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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
78bef8fe-ebec-45c9-b3ba-089d763df2fb | detclipv2-scalable-open-vocabulary-object | 2304.04514 | null | https://arxiv.org/abs/2304.04514v1 | https://arxiv.org/pdf/2304.04514v1.pdf | DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via Word-Region Alignment | This paper presents DetCLIPv2, an efficient and scalable training framework that incorporates large-scale image-text pairs to achieve open-vocabulary object detection (OVD). Unlike previous OVD frameworks that typically rely on a pre-trained vision-language model (e.g., CLIP) or exploit image-text pairs via a pseudo labeling process, DetCLIPv2 directly learns the fine-grained word-region alignment from massive image-text pairs in an end-to-end manner. To accomplish this, we employ a maximum word-region similarity between region proposals and textual words to guide the contrastive objective. To enable the model to gain localization capability while learning broad concepts, DetCLIPv2 is trained with a hybrid supervision from detection, grounding and image-text pair data under a unified data formulation. By jointly training with an alternating scheme and adopting low-resolution input for image-text pairs, DetCLIPv2 exploits image-text pair data efficiently and effectively: DetCLIPv2 utilizes 13X more image-text pairs than DetCLIP with a similar training time and improves performance. With 13M image-text pairs for pre-training, DetCLIPv2 demonstrates superior open-vocabulary detection performance, e.g., DetCLIPv2 with Swin-T backbone achieves 40.4% zero-shot AP on the LVIS benchmark, which outperforms previous works GLIP/GLIPv2/DetCLIP by 14.4/11.4/4.5% AP, respectively, and even beats its fully-supervised counterpart by a large margin. | ['Hang Xu', 'Zhenguo Li', 'Wei zhang', 'Dan Xu', 'Xiaodan Liang', 'Jianhua Han', 'Lewei Yao'] | 2023-04-10 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Yao_DetCLIPv2_Scalable_Open-Vocabulary_Object_Detection_Pre-Training_via_Word-Region_Alignment_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Yao_DetCLIPv2_Scalable_Open-Vocabulary_Object_Detection_Pre-Training_via_Word-Region_Alignment_CVPR_2023_paper.pdf | cvpr-2023-1 | ['open-vocabulary-object-detection'] | ['computer-vision'] | [ 1.87526103e-02 3.73820215e-02 -4.19725269e-01 -1.67935118e-01
-1.30124390e+00 -5.92284560e-01 7.14445174e-01 1.14768319e-01
-6.67555392e-01 1.81711972e-01 -2.03316569e-01 -2.03628257e-01
4.58283126e-01 -3.00185233e-01 -8.42704296e-01 -5.45474231e-01
3.71546179e-01 5.50474942e-01 5.63314378e-01 -3.01540550e-02
8.00832137e-02 5.92427477e-02 -1.28368700e+00 4.86233421e-02
5.28123140e-01 1.32194161e+00 8.22272420e-01 5.53432286e-01
-2.39803448e-01 4.69558567e-01 -1.08970031e-01 -1.55252084e-01
5.98093450e-01 1.98171392e-01 -4.77525473e-01 2.01777086e-01
1.19391799e+00 -4.53148335e-01 -4.53834802e-01 9.45786536e-01
5.71877420e-01 -8.69778693e-02 8.95127237e-01 -1.02898920e+00
-8.22404802e-01 1.90739036e-01 -1.18714046e+00 2.60989040e-01
9.21335537e-03 5.22357881e-01 1.33342111e+00 -1.69567335e+00
7.23478258e-01 1.14930224e+00 5.85474610e-01 2.11718038e-01
-1.40025663e+00 -8.14246297e-01 2.35521317e-01 -7.50957057e-03
-2.02450395e+00 -1.87862381e-01 3.74018461e-01 -5.12152314e-01
1.08500004e+00 -1.87343180e-01 4.40967143e-01 9.43812907e-01
1.62003249e-01 1.14419293e+00 7.79465854e-01 -4.39730555e-01
-1.22957081e-01 2.73007542e-01 2.29751080e-01 9.24649000e-01
3.52387726e-01 -1.15418516e-01 -4.09504205e-01 8.01366866e-02
8.76242757e-01 2.10891902e-01 -2.09792450e-01 -5.91173947e-01
-1.33068371e+00 9.37248528e-01 5.74497283e-01 5.47181629e-02
-1.66863859e-01 3.18631887e-01 5.44717431e-01 1.82078511e-01
5.34715533e-01 9.48653817e-02 -3.71073842e-01 4.78730261e-01
-1.16446102e+00 2.24958941e-01 2.16923356e-01 1.30206645e+00
9.00050521e-01 4.66046147e-02 -6.16037250e-01 9.75774288e-01
5.88081062e-01 1.15309203e+00 2.28098035e-01 -5.52602172e-01
6.19152665e-01 3.52976322e-01 1.65195093e-02 -7.57106960e-01
-1.84285223e-01 -5.50670743e-01 -6.21415317e-01 5.37197106e-02
1.90429643e-01 6.21252581e-02 -1.45124388e+00 1.32796943e+00
3.65474939e-01 1.65626779e-01 8.17044750e-02 1.05950522e+00
1.02424681e+00 8.54548931e-01 3.70722562e-02 -9.50092264e-03
1.75930846e+00 -1.51875782e+00 -3.60916197e-01 -7.19043672e-01
6.32202327e-01 -9.29085255e-01 1.15544808e+00 1.49278268e-01
-9.05436873e-01 -6.81692362e-01 -1.02060437e+00 -4.30034965e-01
-2.21398041e-01 4.63649750e-01 4.58214730e-02 2.62860775e-01
-1.06574714e+00 -3.88121724e-01 -5.13721526e-01 -4.44295019e-01
6.36072457e-01 1.17098600e-01 -4.30900395e-01 -4.06885117e-01
-6.83215559e-01 6.51521623e-01 6.18595958e-01 -2.44777307e-01
-1.39484346e+00 -9.12218928e-01 -1.08871782e+00 -6.88954443e-02
7.25932360e-01 -6.81754470e-01 1.02106023e+00 -6.13803744e-01
-9.13408041e-01 1.11318910e+00 7.41925137e-03 -6.44428670e-01
4.67644095e-01 -2.24867493e-01 -1.85394287e-02 3.53228152e-01
5.53776801e-01 1.23672104e+00 1.12190998e+00 -1.06347322e+00
-8.40993941e-01 -3.32003057e-01 -2.64437944e-01 2.52846450e-01
-1.88136071e-01 5.29445037e-02 -1.33737087e+00 -7.57908165e-01
-7.24186823e-02 -9.19170797e-01 -2.01151922e-01 4.13047880e-01
-4.64555383e-01 -3.24844509e-01 1.00627923e+00 -4.56800848e-01
7.69980729e-01 -2.08676815e+00 -1.60390988e-01 -2.91680563e-02
6.90220833e-01 5.76975167e-01 -5.45659840e-01 1.65851474e-01
3.02070230e-01 -2.32735023e-01 -1.69100955e-01 -7.75952876e-01
-9.03548971e-02 2.00778931e-01 -3.72532576e-01 6.80992603e-01
3.08324426e-01 1.29489887e+00 -7.48124719e-01 -9.42628801e-01
5.18325150e-01 3.32686335e-01 -4.35213387e-01 8.51971731e-02
-4.90102291e-01 6.30233064e-02 -5.41173398e-01 7.59035707e-01
7.35046089e-01 -5.04048049e-01 -1.91141993e-01 -2.44906738e-01
-8.01969469e-02 -3.56345356e-01 -8.24750900e-01 1.95507038e+00
-5.98629415e-01 9.57623959e-01 1.45782188e-01 -1.04063404e+00
1.00546134e+00 4.33255471e-02 3.21702212e-01 -7.97628641e-01
1.48713440e-01 1.45566180e-01 -5.77656567e-01 -2.41132900e-01
4.91925567e-01 3.10528457e-01 -1.14651129e-01 1.24325074e-01
4.91064608e-01 -2.18517780e-01 -1.04481773e-02 5.43474376e-01
8.81367683e-01 1.41240284e-02 4.61847067e-01 -2.74857342e-01
6.20812595e-01 6.68409318e-02 3.10800493e-01 1.14946187e+00
-2.40138710e-01 8.88715446e-01 2.75945924e-02 -2.82912642e-01
-1.32237911e+00 -1.26083720e+00 -4.10505652e-01 1.04171789e+00
5.22763312e-01 -3.85573417e-01 -3.93920809e-01 -7.82355070e-01
7.26939291e-02 4.79915023e-01 -3.98431748e-01 1.48478225e-01
-1.88194498e-01 -3.85303557e-01 6.95310116e-01 5.27793407e-01
5.63253224e-01 -7.26593435e-01 -3.16812545e-01 1.27072766e-01
-2.19164014e-01 -1.66380012e+00 -9.99827027e-01 2.04233378e-01
-3.95142615e-01 -7.20864475e-01 -9.85010743e-01 -1.21089637e+00
5.33741295e-01 9.78386462e-01 1.09443688e+00 -6.87332675e-02
-7.04307914e-01 5.21918833e-01 -4.18988198e-01 -4.63665605e-01
-2.74906065e-02 -2.09327717e-03 -8.93785208e-02 -3.43960673e-02
3.19896787e-01 -1.07693434e-01 -9.08061028e-01 3.51450443e-01
-7.09294975e-01 1.63221285e-01 8.82891715e-01 1.13656127e+00
1.04774654e+00 -6.84523761e-01 2.98072875e-01 -2.95708627e-01
-3.53478342e-02 -4.11328614e-01 -9.05230403e-01 3.89098465e-01
-8.02639723e-01 -8.36238265e-02 2.45268747e-01 -4.89843518e-01
-7.25531995e-01 2.34560400e-01 -7.76868835e-02 -1.22769070e+00
9.63254273e-02 4.90640625e-02 4.96620722e-02 -3.35834444e-01
7.19441354e-01 7.15759933e-01 -2.94815544e-02 -2.56526351e-01
7.84359097e-01 7.17315197e-01 7.24289417e-01 -3.37453455e-01
1.06875610e+00 6.80051386e-01 -5.32528996e-01 -1.02201474e+00
-9.77159679e-01 -1.21395540e+00 -4.14319038e-01 -9.66469757e-03
1.26783657e+00 -1.57280469e+00 -2.73520470e-01 2.07883313e-01
-1.20244884e+00 -3.16576034e-01 -1.63448334e-01 4.35031027e-01
-4.19082046e-01 4.92031336e-01 -3.73191059e-01 -6.62228167e-01
-7.62914479e-01 -1.19612265e+00 1.80782855e+00 -9.46670100e-02
3.50246847e-01 -7.74241269e-01 -1.20932132e-01 6.97862625e-01
1.31336257e-01 -1.18795708e-01 1.79929316e-01 -6.94728136e-01
-8.20371211e-01 -2.60315090e-01 -9.22644556e-01 5.49075961e-01
-2.33716443e-01 -4.60640967e-01 -8.08536589e-01 -5.18713832e-01
-3.76361281e-01 -7.58852005e-01 1.12815464e+00 4.08039689e-01
8.70221436e-01 -2.77209505e-02 -5.29760599e-01 7.76240528e-01
1.70624459e+00 -2.95801878e-01 3.35527867e-01 1.72564477e-01
1.06002176e+00 1.81984380e-01 1.00484276e+00 3.41253728e-01
5.07961512e-01 9.86413419e-01 5.57116568e-01 -4.40639317e-01
-3.45463961e-01 -2.74803966e-01 4.48032439e-01 2.34998986e-01
4.56366628e-01 -3.30958366e-01 -9.32979345e-01 8.39798152e-01
-1.85049701e+00 -6.64956331e-01 -6.48672059e-02 1.93292463e+00
7.26658523e-01 1.43530220e-01 -6.68032914e-02 -5.56432486e-01
5.89083433e-01 5.61115503e-01 -6.39135540e-01 1.20807692e-01
-1.62458345e-01 4.17024009e-02 1.01919556e+00 4.11692232e-01
-1.22121167e+00 1.43019998e+00 4.89848661e+00 1.46324861e+00
-1.09933329e+00 5.16706288e-01 6.10428512e-01 -3.14777821e-01
4.18533646e-02 -1.32136464e-01 -1.23462260e+00 1.59689844e-01
3.73250157e-01 1.31705165e-01 6.85856268e-02 1.06531680e+00
1.11180134e-01 4.76124585e-02 -8.66016150e-01 1.45735753e+00
3.45001310e-01 -1.66094160e+00 1.16236702e-01 4.72960882e-02
9.18265820e-01 8.66949081e-01 1.24684900e-01 4.20801908e-01
2.25787908e-01 -8.21243167e-01 9.66546655e-01 -1.29012033e-01
1.25509739e+00 -3.14049870e-01 5.39377153e-01 4.13573861e-01
-1.59285080e+00 -1.97475739e-02 -5.64616203e-01 3.70278358e-01
2.17103019e-01 4.43265021e-01 -9.88848567e-01 5.74658930e-01
8.30686808e-01 7.66196907e-01 -5.39718151e-01 7.30751455e-01
-2.19827265e-01 6.63492203e-01 -4.28809941e-01 2.25376725e-01
7.54166543e-01 -1.19895190e-01 6.85086429e-01 1.33599555e+00
1.11256473e-01 -2.00423777e-01 8.14939559e-01 9.32377756e-01
-8.30883831e-02 3.12070757e-01 -5.84300518e-01 2.03042865e-01
5.80369532e-01 1.51304817e+00 -5.20727575e-01 -4.87109542e-01
-8.99026513e-01 1.07079411e+00 3.62687528e-01 3.24858069e-01
-1.06690073e+00 -1.55627280e-01 6.36086226e-01 1.77709162e-01
8.80717456e-01 -2.37499654e-01 8.40459242e-02 -1.13412452e+00
9.07907411e-02 -5.53426623e-01 2.32614219e-01 -9.51319933e-01
-1.12088931e+00 5.59573233e-01 -8.46468005e-03 -1.15082693e+00
1.33904099e-01 -6.07798934e-01 -4.97762710e-01 8.14992905e-01
-1.86975765e+00 -1.77974916e+00 -5.21286726e-01 8.68447661e-01
1.04435766e+00 -2.56187677e-01 3.04812610e-01 2.12610319e-01
-4.40606356e-01 8.75185132e-01 2.58030683e-01 3.65380615e-01
8.78141642e-01 -1.12636900e+00 6.96760774e-01 8.75492811e-01
6.24544919e-01 2.40982816e-01 4.08623397e-01 -5.65053463e-01
-1.32114935e+00 -1.76572573e+00 4.58944112e-01 -5.27959883e-01
7.96774447e-01 -7.87654400e-01 -7.42607951e-01 7.99253464e-01
2.65687078e-01 7.02349901e-01 4.75113690e-02 -3.68504524e-01
-7.37389624e-01 -8.16456527e-02 -8.14436376e-01 5.45657992e-01
1.08306515e+00 -6.45341516e-01 -7.03114271e-01 6.41439855e-01
1.17985165e+00 -4.48331386e-01 -5.97601891e-01 2.22320840e-01
2.71439463e-01 -4.69650209e-01 1.33977485e+00 -1.36649450e-02
7.46058822e-02 -5.42612076e-01 -2.78242916e-01 -6.74609184e-01
-2.03018159e-01 -4.59637851e-01 1.56287253e-01 1.03924537e+00
3.07074398e-01 -6.48277521e-01 6.02215886e-01 -2.31198832e-01
-2.32030198e-01 -7.32674658e-01 -8.67536783e-01 -9.14772153e-01
-1.62777409e-03 -6.46246612e-01 -1.57764420e-01 6.43511295e-01
-5.42774856e-01 6.23910666e-01 -4.91686523e-01 3.68377417e-01
9.31704998e-01 5.52146360e-02 1.05235517e+00 -8.64129782e-01
-2.53329664e-01 -2.52662808e-01 -5.50958991e-01 -1.69770110e+00
5.11783846e-02 -1.07404590e+00 2.70184964e-01 -1.35931897e+00
4.97923344e-01 -4.39403385e-01 -2.36689746e-01 5.41733444e-01
-3.20228428e-01 9.08853471e-01 3.31696242e-01 4.33491260e-01
-1.09660280e+00 6.15239918e-01 1.15635860e+00 -4.98165488e-01
-9.96533111e-02 -5.04199922e-01 -5.43954372e-01 5.91732442e-01
3.77589315e-01 -4.68139827e-01 -5.44966280e-01 -3.49594712e-01
-2.14869678e-01 -1.57756954e-02 7.59310365e-01 -7.69077778e-01
3.32437634e-01 1.23757556e-01 2.45437950e-01 -9.53929842e-01
3.93476367e-01 -5.70186555e-01 -4.33436781e-01 2.39920810e-01
-1.88402057e-01 -3.23063135e-01 1.77906871e-01 1.12857330e+00
-1.87441364e-01 1.99526139e-02 8.35121810e-01 7.90221393e-02
-1.08088589e+00 6.99414790e-01 -1.73952028e-01 3.27945381e-01
1.31116629e+00 -2.53651530e-01 -2.47623444e-01 1.66298430e-02
-3.95663947e-01 8.06384444e-01 4.43900824e-01 6.82298422e-01
8.29169869e-01 -1.07941902e+00 -9.44915771e-01 1.72758311e-01
8.08487058e-01 4.60239649e-01 3.74005556e-01 1.01598227e+00
-4.75307912e-01 6.76402450e-01 2.66051114e-01 -1.26290822e+00
-1.47862041e+00 7.65349090e-01 2.01236218e-01 -3.36076498e-01
-1.13252556e+00 1.05959845e+00 1.12036371e+00 -2.69924343e-01
3.84065151e-01 -2.78465122e-01 4.67034988e-02 -1.14216864e-01
6.33741200e-01 -3.55576992e-01 -1.24500103e-01 -6.47486269e-01
-3.88734519e-01 8.86567473e-01 -5.43760717e-01 2.27635074e-02
9.60865080e-01 -4.09463912e-01 3.71970564e-01 4.27082814e-02
1.34152448e+00 -1.68437555e-01 -1.55416799e+00 -9.20720637e-01
-3.26651067e-01 -4.06486511e-01 4.83241856e-01 -5.46056271e-01
-1.12784290e+00 7.71226823e-01 7.11523175e-01 -5.64384937e-01
8.10232043e-01 5.39818525e-01 8.82716656e-01 3.39522749e-01
3.78259182e-01 -9.50208902e-01 5.47499716e-01 4.20518637e-01
8.19903255e-01 -1.73693478e+00 4.14815396e-02 -3.92848819e-01
-8.25215697e-01 7.57646203e-01 5.76629639e-01 -2.45781064e-01
4.32063192e-01 1.06156260e-01 4.97765951e-02 -2.68055588e-01
-7.94654191e-01 -4.37155992e-01 6.09551132e-01 5.43079138e-01
-1.65704519e-01 -1.22411221e-01 -2.25468092e-02 2.65968144e-01
3.95370662e-01 -3.29026997e-01 6.30768910e-02 6.16390705e-01
-6.98544443e-01 -6.58395231e-01 -4.27617580e-01 3.66119325e-01
-1.19999819e-01 -5.03204525e-01 1.52788963e-02 8.73756528e-01
6.81653246e-02 7.70731747e-01 1.92535773e-01 -1.89090863e-01
5.30158281e-02 -2.73599625e-01 2.02501804e-01 -8.16328764e-01
-3.09060812e-01 3.72722059e-01 -2.12461144e-01 -6.45087063e-01
-1.02188893e-01 -5.34377515e-01 -1.30734003e+00 1.05476968e-01
-6.52643383e-01 -1.88016057e-01 5.21905839e-01 9.27529216e-01
4.58608061e-01 3.91475677e-01 4.41445857e-01 -9.16543722e-01
-6.36886418e-01 -9.11206603e-01 -4.66800511e-01 1.05368294e-01
4.66981500e-01 -7.13933229e-01 -2.16571808e-01 5.91949895e-02] | [9.664897918701172, 1.4925458431243896] |
0b1ab59d-ee55-4cd3-85da-d0fb8b2fb7ea | class-attention-network-for-semantic | null | null | https://ieeexplore.ieee.org/abstract/document/9306448 | https://ieeexplore.ieee.org/abstract/document/9306448 | Class Attention Network for Semantic Segmentation of Remote Sensing Images | Semantic segmentation in remote sensing images is beneficial to detect objects and understand the scene in earth observation. However, classical networks always failed to obtain an accuracy segmentation map in remote sensing images due to the imbalanced labels. In this paper, we proposed a novel class attention module and decomposition-fusion strategy to cope with imbalanced labels. Based on this motivation, we investigate related architecture and strategy by follows. (1) we build a class attention module to generate multi-class attention maps, which forces the network to keep attention to small sample categories instead of being flooded by large sample data. (2) we introduce salient detection, which decomposes semantic segmentation into multi-class salient detection and then fuses them to produce a segmentation map. Extensive experiments on popular benchmarks (e.g., US3D dataset) show that our approach can serve as an efficient plug-and-play module or strategy in the previous scene parsing networks to help them cope with the problem of imbalance labels in remote sensing images. | ['Yuchao Dai', 'Mingyi He', 'Zhibo Rao'] | 2020-12-31 | null | null | null | null | ['scene-parsing'] | ['computer-vision'] | [ 5.24042606e-01 2.77578443e-01 -9.10037458e-02 -7.44684100e-01
-5.84577262e-01 -4.54680502e-01 1.22700058e-01 2.05853581e-01
-2.62051374e-01 2.94978350e-01 -1.22720368e-01 -5.55619478e-01
3.53622586e-02 -1.25251269e+00 -6.67929530e-01 -5.47238231e-01
1.53358087e-01 3.71238887e-01 4.74825323e-01 -1.33636177e-01
2.03795746e-01 4.54116225e-01 -1.69738710e+00 2.95056045e-01
1.17326033e+00 7.79366791e-01 5.81053972e-01 5.51978111e-01
-6.73644066e-01 8.39148402e-01 -5.48864782e-01 2.64154701e-03
3.56443971e-01 -4.80103940e-01 -1.25629222e+00 3.77589792e-01
2.75925040e-01 -3.43003601e-01 2.78750598e-01 1.53289580e+00
4.48975921e-01 1.11867763e-01 4.61894184e-01 -1.14428794e+00
-4.32203740e-01 1.09431207e+00 -9.94447172e-01 4.12189722e-01
-2.55689800e-01 -5.06670363e-02 1.14312470e+00 -7.25744367e-01
2.93293118e-01 1.41874051e+00 4.63595599e-01 6.31487966e-01
-8.26013803e-01 -5.88685215e-01 6.37517631e-01 3.56098358e-03
-1.17717433e+00 -1.89384535e-01 8.89115930e-01 -3.89147162e-01
6.23330534e-01 6.10829532e-01 7.15989888e-01 3.62561464e-01
-3.86862010e-01 1.11450875e+00 8.88162374e-01 -2.15502307e-01
2.49608263e-01 -4.12868597e-02 6.98818624e-01 3.66723657e-01
3.36795062e-01 -5.10699511e-01 2.26561338e-01 2.42052406e-01
6.73279464e-01 2.22132504e-01 -2.20002994e-01 1.19735613e-01
-8.72387350e-01 9.38695967e-01 9.30129349e-01 2.48886332e-01
-3.45008463e-01 4.69076969e-02 1.07551746e-01 1.41823769e-01
9.29106772e-01 2.95269608e-01 -5.77511430e-01 6.52122557e-01
-9.03506696e-01 9.05612335e-02 4.67828035e-01 6.50236845e-01
1.15696597e+00 -3.62047032e-02 1.61599403e-03 8.88948023e-01
3.83580178e-01 4.48361039e-01 2.77414471e-01 -7.63202667e-01
4.77246016e-01 9.43934560e-01 -1.49756014e-01 -1.00937331e+00
-8.15051019e-01 -7.29328752e-01 -9.54387069e-01 2.18902424e-01
1.44630015e-01 -1.25561357e-01 -1.35069311e+00 1.52899611e+00
5.22963703e-01 1.13715112e-01 5.83868064e-02 1.18584943e+00
1.28614450e+00 7.09183395e-01 2.62286961e-01 2.96729177e-01
1.54064941e+00 -1.10218227e+00 -3.50897312e-01 -6.09972894e-01
5.70225775e-01 -2.45949730e-01 1.25238252e+00 5.47583625e-02
-7.95226276e-01 -7.47621119e-01 -9.29637730e-01 9.33080353e-03
-5.30035675e-01 1.22459151e-01 6.34640932e-01 6.71332836e-01
-9.38424408e-01 4.54144776e-01 -8.90057147e-01 -3.69991958e-01
7.22581685e-01 8.78307670e-02 1.68940917e-01 2.11291105e-01
-1.03794527e+00 4.06368077e-01 6.85563922e-01 3.31188172e-01
-8.60636950e-01 -5.40562391e-01 -8.78989100e-01 3.54108900e-01
5.81861615e-01 -6.59114063e-01 1.08428574e+00 -1.38070810e+00
-1.12866890e+00 9.99387920e-01 1.27444953e-01 -3.37031990e-01
3.67799610e-01 -1.47963017e-01 -6.93594990e-03 1.99340954e-01
3.19615632e-01 1.14628482e+00 6.15771413e-01 -1.36360931e+00
-8.75058055e-01 -5.13381958e-01 4.51131582e-01 4.59791273e-01
-4.67349514e-02 2.64890231e-02 -3.14755082e-01 -6.02772355e-01
7.49637425e-01 -4.53296006e-01 -6.37531877e-01 -2.98639953e-01
-7.20898986e-01 -7.85990432e-02 9.00067806e-01 -5.33737123e-01
8.43688369e-01 -1.93873692e+00 5.38393781e-02 1.84836283e-01
3.85991991e-01 7.77745247e-02 -1.82473660e-01 -2.28572711e-01
-2.61766523e-01 6.18757188e-01 -6.04846001e-01 -1.44242615e-01
-2.14663967e-01 2.22152039e-01 -3.78260493e-01 3.21168274e-01
5.88479757e-01 8.42174053e-01 -8.92894447e-01 -6.79724038e-01
1.91480175e-01 8.78560692e-02 -6.55197084e-01 1.73557937e-01
-5.27917981e-01 5.63064814e-01 -7.82457471e-01 9.86796737e-01
9.83698905e-01 -4.64199275e-01 -2.36712303e-02 -8.97625685e-02
3.64303924e-02 1.89967752e-01 -1.16439080e+00 1.48963976e+00
-3.29676755e-02 2.30783686e-01 1.57327175e-01 -1.36677289e+00
9.96078134e-01 -2.16889814e-01 3.18092257e-01 -7.25648999e-01
1.92958593e-01 9.47868302e-02 -1.44791201e-01 -5.61870635e-01
5.86154640e-01 -7.48709291e-02 -1.86613917e-01 3.11350346e-01
-3.56801718e-01 -1.81113243e-01 1.22501053e-01 6.16760738e-02
6.11240804e-01 1.79138958e-01 6.03100434e-02 -3.98762345e-01
5.91573000e-01 3.87737632e-01 7.90938139e-01 9.52001870e-01
-1.37911141e-01 6.76423252e-01 5.74535191e-01 -8.17028940e-01
-8.05899322e-01 -5.21042347e-01 -1.55227175e-02 1.46949220e+00
4.51896042e-01 1.36263087e-01 -9.40667152e-01 -8.60876203e-01
-2.98789442e-01 3.88761461e-01 -6.10702038e-01 2.30557367e-01
-4.26608533e-01 -1.38524568e+00 3.75846416e-01 6.12553179e-01
8.07157636e-01 -1.30100763e+00 -9.41122115e-01 2.45955825e-01
-4.85863596e-01 -8.47863615e-01 1.13225661e-01 2.99380511e-01
-8.90311241e-01 -1.06007719e+00 -6.89495981e-01 -7.84604549e-01
7.43155420e-01 7.46891558e-01 1.18690670e+00 4.03297752e-01
-2.13360414e-01 -1.33747220e-01 -6.64443970e-01 -5.03700435e-01
-2.51220286e-01 4.54989314e-01 -6.35973692e-01 -2.77834386e-02
3.29473972e-01 -3.98737550e-01 -7.51182437e-01 3.65458488e-01
-1.23272574e+00 3.06905001e-01 5.76730907e-01 5.91375768e-01
6.27378404e-01 2.77646571e-01 6.82897985e-01 -1.13987195e+00
1.74532890e-01 -5.79321086e-01 -7.64500380e-01 2.97902793e-01
-2.37844154e-01 -1.14824511e-01 4.18108374e-01 -1.92371923e-02
-1.01003087e+00 2.43999496e-01 -3.41670513e-01 -1.03208534e-01
-4.19722915e-01 7.90557861e-01 -3.77641112e-01 6.67283982e-02
4.98004794e-01 5.32878153e-02 -2.73992389e-01 -7.46580958e-01
2.53631741e-01 9.55285609e-01 3.87910426e-01 -4.00573283e-01
4.46746588e-01 6.28516436e-01 -4.39561665e-01 -7.54953861e-01
-1.31306911e+00 -6.23575866e-01 -5.26005149e-01 3.04931626e-02
1.23992896e+00 -1.12178659e+00 -3.29210430e-01 6.56443655e-01
-9.06189203e-01 -5.61569989e-01 -1.39643803e-01 1.49636492e-01
-1.01459779e-01 3.21203887e-01 -4.66928184e-01 -8.81582916e-01
-5.27924716e-01 -1.09433258e+00 1.29749572e+00 5.62441409e-01
3.60098034e-01 -5.64355850e-01 -2.89207131e-01 4.13035601e-01
3.66913617e-01 2.40616128e-01 7.76189625e-01 -7.68774390e-01
-6.35602057e-01 2.02005878e-01 -7.87112534e-01 3.07174802e-01
-5.14145009e-02 -4.81311157e-02 -1.16717196e+00 -8.82862434e-02
-8.47912282e-02 -2.53459185e-01 1.35102367e+00 5.63194633e-01
1.61658823e+00 -1.53963313e-01 -2.23362356e-01 7.59937763e-01
1.49356043e+00 1.28356501e-01 6.59812748e-01 4.48454678e-01
8.72581899e-01 7.74073124e-01 5.45782208e-01 2.38800481e-01
6.10974431e-01 2.70478010e-01 8.04046631e-01 -7.63446212e-01
5.37149087e-02 1.27019078e-01 -1.18382461e-01 5.13001621e-01
1.04979746e-01 -4.05268282e-01 -1.12097740e+00 6.56652391e-01
-1.84259760e+00 -7.05919981e-01 -3.74723911e-01 1.67419219e+00
5.98178685e-01 8.90527740e-02 1.01929635e-01 1.30539939e-01
9.10276651e-01 3.77094656e-01 -6.26322746e-01 1.46916986e-01
-2.73218453e-01 -4.93993461e-02 6.24517143e-01 4.95552689e-01
-1.51019692e+00 1.12227309e+00 5.55401993e+00 7.72138476e-01
-1.13243771e+00 1.51223034e-01 1.19411409e+00 4.37400937e-01
-4.45479661e-01 3.99419293e-02 -8.80266607e-01 3.23026091e-01
3.89936864e-01 3.73403639e-01 -8.87166187e-02 7.77446568e-01
1.17767265e-03 -9.29268897e-02 -5.76227367e-01 5.46182215e-01
-1.88605055e-01 -1.03913403e+00 3.18724483e-01 -1.71302453e-01
5.47848046e-01 4.08565968e-01 -2.93088317e-01 1.98490486e-01
6.01139307e-01 -9.54319417e-01 9.32491660e-01 1.40444413e-01
4.70991135e-01 -6.62198126e-01 9.88550782e-01 5.53009093e-01
-1.37804997e+00 -1.59166306e-01 -6.82419479e-01 -1.40450178e-02
-3.95681597e-02 6.70085371e-01 -6.35190368e-01 7.18828857e-01
9.06208038e-01 6.94918931e-01 -5.75915635e-01 1.03142357e+00
-1.91861704e-01 6.52588069e-01 -4.60648268e-01 2.20410913e-01
5.49447834e-01 -3.23521197e-01 2.96561569e-01 1.15683198e+00
1.97332904e-01 4.07467932e-01 5.78296781e-01 1.07729268e+00
-3.11604105e-02 6.19526207e-02 -1.26654670e-01 2.27017216e-02
2.21160531e-01 1.39307630e+00 -1.47928774e+00 -6.01166606e-01
-2.92324871e-01 7.03556657e-01 1.42089248e-01 2.96997339e-01
-9.37522411e-01 -4.53530312e-01 3.84623945e-01 -9.32328552e-02
3.79328847e-01 2.46654168e-01 -5.97340226e-01 -1.26090276e+00
-9.49255452e-02 -6.67066395e-01 6.65191948e-01 -8.52694929e-01
-9.29867208e-01 6.86674833e-01 -1.43470839e-01 -9.81632769e-01
4.66697454e-01 -2.02600449e-01 -5.47771513e-01 6.84667528e-01
-2.09741664e+00 -1.21752369e+00 -8.44406664e-01 2.40535185e-01
8.34229350e-01 3.71561885e-01 4.09325361e-01 3.71191174e-01
-8.41395259e-01 1.51015311e-01 -3.75235647e-01 4.22434628e-01
4.40875664e-02 -1.39500725e+00 5.90489328e-01 1.04788613e+00
-3.75521891e-02 2.37929951e-02 4.30004567e-01 -6.36374474e-01
-7.07187116e-01 -1.54389536e+00 5.14336705e-01 -1.13757648e-01
2.89186120e-01 -1.78105578e-01 -1.17619348e+00 6.20063722e-01
-2.03512236e-01 -3.92623208e-02 5.42453170e-01 -2.70796061e-01
-1.75075769e-01 -8.28349665e-02 -1.15672779e+00 2.07242399e-01
1.04643619e+00 -1.01296358e-01 -5.21392822e-01 3.31620753e-01
1.04052639e+00 -5.17115295e-01 -3.65134209e-01 6.69297099e-01
1.38280958e-01 -8.53675723e-01 1.08483005e+00 -5.62413573e-01
6.74175024e-01 -5.34600079e-01 -1.57040998e-01 -9.83363390e-01
-4.28342998e-01 -3.71359736e-02 5.90979636e-01 1.14674902e+00
5.54839134e-01 -5.81298470e-01 7.73412108e-01 2.03892425e-01
-4.32833672e-01 -3.31359863e-01 -5.21705508e-01 -4.02542651e-01
3.16893458e-02 -3.94475996e-01 1.00249505e+00 1.09918153e+00
-6.72359228e-01 2.83865482e-01 -2.09527194e-01 5.78696668e-01
6.52500331e-01 6.38445377e-01 5.62434256e-01 -1.47539949e+00
-1.47439241e-01 -6.47831202e-01 -6.75905868e-02 -1.24621987e+00
1.48310084e-02 -9.70152736e-01 2.70764053e-01 -1.77860928e+00
5.15127420e-01 -8.28893185e-01 -2.45303601e-01 9.24525023e-01
-5.16825199e-01 5.00380993e-01 1.17426544e-01 2.74065673e-01
-7.51738310e-01 3.26769024e-01 1.11679208e+00 -4.04169083e-01
-4.16767806e-01 6.35428876e-02 -1.21828008e+00 9.04330909e-01
8.65170181e-01 -6.40826166e-01 -2.86738962e-01 -7.56306171e-01
4.99709785e-01 -6.99577034e-02 5.26982903e-01 -9.24274981e-01
9.04641002e-02 -2.41006523e-01 2.91769296e-01 -9.49886620e-01
-2.50642955e-01 -7.61777699e-01 -4.31416780e-02 5.20541072e-01
-1.90372944e-01 -4.16341007e-01 1.22865103e-01 4.02179599e-01
-2.08612993e-01 -3.14211249e-01 8.99998248e-01 -4.56689328e-01
-1.02774906e+00 2.26623878e-01 -1.69673994e-01 9.53135267e-02
8.05818617e-01 -2.80901521e-01 -4.57011133e-01 -6.06866740e-02
-8.44989240e-01 6.48581982e-01 3.34487438e-01 2.37479255e-01
3.59146029e-01 -8.55485082e-01 -9.80630934e-01 4.42576617e-01
-4.35322449e-02 8.30782235e-01 4.86198604e-01 6.74449623e-01
-8.84643614e-01 6.26307130e-02 -1.58504844e-01 -8.59563768e-01
-1.07466269e+00 2.92156190e-01 5.91349840e-01 -1.72978327e-01
-6.71907425e-01 1.03568149e+00 5.06183863e-01 -7.66247928e-01
1.00511141e-01 -8.18260014e-01 -5.87354600e-01 4.25131470e-01
4.89674389e-01 4.43395525e-02 1.17564663e-01 -5.04932880e-01
-4.87452149e-01 7.52069950e-01 2.31621847e-01 2.71016419e-01
1.59475815e+00 -4.07820821e-01 -2.47015029e-01 1.50566459e-01
7.34603345e-01 -3.35745335e-01 -1.18694675e+00 -3.17815214e-01
-7.36705288e-02 -3.17867249e-01 2.11431041e-01 -6.82932138e-01
-1.64946473e+00 9.48826373e-01 5.14834762e-01 6.97976232e-01
1.33521879e+00 1.63571358e-01 7.28842378e-01 3.02347779e-01
4.80888858e-02 -9.07708704e-01 -2.65118241e-01 4.39893812e-01
3.79946917e-01 -1.45113170e+00 -1.53104424e-01 -6.25886798e-01
-5.44035077e-01 9.59766030e-01 6.18325114e-01 2.06738114e-02
6.54043198e-01 1.17532518e-02 2.61752367e-01 -5.35635531e-01
-2.72114158e-01 -6.69926107e-01 1.75826013e-01 3.31811994e-01
1.74337879e-01 1.96949527e-01 -1.34958297e-01 7.39677548e-01
2.60059759e-02 -2.03761637e-01 5.38923621e-01 8.66692781e-01
-1.08887970e+00 -6.84557140e-01 -3.78919035e-01 3.92792583e-01
-6.16481960e-01 -2.37209633e-01 -3.83515686e-01 4.90135729e-01
3.24895471e-01 8.83891463e-01 4.22968328e-01 -3.17356855e-01
1.72597542e-01 -1.74154252e-01 -1.73507124e-01 -7.98700690e-01
-6.46194518e-01 2.98154145e-01 -5.59891090e-02 -4.49816018e-01
-8.45567346e-01 -2.19420299e-01 -1.39801085e+00 1.01576239e-01
-4.97637153e-01 1.73096910e-01 5.68414867e-01 8.79679859e-01
2.53131747e-01 8.25228214e-01 6.36291742e-01 -8.34833026e-01
-2.67736733e-01 -9.80627716e-01 -5.95565856e-01 4.08331066e-01
2.97037840e-01 -3.31789374e-01 -4.95976031e-01 -1.78330075e-02] | [9.48173999786377, -1.0229955911636353] |
f8afa651-1208-4f84-a9cc-ca145c88573a | similarity-aware-multimodal-prompt-learning | 2304.04187 | null | https://arxiv.org/abs/2304.04187v3 | https://arxiv.org/pdf/2304.04187v3.pdf | Similarity-Aware Multimodal Prompt Learning for Fake News Detection | The standard paradigm for fake news detection mainly utilizes text information to model the truthfulness of news. However, the discourse of online fake news is typically subtle and it requires expert knowledge to use textual information to debunk fake news. Recently, studies focusing on multimodal fake news detection have outperformed text-only methods. Recent approaches utilizing the pre-trained model to extract unimodal features, or fine-tuning the pre-trained model directly, have become a new paradigm for detecting fake news. Again, this paradigm either requires a large number of training instances, or updates the entire set of pre-trained model parameters, making real-world fake news detection impractical. Furthermore, traditional multimodal methods fuse the cross-modal features directly without considering that the uncorrelated semantic representation might inject noise into the multimodal features. This paper proposes a Similarity-Aware Multimodal Prompt Learning (SAMPLE) framework. First, we incorporate prompt learning into multimodal fake news detection. Prompt learning, which only tunes prompts with a frozen language model, can reduce memory usage significantly and achieve comparable performances, compared with fine-tuning. We analyse three prompt templates with a soft verbalizer to detect fake news. In addition, we introduce the similarity-aware fusing method to adaptively fuse the intensity of multimodal representation and mitigate the noise injection via uncorrelated cross-modal features. For evaluation, SAMPLE surpasses the F1 and the accuracies of previous works on two benchmark multimodal datasets, demonstrating the effectiveness of the proposed method in detecting fake news. In addition, SAMPLE also is superior to other approaches regardless of few-shot and data-rich settings. | ['Diana Maynard', 'Xingyi Song', 'Xiaoman Xu', 'Yimin Wang', 'Xiaomin Yu', 'Ye Jiang'] | 2023-04-09 | null | null | null | null | ['fake-news-detection'] | ['natural-language-processing'] | [ 1.50385275e-01 -2.02956244e-01 -4.62058216e-01 -2.00035885e-01
-1.15388799e+00 -6.81087792e-01 9.12532270e-01 8.31345990e-02
-3.40268314e-01 5.20918727e-01 3.69337678e-01 -1.14029199e-01
3.58039707e-01 -5.00615478e-01 -7.21361220e-01 -6.59644485e-01
5.45828402e-01 2.56566912e-01 1.42312482e-01 -5.99058390e-01
5.36965907e-01 -4.29489873e-02 -1.41155279e+00 8.70488524e-01
1.04235041e+00 8.15048814e-01 -1.75976619e-01 4.16154712e-01
-1.96115091e-01 8.18189621e-01 -9.88214672e-01 -8.56475174e-01
-9.08557251e-02 -5.23317933e-01 -5.09876668e-01 2.31064539e-02
4.82429117e-01 -4.80904579e-01 -6.41772211e-01 1.11828637e+00
4.33939517e-01 -4.11249213e-02 5.85563600e-01 -1.32726896e+00
-7.46729016e-01 6.57917678e-01 -6.13175690e-01 2.18630984e-01
5.59275448e-01 2.05769226e-01 7.62905598e-01 -1.04334629e+00
5.29386461e-01 1.36510897e+00 5.64058483e-01 4.26326722e-01
-1.01230335e+00 -7.73927271e-01 -9.67175066e-02 4.89514023e-01
-1.07212520e+00 -5.42059481e-01 1.14782012e+00 -3.41099530e-01
6.01522505e-01 3.38303208e-01 5.16329288e-01 1.89952850e+00
1.64525315e-01 1.00123596e+00 1.39167297e+00 -4.41659540e-01
-1.30050099e-02 5.02886355e-01 2.47421548e-01 7.35590398e-01
2.04745397e-01 1.84810549e-01 -9.16790903e-01 -7.27164149e-01
1.44218519e-01 2.07794886e-02 -4.00804758e-01 4.14567776e-02
-1.40227044e+00 1.18622386e+00 1.68040514e-01 4.08581674e-01
-1.76843792e-01 7.04420498e-03 7.14682460e-01 5.23217857e-01
5.27653694e-01 7.15946972e-01 -5.23031689e-02 -1.63146481e-01
-9.35645401e-01 6.63879365e-02 7.06760466e-01 6.31117821e-01
5.90704501e-01 7.67136039e-03 -4.45663601e-01 8.22445750e-01
-1.75529271e-02 7.29420185e-01 7.24243879e-01 -5.84572494e-01
7.35991478e-01 6.48661315e-01 3.68793070e-01 -1.71704900e+00
-1.85041159e-01 -2.72623479e-01 -5.62259972e-01 -4.20484185e-01
3.90535176e-01 1.18525197e-04 -7.71743476e-01 1.49079049e+00
2.76269883e-01 3.03011894e-01 1.15625791e-01 1.20894909e+00
9.32148218e-01 8.31323504e-01 -8.50398540e-02 -3.17913562e-01
1.42109859e+00 -1.12360513e+00 -1.09856546e+00 -4.24125969e-01
7.11308062e-01 -1.07757366e+00 1.34655070e+00 4.42152023e-01
-4.23463821e-01 -8.54026079e-02 -1.08119869e+00 -1.25553086e-01
-5.39961040e-01 3.94883543e-01 5.68443716e-01 9.00645256e-01
-3.34041715e-01 3.09612870e-01 -4.27972764e-01 -1.15171991e-01
3.70462596e-01 3.08030713e-02 -3.67680103e-01 -2.81166852e-01
-1.66203570e+00 1.03131223e+00 4.49402779e-01 2.95039006e-02
-8.25127721e-01 -3.29157144e-01 -7.78552949e-01 -1.35282680e-01
7.63387918e-01 -5.76771200e-01 1.02150190e+00 -1.29624140e+00
-1.76545846e+00 5.96498072e-01 -1.65208280e-01 -1.49050310e-01
6.65147245e-01 -1.30572334e-01 -5.98672509e-01 4.85293508e-01
2.71549784e-02 2.90699661e-01 1.68213201e+00 -1.52021039e+00
-1.40168563e-01 -8.45608264e-02 9.03000534e-02 1.02681078e-01
-6.49925232e-01 7.63587747e-03 -1.97898209e-01 -8.93166244e-01
1.14877857e-01 -6.70726776e-01 2.92452216e-01 -4.20763254e-01
-5.09056926e-01 9.21351314e-02 1.19686496e+00 -7.80906796e-01
1.24098694e+00 -2.18465710e+00 1.07659986e-02 1.69765815e-01
2.62575716e-01 3.34832698e-01 -3.88267428e-01 5.17050982e-01
1.61260664e-01 -6.23228662e-02 -1.49134025e-01 -4.47725207e-01
1.03914924e-01 9.71747562e-03 -5.95469058e-01 8.00324440e-01
7.06605315e-02 9.06756938e-01 -1.04209757e+00 -4.71792191e-01
2.03056678e-01 3.93174559e-01 -3.64332080e-01 1.10738821e-01
-6.65501729e-02 4.44203198e-01 -1.16862372e-01 9.09765899e-01
7.34774828e-01 -2.04502001e-01 1.40714273e-01 -5.43312132e-01
2.90097684e-01 1.70736641e-01 -6.72505081e-01 1.29707360e+00
-3.64314258e-01 6.25771940e-01 -6.44505918e-02 -1.16442585e+00
7.05662549e-01 3.22816461e-01 2.11323187e-01 -7.73426533e-01
6.00302339e-01 4.21273172e-01 -4.62703615e-01 -8.24953020e-01
6.50807858e-01 -3.17167938e-01 -4.35395300e-01 5.18862426e-01
1.27325431e-01 -1.55158043e-01 -2.92585790e-01 3.84807289e-01
8.77861798e-01 -2.85691410e-01 2.06727788e-01 3.75743657e-01
4.45844412e-01 1.85113505e-01 1.84828088e-01 9.44824398e-01
-2.88952589e-01 2.98103273e-01 6.94531739e-01 -2.18360975e-01
-6.65418804e-01 -7.03757465e-01 1.67353049e-01 1.18471193e+00
7.82378674e-01 -4.33034331e-01 -5.97223461e-01 -1.09668016e+00
2.89578456e-02 8.89474213e-01 -6.10752940e-01 -6.00496411e-01
-3.25154603e-01 -1.00472319e+00 8.45669746e-01 -1.89314932e-01
4.22969550e-01 -5.36060274e-01 -2.18372867e-01 5.34960628e-02
-9.31108475e-01 -1.32497692e+00 -5.28848231e-01 -2.98255175e-01
-5.69980502e-01 -1.01352537e+00 -5.13148844e-01 -4.06806171e-01
5.43207407e-01 6.48913980e-01 6.12641811e-01 2.94828594e-01
2.03990098e-02 3.88351679e-01 -6.11989975e-01 1.68411471e-02
-5.76373518e-01 -1.18122064e-01 -7.32686650e-03 4.14977759e-01
3.27223659e-01 -1.92158688e-02 -2.44956627e-01 3.48559797e-01
-1.03776371e+00 -2.17324402e-02 6.54660225e-01 1.44620359e+00
-5.32071292e-02 -6.48339242e-02 6.80225730e-01 -7.78014004e-01
8.95145297e-01 -8.12508702e-01 -2.18620434e-01 2.31010690e-01
-3.55108559e-01 -2.04666182e-01 6.05493963e-01 -1.01176882e+00
-1.02588665e+00 -2.63140649e-01 2.57479370e-01 -5.36668599e-01
9.10239369e-02 5.79892397e-01 2.50963029e-02 -2.69146055e-01
8.11257601e-01 3.27085167e-01 -8.81498959e-03 -1.34545147e-01
4.90933716e-01 9.90397453e-01 3.92462581e-01 -4.76473123e-01
9.23371315e-01 6.64241433e-01 -4.55720007e-01 -6.67466283e-01
-1.02262962e+00 -5.32697797e-01 -1.35512710e-01 -2.52343625e-01
3.63171160e-01 -8.45339775e-01 -6.95173383e-01 5.63127816e-01
-1.49222684e+00 9.75388288e-02 3.82826298e-01 4.75070059e-01
-3.02148610e-01 8.27543139e-01 -7.79543221e-01 -7.89338827e-01
-8.76681730e-02 -1.04640210e+00 1.15726101e+00 -2.11314425e-01
-5.26116788e-02 -8.58968914e-01 -1.05768822e-01 9.58091080e-01
3.75550985e-01 2.11292252e-01 7.81027734e-01 -8.51957083e-01
-2.46329084e-01 -5.84015012e-01 -4.92093176e-01 2.90240288e-01
1.30868763e-01 -3.27776074e-01 -1.13894546e+00 -2.68870682e-01
1.67591959e-01 -6.34173334e-01 1.04223299e+00 -1.98250189e-01
8.16956580e-01 -7.47380853e-01 -2.36069009e-01 -1.33670457e-02
1.03581369e+00 -3.08078915e-01 4.00360852e-01 3.16331863e-01
8.19420159e-01 7.44277000e-01 7.70849228e-01 4.89378303e-01
4.80972022e-01 8.11067045e-01 4.18135494e-01 7.54077733e-02
1.90321997e-01 -2.51767844e-01 8.40052485e-01 7.47704923e-01
2.22185224e-01 -3.68423939e-01 -7.96571314e-01 3.07127476e-01
-1.88286889e+00 -1.24271083e+00 -1.57067388e-01 1.94667721e+00
8.21191669e-01 -4.89014722e-02 4.40899618e-02 1.35608405e-01
7.77556658e-01 2.06617460e-01 -1.39739856e-01 -2.84132391e-01
-3.99260551e-01 -3.07671547e-01 4.60125774e-01 7.04924643e-01
-1.26450729e+00 1.21519220e+00 5.90428972e+00 1.27004838e+00
-1.48954451e+00 7.06481814e-01 2.28417397e-01 -1.69186398e-01
-5.95312893e-01 -8.90331566e-02 -3.04235220e-01 8.12924445e-01
8.09227645e-01 1.26999348e-01 6.04482532e-01 5.64179778e-01
4.18198526e-01 -2.80224025e-01 -7.07400680e-01 1.21605051e+00
7.39380896e-01 -1.24692380e+00 3.15765709e-01 -2.16057122e-01
7.86326885e-01 -3.53280932e-01 2.42437571e-01 5.41126907e-01
-3.05957824e-01 -9.65388060e-01 9.10247326e-01 5.43902278e-01
4.19176549e-01 -7.04056084e-01 1.28752756e+00 6.15036547e-01
-4.17490244e-01 -2.62441725e-01 -1.65188581e-01 8.10049251e-02
1.95788577e-01 6.98432684e-01 -8.88042629e-01 5.89335263e-01
2.31831700e-01 4.89045709e-01 -7.16743469e-01 6.64131224e-01
-1.83464006e-01 6.77769125e-01 -1.74608603e-01 -1.19094521e-01
3.38203371e-01 1.12561502e-01 7.71183312e-01 1.39899909e+00
2.44050995e-01 -2.37547178e-02 2.57289052e-01 5.85833490e-01
-9.97329503e-02 2.00400636e-01 -5.10659099e-01 -4.92094547e-01
4.12473559e-01 1.06421554e+00 -4.13306326e-01 -6.22923732e-01
-3.35980117e-01 1.28501022e+00 1.38420969e-01 3.08400959e-01
-1.22749949e+00 -3.31479758e-01 1.07497498e-01 -3.77895571e-02
3.33945677e-02 -5.75586297e-02 -4.00212675e-01 -1.56720042e+00
-5.53264543e-02 -1.35045838e+00 3.50249887e-01 -5.23309171e-01
-1.58496499e+00 4.37111259e-01 -5.71242941e-04 -1.33990574e+00
-1.50053697e-02 -4.48562473e-01 -3.18017304e-01 2.96714187e-01
-1.57868373e+00 -1.39354634e+00 -4.24712896e-01 6.04285777e-01
4.13485080e-01 -4.55954187e-02 5.15297472e-01 3.58420074e-01
-5.47541738e-01 7.68567801e-01 -2.55970769e-02 -1.60136551e-03
1.23587620e+00 -6.77426279e-01 -4.06369209e-01 6.73912704e-01
-6.66551888e-02 6.66462541e-01 9.16611791e-01 -7.42220402e-01
-1.58134079e+00 -7.59880245e-01 8.46469283e-01 -6.12999439e-01
8.59886467e-01 -4.63040233e-01 -1.04772186e+00 3.67938787e-01
1.32415771e-01 -7.15981349e-02 6.06854200e-01 -1.76191270e-01
-8.57653320e-01 9.13699865e-02 -1.37886798e+00 5.01800835e-01
6.03148103e-01 -8.24084878e-01 -8.29255342e-01 4.75227207e-01
5.62198758e-01 -5.04769206e-01 -3.56721222e-01 1.93007052e-01
5.92215955e-01 -9.42147851e-01 7.10946620e-01 -6.34333909e-01
3.90836030e-01 -2.75310010e-01 -1.60552979e-01 -1.32100463e+00
7.80740306e-02 -5.96451342e-01 -3.11145633e-01 9.61617053e-01
2.94400841e-01 -7.58424699e-01 3.55745167e-01 2.14533452e-02
2.42158230e-02 -4.64642644e-01 -1.09793496e+00 -7.41392553e-01
-2.67401874e-01 -4.02746975e-01 3.69291663e-01 1.51663733e+00
3.51959974e-01 2.58767277e-01 -1.08017945e+00 3.38790715e-01
5.63739061e-01 8.87761340e-02 8.10512543e-01 -7.58645654e-01
-3.16302001e-01 -3.50949168e-01 -2.64002651e-01 -9.84314501e-01
4.47976708e-01 -6.91920578e-01 4.76066433e-02 -9.34568524e-01
3.46457332e-01 -1.58835892e-02 2.44046412e-02 5.34845114e-01
-3.78707945e-01 5.36054790e-01 5.58824688e-02 4.12793756e-01
-5.48025131e-01 7.91116476e-01 1.32656944e+00 -4.92916584e-01
-1.02323934e-01 -3.35818827e-01 -5.70773304e-01 6.85075939e-01
7.30573356e-01 -7.65444517e-01 -2.26670265e-01 -2.03798398e-01
2.56789982e-01 7.73867816e-02 7.29661047e-01 -4.19801712e-01
2.21480325e-01 -3.31358463e-01 8.94787461e-02 -2.78669685e-01
7.37092376e-01 -6.78138614e-01 -2.85053641e-01 4.04171735e-01
-2.14308441e-01 -3.33407521e-01 1.10133618e-01 8.81696343e-01
-4.28398430e-01 -1.97083846e-01 8.56274426e-01 -2.29563434e-02
-4.59810704e-01 -2.85668731e-01 -6.55739188e-01 -8.03930461e-02
8.96284044e-01 -1.15467243e-01 -8.44918668e-01 -7.61821091e-01
-4.05898392e-01 -1.24211438e-01 3.87805194e-01 5.27220368e-01
7.98048973e-01 -1.28341472e+00 -6.04031384e-01 -7.09887370e-02
4.07028466e-01 -8.36617410e-01 4.33276057e-01 1.38566768e+00
-1.74138084e-01 1.49826631e-01 1.48236006e-01 -5.32263756e-01
-1.10251784e+00 7.07108557e-01 1.32239595e-01 -3.67463492e-02
-3.17803532e-01 5.09055853e-01 -2.48244226e-01 -4.75348800e-01
-3.70634906e-02 1.09587021e-01 -1.32472999e-02 3.73825639e-01
5.59352636e-01 4.11123723e-01 1.15782432e-02 -8.81389022e-01
-3.89671624e-01 2.77203858e-01 -4.39055897e-02 -2.17703849e-01
8.55693996e-01 -2.54194409e-01 -2.62266666e-01 5.29219568e-01
1.16234231e+00 4.25833195e-01 -5.37490308e-01 -2.60018915e-01
6.57872483e-02 -8.47066104e-01 1.71183586e-01 -1.00382757e+00
-6.17097080e-01 6.52562678e-01 2.29961544e-01 1.41831979e-01
8.48691046e-01 7.46717378e-02 9.72284675e-01 4.95037735e-01
3.84046018e-01 -1.10192978e+00 7.84624577e-01 4.08982337e-01
1.01001942e+00 -1.67311299e+00 -1.14845462e-01 -6.23724818e-01
-9.26147819e-01 1.28872299e+00 4.77198690e-01 -3.32557298e-02
9.75976959e-02 -8.85333195e-02 4.13965672e-01 -2.91803986e-01
-4.47176218e-01 2.35345393e-01 4.73141015e-01 2.56717056e-01
2.62086168e-02 5.02856821e-03 -5.06257892e-01 7.79194653e-01
-1.11560903e-01 -2.68158376e-01 7.03403175e-01 6.78265810e-01
-4.45730001e-01 -8.10791135e-01 -8.23986411e-01 2.23504275e-01
-2.72776395e-01 -1.09633312e-01 -7.36776352e-01 5.75929940e-01
2.50333726e-01 1.39674354e+00 -4.56928104e-01 -6.56458497e-01
1.01694077e-01 2.55064480e-02 4.23242360e-01 -3.97292107e-01
-5.54543138e-01 -3.28253843e-02 3.12401146e-01 -7.15072513e-01
-4.51068252e-01 -3.10143352e-01 -8.09079230e-01 -6.02706850e-01
-8.07374418e-01 1.00336790e-01 6.48732126e-01 1.33652389e+00
3.47384512e-01 -9.26133990e-03 8.17324758e-01 -8.00818145e-01
-9.25698876e-01 -1.08272111e+00 -9.16535482e-02 4.81701970e-01
6.10774934e-01 -1.02302408e+00 -8.68480742e-01 -2.53622860e-01] | [8.1884126663208, 10.315967559814453] |
7cde01f3-30fc-4bf4-96ef-4e0f0dd9a031 | sim2sg-sim-to-real-scene-graph-generation-for-1 | 2011.14488 | null | https://arxiv.org/abs/2011.14488v2 | https://arxiv.org/pdf/2011.14488v2.pdf | Self-Supervised Real-to-Sim Scene Generation | Synthetic data is emerging as a promising solution to the scalability issue of supervised deep learning, especially when real data are difficult to acquire or hard to annotate. Synthetic data generation, however, can itself be prohibitively expensive when domain experts have to manually and painstakingly oversee the process. Moreover, neural networks trained on synthetic data often do not perform well on real data because of the domain gap. To solve these challenges, we propose Sim2SG, a self-supervised automatic scene generation technique for matching the distribution of real data. Importantly, Sim2SG does not require supervision from the real-world dataset, thus making it applicable in situations for which such annotations are difficult to obtain. Sim2SG is designed to bridge both the content and appearance gaps, by matching the content of real data, and by matching the features in the source and target domains. We select scene graph (SG) generation as the downstream task, due to the limited availability of labeled datasets. Experiments demonstrate significant improvements over leading baselines in reducing the domain gap both qualitatively and quantitatively, on several synthetic datasets as well as the real-world KITTI dataset. | ['Stan Birchfield', 'Marc T. Law', 'Gavriel State', 'Eric Cameracci', 'Jean-Francois Lafleche', 'Shoubhik Debnath', 'Aayush Prakash'] | 2020-11-30 | sim2sg-sim-to-real-scene-graph-generation-for | http://openaccess.thecvf.com//content/ICCV2021/html/Prakash_Self-Supervised_Real-to-Sim_Scene_Generation_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Prakash_Self-Supervised_Real-to-Sim_Scene_Generation_ICCV_2021_paper.pdf | iccv-2021-1 | ['scene-generation'] | ['computer-vision'] | [ 2.92071700e-01 1.33230776e-01 4.02583927e-02 -2.89017528e-01
-1.00470579e+00 -6.87515318e-01 7.33438492e-01 9.35347006e-02
-3.16054493e-01 9.08989012e-01 1.36712402e-01 -8.70307311e-02
3.02855968e-01 -8.64429653e-01 -9.07430351e-01 -4.72354531e-01
4.45207208e-01 7.49300778e-01 1.23580068e-01 -1.82654291e-01
-1.71036392e-01 2.90056288e-01 -1.58157837e+00 2.03294918e-01
1.12718296e+00 8.97645652e-01 3.46524179e-01 1.43346697e-01
-1.45427093e-01 7.76769102e-01 -6.80965662e-01 -3.36902529e-01
5.20862043e-01 -4.85668749e-01 -7.42711186e-01 4.79210496e-01
7.40371346e-01 -3.38978708e-01 -3.68968636e-01 1.06710625e+00
4.40996289e-01 5.27106225e-02 6.02391005e-01 -1.52112031e+00
-5.66550374e-01 2.12994516e-01 -6.46022022e-01 -4.28763181e-02
2.09556118e-01 4.54232424e-01 9.00568187e-01 -9.01331365e-01
9.74594772e-01 1.13388395e+00 4.27691460e-01 5.03544092e-01
-1.57300544e+00 -6.34717464e-01 9.10161734e-02 5.23457304e-03
-1.23435080e+00 -5.38166881e-01 9.67865348e-01 -5.74523151e-01
4.05943990e-01 -2.26065964e-01 4.38048333e-01 1.59251618e+00
-3.89332235e-01 9.87425566e-01 1.23757195e+00 -3.39345723e-01
3.23623866e-01 2.35261813e-01 -3.50731403e-01 4.12604034e-01
3.57134312e-01 1.31768599e-01 -5.60118556e-01 1.58280909e-01
6.87490582e-01 -2.74751991e-01 -3.71707946e-01 -8.26240420e-01
-1.37149155e+00 7.74691045e-01 5.48031688e-01 -1.04488641e-01
-3.44022959e-01 -3.94453779e-02 4.40589905e-01 3.35950583e-01
7.45272577e-01 7.21097767e-01 -3.95789742e-01 -1.14139367e-03
-8.82835209e-01 5.64051449e-01 4.29330707e-01 1.21880889e+00
7.18482614e-01 1.83272049e-01 -6.63729906e-02 8.16399336e-01
5.20885549e-02 5.91738641e-01 3.78512979e-01 -8.66747379e-01
8.13757420e-01 7.77102947e-01 3.94496083e-01 -9.27852392e-01
-1.62240237e-01 -3.10459316e-01 -9.97637153e-01 2.92456716e-01
9.41878557e-01 -1.59439579e-01 -9.95744050e-01 1.95230770e+00
5.21707356e-01 -2.70401835e-02 1.72168136e-01 8.81264865e-01
9.25387979e-01 5.67372262e-01 -2.45135352e-02 1.92390084e-01
9.72626269e-01 -1.05924356e+00 -3.38357002e-01 -5.74123263e-01
6.68804288e-01 -6.38940454e-01 1.68136215e+00 1.81442454e-01
-7.97477543e-01 -5.04881680e-01 -9.36782598e-01 -2.64037438e-02
-3.99569869e-01 1.15839027e-01 5.53980649e-01 1.56465977e-01
-8.18297505e-01 4.42553967e-01 -7.02178359e-01 -3.70532155e-01
7.87993729e-01 -1.33669138e-01 -6.21553838e-01 -4.67537463e-01
-1.07665956e+00 6.21367335e-01 5.05602717e-01 1.62927415e-02
-9.55556571e-01 -9.26129103e-01 -1.12054634e+00 -9.55913514e-02
4.23220694e-01 -4.95494068e-01 1.23921657e+00 -1.17598438e+00
-9.09677923e-01 1.17678690e+00 2.18579561e-01 -4.09814060e-01
9.68273580e-01 -2.05832981e-02 -3.09685796e-01 -2.75140386e-02
4.70563352e-01 9.28834617e-01 8.32733154e-01 -1.42981517e+00
-4.18662906e-01 -2.79449344e-01 1.60129711e-01 3.46502095e-01
-1.84494257e-01 -4.20960695e-01 -3.72171134e-01 -7.16715991e-01
-1.98403876e-02 -8.92409503e-01 -4.35650378e-01 1.61433443e-01
-5.67639112e-01 -3.94460633e-02 9.48795199e-01 -5.34995258e-01
6.02037430e-01 -2.36410284e+00 -2.14863028e-02 -7.58055672e-02
2.99099207e-01 2.82974035e-01 -3.66667837e-01 3.82316589e-01
-1.17128856e-01 -2.05296129e-01 -3.21070254e-01 -2.77464658e-01
-6.32321909e-02 1.83793470e-01 -4.44029421e-01 3.66912514e-01
3.59170675e-01 9.43769634e-01 -1.12121809e+00 -5.26170254e-01
7.87608624e-02 1.15114242e-01 -3.92463893e-01 4.21679825e-01
-7.43140996e-01 6.26038909e-01 -2.47154236e-01 4.08179879e-01
6.86565757e-01 -5.95569670e-01 3.79982293e-02 -1.47553608e-01
3.53201389e-01 3.54079127e-01 -1.24911690e+00 1.93962049e+00
-5.45562506e-01 8.92670870e-01 -7.73265809e-02 -1.00531971e+00
8.51266325e-01 4.89734858e-02 2.97896475e-01 -9.89130378e-01
-2.10695475e-01 1.60153732e-01 -1.66851640e-01 -4.52332199e-01
4.15765673e-01 -2.91374534e-01 -2.61961520e-01 6.50041103e-01
7.60176480e-02 -4.50548053e-01 5.02348304e-01 5.35281420e-01
1.15704191e+00 3.57444435e-01 9.15873647e-02 -1.63575605e-01
-5.51280305e-02 5.98334193e-01 6.55855834e-01 5.00850439e-01
-8.45364481e-02 9.69557583e-01 5.87027490e-01 -4.28698957e-01
-1.35171294e+00 -1.11592913e+00 1.87319815e-01 7.77280867e-01
3.37849438e-01 -3.13395783e-02 -8.07018638e-01 -1.05267942e+00
1.12935314e-02 6.56035125e-01 -5.73768079e-01 -1.39856055e-01
-2.63354123e-01 -5.36450803e-01 3.44639719e-01 5.51452458e-01
7.50428498e-01 -1.13280809e+00 -5.66974521e-01 1.89253137e-01
-4.99491781e-01 -1.55835390e+00 -3.08411300e-01 -1.25834852e-01
-4.60834116e-01 -1.21980774e+00 -5.68101645e-01 -6.68428361e-01
1.00204480e+00 6.00621402e-01 1.48802805e+00 1.03671163e-01
-3.10456663e-01 4.03830409e-02 -3.81399065e-01 -3.64302993e-01
-5.60004354e-01 1.13477208e-01 -2.75245309e-01 3.00893169e-02
1.29928842e-01 -5.19891858e-01 -6.40697956e-01 3.10529023e-01
-1.12118924e+00 5.70474565e-01 3.55531573e-01 1.00065398e+00
5.23572862e-01 1.94992840e-01 6.30878091e-01 -1.21055436e+00
3.63019139e-01 -4.72510695e-01 -7.45261908e-01 7.79949650e-02
-3.87353390e-01 -8.06488842e-03 7.66695797e-01 -3.81488889e-01
-1.20793009e+00 1.03969455e-01 2.00221717e-01 -4.85522091e-01
-4.02415425e-01 4.14247304e-01 -3.86916578e-01 3.11684519e-01
1.07195306e+00 1.12251632e-01 -3.28755789e-02 -3.97740871e-01
3.90052706e-01 5.81542909e-01 6.22711182e-01 -5.48956811e-01
1.08800030e+00 6.27716720e-01 -2.05927908e-01 -5.59045553e-01
-1.38624525e+00 -3.40961874e-01 -6.73449457e-01 -6.36111274e-02
5.90570092e-01 -1.16386342e+00 1.49469793e-01 6.44100428e-01
-1.06747580e+00 -7.82683790e-01 -5.97090960e-01 1.58333793e-01
-5.82217395e-01 2.11842552e-01 -3.08352917e-01 -3.53582293e-01
-5.98985478e-02 -1.00579357e+00 1.32002771e+00 6.92284107e-02
-2.20178589e-01 -8.67886126e-01 -1.64932787e-01 3.95287484e-01
2.45802790e-01 8.83564830e-01 9.69562948e-01 -5.13644159e-01
-6.38026118e-01 -3.16217244e-01 -6.34547234e-01 3.23464036e-01
2.49342382e-01 -1.13658875e-01 -9.78013098e-01 -3.64852428e-01
-3.37906688e-01 -9.72788334e-01 5.35958290e-01 -7.88366869e-02
1.19199002e+00 -1.58758417e-01 -1.39344260e-01 3.94599438e-01
1.31827629e+00 -1.80746824e-01 4.83442217e-01 2.69288540e-01
9.31967437e-01 9.50628757e-01 8.63588393e-01 3.04665327e-01
5.89907169e-01 7.52659976e-01 4.51963902e-01 -5.68213284e-01
-5.59013009e-01 -7.92428434e-01 3.88219841e-02 3.53774369e-01
5.70068359e-01 -4.36086386e-01 -1.15698874e+00 8.36167514e-01
-1.77309835e+00 -7.48510897e-01 -2.32244313e-01 2.23465466e+00
9.15121734e-01 2.95906901e-01 1.59662709e-01 1.08960964e-01
5.72176397e-01 1.56543568e-01 -7.82571673e-01 2.62784362e-01
-3.88545215e-01 -9.72516611e-02 3.46570939e-01 1.95557863e-01
-1.22996736e+00 1.09786141e+00 5.38122654e+00 8.47239316e-01
-1.22074294e+00 6.28162175e-02 7.76421905e-01 -7.71174058e-02
-4.69385296e-01 1.53903803e-03 -4.62522119e-01 6.31411850e-01
5.75369835e-01 -9.86993536e-02 2.98200220e-01 9.66075242e-01
2.33982846e-01 -1.87873662e-01 -1.38391590e+00 1.10119224e+00
-6.02649115e-02 -1.36847353e+00 1.23541728e-02 5.15633114e-02
9.54284549e-01 1.64429501e-01 1.08925570e-02 2.12986246e-01
7.12542474e-01 -1.11065257e+00 8.80710065e-01 -3.63530964e-02
1.11458266e+00 -4.33580756e-01 4.46855754e-01 6.21153951e-01
-9.02957141e-01 1.95119679e-01 -2.32779637e-01 8.67908150e-02
1.68539003e-01 7.17668235e-01 -9.73482192e-01 3.22366327e-01
5.35876274e-01 7.65733421e-01 -6.76102459e-01 9.57747340e-01
-4.37628359e-01 5.04628778e-01 -3.26795936e-01 2.61041909e-01
2.62143224e-01 -9.78087708e-02 2.55144477e-01 8.63188624e-01
1.84625611e-01 -3.38240594e-01 2.52633512e-01 1.12025797e+00
-3.04820478e-01 -2.28064824e-02 -1.10217822e+00 -2.20408216e-01
4.70373422e-01 1.08293962e+00 -8.40141416e-01 -2.61531323e-01
-3.25934649e-01 9.78533864e-01 5.26281655e-01 5.95184207e-01
-6.22186899e-01 -3.37437600e-01 4.80715603e-01 3.78644615e-01
3.09742969e-02 -1.50702685e-01 -4.62117076e-01 -1.28073168e+00
2.34760493e-01 -1.18617737e+00 2.49306813e-01 -9.37559307e-01
-1.37754965e+00 6.12865269e-01 5.23848981e-02 -1.60831833e+00
-4.14984435e-01 -4.39075768e-01 -3.11011404e-01 6.25697255e-01
-1.44461155e+00 -1.27962363e+00 -6.61320567e-01 5.45396805e-01
7.23403811e-01 -5.29989526e-02 6.48942888e-01 3.07163596e-01
-5.05356312e-01 5.34886241e-01 7.65210390e-02 1.36252463e-01
7.40662813e-01 -1.23360705e+00 7.93005705e-01 9.40431595e-01
3.95402521e-01 9.09566954e-02 7.63567209e-01 -4.45319712e-01
-1.09563136e+00 -1.33679450e+00 5.34471273e-01 -4.01468098e-01
5.67629933e-01 -8.37822139e-01 -9.63489056e-01 4.99152929e-01
-7.62263089e-02 2.76988328e-01 4.99142259e-01 -3.64517160e-02
-6.56722367e-01 -1.03537977e-01 -1.24782741e+00 7.47017443e-01
1.28798032e+00 -4.24042851e-01 -1.34840310e-01 7.84062922e-01
8.10417533e-01 -6.55187368e-01 -5.72178841e-01 4.56259131e-01
8.20164979e-02 -9.27745283e-01 7.56148875e-01 -6.10475838e-01
9.74329352e-01 -3.11888188e-01 -2.01851711e-01 -1.65539527e+00
8.56385529e-02 -4.65931773e-01 1.37328401e-01 1.37926388e+00
7.00921297e-01 -4.24283475e-01 1.00999033e+00 6.72617376e-01
6.00560755e-02 -4.95913297e-01 -6.87579036e-01 -9.08847034e-01
-1.92827042e-02 -3.85265589e-01 8.38282466e-01 1.20075548e+00
-3.95948797e-01 5.49039543e-01 -3.25592607e-01 -1.53106153e-01
6.95642233e-01 3.92070442e-01 1.36026847e+00 -1.20374966e+00
-3.87496471e-01 -1.30934805e-01 -3.58293653e-01 -9.03055131e-01
2.32946604e-01 -8.27450454e-01 2.22310781e-01 -1.65859902e+00
9.35935676e-02 -8.82875502e-01 9.41110179e-02 4.44700867e-01
-1.86965808e-01 3.78927708e-01 3.94386947e-02 1.96894437e-01
-5.73820055e-01 8.15612912e-01 1.50589037e+00 -1.59095630e-01
7.30573200e-03 -1.53785706e-01 -8.24395895e-01 7.48055339e-01
6.98320627e-01 -3.18154186e-01 -8.41955423e-01 -7.18133748e-01
1.59264296e-01 -4.63645048e-02 4.22573417e-01 -9.23632920e-01
-1.34904340e-01 -3.65830600e-01 3.34388137e-01 -5.00978410e-01
3.31643522e-01 -7.87186623e-01 1.08854286e-01 2.23223455e-02
-3.71319771e-01 -5.97253814e-02 2.04356208e-01 6.13982916e-01
-2.77995348e-01 4.03260719e-03 9.54532206e-01 -1.53860882e-01
-7.66047895e-01 4.60925549e-01 1.33676320e-01 6.92906380e-01
1.16619861e+00 -8.53074268e-02 -4.95359391e-01 -5.64441860e-01
-3.11008692e-01 3.38801384e-01 8.94143403e-01 4.92511570e-01
4.96998727e-01 -1.48959327e+00 -6.71777546e-01 3.11020017e-01
6.30366564e-01 7.95424759e-01 2.91273832e-01 4.10080016e-01
-6.01515949e-01 -5.75587563e-02 -7.00802878e-02 -5.98801434e-01
-1.01738465e+00 6.30451918e-01 2.10146993e-01 -4.32759315e-01
-8.18058312e-01 9.38835859e-01 6.27115130e-01 -6.99228048e-01
2.56818235e-01 -6.69414699e-02 2.38113120e-01 -1.61643662e-02
3.44424009e-01 4.40884791e-02 1.19656563e-01 -6.37436330e-01
3.54461186e-02 2.09356863e-02 -7.12327734e-02 -2.04190165e-01
1.22465301e+00 -4.04452533e-02 2.03236848e-01 2.38375559e-01
1.08487022e+00 -2.69335747e-01 -1.70284104e+00 -5.74978769e-01
-4.08753715e-02 -7.48718858e-01 -1.84649661e-01 -8.00150096e-01
-1.18137980e+00 9.78111088e-01 2.20205337e-01 1.79410756e-01
1.04182923e+00 -1.40129970e-02 9.43653286e-01 2.28258520e-01
4.90203440e-01 -1.08403122e+00 4.14176017e-01 1.40588194e-01
1.01064813e+00 -1.60304236e+00 -1.61796167e-01 -5.62271237e-01
-8.19613516e-01 5.19909680e-01 1.09875166e+00 -9.91736352e-02
3.12204570e-01 1.84808344e-01 2.39938691e-01 -2.48195082e-01
-7.27437556e-01 -1.70862645e-01 8.44386965e-02 9.00595903e-01
1.41585305e-01 4.56384085e-02 2.14742497e-01 1.63461789e-02
-3.14028412e-01 -1.71498939e-01 6.04285121e-01 8.46491933e-01
-2.77046859e-02 -1.19271863e+00 -1.51939973e-01 5.70691049e-01
-1.37568340e-01 5.77451289e-02 -5.38509667e-01 8.57434750e-01
-4.30488922e-02 8.47625792e-01 -1.47686731e-02 -2.20372126e-01
5.47576666e-01 -3.44267517e-01 3.64958644e-01 -8.99946213e-01
4.85116057e-02 -9.13448781e-02 2.36030474e-01 -4.46138591e-01
-2.77712405e-01 -5.71850061e-01 -1.02119720e+00 -1.46608368e-01
-4.73021120e-02 -6.56765560e-03 4.92865860e-01 9.60518420e-01
5.08126915e-01 2.91338116e-01 7.25255013e-01 -8.66045356e-01
-5.25924087e-01 -9.05756593e-01 -4.79888558e-01 1.01660001e+00
2.52049357e-01 -8.22996378e-01 -1.68232292e-01 2.07213521e-01] | [9.944472312927246, 1.1957917213439941] |
fa8d7ff2-24a5-48b6-a2a1-748980fdb86c | ad-corre-adaptive-correlation-based-loss-for | null | null | https://ieeexplore.ieee.org/document/9727163 | https://ieeexplore.ieee.org/document/9727163 | Ad-Corre: Adaptive Correlation-Based Loss for Facial Expression Recognition in the Wild | Automated Facial Expression Recognition (FER) in the wild using deep neural networks is still challenging due to intra-class variations and inter-class similarities in facial images. Deep Metric Learning (DML) is among the widely used methods to deal with these issues by improving the discriminative power of the learned embedded features. This paper proposes an Adaptive Correlation (Ad-Corre) Loss to guide the network towards generating embedded feature vectors with high correlation for within-class samples and less correlation for between-class samples. Ad-Corre consists of 3 components called Feature Discriminator, Mean Discriminator, and Embedding Discriminator. We design the Feature Discriminator component to guide the network to create the embedded feature vectors to be highly correlated if they belong to a similar class, and less correlated if they belong to different classes. In addition, the Mean Discriminator component leads the network to make the mean embedded feature vectors of different classes to be less similar to each other.We use Xception network as the backbone of our model, and contrary to previous work, we propose an embedding feature space that contains k feature vectors. Then, the Embedding Discriminator component penalizes the network to generate the embedded feature vectors, which are dissimilar.We trained our model using the combination of our proposed loss functions called Ad-Corre Loss jointly with the cross-entropy loss. We achieved a very promising recognition accuracy on AffectNet, RAF-DB, and FER-2013. Our extensive experiments and ablation study indicate the power of our method to cope well with challenging FER tasks in the wild. | ['Mohammad H. Mahoor', 'Ali Pourramezan Fard'] | 2022-03-03 | null | null | null | ieee-access-2022-3 | ['facial-expression-recognition'] | ['computer-vision'] | [ 1.11542888e-01 -3.94529067e-02 1.13876320e-01 -7.77910829e-01
-3.35691929e-01 -1.96917519e-01 4.59152937e-01 -2.30621994e-01
-4.86475617e-01 5.57089925e-01 -4.52233925e-02 4.13914979e-01
-1.07027128e-01 -7.24365115e-01 -3.89919311e-01 -9.04497683e-01
-2.70795256e-01 -1.35892546e-02 -2.05578119e-01 -2.55379438e-01
-5.50031960e-02 6.65343523e-01 -1.58618355e+00 3.73049557e-01
8.19761634e-01 1.61419988e+00 -2.53465176e-01 1.96827486e-01
1.11241676e-01 6.10631943e-01 -6.24443889e-01 -5.76125681e-01
2.84245640e-01 -6.89615548e-01 -3.94837588e-01 -1.30345896e-01
4.85924304e-01 -8.39574933e-02 -3.01591367e-01 1.13062739e+00
6.60775661e-01 -3.01072863e-03 7.98905849e-01 -1.71441650e+00
-5.59949279e-01 1.44824669e-01 -5.39011061e-01 2.30841599e-02
1.10734202e-01 -3.45483162e-02 1.00104046e+00 -1.06335330e+00
3.85248870e-01 1.31639063e+00 7.30044603e-01 7.08830595e-01
-1.21632147e+00 -9.53629613e-01 1.09166384e-01 3.52400094e-01
-1.58906686e+00 -5.80122888e-01 9.39525068e-01 -2.93739378e-01
5.71709514e-01 1.88328326e-01 6.39220297e-01 1.12158251e+00
3.37722063e-01 7.97732234e-01 1.04548037e+00 -1.82732657e-01
2.71245033e-01 2.26442307e-01 -1.76441565e-01 6.38279796e-01
-2.38604620e-01 1.93649754e-01 -4.39949185e-01 -1.68991387e-01
3.84602934e-01 1.19368702e-01 -3.64015460e-01 -3.52005243e-01
-5.05832672e-01 1.06085300e+00 5.93669176e-01 4.05849278e-01
-3.85246605e-01 6.78648725e-02 3.94963026e-01 3.77318949e-01
4.54832435e-01 2.22416490e-01 -2.66841829e-01 -7.85041228e-02
-7.53538609e-01 2.93895900e-01 5.83881915e-01 4.95957136e-01
8.79112124e-01 -3.83545011e-02 -4.21021879e-01 1.32139564e+00
2.78038412e-01 3.25187474e-01 6.14303172e-01 -8.85965943e-01
3.75800788e-01 6.81437075e-01 -2.46300817e-01 -1.64368916e+00
-3.18370879e-01 -3.43374968e-01 -1.03705227e+00 3.46600711e-01
2.23328203e-01 -2.43725717e-01 -6.40344679e-01 2.33070588e+00
2.10795507e-01 1.13920979e-01 -7.56422058e-02 9.47235227e-01
7.08065152e-01 4.47577953e-01 -2.18363423e-02 -7.69741973e-03
1.04861844e+00 -6.83189511e-01 -6.71733618e-01 -6.88675418e-02
7.24964499e-01 -6.20207489e-01 9.33145344e-01 2.00180054e-01
-8.08992028e-01 -5.04866481e-01 -1.00171614e+00 2.95786291e-01
-3.43040317e-01 2.84219414e-01 5.39733410e-01 4.65348423e-01
-9.25715089e-01 7.57370293e-01 -5.43398499e-01 7.98099637e-02
8.06326926e-01 4.31954801e-01 -7.23501801e-01 3.09008788e-02
-1.21951890e+00 6.01837277e-01 1.53193027e-01 2.76972532e-01
-7.00145245e-01 -6.83215678e-01 -9.57674265e-01 1.80391937e-01
-7.89070725e-02 -3.40823263e-01 7.45017171e-01 -1.59529853e+00
-1.45531309e+00 7.83504605e-01 1.73479766e-01 -2.20834479e-01
5.11607707e-01 -4.24587876e-02 -5.89646637e-01 9.29479897e-02
-4.15226743e-02 7.61462152e-01 9.23934042e-01 -8.83166492e-01
-4.33293045e-01 -4.39477235e-01 -2.05978349e-01 7.33808652e-02
-6.33962631e-01 -1.15885884e-02 -4.35238592e-02 -7.71361053e-01
-1.09711729e-01 -9.01958108e-01 2.23483726e-01 4.65103626e-01
-3.13684106e-01 -2.53701657e-01 1.05563223e+00 -4.22778279e-01
1.14701962e+00 -2.48348856e+00 6.37053102e-02 3.29809129e-01
2.60133684e-01 3.95700872e-01 -5.76704025e-01 3.98035757e-02
-4.20562357e-01 -1.12087734e-01 -2.50051707e-01 -1.63574457e-01
-9.87235736e-03 1.60362393e-01 -7.04440847e-02 4.35020119e-01
7.24147260e-01 4.87520248e-01 -7.65368402e-01 -2.03513637e-01
-1.02305099e-01 8.10780108e-01 -6.88157976e-01 3.76475394e-01
2.73735583e-01 1.32369637e-01 -3.66668850e-01 5.73620975e-01
8.52355361e-01 1.27276093e-01 -9.31650847e-02 -4.47190881e-01
3.60138953e-01 -1.05597816e-01 -1.22737038e+00 1.24924755e+00
-4.84668314e-01 5.42704463e-01 1.55722558e-01 -1.07141960e+00
1.25332367e+00 1.42003849e-01 4.72841531e-01 -7.03117788e-01
3.37455958e-01 2.72790790e-01 2.40024105e-01 -3.04904699e-01
1.55802920e-01 -3.06249380e-01 1.24237470e-01 2.57355064e-01
2.35054255e-01 3.49939555e-01 -7.67283663e-02 -6.80598542e-02
1.04304600e+00 -2.33016193e-01 -1.15139531e-02 -7.81857222e-02
6.96974814e-01 -9.16613758e-01 7.92681932e-01 2.75390714e-01
-5.50833881e-01 6.15922391e-01 7.73963213e-01 -4.69980627e-01
-7.88579881e-01 -8.95771265e-01 -3.32176596e-01 9.17280734e-01
-4.14970554e-02 -4.26126510e-01 -6.67196214e-01 -1.13465130e+00
1.17716312e-01 4.44752097e-01 -8.75553191e-01 -7.70315051e-01
-2.47422278e-01 -6.40056074e-01 6.70527041e-01 5.23499966e-01
7.98937857e-01 -1.01197243e+00 -3.73987973e-01 9.61382017e-02
-8.65151361e-02 -8.34854603e-01 -6.04182422e-01 2.94041783e-01
-3.87406766e-01 -1.00267982e+00 -8.46805036e-01 -7.05551922e-01
7.65858471e-01 -1.15730442e-01 9.25714016e-01 4.75116074e-02
-4.52110231e-01 -1.82016753e-02 -4.43728596e-01 -1.77597180e-01
-8.80224258e-02 -2.71128356e-01 3.83330919e-02 7.14630902e-01
5.79348505e-01 -5.99723637e-01 -7.19423711e-01 6.77283347e-01
-9.42990422e-01 -3.14133644e-01 3.53307456e-01 1.27106106e+00
5.08341432e-01 -3.24250781e-03 4.00237948e-01 -5.13772666e-01
5.12977839e-01 -6.10571444e-01 -1.10808648e-01 9.42130089e-02
-4.58823442e-01 -2.95786615e-02 7.80953288e-01 -6.15915477e-01
-5.45453429e-01 -1.12979971e-01 -2.34708562e-01 -6.59364760e-01
8.49370807e-02 2.56994188e-01 -4.70601380e-01 -1.79744154e-01
4.76467580e-01 6.49490133e-02 2.64370829e-01 -2.69726157e-01
-2.87848916e-02 7.82943428e-01 1.01141125e-01 -4.48428690e-01
4.99317229e-01 3.05774987e-01 8.11460838e-02 -5.67018330e-01
-5.66607714e-01 9.80107710e-02 -4.19110000e-01 -2.17232674e-01
6.75823629e-01 -7.94880986e-01 -4.33282703e-01 6.65890276e-01
-8.90451133e-01 -1.60750270e-01 -3.81036192e-01 3.80029410e-01
-3.13014388e-01 9.26865339e-02 -2.56584257e-01 -5.95701277e-01
-4.17612553e-01 -1.31658244e+00 1.08883381e+00 3.11807126e-01
-2.99151152e-01 -8.81007075e-01 1.02947198e-01 -2.91527640e-02
4.44142431e-01 5.56098819e-01 7.94158697e-01 -6.34788275e-01
2.13510841e-01 -3.90612394e-01 -2.81188756e-01 9.90706801e-01
3.85040283e-01 1.36244044e-01 -8.86367321e-01 -3.95525336e-01
-1.22919656e-01 -6.89518332e-01 9.32676494e-01 5.29206879e-02
1.56653130e+00 -5.02382457e-01 -4.90617417e-02 9.29868758e-01
1.10696220e+00 1.29945546e-01 8.54807079e-01 1.98039129e-01
4.14174646e-01 6.10935986e-01 4.59610522e-01 5.64432740e-01
8.61273110e-02 9.63436067e-01 4.12902445e-01 -1.16882764e-01
2.26337723e-02 -2.37510338e-01 5.09488225e-01 4.41172749e-01
6.66434541e-02 -1.08469047e-01 -5.35559714e-01 3.50274831e-01
-1.56751621e+00 -1.06461680e+00 4.29269493e-01 1.96720719e+00
7.94728994e-01 -2.37232774e-01 1.56930864e-01 2.04946965e-01
6.40113711e-01 2.69501597e-01 -5.19370914e-01 -6.84210539e-01
-4.50989008e-01 3.38788718e-01 -1.09507069e-01 1.70007423e-01
-1.12027013e+00 7.76790917e-01 5.10526991e+00 1.07338715e+00
-1.52975798e+00 -1.03325136e-01 9.89552498e-01 -2.74318188e-01
-2.25573406e-01 -3.11882019e-01 -6.78703368e-01 8.03810954e-01
7.09112585e-01 8.91280845e-02 2.92313784e-01 1.00664914e+00
4.95642759e-02 1.84971556e-01 -1.17272007e+00 1.21941006e+00
1.77636608e-01 -8.47694755e-01 1.87821209e-01 1.17220402e-01
5.64960241e-01 -2.41322652e-01 3.42485785e-01 2.77611166e-01
9.11831670e-03 -1.23661089e+00 5.73115349e-01 4.15079683e-01
8.49848390e-01 -9.87262905e-01 8.98242414e-01 -1.50232583e-01
-1.09365451e+00 -1.35989800e-01 -5.43026507e-01 1.60098329e-01
-2.75931478e-01 5.94076335e-01 -5.22725046e-01 2.88207173e-01
8.15314412e-01 8.93968046e-01 -4.62091446e-01 7.57694304e-01
-1.44931361e-01 3.20692956e-01 -2.54383862e-01 1.95628628e-01
2.38847986e-01 -4.49608177e-01 4.60951000e-01 9.65595424e-01
3.72912258e-01 -2.22170785e-01 -1.12473264e-01 8.48912418e-01
-2.80830950e-01 2.26171553e-01 -5.52765012e-01 -6.64376691e-02
2.72358179e-01 1.32250965e+00 -1.19582415e-01 -1.11871257e-01
-1.47108555e-01 1.21199322e+00 5.34568548e-01 3.11113238e-01
-8.42642844e-01 -8.24910462e-01 1.12409580e+00 7.74472207e-02
3.52073342e-01 1.39435977e-01 4.20604348e-02 -1.07227480e+00
2.17673957e-01 -9.73670602e-01 4.34329778e-01 -4.98437583e-01
-1.61635149e+00 1.02716601e+00 -2.42092744e-01 -1.30566859e+00
-2.16861770e-01 -6.27414346e-01 -7.03802824e-01 9.12309587e-01
-1.55722654e+00 -9.50090945e-01 -4.43108976e-01 8.75868678e-01
5.19320853e-02 -5.28812647e-01 9.02352273e-01 5.13074160e-01
-6.45342529e-01 1.29761028e+00 2.82737941e-01 4.96815920e-01
7.46887743e-01 -8.48989666e-01 -4.16072488e-01 3.91277641e-01
-1.59402341e-02 4.78412271e-01 3.68735790e-01 -2.04363540e-01
-8.46564531e-01 -1.23555768e+00 9.13616598e-01 7.78739676e-02
4.67950135e-01 -5.05309463e-01 -9.17693794e-01 2.33940423e-01
-1.71637446e-01 5.50976098e-01 1.03726065e+00 -4.80224341e-02
-7.68165052e-01 -6.10220075e-01 -1.42036462e+00 4.33219999e-01
9.21803474e-01 -5.88047266e-01 -1.57430425e-01 8.75101984e-02
1.14520676e-01 3.22203189e-02 -8.71193647e-01 6.69635355e-01
8.43832016e-01 -1.16423154e+00 5.63337982e-01 -6.25667334e-01
7.02333093e-01 -1.48247734e-01 -4.13481891e-01 -1.56696200e+00
-3.46851140e-01 -3.17196399e-01 -5.03255464e-02 1.32119179e+00
1.85786769e-01 -6.95042074e-01 7.00671434e-01 4.93745893e-01
1.06273018e-01 -1.14796627e+00 -1.17978108e+00 -9.58935678e-01
1.47511631e-01 -1.83160290e-01 6.72359705e-01 1.06031835e+00
-1.95585221e-01 8.37616473e-02 -3.71855140e-01 -2.66900420e-01
4.10551995e-01 1.43547542e-02 4.78653342e-01 -1.09100342e+00
-2.94465512e-01 -6.30109668e-01 -8.84115338e-01 -6.13366604e-01
4.79370594e-01 -1.04228210e+00 -1.59032375e-01 -7.92659521e-01
2.57214785e-01 -5.59333265e-01 -4.98471171e-01 8.68529975e-01
-1.11493930e-01 5.06746352e-01 3.51897746e-01 6.06929772e-02
-3.82012546e-01 1.25121140e+00 1.03040385e+00 -3.07735473e-01
-2.07145855e-01 -9.64847505e-02 -6.19459987e-01 7.07622290e-01
6.46255732e-01 -5.20887852e-01 -1.91827461e-01 -3.03499609e-01
-3.98996919e-02 -3.92836362e-01 3.91031265e-01 -9.60910678e-01
-2.56169945e-01 -7.37476954e-03 7.47903526e-01 -1.59806199e-02
3.72268617e-01 -1.07647145e+00 6.52414635e-02 4.19979900e-01
-4.51110005e-01 -1.28545210e-01 1.42914504e-01 2.48714224e-01
-5.18646836e-01 7.25876018e-02 1.21066928e+00 3.13572377e-01
-4.14805502e-01 5.83785534e-01 -8.18928704e-02 1.82903334e-01
1.05740154e+00 -3.70550424e-01 3.48507985e-03 -4.51181740e-01
-6.52050793e-01 1.06936678e-01 2.71246403e-01 6.70142829e-01
8.21498394e-01 -1.86008644e+00 -7.87559807e-01 6.03792906e-01
3.53838235e-01 -4.60290521e-01 2.40888372e-01 7.00920522e-01
-1.74244821e-01 -1.05664376e-02 -6.06469929e-01 -5.53668857e-01
-1.32320297e+00 3.43170092e-02 8.16114902e-01 -1.37085438e-01
-1.41639084e-01 1.03912139e+00 4.04906005e-01 -4.58583504e-01
2.27992937e-01 7.77768567e-02 -2.31834069e-01 2.77642429e-01
7.98306048e-01 3.11493307e-01 2.84537300e-02 -9.50161874e-01
-5.44088304e-01 7.21153498e-01 -2.72100300e-01 2.02894241e-01
1.47652555e+00 2.08507806e-01 -1.29947692e-01 1.19520664e-01
1.86821067e+00 -3.62229615e-01 -1.37927556e+00 -2.97664195e-01
-3.28956276e-01 -7.30290890e-01 1.87245354e-01 -8.35462451e-01
-1.56459963e+00 8.80930722e-01 1.06041610e+00 -2.11163193e-01
1.46731484e+00 -1.53446779e-01 6.43934608e-01 -7.37428525e-03
1.51383132e-01 -1.19822347e+00 3.94732982e-01 5.10733366e-01
1.10325146e+00 -1.21544230e+00 -3.15830946e-01 -1.26414135e-01
-6.64947331e-01 1.13598871e+00 7.78253198e-01 -2.40569383e-01
7.88705707e-01 1.22433826e-01 -1.04390360e-01 -1.81275204e-01
-5.35419464e-01 1.95804596e-01 4.48350519e-01 5.13521314e-01
5.08659899e-01 -7.53392950e-02 -3.42729211e-01 7.53933430e-01
-1.82991564e-01 2.46514417e-02 -2.03571492e-03 6.54795349e-01
-5.43752909e-02 -1.12710965e+00 -4.85828109e-02 4.61937666e-01
-3.66472870e-01 9.25362483e-02 -5.98746538e-01 6.00947261e-01
4.36251491e-01 7.74810314e-01 3.76359403e-01 -7.96315789e-01
2.89579153e-01 4.91236560e-02 2.96376526e-01 -2.56797314e-01
-5.33277512e-01 -1.23877250e-01 -2.15468019e-01 -8.35186005e-01
-2.31344968e-01 -4.56483185e-01 -1.11361778e+00 -2.38791063e-01
-2.26101682e-01 1.14846930e-01 5.21964967e-01 6.95458651e-01
5.91118276e-01 2.62880802e-01 1.34120071e+00 -5.96441090e-01
-6.87714875e-01 -8.20693374e-01 -8.55447650e-01 8.27724099e-01
3.25633585e-01 -8.76494586e-01 -6.13050282e-01 -3.18217665e-01] | [13.446634292602539, 1.5421992540359497] |
7158ecc2-1461-437b-a121-d6bf81a4ada9 | robot-cooking-with-stir-fry-bimanual-non | 2205.05960 | null | https://arxiv.org/abs/2205.05960v1 | https://arxiv.org/pdf/2205.05960v1.pdf | Robot Cooking with Stir-fry: Bimanual Non-prehensile Manipulation of Semi-fluid Objects | This letter describes an approach to achieve well-known Chinese cooking art stir-fry on a bimanual robot system. Stir-fry requires a sequence of highly dynamic coordinated movements, which is usually difficult to learn for a chef, let alone transfer to robots. In this letter, we define a canonical stir-fry movement, and then propose a decoupled framework for learning this deformable object manipulation from human demonstration. First, the dual arms of the robot are decoupled into different roles (a leader and follower) and learned with classical and neural network-based methods separately, then the bimanual task is transformed into a coordination problem. To obtain general bimanual coordination, we secondly propose a Graph and Transformer based model -- Structured-Transformer, to capture the spatio-temporal relationship between dual-arm movements. Finally, by adding visual feedback of content deformation, our framework can adjust the movements automatically to achieve the desired stir-fry effect. We verify the framework by a simulator and deploy it on a real bimanual Panda robot system. The experimental results validate our framework can realize the bimanual robot stir-fry motion and have the potential to extend to other deformable objects with bimanual coordination. | ['Fei Chen', 'Miao Li', 'Sylvain Calinon', 'Shixiong Wang', 'Zhipeng Dong', 'Yiting Chen', 'Junjia Liu'] | 2022-05-12 | null | null | null | null | ['deformable-object-manipulation'] | ['robots'] | [-2.18981996e-01 1.40198112e-01 1.51629850e-01 5.41040488e-02
-2.18255296e-01 -7.13652790e-01 5.11530459e-01 -5.22603154e-01
-1.15812957e-01 5.97879291e-01 -2.37304106e-01 2.21815109e-01
-4.19537395e-01 -7.20039904e-01 -1.32175314e+00 -8.34256530e-01
-3.52537662e-01 7.97352374e-01 3.29234660e-01 -8.49530935e-01
-8.54303539e-02 3.67471218e-01 -1.11779618e+00 -4.34981845e-02
8.95830750e-01 3.74548972e-01 9.48798001e-01 7.44563758e-01
4.74379689e-01 9.72139716e-01 -1.89019158e-01 2.43625134e-01
4.23114330e-01 -2.89740592e-01 -1.09595013e+00 1.94890171e-01
1.11273125e-01 -4.28390920e-01 -3.81453663e-01 8.42219174e-01
3.13532740e-01 4.13644850e-01 7.76166201e-01 -1.34626591e+00
-8.86227190e-01 1.01663601e+00 -7.09212244e-01 -4.27629203e-01
5.20844162e-01 3.60710979e-01 6.55837715e-01 -3.03962588e-01
8.53229702e-01 1.75348341e+00 5.47237337e-01 7.30357826e-01
-8.42520714e-01 -5.21336734e-01 3.83638769e-01 3.08354080e-01
-9.53025937e-01 2.39766702e-01 8.74387443e-01 -4.76080567e-01
6.91503584e-01 5.78649640e-02 9.76336002e-01 1.02331603e+00
3.49583745e-01 1.06196272e+00 6.42857015e-01 -5.18521070e-01
-3.71894091e-01 -7.46849060e-01 -2.00270507e-02 8.33579302e-01
1.66095302e-01 6.11875728e-02 6.46762084e-03 4.31072831e-01
1.18139994e+00 2.85892785e-01 -3.29491645e-01 -8.88859689e-01
-1.60452259e+00 4.65168893e-01 8.39755535e-01 4.06955540e-01
-3.59027863e-01 8.14098895e-01 2.61685878e-01 5.06286681e-01
-3.86964828e-01 4.27305937e-01 -6.27947628e-01 1.64429322e-02
-6.12232834e-02 7.69460320e-01 7.62452126e-01 1.34513283e+00
5.74949741e-01 2.67876443e-02 8.22730586e-02 5.79715669e-01
2.15115398e-01 5.33846080e-01 1.48582757e-01 -1.17842245e+00
7.13479042e-01 4.32153463e-01 4.93668318e-01 -1.10467172e+00
-8.32357049e-01 3.54169756e-01 -7.63562977e-01 3.86439562e-01
5.67379951e-01 -6.92616105e-02 -4.27988917e-01 1.61413646e+00
5.19508660e-01 -2.75559992e-01 -2.11596638e-02 1.38094246e+00
4.73130167e-01 7.33960688e-01 -1.79524198e-01 -3.84131772e-03
1.20038581e+00 -1.39233422e+00 -7.57985413e-01 2.31316030e-01
6.66458726e-01 -4.27780092e-01 1.17512608e+00 4.25940424e-01
-1.17873681e+00 -6.67008400e-01 -1.06214714e+00 5.12287626e-03
5.39423488e-02 9.33833495e-02 3.29389662e-01 -2.48666137e-01
-5.47987938e-01 8.97437513e-01 -1.25862193e+00 -3.56240749e-01
-1.30402148e-01 3.83610934e-01 -4.91773486e-01 9.86328647e-02
-1.11758959e+00 1.21484196e+00 6.82814837e-01 4.77411896e-01
-1.13099360e+00 -3.80216688e-01 -8.27990592e-01 -2.59375781e-01
8.29295278e-01 -9.15077984e-01 1.39838362e+00 -7.73717225e-01
-2.16869307e+00 4.56268281e-01 7.02316284e-01 -7.00992346e-02
7.74447620e-01 -6.65745795e-01 2.25105464e-01 2.32293785e-01
9.48838983e-03 6.31446123e-01 6.91594481e-01 -1.47728860e+00
-3.30823720e-01 -1.36710122e-01 4.93516594e-01 6.24270618e-01
-1.68019440e-03 -2.83923447e-01 -3.01927149e-01 -8.00904810e-01
6.24200329e-02 -1.42170596e+00 -2.24618152e-01 2.99525738e-01
-4.07355607e-01 -4.84812140e-01 1.25312281e+00 -5.15432477e-01
4.60864961e-01 -1.98020983e+00 9.44660127e-01 -9.49559882e-02
1.73460711e-02 7.53800347e-02 -2.87830949e-01 6.10430896e-01
8.42675939e-02 -3.47247362e-01 -2.04166338e-01 3.31895977e-01
1.06508918e-01 5.65973818e-01 8.65043774e-02 5.60982347e-01
7.20454752e-02 1.12487340e+00 -1.24208784e+00 -3.48437816e-01
2.90921479e-01 1.96622908e-01 -5.86476624e-01 5.25554955e-01
-4.97995406e-01 8.25142503e-01 -8.94632995e-01 5.26422083e-01
6.26719415e-01 -5.28993420e-02 4.75412339e-01 -3.59196216e-01
-2.12847665e-01 -3.79026443e-01 -1.20388508e+00 1.99870491e+00
-4.47949737e-01 1.78837478e-01 6.54634655e-01 -1.29479349e+00
9.93069708e-01 1.20080791e-01 4.87577915e-01 -4.02743995e-01
3.07588369e-01 1.05442636e-01 1.69114113e-01 -1.06825542e+00
3.57718021e-01 1.91891238e-01 -2.98391074e-01 3.37601364e-01
-1.73980072e-01 -7.84902871e-01 1.48158506e-01 -2.23581925e-01
9.40040529e-01 1.01958823e+00 2.38603950e-01 -4.84876245e-01
5.47032773e-01 3.48217934e-01 2.49058008e-01 4.77521777e-01
1.48669071e-02 3.51336092e-01 4.85743731e-02 -4.80201244e-01
-9.24506545e-01 -1.18295562e+00 4.93720561e-01 1.08351159e+00
7.75429487e-01 1.45190373e-01 -7.78834403e-01 -6.10194981e-01
8.17151293e-02 3.45255405e-01 -4.71804172e-01 -1.58088297e-01
-1.35580981e+00 -2.54217207e-01 2.04390272e-01 6.40540123e-01
6.45070851e-01 -1.51209474e+00 -1.05668604e+00 3.91136289e-01
-3.90184373e-01 -9.79333520e-01 -6.85715675e-01 3.77987549e-02
-6.38693094e-01 -1.52517283e+00 -7.50450492e-01 -1.39869893e+00
5.79352617e-01 4.26929772e-01 6.46135628e-01 6.30410984e-02
-3.63183260e-01 6.89353347e-01 -6.55760169e-01 2.41138153e-02
-5.87267220e-01 -1.70598850e-01 1.00362159e-01 -5.13788998e-01
-5.81524074e-01 -7.75195301e-01 -6.42470002e-01 6.41008914e-01
-7.69752622e-01 1.98050633e-01 6.83986366e-01 9.19498980e-01
3.92791152e-01 -2.91266274e-02 1.93185806e-01 -1.16232961e-01
4.04075980e-01 2.93270014e-02 -6.62161767e-01 1.58764765e-01
1.90829903e-01 -1.60302803e-01 8.73293698e-01 -1.09458601e+00
-1.03079975e+00 4.16876227e-01 3.41727674e-01 -6.06949687e-01
9.28068161e-02 2.31967643e-01 -2.68319875e-01 -7.92282373e-02
4.64092582e-01 7.35944360e-02 2.95827210e-01 -3.40731055e-01
8.55165660e-01 2.77422756e-01 7.27309346e-01 -1.02927530e+00
9.79545474e-01 5.03337741e-01 1.61074981e-01 -5.63083768e-01
-2.09202424e-01 -8.28422308e-02 -1.02026296e+00 -4.79800195e-01
1.03529823e+00 -7.08362937e-01 -1.59009469e+00 7.37940609e-01
-1.43409407e+00 -1.01530349e+00 -6.79468960e-02 5.32518804e-01
-1.14578247e+00 4.83047605e-01 -9.27621484e-01 -3.55160147e-01
-2.86758006e-01 -1.17924333e+00 1.04637504e+00 1.20184571e-01
-2.18555212e-01 -6.96400225e-01 1.47497132e-01 1.74810410e-01
2.97539383e-02 6.15857899e-01 8.24485779e-01 3.95552590e-02
-6.96428955e-01 2.07737312e-02 6.08199649e-02 2.86485292e-02
2.35838279e-01 -1.96340606e-02 1.25591382e-01 -5.58432102e-01
-2.33953610e-01 -5.83598316e-01 3.20805460e-01 3.04515123e-01
1.07767689e+00 -5.88002026e-01 -5.72817504e-01 2.17249721e-01
1.15184438e+00 3.81728083e-01 4.19680387e-01 5.48156500e-01
1.04669285e+00 6.48245633e-01 9.78389978e-01 1.95631579e-01
5.77079535e-01 1.14425325e+00 9.38882709e-01 2.07895681e-01
-4.36539650e-02 -2.41319641e-01 5.68435550e-01 1.02915311e+00
-5.93221247e-01 -3.29205133e-02 -6.02730870e-01 4.89434630e-01
-2.29137087e+00 -8.79858136e-01 -2.79690385e-01 1.48567009e+00
7.56369591e-01 -3.10327679e-01 2.06490591e-01 -1.35491237e-01
8.16060543e-01 -2.25242227e-01 -3.74196678e-01 -1.49458975e-01
3.45843911e-01 -3.09407264e-01 2.41514519e-01 2.89596260e-01
-9.82671559e-01 9.97634172e-01 5.38870192e+00 6.78346217e-01
-1.13832605e+00 -3.68906818e-02 -4.43837255e-01 1.54195055e-01
1.84866711e-01 -1.37441143e-01 -1.60998225e-01 3.26932490e-01
-6.85010552e-02 1.59965888e-01 6.82586432e-01 8.18255246e-01
3.14345002e-01 1.89418882e-01 -1.36125398e+00 7.72594988e-01
-2.48529702e-01 -9.79374528e-01 -4.81772423e-02 -5.37965059e-01
6.37849450e-01 -4.64818358e-01 -2.14527979e-01 2.86773980e-01
8.72631013e-01 -6.17819488e-01 1.19396937e+00 4.98286664e-01
4.26413447e-01 -3.25641394e-01 1.52741700e-01 8.21712255e-01
-1.43598783e+00 -3.38438541e-01 -2.56243557e-01 -2.18628377e-01
3.45178902e-01 -7.61145279e-02 -4.18705493e-01 9.70507205e-01
9.18790758e-01 9.22635019e-01 2.20786810e-01 5.85734129e-01
-2.80797631e-01 -1.28119132e-02 -2.52880633e-01 -5.19425035e-01
3.25436354e-01 -3.94892752e-01 8.74734282e-01 9.26750898e-01
3.23638380e-01 3.08537364e-01 4.89240617e-01 8.33118379e-01
2.57511735e-01 -1.35788083e-01 -5.20305216e-01 3.12410772e-01
1.80805564e-01 1.29962707e+00 -6.61743402e-01 -5.00510745e-02
1.99259017e-02 1.13261783e+00 5.11028707e-01 1.69643044e-01
-1.20729196e+00 -3.61449033e-01 3.26059133e-01 -4.64046858e-02
4.28282678e-01 -5.90347171e-01 4.63900477e-01 -1.00508606e+00
3.00951391e-01 -8.58109891e-01 6.90548122e-02 -1.02919006e+00
-1.22749639e+00 8.10558274e-02 3.68725061e-01 -1.58752942e+00
-1.82141051e-01 -6.54204190e-01 -5.77852428e-01 2.42506474e-01
-1.17854595e+00 -1.51046741e+00 -5.76546252e-01 6.31494820e-01
7.35598505e-01 2.69957900e-01 6.03707135e-01 1.21156750e-02
-2.95684993e-01 1.22732662e-01 -1.24880254e-01 1.61192358e-01
5.06002843e-01 -1.13626337e+00 -3.92613336e-02 2.29024380e-01
-3.52501273e-01 4.08714354e-01 6.10087812e-01 -5.47330797e-01
-1.96341419e+00 -1.11612427e+00 -1.24374598e-01 -3.60656202e-01
8.02006900e-01 -2.31443733e-01 -6.69991612e-01 9.47777390e-01
1.91579521e-01 4.21988703e-02 -4.81956691e-01 -3.29355150e-01
4.24097069e-02 -2.00368837e-01 -8.92389238e-01 8.24800014e-01
1.48569608e+00 2.12757528e-01 -9.80018795e-01 4.57782865e-01
1.01679325e+00 -8.36380482e-01 -6.66683912e-01 6.01828456e-01
8.16611767e-01 -3.97194237e-01 9.83083427e-01 -5.73634565e-01
5.85491180e-01 -7.82004118e-01 6.31313249e-02 -1.62744272e+00
-5.44115961e-01 -8.59069049e-01 -1.31199919e-02 7.49264777e-01
-1.52027473e-01 -3.83063614e-01 4.00622547e-01 -3.55457515e-01
-3.72630298e-01 -6.21569395e-01 -8.84226859e-01 -1.19374979e+00
1.52075842e-01 3.04548107e-02 4.95533288e-01 1.05960596e+00
3.14889759e-01 1.12544887e-01 -7.06553042e-01 4.11471039e-01
2.63733566e-01 3.99619460e-01 1.26090837e+00 -8.81524205e-01
-5.20927310e-01 -4.19817001e-01 -1.91601470e-01 -1.41846263e+00
2.31497541e-01 -9.33811724e-01 5.34917474e-01 -1.70669949e+00
1.06776558e-01 -5.01117766e-01 3.31049353e-01 6.09076500e-01
7.24292845e-02 -1.56522289e-01 6.49121702e-01 2.88274705e-01
-6.25231087e-01 7.68539310e-01 2.25718069e+00 -3.23896140e-01
-4.59794641e-01 -2.22566515e-01 1.37963165e-02 8.45606565e-01
4.92391795e-01 -2.84943819e-01 -2.49876350e-01 -5.29906809e-01
-4.05094028e-02 5.12564301e-01 6.41638756e-01 -8.85163069e-01
2.31407303e-02 -5.84917486e-01 -7.14736432e-02 -4.20258999e-01
1.70527995e-01 -1.23731983e+00 4.10067856e-01 1.12877131e+00
-2.36596271e-01 1.78047478e-01 -2.32966547e-03 6.44850731e-01
6.71188235e-02 6.45495951e-02 6.19853914e-01 -2.50954270e-01
-6.56528413e-01 1.24672189e-01 -3.21080238e-01 -2.80785263e-01
1.50703263e+00 -4.74694557e-02 -1.91400498e-01 -1.39627337e-01
-1.03841293e+00 7.30658114e-01 2.88777798e-01 6.90834761e-01
4.14611101e-01 -1.45071721e+00 -5.89588881e-01 -2.78908134e-01
-2.32555568e-01 4.43543583e-01 4.54038829e-01 9.31577027e-01
-9.28386867e-01 3.41233090e-02 -6.51558042e-01 -9.36382830e-01
-1.21788716e+00 7.66998172e-01 3.78828555e-01 -3.57251436e-01
-1.05146980e+00 2.49064520e-01 3.60945314e-01 -9.50128376e-01
-1.52845144e-01 -5.97156823e-01 -2.99754679e-01 -5.44822216e-01
-1.11730389e-01 6.43456995e-01 -4.71828640e-01 -3.48178208e-01
-1.25670433e-01 1.04167318e+00 2.06091657e-01 2.74629563e-01
1.48523414e+00 -1.19840287e-01 -2.58365542e-01 2.17315942e-01
8.32229435e-01 -3.90081406e-01 -1.52147365e+00 2.27890953e-01
-3.89241546e-01 -1.04047559e-01 -9.41798389e-01 -4.16678846e-01
-9.16293025e-01 5.10044396e-01 2.69212365e-01 9.31281298e-02
8.26018095e-01 2.05607668e-01 8.46308053e-01 9.47821498e-01
7.17545092e-01 -1.17539382e+00 8.10042977e-01 6.54044271e-01
1.59180725e+00 -7.17047811e-01 -1.72984377e-01 -6.75076783e-01
-5.58319330e-01 1.40789664e+00 9.13184524e-01 -7.35270917e-01
5.19763052e-01 4.11560297e-01 1.65463537e-02 -1.10584952e-01
-3.88631314e-01 1.83595359e-01 1.59549221e-01 8.29621255e-01
-1.46896571e-01 2.99713105e-01 -4.28325087e-01 5.69074571e-01
-1.82831675e-01 6.52227402e-02 5.50503194e-01 1.13660097e+00
-4.65775639e-01 -8.67215097e-01 -4.91416454e-01 -2.77470261e-01
2.49025211e-01 7.18214989e-01 4.46331725e-02 1.54750729e+00
1.92815602e-01 6.09187722e-01 -7.05574825e-02 -2.30708629e-01
7.96367705e-01 -5.00620723e-01 1.15326464e+00 -4.01006311e-01
-4.00731653e-01 1.33702084e-01 -1.59070268e-01 -8.23133647e-01
-7.86422968e-01 -7.02057600e-01 -1.73688912e+00 -8.95566344e-02
-3.02893758e-01 -2.24962592e-01 1.37836576e-01 9.29072618e-01
-2.06605628e-01 6.86792552e-01 5.49633563e-01 -1.58680022e+00
-8.69354486e-01 -9.58411038e-01 -4.13532525e-01 7.22727835e-01
4.01849419e-01 -1.09426832e+00 1.23585411e-03 2.88140774e-01] | [4.847192764282227, 0.5923830270767212] |
a958c64c-0258-431f-869f-7709eecc85f3 | multi-agent-deep-reinforcement-learning-for-8 | 2111.02258 | null | https://arxiv.org/abs/2111.02258v1 | https://arxiv.org/pdf/2111.02258v1.pdf | Multi-Agent Deep Reinforcement Learning For Optimising Energy Efficiency of Fixed-Wing UAV Cellular Access Points | Unmanned Aerial Vehicles (UAVs) promise to become an intrinsic part of next generation communications, as they can be deployed to provide wireless connectivity to ground users to supplement existing terrestrial networks. The majority of the existing research into the use of UAV access points for cellular coverage considers rotary-wing UAV designs (i.e. quadcopters). However, we expect fixed-wing UAVs to be more appropriate for connectivity purposes in scenarios where long flight times are necessary (such as for rural coverage), as fixed-wing UAVs rely on a more energy-efficient form of flight when compared to the rotary-wing design. As fixed-wing UAVs are typically incapable of hovering in place, their deployment optimisation involves optimising their individual flight trajectories in a way that allows them to deliver high quality service to the ground users in an energy-efficient manner. In this paper, we propose a multi-agent deep reinforcement learning approach to optimise the energy efficiency of fixed-wing UAV cellular access points while still allowing them to deliver high-quality service to users on the ground. In our decentralized approach, each UAV is equipped with a Dueling Deep Q-Network (DDQN) agent which can adjust the 3D trajectory of the UAV over a series of timesteps. By coordinating with their neighbours, the UAVs adjust their individual flight trajectories in a manner that optimises the total system energy efficiency. We benchmark the performance of our approach against a series of heuristic trajectory planning strategies, and demonstrate that our method can improve the system energy efficiency by as much as 70%. | ['Ivana Dusparic', 'Babatunji Omoniwa', 'Boris Galkin'] | 2021-11-03 | null | null | null | null | ['trajectory-planning'] | ['robots'] | [-3.42374176e-01 2.35961154e-01 -2.43248671e-01 4.64949518e-01
6.22445829e-02 -1.06938255e+00 1.86896041e-01 -3.26625966e-02
-1.80041909e-01 1.07157850e+00 -5.32473087e-01 -7.14487374e-01
-6.20044470e-01 -1.38014793e+00 -4.50488538e-01 -9.29122329e-01
-8.23766470e-01 3.53192210e-01 -8.00971463e-02 -7.52639890e-01
-3.08535784e-01 8.88348341e-01 -1.21816158e+00 -8.62335026e-01
8.17695558e-01 9.75429296e-01 1.70723051e-01 8.04691792e-01
8.49154472e-01 -1.21368002e-02 -5.41652799e-01 6.10531271e-02
6.47592485e-01 -2.92308688e-01 -6.53650880e-01 2.57332504e-01
-3.54814649e-01 -5.25465190e-01 -4.71666127e-01 3.61634731e-01
6.46820068e-01 5.09803116e-01 5.87593257e-01 -1.48052287e+00
9.10330042e-02 -1.09106780e-03 -3.96162331e-01 -8.28056247e-04
1.37718529e-01 2.86266476e-01 8.84170830e-01 -1.25438562e-02
1.63004234e-01 5.15825927e-01 5.59113145e-01 3.25377524e-01
-8.13318372e-01 -2.41366610e-01 1.96338847e-01 -3.58691067e-01
-1.30090356e+00 -6.39518678e-01 1.08263053e-01 8.34806189e-02
1.07947171e+00 3.05512786e-01 1.52498949e+00 2.72311658e-01
5.63026607e-01 1.60973325e-01 1.46234199e-01 -2.18824819e-01
4.95399714e-01 -6.02368474e-01 -1.09281230e+00 8.99928093e-01
5.44759214e-01 2.35934243e-01 2.29469538e-01 -1.16267331e-01
7.81034172e-01 -1.41606480e-01 -5.69084406e-01 -4.91460294e-01
-1.19579232e+00 8.69681776e-01 1.05803752e+00 6.43330738e-02
-1.06048954e+00 6.95782781e-01 -1.46081418e-01 4.95097816e-01
2.07583264e-01 1.04585969e+00 -5.98290682e-01 2.90949978e-02
-9.46417391e-01 5.14282048e-01 7.35206962e-01 1.18718672e+00
3.03871632e-01 4.13648725e-01 -8.64184722e-02 3.46862286e-01
3.95563692e-01 6.52063847e-01 -1.00933202e-01 -1.32603478e+00
8.58121365e-02 3.01072925e-01 5.51650882e-01 -6.73053026e-01
-6.84253395e-01 -8.59081209e-01 -9.45662022e-01 3.00045639e-01
-1.37602285e-01 -1.26262021e+00 -7.47762144e-01 1.52761555e+00
5.74576318e-01 -1.24843121e-01 3.91701311e-01 9.50234354e-01
-2.95825154e-02 1.07506943e+00 -3.08023483e-01 -5.22385776e-01
1.05483007e+00 -8.27586889e-01 -1.67472467e-01 -5.71548082e-02
7.69294024e-01 -3.81551772e-01 2.22233653e-01 1.18152618e-01
-1.15441799e+00 -1.91256851e-02 -1.16456258e+00 7.21384704e-01
-2.93670774e-01 -1.23292126e-01 3.51458251e-01 7.39565074e-01
-1.53562391e+00 5.85305750e-01 -7.38060176e-01 -5.98781109e-01
4.20971900e-01 7.49111533e-01 1.83202013e-01 -3.31577891e-03
-1.16506350e+00 7.38704681e-01 1.76980659e-01 -1.49706095e-01
-1.05101144e+00 -5.14338672e-01 -5.59719265e-01 2.11640686e-01
5.30882955e-01 -1.20720637e+00 1.34365773e+00 -6.39725506e-01
-1.50435925e+00 -5.68089262e-02 4.23009634e-01 -6.88490510e-01
2.60553837e-01 4.37492609e-01 -2.17544846e-02 3.37260336e-01
-4.09526527e-02 1.15496171e+00 6.85558081e-01 -9.11639750e-01
-1.10805702e+00 2.81323176e-02 8.43753576e-01 8.64998937e-01
-3.49646062e-01 -7.09690928e-01 -1.28274225e-03 -4.20890570e-01
-2.79042542e-01 -1.60510731e+00 -6.48139715e-01 1.17799945e-01
-1.33181036e-01 -1.10034660e-01 1.13096452e+00 2.99188066e-02
9.41676557e-01 -1.53891647e+00 5.33652902e-01 2.13622570e-01
1.13236882e-01 1.30643204e-01 1.75400041e-02 8.84101808e-01
5.61049581e-01 2.67624617e-01 -9.96631831e-02 8.56898874e-02
-3.15145224e-01 4.11786109e-01 6.28064722e-02 4.41267699e-01
-1.10213816e-01 5.80792606e-01 -1.03596795e+00 1.43264860e-01
6.02807887e-02 2.89615452e-01 -6.65108025e-01 2.57078912e-02
-3.73277187e-01 2.84480751e-01 -6.77913666e-01 8.12931597e-01
3.38276684e-01 7.39835650e-02 1.56419247e-01 1.26309335e-01
-4.69037980e-01 -3.45678657e-01 -5.65918803e-01 1.32790709e+00
-7.11372912e-01 5.65651238e-01 6.20127499e-01 -8.21223378e-01
4.39647168e-01 5.67925990e-01 1.06883788e+00 -1.74014896e-01
1.31702617e-01 1.85489565e-01 4.63707708e-02 -6.28069565e-02
7.42050290e-01 3.23999114e-02 2.01845076e-02 8.22807550e-02
-2.60383606e-01 -5.84485710e-01 1.36675507e-01 1.57013834e-01
1.34889388e+00 -1.63250297e-01 1.73961803e-01 -4.02201772e-01
1.33113652e-01 2.53343523e-01 3.05716038e-01 2.63094217e-01
-1.26170605e-01 -2.31593326e-01 1.36398688e-01 -4.40026522e-01
-9.90635931e-01 -5.76893926e-01 1.24408036e-01 4.62993175e-01
4.82018083e-01 -1.46247432e-01 -6.76540196e-01 -3.94736886e-01
6.19144924e-02 6.88815117e-01 -3.70706758e-03 -1.59770772e-01
2.63921395e-02 -7.26552844e-01 4.18492198e-01 -2.63505373e-02
7.77236402e-01 -3.68117481e-01 -1.47166193e+00 5.07863224e-01
1.95472571e-03 -9.85314965e-01 -3.72354269e-01 3.21233571e-01
-6.78868175e-01 -9.51292038e-01 -8.39833677e-01 -4.01589721e-01
6.60301268e-01 9.64316130e-01 6.50919676e-01 4.24453229e-01
1.72372162e-01 6.75153434e-01 -6.11883759e-01 -3.63536239e-01
1.59575418e-01 4.86178488e-01 4.10275280e-01 -2.99431741e-01
-6.73112392e-01 -5.22984624e-01 -9.32457209e-01 4.66414183e-01
-6.14066243e-01 -1.63643122e-01 3.46879929e-01 4.70798284e-01
6.41301394e-01 8.34874213e-01 6.96684480e-01 -9.76365358e-02
6.87316656e-01 -9.75907862e-01 -8.90047073e-01 -1.14987984e-01
-5.62312126e-01 -3.35350066e-01 8.90197635e-01 2.39374172e-02
-1.63137972e-01 2.55352885e-01 6.95048124e-02 -4.46830422e-01
2.90066779e-01 4.84725386e-01 1.22522742e-01 -7.87311435e-01
2.88216919e-01 -1.47889286e-01 -6.98937103e-02 4.50460553e-01
2.82395452e-01 4.92629707e-01 -1.90035582e-01 -3.88138086e-01
1.01270390e+00 2.28057623e-01 8.27492297e-01 -1.21431398e+00
-1.58006817e-01 2.84555927e-02 -3.06623012e-01 -6.94080353e-01
6.87544227e-01 -1.04633117e+00 -8.19166660e-01 1.03832811e-01
-8.83303881e-01 -7.56377518e-01 -2.49157280e-01 2.59389192e-01
-6.78491890e-01 -1.35649741e-01 3.37639377e-02 -1.00995064e+00
-3.45304400e-01 -1.06329417e+00 9.34714735e-01 4.90768373e-01
1.02250256e-01 -9.32006538e-01 -1.00005016e-01 -1.98746935e-01
6.52446270e-01 8.21662843e-01 7.15512514e-01 3.14652234e-01
-7.64439404e-01 -1.51787266e-01 3.67222369e-01 -3.54638517e-01
2.81650573e-01 8.97052214e-02 -3.34354579e-01 -1.03305721e+00
-8.07323933e-01 -9.02689919e-02 3.02973628e-01 8.28454912e-01
9.60961223e-01 -5.50952137e-01 -8.20282340e-01 8.62755835e-01
1.55755532e+00 6.20564520e-01 2.67126083e-01 1.57124966e-01
1.09104544e-01 3.05027008e-01 8.38449717e-01 5.80879092e-01
5.10605991e-01 8.26081276e-01 1.48398721e+00 -6.79074451e-02
4.37500000e-01 9.07262936e-02 2.51760572e-01 2.32349947e-01
-5.66874444e-01 -1.24432051e+00 -6.02822185e-01 6.02291942e-01
-1.71976519e+00 -7.03049064e-01 1.30991906e-01 2.32793307e+00
-3.96893695e-02 -1.53470039e-01 3.43304038e-01 2.91892458e-02
3.94695878e-01 7.92924836e-02 -7.92251229e-01 -4.87568885e-01
2.32260808e-01 -2.44895041e-01 1.13897407e+00 1.94027841e-01
-9.83077526e-01 6.78337216e-01 5.57196712e+00 5.09850919e-01
-9.26159382e-01 -2.14918315e-01 4.15829211e-01 -4.01857942e-01
-2.59328246e-01 5.53343631e-02 -2.96458751e-01 3.46600562e-01
1.29705358e+00 -1.42335102e-01 9.28321242e-01 3.97946626e-01
6.32578969e-01 -2.40143597e-01 -5.42502761e-01 5.14191926e-01
-6.60255671e-01 -1.39133358e+00 -2.44924009e-01 8.28523815e-01
8.67305398e-01 2.60637790e-01 -1.21680111e-01 3.78115731e-03
3.37160587e-01 -6.68760300e-01 6.06531441e-01 3.19916666e-01
9.34969902e-01 -1.46023691e+00 5.85569084e-01 6.60424829e-01
-1.50589859e+00 -5.88760495e-01 -3.62360090e-01 -1.88910916e-01
4.12389636e-01 4.83211160e-01 -8.57209980e-01 9.52652454e-01
6.66084826e-01 3.53022963e-01 1.78218961e-01 1.23580837e+00
1.34343877e-01 1.85635686e-01 -6.28030360e-01 -3.97056669e-01
8.00077915e-01 -3.74718457e-01 1.00597060e+00 3.07361096e-01
1.16649330e+00 5.37480831e-01 4.24899846e-01 2.40277305e-01
-3.03614050e-01 -2.39315629e-01 -1.17700100e+00 -2.52462357e-01
8.73666406e-01 1.42678702e+00 -8.07861686e-01 1.62665069e-01
-1.73206329e-01 8.05500209e-01 -1.87230274e-01 2.64679670e-01
-8.15736353e-01 -7.83283770e-01 1.04574203e+00 3.68632466e-01
2.29267478e-01 -7.10784316e-01 4.85304087e-01 -5.99131703e-01
-4.88449842e-01 -4.24552917e-01 -8.81160870e-02 -8.90679300e-01
-5.67322791e-01 7.62514889e-01 -2.01844305e-01 -1.47769940e+00
-2.51261592e-01 -1.41240299e-01 -5.71819186e-01 5.59735119e-01
-1.63336337e+00 -1.02814627e+00 -3.42821479e-01 4.43508208e-01
5.46656847e-01 -3.97739857e-01 9.65308726e-01 9.33109957e-04
-5.98931611e-01 1.84644517e-02 1.25448212e-01 -5.12291789e-01
6.08590990e-02 -9.36094105e-01 3.11597615e-01 9.05298054e-01
-2.63818502e-01 1.71771869e-02 6.15946770e-01 -6.28414154e-01
-1.98476791e+00 -1.45518804e+00 2.68514734e-02 1.51921034e-01
2.42084950e-01 3.12572896e-01 2.48628274e-01 3.66730481e-01
6.74196899e-01 3.79458815e-02 4.36058670e-01 -3.98258656e-01
9.81295764e-01 -1.29607156e-01 -1.32559037e+00 9.02673542e-01
9.57067668e-01 3.74488831e-01 5.43924928e-01 5.85928380e-01
7.80695736e-01 -3.41677189e-01 -8.89020383e-01 4.04383630e-01
5.32269597e-01 -4.94808614e-01 6.82568312e-01 -3.32633346e-01
-1.64614961e-01 -6.14111483e-01 -3.56973678e-01 -2.20355725e+00
-6.12490594e-01 -1.22772884e+00 -3.02982152e-01 6.72799230e-01
4.13511038e-01 -6.81308627e-01 6.23024106e-01 2.93479413e-02
-4.35794055e-01 -1.10021794e+00 -1.15653455e+00 -7.90151179e-01
-7.36553743e-02 2.09797800e-01 8.39308858e-01 5.95484376e-01
-2.07453921e-01 3.11304927e-01 -2.38714233e-01 8.72310281e-01
4.60811079e-01 -1.74928904e-01 6.27317250e-01 -1.37390709e+00
-8.64562839e-02 -1.91851735e-01 -1.70432687e-01 -1.09576702e+00
-8.54313448e-02 -5.79193354e-01 -1.83620341e-02 -1.88554466e+00
-9.41979706e-01 -8.38442206e-01 2.96389312e-01 4.92573619e-01
4.96794224e-01 1.14033408e-01 1.02009155e-01 -1.03981145e-01
-1.94564864e-01 6.85851336e-01 1.35189521e+00 -1.23008631e-01
-3.90826076e-01 7.06617951e-01 -7.20600069e-01 3.48385096e-01
9.66747403e-01 -1.36604905e-01 -1.02073324e+00 -4.86886591e-01
4.17761326e-01 6.31115913e-01 1.13047510e-01 -1.45951450e+00
2.81114280e-01 -4.20482010e-01 3.95409375e-01 -1.95569590e-01
5.55089951e-01 -1.33350754e+00 2.69729435e-01 8.12372983e-01
4.82640624e-01 4.75336015e-01 1.96858630e-01 9.06174004e-01
1.97098538e-01 4.35078703e-02 8.75312924e-01 -1.18907079e-01
-2.49847412e-01 7.62822211e-01 -1.20222819e+00 -4.64183539e-01
1.69814205e+00 -3.00999999e-01 -4.62394617e-02 -8.77372384e-01
-1.64977625e-01 9.78829145e-01 6.88901961e-01 -1.86250489e-02
3.83523196e-01 -1.10862076e+00 -3.89179349e-01 -9.62133929e-02
-3.74786764e-01 6.99329749e-02 2.74833273e-02 7.95170069e-01
-7.53459096e-01 4.99845862e-01 -3.62952948e-01 -3.17359746e-01
-8.50021482e-01 2.05256447e-01 9.22922373e-01 1.73708230e-01
-2.08539262e-01 7.06440687e-01 -4.56300616e-01 -1.97642133e-01
-2.33259022e-01 -2.52005666e-01 -5.74467704e-02 8.55319127e-02
-9.63821113e-02 5.18288970e-01 6.22367375e-02 -4.22577560e-01
-2.61740029e-01 6.99026287e-01 5.90305567e-01 -4.23761867e-02
1.41216004e+00 -4.20347780e-01 4.26162854e-02 -5.00326276e-01
4.07521516e-01 -9.13372487e-02 -1.41530001e+00 5.71002841e-01
-6.82102442e-01 -1.84783265e-01 8.26742828e-01 -5.46259940e-01
-1.31996202e+00 3.23846251e-01 2.57548153e-01 9.53574598e-01
1.51774251e+00 -5.93557358e-01 9.46996152e-01 7.05277324e-01
1.00460327e+00 -9.82082307e-01 -2.72973478e-01 4.15173680e-01
4.80409324e-01 -6.84290171e-01 1.38404518e-01 -1.25439540e-01
-3.68493557e-01 1.33991218e+00 4.46307868e-01 -1.53248593e-01
5.22244394e-01 -2.43412293e-02 -5.94062865e-01 -1.42018482e-01
-9.04152632e-01 -5.50584316e-01 -3.54290783e-01 7.01274157e-01
-8.78849030e-02 3.82756621e-01 4.20156941e-02 6.61135884e-03
-1.02586202e-01 -3.57101142e-01 8.62325549e-01 1.07991350e+00
-9.43139374e-01 -1.12849081e+00 -5.98599873e-02 6.93729877e-01
-1.31883519e-02 2.91085631e-01 -8.80418867e-02 7.97207832e-01
3.30920547e-01 1.25178039e+00 1.04413599e-01 -3.87014151e-01
4.59196478e-01 -5.64975619e-01 4.55118299e-01 -4.70901817e-01
-2.86016911e-01 -1.97366495e-02 3.62996936e-01 -3.33603263e-01
-4.03677911e-01 -6.01893187e-01 -1.29343534e+00 -5.19695580e-01
-3.81983429e-01 4.71763521e-01 6.76687956e-01 8.42687845e-01
6.54433012e-01 8.41261983e-01 1.25301945e+00 -1.11035120e+00
-1.08617656e-01 -3.59212190e-01 -7.00051844e-01 -1.27778196e+00
7.01901615e-01 -7.53453314e-01 6.33052289e-02 -4.60495859e-01] | [5.8354573249816895, 1.5901068449020386] |
f13fe7c5-1b21-4652-bc1d-d2d3b7c34418 | a-new-approach-for-automatic-segmentation-and | 2101.07195 | null | https://arxiv.org/abs/2101.07195v1 | https://arxiv.org/pdf/2101.07195v1.pdf | A New Approach for Automatic Segmentation and Evaluation of Pigmentation Lesion by using Active Contour Model and Speeded Up Robust Features | Digital image processing techniques have wide applications in different scientific fields including the medicine. By use of image processing algorithms, physicians have been more successful in diagnosis of different diseases and have achieved much better treatment results. In this paper, we propose an automatic method for segmenting the skin lesions and extracting features that are associated to them. At this aim, a combination of Speeded-Up Robust Features (SURF) and Active Contour Model (ACM), is used. In the suggested method, at first region of skin lesion is segmented from the whole skin image, and then some features like the mean, variance, RGB and HSV parameters are extracted from the segmented region. Comparing the segmentation results, by use of Otsu thresholding, our proposed method, shows the superiority of our procedure over the Otsu theresholding method. Segmentation of the skin lesion by the proposed method and Otsu thresholding compared the results with physician's manual method. The proposed method for skin lesion segmentation, which is a combination of SURF and ACM, gives the best result. For empirical evaluation of our method, we have applied it on twenty different skin lesion images. Obtained results confirm the high performance, speed and accuracy of our method. | ['Amirmehdi Farshad', 'Melika Farshad', 'Mehran Yazdi', 'Akram Jamshidzadeh', 'Zahra Karimi', 'Sara Mardanisamani'] | 2021-01-18 | null | null | null | null | ['skin-lesion-segmentation'] | ['medical'] | [ 4.8484787e-01 -3.6266869e-01 -4.9018513e-02 -6.2931977e-02
-4.1519034e-01 -4.4051576e-01 1.8689154e-01 5.6709677e-01
-8.0478036e-01 5.2792621e-01 -1.7864580e-01 2.7568065e-02
-1.8724193e-01 -8.5897106e-01 1.6739692e-01 -8.4099752e-01
1.7649649e-01 1.3364848e-01 7.3955095e-01 1.7230432e-03
7.0934564e-01 7.6871955e-01 -1.3898151e+00 -7.8132220e-02
1.0450528e+00 8.9711303e-01 2.5636229e-01 7.6411891e-01
-5.0118798e-01 2.3678756e-01 -3.8047737e-01 2.4319306e-02
1.5744126e-01 -5.0456595e-01 -7.4644053e-01 4.4913062e-01
-1.3680388e-01 -1.3348654e-01 2.5265658e-01 1.0871314e+00
4.5778385e-01 2.4901971e-01 1.0836575e+00 -7.2197551e-01
-8.9914724e-02 -1.0816355e-01 -1.2049574e+00 2.3202512e-01
3.7147176e-01 -1.4990200e-01 2.0463546e-01 -5.7782441e-01
6.6051406e-01 9.5380646e-01 6.5517193e-01 2.5483164e-01
-9.1256219e-01 -3.4524238e-01 -4.6257371e-01 3.8676786e-01
-1.6036484e+00 -2.5161242e-02 6.4667189e-01 -4.9520347e-01
5.1651704e-01 4.4746879e-01 6.7204207e-01 8.5992284e-02
4.3877485e-01 6.7606884e-01 1.4390223e+00 -9.5121723e-01
2.5011542e-01 3.4954074e-01 5.3937411e-01 8.0859125e-01
3.4473842e-01 -2.8606364e-01 2.1946999e-01 -2.0867123e-01
7.8452367e-01 1.7769367e-01 -1.3294384e-01 -5.7917990e-02
-7.3615998e-01 6.2596369e-01 2.7450693e-01 8.3954823e-01
-7.6099151e-01 -1.2142825e-01 3.7609437e-01 -4.3615085e-01
3.0257004e-01 -1.2289291e-01 9.7532503e-02 1.4337167e-01
-1.2710962e+00 -1.6618294e-01 5.8184826e-01 3.1716681e-01
5.3106517e-01 -4.0874791e-01 -2.0035951e-01 7.3830682e-01
4.1056260e-01 4.9606434e-01 6.0307735e-01 -5.0791836e-01
-2.1150400e-01 8.2660788e-01 1.5742768e-02 -1.2005231e+00
-2.8217563e-01 4.1778815e-01 -7.2852522e-01 6.1817229e-01
3.5200381e-01 -4.3512564e-02 -1.4322369e+00 7.2387952e-01
5.8364636e-01 -5.5720333e-02 -6.9467485e-02 8.0875516e-01
7.3333997e-01 7.6135117e-01 4.3155828e-01 -3.7453675e-01
1.4778937e+00 -5.9269232e-01 -1.0079454e+00 4.7829837e-01
1.9216767e-01 -1.1642698e+00 6.3139105e-01 7.1803981e-01
-7.8708369e-01 -4.8911777e-01 -8.9305222e-01 3.1230894e-01
-3.9660722e-01 2.8031543e-01 4.9855390e-01 7.4545175e-01
-9.5182687e-01 5.2877134e-01 -1.0619222e+00 -1.0858666e+00
1.6938788e-01 4.0717784e-01 -3.7981293e-01 1.3770774e-01
-6.5378582e-01 7.8742790e-01 5.5976182e-01 1.4050072e-01
-6.7631081e-02 1.8955401e-01 -6.2483251e-01 -3.3215892e-01
2.3538619e-01 -1.5964381e-01 8.0935889e-01 -1.1839492e+00
-1.4053814e+00 8.2411087e-01 -2.5655186e-01 -8.1226856e-02
3.9390016e-01 1.4718783e-01 -2.8459454e-01 6.6364229e-01
-1.8895249e-01 2.1155511e-01 6.5595412e-01 -1.0236340e+00
-6.5191770e-01 -5.7458597e-01 -4.4190368e-01 2.0196620e-01
-2.1570191e-01 2.0030321e-01 -5.4734522e-01 -2.9956341e-01
3.4662107e-01 -6.6572422e-01 -3.9343640e-01 3.3936460e-02
-4.2298561e-01 -2.5510883e-01 1.0020045e+00 -9.9994391e-01
1.2241237e+00 -2.2429209e+00 -1.4753938e-01 8.7783307e-01
-4.1363150e-02 6.6338831e-01 3.6915463e-01 4.6330848e-01
2.1619362e-01 2.0307340e-01 -4.5582354e-01 1.2657771e-01
-5.0053889e-01 1.4203691e-01 4.2810729e-01 5.4890567e-01
-8.7201707e-02 1.4057592e-01 -5.6535834e-01 -1.2919787e+00
5.0167787e-01 7.4074858e-01 2.2638202e-01 2.8647924e-02
3.2369837e-01 1.8551674e-01 -6.8815303e-01 7.5328696e-01
8.0473673e-01 1.6858289e-01 4.9692042e-02 -2.3326737e-01
-3.3980781e-01 -7.0626283e-01 -1.4171100e+00 1.0944799e+00
-1.1161261e-01 5.1110214e-01 7.4878328e-02 -7.4005055e-01
1.1621788e+00 5.7818097e-01 6.6109461e-01 -4.9187577e-01
4.9971479e-01 3.2966447e-01 -1.7918427e-01 -1.0832669e+00
2.7169093e-01 6.6261888e-02 7.0666307e-01 2.8901967e-01
-2.4323775e-01 -1.3148281e-01 5.5914557e-01 -3.4452628e-02
6.6861093e-01 1.8816918e-01 7.7874368e-01 -1.8026100e-01
9.4329566e-01 2.8311151e-01 2.1107201e-01 1.7994398e-01
-4.9192834e-01 3.3812702e-01 2.5780794e-01 -2.2060642e-02
-7.8589594e-01 -7.8539383e-01 -4.6299911e-01 4.6081999e-01
3.5733140e-01 1.8671963e-01 -1.1320421e+00 -5.2589643e-01
-1.7293361e-03 2.5101161e-01 -6.1155236e-01 4.6322587e-01
-4.9579185e-01 -7.1654105e-01 2.8172055e-01 2.1205820e-01
8.0331862e-01 -1.1582252e+00 -8.9404285e-01 2.5414926e-01
2.6492810e-01 -5.4140192e-01 -5.8316302e-02 -4.6444246e-01
-1.0389522e+00 -1.1035165e+00 -1.1541348e+00 -8.6872888e-01
9.4587612e-01 3.2740808e-01 2.3863921e-01 5.2934778e-01
-9.9644518e-01 2.9034689e-01 -6.5249455e-01 -3.5503209e-01
-2.7473468e-01 -2.6485568e-01 -3.3710626e-01 1.3129246e-01
2.6283666e-01 -1.7042956e-01 -6.2593156e-01 9.9031560e-02
-1.0982552e+00 -3.6158025e-01 8.5677332e-01 3.4889317e-01
7.8783989e-01 4.1917786e-01 1.5012236e-01 -1.0939704e+00
6.5455282e-01 -1.1620725e-01 -5.7062095e-01 3.6410040e-01
-6.0708439e-01 -2.6645812e-01 2.7088419e-01 -1.6298835e-01
-1.0874727e+00 2.9098991e-01 -1.7542475e-01 8.4955476e-02
-4.9157590e-01 2.8297606e-01 2.0041619e-01 -4.9495649e-01
7.0678163e-01 1.8838753e-01 3.3265105e-01 -5.7862312e-01
9.1539532e-02 1.1549584e+00 3.7919191e-01 4.6263952e-02
3.5184136e-01 5.1858157e-01 1.3868767e-01 -1.1875215e+00
-1.7331222e-01 -9.4979155e-01 -7.6426268e-01 -4.9930879e-01
1.3084043e+00 -3.0795552e-02 -5.1868844e-01 7.8068453e-01
-1.0929613e+00 2.3899850e-01 2.3361692e-01 6.5455604e-01
-3.3274192e-01 8.7845576e-01 -5.5323875e-01 -1.3951575e+00
-7.1387011e-01 -1.0126902e+00 5.7728457e-01 8.4275162e-01
-9.4704419e-02 -1.2053726e+00 2.1034023e-01 2.0573916e-01
3.5670245e-01 7.1170050e-01 7.5130469e-01 -4.2747179e-01
-3.0918494e-02 -5.4036796e-01 -2.5253147e-01 3.0541041e-01
5.1578426e-01 6.4067066e-01 -7.4016768e-01 -6.9744617e-02
-7.5958677e-02 1.5103339e-01 8.9121288e-01 5.6209534e-01
1.0933422e+00 1.6020576e-02 -6.6778105e-01 2.6093245e-01
2.1044478e+00 7.7197194e-01 8.8334239e-01 2.5391176e-01
3.0749863e-01 6.6902834e-01 9.9455726e-01 2.9621238e-01
-1.9325878e-01 3.3230677e-01 1.8980272e-01 -6.9326156e-01
-2.1381058e-01 3.2596448e-01 3.8839445e-02 6.6760296e-01
-6.7792553e-01 -1.7029601e-01 -1.0578201e+00 4.5839137e-01
-1.5318227e+00 -8.7176478e-01 -5.1385289e-01 2.3188305e+00
6.7843968e-01 9.4834842e-02 3.2996371e-01 6.6763180e-01
1.1179237e+00 -3.5065100e-01 -2.1032356e-01 -6.7487353e-01
2.6901731e-01 5.2547514e-01 7.2432274e-01 6.8740237e-01
-1.1118375e+00 6.0993809e-01 6.0761304e+00 9.4125289e-01
-1.5152009e+00 7.4062664e-03 3.1347311e-01 6.0471827e-01
1.9199225e-01 -1.4362612e-01 -5.0012702e-01 3.9038765e-01
2.7432749e-01 -1.8407103e-01 1.0181755e-01 4.7600573e-01
3.4092623e-01 -1.0239743e+00 -3.2430080e-01 7.5787544e-01
1.3491488e-01 -8.5193890e-01 -1.5600964e-01 8.6668087e-03
5.3740948e-01 -6.4031565e-01 -1.7558068e-01 -5.1791084e-01
-3.1519178e-01 -8.0817080e-01 2.6594183e-01 8.3585542e-01
5.0974667e-01 -7.6580340e-01 1.1716969e+00 1.3993041e-01
-1.1940038e+00 2.4108003e-01 -1.0927586e-01 2.7763435e-01
7.1424343e-02 5.1283419e-01 -1.1398456e+00 5.7691789e-01
3.3389959e-01 2.6708320e-01 -5.9498078e-01 1.7457064e+00
-9.5751956e-02 6.1766738e-01 -3.4393418e-01 -4.4463170e-01
2.5185463e-01 -4.8558655e-01 4.3960407e-01 1.4270431e+00
2.0414315e-01 3.3149233e-01 -4.9783338e-02 5.5625117e-01
8.1411451e-01 1.0485280e+00 -4.9568132e-01 -2.0680730e-01
2.9663673e-01 1.4246918e+00 -1.5442438e+00 -5.0995809e-01
-1.0545810e-02 9.1759598e-01 -3.7761050e-01 1.8210302e-01
-4.9842733e-01 -1.0343828e+00 -2.9325515e-01 2.8306201e-01
-4.1754194e-02 -2.7508345e-01 -3.6367616e-01 -4.4037244e-01
-2.6762775e-01 -3.0470011e-01 5.2967215e-01 -5.7120556e-01
-7.7983558e-01 5.5254662e-01 7.3795900e-02 -1.0240366e+00
9.6350618e-02 -6.4482129e-01 -8.9907068e-01 9.6155226e-01
-1.0610492e+00 -1.0219655e+00 -4.3976128e-01 6.3414556e-01
5.5030131e-01 2.6300466e-01 6.2989551e-01 9.5387764e-02
-6.5544319e-01 5.6934379e-02 2.1781461e-01 1.5193331e-01
5.4595870e-01 -1.1612296e+00 -4.1954917e-01 8.9275670e-01
-2.7763605e-01 4.2104554e-01 5.9824896e-01 -8.9212143e-01
-8.5034394e-01 -3.6619553e-01 6.0631555e-01 4.2620963e-01
2.1676135e-01 2.9113483e-01 -7.9280424e-01 1.7171526e-02
3.9298001e-01 -3.2355297e-01 6.2409091e-01 -3.9466184e-01
3.4787807e-01 -1.5261240e-02 -1.7236769e+00 2.3115599e-01
8.5285142e-02 4.1206229e-02 -4.6885890e-01 3.2351407e-01
-1.3731524e-01 -3.0700708e-02 -7.7977306e-01 2.5988927e-01
7.3297942e-01 -1.0704219e+00 6.2337941e-01 -4.5634419e-02
-2.4995379e-02 -3.8301766e-01 2.0217250e-01 -9.6535450e-01
-2.2215359e-01 -6.3873440e-02 6.6345608e-01 1.2255498e+00
2.9820213e-01 -5.8171034e-01 6.8652636e-01 3.8942313e-01
3.0843690e-01 -8.2753897e-01 -5.7935274e-01 -3.8993275e-01
-4.5003867e-01 2.5393018e-01 -1.5449609e-02 7.3743260e-01
-1.3448620e-03 -2.4825622e-01 1.6562706e-01 -6.6131011e-02
8.6722356e-01 -1.8898827e-01 3.4224007e-01 -1.2678889e+00
9.1434687e-02 -3.1546220e-01 -7.6160175e-01 1.0391935e-02
-5.5817080e-01 -4.1546315e-01 2.1681446e-03 -2.1008952e+00
1.9923732e-01 -4.2042497e-01 -3.7628382e-01 4.8049611e-01
-2.9698500e-01 4.7833177e-01 1.3774376e-01 1.8801405e-01
-6.2798195e-02 -3.4131712e-01 1.4352225e+00 1.3610435e-01
-4.6089631e-01 1.0813748e-02 -1.2052982e-01 9.1071820e-01
8.3302605e-01 -2.8897676e-01 -1.4485304e-01 3.0184723e-02
-3.6624166e-01 1.1306849e-02 -4.9496248e-02 -1.0829079e+00
2.3560122e-01 -3.4156385e-01 5.6150037e-01 -5.1542306e-01
1.9369315e-01 -8.5499853e-01 1.0715620e-01 9.0229404e-01
-3.6665272e-02 -2.5368226e-01 1.8926147e-02 2.8846729e-01
-1.4677466e-01 -7.5868255e-01 1.1055393e+00 -2.3500657e-01
-9.9388760e-01 -8.8645294e-02 -6.5033972e-01 -6.2822288e-01
1.5318624e+00 -7.1401811e-01 7.3144799e-03 -3.4724057e-02
-9.8633480e-01 -2.9080921e-01 4.9181718e-01 -2.7202979e-01
6.1525935e-01 -9.9623865e-01 -4.6600404e-01 7.9680972e-02
-2.4488862e-01 -2.1040951e-01 2.6837906e-01 1.3280929e+00
-1.4039189e+00 7.7490613e-02 -6.5352547e-01 -4.8747897e-01
-1.8907366e+00 4.0038431e-01 9.8837860e-02 5.1540289e-02
-3.9945775e-01 4.4969437e-01 -4.3570948e-01 3.6562246e-01
1.5011101e-02 -2.7993533e-01 -8.6937541e-01 1.6179335e-01
4.5059046e-01 9.1797632e-01 -1.8205930e-02 -8.3011955e-01
-3.3969846e-01 1.2835628e+00 4.6337891e-02 -3.4538111e-01
8.0562395e-01 1.2084429e-02 -4.3489954e-01 3.0188024e-01
8.9693010e-01 3.9913496e-01 -5.4135817e-01 2.2499451e-01
4.6612374e-02 -6.8422014e-01 3.3807990e-01 -8.4834141e-01
-8.8689154e-01 8.6265236e-01 1.1797029e+00 5.8790183e-01
1.4415543e+00 -5.1068783e-01 8.3859140e-01 8.2695540e-03
3.2064977e-01 -1.2687385e+00 -3.7084293e-01 -1.5041868e-01
4.6514440e-01 -1.0612261e+00 2.3795791e-01 -8.7777770e-01
-6.9521695e-01 1.5083600e+00 3.1014162e-01 -3.4476137e-01
7.0773745e-01 2.3669900e-01 3.5688883e-01 3.3635151e-02
-9.3285210e-02 -5.6838644e-01 1.8100108e-01 5.3338158e-01
5.0072110e-01 1.8429351e-01 -1.3669884e+00 3.0021954e-02
4.3323657e-01 3.8102305e-01 4.6922722e-01 1.1947733e+00
-1.1668706e+00 -1.1180909e+00 -9.7104663e-01 4.2310339e-01
-7.3252100e-01 4.1671473e-01 -4.8734570e-01 1.0338929e+00
3.9667678e-01 9.8008007e-01 -1.7341159e-02 -1.4971654e-01
1.4632660e-01 -1.3771576e-01 8.2595164e-01 -2.2566150e-01
-5.2404320e-01 4.0902892e-01 -1.2033639e-01 -1.5815270e-01
-5.6968856e-01 -5.9860557e-01 -1.6862917e+00 -1.4712351e-02
-5.8569121e-01 3.8233733e-01 1.3385739e+00 8.0445236e-01
-2.9620808e-01 5.9867632e-02 6.8230510e-01 -5.8093959e-01
-1.9780096e-01 -8.4806532e-01 -9.2871296e-01 4.8226771e-01
-1.7978789e-02 -6.4262313e-01 -2.2280006e-01 1.8719895e-01] | [15.191524505615234, -2.927056074142456] |
2449d828-f790-4ea7-87b0-80d8573fba26 | fast-video-object-segmentation-with-spatio | 1903.12161 | null | http://arxiv.org/abs/1903.12161v1 | http://arxiv.org/pdf/1903.12161v1.pdf | Fast video object segmentation with Spatio-Temporal GANs | Learning descriptive spatio-temporal object models from data is paramount for
the task of semi-supervised video object segmentation. Most existing approaches
mainly rely on models that estimate the segmentation mask based on a reference
mask at the first frame (aided sometimes by optical flow or the previous mask).
These models, however, are prone to fail under rapid appearance changes or
occlusions due to their limitations in modelling the temporal component. On the
other hand, very recently, other approaches learned long-term features using a
convolutional LSTM to leverage the information from all previous video frames.
Even though these models achieve better temporal representations, they still
have to be fine-tuned for every new video sequence. In this paper, we present
an intermediate solution and devise a novel GAN architecture, FaSTGAN, to learn
spatio-temporal object models over finite temporal windows. To achieve this, we
concentrate all the heavy computational load to the training phase with two
critics that enforce spatial and temporal mask consistency over the last K
frames. Then at test time, we only use a relatively light regressor, which
reduces the inference time considerably. As a result, our approach combines a
high resiliency to sudden geometric and photometric object changes with
efficiency at test time (no need for fine-tuning nor post-processing). We
demonstrate that the accuracy of our method is on par with state-of-the-art
techniques on the challenging YouTube-VOS and DAVIS datasets, while running at
32 fps, about 4x faster than the closest competitor. | ['Luc van Gool', 'Francesc Moreno-Noguer', 'Albert Pumarola', 'Sergi Caelles', 'Alberto Sanfeliu'] | 2019-03-28 | null | null | null | null | ['one-shot-visual-object-segmentation'] | ['computer-vision'] | [ 7.22597912e-02 -9.17933583e-02 -2.07795948e-01 -2.34316424e-01
-5.86125195e-01 -5.86704552e-01 5.59635341e-01 -1.67901963e-01
-5.52464008e-01 6.79419279e-01 -4.57750618e-01 1.77911520e-02
1.08224064e-01 -5.20633399e-01 -9.59603965e-01 -6.10287607e-01
-1.41554549e-02 3.55874151e-01 8.21429729e-01 9.96016487e-02
9.21353921e-02 4.79963303e-01 -1.60578692e+00 1.69019669e-01
7.41586864e-01 1.27121377e+00 2.66810805e-01 8.05847466e-01
-1.26624882e-01 9.21630859e-01 -3.71773869e-01 -3.64999235e-01
5.27511656e-01 -4.06445056e-01 -8.24905574e-01 4.48074579e-01
5.73767364e-01 -6.37202442e-01 -3.62047136e-01 7.70074666e-01
1.44185483e-01 2.90132016e-01 3.41112137e-01 -1.07211888e+00
-1.31017685e-01 1.68069512e-01 -6.21713221e-01 3.33591610e-01
1.36458755e-01 4.69765812e-01 7.43025661e-01 -6.33041561e-01
7.72974312e-01 8.75192523e-01 6.85017526e-01 6.96157217e-01
-1.34207177e+00 -3.58067185e-01 5.34220636e-01 3.04557234e-01
-1.16644692e+00 -6.10164046e-01 7.41837919e-01 -4.97933537e-01
7.34652817e-01 1.01191439e-01 8.38016331e-01 9.79996502e-01
-8.34354088e-02 8.58818650e-01 9.31112111e-01 -2.44389579e-01
2.13788837e-01 -5.45758568e-02 -2.34930471e-01 8.16922009e-01
-2.53332794e-01 3.37717459e-02 -2.69173831e-01 2.39513829e-01
9.86022532e-01 -3.28889005e-02 -3.78517270e-01 -5.42070031e-01
-9.84010518e-01 5.26612759e-01 4.08438683e-01 4.34303284e-01
-3.99947941e-01 3.90903622e-01 2.83517092e-01 2.40199089e-01
6.15422070e-01 8.45013931e-02 -6.31886125e-01 -2.45925978e-01
-1.47890711e+00 1.73418745e-01 6.05435729e-01 7.28023648e-01
8.74350369e-01 1.75339207e-01 -9.30160731e-02 5.20178854e-01
2.21748397e-01 6.12644888e-02 3.59989971e-01 -1.17542446e+00
3.57869714e-01 3.01961720e-01 1.94665432e-01 -7.96100616e-01
-1.97632834e-01 -3.76157373e-01 -6.64446175e-01 4.87332135e-01
8.31394196e-01 2.11769007e-02 -1.28737843e+00 1.65423512e+00
5.63470840e-01 7.03570664e-01 -2.15332836e-01 9.76776719e-01
3.90855402e-01 6.72585845e-01 6.46147802e-02 -2.96186388e-01
9.99864161e-01 -1.18268740e+00 -5.48362911e-01 -2.57623047e-01
3.60571206e-01 -6.17713034e-01 9.19808447e-01 5.39009571e-01
-1.26501477e+00 -7.61955798e-01 -9.57988560e-01 -1.08251132e-01
-1.73481181e-01 8.51349235e-02 4.78356451e-01 5.52099943e-01
-1.08347642e+00 9.30520296e-01 -1.31760383e+00 -1.68614820e-01
5.57325721e-01 4.56524491e-01 -2.29090393e-01 9.03241523e-03
-8.09369147e-01 7.29723573e-01 3.34132582e-01 2.67558455e-01
-1.01978683e+00 -8.46191227e-01 -7.82313287e-01 1.05975801e-03
6.41651690e-01 -6.44939899e-01 1.14906549e+00 -1.47428155e+00
-1.88567495e+00 6.83109760e-01 -2.29565918e-01 -7.86426842e-01
1.06615448e+00 -3.47734988e-01 1.15401924e-01 3.19727242e-01
-1.85633063e-01 8.37549388e-01 1.19881201e+00 -1.17500758e+00
-6.03740513e-01 -4.67293449e-02 3.70248586e-01 -4.49409187e-02
-1.51110142e-01 8.14208481e-03 -9.84219015e-01 -7.27270126e-01
-1.39611050e-01 -9.68464077e-01 -3.56100470e-01 3.81354779e-01
-1.74812764e-01 -7.42039755e-02 9.50791895e-01 -6.82144344e-01
9.61546242e-01 -2.15381098e+00 1.98293731e-01 -7.99730048e-02
3.28048766e-02 6.41252995e-01 -9.75228995e-02 -1.40937358e-01
9.15804058e-02 6.53871289e-03 -3.72423649e-01 -8.51859629e-01
-3.27418417e-01 2.50901133e-01 -2.74202645e-01 6.29024982e-01
3.90569299e-01 9.84582186e-01 -7.99519718e-01 -5.79725981e-01
7.02450216e-01 6.71862364e-01 -6.59282804e-01 2.82117248e-01
-6.29934311e-01 7.21982479e-01 -2.39863142e-01 3.22642416e-01
5.36979795e-01 -3.37291092e-01 -2.06672046e-02 -1.54237747e-01
-1.97392270e-01 1.55426502e-01 -1.25934660e+00 2.13671279e+00
-4.22562301e-01 6.78031206e-01 1.05460741e-01 -1.19242311e+00
5.84149837e-01 3.06773841e-01 7.84983575e-01 -5.61175525e-01
1.42990664e-01 1.93850502e-01 -1.87639520e-01 -4.75093842e-01
1.44965872e-01 1.37206838e-01 5.31902552e-01 2.14302063e-01
1.42425522e-01 -5.80825694e-02 3.89629394e-01 -8.70715901e-02
8.62060845e-01 8.17576826e-01 -3.90271563e-03 1.15889926e-02
7.49893010e-01 -1.45424441e-01 7.24659443e-01 5.10579288e-01
-1.38092443e-01 9.76838470e-01 4.94790226e-01 -7.38092005e-01
-1.03121650e+00 -7.50777721e-01 7.14987889e-02 7.49948680e-01
2.33734578e-01 -3.56044590e-01 -8.98080409e-01 -8.29483509e-01
-2.65251160e-01 3.43149453e-01 -6.94684565e-01 1.98458269e-01
-9.30458069e-01 -4.83421415e-01 2.63214499e-01 5.86628675e-01
6.36725485e-01 -1.02857363e+00 -1.05043077e+00 4.57418650e-01
-3.31371203e-02 -1.45470667e+00 -4.59156781e-01 -9.97368619e-03
-9.97331500e-01 -1.04437876e+00 -9.46483016e-01 -4.09316301e-01
4.96058017e-01 -3.16305608e-02 1.00306141e+00 2.70974517e-01
-2.74303436e-01 2.52673388e-01 -2.69975990e-01 1.75484717e-02
-2.49595493e-01 4.67219613e-02 -2.70092070e-01 3.17630023e-01
-1.75004527e-01 -5.85625172e-01 -7.77969480e-01 4.28448886e-01
-9.87211406e-01 2.72704571e-01 2.71607071e-01 7.06841409e-01
7.17916012e-01 -8.36337134e-02 1.60882592e-01 -6.11240566e-01
-3.84261876e-01 -1.92970261e-01 -9.61051524e-01 1.64144725e-01
-2.70631224e-01 9.26758919e-04 7.71616161e-01 -5.86973071e-01
-1.03552866e+00 2.93745369e-01 -1.50756598e-01 -9.27133262e-01
-1.03713110e-01 1.48502499e-01 9.35468972e-02 -1.86999097e-01
2.87360430e-01 1.42867312e-01 7.04604387e-02 -5.18387258e-01
2.45157242e-01 1.45719782e-01 5.71656287e-01 -4.38428879e-01
8.22044075e-01 7.63671696e-01 -7.32486844e-02 -6.53184593e-01
-9.60800469e-01 -3.08432817e-01 -8.77007961e-01 -3.34967047e-01
9.84401107e-01 -8.87859166e-01 -6.15603507e-01 5.37037075e-01
-1.19502413e+00 -8.37968051e-01 -4.87914830e-01 3.93325716e-01
-7.69738853e-01 4.30366009e-01 -5.91193736e-01 -7.49792635e-01
-1.18080713e-01 -1.10605717e+00 1.03439403e+00 1.36346936e-01
6.32821117e-03 -9.90422070e-01 -1.00402378e-01 2.84859568e-01
4.46850330e-01 4.47765142e-01 3.89879316e-01 -1.68076485e-01
-1.01395512e+00 8.01982135e-02 -1.73850968e-01 5.40776610e-01
7.56700635e-02 2.73693979e-01 -9.56789434e-01 -2.85132676e-01
1.22092813e-01 -1.75206736e-01 1.05870676e+00 5.74062467e-01
1.34946740e+00 -1.72638699e-01 -1.06458366e-01 9.34621334e-01
1.49316204e+00 1.05106600e-01 6.95492804e-01 3.76817703e-01
9.26832914e-01 5.48683047e-01 6.72588766e-01 3.25323015e-01
1.95469201e-01 8.76383007e-01 5.68997502e-01 -1.53951794e-01
-3.72941673e-01 -3.71947922e-02 3.49726528e-01 2.92410374e-01
-3.13541055e-01 -2.96635687e-01 -7.26550281e-01 6.60853624e-01
-1.94152474e+00 -9.02173877e-01 7.00280890e-02 2.26140475e+00
7.32988536e-01 2.57582366e-01 2.90827185e-01 2.14938313e-01
4.94830877e-01 3.31064284e-01 -6.83160663e-01 -5.75893819e-02
5.85302152e-03 2.13191792e-01 3.74959350e-01 5.98731220e-01
-1.26720667e+00 1.09533405e+00 5.44899607e+00 6.63637936e-01
-1.39248097e+00 1.20249361e-01 9.39826667e-01 -3.34933072e-01
4.62984592e-02 6.04007132e-02 -5.34411430e-01 5.79636514e-01
7.99698889e-01 2.75547445e-01 5.46388865e-01 7.02865422e-01
2.61235505e-01 -2.68681616e-01 -1.18284321e+00 9.71921146e-01
-1.66109167e-02 -1.34702694e+00 -2.97627062e-01 -9.84646082e-02
8.28080595e-01 1.57915071e-01 -6.02236874e-02 3.26506980e-02
1.28932735e-02 -8.88597965e-01 9.97451365e-01 5.06693661e-01
8.01086545e-01 -5.46076834e-01 5.30209363e-01 3.07295084e-01
-1.22928441e+00 1.49516808e-02 -1.49053574e-01 -1.74192432e-02
4.25567865e-01 4.92812514e-01 -3.62947315e-01 5.18387139e-01
7.58333623e-01 8.44733775e-01 -5.10540485e-01 1.04909301e+00
-1.09627672e-01 5.08433640e-01 -5.16192257e-01 4.70980942e-01
3.88453037e-01 -1.66636169e-01 4.09260452e-01 1.05054474e+00
3.15408409e-01 2.43989863e-02 2.50284761e-01 8.43319535e-01
1.13269180e-01 -8.85095000e-02 -2.30202019e-01 1.76143482e-01
-6.22074977e-02 1.14838493e+00 -9.47375774e-01 -4.76186663e-01
-5.29769421e-01 1.13792992e+00 2.93383300e-01 4.03619468e-01
-1.11031628e+00 2.11647987e-01 6.17426932e-01 3.03055376e-01
7.13815212e-01 -4.56164718e-01 -1.85857341e-01 -1.29846859e+00
3.18809628e-01 -6.27421916e-01 2.33664036e-01 -5.22328854e-01
-9.15554643e-01 5.92426598e-01 -1.20129168e-01 -1.21755803e+00
-4.64733928e-01 -3.94620657e-01 -3.68200570e-01 5.56272030e-01
-1.76519299e+00 -1.04831851e+00 -2.11060822e-01 6.47272766e-01
8.27524006e-01 2.72924006e-01 4.27224070e-01 4.47257757e-01
-5.60801923e-01 3.16744626e-01 -1.22349486e-01 1.92989647e-01
5.79908252e-01 -1.18113387e+00 4.57674325e-01 1.06169903e+00
3.28415871e-01 2.43998155e-01 6.58177793e-01 -3.98068219e-01
-1.12150776e+00 -1.15951061e+00 5.89645505e-01 -3.55004787e-01
4.74759102e-01 -3.35513979e-01 -1.12154603e+00 6.93720996e-01
1.52901202e-01 6.63336694e-01 4.46049795e-02 -3.62383783e-01
-1.32885456e-01 -2.33687297e-01 -9.52264488e-01 4.68646705e-01
9.88011599e-01 -2.84780025e-01 -2.49671176e-01 9.02646855e-02
5.96165001e-01 -6.69608116e-01 -6.45967007e-01 4.51662779e-01
6.12958491e-01 -1.32901442e+00 8.71642053e-01 -4.21348482e-01
3.96935552e-01 -5.54519475e-01 1.91989154e-01 -8.54088128e-01
-1.90570636e-03 -9.41397846e-01 -3.97335708e-01 1.21787083e+00
2.23696511e-02 -2.95635611e-01 1.06535482e+00 7.41255462e-01
1.63698196e-02 -8.86872351e-01 -1.06999362e+00 -8.48748744e-01
-2.31923550e-01 -6.46584988e-01 2.80337423e-01 7.10620940e-01
-6.54208839e-01 -1.12411514e-01 -5.40653706e-01 1.27225816e-02
5.16660690e-01 1.46369934e-01 8.95532906e-01 -1.07410932e+00
-5.21390915e-01 -3.82245928e-01 -4.81492996e-01 -1.26445413e+00
2.53784329e-01 -3.88433963e-01 1.35859936e-01 -1.20058095e+00
-1.61554515e-01 -4.56239492e-01 -2.67081320e-01 5.19333899e-01
-1.99148178e-01 5.58542490e-01 3.95806611e-01 1.11556195e-01
-6.63169861e-01 4.60887015e-01 1.23443937e+00 -8.83998871e-02
-4.60307062e-01 3.02250888e-02 -2.33233329e-02 8.93200338e-01
6.21384740e-01 -3.95902306e-01 -3.40617418e-01 -6.28409684e-01
-2.16377620e-02 1.11898206e-01 6.76525891e-01 -1.21928120e+00
1.93049148e-01 -1.44084945e-01 3.75522196e-01 -4.02617663e-01
5.25478423e-01 -8.91110480e-01 2.96120435e-01 4.19705480e-01
-1.44740865e-01 -6.62744120e-02 2.25639269e-01 6.07149243e-01
-2.51404047e-01 -1.62161246e-01 1.01828909e+00 -9.98017192e-02
-7.35065341e-01 6.60712421e-01 -2.11197093e-01 -8.56365412e-02
1.18379569e+00 -3.89936060e-01 1.49466559e-01 -2.14714602e-01
-8.95520687e-01 1.43567622e-01 6.74598396e-01 3.68396670e-01
4.74225998e-01 -9.25625026e-01 -5.07919014e-01 2.21605808e-01
-3.66996408e-01 4.14620548e-01 3.60281885e-01 1.05947876e+00
-6.36217594e-01 1.90917209e-01 -1.69407651e-01 -9.71551180e-01
-1.09439242e+00 6.67095959e-01 4.27010328e-01 -3.65892202e-01
-8.34721148e-01 8.59173238e-01 2.35127300e-01 2.19117761e-01
2.74483323e-01 -4.21165377e-01 -8.99701193e-02 1.04881071e-01
4.26997900e-01 2.34435588e-01 5.24862260e-02 -6.22415662e-01
-2.34314114e-01 9.36567783e-01 -3.78663130e-02 -1.27672702e-01
1.33314693e+00 -2.60209382e-01 1.47344068e-01 5.31082630e-01
1.21490502e+00 -1.39075324e-01 -2.08707976e+00 -2.62864172e-01
-2.08267048e-02 -7.39685237e-01 -1.40493093e-02 -4.28701252e-01
-1.48353112e+00 8.52529228e-01 6.21152163e-01 1.08517222e-01
1.21076286e+00 -1.66651607e-01 8.65069330e-01 4.25967053e-02
2.61250407e-01 -9.79363382e-01 5.83658665e-02 3.47220421e-01
6.08578622e-01 -1.18812895e+00 -4.78819795e-02 -5.10600567e-01
-4.74702597e-01 1.16987646e+00 5.93758345e-01 -3.59643579e-01
4.70734388e-01 2.58901089e-01 9.39803123e-02 1.62991613e-01
-7.38262892e-01 -3.42733294e-01 3.24423283e-01 4.04886812e-01
2.63764888e-01 -5.28318405e-01 -1.25318706e-01 -5.47225475e-02
1.12190284e-01 1.90004155e-01 3.29313278e-01 6.91576123e-01
-5.63769713e-02 -1.10617816e+00 -1.20989703e-01 2.19846621e-01
-7.04551160e-01 1.45376101e-01 1.62892714e-01 8.15644503e-01
2.94030905e-01 7.96693027e-01 1.72084987e-01 3.28972451e-02
9.72239003e-02 -7.08284602e-02 6.57481551e-01 -4.27406609e-01
-4.95522797e-01 3.32803279e-01 -1.55073494e-01 -1.08868194e+00
-9.14781690e-01 -7.72361577e-01 -1.15780687e+00 -2.88688894e-02
-3.48071486e-01 -1.27639145e-01 5.56544065e-01 1.06213176e+00
2.41308764e-01 6.10813081e-01 4.94959682e-01 -1.33425677e+00
-1.98300168e-01 -6.69976532e-01 -1.89793438e-01 4.38840866e-01
5.80021322e-01 -6.43764019e-01 -2.29781702e-01 4.60432440e-01] | [9.062980651855469, -0.17356529831886292] |
e4c9d6c5-33e1-42f4-8ab6-924d7efa4d1b | efficient-and-robust-training-of-dense-object | 2206.12145 | null | https://arxiv.org/abs/2206.12145v1 | https://arxiv.org/pdf/2206.12145v1.pdf | Efficient and Robust Training of Dense Object Nets for Multi-Object Robot Manipulation | We propose a framework for robust and efficient training of Dense Object Nets (DON) with a focus on multi-object robot manipulation scenarios. DON is a popular approach to obtain dense, view-invariant object descriptors, which can be used for a multitude of downstream tasks in robot manipulation, such as, pose estimation, state representation for control, etc.. However, the original work focused training on singulated objects, with limited results on instance-specific, multi-object applications. Additionally, a complex data collection pipeline, including 3D reconstruction and mask annotation of each object, is required for training. In this paper, we further improve the efficacy of DON with a simplified data collection and training regime, that consistently yields higher precision and enables robust tracking of keypoints with less data requirements. In particular, we focus on training with multi-object data instead of singulated objects, combined with a well-chosen augmentation scheme. We additionally propose an alternative loss formulation to the original pixelwise formulation that offers better results and is less sensitive to hyperparameters. Finally, we demonstrate the robustness and accuracy of our proposed framework on a real-world robotic grasping task. | ['Heiko Neumann', 'Markus Spies', 'Andras Gabor Kupcsik', 'David B. Adrian'] | 2022-06-24 | null | null | null | null | ['robotic-grasping', 'robot-manipulation'] | ['robots', 'robots'] | [ 9.15969908e-02 -3.16381186e-01 -1.54821739e-01 -1.82847291e-01
-5.53372145e-01 -4.70067382e-01 4.66750443e-01 2.64746189e-01
-5.25194466e-01 5.63140690e-01 -3.16615641e-01 2.23050207e-01
-5.83241224e-01 -5.12889504e-01 -1.00024486e+00 -8.29851270e-01
5.51351830e-02 5.92267334e-01 3.57048571e-01 -1.66877076e-01
1.40410483e-01 1.06607306e+00 -1.53109729e+00 -1.53911933e-01
5.61011374e-01 1.25232267e+00 8.05356801e-01 2.59663433e-01
2.72274911e-01 5.23861229e-01 -3.62727165e-01 -1.14275843e-01
4.92364079e-01 1.92970604e-01 -5.33262074e-01 1.70158476e-01
5.52317560e-01 -4.96585041e-01 -3.27802688e-01 9.92683470e-01
4.23409373e-01 3.37985337e-01 5.59822619e-01 -1.25072873e+00
-2.99911797e-01 4.20065671e-01 -4.71190214e-01 -2.49595076e-01
-1.56864166e-01 2.55152792e-01 8.04437041e-01 -9.71631467e-01
6.86882496e-01 1.41865075e+00 5.80498457e-01 5.30714214e-01
-1.15301538e+00 -5.69946945e-01 3.12560529e-01 2.24309415e-01
-1.25179648e+00 -3.43763947e-01 8.32026482e-01 -3.90624315e-01
5.71718812e-01 3.47280242e-02 5.66624701e-01 1.07872367e+00
2.79378533e-01 1.00351477e+00 7.98970282e-01 -2.22481474e-01
1.11216441e-01 -3.94412950e-02 7.44701223e-03 6.33881927e-01
3.41232419e-01 -1.00720346e-01 -1.72334164e-01 7.84283318e-03
1.14809167e+00 4.87299323e-01 -2.28965104e-01 -1.16809583e+00
-1.52644539e+00 6.47524357e-01 7.45733380e-01 1.74524218e-01
-6.97050810e-01 4.15862650e-01 2.90048242e-01 5.71874948e-03
1.16324902e-01 4.79567885e-01 -4.97461438e-01 1.31255716e-01
-4.21406537e-01 5.51838279e-01 6.17485464e-01 1.49701059e+00
7.38743544e-01 -6.18592389e-02 -3.87610763e-01 9.16799128e-01
3.07524920e-01 4.67013299e-01 8.25564414e-02 -1.01249516e+00
5.27098119e-01 4.38581675e-01 3.04862916e-01 -9.98518050e-01
-5.08509636e-01 -5.01598656e-01 -8.66242886e-01 4.32298213e-01
3.95421565e-01 1.83777779e-01 -8.86214614e-01 1.69713640e+00
5.20830572e-01 -1.39074609e-01 -1.51128888e-01 1.16186190e+00
4.42761481e-01 3.10138255e-01 -7.50199705e-02 5.74030392e-02
1.25056601e+00 -1.01544023e+00 -6.48440778e-01 -8.39111060e-02
3.06498766e-01 -6.69164896e-01 8.71184528e-01 5.90258420e-01
-1.15495968e+00 -5.11013210e-01 -9.21088159e-01 -2.03519747e-01
-2.53226429e-01 4.53309655e-01 8.28670859e-01 -7.48610348e-02
-5.59196770e-01 7.24629700e-01 -1.12516379e+00 -3.76760513e-01
6.95690334e-01 6.27200842e-01 -5.36961854e-01 -3.05966377e-01
-4.79105115e-01 1.22389603e+00 6.25990629e-01 4.59558845e-01
-1.06977534e+00 -5.92736006e-01 -8.45989168e-01 2.98914826e-03
7.34859824e-01 -5.07204056e-01 1.12782323e+00 -3.01404983e-01
-1.57967663e+00 4.28485662e-01 2.32126176e-01 -3.95253628e-01
6.29986763e-01 -6.45704150e-01 3.78278583e-01 1.36389419e-01
-6.73179999e-02 8.43590796e-01 9.02372777e-01 -1.50112927e+00
-4.40015495e-01 -5.85946202e-01 4.00940269e-01 1.01767205e-01
-3.87491167e-01 -8.35052207e-02 -6.02844834e-01 -6.50345206e-01
2.55551189e-01 -9.38181162e-01 -3.41483027e-01 5.31750917e-01
-2.36773074e-01 -3.29624534e-01 1.10047221e+00 -4.20467019e-01
4.36053723e-01 -2.09839416e+00 6.94475174e-01 2.72126757e-02
1.39860436e-01 2.52041429e-01 -2.29064390e-01 3.64442676e-01
3.59352022e-01 -4.20123667e-01 -2.20638424e-01 -3.92639101e-01
1.61002666e-01 3.59491765e-01 -2.60077894e-01 7.07597315e-01
5.66880941e-01 9.72736597e-01 -7.71008253e-01 -4.30545896e-01
4.95536834e-01 4.89206553e-01 -6.34843826e-01 2.61052132e-01
-4.56402838e-01 6.51530027e-01 -5.96274257e-01 8.57852638e-01
7.05407083e-01 -8.40009749e-02 -1.98977828e-01 -5.66163242e-01
-2.46661723e-01 -2.78877378e-01 -1.31048238e+00 2.10067081e+00
-6.87204480e-01 2.41593003e-01 6.25297129e-01 -1.19130373e+00
9.81043041e-01 3.65327448e-02 6.81257367e-01 -2.20856667e-01
4.26242739e-01 3.19539309e-01 6.80053607e-02 -6.20017648e-01
4.78619397e-01 1.83523431e-01 2.61620600e-02 -2.48301998e-02
3.07898343e-01 -3.71297717e-01 2.11743250e-01 -2.34420210e-01
9.42239106e-01 4.68848944e-01 1.78132161e-01 -2.15111196e-01
3.80955160e-01 5.84018379e-02 4.32724178e-01 5.41958749e-01
-9.06695649e-02 4.04061586e-01 1.64581999e-01 -2.99389120e-02
-1.12256336e+00 -7.67491400e-01 -3.43309104e-01 7.53424764e-01
6.22006595e-01 -1.18132848e-02 -2.92677552e-01 -4.58122760e-01
4.65904891e-01 2.97524601e-01 -2.37453744e-01 -1.10571548e-01
-9.82363701e-01 -4.37027216e-01 1.60067528e-01 6.99363172e-01
3.60518098e-01 -1.14107466e+00 -8.30919087e-01 3.80694300e-01
1.23381011e-01 -1.34863400e+00 -1.00270294e-01 3.12553048e-01
-1.04048085e+00 -1.15716577e+00 -8.07617664e-01 -8.99689078e-01
7.91042030e-01 3.38047057e-01 5.19485056e-01 -1.09824687e-01
-4.91381347e-01 5.82202315e-01 -4.24992472e-01 -3.66048902e-01
-1.00642920e-01 5.63756824e-02 1.30795166e-01 -6.19280599e-02
-2.33294129e-01 -4.50197846e-01 -5.76377332e-01 4.30538505e-01
-9.23905790e-01 -1.46290541e-01 1.12734246e+00 1.00536382e+00
6.31956279e-01 -2.17324823e-01 6.02548897e-01 -3.52602482e-01
2.55845904e-01 -2.86623538e-01 -8.71887267e-01 1.80786759e-01
-1.47050619e-01 -4.85884398e-02 6.55951023e-01 -7.64713883e-01
-8.50676715e-01 4.39807236e-01 -3.12956572e-02 -9.67263222e-01
-1.66716605e-01 2.42347956e-01 -1.42582908e-01 -4.71224427e-01
3.70867401e-01 -5.62011711e-02 2.86293894e-01 -7.51222193e-01
3.33755612e-01 5.18467605e-01 5.63200772e-01 -7.24428654e-01
7.32432067e-01 5.17117500e-01 2.37091005e-01 -7.14828312e-01
-5.54054797e-01 -6.20521188e-01 -9.66239631e-01 -1.46960884e-01
5.97541869e-01 -7.61809111e-01 -9.27131712e-01 4.73000199e-01
-1.38855672e+00 -3.73385668e-01 -2.55487412e-01 7.63019621e-01
-8.69426250e-01 3.10239911e-01 -5.67249238e-01 -8.14411581e-01
-3.03269416e-01 -1.28677285e+00 1.41976202e+00 -1.17663391e-01
2.61063308e-01 -7.19098508e-01 -3.91965270e-01 1.17341734e-01
3.03457737e-01 4.05950069e-01 7.93907285e-01 -4.22267467e-01
-1.10790205e+00 -3.65104705e-01 -3.20638746e-01 3.98692667e-01
2.19117880e-01 -2.62583524e-01 -6.38087273e-01 -6.64998412e-01
-5.54824434e-02 -6.03854597e-01 7.78759360e-01 2.09262654e-01
1.42109823e+00 -9.94015336e-02 -5.70199013e-01 4.62597430e-01
1.58366871e+00 2.45701242e-02 2.34873980e-01 2.97417849e-01
8.71105552e-01 6.65905952e-01 1.11116481e+00 5.02781749e-01
1.18874349e-01 9.70882952e-01 9.55844522e-01 8.23730007e-02
-3.40972878e-02 1.42306417e-01 1.66033059e-01 5.19113600e-01
-6.87946528e-02 -2.94204503e-01 -6.30798459e-01 6.51156604e-01
-2.11665916e+00 -5.56341410e-01 2.85961330e-02 2.06045127e+00
5.89840889e-01 -4.83119637e-02 -2.62652226e-02 4.68426645e-02
6.09448850e-01 -3.33976224e-02 -7.68339932e-01 3.69635701e-01
1.46606013e-01 1.81635171e-01 5.62303901e-01 1.42341673e-01
-1.16499197e+00 9.64470267e-01 5.29837847e+00 7.19307482e-01
-1.17164087e+00 6.25571758e-02 -2.28693068e-01 -4.31555063e-02
2.39684865e-01 -3.05021167e-01 -9.45165992e-01 1.31472185e-01
1.93255618e-01 1.60929814e-01 3.24300140e-01 1.10382402e+00
9.96604636e-02 -6.55343086e-02 -1.36548924e+00 9.98416364e-01
2.44191997e-02 -1.19158351e+00 -6.11162856e-02 -3.83184925e-02
4.82489437e-01 -7.03429570e-03 -2.37066016e-01 2.58795649e-01
1.91389397e-02 -6.54706180e-01 9.67272580e-01 4.29887593e-01
4.32556212e-01 -5.62276781e-01 7.15302646e-01 5.12285769e-01
-1.09221101e+00 -4.55521166e-01 -5.59327006e-01 1.34891361e-01
2.39989370e-01 3.72818053e-01 -6.82564735e-01 7.75590241e-01
6.96714222e-01 8.07607174e-01 -1.75745845e-01 1.42328739e+00
-2.18674056e-02 -1.38547242e-01 -3.92917842e-01 -1.08907923e-01
1.86048746e-01 -6.18740842e-02 6.73752010e-01 8.63653541e-01
1.59442902e-01 -1.20368889e-02 5.97089648e-01 9.12540674e-01
1.11561995e-02 -1.94032155e-02 -6.37240589e-01 2.21222028e-01
3.67799640e-01 1.56789923e+00 -6.91051245e-01 -5.94139621e-02
-2.63502628e-01 6.66471124e-01 6.09591067e-01 1.54623255e-01
-7.34160960e-01 -4.02517945e-01 5.98507822e-01 -8.32418539e-03
5.93773246e-01 -7.60245323e-01 1.00352123e-01 -1.03609717e+00
4.00455415e-01 -6.77132368e-01 -9.81210470e-02 -5.65540373e-01
-1.25910115e+00 2.81105727e-01 3.54993224e-01 -1.41988254e+00
-1.09732985e-01 -1.08937204e+00 -2.47082397e-01 6.17951632e-01
-1.65494156e+00 -1.24564850e+00 -6.22256458e-01 5.29560745e-01
6.82598531e-01 1.32919833e-01 6.30319059e-01 4.14659053e-01
-4.77656037e-01 2.57978022e-01 5.67417219e-02 2.31024735e-02
5.85622013e-01 -1.02723169e+00 -1.45120561e-01 5.14582157e-01
-2.18920693e-01 6.43971443e-01 4.72187221e-01 -5.05777776e-01
-2.01980495e+00 -1.30245030e+00 -1.43272549e-01 -3.31100583e-01
6.59174144e-01 -4.90877956e-01 -7.92797267e-01 7.40098953e-01
-2.07540378e-01 1.89383000e-01 -1.01104498e-01 -2.12211207e-01
2.20685259e-01 -2.71915942e-01 -1.23522353e+00 3.84282887e-01
1.06015253e+00 -3.02207787e-02 -5.00866175e-01 6.30896270e-01
6.19241297e-01 -8.51618528e-01 -1.14852536e+00 7.96419919e-01
7.11935222e-01 -3.59731555e-01 1.05543804e+00 -4.85008508e-01
2.97151893e-01 -4.17359680e-01 -1.82426646e-01 -1.09751523e+00
-4.04044151e-01 -2.80615479e-01 -3.51286352e-01 1.12614572e+00
-1.14578381e-01 -5.50420046e-01 6.79617107e-01 1.62645146e-01
-5.51151037e-01 -9.53558862e-01 -8.39139760e-01 -1.07074535e+00
-1.79092661e-01 -1.71388730e-01 3.46248776e-01 5.57446122e-01
-3.81966561e-01 -1.17427729e-01 -4.07558382e-01 4.17227268e-01
7.26262927e-01 3.84339988e-01 9.91394699e-01 -1.30502021e+00
-1.22320578e-01 -2.20171571e-01 -6.69320226e-01 -1.36251545e+00
2.20079467e-01 -7.62256444e-01 6.78987384e-01 -1.52749515e+00
7.88882747e-02 -7.74260998e-01 -1.45751968e-01 5.42157114e-01
9.83090401e-02 2.50207603e-01 4.33785111e-01 3.29743564e-01
-4.91397470e-01 9.20510352e-01 1.61997271e+00 -2.31194749e-01
-1.00351535e-01 1.69555396e-01 -1.16515532e-01 5.75002313e-01
6.36059940e-01 -2.84620553e-01 -2.06836089e-01 -6.56791508e-01
-5.00397801e-01 -1.28605999e-02 8.59350502e-01 -1.04212224e+00
2.23165259e-01 -2.23296061e-01 2.55665749e-01 -7.62932181e-01
6.99676335e-01 -1.26317179e+00 -4.93455380e-02 5.62279463e-01
-1.41207218e-01 -6.78522810e-02 2.61128664e-01 7.14800000e-01
-1.32740095e-01 -4.21657413e-01 8.89096260e-01 -2.29489282e-01
-7.68710732e-01 6.22633219e-01 2.82289356e-01 -4.99121726e-01
1.31349158e+00 -1.75548699e-02 -2.68745609e-02 1.46799430e-01
-6.62599742e-01 5.07432699e-01 4.93656516e-01 6.85630679e-01
7.24693418e-01 -1.40085971e+00 -5.00987172e-01 6.21436164e-02
1.52395874e-01 6.61468148e-01 6.82233647e-02 1.09526944e+00
-5.09917617e-01 4.58417833e-01 -4.40670758e-01 -1.05886889e+00
-1.09338450e+00 6.32559478e-01 8.22403580e-02 -3.89243918e-03
-9.57524061e-01 6.37381792e-01 1.94477797e-01 -4.85553235e-01
6.31855607e-01 -7.22974300e-01 -6.61346540e-02 -6.64507225e-02
1.61128193e-01 4.79175746e-01 1.49109125e-01 -4.47120488e-01
-3.81247610e-01 7.20826387e-01 -1.00892961e-01 1.93500891e-01
1.61797941e+00 6.00251853e-02 -1.55502483e-01 3.82339418e-01
9.92763281e-01 -3.39032471e-01 -1.64730310e+00 -2.46635854e-01
-7.87568390e-02 -5.55038452e-01 4.16750200e-02 -4.61939335e-01
-1.03940821e+00 7.91748881e-01 5.14983416e-01 -4.28697094e-02
6.96436167e-01 1.40726164e-01 5.98348260e-01 8.17068577e-01
8.55837762e-01 -9.23425019e-01 2.70619422e-01 4.52615529e-01
1.43148065e+00 -1.33747017e+00 1.86049461e-01 -4.74842608e-01
-2.73530513e-01 1.17395401e+00 8.73132467e-01 -2.95657992e-01
4.66269016e-01 6.40627593e-02 -2.84514964e-01 -1.91367194e-01
-4.54849839e-01 -1.37312606e-01 1.79715216e-01 4.57387924e-01
-1.62317142e-01 -2.84161985e-01 -4.32341173e-02 1.88394055e-01
2.28319764e-01 3.92653979e-02 7.92197958e-02 1.32949185e+00
-5.00287950e-01 -9.19397950e-01 -3.61011267e-01 4.15112436e-01
-1.80465728e-01 2.77924210e-01 2.12831721e-02 1.08983147e+00
1.13721147e-01 4.95549202e-01 -3.30711119e-02 -2.65302267e-02
6.78581595e-01 -2.35537171e-01 1.03779674e+00 -8.07770729e-01
-3.77883255e-01 -3.15442644e-02 -2.97095597e-01 -5.76517344e-01
-6.46148443e-01 -7.66036391e-01 -1.14773571e+00 1.23972736e-01
-7.43595064e-01 -2.42089093e-01 1.00073969e+00 9.89322662e-01
3.79985929e-01 5.58520973e-01 4.97844815e-01 -1.63594139e+00
-9.84115660e-01 -1.09192729e+00 -4.74815995e-01 3.05864900e-01
4.67612267e-01 -1.29842818e+00 -4.42080237e-02 -2.16578826e-01] | [5.899233341217041, -0.9601184725761414] |
800c6013-9e80-4c59-bd43-12af5d542248 | the-chamber-ensemble-generator-limitless-high | 2209.14458 | null | https://arxiv.org/abs/2209.14458v1 | https://arxiv.org/pdf/2209.14458v1.pdf | The Chamber Ensemble Generator: Limitless High-Quality MIR Data via Generative Modeling | Data is the lifeblood of modern machine learning systems, including for those in Music Information Retrieval (MIR). However, MIR has long been mired by small datasets and unreliable labels. In this work, we propose to break this bottleneck using generative modeling. By pipelining a generative model of notes (Coconet trained on Bach Chorales) with a structured synthesis model of chamber ensembles (MIDI-DDSP trained on URMP), we demonstrate a system capable of producing unlimited amounts of realistic chorale music with rich annotations including mixes, stems, MIDI, note-level performance attributes (staccato, vibrato, etc.), and even fine-grained synthesis parameters (pitch, amplitude, etc.). We call this system the Chamber Ensemble Generator (CEG), and use it to generate a large dataset of chorales from four different chamber ensembles (CocoChorales). We demonstrate that data generated using our approach improves state-of-the-art models for music transcription and source separation, and we release both the system and the dataset as an open-source foundation for future work in the MIR community. | ['Jesse Engel', 'Curtis Hawthorne', 'Ian Simon', 'Ethan Manilow', 'Josh Gardner', 'Yusong Wu'] | 2022-09-28 | null | null | null | null | ['music-transcription', 'music-information-retrieval'] | ['music', 'music'] | [ 2.88111329e-01 -1.36014909e-01 3.23769331e-01 -3.37840766e-02
-1.31449437e+00 -1.20160389e+00 7.82639742e-01 -2.21897766e-01
1.83114246e-01 4.84197587e-01 5.24856389e-01 1.04106270e-01
-1.44409969e-01 -4.05868262e-01 -4.34965372e-01 -7.34255910e-01
-4.94993888e-02 7.39286244e-01 -1.55829832e-01 -3.90497804e-01
1.83310270e-01 3.14950906e-02 -1.74622118e+00 7.70464063e-01
4.50312227e-01 7.39819527e-01 2.04805601e-02 1.22971046e+00
6.71941265e-02 8.98480237e-01 -1.00158978e+00 -4.38980639e-01
1.80464536e-01 -9.95728493e-01 -7.52295256e-01 5.88549599e-02
4.69622672e-01 1.35687783e-01 9.72305834e-02 4.69891548e-01
8.33449423e-01 1.45240892e-02 9.69238222e-01 -1.18158770e+00
-3.15922588e-01 1.63559973e+00 -2.65492320e-01 -3.27230006e-01
2.77142912e-01 -3.22120339e-02 1.42872775e+00 -7.55115271e-01
7.45248914e-01 8.85383904e-01 8.52728665e-01 5.57569742e-01
-1.57273912e+00 -6.34324789e-01 -5.11837900e-01 -2.67673768e-02
-1.47728324e+00 -8.13766241e-01 6.74793839e-01 -7.33528197e-01
5.75399697e-01 5.75109780e-01 8.37868929e-01 1.36033750e+00
-2.82310158e-01 7.99694479e-01 8.34956527e-01 -6.11175776e-01
1.76818192e-01 -1.21526651e-01 -2.79275864e-01 3.10504228e-01
-4.80869651e-01 -5.41071333e-02 -8.73487830e-01 -2.35350803e-01
8.13164592e-01 -5.81485987e-01 -2.33439356e-01 -6.89992607e-02
-1.69109273e+00 5.41145027e-01 -1.31902233e-01 3.35428357e-01
-5.59928119e-02 2.87224025e-01 4.51709837e-01 4.49987501e-01
2.61456370e-01 8.76738608e-01 -2.15511486e-01 -5.40527940e-01
-1.48428190e+00 6.63209438e-01 1.11346042e+00 1.08873963e+00
1.98754862e-01 2.23287463e-01 -4.28260684e-01 1.22328687e+00
-1.47861645e-01 3.37548554e-01 5.58114469e-01 -1.19712102e+00
2.57858336e-01 5.70765026e-02 -1.23479925e-02 -3.66004407e-01
-3.13852221e-01 -7.54295349e-01 -8.55747283e-01 4.75778878e-02
3.59736711e-01 -1.75523311e-01 -7.96451807e-01 1.73663795e+00
-6.72872066e-02 1.83641091e-01 -1.08752243e-01 7.63935030e-01
9.62050140e-01 5.87084770e-01 -3.87834638e-01 -1.61988884e-01
1.11362839e+00 -1.04325676e+00 -5.47936738e-01 1.63570344e-01
4.88330543e-01 -1.34129477e+00 1.06801724e+00 9.23826575e-01
-1.27753437e+00 -8.53017032e-01 -1.02246213e+00 -6.88728585e-04
1.29526660e-01 5.34540772e-01 4.72072244e-01 5.41655242e-01
-8.30683172e-01 1.04932272e+00 -5.77521086e-01 -1.43387103e-02
1.62926987e-02 1.97475076e-01 -3.56737413e-02 5.17230570e-01
-7.45818496e-01 3.37628037e-01 4.09030288e-01 -1.10977188e-01
-9.50292289e-01 -6.66916966e-01 -5.26967645e-01 -7.93627650e-02
1.11641884e-01 -7.19764888e-01 1.67399764e+00 -9.59259868e-01
-1.81304133e+00 9.20348346e-01 2.65730649e-01 -2.91080326e-01
3.49292666e-01 -2.24483028e-01 -3.95443022e-01 -9.79579091e-02
-7.50479400e-02 6.40895903e-01 9.15137649e-01 -1.28946078e+00
-4.99876171e-01 1.08318083e-01 -3.32632869e-01 1.23242356e-01
-2.56546382e-02 2.11748794e-01 -2.70070285e-01 -1.07893431e+00
5.73656633e-02 -1.30311179e+00 1.23671994e-01 -7.80848265e-01
-9.66029644e-01 1.51773080e-01 2.07490295e-01 -7.78888404e-01
1.61335766e+00 -2.21146989e+00 5.78658938e-01 1.86750144e-01
-1.32369567e-02 -1.08117372e-01 -2.51678318e-01 6.29276395e-01
-1.02710739e-01 1.43518925e-01 -3.81746948e-01 -7.17963576e-01
3.37318152e-01 -5.04297093e-02 -4.70923245e-01 -5.06492704e-03
-4.21385765e-02 6.95246398e-01 -7.66952395e-01 -4.69713062e-01
-3.34409550e-02 4.99055445e-01 -6.62109375e-01 3.47976089e-01
-4.02213156e-01 8.12432706e-01 1.03328571e-01 5.33195555e-01
1.31029025e-01 3.67472246e-02 1.06842943e-01 -4.60444093e-02
-2.44849116e-01 6.26831591e-01 -1.58664274e+00 2.09751081e+00
-5.08448422e-01 6.77189767e-01 2.74321102e-02 -1.80640802e-01
1.03626776e+00 6.71028733e-01 5.80364704e-01 1.19520545e-01
1.48591712e-01 4.41116184e-01 3.07377607e-01 -1.61154240e-01
7.73452699e-01 -1.85698524e-01 -4.98223692e-01 8.06513846e-01
5.11763215e-01 -8.45108807e-01 6.06569588e-01 1.25751182e-01
1.07025313e+00 5.99238515e-01 1.29559189e-01 1.34227052e-02
1.48197889e-01 2.31275074e-02 1.48530856e-01 6.69482648e-01
4.54938561e-01 1.36898804e+00 3.71643454e-01 -6.09149560e-02
-1.48170376e+00 -1.12315202e+00 -5.85377552e-02 1.40048063e+00
-7.52119064e-01 -1.20998931e+00 -9.67180848e-01 2.99691246e-03
-3.84737074e-01 5.26568532e-01 -2.97978103e-01 1.51296437e-01
-6.97031379e-01 -6.47142529e-01 1.13696229e+00 4.01378870e-01
-5.33572445e-03 -1.40311003e+00 -2.36264363e-01 3.46135288e-01
-4.93398815e-01 -8.65217507e-01 -7.51899719e-01 3.74193072e-01
-5.13957441e-01 -7.44980097e-01 -6.60416901e-01 -7.50372410e-01
-2.23039031e-01 -2.96559304e-01 1.68331647e+00 -2.22193316e-01
-3.58807176e-01 4.76284474e-02 -6.74678981e-01 -5.60130835e-01
-1.04716027e+00 4.60133195e-01 2.07882784e-02 -9.69161338e-04
-2.09892541e-01 -9.07624066e-01 -2.88799018e-01 7.74006695e-02
-8.93873215e-01 5.04379034e-01 1.82997927e-01 6.31306946e-01
8.30905855e-01 -2.03319624e-01 4.83944237e-01 -9.71352220e-01
6.90231442e-01 -1.75027698e-01 -2.01005861e-01 -1.43881515e-01
-2.04138309e-01 -6.76975772e-02 6.72846258e-01 -5.55176735e-01
-7.19390213e-01 8.31089020e-02 -3.11977625e-01 -3.05481225e-01
-3.08207422e-01 3.86393309e-01 -1.42970055e-01 5.26941419e-01
9.65493858e-01 3.53225172e-02 -6.21307790e-01 -9.28388298e-01
7.02884316e-01 1.01263106e+00 1.18065214e+00 -8.85519981e-01
6.75286055e-01 -4.69546858e-03 -2.28704140e-02 -7.55272627e-01
-9.99913514e-01 -4.10783887e-01 -8.87937248e-01 -2.09798247e-01
7.19086885e-01 -1.08098280e+00 -4.19621229e-01 3.34158212e-01
-9.84015465e-01 -5.38916945e-01 -7.28034258e-01 3.98437053e-01
-1.08417618e+00 -1.42197594e-01 -9.25666213e-01 -7.99801826e-01
-5.11043668e-01 -7.01258838e-01 1.33656681e+00 -6.07381277e-02
-8.12089205e-01 -4.99535024e-01 6.47945344e-01 3.16613257e-01
6.62335306e-02 6.47732675e-01 7.74900854e-01 -5.28772414e-01
-4.73947793e-01 9.95956585e-02 4.73673433e-01 4.03581679e-01
-2.96590794e-02 5.08683205e-01 -1.38486469e+00 1.08900316e-01
-3.43745440e-01 -4.19037253e-01 8.87966931e-01 3.10709119e-01
9.84887719e-01 -2.36675724e-01 1.88272074e-01 9.12731826e-01
1.00919604e+00 -1.19793266e-01 5.65606833e-01 1.32442564e-01
8.24620426e-01 4.14095372e-01 2.15661183e-01 6.87947392e-01
6.87774494e-02 9.11112547e-01 3.94360349e-02 1.46129191e-01
-5.08270025e-01 -4.59769189e-01 6.21948242e-01 1.54629695e+00
-7.07634509e-01 -3.16383570e-01 -7.01103032e-01 6.01025879e-01
-1.82492399e+00 -1.22635293e+00 -3.64031136e-01 2.11681557e+00
1.25505173e+00 -1.61269546e-01 5.61021805e-01 6.92111850e-01
4.39666539e-01 1.23100160e-02 -1.43492157e-02 -5.29692292e-01
-3.37025523e-01 6.77243233e-01 -1.19473733e-01 2.33369380e-01
-1.25517404e+00 7.88079441e-01 6.53433418e+00 1.05092537e+00
-7.65461326e-01 8.78239982e-03 2.76646435e-01 -4.08983976e-01
-2.40984142e-01 7.51832500e-02 -6.45207465e-01 2.70823687e-01
1.18447995e+00 1.43162161e-01 8.75472605e-01 5.51537812e-01
9.97693241e-02 3.05318326e-01 -1.46397519e+00 1.04809833e+00
2.59786636e-01 -1.30341780e+00 -1.06070548e-01 -8.11958984e-02
1.05366707e+00 -9.29635391e-02 -6.91423640e-02 3.35459948e-01
3.71009290e-01 -1.09950805e+00 1.29066622e+00 4.55561608e-01
1.05844641e+00 -7.50942111e-01 3.99963319e-01 3.35295111e-01
-1.15611720e+00 1.78284228e-01 7.74796456e-02 1.01172566e-01
1.96320161e-01 3.32142919e-01 -9.41495657e-01 6.37793899e-01
3.61911803e-01 7.04473376e-01 -6.85790837e-01 1.04846978e+00
-3.84327084e-01 1.06578493e+00 -3.09731632e-01 3.66827130e-01
-1.64756671e-01 -2.53002375e-01 6.98772907e-01 1.55274403e+00
5.95252931e-01 -2.46858597e-01 2.46253550e-01 9.41361248e-01
-1.24602422e-01 2.93815345e-01 -1.20551392e-01 -4.29468751e-01
3.08600783e-01 1.51108992e+00 -7.30117381e-01 -3.27882320e-01
3.44090968e-01 9.16254342e-01 -6.39742464e-02 3.81539464e-02
-6.77652061e-01 -2.56995797e-01 4.12732691e-01 1.71421796e-01
3.02577078e-01 -1.29435390e-01 -4.15038288e-01 -1.25417411e+00
-3.52558762e-01 -1.32308435e+00 3.24541271e-01 -1.06505382e+00
-1.33298540e+00 8.33865166e-01 -1.16861522e-01 -1.52084637e+00
-8.93747211e-01 -2.78606385e-01 -5.76744258e-01 8.08145702e-01
-6.57675326e-01 -1.47732413e+00 -9.68301147e-02 3.85788083e-01
6.74207926e-01 -3.19330841e-01 1.23623765e+00 3.48174095e-01
-1.19265139e-01 3.82030398e-01 6.35051131e-02 2.33801916e-01
1.14751673e+00 -1.64744413e+00 4.69605237e-01 3.63833010e-01
1.28934991e+00 3.60775679e-01 8.35659981e-01 -2.97165036e-01
-8.30293357e-01 -1.02519119e+00 8.52039814e-01 -8.14939737e-01
6.87942386e-01 -7.61901736e-01 -4.32530016e-01 6.82092369e-01
3.14423203e-01 -6.03397071e-01 1.08780468e+00 4.20745462e-01
-3.78904670e-01 1.82399854e-01 -3.77901882e-01 6.62060082e-01
1.06933677e+00 -5.66401541e-01 -5.00832736e-01 2.56558985e-01
5.19397438e-01 -5.76346397e-01 -1.03509355e+00 2.25391194e-01
8.71450067e-01 -1.09459507e+00 8.23349535e-01 -3.79450619e-01
7.35929549e-01 -5.12222707e-01 -1.58822313e-01 -1.62197006e+00
-3.97464514e-01 -1.13479769e+00 -9.99599844e-02 1.61303723e+00
6.12713516e-01 4.48504388e-01 3.87883812e-01 -2.37971231e-01
-4.89844620e-01 -1.12603240e-01 -4.97512668e-01 -5.53328037e-01
-1.33926412e-02 -7.71508217e-01 7.11634099e-01 7.77382851e-01
8.72161314e-02 8.96241307e-01 -6.43301308e-01 -3.35581809e-01
3.98103923e-01 5.63349783e-01 1.19930315e+00 -1.34308422e+00
-1.09709597e+00 -6.43039167e-01 -6.29744455e-02 -4.94922638e-01
-2.26966992e-01 -1.12689793e+00 1.78002208e-01 -1.23701835e+00
4.03475821e-01 -5.22286355e-01 -1.12525955e-01 4.69066262e-01
7.99102113e-02 8.66631269e-01 6.13581717e-01 4.46937799e-01
-4.96172965e-01 2.27198288e-01 1.18747520e+00 -5.15964553e-02
-4.87622917e-01 2.63954818e-01 -4.32903647e-01 8.60047936e-01
7.37799108e-01 -5.52965820e-01 -1.84422180e-01 -1.42246708e-01
5.30795515e-01 1.57483935e-01 2.41686553e-01 -1.37652540e+00
-1.01312809e-01 1.68964416e-01 3.26986700e-01 -5.14038205e-01
6.40439570e-01 -6.58462048e-02 6.72877491e-01 -1.95394963e-01
-6.59366965e-01 -1.72525167e-01 1.97287314e-02 5.40498225e-03
-3.11069876e-01 -2.94730544e-01 4.07851577e-01 -2.03813344e-01
1.19037163e-02 -8.54901038e-03 -2.30773956e-01 3.00589889e-01
1.87566653e-01 2.69774079e-01 -1.03757009e-01 -6.11520231e-01
-1.10885489e+00 -4.88285869e-01 4.67202932e-01 4.22879577e-01
1.53826132e-01 -1.50189865e+00 -1.18865192e+00 1.92784145e-01
1.88862294e-01 -1.35246590e-01 4.93521579e-02 5.85665464e-01
-4.47237164e-01 1.03802718e-01 -6.83332831e-02 -5.51777720e-01
-1.43332434e+00 2.12921172e-01 2.53839232e-02 -3.93965155e-01
-5.26753724e-01 9.34986591e-01 -9.41094756e-02 -5.03355086e-01
1.70397505e-01 -3.60432416e-01 -1.75126180e-01 2.99800158e-01
5.33618033e-01 1.09516650e-01 1.40651464e-01 -7.23983049e-01
1.19701058e-01 3.87745947e-01 5.70756555e-01 -7.89304435e-01
1.44812572e+00 2.02807724e-01 -6.51601627e-02 1.15256786e+00
6.67404294e-01 6.46163940e-01 -9.54707861e-01 2.05408558e-01
-1.29944965e-01 -7.56784230e-02 -1.88123778e-01 -1.00132465e+00
-5.03837824e-01 8.48210692e-01 2.41370469e-01 4.35467452e-01
9.55453575e-01 9.90766883e-02 7.28351533e-01 2.07071260e-01
2.64510810e-01 -1.15318751e+00 2.68329140e-02 5.70240974e-01
1.00329041e+00 -6.56770766e-01 -1.75317749e-01 -1.29396766e-01
-7.89988995e-01 1.03960323e+00 -2.59148031e-02 -1.94141343e-01
4.21794236e-01 6.49039030e-01 1.86003193e-01 6.57645166e-02
-9.99499679e-01 -3.76302093e-01 6.19800031e-01 3.73776287e-01
9.49080169e-01 4.76629138e-01 -1.89730171e-02 1.10351956e+00
-1.22847080e+00 -2.06250131e-01 7.20912635e-01 5.84945142e-01
-2.50193119e-01 -1.46769488e+00 -6.42319679e-01 2.53896564e-01
-6.25420451e-01 -4.63785410e-01 -7.23803163e-01 4.59901631e-01
6.85414910e-01 1.00851429e+00 -1.87025983e-02 -7.30297387e-01
7.62407482e-02 3.65353703e-01 8.86860788e-01 -9.24228132e-01
-1.07144976e+00 8.29679549e-01 4.34607506e-01 -1.17299952e-01
-5.11900723e-01 -7.57542968e-01 -9.98831749e-01 -6.84162751e-02
-2.14717701e-01 3.10120583e-01 7.75157332e-01 6.56052053e-01
7.45952055e-02 8.26463282e-01 4.54105347e-01 -1.12931526e+00
-2.65318811e-01 -1.49638426e+00 -8.33273888e-01 5.09794474e-01
1.35311440e-01 -1.93344966e-01 -3.13331544e-01 8.25048208e-01] | [15.877348899841309, 5.450754165649414] |
1d1031d2-1883-4f55-9003-a0563c474041 | backretrieval-an-image-pivoted-evaluation | 2105.04971 | null | https://arxiv.org/abs/2105.04971v1 | https://arxiv.org/pdf/2105.04971v1.pdf | Backretrieval: An Image-Pivoted Evaluation Metric for Cross-Lingual Text Representations Without Parallel Corpora | Cross-lingual text representations have gained popularity lately and act as the backbone of many tasks such as unsupervised machine translation and cross-lingual information retrieval, to name a few. However, evaluation of such representations is difficult in the domains beyond standard benchmarks due to the necessity of obtaining domain-specific parallel language data across different pairs of languages. In this paper, we propose an automatic metric for evaluating the quality of cross-lingual textual representations using images as a proxy in a paired image-text evaluation dataset. Experimentally, Backretrieval is shown to highly correlate with ground truth metrics on annotated datasets, and our analysis shows statistically significant improvements over baselines. Our experiments conclude with a case study on a recipe dataset without parallel cross-lingual data. We illustrate how to judge cross-lingual embedding quality with Backretrieval, and validate the outcome with a small human study. | ['Danushka Bollegala', 'Niall Twomey', 'Mikhail Fain'] | 2021-05-11 | null | null | null | null | ['unsupervised-machine-translation', 'cross-lingual-information-retrieval'] | ['natural-language-processing', 'natural-language-processing'] | [ 1.43076763e-01 -3.75183105e-01 -3.12038451e-01 -5.77868521e-01
-1.49836719e+00 -9.56774354e-01 1.19410908e+00 4.67885643e-01
-7.43889272e-01 4.32764500e-01 5.41414857e-01 -1.73709646e-03
2.92648017e-01 -3.62416625e-01 -9.27478313e-01 -2.60891348e-01
1.74542725e-01 5.85764110e-01 -2.23572895e-01 -1.30407631e-01
3.23998109e-02 8.19803849e-02 -9.69480157e-01 5.72724998e-01
4.98285413e-01 7.17912436e-01 -4.10149172e-02 2.80303538e-01
-1.71076789e-01 3.19797397e-01 -2.11478665e-01 -9.07257497e-01
3.34453911e-01 -5.23986161e-01 -7.24281490e-01 1.26985759e-01
1.03106844e+00 -9.86320712e-03 -1.69489354e-01 1.07926309e+00
4.96156037e-01 -2.07751349e-01 8.40788424e-01 -1.20613539e+00
-1.07465518e+00 5.03063381e-01 -5.97180068e-01 2.48687919e-02
2.54378974e-01 2.29340330e-01 1.47420228e+00 -1.01302028e+00
9.88181591e-01 1.32424068e+00 7.67047405e-01 2.55306900e-01
-1.70829320e+00 -4.89856690e-01 -2.72348881e-01 8.22507963e-02
-1.31452608e+00 -4.37628448e-01 5.22559285e-01 -5.75751066e-01
8.74501169e-01 -1.84239328e-01 4.42888774e-02 1.27685571e+00
2.33870409e-02 8.43540728e-01 1.41411459e+00 -6.36831760e-01
-2.44549662e-01 4.51750904e-01 -1.02668844e-01 7.70367444e-01
2.61365771e-01 2.01456219e-01 -6.19833827e-01 5.55383489e-02
4.50478017e-01 -3.32494557e-01 -2.64438272e-01 -6.97421253e-01
-1.59632039e+00 1.07354152e+00 6.08872354e-01 6.13180280e-01
-1.90371290e-01 2.44550303e-01 9.59641159e-01 4.88878936e-01
6.66628480e-01 7.52011001e-01 -4.15265054e-01 9.13380645e-03
-9.59502697e-01 8.21484476e-02 5.15524626e-01 9.49858844e-01
7.56812453e-01 -1.56750172e-01 -3.36574838e-02 9.77225661e-01
9.05093551e-02 5.69488883e-01 7.45211780e-01 -7.28612304e-01
6.76050961e-01 4.13777202e-01 -8.72327164e-02 -1.02849090e+00
4.59932610e-02 -7.15923756e-02 -5.56705236e-01 -4.43357928e-03
3.78117651e-01 2.19936326e-01 -4.32271689e-01 1.82871056e+00
-1.60511136e-01 -3.60553294e-01 2.52847165e-01 8.29808474e-01
5.61700225e-01 7.44466245e-01 1.47593379e-01 5.34728244e-02
1.35090733e+00 -1.10969126e+00 -4.84947413e-01 -2.24718869e-01
9.06401217e-01 -1.16690063e+00 1.49665785e+00 2.01190934e-01
-7.70355642e-01 -5.80492675e-01 -1.00415564e+00 -4.92179632e-01
-6.19295478e-01 2.89165705e-01 3.17593932e-01 2.96314210e-01
-1.03032172e+00 4.76584882e-01 -4.70786929e-01 -7.69900739e-01
5.34390658e-02 -2.41119176e-01 -8.86747241e-01 -3.99044335e-01
-1.09474909e+00 1.22355354e+00 3.21747363e-01 -3.58650595e-01
-7.74709404e-01 -4.88381177e-01 -1.06056178e+00 -3.66276920e-01
1.08955123e-01 -3.88715386e-01 1.02747393e+00 -1.26404011e+00
-9.25712168e-01 1.57660949e+00 1.79495677e-01 -5.48730850e-01
6.75849020e-01 -2.70939082e-01 -5.25018632e-01 3.51552144e-02
5.36254644e-01 9.97615039e-01 6.73730671e-01 -1.20821476e+00
-2.34568313e-01 -3.61512393e-01 8.50197375e-02 2.24461898e-01
-4.27366316e-01 5.73321916e-02 -6.85228229e-01 -7.49184728e-01
-1.71470791e-01 -1.03307867e+00 -8.44190940e-02 2.44012848e-01
-1.56908721e-01 -1.33709818e-01 4.40775037e-01 -7.51411855e-01
7.39725411e-01 -2.30776477e+00 1.42816067e-01 -1.61825538e-01
-1.49703041e-01 -1.52628094e-01 -5.58468640e-01 5.43574274e-01
-1.98949084e-01 2.12978721e-01 -3.79659057e-01 -5.11677623e-01
2.54293680e-01 1.89839020e-01 -2.12265000e-01 6.50614500e-01
2.88208514e-01 1.13397598e+00 -9.96920466e-01 -7.81698346e-01
1.60763562e-01 4.67954695e-01 -3.61240625e-01 1.17072977e-01
-1.39090031e-01 -2.92925327e-03 -1.24501362e-01 1.49081215e-01
9.12135094e-02 -3.14234853e-01 2.00235009e-01 -6.57705843e-01
-5.55736050e-02 3.81867379e-01 -5.04873872e-01 2.21993470e+00
-7.57935345e-01 9.34426606e-01 -1.86941594e-01 -8.70320559e-01
8.44330847e-01 2.31031865e-01 2.29221329e-01 -1.06614304e+00
5.41109294e-02 2.95558393e-01 -2.73864627e-01 -3.23040456e-01
8.09314787e-01 -2.72469074e-01 -4.73538131e-01 7.23326385e-01
2.08164066e-01 -4.82888401e-01 3.88122320e-01 3.12201440e-01
7.69706786e-01 4.75664079e-01 3.35802794e-01 -3.94360900e-01
3.54508698e-01 2.75172532e-01 1.87702645e-02 2.67968804e-01
-3.07061046e-01 7.51590610e-01 2.54542261e-01 -4.60553735e-01
-1.40732396e+00 -1.00443447e+00 -1.90846071e-01 9.98700321e-01
-1.62598751e-02 -6.48759723e-01 -6.51682079e-01 -7.94499159e-01
-8.31494406e-02 6.92573488e-01 -7.46976912e-01 9.18502584e-02
-2.49289215e-01 -5.88212609e-01 7.49146879e-01 3.65453869e-01
4.31678683e-01 -8.12527180e-01 -2.94901997e-01 2.22938079e-02
-3.07761252e-01 -1.50402427e+00 -7.61762857e-01 5.90265356e-02
-6.84751630e-01 -9.17188108e-01 -7.56785274e-01 -9.79390264e-01
5.11967182e-01 3.05411190e-01 1.57985675e+00 -1.89821795e-01
-3.89996201e-01 7.29776561e-01 -3.63201171e-01 8.35993960e-02
-7.11618781e-01 3.17551754e-02 1.15464680e-01 -1.56635731e-01
4.73650843e-01 -2.26196453e-01 -3.62114966e-01 3.57549995e-01
-1.00342000e+00 2.02565596e-01 6.18730545e-01 9.16059613e-01
7.53486812e-01 -5.52965820e-01 1.66835770e-01 -8.33877921e-01
7.48112798e-01 -2.56026775e-01 -6.40721142e-01 4.74650532e-01
-6.92775071e-01 3.25069666e-01 5.02807140e-01 -2.20630601e-01
-5.68139434e-01 -1.00385867e-01 3.82626891e-01 -5.48359990e-01
-1.16234878e-02 6.96633041e-01 1.46769330e-01 2.96850860e-01
9.25086915e-01 2.26563752e-01 -5.23910150e-02 -4.12933975e-01
8.54422033e-01 4.50065702e-01 5.36660075e-01 -8.64156425e-01
6.88060403e-01 3.49996805e-01 -3.54754388e-01 -7.34175384e-01
-7.65849531e-01 -4.23377812e-01 -7.87694752e-01 1.27159236e-02
9.56195891e-01 -1.24539208e+00 -8.84809271e-02 -5.24573363e-02
-1.22515178e+00 -2.14983761e-01 -1.89117879e-01 6.39538527e-01
-6.32571936e-01 3.04832131e-01 -6.73986971e-01 -1.37744337e-01
-2.37209231e-01 -1.23114681e+00 1.33763671e+00 -4.13693011e-01
-3.78246516e-01 -1.21034980e+00 3.92646372e-01 4.52751249e-01
2.14900285e-01 2.32945576e-01 9.38571632e-01 -5.46315432e-01
-4.53081727e-01 -2.92481273e-01 -4.44115847e-01 6.02458060e-01
1.33211702e-01 -3.76007855e-02 -8.44473898e-01 -4.37101036e-01
-3.62323225e-01 -9.11031604e-01 7.79952765e-01 -3.56450975e-02
7.30482221e-01 -6.15884922e-02 -5.18982820e-02 3.55167985e-01
1.64810467e+00 -5.85103452e-01 4.77623165e-01 3.31640691e-01
6.88332558e-01 7.47243166e-01 4.78407711e-01 -1.40544936e-01
4.31793958e-01 8.71500850e-01 -6.06434792e-02 -2.21527219e-01
-4.21497226e-01 -4.01657104e-01 7.17305958e-01 1.46407843e+00
2.79475093e-01 -2.36311361e-01 -8.41882050e-01 8.52306306e-01
-1.44058156e+00 -7.61046886e-01 1.80351269e-02 2.13351679e+00
1.01190078e+00 -9.11861435e-02 -1.02150574e-01 -3.38137031e-01
6.22227609e-01 2.21485540e-01 -1.90951735e-01 -4.33956355e-01
-3.04688185e-01 3.42031755e-02 6.09396040e-01 3.56149346e-01
-1.07716990e+00 1.10356271e+00 6.29316664e+00 7.51472056e-01
-1.21093774e+00 4.56310064e-01 6.47997499e-01 7.18821213e-02
-3.13820511e-01 -1.44471437e-01 -4.80241388e-01 1.17163859e-01
8.40404272e-01 -2.55257308e-01 3.22144538e-01 8.01349521e-01
-1.24943912e-01 1.49808377e-01 -1.67451966e+00 1.21584642e+00
4.71959114e-01 -1.27858377e+00 1.85571760e-01 -1.76565479e-02
9.59360957e-01 5.66741765e-01 1.67484045e-01 2.96355277e-01
4.46908891e-01 -9.54790831e-01 7.44180202e-01 -7.33514428e-02
1.09798622e+00 -4.65820700e-01 4.09180999e-01 -1.01217918e-01
-9.52496886e-01 6.72492087e-01 -2.95961171e-01 3.42570037e-01
3.26974802e-02 2.95479834e-01 -8.34395885e-01 5.18577218e-01
6.25919640e-01 1.00255537e+00 -7.56359160e-01 4.44269001e-01
-1.88829616e-01 3.97496849e-01 -2.23119795e-01 5.54480441e-02
5.97841322e-01 -2.43564978e-01 1.50415733e-01 1.47870004e+00
3.18982840e-01 -5.24483085e-01 2.38169238e-01 9.44532454e-01
-5.18540442e-01 6.03642166e-01 -1.04913414e+00 -5.40788949e-01
8.02543461e-02 1.22324646e+00 -5.60528457e-01 -4.18736309e-01
-7.29592741e-01 1.23390985e+00 4.92099524e-01 1.92246422e-01
-6.95888996e-01 5.59328608e-02 5.56837976e-01 4.58824597e-02
3.35599040e-03 -4.45707798e-01 -1.37075469e-01 -1.33268893e+00
5.34021445e-02 -1.08212769e+00 2.79673487e-01 -9.48905706e-01
-1.78304243e+00 7.71744609e-01 -3.04928776e-02 -1.38510799e+00
-3.02648991e-01 -8.24879348e-01 4.10925411e-02 8.32180917e-01
-1.52277982e+00 -1.28500044e+00 1.67582580e-03 6.10383451e-01
6.61811233e-01 -3.27342063e-01 1.14712346e+00 5.05351782e-01
-1.92750812e-01 7.01348722e-01 3.34380478e-01 4.42797065e-01
1.17497015e+00 -1.05327499e+00 5.89113951e-01 8.10102046e-01
7.96155751e-01 6.27233744e-01 6.62818193e-01 -3.82968992e-01
-1.40560102e+00 -1.12511909e+00 8.33520234e-01 -5.45534134e-01
1.03123116e+00 -4.34124291e-01 -7.91892231e-01 7.42695749e-01
6.77346885e-01 8.13070312e-02 7.33922124e-01 1.49652392e-01
-1.04041159e+00 5.54563254e-02 -8.75623286e-01 6.68147802e-01
7.43222415e-01 -1.13801777e+00 -5.92249930e-01 5.50158799e-01
6.96502209e-01 -4.17094380e-02 -9.88020003e-01 5.46731278e-02
5.66739738e-01 -7.79428601e-01 8.92509401e-01 -5.09465992e-01
8.45559597e-01 -3.02701145e-01 -7.09912479e-01 -1.38151598e+00
-9.85797197e-02 -1.67505428e-01 7.52907813e-01 1.22794950e+00
6.63484097e-01 -2.74892330e-01 3.04677218e-01 2.40119666e-01
8.85872543e-02 -2.12792665e-01 -8.02712560e-01 -9.12247121e-01
2.96057671e-01 -4.99765068e-01 7.65976310e-02 1.48194313e+00
-1.94168463e-01 7.58850932e-01 -3.38083446e-01 -1.70044601e-01
5.97453356e-01 2.36845002e-01 8.87638152e-01 -7.10160792e-01
-3.46389920e-01 -4.38703597e-01 -5.64553201e-01 -8.34354281e-01
3.80210191e-01 -1.40570223e+00 3.43339711e-01 -1.32381487e+00
3.81482542e-01 -4.16899115e-01 -2.75328368e-01 3.79227430e-01
1.56860575e-01 7.02715695e-01 2.12855533e-01 2.27759764e-01
-6.38440311e-01 5.88067412e-01 1.02069533e+00 -4.88334894e-01
4.33828235e-01 -8.92323673e-01 -3.99470389e-01 5.17074943e-01
5.60312390e-01 -4.77694243e-01 -3.57881784e-01 -9.01300967e-01
2.93580413e-01 -3.02768469e-01 3.33973587e-01 -7.13081002e-01
-3.00585359e-01 5.34202233e-02 2.55592942e-01 -1.37461379e-01
4.42330301e-01 -8.45702291e-01 -7.38978311e-02 6.55564815e-02
-8.36067617e-01 6.11454189e-01 2.02236965e-01 4.03061330e-01
-4.77569968e-01 -1.30434752e-01 7.40501463e-01 -1.27215117e-01
-8.55828345e-01 8.55612084e-02 1.02791317e-01 4.35382634e-01
6.85621798e-01 1.55999124e-01 -3.25092554e-01 -2.88072586e-01
-2.73182869e-01 2.18037348e-02 9.84507143e-01 8.03095579e-01
3.99160028e-01 -1.73359561e+00 -9.59205210e-01 -4.94712330e-02
8.07682753e-01 -6.75242901e-01 -3.19597214e-01 6.90236092e-01
-5.90201735e-01 4.00842816e-01 -2.20525935e-01 -8.57997656e-01
-1.19763923e+00 6.32615745e-01 1.08054824e-01 -3.47395301e-01
-3.89163524e-01 5.70357502e-01 4.34156567e-01 -6.29447997e-01
-1.91938095e-02 -3.16761822e-01 2.55175710e-01 2.14692671e-02
4.03400660e-01 -2.01393485e-01 7.35862851e-02 -1.11486804e+00
-2.73043275e-01 7.21743166e-01 -8.79586190e-02 -5.11436582e-01
1.20323455e+00 -1.71866909e-01 -1.26329079e-01 8.66627276e-01
1.85957384e+00 -1.66968539e-01 -8.41586113e-01 -5.94729185e-01
1.61059931e-01 -4.47711229e-01 6.29964173e-02 -7.62315333e-01
-9.20574903e-01 1.09860921e+00 8.05252314e-01 -1.08219996e-01
7.46080816e-01 1.36913553e-01 6.95598602e-01 6.07671678e-01
4.81586456e-01 -9.87977684e-01 2.74977416e-01 3.52983147e-01
1.18005824e+00 -1.62871420e+00 1.65358454e-01 5.72730564e-02
-8.45457792e-01 8.67177248e-01 2.79265463e-01 -1.32796377e-01
2.66393453e-01 -3.04404665e-02 2.77972370e-01 -1.75791964e-01
-6.85301065e-01 -1.29551962e-01 4.57511902e-01 3.92272443e-01
9.38229263e-01 3.18972379e-01 -2.62469113e-01 4.02724603e-03
-1.84419230e-01 -1.20292664e-01 2.16669962e-01 6.53691947e-01
-2.50363685e-02 -1.41470003e+00 -1.97671473e-01 1.00258984e-01
-3.36379468e-01 -4.79377002e-01 -4.46299374e-01 1.00564492e+00
-6.24394715e-02 5.52480996e-01 6.61596656e-02 -1.56773791e-01
2.07086682e-01 1.55439004e-01 8.20475161e-01 -4.98517036e-01
-5.31638026e-01 6.24726042e-02 2.45428354e-01 -5.49285829e-01
-5.99894345e-01 -7.56349504e-01 -8.58389795e-01 -2.35110447e-01
2.99318098e-02 8.79127607e-02 1.02860153e+00 7.16617644e-01
3.68443429e-01 9.20959190e-02 3.82028818e-01 -7.02787519e-01
-3.28026831e-01 -9.48505282e-01 -1.92525566e-01 1.08407032e+00
1.31234899e-02 -4.99169648e-01 -1.58946514e-01 5.76801836e-01] | [11.17158031463623, 1.6574817895889282] |
1e98f1e0-eec5-493f-960e-86ddc3567246 | two-phase-dual-copod-method-for-anomaly | 2305.00982 | null | https://arxiv.org/abs/2305.00982v1 | https://arxiv.org/pdf/2305.00982v1.pdf | Two-phase Dual COPOD Method for Anomaly Detection in Industrial Control System | Critical infrastructures like water treatment facilities and power plants depend on industrial control systems (ICS) for monitoring and control, making them vulnerable to cyber attacks and system malfunctions. Traditional ICS anomaly detection methods lack transparency and interpretability, which make it difficult for practitioners to understand and trust the results. This paper proposes a two-phase dual Copula-based Outlier Detection (COPOD) method that addresses these challenges. The first phase removes unwanted outliers using an empirical cumulative distribution algorithm, and the second phase develops two parallel COPOD models based on the output data of phase 1. The method is based on empirical distribution functions, parameter-free, and provides interpretability by quantifying each feature's contribution to an anomaly. The method is also computationally and memory-efficient, suitable for low- and high-dimensional datasets. Experimental results demonstrate superior performance in terms of F1-score and recall on three open-source ICS datasets, enabling real-time ICS anomaly detection. | ['Jerry Bruce', 'Emmanuel Aboah Boateng'] | 2023-04-30 | null | null | null | null | ['outlier-detection'] | ['methodology'] | [ 1.54050458e-02 -2.69381523e-01 2.35648006e-01 -1.66838229e-01
-4.83939201e-01 -7.77548432e-01 2.94626296e-01 8.31932247e-01
