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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
7f2b52a7-e416-4c69-a27f-5c6d4db5e6dd | coconet-coupled-contrastive-learning-network | 2211.1096 | null | https://arxiv.org/abs/2211.10960v1 | https://arxiv.org/pdf/2211.10960v1.pdf | CoCoNet: Coupled Contrastive Learning Network with Multi-level Feature Ensemble for Multi-modality Image Fusion | Infrared and visible image fusion targets to provide an informative image by combining complementary information from different sensors. Existing learning-based fusion approaches attempt to construct various loss functions to preserve complementary features from both modalities, while neglecting to discover the inter-relationship between the two modalities, leading to redundant or even invalid information on the fusion results. To alleviate these issues, we propose a coupled contrastive learning network, dubbed CoCoNet, to realize infrared and visible image fusion in an end-to-end manner. Concretely, to simultaneously retain typical features from both modalities and remove unwanted information emerging on the fused result, we develop a coupled contrastive constraint in our loss function.In a fused imge, its foreground target/background detail part is pulled close to the infrared/visible source and pushed far away from the visible/infrared source in the representation space. We further exploit image characteristics to provide data-sensitive weights, which allows our loss function to build a more reliable relationship with source images. Furthermore, to learn rich hierarchical feature representation and comprehensively transfer features in the fusion process, a multi-level attention module is established. In addition, we also apply the proposed CoCoNet on medical image fusion of different types, e.g., magnetic resonance image and positron emission tomography image, magnetic resonance image and single photon emission computed tomography image. Extensive experiments demonstrate that our method achieves the state-of-the-art (SOTA) performance under both subjective and objective evaluation, especially in preserving prominent targets and recovering vital textural details. | ['Xin Fan', 'Zhongxuan Luo', 'Risheng Liu', 'Guanyao Wu', 'Runjia Lin', 'JinYuan Liu'] | 2022-11-20 | null | null | null | null | ['infrared-and-visible-image-fusion'] | ['computer-vision'] | [ 4.27562296e-01 -2.15598166e-01 -7.87467584e-02 -2.83641249e-01
-1.09769166e+00 -1.09018378e-01 3.10944676e-01 -6.31105155e-02
-3.14213246e-01 6.18156075e-01 4.21845555e-01 2.90815324e-01
-3.35141689e-01 -5.50361156e-01 -5.40987492e-01 -1.08470833e+00
4.12348717e-01 -2.13274315e-01 6.85233250e-02 -1.78489417e-01
-1.94147035e-01 3.33376557e-01 -1.17780662e+00 4.50330377e-01
1.04579878e+00 1.23249364e+00 4.73224968e-01 1.48012370e-01
-2.68559903e-02 8.72583270e-01 -2.00653389e-01 -1.83118328e-01
2.33768851e-01 -4.78109747e-01 -5.48024595e-01 3.62912327e-01
3.55763912e-01 -2.59635121e-01 -5.70261240e-01 1.38553715e+00
6.85708225e-01 1.48828298e-01 5.10428667e-01 -1.03994524e+00
-7.83262849e-01 2.96707094e-01 -1.16817462e+00 3.73223960e-01
2.99758613e-01 1.56632662e-01 5.53719938e-01 -1.06005776e+00
3.22831869e-01 1.18452609e+00 3.28244209e-01 3.82891953e-01
-1.01773500e+00 -7.91289210e-01 3.91916186e-01 1.13447651e-01
-1.25470233e+00 -5.30063152e-01 1.17329526e+00 -2.29009166e-01
2.38693833e-01 4.02381271e-01 4.06151026e-01 9.47819471e-01
4.35834229e-01 9.07969654e-01 1.29815066e+00 -3.02112639e-01
-3.04150194e-01 8.22910666e-02 -3.43831144e-02 9.05395687e-01
3.14185582e-02 2.04292178e-01 -4.66752082e-01 -8.18473920e-02
7.26958871e-01 6.91968918e-01 -9.23027158e-01 -1.91527978e-01
-1.28162205e+00 3.86973172e-01 9.30658519e-01 6.47309959e-01
-5.51486671e-01 -2.52112687e-01 1.09386332e-01 1.51895255e-01
5.26897609e-01 -8.53759274e-02 -2.45576605e-01 6.63514197e-01
-5.33541501e-01 -2.70142883e-01 -4.26155888e-02 4.78194267e-01
8.96238267e-01 -9.60073248e-02 -4.73207653e-01 8.85585248e-01
6.28521562e-01 4.65606987e-01 6.19420350e-01 -6.83380187e-01
6.51675701e-01 6.57683015e-01 -3.06291115e-02 -1.10663438e+00
-3.22231561e-01 -7.27190614e-01 -1.31673658e+00 4.11842853e-01
-7.40442947e-02 -1.10176384e-01 -1.10261405e+00 1.79847813e+00
5.11079192e-01 3.71162891e-01 6.70552105e-02 1.21178412e+00
1.04894781e+00 5.70328116e-01 9.83453169e-02 -6.31406903e-01
1.45970106e+00 -1.03069282e+00 -8.67526233e-01 -3.32589269e-01
1.45899147e-01 -8.03909600e-01 7.66806126e-01 2.91097850e-01
-1.18810856e+00 -7.77431846e-01 -9.92255449e-01 -1.60416931e-01
-4.08605188e-02 3.98514569e-02 4.35968399e-01 1.48606539e-01
-6.50208414e-01 2.98887044e-01 -8.12898040e-01 1.34004563e-01
5.24216354e-01 2.02932507e-01 -6.52697206e-01 -5.53697288e-01
-1.09003878e+00 8.27226639e-01 4.42712426e-01 2.42479175e-01
-6.35881126e-01 -6.90773845e-01 -9.63446081e-01 -1.44875247e-03
5.16462684e-01 -1.15003669e+00 6.32535696e-01 -1.00438881e+00
-1.22561598e+00 6.57626569e-01 -8.04834533e-03 4.31371816e-02
3.18192899e-01 -1.16858050e-01 -6.89676702e-01 3.00665379e-01
1.40283003e-01 6.01143181e-01 9.43538725e-01 -1.61886585e+00
-7.35687256e-01 -7.23805428e-01 -1.74104020e-01 5.87758183e-01
-3.25759947e-01 -2.36889925e-02 -5.77098191e-01 -8.54846716e-01
3.69678140e-01 -4.87655491e-01 -7.02176020e-02 1.66121468e-01
-4.09846961e-01 3.29882465e-02 9.34196174e-01 -8.60525310e-01
9.73953784e-01 -2.23650312e+00 4.84783381e-01 4.84478399e-02
5.36022663e-01 1.22615457e-01 -1.77012175e-01 -1.81063280e-01
-2.42068216e-01 -1.87206164e-01 -4.84395355e-01 -4.13768142e-01
-4.86267984e-01 1.20540090e-01 -2.59005912e-02 6.49469376e-01
2.92410776e-02 9.87432063e-01 -8.51477563e-01 -7.21552134e-01
5.73308587e-01 8.20324183e-01 -5.99430166e-02 2.19648093e-01
3.75568837e-01 1.01906765e+00 -7.30147421e-01 8.65906954e-01
8.48736703e-01 -2.64313936e-01 -4.09419090e-01 -6.86863482e-01
1.57369792e-01 -3.42225641e-01 -9.04829979e-01 2.10960007e+00
-6.08629704e-01 -4.29809578e-02 5.46472192e-01 -1.04312646e+00
6.98684931e-01 3.14797908e-01 7.78082669e-01 -1.03163624e+00
2.33210340e-01 1.21553838e-01 -3.02519679e-01 -6.68627143e-01
-2.79231463e-02 -5.43782949e-01 1.02936208e-01 -4.83756587e-02
8.63030925e-02 2.39159197e-01 -3.77383232e-01 -1.09056368e-01
6.44156814e-01 -1.26144178e-02 4.97992337e-02 1.48539335e-01
9.59348619e-01 -5.58521450e-01 7.49992609e-01 4.47578698e-01
-7.29217306e-02 8.93919230e-01 -2.23414153e-01 -6.43677413e-02
-5.27785420e-01 -1.06801152e+00 -1.07086800e-01 8.53488147e-01
6.61340356e-01 1.41448781e-01 -3.06531638e-01 -7.10092485e-01
-2.65393794e-01 3.16756546e-01 -5.55968463e-01 -4.73700404e-01
-4.93183434e-01 -7.94835687e-01 -2.77291462e-02 4.06579137e-01
7.48219788e-01 -8.75365794e-01 -2.80005127e-01 2.31903672e-01
-5.13683081e-01 -1.02067769e+00 -7.05124319e-01 -5.04364781e-02
-8.44679654e-01 -9.95780885e-01 -1.03672791e+00 -6.54026985e-01
5.99024117e-01 7.14438319e-01 8.03799808e-01 8.02241489e-02
-3.45780164e-01 3.94315302e-01 -3.01075131e-01 -2.14388534e-01
-1.14795946e-01 -4.45104361e-01 -2.51737922e-01 4.77178812e-01
-1.54576942e-01 -6.33252442e-01 -7.34138846e-01 1.43077865e-01
-1.26810753e+00 1.82940587e-01 7.59460330e-01 9.96649802e-01
6.92485511e-01 2.87978083e-01 3.94256979e-01 -3.64579886e-01
3.15386534e-01 -5.34727573e-01 -2.89388150e-01 5.43279231e-01
-3.29801351e-01 -7.57717416e-02 4.53772575e-01 -2.32264057e-01
-1.35758460e+00 1.66637436e-01 -1.66386604e-01 -8.89608145e-01
-3.59650664e-02 4.86494333e-01 -4.30828899e-01 -2.04069510e-01
2.85312980e-01 4.80320394e-01 8.65959823e-02 -5.41601777e-01
2.56030589e-01 5.09695590e-01 8.52115750e-01 -3.70876610e-01
7.87858367e-01 7.85365701e-01 -8.55419189e-02 -3.70782316e-01
-1.03110981e+00 -5.23635447e-01 -3.02160800e-01 -1.42411888e-01
9.56548035e-01 -1.20158565e+00 -5.87605715e-01 4.79383439e-01
-9.31073308e-01 3.93911511e-01 -9.11594182e-02 7.54791319e-01
-3.29366982e-01 5.69189668e-01 -5.25339246e-01 -5.78189492e-01
-6.59907401e-01 -1.28973711e+00 1.43769550e+00 5.97699583e-01
5.68814993e-01 -8.78937662e-01 -1.54169902e-01 6.26139104e-01
2.66384900e-01 5.26992559e-01 7.47516990e-01 -5.78249171e-02
-4.57805395e-01 -7.40609765e-02 -5.39508760e-01 5.40847838e-01
5.15028894e-01 -4.36079264e-01 -9.70035672e-01 -4.42877203e-01
4.39739525e-01 -2.59372294e-01 1.29464221e+00 4.76911962e-01
1.21598494e+00 -2.06216753e-01 -5.10326564e-01 7.32373774e-01
1.46830511e+00 1.62497893e-01 5.87112963e-01 1.61429681e-02
1.02007401e+00 5.92445552e-01 6.09517813e-01 8.78873318e-02
1.81574211e-01 6.63073659e-01 4.65469718e-01 -6.47863626e-01
-3.32382798e-01 -2.79418211e-02 3.60667318e-01 6.61431193e-01
5.38119068e-03 -1.02895536e-01 -4.56935853e-01 3.53434950e-01
-1.91642237e+00 -8.84751379e-01 1.61773160e-01 2.23340130e+00
6.97584093e-01 -1.62777707e-01 -1.25691980e-01 -1.02866322e-01
8.32607448e-01 1.36276394e-01 -6.36178195e-01 6.18493259e-01
-3.09203297e-01 -2.49055512e-02 2.22086221e-01 3.87373298e-01
-1.16425383e+00 2.60531992e-01 5.10130262e+00 1.08913994e+00
-1.32525337e+00 5.10137856e-01 8.28850091e-01 -1.25723287e-01
-5.84322214e-01 -2.32089490e-01 -2.01088563e-01 6.18070364e-01
2.20839068e-01 2.74035297e-02 2.88556099e-01 2.40719885e-01
9.15552378e-02 -1.21763460e-01 -7.10817099e-01 1.35513318e+00
2.19098061e-01 -1.18969631e+00 -2.23059021e-02 2.43127681e-02
6.21755540e-01 2.20619347e-02 2.09094390e-01 1.05621539e-01
-1.77472085e-02 -9.94922638e-01 6.24147356e-01 9.37678874e-01
6.72708273e-01 -7.82156825e-01 7.00446367e-01 3.38153630e-01
-1.30785632e+00 -3.05601209e-01 -2.36286446e-01 4.87029195e-01
2.31953293e-01 6.98642731e-01 9.78516787e-02 1.25834191e+00
7.38384247e-01 8.76363814e-01 -5.06549895e-01 9.30883706e-01
4.88408022e-02 -8.97910073e-02 -1.75129741e-01 8.60877156e-01
1.80450037e-01 -2.09686771e-01 6.22219026e-01 8.06583583e-01
3.76301646e-01 3.83646756e-01 5.22187889e-01 9.07406330e-01
-7.51933604e-02 -5.65352216e-02 -5.38442969e-01 3.91457349e-01
2.61627853e-01 1.48728573e+00 -4.80669051e-01 -3.12457234e-01
-5.84914029e-01 1.06068361e+00 9.89625528e-02 5.40440500e-01
-8.45690787e-01 -7.47468174e-02 3.65879208e-01 -7.62431100e-02
1.55265793e-01 2.70127535e-01 -7.59282187e-02 -1.44881260e+00
1.43550545e-01 -7.36544073e-01 5.73503435e-01 -9.67015743e-01
-1.50832164e+00 7.23116815e-01 2.41785068e-02 -1.53523028e+00
3.36444557e-01 -2.44884998e-01 -6.76550150e-01 1.10490227e+00
-1.82773745e+00 -1.38663197e+00 -6.25740111e-01 1.07942128e+00
3.28626007e-01 -1.19812503e-01 2.92700171e-01 6.26128733e-01
-5.61578572e-01 5.12173593e-01 1.29045978e-01 -1.00865781e-01
7.25830913e-01 -8.20742428e-01 -4.90194768e-01 8.13609123e-01
-1.01030998e-01 4.17560995e-01 3.12236398e-01 -5.18813550e-01
-1.31938231e+00 -1.15901792e+00 1.00084037e-01 -1.37673825e-01
2.59439111e-01 2.20078051e-01 -1.08263767e+00 3.74695420e-01
3.02861750e-01 5.65256238e-01 4.23602313e-01 -3.18197221e-01
-3.21587265e-01 -3.84604096e-01 -1.29358315e+00 2.77604520e-01
8.17365527e-01 -5.66447735e-01 -5.42533457e-01 3.91036451e-01
8.67424548e-01 -4.74014670e-01 -1.02204251e+00 8.18369150e-01
3.60517442e-01 -1.04130292e+00 1.22330713e+00 -2.42449537e-01
3.74723107e-01 -5.79961896e-01 -1.21133678e-01 -1.19848084e+00
-4.68895823e-01 -4.08017606e-01 -4.88496982e-02 1.32018554e+00
-1.35426773e-02 -6.43294871e-01 3.26741844e-01 3.21458787e-01
-4.02048975e-01 -8.48188579e-01 -1.12376344e+00 -3.80423516e-01
-7.23073184e-02 -6.39694706e-02 3.42967957e-01 1.06522739e+00
-3.43052596e-01 4.67647552e-01 -5.32472432e-01 3.00577492e-01
8.34679842e-01 2.10353836e-01 2.14788839e-01 -1.04148316e+00
-2.75027186e-01 -4.91313934e-01 -1.79888129e-01 -7.05063879e-01
6.87736496e-02 -9.29938376e-01 -3.25739831e-02 -1.62918389e+00
5.91237426e-01 -3.14030617e-01 -8.73137712e-01 5.58110714e-01
-5.98436058e-01 4.67178881e-01 2.47062281e-01 2.95994967e-01
-5.44926822e-01 8.64118874e-01 1.75985694e+00 -4.68803197e-01
5.61440960e-02 -2.10578591e-01 -1.13233232e+00 7.82816887e-01
4.46421057e-01 -4.41153020e-01 -4.61544931e-01 -5.40117264e-01
-2.32498616e-01 5.55689275e-01 6.36978269e-01 -9.63738382e-01
2.91130930e-01 -1.26450643e-01 7.16845989e-01 -4.21073943e-01
4.88937914e-01 -1.07918441e+00 1.32345587e-01 2.66083390e-01
-6.53399527e-02 -3.58002514e-01 3.46113108e-02 8.33573341e-01
-5.59129119e-01 1.17075061e-02 9.99299169e-01 -2.23162189e-01
-5.06196499e-01 6.27198935e-01 2.45534688e-01 -2.92037457e-01
9.82304454e-01 -2.37977698e-01 -3.74091089e-01 -1.14303812e-01
-7.82880962e-01 3.66878659e-01 3.44903976e-01 4.86094415e-01
8.75166297e-01 -1.51830149e+00 -8.45349669e-01 2.26637021e-01
2.57771939e-01 4.20498513e-02 8.48993301e-01 1.35164642e+00
1.22937955e-01 3.88782881e-02 -3.17082971e-01 -6.33864105e-01
-1.25423837e+00 7.79090941e-01 5.29177845e-01 -3.46311539e-01
-8.52993846e-01 8.27350259e-01 7.46776521e-01 -2.32830703e-01
1.59619629e-01 -2.66439795e-01 -3.82445782e-01 -9.06716287e-02
7.71633863e-01 7.38140121e-02 9.59095061e-02 -9.17719960e-01
-4.78342384e-01 8.60221088e-01 -1.51450053e-01 1.12538062e-01
1.22023988e+00 -5.44253051e-01 -1.95468754e-01 1.07117228e-01
1.46907449e+00 -2.86557190e-02 -1.26394355e+00 -6.85450196e-01
-6.67177498e-01 -5.33965290e-01 6.53130233e-01 -8.06573749e-01
-1.54505193e+00 8.39778662e-01 1.09785354e+00 -4.68215272e-02
1.73520732e+00 2.85076275e-02 1.02037215e+00 -9.89001244e-02
1.28362730e-01 -6.09569192e-01 1.84285089e-01 -1.14660367e-01
1.01710463e+00 -1.56699967e+00 1.70882776e-01 -4.35034782e-01
-6.93751514e-01 8.99818420e-01 5.19011199e-01 1.64343953e-01
5.67950428e-01 1.84273764e-01 -1.29973099e-01 -2.92680472e-01
-4.22133058e-01 -1.92365021e-01 6.30963683e-01 5.46307504e-01
2.43575752e-01 -1.21778868e-01 -7.34717958e-03 5.60885847e-01
6.85741961e-01 -5.41614294e-02 1.85201112e-02 8.53264868e-01
-5.02302468e-01 -7.74375081e-01 -7.41946042e-01 3.47261637e-01
-4.68168050e-01 -1.41654447e-01 -9.04684328e-03 4.45529878e-01
3.48585963e-01 1.02744269e+00 -1.05052747e-01 -4.20810133e-01
2.33575612e-01 -2.70282924e-01 6.69635296e-01 -3.21613491e-01
-3.93403828e-01 6.34188175e-01 -3.77368689e-01 -5.36742449e-01
-8.55177402e-01 -2.55871475e-01 -1.19323206e+00 6.38846755e-02
-4.93384153e-01 -2.78996564e-02 2.61597961e-01 1.02376604e+00
2.52444863e-01 8.95091236e-01 8.72448564e-01 -9.81848359e-01
-2.54239351e-01 -8.85120392e-01 -7.09976137e-01 5.31757057e-01
7.16139019e-01 -8.31001282e-01 -1.18077099e-01 -1.36309803e-01] | [10.535764694213867, -1.8855385780334473] |
97eeb369-1881-442f-bf5f-c177d1cb6dd2 | extractive-is-not-faithful-an-investigation | 2209.03549 | null | https://arxiv.org/abs/2209.03549v2 | https://arxiv.org/pdf/2209.03549v2.pdf | Extractive is not Faithful: An Investigation of Broad Unfaithfulness Problems in Extractive Summarization | The problems of unfaithful summaries have been widely discussed under the context of abstractive summarization. Though extractive summarization is less prone to the common unfaithfulness issues of abstractive summaries, does that mean extractive is equal to faithful? Turns out that the answer is no. In this work, we define a typology with five types of broad unfaithfulness problems (including and beyond not-entailment) that can appear in extractive summaries, including incorrect coreference, incomplete coreference, incorrect discourse, incomplete discourse, as well as other misleading information. We ask humans to label these problems out of 1600 English summaries produced by 16 diverse extractive systems. We find that 30% of the summaries have at least one of the five issues. To automatically detect these problems, we find that 5 existing faithfulness evaluation metrics for summarization have poor correlations with human judgment. To remedy this, we propose a new metric, ExtEval, that is designed for detecting unfaithful extractive summaries and is shown to have the best performance. We hope our work can increase the awareness of unfaithfulness problems in extractive summarization and help future work to evaluate and resolve these issues. Our data and code are publicly available at https://github.com/ZhangShiyue/extractive_is_not_faithful | ['Mohit Bansal', 'David Wan', 'Shiyue Zhang'] | 2022-09-08 | null | null | null | null | ['extractive-summarization'] | ['natural-language-processing'] | [ 2.31957987e-01 4.96404916e-01 -4.68209296e-01 -2.18306765e-01
-1.17491281e+00 -9.30847585e-01 7.51036406e-01 6.48584962e-01
-1.69785485e-01 1.02953577e+00 1.29833388e+00 -2.53664941e-01
-2.28627041e-01 -4.36623484e-01 -2.74659514e-01 -2.35475346e-01
5.53717017e-01 5.30654490e-01 2.32292444e-01 -4.76104468e-01
1.13633966e+00 1.82722777e-01 -1.24760497e+00 4.81109619e-01
1.33509779e+00 1.68772757e-01 -1.57503098e-01 8.64137053e-01
-2.07388066e-02 1.15353870e+00 -1.14703238e+00 -5.07201135e-01
-3.39749575e-01 -7.77590513e-01 -1.52111161e+00 -5.28158583e-02
9.42585766e-01 -4.18354511e-01 -1.54313669e-01 1.14480400e+00
5.79368293e-01 9.62622762e-02 8.03131878e-01 -1.09924638e+00
-4.11088943e-01 9.53930438e-01 -5.27346790e-01 7.78074801e-01
9.55393791e-01 -5.50121330e-02 1.16407871e+00 -6.12023592e-01
6.19780540e-01 1.39768016e+00 6.35210335e-01 4.70391572e-01
-7.94257164e-01 -4.35009152e-01 -1.95943713e-01 2.52161413e-01
-7.85145819e-01 -9.46887195e-01 6.44521236e-01 -3.00098270e-01
1.12355590e+00 9.31775451e-01 5.11716008e-01 1.07140374e+00
3.89112204e-01 8.70388687e-01 6.62574828e-01 -4.30802107e-01
-5.71080968e-02 -1.71915323e-01 9.04165745e-01 4.09654349e-01
8.84600043e-01 -4.25750405e-01 -7.55002677e-01 -4.16951984e-01
8.40595737e-02 -4.15353656e-01 -7.07095325e-01 4.97321963e-01
-1.30198717e+00 9.54047620e-01 -3.24422717e-02 5.20415604e-01
-4.03223634e-01 -8.31663162e-02 6.29878998e-01 1.73628509e-01
3.86743009e-01 1.12177861e+00 -6.08649664e-02 -6.37247443e-01
-1.31753099e+00 5.11489630e-01 1.30750406e+00 9.44170296e-01
4.46285456e-01 2.23603398e-02 -4.35055315e-01 7.05316722e-01
-9.46652442e-02 3.93306077e-01 6.70977354e-01 -1.54176533e+00
6.44751251e-01 6.41164660e-01 2.34074295e-01 -1.36155987e+00
-5.05352139e-01 -1.63111374e-01 -7.56401062e-01 -4.55050439e-01
2.60586560e-01 -9.58267599e-02 -2.24885941e-01 1.43382716e+00
-1.92289844e-01 -5.63907921e-01 1.72747895e-01 7.83136964e-01
1.37927723e+00 7.89409459e-01 -2.93271303e-01 -7.35910058e-01
1.56143165e+00 -9.93752420e-01 -1.33378577e+00 -4.44318116e-01
5.65405190e-01 -1.15126967e+00 9.78974342e-01 3.56996655e-01
-1.54411232e+00 -5.22098690e-02 -1.33886826e+00 -3.78423035e-01
6.56294301e-02 -8.67346078e-02 3.40958357e-01 3.46151024e-01
-8.85602593e-01 7.61312842e-01 -5.75859725e-01 -7.07991362e-01
1.38291959e-02 -9.18449089e-02 -2.68412560e-01 1.66907236e-01
-1.11677432e+00 1.34343803e+00 6.29502237e-01 -2.99485952e-01
-1.86084136e-01 -4.97292668e-01 -8.03070366e-01 6.35932013e-02
7.18063533e-01 -7.02723563e-01 1.63220954e+00 -6.69819117e-01
-9.68310118e-01 7.17837632e-01 -2.61740476e-01 -3.83214980e-01
1.64480999e-01 -5.39454579e-01 -3.06895941e-01 4.09883618e-01
4.47639346e-01 6.57810792e-02 3.27571481e-01 -9.99545395e-01
-3.63584042e-01 -1.28426507e-01 -1.08271673e-01 3.02373022e-01
-2.79170543e-01 3.59755576e-01 6.97301992e-04 -7.35241234e-01
1.78142950e-01 -6.03683949e-01 3.99807453e-01 -7.44239390e-01
-9.15951848e-01 -5.84008574e-01 5.80516338e-01 -8.26321065e-01
1.94150984e+00 -1.68426526e+00 3.93958502e-02 -3.32213014e-01
4.05436009e-01 4.25444901e-01 1.00047931e-01 9.74460602e-01
9.21697915e-02 5.83076298e-01 -3.14870030e-01 1.66866228e-01
1.44575417e-01 1.07451700e-01 -4.83465672e-01 2.86278397e-01
3.84720229e-02 7.01778710e-01 -1.20723343e+00 -8.78591180e-01
-1.19746260e-01 -6.47599623e-02 -3.61710936e-01 1.72024131e-01
6.96908012e-02 -9.79304165e-02 -4.16028023e-01 4.30741787e-01
4.63710576e-01 -2.13913620e-01 3.60032693e-02 -4.09403086e-01
-4.08762038e-01 9.51412320e-01 -8.04847002e-01 1.35904872e+00
3.03153187e-01 8.51423502e-01 -2.34699816e-01 -5.56358337e-01
5.99035859e-01 3.66506696e-01 -7.87787810e-02 -2.86417842e-01
1.33403823e-01 4.07717079e-01 8.55895281e-02 -6.72650158e-01
1.37293828e+00 -2.43129790e-01 -4.58829731e-01 8.39642763e-01
6.56330213e-02 -6.62803650e-01 6.65823698e-01 9.56671238e-01
1.18665504e+00 -5.70641875e-01 1.03189349e+00 -5.46312928e-01
3.66280854e-01 3.42300564e-01 6.25729501e-01 9.99739170e-01
-4.09019411e-01 8.81780148e-01 8.78817201e-01 -5.08842543e-02
-8.10406685e-01 -8.37327123e-01 1.04317307e-01 8.56208146e-01
3.43622267e-02 -9.85199094e-01 -8.20532322e-01 -5.02407491e-01
-2.73213983e-01 1.41526103e+00 -3.63021582e-01 -3.37187082e-01
-7.07273185e-01 -4.31923240e-01 8.82003129e-01 3.10169548e-01
4.74417448e-01 -1.08237720e+00 -9.68014836e-01 5.39550856e-02
-1.04102147e+00 -8.33372056e-01 -8.40491295e-01 -1.03174277e-01
-7.46403098e-01 -1.35269928e+00 -3.77417117e-01 -3.33598852e-01
1.78429320e-01 3.55031103e-01 1.38535929e+00 3.99744600e-01
2.49414086e-01 3.48656893e-01 -6.02257848e-01 -5.01110673e-01
-8.50616336e-01 1.49388313e-01 -3.03891990e-02 -7.85088897e-01
4.54027534e-01 -3.00972819e-01 -4.86498654e-01 -9.02533531e-02
-1.04291296e+00 -7.95232952e-02 3.53383213e-01 5.94089806e-01
-9.27535817e-02 -2.33050972e-01 7.25738049e-01 -1.07893765e+00
1.43687165e+00 -4.26840812e-01 1.95417047e-01 3.08397353e-01
-5.91712356e-01 9.77851897e-02 4.07285988e-01 -4.07521389e-02
-1.02178252e+00 -9.18913245e-01 -2.01550752e-01 2.16106340e-01
-3.43558714e-02 6.99597359e-01 1.48702651e-01 4.95502025e-01
9.76270616e-01 1.19385906e-02 -1.14402972e-01 -2.05576837e-01
6.49694428e-02 7.97969520e-01 7.73460686e-01 -5.57627857e-01
4.22034174e-01 1.30022347e-01 -4.89464611e-01 -1.08504379e+00
-1.43060231e+00 -6.92923009e-01 -1.93960115e-01 -2.21131623e-01
6.08929753e-01 -5.85624397e-01 -3.32464159e-01 1.18027985e-01
-1.50837493e+00 2.07561091e-01 -4.16749090e-01 3.65814924e-01
-5.75428188e-01 1.20227456e+00 -8.48841429e-01 -7.27443397e-01
-9.30049121e-01 -7.25126743e-01 7.04743862e-01 4.71341759e-01
-1.32298172e+00 -7.87459970e-01 3.34834635e-01 6.03850424e-01
4.18074131e-01 4.36224520e-01 9.41991985e-01 -1.10220730e+00
1.97747752e-01 -4.53529768e-02 -3.63535457e-03 2.29342103e-01
2.33604535e-01 3.19052190e-01 -4.74386066e-01 -1.23695239e-01
4.40116107e-01 -5.58455527e-01 9.68803406e-01 3.49189162e-01
4.70824718e-01 -1.06934929e+00 -8.07019603e-03 -7.19416216e-02
1.05539799e+00 -1.44831911e-01 7.62958705e-01 3.86369318e-01
2.93735296e-01 6.97958052e-01 8.35816085e-01 6.49058342e-01
4.36092645e-01 3.27623039e-01 1.32701740e-01 4.60802794e-01
-2.77872235e-01 -3.02947134e-01 4.28164989e-01 1.33904290e+00
-1.65106133e-01 -6.72970235e-01 -8.84458780e-01 6.62664890e-01
-1.83052564e+00 -1.45013249e+00 -5.43100417e-01 1.86860502e+00
1.08871102e+00 1.88064054e-01 1.61652759e-01 3.32964808e-01
7.88690507e-01 4.98321921e-01 -7.31706917e-02 -9.37922001e-01
-1.93774521e-01 -1.06555238e-01 5.41894175e-02 7.47957528e-01
-7.69859850e-01 8.58594596e-01 6.26231003e+00 5.55203617e-01
-7.10291743e-01 -1.22673824e-01 7.56044313e-02 -3.49026844e-02
-6.12937629e-01 1.82465643e-01 -5.53913772e-01 3.28172535e-01
8.70676398e-01 -7.18131065e-01 -1.77726343e-01 3.46180022e-01
2.83048302e-01 -5.01400828e-01 -1.02531779e+00 5.34052074e-01
5.72044134e-01 -1.31656015e+00 2.87163585e-01 -2.94203907e-01
5.66209495e-01 -2.84112275e-01 -3.36439669e-01 9.57814083e-02
1.64522007e-01 -7.12068856e-01 7.53576636e-01 5.15412748e-01
3.30175847e-01 -7.17521548e-01 1.01213598e+00 6.11559093e-01
-4.72995639e-01 3.29466015e-01 -3.82314205e-01 -2.26540804e-01
2.10145459e-01 5.89802384e-01 -7.55586684e-01 7.23442495e-01
2.99473673e-01 7.99323022e-01 -6.38952434e-01 9.76637840e-01
-5.08793116e-01 7.15513110e-01 -5.27051324e-03 -3.08525562e-01
1.57446593e-01 5.01599237e-02 1.17015660e+00 1.47052479e+00
4.16112304e-01 5.17022789e-01 -2.68426180e-01 7.39579439e-01
-1.18546747e-02 5.81509918e-02 -6.51834607e-01 -3.20251256e-01
7.89372265e-01 9.74686563e-01 -6.71553910e-01 -6.42262399e-01
4.47564907e-02 8.16925347e-01 1.20055988e-01 5.56940064e-02
-5.34407735e-01 -5.72333038e-01 -1.39370486e-02 1.10081919e-01
-2.19877794e-01 1.34369954e-01 -3.61339211e-01 -1.32852483e+00
-8.00021142e-02 -1.26825523e+00 7.45831490e-01 -8.61341178e-01
-1.23358262e+00 4.33320016e-01 3.34680706e-01 -9.60169792e-01
-4.14161175e-01 5.51106744e-02 -1.06765461e+00 4.17598575e-01
-1.16902161e+00 -6.01164937e-01 -3.13073814e-01 1.46646559e-01
9.16800499e-01 2.27559015e-01 7.38768816e-01 -1.70914903e-01
-5.10833859e-01 2.68921822e-01 -4.48205352e-01 -1.32271871e-01
1.09038877e+00 -1.28952265e+00 1.60810873e-01 1.06533611e+00
-1.87435508e-01 7.97496676e-01 1.47948992e+00 -7.60342717e-01
-1.10778093e+00 -6.97711170e-01 1.55329692e+00 -6.78792298e-01
5.75327814e-01 5.49574196e-01 -1.15663636e+00 7.48220265e-01
9.94601667e-01 -9.71884072e-01 8.27830911e-01 9.51631069e-02
-2.22895235e-01 4.09724921e-01 -8.77376080e-01 5.28140366e-01
7.69741476e-01 -1.99949101e-01 -1.80187237e+00 4.61154670e-01
5.61835587e-01 -2.88520694e-01 -7.01252759e-01 3.80679548e-01
2.89584458e-01 -1.26489496e+00 5.39152324e-01 -4.50765222e-01
8.33764851e-01 1.89636424e-02 3.79705988e-02 -1.40272462e+00
-4.05468583e-01 -8.20310473e-01 -2.36947164e-01 1.40090537e+00
3.90709639e-01 -5.89567840e-01 3.36630553e-01 5.82883894e-01
-6.76413953e-01 -1.77937999e-01 -7.39257395e-01 -6.42755687e-01
3.44826669e-01 9.75996703e-02 3.21278214e-01 1.14208126e+00
9.45756972e-01 1.00590968e+00 -2.88658261e-01 -3.91004622e-01
5.02471924e-01 1.56839982e-01 6.31085396e-01 -1.18673229e+00
2.22946450e-01 -6.41615629e-01 6.24444559e-02 -8.36691618e-01
1.84946775e-01 -7.07704842e-01 1.46836117e-01 -2.11960721e+00
6.49208188e-01 3.86264592e-01 3.60780746e-01 3.72884035e-01
-3.53813827e-01 -9.89135578e-02 2.65952677e-01 4.58866507e-01
-9.06079113e-01 4.25782681e-01 1.26542425e+00 -1.61924288e-01
-1.09267808e-01 -2.20724404e-01 -1.35471225e+00 1.00716352e+00
9.24046993e-01 -5.10135949e-01 -1.32893696e-01 -2.56727606e-01
3.68839383e-01 3.16602498e-01 1.15265027e-01 -9.47817981e-01
5.28182685e-01 -1.74047545e-01 -1.48236111e-01 -9.69538748e-01
-8.52518156e-03 -1.81480721e-01 -3.54395583e-02 5.89508772e-01
-5.38042486e-01 3.67134243e-01 1.52368188e-01 1.94888309e-01
-4.05402869e-01 -7.39759743e-01 5.46238363e-01 -3.04839194e-01
-4.20487344e-01 -5.49951255e-01 -7.50294924e-01 7.95630753e-01
5.49712598e-01 -1.10645011e-01 -1.13039422e+00 -7.32102513e-01
-1.57860845e-01 2.99008936e-01 5.89141548e-01 2.32679173e-01
7.93099761e-01 -9.68457878e-01 -1.13243508e+00 -5.15369713e-01
8.34553987e-02 -3.01375300e-01 2.28182733e-01 1.00973117e+00
-6.12953305e-01 6.18463874e-01 -1.68419138e-01 -2.45772585e-01
-1.38060796e+00 1.49293393e-01 -5.92017584e-02 -3.25246334e-01
-5.73242784e-01 4.24802810e-01 -2.80158788e-01 -1.63672507e-01
1.34375757e-02 -4.04274911e-01 -4.94639248e-01 4.69276100e-01
6.85432196e-01 9.28547025e-01 8.11100379e-02 -7.89411485e-01
-4.73728061e-01 2.77083725e-01 -2.83439100e-01 -1.07594080e-01
9.93471503e-01 -1.58409044e-01 -5.17164528e-01 5.85610807e-01
8.87168646e-01 3.86869311e-01 -3.94413769e-01 1.98361784e-01
2.85067022e-01 -2.13550970e-01 -1.55273899e-01 -8.04567575e-01
-2.44192347e-01 6.02971733e-01 -4.57011372e-01 7.49091685e-01
8.44537973e-01 1.77637652e-01 9.12343502e-01 5.63628852e-01
-1.45770656e-02 -1.27384377e+00 1.87251806e-01 8.25191081e-01
1.45710111e+00 -1.00250006e+00 6.39963627e-01 -4.41385090e-01
-8.37237895e-01 1.28890383e+00 4.82541442e-01 -1.32805929e-01
2.97120027e-02 5.90705685e-02 -8.57273117e-02 -6.42720878e-01
-9.50274467e-01 1.58208951e-01 3.05837095e-01 1.61343962e-01
7.64935970e-01 3.55910100e-02 -1.08334589e+00 6.78894699e-01
-7.38098800e-01 -2.36839026e-01 1.27351403e+00 9.38345551e-01
-9.17520583e-01 -6.45797968e-01 -5.37473917e-01 7.39493549e-01
-6.42265022e-01 -1.08330727e-01 -1.13769007e+00 9.46826160e-01
-6.22600794e-01 1.52180910e+00 -7.96299502e-02 -2.01994881e-01
5.46173096e-01 2.47615110e-02 5.03613710e-01 -7.59518087e-01
-9.63299870e-01 1.84509847e-02 8.44141722e-01 -3.73650134e-01
-7.86793530e-01 -7.15570390e-01 -1.22482049e+00 -8.20207477e-01
-4.59255755e-01 7.02931523e-01 -5.11374958e-02 9.47980404e-01
1.53136715e-01 4.31415975e-01 1.33993611e-01 -5.91448784e-01
-9.19831872e-01 -1.33317626e+00 -3.03246140e-01 4.26508874e-01
5.11489451e-01 -3.39442819e-01 -8.86396766e-01 -1.27435923e-01] | [12.264628410339355, 9.42177963256836] |
179e5ecb-af3c-41ab-81e8-9a67a1d72626 | hin-hierarchical-inference-network-for | 2003.12754 | null | https://arxiv.org/abs/2003.12754v1 | https://arxiv.org/pdf/2003.12754v1.pdf | HIN: Hierarchical Inference Network for Document-Level Relation Extraction | Document-level RE requires reading, inferring and aggregating over multiple sentences. From our point of view, it is necessary for document-level RE to take advantage of multi-granularity inference information: entity level, sentence level and document level. Thus, how to obtain and aggregate the inference information with different granularity is challenging for document-level RE, which has not been considered by previous work. In this paper, we propose a Hierarchical Inference Network (HIN) to make full use of the abundant information from entity level, sentence level and document level. Translation constraint and bilinear transformation are applied to target entity pair in multiple subspaces to get entity-level inference information. Next, we model the inference between entity-level information and sentence representation to achieve sentence-level inference information. Finally, a hierarchical aggregation approach is adopted to obtain the document-level inference information. In this way, our model can effectively aggregate inference information from these three different granularities. Experimental results show that our method achieves state-of-the-art performance on the large-scale DocRED dataset. We also demonstrate that using BERT representations can further substantially boost the performance. | ['Zhen-Yu Zhang', 'Pengfei Yin', 'Fang Fang', 'Yanan Cao', 'Shi Wang', 'Hengzhu Tang', 'Jiangxia Cao'] | 2020-03-28 | null | null | null | null | ['document-level-relation-extraction'] | ['natural-language-processing'] | [ 4.44715284e-02 -1.87848181e-01 -3.26293446e-02 -4.70547110e-01
-1.28764498e+00 -5.80503881e-01 5.72938085e-01 2.14102864e-01
-1.75782263e-01 8.60867500e-01 8.19621444e-01 -2.30232537e-01
-3.10448080e-01 -1.00014460e+00 -6.27145588e-01 -4.28122073e-01
3.61842573e-01 3.81086916e-01 -1.00237485e-02 -1.36621118e-01
2.17530400e-01 3.08676273e-01 -1.26449072e+00 7.07772374e-01
1.29386854e+00 9.37114596e-01 2.54610837e-01 5.84086657e-01
-4.24370199e-01 8.10051382e-01 -5.87081254e-01 -5.49609125e-01
3.15919891e-03 -3.56054574e-01 -8.89620721e-01 1.30654186e-01
5.77571869e-01 -5.19798040e-01 -9.11785872e-04 1.19154179e+00
4.89499331e-01 2.53120482e-01 6.90548718e-01 -6.57364309e-01
-7.43196189e-01 7.77540147e-01 -5.64259589e-01 1.08358651e-01
5.05204499e-01 -4.27161515e-01 1.31266713e+00 -1.36793065e+00
5.68481803e-01 1.52035534e+00 1.98268384e-01 -7.32300477e-03
-9.79677975e-01 -5.58471203e-01 4.66316730e-01 3.17672521e-01
-1.37824106e+00 -2.62982696e-01 7.44305432e-01 -1.28708199e-01
8.41845572e-01 4.53090668e-01 2.01334730e-01 8.25400233e-01
3.99593003e-02 1.00638938e+00 1.26351690e+00 -2.90509611e-01
-1.26147926e-01 3.32516059e-02 4.99843061e-01 5.50957441e-01
1.64626837e-01 -7.54085720e-01 -4.78594720e-01 7.96878263e-02
4.75092709e-01 1.33027866e-01 -2.55017489e-01 2.80663401e-01
-1.29227841e+00 5.29625773e-01 3.72337699e-01 4.21734661e-01
-4.45128411e-01 -2.37836897e-01 4.00676101e-01 1.05415314e-01
6.74741626e-01 1.45708218e-01 -5.56958318e-01 1.01327784e-02
-1.04905128e+00 1.35901451e-01 8.20897281e-01 1.24486327e+00
8.04888010e-01 -3.21901947e-01 -6.93472505e-01 1.06780124e+00
2.91328818e-01 4.92310911e-01 1.52392015e-01 -9.06812727e-01
1.15564871e+00 7.03062475e-01 5.56475632e-02 -1.24471545e+00
-8.03788826e-02 -7.04418898e-01 -1.47525537e+00 -6.24616027e-01
-9.92208272e-02 -1.40558735e-01 -5.67126989e-01 1.59694672e+00
3.18839133e-01 2.59648830e-01 3.47133636e-01 6.77215993e-01
1.04077756e+00 9.15829182e-01 -2.16283694e-01 -4.62652981e-01
1.53980410e+00 -8.85384679e-01 -9.26974237e-01 1.11796945e-01
5.25282979e-01 -6.47756934e-01 8.18535328e-01 2.19424292e-01
-1.24093926e+00 -6.65212154e-01 -9.94957626e-01 -5.80816567e-01
-4.33501273e-01 7.39068270e-01 1.40199944e-01 2.46707350e-02
-6.80681467e-01 2.05938667e-01 -6.13284171e-01 -1.25142232e-01
2.66999530e-04 1.75584301e-01 -2.54527211e-01 -2.95240581e-01
-1.69495618e+00 9.09880161e-01 5.94065726e-01 5.71831405e-01
-2.78140724e-01 -5.94583452e-01 -7.69784868e-01 3.70800942e-01
5.85529566e-01 -9.83574867e-01 8.63632917e-01 3.10032647e-02
-1.33737934e+00 2.27154359e-01 -5.33379018e-01 -1.41295120e-01
2.57776469e-01 -3.25263500e-01 -4.02337313e-01 -4.21571871e-03
1.57983929e-01 2.09209710e-01 4.22962427e-01 -1.22747183e+00
-7.35787749e-01 -6.14471197e-01 4.75488365e-01 5.66150308e-01
-4.27196622e-01 4.30803783e-02 -7.41159499e-01 -6.13363564e-01
2.90598392e-01 -6.02442503e-01 1.15584671e-01 -6.89096749e-01
-7.15619206e-01 -5.94436288e-01 4.52999353e-01 -1.20782280e+00
1.70584559e+00 -1.84471035e+00 6.93531036e-01 -2.72714882e-03
3.18893224e-01 3.44068669e-02 1.12887412e-01 7.25875795e-01
3.40207458e-01 2.58119643e-01 -3.78404260e-01 -5.36893070e-01
3.27520877e-01 5.78492470e-02 -5.01522958e-01 -1.50611624e-01
3.80412750e-02 8.79038870e-01 -8.47030640e-01 -8.56750131e-01
1.76916152e-01 2.78555304e-01 -6.71225011e-01 1.81327805e-01
-1.01877093e-01 3.41734231e-01 -7.69464076e-01 4.42299545e-01
9.52685714e-01 -3.00141573e-01 3.13178360e-01 -8.88114333e-01
-7.46300304e-03 4.51140165e-01 -1.21674848e+00 1.80835295e+00
-8.30128849e-01 1.68203235e-01 -1.37631118e-01 -9.19792354e-01
8.71519804e-01 1.03967100e-01 1.40767589e-01 -3.45453203e-01
-2.31959894e-01 8.97785723e-02 -3.13745409e-01 -5.12228847e-01
8.23827922e-01 1.98611896e-02 -5.28744400e-01 3.09650093e-01
7.75164962e-02 -4.55296077e-02 4.25318956e-01 5.00936210e-01
7.25341797e-01 3.05179153e-02 3.60365987e-01 -1.10889047e-01
9.68828857e-01 -5.51051915e-01 7.02396333e-01 6.13400519e-01
5.02356350e-01 2.91974396e-01 4.97695088e-01 8.83906260e-02
-7.50922084e-01 -1.00949049e+00 -2.72307873e-01 9.49333072e-01
2.43728131e-01 -7.20552146e-01 -7.10744798e-01 -7.78760076e-01
-1.74930289e-01 8.93585324e-01 -3.20665330e-01 1.19805276e-01
-6.67328298e-01 -7.22351551e-01 2.65283942e-01 6.43539250e-01
1.12379992e+00 -5.07176340e-01 2.55941421e-01 2.18399316e-01
-8.78333330e-01 -1.50037396e+00 -5.65571070e-01 -4.46757138e-01
-7.01491892e-01 -5.73401451e-01 -5.71145058e-01 -4.86388922e-01
6.70783222e-01 4.07584310e-02 1.01207829e+00 -2.49946862e-01
1.58941165e-01 1.44935071e-01 -4.55929726e-01 4.21460066e-03
-1.37587175e-01 2.68655717e-01 1.00646906e-01 1.71550557e-01
2.26593316e-01 -5.76613724e-01 -4.59685028e-01 3.15569676e-02
-1.00358963e+00 2.86019236e-01 9.24012363e-01 8.23222637e-01
5.34874320e-01 2.52424866e-01 6.35030627e-01 -8.15013885e-01
9.49003398e-01 -6.06127024e-01 -2.26864904e-01 7.53930211e-01
-1.97781712e-01 2.57745117e-01 8.64321172e-01 -1.28821237e-02
-1.54098701e+00 -4.89246309e-01 -1.83038056e-01 -7.58380368e-02
9.35350358e-02 1.08055305e+00 -5.42788386e-01 7.55889297e-01
1.06712982e-01 6.00992441e-01 -4.97301906e-01 -6.81218624e-01
5.04955709e-01 1.24398899e+00 4.93148178e-01 -9.20516968e-01
6.76181853e-01 1.95358723e-01 6.41993061e-02 -6.70962811e-01
-1.57892430e+00 -3.48711044e-01 -8.74925673e-01 8.55088681e-02
9.30553079e-01 -1.07020426e+00 -5.50406337e-01 2.64290124e-01
-1.47229552e+00 4.65134233e-01 1.44328833e-01 4.70870495e-01
-1.70975953e-01 5.50921977e-01 -6.70775294e-01 -6.14277601e-01
-5.76245785e-01 -1.15623069e+00 1.44958413e+00 7.47867972e-02
2.24891901e-01 -1.06284761e+00 -1.78865463e-01 6.71332657e-01
1.66162729e-01 2.36663353e-02 9.31309760e-01 -4.93879229e-01
-8.73051584e-01 2.04282217e-02 -5.73955178e-01 5.86919665e-01
2.75091648e-01 -2.14847431e-01 -4.83416915e-01 -1.39193475e-01
4.78353165e-02 -9.56935361e-02 9.83105183e-01 -1.03130795e-01
1.52378428e+00 -5.98561406e-01 -6.00960888e-02 4.23350960e-01
1.28384972e+00 -4.33029830e-01 6.26715958e-01 -3.98498029e-02
9.29697514e-01 4.87706155e-01 8.53629947e-01 4.66595173e-01
1.06458330e+00 7.89750338e-01 -7.38764629e-02 2.10443959e-01
-8.95568877e-02 -1.81408525e-01 3.33195865e-01 1.62654912e+00
-2.02659041e-01 -2.99025685e-01 -6.04031503e-01 3.05437654e-01
-2.10523081e+00 -1.19029713e+00 -4.98521188e-03 1.87886810e+00
1.02404976e+00 -1.49375811e-01 -2.86976695e-01 -1.47615284e-01
7.54180491e-01 1.89126670e-01 -3.50176632e-01 -2.92470336e-01
-3.82955233e-03 -1.85888484e-01 4.38297428e-02 5.32328010e-01
-1.01669884e+00 8.06321084e-01 5.19664955e+00 1.14369977e+00
-5.35559356e-01 -4.98508161e-04 3.77307087e-01 1.58985808e-01
-6.36938214e-01 1.24753816e-02 -1.33407080e+00 7.26948440e-01
5.27882218e-01 -3.80885214e-01 4.44300771e-01 3.91960353e-01
1.17585972e-01 8.12329948e-02 -1.03309941e+00 8.91017973e-01
3.92924875e-01 -1.28434110e+00 4.47835475e-01 5.49116880e-02
8.18004429e-01 -4.04794514e-01 -2.50622958e-01 7.06473529e-01
4.36882019e-01 -7.19222009e-01 3.10205519e-01 1.05416119e+00
6.31226897e-01 -8.12147856e-01 9.19395566e-01 7.72859156e-01
-1.49729550e+00 8.74851793e-02 -5.01523912e-01 1.50759131e-01
3.06251407e-01 9.75882828e-01 -5.15209854e-01 1.35054147e+00
4.37188208e-01 9.40292895e-01 -4.82042283e-01 3.73270869e-01
-3.42573285e-01 3.27661484e-01 -3.01452368e-01 -1.31421790e-01
1.06733575e-01 -4.88188505e-01 4.22810435e-01 1.41213191e+00
5.39102912e-01 3.39440048e-01 3.99763137e-01 6.74639642e-01
-3.73199195e-01 5.30690886e-02 -2.41826609e-01 1.65985897e-01
7.92968512e-01 1.44930184e+00 -5.53808548e-02 -6.68284178e-01
-3.79980028e-01 1.17688143e+00 7.38728046e-01 4.63528991e-01
-8.99976075e-01 -5.13540566e-01 3.56846958e-01 -4.78550524e-01
3.70716453e-01 -5.37919044e-01 -7.30275586e-02 -1.86043882e+00
4.83727366e-01 -6.71734214e-01 4.62235898e-01 -7.28173196e-01
-1.51655424e+00 2.53207147e-01 2.44517013e-01 -1.32859755e+00
-3.80795777e-01 -2.96179444e-01 -3.44472259e-01 9.85504568e-01
-1.38561344e+00 -1.53832698e+00 -2.34354153e-01 3.76528025e-01
7.18932867e-01 1.60379529e-01 6.90381825e-01 3.24446589e-01
-7.45458543e-01 6.39271379e-01 2.79716522e-01 2.66030878e-01
5.90704560e-01 -1.39511037e+00 -3.03327404e-02 8.99406254e-01
1.06229834e-01 1.16891563e+00 3.16163272e-01 -6.59344852e-01
-1.60725355e+00 -1.27437305e+00 1.13852143e+00 -5.95658064e-01
6.75223589e-01 -4.39187735e-01 -8.09292555e-01 8.41466546e-01
2.44093761e-01 -2.35252246e-01 7.51463652e-01 4.97400343e-01
-3.31136525e-01 -4.73248333e-01 -9.66810703e-01 7.35569060e-01
1.05928183e+00 -7.56956160e-01 -1.03383756e+00 3.05663526e-01
9.20955479e-01 -6.46417737e-01 -1.54759645e+00 5.93830168e-01
4.33676869e-01 -5.25122643e-01 1.06943083e+00 -5.49448192e-01
8.74923527e-01 -4.28695560e-01 -5.07564723e-01 -1.43385839e+00
-3.38162869e-01 -6.92578927e-02 -4.69150841e-01 1.86698401e+00
3.74818683e-01 -6.35083497e-01 4.84227166e-02 4.00098801e-01
-1.67803243e-01 -9.59944189e-01 -7.42805004e-01 -6.66909218e-01
2.46016726e-01 -2.60423779e-01 9.65029538e-01 7.62755156e-01
1.03927396e-01 8.08240414e-01 -4.30054247e-01 4.06597525e-01
7.40426421e-01 7.02505410e-01 7.76240528e-01 -8.73665273e-01
-3.40747297e-01 -1.42858833e-01 -1.26462936e-01 -1.52216220e+00
3.78466308e-01 -1.17274606e+00 -1.62122265e-01 -2.17857528e+00
6.85331643e-01 -1.24996774e-01 -4.03485149e-01 9.23265070e-02
-5.65951765e-01 -1.25774622e-01 3.90584081e-01 2.40892574e-01
-9.09012616e-01 7.54208326e-01 1.53438067e+00 -1.67975381e-01
2.74601430e-01 -3.75162691e-01 -8.95798147e-01 6.56749845e-01
5.55426002e-01 -9.22922194e-02 -2.84505874e-01 -6.97119355e-01
3.71453226e-01 2.79325187e-01 2.18817398e-01 -7.35093832e-01
5.78887343e-01 -1.77675441e-01 4.35337037e-01 -1.14990687e+00
4.77578670e-01 -6.69323623e-01 -1.36012837e-01 -2.11978242e-01
-5.14817417e-01 -1.27968311e-01 -1.03193030e-01 5.88628650e-01
-5.18183768e-01 -7.67966583e-02 2.82227218e-01 -7.34439120e-02
-4.01066154e-01 3.98590386e-01 3.40686142e-02 1.67581707e-01
6.48683250e-01 2.88263261e-01 -4.39561456e-01 -2.40734965e-01
-7.03541696e-01 5.66487372e-01 -1.07597904e-02 2.83439308e-01
5.36566019e-01 -1.59379864e+00 -1.08056784e+00 -1.65632710e-01
1.16211781e-02 1.87556550e-01 5.78438580e-01 8.44865918e-01
1.28123611e-02 5.63822806e-01 2.59883672e-01 -5.78675330e-01
-1.11855578e+00 3.75509501e-01 -1.94113225e-01 -9.27078664e-01
-3.24403644e-01 4.74450558e-01 2.03891858e-01 -8.35970938e-01
-2.85546780e-01 -4.08075869e-01 -4.10954535e-01 2.67620891e-01
5.19065917e-01 5.64230978e-01 -7.52254650e-02 -5.77804685e-01
-2.81766295e-01 8.76282811e-01 -3.93638015e-01 -1.50772676e-01
1.17858744e+00 -5.13055384e-01 -6.49309814e-01 5.29775262e-01
1.43224955e+00 3.52684498e-01 -5.90574086e-01 -5.33934355e-01
-1.65990308e-01 -3.80502135e-01 9.09061357e-02 -7.53950834e-01
-6.17726088e-01 1.01232159e+00 -1.36748984e-01 2.16603160e-01
1.30434155e+00 3.91705818e-02 1.00593817e+00 6.61202431e-01
4.59018111e-01 -1.09234464e+00 -1.69883728e-01 7.66101480e-01
1.14917147e+00 -1.06351221e+00 2.98807383e-01 -6.54905081e-01
-4.91943955e-01 1.02931845e+00 5.75218678e-01 1.17832487e-02
4.63495016e-01 1.87717244e-01 -6.26342118e-01 3.17916274e-02
-7.54797876e-01 -2.84098029e-01 7.15249717e-01 -8.64190906e-02
4.97784406e-01 2.09322244e-01 -5.86964846e-01 8.66908312e-01
-1.80031955e-01 1.48299858e-02 3.32842499e-01 4.82439131e-01
-4.76582050e-01 -1.12036169e+00 -1.72235772e-01 6.75936401e-01
-4.68202710e-01 -4.34558123e-01 -2.61624157e-01 3.36777270e-01
1.14435833e-02 9.46845472e-01 -4.40668836e-02 -4.07652378e-01
2.01588079e-01 -5.31659611e-02 5.05733728e-01 -6.72457337e-01
-2.04405665e-01 -2.23746657e-01 5.32726109e-01 -2.66133249e-01
-3.75314415e-01 -6.34680510e-01 -1.28120005e+00 -2.41297677e-01
-2.84613550e-01 1.42027736e-01 5.67812324e-01 1.31293476e+00
6.50089443e-01 7.29733407e-01 7.64404714e-01 -5.52953064e-01
-7.73411572e-01 -1.20064688e+00 -3.84238034e-01 2.85783947e-01
2.22415328e-01 -3.83172214e-01 -4.02081132e-01 -6.93790093e-02] | [9.417081832885742, 8.723991394042969] |
deeac490-f47d-4b40-867c-19ac4b22e8e9 | tfix-learning-to-fix-coding-errors-with-a | null | null | http://proceedings.mlr.press/v139/berabi21a.html | https://files.sri.inf.ethz.ch/website/papers/icml21-tfix.pdf | TFix: Learning to Fix Coding Errors with a Text-to-Text Transformer | The problem of fixing errors in programs has attracted substantial interest over the years. The key challenge for building an effective code fixing tool is to capture a wide range of errors and meanwhile maintain high accuracy. In this paper, we address this challenge and present a new learning-based system, called TFix. TFix works directly on program text and phrases the problem of code fixing as a text-to-text task. In turn, this enables it to leverage a powerful Transformer based model pre-trained on natural language and fine-tuned to generate code fixes (via a large, high-quality dataset obtained from GitHub commits). TFix is not specific to a particular programming language or class of defects and, in fact, improved its precision by simultaneously fine-tuning on 52 different error types reported by a popular static analyzer. Our evaluation on a massive dataset of JavaScript programs shows that TFix is practically effective: it is able to synthesize code that fixes the error in 67 percent of cases and significantly outperforms existing learning-based approaches. | ['Martin Vechev', 'Veselin Raychev', 'Jingxuan He', 'Berkay Berabi'] | 2021-07-18 | null | null | null | icml-2021-7 | ['program-repair', 'program-repair'] | ['computer-code', 'reasoning'] | [ 7.84166828e-02 -1.32463828e-01 -5.30890703e-01 -1.70111775e-01
-1.42298877e+00 -8.03494632e-01 1.11831330e-01 3.22097719e-01
1.60122991e-01 3.81977797e-01 -6.47269487e-02 -7.72155643e-01
9.70684290e-02 -7.94566274e-01 -1.02805424e+00 1.32787019e-01
1.91407740e-01 1.47188947e-01 4.09621418e-01 -2.84445077e-01
7.05173671e-01 -2.03569680e-01 -1.61188614e+00 6.94962263e-01
1.09678054e+00 1.53811142e-01 1.06131159e-01 8.81480157e-01
-3.63636404e-01 1.16266751e+00 -9.03343439e-01 -7.37264991e-01
6.57847449e-02 -2.34779090e-01 -1.03651810e+00 -5.55175364e-01
6.76645994e-01 -1.16941042e-01 3.52114528e-01 1.13742769e+00
