918e-02
1.84153020e-01 -5.18498361e-01 1.71180654e+00 6.06504619e-01
-1.04131782e+00 -7.88876891e-01 -1.41033673e+00 -7.75481686e-02
-4.20775980e-01 -4.87952493e-02 3.98402661e-01 7.38915980e-01
-1.18419540e+00 7.87704647e-01 -4.80545193e-01 -1.58381552e-01
3.37732136e-01 3.62977535e-01 -2.18250394e-01 6.77410811e-02
-1.24480999e+00 8.71273637e-01 2.99203008e-01 1.95958421e-01
-4.16073292e-01 -8.51831913e-01 -5.01881003e-01 1.54069081e-01
3.25017750e-01 -3.74238104e-01 8.17734480e-01 -9.82861221e-01
-1.73372281e+00 7.94335961e-01 2.96331048e-01 -4.25383210e-01
7.79348910e-01 -6.04855195e-02 -4.08064187e-01 2.21362814e-01
-2.12509662e-01 6.30911946e-01 8.93671632e-01 -1.43155026e+00
-4.72494155e-01 -6.51583150e-02 1.49527967e-01 3.52301806e-01
-6.65341079e-01 -1.28548622e-01 -5.93571484e-01 -7.17864394e-01
8.07524398e-02 -7.09518671e-01 -1.20900925e-02 2.11491123e-01
-2.43202716e-01 2.08366871e-01 7.97785342e-01 -4.85024571e-01
1.30369580e+00 -2.17362165e+00 -6.44985512e-02 4.17562813e-01
2.72305816e-01 3.91241044e-01 -9.47799161e-02 7.87539184e-02
-1.41975060e-01 -2.53097825e-02 -3.48333180e-01 -9.06780213e-02
-8.49439949e-02 -2.65675753e-01 -4.07814980e-01 2.63381451e-01
1.87191591e-01 6.86273038e-01 -7.51351237e-01 -3.62729520e-01
5.32618999e-01 6.46802962e-01 -4.01052147e-01 2.92704791e-01
-1.58472255e-01 6.82155132e-01 -8.64944712e-04 1.99964136e-01
7.62934208e-01 -2.90182382e-01 1.02743655e-01 -4.78714287e-01
-3.17026794e-01 3.12157813e-02 -1.16083646e+00 1.65235221e+00
-6.39830112e-01 7.16699362e-01 -3.29501688e-01 -8.06275427e-01
1.19208324e+00 2.13698223e-02 4.33843136e-01 -1.39583123e+00
2.08749831e-01 2.74786830e-01 1.83699861e-01 -2.51458079e-01
3.95335376e-01 -1.42688453e-01 7.98401460e-02 4.91817743e-01
-2.38220513e-01 -1.83013663e-01 1.16199382e-01 -8.66326690e-02
9.26788151e-01 7.43437633e-02 1.56463861e-01 -1.59727663e-01
3.05960476e-01 -2.00974479e-01 3.68419290e-01 8.42887700e-01
-1.35850877e-01 1.03916299e+00 4.14778352e-01 -3.91013503e-01
-1.05986917e+00 -1.14914370e+00 -5.33720732e-01 1.01446557e+00
4.26975876e-01 -2.22723454e-01 -7.56563723e-01 -3.89344573e-01
-2.89065927e-01 3.61663759e-01 -8.48539650e-01 -3.34186673e-01
-6.21543050e-01 -8.33394408e-01 2.54426390e-01 3.35680872e-01
8.15955460e-01 -1.21877038e+00 -9.81438935e-01 4.12878161e-03
1.06572285e-01 -1.00367367e+00 -3.10346425e-01 3.21087390e-01
-7.21484303e-01 -7.11639345e-01 -9.05456364e-01 -5.97971141e-01
4.87163961e-01 1.72964633e-01 1.01504290e+00 2.02995151e-01
-4.38658267e-01 -8.14964324e-02 -2.76517093e-01 8.47359560e-03
-1.84853554e-01 -4.66190428e-02 -1.73060507e-01 1.21370628e-01
-1.33520916e-01 -8.28757286e-01 -8.87493014e-01 4.36718911e-01
-1.02947676e+00 2.72914529e-01 6.28364623e-01 7.03641176e-01
7.17457831e-01 1.59711584e-01 5.36012053e-01 -8.21125507e-01
7.22626746e-01 -7.61751384e-02 -6.18148625e-01 1.82120934e-01
-6.35670662e-01 -5.30314771e-03 7.61273980e-01 -6.27594829e-01
-9.83443737e-01 -2.22690701e-01 1.57358378e-01 -4.89787638e-01
4.93002012e-02 3.49284321e-01 -1.49672195e-01 -3.25274676e-01
9.44246233e-01 2.17189908e-01 6.49655908e-02 -2.32026339e-01
3.11110795e-01 4.53447551e-01 7.00852990e-01 -4.26783681e-01
8.96207511e-01 3.53565097e-01 -1.02906212e-01 -6.03151321e-01
-5.38867772e-01 1.39438808e-01 -3.79309714e-01 -4.41490114e-01
5.99985480e-01 -5.24726033e-01 -7.65656114e-01 4.65063542e-01
-5.06416976e-01 -6.77209318e-01 -3.85171145e-01 2.87592113e-01
-4.94998693e-01 1.77577779e-01 -3.77124369e-01 -6.76392078e-01
-2.17950106e-01 -9.80840087e-01 6.66625023e-01 2.53319442e-01
-1.79931130e-02 -7.72299647e-01 1.69587448e-01 -1.96869403e-01
8.77294898e-01 3.52627575e-01 1.01493752e+00 -3.57066765e-02
-5.92120767e-01 -3.37185822e-02 -4.79687154e-01 4.31907713e-01
2.81478971e-01 -2.07230151e-01 -1.08356655e+00 -2.61420310e-01
1.10665843e-01 -3.18533599e-01 9.09609973e-01 5.07694900e-01
1.50786257e+00 1.13713600e-01 -1.72441434e-02 7.52256751e-01
1.52305031e+00 2.43338600e-01 7.51773536e-01 5.43860734e-01
5.98621845e-01 6.43672824e-01 5.18336594e-01 1.06033139e-01
-7.64985979e-02 1.10239661e+00 2.03652471e-01 -4.93754089e-01
-3.62621397e-01 -1.19045205e-01 -3.56531069e-02 3.01682413e-01
-1.56273380e-01 -1.45084888e-01 -7.63822734e-01 2.87256420e-01
-1.65741849e+00 -7.56838918e-01 3.83833289e-01 2.65785360e+00
1.00779295e+00 3.95681560e-01 3.03405941e-01 4.33167875e-01
6.49038494e-01 1.79147601e-01 -6.64672792e-01 -1.97458401e-01
-3.43737364e-01 4.33162928e-01 1.74157470e-01 2.57639498e-01
-9.28948224e-01 6.06817305e-01 6.19227362e+00 8.69665921e-01
-1.60201931e+00 -1.39000878e-01 7.22827256e-01 -3.25431108e-01
-3.40893507e-01 -3.36620212e-01 -1.73459381e-01 5.18888593e-01
6.24618530e-01 -1.76771909e-01 5.57505727e-01 4.89615738e-01
3.16313088e-01 -1.31382942e-01 -9.03752029e-01 1.15479541e+00
5.71067743e-02 -1.16300106e+00 -1.52112201e-01 -2.42699385e-02
7.26852119e-01 -2.22904101e-01 7.28391826e-01 8.04510061e-03
-4.70594205e-02 -1.06772614e+00 7.02913225e-01 4.42215264e-01
1.15334558e+00 -7.50615776e-01 4.07587141e-01 -5.04875444e-02
-9.91711974e-01 -1.55193478e-01 -2.94902116e-01 1.30834341e-01
1.06584541e-02 7.88755953e-01 -4.33078855e-01 3.39675933e-01
9.23337638e-01 4.51416075e-01 -8.52779984e-01 1.21082151e+00
3.32869142e-02 3.52804661e-01 -2.80496627e-01 1.01573490e-01
-1.53780043e-01 -1.61053792e-01 3.33821833e-01 9.52369153e-01
3.01864415e-01 -8.41626897e-02 2.24611796e-02 1.05924594e+00
-3.48196663e-02 2.83463031e-01 -4.33938831e-01 3.14507693e-01
4.29668248e-01 1.32724655e+00 -9.04259801e-01 -9.35579389e-02
-1.90330938e-01 9.33213592e-01 3.47179413e-01 3.87304634e-01
-7.54346609e-01 -5.81404030e-01 3.66833925e-01 2.82757223e-01
3.69863659e-01 -2.36479975e-02 -4.23422664e-01 -8.71660829e-01
3.56792301e-01 -8.28401506e-01 1.25382230e-01 -7.80858934e-01
-1.05515933e+00 9.92365122e-01 -3.42749432e-02 -1.62004149e+00
-1.90088436e-01 -4.13821548e-01 -5.19464552e-01 8.93937349e-01
-1.59946835e+00 -7.71339238e-01 -7.05807507e-01 4.83700067e-01
3.43565494e-02 2.47411970e-02 5.99963844e-01 4.41552967e-01
-4.03954089e-01 8.12949300e-01 -2.23682225e-02 -1.30118713e-01
7.26751745e-01 -1.31737006e+00 1.44069552e-01 8.01467717e-01
1.62863567e-01 5.25363564e-01 7.87918508e-01 -3.43591124e-01
-8.64905775e-01 -7.48896122e-01 3.44082862e-01 -1.19377986e-01
3.79510283e-01 -6.84867382e-01 -1.29673254e+00 5.74006028e-02
-3.23147997e-02 2.45993778e-01 4.56593096e-01 2.23966110e-02
-4.36612755e-01 -4.63869303e-01 -1.07762063e+00 6.30583227e-01
1.03654408e+00 -5.60397446e-01 -3.16805132e-02 -9.40766037e-02
6.09442532e-01 -4.82022911e-01 -7.27346838e-01 4.56548721e-01
6.41385317e-01 -1.37931800e+00 7.96479225e-01 -6.39387369e-02
3.93200964e-01 -5.80648243e-01 8.86848122e-02 -1.31934714e+00
-1.11257821e-01 -4.43645567e-01 1.31994650e-01 1.10521102e+00
3.24846834e-01 -5.36855936e-01 7.05266058e-01 5.09923875e-01
-5.88293001e-02 -7.37323284e-01 -7.95197845e-01 -7.65853584e-01
-9.25435871e-02 -3.04614663e-01 4.29642946e-01 8.69908035e-01
-1.69633642e-01 9.91922393e-02 -4.59363818e-01 -1.13486901e-01
5.15660286e-01 3.99928480e-01 5.08060575e-01 -1.13637376e+00
-4.06001717e-01 -5.63647807e-01 -4.29736316e-01 -8.15061152e-01
-1.63630530e-01 -6.95461154e-01 1.95773542e-01 -1.15419114e+00
1.62422180e-01 -8.38658154e-01 -4.49916750e-01 3.14114153e-01
-2.17271015e-01 8.13784897e-01 3.76326680e-01 1.58433869e-01
-3.74625415e-01 6.29064560e-01 1.44137573e+00 1.27310693e-01
-6.38390720e-01 -3.01495284e-01 -5.90755105e-01 3.73964936e-01
6.65014327e-01 -1.28148288e-01 -5.57074010e-01 -3.16884458e-01
2.12300137e-01 -1.32847115e-01 3.90182465e-01 -1.35271311e+00
1.43855765e-01 2.25722557e-03 6.16894782e-01 -1.28643647e-01
2.64426619e-01 -6.93759024e-01 1.77329481e-01 1.67357773e-01
-5.02715170e-01 -5.31165525e-02 1.59075052e-01 3.26345116e-01
-2.43409798e-01 1.50171593e-01 1.00477302e+00 2.57050216e-01
-7.35885441e-01 2.08636031e-01 4.38623736e-03 -1.76181477e-02
7.34312117e-01 -4.44570690e-01 -2.87114859e-01 -4.83852357e-01
-5.65583885e-01 -2.17642173e-01 6.19099200e-01 4.17470604e-01
4.77914423e-01 -1.30103421e+00 -4.17493075e-01 3.62756729e-01
1.30431801e-01 -1.76097855e-01 3.66977304e-01 7.83554494e-01
-3.00320059e-01 -7.58600142e-03 -6.45951807e-01 -6.55960500e-01
-9.38169181e-01 6.20899558e-01 3.77234101e-01 -9.55367833e-02
-8.73623908e-01 3.76874685e-01 3.44526649e-01 6.45379210e-03
2.25673497e-01 -2.75390267e-01 -3.66429955e-01 -1.85263857e-01
6.89526200e-01 9.86911654e-02 2.48713821e-01 -4.31626439e-01
-9.06997323e-02 7.90212035e-01 -6.52872398e-02 -2.99260020e-01
1.29326987e+00 -2.12630108e-01 2.10986927e-01 6.02903962e-01
1.27248073e+00 -3.14567238e-02 -1.43059814e+00 -2.23376036e-01
-2.78842568e-01 -7.21711993e-01 2.11688444e-01 -9.58126545e-01
-1.24071097e+00 7.99418330e-01 1.27318966e+00 2.50951797e-01
1.65476155e+00 -1.49818137e-01 4.62362081e-01 3.89958248e-02
2.84598529e-01 -1.02968037e+00 3.89000505e-01 5.17084114e-02
8.10004175e-01 -1.17340899e+00 -2.31400549e-01 -5.25056422e-01
-5.08699656e-01 9.05566096e-01 6.10531211e-01 -2.11790711e-01
4.15810406e-01 3.30854416e-01 3.93883795e-01 3.90542746e-02
-3.99264038e-01 -3.32176358e-01 4.01053339e-01 6.64946496e-01
2.88386554e-01 -2.95005500e-01 -2.84071982e-01 1.63388938e-01
-4.65930372e-01 1.08001586e-02 2.10490704e-01 7.95797765e-01
-3.98422003e-01 -1.05459499e+00 -6.55425712e-02 4.79317963e-01
-2.82101929e-01 -4.41639498e-02 5.27126202e-03 7.47999370e-01
8.72929916e-02 6.35873020e-01 2.69919008e-01 -5.30055225e-01
4.84061480e-01 -3.08435947e-01 5.59540927e-01 -2.95062602e-01
-5.28568685e-01 3.86893637e-02 -3.46897036e-01 -6.16854548e-01
-4.24371481e-01 -3.40239972e-01 -1.09440541e+00 -1.26081780e-01
-1.69883445e-01 -2.12779511e-02 5.99455893e-01 6.27361476e-01
2.15138346e-01 6.33691311e-01 8.04516733e-01 -7.26061106e-01
-2.30311736e-01 -7.21071959e-01 -6.56819701e-01 7.59283185e-01
3.92604619e-01 -8.91190290e-01 -3.74686837e-01 -5.80482110e-02] | [11.321791648864746, -2.0205507278442383] |
9f200856-c6f0-4af2-aefe-1faf70fa6e04 | perspective-reconstruction-of-human-faces-by | 2208.07142 | null | https://arxiv.org/abs/2208.07142v1 | https://arxiv.org/pdf/2208.07142v1.pdf | Perspective Reconstruction of Human Faces by Joint Mesh and Landmark Regression | Even though 3D face reconstruction has achieved impressive progress, most orthogonal projection-based face reconstruction methods can not achieve accurate and consistent reconstruction results when the face is very close to the camera due to the distortion under the perspective projection. In this paper, we propose to simultaneously reconstruct 3D face mesh in the world space and predict 2D face landmarks on the image plane to address the problem of perspective 3D face reconstruction. Based on the predicted 3D vertices and 2D landmarks, the 6DoF (6 Degrees of Freedom) face pose can be easily estimated by the PnP solver to represent perspective projection. Our approach achieves 1st place on the leader-board of the ECCV 2022 WCPA challenge and our model is visually robust under different identities, expressions and poses. The training code and models are released to facilitate future research. | ['Jiankang Deng', 'Alexandros Lattas', 'Jinke Yu', 'Jia Guo'] | 2022-08-15 | null | null | null | null | ['3d-face-reconstruction', 'face-reconstruction'] | ['computer-vision', 'computer-vision'] | [-2.08283458e-02 2.62424767e-01 -4.27435264e-02 -4.48207051e-01
-5.03950238e-01 -5.25020123e-01 4.24512148e-01 -9.90767717e-01
2.80875087e-01 2.64607549e-01 1.73270091e-01 3.28609860e-03
2.44602084e-01 -4.27124918e-01 -6.62819743e-01 -3.22134674e-01
1.09789968e-01 9.10737634e-01 -3.62711519e-01 1.57964323e-02
2.06806019e-01 1.05274796e+00 -1.44067907e+00 6.97723255e-02
-1.24460422e-02 8.92642796e-01 -3.39023799e-01 5.65335333e-01
2.65413597e-02 -1.47336751e-01 -1.92798954e-02 -6.22545838e-01
9.02024865e-01 -1.83308274e-01 -4.86707777e-01 5.28331339e-01
9.45033491e-01 -5.42802274e-01 -2.77143717e-01 9.93037462e-01
6.40729725e-01 -2.69326776e-01 4.17189211e-01 -1.30973125e+00
-6.78269714e-02 -3.58351171e-01 -1.19345224e+00 -4.53115970e-01
1.08379114e+00 -2.96556950e-01 3.70317310e-01 -1.62060654e+00
1.11798823e+00 1.67465556e+00 9.18223023e-01 9.26079392e-01
-1.04090941e+00 -7.40071595e-01 7.59439096e-02 -1.09230697e-01
-1.69531679e+00 -8.97081375e-01 9.76032853e-01 -1.24663390e-01
6.30283892e-01 3.79239738e-01 8.54982615e-01 1.00168478e+00
-6.42123967e-02 3.03566039e-01 1.03829467e+00 -3.99667084e-01
-7.40456507e-02 -1.51967639e-02 -3.96928966e-01 9.17431712e-01
-2.59537157e-02 1.35996401e-01 -7.82116294e-01 -5.97224653e-01
1.14008474e+00 -5.30970432e-02 -2.66083896e-01 -7.86912084e-01
-5.82012177e-01 5.53367317e-01 4.72009182e-02 -3.40668708e-01
-8.23286250e-02 -1.36766553e-01 -2.16927931e-01 -2.14598402e-02
5.67131162e-01 -2.14808419e-01 -3.64890814e-01 4.10558134e-02
-9.17277575e-01 3.78075004e-01 9.61789131e-01 1.27468622e+00
5.45772672e-01 7.39497691e-02 4.80656266e-01 6.30506992e-01
8.82843196e-01 7.93133974e-01 -3.18821371e-01 -1.51815164e+00
2.43359059e-01 4.30740446e-01 3.35085467e-02 -1.23265195e+00
-4.15302694e-01 -1.54838353e-01 -5.26422262e-01 4.11702633e-01
3.14716905e-01 -1.38639569e-01 -7.63584316e-01 1.41068232e+00
9.27092433e-01 4.23426330e-01 -1.94703817e-01 1.00959253e+00
8.68549228e-01 5.17158508e-01 -8.21538329e-01 -4.80843842e-01
1.36455119e+00 -5.74084342e-01 -5.93612194e-01 -2.80668974e-01
-2.44310200e-02 -9.97070968e-01 4.10877913e-01 3.79256397e-01
-1.38811803e+00 -1.73082098e-01 -9.04992461e-01 -1.70414284e-01
3.63701075e-01 -2.93934382e-02 3.55366349e-01 8.32523882e-01
-1.32627761e+00 2.05968738e-01 -8.36147845e-01 -3.23778600e-01
5.17069221e-01 7.16535628e-01 -8.97805929e-01 -3.87004554e-01
-3.88305396e-01 9.00375247e-01 -4.26385194e-01 2.98415512e-01
-6.77166641e-01 -7.61240065e-01 -6.99932098e-01 -1.61073491e-01
2.69981325e-01 -7.10617661e-01 1.00112522e+00 -5.92293561e-01
-1.79405439e+00 1.39126229e+00 -6.74900293e-01 1.72154933e-01
7.60090232e-01 4.81713526e-02 -9.20854285e-02 1.85583323e-01
-1.43683240e-01 7.34720826e-01 9.31531906e-01 -1.60750425e+00
-4.53102291e-02 -1.18199956e+00 -2.25227475e-01 5.87841511e-01
4.78964634e-02 2.43105739e-01 -1.13667691e+00 -1.02629103e-01
7.69921303e-01 -1.23896480e+00 -1.15529403e-01 7.34107196e-01
-3.87661994e-01 3.10211420e-01 1.24607038e+00 -8.35003376e-01
3.30152333e-01 -2.06095648e+00 2.29839072e-01 3.94666016e-01
6.64698556e-02 9.04253721e-02 6.96581900e-02 1.53651699e-01
-1.83295682e-01 -9.45669413e-02 4.83992882e-02 -8.92073870e-01
-2.26208344e-01 4.60526086e-02 -1.20933145e-01 9.41490591e-01
-3.78089309e-01 4.43648696e-01 -4.27991688e-01 -4.79450107e-01
5.24467900e-02 1.19704545e+00 -7.89809704e-01 -2.90626604e-02
1.33485824e-01 7.60635555e-01 -3.24410588e-01 7.91293263e-01
1.55324900e+00 1.64847001e-01 4.81922954e-01 -2.39431038e-01
1.96089540e-02 -1.09652482e-01 -1.39789009e+00 1.79583526e+00
-3.81422877e-01 4.11089301e-01 6.70294344e-01 -2.99700171e-01
1.01195037e+00 5.15791774e-01 6.94157243e-01 -3.95455122e-01
5.48592843e-02 -3.83818932e-02 -3.88061702e-01 3.84253934e-02
1.23221546e-01 -3.17628354e-01 4.40639883e-01 3.29720408e-01
-1.78783268e-01 -4.43291485e-01 -5.30461490e-01 -1.00088507e-01
4.93737072e-01 3.68722379e-01 6.02642819e-02 -2.20496118e-01
5.30697584e-01 -4.22601968e-01 7.12264299e-01 -2.66975313e-01
-2.98087392e-02 1.08610308e+00 5.74727178e-01 -5.66208601e-01
-8.83347809e-01 -1.11999941e+00 -4.98135537e-01 1.97966710e-01
3.29539813e-02 -4.74869132e-01 -8.79302800e-01 -6.71396673e-01
-1.17176706e-02 3.70490342e-01 -3.58665764e-01 3.14789683e-01
-8.31919670e-01 -4.30856377e-01 2.56058365e-01 2.68752366e-01
2.69935250e-01 -4.31741744e-01 -2.14341655e-01 -4.74783212e-01
-4.28198874e-02 -1.25189745e+00 -6.49963856e-01 -4.98626173e-01
-8.92559171e-01 -1.17011559e+00 -6.41506910e-01 -8.17171335e-01
1.29325497e+00 3.39855522e-01 7.69854665e-01 1.12053253e-01
-1.57316238e-01 5.18772304e-01 2.59536028e-01 -3.53233665e-01
2.38802023e-02 -6.51870906e-01 4.57371444e-01 8.00219402e-02
2.03055546e-01 -6.65561497e-01 -4.83397275e-01 6.88039184e-01
-7.68480673e-02 1.41703025e-01 -3.29669565e-01 4.81406987e-01
9.88807917e-01 -7.74611160e-02 -1.64832711e-01 -6.81165516e-01
-1.33789510e-01 -9.82382968e-02 -9.66461301e-01 -6.66247085e-02
-2.52635539e-01 -4.17957902e-01 3.53046179e-01 -4.30866219e-02
-1.20373034e+00 6.75911605e-01 -3.62922013e-01 -8.89225364e-01
1.19606405e-01 -1.47849038e-01 -5.73948145e-01 -5.78867555e-01
2.29215816e-01 -1.49761671e-02 2.72820503e-01 -6.72558129e-01
9.36813205e-02 3.23672235e-01 1.65517598e-01 -4.27898794e-01
1.03446555e+00 1.01046395e+00 5.82001030e-01 -9.26656127e-01
-6.90382361e-01 -1.46086607e-02 -9.45441604e-01 -3.95728618e-01
5.53784132e-01 -1.03177786e+00 -1.02140129e+00 3.09630662e-01
-1.39009821e+00 2.56460875e-01 1.62056182e-02 2.82312036e-01
-6.50099218e-01 4.57158566e-01 -2.91265309e-01 -9.32569444e-01
-3.05139571e-01 -1.28975666e+00 1.41805470e+00 1.97411284e-01
-4.26043011e-02 -7.31876850e-01 -3.39579396e-02 4.72057551e-01
3.52537222e-02 3.44591022e-01 5.57492495e-01 1.97966874e-01
-5.33352852e-01 -3.70409548e-01 5.69613650e-02 -9.90528464e-02
-1.68836206e-01 2.11973593e-01 -1.07659817e+00 -4.37899858e-01
4.96248513e-01 1.58626705e-01 4.95492704e-02 5.45470476e-01
1.01656377e+00 -2.39244223e-01 -5.71115613e-01 1.24597776e+00
1.25772047e+00 3.73755097e-02 6.01655662e-01 -3.92005146e-01
7.14809895e-01 7.86636233e-01 4.03919369e-01 5.02236247e-01
5.14431179e-01 1.19058204e+00 7.13559926e-01 1.67488545e-01
-3.18374455e-01 -6.51139319e-01 2.49951646e-01 5.96087217e-01
-3.01980823e-01 1.89873099e-01 -8.82114828e-01 -1.21162146e-01
-1.24024200e+00 -7.85991967e-01 -3.21861804e-01 2.37648988e+00
2.94797987e-01 -3.37067157e-01 9.17040259e-02 -1.19710281e-01
5.81309736e-01 -4.12552021e-02 -4.41421181e-01 -3.45140457e-01
6.57388046e-02 2.88433693e-02 2.69234627e-01 8.56363058e-01
-5.27261198e-01 8.45767260e-01 7.37885141e+00 3.71763170e-01
-1.11525226e+00 1.16660200e-01 5.79904020e-01 -4.00772721e-01
-3.45028371e-01 1.94877759e-01 -1.30527663e+00 2.02217638e-01
4.45756078e-01 -3.77588905e-02 6.71759903e-01 7.39516914e-01
1.58593059e-02 1.35454208e-01 -1.08613014e+00 1.35717595e+00
5.90198517e-01 -1.23848712e+00 -8.25506598e-02 6.20202899e-01
7.49154925e-01 -7.55736306e-02 3.70383829e-01 -2.40371406e-01
-2.36101136e-01 -1.20656180e+00 5.66010773e-01 4.43971217e-01
1.18680191e+00 -8.91152799e-01 1.66653544e-01 5.96860886e-01
-9.56567347e-01 2.63447195e-01 -4.84881848e-01 -3.27916816e-02
3.41271162e-01 1.72605604e-01 -9.85189199e-01 4.16332036e-01
5.67640960e-01 5.37612677e-01 -1.33461773e-01 7.52248287e-01
2.38404591e-02 6.72797263e-02 -6.43047512e-01 6.00573719e-01
-4.60025489e-01 -5.30178547e-01 7.73589551e-01 4.32305783e-01
7.57909000e-01 3.26866835e-01 1.33212674e-02 6.74120724e-01
-2.39230067e-01 -1.01684462e-02 -8.77929032e-01 4.96139109e-01
3.69224817e-01 1.25206602e+00 -7.37346947e-01 4.03912604e-01
-3.80635828e-01 9.35397804e-01 1.10226244e-01 1.91122070e-01
-6.63794875e-01 6.18332446e-01 9.20937181e-01 5.40845454e-01
1.09910682e-01 -3.79775017e-01 -4.14932340e-01 -1.24877870e+00
2.19762996e-01 -7.76684463e-01 -1.02866955e-01 -8.29846442e-01
-8.75489175e-01 6.54518783e-01 -1.84676498e-02 -1.22870648e+00
-1.01603627e-01 -8.48332345e-01 -5.21567464e-01 8.76030922e-01
-1.10133684e+00 -1.30689430e+00 -6.29335344e-02 6.93682849e-01
3.61445308e-01 -1.28473431e-01 1.21974754e+00 3.02695364e-01
-4.71455365e-01 7.81101942e-01 -2.17666790e-01 -1.20634690e-01
4.78577107e-01 -6.90652728e-01 4.91369247e-01 5.50516963e-01
2.51100630e-01 5.61954737e-01 6.54789627e-01 -5.88541985e-01
-2.24539208e+00 -5.97969890e-01 8.36988807e-01 -9.43424940e-01
-9.57724452e-02 -6.70032918e-01 -3.68461847e-01 1.07458222e+00
-1.73865214e-01 5.62727690e-01 4.63408709e-01 1.29394248e-01
-3.60779405e-01 -1.91712920e-02 -1.58875084e+00 5.46790600e-01
1.39979982e+00 -4.31212485e-01 4.94402880e-03 5.17819226e-01
8.80798474e-02 -1.20328498e+00 -6.58436477e-01 2.99510866e-01
8.27731729e-01 -9.30101216e-01 1.32178402e+00 -2.07842514e-01
-9.11933463e-03 -3.67177129e-01 -4.33011472e-01 -9.54250634e-01
1.19769156e-01 -8.43485832e-01 -2.58154273e-01 9.55223143e-01
7.48260692e-02 -3.93858612e-01 1.37583172e+00 6.31015241e-01
1.69888690e-01 -8.99947286e-01 -1.45547390e+00 -4.72078055e-01
-1.23471431e-01 -3.15865427e-01 6.92766070e-01 6.92920566e-01
-1.40730411e-01 2.52604753e-01 -6.55667007e-01 5.24194002e-01
8.05548131e-01 1.66630328e-01 1.07961738e+00 -1.32282782e+00
-1.14417464e-01 3.40920873e-02 -5.68203926e-01 -1.12411332e+00
4.73998159e-01 -1.01366889e+00 -3.45712781e-01 -1.15035069e+00
3.20976108e-01 -3.50396395e-01 6.87095463e-01 1.87533095e-01
3.73816252e-01 5.94621539e-01 3.12712997e-01 -8.69195908e-02
-8.39759782e-02 3.36657137e-01 1.42575467e+00 2.67733961e-01
1.31733060e-01 2.57740617e-01 -6.42968714e-01 1.15212142e+00
2.31391817e-01 -3.58019322e-01 -3.52600127e-01 -7.19040930e-01
1.12734370e-01 5.91147065e-01 2.02165484e-01 -5.89001894e-01
6.93579465e-02 -6.77263662e-02 9.21730518e-01 -7.98528016e-01
1.27492070e+00 -1.22851133e+00 7.82821774e-01 2.41260916e-01
3.01857710e-01 2.52429456e-01 1.52587444e-01 2.29471549e-01
1.67797223e-01 -2.07444325e-01 8.52710962e-01 -1.47447079e-01
1.67548284e-02 8.27837408e-01 1.90241352e-01 -1.97247952e-01
1.11198545e+00 -5.26911557e-01 8.31063315e-02 -6.24482810e-01
-7.86318421e-01 7.76702240e-02 1.11539614e+00 2.23997220e-01
9.67390537e-01 -1.46105266e+00 -8.61205161e-01 8.52811754e-01
-3.73008728e-01 7.67904744e-02 4.22999412e-01 5.98518074e-01
-8.58304739e-01 1.88477978e-01 -1.98927343e-01 -7.54663348e-01
-1.84230125e+00 4.13186938e-01 5.83856523e-01 1.77744061e-01
-7.99922943e-01 9.61882830e-01 1.44780934e-01 -8.39013100e-01
1.76775351e-01 6.03464186e-01 6.72964081e-02 -2.46456489e-01
7.08112001e-01 4.41049784e-01 8.54078233e-02 -1.33775353e+00
-6.12895608e-01 1.32226396e+00 1.61606669e-01 -2.27042750e-01
1.33658719e+00 -8.75503644e-02 -1.90253913e-01 -1.69010878e-01
1.40638471e+00 4.99503672e-01 -1.39217186e+00 1.81672215e-01
-4.94358093e-01 -1.03448439e+00 -3.76809426e-02 -4.19600457e-01
-1.50526977e+00 1.09021235e+00 5.68136752e-01 -5.26509583e-01
9.08732772e-01 -1.40198335e-01 5.82655311e-01 -6.52030706e-02
7.38840520e-01 -6.64592147e-01 -3.49754870e-01 5.26697099e-01
1.21255469e+00 -8.48500907e-01 3.96947831e-01 -1.13732457e+00
-4.21961904e-01 1.22689950e+00 7.25620627e-01 -1.18629046e-01
9.65128899e-01 5.49369335e-01 1.07246481e-01 -3.00759226e-01
-4.82050955e-01 6.62813067e-01 3.37336987e-01 7.97496200e-01
2.98753411e-01 -2.33825631e-02 9.13241059e-02 9.94320065e-02
-4.75371331e-01 -2.38955960e-01 3.54922444e-01 6.07602715e-01
1.68762669e-01 -1.22124219e+00 -7.23300338e-01 -4.63025458e-02
-4.02823985e-01 4.00354832e-01 -4.74163264e-01 5.51450729e-01
3.84598151e-02 6.05271518e-01 1.61902323e-01 -1.04229130e-01
4.28482413e-01 5.03663719e-02 1.03021860e+00 -6.29288912e-01
-2.01884191e-02 3.03070575e-01 3.61168161e-02 -7.73029387e-01
-8.38044956e-02 -9.60164249e-01 -1.33590031e+00 -6.13173425e-01
-2.45451570e-01 -2.04931200e-01 9.85268831e-01 5.61968088e-01
6.53784752e-01 -4.13812935e-01 8.72600496e-01 -1.35701692e+00
-4.30307150e-01 -4.15523767e-01 -6.49820089e-01 3.24792005e-02
9.96724889e-02 -7.33089864e-01 -3.34491521e-01 -1.51443064e-01] | [13.17502212524414, 0.020670443773269653] |
cb8d1482-a9e2-4beb-ad3d-5e9dbb22bc2a | better-datastore-better-translation | 2212.08822 | null | https://arxiv.org/abs/2212.08822v1 | https://arxiv.org/pdf/2212.08822v1.pdf | Better Datastore, Better Translation: Generating Datastores from Pre-Trained Models for Nearest Neural Machine Translation | Nearest Neighbor Machine Translation (kNNMT) is a simple and effective method of augmenting neural machine translation (NMT) with a token-level nearest neighbor retrieval mechanism. The effectiveness of kNNMT directly depends on the quality of retrieved neighbors. However, original kNNMT builds datastores based on representations from NMT models, which would result in poor retrieval accuracy when NMT models are not good enough, leading to sub-optimal translation performance. In this paper, we propose PRED, a framework that leverages Pre-trained models for Datastores in kNN-MT. Better representations from pre-trained models allow us to build datastores of better quality. We also design a novel contrastive alignment objective to mitigate the representation gap between the NMT model and pre-trained models, enabling the NMT model to retrieve from better datastores. We conduct extensive experiments on both bilingual and multilingual translation benchmarks, including WMT17 English $\leftrightarrow$ Chinese, WMT14 English $\leftrightarrow$ German, IWSLT14 German $\leftrightarrow$ English, and IWSLT14 multilingual datasets. Empirical results demonstrate the effectiveness of PRED. | ['ShuJian Huang', 'Mingxuan Wang', 'Zewei Sun', 'Shanbo Cheng', 'Jiahuan Li'] | 2022-12-17 | null | null | null | null | ['nmt'] | ['computer-code'] | [ 1.09648444e-01 -1.94716215e-01 -7.22473323e-01 -2.55751610e-01
-1.36604261e+00 -7.15017200e-01 7.88965404e-01 -1.07143089e-01
-6.36493027e-01 8.98826540e-01 6.29615366e-01 -6.75371230e-01
9.03987437e-02 -7.52709091e-01 -1.00441909e+00 -2.30735973e-01
3.06759089e-01 8.06900799e-01 -3.06093335e-01 -7.81428993e-01
2.51570612e-01 -1.96874719e-02 -7.68573225e-01 6.25429034e-01
1.11914337e+00 4.90626693e-01 3.34583759e-01 2.64875293e-01
-2.26878494e-01 4.16368723e-01 -4.26132649e-01 -6.18757010e-01
5.28390646e-01 -5.74118793e-01 -7.70186365e-01 -7.20857382e-01
5.70611477e-01 -3.59197766e-01 -3.78118157e-01 7.98630059e-01
7.00442076e-01 -5.22917844e-02 7.62645900e-01 -8.37674081e-01
-1.36582482e+00 1.28832161e+00 -3.88786435e-01 2.34092981e-01
1.78321078e-01 -5.36296470e-03 1.22415519e+00 -1.56859732e+00
9.18975413e-01 1.09562540e+00 5.31396627e-01 4.84930962e-01
-1.24948931e+00 -8.47485006e-01 2.71414127e-02 1.08463541e-01
-1.44197941e+00 -7.56615639e-01 1.87812522e-01 1.10793091e-01
1.42096591e+00 2.03215137e-01 4.19068217e-01 1.12532020e+00
3.18959892e-01 7.43121982e-01 1.17563355e+00 -6.33938134e-01
-1.97747409e-01 5.69357537e-03 -1.05992630e-01 4.23194259e-01
1.27083585e-01 9.30718184e-02 -9.55524802e-01 -1.37469828e-01
7.48785198e-01 -1.38147807e-04 -3.87579620e-01 -9.70022678e-02
-1.66514575e+00 7.98178315e-01 6.37438715e-01 6.36706054e-01
-3.73351753e-01 7.27287009e-02 3.24656963e-01 7.91437089e-01
6.52529180e-01 8.18566322e-01 -5.99684656e-01 -1.68413997e-01
-1.03808331e+00 2.95982044e-02 3.87525201e-01 1.37548399e+00
9.29580927e-01 -1.50023565e-01 -2.75669515e-01 1.08252466e+00
-1.21699445e-01 9.94239628e-01 7.53633082e-01 -7.49262393e-01
1.17882943e+00 4.18354958e-01 -3.95101830e-02 -6.06508613e-01
1.11037575e-01 -6.42429113e-01 -8.27291429e-01 -5.42909086e-01
2.67506763e-02 1.47306249e-01 -7.85562634e-01 1.78350019e+00
-1.62900284e-01 -2.24329829e-01 4.61188436e-01 1.01181507e+00
5.66012859e-01 1.00562787e+00 -3.38955581e-01 -1.37581810e-01
9.61009145e-01 -1.32025743e+00 -6.01947725e-01 -4.55404997e-01
1.30291224e+00 -1.36425865e+00 1.26819730e+00 -1.72800019e-01
-1.19938481e+00 -3.82196844e-01 -9.00010288e-01 -2.56426275e-01
-3.59553218e-01 2.73501009e-01 2.95769066e-01 6.65017739e-02
-1.24138689e+00 7.39926875e-01 -6.69706404e-01 -4.11954045e-01
3.46866250e-02 2.80404717e-01 -5.67293525e-01 -6.35768235e-01
-1.46135688e+00 1.24563372e+00 2.46584490e-01 7.46480748e-02
-5.89618385e-01 -7.64679611e-01 -6.36336923e-01 -3.03707600e-01
6.63905591e-02 -7.52050936e-01 1.16327000e+00 -7.53346682e-01
-1.31461227e+00 6.87068343e-01 -3.88502330e-01 -3.86462957e-01
2.02543527e-01 -3.24595243e-01 -3.82172018e-01 -2.86451817e-01
4.90967631e-01 9.29503143e-01 5.49900174e-01 -9.18022037e-01
-3.71009916e-01 -3.28381687e-01 -1.94234729e-01 7.01081097e-01
-4.59799320e-01 -7.79539943e-02 -6.13436699e-01 -1.01626742e+00
5.24659455e-01 -1.20215571e+00 -5.34937084e-02 -4.75994229e-01
-5.21951437e-01 -1.80686742e-01 3.07648987e-01 -7.50835180e-01
1.08941901e+00 -1.93555605e+00 3.80942106e-01 1.66579220e-03
-5.53404726e-02 1.94099873e-01 -7.97214091e-01 7.67121136e-01
1.73969090e-01 2.54881203e-01 -1.99731477e-02 -2.16478869e-01
4.47047455e-03 3.90861303e-01 -5.47520220e-01 1.25713855e-01
1.78545088e-01 1.27830172e+00 -8.46828699e-01 -4.25024807e-01
-3.57078373e-01 2.58010924e-01 -4.28338051e-01 3.41601074e-02
-8.29555765e-02 2.66657352e-01 -3.02969307e-01 7.81697989e-01
4.31598574e-01 -1.73215017e-01 1.53392091e-01 -1.38457835e-01
4.56345379e-02 9.96753216e-01 -3.90712380e-01 2.17248607e+00
-7.59613395e-01 5.06880045e-01 -3.65220994e-01 -4.88539636e-01
9.18083847e-01 2.67690927e-01 2.11628154e-01 -1.28301644e+00
-1.74538389e-01 9.39864993e-01 -7.22712204e-02 1.67681083e-01
1.04291189e+00 -1.74280852e-02 -1.32062718e-01 8.35025907e-01
-6.57302588e-02 -2.02170610e-01 1.42619997e-01 3.02626222e-01
1.04761910e+00 1.99355423e-01 1.21528499e-01 -3.16738248e-01
-1.34478584e-02 3.13858271e-01 6.18636131e-01 7.38840401e-01
1.63772300e-01 5.38968801e-01 -1.83050901e-01 -4.08520252e-01
-1.35728800e+00 -1.00513780e+00 1.07016936e-01 1.31449461e+00
1.82129797e-02 -4.99899089e-01 -6.92395270e-01 -8.04878414e-01
-1.46150038e-01 7.17172623e-01 -3.01203579e-01 -4.19199556e-01
-1.31093276e+00 -6.17343426e-01 7.79885828e-01 4.65278387e-01
4.48322773e-01 -8.38692009e-01 1.77007943e-01 2.23902866e-01
-8.71301055e-01 -1.11588085e+00 -1.12296891e+00 1.95054114e-01
-1.24474573e+00 -3.66513819e-01 -9.71747458e-01 -9.57195461e-01
6.85025096e-01 5.34949481e-01 1.62008464e+00 8.66104811e-02
4.44524646e-01 -2.73507476e-01 -5.33165574e-01 7.36359805e-02
-5.99572957e-01 7.80387461e-01 3.98991495e-01 -7.01762319e-01
5.20401597e-01 -4.33291405e-01 -5.00250220e-01 5.92103541e-01
-7.48607397e-01 1.36809975e-01 1.21445787e+00 1.08309400e+00
8.13504636e-01 -5.70191026e-01 5.01434982e-01 -7.30882585e-01
6.28340542e-01 -4.35322970e-01 -1.61380768e-01 4.85408992e-01
-1.01234674e+00 3.21742535e-01 6.93469703e-01 -5.85135698e-01
-4.69355524e-01 -6.01971865e-01 7.86814764e-02 -5.18450499e-01
6.38739288e-01 7.66856611e-01 6.21660724e-02 2.06552222e-01
8.98915827e-01 5.27464807e-01 -1.35098860e-01 -5.92847228e-01
6.51473343e-01 6.84979498e-01 4.26625729e-01 -7.40297258e-01
8.23601484e-01 -1.02728881e-01 -4.30209577e-01 -1.50974929e-01
-4.32848662e-01 -1.10105626e-01 -5.27526975e-01 3.53318512e-01
4.22256291e-01 -1.13199723e+00 2.96274871e-01 1.15526520e-01
-1.25375354e+00 -3.92486513e-01 -1.33603334e-01 6.31974399e-01
-4.81102645e-01 -6.84572682e-02 -9.50988531e-01 -5.64090572e-02
-8.32671225e-01 -1.24388278e+00 1.17303252e+00 -3.11913282e-01
-3.57409775e-01 -6.38255596e-01 8.69300663e-02 4.85166669e-01
6.47534370e-01 -5.96761882e-01 1.35250676e+00 -7.55357862e-01
-7.67027438e-01 9.29757282e-02 -1.90900847e-01 1.80934966e-01
1.89207539e-01 -5.22725761e-01 -4.14024293e-01 -6.33436799e-01
-2.73538142e-01 -4.25095439e-01 8.27460945e-01 -8.16018283e-02
2.77568609e-01 -6.25074387e-01 -3.55987370e-01 6.03491604e-01
1.33741629e+00 2.20040884e-02 6.12106919e-01 5.53108931e-01
5.88120818e-01 4.12401170e-01 7.76357651e-01 -2.00749651e-01
6.07803464e-01 9.71352994e-01 9.86476913e-02 -1.75724208e-01
-4.86199945e-01 -6.83298409e-01 7.44981289e-01 1.74962342e+00
6.02960661e-02 -1.57682344e-01 -9.34642315e-01 5.43493986e-01
-1.52838039e+00 -5.16505063e-01 1.87450692e-01 2.10660529e+00
1.33601260e+00 -4.84346710e-02 -3.20879906e-01 -3.72193396e-01
6.31389916e-01 -1.14267267e-01 -5.19666255e-01 -5.12748420e-01
-4.57058191e-01 3.93683970e-01 6.71209812e-01 2.86532581e-01
-5.83273649e-01 1.46843565e+00 5.77079439e+00 1.07684803e+00
-1.07700217e+00 3.13578457e-01 5.70049345e-01 -2.42449194e-01
-6.67769432e-01 1.43348068e-01 -9.52589452e-01 4.73005623e-01
1.01832664e+00 -1.00444108e-01 8.22272182e-01 4.92457569e-01
-6.04310408e-02 3.94762605e-01 -1.41514957e+00 8.73708367e-01
5.80797270e-02 -1.61783838e+00 5.93172550e-01 2.58625060e-01
1.08813179e+00 6.03640199e-01 4.28069830e-01 7.48477638e-01
5.38072824e-01 -9.87744570e-01 6.90200686e-01 3.29223454e-01
1.02102304e+00 -7.24573076e-01 7.12184668e-01 4.14517224e-01
-9.52933848e-01 1.39808074e-01 -6.58242643e-01 2.22630575e-01
-1.60324007e-01 5.02815723e-01 -1.12707162e+00 7.61567295e-01
7.42361426e-01 7.16691375e-01 -3.80405456e-01 2.68915474e-01
-1.33379132e-01 3.52291256e-01 -2.59028196e-01 1.67731717e-02
5.21330595e-01 -3.26069087e-01 5.12297392e-01 1.00206053e+00
8.40968966e-01 -2.53839254e-01 1.02888867e-02 7.00916409e-01
-6.09752834e-01 5.53045750e-01 -9.63467836e-01 -2.63748497e-01
9.99713719e-01 8.00487399e-01 -2.03185782e-01 -3.83207589e-01
-3.04117113e-01 1.24285293e+00 6.98087037e-01 4.08889443e-01
-5.60084999e-01 -1.50218487e-01 8.09943557e-01 7.24525005e-02
7.23044276e-02 -3.72146040e-01 -1.67454213e-01 -1.39256322e+00
3.79885197e-01 -1.60134327e+00 1.29654363e-01 -7.41280437e-01
-1.28233504e+00 1.07538152e+00 -2.31119052e-01 -1.51823854e+00
-6.43663406e-01 -2.61228144e-01 -2.07709079e-03 1.19252586e+00
-1.54533923e+00 -1.42758536e+00 5.12739062e-01 4.40611213e-01
6.53262913e-01 -4.52522427e-01 1.01052094e+00 4.17828202e-01
-3.33416015e-01 1.02506387e+00 5.61347604e-01 2.27010950e-01
1.18534446e+00 -9.01458442e-01 8.35047305e-01 8.04757178e-01
5.45918524e-01 1.18939364e+00 3.52408737e-01 -7.66415477e-01
-1.56428385e+00 -1.30191076e+00 1.56647444e+00 -6.27234459e-01
5.17547011e-01 -2.51921773e-01 -7.00730860e-01 7.15895712e-01
5.28368473e-01 -1.88768402e-01 5.13691843e-01 1.94296673e-01
-6.23220086e-01 -1.12495460e-01 -7.48718679e-01 1.05555379e+00
1.32041621e+00 -9.23666835e-01 -5.72140276e-01 4.00710523e-01
1.03160071e+00 -4.46189553e-01 -1.09269655e+00 5.39692819e-01
4.84575510e-01 -4.08361912e-01 9.50394392e-01 -6.20699525e-01
6.07457459e-01 -5.97476885e-02 -7.12542593e-01 -1.76069880e+00
-1.89664230e-01 -5.57150066e-01 1.19433329e-01 1.08416975e+00
9.92762566e-01 -6.76043034e-01 5.07317543e-01 2.34983996e-01
-3.99815232e-01 -8.77158821e-01 -1.24717569e+00 -1.13265133e+00
8.48063409e-01 -1.99589029e-01 1.06951237e+00 1.21810663e+00
6.12895787e-02 5.84282875e-01 -3.41205448e-01 -1.74879372e-01
2.59427339e-01 1.79880962e-01 6.59653842e-01 -5.83768666e-01
-2.15196565e-01 -4.50349391e-01 8.55488256e-02 -1.47573841e+00
1.29540130e-01 -1.40129173e+00 7.11188093e-02 -1.46511281e+00
4.46558475e-01 -8.20085883e-01 -2.98998356e-01 6.58390522e-01
-2.48130128e-01 5.07873654e-01 1.43242657e-01 9.73507226e-01
-2.74167657e-01 5.90739489e-01 1.38907492e+00 -3.73922437e-01
-8.28429684e-02 -4.18510109e-01 -7.15531886e-01 1.20059006e-01
7.79629350e-01 -7.12965786e-01 -3.34715605e-01 -1.44419849e+00
4.42555249e-01 6.97618574e-02 -1.31793365e-01 -5.32565534e-01
3.21282685e-01 8.00698176e-02 2.33388290e-01 -5.96215487e-01
4.45245475e-01 -4.60441351e-01 3.51232546e-03 2.36142755e-01
-6.47708595e-01 8.98453593e-01 -1.26570731e-01 2.48724207e-01
-2.64059961e-01 4.06703278e-02 4.19435740e-01 -1.87712207e-01
-2.41822496e-01 2.34423012e-01 -1.98137999e-01 6.11377284e-02
2.50829369e-01 5.34457192e-02 -5.61672628e-01 -3.69584531e-01
-3.44744176e-01 3.57013419e-02 7.78118551e-01 7.94774354e-01
6.78719163e-01 -1.91493511e+00 -1.08824348e+00 2.05941245e-01
4.03207272e-01 -3.35432708e-01 -2.60490447e-01 9.01404262e-01
-2.02238947e-01 7.82130957e-01 -8.30875933e-02 -4.95524138e-01
-8.98217022e-01 2.55153447e-01 1.67459950e-01 -5.49364030e-01
-2.99120039e-01 7.57443070e-01 -2.07231879e-01 -1.15706754e+00
-1.64054260e-01 -4.53746617e-01 2.22788006e-01 -3.62448424e-01
1.22340262e-01 2.05943331e-01 5.85131288e-01 -8.71173561e-01
-2.78209865e-01 4.11527663e-01 -6.39716923e-01 -4.56908286e-01
9.96115685e-01 -2.52370507e-01 -3.49937528e-01 5.67270219e-02
1.35954046e+00 3.86103317e-02 -3.35551739e-01 -9.52347696e-01
9.85234752e-02 -4.63729709e-01 4.97253314e-02 -9.21473503e-01
-8.36643040e-01 6.87557638e-01 4.22955036e-01 -6.52479291e-01
9.48093891e-01 -1.38593704e-01 1.36863256e+00 9.25919354e-01
9.02949572e-01 -1.13857508e+00 1.53655395e-01 8.81575227e-01
8.88401866e-01 -1.13903451e+00 -2.26214126e-01 -1.04437144e-02
-4.98595595e-01 8.74992192e-01 4.66995537e-01 2.94300735e-01
2.13748857e-01 6.98762685e-02 3.46423090e-01 2.31321380e-01
-9.08605158e-01 1.18222132e-01 4.32670772e-01 3.22135538e-01
7.59795308e-01 1.28299803e-01 -3.42830628e-01 4.32011664e-01
-5.00443935e-01 -3.10976207e-01 3.87979411e-02 1.06002474e+00
-2.43554741e-01 -1.67960167e+00 -2.55000085e-01 4.53060001e-01
-4.16929424e-01 -9.67373133e-01 -6.32060587e-01 6.24462903e-01
-2.79610027e-02 1.00280559e+00 -4.12061065e-02 -6.55181885e-01
1.35600746e-01 1.49962619e-01 3.87679070e-01 -6.50543332e-01
-7.47293115e-01 1.64518505e-01 2.45935544e-01 -5.41699588e-01
6.20234460e-02 -5.28823376e-01 -1.01651514e+00 -6.89257264e-01
-4.61491793e-01 4.95072603e-01 7.65571415e-01 8.69817853e-01
7.46907473e-01 -4.22941297e-02 7.93308794e-01 -3.75633210e-01
-7.62507558e-01 -1.29435444e+00 -3.63576524e-02 1.05804704e-01
4.17961404e-02 -2.37310588e-01 7.64003471e-02 -3.29731286e-01] | [11.67357349395752, 10.198083877563477] |
2c8310ae-9aed-452a-b9c4-39ee149ffe09 | optimizing-kernel-target-alignment-for-cloud | 2306.14515 | null | https://arxiv.org/abs/2306.14515v1 | https://arxiv.org/pdf/2306.14515v1.pdf | Optimizing Kernel-Target Alignment for cloud detection in multispectral satellite images | The optimization of Kernel-Target Alignment (TA) has been recently proposed as a way to reduce the number of hardware resources in quantum classifiers. It allows to exchange highly expressive and costly circuits to moderate size, task oriented ones. In this work we propose a simple toy model to study the optimization landscape of the Kernel-Target Alignment. We find that for underparameterized circuits the optimization landscape possess either many local extrema or becomes flat with narrow global extremum. We find the dependence of the width of the global extremum peak on the amount of data introduced to the model. The experimental study was performed using multispectral satellite data, and we targeted the cloud detection task, being one of the most fundamental and important image analysis tasks in remote sensing. | ['Jakub Nalepa', 'Bertrand Le Saux', 'Bartosz Grabowski', 'Grzegorz Czelusta', 'Filip Szczepanek', 'Jakub Mielczarek', 'Artur Miroszewski'] | 2023-06-26 | null | null | null | null | ['cloud-detection'] | ['computer-vision'] | [ 5.04697382e-01 -2.16772467e-01 -2.69997269e-01 -3.10999900e-01
-6.64037228e-01 -4.87408459e-01 4.43752766e-01 4.75629479e-01
-5.13263643e-01 6.48219287e-01 -5.15949190e-01 -4.29626554e-01
-3.63155633e-01 -9.03970480e-01 -7.06757009e-01 -1.01644981e+00
-7.43697658e-02 2.32172534e-01 1.13137908e-01 -3.42669547e-01
3.71011704e-01 5.00977755e-01 -1.33105755e+00 9.07949135e-02
9.15564954e-01 1.22501707e+00 -6.08054847e-02 5.92213392e-01
3.38888109e-01 1.78264782e-01 -5.02156377e-01 -1.30819485e-01
4.71317917e-01 -3.21156681e-01 -6.63487852e-01 -3.16234171e-01
4.74440932e-01 3.15091014e-01 -4.76295920e-03 1.54787302e+00
5.23889601e-01 -6.59193769e-02 5.84821284e-01 -1.25961423e+00
-5.21121956e-02 8.03712249e-01 -2.95309663e-01 5.45682371e-01
-5.56076646e-01 2.48396005e-02 1.12504268e+00 -1.91308826e-01
5.15557289e-01 7.68382668e-01 3.07195991e-01 2.86800683e-01
-1.47081888e+00 -5.87837100e-01 -4.96749073e-01 3.67666662e-01
-1.63697350e+00 -2.18922451e-01 4.80903178e-01 -2.73939937e-01
8.77941012e-01 2.58312196e-01 5.89701891e-01 5.33190787e-01
4.60252583e-01 1.46711454e-01 1.39507556e+00 -6.34014428e-01
5.17809927e-01 2.71108627e-01 5.66053391e-01 6.79392338e-01
5.33386230e-01 -7.13967311e-04 -3.07367474e-01 -3.70195955e-01
4.14676428e-01 -4.26061064e-01 -2.94465154e-01 -3.53713095e-01
-6.93616033e-01 8.80750537e-01 6.66838467e-01 4.70163971e-01
-6.76708817e-02 4.95202065e-01 7.83941671e-02 4.50055659e-01
3.06155354e-01 6.85237229e-01 -3.63332301e-01 7.16016218e-02
-5.53258538e-01 2.11033478e-01 5.83342195e-01 4.26017255e-01
1.00578904e+00 -2.01053724e-01 -3.63205671e-02 2.25197852e-01
-2.88211733e-01 5.70591986e-01 3.36134851e-01 -7.13322043e-01
3.88650030e-01 5.46185851e-01 8.61494988e-02 -5.03944337e-01
-6.47265375e-01 -7.28558064e-01 -1.01254010e+00 4.32266667e-02
6.19127214e-01 -1.49508476e-01 -2.93187261e-01 1.52877986e+00
1.86278164e-01 -2.99091995e-01 1.38860837e-01 8.73982728e-01
1.61427334e-01 5.58787882e-01 -1.12305291e-01 -1.95574358e-01
1.36891210e+00 -5.29467523e-01 -3.92550230e-01 -9.86218639e-03
8.29874396e-01 -3.68594944e-01 1.10071230e+00 4.10461515e-01
-4.96668220e-01 -2.70337969e-01 -1.17844760e+00 1.15238212e-01
-4.41541672e-01 2.93633312e-01 8.21677327e-01 1.01394022e+00
-6.68414176e-01 9.60076332e-01 -4.92000669e-01 -7.91478679e-02
2.96174526e-01 5.49771845e-01 -3.47236246e-01 2.11131677e-01
-1.17172754e+00 8.34954143e-01 4.65264112e-01 2.81628430e-01
-3.65306914e-01 -2.34103292e-01 -3.77737731e-01 4.29272234e-01
2.13906750e-01 -3.73004287e-01 9.79983091e-01 -1.05107522e+00
-1.56100357e+00 7.92112470e-01 1.99311346e-01 -6.82983994e-01
1.86084539e-01 1.54045045e-01 -1.19045988e-01 -1.19650654e-01
-2.48897925e-01 3.20491672e-01 7.81204700e-01 -3.14916283e-01
-2.53344923e-01 -7.42378056e-01 4.64809826e-03 -2.14379877e-01
-3.84134918e-01 -2.13796988e-01 4.23954844e-01 -6.39829412e-02
5.74101470e-02 -1.01439738e+00 -2.85794228e-01 -4.39533859e-01
-1.54143095e-01 5.03874645e-02 2.82963872e-01 1.01107948e-01
9.43069577e-01 -2.02376318e+00 3.01318049e-01 5.23322999e-01
3.81044410e-02 1.48380280e-01 7.60749429e-02 3.59332502e-01
-4.66957837e-02 1.12486526e-01 -1.86684757e-01 3.68695557e-01
2.50985293e-04 -9.63964239e-02 -4.11155909e-01 7.14442849e-01
1.80464864e-01 7.37006187e-01 -5.25886714e-01 -1.85284421e-01
-1.90492086e-02 1.19853161e-01 -6.72453165e-01 -1.66990444e-01
-2.50476897e-01 1.91744238e-01 -5.30319989e-01 3.24470937e-01
8.16421509e-01 -4.22168612e-01 3.56512934e-01 -2.44028613e-01
-4.78815854e-01 1.86456189e-01 -9.41483438e-01 1.22494256e+00
-2.89857864e-01 7.72892773e-01 -6.58274628e-03 -1.13085437e+00
8.03929746e-01 -7.17800185e-02 2.39358634e-01 -5.56282997e-01
5.62385142e-01 3.94587994e-01 3.19072723e-01 -9.61447954e-02
4.89283085e-01 -2.96602011e-01 -3.73835772e-01 1.06023394e-01
2.79540885e-02 -4.10932362e-01 9.95957702e-02 -2.48597786e-01
8.37964535e-01 -5.40821016e-01 3.83984447e-01 -7.01720059e-01
5.36134124e-01 8.65885466e-02 5.18880114e-02 7.26862848e-01
-6.59133792e-02 7.88011104e-02 1.02156556e+00 -5.16326189e-01
-8.52849722e-01 -5.34382820e-01 -5.52740335e-01 7.12925732e-01
2.05267012e-01 -4.37445372e-01 -1.03157294e+00 -1.01499327e-01
-1.32785335e-01 4.67711180e-01 -4.75567430e-01 -4.00512934e-01
-4.81791824e-01 -1.40030813e+00 6.78566158e-01 -1.57385543e-02
6.37035549e-01 -5.81409097e-01 -1.03303707e+00 -1.04696505e-01
-3.72471511e-02 -1.10592127e+00 -5.67930974e-02 7.96341121e-01
-1.02270079e+00 -8.86687398e-01 -3.52015167e-01 -2.47469470e-01
2.90046334e-01 -1.44343272e-01 7.39595711e-01 -1.87614009e-01
-5.01250148e-01 -1.67352021e-01 -9.06470940e-02 -3.96783441e-01
-3.34491789e-01 6.17179811e-01 -5.41892201e-02 1.55685022e-01
-9.18422639e-02 -3.51531416e-01 -3.65082473e-01 1.98069319e-01
-8.14019740e-01 -2.69506603e-01 6.00196779e-01 8.97981167e-01
5.83941400e-01 1.67529225e-01 2.19143897e-01 -5.62295258e-01
4.90859568e-01 1.14618517e-01 -1.46914828e+00 3.18089277e-01
-5.93788207e-01 6.66509211e-01 8.32900405e-01 -3.02048057e-01
-3.62724662e-01 2.79156804e-01 2.56200463e-01 2.37873383e-03
3.44961733e-01 4.23128784e-01 -4.60690744e-02 -6.83471680e-01
8.32278132e-01 -1.95163290e-03 -3.26301664e-01 -1.90010875e-01
2.08843410e-01 7.70412982e-01 1.82746962e-01 -6.72959983e-01
4.29552794e-01 3.34066838e-01 7.27886498e-01 -1.08630514e+00
-6.76493168e-01 -2.07063526e-01 -3.99619877e-01 -5.32025583e-02
6.31532609e-01 -4.77266520e-01 -1.12024069e+00 4.09846991e-01
-9.29393828e-01 -3.76127571e-01 -3.48418474e-01 4.38309520e-01
-3.99315774e-01 1.09642535e-01 -3.99113148e-01 -8.30515563e-01
-6.11102998e-01 -1.14385223e+00 8.68536353e-01 2.98947364e-01
3.49539340e-01 -5.12070000e-01 -2.24664062e-02 8.03659409e-02
5.53636849e-01 7.70290941e-02 1.25924945e+00 -3.11744273e-01
-7.10105538e-01 -3.55896860e-01 -2.55655348e-01 3.21762353e-01
-4.02211249e-01 -1.26087889e-02 -1.12998271e+00 -1.43109158e-01
1.37435287e-01 -3.15976918e-01 8.75440001e-01 3.65776032e-01
1.07451761e+00 -2.32574083e-02 -1.65532798e-01 5.81138134e-01
1.43764770e+00 -2.51058578e-01 5.48428595e-01 1.19525366e-01
1.28879011e-01 2.98808575e-01 4.81749386e-01 3.18270057e-01
-8.84307176e-02 8.77503633e-01 4.00827020e-01 3.52241963e-01
4.82342541e-01 1.66097984e-01 3.59984249e-01 6.40685737e-01
-1.54792488e-01 -1.06872000e-01 -9.33244705e-01 1.06388420e-01
-1.67279899e+00 -7.85994411e-01 -3.27348202e-01 2.56085324e+00
4.74818051e-01 2.52770334e-01 -1.79472759e-01 1.25486121e-01
5.85757911e-01 -2.07836200e-02 -1.30509660e-01 -5.89037418e-01
-2.34015152e-01 6.56409085e-01 9.45566952e-01 6.15112543e-01
-1.07160163e+00 9.20043766e-01 6.34383059e+00 8.44050586e-01
-1.59836054e+00 1.81584775e-01 5.43482065e-01 1.04279891e-01
4.04160395e-02 2.03035429e-01 -7.26525724e-01 2.44835153e-01
1.07077587e+00 -2.00218722e-01 3.99651468e-01 7.41726220e-01
-3.24510694e-01 -4.57004070e-01 -1.04132056e+00 1.03842223e+00
-5.29096067e-01 -1.29918897e+00 -1.33846924e-01 1.67858288e-01
5.31190574e-01 1.51273757e-01 1.98656499e-01 1.65787321e-02
-3.86607826e-01 -9.60138977e-01 5.65628767e-01 2.33946308e-01
7.17513978e-01 -7.76335299e-01 7.43730485e-01 4.62808281e-01
-9.80482280e-01 -3.25414538e-01 -5.83351791e-01 -3.38263839e-01
-3.96667898e-01 6.75786853e-01 -5.82453728e-01 3.06813955e-01
4.93210644e-01 7.60672754e-03 -6.07685745e-01 9.07399356e-01
-1.89980131e-03 3.62533420e-01 -6.75509095e-01 -6.60194874e-01
1.97070599e-01 -7.26129472e-01 4.32705998e-01 7.73882031e-01
3.56512636e-01 1.32952139e-01 -2.61507243e-01 8.72848809e-01
-7.01864362e-02 2.42155164e-01 -4.80638713e-01 -1.80684954e-01
1.43512607e-01 1.29414761e+00 -8.49340796e-01 -1.19477861e-01
3.86868775e-01 7.03563273e-01 3.58658761e-01 -1.88165039e-01
-8.68101418e-01 -4.74296868e-01 3.54200453e-01 -3.86479534e-02
8.64138156e-02 -2.88765371e-01 -2.36218184e-01 -1.30093551e+00
4.90290560e-02 -5.42213678e-01 1.88896179e-01 -2.89952487e-01
-6.67837501e-01 5.26334941e-01 7.05509307e-03 -9.20456052e-01
3.38300765e-02 -7.47333467e-01 -4.28404033e-01 5.77178955e-01
-1.28421307e+00 -5.53164184e-01 -2.33192801e-01 3.72344971e-01
-3.00335795e-01 9.81712639e-02 1.01575458e+00 3.40214610e-01
-7.37526894e-01 6.31200016e-01 4.62162137e-01 -2.32615337e-01
3.19499731e-01 -9.35426831e-01 5.59341051e-02 5.40160835e-01
1.77856386e-01 2.98959225e-01 8.09265852e-01 -2.33506262e-01
-1.63523567e+00 -6.15300715e-01 9.06169236e-01 -1.64473042e-01
7.33496130e-01 -5.69352210e-01 -3.82462114e-01 1.95877075e-01
-7.88655430e-02 1.74105316e-01 3.69679570e-01 -1.38012692e-01
-2.58352846e-01 -5.89218855e-01 -9.62701440e-01 2.27994204e-01
4.58587140e-01 -6.62769318e-01 2.02284455e-01 5.68475842e-01
2.16319829e-01 -2.55534440e-01 -7.40339458e-01 3.25977743e-01
5.23591816e-01 -8.76039445e-01 5.34108222e-01 -4.83348519e-01
8.89629200e-02 -1.23229995e-01 -4.42830771e-01 -1.11652255e+00
3.13040540e-02 -8.43973458e-01 2.53091276e-01 5.56897938e-01
4.08097863e-01 -8.14576447e-01 5.95951438e-01 1.26090914e-01
2.00592101e-01 -6.42233968e-01 -1.31606781e+00 -7.92515516e-01
1.68059349e-01 -1.50159061e-01 1.76415965e-01 5.64386249e-01
9.96745005e-02 6.16755664e-01 -6.26935586e-02 3.34812760e-01
6.26061440e-01 4.81756866e-01 5.15031695e-01 -1.16419947e+00
-3.72699767e-01 -5.90830326e-01 -6.10788226e-01 -6.00959837e-01
-7.75903836e-02 -1.01304197e+00 -1.03377163e-01 -4.51425612e-01
1.40046448e-01 -5.72528183e-01 -2.31178075e-01 1.59363523e-01
2.29357541e-01 1.63923547e-01 1.35153353e-01 -2.60624569e-02
-3.60634506e-01 3.22312266e-01 8.11121762e-01 -2.16582045e-01
-1.57787174e-01 1.88351572e-01 -1.90994740e-01 4.16925251e-01
9.31147993e-01 -8.48076046e-01 2.91462075e-02 -1.97723091e-01
6.98413670e-01 8.21567401e-02 4.67250049e-01 -1.15519786e+00
9.94232893e-02 1.54360617e-02 6.32231915e-03 -1.42187238e-01
4.51047391e-01 -7.91949511e-01 7.38939047e-02 9.04827774e-01
-3.57589364e-01 -4.00761515e-02 7.20259547e-02 3.21356267e-01
-1.64210841e-01 -5.07899225e-01 1.00966775e+00 2.19908610e-01
-1.36081621e-01 5.52179441e-02 -3.69958192e-01 -8.97163525e-02
9.83659744e-01 3.28879774e-01 -5.59817076e-01 -6.66790158e-02
-3.91438276e-01 -3.85728478e-02 2.70214111e-01 -3.05355787e-02
7.88852125e-02 -9.03226912e-01 -4.14111346e-01 2.57971525e-01
-3.57743651e-02 -3.98985714e-01 2.00625032e-01 9.57956672e-01
-7.17011690e-01 5.84077656e-01 -4.95664537e-01 -6.63321316e-01
-1.15911973e+00 3.04090470e-01 8.00674856e-01 -3.64165455e-01
-1.16791680e-01 6.15377784e-01 -2.20168412e-01 -2.26648495e-01
-3.82001139e-02 -8.22983682e-01 2.82751005e-02 1.99812114e-01
2.67933398e-01 4.54867780e-01 3.99379015e-01 -3.64493012e-01
-4.02562708e-01 7.80627608e-01 2.72076398e-01 -8.32471699e-02
1.07072556e+00 4.01440829e-01 -6.07010901e-01 4.30733681e-01
1.20508838e+00 -1.75821766e-01 -5.10869682e-01 -2.07783673e-02
1.12587959e-01 -6.40769452e-02 3.19250703e-01 -2.75011241e-01
-7.87065983e-01 1.34320033e+00 1.08618712e+00 3.37446809e-01
9.64160621e-01 -7.43755326e-02 2.92881340e-01 1.07876265e+00
7.32167721e-01 -1.06654871e+00 -2.76367784e-01 6.17109179e-01
4.84481394e-01 -9.74389672e-01 8.09154361e-02 -4.63513911e-01
-2.51214325e-01 1.43245292e+00 3.64270806e-01 -1.11201353e-01
6.20346189e-01 1.38088629e-01 -3.86733800e-01 -2.17511997e-01
-6.31809950e-01 -2.04447478e-01 8.53577703e-02 6.05735974e-03
2.33753882e-02 4.81821060e-01 -4.60954636e-01 4.58468348e-01
-2.57410228e-01 -7.82203600e-02 6.37990296e-01 5.83427072e-01
-5.06520271e-01 -1.23151970e+00 -2.39102751e-01 2.40419701e-01
-4.62857038e-01 -2.03248769e-01 -4.64278221e-01 7.42096841e-01
1.63157478e-01 4.86735553e-01 1.65337160e-01 -3.21918249e-01
4.89401296e-02 1.16377316e-01 9.10296917e-01 -9.38945264e-02
-7.12868392e-01 -2.12592617e-01 -9.38597694e-02 -3.33148420e-01
-2.06618950e-01 -2.35255659e-01 -1.13497090e+00 -2.71044374e-01
-7.70162046e-01 3.36144418e-01 9.01831806e-01 7.93343425e-01
2.80091137e-01 1.58313960e-01 5.70662618e-01 -6.06639624e-01
-9.19463515e-01 -7.00207770e-01 -6.23816490e-01 -1.73184350e-01
4.12416190e-01 -2.63086826e-01 -2.42089435e-01 -6.27599478e-01] | [5.628072738647461, 4.943429946899414] |
3c6e36a0-bb55-40ad-9110-2e8d60f8c6ba | segmental-spatiotemporal-cnns-for-fine | 1602.02995 | null | http://arxiv.org/abs/1602.02995v4 | http://arxiv.org/pdf/1602.02995v4.pdf | Segmental Spatiotemporal CNNs for Fine-grained Action Segmentation | Joint segmentation and classification of fine-grained actions is important
for applications of human-robot interaction, video surveillance, and human
skill evaluation. However, despite substantial recent progress in large-scale
action classification, the performance of state-of-the-art fine-grained action
recognition approaches remains low. We propose a model for action segmentation
which combines low-level spatiotemporal features with a high-level segmental
classifier. Our spatiotemporal CNN is comprised of a spatial component that
uses convolutional filters to capture information about objects and their
relationships, and a temporal component that uses large 1D convolutional
filters to capture information about how object relationships change across
time. These features are used in tandem with a semi-Markov model that models
transitions from one action to another. We introduce an efficient constrained
segmental inference algorithm for this model that is orders of magnitude faster
than the current approach. We highlight the effectiveness of our Segmental
Spatiotemporal CNN on cooking and surgical action datasets for which we observe
substantially improved performance relative to recent baseline methods. | ['Austin Reiter', 'Colin Lea', 'Rene Vidal', 'Gregory D. Hager'] | 2016-02-09 | null | null | null | null | ['fine-grained-action-recognition'] | ['computer-vision'] | [ 4.24298316e-01 -2.65189528e-01 -4.90570366e-01 -4.76911098e-01
-7.49516308e-01 -5.09207845e-01 7.73032844e-01 2.06991687e-01
-5.38048089e-01 4.60075617e-01 5.14797390e-01 -6.95391670e-02
7.16327950e-02 -5.15040696e-01 -8.51212323e-01 -5.04666686e-01
-3.03662151e-01 3.20503354e-01 7.88200140e-01 -2.28410698e-02
1.68291301e-01 5.81162930e-01 -1.51828980e+00 6.66230023e-01
4.63786155e-01 1.27307439e+00 -5.71065322e-02 1.10410738e+00
1.97019294e-01 1.43196487e+00 -3.71486366e-01 2.51789123e-01
3.41873281e-02 -4.63155568e-01 -1.35022891e+00 4.07226115e-01
5.39971292e-01 -6.96462154e-01 -4.01510537e-01 6.61703944e-01
-7.48889446e-02 4.94396329e-01 5.39352655e-01 -1.21153831e+00
-2.26270199e-01 2.32500359e-01 -2.54428804e-01 4.05026317e-01
3.63386333e-01 3.61608207e-01 8.23664784e-01 -1.97783530e-01
7.22844064e-01 1.37479925e+00 7.29158044e-01 3.99724096e-01
-1.06019306e+00 -2.66214669e-01 5.13640285e-01 2.07699165e-01
-8.79845560e-01 -3.12988460e-01 3.13566446e-01 -6.87919796e-01
1.47583759e+00 -2.27220789e-01 9.81279552e-01 1.01259708e+00
4.18087929e-01 1.17218578e+00 7.88914621e-01 -4.56343032e-02
1.98869690e-01 -6.94022954e-01 1.20567419e-01 9.70461965e-01
-4.56017524e-01 -1.08296527e-02 -5.44801712e-01 4.95635718e-02
1.13584685e+00 3.11413407e-01 3.12322527e-01 -5.58591843e-01
-1.44612277e+00 6.14223242e-01 3.24377716e-01 3.22641611e-01
-6.20040059e-01 8.02555859e-01 4.27235484e-01 -1.06950700e-01
3.13938528e-01 1.89435229e-01 -7.26888061e-01 -8.44355941e-01
-1.02627611e+00 5.30683994e-01 6.72440469e-01 9.10430372e-01
5.96699834e-01 -4.35645461e-01 -5.99824190e-01 6.09577656e-01
5.87364584e-02 5.58010712e-02 2.62028992e-01 -1.60722947e+00
1.62869602e-01 8.20591450e-01 2.73573995e-01 -6.49464667e-01
-5.30390203e-01 8.28266740e-02 -3.38563085e-01 1.12268120e-01
5.13620853e-01 9.60565731e-02 -1.16428697e+00 1.62637281e+00
4.15377080e-01 7.18021095e-01 -2.81367481e-01 6.79921269e-01
4.55729783e-01 3.63973707e-01 4.20621574e-01 1.99200928e-01
1.33415067e+00 -1.37224889e+00 -5.06355107e-01 -2.50162661e-01
8.97395730e-01 -2.16837615e-01 6.62277400e-01 1.59865782e-01
-1.20228422e+00 -6.69418156e-01 -6.75314784e-01 -3.22516412e-01
-3.81131291e-01 1.22063562e-01 7.82736838e-01 8.12795237e-02
-9.44076359e-01 8.83780301e-01 -1.64693141e+00 -7.92918682e-01
9.13651407e-01 3.44554305e-01 -4.71916348e-01 -3.33683081e-02
-8.67685854e-01 8.68032455e-01 2.73221582e-01 1.48665281e-02
-1.34538937e+00 -5.00007808e-01 -1.20520353e+00 -2.14911457e-02
3.97422940e-01 -4.17297214e-01 1.77564967e+00 -7.94675589e-01
-1.47890687e+00 7.10690737e-01 -3.38007361e-01 -7.43586779e-01
3.45987737e-01 -4.37810212e-01 -8.98618102e-02 5.19574821e-01
3.45931649e-01 8.33432972e-01 7.06645429e-01 -5.43149769e-01
-1.10436010e+00 -2.80640423e-01 3.69629830e-01 2.43545607e-01
2.62294024e-01 2.36433387e-01 -8.46450448e-01 -5.89296520e-01
-7.62832463e-02 -1.08040273e+00 -5.96970737e-01 2.23299101e-01
-1.30130619e-01 -4.21664208e-01 9.25417960e-01 -4.86912817e-01
9.98356402e-01 -2.10805249e+00 2.47369751e-01 -2.54879415e-01
-3.92244756e-02 8.54726806e-02 -2.27172703e-01 2.53441095e-01
1.15474232e-01 -1.91119388e-01 -2.40642399e-01 -3.45614016e-01
5.88123091e-02 2.95488358e-01 9.22207907e-02 5.03143787e-01
6.17558777e-01 1.25517428e+00 -1.09358144e+00 -5.83782375e-01
5.20400286e-01 3.13909918e-01 -6.08989179e-01 2.19412655e-01
-4.57734942e-01 5.31795502e-01 -7.77643383e-01 8.28281701e-01
7.73488581e-02 -4.64982480e-01 8.59374702e-02 -2.10126549e-01
-1.43908918e-01 3.95678103e-01 -9.21482861e-01 2.14753842e+00
-1.03807792e-01 5.16054213e-01 9.26428065e-02 -1.25712824e+00
2.76801676e-01 2.67393410e-01 9.71560299e-01 -5.46960652e-01
1.71120852e-01 -2.14830697e-01 -3.28653246e-01 -5.41075289e-01
4.92527544e-01 8.55705068e-02 -4.87754822e-01 4.51149940e-01
1.73739821e-01 -3.07026207e-01 3.52310777e-01 2.91394085e-01
1.69838262e+00 9.35440838e-01 3.78403395e-01 -7.82193765e-02
1.52786657e-01 4.06914949e-01 6.37446940e-01 7.63558209e-01
-8.75350058e-01 3.83559823e-01 5.64772487e-01 -6.79163516e-01
-5.35837829e-01 -7.92888045e-01 2.47448266e-01 1.39857912e+00
2.49056518e-01 -3.26004475e-01 -8.83071899e-01 -8.72546554e-01
6.58086315e-02 2.02503279e-01 -9.32682216e-01 -1.51950166e-01
-8.68561387e-01 -1.08131520e-01 5.94332218e-01 1.22373867e+00
7.26792216e-01 -1.43849397e+00 -1.08998013e+00 3.62946481e-01
-2.95537591e-01 -1.44953871e+00 -5.97653508e-01 4.28415000e-01
-9.34374571e-01 -1.39769804e+00 -4.16176945e-01 -6.95294976e-01
2.87343979e-01 8.27265531e-02 1.27820873e+00 -1.29517555e-01
-5.97928643e-01 6.19638264e-01 -4.00797755e-01 -1.71907946e-01
-2.42254138e-01 -9.18513387e-02 -1.69230685e-01 -1.51266083e-01
6.03208065e-01 -2.97344565e-01 -6.66262209e-01 3.62351149e-01
-8.16966414e-01 -4.94453609e-02 6.22291386e-01 6.92704797e-01
6.82848752e-01 2.04211418e-02 6.72844946e-02 -7.57012546e-01
1.48407891e-01 -3.17598104e-01 -4.12128359e-01 2.35264391e-01
1.30722240e-01 -1.04015414e-02 1.19116344e-01 -4.06873256e-01
-1.11975515e+00 5.77022612e-01 8.25528130e-02 -5.42486548e-01
-8.04975331e-01 3.00489396e-01 1.11138970e-01 1.39944553e-01
3.12605441e-01 8.25821385e-02 -1.46631762e-01 -3.73571485e-01
4.57927316e-01 3.42014223e-01 7.41714835e-01 -6.23025656e-01
2.13669017e-01 6.47500396e-01 -2.67278906e-02 -8.01950693e-01
-9.56887484e-01 -6.99952364e-01 -1.36597216e+00 -1.37130722e-01
1.64328647e+00 -1.01767766e+00 -8.09734166e-01 9.05737519e-01
-1.05247486e+00 -1.13501525e+00 -4.16949838e-01 4.74046320e-01
-1.12541831e+00 2.01428607e-01 -1.01695919e+00 -4.82772738e-01
4.00199175e-01 -1.22906780e+00 1.71088243e+00 5.37023731e-02
-4.72599179e-01 -9.91518080e-01 6.03552535e-02 6.43496394e-01
6.76217154e-02 4.32707012e-01 5.04155397e-01 -3.06184232e-01
-8.95202875e-01 -1.59282625e-01 -1.09745704e-01 1.40009999e-01
2.61916161e-01 5.47812181e-03 -4.92127866e-01 -2.76683494e-02
-4.60677654e-01 -6.77488804e-01 1.08247209e+00 8.58287334e-01
1.19922030e+00 9.55901071e-02 -7.10641563e-01 4.69483495e-01
8.73245716e-01 3.16175818e-01 4.79934841e-01 2.21074834e-01