3.62428166e-02 6.13728821e-01 -5.64353347e-01 -3.96704257e-01
-5.50194204e-01 -7.32812047e-01 -3.85708809e-01 -7.95296967e-01
-2.06662297e-01 5.62898099e-01 1.89604804e-01 3.58610362e-01
4.85297412e-01 9.14584160e-01 -1.30319560e+00 1.11049533e-01
1.09565187e+00 1.14048564e+00 -3.55195284e-01 4.54528809e-01
2.35650077e-01 9.50961411e-02 -9.10775304e-01 -2.14745358e-01
2.80692905e-01 -1.56606823e-01 -8.72111171e-02 -2.14747161e-01
-2.90718764e-01 -1.50197847e-02 1.92566425e-01 1.17810190e+00
1.19694926e-01 -1.98892936e-01 7.94806600e-01 -1.87009883e+00
-3.41685623e-01 -6.36011064e-02 -5.65030456e-01 7.22161457e-02
3.00054252e-01 2.48415560e-01 9.38644826e-01 -9.31812108e-01
8.60325694e-02 9.43802536e-01 6.14714086e-01 1.02466583e-01
-1.44296539e+00 -6.34772003e-01 3.49162403e-03 -1.18610613e-01
-1.16959250e+00 1.13666534e-01 5.81428826e-01 -4.14752305e-01
1.12813234e+00 5.06433129e-01 4.90628183e-01 9.38790739e-01
9.57610786e-01 4.17973191e-01 5.72909474e-01 -1.25354901e-01
7.82441556e-01 -1.30111456e-01 1.55308515e-01 1.25774845e-01
1.26984024e+00 2.97328383e-02 -2.15019509e-01 -7.04990685e-01
3.71656537e-01 6.24905527e-01 1.00206034e-02 -4.20503557e-01
-8.00071776e-01 6.95539713e-01 4.02951166e-02 3.14232320e-01
-4.82610822e-01 -1.53408498e-01 5.44997394e-01 3.96568388e-01
3.82578611e-01 7.66332924e-01 -5.44110954e-01 -3.33504558e-01
-6.77121162e-01 2.10378870e-01 9.34843123e-01 9.29269254e-01
3.68739009e-01 8.32893476e-02 3.42445355e-03 3.00615966e-01
6.26258314e-01 6.62945569e-01 9.23017189e-02 -2.70439476e-01
3.16043556e-01 1.01671576e+00 5.26269563e-02 -1.03376281e+00
-4.20635283e-01 -1.92025453e-01 -8.37695360e-01 3.54080170e-01
1.04361460e-01 -1.29522337e-02 -9.36587870e-01 1.09359622e+00
2.71527290e-01 1.08195506e-01 -6.19901763e-03 5.35215259e-01
-9.40006599e-02 5.39813995e-01 2.05596606e-03 -3.47015858e-01
1.08506167e+00 -3.43789846e-01 -9.36631441e-01 -9.17736143e-02
5.33857942e-01 -6.62261069e-01 7.87546754e-01 8.71227622e-01
-6.26627445e-01 -6.03510588e-02 -1.43652916e+00 5.71585536e-01
-5.26563227e-01 -1.82090580e-01 4.30954069e-01 9.45806623e-01
-4.81511712e-01 7.63941884e-01 -1.25026560e+00 -1.15484215e-01
5.78188658e-01 6.20483816e-01 -3.31502199e-01 -1.02608092e-01
-6.92121148e-01 4.92057413e-01 3.71575952e-01 2.89251804e-01
-6.87730134e-01 -8.34487498e-01 -7.28242695e-01 1.25376686e-01
1.53019235e-01 -2.90804803e-01 8.49876761e-01 -4.20964301e-01
-1.09015560e+00 7.48772323e-02 9.42749977e-02 -2.86272556e-01
3.78856838e-01 -4.87821877e-01 -8.50236833e-01 -4.77113090e-02
1.06293909e-01 -4.89951104e-01 1.02047765e+00 -1.16858339e+00
-6.14616275e-01 -7.41996944e-01 -6.73065364e-01 -5.56640863e-01
-3.16060036e-01 -5.45564331e-02 -1.29314801e-02 -4.71603036e-01
2.64272928e-01 -8.42166483e-01 -2.85660923e-01 -3.78910840e-01
-6.35617614e-01 1.00722328e-01 1.31696010e+00 -5.49903214e-01
1.48883653e+00 -2.36486793e+00 -3.92490447e-01 1.03260696e+00
5.57002872e-02 3.00435126e-01 1.88446730e-01 7.54952729e-01
-3.66795748e-01 2.45840605e-02 -5.51015854e-01 -1.14065297e-01
2.36705124e-01 2.89196014e-01 -4.54734176e-01 6.17426097e-01
6.25113189e-01 5.16599178e-01 -8.76806378e-01 1.92740798e-01
4.38078582e-01 1.59693405e-01 -4.52497065e-01 1.93093807e-01
8.57071951e-03 2.99874634e-01 -6.82349145e-01 1.05541956e+00
8.16554725e-01 4.29159850e-02 2.26084188e-01 1.19692989e-01
1.67004112e-02 -2.00221121e-01 -1.25338948e+00 1.03660774e+00
-1.62451223e-01 3.41481537e-01 -1.10397540e-01 -7.68312991e-01
1.17484212e+00 2.76143998e-01 5.92413962e-01 -5.76348484e-01
1.51427075e-01 5.99589586e-01 -3.96841951e-02 -3.00907344e-01
1.66833952e-01 7.63419122e-02 -3.49186271e-01 4.32394236e-01
-1.67574361e-01 -2.76172549e-01 1.91140965e-01 5.87416179e-02
1.66927660e+00 -3.27942878e-01 4.59560305e-01 -3.72083008e-01
6.13308609e-01 -2.62124777e-01 8.42901826e-01 3.48406136e-01
2.23333128e-02 5.01527786e-01 9.90339458e-01 -3.64678055e-01
-8.85180354e-01 -1.57730675e+00 -2.43422911e-01 7.52537549e-02
-6.51347041e-02 -4.25690800e-01 -5.36380887e-01 -7.18119264e-01
6.32152200e-01 1.09627104e+00 -4.55258548e-01 -5.19144952e-01
-1.88436985e-01 -8.26033831e-01 2.41695940e-01 6.05735362e-01
-1.53820813e-01 -9.16625023e-01 -5.14188886e-01 3.79972458e-01
5.49109995e-01 -9.20519888e-01 -1.35127068e-01 4.36941683e-01
-9.37420726e-01 -1.38511395e+00 2.23372579e-01 -1.13784485e-01
1.15671074e+00 -2.06543133e-01 8.38084698e-01 2.28439067e-02
-6.72750831e-01 3.55761975e-01 -2.74670959e-01 -9.06431079e-01
-3.51702094e-01 -4.58923668e-01 5.14306724e-01 1.50408313e-01
6.44283652e-01 -6.23682559e-01 -3.99452925e-01 3.80445898e-01
-1.15740728e+00 -1.16680384e+00 7.25801289e-01 7.25269198e-01
6.61503732e-01 3.49665433e-01 8.26272786e-01 -8.83944392e-01
9.59344447e-01 -6.88166320e-01 -9.98630166e-01 1.06977649e-01
-1.08959150e+00 -3.74296844e-01 8.49038541e-01 -3.60555314e-02
-7.36937344e-01 6.90292865e-02 2.30242416e-01 -5.60897529e-01
-3.15718353e-01 3.66309971e-01 -4.95580494e-01 2.08252922e-01
4.52041656e-01 -2.87711084e-01 -5.69084920e-02 -3.21061879e-01
-2.44983137e-01 6.10065341e-01 5.16729236e-01 -4.29079205e-01
1.03936863e+00 3.99248838e-01 2.17912778e-01 -8.40930581e-01
-3.87888551e-01 -5.75369537e-01 -5.83136857e-01 3.12845013e-03
6.33915305e-01 -5.77228129e-01 -9.03925598e-01 4.20823246e-01
-9.61503267e-01 2.58926421e-01 -3.62224996e-01 4.64223444e-01
-2.03681275e-01 3.66304874e-01 -3.77382874e-01 -9.62345839e-01
-4.22380447e-01 -9.20013011e-01 9.97771323e-01 2.24071130e-01
-5.35139501e-01 -1.00916147e+00 2.89362252e-01 -1.52510762e-01
1.94256291e-01 7.84239769e-01 1.05164528e+00 -1.31617606e+00
-5.40076792e-01 -1.01226020e+00 3.36394794e-02 6.58223331e-01
5.31143248e-01 4.10445303e-01 -1.02371466e+00 -7.27257907e-01
2.01024026e-01 2.30741128e-01 1.22439802e-01 3.62265319e-01
1.23540616e+00 -9.64139029e-02 -4.39735800e-01 2.20732331e-01
1.46579957e+00 4.86976594e-01 7.45623171e-01 2.30282485e-01
5.42702496e-01 4.94256318e-01 1.00458789e+00 5.48359871e-01
-3.92452806e-01 1.28019944e-01 8.18809330e-01 8.31908509e-02
8.88922811e-01 -3.36899869e-02 4.01460439e-01 6.92276895e-01
3.19756925e-01 -2.59045511e-01 -1.08610570e+00 6.20941579e-01
-2.05663753e+00 -6.62215710e-01 -5.63950837e-01 2.49422431e+00
3.50446105e-02 4.89557087e-01 -1.06912829e-01 3.73537064e-01
5.88762522e-01 -5.14853656e-01 -6.80611253e-01 -7.58658051e-01
2.61532307e-01 3.32954764e-01 3.48382741e-01 -2.23241542e-02
-9.49536026e-01 1.67066485e-01 6.03620052e+00 5.80554843e-01
-6.08259082e-01 -3.20859194e-01 5.39201736e-01 -6.57512844e-02
-1.29676625e-01 -8.78919289e-02 -3.25100452e-01 6.55564308e-01
1.27273524e+00 -3.03687751e-01 -1.47068650e-02 9.09327507e-01
2.92690396e-01 -1.78053468e-01 -1.42711353e+00 8.37918282e-01
-2.41336688e-01 -8.10569584e-01 -2.07613334e-01 3.34385246e-01
6.40575767e-01 -9.92543101e-02 3.12600359e-02 -1.10733649e-02
1.11506939e-01 -1.02750552e+00 3.22988510e-01 4.85361367e-01
3.18024755e-01 -1.25510716e+00 1.26029265e+00 7.79813621e-03
-9.65048194e-01 -4.99542117e-01 -1.62730247e-01 1.48310930e-01
1.97026432e-01 1.22871649e+00 -9.19184864e-01 9.67508435e-01
8.16300094e-01 4.29115176e-01 -4.05129075e-01 1.11469722e+00
-2.02650949e-01 6.01651907e-01 -4.30026442e-01 -4.34839055e-02
-5.25578596e-02 -6.57956243e-01 7.55317092e-01 7.90649235e-01
6.69429183e-01 -4.07767862e-01 3.20739811e-03 8.66456985e-01
2.67948061e-01 -1.98639587e-01 -9.03620720e-01 -3.55817974e-01
5.82212329e-01 1.33623099e+00 -1.04317236e+00 2.99090147e-01
-3.87997866e-01 6.12158835e-01 -3.81523579e-01 1.97914064e-01
-7.17833042e-01 -7.45869219e-01 9.39757586e-01 -3.23056206e-02
3.49106520e-01 -2.35820875e-01 -6.38747394e-01 -6.70214891e-01
4.85739589e-01 -8.78951013e-01 5.52937627e-01 -1.47687986e-01
-1.81823349e+00 3.45839888e-01 -9.81395096e-02 -1.58189404e+00
-3.08744401e-01 -1.05148673e+00 -9.38748300e-01 5.30144155e-01
-1.03814912e+00 -7.78917909e-01 -1.08263075e-01 4.32619482e-01
1.11497961e-01 -1.76954731e-01 9.74444389e-01 2.46453807e-01
-8.97377431e-01 3.93317938e-01 4.60085750e-01 -8.54382664e-02
6.11593425e-01 -1.42572308e+00 4.91734833e-01 1.01131082e+00
-1.33299798e-01 6.96984112e-01 8.50841463e-01 -9.88281429e-01
-1.57803822e+00 -1.06528127e+00 4.29826081e-01 -6.91673100e-01
1.00879800e+00 -6.20898843e-01 -1.05483711e+00 5.64513743e-01
-1.30557120e-01 2.65842886e-03 1.07753944e+00 1.94117710e-01
-6.79757297e-02 -3.39218050e-01 -1.64058423e+00 4.20694470e-01
3.99698645e-01 -1.87840879e-01 -5.19796789e-01 3.25878114e-01
4.67799187e-01 -1.43137693e-01 -1.00553298e+00 5.16990840e-01
1.85151652e-01 -9.49992537e-01 5.90388417e-01 -5.42206287e-01
-1.48755148e-01 -5.59030533e-01 8.71597789e-04 -1.24988425e+00
8.61538516e-04 -7.65120089e-01 -2.78360188e-01 1.15793622e+00
5.97617388e-01 -1.23840153e+00 4.26658958e-01 7.32042134e-01
-3.01843554e-01 -6.71462595e-01 -1.09409094e+00 -1.07730722e+00
-3.39594752e-01 -6.51232719e-01 8.69640112e-01 8.28766406e-01
1.63398474e-01 -1.26261845e-01 2.06053436e-01 7.41090834e-01
5.69388390e-01 1.42258070e-02 6.83680415e-01 -1.67388880e+00
-1.10839289e-02 -8.40874240e-02 -8.12796772e-01 1.21701501e-01
-1.29398718e-01 -4.70297188e-01 -1.53890308e-02 -1.12041593e+00
-2.04728946e-01 -1.92795455e-01 -6.25953555e-01 5.24674118e-01
-8.66906643e-02 2.72590462e-02 -2.43861139e-01 4.11897637e-02
-5.09594560e-01 5.40402293e-01 4.54765677e-01 4.38684076e-02
-2.03084543e-01 2.41372496e-01 -3.60569149e-01 7.91060984e-01
9.98199463e-01 -7.88243890e-01 -5.17338097e-01 1.09844781e-01
2.34196916e-01 -3.77883077e-01 3.42846841e-01 -1.04523230e+00
1.82163581e-01 -3.07795461e-02 6.13844693e-01 -8.19078803e-01
2.93188542e-02 -1.47074640e+00 2.55472898e-01 8.46845567e-01
5.77875555e-01 6.93088531e-01 6.44351125e-01 9.18204725e-01
-3.60790282e-01 -2.25639150e-01 4.35798436e-01 5.80792069e-01
-1.88250631e-01 2.89510459e-01 -4.26294506e-01 -3.33178431e-01
1.70747209e+00 -2.63120532e-01 -1.77487507e-01 -5.15526682e-02
-4.87990350e-01 4.96541232e-01 5.35944879e-01 5.27880490e-01
9.01286364e-01 -1.34102380e+00 -4.50307220e-01 7.54013360e-01
3.38503927e-01 1.26793653e-01 1.30179301e-01 7.87603259e-01
-7.25541890e-01 3.06803018e-01 -1.54582977e-01 -8.66094589e-01
-1.00504601e+00 5.14010727e-01 -1.04501933e-01 -4.39515769e-01
-3.75530392e-01 3.15413803e-01 -1.57777801e-01 -4.72293466e-01
-1.54707596e-01 -3.28236282e-01 3.69696230e-01 -9.57300663e-02
4.98293787e-01 6.61840856e-01 5.74579120e-01 5.01263700e-03
-6.85199738e-01 1.77429080e-01 -2.35581070e-01 2.14432895e-01
1.52107155e+00 3.61274123e-01 -5.33450902e-01 5.46404481e-01
9.28097010e-01 4.54518199e-02 -9.66618657e-01 5.47155559e-01
7.35355735e-01 -7.84919679e-01 -3.97734672e-01 -6.76281452e-01
-7.78584063e-01 5.43946922e-01 4.95475382e-01 7.98846483e-01
1.31287193e+00 -4.50450361e-01 4.89452451e-01 3.75105858e-01
2.75865227e-01 -1.20958900e+00 1.61254197e-01 3.39409381e-01
6.96714163e-01 -8.84591043e-01 1.98124662e-01 -4.58481908e-01
-3.66685867e-01 1.26109910e+00 6.23529255e-01 -2.06261292e-01
7.74848819e-01 5.88094234e-01 -7.71558359e-02 -3.69380176e-01
-6.46609008e-01 6.34345233e-01 2.59845465e-01 6.95669830e-01
-3.20859551e-02 6.37154430e-02 -4.00737785e-02 7.49412477e-01
1.84417054e-01 -5.74541092e-01 2.67533213e-01 1.13483942e+00
-2.12039381e-01 -1.32625556e+00 -4.64484483e-01 8.16907346e-01
-4.48524117e-01 2.96472907e-01 -4.59768355e-01 9.59612608e-01
-2.09416881e-01 1.09196365e+00 3.49422067e-01 -2.69348025e-01
7.28894353e-01 -9.78776225e-05 -1.14295982e-01 -5.71371973e-01
-5.24481118e-01 -1.76031480e-03 -2.25743309e-01 -9.57919419e-01
3.61638725e-01 -9.29805934e-01 -1.42895508e+00 -2.26881430e-01
-4.93743509e-01 3.17348897e-01 8.51572037e-01 7.70492792e-01
7.30231225e-01 9.18881178e-01 9.13085520e-01 -4.18829948e-01
-7.13113606e-01 -7.32656658e-01 -9.27508235e-01 3.70383739e-01
2.21416786e-01 -6.88519180e-01 -5.89258373e-01 -4.79129404e-01] | [7.239721298217773, 2.6841156482696533] |
f62b79cb-065c-4319-a501-d05133613c25 | weakly-supervised-domain-adaption-for-aspect | 2006.09235 | null | https://arxiv.org/abs/2006.09235v1 | https://arxiv.org/pdf/2006.09235v1.pdf | Weakly-supervised Domain Adaption for Aspect Extraction via Multi-level Interaction Transfer | Fine-grained aspect extraction is an essential sub-task in aspect based opinion analysis. It aims to identify the aspect terms (a.k.a. opinion targets) of a product or service in each sentence. However, expensive annotation process is usually involved to acquire sufficient token-level labels for each domain. To address this limitation, some previous works propose domain adaptation strategies to transfer knowledge from a sufficiently labeled source domain to unlabeled target domains. But due to both the difficulty of fine-grained prediction problems and the large domain gap between domains, the performance remains unsatisfactory. This work conducts a pioneer study on leveraging sentence-level aspect category labels that can be usually available in commercial services like review sites to promote token-level transfer for the extraction purpose. Specifically, the aspect category information is used to construct pivot knowledge for transfer with assumption that the interactions between sentence-level aspect category and token-level aspect terms are invariant across domains. To this end, we propose a novel multi-level reconstruction mechanism that aligns both the fine-grained and coarse-grained information in multiple levels of abstractions. Comprehensive experiments demonstrate that our approach can fully utilize sentence-level aspect category labels to improve cross-domain aspect extraction with a large performance gain. | ['Tao Liang', 'Wenya Wang', 'Fengmao Lv'] | 2020-06-16 | null | null | null | null | ['aspect-extraction'] | ['natural-language-processing'] | [ 2.34606639e-01 -2.19472535e-02 -6.76449776e-01 -7.31702089e-01
-1.30656946e+00 -8.93715739e-01 5.56901753e-01 2.42566228e-01
2.44645141e-02 6.12831950e-01 1.31476492e-01 -2.23978221e-01
2.46245876e-01 -9.25848722e-01 -6.60005987e-01 -5.21997392e-01
5.25971234e-01 5.61717093e-01 1.22119159e-01 -4.03889835e-01
1.64308667e-01 -1.13845719e-02 -1.22270310e+00 7.35104918e-01
1.11090267e+00 1.13531423e+00 -1.25270173e-01 3.47486027e-02
-8.43305409e-01 4.87730324e-01 -6.87092900e-01 -7.95091689e-01
7.94255659e-02 -3.18068236e-01 -8.14198971e-01 3.82079631e-01
1.29874855e-01 9.39407051e-02 5.01933694e-01 1.24018741e+00
1.83535576e-01 -2.44265839e-01 9.30179596e-01 -1.19323587e+00
-8.30983520e-01 7.03468561e-01 -9.30269957e-01 -1.67988185e-02
2.30096400e-01 -1.10098504e-01 1.44621432e+00 -8.83565307e-01
5.08450270e-01 1.09609902e+00 5.48473239e-01 2.92116761e-01
-8.81701589e-01 -7.46686041e-01 7.23569870e-01 9.67635214e-02
-1.13798237e+00 -1.76988423e-01 1.04176307e+00 -3.92195314e-01
9.77028966e-01 -8.54793489e-02 5.49799442e-01 9.56478119e-01
1.67702883e-01 9.16291356e-01 1.24846947e+00 -3.10038120e-01
2.29319528e-01 6.58308446e-01 5.16962528e-01 4.27416921e-01
2.63227105e-01 -4.28120464e-01 -6.46619678e-01 -2.27073759e-01
3.93283546e-01 -1.68360487e-01 1.08942509e-01 -2.39144117e-01
-9.25479650e-01 9.52427864e-01 1.59144193e-01 4.25518632e-01
-4.86908138e-01 -5.07992506e-01 6.75899744e-01 5.19839287e-01
1.06303167e+00 3.83950710e-01 -1.27168465e+00 -1.79894790e-01
-6.56930625e-01 2.53073331e-02 9.86007392e-01 1.33155382e+00
1.07512712e+00 -7.51885623e-02 -1.62026390e-01 9.37956750e-01
3.00137877e-01 6.61981702e-01 4.89015013e-01 -3.70961636e-01
7.25022793e-01 1.09656513e+00 -1.35491922e-01 -8.14642906e-01
7.21703768e-02 -6.23160601e-01 -6.38936460e-01 -2.40566194e-01
5.67066297e-02 -1.98696569e-01 -9.85941231e-01 1.53776670e+00
7.06149220e-01 -1.37147963e-01 9.36394185e-03 6.42543435e-01
8.93133759e-01 5.21996677e-01 4.04854536e-01 -1.06148563e-01
2.09498215e+00 -1.06827211e+00 -6.05237246e-01 -5.19801021e-01
5.91860473e-01 -8.61604512e-01 1.26883781e+00 1.31205112e-01
-5.73380113e-01 -4.80908513e-01 -9.02348518e-01 1.50609881e-01
-6.72626376e-01 1.56429097e-01 8.37796986e-01 6.27324164e-01
-4.78193700e-01 -2.61310302e-02 -5.59533000e-01 -1.27107844e-01
4.49384063e-01 2.33027354e-01 -4.25059229e-01 -1.47762239e-01
-1.42623353e+00 6.35587335e-01 1.94623291e-01 -2.39528000e-01
-6.04105473e-01 -9.27839100e-01 -1.10227001e+00 7.04064369e-02
4.58855361e-01 -8.65233362e-01 1.36897290e+00 -1.41166067e+00
-1.25196826e+00 9.95603621e-01 -5.48973262e-01 1.45573644e-02
-2.42775783e-01 -2.47103661e-01 -6.19760990e-01 -6.20871894e-02
5.51931500e-01 2.76816010e-01 1.22177100e+00 -1.22478402e+00
-9.34955001e-01 -6.39251828e-01 5.02474189e-01 4.98048276e-01
-6.45208478e-01 8.31299722e-02 -5.35406351e-01 -8.16687226e-01
-8.56185257e-02 -6.92569971e-01 -2.44569525e-01 -5.20773351e-01
-2.38244250e-01 -5.63540697e-01 6.09842420e-01 -5.00032485e-01
1.00898015e+00 -1.99301553e+00 -4.68030609e-02 -2.82587670e-02
1.54587984e-01 1.41945004e-01 -3.51338387e-01 2.27687210e-01
-7.62280300e-02 -1.14964172e-02 -2.77331173e-01 -2.55139828e-01
1.04707941e-01 5.45482561e-02 -5.12614727e-01 -1.79667352e-03
5.30266166e-01 1.05820179e+00 -7.49060810e-01 -5.96353173e-01
-2.21669212e-01 3.31419051e-01 -5.38890064e-01 1.63860500e-01
-3.73173565e-01 3.56692433e-01 -1.01584089e+00 1.01727772e+00
8.17062497e-01 -3.49678338e-01 3.49817798e-04 -6.07152462e-01
2.82474577e-01 7.19875693e-01 -8.59656334e-01 1.56045818e+00
-9.35634971e-01 1.41648784e-01 5.96544147e-02 -1.21073747e+00
9.39468682e-01 3.85265857e-01 4.91729289e-01 -5.59297085e-01
-2.82585248e-02 1.70130908e-01 -2.59427667e-01 -1.74846381e-01
5.26292562e-01 -8.89031529e-01 -6.26282692e-01 5.21439612e-01
2.65812904e-01 -4.70599532e-01 3.88696007e-02 1.28873110e-01
9.30656195e-01 1.60049751e-01 6.33726656e-01 -1.65623724e-01
7.75336742e-01 4.18635249e-01 8.48794043e-01 2.48902202e-01
-1.38575971e-01 3.86030495e-01 5.23024559e-01 -1.34780586e-01
-7.81373441e-01 -7.89410055e-01 -1.70417145e-01 1.43556082e+00
1.37015849e-01 -5.75894594e-01 -6.64430499e-01 -1.33962059e+00
-6.46924376e-02 6.29769981e-01 -6.81998849e-01 -1.99818984e-01
-2.64358968e-01 -6.76263213e-01 2.34173849e-01 4.53985363e-01
5.82359910e-01 -1.05463386e+00 2.64376998e-01 1.52917355e-01
-3.85556817e-01 -1.25313485e+00 -5.32569647e-01 1.96638823e-01
-7.95880139e-01 -8.64404559e-01 -5.00782490e-01 -8.73974800e-01
7.54173636e-01 2.94206738e-01 1.67694128e+00 -4.10891652e-01
2.78712422e-01 2.36084774e-01 -7.14364409e-01 -5.46339095e-01
-2.71837771e-01 4.55465794e-01 -1.45287722e-01 5.87445572e-02
1.25015950e+00 -5.73394001e-01 -3.95892411e-01 3.03051978e-01
-9.83764589e-01 -2.79237598e-01 9.13766265e-01 7.34664738e-01
8.37500632e-01 3.17524523e-01 6.96978211e-01 -1.41886902e+00
7.81370878e-01 -5.51206708e-01 -4.60646063e-01 2.65889496e-01
-7.25220084e-01 -8.60706531e-03 6.65948093e-01 -3.83952916e-01
-1.33318794e+00 -1.23526283e-01 -2.35649794e-01 1.03138462e-01
-3.66555035e-01 7.99987912e-01 -6.03126764e-01 2.78131008e-01
3.11461985e-01 3.86933029e-01 -3.95635903e-01 -4.06938314e-01
5.06901681e-01 1.01415253e+00 -1.35609329e-01 -7.73055792e-01
9.77324486e-01 3.63141507e-01 -4.82280731e-01 -5.85258722e-01
-1.62350333e+00 -7.71374822e-01 -6.74889803e-01 2.22323030e-01
6.95010126e-01 -1.33296335e+00 -7.95553923e-02 3.35307658e-01
-1.13800383e+00 2.20164895e-01 -5.32472610e-01 1.68085217e-01
-2.06200063e-01 2.76898205e-01 -4.77945745e-01 -3.54462147e-01
-5.96150637e-01 -1.04409647e+00 1.61538565e+00 1.15306363e-01
-1.74364746e-01 -1.11967981e+00 1.93977118e-01 6.72254205e-01
2.73063123e-01 -2.56137520e-01 1.08892608e+00 -8.56414199e-01
-3.98921162e-01 -3.53259325e-01 -2.82682151e-01 6.00978494e-01
6.68303072e-01 -3.41048181e-01 -9.39161301e-01 3.60173583e-02
4.48024720e-01 -3.26092899e-01 6.37937725e-01 2.53274292e-01
7.92421043e-01 -3.54713529e-01 -2.81518012e-01 2.23972350e-01
1.27825916e+00 -2.28439290e-02 2.76004851e-01 4.31225538e-01
8.74916077e-01 7.06348658e-01 1.24678791e+00 2.94895887e-01
7.33100176e-01 5.67856789e-01 -1.93325981e-01 -5.99312857e-02
-1.44034773e-01 -2.62035310e-01 6.30067289e-01 1.26006985e+00
1.48106128e-01 2.49952842e-02 -5.75822830e-01 8.73371601e-01
-1.53262460e+00 -6.56416416e-01 1.43780917e-01 1.77985334e+00
1.27277768e+00 3.37813079e-01 -2.60379212e-03 -2.46945068e-01
5.81471145e-01 2.60268897e-01 -7.16658354e-01 -4.34085310e-01
-7.70406052e-02 1.75592631e-01 1.34182036e-01 2.50338018e-01
-1.31547141e+00 1.18565238e+00 5.14009714e+00 1.31129861e+00
-8.90401363e-01 4.28045928e-01 5.06959856e-01 3.70635271e-01
-8.12102616e-01 7.97248781e-02 -1.14635909e+00 3.14271092e-01
8.55620682e-01 -3.44470054e-01 -1.62584126e-01 1.23973656e+00
-8.86812657e-02 2.44206190e-01 -9.97276127e-01 6.44064665e-01
1.89230666e-01 -9.31079865e-01 3.73051822e-01 1.19824065e-02
9.52164650e-01 -2.83539407e-02 9.98157412e-02 7.53300786e-01
3.18955570e-01 -6.58931732e-01 4.03936148e-01 -1.47677302e-01
7.48347759e-01 -8.15814495e-01 9.84230280e-01 2.23468095e-01
-1.58206677e+00 3.23202491e-01 -4.26082551e-01 5.60524166e-02
2.18340740e-01 1.21348381e+00 -7.44846821e-01 7.66158462e-01
5.33289552e-01 1.08296549e+00 -2.82420009e-01 3.92868221e-01
-5.16149282e-01 7.64422059e-01 -7.39523321e-02 3.49979214e-02
2.72400767e-01 -3.76785964e-01 3.94891381e-01 1.13916349e+00
1.36925429e-01 1.39526762e-02 2.01922685e-01 5.86005986e-01
-2.37876207e-01 4.26870435e-01 -7.18566656e-01 -2.13469625e-01
1.90442160e-01 1.56840396e+00 -5.35955012e-01 -5.06869674e-01
-8.05279374e-01 9.98800993e-01 2.56542385e-01 3.22285831e-01
-7.97554910e-01 -3.84599030e-01 1.18696213e+00 2.46389136e-02
7.67852902e-01 -7.31352568e-02 -6.85057044e-01 -1.60130632e+00
3.07146341e-01 -1.19431162e+00 4.25173044e-01 -3.32189500e-01
-1.83130956e+00 8.60198677e-01 -2.06843421e-01 -1.67169285e+00
-2.37322122e-01 -5.44227123e-01 -4.08930242e-01 8.82602930e-01
-2.03455830e+00 -1.66675353e+00 6.20609969e-02 6.80447698e-01
9.19610679e-01 -2.59588003e-01 9.23409283e-01 3.38388264e-01
-1.86239749e-01 7.69518852e-01 -2.38704413e-01 1.35475859e-01
8.46617520e-01 -1.34812737e+00 4.41620618e-01 6.01829410e-01
1.98742107e-01 9.21177745e-01 5.04771769e-01 -7.61869550e-01
-1.19372177e+00 -1.22894764e+00 1.16975474e+00 -8.04406166e-01
8.60858917e-01 -4.16349411e-01 -9.03911471e-01 6.46033823e-01
2.73875892e-01 -2.46887445e-01 1.15064967e+00 7.98052549e-01
-7.67168701e-01 -3.65864098e-01 -1.06695914e+00 3.46181720e-01
6.81293309e-01 -9.08447206e-01 -1.03551507e+00 2.85578728e-01
1.12873840e+00 -1.11049950e-01 -1.13698554e+00 5.33793867e-01
3.21949959e-01 -5.56634665e-01 7.37601757e-01 -7.25832820e-01
7.00338364e-01 -4.05134290e-01 -1.68387353e-01 -1.46137702e+00
-1.24064207e-01 -9.82406661e-02 -5.04707098e-02 1.89593625e+00
8.18314552e-01 -6.31186962e-01 7.41390467e-01 5.82258582e-01
4.62551787e-02 -7.06619263e-01 -7.15315461e-01 -6.19276762e-01
2.68596351e-01 -6.18150473e-01 8.42786372e-01 8.94581199e-01
2.72163525e-02 1.05057573e+00 -1.45881372e-02 3.22948873e-01
4.88514811e-01 9.26870048e-01 5.13742447e-01 -1.06618142e+00
-3.26930374e-01 -1.45995185e-01 -3.39567274e-01 -1.18835354e+00
5.31245232e-01 -8.35975468e-01 1.48355216e-01 -1.46081877e+00
4.51426059e-01 -6.23490632e-01 -5.33349037e-01 5.54234922e-01
-5.04273772e-01 2.27447122e-01 -2.04455987e-01 9.20974165e-02
-8.26721907e-01 8.68077993e-01 1.38041413e+00 -4.75956142e-01
-3.92930508e-02 3.85439396e-01 -1.38167739e+00 8.31906021e-01
8.03884983e-01 -6.10529900e-01 -6.28259897e-01 -4.00319397e-01
3.03070992e-01 -3.83867025e-01 -2.07004964e-01 -4.29208040e-01
-8.94501954e-02 -1.86037704e-01 -1.71710923e-02 -6.07175350e-01
2.32450351e-01 -1.00768375e+00 -3.93491358e-01 -9.28374380e-02
-1.57390133e-01 -1.51270807e-01 1.74823821e-01 6.59195960e-01
-9.25764322e-01 -9.78172868e-02 4.52844292e-01 -2.36164868e-01
-8.83679569e-01 4.81638849e-01 -1.54515266e-01 3.81793976e-01
8.44068825e-01 6.22232780e-02 -1.54588044e-01 -2.70605952e-01
-3.96422178e-01 -2.58737467e-02 3.74912262e-01 6.18312001e-01
4.05680180e-01 -1.39541662e+00 -5.83148718e-01 2.49915823e-01
6.68202996e-01 3.38466503e-02 2.84031093e-01 6.25892520e-01
2.69655436e-01 5.43537855e-01 1.14672080e-01 -4.48500097e-01
-1.18883646e+00 5.14992058e-01 -7.78201371e-02 -7.75733113e-01
-1.99092060e-01 9.45467710e-01 6.74998224e-01 -1.00605059e+00
-2.72310495e-01 -2.25147486e-01 -4.08268571e-01 2.50400662e-01
4.81717318e-01 -3.99552613e-01 3.27320039e-01 -5.99954784e-01
-5.20887256e-01 9.42851186e-01 -3.88208061e-01 -5.08712791e-03
1.42587602e+00 -4.30289954e-01 -4.02627468e-01 4.23504114e-01
1.24095225e+00 2.25820810e-01 -8.35868001e-01 -5.41956007e-01
2.24268194e-02 -2.00608611e-01 -6.88230246e-02 -8.55375707e-01
-1.15030527e+00 7.97092795e-01 1.15137391e-01 1.53567329e-01
1.35638428e+00 4.75031734e-01 1.03748369e+00 2.55287081e-01
4.74379927e-01 -1.01840889e+00 -1.11421511e-01 6.66345477e-01
5.80967069e-01 -1.44669211e+00 -9.12847817e-02 -9.05756116e-01
-1.04379988e+00 6.09818697e-01 8.05976272e-01 1.18593775e-01
9.65786994e-01 2.51284063e-01 4.14155424e-01 -3.96521389e-01
-7.63953030e-01 -4.34023708e-01 5.41684866e-01 7.83242345e-01
6.77430809e-01 1.92036092e-01 -3.09451818e-01 1.16073060e+00
-5.94057404e-02 -1.16647281e-01 1.16352178e-01 8.11950922e-01
-4.11332458e-01 -1.57888889e+00 -2.26491783e-02 6.50004506e-01
-6.51077807e-01 -5.64272761e-01 -3.11062574e-01 3.77687901e-01
1.97456509e-01 9.54266250e-01 -3.42629135e-01 -2.59863049e-01
5.32413363e-01 1.05825558e-01 6.74434826e-02 -1.12202382e+00
-6.58175290e-01 4.84547839e-02 3.69354486e-01 -4.23474014e-01
-7.40218639e-01 -6.60372972e-01 -9.60408449e-01 -2.11527161e-02
-3.08404952e-01 5.65571368e-01 5.84790587e-01 1.31139541e+00
5.53798676e-01 5.05669117e-01 7.21920371e-01 -2.73511559e-01
-4.60373640e-01 -9.69820023e-01 -9.12333488e-01 4.67710048e-01
2.24896014e-01 -5.99577427e-01 -1.96576566e-01 2.36125007e-01] | [11.38651180267334, 6.676271915435791] |
8e1ed9dd-1bc4-43e7-b2e4-dfa60f85a742 | octis-comparing-and-optimizing-topic-models | null | null | https://aclanthology.org/2021.eacl-demos.31 | https://aclanthology.org/2021.eacl-demos.31.pdf | OCTIS: Comparing and Optimizing Topic models is Simple! | In this paper, we present OCTIS, a framework for training, analyzing, and comparing Topic Models, whose optimal hyper-parameters are estimated using a Bayesian Optimization approach. The proposed solution integrates several state-of-the-art topic models and evaluation metrics. These metrics can be targeted as objective by the underlying optimization procedure to determine the best hyper-parameter configuration. OCTIS allows researchers and practitioners to have a fair comparison between topic models of interest, using several benchmark datasets and well-known evaluation metrics, to integrate novel algorithms, and to have an interactive visualization of the results for understanding the behavior of each model. The code is available at the following link: https://github.com/MIND-Lab/OCTIS. | ['Antonio Candelieri', 'Pietro Tropeano', 'Bruno Giovanni Galuzzi', 'Elisabetta Fersini', 'Silvia Terragni'] | 2021-04-19 | null | null | null | eacl-2021-2 | ['bayesian-optimisation'] | ['methodology'] | [-4.26773161e-01 2.91652419e-02 -5.21999359e-01 -3.97400558e-01
-8.32601964e-01 -4.75428462e-01 7.55430520e-01 5.39743900e-01
-1.73605040e-01 4.51581031e-01 -1.27395550e-02 -2.00559333e-01
-3.80994946e-01 -6.07922375e-01 -3.84436071e-01 -5.86054444e-01
-1.88972518e-01 7.84456074e-01 5.04869699e-01 1.78853124e-01
6.57074034e-01 1.38178825e-01 -1.71280980e+00 -5.95754161e-02
9.52508152e-01 7.88488030e-01 2.53540635e-01 3.69692504e-01
-1.47538772e-02 -8.18727016e-02 -6.37235343e-01 -3.12514693e-01
-1.18528090e-01 5.56440838e-03 -6.46799564e-01 -3.36633444e-01
2.48505905e-01 1.34857893e-01 7.21915737e-02 1.03339553e+00
2.12688670e-01 2.52442509e-01 8.32855642e-01 -1.29201734e+00
-6.39668778e-02 7.72557259e-01 -1.05486222e-01 3.53248060e-01
3.30962270e-01 5.48719689e-02 9.20824289e-01 -8.34944963e-01
5.99873900e-01 1.24435771e+00 3.97829786e-02 5.16277328e-02
-1.06856871e+00 -6.75737679e-01 2.07561865e-01 4.79139566e-01
-1.67552292e+00 -2.56204605e-01 6.27184570e-01 -5.57509184e-01
6.65201426e-01 3.20226341e-01 5.99578798e-01 1.00018072e+00
3.39893311e-01 6.52224064e-01 1.12271178e+00 -4.15148348e-01
5.16622484e-01 6.18552029e-01 7.10418105e-01 5.37875354e-01
3.58645856e-01 7.37443566e-02 -6.62281573e-01 -4.35145915e-01
2.70669788e-01 -2.86202401e-01 -3.21425319e-01 -3.80282938e-01
-1.04577816e+00 9.17953789e-01 1.86350435e-01 4.07275736e-01
-4.09929782e-01 -4.18976657e-02 1.74149618e-01 -1.24970570e-01
6.86997414e-01 5.50428271e-01 -3.35440069e-01 -4.29201126e-01
-1.07395637e+00 5.16490996e-01 9.79250252e-01 7.68468618e-01
7.20607758e-01 -3.84765238e-01 -3.79209757e-01 8.67992759e-01
8.01274776e-01 2.63653278e-01 5.84859729e-01 -9.21190202e-01
1.31687298e-01 5.55230558e-01 1.90948546e-01 -7.65766203e-01
-3.55078250e-01 -5.37755251e-01 -2.63790905e-01 -2.87557915e-02
3.01130563e-01 -7.01950258e-03 -6.80109799e-01 1.38915980e+00
5.66378951e-01 4.05249387e-01 -1.77874386e-01 7.80535042e-01
7.97941804e-01 9.11158562e-01 8.34899694e-02 -1.15935825e-01
1.64097226e+00 -8.03411901e-01 -6.20885849e-01 2.46079993e-02
4.02054757e-01 -8.86656404e-01 1.31908941e+00 5.31061947e-01
-1.09605634e+00 -1.84664220e-01 -1.09827435e+00 3.60193014e-01
-6.58111751e-01 2.06332967e-01 2.42644727e-01 7.13145852e-01
-8.38911951e-01 5.16970932e-01 -1.08620358e+00 -5.89821339e-01
5.29408753e-02 1.38922989e-01 3.63778085e-01 1.26735464e-01
-1.28426075e+00 1.01331663e+00 6.96281850e-01 -2.45359808e-01
-1.08203173e+00 -9.53196645e-01 -4.66315717e-01 3.28290284e-01
6.08707190e-01 -6.82938576e-01 1.41793263e+00 -2.38490075e-01
-1.48276162e+00 6.02754176e-01 -2.24458575e-01 -3.62691879e-01
2.82094300e-01 -4.22237098e-01 -1.94225594e-01 1.20164111e-01
-1.42651558e-01 4.42720383e-01 5.72818995e-01 -1.30048764e+00
-3.54187399e-01 -4.12455171e-01 -1.25369012e-01 2.21357986e-01
-4.70027506e-01 3.61908376e-01 -9.79731262e-01 -4.87670481e-01
-4.34052162e-02 -8.50068152e-01 -1.47566885e-01 -2.79864669e-01
-5.37848711e-01 -3.74191880e-01 7.49491692e-01 -5.08911610e-01
1.66358650e+00 -1.92678607e+00 1.40893459e-01 4.17360485e-01
-7.74009153e-02 1.49015412e-01 1.11592777e-01 5.39083838e-01
2.68084109e-01 3.41578960e-01 -1.16678253e-01 -4.47228044e-01
2.02061266e-01 -1.82920620e-01 -1.11122422e-01 2.32972026e-01
-1.42376885e-01 2.31565133e-01 -6.58617854e-01 -5.65732121e-01
5.13492405e-01 5.23372829e-01 -2.73304015e-01 4.39938277e-01
-5.69945812e-01 1.48951501e-01 -6.14059269e-01 4.57813531e-01
5.63194811e-01 -3.74141306e-01 2.99775183e-01 -1.36446163e-01
-1.49797410e-01 5.59706986e-01 -1.42958224e+00 1.31951010e+00
-3.18549186e-01 6.52924240e-01 -2.23979995e-01 -7.10579574e-01
1.00485754e+00 2.09705308e-01 4.25480515e-01 -2.61125952e-01
3.84546459e-01 6.81582838e-02 -2.62515038e-01 -2.82264262e-01
6.42000437e-01 4.87976789e-01 1.91624239e-01 6.35494590e-01
1.65347517e-01 -1.98459908e-01 6.97610676e-01 3.52321059e-01
6.25678182e-01 -4.14438248e-02 3.15142184e-01 -5.72420955e-01
4.11751807e-01 -1.00770928e-01 1.99578971e-01 7.56189287e-01
2.29332015e-01 3.13648194e-01 5.63567579e-01 -1.04137838e-01
-6.88818038e-01 -1.08375347e+00 -6.12725616e-01 1.05521560e+00
9.29483920e-02 -9.01522398e-01 -8.60821545e-01 -2.27626443e-01
-1.58131108e-01 1.20216870e+00 -5.27423561e-01 8.87007415e-02
-6.12365529e-02 -7.98109353e-01 2.30684951e-01 1.05490655e-01
1.78333536e-01 -8.31693709e-01 -8.34359288e-01 -2.35386416e-02
-2.81882256e-01 -7.49538898e-01 -1.62668958e-01 5.98932169e-02
-1.05683804e+00 -9.72085714e-01 -6.22789025e-01 -1.12428248e-01
3.36756051e-01 -1.22729847e-02 1.33579588e+00 1.30583927e-01
-1.85786158e-01 4.59620029e-01 -4.11402404e-01 -7.08161592e-01
-5.44476330e-01 4.90513921e-01 -1.99835062e-01 -3.01642507e-01
2.78445244e-01 -4.62329894e-01 -6.10898733e-01 6.77022815e-01
-8.53132427e-01 -2.05107048e-01 1.05715238e-01 3.65877330e-01
5.38934052e-01 -5.57729974e-02 1.06406026e-01 -6.23692513e-01
8.58028710e-01 -7.73415565e-01 -1.22414005e+00 5.19739211e-01
-1.12170839e+00 1.03106750e-02 -5.74556813e-02 -3.24055195e-01
-9.21316385e-01 -6.59623206e-01 2.50196248e-01 -3.28091741e-01
-2.74976462e-01 8.59821439e-01 -1.14915431e-01 2.62105614e-01
6.50868118e-01 -6.57777563e-02 -1.30975306e-01 -5.98732352e-01
4.10680741e-01 6.67920530e-01 -5.87740242e-02 -7.68692315e-01
3.25604647e-01 1.40693665e-01 -3.97753894e-01 -7.75855482e-01
-8.03494990e-01 -7.90642440e-01 -4.22597438e-01 -4.20558006e-01
2.95984507e-01 -5.86999476e-01 -4.41530883e-01 2.97884852e-01
-9.71554995e-01 -4.05496299e-01 1.26091719e-01 6.42842114e-01
-4.58289385e-01 1.54285148e-01 -1.30081058e-01 -7.19986200e-01
-3.97800952e-01 -1.23461223e+00 8.91305685e-01 4.93371308e-01
-4.52976346e-01 -1.19783854e+00 3.89531821e-01 2.88728207e-01
4.41482753e-01 -3.25331628e-01 9.13001478e-01 -8.92644227e-01
-6.65791214e-01 -1.69837922e-01 -4.07139622e-02 8.62960070e-02
-1.98536113e-01 6.62955403e-01 -8.42931092e-01 -1.70792416e-01
-1.82741240e-01 3.12446859e-02 6.86439693e-01 7.70857930e-01
1.25972843e+00 -2.25322977e-01 -6.83305025e-01 4.78908300e-01
1.35051763e+00 1.32905722e-01 5.66820979e-01 5.75470626e-01
-2.82965787e-02 4.70151484e-01 8.72713029e-01 6.94974661e-01
5.64626217e-01 1.08972812e+00 4.72980738e-01 3.09276909e-01
2.91206777e-01 6.56993315e-02 1.24290973e-01 6.77840531e-01
5.84125631e-02 -5.32151043e-01 -1.37328207e+00 4.28810477e-01
-1.95112920e+00 -6.18953347e-01 2.36745756e-02 2.43680048e+00
6.00929737e-01 1.69554457e-01 3.56371582e-01 -2.34857932e-01
7.09260106e-01 1.79030240e-01 -2.96531111e-01 -3.25039715e-01
3.47949475e-01 5.83688803e-02 2.83747554e-01 5.34670293e-01
-1.14715433e+00 8.89137924e-01 6.26779556e+00 9.22074854e-01
-1.11205065e+00 1.84556499e-01 5.47348738e-01 -2.98860431e-01
-1.85804829e-01 1.85833842e-01 -9.84009802e-01 5.13840318e-01
1.40462768e+00 -7.59582222e-01 3.27974647e-01 8.52685988e-01
5.52649617e-01 -4.66709167e-01 -8.47539425e-01 5.01981199e-01
-1.40740806e-02 -1.32630181e+00 -9.49713439e-02 7.10673705e-02
4.72092271e-01 1.85833618e-01 1.80993617e-01 2.05988884e-01
2.66857266e-01 -6.81907177e-01 5.96183479e-01 6.26680374e-01
5.89235239e-02 -5.63029766e-01 6.80343151e-01 1.23910137e-01
-7.93431580e-01 -5.94090037e-02 -2.73658425e-01 4.55647647e-01
1.66675642e-01 7.94772923e-01 -8.13383520e-01 6.80707872e-01
9.47393656e-01 4.49186116e-01 -6.87658668e-01 1.69949245e+00
-3.16264123e-01 1.07972324e+00 -5.92298150e-01 -3.48505855e-01
1.04077244e-02 -3.13890040e-01 7.58699238e-01 1.20176840e+00
4.74071175e-01 -1.04295820e-01 1.80738360e-01 1.08272386e+00
2.59581029e-01 5.47152281e-01 -2.03386873e-01 -1.84495877e-02
8.66166651e-01 1.30178940e+00 -9.22784269e-01 -4.27636296e-01
-2.03597713e-02 1.13355974e-02 2.15222575e-02 2.96369195e-01
-9.19216335e-01 -2.48779297e-01 5.74492812e-01 1.34028196e-01
1.51439205e-01 -2.93898970e-01 -3.20602059e-01 -9.06540394e-01
-1.26863187e-02 -9.59556937e-01 5.62650979e-01 -8.42301250e-01
-6.89601600e-01 6.20882988e-01 1.01339078e+00 -9.50336814e-01
-3.45668465e-01 -4.97344673e-01 -9.74659204e-01 5.60926676e-01
-1.18310654e+00 -8.00055981e-01 -5.18774271e-01 2.72945225e-01
4.16106433e-01 -8.37669596e-02 6.55101120e-01 6.14418872e-02
-8.02940547e-01 3.48592192e-01 2.74730414e-01 -4.66502905e-01
7.24690318e-01 -1.05230534e+00 8.57930705e-02 6.21085286e-01
2.27195889e-01 6.70197308e-01 1.15738750e+00 -5.10090888e-01
-8.87663960e-01 -7.13671327e-01 4.93902862e-01 -2.46050447e-01
7.21410692e-01 -1.84486672e-01 -1.06280649e+00 3.83836299e-01
4.09965634e-01 -7.30350077e-01 8.02389026e-01 6.79491460e-01
-6.59871250e-02 -7.80341169e-03 -7.44833052e-01 5.99043548e-01
2.57724196e-01 1.59094110e-03 -4.18438077e-01 5.43049812e-01
3.69618416e-01 -3.16678494e-01 -1.03883111e+00 2.81873524e-01
5.86852193e-01 -9.77346659e-01 7.91246176e-01 -2.40403563e-01
2.80996323e-01 -8.89446139e-02 -5.28417900e-02 -1.36445153e+00
6.38606027e-02 -4.97523338e-01 -3.27684581e-01 1.28758478e+00
7.91653335e-01 -7.82242417e-01 5.53879380e-01 4.71752256e-01
5.63498996e-02 -1.01788294e+00 -7.06188023e-01 -7.01727271e-01
3.78420502e-02 -6.23341143e-01 7.28137851e-01 5.04956722e-01
-4.09364700e-03 8.24969113e-02 1.05102658e-01 4.15287554e-01
6.13738537e-01 9.16153714e-02 7.82650828e-01 -1.41578484e+00
5.49429609e-03 -8.18314195e-01 -1.47631809e-01 -6.30833626e-01
6.44808263e-03 -6.47168040e-01 -1.12555601e-01 -1.41858912e+00
1.65556490e-01 -4.56811368e-01 -4.49609518e-01 2.41105467e-01
-2.76032746e-01 -2.14049533e-01 3.97125989e-01 2.90462881e-01
-7.02495933e-01 7.35616446e-01 6.91340685e-01 -4.54642214e-02
-3.34198415e-01 3.68994862e-01 -5.40066421e-01 7.21758723e-01
1.10796618e+00 -8.49224985e-01 -4.43855137e-01 -6.64032027e-02
1.79135397e-01 -1.91534832e-01 4.02062804e-01 -1.02902782e+00
3.18950981e-01 -2.49563754e-01 -2.06785649e-01 -7.59664237e-01
4.59012806e-01 -5.16889751e-01 1.77691951e-01 2.07015783e-01
-3.29256237e-01 5.07350489e-02 4.03828204e-01 2.08547220e-01
-2.08060980e-01 -7.87880123e-01 7.00746179e-01 7.97878653e-02
-4.69168097e-01 7.06121325e-02 -5.00655234e-01 5.44542484e-02
1.01003766e+00 1.33165950e-02 -4.74701643e-01 -4.44951236e-01
-7.34859109e-01 4.81797338e-01 3.56989741e-01 5.09466708e-01
4.29938704e-01 -7.60311961e-01 -6.07801676e-01 -8.48165974e-02
1.78523719e-01 -3.56531650e-01 2.18354896e-01 7.45423496e-01
-3.59720111e-01 5.33908546e-01 -1.53759611e-03 -8.25296044e-01
-1.43321395e+00 3.22949469e-01 3.42560261e-01 -6.27633095e-01
-7.84290507e-02 4.42097962e-01 -3.78979556e-02 -3.39441270e-01
3.80239278e-01 -2.54151702e-01 -2.38864064e-01 1.71120763e-01
3.37722123e-01 5.42138815e-01 2.25318581e-01 -1.53178647e-01
-3.59319210e-01 3.92213196e-01 -1.60144448e-01 -3.27395409e-01
1.20876122e+00 -6.28667101e-02 6.92364424e-02 5.97775638e-01
7.43239284e-01 -4.29156572e-01 -8.31284285e-01 -2.06618398e-01
1.78464264e-01 -3.33476990e-01 3.96502227e-01 -9.03706133e-01
-7.40423739e-01 8.00733984e-01 9.12718713e-01 4.06596243e-01
8.93779635e-01 1.66754976e-01 -1.70090273e-02 2.64681071e-01
1.72630832e-01 -1.10090041e+00 1.46531269e-01 4.61957067e-01
8.85721087e-01 -9.97215927e-01 4.05704021e-01 -4.94627595e-01
-5.18657386e-01 1.02508867e+00 4.92360562e-01 1.05793685e-01
1.24145520e+00 2.23649994e-01 1.77336365e-01 -4.32171136e-01
-1.16558778e+00 -3.36169191e-02 5.65255165e-01 3.02905053e-01
5.37963569e-01 9.50078368e-02 -4.87084091e-01 3.47870439e-01
-2.71149397e-01 -1.34417444e-01 3.44660997e-01 6.48536265e-01
-4.40392792e-01 -1.29571319e+00 -5.73783636e-01 4.48760182e-01
-5.00213206e-01 3.72501984e-02 -1.27146348e-01 9.50826883e-01
-4.43624943e-01 9.15342093e-01 2.94258744e-02 -1.07351895e-02
1.52163014e-01 1.85809657e-01 4.86057587e-02 -5.82060277e-01
-4.87307519e-01 2.43220493e-01 1.00555725e-01 -5.55830598e-01
-3.36275816e-01 -7.89909363e-01 -7.16820538e-01 -1.52243167e-01
-5.71732402e-01 4.65611190e-01 1.13786185e+00 7.04527795e-01
5.83651125e-01 4.60646063e-01 3.93155038e-01 -8.29969347e-01
-5.54351926e-01 -1.29376853e+00 -3.14652562e-01 -3.04236561e-02
-5.21253169e-01 -1.00324321e+00 -3.63366276e-01 -1.43357530e-01] | [10.415432929992676, 6.952080726623535] |
87a13781-0699-41f9-9b16-0ba4b9b29946 | vgstore-a-multimodal-extension-to-sparql-for | 2209.02981 | null | https://arxiv.org/abs/2209.02981v1 | https://arxiv.org/pdf/2209.02981v1.pdf | VGStore: A Multimodal Extension to SPARQL for Querying RDF Scene Graph | Semantic Web technology has successfully facilitated many RDF models with rich data representation methods. It also has the potential ability to represent and store multimodal knowledge bases such as multimodal scene graphs. However, most existing query languages, especially SPARQL, barely explore the implicit multimodal relationships like semantic similarity, spatial relations, etc. We first explored this issue by organizing a large-scale scene graph dataset, namely Visual Genome, in the RDF graph database. Based on the proposed RDF-stored multimodal scene graph, we extended SPARQL queries to answer questions containing relational reasoning about color, spatial, etc. Further demo (i.e., VGStore) shows the effectiveness of customized queries and displaying multimodal data. | ['Lei Zou', 'Wenjuan Han', 'Zilong Zheng', 'Yanzeng Li'] | 2022-09-07 | null | null | null | null | ['relational-reasoning'] | ['natural-language-processing'] | [-4.97223496e-01 2.77181953e-01 -1.44402936e-01 -5.03376722e-01
-2.94266403e-01 -7.06435740e-01 6.34170890e-01 7.46431649e-01
-2.01284781e-01 7.03241229e-01 5.77163756e-01 -8.28624815e-02
-5.29194117e-01 -1.36595404e+00 -5.56849301e-01 -1.98936880e-01
-7.26349943e-04 3.93608302e-01 7.17613399e-01 -6.98849380e-01
2.72595078e-01 4.61479783e-01 -2.18223667e+00 5.33718824e-01
1.03698087e+00 9.14088428e-01 1.73328310e-01 1.97046608e-01
-8.75730634e-01 1.23863316e+00 -3.89353186e-01 -8.75718832e-01
-5.70771158e-01 1.22438692e-01 -9.30075884e-01 -3.41255397e-01
4.62686270e-01 -2.58144736e-01 -5.06351590e-01 1.16501760e+00
6.00658476e-01 2.67000645e-01 -4.76435944e-02 -1.57412767e+00
-1.02736092e+00 5.49637437e-01 4.50906418e-02 -2.84004986e-01
1.19544172e+00 -3.47427875e-01 7.48302400e-01 -5.34214139e-01
1.22984242e+00 1.51882279e+00 2.23006636e-01 2.06108302e-01
-5.45032978e-01 -1.51962906e-01 -9.92026627e-02 5.64040959e-01
-1.65603673e+00 -8.10482502e-02 6.88585699e-01 -6.80031022e-03
1.04816806e+00 7.60213017e-01 8.30824375e-01 5.46870768e-01
-2.79448241e-01 4.64149266e-01 9.14633453e-01 -3.19047838e-01
1.23345092e-01 2.37776995e-01 1.02647290e-01 9.79729056e-01
2.92610168e-01 -5.84937692e-01 -7.91706800e-01 -3.61062251e-02
4.86927509e-01 -6.77234754e-02 -5.70471175e-02 -5.97280502e-01
-1.22471201e+00 6.40824556e-01 6.83446705e-01 2.07507387e-01
-2.47478206e-02 9.20799002e-02 5.63201249e-01 9.12365615e-02
-4.02537808e-02 1.74784869e-01 -5.48985461e-03 1.19819969e-01
-7.92692080e-02 3.21736902e-01 7.55243599e-01 1.44944632e+00
1.26696455e+00 -2.60359079e-01 -6.04306422e-02 8.78832400e-01
6.77610517e-01 7.09509790e-01 1.98076800e-01 -1.12803137e+00
6.29866540e-01 1.44332623e+00 1.77732110e-01 -1.50973308e+00
-5.70346415e-01 4.91467029e-01 -3.14382851e-01 -3.56683254e-01
-2.15036366e-02 2.49945119e-01 -7.71887183e-01 1.44154215e+00
7.79035389e-01 -3.77822340e-01 3.14509064e-01 1.09159684e+00
1.85626352e+00 3.92945200e-01 6.67620003e-01 1.76645532e-01
1.49973297e+00 -3.22053134e-01 -1.06689143e+00 1.99462980e-01
7.81791508e-01 -6.11930013e-01 1.26053321e+00 -1.40047818e-01
-9.25896943e-01 -1.56975955e-01 -7.41490662e-01 -6.48641586e-01
-1.30515218e+00 -3.21096182e-01 1.07131910e+00 6.08386159e-01
-1.16370642e+00 -8.73093754e-02 -3.66540760e-01 -1.11646056e+00
9.26976800e-02 9.62979719e-02 -6.31972492e-01 -4.36367989e-01
-1.47827077e+00 7.02568889e-01 1.02466524e+00 -1.15946092e-01
-3.81461024e-01 -2.95865387e-01 -1.06389761e+00 -8.70256871e-02
7.42177248e-01 -8.02885056e-01 5.04674017e-01 -3.43778938e-01
-9.66991186e-01 9.86920714e-01 -1.77218512e-01 1.00923486e-01
-6.92333803e-02 -1.55102611e-02 -1.10589957e+00 4.45991695e-01
1.74328566e-01 6.17621839e-01 -1.41098246e-01 -1.51194751e+00
-5.70616961e-01 -8.91095817e-01 6.15034163e-01 4.53922570e-01
-2.86618710e-01 2.12113634e-02 -9.37499583e-01 -2.13426240e-02
2.95333922e-01 -4.06733930e-01 1.78116620e-01 -1.92607686e-01
-5.82970142e-01 -4.02386844e-01 8.76796365e-01 -6.25091851e-01
1.48276043e+00 -2.16140842e+00 1.11543544e-01 5.42187572e-01
1.63876653e-01 -1.99191466e-01 -6.90751299e-02 1.14258361e+00
2.97139347e-01 4.42925662e-01 3.40232521e-01 3.44956696e-01
4.17863846e-01 6.10922813e-01 -1.73400164e-01 -3.28992074e-03
-4.86612111e-01 1.03929567e+00 -8.74362946e-01 -1.16857314e+00
2.95492977e-01 4.98020321e-01 -3.98895979e-01 1.62879168e-03
-5.81481457e-01 -3.84569131e-02 -8.13798308e-01 1.01427531e+00
7.20975697e-01 -3.18127632e-01 6.67439103e-01 -6.55009985e-01
-3.17255020e-01 -3.83806139e-01 -1.07325494e+00 2.42502165e+00
-1.21171042e-01 3.26541245e-01 -2.07654923e-01 -4.43114758e-01
9.66280937e-01 1.41788021e-01 6.11469865e-01 -1.38987064e+00
-1.39654055e-01 -5.25777368e-03 -8.39715958e-01 -1.25254238e+00
9.40728486e-01 2.02785447e-01 -2.78831124e-01 -1.32333383e-01
-1.09000817e-01 -1.36371985e-01 4.09095377e-01 7.66058266e-01
6.91051543e-01 5.13861775e-01 9.29545686e-02 3.72597277e-02
4.18261349e-01 4.14919078e-01 -4.93199751e-02 3.46812725e-01
2.35597402e-01 -3.32988687e-02 4.90555644e-01 -4.35385376e-01
-7.71450877e-01 -1.36777174e+00 1.21423118e-02 1.07814670e+00
9.85248923e-01 -1.04138970e+00 -4.06755298e-01 -8.10347870e-02
2.56277639e-02 7.10325897e-01 -2.60923982e-01 2.15375975e-01
-5.49617335e-02 -3.26147616e-01 4.89415675e-01 1.60841137e-01
7.58879960e-01 -1.04004765e+00 -6.33471727e-01 -2.79390097e-01
-4.44220841e-01 -1.17577231e+00 3.62080187e-01 -7.09728777e-01
-6.02168024e-01 -1.43694949e+00 3.75957564e-02 -7.33726263e-01
4.77563322e-01 1.47098601e-01 1.18386853e+00 4.78960156e-01
-3.13545793e-01 1.00148571e+00 -6.86027706e-01 -2.37256274e-01
2.46437177e-01 -2.78850764e-01 -4.64125276e-01 -3.23192179e-01
3.03549588e-01 -2.49206737e-01 -5.52448690e-01 1.63556516e-01
-1.33610427e+00 4.93856370e-02 5.46627678e-03 8.37020651e-02
7.73107529e-01 1.07330799e-01 3.43590617e-01 -8.93424988e-01
6.44177675e-01 -5.90914190e-01 -4.11933899e-01 9.83907700e-01
-3.19937855e-01 7.02638701e-02 -7.46490806e-02 2.26844206e-01
-1.33698833e+00 -2.47735932e-01 3.98154231e-03 -1.34220511e-01
-2.52983242e-01 1.01382780e+00 -8.54041338e-01 -6.96314052e-02
2.80707389e-01 -9.75790396e-02 -1.51144668e-01 -3.65882933e-01
9.50851440e-01 2.88560003e-01 5.02236247e-01 -7.20696867e-01
5.73601127e-01 5.68557382e-01 2.97037244e-01 -9.09023225e-01
-3.54384124e-01 -3.52986395e-01 -4.62327808e-01 -5.38134038e-01
1.29675627e+00 -9.59900141e-01 -1.32132983e+00 -3.43985781e-02
-1.08960092e+00 1.46171972e-01 -2.57017404e-01 3.60533834e-01
-5.61842859e-01 5.15164554e-01 -2.47762769e-01 -6.79741204e-01
2.83203006e-01 -5.48048377e-01 1.01772392e+00 4.04185444e-01
2.01341733e-01 -9.88882720e-01 1.28076270e-01 6.05956018e-01
2.97961712e-01 7.08686769e-01 1.44764781e+00 -4.11048681e-01
-9.45822835e-01 -3.05088516e-02 -6.61658227e-01 -6.84814811e-01
-5.43900765e-02 1.36105210e-01 -5.67682147e-01 2.40380272e-01
-9.64210808e-01 -5.34745336e-01 8.68789777e-02 -4.00773674e-01
1.02508783e+00 -3.05127263e-01 -4.06846225e-01 4.67364967e-01
2.06679654e+00 4.10373628e-01 1.36711609e+00 5.43043971e-01
9.34447467e-01 9.00139213e-01 7.30516672e-01 5.95201075e-01
1.03210247e+00 6.76947951e-01 9.12824571e-01 6.99137524e-02
6.95188642e-02 -6.59278154e-01 -4.75376695e-01 8.27315629e-01
-3.97254556e-01 -5.43511271e-01 -9.20211911e-01 1.25572950e-01
-2.15736961e+00 -1.11812103e+00 -4.85237181e-01 1.87987363e+00
3.32375407e-01 -8.03385973e-01 -1.50350034e-01 -3.27459276e-01
5.74170768e-01 1.53317571e-01 -3.43982279e-01 -1.90710619e-01
-6.09016299e-01 -2.01112792e-01 1.60213023e-01 3.79608780e-01
-7.11909711e-01 1.24191153e+00 5.83063030e+00 7.38936722e-01
-5.80259740e-01 1.14415757e-01 -2.15277135e-01 8.46858546e-02
-1.18230796e+00 2.40696475e-01 -3.67702663e-01 1.63721308e-01
5.08487940e-01 -2.83536196e-01 8.78742397e-01 4.70039815e-01
3.08736898e-02 -3.15742224e-01 -5.47096908e-01 1.18488050e+00
2.68615305e-01 -1.51201856e+00 7.93237686e-01 -1.53129905e-01
2.85488248e-01 -3.94241154e-01 -3.93069774e-01 2.30089098e-01
2.81808317e-01 -9.86657917e-01 6.09606862e-01 1.22030270e+00
9.56418693e-01 -8.06800365e-01 5.18913031e-01 -7.51094967e-02
-1.44199491e+00 -7.39715935e-04 -4.04518634e-01 4.14485067e-01
2.52497852e-01 3.49256285e-02 -3.39671165e-01 1.40458786e+00
1.15468049e+00 6.03423834e-01 -1.15582228e+00 1.01522827e+00
-1.29082231e-02 -3.70403767e-01 -2.01864302e-01 -1.76101238e-01
-2.00670853e-01 -4.23311919e-01 1.39601871e-01 7.38629758e-01
5.58497787e-01 3.39059085e-01 -1.31123558e-01 4.76720572e-01
5.62508814e-02 6.46463513e-01 -1.02102113e+00 -2.47060865e-01
7.06123531e-01 1.00385845e+00 -5.87722480e-01 -2.54447758e-01
-6.63070560e-01 5.99667609e-01 3.93235683e-01 7.62474179e-01
-9.23884988e-01 -5.10187149e-01 2.69190937e-01 3.05569738e-01
-6.03662133e-02 -3.21193337e-01 2.29281604e-01 -1.21822107e+00
-1.60938874e-01 -7.20849812e-01 8.48926783e-01 -1.67974901e+00
-9.97269690e-01 4.09865439e-01 3.20261955e-01 -8.46027315e-01
2.49891039e-02 -4.84027177e-01 -2.31455415e-02 4.46220636e-01
-1.11568153e+00 -1.69558060e+00 -8.78978252e-01 1.39475191e+00
-2.24550501e-01 4.68171621e-03 9.36541855e-01 5.89051485e-01
-3.49599630e-01 -1.08036399e-01 -2.77193844e-01 -6.05625659e-02
6.46148682e-01 -9.73388672e-01 -6.17501676e-01 2.49066189e-01
1.91506520e-02 7.80794799e-01 5.44452667e-01 -9.83776271e-01
-2.20646811e+00 -7.19056487e-01 8.41573179e-01 -3.02592486e-01
7.15257704e-01 -4.34636213e-02 -8.49585354e-01 8.72761607e-01
6.01799190e-01 -8.49030688e-02 8.87314260e-01 6.75473511e-02
-3.74848574e-01 -2.36804768e-01 -1.13010609e+00 7.81143129e-01
1.45005155e+00 -8.56127381e-01 -4.12744641e-01 2.87098557e-01
1.04672110e+00 -3.28282267e-01 -1.23791623e+00 4.90704209e-01
4.03800756e-01 -1.14415181e+00 1.23635077e+00 -9.43449080e-01
8.15630704e-02 -5.28281569e-01 -9.15463209e-01 -7.41921902e-01
6.89975545e-02 4.53894325e-02 -3.70812640e-02 1.48682296e+00
3.11882228e-01 -5.55218935e-01 5.60224414e-01 1.02050495e+00
-9.48687866e-02 -1.69806451e-01 -8.27788174e-01 -3.39634329e-01
-6.99772775e-01 -4.47105229e-01 1.22704804e+00 1.08290172e+00
2.55069852e-01 1.09289564e-01 -9.18464512e-02 3.73123169e-01
3.96529704e-01 2.54250795e-01 8.32372546e-01 -1.31953084e+00
3.86955678e-01 -1.67963486e-02 -8.08471560e-01 -3.92673463e-01
-2.11631097e-02 -8.40558469e-01 -6.61478221e-01 -2.40957665e+00
1.89926401e-02 -4.16763276e-01 5.88614158e-02 4.58509922e-01
3.46832067e-01 4.90329042e-02 2.71757036e-01 1.79645672e-01
-1.31881082e+00 4.05148953e-01 1.46291256e+00 -1.86368734e-01
-2.51949336e-02 -9.98433948e-01 -5.00104487e-01 4.95369554e-01
6.22418940e-01 1.01545667e-02 -7.65988350e-01 -6.70271158e-01
1.23438883e+00 4.13924009e-01 6.89341009e-01 -6.88025594e-01
6.19169176e-01 -5.11421919e-01 7.66307786e-02 -7.46395528e-01
5.09360254e-01 -1.13555050e+00 8.21858108e-01 -2.58945692e-02
-1.20638482e-01 2.47211188e-01 1.94434553e-01 5.49373925e-01
-6.64502263e-01 -1.52654797e-01 -1.21231176e-01 -2.07367182e-01
-1.48852026e+00 1.61681011e-01 3.25649865e-02 2.08886132e-01
1.23561692e+00 -1.43358648e-01 -1.00318730e+00 -3.29169571e-01
-1.06685579e+00 4.08651054e-01 7.79166698e-01 4.90114599e-01
7.89991558e-01 -1.64408839e+00 2.76629012e-02 -2.45497748e-01
6.87714517e-01 -1.38598174e-01 7.57293522e-01 3.75303060e-01
-8.95289660e-01 4.63776290e-01 -4.82692689e-01 -3.28668773e-01
-1.19724333e+00 1.00569022e+00 7.06778988e-02 2.93404281e-01
-4.14377511e-01 2.00270712e-01 -2.14155629e-01 -6.46493495e-01
1.27195388e-01 4.66686822e-02 -6.28692210e-01 2.93556094e-01
7.54402801e-02 6.25320494e-01 -2.41988748e-01 -8.24999392e-01
-6.87959909e-01 8.67152870e-01 9.26884830e-01 -1.00495428e-01
8.81638348e-01 -5.07264435e-01 -6.22759938e-01 4.40690607e-01
9.18094516e-01 3.46415073e-01 -2.96229243e-01 1.29173636e-01
2.28596881e-01 -5.91511846e-01 -4.90309030e-01 -8.16360116e-01
-8.92305553e-01 5.22047102e-01 5.22081494e-01 5.59149742e-01
1.05223703e+00 4.15624708e-01 3.04144025e-01 6.88271880e-01
8.82113874e-01 -1.02676237e+00 5.11127450e-02 2.79140085e-01
8.29518259e-01 -1.09443438e+00 1.63432539e-01 -8.72956872e-01
-8.79225552e-01 1.10339916e+00 7.24024057e-01 5.30121684e-01
5.78724623e-01 -5.04669920e-02 8.56685191e-02 -9.50920284e-01
-5.11826754e-01 -6.27835155e-01 1.79570422e-01 7.81870544e-01
2.97151923e-01 9.56585929e-02 -2.33157590e-01 1.48550138e-01
-3.64071429e-02 1.53841585e-01 1.41056880e-01 1.05333900e+00
-5.15074670e-01 -9.39796269e-01 -2.34981343e-01 1.15852743e-01
4.43529040e-02 7.16985911e-02 -4.71361995e-01 1.24306035e+00
3.15744549e-01 1.01237249e+00 3.15721452e-01 -2.45077938e-01
7.19880223e-01 -1.15972474e-01 5.09488940e-01 -2.48015895e-01
-2.96822369e-01 -4.93443727e-01 3.39465618e-01 -1.00077212e+00
-8.58381569e-01 -1.07254997e-01 -1.83092952e+00 -3.74776483e-01
2.57201701e-01 2.47093007e-01 8.39217067e-01 6.87290311e-01
4.30767685e-01 2.19548315e-01 -2.62524307e-01 -1.14047013e-01
6.29456103e-01 -3.13459545e-01 -7.86348224e-01 7.81558990e-01
-2.30455250e-01 -5.80728412e-01 3.62598225e-02 -5.34006860e-03] | [9.27785587310791, 7.914941787719727] |
4408db60-24ec-4b15-928d-d107ac06ef4b | deep-learning-for-android-malware-defenses-a | 2103.05292 | null | https://arxiv.org/abs/2103.05292v3 | https://arxiv.org/pdf/2103.05292v3.pdf | Deep Learning for Android Malware Defenses: a Systematic Literature Review | Malicious applications (particularly those targeting the Android platform) pose a serious threat to developers and end-users. Numerous research efforts have been devoted to developing effective approaches to defend against Android malware. However, given the explosive growth of Android malware and the continuous advancement of malicious evasion technologies like obfuscation and reflection, Android malware defense approaches based on manual rules or traditional machine learning may not be effective. In recent years, a dominant research field called deep learning (DL), which provides a powerful feature abstraction ability, has demonstrated a compelling and promising performance in a variety of areas, like natural language processing and computer vision. To this end, employing deep learning techniques to thwart Android malware attacks has recently garnered considerable research attention. Yet, no systematic literature review focusing on deep learning approaches for Android Malware defenses exists. In this paper, we conducted a systematic literature review to search and analyze how deep learning approaches have been applied in the context of malware defenses in the Android environment. As a result, a total of 132 studies covering the period 2014-2021 were identified. Our investigation reveals that, while the majority of these sources mainly consider DL-based on Android malware detection, 53 primary studies (40.1 percent) design defense approaches based on other scenarios. This review also discusses research trends, research focuses, challenges, and future research directions in DL-based Android malware defenses. | ['Yepang Liu', 'Li Li', 'Chakkrit Tantithamthavorn', 'Yue Liu'] | 2021-03-09 | null | null | null | null | ['android-malware-detection', 'mobile-security'] | ['miscellaneous', 'miscellaneous'] | [ 2.27238704e-02 -3.35679799e-01 -9.64432299e-01 1.65362731e-01
-2.82068819e-01 -6.42599642e-01 6.68199360e-01 -1.92827269e-01
-7.04855844e-02 2.25618169e-01 -2.70749927e-01 -1.08907425e+00
4.47824737e-03 -5.74363172e-01 -5.47157168e-01 -4.62257773e-01
-4.13957499e-02 -3.07989448e-01 -1.37065038e-01 -2.38450795e-01
6.26134872e-01 3.53948742e-01 -1.53218567e+00 4.48274732e-01
9.60338414e-01 9.83636975e-01 -2.21832260e-01 5.51203609e-01
-1.72912166e-01 7.63142109e-01 -1.00278819e+00 -6.79571152e-01
1.03866987e-01 -4.74829972e-02 -5.22199571e-01 -4.75379944e-01
1.39604047e-01 -6.12802148e-01 -3.31754059e-01 1.17069077e+00
3.90779942e-01 -4.35436219e-01 4.06879008e-01 -1.24366462e+00
-1.01707518e+00 3.81872565e-01 -7.90501535e-01 4.41696018e-01
3.78386110e-01 1.90281197e-01 4.95728225e-01 -5.84852278e-01
3.01773306e-02 1.17274106e+00 9.05297935e-01 7.85917103e-01
-5.33944488e-01 -1.01491940e+00 1.21058866e-01 2.30808631e-01
-9.90435302e-01 -2.82461017e-01 9.27057207e-01 -6.65156305e-01
1.06247461e+00 2.35694617e-01 6.39403999e-01 1.75693512e+00
1.03560221e+00 5.97844481e-01 1.16522777e+00 -1.69732481e-01
1.69477969e-01 2.50983059e-01 1.81436002e-01 4.60063398e-01
6.40768588e-01 2.78189927e-01 1.11793075e-02 -7.00364351e-01
7.28561208e-02 3.00126433e-01 1.91950873e-01 1.46231577e-01
-2.65825570e-01 1.23755884e+00 2.18722120e-01 2.79979765e-01
-2.39294916e-01 -1.75480783e-01 1.05016160e+00 1.23805650e-01
6.52379334e-01 4.33078647e-01 -7.40457118e-01 -4.09879178e-01
-7.45328903e-01 -3.87254246e-02 9.06056166e-01 2.12140560e-01
3.50993007e-01 6.34384930e-01 4.76353168e-01 6.53445363e-01
6.89881861e-01 5.63803971e-01 7.96365321e-01 -7.19981968e-01
1.14101142e-01 7.50294566e-01 -5.31792104e-01 -1.59160578e+00
1.00633703e-01 -1.35912120e-01 -5.91154933e-01 3.26771438e-01
-1.61984831e-01 -3.76887798e-01 -7.14492559e-01 1.29686296e+00
1.50673717e-01 2.30648205e-01 6.00728989e-02 1.36049241e-01
7.67574906e-01 6.18478477e-01 1.60333663e-01 -2.25022942e-01
1.37998486e+00 -6.21696532e-01 -6.54615283e-01 -3.86065990e-01
6.11609101e-01 -6.84850991e-01 1.27819800e+00 7.15731859e-01
-5.73060989e-01 -4.12717313e-01 -1.23173368e+00 1.42839715e-01
-9.20069814e-01 -7.41771907e-02 7.15457082e-01 1.51421905e+00
-8.06644320e-01 1.92157626e-02 -1.09390342e+00 -1.91637814e-01
8.73394847e-01 6.43903792e-01 8.55006054e-02 2.17357025e-01
-9.91010010e-01 6.06389523e-01 -7.07344860e-02 -8.56123343e-02
-1.44784391e+00 -3.78485471e-01 -7.01999724e-01 -9.55213457e-02
6.22269869e-01 5.08472733e-02 1.26472425e+00 -1.01713920e+00
-1.46899450e+00 6.06616497e-01 1.05996571e-01 -7.69567370e-01
-1.69077456e-01 -5.97713828e-01 -6.00977719e-01 -1.98133126e-01
8.62879958e-03 2.26635840e-02 1.19695163e+00 -1.12097573e+00
-3.96850914e-01 -5.41926682e-01 4.07046765e-01 -4.26328659e-01
-1.01967418e+00 4.41388935e-01 4.85240519e-02 -5.98692954e-01
-8.32837343e-01 -1.29712701e+00 7.17763603e-02 -6.32056952e-01
-3.49722087e-01 -3.16481084e-01 1.75455737e+00 -7.96309650e-01
1.64885318e+00 -2.07761002e+00 -2.61965513e-01 -1.47037655e-01
4.80376661e-01 1.16887915e+00 1.38046190e-01 2.98022866e-01
1.01181082e-01 7.31994748e-01 -2.39145711e-01 -3.58048864e-02
-4.25353676e-01 1.28302634e-01 -6.66753531e-01 4.52843189e-01
-1.30985528e-01 8.77723992e-01 -6.63804412e-01 -6.24600537e-02
3.11778009e-01 7.38425374e-01 -6.15577877e-01 -6.18885597e-03
-2.52077729e-01 2.03693986e-01 -7.59912789e-01 1.06560349e+00
6.65679932e-01 -4.93695214e-02 3.55053216e-01 3.29884738e-01
-8.39610398e-02 4.81834412e-01 -6.59809485e-02 8.85840535e-01
-7.85350204e-01 9.95602667e-01 2.50400484e-01 -9.33047891e-01
6.61003351e-01 5.02451323e-02 4.84893143e-01 -4.01858360e-01
5.48258066e-01 2.36869112e-01 5.84399819e-01 -6.66715503e-01
1.49281919e-01 5.58969498e-01 -8.28914642e-02 4.56983119e-01
-4.93152350e-01 2.92299122e-01 -4.08028901e-01 -2.37609334e-02
1.21536708e+00 -3.42338122e-02 4.20520455e-01 -2.54128613e-02
9.08859551e-01 -1.17619798e-01 3.70006442e-01 4.51048791e-01
-4.93884772e-01 -3.37467045e-01 7.16424763e-01 -4.69395071e-01
-3.90310436e-01 -5.85343719e-01 8.41786042e-02 1.33298266e+00
-2.07487285e-01 -7.26558387e-01 -1.30264795e+00 -1.34956419e+00
-1.26380995e-01 2.84073442e-01 -8.52859020e-01 -6.93669915e-01
-7.98378885e-01 -7.62855828e-01 1.12883639e+00 1.71690390e-01
8.36810112e-01 -1.21860278e+00 -7.94580042e-01 -2.19071433e-01
2.32197613e-01 -1.04289699e+00 -1.61484465e-01 -1.15907907e-01
-8.61337245e-01 -1.61125350e+00 -2.62305588e-01 -7.59236217e-01
2.88370222e-01 5.95258713e-01 6.78081274e-01 4.29594755e-01
-2.62372077e-01 4.84164327e-01 -3.67036104e-01 -6.60938263e-01
-7.76886404e-01 4.29288447e-01 4.50670242e-01 -2.64831036e-01
8.80666316e-01 -2.28589803e-01 -2.24800453e-01 2.10489199e-01
-8.69504273e-01 -1.06717324e+00 7.46930540e-01 6.06014848e-01
5.26365899e-02 4.77263212e-01 6.89716339e-01 -8.10919344e-01
1.25259161e+00 -9.21740234e-01 -5.71582437e-01 -5.16059110e-04
-8.50483477e-01 -6.04755700e-01 7.56486356e-01 -8.88214588e-01
-9.30275619e-01 -4.80811656e-01 -5.45270562e-01 -3.73096079e-01
-1.87160745e-01 5.70553899e-01 -3.05475801e-01 -5.88027418e-01
8.23643446e-01 2.11496621e-01 4.38758880e-01 -3.77959251e-01
3.83484177e-02 1.15083098e+00 -1.48510784e-01 -3.03837508e-01
5.97831905e-01 3.15677851e-01 -3.71653646e-01 -1.19845152e+00
-4.84121084e-01 1.56942993e-01 -9.10426304e-02 4.52441722e-03
9.86577094e-01 -5.67041874e-01 -9.55452740e-01 7.32987106e-01
-9.78992760e-01 -2.35447004e-01 4.47634369e-01 1.22725688e-01
2.97074988e-02 6.92341149e-01 -6.40463054e-01 -8.04030538e-01
-5.20512819e-01 -1.82013261e+00 6.78627431e-01 1.65129885e-01
-1.96358964e-01 -1.06829965e+00 8.01443011e-02 5.77068090e-01
5.50971150e-01 2.03756645e-01 9.33783293e-01 -8.60936046e-01
-7.31448978e-02 -3.80619735e-01 8.48602853e-04 7.51869321e-01
6.23773038e-01 4.44067568e-01 -9.85469878e-01 -2.51705408e-01
7.66607046e-01 -3.38635817e-02 4.77304816e-01 4.92168814e-01
1.54235601e+00 -8.18808973e-01 -5.56072593e-01 5.17859340e-01
8.59522104e-01 9.60916638e-01 7.17176795e-01 5.88602006e-01
9.52894807e-01 6.42096639e-01 6.18540049e-01 2.37630442e-01
3.02660942e-01 4.32711065e-01 9.55554783e-01 3.62964630e-01
2.05630913e-01 2.75765471e-02 9.45184946e-01 4.92636353e-01
5.60494624e-02 -1.54971167e-01 -1.06179035e+00 9.88889784e-02
-1.24870908e+00 -8.03603828e-01 -1.72979850e-02 2.02461219e+00
4.86351818e-01 4.38211292e-01 2.83661425e-01 1.48576960e-01
5.60294271e-01 4.59941804e-01 -7.62278616e-01 -1.08759236e+00
3.16992939e-01 7.36876056e-02 2.40898341e-01 3.32062155e-01
-1.24054945e+00 1.10725772e+00 6.70725298e+00 8.70523274e-01
-1.60779130e+00 4.43987072e-01 5.95336497e-01 2.53057241e-01
6.65917769e-02 -1.73305303e-01 -8.32269073e-01 8.23362589e-01
1.33160293e+00 1.49741754e-01 4.00636971e-01 1.45207560e+00
1.10763878e-01 1.84149146e-01 -4.50641781e-01 9.31945801e-01
2.69293875e-01 -1.29327631e+00 3.71024795e-02 9.73991692e-01
6.55619144e-01 4.82745022e-02 9.78598714e-01 6.67025089e-01
-1.77966028e-01 -1.02318347e+00 -6.24458827e-02 -2.36814767e-01
6.62756085e-01 -9.16860878e-01 6.98059916e-01 2.77330160e-01
-9.38864827e-01 -8.33594620e-01 -1.73041120e-01 -4.35152113e-01
-3.02903295e-01 2.16872931e-01 -8.29879344e-01 -7.66304694e-03
8.65857959e-01 9.62138414e-01 -7.30191112e-01 3.42202157e-01
-2.65110195e-01 1.17131948e+00 3.27743232e-01 -7.45450258e-02
3.31957847e-01 2.03722883e-02 4.28277552e-01 9.85607326e-01
-2.63396334e-02 -2.63261497e-01 7.87066892e-02 4.72491533e-01
5.69989812e-03 -1.45430788e-01 -1.21694362e+00 -6.62166297e-01
3.92488897e-01 1.19274449e+00 -5.98844707e-01 -1.67187840e-01
-7.66100764e-01 4.79931206e-01 -2.69882619e-01 1.06842376e-01
-1.05645657e+00 -2.58252680e-01 1.29323411e+00 3.66500467e-01
-1.01603106e-01 -4.95876998e-01 -4.93789673e-01 -8.43363464e-01
-3.25377554e-01 -1.73702443e+00 1.41915992e-01 -8.40261653e-02
-1.00793004e+00 6.49054229e-01 1.34340361e-01 -9.05688643e-01
-3.17868203e-01 -1.05895507e+00 -7.56471992e-01 2.54121155e-01
-9.29444015e-01 -1.05086243e+00 3.85401174e-02 3.74073684e-01
9.97585475e-01 -8.75947416e-01 6.25963151e-01 3.77161801e-01
-9.12490070e-01 7.13526845e-01 -1.62940338e-01 3.85736581e-03
2.97439903e-01 -6.37207866e-01 5.39547801e-01 7.96001196e-01
-1.98213890e-01 1.40510583e+00 3.02827328e-01 -1.19514740e+00
-1.72092247e+00 -8.23007047e-01 1.82868451e-01 -6.61969602e-01
7.14291513e-01 -4.80204046e-01 -9.89005804e-01 7.66790450e-01
4.00826007e-01 -4.01788354e-01 1.01313686e+00 -8.96942019e-02
-5.04968703e-01 -4.40999903e-02 -1.18732679e+00 7.14452744e-01
5.60954034e-01 -7.42870629e-01 -1.77985877e-01 1.98619068e-01
7.09880710e-01 -1.34437352e-01 -5.43027520e-01 6.65701985e-01
7.75145769e-01 -9.72360849e-01 1.10139036e+00 -6.12383366e-01
3.75591606e-01 1.89168602e-01 -1.42175525e-01 -6.16049945e-01
1.61587074e-01 -7.20803916e-01 -9.04055297e-01 9.01054800e-01
-1.18855946e-03 -8.76413703e-01 9.34204936e-01 1.46749705e-01
-2.49207482e-01 -1.39635956e+00 -7.02417612e-01 -7.74205506e-01
3.02353501e-01 -5.82843006e-01 5.82368135e-01 9.18119848e-01
-5.38960338e-01 5.82259409e-02 -4.92351472e-01 -5.33825234e-02
2.73403108e-01 -5.61729252e-01 9.06098485e-01 -1.15800500e+00
-2.88510788e-02 -8.19500566e-01 -2.37753913e-01 -7.63499856e-01
7.31739342e-01 -5.18106997e-01 -4.43546295e-01 -1.07659054e+00
3.67181376e-02 -8.83550122e-02 -7.86090866e-02 6.47132158e-01
8.24470371e-02 2.28389740e-01 -2.53167540e-01 2.23177910e-01
-9.98922586e-02 3.53371054e-01 6.20569587e-01 -3.32068861e-01
-4.06884462e-01 4.38855439e-01 -1.25298035e+00 1.34558344e+00
1.24461162e+00 -3.72371465e-01 -6.97240829e-01 -2.74892867e-01
1.90704867e-01 -6.07103467e-01 1.41176388e-01 -6.61141098e-01
-3.49035770e-01 -2.73451447e-01 3.40319499e-02 -2.71295190e-01
3.05333048e-01 -7.51242816e-01 -3.40001762e-01 9.19663548e-01
1.46565452e-01 4.55964684e-01 5.52308857e-01 4.98836458e-01
1.15694124e-02 -1.12315930e-01 6.34849131e-01 -9.37270280e-03
-7.09090412e-01 2.07411334e-01 -9.72111583e-01 -9.47984755e-02
1.32817352e+00 -4.04787213e-01 -5.42173684e-01 -2.39390463e-01
-1.87151000e-01 -4.09618527e-01 4.88496870e-01 9.97249424e-01
8.57367456e-01 -9.37643588e-01 -8.21694285e-02 2.21682489e-01
-2.23980978e-01 -7.50338495e-01 1.39711857e-01 7.16620982e-01
-3.71528506e-01 6.24813855e-01 -2.39726961e-01 -2.79730707e-01
-1.59015048e+00 1.00498402e+00 2.00446188e-01 -1.45865664e-01
-2.03449354e-01 5.50686598e-01 3.42034936e-01 -3.60949397e-01
2.50139147e-01 -2.24663131e-02 -4.93557990e-01 -3.31890211e-02
7.76402295e-01 8.15817833e-01 -3.44972499e-02 -8.45535576e-01
-8.23863924e-01 4.95172948e-01 -3.71449977e-01 4.60415304e-01
8.71719778e-01 1.42397761e-01 -3.65680844e-01 2.65357383e-02
1.28095269e+00 5.93507349e-01 -7.85534382e-01 6.59641743e-01
6.80385903e-02 -4.91918802e-01 -1.36450872e-01 -7.37567484e-01
-1.11247087e+00 1.01010191e+00 7.58560658e-01 7.83365250e-01
1.13709950e+00 -2.79458314e-01 9.74515557e-01 2.84111887e-01
3.22881937e-01 -6.16635740e-01 5.34871340e-01 6.07861578e-01
6.33543432e-01 -1.38626850e+00 -1.19204810e-02 -2.62596756e-01
-3.82632881e-01 8.84148598e-01 1.01504326e+00 -1.94625705e-01
9.99147356e-01 3.46346766e-01 -2.25503985e-02 -2.20722452e-01
-2.48572499e-01 5.03407776e-01 1.95841670e-01 9.66409624e-01
5.28353810e-01 -5.83867282e-02 -4.75643426e-01 6.30969405e-01
5.94368018e-03 -4.28325176e-01 5.52462161e-01 1.07000041e+00
-3.82051945e-01 -1.46647143e+00 -4.18884993e-01 6.14435852e-01
-1.26544273e+00 -1.16688780e-01 -8.91788185e-01 8.26775670e-01
4.65016365e-01 1.28091705e+00 -5.02041459e-01 -1.04489112e+00
-2.21667081e-01 -3.97544891e-01 3.36504616e-02 -8.16125154e-01
-8.10897052e-01 -2.24445134e-01 -2.10620269e-01 -4.60694551e-01
-7.25531206e-02 -4.07630444e-01 -9.00323451e-01 -4.71551389e-01
-2.82432705e-01 -1.55155156e-02 1.00989950e+00 9.47225869e-01
7.64655530e-01 4.90211964e-01 6.40399337e-01 -7.69959748e-01
-5.44612288e-01 -7.91734338e-01 2.41898056e-02 -5.00374556e-01
4.67833996e-01 -9.77527022e-01 -4.64586616e-01 -4.27723944e-01] | [14.423236846923828, 9.68045425415039] |
a9a9bb56-f292-49ac-9a70-b0a08ff37beb | clac-sentipipe-semeval2015-subtasks-10-be-and | null | null | https://aclanthology.org/S15-2081 | https://aclanthology.org/S15-2081.pdf | CLaC-SentiPipe: SemEval2015 Subtasks 10 B,E, and Task 11 | null | ['Canberk {\\"O}zdemir', 'Sabine Bergler'] | 2015-06-01 | null | null | null | semeval-2015-6 | ['twitter-sentiment-analysis'] | ['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] | [-7.354000568389893, 3.7918314933776855] |
f3b91149-6030-465f-abdd-476f5efd3935 | transformer-based-generative-adversarial | 2205.10663 | null | https://arxiv.org/abs/2205.10663v2 | https://arxiv.org/pdf/2205.10663v2.pdf | Transformer based Generative Adversarial Network for Liver Segmentation | Automated liver segmentation from radiology scans (CT, MRI) can improve surgery and therapy planning and follow-up assessment in addition to conventional use for diagnosis and prognosis. Although convolutional neural networks (CNNs) have become the standard image segmentation tasks, more recently this has started to change towards Transformers based architectures because Transformers are taking advantage of capturing long range dependence modeling capability in signals, so called attention mechanism. In this study, we propose a new segmentation approach using a hybrid approach combining the Transformer(s) with the Generative Adversarial Network (GAN) approach. The premise behind this choice is that the self-attention mechanism of the Transformers allows the network to aggregate the high dimensional feature and provide global information modeling. This mechanism provides better segmentation performance compared with traditional methods. Furthermore, we encode this generator into the GAN based architecture so that the discriminator network in the GAN can classify the credibility of the generated segmentation masks compared with the real masks coming from human (expert) annotations. This allows us to extract the high dimensional topology information in the mask for biomedical image segmentation and provide more reliable segmentation results. Our model achieved a high dice coefficient of 0.9433, recall of 0.9515, and precision of 0.9376 and outperformed other Transformer based approaches. | ['Ulas Bagci', 'Daniela Ladner', 'Amir Borhani', 'Debesh Jha', 'Elif Keles', 'Matthew Antalek', 'Bin Wang', 'Zheyuan Zhang', 'Ugur Demir'] | 2022-05-21 | null | null | null | null | ['liver-segmentation'] | ['medical'] | [ 5.49602248e-02 4.79177564e-01 1.40341073e-01 -3.20856929e-01
-6.80808723e-01 -4.78959322e-01 4.50090140e-01 -3.39879207e-02
-2.45854244e-01 7.03805149e-01 1.98927671e-01 -1.12642452e-01
3.13228294e-02 -1.02860034e+00 -4.57539946e-01 -1.04025686e+00
-5.26800677e-02 6.59868062e-01 2.57596433e-01 3.17229256e-02
-1.19954154e-01 5.94536066e-01 -7.63259768e-01 2.85321355e-01
1.28572845e+00 1.15571189e+00 1.71397761e-01 5.41047692e-01
-2.83409685e-01 5.91969788e-01 -6.84557617e-01 -3.86513561e-01
3.16432685e-01 -7.72987425e-01 -8.00868154e-01 -9.19792578e-02
-2.13448942e-01 -8.34257975e-02 -1.31648287e-01 1.16741633e+00
6.30794406e-01 -1.29137665e-01 8.18741798e-01 -8.16445291e-01
-6.06704772e-01 8.64212334e-01 -5.21980166e-01 1.43383309e-01
5.78643717e-02 3.62070471e-01 3.53639007e-01 -3.65776569e-01
5.11198461e-01 7.81791210e-01 7.90777683e-01 6.21351182e-01
-1.00015843e+00 -5.80576479e-01 -3.52536529e-01 1.29963588e-02
-1.16790771e+00 5.58204204e-02 8.62871170e-01 -4.98370707e-01
3.78017038e-01 3.70882392e-01 1.08293939e+00 9.69251812e-01
6.06112540e-01 6.44049525e-01 1.35464752e+00 -1.70349926e-01
-1.55214332e-02 3.11087281e-01 -2.55371451e-01 7.92703092e-01
-8.36731214e-03 4.27235551e-02 1.85733154e-01 2.46529907e-01
1.11986184e+00 1.50772840e-01 -5.02086163e-01 -1.63338602e-01
-1.16808450e+00 8.28457057e-01 1.13324785e+00 8.56634676e-01
-7.84071624e-01 1.09167010e-01 3.22401881e-01 -1.32722966e-02
1.62907302e-01 6.07031345e-01 -4.07126322e-02 2.57918060e-01
-9.86850977e-01 -3.84410799e-01 6.17445886e-01 5.47331095e-01
3.43386352e-01 3.52359235e-01 -5.09991109e-01 5.27103662e-01
3.41070861e-01 3.09765041e-01 1.12030697e+00 -4.95564044e-01
1.31220091e-02 8.52352381e-01 -5.10592937e-01 -8.30213487e-01
-5.60201883e-01 -1.05377758e+00 -1.39430296e+00 1.56302482e-01
3.62934202e-01 -6.98965043e-02 -1.37716913e+00 1.54050231e+00
2.60190159e-01 1.48081645e-01 -5.17771766e-02 1.01899827e+00
8.79945934e-01 5.35515606e-01 1.07631870e-01 -2.09539719e-02
1.43255055e+00 -8.52777481e-01 -7.25684345e-01 6.34608418e-02
4.25961465e-01 -6.50923550e-01 7.97494948e-01 3.03798229e-01
-1.04250610e+00 -4.12123740e-01 -1.08647609e+00 3.12858135e-01
-3.09111238e-01 -2.90285554e-02 5.66623330e-01 7.80272305e-01
-1.18405974e+00 5.97768009e-01 -9.55383360e-01 -2.33566940e-01
7.85393119e-01 5.52217305e-01 -2.69765288e-01 1.93500265e-01
-1.09809208e+00 7.99715757e-01 4.01280195e-01 1.57698691e-01
-9.21994925e-01 -6.96683586e-01 -6.08103633e-01 1.47261331e-02
7.52820596e-02 -8.96321595e-01 8.25524211e-01 -1.35278869e+00
-1.68281436e+00 8.31715345e-01 3.79824162e-01 -7.39548147e-01
7.89800048e-01 3.61928612e-01 -7.03814775e-02 3.28356147e-01
-1.21359348e-01 7.91363418e-01 6.74276888e-01 -1.24387801e+00
-1.10428557e-01 -4.10291076e-01 -2.79211253e-01 7.14535043e-02
-3.02906837e-02 -2.76036650e-01 -1.93377435e-01 -8.63395333e-01
3.31282675e-01 -8.29486132e-01 -4.25487012e-01 -1.86221763e-01
-5.53362429e-01 1.69950724e-01 6.39394701e-01 -9.28575456e-01
8.51577163e-01 -1.84129298e+00 1.09230161e-01 4.99638796e-01
4.76353437e-01 3.55004638e-01 2.86018461e-01 -3.13670188e-02
-2.12005004e-01 1.84639335e-01 -5.43540180e-01 -2.41215657e-02
-4.11555499e-01 1.30928427e-01 2.21606836e-01 4.47529048e-01
1.00989655e-01 1.36410594e+00 -6.71622694e-01 -6.23726428e-01
3.65704536e-01 6.71374440e-01 -4.15839374e-01 2.87984163e-01
6.89687580e-02 1.11187327e+00 -5.85411906e-01 4.80856806e-01
5.85338831e-01 -2.80114144e-01 1.05782971e-01 -4.15702909e-01
3.16569686e-01 -1.11566588e-01 -6.72054529e-01 1.70980644e+00
-3.80985886e-01 3.25026214e-01 6.13703718e-03 -1.13483429e+00
1.04999280e+00 5.74905574e-01 8.16590667e-01 -7.31459200e-01
5.83474994e-01 1.88491121e-01 3.43569934e-01 -4.97823805e-01
-2.31604263e-01 -4.71114218e-01 5.40471040e-02 1.89289778e-01
-2.15703174e-02 -3.11308563e-01 -3.77990961e-01 -4.71555628e-02
8.44536304e-01 -2.11748946e-02 1.97809413e-01 -3.87561023e-01
7.04641759e-01 -7.28559941e-02 4.53725159e-01 2.84737796e-01
-1.24952801e-01 7.28847861e-01 5.09997308e-01 -4.18522209e-01
-9.75243866e-01 -7.61371315e-01 -1.08975098e-01 2.13824352e-03
-2.87272390e-02 1.81314573e-01 -1.17248166e+00 -9.17709410e-01
-3.60936075e-01 4.37280864e-01 -8.20794702e-01 -2.85797358e-01
-6.99350178e-01 -8.91810119e-01 6.14314437e-01 7.27818012e-01
8.57496381e-01 -1.03401113e+00 -6.57081962e-01 3.62856925e-01
-2.08451033e-01 -8.85060132e-01 -4.06436265e-01 1.70626163e-01
-9.75466251e-01 -9.48641002e-01 -1.00850737e+00 -6.19831026e-01
7.60243118e-01 -4.32898462e-01 1.00655162e+00 1.59661770e-01
-4.19059187e-01 7.04645142e-02 -2.84202069e-01 -2.97364742e-01
-7.72824287e-01 5.96171729e-02 -4.35921997e-01 1.25665888e-01
-1.14336655e-01 -5.99371254e-01 -9.72023308e-01 2.22628430e-01
-1.00122881e+00 1.71325013e-01 9.84233677e-01 1.05575645e+00
6.13520861e-01 3.93688194e-02 5.95592737e-01 -9.86212909e-01
4.96855795e-01 -4.50341076e-01 -4.55914170e-01 1.24731869e-01
-6.78990304e-01 -3.22741382e-02 7.32606053e-01 -2.57537752e-01
-9.54975069e-01 1.20407239e-01 -3.92459333e-01 -5.39666235e-01
2.28546700e-03 3.43028694e-01 -1.47945866e-01 -2.65409172e-01
4.75239992e-01 3.14686328e-01 3.39338899e-01 -2.29664251e-01
1.32004181e-02 4.05534446e-01 5.45310438e-01 -2.02024117e-01
5.57363689e-01 3.27995330e-01 2.05888972e-01 -3.38762730e-01
-2.48385206e-01 1.90712791e-02 -6.18354857e-01 -2.87663847e-01
1.40098703e+00 -5.20173430e-01 -6.44634187e-01 4.94521141e-01
-9.16487157e-01 -3.22616756e-01 -5.21486878e-01 5.30024052e-01
-4.01549280e-01 2.97002465e-01 -7.68730283e-01 -6.35068178e-01
-6.99729264e-01 -1.66659105e+00 7.09914267e-01 4.28905189e-01
5.40442020e-02 -1.18489993e+00 -2.73531556e-01 2.75056332e-01
8.28107893e-01 9.31802571e-01 9.52285945e-01 -8.19086671e-01
-6.09527767e-01 -4.05671805e-01 -2.37239257e-01 4.99524087e-01
1.10919781e-01 -3.36048216e-01 -9.74694014e-01 -1.30960524e-01
4.33008045e-01 2.16042131e-01 6.55813634e-01 7.90354311e-01
1.23414850e+00 -3.54499906e-01 -3.66233975e-01 7.77448952e-01
1.59616864e+00 4.20207530e-01 8.84842336e-01 1.97456852e-02
7.48293042e-01 3.70867074e-01 6.40824512e-02 -1.17366398e-02
1.40182942e-01 6.59763932e-01 6.17216110e-01 -5.73207438e-01
-5.27269304e-01 -1.03550836e-01 -5.62473089e-02 8.36345375e-01
-6.20116815e-02 -2.22032011e-01 -1.13500965e+00 3.74218553e-01
-1.41153920e+00 -7.10676908e-01 -2.25052238e-01 2.12332678e+00
6.66420043e-01 7.81387612e-02 -8.21188986e-02 1.61450639e-01
6.74932122e-01 -3.42182845e-01 -3.82496297e-01 -3.50996554e-01
-3.28147365e-03 5.56342721e-01 6.69660807e-01 3.58427316e-01
-9.02929306e-01 5.16722858e-01 5.93191910e+00 8.64230514e-01
-1.45948625e+00 3.10679823e-01 1.02447128e+00 4.12548333e-01
-3.52319449e-01 -3.64184320e-01 -9.97767150e-02 6.73025429e-01
7.66470492e-01 9.79512930e-02 1.00700244e-01 4.55513805e-01
8.64612162e-02 -1.28994837e-01 -7.93939710e-01 7.81245232e-01
2.71925125e-02 -1.21288633e+00 1.03124902e-01 3.24855655e-01
7.21244335e-01 -2.05741465e-01 1.35270461e-01 1.69936121e-02
1.79815158e-01 -1.46196020e+00 3.76679659e-01 7.70782709e-01
9.92966235e-01 -5.62075555e-01 1.24838388e+00 3.53417844e-01
-9.88190055e-01 8.66551977e-03 1.71043258e-02 5.16196072e-01
4.02785130e-02 7.36801445e-01 -1.25409663e+00 8.52601409e-01
3.61942321e-01 3.18599045e-01 -4.53481704e-01 1.12456059e+00
-1.49500370e-01 5.90341389e-01 -2.79974133e-01 1.52575061e-01
4.76099938e-01 -3.43508989e-01 5.93854487e-01 1.01860559e+00
3.74993235e-01 1.05572101e-02 3.84733453e-02 9.97847617e-01
-1.77725881e-01 1.87937945e-01 -3.98979008e-01 1.87671795e-01
7.47254491e-02 1.42565560e+00 -1.33919764e+00 -4.61006016e-01
5.44528216e-02 9.55946207e-01 -1.72129706e-01 -1.36010265e-02
-1.01631463e+00 -1.93836689e-01 -4.25589010e-02 3.54091197e-01
1.12666473e-01 1.90357178e-01 -5.85982323e-01 -8.81853104e-01
-2.07414359e-01 -7.70917058e-01 3.72519821e-01 -6.35660887e-01
-1.13508272e+00 1.03560770e+00 -2.26915091e-01 -1.20461428e+00
-1.73663035e-01 -3.64529282e-01 -8.30964029e-01 1.06650221e+00
-1.34549189e+00 -1.39659822e+00 -6.86258733e-01 4.25451368e-01
4.79980826e-01 -5.54346014e-03 7.98356473e-01 3.46415520e-01
-3.77530634e-01 5.35991132e-01 -1.61061570e-01 3.96679550e-01
4.37936932e-01 -1.51785147e+00 -1.53153732e-01 8.57475579e-01
-1.87474772e-01 3.29714239e-01 4.93045479e-01 -5.83117306e-01
-1.06180108e+00 -9.88404632e-01 1.82575092e-01 -1.11296833e-01
1.95705637e-01 8.64677951e-02 -8.92612815e-01 4.57625628e-01
4.17545229e-01 3.09726298e-02 5.73331535e-01 -7.41029620e-01
4.41065356e-02 -1.91919088e-01 -1.73350930e+00 8.91471729e-02
5.41214168e-01 -7.01759756e-03 -4.06460166e-01 3.51410508e-02
4.19666320e-01 -6.40211284e-01 -1.17168510e+00 5.18544853e-01
5.09805381e-01 -1.00129855e+00 8.14396083e-01 -1.70402005e-01
4.23754871e-01 -2.48129323e-01 2.96944410e-01 -1.39941573e+00
-2.11899444e-01 -3.24098587e-01 3.36422056e-01 9.85889018e-01
4.13956583e-01 -8.45514536e-01 7.84491003e-01 4.66767669e-01
-3.38660270e-01 -1.03703356e+00 -9.35363650e-01 -4.43688184e-01
2.67524093e-01 -4.70259041e-02 8.93429995e-01 1.02861857e+00
-4.20536965e-01 1.51803242e-02 3.79785267e-03 -1.02413900e-01
5.32904387e-01 -6.53535873e-02 3.08694214e-01 -1.18032384e+00
-2.88459152e-01 -7.08357215e-01 -6.96569324e-01 -7.11731195e-01
-2.49640375e-01 -1.25099099e+00 -2.28143483e-01 -1.72714734e+00
5.35037667e-02 -7.00512648e-01 -3.09620112e-01 4.63922381e-01
5.46346828e-02 6.56723142e-01 5.22947274e-02 2.26548597e-01
8.71117413e-02 3.71707678e-01 1.83958375e+00 -1.19185671e-01
-5.98001555e-02 1.23369269e-01 -6.77840650e-01 5.25806010e-01
8.39674652e-01 -3.50712508e-01 -2.45859191e-01 -2.36194521e-01
-2.55030453e-01 3.06700230e-01 4.76414829e-01 -1.18587291e+00
1.72930136e-01 2.96737343e-01 8.57710123e-01 -2.45982766e-01
-2.34210752e-02 -1.16582942e+00 6.03242636e-01 9.23584938e-01
-1.30877271e-01 -1.24375381e-01 2.61430964e-02 3.06905121e-01
-3.45424324e-01 -2.78897174e-02 1.10170662e+00 -5.19272029e-01
-4.50799763e-02 3.68102998e-01 -1.82174817e-01 -1.98872447e-01
1.25767362e+00 -4.74725604e-01 4.13610190e-02 -2.98168242e-01
-9.07309234e-01 -7.45599717e-03 3.39147896e-01 -5.67909107e-02
5.67839980e-01 -1.28437710e+00 -6.60559893e-01 3.68677884e-01
-4.14052993e-01 3.12942296e-01 2.99996585e-01 1.42218733e+00
-7.88027525e-01 3.46626014e-01 -4.31967229e-01 -7.50898957e-01
-8.35210562e-01 4.28245097e-01 8.04820299e-01 -6.75170779e-01
-8.28141987e-01 7.12645531e-01 2.94390798e-01 -9.89266187e-02
-1.62690967e-01 -4.94205028e-01 -3.97111833e-01 -6.66580498e-02
1.07996471e-01 1.10228606e-01 1.55572459e-01 -5.80895424e-01
-2.58645386e-01 6.19275570e-01 2.25074932e-01 1.01000480e-01
1.22987759e+00 1.41666785e-01 -2.25399673e-01 9.36219096e-02
1.09119558e+00 5.04927477e-03 -8.27350318e-01 3.16682532e-02
-2.78622687e-01 -1.66052222e-01 2.59707361e-01 -1.17712557e+00
-1.67359829e+00 9.23324347e-01 8.90119135e-01 5.35383999e-01
1.38331640e+00 -1.00945055e-01 9.41738486e-01 -4.83803570e-01
1.79034978e-01 -5.53988516e-01 -4.38943878e-02 -9.10944566e-02
8.06424022e-01 -9.74613607e-01 -2.79155165e-01 -5.25578618e-01
-7.63949871e-01 1.25243032e+00 4.40196216e-01 -2.29218304e-01
6.01624370e-01 4.88457054e-01 1.55905604e-01 -3.08736295e-01
-1.05963685e-01 -6.29399195e-02 4.92781222e-01 6.33487761e-01
4.37406540e-01 2.59279191e-01 -2.50220597e-01 7.70938873e-01
-2.74759322e-01 -1.32070361e-02 3.77750069e-01 4.57515925e-01
-2.78393090e-01 -1.03925359e+00 -3.22313428e-01 6.08854532e-01
-7.48261809e-01 1.68878287e-02 -6.18653595e-02 6.46702647e-01
2.09650889e-01 4.13295239e-01 -4.43349779e-02 -2.03168929e-01
1.10475622e-01 1.72591299e-01 4.54683453e-01 -3.64910722e-01
-1.02754271e+00 3.12013507e-01 -3.95535469e-01 -4.20668662e-01
-3.29304427e-01 -2.66270489e-01 -1.35990775e+00 -1.29749700e-02
-3.51518154e-01 3.03509533e-01 8.50789309e-01 7.44575560e-01
1.08727790e-01 1.13669002e+00 7.12199867e-01 -5.88085949e-01
-3.25918883e-01 -1.05663478e+00 -3.89021605e-01 4.40974921e-01
1.17830567e-01 -4.74766523e-01 -1.26900256e-01 -7.55774463e-03] | [14.215920448303223, -2.2155873775482178] |
2180e2b7-d462-4b06-ae8c-afd78a1dd97a | variable-selection-for-kernel-two-sample | 2302.07415 | null | https://arxiv.org/abs/2302.07415v2 | https://arxiv.org/pdf/2302.07415v2.pdf | Variable Selection for Kernel Two-Sample Tests | We consider the variable selection problem for two-sample tests, aiming to select the most informative variables to distinguish samples from two groups. To solve this problem, we propose a framework based on the kernel maximum mean discrepancy (MMD). Our approach seeks a group of variables with a pre-specified size that maximizes the variance-regularized MMD statistics. This formulation also corresponds to the minimization of asymptotic type-II error while controlling type-I error, as studied in the literature. We present mixed-integer programming formulations and offer exact and approximation algorithms with performance guarantees for linear and quadratic types of kernel functions. Experimental results demonstrate the superior performance of our framework. | ['Yao Xie', 'Santanu S. Dey', 'Jie Wang'] | 2023-02-15 | null | null | null | null | ['variable-selection'] | ['methodology'] | [ 1.49541885e-01 1.82563022e-01 -6.34621322e-01 -5.54613590e-01
-1.26118207e+00 -2.47619346e-01 -1.00146227e-01 2.16935232e-01
-3.48461688e-01 1.15279973e+00 -5.15320420e-01 -3.53830427e-01
-8.42138827e-01 -7.61998296e-01 -4.60335910e-01 -1.05593538e+00
-3.14036608e-01 6.15511179e-01 -2.42022097e-01 4.35345769e-01
5.48755527e-01 4.10557002e-01 -1.44345772e+00 -2.69204050e-01
1.56374538e+00 1.30536866e+00 5.67901023e-02 4.24926192e-01
5.91732599e-02 1.46448791e-01 -4.59919393e-01 -3.07582736e-01
4.47367102e-01 -4.87314641e-01 -6.62022769e-01 3.71929288e-01
3.31442177e-01 1.38681429e-02 5.05961597e-01 1.36690569e+00
5.67734659e-01 4.06324983e-01 9.96876299e-01 -1.64736438e+00
-6.53404772e-01 2.87727654e-01 -8.93491745e-01 4.56368662e-02
2.02792659e-01 -8.33672062e-02 1.10463679e+00 -7.66402125e-01
2.17245266e-01 1.14651668e+00 4.64766711e-01 3.59172136e-01
-1.80267608e+00 -3.18742961e-01 4.10763025e-02 1.23848803e-01
-1.66949487e+00 -2.21518353e-01 5.29911578e-01 -5.44079661e-01
3.75281036e-01 5.90506077e-01 3.42757314e-01 5.19192398e-01
1.88819945e-01 7.57003427e-01 1.29570210e+00 -4.47718501e-01
6.48147821e-01 4.09853965e-01 4.52436805e-01 6.77252829e-01
6.81236863e-01 -1.38634974e-02 -3.00660998e-01 -7.24672079e-01
6.91181183e-01 -2.95607924e-01 -4.77876991e-01 -6.01231575e-01
-8.60816479e-01 1.22924960e+00 -3.76762062e-01 -2.17921063e-01
-6.30699277e-01 -3.49620014e-01 1.59343496e-01 2.81146586e-01
8.32552671e-01 4.10930812e-01 -4.99150515e-01 1.23510063e-01
-9.41967785e-01 2.69418389e-01 9.26042795e-01 8.69846880e-01
4.73915279e-01 3.76725942e-02 -5.60206294e-01 8.89631331e-01
3.34308237e-01 7.55428255e-01 -5.68105765e-02 -8.79278839e-01
3.77909720e-01 5.23918211e-01 2.22955197e-01 -8.23448122e-01
-2.94720650e-01 -5.54804623e-01 -7.94669151e-01 3.69389087e-01
5.76195896e-01 -2.51966059e-01 -5.04173934e-01 1.57381153e+00
5.52050829e-01 2.14021385e-01 -1.29564837e-01 1.00532472e+00
1.83047950e-01 5.06006241e-01 -2.25502014e-01 -9.20277536e-01
9.78944361e-01 -6.12130940e-01 -6.98732436e-01 1.27921268e-01
4.74077523e-01 -3.85864675e-01 7.58957386e-01 5.52416384e-01
-1.20628726e+00 -1.70419022e-01 -7.62241900e-01 4.19126898e-01
6.62466213e-02 5.98557182e-02 5.26621699e-01 8.88197124e-01
-8.81596208e-01 3.55504334e-01 -5.53378105e-01 2.15128213e-01
2.91797072e-01 6.30830050e-01 -3.53903472e-01 2.23557919e-01
-6.90750778e-01 4.69678283e-01 2.67473429e-01 2.19813675e-01
-3.24956119e-01 -1.14744556e+00 -8.43152642e-01 1.21298224e-01
5.31037033e-01 -5.84294200e-01 8.23913813e-01 -8.74623716e-01
-1.45010900e+00 8.59975815e-01 -3.43858510e-01 -4.51498240e-01
6.37961209e-01 9.40105468e-02 -4.10692282e-02 2.06056938e-01
2.32484221e-01 1.94085259e-02 7.10965633e-01 -9.93458986e-01
-6.85976386e-01 -5.32099009e-01 -2.86612988e-01 -8.90701786e-02
-1.20758913e-01 3.16464864e-02 -1.88140795e-01 -6.46016359e-01
1.16184108e-01 -7.14340389e-01 -5.36802292e-01 -8.13329965e-02
-5.29183745e-01 -2.20354408e-01 2.48145998e-01 -7.27096975e-01
1.29024208e+00 -2.14289546e+00 2.90117294e-01 6.28687680e-01
3.17746997e-02 -1.20085217e-01 -3.68607081e-02 -8.26028362e-02
-3.30930986e-02 7.44640455e-02 -4.30370033e-01 -1.47782311e-01
2.49531254e-01 -9.53543112e-02 -1.11962289e-01 9.17505205e-01
3.03360671e-01 4.28942055e-01 -6.81036532e-01 -4.66637462e-01
-4.83308136e-02 2.06847951e-01 -5.92101038e-01 4.25712347e-01
3.75370681e-02 5.62711172e-02 -6.63119674e-01 8.31127346e-01
1.17830849e+00 -1.43667758e-01 6.96252957e-02 1.33045167e-01
-1.71567887e-01 -2.15646029e-01 -1.70994687e+00 8.11758816e-01
-1.52740002e-01 2.92454064e-01 4.01760429e-01 -1.48613822e+00
1.06128132e+00 -1.67020019e-02 6.51656449e-01 -3.91469300e-01
7.32032284e-02 3.80463660e-01 -3.79443049e-01 -5.35268068e-01
1.95404768e-01 -3.49687368e-01 -1.07180759e-01 1.09693989e-01
-2.69987077e-01 1.71802357e-01 1.72019735e-01 -3.53650808e-01
8.16457331e-01 -3.88074160e-01 6.63464010e-01 -8.38496864e-01
6.44462705e-01 -1.83759287e-01 8.57712269e-01 9.78237331e-01
-2.35840976e-01 3.96784157e-01 9.54553068e-01 7.50994775e-03
-5.15439153e-01 -1.08528852e+00 -7.51507759e-01 6.74371183e-01
-4.67002057e-02 3.80911291e-01 -5.92401624e-01 -5.57616055e-01
5.35240769e-01 8.48442972e-01 -7.49929667e-01 3.86689641e-02
-1.31712750e-01 -9.63360250e-01 6.29219040e-02 2.02354580e-01
2.20224544e-01 -2.40493611e-01 -4.79894161e-01 -9.05104131e-02
8.00677538e-02 -5.94181418e-01 -5.19603074e-01 4.32383299e-01
-7.70871043e-01 -1.13365304e+00 -8.89206409e-01 -8.44237089e-01
8.27660978e-01 1.13718688e-01 9.98765171e-01 -2.46434227e-01
-2.64240265e-01 3.74240428e-01 -9.59103853e-02 -1.37995556e-01
6.36086836e-02 -2.40434676e-01 1.32085364e-02 3.14248413e-01
3.85252982e-01 -7.94580393e-03 -3.56238306e-01 4.93400455e-01
-7.84625173e-01 -4.88887280e-01 5.16980171e-01 1.15390921e+00
1.05514872e+00 2.21976519e-01 6.16952300e-01 -8.95427823e-01
6.55025244e-01 -7.60626674e-01 -1.24678683e+00 5.61552465e-01
-8.87998760e-01 2.80989975e-01 5.34278274e-01 -7.30072379e-01
-6.68360054e-01 -2.23625988e-01 5.01891971e-01 -3.49431664e-01
4.92684431e-02 6.28024459e-01 -3.83046806e-01 -3.48442942e-01
1.09398514e-01 2.62846261e-01 1.53521612e-01 -2.67671049e-01
-1.09029531e-01 6.21340930e-01 3.02638263e-01 -8.53808403e-01
5.85371971e-01 2.17352718e-01 4.71041679e-01 -8.02547693e-01
-7.27556944e-01 -4.78543013e-01 -3.40734214e-01 -4.77621593e-02
8.02084029e-01 -4.38716263e-01 -1.02313781e+00 1.43330932e-01
-8.18574488e-01 -2.57005125e-01 -3.54268849e-01 8.22278142e-01
-9.12413299e-01 2.93712258e-01 -2.25780997e-02 -1.26094186e+00
-1.98659271e-01 -1.29247594e+00 1.05021787e+00 2.00713143e-01
8.17122385e-02 -1.24144483e+00 1.33947641e-01 1.76256508e-01
1.17776230e-01 5.60070872e-01 9.00049210e-01 -8.02205920e-01
-5.33837914e-01 -2.56913215e-01 -1.81245804e-01 3.16431761e-01
-1.33035302e-01 -3.09285913e-02 -4.86256778e-01 -3.70931268e-01
1.55193210e-01 -7.90849403e-02 5.74658096e-01 1.07655716e+00
1.31869102e+00 -5.63349903e-01 -2.96668321e-01 9.18352842e-01
1.61228538e+00 2.34495312e-01 1.99679732e-01 3.52957696e-01
2.77205277e-02 7.50010908e-01 1.05938625e+00 9.26048756e-01
4.60377038e-02 7.10631847e-01 9.31956768e-02 6.57859892e-02
6.79732025e-01 3.49328667e-01 4.39926796e-02 3.37144166e-01
2.10165322e-01 -1.51136354e-01 -6.83424413e-01 4.70832080e-01
-1.91804576e+00 -9.99287367e-01 -3.02271605e-01 2.83701968e+00
8.77080142e-01 -8.48512799e-02 2.06172019e-01 6.86758608e-02
8.71937454e-01 -2.87203074e-01 -6.14637911e-01 -5.85655749e-01
-2.39520833e-01 3.57334465e-01 8.19876790e-01 7.36509919e-01
-1.04896867e+00 1.65818512e-01 7.35402012e+00 1.10275173e+00
-7.89585531e-01 -2.63841301e-01 8.31305563e-01 -1.60794556e-01
-4.22513992e-01 -1.39681092e-02 -9.31312740e-01 6.62676215e-01
8.94847035e-01 -5.82560778e-01 2.94465989e-01 9.14771914e-01
4.81514215e-01 -3.05397958e-01 -9.41871881e-01 8.70055079e-01
-3.96060236e-02 -1.08864701e+00 -4.42445785e-01 4.62971598e-01
8.61303091e-01 -8.12671959e-01 2.55100012e-01 2.67400950e-01
2.29931697e-02 -9.86602187e-01 5.26470721e-01 7.13258684e-01
7.62769699e-01 -1.08584666e+00 7.51767218e-01 1.41515985e-01
-9.55305398e-01 -1.86561551e-02 -5.75482249e-01 1.82525173e-01
-1.76331997e-02 1.37957227e+00 -6.23972595e-01 4.14420545e-01
3.47951591e-01 2.44845092e-01 -1.35737628e-01 1.54794931e+00
1.50613979e-01 5.11578083e-01 -3.33079576e-01 -2.42964610e-01
-9.08358544e-02 -8.79496574e-01 6.83895528e-01 1.00914979e+00
3.30299199e-01 4.62737717e-02 5.01104355e-01 1.00060511e+00
2.86829203e-01 4.26974684e-01 -3.41506124e-01 2.08698124e-01
6.12830043e-01 9.02535915e-01 -7.08724856e-01 -3.67889404e-02
-3.70008588e-01 8.21630895e-01 2.64294356e-01 4.65337366e-01
-8.08849514e-01 -5.93520522e-01 8.75391364e-01 -1.12298116e-01
2.81543463e-01 -4.03455645e-02 -5.23422122e-01 -9.24165189e-01
3.71388346e-01 -7.45761335e-01 6.01323605e-01 -4.28868383e-02
-1.54632950e+00 4.93462980e-02 2.32797533e-01 -1.10889685e+00
-1.14465795e-01 -8.13635409e-01 -5.32882810e-01 1.09146047e+00
-1.40778542e+00 -5.17732024e-01 1.31351918e-01 3.16775203e-01
3.26264173e-01 8.28461722e-03 5.33659279e-01 2.55265534e-01
-9.18789864e-01 9.93604720e-01 6.29832268e-01 -2.97283500e-01
6.30631387e-01 -1.71599305e+00 -5.49036026e-01 6.51925623e-01
-7.48401284e-01 4.93266940e-01 9.19268548e-01 -5.77023804e-01
-1.61194110e+00 -8.67223144e-01 8.05077791e-01 1.42034695e-01
8.19646060e-01 1.02051102e-01 -8.86607707e-01 2.81194359e-01
-3.77803802e-01 6.09435886e-02 1.15219045e+00 2.85657436e-01
4.76147793e-02 -1.52457029e-01 -1.50830042e+00 2.52556622e-01
5.78029633e-01 -1.27610102e-01 -2.00039342e-01 5.45143247e-01
3.71792644e-01 -1.87149242e-01 -1.16367733e+00 3.29475701e-01
3.89937311e-01 -7.42119312e-01 8.89193296e-01 -7.39299953e-01
5.24863005e-02 -1.19122379e-01 -2.56747544e-01 -1.29966664e+00
-4.58192408e-01 -7.54070938e-01 4.77640592e-02 1.16218615e+00
5.10123312e-01 -9.58449423e-01 7.66039193e-01 1.02493060e+00
1.80861861e-01 -9.75215197e-01 -1.26001978e+00 -1.19227898e+00
7.90077075e-02 -3.82457674e-01 2.93040156e-01 8.52422595e-01
1.26619592e-01 -1.87373295e-01 -3.39960188e-01 4.41745162e-01
1.19201791e+00 3.97575378e-01 5.56661129e-01 -1.29971540e+00
-6.04287207e-01 -5.36136329e-01 -6.14274800e-01 -6.85092509e-01
6.72777176e-01 -8.32774580e-01 1.30625993e-01 -1.21876955e+00
4.14434552e-01 -4.62701708e-01 -1.65208682e-01 2.32157648e-01
-5.79456091e-01 -2.85030603e-01 -2.78647870e-01 -2.78924972e-01
-4.97663289e-01 4.86050218e-01 9.21843529e-01 -4.20602560e-02
-5.45953453e-01 4.79918748e-01 -8.11787307e-01 3.30406636e-01
6.81950927e-01 -5.42232215e-01 -3.40389788e-01 1.29042923e-01
-1.68657005e-02 5.15156329e-01 1.50276750e-01 -6.10062659e-01
-1.44788936e-01 -9.61767256e-01 2.79148251e-01 -5.48504233e-01
5.28671555e-02 -8.48248482e-01 1.34622574e-01 4.46626455e-01
-5.58736444e-01 -1.02678768e-01 -2.27828342e-02 5.12308180e-01
-1.38726398e-01 -5.26260376e-01 1.04262304e+00 5.14799476e-01
-2.30671763e-01 2.39300236e-01 -3.61125827e-01 2.09106579e-01
1.48681927e+00 -1.99580103e-01 -2.28168041e-01 -1.11815050e-01
-5.20272374e-01 5.40023565e-01 1.84823021e-01 -1.52140662e-01
6.93359971e-01 -1.35800493e+00 -8.90367150e-01 2.32968047e-01
3.55950259e-02 -2.76341110e-01 1.13015056e-01 1.24212348e+00
-2.27480024e-01 4.50810164e-01 6.48872256e-02 -5.36898017e-01
-1.22251511e+00 6.31490290e-01 2.43438318e-01 -3.22038174e-01
-1.20282628e-01 7.94780850e-01 2.97048921e-03 -2.88741976e-01
1.38997912e-01 -2.50199229e-01 5.18134236e-02 1.71390593e-01
7.70843744e-01 9.40983474e-01 -2.81325243e-02 -2.39718825e-01
-4.09241945e-01 3.15293342e-01 3.20111185e-01 -9.96191576e-02
1.24910676e+00 3.71163562e-02 -2.94910938e-01 3.41579169e-01
1.43623555e+00 1.45330310e-01 -1.12137473e+00 -9.34335515e-02
1.18879206e-01 -8.08506191e-01 1.70385987e-01 -3.72358948e-01
-9.00705576e-01 3.39843512e-01 5.91637135e-01 3.05528462e-01
1.25987613e+00 -2.00899050e-01 1.83973759e-01 1.26062080e-01
3.45661730e-01 -1.19260275e+00 -4.01149631e-01 2.32343346e-01
5.60483694e-01 -1.22838438e+00 4.92732339e-02 -5.30920804e-01
-6.03601038e-01 1.17474711e+00 5.65461218e-01 -1.20680414e-01
7.95342147e-01 4.33982402e-01 -3.45522553e-01 2.21032694e-01
-7.12419808e-01 -2.92694896e-01 6.89861715e-01 7.29767025e-01
2.83285081e-01 4.09351796e-01 -1.02633667e+00 6.98887527e-01
2.05423623e-01 -1.87570035e-01 2.55233437e-01 7.25795686e-01
-6.17342234e-01 -8.83105516e-01 -8.65816295e-01 9.10553396e-01
-4.39295858e-01 1.34083018e-01 -2.32304484e-01 5.96059084e-01
-3.49298120e-01 1.05804920e+00 9.63344276e-02 -5.37093282e-02
2.30897933e-01 7.10520670e-02 4.13266391e-01 -3.33717585e-01
-1.12673275e-01 2.13496506e-01 -1.28743276e-01 -4.43417639e-01
-1.76076040e-01 -7.81365931e-01 -9.06883895e-01 -2.70942181e-01
-5.86998045e-01 5.54288030e-01 5.12445450e-01 8.40708077e-01
1.52370736e-01 1.91659942e-01 9.47792292e-01 -4.08462256e-01
-1.09763420e+00 -5.03104866e-01 -1.09066820e+00 1.35563329e-01
3.48711222e-01 -8.05260718e-01 -6.46282017e-01 -3.62733573e-01] | [7.417232036590576, 4.351412773132324] |
e2053b6a-bd7b-4596-9257-f183416acf46 | logg3d-net-locally-guided-global-descriptor | 2109.08336 | null | https://arxiv.org/abs/2109.08336v3 | https://arxiv.org/pdf/2109.08336v3.pdf | LoGG3D-Net: Locally Guided Global Descriptor Learning for 3D Place Recognition | Retrieval-based place recognition is an efficient and effective solution for re-localization within a pre-built map, or global data association for Simultaneous Localization and Mapping (SLAM). The accuracy of such an approach is heavily dependent on the quality of the extracted scene-level representation. While end-to-end solutions - which learn a global descriptor from input point clouds - have demonstrated promising results, such approaches are limited in their ability to enforce desirable properties at the local feature level. In this paper, we introduce a local consistency loss to guide the network towards learning local features which are consistent across revisits, hence leading to more repeatable global descriptors resulting in an overall improvement in 3D place recognition performance. We formulate our approach in an end-to-end trainable architecture called LoGG3D-Net. Experiments on two large-scale public benchmarks (KITTI and MulRan) show that our method achieves mean $F1_{max}$ scores of $0.939$ and $0.968$ on KITTI and MulRan respectively, achieving state-of-the-art performance while operating in near real-time. The open-source implementation is available at: https://github.com/csiro-robotics/LoGG3D-Net. | ['Clinton Fookes', 'Sridha Sridharan', 'Peyman Moghadam', 'Milad Ramezani', 'Kavisha Vidanapathirana'] | 2021-09-17 | null | null | null | null | ['3d-place-recognition'] | ['computer-vision'] | [-3.17001820e-01 -4.99787867e-01 -1.38498962e-01 -6.24287605e-01
-1.32560229e+00 -6.13162398e-01 5.98168969e-01 3.56590658e-01
-6.49575055e-01 4.93347853e-01 -1.06157094e-01 9.35304072e-03
-3.46917331e-01 -7.00258791e-01 -1.10854411e+00 -2.91641504e-01
-4.74553436e-01 4.96908754e-01 2.18655720e-01 -1.92815512e-02
4.88062501e-01 8.10040772e-01 -1.59139454e+00 -2.71069497e-01
6.03498101e-01 1.17197037e+00 3.41562182e-01 4.99366373e-01
1.82840452e-01 6.32107735e-01 -4.64306064e-02 3.12827863e-02
5.79892278e-01 1.92133352e-01 -7.48594344e-01 -8.20727348e-02
8.66945684e-01 -9.26524475e-02 -3.90625626e-01 9.88656342e-01
5.73243916e-01 4.44452405e-01 3.15433681e-01 -1.17298508e+00
-6.53366268e-01 1.09103890e-02 -4.52337444e-01 3.93479690e-02
4.83059287e-01 6.92076832e-02 1.12625337e+00 -1.14535844e+00
7.32595444e-01 9.82962072e-01 8.72795701e-01 -2.34533064e-02
-1.17881000e+00 -7.76181459e-01 2.55971074e-01 1.16515197e-01
-1.99951267e+00 -7.43075967e-01 6.05591416e-01 -1.30978778e-01
1.26706088e+00 4.31067944e-02 3.02322388e-01 5.21197379e-01
3.45889002e-01 6.76437557e-01 9.46547508e-01 -1.53918624e-01
3.17114294e-01 -1.93779707e-01 -1.48969233e-01 8.85434747e-01
1.22190826e-01 2.87684888e-01 -9.36435103e-01 -4.13335338e-02
7.74297476e-01 1.35657534e-01 3.63714546e-02 -8.70600462e-01
-1.31481409e+00 6.69188619e-01 1.23696673e+00 1.39013469e-01
-5.89721322e-01 6.57953620e-01 1.79376513e-01 3.55897307e-01
4.71525729e-01 4.28352773e-01 -2.74594188e-01 -2.44733155e-01
-8.95474792e-01 4.62860823e-01 4.38652456e-01 1.24896407e+00
1.28706563e+00 -2.91007042e-01 3.70325804e-01 8.12161863e-01
2.85963655e-01 8.03933084e-01 1.84156597e-01 -8.94793630e-01
4.00978446e-01 4.99693602e-01 2.88139015e-01 -1.28175879e+00
-5.84866941e-01 -5.62419593e-01 -5.17622054e-01 1.56284571e-01
-9.16956067e-02 3.82902652e-01 -1.17931736e+00 1.75382888e+00
3.31978768e-01 4.45547640e-01 -8.89873430e-02 9.57413256e-01
5.00423670e-01 5.18547475e-01 -1.67221442e-01 6.03866756e-01
9.00846899e-01 -1.05776966e+00 -1.40824169e-01 -8.24374437e-01
7.22446442e-01 -7.71697104e-01 7.51147211e-01 3.24599408e-02
-6.93869531e-01 -5.02193570e-01 -1.11549938e+00 -3.23905230e-01
-5.32474279e-01 1.99199226e-02 6.41957164e-01 4.02573459e-02
-1.38433540e+00 5.04144192e-01 -1.16037297e+00 -6.68785036e-01
4.84331191e-01 5.65883100e-01 -9.02711630e-01 -4.66698438e-01
-6.24151409e-01 1.03817737e+00 2.60156840e-01 2.35080495e-01
-9.19774413e-01 -5.73998749e-01 -9.94343221e-01 -2.21912608e-01
5.02302684e-02 -4.11352396e-01 1.10974717e+00 -2.87524372e-01
-1.14983463e+00 9.15748775e-01 -2.79116482e-01 -5.67804992e-01
3.59992802e-01 -4.40656334e-01 -2.01134786e-01 -1.38675049e-01
6.08051240e-01 1.01505411e+00 2.50381440e-01 -1.23624265e+00
-1.02310193e+00 -5.28603733e-01 -5.85387386e-02 2.38525912e-01
2.17457160e-01 -2.50691950e-01 -6.71059310e-01 -2.02406049e-01
7.58478105e-01 -1.35097003e+00 -2.80718893e-01 3.82134676e-01
-7.35775605e-02 -1.27458215e-01 4.86018807e-01 -2.93281376e-01
5.07187784e-01 -2.22300720e+00 -1.18085086e-01 3.05936068e-01
4.22121026e-02 -1.24136180e-01 -3.95470142e-01 5.94012082e-01
2.52354234e-01 -1.80965215e-01 -3.26736867e-01 -5.69885254e-01
1.67390049e-01 2.06497982e-01 -1.17629729e-01 1.08651519e+00
9.14484188e-02 9.55058455e-01 -1.05302477e+00 -7.32484832e-03
5.05259275e-01 5.98835528e-01 -5.44656754e-01 5.49975038e-02
1.70716211e-01 2.60106146e-01 -4.49754119e-01 8.42391133e-01
7.80097723e-01 -2.36481354e-01 -2.41269976e-01 1.63172856e-01
-3.14353019e-01 5.57867944e-01 -1.25263047e+00 2.44637656e+00
-7.41947711e-01 6.62340701e-01 -9.23877209e-02 -6.29032671e-01
1.23493361e+00 -3.23468089e-01 4.58025545e-01 -1.15223479e+00
-1.93980455e-01 6.38987362e-01 -3.53228986e-01 1.99595988e-01
8.09702158e-01 2.14921564e-01 -3.30264717e-01 1.30655482e-01
1.78799555e-01 -2.05233604e-01 -5.41628040e-02 -6.17374219e-02
1.23824966e+00 3.14671695e-01 2.58733392e-01 -4.44585949e-01
2.35585153e-01 2.57069647e-01 3.92835557e-01 8.68948221e-01
-1.56992733e-01 6.03651524e-01 -2.47080594e-01 -6.39415324e-01
-8.79881859e-01 -1.07243848e+00 -2.35649481e-01 9.84487534e-01
6.74851775e-01 -3.87967676e-01 -3.84130813e-02 -4.03637826e-01
4.01493073e-01 4.91078258e-01 -5.16393900e-01 -8.38904306e-02
-5.00768661e-01 -1.70143753e-01 4.93575752e-01 5.80690801e-01
6.45287752e-01 -8.78352642e-01 -7.90306866e-01 3.76597315e-01
-1.46790706e-02 -1.13338530e+00 -2.01924011e-01 4.78922963e-01
-6.62456989e-01 -7.80020773e-01 -4.46161628e-01 -8.06327641e-01
7.08552063e-01 7.91569948e-01 9.83490348e-01 1.64288245e-02
-3.30259092e-02 3.52199882e-01 -4.43034112e-01 -2.35570490e-01
2.94661999e-01 4.10868257e-01 2.02194124e-01 -3.51482183e-01
4.61706668e-01 -7.20658183e-01 -6.22712433e-01 3.59912813e-01
-5.88267207e-01 -1.84989750e-01 5.93374670e-01 7.85675645e-01
1.01510704e+00 -4.54806387e-01 1.85476765e-01 -4.57451284e-01
2.58741200e-01 -3.67311537e-01 -9.41029549e-01 3.74238864e-02
-7.62269437e-01 9.23253745e-02 2.67806679e-01 1.35461718e-01
-3.15229356e-01 4.22005087e-01 -1.95786148e-01 -5.84206939e-01
-2.17226535e-01 7.68484175e-01 2.89922893e-01 -6.47906482e-01
7.25553334e-01 4.61913139e-01 -2.12262481e-01 -5.03956020e-01
3.80474031e-01 4.67640758e-01 7.93232381e-01 -6.20038331e-01
1.06550980e+00 5.75208664e-01 -7.21679348e-03 -4.90949303e-01
-7.87355185e-01 -9.27123427e-01 -7.00285137e-01 8.07145759e-02
4.00771499e-01 -1.34506607e+00 -4.67050076e-01 3.95125121e-01
-8.54509711e-01 -4.40407038e-01 -6.15906902e-02 4.52261299e-01
-7.55393386e-01 -4.79371548e-02 -1.09197795e-01 -5.56718290e-01
-2.97218114e-01 -1.06577003e+00 1.51511693e+00 1.41094625e-01
3.42735322e-03 -7.62923777e-01 4.00532097e-01 8.18131045e-02
6.52783513e-01 3.93199593e-01 1.79047719e-01 -4.72684652e-01
-1.08197677e+00 -6.86236680e-01 -3.75013143e-01 -4.42066649e-03
1.36167660e-01 -5.05357742e-01 -1.03019452e+00 -5.97176969e-01
-5.17811954e-01 -4.27749515e-01 8.81490290e-01 1.07756704e-02
8.74763131e-01 7.88478646e-03 -4.36089635e-01 9.08876181e-01
1.77927780e+00 -8.00137296e-02 5.03977418e-01 7.63363242e-01
6.29502177e-01 1.49049729e-01 9.51125145e-01 4.08943713e-01
7.18882561e-01 8.13322008e-01 7.04773128e-01 -3.22194807e-02
-3.57846543e-02 -5.34702361e-01 1.85807332e-01 5.33412457e-01
2.26915747e-01 -1.08487904e-01 -1.31095004e+00 1.02858472e+00
-2.12776208e+00 -6.24901354e-01 2.42323816e-01 2.38992953e+00
3.33044261e-01 3.77089158e-03 -2.71242738e-01 -3.96799058e-01
4.55445707e-01 4.16921943e-01 -5.49400687e-01 -1.01129180e-02
-5.01069464e-02 1.66420564e-01 1.03983319e+00 7.03212023e-01
-1.31937432e+00 1.28878164e+00 4.98618984e+00 6.20430350e-01
-1.39016140e+00 -5.60874166e-03 2.04668105e-01 -1.49113923e-01
4.76054586e-02 1.23469688e-01 -7.32056081e-01 1.64574862e-01
9.30765986e-01 -1.08330421e-01 5.56833029e-01 1.10298336e+00
-1.16498373e-01 -3.03230286e-01 -1.15144503e+00 1.24815702e+00
-1.16030969e-01 -1.56652665e+00 -2.93849230e-01 2.69541264e-01
7.64520049e-01 1.06291568e+00 9.00911074e-03 2.83916235e-01
3.63172174e-01 -9.41861808e-01 1.12826216e+00 2.59945095e-01
8.38218689e-01 -9.56827044e-01 7.53042877e-01 2.95951992e-01
-1.47379100e+00 1.24805160e-01 -5.99069297e-01 -5.40226400e-02
-5.36997952e-02 4.05519605e-01 -9.36965346e-01 7.52052307e-01
9.23810303e-01 9.29031670e-01 -7.85620809e-01 1.32483733e+00
-1.88952267e-01 -3.18308151e-03 -8.13906312e-01 7.30377138e-02
7.30397046e-01 2.23975226e-01 3.90701145e-01 1.14298534e+00
4.84715879e-01 -1.75178215e-01 4.21910077e-01 6.94817305e-01
-2.33464271e-01 -4.35932316e-02 -8.10169995e-01 2.58090079e-01
8.28823686e-01 1.19516277e+00 -6.61278963e-01 4.04067561e-02
-1.92387834e-01 9.02535379e-01 7.38950312e-01 -2.43905839e-03
-6.59353077e-01 -4.76108015e-01 9.10184979e-01 2.02125381e-03
3.76128912e-01 -6.59208596e-01 -2.01521158e-01 -9.79020000e-01
2.37422869e-01 -4.63831037e-01 1.03838779e-02 -6.29659176e-01
-1.19591451e+00 6.46969736e-01 -2.79636025e-01 -1.36511803e+00
-2.54002512e-01 -4.74110276e-01 -1.50716543e-01 8.80385399e-01
-1.87224746e+00 -1.41648352e+00 -6.60053372e-01 5.00683665e-01
2.38870904e-01 1.06885731e-01 9.07173693e-01 4.72088188e-01
-1.91893522e-02 5.23889720e-01 4.24209476e-01 7.59305060e-02
8.11275184e-01 -1.06089020e+00 8.39011431e-01 9.46690261e-01
4.81145531e-01 6.88330710e-01 5.01233876e-01 -4.19103712e-01
-1.74010801e+00 -1.27437139e+00 1.16030729e+00 -5.77248394e-01
5.23056984e-01 -5.22501290e-01 -6.38665497e-01 8.05459976e-01
-2.31977150e-01 5.62137246e-01 2.58711576e-01 3.38205487e-01
-5.34346879e-01 -4.02806997e-01 -1.09743011e+00 3.09333324e-01
1.35194421e+00 -8.51626992e-01 -1.51243299e-01 4.54036206e-01
4.98154312e-01 -8.25399458e-01 -8.17236364e-01 3.87359411e-01
5.17509460e-01 -8.72615933e-01 9.89671111e-01 -1.10407680e-01
1.33373901e-01 -6.38598740e-01 -6.85289860e-01 -1.17740333e+00
-6.16029441e-01 -2.87826538e-01 3.18508625e-01 8.60313535e-01
3.59828919e-01 -7.55777419e-01 6.88248694e-01 5.03082275e-01
-2.77162671e-01 -6.47893608e-01 -1.28270996e+00 -1.05902588e+00
-1.60727873e-01 -5.91394722e-01 6.57677710e-01 8.66423368e-01
-4.24449533e-01 -1.25073397e-03 -2.76271999e-01 6.52438343e-01
6.77351594e-01 3.27435404e-01 1.11603487e+00 -1.18244612e+00
2.75514811e-01 -2.71588057e-01 -9.69291627e-01 -1.26270294e+00
2.19110295e-01 -1.09984577e+00 5.52787244e-01 -1.72325253e+00
-3.41105908e-02 -9.88049686e-01 -6.20812297e-01 7.54568160e-01
2.88558990e-01 4.66803402e-01 2.42778078e-01 4.71189320e-01
-1.14745986e+00 6.56391561e-01 5.56373656e-01 -8.03042129e-02
-1.22069247e-01 -2.89563507e-01 -4.51910973e-01 3.51120919e-01
7.23442674e-01 -7.78091371e-01 -7.09604919e-02 -9.48697090e-01
1.44692108e-01 -2.14107215e-01 6.70564651e-01 -1.58981633e+00
6.08195782e-01 -4.50048223e-02 3.00568104e-01 -8.20635974e-01
5.20627141e-01 -9.32298660e-01 1.81228250e-01 3.36218148e-01
-2.95600712e-01 4.36394244e-01 3.10877919e-01 6.71801984e-01
-3.51027817e-01 1.55244812e-01 7.10868716e-01 -3.70034203e-02
-1.27764595e+00 5.36864161e-01 2.28697926e-01 -2.37831116e-01
7.96185553e-01 -2.05141231e-01 -2.59616643e-01 -2.55078137e-01
-1.17807612e-01 3.85312378e-01 9.73151088e-01 6.82787418e-01
6.86740696e-01 -1.37697089e+00 -6.55631244e-01 1.48932576e-01
6.45821095e-01 3.31278563e-01 1.95575189e-02 7.17249811e-01
-9.78384316e-01 6.38622642e-01 -2.90513456e-01 -9.23285007e-01
-7.66082287e-01 1.99462950e-01 3.80743086e-01 -3.21637876e-02
-6.89631283e-01 1.00113380e+00 -7.00805113e-02 -9.78682160e-01
2.07562163e-01 -3.19411069e-01 6.19044363e-01 -3.66760701e-01
2.94088215e-01 1.50672078e-01 3.50351542e-01 -8.97951305e-01
-1.02160692e+00 7.51940072e-01 -3.93490903e-02 -8.02686885e-02
1.66042852e+00 -3.06362063e-01 -8.53128508e-02 3.17645937e-01
1.45713675e+00 -7.89926648e-02 -1.33345830e+00 -5.19426942e-01
3.03272426e-01 -7.19542623e-01 3.19485575e-01 -7.05706298e-01
-8.85444403e-01 5.06431639e-01 9.35839236e-01 -3.50443095e-01
7.77937591e-01 1.26226172e-01 7.24046946e-01 9.70169246e-01
1.14430678e+00 -8.16333175e-01 -2.56411523e-01 8.09495211e-01
8.41121733e-01 -1.45618761e+00 2.38152072e-01 9.66050662e-03
-2.58323312e-01 7.34847963e-01 4.15883183e-01 -6.91682279e-01
4.78812754e-01 2.63942722e-02 1.85451657e-01 -3.59959424e-01
-5.52802324e-01 -3.59884173e-01 3.35152239e-01 5.42955279e-01
1.74563766e-01 1.24192245e-01 3.20828646e-01 -8.10996294e-02
-2.93050230e-01 3.32173593e-02 -7.71247828e-03 1.31080091e+00
-6.36744976e-01 -8.15411925e-01 -1.73475042e-01 1.04948282e-01
-2.47286912e-02 3.03484499e-02 -2.75785506e-01 8.50807250e-01
-5.91496229e-02 7.43653178e-01 1.84106320e-01 -4.59827423e-01
3.85519236e-01 -1.69634715e-01 3.77478272e-01 -5.14238000e-01
-3.30096632e-01 -1.46151900e-01 3.79506014e-02 -1.10960817e+00
-3.76429975e-01 -6.85225666e-01 -1.34550297e+00 -5.07290840e-01
-1.29964456e-01 1.85239315e-02 1.00527298e+00 7.09063232e-01
9.03236151e-01 7.01358542e-02 5.98065317e-01 -1.24714649e+00
-3.21390748e-01 -7.54696250e-01 -4.59016383e-01 1.47190109e-01
5.19608438e-01 -8.11610401e-01 2.15832535e-02 -3.55905831e-01] | [7.616470813751221, -2.047046422958374] |
9c2472a3-7cf4-4677-ae64-5c418ec77b17 | exploration-on-grounded-word-embedding | 1809.02765 | null | http://arxiv.org/abs/1809.02765v1 | http://arxiv.org/pdf/1809.02765v1.pdf | Exploration on Grounded Word Embedding: Matching Words and Images with Image-Enhanced Skip-Gram Model | Word embedding is designed to represent the semantic meaning of a word with
low dimensional vectors. The state-of-the-art methods of learning word
embeddings (word2vec and GloVe) only use the word co-occurrence information.
The learned embeddings are real number vectors, which are obscure to human. In
this paper, we propose an Image-Enhanced Skip-Gram Model to learn grounded word
embeddings by representing the word vectors in the same hyper-plane with image
vectors. Experiments show that the image vectors and word embeddings learned by
our model are highly correlated, which indicates that our model is able to
provide a vivid image-based explanation to the word embeddings. | ['Ruixuan Luo'] | 2018-09-08 | null | null | null | null | ['learning-word-embeddings'] | ['methodology'] | [-4.71886247e-01 1.18476331e-01 -3.51974458e-01 -2.42427886e-01
-7.17163607e-02 -2.42624134e-01 7.07098424e-01 4.38048169e-02
-3.50701690e-01 1.90941215e-01 7.19358742e-01 -3.01950336e-01
3.49335104e-01 -8.12911630e-01 -5.34183204e-01 -5.82293570e-01
1.11262007e-02 -5.41272275e-02 -8.71994346e-02 -4.04852450e-01
3.20096463e-01 1.98314145e-01 -1.24664891e+00 3.76952052e-01
2.06610814e-01 5.57680607e-01 2.72753358e-01 6.18461072e-01
-7.50445962e-01 2.69964248e-01 -7.95731187e-01 -4.32266742e-01
-1.72024161e-01 -3.09269994e-01 -4.43133533e-01 1.96885705e-01
3.08685213e-01 -3.31925184e-01 -8.60297561e-01 9.88107860e-01
2.02859581e-01 6.68050721e-03 8.66788268e-01 -1.34057784e+00
-1.84229112e+00 1.73981369e-01 -5.94089270e-01 -3.09739541e-02
3.76957297e-01 1.34024218e-01 1.28358746e+00 -1.42474127e+00
6.02483869e-01 1.56420195e+00 2.40736395e-01 5.83912611e-01
-9.13622797e-01 -4.66020793e-01 4.67586637e-01 5.29595852e-01
-1.65452790e+00 3.02977830e-01 8.27140868e-01 -4.87766564e-01
9.63521779e-01 -2.53636260e-02 1.12537432e+00 1.29524207e+00
3.21697444e-01 5.13359547e-01 6.67915463e-01 -5.33734739e-01
3.78320366e-02 3.37153286e-01 4.45339754e-02 8.22338402e-01
5.37840903e-01 3.64069594e-03 -5.17522991e-01 1.34615451e-01
1.20934176e+00 7.18599916e-01 -2.10254595e-01 -5.37886977e-01
-1.36775327e+00 1.42040527e+00 9.84294534e-01 4.42938298e-01
-3.54354054e-01 6.70192838e-01 3.06383312e-01 -1.91786855e-01
4.84281242e-01 3.20639819e-01 -1.02603361e-01 8.14578086e-02
-1.85850665e-01 -6.17316775e-02 1.89281255e-01 8.98102283e-01
8.08850110e-01 3.79212558e-01 -1.07323200e-01 5.52725613e-01
7.61857092e-01 7.19232559e-01 1.00616097e+00 -1.99192926e-01
-1.39640585e-01 8.15270901e-01 -1.28487259e-01 -1.51302826e+00
-1.22428544e-01 -3.78634445e-02 -6.06315672e-01 2.14702323e-01
-3.24109167e-01 2.36620635e-01 -1.21279109e+00 1.46213198e+00
-5.33664599e-02 5.98690033e-01 2.70689785e-01 1.15856791e+00
1.35862792e+00 1.13072968e+00 2.35348985e-01 4.12882060e-01
1.96713412e+00 -1.14399350e+00 -1.19500601e+00 -5.90342760e-01
5.30789554e-01 -5.29948056e-01 1.59170973e+00 -9.78731271e-03
-4.37033802e-01 -7.81293988e-01 -1.36789215e+00 -2.90093929e-01
-9.90285218e-01 -1.73603028e-01 7.14536548e-01 5.47088146e-01
-7.94214249e-01 2.16501709e-02 -4.79420424e-01 -4.93148923e-01
2.87898183e-01 -2.32745647e-01 -7.47726440e-01 -4.23233390e-01
-1.27185810e+00 8.94282937e-01 4.02099788e-01 -7.44823441e-02
-7.32390642e-01 -2.75923729e-01 -1.44942510e+00 1.11107960e-01
-1.88832179e-01 -4.78797793e-01 6.29388988e-01 -6.16788387e-01
-9.20772076e-01 9.09530461e-01 -2.58936733e-01 8.89489353e-02
-3.51032078e-01 -3.36568445e-01 -8.00205708e-01 2.99295008e-01
-1.27458900e-01 8.60415757e-01 1.06212628e+00 -1.71108270e+00
-2.48514675e-02 -3.58785659e-01 1.40473977e-01 6.95973113e-02
-1.00168478e+00 -4.02421594e-01 -4.91838366e-01 -8.82005036e-01
3.75971268e-03 -3.94854218e-01 -3.72089446e-01 4.45336789e-01
-4.72338721e-02 -1.72098786e-01 9.93723333e-01 -4.76091623e-01
1.27685440e+00 -2.49424076e+00 -1.07441738e-01 -6.46979511e-02
5.87514818e-01 2.17825964e-01 -6.52968168e-01 7.23417699e-01
-3.17586660e-01 3.84698153e-01 8.64280164e-02 -1.25025481e-01
4.65819016e-02 8.99987102e-01 -5.68739235e-01 5.05206466e-01
3.49161893e-01 1.42053401e+00 -1.21264899e+00 -5.53304970e-01
5.45819402e-01 9.57078934e-01 -4.47670072e-01 5.42239606e-01
-8.20715427e-02 -2.56572932e-01 -5.18085480e-01 8.18682760e-02
5.38200319e-01 -2.76421756e-01 5.54587878e-02 -4.76393849e-01
2.94692874e-01 -6.84984624e-02 -9.55722153e-01 1.94448197e+00
-6.54368222e-01 9.69502628e-01 -6.82204187e-01 -8.52761388e-01
1.14113176e+00 3.70563537e-01 -1.09145522e-01 -7.11677849e-01
7.55477622e-02 -1.71185493e-01 -3.80624771e-01 -8.74145925e-01
7.68069148e-01 -4.66554552e-01 -1.75730139e-02 4.20125544e-01
2.16577649e-01 -1.91472575e-01 -3.38274300e-01 4.01187122e-01
5.43390453e-01 -7.39756450e-02 4.19083238e-01 -8.03758278e-02
2.40501434e-01 -2.50567824e-01 8.53242725e-02 2.52781510e-01
1.40683874e-01 9.13168252e-01 3.35897684e-01 -7.48340845e-01
-1.12067318e+00 -1.20287335e+00 -6.30870685e-02 8.98970902e-01
5.62502146e-01 -9.62836802e-01 -4.75013554e-01 -5.11188328e-01
-3.38636264e-02 8.65688562e-01 -1.00266743e+00 -4.10947591e-01
-2.66183138e-01 -2.42338002e-01 1.25627220e-01 8.66101623e-01
-1.04247160e-01 -1.10729575e+00 -5.33785582e-01 5.08142412e-02
2.58170485e-01 -1.23284602e+00 -5.49547076e-01 -7.48756453e-02
-5.55035651e-01 -8.73185515e-01 -9.43880558e-01 -1.13790083e+00
1.19498432e+00 6.54654264e-01 1.15647149e+00 3.89447689e-01
-6.45130098e-01 5.59655905e-01 -9.30257976e-01 -3.81718248e-01
1.68715239e-01 -4.92596030e-01 9.60792676e-02 6.32564013e-04
8.41847360e-01 -3.29538405e-01 -8.45428109e-01 2.34965328e-02
-1.31362367e+00 3.95045877e-02 4.25223559e-01 1.02972233e+00
6.26595080e-01 -4.70286101e-01 2.09482566e-01 -4.78707552e-01
8.90785575e-01 -4.65317726e-01 -4.31772525e-04 2.30776012e-01
-6.50178254e-01 2.99819380e-01 4.74300534e-01 -7.38459349e-01
-3.89380902e-01 -3.83198947e-01 -1.30081072e-01 -7.79108584e-01
-8.18015784e-02 4.95056599e-01 -1.17483791e-02 1.07357211e-01
5.03409803e-01 3.22455674e-01 -7.46301413e-02 -4.08349425e-01
1.32184672e+00 5.45728624e-01 4.28710520e-01 -2.01497421e-01
9.31133628e-01 5.23470521e-01 -1.67753071e-01 -1.06774020e+00
-7.51025319e-01 -4.33024138e-01 -6.01129234e-01 -1.36015881e-02
1.48264837e+00 -8.89165521e-01 -3.61389428e-01 -2.23005250e-01
-1.59751725e+00 2.30223432e-01 -5.61931312e-01 6.06074214e-01
-2.56566823e-01 4.41611677e-01 -3.10762167e-01 -6.28342152e-01
-1.20400168e-01 -9.06413913e-01 1.24682486e+00 2.81825989e-01
-2.02641249e-01 -1.36805212e+00 2.30482057e-01 -2.13269770e-01
2.47145757e-01 2.35811189e-01 1.08784664e+00 -5.48069715e-01
-1.26299560e-01 -4.36503500e-01 -4.87966180e-01 2.56855994e-01
1.45926639e-01 -1.44405022e-01 -9.61892724e-01 -1.87149927e-01
-2.23262951e-01 -2.00587705e-01 7.95084834e-01 1.92450210e-01
1.37947452e+00 -3.75096083e-01 -2.36841872e-01 7.87608504e-01
1.71958947e+00 -1.44695938e-01 9.11797643e-01 1.32953197e-01
1.00672448e+00 3.21398258e-01 3.68898213e-01 4.11830723e-01
3.23511034e-01 4.70879674e-01 7.70351410e-01 -6.60570741e-01
-1.07199550e-01 -8.79379690e-01 2.03169674e-01 9.54351366e-01
2.60007530e-01 -2.70490170e-01 -7.18031168e-01 7.98931181e-01
-1.85739064e+00 -6.71731353e-01 -1.61009014e-01 1.62852907e+00
2.40800425e-01 -2.34188754e-02 -3.13311666e-01 -6.03190474e-02
5.11479437e-01 6.89053237e-01 -2.32157663e-01 -8.51522744e-01
-4.49433438e-02 4.49363708e-01 1.98287845e-01 4.72240388e-01
-4.45176661e-01 1.16421151e+00 7.12525034e+00 6.58972442e-01
-1.04000044e+00 1.24637686e-01 1.88204110e-01 1.67015076e-01
-9.59586799e-01 -1.17594436e-01 -2.53629953e-01 1.35228276e-01
6.55514061e-01 -3.62832427e-01 4.57250923e-02 1.00598216e+00
-1.22012839e-01 4.72583801e-01 -1.02876723e+00 1.50513399e+00
6.08968377e-01 -1.43103683e+00 8.12278450e-01 1.58616066e-01
5.31605244e-01 -5.48140824e-01 3.95056605e-01 2.08445549e-01
1.14927396e-01 -1.75523174e+00 4.41357553e-01 3.20842385e-01
1.12041080e+00 -7.17774093e-01 8.02149713e-01 -1.75422385e-01
-1.40982711e+00 2.21921891e-01 -9.53570545e-01 -1.64857253e-01
4.25748348e-01 3.28917503e-01 -5.39622009e-01 2.57519424e-01
3.77505481e-01 8.94913971e-01 -5.87743163e-01 4.09611523e-01
-8.17685127e-01 3.92300427e-01 2.62588382e-01 -3.68768156e-01
6.30603850e-01 -2.05416068e-01 2.60812957e-02 1.40756023e+00
3.83333594e-01 3.65947932e-01 1.58006281e-01 1.12911499e+00
1.70830805e-02 3.05760771e-01 -1.00626850e+00 -5.40616632e-01
1.86284840e-01 1.33585572e+00 -4.52239364e-01 -3.41224223e-01
-7.69030750e-01 1.29264140e+00 1.93151936e-01 4.97359037e-01
-7.16445982e-01 -5.05176187e-01 1.34989715e+00 8.01640823e-02
5.05787015e-01 -6.44975603e-01 -2.24593803e-01 -1.04761755e+00
-2.74695784e-01 -2.52150983e-01 1.28785461e-01 -1.35988164e+00
-1.52047682e+00 6.42494142e-01 -9.30370465e-02 -1.18493068e+00
-6.13520816e-02 -1.24610841e+00 -8.91462743e-01 9.03138757e-01
-1.53246009e+00 -1.27745497e+00 -5.37468672e-01 6.42257214e-01
5.44225931e-01 -1.43462569e-01 1.39291775e+00 -4.97572497e-02
-1.70289069e-01 6.15335882e-01 -9.60428640e-02 3.24551225e-01
4.33270484e-01 -1.21008551e+00 7.73455739e-01 4.15913552e-01
9.14678574e-01 8.22938085e-01 6.26555920e-01 -2.78235525e-01
-1.64588964e+00 -8.25061500e-01 7.28510261e-01 -3.41253757e-01
7.06189036e-01 -5.93200445e-01 -1.10387123e+00 6.12898648e-01
5.44207513e-01 4.38106328e-01 1.12508702e+00 1.50200455e-02
-7.66737878e-01 2.51475781e-01 -7.91897714e-01 8.21834981e-01
7.16025889e-01 -7.93396533e-01 -9.18222308e-01 3.55194151e-01
1.32285273e+00 1.57766446e-01 -6.08579576e-01 -2.57554382e-01
5.75601101e-01 -5.70275545e-01 1.38891602e+00 -1.08161938e+00
8.50151300e-01 -1.90819487e-01 -3.52376580e-01 -1.61020243e+00
-5.62677801e-01 1.74935833e-02 -2.56485611e-01 8.38878810e-01
2.02417240e-01 -5.64174533e-01 4.83833760e-01 1.27603650e-01
3.09109271e-01 -8.95363510e-01 -5.37053466e-01 -7.89878547e-01
1.70195073e-01 -4.05045241e-01 8.50257039e-01 1.06041932e+00
2.40029037e-01 5.33555567e-01 -4.04866368e-01 1.05564721e-01
3.47313732e-01 -1.30343707e-02 6.10070288e-01 -8.45906496e-01
-2.02711415e-03 -3.63167644e-01 -1.06585538e+00 -1.28736448e+00
1.26139909e-01 -8.42941821e-01 -1.48222297e-01 -1.84881282e+00
2.07036048e-01 1.68693185e-01 -5.25599539e-01 3.52627158e-01
-4.68010515e-01 5.04538417e-01 1.97551757e-01 -1.55143321e-01
-2.07685083e-01 1.00493276e+00 1.47702301e+00 -7.59186074e-02
4.05833781e-01 -1.20685816e+00 -7.84266829e-01 6.45115972e-01
5.05817115e-01 -4.18006122e-01 -4.81635928e-01 -8.12570512e-01
2.20623642e-01 -4.08229619e-01 4.19825166e-01 -3.69437724e-01
-1.96922958e-01 -3.30404818e-01 5.94160318e-01 -3.14029366e-01
5.25498450e-01 -1.03382528e+00 -2.69289136e-01 5.22744417e-01
-2.89549083e-01 5.14210343e-01 7.64183551e-02 6.55602038e-01
-4.60307598e-01 -3.12317759e-01 4.21875060e-01 -1.68129772e-01
-1.10879624e+00 4.30044830e-01 -1.69315264e-01 -7.10767582e-02
1.00305402e+00 -4.47648078e-01 -1.46658465e-01 -7.62327075e-01
-4.03145045e-01 3.62636410e-02 3.86583596e-01 9.10557747e-01
1.36537313e+00 -1.98209202e+00 -6.59712076e-01 3.85983944e-01
7.72631526e-01 -2.61420935e-01 4.53914814e-02 -1.68684021e-01
-5.34533739e-01 2.93271899e-01 -3.44568878e-01 -4.15651828e-01
-1.08730423e+00 1.09398413e+00 -1.92629099e-01 2.01816976e-01
-1.03847468e+00 8.92613471e-01 7.48762429e-01 -1.28356412e-01
6.34779185e-02 -1.53505996e-01 -4.81401742e-01 7.67903868e-04
9.02176082e-01 -2.09731072e-01 -6.25361264e-01 -7.34984875e-01
-3.52094680e-01 9.60278332e-01 1.38212591e-01 -2.69038945e-01
1.37966907e+00 5.59751652e-02 1.39532655e-01 6.66780829e-01
1.68334210e+00 -3.66111726e-01 -7.99746037e-01 -2.93754041e-01
-3.95258367e-01 -9.59988296e-01 1.27007753e-01 -2.46169984e-01
-1.21006620e+00 1.66475654e+00 8.44846725e-01 7.16380402e-02
6.36826336e-01 2.22066164e-01 9.49300349e-01 9.20018479e-02
1.85918845e-02 -6.66378796e-01 8.14973414e-01 3.37248772e-01
1.15656579e+00 -1.18263364e+00 6.87899590e-02 -3.10791761e-01
-8.18216026e-01 1.39313734e+00 6.71992660e-01 -4.46073532e-01
7.46402740e-01 -8.41603130e-02 3.55937690e-01 -5.92127025e-01
-5.81475854e-01 -3.51289600e-01 5.49198151e-01 9.30000067e-01
4.76469427e-01 2.32428864e-01 -3.91898274e-01 7.68915713e-01
-1.37081385e-01 -5.99379480e-01 4.95651245e-01 6.65670514e-01
-5.62478721e-01 -9.97317076e-01 -3.63780439e-01 9.05887596e-03
1.72703370e-01 -2.84231007e-01 -3.51230890e-01 6.35999441e-01
-1.04932697e-03 6.90423906e-01 3.85347545e-01 -5.41378081e-01
3.72476697e-01 1.38989002e-01 4.44152027e-01 -8.52992833e-01
-6.04334101e-02 -2.93439895e-01 -4.66371268e-01 -4.84525204e-01
4.08190042e-02 3.25070113e-01 -1.58771276e+00 -1.64396554e-01
-8.82668495e-02 4.94301133e-02 1.01672924e+00 7.26516962e-01
2.46795878e-01 6.10441506e-01 4.65480268e-01 -8.13075483e-01
-1.63310841e-01 -8.27647448e-01 -7.36615777e-01 9.51931000e-01
2.41943985e-01 -7.81033218e-01 -3.85988414e-01 1.41491160e-01] | [10.479748725891113, 2.0036351680755615] |
f7a384d1-35f4-486c-a4fb-74c6a8049a6f | effective-data-augmentation-with-diffusion | 2302.07944 | null | https://arxiv.org/abs/2302.07944v2 | https://arxiv.org/pdf/2302.07944v2.pdf | Effective Data Augmentation With Diffusion Models | Data augmentation is one of the most prevalent tools in deep learning, underpinning many recent advances, including those from classification, generative models, and representation learning. The standard approach to data augmentation combines simple transformations like rotations and flips to generate new images from existing ones. However, these new images lack diversity along key semantic axes present in the data. Current augmentations cannot alter the high-level semantic attributes, such as animal species present in a scene, to enhance the diversity of data. We address the lack of diversity in data augmentation with image-to-image transformations parameterized by pre-trained text-to-image diffusion models. Our method edits images to change their semantics using an off-the-shelf diffusion model, and generalizes to novel visual concepts from a few labelled examples. We evaluate our approach on few-shot image classification tasks, and on a real-world weed recognition task, and observe an improvement in accuracy in tested domains. | ['Ruslan Salakhutdinov', 'Max Gurinas', 'Kyle Doherty', 'Brandon Trabucco'] | 2023-02-07 | null | null | null | null | ['few-shot-image-classification'] | ['computer-vision'] | [ 9.27042484e-01 8.35326910e-02 -2.78977633e-01 -4.44272846e-01
-2.71155927e-02 -7.08750546e-01 1.15320754e+00 3.74889582e-01
-5.88588536e-01 7.28629649e-01 2.27258846e-01 -1.26462206e-01
2.42730621e-02 -9.76723969e-01 -8.67149055e-01 -6.46062970e-01
1.24488972e-01 4.65578943e-01 1.74343511e-01 -4.91037697e-01
2.90039837e-01 6.03751123e-01 -1.82145584e+00 3.93394142e-01
7.50150859e-01 5.59498131e-01 2.47100160e-01 7.15198815e-01
-4.50711221e-01 7.06079960e-01 -5.95512569e-01 -2.36204341e-01
2.93931097e-01 -5.60819328e-01 -6.79112136e-01 3.93504381e-01
3.75437707e-01 -3.82521786e-02 -2.14965537e-01 1.01072001e+00
2.01305822e-01 1.81559682e-01 9.69400585e-01 -1.48705375e+00
-1.24184334e+00 4.07324106e-01 -7.02983677e-01 9.46912542e-02
4.72589768e-02 1.40846416e-01 5.71064413e-01 -6.14049733e-01
1.07027578e+00 1.22032607e+00 6.76678598e-01 5.53989887e-01
-1.58643103e+00 -3.77546400e-01 3.13137293e-01 1.55401006e-01
-9.60402668e-01 -1.40843630e-01 7.86784947e-01 -5.33140242e-01
7.99241841e-01 1.68217346e-01 8.08055818e-01 1.33121288e+00
-1.80134431e-01 6.63918495e-01 1.22444391e+00 -6.75069273e-01
4.46387619e-01 1.38945714e-01 -1.96314871e-01 6.43556774e-01
2.38085195e-01 6.30947016e-03 -4.77620333e-01 -4.97352369e-02
8.92753422e-01 4.22122866e-01 6.35611415e-02 -9.37727392e-01
-1.17035484e+00 1.12019074e+00 8.17868888e-01 2.12767601e-01
-2.74867922e-01 -1.70524251e-02 3.53404850e-01 1.90854028e-01
7.47363865e-01 7.74390996e-01 -5.27664006e-01 5.49796633e-02
-7.51392186e-01 1.65394187e-01 6.53666735e-01 9.40069973e-01
1.00791681e+00 7.16327354e-02 -2.43733421e-01 8.16929758e-01
-6.38117790e-02 3.93167734e-01 7.53072083e-01 -8.68218124e-01
-3.36449221e-03 7.91837037e-01 -1.77768633e-01 -9.06771123e-01
-1.98352337e-01 -3.94957453e-01 -8.74039710e-01 3.26501399e-01
1.93723604e-01 8.89687762e-02 -1.66461694e+00 1.81952250e+00
2.56155998e-01 7.44579211e-02 -7.71704828e-03 4.06200290e-01
6.03141248e-01 5.67186773e-01 2.70878643e-01 1.23617589e-01
1.27113438e+00 -1.08161104e+00 -5.58270276e-01 -5.72158873e-01
8.15459907e-01 -6.30436420e-01 1.13238192e+00 1.74517408e-01
-6.12789750e-01 -6.54725075e-01 -1.01551378e+00 -1.09406494e-01
-1.03294897e+00 -5.83443828e-02 8.84997845e-01 6.08434200e-01
-7.30057895e-01 6.34112537e-01 -6.04853809e-01 -8.12833309e-01
7.63785839e-01 1.22101583e-01 -5.23876965e-01 -3.66777569e-01
-8.65082443e-01 1.02224219e+00 6.26961410e-01 -4.60824102e-01
-1.05053782e+00 -8.92150939e-01 -1.02110696e+00 -2.71626145e-01
5.21338642e-01 -7.51773298e-01 1.09168673e+00 -1.16308022e+00
-1.30149567e+00 1.07715321e+00 4.70270701e-02 -5.59236050e-01
4.24248099e-01 -2.75154501e-01 -1.56458959e-01 -1.31561771e-01
1.10335104e-01 1.33487833e+00 1.22246158e+00 -1.50252855e+00
-5.12035191e-01 -4.20092702e-01 3.60300057e-02 1.98752463e-01
-6.40692890e-01 -3.11944574e-01 -1.60216421e-01 -9.57741916e-01
-4.00899500e-02 -1.15841460e+00 -5.85558474e-01 2.89977252e-01
-3.33239347e-01 2.77632087e-01 1.05888176e+00 -5.06873012e-01
6.76440895e-01 -2.07883739e+00 3.16532493e-01 -5.11910357e-02
-5.23692369e-03 3.47559661e-01 -5.94563544e-01 3.67555052e-01
-3.07711244e-01 2.03183129e-01 -6.89974546e-01 -3.72848958e-01
-2.86674380e-01 6.66831970e-01 -3.57490748e-01 1.53306022e-01
7.03824818e-01 9.91305351e-01 -1.04152226e+00 -7.68728554e-02
4.09432203e-01 5.10304451e-01 -6.67867839e-01 5.56452423e-02
-6.61597431e-01 4.90937859e-01 -5.01027107e-02 3.21359307e-01
5.26200593e-01 -8.20289776e-02 -1.77842066e-01 4.75260913e-02
4.46184874e-02 -1.17600434e-01 -8.56774032e-01 2.23841166e+00
-4.80223268e-01 6.42929316e-01 -6.13058805e-01 -1.15941012e+00
1.05533135e+00 -2.61126012e-01 2.81873226e-01 -4.46143955e-01
-5.60575463e-02 4.97587062e-02 1.61491215e-01 -3.58143896e-01
7.16755569e-01 6.00174218e-02 2.91273147e-02 4.22012657e-01
3.09423566e-01 -5.75842857e-01 5.13565719e-01 2.11692229e-01
9.92003679e-01 4.93915349e-01 4.21982378e-01 -1.14345558e-01
8.29019919e-02 3.49043578e-01 2.00518608e-01 7.91418850e-01
2.00035647e-01 8.78480732e-01 4.51282412e-01 -6.83627784e-01
-1.51332700e+00 -7.85154164e-01 1.88211184e-02 1.23563159e+00
-1.72233418e-01 -2.31214061e-01 -7.34718978e-01 -7.61091948e-01
-4.28356119e-02 8.49372864e-01 -1.17871702e+00 -4.97806758e-01
-1.11898996e-01 -1.03161430e+00 4.27028656e-01 5.47474802e-01
4.35000896e-01 -1.24874461e+00 -7.10153580e-01 1.45994067e-01
1.12172268e-01 -1.11401618e+00 -8.33912715e-02 5.57126939e-01
-9.43895340e-01 -8.91777575e-01 -8.98050308e-01 -7.70982921e-01
1.12132668e+00 2.64191747e-01 9.72835302e-01 -6.87253475e-02
-8.90861034e-01 1.35818824e-01 -6.66863143e-01 -7.24067509e-01
-6.88071251e-01 3.24535817e-01 -1.70578435e-01 -1.01818562e-01
3.85557413e-01 -4.74575281e-01 -3.38657767e-01 1.09870531e-01
-1.38263667e+00 2.40732625e-01 7.28426456e-01 9.20158923e-01
5.77744782e-01 -1.79082438e-01 3.70928854e-01 -1.16043329e+00
6.58240616e-01 -2.39660814e-01 -2.53516376e-01 1.35732055e-01
-3.79147679e-01 4.11964446e-01 4.51438993e-01 -8.12034607e-01
-9.13119256e-01 3.85600001e-01 2.10728243e-01 -2.86110401e-01
-5.20564795e-01 4.40803438e-01 6.69209426e-03 -4.70392443e-02
1.10007703e+00 2.38721371e-01 1.59315184e-01 -3.62917185e-01
1.10353196e+00 2.79414117e-01 6.47095740e-01 -4.10036474e-01
7.82806098e-01 7.04177916e-01 3.16553563e-02 -9.71306086e-01
-9.05856907e-01 -1.84843615e-01 -1.10752547e+00 2.48069093e-01
8.01382720e-01 -7.19807684e-01 3.16408090e-02 5.51102102e-01
-1.17444599e+00 -3.60299587e-01 -1.02582228e+00 2.29336783e-01
-5.67712843e-01 1.12174481e-01 -1.78224221e-01 -3.83577377e-01
-1.12062944e-02 -1.00079477e+00 1.06339502e+00 3.45965266e-01
-3.50347847e-01 -9.26498950e-01 1.32257029e-01 -2.34210026e-02
4.12695259e-01 5.13838112e-01 1.12397683e+00 -7.23192513e-01
-3.35043073e-01 -2.61924118e-01 -2.20695138e-01 5.05531788e-01
4.77234989e-01 1.98799774e-01 -1.06115258e+00 -2.00723290e-01
-2.11413935e-01 -5.81330240e-01 1.16934049e+00 3.19959253e-01
1.39868939e+00 -8.87326971e-02 -4.81182933e-01 6.36002719e-01
1.18720114e+00 2.09025741e-01 7.40729511e-01 4.87825781e-01
7.54068077e-01 6.76239967e-01 2.82320648e-01 3.59267831e-01
8.43700543e-02 4.32415247e-01 6.03513062e-01 -5.21440983e-01
-3.98656368e-01 -4.43005890e-01 4.74905260e-02 2.81310529e-01
-1.10918395e-01 -2.24299133e-01 -8.62487733e-01 5.77101648e-01
-1.73921299e+00 -8.01216781e-01 1.24029957e-01 2.08555913e+00
8.43161106e-01 2.82881167e-02 -7.29631707e-02 5.66786416e-02
5.90652406e-01 2.11948171e-01 -8.16028297e-01 -3.27497691e-01
-2.54625827e-01 4.79283482e-01 5.40202856e-01 2.60757506e-01
-1.17461360e+00 1.32862127e+00 6.75875711e+00 5.87616801e-01
-1.06866944e+00 -8.91394019e-02 6.83245838e-01 4.63255309e-02
-2.74200827e-01 8.82550105e-02 -5.25055945e-01 1.33451894e-01
3.70035559e-01 -8.41546338e-03 5.20683646e-01 9.25009489e-01
-2.26280928e-01 -4.58235629e-02 -1.18785512e+00 6.95427239e-01
5.87654352e-01 -1.51424122e+00 4.71242338e-01 1.61266133e-01
1.03551853e+00 -6.88635185e-02 2.87669748e-01 3.30135018e-01
7.24910915e-01 -1.10647559e+00 3.37265491e-01 4.81369227e-01
6.56883895e-01 -6.84962034e-01 4.59638447e-01 2.23517850e-01
-7.60784090e-01 -8.09493959e-02 -4.21809435e-01 -1.86474398e-01
-1.25881150e-01 3.19321275e-01 -8.86623561e-01 3.38023335e-01
6.63383126e-01 8.67144585e-01 -1.16318309e+00 7.98978269e-01
-4.22318220e-01 1.98091462e-01 -3.34743589e-01 -2.85598394e-02
2.81451434e-01 -2.31731474e-01 2.36734152e-01 9.77789342e-01
2.92483896e-01 -4.21817899e-01 1.97744951e-01 8.89758885e-01
-3.41781765e-01 -1.04373097e-01 -1.07883930e+00 -2.37592146e-01
3.12281102e-01 1.26866651e+00 -9.23106670e-01 -5.24612963e-01
-2.24326208e-01 1.29001987e+00 1.62249908e-01 2.91786999e-01
-5.68274021e-01 -4.09952641e-01 6.25322998e-01 8.35883394e-02
2.96798855e-01 -1.53948829e-01 -2.72464395e-01 -9.46350098e-01
-1.82561994e-01 -1.15925217e+00 1.51439711e-01 -9.46385801e-01
-1.35107458e+00 4.84467268e-01 2.14349344e-01 -9.69861329e-01
-2.48099521e-01 -6.80633724e-01 -4.33508784e-01 7.02136993e-01
-1.44847965e+00 -1.48757160e+00 -6.57610714e-01 4.36801165e-01
7.03534782e-01 -3.81810099e-01 1.12885630e+00 -2.97734812e-02
-1.31867364e-01 1.41723350e-01 1.85642466e-01 2.21838132e-02
7.27608025e-01 -1.23676240e+00 1.01856482e+00 8.87950718e-01
5.25250375e-01 4.36131626e-01 7.33947694e-01 -6.23389244e-01
-8.68705273e-01 -1.13914311e+00 4.31745648e-01 -4.79248315e-01
7.48711407e-01 -5.82293987e-01 -1.04924881e+00 5.98955870e-01
3.67630094e-01 2.40642413e-01 6.92061484e-01 3.16871740e-02
-5.58895528e-01 2.53465891e-01 -1.00085402e+00 8.24483395e-01
1.33116496e+00 -4.78217393e-01 -5.06174088e-01 4.45609331e-01
7.64864981e-01 -1.92209765e-01 -3.45320702e-01 4.90630001e-01
3.54293108e-01 -5.69948077e-01 1.01014602e+00 -1.19150972e+00
7.43727028e-01 -2.26452947e-01 -1.12912893e-01 -1.91992152e+00
-3.34051937e-01 -3.58681709e-01 2.09413841e-01 1.12820625e+00
3.47279757e-01 -3.29691201e-01 9.45298374e-01 2.66437173e-01
3.11786979e-02 -2.84886748e-01 -3.95177007e-01 -7.44046271e-01
1.97963193e-01 -9.94332433e-02 5.92934430e-01 1.16304326e+00
-5.13651609e-01 4.52397108e-01 -4.14644867e-01 -1.52271554e-01
4.37760830e-01 -1.65348753e-01 1.15460634e+00 -1.42525005e+00
-2.87557933e-02 -4.21775490e-01 -6.82487488e-01 -6.54065847e-01
2.34196931e-02 -9.89454508e-01 -7.79965147e-02 -1.71071482e+00
3.57350826e-01 -3.35662276e-01 9.55754425e-03 9.16122854e-01
-1.64141878e-01 3.72153968e-01 2.91672379e-01 5.85680157e-02
-2.03967169e-01 7.96749294e-01 1.17071056e+00 -4.37446833e-01
-3.04670066e-01 -4.56365108e-01 -8.06823254e-01 9.85435665e-01
8.38391006e-01 -6.72011077e-01 -6.00157142e-01 -5.54951906e-01
9.23476219e-02 -8.33940327e-01 3.92345667e-01 -1.01655471e+00
-1.36393413e-01 -1.94787964e-01 6.93400204e-01 -1.63009316e-01
4.31008428e-01 -7.75958717e-01 -1.56573262e-02 6.76778138e-01
-5.63408434e-01 1.22067548e-01 3.57141614e-01 5.43605089e-01
-5.83648086e-02 -3.84735972e-01 8.40736389e-01 -3.38246435e-01
-9.52278435e-01 2.97720820e-01 -1.93477616e-01 1.94569323e-02
1.22966087e+00 -2.11848125e-01 -6.11289740e-01 -2.19396889e-01
-7.61405647e-01 -1.92325518e-01 7.13488400e-01 1.05009425e+00
3.70107085e-01 -1.23324978e+00 -6.58684909e-01 5.18755674e-01
5.65036476e-01 1.45417765e-01 -6.89368919e-02 2.91323334e-01
-6.88370407e-01 -1.11107342e-03 -7.64790297e-01 -7.79841602e-01
-1.35844171e+00 7.63816416e-01 1.08277239e-01 -2.10514084e-01
-5.50681412e-01 6.65734828e-01 3.03227842e-01 -6.48501813e-01
5.75578474e-02 -4.86936808e-01 -1.84876189e-01 2.79792041e-01
4.00308043e-01 -3.19924839e-02 3.98382209e-02 -4.71712828e-01
-3.56710665e-02 5.47989070e-01 -3.36964637e-01 -1.08463012e-01
1.60238492e+00 2.89035589e-01 -1.57268092e-04 5.19012213e-01
9.83880579e-01 -5.12817621e-01 -1.39978695e+00 -4.58407074e-01
3.67086791e-02 -6.22819126e-01 -9.10243019e-02 -7.84077346e-01
-8.61823797e-01 1.09256411e+00 8.46979082e-01 1.92530677e-01
8.71331632e-01 -1.36271164e-01 1.52601764e-01 7.68991768e-01
1.46847844e-01 -1.11298311e+00 5.84341586e-01 5.34668863e-01
1.01869309e+00 -1.46128845e+00 1.64060965e-01 -2.84717113e-01
-8.94236386e-01 8.94574583e-01 5.45196295e-01 -6.67822137e-02
4.70185280e-01 1.94809273e-01 3.35871875e-02 -4.54586744e-02
-5.23769438e-01 -4.62658793e-01 1.64363787e-01 1.16617596e+00
3.67310256e-01 -4.61729877e-02 7.02514648e-02 -1.23880357e-01
-1.01265050e-01 2.51186639e-02 5.39012253e-01 1.17148960e+00
-3.48207146e-01 -1.25639904e+00 -1.96616068e-01 5.31513214e-01
6.41693845e-02 -1.56244412e-01 -7.01357841e-01 9.00953770e-01
3.44690800e-01 5.00645459e-01 2.86431134e-01 -1.40308931e-01
2.91409016e-01 2.35960603e-01 6.69584692e-01 -1.01198423e+00
-2.69317746e-01 -3.47768635e-01 -3.13657016e-01 -9.07996073e-02
-6.50528312e-01 -6.52446091e-01 -9.77798939e-01 -1.63082387e-02
-1.73360750e-01 -2.55454451e-01 1.08255816e+00 9.00713861e-01
4.30648476e-01 8.84302616e-01 3.40684921e-01 -1.04445875e+00
-2.84955978e-01 -1.14857996e+00 -3.22660178e-01 8.84010077e-01
5.07864803e-02 -7.01605797e-01 -8.49595889e-02 6.29210293e-01] | [9.830095291137695, 1.9112056493759155] |
940ee4eb-e55a-4fee-84c5-2f34502a75f8 | intra-inter-camera-similarity-for | 2103.11658 | null | https://arxiv.org/abs/2103.11658v1 | https://arxiv.org/pdf/2103.11658v1.pdf | Intra-Inter Camera Similarity for Unsupervised Person Re-Identification | Most of unsupervised person Re-Identification (Re-ID) works produce pseudo-labels by measuring the feature similarity without considering the distribution discrepancy among cameras, leading to degraded accuracy in label computation across cameras. This paper targets to address this challenge by studying a novel intra-inter camera similarity for pseudo-label generation. We decompose the sample similarity computation into two stage, i.e., the intra-camera and inter-camera computations, respectively. The intra-camera computation directly leverages the CNN features for similarity computation within each camera. Pseudo-labels generated on different cameras train the re-id model in a multi-branch network. The second stage considers the classification scores of each sample on different cameras as a new feature vector. This new feature effectively alleviates the distribution discrepancy among cameras and generates more reliable pseudo-labels. We hence train our re-id model in two stages with intra-camera and inter-camera pseudo-labels, respectively. This simple intra-inter camera similarity produces surprisingly good performance on multiple datasets, e.g., achieves rank-1 accuracy of 89.5% on the Market1501 dataset, outperforming the recent unsupervised works by 9+%, and is comparable with the latest transfer learning works that leverage extra annotations. | ['Shiliang Zhang', 'Shiyu Xuan'] | 2021-03-22 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Xuan_Intra-Inter_Camera_Similarity_for_Unsupervised_Person_Re-Identification_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Xuan_Intra-Inter_Camera_Similarity_for_Unsupervised_Person_Re-Identification_CVPR_2021_paper.pdf | cvpr-2021-1 | ['unsupervised-person-re-identification'] | ['computer-vision'] | [ 3.56621861e-01 -3.76911402e-01 -2.31141672e-02 -7.41019011e-01
-9.48436677e-01 -8.66338968e-01 8.28390777e-01 1.72794014e-01
-7.37060428e-01 5.22037566e-01 2.73734808e-01 3.04574430e-01
2.56749451e-01 -4.01567787e-01 -5.90125740e-01 -6.27277732e-01
5.64288914e-01 6.19057417e-01 1.19209588e-02 4.03293073e-01
8.84952545e-02 9.37970206e-02 -1.63833642e+00 3.49345654e-01
5.03073871e-01 9.69468415e-01 -3.33697498e-01 7.51693010e-01
1.99129686e-01 6.85855567e-01 -7.00872362e-01 -1.01383424e+00
6.30841374e-01 -4.41031277e-01 -6.87191069e-01 3.81493747e-01
1.11346221e+00 -5.01236737e-01 -2.08890259e-01 1.17781818e+00
5.49517632e-01 1.08624034e-01 7.85233021e-01 -1.52923930e+00
-6.66408479e-01 5.24536550e-01 -7.08493948e-01 -1.00023195e-01
4.60307181e-01 4.37622890e-02 1.19950593e+00 -8.30302894e-01
4.16928530e-01 1.05496740e+00 1.00146961e+00 8.34465623e-01
-1.42191291e+00 -9.46767330e-01 1.56883691e-02 2.53554195e-01
-1.62553799e+00 -3.42498034e-01 5.98931015e-01 -6.48622811e-01
6.36950195e-01 1.82239369e-01 4.72521275e-01 1.21815264e+00
-4.27540481e-01 5.52017927e-01 1.23015988e+00 -3.23355198e-01
9.74585116e-02 2.16827393e-01 3.00583303e-01 3.67380232e-01
2.82628626e-01 8.11346844e-02 -6.04903281e-01 -1.35830551e-01
5.42104900e-01 3.60399723e-01 -4.12157476e-02 -2.32303306e-01
-1.37935972e+00 7.35723078e-01 3.27939034e-01 2.07248405e-02
2.42184460e-01 2.60042816e-01 5.70628524e-01 2.57102370e-01
3.44842494e-01 3.64796638e-01 -4.14642751e-01 -1.29912660e-01
-1.05328417e+00 1.57113656e-01 6.10660315e-01 1.16522634e+00
8.61914933e-01 -5.10386169e-01 -4.30002064e-01 1.06960893e+00
1.40958890e-01 4.84495401e-01 5.04997849e-01 -1.20679343e+00
5.46385407e-01 6.04118466e-01 1.32204533e-01 -8.43928397e-01
-4.43964332e-01 -5.11174679e-01 -1.00930977e+00 -1.54834360e-01
7.84753203e-01 -1.67423040e-01 -6.49995089e-01 2.03157449e+00
1.95255488e-01 5.78849614e-01 -2.24896044e-01 8.04132879e-01
7.15271235e-01 1.50146186e-01 1.56510472e-01 2.06247106e-01
1.49542308e+00 -1.43673360e+00 -1.52647763e-01 -1.92213133e-01
8.34426463e-01 -7.13488698e-01 7.27614641e-01 2.72425920e-01
-7.58177519e-01 -9.18577194e-01 -9.22768891e-01 -5.67676127e-02
-3.30097109e-01 6.10439181e-01 2.86188811e-01 7.42970705e-01
-1.36555922e+00 4.45676893e-01 -1.44351900e-01 -5.45848608e-01
4.47943479e-01 5.55756330e-01 -6.80627167e-01 -2.51124650e-01
-8.56574953e-01 5.07717907e-01 2.76492119e-01 -2.24125877e-01
-6.80984676e-01 -7.95843780e-01 -6.24209583e-01 -5.34543730e-02
1.63249686e-01 -7.29160786e-01 1.28284538e+00 -1.29150271e+00
-1.31628919e+00 1.35755265e+00 -1.89500332e-01 -2.31712103e-01
4.90494847e-01 -1.84224721e-03 -4.06874120e-01 2.55723833e-03
5.23935080e-01 1.09457755e+00 9.47559178e-01 -1.43477511e+00
-1.03544128e+00 -3.43845993e-01 8.92839860e-03 2.04225495e-01
-4.44937825e-01 -1.14856362e-01 -6.86607420e-01 -7.82164931e-01
-2.18464047e-01 -1.39897609e+00 8.77075344e-02 -2.55445600e-01
-5.15341043e-01 -4.23748165e-01 1.30223379e-01 -4.40666080e-01
9.05555248e-01 -2.05329609e+00 2.86366232e-02 9.21764374e-02
5.22718787e-01 1.69455469e-01 -3.39129061e-01 7.99403638e-02
-4.79980916e-01 -1.21875204e-01 -1.29543751e-01 -1.10367179e+00
1.27852589e-01 -1.17483713e-01 -1.55934561e-02 4.69248205e-01
5.07642999e-02 8.52306068e-01 -1.01986253e+00 -5.01183629e-01
2.36578077e-01 3.09774101e-01 -4.77991551e-01 3.50687861e-01
4.25213069e-01 5.81684709e-01 1.38462529e-01 6.04202747e-01
7.30572224e-01 -3.66701007e-01 1.05161034e-01 -5.27522445e-01
2.93226033e-01 -1.64316401e-01 -1.12707460e+00 1.69795465e+00
-3.42792153e-01 6.35532022e-01 -3.63687038e-01 -7.94991910e-01
6.49242878e-01 1.59188911e-01 6.99082732e-01 -6.31120205e-01
2.23649412e-01 1.64396182e-01 -4.30360138e-01 8.91702715e-03
4.84425396e-01 -2.62892731e-02 -2.72262156e-01 8.06025743e-01
2.96021581e-01 3.13424617e-01 3.83199424e-01 3.31270725e-01
1.03546810e+00 -1.26851089e-02 -4.71121483e-02 -2.89168227e-02
6.71660900e-01 -2.73964554e-01 5.55631638e-01 9.43470716e-01
-4.69393730e-01 1.15483987e+00 1.14938974e-01 -5.83757997e-01
-1.34847522e+00 -1.04197252e+00 1.11772880e-01 1.27646399e+00
2.45189846e-01 -4.67712492e-01 -1.02508998e+00 -1.01920295e+00
1.65203065e-01 2.55562127e-01 -7.26437509e-01 -2.39119768e-01
-3.02879870e-01 -8.70006025e-01 7.71518350e-01 6.71028912e-01
7.13576853e-01 -6.19352937e-01 -1.23316817e-01 9.38952789e-02
-3.59026283e-01 -1.35411882e+00 -1.09804869e+00 2.90171076e-02
-5.67420959e-01 -1.17363024e+00 -9.73311543e-01 -7.26453006e-01
7.71802247e-01 4.86845136e-01 1.20999622e+00 5.45896776e-02
-2.35710710e-01 7.46788681e-01 -5.35532117e-01 2.18212545e-01
2.18820646e-02 2.69706011e-01 2.51396507e-01 5.44673502e-01
8.87727976e-01 -3.75588387e-01 -6.53287709e-01 4.90606040e-01
-6.19940102e-01 -1.66829135e-02 3.85105222e-01 1.00066936e+00
4.64176685e-01 -1.15961224e-01 3.76982331e-01 -8.85072470e-01
2.95398742e-01 -4.53939527e-01 -2.94396400e-01 4.01552230e-01
-6.68991446e-01 -4.70301844e-02 5.47417343e-01 -6.23388588e-01
-9.11688387e-01 4.20521051e-01 1.03469081e-01 -2.76055753e-01
-2.81783432e-01 -2.74149865e-01 -1.03714034e-01 -1.90559495e-03
5.38828850e-01 1.24187939e-01 -3.08162987e-01 -4.18129325e-01
4.10912842e-01 8.38956952e-01 7.76600063e-01 -4.53278601e-01
1.04950273e+00 5.47002017e-01 -2.07749501e-01 -1.56480193e-01
-1.23162317e+00 -1.00394881e+00 -9.28135514e-01 -1.92518607e-01
1.07937193e+00 -1.27216399e+00 -8.90021622e-01 9.62726116e-01
-1.11258805e+00 -2.71085262e-01 -4.15377855e-01 3.82302672e-01
-2.71283299e-01 4.85313714e-01 -7.02097654e-01 -3.14554155e-01
-3.14995527e-01 -1.15042925e+00 1.42296612e+00 4.32780892e-01
-3.97201151e-01 -8.97152781e-01 1.83211103e-01 6.97565377e-01
1.91531941e-01 -1.21821649e-02 2.61546314e-01 -7.96492279e-01
-3.11673373e-01 -3.46714526e-01 -8.18670869e-01 4.77713585e-01
1.89903334e-01 -3.40183973e-01 -1.35096967e+00 -5.43726981e-01
-3.31905991e-01 -4.87465054e-01 9.16949511e-01 4.24921177e-02
9.79873776e-01 -1.08320825e-01 -2.50595510e-01 7.79794872e-01
1.34533930e+00 -2.51646817e-01 2.78328001e-01 4.21458304e-01
1.14262593e+00 6.27692640e-01 2.09620103e-01 4.59434688e-01
9.43490386e-01 9.89466727e-01 9.13666412e-02 -1.12141557e-01
-4.31113631e-01 -1.92917928e-01 3.88796389e-01 8.14203680e-01
6.37012869e-02 -1.43812522e-01 -7.40871131e-01 4.35207695e-01
-1.90789437e+00 -8.66898000e-01 -1.14919133e-01 2.36087751e+00
7.79823244e-01 -1.22512802e-01 3.61210197e-01 5.45561016e-02
1.08202648e+00 -1.37306958e-01 -5.65716088e-01 2.01768707e-02
-1.69273630e-01 -3.27326171e-02 8.59931290e-01 4.08910245e-01
-1.32536328e+00 7.24030614e-01 5.66672516e+00 8.28255117e-01
-6.18828952e-01 4.54345405e-01 8.81123483e-01 -9.59337056e-02
2.02642515e-01 -8.78377259e-02 -1.24593258e+00 8.55765045e-01
8.72094274e-01 2.55710870e-01 5.20555377e-01 9.18007433e-01
-5.44177592e-01 -1.27532631e-01 -1.56227911e+00 1.70153654e+00
5.97121596e-01 -8.44830096e-01 7.20282868e-02 -1.12980669e-02
1.10388982e+00 4.90714272e-04 3.29432636e-02 2.38270670e-01
4.70811725e-01 -7.94760883e-01 9.43433642e-01 5.24668634e-01
1.13395774e+00 -7.16099620e-01 1.04672754e+00 -8.60430971e-02
-1.44257605e+00 -3.44432056e-01 -3.22613150e-01 5.92155084e-02
1.28376156e-01 6.76104426e-01 -6.64390147e-01 3.55810881e-01
8.58031809e-01 9.73428667e-01 -1.20253623e+00 9.40196812e-01
2.25929637e-02 3.75707000e-01 -1.32903859e-01 4.11041886e-01
-4.49147262e-02 -1.66883707e-01 6.70146346e-02 1.45021296e+00
2.37325355e-01 -4.12693053e-01 2.89349288e-01 5.34050047e-01
-4.50042844e-01 -2.57445931e-01 -1.62654176e-01 4.18648452e-01
5.29170632e-01 1.40356410e+00 -6.69192433e-01 -7.56827891e-01
-6.07538819e-01 1.55053294e+00 4.51332569e-01 3.32881778e-01
-9.12676573e-01 -2.21698135e-01 6.10972703e-01 -9.89489183e-02
3.38960811e-03 5.60304374e-02 -2.93131083e-01 -1.42999125e+00
-4.47183251e-02 -8.47076237e-01 6.66505039e-01 -4.94519770e-01
-1.81702256e+00 4.52732742e-01 6.22482924e-03 -1.29469669e+00
-2.01490551e-01 -5.01093447e-01 -2.69309342e-01 7.07950771e-01
-1.51858020e+00 -1.57949936e+00 -7.71526694e-01 6.67173624e-01
4.25272107e-01 -3.45629245e-01 7.15639174e-01 6.41876936e-01
-7.43462145e-01 1.28645074e+00 2.61023372e-01 5.11442780e-01
1.34828722e+00 -1.54326212e+00 4.62270617e-01 7.93343365e-01
1.87866241e-01 3.98382753e-01 3.46503519e-02 -3.84549886e-01
-7.71097839e-01 -1.22311473e+00 1.06726611e+00 -1.02428877e+00
3.63198757e-01 -6.29391909e-01 -4.42593664e-01 6.91140711e-01
-1.03053473e-01 1.72258094e-01 1.03072691e+00 5.05561754e-02
-9.72746611e-01 -5.59918225e-01 -1.09052598e+00 4.53727216e-01
1.22737467e+00 -8.97011340e-01 -1.64683968e-01 2.61462837e-01
5.41913509e-01 -3.61951180e-02 -8.47553074e-01 1.07221693e-01
7.38729894e-01 -1.04208267e+00 1.22070336e+00 -2.26473168e-01
2.46940956e-01 -4.35773760e-01 -1.43278763e-01 -1.07349968e+00
-5.49174845e-01 -3.95018160e-01 1.62646383e-01 1.73004758e+00
2.35850558e-01 -5.30251801e-01 8.77465725e-01 9.78490472e-01
1.74428374e-01 -2.52369314e-01 -8.75699401e-01 -7.20633388e-01
-1.96067750e-01 -3.50614727e-01 8.35019588e-01 1.13928199e+00
-3.13531905e-01 3.71071100e-01 -5.69275856e-01 -1.67681843e-01
9.83343005e-01 -1.51530102e-01 9.94050860e-01 -1.32477212e+00
-4.07613605e-01 -4.39823598e-01 -2.96053231e-01 -1.10361028e+00
3.07901472e-01 -1.21614909e+00 4.69976775e-02 -9.81196225e-01
9.37713861e-01 -7.02219069e-01 -3.13859314e-01 5.17830789e-01
-3.99869770e-01 9.67648983e-01 3.23814154e-01 5.67761183e-01
-1.06305432e+00 2.03290567e-01 7.50215232e-01 -4.00787950e-01
4.52990420e-02 -1.35579184e-01 -7.69562840e-01 6.62871599e-01
6.69915378e-01 -6.86132193e-01 -1.68940991e-01 -5.50479531e-01
2.39710167e-01 -6.29531264e-01 4.81500685e-01 -1.42140090e+00
4.92063314e-01 3.52638483e-01 7.53686845e-01 -2.66085178e-01
2.78481960e-01 -8.93867552e-01 1.56085059e-01 1.59070835e-01
-5.10819256e-01 1.91289350e-01 -3.45151514e-01 6.11068428e-01
-6.94134310e-02 -2.65969515e-01 8.65918934e-01 -1.37175217e-01
-5.21608472e-01 2.88148046e-01 -4.63473313e-02 1.44553080e-01
8.87376070e-01 -3.85633618e-01 -2.88219899e-01 -3.45368981e-01
-5.42634904e-01 1.08119257e-01 7.07934380e-01 3.50082546e-01
2.01846063e-01 -1.63779700e+00 -7.69928217e-01 1.58730641e-01
3.84815842e-01 -1.29944459e-01 2.55670100e-01 7.08116174e-01
-2.62024045e-01 2.56471157e-01 -1.43360302e-01 -8.19727421e-01
-1.37216270e+00 4.56463456e-01 3.01008761e-01 -5.32050192e-01
-1.42274335e-01 1.15489638e+00 2.16554895e-01 -6.28752887e-01
2.98009157e-01 9.12297592e-02 -2.35658929e-01 4.57185566e-01
6.69276297e-01 6.64428055e-01 -1.95396274e-01 -1.07907403e+00
-2.95010477e-01 1.16185057e+00 -2.65642017e-01 -9.37732160e-02
9.28731918e-01 -3.65537316e-01 -6.80438355e-02 2.80051410e-01
1.61152065e+00 -2.09537625e-01 -1.47195530e+00 -3.72471452e-01
8.49189311e-02 -4.65694427e-01 -3.97946775e-01 -6.73490405e-01
-1.15302205e+00 7.32118845e-01 9.07342255e-01 -2.25431249e-01
1.04599559e+00 -6.14130683e-02 1.06807923e+00 3.15942109e-01
4.49330777e-01 -1.39246047e+00 5.48378408e-01 3.85311097e-01
2.60483980e-01 -1.67057049e+00 7.17490986e-02 -2.33146369e-01
-7.62657583e-01 9.08329070e-01 6.67951047e-01 -2.41095070e-02
5.02749145e-01 -7.21223652e-02 4.38079834e-02 1.68631271e-01
-1.37715086e-01 -2.05028370e-01 3.78584385e-01 6.98490202e-01
2.19499648e-01 2.25648299e-01 1.20050453e-01 6.73191845e-01
-3.64515409e-02 -4.20156792e-02 2.65325695e-01 5.80874860e-01
6.57586707e-03 -1.45065403e+00 -3.77679795e-01 5.11717558e-01
-2.40985364e-01 -1.56635135e-01 -5.29838204e-01 2.55913913e-01
7.04810560e-01 1.16976285e+00 2.20931962e-01 -8.12126935e-01
1.38162807e-01 3.20390500e-02 3.49630713e-01 -6.58452034e-01
-9.03874040e-01 -3.65019917e-01 -5.95188141e-02 -5.87255657e-01
-7.58515000e-01 -7.68045604e-01 -7.34661698e-01 -4.29978907e-01
-4.00476396e-01 -6.10047840e-02 4.51259792e-01 8.19185495e-01
4.49839443e-01 1.55482396e-01 8.60386431e-01 -1.08925188e+00
-4.49514806e-01 -1.03490996e+00 -4.68699127e-01 1.11501980e+00
1.98867112e-01 -6.38566971e-01 -5.68666101e-01 4.47139323e-01] | [14.77997875213623, 1.0400997400283813] |
c5947037-397d-4a3a-90b2-8f49bd801c9f | unsupervised-polychromatic-neural | 2306.15203 | null | https://arxiv.org/abs/2306.15203v1 | https://arxiv.org/pdf/2306.15203v1.pdf | Unsupervised Polychromatic Neural Representation for CT Metal Artifact Reduction | Emerging neural reconstruction techniques based on tomography (e.g., NeRF, NeAT, and NeRP) have started showing unique capabilities in medical imaging. In this work, we present a novel Polychromatic neural representation (Polyner) to tackle the challenging problem of CT imaging when metallic implants exist within the human body. The artifacts arise from the drastic variation of metal's attenuation coefficients at various energy levels of the X-ray spectrum, leading to a nonlinear metal effect in CT measurements. Reconstructing CT images from metal-affected measurements hence poses a complicated nonlinear inverse problem where empirical models adopted in previous metal artifact reduction (MAR) approaches lead to signal loss and strongly aliased reconstructions. Polyner instead models the MAR problem from a nonlinear inverse problem perspective. Specifically, we first derive a polychromatic forward model to accurately simulate the nonlinear CT acquisition process. Then, we incorporate our forward model into the implicit neural representation to accomplish reconstruction. Lastly, we adopt a regularizer to preserve the physical properties of the CT images across different energy levels while effectively constraining the solution space. Our Polyner is an unsupervised method and does not require any external training data. Experimenting with multiple datasets shows that our Polyner achieves comparable or better performance than supervised methods on in-domain datasets while demonstrating significant performance improvements on out-of-domain datasets. To the best of our knowledge, our Polyner is the first unsupervised MAR method that outperforms its supervised counterparts. | ['Yuyao Zhang', 'Jingyi Yu', 'S. Kevin Zhou', 'Hongjiang Wei', 'Ce Wang', 'Lixuan Chen', 'Qing Wu'] | 2023-06-27 | null | null | null | null | ['metal-artifact-reduction'] | ['medical'] | [ 5.62432170e-01 -3.73699963e-02 1.44912034e-01 -2.86539406e-01
-7.11410403e-01 1.35429323e-01 2.31289953e-01 -3.80359083e-01
-3.56753111e-01 5.42387426e-01 3.44085842e-01 -8.11939240e-02
-4.72288668e-01 -7.28074312e-01 -9.33713019e-01 -8.83527040e-01
1.58017948e-01 4.48003531e-01 1.43891111e-01 -3.46339829e-02
-2.03133270e-01 5.25731385e-01 -9.34663117e-01 3.19366902e-01
7.68606603e-01 8.71133029e-01 3.09628874e-01 1.40073135e-01
3.91228199e-01 9.62402105e-01 -1.65996447e-01 -1.17089391e-01
3.54980886e-01 -4.51741368e-01 -6.23710454e-01 3.81770618e-02
8.40170756e-02 -3.91298741e-01 -7.97928810e-01 9.90286648e-01
6.75725520e-01 2.61237353e-01 8.89928341e-01 -5.80842376e-01
-8.84046078e-01 6.28184617e-01 -5.63688874e-01 4.31989096e-02
-2.09906232e-02 -1.15014082e-02 3.69818449e-01 -9.92245078e-01
6.05508447e-01 6.49656713e-01 1.00828612e+00 6.38280451e-01
-1.34485209e+00 -3.83079559e-01 -4.22510028e-01 1.54958144e-01
-1.16923165e+00 -2.69360065e-01 1.06934965e+00 -3.73419911e-01
5.64686477e-01 3.79873723e-01 4.37465787e-01 1.12887001e+00
4.29807067e-01 4.96478468e-01 1.20707655e+00 -4.18519080e-01
2.28053004e-01 -8.34486410e-02 -7.91033823e-03 7.00162470e-01
2.21208498e-01 2.44936928e-01 -3.29750687e-01 -1.71130344e-01
1.10045362e+00 1.50338843e-01 -7.63303995e-01 -2.79449344e-01
-1.03204787e+00 5.97010195e-01 6.18971825e-01 5.73342621e-01
-6.46095455e-01 1.94897234e-01 1.78297207e-01 4.30775806e-02
2.80374616e-01 4.96501446e-01 -1.07440792e-01 4.00176793e-01
-8.74810278e-01 -6.41300678e-02 5.41197836e-01 6.23617709e-01
2.31867254e-01 3.40142608e-01 -2.59506181e-02 1.09264827e+00
4.44713861e-01 4.69967455e-01 8.34703028e-01 -8.93243849e-01
1.40017346e-01 1.97082996e-01 -2.42141932e-01 -9.66283381e-01
-5.50150514e-01 -7.23457515e-01 -1.17398059e+00 1.52716592e-01
3.78429204e-01 9.34215437e-04 -1.09525120e+00 1.69763899e+00
3.86360317e-01 2.91828036e-01 -5.79057783e-02 1.16704607e+00
9.44685400e-01 6.31403148e-01 -1.46333531e-01 -3.70168179e-01
1.21123564e+00 -6.39769435e-01 -9.28698599e-01 -1.91957533e-01
3.26384962e-01 -6.42302930e-01 8.33616674e-01 6.96166754e-01
-1.28767347e+00 -2.74629772e-01 -1.18856478e+00 4.34785821e-02
2.80379742e-01 3.27535085e-02 3.49601060e-01 4.78210032e-01
-8.90790343e-01 8.52124274e-01 -1.09864450e+00 3.25869381e-01
3.38469625e-01 4.76873726e-01 -2.52613187e-01 -4.81636196e-01
-1.07718825e+00 9.29347754e-01 -3.00017707e-02 4.06764567e-01
-7.67773032e-01 -9.86173868e-01 -8.19661796e-01 -3.93992752e-01
4.96187389e-01 -7.40833282e-01 1.16030025e+00 -9.15313303e-01
-1.77326322e+00 4.44906116e-01 2.52295405e-01 -2.60057122e-01
5.32963693e-01 -1.67522162e-01 -3.97334278e-01 3.68077159e-01
-1.15786202e-01 7.21942335e-02 8.89889002e-01 -1.67239702e+00
5.34160495e-01 -2.61732221e-01 -4.87524152e-01 1.57944873e-01
-1.28001049e-01 -1.28324345e-01 -5.14406025e-01 -9.72367167e-01
6.93787038e-01 -1.03264844e+00 -4.84203458e-01 1.17776163e-01
-4.56010431e-01 5.40907979e-01 3.68734509e-01 -9.18091059e-01
9.04405236e-01 -2.02745819e+00 2.20274419e-01 3.59676301e-01
3.85685325e-01 -2.90977396e-02 9.85715017e-02 2.03103855e-01
-3.53474170e-01 -4.37216938e-01 -9.15954828e-01 -2.54618347e-01
-2.44772196e-01 3.57387900e-01 1.20656624e-01 8.05222213e-01
-2.26761907e-01 9.07635510e-01 -5.88915706e-01 -3.63592297e-01
1.36275545e-01 8.85865808e-01 -6.13227427e-01 2.40358841e-02
1.57669365e-01 8.30147386e-01 -4.51959312e-01 4.42272872e-01
1.02546430e+00 -3.61541271e-01 2.33507186e-01 -6.44859970e-01
4.18999232e-02 2.88230907e-02 -7.84639776e-01 1.83838034e+00
-6.47862971e-01 2.23569795e-01 1.75525859e-01 -1.17910814e+00
5.15718877e-01 5.03598392e-01 1.00802672e+00 -1.13561094e+00
3.85130674e-01 4.70023364e-01 2.18831673e-01 -6.30830407e-01
1.69721156e-01 -7.76375055e-01 2.87562340e-01 4.71803427e-01
-5.84719293e-02 -1.50223047e-01 -5.24039805e-01 -8.00292268e-02
1.11807764e+00 1.74127311e-01 5.74990269e-03 -2.53753841e-01
2.65633464e-01 -1.83861002e-01 4.38661695e-01 6.35352373e-01
2.16531102e-02 1.12459862e+00 -7.44375959e-02 -3.43851417e-01
-1.00620949e+00 -1.22989273e+00 -5.28434217e-01 3.23735744e-01
2.96518896e-02 1.20344318e-01 -8.18557501e-01 -3.61220568e-01
-3.44692856e-01 5.59522688e-01 -6.11500859e-01 -3.42450589e-01
-1.16829646e+00 -1.14108229e+00 5.43895662e-01 5.09475827e-01
3.98228586e-01 -1.10528874e+00 -6.00368202e-01 3.57914299e-01
-3.96683782e-01 -1.28748190e+00 -3.39346081e-01 5.25764704e-01
-1.16379309e+00 -7.89990664e-01 -9.85909581e-01 -5.62822819e-01
8.37352157e-01 -1.42247260e-01 9.93327916e-01 -5.17405421e-02
-4.94901896e-01 4.70767140e-01 -1.93676069e-01 -1.94214117e-02
-4.80418682e-01 -5.58151305e-01 4.98772189e-02 4.91670379e-03
-3.29430610e-01 -8.11125219e-01 -8.48831832e-01 3.07017297e-01
-1.26108837e+00 4.37573008e-02 6.46840930e-01 1.03497100e+00
8.96247029e-01 2.88142830e-01 2.53677517e-01 -9.85055804e-01
3.36374819e-01 -6.21663213e-01 -1.95878923e-01 -7.73908719e-02
-4.21485305e-01 3.17449123e-01 7.46814311e-01 -5.79762518e-01
-1.41778779e+00 3.21493186e-02 -5.32947898e-01 -4.81798559e-01
-5.31686023e-02 6.69954896e-01 -2.78474335e-02 -2.76585251e-01
8.12596679e-01 4.94283378e-01 7.52978995e-02 -4.54692900e-01
-1.13625772e-01 1.77913472e-01 6.90670609e-01 -6.91716135e-01
7.48808980e-01 7.35964835e-01 2.61988223e-01 -7.95836627e-01
-6.78341925e-01 -1.17753670e-01 -3.60357165e-01 -3.18488449e-01
7.28140593e-01 -5.58620691e-01 -5.04666269e-01 5.97006798e-01
-8.01395595e-01 -5.50022602e-01 -5.46803951e-01 9.14139211e-01
-5.08439660e-01 7.34993935e-01 -9.55531299e-01 -5.09378493e-01
-4.08463418e-01 -1.35464430e+00 6.90357387e-01 -1.57509223e-01
-1.25311986e-02 -9.39792156e-01 1.55885473e-01 4.39258933e-01
5.31045794e-01 4.41025406e-01 8.61203611e-01 -5.59524819e-02
-4.37968224e-01 9.02257264e-02 -2.75065633e-03 5.29829144e-01
9.01907682e-02 -7.00149536e-01 -8.04445446e-01 -1.41893893e-01
9.61813927e-01 -1.11310929e-01 7.14319229e-01 9.17174757e-01
1.46988833e+00 -2.52519161e-01 1.74338243e-03 9.10585582e-01
1.57493198e+00 2.96934038e-01 8.11096191e-01 2.50279874e-01
7.66824484e-01 4.25078869e-01 1.23428620e-01 3.41950834e-01
6.27147034e-02 7.17853606e-01 4.83083814e-01 -2.04940587e-01
-4.41907376e-01 9.51599032e-02 1.26221254e-01 1.09397078e+00
-3.05576175e-01 -7.34321401e-02 -8.95831585e-01 2.53479809e-01
-1.59962344e+00 -5.57800531e-01 -4.99143213e-01 2.15182662e+00
8.70203674e-01 -2.48217329e-01 -4.26127851e-01 2.02289879e-01
3.32010418e-01 -2.00432137e-01 -5.66178441e-01 2.77398583e-02
1.13752121e-02 6.93486392e-01 5.91058910e-01 2.56717831e-01
-7.32357502e-01 1.96538955e-01 6.21711493e+00 6.56925857e-01
-1.18254721e+00 5.58445215e-01 4.65366840e-01 -4.12781090e-02
-4.33219582e-01 -4.48854953e-01 7.41188526e-02 3.91503781e-01
7.22938240e-01 4.25990224e-01 4.43975359e-01 3.86333585e-01
1.49526551e-01 -9.46993977e-02 -9.29934442e-01 9.81816769e-01
1.22704908e-01 -9.34098363e-01 -1.40587538e-01 1.32769987e-01
7.07488894e-01 9.93948504e-02 3.13740671e-01 -8.04629922e-02
7.06012920e-02 -1.10559309e+00 6.81765735e-01 6.64015591e-01
8.89920771e-01 -5.58514118e-01 5.83408475e-01 2.70733654e-01
-6.66055679e-01 1.27900451e-01 -2.26033270e-01 3.74829233e-01
4.77396876e-01 9.96377409e-01 -5.74676931e-01 8.77165914e-01
6.85499191e-01 5.74259639e-01 -1.05344720e-01 9.67007577e-01
-5.50587401e-02 6.69942796e-01 -3.31446022e-01 5.96989751e-01
1.07386932e-01 -3.36668968e-01 5.65457463e-01 9.53478336e-01
4.97699678e-01 5.05630493e-01 -2.02424526e-01 1.04820228e+00
-1.63261414e-01 4.03484330e-02 -4.15950835e-01 4.27961826e-01
-2.80966550e-01 1.02872682e+00 -7.10205853e-01 -7.86920413e-02
-3.19499820e-01 1.02525890e+00 -9.04938299e-03 5.07219195e-01
-1.05418622e+00 2.32641459e-01 -5.43671325e-02 3.27405006e-01
-4.72707599e-02 -1.89366058e-01 -4.58164066e-01 -1.00728822e+00
2.72885978e-01 -7.39741981e-01 2.52594620e-01 -8.51437569e-01
-1.35321486e+00 7.44107544e-01 1.02500945e-01 -1.25880241e+00
1.33085430e-01 -6.40070856e-01 -3.42489481e-01 5.85886598e-01
-1.56596482e+00 -8.65091920e-01 -3.11060071e-01 8.17067087e-01
2.56609738e-01 2.21031323e-01 5.76873302e-01 6.99214160e-01
-3.03465843e-01 3.67306978e-01 2.46306226e-01 6.98523084e-03
5.61923563e-01 -1.03303254e+00 -1.18029296e-01 7.04510272e-01
-1.81168228e-01 6.77645981e-01 6.21013641e-01 -7.42819488e-01
-1.67004657e+00 -9.45799291e-01 2.01562569e-02 -7.66499788e-02
4.82365966e-01 -5.04382420e-03 -1.12174881e+00 6.29702210e-01
7.80475140e-02 4.29691255e-01 7.43423164e-01 -4.79165167e-01
-1.80712521e-01 7.10767210e-02 -1.30934453e+00 3.22577298e-01
9.75201964e-01 -3.42163622e-01 -4.71144319e-01 4.14704472e-01
4.46250588e-01 -7.99274921e-01 -1.12867653e+00 6.18806064e-01
5.75596750e-01 -9.17070687e-01 1.34939325e+00 -2.69817971e-02
8.10407579e-01 -8.85258019e-02 -2.42719918e-01 -1.28623784e+00
-4.88453865e-01 -1.84181631e-01 -9.43560451e-02 6.13905609e-01
2.68409163e-01 -7.63510585e-01 7.38005280e-01 5.98379910e-01
-6.62090957e-01 -8.64481390e-01 -1.08676529e+00 -6.09492660e-01
2.77738869e-01 -6.49309754e-01 9.11633447e-02 1.14199972e+00
-3.34847599e-01 -3.41523588e-01 -5.79747021e-01 4.09800440e-01
9.58049417e-01 -2.48885632e-01 4.61985581e-02 -9.74811614e-01
-8.93304288e-01 -7.91489263e-04 -2.82920878e-02 -9.86179113e-01
-6.28925487e-02 -1.05217552e+00 4.13900763e-01 -1.46365499e+00
3.31554174e-01 -7.13164210e-01 -4.29859936e-01 2.93397367e-01
1.58025891e-01 5.63196838e-01 -1.07080288e-01 3.49483967e-01
6.88255802e-02 6.14679635e-01 1.75089705e+00 -1.95067495e-01
-1.09468564e-01 -9.55943093e-02 -4.76335227e-01 8.79952848e-01
6.67109728e-01 -8.21322203e-01 -3.25382650e-01 -7.02958643e-01
2.21663177e-01 3.27709526e-01 5.79770446e-01 -1.11048746e+00
2.74856895e-01 1.60181910e-01 6.12609625e-01 -3.13979685e-01
5.49691558e-01 -1.19137049e+00 7.02324331e-01 6.49706483e-01
-8.10843781e-02 -2.50957429e-01 1.43222794e-01 4.24183697e-01
-1.89919800e-01 -4.74837691e-01 1.08878374e+00 -3.94983947e-01
-5.41400500e-02 3.07739407e-01 -4.40508813e-01 -9.12827998e-02
5.20300448e-01 -1.79832622e-01 1.18638232e-01 -1.44401625e-01
-9.24103439e-01 -4.44557816e-01 2.05703259e-01 -9.32423696e-02
9.30104911e-01 -1.30145669e+00 -6.40961766e-01 1.69799417e-01
-2.83090502e-01 2.02028245e-01 7.12085843e-01 1.36946535e+00
-6.71446621e-01 -4.81742248e-02 -1.02189012e-01 -6.84759915e-01
-8.36307168e-01 4.03338909e-01 7.15584338e-01 -5.73080719e-01
-1.21990252e+00 6.93675995e-01 4.46367383e-01 -6.39830947e-01
-1.78167596e-01 -3.78720403e-01 7.10110068e-02 -4.75501031e-01
2.32298166e-01 3.15911442e-01 4.32452947e-01 -7.46128738e-01
-1.98332176e-01 8.31331730e-01 4.01331820e-02 -1.81860089e-01
1.78524482e+00 1.34059250e-01 -5.32241575e-02 2.17012063e-01
1.26830959e+00 -2.68387701e-02 -1.04529202e+00 -4.68763798e-01
-4.09196764e-01 -1.90005198e-01 4.52603489e-01 -9.26090896e-01
-1.49236798e+00 6.23557568e-01 7.30222821e-01 -2.48132050e-01
1.39460397e+00 -6.96416944e-02 1.07009006e+00 1.76381394e-01
3.04964811e-01 -9.00064766e-01 2.82225817e-01 1.57155752e-01
1.02307236e+00 -9.46933091e-01 3.20158035e-01 -5.51830769e-01
-5.33569574e-01 1.07713592e+00 2.97830343e-01 -2.55538613e-01
7.32343495e-01 6.23418927e-01 -5.87838851e-02 -5.36166549e-01
-5.66616915e-02 3.57652694e-01 4.03728932e-01 4.65316504e-01
3.38561982e-01 -1.03835106e-01 -2.39841148e-01 5.67485154e-01
-1.54009804e-01 1.15925543e-01 5.83556771e-01 1.02193451e+00
4.17283587e-02 -9.60234225e-01 -7.10428417e-01 3.15259814e-01
-6.72618508e-01 -1.25752300e-01 -2.99375830e-03 6.77837610e-01
-7.78721785e-03 4.73645359e-01 -3.55000377e-01 -4.49102074e-02
4.63415056e-01 -1.50772721e-01 8.43442798e-01 -4.39999044e-01
-6.60338998e-01 4.44958299e-01 -2.73539662e-01 -5.94567597e-01
-5.16844273e-01 -5.84638834e-01 -1.56602657e+00 1.86346188e-01
-3.40535790e-01 -1.54048577e-01 7.89205909e-01 8.52418542e-01
-6.20628856e-02 1.09877849e+00 4.21923637e-01 -9.81373191e-01
-5.21748006e-01 -7.41994798e-01 -6.60459459e-01 4.79565442e-01
3.10211569e-01 -8.23026061e-01 -2.35610783e-01 -7.58450851e-02] | [13.461910247802734, -2.515967845916748] |
40c31516-9547-4fa6-a67a-ffa2b60e5e9b | sparse-3d-convolutional-neural-networks | 1505.02890 | null | http://arxiv.org/abs/1505.02890v2 | http://arxiv.org/pdf/1505.02890v2.pdf | Sparse 3D convolutional neural networks | We have implemented a convolutional neural network designed for processing
sparse three-dimensional input data. The world we live in is three dimensional
so there are a large number of potential applications including 3D object
recognition and analysis of space-time objects. In the quest for efficiency, we
experiment with CNNs on the 2D triangular-lattice and 3D tetrahedral-lattice. | ['Ben Graham'] | 2015-05-12 | null | null | null | null | ['3d-object-recognition'] | ['computer-vision'] | [-2.96243399e-01 -2.74071991e-01 2.09109426e-01 -1.37423471e-01
5.09955105e-04 -4.26912270e-02 5.88299274e-01 -3.76288816e-02
-5.72653532e-01 3.19609880e-01 1.62035987e-01 -4.59726065e-01
-2.52268046e-01 -1.08147013e+00 -4.61697847e-01 -4.37189728e-01
-9.70627546e-01 7.90464461e-01 2.21338406e-01 -1.06281608e-01
2.97518760e-01 1.37466192e+00 -1.41154015e+00 5.74645400e-01
-1.74235180e-01 1.46963727e+00 -4.59743649e-01 1.04250157e+00
-1.24209642e-01 3.57481450e-01 -3.80921125e-01 3.25266033e-01
5.87209165e-01 1.71654761e-01 -7.44561791e-01 1.23490922e-01
2.11071491e-01 6.17813598e-03 -5.77958167e-01 4.70649809e-01
5.50342083e-01 2.67345220e-01 8.38232398e-01 -8.81834805e-01
-6.64372385e-01 -3.49183768e-01 -3.83110613e-01 4.66117054e-01
2.67919689e-01 6.72946200e-02 5.49620152e-01 -1.38498092e+00
5.80821455e-01 1.44753373e+00 7.79688001e-01 2.71464258e-01
-8.51867497e-01 -1.89900815e-01 -4.55280155e-01 -5.62148727e-02
-1.16050911e+00 -1.79374665e-01 6.37335241e-01 -5.69768548e-01
1.86160934e+00 8.28685425e-03 8.40397537e-01 6.09174371e-01
7.70805776e-01 4.64984894e-01 7.19523251e-01 -4.42290246e-01
3.79337579e-01 -8.07438016e-01 -3.67779424e-03 6.44986808e-01
1.25998661e-01 3.07233363e-01 -4.61663663e-01 -2.25257874e-01
1.49686253e+00 1.49997175e-01 4.75098580e-01 -5.45720458e-01
-1.53362906e+00 7.78592110e-01 8.09206903e-01 5.01114070e-01
-4.01613921e-01 2.73128778e-01 5.69676638e-01 3.17101538e-01
4.85452384e-01 6.09265804e-01 -3.59859854e-01 -1.71851948e-01
-7.46003985e-01 4.64120090e-01 9.81085002e-01 8.02432656e-01
4.35121149e-01 3.70184392e-01 3.60306948e-01 4.30192977e-01
2.17152126e-02 2.19155803e-01 3.02289784e-01 -9.10145581e-01
2.11492762e-01 6.46707475e-01 -8.83898437e-02 -9.27201331e-01
-7.88990080e-01 -2.22318828e-01 -1.68656254e+00 5.43116748e-01
1.29322022e-01 1.30974710e-01 -1.12097406e+00 7.26513386e-01
1.78145185e-01 8.11131522e-02 8.38312656e-02 1.10576177e+00
9.64478135e-01 7.30665326e-01 -3.68622035e-01 1.86529845e-01
1.03137445e+00 -6.32185340e-01 -1.87877148e-01 5.19170426e-02
6.03681445e-01 -5.32115221e-01 5.50576389e-01 4.73432422e-01
-1.18205833e+00 -4.70663548e-01 -1.04569829e+00 -2.28288114e-01
-5.44236779e-01 -2.05621421e-01 1.12422490e+00 -7.07108527e-02
-8.99163604e-01 6.23779416e-01 -1.12056184e+00 -2.30007574e-01
6.17533684e-01 7.06955612e-01 -7.46807337e-01 -4.41793613e-02
-8.53264332e-01 7.79838443e-01 2.77515858e-01 3.57287616e-01
-5.31051636e-01 -2.46801853e-01 -7.75749683e-01 7.99592808e-02
-2.46339232e-01 -5.78631282e-01 1.11209905e+00 -2.82140404e-01
-1.04838729e+00 9.37127292e-01 -8.13977569e-02 -4.11215216e-01
1.69996902e-01 2.77238458e-01 -4.21921998e-01 -3.14644456e-01
-5.41626215e-02 4.23305303e-01 6.40608549e-01 -6.58062458e-01
-2.84147531e-01 -7.14943767e-01 -1.49867550e-01 2.46226251e-01
9.17891562e-02 -3.19721550e-02 -2.18466390e-02 -4.87057507e-01
7.74161875e-01 -8.03053260e-01 -6.64654136e-01 2.73173526e-02
-2.20353246e-01 -3.25181633e-01 1.10160506e+00 3.79992500e-02
6.01120532e-01 -2.13870144e+00 2.78534472e-01 3.31849098e-01
4.37353969e-01 2.81897932e-01 -1.04376599e-01 4.11573648e-01
-2.46609449e-01 1.02942996e-01 1.00236923e-01 -1.05669990e-01
-1.16936758e-01 1.85601279e-01 -8.77210349e-02 6.72612786e-01
5.38937986e-01 1.10882080e+00 -6.52715445e-01 -5.53380698e-02
1.47739887e-01 6.00647926e-01 -6.63372397e-01 4.74430211e-02
-3.20088893e-01 4.20559585e-01 -6.47347391e-01 7.51429737e-01
4.78477687e-01 -6.08098567e-01 -4.86944586e-01 -5.13441339e-02
-3.42673331e-01 4.84774321e-01 -1.17320621e+00 1.67832279e+00
-3.28366965e-01 7.81855643e-01 -8.85496382e-03 -1.13468552e+00
9.39212322e-01 2.52083302e-01 8.28277469e-01 -1.03260612e+00
1.95127040e-01 2.26597592e-01 3.61415833e-01 -2.59709954e-01
6.08371556e-01 -2.33002931e-01 -2.42335662e-01 6.72304094e-01
-2.20498089e-02 -4.15207475e-01 3.79637927e-02 -1.94450408e-01
1.41569781e+00 -3.54573727e-01 2.48227999e-01 -4.02241826e-01
8.97425115e-02 1.26218215e-01 1.12069279e-01 3.85731250e-01
-1.06513768e-01 6.36415005e-01 3.79448086e-01 -1.57860255e+00
-1.48806047e+00 -1.07723272e+00 -3.12085897e-01 6.85916722e-01
-2.76385486e-01 -5.20402975e-02 -4.04096209e-02 9.30099264e-02
2.48225123e-01 -2.34382123e-01 -6.23575509e-01 2.11907238e-01
-1.02029502e+00 -3.52400720e-01 2.33838633e-01 5.74140608e-01
4.31805223e-01 -1.37632549e+00 -8.77414346e-01 3.95904362e-01
6.88613772e-01 -9.24307048e-01 -1.27866939e-01 6.26162589e-01
-1.06361532e+00 -1.02922344e+00 -4.60980773e-01 -9.89705205e-01
4.62635994e-01 2.65427709e-01 1.24526417e+00 -2.03847095e-01
-6.27489269e-01 3.94394130e-01 7.68020228e-02 -5.84385812e-01
9.60628390e-02 1.73516572e-01 5.76357067e-01 -5.40551126e-01
6.09208703e-01 -7.37365067e-01 -3.59821051e-01 6.65288493e-02