3.38160783e-01 -5.93403161e-01 8.65356922e-02 -1.22743440e+00
-4.83042926e-01 1.19896007e+00 -6.32026196e-01 1.93656459e-01
4.91010576e-01 1.43455714e-01 1.29136443e+00 -6.70952320e-01
7.63519585e-01 9.37620580e-01 1.14435232e+00 5.86552620e-01
-1.29469800e+00 -4.61209744e-01 -3.68616968e-01 -1.89135835e-01
-9.07768726e-01 -3.01113427e-01 4.66995925e-01 -8.60853255e-01
1.68017793e+00 3.36104691e-01 1.67902693e-01 1.01456249e+00
4.81927484e-01 5.95584869e-01 8.75243485e-01 -1.00854743e+00
-4.45501022e-02 -3.78665552e-02 3.77295107e-01 1.05457175e+00
3.38676959e-01 -6.13897257e-02 -1.49511412e-01 -6.00971043e-01
5.80014810e-02 4.68050241e-02 -1.89525962e-01 -8.11152756e-02
-1.26468062e+00 8.21028948e-01 5.80719821e-02 5.73209405e-01
2.01055527e-01 5.45696974e-01 8.86781514e-01 5.31325340e-01
2.17170820e-01 9.62018907e-01 -1.05095828e+00 -7.05902517e-01
-1.05897462e+00 5.04162431e-01 9.85531688e-01 1.03194129e+00
9.95714843e-01 -7.16792122e-02 -6.29286245e-02 8.10476720e-01
1.66865304e-01 5.61772406e-01 4.26756054e-01 -8.79900336e-01
9.65534687e-01 1.16823971e+00 -1.13060482e-01 -9.80827570e-01
-1.31519884e-01 -2.98815280e-01 -1.14161673e-03 3.81202757e-01
3.64908785e-01 6.00287393e-02 -5.17266870e-01 1.65037298e+00
1.39640599e-01 -3.25357586e-01 -3.75795752e-01 2.14102134e-01
4.35779512e-01 3.07457924e-01 -1.40462309e-01 2.37692699e-01
1.11616814e+00 -8.50369692e-01 -4.22955126e-01 -5.12060404e-01
1.21956515e+00 -1.05101883e+00 1.45095837e+00 4.79690015e-01
-8.64930212e-01 -4.26623613e-01 -9.62792456e-01 -1.10161193e-01
-2.27864921e-01 4.24938947e-01 5.08198619e-01 7.85271525e-01
-1.08297718e+00 5.42465627e-01 -9.25603926e-01 -2.82941073e-01
1.78808928e-01 1.41733766e-01 -2.53556013e-01 -1.65709600e-01
-6.34917378e-01 8.46618533e-01 3.03873003e-01 -5.36824763e-01
-7.72238731e-01 -1.25153291e+00 -8.50419223e-01 1.96668372e-01
5.38659215e-01 -5.63133299e-01 1.63335168e+00 -9.11620736e-01
-1.04009867e+00 9.67745066e-01 -2.23209381e-01 -1.29735738e-01
3.87291819e-01 -3.07128251e-01 -3.19429964e-01 -4.62869793e-01
4.65129733e-01 -1.67150721e-02 8.00826609e-01 -1.00731325e+00
-7.11037636e-01 -1.09316431e-01 4.87265676e-01 -8.87288630e-01
-4.09919143e-01 5.62018633e-01 -2.55185574e-01 -5.29195368e-01
-5.18095136e-01 -1.01145887e+00 5.38991112e-03 -4.83168572e-01
-4.62287068e-01 -2.08035871e-01 4.66180921e-01 -8.37693393e-01
1.71931016e+00 -1.96763253e+00 2.07924306e-01 4.91008814e-03
5.13554931e-01 4.29755986e-01 -3.25889736e-01 6.59338593e-01
-3.83388191e-01 4.87864345e-01 -5.03005564e-01 -1.81145892e-02
2.69375801e-01 -4.82640229e-02 -4.62271988e-01 1.38411492e-01
3.23037058e-01 8.69308710e-01 -1.10032952e+00 -4.82294172e-01
3.07360850e-02 1.01426832e-01 -1.22852099e+00 1.77828982e-01
-7.43615627e-01 -2.38202348e-01 -4.19025362e-01 8.27392459e-01
5.17726123e-01 -2.66548604e-01 2.56651074e-01 1.64609462e-01
-2.23379612e-01 4.82498229e-01 -7.87179530e-01 1.82807648e+00
-9.25382555e-01 6.78225577e-01 -3.97229642e-01 -6.67562187e-01
9.67080235e-01 1.06294863e-01 2.34483302e-01 -6.00634277e-01
-1.98308647e-01 6.19205773e-01 -1.92806780e-01 -9.92583096e-01
4.20121223e-01 3.67510796e-01 -5.83655298e-01 8.38585377e-01
1.18993513e-01 -1.00711502e-01 4.45028245e-01 3.58318090e-01
2.01285338e+00 3.60956758e-01 3.82366717e-01 -2.22276688e-01
5.20071626e-01 3.62526804e-01 6.02904975e-01 7.82833993e-01
2.46547267e-01 5.02561092e-01 8.93349528e-01 -5.69034934e-01
-1.09190238e+00 -5.22934377e-01 1.21301785e-01 1.28587079e+00
-6.02359354e-01 -1.17749357e+00 -1.16826260e+00 -1.21271777e+00
2.78062597e-02 8.92644465e-01 -6.49316072e-01 -2.92883664e-01
-8.81138086e-01 -5.29062331e-01 8.02647471e-01 5.34220159e-01
8.56887326e-02 -1.02988756e+00 -7.25609779e-01 2.31716871e-01
-5.52926421e-01 -6.66629851e-01 -4.89076138e-01 8.52909237e-02
-5.35238624e-01 -1.55777431e+00 6.62760204e-03 -6.14181221e-01
4.68553424e-01 -4.44960594e-02 1.86396134e+00 8.33597004e-01
-5.27268767e-01 2.20636114e-01 -4.47664201e-01 -2.50221372e-01
-1.12501907e+00 6.38053417e-01 -5.91177404e-01 -7.08767354e-01
6.76498532e-01 -2.89289147e-01 1.93348199e-01 1.11133225e-01
-9.44715559e-01 -4.71957415e-01 3.90080035e-01 8.42652202e-01
-1.02337142e-02 -1.91091541e-02 3.52988839e-01 -1.44070411e+00
4.51353908e-01 -5.90975881e-01 -1.13856542e+00 5.54733038e-01
-6.06450856e-01 2.27613449e-01 6.92854881e-01 -9.99754146e-02
-9.48392212e-01 -5.17594330e-02 -5.47893643e-01 2.22583796e-04
2.54233420e-01 7.77440012e-01 -2.90951822e-02 -2.22154066e-01
1.18618429e+00 -1.41609326e-01 -3.71010333e-01 -4.88210648e-01
1.11773737e-01 6.83215559e-01 3.71473730e-01 -1.03009677e+00
9.51114118e-01 -1.64582267e-01 -3.70854586e-01 -2.22477406e-01
-7.58536339e-01 -2.55513400e-01 -5.21327674e-01 5.26955649e-02
3.00026923e-01 -5.84503114e-01 -5.54149032e-01 3.90793681e-01
-1.31909025e+00 -5.13702393e-01 -3.09369657e-02 -1.27332658e-01
-3.43476266e-01 2.55048424e-01 -5.52065730e-01 -2.83962280e-01
-4.41659182e-01 -1.44157684e+00 1.32348394e+00 -2.18216553e-01
-4.69483346e-01 -9.92160857e-01 7.53720522e-01 2.69380599e-01
6.80609167e-01 3.64027351e-01 1.38960767e+00 -3.75303149e-01
-7.38978744e-01 -2.80727327e-01 -8.04403797e-02 2.34228894e-01
2.35284150e-01 8.16110134e-01 -7.57592559e-01 -3.20960313e-01
-1.82825655e-01 -4.66482431e-01 6.54833376e-01 8.01637676e-03
1.03146195e+00 -2.84764409e-01 -3.60665441e-01 4.48840916e-01
1.85050499e+00 -3.23379040e-01 7.60618687e-01 6.60895646e-01
7.62803972e-01 2.03585193e-01 6.13738298e-01 3.53203058e-01
4.87938106e-01 6.44549429e-01 4.97602195e-01 1.63332328e-01
-4.35732715e-02 -3.05395544e-01 5.08468330e-01 8.55576694e-01
3.52913976e-01 -6.56421855e-02 -1.42822552e+00 7.94035733e-01
-1.65923536e+00 -9.12204623e-01 -4.93314117e-01 2.08550048e+00
1.27511430e+00 1.42670214e-01 6.50553405e-02 2.90197045e-01
5.51667750e-01 -2.57528633e-01 -6.20305240e-02 -6.37742519e-01
4.03696358e-01 4.61176872e-01 2.05457300e-01 4.62411284e-01
-7.92091310e-01 9.26589608e-01 6.75671196e+00 7.23938286e-01
-1.24602020e+00 3.39287490e-01 -1.43830981e-02 3.66221339e-01
-7.08549976e-01 3.72166276e-01 -9.96832490e-01 5.77277780e-01
1.18982577e+00 -2.69817144e-01 4.88818169e-01 1.15580869e+00
-2.69538254e-01 5.48153222e-02 -1.30388308e+00 4.35677499e-01
9.17477459e-02 -1.49962163e+00 -3.86962831e-01 -3.24675173e-01
8.61003816e-01 2.15227500e-01 -1.56704500e-01 7.77297974e-01
6.86855078e-01 -8.40392232e-01 8.17262709e-01 3.87204140e-01
8.29850614e-01 -4.96563822e-01 7.20805943e-01 3.96592826e-01
-8.38719010e-01 -2.77328104e-01 -4.62334335e-01 8.84757191e-02
-4.19067621e-01 9.99306083e-01 -9.93765116e-01 3.91663313e-01
9.51819003e-01 8.35740864e-01 -1.15714574e+00 1.01961315e+00
-2.36623704e-01 8.85209680e-01 -1.28009757e-02 1.24372862e-01
-2.24975511e-01 5.10077298e-01 2.31089920e-01 1.68268013e+00
4.46328431e-01 -5.48485577e-01 2.05323890e-01 1.28838396e+00
-2.44348198e-01 -4.22862917e-02 -4.90191877e-01 -2.22446978e-01
4.83282030e-01 1.32936430e+00 -2.69728005e-01 -2.92297900e-01
-5.35572529e-01 2.88997829e-01 6.65107965e-01 -7.08441734e-02
-1.10137808e+00 -7.99642682e-01 4.16802734e-01 8.41566771e-02
4.28120553e-01 5.99453598e-02 -1.77868351e-01 -1.45080304e+00
4.25838143e-01 -1.71331620e+00 3.79415631e-01 -5.38591921e-01
-8.97460938e-01 6.76670849e-01 -1.87015712e-01 -1.14025784e+00
-3.86908770e-01 -5.68830609e-01 -6.64217651e-01 7.97962785e-01
-1.53357160e+00 -1.10904086e+00 -3.96180689e-01 2.27008000e-01
4.37749982e-01 -1.30379111e-01 8.36392820e-01 6.92015886e-01
-5.16643882e-01 9.63476598e-01 -1.13594048e-01 2.12875083e-01
9.80857074e-01 -1.59958375e+00 8.88779998e-01 1.16738701e+00
-1.20456209e-02 1.22995520e+00 6.99219763e-01 -6.96810722e-01
-1.57366669e+00 -1.37990105e+00 1.31946170e+00 -1.13092494e+00
8.56619060e-01 -3.60425621e-01 -1.13855791e+00 9.28373098e-01
9.99397710e-02 8.49502310e-02 4.28242207e-01 2.05930561e-01
-1.05856085e+00 -1.53874874e-01 -1.00871944e+00 1.90925449e-01
7.42378712e-01 -8.57503891e-01 -6.08511806e-01 4.86320615e-01
6.65012956e-01 -7.33011305e-01 -1.00502467e+00 1.78901821e-01
3.22083414e-01 -1.17169619e+00 7.05259144e-01 -7.17205584e-01
1.15874398e+00 -3.74311954e-01 -1.90950379e-01 -1.26634407e+00
-1.60271972e-01 -6.53628647e-01 2.31495257e-02 1.65827346e+00
6.58274591e-01 -6.02626204e-01 2.72323281e-01 4.10406083e-01
-2.90292203e-01 -3.40320170e-01 -3.87703240e-01 -5.04601479e-01
2.75217474e-01 -4.68659133e-01 8.78763795e-01 1.13663948e+00
2.52454340e-01 -2.63435058e-02 -1.90820843e-01 -8.01301282e-03
3.49763244e-01 2.90330708e-01 1.19372368e+00 -1.00138307e+00
-7.02826083e-01 -2.28247300e-01 -2.84660280e-01 -5.17521739e-01
4.71829295e-01 -1.07149756e+00 2.48626783e-01 -1.01368523e+00
4.97211337e-01 -7.00221241e-01 4.60228417e-03 1.05291140e+00
-5.40436804e-01 -3.17014083e-02 -2.22011939e-01 1.02667622e-01
-5.92115760e-01 -1.67708203e-01 3.59193534e-01 -3.40035796e-01
1.15169279e-01 1.55505529e-02 -7.83522189e-01 5.33184350e-01
5.24248481e-01 -1.04215574e+00 -4.39175963e-02 -8.55053127e-01
9.56905782e-01 1.49467081e-01 2.26559952e-01 -9.05507624e-01
-5.63121289e-02 -7.04416111e-02 -3.30474555e-01 -2.01147467e-01
-6.21880949e-01 -5.40702045e-01 2.01875851e-01 5.45322537e-01
-5.40677726e-01 4.68473047e-01 3.86903703e-01 1.36003166e-01
-1.31115854e-01 -7.31321335e-01 6.80831432e-01 -3.01631451e-01
-5.72312236e-01 -2.13740155e-01 -1.81138009e-01 5.39446473e-01
7.59926796e-01 2.84663647e-01 -8.48497450e-01 4.93522644e-01
7.17777386e-02 -1.91815987e-01 9.69029248e-01 7.62529671e-01
8.78055096e-02 -9.39235806e-01 -7.25189626e-01 2.83555984e-01
6.08338833e-01 -5.40543258e-01 -1.94531664e-01 5.65556765e-01
-7.23206401e-01 3.53542864e-01 -4.33015786e-02 -8.76122832e-01
-1.40064085e+00 5.54058373e-01 4.18396205e-01 -5.59114337e-01
-3.88268501e-01 7.01693356e-01 -4.23983067e-01 -1.15574229e+00
-2.49948889e-01 -6.46118581e-01 1.93440020e-01 -4.39992756e-01
6.70729756e-01 2.13256717e-01 6.93771243e-01 -1.08332776e-01
-4.97716486e-01 6.65010691e-01 -9.66469422e-02 4.46636021e-01
1.52285862e+00 5.54825664e-01 -6.79002166e-01 2.80612469e-01
1.26192784e+00 6.38554931e-01 -7.06368268e-01 -2.94303328e-01
5.69076836e-01 -8.77750516e-01 -1.88776508e-01 -1.04366672e+00
-1.24837434e+00 7.92553127e-01 1.73705220e-01 1.77415431e-01
8.74391913e-01 -1.37832746e-01 6.71442986e-01 5.10325015e-01
7.63394237e-01 -6.36789083e-01 2.38920152e-01 5.44199586e-01
5.43526113e-01 -1.22312832e+00 -2.78439492e-01 -2.99747080e-01
-3.72194797e-02 1.38031006e+00 8.45132291e-01 3.90855409e-02
1.42685577e-01 8.10973525e-01 5.70734590e-02 -1.56832159e-01
-8.96002531e-01 3.64105970e-01 -3.20466384e-02 4.78481621e-01
8.90085220e-01 -1.54261991e-01 -9.72264037e-02 3.68906796e-01
-7.01717809e-02 3.29812378e-01 9.00730789e-01 1.25259709e+00
-2.94349164e-01 -1.59611559e+00 -4.54028606e-01 5.29941440e-01
-7.67454505e-01 -1.92799851e-01 -2.62837410e-01 6.93157256e-01
2.14489803e-01 8.53143215e-01 -5.48885345e-01 -4.61989015e-01
4.04289961e-01 -2.36887857e-02 5.94158113e-01 -9.73818123e-01
-1.19517219e+00 -4.91407007e-01 1.58355251e-01 -8.52771342e-01
-1.11640342e-01 -7.00649500e-01 -9.93905783e-01 -4.49456722e-01
-5.49086869e-01 1.50892287e-01 5.22913754e-01 9.83235896e-01
4.50555533e-01 7.17951119e-01 4.47291493e-01 -2.64007598e-01
-8.34655046e-01 -6.91583812e-01 1.25869215e-01 3.86962682e-01
5.28394163e-01 -4.21969473e-01 -4.47926044e-01 3.28394085e-01] | [7.693808555603027, 7.750820636749268] |
7c2a625f-5bba-44c7-a4e2-1e21d32bdd94 | integration-of-data-and-theory-for | 2109.01634 | null | https://arxiv.org/abs/2109.01634v4 | https://arxiv.org/pdf/2109.01634v4.pdf | AI Descartes: Combining Data and Theory for Derivable Scientific Discovery | Scientists have long aimed to discover meaningful formulae which accurately describe experimental data. A common approach is to manually create mathematical models of natural phenomena using domain knowledge, and then fit these models to data. In contrast, machine-learning algorithms automate the construction of accurate data-driven models while consuming large amounts of data. The problem of incorporating prior knowledge in the form of constraints on the functional form of a learned model (e.g., nonnegativity) has been explored in the literature. However, finding models that are consistent with prior knowledge expressed in the form of general logical axioms (e.g., conservation of energy) is an open problem. We develop a method to enable principled derivations of models of natural phenomena from axiomatic knowledge and experimental data by combining logical reasoning with symbolic regression. We demonstrate these concepts for Kepler's third law of planetary motion, Einstein's relativistic time-dilation law, and Langmuir's theory of adsorption, automatically connecting experimental data with background theory in each case. We show that laws can be discovered from few data points when using formal logical reasoning to distinguish the correct formula from a set of plausible formulas that have similar error on the data. The combination of reasoning with machine learning provides generalizeable insights into key aspects of natural phenomena. We envision that this combination will enable derivable discovery of fundamental laws of science and believe that our work is an important step towards automating the scientific method. | ['Lior Horesh', 'Bachir El Khadir', 'Nimrod Megiddo', 'Kenneth Clarkson', 'Joao Goncalves', 'Tyler Josephson', 'Vernon Austel', 'Sanjeeb Dash', 'Cristina Cornelio'] | 2021-09-03 | null | null | null | null | ['automated-theorem-proving', 'automated-theorem-proving'] | ['miscellaneous', 'reasoning'] | [ 2.44269565e-01 3.14940661e-01 -2.56391197e-01 -5.21466136e-01
1.47421390e-03 -5.59840858e-01 8.00894022e-01 2.68323898e-01
-2.29108021e-01 9.18877542e-01 -3.41631979e-01 -1.13089561e+00
-4.18598890e-01 -1.01692200e+00 -9.21374142e-01 -3.41834724e-01
-3.67668718e-02 5.60352921e-01 2.40502015e-01 -3.57690513e-01
5.87211192e-01 8.19008291e-01 -1.46067071e+00 6.81937709e-02
1.00084853e+00 8.24133098e-01 -9.11528617e-02 5.89978337e-01
-3.75171691e-01 1.13376200e+00 -9.43639800e-02 -3.02221090e-01
1.37792408e-01 -6.59663141e-01 -1.01335490e+00 -4.65933949e-01
1.07272215e-01 -1.41643152e-01 -1.01267412e-01 1.10288703e+00
-4.37491596e-01 4.68407124e-02 8.75996053e-01 -1.53621733e+00
-7.89443612e-01 4.53364611e-01 -1.67620018e-01 -2.68902004e-01
2.88982838e-01 -3.67938951e-02 1.12244642e+00 -3.69262487e-01
4.18433398e-01 1.17452741e+00 5.64381480e-01 6.53890550e-01
-1.63845193e+00 -3.50490212e-01 -3.80338654e-02 3.96318614e-01
-1.39830804e+00 -5.01969397e-01 8.14922750e-01 -4.88082409e-01
9.63857472e-01 5.07833362e-01 5.94578743e-01 4.37018722e-01
4.03424323e-01 2.34902039e-01 1.06523836e+00 -8.10528636e-01
5.40438652e-01 4.49484229e-01 5.35132829e-03 8.45245600e-01
9.47828472e-01 2.54583508e-01 -4.08604145e-01 -2.63979077e-01
9.06504810e-01 -2.03479528e-01 -1.00281704e-02 -4.44543004e-01
-8.30117464e-01 8.01870644e-01 1.40899226e-01 3.56614619e-01
-3.00797135e-01 2.93568343e-01 9.27676037e-02 2.06928551e-01
-2.79492908e-03 1.08598840e+00 -9.28182542e-01 2.29758188e-01
-7.98733115e-01 4.36087161e-01 1.37505805e+00 9.33785737e-01
8.61781240e-01 -3.19550604e-01 9.83648300e-01 -1.38481809e-02
5.98682046e-01 5.94534814e-01 2.16246054e-01 -1.30508959e+00
-2.35313252e-01 6.10555351e-01 6.88973725e-01 -9.13716376e-01
-3.20388407e-01 2.13709623e-01 -3.42329681e-01 3.16461682e-01
6.27867281e-01 -2.48073358e-02 -5.31352162e-01 1.62936437e+00
2.11963996e-01 1.51604516e-02 3.15217376e-01 5.45406222e-01
5.16491771e-01 5.39260626e-01 4.68517505e-02 -5.46680570e-01
9.46455479e-01 3.12599204e-02 -4.36737537e-01 6.22918010e-02
7.94367909e-01 -3.11091900e-01 9.33545828e-01 5.56465566e-01
-9.44631159e-01 -9.01969597e-02 -9.35397148e-01 -2.74814755e-01
-4.37856615e-01 -4.79154795e-01 1.26505184e+00 4.62770671e-01
-7.00783074e-01 8.32270861e-01 -9.93822336e-01 -2.77647913e-01
1.60088241e-01 4.79102433e-01 -2.36064374e-01 2.82367259e-01
-1.08203149e+00 1.28908825e+00 4.76537883e-01 9.00362805e-02
-4.76056218e-01 -7.65651822e-01 -8.04766297e-01 -1.05316557e-01
5.34008920e-01 -6.65035605e-01 1.31784844e+00 -8.90959442e-01
-1.54416847e+00 7.72681952e-01 -4.14856285e-01 -5.98060131e-01
2.53925949e-01 6.97281435e-02 -4.97283369e-01 3.94788943e-02
-2.26576969e-01 3.56649280e-01 3.23054135e-01 -1.18916750e+00
-3.14138830e-01 -1.17687896e-01 3.59759539e-01 -4.94491309e-01
9.73134208e-03 -1.16003439e-01 4.80356038e-01 1.62982211e-01
3.99344176e-01 -7.11513460e-01 -2.22024828e-01 2.69777089e-01
-2.51253188e-01 -4.00859892e-01 4.87481773e-01 -2.49894872e-01
8.85971546e-01 -1.74800944e+00 -2.20852089e-03 5.92574537e-01
1.57427862e-01 -1.51171699e-01 2.89502144e-01 4.18006778e-01
-9.24843028e-02 6.02039993e-01 -2.00092033e-01 5.10140479e-01
2.69756287e-01 5.23821473e-01 -5.54600060e-01 4.72074926e-01
3.13892573e-01 9.68962908e-01 -9.97972131e-01 -5.44851303e-01
2.71526575e-01 1.57721311e-01 -5.91912806e-01 1.05994575e-01
-7.96984732e-01 1.91386208e-01 -5.41878462e-01 1.65640339e-01
4.24830139e-01 -3.77195984e-01 5.37278593e-01 -1.36819825e-01
-4.20822263e-01 6.69989586e-01 -1.01781654e+00 1.21141970e+00
-3.95445913e-01 4.14725870e-01 -2.33457595e-01 -1.25924981e+00
9.85805511e-01 1.40839070e-01 5.02457559e-01 -4.80483532e-01
2.85524964e-01 3.43373269e-01 2.13012859e-01 -9.24552500e-01
1.67098105e-01 -1.07183015e+00 1.96787104e-01 5.62867701e-01
-2.94519007e-01 -8.17414045e-01 9.34034213e-02 -1.72056220e-02
7.63286471e-01 4.94938254e-01 5.22257984e-01 -4.31026638e-01
5.03247023e-01 4.42302018e-01 4.71801072e-01 7.34339952e-01
1.20716758e-01 -2.10344568e-01 5.48899651e-01 -7.44788766e-01
-1.27202427e+00 -8.76352191e-01 -2.69499987e-01 6.79029107e-01
1.54852077e-01 -4.86955166e-01 -3.65406722e-01 -1.39615372e-01
2.20780671e-01 1.28561604e+00 -6.27609193e-01 -1.88391060e-01
-3.23320866e-01 -6.31753445e-01 3.88330370e-01 3.50098610e-01
1.39410153e-01 -7.24573076e-01 -8.44695330e-01 -4.18605655e-02
1.98674351e-01 -1.03005564e+00 5.21232188e-01 3.32187742e-01
-8.29389334e-01 -1.38654280e+00 3.60006273e-01 -2.77278483e-01
9.31016862e-01 -7.87073821e-02 9.19106662e-01 3.91603291e-01
-3.01520377e-01 3.39045316e-01 -1.84633464e-01 -7.51805723e-01
-8.38268757e-01 -4.82032418e-01 3.40606898e-01 -5.03579617e-01
7.01211870e-01 -6.83913946e-01 -1.32835396e-02 1.57251656e-01
-1.07670355e+00 2.49665812e-01 3.57242256e-01 4.19697970e-01
5.14738739e-01 4.97224391e-01 3.13862890e-01 -7.73821056e-01
3.89988512e-01 -3.77342165e-01 -1.01585102e+00 5.63708663e-01
-8.12299848e-01 7.31962502e-01 9.79856312e-01 -3.69758934e-01
-1.08139741e+00 -1.28529938e-02 3.14058870e-01 -5.60326353e-02
-3.24652165e-01 8.12659025e-01 -2.78322041e-01 -4.37665870e-03
8.19985032e-01 1.56106070e-01 -3.93599682e-02 -2.56295174e-01
4.93944645e-01 4.54152733e-01 6.61534190e-01 -1.31720901e+00
7.96991110e-01 6.09010398e-01 7.79149652e-01 -8.71861637e-01
-7.94469833e-01 -1.39511758e-02 -7.91257918e-01 2.19550908e-01
5.79728782e-01 -2.94275105e-01 -9.11579072e-01 -1.43876299e-01
-1.03913534e+00 -4.13931131e-01 -3.89318138e-01 7.53637314e-01
-8.30456316e-01 3.67905080e-01 -8.20356607e-02 -1.25028133e+00
1.71483636e-01 -6.15216076e-01 5.19531369e-01 6.95150346e-02
-6.79711878e-01 -1.32621026e+00 8.47197175e-02 7.44063482e-02
1.08206287e-01 3.39870751e-01 1.26639211e+00 -7.79076397e-01
-4.81833905e-01 -1.34351313e-01 -8.95026177e-02 1.85848266e-01
4.26181585e-01 4.92464155e-01 -7.81541169e-01 1.98591232e-01
2.94661313e-01 -1.82002321e-01 3.94360393e-01 2.94900119e-01
9.90883708e-01 -6.25486434e-01 -2.97587425e-01 3.11428249e-01
1.35851264e+00 1.84256062e-01 5.47226727e-01 1.64088562e-01
3.16855580e-01 7.24569142e-01 2.93262243e-01 1.60914660e-01
1.95242643e-01 2.93298900e-01 1.14481628e-01 3.92190099e-01
4.56048131e-01 -3.52730095e-01 1.83655888e-01 4.22053307e-01
-4.17106032e-01 2.64184028e-01 -1.04113889e+00 1.42776981e-01
-1.71401823e+00 -1.29839373e+00 -3.46597403e-01 2.24533677e+00
1.23785150e+00 4.61349428e-01 1.10353604e-01 1.47300795e-01
3.26649010e-01 -7.58005083e-01 -6.88808560e-01 -6.42224908e-01
3.04620042e-02 2.32583538e-01 5.33556700e-01 9.04176712e-01
-5.54417908e-01 7.80737460e-01 7.30266953e+00 1.87025547e-01
-1.15594983e+00 -4.39053804e-01 3.09437681e-02 1.48905545e-01
-7.68474638e-01 6.66457355e-01 -5.29316425e-01 5.28661944e-02
1.32397842e+00 -5.72502971e-01 8.16197395e-01 5.88568687e-01
3.73908758e-01 -2.48025551e-01 -1.72872055e+00 5.47722936e-01
-2.10865289e-01 -1.35071409e+00 1.65363610e-01 3.86017300e-02
5.37895441e-01 -3.90786111e-01 -4.72038984e-01 -6.46209493e-02
7.13061154e-01 -1.20617104e+00 6.94213033e-01 9.50996161e-01
5.83145201e-01 -4.12153453e-01 4.82228518e-01 7.00926661e-01
-6.20021701e-01 9.84466448e-02 -5.04458845e-01 -8.10562849e-01
-1.90642416e-01 6.57937109e-01 -1.01155734e+00 4.71443355e-01
9.52485129e-02 5.22786081e-01 -2.23730788e-01 6.81458175e-01
-3.90303016e-01 5.55987656e-01 -7.86547244e-01 -2.73993790e-01
-2.48617843e-01 -2.40452692e-01 1.62243322e-01 9.32679296e-01
8.52951705e-02 5.40110111e-01 -3.35687578e-01 1.52615619e+00
3.04864347e-01 -8.25537816e-02 -6.26546621e-01 -4.03263658e-01
3.71792197e-01 7.67509282e-01 -7.44208932e-01 -3.50714147e-01
-5.43048680e-01 1.25603124e-01 1.84889287e-02 2.95331597e-01
-6.77629173e-01 -2.58026630e-01 3.65067035e-01 1.09407641e-01
1.09346412e-01 -4.61876631e-01 -5.97896695e-01 -1.26285815e+00
-3.87205579e-03 -7.31022060e-01 1.50253503e-02 -9.15376782e-01
-1.31246221e+00 -1.90350324e-01 4.50018436e-01 -6.17156982e-01
-3.43219966e-01 -1.15737844e+00 -5.53304255e-01 8.62857938e-01
-1.37706363e+00 -8.52652073e-01 1.31250918e-01 3.18647146e-01
-1.70824915e-01 2.36607268e-01 9.08589184e-01 -4.25371498e-01
-1.38887137e-01 1.92916684e-03 -6.35964051e-02 -1.04918301e-01
4.19882238e-01 -1.18454075e+00 4.98862937e-02 5.67163587e-01
8.00147578e-02 1.09180510e+00 1.27288961e+00 -7.77003765e-01
-1.87188661e+00 -6.73130631e-01 9.12486911e-01 -7.08193481e-01
1.14613569e+00 -1.05537005e-01 -9.46679354e-01 7.78312922e-01
-3.94917220e-01 -8.64126608e-02 8.17077041e-01 3.04418504e-01
-7.14492619e-01 -2.30799485e-02 -1.17258477e+00 5.44923544e-01
7.89151967e-01 -6.51908696e-01 -1.04987824e+00 2.97415406e-01
4.66686994e-01 -3.24960314e-02 -9.59543169e-01 2.54803449e-01
9.00804162e-01 -5.74833751e-01 6.76056147e-01 -1.35650587e+00
4.30355638e-01 -5.93056440e-01 -3.07839721e-01 -7.59792566e-01
-2.05765828e-01 -7.20323563e-01 -1.46142438e-01 7.05185294e-01
6.05743527e-01 -6.93587363e-01 4.41227257e-01 1.61744320e+00
2.13946640e-01 -5.32020986e-01 -5.99892139e-01 -1.07955658e+00
6.53918266e-01 -6.10759020e-01 5.65001667e-01 1.07786667e+00
7.51212120e-01 2.48640701e-01 1.02809332e-02 2.00354531e-01
5.21708667e-01 3.20261508e-01 7.88750112e-01 -1.72927213e+00
-2.22913638e-01 -5.05670786e-01 -5.46260417e-01 -7.39794016e-01
3.87598366e-01 -9.22305405e-01 1.13584258e-01 -1.48986316e+00
1.49310112e-01 -4.87101823e-01 -1.66228592e-01 5.97324669e-01
4.10119295e-01 -3.87409389e-01 -6.68472275e-02 2.29839385e-01
-3.64996374e-01 7.39405900e-02 1.03493369e+00 2.48903222e-02
-2.06166640e-01 -4.65033710e-01 -1.05621481e+00 1.09491897e+00
9.37560201e-01 -4.73734617e-01 -5.15187383e-01 6.19584834e-03
1.04875779e+00 -2.23054662e-01 7.54471183e-01 -6.82446361e-01
4.32929426e-01 -1.10196435e+00 2.61789292e-01 -3.84691730e-02
-2.20902756e-01 -1.00936437e+00 3.04649055e-01 3.73735249e-01
-5.59106290e-01 -3.90394270e-01 2.89726853e-01 2.64645576e-01
3.94775659e-01 -4.83942360e-01 5.65230131e-01 -1.98541448e-01
-6.03913367e-01 -2.66546875e-01 -1.39471397e-01 7.17736408e-02
8.95663559e-01 3.72177847e-02 -3.53009611e-01 -2.61759311e-01
-5.62913597e-01 9.77082923e-02 7.14386821e-01 -8.31706896e-02
5.35128295e-01 -8.63272488e-01 -6.04973972e-01 8.37154090e-02
-1.99966833e-01 -1.20538129e-02 -4.61498737e-01 8.03713202e-01
-6.85651124e-01 6.81694746e-01 -1.10523582e-01 -2.84800231e-01
-8.08027864e-01 8.35166633e-01 5.88466048e-01 2.91919470e-01
-2.94412255e-01 3.96872431e-01 4.90195192e-02 -1.98525771e-01
-2.68341929e-01 -7.90502906e-01 3.56040031e-01 -7.50012338e-01
5.04054904e-01 2.13446677e-01 -3.70970517e-01 -3.19527358e-01
-5.49742937e-01 4.80852962e-01 -1.40192108e-02 -6.84893653e-02
1.45687389e+00 7.75249153e-02 -4.88633513e-01 7.30175018e-01
8.00712705e-01 1.91648543e-01 -9.45102990e-01 -7.11639449e-02
1.85314924e-01 -1.69331774e-01 -1.55450806e-01 -8.56728196e-01
-2.21862793e-01 6.68008447e-01 -2.42910817e-01 5.26958287e-01
7.72840977e-01 3.98416370e-01 2.60612816e-01 1.17873287e+00
4.18860227e-01 -9.47555602e-01 -4.46348310e-01 2.69522011e-01
8.61739755e-01 -1.17508721e+00 6.10028625e-01 -4.23441231e-01
-1.03347532e-01 1.46493208e+00 3.83928657e-01 -1.26501238e-02
7.26723135e-01 4.74979341e-01 -2.43496343e-01 -2.89743811e-01
-8.10154378e-01 8.16095769e-02 1.97104827e-01 4.16038215e-01
5.34352779e-01 1.48767307e-01 -2.95855641e-01 5.50985575e-01
-5.49622297e-01 5.58677435e-01 6.02107346e-01 1.07461739e+00
-7.21582651e-01 -1.12300837e+00 -4.81914163e-01 3.52705836e-01
-1.58594236e-01 -9.42356661e-02 -7.61118770e-01 9.29623604e-01
3.08038741e-01 1.06792629e+00 -1.72973320e-01 -1.17795050e-01
-3.93377617e-03 3.13563824e-01 9.99063909e-01 -4.80605245e-01
5.05520284e-01 -2.60497928e-01 9.33559388e-02 -2.52947867e-01
-7.85292625e-01 -7.47707069e-01 -1.83286822e+00 -6.66587591e-01
-4.01940167e-01 6.53095484e-01 6.33229792e-01 1.45361733e+00
-9.15996730e-04 9.06750113e-02 1.46021783e-01 -3.94804001e-01
-6.50393069e-01 -4.23927128e-01 -5.05992532e-01 3.53329033e-01
3.57153535e-01 -6.37377024e-01 -5.04253507e-01 5.70844173e-01] | [8.742770195007324, 6.663479328155518] |
0c361f04-f6c3-4baf-89f0-7d08dae8ad74 | an-admm-approach-for-multi-response | 2303.11155 | null | https://arxiv.org/abs/2303.11155v1 | https://arxiv.org/pdf/2303.11155v1.pdf | An ADMM approach for multi-response regression with overlapping groups and interaction effects | In this paper, we consider the regularized multi-response regression problem where there exists some structural relation within the responses and also between the covariates and a set of modifying variables. To handle this problem, we propose MADMMplasso, a novel regularized regression method. This method is able to find covariates and their corresponding interactions, with some joint association with multiple related responses. We allow the interaction term between covariate and modifying variable to be included in a (weak) asymmetrical hierarchical manner by first considering whether the corresponding covariate main term is in the model. For parameter estimation, we develop an ADMM algorithm that allows us to implement the overlapping groups in a simple way. The results from the simulations and analysis of a pharmacogenomic screen data set show that the proposed method has an advantage in handling correlated responses and interaction effects, both with respect to prediction and variable selection performance. | ['Manuela Zucknick', 'Theophilus Quachie Asenso'] | 2023-03-20 | null | null | null | null | ['variable-selection'] | ['methodology'] | [ 4.13070589e-01 -3.47380280e-01 -4.02449429e-01 -7.25225031e-01
-5.36354065e-01 -3.55501115e-01 2.81246632e-01 5.21165073e-01
-3.53648007e-01 1.14134014e+00 2.02598855e-01 -2.06987679e-01
-6.49205148e-01 -6.97528839e-01 -6.74509764e-01 -9.51030195e-01
-2.39867210e-01 7.22403765e-01 2.45834012e-02 -9.66204479e-02
1.58548355e-01 4.86309022e-01 -1.31554472e+00 1.63164228e-01
1.15242696e+00 3.33643138e-01 2.21484438e-01 2.92940825e-01
1.89401194e-01 4.54194248e-01 -3.40818614e-01 -1.25111967e-01
7.16484711e-02 -6.03980720e-01 -4.22627062e-01 1.67683919e-03
1.72809944e-01 1.80057734e-01 2.28768393e-01 6.04722559e-01
6.54845297e-01 1.94263116e-01 1.00035203e+00 -1.07652044e+00
-1.55077606e-01 7.01387048e-01 -1.03837383e+00 -7.88601041e-02
2.03729868e-01 -2.64290422e-01 8.98828506e-01 -7.62302399e-01
4.14404362e-01 1.39834833e+00 5.59008658e-01 2.69091040e-01
-1.89235377e+00 -6.94286406e-01 2.19814733e-01 7.14437664e-02
-1.63028896e+00 -2.76242048e-01 5.28274596e-01 -6.36932611e-01
5.62961936e-01 6.16564989e-01 1.53899625e-01 6.73344254e-01
5.40013313e-01 1.91359743e-01 1.32003129e+00 -3.33544493e-01
2.74414688e-01 2.49513015e-01 5.52913666e-01 3.42692405e-01
2.22069085e-01 2.55668283e-01 -2.51612455e-01 -7.56491065e-01
3.93940151e-01 -4.90637915e-03 -1.71940267e-01 -4.24841374e-01
-6.81615829e-01 1.21952045e+00 1.94554657e-01 3.65003407e-01
-5.54209948e-01 5.40772500e-03 -2.20811227e-03 2.62781113e-01
5.59831738e-01 3.37632626e-01 -6.37394369e-01 7.67637849e-01
-8.04690778e-01 3.42900783e-01 5.93979478e-01 6.35527074e-01
7.73410261e-01 -3.26039791e-01 -4.19451058e-01 1.07882130e+00
3.64416242e-01 2.02342331e-01 2.76298225e-01 -2.83088177e-01
3.23416024e-01 7.82157719e-01 1.23891875e-01 -1.01129389e+00
-9.90765989e-01 -4.11378682e-01 -1.13611460e+00 1.44293979e-01
4.65979487e-01 -2.55580485e-01 -6.72796786e-01 2.09413552e+00
7.11367667e-01 1.79739773e-01 -2.97895640e-01 6.79848850e-01
5.73620677e-01 4.38905269e-01 5.72588384e-01 -8.74082565e-01
1.38562691e+00 -5.07543027e-01 -6.72335088e-01 7.61151984e-02
6.47537291e-01 -5.99272966e-01 5.40056348e-01 4.64672089e-01
-9.51567888e-01 -5.30415416e-01 -6.92417204e-01 9.15713701e-03
-9.13324803e-02 2.30515093e-01 6.80842876e-01 5.30398428e-01
-7.52319932e-01 6.09244227e-01 -4.88677680e-01 -1.35717280e-02
-1.71239339e-02 1.03840351e+00 -4.45118994e-01 -2.55720504e-02
-1.24991953e+00 8.76036584e-01 1.08369663e-01 2.69681700e-02
-4.82950717e-01 -8.79190445e-01 -6.45184278e-01 9.53119025e-02
2.84805357e-01 -1.01242399e+00 4.89165038e-01 -9.34606731e-01
-1.06368494e+00 5.41086257e-01 -5.25756896e-01 -1.19111516e-01
4.77286160e-01 1.41124234e-01 -2.95680970e-01 -2.22377673e-01
-1.19528227e-01 1.80027097e-01 6.86867654e-01 -7.73973286e-01
-4.81055230e-01 -7.35787809e-01 -3.64537686e-01 7.16896802e-02
1.20017774e-01 4.18681324e-01 -7.84614608e-02 -5.86910844e-01
1.99964195e-02 -9.13502276e-01 -8.46224368e-01 -3.64829749e-01
-4.96605784e-01 -2.47853413e-01 1.15496047e-01 -6.26571178e-01
1.45180678e+00 -2.03856754e+00 6.92679346e-01 6.94446921e-01
1.53778210e-01 -1.81705087e-01 -9.43546370e-02 4.41247582e-01
-5.76898575e-01 2.36018002e-01 -3.97893250e-01 -2.33356282e-01
-4.09671664e-01 -4.38278466e-02 6.30462095e-02 7.36019373e-01
1.34756431e-01 4.47232634e-01 -4.62305248e-01 -3.58403653e-01
-2.20339429e-02 5.00843048e-01 -6.81445897e-01 2.87319839e-01
-6.11348413e-02 7.84195304e-01 -6.84139788e-01 3.57184529e-01
1.05042577e+00 -1.62424278e-02 6.38179660e-01 -8.94484948e-03
-3.14287186e-01 -1.05028160e-01 -1.57640767e+00 1.00696194e+00
-1.81412444e-01 -9.36674550e-02 3.54648605e-02 -1.03120649e+00
1.10979831e+00 3.60367000e-01 8.83159995e-01 -3.56379628e-01
3.63959372e-02 2.60805041e-01 2.65180409e-01 -2.77532667e-01
1.33459732e-01 -5.11320949e-01 -6.06609695e-02 1.85279757e-01
-1.70061767e-01 3.66324544e-01 1.58211052e-01 -2.67860621e-01
9.32639480e-01 -2.31960922e-01 9.27152872e-01 -5.30403078e-01
9.14758682e-01 -1.70007378e-01 7.52563715e-01 7.01411903e-01
4.44503874e-01 2.84664124e-01 7.72913158e-01 -1.11363858e-01
-8.82866263e-01 -4.84907597e-01 -7.10723281e-01 9.45026040e-01
-1.91760734e-01 -7.39951357e-02 -1.72258765e-01 -3.26901913e-01
4.22378033e-01 6.14015222e-01 -8.16365659e-01 -8.23591650e-02
-5.17829716e-01 -1.50290608e+00 1.02416158e-01 5.78312278e-02
-1.08984597e-01 -3.95824224e-01 9.12140831e-02 5.90079308e-01
1.69471353e-01 -3.12808424e-01 -2.96981156e-01 6.93843603e-01
-1.00712502e+00 -1.17378426e+00 -5.68403184e-01 -6.01960182e-01
7.07993031e-01 3.86605896e-02 6.58472121e-01 1.46285752e-02
2.16024946e-02 -3.12132865e-01 -1.58944398e-01 -1.99153438e-01
-4.20478642e-01 -2.40869179e-01 1.87381934e-02 2.26003900e-01
2.92848140e-01 -6.82665944e-01 -4.49845672e-01 7.57969320e-01
-8.81121576e-01 -2.65707940e-01 4.88263905e-01 1.07769001e+00
8.07472229e-01 4.71942425e-02 8.85175288e-01 -1.27499557e+00
4.89171863e-01 -9.32313561e-01 -7.52519011e-01 3.67019147e-01
-7.64916480e-01 2.35377923e-01 6.20021522e-01 -6.71891451e-01
-9.62498426e-01 3.87688935e-01 -2.02006161e-01 1.23786516e-01
3.33940126e-02 8.96798372e-01 -5.65796196e-01 -1.26796871e-01
6.30600035e-01 -1.75524995e-01 -1.30890822e-02 -8.62442732e-01
-1.21131055e-02 4.90524948e-01 -2.46883437e-01 -5.22531331e-01
4.39878941e-01 6.51482865e-02 7.70711005e-01 -6.09485865e-01
-1.99176446e-01 -4.32255477e-01 -5.60161412e-01 2.73960829e-01
6.30869269e-01 -8.46588135e-01 -8.56941462e-01 2.50326365e-01
-8.33653212e-01 -1.10096723e-01 1.21520668e-01 8.46196234e-01
-4.32746708e-01 4.47793871e-01 -3.14488113e-01 -6.90027654e-01
3.04970425e-02 -1.26343989e+00 5.42697668e-01 -7.63981044e-02
-1.86329827e-01 -9.83309507e-01 4.10785377e-01 3.52981351e-02
2.84885112e-02 3.87667179e-01 1.36299241e+00 -1.16364181e+00
-1.95937499e-01 -1.27674535e-01 1.13358252e-01 -1.05862357e-01
1.64645553e-01 1.54161537e-02 -5.29347122e-01 -3.20882827e-01
1.61678210e-01 6.52322173e-02 8.57593715e-01 8.80859554e-01
1.06786621e+00 -2.22265542e-01 -5.19329846e-01 6.66811764e-01
1.51189756e+00 4.41475600e-01 6.38729334e-01 -1.09997794e-01
5.60195386e-01 8.04006159e-01 7.24360406e-01 5.92309415e-01
-4.80426401e-02 1.17754209e+00 6.56171516e-02 -4.12352949e-01
3.49914849e-01 -5.86165823e-02 6.75641820e-02 2.95432359e-01
6.95846090e-03 -4.42235500e-01 -4.26950186e-01 -1.37585644e-02
-1.96997595e+00 -8.04240644e-01 -1.04875600e+00 2.46210194e+00
1.04920697e+00 -4.35243100e-01 2.51949996e-01 -1.89357847e-01
8.94394457e-01 -3.23550016e-01 -4.52273369e-01 -6.30793989e-01
-4.19616878e-01 3.91024411e-01 6.35863721e-01 7.61606216e-01
-9.23692882e-01 3.80721301e-01 7.18942547e+00 9.29199636e-01
-7.28012621e-01 -7.45030642e-02 8.50189209e-01 1.11748032e-01
-4.52277780e-01 1.53048098e-01 -1.00140965e+00 3.90160918e-01
9.87185895e-01 -2.82484502e-01 2.67288417e-01 4.24950510e-01
7.62202203e-01 -4.07979846e-01 -1.34829986e+00 2.64995724e-01
-2.12918535e-01 -7.50773013e-01 3.15115489e-02 2.75017768e-01
5.89512885e-01 -6.77462280e-01 -7.65353739e-02 8.07495713e-02
2.20057666e-01 -1.18683505e+00 -1.64979964e-03 6.32746100e-01
6.48219287e-01 -7.76591063e-01 5.71025908e-01 4.36017632e-01
-1.00690567e+00 -9.32469293e-02 -5.08935213e-01 -7.48358443e-02
5.82663454e-02 7.48292923e-01 -8.48450720e-01 8.81029844e-01
-5.09423241e-02 4.78726178e-01 -4.61010665e-01 1.24597692e+00
-1.02307037e-01 4.18218255e-01 -1.87441245e-01 -1.10899110e-03
-1.44781455e-01 -6.02935612e-01 4.18903321e-01 1.07945442e+00
4.56538230e-01 2.77761549e-01 1.32569462e-01 7.84051180e-01
4.05041426e-01 8.22632611e-01 -4.36712176e-01 4.43384558e-01
1.97241262e-01 1.01101840e+00 -4.38525289e-01 -1.84094086e-01
-5.24883449e-01 5.39315760e-01 1.79809079e-01 4.32134181e-01
-9.12637174e-01 5.94362654e-02 5.19184530e-01 1.22013085e-01
1.91932350e-01 2.32996359e-01 -3.39116484e-01 -9.20461059e-01
-2.42422640e-01 -1.01365626e+00 8.23831201e-01 -3.09963584e-01
-1.33793032e+00 1.75089464e-01 2.89737821e-01 -1.03534877e+00
-1.14339106e-01 -4.36529934e-01 -2.87938416e-01 1.46676779e+00
-1.24818838e+00 -1.00705647e+00 2.69637108e-01 7.36142397e-01
1.80881634e-01 2.58192630e-03 8.90076876e-01 5.21874607e-01
-6.89197183e-01 6.05609894e-01 3.91539305e-01 -7.00222135e-01
9.89834368e-01 -1.04978228e+00 -5.14186323e-01 4.63257283e-01
-4.84134644e-01 9.23162997e-01 1.01222467e+00 -9.71648276e-01
-1.09286153e+00 -1.01130903e+00 1.17389143e+00 -7.21167251e-02
5.29322505e-01 -2.86021829e-01 -1.03373587e+00 6.02210104e-01
-1.29850999e-01 -4.77056146e-01 1.17130363e+00 5.52570164e-01
-1.51692644e-01 -2.02157557e-01 -1.24326241e+00 3.80950838e-01
5.90314686e-01 1.09494001e-01 -3.64295274e-01 4.11265403e-01
4.99894619e-01 1.99209470e-02 -1.05378127e+00 5.79481900e-01
4.90221709e-01 -6.88995898e-01 1.02059293e+00 -8.94956529e-01
2.00522795e-01 -3.79646659e-01 -6.50288686e-02 -1.45211506e+00
-8.26088548e-01 -4.23855126e-01 2.35357195e-01 1.16988933e+00
7.36561775e-01 -6.18660569e-01 4.37545329e-01 8.61831129e-01
1.27516851e-01 -6.87362731e-01 -1.04714596e+00 -5.21357477e-01
1.75033301e-01 1.46502897e-01 5.08533776e-01 9.14439440e-01
3.89859788e-02 6.66721582e-01 -1.02485883e+00 1.60882264e-01
4.68131363e-01 1.73993558e-01 6.37408316e-01 -1.43509161e+00
-8.12967539e-01 -1.34119689e-01 -2.10395679e-01 -6.52398646e-01
1.61122218e-01 -8.08467090e-01 -3.26134354e-01 -1.07173717e+00
6.59234643e-01 -5.83588064e-01 -3.77258867e-01 3.63473058e-01
-5.45358598e-01 -2.25153446e-01 -2.85181165e-01 -3.06658633e-02
3.28501463e-02 2.42685407e-01 9.75225985e-01 -6.89134747e-02
-6.77323461e-01 5.67647994e-01 -8.51170540e-01 3.48742813e-01
6.74544692e-01 -7.99184382e-01 -2.78787315e-01 2.71744549e-01
2.04423472e-01 6.37363434e-01 6.91063628e-02 -3.33380103e-01
6.04794882e-02 -5.29509008e-01 5.62760413e-01 -5.38799703e-01
1.60298094e-01 -9.88858223e-01 1.09862399e+00 6.89189851e-01
-5.57275951e-01 9.29316953e-02 9.54998806e-02 4.91650760e-01
-6.87883869e-02 -4.81935501e-01 8.85290504e-01 1.89500853e-01
-1.26916409e-01 1.65822566e-01 -5.19216001e-01 -5.85483015e-01
1.00212371e+00 3.09822317e-02 -1.60038888e-01 -3.13895643e-02
-1.08061016e+00 4.61654752e-01 2.06162632e-01 -4.24318574e-03
3.31577808e-01 -1.15100813e+00 -9.54486847e-01 3.70732546e-02
7.15314075e-02 -5.27228355e-01 4.67263043e-01 1.17407572e+00
1.22236133e-01 3.20543319e-01 -1.76272303e-01 -3.39841545e-01
-1.70083272e+00 9.93928730e-01 2.74705857e-01 -6.39980853e-01
-9.60153565e-02 4.70478147e-01 6.31519258e-01 -2.48345152e-01
1.84033681e-02 -1.54504776e-01 -7.52265990e-01 4.10081893e-01
4.69907671e-01 4.66648310e-01 -3.75856906e-02 -5.44168890e-01
-5.11690676e-01 6.43092573e-01 1.00998946e-01 1.66404918e-01
1.44084513e+00 -5.52766882e-02 -4.72915053e-01 4.54755902e-01
1.22067153e+00 4.21788305e-01 -7.07271039e-01 -1.70497835e-01
1.57634035e-01 -4.68707949e-01 -1.51714459e-01 -8.90793800e-01
-7.78699458e-01 2.75363356e-01 5.68778098e-01 -1.39405906e-01
1.19026256e+00 -1.57728344e-01 -2.34401897e-01 -1.75171457e-02
7.41731599e-02 -7.63057947e-01 -4.30229187e-01 1.77026093e-01
9.72521186e-01 -9.18601453e-01 4.93669063e-01 -7.58342385e-01
-4.93548006e-01 9.89838660e-01 3.84347469e-01 -2.93092504e-02
6.06497347e-01 7.93056414e-02 -3.91802877e-01 -9.26243961e-02
-8.46912205e-01 -9.47916806e-02 4.83471125e-01 3.31337303e-01
7.86636710e-01 1.82856932e-01 -1.49382818e+00 6.85519874e-01
3.71481031e-01 -1.78002715e-01 3.66570026e-01 3.60939234e-01
-3.78583878e-01 -1.83122289e+00 -5.20702362e-01 4.91002858e-01
-5.93594193e-01 -6.10705502e-02 -3.82121861e-01 8.70431900e-01
1.30386174e-01 1.11319232e+00 -1.94611326e-01 -1.94659948e-01
5.55971861e-01 4.57425378e-02 2.90446013e-01 -5.66632807e-01
-1.01082909e+00 8.59759390e-01 2.35305130e-01 -3.54949474e-01
-4.35313046e-01 -8.32042933e-01 -9.63350713e-01 -1.28057376e-01
-5.51837146e-01 4.73800302e-01 4.58483875e-01 7.37770438e-01
2.84938663e-01 5.50463915e-01 1.27308929e+00 -3.01382899e-01
-7.39505529e-01 -8.80170465e-01 -1.08290851e+00 2.26574764e-01
2.78798401e-01 -8.40570509e-01 -3.46184462e-01 -2.89051175e-01] | [7.722726345062256, 4.891781330108643] |
c72b04e4-3b06-4798-b7e0-1e627cfe1152 | a-preliminary-study-of-chatgpt-on-news | 2306.10702 | null | https://arxiv.org/abs/2306.10702v1 | https://arxiv.org/pdf/2306.10702v1.pdf | A Preliminary Study of ChatGPT on News Recommendation: Personalization, Provider Fairness, Fake News | Online news platforms commonly employ personalized news recommendation methods to assist users in discovering interesting articles, and many previous works have utilized language model techniques to capture user interests and understand news content. With the emergence of large language models like GPT-3 and T-5, a new recommendation paradigm has emerged, leveraging pre-trained language models for making recommendations. ChatGPT, with its user-friendly interface and growing popularity, has become a prominent choice for text-based tasks. Considering the growing reliance on ChatGPT for language tasks, the importance of news recommendation in addressing social issues, and the trend of using language models in recommendations, this study conducts an initial investigation of ChatGPT's performance in news recommendations, focusing on three perspectives: personalized news recommendation, news provider fairness, and fake news detection. ChatGPT has the limitation that its output is sensitive to the input phrasing. We therefore aim to explore the constraints present in the generated responses of ChatGPT for each perspective. Additionally, we investigate whether specific prompt formats can alleviate these constraints or if these limitations require further attention from researchers in the future. We also surpass fixed evaluations by developing a webpage to monitor ChatGPT's performance on weekly basis on the tasks and prompts we investigated. Our aim is to contribute to and encourage more researchers to engage in the study of enhancing news recommendation performance through the utilization of large language models such as ChatGPT. | ['Edward C. Malthouse', 'Yongfeng Zhang', 'Xinyi Li'] | 2023-06-19 | null | null | null | null | ['fake-news-detection'] | ['natural-language-processing'] | [-2.78237760e-01 1.37993440e-01 -5.12003124e-01 -2.47279778e-01
-6.44221604e-01 -5.17742276e-01 7.18997717e-01 2.92313337e-01
-2.81373650e-01 2.89517730e-01 1.02098799e+00 -6.06173754e-01
-2.33571250e-02 -5.87746680e-01 -3.27004462e-01 8.45148042e-02
1.77520305e-01 3.06910127e-01 1.27288431e-01 -6.00700438e-01
8.04229975e-01 -7.90839717e-02 -1.25387132e+00 9.10000026e-01
9.55501080e-01 4.84496534e-01 1.73316136e-01 5.28288901e-01
-4.92646635e-01 1.11623228e+00 -5.43809652e-01 -5.87872863e-01
8.02714303e-02 -4.17029917e-01 -8.10861051e-01 -1.69529602e-01
8.89652222e-02 -5.05569756e-01 -2.11451799e-01 5.90285003e-01
6.31645024e-01 5.65547168e-01 3.82857859e-01 -7.42814183e-01
-1.25080013e+00 1.11486089e+00 -3.26631159e-01 3.12837392e-01
7.04400718e-01 -1.73157707e-01 1.27430785e+00 -6.95150912e-01
6.05202198e-01 1.06031120e+00 6.78805590e-01 3.89501691e-01
-1.08801484e+00 -4.90547806e-01 3.26116145e-01 -1.21877603e-01
-8.77553761e-01 -5.01553655e-01 4.69716370e-01 -5.33163011e-01
1.11856318e+00 6.06577039e-01 3.57247531e-01 1.53169262e+00
1.96661338e-01 6.72637105e-01 1.32860494e+00 -3.37097049e-01
8.18998292e-02 6.88071370e-01 2.92586416e-01 1.78610653e-01
-3.56280804e-02 -1.86004966e-01 -6.24798119e-01 -4.80500370e-01
3.60535622e-01 4.97899503e-02 -1.79568589e-01 4.88260925e-01
-8.56985688e-01 1.16011047e+00 3.11837923e-02 4.23612893e-01
-4.04130965e-01 -4.57556278e-01 6.01012111e-01 6.34342849e-01
1.31242120e+00 1.06600571e+00 -4.62284446e-01 -6.70138776e-01
-8.05846035e-01 3.56737882e-01 1.22589242e+00 6.79693341e-01
1.10698253e-01 -2.08898500e-01 -4.76886392e-01 1.13900781e+00
2.88339287e-01 2.36253142e-01 6.43797696e-01 -7.03522265e-01
4.29878026e-01 5.16801655e-01 2.97899842e-01 -1.27951264e+00
-4.00664330e-01 -7.42237329e-01 -1.73897352e-02 -4.55928445e-01
3.97689044e-01 -3.37681085e-01 -1.75093815e-01 1.17988253e+00
-4.25465256e-02 -2.13988841e-01 -1.44539654e-01 8.21776628e-01
8.20404589e-01 8.94205809e-01 1.21660583e-01 -3.00376952e-01
1.42535830e+00 -8.91436458e-01 -4.84027177e-01 -3.06697249e-01
1.03789425e+00 -1.49467742e+00 1.45874548e+00 1.62718549e-01
-7.43751287e-01 -2.68751085e-01 -3.84220541e-01 -1.14483438e-01
-1.47547349e-01 2.24450856e-01 6.03223503e-01 7.30900466e-01
-9.57192302e-01 5.12882829e-01 -5.21301568e-01 -7.87827432e-01
2.06282511e-02 -1.55473366e-01 4.91769165e-02 -2.07563937e-02
-1.21230912e+00 1.03700948e+00 -4.36681032e-01 -1.69067949e-01
-1.17344871e-01 -4.57683414e-01 -3.33796203e-01 1.36600435e-01
4.75930005e-01 -3.79312843e-01 1.46346927e+00 -1.00847578e+00
-1.79968727e+00 3.13264370e-01 -2.77250167e-02 -1.87693864e-01
4.02922451e-01 -3.61803681e-01 -5.39994836e-01 -2.36334860e-01
1.95756763e-01 -7.54510388e-02 5.72382689e-01 -7.97243536e-01
-6.07052863e-01 -5.75771602e-03 2.37453222e-01 1.46070510e-01
-6.19161606e-01 6.47795379e-01 -7.64600188e-02 -7.78428018e-01
-2.30667114e-01 -9.18193460e-01 -1.26165599e-01 -6.05357051e-01
-2.66667664e-01 -2.87357330e-01 6.40549839e-01 -8.21070969e-01
1.39954185e+00 -2.00511503e+00 -4.41137910e-01 2.28731990e-01
1.20537460e-01 2.04371184e-01 -2.47368798e-01 1.00141096e+00