7.84733415e-01 4.06044841e-01 1.08626783e+00 4.34253693e-01
5.85455000e-01 7.58532047e-01 4.17287558e-01 -4.05589975e-02
-1.57044843e-01 -5.75853549e-02 2.99404025e-01 1.38743250e-02
-1.83285788e-01 4.98604476e-02 -7.78961360e-01 9.24636126e-01
-2.27647233e+00 -1.29104435e+00 2.87326843e-01 1.54859710e+00
6.67746484e-01 1.68962002e-01 4.07640308e-01 -1.77885100e-01
2.90966332e-01 4.41523910e-01 -7.92047441e-01 -3.53915870e-01
2.78013885e-01 2.41581991e-01 4.66573298e-01 2.91496962e-01
-1.87135243e+00 1.41509354e+00 7.17895746e+00 4.74839866e-01
-7.29075789e-01 -8.88418555e-02 4.17872876e-01 -9.40099880e-02
3.34025115e-01 -1.46399319e-01 -6.79461062e-01 5.86389527e-02
8.20335388e-01 6.54162586e-01 2.48300597e-01 8.08110058e-01
4.30467159e-01 -4.65232670e-01 -1.39325070e+00 6.57766879e-01
-5.04603088e-02 -1.50292814e+00 -2.79066265e-01 2.76591182e-02
8.40889752e-01 1.53412253e-01 -3.54736477e-01 2.76119560e-01
9.33792472e-01 -1.14379060e+00 7.89243698e-01 6.16022050e-01
5.03777981e-01 -4.85288739e-01 3.13626945e-01 3.51720542e-01
-1.62331128e+00 -2.43745476e-01 2.09981993e-01 -2.43904188e-01
2.46705517e-01 -4.22842279e-02 -3.32147449e-01 1.55380696e-01
9.59694684e-01 1.32883084e+00 -2.91146338e-01 6.40297234e-01
-2.39657894e-01 4.91288751e-01 -1.14205495e-01 2.21388638e-01
6.81952834e-01 6.47206679e-02 1.31713301e-01 1.36517179e+00
-1.08644724e-01 4.78151172e-01 7.03318000e-01 4.48496550e-01
1.55642197e-01 -5.48618555e-01 -4.22184080e-01 -4.57507581e-01
2.02483371e-01 9.88722086e-01 -8.67092431e-01 -5.77886343e-01
-6.18866146e-01 1.29048479e+00 4.99071717e-01 3.24068874e-01
-9.32257235e-01 -2.36503825e-01 1.19454825e+00 -2.07263067e-01
8.07982326e-01 -5.70464671e-01 -2.35117562e-02 -1.08782625e+00
-1.39782816e-01 -8.08929622e-01 5.77692389e-01 -7.01254785e-01
-8.72679591e-01 2.15123132e-01 7.76612386e-02 -9.95220363e-01
-4.67523813e-01 -5.51345825e-01 -3.99160415e-01 5.28397977e-01
-1.18642664e+00 -1.49379945e+00 -3.79630506e-01 9.57115173e-01
9.71442819e-01 2.60896236e-01 8.11416626e-01 -9.96679887e-02
-4.58059698e-01 1.27376467e-01 -2.82705545e-01 5.20708680e-01
4.11730796e-01 -1.18993497e+00 5.63028157e-01 8.93736959e-01
5.03338873e-02 3.59481275e-01 3.98364097e-01 -8.44724119e-01
-1.15460587e+00 -1.32931828e+00 7.04008818e-01 -7.86276639e-01
6.89258575e-01 -2.57009387e-01 -5.35966396e-01 1.16252673e+00
-9.33205783e-02 5.20159304e-01 4.23994780e-01 1.31705850e-01
-3.21810216e-01 1.30925879e-01 -9.00442302e-01 5.13273776e-01
1.28348315e+00 -7.20681131e-01 -6.08017743e-01 2.83924341e-01
6.70484185e-01 -4.80223119e-01 -1.05734324e+00 3.30264956e-01
9.29578006e-01 -9.32048023e-01 9.45890129e-01 -9.82065737e-01
5.65247416e-01 -2.37525105e-01 -2.47037843e-01 -1.02814746e+00
-5.58471143e-01 -4.02701497e-01 -3.83166909e-01 4.37036067e-01
-5.90097234e-02 -8.82335603e-02 1.00073814e+00 6.58110261e-01
-2.80428201e-01 -7.59537339e-01 -6.32232130e-01 -7.28036642e-01
-5.09734638e-02 -6.31873906e-01 2.90614873e-01 5.56899905e-01
1.81296572e-01 1.24599248e-01 -2.04587609e-01 6.99012950e-02
4.32550162e-01 4.02928084e-01 7.15166867e-01 -1.05136967e+00
-1.84413642e-01 -4.53560412e-01 -6.98772907e-01 -1.59833574e+00
4.44485366e-01 -2.02997833e-01 6.31507277e-01 -1.83504975e+00
3.02493781e-01 -1.08410165e-01 -4.09691244e-01 8.92571390e-01
-1.14612103e-01 4.28590089e-01 -3.87079902e-02 4.20410819e-02
-1.21291828e+00 3.52018178e-01 1.22755468e+00 -2.06328809e-01
-2.44189680e-01 -1.66717526e-02 -1.31751135e-01 8.36845577e-01
4.78263140e-01 -2.11986631e-01 -3.63868982e-01 -3.39803249e-01
-5.55348277e-01 3.26460123e-01 5.77711940e-01 -1.18581975e+00
1.94418073e-01 -5.92814207e-01 3.36247861e-01 -6.28172278e-01
6.56912982e-01 -6.29926920e-01 -2.79998511e-01 4.83700633e-01
-5.58840811e-01 -2.54050821e-01 2.17261970e-01 7.49298334e-01
-4.39045101e-01 4.52520192e-01 7.99708188e-01 -5.20347655e-01
-1.16754413e+00 7.18641222e-01 -7.99750924e-01 5.97605482e-02
1.30062282e+00 -3.14897895e-01 -3.65588702e-02 -3.04807425e-01
-1.00017011e+00 2.68478185e-01 4.35278624e-01 5.92592061e-01
4.10549849e-01 -1.23463595e+00 -3.45288545e-01 5.10102212e-02
1.69321626e-01 5.85575104e-02 3.13254118e-01 1.09864855e+00
-4.87565607e-01 7.20715463e-01 -4.13980335e-01 -7.67141342e-01
-1.44248807e+00 3.95181656e-01 3.57994080e-01 -5.07140160e-01
-6.03476703e-01 9.34445202e-01 3.18758041e-01 -3.83263469e-01
3.95271540e-01 -8.72855365e-01 -1.95291132e-01 -1.93273589e-01
4.55867767e-01 4.71820116e-01 -2.27332771e-01 -9.13958788e-01
-6.77273571e-01 5.58868885e-01 1.00638188e-01 -8.94988254e-02
1.35391116e+00 -4.30317819e-02 -1.83744356e-02 6.65454090e-01
1.06986475e+00 -6.78740025e-01 -2.03815699e+00 -2.22885564e-01
8.06916505e-02 -2.29714721e-01 -1.89344212e-01 -8.03514123e-01
-6.92775249e-01 9.67990100e-01 3.36000085e-01 6.95311800e-02
1.11499226e+00 2.42871940e-01 9.34263825e-01 3.53205234e-01
4.84319329e-01 -1.19565392e+00 4.46827233e-01 9.31003749e-01
3.94791335e-01 -1.38203633e+00 -8.02540109e-02 -2.58076906e-01
-6.84359133e-01 8.78758311e-01 6.11294031e-01 -2.80510485e-01
8.23306262e-01 4.12907422e-01 -3.86754284e-04 -4.00976747e-01
-8.38182092e-01 -4.30373549e-01 3.68664235e-01 4.59555984e-01
3.39862049e-01 8.46425444e-02 2.26669863e-01 3.09525460e-01
4.16962862e-01 4.98314083e-01 2.90227849e-02 1.41123998e+00
-5.61856508e-01 -8.08415234e-01 2.05854490e-01 4.02119070e-01
-5.20207226e-01 1.09215222e-01 -2.58536845e-01 6.59031808e-01
2.89922357e-01 1.01232564e+00 4.48599100e-01 -3.26790988e-01
3.21659327e-01 4.98174913e-02 7.05636740e-01 -7.22031176e-01
-5.95083952e-01 1.04132101e-01 2.70102262e-01 -1.65514278e+00
-9.16616499e-01 -1.13087499e+00 -1.48797917e+00 1.57359928e-01
2.48523489e-01 -2.25467026e-01 3.35434794e-01 1.41909885e+00
3.35565478e-01 7.20410287e-01 7.23404661e-02 -1.18915498e+00
-1.81981817e-01 -8.84922802e-01 -6.22393489e-01 5.04488051e-01
4.74666029e-01 -8.16239476e-01 2.25937486e-01 4.69466746e-01] | [8.221253395080566, 0.5229828953742981] |
93368f36-d583-416b-adb2-cfed99a88ef4 | robust-object-detection-under-occlusion-with | 2005.11643 | null | https://arxiv.org/abs/2005.11643v2 | https://arxiv.org/pdf/2005.11643v2.pdf | Robust Object Detection under Occlusion with Context-Aware CompositionalNets | Detecting partially occluded objects is a difficult task. Our experimental results show that deep learning approaches, such as Faster R-CNN, are not robust at object detection under occlusion. Compositional convolutional neural networks (CompositionalNets) have been shown to be robust at classifying occluded objects by explicitly representing the object as a composition of parts. In this work, we propose to overcome two limitations of CompositionalNets which will enable them to detect partially occluded objects: 1) CompositionalNets, as well as other DCNN architectures, do not explicitly separate the representation of the context from the object itself. Under strong object occlusion, the influence of the context is amplified which can have severe negative effects for detection at test time. In order to overcome this, we propose to segment the context during training via bounding box annotations. We then use the segmentation to learn a context-aware CompositionalNet that disentangles the representation of the context and the object. 2) We extend the part-based voting scheme in CompositionalNets to vote for the corners of the object's bounding box, which enables the model to reliably estimate bounding boxes for partially occluded objects. Our extensive experiments show that our proposed model can detect objects robustly, increasing the detection performance of strongly occluded vehicles from PASCAL3D+ and MS-COCO by 41% and 35% respectively in absolute performance relative to Faster R-CNN. | ['Yihong Sun', 'Adam Kortylewski', 'Angtian Wang', 'Alan Yuille'] | 2020-05-24 | robust-object-detection-under-occlusion-with-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Wang_Robust_Object_Detection_Under_Occlusion_With_Context-Aware_CompositionalNets_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Wang_Robust_Object_Detection_Under_Occlusion_With_Context-Aware_CompositionalNets_CVPR_2020_paper.pdf | cvpr-2020-6 | ['robust-object-detection'] | ['computer-vision'] | [ 1.28596708e-01 2.65446454e-01 -7.29930848e-02 -4.04960841e-01
-3.54610473e-01 -6.55690134e-01 5.61071992e-01 -9.07607675e-02
-4.93735790e-01 3.85599345e-01 -1.19563602e-01 -3.64328116e-01
5.15259087e-01 -6.87961698e-01 -1.00248599e+00 -5.66622078e-01
3.52618955e-02 3.67872357e-01 8.90588701e-01 -7.53443968e-03
9.06610489e-03 9.08941984e-01 -1.65559387e+00 6.07786119e-01
7.91786492e-01 1.09428060e+00 1.12076581e-01 6.03691697e-01
-2.08781570e-01 7.40701258e-01 -8.27406347e-01 -1.55705869e-01
4.34788704e-01 2.21252441e-03 -5.85519910e-01 1.85254827e-01
9.64727879e-01 -6.91183388e-01 -2.10694194e-01 8.99005711e-01
3.86016932e-03 -1.56628743e-01 6.13747656e-01 -1.15831482e+00
-2.82483786e-01 3.90025705e-01 -8.08654487e-01 3.20259422e-01
2.66507058e-03 2.27973372e-01 8.38409662e-01 -1.04333687e+00
6.35139585e-01 1.55276835e+00 7.38962710e-01 5.71299613e-01
-1.23348522e+00 -7.70629406e-01 7.41295695e-01 -1.25626419e-02
-1.37382114e+00 -2.50924855e-01 7.19810665e-01 -3.91747832e-01
9.04024005e-01 2.77499765e-01 6.19387388e-01 9.24315512e-01
-9.01896954e-02 1.24571228e+00 1.00559270e+00 -1.28196463e-01
1.36768639e-01 1.63861543e-01 2.27891415e-01 6.18497252e-01
5.70164263e-01 2.09476516e-01 9.50930789e-02 2.48567220e-02
5.85690141e-01 9.96363014e-02 -6.90427348e-02 -6.41111612e-01
-8.29430938e-01 7.93229043e-01 8.62563193e-01 1.43546551e-01
-1.45497635e-01 3.98317307e-01 3.71350229e-01 -1.00709230e-01
3.61664146e-01 8.75342637e-02 -5.63511550e-01 5.48979700e-01
-1.03221035e+00 3.81208599e-01 6.09377086e-01 9.01147604e-01
6.88452601e-01 8.67037848e-02 -2.40217596e-01 5.50813317e-01
4.35274303e-01 4.05161291e-01 5.72848916e-02 -6.87871218e-01
4.49595064e-01 9.70917761e-01 1.28907204e-01 -8.08544457e-01
-4.18143660e-01 -6.50347233e-01 -4.75960910e-01 5.22763491e-01
5.50425529e-01 -2.85622235e-02 -1.46666849e+00 1.63891196e+00
5.26902437e-01 1.09393284e-01 -5.06825931e-02 1.13188148e+00
1.03969061e+00 3.29979956e-01 3.92141700e-01 4.12952811e-01
1.38439548e+00 -9.29451346e-01 -3.75431180e-01 -4.78499413e-01
5.54755926e-01 -6.94009364e-01 7.37539887e-01 1.04443677e-01
-9.01394844e-01 -7.16074049e-01 -1.21746159e+00 -2.16517642e-01
-4.88981158e-01 4.55346853e-01 5.30387104e-01 6.53034449e-01
-7.91395366e-01 3.66127998e-01 -8.83215547e-01 -2.15592712e-01
1.03773713e+00 4.68735576e-01 -3.53768378e-01 -1.48530051e-01
-7.56537855e-01 9.32365179e-01 4.04832095e-01 3.31465483e-01
-1.32863224e+00 -5.26296139e-01 -9.91728961e-01 5.93701452e-02
4.40008044e-01 -2.63619095e-01 9.65366900e-01 -1.17691028e+00
-8.39165747e-01 8.48747790e-01 -2.07270876e-01 -6.33777440e-01
7.97694623e-01 -3.93540531e-01 -5.96516505e-02 1.48434281e-01
4.64556217e-02 1.14323616e+00 6.81149840e-01 -1.60182238e+00
-8.75787318e-01 -3.67946893e-01 2.63326406e-01 2.34705415e-02
1.88655779e-01 -1.26106009e-01 -6.04014516e-01 -5.34953415e-01
3.94757628e-01 -1.02431905e+00 -3.50168675e-01 4.61264610e-01
-4.48125124e-01 -2.71586418e-01 1.60631180e+00 -4.07597572e-01
5.61269879e-01 -2.15186667e+00 -2.84070283e-01 5.23466384e-03
2.13798136e-01 6.75434291e-01 -2.82608181e-01 -1.36266530e-01
1.12469107e-01 2.48986095e-01 -2.54094630e-01 -5.18007576e-01
-2.84135014e-01 5.95724463e-01 -4.69847411e-01 7.58839667e-01
7.93635845e-01 1.06943357e+00 -6.24983013e-01 -4.35274869e-01
3.83266956e-01 8.07860315e-01 -4.68977839e-01 7.91834760e-03
-4.38460290e-01 3.33062142e-01 -3.03076386e-01 8.49565446e-01
9.37746167e-01 1.28356993e-01 1.78801678e-02 -1.54834643e-01
-1.70325711e-02 3.06383401e-01 -1.28602755e+00 9.95048285e-01
-1.86101660e-01 8.74939501e-01 2.59641200e-01 -9.15194213e-01
7.68109798e-01 7.72352368e-02 3.91255319e-02 -6.16264999e-01
2.61127502e-01 6.06760681e-02 8.36906135e-02 -4.24621105e-01
4.47224557e-01 3.08423728e-01 1.87675565e-01 8.05985555e-03
-1.27874449e-01 4.36329655e-02 1.40951574e-01 9.24586877e-02
8.55470181e-01 3.94592613e-01 3.55694182e-02 -1.89584360e-01
5.87655485e-01 9.10901353e-02 6.93192542e-01 8.19387794e-01
-4.87331390e-01 7.23796189e-01 5.91570914e-01 -7.64456868e-01
-8.53862584e-01 -8.76078546e-01 -3.31722081e-01 1.02139258e+00
3.46322984e-01 4.47868146e-02 -6.28192365e-01 -1.00058913e+00
2.92292535e-01 3.80627483e-01 -7.79815257e-01 9.49869156e-02
-1.03265786e+00 -6.59928322e-01 4.93161172e-01 9.36399341e-01
4.58004802e-01 -9.70068693e-01 -8.74219179e-01 1.30085945e-01
-8.85124356e-02 -1.59848881e+00 -1.54296637e-01 4.90878791e-01
-1.05405927e+00 -1.29390621e+00 -4.75316167e-01 -7.99502611e-01
1.04151678e+00 4.94448900e-01 9.76116359e-01 3.63319635e-01
-4.32977021e-01 7.46223098e-03 -2.09864199e-01 -5.26611447e-01
-2.90322602e-01 -1.60475492e-01 -3.01688731e-01 -6.78773895e-02
3.12491298e-01 -2.05412880e-01 -7.55267441e-01 7.07368433e-01
-9.58837688e-01 -1.40645010e-02 7.11405754e-01 5.20943940e-01
7.54955411e-01 -1.05387092e-01 3.34904343e-01 -1.01517785e+00
-8.94667804e-02 -2.60687679e-01 -6.99741721e-01 -6.60273805e-02
-6.93397969e-02 -6.60704970e-02 2.86526322e-01 -6.88431740e-01
-8.05739939e-01 5.70132196e-01 -9.02788714e-02 -5.37130117e-01
-3.52405339e-01 -3.78299430e-02 -1.69624433e-01 -2.42591798e-01
5.42726934e-01 -5.14868870e-02 -3.51103455e-01 -4.38022733e-01
4.39808398e-01 3.74431044e-01 5.52714229e-01 -4.50446397e-01
8.97967935e-01 9.57047105e-01 -1.12109324e-02 -6.41625881e-01
-8.04152489e-01 -7.02740610e-01 -7.01217234e-01 -1.79601476e-01
1.02697551e+00 -1.23478699e+00 -4.55026299e-01 1.70981139e-01
-1.27085471e+00 -3.16740334e-01 -2.05868170e-01 1.81099162e-01
-3.82315740e-02 1.81328133e-01 -5.65636814e-01 -8.45659375e-01
-2.00900748e-01 -1.32598424e+00 1.37865484e+00 3.22493672e-01
1.76322239e-04 -4.62937415e-01 -4.63581830e-01 3.63104582e-01
3.91417950e-01 5.62068760e-01 6.47456944e-01 -6.76862717e-01
-9.47360814e-01 -4.22521293e-01 -6.21414542e-01 3.54955882e-01
-2.56640851e-01 2.80046523e-01 -1.21059358e+00 -3.50698292e-01
-2.49392763e-01 -9.42029431e-02 1.22937632e+00 3.16047043e-01
1.00174236e+00 -1.17036119e-01 -6.28197551e-01 4.26370472e-01
1.35017884e+00 -4.90123965e-02 5.73284507e-01 1.32504493e-01
8.66718113e-01 6.60945058e-01 5.88801742e-01 -1.04529716e-01
1.04263805e-01 6.21901691e-01 9.56598878e-01 -3.62377793e-01
-5.57418406e-01 -1.57670841e-01 2.28459239e-01 -1.27374634e-01
1.42778188e-01 -9.78493467e-02 -7.86409676e-01 8.03320646e-01
-1.68722761e+00 -6.36631727e-01 -5.23976266e-01 1.97796583e+00
5.61650097e-01 4.85736847e-01 9.25356522e-02 7.63251781e-02
6.99798703e-01 8.66056532e-02 -4.12462533e-01 -2.47592106e-01
-1.30932868e-01 6.74138889e-02 7.15635180e-01 4.04406041e-01
-1.38848746e+00 1.10198283e+00 6.09415770e+00 4.85656291e-01
-1.13960159e+00 1.78458661e-01 5.58995783e-01 -1.05028458e-01
7.19842836e-02 -1.52988192e-02 -1.06439292e+00 2.24264443e-01
3.36086899e-01 6.16488397e-01 -2.36290336e-01 1.27875638e+00
2.34283768e-02 -2.45529622e-01 -1.17341888e+00 4.39271837e-01
1.85078755e-02 -9.15311515e-01 4.24005203e-02 6.00494035e-02
9.32426453e-01 3.01504076e-01 7.82850906e-02 4.65047419e-01
2.62164026e-01 -1.09483027e+00 1.05988467e+00 -1.41805306e-01
3.15027863e-01 -5.75541317e-01 8.80126595e-01 2.39383876e-01
-1.41643822e+00 -1.42309651e-01 -5.59956729e-01 -8.52338597e-03
-5.94421923e-02 4.76174206e-01 -1.02732003e+00 4.14723568e-02
7.85678267e-01 2.88304210e-01 -8.49601626e-01 1.16595948e+00
-6.43704236e-01 5.31562030e-01 -3.81368786e-01 2.50887185e-01
3.88408840e-01 3.66804481e-01 6.57059789e-01 1.37483537e+00
-2.57972956e-01 -7.49035925e-02 3.87271345e-01 1.15846646e+00
-1.49983466e-01 -1.60581395e-01 -4.24197614e-01 4.12098229e-01
1.69444025e-01 1.23510969e+00 -1.01571095e+00 -4.08676088e-01
-3.00471872e-01 7.45298624e-01 3.82731378e-01 4.77042228e-01
-1.02355850e+00 -9.15999562e-02 5.91897845e-01 2.66829073e-01
8.95403802e-01 -3.67590368e-01 -4.00854021e-01 -7.58933067e-01
2.16777265e-01 -6.29206121e-01 1.67085350e-01 -7.17980206e-01
-7.24111199e-01 6.49797201e-01 -1.08406015e-01 -1.02562726e+00
2.77384520e-01 -8.19793820e-01 -5.94064236e-01 6.63546622e-01
-1.76444542e+00 -1.40982616e+00 -4.93666768e-01 2.65065610e-01
6.49200797e-01 3.18709433e-01 2.94480771e-01 3.66634399e-01
-5.62085450e-01 3.77853960e-01 -4.70603704e-01 4.57399875e-01
3.37759137e-01 -1.22263181e+00 4.54615444e-01 1.08187163e+00
2.68380523e-01 5.28195679e-01 7.66250312e-01 -6.27399147e-01
-1.04140115e+00 -1.49065113e+00 5.98549604e-01 -6.62193179e-01
2.64315099e-01 -6.70618296e-01 -1.02995455e+00 5.62069952e-01
-2.06936628e-01 7.20097303e-01 2.04782218e-01 -6.53454363e-02
-7.83196807e-01 -8.79915133e-02 -1.17731285e+00 5.59419930e-01
9.30341244e-01 -5.29086232e-01 -5.50843894e-01 1.77382857e-01
8.53518188e-01 -5.11188209e-01 -2.59845406e-01 8.79774749e-01
5.79281211e-01 -9.32505310e-01 1.01577640e+00 -3.83787602e-01
2.23174989e-01 -6.49828732e-01 -1.48365170e-01 -5.72720826e-01
1.59702182e-01 2.72261258e-02 -3.58806998e-01 8.88905585e-01
2.44367510e-01 -4.49018061e-01 1.01996875e+00 4.66198117e-01
-1.14665873e-01 -7.92544723e-01 -8.71086478e-01 -6.40488863e-01
-6.49573132e-02 -5.87276936e-01 3.91104996e-01 5.34812987e-01
-6.68040633e-01 3.30310762e-02 3.26835290e-02 6.12928331e-01
4.26105142e-01 1.33999601e-01 8.36771488e-01 -1.23589551e+00
1.45692686e-02 -4.70742792e-01 -6.07378900e-01 -1.23092973e+00
6.10679574e-02 -5.89287579e-01 2.52814353e-01 -1.46678436e+00
1.32388443e-01 -7.10620105e-01 -2.05641553e-01 6.27389848e-01
-2.49023989e-01 7.65476644e-01 1.98039681e-01 2.25567445e-01
-7.83478498e-01 1.25751391e-01 1.25452065e+00 -3.17612082e-01
-2.09910497e-01 3.65994610e-02 -3.78939480e-01 8.74390900e-01
6.18464768e-01 -6.32432759e-01 -1.49809748e-01 -4.71829951e-01
-7.57145658e-02 -3.79618376e-01 7.70709813e-01 -1.06720757e+00
8.74404889e-03 1.49753168e-01 7.45550752e-01 -1.03967845e+00
1.96586102e-01 -1.07890725e+00 -5.11463583e-02 6.53966665e-01
-1.92069337e-01 -3.15025806e-01 5.46508968e-01 7.27874398e-01
-1.10119216e-01 -9.66460556e-02 8.28986704e-01 -1.72705706e-02
-6.54436886e-01 1.25975490e-01 -4.04928595e-01 -2.22952694e-01
1.11004055e+00 -4.12238777e-01 -4.90889788e-01 8.41780007e-02
-6.84356034e-01 3.15583825e-01 3.74347359e-01 6.78253055e-01
4.62472796e-01 -9.54278469e-01 -4.69185233e-01 1.67800844e-01
1.19536385e-01 4.15620446e-01 1.07490890e-01 7.80519366e-01
-6.30989730e-01 5.03164053e-01 1.43170115e-02 -9.33893383e-01
-1.58992624e+00 5.58915615e-01 5.47705352e-01 -6.44723475e-02
-6.21531904e-01 9.30388987e-01 6.53378785e-01 -3.11189413e-01
6.28393114e-01 -9.07488108e-01 -3.18322390e-01 -2.21685059e-02
4.22210306e-01 7.24145398e-02 9.26536694e-02 -8.73256147e-01
-5.44740081e-01 5.74682415e-01 -3.53595495e-01 3.56268078e-01
1.03970885e+00 5.46096452e-02 5.87272234e-02 -6.90929517e-02
1.21137524e+00 -5.78538105e-02 -1.57985330e+00 -2.39347950e-01
-3.68341208e-02 -2.74109900e-01 1.14690810e-01 -8.19757640e-01
-1.34000683e+00 9.00921524e-01 7.37423658e-01 3.25844996e-02
7.38393426e-01 1.54696211e-01 5.83638668e-01 1.25057295e-01
1.37320846e-01 -7.80404806e-01 1.30946517e-01 3.10499370e-01
7.62060523e-01 -1.42579329e+00 8.60230103e-02 -6.72706962e-01
-3.19355816e-01 9.73236203e-01 9.01425540e-01 -2.93648183e-01
3.46356422e-01 3.49686772e-01 2.11151615e-01 -3.36659282e-01
-4.48209673e-01 -6.44707799e-01 5.93474984e-01 5.21856368e-01
1.61495551e-01 8.02573636e-02 -6.31810129e-02 3.76296103e-01
2.44946510e-01 -4.09104586e-01 3.26195568e-01 1.00366187e+00
-6.76568031e-01 -8.28885615e-01 -5.43778718e-01 6.01747818e-02
-7.09942043e-01 2.08337322e-01 -5.85082173e-01 1.04191589e+00
6.20252013e-01 9.23837841e-01 1.47666022e-01 -5.49386591e-02
4.05411780e-01 -1.14394769e-01 2.63334006e-01 -6.18419707e-01
-6.01084530e-01 6.65957900e-03 -1.02964252e-01 -7.63403356e-01
-4.61417884e-01 -4.31296259e-01 -1.37311280e+00 1.56639487e-01
-6.53152883e-01 -2.56643832e-01 8.96972537e-01 9.37475383e-01
1.18090250e-01 6.89785004e-01 9.70606059e-02 -1.04828513e+00
-2.77878076e-01 -7.68297255e-01 1.88521110e-02 3.52562249e-01
6.24996006e-01 -6.38075531e-01 -4.93549794e-01 -1.71910793e-01] | [9.250589370727539, 0.6244004368782043] |
b437ac16-0a81-41fd-bcd4-86a90243d1fd | autotsg-learning-and-synthesis-for-incident | 2205.13457 | null | https://arxiv.org/abs/2205.13457v1 | https://arxiv.org/pdf/2205.13457v1.pdf | AutoTSG: Learning and Synthesis for Incident Troubleshooting | Incident management is a key aspect of operating large-scale cloud services. To aid with faster and efficient resolution of incidents, engineering teams document frequent troubleshooting steps in the form of Troubleshooting Guides (TSGs), to be used by on-call engineers (OCEs). However, TSGs are siloed, unstructured, and often incomplete, requiring developers to manually understand and execute necessary steps. This results in a plethora of issues such as on-call fatigue, reduced productivity, and human errors. In this work, we conduct a large-scale empirical study of over 4K+ TSGs mapped to 1000s of incidents and find that TSGs are widely used and help significantly reduce mitigation efforts. We then analyze feedback on TSGs provided by 400+ OCEs and propose a taxonomy of issues that highlights significant gaps in TSG quality. To alleviate these gaps, we investigate the automation of TSGs and propose AutoTSG -- a novel framework for automation of TSGs to executable workflows by combining machine learning and program synthesis. Our evaluation of AutoTSG on 50 TSGs shows the effectiveness in both identifying TSG statements (accuracy 0.89) and parsing them for execution (precision 0.94 and recall 0.91). Lastly, we survey ten Microsoft engineers and show the importance of TSG automation and the usefulness of AutoTSG. | ['Anurag Gupta', 'Arjun Radhakrishna', 'Sai Pramod Upadhyayula', 'Chetan Bansal', 'Manish Shetty'] | 2022-05-26 | null | null | null | null | ['program-synthesis'] | ['computer-code'] | [-1.46548152e-01 -2.32782945e-01 1.72915637e-01 -3.49443436e-01
-7.83083558e-01 -5.63494742e-01 -1.58518061e-01 5.89429319e-01
-8.35294649e-02 4.21783984e-01 1.82147503e-01 -9.86505330e-01
-3.17907959e-01 -7.40500391e-01 -4.16817605e-01 4.48167436e-02
8.96545574e-02 2.83186674e-01 3.79666500e-02 -1.12632640e-01
7.12634027e-01 4.18964952e-01 -1.14141381e+00 5.98581016e-01
8.36911738e-01 4.84315276e-01 2.21892931e-02 8.19431245e-01
-1.26164258e-01 9.82042730e-01 -1.13739312e+00 -1.73481822e-01
2.46548504e-01 -1.69937778e-02 -1.00234461e+00 7.06902817e-02
1.06232531e-01 -4.68377978e-01 1.03487827e-01 6.72544658e-01
3.36839348e-01 -2.20391035e-01 -2.97298431e-02 -1.60517561e+00
-1.97703466e-01 3.63120079e-01 -4.49515522e-01 4.29014355e-01
6.01770699e-01 4.44179505e-01 8.59290004e-01 -7.10610569e-01
4.96429563e-01 5.72766721e-01 8.03308010e-01 8.24891254e-02
-1.04522622e+00 -7.79531538e-01 -2.08355129e-01 2.87124336e-01
-1.21357560e+00 -5.45877218e-01 9.51498151e-02 -8.36358786e-01
1.95691574e+00 5.84162772e-01 5.11392593e-01 4.13414240e-01
3.54157209e-01 6.52417541e-02 9.91727769e-01 -7.68328905e-01
3.07701916e-01 1.83930069e-01 5.82466841e-01 7.78259516e-01
6.22854292e-01 -3.89314920e-01 -4.33526069e-01 -5.29392540e-01
5.26877880e-01 3.03524643e-01 -2.07012847e-01 4.03889835e-01
-9.02864277e-01 5.32437742e-01 -2.87377954e-01 3.76457363e-01
-5.75559795e-01 -5.17763644e-02 6.23417616e-01 2.82426089e-01
2.62626082e-01 8.88036668e-01 -4.71611649e-01 -7.20359325e-01
-8.08951080e-01 1.03857316e-01 1.15269053e+00 1.14342904e+00
7.73873687e-01 -8.26587342e-03 -3.36386800e-01 4.69622642e-01
-1.77835628e-01 2.52719104e-01 -1.84873059e-01 -8.80513072e-01
6.64931774e-01 9.92319643e-01 3.25172693e-01 -1.14241910e+00
-5.82585037e-01 -3.24361056e-01 -4.15634692e-01 -5.66065125e-02
5.05089853e-03 -1.09705217e-01 -3.22347224e-01 9.14229870e-01
-1.24073565e-01 -5.01902178e-02 -5.68966985e-01 5.43147981e-01
2.29299888e-01 5.56880295e-01 -1.03065595e-01 -3.86845022e-01
1.21035886e+00 -9.42143798e-01 -8.31694067e-01 -4.30747747e-01
1.06315875e+00 -9.02357340e-01 1.41385484e+00 5.44157267e-01
-8.53520870e-01 -1.11242779e-01 -7.03708470e-01 5.34379423e-01
-7.85440579e-02 1.33808255e-01 5.35751045e-01 6.26145840e-01
-9.75118577e-01 5.11165082e-01 -8.71630490e-01 -6.31664753e-01
4.00500298e-01 9.96060595e-02 -3.33565213e-02 -2.60614365e-01
-5.37000835e-01 8.86087477e-01 1.77994400e-01 -3.92369688e-01
-4.85801518e-01 -1.00726926e+00 -3.74036640e-01 2.67349154e-01
6.74600661e-01 -4.28651720e-01 1.58197141e+00 -4.18969654e-02
-7.48747230e-01 3.83796185e-01 -1.89801380e-01 -2.71058176e-02
-4.51730862e-02 -3.93786579e-01 -7.19919741e-01 2.45980211e-02
4.63058650e-01 -4.77181166e-01 3.90092760e-01 -7.59360611e-01
-1.03904748e+00 -1.85149461e-01 2.17395410e-01 -2.99175292e-01
-5.20380914e-01 6.05710506e-01 -2.36326158e-01 -1.99231312e-01
-3.82914275e-01 -7.64526546e-01 -3.07958752e-01 -9.04788911e-01
-4.39733267e-01 1.21044554e-03 4.31755692e-01 -1.06898236e+00
2.36920023e+00 -1.94017112e+00 -5.34993589e-01 1.69576943e-01
4.97574389e-01 3.01953495e-01 1.48323223e-01 8.72774541e-01
-2.30263516e-01 4.84317243e-01 6.24863803e-02 7.70520270e-02
-3.14662568e-02 2.14895466e-04 -3.76823455e-01 7.47147053e-02
2.02637091e-01 7.22133458e-01 -8.88316393e-01 -2.07284436e-01
2.78483838e-01 -2.29765803e-01 -6.06560051e-01 3.70579392e-01
1.53170690e-01 1.42011002e-01 -1.93971992e-01 9.44929659e-01
3.48560184e-01 -5.48070610e-01 1.29578054e-01 4.84480932e-02
-5.02350271e-01 7.15460420e-01 -9.28591490e-01 9.75554645e-01
-7.96066821e-01 5.03823578e-01 -1.45307317e-01 -7.51414537e-01
8.22553456e-01 3.16176951e-01 3.41824114e-01 -6.14057779e-01
-1.68936104e-01 9.74311009e-02 -2.66649455e-01 -9.16238487e-01
3.71306509e-01 2.07234785e-01 -3.12044591e-01 1.16132689e+00
-2.64614850e-01 3.06803379e-02 5.25984406e-01 4.18976486e-01
1.94643700e+00 -6.15854919e-01 8.19347143e-01 7.36225173e-02
3.24190594e-02 5.16859949e-01 5.58377504e-01 5.62796950e-01
-1.41308695e-01 1.96403146e-01 7.05071151e-01 -5.99655509e-01
-1.08297634e+00 -3.69128764e-01 4.41496700e-01 9.33077335e-01
-6.39718354e-01 -1.16818678e+00 -9.62499261e-01 -8.51554692e-01
-2.26432383e-01 1.24730718e+00 -8.53196532e-02 1.32009480e-02
-5.55166721e-01 -6.39302969e-01 5.22998452e-01 7.16381252e-01
2.31195599e-01 -1.04721487e+00 -1.09618807e+00 4.26339477e-01
-4.53963935e-01 -1.36100709e+00 -5.08168280e-01 2.69543510e-02
-5.98298550e-01 -1.29414248e+00 3.03746760e-01 -1.43680200e-01
7.51031458e-01 5.06694317e-01 1.37424910e+00 5.19280136e-01
-8.31504345e-01 3.98330182e-01 -6.11733913e-01 -4.72531706e-01
-4.20828253e-01 -1.84906393e-01 -7.07453713e-02 -5.35895348e-01
7.96772957e-01 -5.34713566e-01 -1.02900296e-01 3.74259651e-01
-6.66710317e-01 -3.03441808e-02 6.14369631e-01 5.45629025e-01
1.52911887e-01 4.74529862e-01 6.18679583e-01 -1.13687849e+00
9.49416161e-01 -3.71293336e-01 -7.52724349e-01 3.98823619e-01
-1.07485223e+00 -5.51758170e-01 7.91934490e-01 2.53323287e-01
-5.61373413e-01 -5.23778319e-01 -6.27670810e-02 -1.52468726e-01
-5.42660356e-01 6.00026309e-01 2.40227863e-01 1.12052657e-01
7.10547924e-01 -1.78002000e-01 -3.19206297e-01 -2.25822642e-01
-3.28468293e-01 9.79088128e-01 1.61544219e-01 -4.49104249e-01
6.64717495e-01 3.36086378e-02 -5.07235706e-01 -6.70111775e-01
-5.35036266e-01 -8.41013014e-01 -1.37340829e-01 -2.16970056e-01
3.56435776e-01 -3.24519336e-01 -8.30246449e-01 -1.98774096e-02
-1.31140172e+00 -3.33731264e-01 3.56323905e-02 1.62377492e-01
-1.73860312e-01 1.95698455e-01 -7.42174566e-01 -8.89417708e-01
-5.42118847e-01 -1.08240104e+00 6.96217954e-01 1.57092083e-02
-9.67206597e-01 -6.55807078e-01 3.55504639e-02 6.78067923e-01
7.28657246e-01 1.90395042e-01 1.14332330e+00 -6.30015910e-01
-7.31307149e-01 -4.89607275e-01 -3.84444922e-01 3.76985133e-01
3.26442301e-01 2.49741524e-01 -6.49108827e-01 -2.60116667e-01
-1.82026047e-02 2.59410530e-01 -1.38649464e-01 9.11619812e-02
1.22764385e+00 -7.02565372e-01 -3.64980251e-01 4.45189208e-01
1.56051254e+00 5.31100154e-01 5.89976966e-01 7.82710373e-01
4.92055267e-01 6.99587524e-01 1.06440783e+00 8.55462551e-01
2.16732115e-01 4.36504185e-01 1.64986402e-01 8.37764367e-02
3.92734677e-01 2.43341923e-02 2.58496195e-01 1.04345870e+00
-3.07183236e-01 -1.50999799e-02 -1.66895652e+00 6.51198864e-01
-1.69340658e+00 -7.68829763e-01 -4.84554499e-01 2.03316283e+00
5.32012641e-01 3.79390180e-01 4.28148620e-02 4.01379317e-01
6.43613398e-01 -3.54047924e-01 -5.98192541e-03 -6.85622633e-01
7.70901918e-01 4.87215936e-01 6.20337725e-01 1.39867947e-01
-5.72089851e-01 5.75514495e-01 5.93205070e+00 3.33815694e-01
-8.72766197e-01 1.30998656e-01 3.00501764e-01 -8.84494260e-02
-7.72530288e-02 2.55777121e-01 -7.72067189e-01 5.19158661e-01
1.36330175e+00 -5.23125947e-01 6.36333644e-01 1.14127040e+00
5.92681885e-01 -1.03534207e-01 -1.02731407e+00 9.37525034e-01
-2.65858144e-01 -1.72261536e+00 -3.57140154e-01 -5.84828146e-02
6.30841434e-01 -2.79381126e-01 -3.45706284e-01 4.13798600e-01
2.44681194e-01 -1.03075182e+00 3.95529687e-01 3.29749763e-01
9.15878296e-01 -9.14738894e-01 9.90090013e-01 4.11584079e-01
-1.00734198e+00 -3.38069856e-01 -9.23299640e-02 -4.99618977e-01
2.57798225e-01 8.81643713e-01 -1.76971865e+00 6.12090290e-01
1.08660030e+00 1.97418109e-01 -5.29775620e-01 1.04530406e+00
-1.84400320e-01 9.55802679e-01 8.04192349e-02 2.99119260e-02
-1.36483788e-01 3.36907953e-02 3.51699173e-01 1.65830505e+00
2.90248632e-01 3.57462853e-01 3.30569267e-01 7.79269874e-01
4.19530630e-01 1.11396993e-02 -3.44572365e-01 -5.19260943e-01
8.92798007e-01 1.39435422e+00 -8.84231746e-01 -3.67315620e-01
-4.99264091e-01 5.80699742e-01 5.12241013e-02 3.88189167e-01
-7.68430233e-01 -8.55127037e-01 7.51127481e-01 6.57805383e-01
-1.17426351e-01 -3.44387144e-01 -6.52319849e-01 -7.29846299e-01
1.75021321e-01 -1.26202214e+00 1.05337493e-01 -5.07238388e-01
-9.72356558e-01 7.48055995e-01 -1.53503701e-01 -1.10480464e+00
-1.84273601e-01 -2.28356943e-01 -7.76340961e-01 8.13532412e-01
-1.08161914e+00 -5.71032941e-01 -7.47637272e-01 2.69163728e-01
4.73855674e-01 6.05477355e-02 8.74833822e-01 7.30041385e-01
-8.06556404e-01 4.40623254e-01 -5.33075929e-01 -4.27950509e-02
8.60502183e-01 -1.14492607e+00 8.00249517e-01 1.07318032e+00
-3.31552863e-01 1.33488655e+00 6.21000469e-01 -9.43361342e-01
-1.30366647e+00 -1.37995517e+00 1.21826315e+00 -7.20588028e-01
7.38229096e-01 -1.85847625e-01 -9.04411733e-01 8.30351055e-01
-4.53546382e-02 -2.98956543e-01 9.11607623e-01 2.47172743e-01
-1.43925905e-01 -1.30815595e-01 -1.00241709e+00 2.80578345e-01
8.50467682e-01 -7.80359745e-01 -4.14336056e-01 5.19286096e-01
7.68406689e-01 -2.47460455e-01 -7.25463033e-01 8.93057510e-02
1.67689901e-02 -1.13348591e+00 4.27912772e-01 -6.36496842e-01
5.07939398e-01 -3.02685559e-01 -3.24259773e-02 -1.19050050e+00
-4.88990366e-01 -6.93899930e-01 1.40616879e-01 1.23561215e+00
1.36582628e-01 -6.71296895e-01 3.80054742e-01 9.18396652e-01
-6.66071534e-01 -7.95898497e-01 -1.72720402e-01 -6.26728475e-01
-9.87973690e-01 -8.67898941e-01 8.11454654e-01 1.35227466e+00
4.10172343e-01 2.03794748e-01 -1.20047651e-01 3.64161491e-01
3.84104997e-01 -7.48798400e-02 9.07414258e-01 -8.96204233e-01
-4.17642802e-01 -1.40410334e-01 -8.92928913e-02 -2.56097317e-01
-5.44520974e-01 -5.80689251e-01 -1.77241057e-01 -1.82481706e+00
2.26496756e-01 -4.12472010e-01 -1.82643887e-02 9.98946548e-01
7.33998045e-02 -1.06431298e-01 4.57271188e-02 1.24493152e-01
-7.93945193e-01 -4.91481543e-01 3.82178456e-01 3.12871456e-01
-3.70417833e-01 1.10744223e-01 -9.86917913e-01 7.34102428e-01
7.96572268e-01 -5.44021606e-01 -1.14009723e-01 -4.16276991e-01
5.79290807e-01 2.19986901e-01 3.33572537e-01 -1.04238963e+00
5.69787025e-01 -4.77447003e-01 -2.90111899e-01 -6.54058397e-01
-4.16149169e-01 -7.79029667e-01 2.52801329e-01 4.04519886e-01
1.82542503e-01 5.75606942e-01 2.53804535e-01 9.62514430e-02
-1.92608356e-01 -2.35553324e-01 2.94562578e-01 -8.21509436e-02
-6.60418391e-01 -2.27652323e-02 -4.96090859e-01 -9.20050666e-02
1.19081008e+00 -1.73663124e-01 -6.83376610e-01 -1.37649179e-01
-2.58655012e-01 5.33477664e-02 6.24626696e-01 5.84441014e-02
6.49542272e-01 -6.28706932e-01 -3.80573064e-01 3.49514604e-01
3.25630099e-01 -2.90663034e-01 3.14717829e-01 1.17250347e+00
-8.15583289e-01 7.80566514e-01 -1.47388339e-01 -4.61725593e-01
-1.32916570e+00 3.92301112e-01 -2.43723586e-01 -3.76992762e-01
-5.48875570e-01 6.62481248e-01 -1.96825892e-01 -3.00186813e-01
1.23233810e-01 -4.65151072e-01 1.69301510e-01 -1.78409517e-01
8.86747301e-01 9.37167704e-01 8.36333632e-01 1.49121314e-01
-5.15998423e-01 1.77231953e-01 -2.23264351e-01 2.64397174e-01
1.59491074e+00 -5.12272567e-02 -4.87583071e-01 1.47611409e-01
7.68143952e-01 2.81706274e-01 -5.41702271e-01 1.30387634e-01
5.21018505e-01 -9.31738317e-01 -2.34860823e-01 -9.85083222e-01
-7.36290276e-01 6.38894379e-01 1.55645907e-02 4.63643074e-01
1.31114674e+00 -2.02631861e-01 9.69023764e-01 4.51931238e-01
8.71456683e-01 -1.03698337e+00 8.92364606e-02 4.67691958e-01
6.00417316e-01 -7.25579381e-01 -1.22987002e-01 -5.95950246e-01
-6.23884201e-01 1.22472930e+00 7.85290718e-01 1.92370623e-01
2.57574439e-01 7.69202352e-01 -1.39334366e-01 -4.77665931e-01
-1.07075894e+00 5.03490150e-01 -3.94102037e-01 4.86153871e-01
4.12109315e-01 2.92530715e-01 -1.88456848e-01 8.85628104e-01
-1.91490099e-01 3.83020073e-01 1.02440405e+00 1.42399347e+00
-4.43032652e-01 -1.01050675e+00 -5.96940339e-01 8.16736460e-01
-4.98347521e-01 -2.30291709e-01 -1.22989357e-01 6.83085263e-01
2.79203393e-02 1.27309120e+00 -2.73428082e-01 -7.88129032e-01
6.86511517e-01 1.58138752e-01 -3.37650478e-02 -1.11756468e+00
-9.12367940e-01 -8.83348212e-02 5.41536391e-01 -9.90065515e-01
2.51092583e-01 -5.06268024e-01 -1.12022042e+00 -7.56156981e-01
-3.52719873e-01 3.73317152e-01 5.91270745e-01 9.43751812e-01
6.19779289e-01 1.13669646e+00 5.24594665e-01 -3.22136357e-02
-5.69757223e-01 -8.28356624e-01 -4.42952096e-01 9.61509049e-02
1.56736702e-01 -3.60979259e-01 -2.69407898e-01 1.48073345e-01] | [7.9870524406433105, 7.01692533493042] |
3427e0ca-ce45-41c0-9cd5-e036153b5ff2 | automatic-trade-off-adaptation-in-offline-rl | 2306.09744 | null | https://arxiv.org/abs/2306.09744v1 | https://arxiv.org/pdf/2306.09744v1.pdf | Automatic Trade-off Adaptation in Offline RL | Recently, offline RL algorithms have been proposed that remain adaptive at runtime. For example, the LION algorithm \cite{lion} provides the user with an interface to set the trade-off between behavior cloning and optimality w.r.t. the estimated return at runtime. Experts can then use this interface to adapt the policy behavior according to their preferences and find a good trade-off between conservatism and performance optimization. Since expert time is precious, we extend the methodology with an autopilot that automatically finds the correct parameterization of the trade-off, yielding a new algorithm which we term AutoLION. | ['Thomas Runkler', 'Steffen Udluft', 'Phillip Swazinna'] | 2023-06-16 | null | null | null | null | ['offline-rl'] | ['playing-games'] | [-2.87811577e-01 1.85814038e-01 -4.69890207e-01 -3.68251711e-01
-5.60846984e-01 -1.06785858e+00 1.53082564e-01 1.66315764e-01
-7.99107134e-01 6.35819972e-01 -2.58398443e-01 -5.65789998e-01
-5.81629634e-01 -7.57300735e-01 -2.28999242e-01 -5.81371903e-01
-1.66308418e-01 5.39126158e-01 2.16892809e-01 -1.20643690e-01
4.96340841e-01 3.57028842e-01 -1.31295669e+00 -6.32646233e-02
1.04403293e+00 1.10216117e+00 1.53975800e-01 7.49930084e-01
1.83439985e-01 3.41823727e-01 -5.15315831e-01 -2.39433616e-01
6.24665737e-01 -4.96590614e-01 -6.16432369e-01 -3.17468494e-02
-5.32444417e-01 -3.31778258e-01 2.75346667e-01 9.82593894e-01
5.51323950e-01 3.83183151e-01 2.56092489e-01 -1.26083684e+00
-2.45881621e-02 8.62054646e-01 -4.08060014e-01 1.11936018e-01
2.98262298e-01 4.45299357e-01 1.14076376e+00 -1.77000016e-01
4.59455848e-01 1.22861278e+00 1.69385329e-01 3.21361691e-01
-1.62389600e+00 -5.94526052e-01 4.59148586e-01 1.69723988e-01
-1.47537982e+00 -2.63731122e-01 6.00773871e-01 -8.75451714e-02
6.00959599e-01 4.24809903e-01 9.73093569e-01 7.14452684e-01
2.48686180e-01 5.68115473e-01 1.40795445e+00 -4.33554500e-01
7.41827548e-01 5.21019638e-01 -1.94272503e-01 4.09434468e-01
4.22543198e-01 3.26154768e-01 -3.01847339e-01 -3.05606097e-01
6.65811777e-01 -4.69276190e-01 -2.16436729e-01 -7.38102973e-01
-6.07007086e-01 9.14995670e-01 1.11781441e-01 1.80594757e-01
-4.43745166e-01 2.06808105e-01 3.29327434e-01 8.26072514e-01
1.37896165e-01 9.42643642e-01 -5.25338709e-01 -4.04442310e-01
-7.11712837e-01 5.14137566e-01 9.30858493e-01 4.76648092e-01
6.31176353e-01 -9.24724713e-02 -2.37625301e-01 7.07211733e-01
2.19405696e-01 1.52308524e-01 2.74643302e-01 -1.48958826e+00
3.69165361e-01 5.96049786e-01 9.90364671e-01 -8.26853454e-01
-4.60639745e-01 -8.12408447e-01 -3.90781052e-02 6.35712385e-01
5.28757513e-01 -6.59521520e-01 -2.26083741e-01 1.76322591e+00
5.71545303e-01 -4.74823982e-01 -6.36022091e-02 1.06129491e+00
-3.13495070e-01 4.98354971e-01 -1.46493196e-01 -5.91248631e-01
8.97246480e-01 -5.68838775e-01 -7.93265998e-01 -3.08167726e-01
5.80765903e-01 -3.61766517e-01 1.25063717e+00 6.74100220e-01
-1.23760962e+00 -2.53880322e-02 -9.75990593e-01 6.81026578e-01
-2.21242443e-01 3.73241276e-01 4.87138659e-01 9.15921688e-01
-1.00653851e+00 9.84975398e-01 -9.67808425e-01 -1.40922546e-01
-6.63690642e-02 6.90454721e-01 3.61212134e-01 4.52652812e-01
-9.88337398e-01 1.12335706e+00 6.73476815e-01 2.07545817e-01
-6.70595109e-01 -4.51692402e-01 -4.29640085e-01 7.66901895e-02
1.05564630e+00 -5.91934323e-01 1.66003990e+00 -1.59352839e+00
-2.29349613e+00 4.64618891e-01 4.01145965e-01 -6.09602809e-01
8.98571730e-01 -1.59523427e-03 -8.45057741e-02 -1.28665313e-01
-6.52760640e-02 3.44211698e-01 6.98534489e-01 -1.22601461e+00
-7.63195872e-01 -4.81955335e-02 5.27736068e-01 4.84002888e-01
-3.17819476e-01 1.43087627e-02 -2.32312486e-01 -1.92338899e-01
-3.01882416e-01 -1.14426160e+00 -4.89567369e-01 -2.11613640e-01
-1.77216962e-01 -1.23929352e-01 2.63959974e-01 -1.69845670e-01
1.57109749e+00 -1.76788938e+00 1.85029909e-01 7.24574029e-01
-2.34160602e-01 1.23633388e-02 -1.76790953e-01 5.27566373e-01
8.66789967e-02 8.65439624e-02 5.19866422e-02 -1.56352273e-03
1.81346267e-01 1.73862070e-01 1.02940360e-02 6.99322999e-01
-3.85792404e-01 4.77140605e-01 -9.76445377e-01 -3.22028905e-01
1.22944660e-01 -1.34089082e-01 -1.00696158e+00 4.52234775e-01
-4.86492604e-01 3.64976525e-01 -8.89709592e-01 4.01827604e-01
3.55902016e-01 -1.28175721e-01 7.75022864e-01 1.72278270e-01
-7.30781019e-01 1.85486764e-01 -1.56040716e+00 1.09834087e+00
-7.29913533e-01 1.17846102e-01 6.25763834e-01 -8.62729788e-01
7.11906135e-01 7.08968714e-02 6.47948265e-01 -3.23226482e-01
5.68325639e-01 2.67334640e-01 2.37467751e-01 -2.58159548e-01
3.74560684e-01 1.04543246e-01 2.43288614e-02 6.70595825e-01
-4.01790529e-01 -2.45686411e-03 1.99581772e-01 -1.85118422e-01
9.62447524e-01 4.92228895e-01 6.71561658e-01 -6.30108953e-01
5.47031879e-01 1.20208241e-01 5.40274978e-01 9.97252107e-01
-8.81889090e-02 -5.23663349e-02 9.03047919e-01 -2.29930252e-01
-8.20537150e-01 -4.67101455e-01 1.69667885e-01 1.40398252e+00
-7.36080855e-02 -2.99533933e-01 -8.25400949e-01 -4.78014499e-01
4.62925136e-02 9.78023469e-01 -8.78388464e-01 -9.90488902e-02
-4.06597227e-01 -3.53693336e-01 -7.86250606e-02 9.48008671e-02
3.22063774e-01 -9.17350113e-01 -1.45998693e+00 3.93598527e-01
9.84966308e-02 -6.67222917e-01 -6.42386496e-01 4.09151226e-01
-7.59294748e-01 -8.09820056e-01 -4.30411786e-01 3.07275262e-02
8.74344826e-01 -1.35070086e-01 8.61545801e-01 -2.43230201e-02
2.10556984e-02 5.83737254e-01 -3.90801072e-01 -2.44533509e-01
-2.30103225e-01 2.38725707e-01 7.72700994e-04 1.31950080e-01
-4.76238281e-01 -5.82809389e-01 -7.96390116e-01 4.82748628e-01
-6.88469470e-01 -6.09926395e-02 3.11003268e-01 5.08166075e-01
5.86369216e-01 1.90069199e-01 6.22435987e-01 -7.57400990e-01
1.13998091e+00 -4.95913476e-01 -1.45924103e+00 4.16078210e-01
-9.92061675e-01 5.60357451e-01 9.72901702e-01 -6.93224609e-01
-9.18183565e-01 2.77140439e-01 2.58573860e-01 -5.23255348e-01
3.79713386e-01 5.39163291e-01 -2.33887956e-01 -2.16578022e-01
3.95853341e-01 4.73538833e-03 -9.19300318e-02 -2.86521435e-01
4.56991136e-01 3.96422029e-01 1.24180891e-01 -8.68978679e-01
4.68971342e-01 -3.67985442e-02 -1.32962883e-01 -1.35031253e-01
-9.98790860e-01 -8.45805481e-02 -3.33441764e-01 -5.39519489e-01
3.57786715e-01 -3.26639056e-01 -1.30559003e+00 -1.06203377e-01
-8.34204078e-01 -7.11805224e-01 -4.67568159e-01 2.99498409e-01
-1.00347149e+00 -2.64674891e-02 2.26232991e-01 -1.25016654e+00
-2.70374864e-01 -9.17423427e-01 3.19638878e-01 2.59604216e-01
-2.75728881e-01 -8.77472639e-01 1.44485399e-01 -1.36183694e-01
4.17817891e-01 2.20470488e-01 4.91703957e-01 -5.99302053e-01
-4.26746428e-01 -9.93533954e-02 2.43753925e-01 -1.24133281e-01
4.06040475e-02 -1.97693571e-01 -3.32633674e-01 -6.71816707e-01
-1.29512131e-01 -2.81431615e-01 1.17223695e-01 4.49865967e-01
1.35894525e+00 -1.03509903e+00 -2.59165853e-01 7.79298067e-01
1.43368042e+00 6.24086738e-01 2.12583050e-01 9.09268916e-01
-5.04601747e-02 6.00705266e-01 1.03751481e+00 9.83373582e-01
2.20937416e-01 1.03094494e+00 4.44352597e-01 3.41824919e-01
7.07218468e-01 -2.77787030e-01 4.23013180e-01 -6.29344732e-02
-1.12377822e-01 -2.65424848e-01 -6.10708773e-01 1.95168778e-01
-2.24531889e+00 -8.24913919e-01 6.69628501e-01 2.77687716e+00
8.75936568e-01 3.09036255e-01 5.71319878e-01 1.89565122e-02
4.49980378e-01 1.26464386e-02 -1.02025211e+00 -8.85533690e-01
5.85448980e-01 -1.57289729e-01 1.17172098e+00 7.28078365e-01
-5.83191037e-01 9.72649634e-01 6.92353201e+00 6.50767624e-01
-1.05106115e+00 -1.16603144e-01 6.89227700e-01 -3.41336280e-01
-4.49396461e-01 3.42655033e-01 -6.36691689e-01 5.21698236e-01
1.03111088e+00 -5.60113132e-01 1.06698823e+00 9.31393743e-01
9.71778095e-01 -5.06408989e-01 -1.01860297e+00 6.59853220e-01
-4.57642585e-01 -8.57701480e-01 -4.81817991e-01 1.66389436e-01
4.97231632e-01 -5.86102724e-01 -1.80636972e-01 1.95376918e-01
7.05704391e-01 -6.74973667e-01 9.18535471e-01 5.65011322e-01
4.39639002e-01 -1.26078212e+00 5.02477467e-01 6.39619827e-01
-9.23214912e-01 -6.57731354e-01 -3.42250131e-02 -5.49899153e-02
7.63644576e-02 3.23360682e-01 -7.21582353e-01 2.81958997e-01
3.23527843e-01 1.05042927e-01 -2.43077666e-01 1.14945316e+00
-2.95465440e-01 3.84287417e-01 -4.32279527e-01 -4.61899370e-01
4.34520572e-01 -7.99787760e-01 6.44833326e-01 8.08215022e-01
3.08079869e-01 3.28611046e-01 4.46429044e-01 1.07798409e+00
5.20296335e-01 3.64545614e-01 -2.62559801e-01 -1.85581535e-01
6.09086335e-01 1.23032820e+00 -9.93949234e-01 -3.40316556e-02
3.72033447e-01 6.80496454e-01 4.39534396e-01 2.88196295e-01
-7.63615549e-01 -2.85631686e-01 6.11123085e-01 1.49809554e-01
3.66951883e-01 -3.13521087e-01 -2.07420662e-01 -6.81655824e-01
-2.87164629e-01 -9.68490243e-01 6.13277197e-01 -4.39866245e-01
-6.41362369e-01 2.88802981e-01 3.77479672e-01 -1.05106306e+00
-5.07004559e-01 -3.58147919e-01 -5.56881189e-01 4.60799783e-01
-9.66071784e-01 -2.92224139e-01 2.18793452e-01 4.68068212e-01
3.63413602e-01 6.11227192e-02 3.51789057e-01 -1.10008352e-01
-7.70487726e-01 6.20368183e-01 6.32125139e-02 -7.20220625e-01
3.00411165e-01 -1.22501624e+00 -3.64403248e-01 4.86301392e-01
-4.32042658e-01 5.91633141e-01 1.30475342e+00 -3.93016368e-01
-1.68342888e+00 -6.46059275e-01 2.54050225e-01 -1.20352812e-01
9.59811032e-01 -1.85029194e-01 -3.21627051e-01 4.64048266e-01
8.35550502e-02 -3.57284337e-01 3.30853492e-01 -7.48867467e-02
2.58489490e-01 -4.11576807e-01 -1.11470330e+00 8.92366052e-01
7.17135310e-01 1.00709073e-01 1.29887904e-03 1.83885083e-01
6.00040615e-01 -5.51019847e-01 -7.05802321e-01 5.15243858e-02
5.77105522e-01 -9.19930875e-01 6.59797251e-01 -3.09486926e-01
-3.95885736e-01 -2.30463639e-01 2.06887648e-01 -1.51763904e+00
-1.55039400e-01 -1.26226568e+00 7.59869516e-02 9.24803436e-01
5.45522869e-01 -9.63111460e-01 5.11055648e-01 8.46053720e-01
2.43171155e-01 -9.53620613e-01 -8.31833780e-01 -9.25486267e-01
-2.49921098e-01 -3.93429369e-01 4.96810287e-01 4.18427914e-01
6.73775524e-02 -8.75379294e-02 -5.32002509e-01 3.36908288e-02
6.81560993e-01 4.46458101e-01 6.38528585e-01 -8.25193882e-01
-7.02849090e-01 -5.37854910e-01 4.58222389e-01 -1.11611760e+00
1.46172345e-01 -3.78972381e-01 7.26300851e-02 -1.22158909e+00
-1.36819884e-01 -6.59870982e-01 -3.28959972e-01 7.04243839e-01
2.23258153e-01 -5.07304132e-01 3.09522599e-01 8.25032219e-02
-7.48495281e-01 4.43669528e-01 1.13790798e+00 4.70641017e-01
-1.10081029e+00 5.14272809e-01 -8.67416024e-01 4.92364705e-01
1.12235510e+00 -6.83062077e-01 -5.45503497e-01 -6.32638410e-02
6.07466340e-01 6.11684263e-01 -8.05332232e-03 -5.93702018e-01
8.51054117e-02 -8.15348148e-01 -1.60132989e-01 -3.77153635e-01
1.14390276e-01 -1.08116043e+00 2.27661490e-01 6.05184257e-01
-7.82014132e-01 2.80635744e-01 1.05827432e-02 4.57346648e-01
4.18223232e-01 -4.96006787e-01 8.60060334e-01 -1.11082315e-01
-8.44678059e-02 6.69632256e-02 -6.47816122e-01 -7.79794082e-02
1.04802942e+00 -2.22444519e-01 1.11254662e-01 -7.60981143e-01
-6.27738893e-01 7.12906480e-01 4.20305729e-01 1.80984780e-01
6.45021722e-02 -8.73034060e-01 -9.31947827e-02 -2.19912887e-01
-3.71397048e-01 -6.80844605e-01 -2.65647829e-01 7.93837070e-01
-1.60883158e-01 4.46113586e-01 -2.18912855e-01 -2.48868559e-02
-1.03708434e+00 4.82019275e-01 8.98076475e-01 -5.53665996e-01
-3.87902260e-01 2.01517150e-01 -3.81817013e-01 -3.31573844e-01
2.51244128e-01 -3.37447196e-01 -1.67925268e-01 2.56710202e-01
2.54537135e-01 5.43144286e-01 -1.10633589e-01 1.15086965e-01
-5.05827427e-01 4.64493275e-01 4.68939662e-01 -7.60948062e-01
1.24961782e+00 -4.14238036e-01 5.08417487e-02 3.53289515e-01
6.82448030e-01 9.85157117e-02 -1.65552568e+00 2.24413589e-01
2.77961522e-01 -7.42574334e-01 3.59139055e-01 -1.11659491e+00
-1.07758009e+00 6.64577857e-02 7.44599938e-01 4.17415410e-01
1.17858624e+00 -4.64265466e-01 9.77213681e-02 5.03893673e-01
5.99474609e-01 -1.60145891e+00 -3.03462725e-02 3.19058388e-01
7.35970855e-01 -7.49378860e-01 2.93128341e-01 1.32635340e-01
-9.00301218e-01 9.49576139e-01 6.58029020e-01 -3.50189805e-01
3.98586065e-01 3.27743649e-01 2.18982790e-02 1.76741704e-01
-1.14777076e+00 -2.14303792e-01 8.75684433e-03 2.81782568e-01
2.82214824e-02 3.53576303e-01 -1.11840129e+00 6.17874205e-01
-2.73314267e-01 4.33703959e-02 4.32512105e-01 9.29976761e-01
-7.70957768e-01 -1.44328129e+00 -6.68739021e-01 1.72047451e-01
-2.75634885e-01 3.81780177e-01 -3.92351031e-01 7.11718798e-01
-9.31952074e-02 1.10251379e+00 -9.01028663e-02 -6.65475726e-02
5.24160266e-01 -2.58502036e-01 4.26533759e-01 -2.43586794e-01
-8.40135932e-01 1.97312027e-01 3.74962181e-01 -1.04552066e+00
-1.07302822e-01 -5.25801301e-01 -1.11938000e+00 -2.54065599e-02
-1.80717587e-01 5.85295796e-01 5.57306170e-01 9.30279195e-01
2.60674953e-01 2.57158041e-01 1.15089273e+00 -6.81096733e-01
-1.07117522e+00 -4.61420506e-01 -4.52059060e-01 -4.63559538e-01
1.21890910e-01 -7.22933471e-01 -4.61637169e-01 -6.31429374e-01] | [4.110484600067139, 2.322951078414917] |
978dce74-e8ee-4d34-8bb8-d2d726f5f455 | improving-factuality-of-abstractive | 2305.14981 | null | https://arxiv.org/abs/2305.14981v1 | https://arxiv.org/pdf/2305.14981v1.pdf | Improving Factuality of Abstractive Summarization without Sacrificing Summary Quality | Improving factual consistency of abstractive summarization has been a widely studied topic. However, most of the prior works on training factuality-aware models have ignored the negative effect it has on summary quality. We propose EFACTSUM (i.e., Effective Factual Summarization), a candidate summary generation and ranking technique to improve summary factuality without sacrificing summary quality. We show that using a contrastive learning framework with our refined candidate summaries leads to significant gains on both factuality and similarity-based metrics. Specifically, we propose a ranking strategy in which we effectively combine two metrics, thereby preventing any conflict during training. Models trained using our approach show up to 6 points of absolute improvement over the base model with respect to FactCC on XSUM and 11 points on CNN/DM, without negatively affecting either similarity-based metrics or absractiveness. | ['Muhao Chen', 'Fei Wang', 'Tanay Dixit'] | 2023-05-24 | null | null | null | null | ['abstractive-text-summarization'] | ['natural-language-processing'] | [ 1.99638575e-01 4.21066940e-01 -5.72780669e-01 -3.58006269e-01
-1.21173513e+00 -5.98221362e-01 1.00889909e+00 6.75285280e-01
-3.78783196e-01 1.12658930e+00 8.99997652e-01 -6.62833229e-02
-1.65826201e-01 -6.76726937e-01 -8.23290050e-01 -1.37920201e-01
1.77856222e-01 -4.25380170e-02 1.71710998e-01 -3.12281758e-01
9.66785312e-01 8.27697366e-02 -1.55377388e+00 6.41059637e-01
1.46886075e+00 8.10400784e-01 -2.23788694e-01 7.16337502e-01
-7.02103302e-02 1.22133029e+00 -1.34778249e+00 -8.08590174e-01
-9.04073790e-02 -5.17698288e-01 -9.33901906e-01 -2.62669057e-01
1.44791269e+00 -4.60124254e-01 -2.36648753e-01 9.13562059e-01
5.20175695e-01 9.40305069e-02 6.58709049e-01 -9.63854492e-01
-5.43599427e-01 1.14892375e+00 -3.65871042e-01 5.35720408e-01
2.48021200e-01 3.26112956e-02 1.50659919e+00 -6.35328889e-01
6.22808218e-01 1.12827837e+00 4.70236391e-01 4.28521663e-01
-9.46175814e-01 -4.40036267e-01 3.30477506e-01 2.55781054e-01
-7.92108178e-01 -6.96855068e-01 8.13997746e-01 -5.08527756e-02
1.24440122e+00 6.47577763e-01 4.79534179e-01 9.61790621e-01
5.36606491e-01 1.09163713e+00 6.54855549e-01 -3.00136894e-01
2.04575911e-01 -3.45894881e-02 4.16923732e-01 3.72214913e-01
1.02348650e+00 -3.44601393e-01 -8.32196712e-01 1.37400078e-02
2.91386306e-01 -3.22163880e-01 -2.99917519e-01 1.46957636e-01
-1.05032587e+00 7.07701504e-01 3.30991983e-01 2.94275969e-01
-4.85546768e-01 4.34876949e-01 6.42889082e-01 2.76782632e-01
7.42413104e-01 1.21824443e+00 -4.61798042e-01 -3.44866037e-01
-1.40580320e+00 7.53614128e-01 6.49672806e-01 6.86863482e-01
3.12977433e-01 3.50845546e-01 -7.37306237e-01 8.88810396e-01
-2.73421824e-01 4.48892057e-01 4.53051150e-01 -1.16887522e+00
7.81855524e-01 8.13720286e-01 9.87504870e-02 -1.18199968e+00
-3.47487420e-01 -9.65955853e-01 -5.90358198e-01 -2.77633846e-01
-1.39721587e-01 -1.77104920e-01 -4.76000071e-01 1.72890031e+00
-3.06193262e-01 1.20154740e-02 1.89426124e-01 4.98338968e-01
1.23582697e+00 6.49079323e-01 -1.63754579e-02 -3.84097755e-01
9.72580969e-01 -9.54740226e-01 -8.69888723e-01 -3.06519806e-01
7.17416763e-01 -7.11978376e-01 1.17897391e+00 5.91860592e-01
-1.61327934e+00 -3.46736133e-01 -1.59934497e+00 -2.22840130e-01
-2.15472922e-01 1.91069052e-01 6.28394425e-01 6.37468934e-01
-1.03067553e+00 1.07176828e+00 -5.76086104e-01 -9.46466029e-02
6.26647890e-01 1.80232614e-01 -6.20131940e-02 1.42076418e-01
-1.07343483e+00 1.08686471e+00 5.72215974e-01 -4.77836430e-01
-5.39851546e-01 -1.10886896e+00 -7.36785889e-01 4.14781302e-01
7.10237324e-01 -1.07931805e+00 1.58053815e+00 -5.14565051e-01
-1.35138404e+00 4.08012509e-01 -1.66742653e-01 -7.40177631e-01
4.87082392e-01 -8.38895202e-01 -3.96887660e-01 1.07217744e-01
1.82845786e-01 6.32623494e-01 4.45787489e-01 -1.21891928e+00
-8.46821845e-01 -1.17017157e-01 3.97308648e-01 4.79192376e-01
-6.86778247e-01 -3.38652551e-01 -1.95717335e-01 -8.02837431e-01
-1.93200663e-01 -4.26153809e-01 6.72358200e-02 -5.98780870e-01
-6.77555621e-01 -4.03647482e-01 6.07464612e-01 -7.15977073e-01
1.76828539e+00 -1.49387300e+00 -4.13459055e-02 -3.09957772e-01
3.32085997e-01 4.50039625e-01 -3.05285037e-01 5.22631168e-01
5.85167259e-02 6.12637997e-01 -2.47081459e-01 -1.53865531e-01
-8.51950943e-02 -1.36329383e-01 -6.02825463e-01 -1.50610954e-01
4.30078477e-01 9.52402890e-01 -1.18266869e+00 -4.04841781e-01
-2.45784372e-02 -2.00519077e-02 -7.81896889e-01 5.60236424e-02
-4.29285705e-01 -2.34527409e-01 -2.15842858e-01 1.54546455e-01
4.63571399e-01 -2.57728457e-01 1.84453979e-01 -1.19380631e-01
-1.54723570e-01 1.22184503e+00 -7.59706795e-01 1.67794681e+00
-7.78973699e-01 7.79367328e-01 -7.08074212e-01 -5.69523215e-01
7.21670926e-01 2.04514593e-01 2.21477553e-01 -9.72468317e-01
8.04150701e-02 2.67097831e-01 -6.39976002e-03 -2.40279123e-01
1.36341572e+00 8.08597505e-02 -9.54618976e-02 5.49733043e-01
3.47097255e-02 -3.19108158e-01 5.73407114e-01 7.31644392e-01
1.16202474e+00 -1.53277412e-01 4.74953741e-01 -3.49046677e-01
3.82521033e-01 1.56491399e-01 3.43465775e-01 8.84683132e-01
3.31460297e-01 6.97506785e-01 9.07716036e-01 7.58576952e-03
-9.28135037e-01 -9.91306961e-01 1.47383600e-01 6.61273241e-01
2.55470891e-02 -9.09706354e-01 -6.71482623e-01 -9.60535824e-01
-4.41860147e-02 1.55033588e+00 -4.37330514e-01 -6.01633847e-01
-7.70169139e-01 -7.42441952e-01 5.99998534e-01 4.79048401e-01
7.75651455e-01 -7.91649580e-01 -8.61572623e-01 6.76789209e-02
-5.61797917e-01 -7.07568944e-01 -4.81241107e-01 -1.87332645e-01
-1.29782999e+00 -7.97462046e-01 -3.79172087e-01 -1.76433455e-02
1.35670170e-01 6.09438121e-01 1.48934615e+00 8.33357498e-02
3.59916776e-01 1.49859861e-01 -2.51216024e-01 -6.35681927e-01
-4.10664856e-01 4.56108004e-01 -1.86342388e-01 -5.33673286e-01
6.85906131e-03 -5.82002759e-01 -6.18988514e-01 -3.20528507e-01
-1.07775640e+00 8.14970136e-02 7.08605409e-01 6.82889879e-01
1.03555560e-01 1.18065700e-02 1.19587064e+00 -1.17206872e+00
1.04969192e+00 -3.52988064e-01 -9.93838608e-02 3.11119437e-01
-1.02777290e+00 1.65320799e-01 7.94824660e-01 -4.92196120e-02
-1.21616709e+00 -9.50752497e-01 -2.68744770e-02 1.52003653e-02
3.97547871e-01 7.63339460e-01 -1.69863760e-01 4.48451877e-01
9.80564415e-01 -1.47873200e-02 -4.36964065e-01 -5.44412583e-02
5.31727135e-01 4.06122237e-01 6.88624203e-01 -4.84879315e-01
4.73280013e-01 2.82687068e-01 -2.80142635e-01 -6.05909646e-01
-1.25468862e+00 -3.59000385e-01 -3.28435183e-01 -7.62623325e-02
2.57262796e-01 -8.43681157e-01 -3.87866169e-01 1.11420467e-01
-1.33554769e+00 8.73986259e-02 -3.44843775e-01 2.70564258e-01
-4.85695601e-01 4.68507230e-01 -3.34930599e-01 -4.55583215e-01
-8.45746160e-01 -4.33061451e-01 8.47095549e-01 2.96478659e-01
-6.41995370e-01 -9.50898588e-01 1.25100642e-01 3.20979089e-01
4.86543387e-01 3.35704237e-01 1.02968228e+00 -8.71181011e-01
-5.90850592e-01 -1.97481215e-01 -2.21130237e-01 4.32034194e-01
3.36214393e-01 7.19733909e-02 -8.88226390e-01 3.23122256e-02
-5.79531491e-02 -2.13210076e-01 1.30446649e+00 5.33061147e-01
1.00529039e+00 -9.77287710e-01 -5.12844995e-02 6.94314986e-02
1.36809254e+00 8.42034072e-02 7.29716122e-01 3.53987187e-01
5.95682502e-01 5.12061715e-01 8.09056759e-01 5.55629671e-01
3.23408008e-01 5.57941198e-01 2.10447490e-01 1.87181070e-01
-4.59119081e-01 -4.01117295e-01 3.57118845e-01 7.46176898e-01
-2.11962797e-02 -6.22721314e-01 -6.18148565e-01 9.01488900e-01
-1.90982890e+00 -1.32178152e+00 -1.32871866e-01 2.28008461e+00
1.05921805e+00 5.94737172e-01 1.24513991e-02 1.63930982e-01
2.96700180e-01 6.69372082e-01 -4.28613722e-01 -7.14055717e-01
-3.36568803e-01 1.83814719e-01 3.37022752e-01 4.88128245e-01
-7.98197865e-01 9.94671524e-01 6.45807791e+00 8.32779288e-01
-1.20859313e+00 -1.37723491e-01 5.79599380e-01 -6.30563974e-01
-8.64370108e-01 -7.68643692e-02 -5.89572608e-01 3.12778324e-01
9.86451626e-01 -7.88779438e-01 -1.45450011e-01 8.24204326e-01
4.17120337e-01 -3.10076624e-01 -1.21264088e+00 6.14437580e-01
4.62085754e-01 -1.88720858e+00 5.74657321e-01 -1.85460508e-01
1.14474583e+00 -3.86023611e-01 1.50005504e-01 3.96304995e-01
4.34900343e-01 -8.21950614e-01 8.82011473e-01 4.26897585e-01
4.75744516e-01 -1.04619777e+00 9.53856230e-01 2.13232085e-01
-4.82225567e-01 1.62706554e-01 -2.77697533e-01 -9.16305184e-02
5.99020571e-02 9.31789458e-01 -1.00342572e+00 1.01274335e+00
2.50746816e-01 7.91549742e-01 -8.57694387e-01 1.03864944e+00
-3.98431808e-01 9.30536866e-01 6.46292418e-02 -2.18330562e-01
1.61484420e-01 3.24429661e-01 7.60415316e-01 1.36635137e+00
2.14006484e-01 -4.29283418e-02 -3.37065965e-01 7.72438288e-01
-5.42721272e-01 1.08516015e-01 -5.83166480e-01 -2.69154385e-02
5.89143276e-01 9.93487775e-01 -4.06424224e-01 -8.26469004e-01
-1.43756390e-01 6.88737988e-01 2.74083555e-01 5.28328829e-02
-8.55424583e-01 -4.96694982e-01 3.92580539e-01 8.42654705e-03
2.46018186e-01 4.52177189e-02 -9.99744952e-01 -1.28090227e+00
3.19870025e-01 -9.26001310e-01 4.64655370e-01 -5.76177597e-01
-9.53859508e-01 4.84950542e-01 2.87175834e-01 -1.17638540e+00
-3.56144011e-01 3.77171077e-02 -1.04197705e+00 4.91985619e-01
-1.72045004e+00 -8.43428016e-01 -1.13791130e-01 -1.76728770e-01
7.41115451e-01 -1.82114244e-02 3.06558847e-01 -2.14178432e-02
-5.41866422e-01 7.79190302e-01 -1.63147762e-01 -4.48884428e-01
8.29432547e-01 -1.37974656e+00 5.56522071e-01 1.17886102e+00
1.50158465e-01 1.03857255e+00 1.15161192e+00 -8.68095160e-01
-9.34613407e-01 -1.20810139e+00 1.30307758e+00 -7.24939048e-01
3.97238195e-01 1.50140122e-01 -8.85102689e-01 5.27874529e-01
5.90808392e-01 -8.10565710e-01 5.26302278e-01 4.85002011e-01
-6.63841605e-01 -2.63607144e-01 -9.21947777e-01 1.03182065e+00
1.13651812e+00 -1.38152108e-01 -1.01525390e+00 2.48007864e-01
9.81541991e-01 -3.33871394e-01 -5.99532723e-01 5.81627309e-01
4.78906780e-01 -1.01707876e+00 7.53348529e-01 -7.03312814e-01
1.27529860e+00 -9.42490175e-02 -6.34935126e-02 -1.59021044e+00
-3.05620998e-01 -4.19487089e-01 -4.55843896e-01 1.38527262e+00
5.92697918e-01 -3.12797129e-01 7.57395387e-01 2.68147379e-01
-6.57469511e-01 -7.22919822e-01 -6.00538552e-01 -1.04608130e+00
3.70179206e-01 -2.41803780e-01 6.08950257e-01 8.44879389e-01
2.77565420e-01 8.55720580e-01 -3.13218147e-01 -1.47097722e-01
1.91760927e-01 1.59935594e-01 8.73497963e-01 -9.61765230e-01
-2.28626858e-02 -9.57323253e-01 1.79927982e-02 -1.04101193e+00
-1.42677389e-02 -8.83084774e-01 -8.68523642e-02 -2.13986254e+00
3.61071140e-01 1.11638196e-01 -2.47491896e-01 3.43106508e-01
-7.06285954e-01 -1.26368836e-01 2.26717666e-01 1.54532626e-01
-9.47222948e-01 7.27113187e-01 1.13235927e+00 -1.93794638e-01
-2.66892850e-01 -1.46209955e-01 -1.35436785e+00 4.54576552e-01
8.66051316e-01 -4.84161794e-01 -6.61644399e-01 -5.42109191e-01
3.37764472e-01 -7.33505748e-03 1.98599845e-01 -1.21558046e+00