-7.63392270e-01 6.59984946e-02 3.29199940e-01 7.21263230e-01
8.42771947e-01 2.46906847e-01 3.29841316e-01 -4.20913249e-01
5.84794223e-01 -4.73722637e-01 -6.60924673e-01 -2.41941601e-01
2.02685907e-01 -1.51714861e-01 5.32016933e-01 -4.11622018e-01
-3.99288714e-01 1.81673914e-01 -3.08640599e-01 -5.63086987e-01
-4.01090264e-01 5.44728100e-01 3.43530893e-01 -2.63905019e-01
6.25302851e-01 -3.34289707e-02 -7.80040622e-02 -3.36783051e-01
1.56186298e-02 4.80940312e-01 2.99103051e-01 -6.34216011e-01
4.67663765e-01 6.27395213e-01 6.00023389e-01 -1.28913784e+00
-5.87097764e-01 -1.99551031e-01 -1.07139409e+00 -1.03895605e-01
8.52578521e-01 -7.09010720e-01 -1.13536131e+00 5.12841284e-01
-1.59474695e+00 -4.03047234e-01 -4.26122367e-01 4.99364376e-01
-8.01745415e-01 -2.23903805e-01 -7.21225619e-01 -5.13379216e-01
-3.43742311e-01 -9.63853657e-01 1.19330800e+00 -1.14717193e-01
-2.55520582e-01 -1.02166867e+00 1.97545692e-01 -4.82702643e-01
6.21271789e-01 2.43173778e-01 1.04128563e+00 -5.05845904e-01
-9.07203138e-01 -4.28102225e-01 -3.79863024e-01 8.57088566e-02
1.85488276e-02 -2.14581549e-01 -6.13525331e-01 -3.24557483e-01
3.31313275e-02 -4.45920110e-01 6.62218213e-01 6.77229464e-01
1.62468100e+00 -1.48083463e-01 -7.24694505e-02 7.08333254e-01
1.30483580e+00 1.29207626e-01 4.89067495e-01 1.69894218e-01
6.01901054e-01 2.01287970e-01 -1.33869676e-02 6.06955051e-01
3.38982157e-02 3.02097619e-01 5.36179006e-01 -2.32915953e-01
7.50682577e-02 2.23044306e-01 -3.96183938e-01 9.99601841e-01
-1.53695270e-01 -1.76703662e-01 -1.55541611e+00 6.21177852e-01
-1.56401408e+00 -7.94902384e-01 -4.58051227e-02 1.91696775e+00
2.60784924e-01 4.70921934e-01 -4.42659445e-02 1.53084517e-01
2.17290476e-01 1.73404261e-01 -7.89090157e-01 -6.09059632e-01
-1.50327384e-01 5.87054551e-01 3.58161151e-01 1.36266500e-01
-1.19633079e+00 5.59163332e-01 7.98848152e+00 2.33143136e-01
-1.51850390e+00 -4.95358825e-01 7.02304065e-01 -2.27605999e-01
8.36706832e-02 -7.04376578e-01 -7.37444699e-01 -2.16506496e-02
9.05455530e-01 -1.82417423e-01 2.40385458e-01 5.48095644e-01
-2.04275902e-02 3.81421074e-02 -1.39510024e+00 1.18448329e+00
-3.12733829e-01 -2.05997038e+00 7.60312304e-02 3.75417411e-01
8.85965586e-01 5.80675781e-01 2.09735230e-01 7.99222663e-02
2.11059406e-01 -1.65179920e+00 4.25428659e-01 4.31355119e-01
1.00768209e+00 -6.90218985e-01 4.89036411e-01 3.61786127e-01
-1.29596770e+00 1.68882489e-01 -4.24492151e-01 -6.73680484e-01
-2.31816564e-02 7.70271838e-01 -7.19052136e-01 -4.99364361e-03
8.64804506e-01 8.89331639e-01 8.93116370e-02 1.17596054e+00
6.39128923e-01 8.69367346e-02 -6.70142353e-01 -1.74570113e-01
6.17003798e-01 -8.56481194e-02 4.32298392e-01 9.89124358e-01
3.34078193e-01 6.11423254e-01 1.75764024e-01 6.76008821e-01
-1.50314629e-01 -2.84252167e-01 -1.20957565e+00 -3.53500485e-01
1.28767490e-01 7.10792661e-01 -1.06033289e+00 6.22413643e-02
-6.01582468e-01 5.83044827e-01 2.64520615e-01 2.39755780e-01
-9.50712711e-02 -4.19702798e-01 1.01984942e+00 1.84508398e-01
4.93563652e-01 -1.24565506e+00 -5.49436688e-01 -8.84601831e-01
-5.72807230e-02 -4.89691883e-01 7.34453425e-02 -7.23774195e-01
-1.55516636e+00 6.35216296e-01 -4.13546175e-01 -1.12079406e+00
-1.53854504e-01 -1.11192381e+00 -6.46863282e-01 7.79888928e-01
-9.81165349e-01 -6.33314013e-01 2.65761279e-02 5.13132513e-01
3.07623953e-01 -2.63939261e-01 1.05719995e+00 4.26672213e-02
3.85337770e-02 -1.78277835e-01 1.89281359e-01 4.32723910e-01
-3.15435976e-02 -8.53776217e-01 1.18840730e+00 1.24485120e-01
1.87645465e-01 3.38066220e-01 2.57755160e-01 -4.64873970e-01
-1.91098881e+00 -1.07860208e+00 7.29264915e-01 -3.02979082e-01
5.72303474e-01 -5.59032917e-01 -1.02780437e+00 6.37924254e-01
-2.38876417e-01 4.40068096e-01 4.52477425e-01 4.76296172e-02
-3.68254930e-01 2.46656641e-01 -1.17346823e+00 3.73641193e-01
1.07770443e+00 -5.79616606e-01 -3.29368114e-01 5.44454038e-01
7.31599331e-01 -8.24790716e-01 -1.23617101e+00 5.01198232e-01
5.49228907e-01 -8.78929436e-01 1.19020498e+00 -1.01406193e+00
4.65575099e-01 -1.55059127e-02 -3.54930878e-01 -1.12999737e+00
-5.90973020e-01 -1.49770916e-01 -3.06177408e-01 -1.22697957e-01
-2.44576335e-02 -4.22617912e-01 1.25835669e+00 3.96037132e-01
-2.62171119e-01 -9.94762778e-01 -1.43478215e+00 -7.68378496e-01
4.11925524e-01 -5.86282313e-01 5.34026802e-01 6.62407815e-01
-1.64577037e-01 3.88999462e-01 1.48611933e-01 4.28536646e-02
5.07164419e-01 3.07272643e-01 4.92175609e-01 -1.46768987e+00
2.79154450e-01 -5.89262784e-01 -9.63361204e-01 -1.19455278e+00
6.63371161e-02 -8.49556506e-01 -1.58182845e-01 -1.55289066e+00
-8.53472799e-02 -7.46997654e-01 -2.14235991e-01 3.80185694e-01
9.22267973e-01 4.23975736e-01 -2.11506680e-01 1.86909646e-01
-7.49534011e-01 6.31002009e-01 1.47448063e+00 -6.87695816e-02
1.16718579e-02 -1.22854106e-01 5.96839413e-02 5.32812417e-01
6.94866180e-01 -9.17716175e-02 1.50833428e-01 -8.60723078e-01
3.75992805e-01 2.43860498e-01 3.81699473e-01 -1.34608948e+00
4.80899096e-01 -2.27786183e-01 7.77010679e-01 -1.05803561e+00
5.77564716e-01 -7.35698164e-01 -5.32825775e-02 2.31155694e-01
-1.39800683e-01 1.87231675e-01 3.27327639e-01 1.85565323e-01
-4.51426893e-01 4.03433740e-01 9.06419396e-01 -4.89422798e-01
-6.19544566e-01 9.56644595e-01 -4.97043878e-01 -2.43788585e-02
8.67238760e-01 -4.25057173e-01 -1.02318764e-01 -2.06903756e-01
-8.66740108e-01 1.15580130e-02 4.89864290e-01 3.34437639e-01
9.22287405e-01 -1.89678442e+00 -4.96321321e-01 8.54252100e-01
-1.25472173e-01 5.84944844e-01 3.19145098e-02 1.96001694e-01
-7.72880971e-01 1.06336868e+00 -7.24589825e-01 -8.05651665e-01
-7.60259926e-01 3.12431008e-01 5.22006571e-01 -1.90928262e-02
-7.66042352e-01 1.00057924e+00 -5.41607291e-02 -5.46414196e-01
4.45626110e-01 -4.62789208e-01 2.96529651e-01 -4.62292075e-01
3.89843494e-01 3.71440798e-01 5.07529497e-01 -6.84605837e-01
-3.23660105e-01 4.22494411e-01 1.33391649e-01 1.32912144e-01
1.97012818e+00 6.52855098e-01 -5.07998109e-01 6.08094811e-01
1.56606328e+00 -7.75501907e-01 -9.25253928e-01 -3.67239237e-01
9.09223706e-02 -2.47724056e-01 9.57675837e-03 -8.83637592e-02
-7.90743053e-01 1.23178995e+00 4.78258371e-01 5.61090827e-01
7.32814014e-01 1.15104392e-01 7.59115577e-01 1.07402682e+00
6.06648028e-01 -6.50770545e-01 3.46785665e-01 1.67012334e+00
1.12102520e+00 -1.11505353e+00 -6.43806625e-03 4.88934554e-02
1.51453421e-01 1.45682073e+00 5.81590295e-01 -8.41145754e-01
1.36056399e+00 4.89370435e-01 -4.79778796e-01 -8.13366115e-01
-9.27500784e-01 9.44488123e-02 4.90099996e-01 5.96571445e-01
7.72560179e-01 1.01887479e-01 4.86253560e-01 2.04170376e-01
-3.11501324e-01 -4.16098163e-02 5.13679385e-01 1.08208454e+00
-5.70048392e-01 -7.09870636e-01 -3.41428071e-01 8.26874614e-01
-1.16614789e-01 1.73258141e-01 -1.85385540e-01 5.80646038e-01
-7.65800253e-02 2.44135082e-01 7.60866106e-01 -1.75716370e-01
4.10497248e-01 -5.43150082e-02 6.83686376e-01 -8.05170059e-01
-2.47469023e-01 -2.19095454e-01 -2.47712091e-01 -7.43614793e-01
-2.10900143e-01 -4.15989786e-01 -1.38358295e+00 -6.06362700e-01
2.79154867e-01 -1.50329307e-01 1.02151251e+00 6.63984299e-01
4.88825530e-01 3.20357710e-01 4.57196921e-01 -1.55705583e+00
-3.21079493e-02 -7.60099888e-01 -7.23662019e-01 -6.83214664e-02
7.11597681e-01 -6.92375898e-01 1.34546071e-01 -3.52382094e-01] | [8.195103645324707, -3.7111027240753174] |
e4c3466d-b3a7-48e1-bb30-81d2577c3969 | mareo-memory-and-attention-based-visual | 2206.04928 | null | https://arxiv.org/abs/2206.04928v5 | https://arxiv.org/pdf/2206.04928v5.pdf | GAMR: A Guided Attention Model for (visual) Reasoning | Humans continue to outperform modern AI systems in their ability to flexibly parse and understand complex visual scenes. Here, we present a novel module for visual reasoning, the Guided Attention Model for (visual) Reasoning (GAMR), which instantiates an active vision theory -- positing that the brain solves complex visual reasoning problems dynamically -- via sequences of attention shifts to select and route task-relevant visual information into memory. Experiments on an array of visual reasoning tasks and datasets demonstrate GAMR's ability to learn visual routines in a robust and sample-efficient manner. In addition, GAMR is shown to be capable of zero-shot generalization on completely novel reasoning tasks. Overall, our work provides computational support for cognitive theories that postulate the need for a critical interplay between attention and memory to dynamically maintain and manipulate task-relevant visual information to solve complex visual reasoning tasks. | ['Thomas Serre', 'Mohit Vaishnav'] | 2022-06-10 | null | null | null | null | ['visual-reasoning', 'visual-reasoning'] | ['computer-vision', 'reasoning'] | [ 3.17647845e-01 1.16457626e-01 3.80376250e-01 -8.12340900e-02
6.57675043e-02 -8.21486473e-01 7.94253170e-01 2.00427428e-01
-2.08646894e-01 1.83789045e-01 1.63931072e-01 -7.60164857e-01
-4.73258972e-01 -6.14082277e-01 -4.85425144e-01 -3.12257856e-01
1.06355911e-02 4.74925399e-01 1.83307543e-01 -4.05499876e-01
8.72817099e-01 8.97405505e-01 -1.75290453e+00 5.04303336e-01
7.47795701e-01 2.80183464e-01 7.26496816e-01 1.11502385e+00
-2.33826116e-01 1.59371865e+00 -3.60069633e-01 -2.42692024e-01
3.22542816e-01 -5.30963600e-01 -1.17044890e+00 2.24181756e-01
4.92937505e-01 -2.07529843e-01 -2.95673162e-01 8.01560760e-01
4.95413579e-02 6.85801148e-01 7.00251281e-01 -1.05082297e+00
-1.36962616e+00 1.59294784e-01 -3.92029077e-01 1.01354074e+00
5.53037703e-01 1.11614943e+00 8.62598538e-01 -9.48163390e-01
7.59228945e-01 1.55423999e+00 2.62048900e-01 6.09891951e-01
-1.55288422e+00 -2.11345211e-01 3.26643556e-01 6.38128042e-01
-9.79362965e-01 -4.90812719e-01 4.95663792e-01 -7.21035182e-01
1.72265697e+00 5.10353148e-01 1.28405631e+00 7.04119980e-01
4.71868813e-01 6.68978214e-01 1.32930267e+00 -4.71369028e-01
2.66012639e-01 -3.07595402e-01 4.03114349e-01 9.74366009e-01
3.74873579e-01 3.26462150e-01 -8.63075912e-01 1.10292390e-01
1.08739412e+00 -5.48391081e-02 -1.63179800e-01 -5.33944726e-01
-1.31356895e+00 5.24787128e-01 6.32752419e-01 1.86772585e-01
-6.76061273e-01 3.41680348e-01 2.06496969e-01 4.21562552e-01
-3.75871807e-01 1.16055501e+00 -1.53459320e-02 1.64640531e-01
-3.74147683e-01 1.70705110e-01 4.25135076e-01 7.65004337e-01
7.35114753e-01 4.05104816e-01 -7.96322465e-01 4.24567521e-01
6.46039486e-01 4.96561527e-01 2.69481391e-01 -1.34470546e+00
2.24351719e-01 8.31123233e-01 -8.01552981e-02 -1.07664740e+00
-4.23722267e-01 2.42333878e-02 -3.14393580e-01 7.34169722e-01
3.66091341e-01 4.42472756e-01 -1.12345219e+00 1.53670955e+00
1.28245935e-01 -3.65083575e-01 1.83636218e-01 1.01738632e+00
9.03630912e-01 5.14109910e-01 7.26469994e-01 -1.25469685e-01
1.42487562e+00 -8.53743315e-01 -4.43339705e-01 -7.19856560e-01
2.34371107e-02 -2.84237772e-01 1.50339305e+00 1.71412721e-01
-1.64863014e+00 -7.88335502e-01 -1.01442420e+00 -7.10271716e-01
-4.72775310e-01 -5.53759098e-01 1.02649534e+00 3.38875830e-01
-1.47959960e+00 1.86762527e-01 -6.20520353e-01 -5.04059553e-01
7.81913698e-01 4.17421162e-02 -1.69739649e-01 -3.93028051e-01
-6.90739393e-01 1.33491802e+00 5.70123792e-01 1.97791785e-01
-1.44440901e+00 -7.84057081e-01 -8.29424798e-01 3.39499325e-01
4.55096334e-01 -1.37251401e+00 1.31192100e+00 -1.11456180e+00
-1.15004373e+00 1.30944479e+00 -1.52761355e-01 -3.50906253e-01
6.50508553e-02 7.53784105e-02 1.06973156e-01 6.45188570e-01
-3.37077715e-02 7.27224708e-01 8.73489141e-01 -1.44628942e+00
-1.76363572e-01 -4.55075681e-01 5.19949675e-01 4.45771039e-01
4.11855370e-01 -1.72071666e-01 -1.71632245e-01 -5.22331178e-01
7.62253106e-02 -5.85813761e-01 -1.95173010e-01 3.02172035e-01
1.66794777e-01 -3.30039918e-01 3.13516051e-01 -6.13240123e-01
7.12485194e-01 -1.99825263e+00 6.10771120e-01 -8.64385860e-04
8.92631292e-01 2.93918967e-01 -2.78538734e-01 4.80144471e-01
-5.19711189e-02 -3.02288271e-02 -2.26345845e-02 9.12477449e-02
9.14348513e-02 3.42274457e-01 -5.07767320e-01 1.31254494e-01
1.95690617e-01 1.74444890e+00 -1.07688951e+00 -3.64464909e-01
4.29519475e-01 3.08394849e-01 -3.53453487e-01 3.38606656e-01
-4.36100036e-01 4.31485862e-01 -7.97165707e-02 4.98736233e-01
2.96653390e-01 -6.03895247e-01 4.63427126e-01 8.77265930e-02
-5.54821566e-02 -3.36145610e-03 -4.80181456e-01 1.59110248e+00
-1.34916335e-01 1.01824868e+00 -7.52383098e-02 -7.40719020e-01
7.55312860e-01 -4.12878543e-02 -4.17270541e-01 -1.30541813e+00
1.22256808e-01 -5.77588856e-01 4.49030370e-01 -7.59011745e-01
3.05008948e-01 -6.56508654e-02 2.42897868e-01 7.48757005e-01
1.18574217e-01 -5.19513011e-01 4.04263556e-01 5.48803031e-01
9.70390558e-01 1.14919633e-01 7.47199535e-01 -4.54389900e-01
5.10577500e-01 3.96289229e-01 -8.14739242e-02 1.31843591e+00
-3.46108079e-01 -2.13446189e-02 1.82661042e-01 -6.85558975e-01
-9.51142251e-01 -1.63490176e+00 5.58826149e-01 1.63490093e+00
1.32840008e-01 -2.18503028e-01 -5.06508589e-01 -6.31313249e-02
-5.04113175e-02 1.23220062e+00 -7.83369064e-01 -4.76279825e-01
-3.60686541e-01 -2.24777803e-01 1.24199554e-01 6.43098474e-01
4.30548698e-01 -1.92131734e+00 -1.54999375e+00 -1.77475452e-01
1.16840340e-01 -5.45400381e-01 8.27340409e-04 -7.73915574e-02
-7.59072244e-01 -1.38473248e+00 -2.06728771e-01 -6.74734771e-01
1.00883758e+00 1.03986859e+00 1.52055562e+00 6.90182567e-01
-1.08797550e+00 1.41642129e+00 -8.06661025e-02 -3.06569993e-01
-3.79052877e-01 -5.60509443e-01 -6.19296849e-01 -5.12757063e-01
2.84731060e-01 -3.70682210e-01 -5.57522118e-01 -1.26816183e-01
-8.00709605e-01 6.39825165e-01 6.05964363e-01 5.27771175e-01
3.90531540e-01 -2.83607930e-01 1.43378079e-01 -5.78866124e-01
1.16506445e+00 -4.05384511e-01 -6.96067095e-01 6.54889524e-01
-5.56885362e-01 2.21702874e-01 3.70511681e-01 -3.85983407e-01
-1.34539711e+00 -1.63978159e-01 6.81344211e-01 -3.29055876e-01
-1.36913851e-01 2.65357107e-01 1.28595918e-01 -2.09305286e-01
8.27412188e-01 6.43274546e-01 -1.56299129e-01 2.23314926e-01
8.25125277e-01 -2.28909045e-01 6.53020382e-01 -8.82834971e-01
7.72400975e-01 4.74548638e-01 1.99282095e-01 -8.13069344e-01
-8.47854197e-01 -1.64028987e-01 -6.40079498e-01 -6.32757246e-01
1.21617186e+00 -7.20527053e-01 -1.47125947e+00 2.05340102e-01
-9.83693302e-01 -7.28201985e-01 -4.51152354e-01 1.52849583e-02
-1.01141715e+00 2.95716286e-01 -4.26531285e-01 -8.92049193e-01
-3.37136060e-01 -9.41476882e-01 6.10619485e-01 5.05737126e-01
-5.93362808e-01 -9.70352232e-01 -2.19178014e-02 6.11154497e-01
6.57217443e-01 9.03465897e-02 1.28964913e+00 -1.75549075e-01
-1.20191741e+00 4.98823553e-01 -6.35805309e-01 -3.27393442e-01
-3.98176759e-01 -1.34226114e-01 -8.63531291e-01 -2.88317770e-01
2.26441529e-02 -6.12072825e-01 1.04271102e+00 1.05802812e-01
1.17280102e+00 -6.82573617e-02 5.95533522e-04 5.79621732e-01
1.38936484e+00 5.08219719e-01 8.36510539e-01 1.61076546e-01
6.20444834e-01 5.22550046e-01 3.03488523e-01 9.82262641e-02
6.56384170e-01 2.75191888e-02 3.27242076e-01 1.67839557e-01
-3.46724302e-01 -1.50927752e-01 3.34526673e-02 1.45206586e-01
-5.29184461e-01 1.19063571e-01 -1.35088968e+00 4.09036875e-01
-1.77008760e+00 -1.56429029e+00 -9.85590890e-02 1.92231166e+00
6.53876185e-01 9.70242098e-02 -2.02182397e-01 -2.01744780e-01
2.24600002e-01 1.18492931e-01 -8.75807226e-01 -7.31428981e-01
1.28344581e-01 3.86406809e-01 -9.81574804e-02 4.92504895e-01
-4.25897837e-01 1.28804743e+00 7.94544363e+00 -8.92469008e-03
-6.02642596e-01 1.28538432e-02 1.40805021e-01 -7.58072957e-02
-2.61017233e-01 9.79285091e-02 -1.44420356e-01 -2.52826065e-01
7.11075723e-01 -5.54315269e-01 1.19460857e+00 5.63899279e-01
-6.54212013e-02 -4.74224120e-01 -1.13962805e+00 1.10126448e+00
4.57432300e-01 -1.45929444e+00 5.06248653e-01 -1.59161568e-01
3.90116215e-01 -5.02329469e-01 3.16269428e-01 5.27856946e-01
8.94650996e-01 -1.49434364e+00 6.71415687e-01 1.11584234e+00
3.09758425e-01 -4.68945891e-01 -1.90176696e-01 3.26833189e-01
-1.05337906e+00 -4.80239749e-01 -4.76083547e-01 -5.44696152e-01
-1.63063794e-01 -2.44866312e-01 -8.96936238e-01 -1.07542247e-01
7.35393047e-01 3.94919872e-01 -1.08450735e+00 8.77871454e-01
-5.87293804e-01 -1.07199356e-01 5.84368050e-01 -8.70963037e-02
-5.27145602e-02 2.09579974e-01 4.58613187e-01 9.42422569e-01
-2.40938365e-01 4.89187926e-01 7.82712251e-02 1.24966383e+00
4.35259998e-01 -2.60593355e-01 -6.66417658e-01 -3.09941202e-01
4.60642964e-01 1.05942595e+00 -9.91305292e-01 -4.78266418e-01
-2.01720014e-01 1.05229747e+00 8.64283264e-01 6.53005183e-01
-6.38032079e-01 9.18369219e-02 5.06808162e-01 -1.00757971e-01
4.07207072e-01 -7.48775959e-01 -4.03509170e-01 -1.02802634e+00
-4.69515890e-01 -9.83369291e-01 4.68154162e-01 -1.71877587e+00
-1.13000834e+00 2.41961271e-01 1.07135326e-01 -3.30207318e-01
-2.01691151e-01 -9.13333058e-01 -4.67039257e-01 9.08425987e-01
-1.20832610e+00 -1.19394851e+00 -7.62896359e-01 1.05202019e+00
7.21281409e-01 -2.65319973e-01 9.15971756e-01 -7.41557419e-01
-1.34481505e-01 -1.70711707e-02 -6.90658927e-01 -4.99858521e-02
6.83960095e-02 -1.32638431e+00 7.40301490e-01 8.17631602e-01
2.00252905e-01 1.20860124e+00 6.25283480e-01 -6.01869524e-01
-1.98983908e+00 -4.82593328e-01 1.84232816e-01 -9.16882694e-01
4.34561253e-01 -4.53114420e-01 -1.04447639e+00 9.84391391e-01
5.36756754e-01 -2.20005170e-01 8.20928693e-01 9.88934562e-02
-8.81286025e-01 3.11490357e-01 -1.07803977e+00 1.07652915e+00
1.38593292e+00 -8.27981770e-01 -1.61720395e+00 1.82036698e-01
5.67926049e-01 -2.05927730e-01 -3.16731870e-01 -1.71461836e-01
5.26300907e-01 -1.15609312e+00 1.47465432e+00 -1.01227474e+00
2.96704501e-01 -4.60473120e-01 -6.17994592e-02 -1.01665747e+00
-1.14170897e+00 -3.24405223e-01 -4.48203206e-01 4.91275728e-01
-1.35395288e-01 -5.56689799e-01 -1.78819187e-02 7.55645037e-01
7.87356645e-02 -1.09089725e-01 -5.28810263e-01 -3.04380238e-01
-2.87049204e-01 -3.91417503e-01 2.27825075e-01 8.84907305e-01
-3.04337181e-02 5.69047153e-01 2.20188767e-01 7.47266412e-02
7.02897072e-01 2.50834793e-01 6.72426760e-01 -1.51622891e+00
-3.28096360e-01 -7.38793075e-01 -3.42486292e-01 -5.86337805e-01
3.10032487e-01 -1.01670694e+00 -1.04975611e-01 -2.05273604e+00
4.70482618e-01 2.87202895e-01 -1.95206255e-01 6.48658872e-01
-1.27353936e-01 -7.67681655e-03 8.07176352e-01 1.75903603e-01
-9.24550235e-01 3.00186500e-03 1.67745018e+00 -6.42682463e-02
-1.30634531e-01 -6.99957609e-01 -1.19668829e+00 6.40390277e-01
5.80186069e-01 7.34957755e-02 -8.92453849e-01 -4.54543680e-01
5.66057980e-01 1.24826498e-01 9.51109767e-01 -8.74437153e-01
2.73265600e-01 -7.79189348e-01 9.97916341e-01 -3.79865468e-01
1.97074994e-01 -7.16542125e-01 4.08412144e-02 6.05272651e-01
-5.03561974e-01 4.79724437e-01 5.70520341e-01 6.14135623e-01
5.45179665e-01 -2.03434937e-02 8.23101640e-01 -8.52075458e-01
-1.63088131e+00 -2.23728940e-01 -8.88438702e-01 2.41623580e-01
1.40066183e+00 -4.20717150e-01 -9.40802813e-01 -6.20909892e-02
-1.04967439e+00 2.99855381e-01 4.52677190e-01 5.33204436e-01
1.04004467e+00 -9.72372174e-01 -4.75627273e-01 2.51666337e-01
3.25881749e-01 -3.79410088e-01 6.20205700e-01 4.05109942e-01
-8.25166345e-01 3.81518960e-01 -7.81257153e-01 -4.35908020e-01
-1.14219427e+00 1.28588223e+00 4.69324380e-01 1.79435492e-01
-8.45943093e-01 8.11914384e-01 4.58036423e-01 1.25549614e-01
-1.49431050e-01 -2.36626506e-01 -2.80816942e-01 -8.81817117e-02
1.07163763e+00 2.23672763e-01 -5.51851988e-01 -3.58660251e-01
-3.01480412e-01 4.58786368e-01 2.18768809e-02 -9.96165574e-02
1.05879068e+00 -3.38639885e-01 -4.96493369e-01 5.95019102e-01
2.83094317e-01 -3.79116207e-01 -1.36605990e+00 4.66515608e-02
-8.17159712e-02 -6.56740725e-01 -2.17512727e-01 -1.15563309e+00
-5.76179683e-01 9.55851555e-01 3.49828869e-01 1.52205691e-01
1.13340163e+00 1.99031979e-01 -2.44628295e-01 8.83327544e-01
1.93257362e-01 -9.11630690e-01 7.43974984e-01 6.68144226e-01
1.27463973e+00 -9.45382953e-01 2.36513242e-01 1.02094859e-01
-8.33441138e-01 1.17820251e+00 9.16818559e-01 -1.84325948e-01
9.70759988e-02 -8.63240734e-02 -6.10385798e-02 -7.06907094e-01
-1.19456244e+00 -4.91751790e-01 4.54192013e-01 1.09989762e+00
2.11033240e-01 -9.60323140e-02 1.95850149e-01 -5.03223874e-02
-1.37033716e-01 -1.17124349e-01 4.30704266e-01 9.65823948e-01
-7.85947382e-01 -1.10434420e-01 -5.96924484e-01 2.52729982e-01
3.72157842e-01 -1.87689647e-01 -6.12571955e-01 9.07622218e-01
-1.00585498e-01 8.69828403e-01 2.89757669e-01 3.14986646e-01
1.51842847e-01 4.44934249e-01 1.14245820e+00 -7.76355267e-01
-7.65914857e-01 -5.96601605e-01 -3.70787442e-01 -7.87756622e-01
-4.55261320e-01 -6.18899822e-01 -1.51014161e+00 -3.39640617e-01
5.92270255e-01 -4.76520479e-01 9.28946659e-02 9.62952614e-01
3.42860729e-01 9.70181525e-01 -3.38148028e-01 -8.30818236e-01
-2.55252957e-01 -4.90236938e-01 -2.40751341e-01 5.97003520e-01
2.83027083e-01 -6.95804536e-01 -2.65116006e-01 2.15563312e-01] | [10.612617492675781, 2.261005163192749] |
1ad23e5c-d343-4319-89d4-2569f754e3a4 | a-semantically-enhanced-dual-encoder-for | 2306.08373 | null | https://arxiv.org/abs/2306.08373v1 | https://arxiv.org/pdf/2306.08373v1.pdf | A semantically enhanced dual encoder for aspect sentiment triplet extraction | Aspect sentiment triplet extraction (ASTE) is a crucial subtask of aspect-based sentiment analysis (ABSA) that aims to comprehensively identify sentiment triplets. Previous research has focused on enhancing ASTE through innovative table-filling strategies. However, these approaches often overlook the multi-perspective nature of language expressions, resulting in a loss of valuable interaction information between aspects and opinions. To address this limitation, we propose a framework that leverages both a basic encoder, primarily based on BERT, and a particular encoder comprising a Bi-LSTM network and graph convolutional network (GCN ). The basic encoder captures the surface-level semantics of linguistic expressions, while the particular encoder extracts deeper semantics, including syntactic and lexical information. By modeling the dependency tree of comments and considering the part-of-speech and positional information of words, we aim to capture semantics that are more relevant to the underlying intentions of the sentences. An interaction strategy combines the semantics learned by the two encoders, enabling the fusion of multiple perspectives and facilitating a more comprehensive understanding of aspect--opinion relationships. Experiments conducted on benchmark datasets demonstrate the state-of-the-art performance of our proposed framework. | ['Hongye Li', 'Kaifang Dong', 'Peiyu Liu', 'Shehui Liang', 'Baoxing Jiang'] | 2023-06-14 | null | null | null | null | ['sentiment-analysis', 'aspect-based-sentiment-analysis', 'aspect-sentiment-triplet-extraction'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 1.27403975e-01 2.83149838e-01 -3.27407688e-01 -7.09377468e-01
-5.94226539e-01 -5.81861854e-01 6.77200317e-01 5.72881103e-01
-1.45338029e-01 3.06719691e-01 8.67008507e-01 -3.90212178e-01
1.24924943e-01 -1.00000525e+00 -4.72816020e-01 -3.81532311e-01
2.43470713e-01 3.79301943e-02 -1.32918179e-01 -6.84112310e-01
1.57764882e-01 -5.08427992e-02 -1.31570542e+00 7.95514166e-01
6.27442539e-01 1.27899158e+00 -2.60435015e-01 1.36334434e-01
-9.05864239e-01 1.20529568e+00 -4.54051703e-01 -9.70017493e-01
-9.41075012e-02 -3.02149087e-01 -6.96013272e-01 2.83449858e-01
3.70454304e-02 7.23291263e-02 8.37880895e-02 1.10530579e+00
9.65833813e-02 -9.93050411e-02 3.47927064e-01 -1.04147851e+00
-6.89334691e-01 1.06387067e+00 -6.22200131e-01 -1.79035962e-01
2.74896324e-01 -1.22870803e-01 1.83695996e+00 -7.22968757e-01
4.62104738e-01 1.17506075e+00 6.37918234e-01 1.79995224e-01
-6.84343219e-01 -2.27205679e-01 6.82561576e-01 -5.97603731e-02
-7.29368508e-01 -2.55712181e-01 1.06062090e+00 -3.60129744e-01
1.22692871e+00 5.87014742e-02 9.78365242e-01 1.00251043e+00
2.21612260e-01 1.15606463e+00 9.22417223e-01 -3.34991276e-01
2.80313455e-02 4.10140336e-01 6.35430813e-01 5.64001203e-01
3.61676395e-01 -4.07335490e-01 -8.89612079e-01 -1.04772523e-02
5.43193482e-02 9.71614942e-02 -2.18022671e-02 -2.94843405e-01
-8.54858160e-01 8.75161350e-01 3.44752848e-01 5.19406497e-01
-6.74940884e-01 -3.88376340e-02 7.61040449e-01 2.13021129e-01
6.26094878e-01 3.79660577e-01 -7.92881548e-01 -2.19213247e-01
-6.33122623e-01 -5.73302396e-02 1.02917457e+00 1.03850615e+00
8.89311969e-01 4.17740047e-02 -2.83585787e-01 4.81628835e-01
7.00638473e-01 2.68526971e-01 3.49709213e-01 -2.99631298e-01
5.82163334e-01 1.32724309e+00 -2.18920946e-01 -1.00649166e+00
-2.67838389e-01 -7.79442906e-01 -6.06855631e-01 -2.87130982e-01
-2.61433780e-01 -3.36624652e-01 -7.76092172e-01 1.64760578e+00
4.06464547e-01 -4.93241638e-01 2.73985714e-01 6.07774794e-01
9.92637396e-01 4.56567317e-01 1.02456614e-01 -4.06345166e-02
1.76788306e+00 -1.14709604e+00 -1.00457907e+00 -6.58428133e-01
8.05629253e-01 -6.57531977e-01 9.53507960e-01 -8.20974410e-02
-9.39588785e-01 -2.92093635e-01 -1.21868396e+00 -2.43007734e-01
-6.88533604e-01 6.43606856e-02 8.59105885e-01 3.58020306e-01
-9.83548760e-01 1.32455826e-01 -5.73452353e-01 -2.36810103e-01
4.20246243e-01 2.52777278e-01 -2.41807893e-01 1.60135850e-01
-1.34885955e+00 7.12904096e-01 2.76484936e-01 3.09924781e-01
-2.01941386e-01 -7.01826632e-01 -1.25831103e+00 2.84365922e-01
6.78646386e-01 -9.71046090e-01 1.09591174e+00 -1.19360638e+00
-1.22526157e+00 5.73079407e-01 -5.11944413e-01 -3.38405013e-01
-1.73425436e-01 -4.14897919e-01 -4.64316994e-01 -1.12450793e-01
3.64114493e-01 2.94995695e-01 6.37003303e-01 -1.25932837e+00
-7.92406797e-01 -8.27292085e-01 6.52452111e-01 4.13944572e-01
-7.84239292e-01 2.08213255e-01 -5.83377123e-01 -4.63451624e-01
-2.14060813e-01 -6.22453094e-01 -1.48672715e-01 -5.24783552e-01
-6.55268133e-01 -1.50671721e-01 8.29749286e-01 -5.81767499e-01
1.56446159e+00 -2.18105340e+00 1.58830971e-01 1.46014526e-01
4.21605587e-01 2.09128723e-01 -1.02404617e-01 7.86936939e-01
-6.11690525e-03 1.00187860e-01 -2.83545434e-01 -6.23941362e-01
2.65229940e-01 1.08837277e-01 -4.60940629e-01 -1.81993917e-01
4.25704122e-01 1.34343469e+00 -7.84892678e-01 -2.23651275e-01
-1.17406934e-01 5.96569657e-01 -4.75334287e-01 1.43928140e-01
-4.92367506e-01 2.11660221e-01 -7.21990466e-01 7.67073035e-01
3.39794457e-01 -4.50602680e-01 2.69208312e-01 -5.24830759e-01
-3.27052362e-02 8.32679451e-01 -7.77910948e-01 1.46327531e+00
-7.86487103e-01 4.45070773e-01 2.14368924e-02 -8.66861045e-01
9.68914986e-01 2.70706475e-01 5.13007462e-01 -7.61235833e-01
3.36657494e-01 1.44128278e-01 -1.51068136e-01 -3.82884234e-01
7.06690371e-01 -4.01222140e-01 -2.57519037e-01 6.73582435e-01
1.84764236e-01 -1.81400479e-04 3.14486861e-01 4.45050269e-01
7.72577226e-01 2.25982890e-01 6.52372956e-01 -9.69628245e-02
8.04684222e-01 -2.18850952e-02 4.71145362e-01 1.87154621e-01
7.34969452e-02 6.04222007e-02 9.98351157e-01 -5.33543527e-01
-6.64902747e-01 -3.53142291e-01 2.97973692e-01 9.80188429e-01
7.05527142e-02 -9.81679261e-01 -4.52907920e-01 -9.30950522e-01
-1.95645466e-01 1.00093257e+00 -1.00538313e+00 -1.70106962e-01
-5.13171792e-01 -6.94157183e-01 1.79685265e-01 6.69344664e-01
5.07405877e-01 -1.02943683e+00 -4.43951577e-01 1.83349133e-01
-2.91169107e-01 -1.65777910e+00 -2.66692698e-01 2.40533650e-01
-6.43837154e-01 -1.07644498e+00 -1.27878919e-01 -5.44158280e-01
4.94880110e-01 3.54240328e-01 1.41941500e+00 8.59272406e-02
3.70230019e-01 4.23635393e-01 -6.62780404e-01 -6.24754012e-01
-2.95660049e-01 2.78673828e-01 -4.61263627e-01 4.80919123e-01
9.85213697e-01 -5.06914139e-01 -4.56306010e-01 -2.47314095e-01
-1.01264751e+00 1.53842449e-01 7.90636182e-01 5.79552174e-01
5.26358128e-01 7.19053717e-03 2.49932274e-01 -1.35700893e+00
9.03425813e-01 -5.62271357e-01 -1.75889060e-01 3.50299865e-01
-5.26640475e-01 1.22414880e-01 7.14172721e-01 2.14804128e-01
-1.23640430e+00 -4.06703562e-01 -2.11029902e-01 9.73384827e-02
5.76640777e-02 1.12320113e+00 -3.61388892e-01 4.11291867e-01
-4.51083248e-03 3.23897362e-01 -2.28669077e-01 -1.24457397e-01
6.16941631e-01 5.90541065e-01 -5.55588491e-02 -5.07625222e-01
3.81761461e-01 6.29108787e-01 -8.74048620e-02 -8.47262740e-01
-1.48334825e+00 -6.84552729e-01 -6.55323744e-01 -1.94167197e-01
9.10680830e-01 -1.03792703e+00 -3.96908283e-01 3.34911585e-01
-1.15314734e+00 2.80412614e-01 -4.90298539e-01 1.30258605e-01
-2.12530896e-01 2.60442376e-01 -5.10088444e-01 -7.31787860e-01
-7.17020988e-01 -1.28599966e+00 1.45655012e+00 1.90577358e-01
-2.52650768e-01 -1.27950370e+00 -5.81903420e-02 5.65511525e-01
4.68913287e-01 1.31613851e-01 1.16493857e+00 -9.21828330e-01
-4.69786942e-01 -3.01895320e-01 -2.89882302e-01 3.96388859e-01
4.33733672e-01 -9.01412964e-02 -9.91898954e-01 5.52818142e-02
2.92322725e-01 -1.97785452e-01 8.48156571e-01 2.02040583e-01
4.67507660e-01 -4.39177811e-01 6.07713917e-03 2.75239766e-01
1.43386853e+00 6.88089803e-02 3.05631310e-01 5.58358848e-01
1.00599217e+00 9.12958741e-01 5.18784702e-01 4.16716248e-01
1.11007130e+00 4.89260018e-01 5.19637108e-01 -9.42832381e-02
-6.92942664e-02 -3.62171263e-01 5.72635531e-01 1.41333926e+00
1.82098836e-01 -1.00922450e-01 -6.52994275e-01 7.10634172e-01
-1.87298274e+00 -5.72236657e-01 -1.90182626e-01 1.63734102e+00
7.14107454e-01 3.16533059e-01 -6.66568056e-02 -1.26805790e-02
2.76858687e-01 7.30774045e-01 -4.34873879e-01 -7.35875666e-01
-2.10681051e-01 -7.73782954e-02 3.89735810e-02 4.10382688e-01
-1.03299582e+00 1.08999145e+00 4.82329369e+00 4.62723672e-01
-1.00608778e+00 -3.33515294e-02 4.40342486e-01 2.14962706e-01
-1.01721799e+00 2.78650224e-01 -9.29612517e-01 3.62972841e-02
8.46016109e-01 -2.15752095e-01 -5.63806593e-02 8.05860996e-01
5.83271384e-02 1.62649199e-01 -9.54827964e-01 3.93661261e-01
3.21095735e-01 -1.23285246e+00 5.22932053e-01 5.62572554e-02
6.29459262e-01 -3.42207067e-02 -1.27171967e-02 3.95013839e-01
1.98308975e-01 -5.59040129e-01 7.45987177e-01 3.67531002e-01
2.38410905e-01 -7.33897269e-01 1.09872878e+00 5.27445525e-02
-1.52080750e+00 -4.51967642e-02 -7.59422556e-02 -1.32037506e-01
3.12435567e-01 8.92036140e-01 -4.90952641e-01 9.93652165e-01
5.87980747e-01 1.31596684e+00 -5.55947185e-01 2.84733623e-01
-6.61603451e-01 4.26772356e-01 1.74678028e-01 -3.16564679e-01
6.96526885e-01 -4.49198127e-01 6.00765347e-01 1.27237296e+00
1.50334288e-03 -1.58292949e-01 -7.70495757e-02 8.11224282e-01
-1.21456474e-01 4.14607048e-01 -7.61941493e-01 -5.46938419e-01
-1.02518015e-02 1.49318373e+00 -5.79089880e-01 -5.10231793e-01
-1.02414310e+00 5.05059779e-01 4.06027943e-01 3.84276122e-01
-4.75136071e-01 -1.89727366e-01 9.18464720e-01 -2.04906046e-01
6.36115313e-01 -1.70556143e-01 -6.26114428e-01 -1.47980416e+00
3.84135157e-01 -1.00791192e+00 3.51836890e-01 -6.33871794e-01
-1.16359115e+00 9.11972046e-01 -2.29530320e-01 -1.05431187e+00
-2.88939744e-01 -5.60713947e-01 -6.71027005e-01 7.05519199e-01
-1.91447246e+00 -1.77294612e+00 5.83187938e-02 3.87124240e-01
6.89687848e-01 -1.00823455e-02 6.99095309e-01 1.71304584e-01
-5.47306240e-01 3.50912333e-01 -6.12205327e-01 1.71795905e-01
1.59807697e-01 -1.29859531e+00 5.82073808e-01 1.02169383e+00
3.09393525e-01 9.90995109e-01 3.91196728e-01 -5.97498298e-01
-1.50739956e+00 -1.03258777e+00 1.53750503e+00 -3.30043107e-01
1.12307858e+00 -3.15771550e-01 -7.56019831e-01 8.71063113e-01
5.88422656e-01 -5.17222464e-01 1.16492581e+00 5.14462113e-01
-6.19270205e-01 -9.61815268e-02 -4.21234012e-01 7.78115809e-01
6.68300271e-01 -8.94821763e-01 -8.28352213e-01 -4.91588749e-02
1.18255937e+00 -3.96788791e-02 -7.21556842e-01 5.54491043e-01
5.82544029e-01 -1.18490505e+00 5.47127604e-01 -8.13792467e-01
9.02006447e-01 -2.18839779e-01 -4.30956423e-01 -1.31695819e+00
-2.90643913e-03 -2.27974519e-01 -3.28617632e-01 1.49852264e+00
7.94680834e-01 -4.80978727e-01 4.88299578e-01 5.90013146e-01
-1.78731993e-01 -1.25142717e+00 -4.19362485e-01 -9.81115624e-02
-3.06180716e-01 -7.38049746e-01 8.83305371e-01 8.09254408e-01
2.29411140e-01 1.04109716e+00 -2.29944989e-01 9.46032256e-02
3.03778887e-01 5.81942260e-01 5.12763202e-01 -9.73356426e-01
-2.42093667e-01 -6.08083367e-01 -2.46738046e-01 -9.28157866e-01
3.42391640e-01 -7.61765838e-01 -2.58778721e-01 -1.88328898e+00
3.02328378e-01 1.01719303e-02 -4.01822418e-01 6.20889843e-01
-3.78721595e-01 -7.14075342e-02 1.75593704e-01 -1.57076225e-01
-8.08016837e-01 9.37799275e-01 1.33908558e+00 -2.84144998e-01
-3.48278470e-02 -3.66040803e-02 -1.52312016e+00 8.59416068e-01
6.53811514e-01 -3.07585031e-01 -6.15763247e-01 -6.74637914e-01
9.80422080e-01 -2.37271860e-01 -4.25204784e-02 -3.74740154e-01
3.42960656e-01 7.56169632e-02 -1.96106777e-01 -5.94387650e-01
2.56462157e-01 -9.40390587e-01 -4.40771282e-01 1.46011099e-01
-4.18412417e-01 1.94987103e-01 1.87201560e-01 5.21644831e-01
-8.35021615e-01 -7.59451240e-02 7.91822001e-02 -2.33513653e-01
-7.13939905e-01 3.10662836e-01 -1.45580456e-01 1.29689679e-01
5.55801392e-01 9.71208094e-04 -2.10362479e-01 -5.48412800e-01
-3.90624732e-01 1.77807271e-01 2.50664324e-01 6.99702799e-01
5.59582889e-01 -1.15529263e+00 -4.22043025e-01 4.08637106e-01
5.30607343e-01 -4.94015515e-02 1.29211605e-01 9.13787901e-01
1.42003223e-02 5.52738547e-01 1.51369974e-01 -2.29402468e-01
-1.08950508e+00 4.26973611e-01 1.84368625e-01 -6.99995160e-01
-4.41816449e-01 7.44971752e-01 3.57492596e-01 -5.10473132e-01
-1.10302962e-01 -4.69316870e-01 -8.74125004e-01 5.51886916e-01
3.51732105e-01 -1.61583424e-01 3.03709477e-01 -1.03102875e+00
-3.81130964e-01 5.51900864e-01 -2.12322205e-01 -3.29978056e-02
1.41919112e+00 -4.18375075e-01 -6.44243777e-01 6.48422480e-01
1.26690161e+00 2.07263127e-01 -6.78356767e-01 -4.82270569e-01
2.26776600e-01 -7.70590454e-02 8.85354877e-02 -6.15414798e-01
-1.20479858e+00 1.04935443e+00 -4.26200688e-01 3.63407433e-01
1.15555465e+00 2.77752941e-03 1.11386466e+00 2.90287048e-01
2.25308433e-01 -8.85940731e-01 3.36660407e-02 8.38898838e-01
6.88331664e-01 -1.13056827e+00 -7.43536130e-02 -4.91719604e-01
-1.05006158e+00 1.13835359e+00 5.35308003e-01 3.01404864e-01
7.62409091e-01 2.82137245e-01 3.82505029e-01 -6.58517003e-01
-1.01298809e+00 -5.50341189e-01 4.27379459e-01 1.05167218e-01
6.25285208e-01 7.94064328e-02 -3.28821123e-01 1.06615973e+00
-2.67617166e-01 -1.92272514e-01 3.86352807e-01 9.97244835e-01
-1.08057134e-01 -1.22664964e+00 3.45550984e-01 4.58697140e-01
-5.78713834e-01 -5.79590082e-01 -6.10483110e-01 5.23726642e-01
-2.14263182e-02 1.00207520e+00 -1.63084671e-01 -4.55440283e-01
5.23175657e-01 7.20569640e-02 -1.54542789e-01 -6.76464498e-01
-9.49419856e-01 1.85493022e-01 4.49874967e-01 -6.90422118e-01
-9.27651227e-01 -4.03907537e-01 -1.05050218e+00 6.45729601e-02
-1.55417413e-01 2.68674970e-01 8.07916105e-01 1.38505852e+00
5.99322796e-01 9.49732602e-01 5.97621858e-01 -9.96657908e-02
-2.08424032e-01 -8.32870781e-01 -5.73307097e-01 2.62299299e-01
4.24622595e-01 -3.37886214e-01 -1.54154494e-01 -2.45664924e-01] | [11.489026069641113, 6.653900146484375] |
2a9a20b4-2c34-4b00-83c7-d1d9b277eb02 | jrdb-act-a-large-scale-multi-modal-dataset | 2106.08827 | null | https://arxiv.org/abs/2106.08827v2 | https://arxiv.org/pdf/2106.08827v2.pdf | JRDB-Act: A Large-scale Dataset for Spatio-temporal Action, Social Group and Activity Detection | The availability of large-scale video action understanding datasets has facilitated advances in the interpretation of visual scenes containing people. However, learning to recognise human actions and their social interactions in an unconstrained real-world environment comprising numerous people, with potentially highly unbalanced and long-tailed distributed action labels from a stream of sensory data captured from a mobile robot platform remains a significant challenge, not least owing to the lack of a reflective large-scale dataset. In this paper, we introduce JRDB-Act, as an extension of the existing JRDB, which is captured by a social mobile manipulator and reflects a real distribution of human daily-life actions in a university campus environment. JRDB-Act has been densely annotated with atomic actions, comprises over 2.8M action labels, constituting a large-scale spatio-temporal action detection dataset. Each human bounding box is labeled with one pose-based action label and multiple~(optional) interaction-based action labels. Moreover JRDB-Act provides social group annotation, conducive to the task of grouping individuals based on their interactions in the scene to infer their social activities~(common activities in each social group). Each annotated label in JRDB-Act is tagged with the annotators' confidence level which contributes to the development of reliable evaluation strategies. In order to demonstrate how one can effectively utilise such annotations, we develop an end-to-end trainable pipeline to learn and infer these tasks, i.e. individual action and social group detection. The data and the evaluation code is publicly available at https://jrdb.erc.monash.edu/. | ['Hamid Rezatofighi', 'Ian Reid', 'Silvio Savarese', 'Fatemeh Saleh', 'Mahsa Ehsanpour'] | 2021-06-16 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Ehsanpour_JRDB-Act_A_Large-Scale_Dataset_for_Spatio-Temporal_Action_Social_Group_and_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Ehsanpour_JRDB-Act_A_Large-Scale_Dataset_for_Spatio-Temporal_Action_Social_Group_and_CVPR_2022_paper.pdf | cvpr-2022-1 | ['action-understanding'] | ['computer-vision'] | [ 4.07682747e-01 1.82698175e-01 2.40312353e-01 -4.59094167e-01
-5.96674085e-01 -4.72698867e-01 6.53235197e-01 -1.24182373e-01
-5.60971141e-01 6.97071970e-01 6.19141459e-01 3.44963551e-01
-1.75238580e-01 -2.09283903e-01 -6.40645981e-01 -5.65465033e-01
-5.83751321e-01 9.46193516e-01 4.32246208e-01 -1.66485786e-01
-1.73279028e-02 3.17242354e-01 -1.67444968e+00 5.08899748e-01
1.90163866e-01 7.47522473e-01 3.16225350e-01 1.05567455e+00
5.30494213e-01 1.23142219e+00 -6.27889633e-01 -2.42037848e-01
4.16444093e-01 -4.88796502e-01 -1.03654313e+00 5.06685913e-01
5.61766386e-01 -5.93069494e-01 -5.47764063e-01 6.40867770e-01
5.28134584e-01 5.84888160e-01 5.75414538e-01 -1.67117846e+00
1.22942738e-02 2.60592669e-01 -2.78070301e-01 2.91760772e-01
9.40237343e-01 5.76795042e-01 1.01873863e+00 -4.14504111e-01
9.68403578e-01 1.59174776e+00 6.49421036e-01 5.77336311e-01
-9.61322188e-01 -3.65401059e-01 1.65739596e-01 2.11467817e-01
-1.12497783e+00 -5.09263992e-01 3.91388893e-01 -6.96242332e-01
1.17371809e+00 8.75128210e-02 9.45994496e-01 1.58063400e+00
-2.21919239e-01 1.11212957e+00 8.18037868e-01 -1.89875752e-01
8.57667327e-02 -3.77942145e-01 -1.56979397e-01 7.96859682e-01
-9.11638513e-02 -2.57456362e-01 -8.58702481e-01 -5.62004931e-02
9.21776831e-01 1.21565908e-01 4.26109023e-02 -6.20296717e-01
-1.76894295e+00 3.21143955e-01 4.92237657e-01 2.59261150e-02
-3.33523542e-01 6.00305736e-01 5.24069011e-01 2.01132029e-01
3.14291179e-01 2.19839424e-01 -4.38695043e-01 -7.01629281e-01
-1.53969511e-01 6.40256941e-01 6.59210742e-01 9.55837965e-01
5.68519592e-01 -4.37274188e-01 -1.28275663e-01 7.15320587e-01
3.97473544e-01 5.04308224e-01 1.43958345e-01 -1.55031252e+00
6.42312706e-01 8.41069043e-01 4.89481986e-01 -9.85261977e-01
-7.35632360e-01 2.20064476e-01 -2.87902564e-01 2.32930407e-01
6.73978329e-01 -1.90317258e-01 -6.72150731e-01 1.80821490e+00
6.38651967e-01 1.79918393e-01 -3.92111689e-02 9.97753203e-01
4.49342132e-01 1.84814423e-01 3.08018804e-01 2.05013677e-01
1.35099685e+00 -1.05099070e+00 -5.05386055e-01 -5.25476635e-01
9.21079874e-01 -2.38582805e-01 9.87270892e-01 3.61431271e-01
-6.85069084e-01 -5.57940722e-01 -5.24306059e-01 -4.22586920e-03
-2.80561894e-01 1.70871004e-01 7.07970500e-01 1.89763114e-01
-8.99714828e-01 3.30858886e-01 -1.20791328e+00 -9.23976779e-01
6.78810954e-01 4.55924124e-01 -8.49945188e-01 -1.24741718e-01
-8.29078257e-01 7.90693164e-01 4.15198326e-01 -3.29386666e-02
-1.29276156e+00 7.53633771e-03 -1.00201726e+00 -4.83323902e-01
8.92961800e-01 -5.43880582e-01 1.53076625e+00 -1.02319145e+00
-1.08672106e+00 9.98058915e-01 1.27020434e-01 -3.05760056e-01
9.03949916e-01 -4.62390065e-01 -2.32153520e-01 4.03522074e-01
6.22721195e-01 9.37175572e-01 5.04980922e-01 -1.05253732e+00
-1.09498620e+00 -5.76894045e-01 3.11654419e-01 7.20706224e-01
-3.17018740e-02 3.44971329e-01 -5.29797018e-01 -5.52460909e-01
-2.58955985e-01 -1.36740482e+00 -2.53596246e-01 -4.06624228e-02
-2.16386095e-01 -4.34777111e-01 7.22667277e-01 -6.61813974e-01
7.63664842e-01 -2.18888927e+00 5.19548059e-01 -1.86702713e-01
1.97563142e-01 -2.47649569e-02 5.09673879e-02 5.63243926e-01
2.27314606e-01 -3.84365678e-01 -2.58488148e-01 -6.20207310e-01
3.45458329e-01 3.47389102e-01 2.35405043e-01 7.84202278e-01
4.48634364e-02 8.80998969e-01 -1.32207072e+00 -6.32489026e-01
3.59499961e-01 2.40061834e-01 -4.73110855e-01 3.55191678e-01
-3.05880547e-01 8.27463150e-01 -5.19150257e-01 7.70266116e-01
-4.50420566e-02 -8.88568759e-02 2.98449427e-01 2.83191353e-01
1.06492385e-01 -1.19692765e-01 -1.22681332e+00 1.93139648e+00
1.72048301e-01 7.23413229e-01 2.15492710e-01 -8.23906302e-01
4.27489668e-01 2.60481179e-01 9.29377913e-01 -2.12690532e-01
2.13814482e-01 1.13252159e-02 -1.38442546e-01 -9.26802039e-01
2.93851256e-01 9.69894323e-03 -6.46321535e-01 5.12569726e-01
2.98084378e-01 2.31920153e-01 4.41998124e-01 4.04198229e-01
1.69875991e+00 5.44690847e-01 3.79051358e-01 2.10119829e-01
2.99252957e-01 2.35762104e-01 4.62785035e-01 7.86573052e-01
-7.76614010e-01 3.80646974e-01 4.85294908e-01 -5.73302925e-01
-8.64124179e-01 -9.96075630e-01 3.43720764e-01 1.70720172e+00
2.38385677e-01 -4.77577925e-01 -6.24879122e-01 -8.75472963e-01
1.12704905e-02 2.92391926e-01 -5.91746569e-01 4.65577431e-02
-8.15936983e-01 -3.60030204e-01 7.93210626e-01 6.59738541e-01
6.56566203e-01 -1.65057766e+00 -1.11344874e+00 9.41280648e-02
-5.64788520e-01 -1.31548047e+00 -1.67086676e-01 -4.44794670e-02
-1.67967588e-01 -1.59466147e+00 -3.64788353e-01 -7.39540517e-01
4.99554634e-01 2.23511606e-01 1.25527716e+00 -1.24595545e-01
-4.92446572e-01 1.07929933e+00 -6.49479866e-01 -4.79825020e-01
-1.73256055e-01 -2.25623503e-01 3.73795301e-01 9.19091254e-02
5.96900702e-01 -3.47498357e-01 -6.38846695e-01 5.73160827e-01
-4.55618203e-01 4.44715209e-02 5.05842030e-01 3.13209862e-01
1.85346991e-01 1.00412443e-01 2.54413724e-01 -5.36084533e-01
1.76534757e-01 -5.49232781e-01 -1.03483237e-01 9.36374366e-02
4.46674258e-01 -4.50616479e-01 4.98245433e-02 -2.76773065e-01
-1.02392614e+00 5.02621233e-01 2.43599549e-01 -2.04777330e-01
-8.46268892e-01 9.05819461e-02 -2.74270356e-01 2.83046842e-01
6.13816082e-01 -1.80591464e-01 -5.01855277e-02 -3.03291708e-01
2.55600661e-01 7.85075784e-01 8.72213125e-01 -5.46668470e-01
4.38301295e-01 7.24883378e-01 -9.20777395e-02 -9.52498972e-01
-8.24360073e-01 -8.87177169e-01 -1.18336582e+00 -9.66116548e-01
1.38586593e+00 -1.22711813e+00 -9.82406676e-01 8.06979239e-01
-7.34490693e-01 -9.17219162e-01 -1.80949226e-01 6.12857819e-01
-1.10511243e+00 3.24196875e-01 -5.51819026e-01 -1.01643634e+00
4.10575807e-01 -8.91936898e-01 1.54729855e+00 -2.29025364e-01
-7.00173199e-01 -7.37331390e-01 -4.43926919e-03 9.35932338e-01
-3.73027593e-01 7.92019069e-01 -5.80656491e-02 -7.02133298e-01
-5.28056920e-01 -2.83076167e-01 6.86557218e-02 9.86276790e-02
-6.96522370e-02 -3.31772178e-01 -6.03274167e-01 -3.74390244e-01
-3.62800390e-01 -8.63156915e-01 4.75706607e-01 1.77958012e-01
6.93720341e-01 2.87098251e-02 -4.55197811e-01 3.95139791e-02
7.18650222e-01 -5.62356599e-02 2.96146721e-01 3.83210301e-01
9.44273770e-01 9.40627456e-01 1.07424188e+00 6.10920131e-01
5.22929966e-01 9.08210039e-01 5.57765782e-01 2.64964074e-01
-6.02490902e-02 -1.87795371e-01 7.26888657e-01 1.28105029e-01
-4.03969347e-01 -3.64963174e-01 -1.27802861e+00 5.00393212e-01
-2.27958798e+00 -1.32708025e+00 -2.25642055e-01 1.82579935e+00
3.78043413e-01 1.16304390e-01 6.65534914e-01 3.60841304e-02
6.56720459e-01 2.11876333e-01 -5.19902945e-01 4.18088675e-01
2.30684936e-01 -6.14498615e-01 4.25891519e-01 3.61207068e-01
-1.67528248e+00 9.55829382e-01 5.77914095e+00 3.49125952e-01
-2.09507287e-01 -1.86501015e-02 1.61029309e-01 -3.42895687e-01
6.30697370e-01 -4.00608391e-01 -7.88816333e-01 3.02318484e-01
7.48988152e-01 3.47872466e-01 4.21555161e-01 8.25926602e-01
3.98653030e-01 -4.34600651e-01 -1.22512233e+00 1.03897977e+00
3.38787884e-01 -6.26406074e-01 -4.63247925e-01 2.38615498e-01
3.86287749e-01 2.18840390e-01 -4.80591714e-01 5.71077704e-01
8.16547632e-01 -8.86375010e-01 9.94590163e-01 5.21596670e-01
6.69608951e-01 -5.92837930e-01 5.84473312e-01 7.52144873e-01
-1.31571174e+00 -5.20330489e-01 1.30177632e-01 -5.43074369e-01
6.03941798e-01 -2.65008509e-01 -9.13392723e-01 2.48333991e-01
1.15537560e+00 1.20885444e+00 -7.51478314e-01 5.51999688e-01
-1.91233903e-01 1.58869192e-01 -3.08249325e-01 3.99269387e-02
3.66881341e-01 -1.20131955e-01 6.91473186e-01 1.05455458e+00
1.38369977e-01 3.28888923e-01 8.99412572e-01 3.17288712e-02
1.27098322e-01 -4.33549702e-01 -8.71673465e-01 -1.91860959e-01
2.27287501e-01 1.02752745e+00 -9.88008201e-01 -3.57829452e-01
-4.10214573e-01 1.23533750e+00 4.43032235e-01 1.41086236e-01
-9.87001777e-01 1.68308601e-01 8.71772230e-01 2.87086636e-01
5.66496067e-02 -4.44862425e-01 5.32316089e-01 -9.95661378e-01
8.72251093e-02 -9.84431267e-01 6.56518579e-01 -9.08461452e-01
-1.16928422e+00 2.80187339e-01 5.22750854e-01 -1.37442899e+00
-4.46769416e-01 -7.28983581e-01 3.25455256e-02 2.36291915e-01
-5.59446275e-01 -1.44964147e+00 -7.22002804e-01 8.10161710e-01
8.91277611e-01 -2.49199986e-01 6.84888184e-01 1.66915163e-01
-5.36760986e-01 -9.80498642e-02 -3.62714499e-01 4.21669304e-01
8.16337645e-01 -1.41026330e+00 2.07396790e-01 6.21126950e-01
9.47570428e-02 2.32996628e-01 8.27975690e-01 -9.16004062e-01
-1.02452958e+00 -1.19366038e+00 5.81848502e-01 -1.32449591e+00
7.48841405e-01 -6.46405697e-01 -4.81054693e-01 1.36412644e+00
-1.27707422e-01 1.48639590e-01 8.01119924e-01 -5.33870980e-02
1.26923144e-01 3.20319861e-01 -1.04965091e+00 6.76275671e-01
1.99092686e+00 -4.93657470e-01 -7.87776947e-01 7.52623796e-01
4.27117854e-01 -4.95493144e-01 -7.36724317e-01 3.71914774e-01
5.63110650e-01 -1.17518997e+00 1.11198211e+00 -6.92587435e-01
3.32023144e-01 -5.22392511e-01 -2.49259368e-01 -1.04319978e+00
-2.28270829e-01 -1.88717097e-01 -1.83045059e-01 9.58642066e-01
-2.36994654e-01 -1.55469030e-01 8.67405295e-01 6.70325816e-01
-1.99622110e-01 -2.97648013e-01 -8.56380761e-01 -6.83223069e-01
-7.25907803e-01 -5.55116773e-01 9.85470712e-02 6.94904447e-01
1.15374483e-01 2.57514358e-01 -5.41097820e-01 1.38668731e-01
6.68794870e-01 -4.54064965e-01 1.41431940e+00 -1.25596106e+00
-4.48643357e-01 -1.40420020e-01 -1.00180233e+00 -9.83478487e-01
3.91032010e-01 -5.04568398e-01 4.49827343e-01 -1.60196924e+00
2.24489659e-01 -1.30022570e-01 9.39703882e-02 6.12297595e-01
-4.27018628e-02 5.08132458e-01 1.70870796e-01 4.44532245e-01
-1.49271691e+00 4.88687783e-01 8.67552221e-01 1.43465502e-02
-2.09896602e-02 -3.60378549e-02 -3.77012454e-02 1.18173790e+00
6.21244490e-01 -3.20703417e-01 -4.24935818e-01 -2.87780404e-01
8.66565853e-02 -7.68368468e-02 7.87215948e-01 -1.45530558e+00
1.10083155e-01 -2.41755217e-01 2.90586323e-01 -4.53810751e-01
6.76072121e-01 -9.73451376e-01 3.01895797e-01 5.92592299e-01
-4.70833212e-01 -7.94491172e-02 -3.50947797e-01 1.06498253e+00
-2.49424148e-02 1.20586492e-01 3.74706417e-01 -6.41976595e-01
-1.38522649e+00 6.28013462e-02 -5.39297938e-01 -2.87805982e-02
1.60964596e+00 -2.80223638e-01 -1.72989711e-01 -4.68539864e-01
-1.05120790e+00 5.52890122e-01 5.12369215e-01 6.07304156e-01
2.56521106e-01 -1.28630793e+00 -6.92613900e-01 1.03783496e-02
4.56026793e-01 2.12006733e-01 3.63799363e-01 8.02092314e-01
-7.07403660e-01 9.95933712e-02 -4.68614578e-01 -6.79076850e-01
-1.68968499e+00 2.42053136e-01 9.38308239e-02 -1.57785848e-01
-8.76407146e-01 9.73701715e-01 3.41911726e-02 -7.54471123e-01
4.36954618e-01 8.43529403e-02 -3.75867128e-01 7.35173151e-02
5.58162689e-01 7.11372316e-01 -4.99460578e-01 -1.33328629e+00
-5.61953187e-01 1.83658347e-01 5.38677990e-01 -2.64981985e-01
1.41265535e+00 -3.31669420e-01 3.25605534e-02 7.10744858e-01
9.21015084e-01 -3.57677966e-01 -1.80157864e+00 -1.93138178e-02
1.75584659e-01 -6.37672603e-01 -6.48711205e-01 -6.36353016e-01
-4.26668316e-01 3.78152221e-01 3.72654110e-01 1.42216980e-01
7.27981985e-01 4.25766408e-01 5.61529994e-01 6.46079361e-01
8.85030508e-01 -1.28672934e+00 7.65466928e-01 5.01217186e-01
9.44527626e-01 -1.46410203e+00 1.08407885e-01 -4.48354542e-01
-1.15383685e+00 5.47500372e-01 9.14703250e-01 -4.10331227e-03
1.30213186e-01 -4.66975570e-02 6.29954413e-02 -6.33935928e-01
-5.64870775e-01 -4.58938271e-01 -1.44187346e-01 7.21377194e-01
-4.31159176e-02 2.09495172e-01 2.94973612e-01 2.28563920e-01
-2.97485352e-01 -1.72757898e-02 3.84316891e-01 1.36775005e+00
-5.33397973e-01 -6.52483940e-01 -5.00927866e-01 3.48765373e-01
-1.44119576e-01 6.65566087e-01 -6.85466051e-01 8.55806589e-01
4.49732989e-01 1.05098915e+00 2.04475030e-01 -3.88647705e-01
5.06197751e-01 1.23197086e-01 4.29105580e-01 -7.87124515e-01
-3.44042927e-01 -2.49004975e-01 4.44661975e-01 -9.41512287e-01
-9.47300136e-01 -1.24878359e+00 -1.38825321e+00 -1.28589422e-01
3.36606741e-01 -2.17821315e-01 1.29390791e-01 9.30246234e-01
3.45039479e-02 4.32081640e-01 1.24911331e-01 -1.54313815e+00
-1.42092764e-01 -1.12439585e+00 -5.89113355e-01 1.13104320e+00
1.21485777e-01 -1.06873965e+00 -3.09553146e-01 4.41706538e-01] | [8.01789379119873, 0.40511301159858704] |
06666952-2fdc-421b-b7f7-cc5882f0fb5e | secure-multiparty-computations-in-floating | 2001.03192 | null | https://arxiv.org/abs/2001.03192v1 | https://arxiv.org/pdf/2001.03192v1.pdf | Secure multiparty computations in floating-point arithmetic | Secure multiparty computations enable the distribution of so-called shares of sensitive data to multiple parties such that the multiple parties can effectively process the data while being unable to glean much information about the data (at least not without collusion among all parties to put back together all the shares). Thus, the parties may conspire to send all their processed results to a trusted third party (perhaps the data provider) at the conclusion of the computations, with only the trusted third party being able to view the final results. Secure multiparty computations for privacy-preserving machine-learning turn out to be possible using solely standard floating-point arithmetic, at least with a carefully controlled leakage of information less than the loss of accuracy due to roundoff, all backed by rigorous mathematical proofs of worst-case bounds on information loss and numerical stability in finite-precision arithmetic. Numerical examples illustrate the high performance attained on commodity off-the-shelf hardware for generalized linear models, including ordinary linear least-squares regression, binary and multinomial logistic regression, probit regression, and Poisson regression. | ['Mark Tygert', 'Chuan Guo', 'Awni Hannun', 'Brian Knott', 'Ruiyu Zhu', 'Laurens van der Maaten'] | 2020-01-09 | null | null | null | null | ['mathematical-proofs'] | ['miscellaneous'] | [ 2.05886945e-01 1.75596491e-01 5.36690801e-02 -6.03364706e-01
-1.49913168e+00 -1.26966238e+00 1.44419178e-01 7.84014881e-01
-8.59787107e-01 7.47982442e-01 -2.18442932e-01 -5.58093846e-01
6.41048774e-02 -1.11653113e+00 -7.41455197e-01 -1.22336924e+00
-3.49588901e-01 4.80581045e-01 -2.22658291e-01 7.11751804e-02
1.42514586e-01 4.84560221e-01 -9.73641396e-01 2.98515826e-01
1.23145707e-01 1.16110456e+00 -8.88212264e-01 6.79446757e-01
3.22584093e-01 7.03320980e-01 -3.02530378e-01 -9.51613069e-01
6.52738869e-01 2.03475147e-01 -5.03087401e-01 -5.19767761e-01
-6.55383691e-02 -6.11189604e-01 -9.33533311e-02 1.16195965e+00
1.73836634e-01 -2.59730339e-01 3.87322903e-01 -1.44550729e+00
-2.73342788e-01 8.01348627e-01 -4.93219346e-01 -4.74076122e-01
2.05782756e-01 2.66829412e-02 8.23734224e-01 -5.04241884e-01
4.65562820e-01 8.13186646e-01 7.15657890e-01 1.06955975e-01
-1.32150352e+00 -7.61563182e-01 -5.01773000e-01 -3.17387879e-01
-1.51433980e+00 -5.97510576e-01 3.08827311e-01 -9.91037115e-02
4.94578391e-01 8.86415243e-01 1.62251234e-01 2.49100208e-01
6.38698399e-01 1.94394022e-01 1.47112906e+00 -3.27055275e-01
5.38677812e-01 7.02699840e-01 3.15060884e-01 6.11200213e-01
7.53044724e-01 1.12885825e-01 -5.16675413e-01 -1.10386860e+00
1.01681210e-01 2.59355456e-01 -4.20719713e-01 -3.41499984e-01
-1.09307098e+00 9.42186117e-01 1.57831833e-01 -9.10714492e-02
-4.65143591e-01 3.42633098e-01 5.57897747e-01 7.09780931e-01
3.83246660e-01 -2.07404405e-01 -7.59250641e-01 4.64808226e-01
-5.85643947e-01 5.24442077e-01 1.23537028e+00 8.80236268e-01
8.33540559e-01 -5.96315384e-01 -8.47513303e-02 -2.53577560e-01
3.86891961e-01 7.35737264e-01 -7.31359422e-02 -7.77846456e-01
7.73887396e-01 2.36711666e-01 5.42463481e-01 -1.03307760e+00
-5.26038669e-02 2.17230931e-01 -1.12133181e+00 5.12605071e-01
7.67757833e-01 -3.91791344e-01 -1.72230899e-01 1.52387524e+00
5.41341603e-01 -6.83607519e-01 4.99024928e-01 9.18690860e-01
4.59380597e-02 8.92090380e-01 1.12545401e-01 -5.42555213e-01
1.55336571e+00 -2.01793805e-01 -3.79803985e-01 1.52950585e-01
5.73906720e-01 -6.14735961e-01 2.37329081e-01 5.09098887e-01
-1.19447839e+00 2.73133427e-01 -8.15472841e-01 -4.68892157e-01
-2.72074640e-01 -5.22577763e-01 9.24506426e-01 9.41539586e-01
-7.05971003e-01 5.34885228e-01 -8.85019779e-01 7.17392862e-01
4.25947338e-01 7.71095157e-01 -9.99767900e-01 -1.16849907e-01
-1.09389770e+00 4.45474267e-01 -4.66065742e-02 2.01000899e-01
-3.39851767e-01 -1.00831819e+00 -6.70289338e-01 2.61596173e-01
5.94963729e-02 -5.87342978e-01 1.11687851e+00 -6.96605563e-01
-1.01544023e+00 1.14768445e+00 -2.02115759e-01 -5.63760459e-01
7.80319273e-01 3.99487764e-01 2.16946632e-01 2.32037567e-02
-1.18547104e-01 -5.89619949e-02 4.47162718e-01 -9.18515801e-01
-6.32657111e-01 -1.18659568e+00 5.53225949e-02 -5.72122410e-02
-9.17799845e-02 2.46567547e-01 9.55911875e-02 -1.85258776e-01
4.19561952e-01 -1.07354021e+00 -5.70704162e-01 4.98322248e-01
-2.81951696e-01 3.78211737e-01 4.67478096e-01 -7.31540680e-01
5.23626566e-01 -2.14545178e+00 -1.75568998e-01 6.22470319e-01
1.43905938e-01 -1.18491195e-01 4.77610916e-01 4.49997455e-01
1.57968059e-01 7.66626149e-02 -4.69259441e-01 -5.35734951e-01
1.42621785e-01 -1.47789568e-01 -6.26544416e-01 1.24185109e+00
-6.05699480e-01 6.77245796e-01 -5.53567648e-01 -1.59805089e-01
-1.91618010e-01 4.52552825e-01 -3.97952169e-01 -1.95647448e-01
5.64732738e-02 2.42960408e-01 -6.06224597e-01 5.11770308e-01
1.16415846e+00 -1.46346867e-01 2.76222497e-01 7.47756436e-02
-3.93956620e-03 1.67707950e-01 -1.46912432e+00 1.10531318e+00
-3.72745782e-01 1.97312772e-01 7.69352674e-01 -3.74049783e-01
3.34960371e-01 6.64282322e-01 3.69654566e-01 -1.72532290e-01
4.30690229e-01 3.31652105e-01 -5.79664528e-01 2.04711288e-01
5.29237866e-01 -6.54529631e-01 -6.56386554e-01 9.46228027e-01
-5.36920667e-01 3.44113051e-03 -7.20545173e-01 9.05360207e-02
8.38476062e-01 -6.48488879e-01 3.12357843e-01 -1.26431793e-01
4.23346490e-01 1.06788918e-01 4.97939020e-01 5.38400114e-01
-1.32552147e-01 2.78832942e-01 8.24950039e-01 -5.08113027e-01
-9.67281938e-01 -8.74478817e-01 -4.41888832e-02 1.01295888e+00
1.74718678e-01 2.26762146e-02 -5.05406559e-01 -4.15985882e-01
5.75367689e-01 4.07277882e-01 -3.53900194e-01 1.02862626e-01
-8.29943493e-02 -6.68747425e-01 5.03856063e-01 6.71201795e-02
4.16837633e-01 -5.11180043e-01 -6.19009852e-01 4.52715866e-02
4.11887355e-02 -8.39021504e-01 -4.38635439e-01 4.65978622e-01
-8.70468676e-01 -8.78517807e-01 -3.06160510e-01 -1.85382754e-01
8.26362133e-01 3.22494328e-01 2.38060325e-01 3.65955569e-02
1.98891051e-02 -4.24268693e-02 3.14832181e-01 -4.62518930e-01
-6.07880831e-01 -4.37823504e-01 -3.51145444e-03 2.19283834e-01
3.68365288e-01 -4.21271920e-01 -6.05171084e-01 -1.95104890e-02
-9.13676977e-01 -3.66924316e-01 1.92251608e-01 3.02393675e-01
7.24828184e-01 2.69443572e-01 -1.00436524e-01 -1.08583200e+00
3.05870086e-01 -4.56650376e-01 -1.26767457e+00 2.67860830e-01
-4.70082432e-01 1.80071279e-01 7.70380676e-01 -2.04186663e-01
-7.62403369e-01 2.55719185e-01 3.48844320e-01 -1.69560730e-01
2.58429140e-01 1.35130763e-01 -2.79899031e-01 -4.58677799e-01
3.88028413e-01 2.17228100e-01 9.06968191e-02 -3.10543090e-01
4.67801362e-01 7.95821488e-01 5.52646041e-01 -3.93195480e-01
8.08332145e-01 9.20086563e-01 6.39443517e-01 -2.00226709e-01
-2.49526724e-01 -7.75053650e-02 -4.49642122e-01 5.04828990e-01
6.76883638e-01 -1.06671894e+00 -1.59631598e+00 7.84508228e-01
-1.18248880e+00 1.44278631e-01 -1.86230466e-01 3.47857863e-01
-2.77473152e-01 3.57333779e-01 -7.39976287e-01 -1.08618784e+00
-6.36284888e-01 -1.08997750e+00 8.49835157e-01 -9.69358012e-02
-2.78882645e-02 -7.74901986e-01 -2.57174075e-01 5.38903415e-01
3.95686299e-01 5.73179543e-01 8.03682089e-01 -1.00163805e+00
-9.97478843e-01 -9.28440511e-01 -1.54692262e-01 2.70597577e-01
-1.67164326e-01 -1.56538680e-01 -1.12188375e+00 -4.26347166e-01
5.93811810e-01 -1.50063992e-01 4.95076269e-01 1.19657911e-01
9.42123294e-01 -1.00214612e+00 -2.76255429e-01 6.82574689e-01
1.46058130e+00 -3.93597722e-01 4.92626041e-01 -1.96469039e-01
4.01440591e-01 5.30393243e-01 3.69036794e-01 7.47242510e-01
5.06004989e-01 2.46302009e-01 2.75495529e-01 -2.07010414e-02
9.14847076e-01 1.08676180e-02 1.75878689e-01 2.08771065e-01
2.43858024e-01 2.03802474e-02 -4.03832287e-01 2.21311614e-01
-1.43793106e+00 -6.52458429e-01 -5.15265703e-01 2.80356646e+00
1.12672496e+00 -1.11065805e-01 -2.80543596e-01 1.47285715e-01
4.44197834e-01 3.61434743e-02 -2.34217495e-01 -7.38105476e-01
7.46990889e-02 4.50107791e-02 1.48770535e+00 7.39196241e-01
-1.03150129e+00 4.45196658e-01 5.18544579e+00 5.02742827e-01
-9.97762203e-01 3.05532604e-01 1.08756137e+00 -2.69535601e-01
-5.89639783e-01 3.75694722e-01 -6.26936257e-01 5.41809440e-01
1.05202317e+00 -4.28381354e-01 4.79255497e-01 8.83812428e-01
-1.38397709e-01 -4.71445441e-01 -1.30050123e+00 5.61992288e-01
-4.80616033e-01 -1.27727675e+00 -6.37288451e-01 4.18488801e-01
6.61265612e-01 -3.00106764e-01 6.76762089e-02 -9.41314697e-02
4.52553809e-01 -7.49392450e-01 6.30080044e-01 5.06354213e-01
8.14486146e-01 -1.01654100e+00 7.55761504e-01 5.13231218e-01
-6.70167148e-01 -7.53044710e-02 -4.73852843e-01 -1.45563036e-01
-9.27315056e-02 7.28616118e-01 -5.68051398e-01 4.96920437e-01
3.46932709e-01 -6.90946877e-01 2.37721384e-01 4.58772510e-01
-1.05450995e-01 2.17780113e-01 -7.09770858e-01 7.97600076e-02
1.05645142e-01 -3.31974715e-01 2.03467086e-01 8.40088189e-01
1.30726650e-01 7.75364876e-01 -1.42313853e-01 6.62410676e-01
-3.95600408e-01 9.72731486e-02 -4.16770279e-01 2.99689442e-01
5.41197658e-01 1.24735439e+00 -4.45145071e-01 -1.82376951e-01
-3.90303403e-01 9.50930953e-01 -3.22479643e-02 8.94526839e-02
-4.34977084e-01 -2.41661206e-01 8.32203746e-01 4.55343798e-02
1.95977271e-01 -2.87386388e-01 -8.61972213e-01 -9.81944561e-01
4.03443694e-01 -7.58953273e-01 6.01598799e-01 -2.43166119e-01
-1.28404462e+00 7.66109526e-02 -2.10140169e-01 -7.48563588e-01
-2.54540443e-01 -2.95133114e-01 -3.78649652e-01 1.45967579e+00
-1.39974320e+00 -9.92784500e-01 3.46888453e-01 7.48690963e-01
-7.13618636e-01 6.34750351e-03 1.15865409e+00 -7.43826404e-02
6.66490495e-02 7.40744412e-01 3.01834553e-01 7.39779472e-02
6.12602353e-01 -8.04780185e-01 2.94244528e-01 7.26623833e-01
-1.49645373e-01 9.56627071e-01 6.72394335e-01 -4.84364867e-01
-2.12669086e+00 -8.36319685e-01 1.34713089e+00 -3.81708801e-01
6.31300807e-01 -6.21807039e-01 -9.44624543e-01 8.29162896e-01
-3.07322413e-01 5.24697423e-01 9.48743641e-01 -1.80183277e-01
-6.58910096e-01 -5.44627368e-01 -1.89106786e+00 2.07088605e-01
-1.03450872e-01 -7.78476775e-01 -1.07238730e-02 4.56099242e-01
6.52579010e-01 -4.98894960e-01 -7.92686641e-01 -6.76372051e-02
8.27721238e-01 -8.10539901e-01 8.13556850e-01 -5.24781704e-01
1.02731854e-01 -2.83548981e-01 -5.64632416e-01 -4.27821785e-01
2.57565856e-01 -1.09020579e+00 2.22909555e-01 1.12785113e+00
4.19592112e-01 -9.57491279e-01 8.34892869e-01 1.89772677e+00
8.15397263e-01 -3.63144845e-01 -1.48419416e+00 -1.73656374e-01
3.57127458e-01 -5.08996069e-01 8.68618548e-01 8.74642372e-01
9.70250145e-02 -4.28490072e-01 -3.64391208e-01 8.68641675e-01
9.36142623e-01 4.42154557e-01 9.78084862e-01 -8.64959836e-01
-2.95433372e-01 2.79210538e-01 -3.01325709e-01 -5.70114613e-01
5.38403206e-02 -8.55393767e-01 -2.08285704e-01 -8.23235393e-01
3.63853574e-01 -9.94941831e-01 -1.98286965e-01 8.21829855e-01
-1.09595269e-01 3.04177046e-01 2.38623872e-01 1.77377686e-01
-3.91846836e-01 -1.77174304e-02 8.12165737e-01 9.02238488e-02
-4.92574722e-02 6.51478171e-01 -1.11128640e+00 5.83704531e-01
4.61642355e-01 -8.70715022e-01 4.23420221e-02 -1.58335179e-01
4.23296154e-01 9.37162340e-01 7.35237777e-01 -5.34704208e-01
7.73689628e-01 -2.39112183e-01 2.44655132e-01 -4.35800910e-01
2.81216800e-01 -1.29460728e+00 5.29182851e-01 5.13511717e-01
-4.89883542e-01 -1.06711321e-01 -1.43598080e-01 5.13365626e-01
1.75406322e-01 -1.98818624e-01 7.87348330e-01 -8.74470081e-03
1.75184801e-01 3.45764399e-01 -2.13950440e-01 -5.01895308e-01
1.36063194e+00 3.76287937e-01 -3.66530180e-01 -6.92371547e-01
-5.51791608e-01 2.52676368e-01 5.75702667e-01 -2.66120434e-01
2.34909162e-01 -7.75530815e-01 -7.74789214e-01 4.69852716e-01
-3.32456738e-01 -3.12809348e-02 3.39126021e-01 6.86240792e-01
-4.11325544e-01 4.95764256e-01 2.26341173e-01 -8.05188045e-02
-1.51982880e+00 7.72943258e-01 8.57869759e-02 -4.46964443e-01
-1.97727904e-01 7.45570421e-01 -1.68603152e-01 -3.93290579e-01
6.01592101e-02 -4.01206523e-01 9.46697772e-01 1.29544780e-01
7.89132178e-01 4.14390534e-01 3.71800035e-01 -3.82667005e-01
-4.23302323e-01 8.41427222e-02 7.46727549e-03 -3.54181528e-01
1.15761304e+00 -1.00775823e-01 -7.40280449e-01 4.33986448e-02
1.73040378e+00 5.72004139e-01 -9.51431811e-01 -3.16678792e-01
-3.87794375e-01 -5.98266959e-01 -6.19674250e-02 -5.94518065e-01
-9.70842361e-01 7.51240611e-01 2.42470473e-01 9.10456777e-02
1.07271755e+00 -2.68115640e-01 8.50715756e-01 5.50386488e-01
1.04185700e+00 -5.57209253e-01 -1.14285505e+00 -9.39151198e-02
4.22946066e-01 -1.31925476e+00 5.34068942e-01 -3.67422849e-01
-7.72195816e-01 9.58354890e-01 -4.60864872e-01 2.57966965e-02
7.87290871e-01 6.21168137e-01 -1.00604996e-01 2.63156176e-01
-8.42256188e-01 8.23735714e-01 -1.49491951e-01 2.61506975e-01
-1.76137492e-01 2.44000390e-01 -2.16568992e-01 9.45567071e-01
-1.95862591e-01 1.79971417e-03 4.85540539e-01 8.93223763e-01
-2.09482953e-01 -1.34902036e+00 -7.41587698e-01 1.09117180e-01
-1.07667482e+00 -1.17559582e-01 7.36248679e-04 6.74652219e-01
-1.68264925e-01 8.74790370e-01 8.73302370e-02 3.13048095e-01
9.85150412e-03 1.81857839e-01 1.45317286e-01 -1.28501698e-01
-1.02736127e+00 -1.86224252e-01 -4.74789590e-02 -5.26985288e-01
1.51716098e-01 -7.83558071e-01 -1.44821346e+00 -9.86469448e-01
-1.18803769e-01 2.15643451e-01 1.24370170e+00 6.92745209e-01
3.45416903e-01 -5.31751096e-01 1.22397554e+00 -4.51241702e-01
-1.18542969e+00 -2.48420775e-01 -7.75273919e-01 7.73938969e-02
6.69647872e-01 5.74392498e-01 -6.20214283e-01 3.57938483e-02] | [5.889193534851074, 6.680034637451172] |
cc35a25e-c03e-4629-bb7f-bcdf10753e74 | multilingual-multimodal-pre-training-for-zero | 2103.08849 | null | https://arxiv.org/abs/2103.08849v3 | https://arxiv.org/pdf/2103.08849v3.pdf | Multilingual Multimodal Pre-training for Zero-Shot Cross-Lingual Transfer of Vision-Language Models | This paper studies zero-shot cross-lingual transfer of vision-language models. Specifically, we focus on multilingual text-to-video search and propose a Transformer-based model that learns contextualized multilingual multimodal embeddings. Under a zero-shot setting, we empirically demonstrate that performance degrades significantly when we query the multilingual text-video model with non-English sentences. To address this problem, we introduce a multilingual multimodal pre-training strategy, and collect a new multilingual instructional video dataset (MultiHowTo100M) for pre-training. Experiments on VTT show that our method significantly improves video search in non-English languages without additional annotations. Furthermore, when multilingual annotations are available, our method outperforms recent baselines by a large margin in multilingual text-to-video search on VTT and VATEX; as well as in multilingual text-to-image search on Multi30K. Our model and Multi-HowTo100M is available at http://github.com/berniebear/Multi-HT100M. | ['Alexander Hauptmann', 'Florian Metze', 'Graham Neubig', 'Junjie Hu', 'Mandela Patrick', 'Po-Yao Huang'] | 2021-03-16 | null | https://aclanthology.org/2021.naacl-main.195 | https://aclanthology.org/2021.naacl-main.195.pdf | naacl-2021-4 | ['text-to-video-search'] | ['natural-language-processing'] | [-3.30624044e-01 -5.57303667e-01 -6.16482019e-01 -1.58127487e-01
-1.82632542e+00 -9.23590541e-01 7.32822061e-01 -6.16188757e-02
-9.84005272e-01 4.97268170e-01 3.65033537e-01 -5.69712698e-01
2.39861190e-01 -1.00939006e-01 -1.15807998e+00 -2.74335861e-01
3.62205744e-01 5.61947227e-01 2.24902958e-01 -1.35118231e-01
-4.34246525e-04 -2.85232335e-01 -1.34298062e+00 5.70996583e-01
9.04155314e-01 5.98741472e-01 7.18946040e-01 9.86696124e-01
7.37096295e-02 8.13282847e-01 -3.88345234e-02 -6.20585442e-01
-4.50797826e-02 -1.67541876e-01 -9.51703668e-01 -1.60697952e-01
1.27507114e+00 -5.56539714e-01 -5.34778118e-01 1.00240397e+00
8.41488779e-01 1.85607433e-01 4.58710521e-01 -1.15715659e+00
-8.34469318e-01 6.17549002e-01 -4.30013657e-01 2.06871465e-01
7.30727315e-01 2.15486765e-01 9.77890432e-01 -1.49195242e+00
1.02618718e+00 1.44756222e+00 3.77151698e-01 3.86937618e-01
-7.98536837e-01 -6.88587725e-01 2.01730996e-01 6.16554141e-01
-1.36514294e+00 -5.84551573e-01 1.71260610e-01 -4.29944992e-01
8.75730097e-01 1.21223629e-01 5.12564778e-01 1.47185028e+00
1.94723532e-01 1.36876738e+00 9.85226870e-01 -6.95083439e-01
-4.05794322e-01 1.39902860e-01 -6.94985241e-02 1.14567637e+00
-3.00490290e-01 -1.69678569e-01 -8.94406557e-01 1.62601173e-01
4.38775569e-01 -3.72660041e-01 -5.45168281e-01 -5.41651189e-01
-1.55982947e+00 7.89487720e-01 9.81360953e-03 3.97676766e-01
1.54590324e-01 2.25225866e-01 7.62234509e-01 6.77895963e-01
3.63864273e-01 1.65646344e-01 -4.30826545e-01 -4.93389368e-01
-7.95128167e-01 -7.96379298e-02 4.20118660e-01 1.32118726e+00
6.64637625e-01 -8.62125009e-02 -2.39733979e-01 1.13208246e+00
2.68476546e-01 9.73097920e-01 7.03413069e-01 -1.10564685e+00
8.30840230e-01 -5.57660824e-03 -1.54508710e-01 -3.94145578e-01
1.92639038e-01 1.71499595e-01 -6.35146350e-02 -3.23080212e-01
3.76259744e-01 -4.26259004e-02 -9.92250204e-01 1.57050467e+00
2.02459931e-01 8.82976726e-02 1.62499651e-01 9.95178640e-01
1.20124876e+00 6.41725659e-01 1.87044486e-01 4.65281457e-02
1.53460336e+00 -1.50345170e+00 -8.59260917e-01 -1.67689711e-01
9.95948136e-01 -1.38532710e+00 1.43948603e+00 6.82137385e-02
-1.19569504e+00 -4.83681947e-01 -6.47911668e-01 -5.11052549e-01
-5.64356685e-01 3.10336292e-01 7.86467195e-02 2.21451566e-01
-1.39896929e+00 -2.77109474e-01 -7.47307003e-01 -8.54827344e-01
-5.73180653e-02 -6.25930727e-02 -6.10478938e-01 -8.26346695e-01
-1.27191496e+00 9.90740001e-01 2.48028502e-01 -3.36356312e-01
-1.38519895e+00 -6.63207471e-01 -1.28872609e+00 -2.93544024e-01
7.31069565e-01 -4.35290724e-01 1.53960323e+00 -5.69167078e-01
-1.13830495e+00 1.05714929e+00 -3.26348454e-01 -1.66055724e-01
5.64837515e-01 -3.62963885e-01 -3.10758680e-01 5.78294039e-01
4.06083107e-01 1.17634165e+00 6.22519314e-01 -9.61232901e-01
-6.64852500e-01 2.59146336e-02 3.46584827e-01 6.85399532e-01
-6.40302122e-01 2.49085471e-01 -1.39627421e+00 -5.70652902e-01
-4.60080147e-01 -1.25396693e+00 1.88837737e-01 -2.36754313e-01
-1.34002060e-01 -3.19030344e-01 7.70234644e-01 -7.41699815e-01
1.09276044e+00 -1.99311805e+00 2.02021420e-01 -4.49234903e-01
-8.94503668e-02 1.22370869e-01 -5.76710045e-01 5.14247835e-01
8.02741796e-02 2.33054371e-03 3.57116014e-01 -3.84085298e-01
2.27981701e-01 7.05466494e-02 5.08009270e-02 2.29470402e-01
-2.56517500e-01 1.20682466e+00 -1.04045212e+00 -1.06427813e+00
3.70932132e-01 5.63661397e-01 -5.50602376e-01 1.05580678e-02
-1.37460142e-01 3.56480986e-01 -3.94830823e-01 1.01212418e+00
1.53022662e-01 -2.33346790e-01 2.63263583e-01 -2.81483889e-01
-2.28585988e-01 6.17187377e-03 -4.73107398e-01 2.42587066e+00
-6.42576277e-01 1.07464635e+00 1.85656339e-01 -6.90059841e-01
-4.42482866e-02 5.51545203e-01 4.26519424e-01 -1.02103627e+00
3.03028636e-02 3.87501508e-01 -4.76988077e-01 -7.38121867e-01
8.53786707e-01 4.30876017e-01 -4.22579288e-01 4.46214974e-01
4.99562353e-01 -5.89232296e-02 5.65859675e-01 6.93526387e-01
7.55045831e-01 2.31525555e-01 -1.73469692e-01 -2.27888286e-01
4.63815957e-01 1.18124843e-01 -2.71962974e-02 7.69969761e-01
-2.97466308e-01 3.01601827e-01 4.16368954e-02 -7.64691979e-02
-8.85833859e-01 -9.32223380e-01 -6.52870387e-02 1.75080919e+00
6.80623204e-02 -6.26475036e-01 -7.94189811e-01 -7.45963097e-01
-7.54418820e-02 5.60080290e-01 -2.39051133e-01 4.99053299e-02
-4.87151742e-01 -1.80315495e-01 5.12364984e-01 2.51437068e-01
3.04834872e-01 -7.87800789e-01 -3.44393253e-01 -1.68191507e-01
-7.14635313e-01 -1.81574368e+00 -1.30550385e+00 1.13002516e-01
-3.77621949e-01 -9.44675565e-01 -1.11791563e+00 -1.33481157e+00
3.91695052e-01 5.50252676e-01 1.23260474e+00 -2.30759099e-01
-3.55949193e-01 1.39709854e+00 -4.07532781e-01 3.75425033e-02
-2.36400649e-01 1.18434541e-01 5.89343980e-02 -4.79474723e-01
4.18406010e-01 1.47573441e-01 -4.12305146e-01 3.29295367e-01
-7.63168454e-01 -8.30033571e-02 4.14086610e-01 8.46246183e-01
6.71506524e-01 -6.80921674e-01 1.11461259e-01 -2.47695342e-01
6.72675490e-01 -3.58991057e-01 -6.43615782e-01 7.40817130e-01
-5.53792298e-01 -7.68772662e-02 6.48284107e-02 -6.65071726e-01
-8.05654526e-01 -1.18334576e-01 1.40203401e-01 -1.24873769e+00
5.98680191e-02 7.95031369e-01 2.26264447e-02 -2.49739230e-01
1.26281247e-01 4.71554846e-01 -3.38647008e-01 -3.31610084e-01
6.72105312e-01 5.28573215e-01 5.75264812e-01 -8.39280486e-01
5.05580842e-01 3.12589332e-02 -5.92387438e-01 -8.69158685e-01
-4.09812331e-01 -6.65116251e-01 -5.53263307e-01 -6.00760281e-01
1.32917130e+00 -1.63138068e+00 -6.56828701e-01 2.89043486e-01
-9.07822847e-01 -7.12163985e-01 2.11174667e-01 9.00025725e-01
-5.99320710e-01 3.93174887e-01 -1.07672048e+00 -4.16157842e-01
-1.01781823e-01 -1.78151393e+00 1.25742555e+00 -1.28471702e-01
1.19589023e-01 -1.22650611e+00 7.98613429e-02 9.00813699e-01
1.78785175e-02 -5.08228481e-01 8.45566571e-01 -5.80636919e-01
-8.38653564e-01 -1.35607406e-01 -7.79987648e-02 3.97230163e-02
-2.84083456e-01 -2.66344756e-01 -6.38158858e-01 -7.48374283e-01
-6.26864195e-01 -1.16582394e+00 8.58536780e-01 3.63132089e-01
8.18061292e-01 2.31281947e-02 -1.69867426e-01 5.11918902e-01
1.30206239e+00 1.16994649e-01 1.37148187e-01 5.17074227e-01
8.26468945e-01 3.07995141e-01 9.30994689e-01 -5.69510795e-02
1.05836201e+00 9.06757772e-01 1.09097183e-01 7.49045461e-02
-4.35101449e-01 -5.10497570e-01 7.85732627e-01 1.64063227e+00
1.78300396e-01 -4.11091566e-01 -1.00848055e+00 9.56269562e-01
-1.77417552e+00 -9.01295364e-01 3.09408963e-01 2.08271885e+00
1.04443705e+00 -3.40671241e-01 -7.42214248e-02 -8.61063838e-01
4.73366350e-01 1.78401038e-01 -3.64870310e-01 -2.63871491e-01
-2.11934224e-01 -1.46279797e-01 5.72868109e-01 9.10618424e-01
-1.12405694e+00 1.47395182e+00 5.94331980e+00 1.17927015e+00
-1.18075967e+00 7.11034119e-01 3.24380994e-01 -4.56688583e-01
-4.54916686e-01 -2.80588388e-01 -9.20544684e-01 3.29109818e-01
8.89188230e-01 -3.41125429e-01 3.78767610e-01 5.97389460e-01
3.68120521e-02 -5.07644191e-02 -1.04281592e+00 1.30958700e+00
6.25273645e-01 -1.07477427e+00 1.98847055e-01 1.09016158e-01
9.33330119e-01 5.13124883e-01 4.01702732e-01 8.07403862e-01
5.79281032e-01 -7.75158525e-01 8.52014899e-01 1.27466381e-01
1.10525012e+00 -5.41975796e-01 2.55330294e-01 1.10110901e-01
-1.48769879e+00 1.58633828e-01 -1.16713747e-01 7.02651918e-01
1.20560043e-01 -2.31717870e-01 -7.00384915e-01 5.39075792e-01
9.67272997e-01 8.87349486e-01 -6.59857571e-01 7.71763265e-01
-1.16290703e-01 2.46559322e-01 -1.27066076e-01 1.23713531e-01
6.90046132e-01 -9.13856924e-02 5.03250480e-01 1.40922415e+00
5.18661857e-01 -3.39810461e-01 7.37799585e-01 1.86035156e-01
-4.19374913e-01 3.92317772e-01 -9.11135972e-01 -1.74339533e-01
4.78959471e-01 1.01844168e+00 -2.52746642e-01 -6.29154921e-01
-7.63920009e-01 1.29583395e+00 3.86391282e-01 8.35820079e-01
-9.25735891e-01 1.39724063e-02 4.25904006e-01 -1.75594583e-01
4.45101738e-01 -2.60120392e-01 7.62916088e-01 -1.68259156e+00
1.61289796e-02 -1.34882379e+00 5.15522778e-01 -1.16379321e+00
-9.43033814e-01 5.69526494e-01 1.98935360e-01 -1.28409612e+00
-5.59828520e-01 -5.36136806e-01 -2.63645705e-02 5.18248022e-01
-1.56427717e+00 -1.48602438e+00 3.16262618e-02 1.03141797e+00
1.21845293e+00 -3.75566512e-01 7.29588032e-01 8.39112997e-01
-3.46713781e-01 9.97670889e-01 2.23479986e-01 3.08405757e-01
1.47925055e+00 -9.52383161e-01 -1.19041614e-01 8.10883701e-01
4.51813191e-01 5.29176235e-01 4.29209977e-01 -7.03406692e-01
-1.86245108e+00 -8.55789125e-01 9.93985415e-01 -7.18079329e-01
1.12472868e+00 -2.95460165e-01 -4.66159731e-01 1.14277899e+00
1.19007671e+00 -1.49275869e-01 5.71717083e-01 3.69557142e-02
-4.71928447e-01 2.21330196e-01 -4.92523581e-01 8.57647777e-01
8.31157446e-01 -1.10139060e+00 -3.60614419e-01 6.74835801e-01
9.78747487e-01 -6.18811011e-01 -1.09645462e+00 2.33361512e-01
7.56217837e-01 -4.16166693e-01 1.17457247e+00 -4.97548103e-01
5.53604543e-01 6.07741103e-02 -5.21548152e-01 -1.29232943e+00
1.09796137e-01 -4.58391845e-01 2.17629597e-01 7.63491094e-01
6.00437641e-01 -3.29270720e-01 2.88892180e-01 1.38231948e-01
-3.84584785e-01 -5.08346856e-01 -1.24734497e+00 -8.62239122e-01
3.89918536e-01 -5.10704219e-01 -1.21440955e-01 1.34945011e+00
1.30284265e-01 4.42812949e-01 -5.85249960e-01 -6.19814619e-02
7.08177686e-01 -5.51674441e-02 7.02291548e-01 -4.06476408e-01
2.75812042e-03 -3.94925207e-01 -4.82506454e-02 -1.35372758e+00
5.71778893e-01 -1.09789813e+00 7.81864151e-02 -1.35982323e+00
5.59170008e-01 1.50719687e-01 -3.24600518e-01 5.99441886e-01
-1.33307561e-01 3.61231118e-01 4.10318494e-01 2.16691166e-01
-1.32327235e+00 5.30692220e-01 1.45612884e+00 -4.75908548e-01
3.49559546e-01 -8.54612887e-01 -8.64963457e-02 4.91325557e-01
1.92217648e-01 -2.00702757e-01 -4.93542612e-01 -1.17593038e+00
-1.32593572e-01 3.45870286e-01 2.37315834e-01 -7.03491747e-01
4.63964969e-01 3.38787921e-02 4.90063801e-02 -7.83409834e-01
7.37958908e-01 -6.94913268e-01 -3.73738587e-01 2.48007864e-01
-5.94750285e-01 6.18625045e-01 3.96669954e-01 3.61948490e-01
-5.08477747e-01 -1.01455972e-01 2.38433152e-01 -2.43453696e-01
-1.07543075e+00 3.37568879e-01 -6.21261775e-01 3.22033435e-01
9.63068128e-01 1.37305588e-01 -6.41465008e-01 -7.41472900e-01
-7.82687306e-01 8.85144413e-01 6.28831446e-01 8.05318773e-01
5.50876677e-01 -1.66546893e+00 -6.01088047e-01 -1.54424921e-01
5.58926046e-01 -6.80865526e-01 3.80704850e-01 1.21245646e+00
-3.71508628e-01 1.15834141e+00 -4.38659592e-03 -8.23375881e-01
-1.72954595e+00 4.22908843e-01 1.15730971e-01 -1.71273485e-01
-2.22669661e-01 8.98338437e-01 2.51779109e-01 -6.65137410e-01
4.81880486e-01 -2.17372164e-01 -4.34561856e-02 3.38969916e-01
3.09878558e-01 -5.21016829e-02 -1.61257625e-01 -9.62956131e-01
-3.46203804e-01 9.00924325e-01 -2.56698579e-01 -5.58500946e-01
8.26958776e-01 -4.35724437e-01 2.09590778e-01 6.79880381e-01
1.56196010e+00 1.17169984e-01 -7.74963319e-01 -3.78134817e-01
-1.94711193e-01 -4.92287070e-01 2.05137223e-01 -7.20953763e-01
-1.00685441e+00 9.75344837e-01 8.38744998e-01 -5.10001123e-01
8.11545849e-01 2.85716325e-01 8.90841126e-01 6.53883040e-01
4.66693461e-01 -1.28730619e+00 3.59453171e-01 8.21270049e-01
6.94949329e-01 -1.76207197e+00 -1.41516417e-01 3.71819027e-02
-8.48372817e-01 8.58685553e-01 7.20419705e-01 6.10334575e-01
4.21788841e-01 -4.52806763e-02 4.18491960e-01 -8.46885368e-02
-1.13222170e+00 -4.31779087e-01 5.50604820e-01 2.59772241e-01
6.46574020e-01 2.35821139e-02 -5.87425418e-02 -1.05279617e-01
5.91403432e-02 -7.15313777e-02 2.53645271e-01 1.16386390e+00
-2.46293560e-01 -1.19518685e+00 -3.59445930e-01 -1.07273191e-01
-4.29405957e-01 -7.41378605e-01 -1.61440060e-01 1.02355361e+00
-3.27067316e-01 8.40858936e-01 -7.08248913e-02 -3.23542327e-01
1.11068614e-01 4.81754005e-01 6.14234746e-01 -3.95037264e-01
-3.80644798e-01 7.86224067e-01 2.32589617e-01 -7.42418528e-01
-4.84782577e-01 -8.11707020e-01 -6.84639335e-01 -2.33173236e-01
-1.08993240e-01 1.96710944e-01 6.87274516e-01 7.17490852e-01
2.68995881e-01 3.68820101e-01 3.63019779e-02 -6.65357053e-01
-1.04313105e-01 -9.96713221e-01 7.72553906e-02 1.47898927e-01
2.36392722e-01 -5.17921925e-01 -1.74118698e-01 3.65667790e-01] | [11.143969535827637, 1.5204808712005615] |
08432383-a8cb-4501-a37e-95618c3c3c4a | a-retrofitting-model-for-incorporating | null | null | https://aclanthology.org/2020.coling-main.111 | https://aclanthology.org/2020.coling-main.111.pdf | A Retrofitting Model for Incorporating Semantic Relations into Word Embeddings | We present a novel retrofitting model that can leverage relational knowledge available in a knowledge resource to improve word embeddings. The knowledge is captured in terms of relation inequality constraints that compare similarity of related and unrelated entities in the context of an anchor entity. These constraints are used as training data to learn a non-linear transformation function that maps original word vectors to a vector space respecting these constraints. The transformation function is learned in a similarity metric learning setting using Triplet network architecture. We applied our model to synonymy, antonymy and hypernymy relations in WordNet and observed large gains in performance over original distributional models as well as other retrofitting approaches on word similarity task and significant overall improvement on lexical entailment detection task. | ['Pushpak Bhattacharyya', 'Sreedhar Reddy', 'Sapan Shah'] | 2020-12-01 | null | null | null | coling-2020-8 | ['word-similarity'] | ['natural-language-processing'] | [ 2.27215827e-01 2.83265412e-01 -6.27041459e-01 -6.68256879e-01
-3.85073662e-01 -6.14908755e-01 6.98647976e-01 6.71505272e-01
-9.98681307e-01 4.45537537e-01 7.55201697e-01 -3.37448597e-01
-4.12325501e-01 -1.04652131e+00 -4.93390441e-01 4.14654724e-02
8.15765113e-02 6.21441364e-01 2.38187332e-02 -5.87312222e-01
4.61036935e-02 2.32530639e-01 -9.18265402e-01 1.83120072e-01
8.66034567e-01 6.72696054e-01 -2.63541341e-01 1.49887934e-01
-2.24118918e-01 5.35776794e-01 -2.33692765e-01 -8.73048306e-01
2.38167152e-01 5.94349168e-02 -9.52178121e-01 -7.02807367e-01
5.64315975e-01 2.36858964e-01 -7.19937980e-01 9.48280573e-01
3.40396672e-01 6.56422615e-01 8.56662273e-01 -1.04946518e+00
-1.59989727e+00 8.62949610e-01 -1.94097638e-01 5.11357784e-01
5.06353319e-01 -4.51708853e-01 1.99297488e+00 -1.16240776e+00
6.44966781e-01 1.26577318e+00 7.05201805e-01 2.35060796e-01
-1.51094198e+00 -6.44490182e-01 -3.02864751e-03 6.36449873e-01
-1.50761616e+00 -2.17822015e-01 3.66053134e-01 -2.87820935e-01
1.96827149e+00 3.51734534e-02 4.22753841e-01 1.05528426e+00
-3.43474969e-02 1.97719008e-01 2.89856255e-01 -6.24956727e-01
-1.14261471e-01 1.11860573e-01 4.81951654e-01 4.44343209e-01
5.25841951e-01 -5.62190637e-02 -5.71568370e-01 -3.61735493e-01
3.50790232e-01 7.78606683e-02 1.80369224e-02 -4.03260589e-01
-1.00248134e+00 1.23314977e+00 8.96627426e-01 3.67929667e-01
-1.49262071e-01 1.84845090e-01 4.05811012e-01 4.83821571e-01
5.84733605e-01 1.12695098e+00 -6.19591296e-01 1.58154890e-01
-3.90094072e-01 1.12674132e-01 8.39168847e-01 9.17774320e-01
8.18662465e-01 -2.04693794e-01 -2.65038639e-01 1.15053010e+00
3.76887947e-01 6.00191593e-01 8.90479445e-01 -5.04664183e-01
5.25886953e-01 8.37290049e-01 -3.56778055e-01 -1.18770373e+00
-4.38321799e-01 -2.06668884e-01 -2.71314621e-01 -4.79698181e-01
-9.94464979e-02 1.57830968e-01 -7.03037500e-01 2.00849915e+00
2.81855583e-01 4.56795216e-01 1.47835493e-01 6.03472054e-01
1.01751709e+00 4.18671012e-01 2.86956698e-01 1.86594337e-01
1.42613506e+00 -7.67673850e-01 -6.07805848e-01 -2.77641356e-01
1.14046562e+00 -6.58728957e-01 1.34150648e+00 -3.28299761e-01
-8.26694906e-01 -1.44800842e-01 -1.10741520e+00 -4.28261548e-01
-8.11728835e-01 -3.31299633e-01 7.51545608e-01 4.23502862e-01
-6.64387763e-01 6.47006392e-01 -5.09783983e-01 -6.91471517e-01
3.12669337e-01 3.18400621e-01 -7.59025574e-01 -2.10606888e-01
-1.66402602e+00 1.69688904e+00 6.19908154e-01 -4.24933016e-01
-5.45342304e-02 -1.17340779e+00 -1.43162322e+00 2.51409322e-01
3.47049147e-01 -8.04275692e-01 8.28397453e-01 -4.33999449e-01
-8.86508167e-01 1.02194929e+00 -5.78881167e-02 -6.31472766e-01
-4.76797611e-01 -2.40975201e-01 -9.55756545e-01 -3.88192922e-01
1.23970523e-01 2.80431360e-01 2.46682927e-01 -4.62568223e-01
-3.55265498e-01 -1.58376426e-01 1.58557713e-01 4.42535758e-01
-7.87371397e-01 1.45054802e-01 -5.39822690e-02 -7.88715661e-01
-1.93294302e-01 -6.97580695e-01 5.79269696e-03 -1.98709235e-01
-3.57076973e-01 -6.97237194e-01 3.82462114e-01 -4.70956296e-01
1.57791102e+00 -1.73381960e+00 2.71887213e-01 6.67446733e-01
2.32775107e-01 3.87819678e-01 -7.14564741e-01 6.42002165e-01
-4.36411113e-01 1.15296088e-01 -4.88419831e-02 1.62009865e-01
3.10992390e-01 4.38411802e-01 -5.01861811e-01 2.04323724e-01
3.85965794e-01 1.29731882e+00 -1.02831912e+00 -1.79026008e-01
-7.22985528e-03 4.44873571e-01 -8.30833912e-01 9.79110152e-02
5.80928251e-02 -6.16625607e-01 -9.53282863e-02 2.85032660e-01
2.52446204e-01 -5.94710857e-02 4.29544359e-01 -6.88233912e-01
6.45923793e-01 9.44738090e-01 -7.89138675e-01 1.65689743e+00
-8.41087937e-01 4.42236543e-01 -8.34372878e-01 -1.10482669e+00
9.49639678e-01 1.17747091e-01 3.92605096e-01 -8.76243830e-01
-1.85245078e-03 1.09811924e-01 2.36887768e-01 -4.64498669e-01
7.59399116e-01 -4.46289897e-01 -1.79530039e-01 5.04752576e-01
4.86965060e-01 -5.31260446e-02 2.21363723e-01 3.68637651e-01
1.32755125e+00 -3.77834849e-02 7.03681767e-01 -2.04459548e-01
3.53208810e-01 -1.86712116e-01 4.31428134e-01 2.94055134e-01
3.02118152e-01 7.99073428e-02 2.05566168e-01 -3.43255132e-01
-1.10194314e+00 -1.37320995e+00 -2.67635316e-01 1.51744521e+00
-4.96750139e-02 -1.03623223e+00 -1.26181236e-02 -8.91512096e-01
5.98167479e-01 1.00460553e+00 -8.64627481e-01 -9.37233031e-01
-7.12040544e-01 -5.54236710e-01 7.65223265e-01 9.12204862e-01
-2.48464882e-01 -8.92078102e-01 1.03535205e-02 2.44864617e-02
1.54968679e-01 -1.25430751e+00 -9.64270949e-01 4.65723842e-01
-4.37687516e-01 -1.15663350e+00 -4.94010821e-02 -1.05163705e+00
5.33400714e-01 2.97897179e-02 1.48766077e+00 7.09055290e-02
-4.68132496e-01 3.51546526e-01 -3.73479903e-01 -1.48356020e-01
2.49100998e-02 2.51517713e-01 4.48844194e-01 -3.51089269e-01
1.22866833e+00 -6.14358544e-01 -2.25114271e-01 3.42652835e-02
-9.64899063e-01 -6.87054992e-01 2.22288072e-01 1.08570373e+00
3.37676197e-01 -5.55175364e-01 6.30423188e-01 -9.47404504e-01
8.02282572e-01 -8.80044758e-01 -2.30704591e-01 4.79662836e-01
-9.76037681e-01 6.05465889e-01 2.96170145e-01 -6.86533928e-01
-7.48080969e-01 -3.98954868e-01 -7.43241832e-02 -2.10827157e-01
3.09842169e-01 8.97757828e-01 -9.64078382e-02 9.44363475e-02
8.51571500e-01 -3.99888664e-01 -4.18203056e-01 -3.78056556e-01
1.10233247e+00 5.12972236e-01 5.60485125e-01 -8.09666455e-01
1.05122173e+00 1.00964069e-01 -1.58542756e-03 -5.53633690e-01
-1.23062444e+00 -6.41785502e-01 -7.88828731e-01 7.74941444e-01
7.46393859e-01 -8.33607733e-01 -5.50422072e-01 -6.49174571e-01
-1.07061827e+00 1.55817017e-01 -6.87346041e-01 6.97176754e-01
-6.58444092e-02 2.56075740e-01 -4.13526267e-01 -6.65763766e-02
-3.69527757e-01 -3.03368211e-01 5.90613663e-01 -1.11524917e-01
-8.06102514e-01 -1.56540227e+00 5.49554110e-01 2.29556244e-02
6.04428113e-01 -1.19215414e-01 1.53359783e+00 -1.50343442e+00
9.56553519e-02 -3.79346579e-01 -2.81577438e-01 3.49415839e-01
4.41308379e-01 -1.61079153e-01 -5.69678247e-01 -1.86804026e-01
-4.47454333e-01 -3.61349702e-01 9.07703817e-01 -1.79488227e-01
5.77359438e-01 -5.22418559e-01 -4.04585570e-01 7.76386619e-01
1.34663749e+00 -1.84059188e-01 3.83824438e-01 2.50888616e-01
8.96315575e-01 5.05432248e-01 3.78182024e-01 1.22521721e-01
6.44982219e-01 8.83592069e-01 9.11594704e-02 3.65606099e-02
-1.23472191e-01 -5.83023489e-01 1.32239491e-01 9.20913815e-01
3.58563840e-01 -8.91720727e-02 -9.33289409e-01 7.53228903e-01
-1.76221204e+00 -9.95522618e-01 7.65029192e-02 2.11155748e+00
1.16335702e+00 -2.07604125e-01 -1.51591733e-01 -2.66963065e-01
5.50435781e-01 4.52936627e-02 -4.20777202e-01 -6.67944849e-01
-2.35645920e-01 7.95416057e-01 4.58097905e-01 8.00283313e-01
-9.78861690e-01 1.34898627e+00 6.37282658e+00 6.71508074e-01
-7.37592340e-01 1.86821863e-01 -2.88800806e-01 -4.09230024e-01
-8.54892969e-01 1.29670635e-01 -5.83415568e-01 7.91074038e-02
8.65278542e-01 -7.18220651e-01 3.91039282e-01 5.05041599e-01
-4.92469937e-01 4.14422631e-01 -1.54992712e+00 9.58426118e-01
4.31840062e-01 -1.29071057e+00 3.83388340e-01 -5.06679453e-02
7.59906530e-01 2.70921499e-01 1.38544694e-01 5.37163198e-01
8.58634055e-01 -1.39313006e+00 -3.48129570e-02 4.16027933e-01
8.50580752e-01 -9.10913169e-01 7.63960958e-01 -2.46477813e-01
-1.19621849e+00 4.52841856e-02 -7.09863782e-01 -3.54033560e-02
9.07669589e-02 4.20717239e-01 -1.08622444e+00 4.51230824e-01
2.29692966e-01 1.10492730e+00 -5.87777495e-01 7.11436033e-01
-5.32628894e-01 5.20111680e-01 -3.42611313e-01 3.01629175e-02
2.66944140e-01 -1.56000733e-01 3.09993982e-01 1.40844786e+00
2.21874893e-01 -6.69477284e-02 -6.78185299e-02 9.14546430e-01
-5.49503267e-01 5.05161107e-01 -9.00626659e-01 -1.55947104e-01
9.65479195e-01 1.31983745e+00 7.54898712e-02 -2.63889760e-01
-6.56811059e-01 9.01724041e-01 8.56514156e-01 3.25438887e-01
-6.60850286e-01 -6.85733736e-01 1.20589364e+00 -1.00873671e-01
2.96085924e-01 3.78278308e-02 -8.44959635e-03 -1.33277988e+00
4.97760177e-02 -4.86429751e-01 8.50492179e-01 -4.60890800e-01
-1.98723578e+00 5.31430721e-01 1.00606056e-02 -6.76083088e-01
-3.03729922e-01 -7.76098609e-01 -6.57350421e-01 1.02452672e+00
-1.62575543e+00 -1.23618531e+00 1.71989664e-01 5.39352417e-01
-9.98427123e-02 -5.74332595e-01 1.28873217e+00 3.17918628e-01
-3.78270000e-01 1.03916895e+00 4.25785594e-02 3.01180393e-01
1.13974345e+00 -1.33932447e+00 7.38112628e-01 5.01301885e-01
7.33394623e-01 1.14782619e+00 3.61829549e-01 -6.51677132e-01
-9.89678860e-01 -1.13258266e+00 1.65109766e+00 -9.31045651e-01
1.21334255e+00 -3.11354220e-01 -1.02530241e+00 1.04398453e+00
2.84465283e-01 3.73916090e-01 1.34416854e+00 7.15364337e-01
-1.36073422e+00 4.12404463e-02 -9.37180281e-01 5.95102012e-01
1.26309323e+00 -1.08138978e+00 -1.39180195e+00 2.33789980e-01
1.10199356e+00 -8.97106156e-02 -1.25274265e+00 3.27603191e-01
5.16924381e-01 6.85892552e-02 1.33920670e+00 -1.67948735e+00
3.86574537e-01 -2.22391821e-03 -6.42072737e-01 -1.65029430e+00
-5.82769930e-01 -4.13460076e-01 -3.70487779e-01 1.15084314e+00
7.51737237e-01 -7.12945938e-01 3.77525210e-01 7.31707752e-01
4.62334640e-02 -8.18828583e-01 -9.05678272e-01 -1.09060633e+00
4.83475983e-01 -2.10422128e-01 8.72919083e-01 1.42795146e+00
7.39175618e-01 9.79175091e-01 -1.49791026e-02 -4.62997071e-02
2.44790182e-01 -9.10229459e-02 1.92378893e-01 -1.29191804e+00
-6.19570017e-02 -4.15686101e-01 -6.89564526e-01 -6.63443923e-01
8.02069187e-01 -1.88426614e+00 -4.05939996e-01 -1.53812730e+00
1.91954821e-01 -3.82741928e-01 -6.43426597e-01 6.81120813e-01
-3.79978538e-01 2.79551774e-01 1.24097787e-01 -1.42608717e-01
-5.17742157e-01 7.18826652e-01 4.16637748e-01 -1.27112567e-01
6.69210777e-02 -4.62149054e-01 -9.54631388e-01 6.71095848e-01
4.34831381e-01 -6.71718657e-01 -5.35864472e-01 -8.03783894e-01
8.90759408e-01 -6.88739717e-01 1.74774483e-01 -2.28301272e-01
3.38778824e-01 -8.69368762e-02 9.58037376e-02 -8.47354010e-02
3.66745681e-01 -9.99525189e-01 -3.30239862e-01 2.73026284e-02
-9.80378032e-01 4.26022857e-01 -8.10580254e-02 5.66820085e-01
-1.84072927e-01 -2.32279152e-01 3.98602039e-01 3.49853218e-01
-6.53863370e-01 1.87890992e-01 1.40132755e-01 7.38803267e-01
7.91654825e-01 -2.02270169e-02 -4.52504992e-01 -7.81687498e-02
-6.59844339e-01 1.87514335e-01 2.58065522e-01 9.87156868e-01
6.94517016e-01 -1.96340752e+00 -8.54623616e-01 8.97735953e-02
6.90270185e-01 -5.31113982e-01 -4.07927364e-01 4.83238488e-01
1.29517049e-01 5.41031182e-01 -5.31970486e-02 7.22178072e-02
-1.16848528e+00 7.60050535e-01 3.50590169e-01 -3.32728624e-01
-4.60554779e-01 9.98993337e-01 -2.40782015e-02 -9.66747880e-01
1.33524165e-01 -4.81426626e-01 -2.81583577e-01 1.62500426e-01
2.49893233e-01 3.23523641e-01 2.07165778e-01 -7.87127674e-01
-7.15518832e-01 6.10926211e-01 -2.54560143e-01 6.92699105e-02
1.45825946e+00 1.17322244e-02 -2.03529626e-01 2.11493924e-01
1.68520963e+00 9.79094952e-02 -1.83463499e-01 -1.15404439e+00
6.16558552e-01 -4.54546183e-01 -1.80531353e-01 -7.51759946e-01
-6.87196076e-01 5.04484951e-01 1.45407721e-01 -1.71014100e-01
4.41184223e-01 3.53066206e-01 8.83310139e-01 8.98157239e-01
-9.34030190e-02 -1.15333593e+00 1.37250215e-01 1.09772754e+00
7.12826967e-01 -1.06926048e+00 8.09777379e-02 -3.10117900e-01
-7.40859807e-01 9.63104069e-01 5.70630312e-01 -4.23678696e-01
9.26509738e-01 2.45844007e-01 -1.39066383e-01 -4.60122555e-01
-1.00557482e+00 -4.27672148e-01 9.94184017e-01 8.69398892e-01
8.21724415e-01 -6.70112967e-02 -3.60856920e-01 7.28162885e-01
-3.30887049e-01 -4.66806471e-01 4.35228460e-02 4.64856565e-01
-1.62299722e-01 -1.20045805e+00 3.05379719e-01 7.24956512e-01
-2.10729852e-01 -8.68208945e-01 -4.58415657e-01 5.31142354e-01
7.90819600e-02 7.24437773e-01 2.21123993e-01 -4.51482236e-01
6.22232199e-01 3.12762111e-01 4.99273509e-01 -1.12366927e+00
-8.01671803e-01 -6.59752131e-01 3.92879814e-01 -6.09252095e-01
-2.28065923e-02 -2.91054815e-01 -1.32146513e+00 -8.83635953e-02
-5.42092681e-01 2.19694331e-01 2.70120323e-01 1.11844158e+00
4.12578672e-01 4.60597873e-01 4.61582065e-01 -3.44675370e-02
-9.25381601e-01 -9.98749316e-01 -4.12469655e-01 9.95756924e-01
-3.56520563e-02 -5.57667732e-01 -1.29359096e-01 -2.30996966e-01] | [10.067839622497559, 8.722655296325684] |
bc81de84-e7c4-4461-9bc7-7459479e79c2 | category-guided-attention-network-for-brain | 2203.15383 | null | https://arxiv.org/abs/2203.15383v1 | https://arxiv.org/pdf/2203.15383v1.pdf | Category Guided Attention Network for Brain Tumor Segmentation in MRI | Objective: Magnetic resonance imaging (MRI) has been widely used for the analysis and diagnosis of brain diseases. Accurate and automatic brain tumor segmentation is of paramount importance for radiation treatment. However, low tissue contrast in tumor regions makes it a challenging task.Approach: We propose a novel segmentation network named Category Guided Attention U-Net (CGA U-Net). In this model, we design a Supervised Attention Module (SAM) based on the attention mechanism, which can capture more accurate and stable long-range dependency in feature maps without introducing much computational cost. Moreover, we propose an intra-class update approach to reconstruct feature maps by aggregating pixels of the same category. Main results: Experimental results on the BraTS 2019 datasets show that the proposed method outperformers the state-of-the-art algorithms in both segmentation performance and computational complexity. Significance: The CGA U-Net can effectively capture the global semantic information in the MRI image by using the SAM module, while significantly reducing the computational cost. Code is available at https://github.com/delugewalker/CGA-U-Net. | ['Sen Zha', 'Meng Ding', 'Chen Chen', 'Hong Yu', 'Jiangyun Li'] | 2022-03-29 | null | null | null | null | ['brain-tumor-segmentation'] | ['medical'] | [ 2.36719713e-01 1.01806954e-01 -1.68667182e-01 -3.89782310e-01
-9.36471224e-01 3.51217277e-02 2.91792035e-01 1.82824478e-01
-5.33780754e-01 5.92471182e-01 1.81431979e-01 -5.47792278e-02
-2.10229725e-01 -7.27690876e-01 -5.70828557e-01 -9.68689740e-01
1.65628552e-01 5.13972759e-01 5.19342959e-01 6.39889836e-02
2.13218108e-01 4.87653226e-01 -1.07935190e+00 1.83383837e-01
1.15771389e+00 1.07891119e+00 7.93443978e-01 1.46268547e-01
-1.61658838e-01 7.63874590e-01 -1.57669753e-01 4.70510647e-02
-9.54427198e-02 -4.37165201e-01 -1.07924819e+00 8.22153986e-02
-5.52630872e-02 -2.03295231e-01 -3.10784578e-01 1.48409653e+00
5.85956454e-01 1.39391035e-01 6.64638221e-01 -7.80225694e-01
-4.75536108e-01 7.60011613e-01 -6.43000126e-01 5.78097820e-01
-2.50426292e-01 -4.78914976e-02 8.34661484e-01 -9.17800069e-01
6.13805354e-01 7.19920158e-01 5.43012321e-01 6.45668089e-01
-9.45201993e-01 -6.24143183e-01 2.61333644e-01 6.29708409e-01
-1.40950155e+00 -8.76489356e-02 7.19779611e-01 -4.23892945e-01
5.71411729e-01 3.37526441e-01 7.56348968e-01 8.33175659e-01
5.91754615e-01 1.06621087e+00 1.04119217e+00 -1.22078337e-01
1.89534158e-01 -2.91470975e-01 5.66702247e-01 7.22655714e-01
1.54370189e-01 -3.47470582e-01 -5.90288900e-02 1.55401617e-01
8.42239380e-01 2.76573569e-01 -6.14644706e-01 -3.36335629e-01
-1.29323888e+00 7.75823116e-01 1.17649913e+00 8.99440229e-01
-6.87853217e-01 2.70716637e-01 3.26047271e-01 -2.83701688e-01
7.04428613e-01 2.14263827e-01 -9.01165605e-02 2.24889711e-01
-8.66521180e-01 -1.21665955e-01 7.11559206e-02 5.44036567e-01
3.17132682e-01 -1.26701087e-01 -4.60021198e-01 8.59538496e-01
9.42271575e-02 1.53350085e-01 1.01546812e+00 -5.68437815e-01
-7.70468339e-02 4.91841465e-01 -2.91759163e-01 -7.21354544e-01
-8.27605963e-01 -8.70018840e-01 -1.07315969e+00 -4.05028537e-02
2.98524976e-01 3.90110910e-02 -1.32525098e+00 1.71924734e+00
2.98112303e-01 2.75784224e-01 -3.40976894e-01 1.15663874e+00
1.09962809e+00 1.96964562e-01 2.41435841e-01 -2.28009343e-01
1.41156435e+00 -1.23529792e+00 -8.45705807e-01 -1.34940505e-01
5.59762299e-01 -3.58938873e-01 8.11010420e-01 -5.63334581e-03
-1.00616252e+00 -3.20475489e-01 -7.52458096e-01 -6.78974912e-02
-2.29466826e-01 1.39186988e-02 7.75197983e-01 4.20881659e-01
-1.14588940e+00 4.11664933e-01 -1.16867638e+00 -2.30166331e-01
8.87708962e-01 5.82508266e-01 -2.65172571e-01 -1.90020755e-01
-1.07286167e+00 7.53789723e-01 4.56572145e-01 1.56004295e-01
-8.76162350e-01 -8.30512404e-01 -6.67642534e-01 1.37320697e-01
3.28466594e-01 -6.50479198e-01 1.20955169e+00 -9.82285738e-01
-1.27433658e+00 7.19909728e-01 -2.95530677e-01 -4.52142656e-01
4.42838639e-01 5.83383068e-02 -3.36510204e-02 4.46183443e-01
2.57628173e-01 8.36628020e-01 4.47675496e-01 -1.08884180e+00
-3.16225797e-01 -6.81214094e-01 -3.81762683e-01 4.01320010e-01
-1.74730182e-01 -1.42864725e-02 -5.22878289e-01 -8.68037164e-01
4.75383848e-01 -7.86262572e-01 -4.95735586e-01 -2.32610911e-01
-4.31190372e-01 -2.24266052e-01 7.10487485e-01 -1.04468846e+00
1.00989330e+00 -1.89496136e+00 3.84267658e-01 2.73048073e-01
5.73593080e-01 1.97533146e-01 2.73568600e-01 -4.12391722e-01
-1.97101191e-01 -1.33578688e-01 -9.08438265e-01 -7.96593577e-02
-3.41677725e-01 1.50827942e-02 2.42426798e-01 5.88569701e-01
-1.82978570e-01 1.25610149e+00 -9.56855595e-01 -5.04350364e-01
2.16625735e-01 5.06300211e-01 -5.90256512e-01 7.44454786e-02
-5.13234995e-02 9.48151767e-01 -5.60013056e-01 7.55431414e-01
4.25687879e-01 -4.20095682e-01 -4.55276631e-02 -2.37676948e-01
1.42773194e-02 -1.48624614e-01 -4.18313205e-01 1.96387994e+00
-3.17039013e-01 4.84298706e-01 1.65250197e-01 -1.31105900e+00
6.06636167e-01 3.07131052e-01 9.69073355e-01 -8.49040508e-01
6.97771728e-01 3.44818383e-01 3.85873824e-01 -4.23224509e-01
6.21557608e-02 -1.45826429e-01 8.43680352e-02 2.56705523e-01
-2.46331356e-02 -9.82788056e-02 -2.25499216e-02 1.35912806e-01
9.88760233e-01 -2.25232393e-01 2.22660899e-01 -5.26440203e-01
6.01862907e-01 -1.47846252e-01 6.25509977e-01 6.55209243e-01
-5.37069321e-01 5.88019013e-01 2.83162802e-01 -1.86282948e-01
-6.46598160e-01 -7.83298671e-01 -4.81174648e-01 6.72571182e-01
1.93611741e-01 -1.66620612e-02 -1.26013148e+00 -7.01663613e-01
-2.49387860e-01 5.95525324e-01 -1.02739215e+00 -1.90691695e-01
-5.44939160e-01 -1.09560072e+00 3.14197503e-02 7.45177507e-01
6.94946349e-01 -1.10891116e+00 -6.24686420e-01 2.35018954e-01
-5.24566472e-01 -8.70084465e-01 -6.50935590e-01 2.79514641e-02
-9.44244266e-01 -9.34602976e-01 -1.30122626e+00 -7.67282486e-01
1.06874132e+00 1.69700876e-01 7.09407151e-01 1.12680465e-01
-5.78325868e-01 2.93529153e-01 -3.76232862e-01 -3.01472336e-01
7.32568428e-02 2.64560789e-01 -3.58056456e-01 1.97367147e-01
2.55489051e-01 -4.29469705e-01 -8.08119357e-01 2.16405183e-01
-9.23994839e-01 3.44552547e-01 8.32458377e-01 9.88207519e-01
9.47263658e-01 -5.57209887e-02 6.92227244e-01 -9.84043956e-01
3.80287975e-01 -5.88105679e-01 -3.47369611e-01 1.53402742e-02
-4.37757909e-01 -2.03792781e-01 4.08164650e-01 -7.84248039e-02
-8.23533058e-01 8.65432099e-02 -4.34441715e-01 -4.05782670e-01
2.73986757e-02 5.49171865e-01 -1.86817035e-01 -3.07777196e-01
2.02628091e-01 3.87370586e-01 7.60217011e-02 -3.91329825e-01
-2.33551506e-02 5.87597668e-01 5.91345489e-01 -1.42234907e-01
1.94821298e-01 6.23188674e-01 8.91739596e-03 -7.57462621e-01
-9.01226163e-01 -5.43666184e-01 -7.18262494e-01 -4.03750777e-01
1.17661190e+00 -5.59020579e-01 -3.10702115e-01 7.40296721e-01
-8.51667523e-01 -4.07275230e-01 -1.60510927e-01 5.50658941e-01
-5.50920784e-01 2.98427612e-01 -5.65816581e-01 -1.67055279e-01
-7.13590920e-01 -1.71220052e+00 8.45467091e-01 3.05565745e-01
5.16959205e-02 -9.76603925e-01 -2.84145772e-01 4.83337879e-01
6.14161491e-01 2.91248351e-01 8.95377278e-01 -6.17622316e-01
-5.75715065e-01 -1.12837188e-01 -5.70055783e-01 7.39793107e-02
2.54962116e-01 -6.45980358e-01 -8.01114023e-01 -3.24913830e-01
5.74867763e-02 1.29440024e-01 1.17514479e+00 1.00838709e+00
1.88655365e+00 6.89809769e-02 -4.31559414e-01 7.07559764e-01
1.30003607e+00 2.67900914e-01 5.52302063e-01 2.33365104e-01
9.59509492e-01 2.90517122e-01 3.11427295e-01 1.84020057e-01
3.97003710e-01 6.59171402e-01 5.31514168e-01 -3.23057532e-01
-5.02938092e-01 3.37751597e-01 -1.54709011e-01 1.05052793e+00
-1.35545015e-01 1.32917970e-01 -1.15525973e+00 9.47136402e-01
-1.84516716e+00 -6.87292457e-01 -1.91408440e-01 2.02125263e+00
7.38830447e-01 -2.91159116e-02 -1.66239694e-01 1.12559028e-01
8.47059786e-01 -7.72316009e-02 -6.57009959e-01 6.19054213e-02
2.17955098e-01 3.39075595e-01 4.83503908e-01 6.07660592e-01
-1.01147842e+00 7.99703777e-01 5.13872719e+00 1.06671929e+00
-1.27008343e+00 7.32585669e-01 9.29886401e-01 2.49463711e-02
-2.58704215e-01 -4.47670102e-01 -4.29535389e-01 6.85643613e-01
6.44438565e-01 -1.57058850e-01 8.23803097e-02 5.75842857e-01
1.59032181e-01 -2.07658380e-01 -5.85026622e-01 8.51379156e-01
2.92949319e-01 -1.31548095e+00 -4.09573689e-03 5.29841939e-03
7.38161564e-01 2.81275928e-01 6.37031859e-03 4.35777009e-02
-5.53331785e-02 -8.60309780e-01 5.59185147e-01 7.15840876e-01
7.58776844e-01 -8.33442211e-01 1.12572634e+00 1.28165975e-01
-9.89676774e-01 -3.53314430e-02 -1.94730267e-01 6.18584156e-01
9.16974470e-02 7.70585656e-01 -6.23194039e-01 6.80255651e-01
7.15160310e-01 8.08742583e-01 -6.17508173e-01 1.41869581e+00
-1.23736233e-01 6.05621099e-01 -1.08340554e-01 1.44428447e-01
4.43562239e-01 -1.11747839e-01 6.72533095e-01 9.89660621e-01
2.94249862e-01 3.47514570e-01 2.86749721e-01 7.67785192e-01
-1.56694829e-01 2.29505748e-01 -9.60866809e-02 3.35272074e-01
4.42694165e-02 1.31089330e+00 -1.34037876e+00 -3.81764531e-01
-1.89780325e-01 9.74533916e-01 3.54743540e-01 1.12300791e-01
-9.14064646e-01 -4.04215306e-01 1.92541748e-01 2.60316044e-01
1.95199654e-01 4.40190807e-02 -5.44247925e-01 -1.00600278e+00
-2.29899347e-01 -4.39483136e-01 4.29924130e-01 -7.49305546e-01
-8.85302186e-01 8.90760422e-01 -1.12554632e-01 -8.41785610e-01
2.43230075e-01 -4.08554107e-01 -5.94203532e-01 6.10762119e-01
-1.63846326e+00 -1.07943678e+00 -7.31147528e-01 7.50365973e-01
7.25410283e-01 7.06203431e-02 6.28416896e-01 3.60178530e-01
-8.46299291e-01 4.71259207e-01 1.23815387e-01 1.26919270e-01
3.99020553e-01 -1.20887089e+00 -1.03572980e-01 7.94082642e-01
-2.96714783e-01 3.49600017e-01 3.25266510e-01 -7.30195522e-01
-8.50979865e-01 -1.26405585e+00 5.15914738e-01 4.12953570e-02
5.97041905e-01 -4.49485183e-02 -1.04547489e+00 6.89443290e-01
2.54513323e-01 3.18879515e-01 5.89265227e-01 -4.64157432e-01
1.60436913e-01 -6.54217275e-03 -1.26155365e+00 3.65376323e-01
7.74316609e-01 -1.96809500e-01 -3.63411993e-01 5.76201022e-01
6.94999814e-01 -5.78367829e-01 -9.31643784e-01 5.55881858e-01
2.14493766e-01 -6.89472973e-01 8.55046570e-01 -8.65082145e-02
3.55790138e-01 -1.16880834e-01 6.94872588e-02 -1.39193463e+00
-6.92819476e-01 -1.00439057e-01 3.78224775e-02 6.55490339e-01
1.80883557e-01 -7.94559360e-01 5.70861161e-01 5.85682213e-01
-6.15270436e-01 -1.16432047e+00 -1.17627454e+00 -5.47902405e-01
2.53369331e-01 -3.03388417e-01 3.85836124e-01 9.02506173e-01
-1.06081374e-01 -8.38499144e-02 7.44625404e-02 1.75797358e-01
8.82415712e-01 -5.38368151e-02 -8.38359594e-02 -1.04967380e+00
-1.28342211e-01 -8.10642540e-01 -5.83936334e-01 -5.37663937e-01
2.55467266e-01 -1.37457490e+00 -2.11093985e-02 -1.89394915e+00
5.45899093e-01 -3.69772404e-01 -7.01280296e-01 5.94177842e-01
-3.30375433e-01 4.39774781e-01 7.18199536e-02 3.22830200e-01
-5.57749629e-01 7.54345715e-01 1.53312910e+00 -2.17175975e-01
1.11562051e-01 -8.87707770e-02 -7.03393698e-01 7.73082316e-01
9.73684430e-01 -3.84176642e-01 -1.69432297e-01 -4.82012540e-01
-6.42619431e-01 -6.15977421e-02 3.57661515e-01 -1.23390913e+00
3.94850105e-01 9.36603770e-02 4.44815397e-01 -5.49115598e-01
9.70357060e-02 -7.81708360e-01 -1.19253971e-01 7.86088765e-01
-2.38745719e-01 -1.85498998e-01 1.84895635e-01 2.96182573e-01
-2.58395851e-01 -2.90635437e-01 1.15775812e+00 -2.96500862e-01
-6.40897393e-01 6.79655552e-01 -4.22472537e-01 -1.87310129e-01
1.29981399e+00 4.86641563e-03 -5.92369661e-02 -8.78594890e-02
-9.84543622e-01 1.75728634e-01 8.35635960e-02 2.09271371e-01
6.64695799e-01 -1.28043401e+00 -6.33668065e-01 1.41004309e-01
-5.70553914e-02 1.71915099e-01 6.50655568e-01 1.42870033e+00
-5.50987542e-01 6.69023037e-01 -2.77104735e-01 -6.87824786e-01
-1.12261701e+00 3.64276856e-01 6.00118577e-01 -4.68362719e-01
-8.73823464e-01 1.02149582e+00 5.28854668e-01 -2.70570993e-01
7.32838586e-02 -5.05458057e-01 -5.31165659e-01 -5.36905676e-02
6.04993999e-01 1.17764913e-01 3.00547361e-01 -8.34950626e-01
-4.05198127e-01 5.13917208e-01 -4.59871441e-01 1.57713622e-01
1.40310061e+00 -9.80228484e-02 -3.65439236e-01 1.00464948e-01
1.04333365e+00 -4.19721931e-01 -1.04891860e+00 -2.75043130e-01
-2.09238738e-01 -2.82372773e-01 8.62378299e-01 -8.92840683e-01
-1.69603825e+00 7.93207824e-01 8.49280715e-01 -1.35839358e-01
1.27135599e+00 8.14949796e-02 1.02116168e+00 -1.27131104e-01
3.90616208e-01 -7.85622776e-01 -1.03321664e-01 3.14092368e-01
1.00002420e+00 -1.14796650e+00 -1.73857555e-01 -4.84717667e-01
-5.90953648e-01 6.98734641e-01 6.95191085e-01 -5.69051094e-02
8.15970957e-01 3.82823460e-02 -8.41158852e-02 -3.86866391e-01
-1.89396456e-01 -3.65756780e-01 3.13291997e-01 2.99738109e-01
3.36017251e-01 2.41073042e-01 -5.28255999e-01 9.61625457e-01
8.77246037e-02 -3.40825655e-02 3.48780662e-01 7.91577816e-01
-4.98280078e-01 -8.21629345e-01 -3.75377446e-01 7.99200952e-01
-6.42498970e-01 -2.59301811e-01 -9.49252024e-02 4.24892247e-01
-5.73953018e-02 5.29045761e-01 -1.54952630e-02 -1.10909529e-01
3.35308984e-02 -6.73069805e-02 4.61733133e-01 -3.99309516e-01
-5.57010770e-01 2.38908887e-01 -3.65911216e-01 -6.12655759e-01
-3.99884254e-01 -7.52549887e-01 -1.59034956e+00 1.54083923e-01
-3.39805573e-01 1.32934541e-01 6.33954227e-01 9.29223359e-01
3.50398928e-01 1.16701114e+00 5.57785809e-01 -8.53879571e-01
-9.25028771e-02 -1.02982259e+00 -5.32596469e-01 3.50514740e-01
9.71510336e-02 -9.41969395e-01 -6.98696449e-02 -2.60629207e-01] | [14.501426696777344, -2.367771625518799] |
79b8c669-5fc9-413b-9b32-2010f34eebc1 | blendmask-top-down-meets-bottom-up-for | 2001.00309 | null | https://arxiv.org/abs/2001.00309v3 | https://arxiv.org/pdf/2001.00309v3.pdf | BlendMask: Top-Down Meets Bottom-Up for Instance Segmentation | Instance segmentation is one of the fundamental vision tasks. Recently, fully convolutional instance segmentation methods have drawn much attention as they are often simpler and more efficient than two-stage approaches like Mask R-CNN. To date, almost all such approaches fall behind the two-stage Mask R-CNN method in mask precision when models have similar computation complexity, leaving great room for improvement. In this work, we achieve improved mask prediction by effectively combining instance-level information with semantic information with lower-level fine-granularity. Our main contribution is a blender module which draws inspiration from both top-down and bottom-up instance segmentation approaches. The proposed BlendMask can effectively predict dense per-pixel position-sensitive instance features with very few channels, and learn attention maps for each instance with merely one convolution layer, thus being fast in inference. BlendMask can be easily incorporated with the state-of-the-art one-stage detection frameworks and outperforms Mask R-CNN under the same training schedule while being 20% faster. A light-weight version of BlendMask achieves $ 34.2% $ mAP at 25 FPS evaluated on a single 1080Ti GPU card. Because of its simplicity and efficacy, we hope that our BlendMask could serve as a simple yet strong baseline for a wide range of instance-wise prediction tasks. Code is available at https://git.io/AdelaiDet | ['Yongming Huang', 'Youliang Yan', 'Chunhua Shen', 'Zhi Tian', 'Hao Chen', 'Kunyang Sun'] | 2020-01-02 | blendmask-top-down-meets-bottom-up-for-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Chen_BlendMask_Top-Down_Meets_Bottom-Up_for_Instance_Segmentation_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Chen_BlendMask_Top-Down_Meets_Bottom-Up_for_Instance_Segmentation_CVPR_2020_paper.pdf | cvpr-2020-6 | ['real-time-instance-segmentation'] | ['computer-vision'] | [ 4.14727867e-01 2.69743502e-01 -4.13197011e-01 -4.44048077e-01
-8.85369599e-01 -3.02073777e-01 4.14912194e-01 -4.87398393e-02
-6.64048493e-01 2.90716290e-01 -3.09426576e-01 -3.98003668e-01
4.57588404e-01 -8.42645943e-01 -9.66948509e-01 -3.85446668e-01
2.39605397e-01 4.68673825e-01 7.59370983e-01 4.41872105e-02
1.53784797e-01 4.24632639e-01 -1.69758856e+00 6.19563639e-01
6.79746985e-01 1.29106247e+00 3.46735001e-01 7.62243092e-01
-1.30798459e-01 5.94163895e-01 -4.92898911e-01 -5.43430328e-01
6.04712963e-01 7.77777936e-03 -6.87969685e-01 -1.04104318e-01
9.20896530e-01 -3.74868065e-01 -4.10949200e-01 8.67117405e-01
5.21749675e-01 -1.16515219e-01 1.41251639e-01 -7.82506585e-01
-2.13366851e-01 6.44778192e-01 -9.72349346e-01 3.90541971e-01
-2.79034287e-01 5.31727076e-01 1.05529201e+00 -9.12449002e-01
2.62022704e-01 1.07897973e+00 8.38452518e-01 6.37654543e-01
-1.21005940e+00 -7.21477687e-01 3.91899228e-01 2.24493816e-01
-1.31251156e+00 -2.33958259e-01 3.48922312e-01 -1.96405277e-01
1.32452691e+00 3.65110427e-01 8.20698559e-01 6.59564376e-01
-4.60241772e-02 1.30802763e+00 1.15630877e+00 -1.13566287e-01
-8.73352885e-02 -1.54932598e-02 1.63387194e-01 1.00955963e+00
3.30084413e-01 -3.58086117e-02 -4.32530373e-01 3.16931665e-01
9.50584173e-01 1.27394691e-01 -2.76965275e-02 7.54407719e-02
-1.23599195e+00 6.57158375e-01 8.29361737e-01 3.12239565e-02
-1.42675906e-01 5.74184954e-01 3.87588114e-01 -2.17089429e-01
6.76565826e-01 3.91488761e-01 -6.80432618e-01 -1.07634425e-01
-1.39686430e+00 1.88490942e-01 6.46696448e-01 8.85663331e-01
9.19203520e-01 -7.38122091e-02 -5.24996459e-01 6.93185687e-01
1.07497737e-01 3.19953978e-01 3.62630963e-01 -9.28232074e-01
3.37634683e-01 7.10718572e-01 -1.98087141e-01 -5.64794540e-01
-4.69671607e-01 -7.50199020e-01 -4.92974758e-01 4.20138866e-01
5.01261473e-01 -1.23369820e-01 -1.39842558e+00 1.28707838e+00
4.45801258e-01 6.24837995e-01 -5.51239252e-01 9.84022439e-01
9.14386213e-01 5.81144691e-01 -2.59575117e-02 3.57237041e-01
1.67605865e+00 -1.60722971e+00 -1.37633592e-01 -7.07643211e-01
4.01775718e-01 -7.51647174e-01 1.06423545e+00 4.89560753e-01
-1.21598876e+00 -7.95309424e-01 -1.11038232e+00 -4.09547240e-01
-4.17081326e-01 3.43322486e-01 9.39217389e-01 7.10316718e-01
-1.08686078e+00 7.32611716e-01 -1.17309189e+00 -1.16624735e-01
1.16436064e+00 5.82073092e-01 -9.69009823e-04 -1.25859618e-01
-5.54064929e-01 6.87712848e-01 3.16421390e-01 1.66150689e-01
-8.48141134e-01 -1.04762924e+00 -6.53897047e-01 2.80066635e-02
6.68201149e-01 -5.56943655e-01 1.43102443e+00 -7.38352954e-01
-1.42046535e+00 1.08190763e+00 -2.54937500e-01 -8.17100883e-01
6.82930470e-01 -5.40491581e-01 7.23520815e-02 4.96807918e-02
1.21516116e-01 1.18003333e+00 9.35907304e-01 -8.52097154e-01
-9.27640676e-01 -3.35032105e-01 -8.47103633e-03 7.86265358e-02
-2.04131361e-02 -8.54382664e-02 -9.77021754e-01 -4.74970251e-01
-1.41055524e-01 -7.39749789e-01 -5.50659835e-01 2.74682999e-01
-5.12840509e-01 -1.76307231e-01 8.18449318e-01 -3.41934681e-01
9.61550176e-01 -1.95043004e+00 -2.52345890e-01 -2.09928825e-01
3.72888148e-01 8.89881194e-01 -4.93512824e-02 -7.76823759e-02
1.76433176e-01 1.62956323e-02 -5.04667401e-01 -7.27321267e-01
-8.33471641e-02 1.38315365e-01 -2.09961727e-01 4.79697317e-01
6.06981516e-01 1.34140587e+00 -6.81026459e-01 -4.04267311e-01
5.10418892e-01 5.69611371e-01 -5.73069096e-01 6.39154539e-02
-4.85007435e-01 3.09028924e-01 -1.96157902e-01 8.45389307e-01
6.82769775e-01 -4.40085351e-01 -2.79189825e-01 -3.48584741e-01
-3.30993921e-01 3.88799965e-01 -1.01111341e+00 1.79869413e+00
-3.39857608e-01 8.48553836e-01 1.41313270e-01 -1.05653405e+00
6.65946603e-01 -9.70913842e-02 2.47153565e-01 -8.53845954e-01
2.78303087e-01 1.66761458e-01 4.76003103e-02 -1.10646918e-01
4.45119113e-01 1.53023750e-01 1.48452997e-01 1.95837438e-01
-3.30009013e-02 -1.20421164e-01 2.01924294e-01 3.08510978e-02
1.19198608e+00 2.54023254e-01 1.60315186e-01 -1.06244884e-01
1.44492537e-01 1.61540240e-01 5.57686508e-01 1.14902699e+00
-3.73670310e-01 8.45471025e-01 2.87772119e-01 -7.42310762e-01
-9.26895678e-01 -8.38303387e-01 -4.40646410e-01 1.11434507e+00
2.56280482e-01 -3.69292915e-01 -1.01482379e+00 -6.44762576e-01
1.16711356e-01 1.83911800e-01 -6.35943651e-01 2.65608102e-01
-7.55952656e-01 -9.57351029e-01 7.76354790e-01 8.61050189e-01
8.70923221e-01 -1.10303223e+00 -9.36864138e-01 3.06531101e-01
3.46398830e-01 -1.31733906e+00 -3.39693487e-01 3.81390452e-01
-8.06817532e-01 -1.06533003e+00 -6.77872300e-01 -5.90466142e-01
5.92031598e-01 3.76431853e-01 1.28550458e+00 2.70855516e-01
-9.63672340e-01 -1.97223341e-03 -1.31314024e-01 -6.23413324e-01
2.16983259e-01 4.11356777e-01 -4.04475898e-01 -1.28144562e-01
4.68223989e-01 -2.30269611e-01 -1.15346026e+00 2.15460807e-01
-7.24917769e-01 4.85253066e-01 8.37458789e-01 6.74354196e-01
9.61519241e-01 -3.48126501e-01 2.09649280e-01 -1.24798703e+00
-6.17825836e-02 -1.33151561e-01 -7.26015270e-01 -5.79527952e-02
-4.78538930e-01 -2.17416421e-01 4.34416711e-01 -2.53027946e-01
-8.57540011e-01 4.12334323e-01 -4.66065198e-01 -3.73649448e-01
-2.84064263e-01 2.25714743e-02 4.63914089e-02 -1.42110884e-01
5.92776418e-01 1.05827026e-01 -8.44563767e-02 -5.16529381e-01
6.28787279e-01 5.12607932e-01 6.52833998e-01 -4.53779221e-01
6.11958921e-01 7.94866085e-01 -1.87795013e-01 -6.21382892e-01
-1.21370697e+00 -7.32203245e-01 -4.92896557e-01 -1.70508400e-02
9.41090167e-01 -1.17090082e+00 -6.69854581e-01 7.74074554e-01
-9.47795510e-01 -9.88286614e-01 -3.94836307e-01 5.63584045e-02
-3.51428568e-01 1.00531414e-01 -8.31136644e-01 -5.51992655e-01
-4.70602304e-01 -1.30673325e+00 1.47019994e+00 4.97233689e-01
5.95586821e-02 -5.53857207e-01 -4.09351230e-01 6.55756891e-01
4.84329492e-01 9.96453613e-02 2.43075877e-01 -4.55411673e-01
-9.67401028e-01 -1.96804553e-01 -7.92514265e-01 2.58521259e-01
-1.03897274e-01 1.36002019e-01 -1.28536642e+00 -1.83125347e-01
-2.64505982e-01 -2.31483936e-01 1.52409482e+00 6.06117427e-01
1.74283218e+00 -9.78604257e-02 -4.91347730e-01 1.00441182e+00
1.55308855e+00 -2.12571561e-01 6.45049989e-01 1.68652594e-01
1.04664254e+00 1.24084085e-01 5.93212247e-01 2.70879477e-01
3.26109231e-01 7.05098212e-01 5.76616228e-01 -6.83055639e-01
-5.44971585e-01 -1.57070030e-02 -3.04513741e-02 2.26897657e-01
-8.30292702e-02 -3.66799831e-02 -9.37582612e-01 5.59421003e-01
-1.85191393e+00 -7.99652994e-01 -2.32634515e-01 1.96293843e+00
9.12396073e-01 4.48645741e-01 2.48624787e-01 -8.08691382e-02
6.80711269e-01 3.01387668e-01 -8.03701103e-01 -4.15507436e-01
1.66965872e-01 8.27026129e-01 9.70084667e-01 4.63855863e-01
-1.36364257e+00 1.25346017e+00 5.62289953e+00 1.12783158e+00
-1.16747797e+00 1.67962909e-01 1.16409910e+00 -3.94589096e-01
2.57397264e-01 -1.85194373e-01 -1.17035365e+00 4.75011826e-01
6.58870757e-01 4.53357071e-01 3.30676079e-01 1.11274350e+00
3.88428289e-03 -2.49321520e-01 -1.07043433e+00 9.69963431e-01
-1.20994918e-01 -1.74291337e+00 -2.22472206e-01 9.45467949e-02
7.20729113e-01 7.47398198e-01 1.40828595e-01 4.77882773e-01
1.20126300e-01 -1.24810553e+00 6.70464993e-01 1.22433402e-01
9.56415534e-01 -6.36696100e-01 5.92523634e-01 3.27093005e-01
-1.39280236e+00 8.15728121e-03 -6.07695758e-01 -3.38595986e-01
5.76058365e-02 8.59108269e-01 -7.19614387e-01 9.45529789e-02
8.70881140e-01 7.37234592e-01 -6.08776808e-01 1.28012657e+00
-2.45513707e-01 8.05233181e-01 -5.00487208e-01 1.50594160e-01
4.46081102e-01 7.76597634e-02 1.82181180e-01 1.72126114e+00
-9.51055065e-03 7.91812316e-02 3.08638662e-01 9.12854791e-01
-3.27765793e-01 -2.38443315e-01 -1.41643569e-01 2.28466704e-01
3.98741990e-01 1.59604108e+00 -1.26245832e+00 -6.48748636e-01
-6.62521005e-01 1.05670035e+00 4.42732006e-01 -9.36659947e-02
-1.10862255e+00 -2.22328559e-01 9.90716040e-01 1.63290545e-01
7.84607470e-01 -7.60089830e-02 -8.48494530e-01 -9.14001226e-01
-7.11198673e-02 -6.81045413e-01 1.02334738e-01 -2.92494386e-01
-1.08381689e+00 4.81060654e-01 -4.25999641e-01 -8.21961999e-01
3.37490380e-01 -1.12259185e+00 -7.46449530e-01 7.26833880e-01
-1.86573911e+00 -1.17910957e+00 -5.23346841e-01 3.24624002e-01
8.76025200e-01 2.73508400e-01 5.56320846e-01 3.75027120e-01
-9.40766871e-01 7.37391770e-01 -2.57564366e-01 4.00363028e-01
3.69326383e-01 -1.50302160e+00 1.12513638e+00 9.18047845e-01
4.61986631e-01 4.27923501e-01 3.00435752e-01 -4.68184203e-01
-1.18621421e+00 -1.53965020e+00 2.93354541e-01 -5.22206724e-01
4.55533594e-01 -5.86576283e-01 -7.74150372e-01 5.91260493e-01
-1.70715638e-02 6.53891504e-01 4.27623004e-01 -4.76140566e-02
-4.27314609e-01 -1.23230860e-01 -1.05420840e+00 6.01150632e-01
1.18478143e+00 -3.65535796e-01 -1.50149971e-01 3.52030575e-01
8.68681669e-01 -9.09666479e-01 -5.88700950e-01 5.04803956e-01
5.13782322e-01 -1.18373930e+00 1.11468840e+00 -4.03755680e-02
4.44061518e-01 -4.35434788e-01 1.31821245e-01 -7.41050482e-01
-2.97691971e-01 -5.37434638e-01 -3.33012640e-01 8.54352951e-01
5.89953840e-01 -5.82717121e-01 1.22256076e+00 6.00939751e-01
-3.84886324e-01 -1.35787106e+00 -9.33348417e-01 -6.13525748e-01
-6.11613691e-02 -7.04803467e-01 6.11312091e-01 3.78751159e-01
-3.63263547e-01 1.33736089e-01 -1.93548664e-01 1.71744600e-01
5.66145480e-01 2.32546851e-01 7.75222421e-01 -9.13745165e-01
-7.30062187e-01 -7.05469966e-01 -4.68484104e-01 -1.52818823e+00
-1.89342126e-01 -8.27474415e-01 2.37087652e-01 -1.58934748e+00
1.68409258e-01 -5.71516752e-01 -3.22950244e-01 8.04119647e-01
-5.34736395e-01 1.05990219e+00 2.55415201e-01 1.20375521e-01
-6.97931290e-01 1.31080836e-01 1.23628521e+00 -1.63581073e-01
-3.86095569e-02 5.04350290e-02 -7.07176805e-01 7.29085207e-01
8.52015018e-01 -3.48354965e-01 -1.91225246e-01 -6.58846259e-01
-2.26162627e-01 -4.30995315e-01 4.80293423e-01 -1.34863186e+00
1.38351098e-01 1.52156174e-01 5.23290455e-01 -5.27526617e-01
4.96157140e-01 -5.20703673e-01 -2.63185501e-01 5.39741755e-01
-1.41049087e-01 -2.78844982e-01 5.19427001e-01 4.46587831e-01
2.19680239e-02 9.65511799e-02 9.64828074e-01 -3.80518883e-01
-9.82808053e-01 6.89750493e-01 -8.88564736e-02 9.87093598e-02
1.03173983e+00 -5.59981823e-01 -4.80549783e-01 2.53731102e-01
-3.89585495e-01 1.63741201e-01 4.04565126e-01 2.99205452e-01
4.48137879e-01 -6.56785846e-01 -6.24098778e-01 2.13652998e-01
-2.12680195e-02 6.09052539e-01 2.24762365e-01 7.52891779e-01
-6.76920116e-01 4.42470640e-01 7.75082782e-03 -9.46277976e-01
-1.07812595e+00 2.91945517e-01 3.08275223e-01 -2.88113981e-01
-1.00736964e+00 1.47528470e+00 4.62043226e-01 -1.98570624e-01
2.94325978e-01 -5.75628877e-01 2.25144655e-01 -2.16852441e-01
8.38860154e-01 2.66345412e-01 2.59993553e-01 -3.37278813e-01
-3.78924370e-01 4.93343860e-01 -3.69523376e-01 1.75001696e-01
1.32088923e+00 2.45256707e-01 7.18157515e-02 7.51335025e-02
8.82489800e-01 -2.30550438e-01 -1.88839018e+00 -1.44225478e-01
-1.57689229e-01 -4.68031704e-01 2.10041046e-01 -9.17697370e-01
-1.39228487e+00 1.09247446e+00 6.35274351e-01 -2.58961692e-02
1.04154801e+00 1.71890825e-01 1.03722811e+00 3.10106240e-02
3.12858194e-01 -9.58583295e-01 -6.17507324e-02 2.98154533e-01
3.11816037e-01 -1.49288511e+00 2.69331392e-02 -6.59270763e-01
-3.33505899e-01 7.50774384e-01 1.00153053e+00 -5.09195685e-01
3.53679359e-01 6.32736564e-01 7.32144266e-02 -2.57276773e-01
-5.80985248e-01 -5.84023833e-01 3.88982445e-01 5.51824689e-01
5.71061790e-01 3.10725600e-01 -5.38260154e-02 4.03715134e-01
-8.41418505e-02 -1.07560173e-01 2.21071199e-01 6.24180853e-01
-6.31065249e-01 -1.10386026e+00 -6.69546723e-02 8.27063382e-01
-6.42914057e-01 -3.32109898e-01 -9.82894599e-02 7.63085663e-01
4.74837095e-01 7.01739430e-01 3.42383653e-01 -2.15324402e-01
1.93538889e-01 -2.07560822e-01 5.09065032e-01 -9.76013184e-01
-8.60124826e-01 -1.64455861e-01 -1.42993137e-01 -1.09783280e+00
-2.35322058e-01 -4.87684250e-01 -1.31722152e+00 -2.25061327e-01
-3.26452464e-01 -3.88068736e-01 6.47089303e-01 9.67271388e-01
5.13139129e-01 6.03077233e-01 7.59090334e-02 -1.30869639e+00
-1.84846193e-01 -7.54728854e-01 -2.10715741e-01 1.12104565e-02
2.66230226e-01 -3.29226911e-01 6.73078150e-02 -1.74780473e-01] | [9.464942932128906, 0.09876348823308945] |
5a311d01-ac57-4bab-92bf-99dd52e4b04f | residual-network-based-aggregation-model-for | 1807.09150 | null | http://arxiv.org/abs/1807.09150v1 | http://arxiv.org/pdf/1807.09150v1.pdf | Residual Network based Aggregation Model for Skin Lesion Classification | We recognize that the skin lesion diagnosis is an essential and challenging
sub-task in Image classification, in which the Fisher vector (FV) encoding
algorithm and deep convolutional neural network (DCNN) are two of the most
successful techniques. Since the joint use of FV and DCNN has demonstrated
proven success, the joint techniques could have discriminatory power on skin
lesion diagnosis as well. To this hypothesis, we propose the aggregation
algorithm for skin lesion diagnosis that utilize the residual network to
extract the local features and the Fisher vector method to aggregate the local
features to image-level representation. We applied our algorithm on the
International Skin Imaging Collaboration 2018 (ISIC2018) challenge and only
focus on the third task, i.e., the disease classification. | ['Yongsheng Pan', 'Yong Xia'] | 2018-07-24 | null | null | null | null | ['skin-lesion-classification'] | ['medical'] | [ 5.77313483e-01 -1.87446713e-01 -3.02720726e-01 -2.21147329e-01
-9.79187906e-01 -4.66306359e-01 6.97048903e-01 -2.20420323e-02
-2.11851373e-01 4.41637963e-01 2.01863632e-01 -3.02412927e-01
-2.87989885e-01 -6.04580283e-01 -1.53382376e-01 -9.80079114e-01
6.50662854e-02 -1.90443501e-01 -5.13797253e-02 2.82468796e-01
7.02505708e-02 5.70773005e-01 -1.45089149e+00 6.76947773e-01
1.03967369e+00 1.59737599e+00 -2.84134179e-01 8.18516254e-01
-1.47313952e-01 7.84631073e-01 -3.10374856e-01 -3.17151636e-01
2.06640467e-01 -4.63973075e-01 -9.76303995e-01 -1.67163566e-01
6.74371362e-01 -2.64793396e-01 -2.34407634e-01 1.18358552e+00
5.68207264e-01 -1.67184964e-01 1.15822029e+00 -1.11003888e+00
-4.84800130e-01 9.77089927e-02 -6.75521672e-01 2.66182423e-01
1.65847823e-01 5.30498587e-02 6.91894591e-01 -7.43195176e-01
5.43849289e-01 9.38612580e-01 9.24203336e-01 3.78940046e-01
-8.14471900e-01 -2.44514257e-01 -3.28367203e-01 6.28147483e-01
-1.54306281e+00 -6.83505237e-02 4.72478092e-01 -6.40282750e-01
6.10254765e-01 5.26027441e-01 4.69044000e-01 1.23612785e+00
3.90626103e-01 8.94396067e-01 1.45440829e+00 -4.96326059e-01
8.93278271e-02 -1.74136907e-01 1.75354213e-01 7.90885746e-01
1.20204508e-01 -2.19284203e-02 -1.38056010e-01 -1.77891895e-01
8.42175841e-01 2.27604032e-01 -1.80230200e-01 3.23823899e-01
-8.05573106e-01 9.24144506e-01 6.01213038e-01 4.56177771e-01
-4.96022522e-01 2.25631565e-01 4.52810675e-01 8.24925154e-02
5.57566583e-01 1.17048388e-02 -3.57756088e-03 2.43182674e-01
-9.71944630e-01 3.99664938e-02 4.64972496e-01 -2.53268868e-01
3.50523829e-01 -3.16704661e-01 -6.30046487e-01 9.77986097e-01
3.28031510e-01 2.45036125e-01 5.49618661e-01 -8.06080222e-01
1.52385077e-02 7.64979780e-01 -4.47456867e-01 -1.00612128e+00
-4.05087262e-01 -5.72074711e-01 -1.19662511e+00 3.30052912e-01
3.70126396e-01 1.53865498e-02 -1.24108696e+00 1.42098653e+00
3.27968180e-01 4.61954176e-01 -1.76058233e-01 8.08281660e-01
9.40500021e-01 5.24302535e-02 3.61890227e-01 1.54917657e-01
1.31225288e+00 -9.18913960e-01 -7.43874431e-01 2.85020232e-01
7.98329830e-01 -6.27313614e-01 3.87297422e-01 3.63644987e-01
-8.10974300e-01 -4.36363876e-01 -1.05908799e+00 -9.75564942e-02
-6.54924691e-01 4.76305127e-01 8.50802779e-01 5.68439901e-01
-1.13877249e+00 6.86618209e-01 -8.65201116e-01 -5.41985512e-01
6.77913427e-01 1.80809706e-01 -7.99802780e-01 -3.29662681e-01
-1.14885104e+00 9.89329457e-01 1.40779773e-02 2.44261324e-01
-5.87523937e-01 -6.03048742e-01 -6.69347584e-01 -1.17041886e-01
-1.64185718e-01 -6.97823465e-01 7.06668854e-01 -1.03714621e+00
-1.10945022e+00 8.57676506e-01 1.19906310e-02 -5.88710070e-01
4.31605250e-01 3.14646028e-02 -4.92883176e-01 5.63875258e-01
-7.20720273e-03 4.61568624e-01 8.05759788e-01 -6.99189723e-01
-8.27434897e-01 -5.14135242e-01 -6.36967048e-02 -3.65866646e-02
-5.01775503e-01 -2.88957562e-02 2.94221076e-03 -4.90982264e-01
2.87183542e-02 -7.09169984e-01 -1.91124186e-01 3.27991217e-01
-5.55018187e-01 -6.79914236e-01 6.82484329e-01 -1.18463838e+00
1.47315824e+00 -2.17350292e+00 9.00641456e-02 4.04641807e-01
4.77496684e-01 4.27008778e-01 -3.91839743e-01 3.79962832e-01
-4.08027232e-01 5.28165996e-01 -1.32435903e-01 -2.49288440e-01
-1.33712262e-01 -1.11452997e-01 1.53339759e-01 6.12418413e-01
4.44221139e-01 9.67084467e-01 -6.11733735e-01 -5.78586221e-01
3.09936643e-01 8.16370547e-01 -1.93178326e-01 -5.05991951e-02
3.14216971e-01 7.58839995e-02 -3.19820166e-01 8.54380608e-01
8.36532176e-01 -4.91052479e-01 3.67518663e-02 -8.13356161e-01
1.97413445e-01 -1.97259694e-01 -9.56106961e-01 1.55460775e+00
-3.19189787e-01 4.68518138e-01 -2.97003482e-02 -8.21794450e-01
3.49360853e-01 3.27090472e-01 7.85442829e-01 -4.43841904e-01
1.85836360e-01 2.35652059e-01 -1.37647986e-02 -7.66372025e-01
-2.82872081e-01 1.42466664e-01 2.41380900e-01 9.85303149e-02
2.65502125e-01 3.62842143e-01 1.25833094e-01 8.41577575e-02
1.50715387e+00 -8.15597828e-03 5.69349289e-01 -5.97388670e-02
5.92008114e-01 -2.00734645e-01 2.29411140e-01 4.41108465e-01
-4.89373714e-01 6.70755029e-01 5.49732029e-01 -3.13352793e-01
-5.56423545e-01 -1.09962153e+00 -5.29459834e-01 7.55728185e-01
-1.34802058e-01 -2.89978325e-01 -9.60870206e-01 -1.21304488e+00
1.66713834e-01 3.38079259e-02 -1.12973785e+00 -2.46274859e-01
-1.26949400e-01 -8.10582340e-01 7.78838634e-01 7.00844586e-01
5.29407740e-01 -5.86189151e-01 -1.81771636e-01 -2.48887226e-01
-1.55583218e-01 -7.55332708e-01 -2.12738439e-01 1.77108437e-01
-1.65985569e-01 -1.47528636e+00 -9.77410734e-01 -7.57147491e-01
5.99083781e-01 1.57123938e-01 4.45861965e-01 1.43006533e-01
-9.70143318e-01 2.98425645e-01 -4.78318065e-01 3.37442197e-02
-1.64652675e-01 -1.25438571e-01 -1.88987732e-01 4.97073233e-01
2.26224840e-01 -3.59951586e-01 -9.69749212e-01 -1.81643859e-01
-1.06097806e+00 -1.15211889e-01 8.89544964e-01 1.01801932e+00
5.53003311e-01 4.76530381e-02 4.88907576e-01 -7.85634995e-01
6.63110733e-01 -6.19080544e-01 2.08450437e-01 5.09594023e-01
-2.65886903e-01 -2.54249841e-01 3.75389695e-01 -1.59504101e-01
-8.39936435e-01 1.60646245e-01 -4.18646127e-01 -4.40406144e-01
-2.28481948e-01 8.29574764e-01 2.07318664e-01 -5.52564442e-01
7.65425563e-01 -7.09681818e-03 6.70416579e-02 -4.89313185e-01
1.57033727e-01 8.23981822e-01 4.35264289e-01 7.68976659e-02
4.82807398e-01 3.61686349e-01 4.92804378e-01 -5.74307680e-01
-9.31167901e-01 -5.60364902e-01 -6.07480526e-01 -2.86733627e-01
1.33485079e+00 -6.62608147e-01 -5.53041875e-01 7.13066399e-01
-1.03718269e+00 1.54125551e-02 -5.49356453e-02 2.61039168e-01
-2.03430131e-01 6.68829799e-01 -5.56455612e-01 -7.57561684e-01
-4.54856873e-01 -1.02231586e+00 1.32996881e+00 4.95185703e-01
-1.24253772e-01 -1.30640006e+00 1.58545315e-01 2.93817282e-01
5.92780232e-01 8.00133586e-01 9.85361278e-01 -6.74049854e-01
-2.30435189e-02 -5.34145951e-01 -6.29028141e-01 7.17309535e-01
3.61787617e-01 2.62633711e-01 -1.15888143e+00 -2.27607831e-01
-3.17148417e-01 -3.76017511e-01 1.47571814e+00 4.61843312e-01
1.56349409e+00 -8.58197957e-02 -2.99095035e-01 8.88675511e-01
1.69215870e+00 -1.17331386e-01 6.83300793e-01 -2.61257589e-01
6.74460411e-01 5.59606314e-01 2.06955418e-01 1.37904137e-01
3.82912964e-01 5.12091219e-01 5.08328080e-01 -3.99722725e-01
-7.97931075e-01 -8.46143067e-02 1.31258979e-01 7.54496813e-01
-3.56669217e-01 3.49830091e-02 -9.13564742e-01 4.44593370e-01
-1.70627332e+00 -8.89931500e-01 -2.23217636e-01 1.95025897e+00
5.38987100e-01 -3.22684199e-01 -5.69732375e-02 2.64889181e-01
7.14878321e-01 1.54756457e-01 -3.44864100e-01 -3.22858185e-01
-1.71775088e-01 6.33009076e-01 3.44623089e-01 2.48167381e-01
-1.50533116e+00 3.99671435e-01 6.67721510e+00 1.23749685e+00
-1.42029524e+00 1.87881276e-01 8.23899627e-01 3.15923870e-01
-3.47370841e-02 -3.57483774e-01 -4.22703922e-01 6.64535403e-01
8.31641078e-01 3.02013785e-01 2.24944711e-01 5.32024384e-01
-2.83984601e-01 -2.48820215e-01 -8.20143521e-01 1.06670535e+00
3.81182730e-01 -1.30323803e+00 6.89364299e-02 1.96889013e-01
6.82336986e-01 -1.29985169e-01 3.98777574e-01 4.67963591e-02
-4.19347845e-02 -1.54671478e+00 -1.19795009e-01 1.01502764e+00
1.21294796e+00 -4.87573594e-01 1.11756468e+00 -7.71449655e-02
-1.18495750e+00 -1.26264364e-01 -2.02623412e-01 4.87619400e-01
-1.31164759e-01 7.52570927e-01 -6.96612239e-01 7.16595590e-01
6.43006206e-01 7.69762635e-01 -9.83723819e-01 1.22270799e+00
-1.18222900e-01 6.48540258e-01 -1.76059499e-01 7.22512901e-02
1.21583432e-01 -1.12827122e-01 2.30956241e-01 1.21903300e+00
3.61660808e-01 -3.26363534e-01 8.06292519e-02 5.21644652e-01
3.03317845e-01 2.00288743e-01 -4.42192644e-01 -2.11292416e-01
-1.62005685e-02 1.78077650e+00 -4.58571672e-01 -1.57307506e-01
-2.93684602e-01 1.16114521e+00 2.83009410e-01 3.64276677e-01
-6.57759607e-01 -5.61168134e-01 7.02582479e-01 -1.49555162e-01
9.11321938e-02 2.82824636e-01 -3.27055335e-01 -1.10262370e+00
-1.03637993e-01 -6.62040412e-01 5.36268950e-01 -5.31675935e-01
-1.77677822e+00 5.84340990e-01 -5.90992928e-01 -1.14261580e+00
-3.11492503e-01 -1.04140425e+00 -9.83476460e-01 9.52258706e-01
-1.71229196e+00 -1.60742426e+00 -5.87802827e-01 7.45655239e-01
4.75668609e-02 -1.13861606e-01 1.11284339e+00 2.76380539e-01
-7.49798000e-01 7.74290681e-01 1.41618788e-01 2.96239227e-01
7.40504861e-01 -1.37049770e+00 -2.29006916e-01 6.15220189e-01
-1.56163245e-01 5.15519023e-01 9.80385691e-02 -5.74446201e-01
-1.19245172e+00 -1.23667896e+00 7.83350706e-01 -2.44335964e-01
6.63649678e-01 -5.14632612e-02 -6.07909679e-01 5.35634160e-02
2.06508562e-01 4.55460608e-01 1.13545096e+00 -9.59642604e-02
-5.46107411e-01 -6.86952770e-02 -1.38807964e+00 3.52002621e-01
8.46866965e-01 -8.18883359e-01 -7.23640844e-02 5.23236275e-01
2.48053417e-01 7.35539943e-02 -1.27560949e+00 6.52199447e-01
7.60370612e-01 -8.75472844e-01 9.67483103e-01 -7.34058022e-01
7.84306288e-01 1.27306907e-02 -4.41863507e-01 -1.30615926e+00
-5.87536275e-01 1.05472980e-02 -3.63989398e-02 1.13323188e+00
9.65758413e-02 -6.63538218e-01 6.72935605e-01 1.47611931e-01
1.47704750e-01 -1.19782567e+00 -1.14621401e+00 -3.61862659e-01
1.94392875e-01 -9.32536945e-02 2.42860928e-01 8.53973091e-01
-1.94784790e-01 -1.36029109e-01 -2.34395713e-01 3.72870602e-02
6.11618459e-01 -2.12402195e-01 1.27432168e-01 -1.17636251e+00
-1.60317257e-01 -6.18916273e-01 -7.96328187e-01 -2.01779589e-01
5.99120893e-02 -1.32087028e+00 -2.64384747e-01 -1.82829762e+00
4.29303408e-01 3.41758039e-03 -8.95041645e-01 6.08041286e-01
-4.08543050e-01 6.18059576e-01 1.02015905e-01 -7.57514387e-02
-4.64173466e-01 -6.09932654e-03 1.26060581e+00 -3.87274384e-01
4.12113428e-01 -6.74831420e-02 -7.54136384e-01 6.20633006e-01
6.08157277e-01 3.41114625e-02 4.64592362e-03 -1.62844166e-01
-1.33353323e-02 -1.49825603e-01 7.35474646e-01 -1.05364954e+00
3.20048571e-01 -1.44527704e-01 8.74339342e-01 -3.27404916e-01
3.13591093e-01 -5.37040412e-01 -1.40172586e-01 6.60801470e-01
-4.28234994e-01 -3.84422690e-01 -1.67334765e-01 3.87792617e-01
-5.06819844e-01 -2.10692629e-01 8.11463892e-01 7.07407668e-02
-6.64023280e-01 5.69717228e-01 -2.28000715e-01 -5.54657459e-01
1.08379018e+00 -7.17630163e-02 -5.32551587e-01 -1.59670785e-01
-7.88686693e-01 -1.02488466e-01 -4.07838710e-02 1.21403903e-01
6.49512172e-01 -1.65150023e+00 -9.11006927e-01 3.21919739e-01
5.90282716e-02 -5.21390378e-01 6.93172872e-01 1.28673422e+00
-5.38687825e-01 3.95135134e-01 -2.65006870e-01 -6.81709349e-01
-1.20700049e+00 2.19385013e-01 4.85465854e-01 -5.94116688e-01
-1.94515169e-01 1.01352942e+00 2.29273111e-01 -9.61293131e-02
1.80043250e-01 -3.80797237e-02 -6.71404183e-01 2.08190754e-01
8.37107182e-01 4.24518049e-01 1.42598301e-01 -4.85492796e-01
-4.67550933e-01 8.05430353e-01 -1.06158994e-01 2.68342435e-01
1.16556633e+00 1.36064187e-01 -3.80589366e-01 -1.24160953e-01
1.73098552e+00 -3.20819378e-01 -8.66526008e-01 -7.48790801e-03
-1.56899706e-01 -3.87724817e-01 4.42989022e-01 -1.21543729e+00
-1.09709096e+00 1.16852534e+00 1.22501981e+00 3.68143201e-01
1.34766257e+00 -1.08090810e-01 7.63564408e-01 -5.72767816e-02
5.12615927e-02 -9.25847530e-01 -9.86287966e-02 1.73981816e-01
8.84955108e-01 -1.13482523e+00 -1.29353523e-01 -6.22748852e-01
-4.49973494e-01 1.35512745e+00 3.72372150e-01 -3.84717077e-01
9.36665416e-01 3.55060846e-01 1.37197301e-01 1.13463752e-01
-6.80071592e-01 -5.64752162e-01 7.98795581e-01 6.96449459e-01
4.06127512e-01 2.58562416e-01 -5.45250416e-01 8.54975343e-01
4.45829272e-01 8.08851123e-02 -9.14485976e-02 5.10450542e-01
-1.70793563e-01 -1.11516082e+00 -6.00264855e-02 9.90458488e-01
-7.40162849e-01 3.24190482e-02 -7.55099237e-01 4.19974238e-01
4.95178580e-01 7.02466726e-01 -8.00408199e-02 -7.78349996e-01
-1.56770006e-01 7.88136721e-02 6.03323102e-01 -2.68084407e-01
-6.86634243e-01 -5.82606392e-03 5.44498153e-02 -8.29128623e-01
-3.94029170e-01 -4.52933401e-01 -8.98348451e-01 -7.97810964e-03
-2.75562465e-01 -2.90857494e-01 9.00349200e-01 1.02056396e+00
2.77573556e-01 6.05544031e-01 8.26655209e-01 -4.95428830e-01
-7.00811923e-01 -9.26556408e-01 -7.86674261e-01 6.03286266e-01
3.51078302e-01 -6.79256082e-01 -4.58547741e-01 -1.72329381e-01] | [15.63988971710205, -2.9378061294555664] |
2118ef8f-4e2f-4dc9-8a96-fd15a81565d3 | event-extraction-in-video-transcripts | null | null | https://aclanthology.org/2022.coling-1.625 | https://aclanthology.org/2022.coling-1.625.pdf | Event Extraction in Video Transcripts | Event extraction (EE) is one of the fundamental tasks for information extraction whose goal is to identify mentions of events and their participants in text. Due to its importance, different methods and datasets have been introduced for EE. However, existing EE datasets are limited to formally written documents such as news articles or scientific papers. As such, the challenges of EE in informal and noisy texts are not adequately studied. In particular, video transcripts constitute an important domain that can benefit tremendously from EE systems (e.g., video retrieval), but has not been studied in EE literature due to the lack of necessary datasets. To address this limitation, we propose the first large-scale EE dataset obtained for transcripts of streamed videos on the video hosting platform Behance to promote future research in this area. In addition, we extensively evaluate existing state-of-the-art EE methods on our new dataset. We demonstrate that such systems cannot achieve adequate performance on the proposed dataset, revealing challenges and opportunities for further research effort. | ['Thien Huu Nguyen', 'Franck Dernoncourt', 'Viet Dac Lai', 'Amir Pouran Ben Veyseh'] | null | null | null | null | coling-2022-10 | ['event-extraction'] | ['natural-language-processing'] | [ 2.68596619e-01 -1.42807141e-01 -1.28916547e-01 -2.58253783e-01
-9.84651446e-01 -6.74152732e-01 7.28329480e-01 2.91607887e-01
-4.37346995e-01 7.77540267e-01 5.20613313e-01 -1.53973162e-01
9.48054194e-02 -5.55270076e-01 -6.29106104e-01 -3.96800131e-01
-1.51837006e-01 -1.87288150e-01 3.85214835e-01 6.54014051e-02
1.42887071e-01 2.88184762e-01 -1.68216050e+00 3.85527164e-01
4.41193372e-01 1.07260036e+00 -1.80125162e-02 4.03032541e-01
-7.39506707e-02 1.05159295e+00 -7.71544814e-01 -3.58838141e-01
-2.37567481e-02 -5.78579128e-01 -6.84197247e-01 3.88965681e-02
1.24014448e-02 -2.80442804e-01 -6.74683928e-01 8.73094499e-01
4.66824085e-01 9.40619111e-02 5.27595699e-01 -1.40412474e+00
-1.13284953e-01 7.99268365e-01 -3.50608855e-01 5.41395783e-01
5.03233910e-01 -2.40634128e-01 1.10924423e+00 -9.30061400e-01
9.56707776e-01 7.67218709e-01 2.68396080e-01 2.90795892e-01
-5.96223056e-01 -8.31756294e-01 8.98132697e-02 4.06429112e-01
-1.36216950e+00 -8.15985799e-01 9.69958842e-01 -2.15485215e-01
8.60482752e-01 5.26839972e-01 5.17874658e-01 1.55546045e+00
-2.55039722e-01 1.32037377e+00 8.78457308e-01 -3.70175421e-01
2.49055281e-01 9.86483544e-02 1.26284242e-01 2.56560713e-01
1.25276625e-01 -3.98289859e-01 -9.32224154e-01 -9.94144902e-02
3.64151806e-01 -1.57872543e-01 -5.59701145e-01 1.85871527e-01
-1.20266390e+00 4.03781563e-01 -2.90250242e-01 4.69954044e-01
-5.21924436e-01 -1.71429753e-01 7.75205255e-01 2.94388175e-01
5.19932091e-01 3.46877247e-01 -4.70303953e-01 -8.36580276e-01
-1.19659853e+00 3.23372871e-01 1.08740258e+00 1.14343536e+00
1.97784558e-01 -1.84806421e-01 -1.48695350e-01 5.09536207e-01
8.89602900e-02 5.23832850e-02 1.39677867e-01 -7.32594430e-01
8.58386636e-01 5.67611217e-01 1.23156019e-01 -1.32600939e+00
-2.19225362e-01 -3.28021586e-01 -7.37206638e-01 -6.03734910e-01
2.72759616e-01 -1.92115605e-01 -2.11182058e-01 1.51697171e+00
2.18020141e-01 5.15498221e-01 5.78315258e-02 8.71130288e-01
1.40260053e+00 8.38637650e-01 5.32618091e-02 -6.47094965e-01
1.52274513e+00 -7.45627880e-01 -1.06570184e+00 -1.09002873e-01
2.47761160e-01 -8.38181496e-01 7.02585697e-01 5.35234153e-01
-1.07056808e+00 -4.25074175e-02 -8.25801551e-01 6.15326688e-02
-3.22309107e-01 2.61132181e-01 7.04225123e-01 2.27258399e-01
-4.82977688e-01 2.43173465e-01 -9.19889987e-01 -5.79675257e-01
5.72227716e-01 -1.44907301e-02 -1.88302308e-01 7.03202412e-02
-1.40929556e+00 3.23320210e-01 1.21584333e-01 1.62792519e-01
-9.07662630e-01 -5.81142306e-01 -6.40781343e-01 2.12567285e-01
1.05471551e+00 -4.93280441e-02 1.50597501e+00 -5.72669744e-01
-1.21629620e+00 6.43426597e-01 -3.19399327e-01 -5.18129528e-01
4.34101641e-01 -4.15334463e-01 -7.72665858e-01 3.72876972e-01
1.66252151e-01 1.48330480e-01 6.69720352e-01 -9.32318211e-01
-8.45946848e-01 -2.05879241e-01 2.89312512e-01 8.10197461e-03
-8.34432423e-01 6.21186256e-01 -9.68780100e-01 -9.12632465e-01
-2.32426405e-01 -8.73914957e-01 4.52570617e-02 -3.29760462e-01
-6.76006317e-01 -3.00372332e-01 9.49981928e-01 -5.92803240e-01
2.01991272e+00 -2.29447842e+00 -4.21486236e-02 -4.28680368e-02
3.85790914e-01 9.58194435e-02 1.12550281e-01 8.27646017e-01
1.64708123e-01 3.89023662e-01 1.05519705e-01 -2.01369852e-01
1.39334928e-02 -7.27507621e-02 -4.51679021e-01 2.84752011e-01
7.41683841e-02 6.38089776e-01 -1.01653636e+00 -7.45757639e-01
-1.18623935e-01 3.20053577e-01 -2.87139475e-01 3.19148988e-01
-1.59618795e-01 3.12334150e-01 -9.45630491e-01 8.21073830e-01
2.36460701e-01 -3.76779228e-01 1.82430804e-01 -4.48893011e-01
-3.14132869e-01 5.89610219e-01 -1.32769239e+00 1.48285389e+00
-2.24693820e-01 1.09467912e+00 8.71630609e-02 -1.08303094e+00
3.99882197e-01 7.56727219e-01 8.85244846e-01 -2.64952093e-01
2.76122928e-01 1.75301135e-02 -3.56589943e-01 -7.67820299e-01
7.07870603e-01 3.32251728e-01 -3.49346876e-01 3.93236399e-01
-1.18198931e-01 2.68920392e-01 6.57288194e-01 4.17964697e-01
1.43137527e+00 4.35532369e-02 5.51699817e-01 2.18554989e-01
3.89588654e-01 1.50748178e-01 7.66012192e-01 6.07495010e-01
-3.43667597e-01 6.63628280e-01 5.56133747e-01 2.78105936e-03
-6.55564070e-01 -5.58179855e-01 -1.15875579e-01 8.68383646e-01
4.68943231e-02 -1.01105416e+00 -8.45013738e-01 -7.35358417e-01
-5.72772682e-01 3.61904263e-01 -2.96120048e-01 3.54351431e-01
-5.87387860e-01 -6.31672740e-01 8.15239131e-01 4.21906024e-01
5.30701399e-01 -1.05018401e+00 -7.76857853e-01 4.74832028e-01
-8.47172618e-01 -1.62065208e+00 -4.18451220e-01 2.56300047e-02
-5.98435640e-01 -1.41910481e+00 -4.57180142e-01 -5.74968576e-01
3.20737004e-01 2.76013851e-01 1.30308211e+00 -8.97078365e-02
-1.38803452e-01 4.18316483e-01 -7.68450141e-01 -5.73973775e-01
-2.63886839e-01 2.53757954e-01 -6.68249205e-02 4.80597876e-02
7.84613788e-01 -5.21728694e-01 -5.41254640e-01 4.85073596e-01
-1.13838637e+00 -4.06045392e-02 3.20462108e-01 4.45752770e-01
4.81161267e-01 3.87310982e-01 7.31112003e-01 -9.34173286e-01
7.55319774e-01 -7.18352079e-01 -4.13899034e-01 1.04335956e-01
-2.53378600e-01 -2.86107421e-01 5.74743569e-01 -4.23377097e-01
-1.15046990e+00 -3.23105693e-01 3.51966619e-02 -9.80412960e-02
-3.55196953e-01 8.11582088e-01 -2.44284153e-01 4.83157396e-01
2.62329072e-01 2.12419376e-01 -7.46925116e-01 -5.04441857e-01
2.87707523e-03 1.03451300e+00 5.28409719e-01 -5.74708819e-01
4.42255884e-01 4.40795153e-01 -1.47957101e-01 -9.31104302e-01
-9.50458467e-01 -7.18960881e-01 -1.93700820e-01 -3.69025916e-01
6.46312892e-01 -1.07642806e+00 -5.76672733e-01 1.28594846e-01
-1.19709694e+00 2.88105547e-01 -6.86884522e-02 6.53475702e-01
-2.55371064e-01 4.20137554e-01 -8.07850361e-01 -8.19406688e-01
-2.17203170e-01 -1.05371261e+00 1.01803005e+00 1.89650640e-01
-4.59348977e-01 -4.85833436e-01 -4.18808572e-02 3.77028227e-01
9.08124000e-02 4.64609623e-01 4.39287871e-01 -6.89992189e-01
-6.93958640e-01 -2.38445267e-01 2.73999962e-04 8.41261260e-03
8.60502794e-02 2.38823086e-01 -9.07106638e-01 -1.47384945e-02
1.46825105e-01 -3.69007170e-01 7.78176546e-01 2.63649076e-01
1.26467884e+00 -2.75395900e-01 -5.16561031e-01 1.95752904e-01
1.03469884e+00 3.99243921e-01 6.70670390e-01 5.93324006e-01
3.48951697e-01 4.93984967e-01 8.27892303e-01 9.59701240e-01
4.88043308e-01 6.05019212e-01 1.87644318e-01 1.78300351e-01
-1.05351070e-03 -2.23128572e-01 4.63631153e-01 1.08297515e+00
-2.31281340e-01 -8.31763506e-01 -7.96301961e-01 8.48421276e-01
-1.90311980e+00 -1.15059721e+00 -3.43386739e-01 1.67515862e+00
7.74105370e-01 1.53413773e-01 1.15687056e-02 4.70967293e-01
5.66062689e-01 4.33336943e-01 -2.49836564e-01 1.64769366e-01
-7.86343664e-02 -9.16457921e-02 1.25686601e-01 -5.13142347e-01
-1.40426469e+00 6.81221485e-01 5.34594631e+00 9.03560162e-01
-8.53021801e-01 5.06964512e-02 4.40148622e-01 -2.20140025e-01
6.73552603e-02 -4.19544093e-02 -8.25686395e-01 5.86593866e-01
1.12773871e+00 -1.72349826e-01 2.25824997e-01 6.94092393e-01
5.39378166e-01 -8.79856274e-02 -1.34687138e+00 1.18731022e+00
3.86195295e-02 -1.26363647e+00 -3.00994575e-01 1.61026989e-03
5.66345274e-01 -1.31142706e-01 -2.66095996e-01 4.35041100e-01
-2.35115021e-01 -6.15812838e-01 7.61617422e-01 2.64879286e-01
6.19760692e-01 -6.37327313e-01 7.97656536e-01 3.44182432e-01
-1.45581639e+00 1.60391062e-01 -1.01518936e-01 -1.00117764e-02
5.21992683e-01 9.15812373e-01 -5.41561306e-01 7.21626341e-01
9.62227702e-01 9.44486260e-01 -2.66577601e-01 9.66793358e-01
-2.50924289e-01 1.25330007e+00 -4.61128175e-01 -1.66375920e-01
1.81691304e-01 1.54271573e-01 6.31083608e-01 1.39208663e+00
4.49463338e-01 5.50062239e-01 1.61082834e-01 5.13949573e-01
-6.59963548e-01 3.09080124e-01 -6.35051370e-01 -6.14782989e-01
5.55458188e-01 1.25606966e+00 -8.59970629e-01 -3.18742573e-01
-8.70739281e-01 6.75223827e-01 2.86962818e-02 4.69236374e-01
-9.87720847e-01 -4.20819998e-01 5.47069192e-01 1.42101616e-01
3.04523528e-01 -1.50335982e-01 1.55803666e-01 -1.59983468e+00
2.15421155e-01 -1.05834627e+00 4.63442922e-01 -4.67040181e-01
-1.13890159e+00 7.03818858e-01 5.92387095e-02 -1.52899373e+00
-3.30367297e-01 -1.38550594e-01 -3.20785373e-01 1.00611642e-01
-1.40744376e+00 -6.84090018e-01 -3.24933141e-01 4.90490675e-01
8.39854658e-01 -8.61119255e-02 5.83688259e-01 8.39381933e-01
-7.63082623e-01 2.51084119e-01 3.50958034e-02 3.88240546e-01
9.45201039e-01 -8.29956949e-01 2.32107222e-01 9.72063601e-01
4.89318073e-01 4.60840642e-01 7.42083430e-01 -5.97114086e-01
-1.82653165e+00 -1.05888474e+00 1.19433081e+00 -4.82353479e-01
7.63020158e-01 -4.97712761e-01 -6.82167828e-01 5.88100135e-01
2.51464009e-01 -9.25029628e-03 8.36269677e-01 2.37363726e-02
3.23840491e-02 -6.35045674e-03 -8.37259650e-01 7.03563750e-01
1.14129496e+00 -6.04456127e-01 -5.45188010e-01 2.15581372e-01
5.59765637e-01 -4.49291587e-01 -9.96307254e-01 3.09196740e-01
3.99780452e-01 -6.76743448e-01 7.06243515e-01 -5.66654027e-01
5.72716236e-01 -2.10459203e-01 -1.84998944e-01 -8.53379309e-01
3.71674180e-01 -8.16240370e-01 -6.12529993e-01 1.88814449e+00
3.94267142e-01 -8.85265507e-03 7.48486936e-01 6.17808938e-01
3.38150673e-02 -6.93054080e-01 -7.20615983e-01 -5.86647272e-01
-7.49055207e-01 -1.03884196e+00 5.74585617e-01 9.66701806e-01
1.64517179e-01 4.05549288e-01 -5.56063950e-01 1.90466955e-01
3.57688934e-01 2.41647065e-01 6.82865620e-01 -1.05224466e+00
-6.77621439e-02 -1.86589092e-01 -2.26582035e-01 -1.16088498e+00
1.55814141e-01 -5.55642664e-01 3.15748677e-02 -1.47008860e+00
5.66244423e-01 -1.70127422e-01 -1.57201648e-01 2.14951932e-01
-2.38545351e-02 1.37431398e-01 1.32812321e-01 2.43754357e-01
-1.03129148e+00 5.43340862e-01 8.25639784e-01 -8.47019404e-02
-1.88581869e-01 -6.95724711e-02 -6.88825071e-01 8.65788877e-01
6.72102690e-01 -7.48581767e-01 -4.58601177e-01 -1.76093832e-01
3.04542065e-01 3.67357314e-01 1.31420478e-01 -8.75704706e-01
3.89902085e-01 -3.17201942e-01 -6.66035339e-02 -7.24364221e-01
1.23204172e-01 -9.52189445e-01 1.77181423e-01 -3.06945920e-01
-2.69060999e-01 9.65010822e-02 4.46007401e-02 7.82329857e-01
-8.83486807e-01 -1.02569185e-01 -3.27135623e-03 -1.12870105e-01
-7.78877616e-01 4.14761752e-01 -6.04488671e-01 4.95024860e-01
9.82025504e-01 5.51528111e-02 -2.25261241e-01 -5.91051102e-01
-4.35917109e-01 3.06562930e-01 1.74612269e-01 5.20398080e-01
6.32898211e-01 -1.12495744e+00 -7.37461448e-01 -2.58483171e-01
1.67664126e-01 -1.12374216e-01 -4.83381236e-03 8.18160594e-01
-1.05215780e-01 5.39839506e-01 3.06539476e-01 -3.61178190e-01
-1.46279597e+00 3.88515681e-01 -3.62689912e-01 -4.28434074e-01
-6.75529122e-01 3.76793087e-01 -6.75647520e-03 2.18308926e-01
6.32688761e-01 -5.31653881e-01 -5.09529054e-01 2.77570337e-01
7.56601214e-01 4.25266027e-01 1.51892707e-01 -5.55421054e-01
-5.20248652e-01 1.24315538e-01 6.08803071e-02 -5.64096794e-02
1.66763914e+00 -3.40866148e-01 8.23673680e-02 4.15510833e-01
1.11206424e+00 2.02142760e-01 -8.51343513e-01 -1.20405339e-01
2.14192376e-01 -3.30630034e-01 1.88171566e-01 -4.62950855e-01
-1.03595114e+00 7.16174960e-01 1.01744860e-01 3.54457110e-01
1.44038165e+00 5.42674139e-02 1.11326385e+00 5.46406329e-01
5.78161299e-01 -1.19931901e+00 -4.16299738e-02 3.85406762e-01
7.01292276e-01 -1.25264311e+00 2.35705599e-02 -6.06759250e-01
-4.75956202e-01 1.09410334e+00 3.78240108e-01 3.89503598e-01
7.21421182e-01 5.54675221e-01 -1.69489697e-01 -3.28449637e-01
-8.39928567e-01 -1.63789168e-01 3.31458569e-01 1.01661846e-01
6.16167903e-01 -2.08055839e-01 -6.50659084e-01 1.17566812e+00
1.43965751e-01 3.50856990e-01 6.60765409e-01 1.21770656e+00
-5.87588688e-03 -1.00435293e+00 -1.74098492e-01 5.44857502e-01
-1.20111597e+00 -6.52839988e-02 -3.15821320e-01 7.89372981e-01
-1.30508646e-01 1.24567568e+00 -3.12037379e-01 -1.64126486e-01
4.89989877e-01 -9.16445106e-02 1.73092753e-01 -4.42552119e-01
-6.37907445e-01 1.36205196e-01 5.20878673e-01 -5.62538922e-01
-9.22825038e-01 -7.59393394e-01 -1.26041722e+00 -2.02247471e-01
-2.84601927e-01 3.91223043e-01 6.03631020e-01 8.96394432e-01
3.66909981e-01 6.43314362e-01 4.22860801e-01 -5.36734700e-01
-1.07873209e-01 -7.65650630e-01 -5.06701708e-01 4.81194288e-01
1.01840356e-02 -5.86130679e-01 -3.49945992e-01 3.76350671e-01] | [8.973428726196289, 9.067146301269531] |
5dbc2289-5f51-48c6-b042-fcd8bddab8b7 | order-flow-and-price-formation | 2105.00521 | null | https://arxiv.org/abs/2105.00521v1 | https://arxiv.org/pdf/2105.00521v1.pdf | Order flow and price formation | I present an overview of some recent advancements on the empirical analysis and theoretical modeling of the process of price formation in financial markets as the result of the arrival of orders in a limit order book exchange. After discussing critically the possible modeling approaches and the observed stylized facts of order flow, I consider in detail market impact and transaction cost of trades executed incrementally over an extended period of time, by comparing model predictions and recent extensive empirical results. I also discuss how the simultaneous presence of many algorithmic trading executions affects the quality and cost of trading. | ['Fabrizio Lillo'] | 2021-05-02 | null | null | null | null | ['algorithmic-trading'] | ['time-series'] | [-4.33211654e-01 -1.77161172e-01 -2.69025743e-01 -2.64558971e-01
-1.60268545e-01 -1.14678884e+00 8.09229374e-01 3.98627788e-01
-3.12175155e-01 5.42563975e-01 7.95169175e-02 -7.55247235e-01
-2.47085586e-01 -7.86719918e-01 -6.49191558e-01 -1.91527400e-02
-5.16320467e-01 9.86902535e-01 2.79610425e-01 -1.81988969e-01
8.57367694e-01 5.47433734e-01 -7.08591402e-01 5.59237450e-02
3.41192365e-01 1.23123229e+00 -5.50274014e-01 4.38244164e-01
-1.02180935e-01 9.71517324e-01 -6.97675228e-01 -9.00667191e-01
1.07695389e+00 -2.57578582e-01 -2.16291919e-01 4.87442642e-01
-3.38494152e-01 -9.62576866e-01 -2.17030928e-01 6.79327071e-01
-2.08318636e-01 9.42851603e-02 5.62692523e-01 -1.20263398e+00
-5.18758297e-01 9.69689608e-01 -4.14498746e-01 9.66986120e-01
-1.19291693e-01 2.89515108e-01 1.43742669e+00 -4.96150941e-01
4.41792578e-01 6.55575573e-01 1.89505264e-01 -1.77034736e-01
-1.27248931e+00 -4.77961302e-01 1.19833775e-01 -3.41134489e-01
-6.77442670e-01 -1.37697533e-01 5.37496805e-01 -5.01375139e-01
1.08289635e+00 3.18388075e-01 1.05408096e+00 4.78345811e-01
9.37772870e-01 7.71284997e-01 1.11826217e+00 -2.50013083e-01
3.59598666e-01 -5.40635921e-03 4.89828229e-01 -4.88804996e-01
8.69293094e-01 5.82355380e-01 -5.74699759e-01 -7.26840615e-01
1.32490170e+00 -3.45275328e-02 2.95984894e-01 -1.20878406e-01
-6.36693180e-01 7.88958728e-01 -1.11944012e-01 2.92624593e-01
-7.45717227e-01 -5.78783378e-02 2.33137041e-01 9.53744411e-01
6.04640424e-01 6.86138809e-01 -9.73385632e-01 -4.60385293e-01
-1.11505961e+00 7.36727357e-01 1.40820885e+00 7.04409480e-01
5.82978614e-02 6.08255193e-02 1.44110367e-01 -1.94788292e-01
2.32585207e-01 2.17372075e-01 2.23568529e-01 -9.75158274e-01
5.21772742e-01 2.30122522e-01 6.69054866e-01 -4.15937543e-01
-1.35583892e-01 -5.69721103e-01 8.79428461e-02 1.81370929e-01
5.62470078e-01 -4.10271227e-01 2.52718404e-02 5.99626601e-01
-3.68975699e-01 -1.63426548e-01 -1.65283218e-01 2.98546940e-01
-5.42054415e-01 4.59811181e-01 -4.37438995e-01 -8.16415489e-01
1.13878334e+00 -8.36131215e-01 -7.68340528e-01 1.95223764e-01
1.19086131e-01 -9.77362096e-01 4.70464975e-01 5.18782377e-01
-1.58390605e+00 6.03637174e-02 -6.84426963e-01 5.03042281e-01
1.75804511e-01 -6.79589510e-01 8.39816689e-01 5.60095727e-01
-6.60737991e-01 9.05925691e-01 -9.96639788e-01 3.92930597e-01
8.86210650e-02 3.71307135e-01 6.66057408e-01 8.27511907e-01
-6.81627274e-01 7.31741846e-01 -1.49339037e-02 1.89180672e-01
-3.60411614e-01 -1.02904367e+00 -2.15480775e-02 3.84344578e-01
7.32250094e-01 -2.70596087e-01 1.89232767e+00 -6.86932445e-01
-1.40875494e+00 2.13810205e-01 2.40159631e-01 -1.04111922e+00
1.07013786e+00 -5.42028606e-01 -4.91838872e-01 -1.30428061e-01
-2.58791715e-01 -3.90768766e-01 7.61719644e-02 -7.54000187e-01
-8.40424776e-01 -3.84627670e-01 -6.67760968e-02 1.70762107e-01
3.04298103e-01 2.88694859e-01 3.96893829e-01 -1.06258404e+00
-2.31467816e-03 -6.94169581e-01 -2.95287430e-01 -8.57926905e-01
-5.12645394e-02 3.94620560e-02 -2.06332635e-02 -5.69241583e-01
1.17712486e+00 -1.79989111e+00 -7.22174823e-01 6.05556428e-01
-1.32968593e-02 -4.12597805e-01 4.53783065e-01 1.14171493e+00
6.28818572e-02 5.25786042e-01 2.80548334e-01 1.65394858e-01
4.85608876e-01 -2.73140054e-02 -8.95899177e-01 2.44291008e-01
-1.85891062e-01 1.29077148e+00 -3.52419883e-01 3.57917041e-01
6.08883472e-03 -3.81524861e-01 -4.37507570e-01 1.53719887e-01
-3.73899549e-01 2.99980104e-01 -3.11005145e-01 6.09823167e-01
5.30776620e-01 -2.72310346e-01 4.57114756e-01 3.70155096e-01
-6.25616074e-01 8.87245595e-01 -8.88837516e-01 2.11079434e-01
1.13548487e-01 3.77075285e-01 -9.53394398e-02 -3.21214795e-01
4.20096427e-01 3.38305593e-01 3.61805677e-01 -9.24535930e-01
2.23983943e-01 4.26230341e-01 4.85150546e-01 1.66838542e-01
4.18568581e-01 -5.16416311e-01 1.18133560e-01 1.29472888e+00
-5.21966100e-01 8.80128145e-02 6.01001918e-01 -6.73860237e-02
7.55308807e-01 -2.83215463e-01 2.78933436e-01 -5.63377142e-01
-2.39620402e-01 9.94782150e-02 4.63708997e-01 8.43773484e-01
-4.34199013e-02 -1.94473282e-01 9.94291008e-01 -7.44650185e-01
-1.21519196e+00 -8.87883961e-01 -2.26204038e-01 8.66907597e-01
2.35387050e-02 -3.77229415e-02 -5.54092467e-01 -2.32710332e-01
6.49151921e-01 9.23899591e-01 -4.55862105e-01 3.10254335e-01
-6.10224545e-01 -1.12789559e+00 -3.46754864e-02 6.76925719e-01
3.15106601e-01 -1.04843295e+00 -9.94409561e-01 3.80113691e-01
5.99993825e-01 -1.03584754e+00 -5.29745162e-01 1.51858568e-01
-1.26367724e+00 -1.09157801e+00 -2.43392155e-01 -1.27488645e-02
4.26773816e-01 -3.28180432e-01 1.11845863e+00 8.15145746e-02
6.64116591e-02 2.31441349e-01 -4.67600636e-02 -5.00717342e-01
-2.21321583e-01 -6.21147081e-02 -1.69206448e-02 9.71201733e-02
3.73725355e-01 -2.92023957e-01 -6.29431069e-01 1.74250796e-01
-9.53889430e-01 -3.70663762e-01 3.75845790e-01 4.05454695e-01
1.04027651e-01 3.51476997e-01 3.14178407e-01 -8.58918965e-01
1.20798814e+00 -1.43789157e-01 -1.49549329e+00 2.55286843e-01
-1.22846174e+00 1.43928051e-01 5.74846230e-02 -2.93244898e-01
-1.16775489e+00 -7.52987266e-01 4.61859733e-01 1.36570677e-01
4.09573406e-01 5.38971007e-01 3.73087287e-01 7.07549825e-02
-3.79293948e-01 1.90355793e-01 1.26341969e-01 -8.00300002e-01
-7.95555785e-02 -1.30851805e-01 -3.69493216e-02 -3.40982288e-01
7.77738690e-01 3.03394020e-01 -1.46013066e-01 -4.35388505e-01
-4.33850521e-03 9.40115899e-02 -5.27463794e-01 1.94516808e-01
1.61358878e-01 -4.20316130e-01 -1.09614539e+00 6.79255247e-01
-8.26016128e-01 -6.78313255e-01 -6.14981592e-01 7.75833547e-01
-6.02959991e-01 -6.84045106e-02 -1.56493294e+00 -1.11705005e+00
-2.91132834e-02 -7.74668634e-01 1.25084564e-01 3.83335091e-02
-4.54609424e-01 -1.26712835e+00 3.62046123e-01 3.01776260e-01
3.85283470e-01 8.79105926e-02 1.06047952e+00 -1.41611087e+00
-1.21108973e+00 -3.84699345e-01 2.26098999e-01 4.55994532e-02
1.42971992e-01 2.19461486e-01 -2.02110767e-01 -1.51603460e-01
6.75541103e-01 3.92029554e-01 2.96836913e-01 5.51742971e-01
-1.25364468e-01 -7.37843394e-01 -2.97450032e-02 9.46609229e-02
1.51335406e+00 8.04140508e-01 1.16898321e-01 7.88393617e-01
-2.49601170e-01 5.81360579e-01 8.63384366e-01 9.09405053e-01
6.74405098e-02 2.40706071e-01 -1.07494220e-01 5.25820255e-01
9.34430897e-01 -2.83010602e-01 2.20194340e-01 8.00013244e-01
-2.13015020e-01 -3.96011956e-02 -7.57779360e-01 2.40271866e-01
-1.59223902e+00 -8.10117424e-01 2.01451689e-01 2.27332640e+00
4.86012608e-01 8.09820890e-01 7.44052947e-01 -1.23611502e-01
6.37328982e-01 -7.48830065e-02 -5.73465824e-01 -4.75069970e-01
-1.91619359e-02 1.18458003e-01 1.17382157e+00 7.11986899e-01
-3.23457360e-01 6.49670184e-01 8.76347065e+00 1.41973764e-01
-8.53580356e-01 -2.93668896e-01 1.21995044e+00 -5.19695699e-01
-4.82431352e-01 3.18989187e-01 -7.68090248e-01 6.21399403e-01
1.19583130e+00 -8.30577791e-01 5.76722443e-01 2.67244756e-01
2.44496107e-01 -1.68359950e-01 -1.29824972e+00 3.05638015e-01
-6.00291789e-01 -1.40840590e+00 1.66583881e-01 9.02989268e-01
6.07897401e-01 -7.22497031e-02 4.51150715e-01 -2.02537879e-01
3.50552350e-01 -5.66844642e-01 1.27385509e+00 4.80671793e-01
-1.73978895e-01 -8.68461668e-01 7.85369813e-01 4.30120200e-01
-8.58042598e-01 -2.28787020e-01 5.04081137e-02 -8.10109198e-01
4.25399572e-01 3.60549718e-01 -4.64605659e-01 4.23394531e-01
3.31448197e-01 -4.62684967e-02 -2.81268716e-01 8.62356305e-01
1.50009513e-01 6.38710380e-01 -3.77190173e-01 8.08028504e-02
3.59553784e-01 -8.23140860e-01 2.66854852e-01 3.07555676e-01
-3.18854637e-02 5.33291698e-01 -2.08935365e-01 1.12656248e+00
6.99113065e-04 -5.42237982e-02 -1.07244544e-01 -3.83776903e-01
3.68346184e-01 6.19585752e-01 -1.39389753e+00 -4.18928504e-01
-5.60268521e-01 5.83635569e-01 -5.88596880e-01 4.15599138e-01
-6.06524169e-01 1.31270424e-01 3.57448876e-01 7.60103106e-01
4.39307094e-01 -4.90557760e-01 -7.90336609e-01 -1.08506656e+00
5.56513608e-01 -8.25617790e-01 1.36481211e-01 2.53422018e-02
-1.18442142e+00 1.22142367e-01 4.53128442e-02 -7.79008865e-01
-5.34656107e-01 -5.08349419e-01 -7.73557901e-01 5.24801493e-01
-1.23701441e+00 1.21720664e-01 9.83079851e-01 -6.40141442e-02
5.77454448e-01 -3.03787768e-01 2.33466774e-02 -2.09274605e-01
-1.75327033e-01 5.64655811e-02 7.00754404e-01 2.81618983e-01
9.14138779e-02 -1.42021549e+00 1.37071466e+00 4.49469030e-01
2.32261419e-01 7.93106198e-01 7.31626749e-01 -1.27885139e+00
-1.21497238e+00 -1.88425168e-01 9.17589426e-01 -6.10888481e-01
1.30511487e+00 -3.34753215e-01 -6.03744388e-01 1.14730310e+00
7.35947609e-01 -6.01187766e-01 6.55901730e-01 -1.48917124e-01
1.37221292e-01 -1.30553022e-01 -9.06265795e-01 4.62602109e-01
4.72668141e-01 -3.95726234e-01 -9.36737895e-01 5.74744493e-03
6.35178328e-01 7.45750144e-02 -9.42003131e-01 9.79439169e-02
9.32792783e-01 -1.31455481e+00 5.03764749e-01 -3.80284011e-01
-1.06218956e-01 5.09132922e-01 2.00535744e-01 -6.08396351e-01
-1.88196033e-01 -1.36765540e+00 9.36612673e-03 8.90430212e-01
6.63042903e-01 -1.29370630e+00 7.18357682e-01 1.46650779e+00
6.84631646e-01 -5.82873523e-01 -9.01452065e-01 -1.07726991e+00
4.09843624e-01 1.20246984e-01 8.24069798e-01 7.05922008e-01
3.02427948e-01 -5.20493723e-02 1.28473341e-01 -3.19593489e-01
8.42723966e-01 6.44993007e-01 1.95276663e-01 -1.24741530e+00
-5.23162544e-01 -6.83305502e-01 1.01689123e-01 -9.73515391e-01
-4.25953567e-01 -2.19344184e-01 -6.32790148e-01 -5.99431634e-01
1.97363392e-01 1.28548771e-01 -5.20283163e-01 -4.78462428e-01
3.74761313e-01 -3.26298207e-01 7.56854057e-01 7.34831393e-01
-3.51728797e-01 1.50819212e-01 8.95026982e-01 3.13928217e-01
-4.36882764e-01 3.69456559e-01 -5.80690801e-01 5.48973560e-01
7.10445762e-01 -2.65988857e-01 -9.84946042e-02 -1.94168799e-02
6.53126359e-01 5.63379049e-01 1.04843475e-01 -2.70952940e-01
1.78851560e-01 -4.90372330e-01 2.43551597e-01 -8.77034664e-01
-1.57121643e-01 -8.14103901e-01 3.65207195e-01 8.95576358e-01
-5.41189015e-01 1.15150392e+00 2.43182871e-02 4.66027468e-01
-4.00890797e-01 -1.79355204e-01 1.92988560e-01 -4.09068912e-01
8.14864114e-02 6.88162744e-02 -7.24551618e-01 1.19174495e-01
1.06242990e+00 -1.04791977e-01 -3.49013694e-02 -3.54422569e-01
-7.93629706e-01 2.49512300e-01 5.53310275e-01 2.20526382e-01
-2.93807015e-02 -8.36883962e-01 -4.64807242e-01 2.05285057e-01
-6.47903085e-01 -6.96650445e-01 -3.42957616e-01 7.51730919e-01
-9.65731502e-01 8.18970978e-01 -3.49694163e-01 1.39245018e-01
-6.46817565e-01 5.08272231e-01 4.25300360e-01 -7.41548061e-01
-3.79192948e-01 3.14137906e-01 2.56411105e-01 3.22754323e-01
-1.74122825e-01 -7.68289387e-01 2.76274383e-01 1.99604079e-01
5.07027626e-01 8.19888532e-01 -6.66256202e-03 -1.29829407e-01
-2.07628869e-02 1.80659279e-01 -3.36121649e-01 -8.09435666e-01
1.38723338e+00 -3.16873938e-01 -5.30888140e-01 1.06490695e+00
4.37279940e-01 1.35475680e-01 -1.41184795e+00 5.93857653e-02
9.65573311e-01 -6.34378731e-01 -4.95319188e-01 -7.22441852e-01
-1.06397176e+00 4.11199629e-01 -1.39232367e-01 9.28203046e-01
4.99434441e-01 -2.26128817e-01 8.74813735e-01 1.82915583e-01
5.31839311e-01 -1.37430024e+00 -1.48467749e-01 2.95799106e-01
6.64994121e-01 -4.68179762e-01 -3.65649723e-02 -2.90911287e-01
-5.67878127e-01 9.60758269e-01 -1.05306387e-01 -7.37251639e-01
1.15456593e+00 5.77010930e-01 1.49015784e-01 -2.14898214e-01
-1.05264008e+00 6.34515941e-01 -1.70587018e-01 -2.50433505e-01
3.21342260e-01 2.58867830e-01 -7.03490734e-01 9.45612609e-01
-4.18339580e-01 -1.75110139e-02 8.86771679e-01 1.37696934e+00
-3.37666959e-01 -1.38518548e+00 -3.61711711e-01 6.60707295e-01
-1.14802110e+00 -4.71450090e-01 -7.35926867e-01 9.68083858e-01
-6.47567868e-01 6.00701809e-01 7.32405663e-01 2.34045133e-01
3.85321409e-01 2.67465502e-01 3.11233699e-01 -3.57808977e-01
-1.11287224e+00 7.94386566e-01 -1.05375119e-01 -2.25292310e-01
3.46854469e-03 -1.10812402e+00 -1.06159067e+00 -8.26827705e-01
-4.49062645e-01 3.90505046e-01 8.74830037e-02 8.95898104e-01
1.21275857e-01 1.41360298e-01 6.98098242e-01 -3.20592046e-01
-1.30338693e+00 -6.41821742e-01 -1.56086767e+00 1.50837436e-01
4.29460734e-01 -2.46050686e-01 -7.06863582e-01 -3.47965509e-02] | [4.760397911071777, 4.053459167480469] |
d244cae0-a4ed-403b-a3c8-af1c10e2fb0c | a-cnn-rnn-framework-with-a-novel-patch-based | 1902.11274 | null | https://arxiv.org/abs/1902.11274v3 | https://arxiv.org/pdf/1902.11274v3.pdf | A Novel Multi-Attention Driven System For Multi-Label Remote Sensing Image Classification | This paper presents a novel multi-attention driven system that jointly exploits Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) in the context of multi-label remote sensing (RS) image classification. The proposed system consists of four main modules. The first module aims to extract preliminary local descriptors of RS image bands that can be associated to different spatial resolutions. To this end, we introduce a K-Branch CNN, in which each branch extracts descriptors of image bands that have the same spatial resolution. The second module aims to model spatial relationship among local descriptors. This is achieved by a bidirectional RNN architecture, in which Long Short-Term Memory nodes enrich local descriptors by considering spatial relationships of local areas (image patches). The third module aims to define multiple attention scores for local descriptors. This is achieved by a novel patch-based multi-attention mechanism that takes into account the joint occurrence of multiple land-cover classes and provides the attention-based local descriptors. The last module exploits these descriptors for multi-label RS image classification. Experimental results obtained on the BigEarthNet that is a large-scale Sentinel-2 benchmark archive show the effectiveness of the proposed method compared to a state of the art method. | ['Begüm Demir', 'Gencer Sumbul'] | 2019-02-28 | null | null | null | null | ['remote-sensing-image-classification'] | ['miscellaneous'] | [ 2.71088213e-01 -2.65197337e-01 -1.60780087e-01 -4.62217957e-01
-1.11519253e+00 -6.38931766e-02 5.29353797e-01 3.86005729e-01
-5.87860942e-01 4.03850645e-01 3.20234776e-01 1.07935406e-01
-5.69926679e-01 -1.14310229e+00 -6.14055037e-01 -9.83024776e-01
-1.92270756e-01 3.85671258e-02 1.67495549e-01 -3.69971067e-01
2.10617408e-01 9.34305906e-01 -1.80244768e+00 4.09265965e-01
4.20126557e-01 1.19359946e+00 5.65425456e-01 6.23728871e-01
-1.09997973e-01 1.23965597e+00 -1.86914980e-01 2.44492605e-01
3.79546210e-02 -1.36104301e-01 -9.72310781e-01 -2.51082152e-01
3.83063942e-01 -2.78565198e-01 3.53611112e-02 1.01861310e+00
6.36567175e-01 4.10415441e-01 4.80043173e-01 -6.59904957e-01
-7.70630598e-01 4.30826902e-01 -4.58425909e-01 4.88904327e-01
-3.34798098e-01 -3.01620930e-01 1.30639744e+00 -8.18037093e-01
3.47846061e-01 1.15445650e+00 8.00449133e-01 1.89581990e-01
-1.01814711e+00 -5.05359292e-01 2.67410904e-01 2.63582826e-01
-1.87986279e+00 -4.75614965e-02 6.82910085e-01 -4.94919360e-01
1.19512558e+00 1.74193308e-01 3.71331871e-01 3.71247619e-01
2.91321099e-01 5.13145685e-01 9.78598952e-01 -1.66405171e-01
-9.72838551e-02 7.43716806e-02 4.82907832e-01 4.81137305e-01
-3.80544782e-01 3.82617228e-02 -1.02125973e-01 -1.25789687e-01
6.53093755e-01 5.16540527e-01 1.28711918e-02 -2.24468410e-02
-9.73787487e-01 1.10390890e+00 1.24857557e+00 7.34265327e-01
-1.06080031e+00 2.44700596e-01 3.80069643e-01 -3.10380943e-02
9.04380739e-01 9.24486145e-02 -3.69440705e-01 6.81115031e-01
-1.02373576e+00 5.45120500e-02 1.50586337e-01 5.10935128e-01
1.33507013e+00 -1.62115708e-01 -3.51894826e-01 1.15727472e+00
4.88100737e-01 6.17072642e-01 3.63732636e-01 -2.95326352e-01
2.49681264e-01 7.70303249e-01 -1.20932877e-01 -1.24649966e+00
-6.32218659e-01 -7.77017713e-01 -1.11446500e+00 -3.32669988e-02
-5.21223605e-01 1.48846999e-01 -9.10330534e-01 1.51257300e+00
1.59611732e-01 1.60848513e-01 2.63889104e-01 8.73960018e-01
1.24910271e+00 1.09331119e+00 5.08218527e-01 2.03270704e-01
1.38246095e+00 -1.09675539e+00 -4.06771868e-01 -3.70539241e-02
4.64603364e-01 -3.87506813e-01 4.60482776e-01 -2.89365023e-01
-6.43155158e-01 -8.67243707e-01 -8.22659969e-01 -6.18368313e-02
-9.17366087e-01 4.12726372e-01 2.31776729e-01 -1.95702929e-02
-1.22792828e+00 3.44936162e-01 -4.67511773e-01 -6.24830246e-01
3.29380393e-01 2.60638267e-01 -3.42689961e-01 1.49070621e-01
-1.27590942e+00 7.86646664e-01 4.47232783e-01 6.15437210e-01
-8.75143647e-01 -5.66651523e-01 -9.56921518e-01 2.86384404e-01
-9.74968076e-02 -2.52829939e-01 7.08844006e-01 -1.43826151e+00
-9.68288660e-01 9.76639330e-01 -2.58576311e-02 -3.69432956e-01
-3.33689421e-01 -9.19434056e-02 -5.20097613e-01 3.22053373e-01
4.41699445e-01 7.61655390e-01 4.83231366e-01 -1.12878287e+00
-8.97060335e-01 -5.01993835e-01 4.66840975e-02 3.73574406e-01
-2.92461842e-01 3.40069652e-01 -1.49068117e-01 -6.22766495e-01
7.89880976e-02 -5.49982011e-01 -3.89653742e-01 -4.48781580e-01
-1.20234683e-01 -2.93713897e-01 5.63250422e-01 -6.05087876e-01
1.05227125e+00 -2.26119518e+00 1.46666408e-01 3.30118001e-01
-3.56019065e-02 2.95620978e-01 -5.12549102e-01 5.59892833e-01
-2.05303594e-01 2.16896743e-01 -1.91030562e-01 -4.91107628e-02
-2.82095671e-01 2.19147354e-02 -1.76977798e-01 4.52228814e-01
6.30117595e-01 8.96422446e-01 -6.57604456e-01 -5.04369736e-01
2.24300846e-01 7.80344784e-01 -1.29944399e-01 2.64190853e-01
-8.40301514e-02 3.22924346e-01 -6.96321607e-01 6.61054134e-01
8.25167537e-01 -2.55078375e-01 -1.46678567e-01 -3.80743533e-01
-6.38139963e-01 -1.14162356e-01 -8.46911013e-01 1.46617043e+00
-6.15617037e-01 3.64810616e-01 -6.31893575e-02 -1.20910871e+00
1.20901847e+00 4.41972688e-02 4.96071517e-01 -9.91979122e-01
3.54756750e-02 1.38655320e-01 -5.89197636e-01 -5.66352248e-01
6.30449593e-01 -1.65772662e-01 7.04058707e-02 3.05114716e-01
-5.04341759e-02 4.18905526e-01 -2.12088928e-01 -2.61690170e-01
7.81246960e-01 1.57643352e-02 3.60946476e-01 -3.95564944e-01
1.03351092e+00 8.02316070e-02 3.90831649e-01 8.24960053e-01
-1.11388013e-01 4.75426674e-01 6.03950173e-02 -9.58407342e-01
-8.45809221e-01 -5.70355594e-01 -3.45927596e-01 1.60185635e+00
1.41485527e-01 1.54757202e-01 -2.24661887e-01 -4.42534655e-01
-1.24049716e-01 2.22224638e-01 -1.11584520e+00 1.55000746e-01
-3.71898502e-01 -1.04249656e+00 5.61542213e-01 6.43788397e-01
8.25895131e-01 -1.44882846e+00 -6.58677518e-01 2.47233957e-01
-1.15750231e-01 -6.72059536e-01 1.93684306e-02 3.53625774e-01
-6.03085637e-01 -8.40223730e-01 -9.46373940e-01 -1.00917053e+00
6.15404069e-01 5.55240989e-01 1.08763254e+00 1.66391760e-01
-2.20418230e-01 3.11924487e-01 -5.53590298e-01 -2.85955053e-02
-9.66197997e-02 4.49288338e-01 -6.64432824e-01 5.55067241e-01
5.48785985e-01 -2.89486378e-01 -6.63471401e-01 2.80013293e-01
-1.09495223e+00 -3.08949977e-01 1.03693438e+00 1.02073908e+00
1.11229777e+00 -6.20500930e-03 5.08923113e-01 -5.42824328e-01
2.74872392e-01 -9.82709289e-01 -6.38391137e-01 4.74129051e-01
-1.67368189e-01 1.03622638e-01 4.67905968e-01 3.62737700e-02
-9.41803932e-01 1.85698524e-01 -2.93234587e-01 -1.32081360e-01
-3.97012025e-01 9.83786881e-01 1.51152536e-01 -9.44983736e-02
4.57080424e-01 5.85275292e-01 -1.16597973e-01 -6.20843291e-01
2.16397569e-01 1.03183210e+00 2.26922780e-01 -1.53140098e-01
3.68991196e-01 5.11180699e-01 -9.08748284e-02 -1.08138955e+00
-1.12518346e+00 -9.54590857e-01 -7.30409503e-01 -2.11182117e-01
1.31928432e+00 -1.20774329e+00 -4.96007025e-01 6.02904022e-01
-1.07245207e+00 -2.33663484e-01 -1.60747945e-01 4.59004790e-01
-2.41595998e-01 -6.31818995e-02 -4.65817392e-01 -7.67433405e-01
-7.30324924e-01 -9.95296061e-01 1.42785716e+00 3.66297811e-01
5.73568463e-01 -1.07847548e+00 5.15698373e-01 1.67798206e-01
8.07405055e-01 2.93764621e-01 6.61438465e-01 -7.37697959e-01
-6.26453817e-01 -2.96602789e-02 -8.41483772e-01 5.47091246e-01
-2.83134487e-02 -1.48026630e-01 -9.47598219e-01 -2.45817140e-01
-3.60681593e-01 -3.99009645e-01 1.24256170e+00 5.97311258e-01
1.11001480e+00 -1.73509434e-01 -2.09027156e-01 6.49805605e-01
2.00590754e+00 -3.04876529e-02 6.58746123e-01 6.42230868e-01
7.86621034e-01 6.35902584e-01 6.99368596e-01 3.35919589e-01
5.21319091e-01 6.18700743e-01 7.16026843e-01 -6.61983609e-01
1.66952774e-01 2.37058744e-01 5.81003763e-02 5.74095488e-01
-2.55064547e-01 6.15737401e-02 -9.99978006e-01 8.74054849e-01
-1.92095268e+00 -1.35373652e+00 -1.28718883e-01 1.86138904e+00
2.00727448e-01 -5.12383461e-01 -3.99923883e-02 -2.68417805e-01
8.94476473e-01 6.93697751e-01 -4.27986860e-01 -3.25197816e-01
-4.03035045e-01 3.64204973e-01 8.49281788e-01 4.55025911e-01
-1.61128843e+00 9.29642797e-01 5.20046806e+00 8.44287157e-01
-1.24625754e+00 3.18923146e-01 5.93896747e-01 2.46053025e-01
-5.81259839e-02 -3.15096736e-01 -9.01265383e-01 9.44055766e-02
1.02749598e+00 3.30444902e-01 6.50434718e-02 7.71576285e-01
8.66474658e-02 3.84779572e-02 -3.01238507e-01 5.46427190e-01
2.84581304e-01 -1.30100393e+00 3.37076753e-01 -1.54427532e-02
7.20981359e-01 8.25782180e-01 6.20703362e-02 2.52783120e-01
1.04995696e-02 -1.03106284e+00 5.19898117e-01 1.11228263e+00
6.38415813e-01 -9.98634040e-01 1.29287791e+00 1.83975071e-01
-1.74270582e+00 -3.52233052e-01 -6.93881273e-01 4.24255073e-01
-3.29613030e-01 3.43288571e-01 -4.04100329e-01 8.05803716e-01
8.64283741e-01 1.15494621e+00 -7.02486992e-01 8.82199466e-01
1.11594170e-01 2.43464068e-01 -1.61360316e-02 -1.35117650e-01
8.48516822e-01 -7.38319457e-02 2.43894875e-01 1.52272534e+00
2.44236842e-01 1.17518276e-01 1.82790801e-01 8.60792100e-01
1.50699550e-02 5.17830729e-01 -7.62663245e-01 2.80153841e-01
1.45287991e-01 1.62435937e+00 -5.20628333e-01 -1.94128916e-01
-4.79638696e-01 7.40628004e-01 4.42721903e-01 3.81902993e-01
-7.93029070e-01 -6.04100704e-01 4.44877863e-01 -2.63994962e-01
6.87711716e-01 9.95389521e-02 4.53034490e-01 -7.26591170e-01
-2.41167739e-01 -4.93677467e-01 5.56348979e-01 -9.02587175e-01
-1.33634937e+00 1.14273739e+00 -1.55229211e-01 -1.27958095e+00
9.17005464e-02 -2.49544322e-01 -2.92102784e-01 1.20833313e+00
-2.33471227e+00 -1.76115394e+00 -6.76574230e-01 7.49388933e-01
3.90571058e-01 -2.55042970e-01 1.00703907e+00 4.78838652e-01
-5.29894710e-01 1.38717100e-01 4.31819081e-01 1.94772378e-01
3.63890350e-01 -9.69446957e-01 1.77859906e-02 6.14634454e-01
-1.33919075e-01 3.26236308e-01 -1.30292073e-01 -4.23573434e-01
-8.34585786e-01 -1.84350932e+00 1.20322824e+00 2.27897152e-01
6.01212680e-01 2.21332207e-01 -9.01963294e-01 6.41963780e-01
8.03892966e-03 4.71178561e-01 8.71784508e-01 -1.42339319e-01
-3.80043894e-01 -6.19969368e-01 -9.56431925e-01 -2.02782974e-01
3.52459282e-01 -8.87147605e-01 -2.82630831e-01 4.62407053e-01
4.78610516e-01 1.34429812e-01 -1.00176179e+00 4.55992669e-01
4.35890138e-01 -8.78348231e-01 1.04784989e+00 -4.90447432e-01
4.21500415e-01 -4.43256527e-01 -4.95252877e-01 -1.16019797e+00
-8.12635005e-01 2.64093727e-01 7.07834899e-01 1.07864749e+00
3.56030911e-01 -5.85950613e-01 1.78034097e-01 -1.86369464e-01