5.68306208e-01 4.90251839e-01 6.74523115e-01 -3.43060493e-01
9.26780105e-02 -2.12827161e-01 -5.35474479e-01 1.05170913e-01
-3.69036674e-01 7.23714590e-01 -8.30317855e-01 -5.63854985e-02
-3.00394028e-01 5.21309674e-01 -9.28969681e-01 1.01153664e-01
-4.05030966e-01 2.51422137e-01 -7.50001967e-01 1.75643742e-01
2.04513177e-01 -5.58109164e-01 -6.58043381e-03 2.31288001e-01
-5.65351069e-01 1.21747303e+00 -6.02325737e-01 8.53417277e-01
-7.15614796e-01 5.42669415e-01 -1.12179466e-01 -3.85001481e-01
9.03245091e-01 4.58672881e-01 3.32455814e-01 -9.47981715e-01
1.51820347e-01 7.41732307e-03 5.62878847e-02 -8.09677005e-01
1.00615644e+00 6.45693094e-02 1.75476253e-01 1.28288960e+00
-4.39019829e-01 4.82495874e-01 -1.69305690e-02 4.30601567e-01
9.29014862e-01 7.60751367e-02 6.24281019e-02 -3.07016730e-01
-7.75516778e-02 1.47731751e-01 -5.42813279e-02 1.11087763e+00
1.67988643e-01 3.05402905e-01 3.81749421e-01 -2.05621034e-01
-6.64866805e-01 -7.96936154e-02 1.41312763e-01 1.93094957e+00
-2.78167784e-01 -7.96176612e-01 -4.70403761e-01 -7.10386336e-01
-1.91324085e-01 1.24183381e+00 -3.78425032e-01 5.68290353e-02
-2.93371588e-01 -6.86431706e-01 2.71768272e-01 1.72455832e-01
1.39331833e-01 -9.02620733e-01 -2.37702310e-01 5.25048018e-01
-6.11858249e-01 -9.76503372e-01 -9.01207089e-01 -2.24749789e-01
-8.33233356e-01 -6.72497869e-01 -6.95513189e-01 -5.47203064e-01
5.56383610e-01 8.26934576e-01 8.80891681e-01 1.52442664e-01
4.83109713e-01 7.19648480e-01 -1.07730615e+00 -2.12734118e-01
-7.04630613e-01 4.19200361e-01 -9.93894413e-02 -1.93320751e-01
3.84016365e-01 -5.65888286e-01 -4.12131935e-01 6.82848632e-01
-7.10595250e-01 2.89277852e-01 3.36081088e-01 4.56322342e-01
-3.43046844e-01 -1.65352881e-01 8.42086256e-01 -1.42910218e+00
1.36788416e+00 -9.00510907e-01 -2.06905738e-01 4.93054725e-02
-8.60468030e-01 -2.34947309e-01 7.34360218e-01 -6.60509765e-01
-1.34432030e+00 -9.48327065e-01 -4.64358121e-01 4.04729396e-01
1.15149915e-01 1.12684655e+00 5.80879748e-01 -1.20978184e-01
1.07097423e+00 -2.17433311e-02 6.59351572e-02 -6.68321669e-01
2.87591159e-01 1.04340601e+00 -3.09919894e-01 -5.87090492e-01
4.62504029e-01 3.15067284e-02 -1.10216153e+00 -7.70106077e-01
-9.71724331e-01 -6.09643698e-01 4.18272525e-01 -2.78755784e-01
3.04460704e-01 -8.24926913e-01 -5.80657423e-01 -8.71524662e-02
-9.93867457e-01 -2.73133934e-01 4.85653691e-02 7.58629382e-01
-7.40133077e-02 3.66701275e-01 -1.30460656e+00 -9.02877808e-01
-5.75914800e-01 -1.00375140e+00 2.97307044e-01 -7.91849848e-03
-8.32969487e-01 -1.04881978e+00 1.75822303e-01 9.79049921e-01
1.13813913e+00 -5.42608798e-01 9.60115910e-01 -1.31148624e+00
-2.42482230e-01 -5.76845765e-01 -1.67105719e-01 1.82811677e-01
7.91439563e-02 -4.05394137e-02 -7.96129525e-01 -2.65356302e-01
1.46362334e-01 -1.95418328e-01 3.23431402e-01 1.50724724e-01
6.19540930e-01 -9.92133021e-01 -1.81949213e-02 -1.30747944e-01
7.57793605e-01 8.89582932e-02 3.34755242e-01 6.94015265e-01
4.48821902e-01 8.28213096e-01 4.98260617e-01 6.76357508e-01
6.06197596e-01 7.13266850e-01 1.81858111e-02 2.65138298e-01
2.15656862e-01 -5.82050681e-01 7.00612664e-01 1.08594525e+00
-2.95305431e-01 -6.68500185e-01 -4.57304746e-01 7.98129365e-02
-1.79175162e+00 -1.09627891e+00 -6.12685718e-02 2.05369306e+00
5.34765899e-01 2.71439850e-01 2.41930977e-01 -4.37669516e-01
6.05072737e-01 2.17642680e-01 -1.21077904e-02 -7.38815010e-01
2.24058807e-01 -2.26642981e-01 4.27715361e-01 4.17688876e-01
-4.41860318e-01 6.76104605e-01 6.05456543e+00 4.77187485e-01
-1.38361800e+00 1.60612673e-01 7.46431768e-01 1.23459801e-01
-6.89352453e-01 1.44409314e-01 -7.19413817e-01 6.65508509e-01
9.85281527e-01 -3.41653258e-01 5.25049448e-01 9.46063638e-01
8.62717211e-01 4.43628542e-02 -7.41637766e-01 2.90085614e-01
1.67472750e-01 -1.40349960e+00 -6.57998864e-03 2.31418893e-01
6.00663483e-01 3.47892076e-01 2.66340435e-01 7.23713934e-01
3.36242169e-01 -5.30592740e-01 6.46054387e-01 3.00540458e-02
1.70492649e-01 -2.43299901e-01 8.71108770e-01 6.72456920e-01
-4.10178185e-01 -2.37876596e-03 -2.69769043e-01 -7.26671875e-01
2.79181391e-01 4.51164991e-01 -1.11099792e+00 2.15001032e-01
3.96321982e-01 7.91199148e-01 -2.89136261e-01 9.13853288e-01
-1.96672627e-03 1.20720124e+00 -1.38639167e-01 -3.91309083e-01
1.81231201e-01 -2.47585237e-01 6.02853775e-01 1.34792507e+00
5.55387497e-01 1.65241823e-01 2.95046747e-01 5.73502421e-01
-2.06397206e-01 7.53195345e-01 -2.54153669e-01 -5.35345614e-01
4.19153064e-01 1.17468393e+00 -8.71890962e-01 -6.55147210e-02
-7.73770511e-01 6.13567710e-01 2.08466962e-01 3.47590983e-01
-5.97314477e-01 1.59486324e-01 4.65050578e-01 8.93419981e-01
3.84203754e-02 -6.62061945e-02 -3.07634324e-01 -1.20825279e+00
-1.95683420e-01 -1.29652190e+00 2.24350110e-01 -7.19893277e-01
-1.47857118e+00 7.29257762e-01 -2.01706082e-01 -9.10631359e-01
-1.55311421e-01 -1.01615019e-01 -8.05128753e-01 7.70376623e-01
-1.20136702e+00 -8.47252131e-01 1.90684468e-01 2.88975090e-01
7.75626898e-01 -1.35281058e-02 7.42492855e-01 5.21821976e-01
-3.94258827e-01 6.40563428e-01 2.18969792e-01 -2.02910796e-01
1.09438217e+00 -6.28341436e-01 4.30305511e-01 6.59157574e-01
1.56541467e-01 1.03577960e+00 1.08370960e+00 -8.40597630e-01
-1.21927142e+00 -8.30885708e-01 1.45448780e+00 -5.73093891e-01
8.67651641e-01 -2.32501715e-01 -8.99159908e-01 8.81372929e-01
1.88473433e-01 -6.38394415e-01 1.02358460e+00 6.96772158e-01
-4.17847216e-01 3.64177316e-01 -9.71674681e-01 7.46907473e-01
7.80822456e-01 -7.12989509e-01 -4.83005464e-01 5.15946090e-01
6.78993642e-01 -4.89320666e-01 -7.95562506e-01 -2.60641426e-01
6.15500152e-01 -7.41411388e-01 4.77738798e-01 -6.16369724e-01
7.10663021e-01 1.97292492e-01 1.33850381e-01 -1.44286191e+00
-5.94918728e-01 -9.50461686e-01 2.82061785e-01 1.47404337e+00
8.01196814e-01 -1.01011634e+00 6.30937994e-01 1.02938032e+00
-1.80624709e-01 -5.58882415e-01 -3.42884928e-01 -2.76797295e-01
-2.20483705e-01 -4.54065710e-01 2.36427516e-01 1.18239772e+00
5.30964971e-01 5.21505833e-01 -7.81347513e-01 -7.65033886e-02
-2.05426052e-01 5.44479117e-02 6.92048252e-01 -9.74059820e-01
-6.06160522e-01 -4.37369168e-01 2.58003235e-01 -1.29145825e+00
-1.43445462e-01 -8.10105026e-01 -1.95945352e-01 -1.49927104e+00
2.84806550e-01 -5.17163634e-01 8.83348659e-02 3.16182107e-01
-5.30503280e-02 1.41136255e-02 1.94157243e-01 5.03020227e-01
-6.71119511e-01 3.09015036e-01 1.25611746e+00 2.58544594e-01
-5.43946981e-01 6.79909408e-01 -1.55808890e+00 4.38443899e-01
7.58256257e-01 -7.22837210e-01 -4.73122597e-01 -3.81899804e-01
7.13583171e-01 6.68338612e-02 -1.92709684e-01 -3.88023585e-01
2.07310542e-01 -2.52471805e-01 -1.90611467e-01 4.09683511e-02
-6.18565828e-02 -6.12227321e-01 2.35486757e-02 -8.48721638e-02
-1.01560581e+00 3.79469037e-01 -3.70497927e-02 3.56949598e-01
-1.00394912e-01 -1.69635087e-01 1.93943501e-01 -2.44127899e-01
-1.60681576e-01 -3.10326703e-02 -1.15410066e+00 7.08491281e-02
4.81059700e-01 -1.81709617e-01 -6.77495360e-01 -1.08117723e+00
-4.77903605e-01 7.92868882e-02 3.82025063e-01 7.07446396e-01
3.10191363e-01 -6.63922250e-01 -7.87200630e-01 -5.17203324e-02
1.75102070e-01 -7.85808086e-01 3.49673659e-01 1.08531654e+00
-1.38537586e-01 4.22384441e-01 7.42037669e-02 8.11795294e-02
-1.08007622e+00 2.34261766e-01 8.69360566e-03 -3.09888005e-01
-8.28486204e-01 6.91811740e-01 2.08036434e-02 -3.66352588e-01
9.38550457e-02 -1.99702740e-01 -5.83511889e-01 3.89922148e-04
7.56478071e-01 4.14348692e-01 4.82971556e-02 -4.62013125e-01
2.42367074e-01 -2.86827207e-01 -5.94571650e-01 -1.19403424e-02
1.40941274e+00 -4.33954388e-01 -9.21224803e-02 3.27109098e-01
8.99984062e-01 7.19642401e-01 -7.49174654e-01 -3.99418354e-01
2.06657037e-01 -5.30230820e-01 2.38040090e-01 -1.04997504e+00
-7.53079772e-01 1.43479332e-01 -5.01228077e-03 7.16051102e-01
5.15078604e-01 -4.40803505e-02 9.58712459e-01 3.11510891e-01
2.44843155e-01 -9.71837580e-01 -1.94225702e-02 6.60599530e-01
7.48592556e-01 -1.08817804e+00 -5.22568375e-02 -2.57668287e-01
-9.24873114e-01 1.12392914e+00 4.78667587e-01 2.20954686e-01
5.75009823e-01 1.12229390e-02 4.04942870e-01 -8.04084986e-02
-9.79372025e-01 2.90956110e-01 2.91410059e-01 6.76785037e-02
1.20311368e+00 1.51434354e-02 -7.27885425e-01 8.74611974e-01
-2.50861436e-01 1.75977901e-01 8.58633280e-01 8.01006675e-01
-5.51602840e-01 -1.08650899e+00 -2.17161387e-01 9.54270482e-01
-9.01117206e-01 -3.80730242e-01 -3.07200760e-01 4.00526971e-01
-4.89563942e-01 1.38334429e+00 -3.63816708e-01 -5.75299263e-01
4.07709777e-01 -1.11291595e-01 -3.13357502e-01 -9.56302822e-01
-1.15677786e+00 7.27711394e-02 7.30592549e-01 -3.82226825e-01
-2.32912436e-01 -5.75030506e-01 -5.40873110e-01 -7.29190469e-01
-5.65938175e-01 8.42931628e-01 6.11191332e-01 1.14422011e+00
7.52414346e-01 8.15865621e-02 5.25665760e-01 -5.11575222e-01
-7.12006688e-01 -1.09813499e+00 -3.50172818e-01 1.64894477e-01
-1.04150429e-01 -2.21149966e-01 -4.49404508e-01 -4.10114348e-01] | [10.435492515563965, 6.003114700317383] |
e75e39c8-972d-4a27-8468-686d00c22456 | self-supervised-learning-of-remote-sensing | 2104.0707 | null | https://arxiv.org/abs/2104.07070v2 | https://arxiv.org/pdf/2104.07070v2.pdf | Self-Supervised Learning of Remote Sensing Scene Representations Using Contrastive Multiview Coding | In recent years self-supervised learning has emerged as a promising candidate for unsupervised representation learning. In the visual domain its applications are mostly studied in the context of images of natural scenes. However, its applicability is especially interesting in specific areas, like remote sensing and medicine, where it is hard to obtain huge amounts of labeled data. In this work, we conduct an extensive analysis of the applicability of self-supervised learning in remote sensing image classification. We analyze the influence of the number and domain of images used for self-supervised pre-training on the performance on downstream tasks. We show that, for the downstream task of remote sensing image classification, using self-supervised pre-training on remote sensing images can give better results than using supervised pre-training on images of natural scenes. Besides, we also show that self-supervised pre-training can be easily extended to multispectral images producing even better results on our downstream tasks. | ['Vladimir Risojević', 'Vladan Stojnić'] | 2021-04-14 | null | null | null | null | ['remote-sensing-image-classification'] | ['miscellaneous'] | [ 7.85270035e-01 -5.08502424e-02 -2.67086983e-01 -4.77811247e-01
-4.34007376e-01 -4.45189208e-01 6.41499698e-01 5.72927892e-01
-7.20234692e-01 5.78154147e-01 -8.76151398e-02 -4.76055771e-01
-2.28592634e-01 -9.67928648e-01 -5.31959951e-01 -7.56558001e-01
-9.58069861e-02 2.72218734e-01 -4.21345513e-03 -2.05770489e-02
-1.45450145e-01 7.25112915e-01 -1.71207845e+00 2.55535245e-01
7.72845626e-01 5.56792140e-01 5.73098063e-01 3.83598000e-01
2.02964898e-02 6.57320261e-01 -3.34276110e-01 1.98061377e-01
3.28983039e-01 -4.94408965e-01 -8.78114283e-01 6.15342021e-01
3.64857823e-01 -3.98383364e-02 9.61953551e-02 1.05026913e+00
3.67022604e-01 2.84830421e-01 9.32229757e-01 -7.12108076e-01
-1.00723922e-01 4.03640658e-01 -5.48915505e-01 3.28323483e-01
-2.58207053e-01 5.43006361e-02 9.68364596e-01 -6.45738780e-01
5.89946926e-01 9.45845068e-01 3.85780007e-01 1.87893033e-01
-1.33003509e+00 -2.65646785e-01 4.46099266e-02 1.24860808e-01
-1.27868283e+00 -3.57589483e-01 7.38539398e-01 -5.29591203e-01
4.76969719e-01 1.52360499e-01 2.04906404e-01 5.08431196e-01
-1.22304894e-01 6.94173992e-01 1.49131298e+00 -7.83340096e-01
4.46889579e-01 4.57951814e-01 4.76783216e-02 3.31359535e-01
2.40797363e-02 3.24374139e-01 -1.53640717e-01 1.56007722e-01
7.80631304e-01 2.39452079e-01 -7.66405016e-02 -3.54963809e-01
-1.15901387e+00 1.09946871e+00 8.67644131e-01 7.03017056e-01
-2.71504194e-01 -2.88611144e-01 2.04168633e-01 4.57359493e-01
9.64495361e-01 5.72623074e-01 -4.23535168e-01 4.76415902e-01
-1.16805029e+00 -3.11444998e-01 2.69675374e-01 3.76410693e-01
1.07500815e+00 -3.08456700e-02 1.57217041e-01 1.11414373e+00
-3.96250449e-02 5.01949370e-01 4.19094115e-01 -4.66043413e-01
1.18456587e-01 5.77400029e-01 -9.65922996e-02 -7.22515821e-01
-6.09670103e-01 -6.85266316e-01 -1.05514610e+00 4.31551933e-01
5.08675337e-01 -2.64040738e-01 -8.97734106e-01 1.14486289e+00
2.13526450e-02 -1.58050850e-01 2.90848672e-01 9.10038471e-01
7.62522697e-01 8.15916955e-01 2.55420327e-01 -3.79245847e-01
1.00643480e+00 -9.24549222e-01 -2.77637780e-01 -5.03577948e-01
6.81440294e-01 -6.08656049e-01 1.02656341e+00 3.83156031e-01
-2.99212784e-01 -7.71218419e-01 -7.02290356e-01 3.08613271e-01
-4.74033713e-01 6.84369087e-01 6.47024930e-01 4.42691833e-01
-8.09883177e-01 6.20751977e-01 -5.77822268e-01 -8.27943623e-01
4.65783328e-01 1.65264040e-01 -5.48238039e-01 -1.78417802e-01
-7.55517066e-01 8.35394740e-01 5.18806517e-01 -3.54892984e-02
-8.34106624e-01 -3.19650501e-01 -8.34006190e-01 1.24570087e-01
3.96442950e-01 -1.36929080e-02 8.83647621e-01 -1.42255104e+00
-9.27726746e-01 1.16343367e+00 7.46204555e-02 -4.17461097e-01
3.87553155e-01 -1.47524979e-02 -2.75980353e-01 1.70834184e-01
1.29443809e-01 6.49492145e-01 9.65377152e-01 -1.30824900e+00
-4.85237777e-01 -5.06919920e-01 -4.35155220e-02 1.79934263e-01
-4.33042228e-01 -2.26477012e-01 1.13070242e-01 -6.33045793e-01
8.46571848e-02 -9.04122651e-01 -6.38712704e-01 1.16200149e-01
-1.97177634e-01 -1.52915781e-02 8.73159289e-01 -2.60755002e-01
5.03045261e-01 -2.62160635e+00 -1.86783642e-01 3.56075168e-01
-1.28692612e-01 5.28925955e-01 -3.18534225e-01 4.80781734e-01
-3.43215823e-01 -3.60979661e-02 -4.74442512e-01 1.61908984e-01
-5.76944768e-01 2.23503768e-01 -2.08698273e-01 4.94209707e-01
4.21139836e-01 7.57999539e-01 -9.50659633e-01 -4.84968662e-01
4.64032978e-01 2.37765118e-01 -8.71440247e-02 2.05990449e-01
-2.42030978e-01 8.26824248e-01 -2.61789888e-01 2.87504613e-01
4.54051018e-01 -5.26465356e-01 2.60521382e-01 1.22578301e-01
-4.96572591e-02 -1.02232955e-01 -8.68033528e-01 1.44228375e+00
-7.12108970e-01 8.77642393e-01 -2.11905926e-01 -1.49676418e+00
1.09660697e+00 1.29968137e-01 5.32604694e-01 -6.24709427e-01
-2.24158075e-02 -8.72349292e-02 7.17106760e-02 -5.66992581e-01
2.15177685e-01 -5.97007096e-01 1.50285348e-01 5.15667737e-01
-1.27472021e-02 -2.61295319e-01 2.95202106e-01 -1.13424063e-01
5.21631002e-01 2.41817608e-02 5.13764560e-01 -3.10224086e-01
5.74883401e-01 5.23497522e-01 3.57618220e-02 5.93576193e-01
2.35709116e-01 4.53829437e-01 9.44778621e-02 -3.85608494e-01
-8.19989443e-01 -8.82221997e-01 -4.71822828e-01 1.25883710e+00
6.56201690e-02 -3.28443907e-02 -4.03529167e-01 -8.09215724e-01
-4.74506207e-02 3.91407311e-01 -4.47956532e-01 1.17486216e-01
-7.05761686e-02 -1.00185668e+00 1.95410639e-01 4.81427580e-01
5.72598338e-01 -1.30270267e+00 -7.24510193e-01 1.65081769e-01
5.27255982e-02 -1.16437066e+00 2.92825669e-01 4.49329495e-01
-1.33859754e+00 -1.06725097e+00 -9.71015036e-01 -7.40102470e-01
9.54575956e-01 6.62748456e-01 8.18170190e-01 5.33538945e-02
-4.01550412e-01 3.22617054e-01 -6.55491173e-01 -4.50892955e-01
-4.70661700e-01 2.57097274e-01 -3.12040508e-01 2.88358450e-01
9.52978060e-02 -5.80660284e-01 -3.08624148e-01 3.77859026e-01
-1.23262584e+00 -6.46895319e-02 7.85560668e-01 8.43070149e-01
4.90717828e-01 5.76101065e-01 6.05414689e-01 -1.38174033e+00
1.30334139e-01 -2.89378792e-01 -5.34955502e-01 1.93137482e-01
-5.79487801e-01 2.09054761e-02 6.34061277e-01 -1.98055059e-01
-1.10480225e+00 4.53374267e-01 -1.76820194e-03 1.17182657e-01
-6.65499270e-01 9.08012867e-01 1.12722091e-01 -4.08927090e-02
1.10873938e+00 1.65590137e-01 1.09893478e-01 -6.03934526e-01
2.59787828e-01 9.16658700e-01 1.91801742e-01 -1.55040607e-01
8.77954900e-01 7.59493947e-01 6.25170246e-02 -1.44034469e+00
-1.04767358e+00 -8.38815153e-01 -8.31358373e-01 -1.90193877e-01
8.96399081e-01 -1.00139916e+00 7.45088831e-02 2.42341489e-01
-4.62438524e-01 -6.21539950e-01 -4.02601331e-01 6.78899050e-01
-3.99051309e-01 4.20165330e-01 -1.34258389e-01 -8.48536491e-01
-1.12253353e-01 -8.01597595e-01 9.18336689e-01 1.23687789e-01
1.43384352e-01 -1.23640192e+00 -1.47286023e-03 4.07229215e-01
3.01809579e-01 -8.21660366e-03 9.61648107e-01 -6.72922790e-01
-2.07336739e-01 2.32115407e-02 -3.73250812e-01 7.37206995e-01
6.07114017e-01 -3.23460698e-01 -1.17098069e+00 -3.27425897e-01
-2.69593239e-01 -4.86138046e-01 1.27507949e+00 1.94516689e-01
1.11313009e+00 6.99315667e-02 -2.72942513e-01 2.87204534e-01
1.69529986e+00 -9.92911160e-02 5.55566669e-01 2.94812560e-01
5.46433508e-01 1.16173375e+00 9.77503896e-01 1.93425477e-01
-1.31929219e-01 4.34471577e-01 3.68936032e-01 -7.13278353e-01
1.07395224e-01 3.04329935e-02 -8.54578044e-04 9.88370925e-02
-2.58487374e-01 2.66716797e-02 -1.03912556e+00 6.02669775e-01
-1.73152196e+00 -8.97414684e-01 -1.94445148e-01 2.30927062e+00
6.61544919e-01 -1.87668294e-01 2.79148668e-01 5.31375706e-01
7.08899081e-01 2.37052053e-01 -3.21818292e-01 1.29155040e-01
-1.41320318e-01 3.23700130e-01 5.45608878e-01 3.20316225e-01
-1.46051323e+00 1.00766993e+00 6.08701992e+00 5.97351611e-01
-1.59567189e+00 1.57632127e-01 8.15063000e-01 2.86538780e-01
2.81263161e-02 -8.09258595e-02 -4.01001930e-01 1.65985480e-01
5.16060948e-01 2.78411001e-01 -1.54006537e-02 7.90469110e-01
5.33359528e-01 -5.44169962e-01 -1.03791213e+00 8.50202501e-01
-1.85060024e-01 -1.05825543e+00 3.90775390e-02 1.01124741e-01
9.59835172e-01 1.39539018e-01 -8.97375122e-02 -2.41702851e-02
6.70333803e-02 -1.19796443e+00 5.16939536e-02 8.18965137e-02
7.29848742e-01 -5.62692523e-01 7.99574316e-01 8.00606668e-01
-7.71698594e-01 -1.34680718e-01 -5.24129808e-01 -2.13653684e-01
-1.18394941e-01 9.94496822e-01 -9.54841971e-01 5.94696522e-01
4.87850159e-01 9.57336664e-01 -7.65785575e-01 1.04141474e+00
-3.98901910e-01 7.67525613e-01 -1.63895696e-01 1.75615057e-01
3.04510176e-01 -3.61975104e-01 1.12955034e-01 1.23751664e+00
-2.79362388e-02 1.09140784e-01 3.27137679e-01 2.96881616e-01
2.77612746e-01 4.39412594e-01 -8.36750567e-01 -2.09785894e-01
-1.62849799e-01 1.23090780e+00 -9.55272853e-01 -3.60117227e-01
-2.81167597e-01 7.97391593e-01 2.34912619e-01 5.15217960e-01
-1.94285080e-01 -7.08112866e-02 5.03331460e-02 3.98216814e-01
4.15466756e-01 -2.57157683e-01 -2.51550853e-01 -1.17428553e+00
-3.14614892e-01 -7.62815177e-01 5.19127965e-01 -8.11226606e-01
-1.27542424e+00 5.50226092e-01 1.82953358e-01 -1.47930026e+00
-3.48782778e-01 -6.66069925e-01 -5.31152070e-01 5.08184314e-01
-1.93190885e+00 -1.15161693e+00 -4.68976080e-01 5.58682382e-01
4.32116151e-01 -2.40155101e-01 8.44031155e-01 6.60415068e-02
-8.82632136e-02 5.33817336e-02 3.85293007e-01 2.81988144e-01
6.50253057e-01 -1.07685363e+00 -2.07704633e-01 8.58394802e-01
7.87713468e-01 1.51582062e-01 5.19889414e-01 -3.03001910e-01
-8.61061990e-01 -1.42692637e+00 6.42622232e-01 2.22002625e-01
3.13924491e-01 1.86114740e-02 -8.78397703e-01 4.60190386e-01
-5.73430471e-02 1.33727267e-01 8.80782843e-01 1.23738870e-01
-2.66681463e-01 -4.40247566e-01 -1.02296400e+00 1.28339857e-01
6.75791979e-01 -5.31925499e-01 -3.67834240e-01 7.81840920e-01
-8.23993236e-02 1.31054506e-01 -7.01916099e-01 4.32715803e-01
1.23646870e-01 -8.61831486e-01 8.67190957e-01 -4.80864316e-01
6.11681044e-01 -3.51838648e-01 6.00714721e-02 -1.50397909e+00
-1.99391320e-01 7.04287663e-02 7.83503711e-01 8.55238795e-01
5.53901613e-01 -4.71310347e-01 8.59716356e-01 -2.60460466e-01
3.58852595e-01 -2.50993222e-01 -4.47621793e-01 -9.08290088e-01
5.76163307e-02 -1.98355436e-01 -9.01532173e-03 1.13325894e+00
-1.78798258e-01 6.60079479e-01 -3.54865015e-01 2.09905520e-01
6.12927318e-01 4.74255055e-01 8.26702178e-01 -1.51314950e+00
-4.10815358e-01 -4.22342122e-02 -4.78885442e-01 -8.04722786e-01
2.51377136e-01 -1.00519884e+00 2.53684103e-01 -1.69723690e+00
2.14245871e-01 -6.47030354e-01 -1.08744040e-01 7.62575150e-01
-1.42091334e-01 7.04465449e-01 3.27434182e-01 3.98207128e-01
-2.86463469e-01 2.15959698e-01 1.15961421e+00 -4.15316522e-01
-3.37074578e-01 3.38413686e-01 -4.54256684e-01 5.67301452e-01
9.95074272e-01 -5.21174908e-01 -5.02286851e-01 -4.17885184e-01
-1.08171053e-01 -7.13927448e-02 4.46905762e-01 -9.24689651e-01
-1.22795314e-01 -2.75018215e-01 3.25301588e-01 -3.15614998e-01
1.09561794e-01 -1.08181441e+00 -5.10120690e-02 5.35356760e-01
-3.66264373e-01 -4.75184381e-01 1.24583736e-01 5.06634414e-01
-4.71826851e-01 -5.92835844e-01 9.54751313e-01 -5.06878316e-01
-1.05956626e+00 -5.16355829e-03 -6.03876710e-01 -2.57683247e-01
1.01707315e+00 -1.07241951e-01 4.87050600e-02 -5.28628707e-01
-1.09686351e+00 1.95156578e-02 4.12548751e-01 7.76886791e-02
4.83234435e-01 -8.08701873e-01 -8.91399443e-01 2.72077233e-01
5.11316776e-01 7.56120160e-02 6.11299388e-02 6.46517456e-01
-5.12012839e-01 2.31800333e-01 -3.27226758e-01 -9.65405703e-01
-1.51988626e+00 4.17522013e-01 8.83965120e-02 -3.68414670e-01
-4.17905062e-01 3.08598578e-01 2.82390475e-01 -4.38791245e-01
8.61431211e-02 -5.63426390e-02 -4.99418348e-01 2.23818824e-01
4.22600538e-01 1.90732241e-01 2.44336814e-01 -7.06554592e-01
-3.07252377e-01 5.45038283e-01 8.43651295e-02 -2.59389579e-01
1.71282756e+00 1.28398627e-01 -9.68191475e-02 6.03632450e-01
1.07078290e+00 -1.45172387e-01 -1.01785183e+00 -4.43693668e-01
9.33366045e-02 -3.16146165e-01 2.99695462e-01 -6.14430308e-01
-1.15000653e+00 1.15899909e+00 7.04646349e-01 2.29287162e-01
1.31706595e+00 1.95971861e-01 -9.06535760e-02 7.63785660e-01
3.75399023e-01 -8.24555755e-01 7.81062320e-02 3.83787900e-01
5.50181925e-01 -1.67900431e+00 2.90306836e-01 -5.62295735e-01
-6.55667305e-01 1.04197717e+00 1.87634155e-01 -1.35297269e-01
7.48548210e-01 -2.29833439e-01 5.03747702e-01 -1.11178815e-01
-2.64171064e-01 -8.31218660e-01 4.65938210e-01 7.68477559e-01
5.73722661e-01 4.76524904e-02 -2.66595691e-01 -2.49637350e-01
7.38962963e-02 -2.76158261e-03 4.98671323e-01 1.01881611e+00
-5.33311188e-01 -1.23670554e+00 -4.05400515e-01 6.79980099e-01
-3.77406567e-01 -1.22946545e-01 -4.99926180e-01 7.00199902e-01
1.05344087e-01 1.12003458e+00 2.00328544e-01 -4.05778065e-02
-1.67002678e-02 -1.07889421e-01 4.26681280e-01 -1.13166940e+00
-5.65107346e-01 1.50246829e-01 1.16505712e-01 3.43221799e-02
-1.17575407e+00 -5.46448588e-01 -9.89643157e-01 1.78944021e-01
-4.04711783e-01 3.83998990e-01 6.70698881e-01 1.14849997e+00
-2.28136014e-02 1.45872623e-01 8.83507192e-01 -7.83204019e-01
-3.94973308e-01 -1.12575805e+00 -7.90055156e-01 5.02234995e-01
3.16376925e-01 -3.47505152e-01 -3.19884926e-01 4.59135979e-01] | [9.599213600158691, -1.3250246047973633] |
44be7d4c-5b37-4919-96ff-aab13838fd92 | kvl-bert-knowledge-enhanced-visual-and | 2012.07 | null | https://arxiv.org/abs/2012.07000v1 | https://arxiv.org/pdf/2012.07000v1.pdf | KVL-BERT: Knowledge Enhanced Visual-and-Linguistic BERT for Visual Commonsense Reasoning | Reasoning is a critical ability towards complete visual understanding. To develop machine with cognition-level visual understanding and reasoning abilities, the visual commonsense reasoning (VCR) task has been introduced. In VCR, given a challenging question about an image, a machine must answer correctly and then provide a rationale justifying its answer. The methods adopting the powerful BERT model as the backbone for learning joint representation of image content and natural language have shown promising improvements on VCR. However, none of the existing methods have utilized commonsense knowledge in visual commonsense reasoning, which we believe will be greatly helpful in this task. With the support of commonsense knowledge, complex questions even if the required information is not depicted in the image can be answered with cognitive reasoning. Therefore, we incorporate commonsense knowledge into the cross-modal BERT, and propose a novel Knowledge Enhanced Visual-and-Linguistic BERT (KVL-BERT for short) model. Besides taking visual and linguistic contents as input, external commonsense knowledge extracted from ConceptNet is integrated into the multi-layer Transformer. In order to reserve the structural information and semantic representation of the original sentence, we propose using relative position embedding and mask-self-attention to weaken the effect between the injected commonsense knowledge and other unrelated components in the input sequence. Compared to other task-specific models and general task-agnostic pre-training models, our KVL-BERT outperforms them by a large margin. | ['Lejian Liao', 'Sicheng Yang', 'Zhanchen Sun', 'Siyi Ma', 'Dandan song'] | 2020-12-13 | null | null | null | null | ['visual-commonsense-reasoning'] | ['reasoning'] | [ 1.90402165e-01 3.04398358e-01 -3.03880684e-02 -2.14735076e-01
-1.41603172e-01 -5.36936522e-01 7.71728754e-01 -1.21528260e-01
-4.15007502e-01 6.17254615e-01 4.14155692e-01 -6.90226555e-01
1.12596035e-01 -8.68571222e-01 -7.60541737e-01 -3.23543191e-01
5.85383892e-01 8.22067335e-02 4.76621002e-01 -4.96742100e-01
3.62031609e-01 3.02487075e-01 -1.46220064e+00 8.74634683e-01
1.09773660e+00 8.56566191e-01 6.82796597e-01 4.05336410e-01
-5.91228783e-01 1.76944280e+00 -5.88808894e-01 -7.17431068e-01
-4.61356975e-02 -7.88347363e-01 -1.12475526e+00 -9.32159051e-02
3.18454921e-01 -3.84803653e-01 -4.64150488e-01 1.28058767e+00
2.33078271e-01 3.34464073e-01 7.59322464e-01 -1.32611811e+00
-1.62949240e+00 5.82038462e-01 -5.37100792e-01 5.54388225e-01
5.38913488e-01 5.20288527e-01 9.28400159e-01 -9.62893486e-01
6.03297174e-01 1.49418569e+00 1.49010226e-01 7.66008556e-01
-1.09541738e+00 -6.34600639e-01 4.38098609e-01 1.03151739e+00
-1.17420948e+00 -1.14933379e-01 1.32195199e+00 -3.31357092e-01
1.07211769e+00 3.11705500e-01 1.08524978e+00 1.10379159e+00
9.05642733e-02 1.11546731e+00 1.51884949e+00 -4.89668339e-01
-5.67024015e-03 4.97748107e-01 1.78844612e-02 9.04284537e-01
1.29201887e-02 7.29469433e-02 -4.76658255e-01 4.52980191e-01
8.27045441e-01 1.10102624e-01 -3.90341580e-01 -2.66223609e-01
-1.32364047e+00 7.95390666e-01 1.18362212e+00 4.30337459e-01
-3.58301669e-01 2.18055934e-01 4.15242642e-01 2.39263818e-01
-3.17362770e-02 2.88962901e-01 1.05648907e-02 2.98816770e-01
-7.01230109e-01 -1.24314114e-01 3.05235147e-01 7.77161300e-01
7.07776904e-01 2.71223813e-01 -5.59487581e-01 6.32558227e-01
4.33116347e-01 5.57160139e-01 6.18333399e-01 -8.31367135e-01
4.95373756e-01 1.07266927e+00 -3.01838249e-01 -1.24730921e+00
-1.58543989e-01 -2.60701001e-01 -7.89967358e-01 4.18947458e-01
1.19394988e-01 3.53729516e-01 -1.06321096e+00 1.81007421e+00
2.97026366e-01 -7.13200644e-02 3.12690675e-01 1.19651246e+00
1.34925175e+00 5.86786747e-01 3.45175862e-01 -7.45693520e-02
1.69730914e+00 -9.76035416e-01 -8.53732169e-01 -5.95846653e-01
2.97006637e-01 -4.05550629e-01 1.76172566e+00 -6.20538509e-03
-8.60678852e-01 -7.19621539e-01 -1.13810050e+00 -7.61618376e-01
-8.18261385e-01 -1.47539238e-02 6.68069899e-01 2.96080947e-01
-1.05226624e+00 4.13572155e-02 -1.28507867e-01 -9.18163583e-02
7.87871778e-01 -4.67341810e-01 -2.63727874e-01 -5.17611444e-01
-1.56199193e+00 1.34642637e+00 5.90206325e-01 2.48174369e-01
-8.95609796e-01 -5.37468076e-01 -1.04897249e+00 1.38629645e-01
6.05538964e-01 -1.00716174e+00 8.32117617e-01 -1.27721584e+00
-1.13831472e+00 1.18567765e+00 -1.81794927e-01 -2.19843641e-01
4.88575757e-01 -1.29105091e-01 -4.32750791e-01 5.16742766e-01
2.38117293e-01 8.41132939e-01 9.53913808e-01 -1.54233551e+00
-1.59481347e-01 -2.84788787e-01 6.50800526e-01 2.91336387e-01
-1.00999385e-01 -1.76172778e-01 -3.04025590e-01 -7.52542973e-01
1.02490574e-01 -4.28398907e-01 2.17689767e-01 3.00299376e-01
-3.46017092e-01 -3.27019721e-01 7.52568007e-01 -9.63985920e-01
8.69082928e-01 -2.15227461e+00 1.96209401e-01 -7.00438172e-02
3.70898962e-01 1.82050899e-01 -1.72892466e-01 1.88745543e-01
-1.49860397e-01 3.02083381e-02 -1.27663955e-01 2.15513811e-01
1.81630239e-01 3.96978050e-01 -6.20655656e-01 1.46703392e-01
3.33673030e-01 1.59271729e+00 -1.11926508e+00 -8.39111447e-01
3.36465210e-01 4.35027599e-01 -3.89020413e-01 1.38445109e-01
-4.65875149e-01 1.25072390e-01 -1.72318801e-01 5.37662089e-01
6.72455788e-01 -5.43637097e-01 -9.92996711e-03 -5.34860134e-01
2.23873168e-01 3.62212360e-02 -7.16868877e-01 1.47263610e+00
-5.69607556e-01 8.29969883e-01 -3.66118371e-01 -1.11152732e+00
6.66198313e-01 4.50927299e-03 -3.94033879e-01 -1.22598219e+00
1.68259338e-01 -2.82877862e-01 2.39700422e-01 -8.35370183e-01
5.84700555e-02 -7.28567839e-01 1.49898902e-01 2.88201928e-01
-5.00962362e-02 -3.03174287e-01 1.34179056e-01 6.72594965e-01
6.95754588e-01 2.59973526e-01 5.99166691e-01 2.30937079e-03
8.89975071e-01 1.64820567e-01 1.60006344e-01 5.63543200e-01
-5.33989310e-01 2.03748986e-01 5.74793816e-01 -2.31159076e-01
-7.06209064e-01 -1.27831376e+00 2.92232245e-01 1.23923612e+00
5.17114460e-01 -1.65955313e-02 -3.84989172e-01 -7.54559278e-01
4.44989875e-02 1.25634122e+00 -9.97621715e-01 -4.16534722e-01
-2.47696310e-01 -1.25223741e-01 3.33517134e-01 7.23458707e-01
1.05064571e+00 -1.63869631e+00 -8.11558783e-01 -1.16995804e-01
-4.50904906e-01 -1.36057115e+00 -1.99023932e-01 -5.73113933e-02
-4.76016760e-01 -1.24304509e+00 -5.80919027e-01 -8.84005427e-01
5.92759132e-01 5.46248317e-01 1.08941901e+00 3.29430997e-01
-5.00410318e-01 5.05096197e-01 -4.45256293e-01 -3.45069468e-01
-2.52298355e-01 -8.46694469e-01 -5.56058466e-01 -1.39080018e-01
4.29222107e-01 -2.87089854e-01 -6.57867551e-01 -1.66321531e-01
-1.03676546e+00 5.24584711e-01 6.80821836e-01 8.42205882e-01
2.88780361e-01 -8.68967101e-02 6.25499427e-01 -5.11676669e-01
7.16745377e-01 -3.03799748e-01 -2.10952461e-02 4.63324964e-01
-2.61187822e-01 1.11391202e-01 6.70777082e-01 -5.47323167e-01
-1.38069522e+00 -4.32508647e-01 -5.72490543e-02 -7.23279655e-01
2.81692408e-02 5.37374735e-01 -3.53827119e-01 7.33254477e-02
5.78307986e-01 7.22709715e-01 -3.32154296e-02 1.28539175e-01
9.38847482e-01 1.53431550e-01 6.11658454e-01 -4.48285341e-01
7.77592480e-01 7.09985077e-01 -1.33912295e-01 -5.67077518e-01
-1.32605040e+00 -3.33309770e-01 -5.24190843e-01 -4.40621793e-01
1.22110605e+00 -9.60963786e-01 -8.31728041e-01 -9.60401893e-02
-1.47513485e+00 -2.06203967e-01 -4.00027215e-01 1.96841091e-01
-6.20865703e-01 5.92597604e-01 -3.35172027e-01 -8.63427103e-01
-1.01093069e-01 -8.65542531e-01 5.83763957e-01 1.50560766e-01
-2.51124911e-02 -9.54513550e-01 -3.92965108e-01 7.00984836e-01
3.41510713e-01 2.24600062e-01 1.34931588e+00 -4.28357452e-01
-6.49612904e-01 2.19007626e-01 -1.07285655e+00 5.48386931e-01
-1.25102341e-01 -3.51672828e-01 -1.01413739e+00 1.97795764e-01
2.53913224e-01 -7.25797176e-01 1.23848081e+00 6.10269681e-02
1.18507266e+00 -3.47002387e-01 -3.75807472e-02 3.28798056e-01
1.50744104e+00 8.15069824e-02 9.16721344e-01 4.50001150e-01
8.38611782e-01 7.05135822e-01 3.99097919e-01 6.78116605e-02
9.02941883e-01 1.83085293e-01 6.30415082e-01 -1.67540595e-01
-6.22785509e-01 -4.23959196e-01 4.49976861e-01 4.29527164e-01
-2.29728013e-01 1.29932910e-01 -9.29867029e-01 6.25609100e-01
-1.66512692e+00 -1.57553959e+00 -8.39529857e-02 1.41292381e+00
1.08967340e+00 2.12712921e-02 -2.53403395e-01 1.59286737e-01
5.45527697e-01 3.60045642e-01 -6.96041584e-01 -4.12513673e-01
-4.14334416e-01 -1.63747460e-01 -1.93951443e-01 4.11906660e-01
-7.06286907e-01 1.32106185e+00 5.25718784e+00 8.42005372e-01
-9.61012006e-01 2.17785537e-01 2.36538827e-01 1.15598992e-01
-7.45039821e-01 -3.47689986e-02 -2.29387030e-01 1.34703457e-01
2.29171246e-01 -1.69174954e-01 5.86718619e-01 6.77162826e-01
-3.70059386e-02 -2.96818793e-01 -9.37962949e-01 1.20977807e+00
6.35521531e-01 -1.33286285e+00 6.15943551e-01 -3.21210951e-01
4.43111867e-01 -4.98215765e-01 2.14227751e-01 8.26315284e-01
2.26908684e-01 -1.20495880e+00 8.92312706e-01 6.52049541e-01
7.53817260e-01 -5.43922484e-01 5.96378982e-01 4.69279796e-01
-1.12288570e+00 -3.04302871e-01 -4.99501050e-01 -2.75400519e-01
1.87705576e-01 4.82387990e-01 -5.95126271e-01 4.21810120e-01
6.59117937e-01 5.35496056e-01 -9.19919133e-01 4.57358003e-01
-9.34350908e-01 2.60045588e-01 2.93431789e-01 -2.05334917e-01
2.68028200e-01 1.62186489e-01 3.25000226e-01 1.09677076e+00
-1.49865881e-01 4.25579846e-01 -8.91375691e-02 1.45317960e+00
-3.55322547e-02 9.97933373e-02 -6.13302827e-01 -1.68803379e-01
1.22326262e-01 1.03313005e+00 -7.20304489e-01 -7.08924174e-01
-5.55521548e-01 1.32988834e+00 7.62672484e-01 6.28982723e-01
-1.02129948e+00 -3.36082965e-01 1.29035279e-01 -9.97484382e-03
3.07884306e-01 3.06701567e-02 -2.89785594e-01 -1.27478254e+00
9.36005786e-02 -7.39870131e-01 4.85911518e-01 -1.62527740e+00
-1.33095396e+00 4.99352038e-01 8.62098578e-03 -9.00239289e-01
8.78322963e-03 -1.04713392e+00 -6.09640896e-01 8.66053164e-01
-2.03911829e+00 -1.55846965e+00 -5.51874399e-01 9.77113485e-01
6.26076519e-01 -1.57163776e-02 4.94052321e-01 -2.90718615e-01
-1.01202786e-01 2.06000417e-01 -6.23636544e-01 2.58737683e-01
4.71794456e-01 -1.29133534e+00 -1.64440528e-01 8.24901819e-01
1.41441002e-01 8.32526445e-01 6.46729767e-01 -7.78347969e-01
-1.19179451e+00 -8.25106800e-01 7.12652266e-01 -6.06342971e-01
8.02461803e-01 -3.23741466e-01 -1.16724932e+00 6.14713132e-01
5.40485859e-01 1.34070545e-01 7.58802295e-01 -1.00238807e-01
-9.91993368e-01 2.16586560e-01 -1.09776783e+00 9.38099742e-01
1.18076825e+00 -9.83838022e-01 -1.59558952e+00 1.80586457e-01
1.02831399e+00 -4.45591360e-02 -2.70658761e-01 1.28831699e-01
3.58983994e-01 -1.04196489e+00 1.22617590e+00 -8.49923730e-01
8.83544683e-01 -4.63804036e-01 -3.95020902e-01 -1.18104327e+00
-5.06613135e-01 1.49402097e-01 -1.84748694e-01 1.01962173e+00
1.56931430e-01 -2.86869854e-01 3.10729772e-01 3.71782064e-01
1.10102713e-01 -4.52461392e-01 -8.13441753e-01 -7.06575274e-01
5.01250811e-02 -5.75417638e-01 4.21474546e-01 1.07301044e+00
3.28730822e-01 8.02131832e-01 -2.15655267e-01 6.31466806e-02
4.87105161e-01 3.90088588e-01 5.13110399e-01 -8.83047581e-01
-1.05384052e-01 -4.86928374e-01 -4.25072253e-01 -8.76163721e-01
5.53491354e-01 -1.15772235e+00 -1.49598226e-01 -2.13866138e+00
6.64844990e-01 2.85655826e-01 -3.61577958e-01 5.93600988e-01
-5.72424948e-01 2.41749525e-01 6.29035056e-01 1.84001878e-03
-6.63906395e-01 7.02156842e-01 1.97109318e+00 -3.81072998e-01
1.56115875e-01 -8.09664786e-01 -9.86583829e-01 8.58943939e-01
6.24058068e-01 1.03255987e-01 -7.22119451e-01 -3.08564693e-01
4.07601446e-01 -2.08288413e-02 1.16747391e+00 -6.16029859e-01
1.73590064e-01 -4.60803539e-01 8.04365814e-01 -6.67930663e-01
3.53011668e-01 -9.72817659e-01 -3.83620113e-01 4.98647451e-01
-3.41427535e-01 2.04459783e-02 3.06434900e-01 5.87088168e-01
-2.81330287e-01 -1.10036023e-01 8.09814274e-01 -4.97180641e-01
-1.38742948e+00 -1.86027288e-01 -2.91815270e-02 3.80119860e-01
1.19950914e+00 -4.62623239e-01 -7.62017846e-01 -3.08973283e-01
-7.71098435e-01 1.67547002e-01 2.02231973e-01 4.62830424e-01
1.24158573e+00 -1.45355940e+00 -5.16892135e-01 -3.09280064e-02
5.26502013e-01 -3.71655852e-01 7.94290960e-01 7.07414508e-01
-2.96279699e-01 4.21884000e-01 -4.28388864e-01 -2.47729808e-01
-9.93211627e-01 1.32603502e+00 1.93905979e-01 1.44314095e-01
-7.62639523e-01 9.50306833e-01 6.81963146e-01 -2.14048728e-01
-5.64644374e-02 -3.26831222e-01 -5.17749131e-01 5.95557764e-02
7.86426246e-01 7.36002848e-02 -5.14576614e-01 -7.30925262e-01
-5.60626447e-01 3.59912544e-01 6.76342249e-02 -1.59327015e-01
9.15018320e-01 -2.93904901e-01 -2.97156811e-01 6.40246809e-01
9.92418587e-01 -9.90341082e-02 -1.01064813e+00 -2.45119929e-01
-3.08259070e-01 -5.21471322e-01 2.29041763e-02 -1.24478531e+00
-8.44746232e-01 1.35653639e+00 2.56755501e-01 -1.29992096e-02
1.11840951e+00 4.62574184e-01 1.78775176e-01 2.28152201e-01
8.94833952e-02 -9.47860181e-01 7.30475068e-01 5.01505792e-01
1.54674435e+00 -1.46308732e+00 5.56292124e-02 -4.75312650e-01
-1.18351793e+00 9.46974576e-01 1.02876568e+00 -1.95207596e-02
2.78268218e-01 -1.90043077e-01 1.72736436e-01 -4.41787094e-01
-7.68738091e-01 -6.68437839e-01 5.47312379e-01 9.62133586e-01
2.86877394e-01 -4.77249883e-02 -4.50692028e-02 6.85261130e-01
-1.87979937e-01 7.05451146e-03 2.91821539e-01 6.84528351e-01
-7.01812804e-01 -1.45574033e-01 -3.91834140e-01 1.90823048e-01
5.62146008e-02 -4.89638031e-01 -4.97476339e-01 9.09927666e-01
3.06372613e-01 7.48665154e-01 -1.15623027e-01 -1.09512299e-01
3.80645603e-01 2.61288881e-01 7.13982522e-01 -4.69774455e-01
-1.88206762e-01 -4.36500996e-01 -1.11331560e-01 -5.43384552e-01
-5.96204400e-01 -1.59036919e-01 -1.60986304e+00 -1.70719385e-01
5.49864061e-02 -3.07910591e-01 1.68308780e-01 1.20732987e+00
-5.83093390e-02 9.46003675e-01 -6.20851070e-02 -3.83762002e-01
-4.02430773e-01 -7.76067615e-01 -4.50933099e-01 7.10253358e-01
3.96207511e-01 -8.86900783e-01 -4.27515537e-01 1.95253611e-01] | [10.795771598815918, 1.7332181930541992] |
5d2acf88-af97-42e4-aa99-350328046c67 | end-to-end-natural-language-understanding | 2107.05541 | null | https://arxiv.org/abs/2107.05541v6 | https://arxiv.org/pdf/2107.05541v6.pdf | End-to-End Natural Language Understanding Pipeline for Bangla Conversational Agents | Chatbots are intelligent software built to be used as a replacement for human interaction. Existing studies typically do not provide enough support for low-resource languages like Bangla. Due to the increasing popularity of social media, we can also see the rise of interactions in Bangla transliteration (mostly in English) among the native Bangla speakers. In this paper, we propose a novel approach to build a Bangla chatbot aimed to be used as a business assistant which can communicate in low-resource languages like Bangla and Bangla Transliteration in English with high confidence consistently. Since annotated data was not available for this purpose, we had to work on the whole machine learning life cycle (data preparation, machine learning modeling, and model deployment) using Rasa Open Source Framework, fastText embeddings, Polyglot embeddings, Flask, and other systems as building blocks. While working with the skewed annotated dataset, we try out different components and pipelines to evaluate which works best and provide possible reasoning behind the observed results. Finally, we present a pipeline for intent classification and entity extraction which achieves reasonable performance (accuracy: 83.02%, precision: 80.82%, recall: 83.02%, F1-score: 80%). | ['MD Abdullah Al Nasim', 'Mohammad Sabik Irbaz', 'Mueeze Al Mushabbir', 'Fahim Shahriar Khan'] | 2021-07-12 | null | null | null | null | ['transliteration'] | ['natural-language-processing'] | [-6.86359406e-01 4.18690383e-01 2.95798063e-01 -4.21380430e-01
-3.28226328e-01 -5.31261683e-01 8.65405798e-01 7.65409097e-02
-5.29704988e-01 6.33331358e-01 1.44794032e-01 -7.41625249e-01
1.23417921e-01 -5.88508546e-01 -2.16278419e-01 -3.30987960e-01
3.24405968e-01 9.13023174e-01 8.86602998e-02 -4.68066424e-01
1.54036820e-01 2.21709847e-01 -9.23449278e-01 3.84013772e-01
8.30131352e-01 1.34028435e-01 4.40959215e-01 9.20370936e-01
-2.61645555e-01 1.40515172e+00 -7.62129664e-01 -1.06322443e+00
-1.80614311e-02 -3.03156644e-01 -1.41848528e+00 -2.71001160e-01
-2.82605320e-01 -6.77612185e-01 -4.40716714e-01 5.59387863e-01
3.40998828e-01 -1.00129351e-01 6.66173041e-01 -1.35063469e+00
-8.10306132e-01 1.01361275e+00 -6.71821088e-02 -7.61274546e-02
4.03034151e-01 1.42619655e-01 9.70328689e-01 -6.98810935e-01
9.64338243e-01 1.06577885e+00 8.00860345e-01 7.27587700e-01
-7.44764745e-01 -3.66011798e-01 -4.67060387e-01 2.02579558e-01
-9.76500988e-01 -3.45267981e-01 4.77295518e-01 -4.39139932e-01
1.47887552e+00 1.90865830e-01 4.40356016e-01 1.52574801e+00
3.24444585e-02 7.95565605e-01 1.03609002e+00 -7.32059062e-01
-2.03154087e-01 1.11393499e+00 3.81358147e-01 7.51099110e-01
7.45731816e-02 -4.98114586e-01 -3.61101419e-01 -3.59142125e-02
3.74439687e-01 -1.48385406e-01 2.87977427e-01 3.58502924e-01
-1.07325959e+00 9.81129706e-01 1.07230559e-01 7.94729114e-01
-4.22334224e-01 -1.50879979e-01 6.67775512e-01 6.23642266e-01
3.41034532e-01 6.12493515e-01 -7.05664098e-01 -1.00284684e+00
-2.97394276e-01 1.80072606e-01 1.51317501e+00 1.34399176e+00
4.33374077e-01 -2.50988156e-01 7.93683622e-03 1.15908754e+00
4.72232431e-01 -6.76881894e-02 6.44198477e-01 -6.59371853e-01
5.35185218e-01 8.92643869e-01 1.34133905e-01 -7.53210247e-01
-4.00494367e-01 1.57757908e-01 -4.37684447e-01 -2.37078860e-01
8.23193789e-01 -5.79047740e-01 -2.62291670e-01 1.09654379e+00
3.03635836e-01 -5.17018080e-01 3.27582449e-01 5.35067618e-01
9.52697515e-01 6.86327279e-01 -3.07782609e-02 5.94525933e-02
1.55922461e+00 -1.34507060e+00 -9.34245288e-01 3.87610570e-02
1.26617455e+00 -1.15620983e+00 1.18178296e+00 2.36475334e-01
-7.81003952e-01 -3.48471224e-01 -8.00179541e-01 -3.81456107e-01
-6.53107285e-01 3.51456285e-01 8.12604308e-01 9.64247406e-01
-8.66472065e-01 5.05740404e-01 -9.11906004e-01 -1.01733637e+00
1.88325688e-01 2.63849378e-01 -6.68319046e-01 2.07435135e-02
-9.96892989e-01 1.49199474e+00 2.79107749e-01 -6.03766888e-02
-5.44601023e-01 -4.23898876e-01 -6.54904425e-01 -3.16372067e-01
2.01316196e-02 -7.23067746e-02 1.40767527e+00 -5.72064281e-01
-1.61789691e+00 9.89249349e-01 1.31619498e-01 -5.62083423e-01
7.12663114e-01 -2.29723394e-01 -3.23467851e-01 -3.53974313e-01
-8.00704360e-02 2.24242717e-01 9.44618434e-02 -7.27590263e-01
-7.89601207e-01 -2.91764766e-01 4.15360004e-01 -1.33534530e-02
-5.57207644e-01 6.00266337e-01 -2.34186321e-01 -1.70608133e-01
-5.00129819e-01 -8.75692844e-01 4.74545434e-02 -5.32521963e-01
-2.01580495e-01 -8.02401304e-01 8.38220775e-01 -1.18457592e+00
1.31752896e+00 -1.92892325e+00 -2.97212869e-01 -3.43482137e-01
1.29955724e-01 4.57899153e-01 1.43718228e-01 9.89262223e-01
4.14597809e-01 1.73306420e-01 2.48897269e-01 -3.96211386e-01
2.44383499e-01 2.81767994e-01 2.76168734e-01 2.99462795e-01
3.49567205e-01 8.43754292e-01 -8.55316281e-01 -5.90911567e-01
3.03998321e-01 5.84119678e-01 -5.36092639e-01 5.35539508e-01
1.69405460e-01 1.08810470e-01 -3.64159077e-01 5.97657681e-01
3.32576215e-01 1.36176825e-01 4.44112867e-01 3.47916745e-02
-5.68396151e-01 6.75705433e-01 -8.49510491e-01 1.31503415e+00
-1.01830804e+00 1.04520643e+00 1.09730937e-01 -8.94657254e-01
1.17209721e+00 5.80131769e-01 2.87972033e-01 -2.89570808e-01
5.10886312e-01 4.27340955e-01 4.06028777e-01 -1.03976166e+00
6.62571371e-01 2.64984190e-01 -9.16172341e-02 5.30922294e-01
4.82180148e-01 -1.14497297e-01 6.31495193e-02 2.12682217e-01
1.36823070e+00 5.21103330e-02 4.56407040e-01 -9.47220400e-02
4.00564551e-01 3.16727042e-01 2.23404840e-02 5.15765786e-01
-4.71860796e-01 1.77204177e-01 7.25969076e-01 -4.44524169e-01
-1.24739993e+00 -1.44310266e-01 -1.34585410e-01 1.29949164e+00
-4.52541739e-01 -6.85895085e-01 -8.73163581e-01 -9.90204692e-01
-4.09966856e-01 7.74034142e-01 -4.55829591e-01 1.72762051e-01
-6.12631440e-01 -5.23679733e-01 8.31194699e-01 2.25719079e-01
5.26222765e-01 -1.34153783e+00 -3.33894700e-01 4.70926464e-01
-1.80469081e-01 -1.18836415e+00 -1.29448414e-01 5.17438412e-01
-4.44561094e-01 -9.19436395e-01 -3.72125685e-01 -1.08715045e+00
2.31139407e-01 -5.32503240e-02 9.36158180e-01 1.15078188e-01
-2.64554560e-01 1.44941747e-01 -6.83578789e-01 -6.25509143e-01
-1.01882780e+00 3.80513012e-01 -5.37318513e-02 -4.81324762e-01
8.20453942e-01 -3.59901696e-01 -7.34659135e-02 1.43687725e-01
-4.37982917e-01 2.04991549e-03 6.48971200e-01 9.54391897e-01
-6.34675145e-01 -2.83523858e-01 4.96516645e-01 -1.27705061e+00
7.95962632e-01 -7.52923250e-01 -9.59845409e-02 2.34080166e-01
-3.90959859e-01 -1.18177995e-01 7.61293888e-01 -5.19740582e-01
-1.05074656e+00 -2.74753645e-02 -6.84345126e-01 2.06536680e-01
-3.72796685e-01 4.83433843e-01 -1.24299392e-01 2.71250188e-01
7.74051487e-01 -2.84077739e-03 5.57217486e-02 -6.62328482e-01
3.74627888e-01 1.81586289e+00 -9.03590322e-02 -2.66705662e-01
3.90065610e-01 -6.20291196e-02 -7.71054327e-01 -1.14136744e+00
-2.43335500e-01 -7.58975625e-01 -8.98370624e-01 -2.61509478e-01
9.87302959e-01 -5.76627612e-01 -1.02338099e+00 5.46341836e-01
-1.51755667e+00 -5.03926694e-01 1.86842486e-01 5.12974322e-01
-2.11629108e-01 3.66566591e-02 -7.63405204e-01 -1.13983583e+00
-4.15726870e-01 -9.85026717e-01 3.88526976e-01 1.61802426e-01
-7.50667870e-01 -1.18070030e+00 7.36151710e-02 9.15687859e-01
5.52976966e-01 -2.27626935e-01 8.19485843e-01 -1.40131581e+00
-5.55589311e-02 -3.21266711e-01 -2.21800968e-01 5.35124063e-01
3.09931695e-01 4.43269908e-01 -1.07503152e+00 1.39124781e-01
-1.69705063e-01 -4.49915469e-01 3.23903514e-03 -2.50571400e-01
3.32063645e-01 -5.81551433e-01 -8.42113122e-02 -5.75692691e-02
9.66515958e-01 5.23190737e-01 4.99636680e-01 5.77780604e-01
6.51233971e-01 8.57407749e-01 5.56706607e-01 5.25489390e-01