1.31045267e-01 -2.38924727e-01 4.04830426e-02 -8.18218708e-01
1.22296385e-01 -1.09873287e-01 -2.87092656e-01 4.61686224e-01
-8.39963257e-01 2.00868934e-01 4.84407961e-01 4.45524663e-01
-3.95693868e-01 -2.33874381e-01 4.64118451e-01 -1.02966290e-03
-4.05740380e-01 -2.20284253e-01 -1.64098904e-01 1.93540335e-01
4.66576427e-01 -9.48094577e-03 -1.10687578e+00 -5.68813086e-01
1.97582424e-01 2.32512742e-01 3.33310932e-01 4.43451405e-01
5.43818295e-01 -1.14995348e+00 -1.11757803e+00 -6.14199698e-01
2.40972236e-01 -1.59036905e-01 3.81237894e-01 6.06735289e-01
-3.11652333e-01 9.60233390e-01 -2.60347337e-01 -2.00494245e-01
-1.21335590e+00 2.08500803e-01 1.89456195e-02 -8.09253216e-01
-3.99483919e-01 6.77558184e-01 -1.05568938e-01 -2.29872987e-01
4.61126268e-02 -5.43484688e-01 -4.02078122e-01 2.71364540e-01
6.42057180e-01 8.35151732e-01 6.80121899e-01 8.47626179e-02
-2.00905174e-01 8.31868276e-02 -5.52854538e-01 -1.96273014e-01
1.19359362e+00 1.79121777e-01 -2.56446912e-03 4.00737196e-01
9.60901499e-01 4.92531687e-01 -9.70586002e-01 -2.82752533e-02
3.72778744e-01 -4.42686707e-01 1.34272411e-01 -1.27642190e+00
-7.84408569e-01 7.50237107e-01 -1.77844748e-01 3.64868730e-01
1.02546108e+00 -2.62096196e-01 9.63364005e-01 5.37140489e-01
-7.75178615e-03 -1.06428492e+00 5.43364465e-01 6.01093233e-01
1.23266518e+00 -1.11339712e+00 5.18122017e-01 -4.89402980e-01
-7.69263625e-01 8.78830791e-01 7.09882498e-01 -2.67407507e-01
-1.60134926e-01 -1.74399197e-01 -2.31549695e-01 -3.18449765e-01
-1.29293621e+00 1.75139904e-01 5.26508510e-01 2.15826362e-01
7.39264727e-01 -7.23776221e-02 -8.67748737e-01 6.70941114e-01
-7.58054912e-01 -1.65856466e-01 1.00053155e+00 7.22390175e-01
-6.80884480e-01 -7.39323199e-01 -3.49085554e-02 9.04716909e-01
-6.76154971e-01 -4.01025921e-01 -5.77477098e-01 8.29546988e-01
-3.25292349e-01 1.17023206e+00 7.92632401e-02 -4.34300959e-01
5.32395422e-01 -1.58691645e-01 5.08876860e-01 -8.72694492e-01
-8.59475136e-01 -2.33210281e-01 7.28070438e-01 -3.58854324e-01
-2.59498775e-01 -6.64933205e-01 -1.18462539e+00 -6.60750270e-01
-2.57765502e-01 1.96143284e-01 4.68015701e-01 9.11237836e-01
5.63304603e-01 8.90619516e-01 5.26546717e-01 -2.17743605e-01
-6.30007863e-01 -1.06098378e+00 -6.16998151e-02 1.84083983e-01
3.68014038e-01 -3.77328634e-01 -2.65227586e-01 -1.93598807e-01] | [12.364200592041016, 9.372175216674805] |
3ade9224-9d55-4a64-bf8b-cee6eaae5cf0 | challenges-in-the-knowledge-base-population | null | null | https://aclanthology.org/L12-1104 | https://aclanthology.org/L12-1104.pdf | Challenges in the Knowledge Base Population Slot Filling Task | The Knowledge Based Population (KBP) evaluation track of the Text Analysis Conferences (TAC) has been held for the past 3 years. One of the two tasks of KBP is slot filling: finding within a large corpus the values of a set of attributes of given people and organizations. This task has proven very challenging, with top systems rarely exceeding 30{\%} F-measure. In this paper, we present an error analysis and classification for those answers which could be found by a manual corpus search but were not found by any of the systems participating in the 2010 evaluation. The most common sources of failure were limitations on inference, errors in coreference (particularly with nominal anaphors), and errors in named entity recognition. We relate the types of errors to the characteristics of the task and show the wide diversity of problems that must be addressed to improve overall performance. | ['Ralph Grishman', 'Bonan Min'] | 2012-05-01 | null | null | null | lrec-2012-5 | ['knowledge-base-population'] | ['natural-language-processing'] | [ 6.33229911e-02 6.00053847e-01 -1.47941768e-01 -3.68747264e-01
-9.46177542e-01 -6.39450669e-01 5.76956511e-01 6.91819608e-01
-7.59323537e-01 1.34185779e+00 4.27923083e-01 -2.66017258e-01
-5.95983446e-01 -6.02390528e-01 -4.89586473e-01 -1.72289893e-01
3.89287949e-01 1.20765996e+00 4.90425110e-01 -2.39698797e-01
5.35859644e-01 3.53015631e-01 -1.52909470e+00 5.72172165e-01
6.57076418e-01 8.27155888e-01 -2.32324898e-01 2.97749698e-01
-7.15458930e-01 9.23089981e-01 -1.00997877e+00 -7.31209695e-01
-2.93233424e-01 -4.82926965e-02 -1.57157063e+00 -3.75595599e-01
3.90894234e-01 2.31997266e-01 -3.94000262e-02 1.04367828e+00
6.45984411e-01 1.84368223e-01 4.95332122e-01 -1.24833632e+00
-1.94952667e-01 8.10516357e-01 -5.56342863e-02 6.05587959e-01
8.42265666e-01 -5.30556858e-01 1.08153772e+00 -8.01943719e-01
1.03629839e+00 1.35619783e+00 8.18947196e-01 4.45508391e-01
-1.15593767e+00 -6.37952089e-01 -2.92676836e-01 3.23766261e-01
-1.55623543e+00 -7.46478319e-01 4.49599884e-02 -5.77462614e-01
1.62026322e+00 4.57148820e-01 1.67959303e-01 6.40501618e-01
-1.90666243e-01 3.32989812e-01 9.19070840e-01 -7.09187865e-01
2.52914429e-01 5.41353166e-01 5.54844677e-01 3.54387313e-01
5.48588574e-01 -4.40175653e-01 -8.62181604e-01 -8.33829284e-01
3.05522114e-01 -8.29163611e-01 -2.49888375e-01 -3.64834704e-02
-1.02218318e+00 7.81065524e-01 -1.86135005e-02 5.24413824e-01
-4.38254178e-01 -3.58348340e-01 4.77439225e-01 1.36594102e-01
1.75908223e-01 1.06136322e+00 -9.31102037e-01 -4.69543010e-01
-7.28247881e-01 6.76686645e-01 1.39155388e+00 1.02139521e+00
5.58179438e-01 -7.04711080e-01 -1.77010566e-01 9.97766256e-01
-1.07611917e-01 1.54032812e-01 3.44250411e-01 -1.17550921e+00
8.70801568e-01 9.02770400e-01 6.23351991e-01 -1.07600105e+00
-7.35940576e-01 -8.26126803e-03 -4.62632775e-02 -4.61552918e-01
1.05859315e+00 -2.14068517e-01 -4.15573388e-01 1.56570470e+00
4.42628413e-01 -4.36198682e-01 2.81854749e-01 4.91394013e-01
1.09452391e+00 3.30101222e-01 5.22148669e-01 -5.06216586e-01
1.73578465e+00 -3.60047162e-01 -1.14828444e+00 -3.96527410e-01
9.53714907e-01 -1.02751350e+00 5.14840543e-01 7.60713369e-02
-1.08739817e+00 -1.89974234e-02 -6.74556434e-01 6.89053489e-03
-5.46777010e-01 -2.31650531e-01 5.40810466e-01 6.29746854e-01
-6.42764390e-01 4.24126148e-01 -3.45320106e-01 -7.95694888e-01
5.94988503e-02 4.78435695e-01 -4.54441935e-01 2.61460662e-01
-1.61210907e+00 1.43210948e+00 6.67928278e-01 -3.70141894e-01
2.02930674e-01 -6.05502546e-01 -5.16515136e-01 1.16793506e-01
7.15400457e-01 -3.97937655e-01 1.34486341e+00 -5.03418207e-01
-6.50510788e-01 1.20121193e+00 -2.88289219e-01 -4.57416445e-01
2.28872448e-01 -3.61388236e-01 -6.63151681e-01 1.95461195e-02
5.97321987e-01 2.93709785e-01 -7.66431466e-02 -8.32612216e-01
-1.01614821e+00 -4.57881898e-01 -1.19246744e-01 3.35467845e-01
-1.81223020e-01 7.19288111e-01 -3.44541848e-01 -1.79445714e-01
2.12789625e-01 -8.00959170e-01 4.98002023e-02 -8.36961389e-01
-3.84391695e-01 -8.14657211e-01 2.77465254e-01 -8.90701294e-01
1.63366282e+00 -1.93175435e+00 -7.71629736e-02 6.47413805e-02
-1.01153485e-01 1.29706278e-01 6.21802151e-01 6.69085741e-01
-1.90800708e-02 4.12681699e-01 1.48348331e-01 1.58208862e-01
1.10779069e-01 2.37389430e-01 -5.47290921e-01 -3.28863342e-03
2.03420557e-02 4.59512204e-01 -7.88586438e-01 -8.85768116e-01
-3.09014469e-01 -5.34542045e-03 -2.77452737e-01 -1.09041721e-01
-1.14706010e-01 -8.47793669e-02 -3.67405385e-01 6.22007191e-01
2.32214704e-01 -1.34673551e-01 4.30458516e-01 -2.04243377e-01
-2.00400531e-01 8.84675264e-01 -1.36276388e+00 1.09765649e+00
7.73404762e-02 4.76496339e-01 7.20706433e-02 -8.03542674e-01
8.60789478e-01 6.69391870e-01 3.97536010e-01 -6.17776334e-01
-1.59461230e-01 7.43668318e-01 -3.27372178e-02 -6.82257414e-01
8.14227879e-01 -2.38566831e-01 -2.83938617e-01 2.95442134e-01
1.91723883e-01 7.14604110e-02 4.27275211e-01 2.50714004e-01
1.16901398e+00 -4.16653335e-01 6.23132527e-01 -5.05941212e-01
4.48506445e-01 6.32049978e-01 1.01309311e+00 9.55197096e-01
-2.85463214e-01 3.27240288e-01 8.20452929e-01 -4.86086577e-01
-9.72750604e-01 -5.80671549e-01 -6.60075068e-01 1.08140135e+00
-5.26959263e-02 -6.87684298e-01 -5.88672400e-01 -6.03358090e-01
1.43352717e-01 8.39181483e-01 -4.27268058e-01 2.46976465e-01
-5.59385836e-01 -8.88809800e-01 8.43963802e-01 5.62952578e-01
2.05124244e-01 -1.29978764e+00 -7.97802389e-01 5.75518727e-01
-8.31735611e-01 -1.41984928e+00 1.77038580e-01 3.51376027e-01
-6.41240716e-01 -1.35224569e+00 -1.54400766e-01 -7.81324804e-01
3.29948604e-01 -5.00815153e-01 1.40445018e+00 4.72627841e-02
-1.31768763e-01 5.12641609e-01 -3.59052956e-01 -6.82561755e-01
-4.18506712e-01 2.36001715e-01 1.83591515e-01 -5.66319227e-01
1.16999102e+00 -7.33834207e-02 -2.98989937e-02 3.66153359e-01
-3.99130553e-01 -3.90497178e-01 2.04454139e-01 9.15024698e-01
4.43882853e-01 2.36318216e-01 6.81724131e-01 -1.27108240e+00
8.30045700e-01 -4.76615697e-01 -2.27537021e-01 4.95661139e-01
-8.49339664e-01 -5.95122911e-02 -2.06339374e-01 -1.62799463e-01
-1.08492732e+00 -1.24885865e-01 -1.23849086e-01 3.79525453e-01
-3.15941334e-01 6.96317911e-01 -6.07406944e-02 1.62066981e-01
9.40830588e-01 -2.17795417e-01 -1.92453623e-01 -6.30972266e-01
-2.75607139e-01 8.65536928e-01 6.31710291e-01 -8.65088224e-01
8.98810774e-02 1.00214869e-01 -3.44404221e-01 -6.82996333e-01
-1.28291118e+00 -8.40376318e-01 -4.16497141e-01 1.25045821e-01
7.30069518e-01 -7.80066431e-01 -6.25837743e-01 4.54424471e-02
-1.29577887e+00 -9.57808271e-03 -3.36014777e-01 3.55373889e-01
-4.00248349e-01 1.87152609e-01 -5.71129084e-01 -9.16765034e-01
-4.53036129e-01 -7.07981646e-01 6.64258718e-01 4.32284266e-01
-9.08333600e-01 -7.54819512e-01 4.83067371e-02 6.10041380e-01
2.96225250e-01 1.24743609e-02 1.10860360e+00 -1.34838569e+00
2.79458165e-01 -2.87586004e-01 -1.33603722e-01 -2.37879023e-01
-7.03457370e-02 -1.65463611e-01 -7.09038675e-01 -3.62748504e-02
-7.91548714e-02 -2.79647827e-01 4.66077983e-01 1.23582035e-01
5.81643343e-01 -4.53542113e-01 -7.63094187e-01 -1.46179691e-01
1.27271271e+00 3.81701827e-01 5.38221657e-01 9.97545838e-01
1.49368867e-01 1.11045921e+00 7.82952487e-01 2.84375638e-01
2.89820075e-01 9.08190191e-01 -1.13537125e-01 5.43244064e-01
2.68961430e-01 -1.41926594e-02 -9.25245956e-02 1.63736239e-01
-5.21102324e-02 -9.41099375e-02 -1.52604568e+00 9.48632956e-01
-1.97689962e+00 -1.04025257e+00 -3.38301927e-01 2.12803054e+00
1.09403360e+00 4.43811864e-01 9.15281326e-02 2.15616956e-01
9.67803180e-01 -3.47514421e-01 -1.72217786e-01 -4.85869855e-01
-2.54961699e-01 1.94658771e-01 4.90154177e-01 4.95897740e-01
-1.10074735e+00 8.79715979e-01 6.76793337e+00 4.10393149e-01
-3.38494480e-01 2.58905124e-02 2.78886229e-01 7.20870569e-02
4.00292547e-03 2.76892006e-01 -1.40888333e+00 3.90760452e-01
1.16966105e+00 -5.02060115e-01 6.20571412e-02 6.53705597e-01
-5.47826707e-01 -3.39697570e-01 -1.01494813e+00 6.97134852e-01
5.63280703e-03 -1.37021399e+00 -2.14731947e-01 6.36649691e-03
4.40326482e-01 -3.35983634e-02 -4.79033351e-01 5.49474895e-01
3.35373372e-01 -9.27481890e-01 6.78925395e-01 4.20739532e-01
3.31367284e-01 -6.30694032e-01 1.08678937e+00 4.30245131e-01
-6.65971577e-01 -3.51724446e-01 -4.00229007e-01 -8.34827125e-02
1.79892495e-01 5.60397863e-01 -8.19434881e-01 4.02014554e-01
1.09036469e+00 1.66236341e-03 -4.20771688e-01 1.14711940e+00
-9.38350037e-02 6.11494839e-01 -6.53479636e-01 -1.66700751e-01
2.21956577e-02 4.15679246e-01 7.41200507e-01 1.34803998e+00
9.84950289e-02 4.48108792e-01 3.64539586e-02 5.12193501e-01
-7.47793615e-02 3.07142556e-01 -2.66240627e-01 -1.02936074e-01
1.16207111e+00 1.06705010e+00 -6.52071178e-01 -5.17073214e-01
-1.29667610e-01 3.83779973e-01 4.81148601e-01 5.43777458e-02
-2.70057887e-01 -5.90256751e-01 3.96875083e-01 2.20179945e-01
1.54176533e-01 4.16253746e-01 -2.83733338e-01 -7.30153024e-01
1.48053944e-01 -9.11874413e-01 1.11160862e+00 -5.74519932e-01
-1.19257247e+00 3.75753403e-01 1.19110532e-01 -4.64091122e-01
-5.18972397e-01 -4.27168757e-01 -1.25009000e-01 8.82615685e-01
-1.07940149e+00 -4.20892984e-01 7.23837037e-03 2.88403720e-01
4.27534468e-02 -2.05337614e-01 1.29655504e+00 5.05196214e-01
-4.41933811e-01 4.70011264e-01 -1.98929504e-01 2.77290553e-01
9.89203691e-01 -1.32359898e+00 1.80866376e-01 3.99005383e-01
-6.05687546e-03 7.11910963e-01 1.18563795e+00 -6.87376738e-01
-7.85227001e-01 -4.39743906e-01 2.00158644e+00 -9.33706880e-01
7.01259673e-01 1.12506680e-01 -1.05035329e+00 7.47557402e-01
1.58609934e-02 -3.60672474e-01 8.29971671e-01 7.03827441e-01
-3.24628651e-01 3.40813667e-01 -1.16494858e+00 1.01758353e-01
8.34150374e-01 -5.41615248e-01 -1.16110694e+00 4.99909103e-01
4.76452798e-01 -7.66339123e-01 -1.05470204e+00 5.01176417e-01
3.55545104e-01 -6.38308108e-01 9.67426836e-01 -1.06792974e+00
1.23734906e-01 -6.63445219e-02 -2.80327111e-01 -8.15128922e-01
-3.34797859e-01 -4.95641708e-01 -1.06964260e-01 1.77856755e+00
6.67941332e-01 -6.30347311e-01 6.50011301e-01 1.31514823e+00
1.53207153e-01 -3.28022212e-01 -1.22047734e+00 -6.00927949e-01
1.45071223e-01 -1.87271252e-01 6.08877897e-01 1.23558605e+00
6.45037413e-01 5.62184572e-01 2.34831870e-03 1.68188512e-01
4.35462683e-01 -1.06865903e-02 4.14900571e-01 -1.61315811e+00
5.17575964e-02 -4.29322749e-01 -4.32153821e-01 -3.52180630e-01
1.03770211e-01 -6.35157824e-01 -3.79804261e-02 -1.54354930e+00
4.03374106e-01 -6.57234550e-01 -2.13177070e-01 7.17300355e-01
-2.65330791e-01 -1.20628856e-01 -2.08931677e-02 5.74008763e-01
-6.05624855e-01 -2.28425503e-01 4.95184213e-01 1.20223284e-01
-1.97529271e-01 -1.58030093e-01 -9.86454189e-01 6.90409482e-01
7.11092472e-01 -8.80994260e-01 2.51150161e-01 -2.27147445e-01
7.35842884e-01 2.13030770e-01 2.25817785e-01 -8.15408528e-01
6.65464640e-01 -4.62505929e-02 1.69747055e-01 -5.18473566e-01
1.07643433e-01 -5.75330198e-01 2.68681884e-01 3.03469628e-01
-4.60355401e-01 9.64988396e-02 3.74284238e-01 2.39312261e-01
-2.97717154e-01 -6.55732095e-01 3.47442836e-01 -1.96144968e-01
-7.26198733e-01 -5.09299994e-01 -2.92439163e-01 4.71909791e-01
8.83225501e-01 3.13173980e-02 -5.09772062e-01 4.56825532e-02
-1.03508937e+00 3.27776641e-01 1.25593677e-01 2.75795072e-01
1.70994513e-02 -9.87909913e-01 -7.20661521e-01 -4.83039111e-01
1.96794614e-01 -9.11408737e-02 4.92636301e-03 8.12088728e-01
-4.34578717e-01 8.67103159e-01 -2.66296178e-01 -1.69239044e-01
-1.30472982e+00 2.63389260e-01 4.04905230e-01 -5.29304266e-01
-5.74892044e-01 8.09270501e-01 -5.41887403e-01 -5.43022692e-01
4.19824690e-01 3.22648972e-01 -7.48323977e-01 5.18434227e-01
7.47829199e-01 4.45609361e-01 3.63393039e-01 -8.02655280e-01
-7.81705320e-01 3.79374065e-02 -2.41963103e-01 -1.38784617e-01
1.34335911e+00 3.15780332e-03 -3.41573983e-01 5.06674170e-01
4.61821049e-01 -5.06001078e-02 -3.07878911e-01 -5.32600164e-01
7.78550327e-01 -2.40675196e-01 -7.06340969e-02 -1.14022744e+00
-4.05687779e-01 4.28744674e-01 2.25852624e-01 6.20156050e-01
4.84348416e-01 2.71170139e-01 5.05952299e-01 6.30983353e-01
3.74475837e-01 -1.60379481e+00 -5.54489434e-01 6.65173709e-01
6.69745207e-01 -1.01914930e+00 6.49636388e-02 -6.86167657e-01
-5.17657280e-01 1.14379930e+00 6.70595765e-01 6.37621105e-01
3.54682267e-01 3.20076913e-01 1.30388746e-02 -6.92513645e-01
-8.04993868e-01 -9.77069214e-02 1.10582300e-02 2.82915801e-01
6.03294790e-01 -3.96630988e-02 -9.31796610e-01 8.62902701e-01
-4.04443264e-01 -1.13050178e-01 4.25539285e-01 1.07010150e+00
-5.99645436e-01 -9.80434239e-01 -5.07380724e-01 7.30753362e-01
-9.80795801e-01 2.00883020e-03 -7.87138522e-01 8.40369940e-01
-5.01543768e-02 1.08070934e+00 2.60198236e-01 -7.82092139e-02
7.49311984e-01 6.49239361e-01 2.65324682e-01 -7.51471102e-01
-8.60545099e-01 -3.85205448e-01 9.79511857e-01 -4.27112222e-01
-5.58804274e-01 -1.26009285e+00 -1.41882408e+00 -3.69087130e-01
-5.49324334e-01 7.71967530e-01 4.65557396e-01 1.19427550e+00
2.32496917e-01 9.75555182e-02 -2.39256471e-01 -6.23720549e-02
-6.49287939e-01 -1.10489857e+00 -5.26774943e-01 6.46523893e-01
-2.61172980e-01 -9.96947706e-01 -3.61206502e-01 6.78207725e-04] | [9.34967041015625, 9.168758392333984] |
e2d0a22a-015e-4587-8c8d-42ab56ea8f29 | building-a-public-domain-voice-database-for | null | null | https://doi.org/10.1145/3487553.3524931 | https://wikiworkshop.org/2022/papers/WikiWorkshop2022_paper_13.pdf | Building a Public Domain Voice Database for Odia | Projects like Mozilla Common Voice were born to address the challenges of unavailability of voice data or the high cost of available data for use in speech technology such as Automatic Speech Recognition (ASR) research and application development. The pilot detailed in this paper is about creating a large freely-licensed public repository of transcribed speech in the Odia language as such a repository was not known to be available. The strategy and methodology behind this process are based on the OpenSpeaks project. Licensed under a Public Domain Dedication (CC0 1.0), the repository currently includes audio recordings of pronunciations for more than 55,000 unique words in Odia, including more than 5,600 recordings of words in the northern Odia dialect Baleswari. No known public listing of words in this dialect was found by the author prior to this pilot. This repository is arguably the most extensive transcribed speech corpus in Odia that is also available publicly under any free and open license. This paper details the strategy, approach, and process behind building both the text and the speech corpus using many open source tools such as Lingua Libre, which can be helpful in building text and speech data for different low-medium-resource languages. | ['Subhashish Panigrahi'] | 2022-08-16 | null | null | null | www-22-companion-proceedings-of-the-web | ['automatic-speech-recognition'] | ['speech'] | [-2.71858931e-01 2.08549440e-01 2.58287638e-01 -2.01199517e-01
-1.17381823e+00 -7.37649322e-01 4.65837121e-01 -8.15438852e-03
-5.53939223e-01 6.06808662e-01 7.02458680e-01 -8.40117574e-01
-9.49924663e-02 -2.99007386e-01 -1.17838368e-01 -4.51076299e-01
2.23873660e-01 6.28188133e-01 -2.88032647e-02 -4.66353476e-01
1.38393193e-01 5.64089239e-01 -1.60303557e+00 4.86545311e-03
4.11887646e-01 3.63302916e-01 7.75200069e-01 7.79153585e-01
-2.37645522e-01 5.11668861e-01 -8.16258192e-01 -1.84757262e-01
-3.20450738e-02 -3.90598804e-01 -1.25805426e+00 5.91724142e-02
7.06397071e-02 -3.45343053e-01 -3.56180221e-01 6.26550615e-01
9.65556324e-01 2.42177233e-01 3.58842835e-02 -6.06702805e-01
-4.49824244e-01 9.41840529e-01 4.30750579e-01 5.68525612e-01
4.98149663e-01 1.22084960e-01 1.03671527e+00 -9.51696634e-01
6.83621466e-01 1.26296639e+00 3.23524594e-01 3.82072896e-01
-6.68444633e-01 -5.80788136e-01 -1.84399650e-01 -1.89105198e-01
-1.52826989e+00 -1.20203853e+00 3.95189494e-01 -4.36268598e-01
1.31628776e+00 3.71960402e-01 5.41049182e-01 1.18307805e+00
-4.62152690e-01 5.41531324e-01 1.03359938e+00 -8.10414791e-01
1.12987071e-01 4.42084968e-01 -6.92822412e-02 2.32185930e-01
-1.14356831e-01 -1.28677040e-01 -6.97813392e-01 -2.84013957e-01
5.51887810e-01 -6.40134335e-01 -3.09351027e-01 4.13143098e-01
-1.19309890e+00 7.13175297e-01 -4.92515057e-01 7.91457474e-01
-2.04623312e-01 -4.11858141e-01 6.90911829e-01 6.47747755e-01
6.16361022e-01 8.67994726e-02 -6.45524204e-01 -9.76867318e-01
-7.69391060e-01 2.22739995e-01 9.16271508e-01 8.99426162e-01
3.39247406e-01 3.82628918e-01 6.16986871e-01 1.66862285e+00
6.53020263e-01 5.18547118e-01 8.11051965e-01 -7.27313459e-01
5.40264964e-01 9.31136534e-02 -1.71112642e-01 -3.28742206e-01
-1.73230946e-01 2.45237909e-03 -1.42379459e-02 -2.22039476e-01
5.30799985e-01 -4.07664716e-01 -7.65197337e-01 1.38481498e+00
4.19248551e-01 -4.08710092e-01 3.76082003e-01 7.12772131e-01
1.11154342e+00 8.51694286e-01 -3.03652734e-01 -2.29408860e-01
1.49913752e+00 -4.17573094e-01 -1.00507247e+00 -2.26974100e-01
5.94402432e-01 -1.58232749e+00 1.05370104e+00 4.06390160e-01
-1.33627009e+00 -2.11924270e-01 -8.65249395e-01 -7.89402425e-02
-5.36028266e-01 -1.56612933e-01 4.28793222e-01 1.17180049e+00
-1.30361569e+00 8.41929093e-02 -6.87691391e-01 -8.33207011e-01
4.68532704e-02 1.45504653e-01 -5.62554359e-01 -3.90216038e-02
-1.23450649e+00 9.86479342e-01 5.60025461e-02 -1.32087767e-01
-5.93104482e-01 -4.12106901e-01 -8.82153749e-01 -4.25327986e-01
2.67574519e-01 2.92629927e-01 1.58679986e+00 -6.72522604e-01
-1.62879932e+00 1.19633961e+00 -9.60591137e-02 -2.63097197e-01
3.40134084e-01 -4.39567678e-02 -8.54337931e-01 2.62629334e-02
1.09402858e-01 1.72310084e-01 3.26327413e-01 -5.95442116e-01
-6.25331461e-01 -4.03132468e-01 -3.73155862e-01 1.31869927e-01
-1.31756514e-01 9.62742269e-01 -2.09877610e-01 -7.29503572e-01
-5.75514808e-02 -9.45987046e-01 3.59599978e-01 -7.67248690e-01
-6.51933551e-02 -2.78677166e-01 6.54114366e-01 -1.18764842e+00
1.23989129e+00 -2.31295133e+00 -4.60606962e-01 -5.71977310e-02
-3.96608680e-01 4.56446350e-01 1.64597049e-01 1.16463578e+00
-3.26539204e-03 2.09592178e-01 -8.72722417e-02 -5.07230014e-02
4.38172407e-02 4.16975141e-01 -3.74415725e-01 7.73383796e-01
-1.47087127e-01 2.55100816e-01 -8.97980630e-01 5.89252897e-02
3.86523902e-01 6.20656192e-01 1.14249270e-02 -2.33857594e-02
3.46927106e-01 2.97743112e-01 9.41936597e-02 4.19922292e-01
5.02125919e-01 6.83298111e-01 2.27238208e-01 6.14927530e-01
-9.16087866e-01 1.39283717e+00 -1.37236369e+00 1.43129992e+00
-6.47062778e-01 9.74025548e-01 5.54917991e-01 -5.68968654e-01
1.00836539e+00 1.00405991e+00 1.02582067e-01 -1.12185664e-01
8.66612867e-02 8.39044750e-01 4.45873678e-01 -3.84721696e-01
7.45073915e-01 -3.24397773e-01 1.89960897e-01 4.62942004e-01
3.02984416e-01 -4.94335383e-01 5.61106317e-02 7.92376976e-03
9.22160327e-01 -1.78739831e-01 3.43946695e-01 -5.48889756e-01
5.56535423e-01 -7.10693151e-02 2.51970083e-01 2.68670917e-01
-2.81533450e-01 5.47227740e-01 1.69499055e-01 1.74928695e-01
-1.21407342e+00 -6.19791925e-01 -8.59470129e-01 1.01080227e+00
-7.98172891e-01 -6.69826031e-01 -6.47189975e-01 2.57891566e-01
-3.81396860e-01 8.88423145e-01 1.74846530e-01 4.86935049e-01
-4.29790080e-01 -7.59579092e-02 1.03335857e+00 -1.01991203e-02
9.13606882e-02 -1.29483664e+00 -2.12104529e-01 3.67367715e-01
-2.37215891e-01 -1.12636709e+00 -4.20359194e-01 9.67694148e-02
-4.34724838e-01 -4.68770057e-01 -7.61498809e-01 -9.64912117e-01
4.47808951e-02 3.00709933e-01 5.32705188e-01 -1.70468569e-01
-1.83683261e-01 5.71488440e-01 -4.28860158e-01 -5.74473560e-01
-1.01764262e+00 3.72321792e-02 2.72869140e-01 -2.96780437e-01
5.27586102e-01 -3.39966059e-01 1.03703909e-01 8.36994424e-02
-5.81280231e-01 -4.26729769e-01 1.93784103e-01 5.27070463e-01
8.13112259e-02 -8.47450718e-02 8.97490621e-01 -5.87157011e-01
7.45725513e-01 -6.42991066e-01 -4.84883279e-01 -3.48303169e-01
-2.87280917e-01 -5.43859899e-01 2.42460623e-01 3.96896638e-02
-1.01124084e+00 -2.48803273e-01 -9.83650148e-01 6.67264387e-02
-5.94188750e-01 6.50692701e-01 -2.25809231e-01 3.85008007e-01
3.27516139e-01 9.61088464e-02 1.32579789e-01 -8.38046312e-01
1.99626431e-01 1.78209996e+00 4.26847667e-01 -3.84209454e-01
4.14955050e-01 1.22430012e-01 -6.78812206e-01 -1.70277786e+00
-1.16948418e-01 -9.27649438e-01 -4.76062953e-01 -2.19073683e-01
5.34526646e-01 -1.13419759e+00 -3.74442816e-01 7.16268182e-01
-8.74968231e-01 -3.77734900e-01 -2.94831246e-01 8.09186935e-01
-5.39991558e-01 1.27345204e-01 -4.20402765e-01 -1.18583453e+00
-9.49398205e-02 -1.18832219e+00 7.42987692e-01 -2.50028372e-01
-6.26871288e-01 -1.09076786e+00 2.31766835e-01 5.29407263e-01
4.40573454e-01 -3.71914685e-01 6.34784758e-01 -1.02795768e+00
-1.10091560e-01 -1.80208609e-01 3.61060321e-01 7.15327978e-01
4.39827561e-01 2.72454679e-01 -1.34970224e+00 -3.38380903e-01
-1.05545945e-01 -3.08957577e-01 2.43004173e-01 1.61455736e-01
1.82450846e-01 -4.74407107e-01 3.20093393e-01 -1.91765860e-01
9.87512171e-01 4.16722089e-01 5.83694696e-01 2.85998613e-01
3.20206404e-01 1.08395481e+00 3.01648498e-01 5.25058627e-01
5.15361309e-01 7.13858128e-01 -1.31479844e-01 5.78888774e-01
-3.06038976e-01 -1.55991837e-01 9.03478384e-01 1.50757360e+00
1.14114389e-01 -1.03712656e-01 -1.30516422e+00 1.41273022e+00
-1.28487194e+00 -9.56182420e-01 -4.02502567e-01 2.49777770e+00
9.58625197e-01 -1.24904089e-01 4.50897604e-01 3.36046845e-01
7.58323610e-01 2.31813222e-01 2.13050663e-01 -8.41437161e-01
-1.38165116e-01 3.82654935e-01 2.63564765e-01 9.59725738e-01
-5.85804760e-01 1.05087340e+00 6.05549479e+00 7.90958285e-01
-1.10381222e+00 3.63730609e-01 1.53687913e-02 -2.21892953e-01
-3.35552096e-01 1.58696532e-01 -8.43363822e-01 2.86072195e-01
1.91727936e+00 -3.58186603e-01 7.54191458e-01 7.69459784e-01
6.37729883e-01 -4.19344082e-02 -5.02587020e-01 9.58327591e-01
-4.73909341e-02 -1.21162498e+00 -4.45766300e-01 3.80904675e-01
2.63070405e-01 7.17270851e-01 -1.31351694e-01 2.02363387e-01
4.07083243e-01 -7.93752968e-01 1.05292737e+00 -1.45139858e-01
7.00857580e-01 -9.58049119e-01 6.14692032e-01 3.77773106e-01
-8.65858436e-01 1.66703090e-01 -3.70144159e-01 -1.95019767e-01
2.93341398e-01 4.22211856e-01 -1.05492771e+00 4.53665018e-01
6.24193430e-01 2.54989386e-01 -1.79674163e-01 9.53866303e-01
-9.26955789e-02 1.35022485e+00 -2.56079227e-01 4.05919068e-02
4.13437665e-01 -1.22589156e-01 9.84063447e-01 1.47879469e+00
4.20464694e-01 -8.88999328e-02 -1.28117636e-01 1.94121629e-01
-9.38089006e-03 6.58161938e-01 -8.38805079e-01 -4.47176814e-01
7.59698093e-01 1.12462437e+00 -4.71337080e-01 -9.59268585e-02
-7.03309953e-01 6.17751837e-01 -1.28611540e-02 4.81819175e-02
-8.31448212e-02 -5.99554121e-01 1.00938344e+00 4.96024221e-01
1.60879329e-01 -5.16576111e-01 -1.78176779e-02 -5.96796274e-01
-6.91804811e-02 -1.17924118e+00 1.24447033e-01 -7.39979506e-01
-1.01214862e+00 7.64062464e-01 1.20292217e-01 -8.54644001e-01
-5.70946157e-01 -6.23450577e-01 -2.91375309e-01 1.47473252e+00
-9.78916526e-01 -8.57069850e-01 2.46922418e-01 3.95607233e-01
8.43268752e-01 -5.31110346e-01 1.09800923e+00 4.60980326e-01
-4.84641075e-01 1.31642625e-01 3.77151191e-01 1.06921099e-01
7.19783306e-01 -1.15685701e+00 6.76740050e-01 4.69314903e-01
1.38284758e-01 9.13201630e-01 6.96210682e-01 -4.62473541e-01
-1.14386344e+00 -7.15287268e-01 1.52484775e+00 -3.74221087e-01
1.30578411e+00 -6.03240728e-01 -8.51781905e-01 9.14122939e-01
6.32558227e-01 -6.19163930e-01 9.99071479e-01 1.21117495e-01
5.75642958e-02 1.79851249e-01 -9.20333385e-01 4.90610242e-01
7.63877451e-01 -9.73100126e-01 -8.85891199e-01 4.71024066e-01
6.07463121e-01 -4.01422650e-01 -8.99108768e-01 -3.38669330e-01
4.09935951e-01 -5.22571802e-01 3.34137827e-01 -3.27047646e-01
-1.49725214e-01 -2.38998830e-01 -4.31276292e-01 -1.45526648e+00
8.17406848e-02 -1.38105285e+00 5.26188791e-01 1.96222556e+00
3.90124649e-01 -1.17081594e+00 -3.98907363e-02 2.35221356e-01
-6.70970738e-01 -2.10398257e-01 -1.41746950e+00 -8.53164554e-01
3.35516512e-01 -1.08177531e+00 5.04383385e-01 9.47028816e-01
2.79975295e-01 1.81336224e-01 -7.68317282e-02 1.33468658e-01
2.79512275e-02 -6.50386214e-01 6.88332736e-01 -9.79459405e-01
-1.47081479e-01 -2.89636821e-01 -5.61573625e-01 -7.62107313e-01
9.54484716e-02 -9.62262034e-01 -4.72636111e-02 -1.55729330e+00
-5.94792128e-01 -3.90164644e-01 3.25661898e-01 5.11603594e-01
4.23343569e-01 -1.09815829e-01 1.22528993e-01 3.78582805e-01
4.42477882e-01 2.51237631e-01 8.50711107e-01 2.94179529e-01
-3.92515898e-01 5.88624366e-02 -8.46512437e-01 6.47558451e-01
8.05018544e-01 -4.41275418e-01 -4.01755244e-01 -2.73737818e-01
-2.08917379e-01 -2.97420584e-02 3.59635754e-03 -7.45144606e-01
-1.69422210e-03 -7.04246238e-02 -3.89039576e-01 -6.40292227e-01
4.02345985e-01 -4.57171053e-01 2.78196990e-01 -2.15192661e-02
-9.33396369e-02 5.45150042e-02 5.94410419e-01 -2.05446228e-01
-3.42187911e-01 -5.58712602e-01 8.42084050e-01 -3.77443939e-01
-4.99899536e-01 -1.40356228e-01 -1.12627208e+00 2.65769184e-01
7.74201453e-01 -1.74891070e-01 -2.26924568e-01 -6.93111241e-01
-6.27980709e-01 -2.15231985e-01 3.15555900e-01 7.28006721e-01
2.59871900e-01 -1.01666749e+00 -9.21423912e-01 3.69464874e-01
2.01572180e-01 -3.56577039e-01 2.21072733e-01 7.55724609e-01
-6.51034892e-01 7.80510485e-01 5.37764886e-03 -4.18739170e-02
-1.29194248e+00 -1.25516966e-01 2.33098119e-01 4.67587143e-01
-9.56650734e-01 6.53012276e-01 -3.25730264e-01 -8.50651562e-01
7.92199224e-02 5.95116355e-02 -7.20403939e-02 2.88152158e-01
6.53356731e-01 4.65438932e-01 3.68010700e-01 -1.39951885e+00
-4.90537614e-01 -1.93539724e-01 -1.53748959e-01 -9.73284364e-01
1.35481703e+00 -5.95764637e-01 -2.32592091e-01 1.16175985e+00
1.15231478e+00 6.30481839e-01 -5.43753088e-01 1.23247638e-01
-1.45484917e-02 -3.98238599e-01 4.29764867e-01 -6.44585788e-01
-4.42696691e-01 7.95602262e-01 3.60385329e-01 3.94021630e-01
5.38545191e-01 9.41992328e-02 8.62381577e-01 1.45310894e-01
2.56422639e-01 -1.36997175e+00 -7.98895180e-01 1.05499077e+00
1.09692955e+00 -7.15887249e-01 -3.83231759e-01 -1.64807782e-01
-7.66611755e-01 1.14999270e+00 -1.10254034e-01 3.90469402e-01
7.59215593e-01 3.01762462e-01 5.04209161e-01 -1.30714759e-01
-6.95838749e-01 -5.51272452e-01 -1.47691891e-01 7.84099638e-01
9.71875429e-01 1.19438045e-01 -5.41958332e-01 4.14381504e-01
-9.21166837e-01 -3.93252999e-01 8.81467700e-01 9.39600229e-01
-3.82704914e-01 -1.32133269e+00 -5.82080424e-01 8.66172686e-02
-8.62600625e-01 -4.12506521e-01 -5.13399541e-01 8.52917075e-01
-2.54359692e-01 1.35275316e+00 9.02661607e-02 -7.86213428e-02
3.38647246e-01 4.53658700e-01 1.85210243e-01 -9.99815345e-01
-7.26553738e-01 3.56678545e-01 8.12452674e-01 1.72123879e-01
-1.40614435e-01 -1.07750595e+00 -1.29224193e+00 -6.11968756e-01
-2.70776391e-01 2.06747115e-01 1.25675666e+00 9.45890903e-01
4.47908081e-02 2.69313663e-01 4.73595977e-01 -6.59659803e-01
-3.70945901e-01 -1.20874548e+00 -9.53660846e-01 -1.61407620e-01
2.76768714e-01 -3.19360763e-01 -5.98589182e-01 -2.69951969e-02] | [14.245469093322754, 6.987270355224609] |
eee0135b-e983-45cd-87d3-8443eed4bd5b | a-video-based-end-to-end-pipeline-for-non | 2303.16867 | null | https://arxiv.org/abs/2303.16867v1 | https://arxiv.org/pdf/2303.16867v1.pdf | A Video-based End-to-end Pipeline for Non-nutritive Sucking Action Recognition and Segmentation in Young Infants | We present an end-to-end computer vision pipeline to detect non-nutritive sucking (NNS) -- an infant sucking pattern with no nutrition delivered -- as a potential biomarker for developmental delays, using off-the-shelf baby monitor video footage. One barrier to clinical (or algorithmic) assessment of NNS stems from its sparsity, requiring experts to wade through hours of footage to find minutes of relevant activity. Our NNS activity segmentation algorithm solves this problem by identifying periods of NNS with high certainty -- up to 94.0\% average precision and 84.9\% average recall across 30 heterogeneous 60 s clips, drawn from our manually annotated NNS clinical in-crib dataset of 183 hours of overnight baby monitor footage from 19 infants. Our method is based on an underlying NNS action recognition algorithm, which uses spatiotemporal deep learning networks and infant-specific pose estimation, achieving 94.9\% accuracy in binary classification of 960 2.5 s balanced NNS vs. non-NNS clips. Tested on our second, independent, and public NNS in-the-wild dataset, NNS recognition classification reaches 92.3\% accuracy, and NNS segmentation achieves 90.8\% precision and 84.2\% recall. | ['Sarah Ostadabbas', 'Emily Zimmerman', 'Rebecca A. Schwartz-Mette', 'Marie J. Hayes', 'Matthew S. Goodwin', 'Emma C. Grace', 'Cassandra B. Rowan', 'Cholpady Vikram Kamath', 'Samuel Zlota', 'Kashish Jain', 'Elaheh Hatamimajoumerd', 'Michael Wan', 'Shaotong Zhu'] | 2023-03-29 | null | null | null | null | ['action-recognition-in-videos'] | ['computer-vision'] | [ 6.37450457e-01 4.35119122e-01 -6.02239788e-01 -5.34592390e-01
-1.04706049e+00 -7.37823248e-01 -2.47447342e-01 5.61235487e-01
-2.73182690e-01 -1.42520860e-01 1.29560158e-01 -5.69758900e-02
-7.86914900e-02 -4.23489362e-01 -1.21767986e+00 -4.31572616e-01
-2.91989952e-01 1.70731336e-01 3.79008502e-01 3.58524621e-01
-6.95981756e-02 3.39582302e-02 -1.37897027e+00 9.09091949e-01
2.49863416e-01 1.35764205e+00 -1.65416077e-01 1.36652589e+00
1.41677961e-01 8.09056938e-01 -4.43372399e-01 -2.02952459e-01
2.98854679e-01 -3.75378966e-01 -3.18891078e-01 -6.50599778e-01
9.35567737e-01 -7.15326428e-01 -1.36672437e-01 4.79976028e-01
7.78859496e-01 -3.86920989e-01 5.75088978e-01 -9.77616429e-01
-1.72849491e-01 4.01027173e-01 -6.80485964e-01 8.18308473e-01
7.61353910e-01 1.39398068e-01 6.25212252e-01 -3.31748068e-01
6.47537947e-01 3.62878025e-01 1.19633651e+00 8.47691476e-01
-9.45242107e-01 -7.57121205e-01 -4.02591377e-01 -8.42020735e-02
-1.00718784e+00 -8.79717112e-01 3.56270730e-01 -7.40152657e-01
1.33432567e+00 2.04299867e-01 1.43241560e+00 1.28596425e+00
-8.60403478e-02 9.44633961e-01 4.30757284e-01 -1.99758857e-01
2.99935043e-01 -5.88845968e-01 -2.26765290e-01 8.40270996e-01
2.38330886e-01 -3.06632500e-02 -1.06401801e+00 -6.37253970e-02
3.88630360e-01 -3.58280353e-02 -6.95559233e-02 -1.71778619e-01
-1.01631463e+00 4.31485355e-01 -2.44914383e-01 1.55200392e-01
-3.75918686e-01 3.19328785e-01 5.62693954e-01 -5.46993166e-02
6.11744881e-01 4.07374531e-01 -7.91370869e-01 -1.05477166e+00
-1.42317605e+00 9.33543816e-02 8.00364554e-01 1.03391862e+00
5.69406375e-02 -3.47896665e-01 -5.92676699e-02 8.43546689e-01
2.78678417e-01 5.51115930e-01 3.17497969e-01 -1.15964532e+00
6.75350785e-01 3.73195022e-01 -3.90829332e-02 -6.74731195e-01
-1.05655050e+00 1.78122167e-02 -2.22462460e-01 -2.01594338e-01
5.06428242e-01 -4.98812914e-01 -8.48153532e-01 1.80984306e+00
5.46136677e-01 5.28503776e-01 -1.74628407e-01 6.18281960e-01
1.33180344e+00 4.98592615e-01 1.26417885e-02 -5.39139211e-01
1.01973605e+00 -6.70202613e-01 -4.57051128e-01 1.15188286e-02
6.83242738e-01 -3.97556245e-01 3.07329595e-01 7.13211298e-01
-1.45351374e+00 -3.19387466e-02 -1.07505870e+00 2.60756493e-01
-1.06832795e-01 -1.59981549e-01 2.86307096e-01 7.92986155e-01
-1.20438254e+00 6.81646645e-01 -1.40292859e+00 -6.07514024e-01
9.48192418e-01 4.12181020e-01 -4.49637115e-01 1.07582234e-01
-6.71343505e-01 5.47720730e-01 -1.30767524e-01 -2.11519957e-01
-9.85048950e-01 -1.41709948e+00 -9.87801492e-01 -2.59056807e-01
2.87623648e-02 -5.73993772e-02 1.79648304e+00 -8.47643256e-01
-1.03890836e+00 1.19791424e+00 -3.14124584e-01 -4.06374663e-01
5.54645658e-01 -6.44963324e-01 -4.75842953e-01 1.03475964e+00
2.78417051e-01 6.40381873e-01 8.29673767e-01 -1.21756934e-01
-6.57093704e-01 -2.84556121e-01 -4.80225652e-01 4.41596247e-02
-1.24303147e-01 5.63631117e-01 -3.96257609e-01 -5.78993499e-01
5.07617652e-01 -5.42653918e-01 2.16685742e-01 4.28709775e-01
1.28259540e-01 -1.05681151e-01 3.65694910e-01 -1.03130126e+00
9.73261178e-01 -1.96864176e+00 -4.89645094e-01 -1.06848367e-01
5.63280642e-01 5.08762181e-01 -2.41479427e-01 1.54776886e-01
-2.53023475e-01 -2.27744371e-01 4.28324491e-02 -1.47618428e-01
-4.06757206e-01 -1.99993104e-01 4.00692493e-01 9.78918314e-01
1.67421728e-01 1.07178748e+00 -1.37057734e+00 -3.73050332e-01
2.78733909e-01 3.52507174e-01 -6.12848043e-01 5.91038108e-01
-5.55832945e-02 1.98258832e-01 -2.84018051e-02 1.30234790e+00
2.79870033e-01 2.13905126e-01 -2.31418401e-01 -1.44413218e-01
-1.43679798e-01 2.94251204e-01 -5.77655017e-01 2.09410787e+00
2.68298090e-01 8.90430152e-01 1.76885445e-02 -1.26461577e+00
3.79552215e-01 6.37133241e-01 1.13936949e+00 -8.02675903e-01
1.77590132e-01 1.38579980e-01 -1.13976903e-01 -1.47568738e+00
-4.63157654e-01 2.86264747e-01 2.05366254e-01 3.17864716e-01
1.75516292e-01 -7.63575584e-02 2.28638560e-01 -2.35325158e-01
1.85909629e+00 4.79523242e-01 2.83226609e-01 -1.81079134e-01
-4.27050620e-01 -1.51694119e-01 6.76800668e-01 8.99133801e-01
-6.04116917e-01 1.27438629e+00 7.27992773e-01 -6.31610930e-01
-8.00827801e-01 -1.04151189e+00 -2.05827326e-01 1.19191992e+00
-4.45029765e-01 -1.65557370e-01 -1.24288690e+00 -6.22345626e-01
-2.40965262e-02 2.66332269e-01 -1.03064430e+00 -2.02321615e-02
-6.47827148e-01 -5.63567758e-01 1.06538475e+00 9.74279344e-01
-1.07715815e-01 -1.03061223e+00 -1.53668332e+00 6.55543506e-01
1.00215852e-01 -1.01770854e+00 -4.57413703e-01 2.21744791e-01
-2.33635053e-01 -1.62099791e+00 -9.51313913e-01 -6.20074689e-01
3.89174402e-01 -8.27429667e-02 8.74636829e-01 -2.43945718e-01
-8.46442461e-01 7.81879246e-01 -6.71352267e-01 -9.55745041e-01
-4.71917093e-01 -3.31237435e-01 2.26706401e-01 -2.28546843e-01
8.89660954e-01 -8.37688625e-01 -1.31815505e+00 5.00057712e-02
-5.12558818e-01 1.40426427e-01 2.05009028e-01 3.19005400e-01
5.92666805e-01 -9.83652949e-01 6.09017670e-01 -2.40845621e-01
-2.26941437e-01 -8.94911408e-01 -6.65497482e-01 2.22409427e-01
-4.04983580e-01 -5.83122432e-01 1.40537187e-01 -6.97570384e-01
-1.15219779e-01 1.84588686e-01 -3.24542880e-01 -6.04169488e-01
-4.71212089e-01 -6.66450383e-03 4.51212227e-01 1.40911266e-01
7.35462427e-01 5.43384328e-02 1.88453630e-01 -3.50209057e-01
-1.29794136e-01 7.30365396e-01 9.04304266e-01 -2.10084215e-01
-7.95878023e-02 3.09409767e-01 -9.56125483e-02 -8.08459699e-01
-9.56166983e-01 -9.27123249e-01 -3.66699576e-01 -4.00893480e-01
1.44282198e+00 -8.46045673e-01 -8.88678670e-01 7.94833899e-01
-1.15329432e+00 -6.58117592e-01 1.32618891e-02 7.14339256e-01
-7.61621475e-01 5.26847132e-03 -6.84288502e-01 -8.76063406e-01
-8.66179407e-01 -7.42823720e-01 1.11720383e+00 3.20488125e-01
-9.08285379e-01 -2.82286376e-01 4.81066257e-01 6.69197679e-01
1.81135789e-01 9.48431313e-01 2.71382362e-01 -1.00796747e+00
1.03326514e-01 -4.23192382e-01 -2.48873562e-01 3.98086071e-01
2.50974111e-02 3.07166964e-01 -1.06098461e+00 -1.14582472e-01
-1.73406556e-01 -2.62973547e-01 3.93043995e-01 8.52765977e-01
9.74982381e-01 -2.07950205e-01 -2.34656826e-01 8.27793360e-01
1.07992017e+00 8.31578374e-01 5.14021032e-02 -7.11353868e-02
6.57369733e-01 6.38299108e-01 7.94181108e-01 5.56850493e-01
3.64809662e-01 3.17169964e-01 5.83658695e-01 7.93983042e-02
-3.63294408e-02 -2.32571125e-01 1.93326071e-01 3.09197545e-01
-1.92192376e-01 -8.43740031e-02 -1.25087798e+00 1.12836540e+00
-1.52386355e+00 -9.96635914e-01 5.05322628e-02 1.91024840e+00
7.88912237e-01 3.96978632e-02 6.01166844e-01 2.13771373e-01
5.36796570e-01 7.58038461e-03 -6.67185724e-01 -6.47405088e-01
2.84960508e-01 4.59280938e-01 5.80764592e-01 -2.58758098e-01
-1.18446410e+00 2.37055585e-01 5.87273264e+00 2.61554241e-01
-1.05611408e+00 2.13415056e-01 7.55502284e-01 -1.11233401e+00
3.52527052e-01 -8.07860017e-01 -5.88258147e-01 9.82892871e-01
1.33331251e+00 6.77340209e-01 5.03523886e-01 9.13491488e-01
4.07490790e-01 -3.57217938e-01 -1.64609969e+00 1.07991552e+00
2.77912021e-01 -1.54323149e+00 -1.06586361e+00 -5.97230613e-01
6.32103086e-01 6.32553697e-01 -4.10948724e-01 -2.98991770e-01
-3.97081316e-01 -1.12999797e+00 1.36126745e+00 3.45077634e-01
1.52756858e+00 -4.86731768e-01 3.22573304e-01 3.20869297e-01
-1.16981423e+00 3.94876711e-02 2.45121390e-01 1.60206318e-01
-6.68611601e-02 2.43970022e-01 -1.02035093e+00 -2.80319065e-01
1.47124088e+00 7.79432714e-01 -1.62634254e-01 1.13262498e+00
1.07552424e-01 1.07609904e+00 -8.61539006e-01 -4.68504637e-01
1.55396521e-01 3.84967208e-01 3.77274901e-01 1.81959665e+00
4.12950516e-01 7.78105021e-01 -7.25677729e-01 4.49481368e-01
1.64355502e-01 -5.94191141e-02 -5.75852692e-01 -1.92564473e-01
2.84562051e-01 7.59644151e-01 -6.30475223e-01 -1.23173438e-01
-7.98474252e-01 5.72823286e-01 1.37842774e-01 3.43009382e-02
-9.68126416e-01 -5.53007185e-01 6.67320609e-01 4.79468495e-01
3.38619828e-01 4.39939201e-01 -2.67849505e-01 -6.36161566e-01
1.97788805e-01 -6.57974243e-01 4.81619328e-01 -7.78949142e-01
-5.42139173e-01 -8.28353018e-02 8.99897590e-02 -1.24183130e+00
-3.20015371e-01 -4.23454076e-01 -8.17136526e-01 2.24551722e-01
-9.57128406e-01 -8.65784407e-01 -4.03421879e-01 -8.56565218e-03
6.52989507e-01 4.01887260e-02 5.73799312e-01 3.89746189e-01
-6.42894626e-01 1.01767397e+00 -3.28222781e-01 3.22888382e-02
2.46481031e-01 -1.12557888e+00 6.21297479e-01 8.09665263e-01
-1.01735055e-01 3.16472083e-01 6.55740023e-01 -6.30260706e-01
-1.48508668e+00 -1.10783398e+00 7.80535877e-01 -9.66170609e-01
4.17921245e-01 -3.00479233e-01 -6.19806945e-01 5.54636657e-01
-2.40734115e-01 2.84968644e-01 9.26585615e-01 -6.41866684e-01
-2.29259193e-01 -1.81684047e-01 -1.59219515e+00 2.09163874e-01
1.46493614e+00 -2.72918224e-01 -6.69441283e-01 4.07341540e-01
7.22174048e-01 -9.99053359e-01 -1.08677208e+00 1.00127399e+00
1.31985855e+00 -9.33023691e-01 8.33470523e-01 -5.00411570e-01
6.88704312e-01 1.02195404e-01 -9.75609869e-02 -5.06458879e-01
2.26110101e-01 -1.08749688e+00 -7.04574943e-01 9.17581081e-01
5.06968319e-01 -4.40517545e-01 1.11026502e+00 8.92768860e-01
-1.94339797e-01 -1.43072331e+00 -1.18375611e+00 -3.03140789e-01
-3.97584707e-01 -1.14031363e+00 5.92482269e-01 7.34765232e-01
3.27780783e-01 -4.42044944e-01 1.81212172e-01 3.24008554e-01
3.61728370e-01 -3.09184283e-01 1.49129808e-01 -6.24193430e-01
-2.78014064e-01 -3.09694022e-01 -6.83675826e-01 -5.43436229e-01
-5.00401318e-01 -4.01278853e-01 1.32140964e-01 -1.49934244e+00
2.20272895e-02 -1.26588538e-01 -4.45829064e-01 8.85344923e-01
4.07253593e-01 5.31242967e-01 -4.70620424e-01 -4.83895481e-01
-9.22277868e-01 -4.49759901e-01 5.07243931e-01 1.04043558e-01
-2.77356625e-01 1.46562368e-01 -3.51110607e-01 8.91084850e-01
6.54936016e-01 -5.36826551e-01 -3.86310667e-01 -4.34765488e-01
2.79530585e-01 5.33564568e-01 1.27755016e-01 -9.67549086e-01
-8.12876672e-02 -1.89469740e-01 6.59314573e-01 -9.37967598e-01
2.62343436e-01 -5.03160357e-01 -2.60357931e-02 6.03456438e-01
-2.84808755e-01 -3.53418857e-01 3.63918155e-01 1.92841321e-01
4.00854647e-01 -1.84636563e-01 8.80590141e-01 8.59972909e-02
-3.14136952e-01 5.93657613e-01 -4.29494917e-01 4.95542943e-01
1.10525262e+00 -5.52209616e-01 -1.41977146e-01 -3.96802090e-03
-7.37123370e-01 1.24075115e-01 2.97560424e-01 7.10711539e-01
6.30388618e-01 -8.09512973e-01 -6.14414275e-01 5.64954638e-01
3.80326033e-01 3.21756124e-01 2.76671648e-01 1.11019182e+00
-1.14311218e+00 3.13281745e-01 2.87819386e-01 -9.67620730e-01
-1.46418869e+00 2.93336034e-01 3.54880959e-01 7.53868103e-01
-7.22788155e-01 1.41183794e+00 -1.18161783e-01 1.79725364e-01
6.21459961e-01 -9.37557697e-01 -1.47110999e-01 1.43338501e-01
8.79443884e-01 7.47414887e-01 2.90386260e-01 -3.72266501e-01
-6.60187125e-01 8.21749389e-01 2.83165485e-01 3.07893064e-02
1.46825755e+00 3.00855875e-01 8.30290914e-02 2.20917568e-01
1.51227164e+00 -3.27964574e-01 -1.40854883e+00 4.38780785e-01
-2.56306082e-01 -1.34674132e-01 -4.45838660e-01 -7.57241189e-01
-9.22669590e-01 7.81998336e-01 9.38654244e-01 3.22386324e-01
9.75326598e-01 4.77734715e-01 1.06043422e+00 1.66973516e-01
9.72850174e-02 -1.09872913e+00 -1.93007082e-01 -1.76111460e-01
8.81904781e-01 -1.27141345e+00 -2.84341842e-01 6.71731727e-03
-1.57094896e-01 9.59949970e-01 6.78574860e-01 8.83043706e-02
4.17003483e-01 3.96083832e-01 2.00491488e-01 -3.59072179e-01
-8.30651522e-01 4.98927057e-01 5.35475910e-01 9.18217957e-01
1.37758359e-01 2.03595266e-01 -1.40428454e-01 9.05812621e-01
-9.80461240e-02 2.71430135e-01 1.75130308e-01 1.08465600e+00
-2.96703577e-01 -4.72773872e-02 1.33484066e-01 1.11310410e+00
-1.11828816e+00 5.73017495e-03 1.15589790e-01 1.93595469e-01
7.15322733e-01 1.03405011e+00 4.56488468e-02 -1.33589953e-01
4.14925784e-01 1.12667702e-01 6.08327568e-01 -7.63945878e-01
-5.55715740e-01 9.18788835e-02 4.45535123e-01 -1.32634485e+00
-2.63594091e-01 -1.37577224e+00 -1.42123330e+00 3.37063700e-01
1.18671030e-01 -3.83371204e-01 9.47786748e-01 8.15645337e-01
1.41534343e-01 7.96593279e-02 5.02054453e-01 -8.00863385e-01
8.41754749e-02 -7.06323981e-01 -4.39694732e-01 -1.26401544e-01
9.61907625e-01 -2.99151391e-01 -3.64929885e-01 1.66080177e-01] | [14.057308197021484, -2.289034605026245] |
0d239958-e8cc-411d-9b37-5fd85a3276b8 | rubik-s-cube-operator-a-plug-and-play | 2203.12921 | null | https://arxiv.org/abs/2203.12921v1 | https://arxiv.org/pdf/2203.12921v1.pdf | Rubik's Cube Operator: A Plug And Play Permutation Module for Better Arranging High Dimensional Industrial Data in Deep Convolutional Processes | The convolutional neural network (CNN) has been widely applied to process the industrial data based tensor input, which integrates data records of distributed industrial systems from the spatial, temporal, and system dynamics aspects. However, unlike images, information in the industrial data based tensor is not necessarily spatially ordered. Thus, directly applying CNN is ineffective. To tackle such issue, we propose a plug and play module, the Rubik's Cube Operator (RCO), to adaptively permutate the data organization of the industrial data based tensor to an optimal or suboptimal order of attributes before being processed by CNNs, which can be updated with subsequent CNNs together via the gradient-based optimizer. The proposed RCO maintains K binary and right stochastic permutation matrices to permutate attributes of K axes of the input industrial data based tensor. A novel learning process is proposed to enable learning permutation matrices from data, where the Gumbel-Softmax is employed to reparameterize elements of permutation matrices, and the soft regularization loss is proposed and added to the task-specific loss to ensure the feature diversity of the permuted data. We verify the effectiveness of the proposed RCO via considering two representative learning tasks processing industrial data via CNNs, the wind power prediction (WPP) and the wind speed prediction (WSP) from the renewable energy domain. Computational experiments are conducted based on four datasets collected from different wind farms and the results demonstrate that the proposed RCO can improve the performance of CNN based networks significantly. | ['Zijun Zhang', 'Zhong Zheng', 'Luoxiao Yang'] | 2022-03-24 | null | null | null | null | ['rubik-s-cube'] | ['graphs'] | [-7.02831894e-02 -5.18543541e-01 3.82767431e-02 -1.95632756e-01
1.74928352e-01 -4.87031460e-01 2.66961575e-01 -1.95403621e-01
-2.36438140e-01 4.76702988e-01 -1.02934673e-01 -3.72554451e-01
-9.53475535e-01 -7.54274070e-01 -7.60136545e-01 -1.08509529e+00
-4.98678029e-01 4.89203967e-02 -2.29357809e-01 -1.54416248e-01
1.88813522e-01 6.42313123e-01 -1.55250955e+00 1.73811600e-01
7.85334527e-01 1.40976477e+00 4.77035820e-01 4.98597980e-01
1.87515095e-02 6.21922970e-01 -6.07365906e-01 4.26253304e-02
7.87678778e-01 6.37372136e-02 -5.10873318e-01 2.77735051e-02
4.43803482e-02 -1.53051883e-01 -2.13715062e-01 1.03862917e+00
4.30963755e-01 3.37815285e-01 3.38741571e-01 -1.43406880e+00
-7.22366214e-01 7.10868239e-01 -3.82461041e-01 2.43823931e-01
-4.27632451e-01 2.73600668e-01 9.11250174e-01 -7.40595639e-01
4.42241013e-01 1.17110455e+00 5.04032314e-01 -2.09605128e-01
-1.01368630e+00 -7.52887487e-01 2.57337660e-01 5.01712441e-01
-1.21040976e+00 1.65937826e-01 1.22636664e+00 -5.36448061e-01
7.79519320e-01 2.68982023e-01 9.23823833e-01 8.01137209e-01
5.83778918e-01 4.24549282e-01 8.24976146e-01 2.56044447e-01
2.49223933e-01 -3.44190150e-01 6.86839735e-03 5.24204493e-01
3.16358149e-01 2.55921066e-01 -4.13846076e-01 1.59793124e-01
8.14388454e-01 3.25295091e-01 -2.39419118e-02 -3.85378480e-01
-1.59358048e+00 6.29190028e-01 7.94734895e-01 1.05187051e-01
-6.33299351e-01 1.27041146e-01 7.28608847e-01 5.83389997e-01
4.11183864e-01 6.42878413e-01 -1.06517804e+00 2.44918298e-02
-4.80049938e-01 1.95098922e-01 6.67273760e-01 7.32223213e-01
5.79239786e-01 3.87535572e-01 -3.74074280e-02 7.36825466e-01
2.26233721e-01 6.16838753e-01 6.34728014e-01 -8.24997485e-01
7.74543166e-01 7.04274714e-01 -6.79338053e-02 -1.46798229e+00
-6.26694322e-01 -6.10587537e-01 -1.45202434e+00 1.17790222e-01
9.93955135e-02 -4.97721493e-01 -7.65936792e-01 1.33499634e+00
3.75162244e-01 1.90517277e-01 -5.58192581e-02 1.11706030e+00
3.96666408e-01 9.71132398e-01 -1.57122359e-01 -2.08894432e-01
1.27372181e+00 -8.14014375e-01 -8.59744966e-01 3.10916990e-01
4.42667395e-01 -7.16277301e-01 8.16645443e-01 5.37186861e-01
-6.41173005e-01 -8.34722877e-01 -1.29461586e+00 2.46429339e-01
-7.00545251e-01 5.20738423e-01 5.52144170e-01 8.51493180e-02
-6.33458257e-01 8.80146205e-01 -9.65204477e-01 5.62511906e-02
4.53042239e-01 3.82663816e-01 -2.98271775e-01 2.95372576e-01
-1.19336712e+00 4.65540618e-01 7.96302497e-01 1.10843384e+00
-7.27932751e-01 -9.54628706e-01 -4.94037688e-01 2.61872381e-01
4.26315397e-01 -6.97784245e-01 6.38744712e-01 -8.27817619e-01
-1.85620642e+00 -1.76610202e-01 5.56111038e-01 -4.08117026e-01
2.67674088e-01 -2.37614289e-01 -4.01873082e-01 -1.83405221e-01
-2.40901083e-01 2.19518483e-01 1.11423922e+00 -9.86886621e-01
-8.31973135e-01 -2.04665363e-01 -8.88159685e-03 1.02270870e-02
-4.85601395e-01 -2.74704069e-01 5.61519749e-02 -9.67008233e-01
3.53379071e-01 -9.99359131e-01 -3.34830791e-01 -2.97471374e-01
-6.74992800e-01 -1.66600287e-01 1.15785563e+00 -7.67167151e-01
1.09213805e+00 -2.20097923e+00 3.75226229e-01 4.41254586e-01
5.49765266e-02 2.25526273e-01 -3.07841450e-01 4.22853380e-01
-4.32146519e-01 -4.35718969e-02 -2.36552179e-01 4.22322266e-02
5.24319895e-02 4.37739164e-01 -2.80111939e-01 3.11459869e-01
3.76729339e-01 7.08892643e-01 -9.63710546e-01 -8.27815663e-03
4.79334861e-01 2.86247760e-01 -2.77903318e-01 1.01370372e-01
-1.83700562e-01 4.75296170e-01 -4.66412872e-01 2.96375543e-01
9.73098338e-01 -2.03862518e-01 3.22239667e-01 -8.65607500e-01
-3.41051549e-01 -3.92041840e-02 -1.42110622e+00 1.35759294e+00
-7.50442505e-01 2.19643980e-01 1.52658507e-01 -1.30161309e+00
9.60085273e-01 2.59062380e-01 9.90907192e-01 -5.67375004e-01
1.77056015e-01 1.09085692e-02 3.32117617e-01 -7.81253695e-01
3.42579544e-01 3.23860824e-01 1.77277088e-01 2.13339582e-01
-9.97457728e-02 -1.17903173e-01 4.02147144e-01 -4.48654890e-01
8.09746981e-01 1.24750927e-01 -2.17815220e-01 -5.03589988e-01
6.24899209e-01 -1.11245662e-02 7.89547741e-01 2.96655566e-01
1.81155637e-01 2.63785213e-01 6.83759153e-01 -1.04330730e+00
-1.14112186e+00 -8.44320893e-01 -1.49284042e-02 8.34075749e-01
2.17990838e-02 1.35977268e-02 -2.75628924e-01 -8.30257714e-01
2.29962260e-01 2.21409053e-01 -5.65406263e-01 -1.76142991e-01
-7.48818398e-01 -1.15050316e+00 6.08637370e-02 4.31973159e-01
7.07795203e-01 -1.09976923e+00 -4.64484960e-01 3.48456115e-01
1.02207541e-01 -9.46715295e-01 -3.45335454e-01 5.69288135e-01
-9.50683773e-01 -1.03400743e+00 -2.35068008e-01 -7.52544940e-01
8.33162427e-01 8.03519338e-02 5.42296946e-01 -2.67973661e-01
-4.49879244e-02 -2.77132630e-01 -4.20356661e-01 -3.87200326e-01
6.89284280e-02 4.43284035e-01 1.43096253e-01 3.47855538e-01
-2.81871617e-01 -7.90114522e-01 -7.11410522e-01 3.19682986e-01
-1.06532645e+00 6.01159409e-02 8.30929458e-01 1.12799275e+00
4.06786710e-01 7.48558164e-01 4.50962275e-01 -5.24285495e-01
7.60727763e-01 -5.60528755e-01 -9.34493840e-01 -8.00413787e-02
-6.60460830e-01 1.06542781e-01 1.25305974e+00 -6.12204254e-01
-6.88744366e-01 2.76027098e-02 2.86001444e-01 -8.14510107e-01
2.32394993e-01 8.90736878e-01 -1.79778576e-01 1.24075569e-01
3.95572521e-02 1.39897048e-01 1.40932694e-01 -5.43572605e-01
2.30591938e-01 3.37778836e-01 3.39697778e-01 -4.28240269e-01
1.15353870e+00 3.20116371e-01 3.33772629e-01 -7.46185482e-01
-5.29646993e-01 -1.55681029e-01 -6.87303424e-01 -1.24502666e-01
7.62009799e-01 -5.95383763e-01 -1.06351578e+00 6.58132136e-01
-1.15614235e+00 -1.47075176e-01 -5.12882292e-01 7.05261528e-01
-2.04404876e-01 1.05731621e-01 -5.87523401e-01 -4.70114499e-01
-4.45805550e-01 -1.25215673e+00 7.92943358e-01 7.11612478e-02
3.05336207e-01 -8.54909480e-01 -1.78542361e-01 1.00945815e-01
4.27619576e-01 3.17077041e-01 1.19719601e+00 -5.49018741e-01
-7.12426603e-01 -1.76249012e-01 -1.11185580e-01 8.81408513e-01
3.62822533e-01 3.30999255e-01 -4.32723492e-01 -3.85490358e-01
-5.94568951e-03 1.55931920e-01 5.26543498e-01 3.67230386e-01
1.53626406e+00 -6.69984818e-01 7.47665167e-02 8.06673944e-01
1.36446702e+00 5.52790821e-01 3.25906754e-01 3.80654931e-01
1.07268918e+00 4.79581207e-01 5.19829988e-01 5.38988054e-01
3.33162755e-01 3.84815574e-01 8.87296855e-01 -4.98968363e-02
2.97344476e-01 1.70736492e-01 2.25561246e-01 1.53033078e+00
-3.41858208e-01 -9.25403014e-02 -5.75465679e-01 5.31578243e-01
-1.93984807e+00 -7.69280493e-01 -3.30698602e-02 2.03831124e+00
4.20645505e-01 3.84007692e-02 -3.01878810e-01 2.88642526e-01
5.53206205e-01 4.20975506e-01 -5.68233132e-01 -3.93061876e-01
7.83927441e-02 2.48654425e-01 8.16311955e-01 1.31030560e-01
-1.35732663e+00 3.77029032e-01 5.18663406e+00 6.09320164e-01
-1.55544519e+00 -3.19248508e-03 3.66726726e-01 -1.37248382e-01
1.53079838e-01 -3.75963598e-01 -2.61985511e-01 8.25786769e-01
6.66695654e-01 8.60998631e-02 8.48973691e-01 8.69732261e-01
7.07690656e-01 3.59643310e-01 -7.46239424e-01 9.07578945e-01
-4.74819630e-01 -1.13648129e+00 1.24389149e-01 7.99197555e-02
9.11698282e-01 1.28498346e-01 3.12065274e-01 1.60956770e-01
3.71706188e-01 -6.80183530e-01 7.18368530e-01 6.74490035e-01
1.49522379e-01 -8.05667818e-01 1.12089801e+00 6.52249949e-03
-1.40043151e+00 -6.92915142e-01 -4.32277232e-01 -6.12853877e-02
3.52339931e-02 9.25960720e-01 -8.07536721e-01 1.23394799e+00
9.99920845e-01 9.04708564e-01 -2.65207887e-01 7.85259902e-01
3.64828557e-02 4.82928276e-01 -3.68080705e-01 -7.38221258e-02
4.29906577e-01 -7.54858255e-01 5.55111051e-01 6.12929344e-01
5.95111310e-01 -4.25789684e-01 3.46458346e-01 5.66124916e-01
-8.61674622e-02 1.03072233e-01 -4.44567978e-01 -1.63542375e-01
3.68820876e-01 1.61176026e+00 -5.80966771e-01 -6.60580099e-02
-1.44293219e-01 4.87064511e-01 -1.26758724e-01 5.20546257e-01
-7.74724305e-01 -5.91269374e-01 9.92165446e-01 -3.06853056e-01
7.20493674e-01 -5.49008906e-01 -4.22354490e-01 -7.74234176e-01
3.51915240e-01 -7.32009470e-01 1.50080487e-01 -7.55945861e-01
-1.46599066e+00 4.49563712e-01 3.00448667e-02 -1.65318155e+00
-2.17044763e-02 -9.98727620e-01 -6.11390352e-01 7.55889952e-01
-1.54789245e+00 -1.10383868e+00 -3.45890909e-01 1.87370524e-01
5.02659798e-01 -1.66698605e-01 2.48644516e-01 4.74100381e-01
-1.09612167e+00 -5.66763505e-02 4.59530741e-01 1.72463462e-01
2.84529895e-01 -1.20525980e+00 2.17230737e-01 8.10146987e-01
-3.34327757e-01 7.08879828e-01 5.47477841e-01 -6.09432638e-01
-1.82263160e+00 -1.60695684e+00 3.71598810e-01 6.96857795e-02
1.00060153e+00 -2.06080809e-01 -6.38591886e-01 5.96647978e-01
1.91349164e-01 3.32178503e-01 1.93139210e-01 -1.48861542e-01
-1.65980887e-02 -8.87939870e-01 -9.20083225e-01 4.91463572e-01
8.02692175e-01 -2.38664910e-01 -2.08222911e-01 6.00329161e-01
8.20428133e-01 -2.91046709e-01 -1.41537094e+00 6.43263578e-01
6.42911971e-01 -4.48934436e-01 1.14900982e+00 -6.97459638e-01
5.34909844e-01 -7.57330120e-01 4.89258021e-03 -1.81091416e+00
-5.06499648e-01 -5.91137469e-01 -5.40007018e-02 9.76930261e-01
2.75509328e-01 -8.73677552e-01 4.49019670e-01 -6.52312301e-03
-4.73018706e-01 -9.05988753e-01 -1.01419103e+00 -8.44273150e-01
-2.11445674e-01 -2.62131810e-01 1.16326892e+00 1.14916050e+00
-4.40609336e-01 9.70786065e-02 -3.30091745e-01 5.88252723e-01
2.60786742e-01 9.19244289e-02 6.84305429e-01 -1.18494451e+00
1.74294010e-01 -4.77461845e-01 -3.22238833e-01 -6.93970263e-01
7.66969053e-03 -8.59006464e-01 1.07254282e-01 -1.24277222e+00
-5.40781975e-01 -4.50815588e-01 -6.30751848e-01 4.42787766e-01
4.20443453e-02 -1.14431612e-01 3.27229917e-01 2.00194865e-01
-1.30758896e-01 8.55025053e-01 1.66076207e+00 -4.64418292e-01
-2.67096877e-01 3.70745622e-02 -1.40463799e-01 5.15224636e-01
8.20933104e-01 -2.08914518e-01 -6.97501242e-01 -6.42558217e-01
3.58401358e-01 -1.78684160e-01 3.07541311e-01 -1.04219401e+00
2.26044998e-01 -2.81383514e-01 4.18140382e-01 -7.89612770e-01
2.70437617e-02 -1.28237164e+00 5.55437744e-01 5.84088564e-01
-7.73286372e-02 5.99735200e-01 7.68645853e-03 3.53772044e-01
-2.41472349e-01 5.40245250e-02 3.49645764e-01 -1.51568064e-02
-5.08949876e-01 7.44667649e-01 -1.00941882e-01 -4.63698030e-01
8.27606022e-01 3.63741852e-02 -3.87614936e-01 2.63144642e-01
-5.49999416e-01 4.43391830e-01 -9.30957049e-02 6.32357597e-01
4.21262860e-01 -1.53810883e+00 -5.52056491e-01 6.16055846e-01
-7.76064172e-02 2.71779060e-01 4.57276821e-01 1.00775576e+00
-8.17770004e-01 3.50974947e-01 -4.54233199e-01 -5.73211789e-01
-6.69739962e-01 7.12898195e-01 2.65162557e-01 -4.59248036e-01
-5.62725246e-01 3.65136892e-01 -2.11675957e-01 -8.08488190e-01
-1.65493131e-01 -1.09116316e+00 -4.12116230e-01 2.63365686e-01
8.78373608e-02 6.32903397e-01 4.99216914e-01 -2.94421643e-01
-1.02570117e-01 5.55111706e-01 1.24754258e-01 2.34489545e-01
1.69947386e+00 -1.09459661e-01 -5.19684851e-01 4.10863847e-01
1.23955560e+00 -3.99555117e-01 -1.32979274e+00 4.26195972e-02
-1.46336660e-01 -2.97642410e-01 1.36184588e-01 -5.90316594e-01
-1.68752468e+00 6.17024243e-01 6.90157473e-01 7.34124660e-01
1.22887492e+00 -7.60688365e-01 9.02023733e-01 6.04442537e-01
8.56662318e-02 -1.11484480e+00 -6.25126883e-02 7.74960041e-01
9.73690212e-01 -8.62073779e-01 -1.10381003e-02 -2.71720588e-01
-4.00445282e-01 1.25552022e+00 5.31568587e-01 -2.92384505e-01
1.08891952e+00 2.94403166e-01 8.69503394e-02 -2.45271176e-01
-6.53004825e-01 1.94239050e-01 1.51721805e-01 3.77619922e-01
-2.30084077e-01 1.34919256e-01 -2.48945415e-01 4.33516741e-01
-4.33346003e-01 -4.55809385e-02 1.44676030e-01 8.98585618e-01
1.41902104e-01 -7.17806101e-01 -3.72060448e-01 6.75002754e-01
-1.69674650e-01 1.30326360e-01 3.42009187e-01 6.92530096e-01
5.26827514e-01 6.46011233e-01 3.32965434e-01 -7.52510488e-01
5.35515726e-01 -1.06578074e-01 4.36680280e-02 -1.65188476e-01
-6.99653327e-01 -3.20478231e-01 -2.19255477e-01 -5.53734422e-01
-3.84994090e-01 -6.15279853e-01 -8.45453382e-01 -2.55140781e-01
-2.52747148e-01 1.04062013e-01 9.30554509e-01 9.23837662e-01
5.32742858e-01 1.13041234e+00 1.26126015e+00 -9.80476558e-01
-7.62364089e-01 -1.05825937e+00 -7.83634424e-01 4.39122826e-01
3.78372997e-01 -7.46591806e-01 -3.88336450e-01 2.24011078e-01] | [7.005475997924805, 2.748657703399658] |
2faeb445-10ab-487e-8bc0-7dee1df149fb | comparative-analysis-of-automatic-skin-lesion | 1904.03075 | null | http://arxiv.org/abs/1904.03075v1 | http://arxiv.org/pdf/1904.03075v1.pdf | Comparative Analysis of Automatic Skin Lesion Segmentation with Two Different Implementations | Lesion segmentation from the surrounding skin is the first task for
developing automatic Computer-Aided Diagnosis of skin cancer. Variant features
of lesion like uneven distribution of color, irregular shape, border and
texture make this task challenging. The contribution of this paper is to
present and compare two different approaches to skin lesion segmentation. The
first approach uses watershed, while the second approach uses mean-shift.