-4.64270152e-02 -5.17178297e-01 -7.51347244e-01 -5.73968291e-01
5.82263954e-02 -1.28281072e-01 7.38697886e-01 8.91884387e-01
-7.33895004e-01 3.21071923e-01 -3.92890751e-01 6.54787898e-01
4.95699614e-01 3.75773728e-01 2.86703289e-01 -1.32314491e+00
1.48266971e-01 -6.10103309e-01 -6.79853678e-01 -7.70009100e-01
4.20285493e-01 -1.12369204e+00 1.13464586e-01 -1.66370630e+00
5.75623930e-01 -5.37065446e-01 -1.02378166e+00 6.09672070e-01
-1.15805231e-01 5.65393090e-01 -4.89428602e-02 3.92563850e-01
-9.05609429e-01 6.61377907e-01 5.80766439e-01 -5.07511854e-01
-1.62834287e-01 -5.38314953e-02 -4.44927990e-01 2.95967460e-01
8.80673110e-01 -5.71337044e-01 4.97057550e-02 -5.18571854e-01
2.92664945e-01 -1.13435976e-01 7.81969130e-01 -1.14235759e+00
3.25758219e-01 2.97133136e-03 1.54752746e-01 -9.99557734e-01
-8.76381341e-03 -8.81737590e-01 2.27373213e-01 4.76767778e-01
-6.14835024e-01 4.96290624e-02 2.70077825e-01 5.76103270e-01
-5.64191997e-01 -1.79222241e-01 9.96247351e-01 -2.50354528e-01
-9.79482234e-01 4.21917409e-01 -3.85632128e-01 -5.24294734e-01
9.90657508e-01 1.71081230e-01 -1.81815237e-01 9.61114764e-02
-7.45065928e-01 2.27717832e-01 -6.98507428e-02 3.91006023e-01
6.29492044e-01 -1.43847609e+00 -1.09931326e+00 2.53833294e-01
5.04125357e-01 -2.28849262e-01 6.54505908e-01 7.19235778e-01
-7.08672523e-01 4.61654902e-01 -3.95485431e-01 -6.98748469e-01
-1.18783426e+00 4.08817828e-01 8.66105855e-01 -5.10867059e-01
-3.66489887e-01 8.46524119e-01 1.55270830e-01 -5.47378302e-01
1.20847742e-03 -1.07873127e-01 -1.00006998e+00 4.84855294e-01
6.59899950e-01 2.55017966e-01 1.89925045e-01 -1.39306962e+00
-6.10132217e-01 1.11841726e+00 9.08267200e-02 2.96130210e-01
1.90278566e+00 -2.49093741e-01 -5.48204422e-01 4.12596107e-01
1.49602568e+00 -4.57016051e-01 -7.72510588e-01 -6.38602912e-01
-9.46461558e-02 3.26502100e-02 5.93655407e-01 -7.41656721e-01
-1.55899620e+00 9.36406255e-01 1.00676084e+00 1.96723133e-01
1.36440563e+00 -6.59428388e-02 5.15992641e-01 4.32078540e-01
4.26823692e-03 -1.04777789e+00 -3.28341544e-01 6.16751075e-01
9.39660430e-01 -1.36312437e+00 -1.19207427e-01 3.74538451e-01
-4.02246684e-01 1.12049139e+00 2.25836009e-01 -1.92417532e-01
9.88622427e-01 -3.80785942e-01 1.03023984e-01 -7.75869310e-01
-4.48140889e-01 -8.89349818e-01 4.13273007e-01 1.87311053e-01
3.77151340e-01 1.05409548e-01 -2.36043066e-01 2.60082901e-01
5.12012124e-01 -1.40508443e-01 -7.77437538e-02 7.51695931e-01
-7.11775243e-01 -6.65262997e-01 -3.97670776e-01 2.67690539e-01
-5.13499796e-01 -3.19984376e-01 -1.09284945e-01 5.50880909e-01
3.50311786e-01 9.25216496e-01 3.09569746e-01 -3.87488693e-01
2.09929973e-01 -1.18791245e-01 -2.73926198e-01 -5.29522359e-01
-1.11896133e+00 3.24286483e-02 -2.90794998e-01 -6.33039415e-01
-1.09588695e+00 -4.48490918e-01 -8.48289430e-01 8.98048952e-02
-1.86736152e-01 2.43816003e-01 7.81408131e-01 7.88856566e-01
5.59020936e-01 6.74937129e-01 9.93674815e-01 -1.03926194e+00
-2.21201763e-01 -1.06515682e+00 -8.33531082e-01 2.34134272e-01
6.76406860e-01 -4.59428072e-01 -2.24753410e-01 -1.54157877e-01] | [9.742246627807617, -1.3571908473968506] |
d2c22761-dc0f-430f-a2dc-ba6b959caa3c | robust-stability-of-gaussian-process-based | 2304.06530 | null | https://arxiv.org/abs/2304.06530v2 | https://arxiv.org/pdf/2304.06530v2.pdf | Robust Stability of Gaussian Process Based Moving Horizon Estimation | In this paper, we introduce a Gaussian process based moving horizon estimation (MHE) framework. The scheme is based on offline collected data and offline hyperparameter optimization. In particular, compared to standard MHE schemes, we replace the mathematical model of the system by the posterior mean of the Gaussian process. To account for the uncertainty of the learned model, we exploit the posterior variance of the learned Gaussian process in the weighting matrices of the cost function of the proposed MHE scheme. We prove practical robust exponential stability of the resulting estimator using a recently proposed Lyapunov-based proof technique. Finally, the performance of the Gaussian process based MHE scheme is illustrated via a nonlinear system. | ['Matthias A. Müller', 'Victor G. Lopez', 'Tobias M. Wolff'] | 2023-04-13 | null | null | null | null | ['hyperparameter-optimization'] | ['methodology'] | [ 1.91471260e-02 4.10446882e-01 3.11702006e-02 3.97990733e-01
-6.77754998e-01 -4.06231046e-01 4.59856540e-01 9.84853432e-02
-4.21087265e-01 8.72793496e-01 -2.07345292e-01 -3.17744434e-01
-6.12621427e-01 -4.16554481e-01 -7.47594357e-01 -1.10740376e+00
-9.80874225e-02 9.35012996e-02 4.19011042e-02 1.87191263e-01
1.50792956e-01 2.92333663e-01 -7.41076410e-01 -7.35528946e-01
7.44066775e-01 1.25386202e+00 8.05352479e-02 9.17804897e-01
6.91457033e-01 4.05918568e-01 -3.97970766e-01 -1.22084334e-01
5.49972177e-01 -5.59042804e-02 -8.89872834e-02 4.87511694e-01
-3.53413522e-01 -2.71484256e-01 -4.31689888e-01 1.21436632e+00
5.77282369e-01 5.90480685e-01 6.21597111e-01 -1.13429666e+00
-5.60983196e-02 3.51818383e-01 -4.29120898e-01 -7.77548924e-02
-1.96303144e-01 2.00502917e-01 5.38815558e-01 -1.08350086e+00
4.57921594e-01 1.10895514e+00 6.16114676e-01 3.82519364e-01
-1.20043588e+00 -5.17516971e-01 -1.36442231e-02 -3.32560055e-02
-1.55497885e+00 -4.43775892e-01 4.51498300e-01 -6.85389459e-01
3.17751944e-01 -6.64405227e-02 6.22398257e-01 6.05554938e-01
7.53607810e-01 6.72009230e-01 1.05807233e+00 -2.38407105e-01
6.59955323e-01 9.31230113e-02 3.03236041e-02 6.28732979e-01
6.06023729e-01 4.14975673e-01 -1.64798573e-02 -3.59402210e-01
1.11117685e+00 7.34705999e-02 -3.62349749e-01 -7.74450541e-01
-1.08744764e+00 7.43579686e-01 -8.59103203e-02 -2.47850195e-01
-6.75871193e-01 3.08775574e-01 3.54566015e-02 5.00599146e-02
4.59110767e-01 2.05317333e-01 -1.22736961e-01 -9.43372548e-02
-6.83863580e-01 4.30039167e-01 1.00927556e+00 1.13730800e+00
3.24611276e-01 3.50025892e-01 -4.70395416e-01 2.87159473e-01
6.63609982e-01 8.98273885e-01 7.82727376e-02 -1.16392338e+00
4.49684829e-01 -3.83142792e-02 8.28102469e-01 -6.39866531e-01
-1.95913792e-01 -5.04892826e-01 -8.15415323e-01 2.16582939e-01
4.72608119e-01 -8.47012997e-01 -5.76768875e-01 1.47632706e+00
5.17806649e-01 5.29487848e-01 2.52230316e-01 5.57573497e-01
-5.67934573e-01 7.68428922e-01 -2.84622550e-01 -6.46466494e-01
7.60834455e-01 -7.97975123e-01 -1.08250785e+00 2.14182645e-01
-2.94657182e-02 -4.67257529e-01 3.32269549e-01 4.94764656e-01
-1.29926884e+00 -2.43774861e-01 -1.13299942e+00 8.03003788e-01
2.68946379e-01 3.15902442e-01 -2.52818435e-01 4.82328087e-01
-8.19410384e-01 7.15318918e-01 -8.70340109e-01 -1.41979024e-01
4.89924150e-03 3.05332690e-01 1.42870158e-01 3.61634851e-01
-1.08217537e+00 8.18790197e-01 7.64607906e-01 5.18504977e-01
-1.07843208e+00 -7.14101315e-01 -6.40430331e-01 1.88887626e-01
7.99055994e-01 -7.04604089e-01 1.52044606e+00 -1.75338119e-01
-2.25306511e+00 -1.58864856e-01 -1.22310720e-01 -8.34348202e-01
8.33128631e-01 -5.65385818e-01 -1.47868916e-02 3.44407141e-01
-4.18416589e-01 -1.73786193e-01 1.48658121e+00 -1.05939460e+00
-7.32526839e-01 2.03223880e-02 -3.21068704e-01 -2.05508489e-02
4.79255430e-02 -4.04552430e-01 -1.85072705e-01 -5.48939288e-01
-8.23967755e-02 -1.28379238e+00 -7.07398534e-01 2.76601687e-02
-2.45601311e-01 2.61841953e-01 6.95660710e-01 -7.31970489e-01
1.48180330e+00 -2.19603562e+00 5.67986816e-02 5.09387493e-01
6.20342325e-03 2.26507872e-01 4.91205275e-01 7.72987366e-01
4.49094146e-01 -2.38347366e-01 -1.04998097e-01 -3.57438564e-01
2.51162410e-01 1.31196082e-01 -6.10823214e-01 7.99438000e-01
1.55958787e-01 4.87532556e-01 -9.70690787e-01 -1.14228442e-01
3.45536232e-01 3.24954659e-01 -1.99435458e-01 2.18300641e-01
8.64726827e-02 5.04423797e-01 -6.92568958e-01 1.15122423e-01
5.54385304e-01 -1.56417377e-02 1.53783917e-01 -4.96150106e-02
-2.51507312e-01 -5.21333158e-01 -1.66816175e+00 8.96162510e-01
-6.42774701e-01 3.36978376e-01 6.52819455e-01 -7.48238504e-01
6.81126237e-01 5.88057458e-01 3.51346940e-01 2.20941290e-01
4.20288652e-01 2.50719577e-01 -1.52896523e-01 -7.11215585e-02
2.79678792e-01 -3.61728430e-01 1.98407397e-01 1.12487964e-01
8.76176953e-02 -3.64869595e-01 -3.08833439e-02 -9.80840549e-02
8.03415120e-01 -8.25472996e-02 7.61915207e-01 -4.09889847e-01
7.51034796e-01 -4.59869474e-01 6.77381456e-01 7.22507358e-01
-4.47682321e-01 3.05065829e-02 2.60965139e-01 3.09627503e-01
-1.29732728e+00 -9.35391486e-01 -1.30473033e-01 2.65147775e-01
-1.98802594e-02 6.74314331e-04 -6.68541491e-01 -2.39397049e-01
7.59635493e-02 8.11753392e-01 -5.87336957e-01 -3.44181746e-01
-2.47891709e-01 -5.53156435e-01 1.94480773e-02 3.44315767e-01
4.36912358e-01 -1.08186260e-01 -5.88683486e-01 7.11423516e-01
5.17806470e-01 -1.12091327e+00 -4.97728288e-01 -4.42141667e-03
-8.65966558e-01 -7.69575775e-01 -8.15316021e-01 -2.66524643e-01
5.61191440e-01 -1.28451869e-01 1.73002362e-01 -6.18629575e-01
1.40984535e-01 9.84241486e-01 1.03464924e-01 -7.11792886e-01
-4.79722530e-01 -4.91317421e-01 4.43678796e-01 5.24939477e-01
-5.40920258e-01 -2.02809185e-01 -6.63190365e-01 3.12176317e-01
-5.87420166e-01 -8.20791349e-02 5.52431047e-01 1.13974273e+00
6.06647670e-01 6.16709650e-01 5.50953448e-01 -4.11239445e-01
8.47597241e-01 -4.21165735e-01 -1.50848138e+00 2.08231285e-01
-8.57201695e-01 1.22063302e-01 4.75634754e-01 -8.08229804e-01
-1.17633080e+00 2.28470072e-01 3.85582924e-01 -7.26365745e-01
4.51327890e-01 2.67014474e-01 -4.03373465e-02 -1.61080152e-01
-1.51880458e-01 1.36791378e-01 1.93844348e-01 -2.94296324e-01
3.73181552e-01 4.97908145e-01 7.57882714e-01 -6.22393608e-01
1.15746665e+00 4.97055113e-01 5.56501925e-01 -7.30279565e-01
-6.98422432e-01 -6.84661210e-01 -4.11380321e-01 -3.15436363e-01
6.38500750e-01 -8.56891394e-01 -1.04912949e+00 6.53371155e-01
-9.72130835e-01 -3.77141237e-01 -5.59300542e-01 8.37787569e-01
-1.21058166e+00 1.59330204e-01 -7.37080991e-01 -1.81915891e+00
-3.56600702e-01 -7.53895223e-01 8.80465567e-01 2.48528138e-01
1.54304117e-01 -1.34825587e+00 1.33935779e-01 -3.38708431e-01
3.88855547e-01 4.26054031e-01 3.46184283e-01 -5.06763875e-01
-4.44497436e-01 -5.69552541e-01 4.14374508e-02 5.39396167e-01
-2.98425645e-01 1.93459669e-03 -5.23849368e-01 -5.92719495e-01
6.91964686e-01 2.78648436e-01 2.44832635e-01 5.86368442e-01
4.78952348e-01 -5.92787027e-01 -3.95274431e-01 2.91545451e-01
1.75778258e+00 4.42171335e-01 1.54883772e-01 3.94397855e-01
5.15951455e-01 4.53450412e-01 9.09372091e-01 1.07995880e+00
5.84686436e-02 3.89537394e-01 2.19168901e-01 4.19569165e-01
8.41961980e-01 -3.26896012e-01 5.96988916e-01 6.15665972e-01
9.97211039e-02 -4.01462615e-02 -7.53166199e-01 4.30903196e-01
-2.15661168e+00 -7.48771071e-01 7.32326359e-02 2.80353999e+00
4.55780178e-01 1.31561503e-01 -1.64194047e-01 4.82306965e-02
8.85706425e-01 -4.02144492e-01 -5.31449974e-01 -2.57164627e-01
7.98208266e-02 -6.49564862e-02 1.24086487e+00 7.02948034e-01
-1.17013395e+00 2.77476132e-01 6.50098181e+00 8.91491175e-01
-8.00742209e-01 -3.74251604e-02 1.16900809e-01 2.51090378e-01
4.68178868e-01 2.09841847e-01 -1.07489920e+00 5.16182423e-01
1.26908851e+00 -1.15282917e+00 4.38542306e-01 8.87161493e-01
8.52178216e-01 -2.14293897e-01 -8.10721278e-01 7.94463038e-01
-4.35755551e-01 -8.18539202e-01 -4.79624361e-01 3.25743526e-01
9.07331169e-01 -5.38316369e-01 2.73088515e-01 3.92361075e-01
3.70562702e-01 -4.74014670e-01 7.96207547e-01 9.21204448e-01
1.93137676e-01 -9.21213984e-01 8.72955799e-01 4.68797207e-01
-1.21868491e+00 -7.46293545e-01 -3.05957556e-01 -1.57580432e-02
7.75885761e-01 6.71802044e-01 -8.71432424e-01 5.70245147e-01
2.86950916e-02 3.88410479e-01 -1.39017045e-01 1.54256368e+00
-2.89206892e-01 8.34810376e-01 -3.40028644e-01 1.43031120e-01
3.22092205e-01 -6.02975428e-01 1.22255778e+00 9.71763372e-01
7.12744117e-01 -1.76802546e-01 3.58208179e-01 5.81593633e-01
3.39637488e-01 -5.07580452e-02 -4.14230675e-01 -1.65730789e-01
4.24524039e-01 1.12318408e+00 -2.72506386e-01 -4.67352241e-01
-1.85472772e-01 7.51254916e-01 -1.84571296e-01 5.13205469e-01
-8.24812174e-01 -3.88426960e-01 3.51585895e-01 -1.54416755e-01
6.44967794e-01 -5.16804099e-01 4.23585400e-02 -8.68084967e-01
-9.97246131e-02 -3.84723216e-01 8.44156444e-02 -4.90066528e-01
-1.11842763e+00 -2.81073004e-02 2.12472662e-01 -1.44473386e+00
-6.47838354e-01 -3.24361026e-01 -6.86818600e-01 9.87365067e-01
-1.03569388e+00 -6.70419335e-01 4.20991302e-01 2.95479834e-01
4.80637044e-01 3.19850780e-02 2.30024070e-01 8.60153213e-02
-7.52875924e-01 3.62912565e-02 1.03089070e+00 -3.06607693e-01
6.14486158e-01 -1.42937446e+00 1.26456589e-01 9.92789090e-01
-7.85872400e-01 8.26301098e-01 1.23451364e+00 -6.36405051e-01
-1.48059750e+00 -1.40953207e+00 2.45057717e-01 1.50728719e-02
1.51392460e+00 -2.15267301e-01 -8.34120095e-01 6.46743774e-01
-4.49650846e-02 -1.21447071e-01 5.50689399e-02 -6.77821636e-01
5.12881577e-01 6.45212680e-02 -1.05908453e+00 5.51647604e-01
2.05619127e-01 -3.24827194e-01 -3.32309633e-01 1.76639277e-02
7.95123041e-01 -4.82901752e-01 -1.24863803e+00 6.05451703e-01
5.64764380e-01 7.34523013e-02 7.42779016e-01 -3.51336956e-01
-3.85500491e-01 -5.53081393e-01 -1.59073472e-01 -1.53629291e+00
-2.28550032e-01 -1.39527416e+00 -8.30598652e-01 1.10094988e+00
2.73976624e-01 -7.86334097e-01 4.09062326e-01 6.59026861e-01
1.70262948e-01 -7.89252222e-01 -1.10826492e+00 -1.34576118e+00
5.06879808e-03 -1.60623938e-01 -5.62033243e-02 2.16397658e-01
8.43463019e-02 1.81626156e-01 -7.19454706e-01 7.25773692e-01
1.16897571e+00 -3.50475639e-01 6.39235079e-01 -8.21535885e-01
-7.63778985e-01 -1.13005891e-01 -2.63866305e-01 -1.01315260e+00
1.24295108e-01 5.39843924e-03 4.63004291e-01 -1.27035809e+00
4.13253438e-03 1.01573355e-01 -3.01080525e-01 -4.84376580e-01
-4.40814346e-01 -4.49675590e-01 3.59253764e-01 -3.71840596e-02
-5.42049408e-01 8.44913304e-01 9.97247338e-01 8.28166455e-02
-3.83502692e-01 6.25166178e-01 3.02259009e-02 7.22246885e-01
8.42031002e-01 -3.31946433e-01 -6.07448399e-01 2.78841645e-01
-8.23879689e-02 5.24560630e-01 2.29778871e-01 -1.11673439e+00
4.07908738e-01 -1.99165374e-01 -1.02186792e-01 -6.39165640e-01
3.53529632e-01 -1.01479566e+00 3.58517259e-01 8.17594707e-01
-1.81452572e-01 -3.76383454e-01 7.33595192e-02 1.37527180e+00
-7.28849992e-02 -3.03152055e-01 1.02246368e+00 5.21446824e-01
-3.54578912e-01 2.43481472e-01 -8.01677525e-01 -2.21028492e-01
1.14288664e+00 6.33279309e-02 4.37705033e-02 -7.16760755e-01
-1.09486270e+00 5.77452362e-01 1.96649745e-01 -1.38840124e-01
4.50098664e-01 -9.91739750e-01 -2.49293163e-01 -3.17014456e-01
-2.77266920e-01 -2.14292303e-01 3.21230710e-01 1.30768692e+00
1.77679285e-01 5.27782440e-01 2.83081174e-01 -3.55994850e-01
-9.73679960e-01 9.02020156e-01 4.79864895e-01 -3.61682206e-01
-4.19716328e-01 3.88467871e-02 1.36025846e-01 4.94424775e-02
1.52485281e-01 -3.40995669e-01 1.15258835e-01 -3.45682710e-01
5.71002543e-01 8.73804331e-01 -3.75762492e-01 -3.27985793e-01
1.73390791e-01 3.79983515e-01 5.08128345e-01 -7.43593037e-01
1.07058299e+00 -8.98989737e-01 1.93420336e-01 9.21391964e-01
9.10771847e-01 -1.39991322e-03 -1.84657514e+00 -2.24538997e-01
2.64243186e-01 -1.48319915e-01 1.41531467e-01 -1.92763552e-01
-5.57491660e-01 5.14616966e-01 7.44927645e-01 -2.37714380e-01
8.71718228e-01 -6.40155435e-01 5.72931767e-01 4.74945962e-01
5.29361188e-01 -1.28641200e+00 -3.26801687e-01 5.96018612e-01
5.99153697e-01 -7.09464908e-01 2.92171612e-02 -4.01103139e-01
-6.25775874e-01 1.00516891e+00 1.68201238e-01 -5.45351446e-01
1.00610507e+00 1.70024186e-01 -3.76187444e-01 3.59738410e-01
-1.01069438e+00 -1.59936007e-02 1.97259232e-01 3.59955609e-01
-2.26773933e-01 9.37373266e-02 -6.65615559e-01 5.88891327e-01
4.27144110e-01 2.64531821e-01 8.75037491e-01 1.07413387e+00
-5.67674816e-01 -6.49801552e-01 -7.28097320e-01 2.83002686e-02
-6.04797781e-01 1.59811988e-01 3.48038524e-01 7.36614704e-01
-5.23616552e-01 9.70172405e-01 -2.78183609e-01 5.69823906e-02
5.09451151e-01 1.31145373e-01 2.81253666e-01 -3.21408331e-01
-8.71187600e-04 5.76785684e-01 -1.33145064e-01 -4.07706827e-01
9.28025618e-02 -5.89587510e-01 -1.09403169e+00 5.24513647e-02
-5.88043451e-01 3.37329447e-01 7.26434648e-01 8.13571870e-01
1.90692753e-01 3.04663926e-01 8.75095010e-01 -7.76978493e-01
-1.44660747e+00 -8.27216268e-01 -1.00673831e+00 -1.62290663e-01
3.10037285e-01 -8.47198963e-01 -6.87489688e-01 -3.11260056e-02] | [5.128200054168701, 2.534946918487549] |
eb23324b-e5a2-40d2-b997-4d41061d600f | shuffle-divide-contrastive-learning-for-long | 2304.09374 | null | https://arxiv.org/abs/2304.09374v1 | https://arxiv.org/pdf/2304.09374v1.pdf | Shuffle & Divide: Contrastive Learning for Long Text | We propose a self-supervised learning method for long text documents based on contrastive learning. A key to our method is Shuffle and Divide (SaD), a simple text augmentation algorithm that sets up a pretext task required for contrastive updates to BERT-based document embedding. SaD splits a document into two sub-documents containing randomly shuffled words in the entire documents. The sub-documents are considered positive examples, leaving all other documents in the corpus as negatives. After SaD, we repeat the contrastive update and clustering phases until convergence. It is naturally a time-consuming, cumbersome task to label text documents, and our method can help alleviate human efforts, which are most expensive resources in AI. We have empirically evaluated our method by performing unsupervised text classification on the 20 Newsgroups, Reuters-21578, BBC, and BBCSport datasets. In particular, our method pushes the current state-of-the-art, SS-SB-MT, on 20 Newsgroups by 20.94% in accuracy. We also achieve the state-of-the-art performance on Reuters-21578 and exceptionally-high accuracy performances (over 95%) for unsupervised classification on the BBC and BBCSport datasets. | ['Youngjune Gwon', 'Jaeseon Park', 'Hoyoung Kang', 'Bogun Kim', 'Kyoungwon Park', 'Seongho Joe', 'Joonseok Lee'] | 2023-04-19 | null | null | null | null | ['document-embedding', 'text-augmentation', 'unsupervised-text-classification'] | ['methodology', 'natural-language-processing', 'natural-language-processing'] | [ 1.30530864e-01 -9.17102024e-02 -2.53583640e-01 -4.45172459e-01
-7.08340883e-01 -5.29571891e-01 1.00999570e+00 5.05992949e-01
-9.39178228e-01 7.64194250e-01 2.22735628e-01 -3.76331598e-01
1.11469693e-01 -5.13435245e-01 -2.58295149e-01 -7.47589588e-01
-2.05752477e-01 9.13251221e-01 1.37622848e-01 -4.10029113e-01
3.22896481e-01 1.51089966e-01 -1.48489964e+00 4.33847159e-01
8.48748267e-01 6.86962128e-01 -5.10661155e-02 8.31202626e-01
-5.62357485e-01 5.87006211e-01 -6.83608949e-01 -6.68220818e-01
6.79613873e-02 -4.51885968e-01 -1.08758521e+00 2.58085638e-01
4.20116186e-01 8.93730596e-02 -1.75905198e-01 8.17629337e-01
2.99307764e-01 1.63997933e-01 9.77337360e-01 -1.30307639e+00
-6.23724699e-01 9.11797404e-01 -7.74535537e-01 5.84751889e-02
8.70134011e-02 -2.62657076e-01 1.15323508e+00 -1.13474989e+00
4.33872968e-01 1.03712440e+00 5.43290555e-01 5.34131289e-01
-1.17527306e+00 -7.22614586e-01 6.54363558e-02 1.44333377e-01
-1.20579624e+00 -4.29358155e-01 7.76903212e-01 -4.02541280e-01
7.77535081e-01 4.41221535e-01 6.76941037e-01 9.30990517e-01
-7.92388767e-02 9.27349269e-01 1.03463256e+00 -9.95374739e-01
2.77862370e-01 3.88061881e-01 7.36659408e-01 5.09675026e-01
1.05666876e-01 -3.61961335e-01 -4.63720500e-01 -1.91818088e-01
9.89261270e-02 -1.02657214e-01 -1.41884893e-01 -2.15317339e-01
-1.20428991e+00 9.76053655e-01 2.73525655e-01 5.22564948e-01
4.82543232e-03 -3.43096554e-02 5.93349695e-01 6.30098343e-01
1.02142453e+00 4.60537255e-01 -6.35260701e-01 -1.68813214e-01
-8.43732417e-01 8.65439773e-02 8.80465388e-01 6.98751926e-01
6.30835593e-01 -2.54853129e-01 9.03178751e-02 1.27328706e+00
1.21659227e-01 2.33002692e-01 1.08907115e+00 -3.18570107e-01
5.40125847e-01 6.17113292e-01 -9.14436802e-02 -9.02456343e-01
-4.74995017e-01 -6.12138927e-01 -1.07749462e+00 -2.08252490e-01
1.52634501e-01 6.55609667e-02 -9.82359231e-01 1.45623398e+00
1.96736872e-01 -1.59938857e-01 1.47328675e-01 3.62700999e-01
7.62764513e-01 9.50287282e-01 -1.86539767e-03 -3.89419675e-01
1.18805003e+00 -1.34509349e+00 -7.21850693e-01 -3.08452368e-01
1.15608001e+00 -8.21140826e-01 1.46673751e+00 2.86324531e-01
-6.93535924e-01 -3.84203732e-01 -1.06929815e+00 1.66788191e-01
-7.36410320e-01 3.02138358e-01 7.44218707e-01 7.68017650e-01
-8.50523174e-01 5.20980299e-01 -5.43898761e-01 -3.44416082e-01
2.13952363e-01 3.91416669e-01 -4.52778608e-01 -6.74219653e-02
-1.21407712e+00 6.74278855e-01 6.56891942e-01 -3.17409992e-01
-3.04452449e-01 -5.03049016e-01 -7.40235031e-01 -8.60948116e-02
2.48318821e-01 -5.80947325e-02 1.10881329e+00 -8.17985058e-01
-1.29585898e+00 1.16163766e+00 -7.57175758e-02 -4.16138291e-01
3.98963720e-01 -8.49308223e-02 -6.43525362e-01 -1.02203302e-01
3.80050316e-02 8.09344947e-01 7.94543982e-01 -1.36564422e+00
-5.34163773e-01 -3.22350562e-01 -3.91217351e-01 2.61372149e-01
-1.24306285e+00 -6.17071651e-02 -5.56561232e-01 -9.35427070e-01
3.31875235e-01 -1.14948392e+00 4.70768996e-02 -3.22531492e-01
-4.49422687e-01 -5.57565212e-01 9.06670690e-01 -3.54189456e-01
1.27522278e+00 -2.28643775e+00 -1.97699398e-01 8.53060409e-02
1.86066642e-01 2.86834598e-01 -4.16280687e-01 4.99518007e-01
-3.41149479e-01 1.25106677e-01 -1.83134034e-01 -6.79654121e-01
2.94457469e-02 1.52294114e-01 -5.01908958e-01 3.26357484e-01
-5.48716970e-02 7.93204010e-01 -9.20520723e-01 -6.14910424e-01
1.61903743e-02 9.18657333e-02 -4.12905246e-01 -1.39395967e-01
-1.33378372e-01 -3.11054904e-02 3.07977796e-02 2.87620395e-01
5.12830555e-01 -2.90936410e-01 1.16610520e-01 7.71342888e-02
1.26804085e-02 4.90782380e-01 -1.04559386e+00 1.38024926e+00
-2.02790931e-01 8.35745513e-01 -4.35027659e-01 -1.28121734e+00
1.02919495e+00 7.58151561e-02 3.97313416e-01 -3.48298281e-01
1.03923023e-01 1.54154822e-01 -9.06045362e-02 -2.67847806e-01
9.64863956e-01 6.62697926e-02 -1.77446246e-01 9.70087469e-01
1.47464797e-01 -4.18284148e-01 6.78833723e-01 4.89473999e-01
8.05287540e-01 -3.23840678e-01 1.67132974e-01 -4.76479888e-01
5.20554125e-01 -2.44456492e-02 8.01398084e-02 5.49464107e-01
9.39916596e-02 4.70982969e-01 4.87147897e-01 -4.64066505e-01
-1.09888554e+00 -6.46233559e-01 -1.80097878e-01 1.56965327e+00
-3.45550328e-01 -7.34330952e-01 -7.52270401e-01 -1.12261200e+00
-1.03766974e-02 8.03976357e-01 -8.02967370e-01 -1.96564153e-01
-2.91395903e-01 -1.14684796e+00 5.14249802e-01 4.11727071e-01
5.92976630e-01 -8.44069898e-01 2.60317266e-01 8.86801779e-02
-2.48470291e-01 -7.71094799e-01 -5.72616279e-01 4.02407497e-01
-7.37314582e-01 -8.67463350e-01 -9.70837533e-01 -1.30779958e+00
9.12778497e-01 3.18675071e-01 8.90163541e-01 4.73677134e-03
-7.16165900e-02 -2.84676347e-02 -6.41229093e-01 -3.85893226e-01
-5.72161078e-01 3.87168229e-01 2.13720739e-01 3.54322717e-02
4.22630727e-01 -2.87927002e-01 -1.10106714e-01 4.84157175e-01
-9.28122342e-01 2.01863244e-01 3.60534042e-01 1.10043621e+00
4.34612691e-01 2.15935081e-01 7.15230167e-01 -1.07607257e+00
8.35277319e-01 -2.75480598e-01 -3.83215278e-01 2.80234665e-01
-8.76146257e-01 -4.81145047e-02 7.18954206e-01 -7.24504650e-01
-5.81787050e-01 -3.70453596e-02 3.93437110e-02 8.63295496e-02
8.83773938e-02 6.73622072e-01 2.64096320e-01 2.22179934e-01
9.23620582e-01 4.77110624e-01 -1.74995005e-01 -3.91981810e-01
3.37182343e-01 1.35724759e+00 5.01763165e-01 -3.69845659e-01
7.75317192e-01 2.48090222e-01 -3.21856380e-01 -9.54390824e-01
-9.67826903e-01 -6.83307827e-01 -7.88397551e-01 -3.61537864e-03
3.87027234e-01 -7.45246053e-01 -3.83755058e-01 6.29109442e-01
-8.65529716e-01 -3.34955364e-01 -2.97980875e-01 4.11575854e-01
-1.46853402e-01 5.58481455e-01 -6.78000689e-01 -8.02679598e-01
-5.29998481e-01 -6.96746111e-01 7.12804914e-01 -2.15076447e-01
-3.95578086e-01 -1.04422832e+00 2.04926312e-01 4.42775339e-01
2.51927316e-01 -2.71873742e-01 1.13209128e+00 -1.22417498e+00
2.59476513e-01 -5.45000017e-01 -2.56685279e-02 5.81726670e-01
2.59797752e-01 -7.70247579e-02 -9.23140526e-01 -4.37092334e-01
-2.38968134e-01 -6.58148646e-01 1.09920371e+00 -1.50398493e-01
1.34280336e+00 -3.47312301e-01 -2.65876681e-01 1.94066986e-01
8.74329925e-01 2.03227326e-01 5.21808088e-01 5.42881489e-01
5.19322693e-01 6.44997597e-01 6.55720949e-01 3.35795611e-01
3.39567244e-01 4.01921034e-01 -1.15353160e-01 -2.14454323e-01
9.86811966e-02 -7.67199472e-02 1.45098880e-01 1.53858626e+00
1.56844571e-01 -5.79820335e-01 -1.19746768e+00 6.60003364e-01
-1.70110404e+00 -6.67513251e-01 -2.93107331e-01 2.13021851e+00
1.41023183e+00 3.65039051e-01 -4.74107154e-02 8.09724927e-01
7.14683771e-01 1.58152103e-01 -1.93427727e-01 -1.38408512e-01
-2.20122755e-01 1.43366441e-01 2.00060591e-01 3.32259655e-01
-1.37303007e+00 1.09004986e+00 6.11172819e+00 8.63580942e-01
-9.43354070e-01 -4.81903739e-02 8.98169875e-01 -7.57462978e-02
-1.23650365e-01 -2.78974742e-01 -8.80533397e-01 6.09720528e-01
9.53125715e-01 -1.41841680e-01 2.29288384e-01 8.09809566e-01
-3.21736306e-01 -6.81853369e-02 -1.11530197e+00 1.06914103e+00
5.21838963e-01 -1.14638114e+00 1.75990298e-01 -1.47731259e-01
1.09711838e+00 4.04763259e-02 8.39912668e-02 5.89844346e-01
4.16579336e-01 -6.98408663e-01 3.31916243e-01 -2.16401458e-01
9.03036654e-01 -1.07493079e+00 8.72810125e-01 4.44710344e-01
-7.55627692e-01 4.11727317e-02 -3.83854538e-01 1.38467833e-01
-2.37041935e-01 9.05633092e-01 -9.23965156e-01 1.27130494e-01
6.39736056e-01 6.04178071e-01 -7.68166661e-01 5.96434236e-01
-1.94430396e-01 6.95754468e-01 -1.64189741e-01 -5.39202392e-01
3.24640751e-01 -2.58848891e-02 1.83331504e-01 1.43791366e+00
-1.69460997e-01 -3.51073556e-02 1.60380825e-01 7.52632692e-02
-5.15581310e-01 5.36960721e-01 -2.34219566e-01 -3.84325564e-01
3.26125085e-01 1.27125704e+00 -9.68335330e-01 -8.21067989e-01
-1.14105016e-01 1.03758359e+00 3.60262334e-01 2.77469158e-01
-5.17159820e-01 -9.12610471e-01 9.03158411e-02 -1.18268505e-01
1.05728567e-01 -2.13862911e-01 -1.74463734e-01 -1.24932802e+00
1.96664557e-02 -9.34570134e-01 4.29814488e-01 -5.63835740e-01
-1.44709206e+00 8.55345368e-01 -1.55981496e-01 -1.24221504e+00
-4.37598854e-01 -5.12186170e-01 -2.81110317e-01 4.19435710e-01
-1.41350186e+00 -8.86073828e-01 -3.54375601e-01 5.16962290e-01
6.13901198e-01 -5.73635101e-01 1.06301415e+00 1.77882642e-01
-4.76321489e-01 8.80770981e-01 6.76008940e-01 3.97640020e-01
9.95191991e-01 -1.24408865e+00 7.07758009e-01 4.07419950e-01
3.93474936e-01 5.50431788e-01 3.44760656e-01 -6.02427363e-01
-9.23784077e-01 -1.07168591e+00 1.33262622e+00 -1.70191273e-01
8.94811511e-01 -7.49005854e-01 -9.70225990e-01 6.55525982e-01
1.59593940e-01 -1.83208823e-01 1.05008483e+00 3.60756457e-01
-4.38561738e-01 -6.43790141e-02 -8.87892246e-01 8.22194934e-01
7.50154018e-01 -5.31776071e-01 -9.44499314e-01 7.97053814e-01
6.04953885e-01 -7.74594769e-02 -5.42369306e-01 1.58500910e-01
3.87478709e-01 -3.59906137e-01 6.96904302e-01 -6.66844606e-01
6.07183993e-01 5.48081659e-03 -7.12950453e-02 -1.63031745e+00
-2.17081010e-01 -6.51820898e-01 9.96289998e-02 1.44676030e+00
5.49596548e-01 -7.81870782e-01 9.82256889e-01 2.45344549e-01
-1.47560274e-03 -6.60914302e-01 -6.70870781e-01 -9.22733128e-01
1.74601302e-01 -5.27116656e-01 3.24450582e-01 1.68827355e+00
3.57030541e-01 6.15298510e-01 -1.96050793e-01 -5.88574231e-01
4.46961522e-01 -6.80296645e-02 8.27270925e-01 -1.44265783e+00
-8.19221288e-02 -3.91175032e-01 -2.09430993e-01 -9.76033330e-01
3.71184975e-01 -1.17794502e+00 -2.74708457e-02 -1.33751011e+00
3.61586124e-01 -6.27076864e-01 -1.85876340e-01 7.76092649e-01
-1.83918491e-01 4.65334207e-01 6.68259617e-03 4.44448531e-01
-5.32772660e-01 7.19976425e-01 9.51199174e-01 -4.74918485e-01
-4.64776844e-01 -1.33691087e-01 -6.06413007e-01 7.58756757e-01
9.12313521e-01 -5.71524858e-01 -3.41784149e-01 -3.68634790e-01
1.16669610e-01 -5.28678715e-01 -1.44955829e-01 -8.52261662e-01
3.88008952e-01 1.29807800e-01 5.02707779e-01 -9.13429558e-01
3.06499869e-01 -5.88953137e-01 -4.86355901e-01 5.42920649e-01
-7.69304693e-01 8.00739974e-02 2.00241581e-01 2.61321425e-01
-2.22271666e-01 -5.10782421e-01 6.98428988e-01 1.72541782e-01
-3.15749496e-01 7.51921833e-02 -4.69419658e-01 5.75690344e-02
7.99734950e-01 -2.52453890e-02 -4.00854319e-01 -3.38235974e-01
-5.80116034e-01 2.84357965e-01 1.71512350e-01 3.98206383e-01
4.92248654e-01 -1.34623420e+00 -9.46949959e-01 4.76152152e-01
3.79684836e-01 -1.57500327e-01 -2.03155875e-01 4.91662115e-01
-3.75328809e-01 5.09081662e-01 1.86698407e-01 -5.57366192e-01
-1.46080065e+00 3.88710827e-01 -2.18565494e-01 -4.27215934e-01
-3.72372866e-01 8.38187873e-01 -1.22213215e-01 -6.98212743e-01
4.79707837e-01 -7.13592693e-02 -3.69388908e-01 5.30462086e-01
6.59219444e-01 2.77992517e-01 3.20369452e-01 -3.09862018e-01
-1.60501227e-01 4.23585951e-01 -7.29875088e-01 -4.21473891e-01
1.44757128e+00 4.60417606e-02 -2.96677947e-01 7.63364911e-01
1.46720088e+00 4.88707684e-02 -4.35555547e-01 -4.85026211e-01
1.40309304e-01 -2.93193549e-01 2.49315664e-01 -7.91308880e-01
-6.77408993e-01 6.55184805e-01 5.21460056e-01 4.19754028e-01
8.96951795e-01 4.60491143e-02 6.41499341e-01 1.03979790e+00
1.14928447e-01 -1.43379068e+00 3.47130537e-01 9.08913195e-01
8.23733926e-01 -1.26888931e+00 2.06567988e-01 -3.75266045e-01
-6.67403281e-01 1.01750660e+00 4.05587763e-01 1.70454487e-01
6.40297234e-01 1.08798660e-01 7.57148042e-02 9.08458009e-02
-8.21242273e-01 8.68189782e-02 3.74228805e-01 2.72343397e-01
6.19492829e-01 -1.51056170e-01 -4.60104793e-01 3.87270719e-01
-4.95306104e-01 -3.72509211e-01 4.53477800e-01 1.07653797e+00
-5.33129156e-01 -1.13618422e+00 -2.43897095e-01 6.51472151e-01
-3.78173202e-01 -3.42151821e-01 -7.58400381e-01 9.61693466e-01
-2.35607848e-01 8.23625326e-01 4.39912677e-01 -4.69810247e-01
2.50035711e-03 4.94871557e-01 1.18353285e-01 -6.14192128e-01
-4.77900058e-01 4.66066599e-02 1.86803088e-01 2.43889719e-01
-2.83829957e-01 -3.88976276e-01 -1.23272085e+00 -3.47437769e-01
-6.95141375e-01 5.03522575e-01 9.77173567e-01 8.38427961e-01
5.16783409e-02 2.81853706e-01 1.01369154e+00 -6.64867103e-01
-5.28133988e-01 -1.47660053e+00 -5.20145416e-01 6.35872900e-01
1.37519136e-01 -3.54187638e-01 -6.73719823e-01 1.78900659e-01] | [10.538511276245117, 7.606536865234375] |
512ef07b-aa70-4f0c-b3d9-5cabdea41d6e | adaptive-behavior-cloning-regularization-for-1 | 2210.13846 | null | https://arxiv.org/abs/2210.13846v1 | https://arxiv.org/pdf/2210.13846v1.pdf | Adaptive Behavior Cloning Regularization for Stable Offline-to-Online Reinforcement Learning | Offline reinforcement learning, by learning from a fixed dataset, makes it possible to learn agent behaviors without interacting with the environment. However, depending on the quality of the offline dataset, such pre-trained agents may have limited performance and would further need to be fine-tuned online by interacting with the environment. During online fine-tuning, the performance of the pre-trained agent may collapse quickly due to the sudden distribution shift from offline to online data. While constraints enforced by offline RL methods such as a behaviour cloning loss prevent this to an extent, these constraints also significantly slow down online fine-tuning by forcing the agent to stay close to the behavior policy. We propose to adaptively weigh the behavior cloning loss during online fine-tuning based on the agent's performance and training stability. Moreover, we use a randomized ensemble of Q functions to further increase the sample efficiency of online fine-tuning by performing a large number of learning updates. Experiments show that the proposed method yields state-of-the-art offline-to-online reinforcement learning performance on the popular D4RL benchmark. Code is available: \url{https://github.com/zhaoyi11/adaptive_bc}. | ['Joni Pajarinen', 'Juho Kannala', 'Alexander Ilin', 'Rinu Boney', 'Yi Zhao'] | 2022-10-25 | adaptive-behavior-cloning-regularization-for | https://openreview.net/forum?id=JVsvIuMDE0Z | https://openreview.net/pdf?id=JVsvIuMDE0Z | null | ['d4rl'] | ['robots'] | [-4.71821785e-01 5.27430028e-02 -3.00879449e-01 -2.69908488e-01
-7.69180715e-01 -7.45457292e-01 4.48586106e-01 2.57192343e-01
-7.49939919e-01 1.08008504e+00 -2.74041802e-01 -9.43154693e-02
-2.16924533e-01 -7.97930956e-01 -9.28731203e-01 -9.87259746e-01
-2.12824583e-01 7.54780412e-01 2.68675655e-01 -2.41742402e-01
6.67040125e-02 2.62953073e-01 -1.50757360e+00 -3.10814708e-01
9.37050760e-01 8.15713048e-01 3.26262802e-01 6.16573989e-01
2.55266160e-01 7.87381411e-01 -8.05999458e-01 -2.53786659e-03
5.60997486e-01 -4.81625468e-01 -4.53505546e-01 1.57772422e-01
-7.28751123e-02 -8.25634241e-01 -4.55353230e-01 1.12613451e+00
5.53039372e-01 4.78388280e-01 2.70495322e-02 -1.13914788e+00
-4.73607749e-01 9.15188372e-01 -5.20796657e-01 7.83610344e-02
4.19850983e-02 6.83517992e-01 8.48846078e-01 -2.74104416e-01
4.85691905e-01 9.77333665e-01 3.43431562e-01 7.65619457e-01
-1.12555611e+00 -7.83723474e-01 5.87248981e-01 8.65277350e-02
-1.09755206e+00 -4.27519977e-01 6.49521232e-01 -1.15393631e-01
8.27530742e-01 -1.79925576e-01 7.73973763e-01 9.96809781e-01
-7.03947917e-02 8.11853766e-01 1.10274577e+00 -4.12963964e-02
7.52843857e-01 1.02411369e-02 -2.98428893e-01 8.12312365e-01
1.35515600e-01 6.06301904e-01 -3.83646637e-01 -3.18612546e-01
8.85482848e-01 -1.67600274e-01 -1.16667725e-01 -6.62767112e-01
-8.87020767e-01 8.64608526e-01 3.66175830e-01 -2.60285698e-02
-4.85276461e-01 5.14185309e-01 6.48757756e-01 7.34467745e-01
1.70862868e-01 6.33607507e-01 -6.92622721e-01 -6.72410071e-01
-4.60039377e-01 4.31992024e-01 6.48317575e-01 7.40735531e-01
9.28237915e-01 1.36688218e-01 -2.01331794e-01 8.87426436e-01
3.29342484e-02 4.01398301e-01 5.57693660e-01 -1.26226771e+00
5.81367373e-01 4.66474265e-01 5.40306568e-01 -4.16341454e-01
-3.13963026e-01 -3.71964216e-01 -4.21127439e-01 6.16209447e-01
6.65611088e-01 -5.74023664e-01 -6.60822093e-01 2.05129790e+00
6.98018491e-01 -9.11344681e-03 2.90691517e-02 1.03609300e+00
-8.85756835e-02 4.45748121e-01 -1.47376895e-01 -5.26736498e-01
6.26594067e-01 -1.19240415e+00 -5.06388843e-01 -1.02082826e-01
6.14053667e-01 -3.16948444e-01 1.34613180e+00 3.98357093e-01
-1.19645238e+00 -4.76707399e-01 -9.87688899e-01 5.32349646e-01
1.34677216e-02 8.18525907e-03 5.28398633e-01 4.74900424e-01
-8.73197377e-01 8.70929956e-01 -1.27221525e+00 -1.64899141e-01
4.22211647e-01 5.71337461e-01 9.80226696e-02 2.80432165e-01
-1.01801443e+00 7.11441696e-01 5.04674673e-01 -1.38318539e-01
-1.39688897e+00 -6.28430188e-01 -4.34505135e-01 -8.32738206e-02
8.62540662e-01 -6.01097286e-01 1.69710338e+00 -1.10587311e+00
-2.23154974e+00 2.39078134e-01 3.01304936e-01 -6.61017418e-01
9.98105645e-01 -1.86518192e-01 -7.86351338e-02 -4.11944091e-03
-5.79235926e-02 5.05328655e-01 1.00673652e+00 -1.23680437e+00
-8.42963874e-01 -3.24061006e-01 5.66649139e-01 5.66958725e-01
-4.04036403e-01 -5.11708736e-01 -3.83868128e-01 -4.62828368e-01
-6.07910573e-01 -1.16207385e+00 -3.22275490e-01 -2.18676224e-01
1.06203355e-01 -3.08248132e-01 7.18613386e-01 -1.49093196e-01
1.20201731e+00 -2.09416485e+00 1.37414439e-02 2.18444169e-01
4.04021069e-02 1.70443654e-01 -3.95949632e-01 5.11361480e-01
3.57145399e-01 -2.39949644e-01 -2.57573184e-02 -1.21111847e-01
1.56336963e-01 3.52591544e-01 -1.91336498e-01 5.97232640e-01
-2.64745265e-01 7.14843392e-01 -1.14304709e+00 -3.73504460e-02
2.03335568e-01 -2.87194736e-02 -9.12594318e-01 5.09515047e-01
-6.30736291e-01 6.59525156e-01 -7.43309617e-01 2.87423223e-01
4.30980444e-01 -2.67316431e-01 4.45863426e-01 3.73976469e-01
-3.29295360e-02 1.56099290e-01 -1.12978148e+00 1.58007050e+00
-6.75106704e-01 2.44543273e-02 2.43034497e-01 -9.15397465e-01
6.76749647e-01 7.78818205e-02 5.80296278e-01 -9.46759343e-01
1.85781911e-01 1.20999388e-01 1.51987135e-01 -3.48384827e-01
3.45555693e-01 2.48347372e-01 2.92156339e-02 7.36110628e-01
-7.75287524e-02 2.30920613e-02 4.90808100e-01 -1.93569064e-01
1.16501343e+00 4.02378738e-01 2.37855554e-01 -1.40334263e-01
3.35937351e-01 8.41308981e-02 6.82831466e-01 9.99838769e-01
-3.73488307e-01 -2.13603452e-01 3.77037823e-01 -3.89592201e-01
-1.18047047e+00 -8.39784741e-01 1.99485719e-01 1.58790493e+00
2.07859918e-01 -1.52673125e-01 -8.25185418e-01 -9.52136278e-01
3.30454707e-01 6.67146444e-01 -6.61721170e-01 -3.43129307e-01
-7.90734410e-01 -6.69802189e-01 3.97487551e-01 4.21875417e-01
5.44736505e-01 -1.14959145e+00 -9.48239803e-01 6.00999594e-01
1.70645475e-01 -7.11644411e-01 -6.87699199e-01 3.66055518e-01
-9.17586923e-01 -9.53525960e-01 -5.08434296e-01 -4.18245673e-01
7.86289573e-01 -1.82797189e-03 8.00624728e-01 1.56182975e-01
5.95866442e-02 4.71693188e-01 -4.01019216e-01 1.56297407e-03
-5.17830431e-01 1.74518287e-01 4.48884070e-01 -3.24609786e-01
-4.93627004e-02 -5.42935729e-01 -7.59128749e-01 4.36528534e-01
-6.77298367e-01 -2.36736715e-01 3.13417286e-01 1.11097169e+00
5.64761817e-01 3.56031358e-01 7.96972811e-01 -7.86939919e-01
7.70742297e-01 -3.63663942e-01 -1.15005553e+00 2.12694436e-01
-7.32851207e-01 3.49020749e-01 1.24584568e+00 -9.80253756e-01
-1.02014887e+00 -1.04729436e-01 9.73147750e-02 -6.46755040e-01
1.23297125e-01 1.59579620e-01 1.47329509e-01 -5.89863434e-02
6.64716423e-01 2.83357590e-01 2.33468965e-01 -2.48263255e-01
4.11689758e-01 3.40388596e-01 1.72184035e-01 -8.97470474e-01
8.72522533e-01 2.38890558e-01 -4.20996785e-01 -2.70988494e-01
-7.83429384e-01 -8.42909813e-02 -9.03615262e-03 -2.88567275e-01
3.14231217e-01 -7.77484894e-01 -1.28671861e+00 5.00828028e-01
-3.45654041e-01 -1.24157012e+00 -5.84566057e-01 3.58405143e-01
-8.56139183e-01 5.74660823e-02 -6.62842274e-01 -7.84945190e-01
-2.07420737e-01 -1.07909131e+00 5.01983047e-01 4.73672181e-01
1.40252575e-01 -9.69551980e-01 3.68638843e-01 1.98858947e-01
3.83384228e-01 -1.20608389e-01 5.76163828e-01 -4.04031366e-01
-4.51461405e-01 2.92927325e-01 2.49877617e-01 2.04470009e-01
3.13131928e-01 -2.31974721e-01 -6.83275938e-01 -9.93650377e-01
-1.01432398e-01 -8.95301223e-01 6.06701136e-01 2.45071441e-01
1.24451816e+00 -6.90815568e-01 -7.54364058e-02 5.05835235e-01
1.35720658e+00 4.24088299e-01 1.37428939e-01 7.53171206e-01
4.54204232e-01 1.35367662e-01 1.04102933e+00 1.14391541e+00
4.09943581e-01 5.48099339e-01 4.92138118e-01 4.55064237e-01
1.24521062e-01 -5.83466768e-01 7.35709965e-01 6.56005859e-01
7.61841312e-02 -5.65778576e-02 -5.46284974e-01 4.19389963e-01
-2.13145304e+00 -8.16925168e-01 7.15052843e-01 2.52018237e+00
1.24991381e+00 3.06621909e-01 4.67821419e-01 -3.43747675e-01
5.08636236e-01 2.87388954e-02 -1.37096202e+00 -3.22465867e-01
2.47993737e-01 -1.34051129e-01 8.62591982e-01 6.16672754e-01
-8.69127452e-01 1.05227208e+00 5.49477530e+00 9.81892645e-01
-1.28444803e+00 1.29444987e-01 6.75884664e-01 -4.93465573e-01
-1.56687438e-01 -1.29857838e-01 -8.71153355e-01 6.55883491e-01
8.89431775e-01 -3.05591702e-01 1.33406222e+00 1.00858665e+00
5.12899697e-01 -2.52018452e-01 -1.10724711e+00 7.20217586e-01
-4.37273443e-01 -1.07374501e+00 -3.15317988e-01 2.06807870e-02
9.98090804e-01 2.81556129e-01 4.77616750e-02 6.99346900e-01
8.10838640e-01 -6.58017516e-01 7.90167511e-01 2.04009950e-01
6.21285856e-01 -9.96029615e-01 3.54240298e-01 7.09334016e-01
-1.08160222e+00 -5.37471831e-01 -4.76330370e-01 -1.37308508e-01
-2.35542402e-01 2.26300806e-01 -8.49997163e-01 1.04066841e-01
7.13890374e-01 2.87745297e-01 -3.72033447e-01 7.83715248e-01
-2.69831628e-01 5.96741855e-01 -4.48401421e-01 -3.16426724e-01
3.50303859e-01 -3.58086675e-01 4.21157956e-01 5.47575414e-01
1.29736915e-01 1.42650604e-01 4.80782539e-01 6.88110471e-01
-5.95729873e-02 -1.04186051e-01 -3.29620987e-01 -2.44077787e-01
7.19852388e-01 9.97547925e-01 -4.46360707e-01 -2.81144828e-01
-1.74495116e-01 7.52116919e-01 9.08124983e-01 4.06455040e-01
-1.06027663e+00 -1.43332720e-01 7.33642817e-01 -1.52939931e-01
5.00971675e-01 -3.55018705e-01 1.29684672e-01 -1.05757272e+00
1.28529000e-03 -1.13012183e+00 4.24398452e-01 -1.78005084e-01
-1.29267538e+00 3.08495343e-01 -2.83513457e-01 -1.31473970e+00
-4.21407312e-01 -7.25045055e-02 -3.14881027e-01 1.89636514e-01
-1.40212071e+00 -5.29808402e-01 -6.93032742e-02 6.98635757e-01
7.64936686e-01 -2.59415776e-01 5.61139286e-01 1.78426728e-01
-6.38556182e-01 8.85930359e-01 6.45017862e-01 -2.52549767e-01
8.14567089e-01 -1.27022529e+00 1.42103166e-03 3.57615024e-01
-1.50166973e-01 3.78477573e-01 7.82453895e-01 -5.57819128e-01
-1.63594639e+00 -1.04575205e+00 -1.42176911e-01 -4.75937091e-02
9.39085424e-01 -1.72972843e-01 -8.04580808e-01 5.10485530e-01
-1.09785153e-02 4.55088206e-02 2.24781647e-01 -6.24107234e-02
-1.50930181e-01 -3.91443104e-01 -1.17004800e+00 8.23000073e-01
9.86268044e-01 -2.36249000e-01 -2.08272502e-01 2.03978658e-01
6.25272751e-01 -6.33011758e-01 -9.42533255e-01 1.56256527e-01
3.47696900e-01 -8.19327533e-01 5.92640102e-01 -4.51000661e-01
5.47099002e-02 -3.71683508e-01 2.16459230e-01 -1.66110063e+00
-3.27136070e-01 -9.26506400e-01 -4.25376922e-01 9.58492815e-01
3.00736427e-01 -9.63449359e-01 9.24317718e-01 4.32595342e-01
1.80441812e-01 -9.20428693e-01 -8.55106413e-01 -1.03600073e+00
1.25695840e-01 -7.14010596e-02 6.16605997e-01 6.64868534e-01
1.26959473e-01 -2.77916193e-02 -3.72664422e-01 2.30565265e-01
7.52655029e-01 3.27423513e-01 9.14470732e-01 -6.17868304e-01
-9.12382782e-01 -4.57037032e-01 2.82458276e-01 -1.42573977e+00
1.38826743e-01 -4.41951245e-01 1.79850698e-01 -9.01390016e-01
1.85806677e-01 -9.20262694e-01 -4.53450769e-01 6.45662308e-01
-1.21538185e-01 -1.74314752e-01 2.88836926e-01 3.17581594e-01
-1.01029289e+00 1.03763390e+00 1.62601495e+00 3.93448509e-02
-6.23204589e-01 2.88798641e-02 -3.75420809e-01 5.07445931e-01
1.09545767e+00 -4.68447208e-01 -5.95627844e-01 -3.86948645e-01
1.65315196e-01 4.37055856e-01 -9.18612182e-02 -7.67048299e-01
1.96137697e-01 -5.99165559e-01 1.75402194e-01 -2.06096340e-02
1.94548756e-01 -6.50425911e-01 -1.48532793e-01 7.93857574e-01
-5.38500130e-01 1.39763579e-02 1.30789012e-01 7.04308152e-01
9.81442332e-02 -1.68217286e-01 1.01442778e+00 -8.68572965e-02
-4.67268556e-01 5.18023729e-01 -4.19543564e-01 3.57498735e-01
1.19223559e+00 -1.08590806e-02 -3.36420387e-01 -3.10021192e-01
-5.87727129e-01 8.29729021e-01 6.99923515e-01 4.54345345e-01
3.04302692e-01 -1.16799378e+00 -3.46739978e-01 1.07414655e-01
-1.10645061e-02 -9.35939550e-02 2.96703696e-01 6.18347287e-01
-2.35431179e-01 8.29350948e-02 -3.22968483e-01 -3.21888357e-01
-8.32996905e-01 6.28729820e-01 4.96952683e-01 -4.35478359e-01
-6.29671454e-01 6.94732130e-01 9.86155793e-02 -6.66889548e-01
5.46566367e-01 -3.10855150e-01 -5.24212606e-03 -1.35682374e-01
3.87195915e-01 4.80039209e-01 -2.16252476e-01 5.23999929e-02
-2.29003638e-01 3.15944254e-01 -2.30616942e-01 -1.75276309e-01
1.35178101e+00 -2.72057801e-01 3.89670581e-01 4.12964046e-01
8.69645000e-01 -1.63685456e-01 -2.11397505e+00 -2.20520779e-01
-2.77341485e-01 -4.82438356e-01 -1.24753471e-02 -8.86397600e-01
-1.08093417e+00 1.60778865e-01 6.28733039e-01 2.25190923e-01
1.11367762e+00 -1.97601989e-01 7.40816712e-01 8.08639884e-01
7.61025846e-01 -1.80204463e+00 4.33109134e-01 6.67014837e-01
6.59244478e-01 -1.26732922e+00 -1.38275817e-01 3.69502664e-01
-7.74086237e-01 1.01428485e+00 8.78573298e-01 -5.01965463e-01
3.41703445e-01 3.52707267e-01 2.91791167e-02 2.35744864e-01
-1.12990844e+00 -6.89710304e-02 -4.46942687e-01 3.69763851e-01
-2.08624378e-01 3.18644345e-01 -1.27073303e-01 2.66437858e-01
-2.38969177e-01 4.91321348e-02 4.83366400e-01 9.96934056e-01
-5.63185453e-01 -1.23504400e+00 -2.33476192e-01 3.62870961e-01
-2.83092171e-01 4.32308555e-01 2.27373261e-02 6.46726370e-01
-1.31348163e-01 8.06922317e-01 -2.99025830e-02 -8.77905041e-02
1.24562785e-01 -3.65278155e-01 7.31172562e-01 -2.89191723e-01
-7.09071338e-01 1.58268973e-01 7.98542351e-02 -8.66083264e-01
-2.44968943e-02 -6.19980276e-01 -1.60667861e+00 -4.94768769e-01
-2.41485953e-01 2.07908750e-01 2.99661368e-01 7.71507204e-01
2.75617510e-01 4.45561230e-01 1.05935073e+00 -6.77618861e-01
-1.39502299e+00 -8.27422440e-01 -4.67077911e-01 2.55209088e-01
4.99400944e-01 -8.17387402e-01 -3.50647926e-01 -2.96877384e-01] | [3.94535756111145, 2.2279961109161377] |
f307f080-0a63-4ba4-8b0d-e6a6f0672404 | dynamic-sensor-placement-based-on-graph | 2211.04019 | null | https://arxiv.org/abs/2211.04019v1 | https://arxiv.org/pdf/2211.04019v1.pdf | Dynamic Sensor Placement Based on Graph Sampling Theory | In this paper, we consider a dynamic sensor placement problem where sensors can move within a network over time. Sensor placement problem aims to select M sensor positions from N candidates where M < N. Most existing methods assume that sensors are static, i.e., they do not move, however, many mobile sensors like drones, robots, and vehicles can change their positions over time. Moreover, underlying measurement conditions could also be changed that are difficult to cover the statically placed sensors. We tackle the problem by allowing the sensors to change their positions in their neighbors on the network. Based on a perspective of dictionary learning, we sequentially learn the dictionary from a pool of observed signals on the network based on graph sampling theory. Using the learned dictionary, we dynamically determine the sensor positions such that the non-observed signals on the network can be best recovered from the observations. Furthermore, sensor positions in each time slot can be optimized in a decentralized manner to reduce the calculation cost. In experiments, we validate the effectiveness of the proposed method via the mean squared error (MSE) of the reconstructed signals. The proposed dynamic sensor placement outperforms the existing static ones both in synthetic and real data. | ['Yuichi Tanaka', 'Junya Hara', 'Saki Nomura'] | 2022-11-08 | null | null | null | null | ['graph-sampling'] | ['graphs'] | [ 4.26351011e-01 2.16065094e-01 -3.06734681e-01 -1.46006057e-02
-4.19024944e-01 -8.18082750e-01 9.97858495e-03 4.70736474e-01
-4.46245015e-01 6.27014041e-01 -1.05788313e-01 3.28920968e-02
-5.30390859e-01 -1.01628351e+00 -9.20862436e-01 -1.00807822e+00
-3.44587296e-01 3.97733539e-01 3.56893927e-01 -2.40393400e-01
4.57158461e-02 4.85462159e-01 -9.20418382e-01 -6.41722918e-01
5.33705652e-01 9.91875410e-01 5.46438932e-01 3.67569923e-01
5.20373225e-01 1.44263789e-01 -6.77989304e-01 2.56258279e-01
5.21469533e-01 -1.88232914e-01 2.01879382e-01 4.47657853e-01
-9.71423015e-02 -9.63184163e-02 -5.94799459e-01 1.48210812e+00
4.04793471e-01 1.42133161e-01 1.12810962e-01 -1.15505946e+00
-8.12207907e-02 7.32990146e-01 -6.68923140e-01 1.21164851e-01
3.49604487e-01 -2.72905439e-01 6.12179399e-01 -3.77042174e-01
5.03132820e-01 8.87361467e-01 5.26752710e-01 3.78334761e-01
-9.60448086e-01 -7.35666275e-01 4.56578135e-01 2.97044851e-02
-1.51909912e+00 -5.16557157e-01 1.12598300e+00 -4.68571745e-02
-5.80490604e-02 2.63940543e-01 7.27705479e-01 6.15083218e-01
1.75998345e-01 2.37738445e-01 3.79744947e-01 -1.42586797e-01
4.51791376e-01 -1.43530175e-01 -2.98694283e-01 5.67272782e-01
9.43647683e-01 -1.22358412e-01 -4.51084346e-01 -2.21987724e-01
5.76921940e-01 6.02228761e-01 -6.25492275e-01 -6.47674441e-01
-1.64645517e+00 6.77937031e-01 5.87213695e-01 3.07350576e-01
-7.39170372e-01 4.12027538e-02 3.27709429e-02 4.14150655e-01
-1.29685208e-01 8.06413740e-02 -3.24167222e-01 2.69928962e-01
-5.88051856e-01 -1.78515449e-01 6.56465292e-01 1.09506202e+00
9.23315525e-01 1.23101220e-01 2.64325231e-01 3.87238383e-01
4.59010005e-01 1.13954914e+00 1.33090645e-01 -8.20272207e-01
7.91894853e-01 3.61300170e-01 4.84724790e-01 -1.81674647e+00
-5.94909370e-01 -4.57108378e-01 -1.17165685e+00 -3.52738857e-01
9.86765251e-02 -8.86858702e-01 -6.02146864e-01 1.62828398e+00
6.98574901e-01 5.53290904e-01 3.23674381e-01 1.08392692e+00
2.36635238e-01 7.25617707e-01 -5.79322994e-01 -8.42925727e-01
6.95497870e-01 -4.57377821e-01 -8.57300043e-01 -4.16460484e-01
2.24273682e-01 -3.61175418e-01 -1.92182269e-02 4.82135683e-01
-8.47928345e-01 -1.77241236e-01 -1.34282935e+00 1.04155922e+00
1.44803882e-01 2.21759140e-01 8.56063813e-02 2.97337025e-01
-8.78314972e-01 2.27747709e-01 -1.22713804e+00 -3.27403456e-01
-1.26142636e-01 5.56399941e-01 -6.42579840e-03 -3.15556407e-01
-1.01529801e+00 2.11923540e-01 4.41417277e-01 3.91659468e-01
-1.11124039e+00 -7.53012374e-02 -8.19202960e-01 -2.09368616e-01
9.20362234e-01 -3.25350016e-01 8.03040206e-01 -6.78011715e-01
-1.24069405e+00 -7.45293200e-02 -3.76725584e-01 -5.21141112e-01
1.70141861e-01 3.17740411e-01 -6.42514467e-01 2.08671153e-01
9.40751359e-02 2.15546906e-01 6.74184024e-01 -1.51875103e+00
-9.75002229e-01 -4.88148868e-01 1.70317024e-01 2.97491401e-01
-4.69981134e-01 -9.02661264e-01 -5.79274595e-01 -4.24101293e-01
8.15727651e-01 -1.29571354e+00 -6.54759705e-01 -8.91340896e-02
-5.62294245e-01 3.07069421e-01 9.33866441e-01 -1.51812702e-01
1.03581059e+00 -2.25743389e+00 3.49963069e-01 7.63919592e-01
2.53647953e-01 -1.31482840e-01 -8.60747620e-02 6.76706731e-01
5.03496230e-01 -2.49683514e-01 -3.24700959e-02 -3.09539348e-01
-3.78887326e-01 5.74671924e-01 -2.20201593e-02 8.25489938e-01
-7.61848271e-01 2.71911174e-01 -1.14381373e+00 -1.09928541e-01
2.12486178e-01 2.28504390e-01 -1.88022792e-01 -2.57043950e-02
-1.11905307e-01 7.67553866e-01 -8.40823889e-01 5.94786048e-01
7.87694633e-01 -2.09923118e-01 5.95169127e-01 -1.35257989e-01
-8.10536221e-02 -1.25454828e-01 -1.58017707e+00 1.61586356e+00
-2.69263238e-01 3.08984309e-01 7.26055443e-01 -1.37581408e+00
9.19996440e-01 5.09097219e-01 1.09696198e+00 -3.54748517e-01
8.55884627e-02 3.01521063e-01 9.89703834e-02 -2.65635818e-01
2.28464663e-01 3.49936306e-01 -2.14621872e-01 1.46338433e-01
-7.06235230e-01 3.12245548e-01 1.01379976e-01 -3.42062600e-02
1.38902259e+00 -9.98031437e-01 2.90039241e-01 4.55491878e-02
4.97004211e-01 4.34508845e-02 1.00200665e+00 8.78888369e-01
6.12900034e-02 -5.99472318e-03 -2.30494261e-01 -1.23997621e-01
-4.94586796e-01 -7.87883461e-01 1.80065334e-01 4.63222414e-01
1.04939544e+00 -7.24752173e-02 -3.28785449e-01 -3.31590533e-01
2.28959218e-01 1.09971002e-01 -3.85997325e-01 -1.32320195e-01
-4.36943203e-01 -4.15329933e-01 2.10237671e-02 -2.26471037e-01
5.10742188e-01 -4.65989798e-01 -9.52181637e-01 5.68421960e-01
-9.02893580e-03 -1.21718991e+00 -4.32175517e-01 3.76547202e-02
-8.34135056e-01 -1.26448464e+00 -4.52177286e-01 -7.89141297e-01
1.10194468e+00 1.11873555e+00 3.22743028e-01 2.26410162e-02
4.64497060e-01 7.30486274e-01 -5.84225178e-01 -4.08588767e-01
-3.25556919e-02 9.23991501e-02 3.55790854e-01 3.90176415e-01
-5.74579760e-02 -5.93275726e-01 -7.49767423e-01 4.92394477e-01
-9.31932509e-01 -3.44215363e-01 5.78147352e-01 5.77942908e-01
9.82741475e-01 8.50294769e-01 5.31127751e-01 -6.29226208e-01
4.51613486e-01 -8.97537172e-01 -9.23236310e-01 7.14336485e-02
-3.53853256e-01 -2.86844313e-01 5.91585517e-01 -7.62342572e-01
-1.79343551e-01 3.72775942e-01 4.15174156e-01 -5.07749498e-01
3.39076668e-01 7.98900366e-01 -4.86153752e-01 -4.70758498e-01
2.27036968e-01 2.12962449e-01 -1.37811139e-01 -2.81840831e-01
-2.06874236e-02 4.00843948e-01 3.40027153e-01 -5.59532158e-02
1.18637669e+00 7.80907750e-01 4.66638297e-01 -8.08902264e-01
-4.62119758e-01 -3.82284611e-01 -1.50287777e-01 -2.60940492e-01
1.46353349e-01 -1.12489927e+00 -7.42117047e-01 2.22849548e-01
-9.64624941e-01 4.28822450e-03 -3.60803716e-02 4.42128628e-01
2.22924352e-01 2.92896509e-01 4.28346172e-02 -1.02030265e+00
2.68277619e-02 -1.00759280e+00 1.06917965e+00 3.08358818e-01
8.28871727e-02 -1.30298269e+00 -6.63693175e-02 -2.42049545e-01
2.92139500e-01 8.61481726e-01 1.50813788e-01 -4.76635665e-01
-8.86399209e-01 -5.06424844e-01 5.46243668e-01 -2.38556072e-01
6.02603197e-01 -4.91880149e-01 -1.64118130e-02 -1.03226972e+00
8.48031342e-02 4.50989962e-01 4.60450500e-01 5.84584475e-01
7.61139691e-01 -7.79293418e-01 -1.05894899e+00 3.81098300e-01
1.50609207e+00 7.06349790e-01 4.18557189e-02 1.23377882e-01
6.98565662e-01 1.44472465e-01 9.59127247e-01 9.23373938e-01
5.39506197e-01 6.23224914e-01 1.13253963e+00 2.58694351e-01
5.67697644e-01 -2.20214829e-01 3.45412880e-01 8.19014549e-01
5.02268612e-01 -8.36909533e-01 -6.32066071e-01 7.86482751e-01
-2.11352396e+00 -4.53819424e-01 5.45774736e-02 2.47209430e+00
2.62420535e-01 2.38164783e-01 -1.54377744e-01 2.62910515e-01
1.04783356e+00 3.15817565e-01 -9.14567113e-01 4.81797755e-01
-5.71704730e-02 -4.92954314e-01 1.29188037e+00 6.53820336e-01
-8.50338042e-01 1.94822565e-01 4.95368147e+00 2.56308705e-01
-1.28556621e+00 7.88392350e-02 -4.03242409e-02 -2.11124614e-01
-3.31550688e-01 7.80413598e-02 -8.75191629e-01 8.03746879e-01
7.04871953e-01 -1.84379250e-01 5.45301616e-01 5.86071551e-01
5.05672872e-01 -1.91074863e-01 -8.80363226e-01 1.14363480e+00
6.89209029e-02 -1.21134198e+00 -2.16455609e-01 2.01712668e-01
9.78379905e-01 2.89080232e-01 -9.39270183e-02 -4.90918756e-01
1.38491005e-01 -3.32546920e-01 5.39384067e-01 4.87580836e-01
4.37468261e-01 -7.42787778e-01 6.51868999e-01 8.30020070e-01
-1.47305179e+00 -4.58474874e-01 -4.87516969e-01 2.26388779e-02
2.97170311e-01 9.95383143e-01 -8.30023170e-01 6.12759590e-01
4.35379922e-01 9.51163113e-01 6.73870668e-02 1.17742324e+00
-1.32208034e-01 4.53148246e-01 -7.30122268e-01 -3.11660200e-01
1.52832970e-01 -3.15751642e-01 1.17724228e+00 4.11731064e-01
9.45767581e-01 3.25679988e-01 8.61442864e-01 1.73846185e-02
-5.95564283e-02 -2.87759393e-01 -8.88808489e-01 3.02019209e-01
1.43177783e+00 9.20739293e-01 -5.98079622e-01 -9.91312265e-02
-3.61315310e-01 5.66997826e-01 -2.84879029e-01 5.53812146e-01
-4.72093225e-01 -3.68880272e-01 4.90070164e-01 2.39028081e-01
4.44503307e-01 -4.52947348e-01 1.55746371e-01 -9.99450326e-01
2.10228458e-01 -7.65117943e-01 1.72483727e-01 -2.15839505e-01
-7.88831949e-01 2.27005839e-01 -1.67271957e-01 -1.63913214e+00
-3.35891955e-02 7.33522996e-02 -3.53378952e-01 1.86671272e-01
-1.17856538e+00 -4.55119222e-01 -5.54908693e-01 8.22750032e-01
1.97360262e-01 -5.26474193e-02 4.24255133e-01 4.71423179e-01
-4.65561569e-01 2.99402297e-01 5.60781658e-01 1.31623834e-01
1.44709721e-01 -6.40161574e-01 -1.62692443e-01 1.02444124e+00
1.13498829e-01 4.25250351e-01 6.65551901e-01 -8.42330813e-01
-1.96243453e+00 -1.28153491e+00 2.98735708e-01 3.02723378e-01
5.84092975e-01 -2.38364384e-01 -4.37696934e-01 7.66578496e-01
-7.37822801e-02 1.83017239e-01 2.41628975e-01 -5.68082392e-01
3.96814704e-01 -6.57172620e-01 -1.24164331e+00 3.78468126e-01
9.68907058e-01 1.97048411e-01 9.32572968e-03 4.65267122e-01
5.14375448e-01 -6.66899741e-01 -7.38623738e-01 7.82631397e-01
1.16681494e-01 -3.96295339e-01 7.25175381e-01 1.74995586e-01
-6.48314595e-01 -7.00506032e-01 -4.67253983e-01 -1.66355050e+00
-1.14563465e-01 -8.96839142e-01 -2.21830279e-01 1.00891817e+00
3.49171907e-01 -1.08234584e+00 1.01129532e+00 2.09885642e-01
7.26675838e-02 -4.38152790e-01 -1.30718601e+00 -5.86603463e-01
-7.40649819e-01 -5.05602099e-02 7.81042695e-01 7.16504514e-01
-1.73442632e-01 3.21788758e-01 -5.85457146e-01 1.01135623e+00
8.24865937e-01 2.23438200e-02 9.38711882e-01 -1.36539650e+00
-3.29020500e-01 2.60298014e-01 -5.09431839e-01 -1.70770073e+00
-7.75020048e-02 -4.34700340e-01 2.80594528e-01 -1.58346987e+00
-1.59849644e-01 -7.90993690e-01 -2.89692104e-01 2.39576712e-01
9.30046216e-02 -2.78739393e-01 -5.28037585e-02 4.44775552e-01
-8.52991402e-01 3.59829158e-01 1.20808172e+00 -4.27972585e-01
-3.74160528e-01 4.76365626e-01 -5.48109591e-01 5.88294685e-01
8.27995181e-01 -6.44464850e-01 -8.46096456e-01 -7.35712230e-01
3.08521867e-01 8.09630811e-01 3.25230174e-02 -1.41129375e+00
5.50302744e-01 -2.65471369e-01 -8.10921192e-02 -6.42900825e-01
5.09798050e-01 -1.71255338e+00 6.48854315e-01 7.93528736e-01
-4.41369507e-03 2.19364494e-01 -2.01738656e-01 1.36408401e+00
-1.22202635e-01 -1.18278019e-01 5.27246714e-01 1.16417415e-01
-4.52153683e-01 6.66213810e-01 -3.30377936e-01 -3.47318828e-01
1.35220790e+00 -2.30436176e-01 9.26685613e-03 -8.79565954e-01
-6.69393957e-01 7.61102319e-01 5.19325495e-01 1.84289724e-01
7.52838373e-01 -1.39283514e+00 -3.61352533e-01 1.39545929e-02
7.31390938e-02 4.59276795e-01 6.87219873e-02 9.77075636e-01
-7.51975179e-02 2.67025262e-01 3.28228712e-01 -6.83355927e-01
-1.12720573e+00 5.22388577e-01 2.82095224e-01 2.89419405e-02
-2.92115808e-01 4.41260040e-01 -3.50056320e-01 -1.20448761e-01
5.02626777e-01 -2.84445256e-01 -2.00131729e-01 -4.33726087e-02
2.50211477e-01 2.85557926e-01 -1.84571937e-01 -5.97203791e-01
-3.45950305e-01 8.31766963e-01 4.32143778e-01 -9.65666324e-02
1.31906438e+00 -7.23050237e-01 -1.45535678e-01 4.98619735e-01
1.04226041e+00 4.21237648e-01 -1.20216334e+00 -5.62565744e-01
-5.67639656e-02 -5.17644227e-01 -1.48855776e-01 -1.32829368e-01
-1.57602096e+00 1.19857289e-01 5.68995416e-01 5.94598532e-01
1.30374289e+00 -2.74147660e-01 9.09333587e-01 7.17362285e-01
1.14037704e+00 -7.84779549e-01 -9.95102674e-02 1.74167573e-01
2.05226868e-01 -1.05637360e+00 -6.51843101e-02 -4.58596021e-01
-1.27632976e-01 9.17747021e-01 2.87653983e-01 -3.19872111e-01
9.66018617e-01 1.32320061e-01 1.46382768e-03 -1.28614351e-01
-5.22580743e-01 5.04243560e-02 -2.99884200e-01 6.64619267e-01
-4.23548698e-01 3.17438364e-01 -8.80732611e-02 1.23934969e-01
3.83869857e-02 -4.81434107e-01 7.64606953e-01 1.02321470e+00
-7.49565184e-01 -9.61446524e-01 -8.05066049e-01 4.35764134e-01
-1.58956245e-01 4.58014607e-01 5.04557602e-03 6.12766385e-01
2.66638011e-01 1.49173832e+00 -1.51743025e-01 -5.65049529e-01
5.15329003e-01 -8.47381473e-01 1.26178890e-01 -6.49924278e-01
3.30564141e-01 2.57712543e-01 -1.03179365e-01 -2.89331168e-01
-6.67179167e-01 -8.43603849e-01 -1.31328750e+00 -1.54375792e-01
-2.99296975e-01 7.82486141e-01 6.19018912e-01 7.95920134e-01
4.56410021e-01 5.57940066e-01 1.15140748e+00 -7.51793027e-01
-5.58287024e-01 -7.15901554e-01 -7.83486903e-01 -8.40118751e-02
7.90513396e-01 -8.01676393e-01 -4.04874861e-01 -2.65689552e-01] | [6.028942108154297, 1.4445114135742188] |
541ee5de-907c-418f-b9fc-7156e6231383 | deep-mining-external-imperfect-data-for-chest | 2006.03796 | null | https://arxiv.org/abs/2006.03796v1 | https://arxiv.org/pdf/2006.03796v1.pdf | Deep Mining External Imperfect Data for Chest X-ray Disease Screening | Deep learning approaches have demonstrated remarkable progress in automatic Chest X-ray analysis. The data-driven feature of deep models requires training data to cover a large distribution. Therefore, it is substantial to integrate knowledge from multiple datasets, especially for medical images. However, learning a disease classification model with extra Chest X-ray (CXR) data is yet challenging. Recent researches have demonstrated that performance bottleneck exists in joint training on different CXR datasets, and few made efforts to address the obstacle. In this paper, we argue that incorporating an external CXR dataset leads to imperfect training data, which raises the challenges. Specifically, the imperfect data is in two folds: domain discrepancy, as the image appearances vary across datasets; and label discrepancy, as different datasets are partially labeled. To this end, we formulate the multi-label thoracic disease classification problem as weighted independent binary tasks according to the categories. For common categories shared across domains, we adopt task-specific adversarial training to alleviate the feature differences. For categories existing in a single dataset, we present uncertainty-aware temporal ensembling of model predictions to mine the information from the missing labels further. In this way, our framework simultaneously models and tackles the domain and label discrepancies, enabling superior knowledge mining ability. We conduct extensive experiments on three datasets with more than 360,000 Chest X-ray images. Our method outperforms other competing models and sets state-of-the-art performance on the official NIH test set with 0.8349 AUC, demonstrating its effectiveness of utilizing the external dataset to improve the internal classification. | ['Pheng-Ann Heng', 'Xi Wang', 'Quande Liu', 'Luyang Luo', 'Lequan Yu', 'Hao Chen', 'Jiaqi Xu'] | 2020-06-06 | null | null | null | null | ['thoracic-disease-classification'] | ['computer-vision'] | [ 2.47477844e-01 -6.39561787e-02 -4.47751433e-01 -6.67380989e-01
-1.36334503e+00 -5.52414060e-01 8.25853944e-02 -7.94453099e-02
-1.84537694e-01 9.43067372e-01 8.78580380e-03 -4.58520889e-01
-2.86037296e-01 -5.85715652e-01 -6.64752364e-01 -7.26643205e-01
3.80716324e-01 7.50170112e-01 -7.24591911e-02 2.09227830e-01
-4.08266485e-01 1.00844644e-01 -9.55381334e-01 5.46261191e-01
9.16549027e-01 1.23706460e+00 8.28214958e-02 4.78639454e-01
4.09675650e-02 1.14455307e+00 -6.10457838e-01 -5.04882038e-01
2.59955913e-01 -3.74690950e-01 -8.97217572e-01 2.13749141e-01
1.19792923e-01 -5.08066356e-01 -4.73603606e-01 9.25759554e-01
7.75044978e-01 -2.62023538e-01 8.38887334e-01 -1.11490989e+00
-7.86537468e-01 3.68366182e-01 -7.61548698e-01 2.17971176e-01
-1.99712798e-01 1.73431173e-01 7.86007047e-01 -6.91935122e-01
4.80784446e-01 5.81062376e-01 9.23764348e-01 8.05611551e-01
-9.76327837e-01 -8.07641029e-01 1.42219886e-01 9.20921266e-02
-1.31781387e+00 1.79476976e-01 7.95990467e-01 -6.07577562e-01
4.03093636e-01 3.23229641e-01 3.05227697e-01 1.51073933e+00
1.66751489e-01 8.12422276e-01 1.29283214e+00 -1.27352476e-01
4.39967811e-02 1.36227563e-01 1.76204935e-01 6.94809735e-01
1.21008798e-01 1.09736234e-01 -3.62509117e-02 -4.06547666e-01
4.31148797e-01 5.31786978e-01 -3.26833874e-01 -2.82931238e-01
-1.10232782e+00 7.28672206e-01 4.42043632e-01 2.80600071e-01
-3.57631594e-01 -5.85386232e-02 5.17991304e-01 2.50124067e-01
5.03956437e-01 3.29121679e-01 -8.25745404e-01 2.62620002e-01
-9.45998788e-01 8.50451663e-02 4.91819054e-01 7.64408588e-01
2.42900491e-01 -2.68820494e-01 -1.85976043e-01 1.02255630e+00
1.47537246e-01 4.42547232e-01 7.98960686e-01 -6.89301074e-01
6.58128917e-01 6.61403120e-01 -2.97232628e-01 -6.72399998e-01
-6.28856480e-01 -6.38045907e-01 -1.36523414e+00 -2.52231155e-02
4.60239291e-01 -3.45788538e-01 -1.38978219e+00 1.88024867e+00
4.35003757e-01 1.94424003e-01 5.21898493e-02 8.50129604e-01
9.51999426e-01 2.59459972e-01 3.15678418e-01 -2.38526523e-01
1.46690679e+00 -8.23451340e-01 -7.50797510e-01 -1.02359891e-01
6.69092119e-01 -5.82393765e-01 9.15102839e-01 4.25348371e-01
-8.44327390e-01 -6.47141397e-01 -1.00645864e+00 1.04082875e-01
-1.59699559e-01 4.17370468e-01 3.89708787e-01 4.49971408e-01
-4.87991571e-01 5.27652919e-01 -8.69353294e-01 1.92926273e-01
8.30233455e-01 3.69549453e-01 -2.77028680e-01 -3.90510827e-01
-1.41563535e+00 7.29241848e-01 2.31820971e-01 -2.53226966e-01
-8.71851861e-01 -1.10589004e+00 -5.13216436e-01 -2.60787278e-01
6.38566732e-01 -9.60399032e-01 1.28312051e+00 -8.89124215e-01
-1.00272560e+00 1.05486202e+00 2.91556746e-01 -2.56291151e-01
9.93604660e-01 -9.10933688e-02 -5.81374168e-01 1.29382148e-01
2.46494591e-01 3.90214950e-01 8.39291453e-01 -1.44624746e+00
-5.24828017e-01 -6.19562149e-01 -1.78060398e-01 4.52872999e-02
-2.29139149e-01 -3.02729398e-01 -4.61890996e-01 -1.08011258e+00
1.27592459e-01 -1.03785443e+00 -2.85257876e-01 1.25581250e-01
-5.15794098e-01 -1.73418149e-01 6.42804027e-01 -8.55384171e-01
1.17421627e+00 -2.12056518e+00 -9.55248550e-02 1.59903448e-02
5.51948369e-01 1.79314211e-01 2.18858659e-01 -2.22920731e-01
-2.44039923e-01 2.40830436e-01 -5.30527472e-01 -3.58485430e-01
-3.30880314e-01 5.19587994e-01 -3.07003498e-01 3.32744718e-01
2.72855163e-01 9.23228800e-01 -6.73480034e-01 -8.84068072e-01
-3.37320790e-02 3.45147789e-01 -3.39295030e-01 3.80973786e-01
-2.17471123e-01 7.96082318e-01 -6.26320422e-01 9.60024416e-01
7.96105504e-01 -9.08785939e-01 1.45708084e-01 -3.03922743e-01
5.66817284e-01 3.37136537e-02 -9.95187700e-01 1.92990875e+00
-4.47785288e-01 1.93054415e-02 -2.51272231e-01 -1.10764956e+00
6.33294165e-01 5.77929258e-01 8.98727059e-01 -5.78482568e-01
2.09686980e-01 3.33019704e-01 6.42477125e-02 -6.53481603e-01
-1.53719410e-01 -6.25187516e-01 -1.15673661e-01 6.54016435e-01
6.19939789e-02 -1.96775585e-01 -3.35395247e-01 5.16645871e-02
1.26339281e+00 -1.22550756e-01 2.55786538e-01 1.71983197e-01
2.67798662e-01 1.98663741e-01 7.24612951e-01 8.19636345e-01
-5.33121526e-01 1.14309931e+00 2.20396474e-01 -5.90571880e-01
-8.97273362e-01 -1.20065761e+00 -5.40077209e-01 6.24260426e-01
2.87824512e-01 -2.06665620e-02 -3.45425040e-01 -1.39059734e+00
6.17261082e-02 3.16149265e-01 -9.61803615e-01 -1.73274294e-01
-5.93891919e-01 -1.23946309e+00 7.24331200e-01 8.47266674e-01
4.36685383e-01 -7.71157384e-01 -4.26946253e-01 2.33000740e-01
-4.11768317e-01 -9.06511188e-01 -2.59506166e-01 4.17662710e-01
-8.62259746e-01 -1.40283537e+00 -1.06633508e+00 -5.72421014e-01
5.96909702e-01 -6.35037571e-02 1.30941343e+00 1.37583911e-01
-7.41768003e-01 3.76471102e-01 -2.93443382e-01 -3.58242124e-01
-5.11192918e-01 -1.32451467e-02 -6.45231307e-02 -1.97606385e-01
2.80897945e-01 -3.06984931e-01 -8.85392904e-01 4.49250907e-01
-9.09064591e-01 1.34757876e-01 6.83600605e-01 1.22568297e+00
1.11487687e+00 7.20614940e-02 8.57457638e-01 -1.25247812e+00
4.04895544e-01 -8.56003940e-01 -5.87743893e-03 6.42876863e-01
-7.56978571e-01 -1.22571453e-01 7.82901406e-01 -5.50458312e-01
-1.22666669e+00 9.55193415e-02 -2.97240347e-01 -6.26829207e-01
-2.56117374e-01 3.62346202e-01 9.35312286e-02 1.98215902e-01
6.29537702e-01 1.95954695e-01 -1.23730794e-01 -4.27331537e-01
1.04046583e-01 7.14203119e-01 4.33208734e-01 -7.09653199e-01
5.65466404e-01 5.82274973e-01 1.43360049e-02 8.39380324e-02
-1.38918269e+00 -4.73374248e-01 -5.70059419e-01 -3.38549688e-02
9.77706075e-01 -1.08417165e+00 -4.19170335e-02 4.12368834e-01
-8.82934570e-01 -6.84374943e-02 -5.01514137e-01 5.77914178e-01
-3.94835442e-01 3.38316560e-01 -7.14394152e-01 -2.70850658e-01
-5.26684642e-01 -1.38494682e+00 9.69194293e-01 -2.34972630e-02
-7.41736293e-02 -9.75505531e-01 3.52106333e-01 6.85922086e-01
1.76323339e-01 4.22219813e-01 1.11234736e+00 -9.73640203e-01
-2.52257138e-01 -1.64428070e-01 -4.40949351e-01 6.60679400e-01
4.56199616e-01 -3.26160729e-01 -9.98922884e-01 -2.80282289e-01
4.42889541e-01 -9.19635117e-01 9.13652718e-01 4.27067339e-01
1.90453017e+00 3.41738574e-02 -4.47258204e-01 6.73327506e-01
1.45069838e+00 2.97049105e-01 2.14983866e-01 5.36401905e-02
7.55778730e-01 3.65255117e-01 7.19945252e-01 5.01349688e-01
4.15020376e-01 2.36556932e-01 4.29916441e-01 -4.88702148e-01
-2.20902994e-01 -6.59194365e-02 -3.84376705e-01 7.44428575e-01
2.13433504e-01 -2.66395122e-01 -1.20968080e+00 4.88431215e-01
-1.60714006e+00 -4.51836467e-01 -4.18256670e-02 1.68769062e+00
1.22354424e+00 -7.80941322e-02 -2.25517198e-01 1.20755471e-02
6.55644357e-01 4.52421531e-02 -1.04608035e+00 3.71574819e-01
9.04718135e-03 3.48509252e-01 3.52118641e-01 -6.48909211e-02
-1.38154089e+00 2.40872920e-01 5.92393732e+00 8.97193551e-01
-1.14093935e+00 5.75004041e-01 1.14377439e+00 -2.42382109e-01
-2.19767109e-01 -3.96752924e-01 -5.14127970e-01 7.03053415e-01
3.44352245e-01 1.54361710e-01 -7.74462372e-02 9.56550360e-01
-3.35085720e-01 1.24479964e-01 -1.16990876e+00 1.01056457e+00
1.42461807e-01 -1.28191340e+00 -4.99111004e-02 6.76286817e-02
1.04097438e+00 2.35569835e-01 4.06681269e-01 3.99585009e-01
3.44603986e-01 -1.01988125e+00 3.11069876e-01 3.05701375e-01
1.22881949e+00 -5.09077549e-01 8.78918827e-01 3.95949692e-01
-8.31149757e-01 1.77443270e-02 -1.39991283e-01 4.24398839e-01
6.92863613e-02 6.48352981e-01 -7.63784349e-01 8.99333060e-01
7.55517483e-01 6.18481338e-01 -5.53374171e-01 6.18328094e-01
8.04103091e-02 5.99319339e-01 -1.95528015e-01 5.71839690e-01
4.57977168e-02 1.81720212e-01 -2.49016713e-02 8.80664349e-01
2.39022464e-01 3.98479074e-01 3.68325263e-01 7.97717392e-01
-4.72962290e-01 -5.61763942e-02 -4.30319995e-01 3.39519680e-01
3.52146238e-01 1.04707551e+00 -4.25437331e-01 -5.44664741e-01
-6.67494953e-01 8.38368654e-01 2.56648481e-01 3.06508504e-02
-1.32351732e+00 1.74173743e-01 2.80290067e-01 -2.78604832e-02
-1.25251725e-01 4.36533183e-01 -7.13303030e-01 -1.29767048e+00
1.67489529e-01 -9.70761061e-01 8.05731714e-01 -5.05974054e-01
-1.79610419e+00 6.77450597e-01 -2.58494824e-01 -1.45613801e+00
-5.16204722e-02 -4.66990739e-01 -3.99088830e-01 7.12690055e-01
-1.82215464e+00 -1.30508494e+00 -4.77000982e-01 8.14027846e-01
5.66994250e-01 -1.87093347e-01 7.25447595e-01 6.69950068e-01
-5.07309854e-01 8.43217731e-01 2.64654726e-01 2.41454720e-01
1.07605267e+00 -1.33526862e+00 -9.93213281e-02 5.02031088e-01
-3.39463130e-02 2.53926456e-01 9.30164754e-02 -6.73407257e-01
-8.84261668e-01 -1.41518497e+00 4.62906718e-01 -7.16197550e-01
4.33429062e-01 1.44974634e-01 -1.25047338e+00 6.02623284e-01
-2.67224789e-01 5.18055081e-01 1.26461494e+00 -1.33011788e-01
-4.45957929e-01 4.71111462e-02 -1.39133799e+00 1.45469531e-02
9.58616793e-01 -4.44809884e-01 -7.12866724e-01 6.41502678e-01
7.19156086e-01 -6.65112674e-01 -1.28966641e+00 9.51000094e-01
3.00171524e-01 -6.78353131e-01 1.09370828e+00 -7.75697470e-01
7.74524629e-01 -1.17927514e-01 -2.26392671e-01 -1.17824364e+00
-7.75539950e-02 1.94991633e-01 -1.04121856e-01 1.04199839e+00
3.63618881e-01 -4.14639562e-01 8.07295620e-01 7.00856566e-01
-9.71659794e-02 -1.15201712e+00 -9.94127333e-01 -5.25767565e-01
4.22369927e-01 -4.24800664e-01 5.65376401e-01 1.31436205e+00
-4.88130897e-01 4.35448773e-02 -5.37923694e-01 2.40348771e-01
5.80556214e-01 4.10993695e-01 3.70104581e-01 -1.18591022e+00
-5.64047694e-01 -2.51755640e-02 1.42210558e-01 -7.71645665e-01
1.48141891e-01 -9.72495019e-01 -1.21906400e-01 -1.42425871e+00
6.36221170e-01 -1.09543622e+00 -7.11730838e-01 6.84172332e-01
-5.82605362e-01 2.66023099e-01 4.94250692e-02 5.47685742e-01
-6.65886700e-01 4.73795682e-01 1.62600565e+00 -4.26199675e-01
1.41529948e-01 2.33555198e-01 -7.99930513e-01 8.36533725e-01
8.49941075e-01 -8.09991717e-01 -4.59337473e-01 -6.59768939e-01
2.38260720e-02 4.26743150e-01 4.11683530e-01 -1.01534033e+00
1.22448951e-01 -2.00699251e-02 7.26426303e-01 -7.73607910e-01
2.03617275e-01 -1.02321112e+00 1.50770456e-01 6.35105789e-01
-4.28397536e-01 -1.89178571e-01 9.00215656e-02 8.47051501e-01
-3.73624951e-01 -7.01826513e-02 1.02239645e+00 -3.92861843e-01
-4.32231605e-01 6.63103163e-01 1.78158104e-01 5.53206027e-01
1.22903574e+00 1.72147959e-01 -2.19326973e-01 1.35762230e-01
-9.27776933e-01 3.90302062e-01 2.05518976e-01 3.86055231e-01
5.87980986e-01 -1.19812286e+00 -7.69289076e-01 1.60656482e-01
3.04143667e-01 4.76825625e-01 7.47865021e-01 7.20541894e-01
-2.34092399e-01 1.22649178e-01 -3.74486111e-02 -9.07831371e-01
-1.25459278e+00 7.66184151e-01 4.77543920e-01 -8.62140834e-01
-4.58818614e-01 8.77416134e-01 5.41223705e-01 -6.21571064e-01
2.11249307e-01 -3.09420794e-01 1.09252296e-01 3.19759622e-02
3.90870094e-01 6.68734983e-02 2.31860012e-01 -2.07074597e-01
-3.44831616e-01 6.64237618e-01 -4.48451638e-01 3.73396665e-01
1.26859069e+00 -7.75493309e-02 2.35557526e-01 2.82643527e-01
1.33768559e+00 -3.22245717e-01 -1.21296597e+00 -5.74597836e-01
-3.72188210e-01 -2.35818997e-01 2.07509683e-03 -1.34846652e+00
-1.26936805e+00 1.06483412e+00 9.42876339e-01 -1.35056630e-01
1.21893430e+00 2.67511994e-01 9.35850501e-01 1.56010702e-01
1.58409342e-01 -1.01185071e+00 3.49119872e-01 7.19724968e-02
6.08248055e-01 -1.87325227e+00 7.00659677e-02 -2.84412712e-01
-9.40400600e-01 8.99589002e-01 8.30021918e-01 1.72953844e-01
8.35654616e-01 3.18209410e-01 2.95785218e-01 -2.49197632e-01
-6.84386969e-01 5.92271611e-02 1.66342825e-01 5.12645304e-01
1.07511953e-01 3.81837428e-01 -1.89318687e-01 1.15027034e+00
3.13041776e-01 2.57380694e-01 8.64337012e-02 8.53825271e-01
4.00531888e-02 -1.21395731e+00 -3.01830530e-01 7.63783991e-01
-9.87283289e-01 -2.65194550e-02 -1.21672556e-01 8.50424945e-01
5.22997737e-01 6.52660251e-01 -2.86262572e-01 -4.71283436e-01
2.30152532e-01 5.82560897e-02 3.32021415e-01 -5.22264004e-01
-5.92482269e-01 -1.75613552e-01 -3.95688653e-01 -2.19965890e-01
-3.04427683e-01 -4.46483552e-01 -1.22748268e+00 4.97805811e-02
-3.88450027e-01 -1.43129662e-01 4.07488286e-01 8.95678699e-01
3.05454701e-01 1.00912309e+00 7.75596321e-01 -9.45315287e-02
-1.11250293e+00 -7.02801704e-01 -4.76395100e-01 8.12352180e-01
5.16329885e-01 -7.56285727e-01 -4.23317760e-01 9.56208631e-02] | [14.887750625610352, -2.093658685684204] |
6a648605-c7f9-442a-a192-1fb68bbdf960 | conversations-are-not-flat-modeling-the | 2106.02227 | null | https://arxiv.org/abs/2106.02227v1 | https://arxiv.org/pdf/2106.02227v1.pdf | Conversations Are Not Flat: Modeling the Dynamic Information Flow across Dialogue Utterances | Nowadays, open-domain dialogue models can generate acceptable responses according to the historical context based on the large-scale pre-trained language models. However, they generally concatenate the dialogue history directly as the model input to predict the response, which we named as the flat pattern and ignores the dynamic information flow across dialogue utterances. In this work, we propose the DialoFlow model, in which we introduce a dynamic flow mechanism to model the context flow, and design three training objectives to capture the information dynamics across dialogue utterances by addressing the semantic influence brought about by each utterance in large-scale pre-training. Experiments on the multi-reference Reddit Dataset and DailyDialog Dataset demonstrate that our DialoFlow significantly outperforms the DialoGPT on the dialogue generation task. Besides, we propose the Flow score, an effective automatic metric for evaluating interactive human-bot conversation quality based on the pre-trained DialoFlow, which presents high chatbot-level correlation ($r=0.9$) with human ratings among 11 chatbots. Code and pre-trained models will be public. \footnote{\url{https://github.com/ictnlp/DialoFlow}} | ['Jie zhou', 'Yang Feng', 'Zhengcong Fei', 'Jinchao Zhang', 'Zekang Li'] | 2021-06-04 | null | https://aclanthology.org/2021.acl-long.11 | https://aclanthology.org/2021.acl-long.11.pdf | acl-2021-5 | ['dialogue-evaluation'] | ['natural-language-processing'] | [-2.95338541e-01 3.08463961e-01 1.38930812e-01 -5.81664205e-01
-7.18354642e-01 -7.01319396e-01 7.34823704e-01 -3.49390268e-01
-1.75751507e-01 9.02276814e-01 8.12431693e-01 -2.74223864e-01
2.81512529e-01 -5.66729426e-01 -6.11274503e-02 -2.36970827e-01
4.12890881e-01 6.32461011e-01 2.39835590e-01 -8.07467401e-01
4.03315753e-01 -2.04135492e-01 -8.76908183e-01 7.86279500e-01
1.12848389e+00 5.63062251e-01 4.23378527e-01 9.74268556e-01
-4.89877194e-01 1.25802279e+00 -1.08760953e+00 -6.98712885e-01
-2.80421317e-01 -1.04931545e+00 -1.59586895e+00 -1.26806453e-01
-2.10074678e-01 -5.70512116e-01 -4.74815011e-01 7.18647897e-01
6.57348633e-01 2.96090782e-01 4.34965104e-01 -1.05599904e+00
-8.05970550e-01 1.02664196e+00 1.46553650e-01 1.84692353e-01
8.54628146e-01 6.47736311e-01 1.19137776e+00 -7.82911479e-01
7.51839757e-01 1.56690419e+00 4.93673146e-01 1.01062500e+00
-9.60103691e-01 -3.21037531e-01 1.10389218e-01 2.62742698e-01
-7.20891953e-01 -4.90484357e-01 7.81570971e-01 -5.88951707e-01
9.06278908e-01 4.29938942e-01 2.72880197e-01 1.43110657e+00
-8.63923579e-02 8.69484544e-01 1.04033232e+00 -1.89071640e-01
-1.72884330e-01 3.62639248e-01 4.22042161e-01 4.62348312e-01
-7.50277877e-01 -2.86412388e-01 -6.48199737e-01 -2.59399176e-01
4.94820803e-01 -4.06354606e-01 -3.10473621e-01 5.23187101e-01
-8.74845922e-01 9.50924277e-01 2.54428476e-01 4.21326607e-01
-1.50938094e-01 -3.95470291e-01 6.05579078e-01 7.02781558e-01
6.27520144e-01 6.37599230e-01 -4.36199129e-01 -8.06298673e-01
-1.18742146e-01 3.30962449e-01 1.25242174e+00 9.18274760e-01
7.24877536e-01 -1.85226873e-01 -7.01339126e-01 1.40619564e+00
1.85305312e-01 1.88429713e-01 6.57620728e-01 -1.29622650e+00
6.27954900e-01 8.04881275e-01 3.27644169e-01 -7.95404851e-01
-3.30624282e-01 1.96395412e-01 -6.63293362e-01 -4.43864733e-01
7.25144327e-01 -6.12869084e-01 1.27703816e-01 1.72212887e+00
3.13080102e-01 -3.44778836e-01 1.81276575e-01 8.23283136e-01
1.23878551e+00 8.83121431e-01 -5.35129197e-02 -5.23838103e-01
1.13545179e+00 -1.35045183e+00 -1.08002841e+00 3.84754705e-04
7.96679139e-01 -9.06569779e-01 1.72435462e+00 3.15975815e-01
-1.09171379e+00 -6.91803753e-01 -5.86042225e-01 -1.60856336e-01
4.95604314e-02 -5.92601113e-02 2.92803138e-01 4.42747414e-01
-9.67910588e-01 4.62271273e-01 -2.17237115e-01 -3.88780892e-01
-2.51403481e-01 -8.14022571e-02 1.94032371e-01 2.58705556e-01
-1.63122189e+00 1.08682752e+00 -2.70297378e-02 1.64194688e-01
-8.16684365e-01 -4.27722871e-01 -4.12242264e-01 -1.05524108e-01
3.37356687e-01 -2.01154977e-01 1.84410167e+00 -7.58828878e-01
-2.27904058e+00 6.59495950e-01 -2.86166608e-01 -2.17608005e-01
6.80459440e-01 -3.09615314e-01 -1.15766063e-01 -8.57697427e-02
-3.91010661e-03 3.53267103e-01 2.47561499e-01 -1.09483588e+00
-4.39322025e-01 1.62898190e-02 4.72799927e-01 4.27634388e-01
-5.01998425e-01 3.86645168e-01 -1.76494166e-01 -1.16226226e-01
-5.09014547e-01 -9.19102848e-01 -1.30429178e-01 -8.47568989e-01
-4.67362761e-01 -8.23488235e-01 6.24755025e-01 -8.42397869e-01
1.69510007e+00 -1.81855369e+00 1.18118659e-01 -4.31514055e-01
1.00986391e-01 4.42035735e-01 -1.40735805e-01 9.68482375e-01
5.30578196e-01 9.52374041e-02 -1.34609625e-01 -2.36626089e-01
1.81042597e-01 -1.23798195e-02 -3.36200774e-01 -1.33411899e-01
2.34914757e-02 9.72821772e-01 -1.09794080e+00 -4.96370107e-01
1.33076534e-01 2.90354937e-02 -6.11675680e-01 1.12956333e+00
-5.96642137e-01 8.70861530e-01 -4.98879701e-01 3.68610136e-02
2.73153722e-01 -2.84753203e-01 3.67994428e-01 3.27046007e-01
-1.41673505e-01 8.07409227e-01 -6.25934005e-01 1.79517078e+00
-6.65857792e-01 5.00144362e-01 2.45557681e-01 -3.69746536e-01
1.28087592e+00 6.39917672e-01 2.60688543e-01 -5.62633574e-01
2.73125499e-01 -1.83717683e-02 2.23528862e-01 -7.63202548e-01
7.70954669e-01 2.13980556e-01 -3.93388242e-01 9.57831740e-01
2.40802184e-01 -1.63545206e-01 1.75967112e-01 6.16940916e-01
1.07099748e+00 -2.81792190e-02 9.88520309e-02 -4.85979877e-02
8.88226986e-01 -9.89791900e-02 4.91305441e-01 6.70525134e-01
-5.51779330e-01 2.90102690e-01 9.47807670e-01 -2.22659290e-01
-7.33144760e-01 -5.41417599e-01 1.71163306e-01 1.59890723e+00
1.19807854e-01 -5.96505344e-01 -1.20431674e+00 -8.39450717e-01
-4.86256510e-01 7.82068670e-01 -2.94993401e-01 -8.95740017e-02
-7.04835236e-01 -6.20370209e-01 7.61780262e-01 2.03733426e-02
7.22367942e-01 -1.52701426e+00 -9.27654728e-02 4.05925244e-01
-1.12527394e+00 -1.00910544e+00 -7.54286230e-01 -4.00122046e-01
-4.24012780e-01 -8.88536632e-01 -3.51611972e-01 -4.96644914e-01
8.12131613e-02 5.16855009e-02 1.17076027e+00 3.59605521e-01
1.54393002e-01 1.56877056e-01 -7.66453147e-01 -1.08583625e-02
-1.01235139e+00 2.93599457e-01 -2.20475703e-01 1.18611949e-02
4.61308897e-01 -4.53215897e-01 -5.27940750e-01 7.98517525e-01
-3.14189166e-01 2.50783354e-01 -3.65797081e-03 1.06595480e+00
-3.40695232e-01 -6.63525701e-01 1.18972111e+00 -1.07369483e+00
1.50294268e+00 -6.72720850e-01 -8.50187764e-02 3.01569402e-01
-5.01371026e-01 -1.99198902e-01 7.21882463e-01 -3.53935242e-01
-1.62727654e+00 -3.98619205e-01 -4.95423287e-01 1.93318188e-01
-2.02274188e-01 2.80241758e-01 -2.62342632e-01 5.56393206e-01
8.70016158e-01 1.53119788e-02 -1.42322276e-02 -6.20215714e-01
5.68317890e-01 1.32141531e+00 2.87981659e-01 -7.85205960e-01
2.55158752e-01 -1.63529083e-01 -8.96677732e-01 -6.72291398e-01
-8.98283005e-01 -4.46298689e-01 -7.07263768e-01 -6.96712852e-01
9.00697470e-01 -5.81169367e-01 -1.12281156e+00 5.10129273e-01
-1.82043421e+00 -8.87621343e-01 9.36918557e-02 1.75842658e-01
-5.90957403e-01 3.68590951e-01 -1.24668562e+00 -1.19095051e+00
-5.43104529e-01 -9.75576520e-01 4.16444242e-01 3.21758389e-01
-7.18244612e-01 -1.03647935e+00 3.21290791e-01 7.64354467e-01
4.06180441e-01 -1.98079601e-01 7.14572728e-01 -1.00538051e+00
-2.95915842e-01 1.86165527e-01 -6.44498914e-02 4.56282198e-01
2.03657717e-01 4.88492362e-02 -1.10142195e+00 5.52948341e-02
2.22236574e-01 -8.26365113e-01 2.63268203e-01 -8.15019533e-02
6.98316872e-01 -7.20176578e-01 2.58814603e-01 -2.15063855e-01
5.94238520e-01 3.75470579e-01 5.29715776e-01 -3.37704457e-02
5.33114672e-01 1.15355957e+00 7.97711670e-01 6.88205540e-01
6.87085748e-01 7.16246009e-01 1.69313446e-01 2.99109966e-01
-9.63034166e-04 -5.47570348e-01 6.07378244e-01 1.41874444e+00
2.70482861e-02 -4.44101870e-01 -7.74879456e-01 3.98851067e-01
-2.05645061e+00 -1.09343076e+00 -4.60816354e-01 1.83866000e+00
1.27050340e+00 4.68629897e-02 4.85329688e-01 -3.82713526e-01
8.37740600e-01 2.95830280e-01 -3.69416654e-01 -8.97020340e-01
6.90604970e-02 -1.84978575e-01 -3.79826516e-01 1.13339794e+00
-4.42200959e-01 1.23261571e+00 5.68926048e+00 6.88583910e-01
-7.34368205e-01 5.19374132e-01 7.42447257e-01 1.36495590e-01
-2.64512450e-01 4.77305800e-02 -7.67781079e-01 6.18364811e-01
1.30515039e+00 -5.76389492e-01 5.30689001e-01 6.63866222e-01
7.10537314e-01 3.43704298e-02 -1.01804364e+00 4.57697093e-01
-9.31113586e-02 -9.95758891e-01 -2.14696974e-01 -5.12514971e-02
5.47634006e-01 -2.53355145e-01 -2.65689492e-01 6.78466439e-01
8.97844613e-01 -7.54218757e-01 1.94598973e-01 5.37086964e-01
5.12109697e-01 -3.65162283e-01 5.97286940e-01 7.62734771e-01
-9.02138770e-01 -1.00741282e-01 -2.28599429e-01 -3.48763138e-01
4.54738319e-01 8.77990574e-02 -1.28136599e+00 5.24856389e-01
3.35999936e-01 5.36613882e-01 -3.20980668e-01 3.44063729e-01
-3.99366915e-01 9.16356862e-01 2.43448198e-01 -4.75080132e-01
1.90118790e-01 -3.45147461e-01 5.02511382e-01 1.49017215e+00
-1.30427197e-01 3.81280363e-01 2.36013025e-01 9.73766029e-01
-2.11818203e-01 3.13838691e-01 -4.65398908e-01 9.73132774e-02
7.57246733e-01 1.29307067e+00 -3.02924849e-02 -3.58588964e-01
-2.64692545e-01 9.11900938e-01 4.38474327e-01 1.76905885e-01
-7.65470445e-01 -4.16437566e-01 6.23609602e-01 -1.01324514e-01
-4.06996459e-01 -5.92003167e-02 -3.12689066e-01 -9.65390444e-01
-1.29641518e-01 -1.06416154e+00 3.17150652e-01 -6.12502575e-01
-1.44821465e+00 1.06054366e+00 -1.71236321e-01 -9.89733577e-01
-7.20116913e-01 -2.25546747e-01 -1.01929390e+00 1.09457493e+00
-9.34729934e-01 -7.86809504e-01 -3.66502762e-01 6.14095986e-01
1.15759897e+00 -1.17449410e-01 9.52555895e-01 1.32083103e-01
-6.59614861e-01 5.76063454e-01 -8.95045325e-02 3.39949071e-01
1.08559108e+00 -1.17053628e+00 4.03194964e-01 3.79388690e-01
-2.08084047e-01 5.81122339e-01 7.59341002e-01 -6.33382976e-01
-8.73295069e-01 -7.85338104e-01 1.28727353e+00 -8.20140839e-01
8.50787282e-01 -4.22625184e-01 -1.16689813e+00 4.95038450e-01
8.85908425e-01 -8.35575938e-01 8.07385564e-01 2.65751094e-01
-1.28179848e-01 1.42716721e-01 -1.02598596e+00 5.23163438e-01
1.05413675e+00 -6.78535700e-01 -7.82700062e-01 5.44513106e-01
1.13872480e+00 -3.87920797e-01 -9.92010593e-01 8.66037980e-02
3.67400616e-01 -1.11854112e+00 3.81293535e-01 -7.55757987e-01
7.56440699e-01 2.58910388e-01 8.43332782e-02 -1.46924102e+00
-2.28315294e-01 -1.17265677e+00 2.50432603e-02 1.75795996e+00
5.62686801e-01 -5.47981381e-01 4.42611814e-01 8.77006710e-01
-3.06745321e-01 -5.83370388e-01 -6.54655933e-01 -4.86909807e-01
2.03723446e-01 -1.13141827e-01 6.28156364e-01 9.48971629e-01
8.22645962e-01 9.96891320e-01 -8.09207439e-01 -4.84371006e-01
2.78757587e-02 -9.21184048e-02 1.18465054e+00 -9.79896665e-01
-4.00039166e-01 -3.72017235e-01 4.86023724e-01 -1.67502427e+00
3.86862308e-01 -6.25930846e-01 3.03756773e-01 -1.41621602e+00
4.38629948e-02 -4.16231990e-01 1.81678236e-01 2.75490321e-02
-5.08677363e-01 -2.49915689e-01 3.58693093e-01 3.61235589e-01
-7.98476815e-01 8.54516089e-01 1.51398301e+00 6.03588074e-02
-5.96214175e-01 2.35664576e-01 -7.63847470e-01 5.51923692e-01
1.00517845e+00 -2.17859641e-01 -4.64408755e-01 -4.80709016e-01
-2.09899291e-01 6.33109748e-01 -8.58535618e-02 -4.99714166e-01
2.12091655e-01 -5.36610901e-01 -4.93185908e-01 -2.76697576e-01
5.25666773e-01 -1.21711895e-01 -2.56637871e-01 1.82645619e-01
-1.01030767e+00 9.60378647e-02 -2.83516824e-01 3.43864799e-01
-2.22063407e-01 -3.80632102e-01 6.49059057e-01 -4.12406713e-01
-2.09039971e-01 6.44292012e-02 -7.18373001e-01 4.71327513e-01
5.32219648e-01 2.55785227e-01 -7.08424687e-01 -9.09784019e-01
-6.21716082e-01 5.19656777e-01 5.76557331e-02 6.88442349e-01
3.58318388e-01 -1.22309780e+00 -9.98990834e-01 -3.19390506e-01
-4.26994637e-02 -2.22642779e-01 4.78489369e-01 6.56022310e-01
-2.21732572e-01 5.39168298e-01 -1.18951350e-01 -3.91001612e-01
-1.33089793e+00 7.75842518e-02 3.68668526e-01 -6.23981476e-01
-2.48682752e-01 9.90325809e-01 2.14236379e-01 -8.46002460e-01
2.39791214e-01 1.77212998e-01 -5.01062512e-01 7.14236125e-02
6.25249028e-01 4.49431717e-01 -1.82973489e-01 -6.55627668e-01
1.81638733e-01 -1.90188333e-01 -9.08556357e-02 -4.44363713e-01
9.39058602e-01 -5.80908358e-01 -3.04843366e-01 8.37979436e-01
1.02305818e+00 -1.09513447e-01 -1.15332556e+00 -5.89354157e-01
1.16181776e-01 -5.40992558e-01 -8.21601748e-01 -1.03921497e+00
-4.37383533e-01 9.69897091e-01 -2.37914808e-02 6.83846235e-01
6.34261250e-01 -1.43956905e-02 9.61452305e-01 5.65096796e-01
2.60892570e-01 -1.33412313e+00 7.04041541e-01 1.20904982e+00
1.19548476e+00 -1.10803711e+00 -6.69720888e-01 -2.90832698e-01
-1.36419535e+00 8.84080589e-01 1.34841752e+00 1.82097927e-01
3.22070301e-01 -1.80622175e-01 4.54806358e-01 1.16678901e-01
-1.34745860e+00 6.05850928e-02 -2.27252379e-01 3.79757971e-01
9.13586020e-01 4.39292006e-02 -7.29858756e-01 8.83906722e-01
-5.59770823e-01 -1.92509234e-01 7.42881656e-01 4.73858684e-01
-5.28364718e-01 -1.40006602e+00 -1.24215491e-01 2.17096329e-01
-2.86115706e-01 7.69178793e-02 -1.14468825e+00 2.32933715e-01
-3.12250078e-01 1.68913484e+00 -2.68303186e-01 -8.50654364e-01
5.25708258e-01 4.57883656e-01 -1.05105966e-01 -7.95431137e-01
-1.31842041e+00 -2.26153269e-01 7.51559854e-01 -4.01408106e-01
-1.84718996e-01 -4.09785032e-01 -1.09932017e+00 -6.91871524e-01
-3.53754580e-01 7.30007172e-01 2.42103010e-01 8.61392379e-01
2.74384379e-01 3.43197048e-01 1.28932977e+00 -5.75388610e-01
-7.50835061e-01 -1.79344618e+00 -1.47322565e-01 6.21805489e-01
6.54661879e-02 -2.47915536e-01 -3.88705313e-01 -3.70378257e-03] | [12.800207138061523, 8.103632926940918] |
d533df23-9393-4789-9075-68a80a66b007 | partial-occlusion-handling-for-visual | null | null | http://openaccess.thecvf.com/content_cvpr_2014/html/Zhang_Partial_Occlusion_Handling_2014_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2014/papers/Zhang_Partial_Occlusion_Handling_2014_CVPR_paper.pdf | Partial Occlusion Handling for Visual Tracking via Robust Part Matching | Part-based visual tracking is advantageous due to its robustness against partial occlusion. However, how to effectively exploit the confidence scores of individual parts to construct a robust tracker is still a challenging problem. In this paper, we address this problem by simultaneously matching parts in each of multiple frames, which is realized by a locality-constrained low-rank sparse learning method that establishes multi-frame part correspondences through optimization of partial permutation matrices. The proposed part matching tracker (PMT) has a number of attractive properties. (1) It exploits the spatial-temporal localityconstrained property for robust part matching. (2) It matches local parts from multiple frames jointly by considering their low-rank and sparse structure information, which can effectively handle part appearance variations due to occlusion or noise. (3) The proposed PMT model has the inbuilt mechanism of leveraging multi-mode target templates, so that the dilemma of template updating when encountering occlusion in tracking can be better handled. This contrasts with existing methods that only do part matching between a pair of frames. We evaluate PMT and compare with 10 popular state-of-the-art methods on challenging benchmarks. Experimental results show that PMT consistently outperform these existing trackers. | ['Narendra Ahuja', 'Kui Jia', 'Yi Ma', 'Tianzhu Zhang', 'Changsheng Xu'] | 2014-06-01 | null | null | null | cvpr-2014-6 | ['occlusion-handling'] | ['computer-vision'] | [ 5.08860983e-02 -6.14752769e-01 -3.13993812e-01 3.10296621e-02
-7.59207904e-01 -4.95339245e-01 5.84813893e-01 -2.90434480e-01
-6.75615966e-02 5.52112639e-01 3.05587709e-01 3.94123703e-01
-1.96590602e-01 -1.50266528e-01 -7.89297462e-01 -7.96981931e-01
-6.33141473e-02 2.57604837e-01 7.09816337e-01 4.37978879e-02
1.16427049e-01 7.94903994e-01 -1.66206062e+00 5.60804494e-02
6.11363649e-01 1.08943295e+00 3.39254290e-01 1.11352608e-01
9.04792827e-03 6.52238786e-01 -4.52146560e-01 -2.76284546e-01
7.60742486e-01 -1.71688274e-01 -2.39633277e-01 4.90348846e-01
9.99960005e-01 -1.16527891e-02 -4.66633081e-01 1.17633557e+00
3.94565403e-01 9.35957879e-02 3.42908382e-01 -1.42912281e+00
-2.76306540e-01 1.82047546e-01 -9.03095543e-01 2.08299130e-01
4.56054568e-01 2.21385628e-01 7.52589285e-01 -9.10107493e-01
6.54494882e-01 1.50992870e+00 8.26559365e-01 2.66931951e-01
-1.23169768e+00 -7.94888377e-01 5.47484457e-01 1.52214125e-01
-1.54774296e+00 -5.24145842e-01 7.36728787e-01 -5.69559216e-01
4.24586743e-01 3.11167538e-01 8.50882053e-01 6.71837747e-01
3.71830225e-01 9.14683938e-01 9.97916996e-01 -3.43466736e-02
-7.51433149e-02 -1.44451335e-01 -4.28090394e-02 6.69050753e-01
6.50096178e-01 3.71332437e-01 -6.61001563e-01 -5.01816273e-01
8.35546672e-01 3.56131762e-01 -4.67037499e-01 -9.06894863e-01
-1.62966716e+00 4.41532731e-01 5.34572184e-01 3.26828241e-01
-4.29276556e-01 1.85912564e-01 3.00168693e-01 5.40659577e-02
8.25069621e-02 1.23939440e-02 -2.59986699e-01 1.58704311e-01
-9.74597514e-01 1.12957515e-01 5.50617814e-01 1.17838848e+00
8.29918921e-01 2.35462800e-01 -4.86034155e-01 6.88678026e-01
7.42116630e-01 8.00185621e-01 4.02959973e-01 -8.76268625e-01
2.33193010e-01 5.65561891e-01 2.30360165e-01 -9.77562845e-01
-8.09996426e-02 -5.13926685e-01 -8.77012789e-01 2.83226132e-01
3.89153719e-01 2.08065540e-01 -7.44996846e-01 1.68915844e+00
6.65186167e-01 4.83147234e-01 -2.62935698e-01 9.23544228e-01
8.04973006e-01 2.32842341e-01 9.98437479e-02 -4.58292365e-01
1.46768188e+00 -9.30463850e-01 -8.57593298e-01 -3.08059603e-01
-1.08961686e-01 -1.25877273e+00 1.18228488e-01 2.43291724e-02
-8.41340899e-01 -9.98300374e-01 -8.97810161e-01 3.73152435e-01
-5.67826219e-02 2.71452129e-01 6.44820035e-01 5.22898197e-01
-9.36149836e-01 3.16750467e-01 -8.44480097e-01 -4.02111441e-01
3.77165020e-01 5.33811927e-01 -6.24774218e-01 -1.31934941e-01
-7.12819934e-01 6.74995005e-01 2.97520965e-01 2.47943386e-01
-6.19137108e-01 -6.14533842e-01 -1.05884147e+00 -1.39389813e-01
4.89209622e-01 -8.27901185e-01 9.62578654e-01 -7.39181578e-01
-1.33539069e+00 5.33763289e-01 -5.39939106e-01 -2.00507820e-01
5.15940607e-01 -3.62712413e-01 -4.00065809e-01 -2.13911906e-02
3.38900745e-01 6.47388995e-01 1.27209425e+00 -1.27149081e+00
-7.17433870e-01 -1.66635290e-01 -2.88291126e-01 1.23212688e-01
-4.32094261e-02 1.13239191e-01 -9.40934718e-01 -9.05930281e-01
4.26410854e-01 -1.00340497e+00 -2.34740183e-01 5.87655842e-01
-1.82582170e-01 -2.29062155e-01 1.13739121e+00 -2.58375645e-01
1.01367271e+00 -2.16917896e+00 6.78003058e-02 8.53975862e-02
1.98091045e-01 4.82762158e-01 -1.58022881e-01 4.31803107e-01
5.11820838e-02 -5.86153448e-01 1.00791559e-01 -3.35877150e-01
-7.82274753e-02 1.71548694e-01 -1.24037281e-01 8.97200167e-01
1.20134339e-01 9.41353083e-01 -8.59298944e-01 -9.49799895e-01
4.29335952e-01 5.91245413e-01 -2.16925353e-01 1.97080836e-01
1.12337336e-01 4.72871661e-01 -6.86336637e-01 9.72274005e-01
9.75636542e-01 -2.69296914e-01 2.42769811e-02 -6.99371338e-01
-3.51688623e-01 -2.54217446e-01 -1.66447842e+00 1.67001843e+00
1.82932884e-01 3.88990611e-01 2.30559602e-01 -6.43836498e-01
8.12299132e-01 2.59961724e-01 9.12923038e-01 -3.93973678e-01
6.50557354e-02 1.36335880e-01 -1.04644299e-02 -7.74645284e-02
2.21980900e-01 1.20590113e-01 1.97836861e-01 1.80384845e-01
-3.59510221e-02 2.64496922e-01 3.73235285e-01 8.62006545e-02
1.04098690e+00 4.44954842e-01 4.40194815e-01 -2.33478874e-01
8.80605996e-01 -2.12714091e-01 1.21891737e+00 7.40759194e-01
-5.10486722e-01 4.26605910e-01 -1.64039984e-01 -4.99255776e-01
-4.86515731e-01 -1.03979099e+00 -2.23210350e-01 6.07441783e-01
4.95869547e-01 -3.45806003e-01 -1.88564971e-01 -5.20013928e-01
3.76960456e-01 -3.05996209e-01 -5.64145386e-01 7.29284137e-02
-8.49144101e-01 -3.19300920e-01 -4.52884138e-02 5.39885759e-01
6.04866803e-01 -8.03153038e-01 -6.01803005e-01 3.25968057e-01
-1.82784855e-01 -1.23018837e+00 -1.08144891e+00 -4.54644524e-02
-9.82120335e-01 -1.30605876e+00 -8.39633405e-01 -8.31407726e-01
7.23344922e-01 1.12144792e+00 1.03940856e+00 2.12974444e-01
-2.52348155e-01 6.35562599e-01 -3.29302698e-01 -1.38721034e-01
-6.61407560e-02 -3.89727473e-01 3.70523810e-01 4.48386967e-01
1.89214066e-01 -4.18826759e-01 -5.99118292e-01 8.89280200e-01
-7.48552024e-01 -2.37008095e-01 1.10299218e+00 8.34015369e-01
1.05927920e+00 3.46588157e-02 5.70713170e-03 -3.87640029e-01
-1.20955139e-01 -1.15281872e-01 -9.58433688e-01 4.67139184e-01
-1.27485022e-01 -1.15452148e-02 1.41818568e-01 -7.85540342e-01
-9.15628254e-01 6.80235803e-01 3.29637975e-01 -9.63253856e-01
1.59300700e-01 -1.05255820e-01 -4.29591745e-01 -8.60623777e-01
4.63272855e-02 4.24355626e-01 1.87143430e-01 -5.74940085e-01
1.88311905e-01 7.23801106e-02 6.23072445e-01 -6.07411146e-01
1.57821631e+00 7.03425229e-01 2.84420848e-01 -7.17513263e-01
-6.98146641e-01 -1.08043265e+00 -8.83110225e-01 -4.52875793e-01
5.48805535e-01 -1.25810707e+00 -8.21533799e-01 4.12141949e-01
-9.46488023e-01 3.37747097e-01 -2.28214115e-01 7.22827613e-01
-3.22859943e-01 9.35135007e-01 -3.71441931e-01 -8.12794924e-01
-2.80233294e-01 -1.20394540e+00 1.29502535e+00 3.95192116e-01
1.02795407e-01 -7.75603473e-01 1.19127385e-01 1.28088862e-01
3.30758959e-01 3.26446593e-01 -8.70021135e-02 -3.61494660e-01
-1.04594207e+00 -3.48999470e-01 -2.16052294e-01 6.38726577e-02
3.72789472e-01 -1.37115240e-01 -6.68351054e-01 -8.26366425e-01
7.61809200e-02 3.66974920e-02 6.35049641e-01 5.33575892e-01
5.99745274e-01 -1.55062973e-01 -7.69644320e-01 5.37816405e-01
1.29753423e+00 -2.49428749e-02 4.40407991e-01 3.20684403e-01
7.88888335e-01 3.57479215e-01 1.06218326e+00 4.31366980e-01
1.64212868e-01 1.14649308e+00 4.09970075e-01 -1.55023653e-02
-4.04426306e-01 -2.05304071e-01 4.65924501e-01 6.61405742e-01
1.07158020e-01 2.82605857e-01 -2.62766600e-01 4.06224519e-01
-2.30956292e+00 -1.17656612e+00 -3.07193696e-01 2.36719394e+00
4.67958033e-01 9.81278047e-02 2.79350817e-01 -1.56161606e-01
9.20227230e-01 2.61884660e-01 -3.91670763e-01 6.85578525e-01
-2.43146002e-01 -2.05210373e-01 5.18178105e-01 8.36235434e-02
-1.39074004e+00 6.85467064e-01 6.12324381e+00 9.28965569e-01
-6.64750576e-01 8.07834938e-02 -5.45085557e-02 1.60388649e-01
2.84158051e-01 1.42972365e-01 -1.22040045e+00 6.03509724e-01
7.80127347e-02 3.30593921e-02 1.48354322e-02 7.68488884e-01
6.63241223e-02 -1.42342046e-01 -9.76954341e-01 1.20604455e+00
1.26237839e-01 -1.06946206e+00 -8.07323530e-02 2.40876809e-01
8.22733998e-01 -9.30318534e-02 -6.13307469e-02 1.93929970e-01
5.53033613e-02 -4.16957170e-01 8.50509882e-01 6.75651670e-01
2.59437948e-01 -3.68744135e-01 6.15476310e-01 2.21940860e-01
-2.13622403e+00 4.29505445e-02 -5.06092548e-01 3.07286024e-01
2.10212901e-01 5.84756434e-01 -2.94068992e-01 8.13441515e-01
7.73682714e-01 1.09111059e+00 -6.12416148e-01 1.88980055e+00
-8.74949917e-02 7.59436265e-02 -5.32880068e-01 3.71110976e-01
-8.31454545e-02 -1.53317392e-01 8.13010216e-01 1.03842509e+00
2.81530321e-01 -1.13580056e-01 8.25689495e-01 5.11347592e-01
2.00892493e-01 -4.68876921e-02 -3.89135063e-01 3.69463801e-01
6.03986204e-01 1.40699470e+00 -6.84736729e-01 -2.60616660e-01
-7.22988069e-01 8.02009940e-01 1.04792342e-01 1.66526094e-01
-7.60274589e-01 1.21400818e-01 9.60107386e-01 9.19002574e-03
9.18388426e-01 -2.28315875e-01 2.80893296e-01 -1.31358564e+00
3.02017957e-01 -9.42225575e-01 5.33027530e-01 -4.56060410e-01
-1.48029256e+00 4.48780835e-01 4.01696609e-03 -2.05344558e+00
7.96368569e-02 -4.09570605e-01 -6.15281641e-01 6.56359315e-01
-1.69401741e+00 -1.55771267e+00 -2.69652009e-01 8.95548344e-01
5.42145908e-01 -1.09317899e-01 3.06393594e-01 4.93043870e-01
-5.34871161e-01 5.45595586e-01 -1.14991263e-01 1.88161552e-01
8.44319582e-01 -9.33017492e-01 2.08116189e-01 1.18840301e+00
3.34564000e-01 7.87021816e-01 6.89334989e-01 -9.53686059e-01
-1.93081117e+00 -1.10252929e+00 7.42983818e-01 -4.85469669e-01
6.17038608e-01 -1.59599736e-01 -8.48398566e-01 5.19144893e-01
-1.15327891e-02 6.06252730e-01 5.96156001e-01 -9.62925255e-02
-5.42905450e-01 -2.75875926e-01 -8.97240162e-01 2.55930722e-01
1.14175344e+00 -2.47421697e-01 -6.27863467e-01 4.13664967e-01
2.93434709e-01 -5.93012393e-01 -1.01265526e+00 6.02864385e-01
6.88653171e-01 -9.47978675e-01 1.36150062e+00 -4.44816872e-02
-6.98359132e-01 -1.09732080e+00 -1.96791589e-01 -7.62660563e-01
-8.40556264e-01 -9.64339614e-01 -4.82600093e-01 1.39482665e+00
-3.14483643e-01 -5.53497553e-01 6.62638128e-01 3.48556072e-01
-3.65361460e-02 -4.80269879e-01 -1.13146865e+00 -1.24409580e+00
-6.18672311e-01 -2.47356426e-02 4.83388245e-01 5.60255110e-01
-5.01830518e-01 -7.28887022e-02 -7.32899129e-01 3.98487091e-01
1.17012393e+00 5.07899106e-01 1.01067793e+00 -1.45069456e+00
-5.09136260e-01 -3.47469091e-01 -7.59151220e-01 -1.32811248e+00
9.13389698e-02 -4.24200892e-01 2.66600877e-01 -1.22395480e+00
5.32512844e-01 -4.33967024e-01 -3.94878626e-01 5.30946195e-01
-4.53082681e-01 4.43077892e-01 4.91700381e-01 7.53245175e-01
-1.06149507e+00 6.75206125e-01 1.25564933e+00 -2.38805220e-01
1.03767850e-01 3.62304062e-01 -4.62071717e-01 5.15711367e-01
1.94780573e-01 -5.37340879e-01 1.86740737e-02 -1.14804722e-01
-4.69908476e-01 -2.45010555e-02 3.82270247e-01 -1.32926023e+00
5.17508686e-01 -1.17108775e-02 7.21845508e-01 -9.18286562e-01
4.73590076e-01 -1.19880891e+00 6.59231126e-01 7.66561687e-01
4.49113399e-01 1.59523696e-01 2.85873443e-01 1.01446569e+00
-3.49012852e-01 7.66692962e-03 8.47893715e-01 -7.45581696e-03
-9.55197811e-01 6.58681273e-01 -1.34996951e-01 -1.00097254e-01
1.08674598e+00 -5.32250464e-01 -1.08180596e-02 -5.91355711e-02
-2.54538298e-01 3.48481417e-01 6.65459156e-01 6.64136946e-01
4.81543988e-01 -1.73676109e+00 -5.85648656e-01 2.06031352e-01
1.66520476e-01 -3.66993248e-01 2.64418006e-01 1.14811516e+00
3.52944545e-02 3.95635754e-01 -2.68176585e-01 -1.03059018e+00
-1.59477353e+00 7.82277107e-01 2.03010052e-01 -3.21686715e-01
-8.35276365e-01 6.34301841e-01 4.51212943e-01 -6.86781555e-02
3.90670151e-01 4.28219587e-02 -2.03153223e-01 -3.59883010e-02
6.54737055e-01 2.29187608e-01 -1.67029396e-01 -1.22678220e+00
-6.30407274e-01 1.21372998e+00 -1.10464767e-01 4.72236693e-01
1.03848720e+00 -7.17723444e-02 -1.74109675e-02 -6.93222275e-03
9.16449606e-01 1.55255258e-01 -1.65331614e+00 -6.36403024e-01
-3.65226865e-02 -1.05717731e+00 -2.14742631e-01 -3.59167784e-01
-1.38459575e+00 2.21427485e-01 6.97441280e-01 -1.35288090e-01
1.01899576e+00 -7.59316161e-02 7.52709448e-01 1.73496276e-01
6.75378382e-01 -7.06605375e-01 1.97978407e-01 4.03110564e-01
7.10218251e-01 -1.27073717e+00 4.82604891e-01 -6.38819396e-01
-2.65388966e-01 9.37665701e-01 6.25205219e-01 -5.56178577e-02
6.75013721e-01 1.27982512e-01 5.98729141e-02 -2.68177092e-01
-4.39839542e-01 -5.88002920e-01 7.94375837e-01 7.24632621e-01
2.54380971e-01 -3.14699531e-01 -2.02150971e-01 4.94287051e-02
3.57192367e-01 -3.17965031e-01 -1.96144685e-01 1.22691655e+00
-5.33620000e-01 -1.42584276e+00 -9.38428700e-01 1.44172639e-01
-2.76118368e-01 3.90054107e-01 -3.26205522e-01 8.76869380e-01
1.71987683e-01 8.12819541e-01 -4.24217790e-01 -1.04130402e-01
3.57800782e-01 -3.18393409e-01 5.40662527e-01 -3.84488136e-01
-6.26574934e-01 6.60138011e-01 -3.39192003e-01 -9.50900316e-01
-9.65610623e-01 -1.16785538e+00 -6.37236595e-01 9.30328593e-02
-4.97578353e-01 1.51718453e-01 3.54031861e-01 8.72777700e-01
4.18857306e-01 3.50821704e-01 5.35448253e-01 -1.34907544e+00
-5.48105478e-01 -6.96841180e-01 -5.26307940e-01 4.94027019e-01
3.75001937e-01 -1.23403728e+00 -3.13635319e-01 -1.18492469e-01] | [6.393035888671875, -2.1303329467773438] |
727143d8-e293-49a9-8693-146e0603265e | real-time-monocular-object-slam | 1504.02398 | null | http://arxiv.org/abs/1504.02398v1 | http://arxiv.org/pdf/1504.02398v1.pdf | Real-time Monocular Object SLAM | We present a real-time object-based SLAM system that leverages the largest
object database to date. Our approach comprises two main components: 1) a
monocular SLAM algorithm that exploits object rigidity constraints to improve
the map and find its real scale, and 2) a novel object recognition algorithm
based on bags of binary words, which provides live detections with a database
of 500 3D objects. The two components work together and benefit each other: the
SLAM algorithm accumulates information from the observations of the objects,
anchors object features to especial map landmarks and sets constrains on the
optimization. At the same time, objects partially or fully located within the
map are used as a prior to guide the recognition algorithm, achieving higher
recall. We evaluate our proposal on five real environments showing improvements
on the accuracy of the map and efficiency with respect to other
state-of-the-art techniques. | ['Juan D. Tardós', 'Dorian Gálvez-López', 'Marta Salas', 'J. M. M. Montiel'] | 2015-04-09 | null | null | null | null | ['object-slam'] | ['computer-vision'] | [-6.87715262e-02 -1.21137522e-01 -4.82985042e-02 -4.83403087e-01
-5.48069060e-01 -6.41634285e-01 6.35188222e-01 4.89441216e-01
-7.00398862e-01 5.96914709e-01 -1.54292077e-01 1.43743947e-01
-2.19113991e-01 -7.32697785e-01 -7.91744173e-01 -4.26376730e-01
-2.57798940e-01 1.32787728e+00 8.35223973e-01 -1.13209300e-01
6.70163631e-01 1.04877675e+00 -1.58314037e+00 -3.58519405e-01
3.38185728e-01 1.34204197e+00 3.63034368e-01 4.73559439e-01
5.87946177e-02 4.15524721e-01 -3.95846367e-01 5.14766090e-02
5.74632108e-01 4.52806681e-01 -3.42138499e-01 1.20309100e-01
5.14903903e-01 -4.14382517e-01 -3.98170888e-01 9.03443158e-01
3.38195115e-01 8.73147622e-02 3.10987055e-01 -1.26678050e+00
-2.20656022e-01 1.60552517e-01 -4.88153249e-01 -8.10814872e-02
5.72451711e-01 -1.86411254e-02 7.55167902e-01 -1.19414473e+00
6.90883219e-01 1.07375574e+00 8.19242358e-01 -9.08737406e-02
-1.13087928e+00 -5.11228204e-01 -7.00132549e-02 2.53407359e-01
-1.81365073e+00 -6.25305533e-01 5.26443541e-01 -4.71672088e-01
1.14874506e+00 1.32715598e-01 8.20697486e-01 2.70403355e-01
4.11096700e-02 2.76340425e-01 8.12838078e-01 -4.69014108e-01
5.79175949e-01 1.76930711e-01 4.75588217e-02 8.82786691e-01
6.15842760e-01 3.41052979e-01 -1.03431439e+00 -3.11354905e-01
5.86261034e-01 5.42702340e-02 6.93780780e-02 -1.26647103e+00
-1.23652959e+00 6.16318166e-01 7.47353852e-01 -3.63280028e-02
-5.44456422e-01 2.92690128e-01 -3.58113833e-02 -7.39259273e-02
1.76337853e-01 3.32187533e-01 -1.10168047e-02 -1.60176620e-01
-8.47638011e-01 4.04970407e-01 7.79326022e-01 1.23572218e+00
1.30580306e+00 -4.10843611e-01 4.60134596e-01 3.64231795e-01
5.46080709e-01 9.53425646e-01 -3.53159048e-02 -6.30559683e-01
3.85133952e-01 9.59659696e-01 5.13611138e-01 -1.17947030e+00
-6.72228992e-01 -2.87459195e-01 -6.62746876e-02 4.33254093e-01
1.17134983e-02 5.38854837e-01 -8.99918437e-01 1.19622731e+00
6.87133312e-01 8.76333788e-02 -1.81427166e-01 1.06334198e+00
4.90344733e-01 2.73227930e-01 -5.27056932e-01 2.40826413e-01
1.24674690e+00 -7.58057714e-01 -3.88315648e-01 -6.71820104e-01
3.90175253e-01 -7.32556999e-01 2.89409161e-01 3.08113277e-01
-7.52872109e-01 -3.85771006e-01 -1.49964631e+00 6.22754693e-02
-6.43345416e-01 1.98358700e-01 8.95735085e-01 6.21035695e-01
-1.26160455e+00 2.20928892e-01 -9.89152968e-01 -6.08331561e-01
2.37019077e-01 7.47331381e-01 -6.17808104e-01 -8.60177800e-02
-4.85905528e-01 1.46730673e+00 7.97626197e-01 7.96508864e-02
-7.95727968e-01 -4.05247808e-01 -1.11828363e+00 -2.87556350e-01
4.51926053e-01 -4.34103429e-01 8.95114541e-01 -9.46633890e-02
-1.27910483e+00 1.24579227e+00 -9.85329971e-02 -6.99689031e-01
5.18691540e-01 -4.35164660e-01 1.52350906e-02 -4.97065252e-03
2.89880782e-01 8.69722307e-01 5.31843126e-01 -1.31658363e+00
-1.04751778e+00 -7.24047065e-01 -2.02326164e-01 3.41730535e-01
3.36792141e-01 -3.39930624e-01 -8.42023253e-01 -2.76371334e-02
1.08008993e+00 -1.13268626e+00 -1.19603753e-01 1.73599780e-01
-1.22278305e-02 7.21395910e-02 8.28218162e-01 -4.03526604e-01
6.78732872e-01 -2.12955713e+00 1.93292186e-01 5.45977652e-01
2.02807769e-01 -2.88365304e-01 1.27477318e-01 5.33360839e-01
4.60419029e-01 -5.86754143e-01 1.17992066e-01 -7.10223198e-01
1.38100937e-01 4.33681458e-01 -3.73896062e-01 1.10929191e+00
-6.96852878e-02 8.58734190e-01 -8.82324219e-01 -3.77933174e-01
6.86970055e-01 8.57476741e-02 -3.01153690e-01 1.23977311e-01
2.45881863e-02 1.36829853e-01 -2.57982105e-01 1.01994073e+00
8.77226114e-01 1.93250835e-01 -7.72046437e-03 2.88833100e-02
-5.67352057e-01 4.44290996e-01 -1.63893378e+00 2.05809236e+00
2.30595004e-02 4.18932527e-01 1.28303528e-01 -5.33443153e-01
1.34936047e+00 -3.05765390e-01 6.58139884e-01 -4.96784359e-01
9.75777507e-02 4.77078855e-01 -4.66989666e-01 3.66558991e-02
1.17591929e+00 3.51878881e-01 -1.07334159e-01 1.64282218e-01
1.68640837e-01 -4.87268955e-01 5.25200628e-02 5.36203608e-02
1.20673788e+00 3.70651543e-01 5.70139408e-01 -3.51692230e-01
2.73060918e-01 3.80617589e-01 2.93532491e-01 9.71508443e-01
-1.96496531e-01 2.43720233e-01 -2.81469971e-01 -6.39150858e-01
-1.08798814e+00 -1.17818582e+00 -2.89650798e-01 6.71278954e-01
9.53571796e-01 -2.50883460e-01 -9.08082649e-02 -3.37890655e-01
7.60850072e-01 3.21933717e-01 -6.97463989e-01 5.15038818e-02
-3.09159577e-01 -3.77091199e-01 9.80896689e-03 4.13227975e-01
2.70113498e-01 -7.59792924e-01 -1.26536810e+00 2.13067487e-01
1.83719397e-01 -1.11084223e+00 2.13891361e-02 6.05833530e-01
-6.81298435e-01 -1.03862500e+00 -5.14553115e-02 -4.76274550e-01
7.48865962e-01 5.00953197e-01 8.26090455e-01 -7.35572949e-02
-3.77864003e-01 4.89288330e-01 -4.71688062e-01 -5.86163819e-01
-1.23279497e-01 2.92275418e-02 5.58883011e-01 -9.71930102e-02
4.51544881e-01 -5.46917975e-01 -1.93353131e-01 4.17436838e-01
-2.82813758e-01 1.04061045e-01 7.32517660e-01 3.72733265e-01
7.31035292e-01 -3.19008589e-01 5.20631894e-02 -2.45417327e-01
-2.18636051e-01 -2.52707034e-01 -1.44269168e+00 9.60351061e-03
-7.03611076e-01 8.47055484e-03 -2.61910826e-01 -2.82535523e-01
-4.17654902e-01 7.42344141e-01 3.55598658e-01 -3.68963510e-01
1.16174437e-01 3.29475641e-01 4.99788579e-03 -8.79285455e-01
7.60526896e-01 3.61046225e-01 -2.62681879e-02 -3.84724528e-01
4.30536211e-01 6.61288142e-01 9.08366501e-01 -4.21327889e-01
1.16041780e+00 9.54708695e-01 2.70026565e-01 -6.14613295e-01
-5.90293229e-01 -1.06542683e+00 -1.14691460e+00 -3.80422950e-01
4.21319187e-01 -9.77104843e-01 -8.91410887e-01 3.77941161e-01
-1.13819659e+00 -8.78330022e-02 -3.84278774e-01 6.49443805e-01
-7.26921439e-01 1.54139116e-01 -1.28190309e-01 -1.12077725e+00
-4.65911627e-02 -8.63314271e-01 1.38394654e+00 1.43874407e-01
1.94302201e-01 -3.08186322e-01 2.69535542e-01 1.51509449e-01
2.59358644e-01 2.60910720e-01 -8.99772644e-02 -6.87528849e-01
-1.17532837e+00 -6.88743114e-01 -3.55366588e-01 -3.90474260e-01
-6.88088238e-02 -3.27442914e-01 -7.94603705e-01 -4.83386129e-01
-1.01474561e-01 -1.40966952e-01 6.65982306e-01 -9.21183825e-03
1.27160206e-01 3.00578088e-01 -7.93443263e-01 7.27953851e-01
1.49584866e+00 2.63377935e-01 5.02730906e-01 8.06013525e-01
2.80397058e-01 3.40227693e-01 1.11129510e+00 6.80006683e-01
5.28894007e-01 1.06498909e+00 9.34827507e-01 1.44217700e-01
1.22681357e-01 -3.06582868e-01 7.37262815e-02 4.99463409e-01
2.54326493e-01 2.36867353e-01 -1.09405637e+00 6.70460343e-01
-2.10067201e+00 -4.75890696e-01 9.34303626e-02 2.44784474e+00
2.99591094e-01 2.88645029e-01 -5.50187342e-02 -3.52105312e-02
5.91326296e-01 9.08052400e-02 -5.23593485e-01 2.89464742e-01
-5.49858063e-02 -1.73133940e-01 1.11177683e+00 7.08119750e-01
-1.26416969e+00 1.14157009e+00 6.02385139e+00 1.40287787e-01
-9.77140307e-01 9.32543501e-02 -4.73300368e-01 -2.93853637e-02
4.40538853e-01 4.63366419e-01 -1.25377798e+00 1.25873059e-01
4.31716591e-01 4.66744453e-02 5.15586913e-01 1.25430512e+00
-3.70624602e-01 -6.51483715e-01 -1.16161549e+00 1.29727900e+00
4.28886712e-01 -1.33349764e+00 -3.59504998e-01 3.55925769e-01
3.80019158e-01 5.71966350e-01 -4.46182728e-01 9.57077518e-02
2.43888974e-01 -5.33754945e-01 1.46312094e+00 5.78823209e-01
5.40584564e-01 -5.40915191e-01 8.98904622e-01 4.08330917e-01
-1.41342807e+00 4.04182486e-02 -5.76741755e-01 -3.48931700e-01
2.46179551e-01 4.45553184e-01 -1.59555876e+00 6.43643439e-01
6.27124369e-01 4.92077917e-01 -7.97547758e-01 1.51334155e+00
-1.06119111e-01 -1.38341278e-01 -9.22336817e-01 -1.61011800e-01
1.33412644e-01 -1.91126093e-01 7.94227242e-01 1.06050122e+00
2.25248158e-01 -3.66059393e-02 6.53069675e-01 7.30609536e-01
2.86862999e-01 -8.32903832e-02 -6.33798420e-01 4.31393772e-01
9.60303962e-01 1.18223870e+00 -9.50741172e-01 -2.33273208e-01
-1.07824169e-01 8.07168305e-01 4.32190448e-01 -2.31762707e-01
-6.94926083e-01 -1.40433952e-01 6.18847251e-01 -2.53670965e-03
3.53694439e-01 -7.69441068e-01 -2.60360837e-01 -8.67109954e-01
2.08513454e-01 -2.37850964e-01 2.25998685e-01 -9.53888595e-01
-6.91819310e-01 3.76408964e-01 -1.70933262e-01 -1.18849945e+00
-2.10061129e-02 -5.80680370e-01 2.67791778e-01 7.83532083e-01
-1.41291273e+00 -1.27954054e+00 -8.38764310e-01 2.33395785e-01
1.56302392e-01 -7.96247646e-02 7.36085474e-01 2.15342700e-01
1.33642718e-01 -1.90193448e-02 7.45624751e-02 -5.42238392e-02
6.22675002e-01 -1.04260004e+00 5.08486390e-01 9.14785981e-01
5.55858910e-01 4.29757386e-01 6.25428021e-01 -1.04858005e+00
-1.87426805e+00 -7.27234364e-01 7.03523219e-01 -8.41761231e-01
6.28167868e-01 -7.58437932e-01 -4.18121248e-01 9.46504712e-01
-6.64601803e-01 8.81266221e-02 -4.11166102e-02 1.49002746e-01
-2.00728133e-01 -3.70969445e-01 -1.06286299e+00 1.40455514e-01
1.08356011e+00 -4.01427031e-01 -7.36105740e-01 4.72828656e-01
4.63004380e-01 -1.16064036e+00 -5.50328374e-01 4.80548590e-01
7.52876222e-01 -9.95529115e-01 1.02080429e+00 -6.62193447e-02
-6.28526449e-01 -8.15735698e-01 -7.39991307e-01 -8.82974386e-01
-3.42477620e-01 -3.58922094e-01 -1.55093655e-01 9.34520066e-01
-4.30224463e-02 -8.50116432e-01 9.28205788e-01 2.95816362e-01
-2.56410301e-01 -3.13351452e-01 -1.45738578e+00 -9.65289772e-01
-1.09668481e+00 -4.96357501e-01 6.28451288e-01 6.01825714e-01
-2.92801440e-01 1.82839914e-03 -2.52150804e-01 8.45831871e-01
8.27596426e-01 2.42231354e-01 1.36046290e+00 -1.46301544e+00
4.21479195e-02 -2.29268745e-01 -1.22783518e+00 -1.14399862e+00
-1.49648905e-01 -8.47384393e-01 4.85076576e-01 -1.23648083e+00
8.58768001e-02 -9.02201295e-01 2.57976651e-02 5.97049773e-01
3.88094395e-01 4.20949429e-01 2.60527581e-01 4.75550652e-01
-9.83888507e-01 3.39587539e-01 2.20381841e-01 8.31699446e-02
-2.57361710e-01 -2.69279093e-01 -5.27976602e-02 6.54933631e-01
2.23389432e-01 -7.03018904e-01 2.76805550e-01 -5.01191974e-01
6.74395040e-02 -1.34935677e-01 5.80308557e-01 -1.54507089e+00
7.17997491e-01 -4.02087756e-02 3.53911132e-01 -1.26427245e+00
7.31840372e-01 -1.13261735e+00 4.81756002e-01 6.98141932e-01
1.67460889e-01 3.55420671e-02 1.08054332e-01 5.50113142e-01
1.59850299e-01 -3.03664833e-01 7.05655813e-01 6.40773699e-02
-1.06649363e+00 1.52379632e-01 2.22014830e-01 -6.05382681e-01
1.20304286e+00 -4.13436204e-01 -2.44124994e-01 -9.09231231e-02
-4.31920797e-01 2.91053444e-01 1.07308769e+00 5.45584321e-01
6.01677537e-01 -1.25479484e+00 -6.22918546e-01 4.19234335e-01
5.20730436e-01 1.78336293e-01 -2.31451467e-01 8.93915951e-01
-9.39473867e-01 6.72223091e-01 -2.40120023e-01 -1.26092064e+00
-1.14295554e+00 7.26729095e-01 2.52597332e-01 1.03320293e-01
-5.03820539e-01 6.85999036e-01 -2.22177893e-01 -4.61092621e-01
4.45949733e-01 -9.47393253e-02 3.83065879e-01 -1.96483023e-02
4.74274188e-01 3.14697057e-01 4.29336667e-01 -8.40638041e-01
-1.01107264e+00 7.96618283e-01 2.73260504e-01 -3.37080508e-01
1.40427947e+00 -3.83812755e-01 -3.32778186e-01 5.22755027e-01
5.62051713e-01 2.21498802e-01 -1.27963734e+00 -5.63050151e-01
4.32022274e-01 -8.83860052e-01 -2.46003289e-02 -7.27301598e-01
-3.89936715e-01 2.78154433e-01 8.83864403e-01 -5.21798581e-02
6.59934938e-01 2.60539979e-01 2.92054474e-01 6.81196451e-01
1.40648592e+00 -9.52239811e-01 -9.12296847e-02 6.09327555e-01
8.41684639e-01 -1.22502124e+00 4.83122081e-01 -2.80318677e-01
-1.54672965e-01 9.90377545e-01 2.68814087e-01 -2.85606205e-01
2.68539429e-01 5.58312237e-01 -4.83193435e-02 -3.59613925e-01
-1.78305760e-01 -4.99491453e-01 2.15189904e-01 6.51620269e-01
-6.07566476e-01 -1.66069288e-02 -1.06080147e-02 1.27740085e-01
-3.56201500e-01 -9.59620625e-02 2.28882935e-02 1.23252070e+00
-1.09893274e+00 -6.07167184e-01 -8.82682025e-01 6.86359033e-02
3.39037091e-01 1.72233090e-01 -4.44798410e-01 8.32012177e-01
1.85107827e-01 6.36090755e-01 1.03875279e-01 -4.47317392e-01
4.52541709e-01 -2.01714821e-02 6.12563848e-01 -5.34833848e-01
-3.44688237e-01 -1.60442442e-01 -1.70436911e-02 -8.81520450e-01
-1.43882006e-01 -8.03697109e-01 -1.22769260e+00 -1.75995961e-01
-6.40098095e-01 3.93877514e-02 1.41016650e+00 7.06130326e-01
5.43626070e-01 -1.20689079e-01 3.78417432e-01 -1.55769980e+00
-5.86224139e-01 -8.76259983e-01 -6.47825181e-01 1.34234577e-01
2.63375074e-01 -1.14590883e+00 -4.11860310e-02 -2.73438990e-01] | [7.34589147567749, -2.2577102184295654] |
a3f9a4c7-5629-4d06-a95f-5e53b42f04ad | cinematic-mindscapes-high-quality-video | 2305.11675 | null | https://arxiv.org/abs/2305.11675v1 | https://arxiv.org/pdf/2305.11675v1.pdf | Cinematic Mindscapes: High-quality Video Reconstruction from Brain Activity | Reconstructing human vision from brain activities has been an appealing task that helps to understand our cognitive process. Even though recent research has seen great success in reconstructing static images from non-invasive brain recordings, work on recovering continuous visual experiences in the form of videos is limited. In this work, we propose Mind-Video that learns spatiotemporal information from continuous fMRI data of the cerebral cortex progressively through masked brain modeling, multimodal contrastive learning with spatiotemporal attention, and co-training with an augmented Stable Diffusion model that incorporates network temporal inflation. We show that high-quality videos of arbitrary frame rates can be reconstructed with Mind-Video using adversarial guidance. The recovered videos were evaluated with various semantic and pixel-level metrics. We achieved an average accuracy of 85% in semantic classification tasks and 0.19 in structural similarity index (SSIM), outperforming the previous state-of-the-art by 45%. We also show that our model is biologically plausible and interpretable, reflecting established physiological processes. | ['Juan Helen Zhou', 'Jiaxin Qing', 'Zijiao Chen'] | 2023-05-19 | null | null | null | null | ['video-reconstruction'] | ['computer-vision'] | [ 4.95845139e-01 2.33758718e-01 2.73168355e-01 -1.98422983e-01
-4.75151658e-01 -4.13320750e-01 6.37743831e-01 -4.89010870e-01
-5.43875992e-01 8.45572293e-01 3.81683886e-01 1.67676598e-01
2.25468315e-02 -2.54203141e-01 -1.02430522e+00 -6.79137588e-01
-2.00153291e-01 2.01823562e-02 1.28222197e-01 1.43089056e-01
2.96995252e-01 3.81586105e-01 -1.43563282e+00 4.46004808e-01
7.24351108e-01 1.09735620e+00 4.50362414e-01 7.03863919e-01
3.53686631e-01 1.12973773e+00 -3.53454888e-01 -4.88969475e-01
2.61766195e-01 -7.36151874e-01 -8.05024683e-01 -4.99305613e-02
5.71341276e-01 -5.60962200e-01 -8.66598606e-01 1.08276367e+00
4.91258234e-01 6.89726099e-02 5.38282156e-01 -1.05881858e+00
-9.77512777e-01 2.40926504e-01 -4.16160196e-01 8.15595210e-01
4.01877314e-01 4.71131355e-01 4.44293410e-01 -8.81791770e-01
8.83167922e-01 1.06666481e+00 4.42651868e-01 8.08257461e-01
-1.62247264e+00 -6.91700459e-01 -8.77997745e-03 5.65226197e-01
-1.14213908e+00 -7.86726236e-01 6.47671521e-01 -6.11128330e-01
1.04613531e+00 -4.79820259e-02 1.10341918e+00 1.56410408e+00
5.88610232e-01 5.47687352e-01 1.51633358e+00 -1.11944750e-01
2.33388692e-01 -2.74604112e-01 -2.70986021e-01 7.97438681e-01
6.58831298e-02 2.77576059e-01 -7.91363120e-01 2.05287412e-01
1.16831255e+00 1.66774407e-01 -6.15625978e-01 -1.49852857e-01
-1.61564279e+00 5.91401458e-01 5.06479263e-01 3.48202407e-01
-6.56778336e-01 3.76746088e-01 1.75416604e-01 3.93215060e-01
3.99771243e-01 3.97358209e-01 -3.45967561e-02 -4.02413011e-02
-1.06786740e+00 -1.52624145e-01 4.80579942e-01 5.49656153e-01
1.44800767e-01 5.52465618e-01 -1.45486251e-01 5.40141881e-01
-1.49574252e-02 6.20600998e-01 7.04711437e-01 -1.70468521e+00
-2.91950852e-02 2.97786202e-02 2.10332014e-02 -8.90615463e-01
-3.47527653e-01 -3.14752549e-01 -8.62924874e-01 4.39556807e-01
5.36405683e-01 -4.19704132e-02 -8.62933934e-01 2.01145959e+00
-1.58359677e-01 6.36591256e-01 -9.99574643e-03 1.11657965e+00
6.11076534e-01 2.87742525e-01 1.22327350e-01 -5.41409552e-01
1.11023104e+00 -8.21043789e-01 -7.68929958e-01 -3.22151780e-01
-1.26688212e-01 -1.93826243e-01 8.76811087e-01 6.44707799e-01
-1.53180540e+00 -5.28840482e-01 -1.00717390e+00 -7.31860697e-02
1.56670406e-01 -4.10764009e-01 6.29986763e-01 4.47189867e-01
-1.42068422e+00 5.82314193e-01 -1.14450336e+00 -2.81319201e-01
9.15962577e-01 2.22228348e-01 -8.46630037e-01 -2.10401103e-01
-1.00070381e+00 9.77790296e-01 8.54239315e-02 2.75704134e-02
-1.56497383e+00 -9.66281414e-01 -5.57974517e-01 -7.35789090e-02
1.22554637e-01 -1.15631223e+00 7.66722441e-01 -1.33270478e+00
-1.50431144e+00 9.96401787e-01 -1.56469107e-01 -7.77898729e-01
6.65375412e-01 -8.72805789e-02 -4.06711280e-01 1.01361465e+00
-4.94321026e-02 1.09954214e+00 1.18243313e+00 -1.01220870e+00
1.23530678e-01 -5.31891048e-01 -3.65010612e-02 2.05272630e-01
-1.77863061e-01 -8.18686932e-02 -7.06660897e-02 -7.45185316e-01
8.69082808e-02 -8.78711522e-01 1.40958697e-01 6.43987834e-01
1.52654961e-01 4.45808947e-01 4.90201265e-01 -1.24034703e+00
5.20107150e-01 -1.83213770e+00 6.07759595e-01 -3.91146451e-01
5.82823217e-01 8.55788812e-02 -2.65915602e-01 -1.03070430e-01
-2.84683257e-01 -7.45804682e-02 -5.12336731e-01 -2.50518829e-01
-3.75687957e-01 -5.65087385e-02 -3.39170396e-01 8.46986294e-01
1.63219333e-01 1.36754310e+00 -1.02959871e+00 -2.60464936e-01
2.65493274e-01 9.28096950e-01 -7.04517186e-01 1.45733491e-01
-5.54352533e-03 9.77219999e-01 -2.96857245e-02 5.66552103e-01
3.95328820e-01 -3.00437033e-01 8.83312002e-02 -2.94725537e-01
2.10455179e-01 -2.66105473e-01 -3.94489318e-01 2.42445779e+00
-2.83374816e-01 9.95587170e-01 3.07062328e-01 -1.42701256e+00
2.69555807e-01 4.62511033e-01 6.18833363e-01 -1.03529871e+00
1.37967765e-01 -1.31120160e-01 1.76594570e-01 -8.39450955e-01
-1.77702278e-01 -3.54543626e-01 5.53897560e-01 4.31156158e-01
4.30047661e-01 -6.83862492e-02 -2.56100953e-01 2.73820251e-01
1.41430616e+00 2.65370101e-01 -1.29589930e-01 -1.08834431e-01
2.61326879e-01 -4.53306735e-01 1.08020425e-01 6.13482833e-01
-5.74867725e-01 8.53239596e-01 4.36812937e-01 -1.81448832e-01
-1.07992291e+00 -1.41385829e+00 -1.30116910e-01 4.70270216e-01
-4.69115078e-02 2.02441603e-01 -9.68426347e-01 -2.93532699e-01
-2.84162581e-01 6.60590291e-01 -8.47352207e-01 -4.56829548e-01
-2.73278803e-01 -5.23196220e-01 6.14653349e-01 4.43911463e-01
6.56474054e-01 -1.44097614e+00 -7.98936009e-01 2.56433368e-01
-5.52988708e-01 -1.42472196e+00 -2.73698747e-01 -2.10499048e-01
-1.02009404e+00 -9.71409857e-01 -1.04431760e+00 -6.85391426e-01
6.35360062e-01 3.79866660e-01 1.01952887e+00 -2.15082109e-01
-6.51627481e-01 7.19541728e-01 -2.68151928e-02 5.84537238e-02
-8.51395354e-02 -5.76613963e-01 1.29674003e-01 1.83179975e-01
-6.59382064e-03 -1.09053481e+00 -1.07625520e+00 1.48707405e-01
-1.08886504e+00 2.79776156e-01 5.71567714e-01 9.08863544e-01
5.50577044e-01 -4.59315389e-01 7.71816611e-01 -4.80513632e-01
3.85756403e-01 -5.74138820e-01 -1.56143039e-01 4.43391614e-02
-4.07722652e-01 -2.16273852e-02 4.84593391e-01 -6.77082002e-01
-1.15907323e+00 -9.25137550e-02 4.46862429e-02 -9.16799009e-01
-1.58449784e-01 1.55586153e-01 1.07734941e-01 -3.00326258e-01
6.99758530e-01 5.85473180e-01 3.45514804e-01 1.02308482e-01
4.53985542e-01 1.51145503e-01 9.36062336e-01 -3.72674048e-01
2.96846926e-01 1.09085357e+00 -1.28364712e-01 -7.76137233e-01
-7.23518193e-01 1.63817778e-02 -7.24181354e-01 -5.81864357e-01
1.19393969e+00 -1.05588734e+00 -8.85459125e-01 5.43510437e-01
-1.01593125e+00 -4.60152358e-01 -1.53734714e-01 6.29069209e-01
-9.45757747e-01 4.02752250e-01 -7.34246433e-01 -5.02802372e-01
-3.25008839e-01 -9.99828696e-01 7.29934514e-01 -8.76292586e-02
-8.23783949e-02 -1.00445855e+00 -1.62800595e-01 8.11623931e-01
5.27315199e-01 5.13127744e-01 5.95097244e-01 -1.31292315e-02
-8.09694946e-01 1.79347485e-01 -2.52246022e-01 4.38709170e-01
-2.07528785e-01 -6.12554133e-01 -1.24317920e+00 -3.50238085e-01
4.53171074e-01 -7.07732856e-01 1.08478332e+00 7.94775307e-01
1.40971446e+00 -2.39975616e-01 -4.40546274e-02 8.20292413e-01
1.28858006e+00 6.88542351e-02 1.04383302e+00 -7.30203241e-02
5.45115054e-01 7.32529581e-01 -1.76749885e-01 2.86423378e-02
2.10118547e-01 2.61693090e-01 5.17000914e-01 9.43619311e-02
-5.80973148e-01 -1.93407267e-01 5.87336779e-01 6.92300856e-01
-3.81429046e-01 5.51569723e-02 -6.11995220e-01 6.29774034e-01
-1.59830666e+00 -1.58557880e+00 2.52005160e-01 2.05297089e+00
6.82601392e-01 1.04212873e-01 -1.51095942e-01 -8.24053958e-02
5.25259256e-01 1.12469591e-01 -1.10916471e+00 9.77654904e-02
-3.59041274e-01 2.55908817e-01 1.85698822e-01 4.19465482e-01
-7.14676559e-01 7.51287758e-01 6.78531361e+00 3.35529357e-01
-1.11569965e+00 5.79887092e-01 7.13820636e-01 -6.46874070e-01
-2.54334301e-01 -2.72037864e-01 1.08462237e-01 4.89561886e-01
1.25459802e+00 -2.65579596e-02 1.09451067e+00 1.82584643e-01
3.11491281e-01 -1.17105365e-01 -1.00181818e+00 1.29853201e+00
6.29108310e-01 -1.36566222e+00 -5.08237556e-02 9.26489085e-02
7.90663898e-01 2.63816595e-01 3.27983111e-01 -1.55828577e-02
-3.60560678e-02 -1.31375575e+00 9.53114152e-01 1.13905323e+00
1.16210520e+00 -3.96720052e-01 9.34110805e-02 3.64514947e-01
-5.97602129e-01 -7.63929486e-02 -3.29779863e-01 6.42971471e-02
1.90965176e-01 3.37586850e-01 -3.03365663e-02 1.25262693e-01
7.47992516e-01 1.08541429e+00 -3.60584229e-01 8.48259926e-01
-1.77181903e-02 5.48056006e-01 5.65834232e-02 4.94581401e-01
-1.25068828e-01 -1.61356270e-01 5.93386948e-01 8.06742430e-01
2.46648505e-01 4.27808821e-01 -3.12260270e-01 1.22853625e+00
-2.44241089e-01 -3.94414932e-01 -7.67299294e-01 -3.41676511e-02
-1.72953103e-02 1.15671241e+00 -7.09251165e-01 -3.26002926e-01
-5.50603092e-01 1.52716279e+00 5.35493314e-01 6.64089561e-01
-8.66775990e-01 4.15606648e-02 5.27020454e-01 1.31886646e-01
2.01719195e-01 -4.21223670e-01 -1.16142862e-01 -1.56444681e+00
9.15637091e-02 -5.36598027e-01 -1.14504673e-01 -1.39519143e+00
-1.15267313e+00 6.96933866e-01 -1.39717788e-01 -8.71760130e-01
-1.48899004e-01 -5.19637227e-01 -2.49433160e-01 7.01662362e-01
-1.47503853e+00 -8.67797554e-01 -5.30428648e-01 9.58724678e-01
6.71464682e-01 -1.14860319e-01 8.11060250e-01 1.92460030e-01
-3.20704758e-01 2.62383938e-01 -1.56979412e-01 -1.07668169e-01
6.04615688e-01 -7.81847894e-01 1.24391705e-01 9.61094141e-01
1.55242771e-01 5.30338407e-01 6.25861526e-01 -5.06731749e-01
-1.51052630e+00 -8.80136013e-01 1.83836311e-01 -4.53567117e-01
8.62267852e-01 -3.35702509e-01 -9.86727774e-01 6.96235418e-01
6.08839333e-01 4.46342558e-01 5.02421200e-01 -8.21359634e-01
-4.46316868e-01 2.43007261e-02 -1.28808534e+00 5.61426818e-01
1.45681036e+00 -9.31685507e-01 -7.39701152e-01 3.38166565e-01
6.06089056e-01 -9.54823941e-02 -1.01346576e+00 -4.00695466e-02
8.03697467e-01 -1.01829994e+00 1.26076174e+00 -5.88397384e-01
5.88298917e-01 -4.91484776e-02 -1.07015580e-01 -1.42214906e+00
-2.99444497e-01 -5.18576980e-01 -3.85421455e-01 6.19133890e-01
-5.96430292e-03 -5.49439669e-01 5.43889463e-01 7.74839640e-01
-1.55912265e-01 -2.63699979e-01 -1.01692152e+00 -6.88981414e-01
1.63952172e-01 -3.75935405e-01 -1.82518080e-01 8.99868667e-01
4.01857756e-02 1.45916760e-01 -4.51640904e-01 -1.47456974e-01
9.02933121e-01 -2.37209842e-01 -8.62278789e-03 -1.05013299e+00
-2.74331659e-01 -4.13522393e-01 -5.07449985e-01 -7.87387073e-01
8.36307168e-01 -1.26451182e+00 -3.36543769e-01 -1.36537218e+00
5.55715859e-01 3.46980661e-01 -4.11721259e-01 3.02810162e-01
2.94970065e-01 5.82093418e-01 5.46910688e-02 3.15902233e-01
-4.56175715e-01 7.17444062e-01 1.63859141e+00 -1.86086610e-01
3.05727899e-01 -7.03340292e-01 -6.63358927e-01 5.82149684e-01
6.43372774e-01 -3.46058339e-01 -6.11396968e-01 -5.43888688e-01
-1.35814339e-01 5.54515421e-01 1.00334489e+00 -1.23010111e+00
2.13658363e-01 1.15504369e-01 7.41431534e-01 1.14643387e-01
6.66811049e-01 -7.13610053e-01 3.07310551e-01 6.90759778e-01
-5.85792780e-01 3.37275788e-02 2.27495253e-01 9.79679048e-01
-3.93595472e-02 1.93294197e-01 9.22873557e-01 -4.09351319e-01
-7.26650894e-01 4.31325048e-01 -4.58127350e-01 3.66960853e-01
1.03900075e+00 -3.31730068e-01 -5.06711781e-01 -4.94552910e-01
-1.18426347e+00 -2.44688272e-01 4.16083723e-01 3.61362040e-01
1.07268810e+00 -1.21582997e+00 -7.72571981e-01 3.04863960e-01
-1.83449924e-01 -7.15164125e-01 5.88177264e-01 1.21765172e+00
-5.50473630e-01 3.89533758e-01 -8.31851304e-01 -5.87642908e-01
-7.84934819e-01 7.68962741e-01 5.28068066e-01 3.83489102e-01
-1.05017507e+00 6.43336594e-01 2.98494250e-01 1.98858812e-01
8.14663321e-02 -7.60431588e-02 -1.17950194e-01 -1.74627289e-01
7.24041760e-01 1.36196226e-01 -6.89067617e-02 -6.92156136e-01
-2.62811750e-01 3.94472659e-01 2.07761779e-01 -4.98473972e-01
1.46758223e+00 -4.05476928e-01 -1.19802067e-02 4.55617249e-01
1.13137960e+00 -4.70786750e-01 -1.91310251e+00 -1.12319477e-01
-5.13778090e-01 -4.93435770e-01 3.81489545e-01 -1.01046526e+00
-1.48766649e+00 1.11060882e+00 8.13127279e-01 -1.58393875e-01
1.13673711e+00 -6.93984330e-02 7.48708904e-01 1.27950162e-01
6.09915316e-01 -5.81707180e-01 5.41510224e-01 2.72162408e-02
1.23771513e+00 -1.20947349e+00 -2.36738816e-01 -1.38404174e-03
-8.09257269e-01 7.30089009e-01 3.90733778e-01 -4.14015561e-01
7.23788261e-01 1.05339512e-01 -2.62522012e-01 -1.91583574e-01
-9.83710766e-01 9.01592374e-02 1.86573952e-01 8.59193087e-01
1.73550263e-01 -1.80918783e-01 -5.70484437e-02 3.58071595e-01
4.04444858e-02 2.38558620e-01 6.69528544e-01 5.69432497e-01
-3.41848999e-01 -2.29477838e-01 -1.40143692e-01 4.67733562e-01
-6.58717930e-01 -3.10559899e-01 -9.28191170e-02 4.43743944e-01
-2.02820346e-01 8.54212403e-01 9.33736041e-02 -1.01052456e-01
-5.63455038e-02 2.04505920e-01 1.13871527e+00 -2.29512930e-01
-1.78553000e-01 -6.46127760e-02 -2.05509335e-01 -9.72483337e-01
-8.22017610e-01 -8.81585896e-01 -1.15368140e+00 -2.86589593e-01
2.47028962e-01 -3.41465473e-01 5.62735915e-01 8.41839731e-01
4.83296990e-01 6.85892165e-01 1.09469920e-01 -1.11295640e+00
-2.35447101e-02 -8.63884926e-01 -6.40717030e-01 6.16360605e-01
4.44567233e-01 -8.41311455e-01 -4.44539279e-01 5.84621608e-01] | [10.74774169921875, 2.502939462661743] |
b27df45d-c5de-49cd-912c-ad20c749ba63 | graph-based-label-enhancement-for-multi | 2304.10705 | null | https://arxiv.org/abs/2304.10705v1 | https://arxiv.org/pdf/2304.10705v1.pdf | Graph based Label Enhancement for Multi-instance Multi-label learning | Multi-instance multi-label (MIML) learning is widely applicated in numerous domains, such as the image classification where one image contains multiple instances correlated with multiple logic labels simultaneously. The related labels in existing MIML are all assumed as logical labels with equal significance. However, in practical applications in MIML, significance of each label for multiple instances per bag (such as an image) is significant different. Ignoring labeling significance will greatly lose the semantic information of the object, so that MIML is not applicable in complex scenes with a poor learning performance. To this end, this paper proposed a novel MIML framework based on graph label enhancement, namely GLEMIML, to improve the classification performance of MIML by leveraging label significance. GLEMIML first recognizes the correlations among instances by establishing the graph and then migrates the implicit information mined from the feature space to the label space via nonlinear mapping, thus recovering the label significance. Finally, GLEMIML is trained on the enhanced data through matching and interaction mechanisms. GLEMIML (AvgRank: 1.44) can effectively improve the performance of MIML by mining the label distribution mechanism and show better results than the SOTA method (AvgRank: 2.92) on multiple benchmark datasets. | ['Chi-Man Vong', 'Jiao Li', 'Rui Yan', 'Daixian Liu', 'Jintao Huang', 'Houcheng Su'] | 2023-04-21 | null | null | null | null | ['multi-label-learning'] | ['methodology'] | [ 5.55892706e-01 1.06128938e-01 -5.55293620e-01 -5.97643375e-01
-6.36842012e-01 -1.11053422e-01 1.08386807e-01 5.52831471e-01
-1.98807448e-01 6.83223367e-01 -2.41709843e-01 1.86635792e-01
-4.43443537e-01 -8.51816595e-01 -5.52200854e-01 -8.80788147e-01
1.60402119e-01 3.02830309e-01 2.49817207e-01 2.86708653e-01
3.17461520e-01 2.86220293e-02 -1.49135780e+00 6.29045308e-01
6.08263016e-01 1.11136842e+00 2.59648710e-01 -3.12625915e-02
-4.81539220e-01 1.13515639e+00 -5.05565345e-01 -2.66617894e-01
5.34518100e-02 -4.18332726e-01 -7.15435326e-01 2.81003267e-01
5.06361365e-01 1.44866168e-01 1.27500325e-01 1.26271546e+00
9.68365893e-02 -9.92480591e-02 6.49200439e-01 -1.49197423e+00
-5.90106308e-01 7.62042046e-01 -1.27501047e+00 -9.32467077e-03
-5.37147261e-02 -3.81216526e-01 1.33461142e+00 -9.14786398e-01
4.63105202e-01 1.48112297e+00 4.19153720e-01 2.03944042e-01
-9.05573666e-01 -1.05805111e+00 5.23489416e-01 5.27910411e-01
-1.47495854e+00 1.64475873e-01 8.21902514e-01 -3.47873032e-01
3.08527261e-01 1.43651262e-01 1.67951137e-01 3.57962072e-01
1.78467575e-02 9.82046545e-01 1.50736630e+00 -4.28975552e-01
-1.63805276e-01 3.39090377e-01 4.19791132e-01 9.48132157e-01
2.41835102e-01 -4.12449747e-01 -6.44450426e-01 1.62124895e-02
3.72692347e-01 2.00647041e-01 5.36698326e-02 -1.28133819e-01
-1.19908357e+00 5.85028470e-01 4.49120015e-01 2.59332418e-01
-6.74428418e-02 5.54006621e-02 3.52868289e-01 3.19590181e-01
5.11099935e-01 1.01383880e-01 -6.26518369e-01 5.20934641e-01
-6.25185966e-01 -8.34510550e-02 2.77271807e-01 1.14804411e+00
1.29591835e+00 -6.24006867e-01 -3.12705785e-01 9.61796820e-01
6.66325748e-01 3.56282920e-01 2.66957611e-01 -6.53427124e-01
4.43242937e-01 1.21693730e+00 -5.21119654e-01 -1.37648964e+00
-5.11033654e-01 -7.04294324e-01 -9.12109017e-01 6.15875423e-03
1.55293256e-01 1.34141222e-01 -5.10871768e-01 1.73781657e+00
5.87087214e-01 4.20081943e-01 -3.15744393e-02 7.29243815e-01
1.06877244e+00 7.63867617e-01 3.35492581e-01 -5.39246798e-01
1.49851239e+00 -1.14757442e+00 -7.87333727e-01 -4.42860842e-01
9.64324474e-01 -7.00157523e-01 1.02029300e+00 5.53837121e-01
-3.84084672e-01 -6.12819195e-01 -1.06251347e+00 2.30498880e-01
-1.57313064e-01 2.64579743e-01 9.23711717e-01 3.56878549e-01
-6.47950649e-01 3.08343291e-01 -3.95905413e-03 -1.39780074e-01
7.21422970e-01 4.09504980e-01 -4.09335256e-01 -5.68037927e-01
-1.16007364e+00 5.40553927e-01 9.50547338e-01 -1.82370376e-02
-4.33896571e-01 -4.30186063e-01 -6.32889867e-01 -6.01751916e-02
8.11298132e-01 -9.06703919e-02 7.45411754e-01 -1.08258021e+00
-7.33534575e-01 1.04905152e+00 -9.78288427e-02 1.87582210e-01
1.77861109e-01 1.70505553e-01 -5.81274033e-01 -9.28425267e-02
4.27466929e-01 5.89599848e-01 6.46884143e-01 -1.40042257e+00
-1.08520591e+00 -3.54413599e-01 1.08400270e-01 4.03466254e-01
-5.52296102e-01 -1.00567050e-01 -5.62553823e-01 -3.86084795e-01
3.75411451e-01 -7.06296384e-01 9.81145818e-03 -1.32484689e-01
-2.86709934e-01 -6.67416275e-01 9.24200058e-01 -4.32702094e-01
1.36502743e+00 -2.18099999e+00 -9.11618955e-03 4.06762034e-01
3.15993369e-01 -3.85165885e-02 -2.48717919e-01 3.92293409e-02
-5.32328077e-02 4.22727652e-02 -2.66312420e-01 -9.27446038e-02
-2.69051284e-01 2.73717105e-01 1.50229380e-01 4.20599818e-01
9.48148742e-02 7.86762416e-01 -1.16774690e+00 -1.10129285e+00
1.02052968e-02 -2.69989014e-01 -2.36583412e-01 9.05145258e-02
-1.77547693e-01 4.31027740e-01 -5.58692098e-01 1.00017512e+00
8.25972319e-01 -7.62129962e-01 3.55632812e-01 -6.97373152e-01
2.08564699e-01 -4.07785833e-01 -1.27798200e+00 1.43148828e+00
-4.77371842e-01 1.39874011e-01 -2.73404747e-01 -1.17392135e+00
1.09119928e+00 2.38473881e-02 7.00404108e-01 -7.99479723e-01
-7.15185702e-02 1.36005387e-01 -2.43398324e-01 -6.41443849e-01
-9.46706012e-02 -2.47127280e-01 -6.90384805e-02 4.79672670e-01
-6.14458695e-02 4.41437542e-01 1.98691577e-01 3.25673372e-01
9.50464547e-01 -1.07276194e-01 4.21628118e-01 -3.54443461e-01
9.36717927e-01 -1.34132475e-01 9.35470998e-01 4.35057640e-01
5.35408445e-02 -1.52988387e-02 4.37158436e-01 -1.16195604e-01
-4.89886940e-01 -7.58096814e-01 -2.22014114e-01 1.09306610e+00
8.00992012e-01 -4.35062259e-01 -3.37768674e-01 -1.24267375e+00
1.54323936e-01 4.05556262e-01 -4.24091786e-01 -4.19774592e-01
-3.77277225e-01 -1.05777586e+00 1.67877585e-01 2.54181564e-01
7.74746478e-01 -1.22264254e+00 2.00043753e-01 1.47385210e-01
-2.01150939e-01 -1.05343139e+00 -4.15253043e-01 3.72177102e-02
-7.03550994e-01 -1.30122983e+00 6.76858574e-02 -1.02929580e+00
1.01366150e+00 3.70443523e-01 1.04986799e+00 3.81161779e-01
-1.89629957e-01 2.38515269e-02 -4.61587548e-01 -2.28636235e-01
-2.28144422e-01 -7.13180453e-02 -1.87666908e-01 4.05864805e-01
5.89592874e-01 -3.37037921e-01 -2.74533361e-01 4.33108538e-01
-9.88428950e-01 4.10231292e-01 9.63604748e-01 9.56450582e-01
9.43242610e-01 6.30309165e-01 9.57013607e-01 -1.34822762e+00
2.64597684e-01 -7.06730366e-01 -5.30024469e-01 6.35889292e-01
-1.15794718e+00 -1.73418913e-02 5.36965430e-01 -4.07164216e-01
-1.01328313e+00 -6.74527064e-02 2.45961577e-01 -9.97426212e-02
1.53433755e-02 7.64888406e-01 -5.89866281e-01 -1.66525021e-01
9.15254578e-02 2.39797570e-02 -2.27607027e-01 -3.98159623e-01
5.56622744e-02 7.40637958e-01 1.84576109e-01 -5.12547910e-01
5.76710284e-01 2.40655735e-01 5.70459485e-01 -2.72511035e-01
-1.57752454e+00 -7.12738276e-01 -4.85350311e-01 -4.36008006e-01
6.53399706e-01 -9.20879602e-01 -7.37991333e-01 5.68714261e-01
-8.79553139e-01 5.40354960e-02 1.19850244e-02 5.09482563e-01
-3.32678258e-02 4.30623055e-01 -5.98713577e-01 -4.86183494e-01
-5.27490415e-02 -1.04779243e+00 9.22726631e-01 4.41953123e-01
2.13213846e-01 -1.07798827e+00 -2.93241680e-01 6.88436687e-01
-2.66661555e-01 1.24724023e-01 1.40878320e+00 -5.77603340e-01
-8.34441662e-01 -1.10774234e-01 -7.20907629e-01 4.53395277e-01
3.81193250e-01 -3.21838707e-01 -9.78583157e-01 -3.86911631e-01
-7.40356222e-02 -4.90355819e-01 7.88067758e-01 9.95066464e-02
1.51051807e+00 -2.46439725e-01 -4.40619528e-01 2.50712842e-01
1.70560598e+00 3.25141370e-01 1.83987468e-01 4.13690269e-01
1.07336998e+00 7.89768517e-01 1.31446564e+00 3.73183310e-01
5.03036857e-01 5.31308413e-01 6.41345024e-01 -2.98875958e-01
-1.51533410e-01 -1.74393177e-01 8.78889561e-02 1.05404770e+00
3.66495073e-01 -2.16197744e-01 -5.89997351e-01 2.44489893e-01
-2.14667845e+00 -5.85889518e-01 -4.68317181e-01 1.94098473e+00
8.51131678e-01 2.36160189e-01 -4.53885198e-01 -1.84444897e-02
1.08427370e+00 1.23901002e-01 -8.25444043e-01 7.35941306e-02
-3.36377233e-01 -1.32190943e-01 6.12165332e-01 2.14729324e-01
-1.15303791e+00 8.27171803e-01 4.97801876e+00 1.28924119e+00
-6.57347441e-01 4.46067840e-01 9.40872669e-01 3.08239549e-01
-3.05215925e-01 2.33933553e-01 -9.93217647e-01 5.50262690e-01
3.20725918e-01 -1.42722547e-01 1.41530052e-01 8.48693013e-01
-5.29630855e-02 -2.54155278e-01 -1.00697756e+00 1.27251315e+00
2.75206029e-01 -9.92580116e-01 2.63417155e-01 1.49478465e-01
9.53605592e-01 -4.57025677e-01 5.80512825e-03 3.51731420e-01
-5.66681772e-02 -9.16915834e-01 4.51800197e-01 5.64320028e-01
8.72057676e-01 -7.04019845e-01 9.50795352e-01 5.08912086e-01
-1.45945561e+00 -1.92333758e-01 -4.98406410e-01 6.17233366e-02
-2.47147143e-01 1.11401951e+00 -6.67213023e-01 7.56410599e-01
4.21595067e-01 1.23061705e+00 -8.77377272e-01 7.88450480e-01
-3.15908253e-01 5.02088666e-01 6.11013956e-02 1.78775545e-02
2.24912629e-01 -2.44091123e-01 -4.35762815e-02 1.08212256e+00
5.77016845e-02 1.39042418e-02 8.64252388e-01 4.92124796e-01
-3.71408135e-01 6.54432178e-01 -2.93821156e-01 2.09617466e-01
4.86156017e-01 1.67362511e+00 -7.49244750e-01 -3.64753366e-01
-6.81939662e-01 6.74235046e-01 3.83245170e-01 3.47402343e-03
-9.34288681e-01 -4.30260897e-02 5.20630255e-02 -1.58637121e-01
-2.58038431e-01 3.05955380e-01 -2.75196373e-01 -9.17355359e-01
1.79250583e-01 -6.13792598e-01 7.51277447e-01 -5.93447924e-01
-1.54525816e+00 1.84911624e-01 -4.04035114e-02 -1.24355090e+00
3.56963634e-01 -5.84246397e-01 -3.34427625e-01 7.05553651e-01
-1.73737180e+00 -1.41131544e+00 -6.49941623e-01 4.76399809e-01
5.23307860e-01 -2.67221481e-01 5.46964109e-01 6.31165028e-01
-6.34125531e-01 7.84847200e-01 -1.86815441e-01 -1.54488817e-01
1.01329434e+00 -1.17890024e+00 -3.35237950e-01 4.78590667e-01
5.90912700e-02 2.53799140e-01 1.13647044e-01 -7.86348581e-01
-1.00507510e+00 -1.49603367e+00 7.25005448e-01 2.98746657e-02
4.68346715e-01 -2.24036202e-02 -1.02124667e+00 5.26115239e-01
-1.53018713e-01 1.13379180e-01 8.56677055e-01 2.34305754e-01
-3.53383034e-01 -6.18123353e-01 -1.21347547e+00 3.48798424e-01
1.11511576e+00 -4.13345337e-01 -2.83390731e-01 7.57619023e-01
7.35341251e-01 2.74988785e-02 -1.02029538e+00 7.59082198e-01
2.79135466e-01 -8.26614499e-01 6.40085936e-01 -2.36881316e-01
4.53862250e-01 -6.73963726e-01 -2.05198899e-01 -9.70139325e-01
-6.83228731e-01 3.57921332e-01 -5.16479164e-02 1.59399843e+00
4.09557819e-01 -6.01986289e-01 7.08806336e-01 2.52566725e-01
5.73673658e-02 -1.00593293e+00 -6.32284820e-01 -6.10312283e-01
-5.22986889e-01 -2.54895627e-01 6.73643112e-01 1.21318591e+00
-2.06888244e-01 4.78182167e-01 -4.56262529e-01 3.22383285e-01
9.51852500e-01 4.12708282e-01 3.98122489e-01 -1.42891479e+00
-3.10680509e-01 -1.91754013e-01 -5.17004907e-01 -8.71207297e-01
4.20321345e-01 -1.48266196e+00 -1.18737034e-01 -1.48883724e+00
8.03227603e-01 -1.01670253e+00 -1.03115821e+00 8.09758246e-01
-4.91913080e-01 3.27410311e-01 -9.46647003e-02 3.79982591e-01
-1.09427786e+00 2.56485224e-01 1.50613630e+00 -5.32334447e-01
3.09457809e-01 -2.38465577e-01 -8.38788450e-01 7.88038015e-01
7.02956736e-01 -6.03669405e-01 -6.87296033e-01 -1.48689568e-01
3.42058361e-01 -9.84995663e-02 1.48013741e-01 -9.40461278e-01
3.47659439e-01 -4.76497740e-01 2.76662022e-01 -5.46389461e-01
-8.85745808e-02 -9.57841456e-01 2.00056091e-01 3.35587025e-01
-5.08667052e-01 -3.03555101e-01 -1.60504207e-01 6.92637265e-01
-2.82545328e-01 -4.36179698e-01 7.60152161e-01 -2.33569711e-01
-1.22697222e+00 4.69894290e-01 3.49025667e-01 6.55188859e-02
1.29457676e+00 -5.52454889e-02 -3.70088786e-01 2.59606093e-01
-4.39502716e-01 5.41188180e-01 2.59939671e-01 3.71920317e-01
6.91918790e-01 -1.58127916e+00 -6.48716152e-01 2.40400001e-01
6.19530857e-01 1.85472026e-01 5.01945853e-01 7.34507978e-01
-3.25592346e-02 1.69237796e-02 2.03831103e-02 -7.44678259e-01
-1.57530165e+00 5.80024660e-01 -3.04011279e-03 -5.58606446e-01
-2.95756370e-01 8.48457277e-01 5.75285912e-01 -4.74175125e-01
6.28818944e-02 2.51831800e-01 -5.37231684e-01 2.78981954e-01
4.12260562e-01 1.88871607e-01 -5.18830679e-02 -6.14970684e-01
-4.31358308e-01 9.08612370e-01 -3.60544860e-01 4.92505550e-01
1.00780833e+00 -1.67297110e-01 -7.95191765e-01 6.37868345e-01
1.34921265e+00 2.77585983e-02 -9.05569017e-01 -5.80634654e-01
3.34769905e-01 -5.82340181e-01 1.21658824e-01 -7.79546678e-01
-1.38001895e+00 4.12142783e-01 5.48504591e-01 -1.22775987e-01
1.34491003e+00 2.26077855e-01 6.45556867e-01 3.17856729e-01
5.78125000e-01 -1.10130036e+00 4.36496019e-01 1.87704802e-01
3.14360172e-01 -1.49850249e+00 2.03368276e-01 -7.87401974e-01
-6.00813389e-01 9.82137501e-01 9.80172813e-01 2.77016699e-01
7.11581230e-01 7.95550924e-03 5.34676686e-02 -5.24544895e-01
-6.07718885e-01 1.56466886e-02 1.95833549e-01 2.21086606e-01
2.33495101e-01 1.23935968e-01 -5.82731485e-01 4.63870764e-01
5.43084562e-01 -1.75057635e-01 1.65281698e-01 6.97835624e-01
-5.37969291e-01 -1.32962394e+00 -3.18239421e-01 8.59148145e-01
-1.94931388e-01 -9.44516659e-02 -4.88178767e-02 4.82373953e-01
8.75753701e-01 1.02676702e+00 -2.35293716e-01 -6.11020267e-01
1.84793249e-01 -8.63489881e-02 3.04258883e-01 -8.13410103e-01
-2.71039516e-01 -2.45743878e-02 -1.24708205e-01 -2.98651189e-01
-8.66859019e-01 -5.31719565e-01 -1.63836312e+00 1.21809252e-01
-5.02460718e-01 1.75313115e-01 3.37516189e-01 1.12638891e+00
-1.61990616e-02 7.51047075e-01 9.27405417e-01 4.86586168e-02
-2.33946964e-01 -7.54596174e-01 -1.05006003e+00 6.50315404e-01
-1.65767401e-01 -7.79245853e-01 -2.38113046e-01 -1.54452061e-03] | [9.712873458862305, 4.032660484313965] |
9408ec90-799d-40d4-b53b-f2743a364188 | deep-cascaded-bi-network-for-face | 1607.05046 | null | http://arxiv.org/abs/1607.05046v1 | http://arxiv.org/pdf/1607.05046v1.pdf | Deep Cascaded Bi-Network for Face Hallucination | We present a novel framework for hallucinating faces of unconstrained poses
and with very low resolution (face size as small as 5pxIOD). In contrast to
existing studies that mostly ignore or assume pre-aligned face spatial
configuration (e.g. facial landmarks localization or dense correspondence
field), we alternatingly optimize two complementary tasks, namely face
hallucination and dense correspondence field estimation, in a unified
framework. In addition, we propose a new gated deep bi-network that contains
two functionality-specialized branches to recover different levels of texture
details. Extensive experiments demonstrate that such formulation allows
exceptional hallucination quality on in-the-wild low-res faces with significant
pose and illumination variations. | ['Sifei Liu', 'Chen Change Loy', 'Xiaoou Tang', 'Shizhan Zhu'] | 2016-07-18 | null | null | null | null | ['face-hallucination'] | ['computer-vision'] | [ 2.58835107e-02 2.42492124e-01 3.01342994e-01 -6.11483097e-01
-6.64875090e-01 -1.75149888e-01 4.58537996e-01 -8.97040129e-01
-5.24481479e-03 7.72021770e-01 2.38175765e-01 4.89751399e-01
1.26451358e-01 -7.01742291e-01 -8.05998802e-01 -4.57181007e-01
2.09542945e-01 4.94206250e-01 -2.87910283e-01 -1.32167965e-01
-3.42616811e-02 1.02279282e+00 -1.80119240e+00 2.88733780e-01
4.92087096e-01 1.05672503e+00 9.38260853e-02 1.65168211e-01
1.87853187e-01 4.32236671e-01 -3.77282083e-01 -6.31546557e-01
5.33848882e-01 -3.33003879e-01 -6.31243587e-01 6.24695599e-01
1.11857069e+00 -6.52258456e-01 -6.20632350e-01 1.22759831e+00
8.30479681e-01 -1.24826320e-01 3.36056203e-01 -9.14584994e-01
-8.29159081e-01 -1.47376791e-01 -9.61112618e-01 -1.74866885e-01
7.65702069e-01 1.33123338e-01 3.69580179e-01 -1.45774114e+00
8.34188700e-01 1.67385995e+00 8.32657754e-01 6.38118207e-01
-1.37968206e+00 -6.97579563e-01 5.06330132e-02 1.17301280e-02
-1.88470232e+00 -1.16479981e+00 7.10949540e-01 -1.19269922e-01
6.70008242e-01 4.02139202e-02 2.65003145e-01 1.43687904e+00
9.84322652e-02 2.15251461e-01 1.37651610e+00 -1.19922653e-01
-1.20609395e-01 4.14622091e-02 -6.05435312e-01 9.27686691e-01
2.61728376e-01 1.29770547e-01 -7.48728275e-01 -7.43535385e-02
1.43900108e+00 8.58019367e-02 -4.46745485e-01 -3.01685274e-01
-1.00158453e+00 5.39827883e-01 2.72916049e-01 -4.48698588e-02
-4.28009689e-01 -1.28499627e-01 1.29692033e-02 1.89197734e-01
6.23894930e-01 2.45925218e-01 -1.76000252e-01 4.61381763e-01
-9.58367407e-01 2.69622564e-01 7.30410457e-01 1.27736413e+00
1.04864144e+00 4.99133557e-01 -2.21453071e-01 8.99574518e-01
2.07228839e-01 6.16686761e-01 1.00266501e-01 -1.00704038e+00
1.28084764e-01 8.11362788e-02 2.33212814e-01 -1.20034730e+00
-2.39145830e-01 -4.33059156e-01 -9.85863924e-01 2.42504612e-01
1.14615083e-01 -1.37591228e-01 -9.75873768e-01 1.84955966e+00
3.90031070e-01 5.57747722e-01 -1.95003659e-01 1.08419871e+00
1.16620743e+00 2.48439848e-01 -2.47920930e-01 -3.91518652e-01
1.44612396e+00 -8.08417499e-01 -1.14724362e+00 -1.48963735e-01
-4.75570679e-01 -9.43690002e-01 9.88840938e-01 2.74188876e-01
-1.44698453e+00 -5.56123376e-01 -7.71931648e-01 -3.84975672e-01
-1.28178284e-01 9.79033485e-02 7.14015484e-01 3.92063797e-01
-1.51387489e+00 5.41890442e-01 -4.85740453e-01 -3.68884861e-01
7.16963053e-01 5.00584006e-01 -8.92591596e-01 -1.39611989e-01
-8.14891696e-01 8.88850689e-01 -2.07308367e-01 5.62214613e-01
-1.19609761e+00 -5.39001942e-01 -1.02020097e+00 7.81285539e-02
3.31866652e-01 -8.37621033e-01 9.27439630e-01 -7.11122811e-01
-1.71854925e+00 1.04312360e+00 -4.49572027e-01 1.27636492e-01
5.11768341e-01 -3.40016365e-01 -4.28251714e-01 1.91416934e-01
1.38352320e-01 7.10440934e-01 1.20270836e+00 -1.55988300e+00
1.13760330e-01 -9.12290037e-01 -8.95051137e-02 4.27977592e-01
-3.73443931e-01 3.01123083e-01 -7.58661509e-01 -6.59189284e-01
1.61833197e-01 -5.01467049e-01 -7.63314441e-02 5.33193111e-01
-5.61484933e-01 4.06828284e-01 9.38659668e-01 -7.62771487e-01
6.99062228e-01 -1.95449495e+00 2.47774258e-01 1.41032524e-02
3.13677639e-01 6.26792014e-02 -2.53384918e-01 9.68189351e-03
-2.08473772e-01 -2.14865386e-01 -1.01213247e-01 -9.04048979e-01
-5.59291095e-02 2.20666006e-01 -3.05079907e-01 7.95304179e-01
2.96907455e-01 9.49674428e-01 -5.78887761e-01 -4.70090449e-01
1.29994586e-01 1.05505311e+00 -6.96414649e-01 5.35562932e-01
4.47365865e-02 6.37224793e-01 -1.59109414e-01 1.06842864e+00
1.36764967e+00 -2.35743448e-01 2.52204895e-01 -4.85084414e-01
-3.76431271e-02 -2.23076344e-01 -1.16230261e+00 1.88705361e+00
-4.94789451e-01 3.84805411e-01 6.77751601e-01 -5.36620736e-01
9.15395737e-01 4.95097339e-01 3.59769464e-01 -6.12383008e-01
1.23502016e-01 -2.83928253e-02 -6.36427402e-01 -4.21196193e-01
3.39877516e-01 -3.87506872e-01 4.78799462e-01 5.60597777e-02
4.30534959e-01 -7.67033771e-02 -4.79564667e-01 -3.61389279e-01
5.49153328e-01 3.64380002e-01 2.79654354e-01 -1.93609148e-01
4.18816477e-01 -7.12853432e-01 5.91068983e-01 1.89161405e-01
1.21990973e-02 1.26637018e+00 4.66302633e-01 -4.00619060e-01
-9.06942487e-01 -1.16251540e+00 -5.09344935e-01 7.86628962e-01
9.82697457e-02 -2.49687746e-01 -8.47591937e-01 -4.14878428e-01
-7.33760446e-02 6.01705676e-03 -7.43600368e-01 2.75067389e-01
-5.31327724e-01 -7.14653611e-01 4.47565556e-01 3.64881158e-01
8.18508208e-01 -1.13234031e+00 3.52559462e-02 -2.02613935e-01
6.63285842e-03 -1.39414966e+00 -6.29829884e-01 -3.19516420e-01
-4.45114195e-01 -8.80770802e-01 -7.63548374e-01 -8.60868990e-01
8.76796544e-01 2.36924380e-01 1.23304522e+00 -2.52058923e-01
-4.79227602e-01 3.06845725e-01 6.00168109e-02 -6.71117604e-02
4.19066399e-01 -5.27903616e-01 2.60463774e-01 4.62362051e-01
1.65947616e-01 -9.22246099e-01 -6.68495834e-01 3.72621208e-01
-6.52506948e-01 -1.29392534e-01 5.61160982e-01 9.34886634e-01
7.24855125e-01 -1.64031997e-01 3.68541867e-01 -7.39605308e-01
2.93999761e-01 -4.49548155e-01 -5.42528033e-01 1.50098830e-01
-6.47040755e-02 -2.89061278e-01 5.68740845e-01 -2.26156205e-01
-1.38352907e+00 1.84654564e-01 -3.18655699e-01 -9.07053232e-01
-2.96756566e-01 -1.51049972e-01 -7.82156944e-01 -5.51183760e-01
5.38319528e-01 2.40859658e-01 2.96543300e-01 -6.39797747e-01
1.61588907e-01 2.19902143e-01 7.51782596e-01 -5.61946809e-01
1.12290287e+00 8.58321011e-01 4.31850702e-02 -1.05475831e+00
-6.80057287e-01 1.06199414e-01 -8.57510388e-01 -1.79338902e-02
7.92423964e-01 -1.22212851e+00 -9.19136703e-01 4.20102507e-01
-1.28286588e+00 9.37914252e-02 -1.57012448e-01 2.89430648e-01
-6.46593094e-01 4.47852105e-01 -5.78164279e-01 -6.12748325e-01
-1.37549803e-01 -1.09021020e+00 1.55502665e+00 2.15757385e-01
1.22623302e-01 -7.73613751e-01 -1.95087254e-01 1.99679300e-01
4.79786098e-01 5.37179470e-01 2.98480541e-01 1.35119826e-01
-7.65224099e-01 1.12173229e-01 -5.56824386e-01 1.48209795e-01
1.50939226e-01 -2.91869164e-01 -1.54091871e+00 -4.72636193e-01
1.96602568e-01 -4.79837775e-01 5.17732978e-01 2.33217537e-01
1.44964898e+00 -4.76718724e-01 -1.66690961e-01 1.32864118e+00
1.32276928e+00 -2.24648699e-01 9.34463918e-01 -2.75981396e-01
8.93586695e-01 8.19521487e-01 3.79484534e-01 6.54036701e-01
2.19083756e-01 1.13016701e+00 5.18663645e-01 -4.69873846e-01
-3.70099872e-01 -4.05285448e-01 2.85222352e-01 1.87346935e-01
-3.70372415e-01 5.21446653e-02 -4.21035171e-01 2.04046845e-01
-1.51949143e+00 -1.05853367e+00 4.77959335e-01 2.05743289e+00
7.07249045e-01 -5.88167310e-01 -1.00057811e-01 -4.54041064e-01
7.09621549e-01 3.13941419e-01 -4.71990496e-01 -8.94630700e-02
-4.65641737e-01 5.67885697e-01 1.11061707e-01 6.46980822e-01
-8.12258899e-01 1.14628339e+00 7.16981459e+00 7.29283214e-01
-1.24649203e+00 3.07812274e-01 6.83702111e-01 -2.66034186e-01
-4.10827428e-01 -3.74631375e-01 -8.57474387e-01 2.41648242e-01
3.64177644e-01 2.11486965e-01 6.69808269e-01 7.16460705e-01
3.57789770e-02 3.37501973e-01 -9.82453287e-01 1.43151236e+00
5.11662602e-01 -1.19799376e+00 2.88310379e-01 2.69410461e-01
7.87737548e-01 -3.69207889e-01 5.64020872e-01 -2.14150008e-02
-7.93750584e-02 -1.60938430e+00 5.24867654e-01 7.19894230e-01
1.61335170e+00 -7.61383176e-01 3.72258961e-01 -1.12115934e-01
-1.16874146e+00 1.43819615e-01 -7.15374827e-01 3.94282937e-02
1.51264757e-01 3.77068549e-01 -4.89362866e-01 6.13352597e-01
7.41510212e-01 6.76843762e-01 -4.92442876e-01 5.66104114e-01
-1.10276639e-01 -1.59415558e-01 -1.89674646e-01 6.97197258e-01
-1.59540012e-01 -3.17591310e-01 4.43758756e-01 9.67931092e-01
5.14092147e-01 2.13883325e-01 7.75535107e-02 1.16050589e+00
-1.98275879e-01 1.38293058e-02 -9.65551734e-01 4.30862546e-01
4.43540931e-01 1.40769422e+00 -3.15474689e-01 2.88411826e-02
-6.25881672e-01 1.34419739e+00 5.67153215e-01 6.63375735e-01
-7.79078364e-01 2.51589082e-02 1.03475404e+00 7.21982956e-01
3.56653959e-01 -1.91853002e-01 -1.15177527e-01 -1.51522672e+00
2.32187524e-01 -7.93820918e-01 -1.53399527e-01 -7.47746646e-01
-1.35343528e+00 1.02417922e+00 -7.55201206e-02 -1.08082461e+00
-1.94950581e-01 -6.04869723e-01 -6.48931205e-01 1.07641137e+00
-1.64321208e+00 -1.57939208e+00 -6.49951816e-01 1.13190365e+00
4.18540061e-01 -3.56739223e-01 1.03834617e+00 4.40889359e-01
-7.25867093e-01 8.67082655e-01 -2.93812424e-01 -3.87767740e-02
8.42720270e-01 -9.31445777e-01 3.03303331e-01 6.14940941e-01
-5.87285087e-02 9.36708331e-01 5.11674464e-01 -4.88173932e-01
-1.61903334e+00 -1.27328825e+00 7.42213428e-01 -3.62791985e-01
5.17766438e-02 -6.84189200e-01 -8.24265957e-01 8.39387953e-01
3.44945163e-01 6.85948133e-01 4.70529407e-01 8.34147632e-02
-4.00941908e-01 -1.26764238e-01 -1.59274721e+00 6.96314037e-01
1.53293622e+00 -7.83088982e-01 -2.13167399e-01 4.54158664e-01
2.12574437e-01 -5.35946608e-01 -8.38927865e-01 6.05609715e-01
6.28252566e-01 -1.36306703e+00 1.20196331e+00 -4.22781736e-01
2.25205019e-01 -4.18127298e-01 -3.14485371e-01 -1.03966379e+00
-4.54579651e-01 -9.92222428e-01 -1.44230932e-01 1.08437502e+00
-1.31308407e-01 -6.00266814e-01 8.10434222e-01 7.72565529e-02
-6.49434552e-02 -8.48341882e-01 -8.29600930e-01 -5.34189224e-01
-3.48488092e-01 1.94684684e-01 7.58080602e-01 1.05598998e+00
-4.30186063e-01 3.40808481e-01 -8.78149033e-01 4.22659487e-01
9.42633569e-01 3.19808684e-02 7.98988760e-01 -1.11853147e+00
-5.08076921e-02 -8.24325457e-02 -5.01742840e-01 -9.99597192e-01
5.98370433e-01 -6.22831523e-01 -7.40870908e-02 -1.03669846e+00
3.14387292e-01 -2.32603010e-02 1.07406147e-01 3.82224649e-01
1.19044356e-01 8.54650199e-01 -2.51406342e-01 -3.21804807e-02
-3.37551415e-01 9.45237219e-01 1.45944858e+00 4.37993437e-01
1.83123782e-01 -2.90225923e-01 -8.23320806e-01 8.34728897e-01
3.27372432e-01 5.77600300e-02 -4.91144150e-01 -6.14464283e-01
-2.55369753e-01 3.43464494e-01 6.21052861e-01 -8.94665420e-01
1.18551612e-01 -1.52488321e-01 9.96470690e-01 -3.68561059e-01
8.99356782e-01 -7.96136856e-01 3.74852061e-01 -1.92063779e-01
9.30160880e-02 2.89916135e-02 2.59387076e-01 5.07622540e-01
-4.21291858e-01 3.55227470e-01 1.00369167e+00 -3.61761123e-01
-5.53964436e-01 9.60057378e-01 2.81882435e-01 -2.22313613e-01
1.02963471e+00 -4.84950423e-01 -2.08791256e-01 -5.16798198e-01
-9.89860296e-01 -2.09343255e-01 7.27570951e-01 4.41491783e-01
9.99277532e-01 -1.61998343e+00 -8.91051888e-01 8.22192609e-01
-9.48922336e-02 3.62490974e-02 6.14240706e-01 6.65996194e-01
-4.56505388e-01 2.90239811e-01 -7.59575307e-01 -4.79266852e-01
-1.15863478e+00 4.67868596e-01 5.19532144e-01 2.10287645e-01
-7.40941882e-01 9.95691478e-01 7.75086224e-01 -5.52873135e-01
1.67318642e-01 4.12929326e-01 -1.92843944e-01 -3.49532589e-02
7.25020707e-01 1.44987047e-01 -2.29989318e-03 -1.07710803e+00
-3.48437518e-01 9.10230756e-01 -7.02028722e-02 -3.30824293e-02
1.41678166e+00 -2.20838830e-01 -3.69231731e-01 -8.39817077e-02
1.16087914e+00 1.91010356e-01 -1.63637733e+00 -1.62556052e-01
-5.45062780e-01 -1.15329719e+00 -7.47369155e-02 -3.53152066e-01
-1.24542212e+00 7.72904873e-01 4.88175154e-01 -5.49257338e-01
1.23715973e+00 -1.54300719e-01 4.64920819e-01 1.75302133e-01
6.75340593e-01 -6.53955638e-01 3.24268460e-01 4.83742207e-01
1.54232264e+00 -1.15637743e+00 1.09579392e-01 -8.89897346e-01
-5.70962667e-01 9.48826790e-01 1.11788893e+00 -1.65775657e-01
7.87280679e-01 5.26114464e-01 -1.80655599e-01 -4.26930368e-01
-5.64172626e-01 -2.11300284e-01 3.19740623e-01 9.81582642e-01
4.90713030e-01 -1.75246283e-01 3.16145152e-01 3.43681395e-01
-2.50109285e-01 -2.51845032e-01 2.48194873e-01 5.53190529e-01
-2.17763320e-01 -6.62197471e-01 -5.96064687e-01 7.79142901e-02
-3.16801429e-01 -8.09362531e-02 -3.09189647e-01 7.97826231e-01
1.81921780e-01 6.35197520e-01 2.58105338e-01 -2.73548424e-01
3.79208207e-01 -1.94325462e-01 9.37446713e-01 -9.07656431e-01
-3.00670952e-01 3.05723757e-01 -7.03061223e-02 -9.64763939e-01
-1.38745591e-01 -4.11010474e-01 -6.38700306e-01 -5.80627680e-01
-5.01153842e-02 -3.27333570e-01 1.78058147e-01 5.06841958e-01
4.34983581e-01 2.11680740e-01 7.81669736e-01 -1.38922906e+00
-4.14896876e-01 -1.12790239e+00 -8.86937618e-01 2.83297807e-01
4.32143599e-01 -9.89774823e-01 -1.55447230e-01 -1.10627033e-01] | [12.865338325500488, -0.05446699634194374] |
bb9d9b61-5e97-438e-9580-bfc76df01a80 | 190503246 | 1905.03246 | null | https://arxiv.org/abs/1905.03246v3 | https://arxiv.org/pdf/1905.03246v3.pdf | End-to-End Wireframe Parsing | We present a conceptually simple yet effective algorithm to detect wireframes in a given image. Compared to the previous methods which first predict an intermediate heat map and then extract straight lines with heuristic algorithms, our method is end-to-end trainable and can directly output a vectorized wireframe that contains semantically meaningful and geometrically salient junctions and lines. To better understand the quality of the outputs, we propose a new metric for wireframe evaluation that penalizes overlapped line segments and incorrect line connectivities. We conduct extensive experiments and show that our method significantly outperforms the previous state-of-the-art wireframe and line extraction algorithms. We hope our simple approach can be served as a baseline for future wireframe parsing studies. Code has been made publicly available at https://github.com/zhou13/lcnn. | ['Haozhi Qi', 'Yi Ma', 'Yichao Zhou'] | 2019-05-08 | end-to-end-wireframe-parsing | http://openaccess.thecvf.com/content_ICCV_2019/html/Zhou_End-to-End_Wireframe_Parsing_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Zhou_End-to-End_Wireframe_Parsing_ICCV_2019_paper.pdf | iccv-2019-10 | ['line-segment-detection', 'wireframe-parsing'] | ['computer-vision', 'computer-vision'] | [-4.26959880e-02 1.93644121e-01 -3.23182136e-01 -4.87832874e-01
-1.26073754e+00 -9.66794014e-01 4.19767976e-01 1.88156441e-01
-8.94587114e-02 4.71211553e-01 -4.97775106e-03 -6.14769816e-01
1.93859577e-01 -8.79016399e-01 -9.24625397e-01 -1.49434498e-02
9.41656902e-03 2.23372176e-01 5.84628522e-01 -1.54227123e-01
3.67097318e-01 5.61428487e-01 -9.67369556e-01 2.03678533e-01
5.44905543e-01 7.92524517e-01 -9.34951473e-03 7.91469753e-01
-1.46939412e-01 4.89639729e-01 -1.85207576e-01 -8.63525569e-01
3.27964455e-01 -3.96208316e-01 -1.11948931e+00 -2.86410693e-02
7.96957433e-01 -4.19468611e-01 -4.30428714e-01 7.94151902e-01
3.71296912e-01 1.17733493e-03 3.71821374e-01 -9.26998854e-01
-3.40012640e-01 7.21462727e-01 -8.72189760e-01 3.66069168e-01
3.78752589e-01 3.75091396e-02 1.34425020e+00 -9.29781854e-01
1.09156227e+00 9.08171594e-01 9.18082595e-01 2.27592155e-01
-1.21303964e+00 -4.56905127e-01 1.54786304e-01 -3.60666998e-02
-1.22321880e+00 -4.09580350e-01 9.80014622e-01 -4.82424945e-01
8.27309310e-01 2.36705750e-01 8.78191411e-01 6.24985039e-01
-5.24366722e-02 1.10093617e+00 6.83911145e-01 -5.42451560e-01
-1.59497365e-01 -2.82009065e-01 3.08170974e-01 1.19960678e+00
-3.57269756e-02 3.68673354e-02 -1.82238460e-01 5.91541715e-02
1.10891926e+00 -4.60424393e-01 -2.63771176e-01 -6.05455458e-01
-1.23878872e+00 8.80204618e-01 5.64714432e-01 1.96468100e-01
-4.91744280e-02 4.49839860e-01 2.98002154e-01 -1.39900461e-01
6.40143991e-01 5.82678139e-01 -2.44477287e-01 -5.14502585e-01
-1.20459116e+00 3.86079133e-01 5.49627423e-01 1.07682037e+00
7.35495865e-01 -1.93576187e-01 -7.89092332e-02 7.62619853e-01
-1.92570791e-01 1.45395622e-01 -1.89268619e-01 -1.22665191e+00
6.52456582e-01 2.34346449e-01 2.04655621e-02 -1.11497033e+00
-4.64449495e-01 -1.95327327e-01 -1.26280323e-01 3.41893673e-01
6.42656565e-01 -4.51369643e-01 -8.75040531e-01 1.23251033e+00
3.17575365e-01 1.18017398e-01 -4.23298091e-01 8.97526324e-01
8.94799888e-01 6.09169662e-01 -1.22234076e-01 4.09449607e-01
1.12726915e+00 -1.27382839e+00 -4.53482360e-01 -1.44832942e-04
7.85277307e-01 -1.24381888e+00 1.07863617e+00 1.61108345e-01
-1.48704493e+00 -3.88628274e-01 -9.97010469e-01 -4.98095661e-01
-1.23791240e-01 3.23437959e-01 1.03310847e+00 2.21993536e-01
-8.98828447e-01 7.34464467e-01 -1.07931280e+00 -1.76969230e-01
6.40618026e-01 8.58150572e-02 -2.50899136e-01 2.83755928e-01
-7.27386475e-01 6.98927224e-01 2.03192562e-01 1.44456238e-01
-2.18522906e-01 -6.25694752e-01 -1.10346413e+00 -8.15579966e-02
2.50238717e-01 -7.26532876e-01 1.53108382e+00 -6.67916536e-01
-1.17505860e+00 9.98621106e-01 -3.23807716e-01 -2.88406521e-01
7.19427824e-01 -4.80976850e-01 1.39346868e-01 4.05082881e-01
1.39344260e-01 9.64094162e-01 3.58815134e-01 -1.33620989e+00
-8.68510127e-01 2.62669414e-01 2.65980333e-01 3.60453278e-02
1.36115551e-01 5.76764829e-02 -9.86736238e-01 -8.78097415e-01
2.24769905e-01 -8.69014919e-01 -2.27473646e-01 3.47809911e-01
-9.47965205e-01 -3.20639253e-01 4.95493025e-01 -6.67743504e-01
1.45150590e+00 -1.74839795e+00 -7.79472850e-03 4.46929395e-01
2.37591982e-01 4.35820781e-02 -2.64066607e-02 4.17146593e-01
-1.31190985e-01 4.16595280e-01 -3.48664314e-01 -3.79987955e-01
-2.74913639e-01 -4.37380224e-01 -2.86874354e-01 3.50235224e-01
3.02036792e-01 1.03508580e+00 -9.05555308e-01 -7.55839586e-01
5.46641171e-01 5.19781053e-01 -2.57955700e-01 -1.77224688e-02
-2.94358004e-02 1.24124654e-01 -5.40196359e-01 7.07701981e-01
4.56960112e-01 -2.72759289e-01 -2.93488503e-01 -4.20499444e-01
-2.76664823e-01 9.88421440e-02 -8.46010745e-01 2.01660895e+00
-2.58789748e-01 1.29380608e+00 -5.72350681e-01 -4.99116242e-01
9.19361115e-01 -2.18217680e-03 5.13346732e-01 -5.87564766e-01
1.97462812e-01 -1.24476969e-01 -3.68595213e-01 -4.42085266e-01
5.81067920e-01 3.81245643e-01 -1.51451245e-01 2.39716813e-01
-3.06643844e-02 -4.18504238e-01 3.57333899e-01 3.14679325e-01
1.00090730e+00 6.34361029e-01 1.25450253e-01 -1.19519569e-01
1.31400384e-03 6.87469602e-01 3.95018905e-01 5.82877696e-01
2.19244920e-02 1.20181584e+00 5.79159617e-01 -9.49311316e-01
-1.63687980e+00 -1.32627213e+00 -1.85476542e-01 6.28294468e-01
4.95712876e-01 -7.33137548e-01 -1.04005766e+00 -6.59956694e-01
-1.77614093e-01 5.91161311e-01 -5.14272809e-01 5.46180785e-01
-9.59870100e-01 -2.75446326e-01 6.42262936e-01 8.53377283e-01
2.51139432e-01 -6.68491721e-01 -5.68537652e-01 1.05653033e-01
-1.86070681e-01 -1.15799308e+00 -6.54104710e-01 1.19089093e-02
-6.53528094e-01 -1.22738111e+00 -8.11222970e-01 -1.05083048e+00
8.11484993e-01 3.12457144e-01 1.30842006e+00 2.78532326e-01
-5.51030278e-01 4.37445305e-02 -4.88369554e-01 -3.03523809e-01
4.51763608e-02 2.94782162e-01 -7.55685508e-01 -5.85866034e-01
1.83334127e-01 -1.00093581e-01 -8.96766961e-01 1.13657981e-01
-4.81693596e-01 6.70798540e-01 1.02889471e-01 5.73992133e-01
9.11439598e-01 -3.71559590e-01 -1.79813504e-01 -1.19605839e+00
6.27352953e-01 -1.59405798e-01 -7.47956395e-01 2.84040511e-01
-3.15655917e-01 1.48190677e-01 3.66157562e-01 7.22197667e-02
-7.92507112e-01 4.08706605e-01 -4.98744041e-01 -2.75620759e-01
-1.92104250e-01 2.03570262e-01 5.47397852e-01 -1.08474597e-01
4.51526076e-01 -2.22524524e-01 -4.15164709e-01 -2.31385991e-01
6.79299831e-01 1.75488353e-01 7.32788384e-01 -5.87078273e-01
1.03738713e+00 5.64236999e-01 -1.04483880e-01 -7.20435143e-01
-8.88091922e-01 -4.66594905e-01 -9.18583989e-01 -3.46205115e-01
9.88822103e-01 -7.47125626e-01 -2.06302553e-01 1.89036742e-01
-1.46047211e+00 -5.54850280e-01 4.45867926e-02 2.40967378e-01
-6.44664943e-01 3.86274159e-01 -9.17808414e-01 -5.26032925e-01
-4.56430078e-01 -1.11026919e+00 1.27902985e+00 3.93256366e-01
-7.02363491e-01 -1.11872232e+00 1.31406710e-01 1.75743937e-01
-1.10607781e-01 7.08812118e-01 8.77085388e-01 -1.10901803e-01
-6.63909614e-01 -2.86732286e-01 -3.24062586e-01 -1.09300040e-01
-1.20922588e-01 7.74714708e-01 -6.29671693e-01 6.63077310e-02
-9.10230458e-01 -1.33441955e-01 8.57888281e-01 6.66315198e-01
1.31300092e+00 -1.41372932e-02 -5.79889894e-01 1.01865828e+00
1.59068489e+00 2.26497352e-01 6.34938240e-01 7.60091841e-01
9.14523542e-01 5.48999667e-01 7.31474996e-01 2.21749887e-01
5.71907163e-01 6.32258058e-01 5.40537015e-02 -8.74192834e-01
-2.79543549e-01 -5.16629279e-01 -3.12684417e-01 4.51819062e-01
-2.78204978e-01 -3.18349302e-01 -9.61104691e-01 7.01016426e-01
-1.90961707e+00 -1.06627250e+00 -4.27881867e-01 1.79984057e+00
5.82216442e-01 5.42172968e-01 2.84614742e-01 7.64168277e-02
7.36725271e-01 1.86656460e-01 -1.61754131e-01 -6.47297084e-01
-2.79558506e-02 3.86135310e-01 7.93342292e-01 7.03366399e-01
-1.41500533e+00 1.44174504e+00 6.92292023e+00 7.46742368e-01
-1.05976999e+00 -2.52185673e-01 9.65474308e-01 4.69859950e-02
-6.17707372e-01 1.97229013e-01 -7.04859078e-01 2.90723711e-01
1.74878299e-01 4.43183780e-02 -1.92172211e-02 9.97828543e-01
3.07827085e-01 -4.26748507e-02 -1.00805926e+00 9.58480179e-01
-2.35543385e-01 -1.86533618e+00 -2.76133150e-01 -3.60227853e-01
7.06193805e-01 5.52742071e-02 -7.19194114e-02 -2.73704112e-01
4.76109892e-01 -8.45015943e-01 8.99912655e-01 4.76579696e-01
8.30934048e-01 -8.02510977e-01 4.35180396e-01 -2.76926577e-01
-1.45331502e+00 5.28521180e-01 -1.86643943e-01 6.63248748e-02
5.70935428e-01 4.70835745e-01 -7.28706241e-01 3.35855484e-01
7.55610585e-01 6.02388501e-01 -7.10813820e-01 1.47204256e+00
-4.77050275e-01 6.93380475e-01 -4.69026834e-01 1.88902318e-02
4.42867041e-01 -1.30764171e-01 2.90790230e-01 1.62483060e+00
2.37331867e-01 6.60914332e-02 1.41472876e-01 8.32112968e-01
4.74092886e-02 3.35970998e-01 -7.86971331e-01 1.37202039e-01
7.09016979e-01 1.36972106e+00 -1.47909117e+00 -2.70575404e-01
-6.18274033e-01 9.88405585e-01 6.67539537e-01 4.52171475e-01
-1.02844620e+00 -9.24105287e-01 3.45563054e-01 1.64702848e-01
3.64389181e-01 -5.04993320e-01 -7.75929809e-01 -9.54736829e-01
2.26443186e-01 -1.28347844e-01 2.00494841e-01 -9.22177553e-01
-8.92556429e-01 9.18491066e-01 -1.74917933e-02 -1.25509357e+00
-2.90882051e-01 -4.58703399e-01 -9.73268509e-01 4.83988732e-01
-1.30446100e+00 -1.29251862e+00 -3.49516481e-01 2.62468785e-01
9.96583760e-01 3.01889062e-01 5.33988953e-01 -4.80339527e-02
-5.67605734e-01 7.51413226e-01 -6.51013777e-02 7.11161256e-01
3.99201304e-01 -1.36041796e+00 1.17687201e+00 1.19975042e+00
4.94486153e-01 5.06129563e-01 7.68468559e-01 -5.75661123e-01
-9.29603517e-01 -9.20827210e-01 7.88175523e-01 -5.02763033e-01
6.45522237e-01 -3.58416498e-01 -6.29931033e-01 8.39583099e-01
2.60920763e-01 -1.27886400e-01 4.62478310e-01 1.26311749e-01
-2.39799410e-01 2.31884271e-01 -7.15259194e-01 7.63210952e-01
1.12994301e+00 -2.18739256e-01 -4.00256634e-01 3.71221006e-01
5.73957384e-01 -8.47738981e-01 -8.93316865e-01 4.26822782e-01
6.74156606e-01 -1.07549942e+00 1.09813130e+00 -3.70875984e-01
8.80874157e-01 -1.20695151e-01 3.88784111e-01 -1.17414534e+00
-3.78082961e-01 -6.77701294e-01 1.65034935e-01 1.06762791e+00
8.96576643e-01 -3.25452536e-02 1.01303065e+00 6.01671755e-01
-2.90885597e-01 -1.30680299e+00 -5.13557851e-01 -4.32873070e-01
1.14740610e-01 -5.83809733e-01 4.09534872e-01 7.74039447e-01
1.50572225e-01 1.71041369e-01 -3.01907659e-01 3.70409228e-02
5.02336323e-01 4.43153322e-01 8.21476758e-01 -8.18192363e-01
-9.49166864e-02 -7.26532102e-01 -4.45092112e-01 -1.31520021e+00
5.16646691e-02 -7.64163613e-01 5.99434264e-02 -1.96394539e+00
-1.48647800e-01 -6.37626171e-01 1.65148124e-01 5.74468672e-01
-2.32033223e-01 6.27592742e-01 2.38710076e-01 4.61440273e-02
-5.56975782e-01 -2.36293506e-02 1.35387123e+00 -5.02255112e-02
-2.52354890e-01 -2.26166919e-01 -4.95339841e-01 9.58707571e-01
8.77854049e-01 -2.74645030e-01 -3.36422622e-01 -8.58296931e-01
2.89208796e-02 3.11198775e-02 2.54459798e-01 -8.26266527e-01
2.16586590e-01 -1.68049425e-01 6.48446977e-01 -7.03792274e-01
2.59145141e-01 -4.35645908e-01 -1.74521774e-01 4.21371125e-02
-4.42499429e-01 5.40631711e-01 2.38545880e-01 -3.75702195e-02
-1.63596004e-01 -2.59962797e-01 5.33460736e-01 -1.28138706e-01
-1.01574206e+00 4.29338455e-01 -1.03642590e-01 1.33030280e-01
1.27346241e+00 -4.11219895e-01 -5.37366986e-01 -4.73909944e-01
-5.79536796e-01 3.22067410e-01 6.85040474e-01 4.38569516e-01
8.54570568e-01 -1.18654883e+00 -6.62251413e-01 -1.03944927e-01
-4.80981879e-02 1.07485250e-01 -9.52671692e-02 3.55903715e-01
-1.40005648e+00 3.42686504e-01 -3.28767411e-02 -5.73014081e-01
-1.40243316e+00 3.60534132e-01 3.00238252e-01 1.97587892e-01
-1.00746286e+00 1.09579122e+00 1.38863809e-02 2.01672167e-01
1.38834953e-01 -4.44386810e-01 4.19522189e-02 -3.19054455e-01
2.06642941e-01 3.07089031e-01 -1.91359982e-01 -6.33442998e-01
-1.08390652e-01 1.18083954e+00 -5.58374673e-02 -1.12344854e-01
1.30721402e+00 -6.80041537e-02 2.79123783e-01 2.31954023e-01
1.41310656e+00 1.02579221e-01 -1.33481312e+00 1.70272440e-01
1.83104854e-02 -5.20936787e-01 4.48856317e-02 -3.40386301e-01
-1.09403098e+00 7.32578576e-01 2.87753671e-01 1.29610956e-01
8.51311505e-01 3.03009212e-01 1.15628195e+00 5.25760315e-02
2.42378414e-01 -1.10953188e+00 -2.88987517e-01 1.85584858e-01
6.24284804e-01 -1.17631340e+00 2.41926573e-02 -1.03929007e+00
-5.55891037e-01 1.57814229e+00 6.15789354e-01 -5.40205479e-01
5.24450302e-01 7.05767095e-01 2.99962193e-01 -2.78824359e-01
-4.14651453e-01 -4.34160709e-01 5.02148151e-01 4.54603970e-01
7.65235126e-01 -3.50910127e-02 -4.07988966e-01 1.90927964e-02
-7.81685114e-01 -1.88842058e-01 3.59439075e-01 9.38595176e-01
-3.20551574e-01 -1.17288637e+00 -3.09722275e-02 3.46137822e-01
-5.81887126e-01 -2.41388127e-01 -3.47541213e-01 1.00596809e+00
-1.11027509e-01 7.41887748e-01 2.64874339e-01 -4.75849956e-01
5.26501060e-01 -2.31204018e-01 4.95087206e-01 -3.69992554e-01
-2.58441180e-01 2.58799583e-01 4.44353580e-01 -8.38981152e-01
-3.20700884e-01 -5.52401781e-01 -1.56279945e+00 -2.22383261e-01
-2.87594914e-01 -5.54507039e-02 5.69691360e-01 7.09041059e-01
2.25429088e-01 4.26839173e-01 4.60861832e-01 -9.19636667e-01
2.82290608e-01 -4.29438025e-01 9.21543315e-02 6.37192309e-01
2.89773941e-01 -4.45481598e-01 2.16977466e-02 3.88287485e-01] | [8.314665794372559, -1.695186734199524] |
02af9bf4-4f23-478c-8224-bb06ffb0e88a | attention-is-all-we-need-nailing-down-object | 1807.11794 | null | http://arxiv.org/abs/1807.11794v1 | http://arxiv.org/pdf/1807.11794v1.pdf | Attention is All We Need: Nailing Down Object-centric Attention for Egocentric Activity Recognition | In this paper we propose an end-to-end trainable deep neural network model
for egocentric activity recognition. Our model is built on the observation that
egocentric activities are highly characterized by the objects and their
locations in the video. Based on this, we develop a spatial attention mechanism
that enables the network to attend to regions containing objects that are
correlated with the activity under consideration. We learn highly specialized
attention maps for each frame using class-specific activations from a CNN
pre-trained for generic image recognition, and use them for spatio-temporal
encoding of the video with a convolutional LSTM. Our model is trained in a
weakly supervised setting using raw video-level activity-class labels.