7.61222363e-01 6.27933323e-01 3.56774293e-02 4.44811136e-01
-1.96490232e-02 -2.66498119e-01 6.87789023e-01 1.38325667e+00
-8.04720223e-02 -8.83699730e-02 -1.20293653e+00 8.22258770e-01
-1.71055186e+00 -8.19382310e-01 -7.06848323e-01 1.60490751e+00
1.04943192e+00 9.75704715e-02 3.81258786e-01 2.32835710e-02
3.24441165e-01 -2.49518365e-01 1.12163015e-01 -1.11136305e+00
3.52910519e-01 8.82884488e-02 2.96564251e-01 5.26617527e-01
-9.89202797e-01 1.09749424e+00 5.45453453e+00 5.14972210e-01
-1.00184059e+00 4.35010493e-01 2.75835633e-01 5.37278354e-01
1.12932079e-01 -1.01087295e-01 -8.45837176e-01 4.78129983e-01
1.45487261e+00 -1.58883676e-01 4.97853458e-01 1.13161993e+00
1.30672738e-01 5.04723005e-02 -1.11868608e+00 7.96186686e-01
-4.25217375e-02 -1.28597426e+00 -5.08704185e-01 -1.62097126e-01
3.10285568e-01 1.92992777e-01 -4.65235919e-01 1.02531731e+00
7.03186810e-01 -9.25472796e-01 4.65715080e-01 5.34487702e-02
2.91478217e-01 -4.57378775e-01 1.22264898e+00 7.59899437e-01
-6.09367788e-01 3.09161413e-02 -2.57544816e-01 -4.42977428e-01
-2.61360198e-01 -2.19530165e-02 -1.66386688e+00 4.23668176e-01
7.57054150e-01 5.60891151e-01 -4.29592162e-01 5.41854560e-01
-1.12108633e-01 8.11580002e-01 -3.13294142e-01 -9.61338222e-01
4.03968722e-01 -2.09123850e-01 2.10324705e-01 1.56712687e+00
1.40146732e-01 -1.16860583e-01 -1.69606298e-01 4.93636638e-01
-1.98681042e-01 4.91200417e-01 -8.21913779e-01 -3.90123993e-01
4.70658869e-01 1.54099417e+00 -5.09238303e-01 -1.12755515e-01
-7.11792111e-01 1.16890907e+00 5.98655581e-01 -2.52306491e-01
-9.37380433e-01 -7.46796191e-01 5.04034877e-01 5.91590777e-02
-9.51690450e-02 -2.25029305e-01 -1.33515000e-01 -9.09105480e-01
3.07815801e-02 -1.15541327e+00 2.69108340e-02 -4.74169463e-01
-1.26021016e+00 8.05737019e-01 -2.79890418e-01 -7.43649960e-01
-2.46288985e-01 -6.97584569e-01 -6.90285563e-01 8.39263797e-01
-8.74645829e-01 -1.52032709e+00 -1.83748811e-01 2.88685799e-01
7.49470472e-01 -3.69965076e-01 1.14117122e+00 6.04058564e-01
-6.59206569e-01 6.49700642e-01 3.36008251e-01 5.49776673e-01
8.49503458e-01 -1.40760505e+00 2.41866648e-01 3.86129081e-01
2.60823131e-01 8.66503716e-01 7.05920696e-01 -3.13104391e-01
-1.55899262e+00 -9.04649615e-01 1.64963353e+00 -9.20996487e-01
1.13299346e+00 -7.16919661e-01 -8.32323253e-01 8.31571341e-01
5.54666877e-01 -5.00378549e-01 5.62308192e-01 4.99899328e-01
-1.11417212e-01 -6.15820214e-02 -1.35821223e+00 5.63022852e-01
5.79790294e-01 -6.23009264e-01 -7.64016092e-01 5.70460379e-01
3.64135832e-01 -2.48675898e-01 -9.55567360e-01 -2.78280258e-01
6.57788634e-01 -8.07048798e-01 2.89454073e-01 -8.70841742e-01
4.72086936e-01 2.29628861e-01 1.15325943e-01 -1.07038975e+00
-1.18068268e-03 -8.82376730e-01 3.83884683e-02 1.73470986e+00
6.94614232e-01 -5.86974800e-01 5.67474782e-01 8.29070389e-01
-3.15697603e-02 -5.70016384e-01 -7.84311950e-01 -5.53272605e-01
2.31871400e-02 -5.56504548e-01 3.00746888e-01 1.19414508e+00
5.15500188e-01 7.42396593e-01 -4.69030917e-01 -1.03473492e-01
-8.19932222e-02 -4.43825215e-01 1.17637980e+00 -9.73154306e-01
-9.74169075e-02 -1.38713539e-01 -3.76720607e-01 -7.65820086e-01
6.51506931e-02 -7.76188195e-01 2.85287593e-02 -1.77990746e+00
2.39223182e-01 -6.22254014e-01 2.73722798e-01 6.78472459e-01
1.40473425e-01 -7.47717591e-03 8.90236199e-02 -4.29167412e-03
-2.98675656e-01 2.06773326e-01 8.90449762e-01 -4.89341654e-02
-2.58044153e-01 2.15331227e-01 -6.57185555e-01 7.18825161e-01
8.84675622e-01 -4.00080264e-01 -6.86873868e-02 -3.37163478e-01
8.63673985e-02 -1.82036623e-01 -1.27215505e-01 -7.22765505e-01
1.74475774e-01 -2.11585704e-02 -2.64176756e-01 -2.43068352e-01
2.01829121e-01 -1.01040387e+00 -2.47002672e-02 3.45630348e-01
-2.58631051e-01 -1.57479811e-02 -6.01823553e-02 -1.33056223e-01
-9.87950340e-02 -7.08136022e-01 5.77024937e-01 2.29140632e-02
-4.13305253e-01 -2.56966144e-01 -8.83722663e-01 -2.22671419e-01
1.24943042e+00 -1.75823718e-01 -3.09937716e-01 -3.52207810e-01
-5.10703921e-01 7.19313622e-02 1.56107843e-01 7.66783416e-01
4.67805006e-02 -9.48530257e-01 -7.78760076e-01 5.21043316e-02
8.54492262e-02 -5.07303953e-01 5.39874658e-02 9.64969218e-01
-1.16487515e+00 5.29494524e-01 -3.77151996e-01 -6.97097331e-02
-1.62924802e+00 1.33748338e-01 8.85503888e-02 -4.30355757e-01
-4.28338826e-01 8.38027060e-01 -6.75422490e-01 -9.31539655e-01
3.19816917e-01 -3.55664521e-01 -4.38759953e-01 3.21948797e-01
2.91595250e-01 5.71961582e-01 3.41207981e-01 -6.54662669e-01
-5.08445263e-01 -3.28669816e-01 -2.72116035e-01 -5.60879670e-02
1.65356386e+00 3.24702375e-02 -2.70334512e-01 7.95701802e-01
1.38498044e+00 2.67499119e-01 -4.29298550e-01 2.09825352e-01
2.34396145e-01 -2.21708506e-01 -3.04044664e-01 -8.33226502e-01
-4.33393627e-01 9.82845724e-01 3.01848352e-01 7.69836426e-01
2.76567221e-01 1.40644953e-01 5.88047922e-01 8.12327445e-01
3.66928458e-01 -1.29622972e+00 -1.70945808e-01 1.13122165e+00
8.22738588e-01 -1.56321073e+00 -2.99647391e-01 2.15686411e-02
-1.00246704e+00 1.23706436e+00 7.16759861e-01 1.13872260e-01
7.58151412e-01 4.26494211e-01 4.82440233e-01 -1.08138725e-01
-9.45149660e-01 -2.87863631e-02 -2.20190495e-01 8.34847689e-01
1.12272668e+00 1.83046788e-01 -5.08710563e-01 1.05422044e+00
-5.21554887e-01 -2.13261142e-01 7.43969917e-01 8.61388087e-01
-1.16140403e-01 -1.26531863e+00 9.77123454e-02 4.59193677e-01
-7.55368948e-01 7.79061541e-02 -6.40284240e-01 1.16968048e+00
7.47638717e-02 1.21324492e+00 -1.15274996e-01 -5.57173193e-01
4.68356639e-01 3.83277863e-01 2.01488346e-01 -7.59118974e-01
-1.19602299e+00 -4.64133352e-01 7.16306210e-01 -9.00156200e-02
-3.64136934e-01 -6.25399470e-01 -9.85047698e-01 -7.40039945e-01
-7.07349598e-01 3.33955854e-01 8.66931319e-01 9.64693844e-01
-4.18814570e-02 3.64310175e-01 7.04911292e-01 -2.98905373e-01
-4.58516717e-01 -1.80496311e+00 -3.23768198e-01 2.98884213e-01
-1.45693079e-01 -1.26682207e-01 -2.30232552e-01 7.35934079e-02] | [12.730609893798828, 7.746260643005371] |
8cd7c694-d45b-4df5-be38-04f3fa5a8527 | with-blinkers-on-robust-prediction-of-eye | null | null | https://aclanthology.org/D13-1075 | https://aclanthology.org/D13-1075.pdf | With Blinkers on: Robust Prediction of Eye Movements across Readers | null | ['Anders S{\\o}gaard', 'Franz Matthies'] | 2013-10-01 | null | null | null | emnlp-2013-10 | ['transition-based-dependency-parsing'] | ['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] | [-7.252848148345947, 3.8093903064727783] |
b80d7455-f999-4fe8-b42d-9ccdfd963986 | unified-transfer-learning-models-for-high | 2307.00238 | null | https://arxiv.org/abs/2307.00238v1 | https://arxiv.org/pdf/2307.00238v1.pdf | Unified Transfer Learning Models for High-Dimensional Linear Regression | Transfer learning plays a key role in modern data analysis when: (1) the target data are scarce but the source data are sufficient; (2) the distributions of the source and target data are heterogeneous. This paper develops an interpretable unified transfer learning model, termed as UTrans, which can detect both transferable variables and source data. More specifically, we establish the estimation error bounds and prove that our bounds are lower than those with target data only. Besides, we propose a source detection algorithm based on hypothesis testing to exclude the nontransferable data. We evaluate and compare UTrans to the existing algorithms in multiple experiments. It is shown that UTrans attains much lower estimation and prediction errors than the existing methods, while preserving interpretability. We finally apply it to the US intergenerational mobility data and compare our proposed algorithms to the classical machine learning algorithms. | ['Shuo Shuo Liu'] | 2023-07-01 | null | null | null | null | ['transfer-learning'] | ['miscellaneous'] | [ 2.52440870e-01 2.30449349e-01 -1.13060141e+00 -2.33125299e-01
-8.58353794e-01 -4.56014931e-01 4.47560936e-01 1.40708849e-01
-1.72787175e-01 1.57468712e+00 -1.76410675e-02 -3.49118382e-01
-5.14348745e-01 -1.01287138e+00 -1.03103137e+00 -6.63203657e-01
-3.29690367e-01 8.03190947e-01 1.13741346e-01 -8.27735439e-02
-3.69092673e-02 -2.32299156e-02 -1.38607609e+00 -6.28940063e-04
1.48062539e+00 8.52290928e-01 -1.21062577e-01 3.83468330e-01
1.82754621e-01 5.24562240e-01 -5.53246260e-01 -3.70825708e-01
2.58002758e-01 -5.94360411e-01 -8.18339169e-01 -2.87441611e-01
1.34985313e-01 -1.41999692e-01 6.84392899e-02 1.16477907e+00
3.62331510e-01 -8.22038800e-02 1.16125500e+00 -1.96503890e+00
-9.81308579e-01 9.46150541e-01 -7.72402406e-01 2.55088240e-01
1.30442306e-01 -6.71068311e-01 1.00062573e+00 -5.26103973e-01
5.55608869e-01 1.34499979e+00 5.66007137e-01 4.11607742e-01
-1.35992467e+00 -9.68986094e-01 -3.30945430e-03 4.90885586e-01
-1.12752128e+00 -2.94542223e-01 6.37520671e-01 -5.25068402e-01
1.63027778e-01 2.28385091e-01 2.60731518e-01 1.44477475e+00
1.34364858e-01 9.28249002e-01 1.03218699e+00 -6.56705439e-01
3.58345330e-01 3.98442388e-01 2.45313242e-01 6.97810948e-01
5.26087046e-01 3.10073942e-01 -5.47605336e-01 -4.12166059e-01
5.87033510e-01 -2.20132425e-01 -9.99652743e-02 -1.83453694e-01
-9.97509062e-01 1.09029210e+00 1.40795246e-01 2.88383156e-01
-9.23042744e-02 -1.41902998e-01 2.46065825e-01 6.90964937e-01
8.75585735e-01 -1.62540764e-01 -8.00316751e-01 -2.38324609e-02
-5.79945743e-01 -3.76780033e-02 7.67259181e-01 1.23731637e+00
9.37407553e-01 -3.94471921e-03 3.56909297e-02 6.84465706e-01
1.49853185e-01 8.25671375e-01 6.80573642e-01 -7.01146781e-01
8.77584159e-01 5.00670493e-01 1.86768800e-01 -8.44056726e-01
-1.72290415e-01 -3.04767251e-01 -9.04471934e-01 -2.03731582e-01
4.95453417e-01 -6.54451668e-01 -7.21176028e-01 2.16841245e+00
4.36245859e-01 5.10217607e-01 9.15005282e-02 4.60150063e-01
3.21538150e-01 5.63185096e-01 8.64161458e-03 -4.77876544e-01
7.03365326e-01 -4.79708463e-01 -7.74058342e-01 -9.72390547e-02
8.70550036e-01 -3.29755872e-01 8.18275571e-01 2.88825035e-01
-6.49740756e-01 -5.03982008e-01 -7.93027699e-01 1.53669491e-01
-4.36110795e-01 2.29900077e-01 6.96369648e-01 1.05548823e+00
-7.66268730e-01 6.18987620e-01 -5.91153800e-01 -5.91813982e-01
5.21247983e-01 5.87619245e-01 -3.28526974e-01 3.45786542e-01
-1.45940149e+00 4.39847678e-01 5.84419608e-01 -2.64879018e-01
-5.56818545e-01 -7.00178742e-01 -7.21304655e-01 -1.06243212e-02
2.31705546e-01 -5.02669036e-01 9.42613602e-01 -1.34671462e+00
-1.41245103e+00 4.78085667e-01 -1.59479409e-01 -2.49290764e-01
6.29528880e-01 -3.25188607e-01 -4.81488556e-01 -1.88600704e-01
5.39232194e-01 7.77710155e-02 8.50991905e-01 -1.06928933e+00
-1.10926616e+00 -4.54649210e-01 -3.01280558e-01 -1.68763936e-01
-6.94489658e-01 -1.93534464e-01 -2.87295636e-02 -7.74168015e-01
-2.20093802e-01 -9.30133581e-01 3.16085309e-01 -1.14513375e-01
-6.03953063e-01 -4.37759787e-01 9.45001304e-01 -5.47674477e-01
1.10337508e+00 -1.87124062e+00 4.20435697e-01 4.14858133e-01
7.71193877e-02 1.61690101e-01 -1.78358138e-01 2.14593768e-01
-2.27294132e-01 2.17085332e-01 -2.37771019e-01 -8.25974569e-02
7.13183060e-02 2.08859950e-01 -5.02765059e-01 5.92387974e-01
-2.18515456e-01 6.75817609e-01 -8.33421469e-01 -7.26547301e-01
-1.04262181e-01 -2.16798007e-01 -3.76083583e-01 1.10793754e-01
1.01534382e-01 3.81839812e-01 -7.63726890e-01 8.07840407e-01
6.07124448e-01 -2.83365220e-01 2.59142578e-01 1.43594757e-01
1.63170844e-01 -8.98039788e-02 -9.81866479e-01 1.08109903e+00
-2.38687903e-01 6.52523637e-01 -1.17782176e-01 -1.47856569e+00
8.25921834e-01 2.36110672e-01 5.28775394e-01 -4.89917338e-01
1.84068859e-01 3.28644097e-01 3.30923609e-02 -5.53122401e-01
4.51617204e-02 -9.66080502e-02 -1.36033520e-01 5.06989062e-01
1.57088324e-01 6.11093760e-01 4.18788008e-02 -6.36065751e-02
7.74814129e-01 1.69995353e-01 4.74859715e-01 -4.06469107e-01
3.24943990e-01 -2.73546696e-01 1.07857633e+00 8.96346629e-01
-2.14749128e-01 4.13920581e-02 7.57899344e-01 -1.68029461e-02
-7.81890512e-01 -1.40676963e+00 -2.39490613e-01 1.32010818e+00
1.70757294e-01 -6.83556050e-02 -6.48747087e-01 -1.02088428e+00
3.80362779e-01 5.21272123e-01 -1.04060209e+00 -2.13280901e-01
-1.47293553e-01 -1.14181554e+00 6.31294310e-01 6.63371205e-01
5.27778506e-01 -4.15991127e-01 1.28211789e-02 -8.55180994e-02
-5.62938988e-01 -7.00385034e-01 -1.91036031e-01 -4.12609391e-02
-9.10611749e-01 -1.21675646e+00 -5.98257482e-01 -7.29376853e-01
5.07987499e-01 -5.07014291e-03 7.26882517e-01 -3.06920916e-01
3.54286879e-01 4.20192406e-02 -5.86077809e-01 -7.67059863e-01
-4.05927867e-01 4.69308376e-01 4.22817826e-01 1.16078436e-01
4.98697340e-01 -6.41854763e-01 -1.17334090e-01 3.83559048e-01
-5.64754486e-01 -1.90479904e-01 6.16540730e-01 9.50417578e-01
2.34689593e-01 8.08397681e-02 1.33265948e+00 -1.18118060e+00
4.69264060e-01 -1.22078037e+00 -6.40092492e-01 5.80250084e-01
-9.49395418e-01 2.10363232e-02 5.54275036e-01 -6.03452623e-01
-1.19254303e+00 -2.83731550e-01 5.57731390e-01 -1.75968349e-01
-3.22022811e-02 4.51265097e-01 -3.75610083e-01 6.82212412e-02
6.29617810e-01 -1.81862933e-03 -1.63892090e-01 -5.52499950e-01
1.13783618e-02 1.03746164e+00 9.78037566e-02 -7.37309039e-01
1.03489614e+00 3.65447402e-01 2.26425081e-02 -8.90017509e-01
-5.64346135e-01 -5.67268655e-02 -8.70553851e-01 1.87421013e-02
4.99398679e-01 -7.21814334e-01 -5.34859300e-01 3.82449687e-01
-8.45578790e-01 -4.84901220e-01 4.85371612e-02 8.56681943e-01
-7.34942257e-01 3.53505671e-01 -6.36105299e-01 -9.09328938e-01
-9.61733386e-02 -7.06560016e-01 9.08112884e-01 7.85881504e-02
-1.67464271e-01 -1.43150568e+00 2.34735698e-01 3.06625813e-01
-2.61089462e-03 3.46514553e-01 1.25181568e+00 -9.65990663e-01
-3.42804521e-01 1.40128702e-01 -2.67978072e-01 -3.42606418e-02
6.95694566e-01 -1.02284037e-01 -8.62127185e-01 -3.29736233e-01
-4.43918377e-01 -3.38851571e-01 6.41965091e-01 4.61914837e-01
1.09873760e+00 -4.65378910e-01 -7.07641959e-01 4.97325391e-01
1.03884232e+00 1.43438384e-01 3.50862116e-01 2.00220123e-01
5.58499217e-01 9.31501210e-01 7.52689898e-01 4.17206019e-01
3.65772754e-01 4.81952637e-01 1.77616123e-02 -1.53386474e-01
3.47695142e-01 -4.44570869e-01 4.98364002e-01 6.75608277e-01
-3.05470198e-01 -6.40377283e-01 -6.84459031e-01 5.84896863e-01
-2.46231890e+00 -1.00803828e+00 -3.59230697e-01 2.37309861e+00
9.06804085e-01 -9.26364809e-02 3.98965806e-01 1.07520923e-01
1.14148366e+00 -3.03402990e-01 -6.86520994e-01 -8.96541625e-02
-1.64265513e-01 1.26770183e-01 6.06605291e-01 4.32883382e-01
-1.22722042e+00 8.16235781e-01 7.25811386e+00 1.03130078e+00
-8.15034151e-01 3.33884984e-01 6.67518914e-01 1.00005776e-01
-2.82498658e-01 -1.98123306e-01 -6.20934188e-01 8.10771763e-01
1.10243630e+00 -6.51793063e-01 5.53139821e-02 8.49169910e-01
5.04072681e-02 7.86275044e-02 -1.37708962e+00 7.02403188e-01
9.21924040e-02 -8.98208201e-01 -7.22363731e-03 2.57305652e-01
7.31574535e-01 -1.01550795e-01 9.19257700e-02 4.15836662e-01
5.59261560e-01 -8.44587028e-01 5.45683503e-01 3.75632763e-01
7.55167782e-01 -8.34103048e-01 8.01294625e-01 4.89541620e-01
-9.02744055e-01 -2.40960911e-01 -4.54545856e-01 -1.28194705e-01
-2.44643927e-01 5.02715111e-01 -7.11383641e-01 9.06883895e-01
5.31198025e-01 8.83766770e-01 -3.98037076e-01 8.38328958e-01
-1.97311401e-01 9.54889357e-01 -3.40481132e-01 -2.88936775e-02
-2.54655808e-01 -3.60538065e-01 3.63150060e-01 7.88091838e-01
7.35172212e-01 -2.00278357e-01 1.20939307e-01 8.70871603e-01
-3.90463680e-01 2.48014271e-01 -6.63470626e-01 3.14859629e-01
6.50424838e-01 7.01483488e-01 -4.51521516e-01 -2.48842642e-01
-5.27426243e-01 9.09804404e-01 3.09548557e-01 7.15963483e-01
-9.22619820e-01 -5.82472861e-01 4.96825069e-01 -1.65996701e-01
1.13555670e-01 2.87147433e-01 -1.60399914e-01 -1.44275737e+00
-1.72346562e-01 -4.02783513e-01 7.17833102e-01 -3.05883050e-01
-1.69159114e+00 4.24948893e-02 4.19837207e-01 -1.22685719e+00
-4.49732482e-01 -5.53756356e-01 -4.56243038e-01 7.30916142e-01
-1.49743199e+00 -1.23188472e+00 1.23389520e-01 7.77094543e-01
1.62702024e-01 -4.56055641e-01 6.58207238e-01 6.46967590e-01
-8.29527497e-01 9.26524580e-01 4.28383261e-01 3.22407544e-01
9.02822614e-01 -1.25059521e+00 -2.21911535e-01 6.97211802e-01
-5.16108610e-02 3.02548319e-01 5.44927478e-01 -1.05218315e+00
-1.07538414e+00 -1.36396968e+00 8.37858737e-01 -4.26170170e-01
7.59545088e-01 -4.00803208e-01 -8.03258419e-01 1.32676983e+00
1.37503492e-02 -1.53215930e-01 1.13154531e+00 5.88839650e-01
-2.63675243e-01 -3.11037809e-01 -1.25805604e+00 3.29307437e-01
9.76179540e-01 -3.41501802e-01 -7.33347058e-01 3.76992136e-01
5.87040246e-01 2.36091558e-02 -7.86396921e-01 3.32868099e-01
5.69052935e-01 -7.22398877e-01 7.14323342e-01 -1.05162585e+00
2.10891634e-01 7.46110827e-02 -2.46132970e-01 -1.33054376e+00
-4.51202571e-01 -5.92797220e-01 -1.31965578e-01 1.56264770e+00
6.42986536e-01 -1.02925229e+00 8.05413485e-01 5.76459646e-01
3.71685266e-01 -1.00426655e-02 -1.35993361e+00 -1.16883576e+00
4.19210345e-01 -1.52306512e-01 6.65256917e-01 1.39210474e+00
2.53651440e-01 2.97384232e-01 -6.97065771e-01 1.28986031e-01
9.97860610e-01 3.49963158e-01 6.27871156e-01 -1.73119700e+00
-9.85776931e-02 -8.52304162e-04 -2.76795387e-01 -5.85775554e-01
7.33783543e-01 -9.24356401e-01 -2.08247855e-01 -9.57471728e-01
3.97037804e-01 -6.80767536e-01 -5.71333647e-01 6.44171059e-01
-1.93204001e-01 7.94165581e-02 -2.56743968e-01 1.60122767e-01
-2.23364651e-01 7.34284222e-01 8.86521339e-01 -4.09809835e-02
-4.10756737e-01 4.38450664e-01 -6.49002671e-01 7.90834069e-01
1.10882914e+00 -5.59038579e-01 -7.40673482e-01 -3.29305232e-01
3.03666666e-02 5.92410155e-02 2.63075113e-01 -7.42988110e-01
-1.25336975e-01 -5.63432455e-01 4.90971774e-01 -2.64830232e-01
-4.73820940e-02 -5.89940250e-01 -6.64348155e-02 4.75761533e-01
-5.00635922e-01 -3.04556370e-01 -1.60417810e-01 8.15371275e-01
1.07823171e-01 -6.88244775e-02 5.79512239e-01 4.13288653e-01
-4.43759859e-01 4.55718994e-01 -3.56735587e-01 1.34890035e-01
1.18643701e+00 1.59001306e-01 -5.22057950e-01 -5.55812418e-01
-4.90663975e-01 1.55137658e-01 8.25916976e-02 4.67575729e-01
3.67785543e-01 -1.51972985e+00 -9.12317216e-01 1.88209236e-01
2.86922082e-02 -5.19690275e-01 -7.05715269e-02 6.98767185e-01
3.85690890e-02 3.96035254e-01 -2.52569616e-01 -3.53291899e-01
-1.17119300e+00 4.85816002e-01 -1.08181499e-02 7.60138556e-02
-7.09425583e-02 4.03416455e-01 3.18380922e-01 -6.74652815e-01
-1.17440209e-01 -2.78994262e-01 -1.67519420e-01 2.55544186e-01
2.74420768e-01 9.24717903e-01 -3.07718933e-01 -5.73522091e-01
-2.41153136e-01 4.60020155e-01 1.25918180e-01 -2.67293770e-02
1.28646243e+00 -3.54689747e-01 -1.18856311e-01 7.00211942e-01
1.18471491e+00 6.12734705e-02 -8.09578955e-01 -4.44481373e-01
1.84626371e-01 -4.21666771e-01 -2.66488284e-01 -5.15497565e-01
-7.83884108e-01 7.02757895e-01 7.11526811e-01 3.87415946e-01
1.10871184e+00 7.29202852e-02 4.94821429e-01 3.61962616e-01
5.73307574e-01 -1.26768637e+00 -2.94401199e-01 1.46639109e-01
4.71967518e-01 -1.47115684e+00 -9.72859934e-02 -6.69351816e-01
-8.99761617e-02 9.14808273e-01 6.20756447e-01 1.84850514e-01
7.18999863e-01 3.69613022e-02 -3.27214479e-01 4.88912880e-01
-7.17937887e-01 -1.59891620e-01 1.77878156e-01 1.02175951e+00
1.68283716e-01 2.29177028e-01 -4.02988017e-01 7.54671872e-01
-1.97022319e-01 3.89191210e-01 4.50438589e-01 5.94398141e-01
-3.34499806e-01 -1.38530505e+00 -4.75117534e-01 5.82232952e-01
-3.88601512e-01 7.19699785e-02 -3.83534163e-01 1.09584546e+00
3.31274450e-01 1.14578998e+00 8.36720169e-02 -4.97389168e-01
-4.87977937e-02 7.67721310e-02 3.86206061e-01 -4.18436110e-01
1.17005363e-01 1.17359487e-02 8.82072002e-02 -2.65504837e-01
-7.45409310e-01 -6.94259703e-01 -1.00515330e+00 -4.71527219e-01
-3.74687761e-01 5.00073850e-01 4.25540432e-02 1.10357106e+00
4.48098719e-01 1.64524615e-01 8.55878234e-01 -3.32693487e-01
-5.10800421e-01 -9.52564418e-01 -9.06091630e-01 1.96360543e-01
5.87417722e-01 -9.34458792e-01 -4.08772379e-01 3.07124227e-01] | [10.240692138671875, 3.310764789581299] |
29e1612b-97a3-43f7-9bec-c25a9f0bbb97 | extracting-linguistic-resources-from-the-web | 1810.13414 | null | http://arxiv.org/abs/1810.13414v1 | http://arxiv.org/pdf/1810.13414v1.pdf | Extracting Linguistic Resources from the Web for Concept-to-Text Generation | Many concept-to-text generation systems require domain-specific linguistic
resources to produce high quality texts, but manually constructing these
resources can be tedious and costly. Focusing on NaturalOWL, a publicly
available state of the art natural language generator for OWL ontologies, we
propose methods to extract from the Web sentence plans and natural language
names, two of the most important types of domain-specific linguistic resources
used by the generator. Experiments show that texts generated using linguistic
resources extracted by our methods in a semi-automatic manner, with minimal
human involvement, are perceived as being almost as good as texts generated
using manually authored linguistic resources, and much better than texts
produced by using linguistic resources extracted from the relation and entity
identifiers of the ontology. | ['Ion Androutsopoulos', 'Gerasimos Lampouras'] | 2018-10-31 | null | null | null | null | ['concept-to-text-generation'] | ['natural-language-processing'] | [ 4.03355733e-02 1.11489785e+00 -1.89017758e-01 -1.81482852e-01
-8.15797448e-01 -6.77127242e-01 9.24168646e-01 5.47799468e-01
-5.98978400e-01 1.40998745e+00 6.17510259e-01 -1.71495348e-01
5.33864908e-02 -1.15817451e+00 -3.62594306e-01 3.01405668e-01
3.06209683e-01 1.06712377e+00 5.12600720e-01 -5.73368073e-01
1.20819762e-01 1.46489784e-01 -1.76969802e+00 3.17779332e-01
1.44820380e+00 5.79550862e-01 2.00639546e-01 1.18488997e-01
-1.09533167e+00 1.40437567e+00 -9.62992907e-01 -6.98477745e-01
-1.21307157e-01 -6.80487216e-01 -1.40533507e+00 1.39009610e-01
-7.07612038e-02 2.86999904e-02 2.46521249e-01 1.06175232e+00
2.21673653e-01 -7.14670792e-02 6.43684387e-01 -1.05280268e+00
-6.12153471e-01 1.47436774e+00 2.91077524e-01 -2.39274129e-01
9.62055743e-01 -1.25903979e-01 1.19865680e+00 -8.03654432e-01
1.28046763e+00 1.37677789e+00 1.62400305e-01 8.03850293e-01
-1.02313042e+00 -2.97298908e-01 -4.13532734e-01 -6.52686357e-02
-1.49507987e+00 -7.38910973e-01 3.71717960e-01 -4.83541280e-01
1.28085327e+00 2.97336489e-01 2.88406163e-01 9.26582038e-01
5.52063202e-03 2.13040978e-01 8.91942799e-01 -1.26331043e+00
3.15467834e-01 8.64515305e-01 -1.33793652e-01 7.63656616e-01
7.08937883e-01 -3.96159023e-01 -6.61638558e-01 -4.25141871e-01
4.27422404e-01 -8.87343764e-01 -5.18685840e-02 1.18280485e-01
-1.05490565e+00 9.49503541e-01 -3.32518369e-01 6.86756015e-01
-4.06482756e-01 -4.97537404e-01 3.61403435e-01 5.67777231e-02
3.55396181e-01 1.09593558e+00 -5.17587483e-01 -1.38021678e-01
-6.16945446e-01 4.71251577e-01 1.45141113e+00 1.77073586e+00
8.23937595e-01 -1.87421083e-01 -1.63904965e-01 7.90145755e-01
3.36202383e-01 6.47818506e-01 8.56811225e-01 -7.24943638e-01
5.58300614e-01 1.05180109e+00 6.97386265e-01 -8.16586375e-01
-2.30168983e-01 1.58417881e-01 -7.07998201e-02 -9.35481042e-02
3.98010403e-01 -4.51363325e-01 -4.58288342e-01 1.42597306e+00
2.83544987e-01 -8.75573039e-01 6.90244853e-01 2.53107607e-01
1.18733180e+00 4.29936051e-01 4.83930856e-01 -3.79480153e-01
1.76424611e+00 -7.47886240e-01 -9.73953366e-01 -4.01551783e-01
7.73664415e-01 -6.73758030e-01 1.08762968e+00 -1.43044621e-01
-1.07899785e+00 -2.03405336e-01 -9.86697853e-01 -3.21051739e-02
-7.80484200e-01 9.74718332e-02 5.44128120e-01 5.80008507e-01
-6.96186483e-01 3.45315516e-01 -3.18201154e-01 -6.05163753e-01
1.36404723e-01 -2.84853607e-01 -4.87184674e-01 3.96536803e-03
-1.65482724e+00 1.19510233e+00 1.01848006e+00 -7.92909503e-01
-3.12789261e-01 -3.86658370e-01 -1.25460279e+00 5.00212796e-02
8.64491403e-01 -6.45706058e-01 1.53282619e+00 -7.09603190e-01
-1.38442826e+00 9.79933441e-01 -2.90826440e-01 -6.92992747e-01
2.95227021e-01 -3.25314589e-02 -7.68529713e-01 2.96219081e-01
7.24787235e-01 6.21542454e-01 2.15633810e-01 -9.20499325e-01
-1.01289868e+00 1.36733249e-01 3.25108409e-01 5.68855405e-02
-3.51816118e-01 6.21134758e-01 -6.71424810e-03 -4.42255080e-01
-4.47958082e-01 -3.28110665e-01 -3.50894064e-01 -2.74383426e-01
-3.93780798e-01 -7.39096880e-01 1.57237962e-01 -8.43052506e-01
1.42571282e+00 -1.36647284e+00 -3.63497049e-01 2.42217302e-01
-6.22082390e-02 2.17605174e-01 6.16306886e-02 8.71518433e-01
2.76537448e-01 6.99847460e-01 -4.09817956e-02 3.04477990e-01
4.54713255e-01 3.79070073e-01 -4.89009649e-01 -5.81466496e-01
3.78542602e-01 6.37310445e-01 -1.38021207e+00 -1.04846108e+00
-1.35741979e-02 -3.98616977e-02 -1.89342543e-01 1.78676367e-01
-8.03921044e-01 5.22456057e-02 -8.17514002e-01 3.75733465e-01
-1.61312625e-01 -3.47942226e-02 3.73066276e-01 1.68396488e-01
3.48729864e-02 9.15866733e-01 -1.12823606e+00 1.31388652e+00
-8.80509436e-01 3.90866518e-01 -6.35819733e-01 -2.59317249e-01
1.03078401e+00 9.36202049e-01 2.48745784e-01 -6.33099616e-01
-5.57253435e-02 5.01361668e-01 -3.47447276e-01 -8.83414626e-01
5.46809852e-01 -5.30925572e-01 -4.05010134e-01 7.48828888e-01
3.93145412e-01 -5.20823240e-01 1.24983859e+00 4.39116508e-01
1.05292857e+00 1.51863530e-01 1.00003409e+00 -2.49568731e-01
8.36802423e-01 6.92340791e-01 6.21188521e-01 6.09608293e-01
4.45301682e-01 1.34664446e-01 5.08722782e-01 -3.06188226e-01
-1.15269160e+00 -4.20195967e-01 -1.47322655e-01 6.72314286e-01
-3.76349956e-01 -1.17466938e+00 -1.07020938e+00 -8.44587147e-01
-3.36397737e-01 1.55111730e+00 -2.13187754e-01 2.82465786e-01
-4.34165895e-01 -1.53656051e-01 7.44904339e-01 2.11853504e-01
1.97069272e-01 -1.63093531e+00 -6.69565678e-01 6.79840744e-01
-6.60487354e-01 -1.77926755e+00 -2.16706142e-01 -3.54909807e-01
-2.62334883e-01 -1.18327534e+00 -2.26763651e-01 -4.90790337e-01
8.66117239e-01 -5.48452914e-01 1.67748523e+00 1.40411779e-01
-1.87977269e-01 -8.07176307e-02 -7.48869896e-01 -8.64822567e-01
-1.20172632e+00 9.77369994e-02 -9.00364742e-02 -3.46045315e-01
7.25770593e-01 -4.04548466e-01 4.94891405e-01 9.42253172e-02
-1.42355120e+00 3.42194825e-01 1.77865073e-01 3.20460856e-01
1.99693069e-01 4.11815643e-01 7.17437863e-01 -1.41048372e+00
1.02455103e+00 -1.65210247e-01 -6.68555737e-01 3.92536372e-01
-5.77241838e-01 6.79000258e-01 1.03353488e+00 -1.29971370e-01
-1.31070614e+00 -6.91017741e-03 -7.64990598e-02 6.31973386e-01
-4.55678076e-01 8.38377178e-01 -4.91941303e-01 3.71597469e-01
1.14806521e+00 3.97452675e-02 -4.94100004e-01 -4.20513988e-01
5.84274292e-01 8.70599270e-01 1.63018763e-01 -1.05426109e+00
1.07936406e+00 1.89438775e-01 -4.15061474e-01 -8.48528564e-01
-1.18223834e+00 -1.68094039e-01 -7.23635316e-01 5.25786355e-02
7.88973629e-01 -7.41012990e-01 -8.74943733e-02 -2.45880798e-01
-1.33251071e+00 -1.34715676e-01 -8.78081679e-01 1.92093521e-01
-4.92109001e-01 1.86449766e-01 -1.43043384e-01 -7.19862044e-01
-5.96906781e-01 -5.95861614e-01 8.27433944e-01 2.58337528e-01
-8.24351132e-01 -1.02464640e+00 5.46790063e-02 3.37632656e-01
3.04276377e-01 4.09070671e-01 1.34303021e+00 -1.01740432e+00
-1.81353301e-01 -4.30908591e-01 3.64202112e-02 1.03609025e-01
6.08422637e-01 -4.22705710e-02 -5.06345153e-01 4.09759998e-01
-3.08746725e-01 -3.28845769e-01 -2.21861035e-01 -5.23918748e-01
2.40248919e-01 -1.17553031e+00 -2.97607213e-01 -2.94064134e-01
1.47570288e+00 1.65430725e-01 5.71925581e-01 4.65684742e-01
3.79600585e-01 9.73396182e-01 6.93282008e-01 4.93230522e-01
6.17330670e-01 6.69735253e-01 -3.70326012e-01 3.05562973e-01
-3.85633558e-02 -6.13270819e-01 2.82546729e-01 6.36414409e-01
6.77401349e-02 -9.28469449e-02 -1.12951243e+00 8.15619111e-01
-1.56836927e+00 -1.06744647e+00 -5.38915163e-03 1.90998495e+00
1.74232519e+00 2.17113286e-01 4.04535830e-02 -1.20968305e-01
5.35897553e-01 -1.95093364e-01 5.33072352e-02 -1.83175117e-01
-3.02014828e-01 6.14627242e-01 4.00204539e-01 7.37178385e-01
-6.09762669e-01 1.32130933e+00 6.08069754e+00 6.19219422e-01
-4.73169059e-01 8.25529993e-02 -3.05234075e-01 2.40762144e-01
-6.98911428e-01 5.89291275e-01 -1.04979646e+00 3.48528385e-01
1.31932271e+00 -1.16842246e+00 2.16131926e-01 9.49159801e-01
3.96381617e-01 -7.62755126e-02 -1.15757024e+00 4.62354600e-01
1.38396889e-01 -1.55412769e+00 6.27237678e-01 -2.40881257e-02
6.70785487e-01 -3.75433445e-01 -1.19701278e+00 2.19414324e-01
9.67777371e-01 -8.87513041e-01 1.21433127e+00 5.14488816e-01
6.67078972e-01 -7.72741973e-01 8.80605102e-01 6.62865520e-01
-1.05948389e+00 1.45220622e-01 -4.99237567e-01 1.12841325e-02
4.81016725e-01 8.92173588e-01 -1.12819147e+00 6.62666738e-01
1.62199378e-01 1.70321673e-01 -8.29187274e-01 5.45527697e-01
-9.42584872e-01 4.11208451e-01 -1.95018172e-01 -5.15610278e-01
9.32714865e-02 8.37772340e-02 5.66829503e-01 1.34507942e+00
2.21009642e-01 8.90746862e-02 4.68023658e-01 9.20361698e-01
-2.11859539e-01 6.47686958e-01 -8.51590991e-01 -5.38405955e-01
5.83666563e-01 1.16345847e+00 -5.56463480e-01 -9.51508880e-01
-2.90736914e-01 4.78664547e-01 1.87817410e-01 1.41999573e-01
-3.58428031e-01 -1.04448020e+00 2.47140527e-02 4.65957552e-01
-9.15248469e-02 -4.66121398e-02 5.28658703e-02 -1.22654557e+00
3.48110050e-01 -1.26245844e+00 3.52290630e-01 -1.03054404e+00
-1.17254090e+00 1.05986798e+00 1.54685646e-01 -1.19161284e+00
-1.02911949e+00 -4.28643882e-01 -1.99748889e-01 9.43591237e-01
-1.28496206e+00 -9.09566104e-01 -7.94972628e-02 1.80757821e-01
5.82299590e-01 -2.58901894e-01 1.36059773e+00 2.46544346e-01
7.79046118e-02 1.51374429e-01 -7.06756175e-01 5.79392493e-01
5.07375479e-01 -1.27403176e+00 4.40691918e-01 9.15398240e-01
4.00415331e-01 8.83627415e-01 9.03890848e-01 -8.72769535e-01
-7.98165321e-01 -1.09393764e+00 2.06996655e+00 -3.80539060e-01
8.56611311e-01 -3.20732743e-01 -7.64043212e-01 6.60034478e-01
6.21043682e-01 -6.03167236e-01 8.29421043e-01 -2.48525828e-01
-7.34319910e-02 2.38139868e-01 -1.24826074e+00 7.08356798e-01
9.00126100e-01 -5.40169597e-01 -1.19369435e+00 9.68147814e-01
7.66395867e-01 -3.15257043e-01 -8.84261072e-01 -4.77286726e-02
1.75513834e-01 -2.60580331e-01 4.36186135e-01 -8.14755619e-01
4.59561527e-01 -5.45698404e-01 4.00176160e-02 -1.16898382e+00
1.91697359e-01 -9.11406040e-01 2.81414509e-01 1.72050405e+00
1.11056721e+00 -6.62301838e-01 2.00590700e-01 1.04923034e+00
1.60244003e-01 -8.95304903e-02 -6.75141454e-01 -1.00551176e+00
-1.31655335e-01 -3.79597783e-01 1.01395237e+00 8.89274299e-01
7.22366631e-01 9.72549260e-01 1.45690486e-01 -3.28914747e-02
3.22467685e-01 -2.34369814e-01 7.86952317e-01 -1.53355634e+00
6.53056204e-02 -2.14139655e-01 -8.19652900e-02 -3.12341958e-01
5.21746993e-01 -9.00215685e-01 3.44863474e-01 -2.14793587e+00
-6.55357093e-02 -4.31054741e-01 5.55035651e-01 9.02853072e-01
6.55744039e-03 -3.93967152e-01 1.37035351e-03 2.41664294e-02
-3.57457817e-01 4.28548068e-01 1.00719368e+00 -1.12582007e-02
-1.63435504e-01 -2.18860477e-01 -1.14789665e+00 1.01125026e+00
5.08365870e-01 -9.85216081e-01 -3.72583956e-01 -1.66266769e-01
5.88010728e-01 -2.54813563e-02 -2.60661662e-01 -1.10035789e+00
1.28089383e-01 -6.99916840e-01 -2.33224764e-01 5.41703254e-02
-1.80117115e-01 -6.55604422e-01 1.11615673e-01 8.93493742e-02
-5.15839100e-01 -4.17237282e-02 1.52553171e-01 -1.72967777e-01
-3.85452628e-01 -1.09447670e+00 5.67612112e-01 -7.53978670e-01
-4.70659822e-01 -1.37847602e-01 -7.98909068e-01 7.02651560e-01
7.39923537e-01 1.09010682e-01 -3.12137187e-01 -4.42533344e-01
-3.67700219e-01 -3.38406898e-02 4.85136122e-01 4.53446090e-01
3.22012603e-01 -1.25859749e+00 -8.71639550e-01 -1.43007100e-01
5.21174610e-01 -1.26812607e-01 -5.43568075e-01 -6.19015582e-02
-6.50871754e-01 8.21766496e-01 -1.62492231e-01 2.36540094e-01
-8.19418192e-01 3.47704530e-01 1.80467755e-01 -7.31304705e-01
-6.87104583e-01 1.72470808e-01 -5.29509127e-01 -5.48725784e-01
-6.88450038e-02 -3.34633797e-01 -7.03125000e-01 8.55852067e-02
9.89434123e-01 -1.77213818e-01 1.54170468e-01 -7.27259278e-01
-3.51894915e-01 9.93098784e-03 1.50633335e-01 -7.59978533e-01
1.25057435e+00 1.04496032e-02 -3.29158127e-01 2.02159449e-01
2.27347121e-01 6.84464812e-01 -3.24173540e-01 -5.83623528e-01
7.13721633e-01 -1.64582998e-01 -4.16437566e-01 -7.95837700e-01
-3.23529989e-01 3.49949002e-01 -3.74606967e-01 5.56966186e-01
6.52080894e-01 1.47489190e-01 9.22430098e-01 7.23221838e-01
1.02442074e+00 -1.33343291e+00 -2.36502215e-01 7.08757102e-01
1.09489632e+00 -8.23021531e-01 -1.93170235e-02 -6.69959486e-01
-7.94542253e-01 1.12320626e+00 6.22698963e-01 4.03294235e-01
3.33379775e-01 3.37997228e-01 2.08666906e-01 -1.47332370e-01
-8.03804994e-01 -5.11315644e-01 8.19118321e-02 7.46129692e-01
5.98497331e-01 -1.60445005e-01 -7.76522577e-01 7.62443304e-01
-6.41477227e-01 1.93457022e-01 1.10745049e+00 1.28409183e+00
-6.42643869e-01 -1.72759223e+00 -1.40650108e-01 3.42754871e-01
-6.48569286e-01 -4.69360143e-01 -7.42350578e-01 7.25418210e-01
7.77995586e-02 1.23515546e+00 -3.92208964e-01 2.25483403e-01
6.19919658e-01 4.65520650e-01 2.58348614e-01 -1.56179011e+00
-4.98423696e-01 -4.30724800e-01 1.15263402e+00 -1.49902135e-01
-5.20468414e-01 -3.50604147e-01 -1.64593852e+00 3.22587267e-02
-3.51693988e-01 8.48455906e-01 4.62336451e-01 1.46687245e+00
1.89661101e-01 2.62849927e-01 2.45513886e-01 -2.46268272e-01
-3.18768084e-01 -1.24929845e+00 -3.37894827e-01 6.88931823e-01
-3.43726635e-01 -4.06224251e-01 -8.71489272e-02 7.03922868e-01] | [11.518828392028809, 9.01388168334961] |
96fb7e56-da2d-45b6-a200-fe8b77f5ea03 | mask-attack-detection-using-vascular-weighted | 2305.1594 | null | https://arxiv.org/abs/2305.15940v1 | https://arxiv.org/pdf/2305.15940v1.pdf | Mask Attack Detection Using Vascular-weighted Motion-robust rPPG Signals | Detecting 3D mask attacks to a face recognition system is challenging. Although genuine faces and 3D face masks show significantly different remote photoplethysmography (rPPG) signals, rPPG-based face anti-spoofing methods often suffer from performance degradation due to unstable face alignment in the video sequence and weak rPPG signals. To enhance the rPPG signal in a motion-robust way, a landmark-anchored face stitching method is proposed to align the faces robustly and precisely at the pixel-wise level by using both SIFT keypoints and facial landmarks. To better encode the rPPG signal, a weighted spatial-temporal representation is proposed, which emphasizes the face regions with rich blood vessels. In addition, characteristics of rPPG signals in different color spaces are jointly utilized. To improve the generalization capability, a lightweight EfficientNet with a Gated Recurrent Unit (GRU) is designed to extract both spatial and temporal features from the rPPG spatial-temporal representation for classification. The proposed method is compared with the state-of-the-art methods on five benchmark datasets under both intra-dataset and cross-dataset evaluations. The proposed method shows a significant and consistent improvement in performance over other state-of-the-art rPPG-based methods for face spoofing detection. | ['Xudong Jiang', 'Jiang Liu', 'Heshan Du', 'Ruibin Bai', 'Jianfeng Ren', 'Chenglin Yao'] | 2023-05-25 | null | null | null | null | ['face-recognition', 'face-alignment', 'face-anti-spoofing'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 3.42675745e-01 -4.00029808e-01 -1.45840496e-01 -2.45224461e-01
-5.37960827e-01 -3.08791995e-01 3.48270714e-01 -6.93533599e-01
2.48611644e-02 3.23524386e-01 3.38170648e-01 2.30310693e-01
1.23683184e-01 -3.84740144e-01 -3.01407427e-01 -1.03208482e+00
-1.96826085e-01 -4.80506331e-01 -8.87928754e-02 -4.97923382e-02
4.29483116e-01 1.03615785e+00 -1.63909125e+00 3.57608557e-01
4.48331267e-01 1.21932924e+00 -3.65666658e-01 4.09540862e-01
2.68944725e-02 4.00806725e-01 -4.69033062e-01 -3.74902308e-01
5.62744617e-01 -4.93513435e-01 -1.45781681e-01 9.64700282e-02
8.08151841e-01 -3.85977179e-01 -7.36626565e-01 1.24120486e+00
9.16326642e-01 -9.89398267e-03 4.11643356e-01 -1.33891761e+00
-5.79352736e-01 -1.43364623e-01 -1.18205762e+00 4.60974872e-01
5.25818825e-01 1.73920110e-01 1.26760572e-01 -1.19691253e+00
5.94313204e-01 1.49323606e+00 6.42209589e-01 8.16426873e-01
-9.59210753e-01 -1.21746171e+00 -2.51818091e-01 3.77595067e-01
-1.45682728e+00 -9.03102398e-01 1.40478122e+00 -3.67657691e-02
6.70965433e-01 2.41660982e-01 4.60419565e-01 1.07943654e+00
2.93740392e-01 4.30927992e-01 1.19689488e+00 -1.50637046e-01
-3.74212414e-01 -1.13203041e-01 -3.93826783e-01 9.99390602e-01
9.57768038e-02 5.44552207e-01 -9.78569031e-01 -1.65522054e-01
9.41681206e-01 9.48686823e-02 -6.04389071e-01 -2.41339713e-01
-7.48844922e-01 5.41787267e-01 3.17958385e-01 3.88734370e-01
-4.56894934e-01 -6.52601048e-02 4.60030735e-01 2.37964153e-01
3.82430255e-01 -2.83436291e-03 -4.04552296e-02 1.65828973e-01
-1.06020474e+00 -3.19505692e-01 5.47678649e-01 4.28397119e-01
4.14085120e-01 4.17466611e-01 -1.20313406e-01 8.38256001e-01
5.43878436e-01 6.55980051e-01 5.93306899e-01 -6.90523684e-01
4.58031178e-01 2.53054053e-01 -3.15243274e-01 -1.73119688e+00
-2.30821833e-01 -2.06149831e-01 -7.58379519e-01 1.64089710e-01
2.00644925e-01 3.76310535e-02 -8.43341470e-01 1.54485679e+00
6.20910525e-01 8.94360662e-01 4.55483794e-02 1.01614034e+00
1.01615143e+00 6.23655140e-01 -1.89427480e-01 -7.18915641e-01
1.34975410e+00 -6.11106575e-01 -8.57951343e-01 -4.06568088e-02
1.71214759e-01 -9.53903615e-01 3.74258012e-01 1.33002222e-01
-9.02055860e-01 -7.45265126e-01 -1.02562582e+00 1.48961097e-01
-1.37195334e-01 1.96911141e-01 2.07493573e-01 1.23796332e+00
-1.05056238e+00 5.92354000e-01 -5.66416919e-01 -1.24232650e-01
6.82823658e-01 5.38195491e-01 -8.89859796e-01 -3.13441724e-01
-1.11350930e+00 6.73578143e-01 -3.20982784e-01 5.41983664e-01
-8.75461280e-01 -6.31344378e-01 -1.12147903e+00 -2.00037137e-01
-6.29876629e-02 -6.02619946e-02 4.31424439e-01 -7.70202100e-01
-1.77804804e+00 1.06878841e+00 -4.68987226e-01 -9.36159864e-02
3.08439285e-01 1.30140826e-01 -8.55922639e-01 7.76703179e-01
-2.55360395e-01 5.25334060e-01 1.82570779e+00 -8.85908484e-01
-1.41293496e-01 -9.16283727e-01 -7.73718715e-01 2.20090449e-01
-4.47882116e-01 5.45278490e-01 -4.28750098e-01 -7.86674619e-01
3.50679874e-01 -5.59483826e-01 2.66167641e-01 1.75426289e-01
-6.81069493e-02 7.04785362e-02 1.46910751e+00 -1.06763661e+00
9.21839118e-01 -2.36838698e+00 -1.68724135e-01 1.92846000e-01
9.69294235e-02 7.70345032e-01 -4.00885582e-01 -1.46952849e-02
-3.58064502e-01 -1.04114838e-01 1.30218174e-02 -3.54049623e-01
-4.44267720e-01 -2.54706204e-01 -2.42145225e-01 1.27086186e+00
1.92690745e-01 6.78932965e-01 -8.75380993e-01 -6.42407954e-01
5.26458383e-01 1.12680995e+00 -2.16618881e-01 4.39607464e-02
6.79527164e-01 6.05289042e-01 -3.75465363e-01 1.13420784e+00
1.32115149e+00 3.97009492e-01 -2.82183639e-03 -4.95639622e-01
1.81401893e-01 -5.47063164e-02 -1.02144015e+00 1.27146554e+00
-1.87154800e-01 8.34766626e-01 3.35115224e-01 -8.15805495e-01
1.30224872e+00 6.95146322e-01 7.57840335e-01 -6.22869134e-01
3.50326180e-01 3.61978650e-01 -2.50014335e-01 -6.69174373e-01
3.97935510e-02 3.93085480e-02 4.80544806e-01 3.72405827e-01
1.77955389e-01 2.06907466e-01 -3.65717292e-01 -3.43072593e-01
7.96285450e-01 4.94843163e-02 3.09693851e-02 -1.92863822e-01
1.16420197e+00 -9.76465285e-01 8.92250955e-01 1.65441826e-01
-8.42511415e-01 6.07447505e-01 2.66934037e-01 -4.33066547e-01
-5.55594444e-01 -6.40339494e-01 -2.09911063e-01 4.33226615e-01
2.72281349e-01 -1.66376308e-01 -7.20363438e-01 -9.64042127e-01
1.92754135e-01 -4.82987650e-02 -6.90792620e-01 -1.78133056e-01
-7.27916718e-01 -4.95843530e-01 8.34379554e-01 7.87530988e-02
1.02910113e+00 -1.01433897e+00 -5.64858973e-01 1.80266574e-01
-3.14797089e-02 -1.26268339e+00 -7.74728298e-01 -6.40317559e-01
-8.04152310e-01 -1.24007952e+00 -9.46212590e-01 -9.32501316e-01
8.71377766e-01 7.80735135e-01 3.87358814e-01 1.22605979e-01
-6.32345676e-01 2.86952764e-01 -1.32535473e-01 -2.51517352e-02
-5.83108105e-02 -7.46088684e-01 3.33928943e-01 7.75500834e-01
5.16060948e-01 -4.56381083e-01 -9.29182172e-01 7.26830840e-01
-6.40753686e-01 -3.25855285e-01 3.67100239e-01 1.00836349e+00
2.65766650e-01 1.76499233e-01 5.06482720e-01 -3.12033951e-01
3.28170836e-01 2.90143257e-03 -4.56854969e-01 1.55499816e-01
-2.46617377e-01 -3.15577418e-01 3.16045284e-01 -5.56244016e-01
-1.02720630e+00 6.63155019e-02 1.51971141e-02 -9.77219284e-01
2.53741685e-02 -5.51698469e-02 -3.26140910e-01 -9.25939918e-01
4.25800353e-01 5.17088890e-01 5.14983535e-01 -2.73902327e-01
7.16405585e-02 7.24360764e-01 6.52272522e-01 -2.10898612e-02
9.88102794e-01 7.29295790e-01 3.31081361e-01 -1.20541883e+00
-6.14298694e-02 -5.21886110e-01 -4.38158423e-01 -4.43933934e-01
4.91730034e-01 -7.28174567e-01 -1.04067302e+00 1.03705537e+00
-1.17996812e+00 2.90607423e-01 2.76282042e-01 6.40623748e-01
-3.12582970e-01 8.86642337e-01 -8.08839083e-01 -1.05269790e+00
-6.61624372e-01 -1.22878897e+00 1.01364923e+00 4.76574183e-01
3.13288569e-01 -6.44678175e-01 -2.80401707e-01 2.63921022e-01
5.23022711e-01 4.14136976e-01 3.37608010e-01 -3.19504082e-01
-2.30499759e-01 -5.94859183e-01 -3.37596506e-01 5.14138162e-01
5.19842148e-01 8.45698342e-02 -1.29322255e+00 -4.30258512e-01
5.57397127e-01 7.14809373e-02 7.84251392e-01 5.04123986e-01
9.69584644e-01 -3.80353510e-01 -4.22535270e-01 8.52390349e-01
1.11981022e+00 3.22796345e-01 8.22197437e-01 -1.83592588e-01
6.49193466e-01 1.02962029e+00 4.40738142e-01 3.56046796e-01
-3.17171752e-01 7.87383080e-01 3.21435750e-01 -3.67454551e-02
-4.19934094e-01 -1.92403704e-01 7.44962156e-01 1.91495463e-01
-4.90110787e-03 8.22766870e-02 -2.41089016e-01 2.37620413e-01
-1.31635642e+00 -1.11495090e+00 2.42039144e-01 2.29221153e+00
5.24148285e-01 -4.07598555e-01 -1.64317921e-01 3.45758677e-01
1.12893820e+00 5.28044403e-01 -4.07847375e-01 -2.31273562e-01
-2.50134110e-01 2.63233602e-01 5.13543725e-01 3.75151545e-01
-1.14743316e+00 8.21254075e-01 5.30980873e+00 7.65073895e-01
-1.55213666e+00 8.09204727e-02 6.12071514e-01 2.15754788e-02
2.97972202e-01 -4.80895847e-01 -9.19746101e-01 6.63008869e-01
5.98295748e-01 1.12590879e-01 4.67908323e-01 5.15420258e-01
2.69840658e-01 3.29960793e-01 -6.67194605e-01 1.58920515e+00
7.79588461e-01 -1.16013813e+00 -3.04038882e-01 2.33916312e-01
4.67202306e-01 -4.10714537e-01 4.56226110e-01 -4.18147683e-01
-4.97712463e-01 -1.08011675e+00 6.56485856e-02 2.00314552e-01
9.92874682e-01 -8.28885496e-01 6.03202879e-01 -4.22095507e-01
-1.47984600e+00 -2.19852045e-01 -4.17295128e-01 5.93629122e-01
1.66795142e-02 3.10106874e-01 -7.33966291e-01 5.22389829e-01
6.61413074e-01 9.87789929e-01 -2.35758185e-01 9.76777136e-01
-3.88440013e-01 2.94565916e-01 -3.69114846e-01 3.66743565e-01
-3.04939374e-02 4.28979881e-02 8.67759407e-01 1.10381246e+00
5.65742373e-01 3.04764926e-01 -3.56047153e-01 5.31380951e-01
-2.13029668e-01 1.28626391e-01 -9.42010045e-01 -1.05348960e-01
4.74484771e-01 1.44172978e+00 -5.70827127e-01 -4.01876681e-02
-3.39722723e-01 1.02203977e+00 -3.37423295e-01 3.18849295e-01
-5.53804219e-01 -5.52647352e-01 9.44377124e-01 2.07055639e-02
3.31997395e-01 -9.36263502e-02 1.93798512e-01 -1.06355631e+00
1.81998797e-02 -8.03050756e-01 5.18310010e-01 -5.52614391e-01
-1.10168481e+00 8.01883519e-01 -4.94967163e-01 -1.38683903e+00
-8.70530382e-02 -6.25158310e-01 -5.98220170e-01 9.62150037e-01
-1.90744472e+00 -1.06276000e+00 -5.01042962e-01 1.01614535e+00
3.07797372e-01 -3.90007108e-01 6.26766384e-01 3.27967793e-01
-8.10629487e-01 9.27313745e-01 -4.08797115e-01 3.53562325e-01
9.59777474e-01 -2.61188447e-01 3.55053604e-01 1.18120670e+00
-6.09606393e-02 5.73905885e-01 3.62123221e-01 -7.99021602e-01
-1.85718274e+00 -8.93356681e-01 8.53010535e-01 4.18646447e-02
1.30178690e-01 -1.30505502e-01 -8.85312200e-01 1.69535756e-01
8.04465860e-02 7.51392663e-01 4.73236442e-01 -7.80578613e-01
-6.79534554e-01 -3.94853711e-01 -1.68841612e+00 3.05066615e-01
8.82416546e-01 -8.13014388e-01 -2.91973680e-01 2.14923784e-01
3.40713151e-02 -3.08886826e-01 -7.17165828e-01 6.10933244e-01
8.96046638e-01 -9.79248881e-01 1.03989315e+00 -2.06277087e-01
-8.23884681e-02 -4.32821155e-01 -7.30480030e-02 -9.56324041e-01
-6.26448402e-03 -1.32582092e+00 -2.55051702e-01 1.26822412e+00
-1.18808120e-01 -9.13894832e-01 1.07870913e+00 2.74948061e-01
1.40297636e-01 -4.01746988e-01 -1.21393108e+00 -6.81630790e-01
-6.84595823e-01 -6.52736500e-02 5.33561051e-01 8.81317079e-01
1.10598721e-01 -2.19483316e-01 -6.45216405e-01 4.19189811e-01
9.92573798e-01 -7.12159798e-02 4.94302690e-01 -9.83437479e-01
2.32764572e-01 -4.03699428e-01 -9.53801155e-01 -6.80377603e-01
3.35660040e-01 -6.07164562e-01 -2.12156549e-01 -6.81643903e-01
-2.28725106e-01 4.43845950e-02 -4.11502153e-01 2.55936325e-01