Pre-processing steps were performed in both approaches for removing hair and
dark borders of microscopic images. The Evaluation of the proposed approaches
was performed using Jaccard Index (Intersection over Union or IoU). An
additional contribution of this paper is to present pipelines for performing
pre-processing and segmentation applying existing segmentation and
morphological algorithms which led to promising results. On average, the first
approach showed better performance than the second one with average Jaccard
Index over 200 ISIC-2017 challenge images are 89.16% and 76.94% respectively. | ['Md. Kamrul Hasan', 'Fakrul Islam Tushar', 'Basel Alyafi'] | 2019-04-05 | null | null | null | null | ['skin-lesion-segmentation'] | ['medical'] | [ 8.48494589e-01 5.37416525e-02 1.79520085e-01 1.07469605e-02
-8.29287469e-01 -7.11564541e-01 5.66003978e-01 6.96940958e-01
-6.88593388e-01 5.90055227e-01 -2.73266405e-01 -1.79946244e-01
7.32825398e-02 -5.75817227e-01 -2.29688752e-02 -9.40809727e-01
1.04317024e-01 2.72502005e-01 8.38021040e-01 5.95488176e-02
6.26958430e-01 7.59737670e-01 -1.11998880e+00 3.76571417e-01
8.84159863e-01 5.75597882e-01 7.42217451e-02 1.28589749e+00
-4.06867027e-01 3.01502854e-01 -3.09478790e-01 -2.32521102e-01
2.24836573e-01 -5.46705663e-01 -1.05450642e+00 2.48799667e-01
2.67493427e-01 1.06639266e-01 4.46907133e-01 1.07228363e+00
5.83452284e-01 -2.16710418e-01 1.15087831e+00 -7.47420669e-01
4.40859199e-02 1.13285594e-01 -1.20572877e+00 3.80259842e-01
3.18664432e-01 -7.58927837e-02 3.16347629e-01 -4.47609305e-01
9.77176905e-01 6.97293818e-01 9.36362028e-01 5.19770861e-01
-1.14613271e+00 -1.23940714e-01 -4.76291478e-01 2.30461285e-01
-1.37133074e+00 -8.02353304e-03 2.38992512e-01 -6.24671996e-01
8.12067389e-01 6.91374600e-01 4.84040946e-01 3.44403088e-01
1.27574548e-01 6.64293706e-01 1.87061763e+00 -8.78945172e-01
4.29488063e-01 2.70399392e-01 2.25577414e-01 7.46561587e-01
6.63210690e-01 -5.11904657e-01 1.56306699e-01 -1.97457433e-01
3.65467638e-01 -1.83550671e-01 9.09566283e-02 7.46724382e-02
-8.43932629e-01 4.09768850e-01 4.65770923e-02 8.29894304e-01
-4.66389745e-01 -7.66372234e-02 5.76001048e-01 -1.98548272e-01
1.33776262e-01 6.59823343e-02 2.27771290e-02 5.29821552e-02
-1.42129385e+00 -2.82895751e-02 6.24799132e-01 3.50754619e-01
2.91400105e-01 -5.00873864e-01 -2.63716072e-01 7.09081352e-01
1.45354241e-01 3.36079955e-01 4.65019196e-01 -5.81353128e-01
-1.91669211e-01 7.40327060e-01 -2.16798931e-01 -6.62183583e-01
-5.32370627e-01 4.65538353e-02 -8.13213289e-01 5.98374486e-01
9.32133675e-01 -1.28424674e-01 -1.67847705e+00 8.36693525e-01
5.39629161e-01 3.54302227e-02 -2.10183859e-02 6.00135744e-01
7.50713110e-01 2.12068796e-01 5.22780776e-01 -1.21273473e-01
1.70832515e+00 -6.91812038e-01 -7.34227359e-01 4.80994731e-01
5.51968396e-01 -1.36959529e+00 4.07701224e-01 3.76227707e-01
-1.10267007e+00 1.12886004e-01 -1.03107178e+00 1.84130564e-01
-7.10787535e-01 3.04300249e-01 3.10411155e-01 9.13556874e-01
-1.15607250e+00 4.46698159e-01 -8.33024740e-01 -1.24609089e+00
7.00525403e-01 3.60574991e-01 -5.52618146e-01 -3.99040729e-02
-3.40136260e-01 8.18441272e-01 2.92969167e-01 -2.21907333e-01
-1.58833236e-01 -6.06553912e-01 -3.89986098e-01 -5.54951668e-01
2.54239887e-02 -2.89088935e-01 9.54486847e-01 -9.33403134e-01
-1.33590543e+00 1.46665108e+00 -3.33189309e-01 -4.88112986e-01
8.27369213e-01 3.45205516e-01 -4.33736145e-02 6.61246061e-01
-1.88561678e-02 4.99871343e-01 3.56840074e-01 -1.39841008e+00
-9.58124936e-01 -3.93341243e-01 -3.91265780e-01 7.54153058e-02
9.92878601e-02 2.27325797e-01 -5.35046637e-01 -4.09886390e-01
1.97567627e-01 -7.65232742e-01 -4.02751237e-01 1.29561692e-01
-6.07733905e-01 -1.29337996e-01 8.30596745e-01 -1.11903620e+00
1.18783557e+00 -1.77381587e+00 -4.97236431e-01 6.07285798e-01
9.04748663e-02 5.84243417e-01 1.34505942e-01 4.75076050e-01
-6.18067347e-02 4.80959296e-01 -6.12340927e-01 1.09721541e-01
-2.11801380e-01 -4.63554934e-02 5.07358670e-01 7.18808651e-01
2.17612371e-01 5.37694216e-01 -5.63958168e-01 -1.08687997e+00
2.89688826e-01 5.79407394e-01 1.06829554e-01 -2.97368824e-01
1.93872005e-01 1.51594505e-01 -1.52825654e-01 1.12125623e+00
1.13585532e+00 2.23645881e-01 1.20972387e-01 -4.59356517e-01
-3.38982046e-01 -5.73688626e-01 -1.29845238e+00 1.27466547e+00
9.00604054e-02 5.94133198e-01 3.35296124e-01 -7.69271433e-01
5.56221962e-01 4.78753120e-01 7.56726086e-01 -5.44692814e-01
3.96570265e-01 3.51579010e-01 6.26001582e-02 -8.00328612e-01
1.48488760e-01 -2.89756775e-01 5.39898753e-01 3.04686755e-01
-1.95089102e-01 -1.50749624e-01 7.29712665e-01 2.27173835e-01
1.10799491e+00 -1.30131829e-03 7.57260859e-01 -5.09633660e-01
1.07404327e+00 4.35515106e-01 2.27077127e-01 4.38394248e-01
-7.87657678e-01 7.17712998e-01 6.89325273e-01 -1.90385938e-01
-9.10333097e-01 -9.28494036e-01 -3.58612806e-01 4.54952747e-01
1.58097371e-02 7.13926703e-02 -1.37020302e+00 -7.51438439e-01
-1.40722126e-01 1.94255888e-01 -8.67218256e-01 5.58547080e-01
-5.56232691e-01 -9.45666313e-01 6.66106761e-01 1.33237615e-01
5.40324092e-01 -6.71983242e-01 -7.51003504e-01 -5.36723882e-02
1.34954065e-01 -7.68539071e-01 6.03821017e-02 -1.31812930e-01
-8.64702523e-01 -1.43191409e+00 -1.27148509e+00 -8.67033780e-01
1.13850629e+00 1.88835226e-02 7.20728040e-01 4.06951398e-01
-1.33428240e+00 1.88019052e-01 -4.32151586e-01 -5.17964363e-01
-3.02507192e-01 -1.51316524e-02 -6.18065774e-01 -1.46550387e-01
5.52555323e-01 -1.65793419e-01 -8.59791219e-01 -3.24991494e-02
-1.25687087e+00 -1.93159357e-01 9.28849995e-01 5.08896768e-01
1.00148070e+00 7.43864626e-02 2.81833038e-02 -1.33619857e+00
4.75541145e-01 -2.58474469e-01 -2.99403876e-01 4.83843744e-01
-3.52075934e-01 -3.39877635e-01 2.33867168e-01 -1.18450923e-02
-9.47366476e-01 4.30847406e-01 -9.62700397e-02 4.75266129e-01
-6.31606817e-01 9.56428200e-02 4.29512471e-01 -4.93057370e-01
6.33066416e-01 2.21264400e-02 2.91435897e-01 -3.86788517e-01
-1.15020238e-01 6.45134449e-01 3.83252531e-01 -3.64818312e-02
6.53238893e-01 9.33068991e-01 4.34934080e-01 -1.14807916e+00
-1.96436659e-01 -1.04759359e+00 -9.11089897e-01 -2.45908663e-01
1.23492944e+00 -1.01400807e-01 -3.04902047e-01 8.63897204e-01
-8.84444594e-01 -2.80663282e-01 5.93608245e-02 2.12862462e-01
-1.20388791e-01 6.86539590e-01 -5.87253869e-01 -1.00481188e+00
-7.26048410e-01 -9.23414171e-01 7.30996370e-01 8.86097610e-01
-2.41829082e-01 -1.24257064e+00 2.60848552e-01 3.77893329e-01
5.60912728e-01 1.00547588e+00 8.28654528e-01 -5.43434024e-01
1.03735179e-01 -5.73569357e-01 -4.22261804e-01 -8.72401446e-02
1.91129565e-01 7.84453750e-01 -1.05974007e+00 -5.20420484e-02
-5.58536351e-01 1.04543962e-01 1.02772343e+00 7.53747880e-01
7.67790258e-01 2.28885099e-01 -6.69318974e-01 3.49744141e-01
2.21287322e+00 2.75161624e-01 9.29949939e-01 3.73306841e-01
9.28700864e-02 8.29013646e-01 6.41444087e-01 9.67355072e-02
-5.31433113e-02 9.25478116e-02 2.47061118e-01 -6.03921950e-01
-5.95766246e-01 3.33519638e-01 -6.50940686e-02 2.48731047e-01
-3.97340685e-01 1.39356196e-01 -1.16315269e+00 8.55169833e-01
-1.21674836e+00 -8.74529958e-01 -7.39623010e-01 2.03379202e+00
6.51576936e-01 -1.53363711e-04 4.86394078e-01 3.34972918e-01
9.49875951e-01 -5.34147739e-01 -2.78799683e-01 -7.52357423e-01
-2.01544598e-01 7.69309044e-01 8.86391461e-01 7.80146778e-01
-1.31020856e+00 7.68323898e-01 6.38771439e+00 9.62725639e-01
-1.25106454e+00 -2.40249243e-02 6.65009141e-01 3.21312785e-01
2.30782136e-01 -1.35366485e-01 -5.75269997e-01 2.62068301e-01
4.74942774e-01 4.66647036e-02 -2.34623134e-01 1.01942919e-01
1.17350295e-01 -1.13750899e+00 -3.62921298e-01 5.31973004e-01
1.60219073e-01 -1.12613761e+00 -2.81853586e-01 1.20118663e-01
8.65032852e-01 -1.41077384e-01 -1.05819784e-01 -4.20370638e-01
1.56423911e-01 -1.19719672e+00 1.17503144e-01 7.98112035e-01
9.26956177e-01 -5.12127697e-01 1.12436032e+00 -9.70516279e-02
-1.26258492e+00 4.15467173e-01 -1.09761924e-01 4.46157336e-01
2.53271777e-02 6.79227889e-01 -1.17874920e+00 5.48875451e-01
5.71509600e-01 1.94050431e-01 -1.06832981e+00 1.80535042e+00
2.08715081e-01 6.65667653e-01 -5.53290725e-01 -1.40614197e-01
3.11290085e-01 -2.94006377e-01 4.88059968e-01 1.87022209e+00
1.20390914e-01 -1.16194218e-01 -1.81315452e-01 3.20481002e-01
6.39070511e-01 6.39514089e-01 -2.61442035e-01 -4.24398482e-03
2.03357682e-01 1.77167869e+00 -1.61198795e+00 -3.47923577e-01
-1.79692775e-01 1.00238180e+00 -2.58932382e-01 1.98462695e-01
-5.42002499e-01 -7.92465806e-01 1.04145445e-01 3.33572149e-01
2.34233271e-02 9.45566148e-02 -5.83802581e-01 -3.01584572e-01
-7.88612813e-02 -5.56000292e-01 7.76462615e-01 -1.84826598e-01
-9.86094058e-01 4.73660141e-01 -1.87940449e-01 -7.82699525e-01
3.44759136e-01 -8.99086475e-01 -1.08264971e+00 9.45163548e-01
-1.59334123e+00 -1.41861963e+00 -4.94736284e-01 2.60490090e-01
4.28176641e-01 1.03111483e-01 8.62045228e-01 -6.41771555e-02
-6.48731887e-01 5.60953736e-01 2.04998985e-01 -7.10224584e-02
8.50856304e-01 -1.78451490e+00 -3.33761305e-01 8.96382451e-01
-5.54009616e-01 2.67723322e-01 6.80922568e-01 -6.61578476e-01
-9.02434170e-01 -7.43114591e-01 6.90486789e-01 -2.66081914e-02
5.27104735e-01 2.23606855e-01 -6.60439312e-01 -8.46274123e-02
7.27026343e-01 -1.17928684e-01 1.14540994e+00 -5.69901228e-01
6.90938383e-02 1.08620100e-01 -1.74919236e+00 5.17757177e-01
4.02313191e-03 2.68424936e-02 -1.82692911e-02 3.85978341e-01
-5.07913709e-01 -1.71532720e-01 -9.49072421e-01 2.23596111e-01
6.65421128e-01 -1.25257707e+00 6.22694135e-01 -1.40790746e-01
1.98112965e-01 -5.25419235e-01 1.82166889e-01 -6.72838926e-01
-1.25425324e-01 -5.18346369e-01 5.80505729e-01 1.22755802e+00
5.91492176e-01 -4.50236380e-01 1.13569236e+00 3.79955053e-01
2.79507101e-01 -9.93707418e-01 -7.19026148e-01 -2.06652239e-01
3.31901789e-01 1.71951398e-01 -1.01860717e-01 7.26977646e-01
-1.02580138e-01 -3.66293699e-01 4.92668241e-01 -1.24971174e-01
9.51384246e-01 -3.36342156e-01 4.49856907e-01 -1.08553505e+00
2.36240745e-01 -7.56448090e-01 -7.85511553e-01 2.32603833e-01
-5.14145076e-01 -9.15538073e-01 -2.47497723e-01 -2.05761170e+00
4.97334659e-01 -3.15015793e-01 -1.51547015e-01 3.69477302e-01
-2.90580451e-01 8.03915560e-01 -6.46207063e-03 2.20408458e-02
-2.63244897e-01 -7.11087584e-01 1.20479608e+00 9.58245480e-04
-3.71273141e-04 -7.32931346e-02 -4.81705487e-01 8.09133232e-01
1.23309374e+00 -2.29278997e-01 6.34582117e-02 3.60242799e-02
-2.52959102e-01 -5.35413206e-01 3.01200211e-01 -1.18597209e+00
4.23624516e-01 -2.88137645e-01 5.05268931e-01 -7.06524014e-01
-1.45849913e-01 -7.01941967e-01 -2.64475791e-04 9.88147736e-01
-5.14238775e-02 -1.71133325e-01 1.83302417e-01 4.95595396e-01
-5.98028749e-02 -7.16835320e-01 1.42337680e+00 -3.26675385e-01
-7.52295554e-01 -2.43305564e-01 -7.97856331e-01 -3.05003434e-01
1.84609866e+00 -8.83532941e-01 -2.61402071e-01 2.22820133e-01
-8.81812871e-01 -8.01107883e-02 6.64627552e-01 -4.93460625e-01
1.87779546e-01 -9.01343405e-01 -9.17825222e-01 -2.13789329e-01
-4.90429997e-02 -2.50259727e-01 4.43363339e-01 1.52371466e+00
-1.54836750e+00 3.43669266e-01 -5.12261808e-01 -4.97330785e-01
-2.16539693e+00 4.82376199e-03 4.76633549e-01 -5.72213233e-01
-3.16722095e-01 9.64113057e-01 -3.10080230e-01 -1.76745821e-02
2.52510235e-02 -1.23153210e-01 -5.63063323e-01 1.05785377e-01
5.38418055e-01 8.38889301e-01 2.17466578e-01 -8.11546862e-01
-5.12678504e-01 9.49265122e-01 -3.26666713e-01 -1.35488048e-01
9.47386980e-01 2.19460353e-02 -5.58278859e-01 1.33746296e-01
1.07998729e+00 1.21051468e-01 -6.57627523e-01 2.16370776e-01
3.65034580e-01 -3.15815657e-01 5.47543541e-02 -1.22060525e+00
-8.37217689e-01 8.83930624e-01 1.08003914e+00 2.96463758e-01
1.31684959e+00 -3.75963598e-01 6.06231630e-01 -1.61375284e-01
-1.68016523e-01 -1.62629032e+00 -4.04707700e-01 1.61525849e-02
5.14450550e-01 -1.11988902e+00 4.64614034e-01 -9.96304095e-01
-6.04663610e-01 1.39948034e+00 3.82135719e-01 -3.09734046e-01
6.38648748e-01 4.86983925e-01 3.59021217e-01 -1.78005978e-01
-3.04538339e-01 -6.64629281e-01 3.39507490e-01 9.39295650e-01
8.79993856e-01 1.22428469e-01 -1.23786080e+00 8.67600739e-02
2.62298614e-01 8.86945240e-03 5.00811279e-01 1.28505099e+00
-5.95764220e-01 -1.15704024e+00 -6.35878086e-01 6.38986111e-01
-1.07745934e+00 2.60098070e-01 -8.83249819e-01 1.00935292e+00
3.72919083e-01 8.51957142e-01 -1.52304247e-01 1.13397673e-01
6.35836646e-02 -1.22481575e-02 6.37112081e-01 -2.81890094e-01
-8.00893247e-01 4.01487201e-01 -5.68654994e-03 -2.79883355e-01
-6.32593930e-01 -7.96939313e-01 -1.47037208e+00 -7.94405118e-03
-2.93173641e-01 -3.29491263e-03 1.16936433e+00 7.40971684e-01
-1.37759641e-01 3.10394615e-01 1.76965788e-01 -3.46572012e-01
-1.01909421e-01 -8.88293982e-01 -8.52793932e-01 4.59313482e-01
-3.14238034e-02 -2.68242568e-01 -8.17124546e-02 7.02362716e-01] | [15.503442764282227, -3.024406671524048] |
9ebb9100-b8cf-4e41-a75e-30f0c50c2109 | juman-a-morphological-analysis-toolkit-for | null | null | https://aclanthology.org/D18-2010 | https://aclanthology.org/D18-2010.pdf | Juman++: A Morphological Analysis Toolkit for Scriptio Continua | We present a three-part toolkit for developing morphological analyzers for languages without natural word boundaries. The first part is a C++11/14 lattice-based morphological analysis library that uses a combination of linear and recurrent neural net language models for analysis. The other parts are a tool for exposing problems in the trained model and a partial annotation tool. Our morphological analyzer of Japanese achieves new SOTA on Jumandic-based corpora while being 250 times faster than the previous one. We also perform a small experiment and quantitive analysis and experience of using development tools. All components of the toolkit is open source and available under a permissive Apache 2 License. | ['Sadao Kurohashi', 'Daisuke Kawahara', 'Arseny Tolmachev'] | 2018-11-01 | null | null | null | emnlp-2018-11 | ['art-analysis'] | ['computer-vision'] | [-8.01275447e-02 -6.72679022e-02 -1.19213670e-01 -1.90759808e-01
-9.41216111e-01 -7.61050880e-01 7.42267221e-02 2.53588855e-01
-8.46128702e-01 6.25734150e-01 3.72984350e-01 -1.19345641e+00
3.50813717e-01 -7.77190208e-01 -1.26560137e-01 -3.73085976e-01
-9.12675932e-02 6.37969255e-01 3.25807154e-01 -4.00578052e-01
1.21173859e-01 5.65040350e-01 -7.04354525e-01 3.86784226e-01
3.55592757e-01 -7.47478288e-03 6.24305196e-02 9.86409485e-01
-5.01253486e-01 7.76242852e-01 -6.52499855e-01 -6.34227753e-01
5.03247499e-01 -1.92515478e-01 -1.09577239e+00 -3.94630611e-01
3.94401550e-01 5.70232309e-02 -5.98845668e-02 1.28687358e+00
4.95190263e-01 -7.61574134e-02 1.83296531e-01 -5.20123839e-01
-6.98601723e-01 1.52976608e+00 -2.45324522e-01 6.72580004e-01
3.05285335e-01 1.76432297e-01 1.19893241e+00 -7.36619413e-01
6.86596215e-01 1.30079544e+00 8.87918234e-01 4.91123110e-01
-1.17085755e+00 -4.82839584e-01 -1.93993643e-01 -3.77332866e-02
-1.28539097e+00 -7.01816618e-01 6.47599041e-01 -1.93738282e-01
2.04730415e+00 3.07691306e-01 3.72858286e-01 7.83939362e-01
1.66894779e-01 5.91302216e-01 1.12096000e+00 -1.14661849e+00
4.54879142e-02 5.75429425e-02 7.80592263e-01 1.06032276e+00
2.08611742e-01 -2.18982443e-01 -2.48068735e-01 -3.76255602e-01
8.39678526e-01 -6.30601108e-01 4.12936985e-01 3.62820297e-01
-9.69892442e-01 9.15103436e-01 -2.86607683e-01 6.13100529e-01
-2.28115648e-01 -1.75516397e-01 8.21961462e-01 5.25907934e-01
3.24639797e-01 5.13791203e-01 -7.84229696e-01 -2.11785153e-01
-8.43723297e-01 3.47884111e-02 1.21350765e+00 8.98336530e-01
4.24159348e-01 6.71750367e-01 3.45191866e-01 9.72774506e-01
3.64514798e-01 3.53040308e-01 8.88883531e-01 -7.12446928e-01
4.82962877e-01 6.41674340e-01 -2.81612754e-01 -4.78491455e-01
-5.90991259e-01 -8.07476193e-02 -2.94489950e-01 2.17836529e-01
6.84849799e-01 -3.08180422e-01 -7.86286473e-01 1.46781945e+00
7.11126328e-02 -5.43214083e-01 8.16629604e-02 1.81973189e-01
7.37554669e-01 5.19116521e-01 2.61562139e-01 -4.00131017e-01
1.56278658e+00 -1.07352555e+00 -9.31596756e-01 -2.06675515e-01
1.10941052e+00 -1.21176159e+00 1.42331386e+00 5.72817564e-01
-1.57054031e+00 -1.59105688e-01 -1.20010257e+00 -2.17571825e-01
-8.18481982e-01 2.55437464e-01 7.57013083e-01 1.19247961e+00
-1.41737211e+00 4.46002960e-01 -1.04814756e+00 -6.48590863e-01
-3.24013203e-01 4.60166186e-01 -1.18615121e-01 6.83229804e-01
-1.03895307e+00 1.03768182e+00 9.54096019e-01 -1.09115526e-01
-1.30750120e-01 -1.64878163e-02 -1.15461016e+00 -2.93426752e-01
2.33415857e-01 4.78106365e-02 1.52821255e+00 -5.41520953e-01
-1.72020769e+00 1.30890131e+00 -8.43380950e-03 -6.66147351e-01
-8.63317922e-02 -2.45646253e-01 -7.79278874e-01 -2.16335639e-01
1.18528605e-02 2.17943847e-01 2.40030020e-01 -4.40177232e-01
-4.74758089e-01 -3.53670061e-01 -3.56641829e-01 -4.76460978e-02
-2.88237244e-01 1.13212073e+00 -2.63035983e-01 -8.90866101e-01
-2.11881727e-01 -6.98766410e-01 -5.29362500e-01 -9.53894258e-01
-5.96802354e-01 -2.95409143e-01 2.03696668e-01 -1.38473463e+00
1.64367056e+00 -1.68502390e+00 -2.98243999e-01 1.41102180e-01
-2.48371005e-01 5.39245248e-01 -2.25668117e-01 7.91701734e-01
-3.40336233e-01 4.59351331e-01 -2.93861896e-01 -3.58461499e-01
7.68479183e-02 1.70109913e-01 -3.89461190e-01 3.56609017e-01
2.29758158e-01 8.79511893e-01 -5.96019208e-01 -5.09821415e-01
1.35997221e-01 -4.38896492e-02 -6.68710396e-02 -3.55131254e-02
-3.78678977e-01 -2.85288483e-01 2.00170472e-01 1.01629961e+00
3.38885814e-01 4.12258089e-01 6.77622557e-01 5.15054226e-01
-5.74707747e-01 1.05543590e+00 -1.30486429e+00 1.82706487e+00
-4.26338941e-01 3.49020123e-01 2.19225734e-01 -4.60719198e-01
8.76414478e-01 4.17807788e-01 -3.55135739e-01 -3.38903487e-01
2.71258235e-01 3.34423006e-01 2.38258600e-01 -3.93398553e-01
8.86085451e-01 -3.45963240e-02 -2.44909897e-01 9.12213206e-01
3.08810532e-01 -4.77698771e-03 7.09714890e-01 1.51910633e-01
1.33972681e+00 3.59182775e-01 1.14546967e+00 -3.23893130e-01
6.99111521e-01 2.41754398e-01 7.81193852e-01 5.89029074e-01
-1.18285194e-01 8.09708238e-02 3.36032599e-01 -8.25395823e-01
-1.32094574e+00 -1.05698776e+00 -2.10436583e-02 1.35402668e+00
-8.54729056e-01 -7.83904254e-01 -8.52248788e-01 -6.48263931e-01
-5.76331317e-01 1.00144124e+00 -2.58633077e-01 5.25579333e-01
-1.28385818e+00 -8.00405085e-01 1.26865494e+00 7.06587911e-01
-2.54076049e-02 -1.98350191e+00 -3.51475984e-01 3.58330071e-01
1.54073104e-01 -1.09433520e+00 -3.47894818e-01 6.26912177e-01
-7.36827791e-01 -6.63120151e-01 1.75232500e-01 -1.27447915e+00
2.76252538e-01 -1.66323677e-01 1.21506286e+00 9.07082334e-02
-2.17398614e-01 -7.87764564e-02 -3.36297810e-01 -6.13966525e-01
-8.76149774e-01 3.88596177e-01 2.39797994e-01 -7.85156190e-01
1.09473109e+00 -6.71135604e-01 4.33002859e-01 -3.92352790e-01
-8.61805320e-01 -2.55843431e-01 5.91994643e-01 4.14277464e-01
2.56767243e-01 -1.80570260e-01 1.00145154e-01 -1.02396452e+00
9.12515521e-01 -9.30956677e-02 -9.42337215e-01 1.38001695e-01
-5.23220003e-01 1.34966597e-01 9.09899056e-01 -4.12925750e-01
-8.91758978e-01 3.64052951e-01 -6.94828749e-01 2.72798747e-01
-4.13409948e-01 7.37058699e-01 -3.30230355e-01 3.42146099e-01
7.67161131e-01 -2.17227228e-02 -1.67902827e-01 -9.36480403e-01
3.74933273e-01 7.07067728e-01 8.86156678e-01 -4.53800440e-01
9.41236913e-01 -4.36738096e-02 -6.90887988e-01 -1.05846560e+00
-2.29958639e-01 -6.78814828e-01 -9.39065278e-01 2.86222965e-01
6.70072198e-01 -5.54875314e-01 -3.50263596e-01 5.75255573e-01
-1.28481412e+00 -6.51675463e-01 -4.21764404e-01 3.28344971e-01
-9.52090323e-02 6.33185804e-01 -1.38105905e+00 -7.01953530e-01
-8.01570892e-01 -8.29498172e-01 6.31454468e-01 2.62285536e-03
-5.94304621e-01 -1.20503151e+00 7.55724549e-01 -5.69511130e-02
1.90215170e-01 -2.26558521e-01 1.36008811e+00 -1.28513765e+00
8.26096535e-02 -4.64651212e-02 1.40600249e-01 3.02819759e-01
-1.44968912e-01 2.81002671e-01 -7.82827497e-01 -1.39604464e-01
1.20582864e-01 -3.21783066e-01 6.50828123e-01 1.83787405e-01
3.52903962e-01 -6.17482364e-01 1.33287534e-01 5.68355143e-01
1.27764177e+00 2.81700641e-01 6.86098039e-01 7.98372209e-01
4.78924334e-01 6.26461983e-01 1.08093835e-01 -1.11852236e-01
1.97156221e-01 4.26684380e-01 -1.05851591e-01 -2.70514004e-02
-2.81082019e-02 -1.38422310e-01 9.86627340e-01 1.76725256e+00
2.26300061e-01 1.65581807e-01 -1.31943488e+00 6.60814345e-01
-1.70254362e+00 -7.03456700e-01 -3.78809094e-01 1.99718964e+00
1.32264793e+00 3.50293696e-01 6.17768407e-01 2.24142432e-01
6.55598223e-01 5.26185110e-02 1.73996463e-02 -1.35434079e+00
-2.69937098e-01 8.08988154e-01 4.90939826e-01 9.99946415e-01
-1.13774860e+00 1.83350575e+00 7.65042686e+00 9.41649020e-01
-8.14459801e-01 3.25360715e-01 4.91424315e-02 5.31524681e-02
-1.53831273e-01 1.43902004e-01 -1.04915011e+00 -9.95204747e-02
1.52209973e+00 -8.57418701e-02 5.70493340e-01 9.11876678e-01
6.20026402e-02 2.00447977e-01 -6.47959888e-01 3.79675567e-01
-3.38146929e-03 -1.33647513e+00 6.47923201e-02 1.68867216e-01
-5.61609119e-03 4.90934372e-01 -3.89731377e-01 4.91134524e-01
9.30706501e-01 -9.95594800e-01 7.45963991e-01 -9.33057591e-02
9.11895037e-01 -7.57124722e-01 7.40527272e-01 1.86630234e-01
-1.05797660e+00 2.33881816e-01 -6.68676674e-01 -5.88017523e-01
1.93719134e-01 3.58076543e-01 -1.15210092e+00 3.01121354e-01
3.71712834e-01 1.81115150e-01 -1.00108540e+00 8.31074774e-01
-5.58663905e-01 1.25131047e+00 -2.98998535e-01 -2.28407443e-01
1.97075307e-01 -4.85606819e-01 8.06236923e-01 2.12596869e+00
-2.14699149e-01 -1.23066865e-01 2.23284855e-01 6.47678912e-01
1.39154851e-01 7.84526646e-01 -7.19433665e-01 -1.36961296e-01
5.07168651e-01 1.59351552e+00 -1.23193288e+00 -3.38102490e-01
-3.51208925e-01 4.71735686e-01 7.99708188e-01 2.32667346e-02
-4.03583318e-01 -8.85406256e-01 2.77846992e-01 -3.58193249e-01
2.04258516e-01 -7.10065544e-01 -6.58351243e-01 -1.05235159e+00
-1.51944002e-02 -1.28693795e+00 7.40433514e-01 -4.23768908e-01
-1.09996545e+00 8.50973606e-01 -2.17596516e-01 -5.10325551e-01
-4.49555695e-01 -1.14084566e+00 -8.28302741e-01 9.26409841e-01
-7.05257893e-01 -1.36133909e+00 5.18805802e-01 3.01173389e-01
6.96917176e-01 -6.63106620e-01 1.61174989e+00 -8.53175968e-02
-9.41994309e-01 5.71148992e-01 -2.56382674e-01 7.16766238e-01
4.86531764e-01 -1.76072633e+00 1.24930024e+00 1.50253880e+00
5.91740727e-01 1.07539880e+00 4.55889791e-01 -9.96842444e-01
-1.05453241e+00 -9.60519552e-01 1.46632409e+00 -6.20625734e-01
1.36677492e+00 -7.03945398e-01 -9.91106749e-01 1.05637467e+00
8.17726076e-01 -6.20815635e-01 9.14326370e-01 4.36684847e-01
-5.58959067e-01 3.13089401e-01 -7.88686216e-01 9.19204116e-01
8.49033415e-01 -7.78839111e-01 -1.20591533e+00 3.40417176e-01
8.75800133e-01 -6.88656196e-02 -6.90004706e-01 2.79346090e-02
4.57783461e-01 -6.16442323e-01 5.26601136e-01 -6.95225239e-01
-1.24120280e-01 -2.54561067e-01 -3.39343771e-02 -1.01091635e+00
-5.02948642e-01 -1.32241213e+00 8.91327113e-02 1.38701916e+00
9.93092120e-01 -5.63931048e-01 6.02638364e-01 1.23795355e-03
-2.50725240e-01 -2.70393193e-01 -7.01034904e-01 -8.49336445e-01
5.18656433e-01 -1.06741381e+00 3.04728419e-01 1.03969312e+00
5.66327333e-01 7.56288171e-01 7.57863652e-03 9.26928297e-02
1.56200886e-01 -3.80983979e-01 5.03974736e-01 -7.89451659e-01
-6.32745206e-01 -6.13329768e-01 -3.65154535e-01 -3.15605611e-01
3.73018861e-01 -1.10802472e+00 -1.29644394e-01 -1.00776207e+00
3.62965092e-02 -6.26258925e-02 -3.33355725e-01 1.01314092e+00
1.69382006e-01 4.90051031e-01 -1.13551788e-01 1.99377254e-01
-2.34689981e-01 -2.75630683e-01 -3.24447453e-02 -4.87972759e-02
-6.03225589e-01 -1.39214441e-01 -5.78187644e-01 1.15244448e+00
1.12182319e+00 -8.69805217e-01 2.37599179e-01 -5.37461042e-01
6.57348216e-01 -3.91963124e-01 -2.89513201e-01 -7.95097172e-01
2.87140638e-01 -6.04125299e-02 8.19064751e-02 -7.68266678e-01
-1.29840627e-01 -5.85971288e-02 -2.85214305e-01 4.61362094e-01
-3.17601383e-01 9.02058303e-01 6.34915590e-01 -2.13325158e-01
1.94324404e-01 -7.31569469e-01 6.18684471e-01 -5.96793354e-01
-5.70911050e-01 -2.00212955e-01 -1.04321134e+00 1.42788887e-01
5.94336152e-01 -8.98105353e-02 -2.73552001e-01 -2.32625734e-02
-6.58558309e-01 -3.13383400e-01 6.31546855e-01 2.80116886e-01
2.83534199e-01 -8.68364513e-01 -7.91358709e-01 4.04529214e-01
-1.81531042e-01 -6.70831859e-01 -6.97772264e-01 5.40368795e-01
-1.02649415e+00 5.63839674e-01 -4.17664260e-01 2.64411390e-01
-1.46467757e+00 6.44935668e-01 7.24933445e-02 -7.62413681e-01
-7.30282903e-01 7.57497668e-01 -6.16224825e-01 -1.06647050e+00
2.84885108e-01 -3.12422991e-01 -4.40928817e-01 -2.42551237e-01
8.19852650e-01 4.39477146e-01 3.94334197e-01 -6.20860040e-01
-4.89936084e-01 7.28665059e-03 -3.61872405e-01 -6.90220654e-01
1.39599240e+00 1.82011649e-01 -7.89991081e-01 9.07484412e-01
6.43239439e-01 8.69743288e-01 -5.72604574e-02 -1.83245465e-01
7.80171037e-01 4.90704000e-01 -2.43823111e-01 -9.02940035e-01
-3.49587351e-01 6.99607253e-01 8.50059986e-02 3.46014231e-01
8.20236206e-01 -4.05541956e-01 9.48118031e-01 8.87355387e-01
9.92372185e-02 -1.45553637e+00 -6.86428010e-01 1.27977550e+00
4.61495876e-01 -5.19454598e-01 -8.20818841e-02 -3.16668749e-01
-4.55572128e-01 1.31543350e+00 2.71882087e-01 -3.20768625e-01
3.59501243e-01 9.74912465e-01 6.77740872e-01 -2.53475219e-01
-6.63867593e-01 -2.92147011e-01 -7.63491169e-02 5.73462129e-01
8.74955475e-01 2.63682932e-01 -6.84909999e-01 6.64138079e-01
-9.01114523e-01 -5.69197237e-01 7.69830108e-01 1.01181746e+00
-2.50383109e-01 -1.59016287e+00 -5.26477337e-01 1.17041454e-01
-1.11510038e+00 -8.15138400e-01 -9.09394622e-01 1.12322450e+00
-1.89211160e-01 8.63095284e-01 -9.09167603e-02 -5.24899483e-01
-3.77963632e-02 7.15453207e-01 3.37777257e-01 -1.10503399e+00
-1.19975770e+00 2.25635096e-01 7.30001390e-01 -4.06562001e-01
4.84149121e-02 -8.42964828e-01 -1.29373419e+00 -1.59761950e-01
-2.36669868e-01 3.22984576e-01 6.85759008e-01 9.25645828e-01
-1.93482488e-01 2.30696462e-02 -9.22737736e-03 -5.62555075e-01
-6.47846222e-01 -1.45547605e+00 -6.65382981e-01 -1.96985260e-01
-3.48767579e-01 1.98869005e-01 -1.94157541e-01 1.44549459e-01] | [10.430002212524414, 10.094178199768066] |
175ecc1e-8cdf-408c-a302-d89dc137aff6 | graph-neural-network-based-log-anomaly | 2307.00527 | null | https://arxiv.org/abs/2307.00527v1 | https://arxiv.org/pdf/2307.00527v1.pdf | Graph Neural Network based Log Anomaly Detection and Explanation | Event logs are widely used to record the status of high-tech systems, making log anomaly detection important for monitoring those systems. Most existing log anomaly detection methods take a log event count matrix or log event sequences as input, exploiting quantitative and/or sequential relationships between log events to detect anomalies. Unfortunately, only considering quantitative or sequential relationships may result in many false positives and/or false negatives. To alleviate this problem, we propose a graph-based method for unsupervised log anomaly detection, dubbed Logs2Graphs, which first converts event logs into attributed, directed, and weighted graphs, and then leverages graph neural networks to perform graph-level anomaly detection. Specifically, we introduce One-Class Digraph Inception Convolutional Networks, abbreviated as OCDiGCN, a novel graph neural network model for detecting graph-level anomalies in a collection of attributed, directed, and weighted graphs. By coupling the graph representation and anomaly detection steps, OCDiGCN can learn a representation that is especially suited for anomaly detection, resulting in a high detection accuracy. Importantly, for each identified anomaly, we additionally provide a small subset of nodes that play a crucial role in OCDiGCN's prediction as explanations, which can offer valuable cues for subsequent root cause diagnosis. Experiments on five benchmark datasets show that Logs2Graphs performs at least on par state-of-the-art log anomaly detection methods on simple datasets while largely outperforming state-of-the-art log anomaly detection methods on complicated datasets. | ['Matthijs van Leeuwen', 'Jiayang Shi', 'Zhong Li'] | 2023-07-02 | null | null | null | null | ['anomaly-detection'] | ['methodology'] | [ 3.01314071e-02 4.29928638e-02 -1.79015547e-01 -9.04608332e-03
2.77335197e-02 -4.56098318e-01 5.72786391e-01 1.11587071e+00
2.91566253e-01 2.15093136e-01 -4.11631260e-03 -7.65186310e-01
-3.39227945e-01 -1.16325498e+00 -6.19132996e-01 -2.22212166e-01
-7.67911553e-01 3.83125901e-01 4.94738519e-01 -1.87112704e-01
1.37201697e-01 7.80477285e-01 -1.18848002e+00 4.18892875e-02
7.38748670e-01 1.17530131e+00 -7.37691879e-01 5.88007152e-01
-3.56270969e-01 1.30085433e+00 -7.11657763e-01 -1.43917978e-01
-9.33875237e-03 -4.39644068e-01 -5.02915502e-01 3.23224843e-01
1.71946958e-01 -3.30555439e-01 -8.93320739e-01 9.18164849e-01
-9.16800126e-02 7.87401944e-02 6.97473586e-01 -1.66353190e+00
-5.65227747e-01 7.84479856e-01 -7.71229148e-01 9.21712637e-01
5.40692508e-01 1.62648231e-01 1.34123671e+00 -6.43018484e-01
1.67105615e-01 9.85442519e-01 5.85712969e-01 -5.70360832e-02
-1.12102365e+00 -6.30526304e-01 6.02449417e-01 3.76838475e-01
-1.26367450e+00 -1.00570396e-02 1.03995514e+00 -2.91735828e-01
1.26072145e+00 2.40831241e-01 7.55564034e-01 8.91907871e-01
4.67255443e-01 8.29308867e-01 2.33189315e-01 -1.39104351e-01
2.34093979e-01 -6.92905366e-01 4.49423432e-01 9.38999832e-01
8.58526409e-01 -1.58239245e-01 -6.93634570e-01 -5.38837254e-01
4.92301673e-01 6.82354450e-01 -8.79368633e-02 -1.58299491e-01
-9.41683352e-01 7.32714355e-01 5.82191825e-01 3.23800564e-01
-6.56692803e-01 3.69054437e-01 7.31720090e-01 5.92283905e-01
6.77344859e-01 3.30247194e-01 -2.12343946e-01 1.27364304e-02
-4.32080954e-01 -2.28296015e-02 8.50485682e-01 6.74381018e-01
6.60692930e-01 5.29430151e-01 -2.46622160e-01 4.50410366e-01
4.50391024e-01 7.05469549e-02 3.40171635e-01 7.67235309e-02
3.44624490e-01 1.35920000e+00 -5.67586005e-01 -1.38842571e+00
-6.47272050e-01 -5.52134931e-01 -1.00215912e+00 -2.03749359e-01
2.13031456e-01 4.41852868e-01 -1.08759665e+00 1.26215231e+00
1.66707128e-01 6.87916517e-01 -3.83058995e-01 2.59441435e-01
6.59435451e-01 4.91312057e-01 -1.93822399e-01 -2.78529406e-01
1.18773532e+00 -4.73675072e-01 -7.79604733e-01 -4.29424077e-01
8.23519111e-01 -1.82561830e-01 1.01705825e+00 2.73949623e-01
-5.41004717e-01 3.00136246e-02 -1.20193768e+00 5.70592701e-01
-6.51392221e-01 -5.83514273e-01 8.94574463e-01 3.50700557e-01
-7.56325662e-01 6.96169496e-01 -1.25648499e+00 -4.02749658e-01
5.72596610e-01 2.14418188e-01 -2.49805704e-01 -1.71265900e-01
-1.00043392e+00 2.80938059e-01 6.31881654e-01 3.73258479e-02
-9.22830403e-01 -5.60404778e-01 -1.26446640e+00 4.17340785e-01
6.85175955e-01 -9.53897014e-02 9.60067153e-01 -1.62936375e-01
-6.94210172e-01 4.32554215e-01 6.09577671e-02 -7.61807740e-01
1.08298197e-01 -3.87564078e-02 -1.02792561e+00 8.00974816e-02
8.70008692e-02 -5.62161684e-01 8.85189593e-01 -7.64181912e-01
-5.87478578e-01 -3.60934794e-01 -2.37545148e-01 -4.20015484e-01
-8.16167295e-01 -1.92471206e-01 -4.82754111e-01 -8.01940441e-01
3.30090433e-01 -4.30842191e-01 -1.46363646e-01 -4.42166030e-01
-7.84114182e-01 -6.51000142e-01 1.22696280e+00 -3.70391488e-01
1.92602170e+00 -2.15773487e+00 -4.38983202e-01 7.44504929e-01
1.06382775e+00 5.45392111e-02 -7.51468241e-02 8.26266348e-01
-3.71678621e-01 7.38399029e-02 -3.14430505e-01 -2.50109524e-01
-1.99668378e-01 3.38024408e-01 -4.75113004e-01 5.78085661e-01
6.63579047e-01 9.98909116e-01 -1.15384865e+00 -1.51687041e-01
8.28743875e-02 -8.22003782e-02 -3.57416570e-01 3.88027370e-01
-2.95320392e-01 9.47878361e-02 -6.27038181e-01 1.09351647e+00
2.55206466e-01 -7.22140372e-01 1.16705053e-01 1.98234946e-01
4.64556277e-01 3.28643709e-01 -9.60139275e-01 9.23516452e-01
3.13953161e-02 6.75420523e-01 -4.85507250e-01 -1.31559265e+00
9.97953176e-01 6.21594861e-02 8.87898564e-01 -6.60107255e-01
3.71847562e-02 1.79221362e-01 9.95662883e-02 -1.95574626e-01
2.16876730e-01 5.02010167e-01 -9.27156061e-02 7.91238129e-01
-4.70417328e-02 3.47292572e-01 5.47796965e-01 8.11302662e-01
2.18336797e+00 -7.55688190e-01 5.85942030e-01 1.71348616e-01
4.03373361e-01 -2.16322660e-01 4.71959621e-01 7.71088183e-01
-1.99400187e-01 4.33135331e-01 1.24881709e+00 -6.07819676e-01
-5.16107559e-01 -1.30081666e+00 3.97658020e-01 1.07131243e+00
-6.04512990e-02 -9.71910179e-01 -1.45138904e-01 -1.19195569e+00
4.91470933e-01 6.19603336e-01 -7.16653883e-01 -7.70462751e-01
-5.54663241e-01 -8.04692805e-01 5.68702579e-01 7.54616559e-01
4.87154024e-03 -1.34317422e+00 1.38400882e-01 4.88960981e-01
2.02886701e-01 -1.13948166e+00 -5.55396736e-01 4.34313267e-01
-9.63452756e-01 -1.69987869e+00 3.19869429e-01 -3.92781466e-01
7.16398478e-01 3.44160259e-01 1.45114195e+00 5.84092736e-01
-3.23247373e-01 5.82273483e-01 -5.38621604e-01 -5.99225998e-01
-4.33689773e-01 5.72366603e-02 3.09158146e-01 3.31695139e-01
6.69918835e-01 -1.13829803e+00 -4.23047304e-01 8.29699859e-02
-1.38739276e+00 -7.51530230e-01 5.93792617e-01 5.46425879e-01
5.51442742e-01 3.40610802e-01 8.88292134e-01 -1.29630578e+00
9.22422469e-01 -9.98388290e-01 -6.30307198e-01 -2.48900093e-02
-1.27147460e+00 -1.51189506e-01 1.08083987e+00 -3.44120622e-01
-1.77048370e-01 -3.17762047e-01 2.14035973e-01 -8.07439983e-01
5.51462173e-02 9.25234675e-01 -7.22945929e-02 5.89703508e-02
8.40288520e-01 3.52301955e-01 3.65453809e-02 -1.90825403e-01
5.65985478e-02 2.92853028e-01 6.19149923e-01 -1.52885512e-01
1.13882899e+00 3.53119314e-01 2.32841089e-01 -7.33640850e-01
-6.70143008e-01 -8.04795146e-01 -4.05674428e-01 -1.68405488e-01
3.53237212e-01 -5.66401541e-01 -7.17859387e-01 5.47872424e-01
-8.16841602e-01 -1.02127299e-01 -4.33073372e-01 9.08196494e-02
-1.17176436e-01 5.19153416e-01 -8.06834519e-01 -6.47820115e-01
-3.26074332e-01 -4.23722327e-01 8.73730659e-01 1.08786644e-02
-2.54282326e-01 -1.32707274e+00 -3.79453227e-02 -2.25367025e-01
3.64850551e-01 6.51924968e-01 1.15824997e+00 -1.53329718e+00
-6.82835162e-01 -1.06778336e+00 -3.73182386e-01 3.85994986e-02
5.87997675e-01 -6.81793690e-02 -5.88368416e-01 -4.81135875e-01
-4.74974155e-01 1.00442708e-01 9.84782517e-01 7.97579437e-02
1.52391434e+00 -3.97450417e-01 -4.97015119e-01 5.18550098e-01
9.19827282e-01 7.92484209e-02 5.03975272e-01 1.82149276e-01
1.27128160e+00 -1.62026118e-02 2.74227202e-01 7.92469978e-01
3.30796450e-01 6.21068142e-02 9.13403928e-01 6.13865675e-03
2.10370362e-01 -5.37317395e-01 5.36024094e-01 8.81615877e-01
1.40214592e-01 -6.53280258e-01 -1.04779172e+00 5.52998126e-01
-2.05732250e+00 -6.65278852e-01 -3.99742573e-01 2.10378742e+00
2.70740271e-01 6.95968330e-01 3.75359446e-01 5.06887615e-01
6.39067709e-01 5.36714792e-01 -7.43513525e-01 -1.85621276e-01
3.84592675e-02 2.00277686e-01 3.99377793e-01 -6.56941682e-02
-1.10253847e+00 6.68184102e-01 5.74511719e+00 6.36285543e-01
-8.38399172e-01 -3.69703054e-01 2.97188014e-01 1.67232394e-01
-3.05441290e-01 -7.80789228e-03 -3.66954386e-01 3.98751736e-01
1.16888535e+00 -4.90069687e-01 3.82264286e-01 9.01705742e-01
1.11800916e-01 3.05897623e-01 -1.41880906e+00 7.88696587e-01
-2.42346134e-02 -1.32290709e+00 2.43903354e-01 3.16782624e-01
3.59169334e-01 3.96506824e-02 -3.11216563e-01 4.43210006e-01
4.19496089e-01 -1.06692576e+00 -4.52328064e-02 2.32233778e-01
4.69485164e-01 -7.57335007e-01 7.63356984e-01 6.43230528e-02
-1.76441407e+00 -2.24814117e-01 -1.76226422e-02 -1.02246381e-01
-2.85922526e-03 1.00829351e+00 -1.09983742e+00 7.12791741e-01
7.40485489e-01 1.34839952e+00 -8.67241204e-01 8.88686121e-01
-1.23632826e-01 1.15830648e+00 -3.60989839e-01 2.19094619e-01
2.23099008e-01 4.44909446e-02 8.24048162e-01 1.14234900e+00
2.32197061e-01 -1.73934460e-01 5.86909175e-01 6.64481401e-01
-4.13656414e-01 -2.98610213e-03 -1.16283739e+00 -8.06348085e-01
4.90893453e-01 1.24708796e+00 -1.10169733e+00 -1.73535958e-01
-6.14411414e-01 8.09695721e-01 4.45522100e-01 3.14212590e-01
-5.75615525e-01 -6.12192631e-01 7.16709137e-01 3.94875705e-01
1.37338862e-01 -2.29052529e-01 -3.15862969e-02 -1.15320337e+00
1.36219591e-01 -7.72668242e-01 1.24205208e+00 -1.63697183e-01
-1.69601309e+00 4.45754290e-01 -1.78629488e-01 -1.54684627e+00
-4.39844519e-01 -6.00945711e-01 -1.35733914e+00 2.29047135e-01
-1.20300496e+00 -1.03073359e+00 -6.30836606e-01 8.42921317e-01
2.93387204e-01 -4.36191261e-01 6.58771873e-01 2.53840804e-01
-8.82092297e-01 6.92997634e-01 -1.75556943e-01 5.75055122e-01
4.79089916e-01 -1.57905841e+00 8.96511972e-01 1.24536252e+00
3.86270761e-01 3.52522731e-01 4.41079646e-01 -9.82624292e-01
-1.63256335e+00 -1.46415138e+00 4.16628510e-01 -4.72299099e-01
1.36369216e+00 -2.67040402e-01 -1.55348527e+00 1.09573436e+00
-3.99726093e-01 6.88427448e-01 5.20934165e-01 3.44389737e-01
-5.23240566e-01 -1.84257984e-01 -7.17620015e-01 5.11082828e-01
1.39485514e+00 -6.17067635e-01 -1.85792670e-01 6.05428576e-01
9.24699485e-01 -2.25300670e-01 -6.14924610e-01 4.99687254e-01
-2.85999943e-03 -7.82618284e-01 8.17226470e-01 -1.01665497e+00
4.08626229e-01 -2.04530522e-01 1.77633852e-01 -1.41398227e+00
-4.46213931e-01 -7.12446332e-01 -1.28841591e+00 1.10487688e+00
2.98082173e-01 -9.93857741e-01 9.22564507e-01 -6.50107414e-02
-3.77478778e-01 -6.93122029e-01 -5.71420193e-01 -1.02562606e+00
-5.33486545e-01 -8.16182077e-01 7.06520200e-01 1.01183975e+00
5.04333945e-03 3.59541535e-01 -1.76078901e-01 5.14961064e-01
4.70121503e-01 -3.50311995e-02 7.16930747e-01 -1.68151295e+00
-1.47195905e-01 -7.29160428e-01 -1.07579947e+00 -4.70983475e-01
2.10647985e-01 -1.12751675e+00 -1.57577470e-01 -1.48912466e+00
-2.07732767e-01 4.11147773e-02 -8.70837986e-01 6.44093215e-01
-4.33696151e-01 1.69594049e-01 -4.25106198e-01 2.49741316e-01
-7.54000902e-01 6.45729303e-01 6.21128857e-01 -2.66639352e-01
-4.75864828e-01 3.35676104e-01 -6.80042624e-01 7.62944400e-01
5.93154669e-01 -5.28433681e-01 -4.70322251e-01 1.89649716e-01
3.33949357e-01 -3.77140306e-02 3.06890339e-01 -9.49902475e-01
3.23944390e-01 -5.78290224e-02 2.97843188e-01 -6.69555247e-01
-2.54575670e-01 -7.48120427e-01 -3.16828370e-01 4.86681104e-01
2.26457082e-02 5.23554146e-01 2.04847515e-01 1.39736128e+00
-4.75570887e-01 3.70644957e-01 2.14231789e-01 3.13895911e-01
-8.62022936e-01 8.93190086e-01 -4.98389781e-01 1.52379245e-01
9.53639507e-01 -9.81330499e-02 -3.64409685e-01 -7.59445012e-01
-5.63449681e-01 5.65182388e-01 5.43759540e-02 5.36210239e-01
8.24718714e-01 -1.58328617e+00 -6.29633725e-01 4.64225709e-01
8.69221091e-01 1.40909001e-01 -1.23571292e-01 8.73734713e-01
-5.18658578e-01 2.26654354e-02 1.51469886e-01 -6.90024734e-01
-9.27146614e-01 6.46374822e-01 2.10099474e-01 -8.09033632e-01
-1.00894201e+00 4.99266982e-01 5.51071241e-02 -1.70294240e-01
1.85163885e-01 -2.63236225e-01 -3.44058007e-01 8.84391516e-02
5.00891745e-01 4.04091120e-01 4.45876271e-01 -1.19275711e-02
-5.43201506e-01 -3.62327769e-02 -4.22329366e-01 6.22779727e-01
1.17634666e+00 2.02849269e-01 -2.97777951e-01 5.81792593e-01
8.39797735e-01 -3.29014920e-02 -8.48597169e-01 -6.55579329e-01
6.12207055e-01 -3.79422456e-01 -2.38090649e-01 -2.09823042e-01
-1.29236925e+00 5.12597740e-01 1.16155855e-01 1.16324008e+00
1.38196898e+00 2.05175623e-01 8.34367037e-01 6.55531287e-01
-1.20074740e-02 -5.74448228e-01 5.93372524e-01 7.13292360e-01
6.73233807e-01 -1.36536419e+00 7.45773837e-02 -5.99813342e-01
-2.93410331e-01 1.21728146e+00 9.21685994e-01 -3.25638145e-01
8.93994927e-01 1.08026147e-01 -3.56627345e-01 -7.93898642e-01
-7.18616605e-01 -1.91869155e-01 5.32142997e-01 5.29561996e-01
3.42524707e-01 8.11148807e-02 3.16439085e-02 5.82674384e-01
8.78634676e-03 -4.58863944e-01 6.71959639e-01 1.03495622e+00
-4.28535372e-01 -7.59864688e-01 -1.42361373e-01 1.26509702e+00
-4.39409584e-01 1.79001931e-02 -6.64789438e-01 8.01060677e-01
-6.33190274e-01 1.00377941e+00 4.55549777e-01 -7.44065464e-01
6.72555566e-01 1.26243398e-01 -1.76921383e-01 -8.82931292e-01
-3.73488218e-01 -2.78936565e-01 1.32220611e-03 -1.02047670e+00
4.08756167e-01 -2.44845957e-01 -1.52367818e+00 -5.33089697e-01
-2.04208836e-01 1.00607425e-01 -2.17926558e-02 1.10973084e+00
5.26387095e-01 1.16306496e+00 8.83897722e-01 -4.10998702e-01
-2.62196004e-01 -1.07908154e+00 -9.37877119e-01 6.73504412e-01
5.28788328e-01 -6.39594138e-01 -6.33854806e-01 -3.50848049e-01] | [6.640012264251709, 5.772704601287842] |
6efb5d7c-c6b9-4ecf-9e66-08532724f0aa | brain2word-decoding-brain-activity-for | 2009.04765 | null | https://arxiv.org/abs/2009.04765v3 | https://arxiv.org/pdf/2009.04765v3.pdf | Brain2Word: Decoding Brain Activity for Language Generation | Brain decoding, understood as the process of mapping brain activities to the stimuli that generated them, has been an active research area in the last years. In the case of language stimuli, recent studies have shown that it is possible to decode fMRI scans into an embedding of the word a subject is reading. However, such word embeddings are designed for natural language processing tasks rather than for brain decoding. Therefore, they limit our ability to recover the precise stimulus. In this work, we propose to directly classify an fMRI scan, mapping it to the corresponding word within a fixed vocabulary. Unlike existing work, we evaluate on scans from previously unseen subjects. We argue that this is a more realistic setup and we present a model that can decode fMRI data from unseen subjects. Our model achieves 5.22% Top-1 and 13.59% Top-5 accuracy in this challenging task, significantly outperforming all the considered competitive baselines. Furthermore, we use the decoded words to guide language generation with the GPT-2 model. This way, we advance the quest for a system that translates brain activities into coherent text. | ['Beni Egressy', 'Roger Wattenhofer', 'Damian Pascual', 'Nicolas Affolter'] | 2020-09-10 | null | null | null | null | ['brain-decoding', 'brain-decoding'] | ['medical', 'miscellaneous'] | [ 6.77807033e-01 3.68749350e-01 1.30695477e-01 -4.53784794e-01
-4.83776659e-01 -4.19846833e-01 1.14558280e+00 2.54482538e-01
-8.73466969e-01 6.31446421e-01 5.15801132e-01 -2.24794626e-01
2.98310250e-01 -6.62261546e-01 -7.84948528e-01 -6.71893060e-01
1.09382153e-01 4.99561727e-01 -4.54405509e-02 -3.78080197e-02
2.93320805e-01 3.82290900e-01 -1.24299550e+00 3.93409401e-01
6.72128260e-01 8.58935177e-01 4.54445988e-01 4.03858602e-01
3.66695002e-02 4.39957708e-01 -6.31863892e-01 -5.19373417e-01
-7.87760839e-02 -8.15822065e-01 -7.30471075e-01 -3.40336591e-01
4.56571519e-01 -1.62173480e-01 -3.44565451e-01 1.01188874e+00
5.10964751e-01 7.42083564e-02 9.75626349e-01 -7.48493671e-01
-7.60852933e-01 8.20103228e-01 -2.52468050e-01 6.44866526e-01
2.20195934e-01 4.86134924e-02 9.67270315e-01 -9.62306440e-01
7.18482137e-01 1.03571069e+00 -4.95168455e-02 8.46071422e-01
-1.42242324e+00 -5.90929925e-01 2.27886170e-01 2.02234402e-01
-1.34024727e+00 -6.17116272e-01 4.49506849e-01 -7.39474177e-01
9.72091794e-01 1.05569303e-01 7.24677742e-01 1.86754775e+00
6.39794230e-01 5.28710961e-01 1.32234788e+00 -3.87197971e-01
3.52185577e-01 -1.56812333e-02 2.34621957e-01 3.83417875e-01
3.95835876e-01 3.46286818e-02 -7.97060430e-01 2.16478422e-01
7.26936758e-01 -1.99427217e-01 -5.00210643e-01 5.77670820e-02
-1.73442674e+00 8.63670349e-01 7.08618820e-01 6.89797580e-01
-8.37969422e-01 2.98421383e-01 1.03407577e-01 -5.17238304e-02
4.79795218e-01 6.57162845e-01 8.87862816e-02 1.79603696e-01
-1.29809892e+00 2.21662775e-01 6.22049451e-01 4.61223036e-01
3.13824475e-01 1.12959795e-01 -6.92086220e-01 5.33156097e-01
3.65372479e-01 4.72530812e-01 9.50033724e-01 -3.27975065e-01
4.40337718e-01 8.86117220e-02 -1.51544780e-01 -9.66327071e-01
-4.61986065e-01 -5.45713723e-01 -7.67383277e-01 -2.20264435e-01
2.45142505e-01 -2.09937487e-02 -9.31310236e-01 1.90700305e+00
-2.59293228e-01 2.69806653e-01 1.33075882e-02 1.09401834e+00
6.28646016e-01 5.84929287e-01 2.91673839e-01 -1.31379083e-01
1.77452397e+00 -8.04806173e-01 -8.12478602e-01 -8.46607983e-01
1.28408656e-01 -2.86832124e-01 8.16905618e-01 5.56730330e-01
-9.99876559e-01 -6.18149877e-01 -1.15898323e+00 -1.63138047e-01
-3.40262800e-01 -2.73265806e-03 5.35734534e-01 5.63047409e-01
-1.37276435e+00 3.35056782e-01 -8.07959735e-01 -4.96170431e-01
6.11534953e-01 1.32123932e-01 -5.33245027e-01 3.66810150e-02
-1.31712592e+00 1.24174178e+00 5.51515937e-01 2.23010853e-01
-1.20509827e+00 -4.83369172e-01 -6.77207589e-01 1.38945505e-01
1.22668467e-01 -8.35231900e-01 9.93324578e-01 -7.20297813e-01
-1.19971275e+00 1.14921200e+00 -4.15615976e-01 -8.75158966e-01
4.10205632e-01 -1.76033363e-01 -4.36310709e-01 2.24607125e-01
4.83507849e-02 1.10838056e+00 1.01951253e+00 -8.84522319e-01
-9.35316905e-02 -5.08598506e-01 -8.50331709e-02 1.40825342e-02
-4.68203962e-01 -2.13740423e-01 -1.76975697e-01 -8.54101717e-01
1.27407938e-01 -7.53622711e-01 -6.77003190e-02 -9.90879908e-02
-2.41161004e-01 -2.05468565e-01 -1.11965537e-01 -8.56535614e-01
9.57466722e-01 -2.13647342e+00 4.87792820e-01 -5.26375845e-02
5.02293646e-01 7.10553750e-02 -3.34645987e-01 3.00618201e-01
-2.55338401e-01 3.24896246e-01 -4.53808576e-01 -2.54733264e-01
1.44495323e-01 4.29210486e-03 -5.74696660e-01 5.30895054e-01
5.46192825e-01 1.34286177e+00 -9.74437773e-01 -1.92754537e-01
-5.97832091e-02 5.80781937e-01 -5.82876325e-01 1.99008361e-01
-1.89506963e-01 6.53858483e-01 -3.14968526e-01 4.07763496e-02
3.20025235e-01 -5.45247719e-02 1.99379310e-01 -2.00652704e-01
7.84049109e-02 3.58574182e-01 -4.15988475e-01 2.03019667e+00
-6.19727075e-01 8.19204390e-01 -3.31491172e-01 -1.33798218e+00
8.80706370e-01 3.85089606e-01 8.23103711e-02 -1.20969641e+00
3.84912699e-01 1.25841275e-01 3.95805895e-01 -5.12402833e-01
1.36628494e-01 -3.54203850e-01 -1.78346053e-01 6.36266410e-01
3.62896234e-01 1.83670558e-02 1.82450101e-01 -1.55338319e-02
1.21526420e+00 -5.24235703e-02 4.25708562e-01 -3.25710505e-01
5.04774034e-01 -4.39254045e-01 -1.22868039e-01 7.25818753e-01
-7.36460509e-03 6.73935294e-01 5.64613581e-01 -1.89492717e-01
-6.85130954e-01 -1.23280752e+00 -2.06164807e-01 9.24080312e-01
-2.37769365e-01 -1.58613712e-01 -9.92457509e-01 -3.06576490e-01
-3.35012645e-01 1.04237187e+00 -9.14076924e-01 -3.97319138e-01
-5.33329129e-01 -7.79416680e-01 6.38819218e-01 3.28777969e-01
3.10362071e-01 -1.32171607e+00 -8.61704826e-01 3.94952476e-01
-4.41599309e-01 -1.43521750e+00 -3.04910451e-01 1.94818571e-01
-8.65866363e-01 -7.22332120e-01 -8.08120072e-01 -7.44907081e-01
7.48049378e-01 -4.60471176e-02 1.21001410e+00 -1.72098204e-01
-3.74803007e-01 8.59304443e-02 -4.48544741e-01 -3.49904627e-01
-3.21292400e-01 1.41345769e-01 3.29958624e-03 4.24810141e-01
4.53638583e-01 -5.29444277e-01 -7.56153643e-01 -2.45634168e-01
-1.15455413e+00 3.80693585e-01 9.28649902e-01 6.91944540e-01
5.39913952e-01 -5.46994567e-01 6.14473462e-01 -7.90742218e-01
9.94116366e-01 -7.32897878e-01 -3.37601244e-01 1.97814494e-01
-4.95043337e-01 4.18451339e-01 6.69532776e-01 -3.86921793e-01
-7.13896394e-01 -8.50110054e-02 -3.26078564e-01 -5.50038065e-04
-3.18388641e-01 5.55597305e-01 5.84430099e-02 2.76934594e-01
8.32669914e-01 5.92815280e-01 -2.36885175e-01 -3.45474452e-01
4.75780547e-01 6.46766603e-01 6.29613996e-01 -3.83941174e-01
4.62552339e-01 3.97852421e-01 -2.33112022e-01 -7.07057297e-01
-7.86062121e-01 -6.12079352e-03 -7.16104388e-01 -2.05056354e-01
1.33377433e+00 -8.65728915e-01 -4.57030565e-01 -3.81352045e-02
-1.43607426e+00 -2.61687249e-01 6.04126640e-02 4.94831532e-01
-5.30557036e-01 1.59087963e-02 -2.63465941e-01 -6.32138073e-01
-4.50033456e-01 -1.19983578e+00 1.17798865e+00 -9.47515219e-02
-5.86765170e-01 -1.01537812e+00 -7.49137327e-02 3.64704341e-01
4.98567104e-01 3.43570232e-01 9.21343386e-01 -8.65669906e-01
-2.93374866e-01 -9.70362052e-02 -1.93840712e-01 2.00712845e-01
-8.80160183e-02 -7.26439893e-01 -1.26334691e+00 -2.17110485e-01
2.48385027e-01 -9.54081342e-02 1.16022706e+00 2.08861709e-01
1.30011177e+00 -1.53001085e-01 -2.51565874e-01 3.70126039e-01
1.21839797e+00 -5.32547310e-02 7.36012340e-01 -6.21598028e-02
3.85348737e-01 7.28204846e-01 9.08393040e-02 1.63666487e-01
2.10440367e-01 6.38993382e-01 3.02299082e-01 6.22914247e-02
-3.04487228e-01 -2.49359444e-01 6.49524689e-01 7.58876264e-01
6.70816451e-02 -5.16877770e-01 -1.09436166e+00 6.68148875e-01
-1.48192048e+00 -8.37697685e-01 -1.96350127e-01 2.04156351e+00
8.52501035e-01 1.57463282e-01 -3.60949069e-01 -1.01552658e-01
4.35335159e-01 2.88566589e-01 -4.98116702e-01 -4.43953514e-01
7.69895092e-02 8.49644005e-01 2.41537228e-01 3.77398282e-01
-8.11109662e-01 8.65986526e-01 5.99920988e+00 4.23275620e-01
-1.31502092e+00 5.77515721e-01 6.20924771e-01 -1.42418221e-01
-2.57782727e-01 -4.75296557e-01 -5.63254893e-01 4.98110592e-01
1.48441148e+00 -3.94164860e-01 7.76983261e-01 1.99941605e-01
3.15471709e-01 -1.81114271e-01 -1.64078808e+00 1.14971769e+00
6.94702268e-01 -1.01605868e+00 2.92393178e-01 2.92040519e-02
4.12354708e-01 1.10541560e-01 1.22781888e-01 2.30268046e-01
-2.61124730e-01 -1.58709502e+00 1.03868890e+00 8.13754082e-01
7.37193584e-01 -1.22121453e-01 5.83905220e-01 5.96893668e-01
-6.76776588e-01 2.07684547e-01 -3.11470658e-01 7.19633922e-02
1.84974223e-01 7.00317383e-01 -9.82720494e-01 3.04692954e-01
2.26829425e-01 5.77128291e-01 -7.63955235e-01 8.26322436e-01
-6.16459966e-01 7.42352307e-01 1.17210336e-01 -1.00440584e-01
1.42507702e-01 2.54063755e-02 2.46154636e-01 1.31259263e+00
6.09342754e-01 2.91835815e-02 -1.28580123e-01 1.36254394e+00
-4.11722362e-01 2.39014730e-01 -6.05219960e-01 -4.84853297e-01
2.83201672e-02 1.28457093e+00 -9.10278738e-01 -2.57779419e-01
-1.28835544e-01 1.33001339e+00 4.13556516e-01 2.88667411e-01
-8.87761414e-01 -5.77121861e-02 4.29100662e-01 6.07475825e-02
2.12210655e-01 -2.98305780e-01 -2.05502778e-01 -1.15327346e+00
1.68430194e-01 -5.99088788e-01 -8.68670568e-02 -8.59090805e-01
-1.23410535e+00 9.90394592e-01 5.18134572e-02 -6.76281452e-01
-4.12339747e-01 -8.66126716e-01 -2.77039230e-01 1.17605650e+00
-1.66042078e+00 -7.17946112e-01 -2.79053360e-01 2.99957454e-01
6.90156341e-01 -3.79405580e-02 1.04448974e+00 3.19012672e-01
-3.22475016e-01 2.53415197e-01 -3.76136094e-01 1.58081338e-01
5.76419413e-01 -9.16251183e-01 6.84518397e-01 9.75032151e-01
6.88331485e-01 9.24223006e-01 6.82681501e-01 -5.22729218e-01
-1.27281690e+00 -9.29792702e-01 1.19660437e+00 -5.51444232e-01
6.95184350e-01 -1.01840723e+00 -1.05588830e+00 4.64474350e-01
4.48314518e-01 -3.85618396e-02 6.09269142e-01 -3.12608033e-01
-2.80042142e-01 1.10069299e-02 -8.07857573e-01 6.41484261e-01
1.28684449e+00 -6.59693420e-01 -8.98077130e-01 5.12928426e-01
5.22277772e-01 -7.09844977e-02 -7.54454255e-01 -1.23403773e-01
4.60527748e-01 -6.93881333e-01 9.81475294e-01 -5.30552506e-01
5.51074803e-01 5.51377842e-03 1.07740201e-01 -1.65346158e+00
-3.99223864e-01 7.52967177e-03 -6.26684129e-02 9.12418664e-01
4.56792682e-01 -6.71654165e-01 2.50095218e-01 2.92390674e-01
2.28669811e-02 -4.63812500e-01 -1.03444803e+00 -5.55503011e-01
2.68132895e-01 -4.32445318e-01 3.96664470e-01 6.54651046e-01
-2.02040195e-01 5.78617990e-01 -2.14543611e-01 -1.11514125e-02
3.61391753e-01 -1.31940767e-01 1.20631814e-01 -1.33079433e+00
-1.48443356e-01 -6.78655207e-01 -3.62444460e-01 -9.71307099e-01
6.09750867e-01 -1.55972683e+00 3.14225614e-01 -1.77260149e+00
2.55487710e-01 7.35802650e-02 -5.59047818e-01 4.70638990e-01
-1.09320715e-01 5.72536886e-01 1.61920100e-01 -1.21807255e-01
-1.96634740e-01 3.25352639e-01 1.16967070e+00 -2.53544986e-01
4.35570776e-01 -5.44403911e-01 -1.02884030e+00 2.25092992e-01
9.05628502e-01 -5.84347188e-01 -4.16043282e-01 -7.89178848e-01
3.22578728e-01 -7.61528835e-02 7.19194293e-01 -1.07589495e+00
9.20424238e-02 1.27928466e-01 5.24390340e-01 2.65707225e-02
2.97520339e-01 -6.18352830e-01 -1.10187931e-02 4.98570204e-01
-6.15522683e-01 6.83359290e-03 1.13546476e-01 4.92353112e-01
-2.10156307e-01 -4.08900589e-01 5.45946240e-01 -7.88066909e-02
-7.14780152e-01 1.73409343e-01 -5.00719845e-01 2.06906751e-01
8.95292461e-01 -3.03742588e-02 -1.83926895e-01 -2.18453720e-01
-8.29332054e-01 -1.27331272e-01 -8.41817930e-02 7.68879533e-01
8.15793872e-01 -1.14616454e+00 -1.04084516e+00 3.59874517e-01
1.42456979e-01 -5.43783009e-01 4.57331240e-02 9.85332370e-01
-3.91112477e-01 8.22199702e-01 -4.42110330e-01 -3.79342347e-01
-7.10605323e-01 3.98814827e-01 2.44724795e-01 -8.36463496e-02
-5.46061456e-01 7.32602179e-01 4.27495986e-01 7.54882917e-02
-7.42654204e-02 -5.61071098e-01 -5.68668067e-01 4.52590048e-01
8.09377849e-01 -3.51941437e-01 4.08808291e-01 -7.70448029e-01
-4.38104689e-01 1.77444190e-01 -2.13599335e-02 -5.14116108e-01
1.49388897e+00 3.24487776e-01 -2.56460160e-01 5.88241875e-01
1.20988488e+00 -1.49690554e-01 -8.03138316e-01 -2.67852135e-02
1.33605555e-01 -1.83946565e-01 2.60828644e-01 -8.67927611e-01
-9.65843260e-01 1.31751037e+00 6.87276721e-01 8.92780051e-02
9.88848269e-01 6.68104663e-02 5.44959486e-01 2.79462337e-01
5.29928625e-01 -7.29102552e-01 7.75403203e-03 4.28696871e-01
1.20761585e+00 -1.09181488e+00 -2.90601462e-01 3.61051075e-02
-5.06821573e-01 1.14563084e+00 3.59772891e-01 -2.52140909e-01
2.77613789e-01 -8.23167562e-02 -2.59850651e-01 -2.97805250e-01
-7.33917058e-01 -2.34457806e-01 5.49508870e-01 4.57947731e-01
7.12238610e-01 2.08206609e-01 -6.62603796e-01 8.58691275e-01
-4.51760978e-01 -9.61421523e-03 4.35483664e-01 4.98104155e-01
-3.88836712e-01 -1.05825758e+00 -3.20779502e-01 6.58123016e-01
-3.73614997e-01 -3.57690483e-01 -3.36516321e-01 4.17872638e-01
5.24734380e-03 8.08295548e-01 1.73380479e-01 -1.35587141e-01
3.17097872e-01 5.99757314e-01 6.32563114e-01 -1.00661743e+00
-5.33693850e-01 -2.19928965e-01 -2.34603643e-01 -4.60107654e-01
-3.42952251e-01 -6.31005168e-01 -1.26195848e+00 2.11179286e-01
1.59272745e-01 -8.35768413e-03 7.30277896e-01 1.09195554e+00
1.96055174e-01 9.45405185e-01 1.22509792e-01 -7.55937934e-01
-3.16633552e-01 -1.10402906e+00 -2.55762964e-01 4.24335450e-01
1.88303337e-01 -6.92437172e-01 -6.09683655e-02 2.47389227e-01] | [10.896322250366211, 2.4996275901794434] |
fdcdf584-5f2e-49c5-abf0-2f04e0c3b1b1 | ukp-square-an-online-platform-for-question | 2203.13693 | null | https://arxiv.org/abs/2203.13693v2 | https://arxiv.org/pdf/2203.13693v2.pdf | UKP-SQUARE: An Online Platform for Question Answering Research | Recent advances in NLP and information retrieval have given rise to a diverse set of question answering tasks that are of different formats (e.g., extractive, abstractive), require different model architectures (e.g., generative, discriminative), and setups (e.g., with or without retrieval). Despite having a large number of powerful, specialized QA pipelines (which we refer to as Skills) that consider a single domain, model or setup, there exists no framework where users can easily explore and compare such pipelines and can extend them according to their needs. To address this issue, we present UKP-SQUARE, an extensible online QA platform for researchers which allows users to query and analyze a large collection of modern Skills via a user-friendly web interface and integrated behavioural tests. In addition, QA researchers can develop, manage, and share their custom Skills using our microservices that support a wide range of models (Transformers, Adapters, ONNX), datastores and retrieval techniques (e.g., sparse and dense). UKP-SQUARE is available on https://square.ukp-lab.de. | ['Iryna Gurevych', 'Gözde Gül Şahin', 'Nils Reimers', 'Jonas Pfeiffer', 'Leonardo F. R. Ribeiro', 'Haritz Puerto', 'Hannah Sterz', 'Clifton Poth', 'Gregor Geigle', 'Max Eichler', 'Rachneet Sachdeva', 'Kexin Wang', 'Tim Baumgärtner'] | 2022-03-25 | null | https://aclanthology.org/2022.acl-demo.2 | https://aclanthology.org/2022.acl-demo.2.pdf | acl-2022-5 | ['explainable-models'] | ['computer-vision'] | [-4.64056879e-01 -4.21093673e-01 2.48620212e-01 -4.68578905e-01
-1.02246451e+00 -1.23696148e+00 4.09775227e-01 3.98869187e-01
-1.92154646e-01 1.61483824e-01 5.07034361e-02 -6.90241873e-01
-5.97587883e-01 -8.96340787e-01 -4.85300541e-01 -2.57547289e-01
5.10122955e-01 1.00344825e+00 4.11407143e-01 -2.97214806e-01
3.22899699e-01 4.03228611e-01 -1.91487646e+00 5.23241460e-02
1.19764745e+00 7.48001337e-01 6.34985805e-01 7.83674717e-01
-5.47877312e-01 3.62449050e-01 -6.58229172e-01 -5.85951746e-01
-1.83170717e-02 6.00056536e-02 -1.12117028e+00 -6.44649029e-01
5.77372551e-01 -2.01947317e-01 -9.75918621e-02 9.47317481e-01
7.48181343e-01 5.32174818e-02 1.12194426e-01 -1.12886500e+00
-8.89719963e-01 3.41507256e-01 1.10928807e-02 3.72490376e-01
9.67124820e-01 2.83480346e-01 8.24801087e-01 -8.18747759e-01
5.57095647e-01 1.10018992e+00 2.99461842e-01 3.53875518e-01
-1.19914031e+00 -6.77025318e-01 -3.28358382e-01 4.17200744e-01
-1.25300825e+00 -6.27597153e-01 2.70537376e-01 -3.87855440e-01
9.30321991e-01 5.81658840e-01 4.28448498e-01 1.12002015e+00
1.00141741e-01 5.73049605e-01 1.16221178e+00 -4.04329866e-01
2.20480487e-01 2.65160888e-01 5.60413897e-01 7.79591262e-01
1.79567412e-02 -5.31332970e-01 -7.20064521e-01 -3.34074318e-01
4.25859541e-01 2.68548466e-02 -1.93845347e-01 -1.06799826e-01
-1.11641502e+00 4.16317195e-01 -1.08492803e-02 2.90486872e-01
-2.15302810e-01 -2.91360170e-01 1.73760056e-01 6.64023638e-01
-4.67386767e-02 8.91592979e-01 -9.48643148e-01 -5.72707295e-01
-7.78652608e-01 3.70044440e-01 1.27471113e+00 1.20549428e+00
1.17420197e+00 -5.78068733e-01 -2.70343751e-01 9.45380270e-01
4.15466815e-01 6.24925554e-01 6.99349701e-01 -9.40371692e-01
4.37220544e-01 7.41333246e-01 3.82387936e-02 -8.08524370e-01
-4.51264530e-01 -1.94728732e-01 -4.28980649e-01 -4.06603873e-01
4.54133987e-01 8.87630507e-02 -7.82610774e-01 1.49086726e+00
6.31632328e-01 -3.67001481e-02 -1.36956424e-01 8.28519940e-01
1.23297274e+00 6.15095437e-01 1.09542660e-01 3.52432579e-01
1.75889015e+00 -9.09133613e-01 -5.42433381e-01 -5.04625201e-01
5.81267536e-01 -9.78891015e-01 1.67043710e+00 5.04956424e-01
-1.40206969e+00 -4.45310801e-01 -5.77368498e-01 -7.13526845e-01
-8.35668564e-01 -3.81415665e-01 5.75578094e-01 6.07014894e-01
-1.40375769e+00 5.16774774e-01 -9.30660188e-01 -7.18109906e-01
-1.61753930e-02 2.18083903e-01 -3.02678496e-01 -4.23934579e-01
-1.17197156e+00 9.50227082e-01 2.32771888e-01 -2.52815306e-01
-7.71012247e-01 -9.19009149e-01 -8.34720492e-01 3.51846904e-01
3.85854393e-01 -1.06721747e+00 1.53920078e+00 -3.72384161e-01
-1.45450830e+00 7.20005572e-01 -1.30243063e-01 1.41519353e-01
-7.49108121e-02 -4.12896782e-01 -2.96663642e-01 2.37805143e-01
3.24315876e-01 4.06062216e-01 3.22886974e-01 -5.21837413e-01
-2.60630220e-01 -6.19488358e-01 3.71322036e-01 3.43897969e-01
-3.46384883e-01 4.84328836e-01 -8.88140202e-01 -4.12292518e-02
-3.77026089e-02 -5.32439768e-01 1.52580321e-01 -1.86597392e-01
-2.75441885e-01 -3.18765759e-01 5.44257045e-01 -9.64863896e-01
1.29800177e+00 -1.98504508e+00 1.65694714e-01 1.28645554e-01
7.63776600e-02 2.94359148e-01 -1.47053018e-01 9.15167511e-01
2.27823287e-01 3.09980690e-01 3.12142707e-02 -1.75762847e-01
4.72888201e-01 2.36477122e-01 -2.84609031e-02 1.36990808e-02
7.96567798e-02 7.56071627e-01 -9.73867595e-01 -6.71858013e-01
2.39652067e-01 3.91847908e-01 -6.05260313e-01 4.55525190e-01
-4.69616532e-01 3.83317441e-01 -6.66179121e-01 8.91537249e-01
6.09558463e-01 -6.78864837e-01 2.76949614e-01 1.49223313e-01
-2.23034859e-01 8.00300300e-01 -1.17527604e+00 2.13249373e+00
-8.84952664e-01 4.52628046e-01 2.83689737e-01 -6.84843719e-01
7.90180027e-01 2.85257697e-01 1.79288555e-02 -7.39085436e-01
-2.43791386e-01 2.76519865e-01 -3.62347811e-01 -7.22510576e-01
7.32726038e-01 4.51539934e-01 -5.19637130e-02 6.41056955e-01
5.31213522e-01 -1.69089645e-01 6.39999688e-01 3.97716612e-01
1.45233846e+00 3.04744989e-01 1.36217862e-01 -3.57920796e-01
2.94234663e-01 -2.80982051e-02 2.76266694e-01 7.33207941e-01
1.18386313e-01 4.44053739e-01 5.00084817e-01 -9.11200568e-02
-7.70725369e-01 -1.11289132e+00 -4.09385651e-01 1.69146025e+00
-6.95067942e-02 -8.03147614e-01 -3.91742915e-01 -2.00809240e-01
-1.04908779e-01 6.02403522e-01 4.29280102e-02 1.11921273e-01
-1.28289059e-01 -2.21026301e-01 5.70934296e-01 2.24535704e-01
4.80999768e-01 -1.14734149e+00 -5.78029573e-01 1.63468838e-01
-3.92496973e-01 -8.29449594e-01 -5.25744036e-02 -1.26954988e-01
-6.47212982e-01 -9.93469059e-01 -4.27079827e-01 -7.46588349e-01
2.79879868e-01 2.43796483e-01 1.78100789e+00 1.85517445e-01
-2.79558420e-01 9.77486610e-01 -4.05231208e-01 -2.86412150e-01
-7.89803714e-02 3.06084692e-01 -2.34213293e-01 -5.70488751e-01
4.77021217e-01 -7.52822042e-01 -8.10265362e-01 2.11813301e-01
-1.25978637e+00 4.63713743e-02 5.66877544e-01 5.27053714e-01
6.26756370e-01 -2.92692244e-01 5.21609664e-01 -7.88726091e-01
1.01341903e+00 -8.79361987e-01 -8.15766037e-01 7.87846744e-01
-9.03941691e-01 4.07411084e-02 5.98010123e-01 -4.59630974e-02
-9.60402250e-01 -4.18906271e-01 -5.57080984e-01 -1.44833878e-01
-5.62818110e-01 9.68409419e-01 -2.76058406e-01 1.25114709e-01
6.76356256e-01 3.56296122e-01 -1.80828422e-01 -8.39073360e-01
5.77345073e-01 8.99309099e-01 3.46869856e-01 -7.15017736e-01
5.24050295e-01 -8.51053670e-02 -5.63485086e-01 -5.76401472e-01
-7.18033433e-01 -7.73371935e-01 -2.32223958e-01 -3.25992033e-02
8.25337768e-01 -7.64180601e-01 -8.16132426e-01 2.78953403e-01
-8.92492771e-01 -3.53807122e-01 6.87524974e-02 1.07448943e-01
-1.67619735e-01 4.92266446e-01 -7.25566864e-01 -4.57835406e-01
-7.70950258e-01 -1.11603951e+00 1.09047651e+00 7.30065584e-01
-3.62375528e-01 -9.33668077e-01 2.51223862e-01 6.38990879e-01
8.52043569e-01 -2.99815446e-01 9.44667757e-01 -8.46577108e-01
-6.64531767e-01 -2.24655151e-01 -1.57306373e-01 7.49003291e-02
-2.98642397e-01 2.68517584e-01 -8.27664316e-01 -2.63055772e-01
-3.27797085e-01 -6.21729910e-01 1.99601576e-01 -8.29565004e-02
1.32420385e+00 -3.17633063e-01 -3.48271310e-01 5.04441202e-01
1.29423356e+00 -5.80833480e-02 7.44690299e-01 4.64299381e-01
1.75060257e-01 5.39908051e-01 4.25149441e-01 9.80754495e-02
6.94982111e-01 4.86688405e-01 1.88947111e-01 4.47637856e-01
4.37046617e-01 -6.31180182e-02 3.45006913e-01 1.15350699e+00
3.04551393e-01 -2.00080991e-01 -1.36136734e+00 5.10872781e-01
-1.71545458e+00 -6.39932334e-01 -1.28950715e-01 2.17322373e+00
1.03216994e+00 -2.38013804e-01 -2.00511411e-01 -3.59052598e-01
3.05363953e-01 -1.42656922e-01 -5.04905224e-01 -4.70116764e-01
3.02687973e-01 8.38419974e-01 -1.12074226e-01 2.66956598e-01
-6.22701585e-01 8.53235126e-01 5.57644749e+00 8.28473330e-01
-8.69232833e-01 1.98641881e-01 1.46868825e-01 -5.56376530e-03
-4.94309723e-01 6.31154552e-02 -8.35410237e-01 5.93956172e-01
1.51104665e+00 -4.52405959e-01 7.64006615e-01 9.64033961e-01
-3.21813613e-01 -2.05691800e-01 -9.87558603e-01 8.04810941e-01
-2.06028894e-01 -1.31232452e+00 -1.78269088e-01 -3.60229462e-01
1.85989290e-01 4.35015798e-01 -1.29011393e-01 7.53161967e-01
5.93297899e-01 -9.50315237e-01 4.08873767e-01 7.65332043e-01
7.20569551e-01 -4.23660964e-01 6.07009888e-01 5.13296783e-01
-7.92749166e-01 1.05754174e-01 -3.95809501e-01 2.33487394e-02
6.67636143e-03 5.29261887e-01 -3.78392488e-01 1.05464792e+00
1.28542650e+00 2.28517458e-01 -8.94793153e-01 1.18767667e+00
-1.72046721e-01 6.14496112e-01 -5.94102263e-01 -1.40263975e-01
-3.23312372e-01 -4.23128664e-01 5.05694188e-02 1.13088369e+00
5.94914615e-01 -7.03912927e-03 7.33354837e-02 7.78139353e-01
1.17413998e-01 2.56234705e-01 -3.00092876e-01 -1.81412667e-01
8.50258231e-01 1.69781852e+00 -2.79286146e-01 -2.64879018e-01
-5.55429757e-01 6.54582739e-01 5.72509766e-01 3.08762252e-01
-5.23283243e-01 -6.93661273e-01 6.06025040e-01 1.64142147e-01
-8.64050984e-02 -3.85541052e-01 1.44005761e-01 -1.26265001e+00
8.98500606e-02 -1.41229630e+00 6.42807424e-01 -1.22864008e+00
-1.47112691e+00 5.06591082e-01 1.70711413e-01 -5.32140255e-01
-3.46354544e-01 -5.52351952e-01 -4.33172405e-01 1.15453732e+00
-1.42545986e+00 -8.30157042e-01 -7.07846761e-01 9.64881599e-01
1.42376691e-01 1.67645335e-01 1.13370109e+00 6.35809600e-01
-6.37763560e-01 1.83283895e-01 6.02379143e-02 2.35639438e-02
9.59826469e-01 -1.38175619e+00 2.72470117e-01 5.28871953e-01
2.85429060e-01 1.31014824e+00 3.98388535e-01 -3.11373293e-01
-1.99145281e+00 -7.42858887e-01 9.29017365e-01 -8.19291532e-01
9.47130620e-01 -4.39452499e-01 -1.03203368e+00 9.13881421e-01
4.64082956e-01 -3.71195376e-01 1.00811779e+00 5.85257888e-01
-4.06728327e-01 -2.76464850e-01 -9.78226006e-01 5.46811044e-01
8.00070465e-01 -7.53281355e-01 -6.76143706e-01 7.43079245e-01
6.19690239e-01 -7.20067143e-01 -1.12890363e+00 4.50929515e-02
3.86808664e-01 -1.03809798e+00 9.01652992e-01 -5.73918283e-01
4.51008230e-01 -2.18289986e-01 7.16792569e-02 -1.08922303e+00
-3.09037387e-01 -6.57938480e-01 -1.55105963e-01 1.37233007e+00
5.84270954e-01 -9.45898294e-01 3.19819778e-01 1.10097444e+00
-3.02717268e-01 -6.94416702e-01 -7.10685015e-01 -5.16981959e-01
-4.88047265e-02 -5.54476380e-01 8.89636815e-01 8.74375463e-01
3.08988899e-01 5.60375392e-01 3.01001400e-01 3.43142867e-01
1.54810488e-01 4.72687721e-01 8.58449340e-01 -1.37853551e+00
-5.90848386e-01 -1.97473213e-01 -1.13811918e-01 -1.24195528e+00
-2.43148267e-01 -1.06072521e+00 -1.68362886e-01 -2.02747536e+00
1.22938551e-01 -6.34974480e-01 -9.79795828e-02 7.53415704e-01
-1.64405331e-01 -4.92278151e-02 -1.00691624e-01 4.64226186e-01
-9.62643325e-01 1.55250996e-01 9.17840302e-01 2.31543928e-01
-5.57098491e-03 -2.32184142e-01 -8.48884881e-01 3.78564745e-01
1.01188850e+00 -3.01269293e-01 -3.07224572e-01 -8.71363103e-01
8.72764349e-01 9.86965895e-02 4.15951520e-01 -1.11160076e+00
5.75864851e-01 1.24304906e-01 9.72072110e-02 -4.42903817e-01
2.53364086e-01 -5.89192748e-01 1.80160969e-01 -2.59581842e-02
-4.50654402e-02 5.00300169e-01 3.53631973e-01 2.00759292e-01
-2.14204371e-01 -5.13307929e-01 1.14366204e-01 -3.60098243e-01
-5.02789617e-01 2.99278706e-01 -3.52952540e-01 3.05795819e-01
7.23012209e-01 3.68679613e-01 -9.32325065e-01 -3.26564163e-01
-6.23631120e-01 6.28119290e-01 4.96102929e-01 3.52644563e-01
3.34160477e-01 -9.13037360e-01 -3.99096400e-01 9.76311937e-02
3.24455351e-01 2.96663851e-01 4.67441469e-01 7.57183850e-01
-7.31486022e-01 5.22937775e-01 -1.16571292e-01 -4.06294674e-01
-1.11486256e+00 4.87584859e-01 8.89948606e-02 -5.82251608e-01
-3.38906199e-01 6.43185794e-01 -2.89762914e-01 -1.03265321e+00
-1.41283602e-01 -7.66191408e-02 -1.41754687e-01 -1.40302842e-02
6.95684135e-01 1.96098283e-01 4.85783547e-01 1.37957111e-01
-4.76045102e-01 1.35171816e-01 -9.07842908e-03 -3.73332202e-02
1.28753901e+00 7.58813247e-02 -4.96539474e-01 3.28801811e-01
7.50022292e-01 8.79059955e-02 -6.49945796e-01 -2.03699932e-01
9.91795734e-02 -4.37335879e-01 -1.20112762e-01 -1.05752957e+00
-7.25075483e-01 7.61394203e-01 3.91729444e-01 5.29610574e-01
1.08470368e+00 2.78852105e-01 6.03729546e-01 6.29719377e-01
5.61162591e-01 -9.64846671e-01 -1.93377614e-01 5.68199933e-01
8.37154150e-01 -9.46806848e-01 -4.64748045e-05 -1.03926085e-01
-2.61366814e-01 9.97975767e-01 6.13214910e-01 2.85675853e-01
5.53235710e-01 1.73170730e-01 9.89950821e-02 -6.02590799e-01
-1.01369643e+00 -3.82199168e-01 1.51413411e-01 5.76860368e-01
6.56533599e-01 -8.41232836e-02 -2.71322429e-01 6.10117555e-01
-5.70558548e-01 1.35967955e-01 3.20222408e-01 1.04540241e+00
-2.26566643e-01 -1.34289408e+00 -4.44094360e-01 6.32231176e-01
-5.46287715e-01 -4.45787996e-01 -1.68800831e-01 6.69377983e-01
-2.46211022e-01 1.18206453e+00 2.22882139e-03 -1.96546122e-01
3.93148959e-01 5.46948552e-01 3.08867007e-01 -9.15734529e-01
-9.29738283e-01 -5.21096349e-01 1.63585737e-01 -6.65312946e-01
-6.42862990e-02 -5.38220406e-01 -8.35072100e-01 -5.22923708e-01
-2.19010606e-01 3.96338671e-01 8.10707331e-01 7.22813666e-01
1.02006924e+00 2.26037204e-01 -1.55818567e-01 -3.27065051e-01
-4.90068585e-01 -1.16879797e+00 -3.66840005e-01 1.88071921e-01
-3.53160411e-01 -3.21805865e-01 -1.13356955e-01 -1.63861230e-01] | [10.860869407653809, 8.330595016479492] |
a1297aaa-8b21-4664-b127-fc75072561b7 | stochastic-precision-ensemble-self-knowledge | 2009.14502 | null | https://arxiv.org/abs/2009.14502v1 | https://arxiv.org/pdf/2009.14502v1.pdf | Stochastic Precision Ensemble: Self-Knowledge Distillation for Quantized Deep Neural Networks | The quantization of deep neural networks (QDNNs) has been actively studied for deployment in edge devices. Recent studies employ the knowledge distillation (KD) method to improve the performance of quantized networks. In this study, we propose stochastic precision ensemble training for QDNNs (SPEQ). SPEQ is a knowledge distillation training scheme; however, the teacher is formed by sharing the model parameters of the student network. We obtain the soft labels of the teacher by changing the bit precision of the activation stochastically at each layer of the forward-pass computation. The student model is trained with these soft labels to reduce the activation quantization noise. The cosine similarity loss is employed, instead of the KL-divergence, for KD training. As the teacher model changes continuously by random bit-precision assignment, it exploits the effect of stochastic ensemble KD. SPEQ outperforms the existing quantization training methods in various tasks, such as image classification, question-answering, and transfer learning without the need for cumbersome teacher networks. | ['Sungho Shin', 'Yoonho Boo', 'Jungwook Choi', 'Wonyong Sung'] | 2020-09-30 | null | null | null | null | ['self-knowledge-distillation'] | ['computer-vision'] | [ 1.88117445e-01 -2.94647133e-03 -2.66435921e-01 -6.49294198e-01
-7.87980020e-01 -4.11461920e-01 1.82079598e-01 1.18535087e-01
-8.59288752e-01 8.96479070e-01 -3.35448116e-01 -5.16534805e-01
-1.45716488e-01 -1.00407112e+00 -9.65138555e-01 -1.03771269e+00
3.54120672e-01 1.77442178e-01 4.67766434e-01 3.81757841e-02
2.61052698e-02 2.24841744e-01 -1.40661895e+00 2.72511125e-01
1.06066895e+00 1.61495340e+00 1.17705323e-01 6.92755222e-01
-1.08747624e-01 9.62393284e-01 -9.57534730e-01 -7.77358890e-01
1.47031307e-01 -3.78746390e-01 -5.68284750e-01 -4.77337241e-01
4.65423435e-01 -4.76489305e-01 -5.21164238e-01 1.55325711e+00
8.86671662e-01 5.74654192e-02 6.05946124e-01 -1.16541600e+00
-1.00002718e+00 7.68243194e-01 -1.15887128e-01 4.36867684e-01
-6.51891351e-01 -7.70853786e-03 7.93646276e-01 -7.67990351e-01
-5.48630171e-02 1.07987952e+00 5.78214765e-01 6.64112449e-01
-9.47187722e-01 -1.13671243e+00 -2.57643819e-01 5.27820230e-01
-1.85141742e+00 -5.02759099e-01 5.02957702e-01 -2.11289227e-01
7.84618556e-01 -3.44129473e-01 3.09494436e-01 5.68622530e-01
1.95000067e-01 7.49402463e-01 7.27292001e-01 -5.44456840e-01
8.49424303e-01 4.81051683e-01 2.70973542e-03 8.44501853e-01
1.27582148e-01 -1.73404999e-02 -8.34776700e-01 -1.02992363e-01
7.81818330e-01 -1.22849509e-01 -2.98771739e-01 -1.63460761e-01
-5.42342126e-01 8.02561879e-01 5.44953644e-01 -3.64778526e-02
-3.08620393e-01 6.25594497e-01 4.53412712e-01 3.30378354e-01
3.31256688e-01 4.42959135e-03 -7.58726418e-01 -2.85347313e-01
-9.36915398e-01 -2.07815006e-01 5.65558910e-01 9.25038815e-01
9.07828450e-01 2.55769730e-01 -5.21727741e-01 6.26765609e-01
3.82931322e-01 4.55539048e-01 7.85560668e-01 -1.13120222e+00
3.36603642e-01 3.68158668e-01 -2.16620550e-01 -5.71213603e-01
2.81644881e-01 -5.46603501e-01 -1.15049195e+00 4.46271934e-02
1.09856732e-01 -4.98783708e-01 -1.11043382e+00 1.67045617e+00
4.50264394e-01 5.84825516e-01 2.42968947e-01 6.33731365e-01
6.85312450e-01 6.65509880e-01 2.15600859e-02 -6.40156791e-02
1.10769796e+00 -8.83272827e-01 -1.06535852e+00 4.83058095e-01
8.89109612e-01 -2.02526107e-01 7.73148477e-01 6.64195836e-01
-1.04325342e+00 -6.67542398e-01 -1.51369667e+00 -2.52811581e-01
-4.85022426e-01 3.39296043e-01 3.62795919e-01 1.21245468e+00
-1.26930475e+00 8.78857613e-01 -9.70263124e-01 4.92600530e-01
9.26985741e-01 8.51596713e-01 4.30627048e-01 2.32225861e-02
-1.51463342e+00 5.24445891e-01 6.00510657e-01 -8.45501944e-03
-7.49777853e-01 -8.10422957e-01 -6.21708393e-01 2.33763322e-01
1.64211057e-02 -7.39685416e-01 1.49727082e+00 -1.05336070e+00
-2.19320655e+00 3.41639608e-01 1.09039694e-01 -8.06442857e-01
1.97454631e-01 -3.49586681e-02 -2.70093828e-01 2.11558670e-01
-3.50378960e-01 9.37567949e-01 1.18666530e+00 -5.33951640e-01
-5.40801764e-01 -4.29794669e-01 -8.29129890e-02 1.52389273e-01
-7.32065678e-01 -5.37295341e-01 -3.63265544e-01 -5.41256785e-01
-7.56729692e-02 -6.07185781e-01 6.13541119e-02 2.84132421e-01
-1.06770378e-02 -7.18173087e-01 8.46671700e-01 -2.43581191e-01
1.38296843e+00 -2.51566744e+00 -4.87407357e-01 2.46031180e-01
1.76854923e-01 7.33432174e-01 1.54905990e-01 -2.53071159e-01
2.53385633e-01 1.02951251e-01 -4.61845584e-02 -4.59578663e-01
1.69152081e-01 5.85322201e-01 -2.65083760e-01 2.20988125e-01
3.36230159e-01 9.76612926e-01 -9.99542117e-01 -5.53145647e-01
-8.30655545e-02 5.66888094e-01 -6.37311935e-01 -2.85114478e-02
-3.39041770e-01 4.36630566e-03 -2.55896032e-01 1.45857126e-01
6.89245999e-01 -6.35863721e-01 1.01463003e-02 -4.07465041e-01
2.61399060e-01 5.69750190e-01 -1.10654449e+00 1.93967962e+00
-3.93350780e-01 5.26425779e-01 -9.89954844e-02 -1.00161028e+00
8.67932320e-01 4.86337572e-01 -4.13270146e-02 -7.82012522e-01
2.24654600e-01 3.51857543e-01 -3.30372192e-02 -1.89373434e-01
4.30173337e-01 -8.15437958e-02 2.41769612e-01 1.50792629e-01
4.93974686e-01 -1.21168695e-01 -2.27724716e-01 2.03970253e-01
7.05078006e-01 -3.96873802e-01 -1.68047994e-01 -9.27345827e-02
3.33902508e-01 -5.16886413e-01 5.38045704e-01 7.24015415e-01
-3.92592490e-01 2.02213541e-01 1.71694368e-01 1.96838193e-02
-7.18462348e-01 -1.31498086e+00 -3.17731321e-01 1.21366704e+00
1.05295338e-01 -2.98437119e-01 -8.72118533e-01 -6.51540279e-01
-7.71418288e-02 6.11523271e-01 -2.75255919e-01 -7.63859272e-01
-1.60208978e-02 -5.59620500e-01 9.88843858e-01 6.14605069e-01
1.11011803e+00 -6.11680448e-01 -4.44049925e-01 1.80862024e-01
2.38144457e-01 -1.01005077e+00 -5.96912920e-01 6.18866563e-01
-1.02376473e+00 -4.17466879e-01 -7.16534615e-01 -8.11647654e-01
5.20360470e-01 -2.50483334e-01 8.40873659e-01 -1.47525176e-01
1.28474861e-01 1.22600399e-01 -1.81098640e-01 -5.24693370e-01
-5.89302257e-02 1.74688920e-01 2.75097966e-01 2.85359118e-02
6.64449334e-01 -7.20307767e-01 -8.30183327e-01 4.10578437e-02
-1.12345910e+00 -3.69048119e-01 6.04062557e-01 9.96114671e-01
8.26450229e-01 4.96833622e-01 9.10152853e-01 -5.55289209e-01
6.92605495e-01 -2.56272763e-01 -6.95779383e-01 3.80636394e-01
-6.83697402e-01 6.18392408e-01 5.58761537e-01 -6.72870874e-01
-7.18730628e-01 -9.49818715e-02 -1.67115018e-01 -8.72112095e-01
2.48791888e-01 1.90905914e-01 -8.21214169e-02 -3.54927421e-01
7.55036533e-01 1.83738306e-01 -2.85786062e-01 -2.87813276e-01
4.33019370e-01 9.58060801e-01 5.82065105e-01 -5.05094171e-01
3.61970693e-01 1.57085553e-01 -1.10850751e-01 -5.05080760e-01
-9.13447440e-01 -9.21749920e-02 -3.85982752e-01 2.28838339e-01
9.48626161e-01 -1.03519189e+00 -1.04586995e+00 5.68769097e-01
-1.18774414e+00 -4.50913101e-01 -5.52488863e-01 5.93982518e-01
-7.75806978e-02 -7.48719275e-03 -6.29174292e-01 -8.22431803e-01
-6.55460656e-01 -1.14257312e+00 9.39703763e-01 5.71488917e-01
3.38795036e-01 -9.84327734e-01 -3.46821487e-01 3.97968367e-02
4.20979977e-01 -4.95622188e-01 1.06193674e+00 -5.45361936e-01
-5.94069660e-01 2.50876039e-01 -3.59441340e-01 1.02018225e+00
1.07959971e-01 -2.63159394e-01 -1.16638935e+00 -1.30995318e-01
8.95562023e-02 -7.65859246e-01 9.63131726e-01 3.56011242e-01
1.81170118e+00 -3.16450149e-01 9.84899178e-02 6.61451161e-01
1.25860834e+00 2.50551522e-01 4.82971251e-01 -1.79727376e-01
4.15891796e-01 -1.79326817e-01 -8.04644544e-03 4.55378473e-01
5.51534593e-01 3.09064031e-01 2.73195714e-01 3.84084940e-01
-1.57356545e-01 -3.09443086e-01 3.57914656e-01 1.02446413e+00
4.22132075e-01 -1.00016490e-01 -8.35763693e-01 3.94110888e-01
-1.51962352e+00 -5.55162609e-01 3.34866166e-01 2.22191095e+00
1.36638021e+00 3.16297054e-01 -5.10853469e-01 3.66150975e-01
5.48669815e-01 -2.59499937e-01 -9.67519343e-01 -5.93079805e-01
2.08277017e-01 6.81634903e-01 8.71722996e-01 3.32506150e-01
-9.61804569e-01 6.75559342e-01 5.31764698e+00 1.39185131e+00
-1.32275009e+00 3.83608609e-01 8.73996437e-01 -1.06172316e-01
-1.11883439e-01 -5.68459392e-01 -1.19999421e+00 6.72300875e-01
1.32863247e+00 3.04578155e-01 3.68780166e-01 7.06705093e-01
-2.15968683e-01 -2.57790715e-01 -1.15358889e+00 1.31821251e+00
-1.62353724e-01 -1.25599062e+00 4.98325601e-02 -2.48680785e-01
1.24855936e+00 2.01367423e-01 6.73643172e-01 5.56026161e-01
4.23435956e-01 -1.00500822e+00 5.67804158e-01 4.59400922e-01
1.02862680e+00 -9.60528970e-01 7.48136640e-01 5.41557193e-01
-8.52616966e-01 -2.27178764e-02 -4.69510585e-01 4.06057835e-02
-2.35500351e-01 9.80229378e-01 -8.30523372e-01 2.16659516e-01
8.51980031e-01 2.57683933e-01 -3.52144718e-01 8.15272391e-01
-2.88953543e-01 9.60685849e-01 -5.31131983e-01 -2.39030480e-01
3.80344450e-01 -1.54267073e-01 -2.07893610e-01 7.65694559e-01
3.15493315e-01 2.98609674e-01 -1.37313247e-01 7.81125546e-01
-7.17385232e-01 -3.44937950e-01 -3.32445242e-02 -1.43596441e-01
8.59039903e-01 7.49871373e-01 -2.66374290e-01 -6.37102187e-01
-3.27669382e-02 1.21329272e+00 4.24503893e-01 3.54190230e-01
-8.18117917e-01 -7.79541016e-01 4.45411980e-01 -4.12446022e-01
6.97214961e-01 -2.17283845e-01 -4.26166624e-01 -6.77991867e-01
6.05295338e-02 -4.77339089e-01 1.64071590e-01 -6.66397810e-01
-1.18579245e+00 1.86749443e-01 -2.68225074e-01 -8.88573825e-01
-6.79768994e-02 -5.36403537e-01 -4.15551662e-01 1.10594285e+00
-1.87339139e+00 -3.23142797e-01 -7.98910186e-02 7.22777963e-01
2.00307146e-02 -1.48838433e-02 7.80240238e-01 6.87971354e-01
-5.60448945e-01 1.27251816e+00 6.58592284e-01 3.12819481e-01
6.62233889e-01 -1.19534993e+00 -6.04287013e-02 1.72341466e-01
-3.00894375e-03 5.83072782e-01 1.58723801e-01 -3.24580818e-01
-1.06194997e+00 -1.46151626e+00 7.44777083e-01 8.88357460e-02
5.66622436e-01 -2.28250369e-01 -9.42679524e-01 1.62887827e-01
1.32555738e-01 4.53695208e-01 9.03535903e-01 -3.51154119e-01
-3.19299489e-01 -5.70725203e-01 -1.24650335e+00 2.29937181e-01
5.56208134e-01 -9.44188833e-01 -2.52077162e-01 2.62501955e-01
1.01952982e+00 -4.07127708e-01 -1.01702225e+00 4.36926872e-01
3.99227679e-01 -6.11042023e-01 6.73623800e-01 -2.81478554e-01
2.23276407e-01 -3.18545580e-01 -2.13582322e-01 -1.30246139e+00
1.46510050e-01 -4.02618766e-01 -5.34408331e-01 1.04779911e+00
3.24645013e-01 -4.75065291e-01 1.12418556e+00 3.91154647e-01
-1.30391017e-01 -1.12628818e+00 -1.27357018e+00 -7.82384872e-01
2.76253790e-01 -2.48173878e-01 6.72787130e-01 6.85522079e-01
-2.09137112e-01 4.94600564e-01 1.73221603e-01 3.35800320e-01
5.28950989e-01 -5.79435349e-01 1.37462333e-01 -9.49376822e-01
-5.00714064e-01 -2.22463205e-01 -4.81683880e-01 -1.64643657e+00
-1.09226190e-01 -8.97820234e-01 1.14097610e-01 -9.58652496e-01
-5.10496080e-01 -6.31166697e-01 -5.44282615e-01 2.74782956e-01
-2.05164522e-01 1.05725072e-01 -1.48746192e-01 -8.81509110e-02
-9.46410239e-01 1.05891371e+00 1.29209256e+00 -2.70773411e-01
-1.83703110e-01 3.22888978e-02 -2.86313504e-01 5.68326473e-01
9.68178689e-01 -6.33394778e-01 -8.44627023e-01 -7.43741095e-01
3.75213116e-01 -1.27240017e-01 2.51225650e-01 -1.23506033e+00
8.89685452e-01 5.09915590e-01 4.87439007e-01 -6.41139746e-01
4.56224889e-01 -8.38859200e-01 -4.44055468e-01 5.00181377e-01
-3.65944266e-01 -1.69385716e-01 1.27091736e-01 6.87119842e-01
-1.85935304e-01 -5.98155022e-01 9.12646472e-01 2.46378079e-01
-4.10600662e-01 5.21274805e-01 -1.81566983e-01 1.39915690e-01
6.37587845e-01 -6.44865558e-02 -1.89935178e-01 -1.74874365e-01
-5.71412802e-01 4.40535069e-01 -2.52436012e-01 -1.11376964e-01
7.43650794e-01 -1.43795764e+00 -1.46222696e-01 3.14725816e-01
-2.93146312e-01 7.37190008e-01 1.10053927e-01 6.88956320e-01
-3.05002749e-01 3.20975363e-01 3.03428739e-01 -6.29860938e-01
-6.90989256e-01 2.60186136e-01 7.01383293e-01 -2.34409899e-01
4.24421206e-02 1.38007140e+00 -2.11767882e-01 -3.21171790e-01
8.30050349e-01 -5.86420596e-01 6.68707341e-02 -1.44096911e-01
6.46948338e-01 3.32863003e-01 4.36093986e-01 2.62813687e-01
-6.63210377e-02 1.78496122e-01 -2.83994585e-01 -1.00223064e-01
8.18768620e-01 9.71563905e-03 3.68962698e-02 6.06594265e-01
1.52666557e+00 -6.49618566e-01 -1.27556229e+00 -6.21419728e-01
-1.42624393e-01 2.25944340e-01 4.89626616e-01 -7.12256074e-01
-1.06542671e+00 1.33105588e+00 1.05919814e+00 -1.60235882e-01
1.24982488e+00 -2.68067241e-01 9.25240219e-01 7.83831954e-01
2.61375308e-01 -1.26449513e+00 1.89114258e-01 8.42829168e-01
8.06281120e-02 -1.19038212e+00 -2.87324190e-01 -4.68596965e-02
-1.60347611e-01 9.33543921e-01 7.09253430e-01 4.49481457e-02
1.19897509e+00 3.13206613e-01 -5.66726066e-02 8.90882313e-02
-8.55048656e-01 1.23752393e-01 9.97826383e-02 4.65852439e-01
6.40161857e-02 -4.09155712e-02 -4.82881740e-02 7.83734024e-01
-2.61516154e-01 3.13119650e-01 5.42888306e-02 7.94719875e-01
-6.17067516e-01 -8.51331055e-01 -1.52541935e-01 6.48952782e-01
-4.17744726e-01 -5.06237328e-01 6.37497529e-02 -1.72320291e-01
6.56953871e-01 7.30947673e-01 3.69860291e-01 -5.76135993e-01
-7.14490339e-02 2.90826172e-01 5.32434344e-01 -6.04280114e-01
-5.57528973e-01 -5.53076982e-01 -3.85698438e-01 -8.55425745e-02
-3.15706551e-01 -1.79823246e-02 -1.59437740e+00 -1.16588078e-01
-7.32897580e-01 5.25699914e-01 8.72845769e-01 9.37326789e-01
5.02566040e-01 8.33472371e-01 5.76331079e-01 -3.35215181e-01
-1.22225726e+00 -7.66308248e-01 -6.94422483e-01 -2.18353659e-01
4.71618831e-01 -5.44590771e-01 -2.74364561e-01 1.48169473e-01] | [8.71484088897705, 3.010143518447876] |
672ea043-83d6-4e48-915a-b380e9a1046a | are-cars-just-3d-boxes-jointly-estimating-the | null | null | http://openaccess.thecvf.com/content_cvpr_2014/html/Zia_Are_Cars_Just_2014_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2014/papers/Zia_Are_Cars_Just_2014_CVPR_paper.pdf | Are Cars Just 3D Boxes? - Jointly Estimating the 3D Shape of Multiple Objects | Current systems for scene understanding typically represent objects as 2D or 3D bounding boxes. While these representations have proven robust in a variety of applications, they provide only coarse approximations to the true 2D and 3D extent of objects. As a result, object-object interactions, such as occlusions or ground-plane contact, can be represented only superficially. In this paper, we approach the problem of scene understanding from the perspective of 3D shape modeling, and design a 3D scene representation that reasons jointly about the 3D shape of multiple objects. This representation allows to express 3D geometry and occlusion on the fine detail level of individual vertices of 3D wireframe models, and makes it possible to treat dependencies between objects, such as occlusion reasoning, in a deterministic way. In our experiments, we demonstrate the benefit of jointly estimating the 3D shape of multiple objects in a scene over working with coarse boxes, on the recently proposed KITTI dataset of realistic street scenes. | ['Michael Stark', 'Muhammad Zeeshan Zia', 'Konrad Schindler'] | 2014-06-01 | null | null | null | cvpr-2014-6 | ['3d-shape-modeling'] | ['computer-vision'] | [-8.16950649e-02 3.82027566e-01 4.76739071e-02 -5.13863206e-01