Nonetheless, on standard egocentric activity benchmarks our model surpasses by
up to +6% points recognition accuracy the currently best performing method that
leverages hand segmentation and object location strong supervision for
training. We visually analyze attention maps generated by the network,
revealing that the network successfully identifies the relevant objects present
in the video frames which may explain the strong recognition performance. We
also discuss an extensive ablation analysis regarding the design choices. | ['Swathikiran Sudhakaran', 'Oswald Lanz'] | 2018-07-31 | null | null | null | null | ['egocentric-activity-recognition', 'hand-segmentation'] | ['computer-vision', 'computer-vision'] | [ 3.53712082e-01 3.83746736e-02 -3.27486932e-01 -2.95811266e-01
-6.66584373e-01 -5.63710690e-01 7.01032162e-01 -2.65155941e-01
-6.29061222e-01 4.13266689e-01 6.62087917e-01 3.60187382e-01
-1.87017415e-02 -3.67759943e-01 -1.12254548e+00 -6.57319665e-01
-2.09417269e-01 3.14052165e-01 6.24073343e-03 2.53738016e-01
3.33920360e-01 4.88326937e-01 -1.57240474e+00 5.75526237e-01
2.07929060e-01 1.16240180e+00 3.01000118e-01 9.16478276e-01
3.96714389e-01 1.33499479e+00 -4.53145623e-01 2.82909330e-02
4.70047258e-02 -4.54560518e-01 -1.11521339e+00 2.80608863e-01
8.58286500e-01 -8.07139397e-01 -8.31039786e-01 6.95159793e-01
2.23797202e-01 4.03227180e-01 7.42348552e-01 -9.91062999e-01
-5.69906116e-01 4.67663080e-01 -4.03055727e-01 9.64116275e-01
4.11737829e-01 3.40799421e-01 1.08428514e+00 -9.22372162e-01
8.37574065e-01 7.64386535e-01 3.37396950e-01 4.84876007e-01
-1.09853590e+00 -1.71076149e-01 5.00420392e-01 4.23130840e-01
-1.22477746e+00 -5.61859548e-01 7.90569842e-01 -7.03501225e-01
1.43682206e+00 -1.16972148e-01 8.50381732e-01 1.62617779e+00
8.09838027e-02 1.26500332e+00 5.44278800e-01 -2.24334806e-01
1.87057525e-01 -4.52106625e-01 4.28756624e-02 6.79894626e-01
-1.55311421e-01 -2.21672148e-01 -8.15736890e-01 2.80599385e-01
1.12348747e+00 2.51991898e-01 -2.64134824e-01 -7.21386671e-01
-1.56999826e+00 5.10302126e-01 7.31274784e-01 3.92592967e-01
-7.39101350e-01 7.75501072e-01 3.33383709e-01 -2.47342274e-01
5.07853806e-01 3.18358839e-01 -4.33319628e-01 -5.12478769e-01
-9.40351963e-01 1.54796809e-01 2.26059347e-01 8.83256018e-01
5.09946704e-01 1.21535584e-01 -5.44019580e-01 4.67335314e-01
6.61821440e-02 1.77971110e-01 4.46053326e-01 -1.24709833e+00
5.43974817e-01 4.72533166e-01 5.64615317e-02 -6.33584201e-01
-3.36408406e-01 -8.08130801e-01 -2.72329655e-02 6.80053979e-02
5.25675654e-01 -7.81023363e-03 -1.03454578e+00 1.86444390e+00
-1.62669063e-01 4.55276102e-01 -1.58874512e-01 1.19002283e+00
4.13965255e-01 2.29267821e-01 5.54671407e-01 2.95253843e-01
1.29667950e+00 -1.15850425e+00 -3.51156443e-01 -4.18962151e-01
7.89075792e-01 -1.55104741e-01 1.26978981e+00 1.36741981e-01
-1.28056836e+00 -4.93982047e-01 -9.10334468e-01 -1.19028501e-01
-1.60555959e-01 3.19227487e-01 7.28892088e-01 2.78757662e-01
-9.83560860e-01 6.24860823e-01 -1.07348061e+00 -6.50183320e-01
9.05062139e-01 3.88214588e-01 -5.06391346e-01 1.98614389e-01
-5.79538167e-01 7.64539480e-01 2.70007670e-01 -3.32257822e-02
-1.65138507e+00 -5.63686609e-01 -9.21158671e-01 3.70274782e-01
2.46823981e-01 -7.51308918e-01 1.29632843e+00 -1.39622211e+00
-1.14014292e+00 1.03685939e+00 -2.95613825e-01 -6.07851565e-01
3.84389699e-01 -5.49724042e-01 -4.39701080e-02 6.06515944e-01
2.65678376e-01 8.93667340e-01 7.48873174e-01 -8.33051741e-01
-6.93517625e-01 -3.83251548e-01 4.74692672e-01 3.20224017e-01
-2.78704524e-01 1.47840977e-01 -6.96037531e-01 -8.51718485e-01
-9.98540968e-03 -7.73725867e-01 1.56625673e-01 -1.78368062e-01
-1.82647839e-01 -2.91949630e-01 8.81442368e-01 -6.45215034e-01
7.49714613e-01 -2.06678391e+00 3.90900552e-01 -1.22035053e-02
1.73877731e-01 -1.03893504e-01 -2.66793430e-01 1.66330755e-01
-3.81145298e-01 -9.80795622e-02 -1.55736476e-01 -3.99484724e-01
-1.84823602e-01 3.50888744e-02 -4.63784248e-01 7.06199944e-01
3.66325319e-01 1.35859740e+00 -1.07910895e+00 -2.37980559e-01
3.62477988e-01 5.71610868e-01 -8.21300089e-01 3.36995691e-01
-2.99400866e-01 5.65853536e-01 -2.81753212e-01 7.79485703e-01
-1.90084949e-02 -3.93320799e-01 2.07198206e-02 -3.54629755e-01
2.26095825e-01 4.23162758e-01 -5.31572938e-01 2.33803248e+00
-4.05000299e-01 9.92454469e-01 -3.16808850e-01 -1.05418921e+00
2.09771514e-01 4.40342307e-01 9.01127279e-01 -6.11000359e-01
2.18243480e-01 -1.48584321e-01 -4.30448018e-02 -7.52665520e-01
2.85505146e-01 3.99811506e-01 7.36651197e-02 7.40203261e-01
6.73433602e-01 6.64596379e-01 1.18571833e-01 3.28273118e-01
1.32041156e+00 7.73102641e-01 -9.35630128e-02 -1.00696512e-01
3.34414303e-01 -5.18490411e-02 2.31350482e-01 8.16177249e-01
-4.96211559e-01 8.84082496e-01 6.20478213e-01 -4.41596776e-01
-1.09445584e+00 -1.07602549e+00 2.54899502e-01 1.60851860e+00
2.57551800e-02 -3.76830310e-01 -9.97855842e-01 -9.95492041e-01
-3.65657985e-01 4.60426718e-01 -1.10595918e+00 -2.20954001e-01
-7.50639796e-01 -2.58874416e-01 5.17484903e-01 1.24099278e+00
5.43392718e-01 -1.45474792e+00 -9.11402583e-01 1.11253023e-01
-2.53440261e-01 -1.33605921e+00 -4.75811809e-01 2.26908714e-01
-8.73859704e-01 -1.27435207e+00 -1.03043425e+00 -8.15625191e-01
8.98638546e-01 3.07204723e-01 1.05504608e+00 -2.88681358e-01
-2.81753898e-01 1.09404016e+00 -2.86544263e-01 -3.40803377e-02
4.01528567e-01 4.20078069e-01 -3.07976659e-02 1.53021172e-01
6.31802678e-01 -6.31333590e-01 -1.01232040e+00 1.59767866e-01
-4.62190151e-01 -6.53206557e-02 4.14675713e-01 5.59309185e-01
4.01152879e-01 -6.61189973e-01 2.18898460e-01 -5.54762423e-01
2.15375796e-01 -5.58691442e-01 -2.22450852e-01 1.66624397e-01
7.03552812e-02 5.56437112e-02 1.22200944e-01 -2.56599128e-01
-8.56525421e-01 5.33612728e-01 1.34842336e-01 -7.87281513e-01
-3.77561152e-01 2.92934835e-01 -4.35500033e-02 7.87078291e-02
7.22928047e-01 3.65212172e-01 -3.57703418e-01 -4.03960079e-01
3.19847345e-01 3.64118487e-01 9.15513515e-01 -5.22969663e-01
3.40663493e-01 7.67757118e-01 -3.42425346e-01 -5.03598452e-01
-1.06782401e+00 -4.13183004e-01 -9.25843537e-01 -3.75647604e-01
1.33451831e+00 -1.17268765e+00 -8.64555001e-01 2.92112261e-01
-1.24947023e+00 -6.33299351e-01 -3.40505272e-01 6.77636504e-01
-1.10464716e+00 9.79917720e-02 -4.79900539e-01 -7.22853422e-01
-6.22070453e-04 -1.01932991e+00 1.33243906e+00 -4.00798433e-02
-5.48446715e-01 -9.35764670e-01 -3.20443287e-02 4.99315053e-01
2.22211272e-01 1.86419070e-01 5.94158649e-01 -4.89176214e-01
-8.83034766e-01 -1.12280741e-01 -2.66801357e-01 1.43617904e-02
1.01991609e-01 -1.85456097e-01 -1.45833147e+00 -2.08495006e-01
-1.70222461e-01 -3.18849206e-01 1.06188238e+00 6.18809283e-01
1.62834096e+00 -2.43071571e-01 -4.00678277e-01 8.52366090e-01
9.63236332e-01 1.24884397e-03 5.72838843e-01 4.23582792e-01
9.07106936e-01 6.88122332e-01 3.50896448e-01 2.25252405e-01
2.51759648e-01 8.04355204e-01 7.63839662e-01 -5.15280850e-02
-9.12371650e-02 -2.74740875e-01 5.18540084e-01 -1.04139887e-01
-4.32947397e-01 -7.20446780e-02 -7.85890579e-01 8.43886793e-01
-2.08684349e+00 -1.39346302e+00 3.96321446e-01 1.95976663e+00
3.47494900e-01 9.05439779e-02 4.42825347e-01 -1.28030285e-01
4.39910144e-01 4.49554414e-01 -7.14047134e-01 -3.92138064e-02
-2.96080485e-03 1.52564242e-01 5.59671521e-01 2.79563427e-01
-1.53404224e+00 1.15364754e+00 7.02294922e+00 2.40590215e-01
-1.20716512e+00 2.55812764e-01 4.96772319e-01 -7.05534399e-01
1.37567416e-01 -3.37155104e-01 -4.65649396e-01 4.26879704e-01
9.01513219e-01 1.94475085e-01 5.75715005e-01 1.15837276e+00
3.00508142e-01 -2.90984735e-02 -1.49154115e+00 1.19120526e+00
3.26417893e-01 -1.42612863e+00 5.47685139e-02 2.31859773e-01
6.70108676e-01 4.27302808e-01 2.99398005e-01 1.60212845e-01
9.92018580e-02 -1.27162182e+00 9.52758133e-01 5.48692703e-01
5.52038670e-01 -5.03293753e-01 3.19483727e-01 2.12616641e-02
-1.08073688e+00 -3.56752276e-01 -1.43577471e-01 -8.22183266e-02
7.97374845e-02 -1.35688275e-01 -6.70037806e-01 -9.16698724e-02
9.00920510e-01 1.22105241e+00 -5.92833817e-01 8.90227735e-01
-3.94764841e-01 7.20118403e-01 -1.24108091e-01 3.68122399e-01
5.60889900e-01 1.80752054e-01 4.27933961e-01 1.31709182e+00
1.36195540e-01 -5.53130032e-03 -2.04675019e-01 1.06711435e+00
-1.83805764e-01 -2.71855354e-01 -7.02035666e-01 -1.61530659e-01
-5.25379516e-02 9.42708671e-01 -6.85186327e-01 -4.80569482e-01
-3.29400003e-01 1.34628749e+00 7.19357431e-01 7.81499982e-01
-1.01395190e+00 -2.44297192e-01 8.33988011e-01 1.68899417e-01
6.04818165e-01 -4.71574306e-01 -1.90713808e-01 -1.28248847e+00
1.89454958e-01 -5.46838999e-01 4.83058572e-01 -1.09480202e+00
-6.82195425e-01 4.67908651e-01 -1.21239968e-01 -1.17224777e+00
-6.26894474e-01 -7.92001486e-01 -6.52411759e-01 6.52961016e-01
-1.11297262e+00 -1.17944539e+00 -6.70709193e-01 8.23490024e-01
8.05154443e-01 -2.78069615e-01 6.53551996e-01 2.41817027e-01
-6.24962747e-01 4.18719381e-01 -3.51000726e-01 5.71976125e-01
3.58414918e-01 -1.19054687e+00 5.75336397e-01 9.42238688e-01
6.19535506e-01 8.22911143e-01 3.84745181e-01 -3.96645993e-01
-1.13170755e+00 -9.95456696e-01 6.89241529e-01 -1.06038308e+00
5.22673666e-01 -5.31479895e-01 -5.49864471e-01 1.27384043e+00
2.06870824e-01 3.08544695e-01 5.78333557e-01 1.34384513e-01
-4.87123430e-01 4.91744168e-02 -7.50596046e-01 6.63405359e-01
1.68186271e+00 -1.08183503e+00 -6.81692064e-01 5.66984653e-01
3.26849520e-01 -3.23914558e-01 -4.87605065e-01 1.31670088e-01
5.74310303e-01 -8.83079290e-01 1.07215834e+00 -1.18547785e+00
6.92918181e-01 -1.71725348e-01 -2.23235399e-01 -9.63310122e-01
-5.04329860e-01 -2.94435978e-01 -4.87498611e-01 7.55373120e-01
1.83513105e-01 -1.17920257e-01 1.23260677e+00 5.90218484e-01
-1.35650426e-01 -5.83175719e-01 -8.96174908e-01 -5.74393690e-01
-3.55786204e-01 -4.61915165e-01 1.57075033e-01 7.54218638e-01
1.35205552e-01 1.53839409e-01 -2.68333018e-01 -8.44900385e-02
5.03083527e-01 -1.28084138e-01 6.70070469e-01 -6.99181914e-01
-2.88331628e-01 -5.12917876e-01 -6.72350824e-01 -1.35486877e+00
3.88999164e-01 -8.77722681e-01 -4.88978475e-02 -1.50406539e+00
3.65288287e-01 5.55101223e-02 -7.37375557e-01 5.78054130e-01
1.31619973e-02 7.18044817e-01 7.66233057e-02 3.52241755e-01
-8.66946578e-01 3.87799680e-01 9.23671722e-01 -2.95121789e-01
-1.21715777e-01 -2.17502370e-01 -5.70434690e-01 6.60811186e-01
6.77745998e-01 -3.18136990e-01 -4.98420000e-01 -8.28266382e-01
-3.32485586e-02 -3.55879098e-01 8.10730875e-01 -1.33703148e+00
1.30691081e-01 -2.82627828e-02 7.14626372e-01 -4.30783063e-01
6.96100473e-01 -8.86212945e-01 -3.31390917e-01 3.79544765e-01
-7.58182049e-01 -2.27793843e-01 2.47238502e-02 5.55201292e-01
-7.98277855e-02 -1.51150348e-02 3.88569713e-01 -1.52220294e-01
-1.06656003e+00 4.47333306e-01 -5.45837462e-01 2.58805393e-03
1.11568904e+00 -5.97171545e-01 -3.74944031e-01 -5.98788321e-01
-1.03881574e+00 -5.79067022e-02 4.81148958e-01 4.84961867e-01
4.26906317e-01 -1.24073279e+00 -3.45931470e-01 1.21712826e-01
3.70638818e-01 -3.26560140e-01 2.58439749e-01 9.98890281e-01
-4.59005803e-01 7.06249177e-01 -5.13520598e-01 -9.00272667e-01
-9.86057997e-01 6.05080366e-01 6.75683081e-01 1.94668785e-01
-7.40902722e-01 9.70034957e-01 6.22161269e-01 1.69709399e-01
4.88263875e-01 -4.26601797e-01 -2.58238763e-01 -1.35657534e-01
4.09630328e-01 2.40661174e-01 -1.15486860e-01 -8.96941721e-01
-5.42084754e-01 6.30200446e-01 -2.96448637e-02 -3.13372135e-01
1.31274939e+00 1.04455359e-01 3.97682905e-01 2.24471048e-01
1.26028681e+00 -2.81076670e-01 -2.03508472e+00 -1.24708787e-01
4.41800021e-02 -5.02627909e-01 8.31253156e-02 -7.42519498e-01
-1.34714365e+00 1.06198967e+00 5.78245580e-01 -4.83923256e-01
8.25980544e-01 2.39151716e-01 5.66365600e-01 4.67317820e-01
2.12599158e-01 -1.06332827e+00 5.00752389e-01 5.02537966e-01
8.60126317e-01 -1.13278723e+00 -1.77600369e-01 -2.11994182e-02
-7.65958905e-01 9.31051970e-01 7.83749342e-01 -4.41465378e-01
2.57871866e-01 -1.09540574e-01 -6.86520636e-02 -3.85180235e-01
-5.01611292e-01 -3.38687539e-01 3.16765904e-01 7.12981224e-01
4.67849404e-01 -3.15232426e-01 2.50757754e-01 4.47839618e-01
9.66878980e-02 5.08046336e-02 2.08746716e-01 9.16825235e-01
-3.84657234e-01 -5.41114330e-01 2.67071184e-02 3.44172567e-01
-5.23178875e-01 1.85636848e-01 -5.24572611e-01 5.00908732e-01
-8.97046626e-02 5.40090263e-01 6.27376676e-01 -3.10051411e-01
1.85820982e-01 1.62333503e-01 7.72891283e-01 -6.71438098e-01
-6.66952252e-01 -7.94503689e-02 -7.12321624e-02 -1.06382418e+00
-6.08918190e-01 -6.93545222e-01 -1.18651390e+00 9.07781422e-02
3.19556564e-01 -1.64274812e-01 5.16613245e-01 1.01927042e+00
4.73740369e-01 7.12811887e-01 2.49809876e-01 -1.28488362e+00
-8.46396536e-02 -8.55224192e-01 -4.43484366e-01 5.52119672e-01
3.74240816e-01 -8.50579202e-01 -1.60426721e-01 3.61424476e-01] | [8.316555976867676, 0.614240825176239] |
4b8209b1-0bc5-42bc-83dd-4e293b22d8ad | active-passive-simstereo-benchmarking-the | 2209.08305 | null | https://arxiv.org/abs/2209.08305v1 | https://arxiv.org/pdf/2209.08305v1.pdf | Active-Passive SimStereo -- Benchmarking the Cross-Generalization Capabilities of Deep Learning-based Stereo Methods | In stereo vision, self-similar or bland regions can make it difficult to match patches between two images. Active stereo-based methods mitigate this problem by projecting a pseudo-random pattern on the scene so that each patch of an image pair can be identified without ambiguity. However, the projected pattern significantly alters the appearance of the image. If this pattern acts as a form of adversarial noise, it could negatively impact the performance of deep learning-based methods, which are now the de-facto standard for dense stereo vision. In this paper, we propose the Active-Passive SimStereo dataset and a corresponding benchmark to evaluate the performance gap between passive and active stereo images for stereo matching algorithms. Using the proposed benchmark and an additional ablation study, we show that the feature extraction and matching modules of a selection of twenty selected deep learning-based stereo matching methods generalize to active stereo without a problem. However, the disparity refinement modules of three of the twenty architectures (ACVNet, CascadeStereo, and StereoNet) are negatively affected by the active stereo patterns due to their reliance on the appearance of the input images. | ['Mohammed Bennamoun', 'Farid Boussaid', 'Hamid Laga', 'Lian Xu', 'Allen Antony', 'Laurent Jospin'] | 2022-09-17 | null | null | null | null | ['stereo-matching-1'] | ['computer-vision'] | [ 4.98986095e-01 2.72710621e-01 2.66502649e-01 -3.92359048e-01
-5.09069085e-01 -7.49202847e-01 8.46567273e-01 -1.90467939e-01
-6.14740968e-01 5.49997568e-01 1.14008449e-01 9.69552249e-02
1.86731234e-01 -8.79583418e-01 -9.50571537e-01 -8.17273080e-01
5.34091711e-01 2.88195699e-01 6.86046958e-01 -3.53620738e-01
3.83034289e-01 5.72766483e-01 -1.72321284e+00 3.97712708e-01
8.00257027e-01 9.23786044e-01 1.66900441e-01 8.21193531e-02
-8.27639326e-02 6.93104446e-01 -4.79467511e-01 -5.65404594e-01
1.05047143e+00 -1.02947049e-01 -7.63134062e-01 3.36348601e-02
1.34115899e+00 -4.14578497e-01 -6.42774701e-01 1.28205264e+00
5.92266083e-01 -7.07134381e-02 3.91652435e-01 -1.23291171e+00
-1.73587739e-01 8.35530981e-02 -5.49658418e-01 1.04267843e-01
4.34571892e-01 4.85524029e-01 7.51263440e-01 -7.77680695e-01
9.55252647e-01 1.22586691e+00 7.87354171e-01 6.76469505e-01
-1.56694674e+00 -9.06729460e-01 -7.38700703e-02 1.53922275e-01
-1.18749964e+00 -9.42315638e-01 8.98465753e-01 -5.75136662e-01
6.72838569e-01 3.42756778e-01 8.24310482e-01 1.06873977e+00
1.82754084e-01 6.47365212e-01 1.28122318e+00 -2.78718263e-01
2.65477061e-01 2.18256749e-02 -1.39621273e-01 5.39074659e-01
2.45098114e-01 9.42679048e-01 -7.22779095e-01 -2.18624219e-01
8.08728814e-01 2.31559435e-03 -4.08691019e-01 -1.10611653e+00
-1.02721417e+00 6.53903842e-01 7.69758105e-01 -3.79384682e-02
-1.79207027e-01 -9.69472155e-02 6.37641698e-02 3.68047446e-01
3.90464902e-01 6.96945727e-01 9.49591622e-02 4.11760360e-01
-1.00292826e+00 2.49803349e-01 5.00787139e-01 7.63482451e-01
1.34756279e+00 7.68334344e-02 4.83607017e-02 7.12292194e-01
1.16199464e-01 3.70361209e-01 1.69130176e-01 -1.28252423e+00
4.14782286e-01 8.70057046e-01 -8.69839266e-02 -1.16560662e+00
-2.04270050e-01 -2.17063203e-01 -8.62167180e-01 1.02095723e+00
4.99701858e-01 1.24339424e-01 -1.01009178e+00 1.59833968e+00
1.29796773e-01 -6.12025671e-02 -2.04264686e-01 1.05422783e+00
1.09192002e+00 1.36316285e-01 -2.78020352e-01 1.65039003e-01
3.57128233e-01 -8.82824838e-01 -1.74065575e-01 -6.20024145e-01
2.06061453e-01 -9.40494359e-01 7.01557636e-01 -2.26659253e-02
-1.22662377e+00 -6.75692379e-01 -1.15504754e+00 -1.69767544e-01
-2.91939914e-01 -7.10222602e-01 6.68714166e-01 4.54387724e-01
-1.35329413e+00 5.60342968e-01 -4.37573761e-01 -3.48501980e-01
4.78056401e-01 5.50263524e-01 -7.21416533e-01 -8.79003927e-02
-1.06401241e+00 7.90135324e-01 2.65215039e-01 -1.17497109e-01
-7.54875600e-01 -8.88246894e-01 -9.94186282e-01 -2.47102797e-01
7.49018323e-03 -9.90120888e-01 7.90191293e-01 -1.68543684e+00
-1.12496686e+00 1.64607847e+00 -2.37884056e-02 -5.75199366e-01
9.72165942e-01 6.86798543e-02 -4.65146527e-02 2.94801481e-02
1.89718261e-01 1.27460849e+00 1.07850409e+00 -1.41855109e+00
-6.50047541e-01 -4.56477404e-01 4.33859140e-01 3.17966491e-01
2.28496909e-01 -3.49624425e-01 -5.58402896e-01 -5.20417392e-01
2.82520056e-01 -1.01199114e+00 -3.92933518e-01 4.56606090e-01
-4.42177147e-01 4.45576578e-01 6.14835918e-01 -2.16679037e-01
5.26755393e-01 -2.58460760e+00 9.16929767e-02 1.52350038e-01
3.51109892e-01 1.37508422e-01 -2.90646136e-01 7.80827850e-02
-3.00523430e-01 -2.61034191e-01 -3.21133643e-01 -3.82649481e-01
-4.47186381e-01 1.82776645e-01 -3.93681973e-01 7.10653722e-01
-3.60825472e-02 7.75003254e-01 -8.14204156e-01 -4.41737324e-01
6.22418940e-01 2.14094490e-01 -7.00605571e-01 3.21356684e-01
2.06274658e-01 6.18102610e-01 1.03924401e-01 2.56234348e-01
1.01464641e+00 1.86588522e-02 -2.48738468e-01 -2.80857682e-01
-1.95424840e-01 1.65493399e-01 -1.05504966e+00 1.96447647e+00
-4.09649014e-01 9.82111335e-01 -2.35907957e-02 -7.81075299e-01
7.82928407e-01 6.85435459e-02 5.27204037e-01 -1.35034382e+00
-2.18638837e-01 2.66415238e-01 1.48308977e-01 2.03137100e-02
4.34544414e-01 1.18003957e-01 2.62497485e-01 2.76251733e-01
-9.17332619e-02 -3.21883857e-01 8.29053074e-02 1.46480247e-01
1.03388584e+00 -1.08140446e-01 1.70760006e-01 -3.90707940e-01
5.32175303e-01 1.58230677e-01 6.48391783e-01 9.84311104e-01
-1.77331001e-01 1.34525168e+00 -6.88330233e-02 -6.73147082e-01
-1.27320886e+00 -1.35730016e+00 -2.88323164e-01 5.46871424e-01
5.15121043e-01 -4.81099263e-02 -5.71994901e-01 -5.72798491e-01
1.49968520e-01 3.00392330e-01 -7.42334604e-01 -2.71692723e-01
-5.47206640e-01 -3.67132813e-01 3.91728878e-01 2.68805087e-01
9.19278085e-01 -9.57670391e-01 -5.85262299e-01 -4.49475572e-02
-8.48804787e-02 -1.00508654e+00 -4.61513191e-01 4.03388180e-02
-8.65471303e-01 -1.39713275e+00 -8.13070059e-01 -7.81083941e-01
9.58910108e-01 6.01759911e-01 1.34972179e+00 2.48016436e-02
-2.78615057e-01 3.31556618e-01 5.01140468e-02 -1.36303708e-01
-5.14365613e-01 -1.47836149e-01 -2.03823879e-01 9.25294608e-02
1.39298677e-01 -7.41264224e-01 -9.83551443e-01 5.07849097e-01
-8.49747241e-01 4.91470605e-01 2.53756464e-01 7.80441701e-01
5.63335836e-01 -4.29566771e-01 -3.37311268e-01 -1.04635608e+00
3.53367589e-02 1.91414356e-03 -8.58214140e-01 -7.96281919e-02
-3.65512192e-01 3.68442945e-02 4.12802041e-01 -2.52403319e-01
-8.62119853e-01 6.20034873e-01 -8.04620758e-02 -4.99741256e-01
-2.00374499e-01 -1.29160911e-01 -2.41592959e-01 -6.21505857e-01
9.34044302e-01 2.13188604e-01 4.30622920e-02 -2.96601206e-01
-2.80168541e-02 2.44228691e-01 6.90872729e-01 -3.63034047e-02
1.25828290e+00 1.08571780e+00 1.00501375e-02 -6.21253490e-01
-9.06843662e-01 -6.30792260e-01 -6.80057883e-01 -2.14085177e-01
6.91398740e-01 -1.12938511e+00 -2.50625134e-01 7.65566111e-01
-1.21430886e+00 -2.91679502e-01 -5.05123854e-01 1.25667736e-01
-6.92721069e-01 4.75070149e-01 -1.65352657e-01 -2.21689418e-01
-2.17571855e-01 -1.37156069e+00 1.02620590e+00 2.18874514e-01
-4.54858333e-01 -1.00360668e+00 2.22670808e-01 4.27308738e-01
4.57304001e-01 2.71215349e-01 7.42015302e-01 -3.21870476e-01
-7.81320691e-01 1.63573902e-02 -1.96446270e-01 3.35890383e-01
6.35139048e-02 -1.79159567e-01 -1.46674085e+00 -4.17973399e-01
8.04789215e-02 -2.38671109e-01 9.73083496e-01 3.11165720e-01
8.69098604e-01 1.65441167e-02 -1.07425913e-01 1.16352820e+00
1.47117877e+00 2.63149053e-01 1.09405339e+00 7.29789734e-01
7.19960809e-01 8.02573562e-01 2.81909764e-01 -1.41628027e-01
-1.26007274e-01 1.00534678e+00 5.79539895e-01 -7.98069179e-01
-3.96376729e-01 -2.59850562e-01 2.84917116e-01 3.41363609e-01
3.11707091e-02 2.39430740e-01 -9.90745127e-01 3.40378851e-01
-1.57955694e+00 -1.16452157e+00 -9.57987756e-02 2.56778264e+00
7.20631003e-01 2.37630829e-01 -2.06209287e-01 2.95666665e-01
8.90778720e-01 3.76607388e-01 -6.47356629e-01 -1.31960094e-01
-6.11320078e-01 2.21999392e-01 6.99054658e-01 4.95547384e-01
-1.23683584e+00 8.02738667e-01 6.39201689e+00 6.02867782e-01
-1.43405366e+00 -7.18696192e-02 5.92323720e-01 -6.40843585e-02
-2.98399299e-01 2.34264880e-01 -4.13991988e-01 3.70302320e-01
3.15012150e-02 -1.28220022e-01 3.26260209e-01 6.54741526e-01
8.93227663e-03 -2.89320856e-01 -1.47539961e+00 1.45561016e+00
7.74938613e-02 -1.57966495e+00 2.09424645e-01 2.41729304e-01
1.06867099e+00 3.64803374e-01 2.41680726e-01 -1.90380245e-01
2.06520811e-01 -9.36190128e-01 9.29067135e-01 3.08927774e-01
8.10655475e-01 -3.83644581e-01 5.65499127e-01 1.42235219e-01
-6.34906709e-01 1.24129578e-01 -4.92236704e-01 -7.01497570e-02
4.04691249e-02 4.90313411e-01 -3.79813164e-01 2.14201733e-01
8.11843693e-01 7.82291412e-01 -8.75350177e-01 1.52213597e+00
-5.39599471e-02 9.68010724e-02 -3.13142329e-01 6.69947684e-01
1.38654932e-01 -1.11590683e-01 8.67929697e-01 7.99769223e-01
-6.12969548e-02 -3.55316222e-01 -1.34520933e-01 1.02202725e+00
2.98609678e-02 -2.37573728e-01 -1.07190442e+00 4.22993481e-01
3.94851446e-01 8.27502668e-01 -4.19906676e-01 -5.70027232e-02
-4.96491849e-01 1.20365691e+00 1.83317512e-01 4.36672211e-01
-4.00650293e-01 3.14072520e-02 8.00094724e-01 4.73628104e-01
-1.32353291e-01 1.45521536e-01 -4.75351423e-01 -1.20580864e+00
-2.02129390e-02 -1.03511989e+00 2.08231345e-01 -8.56501281e-01
-1.32896781e+00 3.99314016e-01 -2.62132794e-01 -1.63298166e+00
-2.91075081e-01 -2.29095593e-01 -6.45730913e-01 9.47116792e-01
-1.29782724e+00 -7.69437373e-01 -6.01160169e-01 1.01654816e+00
3.91347557e-01 -3.48755211e-01 6.83582783e-01 3.00217658e-01
-1.71396554e-01 6.50103629e-01 1.62579119e-01 3.90184015e-01
8.91993225e-01 -1.06131613e+00 8.45164478e-01 8.83222222e-01
2.04605162e-01 5.01016796e-01 6.93643093e-01 -2.48357251e-01
-1.11391318e+00 -8.91697407e-01 6.17292047e-01 -3.56034398e-01
2.29896784e-01 -4.33685184e-01 -6.57730043e-01 3.95249546e-01
1.64461076e-01 3.55770960e-02 2.31401548e-01 -7.26953670e-02
-5.54228246e-01 -3.43253553e-01 -1.26458871e+00 7.10771859e-01
1.40737474e+00 -7.50039220e-01 -7.15002000e-01 -3.12406626e-02
3.69453758e-01 -5.53883672e-01 -2.72166938e-01 6.15764141e-01
4.86561626e-01 -1.85575676e+00 1.25439858e+00 -3.10895473e-01
5.63431442e-01 -1.60625473e-01 -2.31021158e-02 -1.31900239e+00
-4.82791901e-01 -5.04589021e-01 5.13596117e-01 9.71155643e-01
3.82263005e-01 -8.32208574e-01 9.67414618e-01 6.37923479e-01
-1.25921771e-01 -6.37520254e-02 -1.13464785e+00 -6.76591992e-01
-2.18286626e-02 -7.55100027e-02 3.76661628e-01 1.07607722e+00
-4.84709024e-01 2.02036962e-01 3.12664621e-02 -1.47903422e-02
9.15939808e-01 2.25945503e-01 1.08403349e+00 -1.23703992e+00
-3.55421185e-01 -3.56606960e-01 -9.59956229e-01 -1.03091252e+00
1.81730464e-01 -6.93407893e-01 -1.09368242e-01 -1.20794368e+00
3.29241902e-01 -5.43510377e-01 -7.08278641e-02 2.43660480e-01
1.33325905e-01 7.07810521e-01 3.81373107e-01 4.63919252e-01
-2.31239349e-01 2.25568831e-01 1.35829639e+00 -4.99133527e-01
2.77306531e-02 2.34145932e-02 -2.29118377e-01 8.11796844e-01
6.50957286e-01 -4.04119551e-01 -4.09803689e-01 -6.54754400e-01
3.56977105e-01 -1.73999012e-01 7.12588787e-01 -1.24920559e+00
4.78483170e-01 6.41589314e-02 4.16679054e-01 -3.45127851e-01
4.45971698e-01 -1.04058182e+00 5.07694602e-01 5.11979461e-01
-2.98234224e-01 3.00930161e-03 2.64839113e-01 2.18640238e-01
-5.78167737e-01 -5.93587123e-02 1.27407849e+00 -2.84709930e-01
-9.80072439e-01 3.90197784e-01 -1.63850382e-01 3.57148081e-01
7.91330636e-01 -9.65318024e-01 -4.71851081e-01 -3.12976092e-01
-4.34319943e-01 -1.10953994e-01 1.24459827e+00 5.42126954e-01
6.75656855e-01 -1.18200576e+00 -4.76641744e-01 6.14275217e-01
3.18084449e-01 2.64745265e-01 2.73285121e-01 5.69742024e-01
-8.53634059e-01 2.97054887e-01 -7.77625084e-01 -8.93296778e-01
-1.33676922e+00 4.59079206e-01 8.60080183e-01 1.81780025e-01
-6.95195317e-01 7.93890178e-01 9.96401191e-01 -4.36143041e-01
5.04965425e-01 2.15691313e-01 4.85447459e-02 -1.56538248e-01
2.06802890e-01 3.33521292e-02 2.15657398e-01 -7.64771044e-01
-2.11800486e-01 7.06530571e-01 -4.07791398e-02 -1.66197628e-01
1.07179618e+00 -1.01940744e-01 -9.37897563e-02 1.44919649e-01
1.20076573e+00 1.71087414e-01 -1.46374547e+00 -3.08478028e-01
-3.03766578e-01 -7.71383464e-01 1.51031122e-01 -5.11485815e-01
-1.52795947e+00 8.81217003e-01 9.14117455e-01 -1.56659961e-01
1.08326173e+00 -2.83477306e-01 5.36167502e-01 8.91250074e-02
6.98327601e-01 -7.94193745e-01 -1.88547924e-01 4.86608297e-01
8.52624297e-01 -1.40322232e+00 -1.67853996e-01 -5.64624310e-01
-2.77031839e-01 8.76780987e-01 7.97253370e-01 -2.10814834e-01
5.50927520e-01 3.29595268e-01 3.74303907e-01 -1.74079746e-01
-3.87173176e-01 -2.07248554e-01 3.22299182e-01 8.91349196e-01
1.92021832e-01 -4.01035547e-01 -6.87348023e-02 -4.24120694e-01
-3.29003632e-01 -2.84646988e-01 2.43007705e-01 8.17949951e-01
-1.35117933e-01 -1.05247235e+00 -3.84417117e-01 1.28954902e-01
-7.44840652e-02 -2.20477194e-01 -9.84032452e-01 8.22676003e-01
2.75024921e-01 7.81140506e-01 5.70251465e-01 -4.28131133e-01
4.84155893e-01 -4.80137587e-01 6.59874022e-01 -6.44195080e-01
-9.00280714e-01 -3.46858382e-01 -1.71568505e-02 -9.05681252e-01
-6.00267112e-01 -4.37642038e-01 -7.10952997e-01 -5.62850893e-01
5.68909803e-03 -2.87104964e-01 2.19732538e-01 7.25670636e-01
2.79517800e-01 -2.20923036e-01 6.65816069e-01 -8.89712155e-01
-1.69658527e-01 -4.57135439e-01 -4.01776701e-01 9.75664496e-01
2.44582295e-01 -6.03187799e-01 -6.46456540e-01 -9.53491870e-03] | [8.702064514160156, -2.303119421005249] |
a7521747-8b98-444a-af29-0f2e7ce68f72 | capsule-forensics-using-capsule-networks-to | 1810.11215 | null | http://arxiv.org/abs/1810.11215v1 | http://arxiv.org/pdf/1810.11215v1.pdf | Capsule-Forensics: Using Capsule Networks to Detect Forged Images and Videos | Recent advances in media generation techniques have made it easier for
attackers to create forged images and videos. State-of-the-art methods enable
the real-time creation of a forged version of a single video obtained from a
social network. Although numerous methods have been developed for detecting
forged images and videos, they are generally targeted at certain domains and
quickly become obsolete as new kinds of attacks appear. The method introduced
in this paper uses a capsule network to detect various kinds of spoofs, from
replay attacks using printed images or recorded videos to computer-generated
videos using deep convolutional neural networks. It extends the application of
capsule networks beyond their original intention to the solving of inverse
graphics problems. | ['Junichi Yamagishi', 'Isao Echizen', 'Huy H. Nguyen'] | 2018-10-26 | null | null | null | null | ['detect-forged-images-and-videos'] | ['computer-vision'] | [ 5.12201726e-01 -1.91950679e-01 2.57355332e-01 1.79706231e-01
-3.33106726e-01 -1.00412679e+00 6.53923035e-01 -2.93012589e-01
-2.32728466e-01 5.99588215e-01 -1.36363611e-01 -4.68056291e-01
1.40930280e-01 -1.01904726e+00 -9.80783165e-01 -4.58408326e-01
-4.96269375e-01 -3.54160070e-01 2.75828302e-01 -2.82181889e-01
4.53415960e-01 4.73714471e-01 -1.25929081e+00 6.00247800e-01
2.29410946e-01 9.12862003e-01 -1.98171243e-01 1.00389528e+00
3.77012849e-01 9.56593037e-01 -1.21157062e+00 -8.00673425e-01
4.57117319e-01 -4.06208038e-01 -3.80468428e-01 2.74675816e-01
3.79981667e-01 -8.44612002e-01 -1.00044477e+00 1.45636868e+00
3.31544310e-01 -5.02937555e-01 1.51675120e-01 -1.62089062e+00
-1.17067742e+00 5.34862995e-01 -2.89236665e-01 1.15930729e-01
9.95025516e-01 2.42179155e-01 1.62381396e-01 -5.77350616e-01
8.37992966e-01 1.13827574e+00 8.29236865e-01 3.96898568e-01
-8.07238460e-01 -9.05243337e-01 -7.73289800e-01 1.06866926e-01
-1.17170978e+00 -3.29814613e-01 9.87181544e-01 -3.00294936e-01
6.57767415e-01 1.78494126e-01 6.24311924e-01 1.72035563e+00
4.37529653e-01 5.21484137e-01 9.95164752e-01 -5.13973713e-01
-6.40734434e-02 5.30083738e-02 -6.87480628e-01 6.67638540e-01
6.22002125e-01 4.52557266e-01 -4.18019414e-01 -5.54692209e-01
1.00797725e+00 -3.45817022e-02 -6.42794490e-01 -1.30370408e-01
-1.25823128e+00 1.24151480e+00 2.94633172e-02 5.56510568e-01
-1.14816085e-01 2.48932943e-01 5.98556280e-01 8.37464392e-01
6.59913942e-02 8.02231908e-01 1.17768059e-02 1.30942613e-01
-9.97560084e-01 4.25177634e-01 1.09661925e+00 7.37130821e-01
1.60029173e-01 2.57812768e-01 4.81247425e-01 -1.34054134e-02
1.51869655e-01 1.84137732e-01 6.96359754e-01 -7.19038785e-01
3.78271610e-01 -1.51824933e-02 1.51477873e-01 -2.05349374e+00
-4.33298759e-02 -4.87080440e-02 -5.56363225e-01 3.85058045e-01
5.26737928e-01 -4.17029172e-01 -6.33684456e-01 1.05078351e+00
7.67896941e-04 3.62573296e-01 1.15354076e-01 7.74932802e-01
6.12058461e-01 5.93058765e-01 -5.91828823e-01 1.02198347e-01
9.76089597e-01 -7.73719132e-01 -7.71883011e-01 -1.40858427e-01
3.88344139e-01 -1.00115633e+00 1.66566759e-01 1.03052688e+00
-7.83593535e-01 -5.06504238e-01 -1.44521046e+00 4.16810095e-01
-7.60643244e-01 -4.70586628e-01 4.36584145e-01 1.36887407e+00
-1.05166018e+00 6.59746587e-01 -2.62264103e-01 2.78918725e-02
5.77860534e-01 4.23583269e-01 -6.89004838e-01 -1.26080632e-01
-1.49350846e+00 4.71835434e-01 5.85778534e-01 -9.92966071e-03
-1.23483455e+00 -1.79825827e-01 -6.84245825e-01 -2.49621868e-01
4.69431609e-01 -1.22726031e-01 9.73258197e-01 -1.59738743e+00
-1.41012812e+00 8.74373317e-01 8.90899420e-01 -7.39523828e-01
8.65849793e-01 -1.42707884e-01 -9.60689127e-01 6.71622813e-01
-2.85122931e-01 2.51579076e-01 1.80418622e+00 -9.49101746e-01
-2.89757907e-01 3.71057168e-02 2.20163956e-01 -5.58717072e-01
-5.18941462e-01 3.49086195e-01 6.64435923e-02 -1.11828852e+00
-2.08317414e-01 -9.33506191e-01 7.53299147e-02 2.32517421e-01
-4.90972072e-01 4.84505117e-01 1.63478315e+00 -9.44350064e-01
8.09340715e-01 -2.09579349e+00 -1.03733487e-01 1.33172482e-01
2.96513170e-01 8.01720142e-01 -1.42102107e-01 7.69538581e-01
-3.98352683e-01 3.81312937e-01 2.98382137e-02 8.55634958e-02
-2.90440202e-01 -1.68084756e-01 -6.40066445e-01 8.57795060e-01
2.03273352e-02 7.99016416e-01 -9.51445699e-01 -1.32799834e-01
3.56235832e-01 6.44016027e-01 -4.02006418e-01 1.15045100e-01
-4.86619361e-02 4.05768961e-01 -2.69628972e-01 5.82995832e-01
9.00752187e-01 -3.03498302e-02 2.32955351e-01 9.37127098e-02
2.37888172e-01 -9.66179818e-02 -1.03204215e+00 1.29878080e+00
1.22898117e-01 1.12867248e+00 1.08169466e-01 -9.42980945e-01
7.64343798e-01 6.20457470e-01 3.61801416e-01 -2.12717608e-01
4.24876750e-01 2.47048393e-01 -4.08821255e-01 -9.06482220e-01
6.64807260e-01 4.06400487e-02 -2.87316799e-01 4.80929047e-01
2.10981891e-02 -9.22066048e-02 -8.93360898e-02 1.17229939e-01
1.36236596e+00 9.67464820e-02 2.02708200e-01 3.97072077e-01
4.40364748e-01 -2.53813326e-01 -8.72325897e-02 1.04428351e+00
-2.06500322e-01 5.83885431e-01 3.35550725e-01 -6.93100810e-01
-1.31076491e+00 -7.52614081e-01 2.57874787e-01 5.03200531e-01
1.40524760e-01 -4.25534040e-01 -9.88219559e-01 -7.58380115e-01
2.93542761e-02 2.01184675e-01 -3.32097262e-01 -2.35143334e-01
-6.13542855e-01 -4.54276562e-01 1.22685361e+00 -8.00904930e-02
8.12978446e-01 -1.11311996e+00 -8.21844399e-01 3.97175729e-01
-2.22779393e-01 -1.59101653e+00 -1.53283566e-01 -5.31295538e-01
-5.30345201e-01 -1.37772214e+00 -7.47897625e-01 -8.63163412e-01
5.27391613e-01 5.24743855e-01 6.69635832e-01 4.78427321e-01
-4.36801940e-01 3.31211388e-01 -5.68720400e-01 -2.05079094e-01
-1.03202462e+00 -3.99868131e-01 1.81854889e-01 2.56182641e-01
9.22666639e-02 -4.81023073e-01 -2.83283144e-01 3.19837958e-01
-1.58688056e+00 -3.41941297e-01 1.02097780e-01 6.56335652e-01
-4.77404475e-01 6.24694943e-01 2.65038788e-01 -6.04936242e-01
8.44858050e-01 -5.09409964e-01 -8.13606441e-01 7.63192698e-02
-1.09496363e-01 -3.76613945e-01 9.19461370e-01 -9.64067996e-01
-5.47555625e-01 -1.68801725e-01 5.71185797e-02 -7.57230222e-01
-2.95992583e-01 4.28745180e-01 2.15435445e-01 -8.60602558e-01
8.00527692e-01 3.50441128e-01 2.22102981e-02 -8.18020180e-02
2.69515157e-01 8.20571482e-01 8.76761615e-01 1.86639279e-02
1.27392077e+00 6.54548287e-01 -1.91656619e-01 -1.06678021e+00
-7.40053086e-03 6.93083629e-02 -1.64160609e-01 -5.01289129e-01
6.32284045e-01 -7.25390494e-01 -6.98760927e-01 1.25950170e+00
-1.41659212e+00 2.20127150e-01 2.98424453e-01 1.39770761e-01
-3.78649741e-01 1.30458331e+00 -8.33719611e-01 -5.14463961e-01
-5.08316979e-02 -1.13297737e+00 7.51148999e-01 -2.00641528e-02
-7.90679548e-03 -9.37785685e-01 -1.61979169e-01 3.55905265e-01
4.99945700e-01 7.59300351e-01 2.15428263e-01 -5.43436289e-01
-8.13694477e-01 -1.07041609e+00 -1.82897821e-02 4.35614884e-01
2.74374306e-01 1.25978857e-01 -9.05512869e-01 -5.97098529e-01
6.34249926e-01 -3.56598735e-01 4.43315715e-01 1.73036512e-02
1.08009374e+00 -9.11332011e-01 -3.36506099e-01 8.18365633e-01
1.48729300e+00 5.58022738e-01 1.24213886e+00 6.03650689e-01
3.83559912e-01 1.85861185e-01 -1.29971936e-01 3.89727771e-01
-3.80495161e-01 5.65202296e-01 8.74577045e-01 1.92496121e-01
1.86593100e-01 -3.14967573e-01 6.44081056e-01 3.87302697e-01
1.69876039e-01 -7.72120535e-01 -5.11650145e-01 2.43201002e-01
-1.34057641e+00 -1.37167048e+00 -1.13735450e-02 2.20699453e+00
3.55286390e-01 3.15776438e-01 -3.69453914e-02 2.88693845e-01
1.04575622e+00 2.93862909e-01 -9.92180482e-02 -2.53360450e-01
-2.94930875e-01 2.40477035e-03 7.46655285e-01 6.73650578e-02
-1.53341651e+00 7.48507261e-01 7.24925232e+00 5.92330217e-01
-1.25398219e+00 6.76029325e-02 4.10070747e-01 3.11570048e-01
1.88242406e-01 -9.48166773e-02 -1.86469272e-01 7.64868855e-01
9.55142081e-01 7.35917091e-02 6.52662158e-01 9.80821490e-01
-8.55326504e-02 -5.05302101e-02 -6.81423664e-01 1.18312669e+00
5.79813361e-01 -1.71823478e+00 -8.32693577e-02 2.93606877e-01
6.53343856e-01 -3.00719291e-01 2.73358017e-01 -3.10555965e-01
3.21386784e-01 -1.12530053e+00 7.86498368e-01 9.48974490e-02
7.49430239e-01 -5.24331868e-01 7.02530146e-01 2.04658568e-01
-6.77244425e-01 -2.47547731e-01 -2.92826116e-01 -2.15781748e-01
3.25825632e-01 3.11153919e-01 -9.68485236e-01 3.74885648e-01
5.96614778e-01 3.80613565e-01 -5.19707739e-01 1.11943614e+00
-1.69045016e-01 4.11140025e-01 -3.06253552e-01 7.46861771e-02
4.49489683e-01 1.59950390e-01 8.01530004e-01 1.04346442e+00
6.47455692e-01 -4.22170132e-01 -1.17541112e-01 6.21697664e-01
-1.11729763e-01 -4.38192010e-01 -1.29884160e+00 -6.00760102e-01
1.03536159e-01 9.36029911e-01 -9.18665409e-01 -1.43573031e-01
-4.73072648e-01 1.46339262e+00 -2.45831609e-01 1.90621197e-01
-1.16873443e+00 -6.97818696e-01 3.19228381e-01 1.39153987e-01
7.51734555e-01 -5.48360467e-01 5.43532610e-01 -1.37820864e+00
-1.54665977e-01 -1.37128782e+00 2.65850008e-01 -9.83282983e-01
-1.19367218e+00 6.28237188e-01 -2.25089967e-01 -1.32599699e+00
-4.28164214e-01 -8.40077043e-01 -3.71197134e-01 2.07964242e-01
-1.01275790e+00 -8.99521589e-01 -1.85183406e-01 8.18981349e-01
3.85399222e-01 -6.29643738e-01 8.06011677e-01 3.07072490e-01
-6.17086999e-02 4.07851338e-01 1.96515948e-01 7.41423011e-01
5.92259824e-01 -4.08782035e-01 6.92217827e-01 1.20697141e+00
2.03141347e-01 3.57469559e-01 9.91603017e-01 -8.97078812e-01
-1.68043387e+00 -5.87326884e-01 4.08149093e-01 -8.31369460e-02
8.28384757e-01 -3.12254220e-01 -5.69138527e-01 6.97376549e-01
5.41450620e-01 -3.61932814e-02 3.95920753e-01 -1.09177542e+00
-7.16680765e-01 3.99102032e-01 -1.46459532e+00 4.99127299e-01
6.24004126e-01 -8.53780687e-01 -3.93319696e-01 5.19514501e-01
7.27857411e-01 -4.13317680e-01 -5.75377285e-01 -2.07202241e-01
6.95726514e-01 -1.29756367e+00 1.19190109e+00 -5.07105410e-01
6.68471754e-01 -2.18045101e-01 -8.54115337e-02 -9.55361187e-01
6.74614757e-02 -1.48507285e+00 -2.63433546e-01 4.47815359e-01
-1.39699340e-01 -6.22694373e-01 8.11434388e-01 1.41581789e-01
4.03222591e-01 2.08037362e-01 -1.03782141e+00 -8.08974266e-01
-3.50404710e-01 -4.83547240e-01 6.69768751e-01 1.32661486e+00
8.18177313e-02 -3.81141335e-01 -1.16720891e+00 3.53043914e-01
8.32537711e-01 -3.90546799e-01 8.09672415e-01 -8.45001459e-01
-4.24987227e-01 -1.13404453e-01 -1.07763815e+00 -6.36539817e-01
-2.10441664e-01 -4.07087624e-01 -3.65541607e-01 -6.10408723e-01
-4.70878959e-01 3.11310291e-01 1.82473525e-01 9.16469917e-02
5.22632599e-01 7.10169375e-01 4.23805058e-01 2.10570738e-01
-1.45012289e-01 -1.44206733e-01 8.67555916e-01 -2.46148318e-01
1.65798485e-01 -5.20386808e-02 -3.60640019e-01 6.89988077e-01
7.66728699e-01 -8.33904922e-01 -1.58245742e-01 -3.15499187e-01
6.68547571e-01 3.16811383e-01 9.11191344e-01 -1.45466387e+00
1.65652215e-01 1.77291051e-01 4.77054566e-01 -1.36169508e-01
1.66650683e-01 -9.34081793e-01 5.41268170e-01 8.02445590e-01
-1.42424643e-01 2.19485253e-01 -3.52927782e-02 8.00200880e-01
-1.29628092e-01 -5.72408557e-01 6.14360631e-01 -7.31714368e-01
-5.71224332e-01 -1.41061604e-01 -7.52522945e-01 -4.21181679e-01
1.26296175e+00 -6.21793747e-01 -5.01636386e-01 -9.38834488e-01
-6.43152118e-01 -6.99741542e-01 7.16393471e-01 6.44273341e-01
1.09140837e+00 -1.11092758e+00 -4.71529305e-01 3.68088245e-01
-3.83883834e-01 -7.02137291e-01 8.53915140e-02 1.62579507e-01
-1.33908689e+00 3.52698267e-01 -5.37227154e-01 -1.40437528e-01
-1.31300390e+00 1.14440608e+00 3.63082200e-01 -3.91947180e-02
-6.76406264e-01 5.48047960e-01 -1.88647941e-01 -4.44115661e-02
-6.35876507e-03 1.50924578e-01 2.02655733e-01 -1.49323702e-01
1.04787624e+00 3.57393384e-01 -2.22745016e-02 -7.51582265e-01
-8.83637518e-02 8.64631459e-02 1.19306289e-01 6.56372234e-02
1.24833643e+00 -1.34865716e-02 -1.59668058e-01 -3.06700259e-01
1.48983455e+00 -6.19959179e-03 -1.05565429e+00 1.14729464e-01
-2.61982679e-01 -8.07628036e-01 2.93069798e-02 -4.05575186e-01
-1.29783940e+00 4.28684533e-01 6.26170695e-01 9.43794012e-01
9.05575037e-01 -4.82635915e-01 1.21028757e+00 3.72556388e-01
7.30567813e-01 -8.31123829e-01 6.27991378e-01 8.90802070e-02
7.58508265e-01 -9.89236951e-01 6.82018921e-02 -4.22776490e-01
-2.21155554e-01 1.68681729e+00 -5.29829375e-02 -3.29077721e-01
5.91776729e-01 4.76043493e-01 -1.33170679e-01 -1.41233578e-01
6.28188699e-02 7.36392558e-01 -2.57486314e-01 1.07196617e+00
-2.38954842e-01 -1.82693005e-01 2.72360761e-02 6.59691403e-03
-1.75480276e-01 4.87888634e-01 1.34106159e+00 1.30446231e+00
-1.85983792e-01 -9.06315327e-01 -1.13548946e+00 8.30311850e-02
-9.92042303e-01 1.09876402e-01 -4.58095998e-01 7.36375570e-01
1.66501403e-01 1.26503956e+00 -2.28077620e-01 -6.82172358e-01
-4.33024228e-01 -1.43915191e-01 4.35327411e-01 -4.72705588e-02
-6.72339678e-01 -1.06413335e-01 8.91269743e-02 -5.21886468e-01
-6.57078147e-01 -4.83545601e-01 -3.75713915e-01 -7.16054142e-01
-4.34650362e-01 -2.14654759e-01 7.83171356e-01 6.88301325e-01
2.14004010e-01 -2.16658041e-03 7.18810618e-01 -1.09585297e+00
-4.06793475e-01 -4.81126308e-01 -4.12528545e-01 4.42237705e-01
6.65390670e-01 -3.54189724e-01 -5.53539515e-01 4.05056447e-01] | [12.511260032653809, 1.1223548650741577] |
57dabad8-bbc3-483e-ada3-92aa82623fb9 | point-cloud-registration-of-non-rigid-objects | 2212.03856 | null | https://arxiv.org/abs/2212.03856v2 | https://arxiv.org/pdf/2212.03856v2.pdf | Point Cloud Registration of non-rigid objects in sparse 3D Scans with applications in Mixed Reality | Point Cloud Registration is the problem of aligning the corresponding points of two 3D point clouds referring to the same object. The challenges include dealing with noise and partial match of real-world 3D scans. For non-rigid objects, there is an additional challenge of accounting for deformations in the object shape that happen to the object in between the two 3D scans. In this project, we study the problem of non-rigid point cloud registration for use cases in the Augmented/Mixed Reality domain. We focus our attention on a special class of non-rigid deformations that happen in rigid objects with parts that move relative to one another about joints, for example, robots with hands and machines with hinges. We propose an efficient and robust point-cloud registration workflow for such objects and evaluate it on real-world data collected using Microsoft Hololens 2, a leading Mixed Reality device. | ['Manorama Jha'] | 2022-12-07 | null | null | null | null | ['point-cloud-registration', 'mixed-reality'] | ['computer-vision', 'computer-vision'] | [ 1.19667441e-01 4.70631942e-02 4.87232089e-01 -2.14437291e-01
-5.55280983e-01 -6.67003274e-01 6.08071566e-01 -5.56908026e-02
-1.99052140e-01 -2.96880417e-02 -2.80285209e-01 1.47019014e-01
-3.47054631e-01 -4.93515611e-01 -9.63937044e-01 -2.59716600e-01
8.63759890e-02 1.45926929e+00 4.98977274e-01 -4.42130178e-01
2.02224031e-01 1.26137257e+00 -1.46049356e+00 -8.34195390e-02
4.60047692e-01 5.25870025e-01 3.21272016e-01 4.74019945e-01
-1.39434993e-01 -3.19336891e-01 -1.66334286e-01 -4.74590182e-01
7.99556613e-01 4.01353031e-01 -8.29580784e-01 2.58757979e-01
9.16725755e-01 -9.94181260e-02 -4.15882319e-02 1.03778696e+00
3.62593949e-01 5.54364808e-02 4.99911368e-01 -1.50759804e+00
-1.47092178e-01 -1.29542351e-02 -8.82932842e-01 -3.05777848e-01
8.47329319e-01 -1.46691471e-01 2.74374813e-01 -1.01662040e+00
9.37825084e-01 1.47441733e+00 1.08934188e+00 4.46523219e-01
-1.19365644e+00 -2.23302603e-01 -1.87992573e-01 -1.96541041e-01
-1.47769678e+00 -3.56376529e-01 7.35999107e-01 -6.93563700e-01
9.36579943e-01 5.31288087e-01 5.62750041e-01 5.48720300e-01
2.41156757e-01 -1.72113888e-02 6.57911479e-01 -3.93252671e-01
5.04402034e-02 -3.02164018e-01 7.20671341e-02 2.05531642e-01
4.01825160e-01 -1.37449503e-02 -4.82658483e-02 -5.86862862e-01
1.04068410e+00 5.05445600e-01 -1.77360028e-01 -1.03289866e+00
-1.46877086e+00 9.19114649e-02 4.46917415e-01 2.45705590e-01
-5.51315427e-01 1.60513252e-01 -1.75426438e-01 1.58758163e-01
3.20920438e-01 1.63406283e-02 -8.52190197e-01 -1.91302046e-01
-5.09120822e-01 3.43303323e-01 6.71328366e-01 1.50171614e+00
9.46355045e-01 -6.46834195e-01 7.31144011e-01 5.19180834e-01
7.26933837e-01 4.20462012e-01 1.46624535e-01 -8.48212957e-01
6.39825344e-01 7.81412423e-01 3.27155769e-01 -9.43011999e-01
-5.13756514e-01 2.62569815e-01 -4.89544034e-01 5.31794965e-01
2.55783051e-01 4.49813455e-01 -1.08017170e+00 1.03732908e+00
9.12149310e-01 3.32702875e-01 -1.08243629e-01 9.22334909e-01
8.12489033e-01 4.31075171e-02 -4.51961100e-01 3.76156643e-02
1.51669347e+00 -3.48954409e-01 -5.72829366e-01 -1.69281170e-01
2.77387768e-01 -1.37642813e+00 7.56498039e-01 5.33690304e-02
-1.29492295e+00 -4.43905950e-01 -6.83977425e-01 -3.87125522e-01
-1.91577822e-01 -4.73883808e-01 7.83521906e-02 3.12570810e-01
-7.19872653e-01 7.20666111e-01 -1.16234481e+00 -5.72424531e-01
-5.79183064e-02 9.25091684e-01 -8.93936992e-01 -8.39406103e-02
-3.03146273e-01 1.15208149e+00 -8.00945796e-03 3.05596679e-01
3.24967667e-03 -9.54300106e-01 -6.06094360e-01 -4.66543645e-01
4.33019221e-01 -8.11511755e-01 1.23279727e+00 -5.30558527e-01
-1.46724629e+00 1.45819378e+00 -2.28149705e-02 2.57783860e-01
7.26193607e-01 -5.40441513e-01 -2.73289323e-01 -4.15083259e-01
5.01382574e-02 1.68332160e-01 5.56832850e-01 -1.55686247e+00
-1.31837487e-01 -1.04948866e+00 -1.40081048e-01 3.42248440e-01
6.81024075e-01 2.04934821e-01 -7.62238741e-01 -3.09301168e-01
1.07077193e+00 -1.44509971e+00 -3.39811772e-01 1.39997348e-01
-3.71366620e-01 6.26944238e-03 1.38286388e+00 -5.98813772e-01
-7.07120588e-03 -2.30214620e+00 2.42846027e-01 4.88901466e-01
-1.56405970e-01 -5.99437207e-02 6.28408268e-02 1.65294051e-01
-2.51241624e-01 -5.90538718e-02 -1.62053242e-01 -5.16072154e-01
-1.00535937e-01 5.45392871e-01 -2.17694953e-01 8.00595939e-01
-1.91462010e-01 6.15858376e-01 -7.18623281e-01 -3.33898038e-01
2.83748120e-01 7.15996027e-01 -2.30569348e-01 2.62482345e-01
7.50642717e-02 8.35961521e-01 -3.44192356e-01 7.87050247e-01
1.09973097e+00 2.90166825e-01 -1.55697301e-01 -4.59056050e-01
-1.98754728e-01 5.76596968e-02 -1.88872945e+00 2.08628106e+00
-1.98014870e-01 -1.11947209e-01 4.04547274e-01 -2.94327587e-01
1.01298642e+00 6.05382919e-01 8.82040560e-01 -6.87380880e-02
6.98724091e-02 4.46211964e-01 -1.72888875e-01 -3.58472407e-01
9.05018687e-01 -1.56367183e-01 9.04068351e-02 2.51151592e-01
-1.90421626e-01 -9.95114446e-01 -6.30379498e-01 -4.32076484e-01
9.01887000e-01 4.38554138e-01 3.15478593e-01 -1.42709598e-01
2.24509448e-01 7.74859488e-02 3.41133058e-01 1.43404230e-01
2.21814439e-01 1.16506660e+00 -2.93717980e-01 -4.44584012e-01
-1.19840169e+00 -1.15132368e+00 -1.61393777e-01 4.11651254e-01
4.94065046e-01 -1.24715097e-01 -2.78886318e-01 -1.28827959e-01
3.16785872e-01 1.79573998e-01 -1.33263528e-01 4.74201962e-02
-1.01443672e+00 -2.19780311e-01 1.98718235e-02 3.56573045e-01
1.16818876e-03 -5.26043594e-01 -8.20780873e-01 1.23261631e-01
9.05330852e-02 -1.45940971e+00 -6.14592373e-01 -1.19735353e-01
-1.58920372e+00 -1.22186553e+00 -6.19809210e-01 -7.14588702e-01
9.96849835e-01 4.67818797e-01 1.23430884e+00 1.05639562e-01
-2.79164225e-01 1.02779734e+00 -2.57534534e-01 -4.19692039e-01
-4.70458418e-01 -2.21058711e-01 3.64245713e-01 -2.24351048e-01
5.28475530e-02 -7.01443851e-01 -1.72632247e-01 1.00382209e+00
-9.49650466e-01 -2.37977847e-01 -3.95986345e-03 -2.68814480e-03
1.09259617e+00 -2.21172228e-01 -6.02130055e-01 -4.19616580e-01
1.67585149e-01 -3.57322633e-01 -7.54569769e-01 1.85397029e-01
1.21840298e-01 -2.53003567e-01 -2.89944589e-01 -7.80239999e-01
-7.54452884e-01 7.91146755e-01 1.33491412e-01 -8.03471088e-01
-3.40661347e-01 2.71269798e-01 -4.21520114e-01 -3.66524905e-01
4.63877678e-01 -5.54466665e-01 2.02863455e-01 -9.49230552e-01
3.31526011e-01 5.34269571e-01 8.62137616e-01 -6.90762103e-01
1.16394317e+00 8.55159283e-01 4.54845309e-01 -7.46918738e-01
-8.59615281e-02 -6.91872358e-01 -1.42973006e+00 7.11919880e-03
6.02352083e-01 -6.68907702e-01 -7.95152366e-01 4.18385208e-01
-1.54772508e+00 -8.41270611e-02 -7.41211176e-01 6.69994652e-01
-1.01415873e+00 5.27712405e-01 -1.91216066e-01 -5.61411381e-01
-1.53177395e-01 -1.39655983e+00 1.43097091e+00 1.57036930e-02
-1.85765058e-01 -6.35065079e-01 4.59093362e-01 1.58886492e-01
-2.05772631e-02 5.62641382e-01 4.44430262e-01 -4.99704510e-01
-5.76495051e-01 -5.22555292e-01 4.17709321e-01 -3.16745341e-01
4.08448607e-01 4.72996831e-01 -5.96382201e-01 -2.57484168e-01
2.25040153e-01 3.20909560e-01 -3.68486732e-01 2.52302051e-01
5.27263224e-01 8.96184240e-03 -5.38855016e-01 6.22937262e-01
1.31689799e+00 -4.35345899e-03 8.01833272e-01 3.74977618e-01
9.40463006e-01 8.09270620e-01 9.40903068e-01 2.05218513e-02
2.41460696e-01 1.44669867e+00 9.31423426e-01 1.24618962e-01
1.91923395e-01 1.66043758e-01 -4.77798190e-03 8.66844416e-01
-8.06181669e-01 3.54490966e-01 -1.34013093e+00 2.98476726e-01
-1.89902794e+00 -5.91478407e-01 -1.03429639e+00 2.72308421e+00
4.96719748e-01 -2.60525525e-01 -7.13272989e-02 6.98826462e-02
1.03981733e+00 -6.34718835e-01 -2.39176422e-01 -2.90275048e-02
1.69812620e-01 2.11717993e-01 4.52019691e-01 5.00346601e-01
-7.56639898e-01 5.79214990e-01 5.93799019e+00 -9.95655954e-02
-1.02493215e+00 3.09316427e-01 -5.67249596e-01 5.16650826e-02
1.01111017e-01 2.73739338e-01 -4.94174659e-01 7.53766373e-02
7.15330899e-01 8.87171254e-02 1.93513423e-01 7.77665198e-01
-1.42778531e-01 9.95697603e-02 -1.44460094e+00 1.14039397e+00
-2.04999194e-01 -8.92202556e-01 -3.50526273e-01 4.46971506e-01
5.91043830e-01 4.89614755e-01 -2.47095853e-01 -3.55008274e-01
1.76680148e-01 -6.52800739e-01 1.01646066e+00 9.14719641e-01
5.99506080e-01 -2.62728870e-01 5.40351152e-01 4.38142508e-01
-1.16929579e+00 5.15144944e-01 -4.86903489e-01 2.77096152e-01
5.11240065e-01 5.14285564e-01 -9.11874294e-01 8.52418184e-01
9.12241876e-01 2.37934634e-01 -2.31137782e-01 1.41074109e+00
3.25395048e-01 -3.53575855e-01 -8.17375422e-01 8.41853976e-01
-4.72095460e-01 -5.31943500e-01 9.60872293e-01 5.00100434e-01
7.03001201e-01 3.96313071e-01 1.21359386e-01 7.18424082e-01
1.60909370e-01 -6.26330003e-02 -7.87385464e-01 7.22144663e-01
3.01401079e-01 1.17942798e+00 -8.13280702e-01 3.02220080e-02
-2.33213872e-01 9.79894400e-01 -2.53238201e-01 -3.66851129e-02
-6.14054620e-01 2.17628866e-01 7.25730717e-01 5.16142368e-01
-7.78262764e-02 -8.52897227e-01 -1.51773438e-01 -9.56656635e-01
3.89184624e-01 -4.34107751e-01 1.24914885e-01 -1.20799172e+00
-1.30069339e+00 4.59264189e-01 3.30372423e-01 -1.72396433e+00
-5.23139536e-02 -4.34468120e-01 -3.82054538e-01 9.90377963e-01
-9.11986351e-01 -1.39329791e+00 -6.39199495e-01 9.03663456e-01
3.09578001e-01 2.28624552e-01 7.91320980e-01 2.99767405e-01
2.14615211e-01 -1.15288541e-01 -3.05196829e-02 -2.18967766e-01
6.16305232e-01 -9.69307005e-01 8.39012504e-01 4.20600891e-01
8.53517205e-02 8.01244915e-01 9.29856896e-01 -9.95113730e-01
-2.04583430e+00 -7.38604069e-01 4.46918517e-01 -9.62047338e-01
3.27210665e-01 -2.46179178e-01 -1.18869650e+00 1.18314064e+00
-4.71384078e-01 6.62134588e-01 2.66659826e-01 -1.47546276e-01
-1.62743852e-01 2.66359597e-01 -1.69373310e+00 1.91284180e-01
1.18885994e+00 -4.58492368e-01 -1.14535367e+00 4.48738277e-01
5.49472928e-01 -1.49103200e+00 -1.26281023e+00 6.53601944e-01
5.94720602e-01 -5.38814366e-01 1.51844633e+00 -4.47518855e-01
-3.26918215e-01 -7.04260349e-01 -3.89295697e-01 -1.19049859e+00
-2.60520652e-02 -7.77803421e-01 2.50660758e-02 1.01153910e+00
-3.21295440e-01 -6.62794352e-01 7.57826209e-01 1.07604992e+00
-3.77814174e-01 9.24021378e-02 -1.44700277e+00 -9.97353017e-01
-2.17220217e-01 -3.90517682e-01 9.28962648e-01 1.13142490e+00
-4.28054422e-01 -1.94810808e-01 1.10005133e-01 8.84397686e-01
5.64023972e-01 2.06877347e-02 1.39567339e+00 -2.00064230e+00
7.38628060e-02 9.26551595e-02 -9.86973464e-01 -5.79827785e-01
-2.25932896e-02 -5.69341838e-01 2.36417964e-01 -1.27284801e+00
-2.33760446e-01 -8.42758000e-01 6.69908762e-01 3.11571628e-01
3.47131550e-01 2.98736870e-01 2.99287885e-01 8.64492595e-01
3.17375325e-02 2.35471442e-01 1.13833880e+00 1.04160845e-01
-4.80199575e-01 4.55347300e-01 1.86622962e-01 1.08998811e+00
3.90259206e-01 -7.32007921e-01 1.85826778e-01 -6.91596806e-01
2.28928387e-01 1.35891929e-01 6.01535439e-01 -9.98961687e-01
2.85831034e-01 -3.88817996e-01 -1.79675192e-01 -1.00190902e+00
6.98474228e-01 -1.86715102e+00 1.42058754e+00 2.86706865e-01
4.86929566e-01 6.00597143e-01 1.07200451e-01 2.48879716e-01
8.09186250e-02 -5.11360168e-01 5.43289542e-01 -3.04458827e-01
-1.91201180e-01 5.42369902e-01 4.15852875e-01 -4.99235600e-01
1.30819750e+00 -6.92891121e-01 -1.37357011e-01 1.12622879e-01
-1.06897581e+00 -1.88007176e-01 1.25290084e+00 6.79683805e-01
7.02772796e-01 -1.42283988e+00 -4.97302622e-01 1.65700078e-01
1.58658117e-01 9.44223464e-01 -2.98732053e-02 9.87721980e-01
-6.60691738e-01 -7.56011382e-02 -3.23265553e-01 -1.20889997e+00
-1.75537825e+00 5.33794880e-01 5.79700410e-01 3.46316397e-01
-7.24634469e-01 3.24202001e-01 -2.63959140e-01 -1.15698349e+00
-1.92697853e-01 -7.64250100e-01 1.84067369e-01 -2.47594252e-01
6.62605539e-02 7.30767071e-01 8.44793618e-01 -1.36820900e+00
-5.90053320e-01 1.38372386e+00 3.36754233e-01 -1.10640258e-01
1.46222472e+00 -1.71556007e-02 -4.82014775e-01 5.87801456e-01
1.02886212e+00 2.27240875e-01 -9.21087384e-01 -4.05932188e-01
7.91413262e-02 -7.64359713e-01 -3.17650497e-01 -2.27262899e-01
-9.32616174e-01 4.93816882e-01 9.04268742e-01 1.76425830e-01
4.65658128e-01 4.35071707e-01 3.63737971e-01 2.81200677e-01
1.02035046e+00 -6.27541721e-01 -3.83309007e-01 6.71673179e-01
1.43582618e+00 -7.11147308e-01 3.96027565e-01 -8.76220524e-01
-1.37228683e-01 1.21144712e+00 2.96964943e-01 -2.14191154e-01
7.18446076e-01 3.76617014e-01 1.76418182e-02 -5.43663144e-01
1.96086019e-01 1.28129959e-01 5.20164669e-01 8.62228632e-01
1.73463196e-01 4.15382013e-02 3.53873186e-02 -2.58444808e-03
-3.85491520e-01 5.68337590e-02 4.92773861e-01 1.63863385e+00
-4.66527715e-02 -1.25852597e+00 -1.36132193e+00 1.12092480e-01
-1.03075348e-01 6.00599289e-01 -3.63331050e-01 8.29082727e-01
4.23254855e-02 4.40343440e-01 6.41337633e-01 -2.23692432e-01
1.06821728e+00 -1.40632957e-01 9.00208950e-01 -7.81566322e-01
-7.72547781e-01 1.70187771e-01 -4.20644879e-01 -7.88229465e-01
-8.78319919e-01 -1.08380663e+00 -1.53386796e+00 -1.39019996e-01
-5.73655486e-01 -2.34061167e-01 1.28310275e+00 8.07534695e-01
5.72865009e-01 -9.86601189e-02 2.97542661e-01 -1.66551292e+00
-5.86459458e-01 -7.65420318e-01 -6.01125240e-01 7.46700406e-01
2.38850370e-01 -8.34561467e-01 -9.97944325e-02 2.65802324e-01] | [7.779484272003174, -2.8101093769073486] |
4b473ff7-9c35-44b7-85a2-d858a28528b4 | regret-analysis-of-the-stochastic-direct | 2210.05222 | null | https://arxiv.org/abs/2210.05222v1 | https://arxiv.org/pdf/2210.05222v1.pdf | Regret Analysis of the Stochastic Direct Search Method for Blind Resource Allocation | Motivated by programmatic advertising optimization, we consider the task of sequentially allocating budget across a set of resources. At every time step, a feasible allocation is chosen and only a corresponding random return is observed. The goal is to maximize the cumulative expected sum of returns. This is a realistic model for budget allocation across subdivisions of marketing campaigns, when the objective is to maximize the number of conversions. We study direct search (aka pattern search) methods for linearly constrained and derivative-free optimization in the presence of noise. Those algorithms are easy to implement and particularly suited to constrained optimization. They have not yet been analyzed from the perspective of cumulative regret. We provide a regret upper-bound of the order of T 2/3 in the general case. Our mathematical analysis also establishes, as a by-product, time-independent regret bounds in the deterministic, unconstrained case. We also propose an improved version of the method relying on sequential tests to accelerate the identification of descent directions. | ['Aurélien Garivier', 'Olivier Cappe', 'Juliette Achddou'] | 2022-10-11 | null | null | null | null | ['marketing'] | ['miscellaneous'] | [ 3.29779655e-01 1.85484603e-01 -6.39249504e-01 -4.21099007e-01
-8.01350117e-01 -8.12060714e-01 2.01524884e-01 1.70985371e-01
-7.38510132e-01 8.09388340e-01 -3.70991193e-02 -5.50410867e-01
-6.75611556e-01 -7.86843359e-01 -8.35319817e-01 -6.08099341e-01
-1.26009822e-01 7.81389236e-01 -2.72100240e-01 -9.10809636e-02
2.19921842e-01 6.59889758e-01 -1.22542703e+00 -1.67507648e-01
5.59197426e-01 1.43339610e+00 -5.26068658e-02 5.75641930e-01
-1.24005750e-01 2.15062648e-01 -2.89448738e-01 -6.44684374e-01
6.84684038e-01 -2.98404545e-01 -5.37197769e-01 5.35935342e-01
1.08007470e-03 -2.27621615e-01 1.28502220e-01 1.06942558e+00
4.34623420e-01 1.21047720e-01 3.46067399e-01 -9.49869335e-01
-1.29737481e-01 6.12223625e-01 -6.05889618e-01 1.50843546e-01
4.51703854e-02 -1.88160360e-01 1.37751973e+00 -3.23817849e-01
6.94033384e-01 1.00078201e+00 1.48336887e-01 4.02436644e-01
-1.66772783e+00 -2.17961073e-01 6.04967415e-01 -9.73625407e-02
-8.77425849e-01 -4.39521223e-01 5.44052482e-01 -1.78165346e-01
2.85184145e-01 7.23771691e-01 6.35236502e-01 6.44457936e-01
-3.30309495e-02 8.40337038e-01 1.13389087e+00 -6.59800649e-01
6.61478221e-01 3.67705226e-01 1.18648954e-01 5.41581571e-01
4.51509833e-01 3.91379803e-01 -3.63023251e-01 -3.77212048e-01
4.96590137e-01 -1.15943015e-01 -2.16284376e-02 -6.96530044e-01
-6.33041322e-01 1.16574621e+00 1.86925948e-01 -4.86175204e-03
-5.46892405e-01 1.74833417e-01 3.22203636e-02 5.33046544e-01
5.23770571e-01 3.71644914e-01 -3.76984179e-01 1.18191270e-02
-9.23958778e-01 6.15293503e-01 1.18961811e+00 8.58690500e-01
3.48994285e-01 -2.27354139e-01 -3.70043516e-01 6.03133559e-01
7.66015798e-02 6.10338748e-01 -4.60055061e-02 -9.69897509e-01
7.37857640e-01 1.87210679e-01 8.53512943e-01 -6.87801182e-01
-3.07640016e-01 -9.80120659e-01 -3.42428386e-01 2.35961989e-01
7.58458972e-01 -3.82942289e-01 -5.33628583e-01 1.76937020e+00
3.52346092e-01 -5.42137623e-01 -3.18605185e-01 1.06418049e+00
-4.12401527e-01 4.32324290e-01 -2.79935062e-01 -8.54664683e-01
1.18398988e+00 -6.47322655e-01 -5.57488978e-01 -3.40259552e-01
2.33422026e-01 -7.78862000e-01 8.08777213e-01 5.42724133e-01
-1.26061153e+00 1.77011698e-01 -8.10967028e-01 5.27634680e-01
-5.73078692e-02 9.87098217e-02 9.44757640e-01 1.04363942e+00
-5.51219523e-01 6.27150118e-01 -5.88973999e-01 -1.49074802e-02
2.36458167e-01 4.47916061e-01 2.68379211e-01 1.18869148e-01
-6.44783020e-01 6.04524374e-01 -2.83339292e-01 5.25255203e-01
-8.03274035e-01 -4.68662262e-01 -3.08835626e-01 3.32456142e-01
8.94517243e-01 -5.60654640e-01 1.60230505e+00 -1.06624281e+00
-1.55844545e+00 7.25975275e-01 -2.36837402e-01 -5.83723724e-01
1.06818950e+00 8.32281038e-02 1.84614714e-02 -3.98886412e-01
1.36406854e-01 -1.20296948e-01 6.64468646e-01 -7.96074927e-01
-6.84685528e-01 -7.01442242e-01 4.01792318e-01 2.20965803e-01
-1.57482713e-01 8.54179561e-02 -2.13165790e-01 -3.81140053e-01
1.05570167e-01 -1.12311625e+00 -8.91334772e-01 -4.75798510e-02
-4.38965470e-01 3.13882023e-01 -2.61986405e-01 -3.40186566e-01
1.17982602e+00 -1.86537158e+00 1.93885624e-01 6.11466885e-01
-3.12172979e-01 -3.57496977e-01 -8.18564184e-03 3.21896166e-01
2.35033229e-01 8.86624828e-02 -2.32395474e-02 -2.52554327e-01
4.63346243e-01 8.32352117e-02 -3.50081712e-01 5.32917142e-01
-1.90791354e-01 5.14194131e-01 -6.38291359e-01 -6.88112825e-02
-2.19026700e-01 -3.52382034e-01 -7.24773824e-01 -1.47800641e-02
-5.08451402e-01 -2.37663202e-02 -7.93809593e-01 6.27686918e-01
6.33834660e-01 -2.28347927e-01 6.64534390e-01 2.08189413e-01
-4.49386209e-01 6.89590350e-02 -1.55189466e+00 1.19040799e+00
-7.17420936e-01 2.79966056e-01 4.67071295e-01 -1.19596934e+00
4.45471495e-01 -2.09628209e-01 3.81717116e-01 -7.67138422e-01
3.37564498e-01 3.14381868e-01 -3.80817801e-02 -1.53366327e-01
4.64685738e-01 -2.96469212e-01 -4.06935722e-01 6.01372540e-01
-4.57286477e-01 4.30037290e-01 3.24737519e-01 -1.24860831e-01
1.01726341e+00 -2.53570378e-01 3.01932454e-01 -3.68152797e-01
6.83474615e-02 1.24815740e-01 5.47960043e-01 1.04165637e+00
3.56630534e-02 6.12382889e-02 9.60674167e-01 -2.80832201e-01
-9.47958589e-01 -9.16268528e-01 -1.73237115e-01 1.16136253e+00
5.84086701e-02 2.11877033e-01 -4.24221545e-01 -4.53738749e-01
4.52885628e-01 6.90828741e-01 -7.27632523e-01 3.60340238e-01
-3.85003626e-01 -1.11688161e+00 -3.71229082e-01 7.48783574e-02
5.88983558e-02 -3.71343017e-01 -6.52109861e-01 4.64568406e-01
1.87014267e-01 -8.93413186e-01 -5.40918291e-01 4.22918081e-01
-9.43808258e-01 -9.04433966e-01 -7.24453092e-01 -2.65711159e-01
7.29876220e-01 8.52453411e-02 7.41987884e-01 -5.52138329e-01
-3.44401836e-01 4.13179576e-01 1.04980558e-01 -3.91036421e-01
2.91140098e-03 -7.83780888e-02 -3.79311591e-02 3.84162992e-01
1.41764656e-01 -2.54956186e-01 -6.72372162e-01 5.36521018e-01
-5.53260088e-01 -4.21898454e-01 6.03658319e-01 7.85707414e-01
8.32121730e-01 7.32283620e-03 3.65880817e-01 -9.58206773e-01
8.66001308e-01 -4.60667193e-01 -1.50829434e+00 4.40260082e-01
-7.81788707e-01 3.86439085e-01 4.36865062e-01 -6.45432353e-01
-1.06634712e+00 2.56182700e-01 5.09671032e-01 8.52763355e-02
4.07360047e-01 5.19934893e-01 -1.92474365e-01 -3.31278741e-01
4.51758295e-01 1.10353820e-01 4.26927768e-03 -7.03830898e-01
5.22927940e-01 5.14101088e-01 1.61937159e-02 -7.40597606e-01
3.49970013e-01 4.31278050e-01 4.32379574e-01 -5.48723102e-01
-9.31124151e-01 -2.44084373e-01 -3.65235773e-03 -3.02057922e-01
1.93460345e-01 -4.77302700e-01 -1.18578517e+00 -2.74031311e-01
-8.13088179e-01 -1.69012651e-01 -4.76072282e-01 5.04049122e-01
-8.25326025e-01 1.14375971e-01 -1.02056697e-01 -1.38181341e+00
4.56296764e-02 -9.68932927e-01 5.87960124e-01 1.84264332e-01
1.33249998e-01 -6.68150306e-01 -9.28722546e-02 3.44417930e-01
4.98937517e-01 2.36594424e-01 6.68348193e-01 -4.79323775e-01
-8.10678482e-01 -4.29733574e-01 -5.78588098e-02 8.21743384e-02
-2.27254778e-01 -5.78237057e-01 -3.29995364e-01 -1.92592829e-01
1.45476013e-01 2.33245008e-02 6.03833020e-01 7.71632969e-01
1.09645820e+00 -7.81531572e-01 -3.29338729e-01 3.87750655e-01
1.46440184e+00 2.28650644e-01 1.15767509e-01 5.36228478e-01
-2.34902993e-01 5.67004442e-01 7.89314210e-01 8.66322815e-01
-1.35922030e-01 1.08558905e+00 5.03695905e-01 2.16469973e-01
4.96527314e-01 -1.01558852e-03 1.32799670e-01 -1.78373933e-01
5.54371849e-02 -2.70684332e-01 -2.24524155e-01 6.05258882e-01
-1.96295238e+00 -8.30365837e-01 9.49808583e-02 2.81254506e+00
7.16294706e-01 4.37471032e-01 3.75228643e-01 -1.97217226e-01
7.92829573e-01 -8.25126544e-02 -7.96983659e-01 -8.22358966e-01
4.70505059e-02 3.54960382e-01 1.38500679e+00 7.57629037e-01
-8.09746504e-01 4.60291386e-01 6.66862535e+00 8.32088411e-01
-7.27809787e-01 2.30815217e-01 8.45608056e-01 -9.47504044e-01
-3.78763646e-01 1.59325451e-01 -1.00582027e+00 6.09303474e-01
8.50991130e-01 -4.33053911e-01 7.80978560e-01 1.06045735e+00
4.00672406e-01 -3.22175175e-01 -1.07720840e+00 5.45686841e-01
-4.62859184e-01 -1.28382099e+00 -5.37405252e-01 5.91739655e-01
7.58406341e-01 -4.29197073e-01 2.76893765e-01 7.71994963e-02
4.30073023e-01 -6.55114830e-01 8.21646154e-01 3.82728189e-01
4.41020012e-01 -9.08129275e-01 3.62860471e-01 3.21300238e-01
-7.00101137e-01 -5.13161182e-01 -3.41065943e-01 -2.26529434e-01
4.52622652e-01 8.72779071e-01 -4.83096421e-01 3.90600950e-01
1.52550220e-01 -2.74353802e-01 1.68752953e-01 1.31234121e+00
-1.67889059e-01 4.31823730e-01 -7.93822169e-01 -5.94044685e-01
4.79169369e-01 -5.02156794e-01 6.65210247e-01 9.29388046e-01
3.20553303e-01 3.66628282e-02 5.96450455e-02 7.25630760e-01
-7.70592913e-02 3.66722792e-01 -1.46916658e-01 7.94333126e-03
3.17495465e-01 1.06643212e+00 -7.05304682e-01 1.05346605e-01
-2.21960053e-01 7.53322959e-01 2.19656780e-01 2.69697934e-01
-5.99664927e-01 -2.03259394e-01 6.21550798e-01 1.61738411e-01
6.22188032e-01 -9.18611288e-02 -4.38627988e-01 -8.58677864e-01
4.54855919e-01 -4.80875373e-01 5.17449558e-01 -5.42618223e-02
-1.05604696e+00 9.97738466e-02 4.43252996e-02 -8.25434804e-01
-2.56252468e-01 -6.34934723e-01 -2.83308268e-01 8.47827196e-01
-1.40309989e+00 -5.28375745e-01 2.50351787e-01 2.66729981e-01
4.32729781e-01 2.92126667e-02 4.75836545e-01 2.64558971e-01
-5.93568146e-01 6.30318403e-01 6.40291095e-01 -5.43124616e-01
1.05008848e-01 -1.12394309e+00 -1.84210956e-01 9.07691121e-01
-1.07070141e-01 4.28146899e-01 1.05444586e+00 -4.08604443e-01
-1.64301813e+00 -6.81293964e-01 8.89281392e-01 1.10885598e-01
9.58323300e-01 -4.89549130e-01 8.77866335e-03 5.06688118e-01
-3.77444476e-01 -2.65147865e-01 5.09120345e-01 5.56629300e-01
6.73061684e-02 -4.56659615e-01 -1.13782966e+00 5.19845128e-01
1.04041362e+00 6.99004233e-02 3.24030109e-02 7.35789418e-01
3.57029140e-01 -4.32844073e-01 -6.37869537e-01 -1.07138410e-01
7.72084653e-01 -7.13535726e-01 7.99873710e-01 -8.76060009e-01
1.20573312e-01 1.76086113e-01 -3.93764466e-01 -9.27662075e-01
-2.55360454e-01 -1.06942630e+00 3.00401337e-02 7.36154556e-01
8.12717378e-01 -6.67146444e-01 1.14715409e+00 9.01010990e-01
4.53824580e-01 -8.70337725e-01 -1.21083796e+00 -1.12074840e+00
-3.30250204e-01 -2.95879006e-01 4.15729403e-01 3.60002458e-01
3.12659238e-03 6.04814626e-02 -5.69799125e-01 1.23460419e-01
9.56626952e-01 7.19271779e-01 5.43416619e-01 -8.67925406e-01
-8.80651593e-01 -6.67898297e-01 -9.32053924e-02 -1.34106755e+00
-1.92382231e-01 -5.40244997e-01 -1.43120348e-01 -1.06884825e+00
1.52786866e-01 -7.27084816e-01 -1.49843290e-01 4.48370576e-02
3.57401073e-01 -2.64889836e-01 1.97339877e-01 -1.26473516e-01
-4.42022085e-01 1.31187052e-01 1.01490486e+00 -7.70752952e-02
-3.14480126e-01 9.72070038e-01 -1.07860231e+00 1.74972743e-01
6.65304959e-01 -5.30116022e-01 -2.85327941e-01 -1.48250788e-01
5.43482780e-01 4.98102427e-01 1.72773451e-01 -1.90596148e-01
7.31140450e-02 -6.84443057e-01 8.86076912e-02 -3.82933736e-01
3.54655653e-01 -9.85769033e-01 3.69181126e-01 4.92621899e-01
-6.57617033e-01 -1.61763400e-01 -1.54898107e-01 9.15626466e-01
2.65718490e-01 -8.05946887e-01 5.10480642e-01 -2.02828556e-01
-1.08352102e-01 1.94150776e-01 -2.50232697e-01 -1.95326567e-01
9.93237674e-01 8.55085030e-02 -5.86829893e-02 -5.05992830e-01
-1.14757609e+00 3.40060085e-01 6.73240051e-02 -3.62022221e-02
7.61923864e-02 -1.09584403e+00 -4.93053406e-01 -1.64900884e-01
-1.70452371e-01 -5.16360402e-01 1.14893168e-01 8.20472658e-01
-1.54868200e-01 6.72591150e-01 1.04200363e-01 -1.40999675e-01
-1.12234557e+00 5.47630847e-01 3.60663295e-01 -7.04485178e-01
-9.43884701e-02 7.20236421e-01 -5.22889718e-02 1.24864899e-01
4.52164710e-01 -1.46566600e-01 2.82068670e-01 1.04059428e-01
4.84372407e-01 4.15505290e-01 2.17665106e-01 1.97350323e-01
-1.57137990e-01 2.43999660e-01 -1.90782785e-01 -5.68823636e-01
1.45613730e+00 -1.94604740e-01 -9.02657583e-03 2.87676211e-02
1.00967503e+00 1.80045232e-01 -1.13509011e+00 -2.96663523e-01
2.00660631e-01 -7.84512579e-01 1.88687935e-01 -9.82176363e-01
-1.06416512e+00 3.21606964e-01 5.85478425e-01 5.84772110e-01
1.13169265e+00 -1.54353604e-01 1.42712027e-01 2.96836108e-01
4.13834572e-01 -1.21330893e+00 -3.40783358e-01 -6.07978851e-02
7.11998284e-01 -8.65923047e-01 1.80670425e-01 -3.37469995e-01
-3.38964820e-01 9.82593119e-01 3.40744071e-02 -6.65940791e-02
4.99450028e-01 1.65705964e-01 -5.39129376e-01 4.06310186e-02
-6.65733874e-01 -4.02086437e-01 7.18180612e-02 1.20236121e-01
-3.60793769e-02 4.83084202e-01 -1.03429592e+00 6.58674598e-01
-1.94624856e-01 -4.09511365e-02 4.75501776e-01 9.02161121e-01
-5.34866810e-01 -1.40206134e+00 -3.41411859e-01 5.45703769e-01
-7.01394498e-01 7.59290382e-02 -2.89579481e-01 6.63649023e-01
-4.37713325e-01 8.98715317e-01 -1.69430468e-02 2.98075527e-01
6.40228450e-01 -2.05575034e-01 8.13512564e-01 -1.98053375e-01
-3.14780086e-01 2.73260444e-01 4.86272097e-01 -6.24475300e-01
-9.95799303e-02 -8.83802474e-01 -2.82917172e-01 -4.64015841e-01
-4.74396586e-01 4.01445955e-01 1.09020638e+00 7.09100306e-01
2.25073576e-01 3.14353108e-01 1.03799760e+00 -3.85612965e-01
-1.27750957e+00 -4.09324676e-01 -8.18990290e-01 -3.59290168e-02
9.86053422e-02 -5.33455253e-01 -5.01147270e-01 -4.59412456e-01] | [4.541106224060059, 3.3141894340515137] |
fd0d7ce0-0d23-4264-8d07-9eb5b25eecfe | boosting-multiple-sclerosis-lesion | 2304.10790 | null | https://arxiv.org/abs/2304.10790v1 | https://arxiv.org/pdf/2304.10790v1.pdf | Boosting multiple sclerosis lesion segmentation through attention mechanism | Magnetic resonance imaging is a fundamental tool to reach a diagnosis of multiple sclerosis and monitoring its progression. Although several attempts have been made to segment multiple sclerosis lesions using artificial intelligence, fully automated analysis is not yet available. State-of-the-art methods rely on slight variations in segmentation architectures (e.g. U-Net, etc.). However, recent research has demonstrated how exploiting temporal-aware features and attention mechanisms can provide a significant boost to traditional architectures. This paper proposes a framework that exploits an augmented U-Net architecture with a convolutional long short-term memory layer and attention mechanism which is able to segment and quantify multiple sclerosis lesions detected in magnetic resonance images. Quantitative and qualitative evaluation on challenging examples demonstrated how the method outperforms previous state-of-the-art approaches, reporting an overall Dice score of 89% and also demonstrating robustness and generalization ability on never seen new test samples of a new dedicated under construction dataset. | ['Sebastiano Battiato', 'Francesco Pappalardo', 'Davide Maimone', 'Clara Di Lorenzo', 'Giulia Russo', 'Alessandro Ortis', 'Oliver Giudice', 'Francesco Guarnera', 'Elena Crispino', 'Alessia Rondinella'] | 2023-04-21 | null | null | null | null | ['lesion-segmentation'] | ['medical'] | [ 4.43452060e-01 2.58780401e-02 -1.16009563e-01 -4.30761844e-01
-9.21674848e-01 -1.49514705e-01 4.93807107e-01 1.86548874e-01
-7.24278867e-01 6.64349496e-01 -1.53373331e-01 -3.89456861e-02
-4.88045841e-01 -4.99778032e-01 -4.32533026e-01 -5.17188489e-01
-6.70994699e-01 9.03222620e-01 6.15595579e-01 -1.70100741e-02
3.53252113e-01 7.82915890e-01 -1.20486808e+00 4.37030137e-01
7.93769538e-01 8.11708808e-01 4.93588120e-01 6.72392607e-01
-1.60239846e-01 5.92984498e-01 -3.57917935e-01 2.09546804e-01
3.91127020e-02 -2.89754927e-01 -1.26839244e+00 1.90146789e-01
3.01060170e-01 -3.27976108e-01 -1.38876140e-01 7.10290670e-01
6.83546364e-01 -1.76332484e-03 5.65034032e-01 -5.30411303e-01
-6.76697254e-01 4.31354910e-01 -5.94239175e-01 8.71433973e-01
1.70090750e-01 2.39240944e-01 5.66662192e-01 -3.75223994e-01
1.05547047e+00 8.57052267e-01 7.74773180e-01 5.08816004e-01
-1.41119289e+00 -3.79397482e-01 -1.79577544e-01 5.28181791e-01
-9.57145274e-01 3.58537883e-02 3.26618969e-01 -6.39177203e-01
1.06296659e+00 1.50466472e-01 6.70284390e-01 8.61094177e-01
2.69677222e-01 1.02524471e+00 1.66732049e+00 -4.39599335e-01
1.54174566e-01 -2.57230759e-01 6.52975678e-01 5.88402927e-01
-2.05940753e-01 -5.24469465e-03 -1.69513121e-01 1.94754303e-01
9.11600828e-01 -3.02380044e-02 -2.04272881e-01 -3.22275192e-01
-1.73465645e+00 6.53851271e-01 8.54985952e-01 1.12005424e+00
-7.85615683e-01 1.76281989e-01 4.33503658e-01 1.29100978e-01
5.08287966e-01 2.52117217e-01 -3.02977294e-01 2.51589231e-02
-1.51370597e+00 1.69236183e-01 1.61005720e-01 1.28849074e-01
3.51050645e-01 -3.26811641e-01 -4.72242475e-01 7.75583923e-01
-8.39812756e-02 2.71894723e-01 7.99592376e-01 -7.41275668e-01
1.19145215e-01 7.75708854e-01 -3.48067075e-01 -5.30584395e-01
-1.05324531e+00 -7.23859012e-01 -7.64328718e-01 6.95149720e-01
5.00471354e-01 -2.30873913e-01 -1.44081783e+00 1.31253338e+00
1.08554758e-01 2.96909928e-01 -4.22834843e-01 1.07784724e+00
2.82208264e-01 -7.32092336e-02 -1.89194586e-02 -2.93430630e-02
1.24798501e+00 -9.36143339e-01 -6.06461883e-01 -1.49703905e-01
5.60591936e-01 -3.79900008e-01 8.28176856e-01 3.43856156e-01
-1.15960836e+00 -2.45533735e-01 -1.00739884e+00 1.55344054e-01
-4.76264745e-01 -1.92786846e-02 6.88351274e-01 7.42585659e-01
-1.32111788e+00 1.05769134e+00 -1.16855192e+00 -6.19058371e-01
8.37117434e-01 5.27109921e-01 -6.71939373e-01 1.14803061e-01
-1.10099983e+00 1.07210541e+00 3.68387103e-01 6.58788010e-02
-5.34105062e-01 -7.72939503e-01 -3.32826495e-01 -2.47147635e-01
9.47458148e-02 -7.50982523e-01 1.10329914e+00 -8.10863853e-01
-1.31344187e+00 1.03651559e+00 5.62805198e-02 -1.04757237e+00
9.86070514e-01 -1.57839879e-01 -2.83757329e-01 7.32734859e-01
1.14945076e-01 9.13374186e-01 5.41449666e-01 -8.80636215e-01
-5.36760211e-01 -6.69737637e-01 -1.39904127e-01 -2.25873962e-01
-1.86155975e-01 2.72690624e-01 -8.02091137e-02 -7.51515865e-01
3.76858413e-01 -8.03076863e-01 -5.55448472e-01 -2.73209922e-02
-2.75616914e-01 5.72538264e-02 7.54574418e-01 -1.04167783e+00
1.07235706e+00 -1.33967078e+00 1.54091373e-01 2.68635511e-01
2.47058287e-01 5.55700481e-01 7.18463138e-02 -3.98443416e-02
-3.58145744e-01 1.75105348e-01 -6.39621556e-01 -3.34140062e-02
-2.33451039e-01 3.78752640e-03 2.73020476e-01 5.70088327e-01
2.16314331e-01 1.09676349e+00 -8.19774091e-01 -4.03270572e-01
5.15591025e-01 7.25382090e-01 -2.96931684e-01 -2.63182551e-01
-3.95219103e-02 8.18642080e-01 -3.40483040e-01 6.14682496e-01
2.72417068e-01 -3.32745969e-01 4.96144108e-02 -8.60670023e-03
-8.68022218e-02 -2.64707744e-01 -1.09986997e+00 1.93224192e+00
-3.63232255e-01 5.91341376e-01 -1.49597332e-01 -1.31950569e+00
5.55160701e-01 4.57296461e-01 8.50564599e-01 -1.03556085e+00
3.45937133e-01 5.08330941e-01 2.07677007e-01 -6.94292009e-01
-2.11322177e-02 -2.45182708e-01 4.46003318e-01 5.92293322e-01
9.77971703e-02 3.68058860e-01 3.98396194e-01 6.63584890e-03
1.47780371e+00 2.56345749e-01 -5.15574850e-02 -2.86561549e-01
6.94192171e-01 -3.16595510e-02 8.87831151e-02 7.91186929e-01
-6.24958277e-01 9.72680628e-01 3.02887976e-01 -4.37939286e-01
-1.05749500e+00 -9.44717944e-01 -4.81665224e-01 5.94333649e-01
-3.99426967e-01 2.64264047e-01 -1.15463471e+00 -5.67162931e-01
-2.47551858e-01 4.03586328e-01 -1.06945193e+00 3.56875896e-01
-7.52655685e-01 -8.70488405e-01 4.02279377e-01 7.37977684e-01
6.82967246e-01 -1.30180788e+00 -1.10890687e+00 6.31718993e-01
3.67230084e-03 -9.98805225e-01 1.17143549e-01 3.30067091e-02
-1.27680576e+00 -1.15821719e+00 -1.52976871e+00 -6.29263520e-01
4.31533813e-01 -1.25204295e-01 8.77996862e-01 -9.91972536e-02
-6.94450200e-01 2.64176816e-01 -1.93717748e-01 -4.61112894e-03
-1.90458789e-01 2.33021975e-01 -2.88752437e-01 -1.92169473e-03
1.58502012e-01 -8.14541101e-01 -8.63597393e-01 3.65046076e-02
-1.07386315e+00 -8.22618976e-02 9.38769698e-01 6.93424225e-01
8.97420883e-01 -4.36989099e-01 6.86771452e-01 -9.51891005e-01
4.94236290e-01 -3.37783158e-01 -1.16929613e-01 2.72725672e-01
-6.91573262e-01 5.42126670e-02 7.08205476e-02 -1.54698834e-01
-7.12942123e-01 -1.81326140e-02 -1.94663867e-01 -4.31141526e-01
-4.46467608e-01 4.09992695e-01 4.97441292e-01 -3.16259265e-01
5.71029186e-01 1.49296135e-01 2.18092337e-01 -5.40896237e-01
2.63897926e-01 5.76531589e-01 5.88567793e-01 -1.04222663e-01
1.40851483e-01 7.68862426e-01 7.64017850e-02 -7.68163443e-01
-6.20919466e-01 -6.84641838e-01 -1.10745490e+00 -4.36550736e-01
1.19307292e+00 -2.23066419e-01 -3.07645500e-01 6.58803940e-01
-1.06596494e+00 -4.24481422e-01 -1.48619086e-01 4.96234387e-01
-8.53194892e-01 2.29330495e-01 -5.30785441e-01 -4.30743277e-01
-6.08379960e-01 -1.28599358e+00 9.07434225e-01 7.01654926e-02
-3.25418264e-01 -1.03899729e+00 3.17702383e-01 6.37117624e-01
7.45043099e-01 5.43352008e-01 8.14821899e-01 -6.58729374e-01
-2.27531940e-01 -4.81934160e-01 -4.08746421e-01 2.29948193e-01
-6.04142807e-02 -3.92355025e-01 -7.82299876e-01 -1.14715599e-01
-1.59277827e-01 -1.09804429e-01 1.05818141e+00 8.99571717e-01
1.06904960e+00 3.53991866e-01 -4.26057994e-01 2.38054246e-01
1.42462742e+00 1.05254844e-01 8.57773066e-01 7.97683477e-01
3.03466737e-01 6.72294796e-01 1.18466474e-01 -1.82255581e-01
1.49275541e-01 6.64100349e-01 3.16030055e-01 -1.99406922e-01
-5.31779647e-01 7.61854529e-01 -1.58688053e-01 4.16900218e-01
-2.55212545e-01 3.73724967e-01 -1.45526767e+00 9.67507184e-01
-1.78331351e+00 -1.06066334e+00 -3.25876474e-01 1.94331872e+00
5.66537797e-01 3.10714126e-01 1.27123311e-01 4.15256768e-01
8.87904823e-01 -1.40276989e-02 -5.30198336e-01 -1.98606685e-01
-1.25740007e-01 6.70788348e-01 3.61668110e-01 3.09862942e-01
-1.13995600e+00 5.73428512e-01 6.77073526e+00 7.93719769e-01
-1.26993382e+00 4.53975499e-01 6.63959444e-01 -1.44371167e-01
7.02672601e-02 -2.58785069e-01 -1.99864015e-01 2.93010235e-01
9.65165555e-01 1.98988497e-01 1.70396999e-01 2.85296410e-01
3.08866084e-01 -2.93114424e-01 -7.01941013e-01 7.06584632e-01
2.99845319e-02 -1.44923615e+00 -9.60850492e-02 5.26383929e-02
6.65613830e-01 5.13920486e-01 -4.42169718e-02 9.09129381e-02
-1.75835982e-01 -1.18019664e+00 4.02958453e-01 1.03184068e+00
5.25892496e-01 -6.20187044e-01 9.25974429e-01 2.12429389e-01
-8.18361044e-01 1.15255646e-01 1.17135450e-01 1.93797201e-01
3.72729361e-01 6.31180525e-01 -7.25990891e-01 7.60748744e-01
8.25287104e-01 4.65022504e-01 -7.64749825e-01 1.29981232e+00
-5.29415382e-04 5.29110074e-01 -2.85192132e-01 2.45069191e-01
7.87728310e-01 -1.71356946e-01 8.02245796e-01 1.35706055e+00
1.91712633e-01 -1.07950889e-01 1.08794712e-01 8.31511319e-01
3.39702874e-01 2.82088578e-01 -2.82786906e-01 -6.39148504e-02
-4.00551438e-01 1.38336241e+00 -1.30770504e+00 -2.76368231e-01
-2.82565206e-01 1.04093170e+00 2.31097758e-01 2.33082920e-01
-6.74156129e-01 -3.01363021e-01 -1.24254543e-02 3.84691626e-01
2.89030284e-01 -1.61152378e-01 -5.84979892e-01 -6.93692088e-01
-7.24155009e-02 -5.96333265e-01 4.16151911e-01 -7.89574742e-01
-8.90192568e-01 8.10788453e-01 -2.93462127e-01 -8.79109859e-01
-3.29254061e-01 -6.06734931e-01 -5.26413083e-01 9.47104335e-01
-1.45875931e+00 -1.37880635e+00 -6.62662927e-03 5.82063138e-01
3.85468513e-01 -1.95115790e-01 7.82409906e-01 3.63463640e-01
-3.82732332e-01 7.24801123e-02 1.61857501e-01 6.15166500e-02
4.40454304e-01 -1.50681186e+00 1.64424971e-01 8.40530396e-01
-1.06252648e-01 5.62247038e-01 6.45921350e-01 -6.07834280e-01
-7.15912819e-01 -8.76562119e-01 5.98902047e-01 -1.21917062e-01
9.80197966e-01 4.48747456e-01 -1.08079541e+00 6.14504397e-01
3.31869066e-01 2.19403297e-01 5.25495410e-01 -2.03720257e-01
1.05681382e-01 3.71915221e-01 -1.44733000e+00 4.37489271e-01
8.35861504e-01 -3.09863746e-01 -7.43961871e-01 2.98449367e-01
2.51742005e-01 -2.66594917e-01 -1.05571735e+00 5.87507248e-01
6.54203534e-01 -1.18410528e+00 9.44723606e-01 -6.84960723e-01
4.34882015e-01 -5.54016866e-02 1.37687027e-01 -1.14026475e+00
-2.53554285e-01 -1.86035290e-01 -2.43647456e-01 7.53756106e-01
2.57122308e-01 -5.34602344e-01 9.31683958e-01 2.36108795e-01
-1.49407759e-01 -1.19371676e+00 -1.05843973e+00 -7.92274415e-01
4.62164432e-01 -5.76083362e-01 5.85121065e-02 7.58676827e-01
-3.53897661e-01 5.54481223e-02 4.68965955e-02 -1.04644150e-01
7.45350182e-01 6.35280684e-02 -2.46419348e-02 -1.35914004e+00
-5.96528426e-02 -8.55886281e-01 -8.15624416e-01 -1.73816621e-01
1.83264226e-01 -1.29437840e+00 -4.77549642e-01 -2.09537864e+00
2.27349654e-01 -2.57588089e-01 -6.62882030e-01 4.67441082e-01
7.46020824e-02 7.39779949e-01 5.99770434e-03 1.71196833e-01
-3.98678333e-01 3.31044085e-02 1.39539838e+00 -2.85688251e-01
-1.12102531e-01 -7.16051161e-02 -1.68884262e-01 6.86079860e-01
1.05642605e+00 -2.43939415e-01 -7.90306255e-02 -4.03730124e-01
-1.60399586e-01 -1.25496611e-01 7.07182527e-01 -1.64408398e+00
1.95687100e-01 3.37245733e-01 4.76472408e-01 -6.82447970e-01
-7.28241280e-02 -6.29531503e-01 -1.42926322e-02 1.00867021e+00
-3.39665562e-01 -2.74989367e-01 1.98803291e-01 4.03782934e-01
-2.59386480e-01 -3.02841157e-01 1.04382718e+00 -3.24142486e-01
-8.23239386e-01 4.64678049e-01 -3.87152910e-01 -2.02095151e-01
1.23757875e+00 -3.61881047e-01 -3.00963577e-02 1.27666742e-01
-1.32428432e+00 2.24204943e-01 1.56221613e-01 3.80106211e-01
3.96794230e-01 -1.12876177e+00 -8.08577597e-01 -1.60540402e-01
-1.45060316e-01 -5.32595813e-01 6.09455228e-01 1.54488671e+00
-8.58795166e-01 7.48391151e-01 -6.81004524e-01 -9.05264318e-01
-1.21471429e+00 3.58952552e-01 6.47933245e-01 -6.88719869e-01
-1.07210135e+00 5.91247380e-01 -4.70771581e-01 -5.48503935e-01
9.12433714e-02 -2.02870771e-01 -5.67013562e-01 3.24579298e-01
7.39914477e-01 6.65577531e-01 5.08089125e-01 -5.26019931e-01
-4.20582116e-01 6.35197341e-01 -2.87503958e-01 -4.00216997e-01
1.58902216e+00 -1.41509235e-01 -1.66094556e-01 4.53915864e-01
9.89283144e-01 -7.00826883e-01 -8.85713279e-01 -2.68637240e-01
4.63704437e-01 3.82564552e-02 5.50636232e-01 -1.16526783e+00
-1.28229940e+00 1.28301752e+00 1.47480035e+00 2.83223927e-01
9.39167976e-01 -2.42858037e-01 1.05054796e+00 2.58958369e-01
4.87691730e-01 -1.03043437e+00 -6.99135512e-02 4.27107483e-01
8.77053618e-01 -1.16401136e+00 -2.85160899e-01 6.97271107e-03
-4.23513025e-01 1.34671330e+00 1.60994619e-01 -3.07916492e-01
5.67552686e-01 2.07958803e-01 2.23438814e-01 -4.09029692e-01
-1.47910655e-01 -5.61864793e-01 3.26467574e-01 7.25765347e-01
5.14373064e-01 -1.54890949e-02 -7.00747788e-01 5.04971981e-01
2.77171254e-01 4.67041939e-01 3.30177903e-01 9.41901147e-01
-6.09895110e-01 -1.14188635e+00 -4.04037952e-01 7.13633537e-01
-9.67359841e-01 8.56933892e-02 -1.92165434e-01 8.64006400e-01
1.12376459e-01 4.46365505e-01 -9.54747871e-02 2.40656845e-02
2.25056216e-01 3.13206196e-01 8.28504920e-01 -3.46209049e-01
-8.85309756e-01 2.67805997e-03 -2.01654240e-01 -7.64547706e-01
-8.37480187e-01 -7.22357392e-01 -1.62406301e+00 5.16669974e-02
-7.38873407e-02 -2.66837955e-01 7.24647403e-01 1.18485689e+00
2.51428962e-01 1.11027348e+00 1.03141919e-01 -1.05744910e+00
-1.51119843e-01 -1.09399867e+00 -7.32530892e-01 3.76694143e-01
2.48451024e-01 -7.26542294e-01 8.84870216e-02 1.11718560e-02] | [14.217767715454102, -2.104799747467041] |
bfffee7e-47ed-423d-bc60-aa5ea2bebb90 | learning-to-adapt-to-unseen-abnormal | 2203.13610 | null | https://arxiv.org/abs/2203.13610v1 | https://arxiv.org/pdf/2203.13610v1.pdf | Learning to Adapt to Unseen Abnormal Activities under Weak Supervision | We present a meta-learning framework for weakly supervised anomaly detection in videos, where the detector learns to adapt to unseen types of abnormal activities effectively when only video-level annotations of binary labels are available. Our work is motivated by the fact that existing methods suffer from poor generalization to diverse unseen examples. We claim that an anomaly detector equipped with a meta-learning scheme alleviates the limitation by leading the model to an initialization point for better optimization. We evaluate the performance of our framework on two challenging datasets, UCF-Crime and ShanghaiTech. The experimental results demonstrate that our algorithm boosts the capability to localize unseen abnormal events in a weakly supervised setting. Besides the technical contributions, we perform the annotation of missing labels in the UCF-Crime dataset and make our task evaluated effectively. | ['Bohyung Han', 'Junha Kim', 'Jaeyoo Park'] | 2022-03-25 | null | null | null | null | ['supervised-anomaly-detection'] | ['computer-vision'] | [ 7.64078051e-02 -1.52712762e-01 -2.47235194e-01 -4.74734485e-01
-8.30052376e-01 -4.17888463e-01 4.42039967e-01 2.68967841e-02
-6.14504397e-01 3.48962963e-01 1.02911420e-01 2.56683510e-02
1.96200371e-01 -2.78599799e-01 -8.97148371e-01 -4.72009301e-01
-5.42555571e-01 2.60958642e-01 3.09640318e-01 -1.86374187e-02
4.48836237e-02 2.85177797e-01 -1.60902429e+00 7.47190952e-01
6.65095448e-01 1.09258151e+00 -3.41630131e-01 7.69885421e-01
2.65071601e-01 1.18789935e+00 -5.33693731e-01 -4.33617622e-01
4.86969292e-01 -2.42965326e-01 -8.65502000e-01 5.32998323e-01
8.63067806e-01 -6.91465676e-01 -5.99785447e-01 9.88613546e-01
2.12212533e-01 1.09950557e-01 6.08935118e-01 -1.66163850e+00
-3.40178341e-01 9.51508805e-02 -3.50901335e-01 8.83932769e-01
2.63890564e-01 2.80433804e-01 9.15410757e-01 -9.40706432e-01
4.82483476e-01 8.37054312e-01 6.50412500e-01 8.08064163e-01
-7.60972381e-01 -2.51033038e-01 7.14242339e-01 6.37240112e-01
-1.13531649e+00 -5.38545787e-01 6.81222975e-01 -5.12115955e-01
8.25118244e-01 8.12284723e-02 3.60881537e-01 1.89980829e+00
-2.10059032e-01 1.31634867e+00 7.49113798e-01 -1.55108854e-01
4.97334152e-01 -1.60723597e-01 2.07712591e-01 9.20664608e-01
2.75645673e-01 -9.72307175e-02 -4.80256259e-01 -3.43470931e-01
4.50827271e-01 2.75951236e-01 -1.44219592e-01 -5.27003288e-01
-1.08641839e+00 5.93361914e-01 2.52052851e-04 2.73935288e-01
-2.77652621e-01 1.06012546e-01 8.37665617e-01 5.70322275e-01
7.32876837e-01 4.59597528e-01 -7.59200931e-01 -2.75762618e-01
-6.96807563e-01 -9.80713312e-03 5.55437624e-01 7.99182594e-01
3.73698205e-01 1.65612504e-01 -3.70299429e-01 7.96897948e-01
-1.90975647e-02 1.21598214e-01 6.20101452e-01 -1.22113538e+00
4.94624466e-01 4.86393332e-01 1.73145354e-01 -7.52871871e-01
-4.02153403e-01 -5.97431719e-01 -5.45946240e-01 8.03732648e-02
6.43937886e-01 -2.41037503e-01 -9.84458983e-01 1.68721986e+00
8.08568448e-02 9.50737417e-01 5.43552963e-03 7.20814228e-01
5.06350040e-01 3.29826117e-01 1.92639753e-01 -7.58976117e-02
8.61552060e-01 -1.21764302e+00 -6.31741762e-01 -2.96403080e-01
1.05516529e+00 -1.08887076e-01 1.15432167e+00 5.56902587e-01
-8.05011034e-01 -3.92502964e-01 -7.90658772e-01 4.17497009e-01
-2.86901057e-01 1.15406863e-01 5.24228573e-01 3.60709548e-01
-8.70299697e-01 3.84286702e-01 -1.18152487e+00 -5.23387194e-01
8.35702479e-01 6.11217022e-02 -5.02688706e-01 -9.63428915e-02
-7.95324087e-01 6.34907007e-01 4.88312721e-01 7.10880384e-03
-1.38502157e+00 -4.39144254e-01 -9.71977293e-01 -9.01478902e-02
6.94347858e-01 -4.01425332e-01 1.23629272e+00 -1.44779944e+00
-8.43161643e-01 1.01175022e+00 -9.90369692e-02 -7.98280597e-01
7.57006526e-01 -3.82539988e-01 -7.03007221e-01 3.50992769e-01
5.05494252e-02 4.27887529e-01 1.02613366e+00 -9.66929913e-01
-1.10571682e+00 -3.74371350e-01 3.16487789e-01 -6.81560934e-02
-7.62918115e-01 -2.06890583e-01 -5.51236808e-01 -9.22429800e-01
-2.71014012e-02 -8.26330423e-01 -2.57190704e-01 -1.76845387e-01
-1.79732144e-01 -2.46833250e-01 1.19692695e+00 -4.38209593e-01
1.20839822e+00 -2.31943774e+00 2.17021629e-02 4.25057150e-02
8.00856128e-02 3.36437315e-01 -3.88240427e-01 1.28492303e-02
-2.51150936e-01 -7.11721554e-02 -2.33699724e-01 -4.89451677e-01
-3.40285838e-01 6.63877308e-01 -2.27153495e-01 6.00353360e-01
4.60536361e-01 8.07731271e-01 -1.16241574e+00 -4.65777993e-01
-2.12986507e-02 -1.15978703e-01 -7.72105992e-01 3.88432056e-01
-2.17991844e-01 6.31681085e-01 -5.95063627e-01 1.24599874e+00
2.26717263e-01 -2.84093618e-01 -2.50275284e-01 5.60248718e-02
3.21109831e-01 -2.51387715e-01 -9.50745404e-01 1.85320628e+00
-2.80432589e-03 3.89611691e-01 -1.38783837e-02 -1.48855984e+00
2.73209661e-01 4.02427346e-01 8.21976542e-01 -5.52973807e-01
-1.51427343e-01 1.20091207e-01 -1.46623403e-01 -1.09940803e+00
1.90681830e-01 2.92322755e-01 -4.31851571e-04 1.90930262e-01
5.01495957e-01 5.66962779e-01 3.39946687e-01 3.70264947e-01
1.60599947e+00 2.37317011e-01 -6.27115294e-02 1.22637399e-01
7.59060681e-01 -5.51938377e-02 7.25087523e-01 1.18619120e+00
-7.47512698e-01 6.74011409e-01 5.63804507e-01 -9.04855967e-01
-9.86061990e-01 -1.03666258e+00 -3.20089310e-02 1.55184472e+00
-1.78666815e-01 -4.27074075e-01 -8.03242207e-01 -1.68814504e+00
-1.66398793e-01 4.59105313e-01 -7.28283763e-01 -3.21488410e-01
-6.23271227e-01 -1.01510036e+00 5.95924675e-01 7.23452926e-01
4.18975353e-01 -9.32791173e-01 -2.73045689e-01 4.76258807e-02
-3.30679119e-01 -1.51347601e+00 -2.11090460e-01 1.98730156e-02
-9.69383657e-01 -1.45828450e+00 -3.41072381e-01 -6.87972426e-01
7.89377034e-01 2.12461315e-02 1.13461471e+00 3.68445128e-01
-3.04514021e-01 1.03815091e+00 -6.03764892e-01 -3.30205500e-01
-3.87631088e-01 5.64576546e-03 2.37911344e-01 4.13480669e-01
5.21638393e-01 -4.60892588e-01 -4.56478417e-01 3.31684083e-01
-9.41490114e-01 -4.90234375e-01 3.25486898e-01 9.05642867e-01
5.18639147e-01 -4.15995494e-02 6.40236318e-01 -1.00137520e+00
7.36097246e-02 -8.33384633e-01 -4.06908780e-01 5.47895841e-02
-2.30467781e-01 -1.02325454e-01 5.55079579e-01 -4.13894266e-01
-9.41664398e-01 3.46054971e-01 -2.95965135e-01 -6.94499910e-01
-6.54059172e-01 1.42005160e-01 -8.67307410e-02 6.27086777e-03
7.50290990e-01 1.61120042e-01 -3.20775241e-01 -4.53244597e-01
-2.85372343e-02 3.95657569e-01 9.24291372e-01 -7.07808137e-01
7.87612677e-01 8.63051593e-01 -8.91711265e-02 -6.18274808e-01
-1.45523334e+00 -6.35458648e-01 -8.29635143e-01 -1.97018281e-01
7.87527859e-01 -1.16040707e+00 -2.14781359e-01 5.28191984e-01
-7.35461473e-01 -3.09681475e-01 -4.56052303e-01 4.03781116e-01
-7.62244344e-01 6.21971667e-01 -7.15281606e-01 -7.29486406e-01
6.92531019e-02 -9.27535176e-01 1.11079836e+00 -2.45305404e-01
7.21813738e-03 -1.05922604e+00 9.91922021e-02 4.27890003e-01
1.51132151e-01 4.44589078e-01 5.76756656e-01 -1.21868408e+00
-4.70221579e-01 -4.81499285e-01 1.70149863e-01 6.35062814e-01
-9.02858935e-03 -2.11608037e-01 -1.31449080e+00 -4.39293712e-01
-2.14453507e-03 -5.83774090e-01 1.21756434e+00 1.12861156e-01
1.84259510e+00 -2.97311872e-01 -1.61011547e-01 8.66778255e-01
1.01892006e+00 -1.03240564e-01 5.11350036e-01 7.69092739e-01
8.46670508e-01 2.71416038e-01 7.55731404e-01 4.63237584e-01
3.58150691e-01 4.27791059e-01 8.47884119e-01 -1.37531921e-01
1.99834481e-01 -1.45934567e-01 5.70726931e-01 1.72047704e-01
-2.34333128e-01 -3.47046763e-01 -7.70806909e-01 7.87269711e-01
-2.18445516e+00 -1.25168836e+00 3.29611301e-02 2.01227045e+00
3.26472819e-01 2.60763377e-01 3.89184415e-01 1.38902009e-01
4.25672412e-01 1.17005900e-01 -5.29709220e-01 -5.06521650e-02
-2.60835201e-01 -1.91151187e-01 2.79447198e-01 2.25700781e-01
-1.86148870e+00 8.71162355e-01 6.70276356e+00 5.23593724e-01
-8.03967774e-01 4.10651356e-01 8.13056231e-01 -4.90412503e-01
4.58544910e-01 -4.21019047e-01 -5.08661747e-01 7.46228337e-01
1.02245307e+00 4.16888595e-01 1.54231876e-01 9.76905584e-01
1.64209828e-01 2.00356886e-01 -1.46840727e+00 9.06233966e-01
2.78235316e-01 -1.01441097e+00 3.16682532e-02 -1.59406871e-01
7.57520318e-01 4.46345776e-01 1.87252805e-01 5.78575492e-01
8.29883814e-02 -8.33994865e-01 5.18456280e-01 3.85160327e-01
3.11240882e-01 -6.00532293e-01 9.07902777e-01 5.15735209e-01
-7.26061165e-01 -6.98670566e-01 -2.48531088e-01 -1.81270450e-01
5.50086796e-02 3.42881948e-01 -7.90586114e-01 2.01115936e-01
1.02097738e+00 1.02618396e+00 -1.00423050e+00 9.99879658e-01
-2.27652267e-01 9.06147659e-01 -1.79448858e-01 7.36387610e-01
4.09219623e-01 1.07232310e-01 7.85393178e-01 1.27380502e+00
2.17961162e-01 -1.11896075e-01 6.12965226e-01 3.38983893e-01
-1.55675635e-01 -5.80293164e-02 -7.71423519e-01 8.58633742e-02
3.94543782e-02 1.24446523e+00 -3.46013457e-01 -4.19263899e-01
-7.78314114e-01 1.23884261e+00 4.91605282e-01 5.05823016e-01
-9.37547028e-01 3.11785430e-01 6.44039214e-01 2.02783093e-01
2.78306097e-01 -6.32370031e-03 5.24275042e-02 -1.76768744e+00
2.37617806e-01 -9.37978029e-01 1.08885860e+00 -4.57048416e-01
-1.49211097e+00 4.39744085e-01 3.77633423e-02 -1.33892751e+00
-4.81834650e-01 -7.63850689e-01 -6.75080240e-01 1.41928503e-02
-1.61758447e+00 -1.21192956e+00 -2.69159168e-01 8.95428479e-01
7.42416918e-01 -5.36435425e-01 6.84421301e-01 5.19287467e-01
-8.56421351e-01 7.53110051e-01 -1.14225633e-01 6.04175091e-01
8.45312655e-01 -1.47426915e+00 2.31043845e-01 1.19908226e+00
4.07685071e-01 1.73882172e-02 5.50598025e-01 -3.96631449e-01
-1.06610012e+00 -1.45083904e+00 4.01489258e-01 -1.02728808e+00
8.13566864e-01 -3.68746638e-01 -1.13044810e+00 1.16945028e+00
-2.43434176e-01 7.65242219e-01 7.27902174e-01 1.23414606e-01
-3.16267580e-01 7.18375891e-02 -1.24930286e+00 4.07677889e-01
1.39179933e+00 -4.60879117e-01 -6.03090107e-01 7.40537941e-01
4.13656592e-01 -5.73662877e-01 -5.55745542e-01 6.11856103e-01
1.61070898e-01 -9.46546018e-01 1.00009918e+00 -1.39279473e+00
3.45573425e-01 -2.47183233e-01 -1.15467735e-01 -1.20892382e+00
-1.72354966e-01 -2.63459653e-01 -8.02236915e-01 9.63931084e-01
3.13675970e-01 -4.51377243e-01 9.61560726e-01 4.82032031e-01
-3.37132186e-01 -7.03042269e-01 -9.70908046e-01 -8.48395705e-01
-3.33735496e-01 -8.40787411e-01 2.41519853e-01 1.08747172e+00
-1.79376334e-01 -1.02118030e-02 -7.36147404e-01 5.62199950e-01
6.69622064e-01 -4.34497714e-01 6.42825305e-01 -1.20469558e+00
-3.71444702e-01 6.24832138e-02 -7.90093899e-01 -7.00862288e-01
6.30462348e-01 -6.59935355e-01 -2.43616089e-01 -7.17043877e-01
2.96839893e-01 -1.25931993e-01 -7.51466095e-01 7.18522072e-01
-2.33383939e-01 5.66363037e-01 -2.01013744e-01 2.01713115e-01
-1.36320245e+00 2.52576560e-01 7.25935876e-01 -1.88320041e-01
2.34218277e-02 2.16662243e-01 -4.32998478e-01 1.36887825e+00
6.90264285e-01 -4.61760610e-01 -2.26726785e-01 -7.06743538e-01
-1.60592366e-02 -3.90010118e-01 6.24637067e-01 -1.07399321e+00
6.34650886e-02 5.18827662e-02 6.50761366e-01 -3.41709256e-01
1.90939873e-01 -1.06251383e+00 -7.43617594e-01 3.22129965e-01
-4.42506909e-01 1.49397433e-01 6.46293685e-02 1.00731850e+00
-2.01460317e-01 -2.55559325e-01 6.80812120e-01 -1.94322407e-01
-1.17166471e+00 7.38712132e-01 -2.83790886e-01 3.88035566e-01
1.17798805e+00 -1.35115609e-01 -2.18494087e-01 -5.15590489e-01
-1.14639819e+00 4.98596847e-01 5.60996413e-01 6.04638755e-01
3.88515979e-01 -1.21453691e+00 -7.36417294e-01 2.70565718e-01
5.30016363e-01 -2.44889960e-01 1.99080989e-01 9.13145065e-01
-3.05117041e-01 8.30285698e-02 -1.45211041e-01 -8.58593285e-01
-1.14766574e+00 8.49071264e-01 5.49592614e-01 -2.13703126e-01
-6.61163032e-01 7.73822784e-01 3.69887576e-02 -1.67649552e-01
6.43862188e-01 -1.63352996e-01 -2.18822956e-01 -1.55066431e-01
7.24008024e-01 5.46229720e-01 2.37848699e-01 -6.70971215e-01
-3.68532002e-01 5.12908101e-02 -2.74835885e-01 2.22900748e-01
1.38161075e+00 -8.45795721e-02 2.87958026e-01 3.27876300e-01
1.07673323e+00 -1.17063597e-01 -1.48006558e+00 -3.10476124e-01
4.20690745e-01 -5.94033122e-01 -1.28504634e-01 -6.87382042e-01
-1.18286359e+00 6.10000968e-01 8.55697930e-01 6.64020181e-02
1.25551343e+00 1.60500869e-01 7.20040917e-01 8.75620127e-01
2.05111906e-01 -1.31604946e+00 5.54212689e-01 7.07195759e-01
4.27135289e-01 -1.83346820e+00 -2.64829934e-01 -8.13110843e-02
-7.21388400e-01 8.92144084e-01 1.10382783e+00 -1.33201912e-01
3.23392600e-01 2.63118684e-01 2.86274850e-02 -1.89212888e-01
-8.29249501e-01 -2.29532659e-01 3.74017477e-01 7.72391260e-01
2.01039717e-01 -3.71373653e-01 1.24457687e-01 5.27546704e-01
4.32783246e-01 -4.37060744e-02 5.56392014e-01 1.17460263e+00
-4.18174893e-01 -9.68050182e-01 -3.25622976e-01 4.60691363e-01
-1.05157948e+00 1.98815241e-01 -3.08110863e-01 7.22550869e-01
3.14474493e-01 7.62054384e-01 3.43598753e-01 -1.80726975e-01
4.29241955e-01 4.89136219e-01 2.95128226e-01 -7.18635440e-01
-2.74800807e-01 -4.67388779e-02 4.31040581e-03 -1.03526604e+00
-4.97503668e-01 -6.65155232e-01 -9.60669279e-01 3.24174672e-01
7.67880902e-02 8.55804905e-02 2.54920218e-02 1.23595774e+00
3.42982382e-01 4.13547724e-01 6.58934951e-01 -6.06713831e-01
-5.79915702e-01 -8.13234866e-01 -5.26592672e-01 9.23647940e-01
7.49136508e-01 -6.24453127e-01 -5.62445879e-01 2.04852134e-01] | [7.854234218597412, 1.5832692384719849] |
8832bc91-2be0-47a1-bbb3-0b80f6688835 | feature-embedding-in-click-through-rate | 2209.09481 | null | https://arxiv.org/abs/2209.09481v1 | https://arxiv.org/pdf/2209.09481v1.pdf | Feature embedding in click-through rate prediction | We tackle the challenge of feature embedding for the purposes of improving the click-through rate prediction process. We select three models: logistic regression, factorization machines and deep factorization machines, as our baselines and propose five different feature embedding modules: embedding scaling, FM embedding, embedding encoding, NN embedding and the embedding reweighting module. The embedding modules act as a way to improve baseline model feature embeddings and are trained alongside the rest of the model parameters in an end-to-end manner. Each module is individually added to a baseline model to obtain a new augmented model. We test the predictive performance of our augmented models on a publicly accessible dataset used for benchmarking click-through rate prediction models. Our results show that several proposed embedding modules provide an important increase in predictive performance without a drastic increase in training time. | ['Jure Demšar', 'Davorin Kopič', 'Samo Pahor'] | 2022-09-20 | null | null | null | null | ['click-through-rate-prediction'] | ['miscellaneous'] | [-2.20455118e-02 -1.19451163e-02 -5.39995790e-01 -3.99326622e-01
-7.47471631e-01 -2.59882331e-01 8.33392203e-01 2.66146451e-01
-6.36898816e-01 4.39104319e-01 5.77119052e-01 -5.41028976e-01
-3.35716195e-02 -8.18124592e-01 -5.00848949e-01 -1.93036020e-01
-8.96378756e-02 2.81581938e-01 1.68817595e-01 -2.67733663e-01
2.64268547e-01 2.25750089e-01 -1.41824889e+00 4.94154006e-01
7.61240900e-01 1.05933440e+00 -3.61298412e-01 8.29076290e-01
1.61777195e-02 5.99466503e-01 -9.60117057e-02 -7.41541922e-01
4.04309690e-01 2.49451905e-01 -5.67227423e-01 -3.33177447e-01
2.52750516e-01 -4.90268975e-01 -8.22912872e-01 4.00501519e-01
4.23448145e-01 1.21951550e-01 6.97795033e-01 -1.10042238e+00
-9.43823814e-01 3.29404116e-01 -3.26923788e-01 3.14972073e-01
2.49491304e-01 1.75492957e-01 1.58973515e+00 -1.18493724e+00
4.37971860e-01 1.10780549e+00 7.30223656e-01 5.18817186e-01
-1.55567181e+00 -4.45595592e-01 1.50526926e-01 8.75906125e-02
-9.92426932e-01 -1.32702291e-01 3.84659618e-01 -4.08192635e-01
1.35258138e+00 2.79272854e-01 5.36071062e-01 1.12065327e+00
-5.11757936e-03 7.20606506e-01 7.58031249e-01 -5.04422843e-01
6.50035441e-02 5.79377532e-01 2.64954627e-01 6.76742256e-01
6.09724671e-02 3.11205387e-01 -5.20292342e-01 -5.12618303e-01
5.10531306e-01 4.00844693e-01 -3.81537899e-02 -3.74638885e-01
-1.20336056e+00 1.35743010e+00 7.84527540e-01 5.78693561e-02
-2.08595976e-01 5.09917736e-01 3.42524648e-01 4.90770340e-01
8.31175804e-01 5.55719495e-01 -8.61234069e-01 -3.85741264e-01
-6.59041882e-01 5.79627097e-01 8.32507908e-01 3.80178958e-01
7.61680663e-01 -2.44369879e-01 -5.37947834e-01 9.75295663e-01
4.27373499e-01 2.24606663e-01 8.11664522e-01 -5.00527382e-01
6.45319343e-01 9.61566210e-01 2.97151417e-01 -8.54393125e-01
-2.31325909e-01 -5.22219658e-01 -2.32159346e-01 -2.91551322e-01
-2.15166975e-02 6.89451844e-02 -8.59247208e-01 1.44626272e+00
2.96052396e-01 1.56080768e-01 -3.82777721e-01 5.60946286e-01
3.43459010e-01 6.87346876e-01 3.67187172e-01 3.11763376e-01
1.33321023e+00 -1.55203056e+00 -4.53597546e-01 -7.04803318e-02
1.19005239e+00 -9.31021571e-01 1.21166837e+00 1.37567490e-01
-9.69694912e-01 -5.42196572e-01 -1.02425289e+00 -1.70150310e-01
-7.62259245e-01 4.52934325e-01 1.04823983e+00 6.62746370e-01
-8.65500689e-01 9.14488256e-01 -8.06604326e-01 -1.15248216e-02
2.23343879e-01 4.41306174e-01 -3.31410825e-01 -1.20409243e-01
-1.23670590e+00 9.04952168e-01 2.33307868e-01 -3.27211589e-01
-5.32259643e-01 -1.21331263e+00 -6.83878720e-01 3.82708281e-01
-3.56111340e-02 -8.87909591e-01 1.24285924e+00 -3.80569786e-01
-1.19468403e+00 3.89217645e-01 -2.14898512e-01 -5.47485948e-01
5.42612612e-01 -6.43908203e-01 -4.07036066e-01 -2.65034258e-01
-2.44774759e-01 6.00775599e-01 1.03826880e+00 -6.05060816e-01
-7.36822009e-01 -1.12356946e-01 6.42460734e-02 -1.27684370e-01
-8.55502427e-01 -4.85299043e-02 -4.04134363e-01 -7.51933277e-01
-3.47851515e-01 -9.81182754e-01 -4.89163220e-01 -9.62215140e-02
3.83684859e-02 -3.49858612e-01 8.03284347e-01 -7.55954087e-01
1.83206940e+00 -2.04109335e+00 1.17308483e-01 1.17206387e-01
4.77327675e-01 4.66488987e-01 -4.74130601e-01 7.78453887e-01
-1.81411326e-01 5.07285714e-01 1.72647238e-01 -5.85757971e-01
2.27221861e-01 6.23421855e-02 -2.60044903e-01 -3.45985666e-02
6.11566722e-01 1.12889993e+00 -7.07965016e-01 1.06771169e-02
3.75031441e-01 9.15706515e-01 -1.22343159e+00 3.35569054e-01
-2.43579209e-01 -1.49170816e-01 -2.50834286e-01 3.68812293e-01
4.66088325e-01 -4.60891724e-01 -1.17081083e-01 -2.41570041e-01
1.77426293e-01 6.75012112e-01 -1.09469306e+00 1.44761741e+00
-9.31216002e-01 3.38477939e-01 -6.88522398e-01 -5.23419917e-01
8.42380822e-01 1.07630499e-01 3.98823470e-01 -5.00731647e-01
-7.89733380e-02 1.85203820e-01 4.12215851e-02 -3.85639042e-01
7.40878642e-01 1.60246938e-01 1.48540854e-01 7.70967245e-01
4.62025940e-01 5.43724179e-01 2.85885155e-01 5.58252335e-01
1.31560981e+00 -1.35715038e-01 2.37313397e-02 1.84814259e-01
4.41036284e-01 -4.50187862e-01 5.09060100e-02 3.82994235e-01
6.29973263e-02 3.85472268e-01 5.02481699e-01 -7.09797382e-01
-1.39023411e+00 -9.12205994e-01 -4.73357998e-02 1.47236776e+00
-6.45912945e-01 -1.27380919e+00 -2.23138168e-01 -1.12592566e+00
4.99335647e-01 5.94549716e-01 -1.15296590e+00 -5.66306531e-01
-4.61691737e-01 -7.66783416e-01 8.01852643e-02 8.10014963e-01
6.76845014e-02 -3.67116362e-01 5.58174588e-02 2.57070988e-01
3.56225409e-02 -9.67392266e-01 -2.97907174e-01 4.33944225e-01
-9.98846889e-01 -8.38058591e-01 -4.48691815e-01 -3.99606347e-01
5.84427714e-01 3.13046694e-01 9.87802684e-01 3.39284360e-01
-2.63000548e-01 2.88212538e-01 -5.50462782e-01 -1.71254173e-01
-2.74877012e-01 5.16123772e-01 2.43194848e-02 -3.79000492e-02
6.39313936e-01 -4.64431494e-01 -1.01049006e+00 4.85114455e-01
-8.60127091e-01 -1.74421996e-01 4.95697916e-01 1.17619848e+00
1.71197817e-01 -5.83067000e-01 3.93167377e-01 -8.90648186e-01
8.95121872e-01 -5.66817462e-01 -2.37890184e-01 1.35441601e-01
-9.97381091e-01 4.14405406e-01 4.17314202e-01 -4.25150216e-01
-5.69521666e-01 8.78985785e-03 -3.54376584e-01 -3.04229677e-01
3.95492077e-01 5.19323409e-01 3.91024202e-01 -3.43953878e-01
7.94448197e-01 4.84693721e-02 4.06260975e-02 -9.34002280e-01
8.87758136e-01 7.03536928e-01 -1.81028396e-01 -2.54566580e-01
1.22558653e+00 6.73212707e-02 -3.90510887e-01 -2.35219434e-01
-8.42529476e-01 -8.00578833e-01 -8.69511545e-01 2.62491018e-01
4.21915770e-01 -9.91971076e-01 -3.70872438e-01 -3.68889093e-01
-8.59612048e-01 1.73176318e-01 -3.61249387e-01 5.17297029e-01
-2.92438477e-01 1.71738401e-01 -7.22785532e-01 -5.33699334e-01
-4.72633868e-01 -9.66549814e-01 1.15159142e+00 7.76390508e-02
-4.87577803e-02 -1.13342094e+00 5.20109773e-01 4.48434949e-01
9.72478211e-01 -2.57134318e-01 1.09501147e+00 -8.63403678e-01
-2.05932200e-01 -8.50569248e-01 -4.26916569e-01 5.24502635e-01
-2.97711883e-02 2.40421265e-01 -1.10467422e+00 -3.97306859e-01
-4.81592387e-01 -3.18363339e-01 1.45501065e+00 -7.84485117e-02
1.12745619e+00 -3.37250710e-01 -3.95617723e-01 5.57042480e-01
1.44525051e+00 -3.42564970e-01 6.09304726e-01 4.66810226e-01
4.07667041e-01 -4.79425117e-02 5.34523547e-01 5.71061969e-01
2.86895990e-01 6.66028142e-01 4.09394622e-01 2.11328510e-02
4.97404113e-02 -7.71738648e-01 3.51974130e-01 8.38314772e-01
-2.18117982e-01 8.73162895e-02 -5.72844744e-01 3.08922619e-01
-1.88508058e+00 -7.36836374e-01 2.75953412e-01 2.19483685e+00
5.92090368e-01 1.39288619e-01 4.53490466e-01 3.81211936e-01
2.55781144e-01 2.63363034e-01 -2.12548167e-01 -7.90354133e-01
5.39408803e-01 3.69170249e-01 3.52998614e-01 3.35870206e-01
-1.20228755e+00 1.08371902e+00 6.96232986e+00 7.25771010e-01
-1.04041731e+00 2.51392990e-01 4.56959248e-01 -3.84405583e-01
-4.51940238e-01 9.34498087e-02 -1.10377574e+00 4.14694935e-01
1.38841605e+00 -1.09114289e-01 4.17825431e-01 1.14237988e+00
2.52327956e-02 3.96397650e-01 -1.18792593e+00 7.30589628e-01
-1.88718975e-01 -1.57080412e+00 2.67712653e-01 2.65074611e-01
6.10231400e-01 2.91293293e-01 3.53473872e-01 9.86661673e-01
3.49504650e-01 -1.04792809e+00 5.10369651e-02 2.34639362e-01
5.75534165e-01 -6.85697973e-01 7.09757686e-01 -6.79102121e-03
-9.57927585e-01 -4.69350159e-01 -6.45756304e-01 -2.59231001e-01
2.27009729e-01 4.99042153e-01 -1.06134939e+00 3.38828683e-01
3.28236163e-01 7.10144997e-01 -1.11043406e+00 1.05755723e+00
-1.83620825e-01 8.49918962e-01 -2.58149683e-01 -1.18281893e-01
3.64281297e-01 2.65742660e-01 1.52305111e-01 1.09552670e+00
-3.01427804e-02 -4.47352350e-01 -1.33493289e-01 6.49706244e-01
-2.82151878e-01 2.57657140e-01 -3.12953889e-01 -5.35417676e-01
2.42064983e-01 1.64937663e+00 -9.59736407e-02 -1.56694800e-01
-7.50787556e-01 7.81786501e-01 6.35767341e-01 2.58259594e-01
-1.15346003e+00 -3.46879661e-01 8.39796543e-01 3.47354740e-01
6.58161402e-01 -3.67088467e-01 7.01687708e-02 -1.49861813e+00
-1.90119129e-02 -6.38636529e-01 2.98554212e-01 -3.20599139e-01
-1.27354348e+00 3.32771212e-01 -4.50283974e-01 -1.12238586e+00
-2.90952832e-01 -6.52719676e-01 -6.98815107e-01 8.38061213e-01
-1.87731743e+00 -1.32490563e+00 -2.16454998e-01 3.59888047e-01
3.51356775e-01 -2.65668124e-01 1.11698639e+00 6.85054660e-01
-5.97615123e-01 9.49125886e-01 3.72787535e-01 2.08531134e-02
7.34247148e-01 -1.22674835e+00 7.67901063e-01 3.64277691e-01
3.52498949e-01 9.16831195e-01 4.13830847e-01 -3.21362346e-01
-1.30319047e+00 -1.21816564e+00 1.16624022e+00 -6.89842701e-01
1.11988091e+00 -7.30786562e-01 -5.93719363e-01 8.57518494e-01
-2.06323549e-01 4.18983787e-01 1.25294948e+00 7.59312212e-01
-7.64081717e-01 -3.71314888e-03 -8.15234423e-01 3.07152092e-01
7.11115897e-01 -6.71701372e-01 -4.19798404e-01 4.77570564e-01
1.12139773e+00 9.27063376e-02 -1.35959709e+00 1.74486443e-01
9.10292149e-01 -6.71805084e-01 1.27485740e+00 -1.43319058e+00
8.78077865e-01 7.13384449e-02 -2.43516460e-01 -1.33313322e+00
-8.20146620e-01 -3.30098093e-01 -6.84579790e-01 1.06434023e+00
9.14077461e-01 -5.92808127e-01 1.01419199e+00 8.40511024e-01
4.02806103e-01 -1.33704448e+00 -8.05638611e-01 -5.53240418e-01
6.75644204e-02 -4.02521789e-01 5.90146780e-01 5.74799597e-01
5.17025925e-02 6.01396501e-01 -6.50867045e-01 -3.88925433e-01
1.81692809e-01 -1.91125244e-01 1.03795969e+00 -1.09695244e+00
-6.26272082e-01 -2.22072899e-01 -6.57450497e-01 -1.00067663e+00
-2.81520158e-01 -1.09663939e+00 -9.53356028e-01 -1.36358798e+00
3.36121917e-01 -5.66374958e-01 -7.60019600e-01 5.44036984e-01
-4.80439544e-01 3.73282462e-01 4.19234753e-01 2.18227312e-01
-6.61321342e-01 9.23937082e-01 8.88967693e-01 -1.29370779e-01
-1.18365817e-01 2.37309694e-01 -8.74880254e-01 3.02786708e-01
7.20346510e-01 -4.33666587e-01 -2.70714432e-01 -4.84246016e-01
5.29498816e-01 -5.47184229e-01 1.54364422e-01 -8.77237916e-01
-2.57085413e-01 2.24902913e-01 7.13521600e-01 -2.58667469e-01
4.32165653e-01 -7.91029930e-01 -4.37730670e-01 5.92097819e-01
-6.68055117e-01 2.71930605e-01 -2.04238184e-02 8.81176889e-01
-1.11446008e-01 -2.63229199e-02 4.89859730e-01 2.36776158e-01
-3.52878213e-01 4.25976843e-01 1.30413681e-01 -2.69917995e-01
8.67426217e-01 1.63318720e-02 -3.10553223e-01 -1.68853343e-01
-9.04490292e-01 1.45784304e-01 1.24874867e-01 8.37220907e-01
4.24267828e-01 -1.55099130e+00 -5.31895339e-01 4.30752486e-01
3.26171964e-01 -7.47009158e-01 -1.62636787e-02 7.79279351e-01
-2.36458316e-01 7.37678766e-01 -1.87249240e-02 -3.34229618e-01
-1.20654035e+00 6.84636474e-01 6.33384883e-02 -9.32712674e-01
-2.81345308e-01 8.82970333e-01 -2.54784048e-01 -7.20543444e-01
2.78995544e-01 -4.36410636e-01 -2.05357939e-01 -5.00483476e-02
6.93787456e-01 3.72049749e-01 3.04918408e-01 -1.28790200e-01
1.15188276e-02 2.94955283e-01 -6.40723288e-01 1.74907684e-01
1.62184942e+00 1.09312102e-01 3.90537649e-01 2.51678795e-01
1.59684837e+00 -8.14288259e-02 -9.61842358e-01 -4.65424716e-01
-1.03075951e-01 -9.42133605e-01 2.15999410e-01 -9.35169339e-01
-9.33167994e-01 9.86825764e-01 6.89663172e-01 8.80079791e-02
6.90193176e-01 -3.15998733e-01 8.90908003e-01 3.01115692e-01
6.36192635e-02 -1.09381390e+00 3.40504080e-01 5.06943405e-01
6.87263072e-01 -1.44892490e+00 1.92804769e-01 -1.42686218e-01
-5.79406083e-01 1.15703785e+00 6.10697448e-01 -2.82754630e-01
1.02018058e+00 2.80169453e-02 -3.76876205e-01 1.58771053e-01
-1.45103431e+00 -8.38424489e-02 5.89086831e-01 3.41632038e-01
7.99494624e-01 -1.39227092e-01 -5.39444387e-01 6.37343585e-01
-1.88010842e-01 2.12466210e-01 -7.16685280e-02 7.11633682e-01
-3.60089809e-01 -1.47841322e+00 2.73155034e-01 9.72724319e-01
-4.16061789e-01 -2.57875115e-01 -1.68995395e-01 7.08340585e-01
-1.87804386e-01 4.42426622e-01 -9.94021520e-02 -1.07666469e+00
3.06943029e-01 3.36122513e-01 2.15974346e-01 -4.88295704e-01
-1.00630116e+00 -1.68731987e-01 9.72349942e-02 -7.07056820e-01
1.62462756e-01 -2.47046068e-01 -5.55333555e-01 -4.15113240e-01
-6.97013319e-01 2.14377135e-01 9.15145218e-01 5.31727970e-01
6.99722886e-01 5.11825740e-01 8.83680642e-01 -7.47596741e-01
-1.06326783e+00 -1.21722960e+00 -2.21328601e-01 5.01121461e-01
1.20022103e-01 -8.09783459e-01 -5.65635860e-01 -3.96867424e-01] | [10.151095390319824, 5.614904880523682] |
c07ec659-5e4a-4242-9334-60f44b66a63d | transfer-meets-hybrid-a-synthetic-approach | 1901.07199 | null | http://arxiv.org/abs/1901.07199v1 | http://arxiv.org/pdf/1901.07199v1.pdf | Transfer Meets Hybrid: A Synthetic Approach for Cross-Domain Collaborative Filtering with Text | Collaborative filtering (CF) is the key technique for recommender systems
(RSs). CF exploits user-item behavior interactions (e.g., clicks) only and
hence suffers from the data sparsity issue. One research thread is to integrate
auxiliary information such as product reviews and news titles, leading to
hybrid filtering methods. Another thread is to transfer knowledge from other
source domains such as improving the movie recommendation with the knowledge
from the book domain, leading to transfer learning methods. In real-world life,
no single service can satisfy a user's all information needs. Thus it motivates
us to exploit both auxiliary and source information for RSs in this paper. We
propose a novel neural model to smoothly enable Transfer Meeting Hybrid (TMH)
methods for cross-domain recommendation with unstructured text in an end-to-end
manner. TMH attentively extracts useful content from unstructured text via a
memory module and selectively transfers knowledge from a source domain via a
transfer network. On two real-world datasets, TMH shows better performance in
terms of three ranking metrics by comparing with various baselines. We conduct
thorough analyses to understand how the text content and transferred knowledge
help the proposed model. | ['Guang-Neng Hu', 'Qiang Yang', 'Yu Zhang'] | 2019-01-22 | null | null | null | null | ['movie-recommendation'] | ['miscellaneous'] | [ 1.87467456e-01 -2.54637897e-01 -5.02424121e-01 -4.96018350e-01
-5.47494471e-01 -4.44651991e-01 3.13727587e-01 -1.03622779e-01
-2.72342741e-01 6.58414423e-01 5.48985481e-01 -2.78599054e-01
-3.81096780e-01 -8.78844738e-01 -6.99782670e-01 -3.42855722e-01
9.51761305e-02 2.18526855e-01 1.34348646e-01 -6.35763586e-01
3.48785937e-01 -9.37681720e-02 -1.41759658e+00 7.68275201e-01
1.40090442e+00 1.10558820e+00 5.63472092e-01 2.00469092e-01
-3.60157549e-01 7.88964510e-01 -2.57212937e-01 -6.80186570e-01
3.62645566e-01 -2.32738465e-01 -5.12233973e-01 -1.71096548e-01
4.26487207e-01 -4.43713278e-01 -3.73462349e-01 9.35431600e-01
4.97951806e-01 6.50431275e-01 6.32597446e-01 -8.65727365e-01
-1.35643029e+00 1.10718155e+00 -4.86636370e-01 -3.42787132e-02
3.20444882e-01 -5.04306257e-01 1.21206629e+00 -1.10993779e+00
4.65466768e-01 1.12729156e+00 7.04204202e-01 5.48759937e-01
-7.82818496e-01 -7.07991600e-01 6.82607472e-01 5.64769618e-02
-8.43827784e-01 -1.75415814e-01 8.61130834e-01 -1.74481899e-01
6.13588214e-01 1.74490437e-01 3.41605306e-01 1.21515918e+00
3.13712545e-02 1.44831860e+00 8.44308615e-01 -8.67096782e-02
1.19964890e-01 7.17410147e-01 4.71485645e-01 1.90103322e-01
6.05155528e-02 -8.12179744e-02 -6.92181289e-01 -4.96195182e-02
5.91734946e-01 7.93908834e-01 -3.12227696e-01 -1.09648637e-01
-9.58637178e-01 9.57996309e-01 6.26527011e-01 2.91050643e-01
-2.98116565e-01 -5.36279559e-01 2.96519309e-01 7.86246002e-01
7.48030782e-01 6.22120142e-01 -1.05407727e+00 2.05436155e-01
-8.21301281e-01 1.26326710e-01 1.06770945e+00 1.11757088e+00
4.35556591e-01 -1.47529999e-02 -1.83224812e-01 1.11351383e+00
5.07942617e-01 5.21278501e-01 5.75407147e-01 -5.74932396e-01
5.68192899e-01 5.57577133e-01 3.21108758e-01 -1.04774141e+00
-2.81501472e-01 -1.06832016e+00 -9.61082518e-01 -5.31455398e-01
2.11505905e-01 -4.42604810e-01 -4.92655814e-01 1.53979504e+00
2.09192365e-01 2.06502736e-01 -4.90732305e-02 1.05519629e+00
1.07184255e+00 7.16098547e-01 -1.78325593e-01 -3.28169256e-01
1.00239944e+00 -1.46457577e+00 -6.91516221e-01 -1.30517319e-01
6.55778050e-01 -7.20268011e-01 1.17113602e+00 8.45613420e-01
-9.04324234e-01 -8.07269394e-01 -8.80323470e-01 -1.99612260e-01
-5.13790786e-01 5.00287414e-01 7.43957281e-01 4.30132180e-01
-6.60490453e-01 8.81150901e-01 -4.00166631e-01 -1.86329260e-01
3.69379163e-01 4.13205266e-01 -7.61822192e-03 -3.72395098e-01
-1.51049471e+00 5.20431340e-01 -5.61269075e-02 1.50637046e-01
-3.73401284e-01 -9.48357701e-01 -3.85528982e-01 2.34471828e-01
5.47775328e-01 -8.25738668e-01 1.22731745e+00 -1.32529926e+00
-1.78130269e+00 -8.56434032e-02 -3.48023027e-02 -2.60520488e-01
2.04734325e-01 -8.44555974e-01 -7.63482690e-01 -3.20790619e-01
-1.50939897e-01 8.60781223e-02 8.99683714e-01 -1.03312910e+00
-1.05180824e+00 -4.62318987e-01 2.24507809e-01 2.27916166e-01
-9.46626306e-01 -1.48460880e-01 -5.20382464e-01 -9.49296594e-01
-1.70076594e-01 -9.13562477e-01 -1.73586875e-01 -4.06324536e-01
-7.08573386e-02 -4.34728533e-01 7.13192463e-01 -7.10307837e-01
1.42525411e+00 -1.96151054e+00 1.15783788e-01 1.62801102e-01
2.09609732e-01 4.16575879e-01 -3.84391934e-01 5.52021682e-01
3.29659700e-01 -1.16099350e-01 4.35541719e-01 -2.61791557e-01
1.05680779e-01 -4.37437035e-02 -5.74119747e-01 1.28248274e-01
-2.80381680e-01 9.21020150e-01 -1.03683245e+00 -5.95113682e-03
-1.43928319e-01 5.29692352e-01 -9.83737111e-01 2.72993505e-01
-2.81424284e-01 4.93863702e-01 -8.13512206e-01 4.81133699e-01
7.08096266e-01 -5.60494959e-01 1.01233229e-01 -4.46679205e-01
5.13974130e-02 5.61442673e-01 -1.05095291e+00 1.98526275e+00
-7.68663466e-01 9.42846537e-02 8.80039409e-02 -9.57682073e-01
1.01616108e+00 1.46917105e-01 4.74057734e-01 -9.68366802e-01
8.70932937e-02 1.36857092e-01 -6.62909821e-02 -2.99925953e-01
7.41456568e-01 1.82817638e-01 1.44356132e-01 5.62850833e-01
2.11744249e-01 6.54372990e-01 1.57407746e-02 4.79685992e-01
1.08004868e+00 1.02256887e-01 -1.73999351e-02 -2.00307280e-01
4.33071077e-01 -1.65380836e-01 4.60070252e-01 9.52241540e-01
2.02831954e-01 1.98841318e-01 -1.88138828e-01 -1.39308408e-01
-5.19555151e-01 -8.09620917e-01 1.06834136e-01 1.96198177e+00
2.40422517e-01 -5.11316299e-01 -1.95690781e-01 -1.13167715e+00
2.72958189e-01 6.42930627e-01 -6.79150224e-01 -3.99716645e-01
-3.42881292e-01 -4.18515176e-01 -2.13383317e-01 6.71937168e-01
3.53473723e-01 -8.10586572e-01 3.15415591e-01 2.37686187e-01
-1.20809197e-01 -7.38672137e-01 -8.74947488e-01 1.47543192e-01
-1.13831019e+00 -6.39835596e-01 -8.48182261e-01 -6.59039199e-01
5.09330809e-01 1.02818120e+00 1.14978290e+00 -9.17898268e-02
4.47045952e-01 2.46775851e-01 -1.03193879e+00 -2.28864744e-01
1.42072618e-01 3.81276906e-01 2.57996857e-01 1.43152803e-01
7.21667707e-01 -5.86409748e-01 -8.86343896e-01 7.53678024e-01
-6.85307086e-01 -9.85361729e-03 8.31667960e-01 9.50532556e-01
3.32766622e-01 1.29071236e-01 1.05353570e+00 -1.57157886e+00
9.30484056e-01 -1.01607752e+00 -3.14817220e-01 2.47791141e-01
-8.64617229e-01 -2.40855023e-01 1.04879963e+00 -8.32288802e-01
-1.48251605e+00 -2.71189958e-01 -1.44094989e-01 -2.56988019e-01
9.71298758e-03 8.97495449e-01 -2.11805031e-01 2.76217937e-01
7.22517967e-01 2.44029805e-01 -4.72435534e-01 -1.06206357e+00
5.95055521e-01 1.07683027e+00 1.79512829e-01 -5.41603327e-01
5.68087399e-01 2.14763880e-01 -7.82880425e-01 -5.25863707e-01
-1.60154092e+00 -8.90620172e-01 -5.09586394e-01 1.51617797e-02
1.70279518e-01 -1.01707232e+00 -6.18459046e-01 9.08896923e-02
-7.34596074e-01 -1.09700084e-01 -1.48091182e-01 7.70276189e-01
-1.54484391e-01 1.75491869e-01 -7.96163201e-01 -7.07439542e-01
-7.32791364e-01 -7.29809582e-01 8.56630504e-01 4.12888855e-01
2.34516323e-01 -9.57386255e-01 5.11032753e-02 5.85492492e-01
6.78299308e-01 -6.46335244e-01 5.03867269e-01 -1.15807152e+00
-2.58630931e-01 -2.90775955e-01 -4.02592361e-01 4.63158339e-01
2.39746630e-01 -6.43068135e-01 -8.18295002e-01 -5.65377593e-01
1.43718898e-01 -2.35098854e-01 9.74403977e-01 1.76908180e-01
1.14376950e+00 -4.91996288e-01 -3.41153771e-01 4.65999454e-01
1.10601032e+00 2.58491933e-01 4.65274870e-01 7.52542466e-02
9.09944177e-01 5.50207913e-01 9.74601090e-01 5.18887043e-01
5.20862281e-01 5.44267178e-01 -2.22771764e-02 -2.41639204e-02
5.26875816e-02 -5.58018804e-01 5.94589293e-01 1.35084295e+00
6.06455766e-02 -3.82966638e-01 -1.49518549e-01 2.29680404e-01
-2.08430243e+00 -8.37242305e-01 -7.95750022e-02 2.07858729e+00
8.63030434e-01 2.43647564e-02 1.38218641e-01 -1.95888802e-01
5.40832520e-01 -2.11785167e-01 -9.71797585e-01 -8.46806634e-03
1.70239240e-01 5.47460094e-02 3.83336186e-01 -2.87649333e-02
-1.09037161e+00 8.05713654e-01 4.94875383e+00 1.04477787e+00
-9.65721667e-01 2.68788725e-01 2.12982267e-01 -2.57548064e-01
-3.51455182e-01 -2.88060129e-01 -1.16321564e+00 6.33184373e-01
9.93405700e-01 -1.27261713e-01 5.43320239e-01 1.05648327e+00
2.28494510e-01 2.89507657e-01 -1.09985185e+00 7.59960651e-01
1.67889103e-01 -1.22175384e+00 1.92848355e-01 1.17141549e-02
1.06809628e+00 3.14019948e-01 2.18331248e-01 9.11855698e-01
6.40999734e-01 -5.11199772e-01 2.49580026e-01 6.80035532e-01
1.97242171e-01 -6.29455328e-01 8.30766261e-01 5.97859383e-01
-1.09072781e+00 -3.03509921e-01 -5.87591827e-01 5.35923988e-02
1.87760834e-02 7.82161474e-01 -3.36593151e-01 7.86323607e-01
7.87310958e-01 1.47029078e+00 -3.62538278e-01 1.03021121e+00
6.69066235e-02 9.01999116e-01 -7.48459026e-02 -8.17305297e-02
1.11889899e-01 -4.96774733e-01 1.72836781e-01 1.14107478e+00
4.59032565e-01 1.71901628e-01 2.59998202e-01 5.87480962e-01
-4.25290734e-01 5.72991610e-01 -3.32866699e-01 -1.71227887e-01
3.11523438e-01 1.31387329e+00 -3.12965393e-01 -2.72761077e-01
-9.82855618e-01 9.97969329e-01 3.50475907e-01 4.86421913e-01
-5.89359701e-01 -5.04015625e-01 4.42826509e-01 1.83713615e-01
6.24240041e-01 5.06016947e-02 -1.38733715e-01 -1.58465242e+00
-1.11876518e-01 -9.78150308e-01 4.60471481e-01 -4.93585140e-01
-1.96489632e+00 4.45154309e-01 -5.15012741e-01 -1.59282827e+00
2.78806001e-01 -6.08995080e-01 -4.27917093e-01 6.99943125e-01
-1.86100078e+00 -1.18234217e+00 -1.28157839e-01 9.07069683e-01
8.24111879e-01 -3.94830644e-01 4.91697162e-01 8.60791326e-01
-4.60608810e-01 1.00010359e+00 7.16248214e-01 -2.05902406e-03
1.17115343e+00 -1.30620003e+00 1.77014187e-01 4.30507839e-01
1.73513666e-01 1.20409954e+00 3.45199108e-01 -6.97218716e-01
-1.86468995e+00 -1.32080209e+00 5.96521139e-01 -3.96666706e-01
7.29349494e-01 -3.16337734e-01 -1.01751256e+00 5.17897546e-01
7.26134554e-02 -2.21398354e-01 1.05854189e+00 7.23565876e-01
-5.58681726e-01 -3.52051765e-01 -8.49453211e-01 4.09175813e-01
1.13989258e+00 -4.03280199e-01 -6.25772834e-01 4.10175234e-01
1.04337871e+00 -2.15268377e-02 -1.12047231e+00 2.33274311e-01
7.00460553e-01 -7.04348624e-01 9.01673794e-01 -8.64063442e-01
6.26204848e-01 -2.02308565e-01 -2.06353039e-01 -1.63765621e+00
-5.92735946e-01 -4.67515916e-01 -5.50874472e-01 1.26644647e+00
6.29216909e-01 -5.55631280e-01 7.26777494e-01 5.76574683e-01
-2.84599155e-01 -6.07482135e-01 -2.24875677e-02 -5.38549483e-01
8.22217762e-02 -3.47972363e-01 6.15325153e-01 1.11510122e+00
2.73131400e-01 8.34432244e-01 -9.18914795e-01 -1.13614857e-01
3.07132006e-01 5.58435261e-01 8.56830835e-01 -1.58407605e+00
-5.59173703e-01 -1.74781740e-01 4.74570543e-01 -1.90327632e+00
-4.83793356e-02 -1.02323329e+00 -1.85597926e-01 -1.49339175e+00
1.82693839e-01 -6.42127037e-01 -8.81492019e-01 1.75137788e-01
-1.22389272e-01 1.04792938e-01 8.85441452e-02 3.56840372e-01
-1.11880505e+00 6.55787587e-01 1.66161168e+00 -1.05146691e-01
-5.24106205e-01 5.36218047e-01 -1.39899254e+00 6.67097449e-01
6.17418528e-01 -3.65404218e-01 -6.89489424e-01 -5.13359010e-01
7.45086849e-01 1.12039737e-01 -3.92708391e-01 -4.73160952e-01
4.97845650e-01 -9.92135629e-02 4.35183138e-01 -6.45768344e-01
1.14311762e-01 -1.08062077e+00 -2.83569664e-01 -1.87935010e-01
-7.10157752e-01 -3.93777698e-01 -1.60716474e-01 1.06811404e+00
-1.74534962e-01 -1.99457649e-02 3.97565365e-01 -1.18419141e-01
-4.73835945e-01 4.15288091e-01 -7.94845223e-02 -7.26530924e-02
3.17308366e-01 1.26803637e-01 -5.01133382e-01 -5.28778732e-01
-8.45490813e-01 4.08592910e-01 -1.04954336e-02 9.07598674e-01
5.28667927e-01 -1.36188412e+00 -6.43948734e-01 8.30632597e-02
1.93454117e-01 -4.38321590e-01 5.27955174e-01 1.01340270e+00
5.09141803e-01 5.67148447e-01 4.11634445e-02 -6.40464798e-02
-8.89842570e-01 8.33795369e-01 -6.27439693e-02 -2.84888029e-01
-3.95005822e-01 8.08191359e-01 3.79543245e-01 -6.54511094e-01
4.10650939e-01 -2.38923505e-01 -7.29725659e-01 2.59638786e-01
9.75071609e-01 4.40317273e-01 1.72837630e-01 -2.47088209e-01
1.64494544e-01 3.39546174e-01 -7.89833784e-01 4.06173766e-01
1.47467053e+00 -5.33112288e-01 1.82706937e-01 2.79184461e-01
1.23728585e+00 1.02765962e-01 -8.69965732e-01 -1.04150701e+00
-1.54310986e-01 -6.57345057e-01 4.90987390e-01 -1.22248316e+00
-1.42931163e+00 6.38761163e-01 3.84036362e-01 2.22599462e-01
1.19336879e+00 -2.18022153e-01 1.34296966e+00 7.19519973e-01
3.86018544e-01 -1.33793724e+00 5.63584780e-03 5.79589665e-01
7.91533113e-01 -1.48958576e+00 -9.85884666e-02 -1.67562947e-01
-9.12288189e-01 8.59841943e-01 7.51896083e-01 -2.24531338e-01
1.24819720e+00 -2.22740155e-02 -1.35359913e-01 2.48487685e-02
-1.00881827e+00 -2.58942127e-01 7.23443389e-01 3.71055752e-01
6.93747342e-01 -3.50178033e-02 -4.66176599e-01 1.54733288e+00
9.04594809e-02 3.63853067e-01 7.46989921e-02 7.50407398e-01
-5.49300373e-01 -1.19184864e+00 1.32385433e-01 1.15524232e+00
-5.02098441e-01 -3.28087389e-01 -3.37316483e-01 1.01423897e-01
1.04295649e-02 1.25712287e+00 -3.51098299e-01 -7.10372448e-01
3.94521713e-01 -2.34211788e-01 1.11036129e-01 -7.75190651e-01
-9.43085253e-01 4.51322168e-01 8.75001494e-03 -4.81536746e-01
-4.52641875e-01 -4.51993734e-01 -8.09944093e-01 -2.13742375e-01
-9.25680578e-01 6.01211369e-01 5.63173056e-01 9.77654874e-01
7.05565631e-01 5.69863260e-01 9.58057761e-01 -6.86722636e-01
-5.85816324e-01 -1.07764983e+00 -8.74344945e-01 3.44973624e-01
1.18550912e-01 -7.04306304e-01 -3.08201015e-01 -8.74180049e-02] | [10.164776802062988, 5.610990524291992] |
9658362e-8212-4de5-a018-e0f2d09d5ca9 | mutual-gaze-and-linguistic-repetition-in-a | null | null | https://aclanthology.org/2022.lrec-1.296 | https://aclanthology.org/2022.lrec-1.296.pdf | Mutual Gaze and Linguistic Repetition in a Multimodal Corpus | This paper investigates the correlation between mutual gaze and linguistic repetition, a form of alignment, which we take as evidence of mutual understanding. We focus on a multimodal corpus made of three-party conversations and explore the question of whether mutual gaze events correspond to moments of repetition or non-repetition. Our results, although mainly significant on word unigrams and bigrams, suggest positive correlations between the presence of mutual gaze and the repetitions of tokens, lemmas, or parts-of-speech, but negative correlations when it comes to paired levels of representation (tokens or lemmas associated with their part-of-speech). No compelling correlation is found with duration of mutual gaze. Results are strongest when ignoring punctuation as representations of pauses, intonation, etc. in counting aligned tokens. | ['Carl Vogel', 'Maria Koutsombogera', 'Anais Murat'] | null | null | null | null | lrec-2022-6 | ['mutual-gaze'] | ['computer-vision'] | [ 1.14271127e-01 1.20085344e-01 -3.56708288e-01 -2.18589872e-01
-3.97558540e-01 -8.59865427e-01 1.26245534e+00 5.21872520e-01
-5.20432353e-01 4.56632406e-01 9.58057046e-01 -3.83454949e-01
3.58810154e-04 -3.53835195e-01 -3.04834753e-01 -5.86104214e-01
-1.52022541e-01 5.73572926e-02 -1.70894712e-01 -5.73874056e-01
7.38533854e-01 9.58020985e-02 -1.59354353e+00 3.65715504e-01
3.31590503e-01 2.42176540e-02 5.06442934e-02 6.85362220e-01
-7.29573429e-01 8.16557109e-01 -8.80189717e-01 -4.34908658e-01
-3.65349352e-01 -8.14477682e-01 -8.23905289e-01 1.47972897e-01
2.93259531e-01 2.83499211e-01 4.58707623e-02 8.93029630e-01
1.95739269e-01 3.70688550e-02 5.28117001e-01 -9.92919803e-01
-4.06829298e-01 1.01803780e+00 -8.76849711e-01 7.62273371e-01
9.11829293e-01 5.63267954e-02 1.25006425e+00 -5.99249721e-01
6.98904097e-01 1.28250420e+00 5.36719382e-01 4.72847700e-01
-1.07799554e+00 -4.19669658e-01 -1.86129034e-01 -3.34912121e-01
-1.15639913e+00 -7.84541368e-01 5.78405142e-01 -5.62086225e-01
1.06753564e+00 5.15817881e-01 4.39444542e-01 9.92707729e-01
2.36180797e-01 3.59395146e-01 1.17584753e+00 -8.98558199e-01
-3.42716813e-01 2.13325590e-01 6.02331936e-01 3.25802416e-01
-6.09618425e-02 -3.13015543e-02 -7.73744822e-01 -1.34692162e-01
4.14730310e-01 -1.89044341e-01 -3.10295463e-01 7.71321893e-01
-1.52211905e+00 7.05035269e-01 -9.16320458e-02 1.20090139e+00
-3.19608212e-01 8.16280618e-02 4.29832369e-01 2.50625789e-01
4.19757605e-01 3.07501405e-01 -1.11174211e-01 -7.48671055e-01
-7.46248603e-01 -1.08683603e-02 7.64608920e-01 6.89640164e-01
8.74514818e-01 -4.69341040e-01 -7.62442127e-02 8.19252968e-01
2.34749660e-01 3.45507264e-01 6.99853361e-01 -7.16995299e-01
4.12427336e-01 5.16588211e-01 4.05146107e-02 -7.99452364e-01
-4.28326875e-01 1.64252177e-01 -8.09527859e-02 -3.89911324e-01
4.54611957e-01 -2.35417277e-01 -4.17421162e-01 2.01265621e+00
1.42503902e-01 -2.55825520e-01 1.28381640e-01 4.45512116e-01
9.62892532e-01 7.29596019e-01 4.10925239e-01 -8.61576736e-01
1.76482654e+00 -4.03506666e-01 -1.16444087e+00 -1.68445349e-01
8.74113739e-01 -1.16820502e+00 1.11803615e+00 -2.52559751e-01
-1.08355463e+00 -4.75545228e-01 -4.91020292e-01 -1.94668010e-01
-3.80993843e-01 -3.93841892e-01 1.52266413e-01 6.42782569e-01
-7.65899181e-01 6.26554906e-01 -4.35873389e-01 -3.71714473e-01
-5.00878274e-01 4.73727891e-03 -3.94094110e-01 6.14401639e-01
-9.40545022e-01 9.15272295e-01 -2.31943615e-02 4.14640419e-02
2.37540454e-01 -7.32889399e-02 -8.62855375e-01 -2.93070644e-01
6.65185526e-02 4.78877723e-02 1.20006561e+00 -1.35373604e+00
-1.19256008e+00 1.44633389e+00 -7.84634054e-01 9.54066664e-02
-1.61921412e-01 -1.31054550e-01 -9.08893645e-01 -2.79175881e-02
1.87667966e-01 3.95872563e-01 5.63964367e-01 -1.03758538e+00
-5.38322568e-01 -2.51536399e-01 9.12106410e-02 1.84884936e-01
2.38352343e-02 7.51577854e-01 1.40524581e-01 -3.74908924e-01
1.64215311e-01 -8.25536609e-01 2.91651994e-01 -8.56255352e-01
-4.32015955e-01 -8.71446550e-01 6.42350972e-01 -5.49167275e-01
1.60329247e+00 -2.45330119e+00 -1.16953328e-01 2.58118026e-02
2.57266968e-01 -1.21251188e-01 -1.17689446e-01 1.01145399e+00
-5.68748236e-01 5.26592433e-01 2.85943486e-02 -3.23497295e-01
1.52312577e-01 3.61576408e-01 -3.14178824e-01 5.67547917e-01
1.42520204e-01 9.51642394e-01 -8.10020149e-01 -6.00866854e-01
-4.27992195e-02 4.43555623e-01 5.66703863e-02 -6.34209020e-03
5.00584580e-02 2.32100859e-01 2.17049774e-02 3.80349666e-01
5.56007139e-02 -2.01421812e-01 3.37499946e-01 -6.11349903e-02
-8.84648561e-01 1.08177006e+00 -4.87748742e-01 1.00833619e+00
-1.51282847e-01 1.00429893e+00 -3.27852279e-01 -3.76422256e-01
7.23825991e-01 4.54349369e-01 -1.06925942e-01 -7.91908383e-01
4.13903773e-01 4.08294238e-02 6.46308601e-01 -6.38298631e-01
9.57047164e-01 -5.22529006e-01 -4.83277053e-01 8.47185135e-01
-3.25569600e-01 7.19887912e-02 5.02937198e-01 2.54345953e-01
7.78991103e-01 -2.19103083e-01 7.65995979e-01 -3.01845461e-01
4.28539217e-01 -3.86025012e-01 1.15781859e-01 3.25658590e-01
1.41873304e-02 3.65069628e-01 9.54323411e-01 -3.33149219e-03
-8.10264230e-01 -8.51332068e-01 -1.83161721e-01 1.23961186e+00
1.72452971e-01 -8.40811312e-01 -5.29532611e-01 -3.98207545e-01
-3.57792050e-01 1.06713998e+00 -8.88270319e-01 2.44432405e-01
-7.79016733e-01 -4.41776335e-01 5.10941565e-01 1.18639499e-01
-1.56738684e-01 -1.44111729e+00 -8.41973424e-01 -1.72810834e-02
-3.70275676e-01 -1.00758123e+00 -4.94515717e-01 1.72244981e-01
-6.02137208e-01 -8.80793989e-01 -5.35243571e-01 -5.92841208e-01
1.91578686e-01 2.10807875e-01 1.07276571e+00 3.67212057e-01
-7.44416611e-03 5.56899130e-01 -5.48887312e-01 -2.31106922e-01
-7.97395170e-01 -3.38167071e-01 -9.36881006e-02 -3.95197183e-01
5.02474070e-01 -4.69883889e-01 -1.33478582e-01 1.45928547e-01
-7.03011572e-01 -2.79665083e-01 1.56916454e-01 2.87945628e-01
7.30443895e-02 -6.81897879e-01 1.63135201e-01 -8.15380991e-01
9.35317397e-01 -6.09549403e-01 1.37791216e-01 1.87263399e-01
-8.27279612e-02 -7.28826076e-02 -4.13450003e-02 -4.94265199e-01
-6.97816491e-01 -7.80440390e-01 -5.10989614e-02 3.21595594e-02
-4.61131334e-01 5.32228410e-01 4.27625537e-01 2.84040719e-01
5.18738270e-01 1.47606626e-01 1.71796605e-01 -1.85930416e-01
3.83883476e-01 7.60855198e-01 4.56753403e-01 -2.80631781e-01
3.99598807e-01 1.31001309e-01 -2.85672665e-01 -1.40751052e+00
-7.08905935e-01 -7.56757677e-01 -5.42947292e-01 -3.63501549e-01
9.27132547e-01 -7.15298235e-01 -7.90748537e-01 2.46116877e-01
-1.19900894e+00 -1.99530467e-01 -3.01520139e-01 4.94062603e-01
-3.69045466e-01 8.13439965e-01 -7.13772357e-01 -9.56264079e-01
-3.35826315e-02 -7.07109988e-01 7.35296309e-01 4.94432002e-01
-1.10897803e+00 -1.04433298e+00 4.74138767e-01 8.54410455e-02
1.55365150e-02 2.26050287e-01 8.29796970e-01 -8.69589329e-01
1.08742438e-01 -3.91125344e-02 1.43416375e-01 -1.27732769e-01
6.85102999e-01 5.16720653e-01 -6.22022688e-01 1.82556704e-01
-8.14690813e-03 -1.34714097e-01 4.50490117e-01 8.19690749e-02
1.35489628e-01 -5.64470768e-01 -2.27755442e-01 -2.02251688e-01
9.47968960e-01 1.44587815e-01 6.69775009e-01 1.46412596e-01
3.48621666e-01 9.92603362e-01 4.98044521e-01 3.50801110e-01
3.24919164e-01 5.05064547e-01 1.06836911e-02 3.90541732e-01
-1.47469819e-01 -6.03583492e-02 5.77245116e-01 1.10409057e+00
-1.62271306e-01 -5.54516971e-01 -1.10720396e+00 8.57876182e-01
-1.24672842e+00 -1.30819345e+00 -8.75250876e-01 2.07093406e+00
6.76445365e-01 9.14728567e-02 4.14591342e-01 2.66877025e-01
1.08455455e+00 6.65761232e-01 3.93381298e-01 -8.02895129e-01
-2.13422820e-01 1.97822690e-01 3.93992662e-02 7.19934404e-01
-5.37523508e-01 8.94528925e-01 7.32908916e+00 3.42147559e-01
-1.22387195e+00 -4.21422236e-02 3.06981564e-01 -3.76673974e-02
-5.52938640e-01 4.71548960e-02 -6.66665018e-01 6.91732168e-01
1.19622397e+00 1.37378260e-01 2.29016751e-01 1.42002478e-01
1.53266057e-01 -6.90833867e-01 -1.09377551e+00 6.16342723e-01
2.28086323e-01 -9.31666493e-01 -2.56265402e-01 6.55917376e-02
3.65438581e-01 -4.49349359e-02 6.33505955e-02 -4.91266660e-02
-1.27745733e-01 -8.39705288e-01 9.08232212e-01 2.39938125e-01
5.38213730e-01 -5.90434492e-01 4.73731816e-01 3.26715767e-01
-8.39080632e-01 3.10155660e-01 2.90254336e-02 -4.40076619e-01
4.78451312e-01 1.94163069e-01 -7.38405347e-01 1.87998593e-01
3.63458693e-01 3.80601436e-01 -3.16419929e-01 5.33694327e-01
-4.99149948e-01 7.68016160e-01 -4.14655477e-01 -3.37147295e-01
4.06535178e-01 -1.63139686e-01 7.52724469e-01 1.56284714e+00
9.35332850e-02 3.90398145e-01 -4.10152107e-01 4.53750789e-01
3.25235426e-01 2.63912410e-01 -4.76597160e-01 -6.01673484e-01
5.69948256e-01 1.14263976e+00 -1.01228559e+00 -1.90259963e-01
-4.40956056e-01 7.50433147e-01 3.21867466e-01 2.86214594e-02
-6.73524499e-01 -3.04203093e-01 5.85268319e-01 1.47366941e-01
3.28785181e-01 -3.73094201e-01 -2.37692490e-01 -7.19416201e-01
-1.39172137e-01 -7.07595289e-01 7.69676045e-02 -5.28855145e-01
-1.26077974e+00 5.59331775e-01 1.01851925e-01 -7.43019879e-01
-6.62680745e-01 -5.62901981e-02 -1.07328320e+00 9.75620210e-01
-1.10460389e+00 -6.70573652e-01 7.54767209e-02 4.43908513e-01
1.65505931e-01 3.11864585e-01 8.38377774e-01 -4.97508608e-02
-3.54351729e-01 5.81644177e-01 -5.13156533e-01 2.18194887e-01
7.34046340e-01 -1.02756488e+00 3.91496986e-01 5.40865719e-01
5.57747662e-01 9.32690978e-01 1.12459075e+00 -5.47461510e-01
-7.86140978e-01 -2.19896529e-02 1.88092828e+00 -5.11435926e-01
9.68195856e-01 6.24311008e-02 -8.73065591e-01 7.24929154e-01
8.30629587e-01 -5.16000986e-01 1.26526463e+00 5.76166570e-01
-3.68417829e-01 6.03925049e-01 -6.84935629e-01 6.25872433e-01
9.20680583e-01 -1.00967503e+00 -1.31030715e+00 1.95501208e-01
7.60504305e-01 -2.77563661e-01 -6.55904472e-01 1.66319430e-01
4.83968705e-01 -1.23717880e+00 4.43240196e-01 -4.15759206e-01
5.71706533e-01 -3.47380824e-02 -7.29585662e-02 -8.16141248e-01
-1.99931711e-01 -9.53228533e-01 5.48316717e-01 1.62484372e+00
5.23224652e-01 -4.70347345e-01 1.53447598e-01 3.99130046e-01
-2.24007115e-01 -8.94543529e-02 -9.39838052e-01 -4.70491081e-01
-1.48000166e-01 -4.02237564e-01 2.43243515e-01 1.41003227e+00
8.58114123e-01 8.01591873e-01 -9.20502692e-02 -3.33809376e-01
5.59133142e-02 2.82195378e-02 5.72492361e-01 -9.76340592e-01
-1.45601928e-01 -6.71774685e-01 -3.04228902e-01 -7.45992124e-01
2.51878917e-01 -4.29385811e-01 1.00648277e-01 -1.03997350e+00
3.94894890e-02 -2.05068558e-01 1.20893747e-01 4.46237773e-01
-1.57968402e-01 2.94971734e-01 4.35890049e-01 4.08474803e-01
-4.14893568e-01 1.63395435e-01 1.00121486e+00 4.36711669e-01
-6.03169501e-01 -1.58064783e-01 -6.37498021e-01 8.00742447e-01
7.09220588e-01 -4.64471370e-01 -2.06051636e-02 -5.44301346e-02
5.09733617e-01 4.37729180e-01 2.32160792e-01 -5.67354321e-01
7.22762793e-02 -3.35597619e-02 -2.84672558e-01 -6.46784604e-01
3.10907692e-01 -4.40691888e-01 2.49091372e-01 2.45237544e-01
-4.51072454e-01 7.84664035e-01 4.09479499e-01 2.94787228e-01
-2.72754729e-01 -5.20178318e-01 3.95293236e-01 -7.99061954e-02
-3.13428044e-01 -6.68009996e-01 -8.70909870e-01 1.10305712e-01
1.01776147e+00 -5.85941076e-01 -3.29506397e-01 -3.88825715e-01
-9.06203449e-01 -2.05822840e-01 5.15490890e-01 5.33606052e-01
1.49798363e-01 -1.01963413e+00 -5.72871149e-01 -1.30181804e-01
7.29028210e-02 -5.73892891e-01 8.45678803e-03 1.11638534e+00
-1.48412809e-01 2.56881893e-01 5.82186319e-03 -3.86754930e-01
-1.77321732e+00 4.72981602e-01 -1.10777579e-01 -5.97045459e-02
-2.17505246e-01 7.08038747e-01 9.78485271e-02 9.23644304e-02
-7.70154446e-02 -3.33719283e-01 -3.38097543e-01 7.16336012e-01
6.51167452e-01 2.58326709e-01 -4.15386260e-01 -1.28167307e+00
-5.19485056e-01 5.37008524e-01 -4.88605984e-02 -4.61991310e-01
8.81355405e-01 -5.48314512e-01 -6.69508338e-01 1.48460841e+00
1.28971243e+00 7.68553019e-01 -6.04221880e-01 -3.59551497e-02
2.56409258e-01 -1.12403095e-01 -5.62504172e-01 -4.22462851e-01
-3.17294896e-01 7.26301908e-01 4.85009439e-02 7.31892824e-01
3.84249806e-01 5.96851110e-01 3.80460292e-01 -1.42084286e-01
-9.75187868e-02 -7.12925792e-01 9.14603025e-02 8.15799475e-01
6.39814615e-01 -8.07660818e-01 -1.33667096e-01 -3.50254923e-01
-7.47783422e-01 1.16049469e+00 3.53866160e-01 -6.37504831e-02
4.86028075e-01 1.30217880e-01 1.16368197e-01 -5.00761271e-01
-7.14183688e-01 -2.74475724e-01 2.80844152e-01 1.25230625e-01
1.22501159e+00 9.63277444e-02 -1.17981017e+00 1.76624488e-02
-8.13520968e-01 -6.54878795e-01 7.03730047e-01 7.95735717e-01
-5.58224022e-01 -1.01575196e+00 -3.96478504e-01 1.26371115e-01
-7.40168929e-01 -3.88951182e-01 -7.09581912e-01 1.21130967e+00
2.20845208e-01 1.04878330e+00 8.75390053e-01 -3.93992066e-01
-1.49044305e-01 3.85292888e-01 6.32534206e-01 -6.73537314e-01
-1.09419179e+00 2.52117515e-01 5.65872669e-01 -1.73051476e-01
-9.92494583e-01 -1.11107278e+00 -1.21169758e+00 -5.46984017e-01
-4.47311789e-01 3.85517061e-01 5.17292976e-01 1.21317828e+00
-1.97238222e-01 1.10389903e-01 2.88770556e-01 -6.52046680e-01
1.75031442e-02 -1.17499530e+00 -5.55167198e-01 2.68074721e-01
5.82057595e-01 -1.76473334e-01 -6.24312758e-01 -8.79644901e-02] | [10.317985534667969, 9.33016586303711] |
e23048e7-d32c-4812-8f30-3ae6be8be847 | triple-structural-information-modelling-for | 2304.11528 | null | https://arxiv.org/abs/2304.11528v1 | https://arxiv.org/pdf/2304.11528v1.pdf | Triple Structural Information Modelling for Accurate, Explainable and Interactive Recommendation | In dynamic interaction graphs, user-item interactions usually follow heterogeneous patterns, represented by different structural information, such as user-item co-occurrence, sequential information of user interactions and the transition probabilities of item pairs. However, the existing methods cannot simultaneously leverage all three structural information, resulting in suboptimal performance. To this end, we propose TriSIM4Rec, a triple structural information modeling method for accurate, explainable and interactive recommendation on dynamic interaction graphs. Specifically, TriSIM4Rec consists of 1) a dynamic ideal low-pass graph filter to dynamically mine co-occurrence information in user-item interactions, which is implemented by incremental singular value decomposition (SVD); 2) a parameter-free attention module to capture sequential information of user interactions effectively and efficiently; and 3) an item transition matrix to store the transition probabilities of item pairs. Then, we fuse the predictions from the triple structural information sources to obtain the final recommendation results. By analyzing the relationship between the SVD-based and the recently emerging graph signal processing (GSP)-based collaborative filtering methods, we find that the essence of SVD is an ideal low-pass graph filter, so that the interest vector space in TriSIM4Rec can be extended to achieve explainable and interactive recommendation, making it possible for users to actively break through the information cocoons. Experiments on six public datasets demonstrated the effectiveness of TriSIM4Rec in accuracy, explainability and interactivity. | ['Ning Gu', 'Li Shang', 'Peng Zhang', 'Tun Lu', 'Hansu Gu', 'Dongsheng Li', 'Jiahao Liu'] | 2023-04-23 | null | null | null | null | ['collaborative-filtering'] | ['miscellaneous'] | [-1.69292971e-01 -8.52094442e-02 -3.73985976e-01 -1.54152915e-01
1.99396595e-01 -5.08631110e-01 1.62999868e-01 2.26705328e-01
3.11577260e-01 2.11805999e-01 7.58806348e-01 -3.75060916e-01
-5.84662676e-01 -8.36839676e-01 -5.06429195e-01 -2.47859821e-01
-2.68578053e-01 2.23060846e-01 1.58930514e-02 -5.13679266e-01
-8.02795663e-02 4.50087041e-02 -1.20908940e+00 6.38895273e-01
1.23575652e+00 8.84778380e-01 2.01160938e-01 5.89974046e-01
-3.46375287e-01 7.38691568e-01 2.49946639e-02 -6.10218525e-01
8.37277919e-02 -6.21092677e-01 -4.73765463e-01 1.28885537e-01
-1.81134894e-01 -2.08766431e-01 -7.76910484e-01 9.55784440e-01
3.09817761e-01 4.44684684e-01 4.26533252e-01 -1.19213533e+00
-1.05354297e+00 1.30172420e+00 -4.71900523e-01 1.89378336e-01
8.49258840e-01 1.19815409e-01 1.45624185e+00 -1.00809050e+00
5.33784091e-01 1.34500372e+00 6.61955118e-01 -8.58161151e-02
-1.16466212e+00 -5.48175693e-01 7.87131906e-01 5.08846283e-01
-1.27513361e+00 7.44507983e-02 9.75333989e-01 -4.22990143e-01
8.79988074e-01 5.78813732e-01 1.19360209e+00 1.09650731e+00
2.48764306e-01 9.00243521e-01 3.21867585e-01 1.48801999e-02
-1.77461505e-01 -1.06478773e-01 6.87715352e-01 7.19845533e-01
2.55770802e-01 -1.15977637e-01 -5.38490057e-01 -4.02093232e-01
7.98095942e-01 6.58228457e-01 -4.40757453e-01 -6.01504371e-02
-1.14310491e+00 6.98167920e-01 7.71793604e-01 3.44465226e-01
-6.30095482e-01 -3.41965616e-01 1.66387826e-01 5.10342002e-01
3.86002421e-01 2.84538418e-01 -4.44851577e-01 2.11952657e-01
-3.73470962e-01 1.48340642e-01 9.07052696e-01 1.03151453e+00
5.93254268e-01 5.04336599e-03 -5.14855027e-01 6.04669094e-01
7.97492743e-01 2.92185217e-01 5.16520023e-01 -3.44401538e-01
4.80026454e-01 1.03048348e+00 -1.50834145e-02 -1.87125707e+00
-4.90462571e-01 -8.18975747e-01 -1.15966308e+00 -8.33723605e-01
5.82580380e-02 -2.35338777e-01 -5.12812018e-01 1.51742959e+00
4.09610868e-01 7.14169085e-01 -3.29895824e-01 1.04593039e+00
1.15261567e+00 6.69215620e-01 -1.74413100e-01 -5.21082044e-01
1.35344636e+00 -9.97188568e-01 -9.61248159e-01 3.04000266e-02
4.46579188e-01 -4.92983162e-01 1.08312190e+00 3.81449163e-01
-7.26301432e-01 -7.49401987e-01 -9.50496256e-01 1.74464941e-01
4.31228392e-02 -6.40549883e-02 9.45265889e-01 3.37194264e-01
-4.92412329e-01 6.94378614e-01 -5.70116878e-01 -9.80148688e-02
1.75409406e-01 3.66312832e-01 -3.93656902e-02 -1.38637364e-01
-1.46227968e+00 5.86929061e-02 2.10708976e-01 2.94100612e-01
-2.37841740e-01 -7.18618214e-01 -6.65419996e-01 3.50426108e-01
6.54540002e-01 -8.73074651e-01 6.91166520e-01 -9.62083280e-01
-1.31457222e+00 -1.88767716e-01 -2.09589005e-01 -3.43326956e-01
1.47520855e-01 -1.48249269e-01 -8.39416802e-01 -3.53596747e-01
-3.52607310e-01 -4.27376151e-01 7.75031447e-01 -1.07082641e+00
-5.09268880e-01 -3.65674168e-01 2.14990914e-01 4.15691942e-01
-4.74688143e-01 -3.85558456e-01 -9.83881474e-01 -7.97284544e-01
2.34904215e-01 -9.89288986e-01 -3.20540011e-01 -7.35301256e-01
-6.54298306e-01 -2.19969511e-01 4.46383834e-01 -9.03710306e-01
1.96668446e+00 -2.12392616e+00 4.26300168e-01 7.50404000e-01
7.79394686e-01 2.62093276e-01 -3.31496835e-01 6.99226201e-01
-1.98950730e-02 9.66848210e-02 4.27965462e-01 -1.36428893e-01
-8.58546495e-02 1.13347024e-01 -2.19574988e-01 1.64746821e-01
-4.25215453e-01 1.22451293e+00 -1.07668352e+00 -1.35522872e-01
2.00975806e-01 5.20893037e-01 -8.81887913e-01 3.70220751e-01
-1.03433967e-01 5.15129268e-01 -7.15855539e-01 1.98506549e-01
5.85102320e-01 -6.47816598e-01 6.24706805e-01 -7.22634733e-01
2.96942979e-01 2.59919584e-01 -1.62054312e+00 1.37075543e+00
-2.98099279e-01 3.96323986e-02 1.85454655e-02 -7.56328821e-01
7.62994885e-01 1.96277454e-01 7.09114909e-01 -4.64282125e-01
9.24943611e-02 -2.37237394e-01 1.49306819e-01 -4.13024426e-01
4.25847054e-01 5.69968283e-01 1.48323655e-01 4.28765416e-01
1.18289120e-01 6.17080450e-01 1.82667673e-01 7.74848282e-01
1.17583990e+00 -4.38109934e-01 2.96397746e-01 -3.28242779e-02
5.78117132e-01 -5.07186890e-01 6.46454334e-01 7.90167630e-01
2.46401265e-01 3.65464061e-01 3.66442025e-01 -3.93028587e-01
-4.79015589e-01 -9.08316553e-01 3.64326358e-01 1.14516032e+00
5.13030827e-01 -1.15752435e+00 -3.79585475e-01 -8.41269135e-01
2.35789329e-01 5.50741673e-01 -6.52687788e-01 -4.83277380e-01
-2.42717952e-01 -5.17119229e-01 -2.36148179e-01 2.71172136e-01
1.05614409e-01 -8.89860272e-01 6.12622201e-01 5.08892775e-01
-4.04533505e-01 -7.44305313e-01 -1.22009528e+00 -3.19490373e-01
-6.75610960e-01 -1.12350571e+00 -1.00113057e-01 -4.22495931e-01
7.16939807e-01 9.52392578e-01 1.07722068e+00 5.52575588e-01
1.48936167e-01 6.77001655e-01 -8.31450999e-01 -6.72287494e-02
4.51898240e-02 -2.13820517e-01 2.00770214e-01 7.35957682e-01
1.64310798e-01 -8.95714939e-01 -9.23780799e-01 5.20342171e-01
-7.18507528e-01 3.89309049e-01 4.44826335e-01 8.53810787e-01
7.04865634e-01 1.75583765e-01 4.48541164e-01 -1.27675605e+00
1.03825915e+00 -9.38308597e-01 -3.18376757e-02 3.34210038e-01
-9.24332321e-01 -1.80145472e-01 9.47379827e-01 -6.88640118e-01
-9.80819821e-01 -1.92135334e-01 -1.01598792e-01 -4.51895624e-01
3.66668880e-01 1.17729414e+00 -3.47821474e-01 9.04073492e-02
5.17386258e-01 2.82454967e-01 -2.63788551e-01 -6.40260041e-01
6.15887046e-01 5.04862070e-01 1.64387748e-01 -7.07858801e-02
7.32909679e-01 5.21568693e-02 -3.80472392e-01 -5.79445124e-01
-9.34257269e-01 -5.98290503e-01 -3.73549938e-01 -2.93977112e-01
4.31088954e-01 -8.69338334e-01 -1.03600001e+00 1.72655106e-01
-9.50510800e-01 1.29936084e-01 -1.49177521e-01 7.24553704e-01
3.05815395e-02 7.01941371e-01 -8.83815765e-01 -5.98449349e-01
-4.85508829e-01 -8.65717053e-01 4.54600781e-01 3.05386245e-01
-1.12894177e-01 -1.09903395e+00 -5.70272841e-02 4.86357480e-01
2.06886992e-01 -1.42016560e-01 7.96845317e-01 -8.92011523e-01
-5.36862910e-01 -2.42318302e-01 -2.08897218e-01 -2.10149176e-02
2.46280462e-01 -6.15023375e-02 -2.07138777e-01 -3.67835999e-01
-2.82179832e-01 3.77310008e-01 6.36486948e-01 4.84300345e-01
1.15367329e+00 -7.63199329e-01 -4.73103493e-01 5.67791045e-01
9.87735271e-01 1.62640125e-01 4.27827924e-01 -4.64238018e-01
1.40940905e+00 2.65348554e-01 4.90443617e-01 7.30093718e-01
7.85904408e-01 7.20202088e-01 3.68322700e-01 4.78115864e-02
-9.25479308e-02 -7.98051953e-01 3.04219514e-01 1.47647166e+00
-4.04121608e-01 -5.52825272e-01 -4.33092386e-01 1.83131218e-01
-2.44811249e+00 -1.01963735e+00 -5.98159850e-01 2.12712073e+00
3.98075581e-01 -5.95220104e-02 3.01205248e-01 -1.30214632e-01
7.65275240e-01 9.08057839e-02 -6.71883285e-01 -4.18920862e-03
1.37317881e-01 -2.99805909e-01 1.45577982e-01 4.23032612e-01
-7.74579167e-01 6.30672157e-01 5.08678532e+00 8.54788721e-01
-5.59877574e-01 4.67104092e-03 1.75438970e-01 -4.48483452e-02
-8.08046520e-01 -4.43313085e-02 -5.28452396e-01 7.71542966e-01
6.21754587e-01 -5.24015129e-01 1.00296426e+00 6.17169082e-01
4.28783387e-01 5.42615652e-01 -7.95663953e-01 1.21813202e+00
1.50729873e-04 -1.38478947e+00 3.33497047e-01 3.92279029e-02
6.60119236e-01 -1.77055746e-01 -2.34440807e-02 4.93230790e-01
6.28961384e-01 -6.97354198e-01 2.42403671e-01 7.87448704e-01
1.48763880e-01 -8.03332567e-01 6.14175856e-01 3.90650183e-01
-1.82615924e+00 -3.45533967e-01 -3.17082465e-01 -4.88859452e-02
2.37772241e-01 9.19669509e-01 -4.70077097e-01 1.21960139e+00
5.26114285e-01 1.44822729e+00 -3.19396436e-01 1.00833678e+00
-2.25304604e-01 9.07024086e-01 -2.16299623e-01 -1.21922910e-01
-1.44710809e-01 -7.01742530e-01 7.54050314e-01 1.03261089e+00
3.49139869e-01 6.78047121e-01 5.56890368e-01 5.35802066e-01
-1.22332655e-01 5.50006628e-01 -4.67077255e-01 -2.00867429e-01
5.21566331e-01 1.33992779e+00 -5.09217381e-01 -1.27158061e-01
-6.64267719e-01 9.83038962e-01 4.37562943e-01 5.35987496e-01
-8.62099290e-01 -1.83745906e-01 6.30342245e-01 2.05007881e-01
4.47874188e-01 -2.97213495e-01 -6.04553930e-02 -1.69392407e+00
-1.17938682e-01 -1.07736313e+00 6.29394114e-01 -4.93917853e-01
-1.58248591e+00 4.91957575e-01 -4.10235196e-01 -1.29134822e+00
4.74030524e-02 -7.40148202e-02 -8.12677145e-01 6.74386203e-01
-7.32641697e-01 -1.27095175e+00 -5.32600462e-01 9.98090029e-01
4.57318962e-01 -1.57047331e-01 5.35902798e-01 4.40711647e-01
-8.51809382e-01 8.03418636e-01 1.16357267e-01 -1.23322187e-02
3.33811730e-01 -1.01049137e+00 4.10004973e-01 7.62035847e-01
8.50097477e-01 9.09729302e-01 5.69779992e-01 -9.51113999e-01
-2.09581470e+00 -1.12232590e+00 4.53416079e-01 -3.48697275e-01
7.67525256e-01 -3.06726396e-01 -1.13568258e+00 8.74907911e-01
-1.40180811e-01 -1.65087149e-01 9.08616900e-01 7.91040659e-01
-2.37944871e-01 -7.86327645e-02 -6.35775387e-01 8.44222188e-01
1.61056519e+00 -4.14972037e-01 -4.59232092e-01 5.26640892e-01
1.05253255e+00 -3.16092491e-01 -1.09162951e+00 2.59546429e-01
6.08355224e-01 -8.28490257e-01 1.09487474e+00 -8.89865398e-01
1.10004283e-01 -2.54906446e-01 1.62060976e-01 -1.56272495e+00
-1.11226070e+00 -9.22362745e-01 -7.53451645e-01 1.23459005e+00
5.44637263e-01 -5.82092047e-01 5.73561013e-01 7.03634441e-01
-1.22352757e-01 -7.95717478e-01 -3.05166006e-01 -2.89741278e-01
-9.20382559e-01 -5.76837361e-01 8.06234658e-01 1.06154108e+00
3.96856844e-01 8.32121670e-01 -9.96442258e-01 2.22406343e-01
4.25024927e-01 4.97546583e-01 8.71505439e-01 -1.47460175e+00
-9.05261517e-01 -1.27333865e-01 -1.78226456e-01 -1.42961800e+00
-1.01426318e-01 -1.19853258e+00 -4.69351709e-01 -1.80073476e+00
2.93649703e-01 -2.66868919e-01 -6.12801373e-01 2.99271852e-01
-5.21255851e-01 -1.82917818e-01 1.67666584e-01 4.89634454e-01
-8.67492914e-01 6.34178102e-01 1.62275195e+00 -8.77918601e-02
-7.97466218e-01 3.03218395e-01 -1.12417293e+00 5.40988028e-01
2.30546832e-01 -1.96836889e-01 -8.93852592e-01 -9.19253752e-02
5.60762584e-01 3.14844608e-01 -1.24864615e-01 -5.16870499e-01
3.62095654e-01 -3.51176083e-01 3.15553188e-01 -3.84359449e-01
-5.00099137e-02 -8.86473536e-01 6.35839939e-01 2.68958241e-01
-2.74853349e-01 -2.29969531e-01 -1.70989543e-01 1.34704256e+00
2.07135268e-02 3.90571773e-01 2.01294012e-02 1.78507969e-01
-5.89456141e-01 9.68276262e-01 -1.92984506e-01 -3.08820188e-01
7.54825354e-01 -8.75476655e-03 -4.69547026e-02 -6.57213986e-01
-1.18306053e+00 5.11205614e-01 -4.59256163e-03 7.20012486e-01
6.99524581e-01 -1.58532178e+00 -6.92254961e-01 3.59099299e-01
-9.38301999e-03 -4.96432662e-01 8.32491934e-01 9.01008844e-01
8.34684670e-02 6.20115735e-02 1.35657415e-01 -2.49380961e-01
-1.29096437e+00 8.84112775e-01 -2.17250530e-02 -5.70344925e-01
-7.78282821e-01 7.54579961e-01 1.83377996e-01 -3.61187100e-01
3.55788618e-02 -1.17849335e-01 -8.21897328e-01 9.91096273e-02
6.23994470e-01 2.53290325e-01 -1.36806384e-01 -4.99341965e-01
-1.27126008e-01 4.16003168e-01 -2.70650148e-01 4.75828558e-01
1.30836380e+00 -5.87253988e-01 -8.48438293e-02 3.52033317e-01
9.69962060e-01 3.43591690e-01 -8.93400192e-01 -5.44331074e-01
-4.53045219e-01 -6.93847418e-01 1.47849426e-01 -5.56475043e-01
-1.29601085e+00 2.67771661e-01 1.56124949e-01 7.26954043e-01
1.10318863e+00 -1.74071744e-01 1.08172476e+00 2.86473602e-01
4.93604928e-01 -6.07849777e-01 4.04046141e-02 5.30957341e-01
9.27015424e-01 -9.56609666e-01 8.85597467e-02 -8.26343477e-01
-8.41264248e-01 7.58015275e-01 4.96848702e-01 -1.53252482e-01
1.27442396e+00 -2.10945413e-01 -5.32450140e-01 -2.44875208e-01
-8.82968247e-01 -9.38756317e-02 1.00944149e+00 4.08747494e-01
4.00245816e-01 2.20739663e-01 -6.02528751e-01 1.57414651e+00
6.35022298e-02 -8.76263157e-02 1.66648477e-01 2.52365232e-01
-1.50791436e-01 -1.00618029e+00 2.03131407e-01 1.03764129e+00
-1.80708632e-01 -2.26557732e-01 -4.58622158e-01 1.89727843e-01
-1.07315302e-01 1.35722566e+00 -2.47788057e-01 -1.26689303e+00
5.93602479e-01 -4.23654735e-01 3.57859805e-02 -6.78586781e-01
-8.41872275e-01 4.01588492e-02 1.10877752e-01 -9.03946638e-01
8.09914544e-02 -2.72833288e-01 -1.17138588e+00 -5.85389316e-01
-5.94559014e-01 4.58952695e-01 1.17933564e-01 9.48046267e-01
1.00359702e+00 8.14037323e-01 9.70208347e-01 -8.31548035e-01
-1.84653595e-01 -8.55363309e-01 -9.29141521e-01 7.55982220e-01
-5.54128103e-02 -6.65298522e-01 -4.03612316e-01 -1.50702953e-01] | [10.175649642944336, 5.617172718048096] |
8876c320-926a-4f71-9f55-bfa630522196 | convolution-based-channel-frequency-attention | 2210.17310 | null | https://arxiv.org/abs/2210.17310v1 | https://arxiv.org/pdf/2210.17310v1.pdf | Convolution-Based Channel-Frequency Attention for Text-Independent Speaker Verification | Deep convolutional neural networks (CNNs) have been applied to extracting speaker embeddings with significant success in speaker verification. Incorporating the attention mechanism has shown to be effective in improving the model performance. This paper presents an efficient two-dimensional convolution-based attention module, namely C2D-Att. The interaction between the convolution channel and frequency is involved in the attention calculation by lightweight convolution layers. This requires only a small number of parameters. Fine-grained attention weights are produced to represent channel and frequency-specific information. The weights are imposed on the input features to improve the representation ability for speaker modeling. The C2D-Att is integrated into a modified version of ResNet for speaker embedding extraction. Experiments are conducted on VoxCeleb datasets. The results show that C2DAtt is effective in generating discriminative attention maps and outperforms other attention methods. The proposed model shows robust performance with different scales of model size and achieves state-of-the-art results. | ['Tan Lee', 'Yusheng Tian', 'Jingyu Li'] | 2022-10-31 | null | null | null | null | ['text-independent-speaker-verification', 'speaker-verification'] | ['speech', 'speech'] | [-2.58958519e-01 -2.24311337e-01 2.17872620e-01 -4.91540223e-01
-8.13074112e-01 -3.64189856e-02 3.92369002e-01 -1.86511889e-01
-5.19753277e-01 2.05014586e-01 6.61833286e-01 -1.80164546e-01
3.05961788e-01 -4.94422644e-01 -3.94994974e-01 -6.65356755e-01
-1.38502687e-01 -4.06513698e-02 -8.10271502e-03 -1.25657171e-01
2.05164537e-01 6.28079057e-01 -1.51449966e+00 2.60168612e-01
5.10282516e-01 1.00824809e+00 1.81335777e-01 9.46927130e-01
-3.15638036e-01 4.72782791e-01 -9.77191329e-01 -2.61206985e-01
-1.45083562e-01 -1.61786690e-01 -7.31476903e-01 -3.30908567e-01