8.71280953e-02 9.07828510e-01 1.17904760e-01 -4.08832133e-02
2.10717469e-01 5.43604076e-01 1.33536625e+00 -1.34751201e-01
-3.97433579e-01 -6.49541691e-02 -3.09325665e-01 3.75693023e-01
5.06945372e-01 -2.72873968e-01 -2.07973108e-01 -5.18204272e-02
-6.62575543e-01 4.02200103e-01 3.07861507e-01 -1.02876341e+00
4.53483224e-01 2.66038924e-01 9.74384785e-01 -4.75736707e-01
8.03202152e-01 -7.64356256e-01 9.66961309e-02 7.35161841e-01
-2.26360969e-02 -5.65423034e-02 2.22437173e-01 4.36434120e-01
-4.29792136e-01 9.63371098e-02 1.26400864e+00 3.68686020e-02
-5.07711947e-01 7.35167742e-01 8.94690380e-02 -4.88160282e-01
1.06435382e+00 -6.63683176e-01 -5.16126037e-01 -2.79563874e-01
-3.23361248e-01 -3.11393976e-01 2.54009783e-01 5.19669831e-01
1.38137937e+00 -1.32179093e+00 -8.22919309e-01 1.14243305e+00
-1.41773999e-01 -9.38490272e-01 6.40894294e-01 1.05535305e+00
-4.27419037e-01 3.55742931e-01 -7.12991595e-01 -5.01779079e-01
-1.66817784e+00 5.37482023e-01 5.52404225e-01 1.98640898e-01
-7.00334728e-01 1.07090521e+00 2.73824576e-02 5.47679402e-02
2.74967343e-01 3.95117849e-01 -4.48113620e-01 8.40743333e-02
1.04048419e+00 4.61673766e-01 9.14583430e-02 -1.22823012e+00
-5.28394461e-01 1.08375120e+00 -3.84309590e-02 1.10345624e-01
1.05487347e+00 -2.28867888e-01 -1.73030987e-01 -4.85707015e-01
1.63918090e+00 4.07673605e-02 -1.37352180e+00 -2.09432483e-01
-3.81896257e-01 -1.37754047e+00 2.31103212e-01 -2.59698272e-01
-1.79102385e+00 1.00486457e+00 1.06415808e+00 -2.18752980e-01
1.38898182e+00 -4.92295891e-01 7.91668713e-01 -3.83401960e-01
3.24648708e-01 -6.65954947e-01 3.09328735e-01 -1.32357717e-01
9.94878769e-01 -1.11919594e+00 3.96080781e-03 -7.33457923e-01
-5.08972049e-01 1.43082166e+00 3.92703116e-01 -9.44702625e-02
8.78266037e-01 -7.25475475e-02 7.35138804e-02 -1.58101171e-01
-2.24817023e-01 1.46068573e-01 4.16914254e-01 9.02840018e-01
1.28039330e-01 -3.78122687e-01 5.35572991e-02 -1.27255926e-02
1.16491564e-01 -9.41092521e-02 1.27725497e-01 6.20725691e-01
-1.10392764e-01 -9.76412416e-01 -4.93000537e-01 1.53374329e-01
-8.28852296e-01 2.01736644e-01 -1.52158394e-01 5.97313881e-01
-2.09262401e-01 1.18591845e+00 -1.44636586e-01 -5.73757708e-01
4.25089821e-02 -1.11991808e-01 6.61125422e-01 -9.73557457e-02
-3.92575681e-01 3.49788576e-01 -1.97111666e-01 -9.91392612e-01
-5.23053050e-01 -6.60630524e-01 -7.92311966e-01 -4.74642754e-01
-2.33688712e-01 -2.89755519e-02 8.20976257e-01 4.23949420e-01
4.98106271e-01 -1.04270279e-01 1.45374727e+00 -1.21265709e+00
-2.09235713e-01 -7.72675633e-01 -7.93226898e-01 2.50034273e-01
6.62037134e-01 -7.92772293e-01 -6.27056062e-01 -1.50492936e-01] | [13.053637504577637, 1.2325316667556763] |
feddc8b1-b430-4ac4-b6c9-576d57489465 | input-output-balanced-framework-for-long | 2103.14269 | null | https://arxiv.org/abs/2103.14269v1 | https://arxiv.org/pdf/2103.14269v1.pdf | Input-Output Balanced Framework for Long-tailed LiDAR Semantic Segmentation | A thorough and holistic scene understanding is crucial for autonomous vehicles, where LiDAR semantic segmentation plays an indispensable role. However, most existing methods focus on the network design while neglecting the inherent difficulty, imbalanced data distribution in the realistic dataset (also named long-tailed distribution), which narrows down the capability of state-of-the-art methods. In this paper, we propose an input-output balanced framework to handle the issue of long-tailed distribution. Specifically, for the input space, we synthesize these tailed instances from mesh models and well simulate the position and density distribution of LiDAR scan, which enhances the input data balance and improves the data diversity. For the output space, a multi-head block is proposed to group different categories based on their shapes and instance amounts, which alleviates the biased representation of dominating category during the feature learning. We evaluate the proposed model on two large-scale datasets, SemanticKITTI and nuScenes, where state-of-the-art results demonstrate its effectiveness. The proposed new modules can also be used as a plug-and-play, and we apply them on various backbones and datasets, showing its good generalization ability. | ['Yuexin Ma', 'Xinge Zhu', 'Peishan Cong'] | 2021-03-26 | null | null | null | null | ['lidar-semantic-segmentation'] | ['computer-vision'] | [-5.84633835e-02 -2.57721454e-01 -4.90831554e-01 -6.83054209e-01
-1.60323799e-01 -3.16911191e-01 3.29402536e-01 -4.98167835e-02
-4.35092986e-01 6.31346583e-01 -2.81096518e-01 -2.90432721e-01
-3.91683042e-01 -1.11010468e+00 -7.10062742e-01 -7.46558130e-01
2.74927408e-01 6.97548687e-01 6.59910440e-01 -1.72493398e-01
2.22269878e-01 5.37232161e-01 -2.18324327e+00 4.75576101e-03
1.26227200e+00 1.14861524e+00 3.94876897e-01 -6.93747625e-02
-7.52718747e-01 3.30952525e-01 -5.16565263e-01 -2.41684571e-01
3.08223635e-01 2.59016715e-02 -2.19071671e-01 2.20702793e-02
5.46452105e-01 -2.02606618e-01 -3.80644739e-01 1.14677513e+00
4.64990914e-01 3.73510905e-02 8.96129370e-01 -1.72693264e+00
-2.15679422e-01 3.64649832e-01 -9.68428433e-01 1.57124236e-01
-4.68444198e-01 2.85821706e-01 7.74778247e-01 -7.00090647e-01
4.58031654e-01 1.44500351e+00 5.44109702e-01 3.70866328e-01
-8.29586923e-01 -1.20307958e+00 4.99192864e-01 5.83985269e-01
-1.38485038e+00 -2.29560286e-01 9.74475324e-01 -3.90863717e-01
2.59722531e-01 1.97312236e-02 6.08574986e-01 7.50451505e-01
-1.03399366e-01 1.02700329e+00 7.94537723e-01 2.16756210e-01
1.34294704e-01 3.32261413e-01 1.76096573e-01 3.84876519e-01
6.30434036e-01 -7.33700581e-03 -3.01484048e-01 2.12457970e-01
4.45697874e-01 3.05271834e-01 5.37268370e-02 -7.40663290e-01
-8.45353007e-01 8.40594113e-01 6.55945659e-01 -1.65741500e-02
-4.80233021e-02 -1.12807006e-01 5.68605006e-01 1.20121231e-02
3.04078847e-01 -3.05063576e-01 -4.18993175e-01 1.27243653e-01
-9.55473065e-01 3.86477441e-01 4.75343704e-01 1.17882240e+00
9.45256472e-01 -3.86805348e-02 -1.42498195e-01 8.41503084e-01
3.40782464e-01 5.67013502e-01 3.27796876e-01 -5.95494986e-01
7.33750284e-01 9.00693893e-01 -1.99613795e-01 -1.09497392e+00
-5.87161124e-01 -8.51484060e-01 -1.03981507e+00 2.13193417e-01
2.92549402e-01 -1.47274405e-01 -1.17592621e+00 1.76172030e+00
6.40000999e-01 1.39953136e-01 -4.49120736e-04 9.82370496e-01
1.09257162e+00 5.53402543e-01 1.02939941e-01 -4.52861302e-02
1.30243456e+00 -9.19488847e-01 -5.88529468e-01 -2.56955564e-01
3.82215112e-01 -5.14706910e-01 1.08282125e+00 2.91208953e-01
-5.10027230e-01 -7.85438538e-01 -9.78621185e-01 1.21032506e-01
-4.66144949e-01 1.66593239e-01 6.25290692e-01 6.12396359e-01
-4.98010486e-01 4.67584878e-01 -4.46035892e-01 -3.46215636e-01
1.01204836e+00 1.07358642e-01 -8.47021341e-02 -3.59336972e-01
-1.09859192e+00 5.90837717e-01 6.86047673e-01 1.76161602e-01
-6.74168169e-01 -8.41564596e-01 -6.36936188e-01 -2.52168160e-02
6.07189596e-01 -5.73711991e-01 7.68347025e-01 -6.06663764e-01
-9.27846372e-01 5.96541345e-01 -2.00482998e-02 -2.04790622e-01
7.83319712e-01 8.53822380e-02 -4.27321047e-01 -9.44066420e-02
1.24442838e-01 8.24135721e-01 8.26560080e-01 -1.53601515e+00
-1.00571167e+00 -6.50678217e-01 -6.18990278e-03 3.13902915e-01
-4.91589129e-01 -6.11444712e-01 -6.14347219e-01 -4.05785620e-01
2.92508543e-01 -6.59720242e-01 -1.86564535e-01 3.10157090e-01
-3.82985026e-01 -2.86976010e-01 1.09946418e+00 -1.40784025e-01
1.23509359e+00 -2.48046660e+00 -2.30226740e-01 2.76069731e-01
2.19812006e-01 2.24432245e-01 -6.62758276e-02 2.71210849e-01
1.68973222e-01 7.91576207e-02 -2.94988900e-01 -2.90898591e-01
3.62983011e-02 5.12073457e-01 -8.19334667e-03 4.63175386e-01
2.29321226e-01 5.68470418e-01 -7.96247780e-01 -8.01330566e-01
2.98058093e-01 2.24780023e-01 -5.25776625e-01 9.32046846e-02
-3.32044989e-01 5.52686155e-01 -7.46263325e-01 7.89065897e-01
1.35475767e+00 7.18966275e-02 -1.65531725e-01 -4.60781425e-01
-9.02503058e-02 -1.48254171e-01 -1.43223870e+00 1.57658410e+00
-3.48365396e-01 2.29278818e-01 3.62657160e-02 -9.09929395e-01
1.12409997e+00 -3.60308200e-01 3.94812167e-01 -9.03411090e-01
2.49281570e-01 2.73558915e-01 1.92095816e-01 -6.75558090e-01
5.60116589e-01 -1.31554872e-01 1.15632201e-02 -2.11422634e-03
-2.93439716e-01 -7.61949047e-02 3.30793440e-01 8.97921715e-03
4.45699990e-01 1.28578857e-01 -9.78458300e-02 -3.58689934e-01
4.41237718e-01 8.63554776e-02 6.79242790e-01 5.08465648e-01
-2.05775723e-01 5.59557199e-01 6.18974328e-01 -3.23989421e-01
-8.14824402e-01 -9.36549306e-01 -4.67110217e-01 9.67969716e-01
7.96868861e-01 -1.27130728e-02 -5.76710045e-01 -7.36958623e-01
4.77064878e-01 7.27302790e-01 -5.87456822e-01 -2.14971766e-01
-2.42441133e-01 -8.91844094e-01 5.65461576e-01 6.34274065e-01
7.41343319e-01 -1.02729356e+00 -6.29190803e-01 -7.17984289e-02
-8.63998458e-02 -9.64450598e-01 -6.60615116e-02 1.81528315e-01
-8.76145959e-01 -1.15478623e+00 -4.03647989e-01 -6.88247323e-01
5.98719001e-01 5.19128978e-01 1.03270471e+00 1.05191514e-01
-1.72197253e-01 -3.89420390e-01 -2.42118418e-01 -5.73704422e-01
1.67528819e-02 4.02214676e-01 -8.96900818e-02 1.14869170e-01
4.17031437e-01 -8.47770154e-01 -8.17902505e-01 6.22888684e-01
-1.10649467e+00 3.59579101e-02 7.09329844e-01 5.67130387e-01
6.51403129e-01 3.55538040e-01 8.25661719e-01 -9.59631681e-01
2.11808398e-01 -7.73123085e-01 -5.19432008e-01 -1.02433443e-01
-6.78075433e-01 -1.38402760e-01 5.89037001e-01 -3.79577249e-01
-9.31956589e-01 -1.27692983e-01 -2.77241588e-01 -6.04905486e-01
-3.47841650e-01 3.56108487e-01 -7.21808016e-01 8.39079469e-02
2.62481332e-01 1.96066469e-01 1.06112011e-01 -5.74164629e-01
4.30998653e-01 8.50138783e-01 4.51554924e-01 -6.47933125e-01
9.57681239e-01 5.34290671e-01 1.46703720e-01 -6.71418071e-01
-6.68360412e-01 -5.19105256e-01 -3.87512982e-01 -1.69806734e-01
4.97422725e-01 -1.04828691e+00 -6.66132510e-01 6.64532602e-01
-8.58762980e-01 -8.77325702e-03 -3.02568913e-01 2.77876526e-01
-2.59871960e-01 3.90499607e-02 2.21534036e-02 -8.24434161e-01
-1.82526186e-02 -1.20441794e+00 1.00890458e+00 5.80370188e-01
4.43524450e-01 -6.18577302e-01 -4.32407558e-01 2.50512004e-01
5.30420721e-01 3.31344664e-01 1.18340409e+00 -7.11843491e-01
-7.43443787e-01 3.28962132e-02 -6.96285009e-01 3.15696150e-01
6.12734705e-02 1.21522494e-01 -1.10362196e+00 -1.79613352e-01
-2.47500956e-01 -2.78214246e-01 1.03341699e+00 2.68447310e-01
1.59108770e+00 4.74370308e-02 -4.41616535e-01 7.36308515e-01
1.57303393e+00 6.49533123e-02 6.05146408e-01 2.77051926e-01
9.07488108e-01 8.69103014e-01 8.66039515e-01 3.94632131e-01
7.74308264e-01 3.08354735e-01 1.01764381e+00 -2.56636590e-01
-1.55183792e-01 -3.86213422e-01 -2.77262598e-01 6.22236848e-01
3.06347340e-01 -4.84565228e-01 -6.66061044e-01 7.17427194e-01
-1.91380703e+00 -6.57673419e-01 -3.58899385e-01 2.06145191e+00
4.34175551e-01 4.58687782e-01 1.86415434e-01 1.10257417e-01
8.57295156e-01 3.93919736e-01 -9.90588963e-01 1.53735310e-01
-1.83727175e-01 -1.86799839e-01 5.89863718e-01 -1.66861415e-02
-9.73426521e-01 8.21532845e-01 5.15741539e+00 1.63055182e+00
-1.13450634e+00 1.09333068e-01 6.73820615e-01 -8.55701268e-02
-5.02189040e-01 -2.52920777e-01 -9.43320572e-01 7.46604800e-01
3.85844618e-01 3.06731765e-03 8.22289661e-02 9.94080722e-01
1.82013065e-01 -1.80096880e-01 -8.83528352e-01 1.04117072e+00
-5.99168390e-02 -1.11199367e+00 2.34010443e-01 1.31573100e-02
5.86065412e-01 2.43656099e-01 3.54417674e-02 5.26237667e-01
-1.34593979e-01 -9.10383344e-01 7.99393058e-01 3.22595567e-01
9.05942798e-01 -1.05953991e+00 8.86594713e-01 6.70920551e-01
-1.45534742e+00 -2.63846308e-01 -6.11924350e-01 1.01280659e-01
4.20950679e-03 8.15057576e-01 -3.10798466e-01 9.67961550e-01
7.84199655e-01 6.22094452e-01 -7.19502509e-01 1.27452135e+00
1.07911631e-01 4.10112351e-01 -5.49090087e-01 1.01880580e-02
2.20401451e-01 -2.85124451e-01 3.49709332e-01 1.05106902e+00
2.44280651e-01 -3.03239048e-01 4.51851189e-01 7.27637589e-01
-2.15817168e-01 1.80663109e-01 -5.55667639e-01 2.84698039e-01
7.25870013e-01 1.39793885e+00 -9.25179183e-01 -3.45015317e-01
-2.92201608e-01 1.65334985e-01 1.22494236e-01 2.04248860e-01
-9.46074486e-01 -3.53060037e-01 6.74387097e-01 1.78251922e-01
5.50621569e-01 1.61493018e-01 -6.18055284e-01 -8.28373611e-01
2.21742123e-01 -7.98136115e-01 3.55013013e-01 -3.78162950e-01
-1.37868476e+00 5.75232208e-01 2.03783900e-01 -1.57426524e+00
2.22255811e-01 -4.73008066e-01 -6.31703496e-01 7.20835507e-01
-1.90146768e+00 -1.19737422e+00 -8.54386747e-01 5.21181762e-01
6.24669671e-01 -6.86195418e-02 -1.20114246e-02 8.86320829e-01
-6.93527222e-01 6.63702607e-01 3.13885733e-02 -2.37619594e-01
6.38024569e-01 -8.67758155e-01 2.41329327e-01 5.52019536e-01
-2.44288445e-01 3.51757497e-01 6.40484273e-01 -5.25250554e-01
-1.14598119e+00 -1.51044583e+00 1.59952700e-01 -5.99737391e-02
4.07303959e-01 -3.05162489e-01 -1.00318027e+00 1.23424068e-01
-1.30092800e-01 1.33700550e-01 2.67875671e-01 -9.14892703e-02
-2.10890815e-01 -5.73556542e-01 -1.37575293e+00 3.88061106e-01
1.35211182e+00 6.80268509e-03 -3.40651482e-01 1.85486987e-01
8.41918826e-01 -4.44114298e-01 -4.38473046e-01 8.77599597e-01
6.31562352e-01 -1.14885008e+00 6.68241918e-01 -5.03684402e-01
5.43700099e-01 -6.92971647e-01 -2.61415929e-01 -1.22434914e+00
-1.82110757e-01 1.46273160e-02 1.02903545e-02 1.47529626e+00
2.89112091e-01 -7.06684113e-01 9.28806901e-01 -1.48645602e-02
-2.67024040e-01 -1.14649951e+00 -8.51055861e-01 -6.58881068e-01
-1.99829582e-02 -4.40963238e-01 1.08755803e+00 6.55543447e-01
-8.54811549e-01 2.19104707e-01 -1.39983250e-02 2.18015358e-01
5.64123690e-01 3.32611710e-01 1.08984005e+00 -1.57263732e+00
1.15139626e-01 -4.69583303e-01 -5.26536524e-01 -1.20376563e+00
-1.36833385e-01 -8.00975561e-01 7.35994279e-02 -1.44913816e+00
2.48758793e-01 -1.00063396e+00 -2.40707591e-01 1.64027885e-01
-2.28263572e-01 3.08230728e-01 1.84695274e-01 1.14173375e-01
-6.00468397e-01 9.55112576e-01 1.39175570e+00 -2.00032249e-01
-2.39958577e-02 1.83995053e-01 -8.65735173e-01 8.57486188e-01
7.31990576e-01 -5.98755538e-01 -8.16145837e-01 -4.34495986e-01
9.70540047e-02 -3.62788886e-01 3.78782868e-01 -1.21164787e+00
3.14661086e-01 -3.86221945e-01 3.02828401e-01 -1.18929088e+00
1.70549408e-01 -1.10675681e+00 1.28518656e-01 1.96553901e-01
1.65923178e-01 -3.07370666e-02 4.02214646e-01 6.53409779e-01
-2.02845290e-01 -2.49302894e-01 8.56348395e-01 4.85660583e-02
-7.73776889e-01 7.74427176e-01 2.39363208e-01 3.38070095e-01
1.05623877e+00 -4.50867325e-01 -4.69765872e-01 -3.02744452e-02
-1.51406735e-01 7.66457319e-01 5.92010260e-01 4.99252439e-01
4.61691201e-01 -1.36136317e+00 -6.27126455e-01 4.67678756e-01
4.94954228e-01 7.63142288e-01 6.98943079e-01 7.61363268e-01
-4.14160222e-01 -1.88366219e-01 -3.32527131e-01 -1.02460444e+00
-8.74360502e-01 5.29235423e-01 1.82969481e-01 -2.94704083e-02
-5.25826335e-01 6.33913875e-01 4.94499624e-01 -7.68506885e-01
3.53811860e-01 -2.84266591e-01 -4.56675142e-01 4.85656053e-01
2.03183681e-01 3.52173865e-01 4.39897217e-02 -4.80753601e-01
-4.17098731e-01 7.20861197e-01 3.15990299e-02 4.48646992e-01
1.19735932e+00 -2.26325735e-01 1.44379139e-02 5.26446640e-01
9.54398215e-01 -7.61796236e-02 -1.35664058e+00 -1.65515319e-01
-3.62459630e-01 -6.09094501e-01 -1.41734093e-01 -4.90815848e-01
-1.54822803e+00 1.02181566e+00 7.29256690e-01 1.27926379e-01
9.70925748e-01 -1.12600297e-01 9.20857251e-01 1.77451134e-01
5.24016500e-01 -1.04869485e+00 -2.26963833e-01 4.09032375e-01
4.95685518e-01 -1.23252797e+00 -1.39581403e-02 -7.47493386e-01
-6.02334797e-01 8.58160794e-01 1.00761104e+00 -8.26038793e-02
6.87964737e-01 2.16004401e-01 1.31329685e-01 -2.04945490e-01
-4.18757170e-01 -4.07325596e-01 -1.41680557e-02 7.25604773e-01
-1.26235306e-01 5.19230701e-02 -3.30596089e-01 6.49745047e-01
-2.53954619e-01 -7.89448917e-02 1.90332800e-01 6.65739298e-01
-6.03370845e-01 -7.91575372e-01 -2.19022661e-01 7.72327006e-01
-1.03214933e-02 1.41873702e-01 1.60759464e-01 1.02958691e+00
8.42144847e-01 7.92698264e-01 2.45334014e-01 -4.90050644e-01
6.41116381e-01 -1.96801946e-01 1.65928051e-01 -2.95461535e-01
-1.60513490e-01 -1.63818195e-01 4.80998633e-03 -4.30877388e-01
-3.68627578e-01 -3.51878941e-01 -1.23698938e+00 -3.89487803e-01
-3.79320174e-01 -3.60149634e-03 7.39294231e-01 9.58263576e-01
3.48750561e-01 8.99198949e-01 7.61590958e-01 -9.10471797e-01
-5.56900442e-01 -1.01677430e+00 -6.89999640e-01 3.44245166e-01
2.31368706e-01 -1.10567403e+00 -4.27914143e-01 -4.32354987e-01] | [8.02263069152832, -2.818971872329712] |
055ad453-50f5-4243-9546-ddb2e0094f7d | efficient-end-to-end-video-question-answering | 2302.02136 | null | https://arxiv.org/abs/2302.02136v2 | https://arxiv.org/pdf/2302.02136v2.pdf | Efficient End-to-End Video Question Answering with Pyramidal Multimodal Transformer | This paper presents a new method for end-to-end Video Question Answering (VideoQA), aside from the current popularity of using large-scale pre-training with huge feature extractors. We achieve this with a pyramidal multimodal transformer (PMT) model, which simply incorporates a learnable word embedding layer, a few convolutional and transformer layers. We use the anisotropic pyramid to fulfill video-language interactions across different spatio-temporal scales. In addition to the canonical pyramid, which includes both bottom-up and top-down pathways with lateral connections, novel strategies are proposed to decompose the visual feature stream into spatial and temporal sub-streams at different scales and implement their interactions with the linguistic semantics while preserving the integrity of local and global semantics. We demonstrate better or on-par performances with high computational efficiency against state-of-the-art methods on five VideoQA benchmarks. Our ablation study shows the scalability of our model that achieves competitive results for text-to-video retrieval by leveraging feature extractors with reusable pre-trained weights, and also the effectiveness of the pyramid. | ['Xiang-Dong Zhou', 'Yu Shi', 'Chongyang Wang', 'Min Peng'] | 2023-02-04 | null | null | null | null | ['video-question-answering', 'video-retrieval'] | ['computer-vision', 'computer-vision'] | [-9.35454741e-02 -2.75023729e-01 -5.50031066e-02 -3.28616887e-01
-1.15435731e+00 -6.04156077e-01 5.74994802e-01 -4.29100841e-02
-6.57673061e-01 6.63107112e-02 7.16055274e-01 -1.90789986e-03
-1.21146746e-01 -6.56360984e-01 -7.47140884e-01 -4.29637104e-01
-1.82836831e-01 1.71760857e-01 6.29348457e-01 -4.96633559e-01
1.15698747e-01 2.35740945e-01 -1.64326286e+00 1.03943360e+00
4.76896018e-01 1.43953454e+00 4.90973666e-02 9.64918315e-01
-1.64506733e-01 1.08548629e+00 -2.54094183e-01 -6.32613003e-01
1.30217940e-01 -2.39555761e-01 -7.50102162e-01 3.60401124e-02
9.31262136e-01 -9.08610642e-01 -1.01102054e+00 5.32919765e-01
5.38319886e-01 -1.97747406e-02 4.07644689e-01 -1.06188178e+00
-1.06047308e+00 1.22472703e-01 -3.78319800e-01 2.81769127e-01
6.10507250e-01 3.39534342e-01 1.58876300e+00 -1.10326409e+00
8.64476979e-01 1.31691468e+00 5.49311638e-01 3.71039748e-01
-8.66610050e-01 -2.34321207e-01 4.73704487e-01 7.49071717e-01
-1.21458256e+00 -4.70766902e-01 5.73743880e-01 -3.49476874e-01
1.30860829e+00 8.30437616e-02 6.42563820e-01 1.20266724e+00
3.14028472e-01 1.18027890e+00 4.32790577e-01 1.27952278e-01
-6.20273734e-03 -2.79561013e-01 -2.58116145e-02 1.04903662e+00
-4.15423065e-01 -3.06503832e-01 -1.10456443e+00 -1.37255639e-01
6.32782280e-01 3.03019196e-01 -2.86195874e-01 -6.68954015e-01
-1.38491344e+00 7.35009968e-01 6.54485166e-01 2.64982998e-01
-4.19749886e-01 5.71212947e-01 6.95421159e-01 5.72315156e-01
3.31188053e-01 1.58909649e-01 -5.32234013e-01 3.13136540e-02
-1.15242624e+00 2.37148508e-01 4.33917731e-01 9.90523458e-01
6.50235355e-01 -8.37460235e-02 -8.45933676e-01 6.23944163e-01
2.21458197e-01 4.80612695e-01 4.46769059e-01 -1.07587969e+00
7.45834112e-01 6.81109786e-01 5.95226549e-02 -1.01896095e+00
-2.92339474e-01 -1.84479728e-01 -5.97756624e-01 -1.09826744e-01
2.72686929e-01 3.53078693e-01 -1.01979685e+00 1.63636243e+00
1.03722088e-01 1.71841800e-01 -8.93415064e-02 1.12746739e+00
1.08149326e+00 8.00212622e-01 1.57927796e-01 1.10658929e-01
1.62138462e+00 -1.43940794e+00 -7.01858103e-01 8.24833810e-02
4.43772614e-01 -4.74118829e-01 1.41924739e+00 1.23413280e-01
-1.47514558e+00 -5.24958014e-01 -8.68390560e-01 -9.19030070e-01
-5.48790932e-01 1.90226942e-01 5.85580587e-01 1.33080006e-01
-1.50487459e+00 3.49865198e-01 -7.91360497e-01 -4.72849011e-01
5.73070586e-01 3.35252017e-01 -6.27742112e-01 -2.81725705e-01
-1.22863078e+00 4.86504555e-01 -2.06206575e-01 3.59556638e-02
-1.11597955e+00 -7.17966139e-01 -9.43078578e-01 3.58035445e-01
3.70294839e-01 -1.17957902e+00 1.00346863e+00 -9.47138786e-01
-1.60446918e+00 7.00264215e-01 -4.35156107e-01 -4.46879715e-01
3.32077503e-01 -4.56065595e-01 -6.65553212e-02 1.04637706e+00
3.93865481e-02 9.83506441e-01 1.23664761e+00 -7.16300905e-01
-5.68999410e-01 -4.89900351e-01 4.29301649e-01 2.98476309e-01
-8.45260084e-01 2.69125283e-01 -1.15821123e+00 -7.78489769e-01
-7.36754686e-02 -5.07245779e-01 4.35250998e-02 5.39750040e-01
2.03141734e-01 -2.73882508e-01 9.45185781e-01 -9.21987534e-01
1.31160235e+00 -2.13874030e+00 6.05917156e-01 -5.25708757e-02
4.84154433e-01 -1.74940638e-02 -6.33584976e-01 6.64317250e-01
2.63763666e-01 -1.68349415e-01 -8.54855776e-02 -4.46415693e-01
1.29694313e-01 3.02628949e-02 -5.23712099e-01 3.57911408e-01
4.93116677e-01 1.47956896e+00 -9.47647214e-01 -4.36828494e-01
1.47929832e-01 6.94071233e-01 -1.00982702e+00 2.59337246e-01
-2.71188647e-01 -6.93852305e-02 -6.56965077e-01 7.46833563e-01
3.23223293e-01 -4.68591452e-01 6.93437038e-03 -5.42690575e-01
1.13045141e-01 2.79969394e-01 -6.29208624e-01 2.31046700e+00
-2.71939158e-01 6.82508111e-01 3.35946262e-01 -9.15772974e-01
1.92338124e-01 4.47024226e-01 5.91039658e-01 -1.14710486e+00
-1.40213832e-01 1.19661070e-01 -6.62632942e-01 -6.68931961e-01
6.38830841e-01 2.53081053e-01 -5.54388128e-02 1.62165269e-01
7.18865335e-01 3.97433080e-02 3.21741432e-01 7.21203029e-01
1.36164320e+00 4.50472295e-01 -2.12422296e-01 -1.27121434e-01
5.13819456e-01 -1.65319607e-01 5.48427925e-02 6.32010341e-01
-1.84329450e-01 9.50996041e-01 6.41144514e-01 -5.70675671e-01
-9.50814843e-01 -1.17976832e+00 3.26484859e-01 1.74768686e+00
8.10617134e-02 -8.55491400e-01 -6.21211410e-01 -7.45331347e-01
-7.03521371e-02 6.78250343e-02 -8.10868800e-01 -1.74078807e-01
-6.53029621e-01 -1.89250067e-01 6.86736941e-01 5.40005267e-01
6.41359925e-01 -9.30249393e-01 -7.49004304e-01 9.23625529e-02
-3.96789342e-01 -1.39907277e+00 -8.18399966e-01 -2.24478558e-01
-7.04219460e-01 -9.53187585e-01 -1.07160115e+00 -7.74756312e-01
3.99574250e-01 5.01939356e-01 1.30610836e+00 1.15872189e-01
-1.80544376e-01 9.40245450e-01 -5.70746481e-01 1.35203511e-01
3.31538469e-01 7.18231797e-02 -3.56123805e-01 2.31639832e-01
1.24672584e-01 -5.43906033e-01 -1.03169918e+00 2.57624298e-01
-1.30160892e+00 1.36022521e-02 4.07274783e-01 8.89435291e-01
5.36383390e-01 -6.02742195e-01 3.21327686e-01 -3.09994876e-01
5.38104415e-01 -3.20711315e-01 -3.36603343e-01 5.36781609e-01
-3.47778993e-03 1.51012987e-02 5.41861832e-01 -2.22839221e-01
-8.33697617e-01 -9.49696675e-02 -2.77937829e-01 -7.50628948e-01
2.34256715e-01 3.38590294e-01 1.86137315e-02 -6.35319576e-02
4.24151212e-01 3.29349846e-01 -7.96052441e-02 -2.67114997e-01
7.43520379e-01 2.74905622e-01 4.01812583e-01 -4.18697804e-01
7.91917384e-01 9.38222468e-01 -3.09012890e-01 -7.27844775e-01
-8.48697007e-01 -6.97429299e-01 -6.33361280e-01 -1.88441589e-01
1.26053512e+00 -1.26470208e+00 -6.67803645e-01 3.60940158e-01
-1.21952355e+00 -2.08623827e-01 -3.34855765e-01 1.42891273e-01
-6.24701321e-01 5.11714458e-01 -1.09334135e+00 -2.23962098e-01
-5.23813903e-01 -1.11936843e+00 1.65839517e+00 -1.61260396e-01
1.99656710e-01 -7.08429098e-01 -1.32868648e-01 6.32879436e-01
7.47907579e-01 -1.42028421e-01 8.15314472e-01 -2.67078578e-01
-9.81134057e-01 -1.20739518e-02 -5.30482888e-01 1.84526354e-01
-3.86551708e-01 -1.04226947e-01 -9.39472497e-01 -4.88384187e-01
-1.38992220e-01 -6.44102871e-01 1.48568451e+00 2.74502575e-01
1.26481652e+00 -3.24781299e-01 -1.37360608e-02 7.51162648e-01
1.21787143e+00 -4.27863479e-01 8.40999484e-01 2.03587815e-01
7.97908723e-01 5.26261032e-01 2.43563876e-01 2.76124716e-01
7.75647461e-01 6.79943204e-01 5.32872796e-01 -1.75830454e-01
-3.28540802e-01 -2.75806218e-01 7.45399833e-01 8.39940369e-01
-9.02982578e-02 -2.54755110e-01 -6.65752172e-01 7.41666138e-01
-2.14585638e+00 -1.12714398e+00 3.01744878e-01 1.80306208e+00
5.65130353e-01 -1.53678522e-01 1.96093485e-01 -2.43549585e-01
-5.46049774e-02 6.01392806e-01 -3.83703589e-01 -2.04662293e-01
-2.71773279e-01 1.93047062e-01 1.98167115e-01 3.42211038e-01
-1.11089373e+00 1.13281476e+00 6.64430046e+00 9.04701471e-01
-1.11698246e+00 3.09229851e-01 4.27720279e-01 -4.99347299e-01
-5.54105282e-01 -3.79089504e-01 -3.88669491e-01 -1.62273496e-02
8.94750476e-01 2.63676971e-01 4.14473683e-01 3.97717535e-01
1.87353033e-03 2.73451895e-01 -1.04313111e+00 9.57564771e-01
3.38777810e-01 -1.64737105e+00 6.55424714e-01 -2.45298609e-01
5.34568071e-01 3.68813723e-01 1.64794400e-01 4.06489521e-01
-1.95712835e-01 -8.80197227e-01 9.62952197e-01 5.36930203e-01
8.53369832e-01 -4.14405257e-01 4.90355313e-01 -1.49108112e-01
-1.51112986e+00 -3.53181303e-01 -3.89093518e-01 1.56020492e-01
2.81227797e-01 1.22119799e-01 -6.88266158e-02 6.45450234e-01
1.02053618e+00 9.95635152e-01 -6.08945906e-01 7.97731340e-01
-5.82889691e-02 3.58302414e-01 -2.68552691e-01 1.94743976e-01
6.10548079e-01 -2.34439857e-02 4.14154887e-01 1.46172607e+00
2.85177976e-01 -1.07604517e-02 -1.11395024e-01 4.52613145e-01
-4.10063237e-01 2.01154768e-01 -5.64286947e-01 -1.94448888e-01
-1.88004330e-01 1.18722761e+00 -2.81586051e-01 -4.51350927e-01
-8.17888737e-01 1.36938143e+00 5.26780009e-01 8.82492781e-01
-9.13027167e-01 -2.60880679e-01 7.15930462e-01 1.83411688e-01
8.47623706e-01 -3.70027393e-01 1.64768949e-01 -1.51872158e+00
3.81213397e-01 -9.67976868e-01 7.69596815e-01 -9.48958874e-01
-1.20335984e+00 6.77212238e-01 -1.51101783e-01 -1.08351672e+00
1.64757259e-02 -7.86777854e-01 -2.79332846e-01 4.92100209e-01
-1.88571942e+00 -1.61954761e+00 -3.46292406e-01 1.10671580e+00
7.20997334e-01 -1.09761402e-01 7.81925678e-01 5.32038569e-01
-1.36852279e-01 6.54027522e-01 3.80524360e-02 9.64663252e-02
7.73591638e-01 -9.67803538e-01 2.48071581e-01 6.35983825e-01
4.38780546e-01 5.38176954e-01 1.56802014e-01 -5.31364083e-02
-1.86687005e+00 -9.12551761e-01 8.82779181e-01 -6.37937427e-01
9.82999563e-01 -7.44890809e-01 -7.31679142e-01 7.31942058e-01
5.82871318e-01 2.80640781e-01 4.47862685e-01 -8.51740409e-03
-9.09719050e-01 -2.90253043e-01 -8.08632672e-01 7.53250241e-01
1.18122041e+00 -1.10981536e+00 -5.82843721e-01 3.30860674e-01
1.07876778e+00 -2.21569330e-01 -7.29124069e-01 3.75426263e-01
7.93586373e-01 -1.10101581e+00 1.20455194e+00 -9.22599077e-01
6.93927884e-01 -2.73888737e-01 -2.89067805e-01 -7.66871035e-01
-3.88901234e-01 -6.55416131e-01 -3.25297445e-01 7.65013695e-01
3.74986708e-01 -1.33381978e-01 5.61685979e-01 3.74950022e-01
-6.20038845e-02 -9.37082767e-01 -1.12525141e+00 -2.94694334e-01
-8.24243203e-02 -5.39062500e-01 1.23979501e-01 6.13689601e-01
3.90456691e-02 4.77423221e-01 -4.78867114e-01 -7.70837301e-03
3.14623028e-01 7.79598802e-02 4.90642965e-01 -6.09089434e-01
-2.22320423e-01 -5.22001266e-01 -5.04371226e-01 -1.65890396e+00
8.69737007e-03 -7.79613972e-01 -1.66284844e-01 -1.80667567e+00
3.55849028e-01 4.40349907e-01 -6.44701421e-01 3.53350729e-01
-4.58689444e-02 5.93166649e-01 2.95393020e-01 1.40002817e-01
-1.46544278e+00 9.58724439e-01 1.43206477e+00 -2.97202796e-01
6.56511113e-02 -6.95336580e-01 -4.34187442e-01 5.11077642e-01
2.53093541e-01 -1.52014092e-01 -4.95544225e-01 -1.15883684e+00
3.96640658e-01 -2.29617916e-02 5.29019773e-01 -7.28648126e-01
3.99118751e-01 2.75789797e-01 2.02223703e-01 -4.55452532e-01
6.51957810e-01 -7.97540784e-01 -5.07606208e-01 7.66137466e-02
-4.06851798e-01 2.63839811e-01 2.59556741e-01 7.01983571e-01
-6.49681211e-01 4.41639334e-01 2.99459815e-01 -3.99623141e-02
-8.62820268e-01 6.32827818e-01 -3.14019978e-01 3.46724130e-02
7.79002309e-01 -4.08847518e-02 -3.84211749e-01 -6.43599093e-01
-6.83073699e-01 6.75330758e-01 2.66233504e-01 6.99079514e-01
1.00034821e+00 -1.26263094e+00 -7.81327128e-01 5.32595068e-02
3.91955644e-01 -4.89856064e-01 5.84374011e-01 1.06372178e+00
-5.45029640e-01 7.33565152e-01 -1.87019169e-01 -5.95802546e-01
-1.21925342e+00 5.56427538e-01 3.40506524e-01 -5.07122993e-01
-7.03062296e-01 9.25323725e-01 5.09909928e-01 -1.72477752e-01
5.46733439e-01 -3.90147299e-01 -2.16978759e-01 2.10046932e-01
7.26168871e-01 4.14806977e-02 1.07978694e-02 -5.64330578e-01
-4.36014682e-01 7.64386177e-01 2.31569298e-02 -1.60444453e-01
1.39309990e+00 -2.37736374e-01 -1.40506729e-01 2.38036126e-01
1.39235890e+00 -1.34012744e-01 -1.30682695e+00 -3.27672064e-01
-2.47402906e-01 -3.81466120e-01 9.17681605e-02 -5.68921685e-01
-1.19184804e+00 1.21777427e+00 4.55135733e-01 -2.96921190e-02
1.30299520e+00 1.12510473e-01 1.07028842e+00 6.45832181e-01
2.57210523e-01 -1.09664977e+00 5.62190771e-01 6.99748635e-01
1.21693373e+00 -1.14767992e+00 -2.41590783e-01 -2.59130020e-02
-6.67746186e-01 1.05055499e+00 2.96060383e-01 -3.55504036e-01
6.64020658e-01 -1.66839495e-01 -5.80018759e-02 -4.38659906e-01
-1.17359865e+00 -3.69394243e-01 6.44499004e-01 2.22840160e-01
3.72258902e-01 -4.10601974e-01 -1.69121325e-01 5.18744946e-01
4.36875135e-01 3.59475315e-02 -1.30760759e-01 7.60181367e-01
-3.41302335e-01 -8.79405618e-01 1.18923657e-01 3.52964312e-01
-5.90862393e-01 -4.04182374e-01 -2.69336730e-01 6.37507677e-01
-1.77728161e-01 8.34892333e-01 9.81443748e-02 -3.79213691e-01
5.13714492e-01 1.65832207e-01 4.54180449e-01 -2.06149355e-01
-8.49332392e-01 2.04285786e-01 1.43477932e-01 -1.44176602e+00
-5.75773656e-01 -2.47157484e-01 -1.10840595e+00 -5.11781946e-02
2.06140295e-01 -8.67390037e-02 3.01651925e-01 8.83731306e-01
8.40257466e-01 5.43085456e-01 1.87436968e-01 -8.40385556e-01
-3.64039987e-01 -6.70535982e-01 -3.28815848e-01 6.47562563e-01
5.67808092e-01 -3.03882033e-01 -2.31083617e-01 1.70055017e-01] | [10.407843589782715, 1.0252203941345215] |
9259ed49-2b5e-45cb-8270-3f9a57aa559e | deep-causal-learning-representation-discovery | 2211.03374 | null | https://arxiv.org/abs/2211.03374v1 | https://arxiv.org/pdf/2211.03374v1.pdf | Deep Causal Learning: Representation, Discovery and Inference | Causal learning has attracted much attention in recent years because causality reveals the essential relationship between things and indicates how the world progresses. However, there are many problems and bottlenecks in traditional causal learning methods, such as high-dimensional unstructured variables, combinatorial optimization problems, unknown intervention, unobserved confounders, selection bias and estimation bias. Deep causal learning, that is, causal learning based on deep neural networks, brings new insights for addressing these problems. While many deep learning-based causal discovery and causal inference methods have been proposed, there is a lack of reviews exploring the internal mechanism of deep learning to improve causal learning. In this article, we comprehensively review how deep learning can contribute to causal learning by addressing conventional challenges from three aspects: representation, discovery, and inference. We point out that deep causal learning is important for the theoretical extension and application expansion of causal science and is also an indispensable part of general artificial intelligence. We conclude the article with a summary of open issues and potential directions for future work. | ['Daniel Dajun Zeng', 'Hu Tian', 'Xiaolong Zheng', 'Zizhen Deng'] | 2022-11-07 | null | null | null | null | ['selection-bias'] | ['natural-language-processing'] | [ 1.72887579e-01 1.32815227e-01 -9.20035601e-01 -5.38540244e-01
-4.04442370e-01 -1.25020728e-01 6.82769179e-01 1.36973560e-01
2.78910808e-03 1.31474066e+00 9.24252391e-01 -5.56066871e-01
-8.20265889e-01 -1.11741078e+00 -8.30894649e-01 -7.20157444e-01
-5.41947484e-01 4.04515147e-01 -2.86983252e-01 9.68896002e-02
2.09185600e-01 1.57185778e-01 -1.05108476e+00 2.29880214e-01
1.00851059e+00 4.62455511e-01 -1.82395712e-01 4.06255275e-01
4.36617471e-02 1.17518222e+00 -4.74835992e-01 -3.97221595e-01
-3.77841383e-01 -5.57461143e-01 -1.00706673e+00 -9.01129007e-01
2.54353583e-01 -5.12038827e-01 -7.05798805e-01 7.25999832e-01
6.29128337e-01 -2.05169946e-01 7.09209442e-01 -1.41778529e+00
-1.24323773e+00 1.35708797e+00 -6.57732189e-01 4.01024938e-01
3.24382544e-01 -2.05454640e-02 1.18751454e+00 -5.67000985e-01
4.03175384e-01 1.89236975e+00 5.79218209e-01 3.84938121e-01
-9.95833457e-01 -1.21082711e+00 3.87731791e-01 6.82278275e-01
-6.97135150e-01 -1.20783709e-01 7.82362223e-01 -5.35632491e-01
4.52045798e-01 1.20325051e-01 5.10128617e-01 1.59013891e+00
3.94882202e-01 8.44813943e-01 8.87011468e-01 -2.44984418e-01
1.95502326e-01 -7.76500881e-01 2.19396129e-01 6.11415148e-01
4.23471004e-01 8.65256131e-01 -6.43256366e-01 -3.55548054e-01
1.03557086e+00 6.59063309e-02 -8.87170285e-02 -2.07118794e-01
-1.61115956e+00 1.36026859e+00 7.67404318e-01 4.45008650e-02
-3.80621910e-01 9.36917901e-01 3.58741850e-01 1.21789448e-01
2.60516196e-01 5.41917622e-01 -6.45782828e-01 1.64193109e-01
-5.05736947e-01 5.26198566e-01 5.92801690e-01 4.75373238e-01
2.91628867e-01 1.76227182e-01 -3.52406144e-01 5.16764700e-01
2.87089586e-01 6.59642041e-01 -1.36862233e-01 -8.58058035e-01
3.23399365e-01 6.01788640e-01 5.38875498e-02 -1.11888516e+00
-7.59070575e-01 -1.56726107e-01 -1.42318416e+00 -2.41939962e-01
3.10489774e-01 -7.54583001e-01 -7.00883031e-01 1.89851165e+00
4.16107625e-01 5.32455623e-01 -2.47640744e-01 1.08497930e+00
1.32110691e+00 5.62913179e-01 4.73039061e-01 -3.41589421e-01
1.11896038e+00 -3.08785349e-01 -1.21998167e+00 -1.99586395e-02
1.04168653e-01 -4.72168505e-01 7.22863376e-01 7.32511580e-02
-7.40704179e-01 -2.58812875e-01 -6.75203860e-01 -1.93134010e-01
-3.35044563e-01 -1.61628529e-01 1.69632888e+00 5.15863419e-01
-4.24209893e-01 3.78422081e-01 -5.46293497e-01 -3.46934557e-01
6.69618547e-01 5.31303346e-01 -5.84918484e-02 -1.88002989e-01
-2.14514494e+00 5.54088354e-01 1.99454725e-01 1.77192360e-01
-1.33692229e+00 -1.17853999e+00 -7.07273841e-01 9.30602327e-02
5.96750438e-01 -1.21163881e+00 1.11788356e+00 -3.45445216e-01
-1.06573272e+00 2.56128341e-01 -2.61359572e-01 -3.05800319e-01
3.94463837e-01 -3.67815167e-01 -5.67626595e-01 -4.28002805e-01
1.89858228e-01 2.80487299e-01 3.72718573e-01 -1.01037979e+00
-7.98058689e-01 -5.89229107e-01 1.03414297e-01 -2.37444997e-01
-2.75358111e-02 2.02324241e-01 1.42905712e-01 -7.14751720e-01
-2.69455403e-01 -4.80248481e-01 -3.67786944e-01 -4.17685688e-01
-6.31503999e-01 -8.07486117e-01 5.87023675e-01 -2.62049824e-01
1.32155240e+00 -1.64581692e+00 1.37560055e-01 -1.15815677e-01
5.36905587e-01 -7.87271708e-02 -8.13386068e-02 5.40911913e-01
-5.23136377e-01 4.01307255e-01 -1.05646975e-01 5.09051263e-01
-1.79662660e-01 8.62080455e-02 -6.61844015e-01 4.94036257e-01
4.61961597e-01 1.30030453e+00 -1.48679829e+00 -2.77698547e-01
3.34211797e-01 4.40357745e-01 -5.95154107e-01 4.09004927e-01
-3.35701525e-01 6.70377433e-01 -6.08485401e-01 7.00762331e-01
3.46164316e-01 -3.91939938e-01 4.38047022e-01 1.04355160e-02
-3.38008881e-01 6.23616159e-01 -1.21376216e+00 1.28463900e+00
-1.93894461e-01 6.37246847e-01 -3.23643804e-01 -1.42854428e+00
5.93850553e-01 6.34946048e-01 5.41236043e-01 -6.42640233e-01
4.83725080e-03 2.03337125e-03 2.27240384e-01 -8.72753680e-01
-5.09615466e-02 -2.14467391e-01 -2.39697069e-01 3.36814731e-01
-1.74745023e-01 4.46059495e-01 -2.29162797e-02 7.91911110e-02
1.18721461e+00 -1.69211388e-01 5.27792990e-01 -1.31482050e-01
2.32950989e-02 3.06406878e-02 7.34992802e-01 9.75948930e-01
-7.02439994e-02 1.59018353e-01 1.01168609e+00 -9.39610004e-01
-6.27813697e-01 -1.15074086e+00 -2.63353348e-01 9.95488584e-01
-2.18136404e-02 -6.56498745e-02 -1.07153609e-01 -6.20778978e-01
3.05212349e-01 5.92245400e-01 -1.21879458e+00 -2.70967364e-01
-8.46826792e-01 -1.48615932e+00 5.25371611e-01 6.70458913e-01
3.40097040e-01 -1.19214904e+00 -1.61284491e-01 2.13222697e-01
-4.29006964e-01 -4.06862259e-01 7.82608762e-02 2.80007660e-01
-9.22207594e-01 -1.54158163e+00 -3.78398508e-01 -3.53642911e-01
1.49972960e-01 1.31391078e-01 1.30659938e+00 -1.88607290e-01
-3.85473639e-01 -5.17068766e-02 2.29693297e-03 -7.53135502e-01
-1.76785007e-01 -4.86273132e-02 1.15625203e-01 -4.76755977e-01
5.96318781e-01 -5.72309971e-01 -8.92621756e-01 3.41522396e-02
-5.00157475e-01 -2.84653008e-01 6.23439610e-01 1.22690058e+00
1.95552677e-01 2.17295080e-01 1.16032517e+00 -1.15723252e+00
7.71768332e-01 -9.91052866e-01 -5.82815886e-01 1.42656356e-01
-7.42637813e-01 1.51364982e-01 3.50969315e-01 -2.01516420e-01
-1.40601456e+00 -3.98088723e-01 5.29237837e-02 1.53934345e-01
-1.77775532e-01 8.47214639e-01 -2.68392175e-01 6.93786263e-01
5.66471219e-01 -5.99634528e-01 -3.80795985e-01 -3.24817806e-01
7.15008378e-01 2.62684524e-01 3.26108903e-01 -5.85026562e-01
3.70322883e-01 5.32146633e-01 1.48786575e-01 -2.21780047e-01
-1.06703198e+00 -2.24342972e-01 -5.47764063e-01 -6.47735000e-02
9.66299117e-01 -8.91667306e-01 -1.43460989e+00 3.55057269e-02
-1.50965178e+00 -4.15664583e-01 4.90181008e-03 7.47215092e-01
-2.61189669e-01 -3.03534478e-01 -5.79826057e-01 -8.00496042e-01
-1.17130019e-01 -9.20341790e-01 8.78075123e-01 1.52543768e-01
-4.76722062e-01 -1.43104517e+00 3.85685563e-01 1.91037133e-01
-5.73759638e-02 3.84776235e-01 1.45211458e+00 -6.49907514e-02
-5.19240558e-01 1.16978839e-01 -6.30303860e-01 -6.05683208e-01
2.71351457e-01 2.19779447e-01 -8.08449328e-01 3.05165648e-01
-5.94898701e-01 -2.72700548e-01 1.24300373e+00 1.21505737e+00
1.45526493e+00 -2.85002172e-01 -8.30663204e-01 4.80956793e-01
1.13920081e+00 3.44340891e-01 5.66934288e-01 -7.00434372e-02
1.06082380e+00 7.96708763e-01 3.76138985e-01 4.34474528e-01
6.77083135e-01 2.17960924e-01 8.50005090e-01 -6.08164132e-01
-1.89471185e-01 -4.97656941e-01 -6.20603748e-02 2.42924333e-01
-4.19373095e-01 -3.61422986e-01 -9.87393022e-01 6.43909693e-01
-2.12424064e+00 -1.36898983e+00 -9.52680528e-01 1.88592529e+00
9.73884583e-01 -3.33528131e-01 1.36134401e-01 1.49467075e-02
8.84126961e-01 7.05712894e-03 -7.78880596e-01 -1.98967546e-01
-1.09059736e-01 1.89164564e-01 5.37611246e-01 4.13074940e-01
-1.32263982e+00 9.48022068e-01 7.78776979e+00 3.19797665e-01
-8.83781254e-01 1.86351389e-01 6.93125486e-01 1.86294615e-02
-5.29196024e-01 1.87579110e-01 -5.27732432e-01 3.87875974e-01
7.96807528e-01 -8.59555602e-03 2.38656536e-01 5.19803286e-01
8.08789015e-01 5.21143153e-02 -1.37827957e+00 6.06477022e-01
-6.36946321e-01 -1.78980649e+00 9.68626589e-02 4.83249798e-02
1.04998279e+00 4.65437062e-02 7.94321224e-02 2.11041465e-01
1.23440015e+00 -1.43173277e+00 1.43983990e-01 5.12652516e-01
8.26591015e-01 -7.89542854e-01 8.67929816e-01 -2.36836106e-01
-7.69558549e-01 -3.61185342e-01 -4.97882158e-01 -7.22615182e-01
2.00129464e-01 1.35496831e+00 -5.62939286e-01 6.60796106e-01
8.84321809e-01 1.06836486e+00 4.55738455e-02 8.78875136e-01
-7.30440080e-01 1.16907203e+00 3.34519565e-01 -2.13848233e-01
5.98412764e-04 1.47030711e-01 2.09665909e-01 1.17463791e+00
4.25238572e-02 2.53534138e-01 1.80012584e-02 1.15262806e+00
-2.67868489e-01 -3.95639688e-01 -6.48609996e-01 -6.04666285e-02
6.38165534e-01 6.79951012e-01 -2.53214300e-01 -1.63505822e-01
-4.15907353e-01 2.79042274e-01 2.87518024e-01 5.62190294e-01
-1.02825105e+00 -7.56049752e-02 7.70531893e-01 -1.45961508e-01
-3.06562692e-01 7.52822012e-02 -7.77571082e-01 -1.01897323e+00
-6.87780440e-01 -6.52870774e-01 9.75687087e-01 -4.59968835e-01
-1.51797366e+00 -4.01129603e-01 8.68355185e-02 -4.58861321e-01
1.77930237e-03 -4.36062247e-01 -6.88692808e-01 7.80326068e-01
-1.49167967e+00 -1.02268970e+00 -1.43599153e-01 4.98279870e-01
3.48754376e-01 5.76442443e-02 8.00428331e-01 4.34931070e-01
-8.79044473e-01 2.53163248e-01 9.66773033e-02 2.99222291e-01
1.03579295e+00 -1.49406588e+00 3.00704420e-01 5.28566420e-01
-2.79107600e-01 1.00816572e+00 6.77602530e-01 -1.04342365e+00
-1.54047048e+00 -1.25825608e+00 1.10749590e+00 -4.98815894e-01
1.02761590e+00 -3.03697288e-01 -5.64906895e-01 6.56218588e-01
3.18969965e-01 -2.02413514e-01 7.98754811e-01 1.06808424e+00
-3.17820907e-01 -7.63683245e-02 -6.55139685e-01 6.39775455e-01
1.22174799e+00 -3.15337740e-02 -5.91783166e-01 4.01955873e-01
9.64304745e-01 1.72184221e-02 -8.75768721e-01 4.96634305e-01
7.38844275e-01 -7.65339673e-01 1.07893336e+00 -1.16065896e+00
1.17142773e+00 -4.98964675e-02 3.10979068e-01 -1.41591537e+00
-8.80085826e-01 -4.57913756e-01 -2.70652086e-01 1.15125930e+00
2.78162926e-01 -5.03256559e-01 5.46303988e-01 4.58583742e-01
1.12603538e-01 -4.71975297e-01 -7.53241718e-01 -2.59285718e-01
5.20494163e-01 -5.25313854e-01 8.72527897e-01 1.51745605e+00
1.27251431e-01 8.04674089e-01 -7.12239206e-01 2.29269296e-01
8.35814297e-01 2.63638616e-01 5.90236425e-01 -1.67319703e+00
2.48065144e-01 -4.82993394e-01 1.98415026e-01 -6.27140224e-01
3.04801822e-01 -6.84947252e-01 -2.25938544e-01 -1.83961248e+00
6.32171273e-01 -6.04936600e-01 -5.01919806e-01 5.20776570e-01
-6.98058069e-01 -2.97450215e-01 -5.40901780e-01 5.69196083e-02
-2.43108198e-01 6.00405514e-01 1.46243560e+00 -4.12785709e-01
-4.94922288e-02 1.41706094e-02 -9.31289196e-01 5.62339127e-01
8.47074151e-01 -6.18764818e-01 -4.20796484e-01 -6.12443328e-01
7.50646234e-01 4.10770506e-01 8.45878839e-01 -1.64242402e-01
1.07855074e-01 -8.72407377e-01 6.76793754e-01 -5.48325419e-01
-4.72451180e-01 -5.20079970e-01 -6.07907474e-02 6.64535463e-01
-7.88079679e-01 2.15847250e-02 -9.49773192e-02 8.55306149e-01
-1.12508647e-01 3.10123205e-01 3.06011021e-01 -1.59355626e-01
-7.90930331e-01 4.03531045e-01 -3.60260099e-01 -5.65091558e-02
6.54395342e-01 5.09888589e-01 -6.11960471e-01 -3.39857697e-01
-3.67488265e-01 6.41476572e-01 -5.12420833e-01 6.78476691e-01
5.43231189e-01 -1.44499242e+00 -1.02799809e+00 -3.23543012e-01
-2.92500500e-02 -7.04509094e-02 3.33707929e-01 7.63359666e-01
2.01624572e-01 8.64297569e-01 3.26861776e-02 -2.74964571e-01
-7.09941268e-01 9.93832707e-01 8.00633803e-02 -2.41648883e-01
-2.46142268e-01 8.51475239e-01 7.14937925e-01 -5.45030296e-01
3.90680075e-01 -3.05807561e-01 -3.82743955e-01 1.30341589e-01
5.87948084e-01 7.57507801e-01 -3.76876712e-01 2.71680713e-01
-4.99498904e-01 1.20427415e-01 2.23244891e-01 3.87255311e-01
1.47131169e+00 1.44008309e-01 -6.12747550e-01 5.76584637e-01
8.64113152e-01 -3.16369832e-01 -9.18390512e-01 1.14059381e-01
1.12341575e-01 -1.57513767e-01 2.63118982e-01 -1.26794469e+00
-1.03069186e+00 1.06137407e+00 5.39710760e-01 2.12228164e-01
7.31812060e-01 1.97297752e-01 3.86071891e-01 6.47885129e-02
2.62760341e-01 -6.42681718e-01 3.99453603e-02 4.78983790e-01
1.11503327e+00 -1.69187045e+00 1.28664374e-01 -3.96538615e-01
5.57870977e-02 1.02754259e+00 5.57480633e-01 -1.19034439e-01
7.49052584e-01 2.81724542e-01 -1.99546516e-01 -5.42753041e-01
-9.51858997e-01 -1.39971524e-01 1.87395588e-01 7.53048122e-01
1.19759381e+00 5.02482891e-01 -5.67376196e-01 5.64905226e-01
-5.48570119e-02 2.16469571e-01 2.61770606e-01 8.66399035e-02
6.47685453e-02 -1.11024714e+00 -5.18286467e-01 6.46083832e-01
-5.10160506e-01 -4.04239655e-01 -5.30146658e-01 9.26016986e-01
2.92673647e-01 1.28052485e+00 7.00019374e-02 -1.95792392e-01
1.27068236e-01 -3.59154940e-01 3.41798931e-01 -4.02053595e-01
-2.95821637e-01 -1.46224439e-01 6.11537881e-02 -6.26483560e-01
-6.51021123e-01 -7.83937812e-01 -1.19005907e+00 -6.47042334e-01
-2.23077431e-01 6.53114244e-02 3.80955398e-01 9.07703638e-01
1.03929237e-01 1.18290842e+00 2.48802036e-01 -2.25504741e-01
-1.51378095e-01 -8.87753904e-01 -2.02959523e-01 -4.72496860e-02
5.52087247e-01 -1.17616200e+00 -6.55511841e-02 1.02983840e-01] | [7.99501895904541, 5.400814056396484] |
38b4de86-7625-4069-af6f-0a95149160ed | a-hybrid-citation-retrieval-algorithm-for | 1609.01597 | null | http://arxiv.org/abs/1609.01597v1 | http://arxiv.org/pdf/1609.01597v1.pdf | A Hybrid Citation Retrieval Algorithm for Evidence-based Clinical Knowledge Summarization: Combining Concept Extraction, Vector Similarity and Query Expansion for High Precision | Novel information retrieval methods to identify citations relevant to a
clinical topic can overcome the knowledge gap existing between the primary
literature (MEDLINE) and online clinical knowledge resources such as UpToDate.