-3.08740258e-01 -7.57416368e-01 8.46276939e-01 4.42481250e-01
1.64931387e-01 1.23528153e-01 2.14876458e-01 -2.91043043e-01
2.92166360e-02 -9.87920642e-01 -8.79867733e-01 -1.48311988e-01
1.10691287e-01 1.11376023e+00 6.10945404e-01 -8.57340023e-02
3.30763102e-01 1.13062239e+00 -1.83029127e+00 3.99263293e-01
4.89527136e-01 7.55166769e-01 -9.02283490e-02 6.99608684e-01
-5.08918285e-01 5.47675848e-01 -2.93612093e-01 -2.08307356e-01
2.03228757e-01 3.15942764e-01 -6.95863366e-01 6.72291338e-01
8.14618707e-01 -5.63224196e-01 -1.69256300e-01 6.09141946e-01
-1.88034594e-01 1.82863530e-02 7.55766034e-01 -1.20275009e+00
-1.73754111e-01 5.91283888e-02 -6.00941539e-01 -3.62366408e-01
5.89906931e-01 -4.20434251e-02 9.40260708e-01 -1.00070643e+00
8.01017940e-01 1.82244217e+00 4.86192971e-01 2.63423443e-01
-1.74268365e+00 -7.91438147e-02 6.72182202e-01 -2.89868027e-01
-1.27343071e+00 -4.47256416e-01 9.01379585e-01 -7.75023341e-01
1.03246331e+00 4.49963957e-01 7.01033831e-01 4.46373403e-01
-2.49418784e-02 9.55006838e-01 8.66642416e-01 -3.24816555e-01
2.66440690e-01 1.95555434e-01 5.87743759e-01 4.73316550e-01
4.66961384e-01 -1.16837166e-01 -1.84514925e-01 -5.58531761e-01
1.18599999e+00 2.63917726e-02 1.45495415e-01 -1.18594456e+00
-8.87115121e-01 6.52649224e-01 5.10370970e-01 -8.72540921e-02
-1.23876564e-01 5.84637165e-01 8.57319906e-02 -2.58138210e-01
6.83603227e-01 1.68652549e-01 -4.88957465e-01 2.08403230e-01
-3.48095030e-01 7.65977681e-01 9.99097466e-01 1.28854299e+00
8.69835198e-01 -2.94922024e-01 1.84311375e-01 4.74610925e-01
6.40317380e-01 4.64707285e-01 -7.02122092e-01 -1.14210844e+00
3.79309714e-01 8.69643629e-01 5.17076373e-01 -1.06254959e+00
-4.80164915e-01 -7.68027306e-02 -2.69565225e-01 6.06189728e-01
4.90832657e-01 4.09497768e-01 -1.11484551e+00 1.46003342e+00
8.76891255e-01 -5.69582023e-02 -2.18341962e-01 7.29091763e-01
7.75337636e-01 5.06371915e-01 1.34710819e-01 2.66051173e-01
1.47964180e+00 -5.84174752e-01 -3.99343222e-01 -5.19354105e-01
6.77072763e-01 -7.59400249e-01 6.54403746e-01 2.68436402e-01
-1.44081235e+00 -5.67677975e-01 -7.31711626e-01 -6.89648628e-01
-4.50814188e-01 -3.05654258e-01 8.40383828e-01 4.52200711e-01
-7.98205793e-01 2.45176435e-01 -9.60937977e-01 -3.37891281e-01
5.16277671e-01 4.54057217e-01 -1.97382569e-01 -3.29874575e-01
-4.61561978e-01 1.19304192e+00 2.71214664e-01 1.87094565e-02
-7.79657662e-01 -8.84945035e-01 -9.87553477e-01 7.05593377e-02
5.70445120e-01 -1.01135862e+00 1.23294330e+00 -1.84365943e-01
-7.74859428e-01 1.11562967e+00 -5.56330860e-01 -1.54479921e-01
5.34744799e-01 -5.54126024e-01 2.19141766e-01 -1.26916841e-01
-7.20457882e-02 7.14560866e-01 5.70010364e-01 -1.91120458e+00
-2.54345357e-01 -8.19145322e-01 8.57628942e-01 4.18716431e-01
7.07466245e-01 -2.01547191e-01 -5.10188639e-01 -2.43544251e-01
8.12994063e-01 -7.70604193e-01 -5.85000694e-01 6.83747411e-01
-5.67149103e-01 -1.24124199e-01 1.05207551e+00 -2.42122188e-01
6.06202722e-01 -2.26418114e+00 8.02945048e-02 1.86684936e-01
3.85802805e-01 -1.03893213e-01 1.71106264e-01 3.98779005e-01
1.71112064e-02 3.64476740e-01 -1.26421332e-01 -7.29202628e-01
2.34138995e-01 5.35592020e-01 -5.05814552e-01 4.68579203e-01
4.31283236e-01 7.14766741e-01 -7.09010422e-01 -3.76822293e-01
7.07034051e-01 7.27456510e-01 -8.53124499e-01 9.09609422e-02
-7.14560211e-01 3.01810771e-01 -9.81889367e-01 3.54335189e-01
9.75701630e-01 -2.52935886e-01 -1.97482221e-02 -2.40958452e-01
-1.19378321e-01 4.38526332e-01 -1.55725992e+00 1.68831038e+00
-5.68480194e-01 3.65859210e-01 2.00518385e-01 -4.56125677e-01
8.53265584e-01 1.86563984e-01 3.38457614e-01 -1.65860787e-01
-1.19793648e-02 -9.42867845e-02 -3.09735149e-01 -1.55424073e-01
5.28205216e-01 -5.93416728e-02 -9.41454917e-02 5.28918564e-01
-5.64602971e-01 -1.09375322e+00 -3.77497301e-02 2.63859630e-01
8.30345333e-01 4.68010753e-01 4.22963709e-01 -2.31839865e-01
2.49632284e-01 3.03965002e-01 1.59755975e-01 7.85766423e-01
3.47479194e-01 7.44788170e-01 4.76474583e-01 -7.72682786e-01
-1.36569643e+00 -1.19682562e+00 -5.35087824e-01 4.94899541e-01
6.64447606e-01 -4.06485170e-01 -5.25764763e-01 -2.69197881e-01
3.53622347e-01 8.62217069e-01 -6.95256114e-01 2.44073465e-01
-7.56191254e-01 -3.69930446e-01 -1.79682016e-01 6.00317001e-01
1.71100706e-01 -4.79417741e-01 -1.01607275e+00 2.73641407e-01
9.95862931e-02 -1.40207183e+00 4.02703993e-02 7.26346001e-02
-1.17649829e+00 -1.25399590e+00 -8.83808434e-02 -2.22855762e-01
9.32816148e-01 5.95521927e-01 1.38122177e+00 3.61658156e-01
-4.93545443e-01 5.31810164e-01 -1.29837230e-01 -5.51619232e-01
-3.18825781e-01 -5.62936544e-01 -1.62227452e-01 -1.82687327e-01
2.40690395e-01 -5.68820179e-01 -2.99711734e-01 4.20690149e-01
-6.58638477e-01 5.83645642e-01 -8.36602412e-03 2.21808344e-01
7.94988155e-01 5.72981276e-02 -3.61032873e-01 -1.22794831e+00
-7.78170899e-02 -2.11467609e-01 -6.90042615e-01 2.30265468e-01
2.72270173e-01 3.47880572e-02 1.52292654e-01 -3.11719865e-01
-1.28249848e+00 1.52826965e-01 2.71647990e-01 -3.84028852e-01
-7.97624171e-01 5.18449992e-02 -3.80093187e-01 2.06532583e-01
4.76147294e-01 -2.74345487e-01 -4.95876402e-01 -8.32905710e-01
7.02345669e-01 1.17058165e-01 1.96693033e-01 -1.03191411e+00
8.21139932e-01 1.02956939e+00 3.29433918e-01 -8.05128217e-01
-1.02701104e+00 -5.06520450e-01 -1.03997421e+00 -5.08053973e-02
9.14774716e-01 -9.91520941e-01 -6.02646470e-01 1.62763745e-01
-1.79840028e+00 -6.04492389e-02 -4.89164710e-01 1.97411552e-01
-7.60395348e-01 2.95029525e-02 -1.41712472e-01 -1.10098279e+00
5.66242754e-01 -1.16866040e+00 1.60627449e+00 -1.90684557e-01
-5.40544808e-01 -9.40650284e-01 -2.04508379e-01 3.29026133e-01
-6.70221075e-02 6.76787555e-01 1.31074142e+00 -2.27679610e-01
-1.19828808e+00 -3.50906283e-01 -3.95643502e-01 -2.65530378e-01
-7.45728891e-03 1.83843493e-01 -1.24187827e+00 1.13528199e-01
1.26376778e-01 -3.21099274e-02 4.22901899e-01 5.22707462e-01
1.20058906e+00 1.07990518e-01 -6.17697716e-01 3.51955444e-01
1.47242355e+00 6.36551306e-02 5.79804540e-01 -1.28631681e-01
8.05287421e-01 9.35382903e-01 5.69572270e-01 4.96827304e-01
4.90981966e-01 8.30485225e-01 7.59349823e-01 -1.69397905e-01
-2.75752157e-01 -3.49915564e-01 -5.27234912e-01 6.23304099e-02
-1.23072684e-01 -2.81298012e-01 -1.10439503e+00 4.07385886e-01
-1.91948283e+00 -5.56836426e-01 -7.59645402e-01 2.11480832e+00
9.21470076e-02 4.03659523e-01 -1.58887297e-01 9.77493916e-03
5.05713224e-01 1.45543396e-01 -7.59444833e-01 -3.44277233e-01
8.14802498e-02 -2.08349183e-01 2.97174305e-01 8.12133431e-01
-9.48132813e-01 7.84733057e-01 6.78803396e+00 3.70622575e-01
-4.27229196e-01 -1.93960607e-01 5.72473884e-01 2.74406355e-02
-7.26639628e-01 6.06875956e-01 -9.67301905e-01 -2.43715197e-01
2.11638197e-01 1.43281847e-01 1.14008173e-01 9.32688892e-01
2.01701835e-01 -4.39882606e-01 -1.67333424e+00 8.25540304e-01
-1.63001150e-01 -1.30484056e+00 2.87096202e-01 2.89078504e-01
6.13627553e-01 -2.70077229e-01 -2.52272993e-01 4.65573883e-03
5.48230469e-01 -9.90379214e-01 1.19724381e+00 3.97130281e-01
4.49697733e-01 -4.46951509e-01 9.41798761e-02 7.46055186e-01
-1.24208248e+00 2.69498616e-01 -2.73655236e-01 -3.87571692e-01
4.48631406e-01 6.71831071e-01 -6.46694422e-01 4.42263931e-01
4.32392478e-01 4.41295862e-01 -2.34609619e-01 9.15703952e-01
-1.12930037e-01 -4.14641239e-02 -7.56231964e-01 3.24926198e-01
3.35999876e-02 -1.00126453e-01 5.81718802e-01 6.66107714e-01
-2.10040256e-01 5.62420964e-01 3.28808516e-01 1.21652532e+00
2.80625910e-01 -3.10157210e-01 -7.97094285e-01 2.82238752e-01
3.36216807e-01 7.48754025e-01 -9.18409228e-01 -4.12382245e-01
-3.69754851e-01 1.31614059e-01 4.14193541e-01 4.10669744e-01
-7.88325906e-01 2.60592490e-01 9.20306623e-01 5.00863016e-01
2.50425905e-01 -7.14328110e-01 -8.68991733e-01 -1.22368073e+00
1.74096972e-01 -3.89506102e-01 -1.13657624e-01 -1.22640455e+00
-9.53146935e-01 2.38134533e-01 5.95425606e-01 -1.00340509e+00
1.42235696e-01 -8.17227602e-01 -4.31873292e-01 8.19753826e-01
-1.13288319e+00 -1.20965981e+00 -2.32154891e-01 1.79639772e-01
6.24394834e-01 8.80409956e-01 8.40880275e-01 -1.94879457e-01
-2.89049651e-02 -5.04306734e-01 -2.42667690e-01 -2.56548762e-01
-1.63216945e-02 -1.06380391e+00 8.58968377e-01 4.59800154e-01
2.19214872e-01 7.93855131e-01 1.02194595e+00 -6.87750101e-01
-1.47111177e+00 -6.61422729e-01 7.66432762e-01 -1.02025712e+00
3.21328342e-01 -8.88127983e-01 -9.86802161e-01 8.80663276e-01
-5.35901308e-01 1.07673101e-01 2.34009847e-01 4.39756453e-01
-6.73998773e-01 3.61548871e-01 -1.19740677e+00 6.91958845e-01
1.34669149e+00 -5.96128583e-01 -6.00791752e-01 4.15490270e-01
7.24919200e-01 -7.91191459e-01 -8.26104701e-01 5.61451256e-01
5.38995266e-01 -1.02466917e+00 1.53057432e+00 -7.70120382e-01
2.67609537e-01 -5.14951825e-01 -4.65838969e-01 -7.50367761e-01
-1.30184829e-01 -1.70119137e-01 -3.61846179e-01 7.74595797e-01
1.03778414e-01 -2.07965598e-01 9.05570507e-01 1.48523426e+00
-8.04125965e-02 -6.04411781e-01 -7.26423025e-01 -4.81688589e-01
-1.03492022e-01 -9.15617168e-01 8.50819945e-01 6.21546090e-01
-4.92547244e-01 1.38708338e-01 4.50972244e-02 5.90622783e-01
8.06550324e-01 5.42188525e-01 1.20450544e+00 -1.60092378e+00
-1.77479103e-01 -2.42285639e-01 -4.83681023e-01 -1.58636367e+00
1.48939751e-02 -4.28988457e-01 -4.68534678e-02 -1.76281929e+00
1.39355615e-01 -8.16619396e-01 5.54813683e-01 1.86376348e-01
1.86577186e-01 -8.49917158e-02 2.39079937e-01 1.10057019e-01
-2.62221515e-01 3.35478187e-01 1.56547213e+00 -8.79481435e-02
2.11652871e-02 -1.22424811e-01 -4.14151460e-01 1.31697381e+00
3.40676785e-01 -4.38000321e-01 -5.13835549e-01 -1.13486993e+00
1.86064377e-01 2.16726720e-01 8.05602252e-01 -5.53980172e-01
-1.85302347e-02 -4.03284848e-01 4.84846175e-01 -1.14004517e+00
1.10210121e+00 -1.19624805e+00 4.70837265e-01 9.09798518e-02
-1.93938315e-01 -4.14974242e-01 3.82881045e-01 5.76797962e-01
1.25018433e-01 -7.45391399e-02 5.17995834e-01 -4.12640631e-01
-5.46900630e-01 4.03049678e-01 -3.93661633e-02 -1.17300294e-01
1.07458329e+00 -7.35318244e-01 -1.56899139e-01 -1.56890631e-01
-1.07359874e+00 2.53886729e-01 8.62609684e-01 3.43482196e-01
5.50356328e-01 -9.98625875e-01 -3.99177462e-01 2.49237001e-01
1.04387350e-01 8.24133694e-01 4.31744814e-01 1.52410313e-01
-5.73298991e-01 4.87507939e-01 1.64003983e-01 -9.86692250e-01
-1.26368940e+00 5.15538990e-01 2.59918839e-01 5.04624918e-02
-7.66908288e-01 7.50492334e-01 1.15674126e+00 -5.01878560e-01
1.46196291e-01 -6.71979189e-01 1.32749632e-01 -3.09310317e-01
3.32145333e-01 2.56618440e-01 -1.54283106e-01 -7.87161708e-01
-2.89261103e-01 1.13161123e+00 -7.80877843e-02 5.11728860e-02
1.17546761e+00 -2.29602352e-01 -1.36336029e-01 6.27012551e-01
8.74413192e-01 -1.47291541e-01 -1.44337606e+00 -2.27250263e-01
-6.12551533e-02 -9.37122107e-01 -1.57229096e-01 -4.57748801e-01
-4.08996046e-01 1.16268992e+00 -2.51855031e-02 4.36924726e-01
4.10346895e-01 5.10570884e-01 2.55477101e-01 3.75332326e-01
7.24263549e-01 -5.07595122e-01 -1.88547418e-01 5.78253448e-01
9.77151811e-01 -1.06495357e+00 3.70993882e-01 -1.25512898e+00
-9.67919528e-02 1.15101695e+00 6.99344993e-01 -2.30576411e-01
7.36895859e-01 4.49665397e-01 -2.59342909e-01 -6.12006009e-01
-7.42109358e-01 -1.13780916e-01 3.05008888e-01 5.70551634e-01
2.12238356e-01 1.23437166e-01 3.31556112e-01 -4.60161567e-02
-2.59160809e-02 -2.51382172e-01 5.28298080e-01 1.05428076e+00
-5.16725898e-01 -1.04374039e+00 -6.34280324e-01 1.32935554e-01
-1.36051374e-02 3.59997869e-01 -3.15262139e-01 1.10223532e+00
2.08434656e-01 7.15050101e-01 4.01903957e-01 2.36513361e-01
6.57517433e-01 -2.32043207e-01 8.24179649e-01 -1.09224677e+00
1.83792168e-03 2.12643966e-01 3.42434555e-01 -5.84909916e-01
-4.64346260e-01 -6.89570367e-01 -1.38247323e+00 -1.80534884e-01
-3.86706859e-01 -2.17929527e-01 8.30601394e-01 1.09040844e+00
2.01298237e-01 2.97141641e-01 2.56601542e-01 -1.21169782e+00
-1.45657286e-01 -3.67228240e-01 -4.96345609e-01 5.57188809e-01
3.76281053e-01 -9.72975314e-01 -2.47155830e-01 4.75793257e-02] | [8.288081169128418, -2.799570083618164] |
9e9e1314-2751-4c0a-8ca7-582bf807f7d9 | recnet-early-attention-guided-feature | 2302.09409 | null | https://arxiv.org/abs/2302.09409v1 | https://arxiv.org/pdf/2302.09409v1.pdf | RecNet: Early Attention Guided Feature Recovery | Uncertainty in sensors results in corrupted input streams and hinders the performance of Deep Neural Networks (DNN), which focus on deducing information from data. However, for sensors with multiple input streams, the relevant information among the streams correlates and hence contains mutual information. This paper utilizes this opportunity to recover the perturbed information due to corrupted input streams. We propose RecNet, which estimates the information entropy at every element of the input feature to the network and interpolates the missing information in the input feature matrix. Finally, using the estimated information entropy and interpolated data, we introduce a novel guided replacement procedure to recover the complete information that is the input to the downstream DNN task. We evaluate the proposed algorithm on a sound event detection and localization application where audio streams from the microphone array are corrupted. We have recovered the performance drop due to the corrupted input stream and reduced the localization error with non-corrupted input streams. | ['Bashima Islam', 'Subrata Biswas'] | 2023-02-18 | null | null | null | null | ['sound-event-detection'] | ['audio'] | [ 4.30035979e-01 1.56174570e-01 5.29424429e-01 -3.15607637e-01
-8.61520529e-01 -3.45601857e-01 2.95081347e-01 1.81909978e-01
-6.17515802e-01 8.27883005e-01 6.72705054e-01 1.28009677e-01
-5.26725173e-01 -5.92966676e-01 -9.35671747e-01 -8.30145776e-01
-1.29407763e-01 -8.68149325e-02 -7.56936520e-02 4.78801101e-01
-1.33082882e-01 2.68784434e-01 -1.85494614e+00 3.02524447e-01
4.94826764e-01 1.35157490e+00 3.17876637e-01 8.66683960e-01
-4.40009832e-02 8.15212727e-01 -9.38847184e-01 1.07411489e-01
3.79716218e-01 -5.89898586e-01 2.91802865e-02 -4.24846143e-01
1.08379833e-01 -6.11164033e-01 -5.21825850e-01 1.22595870e+00
7.80109048e-01 3.05419207e-01 7.31185734e-01 -1.17765856e+00
1.17191508e-01 1.19306839e+00 -1.98841244e-02 1.67963862e-01
1.73895344e-01 -1.48445427e-01 7.25096524e-01 -1.01383770e+00
1.88859478e-01 1.13836384e+00 7.95425475e-01 4.44890767e-01
-7.22046971e-01 -7.96008229e-01 -2.68405765e-01 7.28878528e-02
-1.25199199e+00 -9.28608000e-01 8.42921734e-01 -3.58523190e-01
7.24462569e-01 1.26745805e-01 3.27714533e-01 8.85614634e-01
-8.78672767e-03 6.27804697e-01 4.82918441e-01 -2.79891789e-01
4.53441590e-01 8.94605368e-02 1.42677622e-02 6.64523244e-02
2.65665531e-01 4.17348087e-01 -1.16432488e+00 -1.00147817e-02
3.43599498e-01 5.73370419e-02 -4.30837065e-01 4.27073240e-01
-1.06000674e+00 1.42746717e-01 3.84538680e-01 4.38373595e-01
-6.81602180e-01 2.78322041e-01 1.65176407e-01 3.29241455e-01
3.85492325e-01 7.18328878e-02 -5.72303414e-01 -4.49370682e-01
-1.20633852e+00 -1.34491265e-01 1.03394163e+00 7.99956620e-01
6.67092681e-01 3.31794471e-01 -1.32631764e-01 6.98256791e-01
7.20334113e-01 8.09370637e-01 4.27785695e-01 -1.06244397e+00
7.96460509e-01 2.13552102e-01 -4.06305790e-02 -1.07059002e+00
-3.66534203e-01 -6.88186467e-01 -1.22492397e+00 9.58491415e-02
5.56986392e-01 -8.30623388e-01 -7.54335344e-01 1.89694250e+00
6.87007830e-02 5.48409939e-01 4.71577018e-01 9.05476928e-01
8.29322338e-01 5.85021794e-01 -3.37904066e-01 -2.81018019e-01
7.04759955e-01 -3.26355398e-01 -1.11361516e+00 -2.03558177e-01
-2.93452968e-03 -4.82478857e-01 2.65590698e-01 6.01712525e-01
-9.60783601e-01 -5.40728569e-01 -1.25896907e+00 3.40492368e-01
-3.07684392e-01 -1.55556258e-02 -8.46987739e-02 4.72580999e-01
-9.57420349e-01 9.93914306e-01 -7.61696339e-01 -3.19389417e-03
3.20020109e-01 5.18665433e-01 -3.39109004e-01 1.31564293e-05
-1.31229472e+00 3.78764391e-01 6.50657594e-01 4.88383293e-01
-9.42155600e-01 -8.49977076e-01 -6.38098836e-01 1.60425767e-01
-2.41060369e-03 -3.40617597e-01 1.12252164e+00 -7.44680643e-01
-1.36224186e+00 -3.13309789e-01 -3.67436379e-01 -6.77325547e-01
5.50823390e-01 -2.69962937e-01 -2.71973521e-01 -9.27300677e-02
-1.46349952e-01 4.94918495e-01 9.01888669e-01 -1.14203525e+00
-8.01103652e-01 -3.50879014e-01 -7.29706585e-01 1.31507069e-01
-3.16236079e-01 -5.06438017e-01 -1.88641921e-01 -5.07449627e-01
6.02643967e-01 -3.26375961e-01 -7.77210249e-03 -2.74406105e-01
-5.01700938e-01 2.41576791e-01 6.32581651e-01 -8.21333706e-01
1.22268260e+00 -2.59578514e+00 -6.22617230e-02 4.27087039e-01
9.25075933e-02 4.96267341e-02 -2.98293754e-02 3.97590131e-01
2.22475417e-02 -2.06242986e-02 -5.22141516e-01 -6.69815838e-01
8.29827338e-02 2.99831361e-01 -3.57972592e-01 4.22781229e-01
5.76150641e-02 3.73550236e-01 -8.77325475e-01 -1.90316826e-01
1.61190987e-01 7.51041889e-01 -4.59720433e-01 3.89993340e-01
-2.12434199e-04 6.75503612e-01 3.77461538e-02 3.41351002e-01
8.47610950e-01 4.18176353e-01 -3.52829754e-01 -3.43670636e-01
-1.33247999e-02 4.31242198e-01 -1.68018138e+00 1.70334876e+00
-4.11270171e-01 8.95047128e-01 2.65565336e-01 -5.70667803e-01
1.03896558e+00 5.44803619e-01 5.90500236e-01 -4.01971996e-01
3.66792351e-01 2.79214561e-01 1.09447852e-01 -4.24870998e-01
4.32657570e-01 8.36745277e-02 7.73385689e-02 3.53290379e-01
2.89102525e-01 2.24049196e-01 -2.90814102e-01 1.83552895e-02
1.38400066e+00 -2.00367436e-01 3.51082115e-03 1.83518663e-01
3.33695948e-01 -8.11183095e-01 6.61433935e-01 1.05002379e+00
-2.43146107e-01 8.48227680e-01 3.30609709e-01 1.06301524e-01
-9.54037368e-01 -1.09176850e+00 -6.21491112e-02 5.74653089e-01
-1.93317667e-01 -3.33379917e-02 -8.21094871e-01 -1.21228635e-01
7.53787830e-02 8.39585662e-01 -3.26839387e-01 -2.65953511e-01
-3.38515878e-01 -4.52524513e-01 9.19846833e-01 4.76315945e-01
6.49566472e-01 -9.06993449e-01 -5.71211755e-01 5.10881424e-01
-2.51918077e-01 -1.00893772e+00 -1.18347071e-01 7.58849621e-01
-8.65757823e-01 -7.22034633e-01 -3.61502647e-01 -2.59367019e-01
4.83693630e-01 -2.08822414e-01 5.50383151e-01 -4.15692627e-01
2.44429469e-01 2.37450689e-01 -1.71600342e-01 -7.00993419e-01
-4.45854515e-01 -2.80308276e-01 5.00952363e-01 2.98487395e-01
2.60329723e-01 -1.01222634e+00 -3.35094541e-01 -3.65989432e-02
-9.47405756e-01 -3.58234733e-01 3.90999079e-01 6.49861455e-01
2.87866145e-01 4.50537920e-01 8.72423232e-01 -3.42259914e-01
7.71808624e-01 -7.19759405e-01 -4.54586208e-01 -2.40494743e-01
-2.97041565e-01 3.85268331e-01 6.50517046e-01 -3.71678144e-01
-1.08923733e+00 3.44129533e-01 -3.00336480e-01 -4.89018321e-01
-3.23796719e-01 4.61583108e-01 -5.30108511e-01 4.53178227e-01
5.91105819e-01 1.04176447e-01 -2.41660699e-01 -7.49284685e-01
1.93923205e-01 1.04592180e+00 8.36356282e-01 -1.58971056e-01
5.01805782e-01 2.42255896e-01 8.18385780e-02 -8.12154770e-01
-5.19888222e-01 -3.74233872e-01 -5.47529817e-01 -3.59958827e-01
5.06537020e-01 -9.73020315e-01 -7.53824294e-01 7.66240239e-01
-1.62480247e+00 2.20194563e-01 -6.11296952e-01 1.06574261e+00
-1.91349119e-01 7.07032308e-02 -3.88812780e-01 -1.40611565e+00
-2.47115970e-01 -8.57749760e-01 8.56529593e-01 1.32378563e-01
-1.02306396e-01 -6.48441672e-01 -1.37042195e-01 -2.94708103e-01
3.98824871e-01 1.89451009e-01 3.83776784e-01 -9.95396435e-01
-6.96603179e-01 -3.85213852e-01 -7.53591061e-02 7.89783835e-01
3.21874350e-01 -1.72688961e-01 -1.62392128e+00 6.53068423e-02
3.96591216e-01 2.82656223e-01 1.01583183e+00 6.08326018e-01
6.56650603e-01 -5.54493666e-01 4.59989011e-02 6.96866095e-01
1.26706266e+00 3.75319391e-01 2.65081853e-01 -1.62431225e-01
6.10648751e-01 5.30208409e-01 2.03481510e-01 8.84760797e-01
-1.10919982e-01 9.55890398e-03 7.34143615e-01 3.80058080e-01
-2.90545255e-01 -4.89473015e-01 6.55109107e-01 1.37470472e+00
3.47193182e-01 -6.77335501e-01 -5.82744658e-01 7.44456768e-01
-1.75659275e+00 -7.31937051e-01 -1.32122010e-01 2.23385644e+00
8.11624467e-01 8.40945318e-02 -4.92679030e-01 6.86464548e-01
7.03771949e-01 3.82837793e-03 -8.97310853e-01 -2.08718423e-02
-1.47999376e-01 1.67680103e-02 7.00102508e-01 7.54831553e-01
-8.16167176e-01 3.42449903e-01 6.17188978e+00 5.00822842e-01
-8.49725306e-01 6.35271484e-04 2.24643707e-01 -2.57354915e-01
-2.58281052e-01 -4.46282834e-01 -7.37773538e-01 6.56615794e-01
1.43095684e+00 -1.11886933e-01 5.79734027e-01 4.63678896e-01
4.05203909e-01 -4.70653206e-01 -1.56809008e+00 1.14880157e+00
-6.73460513e-02 -8.04270685e-01 -1.82606205e-01 -8.95330980e-02
5.13973594e-01 2.33131096e-01 1.72249600e-01 -1.85692579e-01
2.74069011e-01 -8.77259552e-01 8.20685506e-01 1.07046759e+00
5.62347293e-01 -8.28954279e-01 1.00728691e+00 6.62287235e-01
-8.42266440e-01 -2.73152262e-01 -3.50088656e-01 -3.21051717e-01
1.21576183e-01 1.31945133e+00 -1.47967029e+00 3.03389221e-01
7.77585566e-01 7.25976169e-01 -8.80527720e-02 1.22971475e+00
-2.26952925e-01 8.87462020e-01 -8.46394718e-01 -6.79302290e-02
-1.43534884e-01 -1.10600263e-01 9.63531852e-01 1.10560679e+00
9.28225696e-01 -1.56457618e-01 -6.35828555e-01 9.78612661e-01
-3.74848813e-01 -3.76045406e-01 -8.49561095e-01 2.20328063e-01
8.57834697e-01 6.65198863e-01 -1.83536962e-01 -1.78055391e-01
2.46646116e-03 6.59480453e-01 -2.47053161e-01 5.72472930e-01
-5.06523490e-01 -4.71582502e-01 8.84199977e-01 -3.94055247e-01
2.66493291e-01 -4.74982373e-02 -5.17220020e-01 -6.72418654e-01
3.59411329e-01 -3.73856574e-01 -7.15773553e-02 -6.61901772e-01
-1.06526935e+00 4.76562828e-01 -1.49746791e-01 -1.36177242e+00
-7.82409489e-01 -2.84539521e-01 -2.30184346e-01 1.15173984e+00
-1.29285169e+00 -3.36213142e-01 -1.64409518e-01 5.47597349e-01
2.62776554e-01 -2.50929534e-01 6.48575425e-01 6.36848807e-01
-3.66914988e-01 5.89098752e-01 5.58157444e-01 2.75392830e-01
5.95340908e-01 -1.03884363e+00 2.21496001e-01 1.05660868e+00
6.02029078e-02 4.39422637e-01 8.83082211e-01 -5.34646511e-01
-1.27609074e+00 -1.21408367e+00 9.37509596e-01 -3.84321332e-01
5.33049524e-01 -3.85442525e-01 -7.23476410e-01 3.39099914e-01
2.04933211e-01 2.61960551e-02 6.12795353e-01 -5.45914173e-01
-5.00973612e-02 -6.02174103e-01 -1.36166883e+00 1.52481541e-01
8.57663035e-01 -8.01895499e-01 -6.09417975e-01 -1.64464429e-01
1.03287244e+00 -3.55327040e-01 -6.19399965e-01 3.16395700e-01
4.63263214e-01 -1.05500960e+00 6.07243478e-01 -1.16275810e-01
2.12940425e-01 -5.29548585e-01 -7.78092921e-01 -1.64433515e+00
1.50269836e-01 -6.74690187e-01 -1.44953921e-01 1.72329485e+00
6.55499160e-01 -4.66231048e-01 6.28424168e-01 7.37699032e-01
-1.88610494e-01 -2.18333732e-02 -1.38512647e+00 -4.83942449e-01
-2.67208576e-01 -1.14662755e+00 7.63168573e-01 3.46490383e-01
-3.41923833e-02 -3.60613922e-03 -5.41748106e-01 6.04798377e-01
8.60537708e-01 -7.60606825e-01 3.28752637e-01 -1.27503288e+00
-3.29638094e-01 2.37756558e-02 -2.98430443e-01 -1.02735710e+00
-1.54385671e-01 -5.99728286e-01 7.10186422e-01 -1.29383481e+00
-4.82557952e-01 -2.16434956e-01 -6.17127776e-01 1.89717233e-01
2.17312917e-01 -1.06861219e-01 3.70591789e-01 -1.05743483e-01
-2.89606601e-01 4.53024358e-01 6.87826097e-01 -1.37153625e-01
-2.75444776e-01 3.47284555e-01 -5.87619483e-01 7.88525760e-01
7.73688197e-01 -8.92578304e-01 -3.99246633e-01 -6.18669450e-01
3.48702937e-01 2.75213599e-01 3.09701294e-01 -1.74749792e+00
8.85339141e-01 4.25388455e-01 5.18760443e-01 -9.28723633e-01
4.46574241e-01 -1.49809933e+00 3.78630519e-01 2.02412352e-01
-5.39354026e-01 -3.10520083e-01 1.83816791e-01 6.48240268e-01
-4.54912424e-01 -3.94095033e-01 4.32183295e-01 4.24277745e-02
-1.83301181e-01 -5.47683276e-02 -4.78859037e-01 -1.21483095e-01
4.33557183e-01 -3.88365909e-02 8.06350112e-02 -7.17514396e-01
-8.02847505e-01 -7.93053582e-02 -3.20720315e-01 1.53221935e-01
9.12214875e-01 -1.42551255e+00 -6.58174872e-01 5.66425264e-01
-4.89092946e-01 3.62993151e-01 9.01404545e-02 4.62753952e-01
6.37521222e-02 4.34634000e-01 9.67653915e-02 -6.46555781e-01
-9.30538833e-01 1.29279658e-01 4.11050797e-01 3.11782390e-01
-3.52816619e-02 9.51822519e-01 -3.05317432e-01 -2.46586218e-01
9.40027297e-01 -8.94408345e-01 -1.86227053e-01 4.75977123e-01
5.91611922e-01 7.56099284e-01 3.24267417e-01 -5.93772531e-01
-3.26690823e-01 3.52838278e-01 4.74713534e-01 -6.75982773e-01
1.29229558e+00 -4.43065971e-01 -1.38640910e-01 9.46211636e-01
1.52928710e+00 -6.94202930e-02 -1.48939562e+00 -4.97197866e-01
-2.84177177e-02 -1.66097283e-01 3.33850563e-01 -8.45518053e-01
-1.25506032e+00 1.04407358e+00 9.48205769e-01 9.17656869e-02
1.33763051e+00 -3.35061133e-01 7.10398972e-01 5.99141717e-01
2.26826087e-01 -1.05235410e+00 -4.16983813e-01 7.55769730e-01
5.62412858e-01 -1.00678325e+00 -3.32305908e-01 3.70678872e-01
-1.68295026e-01 1.08218050e+00 2.16188312e-01 1.06953748e-01
1.16149366e+00 8.36701870e-01 1.29911587e-01 3.27768952e-01
-7.25647330e-01 1.33861629e-02 3.74838598e-02 6.82232261e-01
1.04177974e-01 -9.67655405e-02 1.82071626e-01 9.57636952e-01
-3.57810467e-01 3.58444043e-02 3.85288656e-01 5.11795700e-01
-5.65421462e-01 -6.76720679e-01 -7.40572751e-01 4.96140987e-01
-4.57410634e-01 -3.52198780e-01 -3.15881252e-01 1.23226769e-01
3.97535145e-01 1.43434310e+00 1.86639726e-01 -8.54758382e-01
5.30637681e-01 2.31323779e-01 -1.74762234e-01 -1.14630654e-01
-5.99492848e-01 1.27150133e-01 -4.00349572e-02 -4.64495540e-01
-3.99761915e-01 -7.97279954e-01 -1.55888963e+00 -1.52731612e-02
-1.64804205e-01 2.23811060e-01 1.17909789e+00 9.43908870e-01
3.69953573e-01 1.07095265e+00 7.51059413e-01 -9.88675296e-01
-5.47372758e-01 -1.19333196e+00 -7.64426529e-01 8.32871869e-02
1.12602520e+00 -1.14158928e-01 -9.27564859e-01 1.51294246e-01] | [15.089173316955566, 5.822480201721191] |
03642d6a-3ba3-484a-b9fe-cf86b1ea7f9a | physics-guided-problem-decomposition-for | 2202.05994 | null | https://arxiv.org/abs/2202.05994v2 | https://arxiv.org/pdf/2202.05994v2.pdf | Physics-Guided Problem Decomposition for Scaling Deep Learning of High-dimensional Eigen-Solvers: The Case of Schrödinger's Equation | Given their ability to effectively learn non-linear mappings and perform fast inference, deep neural networks (NNs) have been proposed as a viable alternative to traditional simulation-driven approaches for solving high-dimensional eigenvalue equations (HDEs), which are the foundation for many scientific applications. Unfortunately, for the learned models in these scientific applications to achieve generalization, a large, diverse, and preferably annotated dataset is typically needed and is computationally expensive to obtain. Furthermore, the learned models tend to be memory- and compute-intensive primarily due to the size of the output layer. While generalization, especially extrapolation, with scarce data has been attempted by imposing physical constraints in the form of physics loss, the problem of model scalability has remained. In this paper, we alleviate the compute bottleneck in the output layer by using physics knowledge to decompose the complex regression task of predicting the high-dimensional eigenvectors into multiple simpler sub-tasks, each of which are learned by a simple "expert" network. We call the resulting architecture of specialized experts Physics-Guided Mixture-of-Experts (PG-MoE). We demonstrate the efficacy of such physics-guided problem decomposition for the case of the Schr\"{o}dinger's Equation in Quantum Mechanics. Our proposed PG-MoE model predicts the ground-state solution, i.e., the eigenvector that corresponds to the smallest possible eigenvalue. The model is 150x smaller than the network trained to learn the complex task while being competitive in generalization. To improve the generalization of the PG-MoE, we also employ a physics-guided loss function based on variational energy, which by quantum mechanics principles is minimized iff the output is the ground-state solution. | ['Anish Arora', 'Viktor Podolskiy', 'Anuj Karpatne', 'Wei-Cheng Lee', 'Samuel Olin', 'Sangeeta Srivastava'] | 2022-02-12 | null | null | null | null | ['problem-decomposition'] | ['miscellaneous'] | [ 7.83923268e-02 7.37279505e-02 2.06659168e-01 -2.27448314e-01
-6.45440757e-01 -4.25526053e-01 3.20728093e-01 -1.26843438e-01
-5.20545542e-01 9.27603185e-01 -2.42989212e-01 -4.77693439e-01
-4.09453243e-01 -8.34865987e-01 -9.70647454e-01 -1.09355438e+00
1.93569168e-01 5.31396866e-01 -1.29576370e-01 -2.68740684e-01
3.01466167e-01 6.97710693e-01 -1.34072471e+00 -1.30538628e-01
1.29091895e+00 1.00872922e+00 3.05486411e-01 4.90104705e-01
-1.00774188e-02 1.67602822e-01 4.71362956e-02 -3.23360294e-01
4.26729709e-01 -5.49714923e-01 -7.63925314e-01 -4.15435493e-01
3.22913736e-01 1.50349988e-02 -6.93297386e-01 1.20620763e+00
5.08679330e-01 6.81651056e-01 8.81834269e-01 -1.03439617e+00
-5.31197011e-01 2.60388702e-01 -2.84594804e-01 1.04656540e-01
-9.06016827e-02 3.39078903e-01 1.10977864e+00 -8.72835398e-01
6.32906258e-01 8.85936022e-01 6.27741396e-01 8.40737343e-01
-1.62401092e+00 -6.39065027e-01 -1.90112740e-01 2.39603683e-01
-1.50436258e+00 -2.65241802e-01 7.62183845e-01 -4.58221346e-01
1.05850101e+00 -6.12657629e-02 4.24824238e-01 1.14135444e+00
5.76887608e-01 2.01136991e-01 1.02326906e+00 -4.43245411e-01
5.03426909e-01 2.72672027e-01 1.90675944e-01 7.77152061e-01
1.71999931e-01 1.86848372e-01 -5.68460226e-01 -2.68435597e-01
6.78647518e-01 -1.85738117e-01 -4.14677411e-01 -4.36041623e-01
-8.82690251e-01 8.06456804e-01 6.65076613e-01 -1.41816542e-01
-4.85328764e-01 2.31780693e-01 1.85324192e-01 2.60743350e-01
3.35370928e-01 6.79483652e-01 -5.01448989e-01 1.99294150e-01
-8.85655463e-01 5.21639526e-01 9.12837505e-01 5.02392769e-01
9.12426174e-01 3.39857787e-02 1.10836655e-01 3.96187633e-01
2.37229824e-01 3.68838608e-01 7.17207044e-02 -8.84511411e-01
3.47217441e-01 4.97535914e-01 2.69981503e-01 -6.63090289e-01
-5.50965130e-01 -7.71915376e-01 -1.36531019e+00 3.16048294e-01
3.61250401e-01 -3.05128068e-01 -9.84105527e-01 2.13967872e+00
4.24099028e-01 1.12275511e-01 1.34107977e-01 1.20146430e+00
5.79176605e-01 9.33918536e-01 2.08633795e-01 -1.92483395e-01
9.32191253e-01 -5.60117006e-01 -1.88752979e-01 -3.04057505e-02
6.29485905e-01 -3.60511422e-01 1.16709661e+00 3.25957865e-01
-1.02187932e+00 -4.34519976e-01 -1.00174248e+00 -2.12027341e-01
-2.85261661e-01 -1.23881316e-02 7.29047835e-01 3.66346270e-01
-9.82299805e-01 1.05334401e+00 -8.70198011e-01 -1.35565147e-01
2.84073949e-01 6.26580238e-01 -2.72524327e-01 2.72456735e-01
-1.27983582e+00 1.06467760e+00 4.58963990e-01 1.33379221e-01
-7.53770769e-01 -1.12679064e+00 -3.29086959e-01 3.40797812e-01
2.00051799e-01 -1.01884437e+00 7.96877503e-01 -6.31983519e-01
-1.73403132e+00 5.59105873e-01 -1.78807840e-01 -3.05733114e-01
3.04668128e-01 9.79981571e-02 -1.77920982e-01 -1.28615666e-02
-1.25997126e-01 4.96358722e-01 7.55102515e-01 -1.12432790e+00
-1.11629710e-01 -2.88252294e-01 2.16994032e-01 1.73541456e-02
-2.88590997e-01 -3.81460011e-01 -3.80719155e-02 -2.91218907e-01
3.36390078e-01 -1.25907338e+00 -4.70635116e-01 -1.05256408e-01
-4.00217205e-01 -2.94212639e-01 4.27130938e-01 -5.71648359e-01
9.79664922e-01 -1.99473202e+00 7.05014169e-01 3.52680862e-01
4.26535815e-01 1.24093980e-01 9.15845186e-02 5.53730667e-01
-2.52298772e-01 -1.16472822e-02 -3.10129672e-01 -1.26868159e-01
1.25006869e-01 3.64212207e-02 -4.63465869e-01 3.91681880e-01
9.75183100e-02 8.51044238e-01 -5.74678600e-01 -2.72762962e-02
-7.73024419e-03 5.63317657e-01 -8.36477697e-01 1.28413022e-01
-6.03693962e-01 9.58060205e-01 -5.54393113e-01 -2.83592157e-02
7.64480829e-01 -6.27447546e-01 6.42395318e-02 -3.67547184e-01
-1.39760539e-01 3.25964779e-01 -1.13827693e+00 1.86539340e+00
-6.25861824e-01 3.17967027e-01 2.37680003e-01 -1.42244577e+00
5.18675148e-01 2.96622783e-01 4.20221925e-01 -5.27107537e-01
1.01718254e-01 4.04052943e-01 3.94127727e-01 -3.82284373e-01
1.76849782e-01 -5.48131108e-01 -1.91201642e-02 3.41495097e-01
1.93767369e-01 -4.59027104e-02 4.80931029e-02 2.96190619e-01
1.01305509e+00 3.66135746e-01 -6.59874873e-03 -6.75502956e-01
6.92077577e-01 7.92309940e-02 5.77255428e-01 8.23366225e-01
1.33084550e-01 1.92169249e-01 4.04895008e-01 -5.45043945e-01
-1.31520283e+00 -1.16181254e+00 -3.84722799e-01 7.86971688e-01
2.07758948e-01 -3.16192657e-01 -7.77545154e-01 -1.97960764e-01
-7.83632100e-02 7.79679954e-01 -1.94319993e-01 -4.13713247e-01
-6.18360400e-01 -1.14662719e+00 2.97059834e-01 2.01917455e-01
4.36907351e-01 -7.92805731e-01 -3.90099913e-01 2.74836749e-01
3.90641578e-02 -9.77228820e-01 -9.94080454e-02 3.31682533e-01
-8.35908234e-01 -7.52094626e-01 -5.12168884e-01 -2.32039720e-01
6.45540535e-01 -2.11439118e-01 8.93912077e-01 -1.64634556e-01
-5.07103741e-01 1.39947906e-01 1.22178398e-01 -1.48991138e-01
-2.65608579e-01 2.72249579e-01 5.62057793e-01 -2.19531748e-02
1.97394922e-01 -1.06523228e+00 -8.68898988e-01 -6.63050637e-02
-7.03020155e-01 2.02663928e-01 7.52174497e-01 9.93567050e-01
5.89680016e-01 1.53333202e-01 6.71642542e-01 -6.62136495e-01
6.08016372e-01 -5.10022759e-01 -7.78364420e-01 2.14527458e-01
-6.07548058e-01 5.94807327e-01 1.19427621e+00 -3.61813724e-01
-1.11424804e+00 5.40012047e-02 -2.74504453e-01 -3.05475742e-01
-1.06736749e-01 5.57543695e-01 -8.86046290e-02 -6.54511392e-01
7.60434330e-01 2.90896595e-01 -2.71987140e-01 -4.67214942e-01
2.40365058e-01 4.19795960e-01 4.08543825e-01 -1.02168691e+00
9.37399685e-01 9.23153982e-02 7.09178388e-01 -8.64168465e-01
-1.14179075e+00 -1.40331134e-01 -4.75464851e-01 -2.18676087e-02
9.67501819e-01 -8.10745478e-01 -1.04658270e+00 1.77537799e-01
-1.21963930e+00 -1.33997738e-01 -2.15291679e-01 7.98398018e-01
-5.71431160e-01 2.52492994e-01 -6.04067326e-01 -8.84784997e-01
-4.64576513e-01 -1.17312503e+00 7.09552407e-01 3.86603773e-01
6.22020625e-02 -9.22997355e-01 2.16521937e-02 2.12705106e-01
5.08123279e-01 1.80461138e-01 1.54322934e+00 -2.88202196e-01
-8.05419564e-01 -1.05816603e-01 -4.05113101e-01 2.29194015e-01
-5.47462523e-01 -3.55031580e-01 -9.82336044e-01 -3.72464329e-01
2.59965777e-01 -5.24375081e-01 9.24972475e-01 3.85567993e-01
1.40971529e+00 2.87515800e-02 -2.88193196e-01 8.28675926e-01
1.65413284e+00 -1.08404495e-01 2.76125014e-01 -1.30404621e-01
7.46474445e-01 3.80432516e-01 -7.44850263e-02 1.99248195e-01
2.60922879e-01 6.02871895e-01 1.20482527e-01 1.07797593e-01
1.63811579e-01 -7.37868696e-02 1.67115808e-01 9.96308804e-01
-4.62653995e-01 -1.56985059e-01 -9.56037641e-01 -4.97372933e-02
-1.84863245e+00 -7.70236373e-01 -2.37957940e-01 2.27470255e+00
7.79750705e-01 5.98410890e-02 -2.07388520e-01 -2.29375005e-01
2.99415886e-01 -1.68331921e-01 -9.70108271e-01 -3.53742957e-01
1.37466371e-01 5.50969541e-01 3.78105372e-01 5.26559412e-01
-8.35401595e-01 7.91991293e-01 5.58126831e+00 9.57145214e-01
-1.17508233e+00 2.16349348e-01 3.20039630e-01 -6.31056428e-02
-1.88778281e-01 2.60766625e-01 -8.66892874e-01 4.17062849e-01
1.23497915e+00 -1.24075949e-01 7.89863169e-01 6.37541234e-01
3.38275790e-01 -1.05749652e-01 -1.19200659e+00 9.87866044e-01
-5.20983934e-01 -1.37090230e+00 3.77890207e-02 2.19663426e-01
8.31345856e-01 9.52304378e-02 2.79097613e-02 3.65624845e-01
-2.10449658e-02 -1.16825354e+00 3.71063709e-01 7.82040954e-01
8.08133781e-01 -7.00301230e-01 4.47996676e-01 7.22731888e-01
-8.64429295e-01 -5.36371022e-02 -7.16800988e-01 -1.87149629e-01
1.00329660e-01 8.26603472e-01 -1.19764239e-01 6.25443995e-01
5.50264060e-01 1.82372898e-01 5.30326590e-02 7.91951299e-01
5.60126379e-02 4.92729157e-01 -5.91408134e-01 -1.24039367e-01
3.74811769e-01 -8.19458067e-01 6.36178076e-01 8.40498388e-01
4.07849461e-01 4.55804467e-01 2.23245367e-01 1.53782034e+00
-3.09395790e-01 -7.65563995e-02 -3.40710282e-01 -1.92943558e-01
1.05431251e-01 1.32498586e+00 -4.70046014e-01 -1.03885964e-01
-2.80488133e-01 8.87655973e-01 6.32736206e-01 5.95466375e-01
-8.17253530e-01 -4.32190239e-01 6.11386895e-01 4.93694656e-02
5.32881655e-02 -4.47905064e-01 -3.39054495e-01 -1.37940049e+00
7.15924799e-02 -5.55630982e-01 -1.13473818e-01 -4.24579918e-01
-1.31916320e+00 5.16148984e-01 -8.76035020e-02 -8.47550929e-01
-9.35302377e-02 -9.29512501e-01 -7.09120870e-01 1.41578841e+00
-1.33929658e+00 -8.85229826e-01 -1.67995170e-01 4.30037647e-01
-5.73104098e-02 -5.78613132e-02 9.62896764e-01 3.57750952e-01
-8.11440349e-01 4.62863684e-01 4.81045783e-01 -2.97559559e-01
3.12209964e-01 -1.28607142e+00 6.49107695e-02 5.89851379e-01
-1.61523312e-01 8.77551675e-01 1.01288092e+00 -5.50756991e-01
-1.85062230e+00 -7.47146547e-01 6.78587735e-01 -3.07734340e-01
7.61437237e-01 -6.16539776e-01 -9.62295115e-01 2.66996384e-01
-2.33134285e-01 9.93266851e-02 6.01855695e-01 1.37810126e-01
-2.46944487e-01 -6.19980833e-03 -1.05762637e+00 4.70879763e-01
1.15133345e+00 -6.52473629e-01 -3.24156195e-01 6.34442091e-01
4.22784895e-01 -2.80304879e-01 -8.92414212e-01 3.08133483e-01
4.27033633e-01 -7.87045300e-01 1.03091061e+00 -9.20246124e-01
6.13039613e-01 -1.56783730e-01 4.91843186e-02 -1.29424465e+00
-3.39434832e-01 -5.83129346e-01 -1.98211089e-01 5.69855690e-01
5.09263873e-01 -5.65147519e-01 9.06556964e-01 8.57278943e-01
-1.28521457e-01 -9.45197642e-01 -1.05142093e+00 -7.14885712e-01
4.65984017e-01 -2.80832827e-01 1.20672502e-01 7.22935796e-01
-7.05818012e-02 5.90766251e-01 -3.68986964e-01 4.25994128e-01
8.95477653e-01 3.03261131e-01 4.98958349e-01 -1.36769688e+00
-6.50333107e-01 -3.85961473e-01 -2.49257445e-01 -1.07634592e+00
4.55172449e-01 -1.19581795e+00 -2.91688964e-02 -1.25741565e+00
2.49620602e-01 -4.59176928e-01 -2.54640043e-01 -2.18539368e-02
-1.35840639e-01 -8.44572112e-02 1.74105652e-02 3.06203485e-01
-2.81475455e-01 7.90384054e-01 1.21771717e+00 1.86825559e-01
-3.00844967e-01 -8.88186973e-03 -3.92130405e-01 5.66815317e-01
7.01654971e-01 -5.32471716e-01 -3.96290332e-01 -1.64725259e-01
5.71043432e-01 2.05513313e-01 6.57280564e-01 -1.03919721e+00
5.22996962e-01 -1.81918219e-01 4.62884992e-01 -2.51324445e-01
5.39126635e-01 -5.06043792e-01 2.42207021e-01 5.22109747e-01
-2.74949223e-01 -3.59243572e-01 4.71794568e-02 6.70714021e-01
7.37684965e-02 -2.92984188e-01 9.29784060e-01 -2.81645298e-01
-3.32632780e-01 5.38989484e-01 -1.36756495e-01 -4.79680486e-02
7.80531526e-01 1.87296420e-01 -1.22312910e-03 -1.45467907e-01
-8.18161249e-01 7.78915659e-02 3.39465111e-01 -2.62680888e-01
3.22757423e-01 -9.66654539e-01 -4.93796021e-01 9.54801738e-02
-2.47298613e-01 1.03721373e-01 5.67368627e-01 9.35059667e-01
-4.75731194e-01 4.79737163e-01 -2.32164711e-01 -3.29047710e-01
-3.99625450e-01 4.84705865e-01 5.59115648e-01 -3.75163108e-01
-7.94990778e-01 9.18204427e-01 3.41483980e-01 -6.76552832e-01
-1.82328802e-02 -9.76443384e-03 3.39467108e-01 -4.58690435e-01
1.53552309e-01 3.97117049e-01 3.09482869e-02 -4.00968939e-01
-1.98749602e-01 6.03385866e-01 -6.81172982e-02 2.96313688e-02
1.46326256e+00 8.92967656e-02 -3.71940047e-01 1.61505520e-01
1.18067133e+00 -2.67611444e-01 -1.19097722e+00 -2.14438275e-01
-6.23489134e-02 1.43492237e-01 3.84969890e-01 -7.17752099e-01
-7.74892092e-01 1.18521547e+00 5.04793644e-01 4.54773828e-02
8.83616686e-01 -1.66297853e-01 9.34883535e-01 8.86819780e-01
3.76559615e-01 -1.10441780e+00 -3.48036259e-01 5.33855319e-01
6.57436430e-01 -1.23081601e+00 -7.81351179e-02 -3.54616046e-01
-1.26619026e-01 1.19274664e+00 6.74936414e-01 -3.75581533e-01
7.91697383e-01 -6.83483901e-03 -4.67195123e-01 -2.69007355e-01
-5.65516531e-01 1.10079527e-01 5.50154328e-01 1.45617366e-01
3.52176487e-01 -6.40188670e-03 -3.50024611e-01 7.01859117e-01
-2.46497706e-01 -2.52720732e-02 2.14537531e-01 5.14453232e-01
-3.41533899e-01 -1.10257626e+00 2.18793489e-02 6.16369784e-01
-1.79234967e-01 -1.38982162e-01 -8.42573494e-02 5.24227679e-01
4.66596521e-02 5.33325613e-01 -2.86868155e-01 -2.70853966e-01
1.16698101e-01 3.93662840e-01 6.43712819e-01 -4.91189957e-01
-3.16756278e-01 -3.21764231e-01 -3.21109653e-01 -5.62499642e-01
-1.29400149e-01 -2.48503208e-01 -1.42386043e+00 -6.28802836e-01
-2.73309618e-01 3.13335240e-01 7.21738935e-01 1.28320575e+00
5.88243723e-01 4.19287115e-01 4.27996457e-01 -1.06457257e+00
-1.15150106e+00 -7.07228005e-01 -5.10596275e-01 3.72963071e-01
2.87132531e-01 -7.59585381e-01 -5.36622047e-01 -4.02520061e-01] | [5.468612194061279, 4.93288516998291] |
522ef1d6-8874-49b5-9291-923db6743dc9 | multivariate-time-series-imputation-with | null | null | http://papers.nips.cc/paper/7432-multivariate-time-series-imputation-with-generative-adversarial-networks | http://papers.nips.cc/paper/7432-multivariate-time-series-imputation-with-generative-adversarial-networks.pdf | Multivariate Time Series Imputation with Generative Adversarial Networks | Multivariate time series usually contain a large number of missing values, which hinders the application of advanced analysis methods on multivariate time series data. Conventional approaches to addressing the challenge of missing values, including mean/zero imputation, case deletion, and matrix factorization-based imputation, are all incapable of modeling the temporal dependencies and the nature of complex distribution in multivariate time series. In this paper, we treat the problem of missing value imputation as data generation. Inspired by the success of Generative Adversarial Networks (GAN) in image generation, we propose to learn the overall distribution of a multivariate time series dataset with GAN, which is further used to generate the missing values for each sample. Different from the image data, the time series data are usually incomplete due to the nature of data recording process. A modified Gate Recurrent Unit is employed in GAN to model the temporal irregularity of the incomplete time series. Experiments on two multivariate time series datasets show that the proposed model outperformed the baselines in terms of accuracy of imputation. Experimental results also showed that a simple model on the imputed data can achieve state-of-the-art results on the prediction tasks, demonstrating the benefits of our model in downstream applications. | ['Yonghong Luo', 'Yuan Xiaojie', 'Jun Xu', 'Ying Zhang', 'Xiangrui Cai'] | 2018-12-01 | null | null | null | neurips-2018-12 | ['multivariate-time-series-imputation'] | ['time-series'] | [ 4.28351015e-01 -2.86749214e-01 -1.98470745e-02 -3.45387101e-01
-9.01498318e-01 -4.82323080e-01 4.04586107e-01 -3.80474299e-01
1.16063818e-01 1.18799567e+00 4.09361959e-01 -1.03043333e-01
-7.59336129e-02 -7.40062237e-01 -1.03991210e+00 -9.38237906e-01
-6.94313571e-02 1.52541935e-01 -8.76445651e-01 -2.16532215e-01
-1.23506568e-01 1.15305439e-01 -1.30804217e+00 3.85114074e-01
9.42180872e-01 9.71136034e-01 -5.20051867e-02 5.23775518e-01
-1.15891993e-01 1.03056931e+00 -9.33219850e-01 -4.05740142e-01
1.35295212e-01 -8.71954560e-01 1.34046488e-02 -5.60733229e-02
-2.81281829e-01 -3.77693027e-01 -4.24308211e-01 5.43537259e-01
5.15922666e-01 -6.52338937e-02 7.08944380e-01 -1.51837873e+00
-6.26646101e-01 5.65235376e-01 -7.52845705e-01 5.34812920e-02
9.28953215e-02 2.74517871e-02 4.10226732e-01 -6.43443704e-01
3.18381697e-01 7.80609250e-01 1.02084684e+00 5.92708707e-01
-1.42522788e+00 -8.41219544e-01 -8.33685175e-02 -1.06110424e-02
-1.14308023e+00 -4.04429406e-01 1.15896809e+00 -5.07881045e-01
6.79779291e-01 2.66779184e-01 5.91783822e-01 1.67799580e+00
4.42513973e-01 6.35335267e-01 1.08282423e+00 -5.17120957e-02
2.61048108e-01 -6.41913891e-01 -4.42601025e-01 -2.03175455e-01
-1.98031038e-01 3.32543641e-01 -5.43914080e-01 -3.48887116e-01
8.93223345e-01 5.01407146e-01 1.83709934e-02 1.23580791e-01
-1.63765466e+00 6.86644912e-01 1.06474347e-01 1.00304862e-03
-7.70254612e-01 3.37378561e-01 4.44422036e-01 5.27551651e-01
7.57214010e-01 4.10277881e-02 -6.00664616e-01 -4.10439551e-01
-1.02615035e+00 2.26436496e-01 5.44140279e-01 9.94522333e-01
2.85272360e-01 6.23348892e-01 -4.95824248e-01 6.32792711e-01
-2.56699234e-01 7.18984723e-01 6.20922863e-01 -8.30563128e-01
8.89193594e-01 2.60094821e-01 3.03800970e-01 -8.81214917e-01
-2.13914409e-01 -5.84889293e-01 -1.79563665e+00 -1.21569067e-01
4.98376757e-01 -5.47304511e-01 -1.13137388e+00 2.10325646e+00
7.10899606e-02 7.52071500e-01 1.54529229e-01 6.46786273e-01
4.35083628e-01 6.93984985e-01 -3.37593369e-02 -6.87720418e-01
8.14261198e-01 -4.16065156e-01 -1.19310212e+00 1.38812914e-01
3.17430675e-01 -6.94485366e-01 5.71158588e-01 3.11609834e-01
-1.00573254e+00 -6.37784600e-01 -6.36237204e-01 1.59279034e-01
-5.09015471e-02 2.53447201e-02 7.27558255e-01 3.23466092e-01
-5.47640741e-01 5.70218444e-01 -1.09612072e+00 3.56323749e-01
5.52747488e-01 2.22974479e-01 -3.18893492e-01 5.00052944e-02
-1.21918738e+00 4.08648729e-01 2.52880435e-02 6.04078293e-01
-9.81507003e-01 -1.09210646e+00 -6.53182864e-01 -4.90876362e-02
-9.52911228e-02 -8.91245306e-01 8.35845709e-01 -1.13325989e+00
-1.01277864e+00 2.23222286e-01 -5.21094739e-01 -5.47981858e-01
7.04692185e-01 2.30504982e-02 -6.79349363e-01 -3.73241097e-01
1.15300611e-01 2.83910304e-01 1.39127707e+00 -8.97980332e-01
-2.92486399e-01 -4.56462324e-01 -5.97720683e-01 -2.32638508e-01
4.82422374e-02 -4.15587127e-01 7.82144219e-02 -1.42644644e+00
2.02298373e-01 -8.01065981e-01 -2.72773683e-01 -5.90446770e-01
-4.11756039e-01 1.94275185e-01 7.99289405e-01 -1.15427828e+00
1.17673147e+00 -2.18662977e+00 2.54208595e-01 5.93827255e-02
7.13857561e-02 -3.50299358e-01 -2.29028597e-01 7.96003401e-01
-4.91148323e-01 4.24380833e-03 -5.56620300e-01 -6.39434755e-01
-1.01362474e-01 4.51961070e-01 -9.45629179e-01 3.85391384e-01
2.99956143e-01 1.23406816e+00 -6.69213533e-01 -2.29539312e-02
9.60059911e-02 8.71529341e-01 -1.76578805e-01 4.76534188e-01
-2.77322620e-01 1.33545721e+00 -2.68218547e-01 7.42464662e-01
7.81371891e-01 -9.72374305e-02 -1.18447341e-01 -1.21174298e-01
2.41386190e-01 -1.20671004e-01 -8.33509862e-01 1.93544197e+00
-3.86629701e-01 3.09892625e-01 -3.44483614e-01 -9.85615492e-01
1.02881920e+00 6.53170586e-01 7.76238203e-01 -8.86354804e-01
-1.58007443e-01 4.85939421e-02 -2.09471490e-02 -4.46892112e-01
1.95254922e-01 -2.41289064e-01 -2.40401253e-01 5.21853745e-01
-2.59054333e-01 -3.68678831e-02 -8.91575739e-02 -1.22348204e-01
1.07758737e+00 4.58393037e-01 -1.50936306e-01 4.24648732e-01
5.53320758e-02 -1.19902365e-01 1.07626927e+00 6.45005405e-01
2.09079862e-01 1.06292677e+00 7.01418400e-01 -6.53322339e-01
-1.43960822e+00 -1.09277368e+00 -5.08335829e-02 5.36787450e-01
-4.36065912e-01 -1.68617904e-01 -5.71321070e-01 -4.17701870e-01
-1.10422499e-01 5.97702265e-01 -8.44635189e-01 -6.95845261e-02
-6.81107819e-01 -1.34074867e+00 4.86897618e-01 7.50098109e-01
2.19042718e-01 -1.21529198e+00 -2.90840060e-01 5.70772171e-01
-7.86269009e-01 -9.11379218e-01 -3.64021182e-01 1.72593877e-01
-1.28943992e+00 -6.43643200e-01 -8.11480463e-01 -3.62037212e-01
7.60958612e-01 -1.52958453e-01 1.12527990e+00 -2.09625259e-01
-6.90302998e-02 1.14479303e-01 -4.27460641e-01 -6.20708942e-01
-3.58308405e-01 -8.71111676e-02 -1.85853750e-01 4.87277240e-01
9.17382389e-02 -1.22006536e+00 -8.29226136e-01 5.23042940e-02
-1.21356690e+00 2.85173744e-01 4.70079541e-01 1.30853021e+00
9.05003309e-01 2.40583569e-01 1.24400973e+00 -8.08651447e-01
4.71926063e-01 -9.49207246e-01 -2.40446478e-01 -3.26777622e-02
-3.67658734e-01 -9.72413048e-02 9.49618578e-01 -6.50370359e-01
-1.04574692e+00 1.95941076e-01 -1.34777918e-01 -6.27682269e-01
-9.43308696e-02 7.52117634e-01 -1.13819331e-01 6.69973910e-01
1.48211926e-01 6.88934326e-01 2.05750823e-01 -5.46918631e-01
1.32177830e-01 3.75566423e-01 7.76355088e-01 -4.29578602e-01
7.19521642e-01 6.23023391e-01 1.56628951e-01 -3.77953082e-01
-6.57339513e-01 4.37299460e-02 -5.58092117e-01 2.40598153e-02
6.34899020e-01 -1.42573667e+00 -3.56973380e-01 9.09421504e-01
-1.20300543e+00 -2.92452663e-01 -5.36664546e-01 4.18429673e-01
-9.44840848e-01 -3.33045330e-03 -5.81783950e-01 -8.49397540e-01
-4.09402072e-01 -6.27200305e-01 1.11905384e+00 -2.23412216e-01
-2.51995921e-01 -9.77834284e-01 9.17048100e-03 3.05068731e-01
5.67521393e-01 1.06482744e+00 1.17820442e+00 -1.55025259e-01
-5.22079647e-01 -4.78712589e-01 2.73218393e-01 3.67155403e-01
3.34732503e-01 -3.50450009e-01 -1.01863158e+00 -1.77153215e-01
4.90752012e-01 -2.92017329e-02 7.58856595e-01 7.12667704e-01
1.73111308e+00 -3.83314729e-01 2.67876219e-03 9.86107707e-01
1.27487564e+00 2.03935236e-01 1.21145630e+00 -5.93393408e-02
8.55797648e-01 4.19421583e-01 4.44456458e-01 9.37893689e-01
4.33865339e-01 3.59934628e-01 7.55521119e-01 -2.49570772e-01
9.53628048e-02 -5.67361236e-01 3.22325408e-01 1.03493667e+00
-1.55368775e-01 -2.52412707e-01 -6.43145800e-01 8.44233513e-01
-2.07781363e+00 -1.35440648e+00 -4.10267115e-01 2.17999363e+00
8.51630867e-01 -2.79537082e-01 1.28038570e-01 3.09444845e-01
5.84966004e-01 1.50838897e-01 -9.19558823e-01 -1.54915424e-02
-5.34348547e-01 1.21001095e-01 1.93940848e-01 -1.77181870e-01
-9.03688669e-01 2.61760563e-01 6.27578449e+00 7.11048365e-01
-9.90901411e-01 5.67009710e-02 9.23792839e-01 -1.62121296e-01
-4.19202834e-01 -2.24030241e-01 -3.90900582e-01 1.14966321e+00
1.18012953e+00 -1.38878912e-01 7.18754947e-01 9.37924683e-02
5.99968135e-01 3.91974717e-01 -1.07304668e+00 1.02025211e+00
-1.48187533e-01 -1.16993546e+00 -3.72479074e-02 6.23840392e-02
1.07692039e+00 -1.34474963e-01 3.04251164e-01 2.80539811e-01
1.92271531e-01 -1.41609538e+00 5.11270344e-01 1.06984007e+00
8.28087449e-01 -8.18157971e-01 7.97431767e-01 6.47395730e-01
-8.42473865e-01 -1.05940148e-01 -1.76838487e-01 -3.02433938e-01
3.81719112e-01 1.02444124e+00 -4.32763040e-01 6.12380922e-01
6.02162361e-01 1.11379421e+00 -2.82260120e-01 7.73164213e-01
-1.16796307e-01 9.69109714e-01 -2.10877761e-01 7.53708601e-01
-2.84598857e-01 -4.14806366e-01 3.58728677e-01 6.35850191e-01
8.57360125e-01 -9.46966261e-02 -2.80832082e-01 9.32138681e-01
-7.60190189e-02 -4.13815677e-01 -6.94917500e-01 -1.82167381e-01
3.30484241e-01 8.85874391e-01 -2.44676098e-01 -1.04997016e-01
-5.25872231e-01 9.96528149e-01 2.47884281e-02 6.88815534e-01
-1.18524468e+00 -6.14226609e-02 4.43721384e-01 -3.40299048e-02
4.13367540e-01 -2.52581865e-01 -6.50425613e-01 -1.44918537e+00
5.27238011e-01 -1.10204017e+00 5.38069665e-01 -9.24215376e-01
-1.76996231e+00 5.49565792e-01 -3.15592825e-01 -1.79943335e+00
-8.26223075e-01 2.99728960e-01 -6.22649252e-01 1.16413939e+00
-1.37681282e+00 -1.50323391e+00 -3.08227360e-01 9.15639341e-01
3.66131812e-01 -2.09420487e-01 1.01585662e+00 2.27787927e-01
-4.06138301e-01 3.56466383e-01 6.32550716e-01 2.84983128e-01
5.18743157e-01 -1.03281009e+00 6.11005843e-01 9.59804475e-01
-1.40238479e-02 3.90407532e-01 5.72968841e-01 -8.11749220e-01
-1.77540791e+00 -1.47236514e+00 8.55681241e-01 -5.34343183e-01
5.83596647e-01 -4.80363697e-01 -1.01056039e+00 8.73181403e-01
1.74001947e-01 1.09613210e-01 8.58066380e-01 -2.08441138e-01
-2.48927981e-01 -3.24848682e-01 -1.15077984e+00 3.64414304e-01
8.50165725e-01 -3.92308265e-01 -4.56814528e-01 2.14651808e-01
7.32715011e-01 -4.37488586e-01 -1.11954391e+00 5.46598196e-01
5.56946993e-01 -6.19232535e-01 1.01673496e+00 -7.02956259e-01
9.55271482e-01 -3.12503219e-01 -2.41445258e-01 -1.57960093e+00
-2.03481779e-01 -7.61142492e-01 -6.30573630e-01 1.58497858e+00
1.93756297e-01 -5.16428411e-01 6.80817544e-01 6.56555593e-01
4.05971371e-02 -3.47587198e-01 -1.16423154e+00 -6.49239480e-01
2.03393370e-01 -5.58228731e-01 1.21237910e+00 9.92040813e-01
-5.76503396e-01 2.91771796e-02 -1.18999636e+00 8.80871266e-02
8.99842560e-01 4.54991102e-01 7.98767030e-01 -9.00729954e-01
-3.32659543e-01 1.37950778e-01 -2.09209159e-01 -5.50263524e-01
3.43885213e-01 -6.28953040e-01 -1.83778256e-01 -1.24242151e+00
-4.82241251e-02 -3.85061055e-01 -5.50281942e-01 4.35318112e-01
-1.35805994e-01 4.15380269e-01 -3.37499715e-02 3.60661030e-01
6.64133355e-02 9.19388294e-01 1.26425707e+00 -2.06024423e-01
-2.22936928e-01 2.63126582e-01 -5.98933041e-01 2.56687909e-01
8.02271903e-01 -6.67212009e-01 -5.75251400e-01 -7.63343632e-01
3.02712470e-01 8.90750229e-01 5.98950982e-01 -7.33797193e-01
3.09788827e-02 -2.77636498e-01 8.94749641e-01 -8.49627137e-01
4.13554072e-01 -1.13449800e+00 9.85590935e-01 9.27533582e-02
-1.83960944e-01 4.32863057e-01 2.03821510e-01 8.20535541e-01
-5.38812816e-01 3.93720925e-01 1.26820073e-01 5.24262451e-02
-3.86181921e-01 5.41692197e-01 -1.18678890e-01 -7.32270628e-03
9.12340283e-01 -5.90972975e-02 -2.06411019e-01 -6.66553199e-01
-9.22966599e-01 1.14371426e-01 1.27886340e-01 5.08095145e-01
7.14124203e-01 -1.65916705e+00 -1.05786061e+00 7.15997219e-01
-8.42287242e-02 1.28792152e-01 7.52742171e-01 1.06635642e+00
1.19356506e-01 -1.88797250e-01 -2.26978540e-01 -5.67304194e-01
-6.82645321e-01 6.42185867e-01 1.16856620e-01 -2.60948360e-01
-7.08845615e-01 9.19735804e-02 1.37243748e-01 -3.51651937e-01
-1.47953387e-02 -2.01259792e-01 6.48336858e-02 1.90383822e-01
5.44482291e-01 2.84318984e-01 2.04498190e-02 -3.89026016e-01
5.00092059e-02 3.55489045e-01 1.78653434e-01 7.28952289e-02
1.85652995e+00 -2.34639645e-01 -2.84154445e-01 7.81827033e-01
9.94795620e-01 -3.43735576e-01 -1.48302341e+00 8.72988403e-02
-4.18020070e-01 -4.60081756e-01 -2.85501152e-01 -1.00688624e+00
-1.22640347e+00 8.18492472e-01 4.08061206e-01 2.49030098e-01
1.71669066e+00 -5.92426896e-01 9.55126464e-01 -3.09366971e-01
4.14518297e-01 -5.71624398e-01 -4.23221529e-01 3.86537731e-01
1.11145401e+00 -1.26335371e+00 -3.65165949e-01 -1.02960743e-01
-6.53122663e-01 8.31187248e-01 2.11005330e-01 -1.33875936e-01
5.56488514e-01 5.84676623e-01 8.14572945e-02 3.33403707e-01
-1.08368194e+00 2.94678301e-01 3.42759527e-02 8.87576163e-01
4.25249487e-01 1.51998192e-01 -9.97157171e-02 9.42371845e-01
-3.62508923e-01 4.41770256e-01 5.92446208e-01 7.16266334e-01
6.39836907e-01 -1.17828059e+00 -4.99182969e-01 6.83736205e-01
-6.81784391e-01 -1.00653730e-01 1.35954097e-01 5.12125075e-01
-1.13115590e-02 9.55783546e-01 1.29695952e-01 -2.21995026e-01
1.95537478e-01 1.67360008e-01 1.18837513e-01 7.74709880e-02
-5.74308813e-01 3.36658895e-01 -3.15632045e-01 -4.22729552e-01
-4.78797495e-01 -7.78416038e-01 -9.41365659e-01 -4.39971626e-01
2.12025449e-01 8.46291799e-03 6.04216516e-01 7.23669708e-01
6.90246820e-01 7.03361809e-01 8.83162081e-01 -6.18810177e-01
-5.79557240e-01 -1.00944495e+00 -6.96061313e-01 9.07727063e-01
7.67652929e-01 -1.76833600e-01 -3.93707842e-01 5.73027432e-01] | [7.054409503936768, 3.2663919925689697] |
fb054a45-65a2-4abc-8a2a-52d1f0ad76e2 | deep-generative-quantile-copula-models-for | 1907.10697 | null | https://arxiv.org/abs/1907.10697v1 | https://arxiv.org/pdf/1907.10697v1.pdf | Deep Generative Quantile-Copula Models for Probabilistic Forecasting | We introduce a new category of multivariate conditional generative models and demonstrate its performance and versatility in probabilistic time series forecasting and simulation. Specifically, the output of quantile regression networks is expanded from a set of fixed quantiles to the whole Quantile Function by a univariate mapping from a latent uniform distribution to the target distribution. Then the multivariate case is solved by learning such quantile functions for each dimension's marginal distribution, followed by estimating a conditional Copula to associate these latent uniform random variables. The quantile functions and copula, together defining the joint predictive distribution, can be parameterized by a single implicit generative Deep Neural Network. | ['Ruofeng Wen', 'Kari Torkkola'] | 2019-07-24 | null | null | null | null | ['probabilistic-time-series-forecasting'] | ['time-series'] | [-4.24948454e-01 -4.77912910e-02 -3.46169740e-01 -6.54726982e-01
-1.00118434e+00 -7.73939788e-01 8.73350441e-01 -1.83936417e-01
2.04960123e-01 1.13538659e+00 1.43417209e-01 -3.83830607e-01
-3.87108207e-01 -1.19993055e+00 -8.63825500e-01 -9.74680662e-01
-5.36702573e-01 1.18235958e+00 -3.61084342e-01 8.97141919e-02
-1.30892515e-01 5.83799779e-01 -1.07300282e+00 -2.10062876e-01
8.86273861e-01 1.00302351e+00 -3.35292548e-01 9.06055272e-01
-1.68155640e-01 3.30521256e-01 -1.02780247e+00 -2.43106738e-01
-2.00642958e-01 -1.80916891e-01 -2.01393917e-01 -2.32142359e-01
-1.09455362e-01 -5.98715484e-01 2.86708046e-02 7.13017285e-01
2.03580216e-01 3.81473064e-01 1.53019357e+00 -1.58346844e+00
-1.07257605e+00 7.63064623e-01 -4.45670158e-01 4.76392992e-02
5.83054237e-02 -2.37716943e-01 9.94435489e-01 -5.58069468e-01
1.14241831e-01 1.42828143e+00 7.59830892e-01 -1.44450709e-01
-1.71894670e+00 -6.47946596e-01 -1.26349434e-01 -4.89957184e-01
-1.33240449e+00 4.45142388e-02 7.56887734e-01 -6.14504874e-01
8.80871296e-01 -2.27099717e-01 4.92935598e-01 1.24906433e+00
8.72068584e-01 5.30940235e-01 6.43451214e-01 2.69774944e-02
4.75300312e-01 -1.18131794e-01 -1.86528444e-01 -1.16137452e-02
6.78596273e-02 3.36168259e-01 -1.02174573e-03 -3.24990064e-01
8.64289463e-01 3.66228789e-01 2.80660510e-01 -3.67434800e-01
-6.66739821e-01 1.27345896e+00 2.57198304e-01 -1.44791648e-01
-4.42121834e-01 6.06693029e-01 1.29498422e-01 7.86371008e-02
8.69770885e-01 1.12527363e-01 -5.29614925e-01 5.48501760e-02
-1.15993822e+00 5.70252359e-01 1.09187019e+00 9.56819713e-01
7.61425853e-01 3.98077190e-01 -3.25140327e-01 4.26553011e-01
5.88099718e-01 1.19701135e+00 -6.98020160e-02 -6.03401363e-01
2.82174826e-01 -4.86121923e-02 5.83607852e-01 -5.88096797e-01
-3.80334765e-01 -7.32443631e-01 -9.64849830e-01 -4.26778384e-02
4.77620929e-01 -6.46964669e-01 -1.13213301e+00 1.89099455e+00
8.34551901e-02 2.82633036e-01 -1.12350374e-01 4.43033963e-01
2.23679304e-01 1.18225729e+00 5.07500112e-01 -1.16930887e-01
8.07437181e-01 -3.00017387e-01 -4.56960976e-01 1.45683959e-01
-1.42696157e-01 -5.14190435e-01 4.89781827e-01 2.28531677e-02
-1.01420820e+00 -5.85327268e-01 -6.54955387e-01 1.92606792e-01
-5.62293470e-01 -8.34511593e-03 6.40381694e-01 6.21085823e-01
-1.10012865e+00 6.58507168e-01 -1.15991354e+00 2.40585878e-01
3.03782016e-01 2.03588828e-01 5.76973706e-02 3.23845267e-01
-1.50030100e+00 6.92718983e-01 2.79004306e-01 -8.09746608e-03
-1.33384645e+00 -9.47911561e-01 -8.16089571e-01 5.90907514e-01
-5.30305684e-01 -1.01287913e+00 1.23055983e+00 -6.59307957e-01
-1.56176937e+00 1.33069515e-01 -1.53288305e-01 -3.95024419e-01
1.66573152e-01 -2.76464432e-01 -4.16144580e-01 -2.32162192e-01
2.64330506e-01 2.37442151e-01 1.19586277e+00 -1.02314413e+00
-3.55339497e-01 -1.93295062e-01 -2.26200968e-01 -2.64834136e-01
1.41509458e-01 -2.78285891e-01 4.99717072e-02 -5.63913882e-01
-1.00079179e-01 -4.70790505e-01 -3.13682437e-01 -6.89175546e-01
-6.68943107e-01 -4.21073735e-01 2.51745194e-01 -7.37176061e-01
1.21680832e+00 -1.91623986e+00 2.00348824e-01 5.38317442e-01
-2.34731082e-02 -7.31913745e-01 -2.73022577e-02 8.06114137e-01
-3.22228193e-01 7.13755935e-02 -2.61657864e-01 -5.57400346e-01
4.34415996e-01 2.48176724e-01 -1.10771978e+00 5.68365216e-01
6.18413091e-01 1.02597487e+00 -9.57092464e-01 2.98546217e-02
8.84131044e-02 8.57354581e-01 -3.95792902e-01 4.32268202e-01
-5.74051917e-01 4.88644481e-01 -4.90375668e-01 5.87904274e-01
9.10497546e-01 -2.35288635e-01 -3.86809907e-03 2.31226519e-01
1.93597167e-03 2.47161463e-01 -8.12119424e-01 9.95706975e-01
-3.10036570e-01 4.09679741e-01 -4.90473121e-01 -1.07718742e+00
1.28905380e+00 2.94809550e-01 5.28316379e-01 -1.14188209e-01
1.05036050e-01 2.08611470e-02 -4.86576676e-01 -3.12221069e-02
5.03605485e-01 -6.27914667e-01 -5.40738285e-01 5.40261507e-01
5.17073750e-01 -2.42580563e-01 2.19474453e-02 -6.57017604e-02
2.85280377e-01 5.33332705e-01 8.88010040e-02 -2.66521662e-01
-1.09298259e-01 -3.13027471e-01 1.41454905e-01 7.32345819e-01
2.60232836e-01 5.29453814e-01 1.13858652e+00 -2.61109948e-01
-1.34078908e+00 -2.21411395e+00 -6.70969427e-01 1.27423787e+00
-4.30829436e-01 5.78157827e-02 -4.30449456e-01 -1.28478914e-01
5.41187942e-01 1.15987563e+00 -1.00607598e+00 -1.09266624e-01
-1.64161563e-01 -1.11664712e+00 4.77180481e-01 9.52567637e-01
-1.42995551e-01 -8.31013262e-01 -1.71061337e-01 4.04689312e-01
1.85221523e-01 -4.36365396e-01 -2.55867064e-01 6.87404573e-01
-8.92916143e-01 -5.62445164e-01 -9.41236198e-01 -3.22966099e-01
3.03847522e-01 -6.48347557e-01 1.46572447e+00 -8.57858539e-01
1.44702569e-01 3.70731622e-01 2.22265586e-01 -3.87525469e-01
-3.37935776e-01 4.92016226e-02 -6.38229027e-02 -1.01942912e-01
1.46565422e-01 -1.14774156e+00 -5.73908687e-01 -9.41031519e-03
-8.88540864e-01 -4.39384788e-01 1.59485146e-01 7.42294550e-01
7.15916574e-01 -1.41503304e-01 9.87218618e-01 -4.80206013e-01
8.80361319e-01 -1.26871121e+00 -9.77398992e-01 1.19085975e-01
-3.13617617e-01 2.94191808e-01 6.46960557e-01 -6.65047765e-01
-1.00386858e+00 -2.84772247e-01 5.14834449e-02 -3.76409978e-01
-1.50733486e-01 8.92944932e-01 -1.14315450e-01 9.25127089e-01
3.58265340e-01 1.66600078e-01 -1.61723286e-01 -4.58406657e-01
8.02902639e-01 2.21338004e-01 9.91106987e-01 -8.46302390e-01
7.50436485e-01 3.15290123e-01 3.86869937e-01 -3.44957680e-01
-8.95337939e-01 -6.27804846e-02 -7.84633458e-01 -9.25020650e-02
1.03349316e+00 -9.00039196e-01 -7.73931205e-01 4.01502788e-01
-1.22089911e+00 -2.53687799e-01 -5.21569729e-01 5.11077940e-01
-7.35412538e-01 -2.69528091e-01 -5.55389225e-01 -1.15343022e+00
-2.02019945e-01 -6.63509727e-01 1.46175325e+00 5.18318594e-01
-1.12955689e-01 -1.65093195e+00 6.18459880e-01 -7.61421740e-01
6.60540938e-01 6.94633305e-01 1.20658708e+00 -8.20522010e-01
-3.72975707e-01 -5.81486225e-01 -1.28621832e-01 4.19456154e-01
-1.22303246e-02 4.62391913e-01 -6.75022483e-01 -1.76720560e-01
-2.62778372e-01 1.00506678e-01 8.05047393e-01 1.11399734e+00
9.24825788e-01 -2.15449426e-02 -4.96415079e-01 9.47654068e-01
1.26892757e+00 2.59834319e-01 4.13384169e-01 8.71061161e-03
2.00931400e-01 -3.80119123e-03 -1.04907807e-02 7.58780122e-01
4.76491541e-01 -1.45868659e-01 5.12348890e-01 1.93129867e-01
5.67599416e-01 -6.28254950e-01 4.22049195e-01 3.63159478e-01
-9.76168290e-02 -5.45375764e-01 -8.88963044e-01 5.72660148e-01
-1.53943539e+00 -1.22522116e+00 8.67246464e-02 2.23778105e+00
6.94148660e-01 8.25857744e-02 2.61570454e-01 -4.50800091e-01
7.84021914e-01 9.04986560e-02 -7.78814197e-01 -4.58627254e-01
-2.38354385e-01 6.17295027e-01 3.17763478e-01 4.76032734e-01
-1.31905484e+00 6.63478076e-01 8.42288971e+00 5.80577433e-01
-1.03690684e+00 -1.39594868e-01 9.07313526e-01 6.30263835e-02
-7.33543754e-01 1.02772631e-01 -9.48558092e-01 5.88966906e-01
1.46774662e+00 -4.67025638e-01 3.50889266e-01 9.42946494e-01
-3.02132536e-02 5.09207360e-02 -1.13658798e+00 3.86218607e-01
-4.02712017e-01 -1.08378994e+00 4.35491614e-02 1.05275949e-02
1.00531304e+00 8.53919089e-02 6.35560811e-01 6.72413170e-01
9.32440341e-01 -1.56092525e+00 5.64754188e-01 1.32146275e+00
9.02604759e-01 -1.16814041e+00 6.17185295e-01 2.66327441e-01
-9.32641149e-01 8.11450407e-02 -6.79076672e-01 2.38937549e-02
4.66084510e-01 7.78409898e-01 -8.06550086e-01 3.07342172e-01
5.18500268e-01 5.35729587e-01 -1.09324850e-01 1.02053475e+00
-2.32019529e-01 9.98785019e-01 -5.08626878e-01 6.28231168e-02
1.10509671e-01 -5.86377144e-01 4.15052116e-01 1.26782131e+00
7.72990286e-01 -3.61925453e-01 -1.11012058e-02 1.41333795e+00
6.95862621e-03 -3.42909247e-01 -2.44840458e-01 -1.51995942e-01
5.14166713e-01 1.05933094e+00 -5.43429971e-01 -4.88204986e-01
3.68891619e-02 4.93414909e-01 2.71982133e-01 1.08425260e+00
-1.28290892e+00 -3.84917527e-01 7.34539509e-01 -1.33285806e-01
5.22708356e-01 -2.90892333e-01 -3.11403602e-01 -1.08358324e+00
-4.55666989e-01 5.39182425e-02 4.82692301e-01 -9.84022319e-01
-1.77013373e+00 5.05363882e-01 4.94282782e-01 -9.98214722e-01
-1.09636521e+00 -5.34559846e-01 -1.07743835e+00 1.72639000e+00
-1.32246828e+00 -1.11255085e+00 2.73768902e-01 6.63453519e-01
-2.21496686e-01 -2.06849024e-01 1.00625873e+00 -1.25468284e-01
-2.61789143e-01 4.03721631e-01 8.12381148e-01 -8.24862998e-03
5.57245135e-01 -1.54590201e+00 5.24049103e-01 3.72667044e-01
-3.86497498e-01 7.12748230e-01 9.57888484e-01 -7.30200291e-01
-1.02594483e+00 -1.11233246e+00 7.24317491e-01 -6.36211693e-01
1.16827142e+00 -3.31257761e-01 -8.42422128e-01 1.07745492e+00
1.92111358e-01 -1.86437279e-01 6.96246624e-01 2.61312515e-01
-5.03334641e-01 7.62169901e-03 -1.07495284e+00 8.86919498e-02
8.95731598e-02 -5.31891108e-01 -4.72193986e-01 4.19045597e-01
8.00324738e-01 -3.94152522e-01 -1.26387155e+00 1.60111189e-01
7.99951613e-01 -8.27166021e-01 1.19142187e+00 -1.19948602e+00
8.43315721e-01 -1.75447762e-01 -2.88383931e-01 -1.53588128e+00
-2.83013493e-01 -4.53645229e-01 -2.66947478e-01 1.25851882e+00
4.41257924e-01 -8.44497502e-01 3.43343049e-01 4.80732143e-01
1.09101697e-04 -6.56466305e-01 -1.16432154e+00 -5.75974524e-01
7.40381002e-01 -7.35885978e-01 1.25363755e+00 5.30587196e-01
-4.66663897e-01 5.95548451e-02 -3.79577130e-01 5.09773493e-01
7.72055030e-01 6.71270669e-01 5.39456785e-01 -1.19399381e+00
-4.81715322e-01 -6.21707380e-01 -1.43429428e-01 -1.11503410e+00
1.44772351e-01 -8.63562107e-01 9.85689908e-02 -1.41258943e+00
1.12588383e-01 -1.80474535e-01 -3.75123054e-01 3.57172862e-02
-1.24031223e-01 -1.29733190e-01 -3.23944986e-01 3.04022562e-02
-1.74953938e-01 7.95099497e-01 1.20629787e+00 1.12031288e-04
-1.75640389e-01 5.42469561e-01 -4.76311177e-01 3.58294576e-01
7.97366023e-01 -4.61423635e-01 -2.75243193e-01 -1.03084765e-01
3.44957560e-01 8.17303181e-01 5.03183007e-01 -4.87473130e-01
2.43362002e-02 -4.80714381e-01 1.06134200e+00 -1.04863930e+00
1.96264118e-01 -3.21173340e-01 4.49584186e-01 -8.75100866e-02
-2.15383768e-01 1.63243756e-01 2.06800252e-01 5.61110079e-01
-3.39929044e-01 6.77863657e-02 4.99997616e-01 4.78205681e-01
-9.34278313e-03 4.40812141e-01 -4.60257798e-01 -4.34471352e-04
7.02309310e-01 1.36348546e-01 -2.18098521e-01 -8.07024181e-01
-1.26792419e+00 3.15814644e-01 -1.29062951e-01 1.20682076e-01
4.00211275e-01 -1.41597545e+00 -7.80190468e-01 3.68134290e-01
-5.01390457e-01 1.49082333e-01 2.62976259e-01 5.42509675e-01
-8.66460577e-02 2.82065302e-01 -2.15607837e-01 -8.00146043e-01
1.69051453e-01 5.42184472e-01 5.98119855e-01 -4.23583806e-01
-1.34398893e-01 6.75064564e-01 2.24231943e-01 -5.75750053e-01
-1.50061265e-01 -5.33070385e-01 -1.74765542e-01 1.72938153e-01
1.46603495e-01 2.72510290e-01 -5.57634294e-01 -4.00139540e-01
-1.14133656e-01 2.83808559e-01 6.40543699e-01 -1.87888399e-01
1.68673062e+00 -5.19089513e-02 -3.52713853e-01 7.42979705e-01
1.28084362e+00 -1.04554154e-01 -1.64220333e+00 1.48892477e-01
-9.14433599e-02 9.99062881e-02 -3.80367786e-01 -8.98904085e-01
-9.08448935e-01 9.52543914e-01 6.97836801e-02 6.92915916e-01
9.95615244e-01 3.54685128e-01 3.43740344e-01 -1.64268762e-01
-9.87720266e-02 -4.58009660e-01 -3.67714673e-01 8.45042050e-01
9.12743032e-01 -8.41814756e-01 -2.27931455e-01 3.73611391e-01
-3.90490383e-01 1.53183472e+00 2.91912723e-02 -7.56518185e-01
1.22676396e+00 6.11158907e-01 -2.24132538e-01 7.31703863e-02
-7.57345319e-01 7.83733428e-02 6.79600656e-01 8.55182827e-01
6.05200589e-01 6.04257584e-01 3.03724468e-01 1.25556564e+00
-7.09567666e-01 -2.33787954e-01 1.45539820e-01 2.04800427e-01
-2.63349205e-01 -8.21188271e-01 -4.73494619e-01 4.83723760e-01
-2.29762971e-01 -1.39521509e-01 5.08672953e-01 9.73448336e-01
-2.26796657e-01 5.95884502e-01 7.61309505e-01 5.78158535e-02
-5.03852591e-02 5.02827287e-01 6.45095259e-02 -3.00837576e-01
-9.18820277e-02 5.38242877e-01 -5.28147995e-01 -1.39574841e-01
1.28584340e-01 -9.22527254e-01 -1.05764127e+00 -3.94077092e-01
6.03923090e-02 2.07672447e-01 6.84011698e-01 8.83487999e-01
2.85997391e-02 7.09296942e-01 6.73245907e-01 -1.07366824e+00
-9.12696302e-01 -1.07885039e+00 -9.15513873e-01 -1.42604746e-02
5.61098874e-01 -5.25902510e-01 -5.33234000e-01 7.59603754e-02] | [7.170009613037109, 3.7383525371551514] |
3e40d991-a516-4349-99da-c1ceda0d4a43 | bio-inspired-spike-based-hippocampus-and | 2305.12892 | null | https://arxiv.org/abs/2305.12892v1 | https://arxiv.org/pdf/2305.12892v1.pdf | Bio-inspired spike-based Hippocampus and Posterior Parietal Cortex models for robot navigation and environment pseudo-mapping | The brain has a great capacity for computation and efficient resolution of complex problems, far surpassing modern computers. Neuromorphic engineering seeks to mimic the basic principles of the brain to develop systems capable of achieving such capabilities. In the neuromorphic field, navigation systems are of great interest due to their potential applicability to robotics, although these systems are still a challenge to be solved. This work proposes a spike-based robotic navigation and environment pseudomapping system formed by a bio-inspired hippocampal memory model connected to a Posterior Parietal Cortex model. The hippocampus is in charge of maintaining a representation of an environment state map, and the PPC is in charge of local decision-making. This system was implemented on the SpiNNaker hardware platform using Spiking Neural Networks. A set of real-time experiments was applied to demonstrate the correct functioning of the system in virtual and physical environments on a robotic platform. The system is able to navigate through the environment to reach a goal position starting from an initial position, avoiding obstacles and mapping the environment. To the best of the authors knowledge, this is the first implementation of an environment pseudo-mapping system with dynamic learning based on a bio-inspired hippocampal memory. | ['Fernando Perez-Pena', 'Gabriel Jimenez-Moreno', 'Angel Jimenez-Fernandez', 'Juan P. Dominguez-Morales', 'Alvaro Ayuso-Martinez', 'Daniel Casanueva-Morato'] | 2023-05-22 | null | null | null | null | ['robot-navigation'] | ['robots'] | [ 2.91398279e-02 -1.15827452e-02 6.08565271e-01 7.99859315e-02
6.47619188e-01 -3.48184973e-01 6.68693721e-01 -7.25153387e-02
-6.68838739e-01 5.86895704e-01 -5.19695640e-01 1.10624842e-01
-1.83354869e-01 -1.00029039e+00 -7.84602821e-01 -8.67824435e-01
-3.46996844e-01 4.34913754e-01 8.48813713e-01 -5.43972135e-01
6.36036575e-01 6.66335285e-01 -2.14804649e+00 8.76049101e-02
4.18791980e-01 9.06762242e-01 1.03842533e+00 3.97914171e-01
-2.22757682e-02 5.43912351e-01 -2.48760998e-01 5.15216708e-01
3.57670307e-01 -5.03721416e-01 -3.21850330e-01 -3.79596591e-01
-2.19774202e-01 2.03617498e-01 -3.56425822e-01 8.66092205e-01
3.72038841e-01 6.28950298e-02 4.21015322e-01 -1.04338622e+00
-9.00152922e-02 2.05956206e-01 5.81666470e-01 1.09215036e-01
2.01851442e-01 1.18259512e-01 1.27850205e-01 -7.72934079e-01
9.66275215e-01 7.52363920e-01 4.57804501e-01 6.65799797e-01
-1.21416438e+00 -4.36767131e-01 -3.98469567e-01 3.55368972e-01
-1.54234838e+00 -4.07583892e-01 3.54098767e-01 -3.82130057e-01
1.31195271e+00 -2.80233622e-01 1.26061833e+00 7.93696046e-01
1.20594573e+00 -1.21681109e-01 1.43750858e+00 -4.66295987e-01
1.14498270e+00 9.40616652e-02 7.43589029e-02 4.57660645e-01
5.13835847e-01 3.76020789e-01 -8.55476201e-01 1.80231154e-01
9.11963701e-01 -1.85051143e-01 -1.84699044e-01 -8.75770390e-01
-1.08110988e+00 2.23301023e-01 6.82959139e-01 8.57378602e-01
-5.58267415e-01 3.47664177e-01 -2.88401954e-02 2.05617577e-01
-7.36829758e-01 4.38767135e-01 -5.70921553e-03 -1.75137430e-01
-6.28446937e-01 2.34311730e-01 1.24966276e+00 4.97365236e-01
6.63115680e-01 2.65407890e-01 4.01164949e-01 4.68738467e-01
4.59747761e-01 6.46446407e-01 9.55217540e-01 -9.71177936e-01
-3.47666413e-01 8.12296867e-01 -1.40486181e-01 -8.12409341e-01
-5.46960533e-01 -4.86390859e-01 -6.22467041e-01 9.86467361e-01
2.23070547e-01 3.30918908e-01 -7.54618168e-01 1.57017195e+00
-6.04512580e-02 3.60990316e-02 2.72300422e-01 8.87756228e-01
1.75356045e-01 6.25709713e-01 -1.44966856e-01 1.43073559e-01
1.14021313e+00 -4.41507459e-01 -4.19237733e-01 -6.24259770e-01
2.23946586e-01 6.82902113e-02 2.80141681e-01 4.48134214e-01
-7.93425858e-01 -3.94782424e-01 -1.52763689e+00 3.81631047e-01
-1.01584673e+00 -2.74165243e-01 2.25179866e-01 4.65017647e-01
-1.55674589e+00 7.68251419e-01 -9.61661577e-01 -9.38418567e-01
4.69770841e-02 5.52568793e-01 -6.87820315e-01 2.36106321e-01
-1.07002068e+00 1.43422568e+00 7.36091435e-01 8.52140933e-02
-8.55616868e-01 1.03923371e-02 -5.18463671e-01 1.89294130e-01
-3.55278850e-01 -6.99701786e-01 7.97170818e-01 -5.52822709e-01
-1.79453456e+00 8.11848342e-01 8.97789225e-02 -8.95896733e-01
2.68787965e-02 3.86074334e-01 -2.11281896e-01 3.72841768e-02
-1.39434963e-01 7.29787230e-01 3.69099438e-01 -9.87654269e-01
-3.14714700e-01 -6.00465655e-01 -4.04294163e-01 4.09688149e-03
-2.37589329e-01 -4.83969748e-01 -1.85909569e-02 -1.34131238e-01
5.23661375e-01 -1.21863818e+00 -1.96463987e-01 5.92871793e-02
3.81530255e-01 5.59032559e-01 6.75380409e-01 -1.98872000e-01
7.25894034e-01 -2.04374504e+00 2.76430994e-01 2.58318186e-01
-2.46748537e-01 4.51794088e-01 -4.94844466e-03 7.21886396e-01
1.84584484e-01 -4.36836511e-01 -5.34766257e-01 2.45713547e-01
-8.97964984e-02 4.62340206e-01 -2.92564392e-01 3.06225151e-01
-3.18083733e-01 7.22731829e-01 -5.66060007e-01 5.04482239e-02
5.99542819e-02 6.80979967e-01 -2.61835814e-01 -1.08121365e-01
-1.92420587e-01 3.58579189e-01 -1.69758245e-01 2.12797433e-01
3.77924114e-01 2.69844770e-01 3.64552081e-01 4.77873355e-01
-7.71146774e-01 3.97302471e-02 -1.19080329e+00 1.89584148e+00
-3.65095586e-01 6.76310718e-01 4.36657369e-01 -9.61225033e-01
1.42594647e+00 1.38677895e-01 1.31186798e-01 -1.45156288e+00
4.31765139e-01 7.32096136e-01 4.34767991e-01 -1.22010320e-01
2.25783065e-01 9.57622775e-04 1.63564727e-01 4.09281075e-01
2.00593993e-01 -3.02460462e-01 2.27086082e-01 -1.54197142e-01
1.55402577e+00 3.29073995e-01 2.11593136e-01 -7.19532609e-01
5.50775588e-01 1.04820743e-01 3.10624272e-01 5.66980958e-01
-2.10844859e-01 2.90995032e-01 4.12221104e-02 -4.43625540e-01
-8.72084498e-01 -1.35013402e+00 -2.57584870e-01 4.08297688e-01
5.11769831e-01 1.24064813e-04 -1.02749181e+00 5.78653455e-01
-4.10188697e-02 6.31945193e-01 -6.20203614e-01 -3.91382664e-01
-5.07225215e-01 -3.86376590e-01 4.02458459e-01 6.08062372e-02
7.98898160e-01 -1.60744035e+00 -1.84813964e+00 7.44499803e-01
4.20161337e-01 -7.42328465e-01 6.54441655e-01 8.27366054e-01
-1.05638134e+00 -9.22414362e-01 -2.76686877e-01 -1.05195165e+00
4.84367788e-01 1.28857866e-01 5.34834385e-01 -8.75380039e-02
-6.20642126e-01 3.80544841e-01 -5.60563095e-02 -3.09779227e-01
-3.69053602e-01 2.11969335e-02 2.16969520e-01 -2.94887185e-01
2.20947996e-01 -1.12492514e+00 -5.56599617e-01 3.79185587e-01
-1.10623944e+00 2.69916952e-02 6.84320211e-01 7.46193826e-01
7.67764866e-01 4.34289463e-02 4.36016321e-01 -1.15395747e-01
5.46591699e-01 -5.38488090e-01 -7.51901686e-01 2.49628269e-04
-6.92746818e-01 2.39647552e-01 5.67366600e-01 -1.39479786e-01
-7.49228835e-01 5.01773238e-01 2.23501325e-02 1.96151868e-01
-2.37782076e-01 4.07048911e-01 -2.16698691e-01 -4.33948100e-01
7.30057478e-01 8.78795028e-01 3.96688133e-01 -2.01567873e-01
-2.13247105e-01 4.75044429e-01 8.54207039e-01 -2.57931352e-01
3.53697181e-01 6.58417523e-01 3.37173820e-01 -7.47529030e-01
4.62068796e-01 -3.58971842e-02 -7.43712246e-01 -4.78531331e-01
5.62238872e-01 -4.11851704e-01 -5.42697847e-01 9.98400629e-01
-9.59544718e-01 -7.91239738e-01 -2.28634611e-01 1.95088282e-01
-9.70456123e-01 -2.74872154e-01 -4.45985496e-01 -5.48565507e-01
-2.22857177e-01 -7.82568991e-01 3.67600590e-01 5.43447971e-01
-7.89435988e-04 -5.64996839e-01 7.28646100e-01 -4.44683105e-01
9.98475552e-01 -1.68073978e-02 4.42043543e-01 -3.01421344e-01
-7.60367274e-01 -1.82793304e-01 3.28174293e-01 -2.26025641e-01
-2.66878247e-01 -3.57984900e-01 -1.00127876e+00 -1.50175914e-01
3.28006268e-01 1.62367538e-01 8.95380139e-01 4.03304212e-02
3.61795276e-01 2.61276066e-01 -6.40038550e-01 1.90140128e-01
1.68075442e+00 6.85057104e-01 1.10416031e+00 9.33156013e-01
-2.96936780e-01 6.98374152e-01 2.68932104e-01 2.11489514e-01
2.56007791e-01 7.36644804e-01 7.98925817e-01 6.90946937e-01
-1.58219114e-01 -8.06613266e-02 4.25972700e-01 5.66898763e-01
2.09525213e-01 1.89395323e-01 -1.14338696e+00 6.42722845e-01
-1.98055923e+00 -9.99476016e-01 1.01560719e-01 2.32378006e+00
3.31199706e-01 1.80363759e-01 -3.98965627e-01 2.07482502e-01
7.02248752e-01 -4.81980056e-01 -5.48061728e-01 -7.11217701e-01
-3.03622901e-01 2.44595721e-01 2.69569933e-01 2.46033207e-01
-5.84346533e-01 9.04580057e-01 5.60426855e+00 9.88464653e-02
-1.54104972e+00 5.03814332e-02 -2.21172005e-01 5.51711060e-02
3.15647423e-01 2.14833960e-01 -5.98968744e-01 7.62938797e-01
1.45237410e+00 -1.63956016e-01 9.59210396e-01 9.01385844e-01
-4.26639095e-02 -6.96286261e-01 -7.34159172e-01 8.56711388e-01
-1.26306852e-02 -1.41207612e+00 -2.85786092e-01 2.82070667e-01
3.35530967e-01 2.36169189e-01 -8.21165890e-02 1.10003255e-01
-2.42056832e-01 -9.67307210e-01 1.06019890e+00 1.14887297e+00
8.56894329e-02 -4.36676413e-01 5.36881685e-01 6.67567790e-01
-9.84063208e-01 -4.42669779e-01 -4.19193029e-01 -5.79334676e-01
1.60641477e-01 3.16631049e-01 -5.19451261e-01 -2.09016949e-01
8.15746248e-01 1.31378621e-01 -5.07250249e-01 1.54503858e+00
4.26228940e-02 -7.86846131e-02 -3.96738142e-01 -4.71720397e-01
-1.23066947e-01 -2.67576486e-01 6.40057385e-01 8.77199054e-01
8.53913963e-01 7.38864299e-03 -4.63169605e-01 1.11223722e+00
4.15364355e-01 -1.58771142e-01 -1.08906746e+00 1.21757753e-01
6.18901074e-01 1.15980339e+00 -1.10801482e+00 -5.64161092e-02
3.72331977e-01 8.95862699e-01 3.15190226e-01 1.84991062e-01
-4.76817280e-01 -6.86115742e-01 4.76161152e-01 2.79116839e-01
3.88036013e-01 -6.90406740e-01 -3.85940135e-01 -6.66408241e-01
-1.19146749e-01 -2.36431286e-01 -4.46067035e-01 -1.05205917e+00
-3.13875437e-01 7.79966474e-01 -5.68011940e-01 -1.03116584e+00
-4.56993282e-01 -8.68628323e-01 -5.29057443e-01 6.93249643e-01
-1.09917057e+00 -6.48904622e-01 -5.95349908e-01 3.92894417e-01
-4.10232954e-02 -4.16043907e-01 1.35280788e+00 -4.71069515e-02
-2.65825212e-01 -3.05951685e-01 2.98259467e-01 -2.95027763e-01
4.42460150e-01 -9.27060485e-01 1.62118956e-01 7.88493454e-01
-7.00193048e-02 8.18783820e-01 8.43333483e-01 -5.55662513e-01
-1.79192674e+00 -8.50180924e-01 7.64584124e-01 9.30860266e-02
3.81736308e-01 -6.27417624e-01 -1.03854680e+00 2.46322423e-01
1.65563941e-01 -2.59697530e-02 3.18399101e-01 -7.92384982e-01
-1.93999782e-01 -3.86460692e-01 -1.34157121e+00 7.96544194e-01
9.68638182e-01 -3.80236149e-01 -7.43090808e-01 -2.75497824e-01
-2.34392345e-01 -1.83301434e-01 -5.33688188e-01 8.05564299e-02
9.05621052e-01 -1.27471185e+00 6.30406857e-01 3.95475000e-01
8.93036462e-03 -7.89606690e-01 -1.73146009e-01 -1.43803120e+00
-3.05834979e-01 -2.13381305e-01 -8.74171853e-02 8.65745246e-01
1.09138615e-01 -1.24786711e+00 6.13520503e-01 4.60112989e-01
-2.15630382e-01 -3.53093386e-01 -1.47553802e+00 -1.04957438e+00
5.64858094e-02 -4.38190326e-02 3.62492472e-01 1.51115775e-01
4.15990561e-01 -1.04578577e-01 4.69507992e-01 1.11705765e-01
4.76025432e-01 4.66048233e-02 4.73665357e-01 -1.28146410e+00
-2.68570215e-01 -6.81880653e-01 -1.01458359e+00 -1.70108780e-01
1.67332530e-01 -9.06556666e-01 5.00578701e-01 -1.83096540e+00
-1.38621882e-01 -3.46581221e-01 -4.47733223e-01 4.58766073e-01
9.70335007e-01 4.01355982e-01 1.13986179e-01 3.29724997e-01
-4.38494116e-01 2.63617098e-01 5.16984224e-01 4.34432141e-02
-2.47929230e-01 -5.34750342e-01 -1.40688017e-01 3.10401797e-01
1.11806989e+00 -4.63891238e-01 -2.42474303e-01 -3.35903138e-01
1.60584882e-01 4.60007614e-05 5.91765404e-01 -2.07422256e+00
1.11906850e+00 3.42425972e-01 2.91447133e-01 -7.79954940e-02
5.62462509e-01 -1.06733048e+00 7.67340779e-01 1.38916159e+00
1.64989065e-02 2.85387374e-02 4.04675990e-01 5.58688343e-01
-2.53738225e-01 -4.46987033e-01 8.69121253e-01 -2.52149016e-01
-1.29103756e+00 -3.87258798e-01 -1.34729648e+00 -5.51217496e-01
1.44363916e+00 -7.50315249e-01 -5.19075572e-01 2.12549418e-01
-7.33352780e-01 -5.43538816e-02 1.05343091e+00 2.39267334e-01
6.62442148e-01 -1.06855452e+00 8.06018636e-02 5.49757898e-01
9.41363573e-02 -6.24590039e-01 1.53139278e-01 4.92633909e-01
-1.05455172e+00 8.13280880e-01 -1.33245039e+00 -5.15168190e-01
-6.48577631e-01 4.09736902e-01 7.52907991e-01 1.66106671e-01
-4.52825606e-01 2.38193378e-01 -2.66778499e-01 -4.67679948e-01
1.99558027e-02 -3.28361765e-02 -3.16009790e-01 -2.85589546e-01
6.05255902e-01 2.63640255e-01 4.06461447e-01 -5.14570534e-01
-6.62282944e-01 6.29408598e-01 6.82672560e-01 -6.17380023e-01
1.43437946e+00 4.46091518e-02 -5.21531820e-01 4.44716573e-01
4.63271856e-01 -3.87735546e-01 -9.85249519e-01 3.35979491e-01
2.31696114e-01 3.94912884e-02 -2.51443330e-02 -1.00726199e+00
-6.57451749e-01 9.89816487e-01 9.76279318e-01 1.03026956e-01
1.05003512e+00 -5.07347286e-01 4.11659539e-01 8.14150333e-01
1.28450680e+00 -1.32575154e+00 4.03119102e-02 9.77330863e-01
9.77750123e-01 -4.69589561e-01 -5.27693212e-01 1.45074621e-01
-1.01313934e-01 1.35279965e+00 5.68587065e-01 -6.50535226e-01
8.68474603e-01 6.44573927e-01 -1.44810930e-01 -3.31339911e-02
-8.45412552e-01 -1.16791449e-01 -3.90350819e-01 8.67794275e-01
-8.30485597e-02 -3.69621329e-02 -5.32109737e-01 4.94398654e-01
-1.57566771e-01 3.52634907e-01 7.63459325e-01 1.27681661e+00
-1.15466332e+00 -1.12332153e+00 -5.62536061e-01 -1.29041178e-02
1.76658198e-01 3.50301921e-01 -4.38163489e-01 6.52808845e-01
2.57242858e-01 6.26124382e-01 1.98488563e-01 -3.78680646e-01
2.43953899e-01 4.00263637e-01 5.97730696e-01 -4.28157955e-01
-5.92537880e-01 -4.03092682e-01 -2.89463133e-01 -7.08124220e-01
7.93896317e-02 -6.22340024e-01 -1.86296213e+00 9.81328636e-02
3.00437421e-01 8.90551060e-02 1.49263453e+00 6.09468758e-01
8.55875075e-01 4.31432575e-01 1.09808192e-01 -1.09853137e+00
-1.53547749e-01 -6.43768847e-01 -8.73609006e-01 -1.15061350e-01
-3.08542758e-01 -8.28167200e-01 -3.77330650e-03 -1.11766227e-01] | [8.15259075164795, 2.505573272705078] |
d876f97e-89a9-46a9-99fb-27f1b1b5ec2f | sparse-filtered-sirt-for-electron-tomography | 1608.01686 | null | http://arxiv.org/abs/1608.01686v2 | http://arxiv.org/pdf/1608.01686v2.pdf | Sparse Filtered SIRT for Electron Tomography | Electron tomographic reconstruction is a method for obtaining a
three-dimensional image of a specimen with a series of two dimensional
microscope images taken from different viewing angles. Filtered backprojection,
one of the most popular tomographic reconstruction methods, does not work well
under the existence of image noises and missing wedges. This paper presents a
new approach to largely mitigate the effect of noises and missing wedges. We
propose a novel filtered backprojection that optimizes the filter of the
backprojection operator in terms of a reconstruction error. This data-dependent
filter adaptively chooses the spectral domains of signals and noises,
suppressing the noise frequency bands, so it is very effective in denoising. We
also propose the new filtered backprojection embedded within the simultaneous
iterative reconstruction iteration for mitigating the effect of missing wedges.