3.64894688e-01 -2.65092283e-01 -5.04791856e-01 7.97620714e-01
1.23371530e+00 2.10303128e-01 6.29015028e-01 -9.98448908e-01
-9.12996888e-01 6.95712030e-01 -4.34367090e-01 8.37696731e-01
2.00216472e-01 -6.16703518e-02 9.43763793e-01 -1.18628180e+00
8.90709553e-03 1.39685559e+00 8.88685048e-01 7.21745133e-01
-1.04831898e+00 -8.64743114e-01 -6.26910664e-03 6.09939039e-01
-1.62830746e+00 -6.57370210e-01 9.51773643e-01 -3.14424753e-01
1.52021646e+00 3.31809223e-01 3.51897955e-01 1.00012493e+00
2.14486960e-02 5.93945444e-01 7.22456634e-01 -6.01656616e-01
4.58993465e-02 4.03359711e-01 3.95199537e-01 4.88790393e-01
-2.55494207e-01 9.75578725e-02 -7.35980868e-01 3.77648398e-02
6.34336412e-01 -3.81384522e-01 -1.87784687e-01 4.77094144e-01
-6.72690928e-01 1.06617653e+00 8.32407176e-01 5.77318311e-01
-3.84956568e-01 2.38297254e-01 5.59962213e-01 1.18645735e-01
5.32942593e-01 1.13479741e-01 -3.16472143e-01 -1.48424447e-01
-9.47202861e-01 1.05047308e-01 3.82141352e-01 7.73932815e-01
2.72391737e-01 6.26306057e-01 -6.09305799e-01 1.23582196e+00
3.17774057e-01 3.01059127e-01 9.49624956e-01 -1.43386289e-01
3.72916818e-01 3.33297312e-01 -2.89367914e-01 -6.49696648e-01
-4.21963722e-01 -6.14582002e-01 -7.94418097e-01 -5.19656464e-02
-2.09497303e-01 -2.52524495e-01 -1.23762226e+00 1.70232761e+00
2.37912789e-01 4.53335583e-01 -1.28870793e-02 9.93912518e-01
1.32620275e+00 7.70813763e-01 4.64290649e-01 4.93346065e-01
1.76657689e+00 -9.27575707e-01 -1.02300560e+00 -1.25004888e-01
2.28911340e-01 -8.98872674e-01 9.05540287e-01 -3.13186079e-01
-1.03900552e+00 -9.78700697e-01 -1.20971465e+00 -3.46281826e-01
-6.00170910e-01 3.17742229e-01 4.19541419e-01 1.15084183e+00
-1.14429176e+00 1.92145318e-01 -5.10231972e-01 -2.33173549e-01
5.80044150e-01 6.63343072e-01 -1.19739808e-01 3.51285458e-01
-1.53534853e+00 1.12653339e+00 -3.03097256e-03 3.28346670e-01
-9.10423458e-01 -7.61980236e-01 -1.08325720e+00 5.49013495e-01
-6.67114735e-01 -4.26268548e-01 1.20803571e+00 -6.29440844e-01
-1.75131321e+00 5.06285906e-01 -3.48850340e-01 -6.68037593e-01
1.08816430e-01 -9.98340696e-02 -7.14015543e-01 2.40343157e-03
-2.05890208e-01 9.81380701e-01 1.03445542e+00 -5.55832207e-01
-3.74762148e-01 -5.12259640e-02 -2.08243579e-01 7.54052699e-02
-8.14590037e-01 4.52481389e-01 -4.02837247e-01 -9.56934035e-01
-9.00835842e-02 -6.41203880e-01 1.55774713e-01 -5.54112434e-01
-3.89539987e-01 -5.45947492e-01 1.13935220e+00 -1.18343818e+00
1.44456041e+00 -2.30105472e+00 -1.18509576e-01 3.31932940e-02
2.59003416e-02 5.31314790e-01 -1.79721043e-01 2.25779369e-01
-2.88489550e-01 7.23362640e-02 -1.86655822e-03 -7.19518721e-01
2.98651934e-01 -2.06425548e-01 -1.29877955e-01 4.04257953e-01
6.04614675e-01 9.29195225e-01 -1.52537793e-01 -4.09818172e-01
4.36234742e-01 1.25629425e+00 -7.12673783e-01 2.73515642e-01
3.59178722e-01 1.10098749e-01 -1.90287903e-01 4.65679526e-01
8.90120506e-01 2.52235264e-01 -2.51566052e-01 -1.86951265e-01
-5.39077558e-02 8.35125446e-01 -8.60441446e-01 1.45100665e+00
-7.56977260e-01 1.12513065e+00 3.94455083e-02 -8.10455084e-01
9.54997063e-01 7.00328887e-01 3.05510163e-02 -5.41211724e-01
4.99002934e-01 -6.90546706e-02 3.18604678e-01 -6.46270871e-01
4.92120713e-01 -2.44717941e-01 1.81589611e-02 3.77048671e-01
5.49602747e-01 7.42760226e-02 -3.55916619e-01 -5.70648797e-02
6.59304500e-01 -2.96413571e-01 -3.62052098e-02 -3.79039973e-01
8.71020019e-01 -6.84390783e-01 4.57527548e-01 3.23464155e-01
-3.27371478e-01 5.34835994e-01 1.04424618e-01 -3.39320809e-01
-1.02487242e+00 -5.87763011e-01 -4.16301072e-01 1.23071694e+00
-3.95479053e-01 -2.46051058e-01 -1.20660782e+00 -3.06239277e-01
-6.15603989e-03 7.67642140e-01 -9.28364575e-01 -2.40657911e-01
-5.25267899e-01 -7.52945960e-01 9.23442841e-01 8.83491218e-01
7.22685516e-01 -1.23679793e+00 -4.04935867e-01 3.44163150e-01
-1.16444103e-01 -1.01326454e+00 -8.16594362e-01 2.28666887e-01
-4.99910474e-01 -3.49648505e-01 -9.60345626e-01 -1.10281897e+00
4.73691255e-01 -4.85436283e-02 7.02470422e-01 -1.88801020e-01
-5.19037426e-01 -8.13867897e-02 -1.87916040e-01 -9.04948473e-01
-5.06186858e-02 4.67157364e-01 -4.42030281e-02 3.40601653e-01
8.52459967e-01 -4.83689219e-01 -5.33703446e-01 1.10306375e-01
-5.50393581e-01 -4.18469787e-01 5.26332557e-01 9.62383866e-01
-5.04572392e-02 -1.82021618e-01 1.06427836e+00 -3.02367181e-01
8.78215790e-01 -1.74421519e-01 -3.66315782e-01 -3.04989129e-01
-1.46141037e-01 2.62232482e-01 3.13348532e-01 -5.39520681e-01
-1.06111729e+00 -1.45154834e-01 -6.07668996e-01 -4.60441500e-01
-1.29872322e-01 2.50043303e-01 -2.66697049e-01 -1.97800979e-01
3.63018185e-01 1.26310706e-01 6.82416093e-03 -6.16240978e-01
-7.82989152e-03 1.26130867e+00 1.52913883e-01 -9.94812623e-02
4.99473155e-01 3.98048386e-02 -6.01000190e-01 -1.29906905e+00
-3.42415810e-01 -4.70782429e-01 -5.17215371e-01 -1.80925325e-01
1.05791569e+00 -9.34810162e-01 -7.62729585e-01 6.20722592e-01
-1.16588402e+00 -7.22765597e-03 -2.73169540e-02 7.41691232e-01
8.75538364e-02 -1.11610547e-01 -7.97319055e-01 -9.25982058e-01
-7.76445925e-01 -1.31441057e+00 9.91793692e-01 3.70445639e-01
-3.23985577e-01 -1.00302732e+00 5.44728935e-02 2.72352725e-01
9.63011920e-01 -4.38958794e-01 8.12795341e-01 -8.41805339e-01
-3.28850806e-01 -2.66103894e-01 -2.17441067e-01 5.66408575e-01
-1.96948722e-02 -9.29312706e-02 -1.91688657e+00 -1.93883657e-01
-4.91219163e-02 6.68581054e-02 1.11844540e+00 6.85264885e-01
1.31433070e+00 -1.87743083e-01 -1.22692086e-01 6.81381941e-01
1.01440382e+00 1.30794391e-01 8.62910271e-01 1.18208081e-01
5.38076758e-01 2.85943300e-01 -1.31063566e-01 2.21360743e-01
2.37591594e-01 8.59014750e-01 7.21840039e-02 -3.33295703e-01
-6.57434404e-01 4.71399836e-02 3.12561661e-01 1.06632662e+00
-1.86806507e-02 4.55956385e-02 -7.35081911e-01 8.08535397e-01
-1.16869283e+00 -1.02082610e+00 1.35160714e-01 1.90481520e+00
6.50466919e-01 1.53536946e-02 1.71166450e-01 4.80945528e-01
8.93945754e-01 1.99290991e-01 -3.00319523e-01 -1.10665214e+00
9.68997627e-02 7.32514977e-01 1.95454404e-01 7.13015020e-01
-1.20292473e+00 8.31056237e-01 6.59170246e+00 7.10011780e-01
-1.49600708e+00 3.32214355e-01 4.29907233e-01 -1.73516333e-01
5.92989288e-03 -6.82777107e-01 -1.11843014e+00 4.74210531e-01
1.32055223e+00 8.11800957e-02 1.07520908e-01 7.23167002e-01
1.12873130e-02 4.56145585e-01 -8.22860181e-01 9.82888877e-01
2.83882618e-01 -1.14947498e+00 8.50562565e-03 1.76390223e-02
4.41827238e-01 -6.05552979e-02 3.89349282e-01 5.25049865e-01
-4.39672656e-02 -1.23332191e+00 6.46086812e-01 7.03800097e-02
8.61826420e-01 -1.16585243e+00 1.20317686e+00 -2.45489910e-01
-1.46499574e+00 -1.97186828e-01 -4.53060061e-01 1.12807602e-01
7.49575794e-02 2.11621091e-01 -1.35647964e+00 2.26661503e-01
8.97719800e-01 3.74063849e-01 -3.70449126e-01 9.40588117e-01
-6.28887266e-02 7.43981600e-01 -2.73398012e-01 -3.53189230e-01
3.36195111e-01 3.70764315e-01 3.85487139e-01 1.73923898e+00
3.18440735e-01 -1.83103830e-01 -5.78653634e-01 8.04261982e-01
-2.75998116e-01 -6.42188499e-03 -1.48167267e-01 9.66653451e-02
5.29055238e-01 1.25250804e+00 -1.66755915e-01 -3.22060913e-01
-3.56867820e-01 8.10343921e-01 3.38504136e-01 4.23340201e-01
-1.18295300e+00 -9.34053540e-01 1.06666076e+00 -3.03591415e-02
8.92248392e-01 1.21688060e-01 -3.31097424e-01 -6.29974663e-01
-1.70341611e-01 -5.89892209e-01 2.63471097e-01 -6.28647506e-01
-1.12529361e+00 1.12190425e+00 -2.85770327e-01 -8.93288612e-01
-5.34779727e-02 -6.69634044e-01 -9.06180382e-01 1.60468125e+00
-1.77529609e+00 -1.21371424e+00 -4.20926124e-01 6.89876854e-01
6.97257876e-01 -3.88221532e-01 1.29439473e+00 7.32922912e-01
-8.50040674e-01 1.15282142e+00 -3.06747377e-01 4.35107350e-01
4.97847676e-01 -1.12482154e+00 7.05503941e-01 7.59348929e-01
3.30815986e-02 6.80770934e-01 4.35230583e-01 -1.46275237e-01
-1.03776026e+00 -1.19743061e+00 1.17100644e+00 -7.02044442e-02
2.74225235e-01 -6.60961747e-01 -8.69483232e-01 6.20542169e-01
7.09484458e-01 4.95299548e-02 9.48686659e-01 1.36372790e-01
-5.46124518e-01 -2.66726226e-01 -1.15495443e+00 2.00735331e-01
4.47408438e-01 -9.64510620e-01 -6.47510469e-01 -1.21445723e-01
6.29747391e-01 -2.60921836e-01 -7.18354225e-01 1.72595724e-01
5.77783704e-01 -6.04729652e-01 1.15788507e+00 -5.27642727e-01
1.19434027e-02 -9.36285555e-02 -4.91989143e-02 -1.37579310e+00
-5.60350060e-01 -4.62085277e-01 -1.29401013e-01 1.41614962e+00
6.62047803e-01 -7.37581432e-01 4.49954093e-01 3.29198748e-01
-3.86036217e-01 -5.83861649e-01 -1.18127429e+00 -5.33888459e-01
-3.26857939e-02 -3.66929114e-01 8.65684867e-01 8.44139695e-01
1.03936428e-02 5.99185348e-01 -1.37915492e-01 4.05037075e-01
4.71448302e-01 -3.06795508e-01 4.15920407e-01 -1.08714402e+00
6.92872331e-02 -5.84298015e-01 -7.80113935e-01 -8.11096370e-01
2.03434691e-01 -8.23127568e-01 -1.23950087e-01 -1.36549807e+00
-1.43634360e-02 -2.97686875e-01 -5.89283645e-01 3.88097435e-01
-2.52045184e-01 4.21358615e-01 -5.32176197e-02 -2.69110501e-01
1.96392000e-01 5.95973253e-01 8.46849620e-01 -2.92707592e-01
-2.52817065e-01 -1.23645194e-01 -4.86968219e-01 3.55826855e-01
1.09708285e+00 -2.09756300e-01 -1.13994732e-01 -4.60896075e-01
-7.24098265e-01 -4.23011661e-01 2.33290762e-01 -1.07702935e+00
2.85375625e-01 6.87992811e-01 6.84844136e-01 -6.93548560e-01
7.89006650e-01 -6.02124870e-01 -1.43838972e-01 3.68789911e-01
-5.07053912e-01 -2.00872608e-02 7.14351952e-01 3.43550630e-02
-5.19843936e-01 -1.26731694e-01 9.07025635e-01 3.71345520e-01
-5.74473739e-01 2.01316074e-01 -3.48255724e-01 -3.70836765e-01
7.24913120e-01 -1.09623775e-01 -7.12241884e-03 -3.83927494e-01
-8.43832374e-01 -1.52113974e-01 -4.82377082e-01 5.72814405e-01
6.32615387e-01 -1.62836492e+00 -9.18007553e-01 7.34357059e-01
-2.53037717e-02 -5.26994348e-01 6.52020097e-01 6.14529967e-01
-4.19780880e-01 9.18606281e-01 -2.01102152e-01 -5.58033705e-01
-1.46186745e+00 4.40382957e-01 5.33511341e-01 1.34021893e-01
-5.19217908e-01 1.42758739e+00 1.32209852e-01 -3.41628909e-01
6.16167486e-01 -6.35947526e-01 -5.41630566e-01 1.77409425e-01
9.06257272e-01 3.07656437e-01 3.34780037e-01 -8.71545672e-01
-6.65345669e-01 4.96321917e-01 -3.66005987e-01 -1.37351409e-01
1.50232530e+00 1.30019367e-01 2.55357206e-01 1.13737427e-01
1.43230891e+00 -1.96667016e-02 -1.06927204e+00 -9.24445093e-02
-3.07652056e-01 -1.94929972e-01 5.85802495e-01 -7.33726621e-01
-1.25625002e+00 1.43806934e+00 8.51139188e-01 2.49158099e-01
1.01775479e+00 -2.18728125e-01 8.57361078e-01 -3.04163665e-01
-2.07501069e-01 -1.05013084e+00 -9.34055522e-02 4.09084409e-01
1.06278038e+00 -9.48095381e-01 -3.27565342e-01 -1.42123073e-01
-4.59474444e-01 9.66875374e-01 6.74494982e-01 -3.29507105e-02
9.04712379e-01 4.98146892e-01 2.39266858e-01 1.53789958e-02
-4.49798971e-01 -1.62509710e-01 5.13454199e-01 6.36565387e-01
7.21725762e-01 1.18702777e-01 2.11689577e-01 5.82842410e-01
-4.15936828e-01 -4.46720690e-01 -6.66391924e-02 6.50335133e-01
-2.98547685e-01 -9.27209139e-01 -5.34772456e-01 1.33133233e-01
-6.34259880e-01 -3.83594990e-01 -2.64319658e-01 6.00661933e-01
-3.82319167e-02 9.05464292e-01 2.90542483e-01 -4.23866242e-01
3.38621587e-01 4.63657230e-01 2.97538787e-01 -3.53742391e-01
-1.00748932e+00 2.11165264e-01 3.59084830e-02 -2.18526438e-01
-2.09655568e-01 -3.73237818e-01 -1.08524156e+00 -3.48878056e-01
-5.42248547e-01 2.38253668e-01 1.08267713e+00 4.66879338e-01
7.16316819e-01 1.28263891e+00 6.25142038e-01 -1.00777900e+00
-3.37335140e-01 -1.75531971e+00 -3.74320656e-01 1.17426887e-01
7.73893774e-01 -7.29421139e-01 -4.05292183e-01 -1.44782432e-04] | [14.367362022399902, 6.127752780914307] |
781a62fe-2fcd-4ae0-8090-ebab7e97adba | vilbert-pretraining-task-agnostic | 1908.02265 | null | https://arxiv.org/abs/1908.02265v1 | https://arxiv.org/pdf/1908.02265v1.pdf | ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks | We present ViLBERT (short for Vision-and-Language BERT), a model for learning task-agnostic joint representations of image content and natural language. We extend the popular BERT architecture to a multi-modal two-stream model, pro-cessing both visual and textual inputs in separate streams that interact through co-attentional transformer layers. We pretrain our model through two proxy tasks on the large, automatically collected Conceptual Captions dataset and then transfer it to multiple established vision-and-language tasks -- visual question answering, visual commonsense reasoning, referring expressions, and caption-based image retrieval -- by making only minor additions to the base architecture. We observe significant improvements across tasks compared to existing task-specific models -- achieving state-of-the-art on all four tasks. Our work represents a shift away from learning groundings between vision and language only as part of task training and towards treating visual grounding as a pretrainable and transferable capability. | ['Stefan Lee', 'Jiasen Lu', 'Dhruv Batra', 'Devi Parikh'] | 2019-08-06 | vilbert-pretraining-task-agnostic-1 | http://papers.nips.cc/paper/8297-vilbert-pretraining-task-agnostic-visiolinguistic-representations-for-vision-and-language-tasks | http://papers.nips.cc/paper/8297-vilbert-pretraining-task-agnostic-visiolinguistic-representations-for-vision-and-language-tasks.pdf | neurips-2019-12 | ['visual-commonsense-reasoning'] | ['reasoning'] | [ 2.68312663e-01 4.51439053e-01 -6.77765906e-02 -5.45134962e-01
-9.54010725e-01 -7.42681861e-01 1.17082608e+00 1.14613906e-01
-5.10966361e-01 3.61266106e-01 5.16449690e-01 -5.39690197e-01
3.46417397e-01 -4.50956523e-01 -1.15695143e+00 -1.18755028e-01
4.31850672e-01 7.71462321e-01 2.67167181e-01 -3.25750083e-01
1.07606046e-01 2.14416310e-01 -1.41085386e+00 1.02656722e+00
3.40204626e-01 9.60499585e-01 9.14332643e-02 9.54228520e-01
-3.58575016e-01 1.83277476e+00 -3.22152972e-01 -6.16170824e-01
-1.72489703e-01 -3.85508627e-01 -1.27951467e+00 2.13495657e-01
9.84630883e-01 -3.91100496e-01 -4.49852943e-01 6.71653807e-01
1.71370842e-02 -1.13355659e-01 8.07827353e-01 -1.54801452e+00
-1.68763638e+00 5.23058295e-01 -4.48143989e-01 2.18587056e-01
3.55053097e-01 5.83778024e-01 1.28232026e+00 -1.00523138e+00
6.24967098e-01 1.62127507e+00 4.43564415e-01 7.77479410e-01
-1.53936338e+00 -1.97803825e-01 4.48659331e-01 4.64891583e-01
-8.57514858e-01 -6.58689797e-01 5.84421694e-01 -6.77727640e-01
1.38879240e+00 -2.48522335e-03 5.44311583e-01 1.27458239e+00
-2.43632093e-01 1.26496601e+00 1.11793828e+00 -4.33730006e-01
-8.64901543e-02 1.38071358e-01 2.18929857e-01 9.61789846e-01
3.40098441e-02 8.79298002e-02 -5.78960359e-01 2.12229878e-01
7.56505489e-01 -1.28044143e-01 -1.16432562e-01 -4.33550417e-01
-1.52632475e+00 8.00563216e-01 1.00552702e+00 5.85962459e-03
-4.15381879e-01 8.87893498e-01 5.03219366e-01 2.88724929e-01
6.96888193e-02 5.00214815e-01 -3.36411595e-01 2.16220975e-01
-7.48961210e-01 2.53564715e-01 6.54254854e-01 1.21307182e+00
7.84456551e-01 1.78056583e-01 -8.34976971e-01 4.67090249e-01
5.25779009e-01 6.98400438e-01 2.46331006e-01 -1.12870038e+00
4.70752299e-01 5.34057975e-01 2.24677264e-03 -5.78053713e-01
-7.58102313e-02 -1.39442489e-01 -4.32854056e-01 1.54621482e-01
4.07751590e-01 7.74331763e-02 -1.44109166e+00 1.88614106e+00
-1.30262733e-01 -4.69013415e-02 2.31413186e-01 8.86037111e-01
1.52416599e+00 6.59125686e-01 7.14398205e-01 3.16582799e-01
1.70970213e+00 -1.30119181e+00 -4.76797700e-01 -6.69496536e-01
3.47344667e-01 -5.21559656e-01 1.54974163e+00 2.11931765e-01
-1.48282456e+00 -6.49870217e-01 -7.87036955e-01 -1.03157926e+00
-6.52658463e-01 1.25780627e-01 4.97958541e-01 -8.04193467e-02
-1.69077063e+00 -8.12274292e-02 -4.95715410e-01 -5.54975033e-01
8.52844834e-01 -1.27472326e-01 -3.48971337e-01 -2.11120725e-01
-8.36475074e-01 1.28606856e+00 1.94320723e-01 -7.37561807e-02
-1.43247533e+00 -7.41327524e-01 -1.14889979e+00 -8.83165002e-02
2.98828989e-01 -1.27712846e+00 1.68964720e+00 -1.27961850e+00
-1.11338389e+00 1.70754576e+00 -2.44526923e-01 -6.50834203e-01
3.18582445e-01 -3.53655368e-01 -1.02276571e-01 4.02586073e-01
2.60840297e-01 1.28171575e+00 9.68748927e-01 -1.66521060e+00
-2.50789613e-01 -1.44028798e-01 6.69172049e-01 1.74935400e-01
-2.15465575e-02 6.85645491e-02 -6.36808872e-01 -3.17315012e-01
-4.97086614e-01 -6.86526477e-01 1.31779492e-01 6.04669750e-01
-4.50407743e-01 -4.24869150e-01 7.13040411e-01 -6.34145975e-01
4.23546553e-01 -1.90141988e+00 3.18125695e-01 -4.96824145e-01
4.06612843e-01 1.36738464e-01 -6.22651577e-01 4.02447045e-01
-2.56166160e-01 -5.65777766e-03 -1.61796764e-01 -5.78006804e-01
1.78697497e-01 4.71658468e-01 -6.75552070e-01 1.82640284e-01
8.39057148e-01 1.75924528e+00 -1.14755583e+00 -7.00473964e-01
3.17819715e-01 3.91284347e-01 -3.98429483e-01 5.60033679e-01
-8.66864443e-01 3.08776014e-02 -1.18262500e-01 5.74269891e-01
1.92993358e-01 -6.77442789e-01 -2.82167673e-01 -4.77076411e-01
1.87641978e-01 1.09531842e-01 -2.35877097e-01 1.90700626e+00
-7.80458987e-01 1.03797460e+00 7.08754435e-02 -1.09493601e+00
5.28770506e-01 3.72157365e-01 -1.08409330e-01 -8.78902733e-01
9.46147367e-02 -3.47133398e-01 -2.31455684e-01 -9.28356886e-01
2.49531403e-01 -2.48110548e-01 -5.65571673e-02 3.95222843e-01
5.26599050e-01 -5.23714781e-01 1.41946509e-01 6.82893693e-01
6.98596418e-01 6.53362870e-01 1.35232568e-01 -1.05879739e-01
4.48521346e-01 3.21983367e-01 -3.14290226e-01 8.20329368e-01
-2.08301485e-01 6.39413595e-01 6.76832914e-01 -3.93210441e-01
-1.10134268e+00 -1.31688678e+00 2.61594892e-01 1.53316808e+00
-6.53259531e-02 -3.31467897e-01 -4.99012381e-01 -5.50570250e-01
1.58925936e-01 1.20785904e+00 -1.03893876e+00 -1.74281955e-01
-2.06192777e-01 -1.60311028e-01 7.19609857e-01 9.22687829e-01
4.48708802e-01 -1.57026982e+00 -8.79288793e-01 -2.43842378e-01
-5.70261590e-02 -1.54049754e+00 -1.91407830e-01 1.10659398e-01
-4.38847572e-01 -1.12040508e+00 -7.28619754e-01 -1.02458024e+00
4.70136791e-01 3.24212730e-01 1.90913510e+00 8.58427808e-02
-4.16984111e-01 1.06828868e+00 -1.15973584e-01 -7.65787959e-01
-4.16042835e-01 -4.58598584e-01 -6.90641463e-01 -1.36364400e-01
4.35645461e-01 -1.89154804e-01 -5.31243563e-01 -2.98551202e-01
-9.27841723e-01 4.16548461e-01 6.25691354e-01 7.71024108e-01
4.25242603e-01 -1.14410067e+00 2.73827970e-01 -8.80073547e-01
5.79496384e-01 -4.97365505e-01 -3.11617821e-01 5.07276535e-01
-1.27030909e-01 2.69372702e-01 2.92306900e-01 -3.06788087e-01
-9.18926179e-01 1.16401389e-02 1.64746404e-01 -8.12007904e-01
-3.77351224e-01 3.61916423e-01 2.59040780e-02 1.28263772e-01
7.72268772e-01 3.26895267e-01 7.19794184e-02 9.29516536e-05
1.52094388e+00 2.68398970e-01 1.08228576e+00 -7.52412021e-01
8.61579597e-01 6.83285534e-01 -1.06090000e-02 -7.54947126e-01
-1.33882129e+00 -5.45149744e-01 -5.75993121e-01 -1.19431287e-01
1.39314723e+00 -1.27985394e+00 -8.67531419e-01 7.47126490e-02
-1.69008923e+00 -6.80905938e-01 -5.76414287e-01 -1.78698316e-01
-9.18966532e-01 9.38813612e-02 -5.88036776e-01 -5.85221410e-01
-3.86911035e-01 -9.50287461e-01 1.38462186e+00 5.27527221e-02
-1.03465058e-01 -1.06689978e+00 -9.21401978e-02 5.11116028e-01
4.31504279e-01 3.29170227e-01 1.10362065e+00 -5.31533062e-01
-6.24376595e-01 1.96979702e-01 -1.06222010e+00 2.90234774e-01
-4.86266553e-01 -1.15778103e-01 -1.38887751e+00 -2.50015259e-02
-2.75604218e-01 -1.40940440e+00 1.26944709e+00 9.10845101e-02
1.23437262e+00 -2.68923730e-01 -1.47011250e-01 5.89496434e-01
1.55028486e+00 -3.72764498e-01 5.80450773e-01 3.45373571e-01
8.66089642e-01 6.64220750e-01 1.13516428e-01 -8.55986774e-02
9.91219759e-01 4.05229419e-01 8.88350010e-01 -5.62969923e-01
-6.40660524e-01 -3.96339178e-01 3.16404432e-01 1.53851807e-01
-2.17237901e-02 -1.64582998e-01 -1.12907290e+00 9.61146116e-01
-2.03397512e+00 -1.07944918e+00 -1.96185678e-01 1.68421507e+00
1.04456580e+00 -1.68154061e-01 1.96242899e-01 -4.73094493e-01
1.56808570e-01 2.33338416e-01 -6.22974575e-01 -6.65276945e-01
-3.17498222e-02 1.33311912e-01 1.51772678e-01 5.17933071e-01
-1.05900455e+00 1.31709719e+00 6.43460226e+00 5.30438311e-02
-1.11305368e+00 2.14164197e-01 5.54382443e-01 -2.93859895e-02
-4.73105401e-01 1.29916444e-01 -2.52410442e-01 -2.81169772e-01
9.18838441e-01 5.22585493e-03 3.45395476e-01 7.39017546e-01
-2.15271875e-01 1.36844143e-01 -1.59511530e+00 1.20415139e+00
6.88453972e-01 -1.43966436e+00 5.88596523e-01 -2.97688723e-01
4.17749703e-01 4.08156008e-01 1.40740380e-01 5.99288225e-01
6.97974682e-01 -1.37664914e+00 1.21102595e+00 6.38122082e-01
1.01503956e+00 -1.36831328e-01 3.68648231e-01 -2.78719254e-02
-8.55246365e-01 -4.35887836e-03 -2.33579919e-01 -1.37013448e-02
1.57001019e-01 -3.54636274e-02 -6.55197620e-01 2.51189232e-01
6.23194456e-01 8.98363650e-01 -9.03507531e-01 7.32192993e-01
-5.65411985e-01 4.29405659e-01 1.24792859e-01 8.48371536e-02
5.62348723e-01 3.49060744e-01 2.42156804e-01 1.53098559e+00
-4.41791534e-01 -3.09392363e-02 2.73433864e-01 1.42098880e+00
-1.95023268e-01 -3.17491621e-01 -7.51606762e-01 -2.26063862e-01
3.96627448e-02 1.18619609e+00 -2.34745011e-01 -6.52408004e-01
-8.42981875e-01 1.14653695e+00 6.46167219e-01 8.39230120e-01
-8.55809569e-01 -2.98098356e-01 3.63826245e-01 8.26290324e-02
4.68745738e-01 -2.18273997e-01 -1.74095839e-01 -1.27662265e+00
-3.03720444e-01 -8.48783791e-01 4.93803978e-01 -1.66330814e+00
-1.47860003e+00 6.43287182e-01 1.45059794e-01 -6.59939110e-01
-4.76740032e-01 -1.11773574e+00 -5.80678046e-01 8.94568026e-01
-1.97877407e+00 -2.10607672e+00 -5.83095610e-01 9.20406938e-01
7.09435940e-01 3.60612795e-02 9.81951296e-01 -3.76868755e-01
-3.63709335e-03 3.35863262e-01 -5.94021678e-01 3.23089361e-01
6.82805777e-01 -1.50487483e+00 5.08110046e-01 6.61362171e-01
5.87345779e-01 4.79866862e-01 6.07082009e-01 -1.70796543e-01
-1.49838352e+00 -1.08866751e+00 8.04674685e-01 -1.16582012e+00
1.04633808e+00 -6.51245594e-01 -7.97050297e-01 1.07279980e+00
9.68110204e-01 2.67333537e-01 4.20875847e-01 2.00944439e-01
-1.05570376e+00 1.47342369e-01 -7.40400374e-01 6.39166951e-01
1.00609171e+00 -1.18949664e+00 -1.25975156e+00 6.27884686e-01
1.01608384e+00 -2.54195422e-01 -3.51898879e-01 -8.10764581e-02
3.69064569e-01 -7.21125960e-01 1.34546447e+00 -1.25862193e+00
9.35792863e-01 -2.74083436e-01 -2.22972259e-01 -1.01230276e+00
-3.38348538e-01 -3.89300853e-01 -1.22234695e-01 1.06337070e+00
4.53052163e-01 -1.74154453e-02 3.15003514e-01 4.55134749e-01
-6.71702847e-02 -3.87460202e-01 -5.27661383e-01 -3.09269905e-01
2.19264090e-01 -6.02683425e-01 7.17155337e-02 7.78657556e-01
-7.49115646e-02 1.08035600e+00 -4.37238179e-02 -1.48886770e-01
5.81734657e-01 8.40392113e-02 8.56189191e-01 -9.14602578e-01
-3.79780531e-01 -4.86328125e-01 -3.09340715e-01 -1.01599848e+00
4.67101663e-01 -1.14299345e+00 2.96122640e-01 -2.18717504e+00
4.67245340e-01 2.59651423e-01 -1.86361045e-01 9.83708739e-01
-8.14415738e-02 4.29013819e-01 5.15155315e-01 2.00653762e-01
-1.05550647e+00 4.87762839e-01 1.33979464e+00 -4.45496857e-01
3.41519088e-01 -6.21175647e-01 -9.48017597e-01 7.81505883e-01
2.42976964e-01 7.78650567e-02 -7.79143691e-01 -1.14624286e+00
3.20601016e-01 -7.13155568e-02 1.16780734e+00 -7.14528739e-01
1.41892150e-01 -1.93817064e-01 4.91782755e-01 -2.62247801e-01
4.52392310e-01 -6.43515348e-01 -7.83358216e-01 1.14155047e-01
-8.96092057e-01 2.55248576e-01 5.69537461e-01 6.67070687e-01
-2.40585104e-01 1.75498277e-01 6.28427148e-01 -5.17455697e-01
-1.08356261e+00 1.53239712e-01 -8.69729072e-02 4.87225503e-01
8.63853514e-01 -1.19125612e-01 -8.69199038e-01 -6.83124006e-01
-8.62086356e-01 4.79557544e-01 4.26750839e-01 5.64466119e-01
7.70462632e-01 -9.93490756e-01 -8.81750464e-01 -1.66502118e-01
7.41416931e-01 9.46335774e-03 -6.19157031e-02 5.26989758e-01
-4.58293617e-01 7.64950335e-01 -2.04777300e-01 -7.93092430e-01
-8.87567699e-01 9.67268825e-01 3.72408748e-01 1.82670265e-01
-5.50695896e-01 1.12509418e+00 5.82940698e-01 -2.29245961e-01
2.51813799e-01 -5.26513696e-01 -3.06633353e-01 -1.08419396e-01
6.07344031e-01 -3.64301682e-01 -2.55669862e-01 -7.76516497e-01
-3.92751843e-01 6.43935263e-01 -1.20736053e-02 -3.41879576e-01
1.18978417e+00 -1.36079878e-01 -2.59173155e-01 6.92008495e-01
1.08075666e+00 -6.35582805e-01 -1.27219975e+00 -3.58278871e-01
7.73603795e-03 1.91160501e-03 1.26317814e-01 -1.20139837e+00
-6.76584482e-01 1.50817072e+00 2.62830973e-01 -2.09190696e-02
8.88166964e-01 6.68439627e-01 3.12326998e-01 6.35619938e-01
-1.13685220e-01 -6.11256123e-01 7.08358169e-01 6.45537496e-01
1.45997930e+00 -1.42851901e+00 -1.62180170e-01 -9.82472580e-03
-1.02016103e+00 9.06766415e-01 8.89997423e-01 -3.48382950e-01
2.91485906e-01 -1.03969991e-01 2.83247054e-01 -5.17221153e-01
-1.24913371e+00 -6.49907708e-01 6.70422792e-01 1.02615011e+00
6.70201182e-01 -2.12922245e-01 4.27500844e-01 3.15394193e-01
1.50532022e-01 2.06417486e-01 3.26566607e-01 7.78876662e-01
-2.64299691e-01 -6.25702381e-01 -1.27024442e-01 1.97123379e-01
7.25174844e-02 -5.68444014e-01 -7.36619651e-01 9.21136320e-01
-7.82044381e-02 8.79676342e-01 3.98650348e-01 1.37012333e-01
4.17353511e-01 2.49679103e-01 8.05085063e-01 -9.32032943e-01
-4.61040795e-01 -3.74533623e-01 3.07256579e-01 -8.32816601e-01
-6.61631763e-01 -3.05225462e-01 -1.17619288e+00 2.64776260e-01
2.38075063e-01 -3.07641208e-01 4.77589607e-01 1.19601333e+00
3.17018002e-01 7.25690007e-01 -2.54474193e-01 -8.34497988e-01
-6.02827430e-01 -7.67508268e-01 1.47174031e-01 8.89129162e-01
6.97126031e-01 -4.11240190e-01 2.31035966e-02 5.50172687e-01] | [10.835336685180664, 1.7728861570358276] |
4e6f512e-6536-4d12-a335-1871887ee2e2 | text-sampling-strategies-for-predicting | 2301.01673 | null | https://arxiv.org/abs/2301.01673v1 | https://arxiv.org/pdf/2301.01673v1.pdf | Text sampling strategies for predicting missing bibliographic links | The paper proposes various strategies for sampling text data when performing automatic sentence classification for the purpose of detecting missing bibliographic links. We construct samples based on sentences as semantic units of the text and add their immediate context which consists of several neighboring sentences. We examine a number of sampling strategies that differ in context size and position. The experiment is carried out on the collection of STEM scientific papers. Including the context of sentences into samples improves the result of their classification. We automatically determine the optimal sampling strategy for a given text collection by implementing an ensemble voting when classifying the same data sampled in different ways. Sampling strategy taking into account the sentence context with hard voting procedure leads to the classification accuracy of 98% (F1-score). This method of detecting missing bibliographic links can be used in recommendation engines of applied intelligent information systems. | ['E. N. Baskakova', 'I. S. Smaznevicha', 'F. V. Krasnova'] | 2023-01-04 | null | null | null | null | ['sentence-classification'] | ['natural-language-processing'] | [ 3.46102297e-01 1.66690901e-01 -3.45680267e-01 -2.28366777e-01
-5.45006514e-01 -5.64407170e-01 8.56324434e-01 7.51805663e-01
-5.66888213e-01 9.86996949e-01 6.00223482e-01 -6.20063066e-01
-3.68019551e-01 -1.14090979e+00 -2.73095846e-01 -3.57350111e-01
4.29495007e-01 4.30378914e-01 5.48798144e-01 -9.31233391e-02
1.23304451e+00 4.79495406e-01 -1.62558162e+00 5.65195739e-01
9.74388421e-01 3.81323814e-01 6.47164702e-01 7.48341501e-01
-8.77252579e-01 5.61767161e-01 -1.16479266e+00 -2.34614223e-01
-6.80320114e-02 -6.99233115e-01 -1.15515840e+00 -3.42736184e-03
5.24561405e-01 4.75409448e-01 3.30638200e-01 1.27118635e+00
4.44710016e-01 9.85581893e-03 9.46249723e-01 -6.54898465e-01
-2.12380528e-01 1.08737278e+00 -3.71744603e-01 5.71099937e-01
9.30631220e-01 -7.39509046e-01 1.02505469e+00 -6.41506910e-01
1.06415880e+00 1.25298262e+00 4.13562268e-01 1.22122683e-01
-1.09804010e+00 -3.71855646e-01 -8.92109200e-02 3.57375622e-01
-1.04849100e+00 -1.91079542e-01 6.78864002e-01 -6.74143851e-01
7.87325382e-01 6.85541332e-01 4.62731272e-01 8.05367887e-01
4.78601009e-01 2.19182625e-01 1.08704805e+00 -9.93355811e-01
4.70743060e-01 4.40034628e-01 9.35135245e-01 3.40446860e-01
7.47647643e-01 -6.84199214e-01 -4.38580602e-01 -5.78120232e-01
1.97748337e-02 -1.94856808e-01 -2.70212471e-01 4.17078227e-01
-1.09830356e+00 6.26641810e-01 -1.27670869e-01 1.05958891e+00
-2.80114889e-01 -7.15833545e-01 6.57272458e-01 4.52488363e-01
4.12393540e-01 5.93975127e-01 -3.79039526e-01 2.83824146e-01
-1.25939870e+00 2.05316365e-01 9.96418297e-01 9.11476016e-01
3.32406700e-01 -6.19261324e-01 -6.01873815e-01 7.63823986e-01
-7.36148953e-02 2.05881342e-01 7.80891299e-01 -6.12424433e-01
5.96426249e-01 9.81943727e-01 5.83857000e-02 -1.15462244e+00
-2.83694565e-01 -3.86665851e-01 -4.20911849e-01 -3.57383862e-02
3.60677093e-01 -1.86029971e-01 -5.21750331e-01 1.12547266e+00
3.38377565e-01 -2.30739549e-01 1.45919874e-01 2.43900254e-01
1.22612607e+00 5.89493752e-01 1.73876852e-01 -6.18850946e-01
1.49749851e+00 -5.48945248e-01 -7.43451595e-01 6.51954293e-01
6.94124639e-01 -1.23984504e+00 6.57956302e-01 4.67957884e-01
-7.37055123e-01 -6.00728989e-01 -1.00599134e+00 2.30596066e-01
-6.63647473e-01 3.33425611e-01 -3.48602906e-02 5.79545438e-01
-9.23986137e-01 8.65428150e-01 -1.58419982e-01 -5.84328890e-01
-1.00439347e-01 2.26752505e-01 -2.01251775e-01 3.25737953e-01
-9.64044333e-01 9.07850802e-01 3.80183488e-01 -5.37207544e-01
2.06356391e-01 -3.21974158e-01 -3.66904467e-01 1.69429407e-01
1.45792529e-01 -7.58958757e-01 6.48279369e-01 -9.78982568e-01
-1.04639494e+00 9.34986055e-01 -4.87802714e-01 -2.71900743e-01
5.79702556e-01 4.06114817e-01 -6.21209681e-01 2.44589657e-01
5.25824368e-01 2.65142079e-02 4.70141947e-01 -8.12660694e-01
-7.35539913e-01 -4.88907307e-01 -3.24021339e-01 2.13082880e-01
-3.66431117e-01 3.23565185e-01 -1.14161298e-01 -6.26670957e-01
1.15229994e-01 -4.52403337e-01 -2.20571533e-02 -6.84801221e-01
-4.04528826e-01 -8.20476770e-01 5.89599431e-01 -8.00770879e-01
1.55027676e+00 -1.44301426e+00 6.43026382e-02 4.66370642e-01
-8.67750123e-02 -2.71579623e-02 1.55268058e-01 5.15201092e-01
2.48153303e-02 4.51876670e-01 1.36824757e-01 4.17553037e-02
-4.12476420e-01 -2.13099927e-01 -1.25949189e-01 -2.92798616e-02
-2.95094907e-01 7.54495040e-02 -8.84546280e-01 -1.06982005e+00
-2.66240514e-03 4.26616035e-02 -1.24011144e-01 -1.34985998e-01
-1.89604133e-01 -1.78784225e-02 -8.35348845e-01 3.61288905e-01
4.08828080e-01 1.44659234e-02 4.07975256e-01 -1.05337694e-01
-2.45644420e-01 5.90763152e-01 -1.28267729e+00 1.05635452e+00
-1.98613495e-01 6.98919475e-01 -3.95941317e-01 -1.10640407e+00
1.20671117e+00 3.02676260e-01 2.53580898e-01 -3.21725696e-01
2.33899429e-01 3.41876209e-01 5.53351128e-03 -6.72255397e-01
7.08077967e-01 2.13054314e-01 9.24664289e-02 2.52579868e-01
5.67197576e-02 1.79795712e-01 8.58417213e-01 4.61641461e-01
1.16306567e+00 -7.15067759e-02 4.93601233e-01 -8.08479428e-01
1.14707673e+00 1.57631636e-01 2.48108879e-01 1.03212929e+00
1.36812866e-01 2.30061963e-01 5.70610166e-01 -1.67170539e-01
-1.10543346e+00 -4.86527681e-01 -5.50294518e-01 8.96815836e-01
-7.67055675e-02 -4.79413629e-01 -8.64766061e-01 -8.27328205e-01
-5.66551695e-03 7.74059176e-01 -5.98968148e-01 3.23997825e-01
-4.07410681e-01 -7.30440795e-01 1.62182581e-02 -9.95562226e-02
2.05830932e-01 -1.08762884e+00 -3.28599215e-01 2.25292206e-01
-2.10690603e-01 -6.88549876e-01 -1.78357556e-01 -1.09593999e-02
-1.06496406e+00 -1.25375593e+00 -5.82342923e-01 -8.71174335e-01
7.65104473e-01 2.23803699e-01 1.16946626e+00 2.09256232e-01
-1.02339581e-01 3.83850746e-03 -5.15183687e-01 -4.25332874e-01
-8.22854757e-01 4.27208602e-01 -2.49487773e-01 -6.26172841e-01
6.24098599e-01 -1.91083655e-01 -3.02015275e-01 -4.44583158e-04
-6.52322054e-01 -1.52892917e-01 2.29457468e-01 6.66110575e-01
1.88892618e-01 1.00675263e-01 9.89096403e-01 -1.27744889e+00
1.12532783e+00 -7.55434632e-01 -3.78437370e-01 5.84004462e-01
-7.48967171e-01 1.64620161e-01 7.85065591e-01 -1.31933540e-01
-1.04232728e+00 -4.40985262e-01 -2.10993782e-01 5.42185783e-01
-1.94957048e-01 4.39769566e-01 1.07712917e-01 1.60242207e-02
6.62992358e-01 8.60138796e-03 -2.30774894e-01 -4.94792014e-01
-3.16341758e-01 1.24111331e+00 2.16436852e-02 -5.10533810e-01
2.07856987e-02 -1.18107572e-02 1.74932852e-01 -9.67674136e-01
-7.28215396e-01 -7.90576458e-01 -5.12937129e-01 -4.88648564e-01
6.29960656e-01 -4.42819983e-01 -5.26499212e-01 -2.11946577e-01
-1.24315989e+00 6.42140210e-01 7.02977329e-02 5.33742785e-01
9.01107639e-02 7.55187273e-01 -3.48085105e-01 -7.08498776e-01
-7.43862391e-01 -8.57416749e-01 7.75870264e-01 2.21390352e-01
-6.16657376e-01 -8.75299037e-01 1.62084058e-01 4.31795239e-01
-6.65449947e-02 -2.23842766e-02 1.07283282e+00 -1.12694240e+00
3.18137482e-02 -3.74083906e-01 9.48181301e-02 1.21910118e-01
1.35196298e-01 3.69070530e-01 -7.45637655e-01 5.51828034e-02
-2.18759887e-02 4.93762285e-01 9.72505152e-01 3.62084478e-01
1.14793313e+00 -2.67089725e-01 -6.55076385e-01 -3.97685140e-01
1.65068591e+00 4.13657963e-01 4.48192090e-01 5.44645369e-01
1.56742945e-01 6.17920041e-01 4.84188139e-01 2.62406498e-01
-5.64126559e-02 2.76506394e-01 -8.96437690e-02 6.03757501e-01
-6.60933331e-02 -1.89276151e-02 1.57411948e-01 1.10067213e+00
-2.64051389e-02 -4.61896449e-01 -7.79584050e-01 5.52627683e-01
-1.53685641e+00 -1.40719390e+00 -5.00622511e-01 2.25934601e+00
9.56373096e-01 4.83897120e-01 1.54938519e-01 5.79940081e-01
1.15375364e+00 -1.76682904e-01 1.59297988e-01 -6.87476993e-01
-6.68741465e-02 3.02973330e-01 4.36396867e-01 8.44304264e-01
-7.94632852e-01 5.35733998e-01 6.45512533e+00 9.09464836e-01
-7.87557423e-01 -1.78437904e-01 5.08756876e-01 1.66737989e-01
-5.37768185e-01 7.82222822e-02 -1.26001382e+00 1.02351236e+00
9.60823953e-01 -4.10218239e-01 -2.14716986e-01 5.06933808e-01
2.77127355e-01 -7.58204520e-01 -7.48328507e-01 3.12940836e-01
2.15507761e-01 -1.63144124e+00 4.41643864e-01 -1.27429456e-01
8.59633923e-01 -2.06688598e-01 -7.51946688e-01 -9.40911248e-02
4.70849797e-02 -3.83307159e-01 5.21890223e-01 7.89703667e-01
1.06086031e-01 -6.06980801e-01 1.05165708e+00 4.98286247e-01
-7.01112390e-01 1.02517091e-01 -4.95462537e-01 -3.69945765e-01
-3.36327136e-01 9.86744523e-01 -9.84357834e-01 7.44777203e-01
6.32929981e-01 4.92499202e-01 -7.18204737e-01 1.13761055e+00
2.57717967e-01 6.98648751e-01 -2.75289923e-01 -1.02494812e+00
-5.46705984e-02 -5.50088286e-01 8.53579044e-01 1.54331791e+00
5.01479745e-01 -1.46422103e-01 -9.76278260e-02 5.50289571e-01
-1.17266573e-01 7.81930923e-01 -6.63448095e-01 4.07857686e-01
8.56042981e-01 9.94087696e-01 -1.12067294e+00 -7.53441393e-01
-2.21656844e-01 5.31163990e-01 1.50068834e-01 -1.34059489e-02
-5.25887497e-02 -8.47350895e-01 -1.80993780e-01 6.67756572e-02
8.15164372e-02 2.38991156e-01 -6.49358571e-01 -9.64208126e-01
1.65147305e-01 -4.21358258e-01 4.65047687e-01 -5.02138853e-01
-1.11524463e+00 6.42010808e-01 -2.27025039e-02 -1.04745388e+00
3.80243175e-04 -6.04588628e-01 -8.17484021e-01 9.73394752e-01
-8.49234998e-01 -3.79328310e-01 -1.96445599e-01 -4.86160032e-02
7.83285320e-01 -4.59241509e-01 8.50542724e-01 5.70631959e-02
-5.74784100e-01 1.13419175e-01 4.47007686e-01 9.95008051e-02
6.76818252e-01 -1.30785382e+00 -2.83556849e-01 7.60189056e-01
6.38981909e-03 7.07555950e-01 1.04583061e+00 -9.25859571e-01
-6.86932206e-01 -6.91548109e-01 1.76079619e+00 -3.58256936e-01
5.35687745e-01 7.44708851e-02 -7.44202554e-01 6.77835643e-02
6.64775133e-01 -8.65549266e-01 7.66852200e-01 2.25913823e-01
1.89906955e-01 -2.68522222e-02 -1.34181535e+00 6.48346841e-01
7.02420354e-01 -2.78534144e-01 -1.17200053e+00 5.22739470e-01
4.10543680e-01 2.08663315e-01 -1.07844961e+00 1.31873205e-01
2.96047449e-01 -8.66441667e-01 7.16148794e-01 -4.62014586e-01
4.89677340e-01 -2.35139519e-01 1.82236388e-01 -1.01138043e+00
-2.96456546e-01 -1.78254366e-01 4.41257834e-01 1.37292254e+00
5.70370197e-01 -5.55921733e-01 7.27783620e-01 1.41091362e-01
3.95796478e-01 -5.55350780e-01 -7.42974997e-01 -3.16010594e-01
3.35982777e-02 1.08163625e-01 3.42302352e-01 1.07808387e+00
5.05779147e-01 7.33492553e-01 4.79389459e-01 -3.48721713e-01
7.09663272e-01 3.41460854e-01 3.20825726e-01 -1.71934378e+00
6.99026059e-05 -8.02157104e-01 -4.72407877e-01 -3.07326525e-01
2.76362330e-01 -1.07473600e+00 -5.01392901e-01 -1.62178457e+00
4.99239922e-01 -3.66809905e-01 -9.73114073e-02 -1.53801786e-02
-3.39898586e-01 -1.30048677e-01 -1.68877274e-01 3.02358091e-01
-2.76540786e-01 -8.00518319e-02 9.88837063e-01 -1.00242108e-01
-5.72226942e-02 2.18399525e-01 -5.45795023e-01 6.23809695e-01
8.50250244e-01 -4.94490236e-01 -2.62130290e-01 8.88460353e-02
4.25404847e-01 3.91912647e-02 -1.13669865e-01 -9.26730990e-01
3.04505557e-01 -2.02621132e-01 5.71080565e-01 -7.95056462e-01
-5.88526547e-01 -7.44369447e-01 1.29989058e-01 5.34308195e-01
-8.26765299e-01 1.25721872e-01 -1.06656283e-01 4.53826904e-01
-1.10056415e-01 -1.23110628e+00 4.65863645e-01 -3.97836298e-01
-2.72348434e-01 -5.77270031e-01 -7.42625117e-01 -1.50940865e-01
8.43847513e-01 -2.80138820e-01 -3.41446072e-01 1.11103565e-01
-7.91247249e-01 -9.53622535e-02 3.32386047e-01 2.48235181e-01
2.47726858e-01 -8.89331579e-01 -7.75891066e-01 -2.70933360e-01
2.70726774e-02 -5.05890787e-01 -3.09436619e-01 6.60746217e-01
-7.61043608e-01 4.34834957e-01 -1.52551010e-01 -2.02267930e-01
-1.95805860e+00 3.36751819e-01 -6.20069169e-02 -3.29063088e-01
-4.94148701e-01 3.51926744e-01 -7.42077827e-01 -1.60687193e-01
2.47889549e-01 -4.28201616e-01 -1.09162199e+00 4.00391638e-01
4.95889783e-01 6.87343478e-01 3.83349329e-01 -4.02633935e-01
-2.34472200e-01 5.27914941e-01 -9.89852250e-02 9.00033023e-03
1.09885037e+00 -5.33644333e-02 -6.55551732e-01 6.25440121e-01
1.17149186e+00 5.01664877e-01 9.15149748e-02 -1.00054108e-01
5.68286896e-01 -2.62257576e-01 1.17666155e-01 -8.06383789e-01
-4.25542474e-01 2.37714484e-01 2.74578750e-01 6.88902378e-01
7.72062719e-01 -2.26942435e-01 1.38776913e-01 4.99824584e-01
2.72995889e-01 -1.30157876e+00 -5.37536323e-01 3.67135108e-01
7.01189041e-01 -1.04442775e+00 4.61301416e-01 -5.97518206e-01
-9.33029577e-02 1.52495015e+00 1.19925261e-01 -2.73578793e-01
7.66275823e-01 1.90186769e-01 -3.65032554e-01 -5.09890579e-02
-7.05047607e-01 6.55798763e-02 4.54191506e-01 7.75640979e-02
9.17019367e-01 8.36930424e-02 -1.67838955e+00 4.92124021e-01
-2.21930072e-01 1.00191392e-01 7.78560400e-01 8.86587739e-01
-9.56783593e-01 -1.30473995e+00 -5.39871335e-01 9.61451530e-01
-7.64215946e-01 -1.96783543e-01 -7.25925565e-01 4.49233174e-01
2.68546969e-01 1.09811544e+00 1.80207536e-01 -1.52281746e-01
1.69905946e-01 3.17367911e-01 4.34453070e-01 -5.15455067e-01
-1.05330122e+00 -7.00009391e-02 5.67491353e-01 2.28835225e-01
-6.64096415e-01 -9.74634945e-01 -1.08750892e+00 -3.90562594e-01
-5.16560674e-01 8.69239926e-01 7.51273215e-01 1.07788086e+00
3.74717474e-01 5.03502667e-01 6.36062980e-01 -4.20914471e-01
-3.77649337e-01 -1.30762982e+00 -2.52783269e-01 4.63954687e-01
1.03805058e-01 -4.15402710e-01 -4.19428617e-01 3.99232283e-02] | [12.071832656860352, 9.562891006469727] |
785e2e21-34bd-48c6-8110-410f8d4e5012 | mild-multi-index-hashing-for-loop-closure | 1702.08780 | null | http://arxiv.org/abs/1702.08780v1 | http://arxiv.org/pdf/1702.08780v1.pdf | MILD: Multi-Index hashing for Loop closure Detection | Loop Closure Detection (LCD) has been proved to be extremely useful in global
consistent visual Simultaneously Localization and Mapping (SLAM) and
appearance-based robot relocalization. Methods exploiting binary features in
bag of words representation have recently gained a lot of popularity for their
efficiency, but suffer from low recall due to the inherent drawback that high
dimensional binary feature descriptors lack well-defined centroids. In this
paper, we propose a realtime LCD approach called MILD (Multi-Index Hashing for
Loop closure Detection), in which image similarity is measured by feature
matching directly to achieve high recall without introducing extra
computational complexity with the aid of Multi-Index Hashing (MIH). A
theoretical analysis of the approximate image similarity measurement using MIH
is presented, which reveals the trade-off between efficiency and accuracy from
a probabilistic perspective. Extensive comparisons with state-of-the-art LCD
methods demonstrate the superiority of MILD in both efficiency and accuracy. | ['Lu Fang', 'Lei Han'] | 2017-02-28 | null | null | null | null | ['loop-closure-detection'] | ['computer-vision'] | [-1.02317110e-02 -5.14495730e-01 -4.12454456e-01 -2.40184397e-01
-8.64445090e-01 -4.64022189e-01 9.06613469e-01 7.87954569e-01
-7.38390863e-01 3.82678181e-01 -1.44598410e-02 1.02129295e-01
-3.03664893e-01 -6.35923982e-01 -5.61641514e-01 -6.44629776e-01
-3.01814467e-01 3.88697386e-01 4.11248386e-01 -7.56219774e-02
5.35277903e-01 6.84752703e-01 -1.88426888e+00 -5.89081049e-01
5.36342382e-01 1.00009298e+00 4.15208310e-01 2.47450098e-01
1.17872693e-01 4.14249986e-01 -2.86194533e-01 2.29638368e-02
1.42629534e-01 -2.14191914e-01 -5.38558602e-01 -3.59268427e-01
4.62723732e-01 -2.09126845e-01 -5.30947447e-01 1.21195304e+00
5.27072668e-01 2.10022762e-01 6.43998802e-01 -1.52254128e+00
-4.03653204e-01 -1.31586030e-01 -7.40425169e-01 -1.64017901e-01
8.72743011e-01 -3.57230663e-01 1.02944720e+00 -1.08215368e+00
7.12430298e-01 1.28399420e+00 9.29013729e-01 -6.77195191e-02
-1.24372947e+00 -4.82261986e-01 -4.80980724e-01 3.11204731e-01
-2.04606938e+00 -2.34991819e-01 4.21866655e-01 -3.22192460e-01
8.71182740e-01 8.01645592e-02 6.06706500e-01 4.60067779e-01
4.96634722e-01 4.50485080e-01 1.07635045e+00 -5.67377090e-01
2.28443876e-01 7.42292851e-02 -3.27074016e-03 1.15351522e+00
6.98503375e-01 3.11906189e-01 -7.65268803e-01 -4.95611370e-01
7.50028133e-01 2.49906242e-01 -1.34071231e-01 -1.21796656e+00
-1.42720950e+00 1.11834264e+00 8.18580210e-01 2.50688106e-01
-2.58323193e-01 4.23298419e-01 4.88396645e-01 2.37362817e-01
-8.57106373e-02 3.10748607e-01 2.48664171e-01 -9.29905772e-02
-8.53516281e-01 4.00848478e-01 4.46616322e-01 1.10270333e+00
1.17557931e+00 -6.06144428e-01 1.61043846e-03 5.92670143e-01
4.98772264e-01 8.24696839e-01 5.65457225e-01 -7.47454703e-01
-5.11776879e-02 5.95144510e-01 2.53863901e-01 -1.54111004e+00
-5.19242227e-01 -9.55087766e-02 -8.88619363e-01 2.16540769e-01
-7.21699744e-02 7.64642417e-01 -5.69806695e-01 1.67005420e+00
4.09390330e-01 -1.30452126e-01 3.29545885e-02 9.42041159e-01
4.83523756e-01 5.56720436e-01 -3.24225992e-01 -8.03696215e-02
1.36381376e+00 -7.20699430e-01 -6.19415641e-01 -6.17877766e-02
7.83755958e-01 -8.70582044e-01 7.18202293e-01 -3.54981348e-02
-5.34607410e-01 -5.14181733e-01 -1.40783656e+00 -2.07886428e-01
-4.97293651e-01 3.87812704e-02 7.29787230e-01 5.20115376e-01
-1.10182655e+00 2.01476753e-01 -7.37057328e-01 -8.77530158e-01
-5.24768643e-02 4.52591598e-01 -8.54273736e-01 -2.65832722e-01
-9.62744117e-01 1.13855827e+00 5.02321839e-01 -3.49615484e-01
-5.47090292e-01 -1.12547666e-01 -1.29492855e+00 -3.53401899e-01
3.82796340e-02 -4.01234388e-01 9.33435619e-01 2.55021639e-02
-1.17159951e+00 1.01332223e+00 -3.48093450e-01 -7.25933194e-01
3.32618088e-01 -2.01175332e-01 -1.49417203e-02 3.92169833e-01
3.56150657e-01 8.06390703e-01 7.66015232e-01 -1.04609442e+00
-7.20776856e-01 -5.55879653e-01 -1.99473426e-02 3.63964289e-01
-4.20624055e-02 -2.39260897e-01 -4.16703075e-01 -2.54015416e-01
8.01461041e-01 -1.09878027e+00 -1.21472389e-01 4.18737382e-01
1.48646580e-03 -3.17886442e-01 9.59354758e-01 -6.18244372e-02
1.13706684e+00 -2.12697625e+00 -1.34607144e-02 2.78084695e-01
2.09490925e-01 1.48452623e-02 1.01816326e-01 7.75655150e-01
5.77390611e-01 -4.25834179e-01 -7.64411613e-02 -4.34641719e-01
1.33987486e-01 3.06543440e-01 -2.97077030e-01 1.19522476e+00
-7.92140216e-02 7.12358177e-01 -1.22468972e+00 -8.17696393e-01
7.62042105e-01 4.36741441e-01 -2.66484410e-01 1.28192246e-01
3.64582092e-01 6.40848354e-02 -3.41512293e-01 6.71570301e-01
8.49263310e-01 -1.89075112e-01 -2.30642147e-02 -2.54787087e-01
-4.18076813e-01 9.01870057e-02 -1.12057257e+00 2.02129292e+00
-3.67467612e-01 5.37459433e-01 -1.45873234e-01 -6.89893484e-01
1.19670701e+00 -7.73984045e-02 3.15662593e-01 -8.84281039e-01
2.11504027e-02 4.54926014e-01 -6.33978605e-01 -4.18585166e-02
1.12930012e+00 2.05467090e-01 -4.76903498e-01 3.45369190e-01
-2.09106449e-02 -4.03933078e-01 -4.64875959e-02 1.99620739e-01
1.03318846e+00 1.02968335e-01 1.00977290e+00 -3.19541961e-01
6.63378000e-01 1.70226693e-01 2.22619504e-01 1.01655591e+00
-5.23575008e-01 5.70203424e-01 -1.66422978e-01 -2.85476685e-01
-1.03122783e+00 -9.93655384e-01 -2.25941285e-01 5.93779325e-01
1.12550271e+00 -7.30529785e-01 -3.62105936e-01 -1.92442447e-01
3.37288290e-01 6.98594376e-02 -5.07801950e-01 -3.10003459e-01
-2.99934030e-01 -3.29249352e-01 6.12263918e-01 9.97268707e-02
6.60093606e-01 -4.88147855e-01 -1.23195601e+00 1.30107300e-02
-3.03623438e-01 -1.08180082e+00 -1.86191663e-01 2.87446797e-01
-6.69720292e-01 -8.99538696e-01 -6.30472720e-01 -9.48093593e-01
6.17789805e-01 1.08837676e+00 5.76860607e-01 -6.71723634e-02
-4.38538790e-01 5.23466825e-01 -5.35951793e-01 1.32111445e-01
-2.51851171e-01 3.38365920e-02 4.97607231e-01 -3.18141788e-01
5.49406350e-01 -3.20804417e-01 -5.75163901e-01 4.52306002e-01
-7.76203752e-01 -1.80176064e-01 8.24883401e-01 9.64059114e-01
7.66203880e-01 -3.11858326e-01 -7.69606745e-03 -1.01661548e-01
3.29852164e-01 -2.72701621e-01 -8.60048175e-01 1.79094419e-01
-7.48818219e-01 3.47085327e-01 2.01867949e-02 -1.05547234e-01
-2.53860205e-01 1.89770862e-01 2.07676336e-01 -4.42863852e-01
1.84413984e-01 4.12042409e-01 2.91978866e-01 -8.68601501e-01
5.28609157e-01 6.35539174e-01 2.97303706e-01 -1.47160143e-01
4.56998676e-01 7.57939816e-01 7.17149734e-01 -3.03122222e-01
9.02947783e-01 7.31887996e-01 4.69004422e-01 -7.88044155e-01
-4.53902811e-01 -1.12603676e+00 -6.23290658e-01 -8.83267447e-03
4.11734760e-01 -1.21835780e+00 -9.60240722e-01 3.77986252e-01
-1.11403322e+00 4.56317663e-01 5.25068566e-02 8.25022578e-01
-8.85347784e-01 6.98776245e-01 -2.14165941e-01 -8.35753620e-01
-2.17218816e-01 -1.06034005e+00 1.36852133e+00 1.58098221e-01
-8.16196278e-02 -5.20993292e-01 3.19400072e-01 -1.15910567e-01
2.68191963e-01 2.63514876e-01 6.08722746e-01 -2.54086435e-01
-6.96558237e-01 -6.22412443e-01 -6.12292588e-01 -2.05058724e-01
5.90809435e-02 -5.03655136e-01 -7.01252341e-01 -7.68974841e-01
-1.79954663e-01 -4.21788573e-01 6.40642107e-01 1.82539448e-01
4.16158080e-01 5.66315539e-02 -5.91556966e-01 5.11202753e-01
1.73170805e+00 -3.96150835e-02 4.65814620e-01 5.65446556e-01
3.41430217e-01 2.56407470e-01 1.32847595e+00 7.40270495e-01
4.53995913e-01 9.58609760e-01 5.61338246e-01 1.69267535e-01
-8.58480632e-02 -5.96453667e-01 2.01203495e-01 7.05214500e-01
4.30661410e-01 7.45522231e-02 -7.97742724e-01 7.05682278e-01
-2.02864981e+00 -6.90762877e-01 1.47237867e-01 2.69080782e+00
3.81804526e-01 -1.98113397e-02 -1.20214984e-01 2.04353333e-01
8.69414985e-01 3.54171187e-01 -6.91087395e-02 -1.25145704e-01
-1.63274363e-01 -2.40333050e-01 8.68939519e-01 5.96224546e-01
-1.20229661e+00 8.79503489e-01 5.63207483e+00 1.00433254e+00
-8.86498868e-01 1.58441320e-01 -1.70142099e-01 6.37374818e-01
7.17610791e-02 2.96823084e-01 -9.18764949e-01 1.39937997e-01
4.42795247e-01 -1.00574657e-01 1.95055276e-01 8.38360012e-01
-2.45351329e-01 -7.53639877e-01 -7.35749066e-01 1.60471582e+00
3.70997816e-01 -1.06956267e+00 1.16941251e-01 2.22830549e-01
4.83962446e-01 1.90581024e-01 1.13558723e-02 -3.88917737e-02
-1.24269143e-01 -7.16970563e-01 8.48560989e-01 2.00043082e-01
9.07398760e-01 -7.21091330e-01 7.44394481e-01 1.72962785e-01
-1.49873435e+00 -1.52967097e-02 -6.80127978e-01 -1.19974837e-01
8.19450244e-02 5.01946747e-01 -1.02863288e+00 6.20581269e-01
5.90516031e-01 5.37474453e-01 -4.83171731e-01 1.31137812e+00
5.81942610e-02 -3.16296041e-01 -6.45886004e-01 -2.81768471e-01
4.55360323e-01 -1.67338654e-01 6.43761992e-01 9.93671715e-01
5.40704310e-01 -3.63660872e-01 3.30071032e-01 5.53936422e-01
3.10113043e-01 1.55224487e-01 -1.02516627e+00 1.90916404e-01
8.93748581e-01 1.00395226e+00 -7.60221124e-01 -2.82256734e-02
-3.47144485e-01 1.19905543e+00 2.54361391e-01 -2.36951321e-01
-6.68864012e-01 -7.94000566e-01 6.33494318e-01 -1.26024522e-02
4.12863791e-01 -8.22514772e-01 1.91759735e-01 -7.56217897e-01
-6.19500019e-02 -2.47231454e-01 1.71546146e-01 -5.44398189e-01
-7.42843330e-01 4.84958678e-01 5.31378500e-02 -1.64308405e+00
-2.60307640e-01 -3.60354215e-01 2.55555481e-01 4.33790267e-01
-1.60966671e+00 -1.25448191e+00 -5.74089944e-01 6.37855351e-01
4.28621136e-02 2.01846153e-01 9.14439261e-01 1.90352738e-01
1.35692954e-01 5.97562671e-01 5.16033471e-01 -1.98924735e-01
8.80875945e-01 -8.90276611e-01 4.09498066e-01 5.72515786e-01
2.80186623e-01 7.27516890e-01 8.78087759e-01 -5.16306281e-01
-1.82485604e+00 -8.79777610e-01 1.22014678e+00 -1.21125013e-01
5.42691708e-01 -4.12875652e-01 -4.11070049e-01 2.03552723e-01
-4.51856941e-01 2.85750002e-01 3.06962639e-01 -1.83698073e-01
-6.95817828e-01 -2.08787724e-01 -1.25559711e+00 4.64201152e-01
9.81636286e-01 -9.27630305e-01 -5.95892787e-01 1.24353997e-01
6.17496848e-01 -3.48092318e-01 -8.10951233e-01 5.75817883e-01
7.62147129e-01 -1.11785758e+00 1.07775128e+00 5.26500106e-01
-4.14479136e-01 -8.07571590e-01 -7.36961126e-01 -6.46712780e-01
-3.29736322e-01 -4.91825819e-01 -2.30520330e-02 8.76268983e-01
-3.28588665e-01 -7.07341254e-01 3.73265505e-01 -2.15932265e-01
1.94955289e-01 -3.92943025e-01 -1.28761816e+00 -1.09585690e+00
-7.23965108e-01 -1.03837945e-01 3.64774317e-01 5.26329339e-01
2.46059030e-01 2.38604352e-01 -4.82988179e-01 3.94117743e-01
8.64662230e-01 5.13553679e-01 9.41509187e-01 -1.29737008e+00
2.00469285e-01 -1.57341197e-01 -1.34645128e+00 -1.31516623e+00
-8.46385360e-02 -6.82162821e-01 3.93508315e-01 -1.22982180e+00
3.31002116e-01 -6.79290533e-01 -3.35441232e-01 1.88819781e-01
2.81702578e-01 6.59342587e-01 1.91515297e-01 6.32063031e-01
-1.18733466e+00 9.01549637e-01 6.36675358e-01 3.80021445e-02
6.37429580e-02 -2.38571495e-01 1.07596498e-02 3.67028177e-01
3.23856950e-01 -6.81993484e-01 -1.42852515e-01 -1.38106018e-01
1.41522720e-01 3.13920677e-02 5.72657526e-01 -1.40179729e+00
5.18141150e-01 3.08428228e-01 -2.64252792e-03 -8.88924181e-01
6.94380999e-01 -8.65286469e-01 5.05851582e-02 9.80191708e-01
-7.61682391e-02 3.76917183e-01 -1.26889825e-01 9.81067538e-01
-4.92738336e-01 -2.99768597e-01 7.58213937e-01 8.21220130e-02
-1.16175973e+00 1.58635467e-01 -2.38741204e-01 -4.83119816e-01
1.07866418e+00 -4.74590182e-01 -1.05502687e-01 -4.23285037e-01
5.44965640e-02 -1.02096461e-01 1.08218253e+00 5.20514429e-01
8.10717642e-01 -1.53575313e+00 -2.59143353e-01 2.90851533e-01
8.50561380e-01 -3.10282677e-01 -2.90210024e-02 9.75030184e-01
-8.28354061e-01 7.33424067e-01 -2.33617738e-01 -1.09721482e+00
-1.19981825e+00 7.01580644e-01 -1.31812304e-01 -9.62562338e-02
-6.05585098e-01 5.39074481e-01 -1.04204174e-02 -5.01938641e-01
3.68355274e-01 -1.49278358e-01 2.54023492e-01 -5.59785292e-02
4.71977562e-01 3.07766736e-01 1.29709858e-02 -1.07978296e+00
-7.30665505e-01 1.03895509e+00 8.06765854e-02 -9.95194763e-02
8.33307147e-01 -6.30119920e-01 -2.79791296e-01 4.64785606e-01
1.37765086e+00 -2.29734197e-01 -8.61171663e-01 -5.50789177e-01
2.77603000e-01 -7.43800998e-01 8.06006193e-02 -2.27790587e-02
-2.57526934e-01 5.82101107e-01 9.83767867e-01 -1.00249179e-01
7.22765148e-01 2.52285954e-02 8.47722173e-01 6.93065763e-01
1.25898457e+00 -7.54579544e-01 -1.78947914e-02 6.04059458e-01
5.54474413e-01 -1.48319256e+00 2.82892823e-01 -2.49738500e-01
-1.66619822e-01 9.61695611e-01 1.42310873e-01 -3.45123976e-01
4.82889444e-01 -2.60368418e-02 -2.17113882e-01 -1.86260924e-01
-9.63339061e-02 -4.05076772e-01 -1.16974525e-02 6.51881099e-01
-1.40835658e-01 -3.95932468e-03 -5.30223906e-01 -2.13624373e-01
-2.19484810e-02 -2.70606548e-01 6.92213848e-02 1.26755023e+00
-8.86023939e-01 -8.93596947e-01 -5.14271200e-01 -1.74940582e-02
2.55722255e-02 4.54162732e-02 5.11998795e-02 9.05858338e-01
-1.13995895e-01 8.05182040e-01 -1.71658229e-02 -5.18027365e-01
-5.07263690e-02 -3.54013950e-01 7.48449683e-01 -1.66001067e-01
-1.30359799e-01 -2.04954147e-01 -4.77059871e-01 -8.86654317e-01
-6.09905660e-01 -5.56180239e-01 -1.27018237e+00 -2.08893478e-01
-5.03073871e-01 4.21777397e-01 9.99136806e-01 6.38106108e-01
5.23049474e-01 -4.10509378e-01 7.55715966e-01 -9.77085173e-01
-7.65476108e-01 -6.74688697e-01 -7.18077958e-01 2.14350790e-01
6.43066108e-01 -1.12717748e+00 -3.36159557e-01 -4.83846754e-01] | [7.435938358306885, -2.1201446056365967] |
919e91eb-8f6e-4b52-9484-53b24d08bb23 | traffic-analytics-development-kits-tadk | 2208.07558 | null | https://arxiv.org/abs/2208.07558v1 | https://arxiv.org/pdf/2208.07558v1.pdf | Traffic Analytics Development Kits (TADK): Enable Real-Time AI Inference in Networking Apps | Sophisticated traffic analytics, such as the encrypted traffic analytics and unknown malware detection, emphasizes the need for advanced methods to analyze the network traffic. Traditional methods of using fixed patterns, signature matching, and rules to detect known patterns in network traffic are being replaced with AI (Artificial Intelligence) driven algorithms. However, the absence of a high-performance AI networking-specific framework makes deploying real-time AI-based processing within networking workloads impossible. In this paper, we describe the design of Traffic Analytics Development Kits (TADK), an industry-standard framework specific for AI-based networking workloads processing. TADK can provide real-time AI-based networking workload processing in networking equipment from the data center out to the edge without the need for specialized hardware (e.g., GPUs, Neural Processing Unit, and so on). We have deployed TADK in commodity WAF and 5G UPF, and the evaluation result shows that TADK can achieve a throughput up to 35.3Gbps per core on traffic feature extraction, 6.5Gbps per core on traffic classification, and can decrease SQLi/XSS detection down to 4.5us per request with higher accuracy than fixed pattern solution. | ['Shuo Dai', 'Xiaobo Liu', 'Weigang Li', 'Jianwei Ma', 'Yingqi Liu', 'Wenjun Zhu', 'Xiahui Yu', 'Ying Wang', 'Harry Chang', 'Kun Qiu'] | 2022-08-16 | null | null | null | null | ['traffic-classification'] | ['miscellaneous'] | [-2.80810058e-01 -7.50729799e-01 -1.62451357e-01 -2.26352677e-01
2.42411017e-01 -4.91794288e-01 1.85751781e-01 -2.47136548e-01
-1.35336921e-01 2.80721664e-01 -5.54035902e-01 -1.26662397e+00
-1.12999722e-01 -1.08975255e+00 -7.46737048e-02 -3.79624218e-01
-1.71393633e-01 6.54105783e-01 8.01676035e-01 -4.84821945e-03
4.27100986e-01 1.00387025e+00 -1.65472734e+00 6.48385286e-01
3.59937340e-01 1.80179858e+00 -3.69511753e-01 7.80077040e-01
-6.95265889e-01 1.10017145e+00 -8.80822182e-01 -3.02084655e-01
6.69878364e-01 3.05920213e-01 -4.39801902e-01 -1.99734062e-01
3.70541275e-01 -6.36634171e-01 -6.75912082e-01 7.54035771e-01
1.84054881e-01 -7.15653777e-01 5.28257787e-02 -2.01450276e+00
2.08068155e-02 3.83725137e-01 -5.09907663e-01 7.99105167e-01
6.41732588e-02 8.09106112e-01 4.69911903e-01 -3.78445923e-01
3.09768349e-01 1.20329463e+00 3.41897696e-01 1.86426520e-01
-7.77824461e-01 -1.27156234e+00 -2.94443548e-01 3.79994780e-01
-1.12951338e+00 -5.79276085e-01 6.59813464e-01 -4.40173775e-01
1.19835448e+00 6.57883823e-01 4.06897604e-01 1.01926839e+00
2.47228801e-01 4.30730015e-01 8.93894613e-01 -3.40755358e-02
2.45032415e-01 3.54984105e-01 8.29153299e-01 6.61235929e-01
7.42842972e-01 3.37517187e-02 -3.52357864e-01 -4.18043494e-01
4.83005702e-01 1.72800541e-01 2.58890480e-01 3.46364766e-01
-9.44420516e-01 4.32247192e-01 -4.93928753e-02 8.77325535e-02
-6.69505656e-01 3.84262741e-01 1.18033922e+00 6.05690300e-01
9.47085693e-02 2.78729379e-01 -5.89938879e-01 -6.05423272e-01
-8.03778887e-01 -2.62256414e-01 1.11617577e+00 1.26219428e+00
7.52440155e-01 7.07130551e-01 -1.16316095e-01 -6.20052181e-02
2.07138687e-01 8.98452818e-01 2.55447239e-01 -9.70071495e-01
3.72986615e-01 8.20426464e-01 -5.80737948e-01 -1.23311293e+00
-2.82172203e-01 -2.97057956e-01 -5.99950671e-01 1.87778518e-01
2.84144759e-01 -3.98248106e-01 -7.05443382e-01 9.78934765e-01
1.58176780e-01 4.45781380e-01 -3.24118227e-01 5.57751477e-01
3.28011125e-01 7.73651779e-01 4.98754866e-02 -1.86706826e-01
1.40111983e+00 -7.24935532e-01 -5.25478184e-01 1.74126849e-01
4.34377462e-01 -7.51481295e-01 1.09320045e+00 6.22682691e-01
-5.70148826e-01 -5.36720991e-01 -7.49428928e-01 4.96512979e-01
-5.78131199e-01 -3.07559878e-01 9.42745268e-01 1.27384722e+00
-8.46771359e-01 2.80233659e-02 -7.06036270e-01 -5.08848310e-01
5.71782708e-01 6.88547909e-01 2.74014045e-02 2.00392023e-01
-7.07783043e-01 2.72384763e-01 4.84282374e-01 -4.87657934e-01
-7.82613695e-01 -1.09733891e+00 -8.97338912e-02 2.87145704e-01
5.77791929e-01 -3.81571919e-01 7.83937275e-01 -6.81563199e-01
-1.48281288e+00 4.64295119e-01 1.50354534e-01 -6.05908513e-01
4.46284302e-02 1.00235239e-01 -1.13114488e+00 2.76186883e-01
-1.09845184e-01 -1.26961604e-01 8.93134713e-01 -5.85608184e-01
-7.43658662e-01 -1.95527703e-01 -1.24646932e-01 -9.02090907e-01
-7.52649426e-01 7.11419761e-01 -1.77629367e-01 -2.31062084e-01
-5.34221947e-01 -6.84679925e-01 1.37100562e-01 5.93162216e-02
-4.48991299e-01 -9.21989009e-02 1.85780644e+00 -1.94090724e-01
1.60939705e+00 -2.11251521e+00 -1.01109064e+00 9.10759628e-01
4.86977100e-01 1.06469977e+00 2.55339921e-01 3.33925188e-02
1.54214546e-01 2.59714812e-01 4.95487362e-01 3.94769609e-01
2.09652167e-03 1.03390723e-01 -7.63935983e-01 -6.09234497e-02
1.79532394e-01 6.11338735e-01 -5.05209804e-01 -4.62786049e-01
4.82130587e-01 -1.00826994e-01 -7.04184890e-01 1.95863813e-01
-1.82452127e-01 -1.06309824e-01 -6.58097863e-01 1.17336595e+00
8.24015081e-01 -3.99238318e-01 -1.77465123e-03 -5.55015087e-01
-3.86221319e-01 3.30205828e-01 -1.00681639e+00 6.58608675e-01
-3.97467166e-01 8.33999336e-01 1.09148733e-01 -7.62368023e-01
1.02151716e+00 1.71561167e-01 6.84451401e-01 -7.78340876e-01
3.90017033e-01 1.84269026e-01 1.80852771e-01 -4.70493466e-01
-9.93697569e-02 4.57018733e-01 4.09125787e-04 9.15646434e-01
-2.14317039e-01 6.31995380e-01 4.87287819e-01 6.84037060e-02
1.76503980e+00 -7.11695194e-01 -1.67039678e-01 -1.65402934e-01
8.42432678e-01 4.91699636e-01 6.93033040e-01 6.92933500e-01
-3.37464184e-01 -4.33681935e-01 6.37831688e-01 -9.46074069e-01
-1.08016789e+00 -1.07529175e+00 4.56742384e-03 1.08806670e+00
-2.28995383e-01 -6.96871459e-01 -4.87722665e-01 -5.83503485e-01
2.43441522e-01 6.01327837e-01 3.18835884e-01 -9.79430694e-03
-8.97018969e-01 -6.33827448e-01 9.75352108e-01 3.15485477e-01
1.16329002e+00 -8.45452666e-01 -5.01783669e-01 4.93958950e-01
5.47894418e-01 -1.63037562e+00 -1.18986979e-01 -1.82382345e-01
-6.92858875e-01 -9.58497107e-01 5.92539012e-01 -2.42686376e-01
2.07458094e-01 4.23597991e-01 9.28858280e-01 2.56888956e-01
-8.11898351e-01 2.00234711e-01 1.68652702e-02 -6.22469783e-01
-2.88970441e-01 -9.34340879e-02 4.24204439e-01 1.96912974e-01
1.02519524e+00 -9.64223027e-01 -2.94496238e-01 5.56790531e-01
-5.54284155e-01 -3.11828494e-01 6.48357809e-01 1.16641857e-01
1.20808803e-01 3.23902875e-01 4.56149250e-01 -9.79370415e-01
5.88719726e-01 -7.66669571e-01 -9.05923545e-01 1.95049215e-02
-8.21675718e-01 -2.44878024e-01 1.15473461e+00 -6.07494473e-01
-5.77232122e-01 -3.31958562e-01 1.30846784e-01 -8.94794524e-01
-2.51460284e-01 -1.03156097e-01 -2.52762347e-01 -3.67026627e-01
5.53526282e-01 5.61534017e-02 3.23157549e-01 -1.88230500e-01
-2.39265323e-01 1.27013981e+00 3.71766627e-01 -6.49321914e-01
9.93069112e-01 6.15141094e-01 1.06236309e-01 -8.05263102e-01
-2.13397697e-01 -2.83116609e-01 3.13994139e-02 -3.30538571e-01
5.26311934e-01 -4.75980103e-01 -1.47850764e+00 2.77334213e-01
-9.56421316e-01 -9.79517475e-02 -1.27942964e-01 3.93481821e-01
-1.60710320e-01 3.88767481e-01 -8.46300721e-01 -6.01988196e-01
-9.93483245e-01 -1.18937004e+00 4.01016742e-01 2.14897320e-01
-2.08983615e-01 -4.02266771e-01 -5.00048101e-01 3.99447888e-01
1.36796272e+00 -6.71680570e-02 1.02967238e+00 -8.71874511e-01
-1.10396254e+00 -3.36266786e-01 -7.95997083e-01 4.22605991e-01
1.03930600e-01 6.97104812e-01 -7.98124433e-01 1.96034741e-02
1.61571994e-01 1.28267378e-01 1.09089866e-01 -2.25294620e-01
1.75604725e+00 -4.77818072e-01 -3.25956047e-01 8.85862410e-01
1.27796161e+00 7.48155236e-01 6.79420650e-01 2.41404131e-01
6.66601181e-01 1.02606095e-01 4.89497721e-01 5.00908494e-01
-1.45961747e-01 5.66109359e-01 5.22793412e-01 2.73250163e-01
-1.14616968e-01 2.67876893e-01 5.83920896e-01 8.19093347e-01
1.39292568e-01 -3.66273612e-01 -1.12410164e+00 -3.69552001e-02
-1.39439356e+00 -1.23324668e+00 -3.88771534e-01 1.87969148e+00
2.31134906e-01 9.23384666e-01 4.73978609e-01 3.66878957e-01
7.08954453e-01 -2.27430075e-01 -6.22226536e-01 -1.01913154e+00
2.44926646e-01 5.55655420e-01 8.77126992e-01 -1.37118623e-01
-6.83847129e-01 9.57637370e-01 6.03769922e+00 1.16355228e+00
-1.65210819e+00 -1.29408687e-02 3.51559967e-01 1.88309215e-02
6.57452792e-02 5.05180135e-02 -9.23068821e-01 1.13742113e+00
1.66499484e+00 -3.21029991e-01 4.47878718e-01 1.31882370e+00
2.25504577e-01 3.18903357e-01 -7.94521928e-01 1.12526751e+00
-4.75841671e-01 -1.59724998e+00 3.50988775e-01 4.43740755e-01
-1.40895545e-01 4.74735796e-02 -2.46756867e-01 4.12193179e-01
9.89599898e-02 -6.05641305e-01 8.61233100e-02 3.03190321e-01
8.37121546e-01 -8.31148207e-01 7.81290829e-01 1.73966738e-03
-1.03159320e+00 -4.31441367e-01 -2.33090699e-01 -1.93695217e-01
1.13363072e-01 7.81645596e-01 -1.02137280e+00 9.93991047e-02
7.35885203e-01 2.17125937e-01 -4.57262248e-01 7.87315667e-01
5.60271323e-01 1.17445838e+00 -8.14092278e-01 -2.21364796e-01
6.57958612e-02 -1.74111858e-01 5.55954516e-01 1.25450313e+00
1.69277757e-01 -4.45425101e-02 1.68246791e-01 6.41594112e-01
-5.19989757e-03 -1.82718307e-01 -4.85593468e-01 -3.24778229e-01
7.97064185e-01 1.49201715e+00 -6.06103063e-01 -5.74727774e-01
-5.64056754e-01 5.56015670e-01 -2.02690214e-01 -4.29760739e-02
-1.05438030e+00 -8.07868361e-01 1.39171195e+00 7.57232189e-01
1.87233374e-01 -4.89544511e-01 -4.39354092e-01 -7.40367711e-01
-3.01615079e-03 -1.09504569e+00 4.15882975e-01 -4.26308095e-01
-1.31511652e+00 6.53388441e-01 -2.68908560e-01 -1.05411291e+00
4.11149301e-02 -1.05531430e+00 -1.09072506e+00 3.30739200e-01
-9.98440504e-01 -7.18979239e-01 -6.94620192e-01 9.42864776e-01
1.30857497e-01 -8.57700288e-01 6.98438346e-01 9.88904297e-01
-1.16253984e+00 6.87273860e-01 -3.40682775e-01 3.59355628e-01
2.68032253e-01 -6.12660408e-01 5.32921970e-01 7.75134504e-01
-3.44156861e-01 5.86875260e-01 3.50023240e-01 -6.27175689e-01
-2.21936178e+00 -9.56116736e-01 1.49875388e-01 -1.33679688e-01
1.10938251e+00 -3.84437889e-01 -6.73883498e-01 6.53292656e-01
-5.63627891e-02 3.75803202e-01 7.62272775e-01 -4.19644356e-01
-4.89233404e-01 -1.06830359e+00 -1.46282637e+00 5.28446853e-01
9.37745988e-01 -5.33938646e-01 2.22311497e-01 3.76263887e-01
8.54277372e-01 2.11955383e-01 -8.55208397e-01 2.85674214e-01
2.50212461e-01 -8.64451826e-01 8.62359524e-01 -1.09568894e+00
-3.99185866e-01 -6.12118065e-01 -1.93166777e-01 -1.26695380e-01
-3.75064701e-01 -1.03893733e+00 -6.17240071e-01 1.38355052e+00
1.00066915e-01 -8.96781027e-01 1.03013992e+00 7.43704915e-01
-9.42777246e-02 -4.51074332e-01 -8.74076247e-01 -8.95162404e-01
-9.19328153e-01 -8.59419048e-01 1.04296398e+00 8.40956509e-01
-6.74998015e-02 4.11308259e-01 1.46444691e-02 2.85485268e-01
7.45703638e-01 -4.50232392e-03 1.20621657e+00 -1.18484664e+00
-2.02092659e-02 -5.65555811e-01 -1.24166465e+00 -6.46381855e-01
-1.65555496e-02 -4.27418917e-01 -1.03084266e+00 -6.01141036e-01
-3.40867341e-01 -7.60872364e-01 -3.15715492e-01 4.03372139e-01
6.72251940e-01 2.52078362e-02 9.73308831e-02 1.05787486e-01
-6.09364629e-01 -1.87735334e-01 4.69178557e-01 7.11938515e-02
1.12708583e-01 7.11174756e-02 -5.30355871e-01 5.55278778e-01
1.29468644e+00 -3.66242677e-01 -5.34726143e-01 -2.79473454e-01
-1.67927921e-01 -2.54995823e-01 2.17530459e-01 -1.48889840e+00
6.48872495e-01 -3.91243011e-01 1.35134071e-01 -5.71484983e-01
1.36516735e-01 -1.29531133e+00 2.60855675e-01 7.37671912e-01
5.00984550e-01 4.22376752e-01 2.51158506e-01 1.96967483e-01
2.02886611e-02 2.99211174e-01 5.75413585e-01 1.30519122e-01
-8.79014313e-01 6.65347099e-01 -8.84232521e-01 1.92624107e-02
1.46578979e+00 -6.03722155e-01 -8.83023739e-01 2.45797914e-02
-1.44021347e-01 1.17141075e-01 3.19594920e-01 9.25981924e-02
5.61487198e-01 -1.16186845e+00 -3.77495050e-01 6.14446998e-01
-2.12978974e-01 -5.07396758e-01 -2.15981618e-01 7.78728604e-01
-1.14393270e+00 7.18439281e-01 -7.33273268e-01 -5.67519665e-01
-1.29392207e+00 1.04551089e+00 1.89673126e-01 -8.52306783e-02
-3.49630237e-01 3.01531762e-01 -3.89376640e-01 3.95510234e-02
2.83877194e-01 8.06644708e-02 3.04940850e-01 -3.05313617e-01
9.30244684e-01 1.05655217e+00 3.48234594e-01 5.19712046e-02
-6.93029761e-01 2.44531170e-01 -2.16245577e-01 4.92145360e-01
9.72444236e-01 1.50058165e-01 -4.64690983e-01 -1.21058553e-01
1.13854778e+00 7.23939165e-02 -2.40238130e-01 -1.93470076e-01
4.16325778e-01 -7.78346360e-01 3.09975177e-01 -3.57705534e-01
-1.52429378e+00 6.71075821e-01 5.95656216e-01 4.57510263e-01
1.26828575e+00 -5.75174510e-01 1.33653784e+00 5.87603569e-01
5.77192426e-01 -8.99805069e-01 7.79734403e-02 5.63252747e-01
6.40086904e-02 -8.23640406e-01 -9.40859690e-02 -8.33297074e-01
-2.79661506e-01 1.53046596e+00 1.19874060e+00 -6.47312552e-02
1.05022919e+00 1.00737512e+00 -1.33924112e-01 -5.29702365e-01
-1.11150122e+00 -1.81857094e-01 -3.92432064e-01 4.76923704e-01
-9.36950818e-02 1.87264040e-01 9.24515575e-02 5.17910197e-02
-1.03081949e-01 3.80328968e-02 5.43057084e-01 9.70841646e-01
-6.46336377e-01 -9.98124957e-01 -2.47153059e-01 9.98029232e-01
-6.49415791e-01 -5.58753349e-02 -1.62988409e-01 5.65496922e-01
1.89510882e-02 9.82347488e-01 7.61294544e-01 -1.03193355e+00
1.28035679e-01 -9.98226777e-02 -3.37563217e-01 -3.54334563e-01
-7.78939724e-01 -5.89714587e-01 4.61565316e-01 -9.87364352e-01
3.70814264e-01 -1.36974916e-01 -1.11289585e+00 -1.48618138e+00
1.12462148e-01 -4.19324860e-02 8.43890905e-01 4.01225835e-01
9.60285306e-01 3.02014917e-01 9.85024393e-01 -9.08025801e-02
-1.55911058e-01 -5.57883203e-01 -2.33459592e-01 -3.98871824e-02
-3.70709077e-02 -4.64986593e-01 -5.07776558e-01 -4.88324285e-01] | [5.127316951751709, 7.212810516357422] |
7be94abe-9daf-4208-9d4c-303bc6edaa2b | interpretable-and-generalizable-deep-image | 1904.10424 | null | https://arxiv.org/abs/1904.10424v4 | https://arxiv.org/pdf/1904.10424v4.pdf | Interpretable and Generalizable Person Re-Identification with Query-Adaptive Convolution and Temporal Lifting | For person re-identification, existing deep networks often focus on representation learning. However, without transfer learning, the learned model is fixed as is, which is not adaptable for handling various unseen scenarios. In this paper, beyond representation learning, we consider how to formulate person image matching directly in deep feature maps. We treat image matching as finding local correspondences in feature maps, and construct query-adaptive convolution kernels on the fly to achieve local matching. In this way, the matching process and results are interpretable, and this explicit matching is more generalizable than representation features to unseen scenarios, such as unknown misalignments, pose or viewpoint changes. To facilitate end-to-end training of this architecture, we further build a class memory module to cache feature maps of the most recent samples of each class, so as to compute image matching losses for metric learning. Through direct cross-dataset evaluation, the proposed Query-Adaptive Convolution (QAConv) method gains large improvements over popular learning methods (about 10%+ mAP), and achieves comparable results to many transfer learning methods. Besides, a model-free temporal cooccurrence based score weighting method called TLift is proposed, which improves the performance to a further extent, achieving state-of-the-art results in cross-dataset person re-identification. Code is available at https://github.com/ShengcaiLiao/QAConv. | ['Shengcai Liao', 'Ling Shao'] | 2019-04-23 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/1369_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123560443.pdf | eccv-2020-8 | ['generalizable-person-re-identification'] | ['computer-vision'] | [-1.69804052e-01 -5.66445887e-01 -1.54141616e-02 -7.96659708e-01
-7.21506119e-01 -3.80683631e-01 5.49931645e-01 7.75427148e-02
-6.61484897e-01 4.97221619e-01 4.07583028e-01 3.29456270e-01
-1.33112431e-01 -9.21179354e-01 -6.11353040e-01 -3.10451418e-01
2.99903676e-02 5.94991803e-01 -9.79706198e-02 -1.09246135e-01
-2.58538034e-02 3.23432654e-01 -1.50993299e+00 2.10763410e-01
7.51634896e-01 8.42085898e-01 -8.14040229e-02 3.61906648e-01
3.02920312e-01 2.19038010e-01 -5.53341866e-01 -8.83310616e-01
3.08878183e-01 -2.43626282e-01 -7.34772444e-01 -7.58894980e-02
1.04244709e+00 -5.83689153e-01 -9.55533981e-01 1.12526333e+00
8.60262036e-01 4.82069373e-01 4.68616605e-01 -1.15832794e+00
-1.02239776e+00 2.18471602e-01 -3.01200926e-01 2.97988474e-01
5.91813207e-01 3.48148823e-01 7.61645615e-01 -1.09739554e+00
7.29849786e-02 1.38133001e+00 9.13123012e-01 6.90608084e-01
-1.10011733e+00 -1.01789176e+00 2.52515435e-01 4.62828696e-01
-1.64928389e+00 -4.92937297e-01 5.70218146e-01 -3.04291248e-01
8.73799980e-01 3.16962779e-01 5.56111932e-01 1.15883398e+00
-2.85797000e-01 8.70449066e-01 7.57316351e-01 -2.18364209e-01
-1.86207339e-01 1.05336756e-02 2.33256161e-01 5.83156824e-01
2.61144996e-01 4.19844747e-01 -5.82858801e-01 -8.51008743e-02
7.60940671e-01 5.62941670e-01 -3.27255845e-01 -2.84285545e-01
-1.39113951e+00 5.77750385e-01 9.76511300e-01 1.00937806e-01
-9.14509743e-02 3.66667330e-01 5.00293434e-01 4.21618134e-01
2.43517876e-01 1.19086608e-01 -2.83457160e-01 -4.98371050e-02
-8.38138640e-01 5.31297743e-01 2.22964764e-01 9.89905655e-01
8.54252160e-01 -2.70524859e-01 -5.44582963e-01 1.00627315e+00
1.31354988e-01 7.32010305e-01 7.66374826e-01 -6.03325009e-01
7.93638647e-01 6.66117966e-01 1.20296322e-01 -9.91892278e-01
-5.21904409e-01 -5.29930651e-01 -1.09862745e+00 -1.53356150e-01
4.74596500e-01 1.80964340e-02 -9.61806297e-01 1.92709219e+00
9.86391008e-02 4.56122935e-01 -2.43279010e-01 1.07330155e+00
7.95862675e-01 2.50987321e-01 6.40014559e-02 3.16773027e-01
1.39892006e+00 -1.00901854e+00 -3.82385463e-01 -2.85651386e-01
6.78280354e-01 -5.19728601e-01 9.78049994e-01 -6.38438528e-03
-8.30697238e-01 -1.04462397e+00 -8.73569906e-01 -1.37703121e-01
-4.36799526e-01 3.97159785e-01 6.44428730e-01 7.05196142e-01
-1.04930389e+00 6.27110720e-01 -7.43209898e-01 -5.71654201e-01
4.85648990e-01 5.13087928e-01 -5.15650928e-01 -4.22298938e-01
-1.27362394e+00 6.93763316e-01 3.32540661e-01 1.97198853e-01
-5.75092196e-01 -7.97808886e-01 -1.03228235e+00 2.36392599e-02
-8.38704184e-02 -9.30604696e-01 1.15775037e+00 -7.15145409e-01
-1.11513007e+00 1.01141226e+00 -4.28418636e-01 -3.87159586e-01
6.55542672e-01 -4.02724296e-01 -7.38669455e-01 -1.13748416e-01
2.73915023e-01 6.54586911e-01 6.57724619e-01 -8.92186463e-01
-5.86086392e-01 -5.80235779e-01 6.57383204e-02 1.30384624e-01
-7.48394728e-01 3.85634489e-02 -7.12168038e-01 -9.62020993e-01
-1.03755943e-01 -1.02973092e+00 -6.09491207e-02 3.07672054e-01
-2.72582918e-01 -3.92498076e-01 3.12223971e-01 -8.47937644e-01
1.12828934e+00 -2.23425198e+00 -1.96089409e-02 3.55973780e-01
1.96677953e-01 2.97141612e-01 -2.76630580e-01 7.25260973e-02
-2.25546896e-01 -2.24569395e-01 -1.35843173e-01 -7.21342325e-01
1.38652027e-01 -1.41173542e-01 -1.96625665e-01 6.34771228e-01
-4.28758338e-02 1.19530892e+00 -7.27728188e-01 -2.28983641e-01
4.11505550e-01 4.88280147e-01 -5.22305667e-01 1.97184265e-01
4.49379385e-01 3.67262840e-01 -4.10473198e-01 6.21795714e-01
9.09147024e-01 -2.49141589e-01 -5.48170358e-02 -4.87096310e-01
2.54854113e-01 5.75791188e-02 -1.21526098e+00 2.01422787e+00
-4.38147664e-01 4.48727131e-01 -6.23183310e-01 -1.01046515e+00
9.56791580e-01 8.54706839e-02 4.18264389e-01 -9.72304046e-01
-1.81601569e-02 -1.02202012e-03 -2.91825682e-01 -3.53476703e-02
4.24279898e-01 3.78099561e-01 -1.09663419e-01 3.66487861e-01
-2.45215259e-02 6.58956766e-01 -1.40580470e-02 1.14273638e-01
6.50574923e-01 -1.68363117e-02 8.38685594e-03 -4.78438400e-02
6.77008510e-01 -4.81643349e-01 6.76793516e-01 6.97683454e-01
-4.45002198e-01 9.40668941e-01 -4.03576732e-01 -7.60762870e-01
-9.53357160e-01 -1.24350905e+00 -2.19086036e-01 1.12488258e+00
5.78068137e-01 -5.23024321e-01 -7.66601861e-01 -4.76187587e-01
3.67354125e-01 3.22928190e-01 -7.37428069e-01 -5.07457614e-01
-8.38306427e-01 -8.75742674e-01 6.37795985e-01 8.77818346e-01
1.02358723e+00 -7.95346200e-01 -4.81859595e-02 1.33587658e-01
-3.74288350e-01 -8.70720983e-01 -1.13299644e+00 -5.47500372e-01
-6.84954405e-01 -1.09365416e+00 -1.07114542e+00 -9.22546804e-01
7.74717033e-01 5.88064075e-01 9.44126010e-01 3.10263306e-01
-3.85429114e-01 5.39520144e-01 -1.89509943e-01 -1.38779730e-03
2.92824686e-01 -2.69687809e-02 4.18388218e-01 2.94968247e-01
8.99117231e-01 -4.72306132e-01 -9.99117434e-01 5.32745361e-01
-4.15410250e-01 -2.92019993e-01 4.53974724e-01 9.72081900e-01
5.54619491e-01 -1.88524663e-01 4.58519518e-01 -3.79170269e-01
5.23721516e-01 -1.58930942e-01 -3.20470870e-01 5.30343413e-01
-6.59770548e-01 8.63482337e-03 4.01975989e-01 -5.77962637e-01
-9.36574280e-01 -1.37033481e-02 -7.58526698e-02 -4.23631519e-01
-1.95807353e-01 1.60556287e-01 -3.57673794e-01 -1.17012076e-01
7.06071377e-01 4.09702182e-01 -5.96994348e-02 -6.93581879e-01
3.21150512e-01 5.75404882e-01 9.21496511e-01 -6.75169230e-01
1.07975817e+00 4.61461902e-01 -4.62504268e-01 -3.40143949e-01
-7.33650744e-01 -5.91791391e-01 -7.90689647e-01 -3.39387543e-02
6.96289778e-01 -1.26719618e+00 -7.79530406e-01 7.46952176e-01
-1.07127440e+00 -1.95890799e-01 -1.24508947e-01 6.16183996e-01
-3.54728967e-01 4.29285944e-01 -6.59548104e-01 -3.25644135e-01
-4.78444695e-01 -9.31450844e-01 1.02062881e+00 5.06188691e-01
-2.00279802e-01 -9.25044894e-01 5.69351502e-02 5.08009911e-01
4.88213331e-01 2.76440214e-02 3.95607769e-01 -5.46452641e-01
-5.12254119e-01 -5.64802408e-01 -4.76193309e-01 7.68124908e-02
2.20123351e-01 -7.53858984e-01 -1.05278456e+00 -9.20479536e-01
-5.10415554e-01 -2.74140000e-01 1.05074060e+00 1.51356444e-01
1.45798028e+00 -1.59319311e-01 -5.52879930e-01 1.03069270e+00
1.04113448e+00 -3.24585736e-01 7.87563622e-01 4.83934402e-01
8.97415161e-01 3.95111412e-01 5.04315376e-01 4.46236491e-01
7.36741483e-01 1.18312788e+00 4.26950306e-02 -1.83632672e-01
-4.38178122e-01 -3.82288039e-01 2.75752425e-01 4.31499690e-01
-2.96435356e-01 -6.00161143e-02 -5.98489106e-01 5.01499593e-01
-1.96727049e+00 -1.19064569e+00 2.39083603e-01 2.51625681e+00
6.11080706e-01 -9.98627990e-02 3.89142454e-01 -5.60248308e-02
1.05410790e+00 -7.25713000e-02 -8.04369032e-01 2.72720695e-01
-1.44484237e-01 1.32066980e-01 4.50491428e-01 3.40702206e-01
-1.39442396e+00 8.49199057e-01 5.21677589e+00 7.20496476e-01
-8.61568689e-01 1.95303440e-01 4.78345990e-01 -2.37418339e-01
-8.98684561e-02 -3.08841377e-01 -8.81617129e-01 6.12869203e-01
6.67971730e-01 -1.86452731e-01 5.31813204e-01 8.00805330e-01
-7.42507912e-03 4.97182965e-01 -1.36241293e+00 1.66130280e+00
2.89749295e-01 -1.07964706e+00 1.80924907e-01 4.91970479e-02
4.45571959e-01 -8.82316232e-02 3.21570665e-01 4.59232122e-01
1.26239955e-01 -1.05524004e+00 5.09101212e-01 7.16457069e-01
9.69906926e-01 -9.76760626e-01 7.92796791e-01 -8.85434598e-02
-1.50586259e+00 -1.84198424e-01 -6.46422446e-01 5.49546741e-02
1.26676649e-01 3.15911084e-01 -2.92973936e-01 5.16245902e-01
9.84819412e-01 1.03865457e+00 -1.06868935e+00 1.34204137e+00
-8.61814152e-03 2.84061074e-01 -8.37851390e-02 2.98246622e-01
-2.60704577e-01 5.62764406e-02 2.09098011e-01 1.25962663e+00
4.22722578e-01 -7.81016648e-02 3.30895275e-01 7.70647407e-01
-2.16628850e-01 -1.44797713e-01 -2.28960857e-01 5.82071900e-01
6.65252984e-01 1.02577674e+00 -1.02450356e-01 -3.02445948e-01
-5.50801277e-01 1.53733206e+00 5.30302584e-01 5.16619384e-01
-8.15387011e-01 -5.05055726e-01 1.12251735e+00 7.63811618e-02
2.27692261e-01 -1.03160784e-01 1.23038224e-03 -1.50435495e+00
3.61941814e-01 -7.32813299e-01 7.15568483e-01 -3.07878137e-01
-1.87286055e+00 5.21109402e-01 -6.29677773e-02 -1.47768784e+00
-2.56208777e-01 -6.32369041e-01 -6.44512296e-01 1.03121269e+00
-1.40689564e+00 -1.52550673e+00 -7.74555206e-01 1.00278723e+00
2.33901426e-01 -4.91024554e-01 9.02612686e-01 8.14980268e-01
-6.25816464e-01 1.64117002e+00 7.50300586e-02 7.65954733e-01
1.17950296e+00 -1.13237929e+00 8.00947607e-01 9.97487307e-01
1.70148954e-01 9.87935364e-01 2.11958274e-01 -4.76287991e-01
-1.19904089e+00 -1.50902319e+00 9.35288012e-01 -5.93885958e-01
2.75332898e-01 -3.62785935e-01 -9.49768364e-01 7.40715981e-01
-1.88997880e-01 3.92887205e-01 8.52637053e-01 2.93118984e-01
-8.36635113e-01 -4.24602836e-01 -1.17039597e+00 4.66209829e-01
1.55581522e+00 -7.74112046e-01 -4.42160666e-01 4.89789635e-01
6.18156552e-01 -3.79030943e-01 -9.14786041e-01 2.77319551e-01
6.50388479e-01 -7.99928248e-01 1.47055197e+00 -7.78033376e-01
-2.04253882e-01 -4.93952006e-01 -2.19911918e-01 -1.18137276e+00
-8.22843790e-01 -3.49219859e-01 -7.04615116e-02 1.32724321e+00
1.22284256e-01 -7.57486165e-01 8.22481871e-01 9.79287028e-01
-9.92750004e-03 -3.49418759e-01 -9.20156181e-01 -9.73339617e-01
9.38067064e-02 -2.09520414e-01 9.78069186e-01 9.53968704e-01
-2.95417637e-01 9.42220259e-03 -6.22826576e-01 4.03873712e-01
8.98599863e-01 2.29808167e-01 8.68226171e-01 -1.15360940e+00
-3.87150288e-01 -3.80213559e-01 -8.72971773e-01 -1.29724288e+00
2.19978660e-01 -1.22337437e+00 -2.93761164e-01 -1.28318071e+00
4.62406754e-01 -5.83357096e-01 -6.62076890e-01 4.92388248e-01
-4.46492076e-01 5.37713647e-01 2.48173848e-01 4.35239911e-01
-6.83428764e-01 7.02275872e-01 9.73281026e-01 -5.61136663e-01
-2.60731876e-02 1.48655996e-01 -7.10967660e-01 3.56099665e-01
8.48967671e-01 -1.62682384e-01 -2.65304774e-01 -8.48878622e-01
-2.78785765e-01 -5.44316113e-01 8.73006403e-01 -1.31202769e+00
5.34274399e-01 1.43227041e-01 1.01528251e+00 -3.87553096e-01
4.31649953e-01 -4.27451074e-01 -2.25403085e-02 4.55258906e-01
-4.41362739e-01 3.26448113e-01 -2.73711067e-02 6.00669622e-01
-1.75942674e-01 4.67178747e-02 6.88054800e-01 4.27488387e-02
-8.89736414e-01 1.03140032e+00 3.11836660e-01 5.66428825e-02
7.77461946e-01 -3.18747938e-01 -2.71060616e-01 -3.63071561e-01
-6.26682699e-01 3.79871547e-01 4.40238565e-01 6.73601270e-01
8.23367298e-01 -1.92808664e+00 -9.09269333e-01 2.06842542e-01
5.53831935e-01 -9.68692750e-02 5.25515974e-01 4.80623454e-01
-1.32018283e-01 3.55149388e-01 -1.99542478e-01 -4.78560716e-01
-1.06064451e+00 6.39711916e-01 6.32500887e-01 1.89790558e-02
-7.44158864e-01 9.12317932e-01 2.63405442e-01 -7.72350013e-01
3.61109704e-01 6.91837370e-02 -7.89829791e-02 -2.55804330e-01
1.14181626e+00 3.40229154e-01 3.05193737e-02 -7.45848596e-01
-5.38292944e-01 7.48976469e-01 -3.77591461e-01 1.63522199e-01
1.07144189e+00 -1.78419530e-01 1.67429030e-01 -5.66976406e-02
1.40197730e+00 -2.76942253e-01 -1.23609877e+00 -7.19221115e-01
-1.51605293e-01 -7.64722168e-01 -2.50375748e-01 -6.56133533e-01
-1.18405712e+00 7.52425611e-01 1.25271630e+00 -5.60141325e-01
9.76247549e-01 -7.68239424e-02 1.13080704e+00 5.46422303e-01
4.01113600e-01 -9.76901591e-01 2.81457156e-01 3.11121613e-01
8.10692012e-01 -1.51073658e+00 -6.86428174e-02 -1.28407255e-01
-3.83379310e-01 9.96990502e-01 8.10290635e-01 -1.91835925e-01
6.24908864e-01 -2.89886743e-01 -4.46182527e-02 1.49822548e-01
-1.17848106e-01 -2.52367705e-01 6.12913132e-01 9.97484863e-01
3.28983843e-01 3.11339140e-01 1.43850878e-01 7.30475545e-01
-3.90384257e-01 -7.18268454e-02 -1.78546101e-01 5.45774341e-01
-1.28367826e-01 -1.22172689e+00 -2.81238914e-01 5.16412556e-01
1.38332928e-02 -8.54925960e-02 -1.41433850e-01 5.90295494e-01
2.43813038e-01 8.00069928e-01 2.77829170e-01 -7.07085073e-01
5.39522886e-01 -3.38952869e-01 5.52693129e-01 -3.91743809e-01
-4.26539809e-01 -5.64964712e-01 -1.15188308e-01 -6.65307343e-01
-3.03674757e-01 -7.99229741e-01 -1.05041289e+00 -5.77663541e-01
-6.34573400e-02 -8.27789307e-02 1.24520727e-01 9.30299520e-01
4.40120637e-01 2.25158289e-01 5.68999887e-01 -7.64836013e-01
-7.03964114e-01 -9.38225031e-01 -2.42850363e-01 9.51957464e-01
2.74216473e-01 -8.62493455e-01 1.89575891e-04 -2.31965110e-02] | [14.705004692077637, 0.9523134827613831] |
b8de4dd8-117d-40dc-8d17-44375bb9661b | lightesd-fully-automated-and-lightweight | 2305.12266 | null | https://arxiv.org/abs/2305.12266v1 | https://arxiv.org/pdf/2305.12266v1.pdf | LightESD: Fully-Automated and Lightweight Anomaly Detection Framework for Edge Computing | Anomaly detection is widely used in a broad range of domains from cybersecurity to manufacturing, finance, and so on. Deep learning based anomaly detection has recently drawn much attention because of its superior capability of recognizing complex data patterns and identifying outliers accurately. However, deep learning models are typically iteratively optimized in a central server with input data gathered from edge devices, and such data transfer between edge devices and the central server impose substantial overhead on the network and incur additional latency and energy consumption. To overcome this problem, we propose a fully-automated, lightweight, statistical learning based anomaly detection framework called LightESD. It is an on-device learning method without the need for data transfer between edge and server, and is extremely lightweight that most low-end edge devices can easily afford with negligible delay, CPU/memory utilization, and power consumption. Yet, it achieves highly competitive detection accuracy. Another salient feature is that it can auto-adapt to probably any dataset without manually setting or configuring model parameters or hyperparameters, which is a drawback of most existing methods. We focus on time series data due to its pervasiveness in edge applications such as IoT. Our evaluation demonstrates that LightESD outperforms other SOTA methods on detection accuracy, efficiency, and resource consumption. Additionally, its fully automated feature gives it another competitive advantage in terms of practical usability and generalizability. | ['Tie Luo', 'Ronit Das'] | 2023-05-20 | null | null | null | null | ['edge-computing'] | ['time-series'] | [-3.61785650e-01 -6.30112410e-01 -2.35716984e-01 -6.46556690e-02
-1.92545533e-01 -3.50379825e-01 1.50155693e-01 5.12438655e-01
-3.08455735e-01 1.19301878e-01 -4.57931459e-01 -5.80141306e-01
-1.00583002e-01 -7.70920336e-01 -4.60878462e-01 -5.65261841e-01
-1.42304122e-01 4.07828599e-01 2.79226989e-01 1.82912126e-01
-1.68822765e-01 7.17152894e-01 -1.25588500e+00 -3.01608294e-01
7.40237415e-01 1.68510890e+00 -4.42189187e-01 3.17805678e-01
-6.42438829e-02 3.76028389e-01 -4.95327890e-01 -2.29327768e-01
3.68987113e-01 -3.65244709e-02 1.94782764e-02 -2.00724155e-02
-2.55370699e-02 -5.07206619e-01 -4.95159656e-01 9.52065408e-01
6.07259750e-01 -4.60077747e-04 2.89504498e-01 -1.87809730e+00
-2.59017885e-01 1.86048523e-01 -6.79445267e-01 2.95776665e-01
1.45020708e-01 3.55871886e-01 6.87457919e-01 -7.36993551e-01
-1.52218267e-02 5.37638724e-01 8.33676577e-01 2.55289644e-01
-1.07601213e+00 -8.58873129e-01 4.86015260e-01 4.11000431e-01
-1.31973135e+00 -3.00201058e-01 9.54061985e-01 -2.59319752e-01
1.10819376e+00 1.56227008e-01 5.57033241e-01 1.27667058e+00
4.09624279e-01 8.86661828e-01 2.56884515e-01 6.25642911e-02
7.99971402e-01 -2.65826255e-01 6.74237087e-02 4.73914772e-01
5.97566009e-01 -6.68388531e-02 -3.40078950e-01 -3.79262209e-01
5.94685614e-01 7.68423378e-01 1.96346089e-01 -3.86255443e-01
-9.45281982e-01 3.53312045e-01 1.22176953e-01 1.18635878e-01
-8.08660865e-01 2.09698364e-01 1.00139475e+00 4.91796166e-01
3.71757001e-01 1.09307319e-01 -8.07497799e-01 -4.64164406e-01
-5.01395166e-01 -3.19375694e-01 7.09651232e-01 9.60533142e-01
3.63810122e-01 6.26853466e-01 1.62413836e-01 4.74973172e-01
4.24322933e-02 5.21560431e-01 6.20145321e-01 -4.06319320e-01
2.01181024e-01 8.29252183e-01 -8.31266120e-02 -1.15910542e+00
-7.28171766e-01 -4.87808108e-01 -1.36137259e+00 1.70113742e-01
4.85415757e-02 -2.34631881e-01 -7.02713847e-01 1.38483357e+00
4.67553109e-01 5.55643082e-01 -2.00392261e-01 7.81304181e-01
2.63367355e-01 3.08090955e-01 1.21989027e-02 -3.63098264e-01
1.06810582e+00 -6.85855508e-01 -6.79060221e-01 -2.57778436e-01
6.60017431e-01 -5.29974759e-01 1.15280390e+00 6.67858899e-01
-5.89357018e-01 -1.90432414e-01 -1.18302786e+00 4.20891583e-01
-3.28028947e-01 -3.02039180e-02 9.28906381e-01 5.78966737e-01
-5.17610431e-01 4.74344373e-01 -1.36288095e+00 -3.75259221e-01
5.70126414e-01 5.77776194e-01 -1.21921711e-01 1.40354216e-01
-6.58967316e-01 2.70366937e-01 3.83017659e-01 -6.03662804e-03
-4.78303403e-01 -5.29834092e-01 -7.09528089e-01 1.89163700e-01
5.98745525e-01 -6.02145612e-01 1.21764660e+00 -8.17263007e-01
-1.72790468e+00 1.75626889e-01 6.74190670e-02 -4.42600012e-01
5.17150223e-01 -4.76512194e-01 -1.22155654e+00 -2.63179094e-01
-2.37564757e-01 -3.85806590e-01 1.10202062e+00 -4.89796579e-01
-5.49148858e-01 -5.19898713e-01 -3.91046017e-01 -3.97513032e-01
-7.98761308e-01 -2.25900516e-01 -4.22710657e-01 -6.61189675e-01
3.73645127e-01 -9.60371375e-01 -1.74237967e-01 -6.83043301e-02
-5.63621998e-01 -3.81838739e-01 1.41000497e+00 -3.51320744e-01
1.35549998e+00 -2.37483978e+00 -5.33627212e-01 5.02308547e-01
2.73642212e-01 4.48261887e-01 4.84745838e-02 3.50878358e-01
-7.87847638e-02 -1.61225900e-01 1.04537988e-02 -4.44469117e-02
8.52269381e-02 1.11398973e-01 -2.62233287e-01 5.47242761e-01
9.50344577e-02 7.36997962e-01 -8.69566321e-01 1.31133527e-01
4.52458739e-01 1.38582692e-01 -4.90747213e-01 1.28632382e-01
-2.50164211e-01 2.78394938e-01 -5.13256550e-01 9.13479805e-01
4.63123798e-01 -3.19007754e-01 5.55955358e-02 -1.54505506e-01
2.57550001e-01 9.67688859e-02 -1.33748782e+00 1.40423691e+00
-6.09031558e-01 5.37502766e-01 -1.63960338e-01 -1.15521336e+00
9.02836800e-01 3.20729434e-01 9.41449642e-01 -8.71706247e-01
2.95396060e-01 2.61767149e-01 -5.89406975e-02 -4.00820851e-01
1.00484036e-01 4.57470953e-01 -3.62591237e-01 8.04702282e-01
-3.18479031e-01 4.23234493e-01 -7.91306570e-02 -1.11985812e-02
1.54186356e+00 -3.35589975e-01 3.19864839e-01 1.21418715e-01
3.34766328e-01 -4.00963902e-01 9.55276966e-01 6.10805988e-01
-2.25747272e-01 3.20980251e-02 3.61556113e-01 -8.04094791e-01
-8.76014829e-01 -1.15532827e+00 2.37729967e-01 1.04988813e+00
2.00863957e-01 -5.48839390e-01 -3.43144268e-01 -9.21028256e-01
2.90234536e-01 6.63364291e-01 -2.15725023e-02 -6.15393758e-01
-4.40973997e-01 -6.01462305e-01 3.08812171e-01 8.12365115e-01
5.53261340e-01 -8.21268737e-01 -5.79033673e-01 4.05517459e-01
2.40875080e-01 -1.38888502e+00 -4.68814135e-01 9.33063328e-02
-9.38282549e-01 -1.05042803e+00 1.15985222e-01 -3.77061397e-01
6.08916879e-01 1.94155023e-01 9.95562136e-01 -9.56129096e-03
-4.24797297e-01 4.71367598e-01 -1.40785724e-01 -8.03306103e-01
-1.20524660e-01 1.85517728e-01 8.47483039e-01 1.96409851e-01
8.18591952e-01 -8.46614599e-01 -6.45893991e-01 4.61809486e-01
-9.98120248e-01 -4.93776649e-01 6.13233089e-01 6.04690909e-01
6.89520359e-01 5.19946098e-01 8.23163986e-01 -5.97416580e-01
5.20107448e-01 -7.24744916e-01 -9.51452732e-01 -1.41731575e-01
-1.00843751e+00 -1.83466405e-01 1.01461661e+00 -5.67289889e-01
-5.26561975e-01 1.28264287e-02 3.51800881e-02 -6.43623590e-01
-2.76133984e-01 2.66363591e-01 -3.44124585e-01 6.16295896e-02
4.36765671e-01 7.63066411e-02 -2.19098106e-02 -4.92568731e-01
-1.40648052e-01 7.58777261e-01 5.40050268e-01 -3.61412674e-01
9.48303580e-01 3.56091231e-01 1.47807002e-01 -8.47523987e-01
-4.97693896e-01 -4.23181117e-01 -3.38645875e-01 1.00744953e-02
1.96871564e-01 -8.73449385e-01 -1.11144209e+00 6.94508791e-01
-7.69887805e-01 -1.70618847e-01 -3.56645077e-01 5.17491043e-01
-1.90152675e-01 3.90242308e-01 -5.85665226e-01 -7.31365860e-01
-7.10024357e-01 -7.71052897e-01 8.29110086e-01 1.38458639e-01
-4.31088507e-01 -9.43662286e-01 -4.26470816e-01 -2.22753525e-01
7.01999664e-01 3.92258525e-01 8.62262785e-01 -1.01159000e+00
-4.58082259e-01 -8.90201807e-01 -1.24252334e-01 3.45322043e-01
4.66801643e-01 -9.42336544e-02 -6.75867081e-01 -7.20245659e-01
1.82937998e-02 7.56480619e-02 2.73472667e-01 2.18876809e-01
1.70763445e+00 -3.90203148e-01 -2.82840550e-01 5.86389184e-01
1.20187819e+00 3.09039950e-01 2.04481438e-01 2.92095274e-01
7.79227853e-01 -1.17221884e-01 4.79366750e-01 7.26924539e-01
7.96405002e-02 6.50620759e-01 7.28685617e-01 -1.34515032e-01
5.26027501e-01 2.13781092e-03 5.21568477e-01 9.04134154e-01
2.97720402e-01 -3.08871716e-01 -8.78970623e-01 3.68885875e-01
-2.04668331e+00 -6.68475330e-01 9.97029021e-02 2.47664595e+00
1.90140605e-01 5.29217064e-01 4.75606233e-01 4.57088977e-01
4.80917692e-01 -2.51083881e-01 -1.38518119e+00 -4.08353567e-01
2.17677325e-01 -8.25644936e-03 4.91856188e-01 -2.20376909e-01
-1.07960820e+00 6.03061974e-01 5.63543272e+00 6.28789008e-01
-1.38866174e+00 1.25612998e-02 4.94891793e-01 -4.79781657e-01
1.95225492e-01 -4.80977327e-01 -3.20844650e-01 8.00216138e-01
1.09313905e+00 -2.16355443e-01 3.29560250e-01 1.38032937e+00
2.91294992e-01 1.68607235e-01 -1.41562498e+00 1.47181344e+00
-2.18430638e-01 -8.95546973e-01 -2.05660447e-01 1.48498595e-01
4.49442685e-01 2.78017312e-01 1.76351257e-02 2.66220272e-01
1.26300946e-01 -6.96893811e-01 2.39168733e-01 1.00774355e-01
6.79311693e-01 -1.02791631e+00 9.96090472e-01 1.85176373e-01
-1.22376990e+00 -4.16836023e-01 -1.18281141e-01 -2.89734602e-01
6.42968491e-02 1.13522065e+00 -6.67017519e-01 4.30816919e-01
8.50015104e-01 4.64417905e-01 -2.57291406e-01 1.00407755e+00
-1.28128054e-02 6.48406446e-01 -7.44154871e-01 8.98264349e-02
-2.03170944e-02 -1.48800269e-01 6.01881981e-01 8.42143834e-01
6.75595939e-01 -2.13609099e-01 6.38580859e-01 3.08450311e-01
-1.58456445e-01 -1.50887156e-02 -6.92113519e-01 -1.06959037e-01
7.20466614e-01 1.24782479e+00 -7.50564814e-01 -1.90168962e-01
-5.79442799e-01 1.15274692e+00 -1.58175543e-01 2.63814986e-01
-8.79809856e-01 -4.57218647e-01 1.18230081e+00 1.46279022e-01
1.28683314e-01 -4.01592642e-01 -4.49495882e-01 -1.12623858e+00
4.52390105e-01 -8.26715350e-01 6.30998850e-01 -8.71542096e-02
-1.55000019e+00 3.47738028e-01 -5.62433720e-01 -1.46989429e+00
-3.31250370e-01 -6.65280163e-01 -8.03017020e-01 1.70534745e-01
-8.97267640e-01 -7.90839076e-01 -5.96298754e-01 7.58650720e-01
5.04970551e-01 -4.46989059e-01 7.34652579e-01 4.85521883e-01
-1.27893221e+00 9.76565778e-01 2.77627051e-01 1.39499322e-01
4.95031595e-01 -1.05009449e+00 8.08320761e-01 1.05741465e+00
4.10920233e-02 2.80590087e-01 5.39768338e-01 -5.01671493e-01
-1.77628601e+00 -1.29502821e+00 1.89495429e-01 -2.62522578e-01
8.37931514e-01 -3.61100078e-01 -1.02288079e+00 6.72692895e-01
-3.79980683e-01 3.98406863e-01 7.99703419e-01 2.66928971e-01
-3.23183835e-01 -6.27198637e-01 -1.11898696e+00 7.20606089e-01
1.07982230e+00 -5.13383329e-01 1.37188062e-01 4.05669868e-01
6.01891220e-01 -3.13691914e-01 -8.44297051e-01 4.90382373e-01
2.51560956e-01 -8.54881823e-01 7.40386605e-01 -5.88949740e-01
-2.88799316e-01 -3.77481580e-01 1.19318999e-02 -1.17974246e+00
-2.84376144e-01 -1.04128945e+00 -9.08498645e-01 1.25799048e+00
1.25654086e-01 -1.13222909e+00 9.07780349e-01 8.35571885e-01
-1.85914978e-01 -7.13841796e-01 -1.04328752e+00 -1.27166903e+00
-6.35878980e-01 -9.22697723e-01 9.98728275e-01 9.86664832e-01
-9.30807143e-02 3.37759137e-01 -2.79716074e-01 5.66927791e-01
5.60001135e-01 2.76791900e-02 9.28113461e-01 -1.51772404e+00
-2.15070337e-01 -3.53314966e-01 -7.68073082e-01 -6.65578604e-01
-1.54195443e-01 -4.72466856e-01 -3.15398097e-01 -9.62407947e-01
-3.45248789e-01 -6.04622066e-01 -8.76053035e-01 6.14714205e-01
7.47585371e-02 1.86399464e-02 -3.54358077e-01 2.23808810e-01
-9.19254839e-01 5.56944132e-01 2.44576111e-01 4.38502058e-03
-5.66888332e-01 3.45231116e-01 -4.25628275e-01 1.07848430e+00
1.03266490e+00 -4.28108841e-01 -4.18005973e-01 -4.22346354e-01
1.53885633e-01 -2.68970311e-01 2.45215580e-01 -1.31567526e+00
3.80936384e-01 -3.60506289e-02 5.16029060e-01 -3.15473169e-01
1.36002712e-02 -1.38027120e+00 9.03915539e-02 6.62130713e-01
4.92403090e-01 6.33878708e-01 3.96830410e-01 9.31977630e-01
3.05666924e-02 3.11473072e-01 4.91006017e-01 4.90208060e-01
-8.64334285e-01 8.88660550e-01 -4.16097909e-01 -1.44186497e-01
1.36828911e+00 -2.10324064e-01 -1.28175681e-02 -3.39505196e-01
-3.52216333e-01 1.87105417e-01 6.45509362e-01 7.44999647e-01
4.60246652e-01 -1.51673639e+00 -8.22061226e-02 6.51783049e-01
2.66008675e-01 1.30759180e-01 1.91759363e-01 8.52348149e-01
-2.45314822e-01 1.41713455e-01 3.01649775e-02 -8.34313869e-01
-9.74365890e-01 8.63225281e-01 8.00682902e-02 -5.34102507e-02
-9.99986112e-01 2.24385828e-01 -1.70972854e-01 -1.15628831e-01
4.28995967e-01 -1.78203151e-01 5.38135052e-01 -2.24965349e-01
5.75049996e-01 7.51346409e-01 5.72509587e-01 5.85087873e-02
-5.64937115e-01 3.24092925e-01 -2.18863517e-01 6.02083206e-01
1.14531434e+00 1.56808756e-02 2.69620679e-02 4.67140108e-01
9.62569535e-01 -1.13335118e-01 -1.22904813e+00 -4.25333470e-01
4.48056340e-01 -2.23524034e-01 2.96478271e-01 -3.76302719e-01
-1.29083622e+00 4.14969385e-01 9.77837086e-01 5.27681231e-01
1.45300436e+00 -2.85365313e-01 1.45930219e+00 6.42327070e-01
4.72049952e-01 -1.29300380e+00 3.28457713e-01 4.46023226e-01
2.67027527e-01 -1.40526664e+00 -1.60264879e-01 -1.66843772e-01
-3.47922146e-01 9.83062983e-01 8.53697717e-01 -4.10030447e-02
7.44943261e-01 4.50778157e-01 -7.13952165e-03 -8.85737538e-02
-6.60112500e-01 1.26623496e-01 4.98622805e-02 6.46794736e-01
-4.42600548e-02 3.76319587e-02 1.81739643e-01 6.66292906e-01
2.23071977e-01 -1.23720489e-01 2.46435136e-01 8.56293499e-01
-1.02656558e-01 -9.17212963e-01 9.56857484e-03 7.65216947e-01
-6.17863894e-01 2.69322366e-01 -8.99262130e-02 5.71611881e-01
-1.94916889e-01 9.70468044e-01 4.84620482e-01 -5.97059727e-01
5.40807307e-01 1.67686105e-01 -2.21771881e-01 -2.33801335e-01
-2.69281954e-01 -2.43896872e-01 -2.77712703e-01 -1.02962160e+00
4.33571756e-01 -4.18659270e-01 -1.13978422e+00 -6.49561465e-01
-2.50644356e-01 -5.34159206e-02 8.20865095e-01 1.02979481e+00
1.12444890e+00 7.30772316e-01 9.80462670e-01 -6.48281336e-01
-5.90367794e-01 -6.35470092e-01 -4.96231079e-01 3.81090999e-01
3.57562184e-01 -5.62887549e-01 -4.08671945e-01 -5.26929915e-01] | [7.397425174713135, 2.7031192779541016] |
4b72f4a2-c4b9-48b5-aad2-d14392f756f6 | ada-nets-face-clustering-via-adaptive-1 | 2202.03800 | null | https://arxiv.org/abs/2202.03800v3 | https://arxiv.org/pdf/2202.03800v3.pdf | Ada-NETS: Face Clustering via Adaptive Neighbour Discovery in the Structure Space | Face clustering has attracted rising research interest recently to take advantage of massive amounts of face images on the web. State-of-the-art performance has been achieved by Graph Convolutional Networks (GCN) due to their powerful representation capacity. However, existing GCN-based methods build face graphs mainly according to kNN relations in the feature space, which may lead to a lot of noise edges connecting two faces of different classes. The face features will be polluted when messages pass along these noise edges, thus degrading the performance of GCNs. In this paper, a novel algorithm named Ada-NETS is proposed to cluster faces by constructing clean graphs for GCNs. In Ada-NETS, each face is transformed to a new structure space, obtaining robust features by considering face features of the neighbour images. Then, an adaptive neighbour discovery strategy is proposed to determine a proper number of edges connecting to each face image. It significantly reduces the noise edges while maintaining the good ones to build a graph with clean yet rich edges for GCNs to cluster faces. Experiments on multiple public clustering datasets show that Ada-NETS significantly outperforms current state-of-the-art methods, proving its superiority and generalization. Code is available at https://github.com/damo-cv/Ada-NETS. | ['Yuqi Zhang', 'Xiuyu Sun', 'Ming Lin', 'Senzhang Wang', 'Fangyi Zhang', 'Yaobin Zhang', 'Yaohua Wang'] | 2022-02-08 | ada-nets-face-clustering-via-adaptive | https://openreview.net/forum?id=QJWVP4CTmW4 | https://openreview.net/pdf?id=QJWVP4CTmW4 | iclr-2022-4 | ['face-clustering'] | ['computer-vision'] | [-2.02693731e-01 -1.12694353e-01 2.89853632e-01 -4.77109313e-01
-9.96064618e-02 -2.62913138e-01 4.89047289e-01 -1.98424220e-01
-8.94976258e-02 1.96399868e-01 -6.63970187e-02 2.59745598e-01
-3.46183479e-01 -1.24568582e+00 -4.98329222e-01 -9.82182860e-01
-2.11929381e-01 4.00549173e-01 -3.37220691e-02 -1.21359952e-01
-4.95643951e-02 7.01243162e-01 -1.84956968e+00 1.95473507e-01
6.85542107e-01 9.31005001e-01 -1.20239176e-01 1.83440968e-01
-4.63728398e-01 3.58773261e-01 -3.27654302e-01 -7.49999046e-01
3.85389149e-01 -4.90687132e-01 -4.94204760e-01 1.79506227e-01
4.34864640e-01 7.91630819e-02 -6.49778605e-01 1.53257453e+00
5.13962328e-01 1.07150130e-01 4.34053957e-01 -1.52714455e+00
-9.62358952e-01 6.95989132e-01 -5.79455972e-01 -1.74908370e-01
6.86169416e-02 1.64188463e-02 7.77958870e-01 -8.40282202e-01
5.88392079e-01 1.63592553e+00 5.82414269e-01 8.98477554e-01
-9.63395715e-01 -1.03158844e+00 3.55646983e-02 5.29668868e-01
-1.72872066e+00 -5.27184427e-01 9.46877778e-01 -5.32253049e-02
4.86295521e-01 2.78279752e-01 7.32844770e-01 8.80348682e-01
-1.20647259e-01 3.05088878e-01 7.25600719e-01 -8.45660791e-02
2.39379674e-01 -2.29438379e-01 -1.88868362e-02 8.22029412e-01
4.47592527e-01 -1.91132560e-01 -4.15417671e-01 2.30026245e-02
4.22803909e-01 3.69853586e-01 -2.86398679e-01 -3.31539780e-01
-4.13984239e-01 7.32898295e-01 1.02183700e+00 6.26643658e-01
-3.26588720e-01 -2.08680425e-02 3.03159058e-01 8.28939304e-02
3.96024555e-01 -2.20287532e-01 -4.05175537e-02 3.97149146e-01
-6.36585891e-01 -1.29248068e-01 6.69394732e-01 8.57089937e-01
1.01069975e+00 -1.53160185e-01 8.59264508e-02 8.30152929e-01
5.37628531e-01 4.09670889e-01 2.70647436e-01 -6.78444564e-01
2.34238759e-01 1.16837895e+00 -7.82477856e-01 -1.74018610e+00
-3.52921069e-01 -4.88608330e-01 -1.54990256e+00 6.82777241e-02
1.60774618e-01 9.35605466e-02 -1.03012252e+00 1.55352116e+00
6.09307110e-01 5.50339997e-01 -1.23629302e-01 8.63322079e-01
1.19690681e+00 5.93460500e-01 -7.96254650e-02 -1.21028163e-01
1.25885665e+00 -6.78986132e-01 -6.48897707e-01 -6.57207444e-02
3.86279911e-01 -5.60675681e-01 5.47972560e-01 3.46746951e-01
-6.30968213e-01 -6.21833503e-01 -7.22186267e-01 3.59537095e-01
-5.31239629e-01 -4.26191501e-02 6.38976395e-01 7.32311547e-01
-1.41810322e+00 7.55695045e-01 -5.83594024e-01 -4.59163398e-01
1.05613089e+00 6.78227901e-01 -7.66492307e-01 -6.20917976e-01
-9.18575525e-01 1.71909809e-01 4.53233331e-01 7.34497547e-01
-7.85300791e-01 -2.69606650e-01 -8.84315789e-01 2.13000998e-01
4.74000394e-01 -3.02549422e-01 5.29993474e-01 -1.05866027e+00
-1.09504652e+00 5.81298649e-01 -2.84869540e-02 2.04445291e-02
3.04048955e-01 2.98366457e-01 -8.13941538e-01 3.14333290e-01
7.64356107e-02 6.30450428e-01 1.00396645e+00 -1.39438200e+00
-4.31467861e-01 -8.09080184e-01 -3.42139781e-01 -1.26384586e-01
-5.89874625e-01 -1.29282683e-01 -8.90954971e-01 -2.86500305e-01
4.57121253e-01 -8.78259003e-01 -7.18715340e-02 -2.27340668e-01
-5.48134267e-01 -5.51386833e-01 1.03866029e+00 -3.96178246e-01
1.12851810e+00 -2.24497175e+00 1.88024540e-03 7.50490308e-01
6.51670158e-01 5.84520757e-01 -4.94737744e-01 3.07729900e-01
-2.01467991e-01 3.26822817e-01 -2.91194737e-01 -3.17957342e-01
-1.47731915e-01 1.69420838e-01 3.53958905e-01 7.11840987e-01
1.64680675e-01 8.07006657e-01 -7.79913008e-01 -4.62027341e-01
2.66414374e-01 8.33104610e-01 -4.77738887e-01 7.82854930e-02
5.51933460e-02 2.92839468e-01 -3.49258989e-01 6.72774076e-01
1.33898091e+00 -1.86581030e-01 5.50250590e-01 -2.11672291e-01
2.99615592e-01 -2.80455261e-01 -1.37938893e+00 1.29246807e+00
-6.54383283e-03 2.87621081e-01 3.23693246e-01 -1.06577754e+00
1.02706397e+00 7.39946142e-02 3.85936946e-01 -6.41352475e-01
4.14253384e-01 5.63245453e-02 1.83420897e-01 -4.34612840e-01
-1.83750778e-01 1.94118768e-01 4.02127475e-01 8.58651921e-02
2.11889371e-01 4.66741592e-01 4.54780579e-01 4.94081229e-01
1.03972018e+00 -4.37671006e-01 -3.46431643e-01 -2.85482049e-01
9.12875593e-01 -6.88889384e-01 7.02000678e-01 4.01360065e-01
-1.25063792e-01 5.03682494e-01 4.56282079e-01 -5.86332500e-01
-5.91469884e-01 -7.36221492e-01 5.47980703e-02 7.64571667e-01
1.50016040e-01 -7.66800165e-01 -1.19883513e+00 -9.05602515e-01
2.93798489e-03 2.89132670e-02 -9.20990467e-01 -2.85821170e-01
-5.22265613e-01 -7.32924700e-01 4.19629753e-01 1.15737937e-01
7.88579345e-01 -1.42229855e+00 3.30173969e-01 1.04238890e-01
-4.64838482e-02 -7.68823326e-01 -3.65691036e-01 -2.95408994e-01
-4.68026608e-01 -1.54191160e+00 -3.27483475e-01 -1.06426024e+00
1.10743690e+00 5.68464458e-01 9.18080688e-01 8.74309659e-01
-4.05676812e-01 1.35476559e-01 -4.79416072e-01 2.29188260e-02
-2.56309479e-01 -2.70874109e-02 1.76460087e-01 6.33645713e-01
7.28030741e-01 -6.96451724e-01 -5.49878359e-01 1.75768003e-01
-1.10945296e+00 -3.94949287e-01 5.09824693e-01 6.39230669e-01
4.33193445e-01 6.88854396e-01 4.96025622e-01 -9.17964518e-01
5.35170436e-01 -5.40434182e-01 -5.82080483e-01 1.41892642e-01
-4.86615062e-01 -1.77307308e-01 9.17077780e-01 -1.15124658e-01
-8.32348108e-01 1.62199855e-01 -2.06672236e-01 -5.76725245e-01
-3.65639508e-01 2.45466515e-01 -7.51984239e-01 -3.04100007e-01
3.43290031e-01 1.82106420e-01 2.10151136e-01 -3.87641490e-01
4.62368518e-01 5.86359382e-01 3.38533521e-01 -1.84400752e-01
1.02878702e+00 5.83633661e-01 1.72921821e-01 -8.66320252e-01
-4.92820859e-01 -3.89366627e-01 -5.40719330e-01 -4.43855971e-01
8.44847143e-01 -6.67230725e-01 -1.03830540e+00 7.12574601e-01
-9.19930756e-01 1.56335324e-01 2.04748333e-01 1.77285939e-01
1.79268137e-01 5.06745577e-01 -6.53290808e-01 -6.36118174e-01
-3.36000800e-01 -1.12899971e+00 6.51861548e-01 5.84607899e-01
3.95165503e-01 -7.07350433e-01 -2.74974763e-01 2.16185302e-01
2.03629136e-01 -8.32435116e-03 6.42214596e-01 -6.58293068e-01
-6.29521847e-01 -1.72497198e-01 -4.25472707e-01 5.29852271e-01
3.65767628e-01 3.24439108e-01 -1.02415740e+00 -5.23698986e-01
-5.09707741e-02 2.48821452e-02 9.86769438e-01 3.86521727e-01
1.37725317e+00 -5.42473733e-01 -5.52377105e-01 7.75016367e-01
1.47427046e+00 1.04269437e-01 8.56445670e-01 5.94909787e-02
1.07923365e+00 6.80392504e-01 -1.23053854e-02 1.98403582e-01
2.15796471e-01 1.79154962e-01 7.35715568e-01 -9.34631005e-02
-1.67450994e-01 -1.41919136e-01 5.14021553e-02 8.80093932e-01
1.00313255e-03 -4.13439423e-01 -7.74389744e-01 4.51570630e-01
-1.84120643e+00 -8.80449176e-01 -3.52019161e-01 1.87187767e+00
3.14285278e-01 -8.69040340e-02 -1.27394395e-02 2.64131635e-01
1.31353593e+00 9.09251627e-03 -3.60151470e-01 7.57412985e-02
-1.62122339e-01 2.48904333e-01 1.44240230e-01 1.69523448e-01
-1.02671027e+00 9.03659761e-01 4.51459885e+00 1.06177521e+00
-9.24861193e-01 -6.89160004e-02 8.04419219e-01 1.26662597e-01
-1.81118414e-01 -1.48636326e-01 -7.86508918e-01 7.04114497e-01
5.93517184e-01 5.84630631e-02 7.35665441e-01 8.26354742e-01
-9.22264308e-02 3.13437968e-01 -7.09815621e-01 1.32967508e+00
3.37114722e-01 -1.17826986e+00 4.88920838e-01 2.33609006e-01
7.51160562e-01 -1.64720982e-01 -3.28280218e-02 5.55506237e-02
3.17215592e-01 -1.17198133e+00 1.69670567e-01 3.14184934e-01
6.03240311e-01 -1.21658611e+00 8.61591637e-01 -3.12472377e-02
-1.60976100e+00 -2.25016639e-01 -7.50930548e-01 2.20630631e-01
-2.68596619e-01 9.15992439e-01 -5.24094284e-01 8.38360965e-01
1.12811720e+00 7.21769512e-01 -8.16793859e-01 1.06840050e+00
-2.32663661e-01 5.12341738e-01 -3.78493816e-01 1.19944662e-01
2.44292632e-01 -6.51732206e-01 3.08832705e-01 9.37572002e-01
4.05157954e-01 2.47688472e-01 1.13422684e-02 7.74941027e-01
-6.72645271e-01 3.25438917e-01 -6.74896359e-01 1.32369418e-02
6.30942285e-01 1.86305082e+00 -1.27937281e+00 -2.58633047e-01
-3.95481855e-01 8.30934405e-01 5.42392135e-01 1.89619169e-01
-6.11952305e-01 -4.69146997e-01 6.96806550e-01 3.45775299e-02
3.98990124e-01 7.57108778e-02 2.50071973e-01 -8.27132583e-01
1.47151738e-01 -9.06243086e-01 5.36675334e-01 -3.30526054e-01
-1.68482947e+00 8.98813009e-01 -5.26594937e-01 -9.85340595e-01
4.21147048e-01 -5.40493190e-01 -6.76293135e-01 5.88285744e-01
-1.21463573e+00 -1.09523928e+00 -6.36704326e-01 8.69037211e-01
1.80319309e-01 -3.26037407e-01 5.17411768e-01 5.91310740e-01
-8.84960294e-01 5.44027388e-01 1.79515228e-01 5.82137764e-01
5.47483027e-01 -8.23395252e-01 4.45358425e-01 9.86405790e-01
2.61059403e-01 6.74613416e-01 7.34119788e-02 -8.60858381e-01
-1.43722773e+00 -1.40642190e+00 6.03625178e-01 -2.16956716e-02
2.75481015e-01 -6.25028491e-01 -1.08196473e+00 3.72573853e-01
4.96253110e-02 3.78535450e-01 6.12396836e-01 -1.73802804e-02
-5.50918221e-01 -4.85066772e-01 -1.31847632e+00 5.81375360e-01
1.32508838e+00 -3.34167749e-01 -7.00991303e-02 3.18176568e-01
5.21594822e-01 1.31517515e-01 -5.79694986e-01 2.97217041e-01
3.63940150e-01 -1.20275819e+00 7.33841538e-01 -3.88049364e-01
1.49041012e-01 -5.61715186e-01 2.11112395e-01 -1.40556729e+00
-6.80191994e-01 -4.62210387e-01 1.88934445e-01 1.66295063e+00
-1.03155710e-03 -7.53419936e-01 8.52729619e-01 3.13803881e-01
5.40210567e-02 -5.07370949e-01 -9.28736746e-01 -7.37444699e-01
-2.22064257e-01 -1.53653428e-01 1.11386120e+00 1.10178399e+00
-3.08928132e-01 3.25949639e-01 -5.87834753e-02 2.35609800e-01
1.00416827e+00 -1.76330298e-01 6.83231711e-01 -1.65739048e+00
3.92509460e-01 -6.06964588e-01 -7.78082252e-01 -3.64765137e-01
4.68605667e-01 -1.25133276e+00 -8.69067535e-02 -1.55025887e+00
2.88601309e-01 -4.30662423e-01 -1.59905836e-01 6.07595086e-01
-1.86395228e-01 7.49717474e-01 1.25674188e-01 -7.52513409e-02
-7.64182150e-01 6.47189796e-01 1.14609206e+00 -2.54178375e-01
-1.26005888e-01 -2.66029626e-01 -7.85149157e-01 6.25688493e-01
9.68332052e-01 -5.68744659e-01 -3.53853464e-01 -3.23453933e-01
3.00116893e-02 -5.25475323e-01 3.52973014e-01 -1.28494310e+00
4.90503907e-01 2.11619452e-01 8.19375098e-01 -4.04190540e-01
9.72747952e-02 -1.08466005e+00 5.67260027e-01 3.63335520e-01
1.18077189e-01 -8.78745541e-02 9.08274204e-02 6.76796854e-01
-1.97562620e-01 -1.85311660e-02 9.17757690e-01 -3.11023802e-01
-5.36715627e-01 8.62628043e-01 -1.05437070e-01 -1.66550234e-01
9.95677829e-01 -2.15376794e-01 -2.86957204e-01 -2.51211673e-01
-6.90290034e-01 2.34584644e-01 3.71998549e-01 5.44592917e-01
8.13451648e-01 -1.56442738e+00 -7.93170393e-01 6.53436363e-01
-9.45606455e-02 1.93441391e-01 7.10428894e-01 3.14484507e-01
-5.60824335e-01 1.02630340e-01 -2.19564766e-01 -5.31268775e-01
-1.52172720e+00 7.99949467e-01 3.31177711e-01 2.52254546e-01
-6.88447237e-01 1.00519073e+00 2.26112366e-01 -5.39803863e-01
2.10292637e-01 2.82410353e-01 -4.22433108e-01 2.06395239e-01
7.19840586e-01 4.83868748e-01 4.19573456e-01 -1.03553367e+00
-5.98169327e-01 6.21645212e-01 -1.26928896e-01 5.29615343e-01
1.50540113e+00 -9.68130976e-02 -8.03769708e-01 -3.98605376e-01
1.39150620e+00 -1.42479166e-01 -8.30622554e-01 -7.52066150e-02
-1.82291314e-01 -6.16089225e-01 -7.72400945e-02 -3.28829736e-01
-1.90093625e+00 7.52480209e-01 8.05595636e-01 4.57100153e-01
1.26486039e+00 4.55236249e-02 6.39787912e-01 2.64988244e-01
2.00547904e-01 -8.96990597e-01 -3.34298834e-02 2.25489229e-01
6.31759882e-01 -1.05471265e+00 -2.46279418e-01 -6.97302103e-01
-1.44947261e-01 1.13702822e+00 9.06828463e-01 -1.31522492e-01
9.76234317e-01 -9.26187914e-03 3.50989252e-02 -6.60666883e-01
-3.02179366e-01 -3.11709583e-01 1.56665713e-01 7.91152358e-01
2.86207534e-02 9.02511254e-02 -4.27332856e-02 4.66160476e-01
-1.70009434e-01 -2.53546447e-01 2.86293417e-01 4.60032701e-01
-3.07574272e-01 -1.15459883e+00 -3.83602977e-01 6.96269512e-01
-4.17779446e-01 -1.36197537e-01 -5.46754539e-01 5.68619847e-01
4.36853409e-01 1.33050692e+00 1.30167052e-01 -5.86150289e-01
1.17329791e-01 -9.25631747e-02 2.72732824e-01 -4.58873242e-01
-4.98772353e-01 -6.16416074e-02 -2.99462438e-01 -6.79536045e-01
-4.63866115e-01 -4.04233396e-01 -1.25728655e+00 -7.66606569e-01
-3.38590026e-01 3.95259768e-01 4.64738697e-01 6.33966148e-01
5.29540181e-01 4.23679739e-01 8.74625742e-01 -8.34500849e-01
9.54822898e-02 -8.71773720e-01 -8.70238423e-01 6.26304150e-01
3.64116915e-02 -6.17797852e-01 -6.51453912e-01 -1.88566044e-01] | [13.443795204162598, 1.070569634437561] |
459bec5d-dd04-4712-ba8e-9bbc99ce924f | hiding-data-in-colors-secure-and-lossless | 2201.07444 | null | https://arxiv.org/abs/2201.07444v1 | https://arxiv.org/pdf/2201.07444v1.pdf | Hiding Data in Colors: Secure and Lossless Deep Image Steganography via Conditional Invertible Neural Networks | Deep image steganography is a data hiding technology that conceal data in digital images via deep neural networks. However, existing deep image steganography methods only consider the visual similarity of container images to host images, and neglect the statistical security (stealthiness) of container images. Besides, they usually hides data limited to image type and thus relax the constraint of lossless extraction. In this paper, we address the above issues in a unified manner, and propose deep image steganography that can embed data with arbitrary types into images for secure data hiding and lossless data revealing. First, we formulate the data hiding as an image colorization problem, in which the data is binarized and further mapped into the color information for a gray-scale host image. Second, we design a conditional invertible neural network which uses gray-scale image as prior to guide the color generation and perform data hiding in a secure way. Finally, to achieve lossless data revealing, we present a multi-stage training scheme to manage the data loss due to rounding errors between hiding and revealing processes. Extensive experiments demonstrate that the proposed method can perform secure data hiding by generating realism color images and successfully resisting the detection of steganalysis. Moreover, we can achieve 100% revealing accuracy in different scenarios, indicating the practical utility of our steganography in the real-world. | ['Lina Wang', 'Liming Zhai', 'Ting Liu', 'Yanzhen Ren'] | 2022-01-19 | null | null | null | null | ['steganalysis', 'image-steganography'] | ['computer-vision', 'computer-vision'] | [ 5.82447171e-01 -1.60476461e-01 1.51522115e-01 1.15026675e-01
-1.37217566e-01 -4.25261468e-01 9.14468616e-02 -6.38913691e-01
-5.42518139e-01 3.16925943e-01 -1.61270559e-01 -6.95078433e-01
5.34540892e-01 -1.12885296e+00 -6.50153279e-01 -1.17512047e+00
-2.98385292e-01 -3.43920499e-01 1.33213654e-01 -2.45777145e-01
3.23377579e-01 3.47958207e-01 -1.17847705e+00 2.98720300e-01
6.11535966e-01 1.05716431e+00 4.71408386e-03 7.76111186e-01
1.59916133e-01 8.12248647e-01 -6.77401662e-01 -4.33640808e-01
5.94771743e-01 -6.42957211e-01 -4.02944326e-01 3.38536322e-01
-3.67276967e-01 -1.04870570e+00 -8.13498080e-01 1.45693648e+00
3.75066936e-01 -5.26529074e-01 3.56096208e-01 -1.50004637e+00
-1.24059892e+00 2.80371606e-01 -7.23467648e-01 -2.23002017e-01
-8.82521048e-02 3.25321823e-01 2.71840483e-01 -5.51363289e-01
3.34189743e-01 1.08797050e+00 4.00544077e-01 8.42232764e-01
-8.68069232e-01 -1.21991730e+00 -2.81202108e-01 3.27947319e-01
-1.50747561e+00 -3.91193509e-01 8.44557464e-01 -7.13064447e-02
4.91115451e-01 4.72720712e-01 7.41966248e-01 3.29513580e-01
4.88084882e-01 7.04993427e-01 1.30295837e+00 -5.02845764e-01
-1.86565056e-01 3.17303091e-01 -5.66074550e-01 8.34267497e-01
6.31991804e-01 5.95396757e-01 4.13462259e-02 1.65462226e-01
8.57521057e-01 5.17169416e-01 -6.21310055e-01 -2.08015576e-01
-1.14427471e+00 9.08510447e-01 4.76608872e-01 2.16709048e-01
3.16228457e-02 4.01870340e-01 2.49782071e-01 7.04930961e-01
1.69581458e-01 -2.08491400e-01 7.03312689e-03 6.09323680e-01
-6.28464401e-01 -2.62241125e-01 8.88697624e-01 1.05445528e+00
7.80753374e-01 3.72634768e-01 3.05620044e-01 8.71605352e-02
6.86158001e-01 9.05382752e-01 5.37918627e-01 -6.31748915e-01
6.48208737e-01 2.90880799e-01 -1.73336327e-01 -1.56203592e+00
1.12223074e-01 2.15057448e-01 -1.34316933e+00 7.46614277e-01
5.23840636e-02 -1.01934947e-01 -1.01488709e+00 1.34172404e+00
5.56138530e-02 -1.82446558e-02 6.49823785e-01 8.30965221e-01
5.76610327e-01 1.13353860e+00 -2.72085637e-01 -2.68163770e-01
1.43799651e+00 -6.81086779e-01 -9.09990609e-01 -5.31318523e-02
5.13029337e-01 -6.97746038e-01 5.93217969e-01 3.66562098e-01
-1.02438498e+00 -3.32787246e-01 -1.53268278e+00 -8.78738686e-02
-5.62703192e-01 -2.32453644e-01 4.61742610e-01 1.15945196e+00
-1.10200000e+00 1.14610806e-01 -5.13726294e-01 3.93065184e-01
3.31604511e-01 7.71219194e-01 -5.08563459e-01 -1.93184003e-01
-1.53204036e+00 4.06748116e-01 8.03381562e-01 3.93304080e-01
-8.38659346e-01 -6.93574697e-02 -9.98574018e-01 1.61040902e-01
1.73063278e-01 -1.80001557e-01 6.18330359e-01 -1.01323724e+00
-1.31699133e+00 9.13048744e-01 1.99258596e-01 -3.80501449e-01
4.94053572e-01 6.49074018e-01 -7.31359005e-01 2.99687415e-01
-3.83607060e-01 6.34304643e-01 1.21005642e+00 -1.53467262e+00
-8.10814679e-01 -8.64618868e-02 -8.54422599e-02 -8.51325989e-02
-5.47324359e-01 -9.22363698e-02 -5.04192889e-01 -4.94864285e-01
2.69756705e-01 -9.89293575e-01 -1.17143691e-01 2.40085289e-01
-5.31619966e-01 6.29540265e-01 1.39705241e+00 -9.37013924e-01
1.12949312e+00 -2.50156021e+00 -1.84022307e-01 4.56660241e-01
5.47202408e-01 4.59789336e-01 -1.24057010e-01 3.37736249e-01
-5.39080054e-02 3.90486717e-01 -5.07683575e-01 -1.15532175e-01
-2.56819976e-03 1.47279412e-01 -3.12551796e-01 8.60577524e-01
-2.03245685e-01 9.19351578e-01 -4.06978250e-01 -6.45669758e-01
1.43041849e-01 8.18332076e-01 -4.88303542e-01 9.73980576e-02
3.94794255e-01 2.44404271e-01 -2.76435733e-01 4.75929886e-01
1.41366339e+00 -1.12426601e-01 2.76911885e-01 -2.45930869e-02
-9.24964175e-02 -3.18734974e-01 -1.03249729e+00 7.37235188e-01
-2.62208670e-01 7.25600004e-01 1.24624968e-01 -6.85011148e-01
9.03022230e-01 3.74566942e-01 2.06298754e-01 -8.18370998e-01
4.90301967e-01 3.20694953e-01 2.98745674e-03 -5.18366694e-01
6.41115427e-01 -2.74738044e-01 -1.33498862e-01 7.33371139e-01
-7.35085249e-01 6.12369888e-02 -3.48080009e-01 -7.83290267e-02
6.08586550e-01 -4.07594413e-01 -9.96386707e-02 1.00041740e-01
7.53115356e-01 -3.13651800e-01 4.62987840e-01 2.94267923e-01
-9.63960290e-02 5.36761463e-01 5.03821075e-01 -4.02840257e-01
-1.34967887e+00 -3.04626465e-01 2.84735352e-01 3.92183602e-01
7.24915743e-01 2.65740216e-01 -8.61587584e-01 -3.60960782e-01
-1.58684388e-01 2.19054535e-01 -4.99941975e-01 -4.88967389e-01
-7.70797670e-01 -6.97967768e-01 7.67247379e-01 2.06187516e-02
1.28594685e+00 -1.24872625e+00 -5.49088061e-01 -4.26951125e-02
-1.22996755e-01 -1.04756975e+00 -6.61460042e-01 -1.49043694e-01
-6.30737722e-01 -1.22052884e+00 -8.63452673e-01 -1.46469617e+00
1.16971219e+00 7.54109442e-01 2.48672262e-01 8.87234867e-01
-1.04087971e-01 -1.49647936e-01 -3.34260046e-01 -1.35165676e-01
-7.81813800e-01 -3.10614765e-01 -3.10883582e-01 2.00088203e-01
2.84990728e-01 -2.50000328e-01 -9.56147671e-01 2.09444433e-01
-1.54669666e+00 3.65335166e-01 8.02144885e-01 9.23160493e-01
2.61053979e-01 9.41814601e-01 -3.20317149e-01 -8.54331851e-01
3.93056154e-01 -2.43181020e-01 -8.45988214e-01 1.54795289e-01
-8.72933030e-01 -1.29566178e-01 7.01502562e-01 -3.83131772e-01
-6.18206561e-01 -3.12218130e-01 -1.24128491e-01 -1.54172480e-01
2.52918512e-01 2.65349865e-01 -4.46458459e-01 -8.58021915e-01
-2.34766260e-01 1.02907479e+00 5.84070504e-01 -8.31858963e-02
8.68453160e-02 8.47704828e-01 4.39181954e-01 2.17384636e-01
1.39009249e+00 8.22315574e-01 9.97365490e-02 -4.64907169e-01
3.61535341e-01 2.65229493e-01 -1.92437962e-01 1.36889502e-01
8.46267223e-01 -7.75014877e-01 -1.42705894e+00 1.19400644e+00
-1.13762867e+00 -2.33366311e-01 3.36873829e-01 2.24056035e-01
-2.23592073e-01 8.47669184e-01 -9.28661525e-01 -7.47649968e-01
-3.80753547e-01 -1.35204661e+00 6.10534012e-01 2.08432004e-02
7.72022843e-01 -1.06480217e+00 -5.36952794e-01 -7.72307962e-02
5.19560575e-01 5.28722942e-01 9.26065922e-01 -7.75791407e-02
-1.05725741e+00 -5.89365780e-01 -7.14617252e-01 4.50609803e-01
2.66738027e-01 -2.44905785e-01 -5.91041565e-01 -7.75458217e-01
3.26210022e-01 1.06283493e-01 9.59479511e-01 -5.49751893e-03
1.31341791e+00 -9.50111985e-01 -1.15248308e-01 1.18444455e+00
1.82026052e+00 7.46886849e-01 1.41860998e+00 7.53031433e-01
8.21350098e-01 6.58563733e-01 1.29920527e-01 3.60780209e-01
4.60544825e-01 1.25072598e-01 6.88393116e-01 -6.48173988e-01
6.75476789e-02 -2.25771368e-01 4.08534139e-01 9.11943376e-01
1.38352498e-01 -6.38908684e-01 -3.76840502e-01 2.19903693e-01
-1.21815288e+00 -9.25492167e-01 -1.32778779e-01 1.96917677e+00
7.95131922e-01 8.15871879e-02 -3.48281443e-01 4.73349541e-01
1.04720438e+00 3.66952777e-01 -3.72368604e-01 -3.05349559e-01
-2.11805388e-01 -2.23196298e-01 1.24870455e+00 5.17021000e-01
-9.85197186e-01 8.47898185e-01 5.90101480e+00 7.86576092e-01
-1.48528826e+00 6.05992004e-02 7.67340541e-01 5.00617862e-01
-5.75655103e-01 6.40582368e-02 -5.19651473e-01 8.91470492e-01
4.91595119e-01 2.86934488e-02 6.58358037e-01 3.31991941e-01
-3.96526791e-02 1.30381957e-01 -4.73623097e-01 9.51717198e-01
1.73995391e-01 -1.23557091e+00 1.95141748e-01 6.17275059e-01
5.83523214e-01 -8.30527365e-01 7.42935777e-01 -1.75616160e-01
9.46035460e-02 -9.39910471e-01 6.69141114e-01 1.61096454e-01
1.36465752e+00 -9.30962801e-01 7.87282228e-01 9.82031971e-02
-1.05159616e+00 -1.40728623e-01 -5.99936068e-01 1.73248097e-01
4.68353033e-02 1.73958167e-02 -4.64751631e-01 3.63375217e-01
5.86114705e-01 4.10921812e-01 -2.14900523e-01 6.70411706e-01
-2.51810968e-01 2.85284787e-01 1.40392808e-02 -6.26611039e-02
4.47356284e-01 -1.31635070e-01 2.40367323e-01 1.00510073e+00
6.70282662e-01 3.96387726e-01 -1.39146775e-01 6.76563382e-01
-1.80058971e-01 -2.63027847e-01 -6.19171560e-01 -1.07715137e-01
4.23550665e-01 7.39670217e-01 -7.55272686e-01 -3.12334239e-01
-3.51480454e-01 1.28393495e+00 -6.28502548e-01 5.51901579e-01
-6.93315327e-01 -1.01680350e+00 3.00813168e-01 -8.48534480e-02
6.68085337e-01 -2.96724111e-01 -2.94961721e-01 -1.08676624e+00
-1.14053100e-01 -9.81337130e-01 1.24891885e-01 -6.07574582e-01
-6.28248096e-01 5.80314398e-01 -3.21829051e-01 -1.63019204e+00
7.68311247e-02 -6.10091507e-01 -5.35952032e-01 7.35653877e-01
-2.23645759e+00 -1.19292808e+00 -2.72783041e-01 9.15127575e-01
-1.32147431e-01 -3.93662214e-01 6.63624823e-01 3.43513489e-01
-4.08581644e-01 8.29119623e-01 4.70457315e-01 6.03324413e-01
3.55900764e-01 -7.52147853e-01 6.38764560e-01 1.17546773e+00
-5.38731337e-01 5.38030922e-01 5.99095643e-01 -5.91829956e-01
-1.67813075e+00 -9.98080850e-01 6.95088506e-01 4.56415981e-01
2.97384858e-01 -4.30995554e-01 -9.14760768e-01 5.96631646e-01
3.88950527e-01 -1.65457904e-01 5.47194719e-01 -1.20695162e+00
-3.95731628e-01 -7.36667737e-02 -1.58702397e+00 6.11849010e-01
5.00922382e-01 -5.46215475e-01 5.61386123e-02 -4.88645434e-02
1.01923859e+00 -5.23956120e-01 -4.36555088e-01 -1.11300088e-02
6.61057770e-01 -1.01384807e+00 9.55648720e-01 -4.46388386e-02
5.82006872e-01 -4.74290341e-01 -1.20426968e-01 -7.14890003e-01
-2.36867338e-01 -8.65746200e-01 1.40589327e-01 9.30459857e-01
4.50361967e-02 -1.03011847e+00 9.33484972e-01 4.57833409e-01
2.60650665e-01 -9.48753208e-02 -7.19752550e-01 -6.57992363e-01
-1.65252313e-02 -8.33497103e-03 1.21963608e+00 9.35252488e-01
-2.37154901e-01 -6.47560596e-01 -1.18427646e+00 5.40177226e-01
9.58658576e-01 -2.77009420e-02 6.55900419e-01 -6.58347189e-01
2.94088461e-02 -2.06630334e-01 -6.28038108e-01 -1.05807292e+00
1.63725298e-02 -6.81185067e-01 7.80596063e-02 -1.13033032e+00
1.17795676e-01 -6.31277263e-01 -3.00336719e-01 4.96110052e-01
5.31390794e-02 9.02973771e-01 2.62321621e-01 5.25188625e-01
-1.73145324e-01 2.96015710e-01 1.59190500e+00 -3.46962005e-01
4.04079705e-02 -2.63017446e-01 -8.65451872e-01 3.15688908e-01
7.42946446e-01 -6.18377566e-01 -3.16777170e-01 -6.61245048e-01
1.45844787e-01 2.43586376e-01 5.78737080e-01 -8.40484142e-01
2.65869141e-01 -1.20789029e-01 3.49185795e-01 -3.50680172e-01
1.33058250e-01 -1.47076213e+00 2.34304756e-01 1.31769395e+00
-6.12863861e-02 -1.51964366e-01 6.66313199e-03 6.02713645e-01
-2.18791038e-01 -1.39100507e-01 7.79436052e-01 -1.28456593e-01
-1.04005945e+00 5.37334204e-01 -5.87284029e-01 -5.34119248e-01
1.15350449e+00 -7.45533347e-01 -3.17385614e-01 -4.75791901e-01
-3.78210157e-01 -3.12563553e-02 7.80323207e-01 4.67245542e-02
1.23292351e+00 -1.53166389e+00 -6.01332545e-01 9.67814803e-01
-9.58291590e-02 -3.27369869e-01 3.62362474e-01 5.36992431e-01
-1.29438496e+00 2.45120987e-01 -3.54196638e-01 -1.55949518e-01
-1.39112914e+00 9.40520227e-01 1.97524354e-01 -3.80372219e-02
-7.80483305e-01 6.01922750e-01 2.98479229e-01 1.63287133e-01
1.88393340e-01 1.49696181e-03 -1.32898182e-01 -4.17183727e-01
8.30852687e-01 1.04996979e-01 -4.32101101e-01 -6.91038072e-01
-3.65975760e-02 6.58732057e-01 -1.11399747e-01 -8.72444436e-02
1.12446749e+00 -7.33669758e-01 -6.26181006e-01 -4.21784729e-01
1.81753886e+00 -8.84224474e-02 -1.21687281e+00 -1.91488534e-01
-5.37069738e-01 -8.93931746e-01 1.17436811e-01 -2.20971003e-01
-1.65940917e+00 9.67639387e-01 7.83486068e-01 6.41727626e-01
1.43845320e+00 -7.04543233e-01 1.61879599e+00 2.15460151e-01
3.61449093e-01 -6.59063041e-01 -1.03359453e-01 9.79455784e-02
4.17846501e-01 -1.23167503e+00 4.13603848e-03 -4.02422369e-01
-5.66002429e-01 1.31517899e+00 2.03464046e-01 -1.64036885e-01
6.30914211e-01 3.15008014e-01 1.19222797e-01 -6.90298900e-02
-2.22891346e-01 9.64215323e-02 -1.40818179e-01 8.16129565e-01
-3.38237256e-01 -5.26148267e-02 -1.48462653e-01 5.24662770e-02
-8.34599286e-02 1.21848909e-02 8.93820226e-01 1.04820466e+00
-5.62721670e-01 -1.15632653e+00 -7.25537658e-01 -2.55482048e-01
-7.31382370e-01 -3.23278308e-01 5.41485325e-02 8.01712275e-01
1.77262455e-01 9.60108101e-01 8.81039724e-02 -8.04231346e-01
-2.02440381e-01 -6.08859003e-01 8.65995660e-02 -3.38731185e-02
-6.19023815e-02 1.67179525e-01 -5.39909542e-01 -9.47159678e-02
-3.56033146e-01 -1.16559640e-02 -1.25390184e+00 -1.21610665e+00
-2.98932910e-01 1.59295738e-01 7.17537522e-01 6.87907159e-01
6.40384331e-02 3.77706796e-01 1.25513208e+00 -6.57813132e-01
-1.79611012e-01 -1.70364097e-01 -9.54886794e-01 1.48367077e-01
9.96447504e-01 6.80514351e-02 -6.31468534e-01 3.38360250e-01] | [4.322116851806641, 8.04431438446045] |
029a3b51-807e-4457-9106-21151235e33b | multi-dimensional-edge-based-audio-event | 2210.15366 | null | https://arxiv.org/abs/2210.15366v2 | https://arxiv.org/pdf/2210.15366v2.pdf | Multi-dimensional Edge-based Audio Event Relational Graph Representation Learning for Acoustic Scene Classification | Most existing deep learning-based acoustic scene classification (ASC) approaches directly utilize representations extracted from spectrograms to identify target scenes. However, these approaches pay little attention to the audio events occurring in the scene despite they provide crucial semantic information. This paper conducts the first study that investigates whether real-life acoustic scenes can be reliably recognized based only on the features that describe a limited number of audio events. To model the task-specific relationships between coarse-grained acoustic scenes and fine-grained audio events, we propose an event relational graph representation learning (ERGL) framework for ASC. Specifically, ERGL learns a graph representation of an acoustic scene from the input audio, where the embedding of each event is treated as a node, while the relationship cues derived from each pair of event embeddings are described by a learned multidimensional edge feature. Experiments on a polyphonic acoustic scene dataset show that the proposed ERGL achieves competitive performance on ASC by using only a limited number of embeddings of audio events without any data augmentations. The validity of the proposed ERGL framework proves the feasibility of recognizing diverse acoustic scenes based on the event relational graph. Our code is available on our homepage (https://github.com/Yuanbo2020/ERGL). | ['Dick Botteldooren', 'Wenwu Wang', 'Yuxin Song', 'Chuang Yu', 'Siyang Song', 'Yuanbo Hou'] | 2022-10-27 | null | null | null | null | ['scene-classification'] | ['computer-vision'] | [ 2.53396273e-01 -2.35683665e-01 2.68533498e-01 -5.39185405e-01
-7.59496570e-01 -4.43434030e-01 3.58919919e-01 4.58768994e-01
-1.95039421e-01 -5.43655008e-02 3.79515231e-01 8.31726417e-02
-1.80781335e-01 -7.93681324e-01 -6.54341042e-01 -5.98027766e-01
-2.28475899e-01 -1.01496078e-01 2.65547425e-01 8.43912885e-02
5.93805872e-02 2.89483964e-01 -1.72590721e+00 3.47933650e-01
6.35336265e-02 1.31411791e+00 4.01424408e-01 7.67211914e-01
-1.00373030e-01 6.49696350e-01 -6.54105842e-01 6.05878839e-03
-1.88300163e-01 -3.94540578e-01 -3.65886509e-01 -5.81210777e-02
3.98350924e-01 5.68707138e-02 -7.92901039e-01 9.65890944e-01
6.06128275e-01 3.86571795e-01 5.29942632e-01 -1.57491481e+00
-5.73503971e-01 6.09685063e-01 -1.77197859e-01 3.61404240e-01
3.83342594e-01 -1.86546460e-01 1.64652741e+00 -1.13483071e+00
1.70113385e-01 1.10150504e+00 5.23153722e-01 1.26134828e-01
-8.21133852e-01 -7.41933048e-01 3.33585441e-01 6.92484260e-01
-1.66338062e+00 -4.98620659e-01 1.40340650e+00 -4.34473902e-01
9.07221138e-01 2.26394981e-01 6.65963948e-01 8.91387761e-01
4.88670953e-02 5.73357046e-01 6.45319521e-01 -3.74339968e-01
3.02448601e-01 -2.13111073e-01 4.39165354e-01 6.23972714e-01
-7.43590072e-02 -1.35133237e-01 -1.05497658e+00 -2.04723135e-01
4.48480219e-01 -2.68900003e-02 -2.63830394e-01 -2.75853902e-01
-8.39441478e-01 7.79151797e-01 5.01084924e-01 3.30480814e-01
-3.30663025e-01 3.97571445e-01 6.70431197e-01 1.68379456e-01
4.14593160e-01 -4.93942061e-03 -9.43014324e-02 -1.26914084e-01
-3.31672341e-01 -1.53216526e-01 5.47518790e-01 7.61971772e-01
8.98779631e-01 5.05940259e-01 1.94369778e-01 1.02036977e+00
4.48481649e-01 3.79060805e-01 5.24491072e-01 -5.09112298e-01
3.30219716e-01 4.28018779e-01 -4.21809524e-01 -1.50909925e+00
-4.38352257e-01 -3.23687732e-01 -4.77591366e-01 -3.94160599e-01
-5.05537167e-02 3.84636596e-02 -6.73276663e-01 1.87767076e+00
3.61153752e-01 8.41201484e-01 -1.11628547e-01 8.79792571e-01
1.29615963e+00 9.66770589e-01 1.10550642e-01 2.77837459e-02
1.41520607e+00 -5.68121195e-01 -7.51987875e-01 -1.52491972e-01
3.09097588e-01 -6.71109080e-01 1.21684480e+00 1.12861007e-01
-6.37901783e-01 -7.65755236e-01 -9.80175793e-01 3.24713625e-02
-4.68630582e-01 1.13434479e-01 5.06130159e-01 3.73993397e-01
-8.81751478e-01 -4.98750210e-02 -7.21723378e-01 -2.31616288e-01
9.46801901e-02 1.13523997e-01 -4.65732902e-01 -7.29257688e-02
-1.30717993e+00 1.08664483e-01 3.81637841e-01 3.43032181e-01
-1.20006275e+00 -6.02421880e-01 -1.26432359e+00 2.13088393e-01
2.92957664e-01 -1.58709198e-01 1.00162232e+00 -4.28671181e-01
-1.05883265e+00 5.04685223e-01 -2.05748484e-01 -2.83448011e-01
-3.76076907e-01 -2.10095599e-01 -7.37558961e-01 4.15703863e-01
-6.53108880e-02 2.98152924e-01 8.87861490e-01 -1.14459205e+00
-5.09587407e-01 -3.07082951e-01 6.06050529e-02 3.44604880e-01
-6.85116053e-01 9.19687599e-02 -4.62788433e-01 -7.96003342e-01
9.39932317e-02 -8.25219512e-01 6.70223385e-02 -1.07177824e-01
-3.81746233e-01 -4.51675773e-01 1.05013609e+00 -4.52721328e-01
1.31479084e+00 -2.82921314e+00 -3.11873723e-02 4.96399291e-02
1.13822043e-01 -9.62528139e-02 -4.22909170e-01 6.90277278e-01
-9.81208682e-02 -1.60614550e-01 -1.66837201e-01 -2.13548511e-01
8.31197500e-02 2.04429001e-01 -4.50197399e-01 4.25000787e-01
1.07465245e-01 6.61089778e-01 -1.05414653e+00 -3.04504454e-01
3.51322562e-01 7.22539604e-01 -3.58717382e-01 4.21173722e-01
1.33168370e-01 2.23174989e-01 -4.40534770e-01 4.61166620e-01
4.03416425e-01 -3.41094173e-02 1.85272008e-01 -5.67230046e-01
1.39781505e-01 5.13347805e-01 -1.33020663e+00 1.84347785e+00
-5.86612105e-01 8.00156116e-01 -4.06096652e-02 -1.16080129e+00
1.18422163e+00 5.15239239e-01 5.49138069e-01 -5.11701047e-01
-5.01097068e-02 2.20467094e-02 -1.17306627e-01 -4.74727035e-01
2.99770325e-01 -1.33857995e-01 -4.89159733e-01 3.78214419e-01
4.53641146e-01 -2.54332244e-01 -2.02777460e-01 3.10093969e-01
1.19950652e+00 -3.82144809e-01 3.34396929e-01 1.25843331e-01
4.80976760e-01 -3.12725663e-01 6.95010424e-01 6.61025167e-01
-3.26906830e-01 6.03223860e-01 1.37981936e-01 -2.49987826e-01
-6.67401075e-01 -1.29632461e+00 -1.06455557e-01 1.34513903e+00
2.99957812e-01 -7.64589071e-01 -3.35880846e-01 -4.94318962e-01
-9.68686789e-02 5.70593834e-01 -4.18862641e-01 -3.27027857e-01
-5.04361033e-01 -4.18561220e-01 7.61511683e-01 6.18388772e-01
2.16158196e-01 -1.18480515e+00 -3.37702304e-01 1.87002584e-01
-1.38845801e-01 -1.51809084e+00 -5.07836878e-01 2.87801057e-01
-3.64569604e-01 -9.52294111e-01 -1.86339989e-02 -1.09101510e+00
3.61155897e-01 3.82315427e-01 8.84119034e-01 -3.57757241e-01
-5.95696211e-01 8.58177185e-01 -6.53016686e-01 -4.82382029e-01
2.69419485e-04 -3.45439345e-01 2.36966848e-01 6.20066524e-01
4.63559687e-01 -7.61779428e-01 -5.39887667e-01 1.80942491e-01
-7.98944831e-01 -1.92088678e-01 1.41825899e-01 5.54559231e-01
8.22014093e-01 3.31715167e-01 8.23057830e-01 -6.77598476e-01
5.04844666e-01 -6.94554031e-01 -1.76023692e-01 -1.66810080e-02
1.00089207e-01 -4.01431382e-01 6.92159891e-01 -5.42102158e-01
-9.06875014e-01 2.48630688e-01 -7.71485642e-02 -7.18204260e-01
-4.31012839e-01 6.35623693e-01 -2.18807042e-01 1.05893552e-01
4.16517109e-01 3.52831244e-01 -6.50084555e-01 -5.33029139e-01
4.39861834e-01 7.56831646e-01 6.01378560e-01 -4.50636864e-01
6.61758602e-01 4.70998913e-01 -1.11470997e-01 -1.17195582e+00
-8.62451911e-01 -7.67064452e-01 -4.72920835e-01 -5.12915134e-01
8.68444800e-01 -1.14842534e+00 -4.67455894e-01 2.87951976e-01
-9.21519339e-01 -1.15446873e-01 -4.00185019e-01 7.49086618e-01
-3.63477021e-01 2.57211715e-01 -5.79749107e-01 -9.56324220e-01
1.05088837e-01 -7.61258006e-01 1.30559444e+00 -8.74235332e-02
-2.23861650e-01 -9.36748385e-01 1.90339103e-01 2.09724352e-01
-9.31812674e-02 2.74442405e-01 1.04033971e+00 -7.65662789e-01
-3.36213320e-01 -2.74659574e-01 2.54021548e-02 3.83344948e-01
5.11274099e-01 -1.45626023e-01 -1.38180280e+00 -2.18696952e-01
1.59018338e-01 -3.06655169e-01 8.53304744e-01 2.98930019e-01
1.46701789e+00 -6.45463318e-02 3.45079154e-02 4.64979261e-01
1.35608983e+00 4.98102516e-01 1.58505201e-01 -2.44456694e-01
1.22961819e+00 5.90467095e-01 4.86548483e-01 6.46515369e-01
4.43869293e-01 7.21847773e-01 5.04389763e-01 1.60369307e-01
-3.84194553e-01 -6.00857794e-01 5.02373576e-01 1.58157504e+00
2.79954374e-01 -2.22290069e-01 -9.82755303e-01 8.18814218e-01
-1.68186700e+00 -6.67034924e-01 -5.55880775e-04 1.90318179e+00
5.90592027e-01 -6.34311698e-03 -6.03911355e-02 5.31580925e-01
8.57810795e-01 5.63722610e-01 -3.94522995e-01 -3.87910008e-01
-3.52688469e-02 3.15714031e-01 -2.39178047e-01 3.56247008e-01
-1.16385138e+00 8.19616735e-01 5.61109781e+00 8.48433912e-01
-1.21365809e+00 2.29934648e-01 1.40251935e-01 -9.25910249e-02
-3.00806671e-01 -1.63641751e-01 -5.52597344e-01 2.99291372e-01
1.14809656e+00 -1.66267782e-01 1.45870999e-01 7.05001235e-01
1.83090866e-01 4.47284013e-01 -1.11279929e+00 1.15322125e+00
2.01527491e-01 -1.21830368e+00 2.53806621e-01 -3.13475728e-01
2.46117666e-01 -4.95458655e-02 1.32320181e-01 2.77935266e-01
2.76955646e-02 -8.81275237e-01 6.80568099e-01 4.26350117e-01
6.79578125e-01 -6.35609627e-01 4.01196152e-01 1.62631236e-02
-1.97275829e+00 -9.52671096e-02 -3.60145062e-01 -6.66998923e-02
1.57674447e-01 5.16235113e-01 -1.13157165e+00 6.18547022e-01
8.86900246e-01 9.81562555e-01 -4.76955295e-01 8.97954702e-01
-2.45103046e-01 1.09594297e+00 -2.45307922e-01 3.49155106e-02
6.75659403e-02 1.14761308e-01 7.11380184e-01 1.44449985e+00
3.26598942e-01 1.26336768e-01 2.56500304e-01 5.08994997e-01
-1.90682068e-01 3.15072030e-01 -8.49774420e-01 -2.37602234e-01
7.63662040e-01 1.23340309e+00 -5.25515676e-01 9.25789028e-03
-6.23002708e-01 5.58906794e-01 3.06798786e-01 4.24502373e-01
-7.04625666e-01 -5.30817688e-01 7.04264700e-01 -1.60640344e-01
4.82828230e-01 -4.25834775e-01 2.02700272e-02 -8.24176967e-01
9.39005837e-02 -6.41751409e-01 6.62647069e-01 -9.24446285e-01
-1.41624999e+00 7.95225263e-01 -1.08603008e-01 -1.36071229e+00
-1.35753319e-01 -4.26526010e-01 -7.14342654e-01 4.68193769e-01
-1.33090866e+00 -1.13081872e+00 -4.22101706e-01 8.91502380e-01
6.99379027e-01 -2.37695962e-01 9.62700009e-01 3.91112745e-01
-5.07702470e-01 5.81988513e-01 -1.73750401e-01 3.68452340e-01
4.22976613e-01 -1.14789891e+00 2.60187298e-01 6.71336949e-01
9.53819573e-01 2.96442479e-01 4.47318941e-01 -3.86159331e-01
-1.46017814e+00 -1.29156792e+00 7.77520359e-01 -1.27978370e-01
1.06272304e+00 -7.60435283e-01 -1.13614082e+00 5.49269378e-01
-2.50650998e-02 5.14314353e-01 1.15073836e+00 1.30259186e-01
-7.10396051e-01 -4.35437977e-01 -5.61192691e-01 2.56789356e-01
1.09202874e+00 -1.32491124e+00 -6.05265379e-01 1.69392318e-01
1.01069617e+00 -1.39470130e-01 -9.53935266e-01 4.79145020e-01
2.50089467e-01 -4.85870808e-01 1.06179821e+00 -4.09318715e-01
8.13704953e-02 -5.50035179e-01 -6.59581780e-01 -1.22604144e+00
-3.29464376e-01 -1.84354395e-01 -1.54308900e-01 1.32680166e+00
9.91517454e-02 -3.46917301e-01 3.35713357e-01 -1.57078102e-01
-3.98239076e-01 -5.25702477e-01 -1.14084899e+00 -7.86190093e-01
-5.50645053e-01 -1.15226638e+00 5.24971008e-01 1.01017058e+00
-9.55329165e-02 4.88293529e-01 -3.94673496e-01 7.94964135e-01
6.05640113e-01 3.32117349e-01 5.01021802e-01 -1.13219261e+00
-4.54610020e-01 -5.00679687e-02 -1.06987381e+00 -6.93173647e-01
4.47175860e-01 -1.04351068e+00 2.91189641e-01 -1.30485237e+00
-6.18108734e-02 -4.23264980e-01 -9.02266502e-01 4.66707885e-01
-1.02824211e-01 4.84347582e-01 2.48407751e-01 1.06640205e-01
-6.69478416e-01 7.62240887e-01 8.63394439e-01 -2.56168395e-01
4.94685695e-02 -1.58524066e-01 -3.58228028e-01 7.97566712e-01
7.52315521e-01 -5.34696877e-01 -7.36542583e-01 -5.45483589e-01
1.63694575e-01 5.59064746e-02 5.77700257e-01 -1.10175610e+00
5.30128539e-01 -2.16415059e-02 -3.92285958e-02 -4.29751456e-01
7.26642549e-01 -6.73768759e-01 1.74019605e-01 -3.75754386e-02
-4.97151226e-01 -1.83294877e-01 3.37324262e-01 1.04152024e+00
-8.47905159e-01 -8.94785300e-02 3.75642449e-01 2.42446661e-01
-1.17591476e+00 3.94600362e-01 -3.35864127e-01 8.14865157e-02
7.74676502e-01 5.55034801e-02 -3.97112519e-02 -5.81301510e-01
-1.02549779e+00 -3.01301777e-01 -2.25781634e-01 6.52619720e-01
9.31604624e-01 -1.58505666e+00 -6.29213452e-01 2.67478883e-01
4.76906270e-01 -1.65283740e-01 5.18953443e-01 2.22041801e-01
-2.02141881e-01 2.87978470e-01 2.97279768e-02 -8.08330655e-01
-1.37658417e+00 3.03599417e-01 1.46242619e-01 3.35574329e-01
-7.69022048e-01 1.18111038e+00 7.69343495e-01 -3.00833702e-01
4.44310129e-01 -2.35312581e-01 -3.70057404e-01 1.52832478e-01
3.17972839e-01 1.64575309e-01 4.21806164e-02 -1.01939273e+00
-6.86930001e-01 7.38498867e-01 2.40345433e-01 -9.77242514e-02
1.36778581e+00 -1.28301606e-01 8.39052722e-02 1.07263815e+00
1.51680350e+00 8.23145583e-02 -1.14213324e+00 -6.10745430e-01
-1.37779832e-01 -5.04993021e-01 2.22609311e-01 -2.07843110e-01
-1.14403737e+00 1.15759587e+00 6.39639556e-01 4.02707815e-01
1.42717719e+00 2.88870215e-01 6.90344334e-01 2.13500410e-01
3.18012506e-01 -7.71628678e-01 4.94486332e-01 4.91755307e-01
1.00372446e+00 -9.46963787e-01 -2.99411118e-01 -5.73349774e-01
-7.55499244e-01 1.00824893e+00 3.47775191e-01 -3.29970360e-01
1.10601902e+00 2.85476536e-01 9.62063968e-02 -4.53705490e-01
-7.75649726e-01 -2.92217344e-01 5.51583648e-01 6.82589531e-01
2.83137828e-01 3.56424838e-01 3.29134434e-01 7.44067132e-01
-4.11714673e-01 -6.91584706e-01 3.02639514e-01 7.50373840e-01
-4.57021594e-01 -8.39856565e-01 -1.85349077e-01 2.17346936e-01
-1.90314591e-01 -1.46643132e-01 -3.65593255e-01 4.20610845e-01
3.71115766e-02 1.17352378e+00 3.40345502e-01 -7.32983708e-01
4.84186500e-01 8.19360912e-02 1.37110487e-01 -9.88646150e-01
-3.58188093e-01 1.90732196e-01 -2.34836452e-02 -5.29712021e-01
-3.79355401e-01 -6.50745213e-01 -1.45214176e+00 1.89964116e-01
1.52484176e-03 2.11160123e-01 6.08321488e-01 6.92539692e-01
3.87421191e-01 8.43833268e-01 9.16469216e-01 -6.57694399e-01
9.78749990e-03 -7.08170116e-01 -9.17812943e-01 5.19331813e-01
5.34630537e-01 -7.10494697e-01 -5.55112123e-01 2.24607125e-01] | [14.988616943359375, 4.990789413452148] |
f8b92e07-ada1-421c-ae35-e0abf631118e | qigen-generating-efficient-kernels-for | 2307.03738 | null | https://arxiv.org/abs/2307.03738v1 | https://arxiv.org/pdf/2307.03738v1.pdf | QIGen: Generating Efficient Kernels for Quantized Inference on Large Language Models | We present ongoing work on a new automatic code generation approach for supporting quantized generative inference on LLMs such as LLaMA or OPT on off-the-shelf CPUs. Our approach is informed by the target architecture and a performance model, including both hardware characteristics and method-specific accuracy constraints. Results on CPU-based inference for LLaMA models show that our approach can lead to high performance and high accuracy, comparing favorably to the best existing open-source solution. A preliminary implementation is available at https://github.com/IST-DASLab/QIGen. | ['Markus Püschel', 'Dan Alistarh', 'Elias Frantar', 'Tommaso Pegolotti'] | 2023-07-07 | null | null | null | null | ['code-generation'] | ['computer-code'] | [-1.77958012e-01 -4.08407524e-02 -4.14718062e-01 -5.37652910e-01
-1.33911407e+00 -5.91024339e-01 7.62285233e-01 -2.93770999e-01
1.91854998e-01 7.69604921e-01 6.79659918e-02 -9.39799666e-01
4.56902504e-01 -7.23866642e-01 -5.60409307e-01 -4.23373938e-01
-1.94811627e-01 8.46157253e-01 1.02313079e-01 5.25838621e-02
3.16237599e-01 3.17947119e-01 -1.53737557e+00 3.29610556e-01
6.79390669e-01 8.70165288e-01 1.49996495e-02 1.42098379e+00
7.60395154e-02 7.94225872e-01 -5.70691943e-01 -2.13650733e-01
8.81570801e-02 -5.31982362e-01 -8.15244198e-01 -3.66982609e-01
4.96832967e-01 -3.96342754e-01 -1.89158916e-01 9.23610091e-01
7.78789639e-01 -3.42104763e-01 1.51119500e-01 -1.21489620e+00
-2.03016028e-01 8.99032831e-01 -3.22722942e-01 3.37665975e-01
3.24763209e-01 5.21890521e-01 9.33140337e-01 -6.57060623e-01
4.95550066e-01 1.23742211e+00 8.80818665e-01 3.13346028e-01
-1.62672102e+00 -6.28507137e-01 -6.22658074e-01 -2.14910299e-01
-1.80960751e+00 -1.10773087e+00 2.04696685e-01 -1.51954904e-01
1.54612732e+00 4.23477083e-01 5.26710272e-01 9.82469082e-01
7.00960219e-01 5.91668904e-01 1.15755451e+00 -6.19792283e-01
6.23725593e-01 2.19298348e-01 9.66391414e-02 1.00484824e+00
3.68582577e-01 1.28803715e-01 -4.88937110e-01 -9.54969406e-01
4.78263885e-01 -9.07562017e-01 1.82600021e-01 -3.93167764e-01
-9.55921590e-01 9.89011228e-01 -4.57254983e-02 3.40122938e-01
8.73490423e-02 1.08303595e+00 5.41775465e-01 -1.09132759e-01
4.33425784e-01 1.08523473e-01 -4.36139852e-01 -7.11495042e-01
-1.76597965e+00 6.76515877e-01 1.15366316e+00 1.07920301e+00
7.48746455e-01 3.75699461e-01 -2.62619048e-01 3.63083243e-01
6.57801390e-01 6.01246059e-01 2.71722257e-01 -1.35073328e+00
1.70057908e-01 -1.08619273e-01 -2.29303136e-01 -4.12368536e-01
-2.75524676e-01 -4.99742329e-01 -5.82348287e-01 8.37615654e-02
4.70922999e-02 -3.19518656e-01 -6.93457484e-01 1.56576753e+00
1.48729786e-01 4.26214993e-01 4.31763530e-02 5.23379803e-01
6.13562167e-01 7.96796441e-01 -3.45751256e-01 -1.10324770e-01
1.10127735e+00 -8.49391162e-01 -4.61898237e-01 -2.56165594e-01
1.03664446e+00 -9.45968688e-01 8.99991810e-01 5.26124239e-01
-1.35944557e+00 -3.07132483e-01 -1.11172795e+00 -2.24979728e-01
6.96585104e-02 4.13300216e-01 1.21784115e+00 1.31253111e+00
-1.73911548e+00 3.10582012e-01 -1.28518999e+00 -3.47865224e-01
1.74668744e-01 4.44533527e-01 5.86216033e-01 4.12734635e-02
-8.65800083e-01 6.15629017e-01 4.08422709e-01 -2.92397648e-01
-9.93555844e-01 -6.08717978e-01 -6.18874192e-01 -1.79567607e-03
5.31132370e-02 -8.37104857e-01 1.72140121e+00 -4.23242539e-01
-1.69868958e+00 6.71845615e-01 -3.34874779e-01 -8.10669899e-01
3.32318127e-01 4.24486734e-02 -3.25613320e-01 -2.09866464e-01
-1.14173457e-01 6.51314557e-01 3.67281944e-01 -9.64915037e-01
-1.68628603e-01 1.40637711e-01 -5.58842793e-02 -1.89950615e-01
1.18149348e-01 -8.35178420e-03 -4.40366775e-01 -3.26553434e-01
-2.56236702e-01 -1.22266269e+00 -2.85255104e-01 -4.56767529e-01
-3.82412642e-01 2.41206754e-02 4.84509498e-01 -5.87009370e-01
1.40424204e+00 -1.71202886e+00 -8.01783875e-02 3.08589458e-01
-7.10137635e-02 7.75756836e-02 1.11078165e-01 6.28503084e-01
3.44072610e-01 1.05325587e-01 -1.25845656e-01 -4.02098060e-01
3.69215310e-01 1.63635105e-01 -4.46852654e-01 3.76703918e-01
-1.39797255e-01 7.76455045e-01 -9.06748712e-01 -6.87396467e-01
1.93940490e-01 4.92375106e-01 -9.46122348e-01 1.01054862e-01
-7.22292066e-01 1.76632330e-01 -3.04867685e-01 7.90106356e-01
6.38859451e-01 -3.71153563e-01 5.24445891e-01 1.17391519e-01
-2.34064505e-01 8.03875089e-01 -1.13119316e+00 2.03084278e+00
-7.73382246e-01 7.76781261e-01 2.95891374e-01 -3.43024671e-01
6.43875241e-01 1.78069845e-01 -1.39994875e-01 -2.54522979e-01
3.25996757e-01 4.44435060e-01 2.37287179e-01 1.22962072e-01
8.39375854e-01 3.27003986e-01 -2.74359316e-01 6.16975367e-01
2.25582123e-01 -6.76611781e-01 4.81355906e-01 6.15694046e-01
1.26840377e+00 3.69398743e-01 3.24228495e-01 -9.47289467e-01
3.67669791e-01 6.54790401e-02 1.87561005e-01 9.76259112e-01
9.43186805e-02 3.46549153e-01 5.54256380e-01 -1.12445243e-01
-1.26073062e+00 -8.23777676e-01 -6.17039979e-01 9.59809482e-01
-5.53140402e-01 -1.17456341e+00 -1.17037737e+00 -1.42091945e-01
-1.90001503e-01 1.22610533e+00 -1.61536068e-01 1.98936865e-01
-2.17644632e-01 -1.01819813e+00 1.14968491e+00 3.10989290e-01
4.43553388e-01 -6.49243295e-01 -6.11599445e-01 1.81096330e-01
1.05148509e-01 -9.19897258e-01 -1.66428328e-01 1.75709844e-01
-9.23171759e-01 -3.93418312e-01 -4.27038632e-02 -1.94671035e-01
3.03544015e-01 -3.62398803e-01 1.65209973e+00 3.25988710e-01
-5.56765020e-01 1.05414473e-01 8.39085132e-02 -1.14208385e-01
-1.02558553e+00 4.37604427e-01 -1.99041720e-02 -7.69336343e-01
-9.54361781e-02 -6.25872493e-01 -3.79894763e-01 -3.87029856e-01
-6.59619510e-01 2.11886749e-01 6.05477273e-01 6.86746418e-01
4.63297039e-01 -1.06691889e-01 3.44039463e-02 -1.08424747e+00
6.50399983e-01 -5.38795471e-01 -1.14351916e+00 5.87327033e-02
-9.00735319e-01 5.16074002e-01 3.93489122e-01 8.71528983e-02
-9.15886402e-01 7.84067884e-02 -6.83369279e-01 -1.29434913e-01
1.16224475e-02 4.14259255e-01 -2.14239657e-02 -1.84437633e-01
6.26043618e-01 3.08833271e-01 -1.76500529e-01 -1.32496774e-01
5.19587994e-01 7.13665903e-01 4.51147139e-01 -1.08609390e+00
3.69637191e-01 1.49582162e-01 7.86131620e-02 -6.23094499e-01
-3.99787307e-01 -1.35566201e-03 -2.30833620e-01 1.39719635e-01
2.33175680e-01 -1.05251503e+00 -6.32572353e-01 3.47915232e-01
-1.15391707e+00 -8.12980711e-01 1.63513273e-01 1.58941761e-01
-8.57202053e-01 1.81795254e-01 -8.40284407e-01 -8.92288566e-01
-7.61837780e-01 -1.40565789e+00 1.51214707e+00 -6.08055145e-02
-5.69886684e-01 -1.15500939e+00 4.92622197e-01 2.61513948e-01
8.22430789e-01 -5.06545864e-02 6.01135612e-01 -2.58825332e-01
-9.00766075e-01 -8.93300995e-02 -6.87324479e-02 -1.04709696e-02
-7.36927211e-01 5.21743834e-01 -1.00954902e+00 -6.36153400e-01
-1.86900392e-01 -5.80874383e-01 6.01251304e-01 3.38399291e-01
1.17472982e+00 -4.41768199e-01 -5.01637399e-01 7.88867593e-01
1.71735728e+00 -4.87267315e-01 6.80571139e-01 -1.10191733e-01
3.75444949e-01 -3.74838561e-01 3.27367127e-01 7.71178603e-01
4.08680946e-01 8.68780315e-01 1.18133768e-01 2.78399467e-01
-1.18779093e-01 -7.89574832e-02 5.63065946e-01 1.04752815e+00
3.73893559e-01 -3.31175983e-01 -1.46353841e+00 3.29651564e-01
-1.77818239e+00 -7.14852571e-01 -4.76736248e-01 2.27786899e+00
1.16796565e+00 3.83948296e-01 -8.42758343e-02 -2.18358517e-01
1.56948537e-01 8.59663039e-02 -1.78925753e-01 -1.08197176e+00
3.69660825e-01 6.87666476e-01 7.17617929e-01 9.40076172e-01
-7.11398959e-01 1.01046538e+00 7.74907064e+00 1.43766797e+00
-8.65982175e-01 5.76451719e-01 8.37368906e-01 -3.98419738e-01
-4.52631384e-01 4.75254804e-01 -1.34840417e+00 6.47845089e-01
2.12558961e+00 -3.19898725e-01 7.61274636e-01 9.45473373e-01
2.38202587e-02 -5.18866003e-01 -1.06292200e+00 5.84999025e-01
1.59528792e-01 -1.66157651e+00 -3.58775377e-01 3.09135884e-01
9.40397084e-01 4.71870363e-01 -1.42703339e-01 4.14376974e-01
8.57000351e-01 -9.24187005e-01 7.73621976e-01 3.47281992e-01
1.01263654e+00 -9.80771899e-01 7.84448206e-01 2.94584870e-01
-1.09962726e+00 6.27525032e-01 -2.20792845e-01 -3.01048398e-01
-4.23566140e-02 8.95846844e-01 -1.13587677e+00 2.21360445e-01
4.86393332e-01 1.15113392e-01 -7.43345439e-01 7.12555647e-01
-2.36786619e-01 1.17725754e+00 -5.50029993e-01 -1.89720288e-01
1.30495071e-01 1.43589810e-01 2.95092255e-01 1.63166416e+00
5.18196642e-01 -6.45661056e-02 1.30920112e-01 1.21870863e+00
1.30518109e-01 -2.57745445e-01 -5.22986650e-01 -1.52250621e-02
7.45908320e-01 1.47748089e+00 -8.26264024e-01 -6.59011722e-01
-1.35084331e-01 7.85967886e-01 2.89014459e-01 -6.94909692e-02
-1.17318392e+00 -1.09477833e-01 4.87948716e-01 -2.15521567e-02
2.62389362e-01 -4.22076583e-01 -5.26160657e-01 -1.11894727e+00
-5.17853439e-01 -1.03381276e+00 1.27791584e-01 -8.19013417e-01
-5.86951435e-01 4.86068428e-01 6.40758276e-02 -5.91086984e-01
-9.55236256e-01 -3.46617550e-01 -4.78374779e-01 9.13794279e-01
-7.87243485e-01 -1.06289279e+00 8.61315653e-02 -3.51842865e-02
2.68802881e-01 -1.22377940e-01 1.09971297e+00 2.26840198e-01
-4.26767230e-01 7.90005922e-01 4.53271747e-01 -3.90255451e-01
3.12993348e-01 -1.18587482e+00 8.94796610e-01 9.82074380e-01
3.55008692e-01 7.41195142e-01 9.98527586e-01 -5.85249126e-01
-1.95852101e+00 -8.09736252e-01 7.51675546e-01 -6.15988493e-01
8.79969776e-01 -8.10561359e-01 -4.26042318e-01 9.56186175e-01
6.57440007e-01 -1.01315491e-01 6.38420224e-01 3.82012635e-01
-6.85057789e-02 1.32842436e-01 -9.40027177e-01 4.90245134e-01
8.32651854e-01 -4.20992196e-01 1.48787335e-01 6.55381083e-01
3.13370526e-01 -1.01273143e+00 -9.21178639e-01 1.90199807e-01
4.11210328e-01 -1.13674474e+00 7.49451995e-01 1.76581636e-01
4.29057539e-01 -2.07592681e-01 -2.83336610e-01 -9.52059567e-01
-1.06550545e-01 -9.10421073e-01 -7.31489241e-01 1.35384190e+00
2.87475020e-01 -5.53520739e-01 7.34913290e-01 4.43933606e-01
-1.34788468e-01 -7.05459714e-01 -1.03927946e+00 -7.20701218e-01
2.35108986e-01 -8.01608682e-01 5.54950237e-01 3.99224520e-01
-1.50279040e-02 4.18638468e-01 -2.52672404e-01 2.02108875e-01
7.50848949e-01 2.92401612e-01 1.01509380e+00 -5.86196423e-01
-8.80195916e-01 -4.22107100e-01 -3.60417098e-01 -7.66963184e-01
3.57010841e-01 -1.16769207e+00 9.99335125e-02 -1.08892143e+00
4.69156027e-01 -6.43122315e-01 3.21929455e-01 6.30907476e-01
3.05519223e-01 6.11966431e-01 -2.17622351e-02 5.02608083e-02
-8.05281937e-01 2.27934942e-01 3.08923811e-01 2.08554342e-01
7.11249411e-02 -2.56615460e-01 -4.87965673e-01 4.35586393e-01
1.17288566e+00 -3.82866651e-01 -3.02149355e-01 -2.10666373e-01
4.55943763e-01 2.85213053e-01 3.74805212e-01 -1.45134449e+00
9.50350687e-02 2.34889805e-01 6.67774305e-02 -7.03916013e-01
2.77224839e-01 -1.99695617e-01 5.69468796e-01 6.88717246e-01
-9.31969360e-02 3.32297571e-02 6.95112348e-01 -7.31588900e-02
2.46885300e-01 -4.28570330e-01 6.28398836e-01 1.39403576e-02
-5.66420615e-01 -1.59178138e-01 -6.83670402e-01 2.10323155e-01
6.62328064e-01 3.66894871e-01 -6.02603197e-01 -4.25270468e-01
-2.70564616e-01 6.79968670e-03 7.75533915e-01 -4.38946746e-02
1.73967764e-01 -1.24512208e+00 -8.12278509e-01 3.83425683e-01
-1.66850179e-01 -5.09339511e-01 -1.72577366e-01 7.47114003e-01
-1.00044334e+00 8.19806099e-01 9.97101292e-02 -6.94672525e-01
-1.29201758e+00 1.96933195e-01 2.56324083e-01 -4.81969744e-01
-2.83513784e-01 8.10027421e-01 -4.89872992e-01 -4.32706684e-01
-2.23052964e-01 -3.73662740e-01 8.59876752e-01 -6.46196008e-01
3.51860821e-01 3.09747040e-01 2.82335758e-01 -4.22101945e-01
-4.81651485e-01 -5.46877459e-02 2.26749122e-01 -4.82923955e-01
8.55669916e-01 4.57415693e-02 -5.18128335e-01 6.23424232e-01
1.07432449e+00 2.05495879e-01 -8.02454114e-01 1.64511248e-01
-2.41669714e-01 -3.22533160e-01 6.77074969e-01 -8.43504667e-01
-8.15651238e-01 8.08009326e-01 5.92562497e-01 -1.56974774e-02
7.77638555e-01 1.47813531e-02 6.82490587e-01 3.21179867e-01
9.77177620e-01 -8.53840411e-01 -4.96865690e-01 5.90115964e-01
5.16813934e-01 -7.92543113e-01 5.46407521e-01 -3.33435774e-01
-5.91173507e-02 9.29009557e-01 3.37340623e-01 -1.08086117e-01
4.50468808e-01 1.15601921e+00 -2.84539938e-01 -4.51772138e-02
-1.39809394e+00 -1.92217026e-02 3.47986748e-03 2.96105206e-01
1.09739852e+00 3.02884519e-01 -4.66713965e-01 2.35043541e-01
-7.04885900e-01 1.99194282e-01 6.54234350e-01 1.01642430e+00
-2.51878053e-01 -1.38076162e+00 -5.75868547e-01 5.90864599e-01
-4.64089274e-01 -7.72231877e-01 -1.78145692e-01 5.56605756e-01
4.41315658e-02 6.05961919e-01 1.89231589e-01 -1.92189142e-01
-5.49376786e-01 1.87230811e-01 8.45301807e-01 -7.17061520e-01
-6.40909672e-01 1.09587841e-01 5.46101332e-01 -7.09546626e-01
9.48711783e-02 -9.57100928e-01 -1.37013853e+00 -9.77641821e-01
-2.64318377e-01 1.75085500e-01 8.32109809e-01 4.92162317e-01
8.01617026e-01 5.92324793e-01 1.67126924e-01 -1.06967795e+00
-4.99050230e-01 -9.15926576e-01 -3.69425416e-01 -4.75943089e-01
-8.34098533e-02 -2.12914318e-01 -2.66347528e-01 3.06229037e-03] | [8.663736343383789, 3.5102200508117676] |
17f75cf4-a0f7-4daf-b019-386b233e728b | deeper-and-wider-siamese-networks-for-real | 1901.01660 | null | http://arxiv.org/abs/1901.01660v3 | http://arxiv.org/pdf/1901.01660v3.pdf | Deeper and Wider Siamese Networks for Real-Time Visual Tracking | Siamese networks have drawn great attention in visual tracking because of
their balanced accuracy and speed. However, the backbone networks used in
Siamese trackers are relatively shallow, such as AlexNet [18], which does not
fully take advantage of the capability of modern deep neural networks. In this
paper, we investigate how to leverage deeper and wider convolutional neural
networks to enhance tracking robustness and accuracy. We observe that direct
replacement of backbones with existing powerful architectures, such as ResNet
[14] and Inception [33], does not bring improvements. The main reasons are that
1)large increases in the receptive field of neurons lead to reduced feature
discriminability and localization precision; and 2) the network padding for
convolutions induces a positional bias in learning. To address these issues, we
propose new residual modules to eliminate the negative impact of padding, and
further design new architectures using these modules with controlled receptive
field size and network stride. The designed architectures are lightweight and
guarantee real-time tracking speed when applied to SiamFC [2] and SiamRPN [20].