Searching the MEDLINE database directly or with query expansion methods returns
a large number of citations that are not relevant to the query. The current
study presents a citation retrieval system that retrieves citations for
evidence-based clinical knowledge summarization. This approach combines query
expansion, concept-based screening algorithm, and concept-based vector
similarity. We also propose an information extraction framework for automated
concept (Population, Intervention, Comparison, and Disease) extraction. We
evaluated our proposed system on all topics (as queries) available from
UpToDate for two diseases, heart failure (HF) and atrial fibrillation (AFib).
The system achieved an overall F-score of 41.2% on HF topics and 42.4% on AFib
topics on a gold standard of citations available in UpToDate. This is
significantly high when compared to a query-expansion based baseline (F-score
of 1.3% on HF and 2.2% on AFib) and a system that uses query expansion with
disease hyponyms and journal names, concept-based screening, and term-based
vector similarity system (F-score of 37.5% on HF and 39.5% on AFib). Evaluating
the system with top K relevant citations, where K is the number of citations in
the gold standard achieved a much higher overall F-score of 69.9% on HF topics
and 75.1% on AFib topics. In addition, the system retrieved up to 18 new
relevant citations per topic when tested on ten HF and six AFib clinical
topics. | ['Siddhartha R. Jonnalagadda', 'Ravi P Garg', 'Kalpana Raja', 'Andrew J Sauer', 'Melanie R Klerer'] | 2016-09-06 | null | null | null | null | ['clinical-knowledge'] | ['miscellaneous'] | [ 3.69915329e-02 1.80675820e-01 -5.29419541e-01 4.25539166e-01
-1.29674697e+00 -6.84880316e-01 2.72792697e-01 1.14880753e+00
-5.53836167e-01 1.18147135e+00 4.54602778e-01 -4.52898175e-01
-8.49067390e-01 -7.36269712e-01 -2.66626418e-01 -1.05078325e-01
-5.66961281e-02 7.28217185e-01 2.45776728e-01 -1.12048581e-01
5.76800704e-01 3.18146408e-01 -1.36514521e+00 2.15002984e-01
1.26676548e+00 5.97096860e-01 3.02288055e-01 6.08256578e-01
-2.85199851e-01 1.09182492e-01 -1.01087308e+00 -8.94272253e-02
-3.29001635e-01 -3.92418802e-01 -9.60703850e-01 -8.82488430e-01
3.95279884e-01 -1.40345199e-02 -1.07354097e-01 6.77964330e-01
8.54045808e-01 -3.73851091e-01 8.12602758e-01 -8.72896314e-01
-3.40594083e-01 4.85317379e-01 -1.61705762e-01 4.63891417e-01
8.02319884e-01 -3.12040359e-01 6.82119548e-01 -6.77318633e-01
1.14608610e+00 8.52937460e-01 5.03473938e-01 3.00932080e-01
-8.02185237e-01 -8.06616127e-01 -1.35230869e-01 2.92012077e-02
-1.39031875e+00 -2.69077551e-02 2.09449202e-01 -6.03408933e-01
1.27989793e+00 5.51207900e-01 1.03995335e+00 6.12988472e-01
5.25582492e-01 2.06532910e-01 5.19269347e-01 -6.22790039e-01
3.63446206e-01 -6.27867877e-03 5.07126570e-01 3.01525027e-01
1.03405166e+00 -1.97145551e-01 -3.72678518e-01 -1.03255129e+00
3.61199528e-01 1.42945498e-01 -6.36130095e-01 3.93498629e-01
-1.14288116e+00 6.77024245e-01 1.73515633e-01 4.21433151e-01
-7.80486286e-01 -3.81299645e-01 5.30732751e-01 8.85318518e-02
8.34297538e-02 9.60358977e-01 -7.36701250e-01 -7.61887729e-02
-1.36168718e+00 6.07979834e-01 1.26222622e+00 9.76703644e-01
1.96333095e-01 -4.97795224e-01 -6.29952133e-01 8.95341873e-01
8.10822323e-02 8.58702898e-01 8.24021637e-01 -8.89654756e-01
2.75700301e-01 9.82865930e-01 -5.59620783e-02 -8.45365763e-01
-4.62410867e-01 -6.84840441e-01 -8.20805848e-01 -6.13559842e-01
-3.18615228e-01 -2.32459128e-01 -9.85039234e-01 1.56459773e+00
-3.95788485e-03 -4.06320482e-01 2.84640521e-01 2.99670041e-01
1.56587350e+00 4.47057903e-01 4.32938457e-01 -9.63845670e-01
1.85994339e+00 -3.71189684e-01 -1.11456704e+00 2.94656008e-01
7.61096954e-01 -9.25097823e-01 3.26249897e-01 2.65491813e-01
-1.18380988e+00 -1.36097893e-02 -9.98543620e-01 1.91216230e-01
-6.13206744e-01 5.18970639e-02 3.38237762e-01 3.17340046e-01
-1.12603533e+00 2.63915509e-01 -4.04382795e-01 -8.09587300e-01
4.59057212e-01 3.69632751e-01 -2.60570437e-01 -8.02746564e-02
-1.50049615e+00 1.21162808e+00 5.98812401e-01 -7.29082406e-01
-4.52867091e-01 -1.31444860e+00 -4.98930842e-01 2.72067845e-01
4.19431508e-01 -1.56839609e+00 9.50381041e-01 2.76471257e-01
-6.70361698e-01 6.83259666e-01 -3.86671305e-01 -5.03697157e-01
6.84275478e-02 -2.67560989e-01 -5.98624051e-01 8.68727446e-01
6.86649859e-01 4.98987049e-01 -3.64135504e-02 -7.25949883e-01
-7.45453954e-01 -4.20500666e-01 -3.65674675e-01 2.71109849e-01
-2.53250092e-01 6.13517091e-02 -5.93866885e-01 -6.82289600e-01
-1.26503427e-02 -6.10348403e-01 -3.50119263e-01 -2.87684441e-01
-3.09017360e-01 -4.94942516e-01 3.81197721e-01 -8.38558495e-01
2.14091873e+00 -1.44153833e+00 -2.09706724e-01 3.12893629e-01
5.04105330e-01 5.98568916e-01 2.79609054e-01 8.22507858e-01
-8.74037072e-02 7.58585513e-01 -1.80603191e-01 7.54924059e-01
-5.84758043e-01 -1.22909047e-01 -3.52884308e-02 -1.16307870e-01
9.84266549e-02 9.91724133e-01 -1.05918944e+00 -8.39989781e-01
-1.41802458e-02 4.51199979e-01 -5.13419271e-01 -1.02568597e-01
1.46731615e-01 -2.79816985e-01 -7.91571498e-01 8.87471020e-01
2.93124378e-01 -4.31589216e-01 1.03463992e-01 -1.09064378e-01
3.03091109e-02 4.08512503e-01 -7.12934017e-01 1.62414694e+00
-1.30307943e-01 6.50275648e-02 -2.89891362e-01 -6.32552028e-01
7.05669880e-01 8.69443715e-01 9.37993348e-01 -6.39108539e-01
-2.12243617e-01 6.32486105e-01 -1.61164567e-01 -6.38430119e-01
2.40885198e-01 4.39967029e-02 -1.28219590e-01 9.56035703e-02
2.55942106e-01 -6.96440190e-02 5.06981909e-01 8.00045490e-01
1.52945161e+00 -3.75186950e-01 8.32769632e-01 -4.52460617e-01
5.58039427e-01 4.21286166e-01 4.15320545e-01 1.08916330e+00
2.09870055e-01 5.90522289e-01 3.74967903e-01 -1.17036894e-01
-6.14945471e-01 -9.63109076e-01 -6.27665699e-01 2.02275202e-01
-2.24488348e-01 -1.03037691e+00 -6.22663260e-01 -3.45786512e-01
1.22213066e-01 4.46351200e-01 -4.12757754e-01 -5.26050746e-01
-2.15199351e-01 -7.79569268e-01 6.99687779e-01 7.09842145e-02
2.84437954e-01 -1.03299725e+00 -8.07353795e-01 5.40716648e-01
-3.50114554e-01 -6.43574595e-01 -2.92185694e-01 -1.11890361e-01
-1.24495590e+00 -1.35017180e+00 -1.35545909e+00 -7.25723207e-01
2.68046647e-01 -2.59288967e-01 1.34388542e+00 6.17556423e-02
-6.32085383e-01 5.52455783e-01 -1.77179396e-01 -6.95834339e-01
-1.57432407e-01 3.09666008e-01 -1.91072568e-01 -1.15236127e+00
5.06284058e-01 -2.61039764e-01 -1.01756012e+00 -2.35151127e-01
-9.43733096e-01 -4.72678065e-01 7.42472291e-01 8.08856070e-01
6.10291421e-01 -4.62113738e-01 1.20885015e+00 -7.42536902e-01
1.16975510e+00 -8.30949247e-01 -2.35953078e-01 5.10442197e-01
-1.38678193e+00 -1.86148956e-01 -2.39223659e-01 -3.78642440e-01
-4.37759787e-01 -2.66440660e-01 4.65361811e-02 -1.63099356e-02
-6.26978651e-02 1.15860128e+00 4.05648351e-01 4.15827096e-01
8.49295080e-01 4.88953963e-02 -8.05385970e-03 -6.09533787e-01
-3.07317870e-03 8.56681347e-01 3.91194612e-01 -3.98590207e-01
5.65530472e-02 -7.49986023e-02 2.88989767e-02 -7.29138851e-01
-3.03052962e-01 -1.05362177e+00 -3.31521705e-02 2.81625658e-01
7.25906074e-01 -1.02236402e+00 -7.53211796e-01 -2.42106706e-01
-1.25250590e+00 7.77750969e-01 -3.04283500e-01 8.33108902e-01
-6.48818016e-02 4.56110477e-01 -3.28844726e-01 -6.44476056e-01
-1.34053802e+00 -8.94383967e-01 1.09449470e+00 2.24528357e-01
-8.65887761e-01 -6.77193820e-01 3.74933928e-01 -1.09925270e-01
3.96019191e-01 3.79930198e-01 1.33954191e+00 -1.07046151e+00
4.07318994e-02 -5.73783338e-01 -5.20296060e-02 -2.55054384e-01
5.18995047e-01 -2.98574954e-01 -3.08148921e-01 -1.22040115e-01
-2.44443700e-01 3.29037994e-01 9.62843299e-01 8.64171445e-01
6.77369416e-01 -3.90128464e-01 -1.08871102e+00 5.03521226e-02
1.47682726e+00 7.81562150e-01 5.23453176e-01 5.05680919e-01
1.73543662e-01 4.45736289e-01 4.06873643e-01 2.76819408e-01
2.56166577e-01 6.40264273e-01 -2.03500435e-01 6.10352755e-02
-7.24281892e-02 -3.82751189e-02 -2.68965423e-01 7.08041310e-01
-9.86935943e-02 -1.71307087e-01 -1.25622475e+00 1.07335126e+00
-1.57634115e+00 -7.51799941e-01 -1.30698189e-01 2.36216879e+00
1.20604300e+00 4.47131135e-02 8.43108669e-02 -1.39799625e-01
6.58214033e-01 -8.98842692e-01 -3.63315523e-01 -5.49459755e-01
-6.34675547e-02 8.35798383e-01 3.88606280e-01 2.83010334e-01
-6.30953491e-01 4.53612834e-01 6.61518955e+00 6.45685136e-01
-7.49537408e-01 -2.04764560e-01 2.00127900e-01 -2.26068407e-01
-2.75815785e-01 8.46752152e-02 -9.85197604e-01 4.84361261e-01
1.23536777e+00 -1.11967635e+00 -4.40819591e-01 5.36699414e-01
1.90372124e-01 -2.93962657e-01 -7.10621417e-01 8.60577166e-01
-1.10441148e-01 -1.71700239e+00 5.34518480e-01 2.42246091e-01
5.76196611e-01 -6.27123639e-02 -4.88971919e-01 3.55198562e-01
-1.32095322e-01 -8.48138034e-01 -1.28036603e-01 8.44541907e-01
1.13534737e+00 -3.77078742e-01 1.10416937e+00 1.21931411e-01
-7.95677364e-01 2.76083291e-01 -9.14114639e-02 2.15309292e-01
1.45415083e-01 7.80896842e-01 -9.88502979e-01 1.11459506e+00
9.53763008e-01 6.22478127e-01 -5.13010919e-01 1.63473904e+00
1.97871000e-01 5.72626650e-01 -6.39927626e-01 -1.70593277e-01
-8.54685679e-02 2.69004583e-01 7.18102753e-01 1.36300266e+00
5.14277041e-01 3.57520610e-01 4.96252216e-02 5.89121580e-01
-1.60625398e-01 6.57880366e-01 -6.15659297e-01 -9.06671435e-02
7.89397955e-01 8.85464370e-01 -4.97482687e-01 -8.52455735e-01
-3.89885008e-02 2.56302148e-01 -5.43166399e-01 4.12118107e-01
-2.39716038e-01 -9.56758261e-01 -1.40714580e-02 3.13465118e-01
7.82035589e-02 7.30560541e-01 6.82877451e-02 -8.97306681e-01
9.23054665e-02 -8.29785109e-01 9.73396361e-01 -5.80845118e-01
-1.18771374e+00 5.90780616e-01 5.80379188e-01 -1.27925348e+00
-5.70343196e-01 -8.48954841e-02 -1.42564014e-01 1.40144742e+00
-1.08437824e+00 -5.28729558e-01 9.16268304e-02 2.54033685e-01
2.31108904e-01 -2.64365435e-01 1.29267418e+00 3.63454312e-01
-1.26354337e-01 3.85402203e-01 7.48943463e-02 -2.34235436e-01
9.67743099e-01 -1.05588949e+00 -2.96725124e-01 3.66681777e-02
-5.86977541e-01 1.41050684e+00 2.56398499e-01 -1.23788846e+00
-9.12450850e-01 -7.67244816e-01 1.53171444e+00 -4.59178239e-01
1.39067709e-01 6.07607782e-01 -1.10087430e+00 -1.51229454e-02
2.56558269e-01 -6.90667570e-01 9.45000410e-01 1.10748567e-01
1.12127922e-01 9.36439559e-02 -1.25802290e+00 4.47928011e-01
6.01445317e-01 -1.59167483e-01 -1.18164468e+00 5.28413236e-01
7.63564229e-01 -1.59749523e-01 -1.56308126e+00 9.36264455e-01
8.30449343e-01 3.98573764e-02 1.32331419e+00 -6.40572071e-01
3.70216221e-01 -1.79825634e-01 2.62720197e-01 -8.32020044e-01
-1.77229345e-01 -6.10888481e-01 -3.34003627e-01 9.15781617e-01
7.31398463e-01 -5.94218791e-01 2.39224210e-01 4.20495152e-01
-6.27038851e-02 -9.72958028e-01 -8.31352711e-01 -3.85386825e-01
3.23734671e-01 2.73750007e-01 1.35020658e-01 8.89733493e-01
5.44208705e-01 7.21237659e-01 5.32952487e-01 -1.65033057e-01
3.40688288e-01 1.58654794e-03 6.95387274e-02 -1.79072642e+00
3.45985442e-01 -6.22756362e-01 -3.35320771e-01 -1.71797335e-01
-5.12466013e-01 -9.44491386e-01 -5.55339277e-01 -2.40875077e+00
6.92758620e-01 3.41090225e-02 -5.63061297e-01 5.53526759e-01
-4.23681408e-01 -2.87070364e-01 -2.56251663e-01 5.60327828e-01
-3.90852928e-01 2.26057339e-02 9.82001603e-01 -1.04259111e-01
-5.66916287e-01 -1.07396603e-01 -1.09463441e+00 3.85541588e-01
7.35351741e-01 -8.38714004e-01 -2.91112959e-01 2.58632243e-01
2.75965333e-01 3.99481624e-01 1.10715386e-02 -7.51935363e-01
4.81849819e-01 8.82274806e-02 3.13873023e-01 -9.95549858e-01
-2.54664660e-01 -5.05776465e-01 3.18154573e-01 1.02873695e+00
-3.59115481e-01 2.84660369e-01 6.90275371e-01 4.78667587e-01
-4.15416211e-01 -2.84578741e-01 -2.17126794e-02 -3.99238139e-01
-2.29390278e-01 -2.37936191e-02 -5.58796763e-01 3.97310346e-01
5.88576555e-01 -1.97832599e-01 -5.15781105e-01 -2.98299104e-01
-8.16035867e-01 5.84165096e-01 -6.61542639e-02 4.79769558e-01
8.17575097e-01 -1.10153866e+00 -9.13886011e-01 -3.58686179e-01
4.72254544e-01 -1.38666227e-01 -4.53135604e-03 8.94134939e-01
-5.17012119e-01 1.07544482e+00 2.53110053e-03 -4.75865394e-01
-1.36375821e+00 4.18290198e-01 -1.38393417e-01 -6.65288389e-01
-5.40441453e-01 3.99700701e-01 -3.86054963e-02 -5.77571392e-02
3.58115703e-01 -3.99196684e-01 -6.06441855e-01 4.21391219e-01
5.63951135e-01 4.77183312e-01 5.42459190e-01 -1.51474535e-01
-8.29030812e-01 8.57239544e-01 -2.99375445e-01 -2.50044852e-01
1.17831075e+00 1.67488694e-01 -4.13661182e-01 1.59833431e-01
1.16833615e+00 1.51447859e-03 4.25890714e-01 8.49220082e-02
3.03014725e-01 3.88970137e-01 1.25145286e-01 -1.49602175e+00
-4.41392004e-01 3.13434809e-01 8.16718400e-01 -1.21457260e-02
1.26140821e+00 1.37729302e-01 5.54215372e-01 5.81003606e-01
8.60731974e-02 -8.14264953e-01 -5.12526512e-01 2.33035505e-01
9.73439038e-01 -7.71661341e-01 5.78075707e-01 -3.47001672e-01
-2.35997409e-01 9.47776616e-01 -5.04416879e-03 1.98755056e-01
8.80346239e-01 -1.38648093e-01 -1.16260171e-01 -9.06834126e-01
-8.67578924e-01 -8.81127864e-02 6.83387101e-01 2.82064915e-01
5.50308168e-01 -2.18659148e-01 -1.40199685e+00 9.31433022e-01
1.26500307e-02 5.13326406e-01 2.28081465e-01 1.22143829e+00
-3.95870954e-01 -1.06923676e+00 -2.52897292e-01 1.06829596e+00
-1.18292606e+00 -5.70446908e-01 -6.00694180e-01 8.07000220e-01
-1.97540298e-01 9.90139484e-01 -3.65231961e-01 1.13996521e-01
6.07246578e-01 2.97171444e-01 1.38319358e-01 -8.59879196e-01
-1.01986408e+00 4.88855541e-01 1.99992925e-01 -2.70468712e-01
-6.34398997e-01 -3.75413179e-01 -1.11248374e+00 3.21535826e-01
-8.11326087e-01 8.83616865e-01 6.18981183e-01 5.72785139e-01
7.54839003e-01 7.60329723e-01 -1.39996007e-01 1.26159653e-01
-1.01137549e-01 -1.10205638e+00 -1.61930501e-01 2.42675886e-01
2.00313121e-01 -5.31111419e-01 -3.09696615e-01 1.07340239e-01] | [8.621833801269531, 8.626053810119629] |
9b830a9d-af58-4e45-8cf0-d89bcc73f3f5 | self-supervised-beat-tracking-in-musical | 2201.01771 | null | https://arxiv.org/abs/2201.01771v1 | https://arxiv.org/pdf/2201.01771v1.pdf | Self-Supervised Beat Tracking in Musical Signals with Polyphonic Contrastive Learning | Annotating musical beats is a very long in tedious process. In order to combat this problem, we present a new self-supervised learning pretext task for beat tracking and downbeat estimation. This task makes use of Spleeter, an audio source separation model, to separate a song's drums from the rest of its signal. The first set of signals are used as positives, and by extension negatives, for contrastive learning pre-training. The drum-less signals, on the other hand, are used as anchors. When pre-training a fully-convolutional and recurrent model using this pretext task, an onset function is learned. In some cases, this function was found to be mapped to periodic elements in a song. We found that pre-trained models outperformed randomly initialized models when a beat tracking training set was extremely small (less than 10 examples). When that was not the case, pre-training led to a learning speed-up that caused the model to overfit to the training set. More generally, this work defines new perspectives in the realm of musical self-supervised learning. It is notably one of the first works to use audio source separation as a fundamental component of self-supervision. | ['Dorian Desblancs'] | 2022-01-05 | null | null | null | null | ['audio-source-separation'] | ['audio'] | [ 4.51985091e-01 2.12326020e-01 5.52617125e-02 -6.99133500e-02
-7.40228832e-01 -8.71836185e-01 3.10241014e-01 1.95787832e-01
-3.01288545e-01 5.13986349e-01 2.30842188e-01 1.38324454e-01
5.27533367e-02 -4.59260225e-01 -6.84817135e-01 -7.00414419e-01
-2.43803084e-01 4.36502516e-01 3.56670141e-01 -4.89556342e-01
1.43474489e-01 2.89991219e-02 -1.68563914e+00 3.93005133e-01
3.63751292e-01 8.58897686e-01 1.10627912e-01 8.41116250e-01
8.85711461e-02 8.17802846e-01 -8.49150896e-01 -2.01772489e-02
1.71959043e-01 -7.76092768e-01 -7.55120397e-01 -2.06937000e-01
4.89050955e-01 1.09732009e-01 2.18701269e-02 8.28838468e-01
6.92842066e-01 7.42913857e-02 3.86290312e-01 -1.00456333e+00
8.24888870e-02 1.16812980e+00 -2.52249509e-01 3.48428100e-01
3.84668946e-01 -4.14933115e-02 1.25996149e+00 -5.82373321e-01
3.72968495e-01 8.00557613e-01 1.11679554e+00 3.18978518e-01
-1.29566038e+00 -7.81844378e-01 -1.83272362e-01 -4.37351204e-02
-1.07081068e+00 -6.45709097e-01 1.05989802e+00 -5.10399222e-01
6.90811753e-01 3.46981853e-01 9.06459570e-01 1.14785123e+00
-1.91003904e-01 7.31599152e-01 8.03203046e-01 -6.94104254e-01
2.62418598e-01 9.55205336e-02 8.11799336e-03 2.96245635e-01
-2.79603988e-01 3.10302496e-01 -8.53434741e-01 -1.83827609e-01
8.11003149e-01 -5.14004409e-01 -2.42217615e-01 -4.37283777e-02
-1.02554762e+00 5.54140806e-01 3.10239851e-01 6.93344057e-01
-2.35389680e-01 9.21080336e-02 6.33525491e-01 6.43427968e-01
3.60735744e-01 8.06715727e-01 -4.37288463e-01 -4.42106813e-01
-1.22558296e+00 1.30911991e-01 9.23365831e-01 2.84339458e-01
5.49892902e-01 3.54381055e-01 -1.79340392e-01 8.20044458e-01
-2.62888640e-01 -1.41149819e-01 8.64096463e-01 -8.90532076e-01
1.49748653e-01 3.77779573e-01 -3.19263525e-02 -6.58503890e-01
-4.95231390e-01 -9.99642074e-01 -6.68773651e-01 2.13323861e-01
7.51923382e-01 -1.67030424e-01 -6.23064756e-01 1.80171227e+00
1.77932158e-01 4.42984432e-01 -7.59964809e-03 1.00138688e+00
5.46657562e-01 4.11023825e-01 -3.56773436e-01 -3.37681144e-01
1.09764194e+00 -8.73493969e-01 -6.06988788e-01 -2.44702205e-01
3.76141250e-01 -9.95462835e-01 1.23962808e+00 6.23544753e-01
-1.10891652e+00 -9.60558951e-01 -1.30333674e+00 3.70547384e-01
-6.66249320e-02 1.95214406e-01 4.50409740e-01 3.88782144e-01
-5.96430540e-01 1.16178298e+00 -6.37057960e-01 -8.22346285e-02
1.52253017e-01 2.19732106e-01 -7.28995875e-02 7.01849401e-01
-1.30022657e+00 6.00613177e-01 3.88106078e-01 -2.08920226e-01
-8.80521357e-01 -6.96772039e-01 -5.13031542e-01 -8.79933592e-03
2.79873699e-01 -2.93353736e-01 1.44680238e+00 -1.40424430e+00
-1.62726915e+00 9.99290347e-01 2.22193912e-01 -7.75006115e-01
4.39182401e-01 -3.50972325e-01 -5.41651845e-01 -2.06631962e-02
1.45318091e-01 2.40782142e-01 1.40437758e+00 -9.75168347e-01
-4.56058294e-01 -1.68560088e-01 -1.94799438e-01 2.00305670e-01
-3.60524029e-01 4.13556444e-03 -9.89943445e-02 -1.10467064e+00
3.26391123e-02 -1.01157701e+00 1.92359254e-01 -7.02053189e-01
-5.91925740e-01 -1.64483204e-01 6.07313395e-01 -3.18022907e-01
1.28321230e+00 -2.54368305e+00 2.46145025e-01 2.85178035e-01
-7.37948483e-03 2.26544365e-01 -3.06074005e-02 4.83460218e-01
-5.22371829e-01 -2.62617737e-01 -2.00437933e-01 -3.34332168e-01
-1.55613378e-01 -7.19321938e-03 -6.26863539e-01 3.17883164e-01
1.12679683e-01 3.72109830e-01 -9.32350755e-01 -2.53290415e-01
-8.46991837e-02 3.62620443e-01 -4.91793603e-01 2.53558964e-01
-2.80207187e-01 7.37703323e-01 2.44401231e-01 2.13610217e-01
2.99965069e-02 1.04145072e-01 -6.03438579e-02 -2.18362018e-01
-1.97555676e-01 7.63543785e-01 -1.51464081e+00 1.76649225e+00
-3.38574678e-01 8.62815678e-01 -1.16478547e-01 -9.24617290e-01
9.28579688e-01 6.66791856e-01 7.29832947e-01 -3.89503837e-01
1.82037771e-01 3.97841603e-01 4.43744361e-01 -4.03922170e-01
4.51122314e-01 -3.48396182e-01 9.37499255e-02 6.15817308e-01
4.06294107e-01 -2.04145178e-01 3.06960940e-01 -9.87691581e-02
1.14100695e+00 3.93363446e-01 1.83782697e-01 -4.03220877e-02
3.65727454e-01 -4.07549404e-02 5.06551802e-01 5.71468353e-01
9.19342190e-02 8.69247437e-01 3.24999511e-01 -4.16534990e-01
-7.77572632e-01 -8.11056495e-01 -1.35328267e-02 1.56817484e+00
-4.16908503e-01 -8.80261719e-01 -8.76719832e-01 -4.02898848e-01
-6.46042675e-02 4.19670224e-01 -4.76670533e-01 -3.10026407e-01
-6.35528684e-01 -4.69604611e-01 1.00111043e+00 4.40807045e-01
1.09697692e-02 -1.36203718e+00 -8.54576945e-01 5.56925595e-01
-1.49112999e-01 -7.22013295e-01 -4.17221576e-01 9.59712207e-01
-7.53777862e-01 -1.14327896e+00 -7.09883451e-01 -7.95570493e-01
5.84469996e-02 -2.00007111e-01 1.10192704e+00 7.15400907e-04
-3.37352574e-01 7.76070058e-02 -2.88176864e-01 -7.69705653e-01
-5.36086202e-01 2.44898275e-01 3.53266865e-01 1.15039580e-01
1.23255707e-01 -1.01035130e+00 -3.56540501e-01 2.10506886e-01
-6.66650236e-01 -2.80020148e-01 2.50485986e-01 7.16577113e-01
5.17370164e-01 2.68018037e-01 6.94599748e-01 -6.49420559e-01
7.13667989e-01 -7.64888972e-02 -2.12819532e-01 -2.79125780e-01
-2.42018372e-01 -2.58370209e-02 8.36037457e-01 -8.95344913e-01
-4.33936507e-01 3.23127419e-01 -1.25174180e-01 -4.96172071e-01
-1.36803344e-01 3.80190521e-01 2.81263173e-01 1.90573663e-01
1.13577867e+00 1.86571188e-03 5.13222627e-02 -7.03213453e-01
1.32612750e-01 5.65660298e-01 8.30761075e-01 -3.18608195e-01
7.47925401e-01 1.74157396e-01 -2.55890191e-01 -8.83024871e-01
-1.18677962e+00 -5.37548959e-01 -7.70265460e-01 -3.05460602e-01
4.54937100e-01 -8.17577600e-01 -2.96386421e-01 3.44801068e-01
-8.16315532e-01 -5.38688779e-01 -8.68701041e-01 5.86349726e-01
-7.13140666e-01 1.17551237e-01 -5.59950769e-01 -9.53566074e-01
-2.22289547e-01 -5.84529757e-01 8.43349576e-01 2.45080143e-02
-7.34325528e-01 -6.66807532e-01 6.10605359e-01 8.33235085e-02
3.05132270e-01 2.43314102e-01 6.09405518e-01 -9.21903551e-01
1.76307097e-01 -2.04006791e-01 5.18386364e-01 4.68356729e-01
4.11445409e-01 -5.39287850e-02 -1.48932028e+00 -2.82356411e-01
2.13200197e-01 -5.29867828e-01 9.43894625e-01 1.22718997e-01
8.14085901e-01 -2.32135102e-01 4.28861640e-02 5.56668699e-01
8.43290389e-01 4.30419147e-02 3.75007182e-01 3.83882105e-01
4.83635306e-01 5.46386957e-01 4.60796177e-01 3.95146966e-01
-1.42978966e-01 8.63452494e-01 3.42246503e-01 -3.14512759e-01
-3.72424871e-01 -3.79927665e-01 5.72074890e-01 9.87962604e-01
-3.19882661e-01 3.44776213e-01 -8.49814355e-01 4.26996857e-01
-1.67927670e+00 -1.18179727e+00 -1.60185352e-01 2.34302354e+00
1.29682052e+00 6.11505389e-01 7.26364732e-01 1.00391698e+00
5.44068277e-01 -4.00179625e-03 -4.53630000e-01 -1.18981950e-01
-1.37284100e-01 5.05374253e-01 4.97703291e-02 2.63199955e-01
-1.37373579e+00 8.57963681e-01 6.02453470e+00 6.60556972e-01
-1.35154307e+00 -6.29905835e-02 2.38481238e-02 -2.07377285e-01
1.63952872e-01 4.49953824e-02 -5.14200985e-01 5.51121116e-01
9.84735072e-01 2.00811237e-01 5.72785914e-01 6.76172972e-01
-5.76668710e-04 1.11023486e-01 -1.36108124e+00 1.03630757e+00
4.66624685e-02 -1.04285014e+00 -4.63797420e-01 -1.72842458e-01
4.11419481e-01 -8.20848718e-02 -9.50302705e-02 3.68417203e-01
1.06842980e-01 -1.05729496e+00 9.28577900e-01 3.73649865e-01
6.76452994e-01 -8.03113043e-01 6.02830410e-01 5.52758038e-01
-1.15707052e+00 -1.62762895e-01 -1.01821154e-01 -3.71534228e-01
6.97767269e-03 7.02746332e-01 -8.81808221e-01 3.03323925e-01
7.53871977e-01 7.57777870e-01 -5.28363705e-01 1.14184177e+00
-1.64933145e-01 1.04693806e+00 -4.99433786e-01 4.39167023e-01
8.63175690e-02 2.45325714e-02 8.25536311e-01 1.35247934e+00
5.36881946e-02 -4.19088274e-01 1.46704778e-01 5.44531524e-01
6.79175854e-02 6.26628995e-02 -5.23080826e-01 3.95979099e-02
3.24299484e-01 1.16386044e+00 -7.93015599e-01 -1.72910199e-01
1.67569429e-01 9.50129092e-01 2.41058007e-01 3.88191454e-02
-6.99989796e-01 -4.72822398e-01 2.85562217e-01 3.36293995e-01
2.81976491e-01 -3.27152312e-02 -1.53294444e-01 -8.99342835e-01
-1.98686659e-01 -1.08756924e+00 4.17899072e-01 -6.73674524e-01
-1.18916154e+00 5.51672816e-01 -3.82613003e-01 -1.55926216e+00
-6.23536944e-01 -1.84743926e-01 -7.97170997e-01 5.57806790e-01
-1.13832235e+00 -6.81744099e-01 -1.35672137e-01 6.17832363e-01
4.57156539e-01 -3.73623878e-01 1.10211301e+00 2.65481085e-01
-3.10219735e-01 4.64761496e-01 -1.47955552e-01 3.31707090e-01
1.05899382e+00 -1.52295911e+00 8.99322703e-02 5.32133281e-01
9.49736714e-01 4.28765446e-01 9.17140305e-01 -3.72450382e-01
-8.41667295e-01 -7.58428633e-01 6.86087728e-01 -3.65054786e-01
8.28100920e-01 -4.57867056e-01 -1.02184665e+00 4.50153321e-01
2.32576489e-01 -2.02957928e-01 9.94306862e-01 2.78156936e-01
-2.98482776e-01 -2.40797907e-01 -4.58464324e-01 2.79664457e-01
8.96072984e-01 -7.49067307e-01 -9.27767515e-01 1.73374608e-01
3.95416766e-01 -5.47747314e-01 -7.34187186e-01 3.36779982e-01
5.98136485e-01 -1.06930077e+00 9.44899261e-01 -4.52410102e-01
2.88910776e-01 -3.37511122e-01 1.61761120e-01 -1.48809934e+00
-2.89518416e-01 -1.04886758e+00 -3.09646785e-01 1.39120209e+00
3.03421229e-01 -3.61180790e-02 6.89422250e-01 -4.36070830e-01
-2.30740756e-01 -3.42311174e-01 -8.67536604e-01 -9.09708202e-01
-1.36517406e-01 -6.03061080e-01 2.43206665e-01 1.12299228e+00
2.27882743e-01 8.04054618e-01 -4.09186929e-01 -2.06427768e-01
4.10586327e-01 1.86348855e-01 7.73189425e-01 -1.62097478e+00
-6.90408528e-01 -4.65716690e-01 -3.00564438e-01 -7.71428525e-01
-6.56766370e-02 -9.62662518e-01 1.44659296e-01 -9.14568007e-01
-3.84358466e-01 -4.00218606e-01 -5.86506128e-01 6.10143483e-01
3.61227407e-03 6.02109611e-01 3.12627435e-01 4.63494420e-01
-2.71982044e-01 1.33389726e-01 8.03095400e-01 -9.51148495e-02
-7.67345071e-01 6.71590388e-01 -5.09042799e-01 1.00610816e+00
9.93177831e-01 -6.71615779e-01 -3.55113477e-01 5.42452000e-02
4.28287536e-01 1.84790362e-02 3.59296709e-01 -1.41249752e+00
2.49342337e-01 4.53327447e-01 3.62965912e-01 -5.43666244e-01
5.09672701e-01 -7.01149464e-01 1.09107353e-01 3.81228685e-01
-5.95944047e-01 -2.76374757e-01 2.93049693e-01 3.67114067e-01
-3.99166822e-01 -4.58802938e-01 6.31256223e-01 -1.99425608e-01
-2.04457402e-01 -3.38604212e-01 -3.66195381e-01 1.97778150e-01
4.97940719e-01 -1.82657391e-01 1.98250964e-01 -4.04111743e-01
-1.06216967e+00 -2.23516986e-01 1.41824812e-01 3.44538301e-01
2.34423086e-01 -1.26171434e+00 -5.95392644e-01 3.69173676e-01
-7.74690583e-02 5.28111905e-02 -7.25780651e-02 8.85843337e-01
3.29600833e-02 -1.39220338e-02 -2.46018365e-01 -6.84776127e-01
-1.36664689e+00 3.09925169e-01 4.53140795e-01 -1.88701347e-01
-8.55274916e-01 9.50651407e-01 -2.83751458e-01 -1.14815243e-01
7.73103178e-01 -4.62364137e-01 -4.23736751e-01 5.19902170e-01
4.51114893e-01 2.18029886e-01 2.69662559e-01 -5.69739163e-01
-2.39449069e-01 5.39415061e-01 2.75064379e-01 -3.85173857e-01
1.45089352e+00 2.84671575e-01 1.33382455e-01 1.29067552e+00
8.08091044e-01 3.85764807e-01 -1.28256023e+00 -2.07600966e-01
3.04510504e-01 -3.51028815e-02 -1.56364515e-02 -8.03922951e-01
-7.61177063e-01 6.90421641e-01 5.66277862e-01 7.34331012e-01
1.10769236e+00 -1.41867384e-01 5.15854955e-01 2.83986390e-01
1.45939082e-01 -1.37915301e+00 4.23745424e-01 7.32797801e-01
9.72562611e-01 -1.07790399e+00 -1.95615232e-01 -5.20026162e-02
-5.74387968e-01 1.40200245e+00 3.58491600e-01 -3.64344507e-01
5.67404330e-01 5.72147489e-01 1.16925299e-01 -1.06965750e-01
-6.06642127e-01 -4.11467522e-01 4.83157098e-01 5.20464063e-01
6.49836302e-01 -2.04041943e-01 2.61034310e-01 7.91208446e-01
-1.07456136e+00 6.74947724e-02 3.39193583e-01 8.07990134e-01
-3.94863695e-01 -1.06842506e+00 -5.56656122e-01 1.69178069e-01
-6.45058095e-01 -1.15380891e-01 -7.36545801e-01 5.72007120e-01
3.70373100e-01 8.39566827e-01 2.38436416e-01 -5.50320566e-01
3.53963464e-01 4.95804489e-01 4.88766670e-01 -8.07672977e-01
-1.24423146e+00 5.35081685e-01 6.60716146e-02 -2.05412388e-01
-5.77135801e-01 -3.93582314e-01 -1.25860906e+00 3.21912020e-01
-3.92938673e-01 2.42052689e-01 4.41345364e-01 8.57298255e-01
-7.39463717e-02 9.56730902e-01 7.02887714e-01 -1.01439428e+00
-5.12566745e-01 -1.22786438e+00 -5.15960574e-01 4.97912943e-01
6.71354830e-01 -4.43553239e-01 -4.81637627e-01 3.29803079e-01] | [15.838171005249023, 5.273606300354004] |
a63badf7-f88d-4629-82cb-cf7426d5cc06 | collaborative-residual-metric-learning | 2304.07971 | null | https://arxiv.org/abs/2304.07971v1 | https://arxiv.org/pdf/2304.07971v1.pdf | Collaborative Residual Metric Learning | In collaborative filtering, distance metric learning has been applied to matrix factorization techniques with promising results. However, matrix factorization lacks the ability of capturing collaborative information, which has been remarked by recent works and improved by interpreting user interactions as signals. This paper aims to find out how metric learning connect to these signal-based models. By adopting a generalized distance metric, we discovered that in signal-based models, it is easier to estimate the residual of distances, which refers to the difference between the distances from a user to a target item and another item, rather than estimating the distances themselves. Further analysis also uncovers a link between the normalization strength of interaction signals and the novelty of recommendation, which has been overlooked by existing studies. Based on the above findings, we propose a novel model to learn a generalized distance user-item distance metric to capture user preference in interaction signals by modeling the residuals of distance. The proposed CoRML model is then further improved in training efficiency by a newly introduced approximated ranking weight. Extensive experiments conducted on 4 public datasets demonstrate the superior performance of CoRML compared to the state-of-the-art baselines in collaborative filtering, along with high efficiency and the ability of providing novelty-promoted recommendations, shedding new light on the study of metric learning-based recommender systems. | ['Tommy W. S. Chow', 'Jianghong Ma', 'Tianjun Wei'] | 2023-04-17 | null | null | null | null | ['metric-learning', 'metric-learning', 'collaborative-filtering'] | ['computer-vision', 'methodology', 'miscellaneous'] | [ 1.52620107e-01 -4.53067303e-01 -1.41854540e-01 -5.24038553e-01
-4.41509813e-01 -4.96442646e-01 4.23984498e-01 1.39667615e-01
-3.18158776e-01 3.34698349e-01 6.50981069e-01 -2.50843704e-01
-8.96837413e-01 -6.43037915e-01 -4.56479818e-01 -7.34997034e-01
-4.66788739e-01 -8.30513611e-02 -1.26434803e-01 -3.84759963e-01
3.49999189e-01 1.12000071e-01 -1.48209834e+00 4.31707948e-01
8.01883936e-01 1.10216522e+00 -9.20280814e-02 2.87836134e-01
1.41890943e-01 5.70866764e-01 -5.44757605e-01 -3.70938927e-01
4.74492103e-01 -6.53511822e-01 -2.08763927e-01 -1.63720682e-01
3.58244568e-01 -1.41219227e-02 -3.49928558e-01 8.83859456e-01
5.57276785e-01 4.83772904e-01 5.35750329e-01 -1.08313107e+00
-1.11516702e+00 9.82360959e-01 -5.56388855e-01 3.25083137e-01
5.11620820e-01 -4.03558195e-01 1.60770106e+00 -1.03969169e+00
2.72011429e-01 1.03785789e+00 8.68610740e-01 1.09047249e-01
-1.18338025e+00 -7.43832409e-01 4.99852598e-01 3.64740223e-01
-1.45473254e+00 -4.28518802e-02 7.60784090e-01 -4.87145543e-01
4.76966769e-01 4.52073067e-01 4.83442605e-01 1.07875836e+00
-4.41074297e-02 7.74705589e-01 9.65033770e-01 -3.71612102e-01
1.09243281e-01 2.18698293e-01 1.96623757e-01 3.50660175e-01
1.12357117e-01 3.43646884e-01 -7.13112116e-01 -2.87681818e-01
3.20937812e-01 4.30694461e-01 -4.73794162e-01 -4.19284135e-01
-1.19975626e+00 9.39800620e-01 5.38944006e-01 6.10153258e-01
-3.08091968e-01 -2.17455760e-01 2.01333895e-01 6.73055470e-01
6.10940874e-01 6.61035120e-01 -5.98822296e-01 -1.18464947e-01
-8.84854019e-01 5.50512634e-02 6.88463449e-01 5.30660033e-01
4.05922830e-01 -7.47156236e-03 -3.80212247e-01 8.66510093e-01
3.68889093e-01 3.32092792e-01 5.14927924e-01 -5.90563774e-01
2.37773851e-01 7.76939034e-01 1.95032954e-01 -1.32600510e+00
-4.52030092e-01 -1.00447476e+00 -8.65075886e-01 -1.93390772e-01
5.97869873e-01 -1.98467389e-01 -3.13892365e-02 1.76030648e+00
2.61252075e-01 6.81822658e-01 -3.13445568e-01 1.05546582e+00
4.29219633e-01 2.84598559e-01 -3.79046023e-01 -3.31183106e-01
1.04829788e+00 -7.52656639e-01 -5.66048861e-01 3.56180847e-01
7.34122217e-01 -6.97849691e-01 1.10667443e+00 8.27284455e-01
-5.53390801e-01 -7.81670272e-01 -1.15934908e+00 3.91038269e-01
-2.19929382e-01 2.71989286e-01 1.08920348e+00 1.06341004e+00
-7.43839204e-01 7.38933682e-01 -4.31854904e-01 -1.49044022e-01
1.46361247e-01 1.90871269e-01 -1.83292963e-02 1.13244221e-01
-1.35831940e+00 4.83922213e-01 -1.84518054e-01 1.40895233e-01
-2.17940450e-01 -1.10438585e+00 -4.44331646e-01 7.12307841e-02
4.20992851e-01 -5.58053792e-01 9.11310732e-01 -9.76023078e-01
-1.45499408e+00 -9.56932604e-02 -1.69765800e-02 -5.80038011e-01
2.96259582e-01 -3.12474132e-01 -1.00219119e+00 -3.62130195e-01
-1.46356672e-01 -9.04643834e-02 7.19293058e-01 -9.91301417e-01
-9.82483983e-01 -4.54755455e-01 3.05067688e-01 -3.61671187e-02
-7.38907158e-01 -2.38430738e-01 -1.46158844e-01 -9.25411940e-01
2.51299709e-01 -9.64826345e-01 -2.55311638e-01 -2.28992507e-01
7.40917474e-02 -2.17414856e-01 4.73381579e-01 -3.41804773e-01
1.85744619e+00 -2.28464603e+00 5.43256365e-02 6.75591528e-01
1.98409393e-01 1.06248967e-01 -3.66345197e-01 6.91792250e-01
-2.13855971e-02 -1.79605573e-01 3.99210714e-02 -3.08623090e-02
1.83363408e-01 -6.33739680e-02 -5.31990767e-01 5.31265080e-01
-2.66446084e-01 6.59758329e-01 -1.15495872e+00 1.84223071e-01
1.41424909e-01 7.29680359e-01 -8.29343021e-01 -9.79003906e-02
1.94831729e-01 4.79366571e-01 -2.06685841e-01 3.76639277e-01
5.86526155e-01 -9.07829180e-02 3.94734859e-01 -5.08828759e-01
-1.05846174e-01 3.49626631e-01 -1.62901187e+00 1.76533222e+00
-5.21422267e-01 3.89157385e-01 -2.49078780e-01 -1.19615078e+00
9.49444532e-01 2.21940830e-01 9.40177619e-01 -7.17455924e-01
1.97006240e-02 1.67145804e-02 4.22123164e-01 -1.03736110e-01
4.49673772e-01 1.48489282e-01 6.61275210e-03 7.22717762e-01
-6.02056943e-02 5.92510521e-01 2.97732174e-01 3.03274155e-01
1.05119872e+00 -2.27286182e-02 1.06323563e-01 -1.73873514e-01
6.14353299e-01 -6.23230994e-01 4.02826846e-01 9.55478132e-01
1.82654783e-01 2.90883303e-01 1.59632601e-02 -1.69313878e-01
-3.50287408e-01 -1.03045917e+00 -2.71214664e-01 1.43212044e+00
1.35312825e-01 -8.22670162e-01 -2.26871103e-01 -7.46815860e-01
3.64468515e-01 4.37438071e-01 -8.79094064e-01 -3.55522186e-01
-2.40393147e-01 -1.02420366e+00 3.19913864e-01 5.27715504e-01
-4.45092469e-02 -2.98469305e-01 1.20707313e-02 2.63815910e-01
-1.39573961e-01 -6.54392540e-01 -8.90212774e-01 8.40567350e-02
-9.39700663e-01 -9.89645362e-01 -3.11924338e-01 -3.05602938e-01
5.33522904e-01 8.89500320e-01 8.19064617e-01 1.53675660e-01
7.62643740e-02 4.11335349e-01 -8.37273061e-01 -2.57802606e-01
1.94266856e-01 -5.20837009e-02 4.50757980e-01 7.17022538e-01
6.09319150e-01 -8.07713449e-01 -8.51687729e-01 7.58521378e-01
-7.39029825e-01 -5.34115970e-01 5.67283630e-01 8.65520000e-01
2.86176562e-01 2.57500827e-01 9.07167852e-01 -9.55348909e-01
9.16036487e-01 -8.10009122e-01 -6.96730912e-02 3.09793931e-02
-1.11052716e+00 1.35660589e-01 6.94304943e-01 -7.83652961e-01
-8.91594231e-01 -3.56916696e-01 2.62141461e-03 -1.22575304e-02
2.54386157e-01 7.65693307e-01 -3.18095386e-02 -1.36177182e-01
7.83229053e-01 -2.23306417e-02 -2.65818596e-01 -7.08470404e-01
7.24732220e-01 6.95402026e-01 1.72904849e-01 -5.40164053e-01
7.62656391e-01 4.80582029e-01 -1.73937470e-01 -5.96765220e-01
-1.10357499e+00 -9.10931587e-01 -5.95974982e-01 -5.30177392e-02
1.37073323e-01 -7.41052568e-01 -8.51735830e-01 -1.53275296e-01
-4.59304214e-01 1.88060746e-01 -4.42732394e-01 9.34038103e-01
-9.51741636e-02 5.85955083e-01 -3.86599034e-01 -8.50333154e-01
-1.39831126e-01 -6.30510867e-01 4.90285128e-01 -1.05473123e-01
-2.79986620e-01 -9.83862579e-01 3.24344218e-01 3.30762237e-01
6.45795941e-01 -1.58076152e-01 6.46738231e-01 -8.25592279e-01
-8.23686197e-02 -3.40317011e-01 -1.13664508e-01 3.22385281e-01
4.56116676e-01 -2.66835272e-01 -6.65555358e-01 -4.91967618e-01
9.03046355e-02 3.10843825e-01 8.10170650e-01 3.79804999e-01
9.20231462e-01 -1.74512073e-01 -2.42097467e-01 5.07502973e-01
1.09141207e+00 2.98038006e-01 3.51279914e-01 8.11846033e-02
5.27007878e-01 4.02845442e-01 7.29140818e-01 7.80732751e-01
2.65753239e-01 9.53292906e-01 1.26307100e-01 -5.04631475e-02
1.40507787e-01 -3.15261930e-01 3.47082615e-01 1.00393164e+00
-1.59969211e-01 -1.36033073e-01 -9.59159881e-02 1.93849489e-01
-2.06997609e+00 -1.07549858e+00 -2.85778433e-01 2.42610264e+00
5.09456754e-01 6.01014756e-02 4.38918710e-01 5.12957573e-01
3.70356381e-01 -5.50329797e-02 -4.15497720e-01 -1.99578181e-01
-8.68269876e-02 1.62568510e-01 3.45424742e-01 3.82725924e-01
-9.22393084e-01 3.77473623e-01 5.78143311e+00 6.78634286e-01
-1.02827895e+00 2.29504049e-01 9.99386888e-03 -3.41478497e-01
-3.12184542e-01 -7.49671087e-02 -5.67829490e-01 6.31352603e-01
9.45505261e-01 -1.29920527e-01 6.70707583e-01 6.44291580e-01
5.13812423e-01 2.20209524e-01 -1.26273692e+00 9.27913010e-01
2.69680411e-01 -8.88742685e-01 6.69405535e-02 2.24088892e-01
8.73710692e-01 -1.20822847e-01 2.95137972e-01 5.03119469e-01
2.07740575e-01 -7.18064427e-01 6.12429023e-01 8.33239436e-01
7.58733824e-02 -6.96118414e-01 7.80495703e-01 2.00587943e-01
-1.28744447e+00 -4.81045127e-01 -4.76347119e-01 -4.91844386e-01
1.64744686e-02 1.03027284e+00 -6.68696225e-01 9.19594169e-01
5.74447989e-01 9.50277865e-01 -4.68815804e-01 1.14998078e+00
-2.94965021e-02 8.32725883e-01 -8.17584470e-02 6.44613430e-02
3.54496129e-02 -6.47724807e-01 2.90946573e-01 1.15685773e+00
6.50328517e-01 -6.11990131e-02 1.98193148e-01 5.47288954e-01
1.43701620e-02 5.55365503e-01 -3.40706468e-01 4.11370546e-02
5.44708848e-01 1.30723143e+00 -4.19331074e-01 9.75044668e-02
-7.53388762e-01 7.88835347e-01 1.03661511e-02 2.59203225e-01
-8.36335301e-01 -3.43841910e-01 8.13153684e-01 1.65163040e-01
4.85882699e-01 -3.20370048e-01 -8.79701748e-02 -1.18298113e+00
2.16394272e-02 -9.36067700e-01 4.68372762e-01 -1.67934090e-01
-1.60747457e+00 2.57590681e-01 -3.10150623e-01 -1.64371824e+00
1.00075155e-02 -4.75489348e-01 -4.47745174e-01 5.80295146e-01
-1.01587081e+00 -9.52269077e-01 -7.78917363e-03 6.50052011e-01
2.54528850e-01 -2.68451482e-01 7.30592489e-01 7.67587662e-01
-3.38398457e-01 9.24732089e-01 5.54763377e-01 -1.78818256e-01
9.53054965e-01 -1.19889152e+00 9.26218405e-02 6.96287513e-01
9.34361637e-01 1.13314521e+00 6.07695699e-01 -3.33274901e-01
-1.62418902e+00 -8.23909700e-01 7.39922941e-01 -7.63679504e-01
8.79421055e-01 -4.04756457e-01 -7.28511274e-01 3.34163994e-01
-2.51842588e-01 -8.43169689e-02 1.43029714e+00 7.37185180e-01
-6.39925897e-01 -3.64588201e-01 -8.80178928e-01 4.54236418e-01
1.34631932e+00 -5.16315579e-01 -3.27827930e-01 5.79877235e-02
5.03185630e-01 8.82680714e-02 -1.01973867e+00 1.85604021e-01
1.00607347e+00 -9.93916988e-01 1.14225626e+00 -5.87007165e-01
-1.75506681e-01 -5.90908349e-01 -5.20843029e-01 -1.52149069e+00
-7.98563600e-01 -4.06236589e-01 -4.43203360e-01 1.22509432e+00
4.45621401e-01 -4.27288502e-01 4.84738916e-01 1.43132821e-01
1.96019020e-02 -7.81649411e-01 -6.44199729e-01 -8.11455905e-01
-2.66984791e-01 -7.16094911e-01 7.05647647e-01 1.09149492e+00