Our numerical study is presented to show the performance gain of the proposed
approach over the state-of-the-art. | ['Chiwoo Park', 'Chen Mu'] | 2016-08-04 | null | null | null | null | ['electron-tomography'] | ['medical'] | [ 4.03623939e-01 -4.26990718e-01 7.48940587e-01 -1.22418433e-01
-4.65868920e-01 1.84113204e-01 2.01997772e-01 -4.04210240e-01
-8.18329990e-01 5.63923240e-01 1.34077340e-01 -5.00178747e-02
-2.47408748e-01 -6.05878115e-01 -3.29097956e-01 -9.82519507e-01
3.24989736e-01 2.49961480e-01 5.57971478e-01 5.12488410e-02
5.44927239e-01 5.16370058e-01 -1.39012396e+00 2.27330867e-02
6.61599040e-01 7.37075448e-01 7.52220571e-01 4.76204991e-01
2.97354683e-02 2.39415944e-01 -3.98328632e-01 2.70679537e-02
2.92559803e-01 -5.19255340e-01 -3.17708999e-01 4.25927758e-01
7.35972226e-02 -3.92953128e-01 -3.08540225e-01 1.42022014e+00
7.31465101e-01 3.06174122e-02 5.17292142e-01 -4.29373056e-01
-4.36507285e-01 1.84674766e-02 -7.78012335e-01 4.67743754e-01
1.77719325e-01 2.06180468e-01 2.19987705e-01 -1.36641133e+00
9.17329192e-01 9.30887043e-01 8.53188276e-01 3.70131761e-01
-1.44007909e+00 -4.10736740e-01 -5.14980137e-01 4.69263315e-01
-1.19808054e+00 -2.23472208e-01 1.02930367e+00 -3.71660531e-01
6.48306131e-01 3.61509383e-01 6.06311023e-01 4.41244096e-01
6.14650011e-01 8.09403658e-02 1.52030349e+00 -7.44404972e-01
1.21797748e-01 1.37847464e-03 3.67797107e-01 4.91902351e-01
1.82574734e-01 1.96482912e-02 -4.83852684e-01 -3.79600346e-01
8.58055711e-01 2.81063974e-01 -9.68833983e-01 -2.21377984e-01
-1.07369816e+00 2.25571617e-01 -1.75327435e-02 4.63914901e-01
-5.95440924e-01 -2.57018059e-01 6.93446100e-02 2.32381403e-01
6.47352755e-01 1.95923999e-01 -3.07419896e-02 1.82025522e-01
-1.02265573e+00 6.49425462e-02 4.62454200e-01 5.87560534e-01
6.97027802e-01 -5.12947738e-02 1.56681046e-01 7.61045218e-01
3.97019625e-01 4.49935168e-01 5.39688289e-01 -8.01137149e-01
2.79844314e-01 2.97359884e-01 1.93943799e-01 -8.17699492e-01
-4.76153821e-01 -2.02147886e-01 -8.30811203e-01 5.06639242e-01
2.34174877e-01 -1.90236822e-01 -8.55847597e-01 1.12447858e+00
4.84598190e-01 1.94340691e-01 -2.00930446e-01 1.09577286e+00
5.74737370e-01 8.32038462e-01 -4.72479880e-01 -1.05237949e+00
1.03709054e+00 -4.61858034e-01 -1.20102251e+00 2.59247154e-01
-1.79279044e-01 -1.19287550e+00 7.51987815e-01 6.25323713e-01
-1.15232646e+00 -4.19406801e-01 -1.09570777e+00 1.33786827e-01
4.14041907e-01 9.13356021e-02 -5.22444323e-02 3.82754385e-01
-7.82235265e-01 1.01005769e+00 -1.04545879e+00 -1.33611947e-01
-5.12815975e-02 1.62981749e-01 -3.38874638e-01 1.07141286e-01
-5.84284604e-01 9.45101023e-01 -1.79425851e-01 2.16885149e-01
-4.11700100e-01 -6.96483493e-01 -1.09259687e-01 -1.17737547e-01
2.37516999e-01 -4.50427115e-01 7.97528327e-01 -5.07649660e-01
-1.64543426e+00 6.48530841e-01 -2.07444772e-01 -1.34449005e-01
6.34155810e-01 -1.45599484e-01 -3.94455642e-01 4.91528869e-01
7.15050250e-02 -3.89805496e-01 1.03904819e+00 -1.13265717e+00
-1.16432659e-01 -7.72564292e-01 -8.84875238e-01 3.75806391e-01
-3.39090347e-01 1.22343302e-02 -3.90394926e-01 -5.28069556e-01
9.44724321e-01 -5.38406551e-01 -4.76236492e-01 -2.14781035e-02
-3.01595718e-01 2.02642530e-01 9.62164283e-01 -7.66498387e-01
1.02105916e+00 -2.22744942e+00 2.04208583e-01 7.41080344e-02
1.48386940e-01 1.29378989e-01 4.31834310e-01 7.05052555e-01
-7.62677416e-02 -3.04099590e-01 -4.82077152e-01 -5.18942654e-01
-6.70637727e-01 -8.56385976e-02 -1.07798666e-01 1.08686054e+00
-3.87564063e-01 -1.65557161e-01 -4.04300004e-01 -2.07183853e-01
3.22342604e-01 7.13177085e-01 -3.62367153e-01 2.71641046e-01
3.48507375e-01 8.04113090e-01 -4.05403644e-01 8.32045749e-02
1.20236242e+00 5.71283847e-02 1.70450374e-01 -5.12323678e-01
-6.27440691e-01 1.33073945e-02 -1.50290406e+00 1.14217842e+00
-6.65342212e-02 4.39079940e-01 6.53119266e-01 -7.93553591e-01
1.05288661e+00 5.22131920e-01 6.50769114e-01 -5.56292951e-01
1.11131318e-01 4.58401412e-01 -2.76472569e-01 -8.77489984e-01
4.05285835e-01 -1.00540650e+00 6.66071475e-01 3.16722959e-01
1.16030850e-01 -3.50897402e-01 -3.95291187e-02 -1.71629995e-01
1.01728535e+00 3.01679224e-02 3.39579254e-01 -6.19453669e-01
5.69876969e-01 -2.96595693e-01 7.69331336e-01 3.66413236e-01
-1.28932491e-01 1.05426681e+00 5.64187765e-02 -5.65555871e-01
-1.33909655e+00 -7.56927311e-01 -6.54201627e-01 1.31055936e-01
4.36287999e-01 -6.86786026e-02 -7.38058090e-01 -4.87578213e-02
-2.63242841e-01 4.05804783e-01 -2.06373721e-01 2.11460650e-01
-7.90384531e-01 -1.05716181e+00 -5.01136333e-02 -9.39406529e-02
7.25946069e-01 -7.67592013e-01 -7.92157054e-01 2.88532615e-01
-2.42873177e-01 -8.44924092e-01 -4.44356859e-01 2.48369366e-01
-1.17242420e+00 -1.23071623e+00 -9.80245471e-01 -5.96057296e-01
8.60574186e-01 4.53838646e-01 6.58611834e-01 3.80744822e-02
-2.09278554e-01 2.07458436e-01 -1.49428874e-01 5.33129573e-02
-3.55721086e-01 -7.73971736e-01 2.84826338e-01 1.85072824e-01
8.83289650e-02 -9.33453143e-01 -6.52619720e-01 5.21628082e-01
-8.60697746e-01 1.26764625e-01 2.10248351e-01 1.05956435e+00
9.69094157e-01 5.44468880e-01 1.17572285e-01 -1.04265928e+00
6.66265547e-01 -1.68040439e-01 -7.78313100e-01 -2.40687668e-01
-5.54113507e-01 -1.15047909e-01 1.06226337e+00 -2.38957062e-01
-1.46785593e+00 -1.07402638e-01 -3.89470100e-01 -3.47644925e-01
-5.06809093e-02 2.06808478e-01 5.17981946e-02 -4.10301715e-01
7.64687181e-01 5.67822218e-01 1.41251594e-01 -1.09366858e+00
-2.89942503e-01 5.55989683e-01 5.93733490e-01 -1.12939239e-01
6.27768219e-01 8.70154321e-01 2.82180876e-01 -1.33356416e+00
-2.37932146e-01 -8.37505698e-01 -4.30210352e-01 -3.29523265e-01
7.32288837e-01 -4.97449487e-01 -5.66321492e-01 7.38431871e-01
-1.17607868e+00 2.64809519e-01 -2.42000252e-01 8.55388999e-01
-6.60159349e-01 9.58004951e-01 -8.56908143e-01 -7.50707984e-01
-4.80855525e-01 -1.24301732e+00 5.02640009e-01 2.13269353e-01
2.99923033e-01 -6.03428662e-01 3.03184956e-01 1.08525068e-01
2.66223937e-01 6.92674285e-03 7.01579571e-01 5.77560849e-02
-4.34638977e-01 -7.25164115e-02 1.56358302e-01 5.67440748e-01
-9.39199980e-03 -1.96020335e-01 -7.10692525e-01 -3.54340822e-01
1.30062163e+00 1.89154208e-01 7.89201379e-01 5.98117888e-01
8.14680040e-01 -9.06540528e-02 -2.23958924e-01 8.76407087e-01
1.97276032e+00 2.41897136e-01 7.93348849e-01 2.23384544e-01
2.28311419e-01 2.88307220e-01 3.39230448e-01 5.89548886e-01
-3.44663262e-01 7.06013501e-01 2.76521027e-01 -1.96549091e-02
-1.02387860e-01 8.21859017e-02 -4.96697053e-02 1.07634068e+00
-4.26795095e-01 -6.04367396e-03 -6.44324422e-01 4.03784484e-01
-1.52665722e+00 -8.66611958e-01 -6.47179663e-01 2.25850058e+00
6.34990990e-01 5.72433472e-02 -3.00402641e-01 4.97631967e-01
7.64128506e-01 -5.20276800e-02 -2.67772585e-01 -3.44482735e-02
-4.03804593e-02 2.85466552e-01 7.17148185e-01 5.29976726e-01
-7.53679693e-01 2.31806248e-01 6.57533884e+00 7.08545625e-01
-1.23344195e+00 1.16931744e-01 3.46289203e-02 3.30825180e-01
-3.21711928e-01 -7.71551654e-02 -6.19439900e-01 6.77027047e-01
4.82895017e-01 -1.17913157e-01 1.78902879e-01 5.25018573e-01
4.81980145e-01 -5.31129718e-01 -5.18788218e-01 9.52224910e-01
1.04144514e-01 -1.18713391e+00 -1.29092500e-01 -6.35370146e-04
5.26058078e-01 -9.20626447e-02 -4.14153785e-02 -6.29890859e-01
-3.24434042e-01 -4.93609518e-01 5.74559271e-01 7.35653698e-01
3.43495041e-01 -5.63787520e-01 7.74835825e-01 6.95976198e-01
-6.82613254e-01 2.57160544e-01 -6.45422637e-01 -8.88094977e-02
6.40602291e-01 1.12648511e+00 -5.89777470e-01 5.42785347e-01
6.92024052e-01 5.19819438e-01 1.21811658e-01 1.26762903e+00
-7.58142918e-02 5.90314567e-01 -5.09488463e-01 1.12082869e-01
-2.16097683e-01 -7.62585878e-01 1.17919314e+00 8.92941713e-01
7.29177177e-01 6.84544086e-01 -9.11032185e-02 7.81148255e-01
2.06225365e-01 8.66223946e-02 -5.25119066e-01 5.61881423e-01
3.52946550e-01 9.73538637e-01 -7.86018789e-01 -1.64397001e-01
-4.02089268e-01 8.75832975e-01 -1.32215649e-01 4.59786206e-01
-2.60911405e-01 -2.33730406e-01 1.03202879e-01 6.64828122e-01
3.28427643e-01 -2.84167469e-01 -5.26392102e-01 -1.16071498e+00
2.45684385e-01 -5.96957862e-01 9.26704407e-02 -6.60890281e-01
-1.22712338e+00 5.06321311e-01 -1.01096593e-01 -1.38016713e+00
3.45316887e-01 -5.20390391e-01 -6.47167504e-01 1.01749480e+00
-1.19508135e+00 -4.22444075e-01 -1.01217702e-01 6.69365585e-01
4.66635048e-01 1.03986531e-01 6.68160617e-01 4.60449547e-01
-2.10197240e-01 -2.76598066e-01 4.87969756e-01 -5.21471024e-01
5.67742527e-01 -9.86744046e-01 -6.34399876e-02 1.06813753e+00
-4.41258907e-01 7.30395913e-01 1.30676365e+00 -7.87349403e-01
-1.35895813e+00 -3.96913439e-01 6.61076367e-01 3.76564711e-01
1.84819758e-01 9.84448642e-02 -1.10654521e+00 4.42602426e-01
2.28061676e-01 8.85582566e-02 4.74306077e-01 -4.66025621e-01
3.25592697e-01 -8.45409781e-02 -1.44915628e+00 3.09825480e-01
5.86794257e-01 -2.68329769e-01 -6.35329187e-01 2.19777346e-01
1.77592918e-01 -4.38812494e-01 -7.53647864e-01 5.19479156e-01
3.09887499e-01 -1.61774552e+00 9.22021091e-01 2.82618195e-01
3.21598589e-01 -4.53346789e-01 5.08426540e-02 -1.38882017e+00
-4.01936412e-01 -6.68899298e-01 4.39612269e-01 7.33111560e-01
-1.05442926e-01 -7.26451576e-01 7.69809842e-01 1.70534194e-01
-2.95469791e-01 -5.66410899e-01 -1.27059710e+00 -5.46922088e-01
-2.79528946e-01 -5.87282889e-02 -7.48795867e-02 5.60161352e-01
1.87087342e-01 2.31486335e-01 -6.05375767e-01 1.69133410e-01
1.16523409e+00 1.83782401e-03 4.86907065e-01 -1.02735734e+00
-5.00460684e-01 8.34679976e-02 -4.07920271e-01 -1.07596278e+00
-5.66587210e-01 -3.73911202e-01 3.27959359e-01 -1.42354918e+00
3.44188720e-01 -2.15775639e-01 1.13699622e-01 -3.10389906e-01
-2.78591663e-01 2.85695761e-01 1.47992760e-01 6.47820115e-01
8.53428692e-02 5.93531132e-01 1.62135434e+00 3.30039769e-01
-1.62811860e-01 2.23740935e-01 3.38163897e-02 1.11033332e+00
3.43245745e-01 -6.72350347e-01 -1.54651716e-01 -4.21306163e-01
-1.00963481e-01 3.72492313e-01 1.50317892e-01 -1.21924257e+00
5.49376726e-01 4.71749753e-02 2.67440468e-01 -9.04878020e-01
4.88742054e-01 -1.00545669e+00 6.75410509e-01 5.03720045e-01
1.61247298e-01 -8.12119525e-03 -8.99262205e-02 7.45490968e-01
-3.34237248e-01 -5.38347125e-01 1.29855347e+00 -4.84602690e-01
-2.65080512e-01 -1.68752149e-02 -6.44332707e-01 -4.08947200e-01
8.18525851e-01 -3.09164226e-01 -2.80516297e-02 -1.64967939e-01
-8.44327033e-01 -1.91998616e-01 6.44730628e-01 -5.86171865e-01
9.92337942e-01 -9.60302234e-01 -5.68884075e-01 5.76194823e-01
-5.02550602e-01 -1.81112066e-01 6.38976455e-01 1.00279033e+00
-9.04776812e-01 -6.61427453e-02 -3.58039439e-01 -5.21963894e-01
-1.48443258e+00 3.22883427e-01 4.35475707e-01 -3.88009995e-01
-1.27435529e+00 8.40182662e-01 4.00154814e-02 -3.09384614e-01
-1.19574793e-01 -1.32816490e-02 -2.74923891e-01 -3.75955164e-01
6.49896204e-01 7.38967955e-01 2.58241504e-01 -4.84161884e-01
-1.22462526e-01 9.20938492e-01 1.98363379e-01 -4.15630102e-01
1.57150602e+00 -3.59532177e-01 -3.84730428e-01 3.69179606e-01
1.07992947e+00 1.53803572e-01 -1.19270933e+00 -1.83690801e-01
-2.46479586e-01 -7.16576636e-01 2.34498799e-01 -3.09256017e-01
-8.05903852e-01 8.70153427e-01 5.60689628e-01 1.99776456e-01
1.33613706e+00 -5.42143106e-01 9.35168445e-01 -2.83010285e-02
4.63192433e-01 -1.25065267e+00 -1.02393655e-02 2.60227859e-01
7.31112599e-01 -7.10962951e-01 4.38850164e-01 -7.58716285e-01
-9.08310190e-02 1.32523930e+00 1.40983865e-01 -5.03023028e-01
9.25866961e-01 4.83664453e-01 6.41393941e-03 -3.16157907e-01
-4.95436400e-01 2.25350738e-01 -1.95322677e-01 4.25582469e-01
2.72873312e-01 -6.50018379e-02 -1.14158046e+00 3.28244746e-01
2.35650793e-01 2.93184280e-01 8.32085073e-01 1.04004550e+00
-7.53187776e-01 -9.56954777e-01 -8.68873835e-01 3.08962703e-01
-7.38915741e-01 8.37276801e-02 1.85104266e-01 3.98317069e-01
1.80991620e-01 7.55777359e-01 -1.47586823e-01 -1.13052323e-01
5.29965639e-01 -1.21806338e-01 5.48032522e-01 -4.46785599e-01
-3.83777648e-01 5.48264444e-01 -1.16873220e-01 -3.74944717e-01
-5.99230945e-01 -5.54071486e-01 -1.27112865e+00 -2.07949281e-01
-5.16895473e-01 2.65672088e-01 5.91267526e-01 7.57668257e-01
-5.29203005e-02 4.02643234e-01 7.96679258e-01 -9.98381257e-01
-8.74026537e-01 -9.96583402e-01 -1.12479389e+00 5.37257731e-01
5.03929973e-01 -5.46445072e-01 -6.88055217e-01 1.32664889e-01] | [12.648848533630371, -2.742842435836792] |
a925f99d-09ca-4282-8397-7b927bf2c0a5 | discriminative-region-attention-and | 2204.13323 | null | https://arxiv.org/abs/2204.13323v1 | https://arxiv.org/pdf/2204.13323v1.pdf | Discriminative-Region Attention and Orthogonal-View Generation Model for Vehicle Re-Identification | Vehicle re-identification (Re-ID) is urgently demanded to alleviate thepressure caused by the increasingly onerous task of urban traffic management. Multiple challenges hamper the applications of vision-based vehicle Re-ID methods: (1) The appearances of different vehicles of the same brand/model are often similar; However, (2) the appearances of the same vehicle differ significantly from different viewpoints. Previous methods mainly use manually annotated multi-attribute datasets to assist the network in getting detailed cues and in inferencing multi-view to improve the vehicle Re-ID performance. However, finely labeled vehicle datasets are usually unattainable in real application scenarios. Hence, we propose a Discriminative-Region Attention and Orthogonal-View Generation (DRA-OVG) model, which only requires identity (ID) labels to conquer the multiple challenges of vehicle Re-ID.The proposed DRA model can automatically extract the discriminative region features, which can distinguish similar vehicles. And the OVG model can generate multi-view features based on the input view features to reduce the impact of viewpoint mismatches. Finally, the distance between vehicle appearances is presented by the discriminative region features and multi-view features together. Therefore, the significance of pairwise distance measure between vehicles is enhanced in acomplete feature space. Extensive experiments substantiate the effectiveness of each proposed ingredient, and experimental results indicate that our approach achieves remarkable improvements over the state- of-the-art vehicle Re-ID methods on VehicleID and VeRi-776 datasets. | ['Li Ge', 'Lin Wang', 'Ying WEI', 'Yuefeng Wang', 'Huadong Li'] | 2022-04-28 | null | null | null | null | ['vehicle-re-identification'] | ['computer-vision'] | [-2.91030049e-01 -4.18843716e-01 -2.88719743e-01 -5.27579010e-01
-7.22184300e-01 -6.30116463e-01 8.80229831e-01 -3.45446825e-01
-1.70803681e-01 4.21330005e-01 -2.38689221e-02 -1.01077497e-01
1.37543827e-01 -6.00318670e-01 -6.99151933e-01 -8.31263781e-01
4.87639010e-01 3.71936321e-01 3.53572130e-01 -2.42486104e-01
2.16247365e-01 6.92113101e-01 -1.98037899e+00 1.22019775e-01
9.72018361e-01 9.60785329e-01 2.67326802e-01 2.63429642e-01
-1.35132536e-01 2.43740052e-01 -1.97304130e-01 -5.50438285e-01
3.67901415e-01 -9.46668833e-02 -3.39372903e-01 2.96457946e-01
7.51124203e-01 -5.79373240e-01 -3.68263841e-01 1.28164101e+00
2.27565840e-01 7.27103427e-02 8.14965010e-01 -1.85153234e+00
-4.29743797e-01 2.56259516e-02 -8.37974012e-01 3.37091178e-01
1.54116496e-01 1.13334313e-01 7.23386526e-01 -1.32162082e+00
4.89919007e-01 1.44874072e+00 4.38038528e-01 4.08285141e-01
-1.07690930e+00 -1.16403294e+00 4.45338011e-01 8.53303373e-01
-1.63879609e+00 -7.78568864e-01 9.11037028e-01 -6.30100906e-01
3.33612889e-01 2.70787627e-01 3.09987187e-01 9.54275787e-01
-2.06260338e-01 7.93039799e-01 8.74090195e-01 9.29658264e-02
-3.08267802e-01 4.58921999e-01 3.15563008e-02 4.19863254e-01
4.84697908e-01 2.98075855e-01 -1.70763031e-01 1.10771157e-01
3.47377717e-01 3.81603867e-01 -3.72530404e-03 -6.04412019e-01
-1.13455832e+00 6.88659489e-01 1.99580610e-01 5.71895353e-02
-2.05892056e-01 -1.55091599e-01 4.69360054e-01 1.71893880e-01
3.00440103e-01 -1.34513780e-01 -4.85567786e-02 1.58214226e-01
-5.44149399e-01 2.78442770e-01 -1.19851502e-02 1.15921891e+00
1.12249458e+00 9.55798477e-02 -1.06404498e-01 8.08473170e-01
4.63940322e-01 9.62566376e-01 6.51249737e-02 -7.88739085e-01
7.42275000e-01 5.85184038e-01 1.99537933e-01 -1.47465491e+00
-1.12204939e-01 -1.69974819e-01 -7.76110768e-01 2.87863851e-01
2.69551098e-01 1.31434709e-01 -6.14805520e-01 1.53897142e+00
5.20816267e-01 3.86216253e-01 9.28190276e-02 1.02042925e+00
1.19002390e+00 4.96120512e-01 3.49806063e-02 -9.49476436e-02
1.31787252e+00 -9.65270400e-01 -5.89594483e-01 -1.19889855e-01
5.70793927e-01 -8.16223025e-01 4.06868756e-01 -7.97059834e-02
-5.64280033e-01 -9.80768800e-01 -9.75681543e-01 3.33440512e-01
-3.49255323e-01 2.42725894e-01 1.51951507e-01 4.87394750e-01
-7.43996739e-01 -1.33759622e-02 -1.76539630e-01 -2.19795272e-01
4.28329855e-01 2.08453983e-01 -8.29316735e-01 -3.95315260e-01
-1.05632579e+00 7.14404464e-01 1.86447531e-01 3.49421829e-01
-1.09835005e+00 -5.67412257e-01 -8.96847725e-01 -2.55823582e-01
3.94726157e-01 -3.02344441e-01 7.83927858e-01 -8.93939614e-01
-8.79489064e-01 8.59270930e-01 -4.42468464e-01 -4.90386263e-02
6.38407707e-01 2.02650666e-01 -9.97186005e-01 3.41096781e-02
5.67334056e-01 6.66597009e-01 8.79644275e-01 -1.84641552e+00
-1.16178751e+00 -4.04078931e-01 -1.20011806e-01 2.15969399e-01
4.64865938e-03 -2.06639990e-02 -7.18722343e-01 -3.52380186e-01
9.33674444e-03 -9.84083712e-01 1.05577014e-01 -8.44673663e-02
-3.91960859e-01 -3.77897501e-01 1.26082778e+00 -6.22927904e-01
8.97435725e-01 -2.30099511e+00 -2.92351931e-01 2.76871949e-01
3.76296371e-01 4.28881139e-01 -3.90650690e-01 1.42841518e-01
-1.74253359e-01 -4.86244746e-02 4.89665478e-01 -9.73992422e-02
-1.35834455e-01 1.60085276e-01 -2.24191681e-01 6.55013144e-01
6.15845956e-02 7.57822931e-01 -9.00976539e-01 -6.86700523e-01
5.71073353e-01 2.40691051e-01 -1.80723324e-01 2.26547629e-01
4.19579327e-01 5.18880069e-01 -5.22593975e-01 7.17280746e-01
1.17609179e+00 8.09603557e-02 -2.03111932e-01 -7.24389076e-01
-2.11027816e-01 -3.00162882e-01 -1.20568907e+00 8.99500966e-01
-3.80434036e-01 6.95913315e-01 -8.43781978e-03 -9.62878406e-01
1.00073111e+00 1.15921438e-01 4.65234518e-01 -8.50834310e-01
1.34182811e-01 1.70805275e-01 -4.18525524e-02 -7.24256754e-01
4.78307515e-01 2.69217074e-01 -7.93817714e-02 5.03300279e-02
-3.63451600e-01 5.40509939e-01 1.35345757e-01 9.77549553e-02
3.64841849e-01 -2.24023312e-01 8.91998038e-02 -1.05292253e-01
1.02152646e+00 -1.72423720e-01 8.65390301e-01 3.94634277e-01
-5.22644520e-01 4.29927498e-01 1.30976856e-01 -4.82067525e-01
-1.01061916e+00 -9.99109089e-01 -1.72799975e-01 8.41331542e-01
1.03065777e+00 2.50028577e-02 -4.59568143e-01 -9.88000691e-01
2.56582201e-01 5.80977798e-01 -5.81948936e-01 -1.05851859e-01
-5.57242095e-01 -3.34301323e-01 4.30545092e-01 6.09953225e-01
6.46034896e-01 -4.16223943e-01 1.07168611e-02 -8.54438692e-02
-4.44989890e-01 -1.52293360e+00 -7.35783637e-01 -7.19412446e-01
-3.16279590e-01 -1.12834322e+00 -5.61833322e-01 -9.28468168e-01
9.07127917e-01 1.23234999e+00 5.82225919e-01 1.99349314e-01
1.01621851e-01 1.45749971e-01 -2.80566216e-01 -1.34437427e-01
-4.92650867e-01 -3.24479312e-01 4.18024033e-01 6.98167682e-01
6.71493351e-01 -3.96921158e-01 -7.44945526e-01 9.14768815e-01
-2.24461272e-01 1.71392471e-01 5.72177887e-01 6.71679258e-01
6.77045465e-01 6.84100091e-02 6.98636770e-01 -4.99586344e-01
-4.07536887e-03 -6.21232271e-01 -6.65106714e-01 2.30601594e-01
-7.47299612e-01 -2.13486820e-01 5.92701316e-01 -3.42174709e-01
-9.99841511e-01 -4.35678251e-02 8.20491016e-02 -7.43527591e-01
-3.94233853e-01 -9.21539888e-02 -6.92385256e-01 -2.62807518e-01
3.21135484e-02 3.46820205e-01 1.61862060e-01 -3.09570074e-01
4.10117567e-01 9.00346220e-01 5.08029521e-01 -1.26329660e-01
1.09714746e+00 6.46316111e-01 -4.13417406e-02 -6.79244936e-01
-3.43070447e-01 -7.75126457e-01 -5.49592495e-01 -5.97554147e-01
8.62166762e-01 -1.27940059e+00 -7.46007621e-01 4.89932656e-01
-1.22715783e+00 4.91279244e-01 3.00185710e-01 5.42477489e-01
-1.75130516e-01 6.48911774e-01 -4.45349738e-02 -5.82227767e-01
5.53312674e-02 -1.47323132e+00 1.09185135e+00 4.14444476e-01
2.14122817e-01 -6.74248993e-01 -2.52462804e-01 5.84781170e-01
1.28760472e-01 2.86526680e-02 7.70418525e-01 -6.69559956e-01
-7.94700742e-01 -2.27500722e-01 -7.56698430e-01 2.02008501e-01
2.05670401e-01 6.81446716e-02 -1.02322757e+00 -2.27557674e-01
-5.45427620e-01 2.71608621e-01 7.29530454e-01 8.07462633e-02
9.73246634e-01 -2.06615373e-01 -8.02957773e-01 5.43357909e-01
1.21051741e+00 3.83809000e-01 4.70357150e-01 2.80807793e-01
1.13469398e+00 7.44120359e-01 1.05474234e+00 2.43097201e-01
8.70736003e-01 1.05165517e+00 7.09617555e-01 -1.67051345e-01
-2.51975507e-01 -3.67653817e-01 3.57708454e-01 7.37897038e-01
-6.24139830e-02 -1.00523002e-01 -6.39657080e-01 9.52437639e-01
-1.93598747e+00 -1.30488825e+00 -4.81547803e-01 2.06994343e+00
8.64569992e-02 -9.52004194e-02 2.51566142e-01 -1.27154484e-01
1.23507297e+00 1.45228878e-01 -5.50574303e-01 -1.33161530e-01
-6.81971237e-02 -9.76732314e-01 6.49316847e-01 3.92171800e-01
-1.18001270e+00 7.36824334e-01 5.11735296e+00 1.06140709e+00
-9.22988176e-01 3.59455235e-02 5.48210740e-01 3.82864773e-01
-5.93136668e-01 -1.62900820e-01 -1.21023881e+00 6.50823355e-01
2.82516718e-01 -3.16291273e-01 3.45623970e-01 1.05364096e+00
2.65409619e-01 1.31677672e-01 -1.01785517e+00 1.39288080e+00
5.04151642e-01 -1.21119404e+00 2.60282189e-01 2.15782061e-01
7.50376821e-01 3.86051163e-02 2.73779899e-01 2.29328007e-01
1.02142960e-01 -7.46557951e-01 7.01913953e-01 3.95505399e-01
9.53339219e-01 -9.09586430e-01 8.55694354e-01 1.34570196e-01
-1.78097451e+00 -1.50809407e-01 -2.73801744e-01 4.57416654e-01
2.74426967e-01 2.01070234e-01 -8.29001963e-01 8.00214231e-01
7.24886417e-01 9.87668335e-01 -7.52799630e-01 8.60113800e-01
1.86130419e-01 1.14928916e-01 5.45612872e-02 2.59562373e-01
9.82058421e-02 -3.36609155e-01 7.34367907e-01 9.12177920e-01
4.11335260e-01 -1.90557376e-01 2.55459994e-01 7.31712222e-01
1.80967495e-01 -6.05153106e-02 -8.62217963e-01 3.15176845e-01
7.90464342e-01 1.48560178e+00 -4.30648029e-01 -2.76369154e-01
-6.87134027e-01 6.66129112e-01 -1.53329745e-02 6.08485997e-01
-1.15024567e+00 -2.25220859e-01 1.22541666e+00 2.20254421e-01
6.83221817e-01 4.17266339e-02 4.83331387e-04 -9.38287854e-01
4.07916643e-02 -6.44617558e-01 1.72566593e-01 -7.35604644e-01
-1.39310575e+00 6.69369578e-01 1.64229542e-01 -1.93334210e+00
-2.77371317e-01 -3.87368560e-01 -5.25210261e-01 7.14628160e-01
-1.81792271e+00 -1.37609243e+00 -6.11892581e-01 5.80609024e-01
6.67833805e-01 -5.23571968e-01 2.15508446e-01 6.77991331e-01
-6.84632003e-01 9.64759290e-01 2.44424090e-01 2.74061143e-01
7.22895324e-01 -5.62217593e-01 3.18004131e-01 9.57682371e-01
-1.50075018e-01 2.41247118e-01 4.64922160e-01 -5.73555708e-01
-1.31257069e+00 -1.53143668e+00 8.71814728e-01 -3.78970563e-01
3.81873578e-01 -2.01499358e-01 -7.15004206e-01 4.44936454e-01
-1.25255585e-01 3.09178650e-01 4.61731553e-01 -2.98454076e-01
-5.10001123e-01 -5.93442857e-01 -9.19943333e-01 5.23210168e-01
1.02447450e+00 -5.67739010e-01 -1.75487012e-01 -1.03177972e-01
4.55039144e-01 -8.98067728e-02 -5.72211504e-01 5.39796293e-01
6.64317489e-01 -9.24891531e-01 1.22241759e+00 -4.16086316e-01
-5.70731796e-03 -8.50384533e-01 -2.83188790e-01 -9.80383515e-01
-4.02089030e-01 -9.12731364e-02 1.23749986e-01 1.68263113e+00
5.02048321e-02 -7.73732543e-01 3.22813481e-01 3.81026864e-01
-8.67065340e-02 -4.69883680e-01 -9.76312459e-01 -8.28664482e-01
-2.99431771e-01 -2.61069119e-01 1.04348838e+00 8.39608192e-01
-5.16503990e-01 2.73459494e-01 -4.51881260e-01 5.34035563e-01
7.31380939e-01 4.55314785e-01 1.20138955e+00 -1.38783300e+00
1.91988423e-01 -4.21233445e-01 -8.36757362e-01 -1.07267988e+00
4.01316345e-01 -8.56724322e-01 6.78802803e-02 -1.21613073e+00
5.12569249e-01 -6.88784420e-01 -2.21407935e-01 1.14707790e-01
-2.27166310e-01 2.76995093e-01 2.24988341e-01 3.24390173e-01
-7.41775632e-01 6.04713857e-01 1.23995042e+00 -3.25973064e-01
2.92999625e-01 6.92642480e-02 -6.91127777e-01 6.67026222e-01
5.42167902e-01 -3.72814447e-01 -4.94496047e-01 -3.71641725e-01
-1.52469784e-01 1.28387725e-02 7.82702863e-01 -7.01016605e-01
1.92969173e-01 -3.71942163e-01 2.91551054e-01 -9.99018490e-01
2.17228606e-01 -1.12159467e+00 3.38025928e-01 8.71970728e-02
1.93282813e-01 2.93702602e-01 -5.11532277e-02 9.30277288e-01
-3.33404034e-01 4.98532467e-02 6.05386674e-01 2.16458812e-01
-1.36682081e+00 6.22304618e-01 -2.75107950e-01 7.77936876e-02
1.42558491e+00 -6.71645105e-01 -3.07668120e-01 -3.13960493e-01
-1.16633609e-01 4.60889757e-01 6.54168844e-01 9.30983126e-01
5.66664457e-01 -1.91043270e+00 -8.94027829e-01 5.26105046e-01
6.42599285e-01 -1.66819692e-01 7.30596244e-01 9.41918373e-01
-8.40239376e-02 3.11477035e-01 -2.62793988e-01 -9.54262197e-01
-1.62937355e+00 7.98685551e-01 2.47863457e-01 2.14367256e-01
-4.51975763e-01 4.78877217e-01 8.25990736e-01 -3.23519051e-01
-1.02596693e-01 2.30106920e-01 -6.47617757e-01 2.56237298e-01
6.81981564e-01 6.22525394e-01 -8.57583210e-02 -1.66383076e+00
-6.57650352e-01 9.71379519e-01 -3.56060952e-01 3.42194825e-01
8.63783300e-01 -7.10093737e-01 2.42532268e-01 7.70119727e-02
1.36777985e+00 5.17853117e-03 -1.41366434e+00 -3.15160960e-01
-4.78248805e-01 -8.18209469e-01 -1.70224264e-01 -1.56868413e-01
-1.23084092e+00 8.65438521e-01 7.64287710e-01 -1.00388013e-01
7.75537133e-01 1.57589149e-02 9.06268597e-01 3.01142354e-02
4.60180461e-01 -1.12470388e+00 -4.44673225e-02 2.33864620e-01
5.80197453e-01 -1.58035684e+00 -1.44628048e-01 -6.43164575e-01
-8.56302500e-01 9.77872610e-01 8.92296970e-01 1.61670089e-01
5.18458009e-01 -1.34868428e-01 1.48649707e-01 -7.55896093e-03
-2.97537178e-01 -4.03230369e-01 4.06280607e-01 8.43219280e-01
-2.53338367e-01 9.85520259e-02 1.10512666e-01 5.48633456e-01
1.55999258e-01 -5.46762109e-01 2.09227771e-01 2.45411485e-01
-3.10307890e-01 -8.39587212e-01 -4.31119800e-01 1.77599832e-01
1.17420629e-01 2.74261355e-01 -3.92365530e-02 7.31875539e-01
3.91435564e-01 1.19730771e+00 1.60380289e-01 -9.00099695e-01
2.48716041e-01 -4.26184267e-01 6.55668601e-02 -1.07697569e-01
4.72653024e-02 -1.54512241e-01 1.06586866e-01 -4.69124585e-01
-5.07846653e-01 -8.09991121e-01 -1.04771888e+00 -5.48469484e-01
-5.54191172e-01 -9.68843699e-03 3.97269368e-01 1.04321039e+00
6.80941701e-01 2.80916959e-01 1.20231318e+00 -9.76694584e-01
-3.51242423e-01 -4.63649482e-01 -3.17168623e-01 7.41956890e-01
5.32964706e-01 -1.08465576e+00 -3.30693632e-01 6.97664693e-02] | [8.125836372375488, -1.0086445808410645] |
85d54fda-8328-4547-8e4a-4f3491131c3c | language-matters-a-weakly-supervised-pre | 2203.03911 | null | https://arxiv.org/abs/2203.03911v3 | https://arxiv.org/pdf/2203.03911v3.pdf | Language Matters: A Weakly Supervised Vision-Language Pre-training Approach for Scene Text Detection and Spotting | Recently, Vision-Language Pre-training (VLP) techniques have greatly benefited various vision-language tasks by jointly learning visual and textual representations, which intuitively helps in Optical Character Recognition (OCR) tasks due to the rich visual and textual information in scene text images. However, these methods cannot well cope with OCR tasks because of the difficulty in both instance-level text encoding and image-text pair acquisition (i.e. images and captured texts in them). This paper presents a weakly supervised pre-training method, oCLIP, which can acquire effective scene text representations by jointly learning and aligning visual and textual information. Our network consists of an image encoder and a character-aware text encoder that extract visual and textual features, respectively, as well as a visual-textual decoder that models the interaction among textual and visual features for learning effective scene text representations. With the learning of textual features, the pre-trained model can attend texts in images well with character awareness. Besides, these designs enable the learning from weakly annotated texts (i.e. partial texts in images without text bounding boxes) which mitigates the data annotation constraint greatly. Experiments over the weakly annotated images in ICDAR2019-LSVT show that our pre-trained model improves F-score by +2.5\% and +4.8\% while transferring its weights to other text detection and spotting networks, respectively. In addition, the proposed method outperforms existing pre-training techniques consistently across multiple public datasets (e.g., +3.2\% and +1.3\% for Total-Text and CTW1500). | ['Shijian Lu', 'Wenqing Zhang', 'Song Bai', 'Philip Torr', 'Yu Hao', 'Chuhui Xue'] | 2022-03-08 | null | null | null | null | ['scene-text-detection'] | ['computer-vision'] | [ 7.11598277e-01 -2.07691580e-01 -2.46454656e-01 -2.54992247e-01
-7.07009315e-01 -4.50672448e-01 7.02165365e-01 7.81094597e-04
-6.75880909e-01 3.89962375e-01 2.07540959e-01 -2.26328313e-01
5.44567168e-01 -4.95914429e-01 -1.19331694e+00 -5.94511867e-01
5.80993772e-01 4.14132178e-01 2.93143183e-01 1.51779145e-01
4.16523099e-01 3.89913954e-02 -1.42831564e+00 7.40035892e-01
1.09974241e+00 9.43812191e-01 9.11465108e-01 8.38516057e-01
-4.58715558e-01 1.26241601e+00 -4.52201515e-01 -5.44152558e-01
-2.06895601e-02 -2.39857480e-01 -5.13274431e-01 7.07512379e-01
9.40859675e-01 -6.63939714e-01 -7.58638382e-01 9.84400928e-01
3.48948121e-01 -1.41832009e-01 7.11042166e-01 -9.63561475e-01
-1.08905518e+00 5.25954902e-01 -9.57416475e-01 -9.67626274e-02
1.09203324e-01 3.22657228e-01 8.79651487e-01 -1.22134054e+00
4.57547158e-01 9.72663224e-01 4.17995006e-01 6.60434723e-01
-8.25780630e-01 -4.78765190e-01 1.32188916e-01 2.96805292e-01
-1.13704550e+00 -5.37296236e-01 6.60513163e-01 -5.50810039e-01
1.09220040e+00 2.07216963e-01 4.25741971e-01 1.25782454e+00
-4.79547195e-02 1.54550743e+00 8.31381857e-01 -6.46461189e-01
-3.27781498e-01 4.04110342e-01 1.36976363e-03 8.22875500e-01
2.83796430e-01 -3.23339939e-01 -5.83137095e-01 5.85013032e-01
7.41662681e-01 1.34736776e-01 -2.70164669e-01 -6.26921430e-02
-1.18175447e+00 4.14144903e-01 1.13570102e-01 1.91685528e-01
-3.70264165e-02 2.46236786e-01 6.21463716e-01 -2.13377833e-01
2.12505683e-01 8.02198052e-02 -2.10602611e-01 1.29311187e-02
-1.11996388e+00 -3.90394211e-01 3.36908907e-01 1.44996870e+00
6.19400561e-01 4.40278023e-01 -4.66542721e-01 1.11272669e+00
3.94605160e-01 1.03548384e+00 4.26525354e-01 -2.86346942e-01
1.04791391e+00 6.27361715e-01 -2.25882426e-01 -8.63542914e-01
6.75531551e-02 -1.56122923e-01 -1.08505654e+00 -4.13358599e-01
3.38571072e-01 6.57577515e-02 -1.36959052e+00 1.21639490e+00
-2.62179703e-01 2.30116453e-02 1.22237727e-01 8.78504217e-01
8.89790177e-01 9.81467426e-01 2.62350291e-02 -1.05457611e-01
1.50722468e+00 -1.36179614e+00 -8.51413012e-01 -8.29464853e-01
5.71821272e-01 -1.14961112e+00 1.43901050e+00 2.87354738e-01
-9.62912023e-01 -7.54015446e-01 -1.01465833e+00 -5.21190047e-01
-2.23631367e-01 8.94055188e-01 3.25992495e-01 5.50159454e-01
-8.93794417e-01 -1.05215289e-01 -5.76507032e-01 -3.79291415e-01
6.97791934e-01 1.23561025e-01 -1.26023397e-01 -4.56568748e-01
-7.16899931e-01 6.16558135e-01 5.50811827e-01 1.75848842e-01
-1.27217114e+00 -3.31526339e-01 -9.40263748e-01 5.62790744e-02
6.64355040e-01 -2.24620312e-01 8.58534753e-01 -1.16908848e+00
-1.24612749e+00 9.87737775e-01 -3.70966077e-01 -4.97055501e-01
5.90323329e-01 -4.13516104e-01 -2.97534138e-01 4.98499662e-01
1.43495157e-01 8.19152534e-01 1.20929801e+00 -1.24304378e+00
-6.57256126e-01 -1.66067198e-01 -3.76758128e-01 3.97970885e-01
-8.35207999e-01 -2.63513438e-02 -1.31504643e+00 -7.46460438e-01
-2.61928439e-01 -5.90658128e-01 3.45940255e-02 2.34133571e-01
-6.78608119e-01 -1.36931449e-01 1.07544398e+00 -8.11540782e-01
9.40150619e-01 -1.97466242e+00 -1.49853006e-01 -3.09278756e-01
1.91292435e-01 4.99969214e-01 -4.01158541e-01 2.99108505e-01
1.83972314e-01 2.55899001e-02 -9.58205089e-02 -6.60011768e-01
-1.10584259e-01 2.02269912e-01 -5.44223189e-01 3.40301275e-01
4.31117266e-01 1.20410311e+00 -5.40800512e-01 -9.90036488e-01
7.52571762e-01 3.06830138e-01 -3.59659672e-01 2.28528693e-01
-5.58709681e-01 4.86859295e-04 -3.52943331e-01 8.04480731e-01
6.26931727e-01 -5.48718929e-01 1.16475202e-01 -2.07234144e-01
-1.20105572e-01 -1.22639202e-01 -7.84844100e-01 1.59943962e+00
-3.96965265e-01 1.27116370e+00 -2.37051964e-01 -1.18771970e+00
9.04630542e-01 2.11622864e-01 2.08952531e-01 -1.05802429e+00
1.85700625e-01 4.27871235e-02 -4.10814106e-01 -8.01198423e-01
7.09251702e-01 4.05949295e-01 9.65915713e-03 2.20709130e-01
9.33685154e-02 -1.00894302e-01 3.06475967e-01 4.84525859e-01
4.81309891e-01 8.02457407e-02 -1.66241154e-01 6.74702674e-02
7.38762140e-01 3.01746768e-03 1.81177124e-01 8.63208354e-01
-3.25507037e-02 6.40175343e-01 2.54753560e-01 -3.01170290e-01
-1.39075160e+00 -6.93492413e-01 -1.22796178e-01 1.18747258e+00
3.92588913e-01 -4.64053780e-01 -5.73857307e-01 -5.80294311e-01
-2.61975825e-01 5.61113596e-01 -4.78653580e-01 -5.40139265e-02
-5.95202506e-01 -4.47259039e-01 8.01440001e-01 8.13917518e-01
8.51541758e-01 -9.38867629e-01 -3.35667789e-01 -9.65737328e-02
-4.32639539e-01 -1.83159888e+00 -8.62870812e-01 1.73835456e-01
-7.05588341e-01 -7.67986357e-01 -8.22617948e-01 -1.26327133e+00
9.02252018e-01 6.05023205e-01 7.56783605e-01 1.79710835e-01
-4.99642462e-01 4.17567194e-01 -4.69943106e-01 -3.10640663e-01
-2.54685253e-01 -1.91877216e-01 -2.46803164e-01 2.06636935e-01
4.53090549e-01 1.77156538e-01 -4.42755967e-01 2.15210572e-01
-1.03419268e+00 6.36650980e-01 9.66357231e-01 1.09393978e+00
5.27864814e-01 -1.51816115e-01 3.10609248e-02 -6.95529699e-01
1.12185314e-01 6.30164072e-02 -7.02666163e-01 6.73310816e-01
-3.73466223e-01 -1.66167822e-02 7.28501916e-01 -6.37568653e-01
-1.10030878e+00 3.44588876e-01 8.00649896e-02 -6.22244537e-01
-1.87099203e-01 2.43238896e-01 -2.51466781e-01 1.01275675e-01
4.11988497e-01 1.09374559e+00 -3.92761558e-01 -1.93533793e-01
3.81258696e-01 1.06209159e+00 7.10193872e-01 -6.26622915e-01
8.77592742e-01 4.77049321e-01 -4.51769978e-01 -1.29861999e+00
-8.99098456e-01 -5.39766669e-01 -6.36928797e-01 -2.73887098e-01
1.16382086e+00 -1.21431780e+00 -5.88207781e-01 8.00148427e-01
-1.12595654e+00 -4.19539779e-01 2.12182961e-02 4.87394333e-01
-4.79644448e-01 9.86798286e-01 -6.30655229e-01 -8.39541852e-01
-3.91638398e-01 -1.22074354e+00 1.37418818e+00 2.56256282e-01
4.39986080e-01 -8.76272440e-01 -5.88202536e-01 8.47835898e-01
1.74896941e-01 -1.44124344e-01 7.53042161e-01 -4.18780476e-01
-8.56276929e-01 -1.19884469e-01 -8.54510188e-01 5.66696465e-01
-1.17732227e-01 4.37166058e-02 -1.17431831e+00 -3.25109243e-01
-3.63913566e-01 -7.42907047e-01 1.13251972e+00 2.98960984e-01
1.36104858e+00 -3.21909219e-01 -2.87079781e-01 6.15042567e-01
1.53176677e+00 2.19300658e-01 8.36814940e-01 3.12356412e-01
1.30384612e+00 4.88081336e-01 4.13902551e-01 4.12569702e-01
3.52449417e-01 5.26734948e-01 4.28458750e-01 -4.01755273e-01
-3.76962811e-01 -4.63791400e-01 6.12232625e-01 7.63958097e-01
2.22446874e-01 -5.15550375e-01 -1.01902199e+00 5.41936636e-01
-1.75829494e+00 -8.24773014e-01 -3.89287353e-01 1.88394606e+00
1.14613807e+00 2.62465000e-01 -2.77010471e-01 -6.01525269e-02
9.28208053e-01 2.72012919e-01 -6.54831529e-01 -1.42311677e-01
-4.00216669e-01 -7.50478283e-02 7.40444243e-01 2.04739183e-01
-1.20658386e+00 1.45609713e+00 4.77261972e+00 1.07599044e+00
-1.16018355e+00 -2.57023275e-01 6.43473923e-01 1.03178538e-01
-2.09426321e-02 -1.35326296e-01 -9.93355691e-01 5.69841743e-01
5.27030289e-01 5.94062917e-02 2.28336647e-01 8.73304546e-01
4.97551374e-02 -3.36056687e-02 -1.20760202e+00 1.28334630e+00
6.21560574e-01 -1.48531175e+00 6.11022115e-01 6.15661629e-02
9.23091590e-01 -2.90821046e-02 3.19839180e-01 2.95530885e-01
4.30620313e-02 -1.19183600e+00 9.03602481e-01 3.12310129e-01
1.33613098e+00 -4.60549831e-01 5.43376148e-01 3.64648372e-01
-1.27317941e+00 -1.57505665e-02 -5.62319875e-01 3.47748637e-01
-9.89586022e-03 4.48556960e-01 -8.09610903e-01 4.23244864e-01
6.53989255e-01 1.10474837e+00 -7.73366988e-01 7.96556234e-01
-2.56286770e-01 6.41165376e-01 -1.53043985e-01 -2.02692643e-01
4.53036547e-01 -3.03339995e-02 1.47760794e-01 1.50750160e+00
2.97592080e-04 -1.37666032e-01 2.87445486e-01 8.55267346e-01
-3.54653358e-01 1.43107489e-01 -4.03227240e-01 -4.67794269e-01
2.81514645e-01 1.12792242e+00 -6.14703357e-01 -5.40163755e-01
-7.17800677e-01 1.32483637e+00 8.56896341e-02 4.44315195e-01
-9.96066511e-01 -6.08971715e-01 -4.04031761e-02 -1.85809970e-01
4.89280105e-01 -2.58836806e-01 -2.74844825e-01 -1.49467063e+00
2.04100922e-01 -9.74684417e-01 1.34589151e-01 -1.09756434e+00
-1.00879729e+00 5.33203363e-01 -4.92080510e-01 -1.28348482e+00
1.41616806e-01 -9.98529673e-01 -3.75768095e-01 6.50303304e-01
-1.68793476e+00 -1.57802558e+00 -4.88119215e-01 1.01781273e+00
1.21088052e+00 -2.19015226e-01 1.71658263e-01 3.42416674e-01
-7.97258377e-01 8.94168854e-01 2.77772367e-01 7.23284602e-01
8.59603763e-01 -1.17650127e+00 2.65546262e-01 1.22990322e+00
3.62423956e-01 3.59388053e-01 2.74135292e-01 -7.27010190e-01
-1.72540903e+00 -1.29383481e+00 5.96835196e-01 -4.72206891e-01
5.70639789e-01 -7.26428986e-01 -9.02787268e-01 6.44243181e-01
3.83135855e-01 -1.23853743e-01 2.74044394e-01 -4.67683166e-01
-4.89172846e-01 -5.51538216e-03 -5.12210608e-01 8.20683658e-01
8.06211889e-01 -8.09158444e-01 -5.85071623e-01 6.10685527e-01
6.71641648e-01 -4.80945915e-01 -3.12114656e-01 1.92638025e-01
4.30837065e-01 -5.73074460e-01 9.79503632e-01 -2.69615322e-01
8.02728534e-01 -2.82599062e-01 -2.82434553e-01 -4.69358444e-01
2.41016567e-01 -2.68077791e-01 -1.09326653e-02 1.37236083e+00
2.54744560e-01 -1.36749774e-01 7.16855824e-01 2.03796804e-01
-2.54768431e-01 -4.37657058e-01 -5.33298492e-01 -5.57289004e-01
-5.21197058e-02 -6.02795720e-01 -5.73659241e-02 9.58374023e-01
-2.05104902e-01 5.36313117e-01 -8.99310529e-01 1.49998829e-01
7.23280668e-01 4.51628454e-02 8.47727180e-01 -7.66603410e-01
-1.99506447e-01 -3.02287221e-01 -2.24314183e-01 -1.68052995e+00
1.11884549e-02 -8.36821675e-01 2.66948134e-01 -1.56728339e+00
6.00270510e-01 -1.27533227e-01 1.06445782e-01 4.92353439e-01
-3.16847712e-01 3.05563509e-01 3.81575435e-01 3.37826610e-01
-9.46770191e-01 6.08538866e-01 1.47801435e+00 -4.80224371e-01
1.34725064e-01 -4.52728957e-01 -4.55754995e-01 7.39180267e-01
5.89432597e-01 -2.37719551e-01 -3.52276564e-01 -1.01669765e+00
1.16759306e-02 8.51757973e-02 4.45896477e-01 -8.41034651e-01
5.67139387e-01 -1.30820289e-01 9.73558545e-01 -1.09842050e+00
3.27699989e-01 -7.94154704e-01 -7.10023463e-01 3.68170857e-01
-5.48916101e-01 -3.97932857e-01 2.92106748e-01 7.81814396e-01
-1.09226018e-01 -3.41408163e-01 7.60921121e-01 -5.71601652e-02
-1.00337315e+00 3.14846069e-01 -4.61441666e-01 1.90236211e-01
9.34717953e-01 -5.10874152e-01 -7.83594549e-01 -2.84608960e-01
-2.74328232e-01 6.03290141e-01 3.95162791e-01 5.05394936e-01
1.05242157e+00 -9.71720278e-01 -7.61980951e-01 2.40779355e-01
5.04315019e-01 5.65679297e-02 3.42167795e-01 7.40546107e-01
-6.62991464e-01 5.84354818e-01 -1.82994202e-01 -9.87912476e-01
-1.40978110e+00 6.56107962e-01 1.59433588e-01 -1.49088472e-01
-6.92962170e-01 7.52820730e-01 5.19972026e-01 -4.75283572e-03
6.61607206e-01 -3.18509430e-01 -1.24997132e-01 -1.53424382e-01
6.57131970e-01 2.70972848e-02 -2.24916354e-01 -6.70982420e-01
-2.30611876e-01 9.93239343e-01 -5.13410211e-01 2.59771030e-02
1.15204608e+00 -3.95515680e-01 5.97813204e-02 1.59027413e-01
1.22676551e+00 4.67587379e-04 -1.43362570e+00 -6.36031806e-01
-1.04245357e-02 -4.81391400e-01 -3.98741923e-02 -7.49791145e-01
-1.08435225e+00 1.52943170e+00 4.10574555e-01 -2.89430618e-01
1.18774664e+00 8.45279545e-04 8.40226114e-01 7.97466338e-01
-8.29083025e-02 -1.30034089e+00 8.47171247e-01 5.69828212e-01
6.05200469e-01 -1.44461226e+00 -2.97559611e-02 -2.83676505e-01
-1.13174963e+00 1.30567443e+00 9.59559083e-01 5.17866462e-02
-1.73468783e-01 2.94583499e-01 -2.91591380e-02 5.98325171e-02
-7.74768591e-01 -3.95217896e-01 4.71252620e-01 7.24166572e-01
3.16955328e-01 -2.51377583e-01 2.83099383e-01 3.23298067e-01
2.66878903e-01 -3.22397828e-01 6.87969446e-01 6.94548726e-01
-6.57342017e-01 -7.53257513e-01 -2.56782383e-01 4.89044279e-01
-4.15216953e-01 -5.82241833e-01 -4.80824828e-01 7.57583082e-01
-6.64060488e-02 9.69776273e-01 1.88883126e-01 -2.52903372e-01
3.95081788e-02 -1.16233118e-01 3.84596944e-01 -6.22814357e-01
-1.51618049e-01 4.53688562e-01 -1.05112739e-01 -2.42353544e-01
-3.01882535e-01 -3.54272008e-01 -1.14670682e+00 -4.58164476e-02
-6.50218666e-01 -2.06901208e-01 5.85759401e-01 8.51646483e-01
-2.53524780e-02 5.64990044e-01 6.01896346e-01 -5.53953290e-01
-3.23985308e-01 -8.48022759e-01 -4.12220299e-01 3.29799682e-01
3.29143554e-01 -1.46530375e-01 -2.10001409e-01 7.42589653e-01] | [11.830986976623535, 2.2103750705718994] |
1de088e0-69fd-4e8e-8de3-956b9ea327a2 | sequence-to-sequence-models-for-cache | null | null | https://aclanthology.org/P18-1171 | https://aclanthology.org/P18-1171.pdf | Sequence-to-sequence Models for Cache Transition Systems | In this paper, we present a sequence-to-sequence based approach for mapping natural language sentences to AMR semantic graphs. We transform the sequence to graph mapping problem to a word sequence to transition action sequence problem using a special transition system called a cache transition system. To address the sparsity issue of neural AMR parsing, we feed feature embeddings from the transition state to provide relevant local information for each decoder state. We present a monotonic hard attention model for the transition framework to handle the strictly left-to-right alignment between each transition state and the current buffer input focus. We evaluate our neural transition model on the AMR parsing task, and our parser outperforms other sequence-to-sequence approaches and achieves competitive results in comparison with the best-performing models. | ['Xiaochang Peng', 'Giorgio Satta', 'Linfeng Song', 'Daniel Gildea'] | 2018-07-01 | null | null | null | acl-2018-7 | ['hard-attention'] | ['methodology'] | [ 7.01622188e-01 3.94219279e-01 -4.97362465e-01 -5.67156017e-01
-9.66451883e-01 -4.87581879e-01 4.37287867e-01 1.92175925e-01
-3.51229638e-01 1.97345853e-01 7.24820614e-01 -8.48566413e-01
5.71670115e-01 -8.36041689e-01 -9.73532498e-01 6.74771219e-02
2.42753118e-01 5.77863336e-01 2.34523252e-01 -2.16899246e-01
7.28388131e-02 2.33053371e-01 -7.66449153e-01 6.95946574e-01
3.28513652e-01 6.08741283e-01 4.97330815e-01 1.19688118e+00
-8.32000494e-01 1.29183328e+00 -3.31566751e-01 -5.14630139e-01
1.62663355e-01 -6.44569457e-01 -1.09911191e+00 -1.69436812e-01
3.80719870e-01 -2.62762487e-01 -8.00753772e-01 1.11926568e+00
1.33398786e-01 1.78217411e-01 3.11744153e-01 -7.95771956e-01
-1.11996794e+00 1.26031053e+00 -4.41749126e-01 6.06414199e-01
6.15159571e-01 8.38499740e-02 1.55872762e+00 -7.75031269e-01
7.43276656e-01 1.50527263e+00 2.39011347e-01 8.87538612e-01
-1.21963227e+00 -2.28849143e-01 6.35970592e-01 7.09015355e-02
-8.96960497e-01 -5.17491281e-01 5.16977906e-01 -3.77821326e-01
1.94630802e+00 -7.82983899e-02 4.74499375e-01 1.09611785e+00
7.16575682e-01 7.74302959e-01 5.18492460e-01 -4.53114331e-01
2.54135519e-01 -6.49166703e-01 7.74242818e-01 9.65479314e-01
-1.51075453e-01 -4.28277999e-01 -5.64885318e-01 -1.87702104e-01
7.19580591e-01 -2.78220892e-01 2.06968918e-01 6.22578524e-02
-8.52794945e-01 1.01301336e+00 3.18085074e-01 1.74059972e-01
-2.59184182e-01 8.89156163e-01 5.84560573e-01 5.03173709e-01
3.87772292e-01 2.41529822e-01 -5.79421520e-01 -3.42087686e-01
-4.05415028e-01 -4.98392209e-02 9.70045507e-01 1.08647203e+00
7.92620540e-01 3.40487719e-01 -3.70562971e-01 5.32200575e-01
6.26021564e-01 2.20597357e-01 5.36061227e-01 -8.93950343e-01
8.55630696e-01 4.79200631e-01 -3.22046608e-01 -6.01904154e-01
-2.98348278e-01 -2.19172180e-01 -4.60874319e-01 -4.93386507e-01
1.50403038e-01 8.52789655e-02 -1.12300062e+00 1.83102596e+00
-2.71405652e-02 2.18868077e-01 3.14094931e-01 6.68504596e-01
7.85358131e-01 1.09319341e+00 2.99941748e-01 -8.03686008e-02
1.41201198e+00 -1.21451569e+00 -8.50380123e-01 -1.02373588e+00
1.12744963e+00 -3.85658950e-01 1.30708504e+00 -2.39963010e-01
-1.35596097e+00 -4.39285070e-01 -8.06886911e-01 -4.61019963e-01
-2.48011737e-03 -3.23725380e-02 4.12215441e-01 2.68607527e-01
-1.44995522e+00 2.92361885e-01 -1.12021089e+00 -2.64750510e-01
6.21898063e-02 2.12084949e-01 -4.43073988e-01 -1.83022171e-01
-1.35179663e+00 9.39170361e-01 4.70972955e-01 6.00766018e-02
-8.49721372e-01 -4.70330924e-01 -1.37446880e+00 3.28297138e-01
3.92262131e-01 -8.14269125e-01 1.67703187e+00 -8.64356935e-01
-1.70312285e+00 7.56508946e-01 -5.89741647e-01 -1.01414478e+00
-1.64937258e-01 -3.55774879e-01 -2.84529805e-01 6.81403279e-02
-1.09007716e-01 5.29193103e-01 7.00983286e-01 -6.31809711e-01
-2.78397799e-01 -1.67820677e-01 6.08268082e-02 2.28774473e-01
8.12778994e-02 4.49144095e-01 -7.28752673e-01 -5.76209843e-01
1.24946117e-01 -9.53791022e-01 -5.80250680e-01 -6.89777851e-01
-3.83657068e-01 -2.72219539e-01 3.68812919e-01 -8.75005007e-01
1.47015560e+00 -1.94262302e+00 4.06607538e-01 -1.29053846e-01
1.89624622e-01 1.15473755e-01 -5.18348813e-01 5.67144156e-01
-2.68909723e-01 2.57072240e-01 -3.97177339e-01 -3.05591464e-01
5.76519631e-02 3.22113097e-01 -5.53930104e-01 2.39746436e-01
5.16093254e-01 1.41981423e+00 -8.31107199e-01 -3.02595347e-01
-5.78868948e-02 2.84641802e-01 -6.45006895e-01 5.56037068e-01
-6.70339286e-01 1.81810424e-01 -5.30867517e-01 2.44781941e-01
8.15049931e-02 -3.98918092e-01 5.50407231e-01 1.76445827e-01
3.68635237e-01 1.08150792e+00 -5.87897241e-01 2.02328348e+00
-7.13400543e-01 5.26466489e-01 -1.63075164e-01 -8.03507507e-01
1.09290981e+00 2.10723728e-01 5.54185435e-02 -8.59659612e-01
-5.57175167e-02 -7.38257766e-02 2.78477877e-01 -1.21701084e-01
8.41619432e-01 6.85216263e-02 -6.47833049e-01 5.13580739e-01
1.58639580e-01 1.84090331e-01 -6.30477350e-03 6.04215920e-01
1.62163281e+00 2.86019444e-01 6.22254729e-01 -2.11384445e-01
4.45670784e-01 -6.01853319e-02 4.79774237e-01 8.27088475e-01
-3.51460315e-02 4.24847871e-01 8.46282184e-01 -5.42439580e-01
-1.09266782e+00 -1.04798102e+00 5.40314794e-01 1.46217215e+00
-3.29634309e-01 -8.08979928e-01 -9.08042431e-01 -7.75258958e-01
-2.51578301e-01 9.60030735e-01 -5.01940131e-01 -4.12341595e-01
-1.31473696e+00 -2.55130500e-01 5.71336627e-01 1.07828259e+00
1.15947295e-02 -1.26332700e+00 -3.63212347e-01 6.22832179e-01
-7.34807029e-02 -1.52788174e+00 -9.69484031e-01 4.85144228e-01
-7.55790949e-01 -4.57040012e-01 -8.73904899e-02 -8.88422430e-01
3.97468120e-01 -2.67726511e-01 1.44062960e+00 -2.10303254e-02
1.31847793e-02 3.32950413e-01 -3.31451774e-01 1.84472576e-01
-1.14918733e+00 4.18034464e-01 -3.15803200e-01 -3.32949758e-01
4.57856864e-01 -1.62879035e-01 -1.03500269e-01 -1.75643310e-01
-7.38904178e-01 1.89319134e-01 4.03664231e-01 6.61392272e-01
6.27358973e-01 -7.18108356e-01 4.63662535e-01 -1.17710483e+00
6.42192066e-01 -3.33901286e-01 -5.96592546e-01 1.62544534e-01
-1.80780604e-01 5.52782953e-01 1.03055573e+00 -1.06309414e-01
-9.50319648e-01 3.94102812e-01 -5.52733183e-01 -3.65383267e-01
-1.78192511e-01 6.35564387e-01 -5.18077493e-01 5.58179677e-01
4.13474530e-01 4.07869041e-01 -2.93994248e-01 -4.06820416e-01
9.15519655e-01 5.22315443e-01 7.39353001e-01 -7.44839191e-01
5.99616349e-01 -1.98017597e-01 -3.06643881e-02 -5.12605309e-01
-7.83810854e-01 -2.41622180e-01 -4.99497920e-01 3.02508771e-01
1.31479335e+00 -1.06526637e+00 -2.36433074e-01 5.10874279e-02
-1.53811526e+00 -8.01080942e-01 -3.03146422e-01 6.47128597e-02
-7.71162689e-01 4.38675821e-01 -1.30814600e+00 -6.40077472e-01
-4.81475025e-01 -1.29849219e+00 1.11048746e+00 -1.51562259e-01
-1.71207234e-01 -1.01894677e+00 1.84960365e-01 2.94719934e-02
3.30035478e-01 -2.64125913e-01 1.28405869e+00 -1.06379747e+00
-7.87686586e-01 3.69172767e-02 -1.86561882e-01 1.79679915e-01
-1.15121566e-01 -3.57798845e-01 -5.97461998e-01 -2.67255783e-01
-1.70054927e-01 -1.35044187e-01 1.14712584e+00 3.56619895e-01
1.08096671e+00 -4.07197893e-01 -3.27114791e-01 5.54677308e-01
1.27806163e+00 2.63486266e-01 6.91753805e-01 9.46640223e-02
1.09669590e+00 2.41096526e-01 3.59269500e-01 3.78002599e-02
5.60929239e-01 6.56654596e-01 3.81869435e-01 1.36744484e-01
-4.32888061e-01 -6.84535682e-01 9.18589056e-01 1.16568756e+00
7.16686785e-01 -6.59739852e-01 -1.03864908e+00 3.72993469e-01
-1.89820135e+00 -4.13165808e-01 -1.24806933e-01 1.95707357e+00
6.57659411e-01 3.44399691e-01 -1.63307518e-01 -5.98330855e-01
8.31405938e-01 5.30756056e-01 -3.37420732e-01 -1.35765743e+00
2.63373494e-01 3.62950623e-01 7.18023360e-01 9.91143942e-01
-7.86031663e-01 1.67819667e+00 6.97048950e+00 3.77587765e-01
-7.39758968e-01 2.50765204e-01 5.00471652e-01 2.72648335e-01
-6.74271226e-01 2.59720445e-01 -1.20135605e+00 2.17560709e-01
1.85748291e+00 -5.58236875e-02 7.70392299e-01 7.10363626e-01
-1.43096358e-01 4.30767089e-01 -1.51828873e+00 6.25986516e-01
2.82262843e-02 -1.43871772e+00 3.07519615e-01 -2.12707177e-01
2.42388755e-01 2.06583709e-01 -1.66216165e-01 3.95012558e-01
9.93620455e-01 -1.24878132e+00 6.14849091e-01 2.57791132e-01
8.30867410e-01 -7.29217172e-01 3.62752914e-01 5.39699495e-02
-1.74726260e+00 -3.92033793e-02 -5.00465631e-01 -1.48317233e-01
5.07233739e-01 3.61421764e-01 -9.22449052e-01 4.87955123e-01
3.59026790e-01 7.22924113e-01 -4.55793530e-01 1.09886147e-01
-5.05903900e-01 8.08037281e-01 1.19319931e-01 -2.54163183e-02
4.39529657e-01 -1.95411205e-01 6.51278079e-01 1.58940113e+00
9.92806479e-02 -8.43107104e-02 3.57585192e-01 9.33679342e-01
-4.90686268e-01 4.38829064e-02 -8.59207869e-01 -4.01794314e-01
4.12127107e-01 9.32510197e-01 -6.04722857e-01 -4.89137053e-01
-8.37214828e-01 1.20563138e+00 8.26154113e-01 2.71748424e-01
-7.45530367e-01 -2.55646497e-01 7.64373004e-01 -1.19320013e-01
5.67871571e-01 -4.62432712e-01 -3.05830427e-02 -1.25080609e+00
-7.64101073e-02 -9.28860903e-01 5.70910513e-01 -7.42566884e-01
-1.05078554e+00 7.70894051e-01 -1.31123751e-01 -2.92256683e-01
-6.47597671e-01 -5.30413866e-01 -6.40485108e-01 1.04964685e+00
-1.32991302e+00 -9.21923041e-01 4.56308961e-01 2.31757522e-01
8.83127570e-01 -2.28359446e-01 9.63678122e-01 -1.03096679e-01
-5.94036102e-01 6.25571668e-01 -4.20605630e-01 5.02496183e-01
1.96738496e-01 -1.42853677e+00 1.85655665e+00 1.32359421e+00
4.13268715e-01 8.14835966e-01 4.30116057e-01 -8.15930605e-01
-1.89991152e+00 -1.43358648e+00 9.87457097e-01 -6.57259464e-01
9.36209381e-01 -9.27051425e-01 -1.16032040e+00 1.55979538e+00
4.39410239e-01 3.83385122e-02 5.48932314e-01 6.86229840e-02
-5.84285140e-01 3.63135040e-01 -4.87690300e-01 6.98250473e-01
1.30460095e+00 -8.73178720e-01 -9.96763766e-01 -5.04864492e-02
1.48279238e+00 -7.28542507e-01 -7.59281158e-01 -9.25469026e-02
2.73789242e-02 -1.89260572e-01 7.43451178e-01 -1.12727571e+00
6.36378467e-01 -2.54806429e-01 -4.55241382e-01 -1.28163481e+00
-6.48820877e-01 -7.82669306e-01 -5.37534535e-01 1.01866961e+00
7.75394142e-01 -5.04643977e-01 7.50719368e-01 3.55418414e-01
-6.67574108e-01 -4.95640963e-01 -9.87476826e-01 -6.31146491e-01
2.97169477e-01 -4.61447775e-01 4.49313581e-01 5.80132723e-01
-1.28439171e-02 9.39525127e-01 -3.43425870e-01 2.30199769e-01
1.41377658e-01 2.27683023e-01 4.87395346e-01 -7.51546741e-01
-8.10730278e-01 -1.95185959e-01 -3.06853265e-01 -1.49618649e+00
8.06272507e-01 -1.52343726e+00 4.00761664e-01 -2.00740910e+00
3.03802788e-01 3.86552632e-01 -3.12178671e-01 6.20981276e-01
-2.56447375e-01 -1.95018515e-01 3.05760890e-01 -2.17694551e-01
-6.81150675e-01 2.73829401e-01 8.94298673e-01 -1.42822534e-01
-1.79962203e-01 -5.04190028e-01 -7.74949610e-01 3.21281374e-01
7.21866012e-01 -8.02632034e-01 -3.69622916e-01 -8.56743932e-01
4.77978230e-01 6.27302408e-01 -1.93298563e-01 -5.74570656e-01
1.90120101e-01 -1.08935289e-01 -1.59947932e-01 -3.59508187e-01
1.77002981e-01 -4.15271312e-01 -9.78872031e-02 4.50069666e-01
-9.51479554e-01 6.30466461e-01 1.49426550e-01 7.84206688e-01
7.89743066e-02 -2.31777459e-01 6.98318779e-01 -2.63221890e-01
-8.82178724e-01 3.74943614e-01 -7.85573304e-01 5.03002226e-01
5.54202735e-01 1.37670606e-01 -5.35981417e-01 -5.06829739e-01
-9.09280121e-01 1.07283100e-01 4.88212943e-01 7.26083994e-01
6.80581152e-01 -1.05045223e+00 -6.35642111e-01 2.01542854e-01
9.19319466e-02 -1.86170369e-01 -9.49376449e-02 4.24986482e-01
-4.47248876e-01 4.86327469e-01 -7.22613484e-02 -2.96091199e-01
-1.02980435e+00 7.71920025e-01 3.69120985e-01 -5.52124500e-01
-1.09930778e+00 7.18149006e-01 4.11485463e-01 -4.63036031e-01
4.20468412e-02 -6.87229335e-01 5.62677309e-02 -5.70145667e-01
6.05912626e-01 -1.01865493e-01 1.04637779e-01 -6.91399217e-01
-6.43923998e-01 2.62784392e-01 -4.04671669e-01 -1.88269049e-01
9.87729669e-01 -2.60033943e-02 -1.43060967e-01 5.00628650e-01
1.33167505e+00 -7.60173723e-02 -1.09746981e+00 -3.93427938e-01
3.58580709e-01 -2.14154292e-02 -1.20349163e-02 -4.83565301e-01
-9.47979689e-01 9.57063138e-01 2.30353311e-01 4.41854857e-02
6.05877280e-01 3.39510351e-01 1.04001832e+00 6.11602306e-01
-3.22223939e-02 -8.38096440e-01 1.68868139e-01 1.19209933e+00
3.95399034e-01 -7.56789267e-01 -5.41506112e-01 -3.73631150e-01
-8.26707363e-01 1.13837302e+00 7.02973485e-01 -3.34482342e-01
2.50370204e-01 8.18872571e-01 -3.88512909e-02 6.32203743e-02
-1.25891697e+00 -2.52561182e-01 3.49756517e-02 5.79770625e-01
6.87816560e-01 1.31471872e-01 -2.42904171e-01 7.70621061e-01
-9.22961682e-02 -1.97934449e-01 8.15850079e-01 9.28833485e-01
-6.51588619e-01 -1.51227427e+00 2.11427778e-01 4.50883567e-01
-7.09130287e-01 -6.10596478e-01 -4.84423459e-01 5.42289793e-01
-9.00480568e-01 8.16096425e-01 4.57424730e-01 -3.41256827e-01
2.68927425e-01 5.63708246e-01 4.43403453e-01 -1.25639462e+00
-7.47606158e-01 1.80709496e-01 5.32990873e-01 -8.39061260e-01
2.27283895e-01 -4.17548597e-01 -1.84056973e+00 -2.61633042e-02
6.77652434e-02 5.97902983e-02 4.37937379e-01 8.53223503e-01
4.92992222e-01 1.01861155e+00 3.67874831e-01 -1.41158313e-01
-5.20487249e-01 -7.44634390e-01 -1.36134222e-01 2.21091628e-01
1.35848835e-01 1.30478680e-01 -1.30245797e-02 -1.11092821e-01] | [10.438475608825684, 9.082802772521973] |
b7ca4a80-1865-4212-b9ae-704416e02e73 | sketch-based-3d-shape-retrieval-using | 1504.03504 | null | http://arxiv.org/abs/1504.03504v1 | http://arxiv.org/pdf/1504.03504v1.pdf | Sketch-based 3D Shape Retrieval using Convolutional Neural Networks | Retrieving 3D models from 2D human sketches has received considerable
attention in the areas of graphics, image retrieval, and computer vision.