Experiments show that solely due to the proposed network architectures, our
SiamFC+ and SiamRPN+ obtain up to 9.8%/5.7% (AUC), 23.3%/8.8% (EAO) and
24.4%/25.0% (EAO) relative improvements over the original versions [2, 20] on
the OTB-15, VOT-16 and VOT-17 datasets, respectively. | ['Houwen Peng', 'Zhipeng Zhang'] | 2019-01-07 | deeper-and-wider-siamese-networks-for-real-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Zhang_Deeper_and_Wider_Siamese_Networks_for_Real-Time_Visual_Tracking_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Zhang_Deeper_and_Wider_Siamese_Networks_for_Real-Time_Visual_Tracking_CVPR_2019_paper.pdf | cvpr-2019-6 | ['real-time-visual-tracking'] | ['computer-vision'] | [-3.14487785e-01 -2.35293224e-01 -4.00850177e-01 1.93347633e-02
6.49286434e-02 -5.72247744e-01 5.18755138e-01 -3.32648784e-01
-8.40809166e-01 5.66907287e-01 -7.98157007e-02 -9.09747258e-02
9.17471126e-02 -5.50785601e-01 -7.55712628e-01 -4.23456818e-01
-3.47986430e-01 -2.91692108e-01 7.62571752e-01 -3.64740044e-01
4.84418049e-02 7.26778805e-01 -1.52459991e+00 -2.44867325e-01
9.16833937e-01 1.18353534e+00 -1.55043537e-02 5.28291404e-01
-4.71113026e-02 6.33405924e-01 -7.59787619e-01 -4.26573247e-01
5.95389009e-01 4.40831780e-02 -1.87819541e-01 -4.94439542e-01
8.61751139e-01 -3.97852540e-01 -7.79526055e-01 1.13582563e+00
4.76422936e-01 5.64787872e-02 3.75337154e-01 -1.43219352e+00
-4.93620306e-01 5.43067276e-01 -8.14781785e-01 2.76691258e-01
-3.22013378e-01 3.02898526e-01 8.06545079e-01 -6.93448305e-01
4.52792913e-01 1.08706057e+00 1.03799987e+00 7.34746873e-01
-8.86927485e-01 -1.10099280e+00 2.42920026e-01 -4.27450389e-02
-1.35067868e+00 -5.21146655e-01 4.84326005e-01 -2.93260127e-01
8.72244835e-01 6.61416203e-02 7.15505719e-01 1.16735983e+00
3.61677021e-01 7.10287869e-01 6.96572602e-01 -1.20524995e-01
-2.11138967e-02 1.14792213e-01 -3.10616102e-02 6.11974359e-01
8.11727405e-01 4.61239338e-01 -3.35822701e-01 1.82210624e-01
9.46389616e-01 2.15859681e-01 -2.65636921e-01 -3.52589905e-01
-1.17867935e+00 7.05306709e-01 1.23059309e+00 2.70622015e-01
-2.64992535e-01 5.61177552e-01 5.30028939e-01 1.24660335e-01
1.11435942e-01 5.31125903e-01 -2.71552861e-01 -8.02684128e-02
-1.01136696e+00 2.01090395e-01 4.19255674e-01 1.08046651e+00
4.79896933e-01 5.53706825e-01 -1.13210738e-01 5.27953684e-01
4.39193040e-01 8.07987869e-01 5.46522379e-01 -7.57400393e-01
4.61831748e-01 5.63326538e-01 2.11923987e-01 -1.10762143e+00
-6.05811894e-01 -9.77185547e-01 -6.47271097e-01 3.91800672e-01
6.00399375e-01 -2.24391833e-01 -9.55414414e-01 1.89993203e+00
2.01643333e-01 7.41589069e-02 2.94628572e-02 1.06968188e+00
8.41902137e-01 2.26592824e-01 2.49903843e-01 3.45940232e-01
1.38427913e+00 -1.15259099e+00 -4.95147407e-01 -4.20570910e-01
4.93462056e-01 -7.73030639e-01 8.44788790e-01 -6.52892292e-02
-8.56203914e-01 -9.21298981e-01 -1.29796016e+00 1.27220854e-01
-4.39190596e-01 4.14565593e-01 6.47694051e-01 7.89377868e-01
-1.07757235e+00 7.03094006e-01 -9.75780487e-01 -4.18605566e-01
4.79448467e-01 6.66565001e-01 -4.56808284e-02 3.80118698e-01
-1.22710061e+00 7.07286060e-01 2.98314303e-01 2.49314725e-01
-7.60858297e-01 -7.71857619e-01 -5.84529877e-01 1.46315873e-01
3.48961502e-01 -3.94926280e-01 1.01760268e+00 -9.12746549e-01
-1.44228637e+00 3.81480664e-01 2.02293694e-01 -9.00751650e-01
5.77325404e-01 -4.69410241e-01 -5.53038001e-01 -5.29744551e-02
-5.97776808e-02 1.26149690e+00 8.10309947e-01 -8.16730082e-01
-9.13147867e-01 -2.10898504e-01 2.74849951e-01 -8.11956301e-02
-5.60478330e-01 -1.49457097e-01 -4.82354790e-01 -8.29865754e-01
-1.89142957e-01 -1.10478461e+00 -1.83153331e-01 4.07561094e-01
-1.87223762e-01 -1.63097680e-01 1.07679486e+00 -3.52837831e-01
1.19932735e+00 -2.18565941e+00 -3.57294589e-01 -3.96547578e-02
5.21481216e-01 8.85952115e-01 -2.68623263e-01 8.31345767e-02
2.31055290e-01 -1.25000566e-01 3.09637100e-01 -1.14413410e-01
-7.81989470e-02 8.26580599e-02 -2.68974364e-01 7.14336097e-01
9.61012859e-03 9.46967125e-01 -7.26821184e-01 -3.62016827e-01
3.74510407e-01 6.64964557e-01 -6.07546628e-01 -1.37687460e-01
-1.25127405e-01 2.10813671e-01 -4.42107707e-01 6.04164243e-01
8.40323031e-01 -1.53624699e-01 -1.82387352e-01 -2.76283264e-01
-4.70160663e-01 1.85513481e-01 -9.99662757e-01 1.55193198e+00
-2.95195103e-01 1.00489414e+00 6.47443607e-02 -4.40421581e-01
9.89758611e-01 -8.37632827e-03 3.54350150e-01 -1.01599133e+00
3.27181906e-01 3.54406923e-01 4.08380121e-01 7.99386937e-04
8.08552444e-01 3.20891172e-01 -1.63242854e-02 -1.75302133e-01
4.52321582e-02 7.12881029e-01 1.88133270e-01 1.24373555e-01
9.01229560e-01 1.77192017e-01 -7.02749640e-02 -4.01358724e-01
5.91215670e-01 6.34462908e-02 6.94647014e-01 7.21578479e-01
-6.69942915e-01 2.04343230e-01 2.00757980e-01 -4.63827610e-01
-8.77626657e-01 -9.18373406e-01 -1.35982752e-01 1.09468877e+00
4.23366755e-01 -4.61413652e-01 -6.43890440e-01 -6.55943811e-01
2.32336685e-01 1.93335533e-01 -6.00254118e-01 -2.43920907e-01
-8.48620474e-01 -3.52299333e-01 1.08862698e+00 9.52971637e-01
9.64765489e-01 -7.81828642e-01 -9.88302350e-01 1.75629303e-01
2.52395749e-01 -1.21473241e+00 -6.13200426e-01 -3.59430760e-02
-9.11512613e-01 -9.54382300e-01 -8.42421532e-01 -5.90364337e-01
5.23463011e-01 5.23674667e-01 6.10987961e-01 3.78720500e-02
-5.89238331e-02 -7.79356435e-02 -1.04139350e-01 -3.40644777e-01
-2.88736727e-02 4.27555561e-01 2.74670988e-01 -1.05664410e-01
2.43187681e-01 -2.42878005e-01 -9.55509782e-01 6.37156963e-01
-5.38213968e-01 -1.79904610e-01 7.23548770e-01 6.69659972e-01
2.33861044e-01 -3.32286716e-01 3.67022306e-01 -5.19744694e-01
9.33382437e-02 -1.10296972e-01 -1.16580009e+00 -1.11197077e-01
-7.30302989e-01 6.11241050e-02 9.87856150e-01 -6.42722964e-01
-7.35468149e-01 -6.84245601e-02 -1.24173388e-01 -8.19970846e-01
1.22152038e-01 9.91890673e-03 1.94231689e-01 -7.01181352e-01
8.31747591e-01 2.34119166e-02 1.06190681e-01 -4.40447152e-01
1.63564265e-01 4.66824859e-01 5.55413365e-01 -6.22853190e-02
1.01720583e+00 6.13574386e-01 -3.23502086e-02 -6.27592444e-01
-5.80749929e-01 -3.30327749e-01 -2.81394899e-01 -2.00151742e-01
8.02683592e-01 -1.18917906e+00 -1.10366118e+00 3.32623780e-01
-7.99180269e-01 -2.14425325e-01 -2.35159978e-01 7.56368518e-01
-2.75128409e-02 1.43083423e-01 -4.59042758e-01 -5.56583405e-01
-5.68283200e-01 -1.17685521e+00 8.03921044e-01 8.83122563e-01
-2.26641558e-02 -8.34717691e-01 -1.79596737e-01 -6.15441129e-02
1.18207490e+00 2.99789637e-01 2.21852466e-01 -4.89109784e-01
-6.68612540e-01 -1.19948044e-01 -4.90340978e-01 2.23268703e-01
-8.79609063e-02 -1.34664471e-04 -9.02506709e-01 -7.78366566e-01
-5.33667922e-01 -3.45588997e-02 9.11117375e-01 4.07123715e-01
8.34067881e-01 -1.87702373e-01 -5.11697114e-01 9.34719980e-01
1.40753901e+00 2.33324200e-01 4.75049913e-01 4.82374400e-01
7.75241971e-01 1.93958446e-01 4.78088707e-01 1.31189376e-01
2.58120388e-01 9.66311336e-01 7.72003233e-01 -9.06911418e-02
-5.88253021e-01 -4.50943679e-01 6.10579371e-01 5.03443062e-01
-2.62416661e-01 -9.63039398e-02 -6.68430328e-01 5.80023825e-01
-1.67310905e+00 -7.85350680e-01 -1.42086685e-01 2.26643109e+00
3.44052970e-01 4.46080148e-01 3.46652925e-01 -2.82681942e-01
7.01629341e-01 3.92484486e-01 -6.43491566e-01 -1.38120010e-01
2.64170719e-03 9.93854553e-02 1.30086362e+00 2.16988996e-02
-1.25156271e+00 1.04199922e+00 5.58722448e+00 9.15286720e-01
-1.53494287e+00 1.68773998e-02 7.79515505e-03 -3.39776844e-01
2.90632188e-01 -1.54814839e-01 -1.26840425e+00 5.72454393e-01
1.09745479e+00 7.53127038e-02 2.44114026e-01 1.01175511e+00
-1.48888692e-01 2.11886242e-01 -7.64204443e-01 9.62555051e-01
-1.60254374e-01 -1.32424903e+00 -2.77077317e-01 2.15206593e-01
6.87624097e-01 5.21522999e-01 3.33349168e-01 5.63401759e-01
1.71191394e-01 -9.51811969e-01 9.52465713e-01 1.38683483e-01
6.99990273e-01 -7.90287495e-01 7.64791548e-01 -4.29304019e-02
-1.59611666e+00 -1.11090638e-01 -6.18207932e-01 8.60219747e-02
-1.47687256e-01 1.73749730e-01 -4.83369082e-01 3.39054316e-01
9.14016545e-01 6.75328970e-01 -8.18907440e-01 1.28876591e+00
-2.14842960e-01 5.72677433e-01 -4.74154741e-01 -3.60294133e-01
6.71468854e-01 3.88332307e-01 6.49019361e-01 1.15029883e+00
1.21014230e-01 -7.09083021e-01 -1.61611978e-02 8.53740811e-01
-2.73070425e-01 -1.98942825e-01 -4.80196148e-01 9.64108482e-02
6.35644138e-01 1.37261045e+00 -6.85651004e-01 -1.36644244e-01
-5.33423901e-01 4.92665619e-01 2.25295037e-01 3.50638807e-01
-1.36408901e+00 -7.97764659e-01 1.00943184e+00 1.60893247e-01
7.06826925e-01 -2.56492436e-01 -7.87168518e-02 -9.65449631e-01
-3.73477824e-02 -7.07592726e-01 1.19494282e-01 -2.21943900e-01
-8.70279968e-01 7.32252300e-01 -2.44331315e-01 -1.45695257e+00
-6.33026361e-02 -8.93858612e-01 -5.23545742e-01 6.71070695e-01
-1.56328821e+00 -1.02488577e+00 -4.43484873e-01 5.63078165e-01
3.43885422e-01 -3.03043693e-01 2.52998233e-01 6.29105091e-01
-6.18100703e-01 1.18268204e+00 1.01816118e-01 4.66862261e-01
7.07383752e-01 -9.85939205e-01 6.81098938e-01 9.75399435e-01
1.74792483e-02 8.65028203e-01 4.78129566e-01 -3.89490604e-01
-1.61131954e+00 -1.36192775e+00 5.95805407e-01 -2.57976741e-01
6.77057028e-01 -4.51487124e-01 -7.09156573e-01 5.99270761e-01
1.29396275e-01 4.17417735e-01 1.60729825e-01 1.39311142e-02
-4.98993009e-01 -4.99670774e-01 -1.12529182e+00 9.00713205e-01
1.19814157e+00 -1.16612434e-01 -1.82819709e-01 -1.37179494e-01
7.42401898e-01 -5.80053210e-01 -8.19001913e-01 4.42856133e-01
9.01107669e-01 -9.60045159e-01 1.02182364e+00 -3.01279753e-01
-1.11866899e-01 -4.73329842e-01 -1.34830978e-02 -9.94034588e-01
-3.60940725e-01 -5.60333252e-01 -2.88162172e-01 1.04548180e+00
3.34322721e-01 -1.09894824e+00 9.87701118e-01 1.51392698e-01
-1.48969203e-01 -6.18833840e-01 -1.07240963e+00 -1.09092307e+00
4.22401819e-03 -1.35087818e-01 6.14183426e-01 6.16198063e-01
-5.58342278e-01 6.82141930e-02 -4.09063339e-01 9.15012211e-02
7.41185248e-01 -1.82834733e-02 9.38071549e-01 -1.25499547e+00
-3.22536640e-02 -6.61853611e-01 -6.38979256e-01 -1.25755072e+00
-1.68555498e-01 -6.67764366e-01 -1.74649686e-01 -1.06574750e+00
-2.18478665e-01 -7.09248483e-01 -5.15794337e-01 5.06772935e-01
6.42882809e-02 4.58191395e-01 5.11868000e-01 3.74134958e-01
-6.49990439e-01 5.46319544e-01 1.19676936e+00 2.95795687e-02
-1.19275458e-01 1.03539638e-02 -5.75549603e-01 6.87531888e-01
8.51564944e-01 -4.60842729e-01 -2.45486379e-01 -5.14321029e-01
-5.71373664e-02 -3.87593150e-01 5.90547681e-01 -1.46437871e+00
5.32008708e-01 3.04027617e-01 6.69469774e-01 -5.50689995e-01
2.23845750e-01 -9.03241277e-01 8.56808051e-02 8.97596121e-01
-9.75918621e-02 1.45480052e-01 6.00527704e-01 5.41504323e-01
-2.54600883e-01 9.90736634e-02 9.63324904e-01 2.31861264e-01
-9.65584397e-01 3.48885179e-01 -6.07573986e-02 3.16872485e-02
9.91401255e-01 -3.76566261e-01 -7.48609006e-01 -3.57604958e-02
-1.53528884e-01 4.72688824e-01 3.91271591e-01 7.97461212e-01
3.51277262e-01 -1.47053194e+00 -3.04849476e-01 2.53166437e-01
-2.05927454e-02 -2.81811684e-01 1.44004002e-01 1.13755870e+00
-6.30176544e-01 8.42725515e-01 -5.73679328e-01 -7.72202849e-01
-9.83057082e-01 3.89412522e-01 4.60687846e-01 -2.52988130e-01
-7.17474222e-01 7.74972260e-01 3.00632000e-01 -1.64128542e-01
6.12387121e-01 -4.63734657e-01 -1.82556435e-01 -1.20367184e-01
3.88148278e-01 5.25188684e-01 -9.10000280e-02 -7.03090549e-01
-6.42498374e-01 6.71793222e-01 -3.29717487e-01 3.52405310e-01
9.34003174e-01 1.11595862e-01 5.77834785e-01 -2.94272870e-01
1.04467356e+00 2.01328546e-02 -1.62497282e+00 -2.19611630e-01
-3.54011469e-02 -4.40915048e-01 1.49332494e-01 -5.71433246e-01
-1.62021911e+00 8.23205411e-01 8.78795743e-01 -3.82886492e-02
1.00310731e+00 -2.62249649e-01 9.94608760e-01 2.45276004e-01
3.87574613e-01 -8.68009508e-01 -1.16515346e-01 5.07751942e-01
4.84947443e-01 -9.64898705e-01 -1.42115280e-01 -9.87104252e-02
-3.91218722e-01 9.57340539e-01 1.07469308e+00 -3.90015811e-01
2.74699181e-01 3.22810531e-01 -6.02425658e-04 -2.05857247e-01
-5.03598750e-01 -4.29281384e-01 3.66920918e-01 4.25317287e-01
3.40008110e-01 -9.79861990e-02 -2.10496902e-01 2.43676573e-01
-1.72789454e-01 -2.02010833e-02 8.99351910e-02 8.40658307e-01
-5.28468609e-01 -7.33387232e-01 -4.05140758e-01 2.35940486e-01
-6.22784674e-01 9.98486802e-02 -5.06715402e-02 1.23432469e+00
2.98001766e-01 6.57859683e-01 6.46359399e-02 -5.26348531e-01
5.96254528e-01 -4.14589316e-01 3.25845420e-01 -5.56785008e-03
-9.12492037e-01 6.40804321e-02 -7.60865808e-02 -7.36814618e-01
-2.99036890e-01 -2.32428700e-01 -1.28360808e+00 -6.07346356e-01
-5.02611816e-01 7.39635676e-02 7.04962611e-01 5.03792584e-01
5.55930197e-01 7.89031088e-01 2.62285888e-01 -7.91622281e-01
-7.29606152e-01 -9.54445839e-01 -4.02143031e-01 1.08168125e-02
3.98837149e-01 -1.06170332e+00 -2.91238695e-01 -4.79838401e-01] | [6.259329795837402, -2.1181745529174805] |
e85b701d-62c2-43f6-91e8-3d8d8e87bbb8 | multimodal-semi-supervised-learning-for3d | 2110.11601 | null | https://arxiv.org/abs/2110.11601v2 | https://arxiv.org/pdf/2110.11601v2.pdf | Multimodal Semi-Supervised Learning for 3D Objects | In recent years, semi-supervised learning has been widely explored and shows excellent data efficiency for 2D data. There is an emerging need to improve data efficiency for 3D tasks due to the scarcity of labeled 3D data. This paper explores how the coherence of different modelities of 3D data (e.g. point cloud, image, and mesh) can be used to improve data efficiency for both 3D classification and retrieval tasks. We propose a novel multimodal semi-supervised learning framework by introducing instance-level consistency constraint and a novel multimodal contrastive prototype (M2CP) loss. The instance-level consistency enforces the network to generate consistent representations for multimodal data of the same object regardless of its modality. The M2CP maintains a multimodal prototype for each class and learns features with small intra-class variations by minimizing the feature distance of each object to its prototype while maximizing the distance to the others. Our proposed framework significantly outperforms all the state-of-the-art counterparts for both classification and retrieval tasks by a large margin on the modelNet10 and ModelNet40 datasets. | ['Bing Li', 'YingLi Tian', 'Yang Liang', 'Longlong Jing', 'Zhimin Chen'] | 2021-10-22 | null | null | null | null | ['3d-classification'] | ['computer-vision'] | [ 1.34349670e-02 3.30294445e-02 -4.19045568e-01 -6.88586712e-01
-1.14056933e+00 -4.85061347e-01 7.90287733e-01 3.67283612e-01
-2.70059913e-01 3.78370821e-01 -9.31231305e-02 1.91742226e-01
-2.97823876e-01 -6.46397114e-01 -8.21078420e-01 -8.07724178e-01
-3.40389013e-02 7.47769713e-01 1.17536470e-01 6.45473823e-02
2.14011803e-01 7.69090831e-01 -1.91534805e+00 4.28547502e-01
6.41094446e-01 1.49139786e+00 3.49593431e-01 8.01310465e-02
-3.84433866e-01 3.66662115e-01 -2.54206479e-01 -1.27281502e-01
5.06325126e-01 6.73356429e-02 -7.12275445e-01 3.42661798e-01
9.23143923e-01 -9.04605761e-02 -2.91279793e-01 9.09053087e-01
7.55628347e-01 1.25031918e-01 1.00339842e+00 -1.53511000e+00
-6.69589341e-01 -1.05430009e-02 -6.94777727e-01 -3.40776771e-01
1.71831578e-01 -2.87634075e-01 1.04951870e+00 -1.41604078e+00
7.02019811e-01 1.48566866e+00 2.27367401e-01 4.86148417e-01
-1.10297096e+00 -5.46388626e-01 1.65738314e-01 9.48970914e-02
-1.73904335e+00 -2.19974637e-01 9.71820652e-01 -2.59832948e-01
8.77632260e-01 3.24149206e-02 3.08245122e-01 8.27107549e-01
-1.43356666e-01 9.39485312e-01 9.25907731e-01 -3.33627820e-01
1.94091484e-01 4.05395567e-01 1.18272072e-02 8.68149519e-01
-1.81635711e-02 -2.12521404e-01 -8.88615787e-01 -3.02153200e-01
5.69622517e-01 1.47056013e-01 -3.34546268e-02 -1.08055997e+00
-1.17796338e+00 7.55546391e-01 6.81833744e-01 -1.55980796e-01
-1.03336900e-01 -1.32663757e-01 2.99820006e-01 3.13272476e-01
6.02678537e-01 -2.81659719e-02 -3.64429057e-01 2.32358351e-01
-7.42355883e-01 2.17752144e-01 5.24530113e-01 1.28774440e+00
7.68512070e-01 -3.33915114e-01 6.98892325e-02 1.28863633e+00
7.30118811e-01 8.94760549e-01 1.28252521e-01 -8.03236067e-01
6.92605555e-01 1.03577900e+00 -1.30133450e-01 -9.19817686e-01
-3.37640971e-01 -2.64182985e-01 -9.82796252e-01 1.67218804e-01
1.76574454e-01 4.25580114e-01 -8.44356954e-01 1.77021945e+00
6.18679225e-01 -8.89629722e-02 -5.39141074e-02 1.04768443e+00
1.34054458e+00 6.50249064e-01 -8.66093040e-02 9.50525478e-02
9.44909990e-01 -8.35151017e-01 -1.87410831e-01 8.90091620e-03
5.85658550e-01 -7.90286541e-01 8.67217720e-01 1.56032816e-01
-1.19345379e+00 -6.33617461e-01 -9.99809623e-01 -2.76328564e-01
-3.82318377e-01 1.22556351e-01 3.56605887e-01 3.88124526e-01
-8.62746358e-01 2.99085647e-01 -6.82021737e-01 -4.22909796e-01
6.30641997e-01 4.39381182e-01 -8.93763900e-01 -3.95804852e-01
-8.24521601e-01 8.88139844e-01 3.37622523e-01 -2.46891789e-02
-8.06596041e-01 -6.96603954e-01 -1.04833663e+00 -1.32770076e-01
-8.14350508e-03 -6.81791425e-01 6.76199496e-01 -6.77090049e-01
-1.15576935e+00 1.50589240e+00 1.14867114e-01 2.60130484e-02
4.64748859e-01 -3.93335335e-02 -1.70021340e-01 3.58855546e-01
2.50879806e-02 1.24920428e+00 1.08371162e+00 -1.64806223e+00
-5.16834676e-01 -7.88989723e-01 -1.42272845e-01 5.62593877e-01
-3.76967669e-01 -4.41932648e-01 -6.59118295e-01 -3.22260380e-01
6.16136789e-01 -1.24730504e+00 1.73334256e-01 5.68730116e-01
-3.78257930e-01 -5.94098270e-01 7.91667223e-01 3.30652073e-02
6.57479107e-01 -2.19278574e+00 4.96410549e-01 4.51328248e-01
7.35465065e-02 5.37508773e-03 -4.57654774e-01 3.34624410e-01
-4.75627929e-02 -2.08157495e-01 -2.69892812e-01 -8.28063548e-01
1.90285578e-01 3.11880887e-01 -1.19939104e-01 7.74967194e-01
4.20090795e-01 6.95477664e-01 -7.42637455e-01 -7.25291252e-01
3.91389817e-01 7.46948361e-01 -4.22971219e-01 3.08288932e-01
1.95431244e-02 2.74299115e-01 -3.68813008e-01 1.06596875e+00
1.01993239e+00 -5.70837796e-01 -6.08286522e-02 -4.34793055e-01
1.26278162e-01 -3.99767011e-02 -1.22254479e+00 2.12493014e+00
-4.47535187e-01 3.00185800e-01 -9.94126275e-02 -1.14680469e+00
1.11474097e+00 8.82569104e-02 5.57963610e-01 -7.63536870e-01
-6.85861930e-02 4.19111937e-01 -5.45554757e-01 -4.17641431e-01
4.72327083e-01 2.02460140e-02 -1.51832039e-02 3.75249118e-01
2.82325894e-01 -2.74140298e-01 -1.28022581e-01 2.27571443e-01
5.22505581e-01 1.69302553e-01 -1.73210695e-01 -2.45819062e-01
5.03163755e-01 -1.33842066e-01 1.76269621e-01 7.65238225e-01
1.45785585e-01 8.89636636e-01 8.85551050e-02 -1.83043644e-01
-9.88094151e-01 -1.30319750e+00 -5.00802040e-01 8.56680334e-01
6.64549291e-01 -5.16514629e-02 -8.56499225e-02 -7.09763885e-01
4.46162671e-01 3.21356237e-01 -5.08502722e-01 -3.47701818e-01
-2.40759939e-01 -4.37952667e-01 3.12862217e-01 3.73727590e-01
5.02325714e-01 -5.54689050e-01 -2.62264907e-01 -3.03764254e-01
-1.05011322e-01 -1.15912068e+00 -2.79065728e-01 1.44889355e-01
-1.08737016e+00 -8.74886215e-01 -9.27280128e-01 -1.13319790e+00
1.11820018e+00 5.97319663e-01 1.19121516e+00 5.83038330e-02
-1.44431859e-01 7.66548932e-01 -4.36850578e-01 -2.89886951e-01
-1.96623653e-01 2.06987679e-01 1.01175830e-01 1.08275399e-01
2.99547464e-01 -3.05465221e-01 -6.77766025e-01 4.96405691e-01
-1.05762780e+00 -9.39770788e-02 5.84925175e-01 7.69535899e-01
1.05117369e+00 -2.55170077e-01 3.00166279e-01 -4.70886230e-01
2.24316910e-01 -6.44252539e-01 -3.52547139e-01 3.77559543e-01
-5.51380634e-01 8.38920549e-02 1.44385353e-01 -5.06580472e-01
-7.55591512e-01 1.17857486e-01 2.70029068e-01 -7.92196333e-01
-1.38549298e-01 4.53324109e-01 -3.12640429e-01 -2.89948761e-01
4.48059887e-01 1.49142250e-01 3.10632825e-01 -6.44634187e-01
4.12514955e-01 7.61675775e-01 2.02587306e-01 -7.37679064e-01
7.27978051e-01 7.84825742e-01 4.99945968e-01 -1.02228379e+00
-8.56821418e-01 -5.25760829e-01 -6.93950415e-01 -2.11145371e-01
5.20532727e-01 -1.14472616e+00 -5.52157700e-01 5.16728044e-01
-1.03316855e+00 2.93962769e-02 -1.24394983e-01 5.28407395e-01
-4.33597028e-01 3.81170630e-01 -2.89407820e-01 -5.90892971e-01
-2.75224328e-01 -1.07302356e+00 1.70111525e+00 2.94151325e-02
2.02554286e-01 -7.77072549e-01 -1.01701450e-02 3.08967173e-01
1.11277871e-01 1.02782011e-01 1.17007029e+00 -7.04033554e-01
-5.71069837e-01 -4.20963168e-01 -5.76234043e-01 4.20027792e-01
5.97384237e-02 -8.25178549e-02 -8.97967160e-01 -4.16743815e-01
-1.96888372e-01 -8.06418717e-01 9.02555048e-01 2.39307553e-01
1.05745113e+00 7.35275596e-02 -2.76508749e-01 4.23315734e-01
1.27371037e+00 -3.33569288e-01 2.93177187e-01 -5.72275892e-02
7.15425372e-01 8.18790376e-01 7.29609847e-01 4.32972938e-01
5.09429336e-01 7.18437374e-01 7.41515636e-01 -5.01340143e-02
-1.45055979e-01 -1.87820435e-01 -4.97486666e-02 7.63248205e-01
1.76298529e-01 -2.64103889e-01 -8.86370480e-01 5.34095287e-01
-1.89511311e+00 -6.26245618e-01 1.16824687e-01 2.39941359e+00
5.28881550e-01 -1.39070461e-02 -8.71862248e-02 -2.20063254e-02
7.49073029e-01 7.57029057e-02 -5.49648881e-01 1.60738736e-01
-4.01429564e-01 -7.45082349e-02 2.30395600e-01 1.86546296e-01
-1.22653985e+00 4.62705612e-01 5.32656908e+00 9.36653197e-01
-1.11603534e+00 -1.33419737e-01 5.21669090e-01 -2.72357851e-01
-3.06831449e-01 -3.75454247e-01 -8.58037412e-01 1.82214350e-01
4.18595165e-01 2.49909446e-01 1.40448585e-02 8.36740732e-01
-5.25109954e-02 -7.02369288e-02 -1.56859100e+00 1.53492224e+00
4.56512094e-01 -1.27881455e+00 5.92823029e-01 1.76401421e-01
8.60219300e-01 1.72710851e-01 3.89276594e-01 8.04083273e-02
-4.18533176e-01 -9.77194369e-01 8.20487320e-01 6.47904634e-01
7.83860743e-01 -7.87850201e-01 6.02387726e-01 3.25290054e-01
-1.10985470e+00 1.02693669e-01 -4.87249404e-01 4.71269071e-01
-4.98298444e-02 4.54999954e-01 -4.16974932e-01 5.37389874e-01
7.01801956e-01 8.19938064e-01 -5.83996952e-01 1.06329477e+00
1.75368801e-01 4.93948814e-03 -6.01864338e-01 -1.50480056e-02
3.91230762e-01 -1.98908642e-01 6.43392086e-01 9.00090992e-01
2.96822995e-01 -2.16036767e-01 4.33205724e-01 7.35782981e-01
-2.85770923e-01 2.94242024e-01 -6.14859998e-01 1.64314851e-01
6.90601051e-01 1.11103535e+00 -4.38073128e-01 -1.70174390e-01
-4.18468326e-01 7.37209678e-01 5.31355858e-01 1.21900328e-01
-5.60436845e-01 -1.22479849e-01 4.11087245e-01 2.71687526e-02
2.29356244e-01 -3.13828439e-01 -2.45878637e-01 -9.21234250e-01
2.35825270e-01 -4.39630210e-01 6.09477878e-01 -8.27799320e-01
-1.74964404e+00 5.40150166e-01 2.54692048e-01 -1.68417478e+00
9.55738425e-02 -6.74684405e-01 -7.67246783e-02 6.65872157e-01
-1.67966533e+00 -1.41693604e+00 -4.76069599e-01 7.93626606e-01
2.85087317e-01 -3.89074892e-01 7.66030729e-01 5.12101829e-01
-7.31958523e-02 5.93750715e-01 1.95791036e-01 -1.07383072e-01
9.18289661e-01 -1.07794797e+00 -2.83538043e-01 -5.34746945e-02
1.51109070e-01 5.08681118e-01 1.84492156e-01 -3.40332717e-01
-1.69948721e+00 -1.01695323e+00 8.39049578e-01 -4.04570132e-01
2.69428402e-01 -2.93503433e-01 -8.00958574e-01 4.02816422e-02
-2.02405751e-01 2.97690123e-01 6.82182610e-01 -1.40869349e-01
-7.08810091e-01 -2.96896815e-01 -1.26906466e+00 4.08579111e-01
1.04174101e+00 -7.44720578e-01 -2.98626781e-01 5.93728721e-01
4.96195972e-01 -4.80236351e-01 -9.82591450e-01 6.96623027e-01
5.76229870e-01 -7.96198487e-01 1.15937555e+00 -6.17726624e-01
5.23933887e-01 -2.26439267e-01 -6.71289086e-01 -9.86250162e-01
1.66118462e-02 -4.47125584e-02 -2.67650396e-01 1.01581216e+00
4.29870248e-01 -2.49063343e-01 7.79530227e-01 7.04778075e-01
-1.64662734e-01 -9.12898600e-01 -1.29123354e+00 -8.19664836e-01
2.89996594e-01 -4.40699041e-01 3.86082768e-01 8.03434014e-01
-2.06882566e-01 1.72420755e-01 -2.24602953e-01 2.81234026e-01
9.40732718e-01 5.42717099e-01 7.72537172e-01 -1.30168247e+00
2.01435566e-01 -4.18363899e-01 -6.27541363e-01 -1.36187673e+00
3.45919520e-01 -1.40029180e+00 -1.31431758e-01 -1.29293728e+00
4.24793184e-01 -7.37151563e-01 -3.06370795e-01 3.53603542e-01
2.13141114e-01 5.54914594e-01 1.57314286e-01 4.19652045e-01
-8.43623519e-01 8.85790408e-01 1.30318689e+00 -4.18204755e-01
-1.89591631e-01 -1.58562183e-01 -2.34237567e-01 4.01748657e-01
5.04117727e-01 -4.44187343e-01 -4.73356783e-01 -6.64821029e-01
2.21864283e-01 -1.54483274e-01 6.67398453e-01 -6.14117503e-01
3.62605900e-01 1.74236149e-01 4.55012321e-01 -9.84270990e-01
7.28179991e-01 -1.18203592e+00 -4.76102084e-02 -1.76464375e-02
-6.40448391e-01 -1.16485797e-01 7.05452934e-02 7.15626478e-01
-3.17336977e-01 -1.79586217e-01 8.84635687e-01 -2.18714904e-02
-4.89872277e-01 6.21732295e-01 2.65245020e-01 1.29684538e-01
8.09608221e-01 -2.35154212e-01 -2.46580437e-01 -2.00169012e-01
-5.03881931e-01 3.78850937e-01 4.43669170e-01 7.21281469e-01
9.42081273e-01 -1.76248670e+00 -7.82111585e-01 3.16071033e-01
7.29997456e-01 2.70489126e-01 4.08647209e-01 6.59135580e-01
-1.97254047e-01 4.00374115e-01 -1.49942279e-01 -1.27247238e+00
-1.32527292e+00 2.37165630e-01 2.94807851e-01 8.72531161e-02
-3.91454577e-01 9.03047979e-01 5.30121066e-02 -8.40244651e-01
5.97263932e-01 1.06309056e-01 1.19966760e-01 1.55802384e-01
2.82792747e-01 3.28351796e-01 2.13143125e-01 -1.01519763e+00
-5.75730085e-01 8.23019564e-01 -2.05182180e-01 1.18889458e-01
1.41897571e+00 -9.37942043e-02 -1.56208396e-01 5.68874776e-01
1.77519596e+00 -5.96555710e-01 -1.11918569e+00 -6.02707028e-01
-2.43754253e-01 -6.22166812e-01 1.83947548e-01 -5.37859559e-01
-1.04100394e+00 8.98250937e-01 8.65630805e-01 -7.94569328e-02
7.84535050e-01 4.91544962e-01 5.14027476e-01 6.83119118e-01
4.54164475e-01 -1.18569708e+00 3.11924934e-01 4.41099584e-01
9.68640387e-01 -1.60377467e+00 9.98486504e-02 -3.66757989e-01
-6.07155919e-01 9.63609159e-01 4.82335746e-01 -2.00251728e-01
8.86751354e-01 -3.33270758e-01 -1.03151880e-01 -4.30369794e-01
-6.00188673e-01 -4.33678031e-02 8.28300536e-01 5.79656124e-01
2.34713808e-01 -1.81190699e-01 2.03030974e-01 1.39221981e-01
3.31621945e-01 -4.34696257e-01 -2.44747251e-01 1.05668426e+00
-2.93572605e-01 -9.60430741e-01 -3.55138987e-01 4.47238892e-01
1.21874228e-01 1.94341645e-01 -5.06815493e-01 8.35363686e-01
-8.01210478e-02 7.35928237e-01 2.14864895e-01 -2.19120473e-01
3.37293535e-01 -4.67745662e-02 8.66216481e-01 -4.77784246e-01
-1.20600432e-01 7.47583732e-02 -2.03787714e-01 -4.25103813e-01
-9.48857903e-01 -5.74703693e-01 -1.24529314e+00 -5.29057607e-02
-3.61183256e-01 -6.06513917e-02 9.26392317e-01 7.64397502e-01
5.76678395e-01 -2.31394514e-01 8.84745419e-01 -1.14517164e+00
-6.68088794e-01 -6.43674850e-01 -7.29149640e-01 7.52203465e-01
1.90122992e-01 -1.07036495e+00 -3.95147979e-01 -2.48455793e-01] | [8.148521423339844, -3.579679250717163] |
dfdc85b5-dd05-4d02-a6c9-aa5c2cbe236f | multimodal-image-to-image-translation-via | 2008.03529 | null | https://arxiv.org/abs/2008.03529v7 | https://arxiv.org/pdf/2008.03529v7.pdf | Multimodal Image-to-Image Translation via Mutual Information Estimation and Maximization | Multimodal image-to-image translation (I2IT) aims to learn a conditional distribution that explores multiple possible images in the target domain given an input image in the source domain. Conditional generative adversarial networks (cGANs) are often adopted for modeling such a conditional distribution. However, cGANs are prone to ignore the latent code and learn a unimodal distribution in conditional image synthesis, which is also known as the mode collapse issue of GANs. To solve the problem, we propose a simple yet effective method that explicitly estimates and maximizes the mutual information between the latent code and the output image in cGANs by using a deep mutual information neural estimator in this paper. Maximizing the mutual information strengthens the statistical dependency between the latent code and the output image, which prevents the generator from ignoring the latent code and encourages cGANs to fully utilize the latent code for synthesizing diverse results. Our method not only provides a new perspective from information theory to improve diversity for I2IT but also achieves disentanglement between the source domain content and the target domain style for free. | ['Qijiang Xu', 'Ailin Li', 'Zhizhong Wang', 'Lei Zhao', 'Haibo Chen', 'Zhiwen Zuo', 'Wei Xing', 'Dongming Lu'] | 2020-08-08 | null | null | null | null | ['mutual-information-estimation'] | ['methodology'] | [ 4.19554979e-01 3.64251584e-01 -1.29762128e-01 -1.45768762e-01
-8.96705627e-01 -7.79718876e-01 6.65362656e-01 -8.05842280e-01
1.61279038e-01 8.42997968e-01 3.16929311e-01 -3.51544544e-02
2.71234423e-01 -7.72646606e-01 -9.76663232e-01 -1.15548754e+00
6.03559434e-01 2.39680126e-01 -4.45026666e-01 -8.93329829e-03
-1.69324175e-01 9.19764191e-02 -1.19208276e+00 1.97107032e-01
1.15193272e+00 8.12116683e-01 4.85102475e-01 5.69440186e-01
-1.71379194e-01 8.75961304e-01 -8.62897813e-01 -6.57097757e-01
1.50628507e-01 -1.22923112e+00 -3.41927677e-01 2.54508018e-01
2.47339889e-01 -3.01045299e-01 -4.12199169e-01 1.44063258e+00
3.54191005e-01 -2.37381011e-01 1.12503409e+00 -1.58910894e+00
-1.11135256e+00 4.59212035e-01 -5.87539256e-01 -5.70316732e-01
1.71487212e-01 4.35182005e-01 8.57806861e-01 -5.89139998e-01
7.69624650e-01 1.30448234e+00 7.87599236e-02 7.53294230e-01
-1.40337002e+00 -8.73586416e-01 -1.02027640e-01 -1.33843303e-01
-1.38528419e+00 -2.53536344e-01 9.59673285e-01 -5.42265654e-01
9.37307328e-02 2.67930388e-01 5.55252194e-01 1.53200257e+00
1.48969755e-01 1.06419349e+00 1.23492455e+00 -5.02780974e-01
1.36058226e-01 3.85999113e-01 -8.75116110e-01 5.51097333e-01
-1.31279111e-01 3.81331801e-01 -5.11971056e-01 1.85744837e-01
1.12694645e+00 -1.44098923e-01 -3.89100641e-01 -4.02746707e-01
-1.24998951e+00 1.01117361e+00 5.12774408e-01 1.94682583e-01
-1.79375395e-01 1.33499429e-01 -5.62944338e-02 3.36911350e-01
8.55943263e-02 4.72635686e-01 -1.29579663e-01 -3.63226794e-02
-8.55742872e-01 7.67756179e-02 6.03981435e-01 1.10203099e+00
8.37079227e-01 3.09345663e-01 -5.16450107e-01 7.21075177e-01
3.09789836e-01 1.01296556e+00 2.94901043e-01 -1.09260309e+00
5.61388969e-01 4.84323859e-01 1.13731518e-03 -8.81283224e-01
4.61198270e-01 -4.95006293e-01 -1.13961470e+00 4.61154491e-01
3.86927098e-01 -4.38731313e-01 -9.47687984e-01 2.36744118e+00
-8.02820176e-02 -2.31253594e-01 1.47113979e-01 8.64118040e-01
5.45506179e-01 8.16757858e-01 -1.70416549e-01 -2.44717181e-01
8.76352966e-01 -9.38616574e-01 -9.67473924e-01 -2.45539770e-01
3.40282149e-03 -8.15460443e-01 1.23847568e+00 1.31623074e-01
-1.10598838e+00 -4.71617728e-01 -1.10982037e+00 4.22201343e-02
6.14770874e-02 3.06220084e-01 3.80493492e-01 6.27470255e-01
-9.85783160e-01 2.80689709e-02 -5.17150164e-01 4.65455502e-02
5.12067378e-01 7.06528574e-02 -3.53717953e-01 -3.68841588e-01
-1.14905071e+00 7.88639724e-01 2.79589266e-01 -2.17791140e-01
-1.06654489e+00 -4.67194110e-01 -8.62119973e-01 7.65729994e-02
1.83767036e-01 -8.44920039e-01 7.78260946e-01 -1.72772360e+00
-1.88256919e+00 6.81659281e-01 -1.42077923e-01 5.18580452e-02
6.78880036e-01 8.34768936e-02 -1.46987990e-01 1.93962560e-03
8.05687755e-02 1.12393486e+00 1.45483029e+00 -1.62503564e+00
-7.18320832e-02 1.58789959e-02 -4.09983806e-02 3.74352902e-01
-1.27016574e-01 -4.88187075e-01 -5.67022622e-01 -9.76092756e-01
-6.03030398e-02 -1.03963184e+00 8.53001252e-02 7.13176355e-02
-9.22484696e-01 2.48597816e-01 6.64469123e-01 -6.89281106e-01
7.91488588e-01 -2.43984246e+00 6.63107574e-01 1.16375253e-01
7.71725252e-02 -1.69752583e-01 -3.02164644e-01 3.84836048e-01
-7.09077492e-02 3.80278267e-02 -4.98424023e-01 -2.97327191e-01
1.16247423e-01 3.14894259e-01 -4.93490428e-01 1.70533255e-01
3.42058182e-01 1.19994986e+00 -7.68466532e-01 -4.38468158e-01
6.68371692e-02 6.27620339e-01 -6.46141410e-01 5.98446190e-01
-4.48362201e-01 1.15314269e+00 -2.63408929e-01 4.43090856e-01
8.90151143e-01 -2.26526842e-01 9.73090306e-02 -2.37779975e-01
1.95943236e-01 -2.50081599e-01 -7.65319705e-01 1.83747804e+00
-3.43523979e-01 7.93994904e-01 -7.76956677e-02 -7.41127133e-01
9.67435181e-01 2.01506227e-01 2.55877614e-01 -6.49396181e-01
1.10273406e-01 4.11675833e-02 -7.21390918e-02 -4.08399165e-01
1.71461955e-01 -5.62766790e-01 -2.31706545e-01 3.27155381e-01
3.08708698e-01 -3.68021905e-01 -1.04348138e-01 2.73953676e-01
5.97174883e-01 3.05697262e-01 -1.71424169e-02 -1.57109559e-01
2.49044001e-01 -4.17593569e-01 5.82043707e-01 7.00549781e-01
1.72473490e-01 1.08219278e+00 9.13735151e-01 1.74295202e-01
-1.23736966e+00 -1.60265601e+00 1.76294446e-01 5.55163145e-01
2.26897240e-01 4.81359176e-02 -8.66472781e-01 -6.39154971e-01
-2.67259300e-01 9.64695334e-01 -6.54888988e-01 -3.56048077e-01
-9.89324972e-02 -3.06520104e-01 5.62141895e-01 2.37972572e-01
8.60140562e-01 -8.06175351e-01 -7.25224242e-02 -1.58146665e-01
-7.77143836e-01 -8.86941791e-01 -9.00268972e-01 -8.99390802e-02
-4.91886079e-01 -7.83615410e-01 -1.25362408e+00 -7.21900105e-01
9.89066184e-01 -2.56530404e-01 1.02424705e+00 -4.86942023e-01
-7.90674798e-03 2.90091932e-01 -1.35702193e-01 -2.69838601e-01
-8.11985254e-01 -1.19249046e-01 -3.05588514e-01 2.41885722e-01
-4.85888161e-02 -6.32943630e-01 -6.19550467e-01 3.51294577e-01
-1.13809729e+00 6.05602264e-01 7.51402199e-01 1.05138457e+00
3.79645675e-01 1.01872936e-01 5.29460907e-01 -5.95889509e-01
4.34073508e-01 -4.62724358e-01 -5.75348258e-01 3.84080946e-01
-2.68972605e-01 4.09640282e-01 4.70652372e-01 -6.83948100e-01
-1.35438526e+00 7.08761066e-02 -1.15616836e-01 -6.33746982e-01
7.02696741e-02 3.86048645e-01 -7.24892676e-01 1.86951667e-01
4.90816176e-01 5.88590741e-01 1.70153961e-01 -9.08866823e-02
6.76131904e-01 3.32112134e-01 6.74659252e-01 -7.21080005e-01
1.03962171e+00 3.89979243e-01 -9.22480971e-03 -4.90824819e-01
-6.27082944e-01 3.57162863e-01 -3.62366945e-01 -2.40492359e-01
1.25122845e+00 -1.11219895e+00 -4.75163788e-01 5.76283753e-01
-1.21235204e+00 -2.66584158e-01 -4.35952157e-01 4.16664690e-01
-7.15746820e-01 1.19533285e-01 -3.49585205e-01 -7.48202860e-01
1.45284593e-01 -1.35558021e+00 1.02583075e+00 4.09943908e-01
4.54546288e-02 -9.76538777e-01 -3.10571082e-02 2.67721772e-01
4.66720045e-01 4.82117802e-01 1.12981701e+00 7.96295982e-03
-9.96776283e-01 -4.78045940e-02 -3.25108558e-01 6.71130300e-01
2.47980177e-01 -3.82232964e-02 -8.94733131e-01 -3.32610369e-01
1.47046298e-01 -4.42527145e-01 7.49074936e-01 4.74335670e-01
9.16992843e-01 -5.88809967e-01 4.60388139e-02 7.36907303e-01
1.47578537e+00 2.28803053e-01 1.07727957e+00 -1.69336781e-01
8.57216716e-01 5.08636177e-01 2.50729889e-01 2.71458209e-01
2.25110441e-01 5.71310461e-01 4.20726418e-01 -3.12946349e-01
-2.53173620e-01 -8.19700301e-01 6.26261234e-01 8.29445302e-01
2.40814984e-01 -5.26031435e-01 -4.03232515e-01 4.24266458e-01
-1.46710503e+00 -8.61591578e-01 4.10245270e-01 2.16725898e+00
1.17603159e+00 -1.78620428e-01 -2.54783124e-01 -3.20004910e-01
1.05774105e+00 6.34629428e-02 -7.95132399e-01 -1.34477258e-01
-4.56530899e-01 3.24549675e-02 2.54470021e-01 4.52013016e-01
-7.18206763e-01 7.46158957e-01 6.13746119e+00 1.24477112e+00
-1.17420328e+00 -1.17064156e-02 9.00269926e-01 1.27366912e-02
-8.14759135e-01 9.13467333e-02 -2.48931497e-01 7.79156089e-01
4.41958576e-01 -1.40694588e-01 6.37288153e-01 6.95799768e-01
-1.85219765e-01 -7.09665716e-02 -1.06761491e+00 1.13868380e+00
1.01217337e-01 -1.22516382e+00 2.65722871e-01 3.46718550e-01
1.33549523e+00 -4.49167609e-01 6.29783154e-01 2.11752117e-01
5.42898357e-01 -1.27946651e+00 7.96227515e-01 6.89533532e-01
1.40897167e+00 -8.66093814e-01 4.21607167e-01 3.97566557e-01
-6.93902612e-01 7.72772208e-02 -2.15656489e-01 3.07116091e-01
9.90295261e-02 6.16210163e-01 -4.81562018e-01 4.92250025e-01
2.06947044e-01 3.91687363e-01 -3.71904373e-01 5.53472996e-01
-7.17664063e-01 3.78298432e-01 -3.04399803e-02 3.56866002e-01
8.99844319e-02 -5.22010386e-01 7.65729010e-01 8.09868515e-01
6.64857149e-01 -3.55652452e-01 -2.80642025e-02 1.51038635e+00
-3.10195237e-01 -1.62585825e-01 -8.38761270e-01 -3.72985989e-01
3.23089808e-01 8.36049676e-01 -3.55993479e-01 -1.63118497e-01
-1.90549284e-01 1.48607004e+00 9.10304412e-02 7.97921479e-01
-9.62944865e-01 -1.98845282e-01 5.17184198e-01 -1.73699036e-01
3.42005134e-01 -1.54016316e-01 -5.27029872e-01 -1.36095655e+00
8.53420980e-03 -1.09139276e+00 -6.34902641e-02 -1.05815911e+00
-1.32419002e+00 4.55919772e-01 1.73948389e-02 -1.48137045e+00
-3.73384625e-01 -2.55827695e-01 -6.35151446e-01 1.17815566e+00
-1.27624941e+00 -1.40009332e+00 -1.65021017e-01 7.00954914e-01
3.13345671e-01 -4.47569400e-01 7.58696735e-01 9.29381400e-02
-3.20854038e-01 8.72013509e-01 3.52387160e-01 1.36434942e-01
8.34644437e-01 -1.23291862e+00 5.51581383e-03 8.36396277e-01
4.87349778e-02 6.36297584e-01 7.07187235e-01 -5.92173278e-01
-1.23043966e+00 -8.60873818e-01 4.69432265e-01 -3.67219806e-01
2.95700639e-01 -3.22146446e-01 -5.42040527e-01 5.60894132e-01
6.36656046e-01 -4.60670233e-01 6.89281762e-01 -3.12044144e-01
-4.86101598e-01 6.13647848e-02 -1.17131341e+00 8.59114170e-01
6.92288458e-01 -7.23148048e-01 -6.38119131e-02 7.08158612e-02
8.48243654e-01 -3.44044656e-01 -6.68274939e-01 2.11504385e-01
4.13579404e-01 -1.03527236e+00 7.67423391e-01 -1.42905787e-01
1.08538067e+00 -2.73352236e-01 -3.36626321e-01 -1.54800951e+00
-5.80681227e-02 -6.88519955e-01 8.25743303e-02 1.41256547e+00
4.64179009e-01 -4.69436079e-01 6.22401893e-01 5.79264760e-01
1.54944807e-01 -3.15651774e-01 -9.14163053e-01 -7.24514306e-01
3.67164820e-01 -1.47702560e-01 5.94592690e-01 9.13670003e-01
-2.86989450e-01 2.94111401e-01 -8.74844432e-01 1.13505404e-02
8.05735826e-01 1.84167728e-01 7.15665579e-01 -6.09192610e-01
-5.73930979e-01 -4.17682230e-01 -2.24807248e-01 -1.37077570e+00
1.49036705e-01 -8.77199650e-01 2.18612462e-01 -1.21537220e+00
5.12277663e-01 -2.06318781e-01 3.67372669e-03 2.10239589e-01
-1.34048477e-01 1.54998288e-01 3.82302254e-01 3.67587984e-01
-1.27979264e-01 9.66776729e-01 1.98295987e+00 -3.40881705e-01
5.05993813e-02 -1.10405073e-01 -9.82595921e-01 3.55229914e-01
5.59306979e-01 -5.10354996e-01 -7.78577745e-01 -5.95047235e-01
2.82343388e-01 3.51691931e-01 4.15936142e-01 -7.65101135e-01
3.60295363e-02 -3.49838972e-01 4.64507997e-01 -3.07324022e-01
3.20969403e-01 -7.85565853e-01 4.73423243e-01 2.97550023e-01
-4.59096938e-01 -3.24316531e-01 -1.88562036e-01 7.16940701e-01
-4.66973811e-01 -1.62104055e-01 1.02605534e+00 -2.43009925e-01
-1.96799472e-01 2.61920482e-01 -2.61424959e-01 1.05981298e-01
8.59362423e-01 -1.77016836e-02 -2.41535246e-01 -9.19748247e-01
-6.27986789e-01 9.02820844e-03 7.58998334e-01 3.64427209e-01
5.59292495e-01 -1.80955052e+00 -8.22588444e-01 4.86307353e-01
6.03324920e-02 -3.84268053e-02 3.24320048e-01 4.79929507e-01
-3.25570554e-01 1.30961880e-01 -3.77371997e-01 -6.94430053e-01
-7.05419064e-01 3.99870276e-01 2.61687726e-01 -2.89952427e-01
-2.39439473e-01 9.15280521e-01 9.40864325e-01 -4.23171580e-01
-3.30395997e-02 2.20411390e-01 1.45557284e-01 -1.30714551e-01
1.68629915e-01 -2.05654606e-01 -6.14739358e-01 -6.57381415e-01
1.27794132e-01 3.10009629e-01 1.44984931e-01 -5.21743596e-01
8.26571703e-01 -3.54454666e-01 -8.65751281e-02 3.62188399e-01
1.51532876e+00 -1.99721265e-03 -1.86066675e+00 -7.07758218e-02
-6.16230667e-01 -7.07652271e-01 -8.55511725e-02 -9.39204335e-01
-1.27268612e+00 9.35180902e-01 5.01199603e-01 -3.72562222e-02
1.28129470e+00 3.79757956e-02 6.76733553e-01 -1.06696360e-01
5.44574969e-02 -8.66549551e-01 6.77405179e-01 3.72153848e-01
1.14749229e+00 -1.19294846e+00 -3.83257061e-01 -2.10324556e-01
-1.13983750e+00 9.49558794e-01 5.92207730e-01 -3.59061211e-02
3.94504160e-01 2.14266211e-01 8.81086737e-02 5.75624891e-02
-4.21200007e-01 -2.36598589e-02 3.14303160e-01 9.31184530e-01
2.13234991e-01 2.52551854e-01 7.27185160e-02 3.70679826e-01
-2.15270415e-01 -2.06079766e-01 3.64274561e-01 5.36098719e-01
5.59250638e-02 -1.26459396e+00 -3.68346334e-01 1.05519956e-02
-2.68429011e-01 -2.09082395e-01 -3.94367725e-01 5.59763193e-01
2.62412846e-01 7.61692464e-01 -9.95327458e-02 -4.06936914e-01
-2.03059167e-01 9.44668427e-02 6.72987223e-01 -2.61794001e-01
1.41896069e-01 2.91729122e-01 -3.91525537e-01 -2.46369854e-01
-3.05301070e-01 -5.55558860e-01 -6.78090394e-01 -2.09776312e-01
-2.87956923e-01 -2.03259028e-02 7.64765561e-01 8.85134161e-01
2.18224809e-01 6.42763376e-01 8.94756734e-01 -6.74843550e-01
-3.48613113e-01 -7.78512537e-01 -5.86298108e-01 5.18813074e-01
3.93764287e-01 -4.66744035e-01 -4.13804770e-01 3.31002265e-01] | [11.629826545715332, -0.29737934470176697] |
21eff4f7-23cf-4b24-aba0-c2f69badba3f | adam-few-shot-image-generation-via-adaptation | 2307.01465 | null | https://arxiv.org/abs/2307.01465v2 | https://arxiv.org/pdf/2307.01465v2.pdf | AdAM: Few-Shot Image Generation via Adaptation-Aware Kernel Modulation | Few-shot image generation (FSIG) aims to learn to generate new and diverse images given few (e.g., 10) training samples. Recent work has addressed FSIG by leveraging a GAN pre-trained on a large-scale source domain and adapting it to the target domain with few target samples. Central to recent FSIG methods are knowledge preservation criteria, which select and preserve a subset of source knowledge to the adapted model. However, a major limitation of existing methods is that their knowledge preserving criteria consider only source domain/task and fail to consider target domain/adaptation in selecting source knowledge, casting doubt on their suitability for setups of different proximity between source and target domain. Our work makes two contributions. Firstly, we revisit recent FSIG works and their experiments. We reveal that under setups which assumption of close proximity between source and target domains is relaxed, many existing state-of-the-art (SOTA) methods which consider only source domain in knowledge preserving perform no better than a baseline method. As our second contribution, we propose Adaptation-Aware kernel Modulation (AdAM) for general FSIG of different source-target domain proximity. Extensive experiments show that AdAM consistently achieves SOTA performance in FSIG, including challenging setups where source and target domains are more apart. | ['Ngai-Man Cheung', 'Henghui Ding', 'Ruoteng Li', 'Tianyu Pang', 'Chao Du', 'Abdollahzadeh Milad', 'Keshigeyan Chandrasegaran', 'Yunqing Zhao'] | 2023-07-04 | null | null | null | null | ['image-generation'] | ['computer-vision'] | [ 5.83728254e-01 3.46069261e-02 -4.09274340e-01 -1.93269029e-01
-9.54933405e-01 -6.18717194e-01 8.60823452e-01 -5.13193905e-01
-1.39304623e-01 1.14062619e+00 1.00226127e-01 3.13495845e-01
-1.21514007e-01 -7.34154522e-01 -8.84311676e-01 -7.70704865e-01
3.38430375e-01 4.73158509e-01 4.60152030e-01 -2.48264208e-01
-5.10914288e-02 2.16730669e-01 -1.54733372e+00 2.31352746e-01
1.03067541e+00 9.55156565e-01 2.96315491e-01 6.19483232e-01
1.70948654e-01 9.07855570e-01 -1.04286528e+00 -2.92338252e-01
5.20173490e-01 -1.09716153e+00 -8.36726189e-01 1.65757790e-01
7.93858647e-01 -2.69614726e-01 -2.45906442e-01 1.01124501e+00
7.10069120e-01 2.57742137e-01 8.87983084e-01 -1.42132032e+00
-1.24903452e+00 2.41549194e-01 -2.67477721e-01 2.62266427e-01
1.55943677e-01 3.15096766e-01 5.57238996e-01 -8.82666528e-01
9.10799086e-01 9.79281843e-01 5.77531695e-01 8.96995604e-01
-1.25225794e+00 -6.67075098e-01 2.76987165e-01 4.53203708e-01
-1.43500865e+00 -6.53595328e-01 8.15970659e-01 -2.10334107e-01
8.69380951e-01 1.83139533e-01 4.38673556e-01 1.55811882e+00
-1.80738911e-01 7.27739692e-01 1.38111424e+00 -6.62800968e-01
6.10689342e-01 4.16072398e-01 -2.93326795e-01 2.90384471e-01
1.03044622e-01 2.60493547e-01 -9.22742605e-01 -2.50587255e-01
8.56640995e-01 -4.25891548e-01 -4.16243970e-01 -5.60032129e-01
-1.39741659e+00 8.13734055e-01 1.14808410e-01 2.28922427e-01
-2.33977512e-01 -3.06264162e-02 2.26094350e-01 7.17923522e-01
5.46588838e-01 3.93334985e-01 -3.11919630e-01 -6.87853470e-02
-1.03897977e+00 1.81681901e-01 7.94962168e-01 1.35812068e+00
8.82221282e-01 4.55380112e-01 -5.41325569e-01 9.86311197e-01
-4.15701002e-01 7.07432091e-01 7.33273268e-01 -1.05100417e+00
2.69047588e-01 1.53732702e-01 1.39736399e-01 -4.39068526e-01
3.93398076e-01 -3.58453155e-01 -8.29951823e-01 3.15655857e-01
3.40033799e-01 -1.63285941e-01 -1.16435027e+00 1.99442887e+00
2.99537897e-01 5.68026602e-01 3.26521516e-01 7.95685709e-01
6.68245375e-01 5.34182847e-01 -2.00688779e-01 -2.60776371e-01
1.02729428e+00 -1.07124269e+00 -4.58547473e-01 -2.91236430e-01
-1.36871506e-02 -6.56383276e-01 1.36837363e+00 2.94021130e-01
-1.02155745e+00 -7.29558766e-01 -9.53316033e-01 2.51143306e-01
-4.91674632e-01 3.13953422e-02 3.58009249e-01 6.94396257e-01
-1.27474523e+00 4.73669887e-01 -2.68191487e-01 -6.61345363e-01
6.15240991e-01 5.74817657e-02 -2.35362619e-01 -3.60590428e-01
-1.41797066e+00 9.29431677e-01 5.38299739e-01 -5.40300071e-01
-1.20802629e+00 -1.00530565e+00 -7.38307595e-01 -1.58496991e-01
5.07107913e-01 -1.04911757e+00 1.23580146e+00 -1.54460359e+00
-1.85108364e+00 8.55441988e-01 -1.30244330e-01 -5.57885051e-01
6.22829199e-01 -1.40179753e-01 -6.37219489e-01 2.94752836e-01
2.01741785e-01 8.83267999e-01 1.51629579e+00 -1.39485943e+00
-6.34398103e-01 -2.06911359e-02 -2.50332039e-02 3.35280955e-01
-4.43354845e-01 -1.74520478e-01 -3.91389102e-01 -9.32113349e-01
-6.16944313e-01 -7.79269695e-01 7.33045712e-02 9.23792925e-03
-2.97719359e-01 -9.68104899e-02 9.90875244e-01 -4.19649661e-01
1.06096113e+00 -2.06263614e+00 1.58066705e-01 -4.22648713e-02
-8.68967548e-02 5.70524454e-01 -4.09860492e-01 5.74729085e-01
1.97250489e-03 -2.75434852e-01 -4.63560283e-01 -1.95643842e-01
3.75389792e-02 3.38929564e-01 -6.77189767e-01 1.94729343e-01
3.03421438e-01 1.08278608e+00 -9.99793947e-01 -4.84765857e-01
1.75309509e-01 4.47014034e-01 -1.62453637e-01 2.27247179e-01
-2.82114655e-01 4.87718821e-01 -1.90808088e-01 5.75686336e-01
7.36913502e-01 -2.67095923e-01 -1.29974917e-01 -4.88880426e-02
2.48159260e-01 -1.89488530e-01 -1.07555223e+00 1.75618613e+00
-3.65235597e-01 6.06853664e-01 -2.66387284e-01 -1.00678575e+00
9.13274467e-01 1.36630818e-01 2.03088954e-01 -7.16018796e-01
-3.23140770e-01 1.79916382e-01 -2.53409803e-01 -1.30196303e-01
1.95792422e-01 -2.99320668e-01 -1.73479132e-02 2.98273563e-01
4.00401473e-01 -2.23722249e-01 1.58571437e-01 2.84422964e-01
1.19140947e+00 2.99362808e-01 4.83379841e-01 -1.84408903e-01
2.58993685e-01 1.34955391e-01 6.31710529e-01 1.11294460e+00
-4.02628213e-01 1.00715327e+00 2.87352920e-01 -2.12333933e-01
-1.22071719e+00 -1.37989509e+00 9.97224450e-02 9.13059592e-01
1.93535715e-01 1.05602659e-01 -8.21262479e-01 -1.10871017e+00
-6.33626804e-02 1.03544176e+00 -8.83142173e-01 -4.98822123e-01
-4.40810740e-01 -5.47642946e-01 6.12720370e-01 5.37771583e-01
8.97459269e-01 -1.11624146e+00 -6.07259035e-01 1.79257686e-03
-5.99687845e-02 -1.10958946e+00 -6.74464464e-01 -4.73308042e-02
-6.90429091e-01 -9.46575522e-01 -1.33697438e+00 -9.09693480e-01
6.46591485e-01 3.29392642e-01 1.31482065e+00 -5.87029159e-01
7.52405226e-02 7.90803075e-01 -5.08214056e-01 -3.32288682e-01
-5.36360264e-01 5.96321188e-02 -1.66555211e-01 1.45627141e-01
3.94270837e-01 -6.63492143e-01 -5.50955772e-01 4.13401783e-01
-1.14866078e+00 3.39590497e-02 6.55183017e-01 1.06580269e+00
7.56913066e-01 -1.85638759e-02 1.09030068e+00 -1.10861170e+00
5.62503815e-01 -6.51553571e-01 -2.81018198e-01 5.87029040e-01
-6.46135986e-01 -7.68552572e-02 9.94514048e-01 -8.00398648e-01
-1.43779826e+00 -1.01231277e-01 3.61257553e-01 -8.06031525e-01
-4.62172300e-01 4.46073823e-02 -2.91198373e-01 -1.29514679e-01
1.10964191e+00 7.41532445e-01 -1.03881069e-01 -2.08733156e-01
4.11354154e-01 4.50568825e-01 5.83868265e-01 -4.89241272e-01
9.40167546e-01 6.57081246e-01 -3.29837054e-01 -7.34788299e-01
-8.64240766e-01 -2.75154918e-01 -5.65109611e-01 -2.42766179e-03
4.23017085e-01 -1.00103068e+00 1.04176879e-01 7.62300730e-01
-8.43253255e-01 -7.95253634e-01 -9.04369473e-01 1.45868510e-01
-9.37601089e-01 3.26227665e-01 -7.67903775e-02 -4.09634501e-01
-3.26423913e-01 -6.35405123e-01 1.00815034e+00 2.60642201e-01
-1.65668324e-01 -1.11739337e+00 2.28740573e-01 1.37436077e-01
8.07428181e-01 4.44478482e-01 6.70657098e-01 -5.89872003e-01
-4.18558806e-01 1.92176357e-01 -1.75759718e-01 7.13467896e-01
4.28987235e-01 -5.46687603e-01 -1.06272793e+00 -4.14745480e-01
1.28442362e-01 -5.87080002e-01 1.05898309e+00 3.90106112e-01
9.28097546e-01 -4.74129081e-01 -1.94674850e-01 6.40519798e-01
1.50585234e+00 1.89922318e-01 8.26910198e-01 2.45443538e-01
4.68560338e-01 3.79906148e-01 8.03078949e-01 3.75385255e-01
2.39625826e-01 7.47925401e-01 -1.42337848e-02 -1.96321625e-02
-8.78404140e-01 -3.29415828e-01 7.12688982e-01 3.88690680e-01
9.06695947e-02 -2.66579926e-01 -4.85994577e-01 1.07867742e+00
-1.89556098e+00 -1.10367668e+00 6.13055348e-01 2.43375278e+00
1.25975728e+00 -1.39744341e-01 3.20846230e-01 -2.87183493e-01
8.73225629e-01 1.22516692e-01 -1.10082734e+00 -1.00913160e-01
-5.24479449e-01 5.34126759e-01 3.89398366e-01 3.47693801e-01
-9.71129775e-01 1.09092391e+00 5.93815231e+00 1.37502348e+00
-1.03930223e+00 4.96806145e-01 5.45447826e-01 -3.90071481e-01
-2.66375810e-01 -3.59188911e-04 -7.11490512e-01 5.95587492e-01
8.71475756e-01 -4.15335000e-01 5.71806490e-01 9.09108579e-01
-2.95023054e-01 -1.82904139e-01 -1.02636349e+00 9.44566429e-01
5.00351608e-01 -1.29765677e+00 1.83377668e-01 -2.31475621e-01
1.36933899e+00 -8.89722332e-02 3.73251945e-01 4.24046099e-01
5.49535394e-01 -9.18806791e-01 7.07733095e-01 5.65781176e-01
1.29138184e+00 -6.78247750e-01 4.89065439e-01 2.71967530e-01
-8.07330847e-01 9.37411189e-02 -5.81548989e-01 2.13656723e-01
-4.25381847e-02 6.72942162e-01 -9.24764574e-01 5.57801783e-01
7.08450794e-01 6.74532533e-01 -4.72185165e-01 8.11956227e-01
-3.50673884e-01 8.03658247e-01 -2.03752875e-01 3.27392876e-01
-3.74359675e-02 1.30894095e-01 6.56510592e-01 1.13461483e+00
4.90611851e-01 -2.17790753e-01 -1.95342842e-02 1.08315754e+00
-3.66078224e-03 -2.29261234e-01 -8.87447834e-01 1.07612059e-01
7.36500680e-01 7.13061094e-01 -4.44361538e-01 -6.31058335e-01
-3.08838964e-01 1.59881318e+00 3.41867864e-01 8.10530663e-01
-8.84451032e-01 -5.81094801e-01 7.68384814e-01 7.40740523e-02
7.00382233e-01 6.46755695e-02 -1.38600513e-01 -1.36946213e+00
2.52151966e-01 -9.28999603e-01 5.40515482e-01 -6.15883112e-01
-1.67204821e+00 5.40855467e-01 3.17109615e-01 -1.38737559e+00
-3.66285473e-01 -1.76835939e-01 -5.89023173e-01 8.30294430e-01
-1.88167250e+00 -1.52618361e+00 -3.57975125e-01 1.05398798e+00
7.96727777e-01 -5.28362095e-01 9.47851956e-01 -5.79710416e-02
-1.85590759e-01 8.49599123e-01 4.51987773e-01 -2.22455427e-01
1.08856380e+00 -1.23204219e+00 4.48701292e-01 9.64695096e-01
2.58105516e-01 3.92157793e-01 4.70208645e-01 -7.63404191e-01
-1.11266565e+00 -1.36335814e+00 7.24312961e-01 -6.18045509e-01
2.43609875e-01 -4.10110116e-01 -8.44597757e-01 6.80157006e-01
4.83397067e-01 3.19522589e-01 6.25688791e-01 -3.69533420e-01
-5.51389515e-01 -2.28536427e-01 -1.43564045e+00 5.34837723e-01
1.31750047e+00 -3.57804298e-01 -4.71255779e-01 1.61859930e-01
5.95169127e-01 -3.37563932e-01 -7.79418886e-01 3.06196213e-01
1.95063889e-01 -1.17373991e+00 9.61791039e-01 -4.00398165e-01
2.94895113e-01 -4.67678070e-01 -1.06946103e-01 -1.75965118e+00
-2.93490857e-01 -7.52220035e-01 -5.01202166e-01 1.28458190e+00
6.92766756e-02 -7.88605452e-01 6.65696859e-01 2.83139557e-01
2.13656351e-02 -3.69585872e-01 -1.09996843e+00 -1.43062508e+00
2.75581211e-01 -9.03720856e-02 5.86945057e-01 9.96778846e-01
-3.34596932e-01 2.47360811e-01 -6.75937533e-01 -9.22519863e-02
7.54548669e-01 2.35833749e-01 9.14035916e-01 -9.32211876e-01
-4.36036259e-01 -1.88207418e-01 -2.25492865e-01 -8.64484966e-01
1.42414421e-01 -6.47251010e-01 -3.50669436e-02 -1.45128667e+00
2.32140705e-01 -2.57640302e-01 -4.02578712e-01 6.02911770e-01
-3.52015853e-01 3.42611998e-01 2.04028726e-01 3.88774186e-01
-6.87482417e-01 8.27228248e-01 1.16680932e+00 -8.65955278e-02
-1.88823447e-01 -1.48235515e-01 -8.68921101e-01 3.22110742e-01
8.79057765e-01 -3.99118394e-01 -9.67467427e-01 -3.61010134e-01
-1.75227329e-01 -4.30587709e-01 6.44871950e-01 -1.24252975e+00
9.59238708e-02 -3.92470002e-01 4.86134589e-01 4.64890637e-02
2.72191077e-01 -6.07021391e-01 3.21902841e-01 1.28851503e-01
-2.75220901e-01 -5.41760027e-01 2.47652724e-01 7.45132148e-01
-3.70311260e-01 -1.87491760e-01 1.11614597e+00 -1.73110932e-01
-1.22308266e+00 1.97800800e-01 -4.49010767e-02 4.91129130e-01
1.34920883e+00 -5.20977080e-01 -4.78644848e-01 -5.20327568e-01
-4.40843254e-01 -1.58352971e-01 7.72472680e-01 4.25354123e-01
8.18390012e-01 -1.60903990e+00 -9.46898997e-01 2.66937166e-01
4.30600196e-01 -4.05141432e-03 3.79347593e-01 5.68746388e-01
1.68028078e-03 1.29530177e-01 -3.14040959e-01 -3.30035836e-01
-9.77067471e-01 5.83514273e-01 2.28603259e-01 -1.98779091e-01
-5.16883373e-01 1.03335762e+00 4.41164047e-01 -3.18339467e-01
-6.10189978e-04 6.63777888e-02 1.39446527e-01 -1.66702986e-01
6.07034683e-01 4.13577348e-01 -6.68975934e-02 -4.28715199e-01
-2.06786379e-01 5.02479017e-01 -2.10223928e-01 -1.35071844e-01
1.13131869e+00 -1.66890010e-01 3.87883693e-01 3.46347183e-01
8.93177330e-01 -1.23586327e-01 -1.67487204e+00 -5.66384137e-01
-3.52215201e-01 -7.10445702e-01 -3.14294130e-01 -1.19113100e+00
-1.00034165e+00 7.33632922e-01 5.92127681e-01 -1.53068781e-01
1.63172495e+00 1.55581072e-01 9.04912829e-01 1.26812607e-01
5.93749881e-01 -1.42385006e+00 4.04802114e-01 4.37570959e-01
1.02488232e+00 -1.28354931e+00 -2.99846172e-01 -2.12968782e-01
-1.13744903e+00 8.44425976e-01 9.03586030e-01 -1.20882146e-01
3.29533011e-01 -3.85538340e-02 8.71674437e-03 1.03502594e-01
-7.75270224e-01 -4.30889308e-01 2.05921754e-01 1.42737305e+00
-1.93702027e-01 -1.56581894e-01 1.39700755e-01 5.26770949e-01
-4.94482629e-02 4.23460037e-01 3.49282950e-01 8.44670594e-01
-3.28560710e-01 -1.11013389e+00 -3.57335418e-01 3.72688890e-01
-1.41389802e-01 -2.29443327e-01 -6.27341509e-01 6.81650162e-01
3.07199687e-01 8.38696957e-01 -1.21402599e-01 -1.52389869e-01
3.35757107e-01 1.79593191e-01 7.10729420e-01 -7.70821035e-01
-3.44669998e-01 -2.62425691e-01 -1.07736982e-01 -4.36688751e-01
-5.38627505e-01 -6.34028673e-01 -8.13101053e-01 -9.64392275e-02
-1.62354499e-01 1.05011184e-02 5.19008487e-02 6.47226751e-01
8.23078573e-01 1.76004559e-01 5.22845984e-01 -6.18597806e-01
-6.21915281e-01 -8.36271584e-01 -6.93552315e-01 6.12856448e-01
2.70828307e-01 -6.73045754e-01 -2.25365832e-01 4.97732013e-01] | [10.24509334564209, 2.7154366970062256] |
4b994a5d-12d5-4174-af23-e7f212bf04ef | the-manipulation-problem-conversational-ai-as | 2306.11748 | null | https://arxiv.org/abs/2306.11748v1 | https://arxiv.org/pdf/2306.11748v1.pdf | The Manipulation Problem: Conversational AI as a Threat to Epistemic Agency | The technology of Conversational AI has made significant advancements over the last eighteen months. As a consequence, conversational agents are likely to be deployed in the near future that are designed to pursue targeted influence objectives. Sometimes referred to as the "AI Manipulation Problem," the emerging risk is that consumers will unwittingly engage in real-time dialog with predatory AI agents that can skillfully persuade them to buy particular products, believe particular pieces of misinformation, or fool them into revealing sensitive personal data. For many users, current systems like ChatGPT and LaMDA feel safe because they are primarily text-based, but the industry is already shifting towards real-time voice and photorealistic digital personas that look, move, and express like real people. This will enable the deployment of agenda-driven Virtual Spokespeople (VSPs) that will be highly persuasive through real-time adaptive influence. This paper explores the manipulative tactics that are likely to be deployed through conversational AI agents, the unique threats such agents pose to the epistemic agency of human users, and the emerging need for policymakers to protect against the most likely predatory practices. | ['Louis Rosenberg'] | 2023-06-19 | null | null | null | null | ['misinformation'] | ['miscellaneous'] | [ 4.72515494e-01 8.93958867e-01 -6.49849400e-02 -1.81685343e-01
-3.43276143e-01 -9.84668911e-01 1.29781950e+00 -1.02726541e-01
-3.12254459e-01 5.18581510e-01 6.37007236e-01 -4.55075920e-01
2.17641905e-01 -6.27868712e-01 -2.78235096e-02 -3.92800331e-01
4.03120369e-01 4.95426238e-01 -2.10097544e-02 -7.09037781e-01
4.12426382e-01 1.57310039e-01 -1.17887127e+00 1.86498046e-01
4.52664673e-01 3.52580935e-01 -3.45184863e-01 6.23142123e-01
-8.66443738e-02 1.22906709e+00 -8.79162848e-01 -7.27420866e-01
1.20553039e-01 -4.09307331e-01 -7.10946500e-01 7.33810216e-02
1.72774941e-01 -6.29545391e-01 3.99534889e-02 1.09757388e+00
3.80633771e-01 -1.74734265e-01 1.86586902e-01 -1.46359122e+00
-6.29606843e-01 1.09092081e+00 -4.65947509e-01 1.34339750e-01
6.71674132e-01 8.79580200e-01 7.67499030e-01 -7.58769438e-02
9.41500664e-01 1.78342247e+00 4.70644921e-01 7.71884978e-01
-1.21048605e+00 -7.30087996e-01 1.38374522e-01 -2.61909664e-01
-9.17900383e-01 -5.33753574e-01 9.40742433e-01 -4.37165588e-01
5.29673159e-01 9.18441296e-01 8.72729838e-01 1.82773340e+00
9.18050110e-02 8.03063929e-01 1.08482337e+00 -2.00565055e-01
1.10732757e-01 5.76109052e-01 7.32986769e-03 4.03804272e-01
4.03583050e-01 -1.19457781e-01 -3.23974282e-01 -9.83099818e-01
2.76017457e-01 -5.96696854e-01 -6.93190172e-02 3.62933911e-02
-1.34406948e+00 1.10151660e+00 1.76669523e-01 5.99259973e-01
-7.33182788e-01 2.30384246e-01 5.98162472e-01 1.83307201e-01
5.18244088e-01 1.06687510e+00 1.40530512e-01 -9.20341790e-01
4.43441495e-02 8.39739859e-01 1.20448530e+00 5.12442589e-01
-8.65310282e-02 -6.30290136e-02 1.82989672e-01 5.86086452e-01
6.47292078e-01 6.16403878e-01 1.64322764e-01 -1.28512681e+00
5.42293452e-02 7.82686889e-01 5.08923888e-01 -1.76052248e+00
-2.39031255e-01 2.21409425e-02 -1.09156137e-02 4.75870162e-01
5.17562449e-01 -5.23470581e-01 -3.03349465e-01 1.53530276e+00
7.89143264e-01 -4.68043238e-01 -1.64958403e-01 9.39563692e-01
3.16064537e-01 7.79477417e-01 1.94752336e-01 -2.01655135e-01
1.49808431e+00 -6.98301047e-02 -1.08408368e+00 -3.30416054e-01
5.41142166e-01 -7.10330486e-01 1.24828267e+00 4.58885670e-01
-1.00452614e+00 4.39713180e-01 -1.22099960e+00 2.08662853e-01
-2.56614834e-01 -1.00816631e+00 5.52514911e-01 1.04986799e+00
-3.18336159e-01 2.09311187e-01 -5.14897585e-01 -4.75056559e-01
4.47619140e-01 3.51005048e-02 2.11380586e-01 4.52479124e-01
-1.38829947e+00 1.10332298e+00 -2.77457744e-01 1.59766123e-01
-7.07901537e-01 -5.27087986e-01 -4.56312418e-01 -2.78981537e-01
7.11697936e-01 -5.11775553e-01 1.46433723e+00 -1.39578772e+00
-1.70244622e+00 9.43920255e-01 7.08126843e-01 -4.13629234e-01
8.70530963e-01 -2.56513268e-01 -4.91062075e-01 1.37503287e-02
1.53782526e-02 5.47175050e-01 9.85592365e-01 -1.32564819e+00
-5.10430336e-01 -4.02836084e-01 5.44098794e-01 3.78119528e-01
-3.52116227e-01 7.31355608e-01 7.49951899e-01 -3.56629372e-01
-3.46956015e-01 -1.12163627e+00 -1.20345332e-01 -3.17423604e-02
-8.31009746e-01 -3.05473089e-01 1.27825713e+00 -2.70284146e-01
9.20810223e-01 -2.02979398e+00 -1.77072734e-02 3.32723670e-02
6.09178245e-01 6.44384563e-01 2.62602389e-01 7.15862513e-01
6.38505578e-01 5.90083539e-01 4.56756979e-01 2.65197873e-01
3.46048295e-01 -7.38897622e-02 -2.53438979e-01 4.73801643e-01
-2.26312622e-01 6.97480261e-01 -1.05649137e+00 -1.47355631e-01
-2.05749571e-02 4.61572379e-01 -3.91590148e-02 -1.15344912e-01
-5.93037009e-01 3.71680647e-01 -8.51334572e-01 2.91016847e-01
2.09882289e-01 -1.75459102e-01 4.29752946e-01 3.94110471e-01
-5.22932231e-01 5.96095979e-01 -4.83592808e-01 6.00233555e-01
-3.99397947e-02 9.05627847e-01 7.71533489e-01 3.98801863e-02
6.33205891e-01 3.30500752e-01 7.22675547e-02 -2.42694482e-01
7.97011971e-01 -1.90267757e-01 4.46558535e-01 -5.33537388e-01
4.60403711e-01 -2.74657570e-02 -4.02904123e-01 1.04887617e+00
-1.02730668e+00 -2.80642360e-01 -7.08535016e-01 4.05666769e-01
9.83362734e-01 -6.23540729e-02 1.24443501e-01 -5.03918417e-02
2.86046594e-01 6.11869574e-01 1.79645345e-01 6.30582929e-01
-6.96770072e-01 -3.44053924e-01 6.42105877e-01 -7.57302821e-01
-9.35220182e-01 -5.66632688e-01 4.80798662e-01 1.27258301e+00
2.30777547e-01 -3.20138365e-01 -8.26492608e-01 -5.69400847e-01
1.25369802e-01 1.45607805e+00 -2.85288572e-01 -3.93761218e-01
-4.83698189e-01 -7.71915689e-02 8.37410629e-01 -2.53756553e-01
5.19781113e-01 -1.17446876e+00 -1.17833209e+00 4.17789876e-01
-1.66258976e-01 -8.84516895e-01 -2.92912781e-01 -7.74579942e-01
-1.38045236e-01 -7.52886474e-01 -2.92349935e-01 1.85095444e-02
2.17585027e-01 4.52306032e-01 4.63855594e-01 2.52704881e-02
-3.29429470e-02 5.41146517e-01 -2.99746543e-01 -1.04515254e+00
-1.20054328e+00 -1.75586134e-01 6.18981197e-02 2.62317836e-01
6.83493972e-01 -5.78202963e-01 -4.48471427e-01 5.56703210e-01
-6.65242732e-01 3.00078541e-01 -8.56531113e-02 2.08873808e-01
-5.05212247e-01 -4.07635421e-01 6.14328444e-01 -1.19950092e+00
1.21798825e+00 -6.11991584e-01 -2.25952700e-01 -4.66159545e-02
-5.78355193e-01 -6.35335565e-01 4.02256727e-01 -8.99232984e-01
-1.16762471e+00 -3.98344427e-01 4.17010248e-01 7.62806907e-02
-2.16995955e-01 1.47554547e-01 -5.74596152e-02 -4.90784049e-02
1.11729395e+00 -4.29067284e-01 4.76287663e-01 -1.82651684e-01
5.27725816e-01 1.14104021e+00 -2.37700231e-02 -3.40027183e-01
7.96254396e-01 7.15935946e-01 -6.30486369e-01 -1.19517875e+00
-2.95216143e-01 8.24032128e-02 3.81009072e-01 -1.00510609e+00
8.42399001e-01 -4.46653992e-01 -1.41700709e+00 5.73477805e-01
-1.13195968e+00 -2.75102317e-01 2.44503528e-01 -3.81138921e-02
-1.93373814e-01 2.32728437e-01 -5.00290155e-01 -1.15211940e+00
-4.36815977e-01 -7.80174196e-01 4.71867800e-01 2.45542064e-01
-1.39889503e+00 -5.95897973e-01 -6.77193180e-02 9.66060758e-01
9.21582520e-01 4.20322508e-01 7.49886394e-01 -1.04128647e+00
-4.61476505e-01 -6.29262626e-01 1.66469991e-01 -2.08835048e-03
3.08113366e-01 1.19011007e-01 -8.03663373e-01 1.49956599e-01
2.61870146e-01 -3.50071222e-01 -4.95548010e-01 -8.16061869e-02
6.08231723e-02 -1.09758484e+00 -6.33474231e-01 -5.57095826e-01
3.94740820e-01 7.04605341e-01 4.21831578e-01 2.91844279e-01
6.21972919e-01 9.15074527e-01 7.69771874e-01 4.75720584e-01
4.17629898e-01 6.35523319e-01 4.70134854e-01 4.07900631e-01
2.84978181e-01 -6.79215789e-01 4.91524190e-01 1.82681203e-01
1.33336067e-01 -2.07417548e-01 -8.56514931e-01 7.86447302e-02
-1.94984138e+00 -1.11720467e+00 -2.96264812e-02 1.69527531e+00
5.17245173e-01 6.75593913e-01 4.29336160e-01 -1.26430795e-01
9.72801089e-01 3.95479649e-01 -6.67795360e-01 -9.90019023e-01
3.60156208e-01 -6.34499967e-01 4.50512499e-01 6.36109054e-01
-6.13635838e-01 9.27995205e-01 6.03439856e+00 1.80583999e-01
-9.59198177e-01 2.31700659e-01 7.86073387e-01 -7.09808767e-02
-7.65083611e-01 -4.70471904e-02 -5.43948948e-01 3.76608163e-01
9.03038621e-01 -5.38978875e-01 5.11710584e-01 8.84612799e-01
7.29994118e-01 -3.17480713e-01 -7.38702774e-01 3.92327398e-01
1.17296748e-01 -1.25644028e+00 -4.17243212e-01 2.70506293e-01
2.22111136e-01 -2.77154714e-01 7.68775344e-02 -6.89210519e-02
8.21287155e-01 -8.23246717e-01 5.75406611e-01 1.03189267e-01
1.08465150e-01 -5.27578831e-01 1.26577571e-01 5.56990087e-01
-6.68325201e-02 -3.66447031e-01 1.27293065e-01 -5.33774793e-01
6.86121404e-01 -1.71860550e-02 -1.19879019e+00 -8.04021239e-01
2.70292133e-01 -1.78439673e-02 1.35835662e-01 2.27081701e-01
-1.16363086e-01 7.03218341e-01 -5.01333416e-01 -9.66682494e-01
4.39232171e-01 -1.87775552e-01 1.37395084e+00 5.67223370e-01
-2.69130051e-01 4.15500879e-01 -2.42263138e-01 9.10572290e-01
2.27879643e-01 -2.20982715e-01 -1.15990412e+00 -1.01883960e+00
7.20984042e-01 1.26764905e+00 -5.87899208e-01 -3.69006842e-02
-3.64451855e-03 6.45634353e-01 -3.55933875e-01 -8.38111266e-02
-7.22930670e-01 -1.93130672e-02 9.45860267e-01 4.14565593e-01
-3.17286938e-01 -7.23591968e-02 -2.99872845e-01 -4.05531228e-01
-5.02049267e-01 -1.71856761e+00 -4.61447164e-02 -8.31773221e-01
-1.18568575e+00 3.54916930e-01 -1.15058042e-01 -4.88993943e-01
-4.64035571e-01 -1.16401561e-01 -5.86855054e-01 2.81559825e-01
-2.14480624e-01 -1.07323837e+00 2.33612180e-01 -2.63381880e-02
4.44281280e-01 -5.54024689e-02 7.50068843e-01 -1.32232100e-01
-1.43496543e-01 4.78581935e-01 -6.39714062e-01 -1.64319277e-01
5.24264932e-01 -3.43306303e-01 5.93068361e-01 2.64096439e-01
-1.02523610e-01 6.84454381e-01 1.46636331e+00 -7.79521525e-01
-1.77599835e+00 -1.16604909e-01 7.54713655e-01 -8.64848673e-01
1.07372415e+00 -5.90576351e-01 -4.81075883e-01 9.02465641e-01
3.07568967e-01 -9.48639154e-01 4.93526280e-01 2.08870292e-01
-4.56533223e-01 2.00054392e-01 -1.54831004e+00 1.34622216e+00
1.04922271e+00 -6.23512268e-01 -6.16926372e-01 7.01978564e-01
1.06397808e+00 -6.28375113e-02 -2.81110555e-01 -2.81795144e-01
7.59751678e-01 -7.57177889e-01 5.01429319e-01 -1.00037944e+00
2.00701341e-01 -4.16679457e-02 1.39298469e-01 -1.13978910e+00
-1.31922692e-01 -1.60863793e+00 3.30999821e-01 1.31332910e+00
5.18771708e-01 -9.68268692e-01 8.03022087e-01 1.56249785e+00
2.90751696e-01 -1.15522496e-01 -6.57067776e-01 -1.17333695e-01
-2.75449544e-01 -3.35280985e-01 4.38301742e-01 1.42271972e+00
8.12515140e-01 6.77355587e-01 -6.70189381e-01 1.24201670e-01
3.82038206e-01 -5.32906234e-01 1.04180086e+00 -9.97869313e-01
-1.62836760e-01 -4.17188883e-01 -4.37188148e-01 -6.00032270e-01
-4.09134865e-01 -2.79069722e-01 1.67487953e-02 -8.86917472e-01
-3.78611907e-02 -9.99372452e-02 6.03681564e-01 2.71905929e-01
2.69299209e-01 -2.19501495e-01 4.94748145e-01 3.99461649e-02
-2.93457121e-01 4.29787003e-02 1.41292000e+00 -1.04837976e-02
-6.21022344e-01 1.49425149e-01 -1.39591479e+00 9.63279963e-01
7.39860058e-01 -4.01455790e-01 -5.05112350e-01 3.45634043e-01
9.10148442e-01 4.78841811e-02 3.34000856e-01 -3.40880245e-01
2.15369135e-01 -5.96619725e-01 -3.74476343e-01 2.41562482e-02
4.65851992e-01 -8.75291526e-01 5.60884774e-01 5.86412847e-01
-7.24810302e-01 -6.02733940e-02 -2.06353869e-02 5.94549179e-01
4.83402342e-01 -2.04617102e-02 5.67973852e-01 -3.46219204e-02
7.17652589e-02 -5.20350218e-01 -1.35577047e+00 -1.07109040e-01
1.25018418e+00 -1.73539534e-01 -7.94230700e-01 -1.32264495e+00
-3.23940694e-01 3.47035646e-01 4.51293498e-01 7.61503696e-01
1.65258154e-01 -9.21011150e-01 -6.70209348e-01 -4.01431233e-01
-2.67725065e-02 -4.76384819e-01 -1.84830092e-02 5.02438128e-01
-4.97885197e-01 1.79659817e-02 -1.26798347e-01 -1.56886905e-01
-1.44590807e+00 5.72042704e-01 2.00313210e-01 5.10525465e-01
-9.56135392e-01 7.71618783e-01 4.92139161e-02 -8.60890076e-02
1.48832992e-01 3.89334172e-01 -1.76855419e-02 1.48145214e-01
1.03923059e+00 5.27607560e-01 -7.98174679e-01 -5.96964061e-01
-2.76483417e-01 -5.53127885e-01 -4.34473217e-01 -4.95977968e-01
1.04123783e+00 -2.87947655e-01 -7.43732080e-02 5.93898535e-01
6.66048408e-01 1.97419822e-01 -1.00600708e+00 -4.04588729e-02
-2.61300113e-02 -7.44646192e-01 -2.33307093e-01 -1.20020163e+00
-3.97813052e-01 1.36144832e-01 1.14000916e-01 1.11102068e+00
2.18679681e-01 -3.33489217e-02 9.08928037e-01 3.58906120e-01
3.39258641e-01 -1.12288249e+00 5.97764030e-02 -1.17469624e-01
1.03963315e+00 -8.43337119e-01 3.97572629e-02 -6.61044657e-01
-1.08709097e+00 7.57159710e-01 4.83781844e-01 9.28338394e-02
2.65609831e-01 2.61497557e-01 5.15299797e-01 -6.98202074e-01
-9.52961504e-01 5.27284741e-01 -4.85077828e-01 6.54676437e-01
1.15482859e-01 3.90633464e-01 -6.30839050e-01 1.90498129e-01
-1.90525681e-01 -2.36269623e-01 9.28088129e-01 7.21656442e-01
-6.54869676e-01 -7.66053557e-01 -5.77176750e-01 1.73357412e-01
-5.83624184e-01 5.11342943e-01 -1.35610402e+00 9.43064094e-01
-4.08236027e-01 1.34511471e+00 -9.93171185e-02 -4.31380242e-01
1.85801685e-01 1.64099574e-01 -1.44530386e-01 -2.39447609e-01
-1.01972508e+00 -4.63329554e-02 1.24892879e+00 -5.46454787e-01
-1.30788326e-01 -8.99018109e-01 -9.47779536e-01 -9.24710810e-01
-2.91161686e-01 -1.61847562e-01 7.94247329e-01 8.16044927e-01
4.77382541e-01 -2.81062335e-01 7.70716548e-01 -5.12445509e-01
-8.24779093e-01 -9.54274476e-01 1.25989029e-02 3.01232159e-01
1.08034136e-02 -3.83324951e-01 -5.41681767e-01 -7.26288378e-01] | [9.235128402709961, 6.405200481414795] |
b32da83a-8241-4b32-8fa7-e2cec9e4c8c0 | deepphysics-a-physics-aware-deep-learning | 2109.09491 | null | https://arxiv.org/abs/2109.09491v1 | https://arxiv.org/pdf/2109.09491v1.pdf | DeepPhysics: a physics aware deep learning framework for real-time simulation | Real-time simulation of elastic structures is essential in many applications, from computer-guided surgical interventions to interactive design in mechanical engineering. The Finite Element Method is often used as the numerical method of reference for solving the partial differential equations associated with these problems. Yet, deep learning methods have recently shown that they could represent an alternative strategy to solve physics-based problems 1,2,3. In this paper, we propose a solution to simulate hyper-elastic materials using a data-driven approach, where a neural network is trained to learn the non-linear relationship between boundary conditions and the resulting displacement field. We also introduce a method to guarantee the validity of the solution. In total, we present three contributions: an optimized data set generation algorithm based on modal analysis, a physics-informed loss function, and a Hybrid Newton-Raphson algorithm. The method is applied to two benchmarks: a cantilever beam and a propeller. The results show that our network architecture trained with a limited amount of data can predict the displacement field in less than a millisecond. The predictions on various geometries, topologies, mesh resolutions, and boundary conditions are accurate to a few micrometers for non-linear deformations of several centimeters of amplitude. | ['Stéphane Cotin', 'Ryadh Haferssas', 'Alban Odot'] | 2021-09-17 | null | null | null | null | ['cantilever-beam'] | ['miscellaneous'] | [ 1.30425125e-01 3.64830464e-01 1.81876585e-01 -1.52986020e-01
-5.99027574e-01 -1.48636401e-01 1.19026765e-01 1.00482695e-01
-4.63166296e-01 8.87308896e-01 -3.83461803e-01 -8.28431547e-02
-3.95753890e-01 -1.01589978e+00 -1.01492882e+00 -8.86103332e-01
-1.43611282e-01 8.11724961e-01 2.23400861e-01 -5.12117445e-01
1.59108952e-01 7.15028703e-01 -1.26793921e+00 2.69731171e-02
7.29297698e-01 1.24124646e+00 -5.41007742e-02 4.51453358e-01
2.34790280e-01 2.11587980e-01 -8.75290260e-02 3.07999942e-02
1.53167903e-01 -2.59868771e-01 -1.02687800e+00 -3.53580028e-01
7.48159038e-03 -4.10256803e-01 -2.36854404e-02 6.06621087e-01
8.67938042e-01 2.62164593e-01 8.41289461e-01 -8.18185329e-01
-1.98914573e-01 6.26401961e-01 -2.43793771e-01 -2.61658311e-01
2.02359915e-01 1.44759595e-01 5.87123334e-01 -8.80429506e-01
8.33215117e-01 7.59727716e-01 1.09890389e+00 8.43194723e-01
-1.43169177e+00 -3.25783670e-01 -4.83884931e-01 -1.01037137e-02
-1.18918598e+00 7.81390369e-02 1.15464616e+00 -7.54005730e-01
8.20286632e-01 2.62285441e-01 8.35302234e-01 1.00640905e+00
9.92268741e-01 5.44971153e-02 8.02882135e-01 -2.65796363e-01
4.38622147e-01 -8.76309648e-02 -2.38747850e-01 7.00926244e-01
5.60235530e-02 3.99423808e-01 -1.92020103e-01 -3.48791242e-01
9.03413475e-01 -3.67336571e-01 -2.40614370e-01 -3.59368294e-01
-9.86689091e-01 8.17502856e-01 4.94085282e-01 2.98623502e-01
-3.69217604e-01 3.78216743e-01 4.54938620e-01 3.28228772e-02
4.87415165e-01 8.21669400e-01 -5.43333709e-01 6.39949962e-02
-7.71726191e-01 6.51757598e-01 9.16944623e-01 3.12472582e-01
3.90419900e-01 8.75060260e-02 2.51462698e-01 3.50629240e-01
3.50136399e-01 5.10016024e-01 3.49813491e-01 -9.85666215e-01
-1.03803255e-01 3.12313646e-01 2.25118682e-01 -1.19776118e+00
-8.75405788e-01 -1.66939020e-01 -1.07576501e+00 6.52085841e-01
3.70897681e-01 -4.63341117e-01 -7.63413906e-01 1.59682524e+00
6.96693718e-01 2.19642922e-01 -2.07965329e-01 1.13275099e+00
1.00738573e+00 4.90456998e-01 -2.42412597e-01 -2.59227902e-01
9.13773179e-01 -3.83057714e-01 -4.78672326e-01 9.67367217e-02
5.12984693e-01 -6.01151884e-01 8.85305703e-01 3.64880800e-01
-1.66314423e+00 -3.84081453e-01 -8.92127454e-01 1.27247378e-01
-1.28155410e-01 -2.05036893e-01 1.68039560e-01 7.51448572e-02
-9.84904647e-01 1.44856632e+00 -1.06911504e+00 1.60296425e-01
2.52057344e-01 5.85199654e-01 -1.09229766e-01 6.13235474e-01
-1.29894280e+00 9.81111526e-01 1.17988549e-01 4.71480817e-01
-3.78999978e-01 -1.13165784e+00 -5.24348557e-01 1.85466215e-01
5.79364337e-02 -1.02008760e+00 9.24674153e-01 -5.25406778e-01
-2.01754856e+00 9.55265105e-01 3.52388710e-01 -3.74863893e-01
7.45985389e-01 7.97019748e-04 1.25421107e-01 -1.21920863e-02
-4.09678131e-01 2.22668663e-01 5.48240125e-01 -1.31774855e+00
2.32476950e-01 1.15612745e-01 -1.15320861e-01 -2.46640012e-01
-8.18647891e-02 -3.73344243e-01 2.37582430e-01 -7.21619248e-01
3.20098162e-01 -1.16652441e+00 -6.04712129e-01 9.61348712e-02
-6.22964919e-01 1.23768553e-01 4.76418525e-01 -4.91810739e-01
7.94781625e-01 -1.78268182e+00 6.75266802e-01 5.86765707e-01
1.76400170e-01 -1.21699749e-02 2.87252665e-01 6.58870518e-01
-6.09652512e-02 -4.92537692e-02 -7.58760214e-01 -8.80906507e-02
-1.62865222e-01 -8.05962235e-02 -2.95978516e-01 4.72441643e-01
8.00717995e-02 8.42191756e-01 -5.41125953e-01 -3.83325607e-01
-2.26009991e-02 6.14959776e-01 -7.02465951e-01 1.65593207e-01
-5.00159025e-01 7.72651374e-01 -4.11215812e-01 1.71467006e-01
7.13452816e-01 -3.11889291e-01 1.96300432e-01 -5.81530869e-01
-1.77187726e-01 -7.12748468e-02 -1.12134612e+00 1.61209381e+00
-4.68006492e-01 3.56913179e-01 4.77053016e-01 -1.25467396e+00
9.21701968e-01 4.34136420e-01 1.04738772e+00 -4.08711731e-01
6.96537435e-01 5.56364715e-01 1.29262403e-01 -5.95681787e-01
1.94896795e-02 -5.84444106e-01 -1.30174339e-01 5.27741492e-01
-2.62926251e-01 -6.51945770e-01 -1.01425946e-01 -2.01966390e-01
9.43517148e-01 1.64710253e-01 -2.28250727e-01 -6.02929711e-01
5.01280904e-01 2.35570446e-01 5.82781553e-01 2.62436181e-01
3.49983513e-01 6.49765253e-01 3.57471138e-01 -9.07696187e-01
-1.27744043e+00 -8.29279184e-01 -4.75073427e-01 4.14991230e-01
2.73523182e-01 1.88186243e-01 -8.74917507e-01 9.40087140e-02
1.92643955e-01 2.83483416e-01 -5.69449008e-01 -4.27027076e-01
-1.22916842e+00 -7.19391465e-01 1.42476574e-01 6.19453013e-01
8.15761313e-02 -1.36013770e+00 -8.89870167e-01 4.95768607e-01
6.37669116e-02 -1.04965293e+00 9.34732929e-02 1.40520439e-01
-9.37955022e-01 -9.53552485e-01 -4.68669921e-01 -8.52645576e-01
6.91136420e-01 -7.49411225e-01 1.10377574e+00 4.09269482e-01
-6.53748691e-01 1.41338199e-01 1.06385924e-01 -2.46495515e-01
-8.43743801e-01 1.22874416e-02 2.34524667e-01 -3.03492874e-01
-6.26030326e-01 -9.31401551e-01 -8.23936045e-01 2.25132227e-01
-8.54394078e-01 1.46110922e-01 3.69857758e-01 6.83593035e-01
8.99332106e-01 -7.24578276e-02 7.00365365e-01 -8.96910250e-01
4.93207484e-01 -3.65063131e-01 -6.48597896e-01 -1.87759638e-01
-3.54419291e-01 1.29125893e-01 9.97262001e-01 -6.65141046e-01
-7.54522026e-01 3.02855790e-01 -6.45251632e-01 -4.00210381e-01
3.26833576e-02 8.16827714e-01 2.04243943e-01 -5.13499558e-01
7.26863027e-01 -2.44294722e-02 3.46231818e-01 -2.99613506e-01
-3.58285122e-02 -1.00363359e-01 4.22346771e-01 -8.58179688e-01
7.98148870e-01 5.82307935e-01 8.03614974e-01 -6.17173254e-01
-4.60986495e-01 1.78708985e-01 -5.08047938e-01 -4.51794237e-01
7.40672648e-01 -9.79082361e-02 -1.32897699e+00 6.32064819e-01
-1.13282096e+00 -7.60016739e-01 -3.42415482e-01 4.70698655e-01
-8.82415891e-01 1.60852790e-01 -8.69048119e-01 -4.17262852e-01
-6.78295732e-01 -1.25705850e+00 9.80977893e-01 5.27659431e-02
-2.44264677e-01 -1.11020696e+00 2.35719815e-01 -7.97068998e-02
7.09703624e-01 1.24265063e+00 1.10009229e+00 -1.80411577e-01
-3.95200968e-01 -3.01667631e-01 2.90333986e-01 6.61912188e-02
-1.21246725e-01 3.49267721e-01 -7.06113517e-01 -4.37736779e-01
3.11954230e-01 -3.65547299e-01 5.74942708e-01 8.46106052e-01
1.40374506e+00 -2.50862360e-01 -6.05955720e-01 6.39347255e-01
1.64773750e+00 -1.41961202e-02 3.28822911e-01 -1.54655576e-01
5.66295266e-01 4.84150350e-01 2.04583496e-01 5.89333892e-01
-1.14674233e-01 7.97680140e-01 6.88990653e-01 -1.08105324e-01
2.49573752e-01 3.91193748e-01 -2.24329650e-01 8.77720892e-01
-3.16841304e-01 -1.81829054e-02 -1.15243256e+00 1.71265647e-01
-1.65571833e+00 -7.62960196e-01 -2.62910217e-01 2.10904431e+00
1.02425313e+00 3.88320565e-01 -5.14693223e-02 1.58450767e-01
3.76145154e-01 -3.68061036e-01 -8.05143237e-01 -4.33638245e-01
1.63812131e-01 5.94650865e-01 3.16924304e-01 6.74139678e-01
-8.36751521e-01 3.96509290e-01 6.45809412e+00 4.42233384e-01
-1.77698624e+00 -2.45728806e-01 5.07137418e-01 7.13716596e-02
-4.19485569e-01 -4.46064949e-01 -4.67471242e-01 6.73484325e-01
1.05234623e+00 -2.25344837e-01 2.76789546e-01 6.86823964e-01
3.20486993e-01 -3.34823243e-02 -1.13473511e+00 5.03564954e-01
-3.70557189e-01 -1.98005307e+00 -3.79248649e-01 5.40771335e-02
6.72862828e-01 -1.49663985e-01 5.88437021e-02 -1.87736735e-01
-1.91640407e-01 -1.15025926e+00 7.50647843e-01 9.63121474e-01
7.99531281e-01 -6.72338367e-01 6.36174679e-01 5.74358702e-01
-1.04606938e+00 1.12003736e-01 -2.38872930e-01 7.76148960e-03
6.09666884e-01 8.56532454e-01 -6.23242617e-01 3.73871356e-01
7.11769640e-01 2.65246004e-01 3.31368983e-01 8.51645768e-01
4.57395226e-01 4.33538049e-01 -5.46680093e-01 -1.12956256e-01
-7.56518170e-02 -2.54383326e-01 7.01897502e-01 6.67922497e-01
4.40432757e-01 4.74063843e-01 1.30345851e-01 1.16598904e+00
2.08495697e-03 5.64731024e-02 -3.56901526e-01 3.47460687e-01
2.55812347e-01 1.15468514e+00 -8.58923078e-01 1.46874711e-01
1.96013749e-01 3.39827299e-01 1.78520948e-01 1.38488740e-01
-1.07084274e+00 -3.17785054e-01 4.04052913e-01 6.54130518e-01
3.04477420e-02 -3.72745097e-01 -4.94076043e-01 -5.79981089e-01
1.03869520e-01 -3.27657491e-01 -1.15999961e-02 -6.13386154e-01
-1.23091650e+00 4.95328188e-01 7.19165653e-02 -1.10109258e+00
-4.29999590e-01 -9.79801357e-01 -8.34538758e-01 8.42033446e-01
-1.20004630e+00 -7.41855383e-01 -2.89706230e-01 1.91338480e-01
-5.43166585e-02 1.32213667e-01 9.09025311e-01 3.70032459e-01
-2.96530187e-01 1.79280117e-01 1.61531180e-01 1.77905351e-01
3.62089217e-01 -1.01607001e+00 2.44046926e-01 2.31733993e-01
-7.50084400e-01 2.81964183e-01 1.21674776e+00 -6.66420996e-01
-1.62536395e+00 -6.44011199e-01 5.20637691e-01 2.13907845e-02
5.14661014e-01 -1.53084591e-01 -1.31324017e+00 1.52885303e-01
-1.00280687e-01 4.56471205e-01 4.18733507e-01 -4.02820379e-01
5.12637794e-01 -3.60938348e-02 -1.40899253e+00 3.70122463e-01
8.14571202e-01 3.66519131e-02 -6.16124310e-02 5.33079922e-01
6.12028301e-01 -1.10796833e+00 -1.20943379e+00 9.26135957e-01
6.35595858e-01 -8.27306211e-01 1.10728765e+00 -6.53204679e-01
1.06076670e+00 1.97780594e-01 4.07594562e-01 -1.22413206e+00
-1.92784280e-01 -7.82443523e-01 -1.09312668e-01 4.75575954e-01
3.36861670e-01 -6.73964918e-01 9.47347522e-01 8.36993277e-01
-4.76583332e-01 -1.64288700e+00 -1.14947629e+00 -5.91607690e-01
6.61201775e-01 -1.69369847e-01 3.54146004e-01 7.50866890e-01
-4.39008139e-02 -1.76838800e-01 -7.19593000e-03 -1.60188496e-03
5.21764874e-01 3.12157184e-01 3.05811435e-01 -1.52361584e+00
-2.19254464e-01 -5.54129720e-01 -1.98858336e-01 -5.93127251e-01
4.77543086e-01 -6.13558233e-01 3.88291597e-01 -1.51883352e+00
-2.98558682e-01 -7.53828585e-01 1.15171403e-01 2.74006999e-03
2.23536730e-01 5.74605614e-02 -3.60957325e-01 1.52466372e-02
4.11088109e-01 4.75470454e-01 1.48688900e+00 1.42657082e-03
-2.55990177e-01 1.26012623e-01 -1.09573165e-02 8.59340906e-01
7.58652985e-01 -4.12185490e-01 -1.09556682e-01 -3.44053298e-01
7.33259499e-01 3.84257227e-01 5.01560926e-01 -1.13579154e+00
3.02209258e-01 -4.02954161e-01 1.06331334e-01 -2.93366581e-01
4.14367110e-01 -9.22387183e-01 3.80121320e-01 7.55259037e-01
-4.21943992e-01 -6.48088232e-02 3.57565373e-01 2.12235183e-01
-1.25516970e-02 -2.34216973e-01 1.23890686e+00 1.36297895e-04
4.58230078e-02 4.06883985e-01 -3.55597466e-01 1.12748168e-01
9.56598043e-01 -1.08416565e-01 4.25916053e-02 -4.65074852e-02
-1.01164973e+00 -2.43866779e-02 3.48905921e-01 -2.52313018e-01
6.98046565e-01 -1.39313614e+00 -7.75170922e-01 3.57397735e-01
-5.87124825e-01 4.35084075e-01 4.94765759e-01 8.13514471e-01
-1.06693769e+00 3.14839445e-02 -2.40206748e-01 -6.79178894e-01
-9.69187975e-01 3.07411522e-01 7.24572837e-01 -2.63867676e-01
-5.68551421e-01 1.01041234e+00 -2.89002359e-01 -4.41236436e-01
-1.30552769e-01 -7.01287508e-01 2.14145146e-02 -3.30924869e-01
-5.12190610e-02 3.83457899e-01 3.74781281e-01 -6.39280975e-01
-3.25593233e-01 9.42692816e-01 5.18808842e-01 1.81613714e-02
1.57636333e+00 3.08581769e-01 -1.55863538e-01 2.31730729e-01
1.13977134e+00 -1.76831275e-01 -1.19796824e+00 -2.78659910e-02
-2.90701628e-01 -1.19474065e-03 -1.75720584e-02 -5.26330471e-01
-1.27660286e+00 7.58420467e-01 2.80434012e-01 4.95390922e-01
1.00013137e+00 -2.49209590e-02 1.32502031e+00 3.89741600e-01
1.77973166e-01 -1.13932168e+00 1.67326331e-01 4.82021213e-01
1.14579809e+00 -1.00124085e+00 1.73540115e-01 -6.43767178e-01
1.43464226e-02 1.35511744e+00 6.04826987e-01 -4.13388461e-01
1.24041748e+00 7.86886334e-01 -4.37660888e-02 -1.33982643e-01
-4.77318704e-01 4.89603251e-01 5.02143979e-01 9.61493775e-02
5.20929217e-01 -1.70127004e-01 -4.79804873e-01 5.73288083e-01
-3.40677083e-01 1.06254376e-01 4.39805180e-01 1.09643185e+00
-4.21430081e-01 -1.01571524e+00 -2.64921278e-01 3.40873301e-01
-5.07080913e-01 3.04540187e-01 -1.07919417e-01 6.43297255e-01
-2.07531056e-03 2.22162306e-01 4.78209257e-02 -2.60865301e-01
4.45828259e-01 -4.96451370e-02 5.79413056e-01 -4.09604639e-01
-5.98391891e-01 -1.42904976e-02 -2.03763902e-01 -6.52075529e-01
-3.72011781e-01 -2.95996815e-01 -1.89810848e+00 -4.85915631e-01
-1.58686697e-01 3.38409171e-02 6.61737561e-01 1.00481737e+00
4.54805762e-01 6.51907861e-01 4.98063505e-01 -1.49111438e+00
-5.68365157e-01 -5.33654928e-01 -4.52086210e-01 4.98173982e-01
2.58679211e-01 -8.60846519e-01 -4.09167767e-01 4.82679717e-02] | [6.366211891174316, 3.379960775375366] |
7b5950f4-37ba-4546-a93e-66b0652caaf4 | speeding-up-one-vs-all-training-for-extreme | 2109.13122 | null | https://arxiv.org/abs/2109.13122v1 | https://arxiv.org/pdf/2109.13122v1.pdf | Speeding-up One-vs-All Training for Extreme Classification via Smart Initialization | In this paper we show that a simple, data dependent way of setting the initial vector can be used to substantially speed up the training of linear one-versus-all (OVA) classifiers in extreme multi-label classification (XMC). We discuss the problem of choosing the initial weights from the perspective of three goals. We want to start in a region of weight space a) with low loss value, b) that is favourable for second-order optimization, and c) where the conjugate-gradient (CG) calculations can be performed quickly. For margin losses, such an initialization is achieved by selecting the initial vector such that it separates the mean of all positive (relevant for a label) instances from the mean of all negatives -- two quantities that can be calculated quickly for the highly imbalanced binary problems occurring in XMC. We demonstrate a speedup of $\approx 3\times$ for training with squared hinge loss on a variety of XMC datasets. This comes in part from the reduced number of iterations that need to be performed due to starting closer to the solution, and in part from an implicit negative mining effect that allows to ignore easy negatives in the CG step. Because of the convex nature of the optimization problem, the speedup is achieved without any degradation in classification accuracy. | ['Rohit Babbar', 'Erik Schultheis'] | 2021-09-27 | null | null | null | null | ['extreme-multi-label-classification'] | ['methodology'] | [ 3.48370701e-01 1.66947648e-01 -1.46658465e-01 -7.27221131e-01
-9.99881148e-01 -7.00605214e-02 8.83709788e-02 7.60076284e-01
-8.90713573e-01 7.85841525e-01 -2.65066773e-01 -3.63988549e-01
-4.20677304e-01 -4.97676402e-01 -3.65258723e-01 -1.01512468e+00
-2.48020515e-01 6.96111858e-01 -3.99054550e-02 -2.20577180e-01
3.50507408e-01 5.30410290e-01 -1.51204717e+00 2.76301533e-01
6.76539242e-01 1.23281801e+00 -2.10450977e-01 6.20540857e-01
-1.06716799e-02 6.90843344e-01 -3.86614323e-01 -5.38804770e-01
4.40095425e-01 -4.40299183e-01 -8.10516953e-01 9.21051279e-02
2.54784286e-01 2.26606000e-02 5.65052211e-01 8.68377864e-01
6.58968866e-01 1.25691175e-01 1.00301194e+00 -1.41043782e+00
2.58280516e-01 3.35235298e-01 -1.00538695e+00 -3.21161449e-02
1.95911452e-02 -2.55337241e-03 1.33285093e+00 -6.33000851e-01
4.41279203e-01 9.30835128e-01 9.62614357e-01 4.01225328e-01
-1.41280031e+00 -4.29523259e-01 6.24457672e-02 -6.95285127e-02
-1.29727590e+00 -3.97414863e-01 6.65339231e-01 -4.17704910e-01
9.09215271e-01 6.58911467e-01 4.77756649e-01 3.41506213e-01
5.42338751e-02 5.66153109e-01 9.82063949e-01 -7.45125532e-01
4.12838072e-01 4.56724584e-01 2.64501989e-01 6.01881623e-01
1.98659495e-01 -1.69118062e-01 -1.00312494e-01 -4.42861229e-01
-1.16025515e-01 -1.82169497e-01 -1.38177648e-01 -5.88612020e-01
-6.31605029e-01 1.12052953e+00 3.60528916e-01 6.29842281e-02
-3.69225144e-01 1.40646294e-01 6.40720427e-01 5.03887653e-01
6.03701830e-01 5.41313469e-01 -6.00971758e-01 -7.30466470e-02
-1.09755850e+00 3.09894711e-01 9.29020286e-01 4.68690872e-01
7.02758312e-01 -3.86498421e-01 1.86525866e-01 1.07985997e+00
1.69261962e-01 2.79580772e-01 5.01457512e-01 -7.58701682e-01
6.64719999e-01 6.45351887e-01 1.39361052e-02 -7.59907544e-01
-5.43046355e-01 -5.71067870e-01 -6.39671147e-01 5.74512005e-01
6.79467738e-01 -3.03573638e-01 -4.65257525e-01 1.71738207e+00
5.16273558e-01 -6.35059774e-01 -2.02522725e-02 6.23424768e-01
1.24185048e-01 3.90165687e-01 1.62880749e-01 -5.20140290e-01
1.33877957e+00 -4.99323398e-01 -2.69221604e-01 -4.22337443e-01
1.27214861e+00 -7.03637004e-01 8.73916864e-01 3.90218258e-01
-1.01634717e+00 7.13369169e-04 -1.21126914e+00 -1.44051416e-02
-4.20868695e-01 -1.36533799e-02 4.83629227e-01 7.21643627e-01
-7.81251371e-01 8.26837122e-01 -4.33359861e-01 2.27842554e-02
4.38039750e-01 7.23131001e-01 -3.97649199e-01 1.49850212e-02
-9.39297199e-01 1.12745142e+00 5.06698370e-01 1.16677232e-01
9.89786908e-02 -7.32233822e-01 -6.28406703e-01 1.01560876e-01
1.94167718e-01 -2.55759597e-01 8.55548322e-01 -1.30286777e+00
-9.97362256e-01 1.09488356e+00 7.03342184e-02 -3.70836794e-01
9.42600906e-01 1.70397207e-01 1.15341634e-01 -4.49960306e-02
-1.44403026e-01 7.02605665e-01 6.81438148e-01 -1.12638915e+00
-8.25816154e-01 -6.13349855e-01 -2.47105747e-01 4.17769939e-01
-3.22595507e-01 9.25500244e-02 -8.93820524e-02 -2.86924660e-01
1.20465115e-01 -1.00306141e+00 -3.47739369e-01 -1.24486478e-03
-1.63236067e-01 -2.75898963e-01 3.98615420e-01 -6.09913647e-01
1.20220101e+00 -2.17064333e+00 -8.63728449e-02 5.77971816e-01
1.61998734e-01 7.42190033e-02 1.51967093e-01 3.02219093e-01
-5.26190162e-01 1.24213435e-02 -2.30644867e-01 -5.25743186e-01
-3.87241296e-03 4.74573188e-02 -4.08348404e-02 7.99471378e-01
1.68380111e-01 5.23315847e-01 -6.38631999e-01 -7.11407244e-01
-1.69582441e-01 1.55852318e-01 -6.00830197e-01 -5.55414474e-03
-1.75688919e-02 -2.40200758e-01 -1.62210941e-01 2.63853312e-01
6.24467492e-01 -2.77125537e-01 2.51091778e-01 8.27173218e-02
1.28220960e-01 -3.32073644e-02 -1.52690709e+00 8.95578086e-01
-4.20311302e-01 4.09438491e-01 1.61329612e-01 -1.36909449e+00
7.88154781e-01 1.68994039e-01 7.17233717e-01 -4.60690200e-01
2.71447986e-01 5.16319573e-01 1.24616036e-03 -1.79443210e-01
1.55276105e-01 -6.44346535e-01 6.44912124e-02 5.87719917e-01
-2.79308826e-01 -3.33535187e-02 2.66283482e-01 -1.35719761e-01
7.44977593e-01 -2.55738974e-01 3.88314545e-01 -4.86409038e-01
4.11930621e-01 1.36475975e-03 6.53005600e-01 3.82487118e-01
-2.08207130e-01 4.82946992e-01 9.14071977e-01 -4.60954964e-01
-1.00376904e+00 -5.09586513e-01 -4.71862733e-01 1.13711762e+00
-2.33147487e-01 -9.56028402e-02 -6.22702360e-01 -7.13248432e-01
2.63685197e-01 7.54144967e-01 -5.92418015e-01 -1.70193434e-01
-7.03131020e-01 -1.30482912e+00 1.21046498e-01 3.22879165e-01
2.40297094e-02 -8.03464234e-01 -7.12706625e-01 1.36085868e-01
-4.83347960e-02 -4.78531361e-01 -3.27874720e-01 1.03119540e+00
-1.08212852e+00 -1.08667946e+00 -5.21413267e-01 -6.49502933e-01
9.99296546e-01 -1.24693699e-01 1.08236361e+00 1.67160600e-01
-2.89487213e-01 1.67212058e-02 -7.91134760e-02 -5.48441589e-01
-3.18197638e-01 1.09113455e-02 -1.64108381e-01 -6.07890263e-02
3.65403652e-01 -3.43628973e-01 -4.30989116e-01 2.51750499e-01
-8.36608887e-01 -3.02665327e-02 4.88719910e-01 9.62804973e-01
4.77762967e-01 2.83672184e-01 6.44191384e-01 -1.03477609e+00
4.63111460e-01 -4.86969531e-01 -6.01089835e-01 3.95018667e-01
-1.07363391e+00 3.09947759e-01 6.33110464e-01 -4.06975508e-01
-5.45359612e-01 2.10477009e-01 -2.98462451e-01 -2.66245119e-02
2.07609907e-01 1.67910069e-01 3.64807770e-02 -1.88986987e-01
7.75917232e-01 -3.46674770e-01 3.61927986e-01 -3.19110960e-01
1.85363024e-01 8.93689871e-01 -5.63734025e-02 -5.31690538e-01
3.37502450e-01 2.57998496e-01 4.05177712e-01 -5.96958160e-01
-8.69609177e-01 -6.99925303e-01 -5.34694791e-01 -1.78013191e-01
5.39448321e-01 -6.21298015e-01 -8.51937890e-01 1.63818419e-01
-5.95460474e-01 -4.27533597e-01 -4.92748260e-01 3.35292608e-01
-6.86556041e-01 1.34646595e-01 -4.66377765e-01 -1.00265026e+00
-4.20196712e-01 -1.05581331e+00 8.73514831e-01 -6.06040061e-02
-3.49644572e-01 -1.16421747e+00 -1.54749185e-01 3.55757177e-01
1.45232782e-01 4.62907583e-01 1.27419984e+00 -8.23069572e-01
2.40406990e-02 -6.50174022e-01 -2.20896099e-02 5.76449275e-01
-1.66494519e-01 -1.75426558e-01 -8.71529877e-01 -6.12271190e-01
2.70329505e-01 -5.06032526e-01 7.40185022e-01 2.93826491e-01
9.88650441e-01 -2.22614318e-01 -4.07924861e-01 5.78485370e-01
1.67935157e+00 4.90943491e-02 4.19968158e-01 4.52246875e-01
3.54011059e-01 7.92058825e-01 7.01393545e-01 3.29647988e-01
8.74796435e-02 7.01253355e-01 2.98546433e-01 -4.46800530e-01
3.20461154e-01 2.62850851e-01 -1.41645521e-01 4.43159521e-01
3.06319714e-01 -2.55005043e-02 -8.92914414e-01 4.80518907e-01
-1.80391335e+00 -6.99294984e-01 -1.79581493e-01 2.48098588e+00
1.13260651e+00 4.35091764e-01 3.45480651e-01 5.53095222e-01
6.35383725e-01 -1.40616760e-01 -4.42657590e-01 -8.03446472e-01
1.18228689e-01 1.20464265e-01 7.44869053e-01 7.27803886e-01
-1.12744904e+00 1.80027872e-01 5.90476704e+00 9.75620985e-01
-1.08925450e+00 -7.76305348e-02 1.30643690e+00 -3.97801965e-01
1.03916526e-02 -5.76138683e-02 -9.96155500e-01 5.65116346e-01
1.04866135e+00 2.49713987e-01 1.93119079e-01 9.70605910e-01
-7.90069699e-02 -3.87519717e-01 -1.27016866e+00 9.07505035e-01
-7.68709555e-02 -8.10685158e-01 -4.89099324e-01 1.90694615e-01
6.47141039e-01 -2.47367829e-01 5.84070059e-03 2.46057510e-01
1.01751916e-01 -8.87622118e-01 6.98436677e-01 1.31129593e-01
7.66718268e-01 -1.09950101e+00 8.50572228e-01 5.81697941e-01
-6.06275082e-01 -4.30559963e-01 -3.90496939e-01 -2.26483983e-03
-1.65817857e-01 1.18197429e+00 -7.78298974e-01 2.04040855e-01
3.44920784e-01 -5.50610460e-02 -4.12964731e-01 9.77778852e-01
2.76498765e-01 3.54520798e-01 -7.08908617e-01 -2.69858778e-01
2.21185386e-01 -2.59451240e-01 1.23157308e-01 1.28566492e+00
7.97721744e-03 -9.98324454e-02 -6.88472018e-02 3.61086428e-01
5.61102219e-02 5.96360564e-01 -3.58634084e-01 5.27197540e-01
5.34124784e-02 1.23313868e+00 -9.30561483e-01 -1.50327861e-01
-2.22936288e-01 7.26557672e-01 6.37891412e-01 -6.89517427e-03
-6.46433055e-01 -6.41116202e-01 3.37998807e-01 1.87822863e-01
4.42588985e-01 2.37459168e-01 -6.26617908e-01 -7.39717007e-01
3.17621410e-01 -8.43293905e-01 6.40479386e-01 -2.31669232e-01
-1.15894830e+00 3.92324179e-01 -1.22255394e-02 -1.03667271e+00
-4.54675496e-01 -5.69854617e-01 -4.10378188e-01 1.05547559e+00
-1.43927491e+00 -5.02549052e-01 1.85852692e-01 6.90434426e-02
3.19786482e-02 2.68265963e-01 7.37989902e-01 4.74017590e-01
-4.21879530e-01 8.67301166e-01 3.57666731e-01 -1.22047789e-01
7.52448142e-01 -1.47810090e+00 -1.68580800e-01 4.09490436e-01
-7.48469681e-02 3.73435020e-01 7.65076995e-01 -2.99438834e-01
-7.41939068e-01 -7.90295362e-01 1.30357897e+00 -3.73839647e-01
3.46784800e-01 -3.56484294e-01 -6.47839308e-01 2.81774163e-01
-4.11630034e-01 1.60578433e-02 8.78412366e-01 3.37319791e-01
-1.00541681e-01 -4.55427378e-01 -1.49670124e+00 3.47941995e-01
4.21821117e-01 -1.23796515e-01 -1.70660228e-01 6.22335672e-01
7.02731311e-02 -2.68191695e-01 -9.37693059e-01 3.05087388e-01
5.55085063e-01 -9.31332767e-01 9.49465752e-01 -7.25425661e-01
3.57054621e-01 4.92472714e-03 -4.52123471e-02 -1.15335953e+00
-1.56807452e-01 -3.42011750e-01 1.93632185e-01 1.07641578e+00
7.46223629e-01 -6.95340514e-01 8.03535104e-01 9.50521410e-01
4.45153832e-01 -1.39203870e+00 -9.25595522e-01 -4.98153389e-01
2.61447459e-01 -3.52482557e-01 2.11757168e-01 8.64381075e-01
5.48351854e-02 2.98430204e-01 -1.31425440e-01 -3.02796096e-01
6.37412250e-01 2.04719156e-01 3.22854340e-01 -1.40898943e+00
-4.85421658e-01 -4.22314197e-01 -4.88140076e-01 -6.82059467e-01
-1.62921876e-01 -9.22825217e-01 6.97607398e-02 -9.86446083e-01
2.20092177e-01 -8.94964278e-01 -2.71961719e-01 5.55608869e-01
-4.73373413e-01 2.71002948e-01 2.58926153e-01 2.61493176e-01
-4.97705400e-01 1.23551199e-02 7.27394581e-01 1.43349305e-01
-1.83716446e-01 2.63508409e-01 -8.75573337e-01 5.79028726e-01
7.73645759e-01 -7.00453401e-01 -3.34947586e-01 4.77722026e-02
6.25269711e-01 1.14555731e-01 -1.11343578e-01 -6.55742407e-01
-3.87879461e-02 -1.02284163e-01 6.18767262e-01 -2.55940616e-01
2.15102941e-01 -8.87030125e-01 -6.76042885e-02 6.16429508e-01
-6.85495317e-01 8.74199420e-02 -3.21957804e-02 2.50647277e-01
1.32411823e-01 -8.95996273e-01 1.25459623e+00 -2.78219767e-02
-7.34025910e-02 -2.05994219e-01 4.75435033e-02 1.87571123e-01
1.16207337e+00 -1.38575003e-01 5.19912392e-02 -2.70568043e-01
-6.80686533e-01 3.96257222e-01 4.71468657e-01 -1.49611473e-01
1.07140094e-01 -1.25336730e+00 -6.27845705e-01 2.17668250e-01
6.61409646e-02 -3.52886394e-02 -2.10649624e-01 1.12703705e+00
-5.84261835e-01 2.66189337e-01 4.81805205e-02 -4.09667134e-01
-1.47802544e+00 5.17397523e-01 5.40163219e-01 -5.91268122e-01
-5.09170413e-01 9.86558378e-01 -1.89942211e-01 -2.79343516e-01
4.73526001e-01 -1.43505827e-01 -1.20490557e-02 3.58532339e-01
3.95726115e-01 5.44792175e-01 5.26432931e-01 -3.93653274e-01
-3.83017242e-01 4.00021493e-01 -2.57641584e-01 -2.31877174e-02
1.50642169e+00 1.02289326e-01 -1.97033256e-01 4.11936194e-01
1.70649993e+00 -9.35583934e-02 -1.26397467e+00 1.13537721e-01
2.12041780e-01 -3.32013279e-01 6.06944077e-02 -7.49360085e-01
-9.76425171e-01 7.70273864e-01 7.25082278e-01 3.40748996e-01
1.11520743e+00 -2.80698597e-01 5.11762500e-01 3.34438413e-01
1.37170359e-01 -1.51255786e+00 -3.30728799e-01 7.36677274e-02
5.93066096e-01 -1.32591152e+00 4.34065193e-01 -1.72752082e-01
-6.47651017e-01 1.14480710e+00 2.77332097e-01 5.23812063e-02
7.10023701e-01 3.34345490e-01 8.58743489e-02 -1.93100512e-01
-7.13661015e-01 5.02276383e-02 8.73825476e-02 2.09626913e-01
3.39560121e-01 6.71090111e-02 -6.69119358e-01 3.41954306e-02
-2.15690970e-01 -3.00471097e-01 1.86396137e-01 9.30813313e-01
-5.27449012e-01 -1.19841325e+00 -3.23385566e-01 9.19220984e-01
-6.90782189e-01 9.27185044e-02 -1.59725398e-01 7.24380493e-01
1.99193835e-01 9.38558280e-01 5.19263521e-02 1.38598755e-01
1.89534560e-01 5.22185385e-01 3.19387853e-01 -2.05418721e-01
-5.82295716e-01 -8.50391090e-02 2.47194245e-01 -3.98436099e-01
3.73453880e-03 -7.84714580e-01 -1.23246956e+00 -3.49381268e-01
-8.36384118e-01 3.65510464e-01 9.26467121e-01 8.20480108e-01
-3.92801464e-02 1.57090917e-01 8.01245868e-01 -6.26195550e-01
-1.13966763e+00 -7.10039079e-01 -7.31480777e-01 6.45017862e-01
3.43437850e-01 -5.17498374e-01 -8.36423159e-01 -1.57901764e-01] | [8.694286346435547, 4.226199626922607] |
98794c11-d5f7-4e3e-8baf-da041a8df630 | plasma-making-small-language-models-better | 2305.19472 | null | https://arxiv.org/abs/2305.19472v1 | https://arxiv.org/pdf/2305.19472v1.pdf | PlaSma: Making Small Language Models Better Procedural Knowledge Models for (Counterfactual) Planning | Procedural planning, which entails decomposing a high-level goal into a sequence of temporally ordered steps, is an important yet intricate task for machines. It involves integrating common-sense knowledge to reason about complex contextualized situations that are often counterfactual, e.g. "scheduling a doctor's appointment without a phone". While current approaches show encouraging results using large language models (LLMs), they are hindered by drawbacks such as costly API calls and reproducibility issues. In this paper, we advocate planning using smaller language models. We present PlaSma, a novel two-pronged approach to endow small language models with procedural knowledge and (counterfactual) planning capabilities. More concretely, we develop symbolic procedural knowledge distillation to enhance the implicit knowledge in small language models and an inference-time algorithm to facilitate more structured and accurate reasoning. In addition, we introduce a novel task, Counterfactual Planning, that requires a revision of a plan to cope with a counterfactual situation. In both the original and counterfactual setting, we show that orders-of-magnitude smaller models (770M-11B parameters) can compete and often surpass their larger teacher models' capabilities. | ['Yejin Choi', 'Xiang Ren', 'Keisuke Sakaguchi', 'Soumya Sanyal', 'Hirona J. Arai', 'Xiang Lorraine Li', 'Jena D. Hwang', 'Valentina Pyatkin', 'Chandra Bhagavatula', 'Faeze Brahman'] | 2023-05-31 | null | null | null | null | ['common-sense-reasoning'] | ['reasoning'] | [ 2.83906341e-01 9.70472932e-01 -2.29162976e-01 -3.23533952e-01
-8.66309345e-01 -5.33097506e-01 1.02793729e+00 1.16265498e-01
-3.45226735e-01 1.35213172e+00 6.06595039e-01 -8.78878236e-01
-4.00698364e-01 -7.96988726e-01 -8.19356501e-01 -2.72214293e-01
-2.37746775e-01 7.84480870e-01 8.97780061e-02 -2.14061931e-01
3.03600818e-01 3.85903865e-01 -1.40684521e+00 4.10496235e-01
1.08157265e+00 3.05935740e-01 2.28535593e-01 5.67945600e-01
-1.77943051e-01 1.36058378e+00 -5.54128826e-01 -3.14454108e-01
2.21255906e-02 -3.06543231e-01 -1.54586327e+00 -1.62223414e-01
-7.84054473e-02 -1.53878480e-01 9.82814580e-02 7.86129475e-01
8.28685835e-02 5.95125020e-01 2.56951898e-01 -1.25064099e+00
-1.56995282e-01 1.18997598e+00 -1.24567322e-01 -1.58789456e-01
8.67931008e-01 4.40616906e-01 7.39050806e-01 -8.87788013e-02
5.13641596e-01 1.62410486e+00 5.82399845e-01 6.43395007e-01
-1.56781077e+00 -4.49685842e-01 6.08855069e-01 1.47933602e-01
-1.15873230e+00 -5.02760887e-01 5.82769156e-01 -2.68006086e-01
1.43549681e+00 5.34587264e-01 7.63998747e-01 8.92594814e-01
4.50079471e-01 7.51385629e-01 1.48616076e+00 -5.57158530e-01
4.16986823e-01 -7.74096772e-02 -7.95432404e-02 6.80449307e-01
1.45567238e-01 4.56439108e-01 -4.94716138e-01 -3.24057251e-01
6.25635922e-01 -2.70488828e-01 -1.54818177e-01 -2.98138052e-01
-1.18395746e+00 7.52657652e-01 6.96520507e-02 3.96104008e-01
-5.55052340e-01 3.71449590e-01 1.78290203e-01 2.13851929e-01
1.27638906e-01 1.02316988e+00 -4.85926569e-01 -3.66652071e-01
-9.37348187e-01 6.97354019e-01 1.26044548e+00 8.83272588e-01
5.77035844e-01 -3.80805284e-02 -3.08985382e-01 2.79738456e-01
4.68472958e-01 1.61845431e-01 3.95500720e-01 -1.58419895e+00
5.03060281e-01 4.87956703e-01 7.41745949e-01 -7.07005799e-01
-7.09489346e-01 -1.87575206e-01 -4.09152657e-01 3.02226692e-01
5.71021378e-01 -1.44570902e-01 -7.91445017e-01 2.01777887e+00
4.28860903e-01 6.42102242e-01 3.10816228e-01 5.57277441e-01
2.69647270e-01 5.68973958e-01 5.74080348e-01 -6.28960490e-01
1.38601792e+00 -1.04753804e+00 -5.71255922e-01 -5.32475054e-01
8.50013494e-01 -4.41689938e-01 1.22501171e+00 4.64342326e-01
-1.43789208e+00 -1.83787867e-01 -7.41522849e-01 1.17680982e-01
-2.30167657e-01 -5.29643595e-01 1.23515725e+00 5.91595173e-01
-1.08633912e+00 6.44154191e-01 -9.36603665e-01 -8.62328261e-02
3.26521903e-01 3.12177688e-01 -1.84351221e-01 -1.28101736e-01
-1.27541816e+00 1.17571819e+00 7.57120728e-01 -6.50529042e-02
-1.12568903e+00 -8.81074071e-01 -1.14018261e+00 2.23905951e-01
1.11801589e+00 -9.46917057e-01 1.93344951e+00 -6.47345483e-01
-2.03075147e+00 5.53478897e-01 -1.39973596e-01 -8.62041354e-01
4.92662102e-01 -1.40571311e-01 -2.91580349e-01 -2.06295997e-01
2.87367314e-01 5.23875594e-01 2.76558548e-01 -1.16467333e+00
-4.94731367e-01 -1.86167777e-01 7.49276936e-01 5.34781873e-01
8.11759233e-01 -5.17451204e-02 6.89001381e-03 -3.94665897e-01
4.51777279e-02 -9.36401725e-01 -8.78718078e-01 -6.62261784e-01
-4.25998479e-01 -3.54120672e-01 3.29917409e-02 -3.97437394e-01
1.35660899e+00 -1.63460422e+00 -1.51451886e-01 9.96306911e-02
-1.36550711e-02 2.95849331e-02 -5.74806109e-02 4.56373870e-01
-1.92613065e-01 3.03921759e-01 -2.87104458e-01 -3.87095623e-02
3.26803714e-01 4.61175263e-01 -5.56625664e-01 -7.18666613e-02
-4.19113785e-02 1.09174895e+00 -1.18795538e+00 -6.63196445e-01
2.82134116e-01 -2.93537915e-01 -9.05524671e-01 4.71803397e-02
-8.69757116e-01 3.65186989e-01 -5.13741076e-01 2.15644777e-01
2.97737241e-01 -2.08810031e-01 7.81069160e-01 4.24495727e-01
-2.17675537e-01 9.33319390e-01 -1.31857800e+00 1.96832752e+00
-7.69504011e-01 6.80485591e-02 2.60140169e-02 -7.78104842e-01
2.49766290e-01 5.49147189e-01 1.12700045e-01 -5.75232208e-01
-4.96361554e-02 -1.96720976e-02 1.38156116e-01 -4.99148041e-01
5.44759929e-01 -8.71733129e-01 -4.74133492e-01 8.13763559e-01
-2.44485691e-01 -8.90523612e-01 3.10999274e-01 2.16932774e-01
8.86791945e-01 5.36181867e-01 1.08850420e+00 -4.65004861e-01
4.13843364e-01 3.87586176e-01 7.38143325e-01 1.09275198e+00
-7.32843801e-02 -1.16389900e-01 5.57309270e-01 -6.38346374e-01
-3.71594995e-01 -9.47316766e-01 2.77561247e-01 1.04401398e+00
-2.96421144e-02 -4.95184064e-01 -5.51173508e-01 -5.73734581e-01
-1.74345896e-01 1.65615022e+00 -2.97606021e-01 -1.34816989e-01
-7.39954710e-01 -6.13399923e-01 7.24254727e-01 3.49694729e-01
4.87683117e-01 -1.36022019e+00 -1.06365836e+00 5.10801494e-01
-5.12567639e-01 -7.34870553e-01 -2.90833145e-01 -1.66054927e-02
-9.13280427e-01 -1.27304232e+00 1.20522372e-01 -2.59745181e-01
3.19565475e-01 -1.42848445e-02 1.44847405e+00 -1.08708613e-01
2.62419164e-01 4.57139492e-01 1.34702668e-01 -6.43519044e-01
-5.34942031e-01 -3.90997678e-01 8.16088468e-02 -6.27283335e-01
6.24908358e-02 -7.51182079e-01 -1.24531277e-01 -6.74856529e-02
-6.98597312e-01 7.32174635e-01 4.30116564e-01 7.76305676e-01
4.86875564e-01 3.20307136e-01 4.71661866e-01 -1.15841472e+00
8.90048265e-01 -3.74901414e-01 -6.19403660e-01 4.99187350e-01
-6.86611116e-01 6.09239340e-01 6.94520414e-01 -3.45207989e-01
-1.86382401e+00 -1.35672405e-01 7.95795545e-02 2.36665279e-01
-3.61619532e-01 8.52426231e-01 -1.46657646e-01 3.62072259e-01
7.34733224e-01 1.76718235e-01 -1.82190135e-01 -1.46659419e-01
7.88579106e-01 -9.55082756e-03 5.79445720e-01 -1.42993677e+00
5.33916295e-01 3.15055102e-01 4.73847352e-02 -2.45451778e-01
-1.01864433e+00 -3.41919484e-03 -3.15621316e-01 4.70187664e-02
6.16323352e-01 -8.47236097e-01 -1.32005370e+00 5.65636232e-02
-1.14409959e+00 -1.10487807e+00 -7.34942436e-01 4.44416612e-01
-1.28149211e+00 1.13790587e-01 -3.86543542e-01 -1.10556006e+00
-2.01594234e-02 -1.02610993e+00 7.15136111e-01 2.59653240e-01
-6.51275337e-01 -1.04570317e+00 8.23320374e-02 5.52694917e-01
2.21517384e-01 2.14673802e-01 1.03901923e+00 -6.42942846e-01
-7.76482165e-01 3.78254116e-01 1.83652729e-01 -3.11638564e-01
-1.23495407e-01 -4.15607244e-01 -6.20037079e-01 1.00194305e-01
2.50415415e-01 -3.94905508e-01 1.95648402e-01 2.82791257e-01
1.09288561e+00 -9.88609910e-01 -4.36159790e-01 3.74217406e-02
1.13399911e+00 5.97909987e-01 6.45182073e-01 3.12108696e-01
4.90133949e-02 6.71026349e-01 9.79056239e-01 3.59804034e-01
7.98322678e-01 6.59965098e-01 8.19924772e-02 4.41866338e-01
1.77486554e-01 -7.42975891e-01 2.69400388e-01 1.80607110e-01
-3.43470305e-01 1.75440654e-01 -1.28427017e+00 6.57430768e-01
-2.02772069e+00 -1.45745981e+00 1.79559663e-01 1.95956552e+00
1.32088840e+00 3.05506438e-01 -1.38090327e-01 -9.78894904e-02
1.71392396e-01 2.37142712e-01 -4.45572883e-01 -7.17714548e-01
4.63056296e-01 4.18936461e-01 1.17025226e-01 1.17154527e+00
-6.82986736e-01 1.22164106e+00 6.75770807e+00 7.94317603e-01
-6.31095290e-01 1.56747088e-01 4.21764314e-01 -2.14981869e-01
-6.18153691e-01 4.23813283e-01 -3.61617357e-01 1.83534607e-01
1.16745281e+00 -6.45522773e-01 7.74101913e-01 7.01059937e-01
5.03524661e-01 -6.54523432e-01 -1.39627314e+00 5.94352245e-01
-3.37248892e-01 -1.65999877e+00 5.57005964e-02 -2.84452111e-01
7.88006544e-01 -3.46364200e-01 -3.16332012e-01 9.27101374e-01
1.25065386e+00 -1.35506368e+00 1.01582479e+00 6.55209303e-01
4.17910278e-01 -5.89740098e-01 5.34735084e-01 1.04329634e+00
-1.03091598e+00 -1.79650173e-01 1.62828401e-01 -8.75109136e-01
3.98069412e-01 2.77285904e-01 -9.95341659e-01 7.73204863e-01
3.13542187e-01 1.62285358e-01 -1.12549849e-02 6.61570251e-01
-7.27590621e-01 3.45547974e-01 -3.63281548e-01 5.56054711e-02
3.75283122e-01 -1.72825843e-01 4.02319849e-01 8.64283442e-01
3.45938616e-02 7.68744230e-01 3.82808298e-01 1.12003875e+00
2.08920971e-01 -4.51646894e-01 -7.37182975e-01 1.82011381e-01
6.02357566e-01 6.59977615e-01 -3.86943966e-01 -6.46962941e-01
-5.23998067e-02 4.19095159e-01 4.54125166e-01 5.11139154e-01
-8.94830704e-01 1.78694740e-01 7.35293210e-01 -2.34954376e-02
-3.35088611e-01 -1.84588462e-01 -5.05899131e-01 -1.08317125e+00
-2.99789548e-01 -1.14130855e+00 4.57944781e-01 -1.06740022e+00
-6.77446902e-01 5.95592558e-02 5.73317528e-01 -5.55778980e-01
-7.48941004e-01 -2.29720220e-01 -8.39770138e-01 8.95603240e-01
-1.31413412e+00 -1.04404247e+00 1.91688567e-01 4.86262411e-01
6.10685289e-01 4.23147470e-01 1.09148145e+00 -2.16709778e-01
-2.15820909e-01 2.97119226e-02 -7.44269311e-01 -4.20019984e-01
2.10261285e-01 -1.22179413e+00 2.22847596e-01 8.89944375e-01
3.16493846e-02 9.96294975e-01 9.19676185e-01 -6.59593046e-01
-1.09470046e+00 -9.57191050e-01 1.19284797e+00 -6.06761038e-01
5.85285962e-01 6.40469268e-02 -5.53571701e-01 1.30027580e+00
1.34733245e-01 -5.36616981e-01 6.21138573e-01 4.01283890e-01
-1.31879598e-01 3.33061516e-01 -1.28571308e+00 1.22660661e+00
1.35960805e+00 -5.60492575e-01 -1.40941882e+00 2.46064618e-01
1.04157972e+00 -6.48094833e-01 -5.85689247e-01 1.87711105e-01
4.08904254e-01 -8.66930723e-01 1.02589393e+00 -1.12866676e+00
2.66974896e-01 -4.67610717e-01 2.68181823e-02 -1.35624480e+00
-2.67257571e-01 -1.13716817e+00 -1.62671641e-01 7.85231233e-01
3.70757490e-01 -8.03100824e-01 6.60641372e-01 1.25646222e+00
-3.10770661e-01 -8.45577955e-01 -1.02127874e+00 -7.79822290e-01
4.67998572e-02 -1.04554868e+00 1.07597482e+00 9.53030705e-01
6.90241098e-01 2.87585944e-01 -1.50120422e-01 2.12820098e-01
3.19817007e-01 4.63234276e-01 6.14471436e-01 -9.70343530e-01
-3.27963650e-01 -4.45654422e-01 3.24292511e-01 -1.02073550e+00
5.60538530e-01 -8.55397761e-01 2.82815725e-01 -1.72947431e+00
-3.09412349e-02 -7.96058357e-01 -2.97839148e-03 8.34889054e-01
4.12272550e-02 -5.27387977e-01 3.55024964e-01 -2.46565714e-01
-7.26891339e-01 3.93901080e-01 1.28477621e+00 -5.77693172e-02
-5.24365306e-01 1.30554572e-01 -1.07621777e+00 1.10220337e+00
8.18143487e-01 -3.92853558e-01 -7.54443884e-01 -1.70537308e-01
3.59831452e-01 8.61405373e-01 4.24859107e-01 -8.16945732e-01
4.37760711e-01 -1.09522629e+00 -3.62923801e-01 -2.05064461e-01
2.10475132e-01 -7.15724885e-01 6.16218090e-01 6.64262593e-01
-4.86913145e-01 -3.47498097e-02 5.91912866e-01 4.72341359e-01
-1.92964643e-01 -2.69995689e-01 3.66526544e-01 -5.10835648e-01
-8.45975041e-01 -1.24826096e-01 -6.54677272e-01 2.06886500e-01
1.35941267e+00 -5.47887981e-02 -3.56419832e-01 -3.94466966e-01
-9.30846512e-01 5.22791743e-01 2.07301527e-01 9.14267898e-02
4.12482798e-01 -9.88341033e-01 -4.35468912e-01 -4.28582802e-02
-2.09766313e-01 4.45282191e-01 3.69812071e-01 9.40127552e-01
-4.33118284e-01 9.21624482e-01 -1.71472266e-01 4.18166704e-02
-9.02021408e-01 7.04888105e-01 4.03767526e-01 -1.06695330e+00
-3.45513463e-01 7.24296629e-01 2.27537945e-01 -8.22734714e-01
-1.22724473e-01 -7.09999919e-01 -8.04057866e-02 -2.24447817e-01
4.45467204e-01 2.84039497e-01 -2.96435982e-01 2.69570556e-02
-4.76865679e-01 -5.67526221e-02 2.42145702e-01 -4.10959601e-01
1.18319976e+00 -8.10857490e-02 -1.99847221e-01 3.34613532e-01
9.02801529e-02 5.40212914e-02 -1.21143746e+00 -1.13750845e-01
3.87167364e-01 -2.25271046e-01 -3.75303656e-01 -1.29625797e+00
-5.41480146e-02 5.46038628e-01 -2.59180605e-01 2.11338028e-01
9.77505207e-01 8.86283070e-02 2.02395663e-01 5.89241087e-01
1.04930270e+00 -1.05601311e+00 -2.73993492e-01 7.19430327e-01
1.00042832e+00 -9.28293228e-01 1.97699115e-01 -4.06256914e-01
-6.94277287e-01 6.06436133e-01 5.75139701e-01 3.21774602e-01
1.63888112e-01 2.66084373e-01 -2.21672699e-01 -2.48005182e-01
-1.26419675e+00 -8.22814852e-02 -3.54773290e-02 5.66637337e-01
2.98427552e-01 7.12253273e-01 -3.30694735e-01 9.32881236e-01
-7.17740595e-01 3.79851162e-01 7.08457470e-01 9.61713731e-01
-2.65459269e-01 -1.03736985e+00 -4.71212029e-01 3.56187105e-01
-3.31288487e-01 -1.50709122e-01 4.82392497e-02 1.04227531e+00
7.92970285e-02 1.15220010e+00 -1.59759164e-01 1.08041480e-01
3.30963254e-01 4.79763806e-01 6.90558136e-01 -1.09699726e+00
-5.93330681e-01 -5.48055649e-01 6.59647048e-01 -1.04812038e+00
-6.38312876e-01 -6.05389297e-01 -1.54678512e+00 -4.82152998e-01
1.70244604e-01 3.68489742e-01 1.21267498e-01 1.24735689e+00
1.97108582e-01 6.08341157e-01 -1.68403968e-01 -8.01377177e-01
-8.84609640e-01 -7.09460557e-01 -3.15754235e-01 1.39102533e-01
3.91688384e-02 -5.96081793e-01 1.71749573e-02 -7.47175887e-03] | [4.202368259429932, 1.2268503904342651] |
df7feee3-3a78-4623-8672-cb37530432af | adversarial-learning-based-stance-classifier | 2209.04631 | null | https://arxiv.org/abs/2209.04631v3 | https://arxiv.org/pdf/2209.04631v3.pdf | Adversarial Learning-based Stance Classifier for COVID-19-related Health Policies | The ongoing COVID-19 pandemic has caused immeasurable losses for people worldwide. To contain the spread of the virus and further alleviate the crisis, various health policies (e.g., stay-at-home orders) have been issued which spark heated discussions as users turn to share their attitudes on social media. In this paper, we consider a more realistic scenario on stance detection (i.e., cross-target and zero-shot settings) for the pandemic and propose an adversarial learning-based stance classifier to automatically identify the public's attitudes toward COVID-19-related health policies. Specifically, we adopt adversarial learning that allows the model to train on a large amount of labeled data and capture transferable knowledge from source topics, so as to enable generalize to the emerging health policies with sparse labeled data. To further enhance the model's deeper understanding, we incorporate policy descriptions as external knowledge into the model. Meanwhile, a GeoEncoder is designed which encourages the model to capture unobserved background factors specified by each region and then represent them as non-text information. We evaluate the performance of a broad range of baselines on the stance detection task for COVID-19-related health policies, and experimental results show that our proposed method achieves state-of-the-art performance in both cross-target and zero-shot settings. | ['Yusong Tan', 'Lei Tian', 'Jiaying Zou', 'Haiyang Wang', 'Feng Xie', 'Bin Zhou', 'Xuechen Zhao', 'Zhong Zhang'] | 2022-09-10 | null | null | null | null | ['stance-detection'] | ['natural-language-processing'] | [-1.74036957e-02 3.47351640e-01 -5.05001664e-01 -3.54480147e-01
-7.11902142e-01 -3.74718249e-01 7.10485578e-01 4.08031583e-01
-4.23732579e-01 6.27181172e-01 9.60210383e-01 -5.02533495e-01
5.69745243e-01 -9.36603069e-01 -6.56378448e-01 -5.96583366e-01
1.13062061e-01 8.08738708e-01 -4.54252958e-02 -5.62341928e-01
-3.59429836e-01 -2.63611257e-01 -6.75211251e-01 4.16080654e-01
1.08862722e+00 6.60243034e-01 -8.28553140e-02 3.77581388e-01
1.46696597e-01 1.27442944e+00 -7.58256316e-01 -4.60261434e-01
2.07749698e-02 -2.32292280e-01 -6.04709029e-01 -2.34607369e-01
-9.14093629e-02 -6.70248806e-01 -3.97286296e-01 1.17654681e+00
6.25656486e-01 -5.82239516e-02 8.84635687e-01 -1.04489696e+00
-9.03407097e-01 7.31112480e-01 -4.74957794e-01 3.55664700e-01
1.94410041e-01 1.55500576e-01 7.27770329e-01 -6.25805259e-01
6.31719708e-01 1.52754796e+00 7.83094883e-01 5.28921306e-01
-8.93594503e-01 -8.36970568e-01 5.66581368e-01 -5.51055633e-02
-7.76221037e-01 -1.18094899e-01 8.35035443e-01 -7.61631489e-01
6.08667672e-01 1.59839571e-01 5.01505256e-01 1.98247755e+00
3.16781878e-01 8.56035233e-01 1.02229989e+00 2.40992568e-02
3.46775234e-01 2.40991890e-01 8.41401815e-02 6.08273447e-01
1.98430151e-01 -7.32516646e-02 1.15229025e-01 -6.79028213e-01
2.84877211e-01 4.29533422e-01 -2.95907557e-01 3.48881364e-01
-1.06421471e+00 1.45256817e+00 4.63862598e-01 1.99872211e-01
-7.88024485e-01 -3.29462618e-01 6.01959646e-01 -1.70066301e-02
1.21583045e+00 2.50646770e-01 -5.14865398e-01 1.73111841e-01
-6.30169094e-01 5.32188296e-01 7.54320204e-01 5.70318699e-01
2.88456351e-01 5.04818484e-02 -7.17408061e-01 5.74072719e-01
2.25663841e-01 1.11351812e+00 2.41939873e-01 -6.16603613e-01
7.10054517e-01 5.07829428e-01 1.87954813e-01 -1.43893278e+00
-6.19108677e-01 -4.67882693e-01 -1.19764864e+00 -5.41921079e-01
1.99576199e-01 -8.29716802e-01 -1.04184663e+00 2.04815197e+00
6.26315355e-01 4.75457668e-01 1.31622016e-01 7.54295647e-01
8.36649895e-01 1.06203604e+00 4.29864258e-01 -3.09960306e-01
1.66885149e+00 -6.49677455e-01 -9.36021030e-01 -3.83133113e-01
6.04105413e-01 -5.00908852e-01 7.87868202e-01 -6.01330623e-02
-6.32147074e-01 -1.28680155e-01 -4.04019058e-01 3.08507830e-01
-5.48778057e-01 -3.18706185e-01 1.33355130e-02 3.28518480e-01
-5.89929581e-01 -4.00496162e-02 -8.33333611e-01 -3.11771184e-01
5.37443876e-01 -2.60675937e-01 1.95641205e-01 -1.55020086e-02
-1.89005470e+00 8.85422349e-01 3.29265445e-01 -2.84713030e-01
-1.17065585e+00 -8.87215436e-01 -8.67675424e-01 -9.14861560e-02
5.45038104e-01 -8.44745994e-01 1.23257065e+00 -6.55057907e-01
-9.11451519e-01 8.79352927e-01 -1.98272001e-02 -5.18895328e-01
5.02318323e-01 -1.54841602e-01 -7.84969330e-01 1.15110986e-01
4.16317254e-01 2.78510869e-01 7.91790068e-01 -1.04994071e+00
-5.50101340e-01 -3.74894321e-01 2.35793203e-01 9.17683095e-02
-5.61038554e-01 4.27318066e-01 -7.06709102e-02 -1.10065234e+00
-6.61055863e-01 -1.11194074e+00 -3.72319490e-01 -7.49739528e-01
-6.22624755e-01 -3.09945077e-01 9.36202765e-01 -1.11921871e+00
1.19133747e+00 -1.93221009e+00 -1.79397373e-03 -4.00523283e-02
3.92609596e-01 3.33801329e-01 9.81908441e-02 3.78503948e-01
1.11553051e-01 1.24388374e-01 -1.25092566e-01 -2.33025372e-01
-5.81966341e-02 2.77615935e-01 -7.23093092e-01 5.00543654e-01
9.86902714e-02 8.61242235e-01 -1.09987259e+00 -3.57646435e-01
-1.87121347e-01 5.54190457e-01 -9.82563853e-01 4.37484473e-01
-6.78152084e-01 8.11519027e-01 -9.55785453e-01 4.14373785e-01
6.23687625e-01 -6.38088048e-01 2.95195431e-01 -3.10704578e-02
1.56419545e-01 3.29959869e-01 -2.74400026e-01 9.44174588e-01
-2.26876184e-01 8.59134644e-02 1.87835872e-01 -9.93468344e-01
4.03353095e-01 6.33866787e-01 8.32873225e-01 -4.80900526e-01
4.86266971e-01 -2.77243376e-01 -2.59854615e-01 -6.21449351e-01
2.99834192e-01 -2.74253428e-01 -5.20150721e-01 5.90172529e-01
-3.52688611e-01 4.76580590e-01 -4.83038038e-01 4.58236188e-01
7.49151051e-01 -4.94220853e-01 4.28454310e-01 -4.13907319e-01
3.44369709e-01 5.92393056e-02 9.47025120e-01 5.77708781e-01
-4.41899449e-01 1.82441264e-01 4.18395936e-01 -5.21847308e-01
-7.83988714e-01 -7.06220508e-01 6.91732690e-02 1.41039872e+00
-1.36440381e-01 -1.47472844e-01 -9.21105385e-01 -9.94862080e-01
-6.06364496e-02 8.57410014e-01 -1.07576203e+00 -1.32007271e-01
-6.58299208e-01 -1.04812920e+00 4.44608212e-01 3.88411492e-01
5.74943006e-01 -1.06961370e+00 -4.29294199e-01 2.35939547e-01
-8.36890817e-01 -1.19856167e+00 -6.70011997e-01 -3.38994026e-01
-1.54697195e-01 -1.03562057e+00 -9.48844552e-01 -5.33876300e-01
5.83280623e-01 7.91744739e-02 9.60771859e-01 -2.76191950e-01
3.90128285e-01 3.10662478e-01 -3.04914892e-01 -8.65448534e-01
-6.63328469e-01 7.97904357e-02 3.14324468e-01 1.00701399e-01
3.96463573e-01 -1.09742261e-01 -8.70146453e-01 6.77765757e-02
-7.86446691e-01 -8.92004892e-02 3.33187543e-02 7.15290964e-01
2.55750090e-01 -2.45194972e-01 7.45961189e-01 -1.35826266e+00
9.22657728e-01 -1.29611254e+00 -3.28001738e-01 1.66200742e-01
-3.05433631e-01 -2.75483012e-01 7.17149138e-01 -5.47397137e-01
-1.14463139e+00 -8.22026372e-01 -3.17680299e-01 -4.32513177e-01
-9.04523134e-02 7.34372675e-01 1.52342126e-01 8.24281335e-01
7.15490520e-01 4.43960838e-02 -1.89596564e-01 -3.33571523e-01
4.27451909e-01 1.01699686e+00 3.30784887e-01 -3.98177773e-01
7.04148114e-01 6.62605405e-01 -7.74771333e-01 -6.95501089e-01
-1.60777199e+00 -4.28070009e-01 2.66761839e-01 -1.33218095e-01
1.49089682e+00 -1.37779367e+00 -5.50255120e-01 5.80158889e-01
-1.22880733e+00 -3.66229355e-01 1.79575950e-01 4.51222897e-01
-2.17540413e-01 6.11871947e-03 -8.90142381e-01 -5.97904921e-01
-6.45532131e-01 -1.11186993e+00 1.05049467e+00 -8.11199564e-03
-3.14711928e-01 -1.43734074e+00 6.73090041e-01 4.22125280e-01
4.85561550e-01 5.82606792e-01 1.09957814e+00 -1.16285312e+00
8.68578628e-02 1.24284744e-01 -8.40525553e-02 1.43357903e-01
4.76659924e-01 -5.15162945e-01 -9.42184687e-01 -5.82076669e-01
3.83951545e-01 -3.93988103e-01 6.95213735e-01 5.90650737e-01
1.04312623e+00 -1.19066155e+00 -5.14917374e-01 3.46671760e-01
8.90606105e-01 1.95934564e-01 1.55867800e-01 9.01929364e-02
7.12418735e-01 6.56150043e-01 3.85850728e-01 6.65793478e-01
8.77433062e-01 4.38341051e-01 3.46551895e-01 -5.26695371e-01
3.68813843e-01 -4.50933367e-01 3.24202448e-01 9.98695016e-01
9.74285677e-02 -5.16335666e-01 -1.17783415e+00 4.45762157e-01
-1.78999245e+00 -1.11223400e+00 4.07941788e-01 1.71196675e+00
9.65997994e-01 8.97899121e-02 4.01370883e-01 -5.98394632e-01
7.86657870e-01 6.94485188e-01 -5.94702065e-01 -1.65911302e-01
6.23049885e-02 -2.28561580e-01 3.19842756e-01 5.11765003e-01
-1.43013966e+00 9.04023647e-01 6.10162497e+00 4.09016043e-01
-1.42842388e+00 6.02110922e-01 9.17631686e-01 1.80008024e-01
-4.82398987e-01 -6.01769924e-01 -6.82911992e-01 8.28191340e-01
1.04915786e+00 -1.62972853e-01 1.96066156e-01 7.47645199e-01
3.66863698e-01 7.03734219e-01 -4.15752172e-01 5.38584054e-01
1.66593730e-01 -1.41210389e+00 -1.21894563e-02 1.12507090e-01
1.10966945e+00 5.43693721e-01 3.52169245e-01 6.33550346e-01
9.30573821e-01 -6.31695628e-01 5.07497013e-01 3.60465378e-01
8.29402328e-01 -7.18260586e-01 6.62484169e-01 7.12585807e-01
-7.51462281e-01 -2.46202216e-01 -1.67195886e-01 1.36914745e-01
2.80322939e-01 4.65662360e-01 -1.11286557e+00 3.87862742e-01
6.35001600e-01 7.75425315e-01 -3.14811096e-02 1.05117939e-01
-1.11038916e-01 1.05732632e+00 -5.43278269e-02 1.24357916e-01
5.84169567e-01 4.61389236e-02 6.78950310e-01 1.25810993e+00
2.07363680e-01 5.09210646e-01 7.74066269e-01 6.54202640e-01
-3.44967067e-01 2.03272536e-01 -8.58650744e-01 -1.23319671e-01
4.40667033e-01 6.84376657e-01 -2.36521736e-01 -8.09640527e-01
-4.71988738e-01 6.02274597e-01 2.81993866e-01 6.89692736e-01
-1.20518613e+00 2.85349041e-01 8.96272361e-01 2.85816163e-01
2.47731552e-01 3.61591071e-01 2.57378280e-01 -1.42051756e+00
-5.18855810e-01 -1.39683044e+00 6.04899883e-01 -4.15249467e-01
-1.41252136e+00 6.42203808e-01 1.24120519e-01 -8.95669639e-01
-5.49151778e-01 -3.56894642e-01 -7.98087120e-01 6.97614014e-01
-1.33300400e+00 -1.25923705e+00 1.02405436e-01 7.87386239e-01
4.70383018e-01 -2.54829496e-01 8.76390576e-01 1.97470009e-01
-7.87980080e-01 4.20140356e-01 1.85436994e-01 5.99071681e-01
7.73949325e-01 -8.34798634e-01 4.10231829e-01 6.48123503e-01
-3.89732748e-01 7.52083361e-01 8.79138708e-01 -1.09335768e+00
-8.80866587e-01 -1.77007592e+00 8.31116676e-01 -5.19543111e-01
7.60983646e-01 -4.64661598e-01 -8.79962087e-01 1.12418127e+00
4.30230469e-01 -4.29338306e-01 8.70931745e-01 9.18440372e-02
-5.36838770e-01 3.00229222e-01 -1.16728783e+00 7.66911626e-01
6.03619695e-01 -6.22792602e-01 -9.52073157e-01 7.73813009e-01
1.21592414e+00 -3.16810876e-01 -6.55361235e-01 4.76614028e-01
1.13530993e-01 -4.33662981e-01 1.03820288e+00 -1.25523949e+00
5.48810482e-01 1.67845160e-01 -1.73451796e-01 -1.55946434e+00
-4.93559897e-01 -5.65322280e-01 -2.74778545e-01 8.04984629e-01
3.47627282e-01 -9.94082928e-01 3.52570951e-01 2.85742670e-01
4.92310375e-02 -7.56941199e-01 -5.07262886e-01 -2.48240069e-01
2.88709104e-01 -1.60206228e-01 7.48042703e-01 1.59003282e+00
-7.34855235e-02 4.28253263e-01 -9.18384731e-01 5.30274451e-01
4.98732299e-01 2.02803850e-01 5.62209547e-01 -9.76316035e-01
-2.45030224e-01 -1.09912351e-01 2.48162180e-01 -8.99159551e-01
3.58822972e-01 -7.23458111e-01 -2.56419718e-01 -1.36355591e+00
3.93503040e-01 -2.99129874e-01 -5.82801759e-01 4.85904753e-01
-5.70384264e-01 -9.23541114e-02 2.60072891e-02 2.05048740e-01
-5.31610131e-01 6.14569306e-01 1.17714942e+00 -5.68192422e-01
-4.15774994e-02 1.18850224e-01 -1.06527984e+00 9.57112074e-01
8.34302843e-01 -7.24524081e-01 -4.74710107e-01 -3.90651017e-01
2.98977882e-01 1.62923008e-01 1.53010160e-01 -4.26662683e-01
-5.73604107e-02 -5.20327270e-01 5.06708026e-02 -6.20549321e-01
1.86918527e-01 -5.78248262e-01 -2.17149854e-01 7.41860151e-01
-4.57069576e-01 8.02251995e-02 6.01751767e-02 9.37523007e-01
2.15414949e-02 4.40220147e-01 6.44482195e-01 -1.97661385e-01
-9.65752527e-02 6.13188565e-01 -5.38385928e-01 7.38478482e-01
9.30495024e-01 6.36774778e-01 -7.84711123e-01 -7.07676411e-01
-6.75542772e-01 5.16125441e-01 3.46023381e-01 5.50257504e-01
3.31657469e-01 -1.19962537e+00 -1.21633255e+00 1.76175172e-03
6.62764385e-02 -1.26959175e-01 5.05731225e-01 8.63866031e-01
-2.04399049e-01 4.61624563e-01 1.06747583e-01 -5.03530741e-01
-8.58387053e-01 1.02505136e+00 2.52537489e-01 -7.09621489e-01
-5.81205904e-01 6.19942605e-01 8.46399128e-01 -7.40561604e-01
1.41207904e-01 -3.60402048e-01 -3.76084507e-01 1.90084502e-01
7.95026898e-01 1.05258666e-01 -3.80986810e-01 -7.44672060e-01
-4.51846689e-01 7.95022696e-02 -3.07138145e-01 2.02097297e-01
1.28561842e+00 1.88229293e-01 2.64784358e-02 2.59303659e-01
1.20683825e+00 2.30864942e-01 -1.09662688e+00 -5.58457255e-01
-3.86224359e-01 -4.96036299e-02 -7.60507304e-03 -8.71982455e-01
-9.75606918e-01 7.22147703e-01 3.34568232e-01 2.74056286e-01
8.61458957e-01 7.04866797e-02 1.04170799e+00 1.06904775e-01
7.50124082e-02 -9.42802966e-01 1.21509776e-01 6.51671886e-01
8.38615239e-01 -1.43547451e+00 -2.79746026e-01 -1.25928819e-02
-8.72764528e-01 2.07525864e-01 2.93531150e-01 8.66856724e-02
9.14672673e-01 1.94260418e-01 3.62152159e-01 -3.68484497e-01
-8.30940187e-01 1.59728631e-01 1.82737648e-01 4.65416938e-01
1.95682809e-01 5.64710200e-01 3.26335207e-02 5.95601499e-01
7.32170343e-02 -2.12453857e-01 5.73039539e-02 5.98053753e-01
-4.79864031e-01 -6.38438284e-01 -4.29470658e-01 4.26268488e-01
-8.85780275e-01 -2.90594071e-01 -2.97979712e-01 5.08542061e-01
1.74320161e-01 8.96840513e-01 -5.71570881e-02 -3.59726161e-01
2.19643474e-01 1.40769677e-02 -2.36950248e-01 -6.26223147e-01
-6.81706548e-01 1.33977458e-01 1.42697752e-01 -3.51685464e-01
-2.45071739e-01 -5.28279305e-01 -1.02812195e+00 -4.97147828e-01
1.97175637e-01 2.08646908e-01 1.57212988e-01 1.05349529e+00
4.73616958e-01 6.48275733e-01 7.35040963e-01 -2.40278631e-01
-9.12572742e-01 -1.26055336e+00 -8.84262554e-04 7.78116047e-01
5.94814360e-01 -5.60268223e-01 -7.97338784e-02 -9.70498249e-02] | [8.506062507629395, 9.501327514648438] |
b198824c-35c4-4c58-90c2-e103e9abc245 | stock-market-prediction-using-natural | 2208.13564 | null | https://arxiv.org/abs/2208.13564v1 | https://arxiv.org/pdf/2208.13564v1.pdf | Stock Market Prediction using Natural Language Processing -- A Survey | The stock market is a network which provides a platform for almost all major economic transactions. While investing in the stock market is a good idea, investing in individual stocks may not be, especially for the casual investor. Smart stock-picking requires in-depth research and plenty of dedication. Predicting this stock value offers enormous arbitrage profit opportunities. This attractiveness of finding a solution has prompted researchers to find a way past problems like volatility, seasonality, and dependence on time. This paper surveys recent literature in the domain of natural language processing and machine learning techniques used to predict stock market movements. The main contributions of this paper include the sophisticated categorizations of many recent articles and the illustration of the recent trends of research in stock market prediction and its related areas. | ['Saravanakumar kandasamy', 'Om Mane'] | 2022-08-26 | null | null | null | null | ['stock-market-prediction'] | ['time-series'] | [-6.71528876e-01 -2.79929429e-01 -7.22520888e-01 -2.17859477e-01
1.11809716e-01 -7.05710649e-01 4.82433796e-01 4.99402024e-02
-4.51862276e-01 9.06694770e-01 1.53999543e-02 -6.78564131e-01
-7.82772526e-02 -1.16625679e+00 -4.09165546e-02 -4.17016655e-01
-3.68814737e-01 3.61758828e-01 3.27196956e-01 -6.28633142e-01
8.91989827e-01 7.89755821e-01 -1.24714410e+00 3.19710770e-03
1.57231688e-01 1.46074891e+00 -1.62766933e-01 1.07164495e-01
-9.52289224e-01 1.38197649e+00 -6.31403565e-01 -7.00195849e-01
8.01039875e-01 -3.56234729e-01 -2.89248705e-01 -5.45125082e-02
-4.29985881e-01 -2.13600919e-01 -4.78497669e-02 1.08478439e+00
3.91710130e-03 -3.55087705e-02 3.10711414e-01 -1.41163075e+00
-4.39644784e-01 1.13591421e+00 -8.32268894e-01 8.46138954e-01
-1.69857055e-01 6.86953068e-02 1.39596307e+00 -5.70201099e-01
4.19901878e-01 7.93287098e-01 4.81113613e-01 1.21906653e-01
-6.41413808e-01 -9.17601645e-01 2.09449664e-01 -9.37972143e-02
-8.96787584e-01 -3.01970504e-02 6.98721349e-01 -4.62641180e-01
1.02287769e+00 3.57995480e-01 1.13625491e+00 3.14732641e-01
6.47413075e-01 9.99693751e-01 1.01556134e+00 -1.36773065e-01
4.93919134e-01 3.22656363e-01 1.19776435e-01 -6.31652400e-02
7.17064619e-01 1.28691465e-01 -5.65976620e-01 -9.31178406e-02
9.30656850e-01 4.56339717e-01 1.85634479e-01 1.34264544e-01
-1.08265567e+00 1.11253536e+00 1.02994561e-01 6.84166670e-01
-7.51736820e-01 8.97792447e-03 4.62922633e-01 9.50940907e-01
6.30702376e-01 5.92130899e-01 -7.65115142e-01 -6.05949163e-01
-1.43667781e+00 5.90785325e-01 1.29448950e+00 5.95236123e-01
3.29358339e-01 5.22645295e-01 4.17347759e-01 1.99563041e-01
3.23112071e-01 1.42195225e-01 9.23054278e-01 -7.13541627e-01
3.77004772e-01 7.52202928e-01 1.87786236e-01 -1.00947821e+00
-5.03763437e-01 -4.25982475e-01 -7.49785602e-01 7.26594687e-01
6.24027967e-01 -4.49741870e-01 -2.38266870e-01 9.26149428e-01
-2.17053592e-01 -5.86331822e-02 3.55046540e-02 6.18933916e-01
-4.83120158e-02 9.13027644e-01 -2.84584433e-01 -6.60334885e-01
1.16259408e+00 -7.95929730e-01 -8.74016166e-01 -2.84170151e-01
9.04809237e-02 -9.15270805e-01 2.06446350e-01 4.54771042e-01
-1.25835764e+00 -1.12491958e-02 -8.08143914e-01 6.25611663e-01
-6.15144908e-01 -5.02287447e-01 8.19164038e-01 5.76484203e-01
-8.16510558e-01 1.04725993e+00 -6.37380123e-01 1.63777173e-01
5.51305786e-02 2.24778399e-01 2.58248568e-01 7.72216439e-01
-1.28206277e+00 1.16550839e+00 4.27986771e-01 5.48830330e-02
2.54841030e-01 -7.30006874e-01 -4.36601698e-01 9.62840989e-02
4.91913974e-01 -2.26432115e-01 1.41155410e+00 -1.07061553e+00
-1.45507121e+00 6.54494762e-01 2.16967851e-01 -1.24114215e+00
8.39935124e-01 7.64104202e-02 -5.60767829e-01 -1.83720753e-01
4.50380007e-03 -5.44603867e-03 6.68984950e-01 -1.50208473e-01
-9.69700992e-01 -1.85342327e-01 -2.72245854e-01 -1.56223625e-01
-1.91688873e-02 4.88071769e-01 3.96086574e-01 -1.21755552e+00
2.15621561e-01 -5.36867142e-01 -4.59917635e-01 -2.19237253e-01
2.61426400e-02 -2.02174753e-01 4.56669629e-01 -4.19594795e-01
1.52251208e+00 -1.79282820e+00 -6.86258316e-01 6.74989462e-01
-5.37956469e-02 -9.84658375e-02 5.45126081e-01 9.80337203e-01
-4.67690200e-01 4.76217747e-01 4.59305104e-03 2.17538729e-01
4.15747404e-01 -5.19367941e-02 -9.11605418e-01 3.54382634e-01
3.38177383e-02 9.97661948e-01 -4.27121162e-01 1.81919783e-02
5.19558564e-02 -3.08431000e-01 6.43308908e-02 -2.15108141e-01
-1.93717599e-01 -3.56593519e-01 -4.51445758e-01 8.08458030e-01
4.04617518e-01 -3.37164909e-01 -1.02703944e-01 4.75751817e-01
-9.23820257e-01 4.27751303e-01 -1.52929890e+00 4.99274284e-01
7.36558959e-02 6.95529461e-01 -1.46572873e-01 -8.69661391e-01
1.09234846e+00 3.65437627e-01 7.55547583e-01 -6.55615151e-01
1.73381194e-01 6.87049508e-01 2.61673838e-01 -9.13468823e-02
6.98836148e-01 -7.78070688e-01 -5.35389818e-02 1.15633368e+00
-7.97314644e-01 -9.60604399e-02 6.11007690e-01 -7.18963966e-02
8.49298418e-01 -4.33992058e-01 7.30798006e-01 -4.08059239e-01
1.63848847e-01 2.70979796e-02 5.88608146e-01 4.21254545e-01
-2.85737574e-01 9.24287643e-03 7.83829629e-01 -8.72126222e-01
-9.42787051e-01 -5.69222152e-01 4.70466241e-02 7.37047970e-01
-3.45734894e-01 -5.48345372e-02 -2.48433173e-01 -8.75135139e-02
4.42290485e-01 6.74575150e-01 -5.03383815e-01 4.23092902e-01
-3.76648635e-01 -7.67044187e-01 -2.91069280e-02 5.83193958e-01
5.21832585e-01 -1.30989003e+00 -9.27883208e-01 5.09363651e-01
4.59968418e-01 -6.75520778e-01 -3.34957689e-01 2.64390916e-01
-1.03543293e+00 -8.61787438e-01 -9.51952815e-01 -2.59011686e-01
1.76017761e-01 1.62020087e-01 1.11984873e+00 5.24183214e-02
-3.36164758e-02 -1.04723033e-02 -1.71522126e-01 -1.08976746e+00
-1.96258470e-01 3.21403258e-02 8.79683495e-02 1.35298684e-01
6.66311204e-01 -4.52054232e-01 -4.29845095e-01 8.42408314e-02
-8.57652843e-01 -4.49269146e-01 3.85631591e-01 4.72368687e-01
2.83425897e-01 6.91959083e-01 9.17727947e-01 -7.21586704e-01
8.74504030e-01 -8.38206410e-01 -1.32067418e+00 9.19333026e-02
-1.12397611e+00 2.99568325e-02 -3.55032808e-03 -6.29164502e-02
-9.03752208e-01 -3.24875474e-01 1.55055851e-01 8.96232054e-02
3.84338409e-01 9.06703651e-01 4.80821282e-01 1.38425574e-01
-2.32566241e-02 1.58826575e-01 2.94620454e-01 -4.74920839e-01
-8.93471986e-02 2.67455101e-01 1.45857101e-02 1.08130142e-01
8.92645180e-01 4.38226521e-01 -3.15763764e-02 -8.45250010e-01
-4.94180501e-01 -5.97685754e-01 -4.97151196e-01 -2.60653198e-01
4.95603293e-01 -6.84486806e-01 -5.71361542e-01 7.64718711e-01
-8.29454184e-01 3.79666500e-02 -6.23674273e-01 5.15445232e-01
-2.89120495e-01 2.68860795e-02 -8.66068184e-01 -1.32071853e+00
-3.06829512e-01 -6.66048825e-01 1.82213131e-02 4.23523456e-01
-7.08293319e-01 -1.29984236e+00 3.09256941e-01 6.15399405e-02
7.26952791e-01 2.85226911e-01 3.54553580e-01 -1.19274092e+00
-8.82653832e-01 -5.76503456e-01 4.88395393e-02 2.59676486e-01
2.81633228e-01 2.66685635e-01 -4.24601585e-01 1.59455255e-01
4.47457641e-01 1.79405302e-01 8.58905435e-01 6.33023858e-01
2.99258113e-01 -4.07230347e-01 1.22892149e-01 8.25949833e-02
1.51833344e+00 7.67859995e-01 6.18830025e-01 1.00891781e+00
-1.85902223e-01 8.01583588e-01 7.25509465e-01 9.62213457e-01
1.06762193e-01 1.80287976e-02 2.34957099e-01 2.12072909e-01
9.00324345e-01 9.27595189e-04 3.55714828e-01 7.73226678e-01
-1.92512542e-01 1.77096710e-01 -9.35671210e-01 3.74802411e-01
-1.74657691e+00 -1.60324001e+00 1.13138750e-01 1.72568655e+00
7.59813905e-01 6.97636783e-01 6.07816160e-01 4.13037211e-01
5.16828299e-01 2.90066957e-01 -5.44250071e-01 -4.20684308e-01
-3.77521783e-01 9.65140462e-02 9.48872805e-01 1.73464373e-01
-9.17299807e-01 7.92788923e-01 6.79886103e+00 5.40898204e-01
-1.31310368e+00 -5.54745853e-01 8.15866947e-01 -1.07798494e-01
-3.87278229e-01 -6.94588348e-02 -8.70704114e-01 6.42677665e-01
8.97501290e-01 -9.05611515e-01 1.22955419e-01 1.01234746e+00
5.64830780e-01 -2.50772953e-01 -5.51887095e-01 8.63694489e-01
-3.28476816e-01 -1.82293737e+00 -1.89098030e-01 2.45597452e-01
6.31780982e-01 9.83143523e-02 1.65066645e-01 -7.47440606e-02
2.41316870e-01 -7.15043962e-01 1.14942849e+00 7.22090065e-01
-3.35964352e-01 -7.85278797e-01 9.18012142e-01 3.78558636e-01
-1.22378266e+00 -3.07319105e-01 -3.69773924e-01 -8.20523441e-01
4.52967048e-01 6.56394541e-01 -3.19219530e-01 1.69225112e-01
6.95918918e-01 7.49250650e-01 -1.19371628e-02 1.23327661e+00
1.71024710e-01 5.08457482e-01 -4.24828619e-01 -5.78104615e-01
6.84101641e-01 -6.52651310e-01 3.28907937e-01 8.74101579e-01
4.45942104e-01 2.62077719e-01 -1.92748427e-01 6.84682846e-01
1.39017880e-01 2.82053500e-01 -6.44133151e-01 -6.48608685e-01
3.73981297e-01 8.51690471e-01 -1.24968040e+00 -2.83765882e-01
-8.26248348e-01 4.95201260e-01 -6.11161768e-01 5.15451245e-02
-2.01526001e-01 -4.41539258e-01 8.51968288e-01 6.37673914e-01
4.33843881e-01 -2.75046170e-01 -7.89421499e-01 -1.10304713e+00
1.54161960e-01 -7.86916077e-01 5.64334214e-01 -2.37660125e-01
-1.54635119e+00 2.69501060e-01 -1.48137808e-01 -1.33870935e+00
-7.65159249e-01 -7.97767520e-01 -9.69132662e-01 8.67953420e-01
-1.57057929e+00 -1.87771291e-01 7.05538332e-01 2.89526820e-01
6.00936770e-01 -8.40409338e-01 1.98919564e-01 -3.05853356e-02
-2.95330107e-01 -1.95140868e-01 2.82324255e-01 3.85407567e-01
1.35237336e-01 -1.27785766e+00 9.39978600e-01 7.75275946e-01
3.18786025e-01 5.60377538e-01 6.98854208e-01 -8.29726756e-01
-9.73921955e-01 -4.98183608e-01 1.37834632e+00 -1.98766351e-01
1.62410712e+00 3.56096834e-01 -8.40626001e-01 6.04823768e-01
4.59703296e-01 -6.84098423e-01 7.03115284e-01 -4.00742710e-01
-7.13644773e-02 -3.52270991e-01 -8.47521603e-01 6.34783447e-01
1.29656047e-01 -1.54809266e-01 -8.06434095e-01 -3.44406925e-02
2.66517401e-01 9.96519774e-02 -6.42302752e-01 -2.68377841e-01
5.82995355e-01 -1.21529019e+00 7.49176443e-01 -4.91112381e-01
-5.64992987e-03 1.23688705e-01 2.56022692e-01 -1.00001442e+00
-1.20753087e-01 -1.12460423e+00 2.64357507e-01 1.04670584e+00
7.51417696e-01 -1.08580339e+00 1.08081889e+00 1.30746889e+00
4.05981183e-01 -6.80753529e-01 -9.69433010e-01 -8.86267066e-01
2.20311373e-01 -6.24519408e-01 9.75000381e-01 1.06731784e+00
5.03585160e-01 -1.05652496e-01 -1.97055265e-01 -5.54792821e-01
7.36441433e-01 5.16183257e-01 2.88949162e-01 -1.41304207e+00
1.65207032e-02 -1.21514964e+00 -4.06081378e-01 -6.42305851e-01
1.68474298e-02 -4.89440620e-01 -5.64993322e-01 -1.27826536e+00
-4.52117234e-01 -1.93350822e-01 -4.74942327e-01 2.30362877e-01
5.68071008e-01 -1.44170463e-01 4.99370456e-01 4.10290211e-01
-1.15355581e-01 -1.04716243e-02 9.71783221e-01 -8.75487402e-02
-3.43940675e-01 7.59116888e-01 -6.95000529e-01 9.36163545e-01
1.32888496e+00 -2.92981595e-01 -2.59589553e-01 3.25316340e-01
9.76781428e-01 2.51676291e-01 -1.30559117e-01 -5.09994328e-01
5.80499172e-01 -7.41048336e-01 9.18215364e-02 -1.03718174e+00
8.42923969e-02 -7.88004637e-01 2.61638343e-01 7.13456333e-01
-1.94047958e-01 8.02431345e-01 5.13399765e-02 3.04483771e-01
-7.18006253e-01 -6.45256877e-01 5.85147560e-01 -7.00831175e-01
-8.50766957e-01 2.13987738e-01 -7.93128848e-01 1.20149933e-01
1.11267638e+00 -4.53051656e-01 4.90670912e-02 -6.56706095e-01
-4.15599018e-01 3.29230011e-01 1.07852563e-01 5.02433419e-01
5.35627127e-01 -1.00464416e+00 -5.54619789e-01 1.03128359e-01
-5.64672768e-01 -6.40195012e-01 -2.76445955e-01 4.97656733e-01
-8.89143229e-01 6.51502490e-01 -3.89965355e-01 3.98974806e-01
-7.43673503e-01 4.75892007e-01 3.94689828e-01 -3.89544010e-01
-5.24866045e-01 7.00659156e-01 -3.97224754e-01 4.37482476e-01
2.65182674e-01 -6.61190689e-01 -3.16425532e-01 9.05010581e-01
1.05165696e+00 6.67306244e-01 -2.54547179e-01 -3.31612498e-01
-1.82314456e-01 3.91616940e-01 -3.28592733e-02 -4.13270473e-01
1.76731789e+00 -1.20028295e-01 -6.24033630e-01 1.10776424e+00
6.92046165e-01 -2.29352266e-02 -9.22001719e-01 -1.28999576e-01
1.06056559e+00 -4.23728585e-01 -1.37986466e-01 -3.85319948e-01
-1.39948380e+00 6.17485642e-01 -4.52263430e-02 1.02848756e+00
7.75882721e-01 -3.33095104e-01 8.70582521e-01 5.47717512e-01
3.79548311e-01 -1.58772933e+00 -2.67365634e-01 6.36442304e-01
8.46070826e-01 -1.21538150e+00 1.29840285e-01 1.41830400e-01
-7.61623919e-01 1.38766348e+00 -1.58847973e-01 -3.66208524e-01
1.60849690e+00 6.59147263e-01 4.41879034e-01 -2.61583626e-01
-8.05444837e-01 -6.85628504e-02 -1.72464579e-01 -2.70321630e-02
3.83846223e-01 7.65902698e-02 -4.73035455e-01 9.04575288e-01
-5.33255398e-01 8.35683271e-02 7.22963393e-01 1.15240848e+00
-7.28007138e-01 -1.26690316e+00 -4.57092673e-01 6.67494237e-01
-1.14146864e+00 -2.36153930e-01 -5.22210419e-01 1.10869825e+00
-4.32574898e-01 5.75392962e-01 5.30180991e-01 -3.84404585e-02
3.40846509e-01 1.73369572e-01 -3.33118796e-01 -3.71079087e-01
-8.56854439e-01 1.99941650e-01 -8.86165276e-02 -2.34919891e-01
-5.17381728e-01 -1.24528027e+00 -1.30796981e+00 -8.84736300e-01
4.86021787e-02 3.52594912e-01 5.60765922e-01 7.36577153e-01
-2.02026397e-01 -6.13511913e-02 6.88896954e-01 -5.22250295e-01
-1.10851061e+00 -5.61529875e-01 -1.59119070e+00 -3.18626106e-01
2.98806161e-01 -3.78673732e-01 -5.48029721e-01 -8.66749976e-03] | [4.546178817749023, 4.199467182159424] |
111be701-0125-4436-a085-8d832c362674 | estimating-post-ocr-denoising-complexity-on | 2307.01020 | null | https://arxiv.org/abs/2307.01020v1 | https://arxiv.org/pdf/2307.01020v1.pdf | Estimating Post-OCR Denoising Complexity on Numerical Texts | Post-OCR processing has significantly improved over the past few years. However, these have been primarily beneficial for texts consisting of natural, alphabetical words, as opposed to documents of numerical nature such as invoices, payslips, medical certificates, etc. To evaluate the OCR post-processing difficulty of these datasets, we propose a method to estimate the denoising complexity of a text and evaluate it on several datasets of varying nature, and show that texts of numerical nature have a significant disadvantage. We evaluate the estimated complexity ranking with respect to the error rates of modern-day denoising approaches to show the validity of our estimator. | ['Jean-Marc Ogier', 'Mickaël Coustaty', 'Jérôme Brachat', 'Arthur Hemmer'] | 2023-07-03 | null | null | null | null | ['optical-character-recognition'] | ['computer-vision'] | [ 3.16767931e-01 -3.85092795e-01 4.09275979e-01 -2.58558929e-01
-1.01982248e+00 -7.30611205e-01 6.92145109e-01 5.81154585e-01
-9.32192564e-01 5.91319680e-01 2.43591577e-01 -3.98441166e-01
-3.60471189e-01 -6.65272593e-01 -4.79529589e-01 -5.74289441e-01
1.28698677e-01 1.09327123e-01 5.45886829e-02 -3.01181436e-01
7.99341559e-01 4.03404891e-01 -1.44009185e+00 1.34943128e-01
9.59724426e-01 7.66282916e-01 2.15359941e-01 8.45125914e-01
-2.77136862e-01 2.94338763e-01 -1.12758422e+00 -8.41341853e-01
1.50427252e-01 -2.47100115e-01 -4.53220069e-01 1.33083209e-01
3.77750665e-01 -1.70628503e-01 -4.91797954e-01 1.30097687e+00
4.40189719e-01 -4.06892784e-02 8.92528653e-01 -6.83302879e-01
-5.94906330e-01 4.36108291e-01 -5.03101170e-01 2.71368682e-01
5.66607833e-01 -1.64798617e-01 8.75467420e-01 -7.13551342e-01
7.24601984e-01 1.20414662e+00 6.78853214e-01 7.02034496e-03
-8.38590682e-01 -3.35677594e-01 -1.13460250e-01 1.75793320e-01
-1.42767966e+00 -3.38475108e-01 5.37657142e-01 -2.20337704e-01
4.47316051e-01 4.09749627e-01 2.36151919e-01 7.18885183e-01
3.63825798e-01 4.72077370e-01 1.21161079e+00 -7.55593061e-01
1.24314874e-01 6.01324104e-02 4.28366125e-01 4.55029786e-01
4.08024967e-01 -5.29689789e-01 -4.45851654e-01 -3.51724699e-02
2.63306618e-01 -1.42891586e-01 -3.23634446e-01 3.65987718e-01
-9.27856207e-01 6.68851018e-01 -2.00494289e-01 5.62023997e-01
-3.13515365e-02 8.70023817e-02 5.26017725e-01 5.79799175e-01
6.90799832e-01 2.77710736e-01 -2.74486601e-01 -3.63863200e-01
-1.07761419e+00 2.40418792e-01 7.65212297e-01 1.04660225e+00
4.66785818e-01 -2.86788970e-01 -2.74647027e-04 9.70471799e-01
3.49805504e-03 4.16302770e-01 4.63579297e-01 -6.38177991e-01
7.35604048e-01 2.49175459e-01 1.35670766e-01 -1.14278007e+00
-6.72497153e-02 -1.37168527e-01 -9.82947469e-01 -2.38572247e-02
8.69081974e-01 1.18753299e-01 -7.89834857e-01 1.06020737e+00
-3.42714787e-02 -6.58820152e-01 1.81218702e-03 4.47691381e-01
5.25960386e-01 8.46737325e-01 3.19084972e-02 -4.22588706e-01
1.55345535e+00 -5.63576519e-01 -1.11323285e+00 3.34988773e-01
4.88309115e-01 -1.17782521e+00 1.21468031e+00 9.75675106e-01
-9.96206343e-01 -4.79925334e-01 -8.72735918e-01 -1.54377893e-01
-6.10822439e-01 4.04390246e-01 1.94717318e-01 1.23606491e+00
-8.23563218e-01 9.68629777e-01 -4.59445894e-01 -2.52099186e-01
1.35369986e-01 -5.89359924e-02 3.20406109e-02 2.03346089e-03
-1.21717000e+00 9.19263005e-01 2.38896936e-01 1.02487922e-01
-4.29365605e-01 -4.64866817e-01 -6.17743373e-01 1.29693240e-01
3.09979230e-01 2.68276073e-02 8.34590912e-01 -6.93978190e-01
-1.16114080e+00 7.37090826e-01 -2.85120934e-01 -3.79270911e-01
9.42270637e-01 -2.86949307e-01 -7.13890195e-01 2.50484109e-01
-3.63823384e-01 -6.48147166e-02 1.14547515e+00 -1.01665533e+00
-6.09758139e-01 -2.79237956e-01 3.61823216e-02 -3.93195190e-02
-7.16678619e-01 2.78432786e-01 -6.16153002e-01 -1.01918364e+00
1.77109279e-02 -6.22916043e-01 -1.63937241e-01 -6.80567846e-02
-1.91426516e-01 -1.21664122e-01 6.42420530e-01 -1.11185312e+00
1.58585000e+00 -2.48746181e+00 -1.16365068e-01 3.57380331e-01
-8.38743597e-02 -1.12263553e-01 9.79539901e-02 6.72187746e-01
1.19438268e-01 4.91519630e-01 -3.63192528e-01 -1.47979051e-01
8.41009021e-02 -1.40492786e-02 -3.94855499e-01 6.56553984e-01
-3.53499204e-02 2.07374483e-01 -6.44350529e-01 -7.82266140e-01
1.84555333e-02 4.37781811e-01 -4.04766500e-02 -1.25490695e-01
9.32690594e-03 -2.41819024e-01 -2.58751243e-01 5.55930197e-01
6.83857322e-01 9.08411741e-02 2.09620953e-01 -1.35790557e-01
-2.66646087e-01 1.46677241e-01 -1.35646343e+00 1.28084660e+00
-4.04009134e-01 9.50085759e-01 6.67859092e-02 -4.92700517e-01
7.80176878e-01 1.61150724e-01 1.16247743e-01 -5.02991319e-01
1.91364497e-01 1.51432067e-01 -1.75465271e-01 -3.17435950e-01
1.13615620e+00 4.40622419e-02 7.44223744e-02 2.34467775e-01
-4.82488394e-01 -3.82720292e-01 8.63234401e-01 3.77757460e-01
1.02656734e+00 -1.00627050e-01 1.40677288e-01 -4.50321913e-01
6.97637558e-01 -1.59572154e-01 2.04390753e-02 7.17645347e-01
-7.37277269e-02 9.41368163e-01 7.09796309e-01 -1.19429201e-01
-1.26559067e+00 -8.14202845e-01 -4.57483917e-01 7.83138394e-01
1.23237893e-01 -5.92156291e-01 -1.01177788e+00 -3.77954423e-01
-1.62742898e-01 5.17063200e-01 -4.30507839e-01 2.83564925e-01
-5.76868534e-01 -8.73900831e-01 5.88728964e-01 1.47863314e-01
1.88167393e-01 -7.83012450e-01 -4.88877982e-01 1.01392269e-01
-1.10080048e-01 -1.24779832e+00 -7.26852298e-01 8.74296129e-02
-9.34472919e-01 -8.89012754e-01 -9.97155190e-01 -7.18929350e-01
8.71003330e-01 1.31181180e-01 1.02961612e+00 3.77639323e-01
-2.63880581e-01 3.42900753e-01 -7.39669681e-01 -3.98942173e-01
-6.93415880e-01 5.05971946e-02 -1.50866911e-01 -6.49653301e-02
1.91146433e-01 -1.40184358e-01 -5.23892105e-01 1.49394169e-01
-1.45145679e+00 -5.38652301e-01 5.15676379e-01 5.66623271e-01
2.67326713e-01 6.05496228e-01 1.19948305e-01 -9.33051348e-01
1.12191284e+00 -1.00577757e-01 -7.80902565e-01 3.87302011e-01
-9.57845569e-01 2.15868875e-01 8.38882625e-01 -4.15109009e-01
-1.00697875e+00 -2.09115148e-02 -3.26109856e-01 2.06757009e-01
-6.48698956e-02 5.43846369e-01 -3.48347053e-02 2.49419641e-03
4.33659673e-01 4.53440011e-01 -1.80685937e-01 -7.28147030e-01
1.09334499e-01 9.58142877e-01 6.44310355e-01 -5.52420616e-01
9.54480469e-01 3.62108707e-01 -3.51482518e-02 -1.28253901e+00
-4.77439642e-01 -5.21588743e-01 -3.45564187e-01 -2.84544915e-01
6.23501539e-01 -7.33460426e-01 -6.17701530e-01 6.75005555e-01
-1.29879081e+00 2.11833760e-01 -4.45168912e-02 3.61509293e-01
-2.03959346e-01 1.09466898e+00 -8.75254393e-01 -9.20720339e-01
-3.60153407e-01 -1.01008558e+00 9.29567099e-01 7.17228279e-03
-2.60421604e-01 -9.20935988e-01 -1.47354960e-01 2.27655947e-01
1.16586924e-01 1.19344229e-02 1.25061417e+00 -3.62199903e-01
-2.10166454e-01 -3.97616059e-01 -2.66345024e-01 5.28632522e-01
3.95349637e-02 5.33035338e-01 -6.75454021e-01 -3.69158626e-01
3.67357507e-02 5.20016961e-02 8.45265150e-01 2.70983160e-01
1.39575338e+00 -1.62221327e-01 -5.23161143e-02 2.14709103e-01
1.63857043e+00 1.11978635e-01 1.06119537e+00 4.58735406e-01
1.52120277e-01 6.80838585e-01 7.35130489e-01 6.38188720e-01
-5.46771809e-02 2.67813087e-01 1.67206079e-01 1.03230253e-01
1.29335150e-01 3.00582908e-02 3.60968441e-01 1.34215689e+00
-6.80717751e-02 -4.66430664e-01 -8.05214047e-01 5.49988389e-01
-1.10866868e+00 -6.85358167e-01 -5.57054937e-01 2.18853784e+00
1.02388084e+00 4.95649576e-01 3.88029292e-02 7.20724583e-01
8.31727982e-01 2.25168452e-01 1.37098968e-01 -6.57477856e-01
-3.05631995e-01 2.22521409e-01 7.82624125e-01 4.80020016e-01
-8.64655197e-01 4.84795511e-01 7.41672421e+00 1.31659949e+00
-6.82403207e-01 -5.49937263e-02 7.04533041e-01 1.96889043e-01
-4.44988191e-01 -5.72556555e-02 -6.93024397e-01 8.21726143e-01
9.61582839e-01 -1.88644230e-01 2.71782935e-01 6.34996176e-01
1.63688809e-01 -4.14488494e-01 -8.34830642e-01 9.99384940e-01
8.12189095e-03 -9.96166646e-01 3.57294343e-02 4.33810279e-02
5.09250104e-01 -5.57691693e-01 1.37454987e-01 -1.59293506e-03
-1.78619802e-01 -8.14653993e-01 8.80542994e-01 3.21078092e-01
7.72911072e-01 -8.68184984e-01 7.98046947e-01 1.72776729e-01
-8.92589033e-01 1.42586604e-01 -5.92281640e-01 -3.78451049e-02
1.20014511e-03 9.62956846e-01 -3.04863334e-01 5.39658606e-01
6.22255445e-01 1.19459219e-01 -7.08342552e-01 1.05656672e+00
-4.47514132e-02 6.06496453e-01 -3.87657493e-01 -4.14162368e-01
1.71372041e-01 -5.38132429e-01 2.33309507e-01 1.57971275e+00
4.59149837e-01 1.80622578e-01 -4.99079525e-01 2.91723430e-01
-2.68949509e-01 5.15529931e-01 -1.29931748e-01 -3.75717610e-01
1.81076735e-01 1.12065673e+00 -1.09186375e+00 -3.94797057e-01
-3.69771838e-01 9.54064369e-01 -1.72626898e-01 3.18638414e-01
-7.52082109e-01 -8.55263531e-01 2.54149914e-01 2.24079043e-01
2.16447875e-01 -5.66880822e-01 -5.03925562e-01 -8.09024930e-01
2.81524539e-01 -1.16353655e+00 1.61531448e-01 -4.70604986e-01
-1.00415802e+00 5.52492857e-01 -8.97692367e-02 -1.36797440e+00
7.99376220e-02 -8.26511681e-01 -1.08019426e-01 7.97704697e-01
-1.26948953e+00 -3.29788327e-01 -1.58527866e-01 3.36184502e-01
6.95335627e-01 2.62273476e-02 3.82854581e-01 5.78475118e-01
-2.94152886e-01 7.87649632e-01 8.27580512e-01 2.15448245e-01
7.46059358e-01 -1.17019582e+00 4.93784785e-01 8.52959037e-01
9.36978385e-02 5.65807939e-01 1.05592918e+00 -5.10780573e-01
-1.38060355e+00 -5.60335040e-01 1.00577927e+00 -4.20855939e-01
6.51099384e-01 -4.05134767e-01 -9.34401333e-01 -7.40302950e-02
2.09628612e-01 -5.38248658e-01 2.21007422e-01 -1.65006071e-02
-3.16177249e-01 -2.24286065e-01 -1.03362191e+00 7.71890759e-01
7.05828547e-01 -3.94585848e-01 -5.75607181e-01 4.39423680e-01
5.64275146e-01 -3.79316151e-01 -8.89492631e-01 -5.43441288e-02
6.57396197e-01 -7.32634723e-01 7.33118176e-01 -2.48513609e-01
7.42696643e-01 -1.13921329e-01 -1.20115697e-01 -1.02705479e+00
3.26301306e-02 -7.00722218e-01 2.45308653e-01 1.26523685e+00
4.30067629e-01 -4.76029098e-01 7.22263336e-01 3.72564018e-01
1.40998602e-01 -3.29633296e-01 -9.48104382e-01 -1.05000567e+00
7.17897266e-02 -3.78024280e-01 2.21940100e-01 6.00694180e-01
-2.61711180e-01 -9.29988921e-02 -4.02348697e-01 -5.59408814e-02
6.72727764e-01 -8.44257325e-02 5.00983477e-01 -1.03282714e+00
-5.72239831e-02 -5.47618389e-01 -3.81956130e-01 -9.10000801e-01
-1.44181460e-01 -3.43309671e-01 2.23910660e-01 -1.24497914e+00
1.44770995e-01 -2.36457616e-01 -9.18646157e-02 -9.41735730e-02
-3.23124230e-01 2.70126551e-01 3.31967354e-01 2.84366131e-01
-2.99058706e-01 2.18663156e-01 1.03925478e+00 -3.44366640e-01
6.02587350e-02 -2.12808445e-01 -3.76842886e-01 7.19296157e-01
5.25839567e-01 -7.86701143e-01 -3.35246623e-02 -3.22775424e-01
6.20483577e-01 -2.08969951e-01 -4.31837179e-02 -1.00796175e+00
1.22321799e-01 -3.41964513e-03 2.44910985e-01 -6.66188240e-01
7.04820305e-02 -9.33152199e-01 -2.20989361e-02 4.96604919e-01
-4.42967594e-01 2.10965693e-01 5.39203584e-02 6.34070396e-01
-4.30918217e-01 -8.46867979e-01 6.94465280e-01 -2.89350059e-02
-2.70969570e-01 -2.60397971e-01 -6.69492066e-01 6.80063739e-02
6.31983817e-01 -4.20422137e-01 -2.80295968e-01 -6.09666944e-01
-2.76942015e-01 -2.39287853e-01 5.77243984e-01 9.07574445e-02
6.26451015e-01 -7.50915468e-01 -8.23060274e-01 -1.48642808e-01
-2.05620956e-02 -4.00904894e-01 6.05599582e-02 6.07480049e-01
-1.21349680e+00 2.16887131e-01 -4.96208435e-03 -2.18241230e-01
-1.58852935e+00 5.66901028e-01 -2.74822176e-01 -2.59930819e-01
-5.56960106e-01 4.24015701e-01 -2.16003329e-01 1.38607189e-01
5.37098885e-01 -5.20944655e-01 -1.79780215e-01 3.38550031e-01
5.45407712e-01 7.72928357e-01 4.72208440e-01 -2.57440269e-01
-1.67651653e-01 6.90288901e-01 -1.90673336e-01 -3.26285690e-01
1.18994331e+00 -2.67987281e-01 -2.76075542e-01 3.23483407e-01
1.18788743e+00 4.31846380e-01 -8.21473300e-01 4.96503413e-02
2.64639646e-01 -6.62217915e-01 -8.34668279e-02 -4.39523757e-01
-7.83619285e-01 6.71179473e-01 5.89438856e-01 6.06984317e-01
1.40876007e+00 -4.20803130e-01 7.70668745e-01 4.20684606e-01
2.61824191e-01 -1.53370643e+00 -1.09105006e-01 4.45428103e-01
8.10790062e-01 -8.57272625e-01 5.65975070e-01 -5.78423619e-01
-3.46625805e-01 1.36944807e+00 -2.36887529e-01 3.32230330e-02
5.62415361e-01 4.14388627e-01 7.20896870e-02 2.02817678e-01
-4.88355845e-01 -6.78556040e-04 2.04218626e-01 3.30916137e-01
5.77220082e-01 -3.07389140e-01 -1.13854384e+00 2.04620004e-01
-2.10101068e-01 -2.67134190e-01 1.02843535e+00 9.63695526e-01
-3.62175405e-01 -9.64876771e-01 -8.02766323e-01 5.62203526e-01
-1.03655219e+00 -4.64190841e-01 -4.21995908e-01 6.79245710e-01
-2.33392924e-01 1.06349707e+00 -1.03029534e-01 -4.63332869e-02
2.34388173e-01 8.49144533e-02 4.64721799e-01 -3.28282416e-02
-7.53931642e-01 2.07558721e-01 1.83140546e-01 2.15473752e-02
-2.99245328e-01 -6.22306347e-01 -9.08369541e-01 -4.98546004e-01
-3.67387027e-01 2.42314264e-01 1.15440989e+00 6.87522352e-01
-2.67639607e-01 3.33158165e-01 4.88905847e-01 -4.01217192e-01
-8.98267388e-01 -1.01106620e+00 -7.99096167e-01 5.55345953e-01
1.38415843e-01 -1.63170137e-02 -7.57668972e-01 4.28203583e-01] | [11.903037071228027, 2.7341768741607666] |
04c0b1c9-607d-4018-af91-8ff15d79afa4 | random-sampling-for-fast-face-sketch | 1701.01911 | null | http://arxiv.org/abs/1701.01911v2 | http://arxiv.org/pdf/1701.01911v2.pdf | Random Sampling for Fast Face Sketch Synthesis | Exemplar-based face sketch synthesis plays an important role in both digital
entertainment and law enforcement. It generally consists of two parts: neighbor
selection and reconstruction weight representation. The most time-consuming or
main computation complexity for exemplar-based face sketch synthesis methods
lies in the neighbor selection process. State-of-the-art face sketch synthesis
methods perform neighbor selection online in a data-driven manner by $K$
nearest neighbor ($K$-NN) searching. Actually, the online search increases the
time consuming for synthesis. Moreover, since these methods need to traverse
the whole training dataset for neighbor selection, the computational complexity
increases with the scale of the training database and hence these methods have
limited scalability. In this paper, we proposed a simple but effective offline
random sampling in place of online $K$-NN search to improve the synthesis
efficiency. Extensive experiments on public face sketch databases demonstrate
the superiority of the proposed method in comparison to state-of-the-art
methods, in terms of both synthesis quality and time consumption. The proposed
method could be extended to other heterogeneous face image transformation
problems such as face hallucination. We release the source codes of our
proposed methods and the evaluation metrics for future study online:
http://www.ihitworld.com/RSLCR.html. | ['Nannan Wang', 'Jie Li', 'Xinbo Gao'] | 2017-01-08 | null | null | null | null | ['face-sketch-synthesis', 'face-hallucination'] | ['computer-vision', 'computer-vision'] | [ 5.46756573e-02 -1.93236396e-01 -2.31329769e-01 -4.27953005e-01
-6.14819705e-01 -2.06420392e-01 4.63194579e-01 -4.51969266e-01
-7.57918954e-02 5.73265791e-01 -1.52222009e-03 -3.47011983e-02
-2.54155874e-01 -9.69117284e-01 -4.79212612e-01 -5.70606887e-01
3.18619937e-01 3.98493052e-01 2.35347878e-02 -2.51480967e-01
3.68995279e-01 9.68288600e-01 -1.67104983e+00 1.30722607e-02
6.79011226e-01 1.27774060e+00 5.68730943e-03 3.67555171e-02
-2.89074570e-01 2.79962569e-01 -4.73342627e-01 -9.28094327e-01
4.70571548e-01 -4.58128214e-01 -2.09810808e-01 1.41390830e-01
6.27281070e-01 -5.55898309e-01 -7.75973618e-01 1.08951294e+00
1.01713347e+00 2.33253822e-01 4.69283938e-01 -1.51904690e+00
-6.84935808e-01 2.91453719e-01 -6.34361923e-01 -2.27101326e-01
3.26582223e-01 -8.68179128e-02 4.80237246e-01 -1.75833261e+00
6.83604300e-01 1.38173723e+00 5.76128542e-01 7.76019752e-01
-8.35777879e-01 -1.42533445e+00 -6.08988218e-02 4.29125994e-01
-1.98799670e+00 -1.10059845e+00 1.19843984e+00 3.80678326e-02
5.35815418e-01 2.96144873e-01 7.23738790e-01 7.61270285e-01
-3.19072872e-01 5.78140795e-01 8.55002761e-01 -3.45607966e-01
2.50213802e-01 1.13183565e-01 -3.70564729e-01 1.09178901e+00
1.56590149e-01 4.72772159e-02 -8.51953149e-01 -5.45266509e-01
1.15601563e+00 9.40118060e-02 -2.29655996e-01 -2.57620126e-01
-8.45801651e-01 7.78760433e-01 1.59308106e-01 8.37869197e-02
-2.15000495e-01 2.89451540e-01 3.16529363e-01 4.14079964e-01
3.65433753e-01 -8.62575974e-03 -1.09399341e-01 8.61016065e-02
-1.22154343e+00 2.39835501e-01 5.99729240e-01 1.30284739e+00
7.11013138e-01 1.43011555e-01 -4.30686921e-02 1.30203068e+00
2.39507362e-01 6.12825871e-01 2.58661836e-01 -1.15782070e+00
3.64431739e-01 2.74129868e-01 2.68372847e-03 -1.48990023e+00
6.26553893e-02 -9.78064165e-02 -9.24262583e-01 1.99331325e-02
2.64523178e-01 8.02445635e-02 -6.40824437e-01 1.64419544e+00
6.09610319e-01 4.82123017e-01 -2.41122454e-01 6.96843743e-01
1.03928483e+00 6.16830707e-01 -2.93452561e-01 -5.21962464e-01
1.31829691e+00 -8.66652548e-01 -9.61575270e-01 1.64025113e-01
-7.74525180e-02 -1.24216938e+00 1.11041820e+00 3.48191381e-01
-1.31669152e+00 -5.93674421e-01 -9.27394629e-01 3.07772998e-02
-7.62010664e-02 6.52901411e-01 8.06901038e-01 8.23314250e-01
-1.01545632e+00 5.08578300e-01 -3.39312166e-01 -3.07189077e-01
8.49325061e-01 5.72736859e-01 -4.35787022e-01 -4.45936441e-01
-9.33532357e-01 3.45759481e-01 -1.06016725e-01 1.70326576e-01
-5.93822420e-01 -6.78299725e-01 -5.36692441e-01 4.11859415e-02
5.72140932e-01 -3.57034445e-01 9.72899377e-01 -6.48476362e-01
-1.79720056e+00 4.66263384e-01 -4.22016978e-01 2.35193804e-01
4.74655241e-01 1.98272005e-01 -5.71309745e-01 2.68729836e-01
-1.45618230e-01 7.17678607e-01 1.37508476e+00 -1.01318336e+00
-4.28836226e-01 -4.81795907e-01 -2.00552762e-01 8.16751942e-02
-6.00636423e-01 -5.48389927e-02 -1.14376473e+00 -1.03691840e+00
3.48150611e-01 -9.57770944e-01 2.01179227e-03 7.83191562e-01
1.11529082e-01 -4.18720424e-01 1.06645572e+00 -5.18073499e-01
1.39073944e+00 -2.23688316e+00 -1.90765306e-01 4.23040986e-01
1.00935169e-01 3.66184711e-01 -3.64085793e-01 5.75339794e-01
1.07353546e-01 -1.26997307e-01 -4.73309346e-02 -2.65692115e-01
5.26201613e-02 -1.70808420e-01 -3.48168015e-01 6.92538202e-01
-1.15544900e-01 6.58069432e-01 -6.33875012e-01 -8.31363916e-01
1.27834886e-01 7.65561938e-01 -6.59734428e-01 1.34403497e-01
2.05515265e-01 -9.69193131e-03 -3.31695437e-01 1.01779997e+00
9.78837252e-01 1.08613014e-01 1.95831731e-01 -5.80182791e-01
2.72955656e-01 -2.04450130e-01 -1.51510239e+00 1.93564856e+00
-4.90056783e-01 5.01259625e-01 3.34300995e-02 -7.60144114e-01
1.16679585e+00 4.36353296e-01 4.42978054e-01 -6.25065506e-01
6.19537383e-02 5.30924976e-01 -2.28770986e-01 -2.04935372e-01
2.83810973e-01 -8.92207026e-02 3.66268039e-01 4.67376977e-01
5.56054972e-02 -1.75875291e-01 1.87841967e-01 -4.55756821e-02
7.00666547e-01 3.64608206e-02 3.73751760e-01 -2.09657565e-01
8.10773015e-01 -3.94035131e-01 6.63904130e-01 3.52824509e-01
-1.59413323e-01 4.21759665e-01 2.29825348e-01 -4.40487564e-01
-8.75216424e-01 -7.97861636e-01 -1.79027662e-01 6.46492481e-01
2.06315398e-01 -5.12010753e-01 -8.41439903e-01 -6.16556346e-01
-9.46987048e-03 4.15144235e-01 -3.76372963e-01 -7.34930187e-02
-7.89880812e-01 -2.26400569e-01 5.88354111e-01 2.18198746e-01
6.84290946e-01 -1.19349635e+00 -2.48742253e-01 9.27054584e-02
1.02617601e-02 -9.28282499e-01 -1.02087700e+00 -8.34114075e-01
-8.51893365e-01 -9.95308995e-01 -9.26672101e-01 -8.64527047e-01
1.03388894e+00 3.85934174e-01 7.02437818e-01 4.12883341e-01
-5.58587015e-01 4.00965691e-01 -2.29227528e-01 -2.93957621e-01
7.58224400e-03 -1.33033946e-01 1.76379874e-01 2.81606317e-01
2.35135436e-01 -7.06382394e-01 -9.75918353e-01 6.29186273e-01
-6.79658353e-01 -1.88000008e-01 5.75204372e-01 9.04655874e-01
7.27237940e-01 2.94241369e-01 6.47833169e-01 -8.14986050e-01
6.99905694e-01 -2.14357838e-01 -7.19963431e-01 3.95028591e-01
-6.44029975e-01 -9.73202810e-02 6.69942558e-01 -6.21591151e-01
-1.06000602e+00 1.90613002e-01 -1.11505762e-01 -7.88625419e-01
2.70254344e-01 1.29337534e-01 -3.84873003e-01 -5.49143970e-01
3.71848047e-01 4.94363189e-01 2.66351134e-01 -3.88253123e-01
2.43438289e-01 6.48441434e-01 2.81162441e-01 -5.89001238e-01
1.04023659e+00 5.19855022e-01 1.72143653e-01 -1.02927160e+00
-1.99565943e-02 -7.03152269e-02 -7.59849921e-02 -2.39390224e-01
9.63634625e-02 -7.30081737e-01 -1.00076509e+00 3.64405066e-01
-1.17094946e+00 1.34526715e-01 -1.35827824e-01 3.09382498e-01
-5.17758965e-01 3.52845460e-01 -3.53223562e-01 -8.07353795e-01
-5.90094864e-01 -1.28850663e+00 9.39444184e-01 2.26244122e-01
9.35680941e-02 -4.18548822e-01 -3.23431581e-01 2.47683272e-01
5.18388569e-01 -2.16077909e-01 8.43377531e-01 -1.12413689e-01
-6.31429732e-01 -2.80344725e-01 -4.21585262e-01 -1.44327477e-01
2.83819228e-01 -6.35125190e-02 -6.31310046e-01 -4.54214752e-01
-1.85829446e-01 -1.61631852e-01 3.76556456e-01 1.21879578e-01
1.42202425e+00 -5.48535228e-01 -2.78072685e-01 7.18690872e-01
1.41924977e+00 4.24897879e-01 8.14910352e-01 -3.29097867e-01
4.89746839e-01 5.84125459e-01 8.54869246e-01 7.67197192e-01
6.77494034e-02 9.59521592e-01 -3.11366897e-02 4.26738374e-02
-4.80329305e-01 -5.37882507e-01 1.15390107e-01 7.72063434e-01
-6.40125275e-02 -7.27032423e-02 -4.72069919e-01 2.55172133e-01
-1.68702662e+00 -1.03036737e+00 6.11481547e-01 2.28569889e+00
8.68799508e-01 -4.55827713e-01 2.39569899e-02 3.90531570e-01
7.55658448e-01 7.95965791e-02 -6.81347191e-01 -1.31548628e-01
2.39481881e-01 5.84631920e-01 1.96009025e-01 3.68628889e-01
-6.46365881e-01 1.06144893e+00 5.31920815e+00 1.54265964e+00
-1.09057426e+00 1.94788188e-01 5.91064095e-01 -2.81984329e-01
-2.28720397e-01 -1.02936931e-01 -8.29261124e-01 4.31519181e-01
4.13233548e-01 -9.58500281e-02 1.01313365e+00 9.47670579e-01
-6.69977162e-03 1.71881571e-01 -8.57444167e-01 1.82007575e+00
2.64963984e-01 -1.61111069e+00 3.92560422e-01 -4.74494770e-02
4.76510435e-01 -6.21566594e-01 1.27585933e-01 2.72963922e-02
-4.48078781e-01 -9.39882934e-01 6.43080473e-01 4.02712315e-01
1.38342834e+00 -1.17088056e+00 1.97338924e-01 3.81444655e-02
-1.56756628e+00 -8.64603836e-03 -6.50498867e-01 4.83503520e-01
-1.66202083e-01 4.81761187e-01 -4.64923352e-01 1.41980544e-01
5.99391818e-01 3.54223043e-01 -2.40728229e-01 6.86748326e-01
3.71165834e-02 1.21057011e-01 -3.52244824e-01 -6.58737645e-02
-3.00100803e-01 -4.92240191e-01 3.17350566e-01 7.15191722e-01
7.60623932e-01 5.63965678e-01 -1.05802208e-01 6.79092228e-01
-4.47601765e-01 4.56963122e-01 -6.22751176e-01 -1.58552453e-01
1.15553129e+00 1.17192161e+00 -7.11472869e-01 -2.31462419e-01
-4.57281947e-01 1.14264238e+00 1.56473979e-01 2.96622694e-01
-9.21601057e-01 -8.66005957e-01 6.93029702e-01 3.67566466e-01
4.46742833e-01 -1.64821282e-01 7.08091334e-02 -9.54316616e-01
1.18866272e-01 -1.16428018e+00 4.55591343e-02 -4.15405899e-01
-1.08000624e+00 7.90679753e-01 -1.12811744e-01 -1.32115984e+00
-6.28955960e-02 -2.88203716e-01 -2.33076990e-01 5.49386561e-01
-1.35200226e+00 -1.18314159e+00 -4.57482338e-01 1.00204933e+00
8.56944740e-01 -6.17243409e-01 8.38372648e-01 7.48954296e-01
-5.15688598e-01 1.35249937e+00 -1.84227973e-01 8.96956120e-03
7.19584644e-01 -2.80612677e-01 3.41251224e-01 4.77907956e-01
2.63161391e-01 7.55400479e-01 3.69238675e-01 -6.79551065e-01
-1.92401850e+00 -9.10185993e-01 7.23002136e-01 3.11701655e-01
2.01891914e-01 -3.82656813e-01 -4.49416697e-01 7.37837106e-02
-1.08969934e-01 5.25561154e-01 6.61398351e-01 -3.16831678e-01
-3.34071428e-01 -6.21226311e-01 -1.62758112e+00 1.02095509e+00
1.63952184e+00 -5.71807861e-01 1.78187445e-01 2.58575499e-01
3.58369678e-01 -3.26389104e-01 -8.71472061e-01 3.53870869e-01
9.37101543e-01 -8.61662626e-01 1.11931765e+00 2.83768736e-02
1.83725387e-01 -3.31875145e-01 -3.66220683e-01 -7.90186286e-01
-1.92218855e-01 -9.10057247e-01 -2.43891671e-01 1.29419291e+00
1.83353007e-01 -7.34664023e-01 9.68745828e-01 5.89762807e-01
4.11075890e-01 -1.08561718e+00 -1.11331427e+00 -6.08918905e-01
-4.60382700e-01 -2.33352765e-01 1.13909566e+00 8.31353605e-01
-3.45946342e-01 4.76254299e-02 -5.00836015e-01 -2.27997098e-02
8.89940023e-01 1.80852219e-01 8.62925529e-01 -9.66575742e-01
-1.61454052e-01 -2.38412365e-01 -3.45170796e-01 -8.36365700e-01
2.01769695e-01 -6.71720803e-01 -4.23318028e-01 -1.09542465e+00
1.25212837e-02 -8.59092414e-01 7.17158765e-02 3.28744978e-01
8.53238776e-02 5.65700889e-01 3.27548832e-01 3.44982773e-01
-1.48994088e-01 8.45627248e-01 1.36766446e+00 -5.40433750e-02
-1.05123974e-01 -8.54443610e-02 -4.75739151e-01 4.83611971e-01
8.07756782e-01 -4.82016683e-01 -9.29928303e-01 -2.56541908e-01
-1.30645931e-01 4.44472730e-01 -1.05352767e-01 -8.91241193e-01
3.46515328e-01 -3.53731632e-01 4.85551536e-01 -4.41409558e-01
8.25846970e-01 -1.09141207e+00 4.73482162e-01 4.14638191e-01
-2.80182087e-03 3.40941958e-02 6.02751374e-02 3.79502326e-01
-1.42083570e-01 -2.05812082e-01 9.47896421e-01 -8.20967406e-02
-3.95791531e-01 9.11797106e-01 1.14292860e-01 -2.33229488e-01
1.00028968e+00 -4.35140967e-01 -5.62566742e-02 -5.61752617e-01
-3.11241299e-01 -3.85564655e-01 3.71124297e-01 3.11963439e-01
1.22964334e+00 -1.75671709e+00 -6.65574133e-01 6.15395546e-01
-3.52554359e-02 -4.74519461e-01 2.36384079e-01 5.27676761e-01
-5.70661008e-01 2.60754019e-01 -2.07774520e-01 -7.09040985e-02
-1.38286257e+00 4.94475961e-01 3.47986855e-02 2.36799970e-01
-4.99656826e-01 8.23599041e-01 -8.76397938e-02 -3.00700039e-01
4.01651502e-01 3.67211848e-01 7.76322260e-02 8.00498994e-04
7.26823449e-01 6.76038980e-01 6.76863780e-03 -5.47040880e-01
-3.91721964e-01 8.88398826e-01 3.05563919e-02 -2.08564103e-01
1.12361455e+00 5.12610674e-02 -2.56282151e-01 -2.73598999e-01
1.32912016e+00 1.24823451e-01 -8.48381579e-01 -2.07732290e-01
-5.27350187e-01 -1.11150849e+00 6.87125325e-02 -3.72547865e-01
-1.63945699e+00 5.11110783e-01 7.79172122e-01 -5.30942559e-01
1.43653667e+00 -2.77888656e-01 9.68792915e-01 4.85451162e-01
7.76919067e-01 -1.11206436e+00 7.81836957e-02 -1.32288709e-01
1.31010878e+00 -1.08045280e+00 3.13573748e-01 -8.52316618e-01
-3.28048021e-01 1.12651062e+00 6.62101150e-01 -1.68461144e-01
1.06184149e+00 1.88292444e-01 -1.94244951e-01 -1.49475664e-01
-5.57695746e-01 3.37779105e-01 2.70957351e-01 4.07867670e-01
3.26715112e-01 -9.64243785e-02 -7.63396561e-01 5.09942114e-01
-7.86052570e-02 2.66873419e-01 4.30568270e-02 8.08196664e-01
-9.65937227e-02 -1.48526657e+00 -3.90946478e-01 5.38602829e-01
-1.36333898e-01 -1.57527477e-01 -1.76154599e-01 5.80403626e-01
1.21000908e-01 9.81709063e-01 -5.86344190e-02 -3.04119855e-01
4.27071512e-01 -1.85496539e-01 9.02204633e-01 -2.03545183e-01
-2.11168513e-01 6.24453016e-02 -1.03353791e-01 -7.73239553e-01
-1.37704223e-01 -4.89620954e-01 -1.15146959e+00 -8.03655446e-01
-4.90053713e-01 1.09144926e-01 8.23663771e-01 3.17888677e-01
6.57883584e-01 -1.00650467e-01 1.04734969e+00 -8.59448552e-01
-5.31785250e-01 -7.09403157e-01 -5.23588598e-01 1.63962483e-01
-1.83410719e-01 -7.93932915e-01 -9.30955783e-02 -1.45933568e-01] | [12.67884349822998, 0.027391087263822556] |
1b1bee1a-0022-4bb5-9b5d-3a8d002d7a52 | ai-generated-incentive-mechanism-and-full | 2303.01896 | null | https://arxiv.org/abs/2303.01896v2 | https://arxiv.org/pdf/2303.01896v2.pdf | AI-Generated Incentive Mechanism and Full-Duplex Semantic Communications for Information Sharing | The next generation of Internet services, such as Metaverse, rely on mixed reality (MR) technology to provide immersive user experiences. However, the limited computation power of MR headset-mounted devices (HMDs) hinders the deployment of such services. Therefore, we propose an efficient information sharing scheme based on full-duplex device-to-device (D2D) semantic communications to address this issue. Our approach enables users to avoid heavy and repetitive computational tasks, such as artificial intelligence-generated content (AIGC) in the view images of all MR users. Specifically, a user can transmit the generated content and semantic information extracted from their view image to nearby users, who can then use this information to obtain the spatial matching of computation results under their view images. We analyze the performance of full-duplex D2D communications, including the achievable rate and bit error probability, by using generalized small-scale fading models. To facilitate semantic information sharing among users, we design a contract theoretic AI-generated incentive mechanism. The proposed diffusion model generates the optimal contract design, outperforming two deep reinforcement learning algorithms, i.e., proximal policy optimization and soft actor-critic algorithms. Our numerical analysis experiment proves the effectiveness of our proposed methods. The code for this paper is available at https://github.com/HongyangDu/SemSharing | ['Dong In Kim', 'Zehui Xiong', 'Jiawen Kang', 'Dusit Niyato', 'Jiacheng Wang', 'Hongyang Du'] | 2023-03-03 | null | null | null | null | ['mixed-reality'] | ['computer-vision'] | [-4.47864920e-01 2.92332917e-01 -3.72138172e-01 -5.29353954e-02
-6.60131156e-01 -4.20108348e-01 1.90029472e-01 -4.88213241e-01
-3.70794654e-01 8.92842948e-01 2.73413777e-01 -4.51574892e-01
-1.43595949e-01 -8.17187965e-01 -5.36967158e-01 -7.96669245e-01
-2.70056069e-01 7.78989643e-02 -2.54220545e-01 -1.81602836e-01
-5.45970304e-03 2.36946404e-01 -1.06271207e+00 -4.25184220e-01
1.14506185e+00 1.41192245e+00 7.89359629e-01 5.93690753e-01
-2.02026567e-03 7.32905388e-01 -3.67271394e-01 -5.12385607e-01
3.61100376e-01 -2.38045737e-01 -2.71842062e-01 1.49900531e-02
-6.94062352e-01 -1.16856432e+00 -8.84406745e-01 9.39511180e-01
9.35371161e-01 3.98819987e-03 1.76575676e-01 -1.45125639e+00
-8.07850420e-01 3.35803926e-01 -7.04777718e-01 1.55187249e-01
6.35334909e-01 8.58332869e-03 5.82982063e-01 -7.02228725e-01
7.10502148e-01 9.46693838e-01 3.45383137e-02 6.71040535e-01
-4.99904901e-01 -6.97079778e-01 1.82388828e-03 2.30771348e-01
-1.43147457e+00 -5.37134647e-01 8.39371920e-01 7.79498667e-02
3.52430880e-01 3.99727821e-01 8.23758304e-01 8.84746313e-01
1.69155374e-02 1.17675889e+00 1.00457013e+00 -1.46423280e-01
5.67282736e-01 4.65318859e-01 -6.21959567e-01 5.86798906e-01
8.50543305e-02 1.27988264e-01 -4.55600411e-01 -2.25834832e-01
1.07176960e+00 2.07057167e-02 -4.93963212e-01 -3.44758958e-01
-9.23926413e-01 5.52268326e-01 4.39201713e-01 1.16753593e-01
-8.00555885e-01 2.21866295e-01 4.29430744e-04 4.09584820e-01
5.37830114e-01 -4.00998175e-01 -2.65042305e-01 -4.22783405e-01
-3.81283402e-01 7.29610096e-04 7.79051602e-01 1.41146779e+00
2.69753397e-01 -1.22948505e-01 -4.81065884e-02 6.17157161e-01
5.20336270e-01 8.26566637e-01 2.27413073e-01 -1.30447853e+00
6.72158062e-01 3.43081020e-02 6.49218917e-01 -1.00510061e+00
-1.90297008e-01 -4.87072170e-01 -1.10831523e+00 -1.67262629e-01
8.93579039e-04 -8.67822945e-01 -2.46605411e-01 1.63868999e+00
5.74241936e-01 4.22771364e-01 2.04041749e-01 1.34717262e+00
5.19131064e-01 8.05397570e-01 -7.62729347e-02 -6.08172238e-01
1.05600286e+00 -5.70514679e-01 -9.42336619e-01 2.60877758e-01
6.49306118e-01 -3.73994797e-01 7.41050482e-01 1.56640902e-01
-1.51033628e+00 3.04332133e-02 -8.88286829e-01 4.25833315e-01
6.79072738e-02 -2.11615078e-02 4.86987472e-01 9.56888199e-01
-1.06630611e+00 2.17600018e-01 -5.06421864e-01 -9.56594199e-02
6.32780731e-01 5.45849204e-01 1.14580676e-01 -6.44000843e-02
-1.38398981e+00 3.73582572e-01 -6.81342557e-02 -9.46506038e-02
-6.63643837e-01 -5.69049895e-01 -3.70754838e-01 1.55670196e-01
6.26704156e-01 -9.08259153e-01 1.27806783e+00 -8.46526742e-01
-1.93639433e+00 5.92891395e-01 1.70050904e-01 -2.28581876e-01
8.31730962e-01 -9.90238637e-02 -3.62697572e-01 4.07413602e-01
-1.46768659e-01 4.30188060e-01 5.36716402e-01 -1.25204670e+00
-8.24903071e-01 -4.90126014e-01 3.01810920e-01 8.32380474e-01
-6.11001968e-01 -1.23261452e-01 -8.51542830e-01 -4.73242730e-01
-6.02239035e-02 -1.02622294e+00 -3.22989434e-01 4.28401619e-01
-2.93173641e-01 1.25767171e-01 8.04400265e-01 -8.30561280e-01
1.12469792e+00 -2.27592516e+00 7.50896484e-02 2.22036391e-01
2.61774600e-01 -7.77600333e-02 1.39758676e-01 1.86382771e-01
8.63840878e-01 1.01174414e-01 1.64644986e-01 -3.46958786e-01
8.99253506e-03 1.72457114e-01 1.09123491e-01 4.57295477e-01
-7.65208781e-01 8.81873071e-01 -8.87004375e-01 -4.20737177e-01
2.03817651e-01 2.57019699e-01 -4.84693170e-01 3.85985106e-01
5.15169203e-02 6.82935536e-01 -1.10945666e+00 5.89469671e-01
1.03436279e+00 -4.12575126e-01 3.23118567e-01 2.22391337e-01
5.90854995e-02 -1.02334030e-01 -1.20977807e+00 1.90190637e+00
-1.03321886e+00 1.80335924e-01 7.03298748e-01 -9.34095979e-01
5.25856733e-01 5.37526488e-01 8.43966186e-01 -1.39232421e+00
2.37101674e-01 3.70470822e-01 -6.32817566e-01 -5.54350793e-01
4.21711475e-01 4.09881115e-01 -1.14128634e-01 4.94705409e-01
-4.90918249e-01 2.87852734e-01 -4.63873863e-01 6.36687875e-01
8.57313156e-01 -8.45950022e-02 7.03582093e-02 -2.48038433e-02
2.57269919e-01 -6.47142112e-01 5.15259147e-01 7.77102947e-01
-3.79933745e-01 -2.22206232e-03 1.76385224e-01 9.05443504e-02
-1.19669271e+00 -1.08161044e+00 1.40270233e-01 7.25661695e-01
1.11002576e+00 1.28834203e-01 -6.94127858e-01 -4.18179452e-01
1.59944549e-01 6.92440212e-01 -2.41795275e-02 3.27388532e-02
-9.44155008e-02 -3.56724292e-01 6.57346770e-02 1.49590209e-01
1.05954146e+00 -6.76711798e-01 -6.33922935e-01 3.30283999e-01
-4.39734012e-01 -1.24188924e+00 -6.53359592e-01 -6.24805152e-01
-6.91757083e-01 -4.42161739e-01 -1.29488266e+00 -4.34608459e-01
4.21999872e-01 8.45325947e-01 6.47232294e-01 -8.22265148e-02
1.55970365e-01 9.25280154e-01 -3.72751325e-01 -1.33990273e-01
-1.17871903e-01 -1.66051745e-01 1.21662140e-01 1.23519816e-01
-1.81257531e-01 -6.09188259e-01 -1.32178295e+00 6.45425141e-01
-5.88988602e-01 3.47296625e-01 4.89560455e-01 3.88741583e-01
4.21769291e-01 5.20788133e-03 7.97892690e-01 -3.01880747e-01
8.80213261e-01 -9.88709629e-01 -5.93297482e-01 1.62155956e-01
-3.43838304e-01 -3.09254408e-01 5.68484128e-01 -1.78639308e-01
-1.27083182e+00 -1.79391772e-01 -1.05405203e-03 -3.37028801e-01
3.02125007e-01 2.57002354e-01 -4.14728969e-01 -2.23409563e-01
8.29785615e-02 3.28436315e-01 4.69377525e-02 -1.44812077e-01
4.85096306e-01 1.34365797e+00 1.33023918e-01 -3.74515653e-01
5.86039126e-01 5.66128671e-01 -2.34622344e-01 -9.66059029e-01
-2.28815839e-01 -2.64388204e-01 3.14712286e-01 -6.35025263e-01
6.99328899e-01 -1.40080392e+00 -1.20453095e+00 3.12944978e-01
-1.07299459e+00 -2.41730615e-01 2.90361214e-02 7.35327899e-01
-7.33449519e-01 4.13276523e-01 -4.88433838e-01 -1.19168031e+00
-2.65740335e-01 -8.74830484e-01 8.64264309e-01 6.19721174e-01
4.77982283e-01 -6.85833156e-01 -4.17519033e-01 4.70841736e-01
6.67216361e-01 -5.61135896e-02 2.80224532e-01 1.31328940e-01
-1.10440409e+00 -6.67415783e-02 -3.61980796e-01 -1.19700126e-01
-5.79679534e-02 -7.23081470e-01 -6.28821015e-01 -4.16989118e-01
-1.24264605e-01 -2.16862515e-01 5.85483760e-02 7.63733149e-01
1.56254733e+00 -5.27307987e-01 -5.18436551e-01 4.15152043e-01
1.27088845e+00 4.02069300e-01 7.49305665e-01 1.18887030e-01
2.80003637e-01 3.41234416e-01 7.76110470e-01 1.24848771e+00
9.33106601e-01 9.03129041e-01 7.07624078e-01 -2.02161282e-01
2.52136648e-01 -2.54949033e-01 2.06593439e-01 7.19443440e-01
-2.29716957e-01 -8.19657683e-01 -3.21351379e-01 3.53048146e-01
-1.99635935e+00 -6.51149929e-01 1.47275820e-01 2.34641385e+00
2.75135636e-01 5.53094745e-02 1.50144756e-01 -2.15377942e-01
9.05215800e-01 -1.10851318e-01 -9.13072646e-01 -1.02050506e-01
-3.05276681e-02 -3.52686107e-01 9.93918180e-01 2.09194183e-01
-4.79862154e-01 5.68202674e-01 4.68714952e+00 1.08459401e+00
-7.00302064e-01 4.21994090e-01 9.44989383e-01 -3.18767637e-01
-6.95614219e-01 -2.80645758e-01 -1.37629002e-01 8.88624609e-01
9.66544092e-01 -5.51723838e-01 9.10130620e-01 8.19363356e-01
9.52433288e-01 -4.23800588e-01 -3.65030617e-01 1.36532605e+00
-4.35145885e-01 -1.49861276e+00 -4.32688594e-01 4.87748951e-01
7.25766778e-01 -1.05520912e-01 1.34679988e-01 -4.83853407e-02
3.54874104e-01 -3.14788938e-01 5.61787069e-01 5.78448713e-01
9.00886655e-01 -8.59082222e-01 3.89041245e-01 5.16964555e-01
-1.12661850e+00 -4.97493744e-01 -2.10873693e-01 1.10870019e-01
6.85660005e-01 5.76162875e-01 -1.40104920e-01 7.12618470e-01
6.06404483e-01 2.49272615e-01 4.06207144e-01 9.65792477e-01
6.75442293e-02 1.31824449e-01 -3.58417064e-01 -3.64987135e-01
1.03999332e-01 -4.16177601e-01 6.54352725e-01 4.54451561e-01
9.29198980e-01 7.60072947e-01 6.06840923e-02 5.59671581e-01
-5.03028631e-01 2.49878600e-01 -6.19365335e-01 8.26824829e-02
8.51415575e-01 1.11490953e+00 -3.21460843e-01 -3.79474819e-01
-6.79635882e-01 1.34996736e+00 -1.46815374e-01 5.09274602e-01
-1.01660597e+00 -2.34845757e-01 7.12607801e-01 1.06086224e-01
1.44637704e-01 -4.87608582e-01 -2.29408383e-01 -1.08090472e+00
3.47873151e-01 -3.70924920e-01 1.55208886e-01 -1.00084031e+00
-7.73754358e-01 8.73205513e-02 -3.79838318e-01 -1.32577288e+00
1.08437717e-01 1.00160599e-01 -4.11260515e-01 6.14974737e-01
-1.60505641e+00 -7.07994640e-01 -5.14444172e-01 8.61492395e-01
3.76895249e-01 -1.81186184e-01 3.79491359e-01 7.63868034e-01
-3.87386054e-01 7.47053862e-01 7.20116675e-01 -2.86313027e-01
2.23294571e-01 -7.32647300e-01 5.51649928e-02 3.77019644e-01
-4.13305044e-01 -5.01870327e-02 4.61818308e-01 -5.60080767e-01
-1.97692323e+00 -7.42620945e-01 2.33768329e-01 4.16126698e-01
4.43123221e-01 -3.24078500e-01 -2.69977927e-01 3.84618044e-02
2.38046005e-01 1.03170574e-01 4.65739697e-01 -5.80980599e-01
5.16414940e-01 -1.60280779e-01 -1.47057021e+00 8.15023184e-01
1.51199603e+00 -4.51207638e-01 5.37353694e-01 3.49867046e-01
8.76676917e-01 -5.43739855e-01 -7.69671857e-01 -6.77160993e-02
6.58229530e-01 -7.95035303e-01 7.95501888e-01 -8.40574503e-02
3.36201489e-01 8.06853771e-02 -2.78622031e-01 -1.37767184e+00
7.20036104e-02 -1.07816219e+00 -4.72606152e-01 7.81279743e-01
1.22907586e-01 -8.24030876e-01 1.06269395e+00 9.99284744e-01
1.77362353e-01 -6.39141619e-01 -1.14212632e+00 -7.09783912e-01
-3.52637142e-01 -2.32457250e-01 6.54509008e-01 5.89973092e-01
1.71078637e-01 -1.62584558e-01 -7.03793049e-01 4.11218584e-01
8.58495891e-01 -9.94609147e-02 7.09131181e-01 -5.98377824e-01
-4.67136979e-01 -1.09715238e-02 -1.54059529e-01 -1.81807780e+00
-9.01497155e-02 -6.02237225e-01 -4.16484416e-01 -1.49105692e+00
7.36757517e-02 -9.99354243e-01 -2.02438444e-01 -1.71951175e-01
2.57839054e-01 -3.19877639e-02 3.28166127e-01 4.39100146e-01
-8.59274566e-01 9.96148169e-01 1.75969946e+00 1.23461239e-01
-3.98007810e-01 3.37520093e-01 -7.75501192e-01 3.15101802e-01
9.51242447e-01 -2.22797781e-01 -6.61596060e-01 -4.80717987e-01
3.11380416e-01 1.18534732e+00 3.73140633e-01 -7.00702906e-01
3.45558554e-01 -2.94443071e-01 1.66292831e-01 -2.35926971e-01
4.09810215e-01 -1.12975490e+00 1.11287072e-01 5.77872038e-01
-1.13633730e-01 -3.83945793e-01 -1.95224121e-01 8.91210258e-01
3.20910424e-01 2.12514877e-01 4.50338602e-01 -8.76002386e-02
-6.06752574e-01 6.04343295e-01 -5.70569098e-01 -2.46105641e-01
1.47828937e+00 -2.30574667e-01 -2.97174364e-01 -1.36332273e+00
-7.64615059e-01 7.68192649e-01 2.38244787e-01 3.24871272e-01
6.58157706e-01 -1.51446450e+00 -4.82255578e-01 -2.61667430e-01
-2.31878147e-01 -5.38201034e-01 7.64087856e-01 9.00824726e-01
-2.75160134e-01 3.31317723e-01 -1.57370836e-01 -3.36082846e-01
-1.04270267e+00 3.84509683e-01 2.05635741e-01 1.26777440e-01
-4.60439146e-01 4.25052226e-01 5.95446602e-02 -9.99829397e-02
2.44132802e-01 3.50983530e-01 4.60316427e-02 -3.25403064e-01
3.11449021e-01 6.53648973e-01 -3.58660817e-01 -3.11031282e-01
-1.45628795e-01 1.76079869e-01 1.42940179e-01 -4.06891853e-01
1.10568678e+00 -1.06025589e+00 5.33679903e-01 -2.83234388e-01
1.23006380e+00 -1.27266690e-01 -1.40520072e+00 -3.31335247e-01
-5.99906385e-01 -9.18311000e-01 6.09638631e-01 -6.22103512e-01
-1.36078870e+00 4.25302684e-01 9.95280504e-01 3.79769504e-01
1.25999236e+00 -7.97766000e-02 1.22770250e+00 2.86975075e-02
9.16546047e-01 -1.46956968e+00 -1.97605975e-02 -3.44483465e-01
6.77230060e-01 -1.24121118e+00 -2.64896065e-01 -4.80082512e-01
-8.61197531e-01 6.61705613e-01 3.97151232e-01 2.04354510e-01
7.98112869e-01 1.66414306e-01 -2.54564524e-01 3.98476720e-02
-5.64208686e-01 -2.17981711e-01 -4.75159287e-01 4.98267233e-01
-4.58767898e-02 4.11408663e-01 -5.50748765e-01 4.05138940e-01
2.43625924e-01 2.62620926e-01 7.71936536e-01 1.00659108e+00
-3.11020464e-01 -9.64294851e-01 -1.59225047e-01 4.98097897e-01
-3.21968943e-01 1.78828239e-01 1.02989338e-01 3.09970975e-01
-2.79448092e-01 1.07907796e+00 -5.36090061e-02 -2.37010285e-01
1.88081712e-01 -7.72609353e-01 3.64564717e-01 2.68241465e-01
2.26862609e-01 3.45422700e-02 1.49316847e-01 -6.51424885e-01
-3.32017809e-01 -3.21254790e-01 -1.25423884e+00 -8.72510910e-01
-2.55866885e-01 1.21231325e-01 1.05370677e+00 7.00506270e-01
9.11610663e-01 2.20511168e-01 1.54997599e+00 -7.09836543e-01
-3.99319649e-01 -3.34894598e-01 -8.89066935e-01 1.67398244e-01
9.88276452e-02 -4.35365111e-01 -1.20824039e-01 -4.53914583e-01] | [5.989175796508789, 1.599082589149475] |
02145ca8-a57d-46ce-9634-f0dd225580d0 | open-challenges-for-monocular-single-shot-6d | 2302.11827 | null | https://arxiv.org/abs/2302.11827v1 | https://arxiv.org/pdf/2302.11827v1.pdf | Open Challenges for Monocular Single-shot 6D Object Pose Estimation | Object pose estimation is a non-trivial task that enables robotic manipulation, bin picking, augmented reality, and scene understanding, to name a few use cases. Monocular object pose estimation gained considerable momentum with the rise of high-performing deep learning-based solutions and is particularly interesting for the community since sensors are inexpensive and inference is fast. Prior works establish the comprehensive state of the art for diverse pose estimation problems. Their broad scopes make it difficult to identify promising future directions. We narrow down the scope to the problem of single-shot monocular 6D object pose estimation, which is commonly used in robotics, and thus are able to identify such trends. By reviewing recent publications in robotics and computer vision, the state of the art is established at the union of both fields. Following that, we identify promising research directions in order to help researchers to formulate relevant research ideas and effectively advance the state of the art. Findings include that methods are sophisticated enough to overcome the domain shift and that occlusion handling is a fundamental challenge. We also highlight problems such as novel object pose estimation and challenging materials handling as central challenges to advance robotics. | ['Markus Vincze', 'Jean-Baptiste Weibel', 'Peter Hönig', 'Stefan Thalhammer'] | 2023-02-23 | null | null | null | null | ['occlusion-handling', '6d-pose-estimation'] | ['computer-vision', 'computer-vision'] | [ 9.72731262e-02 -2.16971576e-01 -6.42241418e-01 -8.83272663e-02
-2.43298516e-01 -5.13042629e-01 4.05127138e-01 -1.48175552e-01
-3.22220355e-01 5.14803410e-01 -2.19437689e-01 2.09369510e-01
-4.90767002e-01 -3.52191269e-01 -9.16762114e-01 -5.82241893e-01
-2.19363477e-02 6.86539829e-01 2.04004511e-01 -6.04598178e-03
5.62158644e-01 1.06530368e+00 -1.90685856e+00 -2.41093233e-01
4.39297736e-01 1.15248096e+00 6.72018468e-01 2.25561395e-01
1.71315789e-01 1.98193982e-01 -3.11614782e-01 -1.83786854e-01
4.56175596e-01 2.95895725e-01 -5.77081859e-01 2.09711984e-01
7.26233125e-01 -5.48566401e-01 -4.92124230e-01 9.63992953e-01
5.94233096e-01 1.64980665e-02 5.44709444e-01 -1.48922062e+00
-4.46650207e-01 2.52020270e-01 -4.63018209e-01 -1.05087116e-01
4.02890116e-01 8.39864835e-02 8.09814095e-01 -9.82634187e-01
9.09121215e-01 1.14435744e+00 7.39692807e-01 3.86120141e-01
-7.38031447e-01 -3.98875743e-01 4.43032265e-01 5.23164451e-01
-9.87830579e-01 -2.56703496e-01 9.43515897e-01 -5.73584616e-01
7.88855672e-01 -3.44813289e-03 7.19637096e-01 9.62939084e-01
3.66629273e-01 1.17475176e+00 8.50733936e-01 -3.40489745e-01
1.84113801e-01 -1.71885595e-01 5.38445041e-02 4.76233453e-01
8.17549169e-01 1.21547580e-01 -8.08583081e-01 2.22853214e-01
9.89645481e-01 2.17277616e-01 -2.16544673e-01 -1.32544434e+00
-1.63297200e+00 5.90637684e-01 6.42333567e-01 1.72106117e-01
-4.70541120e-01 3.00490737e-01 2.14756191e-01 1.07862540e-02
1.10788353e-01 8.29097390e-01 -5.13595521e-01 -2.90848106e-01
-3.12146485e-01 7.51966238e-01 9.11640227e-01 1.50589883e+00
5.41393220e-01 -8.15802291e-02 2.90378541e-01 6.58843577e-01
3.08324158e-01 5.84444821e-01 -4.47087223e-03 -1.26632655e+00
2.64014781e-01 5.07977724e-01 5.02954483e-01 -9.24966037e-01
-5.88276684e-01 -4.92089480e-01 -3.26727480e-01 3.36817384e-01
5.07295072e-01 3.46759856e-02 -7.29822636e-01 1.28628004e+00
5.30702055e-01 -3.70744467e-01 -4.14285779e-01 1.12881243e+00
8.71459126e-01 6.77816570e-02 -3.52420300e-01 1.65604260e-02
1.21621311e+00 -1.00960934e+00 -8.09311211e-01 -5.01990139e-01
2.23538250e-01 -1.03897178e+00 7.96273410e-01 5.49373806e-01
-9.67898071e-01 -3.77088755e-01 -1.15546227e+00 -3.06754500e-01
-4.63729858e-01 2.27024108e-01 1.11572075e+00 2.24567682e-01
-5.04200101e-01 4.88723546e-01 -9.37826633e-01 -6.36521220e-01
4.49761182e-01 4.62199241e-01 -3.04782152e-01 -4.85604584e-01
-6.98030055e-01 1.64482307e+00 3.85977507e-01 2.66626209e-01
-6.40895605e-01 -7.51507282e-01 -8.45405221e-01 -5.86066842e-01
8.99634182e-01 -8.26314509e-01 1.40661108e+00 9.81909484e-02
-1.63447177e+00 9.06636715e-01 7.09330738e-02 -3.02560776e-01
6.58749044e-01 -7.39199460e-01 2.86424249e-01 6.48674443e-02
6.62312657e-02 7.62175024e-01 7.28245974e-01 -1.32412207e+00
-7.24466324e-01 -7.30475187e-01 2.29064971e-01 5.05349100e-01
-2.63150949e-02 -1.41637281e-01 -4.07831043e-01 -2.83486217e-01
6.89235032e-01 -1.22569394e+00 -1.39954761e-01 6.22427464e-01
-1.92625806e-01 -4.28745717e-01 9.34514284e-01 -2.60965884e-01
4.88672614e-01 -1.90023589e+00 3.44877124e-01 -4.48006213e-01
1.38159469e-01 1.00323595e-01 2.39205793e-01 4.58907664e-01
4.10668641e-01 -5.11873662e-01 4.91195358e-02 -4.27752398e-02
3.35372657e-01 2.22779438e-01 -2.15575546e-01 9.41640019e-01
8.03860426e-02 1.14325452e+00 -9.69611466e-01 -1.37909666e-01
7.75768995e-01 3.60777259e-01 -3.54159713e-01 1.17777452e-01
-3.30934048e-01 4.94816631e-01 -4.77546632e-01 9.98680055e-01
7.23649025e-01 6.41278923e-03 -1.34797066e-01 -5.93574107e-01
-3.85045618e-01 2.87957400e-01 -1.52887642e+00 1.90430152e+00
-2.61282176e-01 7.09324956e-01 4.22014624e-01 -1.05809247e+00
8.83019447e-01 4.65232097e-02 7.69440889e-01 -1.84111029e-01
4.42975610e-01 4.63349134e-01 -7.51654208e-02 -6.53938293e-01
7.38792360e-01 -3.52943386e-03 -3.66833364e-03 5.75387031e-02
-9.45003703e-02 -9.71188426e-01 8.18512216e-02 -4.21253115e-01
6.43394113e-01 6.16723955e-01 5.73556542e-01 -8.69245604e-02
3.39230224e-02 3.62620234e-01 1.57381147e-01 8.19094896e-01
-4.87085313e-01 4.20708179e-01 -1.28738850e-01 -5.03946364e-01
-1.01165760e+00 -1.14259946e+00 -4.56646651e-01 7.92138040e-01
7.40061402e-01 1.98306307e-01 -2.85606027e-01 -1.92833126e-01
6.20698333e-01 1.40864447e-01 -2.59768903e-01 4.68087159e-02
-7.54211426e-01 -4.61726308e-01 -3.10388505e-01 7.93523729e-01
2.98569232e-01 -1.04005659e+00 -9.99529660e-01 2.08989769e-01
-1.06863528e-01 -1.53888512e+00 2.60485947e-01 2.05389142e-01
-1.06987274e+00 -1.27682555e+00 -9.21121240e-01 -9.00336623e-01
4.12665904e-01 8.26559126e-01 9.01529610e-01 -4.48878288e-01
-5.38989544e-01 7.33339965e-01 -3.83690119e-01 -9.00663614e-01
1.24773808e-01 2.11845919e-01 4.43189532e-01 -6.97213233e-01
4.34732616e-01 -5.21877229e-01 -5.26501060e-01 5.11137247e-01
-5.86280882e-01 -2.09317565e-01 8.04941893e-01 4.65585649e-01
4.99085248e-01 -3.76913100e-01 4.07452136e-01 -2.39594698e-01
1.35457695e-01 -2.53184438e-01 -8.24062765e-01 -1.13801822e-01
-2.58655727e-01 -3.47537428e-01 5.42564725e-04 -4.89776254e-01
-7.81343997e-01 3.65824819e-01 2.11638466e-01 -3.59670609e-01
-4.72268701e-01 4.81096983e-01 -1.67640790e-01 -5.30044138e-01
6.35735989e-01 -6.52302727e-02 1.00095138e-01 -6.14574075e-01
5.18418193e-01 6.03747904e-01 6.11963987e-01 -5.67464769e-01
7.81222582e-01 8.14228654e-01 3.63215655e-01 -1.05974030e+00
-1.00263572e+00 -7.64145195e-01 -1.00821483e+00 -3.61512303e-01
5.09676397e-01 -8.07061434e-01 -9.71721709e-01 4.83171076e-01
-1.33128560e+00 4.16699909e-02 -2.84864008e-01 7.76541114e-01
-9.60764468e-01 2.60894239e-01 -4.87976313e-01 -9.63629067e-01
2.20110714e-02 -1.34587419e+00 1.41342580e+00 1.97452620e-01
-2.09692881e-01 -5.34761548e-01 -3.00435781e-01 4.42676723e-01
1.83985487e-01 3.89440209e-01 4.04908240e-01 -1.62357867e-01
-1.17259049e+00 -5.06142676e-01 -2.70348579e-01 -4.59324494e-02
1.26870871e-01 -2.08291039e-01 -8.83178115e-01 -3.20972145e-01
3.78937498e-02 -3.47937107e-01 5.29163539e-01 6.37581289e-01
1.02162480e+00 4.69139069e-01 -7.18167484e-01 3.73768806e-01
1.12663651e+00 1.39184266e-01 2.47475281e-01 6.78312600e-01
7.24915385e-01 8.11532080e-01 1.13585746e+00 2.44508356e-01
3.45595717e-01 1.12526953e+00 9.40038562e-01 2.64675796e-01
-1.11049637e-01 6.19306751e-02 -1.18117325e-01 6.91799343e-01
-1.03037313e-01 2.10431874e-01 -8.71998966e-01 5.13664722e-01
-1.92103767e+00 -5.62850654e-01 -2.67697841e-01 2.06814861e+00
4.05142039e-01 1.09545432e-01 -6.24427162e-02 2.56015122e-01
5.77337325e-01 -4.35020849e-02 -7.94528902e-01 1.93659037e-01
1.13458589e-01 -3.79599072e-02 6.22003615e-01 2.09176525e-01
-1.34874368e+00 1.09595025e+00 6.65878534e+00 3.78303051e-01
-1.21906435e+00 -1.59974948e-01 -3.12975317e-01 -5.50741702e-02
3.31125230e-01 -2.67320544e-01 -1.15004134e+00 3.78374159e-02
4.99477312e-02 -2.52379272e-02 3.02820325e-01 1.35636461e+00
-1.30122066e-01 -4.89464313e-01 -1.48769307e+00 1.35869658e+00
3.31683815e-01 -1.10840142e+00 -4.00217235e-01 4.72755991e-02
7.97657311e-01 2.80896574e-01 1.71768233e-01 4.58689295e-02
2.82968339e-02 -6.09232664e-01 7.73601055e-01 2.69532353e-01
2.70299107e-01 -2.22040132e-01 6.64470196e-01 4.71864730e-01
-1.04579973e+00 -2.71243364e-01 -6.97966874e-01 -5.81889153e-01
4.02090102e-01 7.23859251e-01 -9.74746168e-01 6.75911009e-01
7.28670776e-01 9.19163227e-01 -2.00930871e-02 1.71608520e+00
-3.01103294e-01 -1.28668368e-01 -5.09979904e-01 -4.13980603e-01
-7.78552890e-02 3.01968008e-02 9.28451002e-01 6.09876215e-01
1.39021456e-01 -2.48210177e-01 4.70693201e-01 7.81722486e-01
1.93968266e-02 -1.92835286e-01 -8.43946159e-01 6.15534894e-02
4.50269789e-01 1.17012012e+00 -8.54203761e-01 8.33210051e-02
-3.45741689e-01 5.87748826e-01 2.13540450e-01 1.02269605e-01
-4.58884984e-01 -2.20510572e-01 7.71149755e-01 3.04563642e-02
3.93070042e-01 -9.29581225e-01 -7.23918140e-01 -1.06936657e+00
3.98534715e-01 -5.67134082e-01 -1.54649317e-01 -8.43505263e-01
-1.21554732e+00 -4.92658876e-02 3.58567894e-01 -1.49873030e+00
-1.40970618e-01 -1.26433265e+00 2.31278911e-01 3.76552403e-01
-1.84532201e+00 -1.11628425e+00 -6.01228774e-01 -4.52033766e-02
8.16068232e-01 1.56666160e-01 6.61290884e-01 2.71864712e-01
-5.39889261e-02 9.69063044e-02 2.32946016e-02 -3.08200568e-01
6.35509670e-01 -8.77663136e-01 2.46837452e-01 6.01100445e-01
1.24354772e-02 6.80184960e-01 9.36391592e-01 -6.47909701e-01
-2.26023126e+00 -6.81054711e-01 6.01679862e-01 -9.25127685e-01
5.78256071e-01 -3.98257375e-01 -4.68600303e-01 7.06698418e-01
-5.14220119e-01 9.59765241e-02 1.24395750e-01 1.31011441e-01
5.76129854e-02 2.85322312e-02 -9.64865744e-01 6.48245454e-01
1.35504901e+00 -2.06552222e-01 -8.80608082e-01 4.83908147e-01
5.56838095e-01 -1.08769202e+00 -7.63124287e-01 8.62976074e-01
9.76105928e-01 -6.43743753e-01 1.14640605e+00 -2.85576910e-01
2.55920947e-01 -3.03129941e-01 -2.82836348e-01 -9.47140396e-01
-2.45561481e-01 -4.30701792e-01 -5.61987996e-01 5.67436039e-01
-9.60838571e-02 -5.88660240e-01 1.09870720e+00 5.50785005e-01
-5.48049629e-01 -9.00499880e-01 -8.99051249e-01 -1.03905225e+00
-1.49907172e-01 -4.87690538e-01 3.12968999e-01 6.69325292e-01
-5.66997007e-02 1.03963196e-01 -2.33339697e-01 2.13456139e-01
6.26640856e-01 5.49234152e-01 1.29075861e+00 -1.76798975e+00
3.06889236e-01 -4.27956432e-01 -6.30842388e-01 -1.67391598e+00
8.98937583e-02 -5.16952813e-01 3.49505574e-01 -1.79705536e+00
1.44882435e-02 -3.12130302e-01 1.40321791e-01 7.48568401e-02
1.16327144e-01 4.73516345e-01 2.80451179e-01 1.58139929e-01
-5.86025357e-01 5.91396391e-01 1.68111658e+00 5.16472459e-02
7.88029842e-03 2.88262844e-01 -4.98896986e-01 9.78368640e-01
4.50564981e-01 -1.04911096e-01 -8.36742222e-02 -6.17667973e-01
2.37749368e-01 -2.58113593e-01 4.73122180e-01 -1.07609582e+00
3.46466362e-01 -2.98818022e-01 3.68806541e-01 -9.72504854e-01
8.57973099e-01 -1.19761884e+00 -2.41333127e-01 5.16891062e-01
3.44012529e-02 -2.03192607e-01 1.46220222e-01 6.21213317e-01
3.96458358e-02 -3.30852419e-01 6.18286550e-01 -4.42494899e-01
-1.13115394e+00 5.04781723e-01 -6.79755732e-02 -1.92590758e-01
1.22188246e+00 -7.24435866e-01 -1.36057645e-01 -2.43973702e-01
-5.59408784e-01 1.33081242e-01 4.67505425e-01 7.81549633e-01
5.71310341e-01 -1.23717844e+00 -2.97945410e-01 -6.55469149e-02
3.53864580e-01 4.07418370e-01 2.45270059e-02 8.09402764e-01
-6.10226035e-01 7.42861509e-01 -3.29142004e-01 -1.03826165e+00
-9.34114873e-01 4.91568297e-01 -4.75939512e-02 3.56054246e-01
-6.87071264e-01 8.85607421e-01 -1.02208570e-01 -6.29129887e-01
7.57150292e-01 -4.70149636e-01 -1.46446098e-02 -1.71685964e-01
2.06029847e-01 8.46580446e-01 2.32236356e-01 -4.65471834e-01
-4.17121291e-01 9.54379737e-01 1.57651126e-01 2.38831818e-01
1.61016476e+00 -1.77784950e-01 -8.65283385e-02 6.50326967e-01
8.28958511e-01 -4.17867601e-01 -1.39492369e+00 -2.66347796e-01
6.47177175e-03 -7.32348680e-01 -4.70352285e-02 -5.61468244e-01
-6.13743186e-01 9.57555830e-01 5.57870924e-01 -5.73400892e-02
5.15436709e-01 2.75998682e-01 6.63989663e-01 6.98733151e-01
1.00572991e+00 -1.12380588e+00 3.46893787e-01 8.91466737e-01
1.24375248e+00 -1.45153022e+00 4.86875117e-01 -9.86862183e-01
3.86806508e-03 1.15227759e+00 7.91599095e-01 -2.94137269e-01
7.12479055e-01 3.48742932e-01 -1.49455249e-01 -2.99366534e-01
5.72905429e-02 -2.71984190e-01 4.35094386e-01 9.50297356e-01
2.42385328e-01 -1.51398452e-02 -2.48618647e-01 6.14362359e-02
-3.89375806e-01 4.61670868e-02 1.54962897e-01 1.43886518e+00
-7.39073694e-01 -9.28920388e-01 -5.28811336e-01 3.81838828e-01
-4.99574430e-02 4.45544153e-01 -3.68586421e-01 1.02136612e+00
-6.59111748e-03 7.29686797e-01 -9.05564204e-02 -2.27889437e-02
5.70538640e-01 -1.78028122e-01 1.17583609e+00 -7.25010395e-01
7.58048939e-03 -1.56682670e-01 -1.22594684e-01 -6.39559448e-01
-5.87148070e-01 -7.68583059e-01 -7.89275110e-01 2.18837261e-02
-7.37792909e-01 -4.69527483e-01 1.25699031e+00 1.19758213e+00
2.55138576e-01 4.05449241e-01 1.13829598e-01 -1.79818690e+00
-8.45537901e-01 -8.07737112e-01 -3.79270226e-01 1.06769122e-01
2.38457024e-01 -1.51602960e+00 3.43841165e-02 -1.43624291e-01] | [7.216326713562012, -2.3118128776550293] |