2.97180235e-01 4.76088554e-01 -8.42737496e-01 1.98912993e-01
3.30279738e-01 1.83723181e-01 7.13006556e-01 -1.67194319e+00
-7.04440415e-01 -5.14181614e-01 -4.08222437e-01 -1.27887368e+00
-9.13709849e-02 -1.09656942e+00 -3.98450553e-01 -1.18059897e+00
-9.27366316e-02 -4.15121019e-01 -1.04763198e+00 1.47159956e-02
-1.57259896e-01 5.06528378e-01 2.01400332e-02 6.68934435e-02
-7.06310987e-01 4.20385987e-01 1.00745964e+00 -1.71497613e-01
-5.02664149e-01 4.63335663e-01 -1.38746715e+00 4.28809673e-01
4.47343379e-01 -3.75347972e-01 -7.48969972e-01 -1.22082643e-01
6.14864647e-01 -2.92981565e-01 -1.49846271e-01 -9.38194275e-01
2.46553451e-01 1.87250525e-02 3.74302506e-01 -2.69553900e-01
7.05698431e-02 -1.03220022e+00 2.03379646e-01 2.28595808e-01
-4.97100532e-01 -1.12492360e-01 -3.44413340e-01 9.92541790e-01
-1.51997805e-01 8.46161693e-02 1.88260138e-01 2.63815761e-01
-4.16415423e-01 1.46661520e-01 -1.87040076e-01 -2.95300156e-01
6.59545422e-01 -1.73381433e-01 1.49428248e-01 -4.71929789e-01
-6.62014663e-01 -3.53495404e-02 -8.67937952e-02 8.34679782e-01
4.06915158e-01 -1.47897732e+00 -6.64129138e-01 2.56313145e-01
1.98902890e-01 -1.00797474e+00 3.48061204e-01 1.22119904e+00
4.88201559e-01 3.64214718e-01 1.87794954e-01 -3.00391227e-01
-9.97138679e-01 6.81135893e-01 2.31465138e-02 -2.55448341e-01
-4.50289190e-01 6.13277733e-01 -4.44861092e-02 -4.37851250e-01
4.46296155e-01 -5.36360443e-01 -2.82248825e-01 4.41361547e-01
7.86777675e-01 6.43442273e-01 3.68262917e-01 -3.89108300e-01
-3.99877131e-01 5.10856032e-01 -1.11941889e-01 7.11706430e-02
1.27167642e+00 -3.66483718e-01 1.19525760e-01 5.19929409e-01
1.03366876e+00 5.10739088e-01 -9.10652459e-01 -6.34467840e-01
2.68731713e-01 -7.87510216e-01 1.51587591e-01 -9.20509815e-01
-1.10897970e+00 4.46365833e-01 9.19076324e-01 2.77996153e-01
1.06258249e+00 -3.48164618e-01 7.45579302e-01 3.90449673e-01
5.78077435e-01 -9.72441971e-01 1.09610669e-01 3.75669420e-01
6.27955854e-01 -1.22760022e+00 1.82087086e-02 -5.90338483e-02
-4.59206492e-01 8.48025799e-01 1.22508712e-01 -1.38921455e-01
1.15291619e+00 2.86721792e-02 -1.44762089e-02 4.94793542e-02
-5.80513239e-01 -2.59452611e-01 7.67007470e-01 5.65838754e-01
8.65983367e-01 2.26786792e-01 -7.19654918e-01 1.18932045e+00
-1.73207551e-01 -7.60364532e-02 1.35187641e-01 5.01796544e-01
-3.03252518e-01 -1.44357669e+00 -5.63994460e-02 6.38164639e-01
-4.38232034e-01 -2.26259276e-01 -2.59436399e-01 3.87798786e-01
2.68869936e-01 1.44372845e+00 -1.87801525e-01 -9.50827360e-01
6.37623072e-01 -1.07847929e-01 3.30743104e-01 -5.54527044e-01
-8.03074181e-01 -4.73705307e-02 -2.16462374e-01 -4.98881519e-01
-5.42848825e-01 -5.52378356e-01 -7.64466345e-01 -1.13439888e-01
-7.02588201e-01 6.77112699e-01 6.82688892e-01 8.81150842e-01
5.12731135e-01 4.43396151e-01 1.07664311e+00 -5.70365548e-01
-6.91997647e-01 -1.06418455e+00 -8.90363514e-01 5.60267508e-01
5.88480011e-02 -7.80347645e-01 -5.10409355e-01 -3.07588935e-01] | [10.078185081481934, 5.603568077087402] |
c6420f2b-2af6-4dbf-8d49-f2344f4d3300 | background-foreground-segmentation-for | 2109.0941 | null | https://arxiv.org/abs/2109.09410v1 | https://arxiv.org/pdf/2109.09410v1.pdf | Background-Foreground Segmentation for Interior Sensing in Automotive Industry | To ensure safety in automated driving, the correct perception of the situation inside the car is as important as its environment. Thus, seat occupancy detection and classification of detected instances play an important role in interior sensing. By the knowledge of the seat occupancy status, it is possible to, e.g., automate the airbag deployment control. Furthermore, the presence of a driver, which is necessary for partially automated driving cars at the automation levels two to four can be verified. In this work, we compare different statistical methods from the field of image segmentation to approach the problem of background-foreground segmentation in camera based interior sensing. In the recent years, several methods based on different techniques have been developed and applied to images or videos from different applications. The peculiarity of the given scenarios of interior sensing is, that the foreground instances and the background both contain static as well as dynamic elements. In data considered in this work, even the camera position is not completely fixed. We review and benchmark three different methods ranging, i.e., Gaussian Mixture Models (GMM), Morphological Snakes and a deep neural network, namely a Mask R-CNN. In particular, the limitations of the classical methods, GMM and Morphological Snakes, for interior sensing are shown. Furthermore, it turns, that it is possible to overcome these limitations by deep learning, e.g.\ using a Mask R-CNN. Although only a small amount of ground truth data was available for training, we enabled the Mask R-CNN to produce high quality background-foreground masks via transfer learning. Moreover, we demonstrate that certain augmentation as well as pre- and post-processing methods further enhance the performance of the investigated methods. | ['Thomas Kurbiel', 'Klaus Friedrichs', 'Hanno Gottschalk', 'Matthias Rottmann', 'Claudia Drygala'] | 2021-09-20 | null | null | null | null | ['foreground-segmentation'] | ['computer-vision'] | [ 3.13498080e-01 1.94342807e-01 1.05892427e-01 -2.25795105e-01
-2.52971500e-01 -4.60882217e-01 5.40270269e-01 1.62458241e-01
-6.09424114e-01 6.41204417e-01 -7.06268370e-01 -4.55883682e-01
-7.48222321e-02 -8.47540259e-01 -7.54662514e-01 -9.40973043e-01
2.50115305e-01 5.97864091e-01 5.43372691e-01 -1.90960228e-01
7.62333795e-02 1.03610253e+00 -2.08116031e+00 -7.92297870e-02
6.53921247e-01 1.08923876e+00 4.28831041e-01 5.63626945e-01
-1.98696524e-01 3.34296525e-01 -5.85784316e-01 -1.56730473e-01
2.38981381e-01 -6.97473064e-02 -2.87355572e-01 4.33089286e-01
3.57451856e-01 -1.08106382e-01 -4.97157723e-02 1.11414039e+00
6.17152639e-02 2.56046534e-01 6.46636367e-01 -1.09453690e+00
1.45753607e-01 2.25027099e-01 -2.59586245e-01 1.47289291e-01
7.62854740e-02 1.84439838e-01 2.33988643e-01 -7.24293232e-01
4.40832466e-01 9.45332050e-01 5.11705399e-01 4.64476734e-01
-1.10991108e+00 -4.72040415e-01 2.48828515e-01 3.57910037e-01
-1.49122679e+00 -3.28910977e-01 1.11494172e+00 -6.41089678e-01
1.79870829e-01 4.37198430e-01 7.43833423e-01 9.25401688e-01
3.90676036e-02 6.45696282e-01 1.24367249e+00 -4.04715985e-01
5.00118852e-01 6.57861590e-01 1.64963022e-01 5.90433657e-01
3.56368303e-01 4.40647034e-03 1.78164423e-01 1.21364862e-01
6.02428734e-01 -1.14243655e-02 -1.85796738e-01 -5.13047099e-01
-8.97365570e-01 6.45670295e-01 2.07030222e-01 7.44720399e-01
-3.73404950e-01 5.89706749e-02 9.39870030e-02 -3.61196160e-01
2.91853786e-01 -6.74479082e-02 -2.14906782e-01 1.28214955e-01
-1.18225956e+00 6.31147474e-02 7.47763872e-01 6.14474475e-01
1.00795054e+00 1.43465936e-01 5.37691638e-02 2.89006889e-01
2.28025660e-01 6.05604887e-01 1.73029080e-01 -8.64503205e-01
1.81491479e-01 4.03570235e-01 2.22762376e-01 -1.23899686e+00
-5.00944853e-01 -5.08568287e-01 -8.77941966e-01 6.79105520e-01
6.77788794e-01 6.68767327e-03 -8.78849030e-01 1.50863922e+00
5.92379928e-01 3.27003062e-01 1.93583649e-02 8.79762709e-01
5.79091132e-01 4.76050735e-01 -4.24904786e-02 -2.72918046e-01
1.47530186e+00 -4.30392206e-01 -9.81706560e-01 -3.75000030e-01
9.01321247e-02 -6.29654467e-01 6.65205896e-01 7.54860103e-01
-1.01136565e+00 -1.00138998e+00 -1.13428533e+00 3.81310016e-01
-7.64864981e-01 4.59139585e-01 2.77862519e-01 1.00186169e+00
-8.44684660e-01 6.36058509e-01 -9.34940100e-01 -1.78372458e-01
1.85818657e-01 4.24101710e-01 -3.01082999e-01 2.24009261e-01
-9.76625144e-01 9.91319299e-01 3.49845320e-01 5.99384904e-01
-8.83682489e-01 -2.68714517e-01 -7.13396072e-01 7.79451511e-04
5.26819527e-01 -4.04561669e-01 8.36986303e-01 -1.05789268e+00
-1.38566351e+00 9.94923234e-01 -5.81589416e-02 -5.78238010e-01
1.04425335e+00 3.90038081e-02 -2.95565456e-01 1.68326795e-01
-2.47766390e-01 5.46615005e-01 1.07201493e+00 -1.72019708e+00
-6.43712282e-01 -3.89140487e-01 3.65322419e-02 -3.46708328e-01
-6.47150576e-02 -2.11647317e-01 -4.34599549e-01 -1.92619696e-01
-3.30329873e-02 -8.53274167e-01 -4.77606207e-01 -1.21891387e-01
-3.83043259e-01 2.47131623e-02 1.02367711e+00 -8.53006542e-01
7.39164650e-01 -2.22508025e+00 7.38385618e-02 4.02123064e-01
5.91306649e-02 4.79641140e-01 3.68593007e-01 -1.82939321e-01
3.02712042e-02 -2.15920627e-01 -5.70227981e-01 -5.38284659e-01
-6.05564378e-02 4.47792798e-01 1.17444910e-01 7.49611735e-01
2.03648776e-01 5.16450226e-01 -5.73153138e-01 -7.53392458e-01
8.87432992e-01 7.21336603e-01 -1.18784029e-02 1.62284032e-01
-2.04186216e-01 8.57331991e-01 -2.71347672e-01 4.30900186e-01
9.01715398e-01 3.48372042e-01 5.06538786e-02 -2.32817978e-01
-4.18479443e-01 -3.74123394e-01 -1.52338231e+00 1.25395501e+00
-6.09837294e-01 5.96323550e-01 6.55286968e-01 -1.33189809e+00
1.14258742e+00 3.63891304e-01 6.25701725e-01 -5.37835658e-01
4.90234286e-01 3.83920908e-01 -2.10791230e-01 -5.21428764e-01
5.74347675e-01 -6.54465482e-02 2.15753406e-01 -2.10445434e-01
-3.47814381e-01 -1.97117269e-01 3.38232994e-01 -4.76545602e-01
6.24957263e-01 8.13224465e-02 2.19355464e-01 -1.49523348e-01
1.05712140e+00 5.50043769e-02 4.56070125e-01 5.39362967e-01
-1.06700987e-01 5.32890916e-01 2.72589147e-01 -2.66663164e-01
-7.68071413e-01 -8.81768525e-01 -3.68559748e-01 4.00569737e-01
3.31075311e-01 4.03801203e-01 -1.16073489e+00 -4.08488631e-01
-5.65824807e-02 7.86822498e-01 -5.43200016e-01 -2.81481706e-02
-7.76700199e-01 -6.83515906e-01 7.44086206e-02 3.23796183e-01
5.12080491e-01 -1.01208687e+00 -1.12016547e+00 3.11961353e-01
-7.54933059e-02 -1.55967963e+00 2.24475667e-01 3.20733160e-01
-9.17647362e-01 -1.03702962e+00 -5.81253409e-01 -4.67156827e-01
5.25498688e-01 1.90278918e-01 1.02653849e+00 3.96309160e-02
-4.32057410e-01 4.20248568e-01 -1.27940506e-01 -3.91416490e-01
-8.35009515e-01 -1.49422631e-01 -1.72298357e-01 5.50148606e-01
1.77166000e-01 -5.25117993e-01 -4.66207445e-01 4.77392435e-01
-8.15443814e-01 -5.71776181e-02 6.65377736e-01 1.88339233e-01
6.85568571e-01 2.79180080e-01 7.97193050e-02 -7.24502504e-01
5.96052920e-03 -1.68161571e-01 -1.03530276e+00 -5.15655987e-02
-2.64743567e-01 -3.86491299e-01 6.87448025e-01 -3.38519573e-01
-9.25376058e-01 5.15663981e-01 -4.27849382e-01 -6.41414881e-01
-9.91391718e-01 -7.74003938e-02 -5.63296378e-01 -9.74770486e-02
4.12513226e-01 2.89101779e-01 9.24796239e-02 -4.78107542e-01
2.80637771e-01 5.25733173e-01 6.77340984e-01 -2.96213984e-01
9.41901624e-01 9.28260207e-01 2.88546592e-01 -1.23958564e+00
-2.87772745e-01 -5.40429711e-01 -7.78734326e-01 -7.16435313e-01
1.36924863e+00 -5.72066128e-01 -8.37684095e-01 2.77860761e-01
-1.29413259e+00 -3.50804359e-01 -4.36830759e-01 3.97949487e-01
-6.84291542e-01 4.41730320e-01 -2.16349095e-01 -1.38682663e+00
1.72855303e-01 -1.34954846e+00 9.59427536e-01 2.13691816e-01
1.88406587e-01 -9.62751269e-01 -3.69957387e-01 5.50038517e-01
1.94037259e-01 4.98631060e-01 5.75330377e-01 -4.22567159e-01
-7.53997564e-01 -3.12654436e-01 9.87377912e-02 5.52834034e-01
1.02811150e-01 2.76461393e-01 -1.28141809e+00 1.53548243e-02
2.78324723e-01 4.56959277e-01 7.54599333e-01 6.45794690e-01
1.22566402e+00 2.19772235e-01 -4.62438464e-01 3.04241449e-01
1.33123791e+00 4.72227991e-01 7.20784605e-01 2.42370784e-01
5.13104260e-01 9.50792670e-01 8.73271108e-01 2.62964219e-01
1.38494326e-02 1.07903874e+00 7.89996207e-01 -5.21741211e-01
-8.50504637e-02 2.74543613e-01 3.16072166e-01 3.31730664e-01
-2.92765409e-01 -1.36325462e-02 -7.91544497e-01 4.19392407e-01
-1.67937016e+00 -8.74935567e-01 -5.46273351e-01 2.32146740e+00
1.99933782e-01 4.44190115e-01 1.50880218e-01 6.09808028e-01
9.21698451e-01 -8.04234967e-02 -3.01075310e-01 -3.07113260e-01
-5.51807843e-02 1.17372595e-01 6.56205535e-01 6.05794370e-01
-1.36084294e+00 4.55654919e-01 4.80029678e+00 8.43819737e-01
-1.12971330e+00 3.37378174e-01 6.03551209e-01 3.56581718e-01
-4.39022668e-02 -2.63815343e-01 -9.56957936e-01 5.11013925e-01
7.78374255e-01 4.94998485e-01 3.29527110e-01 9.18720841e-01
4.48945969e-01 -4.81894791e-01 -7.68964827e-01 9.77080405e-01
9.58018154e-02 -1.01672149e+00 -5.10248542e-01 1.91016346e-01
3.23370397e-01 -3.30722213e-01 -1.63891967e-02 5.25432825e-02
-2.97550380e-01 -7.45670438e-01 1.01620150e+00 6.68179214e-01
2.87672162e-01 -7.68521965e-01 8.55661094e-01 6.16854787e-01
-1.28997982e+00 -1.31476730e-01 -2.26753816e-01 9.99288410e-02
4.13760126e-01 8.58396649e-01 -6.19693220e-01 7.31166601e-01
5.02054751e-01 1.70078889e-01 -5.41669250e-01 7.81324565e-01
-1.09370194e-01 4.65134323e-01 -4.12070811e-01 1.23616210e-05
2.07529381e-01 -5.91150999e-01 6.19906843e-01 1.26017046e+00
2.46873975e-01 -2.63187468e-01 2.39401281e-01 1.26700580e+00
5.73525786e-01 4.42949310e-02 -5.93545079e-01 4.23670560e-01
-1.64973930e-01 1.69083190e+00 -1.19947779e+00 -2.91962147e-01
-3.63192886e-01 5.87097347e-01 -2.88968086e-01 3.10772240e-01
-1.03736222e+00 1.26212230e-02 5.78332901e-01 5.19305289e-01
4.54181671e-01 -4.67603236e-01 -2.35131472e-01 -7.06837714e-01
-2.48754323e-02 -4.89113271e-01 -1.62056983e-02 -5.42756319e-01
-6.98610961e-01 5.28173506e-01 2.74961025e-01 -1.09185147e+00
-1.12673655e-01 -9.41983581e-01 -3.97772610e-01 6.45522416e-01
-1.65740597e+00 -9.59493637e-01 -5.22229671e-01 4.56797421e-01
4.78495687e-01 1.49286538e-01 3.27099383e-01 5.29329658e-01
-6.45628393e-01 4.17108163e-02 -3.72837707e-02 7.33380392e-03
6.93146288e-02 -1.27865934e+00 1.29405707e-02 1.04902971e+00
1.12067379e-01 1.85167193e-01 9.62748170e-01 -3.93689036e-01
-1.12059259e+00 -1.00211692e+00 4.12877291e-01 -2.65599132e-01
3.17931563e-01 -5.28315663e-01 -1.05582118e+00 3.83190542e-01
1.31091550e-01 1.51222244e-01 1.11827195e-01 -4.52779025e-01
4.38387424e-01 -3.58400673e-01 -1.14711642e+00 4.29986268e-01
5.99956989e-01 -2.82658666e-01 -5.04693329e-01 2.58421004e-01
2.41773784e-01 -3.81895006e-01 -6.41788483e-01 5.03509760e-01
1.60499111e-01 -1.41236699e+00 1.15872359e+00 -4.37559113e-02
-1.41652316e-01 -5.37006497e-01 2.02150587e-02 -7.45537162e-01
1.21433780e-01 -2.77455181e-01 -3.90051156e-02 1.24289489e+00
8.61058161e-02 -6.06433094e-01 1.02935612e+00 5.00743985e-01
-2.90469855e-01 -3.53361666e-01 -1.27323711e+00 -7.68421233e-01
-7.13573918e-02 -7.08987117e-01 3.21784854e-01 5.36329329e-01
-7.52335727e-01 -1.23949014e-01 -6.96089817e-03 4.84602600e-01
6.60305798e-01 1.61862582e-01 9.71618116e-01 -1.45656741e+00
-2.52646089e-01 -4.95575786e-01 -6.06233835e-01 -7.07503617e-01
3.08464438e-01 -5.48274279e-01 2.53349960e-01 -1.40857160e+00
-2.41761789e-01 -4.97348011e-01 -1.87629849e-01 3.49228382e-02
4.87782769e-02 2.30851471e-01 2.58887652e-02 -2.20974043e-01
-1.63662553e-01 1.93679765e-01 1.15259397e+00 -2.13327169e-01
-6.29222542e-02 4.75411385e-01 1.50605157e-01 9.03970182e-01
9.12679076e-01 -3.05377662e-01 -1.64237201e-01 2.33465582e-01
3.35319666e-03 9.35934559e-02 7.91725457e-01 -1.40720534e+00
1.82232708e-01 3.79103906e-02 2.92588562e-01 -8.38294387e-01
7.56506205e-01 -1.40003085e+00 4.18059468e-01 7.44393647e-01
1.81157008e-01 -1.83115095e-01 2.82878160e-01 5.19153476e-01
-2.48240620e-01 -5.42984784e-01 9.53989208e-01 -3.00221175e-01
-7.01496720e-01 -1.04117505e-01 -8.44293952e-01 -3.68116587e-01
1.21357381e+00 -7.46045411e-01 2.55548030e-01 -2.12649196e-01
-8.93727481e-01 -2.95680404e-01 3.36685926e-01 4.51109856e-02
4.29074168e-01 -8.66593540e-01 -3.61777157e-01 2.02836081e-01
-8.05490762e-02 1.04655452e-01 4.34020460e-01 1.28516483e+00
-4.20959651e-01 3.23207676e-01 -1.05137251e-01 -8.40671003e-01
-1.40718794e+00 1.03010619e+00 5.64616799e-01 2.96137165e-02
-4.58925486e-01 1.47809759e-01 1.15639225e-01 -2.02059820e-01
1.00307897e-01 -7.37200439e-01 -4.91986692e-01 1.20333344e-01
3.60215634e-01 4.28694665e-01 3.26649785e-01 -8.92757952e-01
-4.44492966e-01 8.64771903e-01 6.21709585e-01 8.20463989e-03
8.93610299e-01 -1.36671722e-01 -2.21616868e-02 5.70711136e-01
8.38583529e-01 2.09659457e-01 -1.26887798e+00 2.63379782e-01
-7.46393204e-02 -2.19954848e-01 1.84307843e-01 -2.40821064e-01
-1.15908825e+00 1.14660561e+00 9.01416361e-01 5.74880004e-01
1.15590942e+00 -1.61643311e-01 3.89220387e-01 1.70609340e-01
4.20617551e-01 -1.09573233e+00 -3.82766247e-01 -9.79165286e-02
5.21882534e-01 -1.24405789e+00 -2.29501367e-01 -7.38184512e-01
-2.47180700e-01 1.35662377e+00 3.06115717e-01 1.02484897e-01
6.32449090e-01 5.05416095e-01 2.30343789e-01 3.83816957e-02
3.20348255e-02 -6.97756946e-01 1.64960817e-01 5.87505341e-01
-3.44719328e-02 2.31041268e-01 -1.55540138e-01 3.39378387e-01
2.39059404e-02 -3.26722153e-02 4.38382119e-01 6.80897355e-01
-5.62295794e-01 -8.70542288e-01 -1.06449819e+00 -7.45958239e-02
-2.82681048e-01 5.16947865e-01 -2.66053647e-01 1.23117888e+00
6.98781073e-01 1.24393141e+00 1.41542345e-01 -1.73984200e-01
5.16431153e-01 8.14715307e-03 4.24961627e-01 -2.46400699e-01
-5.79893172e-01 7.23324567e-02 -9.13576335e-02 -4.36905205e-01
-7.25527644e-01 -8.06553185e-01 -1.13602054e+00 9.63738468e-03
-4.15004998e-01 1.15118690e-01 1.17886376e+00 1.02717483e+00
-1.88681290e-01 7.23408759e-01 4.25992101e-01 -1.22686958e+00
-2.41575330e-01 -6.89030230e-01 -6.14421010e-01 2.91311681e-01
2.68708289e-01 -9.20044482e-01 -4.91609544e-01 5.96218817e-02] | [8.370895385742188, -1.0405975580215454] |
735af96d-67ea-492e-ac4a-64491b6673bc | dhrl-fnmr-an-intelligent-multicast-routing | 2305.19077 | null | https://arxiv.org/abs/2305.19077v1 | https://arxiv.org/pdf/2305.19077v1.pdf | DHRL-FNMR: An Intelligent Multicast Routing Approach Based on Deep Hierarchical Reinforcement Learning in SDN | The optimal multicast tree problem in the Software-Defined Networking (SDN) multicast routing is an NP-hard combinatorial optimization problem. Although existing SDN intelligent solution methods, which are based on deep reinforcement learning, can dynamically adapt to complex network link state changes, these methods are plagued by problems such as redundant branches, large action space, and slow agent convergence. In this paper, an SDN intelligent multicast routing algorithm based on deep hierarchical reinforcement learning is proposed to circumvent the aforementioned problems. First, the multicast tree construction problem is decomposed into two sub-problems: the fork node selection problem and the construction of the optimal path from the fork node to the destination node. Second, based on the information characteristics of SDN global network perception, the multicast tree state matrix, link bandwidth matrix, link delay matrix, link packet loss rate matrix, and sub-goal matrix are designed as the state space of intrinsic and meta controllers. Then, in order to mitigate the excessive action space, our approach constructs different action spaces at the upper and lower levels. The meta-controller generates an action space using network nodes to select the fork node, and the intrinsic controller uses the adjacent edges of the current node as its action space, thus implementing four different action selection strategies in the construction of the multicast tree. To facilitate the intelligent agent in constructing the optimal multicast tree with greater speed, we developed alternative reward strategies that distinguish between single-step node actions and multi-step actions towards multiple destination nodes. | ['Qiuxiang Jiang', 'Yejin Yang', 'Hongwen Hu', 'Jinqiang Li', 'Xingsi Xue', 'Chenwei Zhao', 'Miao Ye'] | 2023-05-30 | null | null | null | null | ['combinatorial-optimization', 'hierarchical-reinforcement-learning'] | ['methodology', 'methodology'] | [-1.06139459e-01 3.02588075e-01 -4.79172260e-01 -1.55589968e-01
-3.75233358e-03 -2.90573210e-01 -6.17030784e-02 -1.69151619e-01
-1.07245252e-01 1.08249116e+00 -3.22673589e-01 -2.48477951e-01
-6.07704937e-01 -1.03706741e+00 -4.20358405e-02 -8.28451693e-01
-5.43381095e-01 4.93723422e-01 5.31656504e-01 -1.67466309e-02
5.71412504e-01 5.16514182e-01 -1.32926083e+00 -2.95014858e-01
9.36536312e-01 1.14955330e+00 2.07059324e-01 5.35913765e-01
-6.05257630e-01 7.95350850e-01 -6.81987166e-01 1.73061952e-01
4.09694344e-01 -7.56605804e-01 -8.99552584e-01 2.36866727e-01
-4.73847568e-01 -5.22593498e-01 7.49497190e-02 8.25617135e-01
4.43562329e-01 -7.86227658e-02 2.96811257e-02 -1.96809804e+00
-1.86557695e-01 1.01338935e+00 -3.79156858e-01 1.41142517e-01
4.35363874e-02 5.10160983e-01 1.18660164e+00 7.48070329e-02
7.13775337e-01 1.31648684e+00 2.34585211e-01 5.92894971e-01
-1.28439200e+00 -9.39084351e-01 5.95702827e-01 2.86947370e-01
-8.39981973e-01 -3.89255047e-01 8.18426669e-01 -1.78030089e-01
5.53187013e-01 -7.33271316e-02 8.41150284e-01 7.65484750e-01
3.01338226e-01 1.34190738e-01 7.49625921e-01 -1.11902550e-01
7.29105651e-01 -1.87847883e-01 -3.14820528e-01 6.46746039e-01
1.18401438e-01 4.69277591e-01 3.61758545e-02 -2.21939400e-01
1.16184509e+00 -2.23536074e-01 -4.31126058e-02 -7.81319559e-01
-1.04315007e+00 9.51059580e-01 6.18344009e-01 1.71549529e-01
-7.37979770e-01 4.79100972e-01 4.42704886e-01 6.08903766e-01
-2.98831880e-01 6.29391491e-01 -7.25521743e-01 -1.22295342e-01
-3.82395864e-01 -1.60996914e-02 1.06069338e+00 9.23191547e-01
1.06035447e+00 2.67948121e-01 1.25941755e-02 6.22516572e-01
6.15820348e-01 1.69818178e-01 -1.28985748e-01 -1.64455366e+00
3.83737713e-01 6.40466571e-01 3.63611057e-02 -9.06085074e-01
-6.15206301e-01 -4.88657266e-01 -8.39859247e-01 6.06893122e-01
1.85666144e-01 -7.23657012e-01 -6.96113825e-01 2.00802064e+00
6.19276285e-01 3.85353446e-01 1.32848397e-01 8.00768673e-01
1.45558849e-01 8.93845856e-01 2.92293280e-01 -8.20858121e-01
7.10110128e-01 -1.01124096e+00 -7.06585407e-01 -3.87765877e-02
6.31372154e-01 -3.64644051e-01 4.15608644e-01 1.13671850e-02
-9.85854030e-01 -2.08261266e-01 -1.01709998e+00 8.21379185e-01
-1.46252692e-01 -4.31699812e-01 5.94293773e-01 4.42040890e-01
-1.15013361e+00 6.05398953e-01 -5.59678197e-01 -4.26760018e-01
3.04968625e-01 5.10773599e-01 1.74821064e-01 1.51104704e-01
-1.19655490e+00 5.71387231e-01 6.01954043e-01 -7.21014664e-02
-1.15559721e+00 -6.11887932e-01 -4.23205525e-01 3.48787040e-01
9.45930600e-01 -6.97730303e-01 1.17340875e+00 -1.22337055e+00
-2.05481219e+00 1.05724260e-01 5.20983815e-01 -2.26320192e-01
1.80362627e-01 5.74594021e-01 -5.66175282e-01 1.91747561e-01
1.97733000e-01 6.08240247e-01 9.13238227e-01 -1.41637087e+00
-1.03496075e+00 -2.02674732e-01 3.40582281e-01 6.69155419e-02
9.27911103e-02 -1.90796465e-01 -2.33089831e-02 -3.74642223e-01
3.05383593e-01 -7.66158342e-01 -7.58294582e-01 1.12001352e-01
-2.47269854e-01 -2.48513728e-01 9.77978170e-01 -1.73044729e-03
1.37034011e+00 -1.97627223e+00 2.77039587e-01 6.12235904e-01
3.85012925e-01 5.77981062e-02 -4.69649136e-01 4.47003871e-01
2.37641737e-01 3.94753188e-01 9.34985131e-02 4.29610044e-01
-1.66690737e-01 5.03771245e-01 1.89560652e-01 -1.15093067e-01
-1.68703813e-02 1.51503742e-01 -1.01299262e+00 -5.58357537e-01
-9.04120505e-02 -1.98385417e-01 -8.35150242e-01 2.27403805e-01
-5.09196818e-01 5.68786204e-01 -9.40596461e-01 4.25463408e-01
6.51447892e-01 -1.76700473e-01 6.19813621e-01 3.98898832e-02
-5.10535479e-01 1.91935882e-01 -1.50475669e+00 1.20278144e+00
-4.88871902e-01 -4.30184938e-02 7.14146614e-01 -1.11034298e+00
1.01511335e+00 3.91427845e-01 1.10009277e+00 -6.79295182e-01
8.18154737e-02 2.57929772e-01 3.05669427e-01 -5.28236389e-01
-1.40596166e-01 -2.49216139e-01 2.57925063e-01 6.42843604e-01
-9.99879763e-02 2.49465451e-01 2.53967792e-01 -9.13577527e-02
1.37031233e+00 -1.13720350e-01 1.17547652e-02 -1.26466211e-02
7.45086432e-01 -1.70416623e-01 1.24216068e+00 5.65983891e-01
-7.00354338e-01 -3.73118281e-01 1.22165239e+00 -3.30106765e-01
-8.46229672e-01 -1.17670429e+00 3.11957002e-01 1.03979897e+00
3.96764755e-01 -1.91471174e-01 -6.38127029e-01 -7.03156292e-01
-6.12934828e-02 5.00359595e-01 -2.42446288e-01 -3.05458277e-01
-6.80809021e-01 -3.13821912e-01 1.28183812e-01 2.51153130e-02
6.57830775e-01 -1.44449472e+00 -6.45609081e-01 7.80265808e-01
8.67788494e-02 -1.08842468e+00 -3.37025017e-01 2.45398149e-01
-1.05686426e+00 -1.22232592e+00 -2.59197485e-02 -1.03595245e+00
5.14235914e-01 -3.23501974e-02 7.26985276e-01 3.62213045e-01
-1.02754302e-01 1.43930003e-01 -2.88536370e-01 2.20336705e-01
-7.67911196e-01 1.89286634e-01 4.34065843e-03 4.44001444e-02
-2.60373235e-01 -9.19251919e-01 -5.82528889e-01 6.38046086e-01
-5.39922059e-01 -5.17978035e-02 7.62486577e-01 6.18946016e-01
5.21243632e-01 3.98828924e-01 1.19240141e+00 -5.59593737e-01
4.38047469e-01 -6.10031247e-01 -9.65318382e-01 7.13849664e-02
-8.46227348e-01 2.17286229e-01 9.65949237e-01 -2.08431810e-01
-7.47666061e-01 -8.49607438e-02 1.19010061e-01 -2.58410603e-01
1.09276593e-01 2.50517458e-01 -5.62596142e-01 -1.81045581e-03
1.71394214e-01 -5.39632253e-02 4.93111193e-01 -2.40988389e-01
1.42577320e-01 5.23364007e-01 -3.77424777e-01 -5.90791404e-01
7.46039212e-01 8.55190977e-02 3.43889475e-01 -4.54161227e-01
-2.98047394e-01 2.21238032e-01 -2.75487423e-01 -4.94925410e-01
7.82473147e-01 -3.55543762e-01 -1.21673608e+00 2.25193515e-01
-1.02311611e+00 -3.77431005e-01 -1.49880201e-01 2.55358040e-01
-8.87813509e-01 1.08662277e-01 -6.91586673e-01 -6.82822585e-01
-4.51400429e-02 -1.31071115e+00 1.72199011e-01 3.24113101e-01
2.21658051e-01 -6.41696990e-01 5.05999997e-02 2.12343074e-02
6.72214329e-01 2.10459784e-01 1.53483927e+00 -3.03438157e-01
-1.16702628e+00 4.21393663e-01 -3.31956625e-01 2.25988433e-01
1.57353029e-01 4.05782193e-01 1.51639078e-02 -3.96719038e-01
-2.67506897e-01 -6.60066158e-02 7.37091526e-02 3.87306839e-01
9.71675217e-01 -5.57682455e-01 -3.03361356e-01 5.63852191e-01
1.63550186e+00 8.29169929e-01 5.62368929e-01 5.95289409e-01
4.65622962e-01 6.71821415e-01 4.23863739e-01 7.73592830e-01
3.88394028e-01 4.43358570e-01 1.20010734e+00 3.46273422e-01
1.51951373e-01 -9.34270471e-02 5.00921726e-01 5.85672259e-01
1.12132452e-01 -3.29701424e-01 -5.32362401e-01 6.38752654e-02
-1.86738873e+00 -1.03741336e+00 2.47101292e-01 2.15708375e+00
3.65027398e-01 4.93292153e-01 2.87991345e-01 3.86748016e-02
9.26004291e-01 1.14737727e-01 -1.03191137e+00 -6.81171477e-01
3.04894775e-01 -2.25875497e-01 3.74959320e-01 3.06451857e-01
-4.98518735e-01 8.78429711e-01 5.97154474e+00 3.67994726e-01
-1.06165826e+00 -4.08822447e-01 2.18380809e-01 2.36923322e-01
-7.26549402e-02 5.21065295e-01 -5.11591077e-01 6.54098153e-01
8.42959940e-01 -1.03492290e-01 9.97511864e-01 6.90503895e-01
7.72383153e-01 7.50970989e-02 -9.43132818e-01 7.25207984e-01
-8.83391798e-01 -1.35991037e+00 1.67735621e-01 1.23369530e-01
5.38979888e-01 -1.38172567e-01 -4.15308475e-01 4.13215280e-01
4.78235453e-01 -5.57905078e-01 2.76372552e-01 1.95722669e-01
4.26511019e-01 -8.79458845e-01 3.58681917e-01 2.19814256e-01
-1.15738392e+00 -7.80932188e-01 1.79079780e-03 -3.42781097e-02
2.07519427e-01 3.28863204e-01 -3.99781257e-01 3.16276968e-01
4.90807772e-01 5.92599392e-01 -4.77271080e-02 1.12473845e+00
-7.14892894e-02 3.01088393e-01 -9.04770344e-02 -1.76645126e-02
3.70766610e-01 -5.40934145e-01 9.00712550e-01 3.47265452e-01
1.30871370e-01 -2.61787022e-03 6.18178189e-01 1.00466061e+00
-1.40014097e-01 1.11447439e-01 -1.93144113e-01 3.14077176e-03
8.53514612e-01 1.17616701e+00 -6.96153045e-01 -8.94642696e-02
-2.45463923e-01 4.36729759e-01 3.35213602e-01 6.95157945e-01
-9.34424460e-01 -6.73097551e-01 1.22096443e+00 2.03931883e-01
4.28521007e-01 -1.33971974e-01 -1.68011114e-01 -5.25685370e-01
-3.14176649e-01 -9.49957728e-01 3.67085904e-01 -4.78917867e-01
-1.05276787e+00 4.19972897e-01 -3.26295108e-01 -1.26512361e+00
-1.00134090e-01 -1.56225979e-01 -8.84185791e-01 3.62690955e-01
-1.40632570e+00 -3.93485397e-01 -2.36709595e-01 6.34777308e-01
3.90891790e-01 -4.06336695e-01 7.39854157e-01 3.29989225e-01
-1.05250776e+00 2.74302870e-01 7.34267086e-02 -2.80469302e-02
1.81213990e-01 -8.26219261e-01 -4.28523183e-01 5.90121865e-01
-7.30683327e-01 7.45839253e-02 5.51162779e-01 -4.70275968e-01
-1.40783370e+00 -8.10819149e-01 2.68172115e-01 5.30480504e-01
8.89057159e-01 8.19184631e-02 -5.10705173e-01 4.39921647e-01
-5.03688678e-02 -1.65168330e-01 2.59341568e-01 -8.74738172e-02
7.61339739e-02 -5.64552367e-01 -1.37802947e+00 8.47716331e-01
1.00251424e+00 2.26190016e-01 2.11971521e-01 9.51326862e-02
9.42320764e-01 9.83516723e-02 -9.14659262e-01 3.45183194e-01
3.05439264e-01 -9.56333339e-01 5.92903793e-01 -8.49907219e-01
2.85375118e-01 -4.52223003e-01 4.12971079e-02 -1.34061313e+00
-7.03474343e-01 -8.47649693e-01 -1.20460443e-01 1.27823949e+00
3.28511268e-01 -8.51584733e-01 9.01786566e-01 8.16341937e-02
-9.48684812e-02 -9.01329219e-01 -1.27165484e+00 -8.40721607e-01
-2.88412452e-01 7.16352835e-02 8.83983612e-01 7.12014258e-01
-9.60032865e-02 6.73868537e-01 -5.53861335e-02 3.71177912e-01
8.74998689e-01 3.13360900e-01 7.46709406e-01 -1.44898081e+00
-2.46617034e-01 -8.04614604e-01 -2.21140876e-01 -9.61381674e-01
2.44545922e-01 -6.00125790e-01 -3.51298675e-02 -1.61719668e+00
-3.67650628e-01 -1.19060171e+00 -4.58752185e-01 2.72251457e-01
4.43535894e-01 -8.46133232e-01 4.02252018e-01 1.14091009e-01
-7.03498244e-01 5.98070264e-01 1.78831625e+00 1.74240544e-01
-6.15304649e-01 3.33379537e-01 -4.84898806e-01 5.46610117e-01
1.18925691e+00 -5.56759000e-01 -7.02097178e-01 -3.03420901e-01
-9.43091791e-03 9.58748817e-01 1.20157860e-01 -9.41471279e-01
5.89519478e-02 -8.46898794e-01 -1.54229015e-01 -2.70762861e-01
-2.01294199e-02 -1.19479012e+00 1.25773624e-01 9.66550469e-01
-3.56413931e-01 7.85765573e-02 -2.84831583e-01 7.63855517e-01
4.21024673e-02 -8.61339346e-02 1.15729380e+00 -1.03819892e-01
-7.19778538e-01 6.57663584e-01 -8.39064062e-01 1.93981186e-01
1.37574303e+00 -3.58310610e-01 -1.35186523e-01 -3.75320941e-01
-9.11067486e-01 1.02819848e+00 2.20971197e-01 4.82974499e-01
5.28583825e-01 -1.12047410e+00 -4.87962186e-01 1.43853873e-01
-3.32218230e-01 -2.33339652e-01 1.51303977e-01 7.48703957e-01
-5.87490678e-01 7.51936957e-02 -7.67553508e-01 -2.88443327e-01
-7.61841953e-01 6.19988501e-01 6.94480062e-01 -4.30548877e-01
-3.86789590e-01 3.62934113e-01 -2.11318478e-01 -3.08847070e-01
4.30069417e-01 -1.01189785e-01 -3.09632480e-01 -5.76924868e-02
-3.85800228e-02 8.22505474e-01 -5.52920222e-01 -1.78792626e-01
-4.08362091e-01 4.53956842e-01 7.49858543e-02 8.11307803e-02
1.29061198e+00 -6.25175595e-01 -5.16684234e-01 -9.08840746e-02
9.84228849e-01 -5.06142557e-01 -1.12269986e+00 2.72641648e-02
1.63873240e-01 -3.13698560e-01 1.82616293e-01 -8.01925898e-01
-1.45194793e+00 2.42979482e-01 6.22495174e-01 6.22099221e-01
1.30768800e+00 -4.05278713e-01 9.64137137e-01 2.59174466e-01
6.62967682e-01 -1.29142118e+00 3.17385316e-01 7.03004122e-01
3.84264320e-01 -9.41219568e-01 -2.91696399e-01 -5.98633170e-01
-3.48380297e-01 1.28927898e+00 1.30643725e+00 -4.42959219e-02
8.69113863e-01 6.23111911e-02 9.11099091e-02 -1.60898522e-01
-1.03830850e+00 -9.62850079e-02 -7.99842060e-01 6.09029472e-01
-2.36684382e-01 -6.22555893e-03 -2.79335082e-01 7.91708753e-02
2.57655114e-01 3.55365761e-02 7.77607143e-01 7.64869153e-01
-1.00605476e+00 -1.41699350e+00 -1.75713360e-01 4.53863651e-01
6.06477149e-02 3.92161489e-01 1.31695881e-01 5.49580932e-01
2.81132221e-01 1.04773378e+00 9.07100067e-02 -3.37241352e-01
2.91925400e-01 -1.68722063e-01 1.22180387e-01 -5.86717129e-01
-3.85162979e-01 -1.88328058e-01 1.80861354e-02 -7.26195753e-01
-2.48382673e-01 -2.51520038e-01 -1.78806710e+00 -5.52305520e-01
-1.32895052e-01 2.66735345e-01 5.76495528e-01 7.58287489e-01
5.96706927e-01 6.82990432e-01 1.39979601e+00 -5.27222812e-01
-6.52459383e-01 -4.59078878e-01 -6.67359591e-01 -1.22903042e-01
5.09745181e-01 -7.76200652e-01 -4.86938179e-01 -5.14738858e-01] | [5.762739181518555, 1.7593961954116821] |
344585e7-27a8-427a-b0b4-5c8c1928d2d0 | dex-nerf-using-a-neural-radiance-field-to | 2110.14217 | null | https://arxiv.org/abs/2110.14217v1 | https://arxiv.org/pdf/2110.14217v1.pdf | Dex-NeRF: Using a Neural Radiance Field to Grasp Transparent Objects | The ability to grasp and manipulate transparent objects is a major challenge for robots. Existing depth cameras have difficulty detecting, localizing, and inferring the geometry of such objects. We propose using neural radiance fields (NeRF) to detect, localize, and infer the geometry of transparent objects with sufficient accuracy to find and grasp them securely. We leverage NeRF's view-independent learned density, place lights to increase specular reflections, and perform a transparency-aware depth-rendering that we feed into the Dex-Net grasp planner. We show how additional lights create specular reflections that improve the quality of the depth map, and test a setup for a robot workcell equipped with an array of cameras to perform transparent object manipulation. We also create synthetic and real datasets of transparent objects in real-world settings, including singulated objects, cluttered tables, and the top rack of a dishwasher. In each setting we show that NeRF and Dex-Net are able to reliably compute robust grasps on transparent objects, achieving 90% and 100% grasp success rates in physical experiments on an ABB YuMi, on objects where baseline methods fail. | ['Ken Goldberg', 'Justin Kerr', 'Yahav Avigal', 'Jeffrey Ichnowski'] | 2021-10-27 | null | null | null | null | ['transparent-objects'] | ['computer-vision'] | [ 1.76546406e-02 2.51645386e-01 6.31003082e-01 -9.76568907e-02
-3.69574100e-01 -9.59731877e-01 -2.22917497e-02 -2.69147325e-02
-3.73928919e-02 3.12754571e-01 -1.11197501e-01 3.08086369e-02
7.81078711e-02 -7.08390713e-01 -1.29222119e+00 -5.48824966e-01
-4.51048821e-01 8.45242620e-01 3.54363352e-01 -4.57748882e-02
3.11951607e-01 7.19061971e-01 -1.44661224e+00 4.58889157e-01
4.78695393e-01 1.03118670e+00 7.12223709e-01 9.18680489e-01
5.12526512e-01 6.74598396e-01 -4.19309884e-01 -1.12264723e-01
9.05563533e-01 5.95411003e-01 -5.38888633e-01 5.19871451e-02
6.69459581e-01 -1.18602157e+00 -3.58357966e-01 6.38155460e-01
7.32997991e-03 4.33467068e-02 6.87939644e-01 -1.07314789e+00
-6.91590428e-01 3.85030001e-01 -7.23660767e-01 -3.91426891e-01
7.19734430e-01 5.16264796e-01 4.78186935e-01 -5.68361104e-01
6.32284641e-01 1.64222157e+00 4.69767898e-01 6.20486915e-01
-9.25474286e-01 -2.64776260e-01 1.90820843e-01 -4.69664335e-01
-8.49347234e-01 -2.49200389e-01 3.45102996e-01 -4.18165028e-01
9.93147075e-01 5.75496666e-02 5.58434606e-01 1.20973790e+00
5.14887333e-01 5.10131478e-01 9.03637528e-01 -2.16014579e-01
5.27201951e-01 -9.92125943e-02 -1.37100175e-01 1.08596277e+00
5.35238326e-01 1.04725391e-01 -6.63900673e-01 -2.51381129e-01
1.40796864e+00 1.62621364e-01 -4.63755846e-01 -9.52627242e-01
-1.44139969e+00 9.00722072e-02 9.19024289e-01 -5.48196495e-01
-3.79447550e-01 4.45575804e-01 -1.98566124e-01 -7.12394565e-02
1.21646963e-01 7.09293008e-01 -4.91082609e-01 -4.70714979e-02
3.25871527e-01 3.04783434e-01 1.15213370e+00 1.49181426e+00
6.02592707e-01 -3.39869291e-01 2.95483053e-01 2.42109731e-01
1.77003250e-01 9.24089432e-01 -4.75267380e-01 -1.59797144e+00
7.36495912e-01 3.26621741e-01 8.86594534e-01 -9.05801415e-01
-5.01273632e-01 4.15951699e-01 -7.94975311e-02 7.81227529e-01
5.79492688e-01 -2.10102245e-01 -1.07514739e+00 1.18784344e+00
4.51886028e-01 -3.06442529e-01 1.41566962e-01 1.30927944e+00
4.69232649e-01 4.56476092e-01 -5.96397817e-01 5.32237709e-01
1.13335836e+00 -8.50159645e-01 -1.64392322e-01 -4.33005154e-01
3.51442993e-01 -5.82302809e-01 1.13498676e+00 8.70718896e-01
-1.08465803e+00 1.06913358e-01 -9.96283948e-01 -2.96785146e-01
-4.40875292e-02 1.97831735e-01 1.19752932e+00 2.30236173e-01
-8.12722623e-01 5.85816622e-01 -1.41601574e+00 -2.03275070e-01
4.08855528e-01 4.89024401e-01 -6.63688719e-01 -5.62928140e-01
-1.15482397e-02 9.52922463e-01 6.26383498e-02 3.55750352e-01
-1.44853556e+00 -6.24230385e-01 -7.43132472e-01 -8.79292414e-02
5.32820880e-01 -6.66569293e-01 1.16673660e+00 -2.46490479e-01
-1.57536209e+00 5.90725362e-01 2.16301948e-01 -5.43663949e-02
3.96368921e-01 -8.74073207e-01 7.06916273e-01 6.90073550e-01
-1.61404759e-01 7.97272444e-01 5.85310340e-01 -2.02099848e+00
-9.71984491e-02 -7.03490734e-01 7.08300889e-01 3.89637917e-01
7.95984715e-02 -3.71980637e-01 1.03489354e-01 2.02574749e-02
6.33036852e-01 -9.65118647e-01 3.67561392e-02 7.24700928e-01
-5.85195005e-01 2.53746778e-01 9.60571945e-01 -4.67250049e-01
-5.06591558e-01 -1.97149765e+00 1.70159400e-01 1.17037617e-01
2.18085855e-01 -4.24572051e-01 -1.54725492e-01 4.24757212e-01
5.57639897e-01 -3.96010011e-01 -2.05233861e-02 -3.94294083e-01
1.06464908e-01 2.52327383e-01 -5.43693721e-01 7.35669792e-01
-1.13597117e-01 5.39898813e-01 -9.05991554e-01 2.04707399e-01
2.24827826e-01 5.37408173e-01 -6.39185309e-01 3.26063156e-01
-6.41965747e-01 4.44832414e-01 -5.45266986e-01 1.08978724e+00
9.50233579e-01 8.92288610e-02 1.44003227e-01 -2.45061070e-01
-7.68021345e-02 1.97257489e-01 -1.03329349e+00 1.89341164e+00
-5.19244134e-01 3.54600251e-01 9.42159772e-01 -1.49712786e-01
9.68929768e-01 -1.39220849e-01 2.59545177e-01 -3.06862026e-01
6.41762391e-02 2.75104761e-01 -4.92856950e-01 -8.22807610e-01
4.74698275e-01 1.79574251e-01 1.53752312e-01 4.50732678e-01
-2.30034232e-01 -8.81406724e-01 -3.83067518e-01 1.96989909e-01
1.52016771e+00 6.53736174e-01 -7.37755179e-01 -2.59027541e-01
-7.10186243e-01 -4.54600118e-02 7.65814036e-02 5.44066548e-01
4.27428991e-01 8.83366704e-01 2.26795346e-01 -5.43467045e-01
-9.33129966e-01 -1.53472507e+00 2.31656849e-01 8.83766115e-01
9.48031604e-01 9.35721993e-02 -7.02665269e-01 -3.90869558e-01
5.58238626e-01 5.11960089e-01 -4.80507702e-01 -9.31073874e-02
-5.93240023e-01 -2.69998282e-01 -8.82662535e-02 5.96366405e-01
3.38087022e-01 -8.99360597e-01 -1.49113977e+00 -1.63387254e-01
-1.14146121e-01 -1.30922818e+00 1.47258723e-02 4.71904278e-01
-7.72609353e-01 -1.36219990e+00 -5.70459485e-01 -5.71748972e-01
1.02824652e+00 7.84596562e-01 9.63591874e-01 -2.10751384e-01
-6.19954646e-01 8.14742386e-01 -3.05323303e-01 -5.57238877e-01
-2.19278540e-02 -3.47664684e-01 2.64659733e-01 -6.30352736e-01
-9.93784666e-02 -6.23280644e-01 -7.50279665e-01 3.49096358e-01
-5.23861170e-01 -1.13089368e-01 3.93849283e-01 8.70842785e-02
4.20071892e-02 -2.57948816e-01 -4.42243427e-01 -3.10510630e-03
9.37697366e-02 -3.09064895e-01 -1.08660984e+00 4.02781554e-02
3.44691187e-01 -5.88018000e-02 1.04607217e-01 -3.82042080e-01
-9.44371283e-01 1.74257383e-01 7.41444707e-01 -7.21651077e-01
-7.73240477e-02 -3.41779262e-01 -6.78936169e-02 -4.90277141e-01