Almost always in state of the art approaches a large amount of "best views" are
computed for 3D models, with the hope that the query sketch matches one of
these 2D projections of 3D models using predefined features.
We argue that this two stage approach (view selection -- matching) is
pragmatic but also problematic because the "best views" are subjective and
ambiguous, which makes the matching inputs obscure. This imprecise nature of
matching further makes it challenging to choose features manually. Instead of
relying on the elusive concept of "best views" and the hand-crafted features,
we propose to define our views using a minimalism approach and learn features
for both sketches and views. Specifically, we drastically reduce the number of
views to only two predefined directions for the whole dataset. Then, we learn
two Siamese Convolutional Neural Networks (CNNs), one for the views and one for
the sketches. The loss function is defined on the within-domain as well as the
cross-domain similarities. Our experiments on three benchmark datasets
demonstrate that our method is significantly better than state of the art
approaches, and outperforms them in all conventional metrics. | ['Yi Li', 'Le Kang', 'Fang Wang'] | 2015-04-14 | sketch-based-3d-shape-retrieval-using-1 | http://openaccess.thecvf.com/content_cvpr_2015/html/Wang_Sketch-Based_3D_Shape_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Wang_Sketch-Based_3D_Shape_2015_CVPR_paper.pdf | cvpr-2015-6 | ['3d-shape-retrieval'] | ['computer-vision'] | [-5.90439048e-03 -3.50795865e-01 -1.99172333e-01 -4.24633086e-01
-7.57974625e-01 -8.56820941e-01 8.32237005e-01 -3.05647820e-01
-2.04455480e-01 5.82268760e-02 6.13253228e-02 -4.31298763e-02
-2.05033809e-01 -7.14784026e-01 -4.79949534e-01 -5.05706251e-01
1.85297921e-01 6.51057005e-01 2.41128162e-01 -1.22626528e-01
5.98021805e-01 8.29636335e-01 -1.62409246e+00 2.94822693e-01
4.34581637e-01 1.15458751e+00 5.04261889e-02 3.62935096e-01
-3.56052995e-01 -4.10843492e-02 -4.14325178e-01 -7.46556103e-01
6.84037626e-01 -1.34795621e-01 -4.89441127e-01 3.99126142e-01
8.59406352e-01 -4.11467284e-01 -4.37194049e-01 1.09687245e+00
4.36063766e-01 -1.24990242e-02 5.75804472e-01 -1.17558038e+00
-5.86191833e-01 -3.46095860e-02 -7.23456800e-01 -3.04294258e-01
4.89051253e-01 4.97550964e-02 1.26666915e+00 -1.41614151e+00
8.85936677e-01 1.37694371e+00 5.81272602e-01 4.33939695e-01
-1.20558739e+00 -5.06888270e-01 3.74089509e-01 -8.03419799e-02
-1.46177959e+00 -4.48182642e-01 1.19805110e+00 -5.18784404e-01
5.06845891e-01 3.77139777e-01 7.26578355e-01 9.67189670e-01
-8.46701786e-02 8.47419739e-01 9.17077482e-01 -3.70774865e-01
2.58734941e-01 2.07411304e-01 -2.63782263e-01 7.84659743e-01
2.60315612e-02 4.51811850e-02 -6.30216062e-01 -3.67589027e-01
1.12350488e+00 3.80834460e-01 -2.25936547e-01 -1.17091036e+00
-1.30429804e+00 7.59335101e-01 1.07301556e-01 3.04080606e-01
-3.84060264e-01 -5.70839606e-02 2.50203341e-01 4.49564934e-01
4.69594270e-01 4.86375958e-01 -2.57229954e-01 5.80393635e-02
-9.79408026e-01 3.28321785e-01 7.71719575e-01 1.08049250e+00
7.69031227e-01 -4.54795778e-01 1.59766227e-01 9.44455564e-01
1.46231711e-01 4.68253851e-01 1.49990534e-02 -9.50888276e-01
6.32932782e-01 8.11067939e-01 3.17344666e-01 -1.30630708e+00
1.42659992e-01 -9.17066112e-02 -8.10154796e-01 5.71511924e-01
5.38932741e-01 3.98930579e-01 -9.09466624e-01 1.52473295e+00
2.56484032e-01 -3.92994016e-01 -4.00025427e-01 1.21143472e+00
5.62379539e-01 2.39264593e-01 -4.60375518e-01 3.35945070e-01
1.04968357e+00 -7.30459213e-01 -2.20719829e-01 -1.21134005e-01
3.78274638e-03 -9.43166971e-01 1.19126141e+00 5.76180458e-01
-1.25393009e+00 -5.86610436e-01 -1.19043422e+00 -2.26677343e-01
-4.73779947e-01 3.91823798e-01 4.94637430e-01 5.63108146e-01
-8.97027314e-01 7.10610867e-01 -5.00863731e-01 -3.74982327e-01
3.94894689e-01 3.51667911e-01 -7.10124969e-01 -2.73321539e-01
-7.88838148e-01 7.48071790e-01 -1.56766832e-01 1.52753472e-01
-6.72635794e-01 -5.15348375e-01 -7.05467045e-01 5.47968037e-02
4.69907910e-01 -7.00707793e-01 7.01914251e-01 -8.59638810e-01
-1.35849845e+00 1.19669914e+00 -6.99097887e-02 2.58009166e-01
8.85860920e-01 -1.78310633e-01 -7.62079135e-02 1.96805686e-01
-1.93697698e-02 6.02136135e-01 1.14584088e+00 -1.60827386e+00
-4.75632459e-01 -6.44164681e-01 2.98645556e-01 2.64328778e-01
-9.22443420e-02 -2.86099255e-01 -9.72992420e-01 -5.35045624e-01
6.49519682e-01 -9.69315529e-01 -1.49941012e-01 8.56097162e-01
-2.21082360e-01 -1.32671922e-01 7.53148615e-01 -3.92909467e-01
8.76080692e-01 -2.21084714e+00 3.22416753e-01 4.91697282e-01
3.96914333e-01 4.10358794e-02 -1.76825047e-01 4.59200293e-01
1.10028103e-01 8.73252451e-02 -7.38293827e-02 -4.82696533e-01
2.05350012e-01 1.31843895e-01 -4.43692058e-01 5.19378245e-01
2.00933158e-01 6.38652086e-01 -8.87458324e-01 -2.72467345e-01
4.15846348e-01 5.38135707e-01 -4.38394070e-01 3.50968868e-01
-5.50100543e-02 2.16415152e-01 -6.45014465e-01 5.88458121e-01
9.82820511e-01 -3.16850394e-01 2.47836366e-01 -2.08606333e-01
-9.47621763e-02 2.75810063e-01 -1.54952645e+00 2.06142473e+00
-4.19223309e-01 3.01228315e-01 -5.62388375e-02 -8.28569055e-01
1.15865815e+00 2.19749883e-01 6.47833824e-01 -5.22582829e-01
-2.72010624e-01 3.74390781e-01 -4.98449296e-01 -3.58558238e-01
2.79631466e-01 -5.61825559e-02 -2.77506839e-02 6.20554507e-01
4.38535474e-02 -4.83161449e-01 -8.65372047e-02 7.54283220e-02
8.26786816e-01 2.58906573e-01 2.87850171e-01 -5.77508993e-02
5.73973596e-01 -2.58283794e-01 3.09161305e-01 6.33233666e-01
4.77826707e-02 1.10631835e+00 6.98979318e-01 -8.68761659e-01
-1.21875679e+00 -1.22473049e+00 1.93731040e-01 8.07993829e-01
3.80700141e-01 -4.26583320e-01 -4.05686826e-01 -1.11177552e+00
3.53460610e-01 2.53564805e-01 -7.69852221e-01 7.06812814e-02
-5.54776132e-01 1.07013039e-01 1.83695778e-01 4.68248695e-01
1.55517474e-01 -7.37999558e-01 -6.80462122e-01 -8.13219398e-02
2.61922270e-01 -8.32321167e-01 -8.28499436e-01 -2.95343883e-02
-7.59420216e-01 -1.08222520e+00 -9.86242414e-01 -5.55107474e-01
9.21061218e-01 6.76401079e-01 1.24514985e+00 2.40399197e-01
-1.02501154e-01 3.59180033e-01 -1.44616321e-01 -1.72095031e-01
-4.47546802e-02 -8.44483301e-02 -9.07462463e-02 1.95718423e-01
3.40602010e-01 -7.48163223e-01 -6.91978812e-01 4.60804284e-01
-8.30729425e-01 6.73420429e-02 7.08944857e-01 9.22083914e-01
6.76014125e-01 -3.54395539e-01 2.36101583e-01 -7.62828410e-01
4.11140889e-01 2.03616321e-02 -6.70110703e-01 5.61344981e-01
-5.47618389e-01 1.85729951e-01 6.20425642e-01 -5.45856535e-01
-6.58958435e-01 2.32138366e-01 1.33386478e-01 -1.12559962e+00
-1.48700491e-01 8.31888467e-02 -4.43035364e-01 -1.18246123e-01
3.91959786e-01 2.37729073e-01 1.35189086e-01 -6.45687938e-01
3.75955254e-01 3.74917060e-01 2.57438481e-01 -6.49187803e-01
8.22345138e-01 7.84302294e-01 -2.79157888e-02 -6.67764187e-01
-6.65630698e-01 -5.15079677e-01 -8.87503028e-01 -1.73737884e-01
4.99462843e-01 -5.67753196e-01 -5.63353062e-01 1.43369868e-01
-1.17398000e+00 2.79854596e-01 -2.72291958e-01 2.92665422e-01
-6.77491069e-01 4.51227635e-01 -7.19821751e-02 -7.44820833e-01
-2.35887378e-01 -1.44774556e+00 1.54750669e+00 -4.70274612e-02
-2.27351606e-01 -7.22648561e-01 -1.47813011e-03 5.52781820e-02
3.24059874e-01 1.39108956e-01 1.13152385e+00 -5.18703938e-01
-7.70296514e-01 -5.05236804e-01 -3.75609219e-01 2.00032398e-01
1.40897036e-01 -4.61676829e-02 -1.06564426e+00 -3.06976497e-01
-5.98126501e-02 -2.39106834e-01 7.05417454e-01 2.55658589e-02
1.16714430e+00 -4.96005267e-03 -3.74748707e-01 4.77912754e-01
1.51415622e+00 2.00368240e-01 4.69564587e-01 -8.70755538e-02
5.27259111e-01 6.99666917e-01 3.05051357e-01 5.37546754e-01
1.85514927e-01 8.05314362e-01 3.76367837e-01 1.07798852e-01
7.16957450e-02 -5.36474049e-01 -6.41397014e-02 6.39299154e-01
-1.67453047e-02 -1.26546860e-01 -7.12652206e-01 5.98978698e-01
-1.80581939e+00 -8.31192970e-01 5.04248857e-01 2.54744697e+00
4.21605885e-01 2.01566845e-01 3.94298024e-02 -4.68038358e-02
6.47703171e-01 4.35718775e-01 -5.25618911e-01 -1.77446976e-01
2.58767307e-02 2.65218288e-01 7.52952099e-02 2.61041373e-01
-1.14732099e+00 5.35319388e-01 5.66128683e+00 5.90758383e-01
-1.28850639e+00 -3.84821653e-01 4.34121579e-01 -1.45567417e-01
-5.03956735e-01 1.82669386e-01 -5.80145061e-01 3.08087885e-01
-5.19982651e-02 1.87478572e-01 5.08404493e-01 8.12735736e-01
-2.26032287e-01 5.23910224e-02 -1.54472399e+00 1.25728858e+00
1.16128355e-01 -1.35511100e+00 2.88525790e-01 1.47182435e-01
5.98798037e-01 -2.62236208e-01 1.53425395e-01 -5.44089675e-02
-6.70586973e-02 -9.13694263e-01 9.10045445e-01 8.14898670e-01
9.78897929e-01 -7.32704043e-01 4.35901254e-01 3.50993544e-01
-1.32596886e+00 2.04720393e-01 -4.58451897e-01 2.62992382e-01
-1.28148034e-01 3.75147372e-01 -3.99882376e-01 5.30389309e-01
5.68774402e-01 6.75448775e-01 -4.28121120e-01 9.17738020e-01
1.59689397e-01 -1.47601604e-01 -4.33764845e-01 -1.74191557e-02
2.93329865e-01 -3.43524754e-01 5.41440666e-01 7.86203146e-01
3.94436985e-01 -4.76508066e-02 2.94501394e-01 1.10165799e+00
-9.98198539e-02 -2.98854429e-02 -9.45329964e-01 2.06772592e-02
5.11682391e-01 1.19181418e+00 -6.55467153e-01 -3.24921876e-01
-5.20788252e-01 1.07599580e+00 3.72946292e-01 2.21874595e-01
-3.39920223e-01 -4.80720580e-01 6.20785534e-01 1.33061394e-01
5.75145900e-01 -2.87453294e-01 -4.48668599e-01 -1.25952208e+00
5.73355198e-01 -8.12061191e-01 3.03903610e-01 -4.91353571e-01
-1.65600955e+00 5.40164113e-01 -1.05490930e-01 -1.59753656e+00
-1.74752221e-01 -7.05256701e-01 -6.17572367e-01 9.86696005e-01
-1.26321924e+00 -1.15107512e+00 -2.53470600e-01 3.74753773e-01
6.91971838e-01 -5.43174781e-02 7.50097752e-01 2.80357480e-01
-4.52083424e-02 5.95745206e-01 5.37603982e-02 9.40025970e-02
7.86562979e-01 -1.16900408e+00 7.59005427e-01 3.81328881e-01
5.71226060e-01 8.08232963e-01 3.50408524e-01 -3.30779076e-01
-1.79377365e+00 -4.57561702e-01 1.09456754e+00 -4.84363079e-01
3.28909308e-01 -6.19103312e-01 -8.52343738e-01 2.73961455e-01
1.31543623e-02 2.05505893e-01 3.19606066e-01 3.31597519e-03
-6.71711326e-01 -5.48946820e-02 -1.04127073e+00 6.55030370e-01
1.16661727e+00 -7.31882274e-01 -4.43844318e-01 5.03108986e-02
9.30506662e-02 -3.91279310e-01 -8.18515718e-01 2.85533786e-01
1.16608059e+00 -1.16744614e+00 1.24670100e+00 -6.83622301e-01
3.87965947e-01 -2.47309580e-01 -3.83434534e-01 -1.02485180e+00
-1.22827962e-01 -4.37010229e-01 4.93377782e-02 8.67588162e-01
2.35454246e-01 -2.82106042e-01 9.71797466e-01 6.44668460e-01
2.24009827e-01 -1.16932702e+00 -9.19180393e-01 -6.38589025e-01
-8.51730630e-02 -2.65152574e-01 7.91696310e-01 9.23603714e-01
-3.29701632e-01 2.87476629e-01 -2.85244823e-01 5.99116366e-03
5.07099092e-01 7.61548221e-01 1.06075394e+00 -1.46908569e+00
-9.31235701e-02 -7.03017175e-01 -4.17863220e-01 -1.31840205e+00
3.85046341e-02 -5.91509223e-01 -1.30334780e-01 -1.26021039e+00
1.01954311e-01 -6.77973330e-01 -1.56795755e-01 1.79036647e-01
2.39393637e-02 2.38326211e-02 3.07014227e-01 2.68346190e-01
-4.49985355e-01 4.45941836e-01 1.38040006e+00 -2.85583287e-01
-5.45033067e-02 1.81594998e-01 -3.79480779e-01 7.13429749e-01
3.21198970e-01 -2.66325086e-01 -4.03988540e-01 -6.38741076e-01
1.84741363e-01 2.21772119e-01 5.53892016e-01 -7.43742704e-01
2.22621977e-01 -2.43788585e-01 6.33102775e-01 -8.44484985e-01
7.01607347e-01 -1.04783189e+00 1.51046664e-01 -9.69066843e-02
-4.13592130e-01 2.65298903e-01 -3.03175479e-01 5.74698687e-01
-1.31561011e-01 -2.34607548e-01 6.78451657e-01 -5.31483829e-01
-4.51587588e-01 5.81532657e-01 3.20816010e-01 -6.18770607e-02
6.32225037e-01 -4.93480504e-01 1.24738283e-01 -4.65685189e-01
-5.63286722e-01 -4.05926108e-02 8.95583630e-01 5.91504693e-01
9.24551189e-01 -1.50585163e+00 -4.69037473e-01 5.77402472e-01
3.78013164e-01 5.17053716e-02 7.58440420e-02 5.22721648e-01
-2.97217727e-01 3.74517113e-01 -2.21481249e-01 -6.40934169e-01
-1.02649832e+00 6.89263523e-01 1.93278134e-01 -2.38963649e-01
-7.93913126e-01 6.76324904e-01 3.92924547e-01 -6.28646076e-01
4.17111516e-01 -1.57577962e-01 1.19314544e-01 -9.17343795e-03
2.72517085e-01 1.63605437e-01 5.78260273e-02 -4.71174121e-01
-3.28678548e-01 8.32798243e-01 -1.72331408e-01 -1.37239546e-01
1.32003760e+00 2.06238423e-02 7.44206384e-02 4.19189870e-01
1.42811096e+00 9.80423689e-02 -1.30237103e+00 -3.69457096e-01
1.38875902e-01 -1.01821971e+00 -2.83267736e-01 -5.54596245e-01
-1.16155541e+00 1.23656511e+00 4.23202515e-01 2.51116604e-01
9.39676583e-01 1.24107683e-02 7.90846944e-01 3.89966637e-01
5.51322520e-01 -9.08184230e-01 1.41255125e-01 3.13979059e-01
1.14180207e+00 -1.27672529e+00 1.05487652e-01 -1.99240774e-01
-5.76502323e-01 1.32206202e+00 5.35293519e-01 -4.31602567e-01
6.41358852e-01 4.70072515e-02 -4.83346581e-02 -4.47207451e-01
-6.78648829e-01 4.10974436e-02 6.67293131e-01 3.01523834e-01
2.61425942e-01 -6.34090081e-02 -4.08675335e-02 3.48047018e-01
3.50035690e-02 -1.59569815e-01 -5.80006614e-02 9.75848198e-01
-2.00230345e-01 -1.27010107e+00 -3.23529005e-01 4.91179019e-01
-7.62310401e-02 1.97336704e-01 -8.03291917e-01 8.86671364e-01
6.58784881e-02 3.69992942e-01 8.25869069e-02 -3.66880983e-01
6.18354738e-01 1.36205345e-01 6.34736657e-01 -3.51502120e-01
-4.09976393e-01 7.49984905e-02 -2.60875285e-01 -5.45725644e-01
-2.99895883e-01 -5.19448757e-01 -5.50469995e-01 -2.19198078e-01
-3.68532240e-01 -6.38742745e-02 5.71244180e-01 7.16121495e-01
4.73572552e-01 -2.75105536e-01 9.99296308e-01 -1.08184183e+00
-8.96996379e-01 -5.49362898e-01 -5.00000954e-01 6.38302565e-01
2.48793006e-01 -1.02721488e+00 -2.13574812e-01 -1.19778253e-01] | [8.568814277648926, -3.4970202445983887] |
0b4ef07f-8ffd-47fd-9e06-6f61bc9ac715 | hierarchical-confusion-matrix-for | 2306.09461 | null | https://arxiv.org/abs/2306.09461v1 | https://arxiv.org/pdf/2306.09461v1.pdf | Hierarchical confusion matrix for classification performance evaluation | In this work we propose a novel concept of a hierarchical confusion matrix, opening the door for popular confusion matrix based (flat) evaluation measures from binary classification problems, while considering the peculiarities of hierarchical classification problems. We develop the concept to a generalized form and prove its applicability to all types of hierarchical classification problems including directed acyclic graphs, multi path labelling, and non mandatory leaf node prediction. Finally, we use measures based on the novel confusion matrix to evaluate models within a benchmark for three real world hierarchical classification applications and compare the results to established evaluation measures. The results outline the reasonability of this approach and its usefulness to evaluate hierarchical classification problems. The implementation of hierarchical confusion matrix is available on GitHub. | ['Martin Hemberg', 'Michael Neunteufel', 'Kevin Riehl'] | 2023-06-15 | null | null | null | null | ['classification-1'] | ['methodology'] | [ 3.04314733e-01 2.50428736e-01 -5.11056818e-02 -2.57501125e-01
-4.60403651e-01 -8.23160052e-01 5.25077105e-01 6.56298578e-01
-2.11928114e-01 8.55829000e-01 -1.01611063e-01 -5.29222310e-01
-1.02067578e+00 -8.82210553e-01 3.30315232e-01 -7.54455566e-01
-5.10286808e-01 7.35622764e-01 6.21850193e-01 -1.92062691e-01
4.69295800e-01 7.63038874e-01 -1.69700015e+00 4.55104798e-01
6.50015235e-01 1.01242447e+00 -1.64642334e-01 9.42944288e-01
9.99984816e-02 9.07722294e-01 -6.71642244e-01 -5.28351903e-01
-3.47186625e-02 -9.25877690e-02 -1.70571005e+00 2.36110121e-01
1.02793619e-01 4.01736826e-01 8.90006498e-02 7.20213056e-01
2.53658801e-01 4.94096391e-02 1.08491421e+00 -1.62571979e+00
-1.28728807e-01 4.09280568e-01 -7.89251700e-02 3.48094165e-01
5.31110406e-01 -4.57545787e-01 1.23944497e+00 -5.58112621e-01
5.04103422e-01 1.06205821e+00 1.02848339e+00 1.36720940e-01
-1.55215287e+00 -3.11123461e-01 -1.90856963e-01 6.65971398e-01
-1.66198111e+00 -6.98769689e-02 1.00354813e-01 -7.98521399e-01
7.77142823e-01 8.92483473e-01 3.82811695e-01 5.90769291e-01
2.42003962e-01 4.30575937e-01 1.69008231e+00 -5.23082674e-01
1.08995467e-01 5.29533848e-02 9.43008125e-01 9.68105376e-01
3.50768626e-01 -6.96860179e-02 1.66122943e-01 -3.66545945e-01
4.68007661e-02 -4.50585812e-01 1.25640526e-01 -7.30142355e-01
-9.44168508e-01 9.41560745e-01 5.40884912e-01 7.89772928e-01
6.66911229e-02 -4.19071883e-01 5.63940525e-01 2.41321355e-01
1.76197365e-01 6.06295049e-01 -2.58825451e-01 1.39233977e-01
-7.48132765e-01 -5.91025576e-02 1.16573048e+00 6.83527768e-01
6.63430393e-01 -4.62104261e-01 -3.50128472e-01 8.57867420e-01
8.40259194e-02 -1.52877405e-01 3.31022739e-01 -9.29339111e-01
-3.62652494e-03 9.68738914e-01 -3.30452979e-01 -1.05164957e+00
-1.17509532e+00 -5.54941475e-01 -1.06705678e+00 7.58350864e-02
3.20866168e-01 3.38789642e-01 -4.30192828e-01 1.29361439e+00
3.30261141e-01 -4.28956151e-01 3.49743478e-02 3.01787734e-01
1.03629851e+00 3.13151181e-01 -5.64766340e-02 -3.80329251e-01
1.31523097e+00 -8.34285796e-01 -3.03857148e-01 4.66229409e-01
1.24033558e+00 -6.98156953e-01 5.60170174e-01 5.68356931e-01
-7.22522736e-01 -3.97545427e-01 -1.09353960e+00 2.18824163e-01
-6.97228730e-01 6.02197461e-02 9.19735372e-01 1.06890559e+00
-1.30537808e+00 6.87996686e-01 -4.33668852e-01 -7.68111825e-01
-3.58489044e-02 7.58505225e-01 -6.51053131e-01 -1.75974235e-01
-1.11793804e+00 1.20314181e+00 7.97839820e-01 -2.03894246e-02
-5.44710279e-01 7.32237250e-02 -7.02281237e-01 1.25702351e-01
1.72824413e-01 -5.15748322e-01 1.08489370e+00 -4.71433640e-01
-1.05438483e+00 9.56063628e-01 3.13939601e-01 -9.13436785e-02
2.37327278e-01 7.19441235e-01 -2.77326852e-01 1.77861135e-02
4.22963619e-01 4.95082736e-01 5.32739758e-02 -1.28160655e+00
-7.89479375e-01 -3.62161636e-01 4.21715915e-01 4.91137728e-02
-4.78301257e-01 2.79979524e-03 -3.23129386e-01 -7.06078485e-02
4.02341336e-01 -1.10213327e+00 -4.48084801e-01 -7.40077376e-01
-6.82030499e-01 -5.75918555e-01 3.70861441e-01 -2.97376007e-01
1.62402117e+00 -1.76536202e+00 4.87669408e-01 6.12300754e-01
6.23234391e-01 -4.29030880e-02 1.84117518e-02 6.80327415e-01
-3.88081759e-01 1.63738921e-01 -5.89294910e-01 7.41432682e-02
-1.31158769e-01 3.60393435e-01 5.17067254e-01 4.14295107e-01
-8.80129337e-02 3.98301780e-01 -8.47466528e-01 -8.97176862e-01
2.52032578e-01 1.01432770e-01 -2.74186403e-01 -1.27879549e-02
6.19090676e-01 1.12307116e-01 -1.93709552e-01 4.36644524e-01
4.26473916e-01 -3.83059829e-01 6.64370775e-01 -1.21730857e-01
-1.05214268e-01 1.93020880e-01 -1.15840912e+00 9.43294287e-01
-4.34213966e-01 5.58424413e-01 -2.78690964e-01 -1.11091900e+00
7.24176764e-01 1.14811026e-01 5.40454805e-01 -1.65401176e-01
3.46144974e-01 -1.41557217e-01 4.76070166e-01 -1.58202365e-01
3.53285998e-01 -7.94380531e-02 -1.95953563e-01 2.64213949e-01
3.44312102e-01 -5.12507856e-02 8.98852885e-01 4.11417276e-01
1.52519906e+00 -4.44129497e-01 9.02748287e-01 -7.33621001e-01
1.07231116e+00 8.65720212e-03 2.78861336e-02 7.77349949e-01
-1.13651089e-01 2.93712229e-01 9.20987546e-01 -3.14564258e-01
-7.12480009e-01 -9.71414447e-01 -7.73383617e-01 1.15911782e+00
-1.86509088e-01 -1.08417606e+00 -7.26955533e-01 -8.98716867e-01
-2.44363323e-01 2.42841542e-01 -1.06364155e+00 7.06185848e-02
2.31322721e-02 -1.21305537e+00 6.29457891e-01 2.67212808e-01
3.26809913e-01 -6.26232028e-01 -1.61480129e-01 5.05111963e-02
-1.98259532e-01 -9.45923388e-01 4.20623153e-01 6.06630564e-01
-9.21634912e-01 -1.52949917e+00 -2.77003795e-01 -7.17928946e-01
3.76279771e-01 1.84542522e-01 1.19764125e+00 3.09158027e-01
-4.71781015e-01 5.22898138e-01 -6.02906525e-01 2.22246140e-01
-8.59086096e-01 6.92808032e-01 -3.31669077e-02 -2.04447925e-01
1.28904749e-02 -4.14770991e-01 -2.14641035e-01 8.52615595e-01
-1.03629518e+00 -3.55089866e-02 4.01974022e-01 7.35615969e-01
3.90532352e-02 4.24388796e-01 4.79077436e-02 -1.02392828e+00
7.62816548e-01 -4.84596401e-01 -5.00208080e-01 3.41608763e-01
-1.10388458e+00 2.07568362e-01 2.36074388e-01 1.63418561e-01
-5.87412298e-01 1.35696204e-02 -3.32081825e-01 3.67177427e-01
-1.87027857e-01 5.58505058e-01 -6.67488202e-02 -3.75384659e-01
8.03380549e-01 -2.10471153e-01 -2.32704431e-01 -2.86911070e-01
2.89291680e-01 7.62109697e-01 1.58454329e-01 -5.16844571e-01
5.12791693e-01 2.37166822e-01 7.92019904e-01 -8.79378259e-01
-7.58575082e-01 -8.49945545e-01 -1.21942329e+00 -2.58294910e-01
7.22501516e-01 -3.24263543e-01 -1.01337910e+00 1.61555797e-01
-1.02284229e+00 6.36177287e-02 1.44362316e-01 1.04354277e-01
-5.77952504e-01 8.46045554e-01 -7.92017043e-01 -7.49119163e-01
-6.05521575e-02 -1.08089650e+00 7.86694109e-01 -3.34871024e-01
-5.36037207e-01 -1.22628105e+00 2.77537912e-01 3.56262326e-01
6.94877356e-02 -3.80646996e-02 1.29365396e+00 -8.18796098e-01
-7.72637501e-02 -4.27189261e-01 -4.47347254e-01 3.75745088e-01
3.22606340e-02 2.15447456e-01 -7.33651102e-01 -1.66263521e-01
-2.10871354e-01 -1.94956604e-02 8.85665655e-01 -5.33206984e-02
1.08416855e+00 -1.92195014e-03 -6.57389045e-01 1.26408398e-01
1.31889939e+00 3.22054684e-01 6.60874486e-01 3.53824377e-01
6.01909637e-01 1.05145037e+00 4.98511881e-01 3.18494946e-01
5.11826217e-01 9.48669434e-01 4.26671445e-01 8.64859857e-03
1.41300395e-01 4.54376101e-01 -1.74565852e-01 1.22442222e+00
-3.20303857e-01 -1.58952907e-01 -1.31871068e+00 5.20265661e-02
-1.69400048e+00 -5.45558631e-01 -8.68871808e-01 2.16502619e+00
5.10804236e-01 3.33228528e-01 3.27766657e-01 9.54817533e-01
8.62633109e-01 -2.87245721e-01 3.90228450e-01 -6.67121351e-01
3.11598238e-02 1.74876288e-01 4.54691708e-01 7.81270385e-01
-1.42412281e+00 6.54676020e-01 7.47757149e+00 1.03686261e+00
-6.04095042e-01 3.61838162e-01 4.86323863e-01 5.77933133e-01
3.66512805e-01 6.75758943e-02 -7.91427732e-01 1.27224275e-03
1.17942929e+00 3.24370116e-02 -1.79254353e-01 7.58112013e-01
-5.23992717e-01 -2.68982977e-01 -1.08196616e+00 8.21394086e-01
9.31426212e-02 -9.30885136e-01 -1.93483874e-01 8.39650407e-02
3.02411288e-01 -1.82288453e-01 -2.10904419e-01 3.05128366e-01
5.82278252e-01 -7.86915839e-01 1.17451847e-01 1.98787712e-02
5.21329522e-01 -4.29509372e-01 8.32204938e-01 -4.78137508e-02
-1.25567734e+00 -2.71257579e-01 -1.08509660e-02 -3.63817483e-01
-3.87795359e-01 4.60862696e-01 -1.08913839e+00 9.27409112e-01
5.21302342e-01 6.59666777e-01 -1.29986179e+00 1.39293301e+00
2.32252181e-02 4.17482972e-01 -1.23721287e-01 -5.57834618e-02
4.46245342e-01 -1.11451879e-01 2.91362315e-01 1.47796035e+00
1.02195807e-01 -2.03029096e-01 3.44612217e-03 1.09534860e-01
5.28902948e-01 4.72476572e-01 -6.76923931e-01 3.93861026e-01
-2.33095177e-02 1.61735332e+00 -1.57489121e+00 -2.80818194e-01
-2.33741820e-01 7.93139577e-01 3.81313354e-01 -2.92791754e-01
-6.68985426e-01 -4.72863048e-01 6.71657994e-02 -5.92464302e-03
-1.49642885e-01 -1.29011363e-01 -2.42824823e-01 -8.79390240e-01
-3.47188830e-01 -6.56950057e-01 9.65218067e-01 -5.71335375e-01
-1.23425996e+00 1.11634636e+00 5.65112889e-01 -1.40241158e+00
-9.43795741e-02 -1.01443827e+00 5.94112426e-02 3.01027924e-01
-7.07259834e-01 -8.21937501e-01 -3.68062347e-01 4.07384694e-01
1.31210998e-01 -4.63874079e-02 1.22701108e+00 4.98328745e-01
-7.92538822e-01 2.54284471e-01 2.23875672e-01 8.18641037e-02
4.03535336e-01 -1.62606132e+00 1.10678218e-01 5.19760430e-01
1.21192664e-01 3.35512131e-01 6.28554821e-01 -1.13365211e-01
-4.40544546e-01 -7.61751592e-01 9.46391702e-01 -9.23238158e-01
8.10157657e-01 -3.26540112e-01 -6.25022173e-01 4.54372793e-01
9.45551097e-02 -3.16118151e-01 1.13824248e+00 6.34520829e-01
-4.66932714e-01 -6.23764284e-02 -1.00564134e+00 3.25974673e-01
1.02507627e+00 -1.66948915e-01 -2.76168853e-01 5.91118872e-01
8.82715657e-02 1.40250221e-01 -1.32945704e+00 6.37484193e-01
6.08324409e-01 -1.40017843e+00 7.37832189e-01 -6.84324563e-01
3.28249812e-01 -1.97421014e-01 -2.86566049e-01 -1.21827233e+00
-8.06148529e-01 -2.68210202e-01 2.77479738e-01 1.11929083e+00
6.12037361e-01 -7.12284744e-01 6.34552717e-01 2.68842280e-01
-1.02041967e-01 -7.13456452e-01 -1.17242241e+00 -7.92274535e-01
1.91162810e-01 -2.84245968e-01 1.67737946e-01 1.00407219e+00
5.02978206e-01 8.78651023e-01 9.55603793e-02 -1.24199537e-03
6.69410229e-01 8.73270705e-02 4.44691956e-01 -1.99301600e+00
-2.12772742e-01 -6.41664326e-01 -1.26847291e+00 -3.29534501e-01
2.15676337e-01 -1.27474856e+00 -2.06708714e-01 -1.63382614e+00
4.66495901e-01 -6.98205888e-01 -4.29292411e-01 4.45198685e-01
-5.56297489e-02 5.72258472e-01 3.06768626e-01 2.11401224e-01
-1.11366427e+00 -1.34728700e-02 7.81583250e-01 -9.72710252e-02
2.78966725e-01 2.65050799e-01 -6.15837276e-01 7.70923853e-01
7.61198699e-01 -6.85613215e-01 -2.30884120e-01 3.62718791e-01
2.78182477e-01 1.33309901e-01 3.40179019e-02 -1.52831066e+00
1.13591636e-02 1.75066844e-01 2.20031470e-01 -4.10160363e-01
2.33548969e-01 -9.15124834e-01 3.39319915e-01 8.32073271e-01
-3.07159752e-01 2.67601669e-01 -6.67766258e-02 2.29145795e-01
-2.44018048e-01 -5.70240498e-01 8.83555055e-01 -1.08474446e-02
-7.43376136e-01 -4.97443117e-02 -7.47073829e-01 -2.85767466e-01
1.29436374e+00 -2.98842490e-01 -3.40053797e-01 -7.88582563e-02
-1.43970668e+00 1.38673037e-01 2.26970062e-01 2.67413110e-01
1.12221204e-01 -1.27391434e+00 -4.49219495e-01 -1.92473471e-01
5.42273581e-01 -7.51980662e-01 5.06178699e-02 1.15726781e+00
-6.88338041e-01 9.68685687e-01 -5.49851298e-01 -8.02268505e-01
-1.97841382e+00 8.30480516e-01 3.50322604e-01 -7.92534173e-01
2.83569214e-03 4.57562625e-01 3.43691528e-01 -6.08145773e-01
8.73834491e-02 -2.95228213e-01 -7.33711541e-01 4.30498719e-01
1.28650635e-01 7.54653633e-01 6.08161151e-01 -8.51558983e-01
-5.42261064e-01 4.17293102e-01 1.25209019e-01 9.78064239e-02
7.69026935e-01 -2.00000659e-01 -6.55872464e-01 5.23809075e-01
1.32023931e+00 -4.29110199e-01 -2.03955293e-01 6.90110475e-02
5.90205491e-01 -8.25400427e-02 -1.91407204e-01 -6.49718285e-01
-6.15807176e-01 8.30997169e-01 7.39694118e-01 1.17301738e+00
1.11214244e+00 4.97759245e-02 -3.52911472e-01 6.72309399e-01
4.31562603e-01 -5.51997900e-01 -2.75600314e-01 6.53342962e-01
7.38449872e-01 -9.20994818e-01 1.48474470e-01 -1.13159990e+00
-4.54410255e-01 1.26405406e+00 3.33073407e-01 2.96232581e-01
9.30367649e-01 3.25704068e-01 -2.56113768e-01 -2.35927507e-01
-7.60421097e-01 -7.38973260e-01 1.62176073e-01 7.30673611e-01
6.83765233e-01 2.95962155e-01 -8.80080044e-01 2.30670691e-01
-1.22333072e-01 -3.49590659e-01 7.11396396e-01 7.73807645e-01
-3.69004935e-01 -1.39827597e+00 -4.64141220e-01 7.38428950e-01
-3.43704820e-01 -5.87238483e-02 -7.82764256e-01 8.81454110e-01
1.85426265e-01 1.11027479e+00 -2.24337757e-01 -8.28636110e-01
3.41816455e-01 1.72230840e-01 6.64459467e-01 -7.88869441e-01
-7.85516798e-01 -5.06294429e-01 5.72399735e-01 -3.55383039e-01
-6.12939298e-01 -6.21479511e-01 -5.73954999e-01 -4.13273841e-01
-5.97735763e-01 6.86221063e-01 4.60977614e-01 8.46356153e-01
-3.12175918e-02 5.22892714e-01 6.48538530e-01 -8.09552729e-01
-7.25297108e-02 -1.16675174e+00 -7.70139337e-01 1.76107496e-01
-1.06143899e-01 -1.14838302e+00 -4.04920131e-01 -2.77163088e-01] | [7.6675705909729, 4.618402004241943] |
10a9b152-158e-488c-9292-659d47bd8612 | bridging-the-covid-19-data-and-the-1 | 2301.13692 | null | https://arxiv.org/abs/2301.13692v1 | https://arxiv.org/pdf/2301.13692v1.pdf | Bridging the Covid-19 Data and the Epidemiological Model using Time-Varying Parameter SIRD Model | This paper extends the canonical model of epidemiology, the SIRD model, to allow for time-varying parameters for real-time measurement and prediction of the trajectory of the Covid-19 pandemic. Time variation in model parameters is captured using the generalized autoregressive score modeling structure designed for the typical daily count data related to the pandemic. The resulting specification permits a flexible yet parsimonious model with a low computational cost. The model is extended to allow for unreported cases using a mixed-frequency setting. Results suggest that these cases' effects on the parameter estimates might be sizeable. Full sample results show that the flexible framework accurately captures the successive waves of the pandemic. A real-time exercise indicates that the proposed structure delivers timely and precise information on the pandemic's current stance. This superior performance, in turn, transforms into accurate predictions of the confirmed and death cases. | ['Yasin Simsek', 'Cem Cakmakli'] | 2023-01-31 | null | null | null | null | ['epidemiology'] | ['medical'] | [ 5.83564676e-02 -1.05961792e-01 -3.60471934e-01 -2.00532779e-01
-6.14574611e-01 -4.63219941e-01 6.92826033e-01 1.32459655e-01
-4.16104048e-01 5.65267980e-01 6.28037274e-01 -6.18354499e-01
-5.10157228e-01 -8.03574562e-01 -2.05641493e-01 -6.53434694e-01
-5.16685724e-01 6.87859118e-01 -1.33202061e-01 -1.06597468e-01
-1.33133277e-01 2.92816371e-01 -8.04441452e-01 -2.86866516e-01
7.32327044e-01 2.05306202e-01 4.34231937e-01 6.20026350e-01
4.63632554e-01 5.12142301e-01 -6.58124983e-01 -1.87444508e-01
2.57107139e-01 -2.22582440e-03 3.43453437e-02 -1.60443828e-01
-6.53876483e-01 -7.28665769e-01 -1.41161859e-01 2.88541675e-01
3.10580134e-01 -1.25700131e-01 1.07688284e+00 -1.15963590e+00
-2.49432415e-01 1.95472628e-01 -3.18356723e-01 4.70262438e-01
5.08607328e-01 2.34852791e-01 5.63530684e-01 -3.65556121e-01
7.03145444e-01 1.28377724e+00 1.16902351e+00 3.20634097e-02
-1.24544382e+00 -5.92965961e-01 -1.40293539e-01 -3.68475348e-01
-1.30331945e+00 -2.14267716e-01 2.96492100e-01 -9.26237524e-01
1.00735748e+00 2.31784686e-01 5.81308544e-01 1.41215456e+00
6.20251417e-01 1.13801911e-01 9.56709802e-01 2.19020117e-02
-3.08265965e-02 2.00448427e-02 3.38966250e-01 5.60279004e-02
7.25040019e-01 5.03076732e-01 3.15195739e-01 -9.97337103e-01
7.49108553e-01 6.58441305e-01 -8.81802514e-02 2.42476165e-01
-8.78277719e-01 1.02308953e+00 -1.64934814e-01 3.17955971e-01
-1.21749246e+00 -3.04292798e-01 2.71862686e-01 2.28307620e-01
8.01025331e-01 -6.49810955e-02 -6.57555640e-01 -2.32934177e-01
-9.19808805e-01 4.26631242e-01 8.00474226e-01 4.49981570e-01
3.03946286e-01 3.33580971e-01 -4.65298921e-01 3.17782402e-01
2.63084620e-01 1.24789786e+00 4.37346511e-02 -7.00961471e-01
2.44814962e-01 3.10173512e-01 5.81308126e-01 -1.04187536e+00
-9.00363028e-01 -5.44587672e-01 -8.93267214e-01 -4.93121833e-01
4.97933894e-01 -5.32634377e-01 -8.47161531e-01 1.88231695e+00
3.73427838e-01 7.21057951e-01 -7.06486627e-02 1.74340039e-01
6.77091479e-02 1.10266769e+00 4.69629735e-01 -7.88731813e-01
1.47971225e+00 -4.97239502e-03 -7.63280690e-01 2.33617947e-02
8.03741097e-01 -3.57568830e-01 3.81546497e-01 8.72882828e-02
-8.70099783e-01 -1.77485138e-01 -1.16891526e-01 6.68915391e-01
-3.89568992e-02 -1.99038982e-01 5.27917147e-01 6.92573607e-01
-1.05636823e+00 3.98820527e-02 -9.51932013e-01 -6.27662003e-01
-1.18779249e-01 1.31198019e-01 3.96323949e-02 -8.12574551e-02
-1.35640156e+00 9.29531157e-01 1.04078077e-01 -1.17662050e-01
-7.41448462e-01 -1.02441990e+00 -6.07122064e-01 2.00616986e-01
4.91781123e-02 -8.99668455e-01 9.25188065e-01 -4.07035738e-01
-6.78283632e-01 4.42994475e-01 -4.92773414e-01 -5.51970720e-01
2.70976037e-01 3.32423776e-01 -6.64620399e-01 8.66710097e-02
3.95015419e-01 -2.93126583e-01 3.73068839e-01 -9.93685782e-01
-3.91055554e-01 -5.71628332e-01 -5.02471387e-01 -1.45825744e-01
1.88986912e-01 3.86598766e-01 7.18602315e-02 -9.20600116e-01
-6.02468550e-01 -1.01016331e+00 -5.02455950e-01 -9.59581375e-01
-6.20926023e-02 -1.01185210e-01 1.99378103e-01 -1.04745483e+00
1.58008778e+00 -1.89767003e+00 -3.31439465e-01 3.85114580e-01
-2.56002516e-01 8.13828409e-02 -2.00679556e-01 1.14305949e+00
5.66611290e-02 -3.31368558e-02 -3.21728796e-01 -2.30990082e-01
-2.74864316e-01 4.00949478e-01 -5.48184633e-01 6.35019779e-01
-4.35071625e-03 7.18613029e-01 -7.37398148e-01 -3.34640010e-03
3.78342904e-02 7.10631967e-01 -6.15517676e-01 2.68838108e-01
2.78973371e-01 5.32472372e-01 -5.47955155e-01 2.62056082e-01
6.67274892e-01 -3.58243257e-01 3.35530400e-01 5.38528085e-01
-1.78529590e-01 1.31415904e-01 -7.16123581e-01 5.21984637e-01
-2.07797423e-01 1.11149594e-01 1.77463427e-01 -8.80181253e-01
7.46115863e-01 6.53232634e-01 6.13836169e-01 -4.90337342e-01
-2.17414070e-02 -1.50693268e-01 8.68382379e-02 -5.15443981e-01
4.16644603e-01 -2.68678814e-01 -7.62092881e-03 6.67340696e-01
-4.54641610e-01 2.34183550e-01 2.92473733e-02 1.04409553e-01
9.87362266e-01 -3.42001736e-01 6.30440474e-01 -3.21126819e-01
-1.51516855e-01 1.89018250e-01 8.48583996e-01 9.06344414e-01
1.10818602e-01 -4.66996878e-02 4.18234378e-01 -8.72373804e-02
-9.77301359e-01 -1.15337622e+00 -4.12895679e-01 8.87669981e-01
-6.56703949e-01 -5.48817078e-03 -4.87781435e-01 1.41737163e-01
8.46619159e-02 9.94289577e-01 -9.20049667e-01 9.91156399e-02
-5.39184570e-01 -1.40343475e+00 3.90826672e-01 2.58138627e-01
-7.93756470e-02 -8.35993350e-01 -7.65097260e-01 4.41728324e-01
-1.17493264e-01 -7.23278940e-01 -1.28496468e-01 -2.60962665e-01
-8.19444597e-01 -1.32378542e+00 -7.68505275e-01 -2.28258520e-01
5.24733126e-01 3.98119420e-01 8.55766296e-01 7.09647387e-02
1.80119388e-02 8.10805142e-01 -1.15344651e-01 -6.64795935e-01
-4.67454731e-01 -3.51236671e-01 1.88698277e-01 -9.27248225e-02
5.80993712e-01 -1.69525906e-01 -7.43071616e-01 -2.84842681e-03
-1.06286633e+00 -4.94033992e-01 1.97786674e-01 6.77477777e-01
1.19111128e-01 -2.08157778e-01 1.18097007e+00 -8.31091464e-01
8.79761636e-01 -1.26585579e+00 -4.98842925e-01 2.81041503e-01
-7.73103654e-01 -5.02159297e-01 2.50684708e-01 -6.35610402e-01
-1.33924425e+00 -6.38842702e-01 9.95733738e-02 2.01882884e-01
-2.21359670e-01 1.00259066e+00 6.04569197e-01 7.02705264e-01
3.40756118e-01 2.33484700e-01 2.09620908e-01 -4.96196657e-01
-3.77473146e-01 7.19334126e-01 2.25129947e-01 -9.23772752e-02
7.71674216e-01 5.77293158e-01 -6.19706847e-02 -1.06931794e+00
-1.70152500e-01 -8.05418015e-01 -2.10695371e-01 1.68574989e-01
6.94987297e-01 -1.42475080e+00 -6.43728495e-01 4.50583577e-01
-9.60809648e-01 -6.86614931e-01 1.77353859e-01 9.98806000e-01
-3.97932887e-01 1.16164185e-01 -9.46368456e-01 -1.43503833e+00
-1.52016595e-01 -1.01776731e+00 1.03233814e+00 -2.41117567e-01
-5.61228395e-01 -1.71826088e+00 9.15350556e-01 1.26961365e-01
6.07250273e-01 4.03418630e-01 1.02829885e+00 -1.10675251e+00
-1.81786790e-01 -4.28537577e-01 2.81618517e-02 -5.57111323e-01
2.87021607e-01 2.10744768e-01 -5.04115105e-01 -6.26868010e-01
1.67329893e-01 3.93296033e-01 6.13648653e-01 8.92086029e-01
2.50043869e-01 -8.53879035e-01 -5.62631011e-01 4.14449006e-01
1.30332911e+00 6.27176285e-01 3.67015660e-01 3.51108313e-02
1.29594624e-01 9.16284919e-01 5.38087785e-01 1.01950824e+00
7.96774089e-01 7.28289425e-01 2.10495189e-01 -2.99785864e-02
6.73770010e-01 -4.69107896e-01 2.27479175e-01 8.48748982e-01
-2.18439430e-01 -4.84941274e-01 -1.28546321e+00 8.11222315e-01
-1.46501970e+00 -1.60821199e+00 -4.69344497e-01 2.43426919e+00
3.62441957e-01 -2.91289181e-01 7.07626343e-01 -2.11378798e-01
4.24392015e-01 1.55433118e-01 -2.42539216e-02 -4.91854787e-01
-2.49826368e-02 -2.53792673e-01 5.60732305e-01 6.87335730e-01
-5.24443924e-01 2.30473086e-01 8.15592098e+00 5.25364339e-01
-7.25585639e-01 1.03941709e-01 7.13413119e-01 2.74008363e-01
-5.88039041e-01 -6.71524256e-02 -5.43619394e-01 6.53556049e-01
1.76096642e+00 -4.25623208e-01 1.37461871e-01 2.09049657e-01
1.08821869e+00 -7.29731023e-02 -4.07977015e-01 4.07024652e-01
-1.67673126e-01 -9.13387835e-01 -1.61177278e-01 6.03879392e-01
7.73960292e-01 3.33890207e-02 2.50794739e-01 3.74915481e-01
5.28197050e-01 -7.89600134e-01 6.18392080e-02 7.82450914e-01
7.72206903e-01 -9.33750093e-01 8.14518988e-01 6.98705852e-01
-8.57761741e-01 -4.29569274e-01 -1.97810397e-01 -2.89348722e-01
6.37500823e-01 5.37990987e-01 -1.33047342e+00 1.85475305e-01
1.06329113e-01 4.78302985e-01 -1.55989736e-01 1.00272441e+00
3.80543917e-01 1.35168731e+00 -2.00568110e-01 4.56575036e-01
2.69639522e-01 -4.81269926e-01 7.07946837e-01 1.34171557e+00
8.37407529e-01 3.70214731e-01 6.42819703e-02 4.07883644e-01
5.40290892e-01 -9.45770741e-02 -1.30421174e+00 -3.43638808e-02
6.35388196e-01 6.45031929e-01 -4.80812162e-01 -3.96436214e-01
-3.97250831e-01 3.75811666e-01 -1.80665031e-01 8.68378043e-01
-5.86904526e-01 1.70656398e-01 5.15680134e-01 4.64334607e-01
1.89647719e-01 -2.10653007e-01 -8.43300521e-02 -8.92497480e-01
-5.35323739e-01 -6.74957514e-01 5.56115806e-01 -4.73602116e-01
-1.18153644e+00 1.64397433e-01 4.40816879e-01 -7.10765541e-01
-1.14858389e+00 -2.47077659e-01 -9.77909684e-01 9.09762561e-01
-1.16913903e+00 -1.16261935e+00 3.72358888e-01 5.44466972e-01
3.13767642e-01 2.10828915e-01 9.89500582e-01 -4.82261516e-02
-4.53799903e-01 2.86478728e-01 5.50124407e-01 -2.14540914e-01
3.45864028e-01 -8.89871955e-01 2.95983732e-01 5.73485374e-01
-5.38216650e-01 1.04746413e+00 8.61749291e-01 -1.39816713e+00
-1.03410041e+00 -1.11309969e+00 1.32116783e+00 -5.43563843e-01
7.87914455e-01 -1.27290189e-01 -9.62811053e-01 9.86616194e-01
-2.15716194e-02 -9.57404494e-01 8.43273163e-01 4.63701151e-02
-1.87439308e-01 2.28418976e-01 -1.42263293e+00 4.07111347e-01
5.22060156e-01 -4.41917986e-01 -6.39168918e-01 3.88929844e-01
7.24053264e-01 4.38594848e-01 -9.72338617e-01 3.61473054e-01
6.27095222e-01 -7.91720331e-01 1.07964182e+00 -7.74483740e-01
-2.62844533e-01 3.41122538e-01 -3.83715443e-02 -1.30626202e+00
-7.15426981e-01 -6.99686110e-01 2.66129017e-01 9.70710754e-01
3.60124618e-01 -1.08206892e+00 8.99030641e-03 6.34474933e-01
2.15987578e-01 -4.59214032e-01 -1.24337494e+00 -6.48538411e-01
1.39703780e-01 -3.57410491e-01 8.23058665e-01 1.14641523e+00
-2.77736306e-01 -2.21040487e-01 -7.94625878e-01 4.79728013e-01
6.32223487e-01 -1.69726998e-01 6.36319220e-01 -1.71942317e+00
-2.80115575e-01 1.20304510e-01 -1.18661709e-01 -5.54765344e-01
1.45262137e-01 -2.02777043e-01 -3.87652665e-01 -1.25661194e+00
6.70446217e-01 -3.46381634e-01 -7.74830580e-02 1.20519018e-02
-2.24812552e-01 -1.61637694e-01 1.12025902e-01 1.56178147e-01
6.18518107e-02 2.03114882e-01 6.39746487e-01 3.26045990e-01
-4.99120682e-01 5.43833435e-01 -3.93330216e-01 6.23206556e-01
8.96840274e-01 -5.52649438e-01 -5.81873715e-01 -1.20794535e-01
3.32791567e-01 8.86732042e-01 6.26645863e-01 -2.16693223e-01
-3.75325680e-02 -6.94451392e-01 1.16457708e-01 -6.80671990e-01
2.89948344e-01 -9.55511272e-01 8.99172008e-01 9.29911375e-01
-1.96187403e-02 8.33790362e-01 1.99663550e-01 9.51531470e-01
3.13456744e-01 1.46758616e-01 1.54694647e-01 1.94161713e-01
3.35146725e-01 4.30164933e-01 -8.43315542e-01 1.77300259e-01
1.01420009e+00 -3.68597716e-01 -3.24381173e-01 -7.33910322e-01
-5.62781096e-01 1.07431650e-01 4.99446034e-01 3.37149240e-02
2.90688336e-01 -8.73934209e-01 -9.20278132e-01 1.20800883e-01
-3.33335459e-01 -7.11310685e-01 5.46021402e-01 1.23096538e+00
-8.39974657e-02 9.22360480e-01 9.03089270e-02 -2.59915441e-01
-1.03871799e+00 8.27503562e-01 8.34576562e-02 -6.17698610e-01
-3.45565408e-01 -8.82215500e-02 4.80037332e-01 -2.51337707e-01
-2.03047574e-01 -7.91340470e-02 -3.04554641e-01 4.21265602e-01
7.99315512e-01 8.63690615e-01 -4.76337910e-01 -1.08362985e+00
-3.74500662e-01 3.37948829e-01 3.06671232e-01 -1.18551143e-01
1.57191658e+00 -6.91524148e-01 1.01308385e-02 7.85619557e-01
7.83786952e-01 1.46239057e-01 -1.19788611e+00 5.34929968e-02
8.89844820e-02 6.25771806e-02 -3.52648258e-01 -7.02512443e-01
-4.69305724e-01 4.84069824e-01 4.89618748e-01 3.04423839e-01
8.46629202e-01 -3.80167127e-01 5.42734861e-01 -3.70286740e-02
5.00052214e-01 -2.95752764e-01 -6.92798615e-01 2.89024502e-01
4.76118922e-01 -8.20445657e-01 -2.06590071e-01 1.01664670e-01
-5.27495980e-01 5.23959875e-01 -2.34705225e-01 -8.82928632e-03
8.70267868e-01 5.82162365e-02 -8.92125815e-02 -2.11740166e-01
-1.07353866e+00 1.42186388e-01 2.11954132e-01 1.04730189e+00
2.30144680e-01 5.54896295e-01 -6.49980307e-01 6.15077257e-01
1.41499788e-01 -1.12997577e-01 7.51376510e-01 3.05339128e-01
-3.10521126e-01 -5.49906373e-01 -6.31105185e-01 7.66602695e-01
-9.26279068e-01 -2.46352181e-01 -7.21776485e-03 9.73031282e-01
-2.29659036e-01 1.25322998e+00 5.77726543e-01 1.43929735e-01
3.22832316e-02 1.76374093e-01 -2.70413667e-01 -3.59372735e-01
-6.36017025e-01 3.04948896e-01 3.56221981e-02 -2.89480478e-01
-3.75777990e-01 -7.87197351e-01 -8.65989268e-01 -6.98243201e-01
-1.10225077e-03 2.52850890e-01 1.69130564e-01 8.63726676e-01
5.93145907e-01 4.57429029e-02 8.16343844e-01 -6.92153037e-01
-8.96326005e-01 -1.06635690e+00 -7.30649054e-01 3.75951022e-01
6.01677477e-01 -5.21368742e-01 -6.98617935e-01 -4.13750648e-01] | [6.011052131652832, 4.389005184173584] |
51a7f7a4-dc28-41ca-8188-8ec1b799721d | rankcse-unsupervised-sentence-representations | 2305.16726 | null | https://arxiv.org/abs/2305.16726v1 | https://arxiv.org/pdf/2305.16726v1.pdf | RankCSE: Unsupervised Sentence Representations Learning via Learning to Rank | Unsupervised sentence representation learning is one of the fundamental problems in natural language processing with various downstream applications. Recently, contrastive learning has been widely adopted which derives high-quality sentence representations by pulling similar semantics closer and pushing dissimilar ones away. However, these methods fail to capture the fine-grained ranking information among the sentences, where each sentence is only treated as either positive or negative. In many real-world scenarios, one needs to distinguish and rank the sentences based on their similarities to a query sentence, e.g., very relevant, moderate relevant, less relevant, irrelevant, etc. In this paper, we propose a novel approach, RankCSE, for unsupervised sentence representation learning, which incorporates ranking consistency and ranking distillation with contrastive learning into a unified framework. In particular, we learn semantically discriminative sentence representations by simultaneously ensuring ranking consistency between two representations with different dropout masks, and distilling listwise ranking knowledge from the teacher. An extensive set of experiments are conducted on both semantic textual similarity (STS) and transfer (TR) tasks. Experimental results demonstrate the superior performance of our approach over several state-of-the-art baselines. | ['Rui Yan', 'Kai Chen', 'Dongyan Zhao', 'Yunsen Xian', 'Wei Wu', 'Jingang Wang', 'Qifan Wang', 'Jiahao Liu', 'Jiduan Liu'] | 2023-05-26 | null | null | null | null | ['semantic-textual-similarity'] | ['natural-language-processing'] | [ 6.57461107e-01 -2.22953379e-01 -1.08911484e-01 -8.71165931e-01
-1.00193810e+00 -4.70202833e-01 7.20346630e-01 8.41473639e-01
-7.68173277e-01 5.50876439e-01 7.53084421e-01 -6.35770187e-02
-3.15298915e-01 -6.16415620e-01 -3.40468675e-01 -6.04074836e-01
3.15735638e-01 4.24586982e-01 3.52000266e-01 -5.05083084e-01
4.83065993e-01 -4.72739674e-02 -1.40976131e+00 6.64406717e-01
1.01928437e+00 8.97443712e-01 4.95526880e-01 2.26245284e-01
-3.61666977e-01 9.87021983e-01 -5.37457108e-01 -3.64443004e-01
-1.85979366e-01 -6.06672168e-01 -9.75950718e-01 -2.23877013e-01
4.56406772e-01 -1.16777368e-01 -2.40000069e-01 1.23827541e+00
5.04010081e-01 5.43609679e-01 7.10857153e-01 -6.71300650e-01
-9.43873107e-01 7.46817887e-01 -5.77609122e-01 5.41484475e-01
4.01853055e-01 -3.65683049e-01 1.44965601e+00 -9.48443234e-01
3.64345282e-01 1.45655739e+00 1.05730161e-01 5.38847506e-01
-1.11386871e+00 -5.50363600e-01 4.67834830e-01 3.44950497e-01
-1.04107022e+00 -4.20119911e-01 9.11857307e-01 -2.85733104e-01
8.65409732e-01 2.71950454e-01 -1.30720928e-01 8.55080962e-01
-7.46112838e-02 8.76769066e-01 1.04926944e+00 -4.75819141e-01
2.61199176e-01 1.57556787e-01 6.42267823e-01 3.19481760e-01
3.80705968e-02 -4.55480397e-01 -4.00682330e-01 -9.55537036e-02
2.55980849e-01 3.21704894e-01 -3.26253623e-01 -2.66376942e-01
-1.14065886e+00 9.24265087e-01 7.36006916e-01 5.06318450e-01
-2.54931211e-01 -3.76956537e-02 8.93331468e-01 5.28162420e-01
7.57929385e-01 3.78424913e-01 -4.77810651e-01 1.64976507e-01
-7.54630983e-01 1.25959694e-01 2.37036139e-01 8.17598403e-01
6.69442236e-01 -3.67150426e-01 -7.56425321e-01 1.28719771e+00
3.25353235e-01 1.88046664e-01 9.22187328e-01 -3.74555379e-01
1.02972209e+00 6.48236990e-01 -1.64125025e-01 -1.19096732e+00
-1.55035853e-01 -4.05750155e-01 -1.01022303e+00 -4.43580061e-01
-2.03313202e-01 1.23629555e-01 -5.97378135e-01 1.84496880e+00
6.72951788e-02 2.01040953e-01 4.38315898e-01 1.11160886e+00
1.27918661e+00 8.65054727e-01 3.13853770e-01 -2.80167013e-01
1.42151344e+00 -9.34517860e-01 -5.54434896e-01 -3.82544219e-01
7.69055188e-01 -6.80948675e-01 1.23420727e+00 4.37301444e-03
-9.43572462e-01 -7.10026860e-01 -9.52062070e-01 -4.41052496e-01
-2.71626204e-01 1.66794136e-01 4.36593920e-01 1.53985098e-01
-8.52769673e-01 6.68178141e-01 -4.07725692e-01 -3.00524831e-01
3.37731659e-01 3.10102791e-01 -3.81919056e-01 -2.64779329e-01
-1.79079294e+00 9.16418970e-01 5.49173057e-01 3.82636525e-02
-4.39243793e-01 -4.10479724e-01 -1.08763206e+00 3.02693993e-01
2.10873768e-01 -6.57950878e-01 1.19863105e+00 -1.01665223e+00
-1.28953230e+00 1.18284261e+00 -3.00359786e-01 -3.50261539e-01
1.18207425e-01 -3.06786150e-01 -3.70027125e-01 2.46325344e-01
3.62347424e-01 3.71771961e-01 5.45702696e-01 -9.37053382e-01
-4.66171205e-01 -4.96654928e-01 3.84248853e-01 6.29696310e-01
-7.25302339e-01 4.97544885e-01 -1.44081920e-01 -6.96722388e-01
2.52185911e-01 -4.01911885e-01 -3.20549160e-01 -1.91469669e-01
-2.53979415e-01 -8.60875010e-01 4.23591346e-01 -5.13619959e-01
1.24597728e+00 -2.26115870e+00 2.52740502e-01 -3.70568596e-02
2.11962927e-02 2.24025220e-01 -4.13206965e-01 5.07596254e-01
-2.38146603e-01 7.20300525e-02 -4.65953499e-01 -3.95466030e-01
-3.25103588e-02 9.88624096e-02 -6.26185238e-01 1.59750506e-01
4.97133136e-01 6.78448677e-01 -1.50811064e+00 -6.33241177e-01
-1.09310150e-01 2.86945879e-01 -4.41461563e-01 2.78462321e-01
4.68339063e-02 3.21994752e-01 -7.16762722e-01 7.52287954e-02
5.99426150e-01 -2.82346427e-01 1.96015671e-01 -1.36367112e-01
1.65628389e-01 9.87190306e-01 -9.47085142e-01 1.88812459e+00
-6.03061736e-01 3.45927387e-01 -3.60280603e-01 -1.42248142e+00
1.14862525e+00 2.12233230e-01 8.90933424e-02 -8.38772774e-01
2.55677034e-04 1.76989093e-01 -4.30926122e-02 -5.00732303e-01
7.57290304e-01 -5.51409781e-01 -3.53621662e-01 5.33624291e-01
2.14267634e-02 -6.49267137e-02 3.35724086e-01 5.32411218e-01
9.38085914e-01 -1.39364734e-01 4.25787032e-01 -3.29318106e-01
9.55013037e-01 -4.84807044e-01 5.38601220e-01 5.37765205e-01
-8.47776160e-02 8.55308473e-01 6.00314021e-01 4.95275632e-02
-4.94636685e-01 -1.18385863e+00 8.62562284e-02 1.56777000e+00
4.26589042e-01 -4.22775716e-01 -3.41401070e-01 -8.89586389e-01
-1.68568343e-01 6.28164828e-01 -5.58072031e-01 -6.12674356e-01
-5.91734529e-01 -5.00299871e-01 1.64997682e-01 6.43279910e-01
4.42278951e-01 -1.27493930e+00 -8.99603292e-02 1.45374253e-01
-3.70225698e-01 -8.22012126e-01 -6.60062075e-01 1.27885789e-01
-9.41058218e-01 -7.49904215e-01 -6.27289832e-01 -1.23249829e+00
8.79302204e-01 8.05085123e-01 1.32047093e+00 2.09215179e-01
1.93359360e-01 -3.49042267e-02 -6.59483433e-01 4.58751768e-02
-4.32444550e-02 4.25905064e-02 -1.27018867e-02 1.39402017e-01
4.07234550e-01 -3.35215032e-01 -5.98005891e-01 -1.00871839e-01
-1.09706950e+00 -2.12595075e-01 6.45186424e-01 1.09305251e+00
5.32032907e-01 -4.12992612e-02 9.40395653e-01 -1.21386874e+00
9.87876475e-01 -7.25295424e-01 1.48084825e-02 3.33645940e-01
-2.39943326e-01 2.38084555e-01 8.82550657e-01 -2.03133434e-01
-1.15486693e+00 -2.73917407e-01 -1.44131973e-01 -1.59293905e-01
2.77888812e-02 8.17983270e-01 -2.70017058e-01 6.64657116e-01
3.89732689e-01 4.36425358e-01 -2.69999683e-01 -4.45181280e-01
3.78643543e-01 9.08675253e-01 3.84182513e-01 -7.67757595e-01
6.60773873e-01 2.01039195e-01 -4.80731070e-01 -6.41911685e-01
-1.44751310e+00 -8.26659262e-01 -6.54199660e-01 1.86209157e-01
5.86484551e-01 -9.21826839e-01 -1.83170661e-01 6.23704679e-02
-1.28206515e+00 2.33027816e-01 -1.65100783e-01 4.53660518e-01
-2.60619611e-01 5.93115032e-01 -4.86700505e-01 -5.50173879e-01
-5.55122316e-01 -9.17492807e-01 1.19065511e+00 2.62175322e-01
-1.64927512e-01 -9.26361322e-01 8.25742930e-02 4.63703662e-01
3.30091923e-01 -2.13354751e-01 1.02981782e+00 -1.03009164e+00
4.63435873e-02 -1.25426471e-01 -4.85521019e-01 4.97581929e-01
3.62952381e-01 -3.75167787e-01 -8.92969131e-01 -4.16358620e-01
2.43459102e-02 -6.44837856e-01 1.34450865e+00 -5.83554097e-02
1.36321783e+00 -2.43913159e-01 -1.19266525e-01 6.05118871e-02
1.16813111e+00 -2.29629770e-01 5.27799726e-01 2.25968018e-01
5.83113372e-01 9.77793097e-01 8.33085537e-01 1.91299662e-01
3.41332436e-01 4.63201255e-01 2.14464322e-01 6.88294768e-02
1.11092798e-01 -2.72770971e-01 4.93430734e-01 1.20800960e+00
3.04213107e-01 -9.34700221e-02 -4.93302226e-01 4.70707029e-01
-1.87683308e+00 -1.20887065e+00 2.44946190e-04 2.27153897e+00
1.09377289e+00 4.05924112e-01 -1.92528203e-01 1.17699772e-01
1.03592551e+00 4.81941462e-01 -3.94750208e-01 -4.20103163e-01
-5.15110381e-02 2.59132564e-01 -1.22198038e-01 3.24260175e-01
-1.29502499e+00 9.91707861e-01 4.83382082e+00 8.00936043e-01
-1.02220607e+00 1.11200489e-01 7.33895421e-01 5.46079651e-02
-7.15034246e-01 2.75094667e-03 -4.87135619e-01 5.17126739e-01
6.38606787e-01 -4.37826157e-01 -7.77719766e-02 6.89838290e-01
2.24101827e-01 2.45459437e-01 -1.10201466e+00 9.14447904e-01
4.41647410e-01 -1.06431007e+00 3.87715608e-01 -6.28865361e-01
6.14445925e-01 -9.37311873e-02 -3.52612347e-03 7.34812379e-01
1.52216554e-01 -8.61989617e-01 6.12335980e-01 4.22986001e-01
5.61160326e-01 -7.42310226e-01 1.01535773e+00 2.62354165e-01
-1.33568406e+00 3.80198285e-02 -6.77841663e-01 -3.07572335e-01
-1.12415686e-01 5.61524212e-01 -4.58697259e-01 7.74564981e-01
5.55481970e-01 1.20180893e+00 -6.91937268e-01 7.46376455e-01
-5.04657865e-01 5.65836310e-01 1.63876325e-01 -5.18364906e-01
3.05903465e-01 -1.76678658e-01 1.70251593e-01 1.38348365e+00
1.45782549e-02 2.04007432e-01 3.82483691e-01 5.96657097e-01
-3.67898703e-01 4.58102494e-01 -5.63928723e-01 2.48154420e-02
5.46815813e-01 1.20466089e+00 -5.12018383e-01 -5.19855797e-01
-4.26600605e-01 1.04523158e+00 7.11858690e-01 2.04920530e-01
-3.48513991e-01 -7.71221161e-01 6.20416403e-01 -2.47952536e-01
8.15531537e-02 2.88661513e-02 -9.65671167e-02 -1.34555221e+00
2.57976025e-01 -4.62085813e-01 7.09542632e-01 -5.66683710e-01
-1.95451164e+00 5.54679215e-01 -1.79145530e-01 -1.54445386e+00
-2.19010442e-01 -5.04205942e-01 -9.12758648e-01 9.52116191e-01
-1.91342223e+00 -8.81011128e-01 3.18070687e-02 4.15438235e-01
9.06345844e-01 -6.78304285e-02 6.37497663e-01 1.93014294e-01
-4.31336075e-01 5.81246316e-01 2.87687451e-01 3.51056904e-01
7.61599541e-01 -1.36229217e+00 1.31406158e-01 5.91221988e-01
5.22220694e-02 1.05599546e+00 4.49362397e-01 -2.50907093e-01
-1.04329729e+00 -1.22689557e+00 1.36005855e+00 -1.62722200e-01
6.93500400e-01 -3.36173445e-01 -1.24230051e+00 2.89359093e-01
2.63824195e-01 1.02423511e-01 9.10465956e-01 3.43924522e-01
-5.57636619e-01 -3.24287862e-01 -8.28844786e-01 6.17434621e-01
9.31229591e-01 -8.21089745e-01 -1.27178490e+00 4.44070309e-01
9.32915092e-01 4.00065668e-02 -5.45460761e-01 2.99896657e-01
7.93319270e-02 -6.92100763e-01 9.81025875e-01 -9.61175561e-01
8.65070939e-01 -4.14985746e-01 -1.78679183e-01 -1.38633132e+00
-4.42327291e-01 -1.77531570e-01 1.81884512e-01 1.61448193e+00
3.62373382e-01 -5.30682564e-01 3.72243106e-01 1.06680036e-01
-3.65531147e-01 -8.05394769e-01 -8.16475689e-01 -7.73577809e-01
3.11714530e-01 -1.91095341e-02 4.12826985e-01 1.11077452e+00
2.98546076e-01 1.00369811e+00 -5.64700412e-03 1.25208870e-01
2.97625631e-01 4.82140929e-01 2.85441875e-01 -1.28833377e+00
-1.35906935e-01 -5.97856104e-01 -4.54403222e-01 -1.38423908e+00
5.70828080e-01 -1.43632209e+00 4.04385835e-01 -1.93181777e+00
6.90670788e-01 -1.36346310e-01 -1.01888704e+00 2.41587251e-01
-7.69391239e-01 -2.29754187e-02 9.62712169e-02 2.37534419e-01
-1.14969039e+00 9.41064656e-01 1.09003758e+00 -5.25401354e-01
1.40732318e-01 -1.70950174e-01 -1.01973283e+00 5.91964304e-01
7.81841159e-01 -4.95891094e-01 -7.34195113e-01 -7.85885155e-01
2.69072086e-01 -2.24370345e-01 1.56016275e-01 -6.04202688e-01
2.71827072e-01 -1.24973379e-01 1.74131721e-01 -4.73271668e-01
1.10384502e-01 -4.56833512e-01 -8.70934665e-01 3.20520580e-01
-1.11582530e+00 3.57103460e-02 -2.18698651e-01 6.36581182e-01
-6.30484462e-01 -5.47177315e-01 6.69542789e-01 -8.08796063e-02
-8.37487638e-01 2.04790443e-01 9.69382897e-02 4.90157098e-01
5.67242622e-01 1.92836240e-01 -4.87186760e-01 -1.30298853e-01
-3.88375133e-01 4.15819138e-01 4.91646305e-02 7.59670675e-01
1.06090665e+00 -1.47559047e+00 -1.23782170e+00 -7.01987594e-02
6.92570984e-01 1.63820330e-02 3.59447151e-01 4.76071835e-01
3.59066995e-03 3.35966408e-01 2.05189735e-01 -4.19617474e-01
-1.24720931e+00 4.55206811e-01 -4.22245376e-02 -4.60809201e-01
-4.47826773e-01 9.58392620e-01 3.98052037e-01 -7.00457752e-01
9.89975557e-02 -2.33276978e-01 -8.59367728e-01 1.87887654e-01
6.97278976e-01 -5.52574657e-02 1.22502722e-01 -6.95111215e-01
-4.66093272e-01 6.26504838e-01 -5.77481091e-01 1.94747880e-01
1.22615623e+00 -8.93168822e-02 -4.53077406e-01 7.00079918e-01
1.49203455e+00 -1.72327548e-01 -7.00628579e-01 -6.42094731e-01
4.50815469e-01 -4.74162310e-01 -7.58680981e-03 -4.24691647e-01
-8.30076098e-01 1.04212415e+00 2.52350956e-01 2.20126346e-01
1.07626736e+00 2.99979806e-01 7.88837194e-01 6.17101133e-01
1.09486729e-01 -1.06661451e+00 3.61848474e-01 8.00256550e-01
1.05701292e+00 -1.45906830e+00 -1.17246896e-01 -3.31635684e-01
-7.55048633e-01 9.58859921e-01 7.29477108e-01 -3.43958974e-01
3.37441295e-01 -4.80479509e-01 -9.46224034e-02 -1.29775584e-01
-8.61469865e-01 -3.06163251e-01 4.49011177e-01 1.66304767e-01
1.05520833e+00 -5.22231087e-02 -6.97693527e-01 8.46797645e-01
4.71852906e-02 -4.83420312e-01 1.92549333e-01 9.50450301e-01
-6.66897476e-01 -1.00382984e+00 1.36846034e-02 6.38503551e-01
-2.80767858e-01 -4.78632957e-01 -3.83539408e-01 1.54683486e-01
-2.04922870e-01 1.24014902e+00 1.91154197e-01 -1.68345556e-01
5.14982998e-01 -1.89498872e-01 1.50973484e-01 -1.17422831e+00
-6.99476421e-01 -3.31430644e-01 -3.36388126e-02 -6.45722151e-02
-3.92378807e-01 -6.49682105e-01 -1.56647849e+00 2.94258386e-01
-3.36055875e-01 4.72478092e-01 2.00508952e-01 1.16994774e+00
2.42433891e-01 6.66385293e-01 1.11361277e+00 -4.70433652e-01
-9.51126516e-01 -1.11892116e+00 -4.86221731e-01 9.19636786e-01
2.53374517e-01 -5.36514997e-01 -3.58075917e-01 -8.90711173e-02] | [11.046395301818848, 8.52261734008789] |
9c6ccf10-6de0-4e0e-a344-1d8281b7df4a | the-causal-neural-connection-expressiveness | 2107.00793 | null | https://arxiv.org/abs/2107.00793v3 | https://arxiv.org/pdf/2107.00793v3.pdf | The Causal-Neural Connection: Expressiveness, Learnability, and Inference | One of the central elements of any causal inference is an object called structural causal model (SCM), which represents a collection of mechanisms and exogenous sources of random variation of the system under investigation (Pearl, 2000). An important property of many kinds of neural networks is universal approximability: the ability to approximate any function to arbitrary precision. Given this property, one may be tempted to surmise that a collection of neural nets is capable of learning any SCM by training on data generated by that SCM. In this paper, we show this is not the case by disentangling the notions of expressivity and learnability. Specifically, we show that the causal hierarchy theorem (Thm. 1, Bareinboim et al., 2020), which describes the limits of what can be learned from data, still holds for neural models. For instance, an arbitrarily complex and expressive neural net is unable to predict the effects of interventions given observational data alone. Given this result, we introduce a special type of SCM called a neural causal model (NCM), and formalize a new type of inductive bias to encode structural constraints necessary for performing causal inferences. Building on this new class of models, we focus on solving two canonical tasks found in the literature known as causal identification and estimation. Leveraging the neural toolbox, we develop an algorithm that is both sufficient and necessary to determine whether a causal effect can be learned from data (i.e., causal identifiability); it then estimates the effect whenever identifiability holds (causal estimation). Simulations corroborate the proposed approach. | ['Elias Bareinboim', 'Yoshua Bengio', 'Kai-Zhan Lee', 'Kevin Xia'] | 2021-07-02 | null | http://proceedings.neurips.cc/paper/2021/hash/5989add1703e4b0480f75e2390739f34-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/5989add1703e4b0480f75e2390739f34-Paper.pdf | neurips-2021-12 | ['causal-identification'] | ['reasoning'] | [ 4.70952123e-01 6.06740654e-01 -6.17893457e-01 -2.37690762e-01
1.42493173e-01 -5.52755713e-01 9.98874784e-01 2.03908041e-01
-1.14454843e-01 1.12124181e+00 3.66026700e-01 -7.43691206e-01
-8.70049477e-01 -1.09744728e+00 -1.26459324e+00 -6.87190413e-01
-5.15184283e-01 2.22356901e-01 -1.86492786e-01 -7.18220994e-02
1.18332684e-01 5.99716008e-01 -1.55865669e+00 3.24679837e-02
9.38127041e-01 6.44594073e-01 4.11986746e-02 4.95209664e-01
2.75216401e-01 1.02668822e+00 -2.12722987e-01 -1.81258529e-01
-5.41636534e-02 -6.15692079e-01 -9.76341188e-01 -5.92729509e-01
3.03225607e-01 -5.09772182e-01 -3.47118258e-01 1.03483105e+00
-6.14096411e-03 -4.87041175e-02 9.55894530e-01 -1.45354116e+00
-7.84234107e-01 1.24064171e+00 3.70697421e-03 4.34579141e-02
3.09536248e-01 3.71296220e-02 1.31604421e+00 -4.92648929e-01
4.78081644e-01 1.46737218e+00 6.60933435e-01 5.89624703e-01
-1.62382793e+00 -6.20898485e-01 1.75517708e-01 4.34899256e-02
-9.19368804e-01 -2.71584868e-01 5.96398473e-01 -6.08974040e-01
4.80060875e-01 3.71582866e-01 7.52980828e-01 1.56935513e+00
2.97843456e-01 5.27850330e-01 1.27007473e+00 -4.83919591e-01
3.71117979e-01 -1.46916315e-01 2.43017748e-01 6.28599882e-01
6.39189482e-01 8.51093411e-01 -6.31150365e-01 -2.34629959e-01
1.01831329e+00 -1.51095986e-01 -5.03217876e-01 -3.37837934e-01
-1.11834288e+00 9.63428736e-01 6.01172864e-01 3.33777457e-01
-4.34161127e-01 5.10536313e-01 2.27537215e-01 2.98331171e-01
1.24366712e-02 6.39943063e-01 -3.99110019e-01 4.18437749e-01
-5.78865826e-01 5.20399511e-01 7.89993763e-01 5.24403691e-01
4.43762839e-01 6.79726899e-02 7.83459283e-03 2.28814393e-01
2.69617110e-01 5.47978103e-01 2.32822716e-01 -8.93220603e-01
-9.27402526e-02 5.83435357e-01 2.36388370e-01 -9.88163114e-01
-5.75299621e-01 -3.65703225e-01 -1.17145813e+00 3.82846482e-02
5.86669743e-01 -3.11472952e-01 -6.20449960e-01 2.52506661e+00
1.59921512e-01 4.32985514e-01 9.17136073e-02 7.63746500e-01
5.27671099e-01 4.91374701e-01 2.16058269e-01 -5.02486527e-01
8.66750717e-01 -6.17966242e-03 -8.28753114e-01 2.92759445e-02
3.93780559e-01 1.27285078e-01 9.59058225e-01 3.45238954e-01
-9.72275615e-01 -3.48985165e-01 -1.22685194e+00 2.52968341e-01
-4.40421164e-01 -2.92018086e-01 1.04921234e+00 5.07213354e-01
-8.84058952e-01 9.02688980e-01 -7.22847223e-01 -3.18487823e-01
2.43037775e-01 5.22577584e-01 -2.77386725e-01 1.83642820e-01
-1.69896448e+00 9.68720138e-01 7.37141073e-01 2.84455925e-01
-1.46381760e+00 -1.05896378e+00 -4.63516891e-01 2.41116300e-01
4.29180771e-01 -9.58179116e-01 1.16047561e+00 -1.18346477e+00
-1.05867541e+00 4.55362529e-01 5.07310405e-02 -6.72570527e-01
4.99462575e-01 -1.74567118e-01 -3.13215077e-01 -1.16301820e-01
-6.68551102e-02 4.49352115e-01 6.86877310e-01 -1.35607755e+00
-4.67211872e-01 -3.95216107e-01 5.06670117e-01 -3.41074318e-01
-2.69935489e-01 -1.96225375e-01 4.98675525e-01 -4.21001971e-01
-1.63759619e-01 -5.62215805e-01 -6.38445988e-02 -2.15523884e-01
-7.40692019e-01 -3.25491726e-01 3.39855075e-01 -2.88523138e-01
1.18103099e+00 -1.76035953e+00 4.39142376e-01 4.38004285e-01
4.42534178e-01 -5.66801280e-02 -3.69493403e-02 3.87137651e-01
-7.42237747e-01 5.40242851e-01 -4.73462105e-01 2.21796513e-01
2.96470165e-01 4.29702848e-01 -6.27609134e-01 4.96949047e-01
4.05328900e-01 9.76058841e-01 -1.06532800e+00 -1.21160939e-01
1.67544484e-01 3.27502757e-01 -7.02099085e-01 3.27721775e-01
-4.48049992e-01 5.57731748e-01 -3.72233927e-01 3.22337188e-02
3.75143498e-01 -1.34144351e-01 5.65318823e-01 -9.12479125e-04
-2.49084145e-01 4.36981261e-01 -1.33718777e+00 1.12838936e+00
-5.09877801e-01 3.97985160e-01 -2.71522880e-01 -1.42451453e+00
5.35838842e-01 6.06680870e-01 2.15824857e-01 -3.85984540e-01
2.03249738e-01 2.51638383e-01 3.09265375e-01 -6.02500379e-01
-1.19778968e-01 -4.92014050e-01 2.94363052e-02 4.25172806e-01
2.03232497e-01 3.54909748e-01 1.34187594e-01 7.82733783e-02
1.16161048e+00 -6.17878698e-03 5.54616809e-01 -6.36820018e-01
4.30184990e-01 -8.75946432e-02 5.20131230e-01 1.12368762e+00
2.75668055e-01 -5.66762574e-02 1.03730190e+00 -5.12163162e-01
-1.01555884e+00 -1.38140404e+00 -3.66622657e-01 7.73935795e-01
-3.01528096e-01 -3.99312712e-02 -6.83455467e-01 -4.56100464e-01
8.69237781e-02 8.97835553e-01 -1.23598552e+00 -3.46210182e-01
-4.54773039e-01 -8.35904181e-01 6.84982419e-01 4.36320394e-01
2.03503996e-01 -1.03547645e+00 -7.30238914e-01 -5.15100211e-02
-6.31158054e-02 -5.01259923e-01 3.19240093e-01 4.92935896e-01
-8.34725380e-01 -1.37668002e+00 -1.73148289e-01 -2.39873588e-01
4.99281377e-01 -4.25844520e-01 1.09540057e+00 1.19706430e-01
1.13005050e-01 1.76957756e-01 4.42488194e-02 -4.18607473e-01
-8.08113873e-01 -1.39111385e-01 4.66026306e-01 -4.97357547e-02
8.40393752e-02 -1.07852590e+00 -3.59722704e-01 -5.82379140e-02
-1.09792829e+00 7.72334486e-02 4.64580804e-01 9.76328373e-01
2.43428990e-01 3.09042275e-01 9.81055140e-01 -9.62031543e-01
6.96680665e-01 -7.52458155e-01 -7.83996463e-01 2.65476763e-01
-7.79279351e-01 5.17575622e-01 8.50778520e-01 -5.00548959e-01
-1.02738547e+00 -7.26800784e-02 -2.08341293e-02 -1.79130092e-01
-3.69734734e-01 9.61066246e-01 -4.24546212e-01 3.89541775e-01
7.54161358e-01 -2.46414691e-02 -1.52274713e-01 -3.09222162e-01
5.67695737e-01 2.73056358e-01 7.15318859e-01 -9.10421371e-01
6.48612440e-01 4.55459982e-01 4.24502611e-01 -4.39535499e-01
-1.02516532e+00 2.29088932e-01 -6.64385438e-01 -3.23438913e-01
8.72240603e-01 -4.15103465e-01 -1.35380840e+00 -2.66971011e-02