951077f3-2981-46d6-8dfc-cf72ca94497e | adaptive-real-time-exploration-and | 2211.05495 | null | https://arxiv.org/abs/2211.05495v2 | https://arxiv.org/pdf/2211.05495v2.pdf | Safe and Adaptive Decision-Making for Optimization of Safety-Critical Systems: The ARTEO Algorithm | We consider the problem of decision-making under uncertainty in an environment with safety constraints. Many business and industrial applications rely on real-time optimization to improve key performance indicators. In the case of unknown characteristics, real-time optimization becomes challenging, particularly because of the satisfaction of safety constraints. We propose the ARTEO algorithm, where we cast multi-armed bandits as a mathematical programming problem subject to safety constraints and learn the unknown characteristics through exploration while optimizing the targets. We quantify the uncertainty in unknown characteristics by using Gaussian processes and incorporate it into the cost function as a contribution which drives exploration. We adaptively control the size of this contribution in accordance with the requirements of the environment. We guarantee the safety of our algorithm with a high probability through confidence bounds constructed under the regularity assumptions of Gaussian processes. We demonstrate the safety and efficiency of our approach with two case studies: optimization of electric motor current and real-time bidding problems. We further evaluate the performance of ARTEO compared to a safe variant of upper confidence bound based algorithms. ARTEO achieves less cumulative regret with accurate and safe decisions. | ['Mehmet Mercangöz', 'Marta Zagórowska', 'Buse Sibel Korkmaz'] | 2022-11-10 | null | null | null | null | ['decision-making-under-uncertainty', 'decision-making-under-uncertainty'] | ['medical', 'reasoning'] | [ 4.10400741e-02 3.21053684e-01 -2.37914637e-01 -1.97495237e-01
-1.07565093e+00 -7.73687184e-01 2.31830180e-01 3.53066236e-01
-5.66550016e-01 1.13187444e+00 -1.84012592e-01 -5.34770906e-01
-9.69007015e-01 -7.95027137e-01 -1.02284026e+00 -9.32704628e-01
-2.46561110e-01 7.99264371e-01 -2.99328089e-01 7.88305048e-03
2.74681270e-01 3.95298749e-01 -1.05287862e+00 -3.36146772e-01
9.62218106e-01 1.50118744e+00 1.54821789e-02 4.06029820e-01
3.06738675e-01 3.51641595e-01 -5.01797438e-01 -4.62518603e-01
6.31306231e-01 1.85948789e-01 -3.77235621e-01 1.18007272e-01
-6.71016216e-01 -1.35997444e-01 3.30378115e-01 1.24681473e+00
3.97466600e-01 3.58121604e-01 6.54704809e-01 -1.51756287e+00
-3.38712409e-02 6.86844647e-01 -7.61413157e-01 8.02285373e-02
-1.51154935e-01 1.56980693e-01 9.48060155e-01 -1.88951507e-01
2.86077484e-02 1.15925860e+00 2.51103759e-01 3.96175534e-01
-1.43137944e+00 -5.59084117e-01 4.72039759e-01 -7.33056366e-02
-1.06121421e+00 -2.35024542e-01 4.27230239e-01 -4.01092708e-01
3.76353681e-01 3.64331245e-01 4.27878797e-01 8.95501494e-01
3.59915406e-01 8.20558906e-01 1.00745201e+00 -2.93118984e-01
8.06854963e-01 2.75300980e-01 6.19510980e-03 1.67394802e-02
6.45754337e-01 5.22321522e-01 -3.44763726e-01 -4.27819788e-01
2.41920099e-01 -5.89825958e-02 -1.84162661e-01 -6.08027160e-01
-6.45894170e-01 1.11416852e+00 -1.47538975e-01 -2.69528627e-01
-7.60239840e-01 4.11971837e-01 1.59007937e-01 2.20009550e-01
6.73887253e-01 4.31164473e-01 -4.82517242e-01 -1.65063024e-01
-5.11051178e-01 2.30499730e-01 9.08865750e-01 1.04295075e+00
6.77897260e-02 6.45731250e-03 -5.71691096e-01 4.42500442e-01
2.68216521e-01 4.73189563e-01 -1.33346528e-01 -8.26078236e-01
8.21515679e-01 -2.26189062e-01 1.07296693e+00 -4.97545570e-01
-2.35557452e-01 -7.62013316e-01 -2.57646829e-01 5.80382943e-01
5.03381491e-01 -5.64901769e-01 -7.09887445e-01 1.80546701e+00
4.08465743e-01 -1.43318862e-01 -1.74670871e-02 6.02512836e-01
-7.35785127e-01 7.47565866e-01 -4.71573230e-03 -8.03832889e-01
9.13324594e-01 -4.72849518e-01 -1.01653004e+00 5.74360415e-02
3.23515296e-01 -2.84943074e-01 4.74320620e-01 9.17144835e-01
-1.17755449e+00 1.39434680e-01 -9.75132287e-01 8.84946764e-01
1.12613678e-01 -1.87297404e-01 4.18650210e-01 9.68040764e-01
-3.97591770e-01 6.13780618e-01 -7.54375100e-01 4.23243731e-01
5.79645395e-01 3.98609310e-01 2.00121060e-01 3.48553926e-01
-1.08935070e+00 9.76427436e-01 5.43243468e-01 3.79927516e-01
-1.19468355e+00 -9.63172257e-01 -6.52293682e-01 2.15022579e-01
1.17422330e+00 -4.75106210e-01 1.45227885e+00 -6.02300048e-01
-1.53060162e+00 1.43288702e-01 3.60286891e-01 -9.65530813e-01
1.08384860e+00 -4.23429012e-01 2.67558210e-02 -6.54135644e-02
4.33183201e-02 -2.13473752e-01 9.48351502e-01 -1.22271812e+00
-1.06855714e+00 -4.38783765e-01 2.34248698e-01 1.60254911e-01
5.11684231e-02 -2.30764225e-01 1.66068256e-01 -4.45799291e-01
-3.64435911e-01 -9.52869773e-01 -6.95051908e-01 -3.44677299e-01
-3.50856960e-01 -9.60384160e-02 2.13842303e-01 -4.91507530e-01
1.08239269e+00 -1.97217894e+00 -8.19749385e-02 6.23986959e-01
-2.82208413e-01 -2.89410770e-01 3.74839604e-01 2.98274994e-01
1.65486678e-01 1.34425700e-01 -2.73292005e-01 -2.41584212e-01
4.41129506e-01 3.40737373e-01 -4.80881304e-01 8.58439744e-01
1.22943193e-01 5.16799092e-01 -7.33959317e-01 3.92786637e-02
1.52301742e-02 -2.86194235e-01 -2.45680884e-01 1.65523425e-01
-6.54117942e-01 1.67356104e-01 -9.02960837e-01 4.19867426e-01
4.05879110e-01 1.09306283e-01 -5.69709130e-02 2.85073251e-01
-2.01928660e-01 -4.34414148e-01 -1.34849143e+00 1.09312844e+00
-8.31598759e-01 -1.28632128e-01 5.59694827e-01 -1.31606781e+00
5.10811210e-01 1.74975261e-01 4.19172168e-01 -2.67753661e-01
3.36535662e-01 1.47297099e-01 -2.13461190e-01 -2.40397111e-01
2.13489428e-01 -5.34893155e-01 -3.35672200e-01 3.97293150e-01
-3.16505641e-01 -3.25783461e-01 -2.01691426e-02 -1.13377109e-01
8.12900186e-01 2.87297498e-02 2.60735899e-01 -6.89664245e-01
7.19060227e-02 -2.05501348e-01 7.44860888e-01 9.52498853e-01
6.71912506e-02 -1.46089320e-03 9.90076303e-01 3.19602415e-02
-8.85229826e-01 -8.38755846e-01 -1.51283607e-01 8.01081538e-01
1.38626555e-02 3.59033287e-01 -2.74550140e-01 -6.58647001e-01
7.13211894e-01 1.48673236e+00 -1.02567589e+00 -6.63141757e-02
5.55396006e-02 -1.04174030e+00 -2.49810860e-01 3.54384720e-01
-3.75443548e-02 -3.06638092e-01 -9.29771841e-01 4.50823665e-01
4.10582721e-01 -8.49163115e-01 -3.69636208e-01 5.78213692e-01
-4.03667569e-01 -8.58915091e-01 -5.31461716e-01 2.24157155e-01
5.89534283e-01 -3.56983662e-01 6.66649163e-01 -6.74171507e-01
-1.19155847e-01 4.99342918e-01 -2.78426576e-02 -1.24518692e+00
-2.74263263e-01 -4.53967482e-01 9.31569189e-02 3.58517319e-01
-3.12896878e-01 -3.19782943e-01 -4.08740669e-01 2.90428370e-01
-6.80588603e-01 -3.16466719e-01 3.77489835e-01 8.85261893e-01
6.67492807e-01 6.74571455e-01 8.56546760e-01 -7.97508955e-01
8.67370069e-01 -5.85633337e-01 -1.64115214e+00 5.09146094e-01
-7.96404421e-01 4.46512401e-01 2.77847290e-01 -5.57575285e-01
-1.29563010e+00 1.17902368e-01 5.90010941e-01 -3.85039359e-01
4.44641262e-01 5.55128634e-01 -3.54018182e-01 7.72203729e-02
1.55414909e-01 -2.17350796e-01 5.82944974e-02 -2.89034754e-01
2.89998114e-01 3.91178161e-01 2.72899359e-01 -1.29263067e+00
6.83070898e-01 5.40524364e-01 5.30389249e-01 -1.62538543e-01
-1.13087893e+00 -2.51969218e-01 1.00031801e-01 -2.60648310e-01
4.93931204e-01 -6.75279021e-01 -1.24298465e+00 8.41713324e-03
-7.57220268e-01 -1.44278944e-01 -6.57194495e-01 6.22790337e-01
-1.23399448e+00 3.77974547e-02 -6.30194396e-02 -1.81499815e+00
-2.14213029e-01 -1.01391053e+00 6.78125143e-01 -1.74753200e-02
1.18745103e-01 -7.88300812e-01 -2.24017829e-01 2.30251998e-01
1.55409351e-01 6.41669691e-01 7.41545379e-01 -6.76734149e-01
-5.28505564e-01 -1.27139613e-01 1.69264570e-01 3.14913541e-01
-1.68201640e-01 -3.92435253e-01 -5.39821029e-01 -2.67917395e-01
5.13463616e-01 -1.86314955e-01 4.84996438e-01 6.50333107e-01
1.39419663e+00 -6.35112166e-01 -3.58496428e-01 3.16659391e-01
1.47927213e+00 6.81064427e-01 9.45428107e-03 6.55703604e-01
-9.75887999e-02 9.54642117e-01 1.23553574e+00 1.25600779e+00
-1.38020664e-01 5.18512666e-01 8.95423234e-01 6.02493286e-01
1.17849147e+00 -7.56038204e-02 6.85651079e-02 -3.78158689e-01
8.17447528e-02 -4.06146824e-01 -7.80547261e-01 7.95704246e-01
-2.30624580e+00 -7.79537737e-01 2.15694502e-01 2.65661716e+00
9.34650779e-01 5.38102508e-01 2.28494465e-01 -3.42274234e-02
8.27780366e-01 -5.09205401e-01 -8.44280720e-01 -6.76270306e-01
6.47673309e-02 -5.39362207e-02 1.33311248e+00 5.04245579e-01
-9.16581869e-01 3.58554423e-02 5.60175610e+00 1.17439055e+00
-5.68410695e-01 1.28383160e-01 1.18785512e+00 -6.63621128e-01
-4.04386461e-01 -1.52451366e-01 -9.22954381e-01 6.60613179e-01
9.93808150e-01 -7.67564535e-01 4.33632463e-01 9.02030230e-01
5.25946498e-01 -3.99675220e-01 -1.15620172e+00 7.08432257e-01
-6.00130081e-01 -1.09687805e+00 -6.56247914e-01 3.10545117e-01
9.58430588e-01 -3.20800841e-01 2.78205395e-01 1.63239524e-01
7.84029782e-01 -9.84543502e-01 1.31187665e+00 6.67842567e-01
2.95310050e-01 -1.45544600e+00 7.26411760e-01 5.44786513e-01
-5.40288031e-01 -7.42960751e-01 -2.47085299e-02 1.35588408e-01
6.18536413e-01 9.38569248e-01 -5.15749335e-01 6.58027232e-01
4.04768944e-01 -7.22097605e-02 4.42499310e-01 1.28022754e+00
-2.46576205e-01 4.94181097e-01 -7.14038908e-01 -3.14348578e-01
5.21739721e-01 -5.62792361e-01 6.52234018e-01 6.48311913e-01
4.65280622e-01 6.28206879e-02 2.12861329e-01 1.03399849e+00
1.93746492e-01 -9.36701745e-02 -2.66056448e-01 9.40575497e-04
3.59916836e-01 8.78903508e-01 -5.21198153e-01 -5.74175715e-02
-1.27856195e-01 2.80651152e-01 -4.23872285e-02 2.99191207e-01
-1.01186657e+00 -1.69892967e-01 5.63159883e-01 -2.78345913e-01
4.55355585e-01 6.32109568e-02 -6.05794072e-01 -3.93996209e-01
3.03132772e-01 -6.01867437e-01 6.73486114e-01 -3.46619070e-01
-1.30194581e+00 -3.32571492e-02 4.52682495e-01 -7.28959262e-01
-2.29036003e-01 -5.11586130e-01 -4.69176590e-01 1.01712537e+00
-1.32161200e+00 -6.92460358e-01 6.68471754e-01 3.13869298e-01
4.93770331e-01 1.18707843e-01 1.37262821e-01 -3.34064811e-01
-6.44541800e-01 5.13404571e-02 7.09409595e-01 -6.37472570e-01
1.89122543e-01 -1.41028523e+00 -1.38052702e-01 6.54035985e-01
-3.96456182e-01 2.51509130e-01 1.25032473e+00 -6.38717532e-01
-1.26690996e+00 -9.70378935e-01 1.54620320e-01 -3.44885021e-01
1.26262701e+00 -3.39894116e-01 -4.39641833e-01 3.61987144e-01
-1.04579091e-01 5.28022763e-04 2.88102984e-01 1.33528978e-01
1.25770450e-01 -2.75143296e-01 -1.37834656e+00 2.71719158e-01
6.30933166e-01 3.62601131e-02 -4.66230541e-01 4.61866140e-01
8.29474270e-01 -2.28521958e-01 -8.79144371e-01 5.18490374e-01
4.16945279e-01 -1.60267293e-01 5.71896672e-01 -8.73189807e-01
-2.73112893e-01 -5.92814088e-02 -1.53605700e-01 -1.58007646e+00
-6.04933687e-02 -1.37222803e+00 -2.03437060e-01 9.97270346e-01
6.53665364e-01 -7.90176153e-01 6.64576828e-01 9.94566917e-01
1.30412847e-01 -7.74430156e-01 -1.44473672e+00 -1.19870472e+00
1.06384322e-01 -6.37098968e-01 6.56891167e-01 3.73813838e-01
9.46241617e-02 -2.10738540e-01 -5.33204019e-01 4.72306103e-01
1.14710510e+00 2.56303042e-01 1.96432486e-01 -8.57190967e-01
-5.75970650e-01 -2.66305119e-01 1.25645980e-01 -4.24110800e-01
3.26789677e-01 -7.32455403e-02 4.89182353e-01 -9.16380286e-01
2.90682111e-02 -6.50135577e-01 -4.90041882e-01 1.74735337e-01
-7.17350617e-02 -5.77540040e-01 2.64981270e-01 -4.07041907e-01
-6.05849683e-01 7.23838329e-01 9.01682913e-01 -2.26040378e-01
-1.94599181e-01 8.64979625e-01 -7.82565475e-01 4.99591440e-01
9.51030850e-01 -4.67454016e-01 -6.14135623e-01 -3.06810811e-03
6.06595218e-01 4.93005842e-01 1.54625118e-01 -4.56391782e-01
-2.70818491e-02 -6.90972984e-01 -5.46412691e-02 -7.30798542e-01
3.29425931e-01 -1.21859431e+00 4.65749323e-01 6.68638587e-01
-5.08907795e-01 -9.79177654e-02 1.69634432e-01 1.26939487e+00
2.57002860e-01 -5.42931259e-01 6.82168543e-01 3.23249772e-02
-2.87771493e-01 3.00782800e-01 -3.63488257e-01 -1.49205789e-01
1.47679651e+00 2.98758268e-01 -1.82218514e-02 -7.75986016e-01
-1.02161932e+00 9.16296422e-01 -1.45813927e-01 1.67221591e-01
2.77797461e-01 -8.60750854e-01 -5.54640889e-01 -2.85291344e-01
-1.17656991e-01 1.68386754e-02 1.50305986e-01 7.90036261e-01
1.40649423e-01 5.53426921e-01 1.78466231e-01 -3.28161776e-01
-9.13197637e-01 9.67598081e-01 2.53630400e-01 -3.54603469e-01
2.85630934e-02 5.69388509e-01 3.24718922e-01 3.86250317e-01
3.90564650e-01 -2.91803718e-01 1.11548662e-01 1.16429076e-01
3.83481860e-01 6.66514814e-01 1.06571965e-01 8.87015685e-02
-2.48192810e-02 -5.44825420e-02 1.72813490e-01 -5.08711696e-01
1.41851830e+00 -4.31823790e-01 2.17131644e-01 3.53306979e-01
5.32462716e-01 -1.75997838e-02 -1.55962384e+00 2.84693111e-02
4.55945849e-01 -5.17310619e-01 2.08368883e-01 -1.00630653e+00
-7.55290747e-01 5.10251582e-01 3.43945295e-01 3.20465356e-01
9.12623644e-01 -2.66445756e-01 1.55548409e-01 2.59470165e-01
7.33061254e-01 -1.41812074e+00 -3.82894397e-01 6.26007766e-02
9.52920198e-01 -8.63791943e-01 -1.13822401e-01 -2.64337659e-01
-7.39942789e-01 7.51317620e-01 1.87747762e-01 1.29478676e-02
6.46554232e-01 3.80119294e-01 -4.67064768e-01 3.48391324e-01
-1.01992536e+00 -1.52404591e-01 1.62331671e-01 4.01878774e-01
-5.11580408e-01 4.47559834e-01 -6.37200713e-01 9.67573762e-01
4.06190492e-02 -2.00214729e-01 4.53336298e-01 9.84927952e-01
-5.11961281e-01 -8.38695824e-01 -7.54338145e-01 5.13899386e-01
-1.01217866e+00 2.50423014e-01 3.10146481e-01 7.25286543e-01
-1.65860295e-01 1.08429468e+00 -2.68723309e-01 3.36638272e-01
4.11015987e-01 -1.13462381e-01 3.66420716e-01 -4.81416255e-01
-1.84999257e-01 2.81732619e-01 5.25689304e-01 -4.77781326e-01
2.03527641e-02 -7.72386551e-01 -7.46143639e-01 7.23315179e-02
-7.42200255e-01 7.41212487e-01 1.02869225e+00 9.46845531e-01
-5.52785583e-03 4.60126579e-01 9.61525023e-01 -3.41330260e-01
-1.94428360e+00 -5.35782158e-01 -1.02217031e+00 -2.21756659e-02
3.42767775e-01 -7.57912815e-01 -5.37069321e-01 -5.17831028e-01] | [4.567654609680176, 3.1244962215423584] |
74c8282e-d975-4411-bde5-04cc954058f2 | simco-similarity-based-object-counting | 1904.07092 | null | https://arxiv.org/abs/1904.07092v2 | https://arxiv.org/pdf/1904.07092v2.pdf | SIMCO: SIMilarity-based object COunting | We present SIMCO, the first agnostic multi-class object counting approach. SIMCO starts by detecting foreground objects through a novel Mask RCNN-based architecture trained beforehand (just once) on a brand-new synthetic 2D shape dataset, InShape; the idea is to highlight every object resembling a primitive 2D shape (circle, square, rectangle, etc.). Each object detected is described by a low-dimensional embedding, obtained from a novel similarity-based head branch; this latter implements a triplet loss, encouraging similar objects (same 2D shape + color and scale) to map close. Subsequently, SIMCO uses this embedding for clustering, so that different types of objects can emerge and be counted, making SIMCO the very first multi-class unsupervised counter. Experiments show that SIMCO provides state-of-the-art scores on counting benchmarks and that it can also help in many challenging image understanding tasks. | ['Andrea Giachetti', 'Marco Godi', 'Marco Cristani', 'Christian Joppi'] | 2019-04-15 | null | null | null | null | ['object-counting'] | ['computer-vision'] | [ 3.16692173e-01 -7.25800470e-02 1.04583167e-01 4.07638997e-02
-3.23224813e-01 -6.89986885e-01 1.06732869e+00 3.73565018e-01
-7.74189174e-01 2.47854561e-01 -1.28435582e-01 -1.62132122e-02
3.99421781e-01 -7.65034914e-01 -8.06843340e-01 -7.43947089e-01
-1.25683621e-01 1.12107742e+00 9.14046466e-01 1.23184569e-01
5.54100633e-01 8.13936114e-01 -1.77598834e+00 3.96133810e-01
1.68916121e-01 6.85063064e-01 5.79222813e-02 9.04440939e-01
-2.71808475e-01 9.20854926e-01 -8.62281144e-01 -7.40538239e-01
3.15324396e-01 -1.76284432e-01 -6.14118099e-01 -7.32240975e-02
8.60288680e-01 -2.33931169e-01 -1.07161496e-02 1.09868979e+00
4.10135180e-01 -1.22286446e-01 1.23728621e+00 -1.13673615e+00
-6.00305676e-01 5.77234268e-01 -1.03676128e+00 5.10474384e-01
1.39668480e-01 1.60285294e-01 8.62011075e-01 -1.16717339e+00
5.22121966e-01 1.57101965e+00 7.34380782e-01 6.85269177e-01
-1.50811982e+00 -4.96285766e-01 -3.75195555e-02 -7.47083798e-02
-1.27157533e+00 8.44702795e-02 4.73660439e-01 -4.58118051e-01
7.10426033e-01 3.12456280e-01 8.18876624e-01 1.15379620e+00
-2.62456447e-01 1.23061180e+00 1.13444400e+00 -4.69615161e-01
3.48836452e-01 1.55289888e-01 2.04702884e-01 9.39132750e-01
6.88623905e-01 -9.33502093e-02 -7.46471137e-02 6.31324388e-03
5.84860325e-01 1.05646096e-01 3.09796035e-01 -7.40333736e-01
-1.38740396e+00 7.89512873e-01 6.77027404e-01 6.37549341e-01
4.52659006e-04 3.97013843e-01 4.00868297e-01 -2.65933394e-01
3.26039314e-01 5.26698053e-01 -1.04251564e-01 1.14715897e-01
-1.15285611e+00 2.66351044e-01 6.46483898e-01 7.19072223e-01
5.68537652e-01 -1.18182912e-01 -6.20207965e-01 5.34708500e-01
-1.45724013e-01 6.19141817e-01 1.20508999e-01 -1.05928957e+00
1.23967104e-01 9.34238136e-01 2.86581442e-02 -1.02111900e+00
-5.61926842e-01 -3.42543364e-01 -9.67197001e-01 6.60647035e-01
7.27921724e-01 3.70498985e-01 -9.99420702e-01 1.33778560e+00
4.28442866e-01 3.36081415e-01 -2.79228300e-01 7.96206534e-01
7.93095052e-01 4.44100797e-01 2.81862691e-02 2.66690135e-01
1.74524212e+00 -1.15703630e+00 -6.50728419e-02 -3.48240465e-01
2.93227226e-01 -5.79081714e-01 8.90976608e-01 4.77997422e-01
-1.31507599e+00 -6.97681606e-01 -1.02463448e+00 -1.11211374e-01
-8.93769920e-01 2.97277153e-01 5.90591073e-01 8.47325563e-01
-7.59859562e-01 7.51993418e-01 -6.86503708e-01 -1.76383927e-01
9.74738657e-01 1.66128457e-01 -1.55708432e-01 1.45747298e-02
-3.60755444e-01 9.25755382e-01 5.71383238e-01 -2.77015209e-01
-9.88589168e-01 -5.71756124e-01 -7.38808870e-01 1.18231826e-01
4.32485491e-01 -5.97974181e-01 7.45109797e-01 -7.84002185e-01
-1.03145754e+00 1.69277871e+00 -8.47344026e-02 -4.85457540e-01
7.18595326e-01 -7.26001412e-02 8.63941945e-03 3.43841910e-01
4.62534964e-01 8.40912163e-01 1.25647759e+00 -1.78999603e+00
-8.03477943e-01 -3.93268615e-01 -1.19241022e-01 -1.62760228e-01
-1.93027571e-01 3.60014290e-01 -6.23178720e-01 -5.94726682e-01
-2.37417102e-01 -6.57172382e-01 -1.42374635e-01 1.92301586e-01
-5.63330948e-01 -4.66121644e-01 6.23542726e-01 -9.56536680e-02
9.17205513e-01 -2.02945614e+00 1.63344398e-01 1.13789424e-01
7.47650146e-01 4.37956601e-01 -2.43663684e-01 -5.46324998e-03
-7.67382756e-02 1.02431290e-01 -4.92240667e-01 -7.60420322e-01
4.77946214e-02 5.10017462e-02 1.19091582e-03 6.71780229e-01
7.43328035e-01 9.62284446e-01 -1.33054101e+00 -8.20221126e-01
5.88974178e-01 3.55766416e-01 -2.48161361e-01 -8.56372118e-02
-1.95781946e-01 1.08050697e-01 -1.67575732e-01 8.75245035e-01
1.19300282e+00 -6.76533580e-01 -1.38912618e-01 -1.31861374e-01
-2.60981709e-01 -4.96467739e-01 -1.24039519e+00 1.09715450e+00
-1.95810422e-01 4.83836055e-01 -2.48328567e-01 -9.53711927e-01
7.19747603e-01 -4.91190612e-01 3.01897138e-01 -5.94533324e-01
5.77572048e-01 2.06308931e-01 -1.15225203e-01 -7.98552185e-02
3.99576813e-01 4.34790850e-01 -1.02517076e-01 5.98562598e-01
2.48243064e-01 -4.05295432e-01 7.46922076e-01 4.16299433e-01
1.13102245e+00 -1.70615673e-01 4.05528009e-01 -2.93174267e-01
5.24824023e-01 -1.01915531e-01 1.90435335e-01 1.13304162e+00
-4.25650656e-01 1.08232749e+00 6.26029491e-01 -5.86719751e-01
-1.30035567e+00 -1.39159846e+00 -3.14999640e-01 1.18774998e+00
1.93068370e-01 7.33987093e-02 -7.89748669e-01 -9.75964665e-01
3.06873471e-01 5.32507718e-01 -1.14887023e+00 1.03384629e-01
-9.84147489e-01 -6.50757909e-01 5.68789244e-01 7.41841018e-01
3.13851535e-01 -1.24782658e+00 -8.85854959e-01 5.02392612e-02
1.93248555e-01 -1.13653886e+00 -4.99413788e-01 6.28361106e-01
-5.42200387e-01 -1.36193657e+00 -1.06772947e+00 -9.18514788e-01
7.81527042e-01 5.08925080e-01 1.68825197e+00 4.72591013e-01
-8.34192932e-01 2.97993988e-01 -2.13153511e-01 -4.59098458e-01
-3.14014405e-01 -2.29555950e-01 -5.70054539e-02 1.07076384e-01
6.15585506e-01 -2.31159478e-01 -6.84170663e-01 1.43285826e-01
-8.68391395e-01 -2.70834297e-01 5.00394583e-01 7.71579385e-01
4.49577302e-01 -4.64141041e-01 -8.51712376e-02 -7.57693470e-01
3.24762672e-01 -1.28637239e-01 -7.45748043e-01 3.22741151e-01
-1.84127726e-02 -1.34136686e-02 6.63962841e-01 -7.05447257e-01
-3.88810337e-01 2.35955596e-01 3.27889264e-01 -5.52244425e-01
-1.86943442e-01 -6.23357654e-01 3.19311351e-01 -9.49411318e-02
7.71253228e-01 3.45690608e-01 -3.05700153e-01 -2.44897708e-01
7.53576338e-01 2.03586176e-01 8.29201460e-01 -3.04806083e-01
1.10021663e+00 1.04433799e+00 2.56197751e-01 -7.20208764e-01
-1.07370043e+00 -8.91287088e-01 -1.04923964e+00 -2.24274606e-01
1.29792202e+00 -6.31105185e-01 -1.00609159e+00 5.83429217e-01
-1.31767309e+00 -2.57311136e-01 -4.42188919e-01 1.40899613e-01
-4.60940152e-01 4.90070075e-01 -5.99194229e-01 -1.16196012e+00
-2.89352685e-01 -7.23259211e-01 1.44452357e+00 3.76414597e-01
-4.84115705e-02 -6.22088194e-01 2.44425580e-01 5.20283766e-02
5.36925234e-02 3.25832158e-01 8.01689863e-01 -7.66461194e-01
-7.06015468e-01 -2.22795889e-01 -8.73861372e-01 3.44293296e-01
-3.08601350e-01 3.50715965e-01 -1.09325767e+00 -2.26547971e-01
-4.00403470e-01 -5.93051851e-01 1.44119358e+00 3.93085450e-01
1.55857527e+00 2.19513223e-01 -5.37487686e-01 6.27134323e-01
1.31561708e+00 -2.15302318e-01 4.84445542e-01 2.73285806e-01
5.11048794e-01 1.98212653e-01 2.15759397e-01 2.77124166e-01
2.10836396e-01 4.02180493e-01 7.67706454e-01 -4.04318303e-01
-2.37949535e-01 -3.84012833e-02 -3.20545137e-02 4.22542453e-01
-2.37428606e-01 -1.91681594e-01 -7.45698452e-01 6.04606330e-01
-1.55133367e+00 -1.25180066e+00 -2.97958761e-01 2.17292404e+00
5.41896641e-01 3.97437811e-01 5.63381612e-01 2.39798412e-01
9.29471791e-01 1.84699431e-01 -3.93806547e-01 -2.70252675e-01
-6.91109121e-01 7.18019187e-01 7.18242764e-01 6.38463795e-02
-1.50093901e+00 1.03008175e+00 6.19500494e+00 1.04719651e+00
-5.88242114e-01 2.73411553e-02 8.23828518e-01 1.56788498e-01
2.38348693e-01 -2.57112712e-01 -9.12439466e-01 4.65298504e-01
8.03393498e-02 3.53195548e-01 3.11280251e-01 8.37907970e-01
-4.68006790e-01 -4.71634567e-01 -1.06765497e+00 1.08279753e+00
3.52767915e-01 -1.49758780e+00 9.80409086e-02 -2.68093437e-01
7.01540112e-01 -3.73717397e-02 8.56747404e-02 4.51385021e-01
5.07995546e-01 -9.93030250e-01 8.88158441e-01 2.71948874e-01
8.13544810e-01 -5.83464801e-01 6.02796197e-01 2.06205562e-01
-1.33405638e+00 -2.46754482e-01 -6.60905719e-01 2.74740368e-01
7.73678795e-02 5.52352488e-01 -5.03404856e-01 7.30841905e-02
7.54517972e-01 4.04198259e-01 -1.11802256e+00 1.29071987e+00
-7.13113770e-02 2.83808202e-01 -3.91720980e-01 -5.21997392e-01
3.09217811e-01 -2.72448123e-01 3.84764016e-01 1.83917499e+00
1.64891146e-02 -1.29400447e-01 8.08623955e-02 1.17976427e+00
-1.98238015e-01 -1.88446403e-01 -6.08890712e-01 1.85211599e-01
4.10058588e-01 1.50237191e+00 -1.56454194e+00 -7.58471012e-01
-2.23577634e-01 1.31264424e+00 4.41124111e-01 -2.05271348e-01
-9.70265210e-01 -3.12582284e-01 1.97367698e-01 -4.75937203e-02
7.22209752e-01 1.14697620e-01 -3.93906593e-01 -1.09461045e+00
-1.03777029e-01 -6.26056731e-01 4.28327829e-01 -7.29084671e-01
-1.51463628e+00 4.00037169e-01 -1.43178061e-01 -1.13376462e+00
2.07985774e-01 -1.07544959e+00 -7.91715682e-01 2.14426547e-01
-1.28613758e+00 -1.17136002e+00 -2.97536254e-01 3.98142159e-01
4.02797580e-01 -1.56553909e-01 7.42230058e-01 1.24778107e-01
-3.38702917e-01 6.09159350e-01 2.28958771e-01 5.55996835e-01
4.61259693e-01 -1.78974032e+00 6.77270114e-01 5.84422529e-01
5.43659866e-01 4.86505300e-01 3.04511636e-01 -4.44996566e-01
-8.55096579e-01 -1.15756285e+00 7.96562195e-01 -1.05618310e+00
5.33660710e-01 -8.36636007e-01 -7.50315428e-01 1.43613219e-01
7.73441717e-02 4.55178350e-01 7.46549964e-02 -2.45543435e-01
-7.69529343e-01 2.58158594e-01 -1.28741360e+00 8.55499446e-01
1.26056325e+00 -2.86900401e-01 -5.89768112e-01 4.59406167e-01
4.47103590e-01 -3.01991161e-02 -2.40395233e-01 2.36470416e-01
4.60508168e-01 -1.36759114e+00 1.45061505e+00 -5.67997515e-01
6.49349391e-01 -4.59161609e-01 1.15453579e-01 -9.01032805e-01
-5.07190108e-01 -2.31446460e-01 -4.08500463e-01 7.28337586e-01
1.52962267e-01 -1.82913393e-01 9.09322143e-01 -4.47982877e-01
1.37046590e-01 -5.44461310e-01 -1.02012646e+00 -9.80464280e-01
3.05439472e-01 -1.08216107e-01 4.24847841e-01 7.89057791e-01
-5.61993241e-01 3.16621393e-01 -3.35759856e-02 2.82375433e-04
1.03146577e+00 8.10452104e-02 1.09710383e+00 -1.45785046e+00
-1.69076383e-01 -1.04130006e+00 -6.76867545e-01 -1.10646868e+00
-1.84458002e-01 -8.35809469e-01 -1.45605534e-01 -1.09417951e+00
8.61969233e-01 -1.95620805e-01 -1.56499863e-01 6.14854209e-02
-4.61569846e-01 6.94608450e-01 6.01973832e-01 -1.54507264e-01
-1.12184238e+00 2.77132779e-01 1.12232471e+00 -4.82342899e-01
1.99534923e-01 5.30392714e-02 -2.46204898e-01 9.77613091e-01
5.01510322e-01 -7.85628140e-01 5.14458418e-01 -2.73185194e-01
7.24150687e-02 -5.10647058e-01 7.82366455e-01 -1.32316923e+00
4.82950918e-02 1.39356479e-01 8.37271452e-01 -9.82080102e-01
4.89967257e-01 -7.70876408e-01 -6.19702041e-01 7.78336048e-01
1.27183557e-01 7.67921156e-04 1.52160823e-01 6.10397160e-01
1.82298645e-01 -4.62644249e-01 1.17696416e+00 -4.81038839e-01
-4.19219315e-01 4.66169454e-02 -2.46194497e-01 3.89856011e-01
1.08767509e+00 -6.65347934e-01 -3.77620131e-01 2.53346324e-01
-6.35635257e-01 1.21428944e-01 4.69055951e-01 2.11758658e-01
4.60493833e-01 -1.32974517e+00 -7.48012483e-01 1.82789013e-01
3.32409441e-01 8.45121369e-02 -6.25531524e-02 4.84735429e-01
-6.42007351e-01 1.10515535e-01 -1.87365144e-01 -1.03281212e+00
-1.27023160e+00 8.52281749e-01 2.35034749e-01 -5.42627037e-01
-7.18854070e-01 1.23552072e+00 3.21448147e-01 -6.30566955e-01
3.53111744e-01 -3.90705228e-01 -3.03343207e-01 2.80873537e-01
8.26784134e-01 5.77792525e-01 -1.50305644e-01 -3.25870275e-01
-4.60823029e-01 9.31295395e-01 1.19223304e-01 -1.06569193e-02
1.32939577e+00 4.12646800e-01 -1.89569872e-03 5.73093653e-01
1.17127144e+00 1.87186718e-01 -1.25430930e+00 -2.11533159e-01
1.32099852e-01 -6.13481462e-01 -4.99181151e-01 -5.83086967e-01
-9.94504809e-01 9.86416280e-01 7.66095877e-01 3.47012609e-01
6.65208578e-01 2.29577363e-01 3.10205132e-01 4.66721117e-01
2.76019722e-01 -1.23619401e+00 7.91551471e-01 7.23059535e-01
6.77502990e-01 -1.51564360e+00 8.38181376e-02 -4.59419265e-02
-5.46291649e-01 1.24414623e+00 5.18952012e-01 -5.96913815e-01
3.41569364e-01 5.01314521e-01 -1.22104928e-01 -4.36356068e-01
-2.84771949e-01 -7.99034059e-01 1.97898746e-01 9.24799800e-01
1.23342022e-01 8.27139616e-02 4.44431268e-02 1.15258299e-01
1.34504721e-01 -4.44338351e-01 4.41058040e-01 7.55658627e-01
-6.34437621e-01 -5.75761616e-01 -6.15847945e-01 6.47665143e-01
-3.22318554e-01 -3.74019891e-02 -7.19362497e-01 1.00928128e+00
4.75750983e-01 4.80347812e-01 6.03101909e-01 4.52582426e-02
2.20850900e-01 -2.23735005e-01 7.71144331e-01 -6.11523092e-01
-7.00909555e-01 -3.17333490e-01 -4.48925555e-01 -5.65337837e-01
-3.80242199e-01 -6.28769577e-01 -1.05625546e+00 -2.41624847e-01
-5.12693048e-01 -4.57392484e-01 4.56674606e-01 5.03413498e-01
-2.34929249e-01 4.60022062e-01 7.38113642e-01 -1.24938524e+00
-5.58501005e-01 -8.09967935e-01 -3.95557672e-01 6.91975296e-01
2.63615519e-01 -8.10716152e-01 -5.06368279e-01 -1.36353835e-01] | [9.0451078414917, 0.49930113554000854] |
df214a42-30bd-4b34-bf49-eed26d9e99a2 | df-net-unsupervised-joint-learning-of-depth | 1809.01649 | null | http://arxiv.org/abs/1809.01649v1 | http://arxiv.org/pdf/1809.01649v1.pdf | DF-Net: Unsupervised Joint Learning of Depth and Flow using Cross-Task Consistency | We present an unsupervised learning framework for simultaneously training
single-view depth prediction and optical flow estimation models using unlabeled
video sequences. Existing unsupervised methods often exploit brightness
constancy and spatial smoothness priors to train depth or flow models. In this
paper, we propose to leverage geometric consistency as additional supervisory
signals. Our core idea is that for rigid regions we can use the predicted scene
depth and camera motion to synthesize 2D optical flow by backprojecting the
induced 3D scene flow. The discrepancy between the rigid flow (from depth
prediction and camera motion) and the estimated flow (from optical flow model)
allows us to impose a cross-task consistency loss. While all the networks are
jointly optimized during training, they can be applied independently at test
time. Extensive experiments demonstrate that our depth and flow models compare
favorably with state-of-the-art unsupervised methods. | ['Jia-Bin Huang', 'Yuliang Zou', 'Zelun Luo'] | 2018-09-05 | df-net-unsupervised-joint-learning-of-depth-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Yuliang_Zou_DF-Net_Unsupervised_Joint_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Yuliang_Zou_DF-Net_Unsupervised_Joint_ECCV_2018_paper.pdf | eccv-2018-9 | ['depth-and-camera-motion'] | ['computer-vision'] | [ 1.98977083e-01 1.55640483e-01 -5.00695229e-01 -5.85267007e-01
-3.17201316e-01 -6.58063948e-01 6.45243406e-01 -6.28511488e-01
-3.24928313e-01 7.14000285e-01 4.15597111e-01 -4.58620638e-02
4.25978035e-01 -4.20436740e-01 -7.55017400e-01 -5.64492226e-01
8.48240182e-02 2.00929493e-01 1.99930206e-01 4.30838913e-01
4.26788598e-01 4.62117881e-01 -1.24374843e+00 3.14433724e-02
6.62843108e-01 6.18710637e-01 2.83001959e-01 1.07287848e+00
-8.18636119e-02 1.48986256e+00 -1.07443035e-01 3.30980532e-02
6.59119248e-01 -5.68534374e-01 -1.03812194e+00 6.27985537e-01
1.06132388e+00 -1.12144399e+00 -9.07101929e-01 9.81655955e-01
1.29618481e-01 6.30519211e-01 4.71474588e-01 -1.17027080e+00
-4.07786608e-01 -2.66494472e-02 -5.77415168e-01 4.28586930e-01
4.67938453e-01 3.95210266e-01 8.38055193e-01 -8.38657260e-01
1.15565491e+00 1.08184993e+00 2.37847805e-01 8.50564599e-01
-1.39245784e+00 -2.47193500e-01 4.03341115e-01 -4.01354358e-02
-8.70730698e-01 -7.86115110e-01 1.02454269e+00 -7.41395295e-01
8.82557929e-01 -2.71456152e-01 5.12696922e-01 1.03728306e+00
1.31869987e-01 8.80140364e-01 8.51806998e-01 -2.87368804e-01
1.84203178e-01 -6.92127123e-02 -3.59334022e-01 1.15225434e+00
-6.71923161e-02 4.40533012e-01 -6.77162826e-01 2.94276208e-01
1.41373062e+00 5.41675203e-02 -6.22371018e-01 -6.51599169e-01
-1.18215203e+00 6.35935783e-01 3.72000903e-01 -1.98310152e-01
-7.43078738e-02 3.82285714e-01 9.28366333e-02 1.05658539e-01
6.39029801e-01 2.44366288e-01 -4.22813982e-01 -2.25448474e-01
-1.12169790e+00 7.31479451e-02 6.66145504e-01 1.01907074e+00
1.18323004e+00 3.96082431e-01 4.22368310e-02 4.09581095e-01
5.42665958e-01 4.41077262e-01 3.43391955e-01 -1.77108467e+00
5.58267832e-01 6.85747266e-02 2.35011727e-01 -9.38419223e-01
-1.27477854e-01 7.53333345e-02 -6.10047281e-01 4.23062801e-01
7.55379379e-01 -3.20678174e-01 -1.07220459e+00 1.78564775e+00
3.20193738e-01 8.88409853e-01 1.03958182e-01 1.28419673e+00
5.44033408e-01 5.63930690e-01 -2.24669173e-01 -1.06367506e-01
5.17911434e-01 -1.40378225e+00 -5.64868033e-01 -5.99211693e-01
6.96339965e-01 -7.64150202e-01 6.43589914e-01 1.21991448e-01
-1.44396281e+00 -8.27174783e-01 -8.42068791e-01 -4.70184535e-01
3.38102162e-01 -1.34336855e-02 5.35901427e-01 3.02596897e-01
-1.25837433e+00 6.59955204e-01 -1.26882923e+00 -2.13337228e-01
2.06820667e-01 2.17091143e-01 -6.97646856e-01 -1.95312902e-01
-7.57574737e-01 7.19664872e-01 2.48810083e-01 1.26074642e-01
-1.22647393e+00 -6.44939005e-01 -1.31040967e+00 -2.59397686e-01
9.30971652e-02 -1.08783591e+00 1.16772699e+00 -1.11320651e+00
-2.03929257e+00 8.89397144e-01 -6.44424498e-01 -3.03923100e-01
6.06212497e-01 -4.63859916e-01 2.35728651e-01 6.31289124e-01
2.31038809e-01 1.07487214e+00 9.09145474e-01 -1.25138593e+00
-7.19870508e-01 -8.32820982e-02 1.66511580e-01 4.92707431e-01
6.99637607e-02 -4.92469579e-01 -6.36592209e-01 -4.84524667e-01
3.26570630e-01 -1.01514995e+00 -2.91382194e-01 2.61251360e-01
-4.37595069e-01 3.27334583e-01 8.97480607e-01 -5.91550529e-01
7.11794257e-01 -1.80125272e+00 3.94943684e-01 1.89589011e-03
1.54461369e-01 -1.40778154e-01 -1.08634926e-01 -1.88813165e-01
-6.89690635e-02 -2.25510374e-01 -3.84564400e-01 -8.18483114e-01
-4.80388016e-01 5.04806578e-01 -4.95351970e-01 9.08002019e-01
3.71922255e-01 9.35660303e-01 -1.22905827e+00 -4.49985832e-01
5.81075788e-01 4.64135945e-01 -9.26443994e-01 6.12253129e-01
-1.86785832e-01 1.28047836e+00 -3.83051872e-01 2.60685503e-01
6.59196973e-01 -4.44221139e-01 1.65781587e-01 -8.54444653e-02
-5.33823331e-05 5.58768272e-01 -1.07450259e+00 2.21652341e+00
-3.05894196e-01 9.50605869e-01 4.15975302e-02 -9.56063867e-01
6.56258583e-01 2.21683860e-01 7.41368294e-01 -4.45036709e-01
-7.80237764e-02 -2.07850292e-01 -3.35378468e-01 -6.15820229e-01
3.69189113e-01 -2.08984986e-01 5.33754706e-01 6.22081220e-01
5.30701220e-01 -3.72418046e-01 7.47788846e-02 1.40177310e-01
8.64247024e-01 9.85624373e-01 -1.36412933e-01 -2.38829814e-02
6.27134562e-01 -2.37322018e-01 8.00998747e-01 5.54371059e-01
-4.34426278e-01 9.43155348e-01 3.99537683e-01 -5.47210991e-01
-1.16189432e+00 -1.21331096e+00 1.07921712e-01 7.78801441e-01
3.81834030e-01 -1.17394067e-01 -4.27087009e-01 -9.14725780e-01
-2.63069183e-01 2.68446296e-01 -5.43624401e-01 2.50242859e-01
-8.54627907e-01 1.03524246e-03 2.82856017e-01 6.68548942e-01
6.29222095e-01 -7.60778725e-01 -7.07408488e-01 2.02608854e-01
-4.00466710e-01 -1.74548626e+00 -7.56461620e-01 6.05038553e-03
-1.23901379e+00 -9.68774140e-01 -7.06341922e-01 -7.30472088e-01
9.00928915e-01 5.16711235e-01 1.19259202e+00 -9.62922815e-03
6.11666106e-02 6.56082630e-01 2.18424723e-02 2.80540675e-01
-2.36059859e-01 -1.73108235e-01 6.12466298e-02 1.62786558e-01
-8.81867930e-02 -7.22791791e-01 -8.59498322e-01 2.30015039e-01
-9.12368417e-01 2.21296668e-01 -2.23757848e-02 7.14656830e-01
5.75715005e-01 -5.16319156e-01 -1.67071119e-01 -8.71311784e-01
-2.12027401e-01 -1.44715786e-01 -8.03239226e-01 -1.31911650e-01
-3.76684636e-01 2.93877393e-01 4.71202374e-01 -1.73536822e-01
-1.46522653e+00 5.62101662e-01 1.49564847e-01 -1.14389277e+00
-3.31157655e-01 2.63478439e-02 3.20031606e-02 -1.59687653e-01
4.14385885e-01 1.28841385e-01 1.14468284e-01 -1.94368124e-01
6.48137450e-01 6.37154933e-03 7.89114535e-01 -6.54136658e-01
8.88521075e-01 1.12060738e+00 9.16514173e-02 -6.62961543e-01
-1.12769246e+00 -7.28166878e-01 -1.14167607e+00 -2.56279200e-01
1.29013133e+00 -1.16960311e+00 -4.81517285e-01 4.44221884e-01
-1.40417659e+00 -7.93618977e-01 -2.26287901e-01 8.99485528e-01
-9.78172183e-01 7.56907046e-01 -8.48775983e-01 -6.84851944e-01
1.13291964e-01 -1.24488628e+00 1.23483157e+00 1.97115973e-01
-5.67336753e-02 -1.63673460e+00 3.11906189e-01 2.84522623e-01
-4.19842117e-02 2.76240677e-01 1.53789818e-01 1.51697949e-01
-9.94615495e-01 3.09114665e-01 -1.63001940e-01 4.85268652e-01
3.90578002e-01 1.88931078e-01 -1.18161297e+00 -1.87087402e-01
2.73040503e-01 -3.03641438e-01 1.01291704e+00 7.15400696e-01
9.70365584e-01 -1.59106448e-01 -5.73375747e-02 1.26908088e+00
1.38164854e+00 -1.65794551e-01 6.64093137e-01 2.19008580e-01
1.20995247e+00 8.24972510e-01 3.42769355e-01 2.92254180e-01
4.04554814e-01 4.34071749e-01 3.92871141e-01 -6.04892373e-02
-1.64586917e-01 -4.44831967e-01 6.41085565e-01 6.56297684e-01
-2.02380016e-01 -5.10170124e-03 -7.04848945e-01 4.85599130e-01
-1.87223577e+00 -1.24698901e+00 4.14775126e-03 2.18632627e+00
6.61393166e-01 1.79821208e-01 -1.31608471e-01 -3.25962335e-01
4.46032345e-01 4.69280183e-01 -5.21452010e-01 -1.34848714e-01
-9.57163349e-02 7.46604428e-02 6.35105908e-01 1.21377432e+00
-1.18304861e+00 1.15574372e+00 6.76519728e+00 -1.22752562e-01
-1.31799328e+00 -4.70110066e-02 7.88509488e-01 -1.63833991e-01
-3.81883025e-01 3.19777280e-01 -6.71273053e-01 2.18565181e-01
6.04008377e-01 2.72808760e-01 3.30975056e-01 5.92476130e-01
5.72911322e-01 -2.16661796e-01 -1.40596175e+00 9.88334358e-01
1.29467055e-01 -1.46976399e+00 -5.63978292e-02 2.58769747e-03
1.30522811e+00 2.40772665e-01 -1.11555330e-01 -2.25535020e-01
4.87632006e-01 -8.11257064e-01 5.71716845e-01 5.65420628e-01
7.25997388e-01 -1.95549503e-01 3.10680509e-01 2.25422859e-01
-1.02516603e+00 2.04817474e-01 -2.04795182e-01 -3.12529057e-01
7.05601752e-01 4.90520239e-01 -4.92252588e-01 4.51749980e-01
5.33324957e-01 1.48879683e+00 -2.24037170e-01 7.50890374e-01
-7.13992774e-01 4.30223733e-01 -2.54495472e-01 7.83683181e-01
4.73657340e-01 -5.65837681e-01 5.70435882e-01 1.00361001e+00
2.34128144e-02 -5.48117310e-02 3.42551470e-01 1.04295301e+00
-4.90934961e-02 -3.95392925e-01 -8.17123771e-01 4.13272917e-01
-6.87413216e-02 1.02878773e+00 -6.46332622e-01 -6.15858734e-01
-7.30022430e-01 1.38712502e+00 3.70712191e-01 8.88660073e-01
-6.30941391e-01 2.26472929e-01 8.48467350e-01 -1.50298299e-02
2.74274498e-01 -5.91170490e-01 -2.96003729e-01 -1.83818650e+00
-1.46750687e-02 -1.64650619e-01 1.93234995e-01 -1.01171970e+00
-1.23976088e+00 1.44644871e-01 -1.33253887e-01 -1.45872688e+00
-6.49863720e-01 -6.52077913e-01 -8.29318345e-01 8.25999439e-01
-2.06666613e+00 -7.12745011e-01 -4.16373104e-01 8.01790595e-01
7.10212111e-01 1.50928006e-01 3.89291078e-01 8.11262429e-02
-3.46274793e-01 9.93725359e-02 -1.21202886e-01 4.61035669e-01
9.71485198e-01 -1.33220375e+00 3.98190767e-01 1.18221557e+00
3.91004890e-01 5.00497699e-01 5.12242377e-01 -4.49165821e-01
-1.14574420e+00 -1.07038856e+00 6.12905324e-01 -8.07194531e-01
5.46331644e-01 -1.36654437e-01 -8.04356039e-01 1.00672293e+00
2.16262415e-01 6.93364024e-01 3.55959058e-01 -4.23841894e-01
-3.65362853e-01 8.75433087e-02 -8.30040038e-01 4.84414965e-01
1.09097540e+00 -8.69840086e-01 -3.54686111e-01 1.33958623e-01
5.48833489e-01 -7.49730825e-01 -6.58033311e-01 1.64581746e-01
5.55126131e-01 -1.27295232e+00 9.04412091e-01 -7.40514696e-01
8.62906933e-01 -4.19014812e-01 -3.74201871e-02 -1.10039115e+00
-7.89184645e-02 -9.34581935e-01 -4.31149691e-01 7.91834474e-01
1.85278594e-01 -3.42083871e-01 1.25899839e+00 8.45404029e-01
-1.23011723e-01 -1.94823578e-01 -7.13313043e-01 -4.29895043e-01
-6.23806082e-02 -4.59071517e-01 -2.56160527e-01 1.13626397e+00
-1.46271676e-01 3.79714757e-01 -4.88330543e-01 2.48496696e-01
9.15621042e-01 7.84666836e-03 1.08672929e+00 -8.50801945e-01
-6.44983888e-01 -2.54563510e-01 -4.33281958e-01 -1.92135429e+00
5.61366618e-01 -6.49409235e-01 3.08502734e-01 -1.32341659e+00
-1.35954231e-01 3.39410082e-02 -8.55858624e-02 2.02701479e-01
-1.54903129e-01 1.43911555e-01 2.55551875e-01 4.40305501e-01
-4.08352315e-01 5.25577128e-01 1.69872272e+00 2.31049247e-02
-5.40557802e-01 -1.76994950e-01 -6.77298978e-02 9.78255808e-01
5.98805249e-01 -2.04963595e-01 -8.30474496e-01 -8.75738621e-01
-2.83262253e-01 3.36282879e-01 4.34504867e-01 -7.37830460e-01
3.54696721e-01 -3.47040147e-01 6.23851657e-01 -1.81732699e-01
2.64068395e-01 -6.57554030e-01 -4.44256693e-01 1.61207169e-01
-3.62445295e-01 -5.41554727e-02 9.53235757e-03 7.93871224e-01
-2.76070952e-01 -8.38698149e-02 7.63811946e-01 -2.71557897e-01
-8.93927753e-01 6.24712825e-01 -2.89433002e-01 3.16709369e-01
6.51534677e-01 -4.21142310e-01 -1.64535120e-01 -7.26110041e-01
-8.61284912e-01 7.27722496e-02 6.36940956e-01 4.80802029e-01
8.91687036e-01 -1.09384274e+00 -5.49540937e-01 5.19721925e-01
-1.07629389e-01 4.39772189e-01 1.18228197e-01 8.41094911e-01
-8.54095101e-01 3.24819326e-01 -3.17027032e-01 -9.70575571e-01
-7.19791472e-01 3.68976802e-01 5.65025628e-01 -2.21048087e-01
-8.74303222e-01 8.98374498e-01 7.66956985e-01 -3.40464443e-01
2.61469901e-01 -3.01019788e-01 2.58326232e-01 -5.47953844e-01
4.38032746e-01 1.76054955e-01 -4.76259619e-01 -7.09046841e-01
-1.93833008e-01 8.84096622e-01 1.09917521e-01 -5.25121272e-01
1.17778850e+00 -5.36853254e-01 9.17037651e-02 3.90665054e-01
1.56505072e+00 -2.17016973e-02 -2.26948881e+00 -1.35042146e-01
-1.71775594e-01 -7.69510508e-01 1.58511564e-01 -1.33090869e-01
-1.36238420e+00 1.00809264e+00 3.19957435e-01 -2.64572501e-01
9.20958042e-01 -2.69544274e-01 6.02849185e-01 3.09225887e-01
9.63301808e-02 -9.55313087e-01 3.48907113e-01 6.27884328e-01
2.41118684e-01 -1.48978364e+00 2.73709875e-02 -4.67297226e-01
-6.17732942e-01 1.26079762e+00 8.49187076e-01 -5.41259348e-01
5.96496344e-01 2.66677201e-01 1.95013016e-01 1.39667436e-01
-7.26888776e-01 -2.53451675e-01 3.77466470e-01 7.22143173e-01
5.04743993e-01 -5.76499879e-01 3.99666369e-01 -5.07353663e-01
1.65297970e-01 1.66557670e-01 8.05591822e-01 8.02958786e-01
-3.22357267e-01 -1.01157331e+00 -1.12826228e-01 -1.00789845e-01
-3.71559888e-01 -6.51329979e-02 -1.92287803e-01 5.88119268e-01
-1.39957443e-01 7.93245018e-01 4.02295768e-01 -8.18194449e-02
-6.35887384e-02 -6.09899461e-02 8.44836295e-01 -6.31519556e-01
5.45974784e-02 3.08891594e-01 -1.11537270e-01 -9.52914536e-01
-1.06679928e+00 -4.12294060e-01 -1.37389231e+00 -1.56668872e-01
7.26466114e-03 -1.78289518e-01 2.67040104e-01 1.04055452e+00
2.01617599e-01 1.54754952e-01 8.27400863e-01 -1.21782303e+00
1.79074949e-03 -6.09369576e-01 -3.93969417e-01 6.58293366e-01
8.33523095e-01 -5.86049616e-01 -7.13180125e-01 9.09511328e-01] | [8.645195960998535, -2.0286357402801514] |
18e846fd-9eed-4c71-a2d5-16692cd173cd | abode-net-an-attention-based-deep-learning | 2212.11396 | null | https://arxiv.org/abs/2212.11396v1 | https://arxiv.org/pdf/2212.11396v1.pdf | ABODE-Net: An Attention-based Deep Learning Model for Non-intrusive Building Occupancy Detection Using Smart Meter Data | Occupancy information is useful for efficient energy management in the building sector. The massive high-resolution electrical power consumption data collected by smart meters in the advanced metering infrastructure (AMI) network make it possible to infer buildings' occupancy status in a non-intrusive way. In this paper, we propose a deep leaning model called ABODE-Net which employs a novel Parallel Attention (PA) block for building occupancy detection using smart meter data. The PA block combines the temporal, variable, and channel attention modules in a parallel way to signify important features for occupancy detection. We adopt two smart meter datasets widely used for building occupancy detection in our performance evaluation. A set of state-of-the-art shallow machine learning and deep learning models are included for performance comparison. The results show that ABODE-Net significantly outperforms other models in all experimental cases, which proves its validity as a solution for non-intrusive building occupancy detection. | ['Sihua Shao', 'Jun Zheng', 'Qingqing Li', 'Ruobin Qi', 'Zhirui Luo'] | 2022-12-21 | null | null | null | null | ['energy-management'] | ['time-series'] | [-1.30227000e-01 -1.17625564e-01 1.32765114e-01 -3.00447941e-01
-7.66326845e-01 1.55674517e-01 6.67314112e-01 2.20289409e-01
-4.17149633e-01 7.40039051e-01 5.42135239e-01 -2.89200306e-01
-1.42686680e-01 -1.15028000e+00 -2.32908487e-01 -1.25166237e+00
-2.44300634e-01 3.56783599e-01 -2.14849725e-01 -1.30094262e-02
3.38690467e-02 3.62071455e-01 -1.55748284e+00 5.07847406e-02
6.49058163e-01 1.22341430e+00 1.98830947e-01 6.58964217e-01
1.99559927e-01 1.12710512e+00 -7.98410475e-01 3.69988024e-01
9.82828252e-03 1.10914242e-02 -6.98823571e-01 -1.43524691e-01
-1.87197775e-02 -7.59826958e-01 -4.10414666e-01 6.88207150e-01
1.08150518e+00 1.85453385e-01 5.28775036e-01 -1.46247661e+00
-3.40934962e-01 1.16691840e+00 -5.20533264e-01 7.95601964e-01
2.14634836e-01 2.03520134e-01 1.09436262e+00 -3.99100274e-01
-8.01049948e-01 8.63895178e-01 7.39749968e-01 3.10621914e-02
-1.13214326e+00 -9.44772959e-01 2.13446736e-01 7.80483723e-01
-1.95763612e+00 -2.28336915e-01 9.03459668e-01 -2.10713118e-01
1.62113893e+00 7.22087741e-01 1.08041298e+00 1.09216523e+00
1.82873800e-01 1.09399605e+00 7.56644130e-01 -1.44479573e-01
7.90339708e-01 -2.55581617e-01 4.69829738e-02 1.64385200e-01
2.98204184e-01 5.71743734e-02 -4.99485061e-02 -3.18597294e-02
2.93678999e-01 3.27768207e-01 -1.16424225e-01 5.57935648e-02
-8.66729915e-01 9.18732226e-01 8.67084384e-01 8.36788595e-01
-6.85798943e-01 4.96416777e-01 4.41645026e-01 -3.25644970e-01
3.88429105e-01 2.32534587e-01 -4.40588385e-01 -1.01550311e-01
-1.02792156e+00 -1.73547819e-01 6.66258574e-01 5.99764168e-01
6.30244732e-01 4.34819937e-01 -3.56182277e-01 4.37685192e-01
4.30888534e-01 5.58675706e-01 8.87769580e-01 -3.26051503e-01
2.57135332e-01 5.60692012e-01 7.87325278e-02 -6.10164404e-01
-1.04233336e+00 -4.16624457e-01 -1.55893826e+00 -1.84992909e-01
-5.05619049e-01 2.54646987e-02 -8.16214919e-01 1.39039671e+00
2.06222489e-01 4.39334780e-01 -1.37179554e-01 4.35159385e-01
9.06801522e-01 7.20202327e-01 5.12447774e-01 -1.52987882e-01
1.58158135e+00 -8.05837154e-01 -8.35981250e-01 -8.73436481e-02
5.83857536e-01 1.21881656e-01 9.11331475e-01 2.25268930e-01
-4.62233931e-01 -5.31047344e-01 -1.35527956e+00 -1.05398595e-01
-8.78697395e-01 -3.53005230e-01 4.49302644e-01 6.90704048e-01
-9.06589270e-01 4.26932603e-01 -1.04046369e+00 -1.82250708e-01
8.82825494e-01 5.03458440e-01 2.31987223e-01 3.87919515e-01
-1.28959274e+00 8.37379992e-01 6.10817969e-01 1.73836097e-01
-1.04555404e+00 -5.27572870e-01 -9.91832733e-01 5.90978086e-01
4.76939976e-02 -4.64775771e-01 1.40167630e+00 -2.13566869e-01
-1.16343939e+00 3.02854598e-01 3.52338761e-01 -7.68860281e-01
3.72128278e-01 -2.30738521e-01 -5.69973886e-01 -4.21564966e-01
1.56668067e-01 4.73518699e-01 6.01418555e-01 -8.97433341e-01
-1.00947988e+00 -3.57891172e-01 -2.27788493e-01 -6.87538832e-02
-2.37706423e-01 -7.11993992e-01 2.99858779e-01 -1.74689353e-01
-1.74673617e-01 -3.78545463e-01 -3.29167724e-01 -9.79896247e-01
-7.73707807e-01 -6.22640431e-01 1.21792543e+00 -8.57570112e-01
1.58131623e+00 -2.02614880e+00 -6.26293421e-01 3.03119004e-01
2.52196550e-01 3.00130099e-01 4.47219849e-01 3.52965385e-01
-1.76412433e-01 1.09044716e-01 -3.30976725e-01 -6.35636032e-01
7.41152942e-01 3.57464135e-01 -1.92269199e-02 5.63539565e-01
-2.56983101e-01 1.46602333e+00 -8.47335637e-01 -4.43014681e-01
1.27038574e+00 9.58979070e-01 5.58869801e-02 2.93590933e-01
6.27663806e-02 2.58064985e-01 -4.86266434e-01 7.19758928e-01
4.54551250e-01 -3.25589687e-01 -8.40399861e-02 -3.34086895e-01
-2.09839225e-01 5.76615095e-01 -9.35367942e-01 1.31466401e+00
-7.53244042e-01 5.89498699e-01 -3.21804315e-01 -1.21455646e+00
6.26273036e-01 4.71602827e-01 7.62257218e-01 -1.43864226e+00
4.83349591e-01 -2.09346637e-01 -3.03804964e-01 -3.34276199e-01
4.90181655e-01 1.36784906e-03 -7.03426242e-01 2.17554927e-01
-4.50788051e-01 1.20155454e-01 -2.81823007e-03 -2.93262094e-01
1.51665246e+00 -6.55151427e-01 9.32270288e-01 -4.60600823e-01
6.10091448e-01 -6.96667552e-01 3.53182882e-01 8.19525719e-01
-5.54041743e-01 -1.01818396e-02 -1.88252404e-01 -8.66184950e-01
-8.63625407e-01 -9.09724891e-01 -3.47387433e-01 1.05978978e+00
-3.24008971e-01 -2.79117882e-01 -7.35493660e-01 -4.37540323e-01
8.07392150e-02 1.15202165e+00 -8.12300026e-01 2.42735092e-02
-4.16286439e-01 -1.06622636e+00 4.59731638e-01 1.03575110e+00
1.13167489e+00 -1.50577545e+00 -1.19263005e+00 6.28544569e-01
-3.79427403e-01 -8.43523383e-01 4.19757441e-02 9.88410532e-01
-2.05245733e-01 -9.83938396e-01 -2.20762849e-01 -4.03639048e-01
8.37029964e-02 3.71248350e-02 1.34498632e+00 9.22267437e-02
-5.51424623e-01 1.88629821e-01 -1.19321547e-01 -5.34026682e-01
1.91563573e-02 5.30449986e-01 -3.76044475e-02 -2.00732082e-01
1.08824396e+00 -1.10131800e+00 -8.57405901e-01 9.71072391e-02
-7.33873308e-01 -1.52474940e-01 5.80797732e-01 3.10870856e-01
5.00510752e-01 7.12316334e-01 7.09297717e-01 -2.65144646e-01
4.68192935e-01 -6.77669466e-01 -7.80482054e-01 -2.64307231e-01
-8.02083492e-01 -1.21620014e-01 7.59506464e-01 7.44942650e-02
-4.21209097e-01 -1.60593949e-02 -6.69678271e-01 7.64239728e-02
-5.67903996e-01 -8.37641419e-04 -7.31670856e-01 4.43600297e-01
1.06264792e-01 3.51943880e-01 -1.05267167e+00 -6.70875013e-01
1.37410700e-01 8.97487581e-01 7.43488610e-01 1.39494240e-01
7.35433340e-01 4.05908883e-01 -9.88214165e-02 -9.54171777e-01
-8.47050488e-01 -7.69916117e-01 -6.82728767e-01 1.14963494e-01
1.10899162e+00 -1.07706499e+00 -1.22551548e+00 5.32058597e-01
-6.17425084e-01 -5.02673745e-01 -6.75713539e-01 5.35460934e-02
-4.66327697e-01 1.71349213e-01 -4.94875640e-01 -1.20702660e+00
-8.98995221e-01 -8.46337199e-01 1.21978915e+00 1.88320056e-01
-3.74600530e-01 -9.07318592e-01 2.97730625e-01 2.56108552e-01
6.37825966e-01 4.32959825e-01 6.12906218e-01 -6.88075006e-01
-6.36355340e-01 -3.36098611e-01 -1.34425193e-01 2.73646921e-01
2.86094487e-01 -6.08408630e-01 -1.49149752e+00 -4.16642964e-01
-4.15464230e-02 -1.02994569e-01 8.94383490e-01 6.70315266e-01
1.62629914e+00 -5.60579598e-01 -4.74891633e-01 5.49047232e-01
1.67935014e+00 3.20360363e-01 1.11721158e+00 7.40685046e-01
8.97802770e-01 -3.50431979e-01 -9.02079344e-02 8.76598835e-01
6.30963504e-01 4.94143546e-01 8.28813314e-01 -3.77860546e-01
4.87806499e-01 -2.28659660e-01 1.01683317e-02 8.09490383e-01
2.98398972e-01 -5.15833616e-01 -8.31887066e-01 8.17038000e-01
-2.01680040e+00 -1.25042689e+00 1.70997098e-01 1.83320129e+00
3.67166758e-01 -6.65637404e-02 1.66096538e-01 9.73771334e-01
3.88317049e-01 5.81764102e-01 -5.69663167e-01 -3.55481088e-01
1.31603003e-01 3.38917404e-01 7.32658565e-01 2.85372883e-01
-1.57855308e+00 2.89843738e-01 6.18468285e+00 6.70219243e-01
-7.27816701e-01 4.92610723e-01 7.60799348e-01 -2.76594162e-01
8.92466307e-03 -7.38298059e-01 -7.43481994e-01 9.56596315e-01
1.37089229e+00 8.62860903e-02 2.66950101e-01 1.06320453e+00
3.70417923e-01 -3.35478753e-01 -1.20791030e+00 1.20293021e+00
-1.68037429e-01 -1.00084257e+00 -4.27014858e-01 4.34534699e-01
5.78896523e-01 5.66661716e-01 -1.54744163e-01 6.98184788e-01
6.15425050e-01 -1.31751502e+00 3.50096762e-01 2.84890205e-01
3.10763478e-01 -1.12905049e+00 1.29961514e+00 2.99075007e-01
-1.88853014e+00 -5.65329075e-01 -9.83893350e-02 -2.94860750e-01
4.30893749e-01 6.53890729e-01 -9.00645614e-01 3.87931108e-01
1.15711904e+00 5.45669973e-01 -5.26608169e-01 8.35482538e-01
-1.66529283e-01 5.85071445e-01 -7.29337156e-01 -9.11813155e-02
5.44937313e-01 2.08600044e-01 -2.10151464e-01 1.26264453e+00
1.97302908e-01 1.11621924e-01 4.08500195e-01 8.48447263e-01
2.83812620e-02 -2.30050996e-01 -6.39663875e-01 4.21576768e-01
5.71801066e-01 1.36581230e+00 -5.91765404e-01 -4.07390624e-01
-3.64644259e-01 8.43718767e-01 9.95411128e-02 1.71934608e-02
-1.09565973e+00 -1.66086972e-01 6.37005746e-01 9.63731669e-03
8.15944135e-01 1.99025273e-01 -1.98220059e-01 -5.96372962e-01
-3.83192390e-01 -3.07612717e-01 4.51185882e-01 -7.34571993e-01
-1.17762554e+00 1.72670051e-01 -6.74807131e-02 -5.87690651e-01
-2.37764627e-01 -1.26069132e-02 -8.52725387e-01 4.33947086e-01
-1.71275520e+00 -9.89063859e-01 -7.67588794e-01 7.61926711e-01
5.79586625e-01 3.13886225e-01 1.06288338e+00 6.52512610e-01
-9.95196939e-01 3.78060281e-01 1.06548831e-01 4.38088685e-01
-4.71499175e-01 -1.59904003e+00 7.83475518e-01 6.87273622e-01
1.26997948e-01 -3.99236053e-01 6.36680901e-01 -2.45850623e-01
-9.05080616e-01 -1.45186555e+00 7.65196621e-01 -3.37651610e-01
3.77719015e-01 -4.13638175e-01 -7.04186678e-01 9.34154868e-01
3.98951054e-01 -2.75719941e-01 8.84883165e-01 -4.64675128e-02
4.31622453e-02 -2.60313570e-01 -1.28622377e+00 1.68164879e-01
6.06619239e-01 -5.20383418e-01 -5.46346486e-01 1.85695514e-01
4.70720351e-01 3.32238078e-01 -9.14369345e-01 4.01699483e-01
2.14816347e-01 -1.00206840e+00 9.54150796e-01 1.89676821e-01
-3.16697776e-01 -3.91564429e-01 -6.29760742e-01 -1.13914275e+00
-9.91084337e-01 -3.89817566e-01 -7.61929333e-01 1.15877354e+00
-2.76539065e-02 -5.99901259e-01 6.36946499e-01 6.02498710e-01
-1.87214136e-01 -4.06893164e-01 -1.41090035e+00 -4.43064600e-01
-3.31673235e-01 -5.70662022e-01 1.31816983e+00 8.50205183e-01
6.51885048e-02 7.37927735e-01 -4.30125147e-01 4.15001243e-01
7.23081887e-01 -2.36311793e-01 6.27675533e-01 -1.31768894e+00
1.32300809e-01 -5.43671668e-01 -7.72824883e-01 -1.00230324e+00
1.15657207e-02 -4.70990747e-01 2.85502493e-01 -1.97266686e+00
3.24335456e-01 -1.40021313e-02 -9.39157486e-01 7.75653720e-01
1.29201904e-01 3.39259326e-01 -3.11753839e-01 -2.61866122e-01
-8.07634711e-01 9.58188415e-01 2.02622205e-01 -6.16152108e-01
-1.19513333e-01 -6.46839440e-02 -5.74049950e-01 6.38414979e-01
1.32000041e+00 -1.67094797e-01 -3.12617838e-01 -1.11158425e-02
-1.78148910e-01 -5.38999975e-01 4.34501559e-01 -1.47824883e+00
1.93829164e-02 3.10168386e-01 7.83371866e-01 -1.42098439e+00
3.51766080e-01 -1.18153310e+00 2.13261455e-01 7.28689015e-01
1.93729535e-01 1.81860104e-01 2.43203416e-01 5.02853394e-01
2.33965963e-01 3.57059300e-01 6.99751854e-01 -1.64686501e-01
-1.07593310e+00 3.75291765e-01 -6.10982001e-01 -3.39949161e-01
9.73044515e-01 -9.03570354e-02 -2.33734578e-01 -4.07594800e-01
-5.02131820e-01 5.25794268e-01 3.96058708e-02 4.09842342e-01
2.26881415e-01 -1.61587155e+00 -5.27346194e-01 4.57586527e-01
9.18807611e-02 1.78339317e-01 3.91977996e-01 7.89612055e-01
-2.02194974e-01 8.54072988e-01 2.02707961e-01 -8.11604679e-01
-9.61680770e-01 8.51636231e-01 6.26191497e-01 -5.11359751e-01
-1.15069747e+00 4.64720875e-01 1.76033184e-01 -3.23333405e-02
6.10443830e-01 -8.88660312e-01 -2.29680777e-01 5.93747646e-02
1.03600264e+00 6.42587543e-01 2.41809621e-01 -6.75591230e-01
-5.70116341e-01 7.57375062e-02 1.67054534e-01 3.63741994e-01
1.42063522e+00 -4.44852889e-01 2.89711982e-01 6.57911599e-01
1.22913182e+00 -7.87631512e-01 -1.14621425e+00 -5.74961044e-02
5.38922399e-02 -1.21828616e-01 5.84254086e-01 -8.30733180e-01
-1.14032817e+00 7.05102682e-01 1.13185978e+00 9.25126910e-01
1.31143963e+00 2.95811729e-03 1.03167570e+00 4.14702564e-01
3.86071980e-01 -1.48361099e+00 -3.24528158e-01 1.66428313e-01
5.42478502e-01 -1.31275380e+00 7.81078264e-02 4.80253339e-01
-3.66225839e-02 2.94555783e-01 4.28990245e-01 1.90905556e-01
9.53462303e-01 5.23514211e-01 -2.17157945e-01 -4.57721204e-01
-4.90676880e-01 -4.68381852e-01 -2.65709579e-01 7.36958802e-01
7.37882331e-02 6.53195322e-01 4.62069929e-01 4.99128968e-01
-4.88142312e-01 -2.54843950e-01 1.65347140e-02 9.01463747e-01
-7.92788625e-01 -3.50853413e-01 -2.93262511e-01 4.83223796e-01
-2.87996769e-01 -2.07376838e-01 8.83888751e-02 9.03254032e-01
3.03743452e-01 1.24469078e+00 4.47206050e-01 -4.38406438e-01
3.44284952e-01 1.94435000e-01 7.76573736e-03 -7.73680061e-02
-6.23422921e-01 -2.47373730e-02 -1.11019783e-01 -6.60946608e-01
-3.16991240e-01 -3.60896409e-01 -1.15421772e+00 -8.46805334e-01
-3.30673456e-01 8.69794339e-02 5.01141608e-01 1.16126370e+00
7.89572001e-02 1.28403938e+00 8.27636719e-01 -1.18328536e+00
-3.22153956e-01 -1.33908391e+00 -1.08784318e+00 3.26412350e-01
6.03428185e-01 -5.90020359e-01 -5.97473562e-01 -5.50661504e-01] | [16.062183380126953, 7.577486038208008] |
6ed1e62a-8857-4b98-a3c4-85ae93d18425 | unsupervised-shadow-removal-using-target | 2010.01291 | null | https://arxiv.org/abs/2010.01291v2 | https://arxiv.org/pdf/2010.01291v2.pdf | Unsupervised Shadow Removal Using Target Consistency Generative Adversarial Network | Unsupervised shadow removal aims to learn a non-linear function to map the original image from shadow domain to non-shadow domain in the absence of paired shadow and non-shadow data. In this paper, we develop a simple yet efficient target-consistency generative adversarial network (TC-GAN) for the shadow removal task in the unsupervised manner. Compared with the bidirectional mapping in cycle-consistency GAN based methods for shadow removal, TC-GAN tries to learn a one-sided mapping to cast shadow images into shadow-free ones. With the proposed target-consistency constraint, the correlations between shadow images and the output shadow-free image are strictly confined. Extensive comparison experiments results show that TC-GAN outperforms the state-of-the-art unsupervised shadow removal methods by 14.9% in terms of FID and 31.5% in terms of KID. It is rather remarkable that TC-GAN achieves comparable performance with supervised shadow removal methods. | ['Xin Feng', 'Chao Tan'] | 2020-10-03 | null | null | null | null | ['shadow-removal'] | ['computer-vision'] | [ 8.41682374e-01 5.06823242e-01 2.87030607e-01 -3.83909434e-01
-5.44943988e-01 -3.21352541e-01 6.26321912e-01 -1.07778871e+00
1.10988855e-01 1.16636336e+00 -4.76905219e-02 -3.17472160e-01
4.45125222e-01 -8.47883999e-01 -8.09727907e-01 -1.26793098e+00
3.98680747e-01 5.37565649e-01 3.43410254e-01 -2.07655638e-01
-2.63172597e-01 4.59795505e-01 -1.10617912e+00 -9.62245166e-02
1.06793451e+00 7.03278005e-01 2.75617987e-01 7.26106822e-01
1.41340777e-01 7.89348185e-01 -9.53146577e-01 -8.62699226e-02
3.71237874e-01 -1.05792892e+00 -1.65527239e-01 2.44711619e-02
4.31144506e-01 -4.90228444e-01 -4.22587425e-01 5.40773749e-01
5.20781577e-01 -4.88919355e-02 7.97344983e-01 -1.42368519e+00
-7.11105824e-01 -1.09089263e-01 -6.00049138e-01 -4.55762029e-01
-3.17850858e-02 -3.43546942e-02 4.29731905e-01 -7.56560981e-01
7.27529645e-01 1.03242922e+00 7.70161211e-01 5.25905490e-01
-1.28523314e+00 -7.33954668e-01 -1.21396355e-01 -1.56563684e-01
-1.22681117e+00 -2.48000413e-01 1.00552845e+00 1.11849010e-02
4.07151639e-01 6.71353877e-01 5.52056968e-01 1.18879223e+00
5.99372387e-01 4.96043563e-01 1.92373466e+00 -5.61674356e-01
1.72723889e-01 -2.42490992e-02 -4.71824795e-01 8.29553068e-01
1.60329401e-01 5.28807640e-01 -8.07576001e-01 7.43577373e-04
6.52549088e-01 -9.57695842e-02 -3.68878424e-01 -5.21557629e-01
-7.78078258e-01 5.43569803e-01 6.26576900e-01 7.66157582e-02
1.42028723e-02 5.40984631e-01 -1.75877377e-01 1.28881976e-01
6.07377172e-01 2.31972501e-01 -1.43400297e-01 4.55096185e-01
-1.04166293e+00 -1.10317469e-01 6.61265671e-01 1.16485226e+00
1.01497447e+00 6.79337680e-01 -4.97550815e-01 4.61289495e-01
1.56025127e-01 1.28762305e+00 -3.17906626e-02 -8.11033368e-01
-5.69101721e-02 1.62912861e-01 1.15058295e-01 -7.49600053e-01
3.71019728e-02 -4.23993647e-01 -1.06477833e+00 7.95050204e-01
4.45309654e-02 -1.98032543e-01 -1.51725411e+00 1.63922703e+00
1.94447473e-01 3.16091746e-01 1.07535399e-01 7.14741766e-01
9.30262506e-01 5.86866796e-01 -3.90769005e-01 -3.92712653e-01
8.14366221e-01 -1.19001913e+00 -1.02328134e+00 -4.59729284e-01
-2.59026997e-02 -8.12283993e-01 1.25310159e+00 -6.62009604e-03
-6.84565842e-01 -3.06182027e-01 -1.28123438e+00 6.05383106e-02
-3.25417221e-01 2.49434739e-01 6.08568072e-01 9.82846081e-01
-9.79252458e-01 1.55012980e-01 -5.02891421e-01 -2.14199021e-01
5.60221195e-01 2.16382533e-01 -1.02199256e-01 -1.77284569e-01
-7.55379796e-01 6.62382245e-01 -1.38181895e-01 9.56171304e-02
-1.30218387e+00 -7.04613626e-01 -6.84617877e-01 -1.63895637e-01
6.65569007e-01 -5.85429370e-01 8.54088247e-01 -1.16755688e+00
-1.73987103e+00 7.84482300e-01 -4.35564458e-01 -2.24291071e-01
7.58658409e-01 -1.96761444e-01 -2.62958676e-01 -1.83792233e-01
7.77769685e-02 3.46775025e-01 1.33219230e+00 -2.31521702e+00
-1.74456030e-01 6.27283612e-03 -1.54655099e-01 3.95501703e-01
7.06596393e-03 -4.91717756e-01 -5.13220847e-01 -8.49357069e-01
3.47939739e-03 -1.36346114e+00 1.50230527e-01 2.69888435e-02
-9.96282160e-01 5.93623042e-01 1.70742643e+00 -7.12029934e-01
7.08760321e-01 -1.88066244e+00 -1.17295166e-03 3.03529382e-01
2.04666734e-01 1.31204784e-01 1.29732728e-01 3.77578169e-01
3.29382420e-01 -2.55432755e-01 -1.04043150e+00 -8.92904639e-01
-1.07968301e-01 7.07607210e-01 -6.72080755e-01 5.72163939e-01
-2.38728210e-01 1.17117131e+00 -8.06672573e-01 -5.72296560e-01
2.89052784e-01 6.77988768e-01 5.01090065e-02 5.67567110e-01
-3.86847630e-02 8.91941607e-01 -9.94508415e-02 9.95576262e-01
1.11228418e+00 1.32882789e-01 5.15902460e-01 -5.61843254e-02
2.69703984e-01 -1.81703910e-01 -3.24569166e-01 1.43966985e+00
-5.85041285e-01 1.21984529e+00 2.07138985e-01 -2.76115656e-01
8.53219450e-01 8.33609607e-03 2.49913633e-01 -1.08767390e+00
-2.54840553e-02 8.01560730e-02 -3.08335423e-01 2.12824866e-01
4.54050153e-01 -4.33027714e-01 -2.66183317e-01 5.17428637e-01
-1.88470379e-01 -6.18040860e-01 -5.00320137e-01 1.13811292e-01
1.19498861e+00 3.09191406e-01 6.63830852e-03 -4.47392046e-01
2.58854628e-01 -3.47255796e-01 5.88559330e-01 8.36855352e-01
2.51007169e-01 9.96249557e-01 1.67816728e-01 9.42638069e-02
-7.81177402e-01 -1.53911233e+00 1.21744648e-01 8.82390738e-01
5.24740040e-01 -2.61818003e-02 -8.96286726e-01 -9.26526904e-01
8.58207867e-02 9.66523767e-01 -8.06823432e-01 -1.56537339e-01
-6.71989262e-01 -6.09283626e-01 6.49000168e-01 3.42663497e-01
9.82316911e-01 -1.20764196e+00 -5.98843813e-01 -3.74419034e-01
-6.72481135e-02 -1.06396008e+00 -4.92284328e-01 5.43056428e-01
-4.70508307e-01 -9.20942426e-01 -8.90341341e-01 -6.55816793e-01
1.18786478e+00 4.60743189e-01 1.22211957e+00 2.82144815e-01
-3.55259776e-01 3.00088733e-01 -2.43426770e-01 -6.85418308e-01
-4.21455264e-01 -2.44841412e-01 -3.01712692e-01 1.71055600e-01
-4.50394452e-01 -8.21375906e-01 -7.18227029e-01 5.36948383e-01
-1.05853164e+00 3.11371744e-01 7.55862296e-01 1.09993875e+00
7.61587381e-01 1.42289117e-01 2.62934640e-02 -1.50209463e+00
4.38578159e-01 1.76072877e-03 -6.06245518e-01 2.16265142e-01
-1.16579270e+00 -4.55927700e-02 7.26997137e-01 -7.46926591e-02
-1.75701344e+00 2.21457377e-01 4.07953680e-01 -3.62805665e-01
-2.43690349e-02 -4.37687039e-01 -6.03814721e-01 -5.63456416e-01
5.38273692e-01 5.13643324e-01 -3.66530985e-01 -1.51835769e-01
5.26224613e-01 1.37554958e-01 8.66552711e-01 -3.62165630e-01
1.65647006e+00 9.72935557e-01 4.05879766e-01 -7.52164721e-01
-8.18365633e-01 -9.17868093e-02 -6.77963912e-01 -3.66658062e-01
9.24558520e-01 -5.74844003e-01 -2.05324873e-01 8.64409268e-01
-8.48683834e-01 -1.28411806e+00 -4.69054610e-01 -3.95936370e-01
-6.27340794e-01 2.97640055e-01 -1.62183374e-01 -7.78500378e-01
-3.94375622e-01 -7.47981369e-01 1.23558271e+00 2.66271979e-01
2.92077631e-01 -9.27973032e-01 5.42090349e-02 4.19167697e-01
5.05160749e-01 6.50773168e-01 6.45438194e-01 3.67928922e-01
-9.52307761e-01 4.39645462e-02 -2.87318498e-01 6.61907315e-01
4.83432412e-01 -4.66803551e-01 -1.29550827e+00 -3.51197481e-01
7.18609840e-02 -1.92774206e-01 1.11542308e+00 4.05084014e-01
1.14536262e+00 -2.91719168e-01 -4.12626803e-01 1.02788401e+00
1.71937740e+00 2.45015129e-01 1.28571999e+00 5.15361540e-02
1.15696704e+00 1.03334062e-01 8.42231512e-01 7.56996050e-02
3.37065384e-02 6.05796754e-01 5.14074028e-01 -8.00484478e-01
-8.22623968e-01 -3.63706797e-01 2.48288363e-01 4.18132752e-01
-2.16880202e-01 -8.64183426e-01 -4.63281423e-01 2.76637405e-01
-1.68579257e+00 -6.71779454e-01 -1.89440101e-01 2.16550183e+00
6.61370218e-01 1.25668168e-01 -5.86928248e-01 -2.89540272e-02
4.23016876e-01 5.75267911e-01 -5.58175981e-01 -2.89941043e-01
-5.13513327e-01 4.12334651e-01 1.13297594e+00 7.99288511e-01
-8.09488714e-01 1.25047851e+00 6.72054052e+00 9.61353004e-01
-9.27097976e-01 4.03921038e-01 5.22722840e-01 2.61948526e-01
-6.64075971e-01 3.07591945e-01 -2.73288190e-01 5.64175069e-01
3.85670483e-01 3.67883354e-01 6.40355825e-01 5.08827567e-01
-1.12637527e-01 -7.64417231e-01 -6.39608502e-01 7.32623875e-01
3.69863570e-01 -1.08509362e+00 -3.58287841e-01 2.42505282e-01
1.33048928e+00 -4.08228666e-01 1.08519375e-01 7.87829235e-02
3.73445362e-01 -1.30980718e+00 5.98096550e-01 7.94024825e-01
1.37761104e+00 -6.68931723e-01 6.79414749e-01 1.42956197e-01
-1.20011330e+00 3.08439076e-01 -6.62967712e-02 3.34963471e-01
2.43993893e-01 8.42751622e-01 -9.45269108e-01 6.39422655e-01
7.03735352e-01 9.81198028e-02 -4.50995356e-01 2.81903476e-01
-9.41149116e-01 7.68255949e-01 -9.06601325e-02 5.09654880e-01
-1.33976847e-01 -4.73429710e-01 5.07720768e-01 1.12081254e+00
2.04251960e-01 9.46675688e-02 -1.03896305e-01 8.98287416e-01
-2.16677338e-01 -5.96050143e-01 -7.94724941e-01 2.61012793e-01
3.29135001e-01 1.19846296e+00 -1.05075896e+00 -3.25391918e-01
1.60468757e-01 1.83152664e+00 -1.86728224e-01 6.19145930e-01
-1.26165915e+00 -3.47741932e-01 3.24647069e-01 1.72297031e-01
3.07728082e-01 -1.48037419e-01 -5.59473753e-01 -6.45185113e-01
8.31231996e-02 -5.82067966e-01 -2.69267172e-01 -1.28365123e+00
-8.75384212e-01 5.55246353e-01 -7.67117292e-02 -9.89184916e-01
-3.42468359e-02 -1.75106466e-01 -8.94869506e-01 8.27794611e-01
-1.73473322e+00 -1.79153764e+00 -1.07312191e+00 7.69897640e-01
3.49177629e-01 -2.03690559e-01 1.00106943e+00 -1.68951482e-01
-1.35682732e-01 7.39656150e-01 5.03188789e-01 -1.22270413e-01
9.97470677e-01 -1.44546509e+00 3.58989835e-01 1.09587455e+00
-1.61862195e-01 -4.32938971e-02 8.63551915e-01 -1.00803959e+00
-1.36648774e+00 -1.15845644e+00 4.73579973e-01 -4.02278543e-01
-2.34003719e-02 -7.45910585e-01 -5.71462154e-01 5.37080884e-01
4.56820369e-01 9.41134319e-02 3.20948184e-01 -3.86161894e-01
-3.80139977e-01 -4.28033918e-01 -1.40466225e+00 6.64499879e-01
1.27858782e+00 -7.23780751e-01 1.92267261e-02 4.81018215e-01
7.91306198e-01 -5.91670036e-01 -3.67423117e-01 5.45845509e-01
4.13959831e-01 -1.44714415e+00 9.93932068e-01 3.21166962e-01
3.54990035e-01 -5.05191624e-01 -2.54333347e-01 -1.24222589e+00
-1.38406500e-01 -8.51670384e-01 -2.35935688e-01 1.28816640e+00
2.58833580e-02 -8.57780695e-01 9.56507206e-01 -2.85674203e-02
-5.69188535e-01 -7.31346548e-01 -7.82669961e-01 -9.34947908e-01
-2.39731714e-01 -3.72212939e-02 4.50132549e-01 5.58220208e-01
-1.07841241e+00 2.22129583e-01 -9.15818691e-01 2.78974295e-01
1.12037051e+00 5.59987426e-01 1.14169466e+00 -8.00622940e-01
-4.77319509e-01 1.50378376e-01 9.28521007e-02 -7.54806817e-01
3.97825867e-01 -4.74200368e-01 6.67347133e-01 -1.72683501e+00
2.91084886e-01 -6.50601029e-01 -1.41372547e-01 6.33484304e-01
-2.11659875e-02 9.07793641e-01 1.05994523e-01 3.50787193e-01
-2.65525818e-01 8.71773481e-01 1.58901238e+00 -2.51073748e-01
-1.88321099e-01 6.39728680e-02 -3.74992639e-01 4.73135948e-01
7.85906553e-01 -6.90443337e-01 -7.52562344e-01 -2.23584086e-01
-3.63888502e-01 -1.37165755e-01 5.94511271e-01 -9.95392978e-01
-1.79036304e-01 -3.73414665e-01 5.45300364e-01 -7.66878486e-01
9.14129615e-01 -8.87465656e-01 4.28454459e-01 3.95252824e-01
2.89391935e-01 -6.37945533e-01 5.59352487e-02 8.24316144e-01
1.02268919e-01 3.81161422e-01 8.38608325e-01 1.89332422e-02
-5.34852147e-01 -3.63290496e-02 -3.50967318e-01 -9.54662859e-02
1.06910002e+00 -4.52302307e-01 -5.59348762e-01 -7.58871794e-01
-3.00255150e-01 -2.60835290e-01 8.86997104e-01 -2.20422186e-02
6.72007501e-01 -1.27262330e+00 -3.99512202e-01 3.28793764e-01
-9.29462463e-02 2.16208726e-01 2.40546018e-01 6.46354616e-01
-5.50868809e-01 3.05554748e-01 -3.52829307e-01 -4.34443146e-01
-1.42054439e+00 -4.05794382e-02 3.40639323e-01 -3.81908745e-01
-7.47886002e-01 8.03403974e-01 9.29198027e-01 -4.55989182e-01
3.01594287e-02 8.35051388e-02 6.96579933e-01 -5.97188771e-01
-1.98604926e-01 3.11732858e-01 4.64937612e-02 -6.26903176e-01
-3.35350573e-01 4.10428256e-01 5.94766319e-01 -1.46776602e-01
1.14343393e+00 -9.18874051e-03 -2.44923577e-01 2.49660730e-01
8.73001575e-01 5.69771290e-01 -1.66672754e+00 2.94753145e-02
-6.69575632e-01 -6.49381280e-01 -1.39004756e-02 -1.13563347e+00
-1.37249911e+00 3.53918284e-01 8.84309292e-01 -1.37534931e-01
1.54629672e+00 2.81118695e-02 8.83927703e-01 1.85862437e-01
4.28510398e-01 -1.03314972e+00 3.19971323e-01 4.04195637e-01
1.15578568e+00 -1.06303084e+00 4.43681508e-01 -7.92197406e-01
-6.33291066e-01 5.97313821e-01 5.65114677e-01 -3.23566794e-01
5.77145934e-01 7.17821777e-01 1.37607351e-01 -2.08767816e-01
-9.80981514e-02 -1.71227068e-01 4.63196605e-01 1.11483169e+00
-6.58519939e-03 4.30737823e-01 3.64063941e-02 -2.92585474e-02
-3.08276802e-01 -4.78959978e-01 4.32526916e-01 1.05765378e+00
-8.96267965e-02 -1.32617307e+00 -6.40567124e-01 3.28933418e-01
1.36953324e-01 -1.56584218e-01 -9.21850324e-01 1.01565397e+00
2.48800606e-01 8.79169643e-01 -1.89313993e-01 -4.65002656e-01
-7.37291053e-02 -9.25102010e-02 6.10426605e-01 -4.30935174e-01
-1.98495671e-01 2.29602754e-01 1.18325487e-01 -5.87681890e-01
-4.24133867e-01 -2.11846054e-01 -1.29950213e+00 -2.65640140e-01
-4.01348829e-01 -2.84029335e-01 6.33344471e-01 9.24730241e-01
1.94857225e-01 1.02837801e+00 7.08348334e-01 -1.07780039e+00
2.01435223e-01 -7.40284026e-01 -7.44436800e-01 1.15080044e-01
3.98667097e-01 -7.33723223e-01 -4.98787493e-01 1.56894416e-01] | [10.845298767089844, -4.103167533874512] |
67fcec12-b54e-46eb-b928-581328b0b69f | language-models-are-weak-learners | 2306.14101 | null | https://arxiv.org/abs/2306.14101v1 | https://arxiv.org/pdf/2306.14101v1.pdf | Language models are weak learners | A central notion in practical and theoretical machine learning is that of a $\textit{weak learner}$, classifiers that achieve better-than-random performance (on any given distribution over data), even by a small margin. Such weak learners form the practical basis for canonical machine learning methods such as boosting. In this work, we illustrate that prompt-based large language models can operate effectively as said weak learners. Specifically, we illustrate the use of a large language model (LLM) as a weak learner in a boosting algorithm applied to tabular data. We show that by providing (properly sampled according to the distribution of interest) text descriptions of tabular data samples, LLMs can produce a summary of the samples that serves as a template for classification and achieves the aim of acting as a weak learner on this task. We incorporate these models into a boosting approach, which in some settings can leverage the knowledge within the LLM to outperform traditional tree-based boosting. The model outperforms both few-shot learning and occasionally even more involved fine-tuning procedures, particularly for tasks involving small numbers of data points. The results illustrate the potential for prompt-based LLMs to function not just as few-shot learners themselves, but as components of larger machine learning pipelines. | ['J Zico Kolter', 'Yiding Jiang', 'Hariharan Manikandan'] | 2023-06-25 | null | null | null | null | ['few-shot-learning'] | ['methodology'] | [-6.19160496e-02 2.55602300e-01 -6.05401278e-01 -5.66380501e-01
-1.52649355e+00 -4.98974741e-01 9.90680814e-01 5.88858187e-01
-3.95197451e-01 6.99414790e-01 2.53722787e-01 -8.16707730e-01
1.01083949e-01 -9.07779455e-01 -8.58088493e-01 -6.96449161e-01
1.27488211e-01 6.37440860e-01 2.28338927e-01 -3.89611572e-01
1.96294457e-01 1.02634169e-01 -1.75611615e+00 7.79391050e-01
7.03222811e-01 8.82630825e-01 1.75771192e-01 7.02301741e-01
-6.59773111e-01 1.25514543e+00 -6.90690935e-01 -4.82747346e-01
-1.22802153e-01 -2.14512035e-01 -5.55154085e-01 -1.76310852e-01
5.47641873e-01 -1.84297383e-01 1.38316333e-01 6.91197395e-01
4.84941721e-01 2.99475849e-01 6.98769033e-01 -1.00373411e+00
-2.25569129e-01 1.02218163e+00 -4.02207196e-01 2.05270931e-01
3.45451146e-01 1.62378192e-01 1.27391386e+00 -1.15973544e+00
2.07905293e-01 1.32569361e+00 1.04480243e+00 6.90737486e-01
-1.45804584e+00 -6.34372056e-01 2.39323024e-02 6.90509006e-02
-9.58826721e-01 -6.83863640e-01 4.90960211e-01 -5.22921801e-01
7.95135200e-01 4.99460787e-01 3.53960931e-01 9.70081151e-01
1.32321462e-01 1.25489366e+00 9.50608492e-01 -8.16552401e-01
6.76473439e-01 5.21065652e-01 7.93125153e-01 5.62107444e-01
9.01414007e-02 1.00760140e-01 -8.52390468e-01 -6.56756878e-01
-4.31372859e-02 -5.81481270e-02 1.62537675e-02 -4.38391089e-01
-6.77397013e-01 1.23331630e+00 3.74280602e-01 7.49076530e-02
-1.35733098e-01 1.36891216e-01 6.19698524e-01 2.35259295e-01
8.65327954e-01 3.05086255e-01 -7.10708678e-01 -2.65561342e-01
-9.87426639e-01 4.18893605e-01 7.79970229e-01 7.52125919e-01
8.63653541e-01 -1.17539167e-01 -3.76921713e-01 9.89786804e-01
2.30322499e-02 9.32929367e-02 6.90068364e-01 -3.63885880e-01
3.46004665e-01 4.02393997e-01 2.19862670e-01 -3.24713103e-02
-2.55370766e-01 -4.95624542e-01 -3.96522284e-01 2.41480857e-01
5.49794316e-01 -9.32503343e-02 -9.15106773e-01 1.72884047e+00
5.26781738e-01 7.07205385e-02 -1.93604290e-01 4.09579754e-01
6.07201278e-01 7.80104756e-01 4.61731672e-01 -2.90837318e-01
1.09150732e+00 -7.83618152e-01 -4.31005567e-01 -4.83569324e-01
1.28649974e+00 -5.00358343e-01 1.53729486e+00 5.45741320e-01
-9.69412744e-01 -5.15010118e-01 -8.74678075e-01 6.99537173e-02
-4.76789743e-01 -4.44290519e-01 8.20349455e-01 9.23398852e-01
-8.81898582e-01 4.93070304e-01 -4.30323094e-01 -2.70970494e-01
5.41088760e-01 -1.10529875e-03 5.42142577e-02 -3.13036352e-01
-1.05232704e+00 1.06087077e+00 3.30900908e-01 -3.72946173e-01
-8.24471176e-01 -1.04622066e+00 -8.32481086e-01 2.67967302e-02
1.84535831e-01 -4.89516139e-01 1.83013487e+00 -9.29372609e-01
-9.94120002e-01 9.38571274e-01 -3.14018577e-01 -7.26031125e-01
5.57954192e-01 -1.72650680e-01 5.30313812e-02 -3.96146238e-01
5.64248897e-02 3.72540474e-01 9.62069750e-01 -1.23021150e+00
-1.09003854e+00 -5.96665740e-01 2.98998449e-02 -2.80209146e-02
-2.37897664e-01 1.42798573e-01 -6.05085492e-02 -4.50215697e-01
-2.39864379e-01 -4.89243954e-01 -3.47617775e-01 -2.67884791e-01
6.43211529e-02 -7.17054486e-01 5.57412088e-01 -4.41170424e-01
1.35556626e+00 -1.96718061e+00 -3.79794717e-01 -4.32333397e-03
4.91891652e-02 3.31605673e-01 3.72621068e-03 4.57295686e-01
-1.02577336e-01 5.53827733e-02 -2.77313646e-02 -3.34266663e-01
4.48310114e-02 8.09366181e-02 -6.90836728e-01 4.11005527e-01
-9.45118442e-03 1.06473672e+00 -9.40230787e-01 -2.09649384e-01
2.92922914e-01 -1.04062045e-02 -3.68757129e-01 3.52121562e-01
-5.43908477e-01 -1.56929851e-01 -9.90060717e-02 3.02978128e-01
3.47424746e-01 -1.87824056e-01 -2.02666459e-04 3.46552104e-01
7.03696311e-02 5.65703809e-01 -8.92610550e-01 1.35035717e+00
-7.13252783e-01 4.95066524e-01 2.40146443e-01 -1.31386697e+00
8.75029445e-01 1.18868247e-01 8.20101127e-02 -4.22515929e-01
-3.56994681e-02 2.65157551e-01 -9.57041420e-03 -1.44155830e-01
2.49638930e-01 -7.23107398e-01 -1.89776018e-01 7.04675317e-01
2.00657770e-01 -1.89188913e-01 2.18062684e-01 3.73861760e-01
1.08429289e+00 -9.39789116e-02 6.41036332e-01 -4.42676425e-01
3.49054068e-01 7.76406452e-02 2.23129585e-01 1.52737856e+00
-1.26549369e-02 2.52224684e-01 3.03138435e-01 -7.50091255e-01
-1.15283036e+00 -8.82032394e-01 -5.32505512e-01 2.13781548e+00
-3.86345476e-01 -6.89373910e-01 -5.30362070e-01 -8.90970409e-01
3.54186833e-01 1.22335076e+00 -7.99037337e-01 -2.36474186e-01
-2.13774413e-01 -9.29444075e-01 2.86783904e-01 5.86842120e-01
-2.58467138e-01 -9.06730175e-01 -5.84114730e-01 3.56334239e-01
2.17080906e-01 -4.31008726e-01 -3.37111093e-02 1.03868878e+00
-8.56411099e-01 -6.69484258e-01 -3.55677426e-01 -5.47090530e-01
3.41978371e-01 2.26240784e-01 1.42428458e+00 5.59964888e-02
-2.70768106e-01 3.06727916e-01 -5.01205802e-01 -8.30859840e-01
-7.71220148e-01 6.43440261e-02 -8.08349624e-02 -2.43423447e-01
8.36605191e-01 -3.36182415e-01 -1.44263268e-01 1.31755963e-01
-7.07201242e-01 7.16073439e-02 3.75541061e-01 1.39132702e+00
-8.74782205e-02 -2.01785639e-01 7.56375194e-01 -1.39503300e+00
8.33143353e-01 -6.80822492e-01 -4.04257089e-01 3.64123523e-01
-5.84202945e-01 2.86453247e-01 8.16799045e-01 -5.52878261e-01
-1.00583386e+00 -2.87612736e-01 -3.53350341e-01 1.86073780e-02
-1.17929801e-01 7.93359876e-01 -7.85406213e-03 2.24269629e-01
1.26597834e+00 3.35089058e-01 -9.52055864e-03 -5.59820116e-01
4.72365201e-01 1.14471436e+00 1.23173058e-01 -9.51805115e-01
5.88530660e-01 1.95910588e-01 -3.62244785e-01 -7.40976810e-01
-1.45217764e+00 -7.57430911e-01 -4.45151985e-01 -1.42770797e-01
1.41179010e-01 -9.22933936e-01 -3.71117860e-01 2.18486160e-01
-8.15864801e-01 -7.45760739e-01 -5.11761367e-01 6.80431277e-02
-5.71769297e-01 -1.47221342e-01 -4.12398279e-01 -1.28938794e+00
-4.52498436e-01 -5.98433137e-01 1.02797282e+00 -3.92578095e-02
-4.82664943e-01 -1.08465588e+00 -8.24540779e-02 2.66997933e-01
2.96782196e-01 -3.59612256e-01 1.27913535e+00 -1.26711953e+00
-1.54154554e-01 -5.87509573e-01 2.01626971e-01 2.97048122e-01
-2.00541019e-01 -2.81766295e-01 -1.31312537e+00 -3.87518674e-01
1.43760338e-01 -9.36329901e-01 9.97570336e-01 4.73289341e-01
1.28207624e+00 -1.61278501e-01 -2.24572226e-01 4.52196479e-01
1.21908772e+00 -4.36355434e-02 2.61977851e-01 3.41824651e-01
2.65236974e-01 7.85222411e-01 8.26043069e-01 5.62610686e-01
6.13413155e-02 4.47159767e-01 -4.40764502e-02 5.86195998e-02
3.17070149e-02 -5.40533662e-01 3.47856343e-01 5.95265806e-01
4.88599151e-01 1.96105972e-01 -1.27280879e+00 5.29060781e-01
-2.10539460e+00 -1.03173649e+00 -8.84301737e-02 2.48092794e+00
1.19667959e+00 5.51143825e-01 2.32650086e-01 6.55671507e-02
5.13076365e-01 1.83076844e-01 -6.56611919e-01 -5.36855042e-01
2.17947930e-01 5.06978929e-01 3.89518887e-01 6.52250171e-01
-9.53784525e-01 7.59201407e-01 6.85162020e+00 1.39083040e+00
-8.69172335e-01 3.67223710e-01 9.80758607e-01 -2.28189617e-01
-4.83077824e-01 6.71851113e-02 -1.19424927e+00 4.50074881e-01
1.34450376e+00 -4.76992100e-01 4.21484381e-01 1.19121635e+00
9.64768827e-02 -2.62035161e-01 -1.37015176e+00 9.62521911e-01
-1.13224015e-02 -1.55467188e+00 9.49375406e-02 -1.33313894e-01
5.45192361e-01 9.05222893e-02 -1.16577342e-01 1.08026528e+00
8.42932642e-01 -1.20138705e+00 8.59037757e-01 -5.33332080e-02
6.78542912e-01 -7.69241631e-01 5.68682492e-01 1.25248814e+00
-7.71713674e-01 -5.15822828e-01 -4.95568156e-01 -3.74664843e-01
-2.68279910e-01 6.71108067e-01 -9.77688193e-01 1.23706914e-01
5.30989230e-01 3.94124657e-01 -5.20499289e-01 7.92690098e-01
-9.83212963e-02 1.04516435e+00 -1.96065620e-01 -5.32733500e-01
2.30706498e-01 2.03038260e-01 1.32756338e-01 1.37142336e+00
3.84373739e-02 3.12779732e-02 3.26884657e-01 4.13287103e-01
-5.56835681e-02 3.16301286e-01 -6.18331611e-01 2.18846574e-01
5.83165050e-01 1.20067096e+00 -2.82271743e-01 -7.90484965e-01
-5.50093830e-01 1.00048572e-01 7.81403065e-01 1.57734320e-01
-4.04270738e-01 1.07541047e-02 4.84416336e-01 2.72214711e-01
2.33069360e-01 1.77129418e-01 -7.33684123e-01 -1.20613205e+00
-1.92326620e-01 -1.24962425e+00 6.45437717e-01 -6.85523272e-01
-1.58728600e+00 1.17662027e-01 -8.13251585e-02 -8.04048598e-01
-5.58286726e-01 -7.31810927e-01 -7.26244628e-01 1.08738124e+00
-1.12282681e+00 -1.10526097e+00 -6.37006536e-02 4.01242077e-01
9.30018783e-01 -2.14930460e-01 9.51292753e-01 -3.36463541e-01
-2.04683453e-01 5.06949365e-01 4.56936121e-01 -3.99595276e-02
7.33030617e-01 -1.54253459e+00 4.46141005e-01 6.09283924e-01
3.94922942e-01 8.48119617e-01 1.02496147e+00 -5.40647149e-01
-1.33887970e+00 -9.75148022e-01 7.91773081e-01 -8.35774541e-01
8.55258226e-01 -8.66456389e-01 -1.09344196e+00 7.15646803e-01
-3.74004394e-02 1.40341684e-01 9.50113595e-01 9.59852397e-01
-5.74749291e-01 -3.87243241e-01 -9.91661012e-01 3.94366711e-01
6.71727300e-01 -5.66599429e-01 -9.38956916e-01 3.05243701e-01
3.76235425e-01 -6.25004321e-02 -3.65522504e-01 1.62175879e-01
7.16109693e-01 -1.01789510e+00 8.07032585e-01 -1.18576920e+00
4.55064893e-01 2.49615505e-01 -2.04855472e-01 -1.55768263e+00
-1.57522976e-01 -6.56014025e-01 -3.64170104e-01 1.11754513e+00
4.06956762e-01 -4.83945489e-01 8.01525772e-01 9.10530329e-01
1.23852029e-01 -7.26923585e-01 -9.31118488e-01 -7.83874571e-01
6.04246378e-01 -1.01521111e+00 3.55391860e-01 7.50315547e-01
3.96564424e-01 6.32547915e-01 -2.62351543e-01 -4.68145013e-01
8.42617869e-01 1.19067514e-02 1.12278342e+00 -1.29699886e+00
-3.67358387e-01 -3.96500885e-01 -1.22231409e-01 -1.16958928e+00
2.73250759e-01 -1.09292877e+00 4.38800842e-01 -1.16144049e+00
5.03082514e-01 -8.54236960e-01 -4.93897468e-01 3.97424459e-01
-6.27189755e-01 -2.08844513e-01 1.96149126e-02 -1.51604023e-02
-3.98881555e-01 2.45094642e-01 5.46014190e-01 -1.73599035e-01
-1.03703715e-01 3.19051057e-01 -1.06598663e+00 7.55792141e-01
6.00650668e-01 -5.13764918e-01 -3.29260737e-01 2.87020020e-02
2.02502951e-01 1.25338867e-01 6.27725869e-02 -6.30146742e-01
3.02962720e-01 -1.71088740e-01 4.11203593e-01 -5.01788378e-01
1.06587984e-01 -4.22662944e-01 -3.21279973e-01 3.39735419e-01
-9.41423237e-01 -4.50054526e-01 1.63356632e-01 4.88247097e-01
-3.06734350e-02 -6.39597237e-01 8.99290264e-01 -3.94870102e-01
-5.88430643e-01 1.42419949e-01 -3.12202275e-01 2.04415560e-01
9.55343902e-01 1.40791833e-01 -3.50479215e-01 -4.80406076e-01
-4.24995363e-01 4.10147831e-02 3.06099862e-01 2.63166487e-01
2.33300418e-01 -1.18498814e+00 -8.86859953e-01 3.07178766e-01
4.27276552e-01 -5.45620918e-02 -2.33220115e-01 4.93820906e-01
8.99215564e-02 5.04020154e-01 2.47593164e-01 -6.32640064e-01
-1.21686816e+00 7.18315601e-01 1.39902607e-01 -2.89930820e-01
-4.60650623e-01 1.02420580e+00 4.18453217e-01 -6.63206577e-01
4.15526450e-01 -3.39282304e-03 2.63960898e-01 3.10029864e-01
9.97487843e-01 4.22772998e-03 3.13918084e-01 -1.83040816e-02
-2.43769124e-01 -2.32986480e-01 -2.71500051e-01 -2.47334823e-01
1.51165473e+00 1.81717426e-01 2.72483081e-02 1.07677770e+00
8.91736329e-01 1.64962620e-01 -1.19492960e+00 -4.68676925e-01
4.10076261e-01 -5.85495174e-01 6.97796121e-02 -9.99597788e-01
-4.87377308e-02 1.15386248e+00 1.36090234e-01 4.86662984e-01
7.39906192e-01 2.52559453e-01 2.84040451e-01 3.64363313e-01
5.51730752e-01 -9.59557772e-01 -1.89010203e-01 4.83512640e-01
4.82557267e-01 -1.40649354e+00 -1.76172554e-01 1.43704772e-01
-4.48335856e-01 1.19413340e+00 4.60428506e-01 7.67070726e-02
5.30797601e-01 5.18512905e-01 2.01529831e-01 2.05874711e-01
-1.60027277e+00 -2.23328352e-01 2.25518882e-01 7.42067695e-01
7.57418036e-01 2.37883106e-02 -2.72614449e-01 7.13754475e-01
-2.24388197e-01 1.85156446e-02 1.99750587e-01 9.70746636e-01
-1.09115028e+00 -1.09590006e+00 -4.97187942e-01 9.44466889e-01
-3.51480126e-01 -4.26601946e-01 -4.58979905e-02 5.54426134e-01
-4.27934043e-02 9.70470965e-01 7.54273608e-02 -2.32655361e-01
-7.06772581e-02 7.23958313e-01 2.98151910e-01 -9.78470504e-01
-7.25675702e-01 -1.41173571e-01 3.23940516e-01 -1.76984280e-01
1.28313854e-01 -7.87463129e-01 -6.93993926e-01 -3.86800915e-01
-2.86845773e-01 4.03335601e-01 7.12119162e-01 1.11803067e+00
-1.97189465e-01 5.03431670e-02 6.93561137e-01 -6.18554115e-01
-1.33474767e+00 -1.28196657e+00 -6.05242908e-01 5.42393088e-01
4.75782365e-01 -3.74498844e-01 -5.11365116e-01 -9.09367427e-02] | [10.8049898147583, 8.103851318359375] |
1d1a47bc-388e-44e5-b82a-1821d18d6e16 | image-to-gps-verification-through-a-bottom-up | 1811.07288 | null | http://arxiv.org/abs/1811.07288v1 | http://arxiv.org/pdf/1811.07288v1.pdf | Image-to-GPS Verification Through A Bottom-Up Pattern Matching Network | The image-to-GPS verification problem asks whether a given image is taken at
a claimed GPS location. In this paper, we treat it as an image verification
problem -- whether a query image is taken at the same place as a reference
image retrieved at the claimed GPS location. We make three major contributions:
1) we propose a novel custom bottom-up pattern matching (BUPM) deep neural
network solution; 2) we demonstrate that the verification can be directly done
by cross-checking a perspective-looking query image and a panorama reference
image, and 3) we collect and clean a dataset of 30K pairs query and reference.
Our experimental results show that the proposed BUPM solution outperforms the
state-of-the-art solutions in terms of both verification and localization. | ['Prem Natarajan', 'Wael Abd-Almageed', 'Jiaxin Cheng', 'Yue Wu'] | 2018-11-18 | null | null | null | null | ['image-to-gps-verification'] | ['computer-vision'] | [ 3.10724974e-01 -2.26996750e-01 -3.21809202e-01 -4.93296415e-01
-1.35147214e+00 -6.73472345e-01 6.52796447e-01 -1.12026319e-01
-3.55741829e-01 2.01064751e-01 -1.84603691e-01 -4.50770587e-01
-5.23582436e-02 -6.84165120e-01 -1.45649040e+00 -5.35364330e-01
-1.64673060e-01 1.01439901e-01 3.78754079e-01 1.95580259e-01
4.07313943e-01 7.35126972e-01 -1.15446544e+00 1.12687670e-01
2.53936082e-01 1.60036612e+00 6.20158529e-03 6.74849093e-01
6.39775515e-01 8.38223219e-01 -3.23481232e-01 -7.05533564e-01
7.05141187e-01 -2.16885633e-03 -9.72718239e-01 1.92874834e-01
1.25111365e+00 -6.08901799e-01 -9.26958203e-01 1.21258867e+00
3.61205667e-01 -9.00773481e-02 1.12872757e-01 -1.49941742e+00
-1.18684244e+00 1.57763585e-02 -8.45108867e-01 1.95213065e-01
5.01073420e-01 -1.13093346e-01 7.50035107e-01 -1.00788856e+00
6.92995489e-01 8.07671785e-01 1.05804491e+00 5.95001020e-02
-8.49186540e-01 -7.61927366e-01 -1.41617104e-01 4.53445822e-01
-1.94064200e+00 -5.92302203e-01 4.33101058e-01 -7.73951635e-02
7.13589191e-01 1.30815804e-01 1.99036106e-01 8.01539660e-01
3.35602403e-01 6.87714696e-01 9.57140386e-01 -2.84503818e-01
-1.19285211e-01 -5.59498817e-02 -2.88367659e-01 8.25668633e-01
9.09288451e-02 4.15202260e-01 -4.96774107e-01 -1.25576109e-01
5.41370749e-01 2.31853034e-02 -4.87229258e-01 -6.81127369e-01
-1.34997368e+00 4.65634704e-01 9.39072311e-01 3.76086205e-01
-3.53601813e-01 5.54766238e-01 6.10483997e-02 3.16411942e-01
1.15352934e-02 2.98624635e-01 -2.88531959e-01 1.05136223e-01
-1.36731684e+00 1.90113664e-01 5.39311349e-01 1.36580813e+00
9.65604365e-01 -1.96378767e-01 6.23886846e-02 3.31644356e-01
2.35070661e-01 1.04580092e+00 2.24880680e-01 -8.56279016e-01
6.62784100e-01 2.14414075e-01 5.14633238e-01 -1.54682302e+00
-7.01287612e-02 -3.93811874e-02 -6.82208002e-01 -5.90104461e-02
7.48308674e-02 -2.75250361e-03 -7.67625451e-01 1.35236752e+00
2.78113127e-01 4.93500531e-01 -1.60172973e-02 9.90245402e-01
8.88338685e-01 5.98542631e-01 -4.34842467e-01 4.09721464e-01
1.39184844e+00 -1.14157677e+00 -2.80191869e-01 -3.81972849e-01
3.44617963e-01 -8.18358243e-01 3.51248533e-01 -1.50610935e-02
-8.33917797e-01 -8.72913003e-01 -1.22209227e+00 -7.74754435e-02
-4.63332802e-01 4.23693061e-01 2.11034745e-01 6.71053350e-01
-1.46580672e+00 4.82489586e-01 -5.13178051e-01 -4.26230282e-01
2.28657052e-01 4.68038738e-01 -1.00179696e+00 -3.25914651e-01
-1.11336982e+00 6.90573514e-01 1.62614644e-01 2.93341160e-01
-8.85812461e-01 -3.74389976e-01 -1.12097919e+00 1.10097438e-01
3.08076739e-01 -6.45602703e-01 1.44507051e+00 -7.57689178e-01
-8.92651260e-01 1.33243334e+00 -4.94798005e-01 -7.23775685e-01
4.69798356e-01 -7.32415244e-02 -7.60963976e-01 2.66599476e-01
5.69154978e-01 7.86212206e-01 9.70738471e-01 -1.20963728e+00
-1.10887706e+00 -2.89579391e-01 -1.52522683e-01 -1.10615544e-01
4.58407223e-01 1.07284285e-01 -1.14309371e+00 -5.16600132e-01
4.10890877e-01 -1.13154531e+00 -2.36191098e-02 2.33170465e-01
-5.37409246e-01 4.52484220e-01 9.39539790e-01 -5.77402830e-01
7.78738022e-01 -2.23768735e+00 -6.85567915e-01 4.13188815e-01
-8.11256543e-02 3.27765167e-01 -1.18915729e-01 5.70301652e-01
-2.43203968e-01 -3.46894190e-02 2.46956432e-03 -6.00654781e-01
9.33003053e-02 1.57741308e-02 -9.79725599e-01 8.84608984e-01
-7.57888481e-02 1.30989897e+00 -7.98343182e-01 -5.10364294e-01
1.94343239e-01 3.09510320e-01 2.21208557e-02 1.80520639e-02
3.73116076e-01 2.64623851e-01 -1.47741744e-02 1.05272281e+00
1.32332563e+00 -5.34248292e-01 -2.53475290e-02 -2.21703351e-01
8.79007578e-02 1.72093809e-01 -1.20321167e+00 1.63133466e+00
-3.34442705e-01 8.94533634e-01 -9.43381526e-03 -5.24604678e-01
8.63868833e-01 1.97191462e-01 1.54213965e-01 -1.21441245e+00
-1.03619777e-01 3.00165176e-01 -5.82013488e-01 -2.52908677e-01
1.23250830e+00 3.86953026e-01 -4.05229390e-01 3.12425584e-01
-1.22718282e-01 9.03423131e-02 -3.11393261e-01 6.11300096e-02
9.34354126e-01 -5.07824123e-02 1.34815916e-01 1.23798110e-01
6.87335253e-01 -9.05177295e-02 5.62121987e-01 1.41361606e+00
-3.13363224e-01 8.95192266e-01 2.69523948e-01 -6.51969671e-01
-1.06535447e+00 -9.95651603e-01 -4.15303111e-02 6.33317173e-01
6.75930679e-01 -1.44205377e-01 -4.36729968e-01 -6.88821971e-01
2.45585263e-01 -1.25155353e-03 -5.94136000e-01 2.86694258e-01
-5.47005296e-01 -7.26182237e-02 8.70047033e-01 5.58301330e-01
1.13908231e+00 -7.15750456e-01 -4.15528685e-01 -3.05832714e-01
-2.08565027e-01 -1.42605472e+00 -9.88264441e-01 -3.26858908e-01
-3.70209515e-01 -1.28531325e+00 -8.02203000e-01 -1.14403141e+00
7.82275200e-01 1.00163639e+00 9.60969925e-01 3.68201971e-01
1.46322533e-01 5.07870197e-01 -1.97905764e-01 2.53479593e-02
-1.93233028e-01 -7.55358189e-02 7.24685714e-02 4.40588623e-01
3.21327060e-01 -3.41401219e-01 -7.67695010e-01 6.77730381e-01
-7.74444580e-01 -2.57225901e-01 7.03911424e-01 6.03236735e-01
9.99218702e-01 4.96766195e-02 -1.02450978e-02 -4.75607634e-01
1.94020286e-01 -1.82742923e-01 -1.16220725e+00 5.96822798e-01
-6.28346860e-01 -3.73816401e-01 9.42475200e-02 -1.23975262e-01
-3.90980810e-01 4.65520322e-01 -8.84139687e-02 -7.10497379e-01
-1.39372155e-01 1.90089807e-01 -1.42386332e-01 -9.30614829e-01
2.78124392e-01 7.72376120e-01 -2.34362230e-01 -8.57767910e-02
3.66631269e-01 7.33938456e-01 1.31198394e+00 -1.80854738e-01
1.12257755e+00 8.63957226e-01 1.46205187e-01 -4.38978940e-01
-5.56881309e-01 -8.72703493e-01 -6.08172894e-01 5.24141453e-02
5.60782075e-01 -1.20948720e+00 -9.99230742e-01 7.46602595e-01
-1.11671782e+00 -4.94473018e-02 2.09607676e-01 1.36321008e-01
-5.55225849e-01 9.17988300e-01 -4.11931366e-01 -6.13706172e-01
-4.94816482e-01 -1.23736668e+00 1.68997228e+00 3.90133768e-01
2.48780519e-01 -5.61672986e-01 -1.46711200e-01 2.19635442e-01
3.76978040e-01 1.16747096e-01 -5.24586663e-02 -3.73943478e-01
-1.27062762e+00 -7.51899302e-01 -8.08989406e-01 -1.16651922e-01
-8.07314068e-02 -2.52028942e-01 -1.00777042e+00 -4.93152171e-01
-6.99179713e-04 -1.15283772e-01 6.61992490e-01 2.49831989e-01
8.98980856e-01 -3.78778458e-01 -6.36061668e-01 1.13770795e+00
1.55962944e+00 2.06832916e-01 8.81268978e-01 8.04356694e-01
4.31634039e-01 -5.35480678e-02 8.83074760e-01 -8.40069056e-02
5.52179754e-01 9.25184131e-01 5.49021125e-01 -1.52175039e-01
1.95085377e-01 -7.38375843e-01 6.76380247e-02 -5.12488373e-02
7.08124638e-01 -1.33305609e-01 -8.47492337e-01 1.02681339e+00
-2.13345098e+00 -1.01814830e+00 -1.48610428e-01 2.19204688e+00
2.03263182e-02 -1.79906711e-01 -2.44677126e-01 -2.40399465e-01
7.95116842e-01 4.58180457e-01 -3.93589675e-01 2.15234254e-02
-2.72764713e-01 -1.13506339e-01 1.19446266e+00 4.76774782e-01
-1.72567368e+00 8.06245506e-01 6.65235615e+00 7.15573013e-01
-1.34801984e+00 1.59192950e-01 5.31837881e-01 2.86486655e-01
3.14841688e-01 2.10673045e-02 -8.80098462e-01 6.48985207e-01
6.97722793e-01 1.36720940e-01 1.18257105e-01 1.25723243e+00
-2.94228852e-01 -3.37780207e-01 -1.01867080e+00 1.41755283e+00
3.58975023e-01 -1.76809824e+00 -3.19705606e-01 2.20779508e-01
8.10360909e-01 4.84363616e-01 3.23260635e-01 5.94343320e-02
-5.22575751e-02 -7.99518049e-01 1.01988959e+00 5.22625387e-01
9.47494626e-01 -5.59318721e-01 9.25112963e-01 1.50505826e-01
-1.43622196e+00 9.33518708e-02 -3.11882555e-01 3.69640738e-01
2.51994431e-01 1.40855536e-01 -9.75730181e-01 9.03992176e-01
9.88738835e-01 4.51100081e-01 -8.30084622e-01 1.43560171e+00
-4.51197654e-01 1.19695127e-01 -2.64319062e-01 4.04147506e-01
4.97859657e-01 1.09248728e-01 3.29529762e-01 1.16131222e+00
6.05773151e-01 -4.33047593e-01 -1.31153584e-01 8.73196125e-01
-2.48174459e-01 -4.26495433e-01 -9.60527539e-01 4.86087292e-01
7.05985129e-01 1.03430390e+00 -4.56702948e-01 -1.62901446e-01
-2.54666775e-01 1.52815938e+00 -6.55449107e-02 2.20640168e-01
-1.02727079e+00 -5.28441906e-01 6.07016027e-01 -1.11056678e-01
9.78264749e-01 -1.03093259e-01 1.29589483e-01 -1.04203880e+00
5.32133639e-01 -5.66781759e-01 3.76760364e-01 -1.44611132e+00
-1.08786571e+00 7.34260678e-01 -5.41268229e-01 -1.57584941e+00
-2.48913318e-01 -5.42157412e-01 -4.68654215e-01 1.03494740e+00
-1.80440617e+00 -1.59373307e+00 -4.16677326e-01 6.39036536e-01
-2.44065106e-01 -1.14899680e-01 7.04179108e-01 7.20437169e-01
-1.43073514e-01 1.03071475e+00 3.35127473e-01 8.35948050e-01
7.02815771e-01 -8.63273740e-01 1.21781301e+00 1.39594126e+00
6.38172626e-01 5.29974282e-01 1.88344061e-01 -4.47797328e-01
-1.63004863e+00 -1.35753632e+00 1.34897220e+00 -7.46120036e-01
6.36038125e-01 -2.53826231e-01 -5.86243331e-01 8.78008485e-01
1.15410358e-01 5.33892691e-01 2.83923507e-01 -3.80153835e-01
-5.52213132e-01 -2.76835740e-01 -1.24247503e+00 2.55215436e-01
7.24791825e-01 -1.18909597e+00 -4.15471733e-01 2.95820117e-01
6.91249847e-01 -1.14820862e+00 -6.73324943e-01 2.82488227e-01
6.59638703e-01 -1.11766219e+00 1.16964555e+00 -1.66659832e-01
5.93879856e-02 -8.12669098e-01 -5.15089750e-01 -6.54836953e-01
-1.45509928e-01 -6.65181458e-01 9.52445790e-02 1.05311704e+00
2.47687653e-01 -6.41239643e-01 9.42905247e-01 5.78839362e-01
1.59659088e-01 -6.63286984e-01 -1.35109532e+00 -9.36630368e-01
-6.03013933e-01 -6.30108774e-01 1.05778861e+00 8.80417228e-01
-3.23193520e-01 -2.72540361e-01 -8.21521282e-01 9.52491224e-01
3.50024164e-01 6.88218236e-01 1.15836418e+00 -4.18565005e-01
-2.92258114e-01 -1.28963009e-01 -8.57915819e-01 -1.67550826e+00
-1.12895004e-01 -4.98179108e-01 2.21945062e-01 -1.32782888e+00
-4.05218452e-02 -3.31066936e-01 -1.18961647e-01 3.78260404e-01
1.20786510e-01 6.51316047e-01 1.91876963e-01 4.14006889e-01
-1.01351070e+00 4.89363410e-02 5.40533364e-01 -2.33337164e-01
4.24160361e-01 2.40285099e-01 -7.36162007e-01 2.53525287e-01
2.73229510e-01 -6.49204254e-01 5.25990054e-02 -6.24238491e-01
2.43863225e-01 3.63118052e-01 8.16363156e-01 -1.23081589e+00
6.45883501e-01 3.08472455e-01 3.34710956e-01 -1.08135319e+00
4.46446627e-01 -1.14008117e+00 3.31963331e-01 2.82490700e-01
1.60106614e-01 4.99046326e-01 2.46569797e-01 8.46069455e-01
-5.11355102e-01 -1.10399209e-01 5.10932326e-01 1.71501160e-01
-1.21452498e+00 5.36271214e-01 1.73397884e-01 -4.06770617e-01
8.94818187e-01 -5.00434101e-01 -5.75644791e-01 -6.52006209e-01
-1.54272243e-01 4.01437998e-01 8.73646498e-01 6.73741281e-01
6.34454787e-01 -1.67033494e+00 -2.16144606e-01 3.01924258e-01
4.11727041e-01 -9.45762098e-02 3.76163453e-01 9.12491024e-01
-7.41893172e-01 8.28318536e-01 1.63189486e-01 -8.09084892e-01
-1.41811919e+00 9.38769639e-01 6.12336099e-01 -3.73165309e-02
-5.82498610e-01 8.56261611e-01 3.91001441e-02 -6.14045441e-01
3.91750902e-01 -2.95339614e-01 5.80407381e-01 -4.77321118e-01
9.32474673e-01 -1.47575125e-01 2.10132718e-01 -1.30533242e+00
-6.32119954e-01 8.21818590e-01 1.51229516e-01 -4.41597030e-03
9.42298949e-01 -2.44718149e-01 -6.54129311e-02 -2.49755815e-01
1.50355113e+00 1.71386167e-01 -9.90844369e-01 -4.86485690e-01
3.47692257e-04 -1.05293107e+00 -2.59870708e-01 -5.80945313e-01
-1.21901214e+00 5.19678295e-01 8.91174972e-01 7.59343207e-02
1.03742039e+00 -5.27374540e-03 9.69717264e-01 6.28294349e-01
6.91170692e-01 -7.77493000e-01 -4.35585886e-01 4.52769935e-01
6.49476230e-01 -1.52473187e+00 -1.33033440e-01 4.15271148e-02
-4.84999567e-01 8.25227678e-01 3.20680112e-01 -2.06200480e-01
5.78088164e-01 -1.63977623e-01 9.36462805e-02 -2.93738127e-01
-6.99796975e-02 -1.43865421e-01 5.07621527e-01 6.23990119e-01
-2.15584308e-01 -1.87248036e-01 6.07268572e-01 1.90533876e-01
-1.31335080e-01 1.83937162e-01 3.48959625e-01 1.04777837e+00
-3.05903424e-02 -9.23564136e-01 -6.90589845e-01 -1.19932234e-01
-4.37630981e-01 -1.61689475e-01 -5.56881428e-01 9.17739511e-01
2.13868216e-01 9.12735760e-01 1.39963776e-01 -6.60111785e-01
3.82164091e-01 -2.42529675e-01 2.16603458e-01 -1.95771120e-02
-2.52690822e-01 -4.31162745e-01 -1.32223576e-01 -8.23984563e-01
-4.99666989e-01 -4.95760471e-01 -8.48262668e-01 -5.23090124e-01
-3.48814249e-01 1.24853939e-01 6.71660900e-01 6.68083906e-01
9.49460506e-01 -1.73633844e-01 6.89213276e-01 -9.94815469e-01
-2.62996584e-01 -4.66926485e-01 -4.63869333e-01 1.38290137e-01
8.44777405e-01 -3.23009521e-01 -4.11988907e-02 -1.16079018e-01] | [7.6977033615112305, -1.9689515829086304] |
2ca705e8-0c35-439e-9606-1d514826050d | meta-learning-with-maml-on-trees | 2103.04691 | null | https://arxiv.org/abs/2103.04691v1 | https://arxiv.org/pdf/2103.04691v1.pdf | Meta-Learning with MAML on Trees | In meta-learning, the knowledge learned from previous tasks is transferred to new ones, but this transfer only works if tasks are related. Sharing information between unrelated tasks might hurt performance, and it is unclear how to transfer knowledge across tasks with a hierarchical structure. Our research extends a model agnostic meta-learning model, MAML, by exploiting hierarchical task relationships. Our algorithm, TreeMAML, adapts the model to each task with a few gradient steps, but the adaptation follows the hierarchical tree structure: in each step, gradients are pooled across tasks clusters, and subsequent steps follow down the tree. We also implement a clustering algorithm that generates the tasks tree without previous knowledge of the task structure, allowing us to make use of implicit relationships between the tasks. We show that the new algorithm, which we term TreeMAML, performs better than MAML when the task structure is hierarchical for synthetic experiments. To study the performance of the method in real-world data, we apply this method to Natural Language Understanding, we use our algorithm to finetune Language Models taking advantage of the language phylogenetic tree. We show that TreeMAML improves the state of the art results for cross-lingual Natural Language Inference. This result is useful, since most languages in the world are under-resourced and the improvement on cross-lingual transfer allows the internationalization of NLP models. This results open the window to use this algorithm in other real-world hierarchical datasets. | ['Alberto Bernacchia', 'Ye Tian', 'Da-Shan Shiu', 'Tim Nieradzik', 'Jamie McGowan', 'Feng-Ting Liao', 'Federica Freddi', 'Jezabel R. Garcia'] | 2021-03-08 | null | null | null | null | ['cross-lingual-natural-language-inference'] | ['natural-language-processing'] | [ 1.85887203e-01 4.40091379e-02 -2.62678444e-01 -4.44590449e-01
-5.86921811e-01 -7.79963732e-01 7.15221524e-01 4.13620695e-02
-7.29949951e-01 9.03251231e-01 2.14384064e-01 -2.87514240e-01
-1.67031527e-01 -5.04513979e-01 -9.44525540e-01 -5.99630415e-01
-1.14262544e-01 8.12568486e-01 4.59458798e-01 -7.15246201e-02
1.99156851e-01 -3.85113843e-02 -1.54708266e+00 8.19342852e-01
1.02934515e+00 8.23048651e-02 6.82196975e-01 3.36894840e-01
-3.36076945e-01 6.67676210e-01 -4.70119685e-01 -4.09905493e-01
1.02679320e-01 -4.53705877e-01 -1.33846903e+00 -1.38702393e-01
3.97924691e-01 2.60358661e-01 3.71485084e-01 7.53318012e-01
2.01642379e-01 2.06839412e-01 8.20435286e-01 -1.25903940e+00
-4.86360401e-01 1.08495474e+00 -5.05004466e-01 -1.30429506e-01
-9.08084437e-02 -3.08458477e-01 9.49994147e-01 -5.43511450e-01
7.47052431e-01 1.46734548e+00 7.63529480e-01 7.47118235e-01
-1.59535396e+00 -7.14426994e-01 3.68118197e-01 2.66781211e-01
-1.20553899e+00 -2.84213513e-01 5.11607289e-01 -6.46237075e-01
1.10068440e+00 -1.71032429e-01 3.23599577e-01 1.13689590e+00
3.24577987e-01 7.28399098e-01 1.57063794e+00 -6.96627319e-01
1.42528385e-01 4.31504399e-01 2.20129296e-01 8.06630552e-01
2.92778254e-01 5.45480438e-02 -6.71958625e-01 -2.20347151e-01
4.64569926e-01 -4.65673596e-01 6.88152835e-02 -5.88891566e-01
-1.32203650e+00 8.34355056e-01 3.13678950e-01 7.93156505e-01
-2.79340837e-02 4.45804521e-02 6.20770991e-01 6.27246976e-01
7.14716315e-01 6.18336260e-01 -8.82926822e-01 3.92617099e-02
-9.40207422e-01 7.61467069e-02 1.01256227e+00 7.57650614e-01
1.14937687e+00 -3.36986244e-01 3.14019434e-02 1.15782511e+00
1.16016209e-01 2.24925820e-02 8.38451743e-01 -1.33803535e+00
3.51432651e-01 4.62359339e-01 -1.28795862e-01 -4.42281991e-01
-5.16312420e-01 -2.59410977e-01 -6.42810643e-01 7.98047930e-02
6.61449313e-01 -2.87297279e-01 -7.86337793e-01 2.24395275e+00
1.45456463e-01 6.07619733e-02 1.78032026e-01 3.00419331e-01
2.56840795e-01 6.44162059e-01 2.06347242e-01 -3.80299807e-01
1.34682345e+00 -9.82471585e-01 -4.60221320e-01 -4.55499500e-01
1.25042272e+00 -6.91562951e-01 1.33380747e+00 4.69520718e-01
-8.15592289e-01 -8.08182836e-01 -7.98043728e-01 -1.15547061e-01
-7.59500980e-01 -1.21285774e-01 8.08969855e-01 6.84898973e-01
-1.20147097e+00 5.68239450e-01 -8.23463261e-01 -7.36750543e-01
6.03204146e-02 3.47824067e-01 -3.78637463e-01 -1.29314467e-01
-1.28174841e+00 1.24017560e+00 9.11146402e-01 -3.61879855e-01
-7.49291301e-01 -7.18865156e-01 -8.17754805e-01 -1.13071568e-01
4.64547843e-01 -1.02870548e+00 1.27727008e+00 -1.11763251e+00
-1.34529829e+00 1.13148665e+00 -4.57432091e-01 -3.82335186e-01
2.41745204e-01 -9.34001282e-02 4.51062247e-02 -4.80501950e-01
2.39995703e-01 9.66695189e-01 6.68913901e-01 -1.54222131e+00
-7.31760919e-01 -3.88019651e-01 -5.27397729e-02 2.18247309e-01
-4.00896788e-01 -4.40382957e-02 -3.20344001e-01 -5.94375789e-01
-3.00108582e-01 -1.17484784e+00 -1.31685048e-01 -6.50371015e-01
-2.77379202e-03 -5.26121795e-01 4.02579606e-01 -4.05647427e-01
1.10537171e+00 -1.92203856e+00 4.91893858e-01 -4.58827019e-02
-6.94184452e-02 -1.78629942e-02 -2.13011384e-01 4.74223793e-01
2.60988772e-02 2.86968887e-01 -5.80904007e-01 -5.34685671e-01
-1.56055959e-02 6.67528391e-01 -1.24182649e-01 -5.31509295e-02
-2.29718849e-01 7.16129124e-01 -8.97943556e-01 -7.57919669e-01
-1.03569798e-01 7.76150730e-03 -7.69653440e-01 6.80829883e-02
-4.08144504e-01 3.46035480e-01 -2.14674175e-01 -1.50249507e-02
4.76536304e-01 -1.80820763e-01 6.05637431e-01 -1.59978718e-01
-2.02318028e-01 5.21070123e-01 -8.25974405e-01 2.05342507e+00
-8.22850585e-01 5.50873339e-01 1.69116892e-02 -1.08707368e+00
6.62493646e-01 2.62634754e-01 2.65479207e-01 -4.16365862e-01
-3.13398987e-01 1.07171543e-01 3.46301168e-01 -3.84642750e-01
2.72486717e-01 -5.02713561e-01 -1.98626518e-01 7.40420043e-01
2.68407911e-01 -3.61675978e-01 4.39449877e-01 3.65736276e-01
7.65412390e-01 4.23705935e-01 3.79578859e-01 -5.63818872e-01
3.34680438e-01 2.59066492e-01 4.19720054e-01 9.10279989e-01
1.19708680e-01 -8.54640827e-03 3.59971792e-01 -2.58108616e-01
-9.76793051e-01 -1.08037221e+00 -3.42878163e-01 1.95341921e+00
-3.53981763e-01 -5.59404194e-01 -7.98930585e-01 -8.66238236e-01
7.04491064e-02 9.24163640e-01 -9.37425852e-01 -1.15745522e-01
-5.48863471e-01 -1.01102281e+00 6.01409078e-01 3.25324386e-01
3.81366432e-01 -1.26181996e+00 -3.67696971e-01 3.22443306e-01
-4.15260226e-01 -1.03565574e+00 -2.03928083e-01 5.62660635e-01
-1.05776393e+00 -8.94007683e-01 -4.37300593e-01 -9.78728533e-01
3.77449393e-01 1.89712215e-02 1.43757665e+00 -1.21163987e-01
8.46893489e-02 2.86090791e-01 -2.68293798e-01 -4.66164201e-01
-8.55375886e-01 7.43244886e-01 1.72617480e-01 -2.02115312e-01
3.89465988e-01 -6.88310564e-01 1.69062302e-01 4.41044271e-01
-8.98189962e-01 3.37448150e-01 5.14977992e-01 9.67111230e-01
3.39104980e-01 1.37581885e-01 7.65971184e-01 -1.47978902e+00
7.29153693e-01 -4.79383945e-01 -5.68411529e-01 5.32364190e-01
-7.05866039e-01 6.13620222e-01 6.37849689e-01 -4.53643024e-01
-1.41871667e+00 2.16437295e-01 2.27164134e-01 -5.65543994e-02
-2.76431710e-01 6.33088171e-01 -7.54582286e-02 2.41905808e-01
7.86603749e-01 1.73544735e-01 -1.03353359e-01 -7.52823830e-01
6.36496842e-01 5.68306148e-01 3.29736292e-01 -1.12546670e+00
5.64316511e-01 3.96191061e-01 -1.51635170e-01 -7.54226089e-01
-1.19250405e+00 -2.95942307e-01 -1.15350056e+00 1.61543980e-01
9.25699174e-01 -8.01403642e-01 -5.92283428e-01 2.67020047e-01
-1.26124156e+00 -9.82585728e-01 3.10570747e-02 5.04663527e-01
-6.95079803e-01 2.79620618e-01 -7.26760745e-01 -4.41042095e-01
1.03232399e-01 -1.01840580e+00 9.03252780e-01 -3.46316665e-01
-4.37190384e-01 -1.54297781e+00 4.15896654e-01 3.22123826e-01
2.52425492e-01 -3.26279610e-01 1.47069347e+00 -7.17976451e-01
-1.20940544e-01 5.90020478e-01 6.41970411e-02 3.82732034e-01
3.99742067e-01 -1.98333889e-01 -1.10563850e+00 -3.75933260e-01
7.83952996e-02 -6.06693149e-01 1.15523636e+00 2.86763221e-01
1.06585503e+00 -1.67748824e-01 -5.62713802e-01 4.58038628e-01
1.11022723e+00 -6.49855807e-02 3.28100681e-01 5.26375771e-01
7.89529741e-01 1.08600473e+00 4.07467157e-01 -1.28894165e-01
6.52017653e-01 7.09457755e-01 -1.01929337e-01 9.80096869e-04
5.94535470e-03 -2.82140851e-01 6.47350848e-01 1.27451587e+00
-7.50222337e-03 9.67182312e-03 -1.11666369e+00 5.84009945e-01
-1.95606601e+00 -7.68985868e-01 4.28550877e-02 2.24895334e+00
1.36734736e+00 1.37561843e-01 1.41907364e-01 -2.37479910e-01
5.23432374e-01 -1.11054346e-01 -4.19962555e-01 -6.60279274e-01
6.43638149e-02 2.02381521e-01 2.31783241e-01 8.19756389e-01
-9.70615506e-01 1.50749826e+00 6.88197184e+00 8.79743218e-01
-9.43142653e-01 4.15354937e-01 3.00087273e-01 1.84602112e-01
-1.23623386e-01 1.41372263e-01 -8.34022105e-01 2.49185815e-01
1.14462078e+00 -4.32938278e-01 5.27275145e-01 6.65282607e-01
-7.60213584e-02 -1.62121162e-01 -1.52514458e+00 3.90756130e-01
1.33097023e-01 -9.22828555e-01 2.73109645e-01 -6.95574954e-02
6.93141937e-01 2.57936329e-01 -2.40783736e-01 6.21409774e-01
1.05808377e+00 -8.05600345e-01 2.93502897e-01 1.06409289e-01
3.33438456e-01 -6.68591797e-01 4.65534300e-01 8.06805372e-01
-1.11233997e+00 9.38546434e-02 -6.05248034e-01 -1.33591473e-01
4.56996873e-04 3.25028390e-01 -1.14892161e+00 6.70740187e-01
7.01979578e-01 6.89002991e-01 -8.90509725e-01 6.82469010e-01
-2.37983420e-01 7.03529000e-01 -9.12281126e-02 2.95309007e-01
1.55100197e-01 -1.30045399e-01 2.52177745e-01 1.53204906e+00
7.97494575e-02 -3.93348396e-01 4.34390932e-01 7.71851897e-01
-8.58967658e-03 2.04288393e-01 -1.04863465e+00 1.77691698e-01
4.18767810e-01 1.06906736e+00 -6.20382428e-01 -5.64666629e-01
-3.95480067e-01 1.03841972e+00 7.18795896e-01 3.08919817e-01
-5.16974688e-01 -8.42962489e-02 3.67047012e-01 -4.11741398e-02
-4.83389571e-03 -3.41963321e-01 -3.41917574e-01 -1.08321857e+00
-4.07957464e-01 -9.06359076e-01 7.41032183e-01 -7.26229131e-01
-1.78023052e+00 4.20303345e-01 6.25166297e-01 -7.82448232e-01
-7.09448457e-01 -6.33523941e-01 -1.79601759e-01 8.93551588e-01
-1.20347285e+00 -1.07920659e+00 9.99652222e-02 6.51015878e-01
7.12774038e-01 -1.13783501e-01 9.04662549e-01 -1.30205313e-02
-3.02966893e-01 5.40206015e-01 4.93386090e-02 4.18461720e-03
1.18662179e+00 -1.30139589e+00 3.96039963e-01 4.32054281e-01
3.82858813e-01 8.89852583e-01 4.18663740e-01 -7.54193187e-01
-8.45272005e-01 -1.06557751e+00 1.07796443e+00 -8.70192945e-01
8.45835388e-01 -8.12126577e-01 -1.33407617e+00 1.12999296e+00
5.47863245e-01 -6.07670009e-01 8.12429309e-01 7.71537662e-01
-6.00126863e-01 -1.05412044e-01 -7.29150832e-01 5.10109067e-01
1.17427683e+00 -5.83318651e-01 -1.15947068e+00 3.57076555e-01
8.81348789e-01 2.49464903e-02 -8.86534691e-01 3.18578005e-01
5.38593829e-01 -8.26299727e-01 7.75388122e-01 -8.44848752e-01
2.76789278e-01 -1.93993196e-01 -8.05896819e-02 -1.84740925e+00
-5.44605494e-01 -1.99795634e-01 4.32185084e-01 1.28183484e+00
9.00633514e-01 -9.03078675e-01 3.72764498e-01 3.27726692e-01
-1.67510957e-01 -1.12253472e-01 -8.49268317e-01 -1.14627194e+00
8.33542824e-01 -4.07865465e-01 2.00000793e-01 1.52563512e+00
1.48714304e-01 9.12152052e-01 -2.89228737e-01 -6.38515502e-02
6.66226029e-01 2.43430465e-01 8.12214494e-01 -1.54100454e+00
-5.74969828e-01 -3.65485698e-01 3.44164312e-01 -7.49674082e-01
8.25836360e-01 -1.41043234e+00 7.57414103e-02 -1.35480368e+00
5.49401641e-01 -6.75870001e-01 -2.76558578e-01 8.82884681e-01
-3.29669952e-01 -3.80340330e-02 3.81168365e-01 2.72205144e-01
-4.92256820e-01 2.44483486e-01 1.11517596e+00 -6.17492795e-02
-3.72263610e-01 -1.41363904e-01 -6.35720015e-01 1.02602804e+00
8.90843272e-01 -9.44001794e-01 -5.49651325e-01 -7.10398674e-01
1.52602389e-01 -4.11775082e-01 -8.56663659e-02 -8.49344909e-01
3.06170791e-01 -1.05172820e-01 2.34696984e-01 -1.46002620e-01
2.62924999e-01 -7.97230601e-01 1.99129730e-01 4.80969250e-01
-5.43892324e-01 1.53779000e-01 4.71838564e-01 2.29214489e-01
-9.23909992e-02 -4.23648328e-01 7.18847096e-01 -4.25363690e-01
-5.87258220e-01 -1.82651207e-01 -6.02787197e-01 3.40208024e-01
6.63021564e-01 4.42839228e-02 -2.66689658e-01 -1.29282763e-02
-8.76749635e-01 2.35885665e-01 6.47754490e-01 5.40704370e-01
-5.02773672e-02 -1.15357399e+00 -8.99968326e-01 -1.99965052e-02
1.82435542e-01 -2.17734054e-01 -3.22017185e-02 6.62029982e-01
1.44423380e-01 4.92057711e-01 -2.48791158e-01 -8.07982266e-01
-1.19411337e+00 8.19360614e-01 2.89002389e-01 -5.80332160e-01
-1.88259602e-01 5.56488752e-01 8.18117857e-01 -8.95973206e-01
7.06571117e-02 -2.78045565e-01 -1.71364889e-01 3.85975718e-01
1.87634960e-01 6.67898357e-02 2.76928134e-02 -3.33538264e-01
-2.90856302e-01 7.88323283e-01 -3.83433044e-01 -4.60050106e-01
1.15610218e+00 -1.52301744e-01 -4.34978276e-01 1.17565262e+00
9.72523689e-01 2.50370093e-02 -9.40750718e-01 -4.48238224e-01
4.77139771e-01 -7.35917762e-02 -2.36205801e-01 -9.05053616e-01
-5.80025434e-01 1.01086533e+00 2.31042117e-01 8.37047324e-02
1.00046206e+00 2.48347715e-01 1.19042270e-01 7.05140591e-01
6.10947132e-01 -1.00798404e+00 2.21248195e-02 9.00092125e-01
6.52766585e-01 -1.08541274e+00 -4.49785292e-02 -3.18181008e-01
-7.53063142e-01 7.82407165e-01 7.06625342e-01 3.23436558e-01
5.55865288e-01 3.63721281e-01 8.42741877e-02 -2.45247465e-02
-1.26974392e+00 -3.44287008e-01 1.44023076e-01 6.98188066e-01
8.09311628e-01 1.71166778e-01 -3.19421232e-01 5.16321778e-01
-4.05844927e-01 1.20000273e-01 2.98520833e-01 7.98396468e-01
-4.42012072e-01 -1.64415610e+00 -3.25717628e-01 2.45560572e-01
-3.69366556e-01 -4.66808230e-01 -7.20366478e-01 1.13041341e+00
3.78770202e-01 8.25847149e-01 1.65889844e-01 -2.39378542e-01
1.78953603e-01 6.92100167e-01 8.44832063e-01 -1.03603530e+00
-7.95639634e-01 1.13585917e-02 2.88009584e-01 -2.53243893e-01
-6.86820149e-01 -6.11870170e-01 -1.15527439e+00 -2.23007664e-01
-9.06942338e-02 4.91976768e-01 5.59478045e-01 1.00068176e+00
2.62542307e-01 4.01156455e-01 3.71855646e-01 -8.49807918e-01
-1.49612814e-01 -1.16438019e+00 -3.32089186e-01 4.61806834e-01
-9.44013968e-02 -7.79723942e-01 -2.42648438e-01 5.02297997e-01] | [10.95283031463623, 9.30573844909668] |
f9857153-d1a0-415c-9f10-67418925f331 | sscbench-a-large-scale-3d-semantic-scene | 2306.09001 | null | https://arxiv.org/abs/2306.09001v1 | https://arxiv.org/pdf/2306.09001v1.pdf | SSCBench: A Large-Scale 3D Semantic Scene Completion Benchmark for Autonomous Driving | Semantic scene completion (SSC) is crucial for holistic 3D scene understanding by jointly estimating semantics and geometry from sparse observations. However, progress in SSC, particularly in autonomous driving scenarios, is hindered by the scarcity of high-quality datasets. To overcome this challenge, we introduce SSCBench, a comprehensive benchmark that integrates scenes from widely-used automotive datasets (e.g., KITTI-360, nuScenes, and Waymo). SSCBench follows an established setup and format in the community, facilitating the easy exploration of the camera- and LiDAR-based SSC across various real-world scenarios. We present quantitative and qualitative evaluations of state-of-the-art algorithms on SSCBench and commit to continuously incorporating novel automotive datasets and SSC algorithms to drive further advancements in this field. Our resources are released on https://github.com/ai4ce/SSCBench. | ['Chen Feng', 'Zhiding Yu', 'Hang Zhao', 'Yue Wang', 'Fisher Yu', 'Tao Jiang', 'Zhiheng Li', 'Zijun Wang', 'Nuo Chen', 'Kenan Li', 'Moonjun Gong', 'Xinhao Liu', 'Sihang Li', 'Yiming Li'] | 2023-06-15 | null | null | null | null | ['3d-semantic-scene-completion', 'scene-understanding'] | ['computer-vision', 'computer-vision'] | [ 4.12664041e-02 -1.80013612e-01 1.64620895e-02 -6.91644192e-01
-9.02173460e-01 -7.51605451e-01 6.85018361e-01 -1.44646361e-01
-1.82164222e-01 3.60298783e-01 3.02053690e-01 -1.97055623e-01
1.95586577e-01 -6.63313091e-01 -7.82293499e-01 -5.87933138e-02
1.68015853e-01 5.90440869e-01 6.47403479e-01 -5.57993591e-01
2.81473249e-01 6.10962152e-01 -2.06007934e+00 2.50864923e-01
5.63324988e-01 9.12352860e-01 5.23629308e-01 4.69152063e-01
-4.56072018e-02 5.60545862e-01 7.76345506e-02 -3.28756899e-01
3.58271450e-01 -2.42371168e-02 -6.33722425e-01 2.00860009e-01
7.22358167e-01 -1.05976440e-01 -5.39164484e-01 8.63030076e-01
1.75765172e-01 2.26647511e-01 2.46148676e-01 -1.57225561e+00
-3.94238420e-02 -5.07537611e-02 -4.34603721e-01 -4.95441607e-04
4.65899915e-01 4.22961533e-01 8.34135771e-01 -1.09300494e+00
1.08057928e+00 1.28962219e+00 8.17507744e-01 1.80194840e-01
-9.59248185e-01 -7.25354791e-01 1.72911659e-01 5.15065312e-01
-1.57542050e+00 -7.80799508e-01 8.42317700e-01 -3.78117800e-01
1.02977896e+00 3.60420853e-01 5.90395570e-01 1.09400356e+00
1.65859833e-02 7.38821208e-01 1.00640321e+00 1.35924644e-03
3.23113710e-01 1.56132989e-02 8.00177082e-02 5.44727743e-01
3.41099590e-01 3.24594021e-01 -9.43651974e-01 2.95350611e-01
4.23923045e-01 -2.20627755e-01 2.12038189e-01 -9.85000789e-01
-1.30576944e+00 5.98524988e-01 4.09370393e-01 -2.70668417e-01
-1.22184187e-01 2.53045708e-01 4.30314869e-01 4.55493778e-02
4.95491236e-01 3.40504676e-01 -4.15570647e-01 -5.01974583e-01
-8.24159980e-01 7.29574919e-01 5.96026421e-01 1.54443073e+00
1.06888723e+00 -1.04641281e-01 5.26428401e-01 5.39415181e-01
3.95763218e-01 7.25453675e-01 -4.30797338e-01 -1.58199239e+00
7.07437038e-01 5.20402193e-01 1.74788475e-01 -8.39397728e-01
-4.18866098e-01 -2.98443645e-01 -2.42495358e-01 3.37112546e-01
2.03695416e-01 2.22886845e-01 -6.86273396e-01 1.34694004e+00
5.49124956e-01 4.26005870e-01 -4.94625941e-02 1.14376426e+00
1.14170873e+00 3.42921346e-01 -8.60527977e-02 5.66179633e-01
1.17984354e+00 -1.01106739e+00 -5.31667411e-01 -6.87833726e-01
6.64911628e-01 -8.66164923e-01 1.12799919e+00 4.93014842e-01
-7.94351101e-01 -6.59465075e-01 -1.18523717e+00 -5.02097487e-01
-6.33562565e-01 -1.41472481e-02 6.09552085e-01 3.47529352e-01
-9.22163844e-01 5.47220185e-02 -8.06287408e-01 -7.56607294e-01
7.11779296e-01 -1.42891973e-01 -6.17245972e-01 -5.93567610e-01
-9.64145839e-01 1.00573659e+00 3.00919801e-01 9.05971229e-02
-1.16011012e+00 -9.47462082e-01 -1.22043073e+00 -4.74874139e-01
8.24311972e-01 -5.60275137e-01 1.43779981e+00 -3.66794288e-01
-1.16185546e+00 1.15344512e+00 -2.30229348e-01 -5.18044949e-01
6.98726833e-01 -5.45298815e-01 -4.99884307e-01 3.39490473e-02
3.12731504e-01 1.14085031e+00 2.69499689e-01 -1.63792479e+00
-7.17928588e-01 -3.09719473e-01 1.02116607e-01 1.41211823e-01
3.43072414e-01 -1.17643729e-01 -7.42169917e-01 -1.11991599e-01
1.91900134e-01 -1.04414785e+00 -4.01790857e-01 1.05165631e-01
-4.21798110e-01 1.56200349e-01 1.15216160e+00 -3.62404495e-01
7.05749512e-01 -2.36753178e+00 -1.05432406e-01 -8.86332616e-02
2.35790402e-01 -1.01609379e-01 -2.54604787e-01 7.79636681e-01
1.96808577e-01 -2.74727702e-01 -3.28316569e-01 -8.22056413e-01
1.07036881e-01 2.52541631e-01 -3.67450476e-01 6.49678946e-01
2.16925755e-01 1.00936723e+00 -9.17550027e-01 -4.63742495e-01
9.28021610e-01 3.29256982e-01 -5.25652587e-01 -5.06569743e-02
-4.49221551e-01 5.30447662e-01 -4.53521192e-01 8.07321191e-01
9.50680315e-01 1.09833181e-01 -5.63313700e-02 -1.57680884e-01
-4.57792401e-01 2.31332287e-01 -1.29874277e+00 2.32117462e+00
-3.50978076e-01 8.94450843e-01 1.04430027e-01 -5.82708955e-01
8.76147091e-01 -2.01781064e-01 4.21600670e-01 -8.64538133e-01
1.30097762e-01 2.16269031e-01 -5.60461164e-01 -4.24304396e-01
9.82280791e-01 1.87474281e-01 -2.60833651e-01 -6.27910644e-02
-1.10554136e-01 -7.93340743e-01 1.82680324e-01 4.24516022e-01
1.01712489e+00 6.08148754e-01 1.51554853e-01 -3.83461297e-01
3.50339741e-01 7.64895439e-01 4.48009968e-01 3.89889985e-01
-3.55139613e-01 9.44889367e-01 2.22698167e-01 -3.25765014e-01
-1.06921422e+00 -1.30332553e+00 -1.44766942e-01 6.72668576e-01
6.70277238e-01 -7.52271295e-01 -5.58040261e-01 -4.61702943e-01
3.57344657e-01 1.00658584e+00 -4.46940303e-01 5.79057038e-02
-3.81275445e-01 2.78770514e-02 2.93343395e-01 6.37065530e-01
6.09999061e-01 -8.50764036e-01 -8.81969333e-01 -3.78378220e-02
-1.94002405e-01 -1.81694019e+00 -4.38583456e-02 -1.67062610e-01
-6.34021342e-01 -1.23563385e+00 1.30776241e-01 -2.89730102e-01
2.60470212e-01 9.21712101e-01 1.40518403e+00 -2.36436591e-01
-3.87652874e-01 5.10092139e-01 -3.14067572e-01 -5.56866169e-01
-2.53017426e-01 1.24512218e-01 -9.82664898e-02 -2.99350262e-01
4.37163353e-01 -3.54851902e-01 -5.92850387e-01 6.03509665e-01
-7.31687307e-01 4.35938179e-01 3.68599474e-01 -4.49630716e-05
8.08531225e-01 -2.31868863e-01 3.45267713e-01 -8.00868571e-01
-6.84276894e-02 -6.99811399e-01 -8.44277322e-01 -3.99034768e-01
-2.82148987e-01 -3.98417532e-01 7.24811703e-02 3.27646405e-01
-1.24094141e+00 2.55776912e-01 -2.10454345e-01 -5.48061192e-01
-6.83803558e-01 3.50903213e-01 -3.25343996e-01 -1.20242778e-02
6.05068326e-01 -3.41090448e-02 6.00542948e-02 -5.61742246e-01
7.97948956e-01 4.71419036e-01 8.35380375e-01 -3.75894487e-01
8.21408570e-01 1.01707220e+00 2.04174131e-01 -9.32673216e-01
-8.73009324e-01 -8.79651189e-01 -7.18525112e-01 -5.54653049e-01
7.71307290e-01 -1.24059963e+00 -2.36875191e-01 3.63155752e-01
-9.40947294e-01 -5.31127751e-01 -3.16217721e-01 4.70435977e-01
-7.80497551e-01 2.10288003e-01 2.73972340e-02 -5.81884444e-01
3.83809835e-01 -1.24685383e+00 1.62459958e+00 -1.50868922e-01
-4.53806311e-01 -7.49934316e-01 1.11045882e-01 7.90082335e-01
2.36775860e-01 5.50457656e-01 2.01499194e-01 -8.64796937e-02
-9.74149168e-01 -3.28276783e-01 -2.90269822e-01 1.52266115e-01
-8.94231349e-02 1.32991709e-02 -1.16299820e+00 -9.23542604e-02
-3.96933049e-01 -4.83916283e-01 8.54438305e-01 1.55201256e-01
1.09320450e+00 5.45186698e-01 -4.00827199e-01 6.71837568e-01
1.45823932e+00 -2.37611338e-01 6.98415399e-01 5.48967540e-01
7.79454947e-01 8.53522718e-01 1.21744764e+00 3.53336513e-01
9.87508118e-01 7.92766809e-01 8.87684703e-01 2.46802475e-02
-6.14321411e-01 -4.03657973e-01 1.22023433e-01 6.87338352e-01
1.18591286e-01 -1.75195560e-01 -1.28838193e+00 7.00224519e-01
-1.83123314e+00 -7.20730782e-01 -7.31663465e-01 1.98117805e+00
2.73662060e-01 3.98467094e-01 2.53125802e-02 -4.18785214e-02
1.91581815e-01 3.45284373e-01 -5.65666020e-01 -1.43684343e-01
-2.68623739e-01 -1.01261579e-01 6.48119152e-01 4.96514022e-01
-1.04657972e+00 1.24326611e+00 5.94951916e+00 8.39657128e-01
-8.65908861e-01 2.61216938e-01 3.98826063e-01 -2.59403586e-01
-4.11437064e-01 1.79982483e-01 -8.93836498e-01 1.04931638e-01
8.71397614e-01 -4.03918093e-03 2.78245777e-01 1.05754435e+00
4.06889081e-01 -4.95642811e-01 -1.07651055e+00 9.95881319e-01
6.57348335e-02 -1.66521585e+00 -4.63821918e-01 2.35414375e-02
8.43225360e-01 7.80564010e-01 3.92486453e-02 2.25786254e-01
3.45253289e-01 -8.11097085e-01 1.18849242e+00 4.58337337e-01
1.03246522e+00 -6.04167879e-01 5.26951373e-01 2.22498074e-01
-1.42666304e+00 1.77357569e-01 -1.98745832e-01 -4.59246710e-02
3.17013055e-01 4.84107852e-01 -6.57037675e-01 8.39350820e-01
9.69058156e-01 1.16045606e+00 -7.76212573e-01 9.02873814e-01
-5.82711361e-02 4.26445037e-01 -4.60914582e-01 4.19138134e-01
3.12760979e-01 -1.94656998e-01 6.58790529e-01 1.19168746e+00
2.57105201e-01 -1.67355224e-01 2.41549879e-01 8.48474562e-01
1.24851435e-01 -2.79896349e-01 -1.03001249e+00 2.50104845e-01
6.30315721e-01 1.38579357e+00 -6.97210431e-01 -1.71325967e-01
-5.55859447e-01 4.79249805e-01 1.12254597e-01 2.57900864e-01
-1.04914951e+00 -2.75580510e-02 1.20169604e+00 5.01232147e-01
2.48801887e-01 -8.30636919e-01 -7.19735742e-01 -7.67534077e-01
4.54426482e-02 -7.12745309e-01 3.90731581e-02 -1.11055493e+00
-9.42727506e-01 3.72753590e-01 4.97194558e-01 -1.48214269e+00
4.80311029e-02 -4.70142931e-01 -3.02399695e-01 4.90573168e-01
-1.80810583e+00 -1.37595761e+00 -9.24811959e-01 4.99945939e-01
9.61570203e-01 5.03621139e-02 4.21483785e-01 2.30289653e-01
-3.78268272e-01 -7.62255117e-02 -1.13469481e-01 -5.04973352e-01
7.03477621e-01 -9.51107323e-01 1.04930043e+00 7.89705813e-01
1.07169457e-01 1.87772661e-01 8.32190037e-01 -7.30019093e-01
-1.90888858e+00 -1.43495715e+00 5.11623681e-01 -1.01214278e+00
7.34350979e-01 -7.54405141e-01 -5.80524147e-01 7.21434534e-01
1.43919379e-01 1.83449730e-01 2.31098682e-01 5.08010015e-02
-3.70545506e-01 -2.99916863e-01 -9.89961207e-01 7.13457882e-01
1.69025862e+00 -4.90549475e-01 -2.29594544e-01 1.93934828e-01
7.41740584e-01 -7.23984957e-01 -6.69043064e-01 5.37173390e-01
5.16281486e-01 -1.26583040e+00 1.12640440e+00 -2.08595112e-01
4.93369550e-01 -5.71813345e-01 -6.44313216e-01 -1.13899541e+00
-1.12429716e-01 -3.88886124e-01 -5.84404841e-02 9.91921008e-01
1.86648890e-01 -6.59554660e-01 8.62253308e-01 1.97619990e-01
-7.27825463e-01 -5.58403552e-01 -9.29295003e-01 -8.08243155e-01
-1.90113544e-01 -1.21215022e+00 7.28813946e-01 5.12449145e-01
-4.92199928e-01 5.16084135e-02 -7.97980577e-02 1.79639429e-01
8.09407055e-01 1.39165834e-01 1.41157925e+00 -1.07826006e+00
2.67335951e-01 -2.83840239e-01 -7.10570097e-01 -9.61793661e-01
6.11370988e-02 -7.23202705e-01 2.45058864e-01 -1.56975472e+00
-1.15336217e-01 -5.25745690e-01 3.61710675e-02 1.57162637e-01
1.21869966e-01 6.21032834e-01 1.42819703e-01 2.72408780e-02
-1.02899849e+00 6.69410110e-01 1.13549519e+00 1.01932652e-01
1.70974255e-01 -3.48586828e-01 -5.61683357e-01 8.23767662e-01
1.12077427e+00 -2.74861395e-01 -6.54188752e-01 -5.42563856e-01
2.84427285e-01 -2.52062261e-01 6.68752730e-01 -1.32864952e+00
1.53389171e-01 -2.21332282e-01 9.43058655e-02 -1.20872617e+00
7.63347447e-01 -7.71787107e-01 4.23137575e-01 4.28709276e-02
8.96652862e-02 1.07501984e-01 5.78707516e-01 5.47497034e-01
-3.13385427e-01 3.02216709e-01 4.58522499e-01 -1.08664215e-01
-1.55355620e+00 3.21920037e-01 -1.64492756e-01 2.44642705e-01
1.19120181e+00 -5.69403291e-01 -4.20092821e-01 -2.52273589e-01
-3.01799059e-01 5.88914096e-01 8.10271800e-01 9.32683706e-01
7.97229707e-01 -1.32448602e+00 -7.88891912e-01 3.27124059e-01
8.15879941e-01 3.68971705e-01 4.70953107e-01 9.39074099e-01
-8.70465279e-01 5.19321978e-01 -1.42870858e-01 -9.12623525e-01
-1.26822996e+00 3.44658434e-01 4.94373888e-02 2.61253268e-01
-7.49302268e-01 6.40563369e-01 3.08777243e-01 -7.13444114e-01
-6.99887276e-02 -4.03463632e-01 2.02579692e-01 -3.55030209e-01
1.67616844e-01 6.65347874e-01 3.00354004e-01 -9.48153138e-01
-6.30150378e-01 5.34262061e-01 3.69295090e-01 -1.55977532e-01
1.27467799e+00 -4.12016064e-01 1.54637754e-01 5.38015723e-01
1.14276946e+00 -5.75116836e-03 -1.65859389e+00 -9.97777879e-02
2.31142882e-02 -7.03769684e-01 2.52546340e-01 -5.72097600e-01
-8.82650256e-01 7.81469941e-01 4.39929992e-01 -1.02688283e-01
8.00665855e-01 2.59169459e-01 8.18794429e-01 3.13642859e-01
8.18120182e-01 -1.10707545e+00 -1.23518698e-01 6.07830048e-01
1.11815155e+00 -1.50463772e+00 1.42394707e-01 -9.25609291e-01
-8.05062234e-01 7.55940855e-01 7.10182071e-01 -1.25830784e-01
7.00525761e-01 3.73652637e-01 1.69296667e-01 -6.54221356e-01
-7.29207695e-01 -3.78855973e-01 3.19328606e-01 6.23051107e-01
-9.55315889e-04 9.27230567e-02 3.64443660e-01 1.85842589e-01
-4.81911749e-01 9.35544521e-02 3.30982178e-01 1.01907361e+00
-3.38141590e-01 -8.13383341e-01 -3.18405986e-01 1.90529287e-01
2.10992903e-01 1.97382629e-01 -3.68122697e-01 1.04619622e+00
2.97951162e-01 1.16647160e+00 7.39246011e-02 -4.94686097e-01
7.18988359e-01 -3.49021941e-01 5.52144766e-01 -6.23897731e-01
6.79018535e-03 -1.65702149e-01 6.38642013e-01 -1.30074358e+00
-3.49686503e-01 -1.04265261e+00 -1.14382434e+00 -4.82146919e-01
7.68697038e-02 -3.34787309e-01 1.02388203e+00 6.95998728e-01
6.29651129e-01 5.30202925e-01 3.76971185e-01 -1.14473188e+00
-4.14974429e-02 -6.52157545e-01 -3.46497983e-01 3.26720387e-01
8.82779509e-02 -1.05327964e+00 -1.10994242e-01 -2.96176318e-02] | [7.984550952911377, -2.129031181335449] |
720c4dad-e7a6-446d-991f-14a74e324b8e | deep-learning-approaches-to-osteosarcoma | null | null | https://doi.org/10.3390/cancers15082290 | https://doi.org/10.3390/cancers15082290 | Deep Learning Approaches to Osteosarcoma Diagnosis and Classification: A Comparative Methodological Approach | Background: Osteosarcoma is the most common primary malignancy of the bone, being most prevalent in childhood and adolescence. Despite recent progress in diagnostic methods, histopathology remains the gold standard for disease staging and therapy decisions. Machine learning and deep learning methods have shown potential for evaluating and classifying histopathological cross-sections. Methods: This study used publicly available images of osteosarcoma cross-sections to analyze and compare the performance of state-of-the-art deep neural networks for histopathological evaluation of osteosarcomas. Results: The classification performance did not necessarily improve when using larger networks on our dataset. In fact, the smallest network combined with the smallest image input size achieved the best overall performance. When trained using 5-fold cross-validation, the MobileNetV2 network achieved 91% overall accuracy. Conclusions: The present study highlights the importance of careful selection of network and input image size. Our results indicate that a larger number of parameters is not always better, and the best results can be achieved on smaller and more efficient networks. The identification of an optimal network and training configuration could greatly improve the accuracy of osteosarcoma diagnoses and ultimately lead to better disease outcomes for patients. | ['George K. Matsopoulos', 'George I. Lambrou', 'Ioannis A. Vezakis'] | 2023-04-13 | null | null | null | cancers-2023-4 | ['tumour-classification'] | ['medical'] | [ 2.25293059e-02 -2.17179313e-01 -3.53599668e-01 5.57999089e-02
-8.96525264e-01 5.41524887e-02 -3.48462234e-03 4.85676169e-01
-1.00863218e+00 6.99438095e-01 -3.50893326e-02 -4.56379414e-01
-1.94315597e-01 -8.43707561e-01 -2.36695912e-02 -9.83859897e-01
-2.00965386e-02 7.70863831e-01 3.59009773e-01 -7.13588223e-02
-4.52217087e-02 8.66837621e-01 -1.21519721e+00 1.84218362e-01
5.60764432e-01 8.30753803e-01 5.23263872e-01 6.58639848e-01
4.84134145e-02 8.88654888e-01 -6.39668882e-01 -1.79108679e-01
-1.25784487e-01 -1.28679991e-01 -1.02720010e+00 -5.07008173e-02
-5.19648492e-02 -3.86834711e-01 -4.18774068e-01 4.35193270e-01
9.24315214e-01 -1.30081609e-01 8.43079567e-01 -7.59996116e-01
-2.61419296e-01 5.48163533e-01 -3.71028394e-01 7.42129803e-01
-3.22976738e-01 4.48676080e-01 7.56554067e-01 -6.20657265e-01
6.21636450e-01 3.84778321e-01 9.77755547e-01 5.05082190e-01
-1.11986065e+00 -6.10819936e-01 -6.65233135e-01 3.21883947e-01
-1.71735632e+00 -3.55109006e-01 2.11487755e-01 -4.64146525e-01
9.03158009e-01 1.23080693e-01 9.91264343e-01 8.14540565e-01
6.93480253e-01 4.15870875e-01 8.20615530e-01 -3.74792784e-01
1.41764894e-01 -6.50417581e-02 -1.94918022e-01 8.21688950e-01
4.53057468e-01 -1.51383072e-01 -2.03031197e-01 -1.35117862e-02
9.46529627e-01 -8.52351077e-03 -1.71676174e-01 1.15168750e-01
-1.24993420e+00 7.91565597e-01 4.83587414e-01 8.32353055e-01
-4.70218301e-01 5.78985929e-01 7.13303685e-01 6.31555542e-02
2.45654702e-01 3.63540173e-01 -1.43321663e-01 -1.70239389e-01
-8.77255797e-01 -2.19442099e-01 3.55323792e-01 -2.79666129e-02
2.37342827e-02 -7.83086345e-02 2.31406316e-01 1.22048628e+00
2.71859020e-01 1.99527845e-01 9.89571214e-01 -7.70489514e-01
-2.04650797e-02 4.88548279e-01 -4.96871889e-01 -6.88196063e-01
-9.96934712e-01 -9.14464891e-01 -9.07493591e-01 3.17924172e-01
7.53819823e-01 4.73011322e-02 -9.69704449e-01 1.26637590e+00
-8.25764239e-03 -3.53306472e-01 -8.79065320e-02 7.71796703e-01
8.16474378e-01 9.81034636e-02 3.63356739e-01 2.90245444e-01
1.60242438e+00 -4.94326860e-01 -2.81902075e-01 -1.95895031e-01
1.19238019e+00 -6.57094955e-01 7.28125215e-01 9.50474590e-02
-8.25713873e-01 -5.40331826e-02 -1.07994938e+00 1.18319176e-01
-2.77364463e-01 4.07919705e-01 7.23948002e-01 7.48919010e-01
-1.25266528e+00 5.43002784e-01 -1.35733807e+00 -8.37033093e-01
6.34193301e-01 9.54000771e-01 -7.72369921e-01 7.68529475e-02
-9.23544288e-01 1.09196341e+00 5.23616612e-01 1.11943960e-01
-7.19941914e-01 -5.31497121e-01 -4.59755182e-01 -1.83123257e-02
-4.66490090e-02 -1.00744319e+00 1.32177389e+00 -8.04511309e-01
-1.16965771e+00 1.12106276e+00 1.39550403e-01 -5.29002309e-01
3.31146836e-01 5.21000564e-01 -1.25224546e-01 4.43310261e-01
5.70397899e-02 6.82194591e-01 3.56062017e-02 -7.52469718e-01
-8.13972116e-01 -4.13015217e-01 -4.14809048e-01 9.48800966e-02
-3.44917029e-01 -1.19416296e-01 -4.20522928e-01 -5.58127522e-01
9.65317115e-02 -9.91974831e-01 -4.14829433e-01 2.01968968e-01
1.17076710e-02 -2.11956903e-01 4.65029985e-01 -7.43125558e-01
9.95779753e-01 -1.94950151e+00 -7.27100819e-02 3.03157032e-01
4.98085946e-01 -2.96124537e-02 1.22915603e-01 2.52912521e-01
-6.24426268e-02 3.19801241e-01 3.09612036e-01 -1.04698598e-01
-4.17867571e-01 1.50502115e-01 9.22421813e-01 8.74623418e-01
7.47461021e-02 8.86942565e-01 -6.75509453e-01 -9.45587635e-01
2.61456996e-01 5.68278551e-01 -3.27586532e-01 -3.44723433e-01
3.44477922e-01 2.63070583e-01 -3.16279203e-01 6.53608620e-01
-1.30794555e-01 -8.08150589e-01 1.48845807e-01 -7.44568780e-02
4.38243836e-01 -3.42841297e-02 -6.73628688e-01 1.46945512e+00
-5.77137768e-01 1.09948170e+00 -1.07331291e-01 -8.06831419e-01
6.70861304e-01 5.55298090e-01 9.34241712e-01 -7.90356576e-01
6.28587663e-01 5.46377659e-01 5.75624943e-01 -8.09321880e-01
-8.23895186e-02 -4.24037814e-01 3.30671102e-01 3.54307979e-01
-1.72727436e-01 1.77026875e-02 3.29979807e-01 -4.67296727e-02
1.51343715e+00 -6.87342227e-01 3.82513821e-01 -2.00058624e-01
3.62296194e-01 1.21433906e-01 2.51316637e-01 4.78486866e-01
-3.21941406e-01 6.61819041e-01 3.71799231e-01 -4.91632491e-01
-9.34215009e-01 -9.52466726e-01 -3.93166423e-01 7.97036350e-01
-3.30493629e-01 1.49036601e-01 -4.52206165e-01 -5.24247885e-01
5.57813682e-02 2.36931458e-01 -8.50103736e-01 -1.77258775e-02
-4.84874427e-01 -9.18918669e-01 8.71746540e-01 5.76765537e-01
2.78947502e-01 -9.04289007e-01 -7.30933011e-01 4.21460927e-01
-8.74597430e-02 -8.61390889e-01 1.92296710e-02 5.39888859e-01
-1.09430444e+00 -1.20048869e+00 -1.18347740e+00 -9.96742845e-01
8.89778852e-01 -5.97818382e-02 8.62710238e-01 5.28762877e-01
-4.74622518e-01 1.16857924e-01 -7.17471167e-02 -3.78643721e-01
-7.13066220e-01 5.65423846e-01 -2.77114213e-01 -4.70560431e-01
3.80650401e-01 -4.75505710e-01 -7.62801468e-01 1.15730137e-01
-9.30581570e-01 1.71162084e-01 1.16337550e+00 1.03160143e+00
6.08630776e-01 3.49504441e-01 4.48865265e-01 -6.46147370e-01
4.77912277e-01 -4.99746561e-01 1.75641075e-01 -5.49403615e-02
-6.41231775e-01 -9.55192558e-03 5.53346992e-01 -1.57921672e-01
-4.77264941e-01 -2.28361860e-01 -5.45186102e-01 4.07996140e-02
-2.38131255e-01 7.77112603e-01 5.46044648e-01 -3.81566644e-01
8.24981928e-01 -1.40406877e-01 5.16664267e-01 -4.39104363e-02
-6.00329340e-01 6.80069566e-01 2.66346753e-01 -2.39105716e-01
6.37437403e-02 8.40941310e-01 3.04216862e-01 -9.32771087e-01
-2.78411269e-01 -5.87975919e-01 -4.81467485e-01 -5.37831664e-01
9.37021077e-01 -6.42935932e-01 -6.02262020e-01 6.30310178e-01
-5.70759952e-01 -5.22929192e-01 -2.24959970e-01 6.29039884e-01
-3.27535748e-01 1.55211806e-01 -9.31922078e-01 -7.55702704e-02
-6.75750494e-01 -1.58979964e+00 1.06748474e+00 2.12826908e-01
-5.83248198e-01 -1.12461936e+00 2.19898611e-01 4.42773581e-01
6.29961073e-01 2.54015177e-01 1.13407779e+00 -6.58686936e-01
-9.57178175e-02 -4.62056041e-01 -5.05683869e-02 7.69486725e-02
3.97109360e-01 2.91623265e-01 -6.86594009e-01 -1.94234237e-01
-5.80288053e-01 -1.12747923e-01 6.23023748e-01 6.16732419e-01
1.14182925e+00 3.44948679e-01 -7.32723475e-01 4.27423507e-01
1.64662099e+00 4.55793530e-01 5.97464561e-01 1.00175786e+00
4.48236585e-01 3.16035807e-01 -5.49960546e-02 8.96278247e-02
3.34431738e-01 4.24874693e-01 4.83133405e-01 -3.73100340e-01
-4.46834475e-01 2.86490411e-01 -2.96671063e-01 6.59738719e-01
-2.67189711e-01 -2.76850313e-01 -1.66673470e+00 7.70816505e-01
-1.24458289e+00 -8.69109154e-01 -5.15686199e-02 1.89960623e+00
5.58839321e-01 3.23663950e-01 1.05262876e-01 5.30557275e-01
5.85050523e-01 -2.65965015e-01 -2.60640472e-01 -2.26690918e-02
-7.82428756e-02 3.51504177e-01 7.09037364e-01 -1.72251929e-02
-7.20721066e-01 2.16953844e-01 7.02115631e+00 8.37795794e-01
-1.65640366e+00 2.81726271e-01 8.57866347e-01 -4.63099450e-01
1.21008873e-01 -5.05206108e-01 -2.81386554e-01 4.64908957e-01
1.07450271e+00 -3.40047404e-02 -2.45456547e-01 4.65487152e-01
5.06332755e-01 -4.91526484e-01 -9.24209774e-01 8.50748599e-01
-1.74263492e-01 -1.65490532e+00 -3.31215620e-01 5.36569417e-01
3.28322709e-01 4.00118560e-01 -2.33678259e-02 -6.16980344e-02
7.86873698e-02 -1.23853171e+00 4.31173176e-01 2.83391833e-01
8.43036830e-01 -7.12805569e-01 1.50566816e+00 1.23788938e-02
-8.59426796e-01 -4.70319577e-02 -6.44242857e-03 2.52023548e-01
-7.65956566e-02 3.80667061e-01 -1.25658119e+00 1.58985034e-01
7.25626349e-01 5.03642023e-01 -7.81815350e-01 1.33880532e+00
2.77842671e-01 6.79125071e-01 -5.54668188e-01 -5.22916496e-01
1.48125216e-01 4.12151307e-01 2.72185076e-02 1.12858438e+00
3.05238336e-01 -2.77800448e-02 -1.76080018e-01 -3.42468917e-02
-2.02962011e-02 3.56763959e-01 -2.23838627e-01 -2.18384862e-01
4.45741743e-01 9.93226111e-01 -1.36977065e+00 2.48660818e-02
-4.36911255e-01 3.51812810e-01 4.90927577e-01 6.44910634e-02
-5.98780453e-01 -1.79780975e-01 3.48507762e-01 5.37242651e-01
-9.44256783e-03 -1.95130501e-02 -5.81021130e-01 -5.16899526e-01
-3.92868698e-01 -7.58914828e-01 7.60536373e-01 -6.38926744e-01
-9.59515274e-01 3.46818089e-01 -3.17806244e-01 -1.04362631e+00
-2.61699315e-03 -6.79577172e-01 -5.65653682e-01 4.17785704e-01
-9.61133718e-01 -9.34439600e-01 -3.60188484e-01 9.74243581e-02
4.97668415e-01 -4.40015532e-02 9.53381717e-01 3.21035355e-01
-5.91458261e-01 6.25491202e-01 3.53101820e-01 3.41759503e-01
5.26352286e-01 -1.10250306e+00 -4.14825559e-01 2.98087537e-01
-3.43168288e-01 2.61057317e-01 6.10811055e-01 -4.11434203e-01
-1.00825179e+00 -7.78777540e-01 4.74017292e-01 5.73756592e-03
8.35302055e-01 4.49328512e-01 -5.37500501e-01 5.59868634e-01
1.68523997e-01 -2.59518594e-01 1.23299718e+00 -1.50479957e-01
2.91734606e-01 -1.85671866e-01 -1.20627654e+00 6.33716106e-01
4.16939259e-01 -3.59481186e-01 -7.56033957e-02 2.99594820e-01
-6.19273521e-02 -4.86748666e-01 -1.40266728e+00 5.13678133e-01
8.95512581e-01 -7.47192860e-01 7.90171623e-01 -1.88586250e-01
5.26220620e-01 -1.85451105e-01 1.15095571e-01 -1.15904069e+00
-5.73391974e-01 3.20492834e-01 4.15000647e-01 6.41811907e-01
8.27454567e-01 -7.31243849e-01 1.44543433e+00 3.53970349e-01
-1.09658152e-01 -1.62513721e+00 -1.08676493e+00 -5.05205691e-01
3.92668635e-01 -3.79699230e-01 3.64993691e-01 5.17614901e-01
-1.51114285e-01 -1.63078848e-02 4.88166332e-01 1.78597923e-02
4.38914061e-01 -5.99556506e-01 3.54636550e-01 -1.09260523e+00
-2.12884948e-01 -1.02456808e+00 -1.16624832e+00 -3.02330889e-02
-1.14457034e-01 -9.86043990e-01 -2.51137286e-01 -2.01886868e+00
4.62095708e-01 -6.16265357e-01 -3.67166311e-01 5.03765047e-01
6.06707111e-02 6.37228370e-01 -2.01853245e-01 4.48736906e-01
-5.80571368e-02 1.00754663e-01 1.22371006e+00 -4.46787894e-01
6.60274401e-02 -5.30186575e-03 -5.71005106e-01 5.05063355e-01
9.54450846e-01 -5.70635915e-01 -2.24326804e-01 -4.97777253e-01
5.63433468e-02 -9.39321890e-03 2.48468250e-01 -1.35141361e+00
4.67504412e-01 -8.04724842e-02 7.22343087e-01 -5.34550548e-01
3.19912940e-01 -8.80039573e-01 7.26730287e-01 1.07224381e+00
-1.10029057e-01 9.34596360e-02 2.31596619e-01 2.09388286e-01
-1.13488443e-01 -3.09374332e-01 9.70449805e-01 -2.37011090e-01
-6.65369868e-01 2.21873373e-01 -9.47394013e-01 -4.50605631e-01
1.35766172e+00 -6.68065071e-01 -3.63633782e-01 -1.73081875e-01
-8.93956482e-01 3.72118279e-02 6.34941339e-01 -1.43614002e-02
4.29156572e-01 -1.18770695e+00 -6.76760852e-01 -3.62886101e-01
1.67838588e-01 6.96570054e-02 3.39715332e-01 1.31968224e+00
-1.44371474e+00 3.74241859e-01 -2.78294891e-01 -8.46832514e-01
-1.47219348e+00 -1.52919799e-01 9.45841074e-01 -6.38769388e-01
-5.90164602e-01 1.02661443e+00 -2.46161386e-01 -2.74289679e-02
2.40398303e-01 -5.92998266e-02 -2.89091021e-01 -7.78618902e-02
2.39117280e-01 5.74431717e-01 4.65936810e-01 -6.12868309e-01
-3.80869746e-01 2.68165678e-01 -4.21183586e-01 -1.52547155e-02
1.32360208e+00 1.92087382e-01 -9.27064344e-02 3.49661529e-01
1.30968142e+00 -2.84789681e-01 -4.81797814e-01 1.67502299e-01
-1.69029281e-01 -1.38718218e-01 4.68073845e-01 -5.89817107e-01
-1.57224834e+00 5.70809960e-01 9.42656457e-01 9.45910364e-02
1.09336281e+00 2.32653379e-01 7.43942797e-01 7.96704367e-02
4.78604913e-01 -9.41163540e-01 -1.20664639e-02 2.55785231e-02
3.42645466e-01 -1.10594726e+00 1.01264484e-01 -1.35819182e-01
-1.20879896e-01 1.37928176e+00 6.30731344e-01 7.44007304e-02
7.22024322e-01 4.44069296e-01 2.84628928e-01 -5.63311219e-01
-6.35888278e-01 -1.65787354e-01 -2.36449629e-01 3.59873354e-01
6.92347407e-01 1.53972149e-01 -4.11057979e-01 2.36726880e-01
-3.57580304e-01 2.21778542e-01 4.61310744e-01 1.10134161e+00
-4.99874145e-01 -9.62982595e-01 -4.35255885e-01 1.13004601e+00
-9.02297020e-01 2.70451277e-01 -2.21231610e-01 1.19446039e+00
6.85496628e-02 7.18846798e-01 9.58228782e-02 -2.69696653e-01
2.17342630e-01 -1.50356695e-01 4.66974914e-01 -4.50014085e-01
-7.58635700e-01 -2.11357474e-01 2.55546510e-01 -1.09748557e-01
-3.75800371e-01 -6.65398002e-01 -1.54372084e+00 -6.06581688e-01
-5.90115070e-01 -2.38419808e-02 8.45035076e-01 1.09485292e+00
1.24449216e-01 9.55680013e-01 2.37567440e-01 -5.29361546e-01
-2.20002890e-01 -9.87093508e-01 -7.33265460e-01 -4.26892079e-02
8.02935734e-02 -6.24106944e-01 -4.13978875e-01 3.02994102e-02] | [15.098892211914062, -2.880466938018799] |
cb1aa9c9-a695-4912-bcb9-8f655c3707ed | a-serial-dual-channel-library-occupancy | 2306.16080 | null | https://arxiv.org/abs/2306.16080v1 | https://arxiv.org/pdf/2306.16080v1.pdf | A serial dual-channel library occupancy detection system based on Faster RCNN | The phenomenon of seat occupancy in university libraries is a prevalent issue. However, existing solutions, such as software-based seat reservations and sensors-based occupancy detection, have proven to be inadequate in effectively addressing this problem. In this study, we propose a novel approach: a serial dual-channel object detection model based on Faster RCNN. Furthermore, we develop a user-friendly Web interface and mobile APP to create a computer vision-based platform for library seat occupancy detection. To construct our dataset, we combine real-world data collec-tion with UE5 virtual reality. The results of our tests also demonstrate that the utilization of per-sonalized virtual dataset significantly enhances the performance of the convolutional neural net-work (CNN) in dedicated scenarios. The serial dual-channel detection model comprises three es-sential steps. Firstly, we employ Faster RCNN algorithm to determine whether a seat is occupied by an individual. Subsequently, we utilize an object classification algorithm based on transfer learning, to classify and identify images of unoccupied seats. This eliminates the need for manual judgment regarding whether a person is suspected of occupying a seat. Lastly, the Web interface and APP provide seat information to librarians and students respectively, enabling comprehensive services. By leveraging deep learning methodologies, this research effectively addresses the issue of seat occupancy in library systems. It significantly enhances the accuracy of seat occupancy recognition, reduces the computational resources required for training CNNs, and greatly improves the effi-ciency of library seat management. | ['Xin Chen', 'Min Yang', 'Zitong Wang', 'XiaoWen Chang', 'Guoqiang Yang'] | 2023-06-28 | null | null | null | null | ['transfer-learning'] | ['miscellaneous'] | [-1.69806808e-01 -4.39334571e-01 -4.41596322e-02 -4.67668116e-01
-6.72687411e-01 -5.24353385e-01 2.21336395e-01 1.51885469e-02
-7.16842353e-01 7.22405910e-01 -9.89321433e-03 -8.50934148e-01
-4.93518747e-02 -1.23562157e+00 -6.21358812e-01 -5.56406319e-01
4.30239767e-01 1.79518253e-01 2.96635740e-02 -1.21356465e-01
3.43249500e-01 6.97090149e-01 -1.78743756e+00 2.14735586e-02
7.49824941e-01 1.22115433e+00 4.42966938e-01 5.75839758e-01
-9.19504240e-02 4.58484888e-01 -7.36182630e-01 -1.02088628e-02
1.55849323e-01 1.95760623e-01 -3.39514017e-01 6.69413358e-02
5.11135995e-01 -6.54352486e-01 -6.83790147e-01 7.55975306e-01
8.74498725e-01 2.99855381e-01 3.23510706e-01 -8.85243475e-01
-2.97140509e-01 1.39581263e-01 -2.77355969e-01 6.53350115e-01
4.40235466e-01 2.23759189e-01 4.53143299e-01 -6.82347655e-01
-2.60524720e-01 8.61457407e-01 7.16166615e-01 2.36878514e-01
-7.01783419e-01 -9.69041646e-01 3.47425416e-02 3.63774598e-01
-1.70125377e+00 -6.31090164e-01 6.37516737e-01 -1.67796582e-01
6.56530559e-01 8.89297962e-01 9.87589061e-01 8.73288989e-01
1.40278831e-01 9.44904447e-01 1.01921761e+00 -5.97696304e-01
3.45437825e-01 5.04329205e-01 3.90052259e-01 8.83888900e-01
1.63587078e-01 -1.58106014e-01 -1.39437258e-01 4.04049531e-02
1.01909053e+00 4.62862819e-01 -2.21897662e-01 7.86836743e-02
-5.61908603e-01 5.83375573e-01 3.20379227e-01 2.84012854e-01
-4.30711448e-01 -1.81362465e-01 2.44162783e-01 -2.65922517e-01
1.15899906e-01 8.63788277e-02 -8.01704004e-02 -2.17314854e-01
-8.00953388e-01 2.44127717e-02 7.88846672e-01 1.05370629e+00
6.15838885e-01 1.13595054e-02 -1.67980224e-01 6.67101443e-01
3.83058071e-01 4.24747765e-01 6.08756423e-01 -5.39499104e-01
2.96562165e-01 7.58088589e-01 1.77177906e-01 -9.11404967e-01
-5.60900450e-01 -6.08495653e-01 -5.86716056e-01 -1.91591874e-01
1.90180466e-01 1.47963027e-02 -8.22327077e-01 1.24809432e+00
1.64117679e-01 3.33843857e-01 -2.57106334e-01 1.06077909e+00
1.12790298e+00 3.89963359e-01 1.21356912e-01 7.18878880e-02
1.75129473e+00 -7.61834681e-01 -7.64644325e-01 -6.93639517e-02
2.92999476e-01 -6.97186589e-01 1.29641807e+00 2.01836497e-01
-8.80792320e-01 -7.59941041e-01 -1.32291555e+00 1.14251308e-01
-8.21293592e-01 1.40291885e-01 9.04949427e-01 1.37064207e+00
-7.77814329e-01 8.71804580e-02 -5.19180119e-01 -2.77713895e-01
3.78219754e-01 8.22591722e-01 8.48867968e-02 -2.82365270e-02
-1.15599930e+00 8.06146264e-01 -3.77891026e-02 3.27885568e-01
-6.78832412e-01 -6.27413690e-01 -9.16575730e-01 3.77493918e-01
8.34971294e-02 -4.79624271e-01 1.30407679e+00 -5.78141987e-01
-1.31432307e+00 6.86561286e-01 -9.10291746e-02 -1.02589667e-01
5.09079814e-01 2.12858960e-01 -6.31787479e-01 -3.69339675e-01
5.03015816e-02 7.60664567e-02 2.22127348e-01 -1.03605866e+00
-9.55295026e-01 -7.80834854e-01 1.17256939e-01 3.35276306e-01
-5.60505450e-01 -2.59667039e-01 -7.43516684e-01 2.08919868e-01
-5.21642342e-02 -7.06711352e-01 -1.30692832e-02 -4.41354215e-01
-1.38511151e-01 -2.55400509e-01 8.50952327e-01 -6.66896403e-01
1.36723614e+00 -2.15116048e+00 -9.27136302e-01 4.44277912e-01
8.10501128e-02 3.30024093e-01 4.57169443e-01 -2.07513973e-01
1.96187824e-01 -3.80493462e-01 4.98765498e-01 -3.64105165e-01
1.02052018e-01 2.15569124e-01 5.53439893e-02 6.07458532e-01
-5.05459845e-01 7.83916414e-01 -5.51683068e-01 -6.36886060e-01
9.53111172e-01 5.31526983e-01 -2.57320017e-01 3.53234619e-01
6.58216059e-01 6.63533155e-03 -3.16071063e-01 8.91621768e-01
8.72157574e-01 1.38008287e-02 1.52339503e-01 -2.13678733e-01
-5.42493582e-01 2.26461336e-01 -1.32468426e+00 1.36126232e+00
-9.65434074e-01 6.40195072e-01 4.76345494e-02 -8.11837852e-01
9.55054283e-01 2.27772042e-01 1.46976471e-01 -1.24421883e+00
5.73292613e-01 8.86114314e-02 -2.93584496e-01 -6.75364852e-01
8.54011893e-01 1.64960235e-01 -3.33837658e-01 -5.69204129e-02
-3.44746768e-01 4.88670081e-01 -3.07913512e-01 -3.43424380e-01
9.63013113e-01 -1.23535231e-01 1.20745324e-01 -1.96359679e-01
5.99941373e-01 -3.42965484e-01 3.63644809e-01 8.23136330e-01
-8.10387254e-01 3.26304704e-01 -2.75518000e-01 -5.69700241e-01
-8.57083797e-01 -1.05374670e+00 -4.30603743e-01 1.25800502e+00
2.69187540e-01 1.17678456e-01 -7.83796668e-01 -2.48486772e-01
2.25302622e-01 5.79662383e-01 -3.73365760e-01 -7.74480477e-02
-4.70601857e-01 -6.84930146e-01 3.89966458e-01 6.78716242e-01
8.67701173e-01 -9.23147857e-01 -7.79929996e-01 2.14649945e-01
-2.91295946e-01 -1.02940059e+00 -4.36845809e-01 2.16591313e-01
-5.29567838e-01 -8.98471117e-01 -5.97090304e-01 -8.85922909e-01
5.40611207e-01 6.56383395e-01 8.58234823e-01 1.43722847e-01
-6.94666743e-01 3.85684192e-01 2.83862978e-01 -5.81633925e-01
3.04756939e-01 2.72078276e-01 2.28534758e-01 -9.86144245e-02
1.08611870e+00 -3.81010443e-01 -1.12179840e+00 3.72996986e-01
-4.12929416e-01 -2.00624704e-01 5.43448508e-01 1.47265449e-01
3.36027712e-01 1.48494029e-02 7.22643316e-01 -6.85334325e-01
8.00915599e-01 -4.73534256e-01 -7.21870542e-01 1.38534084e-01
-6.91458702e-01 -6.15659058e-01 6.45039499e-01 -7.61652812e-02
-8.52429211e-01 1.16397545e-01 -2.96951532e-01 -3.32127094e-01
-4.24023688e-01 -1.28525903e-03 -3.04213613e-01 -1.44175723e-01
4.64395553e-01 5.56173980e-01 -2.85848826e-01 -4.31225240e-01
-1.07100867e-01 1.48535717e+00 8.45602691e-01 -4.22201782e-01
1.45170361e-01 2.94189990e-01 -3.33568573e-01 -7.90267885e-01
-4.72380340e-01 -9.96537089e-01 -3.32916558e-01 -5.51257253e-01
7.77531862e-01 -1.05323637e+00 -1.61041117e+00 1.39096066e-01
-7.14647412e-01 2.50122905e-01 2.60085434e-01 3.78397226e-01
-3.74513358e-01 3.06885779e-01 -4.30820554e-01 -1.33672285e+00
-5.71545005e-01 -1.05464292e+00 8.27985048e-01 8.45727026e-01
-1.17236726e-01 -6.47408247e-01 -3.12374473e-01 9.41808462e-01
5.99584758e-01 -1.96411610e-01 5.54253399e-01 -3.99351507e-01
-6.04810953e-01 -6.73581600e-01 -3.49743813e-01 -1.07743174e-01
1.50381997e-01 -4.15040463e-01 -1.38241124e+00 -2.25008607e-01
-3.47261131e-01 1.68054804e-01 3.47651958e-01 5.74492216e-01
1.70392263e+00 1.36215210e-01 -3.53014827e-01 5.05964339e-01
1.34218168e+00 6.43660069e-01 8.82923961e-01 8.39344442e-01
3.50348562e-01 1.58629358e-01 5.35640240e-01 5.92242002e-01
7.10207283e-01 5.75011194e-01 5.16026139e-01 -5.38765311e-01
7.51958936e-02 -1.56642318e-01 -1.11544922e-01 7.63333559e-01
-3.44082147e-01 2.83731483e-02 -8.69399548e-01 3.57247174e-01
-1.91111016e+00 -6.79410219e-01 8.60646591e-02 2.33270550e+00
4.20489311e-01 1.05497301e-01 1.01217724e-01 2.18345374e-01
6.92135334e-01 -2.42908493e-01 -4.25819963e-01 -4.33509201e-01
2.61709183e-01 3.55275542e-01 7.84809947e-01 4.17130828e-01
-1.27744663e+00 7.55980253e-01 6.12209368e+00 4.49980080e-01
-1.07864618e+00 1.46561995e-01 7.17774212e-01 -1.68691710e-01
1.09985076e-01 -6.74551666e-01 -1.11024189e+00 7.35962212e-01
1.10470176e+00 1.96106315e-01 3.71056169e-01 1.32737601e+00
4.06413168e-01 -3.82136375e-01 -9.38960314e-01 1.24676299e+00
2.75183231e-01 -1.24154818e+00 -4.06522423e-01 2.52627522e-01
1.62905514e-01 -3.49615157e-01 1.50688991e-01 9.03177440e-01
-1.91633254e-01 -1.02705848e+00 4.14797813e-01 6.34394169e-01
6.16285861e-01 -1.15978158e+00 1.19136894e+00 3.46055418e-01
-1.29827118e+00 -1.47911355e-01 -4.23060089e-01 -3.30635935e-01
-3.21088552e-01 4.46239114e-02 -1.00412202e+00 2.58318692e-01
9.22016859e-01 -2.43844554e-01 -3.97810996e-01 1.17299700e+00
3.86765391e-01 1.35750487e-01 -1.51309192e-01 -3.56011450e-01
1.21197671e-01 -4.42670621e-02 -2.00021282e-01 1.33328283e+00
9.28214490e-02 5.21693192e-02 3.24674070e-01 6.35821521e-01
4.80858423e-03 3.57982554e-02 -4.82134551e-01 5.55967152e-01
5.60625792e-01 1.54192853e+00 -4.89307046e-01 -3.04539979e-01
-5.28140903e-01 8.21543396e-01 1.22534931e-01 2.72735596e-01
-1.03415024e+00 -5.33123016e-01 4.58925754e-01 3.96646231e-01
6.20721020e-02 -1.10127769e-01 -3.86537969e-01 -7.46962547e-01
-1.60051733e-01 -5.60048580e-01 1.43972561e-01 -6.60100520e-01
-5.28184712e-01 1.88587189e-01 -4.31710362e-01 -8.76767397e-01
1.88764900e-01 -6.31976843e-01 -5.85748255e-01 1.17931354e+00
-1.69445121e+00 -9.32393253e-01 -1.03056121e+00 7.96132743e-01
6.98822439e-01 -1.94553569e-01 1.14344287e+00 1.04726148e+00
-1.06096148e+00 8.45849931e-01 -4.50794660e-02 2.36082435e-01
4.38699484e-01 -1.09643149e+00 -7.75637552e-02 4.74106699e-01
-6.17174566e-01 7.67134070e-01 5.16134262e-01 -3.00402224e-01
-1.79036343e+00 -9.81478453e-01 4.77590650e-01 -3.68703216e-01
2.64970064e-02 -5.76888084e-01 -5.34712613e-01 7.18594790e-01
-1.17827281e-01 -1.04539074e-01 1.10156071e+00 4.00159180e-01
1.99677154e-01 -2.54860252e-01 -1.18392193e+00 6.78691030e-01
6.32007122e-01 -6.15007043e-01 -3.56788993e-01 2.97609627e-01
2.18640789e-01 -4.82815266e-01 -7.67978191e-01 1.93253845e-01
9.57733452e-01 -7.98561394e-01 1.07261848e+00 -3.72675180e-01
1.07960686e-01 -5.91953956e-02 -5.18394947e-01 -7.21315801e-01
-5.72353601e-01 -1.77338924e-02 -6.68428466e-02 8.55722308e-01
2.70266235e-01 -4.66135502e-01 1.39103889e+00 1.15514421e+00
-4.26201493e-01 -8.31043541e-01 -9.94355738e-01 -3.85250628e-01
-5.20096838e-01 -3.32295954e-01 8.21987271e-01 9.43935990e-01
9.80866030e-02 2.50817180e-01 -1.27180308e-01 2.77767807e-01
5.29914320e-01 -1.31262830e-02 9.85963106e-01 -1.02255595e+00
-3.94811071e-02 -3.29943240e-01 -5.36756694e-01 -1.13220692e+00
-2.10967466e-01 -4.93394285e-01 6.84976205e-02 -1.45562458e+00
4.38028663e-01 -4.65898395e-01 -6.50310874e-01 3.24111938e-01
-5.87979183e-02 2.35952988e-01 -1.31599680e-01 8.13320652e-02
-7.83911049e-01 2.54144490e-01 9.78184044e-01 -1.16978176e-01
-2.66654611e-01 3.64859700e-01 -6.29287899e-01 5.70226431e-01
7.46489048e-01 2.35509560e-01 -8.57707858e-02 -7.87977967e-03
-2.00846553e-01 3.69701236e-02 2.32399806e-01 -1.07658732e+00
5.05543768e-01 1.15560316e-01 1.07514691e+00 -8.12378407e-01
6.07000530e-01 -1.09437871e+00 -2.34326929e-01 4.94646013e-01
3.36022645e-01 -1.00848237e-02 5.22795081e-01 2.44782463e-01
2.37493217e-01 -1.08026333e-01 5.18586576e-01 -3.73184144e-01
-8.98284137e-01 1.47943184e-01 -6.65369153e-01 -9.53856230e-01
1.11814427e+00 -5.91849506e-01 -2.52695709e-01 -3.02294284e-01
-5.62042475e-01 2.24172279e-01 7.26812403e-04 5.22395372e-01
6.48166716e-01 -1.28349555e+00 1.74127892e-02 5.21113873e-01
3.12181637e-02 -2.19890103e-01 4.15538073e-01 6.63813174e-01
-6.05732501e-01 9.17542100e-01 -2.26430774e-01 -6.34885550e-01
-1.57454693e+00 2.58885175e-01 5.50951362e-01 2.22940013e-01
-5.19440293e-01 5.08589745e-01 1.12112850e-01 -7.01390803e-01
7.54618049e-01 -1.71145454e-01 -4.45076466e-01 -1.98675081e-01
6.23956323e-01 4.11204666e-01 3.95488292e-01 -3.11538100e-01
-4.24853802e-01 1.63117602e-01 -2.60866821e-01 2.40328729e-01
1.00206268e+00 -2.59813875e-01 4.13454354e-01 3.19787443e-01
9.46951032e-01 -2.30256975e-01 -9.22205567e-01 -4.97167371e-02
-4.67619509e-01 -8.39454412e-01 6.51450634e-01 -7.55215168e-01
-7.89736390e-01 6.53181255e-01 1.41099346e+00 1.44029692e-01
1.04073572e+00 -3.77487630e-01 7.53858328e-01 3.04601282e-01
3.86451602e-01 -1.37444413e+00 -2.80679494e-01 2.86072791e-01
2.77032435e-01 -1.53382456e+00 -1.07047722e-01 -3.38363200e-01
5.82676893e-03 9.97632742e-01 9.10209596e-01 1.68072611e-01
5.52102447e-01 2.24957272e-01 8.23278502e-02 -2.04183146e-01
-6.55527576e-04 -8.45344141e-02 6.01819269e-02 2.44771644e-01
4.84450758e-01 6.28433943e-01 -7.95734450e-02 8.90654624e-01
-6.08700514e-01 1.16235726e-01 3.03402603e-01 1.15014434e+00
-9.23480332e-01 -4.05883968e-01 -7.69608974e-01 5.01128078e-01
-3.37489069e-01 -5.22576869e-02 9.77125466e-02 8.49665582e-01
2.49505371e-01 1.01095092e+00 5.94258070e-01 -4.87370014e-01
7.22032666e-01 -5.75951766e-03 4.33315903e-01 -3.61885041e-01
-7.13239253e-01 1.77333429e-02 4.29878049e-02 -1.88981891e-01
1.44769967e-01 -3.20529670e-01 -9.45647717e-01 -1.08504248e+00
-5.78483701e-01 -4.09491137e-02 1.21440685e+00 8.78367722e-01
2.38381058e-01 1.09195960e+00 6.60053611e-01 -7.69670725e-01
-3.13269168e-01 -9.74754035e-01 -6.96772814e-01 -9.67726298e-03
5.38368598e-02 -6.43333197e-01 -1.26268700e-01 -5.48649549e-01] | [8.185853958129883, -0.8962308168411255] |
c2634d30-bd6f-43ee-b89b-c742bffa426a | cross-modal-learning-for-image-guided-point | 2209.09552 | null | https://arxiv.org/abs/2209.09552v1 | https://arxiv.org/pdf/2209.09552v1.pdf | Cross-modal Learning for Image-Guided Point Cloud Shape Completion | In this paper we explore the recent topic of point cloud completion, guided by an auxiliary image. We show how it is possible to effectively combine the information from the two modalities in a localized latent space, thus avoiding the need for complex point cloud reconstruction methods from single views used by the state-of-the-art. We also investigate a novel weakly-supervised setting where the auxiliary image provides a supervisory signal to the training process by using a differentiable renderer on the completed point cloud to measure fidelity in the image space. Experiments show significant improvements over state-of-the-art supervised methods for both unimodal and multimodal completion. We also show the effectiveness of the weakly-supervised approach which outperforms a number of supervised methods and is competitive with the latest supervised models only exploiting point cloud information. | ['Enrico Magli', 'Diego Valsesia', 'Emanuele Aiello'] | 2022-09-20 | null | null | null | null | ['point-cloud-completion', 'point-cloud-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 2.78982699e-01 2.35472515e-01 -1.04801036e-01 -1.91066191e-01
-1.32314777e+00 -6.40011728e-01 9.74760473e-01 7.14453682e-03
-1.83674246e-01 4.97870982e-01 1.73682854e-01 6.98939860e-02
-8.14746618e-02 -5.02712250e-01 -1.14330268e+00 -7.46726692e-01
3.21336418e-01 8.14419925e-01 8.22172612e-02 -1.12169394e-02
2.67961681e-01 5.01527190e-01 -1.57749689e+00 3.67799640e-01
6.60566270e-01 5.69757581e-01 4.15662110e-01 6.03281975e-01
2.73530558e-02 3.53912055e-01 9.76732522e-02 -1.52987421e-01
3.80379438e-01 -1.76416546e-01 -7.81342983e-01 5.73131979e-01
7.90287316e-01 -1.99535847e-01 -2.30674312e-01 8.80176604e-01
2.02341214e-01 -9.64169670e-03 6.82509363e-01 -1.03484714e+00
-4.13935691e-01 -2.34421939e-01 -5.84646285e-01 -4.23456550e-01
6.60682142e-01 -1.64536163e-01 9.82108831e-01 -1.17270124e+00
1.06006765e+00 1.09370220e+00 5.45477808e-01 3.37386698e-01
-1.65945041e+00 -9.28826109e-02 1.10744461e-01 -1.65041909e-01
-1.09701049e+00 -6.02790177e-01 9.31762874e-01 -6.42961860e-01
7.68642306e-01 4.93370965e-02 3.46929669e-01 9.44812834e-01
-2.48966187e-01 6.89629614e-01 1.35616577e+00 -7.41762698e-01
2.94191211e-01 2.15087205e-01 -2.97844589e-01 9.17357802e-01
-2.96898216e-01 2.57786542e-01 -7.16898263e-01 -4.05256182e-01
1.07058525e+00 2.53559411e-01 -3.38359207e-01 -1.00941026e+00
-1.62263668e+00 8.36353838e-01 4.26060021e-01 2.46571660e-01
-4.73200113e-01 2.41447493e-01 4.22972342e-04 2.42631808e-01
8.90652955e-01 8.17178637e-02 -2.90366769e-01 -1.04290746e-01
-1.20821953e+00 1.07552754e-02 7.90761471e-01 1.11887109e+00
1.02770638e+00 -1.06704235e-01 2.60064960e-01 5.48533916e-01
5.94701469e-01 8.03082883e-01 -1.97853863e-01 -1.38856041e+00
6.68536067e-01 3.50057364e-01 2.99993813e-01 -6.20998621e-01
2.35534012e-02 -4.07555252e-01 -6.40197396e-01 6.44642830e-01
2.67353266e-01 4.05981690e-01 -7.92244434e-01 1.48155463e+00
3.70424211e-01 3.91065896e-01 4.67167795e-02 7.33023584e-01
2.89457709e-01 6.06403589e-01 -3.02467763e-01 -1.71094298e-01
8.48203123e-01 -1.00489902e+00 -4.58931297e-01 3.69070359e-02
4.15008634e-01 -9.58128154e-01 1.00653458e+00 5.96733451e-01
-1.33315384e+00 -6.18392348e-01 -9.86679733e-01 -1.89087108e-01
-1.07361026e-01 2.05861390e-01 5.16100466e-01 2.90544569e-01
-1.27442634e+00 8.35765183e-01 -1.30830371e+00 -4.84131366e-01
3.65168363e-01 1.67190582e-01 -8.27715099e-01 -4.28468972e-01
-3.47200006e-01 8.89313519e-01 -1.57087594e-01 -2.10228920e-01
-1.16998279e+00 -6.90444946e-01 -9.70497549e-01 -2.34904289e-01
2.15525806e-01 -1.05580294e+00 9.32994664e-01 -7.99695909e-01
-1.52467489e+00 1.05718863e+00 -3.44749421e-01 2.35805437e-02
7.81592727e-01 -3.34334582e-01 1.74019143e-01 6.73979759e-01
2.67657131e-01 7.97219217e-01 1.07831669e+00 -1.98598254e+00
-3.28246593e-01 -4.29214895e-01 8.82730708e-02 3.99007916e-01
1.31325141e-01 -3.96361917e-01 -6.14424169e-01 -2.79818028e-01
4.32338446e-01 -1.20761478e+00 -1.08561687e-01 3.55578154e-01
-1.60954237e-01 1.74388979e-02 9.75390315e-01 -6.44031465e-01
3.23661000e-01 -2.03209019e+00 7.79884517e-01 7.53402859e-02
8.87675583e-02 -2.77897865e-01 -2.12155908e-01 8.31785023e-01
-1.91877410e-01 -1.77235976e-01 -4.51053083e-01 -1.44372582e+00
1.92203134e-01 5.58226705e-01 -5.42081594e-01 9.64296162e-01
1.29347175e-01 8.58434737e-01 -1.04845798e+00 -3.79650086e-01
7.50834644e-01 6.73296988e-01 -4.23930585e-01 3.40322524e-01
-1.42381236e-01 1.05271995e+00 -2.63193369e-01 6.61454856e-01
6.65646374e-01 -3.80330741e-01 -1.70539737e-01 -1.21359237e-01
-1.29254982e-01 1.60771310e-01 -1.03823853e+00 2.74526167e+00
-7.11232066e-01 3.21504682e-01 3.48666519e-01 -1.09523976e+00
6.09450102e-01 6.07959628e-01 4.87465888e-01 -8.21794644e-02
-3.74884218e-01 2.78086543e-01 -8.12067270e-01 -3.30480963e-01
2.59970725e-01 -5.77810109e-01 2.59084702e-01 4.09241289e-01
3.34842056e-01 -5.32735467e-01 -3.48554343e-01 4.31414306e-01
9.35047805e-01 8.89472067e-01 -3.32046933e-02 -5.91683760e-02
4.31626916e-01 -4.63406444e-02 -6.75054593e-03 6.73240185e-01
2.59534985e-01 1.18577015e+00 1.18334018e-01 -6.13434613e-02
-1.27027583e+00 -1.29490590e+00 -2.09079102e-01 7.24216163e-01
6.61681965e-02 -3.30142468e-01 -4.21164960e-01 -6.61966383e-01
-1.17046542e-01 5.88287890e-01 -4.80264992e-01 2.75044829e-01
-3.16189289e-01 -7.68155931e-03 1.61853924e-01 5.05794764e-01
2.41880193e-01 -5.43687224e-01 -1.75546810e-01 -1.32049710e-01
-3.80747646e-01 -1.49549198e+00 -1.55220076e-01 -6.64933398e-02
-1.39235198e+00 -7.00558960e-01 -8.28314722e-01 -5.24917722e-01
9.47711349e-01 4.71841186e-01 1.10720193e+00 7.74769066e-03
2.78309554e-01 1.17673564e+00 -2.89837658e-01 9.51091722e-02
-4.65994239e-01 3.78790423e-02 1.07251294e-01 1.93880096e-01
-2.63941884e-01 -1.09988141e+00 -3.87649804e-01 1.04922630e-01
-1.01897931e+00 3.26791078e-01 5.48088014e-01 7.62284279e-01
8.40718508e-01 -2.79887170e-01 6.67056814e-02 -8.61531675e-01
-1.44356698e-01 -4.04540062e-01 -5.40062606e-01 1.60094440e-01
-4.59646463e-01 3.35703224e-01 4.05042470e-01 -1.37842640e-01
-1.02471542e+00 6.00588620e-01 4.17239554e-02 -9.12161708e-01
-2.76305437e-01 4.35968995e-01 -2.48757884e-01 -3.93075585e-01
3.54020447e-01 1.69951096e-01 2.16942847e-01 -7.31670439e-01
7.01582849e-01 2.95329303e-01 7.49561310e-01 -8.00009429e-01
1.19668972e+00 1.34782457e+00 3.91538620e-01 -7.45784521e-01
-6.20024979e-01 -9.52393591e-01 -1.15486217e+00 -1.59247652e-01
8.29284489e-01 -1.06953263e+00 -3.81901652e-01 9.91999954e-02
-1.49060738e+00 -9.80990976e-02 -5.18980920e-01 6.50826216e-01
-1.05367184e+00 7.93894410e-01 -4.89322543e-01 -6.60345197e-01
1.11428969e-01 -1.13049638e+00 1.90408945e+00 -3.73505294e-01
2.77458191e-01 -1.35315812e+00 3.66832048e-01 5.95578492e-01
2.44903211e-02 2.83286393e-01 3.67356569e-01 -1.87977687e-01
-1.14486158e+00 -3.28352600e-01 -9.56871584e-02 3.93393755e-01
-7.61147812e-02 -3.09413165e-01 -1.33822978e+00 -3.77169967e-01
3.08427393e-01 -3.08727741e-01 1.04712045e+00 1.99621707e-01
5.36129951e-01 9.31450129e-02 -2.91792244e-01 8.17647994e-01
1.66120374e+00 -6.45231068e-01 7.45031834e-01 9.20531973e-02
9.89231765e-01 7.17308700e-01 5.11706173e-01 1.60291925e-01
4.79093283e-01 7.49816477e-01 7.19609678e-01 -2.57596374e-01
-5.06979823e-02 -6.38877332e-01 3.36768329e-01 9.72543538e-01
-5.00928938e-01 1.66039929e-01 -8.11494410e-01 5.93833864e-01
-2.18477416e+00 -7.72259474e-01 -1.29489467e-01 2.37730026e+00
4.56725508e-01 -1.73430681e-01 -2.84821779e-01 1.38096750e-01
3.51226866e-01 3.58815312e-01 -3.15363854e-02 1.40979975e-01
1.47533361e-02 1.30507812e-01 4.02375013e-01 9.72541690e-01
-1.11934805e+00 6.94791257e-01 6.32827759e+00 6.16880238e-01
-7.13964522e-01 5.36461473e-01 5.79982363e-02 2.31474265e-01
-6.33120358e-01 5.70566237e-01 -3.88579160e-01 7.54062012e-02
6.81048751e-01 5.46053588e-01 5.51638544e-01 7.55726099e-01
5.26337363e-02 -1.87550932e-01 -1.63033259e+00 1.08564770e+00
4.22231346e-01 -1.42600965e+00 -9.20799598e-02 4.41343874e-01
1.02576935e+00 4.08208489e-01 1.17229506e-01 -1.00103341e-01
9.31612551e-02 -7.77360260e-01 6.84657097e-01 1.01933360e+00
9.20209885e-01 -2.47704163e-01 4.23670143e-01 7.10802138e-01
-9.51704144e-01 4.31740791e-01 -3.69835228e-01 -4.61018318e-03
4.50091720e-01 5.79574645e-01 -3.12629968e-01 1.06016886e+00
4.16635990e-01 1.07045650e+00 -3.76473665e-01 6.91853046e-01
-2.67601192e-01 4.24828142e-01 -5.37580669e-01 8.01739216e-01
2.01143876e-01 -5.46286702e-01 8.94760787e-01 8.81045699e-01
3.32781762e-01 5.73008670e-04 3.85415018e-01 9.03094292e-01
1.44750297e-01 -2.17911094e-01 -8.69263053e-01 3.17339838e-01
-1.38197765e-01 1.27277279e+00 -5.70952892e-01 -3.48707914e-01
-5.18930912e-01 1.25986934e+00 3.19051802e-01 5.17019331e-01
-3.73820394e-01 3.33460748e-01 -2.92976592e-02 1.28094852e-01
4.20237064e-01 -6.73929751e-01 -1.97152629e-01 -1.52964926e+00
2.82007098e-01 -4.10366356e-01 3.27850170e-02 -1.22646904e+00
-1.43720448e+00 3.72501403e-01 2.44959593e-01 -1.67183745e+00
-3.36057246e-01 -4.95397776e-01 -4.38106984e-01 1.04527366e+00
-1.78223777e+00 -1.73537540e+00 -2.93115258e-01 7.44650424e-01
3.49805951e-01 -1.00345328e-01 9.89348710e-01 3.60227488e-02
2.64874160e-01 -1.66779593e-01 3.56301904e-01 -4.16792035e-01
8.22188854e-01 -1.52474308e+00 2.23583788e-01 7.68889070e-01
5.16768396e-01 5.97300112e-01 6.42293990e-01 -5.67326963e-01
-1.77647257e+00 -6.86853647e-01 6.66085541e-01 -9.62622881e-01
5.20214081e-01 -4.25000548e-01 -8.50321352e-01 8.55943263e-01
1.86884686e-01 4.35396671e-01 3.54614377e-01 4.05302569e-02
-5.18511593e-01 3.21210101e-02 -9.29648161e-01 8.93751979e-02
9.22449172e-01 -1.02577925e+00 -7.01732576e-01 5.07021248e-01
6.06487751e-01 -4.97182280e-01 -6.98398352e-01 2.41692200e-01
1.89156651e-01 -8.56435537e-01 1.20377612e+00 -3.26957077e-01
5.92720330e-01 -3.53807867e-01 -5.02456188e-01 -1.31978512e+00
-3.06928493e-02 -6.35367930e-01 -1.90009117e-01 1.05759919e+00
4.02582176e-02 -4.77742642e-01 8.03920507e-01 3.44882995e-01
-3.45854670e-01 -3.17929000e-01 -1.14629900e+00 -5.35646319e-01
2.03157887e-02 -6.07002079e-01 1.40042081e-01 9.20899212e-01
-1.49200866e-02 3.55017960e-01 -5.19314885e-01 4.51741487e-01
1.08365452e+00 2.69046009e-01 8.44547749e-01 -1.22907162e+00
-6.26111329e-01 8.46834555e-02 -5.50229251e-01 -1.41463244e+00
3.55169505e-01 -1.08925569e+00 1.58271249e-02 -1.67344069e+00
2.25489527e-01 -3.31144810e-01 3.11515667e-02 3.19723755e-01
1.48053825e-01 3.98188144e-01 2.95220852e-01 6.69315398e-01
-7.40556300e-01 7.09696829e-01 1.17718089e+00 2.35666037e-02
7.08186254e-02 -5.00998721e-02 -3.25963378e-01 8.19384456e-01
2.08898038e-01 -5.31780779e-01 -3.94041121e-01 -6.02965057e-01
2.36547768e-01 4.49909091e-01 8.78194094e-01 -8.37697387e-01
4.65781599e-01 6.65230304e-02 1.77906409e-01 -7.29679227e-01
8.33649457e-01 -1.16784060e+00 2.77490348e-01 -1.24540124e-02
-1.72801226e-01 -1.39701784e-01 -8.19035023e-02 1.05527985e+00
-1.58386618e-01 -2.41005421e-01 4.44181263e-01 -1.18741289e-01
-1.30378053e-01 2.95808375e-01 2.18994424e-01 -3.53659362e-01
6.30864501e-01 -7.77851045e-02 -6.81921542e-02 -6.07278585e-01
-9.38122213e-01 -7.48767480e-02 1.01202428e+00 1.55602261e-01
6.76282227e-01 -1.41802680e+00 -7.51071453e-01 1.47158459e-01
3.41351241e-01 1.70131177e-01 1.25385612e-01 1.02738595e+00
-5.67568958e-01 3.08362335e-01 3.71632501e-02 -1.25472665e+00
-1.08888125e+00 6.18448317e-01 1.96344733e-01 -1.96712404e-01
-9.06989038e-01 2.70436168e-01 2.66743243e-01 -8.22749138e-01
1.42090186e-01 -1.41074017e-01 3.34507465e-01 -3.88722777e-01
1.52855173e-01 3.27391833e-01 1.97136819e-01 -9.54473674e-01
-9.61567163e-02 9.21773016e-01 2.29716927e-01 -8.08971405e-01
1.52213037e+00 -3.18165362e-01 -2.12358937e-01 7.91503787e-01
1.32951808e+00 1.16405368e-01 -1.65701914e+00 -5.22538245e-01
-2.54660279e-01 -6.94857240e-01 1.77921936e-01 -4.92488891e-01
-7.54129469e-01 1.16009820e+00 4.26184326e-01 -6.01628348e-02
8.62361550e-01 3.97631615e-01 3.22881222e-01 3.30127120e-01
6.85278356e-01 -7.52762675e-01 1.09677473e-02 8.71141627e-02
1.10336268e+00 -1.50697899e+00 4.32707518e-01 -5.39512038e-01
-6.25140727e-01 9.53288555e-01 -1.41449034e-01 -3.95960361e-01
7.92423427e-01 -7.12948851e-03 -1.53035119e-01 -3.84979665e-01
-7.08702207e-01 -1.68126315e-01 5.61193407e-01 7.77055621e-01
5.23367226e-02 -9.87373888e-02 2.19729081e-01 -9.83228534e-02
2.32689500e-01 -5.07502630e-02 3.22918087e-01 1.13370633e+00
-4.53375392e-02 -1.40343630e+00 -5.18610895e-01 -1.48492958e-02
-1.18038580e-01 -6.99946135e-02 -2.16358334e-01 8.29577267e-01
-6.94744885e-02 8.15569580e-01 -1.18980192e-01 -2.87502389e-02
1.93376794e-01 -5.76866651e-03 8.35787177e-01 -7.99210846e-01
-1.63228557e-01 5.87459862e-01 -1.77505374e-01 -7.34385252e-01
-1.02226460e+00 -9.03055727e-01 -1.01981258e+00 4.55253273e-02
-3.27697843e-01 6.27649650e-02 9.02420521e-01 1.05294132e+00
2.68866360e-01 4.72754836e-02 7.22641408e-01 -1.66778636e+00
-5.53999126e-01 -7.94138849e-01 -5.74152112e-01 5.55402279e-01
5.98278642e-01 -7.53830254e-01 -7.00505495e-01 4.39436734e-01] | [8.60417366027832, -3.0723557472229004] |
f46b520f-e0e6-436a-a0ae-9d0669959d11 | leveraging-real-conversational-data-for-multi | 2204.03232 | null | https://arxiv.org/abs/2204.03232v1 | https://arxiv.org/pdf/2204.03232v1.pdf | Leveraging Real Conversational Data for Multi-Channel Continuous Speech Separation | Existing multi-channel continuous speech separation (CSS) models are heavily dependent on supervised data - either simulated data which causes data mismatch between the training and real-data testing, or the real transcribed overlapping data, which is difficult to be acquired, hindering further improvements in the conversational/meeting transcription tasks. In this paper, we propose a three-stage training scheme for the CSS model that can leverage both supervised data and extra large-scale unsupervised real-world conversational data. The scheme consists of two conventional training approaches -- pre-training using simulated data and ASR-loss-based training using transcribed data -- and a novel continuous semi-supervised training between the two, in which the CSS model is further trained by using real data based on the teacher-student learning framework. We apply this scheme to an array-geometry-agnostic CSS model, which can use the multi-channel data collected from any microphone array. Large-scale meeting transcription experiments are carried out on both Microsoft internal meeting data and the AMI meeting corpus. The steady improvement by each training stage has been observed, showing the effect of the proposed method that enables leveraging real conversational data for CSS model training. | ['Takuya Yoshioka', 'Sefik Emre Eskimez', 'Naoyuki Kanda', 'Dongmei Wang', 'Xiaofei Wang'] | 2022-04-07 | null | null | null | null | ['speech-separation'] | ['speech'] | [ 5.98644197e-01 6.20856322e-02 4.38620716e-01 -5.01593411e-01
-1.65856278e+00 -3.93199742e-01 3.81836444e-01 -1.93206519e-01
-2.47512937e-01 4.43883479e-01 4.65375215e-01 -4.26071346e-01
9.54848342e-03 -1.31537259e-01 -5.43720901e-01 -8.92840624e-01
1.49286119e-02 5.23350716e-01 5.51843680e-02 -3.27121586e-01
-6.35657907e-02 1.78961575e-01 -1.51599503e+00 5.98009646e-01
9.92400467e-01 8.62418890e-01 4.38727945e-01 1.13233078e+00
-2.42278323e-01 4.25527990e-01 -9.74421561e-01 5.41661195e-02
2.25319684e-01 -5.48427343e-01 -2.81685591e-01 2.13766217e-01
9.78329107e-02 -4.91635539e-02 -3.39076251e-01 5.16862869e-01
1.21751940e+00 8.36966559e-02 3.77364457e-01 -1.14085400e+00
9.21972916e-02 6.75814569e-01 -3.58780563e-01 1.20683052e-01
5.22730649e-01 -2.96699498e-02 6.11949384e-01 -9.16175187e-01
-1.40746191e-01 1.10175002e+00 8.27606320e-01 5.71960807e-01
-8.85415435e-01 -9.57432389e-01 -9.01440829e-02 -1.66054711e-01
-1.44101346e+00 -9.77483511e-01 1.02350485e+00 -3.17056417e-01
1.00357294e+00 4.14012581e-01 4.79667127e-01 1.25843227e+00
-3.23961318e-01 9.15921330e-01 1.18091178e+00 -8.01930189e-01
1.28158569e-01 4.11621153e-01 -9.92541760e-02 1.44099444e-01
-5.81867814e-01 3.43799621e-01 -7.89262950e-01 -1.94990277e-01
4.82954443e-01 -1.95316136e-01 -5.94098449e-01 -6.18933514e-02
-1.02455318e+00 3.69060457e-01 -1.72509566e-01 3.89773101e-01
6.95760036e-03 -4.89540011e-01 4.40346360e-01 6.80570245e-01
5.96964240e-01 3.05780768e-01 -6.29602909e-01 -4.39274490e-01
-1.31121063e+00 -1.23717517e-01 1.03341901e+00 1.11034548e+00
4.41638678e-01 1.63452461e-01 1.26671776e-01 1.34158897e+00
3.89253139e-01 8.32884908e-01 8.70108962e-01 -4.35163021e-01
9.34924245e-01 2.68653661e-01 -1.01391457e-01 -8.27306390e-01
-3.20649326e-01 -5.97569883e-01 -6.87831998e-01 -2.81023532e-01
2.01744512e-01 -5.09505272e-01 -7.10457981e-01 1.57209444e+00
2.26857617e-01 6.32999659e-01 4.83431518e-01 7.57719457e-01
8.55923355e-01 7.94353068e-01 -6.60270452e-01 -5.46089113e-01
7.25152314e-01 -1.29552126e+00 -1.02901554e+00 -7.08095506e-02
6.63354814e-01 -9.82987225e-01 1.21002781e+00 7.67932057e-01
-1.11329818e+00 -9.15538847e-01 -1.24080086e+00 4.57258731e-01
-2.29209989e-01 3.26465845e-01 1.38093591e-01 9.45067585e-01
-1.08137584e+00 1.70921773e-01 -6.71726704e-01 -2.95625687e-01
-1.91646397e-01 5.65247953e-01 -2.49131411e-01 -1.41733840e-01
-1.10905766e+00 2.10550010e-01 1.42597342e-02 5.49596250e-01
-8.43162060e-01 -5.02848923e-01 -7.55158842e-01 1.39025245e-02
2.60217667e-01 -6.77704960e-02 1.44600010e+00 -1.23954928e+00
-2.09507108e+00 4.42722291e-01 -3.57201099e-01 -1.40896454e-01
5.21361470e-01 -3.26583266e-01 -1.05004394e+00 -6.07456118e-02
-1.28309280e-01 1.65159270e-01 9.78820443e-01 -1.62931073e+00
-5.05578697e-01 -3.42085153e-01 -4.13258553e-01 4.54260886e-01
-5.25481045e-01 -1.41024038e-01 -5.15340269e-01 -5.87782621e-01
2.26611182e-01 -1.01727414e+00 -5.24993911e-02 -8.20814073e-01
-5.57065368e-01 2.28584573e-01 8.50189328e-01 -7.74304807e-01
1.38998556e+00 -2.40881252e+00 -1.17418341e-01 4.59669054e-01
-2.75243491e-01 4.82198119e-01 -3.28152865e-01 6.97596490e-01
-2.39174947e-01 -3.58298451e-01 -2.66928017e-01 -8.99678409e-01
-1.86423346e-01 7.38518089e-02 -3.19381028e-01 1.63671434e-01
4.61681820e-02 3.34155440e-01 -6.80278361e-01 -4.20877457e-01
1.98661998e-01 3.19161952e-01 -4.88419622e-01 1.00705862e+00
3.03652912e-01 8.30047369e-01 -2.30928604e-02 3.99484992e-01
7.88809180e-01 2.07400158e-01 2.35055819e-01 -1.33560494e-01
-9.03686211e-02 4.53529835e-01 -1.41618907e+00 2.05282092e+00
-8.34892511e-01 6.20434523e-01 4.56810802e-01 -1.18128610e+00
1.18113732e+00 8.74864042e-01 3.26994210e-01 -5.76551914e-01
-9.93505027e-03 3.99678916e-01 4.49776277e-02 -7.49584675e-01
3.10360432e-01 -1.65733710e-01 -1.29576951e-01 3.52250606e-01
1.94611773e-01 -3.74254167e-01 -3.43888670e-01 4.56301123e-02
1.01044989e+00 -4.54136245e-02 -2.32712209e-01 -5.64214177e-02
8.45747709e-01 -1.86906293e-01 5.04083991e-01 7.89855123e-01
3.08842137e-02 1.04818130e+00 8.35087597e-02 3.16118956e-01
-8.30096364e-01 -9.60260928e-01 2.44568195e-02 9.27436471e-01
-9.85523090e-02 -4.60401028e-01 -8.92807186e-01 -5.66406131e-01
-3.94464940e-01 4.44475353e-01 -2.29848012e-01 -4.88475077e-02
-5.32271624e-01 -6.68736219e-01 7.52905309e-01 5.09005070e-01
4.37495142e-01 -7.76555598e-01 4.63803597e-02 4.26934272e-01
-3.34034204e-01 -1.26263869e+00 -5.55459023e-01 5.79696953e-01
-6.64167523e-01 -7.01635838e-01 -6.74985647e-01 -9.88020241e-01
4.72714841e-01 4.24031377e-01 7.10887015e-01 2.44803149e-02
8.90538692e-02 5.92067838e-01 -3.85857731e-01 -5.15044749e-01
-7.22030699e-01 1.26568437e-01 2.87597150e-01 3.44635576e-01
3.24475378e-01 -8.31031382e-01 -3.31630588e-01 6.07225120e-01
-5.65091372e-01 -1.40665406e-02 7.23478436e-01 1.08174121e+00
1.46564528e-01 1.24636687e-01 1.17092097e+00 -7.84227967e-01
6.54310405e-01 -3.18604082e-01 -1.47938326e-01 2.41492465e-01
-3.05649519e-01 -2.73523182e-01 7.47727454e-01 -6.47052824e-01
-1.42430532e+00 1.08544804e-01 -6.29684746e-01 -4.00908858e-01
-3.74845505e-01 5.19490898e-01 -7.24168181e-01 2.75385052e-01
4.88254637e-01 2.97629684e-01 9.60110277e-02 -4.98928696e-01
-1.18987046e-01 1.69714105e+00 6.24890089e-01 -5.72413325e-01
7.16393590e-01 9.11527574e-02 -8.77776980e-01 -1.11947620e+00
-6.12989962e-01 -9.98395979e-01 -6.59392416e-01 -5.95208257e-02
6.41383886e-01 -1.18566644e+00 -3.49190235e-01 7.08835065e-01
-9.52412009e-01 -5.55301130e-01 -2.29696129e-02 8.43885005e-01
-4.12927538e-01 3.08663964e-01 -5.79843521e-01 -1.13723588e+00
-8.97449329e-02 -1.27333415e+00 1.17942822e+00 -4.13374715e-02
-1.65487900e-02 -9.98635292e-01 2.99756438e-01 7.61953533e-01
4.74682510e-01 -2.98614711e-01 3.67083400e-01 -1.14408123e+00
-1.03691891e-01 -3.63815516e-01 2.33640909e-01 8.23283076e-01
3.37066531e-01 -2.36391693e-01 -1.65252972e+00 -5.44816017e-01
3.97555411e-01 -5.46642601e-01 3.59608412e-01 1.69808537e-01
1.04650462e+00 3.41846049e-03 -3.62349719e-01 3.54519516e-01
9.57140386e-01 2.92061567e-01 5.07512510e-01 -1.51930407e-01
6.77401781e-01 5.44496357e-01 6.37425721e-01 2.88936645e-01
2.74335295e-01 6.28806949e-01 -9.88274589e-02 -5.37924051e-01
5.81570789e-02 -3.38924766e-01 5.26578069e-01 2.02313375e+00
3.86468828e-01 -3.60311031e-01 -8.81378829e-01 5.46927154e-01
-1.84450901e+00 -6.52384400e-01 -7.21332803e-02 2.33107352e+00
9.52142000e-01 2.64301658e-01 4.16504629e-02 5.93325257e-01
6.23104036e-01 -5.49179725e-02 -2.94136435e-01 -4.01330441e-01
-2.83424612e-02 1.31807551e-01 4.44988199e-02 5.69395721e-01
-9.42367971e-01 5.20969748e-01 5.98137426e+00 9.89275038e-01
-1.43006098e+00 1.63071781e-01 4.56897587e-01 -3.61118093e-02
-2.48064503e-01 -3.31542969e-01 -6.13917172e-01 4.10315216e-01
1.42626071e+00 2.58422375e-01 3.70493293e-01 5.37656963e-01
5.66860199e-01 -1.22201085e-01 -1.21564400e+00 1.20833397e+00
3.72802019e-01 -8.93199503e-01 -2.59031594e-01 -1.14770472e-01
5.52595437e-01 -1.75739944e-01 -1.09429292e-01 6.21620893e-01
-9.56445001e-03 -1.02543938e+00 4.31816697e-01 1.95767537e-01
8.71346951e-01 -5.33331871e-01 1.01000285e+00 7.41127729e-01
-1.21764445e+00 -1.60839126e-01 -2.39560511e-02 7.35815838e-02
7.65881613e-02 5.48992038e-01 -1.24472821e+00 8.66057217e-01
5.93475103e-01 4.25683677e-01 -3.13408732e-01 8.49049568e-01
1.76443234e-01 1.25140178e+00 -3.67326796e-01 1.86070010e-01
-1.26795014e-02 -1.24284983e-01 4.16709483e-01 1.69282949e+00
3.94166440e-01 5.93929477e-02 3.69385839e-01 2.20954984e-01
2.57318199e-01 1.34319976e-01 -5.07996798e-01 2.26590186e-01
7.09778607e-01 1.12934244e+00 -2.59478420e-01 -1.71849117e-01
-5.41824520e-01 9.20965970e-01 -3.59999053e-02 5.63422620e-01
-7.90119708e-01 -3.51492822e-01 1.87048122e-01 1.01169467e-03
2.25871176e-01 -3.21305007e-01 -1.80681974e-01 -9.60911691e-01
-2.01193560e-02 -1.35534024e+00 -1.66272721e-03 -7.03581750e-01
-1.22159278e+00 8.96821976e-01 -1.85304955e-01 -1.46651173e+00
-4.75544959e-01 -3.08684796e-01 -8.32947493e-01 1.03168595e+00
-1.36860895e+00 -1.06747019e+00 -2.27116108e-01 7.44184732e-01
8.74417484e-01 -5.25536358e-01 1.00304699e+00 6.85408235e-01
-6.01899087e-01 1.04330969e+00 4.09399569e-01 7.38495290e-02
9.98945951e-01 -1.22990572e+00 1.23665459e-01 4.84695762e-01
1.99124902e-01 3.98306698e-01 6.54018819e-01 -1.52864859e-01
-1.26024950e+00 -7.83242404e-01 6.64543450e-01 -2.48120129e-01
4.37331706e-01 -9.65211332e-01 -1.03624177e+00 4.97047901e-01
2.24220440e-01 -2.21569568e-01 1.28553450e+00 1.57503560e-01
1.05361603e-01 -3.82574886e-01 -9.18558836e-01 2.82372504e-01
8.61362696e-01 -7.16076791e-01 -7.50973105e-01 7.67427161e-02
7.51674294e-01 -4.71152902e-01 -8.36524189e-01 3.73284787e-01
4.53740656e-01 -8.05714190e-01 6.93687618e-01 -1.50097176e-01
1.38410613e-01 -1.79372475e-01 -3.64170194e-01 -1.74567509e+00
3.20358694e-01 -9.23316598e-01 3.06850135e-01 1.66039002e+00
8.13289046e-01 -5.14550209e-01 5.70885420e-01 1.56880379e-01
-6.11499786e-01 -5.47752142e-01 -9.32474375e-01 -6.56139851e-01
-1.63842335e-01 -7.53467619e-01 6.12872601e-01 1.05573285e+00
2.50258237e-01 6.82412803e-01 -3.14956337e-01 4.60220367e-01
2.84855813e-01 -2.48818621e-01 1.26784396e+00 -8.54925156e-01
-7.10936487e-01 2.38263711e-01 -7.87132308e-02 -1.60975075e+00
6.83584884e-02 -4.58600432e-01 5.75133920e-01 -1.12699163e+00
-2.70054907e-01 -8.70376527e-01 -1.22090772e-01 -3.68889682e-02
-1.05031610e-01 -1.63376451e-01 -1.09856717e-01 1.06202044e-01
-5.12732446e-01 7.25204766e-01 1.04540467e+00 -3.03472709e-02
-5.84774673e-01 3.51489693e-01 -3.60066742e-01 5.11635184e-01
5.97327650e-01 -3.81494164e-01 -8.31034422e-01 -2.54900336e-01
-3.68254602e-01 6.13358438e-01 -4.36181761e-02 -1.34757090e+00
4.18063730e-01 3.12634557e-01 7.91058540e-02 -6.36618435e-01
6.03583157e-01 -1.01393795e+00 -1.26441136e-01 -6.66393265e-02
-5.10836720e-01 -6.70263991e-02 4.92106736e-01 6.98346615e-01
-6.12587690e-01 1.13366945e-02 3.07851255e-01 2.50361264e-01
-1.59788325e-01 -1.84131220e-01 -3.01290005e-01 9.47145000e-02
5.27178645e-01 -2.68133491e-01 1.22516297e-01 -7.14788258e-01
-9.58693564e-01 2.28686199e-01 -1.08290754e-01 4.10728991e-01
5.34150302e-01 -1.22107363e+00 -7.95203030e-01 6.80720925e-01
3.80429626e-02 3.46371621e-01 2.60521680e-01 9.77502108e-01
1.14258207e-01 2.41140857e-01 3.77310097e-01 -1.11038470e+00
-1.38010240e+00 2.79494077e-01 4.32683736e-01 -2.27356553e-01
-3.17502856e-01 8.10118258e-01 1.47873253e-01 -1.16099167e+00
6.31525278e-01 -3.27797085e-01 -3.02150678e-02 -1.97515205e-01
4.05501753e-01 8.13895911e-02 5.08834183e-01 -4.24009800e-01
-9.31468830e-02 3.83910447e-01 -4.25936282e-02 -5.72403908e-01
1.21476316e+00 -4.70434129e-01 3.79781544e-01 1.05697250e+00
1.35395372e+00 4.84565824e-01 -1.22609544e+00 -3.96203756e-01
-1.23941481e-01 -3.67222130e-01 -1.05400607e-01 -8.55838656e-01
-8.06333125e-01 1.10791802e+00 7.80000269e-01 3.38637918e-01
1.23081291e+00 -2.58996964e-01 8.56490374e-01 2.90543228e-01
3.12637001e-01 -1.26784396e+00 4.07430440e-01 2.50746369e-01
9.29435372e-01 -1.40456688e+00 -5.87112308e-01 -4.98236001e-01
-6.84945107e-01 9.00280118e-01 6.30572319e-01 1.85666665e-01
8.95642877e-01 7.60013103e-01 6.98739469e-01 2.17980757e-01
-6.73462331e-01 1.87522039e-01 1.55352861e-01 6.57989383e-01
5.38826346e-01 -5.07782027e-02 2.84369558e-01 1.08733904e+00
-5.03312111e-01 -7.11467266e-02 3.26875359e-01 8.96183074e-01
-1.86807930e-01 -1.10372031e+00 -6.88440740e-01 3.29707682e-01
-1.08458310e-01 -1.98700681e-01 -3.06124389e-01 6.04529977e-01
7.45924097e-03 1.55088508e+00 8.20740014e-02 -7.17661083e-01
5.77651799e-01 2.62428433e-01 1.62852794e-01 -6.67081058e-01
-8.25046539e-01 6.55455709e-01 3.55854005e-01 -1.18304335e-01
-6.37887836e-01 -7.82940805e-01 -1.18025482e+00 -1.58231109e-02
-9.11398888e-01 6.71520174e-01 7.01013744e-01 1.10053432e+00
2.51338869e-01 8.16862583e-01 1.15341377e+00 -9.66052949e-01
-6.40935302e-01 -1.28626597e+00 -5.90756178e-01 2.03072920e-01
6.53159976e-01 -3.07780057e-01 -6.58003747e-01 8.12139735e-02] | [14.666594505310059, 6.3971686363220215] |
abb3a7d6-88dc-4d7c-9a31-8d08b71527a0 | an-ensemble-of-machine-learning-and-anti | 1607.06190 | null | http://arxiv.org/abs/1607.06190v1 | http://arxiv.org/pdf/1607.06190v1.pdf | An ensemble of machine learning and anti-learning methods for predicting tumour patient survival rates | This paper primarily addresses a dataset relating to cellular, chemical and
physical conditions of patients gathered at the time they are operated upon to
remove colorectal tumours. This data provides a unique insight into the
biochemical and immunological status of patients at the point of tumour removal
along with information about tumour classification and post-operative survival.
The relationship between severity of tumour, based on TNM staging, and survival
is still unclear for patients with TNM stage 2 and 3 tumours. We ask whether it
is possible to predict survival rate more accurately using a selection of
machine learning techniques applied to subsets of data to gain a deeper
understanding of the relationships between a patient's biochemical markers and
survival. We use a range of feature selection and single classification
techniques to predict the 5 year survival rate of TNM stage 2 and 3 patients
which initially produces less than ideal results. The performance of each model
individually is then compared with subsets of the data where agreement is
reached for multiple models. This novel method of selective ensembling
demonstrates that significant improvements in model accuracy on an unseen test
set can be achieved for patients where agreement between models is achieved.
Finally we point at a possible method to identify whether a patients prognosis
can be accurately predicted or not. | ['John Scholefield', 'Durga Suryanarayanan', 'Christopher Roadknight', 'Uwe Aickelin', 'Lindy Durrant'] | 2016-07-21 | null | null | null | null | ['tumour-classification'] | ['medical'] | [ 5.05629361e-01 -6.84451908e-02 -4.28728908e-01 -2.22050756e-01
-8.90845180e-01 -3.41900438e-01 5.04736841e-01 1.00873387e+00
-7.04680145e-01 7.81910658e-01 3.53043616e-01 -5.28302729e-01
-4.77192461e-01 -7.04420090e-01 1.18403696e-01 -9.45460021e-01
-5.21487772e-01 7.35511959e-01 -9.33385491e-02 -3.49790484e-01
1.47014782e-01 6.94514751e-01 -9.89086807e-01 4.65528071e-01
4.38820064e-01 7.11911142e-01 1.08466730e-01 1.01600277e+00
4.15524617e-02 6.31945014e-01 -4.63123679e-01 1.42078832e-01
4.33585532e-02 -5.79601884e-01 -8.65290165e-01 -6.50203973e-02
-1.03476502e-01 1.18430965e-01 -3.11923902e-02 5.63362598e-01
3.78733873e-01 -2.81957507e-01 9.10427928e-01 -7.69099176e-01
1.17707014e-01 1.07713163e-01 -1.91080078e-01 9.75580364e-02
3.00500065e-01 3.02792657e-02 6.81750476e-01 -3.86887908e-01
3.26070279e-01 6.42592967e-01 9.09788370e-01 5.88111579e-01
-1.56433511e+00 -1.46968737e-01 -4.29243296e-01 -2.58867115e-01
-1.29907703e+00 -6.19593620e-01 1.05891556e-01 -6.37392581e-01
6.55553877e-01 9.36509907e-01 8.15058827e-01 3.69398415e-01
7.53144503e-01 1.32230043e-01 1.16892266e+00 -5.18997073e-01
1.41893208e-01 3.63284200e-01 -1.48802646e-03 8.40040028e-01
2.33152837e-01 3.71860474e-01 -3.71367127e-01 -4.46900308e-01
2.89626449e-01 1.40227467e-01 -3.60534042e-01 -2.74712563e-01
-1.42393517e+00 6.83194578e-01 2.95261234e-01 5.17415404e-01
-1.52298599e-01 -8.75605121e-02 7.00150788e-01 5.66611230e-01
4.39527094e-01 8.44063997e-01 -8.23070765e-01 1.65482312e-01
-8.62293005e-01 -2.37087697e-01 9.07378137e-01 2.07391143e-01
3.67763311e-01 -5.01425922e-01 2.05957517e-01 8.72057319e-01
1.21063314e-01 -9.11264196e-02 1.04048204e+00 -4.16399240e-01
-5.21110237e-01 7.22393155e-01 3.48367132e-02 -6.11679077e-01
-8.80779982e-01 -5.51218808e-01 -8.46050143e-01 2.69160926e-01
5.56500912e-01 5.53723723e-02 -8.10857236e-01 1.36919856e+00
3.33609693e-02 -3.29748780e-01 3.61503273e-01 5.16482532e-01
6.96882665e-01 7.02345967e-02 2.89858282e-01 -3.14513028e-01
1.46256304e+00 -3.85172009e-01 -2.25977935e-02 -2.57198483e-01
1.31474721e+00 -5.72413683e-01 3.51707786e-01 4.65378374e-01
-6.93174124e-01 -9.90998894e-02 -7.82768011e-01 1.70225441e-01
-4.87753987e-01 1.30499154e-01 9.76777136e-01 7.14551032e-01
-1.10556483e+00 7.98539102e-01 -9.03806567e-01 -1.07264364e+00
6.50878027e-02 8.71957839e-01 -8.29590738e-01 8.49308632e-03
-9.79517281e-01 1.22106516e+00 4.05527741e-01 -1.16292454e-01
-5.58638334e-01 -5.66547692e-01 -7.03033626e-01 -2.59785235e-01
-1.22043043e-01 -9.32640374e-01 8.61334860e-01 -8.57568800e-01
-9.96502101e-01 1.08209157e+00 -2.09748641e-01 -3.94036919e-01
2.15318158e-01 7.40889609e-01 -3.53138030e-01 -5.67075051e-02
-3.59836891e-02 3.17635149e-01 1.47632226e-01 -6.47042632e-01
-7.76760757e-01 -4.40690279e-01 -3.69446278e-01 3.66168678e-01
-2.01284423e-01 7.18843788e-02 1.00531816e-01 -3.01223367e-01
2.98147023e-01 -9.53444004e-01 -6.68232679e-01 1.12527013e-01
1.10397741e-01 8.15547258e-02 1.21661082e-01 -5.91834426e-01
8.62168789e-01 -1.88653076e+00 2.24674344e-01 1.88749760e-01
3.11508238e-01 3.37917954e-02 -3.04228235e-02 8.13136101e-01
-1.98917821e-01 5.72664142e-01 9.83872563e-02 1.04763731e-01
-4.74347204e-01 1.51408285e-01 4.90961760e-01 7.68189371e-01
1.62598729e-01 7.65984416e-01 -8.56589854e-01 -4.85867083e-01
2.31314272e-01 2.78085053e-01 -2.86509514e-01 -1.02756999e-01
1.65647358e-01 4.10988450e-01 -2.76631892e-01 6.45453453e-01
-3.65421409e-03 -2.01544032e-01 5.16335726e-01 1.38766006e-01
1.62599504e-01 1.34350196e-01 -3.35127950e-01 1.34943640e+00
-4.55136836e-01 5.34303129e-01 1.26475096e-01 -9.00476992e-01
9.03525114e-01 4.72814113e-01 6.63775742e-01 -3.64989847e-01
2.16085896e-01 3.37226927e-01 3.91633779e-01 -2.20472112e-01
1.99282870e-01 -9.03722227e-01 -2.63983250e-01 1.01381302e-01
-2.48398021e-01 5.27049601e-02 6.85782284e-02 5.33614121e-02
1.64048970e+00 -5.59335649e-01 8.11369538e-01 -6.25358284e-01
7.37889946e-01 4.14067656e-01 4.15074676e-01 5.36443293e-01
-2.95912445e-01 3.69472623e-01 6.45988464e-01 -4.93226856e-01
-7.45751798e-01 -6.72192931e-01 -6.21735811e-01 7.74625659e-01
-6.87701181e-02 -1.68947175e-01 1.50290504e-01 -5.47071993e-01
2.84121603e-01 3.57334107e-01 -1.12719107e+00 -3.69031280e-01
8.82478133e-02 -1.48031986e+00 4.15795773e-01 2.67110258e-01
-1.79274976e-01 -4.54114050e-01 -2.65475124e-01 2.22969428e-01
9.40102190e-02 -2.27230012e-01 1.72513068e-01 7.79680133e-01
-1.24451637e+00 -1.37899446e+00 -5.51737607e-01 -8.30619097e-01
1.02822113e+00 -1.27124771e-01 9.42632198e-01 7.19823241e-01
-3.83808613e-01 5.27612753e-02 -2.96986490e-01 -3.74659002e-01
-9.02626812e-01 1.24004982e-01 2.13654358e-02 -2.12397501e-01
3.80080432e-01 -9.47183520e-02 -6.46506310e-01 4.46226925e-01
-6.68103039e-01 -3.75084095e-02 8.96791756e-01 1.09621811e+00
3.98583859e-01 4.11336273e-01 4.14845675e-01 -9.77298856e-01
3.89379531e-01 -6.32708132e-01 1.88674018e-01 1.31904647e-01
-7.74574101e-01 1.58339161e-02 4.89094675e-01 -2.21996620e-01
-5.18520892e-01 1.58824414e-01 -3.51952426e-02 3.98680001e-01
-3.20575684e-01 8.05132926e-01 2.71418542e-01 -3.47442567e-01
7.49247134e-01 1.85388610e-01 5.43305218e-01 -4.01505828e-01
-4.44430143e-01 6.89082503e-01 3.40458959e-01 -1.89632535e-01
2.70745844e-01 4.92872119e-01 4.62516904e-01 -8.33798885e-01
-5.11817276e-01 -9.42676723e-01 -6.49176478e-01 4.97062579e-02
3.59116435e-01 -6.82217181e-01 -6.35009885e-01 4.67346817e-01
-2.68770874e-01 -4.62345153e-01 -1.31982744e-01 7.20432699e-01
-4.19763446e-01 2.50792623e-01 -6.77823663e-01 -6.16212606e-01
-2.86652684e-01 -8.17750216e-01 8.33515286e-01 -2.35292405e-01
-7.70798981e-01 -1.53853989e+00 2.23554730e-01 2.98837930e-01
4.26246256e-01 5.08723021e-01 1.28083801e+00 -1.11553454e+00
7.73993433e-02 -9.02011991e-01 1.37282625e-01 -1.10860735e-01
3.59409809e-01 -7.64615685e-02 -5.29508829e-01 -6.63406014e-01
-5.25574572e-02 -1.82954639e-01 8.89108121e-01 1.44369468e-01
5.39947927e-01 -5.28996959e-02 -8.96601737e-01 4.52252567e-01
1.57883072e+00 2.46397108e-01 4.99996662e-01 4.87302423e-01
3.30404826e-02 8.12429428e-01 4.99709159e-01 -3.97079512e-02
6.64410228e-03 4.34023589e-01 2.20393389e-01 -1.47978529e-01
7.03741536e-02 8.48274529e-02 -3.68932486e-02 4.67090100e-01
1.20809317e-01 -2.59255767e-01 -1.17084515e+00 6.72439098e-01
-1.25353968e+00 -6.99359357e-01 -3.64005417e-01 2.41772199e+00
8.33113849e-01 1.84470385e-01 7.57916346e-02 4.79832798e-01
4.85290289e-01 -2.99289882e-01 -2.57693410e-01 -5.76931536e-01
1.56480804e-01 -3.55452076e-02 7.46130943e-01 5.71784556e-01
-6.79106116e-01 4.49366383e-02 7.74190283e+00 5.83023965e-01
-1.21520162e+00 -4.25838530e-01 9.35737908e-01 -1.64080903e-01
4.92681153e-02 2.93692946e-01 -4.06174123e-01 3.39480996e-01
1.25811756e+00 -3.04205209e-01 8.58371258e-02 2.21630141e-01
2.21751973e-01 -6.42686725e-01 -1.23682141e+00 3.28920841e-01
2.25718543e-02 -1.05561817e+00 -4.45423126e-01 5.03584921e-01
3.47933561e-01 -2.56952606e-02 -2.28055999e-01 2.30423316e-01
2.07673028e-01 -1.06819320e+00 1.96469761e-02 8.28167021e-01
9.84233677e-01 -5.07470429e-01 1.21689224e+00 6.45669103e-01
-6.68998778e-01 -6.79431260e-02 4.45643030e-02 -2.90303886e-01
-3.54401261e-01 3.93567383e-01 -1.57873380e+00 4.47204977e-01
1.42117843e-01 4.13024992e-01 -9.13982570e-01 1.13101768e+00
4.78753448e-01 3.33281904e-01 -3.48156929e-01 -3.32716584e-01
-1.75865680e-01 3.74839455e-01 1.53367952e-01 9.32079434e-01
2.70768911e-01 2.36974820e-01 -1.49795726e-01 -5.66395838e-03
5.41293442e-01 4.63237882e-01 -2.91928411e-01 -1.27631769e-01
9.96484980e-02 1.24649012e+00 -1.10566175e+00 -2.60984212e-01
-2.36774743e-01 7.85000861e-01 1.96448892e-01 -2.32275099e-01
5.11928499e-02 -1.57134593e-01 3.08408260e-01 3.62829685e-01
-1.95436448e-01 1.77195996e-01 -3.80243421e-01 -1.01117241e+00
-8.36641371e-01 -6.77936912e-01 6.59922123e-01 -5.32752275e-01
-1.12445247e+00 3.32985759e-01 -3.51493269e-01 -9.58743036e-01
-2.34672576e-01 -7.86640108e-01 -6.67397797e-01 1.07824647e+00
-9.81921732e-01 -6.52016640e-01 -1.19633533e-01 -1.64321512e-01
1.76242113e-01 -7.06393197e-02 1.37913823e+00 -3.01567525e-01
-4.50836867e-01 3.11891735e-01 4.15789843e-01 5.79164326e-02
8.29835534e-01 -1.44876945e+00 -2.13649496e-01 1.70013517e-01
-4.34920758e-01 5.37573338e-01 7.48609126e-01 -7.62762904e-01
-1.20993507e+00 -9.60395396e-01 1.07836390e+00 -7.00779438e-01
5.22352517e-01 5.07358611e-02 -6.34065270e-01 3.44903499e-01
-2.17576921e-01 -2.24701270e-01 1.39715838e+00 4.28039074e-01
3.66704315e-01 -1.07962362e-01 -1.40480340e+00 2.74305999e-01
3.42829436e-01 -3.56012464e-01 -3.00412744e-01 3.85243058e-01
-1.56561822e-01 -2.29537144e-01 -1.43006432e+00 5.06174743e-01
7.58499086e-01 -9.26759303e-01 8.26732278e-01 -6.93053067e-01
2.16612697e-01 -2.32749090e-01 -9.38840806e-02 -1.39685440e+00
-5.89782417e-01 -1.45872727e-01 7.01190293e-01 7.78827488e-01
7.38340914e-01 -7.51266241e-01 9.73894715e-01 9.79556739e-01
3.43405843e-01 -1.24269366e+00 -5.89429259e-01 -5.11633277e-01
3.31177771e-01 1.44960299e-01 3.82224321e-01 8.77342641e-01
4.19457406e-01 1.22249097e-01 1.89858750e-01 -1.71458572e-01
3.75898570e-01 -1.62747148e-02 5.60407758e-01 -1.47798312e+00
-2.88284063e-01 -4.33294505e-01 -1.08544970e+00 -9.60610434e-02
-3.04111391e-01 -1.10490835e+00 -2.72908300e-01 -1.56507683e+00
4.33535308e-01 -7.56587148e-01 -6.27293110e-01 3.65912288e-01
-4.22517508e-02 3.98985893e-01 -2.30596617e-01 4.95197952e-01
-1.26774222e-01 -1.44521683e-01 7.52909005e-01 1.11665666e-01
-2.39716932e-01 1.26941845e-01 -9.47711170e-01 6.06415927e-01
8.99146318e-01 -5.37916899e-01 -2.27813646e-01 2.11583629e-01
1.10685177e-01 6.43387198e-01 4.26869094e-01 -8.75366211e-01
1.34690210e-01 -3.18939775e-01 8.93783391e-01 -2.86467791e-01
3.11623693e-01 -7.74646461e-01 5.93162835e-01 1.05394721e+00
-5.71495533e-01 -3.84478480e-01 2.15826109e-01 6.64717257e-01
6.44837692e-02 -4.35735017e-01 7.54491270e-01 -4.74626452e-01
-5.40154815e-01 4.27142829e-02 -7.51022100e-01 -4.43504900e-01
1.22072470e+00 -4.14536834e-01 -7.60486424e-02 -1.99799374e-01
-1.12927675e+00 1.86614513e-01 1.10423625e+00 -1.52042389e-01
2.25011721e-01 -1.07943892e+00 -9.25708592e-01 2.77468771e-01
3.41236830e-01 -2.77290434e-01 -1.01463221e-01 1.39614367e+00
-7.99465716e-01 7.33641088e-01 -5.77368364e-02 -4.15478289e-01
-1.63439691e+00 6.09329462e-01 7.97777414e-01 -4.48028564e-01
-1.29896626e-01 8.53933930e-01 -4.94145714e-02 -1.97599202e-01
-2.41706148e-01 6.01801276e-02 -1.77409649e-01 1.81040883e-01
2.29264945e-01 1.88846141e-01 2.64011562e-01 -5.70407033e-01
-4.14428771e-01 1.21597797e-01 -3.72208714e-01 5.97605966e-02
1.17580080e+00 -1.37725770e-01 -3.65853786e-01 5.76352596e-01
1.41070127e+00 -9.14958399e-03 -5.45862079e-01 5.69876991e-02
-4.67523001e-02 -4.36982095e-01 3.16815346e-01 -1.15846729e+00
-6.45883739e-01 4.69714075e-01 5.60293317e-01 3.29265565e-01
1.13185203e+00 4.58131507e-02 3.61061096e-01 4.82936017e-02
2.23847702e-01 -4.28234756e-01 -5.38370669e-01 -8.64047259e-02
5.87172687e-01 -1.25734842e+00 3.63083601e-01 -2.03517050e-01
-3.57318103e-01 1.11461270e+00 9.50503722e-02 -3.17847617e-02
5.19562006e-01 8.52310956e-02 2.55409807e-01 -1.33640006e-01
-1.22526944e+00 -2.91497577e-02 9.74793360e-02 5.40064633e-01
6.59878910e-01 2.71439135e-01 -6.55607998e-01 3.28157336e-01
-3.03328007e-01 6.85083866e-02 6.11648083e-01 8.80238056e-01
-5.71679890e-01 -1.45328987e+00 -3.16449493e-01 1.16445363e+00
-7.08030045e-01 -8.68817568e-02 -8.17839861e-01 8.43072474e-01
-1.17690586e-01 7.81116545e-01 -8.61591175e-02 -2.83267617e-01
6.26454949e-02 2.43174106e-01 4.36101854e-01 -7.49406457e-01
-6.34358704e-01 -2.73432322e-02 5.21933973e-01 9.20111760e-02
-2.33326137e-01 -9.84162390e-01 -9.57306683e-01 -3.78273696e-01
-6.78559601e-01 4.91231561e-01 6.65757537e-01 7.19830096e-01
-6.64441437e-02 3.90036136e-01 6.42262518e-01 -4.30989176e-01
-3.71821523e-01 -7.10079730e-01 -9.23572719e-01 1.18189171e-01
5.95550358e-01 -3.43624264e-01 -7.21006930e-01 -2.22503748e-02] | [15.165712356567383, -3.065053701400757] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.