-1.25690496e+00 -4.93287057e-01 -2.94276923e-01 4.40593034e-01
-7.30342388e-01 -2.98698127e-01 -3.37371558e-01 -1.06151891e+00
1.83368191e-01 -8.40465307e-01 4.20873731e-01 -1.84689477e-01
-4.27850217e-01 -1.33157170e+00 1.45940855e-01 -4.06728923e-01
8.62767361e-03 5.09721935e-01 1.46289313e+00 -5.09359479e-01
-2.91517764e-01 -8.17959681e-02 -1.43416733e-01 1.86264306e-01
6.10047542e-02 1.42470628e-01 -8.87065411e-01 1.79687336e-01
5.67646734e-02 -1.23410933e-01 9.11143124e-01 7.86774457e-01
1.07745981e+00 -8.29742789e-01 -2.52643645e-01 2.65598118e-01
1.55145860e+00 1.41816258e-01 4.99939620e-01 9.48348045e-02
6.43754482e-01 8.63826096e-01 6.26661554e-02 1.76730201e-01
1.74855918e-01 4.34117645e-01 7.83259749e-01 3.04040909e-02
3.22176039e-01 -7.60459900e-01 3.26314718e-01 2.57722855e-01
-4.41262335e-01 -1.32023081e-01 -9.00753438e-01 4.22546655e-01
-1.96314824e+00 -1.14800608e+00 -5.33038080e-01 2.39398122e+00
1.14383662e+00 1.37104765e-01 3.04126054e-01 1.71884522e-01
5.80799282e-01 -2.90869355e-01 -5.08898139e-01 -4.77698684e-01
-2.15478346e-01 2.00048670e-01 6.02148235e-01 7.75514424e-01
-8.62063348e-01 4.02164519e-01 6.85251951e+00 4.15956557e-01
-7.75325537e-01 7.47929811e-02 4.74391580e-01 2.70815670e-01
-6.57630205e-01 6.18894845e-02 -4.11801666e-01 3.39108378e-01
1.32941806e+00 -2.47606203e-01 5.37141919e-01 3.67367566e-01
6.08174443e-01 -2.43061092e-02 -1.76479971e+00 2.41216034e-01
-4.07682925e-01 -1.43296659e+00 2.67213732e-01 1.16789833e-01
7.57978857e-01 -4.05942291e-01 7.30248839e-02 1.04589276e-01
9.26564574e-01 -1.53801811e+00 7.35712707e-01 7.39322126e-01
6.37018621e-01 -6.53629959e-01 5.33738613e-01 5.54476857e-01
-6.04862094e-01 -4.02747542e-01 -2.54204065e-01 -4.88697767e-01
-3.62268053e-02 7.21128881e-01 -4.67586964e-01 6.36899114e-01
3.17533255e-01 6.42507374e-01 -1.13209002e-01 6.58575356e-01
-8.07541192e-01 8.97510827e-01 -3.30553293e-01 4.73399274e-02
9.47382897e-02 8.12862515e-02 4.43278730e-01 9.87266123e-01
2.41209134e-01 9.74627361e-02 -4.20578182e-01 1.45859885e+00
3.12767103e-02 -3.86720002e-01 -8.88774097e-01 1.00479104e-01
4.22052950e-01 7.22104132e-01 -3.20249408e-01 -2.92888105e-01
-1.39843062e-01 3.56936932e-01 3.44670713e-01 4.73673522e-01
-8.28878760e-01 3.73376846e-01 5.11423171e-01 -9.84294862e-02
-1.80556312e-01 1.43514335e-01 -5.84151685e-01 -8.93683672e-01
-3.03414851e-01 -8.86550546e-01 4.46364701e-01 -6.18172646e-01
-1.27109790e+00 -9.76921022e-02 3.82810622e-01 -6.80228472e-01
-4.29338664e-01 -7.54223526e-01 -4.94869292e-01 9.66047227e-01
-1.17224669e+00 -9.60330307e-01 1.18953973e-01 5.87639391e-01
-6.51481599e-02 3.76397699e-01 9.23766553e-01 7.38636926e-02
-6.83911085e-01 2.65766203e-01 -1.35042384e-01 -7.00442791e-02
2.93756962e-01 -1.55388856e+00 -1.02352448e-01 8.59534919e-01
-6.18103594e-02 1.06783819e+00 1.03583682e+00 -7.10538626e-01
-1.44048774e+00 -1.03530657e+00 9.92942750e-01 -6.26390159e-01
1.16528213e+00 -3.87658089e-01 -7.90423036e-01 9.14132118e-01
9.33912396e-02 -3.39525849e-01 2.82145858e-01 5.51117957e-01
-5.72158158e-01 -1.63111149e-03 -7.51088738e-01 7.90972173e-01
1.16048586e+00 -3.54686379e-01 -8.26595485e-01 1.69404522e-01
8.34384680e-01 -2.02394854e-02 -8.94455791e-01 5.67781806e-01
7.28907824e-01 -1.01900685e+00 1.02679384e+00 -1.18762898e+00
1.09301531e+00 -1.09983727e-01 -1.57192722e-01 -1.45265830e+00
-2.47783586e-01 -3.31894726e-01 -2.95566946e-01 1.09676504e+00
5.33329129e-01 -7.81737149e-01 1.91784859e-01 7.21434057e-01
6.64830506e-02 -5.24470210e-01 -8.80191386e-01 -8.19146156e-01
5.71022272e-01 -7.23360956e-01 6.44297183e-01 1.33227110e+00
2.22582012e-01 4.22198713e-01 -4.18234408e-01 5.10806680e-01
6.66980624e-01 -5.23265190e-02 3.97481799e-01 -1.53826058e+00
-4.01623130e-01 -6.86628163e-01 -1.11658692e-01 -6.16109669e-01
5.31057000e-01 -9.43397284e-01 -7.62342811e-02 -1.34869611e+00
3.23727995e-01 -3.84034276e-01 -4.16727066e-01 6.01568341e-01
-5.71365878e-02 -2.36031607e-01 -5.35759144e-02 -9.02378839e-03
1.75382584e-01 3.07743371e-01 1.09125733e+00 -2.77540758e-02
-7.12192059e-02 -1.74682718e-02 -7.12224007e-01 9.19168413e-01
8.11584175e-01 -5.55679142e-01 -4.96186435e-01 -5.93953393e-02
8.46939385e-01 4.20485526e-01 1.20754015e+00 -4.56488818e-01
1.55598536e-01 -4.74724323e-01 1.27072155e-01 -1.64767936e-01
-2.06500560e-01 -8.48013639e-01 3.75247747e-01 6.55905962e-01
-9.48396444e-01 -1.63803548e-01 1.21550202e-01 4.77531254e-01
2.44283024e-02 -3.73727113e-01 4.28855002e-01 1.04816584e-02
-4.25270081e-01 -1.01852126e-03 -4.31109041e-01 -4.87795994e-02
4.66253817e-01 2.48014808e-01 -4.00246233e-01 -3.15320671e-01
-7.54330337e-01 5.65505810e-02 -8.88145715e-02 2.13476032e-01
3.85226011e-01 -1.37538445e+00 -7.54483879e-01 -9.54863802e-02
-1.00605875e-01 -2.88850486e-01 1.03663646e-01 1.02292180e+00
1.02819182e-01 6.18926167e-01 -3.44160870e-02 -2.87624300e-01
-6.21837378e-01 8.26578081e-01 5.27130663e-01 -1.44889429e-01
-3.96110445e-01 4.16228652e-01 5.91396511e-01 -3.31076294e-01
9.05236006e-02 -4.98532295e-01 -1.79612413e-01 -4.23361324e-02
4.08649057e-01 1.69685945e-01 -2.71790713e-01 -1.36791155e-01
-1.72904074e-01 1.36620132e-03 4.71571952e-01 -2.20263749e-01
1.32022381e+00 1.32225484e-01 -4.42761898e-01 8.67810369e-01
8.84478986e-01 -3.52273822e-01 -1.10068405e+00 1.77971706e-01
1.65767428e-02 2.10312679e-02 2.56198347e-02 -1.08629525e+00
-6.03279710e-01 8.91229987e-01 4.60613042e-01 6.33308530e-01
1.15880191e+00 -2.06626523e-02 -1.47568926e-01 3.44796479e-01
9.95747522e-02 -5.99299729e-01 -4.91286576e-01 2.78856903e-01
1.15052474e+00 -9.44072723e-01 -2.88411856e-01 -1.67833760e-01
1.60423815e-01 1.06298888e+00 2.35734418e-01 -2.84973085e-01
6.21515274e-01 2.87777364e-01 -6.63474500e-01 -2.90226489e-01
-9.45583761e-01 -1.49834603e-01 2.82803595e-01 5.02997160e-01
4.53909338e-01 4.23326552e-01 -5.61030447e-01 7.02836394e-01
-2.91693687e-01 1.50360599e-01 5.42678058e-01 3.27119559e-01
-1.80496663e-01 -7.79919744e-01 -4.01682556e-01 3.92225653e-01
-3.11156303e-01 -3.11165273e-01 -4.97074485e-01 1.15955412e+00
2.12634459e-01 8.83362591e-01 8.17154497e-02 -2.71002233e-01
1.61867127e-01 1.04081906e-01 5.20794630e-01 -3.46901864e-01
-2.15012982e-01 -3.21806490e-01 8.53043273e-02 -6.07164800e-01
-8.45007777e-01 -6.47820950e-01 -1.07110250e+00 -4.83180612e-01
-1.03995644e-01 -1.14158122e-02 3.65825236e-01 1.19737363e+00
-2.08340973e-01 7.83608019e-01 4.15674776e-01 -4.34773952e-01
-5.20639718e-01 -9.47886467e-01 -4.93425250e-01 1.93480954e-01
5.78934252e-01 -9.55192447e-01 -6.11986101e-01 2.53698885e-01] | [8.136590003967285, 5.437344074249268] |
68c0bd20-760c-4cf9-9441-24904927f22b | tinysiamese-network-for-biometric-analysis | 2307.00578 | null | https://arxiv.org/abs/2307.00578v1 | https://arxiv.org/pdf/2307.00578v1.pdf | TinySiamese Network for Biometric Analysis | Biometric recognition is the process of verifying or classifying human characteristics in images or videos. It is a complex task that requires machine learning algorithms, including convolutional neural networks (CNNs) and Siamese networks. Besides, there are several limitations to consider when using these algorithms for image verification and classification tasks. In fact, training may be computationally intensive, requiring specialized hardware and significant computational resources to train and deploy. Moreover, it necessitates a large amount of labeled data, which can be time-consuming and costly to obtain. The main advantage of the proposed TinySiamese compared to the standard Siamese is that it does not require the whole CNN for training. In fact, using a pre-trained CNN as a feature extractor and the TinySiamese to learn the extracted features gave almost the same performance and efficiency as the standard Siamese for biometric verification. In this way, the TinySiamese solves the problems of memory and computational time with a small number of layers which did not exceed 7. It can be run under low-power machines which possess a normal GPU and cannot allocate a large RAM space. Using TinySiamese with only 8 GO of memory, the matching time decreased by 76.78% on the B2F (Biometric images of Fingerprints and Faces), FVC2000, FVC2002 and FVC2004 while the training time for 10 epochs went down by approximately 93.14% on the B2F, FVC2002, THDD-part1 and CASIA-B datasets. The accuracy of the fingerprint, gait (NM-angle 180 degree) and face verification tasks was better than the accuracy of a standard Siamese by 0.87%, 20.24% and 3.85% respectively. TinySiamese achieved comparable accuracy with related works for the fingerprint and gait classification tasks. | ['Adel M. ALIMI', 'Habib Chabchoub', 'Tarek M. Hamdani', 'Islem Jarraya'] | 2023-07-02 | null | null | null | null | ['face-verification'] | ['computer-vision'] | [-1.48398150e-02 -3.78497601e-01 6.56705871e-02 -3.20699722e-01
-9.54942033e-02 -3.00749153e-01 8.00741613e-02 -3.22121888e-01
-7.17664361e-01 5.15115440e-01 -7.56497085e-01 -3.21734190e-01
-8.83295164e-02 -8.87784004e-01 -6.10486865e-01 -7.83612728e-01
-1.73783749e-01 1.66604757e-01 8.84199291e-02 -1.42602682e-01
3.45423818e-01 1.06740963e+00 -1.82513523e+00 -1.21539518e-01
7.26576924e-01 1.49582613e+00 -2.93623269e-01 6.05290115e-01
1.19097978e-01 8.80616978e-02 -6.86379552e-01 -8.46211553e-01
3.16632926e-01 8.06128830e-02 -3.51106644e-01 -2.85587817e-01
7.92415023e-01 -2.03219980e-01 -2.02560320e-01 9.70013320e-01
8.60018313e-01 -2.20690295e-01 4.70632344e-01 -1.51036453e+00
-3.49730432e-01 4.69517568e-03 -9.09142494e-01 -6.06447831e-02
2.31958479e-01 -2.84356382e-02 1.98851481e-01 -8.69607270e-01
3.46767068e-01 9.97887194e-01 1.15393150e+00 6.89206839e-01
-7.33092308e-01 -1.30153573e+00 -7.27643669e-01 2.19614774e-01
-1.63805079e+00 -2.92304337e-01 5.95073164e-01 -2.53663838e-01
8.40211093e-01 2.03663826e-01 6.70499265e-01 8.49229455e-01
3.00897062e-01 1.84758380e-01 9.62260365e-01 -3.69185090e-01
-6.99417591e-02 -8.83679315e-02 2.85016447e-01 1.11156845e+00
6.70937598e-01 2.32090607e-01 -3.58978629e-01 -1.91224545e-01
8.35752249e-01 -3.68858725e-02 6.81347121e-03 1.09954163e-01
-7.82080531e-01 4.79137808e-01 2.49250486e-01 2.65518039e-01
-3.11637580e-01 7.04406723e-02 7.29106545e-01 2.38878831e-01
-1.46712467e-01 9.25044436e-03 -4.46444660e-01 -3.28030616e-01
-1.00963545e+00 -5.11948504e-02 7.62526214e-01 8.12499821e-01
6.76163316e-01 3.07416081e-01 2.73343593e-01 9.11955953e-01
2.19323754e-01 1.19172394e+00 2.88227707e-01 -5.63512683e-01
5.05676985e-01 6.12244785e-01 -2.45446544e-02 -1.58629096e+00
-5.30527711e-01 -2.19576195e-01 -1.40312207e+00 3.08805376e-01
7.70888388e-01 -1.22966096e-01 -9.37735558e-01 1.36256349e+00
1.10604666e-01 5.73306195e-02 -1.73602745e-01 8.48434508e-01
8.62144530e-01 4.38480526e-01 1.04161631e-03 2.18786672e-01
1.50677049e+00 -5.80303788e-01 -5.29618084e-01 2.86802173e-01
3.18850398e-01 -1.09167516e+00 1.05421591e+00 4.56897438e-01
-7.53059208e-01 -1.00345659e+00 -1.31392050e+00 2.01695338e-01
-6.23785734e-01 8.12369585e-01 7.81153202e-01 1.44010866e+00
-9.00157928e-01 7.65559256e-01 -6.23904765e-01 -3.76759261e-01
6.53323889e-01 1.00648487e+00 -6.92944288e-01 5.86211681e-02
-1.02982438e+00 6.02395475e-01 1.80418000e-01 5.12361228e-01
-2.56858706e-01 -4.62069988e-01 -8.37731659e-01 1.59255758e-01
-1.22380443e-01 -1.28996810e-02 5.12993634e-01 -7.91278541e-01
-1.57325268e+00 8.35597754e-01 -1.30491003e-01 -1.65572897e-01
3.54210258e-01 1.54480249e-01 -8.58800054e-01 6.82436675e-02
-1.15336493e-01 4.24432725e-01 7.23906159e-01 -4.35118288e-01
-4.02992547e-01 -5.87099135e-01 -3.97910297e-01 -3.86562049e-01
-4.20638144e-01 1.03062123e-01 -5.27597189e-01 -4.61810142e-01
1.22976601e-01 -1.13185191e+00 2.96222627e-01 1.20124631e-01
-2.10269868e-01 -1.45498648e-01 1.02448702e+00 -7.68324673e-01
9.52646673e-01 -2.07891798e+00 -5.98589838e-01 7.49163449e-01
-1.90026715e-01 9.04100239e-01 -1.91871136e-01 -1.24122743e-02
-1.85105070e-01 -8.68805051e-02 6.38663247e-02 6.64903894e-02
-1.02740332e-01 6.05473556e-02 1.60205022e-01 6.66764259e-01
3.33600752e-02 7.87907660e-01 -3.60492647e-01 -5.99090695e-01
1.88444838e-01 6.87549710e-01 -3.48960876e-01 -2.05070496e-01
6.88321471e-01 -4.91871359e-03 -1.99939221e-01 1.20339799e+00
1.38745832e+00 -1.07425004e-01 -5.29530868e-02 -5.77933431e-01
-7.79340714e-02 -4.68814313e-01 -1.45567214e+00 1.28417909e+00
-3.61759365e-01 6.73387885e-01 2.79294029e-02 -1.09881389e+00
1.16359544e+00 3.10785592e-01 4.60882068e-01 -8.72791171e-01
3.72727990e-01 6.00645721e-01 5.04164249e-02 -6.54948175e-01
2.32780054e-01 1.51820198e-01 3.19029875e-02 4.34406340e-01
1.55800954e-01 3.73335600e-01 1.18662380e-01 -2.41285846e-01
5.48632443e-01 -3.83463204e-02 -8.41941535e-02 -2.75640428e-01
1.01046920e+00 -3.43676448e-01 7.08807409e-01 5.25105059e-01
-1.79085791e-01 3.70324492e-01 3.24028909e-01 -6.89489841e-01
-1.04924667e+00 -7.53839016e-01 -4.08629894e-01 4.88386154e-01
1.05577791e-02 -1.27549291e-01 -8.17916930e-01 -5.50277412e-01
2.64080405e-01 -3.26261580e-01 -5.59723377e-01 9.72726196e-02
-7.40443766e-01 -7.85534799e-01 1.32167041e+00 5.59135973e-01
1.23798573e+00 -9.10134733e-01 -5.50937772e-01 -2.11459026e-02
2.49918163e-01 -1.14808452e+00 -5.31474769e-01 -2.64805347e-01
-8.66624951e-01 -1.51150656e+00 -9.08299923e-01 -9.87624884e-01
8.07861030e-01 -7.20332265e-02 7.05126584e-01 3.05632114e-01
-6.47338808e-01 -8.36709812e-02 9.38610509e-02 -4.43625957e-01
2.03996137e-01 1.55021340e-01 2.88840383e-01 1.80971965e-01
6.40053749e-01 -3.64266783e-01 -5.00962198e-01 5.04453182e-01
-5.22309721e-01 -2.13730693e-01 6.05077446e-01 1.23782372e+00
2.14562237e-01 -3.56903151e-02 5.05447865e-01 -6.07605875e-01
3.50974351e-01 5.85669316e-02 -9.57067072e-01 3.48259598e-01
-7.94491589e-01 -2.37303421e-01 8.32087219e-01 -4.96681184e-01
-7.02490926e-01 2.65223473e-01 -1.86424375e-01 -3.28499407e-01
8.12265351e-02 2.19093040e-01 -7.26000369e-02 -8.63265693e-01
4.81592923e-01 2.50976533e-01 3.48573804e-01 -4.17710125e-01
-2.61007339e-01 8.86982083e-01 6.52925432e-01 -5.90180576e-01
8.14357102e-01 2.69981951e-01 4.56337839e-01 -1.03405845e+00
-2.42174007e-02 -1.06872439e-01 -5.01355648e-01 -2.80838639e-01
6.39529526e-01 -4.74487990e-01 -1.72202015e+00 1.20553267e+00
-1.04840708e+00 2.59505481e-01 2.48146519e-01 5.61897218e-01
7.97157884e-02 6.86128318e-01 -6.84510112e-01 -7.14713514e-01
-8.53537261e-01 -1.12127864e+00 8.92536044e-01 7.48903513e-01
8.32343698e-02 -6.55872107e-01 -5.39849460e-01 4.50805545e-01
7.62184203e-01 4.83129710e-01 6.40815616e-01 -1.37124643e-01
-1.75054297e-01 -7.84526646e-01 -6.09645844e-01 3.41106951e-01
1.78362966e-01 4.17045474e-01 -9.81644750e-01 -4.71027136e-01
-2.70097166e-01 -3.36779147e-01 3.48629087e-01 2.20520020e-01
1.36258900e+00 -1.78944185e-01 -9.83047262e-02 9.33307409e-01
1.47247016e+00 6.60364985e-01 8.91559780e-01 2.30831370e-01
4.50385362e-01 4.49061841e-01 4.99852687e-01 2.28555530e-01
1.24511451e-01 7.17181683e-01 4.01102938e-02 -3.79765719e-01
4.57829908e-02 9.93956774e-02 -1.25129567e-02 6.82573140e-01
-6.38177693e-01 2.38018885e-01 -1.05670595e+00 1.40021876e-01
-1.27522600e+00 -9.27681506e-01 -2.58768201e-01 2.18841386e+00
4.44923490e-01 -2.54371017e-01 4.74411510e-02 6.66895866e-01
8.55863452e-01 -2.22407296e-01 -3.85419577e-01 -5.95857739e-01
-1.58161059e-01 9.56002295e-01 7.55517423e-01 1.63378641e-01
-1.11647904e+00 7.37513244e-01 5.51876497e+00 8.36589813e-01
-1.62156153e+00 -7.25855604e-02 5.55358112e-01 1.37637571e-01
7.82276154e-01 -5.18023372e-01 -8.47890615e-01 7.39912689e-01
9.36452925e-01 3.64358515e-01 2.58341253e-01 7.76309252e-01
-3.48862186e-02 -2.98991084e-01 -7.20316589e-01 1.55659139e+00
7.85070646e-04 -1.18019736e+00 -2.63167500e-01 1.32984787e-01
4.93354350e-01 -3.55570108e-01 1.93249062e-01 5.94347939e-02
-4.37148958e-01 -1.27826023e+00 1.96221828e-01 3.71971905e-01
1.27197134e+00 -9.48687851e-01 1.33655834e+00 -8.96707848e-02
-1.37553298e+00 1.33265421e-01 -6.53910398e-01 1.76048115e-01
-3.97687316e-01 3.73669297e-01 -3.31859916e-01 5.15866995e-01
9.00855064e-01 3.52903485e-01 -5.60886502e-01 9.04453695e-01
2.32947454e-01 3.89798701e-01 -5.22419691e-01 -3.42395037e-01
1.09402388e-01 -2.33135015e-01 -7.44315237e-02 1.21751261e+00
5.67786157e-01 1.97241083e-02 -4.45163965e-01 3.22469831e-01
-3.92791908e-03 1.28584802e-01 -2.51987040e-01 4.28334773e-02
3.82802963e-01 1.17723572e+00 -5.86609542e-01 -3.59057516e-01
-2.93236643e-01 8.88464451e-01 -2.79349923e-01 1.53441235e-01
-1.03295267e+00 -1.19361520e+00 4.84845072e-01 -8.81770775e-02
2.39393309e-01 -7.10152835e-02 -4.07108665e-01 -1.19940007e+00
3.17258269e-01 -1.04289806e+00 1.70472577e-01 -3.17978412e-01
-9.47067857e-01 7.35704064e-01 -4.57444131e-01 -1.30199575e+00
-8.83911997e-02 -1.08683813e+00 -5.42715788e-01 1.20767272e+00
-1.26657772e+00 -1.06679893e+00 -9.84788716e-01 9.56330657e-01
-2.07788348e-01 -6.48594677e-01 1.07126534e+00 9.29194570e-01
-7.85053372e-01 1.34613085e+00 -4.53463569e-02 6.64950013e-01
5.33005357e-01 -6.35025084e-01 3.46937299e-01 7.79155850e-01
-2.97526807e-01 8.67162883e-01 9.64245647e-02 -4.15024221e-01
-1.68159902e+00 -6.95774317e-01 1.00398767e+00 1.83193833e-01
1.74566910e-01 -3.58382761e-01 -6.72683656e-01 3.84053700e-02
-2.83601314e-01 4.42841470e-01 7.42572248e-01 -2.25131270e-02
-4.27511901e-01 -6.13560617e-01 -1.61927605e+00 2.78796524e-01
9.10608292e-01 -6.62137985e-01 1.29030153e-01 -2.88871909e-03
-3.74424517e-01 -4.59604532e-01 -1.11768806e+00 4.20236588e-01
1.41781831e+00 -9.06712174e-01 8.97308528e-01 -2.98181415e-01
6.88728467e-02 -3.44867438e-01 -5.61049953e-02 -4.28709507e-01
1.68465693e-02 -2.48673275e-01 5.33372611e-02 1.21247041e+00
3.29750299e-01 -1.02724338e+00 1.28381646e+00 5.78547359e-01
4.34630930e-01 -8.50507617e-01 -9.75269258e-01 -1.06720066e+00
-2.77584702e-01 -2.23382354e-01 7.88925886e-01 8.97853434e-01
-4.16885525e-01 -2.11167544e-01 -5.11261046e-01 1.53947845e-01
8.54051054e-01 -5.82597032e-02 9.36204255e-01 -1.38011074e+00
7.03056529e-02 -2.12517604e-01 -9.78381336e-01 -5.68174720e-01
-1.56482920e-01 -5.88516653e-01 -6.00420117e-01 -9.21883166e-01
-6.62729219e-02 -7.84244001e-01 -2.13128284e-01 8.03555250e-01
2.27456793e-01 7.77251720e-01 8.59475285e-02 1.83596641e-01
1.94436431e-01 1.47824481e-01 1.20368063e+00 -2.87638754e-01
-3.05575654e-02 1.71197072e-01 -2.53386229e-01 4.06767219e-01
8.50723028e-01 -2.86419004e-01 -1.19043157e-01 -1.77518323e-01
-1.80446774e-01 -2.20531672e-02 2.78028220e-01 -1.26888275e+00
5.23078680e-01 2.26202250e-01 9.35857534e-01 -6.65141404e-01
4.57936406e-01 -8.46683502e-01 4.26764339e-01 9.14753973e-01
3.59496981e-01 1.84349492e-01 4.32369471e-01 -2.61469912e-02
-4.08238143e-01 -9.31965411e-02 9.08633471e-01 4.51306522e-01
-5.87728918e-01 5.12714326e-01 -1.41971931e-01 -5.26138008e-01
9.78199542e-01 -8.33282053e-01 -3.37555349e-01 4.44098143e-03
-4.28641915e-01 -1.95349827e-01 2.04662874e-01 2.79529661e-01
6.53192759e-01 -1.57266355e+00 -4.60216612e-01 7.11049199e-01
-1.85968369e-01 -3.30469072e-01 4.07399952e-01 8.30535233e-01
-1.15377665e+00 7.12834954e-01 -8.76557529e-01 -6.57376349e-01
-1.76336324e+00 7.32678398e-02 4.99641746e-01 -6.23979664e-04
-2.38795072e-01 6.64170623e-01 -5.71871519e-01 -4.75850165e-01
3.03667039e-01 -1.39159799e-01 -2.48126552e-01 4.50479146e-03
4.55153763e-01 7.20433533e-01 3.11339796e-01 -7.54967034e-01
-7.50439107e-01 1.22580707e+00 1.20401695e-01 2.50919431e-01
1.24055016e+00 5.64381659e-01 -2.85342157e-01 -4.16029781e-01
1.52781177e+00 -2.23546363e-02 -7.04359412e-01 1.99607402e-01
-2.15286743e-02 -6.10335708e-01 -2.87376434e-01 -7.08208621e-01
-1.65198874e+00 9.85881925e-01 1.32668281e+00 -1.37633502e-01
1.28611696e+00 -7.02522874e-01 8.83709252e-01 5.17831743e-01
6.00168049e-01 -1.21182024e+00 -2.36070842e-01 4.82685328e-01
7.40757167e-01 -1.18170619e+00 -4.61068898e-02 -3.99257600e-01
-1.00541845e-01 1.72025740e+00 7.45032191e-01 -2.63601184e-01
8.41301262e-01 3.60253632e-01 8.38905424e-02 -1.34738311e-01
1.66096821e-01 2.87860692e-01 6.03661537e-01 5.54929614e-01
5.17423153e-01 4.96223271e-02 -5.71697891e-01 4.15375233e-01
-3.82342666e-01 2.10983813e-01 1.09126747e-01 9.00711775e-01
2.14366809e-01 -1.19227481e+00 -4.98756289e-01 5.53509057e-01
-7.39828646e-01 1.76422790e-01 -7.06019700e-02 1.01925945e+00
5.71384370e-01 8.34977150e-01 5.30101657e-02 -7.67394722e-01
3.48083973e-01 -7.88511112e-02 5.63902497e-01 1.17388293e-01
-7.30478883e-01 -3.11100483e-01 -9.68465433e-02 -7.11158931e-01
-4.66857672e-01 -3.60852152e-01 -9.55106914e-01 -9.39351022e-01
-3.89244199e-01 1.59729704e-01 1.04843664e+00 5.63914597e-01
3.91240329e-01 1.25700399e-01 5.57517529e-01 -5.98811507e-01
-3.66488278e-01 -9.37829792e-01 -7.60160327e-01 1.12843603e-01
-4.54503745e-02 -6.65331602e-01 -6.80048242e-02 -2.13622972e-01] | [13.302651405334473, 0.9484583139419556] |
a5826382-126e-4c74-959c-cadc1b1a1a8a | tmr-text-to-motion-retrieval-using | 2305.00976 | null | https://arxiv.org/abs/2305.00976v1 | https://arxiv.org/pdf/2305.00976v1.pdf | TMR: Text-to-Motion Retrieval Using Contrastive 3D Human Motion Synthesis | In this paper, we present TMR, a simple yet effective approach for text to 3D human motion retrieval. While previous work has only treated retrieval as a proxy evaluation metric, we tackle it as a standalone task. Our method extends the state-of-the-art text-to-motion synthesis model TEMOS, and incorporates a contrastive loss to better structure the cross-modal latent space. We show that maintaining the motion generation loss, along with the contrastive training, is crucial to obtain good performance. We introduce a benchmark for evaluation and provide an in-depth analysis by reporting results on several protocols. Our extensive experiments on the KIT-ML and HumanML3D datasets show that TMR outperforms the prior work by a significant margin, for example reducing the median rank from 54 to 19. Finally, we showcase the potential of our approach on moment retrieval. Our code and models are publicly available. | ['Gül Varol', 'Michael J. Black', 'Mathis Petrovich'] | 2023-05-02 | null | null | null | null | ['motion-synthesis', 'moment-retrieval', 'text-to-3d'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 4.58274223e-02 -5.35583496e-01 -4.53217953e-01 9.38486960e-03
-1.35623527e+00 -7.11462319e-01 1.02787590e+00 -2.34053940e-01
-5.45866370e-01 3.73061657e-01 6.74142659e-01 6.60138801e-02
-1.51847541e-01 -2.40462720e-01 -5.00966012e-01 -6.86478317e-01
-7.12386519e-02 6.81412399e-01 4.90362734e-01 -2.06705406e-01
1.10831790e-01 5.51980257e-01 -1.49319947e+00 2.09679037e-01
2.49814585e-01 6.67553544e-01 -9.09658819e-02 9.62769628e-01
3.08537334e-01 8.29384029e-01 -5.60079575e-01 -3.28734070e-01
1.05962083e-01 -2.73517460e-01 -1.00798488e+00 -1.64893538e-01
6.39133692e-01 -6.36562467e-01 -6.09265745e-01 4.68943685e-01
1.08057308e+00 4.37113434e-01 9.35587525e-01 -1.12761879e+00
-2.02103198e-01 1.80800930e-01 -3.34452063e-01 1.57114238e-01
8.19371343e-01 1.57480910e-01 1.16749167e+00 -9.38924372e-01
1.08400524e+00 1.50795674e+00 4.78445858e-01 7.04497337e-01
-1.03589821e+00 -7.36090019e-02 -8.63085464e-02 1.37155369e-01
-1.42262518e+00 -6.76923275e-01 5.42182624e-01 -4.48059887e-01
1.15619743e+00 4.14680332e-01 4.89139736e-01 1.34965730e+00
1.84727043e-01 1.25368142e+00 6.70737684e-01 -4.79018420e-01
-5.67765757e-02 -3.31181318e-01 2.53033936e-02 6.80388689e-01
2.98892614e-03 8.26265067e-02 -8.65459442e-01 -2.64132261e-01
5.59215009e-01 -4.54850942e-01 -1.31569088e-01 -6.80023670e-01
-1.57870233e+00 6.31286860e-01 2.37098578e-02 2.76967883e-01
-1.26882687e-01 6.61782563e-01 5.25407612e-01 1.04443416e-01
5.79043567e-01 -1.94339287e-02 -1.32470429e-01 -5.45714378e-01
-1.34897411e+00 7.30679333e-01 6.91040874e-01 7.93079257e-01
3.21021497e-01 -2.39703968e-01 -4.67380434e-01 9.15154576e-01
4.53839809e-01 7.71031022e-01 5.03035724e-01 -1.40752983e+00
4.05273676e-01 -2.33581699e-02 2.48260334e-01 -1.08832467e+00
-2.04665527e-01 -1.82541296e-01 -4.50768530e-01 -1.18167326e-01
3.54727715e-01 2.62282670e-01 -8.75346780e-01 1.71647704e+00
2.55329818e-01 3.02502699e-03 -7.63876513e-02 1.08282685e+00
7.72633851e-01 6.93069041e-01 -4.17566486e-02 2.03532465e-02
1.19332552e+00 -1.28352571e+00 -8.34050417e-01 -1.05475679e-01
9.39248800e-01 -1.04637706e+00 1.17392182e+00 3.12528342e-01
-1.41048658e+00 -2.33679667e-01 -1.02692461e+00 -4.91922259e-01
-2.24112481e-01 2.51302779e-01 6.31724536e-01 4.22502697e-01
-1.22131717e+00 7.59336531e-01 -1.07212412e+00 -6.75135493e-01
-8.33851174e-02 1.74266160e-01 -4.75268275e-01 -8.86776224e-02
-1.27448452e+00 8.88140857e-01 1.31115705e-01 -1.11313917e-01
-8.45575988e-01 -4.14829642e-01 -7.32150614e-01 -2.85519570e-01
3.53297025e-01 -1.24517810e+00 1.44751227e+00 -8.20057094e-02
-1.70408332e+00 1.10202730e+00 -2.27848545e-01 -3.86127502e-01
1.14096642e+00 -6.44957006e-01 -1.43755093e-01 6.72394872e-01
8.74944031e-02 1.06124115e+00 7.33911991e-01 -1.10870528e+00
-3.42708528e-01 5.47244959e-02 1.16947539e-01 3.93844515e-01
-1.13838308e-01 1.61034778e-01 -1.18273842e+00 -8.78749907e-01
-1.35989413e-01 -1.30796802e+00 1.66590232e-02 -1.05140742e-03
-2.28612199e-01 -5.09126522e-02 6.61066711e-01 -5.54245591e-01
1.53145909e+00 -1.92144454e+00 4.85199064e-01 9.17464271e-02
-1.49989268e-02 -1.54295582e-02 -2.22285300e-01 6.36897922e-01
1.27065390e-01 7.11177662e-02 -2.93387145e-01 -8.62239480e-01
3.75130147e-01 1.96696147e-01 -3.47141296e-01 6.80499315e-01
-9.14054289e-02 1.04949927e+00 -8.81100595e-01 -9.05659556e-01
4.70525146e-01 6.38356030e-01 -6.87024295e-01 -1.20310932e-01
-1.88802257e-01 2.61634409e-01 -2.49188587e-01 6.34961605e-01
1.87325537e-01 -2.87809104e-01 -1.96591094e-01 -1.24688014e-01
1.17630452e-01 3.44931364e-01 -1.03657663e+00 2.25529218e+00
-2.82153130e-01 7.48253047e-01 -3.33083391e-01 -3.35068107e-01
3.37341368e-01 3.55537772e-01 7.86657989e-01 -6.67427003e-01
-3.21214125e-02 2.74679065e-01 -3.82406443e-01 -3.71863127e-01
1.05425286e+00 1.44888043e-01 -4.06741232e-01 4.57636654e-01
-1.20892160e-01 -1.41311511e-01 3.70468378e-01 4.42258209e-01
1.05074573e+00 5.68181396e-01 1.92666519e-02 -1.48288175e-01
4.32227284e-01 1.35695785e-01 5.66121489e-02 7.63362288e-01
-2.09080175e-01 1.06470466e+00 3.64107996e-01 1.67387724e-02
-1.14982772e+00 -1.03734851e+00 -2.18477957e-02 9.99754012e-01
5.78398742e-02 -6.75894201e-01 -7.49572814e-01 -5.79681695e-01
-8.51287395e-02 4.96633768e-01 -5.28772116e-01 7.41411373e-02
-8.30649555e-01 -6.55255020e-01 8.97814035e-01 4.60689455e-01
2.37362534e-01 -8.16254497e-01 -8.12565088e-01 -3.99083681e-02
-5.87715209e-01 -1.25656807e+00 -7.94954658e-01 -3.51596117e-01
-8.19801807e-01 -5.84321320e-01 -1.30013311e+00 -4.70377177e-01
1.50772631e-01 3.20836753e-01 1.13196599e+00 6.32957891e-02
-3.82267296e-01 8.37860405e-01 -4.64734316e-01 9.69687477e-02
-2.76650578e-01 3.48406345e-01 3.45322825e-02 -4.89814728e-01
-1.53312329e-02 -2.53301233e-01 -8.21332455e-01 3.37875724e-01
-1.16938722e+00 -5.35844117e-02 3.81709307e-01 7.44572282e-01
5.09736776e-01 -4.39152271e-01 -3.29285935e-02 -4.85112727e-01
5.60396552e-01 -1.86763704e-01 -3.54791701e-01 9.17803496e-02
-4.14656729e-01 3.45294148e-01 -1.43783882e-01 -4.04417485e-01
-7.93991327e-01 2.12855078e-02 -2.33731210e-01 -6.66835546e-01
1.35586709e-01 5.06812096e-01 -2.09975317e-02 1.49998456e-01
5.03551483e-01 9.62860063e-02 -1.83770627e-01 -5.41052818e-01
7.07208157e-01 2.64764279e-01 5.98458171e-01 -6.56594038e-01
8.01549256e-01 8.06271136e-01 2.87217945e-01 -8.07737291e-01
-4.68380123e-01 -7.66505122e-01 -6.93934619e-01 -2.62408733e-01
7.56210744e-01 -9.13181901e-01 -5.63316822e-01 3.81842583e-01
-1.09375441e+00 -5.80488801e-01 -2.19135329e-01 5.49280107e-01
-8.89794827e-01 7.23078012e-01 -8.97762060e-01 -6.81075454e-01
-2.96069950e-01 -1.08827066e+00 1.74039662e+00 -4.67125177e-01
-5.51242709e-01 -1.06290483e+00 5.14990509e-01 3.60990137e-01
2.85874605e-01 2.78791934e-01 5.56664526e-01 -4.35960501e-01
-6.87076688e-01 -1.98061943e-01 1.07676640e-01 -7.68327788e-02
-2.34177068e-01 1.43611014e-01 -8.59178245e-01 -3.91564399e-01
-4.73870337e-01 -2.45325163e-01 1.30278707e+00 4.15922940e-01
5.81950963e-01 -5.68649881e-02 -3.78163397e-01 5.42148709e-01
1.21615589e+00 -2.21940413e-01 8.94719362e-01 5.79841673e-01
6.67673230e-01 6.50472522e-01 8.64104152e-01 4.80722487e-01
4.03145820e-01 1.14774537e+00 -5.44339493e-02 -1.06017785e-02
-4.27189916e-01 -2.12305576e-01 5.46783864e-01 1.05210936e+00
-1.87245727e-01 -5.82420826e-01 -9.32025790e-01 5.55778384e-01
-2.02398229e+00 -1.10819161e+00 4.94169593e-02 2.09853792e+00
6.04872823e-01 7.32924491e-02 3.22172850e-01 2.49428004e-01
2.35265031e-01 6.15914881e-01 1.05520226e-02 -1.47464603e-01
-2.22456411e-01 2.81157829e-02 4.39807832e-01 9.08009768e-01
-1.22911501e+00 1.32171154e+00 7.31812191e+00 1.06899011e+00
-9.32609379e-01 2.90600248e-02 1.40725821e-01 -5.04075408e-01
-2.57871181e-01 1.40203219e-02 -7.87178576e-01 1.54666871e-01
8.89852524e-01 8.94561782e-02 1.83452725e-01 4.93942678e-01
3.18722188e-01 -2.55295366e-01 -1.15215790e+00 1.10250139e+00
2.78271139e-01 -1.09114707e+00 3.36792767e-01 1.43586770e-01
5.54764688e-01 -7.03787282e-02 2.61956811e-01 9.65217575e-02
-9.23757348e-03 -9.14242208e-01 1.08583748e+00 8.13315809e-01
7.23620176e-01 -7.10757494e-01 5.28118134e-01 8.14107731e-02
-1.27915478e+00 4.84074950e-01 -2.22822148e-02 3.39705288e-01
5.22558868e-01 3.33105236e-01 -4.05849516e-01 7.98906624e-01
5.39549708e-01 7.32650280e-01 -7.22463250e-01 9.72266853e-01
-1.04656257e-01 4.04147863e-01 -4.81485695e-01 1.21039070e-01
3.00269157e-01 1.09113172e-01 9.50366199e-01 1.59384775e+00
2.99609452e-01 -2.04893872e-01 1.23381324e-01 3.89422774e-01
1.50141582e-01 1.96411416e-01 -6.10561132e-01 -1.27631083e-01
1.54334202e-01 9.32578325e-01 -5.79876065e-01 -3.85584205e-01
-4.26681750e-02 1.36317384e+00 -7.88090676e-02 3.59095931e-01
-9.68751788e-01 -3.05431783e-01 3.65494192e-01 1.09908611e-01
2.68925130e-01 -6.05236530e-01 2.21157789e-01 -1.27666795e+00
6.49654865e-02 -7.93762624e-01 4.33314264e-01 -8.21769655e-01
-9.92025018e-01 3.94913286e-01 5.58976114e-01 -1.44476688e+00
-9.11815107e-01 -4.97299045e-01 8.65295529e-02 5.79934776e-01
-1.43055296e+00 -1.26280248e+00 -7.85078853e-02 5.02906978e-01
3.68671328e-01 1.63462713e-01 6.93338156e-01 6.64528966e-01
-3.52607310e-01 6.84643447e-01 4.40294705e-02 -5.57054728e-02
1.09379303e+00 -1.01169932e+00 6.37958288e-01 7.85967350e-01
2.59763539e-01 7.81816125e-01 9.78629470e-01 -6.01535022e-01
-1.52298224e+00 -6.64914370e-01 9.22562122e-01 -8.48148465e-01
6.69502735e-01 -2.63076663e-01 -6.18022025e-01 5.98322570e-01
1.00039743e-01 -1.67277098e-01 4.03665721e-01 -1.86013520e-01
-2.32921764e-01 2.99540818e-01 -6.78914666e-01 9.73798394e-01
1.39072418e+00 -7.14044273e-01 -6.13338709e-01 2.14609131e-01
7.28879333e-01 -6.20592713e-01 -8.10668647e-01 5.89488089e-01
9.36526775e-01 -8.11356723e-01 1.35645175e+00 -3.57619762e-01
4.32985753e-01 -3.65738481e-01 -3.28505397e-01 -8.09861779e-01
-1.32393345e-01 -8.92595768e-01 -3.53175968e-01 9.93488014e-01
1.37849331e-01 -1.60071969e-01 8.78980339e-01 3.14521909e-01
2.70072408e-02 -5.42709172e-01 -8.86771142e-01 -8.00816834e-01
1.65633425e-01 -6.89548850e-01 6.01223931e-02 6.96876347e-01
-2.66797453e-01 3.25780779e-01 -7.65440524e-01 -2.01213107e-01
5.93480766e-01 -1.75711766e-01 1.08845019e+00 -8.09655190e-01
-4.17209238e-01 -5.18261492e-01 -4.10254091e-01 -1.56021786e+00
1.97190836e-01 -8.25920045e-01 9.14402083e-02 -1.66375732e+00
2.93235928e-01 7.96583220e-02 -1.20495283e-03 1.74107030e-01
4.74160649e-02 3.97822976e-01 3.74621540e-01 5.12555480e-01
-9.80966747e-01 5.30777812e-01 1.18692338e+00 -1.31871536e-01
-1.81797132e-01 -2.62956589e-01 -7.17471493e-03 6.02745712e-01
4.28286821e-01 -4.33849812e-01 -4.58288550e-01 -4.82954293e-01
2.03103006e-01 8.03657398e-02 5.45444846e-01 -1.08112633e+00
3.06583881e-01 1.42195061e-01 1.59414843e-01 -1.15676165e+00
8.60292971e-01 -4.35076624e-01 2.54588991e-01 4.23949212e-01
-4.91912603e-01 2.15133056e-01 2.65538782e-01 4.77360040e-01
-2.08328202e-01 -6.56365082e-02 3.90298456e-01 1.26055658e-01
-6.52013183e-01 2.63016373e-01 -4.26457852e-01 1.07393175e-01
3.97738338e-01 -6.07821532e-02 -3.17316473e-01 -7.09117353e-01
-6.82174623e-01 1.44570202e-01 6.92371786e-01 4.80953753e-01
5.28694868e-01 -1.50112033e+00 -7.40920424e-01 -3.26906741e-01
2.83532381e-01 -4.84738410e-01 2.67848283e-01 7.92213798e-01
-8.50380480e-01 8.62735987e-01 1.48000881e-01 -9.17898417e-01
-1.51249313e+00 4.02779043e-01 1.34548888e-01 -5.70654988e-01
-5.84764421e-01 3.75078410e-01 -7.40813389e-02 -9.51401740e-02
4.58690703e-01 -2.58155376e-01 4.76829037e-02 1.08142264e-01
4.04113024e-01 6.14685535e-01 2.37760730e-02 -9.47707236e-01
-6.69398367e-01 9.78696465e-01 1.01547569e-01 -9.25428092e-01
1.05012810e+00 -3.36827993e-01 1.40124157e-01 6.24715805e-01
1.53350532e+00 1.85626313e-01 -9.49483156e-01 -7.87996277e-02
9.10056606e-02 -4.41100627e-01 -1.44372076e-01 -5.19373298e-01
-5.82814217e-01 8.56426060e-01 6.12867892e-01 -1.63360089e-01
9.90091085e-01 7.06359819e-02 9.32566762e-01 5.25925577e-01
3.53795558e-01 -1.11544323e+00 4.06828135e-01 7.60155439e-01
8.40372443e-01 -9.21849132e-01 3.82206589e-01 -2.99785227e-01
-5.28188586e-01 7.46689081e-01 -1.06050996e-02 -1.73471928e-01
4.04284716e-01 1.49315551e-01 -8.38893205e-02 -4.75086533e-02
-8.40480566e-01 -2.78438330e-01 8.34192872e-01 1.93556681e-01
5.44177711e-01 -2.31439471e-01 -3.88371348e-01 1.77184143e-03
-3.04793030e-01 9.65252444e-02 1.22535668e-01 1.27333808e+00
-1.46530584e-01 -1.41951001e+00 -4.15352851e-01 -2.19584048e-01
-7.55782187e-01 -1.64247990e-01 -3.87274116e-01 1.06886542e+00
-4.75516021e-01 6.41022205e-01 -1.51393890e-01 -3.70985389e-01
4.50041085e-01 5.99592775e-02 8.29372823e-01 -2.19402805e-01
-3.18516403e-01 5.58037281e-01 3.44535053e-01 -8.83732915e-01
-7.90899038e-01 -6.35569215e-01 -1.11028969e+00 -4.66637135e-01
-4.38391194e-02 -2.37427264e-01 6.02398634e-01 7.41315961e-01
2.91414589e-01 3.57074499e-01 1.26993880e-01 -1.06924188e+00
-3.80301356e-01 -7.50730932e-01 -3.48778129e-01 5.61901748e-01
4.48942065e-01 -7.68121660e-01 -4.95541900e-01 2.26580560e-01] | [10.151657104492188, 0.8134397864341736] |
257f2cd6-c672-417a-a3c5-e603ce902d77 | maximum-volume-inscribed-ellipsoid-a-new | 1708.02883 | null | http://arxiv.org/abs/1708.02883v3 | http://arxiv.org/pdf/1708.02883v3.pdf | Maximum Volume Inscribed Ellipsoid: A New Simplex-Structured Matrix Factorization Framework via Facet Enumeration and Convex Optimization | Consider a structured matrix factorization model where one factor is
restricted to have its columns lying in the unit simplex. This
simplex-structured matrix factorization (SSMF) model and the associated
factorization techniques have spurred much interest in research topics over
different areas, such as hyperspectral unmixing in remote sensing, topic
discovery in machine learning, to name a few. In this paper we develop a new
theoretical SSMF framework whose idea is to study a maximum volume ellipsoid
inscribed in the convex hull of the data points. This maximum volume inscribed
ellipsoid (MVIE) idea has not been attempted in prior literature, and we show a
sufficient condition under which the MVIE framework guarantees exact recovery
of the factors. The sufficient recovery condition we show for MVIE is much more
relaxed than that of separable non-negative matrix factorization (or pure-pixel
search); coincidentally it is also identical to that of minimum volume
enclosing simplex, which is known to be a powerful SSMF framework for
non-separable problem instances. We also show that MVIE can be practically
implemented by performing facet enumeration and then by solving a convex
optimization problem. The potential of the MVIE framework is illustrated by
numerical results. | ['Chong-Yung Chi', 'Wing-Kin Ma', 'Chia-Hsiang Lin', 'Yue Wang', 'Ruiyuan Wu'] | 2017-08-09 | null | null | null | null | ['hyperspectral-unmixing'] | ['computer-vision'] | [ 6.43368244e-01 2.20193341e-01 -1.32957339e-01 1.63879186e-01
-4.03326064e-01 -7.56226063e-01 2.98257768e-01 -3.57572764e-01
-1.32584810e-01 8.06652784e-01 8.74982476e-02 -4.28950340e-01
-7.22421288e-01 -6.32771254e-01 -6.75279558e-01 -1.33718038e+00
-1.24901086e-01 6.49106026e-01 -3.77626479e-01 -2.01038703e-01
6.38054386e-02 7.57516682e-01 -1.52134299e+00 -1.48016691e-01
1.23616552e+00 8.49056780e-01 3.40607852e-01 4.04051661e-01
1.28955441e-02 2.29485836e-02 -2.20886350e-01 -1.32054552e-01
6.74140871e-01 -1.83325648e-01 -6.35279357e-01 7.62127757e-01
2.56948292e-01 1.56391874e-01 1.89956561e-01 1.36974013e+00
9.26435888e-02 3.46872300e-01 7.90900111e-01 -1.43041277e+00
-2.99075633e-01 3.06513011e-01 -1.08339953e+00 -1.38523191e-01
3.48132811e-02 -6.68873966e-01 9.91428792e-01 -1.08160472e+00
4.85894203e-01 1.11223662e+00 4.63792652e-01 9.13420022e-02
-1.29770553e+00 -3.67576152e-01 1.38247460e-01 -1.26011789e-01
-1.53382027e+00 -2.86351413e-01 4.94937807e-01 -6.42505348e-01
5.65044701e-01 7.51939535e-01 6.14017606e-01 1.44755706e-01
-1.43691033e-01 6.86705887e-01 1.29257727e+00 -6.48061454e-01
1.80648580e-01 2.10594341e-01 3.49788874e-01 5.56637108e-01
6.11962497e-01 2.91228369e-02 -2.38016590e-01 -4.37903166e-01
7.71267653e-01 9.55015421e-02 -7.95740783e-01 -7.87287414e-01
-1.35920000e+00 1.12322474e+00 2.04156607e-01 2.24995106e-01
-3.37126106e-01 -2.90566504e-01 -3.56145483e-03 2.91702390e-01
6.40301406e-01 3.87064725e-01 -1.54701233e-01 3.34446251e-01
-1.10275185e+00 1.91071525e-01 1.00143731e+00 7.39435792e-01
8.89759660e-01 2.68744789e-02 1.53248593e-01 6.72192931e-01
3.37307066e-01 9.06302571e-01 -2.97403604e-01 -7.61258960e-01
5.67138731e-01 3.80483031e-01 4.68281865e-01 -9.86486256e-01
-4.52418953e-01 -5.04464447e-01 -1.11584425e+00 2.04867244e-01
4.09132838e-01 -5.27661294e-02 -5.72899222e-01 1.40840745e+00
4.63383943e-01 3.18590909e-01 1.18336640e-01 1.04063153e+00
2.88803607e-01 9.59586799e-01 -6.45440876e-01 -8.38532627e-01
1.16279411e+00 -6.24553561e-01 -6.38078928e-01 4.56552617e-02
4.05328035e-01 -6.27980411e-01 6.42455935e-01 8.10120106e-01
-7.79532552e-01 -2.39047129e-02 -1.01312959e+00 3.31752867e-01
6.09210581e-02 3.30997258e-01 9.05201554e-01 9.47444499e-01
-9.78118420e-01 2.71477312e-01 -6.98533416e-01 -1.94380552e-01
9.84953120e-02 4.83137697e-01 -6.17463291e-01 -1.87466666e-01
-7.41872609e-01 5.41092992e-01 2.31579930e-01 5.46084344e-01
-3.85445923e-01 -6.43727720e-01 -7.36659110e-01 1.49043668e-02
7.07946122e-01 -6.17356002e-01 6.56765759e-01 -9.27132607e-01
-1.09853971e+00 6.78731143e-01 -6.59221709e-01 -2.62250572e-01
2.73410022e-01 -2.36735418e-02 -3.56525570e-01 3.03877532e-01
1.74328476e-01 9.60712880e-03 1.23552835e+00 -1.43912959e+00
-2.46643782e-01 -7.23076224e-01 7.98370093e-02 1.15354680e-01
-1.68977365e-01 -4.35821712e-02 -1.47338569e-01 -5.70215404e-01
6.18494272e-01 -1.27055550e+00 -6.78093374e-01 -2.23295793e-01
-3.98967445e-01 -6.88618124e-02 6.08350635e-01 -5.00857949e-01
1.23662210e+00 -2.05552530e+00 6.03504062e-01 9.16609824e-01
3.23885411e-01 1.65196165e-01 7.08431974e-02 5.12074351e-01
-4.68923509e-01 -2.51791663e-02 -6.69375360e-01 -1.59105346e-01
-7.59621188e-02 2.13192597e-01 -5.63985825e-01 9.77774441e-01
-8.58473331e-02 4.11968678e-01 -7.93946683e-01 3.69043201e-02
1.26399964e-01 4.02344942e-01 -5.28278589e-01 -2.76680917e-01
-9.89564583e-02 2.68928707e-01 -3.34122032e-01 7.02592075e-01
1.29295468e+00 -2.81846523e-01 1.12130433e-01 -3.65859777e-01
-4.09236878e-01 -3.38366568e-01 -1.91338432e+00 1.33762205e+00
-2.87602007e-01 3.86094928e-01 8.82832706e-01 -1.27622020e+00
5.87686002e-01 2.84157246e-01 7.81743467e-01 5.89710427e-03
-1.99454501e-01 1.76237687e-01 -2.63621300e-01 -2.50172019e-01
6.49179757e-01 -6.29433811e-01 2.14501664e-01 4.72559482e-01
-3.77862573e-01 -6.23815656e-02 3.88015389e-01 3.25529933e-01
5.07118583e-01 -2.81476051e-01 4.46339637e-01 -8.15989316e-01
6.23092890e-01 5.75740747e-02 4.73490059e-01 4.56739455e-01
2.42420450e-01 4.59740579e-01 4.72803801e-01 1.56671181e-02
-6.10010803e-01 -1.00833607e+00 -5.98711729e-01 6.90616548e-01
2.47921899e-01 -4.82789665e-01 -5.48267961e-01 -1.67851686e-01
7.45106712e-02 2.89911747e-01 -6.71419621e-01 4.98681873e-01
-1.74615428e-01 -1.13369870e+00 -1.81099121e-02 1.42296061e-01
2.01608092e-01 -2.43562624e-01 -1.89841151e-01 5.52029684e-02
-2.26905406e-01 -8.42788994e-01 -3.06508601e-01 3.40758115e-01
-9.91239667e-01 -1.15135181e+00 -6.76795900e-01 -5.09337723e-01
8.86951149e-01 8.90925646e-01 8.89795542e-01 -2.83872634e-01
-9.60944146e-02 4.99141216e-01 -2.55323291e-01 -3.42388064e-01
4.35949750e-02 -3.01846415e-01 3.68933767e-01 4.48548317e-01
1.52629510e-01 -5.01896322e-01 -3.63490552e-01 6.13305867e-01
-1.32712972e+00 4.27789055e-02 3.06658864e-01 7.98725247e-01
8.36910665e-01 4.19943094e-01 4.05170500e-01 -9.99627590e-01
4.34034437e-01 -6.20085657e-01 -9.11894083e-01 3.94667536e-01
-6.03269219e-01 5.06361350e-02 2.22539470e-01 8.22336879e-03
-8.52387369e-01 3.47070783e-01 1.26445010e-01 -4.58393008e-01
2.48056561e-01 8.22578669e-01 -5.61566174e-01 -3.13317329e-01
4.34084207e-01 1.67453229e-01 7.92419314e-02 -4.91118252e-01
4.71366674e-01 6.47354901e-01 4.20024723e-01 -6.98296249e-01
1.13626659e+00 1.03621185e+00 4.72418815e-01 -1.30481672e+00
-7.97915161e-01 -1.05370092e+00 -6.38724089e-01 -1.81604862e-01
6.92422926e-01 -9.54907775e-01 -6.94185138e-01 2.33830418e-02
-8.01851153e-01 1.30257010e-01 -1.90229177e-01 6.05579197e-01
-5.46941400e-01 6.95812643e-01 -3.60570140e-02 -1.18810475e+00
-9.14138854e-02 -1.06354415e+00 1.05685425e+00 -3.19250934e-02
2.35496596e-01 -1.00578117e+00 2.75702655e-01 5.43958664e-01
2.89590713e-02 3.94331396e-01 7.84503520e-01 -8.24156776e-02
-4.48275477e-01 2.09005997e-02 -3.19407135e-01 3.21532577e-01
9.25137401e-02 -2.28745281e-03 -7.35147059e-01 -4.16242033e-01
3.73283267e-01 1.96742356e-01 1.18798327e+00 7.34559476e-01
8.59205902e-01 -2.12214291e-01 -7.31261849e-01 8.56262505e-01
1.70536685e+00 -3.84447388e-02 4.96306986e-01 1.19735926e-01
7.95469463e-01 5.89631617e-01 6.19079471e-01 5.97513914e-01
-2.84666102e-02 5.79192102e-01 5.53220272e-01 -2.24676892e-01
6.33806765e-01 3.03450018e-01 1.81758299e-01 4.19322044e-01
-5.21948576e-01 -2.82996058e-01 -6.64162755e-01 4.72244948e-01
-1.92898691e+00 -1.14629185e+00 -6.16557479e-01 2.43827987e+00
2.81379461e-01 -4.70689118e-01 2.00966090e-01 4.96885955e-01
7.65211105e-01 1.07166022e-01 -2.47781605e-01 -3.49887997e-01
-3.89923424e-01 1.43006846e-01 7.34901845e-01 8.84036243e-01
-1.17816210e+00 3.77811074e-01 5.94646168e+00 7.82478750e-01
-7.95221627e-01 1.98284298e-01 6.26567379e-02 -7.71715716e-02
-6.92400277e-01 9.22497362e-02 -8.17098558e-01 7.80860856e-02
3.80252212e-01 -7.62440711e-02 5.82619846e-01 5.74054122e-01
3.61491948e-01 -2.82502681e-01 -7.50479281e-01 1.19048023e+00
1.37247607e-01 -1.20171785e+00 -1.13907121e-01 6.40583754e-01
1.16099060e+00 -1.71869189e-01 1.20547459e-01 -1.81820080e-01
-2.09620707e-02 -1.07126081e+00 3.23921323e-01 2.99875170e-01
8.57447684e-01 -1.03102195e+00 2.86186934e-01 4.47220922e-01
-1.44964242e+00 -2.94763986e-02 -5.77829182e-01 -1.60207734e-01
3.52214336e-01 1.30101478e+00 -5.89168251e-01 1.01620364e+00
3.67228806e-01 7.62263060e-01 5.18761538e-02 1.15487897e+00
1.89282253e-01 5.43222249e-01 -6.68801188e-01 3.01354975e-01
3.11666369e-01 -1.29514444e+00 1.07231045e+00 1.01709890e+00
5.03983498e-01 3.27607781e-01 3.13113302e-01 5.86667836e-01
2.29646504e-01 3.61674726e-01 -6.12000227e-01 -2.16786712e-01
-3.21592875e-02 1.24938262e+00 -1.00459456e+00 -9.30482447e-02
-6.00353777e-01 8.47132266e-01 -4.62689362e-02 5.73659122e-01
-4.74355400e-01 1.27316033e-02 8.34866345e-01 3.32609624e-01
4.53673571e-01 -3.58383745e-01 -2.36969694e-01 -1.55965602e+00
1.76967680e-01 -7.54686534e-01 3.73730004e-01 -3.17978472e-01
-1.18235266e+00 4.68592048e-01 1.73141360e-01 -1.30168164e+00
-9.20844078e-03 -8.93643200e-01 -4.72898364e-01 1.05255032e+00
-1.39834440e+00 -7.94074953e-01 -1.05064943e-01 9.17810798e-01
5.10151908e-02 4.88343760e-02 7.62531400e-01 1.30068094e-01
-6.14847422e-01 -8.32540318e-02 6.97757542e-01 -3.09568733e-01
2.39667401e-01 -1.35569251e+00 -3.34377766e-01 1.09009123e+00
2.42069975e-01 8.29777598e-01 8.21993470e-01 -5.52471936e-01
-1.76645887e+00 -9.84552443e-01 5.93755901e-01 -2.06391379e-01
8.03781688e-01 -2.73213118e-01 -6.92864597e-01 5.91251791e-01
-2.21334212e-02 4.30372655e-02 9.99005914e-01 2.80582041e-01
-1.95079610e-01 -3.89972813e-02 -8.74710917e-01 2.10537404e-01
9.13415074e-01 -2.71809161e-01 -2.27017969e-01 8.76161158e-01
3.17225367e-01 -1.98306143e-01 -8.05332601e-01 5.06383836e-01
4.06467795e-01 -8.13290358e-01 1.12076557e+00 -4.90445286e-01
6.19277954e-02 -8.13360870e-01 -5.77159166e-01 -1.24269974e+00
-2.93667704e-01 -7.65694976e-01 -2.16720417e-01 7.28937924e-01
4.32442456e-01 -7.67520189e-01 7.72965431e-01 2.71270961e-01
-1.14152253e-01 -8.75129998e-01 -1.01398253e+00 -9.34718847e-01
-8.48587006e-02 -4.26870227e-01 3.45750898e-01 9.85715806e-01
8.11645910e-02 2.72299021e-01 -4.83026743e-01 5.38992107e-01
7.82212317e-01 6.14418626e-01 5.28946996e-01 -1.56876791e+00
-5.70651174e-01 -2.75606036e-01 -1.13838695e-01 -1.13012624e+00
2.81721860e-01 -1.02600157e+00 -1.20163597e-01 -1.67375851e+00
1.32975370e-01 -7.13623226e-01 1.13698944e-01 6.50864020e-02
-4.36505713e-02 2.20081404e-01 2.73228347e-01 2.07041711e-01
-1.86603993e-01 4.91451174e-01 1.27406323e+00 -7.18615055e-02
-4.26839471e-01 4.70077813e-01 -8.00950229e-01 5.73370159e-01
4.01962966e-01 -2.30947062e-01 -6.51211679e-01 -1.08835280e-01
6.73171222e-01 3.04170489e-01 2.18860656e-01 -6.59553051e-01
-9.06707197e-02 -2.52787471e-01 7.85413384e-02 -6.91581011e-01
4.99848008e-01 -1.16272044e+00 5.72066545e-01 2.14757875e-01
3.13806593e-01 -1.55710012e-01 5.39698564e-02 8.27527761e-01
-1.46938011e-01 -5.32922804e-01 5.11963606e-01 -6.37683794e-02
-4.21742827e-01 3.31136554e-01 -3.72625798e-01 -3.10216218e-01
1.03648436e+00 -4.68489856e-01 5.36894575e-02 -2.68530965e-01
-1.04348576e+00 7.73297995e-02 3.03659081e-01 -1.31959170e-01
3.57127845e-01 -1.17893994e+00 -8.82563412e-01 2.63206303e-01
7.65133277e-03 -4.94751036e-02 2.55557418e-01 1.27872431e+00
-3.12497288e-01 5.69033027e-01 2.48310536e-01 -8.44135404e-01
-1.21561158e+00 7.10508049e-01 1.92772329e-01 -2.41018444e-01
-5.17668307e-01 8.76410604e-01 4.71039146e-01 -1.84522852e-01
-6.95627406e-02 -3.77201021e-01 -3.79616357e-02 3.68231833e-01
7.16234565e-01 6.27307594e-01 2.26141915e-01 -8.13079476e-01
-2.45194584e-01 6.49782896e-01 2.54436255e-01 -2.70264000e-01
1.41822183e+00 -1.70703232e-01 -7.57564068e-01 1.90928102e-01
1.04455483e+00 5.85030973e-01 -1.03055799e+00 -2.97534585e-01
-4.64736074e-01 -5.16018391e-01 1.53752387e-01 -2.85124719e-01
-1.14382958e+00 6.29395902e-01 2.42113158e-01 4.42350596e-01
1.47635758e+00 -1.38160869e-01 3.00670683e-01 4.23277289e-01
4.68130350e-01 -7.10636497e-01 -6.63743019e-01 3.42271656e-01
1.01101613e+00 -1.09888375e+00 1.28046021e-01 -1.03664041e+00
-3.58723044e-01 1.10607505e+00 6.05759397e-02 -1.67567834e-01
1.05475104e+00 3.48717749e-01 -4.11351025e-01 -2.00361356e-01
-3.18845838e-01 -5.28787673e-01 4.94314224e-01 5.62166572e-01
1.28354713e-01 5.19054174e-01 -4.08087611e-01 4.78301674e-01
-1.55620784e-01 -2.28518814e-01 5.83910465e-01 6.29417002e-01
-5.83416760e-01 -9.73241985e-01 -9.72645402e-01 7.22962737e-01
-1.73455343e-01 -2.53007591e-01 -1.42403990e-01 5.58064699e-01
1.79281160e-01 1.04839456e+00 -1.29433811e-01 -4.41585388e-03
-1.67600885e-02 -2.52455473e-01 6.62473559e-01 -6.10425413e-01
-1.37465954e-01 3.55932027e-01 -7.28730261e-02 -2.80130893e-01
-8.34685624e-01 -8.36102009e-01 -1.04217374e+00 -1.88641414e-01
-7.63057351e-01 6.42326176e-01 6.28584564e-01 1.13838601e+00
-4.24763560e-02 -7.00203627e-02 6.99268579e-01 -9.66029286e-01
-2.72868246e-01 -7.38433719e-01 -1.20976758e+00 -2.40575001e-01
5.86127520e-01 -8.58217180e-01 -6.56488895e-01 -2.53093522e-02] | [10.073305130004883, -1.9274266958236694] |
32049717-af72-48a8-8e1f-c0733ff07be2 | hotgp-higher-order-typed-genetic-programming | 2304.03200 | null | https://arxiv.org/abs/2304.03200v1 | https://arxiv.org/pdf/2304.03200v1.pdf | HOTGP -- Higher-Order Typed Genetic Programming | Program synthesis is the process of generating a computer program following a set of specifications, which can be a high-level description of the problem and/or a set of input-output examples. The synthesis can be modeled as a search problem in which the search space is the set of all the programs valid under a grammar. As the search space is vast, brute force is usually not viable and search heuristics, such as genetic programming, also have difficulty navigating it without any guidance. In this paper we present HOTGP, a new genetic programming algorithm that synthesizes pure, typed, and functional programs. HOTGP leverages the knowledge provided by the rich data-types associated with the specification and the built-in grammar to constrain the search space and improve the performance of the synthesis. The grammar is based on Haskell's standard base library (the synthesized code can be directly compiled using any standard Haskell compiler) and includes support for higher-order functions, $\lambda$-functions, and parametric polymorphism. Experimental results show that, when compared to $6$ state-of-the-art algorithms using a standard set of benchmarks, HOTGP is competitive and capable of synthesizing the correct programs more frequently than any other of the evaluated algorithms. | ['Emilio Francesquini', 'Fabrício Olivetti de França', 'Matheus Campos Fernandes'] | 2023-04-06 | null | null | null | null | ['program-synthesis'] | ['computer-code'] | [ 2.36951739e-01 2.70932317e-01 -4.71716732e-01 -3.91061604e-01
-5.29720306e-01 -7.91692734e-01 3.60750854e-01 -3.29996599e-03
1.37395084e-01 6.85143113e-01 -3.76301467e-01 -8.25836658e-01
9.14682895e-02 -1.30625093e+00 -8.98121953e-01 -2.91521519e-01
-3.27256799e-01 5.48771262e-01 3.91983151e-01 -3.02925557e-01
4.57834363e-01 5.37982434e-02 -2.07440782e+00 2.87259251e-01
1.08231044e+00 6.75302446e-01 1.85519010e-01 8.13833714e-01
-6.07156932e-01 3.28184068e-01 -5.45993745e-01 -3.30686748e-01
2.54518569e-01 -6.16661549e-01 -7.98917770e-01 -2.37298474e-01
-6.66545928e-02 1.31240755e-01 1.95724860e-01 1.29162610e+00
1.91582665e-01 -2.38992766e-01 6.51476607e-02 -1.65107679e+00
-2.47787714e-01 8.95256639e-01 -1.12221263e-01 -3.59644860e-01
4.01332021e-01 3.41759324e-01 1.09244049e+00 -6.59995794e-01
6.28915548e-01 1.37324274e+00 4.44192231e-01 5.93574226e-01
-1.61640525e+00 -4.36969519e-01 -2.54954517e-01 -5.76474488e-01
-1.47570920e+00 -4.02542710e-01 2.86803603e-01 -5.49699724e-01
1.54814756e+00 5.69034398e-01 5.97706556e-01 2.83794045e-01
4.18963730e-01 2.72889972e-01 8.96610737e-01 -8.86446595e-01
4.17417347e-01 9.08136144e-02 2.33454093e-01 1.10615993e+00
4.13380235e-01 5.44670403e-01 -2.56224990e-01 -8.10096145e-01
3.94590914e-01 -6.43060446e-01 -1.22156732e-01 -6.05812073e-01
-1.13661885e+00 8.22753072e-01 -6.38871565e-02 1.31187635e-03
1.74580634e-01 6.56961918e-01 5.58685184e-01 4.83235180e-01
-1.72667667e-01 7.26672351e-01 -4.91107523e-01 -2.18635947e-01
-7.79956222e-01 6.16013527e-01 1.43421555e+00 1.27428651e+00
1.04151297e+00 2.50368655e-01 6.86407536e-02 5.08014083e-01
4.69699025e-01 4.25823480e-01 1.75008491e-01 -9.12242651e-01
4.32826161e-01 1.12251866e+00 -1.42273843e-01 -6.36466086e-01
-3.01765697e-03 -4.93985936e-02 -9.81152356e-02 6.12055659e-01
2.28975460e-01 -2.49128342e-01 -7.71590233e-01 1.68100858e+00
2.35964298e-01 -2.99932182e-01 1.47167727e-01 4.48286533e-01
5.12258470e-01 9.97566581e-01 -2.71935642e-01 -3.32579315e-01
1.07277083e+00 -9.29769635e-01 -2.33140692e-01 -4.75500882e-01
1.07051957e+00 -4.86204982e-01 9.90150869e-01 2.83624738e-01
-1.34441829e+00 -2.91649580e-01 -1.29909849e+00 1.81941807e-01
-3.15891773e-01 -1.33767590e-01 9.51929152e-01 1.17684841e+00
-1.44510698e+00 5.17830968e-01 -7.38496065e-01 -1.57357588e-01
-9.56078917e-02 7.99981713e-01 -2.10241508e-02 -1.10112712e-01
-7.86009967e-01 6.03055537e-01 8.71674538e-01 -1.03823379e-01
-6.29151225e-01 -8.01608324e-01 -9.89286959e-01 1.44997612e-01
4.69027698e-01 -7.10466564e-01 1.29087484e+00 -1.04867780e+00
-1.35820377e+00 9.91168022e-01 -6.09960482e-02 -3.09294403e-01
2.12084740e-01 7.02620983e-01 -3.62993479e-01 -4.45862532e-01
-3.02632339e-02 5.76958776e-01 4.40878183e-01 -1.08537340e+00
-6.63447857e-01 -5.52457571e-02 2.42410719e-01 -1.52304038e-01
3.75616848e-01 3.30278546e-01 -5.54401755e-01 -3.53971541e-01
-1.35153875e-01 -1.19826639e+00 -2.58566320e-01 -2.53168941e-01
-2.98120737e-01 -6.93841949e-02 5.73317409e-01 -3.83317769e-01
1.62749851e+00 -2.33333421e+00 1.89255297e-01 7.07502007e-01
-1.27346560e-01 1.70697987e-01 1.79561943e-01 5.44853747e-01
-6.29436746e-02 5.71424305e-01 -5.34457445e-01 4.40508366e-01
5.67830145e-01 5.10197759e-01 -1.55785963e-01 2.35514328e-01
1.05108231e-01 8.28603387e-01 -8.93300176e-01 -4.71059918e-01
-2.22043544e-01 -1.19937457e-01 -1.04571068e+00 1.94668919e-01
-1.19371271e+00 -1.94224909e-01 -4.85515356e-01 6.83024228e-01
4.53029752e-01 -1.02233581e-01 7.20689416e-01 3.97075355e-01
-4.23180431e-01 4.70403433e-01 -1.40329170e+00 1.59786344e+00
-5.09477556e-01 1.65815026e-01 1.72031611e-01 -6.84268832e-01
1.24295366e+00 1.61083043e-01 -2.09236741e-01 -4.00467217e-01
-1.14852123e-01 7.22269833e-01 3.29569459e-01 -3.56491476e-01
2.67365396e-01 1.84823513e-01 -5.95330775e-01 9.50442195e-01
-2.28434622e-01 -6.59182549e-01 8.79553318e-01 -6.75042868e-02
1.18760777e+00 3.84221047e-01 4.34300005e-01 -7.80765355e-01
6.31475687e-01 4.58973289e-01 8.46306503e-01 8.77784550e-01
5.44341445e-01 8.06747302e-02 1.00894713e+00 -4.72841561e-01
-1.15098548e+00 -6.83267474e-01 2.11326610e-02 1.22934794e+00
-3.02511930e-01 -9.74005759e-01 -9.03050661e-01 -5.25743484e-01
-3.04836016e-02 1.04985094e+00 -2.82629788e-01 -5.53381369e-02
-9.16012049e-01 -7.29865849e-01 7.67955720e-01 2.70119548e-01
3.90910476e-01 -1.21393514e+00 -9.12350833e-01 4.33123827e-01
3.20358604e-01 -5.24565458e-01 -5.46795130e-01 3.54385197e-01
-8.41260433e-01 -1.02039289e+00 2.24523365e-01 -1.12963021e+00
1.03945160e+00 -4.96323943e-01 1.51410484e+00 7.64671028e-01
-4.33134288e-01 -9.67126936e-02 -1.06894404e-01 -1.84161052e-01
-1.21050572e+00 -4.23697978e-02 -6.49657845e-01 -4.94346023e-01
-2.99855042e-02 -4.62314904e-01 1.97706029e-01 1.96914643e-01
-9.49007750e-01 1.84008449e-01 3.99142504e-01 1.15653324e+00
5.07076323e-01 6.52627587e-01 2.12831646e-01 -1.18871939e+00
6.00877404e-01 -1.26648098e-01 -1.34212375e+00 5.44812560e-01
-7.33378112e-01 6.67310119e-01 7.09351182e-01 -3.46866809e-02
-8.56832266e-01 3.58870506e-01 -2.82936022e-02 1.83692634e-01
1.18131995e-01 8.79064560e-01 -5.85949540e-01 -3.21356237e-01
8.26777101e-01 3.43614578e-01 2.47449756e-01 -1.25276566e-01
2.25839704e-01 2.54137695e-01 4.22970235e-01 -1.39033628e+00
5.72361946e-01 -3.12499046e-01 1.42325431e-01 -3.59272271e-01
1.75204962e-01 2.07099885e-01 -1.87384129e-01 2.45718494e-01
1.48772806e-01 -2.95146823e-01 -5.77660680e-01 2.38652870e-01
-1.06043506e+00 -7.15486884e-01 -1.69865206e-01 -2.73923695e-01
-6.72732890e-01 -8.50322749e-03 -2.72454083e-01 -6.81442380e-01
-3.13751817e-01 -1.81158972e+00 8.42198968e-01 -9.79978144e-02
-5.36350369e-01 -8.05080652e-01 2.79428586e-02 -3.17661136e-01
4.31928039e-01 3.50647539e-01 1.80113935e+00 -5.08268416e-01
-8.66023481e-01 8.06399807e-03 1.13826700e-01 5.21163084e-02
-1.72823101e-01 5.06702602e-01 -3.09551299e-01 -1.60449266e-01
-2.91339040e-01 -3.57931405e-02 1.39911786e-01 -6.13103807e-02
1.16118288e+00 -6.69638216e-01 -5.69879174e-01 7.93643832e-01
1.66948020e+00 4.50310826e-01 7.96094596e-01 3.13292325e-01
2.01595128e-01 5.50133288e-01 4.52523828e-01 3.73970360e-01
2.16012567e-01 6.04609549e-01 3.74622494e-01 4.20188248e-01
1.86288804e-01 -2.19050720e-01 3.59047323e-01 3.45546931e-01
2.43170202e-01 3.79086137e-02 -1.43875289e+00 4.20495868e-01
-1.84492469e+00 -7.57737815e-01 -1.46580786e-01 2.38184929e+00
1.26737118e+00 2.00328290e-01 2.56514877e-01 3.70846361e-01
7.51673281e-01 -2.36732364e-01 -2.06251055e-01 -1.15821862e+00
2.35032260e-01 5.91398656e-01 6.21373773e-01 6.24975920e-01
-6.61849737e-01 8.32903564e-01 6.86682320e+00 5.73406458e-01
-1.22159040e+00 -3.40367734e-01 2.90360749e-01 1.22864857e-01
-7.58083820e-01 3.82433504e-01 -1.05125320e+00 4.33861345e-01
1.24157786e+00 -9.89029646e-01 9.86082733e-01 9.77488995e-01
-2.93267727e-01 -1.13547049e-01 -1.50971985e+00 4.41971779e-01
-1.19556457e-01 -1.52300477e+00 -3.01515967e-01 5.14166877e-02
7.23013163e-01 -3.83307219e-01 -1.03794828e-01 5.79764247e-01
7.22846210e-01 -1.22014499e+00 9.57827210e-01 -1.32610993e-02
9.00703788e-01 -1.00962448e+00 3.38156283e-01 4.26352292e-01
-1.04893041e+00 -1.91255108e-01 4.15873006e-02 -1.17131341e-02
-1.98091283e-01 3.86287570e-01 -9.59738791e-01 4.82087672e-01
5.11254609e-01 1.39144510e-01 -4.76307809e-01 1.13927710e+00
-3.20892632e-01 3.80982876e-01 -3.40227127e-01 -2.86633402e-01
5.21763712e-02 -9.50828046e-02 3.31991971e-01 1.40141594e+00
6.41456485e-01 1.76525533e-01 3.82810265e-01 1.26095128e+00
1.89544052e-01 4.93823476e-02 -7.08934665e-01 -4.02417421e-01
4.59309399e-01 8.12631667e-01 -6.94589496e-01 -3.30099821e-01
-4.51484740e-01 1.98471367e-01 2.65180059e-02 2.29594260e-01
-7.68765807e-01 -6.97851598e-01 4.06565249e-01 8.28426480e-02
3.54090035e-01 -1.71620771e-01 -5.60114264e-01 -7.98065364e-01
-2.61581037e-03 -1.52433324e+00 3.80235434e-01 -5.74549496e-01
-5.71900487e-01 5.95929325e-01 1.94820642e-01 -4.79843259e-01
-7.15126991e-01 -5.24565458e-01 -6.25355721e-01 1.18431544e+00
-1.02115130e+00 -6.90176964e-01 1.22262254e-01 1.74471155e-01
1.49503708e-01 -9.18646455e-02 1.09245002e+00 3.72884981e-02
-4.20527816e-01 5.51517606e-01 -3.54498863e-01 -2.41383016e-01
1.49825364e-01 -1.12736678e+00 5.99593043e-01 1.02158785e+00
-6.18979931e-01 1.00621998e+00 8.51125240e-01 -7.51345158e-01
-2.24418354e+00 -1.14319932e+00 1.03777516e+00 1.04505010e-01
7.32893407e-01 -6.02300048e-01 -8.11220229e-01 8.30070972e-01
-2.64330655e-02 5.35271540e-02 3.93984437e-01 -1.40527397e-01
-3.97448599e-01 -1.46089509e-01 -1.14540315e+00 6.65126204e-01
1.10432041e+00 -3.06275785e-01 -3.41792643e-01 1.67366296e-01
6.91161275e-01 -8.68745685e-01 -7.46691227e-01 3.79415810e-01
4.87687349e-01 -8.94436777e-01 8.03812385e-01 -2.76480973e-01
3.54169786e-01 -5.19562900e-01 -3.79660606e-01 -1.20426202e+00
-9.59561765e-02 -9.50352907e-01 -7.53733963e-02 1.20474625e+00
8.05580258e-01 -8.07627678e-01 6.88339591e-01 9.24675226e-01
-4.98974741e-01 -6.62440181e-01 -4.22424227e-01 -8.99617195e-01
7.51645043e-02 -2.98646688e-01 1.24686050e+00 6.13901734e-01
4.02279317e-01 1.22851908e-01 1.05226979e-01 1.15954012e-01
5.15769839e-01 7.39775538e-01 8.03416371e-01 -1.04482234e+00
-8.40865135e-01 -7.56998241e-01 -1.56038284e-01 -4.65624988e-01
4.69040602e-01 -1.32120490e+00 3.47696304e-01 -1.17690957e+00
-2.57863700e-01 -1.00366735e+00 5.30388832e-01 7.90401220e-01
2.28456557e-01 -3.43597472e-01 -3.47390249e-02 -1.42286822e-01
-1.20509841e-01 -1.08119264e-01 8.41517448e-01 -1.43428504e-01
-4.19568330e-01 -1.27946794e-01 -7.08278656e-01 4.18134034e-01
7.18424439e-01 -5.55702388e-01 -3.95545423e-01 -3.94681960e-01
7.81344354e-01 5.69409311e-01 1.51039079e-01 -7.78940916e-01
2.98220396e-01 -4.95543391e-01 -4.98597592e-01 -1.77370727e-01
-1.93629310e-01 -5.96846700e-01 9.87344027e-01 7.09735930e-01
-4.27230537e-01 3.46108258e-01 3.71926546e-01 5.36106452e-02
-1.04121849e-01 -6.90092146e-01 6.80423617e-01 -3.68902862e-01
-1.02257836e+00 5.69841750e-02 -2.62752503e-01 2.32870489e-01
1.13141966e+00 -2.13205829e-01 -5.34081042e-01 2.64727801e-01
-4.78754461e-01 3.50904375e-01 1.10480261e+00 7.49016106e-02
3.03274751e-01 -9.91297483e-01 -3.56672794e-01 6.29298449e-01
2.36541301e-01 -6.28892332e-02 -4.85184669e-01 4.71441805e-01
-9.51858699e-01 4.23674405e-01 -1.75958216e-01 -3.87778401e-01
-1.22260308e+00 6.33244157e-01 4.69592094e-01 -1.87274724e-01
-3.57841015e-01 5.39441109e-01 1.33978546e-01 -7.49665260e-01
-8.31224471e-02 -4.04574662e-01 5.48340678e-01 -6.28548920e-01
4.34794068e-01 1.11359566e-01 2.28071526e-01 -2.44262129e-01
-5.80538690e-01 2.46645957e-01 3.93506289e-01 -2.50990212e-01
1.22211790e+00 5.09214342e-01 -9.24778223e-01 1.67300880e-01
9.45398569e-01 -1.04108140e-01 -6.09378695e-01 3.87284979e-02
1.84061080e-01 -5.50110281e-01 -8.74104947e-02 -7.93061972e-01
-9.17373896e-01 5.47411382e-01 -1.61769867e-01 3.38604808e-01
1.03411925e+00 -2.33241066e-01 3.02591860e-01 4.37489718e-01
9.02047098e-01 -7.77180910e-01 -3.59471023e-01 6.10131264e-01
7.26346970e-01 -4.89222825e-01 -1.38273805e-01 -8.37559462e-01
-1.86237127e-01 1.25038528e+00 6.26827419e-01 -1.44567043e-01
3.21122110e-01 9.64646578e-01 -4.07288015e-01 -1.54133094e-02
-1.07710981e+00 1.02810360e-01 -4.07684594e-02 5.03534555e-01
4.47582126e-01 -3.16346660e-02 -5.93551874e-01 5.58044732e-01
-4.49814320e-01 2.10798681e-01 5.96432149e-01 1.53902650e+00
-4.91100609e-01 -1.87498736e+00 -5.48637390e-01 5.57846129e-01
-3.89720857e-01 -1.89778611e-01 -1.95781857e-01 8.59408498e-01
2.19606936e-01 7.31053233e-01 -1.47606224e-01 -6.70928657e-02
1.43033415e-01 2.73698062e-01 7.00234711e-01 -1.18248594e+00
-9.18083549e-01 -8.64907205e-02 7.23567486e-01 -5.33860326e-01
9.07162279e-02 -4.86489028e-01 -1.55915785e+00 -5.87567747e-01
-1.24735340e-01 4.28435385e-01 7.20486999e-01 6.04810596e-01
4.08934563e-01 3.80954444e-01 2.33181044e-01 -4.66927826e-01
-5.31920910e-01 1.52717382e-01 -4.59654599e-01 -2.56003946e-01
-9.72948968e-02 -3.89526039e-01 -1.57003760e-01 2.00134262e-01] | [8.131832122802734, 7.313465118408203] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.