7.45692790e-01 -5.69373846e-01 -2.88474232e-01 -4.41833138e-01
1.28096282e-01 2.87381679e-01 6.68205738e-01 -1.15695381e+00
7.16457784e-01 1.04994678e+00 1.71926409e-01 -6.48260295e-01
-4.96213496e-01 -9.42700133e-02 -6.65364802e-01 -1.33120179e-01
7.43952036e-01 -8.49361181e-01 -1.68177271e+00 5.10271907e-01
-1.50113201e+00 -9.78854120e-01 -6.64283633e-02 6.90054774e-01
-8.12421560e-01 2.95753274e-02 -8.29376221e-01 -1.04402936e+00
5.48445061e-02 -1.22537458e+00 1.68848848e+00 2.15226695e-01
2.31001407e-01 -3.93584192e-01 -3.33786994e-01 4.22051221e-01
2.65421718e-01 3.88514847e-01 7.64428437e-01 3.30402106e-01
-1.53771484e+00 2.03455418e-01 -3.45953047e-01 -1.69803500e-01
3.08477551e-01 -2.89491192e-02 -1.25141752e+00 -4.24270213e-01
7.61844777e-03 -8.62710893e-01 8.07164669e-01 3.12234581e-01
1.08712208e+00 -3.18949699e-01 -6.31433725e-01 6.80554986e-01
1.28111768e+00 2.63220035e-02 4.62268442e-01 2.41507858e-01
8.30462039e-01 7.81632662e-01 7.36074924e-01 4.31435049e-01
4.31311756e-01 5.89019001e-01 1.30757344e+00 2.41358221e-01
5.34126908e-02 -1.52374115e-02 4.64203447e-01 8.33956897e-02
-4.06282216e-01 -5.51047087e-01 -8.95263553e-01 2.22794592e-01
-1.39078844e+00 -2.75001675e-01 8.75581205e-02 2.23552752e+00
4.29070741e-01 4.16640304e-02 -2.62420863e-01 -4.97579396e-01
2.54837513e-01 -4.11485970e-01 -6.65736496e-01 -2.64414549e-01
2.00959697e-01 -6.55361414e-02 8.46682727e-01 6.97430551e-01
-8.08760226e-01 8.17902863e-01 5.95690823e+00 -3.33488494e-01
-9.16940033e-01 -2.98772722e-01 1.60635449e-02 -3.97825450e-01
-3.95144016e-01 -1.53486773e-01 -4.61692721e-01 -8.78802761e-02
2.19524309e-01 7.99648523e-01 9.73985076e-01 9.00900662e-01
-2.59175658e-01 -8.60957861e-01 -1.54765272e+00 7.92025089e-01
8.91225263e-02 -6.67337954e-01 -2.75643706e-01 5.39722592e-02
5.00475764e-01 2.66795725e-01 2.11607561e-01 -2.09111974e-01
8.73151243e-01 -9.20576394e-01 9.20087278e-01 4.98753488e-01
4.20154035e-01 -2.69983977e-01 3.93235564e-01 4.33322757e-01
-6.98810518e-01 -3.34455818e-01 -6.81527972e-01 -7.01649636e-02
-5.07036634e-02 5.56858599e-01 -1.14087141e+00 2.58306414e-01
1.14673352e+00 1.38997063e-01 1.55728847e-01 8.94166946e-01
-1.52325720e-01 2.11609881e-02 -9.20085311e-01 5.11126034e-02
2.73851603e-02 -1.15798645e-01 5.80764174e-01 5.43068051e-01
1.93435848e-01 4.24385160e-01 2.29832560e-01 1.32026982e+00
-9.41552743e-02 -8.29299629e-01 -7.35795498e-01 3.24579060e-01
1.71306267e-01 1.10156882e+00 -9.25487995e-01 3.46776009e-01
1.70297742e-01 1.13345110e+00 4.31732386e-01 5.56017578e-01
-4.62264240e-01 -2.62590855e-01 6.80621803e-01 3.10800225e-02
2.84018070e-01 -8.05822253e-01 -2.18935788e-01 -1.18350863e+00
7.24547029e-01 -3.98771405e-01 -5.45549512e-01 -1.52813113e+00
-1.02992487e+00 1.19236290e-01 1.64922863e-01 -7.89429367e-01
3.13438207e-01 -1.18209279e+00 -1.69485889e-03 6.23327911e-01
-1.24732411e+00 -1.10209930e+00 -9.40459430e-01 1.90954477e-01
3.81268322e-01 4.65522140e-01 7.91419268e-01 -4.42200154e-01
1.17923304e-01 -1.48142427e-01 -2.67178025e-02 -1.20559685e-01
6.15168452e-01 -1.25163043e+00 3.11504513e-01 3.21924418e-01
-2.72177190e-01 6.83799148e-01 6.22326910e-01 -8.80431712e-01
-2.33808422e+00 -6.10712528e-01 -5.08800566e-01 -9.69187200e-01
2.43967667e-01 -1.03180504e+00 -7.65345991e-01 1.12826049e+00
-1.67970777e-01 1.72409907e-01 -2.81341344e-01 -1.40089273e-01
-6.68712080e-01 -7.37753883e-02 -1.70459688e+00 4.62485343e-01
1.33734107e+00 -2.50111759e-01 -5.35810530e-01 8.25630784e-01
7.88547695e-01 -1.13862431e+00 -5.99501371e-01 4.95034158e-01
9.70222712e-01 -1.16631448e+00 1.06406748e+00 -1.07877627e-01
5.58830321e-01 -1.17660008e-01 -4.53248203e-01 -1.39111459e+00
1.68863699e-01 -7.26759255e-01 -1.94428548e-01 5.98702848e-01
-3.64132263e-02 -7.20392227e-01 9.56304848e-01 9.63722348e-01
-3.87015522e-01 -4.32582319e-01 -6.64662480e-01 -6.98692083e-01
-1.99052125e-01 7.02193454e-02 4.66197848e-01 4.89969641e-01
9.27796736e-02 -3.08175087e-01 3.40024203e-01 9.40667331e-01
9.75706398e-01 3.16711217e-01 8.28023076e-01 -1.08023775e+00
-2.56011575e-01 1.83713004e-01 -3.12780619e-01 -1.25775433e+00
1.92722194e-02 -4.21878785e-01 6.98054552e-01 -1.78860176e+00
3.06397408e-01 -1.02236760e+00 5.73116541e-01 5.52785873e-01
4.19771373e-01 -2.00007603e-01 2.07857847e-01 1.21155836e-01
-3.77978921e-01 4.41163033e-01 1.44153214e+00 -9.40274298e-02
-4.02848683e-02 -1.78425640e-01 -4.49406132e-02 9.20548975e-01
5.61425507e-01 -2.44851053e-01 -3.43463480e-01 -1.35555601e+00
2.31593758e-01 4.78983372e-02 8.60096633e-01 -1.21716523e+00
-1.15806527e-01 -3.08134526e-01 5.78824282e-01 -3.76850247e-01
1.06043983e+00 -1.17086673e+00 -3.80920678e-01 4.05009359e-01
-2.27393270e-01 -1.38295023e-02 4.09853101e-01 5.63671052e-01
7.94623494e-01 -1.14117831e-01 3.96263063e-01 -6.04215801e-01
-2.22684830e-01 9.22179222e-02 -5.06520532e-02 -3.35595638e-01
1.07272112e+00 -2.10065335e-01 -8.95302415e-01 -1.78223908e-01
-3.31222594e-01 2.58011460e-01 1.13005412e+00 3.79545808e-01
9.01560783e-01 -5.95253706e-01 -2.71297634e-01 2.61194080e-01
-6.42731786e-02 8.65402520e-01 -1.11055404e-01 2.84663022e-01
-1.25994194e+00 9.66485068e-02 -1.72411427e-01 -8.82502496e-01
-9.29903150e-01 5.45851707e-01 3.73942703e-01 5.92862368e-01
-9.13636148e-01 1.18999457e+00 6.07821524e-01 -7.17213035e-01
4.51379806e-01 -1.01378036e+00 7.25568593e-01 -1.04019678e+00
3.27403069e-01 5.49740076e-01 9.38310400e-02 2.39604235e-01
-1.63322315e-01 6.60917640e-01 1.04707547e-01 -1.34375423e-01
1.47700131e+00 -2.77968366e-02 -2.37434566e-01 2.56053716e-01
6.64768219e-01 2.13860512e-01 -1.86986995e+00 3.14247519e-01
-6.09094262e-01 -6.67711139e-01 -2.32643336e-01 -9.30621862e-01
-9.11896348e-01 9.76941288e-01 3.44463766e-01 -8.17635357e-02
5.41166186e-01 3.48400801e-01 6.24003828e-01 1.19990468e+00
1.37870932e+00 -6.56019568e-01 4.76692230e-01 5.23148119e-01
1.09661067e+00 -9.40537214e-01 -8.04591272e-03 -9.93969262e-01
-2.50879109e-01 1.38619447e+00 1.04938364e+00 -3.35400432e-01
1.58434376e-01 9.81208742e-01 1.98759250e-02 -5.10248065e-01
-4.98391211e-01 5.30584157e-01 -3.62505615e-01 1.03751123e+00
-4.63971376e-01 5.22409938e-02 1.32608438e+00 -2.29600802e-01
-2.73710698e-01 -2.01206550e-01 6.98637843e-01 1.48524368e+00
-7.60788798e-01 -8.50160122e-02 -6.64969623e-01 8.10800716e-02
7.96409149e-04 2.70341814e-01 -3.76348704e-01 5.07718682e-01
-2.51203477e-01 8.04087698e-01 2.91240901e-01 -2.47950077e-01
3.39800149e-01 -6.03202224e-01 1.42098570e+00 -7.96255648e-01
-2.00652570e-01 -4.23265010e-01 -1.53673505e-02 -1.11427748e+00
-2.17973888e-01 -3.51628393e-01 -1.58213818e+00 -8.45712796e-02
-3.10508162e-01 -3.02920431e-01 1.04005504e+00 7.27865517e-01
3.18582267e-01 9.87344012e-02 5.36527693e-01 -1.62884438e+00
-5.97533882e-01 -7.92634547e-01 -5.76935530e-01 -3.88851273e-03
6.10032558e-01 -9.06652749e-01 -5.10349274e-01 -2.14316010e-01] | [5.926169395446777, -1.0256898403167725] |
560c5c98-ebe8-4689-a33b-5946bb8e0b0c | simultaneous-contact-rich-grasping-and | 2207.01418 | null | https://arxiv.org/abs/2207.01418v2 | https://arxiv.org/pdf/2207.01418v2.pdf | Simultaneous Contact-Rich Grasping and Locomotion via Distributed Optimization Enabling Free-Climbing for Multi-Limbed Robots | While motion planning of locomotion for legged robots has shown great success, motion planning for legged robots with dexterous multi-finger grasping is not mature yet. We present an efficient motion planning framework for simultaneously solving locomotion (e.g., centroidal dynamics), grasping (e.g., patch contact), and contact (e.g., gait) problems. To accelerate the planning process, we propose distributed optimization frameworks based on Alternating Direction Methods of Multipliers (ADMM) to solve the original large-scale Mixed-Integer NonLinear Programming (MINLP). The resulting frameworks use Mixed-Integer Quadratic Programming (MIQP) to solve contact and NonLinear Programming (NLP) to solve nonlinear dynamics, which are more computationally tractable and less sensitive to parameters. Also, we explicitly enforce patch contact constraints from limit surfaces with micro-spine grippers. We demonstrate our proposed framework in the hardware experiments, showing that the multi-limbed robot is able to realize various motions including free-climbing at a slope angle 45{\deg} with a much shorter planning time. | ['Dennis Hong', 'Varit Vichathorn', 'Hayato Kato', 'Yusuke Tanaka', 'Alexander Schperberg', 'Xuan Lin', 'Yuki Shirai'] | 2022-07-04 | null | null | null | null | ['distributed-optimization'] | ['methodology'] | [-6.31893873e-02 1.89602181e-01 -4.06344682e-01 2.94194324e-03
-3.05129379e-01 -5.91687560e-01 -6.33092895e-02 -3.78597051e-01
-4.26799744e-01 8.90204132e-01 -3.99805784e-01 -2.19614562e-02
-6.08204663e-01 -7.97738910e-01 -9.87806499e-01 -8.41196358e-01
-5.94155252e-01 7.87755609e-01 1.14610769e-01 -5.64710796e-01
5.93898714e-01 3.22247386e-01 -1.43464577e+00 -4.33826633e-02
1.19025266e+00 6.16151273e-01 1.09880626e+00 5.81674159e-01
4.00465310e-01 3.44196260e-01 8.48018900e-02 1.39507890e-01
2.75810242e-01 1.93332314e-01 -8.33170295e-01 1.61668271e-01
-1.83085740e-01 -5.58457136e-01 -4.95666713e-02 7.89101779e-01
3.84525955e-01 2.20741898e-01 6.47400022e-01 -1.72653532e+00
-2.98667192e-01 4.09151554e-01 -8.32492828e-01 -7.41154730e-01
1.27455965e-01 2.45505899e-01 7.32801795e-01 -7.92681813e-01
9.37095940e-01 1.49103427e+00 4.42075372e-01 5.60979724e-01
-7.11535096e-01 6.67231977e-02 2.15718579e-02 1.62210688e-01
-8.34531963e-01 3.55234444e-02 5.40942192e-01 -2.99667716e-01
9.59653139e-01 3.09428554e-02 6.36439323e-01 9.17004824e-01
8.19633186e-01 7.05953717e-01 4.89042789e-01 -6.07212894e-02
3.41481954e-01 -6.97551727e-01 -1.02883004e-01 7.91857779e-01
8.08051884e-01 -4.88903105e-01 -3.40857267e-01 -1.88227504e-01
1.26717293e+00 -5.76934516e-02 4.06454094e-02 -7.06816673e-01
-1.61019242e+00 7.28778243e-01 3.26744884e-01 -2.78670639e-01
-6.88079119e-01 6.46863341e-01 1.96474567e-01 -7.34536871e-02
-2.96594381e-01 3.69615078e-01 -6.58223152e-01 -2.73723364e-01
-1.80846497e-01 1.03967655e+00 1.21882904e+00 1.42436111e+00
5.01488328e-01 -1.05897158e-01 1.80134892e-01 8.29408586e-01
5.72523892e-01 7.09938467e-01 -1.22905061e-01 -1.65955496e+00
1.08733869e+00 5.18935561e-01 7.29037523e-01 -1.21418023e+00
-8.54383409e-01 3.12028021e-01 -8.23146224e-01 4.84837025e-01
4.89486337e-01 -4.78667110e-01 -6.40169203e-01 1.46916389e+00
3.27605784e-01 -6.48854971e-01 9.92316753e-02 1.12253010e+00
8.41688365e-02 7.33882427e-01 -2.17121020e-01 -2.65884012e-01
1.06681347e+00 -1.21232891e+00 -5.33586800e-01 -3.74475300e-01
4.58160490e-01 -3.39836299e-01 7.63617754e-01 5.37518561e-01
-1.25852466e+00 -6.72712103e-02 -1.06979430e+00 -3.16920638e-01
1.26818225e-01 3.98794383e-01 6.25918150e-01 -1.27467066e-01
-7.36520827e-01 1.02528584e+00 -1.23792791e+00 -3.91027629e-01
-1.49745628e-01 6.85193121e-01 -2.29978383e-01 -5.54062426e-02
-6.98233187e-01 1.05718970e+00 3.20615053e-01 4.37157482e-01
-6.81169868e-01 -2.21634001e-01 -4.27833319e-01 -3.59305978e-01
5.74938655e-01 -9.14245903e-01 1.09941089e+00 5.67936450e-02
-1.91260469e+00 6.15917683e-01 5.38240857e-02 -1.17226318e-01
8.18656921e-01 -4.87162113e-01 7.47885525e-01 2.69110978e-01
3.15206140e-01 8.26193869e-01 6.81351662e-01 -1.25439107e+00
-3.46796066e-01 -3.07853311e-01 1.62103400e-01 6.07280791e-01
-1.86490670e-01 -4.73918855e-01 -3.05175096e-01 -4.64162529e-01
5.21932781e-01 -1.37147689e+00 -6.32619977e-01 3.43096912e-01
-4.89875823e-01 -3.02066743e-01 8.46822977e-01 -7.77389944e-01
7.95247138e-01 -1.52870119e+00 1.03034902e+00 2.95239519e-02
-4.02456731e-01 -3.23887467e-01 -4.95149903e-02 8.22597086e-01
8.41158628e-01 -1.61426648e-01 -5.79812765e-01 4.55835927e-03
1.42657712e-01 8.11231434e-01 -5.87712154e-02 3.19261134e-01
8.32484439e-02 8.33627760e-01 -1.11270118e+00 -4.57163095e-01
-3.03661913e-01 -1.43051803e-01 -9.60351825e-01 6.83501661e-02
-5.32801270e-01 3.34600896e-01 -9.31813836e-01 1.13412893e+00
7.67300487e-01 9.22948718e-02 4.68539387e-01 -2.71350861e-01
-8.42221618e-01 -4.74787146e-01 -1.33803153e+00 2.04980183e+00
-3.39905888e-01 -1.04413442e-01 9.26100135e-01 -9.70040083e-01
9.48123157e-01 -2.62805969e-02 8.97237301e-01 -1.69908509e-01
-1.74898673e-02 6.93556309e-01 -1.91067517e-01 -1.04906535e+00
7.12955534e-01 4.31005329e-01 -1.08896159e-01 2.11888567e-01
-3.02535415e-01 -6.06542766e-01 2.22121775e-01 -3.72552216e-01
1.13886166e+00 7.48861253e-01 1.81214660e-02 -7.45895803e-01
2.40425035e-01 7.68109620e-01 7.56449342e-01 3.99153560e-01
2.33597875e-01 2.77276188e-01 4.13631290e-01 -1.99497774e-01
-1.52615249e+00 -9.88806903e-01 1.71140566e-01 8.16830993e-01
8.53656709e-01 8.60313848e-02 -7.19969094e-01 3.63483012e-01
7.23642111e-01 1.55538768e-02 1.11065365e-01 3.67996901e-01
-1.32571673e+00 -9.34359372e-01 9.06075388e-02 5.97030461e-01
5.42227268e-01 -1.12409723e+00 -1.19766676e+00 8.15375030e-01
-4.67548102e-01 -1.10729563e+00 3.48272957e-02 1.14927329e-01
-1.19046354e+00 -1.04022145e+00 -1.13872588e+00 -1.20728505e+00
6.45421386e-01 2.97599733e-01 5.23974180e-01 7.05792010e-02
-5.55644691e-01 2.12328687e-01 -4.69572246e-01 8.27687457e-02
1.33319438e-01 -1.77427102e-02 4.90447730e-01 -7.70308912e-01
-6.65165722e-01 -5.93662977e-01 -5.63433886e-01 8.87433112e-01
-5.38199961e-01 2.84570396e-01 8.27164650e-01 9.42616105e-01
8.63225460e-01 -6.02088347e-02 4.50862736e-01 1.86426006e-02
5.88841736e-01 -6.87769055e-01 -5.68264484e-01 2.32652768e-01
4.84523773e-02 -1.07379101e-01 5.06121218e-01 -5.58089852e-01
-9.25679147e-01 3.69452327e-01 2.17327967e-01 -1.32980421e-01
5.17724693e-01 4.98169243e-01 -1.34756854e-02 -4.29742783e-01
1.58723384e-01 9.23454985e-02 2.11027697e-01 -3.91705900e-01
4.23093140e-01 5.88110626e-01 6.08367920e-01 -1.09339571e+00
5.35296381e-01 5.96828461e-01 5.33697724e-01 -9.66166914e-01
2.70076096e-02 -1.22097507e-01 -8.29823017e-01 -3.02798033e-01
1.15184689e+00 -6.18972778e-01 -1.26628757e+00 9.24389660e-01
-1.40439200e+00 -7.46879578e-01 3.48513633e-01 3.31187099e-01
-1.41547108e+00 6.43838286e-01 -7.11536407e-01 -8.64437163e-01
-4.70963925e-01 -1.03006721e+00 1.05516458e+00 1.73930526e-01
2.36778911e-02 -3.32507879e-01 -1.71191096e-01 1.62264630e-01
2.53983647e-01 9.54351187e-01 8.48352313e-01 5.95009565e-01
-8.75116885e-01 6.14205115e-02 -2.80077737e-02 -2.33888909e-01
-2.84316652e-02 1.64142251e-01 1.09092467e-01 -5.79327822e-01
-1.72794238e-01 -4.08237666e-01 2.14685887e-01 5.45020938e-01
7.23583400e-01 -7.04232991e-01 -5.88772416e-01 4.92434770e-01
1.51328850e+00 2.42931277e-01 5.06659746e-01 7.38541782e-01
6.83356106e-01 8.41563761e-01 1.39544225e+00 8.44270587e-01
5.66342175e-01 9.03423369e-01 9.47751939e-01 5.17683804e-01
5.65920055e-01 9.42384228e-02 5.92711389e-01 7.41310120e-01
-6.40786767e-01 -1.61790609e-01 -1.05632973e+00 6.66225374e-01
-2.38737345e+00 -5.75830817e-01 -5.18555939e-01 1.80989242e+00
5.81119895e-01 -2.89042532e-01 1.00565322e-01 -5.97137883e-02
6.80688620e-01 -3.40102285e-01 -9.28489566e-01 -4.30922002e-01
1.07373133e-01 -2.97117025e-01 9.34696794e-01 4.35581505e-01
-1.00083578e+00 9.66916978e-01 5.25245142e+00 4.13005650e-01
-8.21167946e-01 -3.21524113e-01 -3.29002589e-01 1.67148784e-01
4.67761979e-02 2.19463222e-02 -6.45797670e-01 5.79986870e-01
5.42204157e-02 -2.06239834e-01 6.77885115e-01 1.10554433e+00
4.36288983e-01 -3.70068938e-01 -8.18780780e-01 7.71368146e-01
-4.80761081e-01 -1.01222134e+00 -2.31422514e-01 1.51800230e-01
7.89826155e-01 -1.05865471e-01 -3.74310017e-01 -5.62500358e-02
1.63961023e-01 -6.36164308e-01 9.36624169e-01 5.50307393e-01
5.12214065e-01 -4.34809387e-01 3.58511150e-01 8.66794467e-01
-1.32712448e+00 -6.00505948e-01 -7.41832256e-01 -3.80464733e-01
9.05831993e-01 5.23065269e-01 -2.95489132e-01 6.58977687e-01
7.03819096e-01 4.42424983e-01 4.94283825e-01 1.00132060e+00
-1.06984250e-01 -2.49522761e-01 -6.42386377e-01 -6.19756401e-01
3.48002493e-01 -7.26785839e-01 9.85727608e-01 5.14189005e-01
5.18457770e-01 3.35683227e-01 4.70429391e-01 9.57334220e-01
6.22550070e-01 5.89160994e-02 -3.62996697e-01 -7.23916069e-02
3.09745550e-01 1.26803160e+00 -7.87324131e-01 3.04106593e-01
1.25977501e-01 1.07876754e+00 1.36379421e-01 9.48728547e-02
-7.71799326e-01 -5.48093915e-01 7.79288530e-01 -9.50474143e-02
5.67369983e-02 -1.20538104e+00 -5.97411215e-01 -1.06180513e+00
5.51901460e-01 -2.59998322e-01 -2.25075498e-01 -7.15501964e-01
-9.84362662e-01 -1.55927464e-01 2.33186841e-01 -1.29359937e+00
-3.45327348e-01 -9.66806293e-01 -5.24464667e-01 6.09215319e-01
-1.19562244e+00 -9.73864198e-01 -5.81393063e-01 3.20113629e-01
6.06615543e-01 2.62474954e-01 5.70365608e-01 -4.75773169e-03
-3.83107096e-01 -1.56884447e-01 1.48483887e-01 -3.75452548e-01
2.72529006e-01 -1.06980193e+00 2.14029834e-01 3.21667373e-01
-1.28077257e+00 6.32138610e-01 8.06355953e-01 -1.02444077e+00
-2.55193377e+00 -8.48385870e-01 3.78337801e-01 2.66272366e-01
6.93237841e-01 1.64398570e-02 -5.30469060e-01 3.61715347e-01
-4.26460683e-01 -4.78109539e-01 -2.90822208e-01 -5.87452471e-01
7.11765170e-01 2.34046966e-01 -1.42403233e+00 6.37815714e-01
1.44997013e+00 4.48437721e-01 -4.26885992e-01 6.45086646e-01
5.61455905e-01 -8.55377614e-01 -9.29577708e-01 7.81920075e-01
1.07413495e+00 -1.13546960e-01 1.16360438e+00 -4.23302382e-01
7.31299877e-01 -2.87772268e-01 -3.71543020e-01 -8.97295594e-01
-3.33357960e-01 -8.16857398e-01 -3.22004825e-01 8.61056507e-01
1.47733063e-01 -4.91898775e-01 9.86104488e-01 3.65535736e-01
-3.18379998e-01 -1.24989343e+00 -1.07430756e+00 -1.05448377e+00
2.69289941e-01 4.04426843e-01 -2.62049623e-02 6.97127044e-01
4.52798456e-01 -2.66805381e-01 -7.77231276e-01 5.52113116e-01
9.10853922e-01 4.52781886e-01 8.63067210e-01 -1.03299630e+00
-4.41189498e-01 -3.61765213e-02 -1.44091651e-01 -1.17813599e+00
-7.72465244e-02 -6.35876119e-01 6.71070278e-01 -2.09143233e+00
3.05552166e-02 -6.92460537e-01 5.27612567e-01 5.87875366e-01
1.58132836e-01 -3.40618819e-01 3.00442964e-01 4.62733746e-01
-1.92337424e-01 5.65220535e-01 1.72212279e+00 8.36051106e-02
-4.70390528e-01 -2.34775105e-03 2.74416581e-02 6.25091553e-01
8.76252770e-01 -1.38391450e-01 -1.24536082e-01 -8.71746600e-01
3.29037756e-01 7.52029538e-01 2.74030626e-01 -1.10337090e+00
4.13001806e-01 -7.87226737e-01 -2.08306551e-01 -6.76443219e-01
4.17256474e-01 -5.73031247e-01 1.38306186e-01 1.09592855e+00
1.63248762e-01 2.77893871e-01 -6.00964688e-02 5.47315538e-01
3.83700952e-02 -3.46494496e-01 4.82629329e-01 -4.00052845e-01
-8.34365368e-01 2.55949587e-01 -4.75529850e-01 -2.84851164e-01
1.30823147e+00 -1.79644987e-01 -4.44277048e-01 1.72138605e-02
-8.52168739e-01 1.07386172e+00 6.95076466e-01 3.66479307e-01
6.28698707e-01 -1.08689594e+00 -5.31537175e-01 -5.52043378e-01
-3.90378863e-01 4.86818910e-01 2.81272858e-01 8.63614440e-01
-1.31416094e+00 2.75323898e-01 -8.57861757e-01 -7.41904080e-01
-8.27466071e-01 7.35694468e-02 -1.24353282e-01 -1.64716363e-01
-6.22445107e-01 7.05808818e-01 -4.02798951e-01 -7.09497750e-01
-1.72006488e-02 -5.24453223e-01 -5.22560291e-02 -1.87086359e-01
-3.06310743e-01 1.11273122e+00 -3.05983901e-01 -4.19311561e-02
-5.02993345e-01 1.15657663e+00 5.53670764e-01 -1.70597494e-01
1.73873079e+00 -1.85486581e-02 -4.82878208e-01 -6.74705431e-02
8.33743989e-01 -5.24641752e-01 -1.49995422e+00 6.87694550e-01
8.46556649e-02 -9.97591242e-02 -5.68388462e-01 -3.57557058e-01
-5.17643750e-01 5.60387552e-01 3.98893245e-02 -2.18592495e-01
6.24502957e-01 -4.22502041e-01 1.00333083e+00 1.13267839e+00
1.37845135e+00 -1.50636959e+00 1.43471420e-01 8.87215912e-01
1.40120876e+00 -7.40956724e-01 3.35481197e-01 -9.86925423e-01
-5.18385291e-01 1.55270743e+00 6.11092567e-01 -7.42598772e-01
3.13179225e-01 3.27810735e-01 -3.77613127e-01 6.69300333e-02
-5.17856479e-01 2.64001220e-01 -3.29563648e-01 3.47598881e-01
-3.23906213e-01 2.53155142e-01 -1.03054249e+00 4.82344031e-01
-2.63518803e-02 2.38306582e-01 6.11840427e-01 1.82688284e+00
-9.40056443e-01 -9.95747507e-01 -3.73199105e-01 1.46957442e-01
3.76571529e-02 5.83451450e-01 -1.85861364e-01 7.77917624e-01
8.19792673e-02 8.09162080e-01 -1.38246030e-01 -3.16483170e-01
4.03999507e-01 -4.31207836e-01 8.44347894e-01 -3.92474979e-01
-1.76158585e-02 -2.94536680e-01 5.65172993e-02 -9.52935398e-01
-3.44636887e-01 -7.06878841e-01 -1.79843199e+00 1.19207695e-01
-9.95080844e-02 -2.20245600e-01 9.76076782e-01 7.13714302e-01
1.43863976e-01 5.29850014e-02 4.74866211e-01 -1.69373226e+00
-9.52002585e-01 -7.27970660e-01 -6.35351956e-01 -4.23174091e-02
-1.28806725e-01 -1.17943430e+00 -8.67162421e-02 -8.06193948e-02] | [4.786753177642822, 1.1327108144760132] |
58926f58-46de-4328-9b6f-d7a087cf29bf | on-the-benefits-of-selectivity-in-pseudo | 2202.00796 | null | https://arxiv.org/abs/2202.00796v3 | https://arxiv.org/pdf/2202.00796v3.pdf | On Balancing Bias and Variance in Unsupervised Multi-Source-Free Domain Adaptation | Due to privacy, storage, and other constraints, there is a growing need for unsupervised domain adaptation techniques in machine learning that do not require access to the data used to train a collection of source models. Existing methods for multi-source-free domain adaptation (MSFDA) typically train a target model using pseudo-labeled data produced by the source models, which focus on improving the pseudo-labeling techniques or proposing new training objectives. Instead, we aim to analyze the fundamental limits of MSFDA. In particular, we develop an information-theoretic bound on the generalization error of the resulting target model, which illustrates an inherent bias-variance trade-off. We then provide insights on how to balance this trade-off from three perspectives, including domain aggregation, selective pseudo-labeling, and joint feature alignment, which leads to the design of novel algorithms. Experiments on multiple datasets validate our theoretical analysis and demonstrate the state-of-art performance of the proposed algorithm, especially on some of the most challenging datasets, including Office-Home and DomainNet. | ['Gregory Wornell', 'Yuheng Bu', 'Maohao Shen'] | 2022-02-01 | null | null | null | null | ['source-free-domain-adaptation'] | ['computer-vision'] | [ 4.35264707e-01 1.18946601e-02 -4.93238598e-01 -6.21980131e-01
-8.32019031e-01 -6.70747101e-01 5.59134245e-01 2.21833453e-01
-4.43555564e-01 9.02832925e-01 7.02892020e-02 -8.51200521e-02
-3.41507971e-01 -5.89003742e-01 -6.59390330e-01 -7.08954155e-01
2.80277401e-01 5.33262491e-01 1.68143734e-01 8.85338113e-02
4.99751568e-02 2.84613252e-01 -1.26876616e+00 -6.06435575e-02
1.18118072e+00 1.09446466e+00 1.18245937e-01 1.09921210e-01
-1.79831594e-01 3.10161442e-01 -4.24033940e-01 -5.25426924e-01
4.59979713e-01 -5.49248576e-01 -6.99711382e-01 2.88544506e-01
1.57866225e-01 -1.80641457e-01 -3.92555222e-02 1.25617003e+00
4.56059039e-01 2.79380322e-01 7.95335948e-01 -1.37660837e+00
-6.77352846e-01 4.04332370e-01 -3.16566080e-01 1.57818757e-02
-8.52202922e-02 -2.70363837e-01 8.53956342e-01 -7.48276532e-01
6.44546032e-01 8.97643328e-01 4.04163599e-01 7.50402093e-01
-1.22734058e+00 -7.58155763e-01 2.99967676e-01 1.91893548e-01
-1.50459099e+00 -6.26186073e-01 8.46956253e-01 -4.46030289e-01
3.44805956e-01 -1.91284698e-02 3.53123285e-02 1.23706341e+00
-3.79506916e-01 7.85564125e-01 1.20298326e+00 -7.10563242e-01
6.20378613e-01 7.86616325e-01 1.65669903e-01 3.36828142e-01
5.16791642e-01 -2.70360205e-02 -5.79199016e-01 -4.48971242e-01
7.11547732e-01 -1.26105577e-01 -2.78620422e-01 -1.12301219e+00
-9.57796097e-01 8.16329598e-01 1.03572011e-01 5.75090237e-02
-2.59754270e-01 -5.62905967e-01 3.29369992e-01 4.09306198e-01
5.38039923e-01 3.83910418e-01 -6.46787405e-01 2.16893926e-01
-8.28460455e-01 8.04767907e-02 6.57021642e-01 1.35704052e+00
8.18979263e-01 -2.94823259e-01 1.14797652e-01 1.00630713e+00
2.53258973e-01 4.55592513e-01 5.96964896e-01 -6.56760752e-01
6.16875827e-01 5.56569040e-01 3.23044479e-01 -5.85866511e-01
-2.42900312e-01 -5.61428189e-01 -7.84123480e-01 -1.13406710e-01
6.20219827e-01 -1.19354121e-01 -7.13270009e-01 2.07065511e+00
5.88711560e-01 5.58467582e-02 3.23014885e-01 8.15602660e-01
1.83002844e-01 2.42193252e-01 1.91883966e-01 -4.61289674e-01
9.74753439e-01 -8.97604823e-01 -6.12765372e-01 -3.60566854e-01
7.11512268e-01 -3.10884148e-01 1.18042922e+00 2.60264315e-02
-6.71971738e-01 -3.72906089e-01 -1.14873004e+00 1.42190903e-01
-4.29927975e-01 1.05943727e-05 2.90496349e-01 7.46393383e-01
-6.01477623e-01 3.95078242e-01 -6.89802647e-01 -7.27860630e-01
5.56578100e-01 2.84757525e-01 -4.47967172e-01 -2.18322977e-01
-1.13232756e+00 6.99666560e-01 5.57016730e-01 -3.88850063e-01
-5.17517388e-01 -6.76156878e-01 -5.44697821e-01 -3.14253122e-02
5.66288710e-01 -6.19397759e-01 1.22644973e+00 -1.16742396e+00
-1.52696049e+00 9.02435303e-01 -2.22225890e-01 -5.31123996e-01
6.67947054e-01 -1.86266601e-01 -6.61530256e-01 2.07905881e-02
1.72337025e-01 2.68674105e-01 8.22488010e-01 -1.31220496e+00
-8.91388655e-01 -5.78681171e-01 -1.34322882e-01 1.75775930e-01
-9.06347871e-01 -1.07698940e-01 -2.79381424e-01 -6.51658058e-01
8.37551504e-02 -8.92719448e-01 -3.57135057e-01 -2.88824160e-02
-3.24023932e-01 -1.24757893e-01 6.64157867e-01 -3.06139618e-01
1.24874628e+00 -2.35364175e+00 1.01166748e-01 4.45531487e-01
1.11889929e-01 3.93943667e-01 -1.43781394e-01 2.21327648e-01
2.51203217e-02 8.73820335e-02 -3.83303434e-01 -4.26341444e-01
1.06726810e-02 2.23282278e-01 -4.35522407e-01 4.50534105e-01
8.28799903e-02 5.01652122e-01 -8.35264504e-01 -4.79957700e-01
3.15792449e-02 1.82906628e-01 -4.73760158e-01 2.67808110e-01
-2.23603919e-01 6.66920245e-01 -7.52814710e-01 4.93156582e-01
8.11172426e-01 -5.30114532e-01 5.04777670e-01 3.70790996e-02
2.28173092e-01 3.50889027e-01 -1.24028051e+00 1.77847338e+00
-4.09523726e-01 9.30578038e-02 4.29962017e-02 -1.21924734e+00
8.76602411e-01 1.36060387e-01 5.67992926e-01 -6.13694906e-01
6.28515035e-02 5.66329837e-01 -3.36756080e-01 -2.06206292e-01
1.18886389e-01 -6.42594621e-02 -1.86932817e-01 4.75293994e-01
2.25635082e-01 4.92638081e-01 3.88410757e-03 -3.07846945e-02
8.65239084e-01 4.76557128e-02 5.88146269e-01 -3.99463028e-01
5.71975708e-01 1.47623625e-02 6.66565061e-01 7.87331522e-01
-3.30470890e-01 4.72074836e-01 1.90369308e-01 -2.14013010e-01
-9.89146054e-01 -1.10972488e+00 -2.25938469e-01 1.24108708e+00
2.71422386e-01 -2.04021111e-01 -8.86427402e-01 -1.08763134e+00
-7.45827407e-02 7.86832690e-01 -4.69813526e-01 -3.53069693e-01
-3.74714702e-01 -6.56391859e-01 3.13980758e-01 3.82979691e-01
6.19024098e-01 -4.98460352e-01 -4.84961927e-01 2.25194663e-01
-2.99700737e-01 -1.22240734e+00 -4.32713032e-01 2.99597472e-01
-9.86141860e-01 -8.74801755e-01 -8.10495675e-01 -6.16430998e-01
8.27549517e-01 3.70402932e-01 8.34263146e-01 -3.92828494e-01
3.58447254e-01 2.99417406e-01 -3.87552857e-01 -4.47079122e-01
-4.64355856e-01 5.76559603e-01 4.18915838e-01 3.19092959e-01
6.46216154e-01 -9.18605328e-01 -3.56027365e-01 4.73898381e-01
-8.67514849e-01 -1.49247810e-01 8.01290929e-01 8.32836449e-01
8.14705789e-01 -6.56331256e-02 9.13953006e-01 -1.22762966e+00
5.38402438e-01 -7.35245228e-01 -7.76621044e-01 6.56337202e-01
-1.01178479e+00 2.49891609e-01 8.67747009e-01 -6.34726405e-01
-1.21503901e+00 3.97906125e-01 2.89050192e-01 -5.20240128e-01
-2.87826151e-01 3.67829174e-01 -6.79553151e-01 -8.56666118e-02
8.94363463e-01 4.74645108e-01 -4.98206802e-02 -7.44780362e-01
4.73825246e-01 9.21954274e-01 5.62230110e-01 -6.93479002e-01
9.31768417e-01 4.69506979e-01 -1.72761202e-01 -7.71080256e-01
-1.04821718e+00 -5.73592067e-01 -9.15554583e-01 2.33905092e-01
4.24039066e-01 -9.38308477e-01 -1.25097344e-02 2.90462434e-01
-7.88599789e-01 -1.95744589e-01 -4.17883694e-01 4.36763316e-01
-7.30864763e-01 3.57427746e-01 1.16504803e-01 -7.26490498e-01
-1.53175682e-01 -8.35902989e-01 6.26687825e-01 2.33537212e-01
-5.94293326e-02 -9.49999750e-01 4.91724175e-04 2.35984221e-01
3.89210045e-01 7.30120949e-03 1.05136406e+00 -1.27844059e+00
-4.73953247e-01 -1.37180045e-01 -2.34883085e-01 4.17304307e-01
3.19675595e-01 -7.42327929e-01 -9.79327619e-01 -3.55537057e-01
7.90437758e-02 -1.91683874e-01 5.70124388e-01 1.71715796e-01
1.08474243e+00 -5.28747439e-01 -5.27926087e-01 7.42143095e-01
1.43681157e+00 1.15754016e-01 2.37798885e-01 4.24722403e-01
5.49322128e-01 5.69694817e-01 8.32189679e-01 5.52618682e-01
2.95541435e-01 8.09711158e-01 9.06373784e-02 1.49791181e-01
7.40590468e-02 -5.62629938e-01 1.03884540e-01 5.44729412e-01
2.66684800e-01 -2.10566267e-01 -7.47421682e-01 7.84113109e-01
-1.95662165e+00 -5.77296615e-01 3.26970816e-01 2.75292158e+00
9.36585903e-01 -3.91901582e-02 3.69685471e-01 -8.37368369e-02
7.47602344e-01 -1.25555992e-01 -9.72551525e-01 1.66603364e-02
-1.27603963e-01 6.11724481e-02 7.87274241e-01 1.46132976e-01
-1.28635943e+00 8.71946514e-01 6.45547390e+00 8.56596291e-01
-1.07223964e+00 3.03470910e-01 4.31980163e-01 -8.83609336e-03
-1.55708566e-01 -7.02578947e-02 -8.95873010e-01 5.30848563e-01
1.09183478e+00 -7.19346762e-01 5.89815915e-01 1.31788313e+00
-1.46586254e-01 2.44450718e-01 -1.24198294e+00 8.98675084e-01
-1.25447899e-01 -1.05785227e+00 -1.81839447e-02 2.57862210e-01
7.78562307e-01 -2.16522850e-02 -7.69306254e-03 2.29669943e-01
3.05838376e-01 -4.57213342e-01 5.49267173e-01 9.91559029e-02
9.81101096e-01 -7.18522012e-01 3.79811943e-01 5.38088143e-01
-8.72968614e-01 -3.21832955e-01 -4.29262787e-01 2.31030390e-01
-1.82119161e-02 4.98422265e-01 -8.20899129e-01 7.30338752e-01
5.58036506e-01 5.56732059e-01 -5.25774181e-01 9.75261152e-01
-5.38862869e-02 5.68349063e-01 -4.65031803e-01 1.18750565e-01
-1.86872661e-01 -1.38786823e-01 4.77023959e-01 9.95582640e-01
2.87716836e-01 -7.53662065e-02 5.72948158e-02 7.41802752e-01
-2.57667750e-01 2.98117459e-01 -5.71641982e-01 -1.09937258e-01
9.78854775e-01 8.96000445e-01 -4.87339616e-01 -3.16711992e-01
-5.79265356e-01 1.00045955e+00 4.40805495e-01 4.39095438e-01
-7.22258091e-01 -2.53040165e-01 8.68626177e-01 2.07661197e-01
2.60519952e-01 -1.34150326e-01 -5.42992949e-01 -1.38554800e+00
3.44839513e-01 -9.34641182e-01 6.14471734e-01 -8.07979926e-02
-1.49821234e+00 5.31744480e-01 3.01509649e-01 -1.47053182e+00
-3.08128387e-01 -2.82419831e-01 -8.74974951e-02 7.61769295e-01
-1.59718299e+00 -1.01885593e+00 -1.26400441e-01 7.28245974e-01
4.11292404e-01 -2.82971233e-01 9.69798446e-01 5.21750450e-01
-4.59238857e-01 1.12025356e+00 7.17829704e-01 5.23943156e-02
9.62542295e-01 -9.04860377e-01 2.96082526e-01 1.00244248e+00
1.16478764e-01 6.85169756e-01 5.27165174e-01 -4.29107338e-01
-1.08881688e+00 -1.33447385e+00 8.02780926e-01 -4.17978257e-01
4.70327884e-01 -4.96617824e-01 -9.88271534e-01 7.48157501e-01
-2.80355006e-01 5.69576956e-02 8.60143840e-01 1.91658244e-01
-5.97610354e-01 -2.86940873e-01 -1.47561622e+00 4.05623674e-01
1.17841673e+00 -3.84675592e-01 -3.68244261e-01 3.43994081e-01
6.77722514e-01 -6.03909791e-02 -7.51594067e-01 3.10776919e-01
4.76156801e-01 -7.76938140e-01 9.93257761e-01 -8.60517025e-01
1.17995024e-01 -1.45960912e-01 -3.86088282e-01 -1.34724534e+00
-2.70213544e-01 -4.55635190e-01 -1.94000214e-01 1.42274594e+00
4.14265990e-01 -9.28402781e-01 7.57698119e-01 9.01356757e-01
2.08366215e-01 -3.70709449e-01 -1.00860417e+00 -1.17531538e+00
9.08774212e-02 -1.49998903e-01 8.40059519e-01 1.26984179e+00
-8.20544437e-02 3.07541639e-01 -4.84414518e-01 2.89555907e-01
9.43327367e-01 1.25373319e-01 8.00242543e-01 -1.35734034e+00
-1.73693508e-01 -1.01805739e-01 -1.61483735e-01 -1.30031359e+00
1.90248206e-01 -6.71334803e-01 -4.95236479e-02 -1.12865782e+00
1.10157281e-01 -7.77022004e-01 -6.39425218e-01 4.89433914e-01
-1.10780507e-01 -2.28574350e-01 4.05967943e-02 5.45373261e-01
-6.69737101e-01 5.46912193e-01 8.46530557e-01 1.61830410e-01
-3.61863703e-01 2.23969340e-01 -1.04433203e+00 6.05115414e-01
7.93133795e-01 -6.63057864e-01 -6.97028279e-01 -4.41712290e-01
-2.30967879e-01 -2.48004720e-01 1.24671072e-01 -1.03287077e+00
1.67264163e-01 -4.98822868e-01 1.63430855e-01 -1.00979984e-01
2.08399728e-01 -1.16509151e+00 -4.47402053e-05 1.08405180e-01
-5.77623546e-01 -4.50240284e-01 -1.46354213e-01 9.77520525e-01
-7.63392523e-02 -1.20397322e-01 1.06799233e+00 6.32906333e-02
-7.85035431e-01 3.52083951e-01 5.02103418e-02 1.32450461e-01
1.05658317e+00 -1.36286557e-01 -2.02145249e-01 -3.00271988e-01
-5.35322964e-01 7.80657455e-02 5.90417504e-01 5.39040327e-01
2.01363906e-01 -1.45623767e+00 -4.85032797e-01 4.42097127e-01
4.68908012e-01 -2.78665312e-02 1.03919737e-01 5.84938109e-01
1.96291178e-01 4.65246558e-01 -1.97409630e-01 -4.40861553e-01
-9.73058522e-01 8.65325987e-01 8.21573436e-02 -3.12939763e-01
-3.67366344e-01 6.45844221e-01 3.53773981e-01 -5.14698863e-01
2.53520101e-01 3.49637046e-02 3.30233276e-02 -1.47151634e-01
4.31772500e-01 1.88370109e-01 1.05125301e-01 -5.58751762e-01
-5.25748491e-01 2.36878887e-01 -2.41878822e-01 -8.79046246e-02
1.13620794e+00 -5.00427544e-01 3.50349605e-01 2.90788740e-01
1.17496991e+00 -1.28290892e-01 -1.29669464e+00 -9.12840426e-01
2.87917793e-01 -6.72728121e-01 -1.97167173e-01 -8.70970607e-01
-8.30971897e-01 7.26270318e-01 7.73887336e-01 1.44834563e-01
1.44656301e+00 -4.87014167e-02 7.80712783e-01 4.91868883e-01
6.89943731e-01 -1.26361132e+00 -2.24327743e-01 2.75060207e-01
4.27809209e-01 -1.30579174e+00 -1.89723402e-01 -4.05412793e-01
-6.85759366e-01 7.35769689e-01 5.52366793e-01 9.98484567e-02
6.26755238e-01 -1.70589179e-01 3.71525362e-02 3.81585807e-01
-5.23415446e-01 -1.84228316e-01 1.21835425e-01 9.16538477e-01
-1.75017528e-02 6.41673878e-02 -3.40253502e-01 1.07154226e+00
8.13368857e-02 2.23402247e-01 1.40398338e-01 8.67464602e-01
-2.55982488e-01 -1.49098790e+00 -1.95725590e-01 3.11170846e-01
-4.18086350e-01 5.84435612e-02 -4.53352869e-01 6.95244670e-01
-1.10618308e-01 8.65221620e-01 -3.17366540e-01 -3.29892427e-01
3.00788432e-01 4.75833476e-01 2.27364376e-01 -4.96662676e-01
5.89127168e-02 -1.19960243e-02 -1.07784614e-01 -2.89516628e-01
-4.80758488e-01 -6.27524137e-01 -8.65699053e-01 -6.23975694e-02
-4.21488464e-01 2.58051395e-01 6.47658587e-01 9.97335553e-01
8.62946332e-01 7.65708834e-02 7.85118520e-01 -2.97326207e-01
-1.03064370e+00 -8.87713969e-01 -6.09591067e-01 5.70996404e-01
2.36922860e-01 -8.38189781e-01 -3.16372961e-01 1.76829815e-01] | [10.36361312866211, 3.1934139728546143] |
24668202-8bc4-4b90-b79c-5be694751715 | learning-markerless-robot-depth-camera | 2212.07567 | null | https://arxiv.org/abs/2212.07567v1 | https://arxiv.org/pdf/2212.07567v1.pdf | Learning Markerless Robot-Depth Camera Calibration and End-Effector Pose Estimation | Traditional approaches to extrinsic calibration use fiducial markers and learning-based approaches rely heavily on simulation data. In this work, we present a learning-based markerless extrinsic calibration system that uses a depth camera and does not rely on simulation data. We learn models for end-effector (EE) segmentation, single-frame rotation prediction and keypoint detection, from automatically generated real-world data. We use a transformation trick to get EE pose estimates from rotation predictions and a matching algorithm to get EE pose estimates from keypoint predictions. We further utilize the iterative closest point algorithm, multiple-frames, filtering and outlier detection to increase calibration robustness. Our evaluations with training data from multiple camera poses and test data from previously unseen poses give sub-centimeter and sub-deciradian average calibration and pose estimation errors. We also show that a carefully selected single training pose gives comparable results. | ['Baris Akgun', 'Bugra C. Sefercik'] | 2022-12-15 | null | null | null | null | ['camera-calibration', 'keypoint-detection'] | ['computer-vision', 'computer-vision'] | [-1.26072541e-01 1.42756224e-01 -2.87680477e-01 -2.36116707e-01
-1.25428665e+00 -8.38381112e-01 5.87968469e-01 8.34813789e-02
-6.15422070e-01 6.34262145e-01 -3.30215633e-01 -1.64587915e-01
9.09637436e-02 -4.95467819e-02 -9.90658462e-01 -3.74311566e-01
-1.59869082e-02 5.96333623e-01 4.01247501e-01 1.42850000e-02
4.46750432e-01 7.31084168e-01 -6.33126199e-01 -6.02023005e-01
6.97200239e-01 5.86855471e-01 -9.87384692e-02 1.32759142e+00
6.82478070e-01 2.28098303e-01 -6.17186904e-01 -2.52525955e-01
5.91472030e-01 -2.04264909e-01 -1.68866411e-01 6.10050745e-02
7.76226699e-01 -4.05798733e-01 -2.09797859e-01 7.10864305e-01
7.69043505e-01 -3.64084281e-02 4.29392457e-01 -1.04276204e+00
3.10078591e-01 -9.66912583e-02 -5.37736356e-01 -2.63605952e-01
8.81537735e-01 8.25189650e-02 1.15732938e-01 -8.31845284e-01
8.34810793e-01 7.86973059e-01 1.43122506e+00 4.76561368e-01
-1.23269582e+00 -4.56361771e-01 -4.16557580e-01 -5.18821239e-01
-1.37019372e+00 -3.53830844e-01 6.58238113e-01 -4.09995049e-01
6.31424189e-01 1.77228063e-01 6.99239910e-01 1.05073214e+00
7.39514172e-01 2.99213201e-01 9.20435488e-01 -6.66672707e-01
2.14613959e-01 8.70141312e-02 -2.74373621e-01 7.86667228e-01
4.28787351e-01 6.83557272e-01 -2.65376031e-01 -3.10990304e-01
1.44180906e+00 -2.45083451e-01 -4.76699412e-01 -1.03902197e+00
-1.64002359e+00 6.02332115e-01 2.49399871e-01 -4.50274050e-01
-8.21016356e-02 6.84381187e-01 1.80056989e-02 1.31098982e-02
-1.42575979e-01 6.10176146e-01 -6.87950552e-01 -4.13117051e-01
-9.06359076e-01 2.79793471e-01 1.12931395e+00 1.51873505e+00
6.51089847e-01 1.30171493e-01 4.34456944e-01 3.21425170e-01
5.89046717e-01 7.56464899e-01 5.79481363e-01 -1.48604548e+00
4.48992848e-01 1.65695861e-01 6.20648742e-01 -9.32121336e-01
-4.68994111e-01 -2.43251890e-01 -5.90903088e-02 4.78641689e-01
6.34969652e-01 -5.68059444e-01 -1.03591800e+00 1.27281678e+00
6.59302413e-01 4.97159839e-01 3.47355101e-03 1.00021565e+00
2.21495867e-01 -5.41659556e-02 -2.64941067e-01 2.87792441e-02
7.42785394e-01 -6.97691202e-01 -3.26496959e-01 -1.67604297e-01
6.51004732e-01 -1.21854544e+00 6.74301565e-01 7.08526134e-01
-8.12952340e-01 -5.89929938e-01 -1.34681535e+00 1.08282171e-01
1.32200435e-01 3.27282846e-01 4.47876513e-01 6.87067986e-01
-7.09805191e-01 9.09027457e-01 -1.27517784e+00 -3.54336917e-01
-2.70863712e-01 8.35338354e-01 -4.36002016e-01 4.56068426e-01
-6.17855489e-01 1.12061095e+00 2.68152505e-01 -1.51829556e-01
-9.13110495e-01 -7.49099553e-01 -9.83727515e-01 -8.07936549e-01
3.70927691e-01 -7.17229128e-01 1.53572476e+00 -5.68264842e-01
-2.11450124e+00 7.08019316e-01 1.85201854e-01 -5.55187464e-01
9.00502503e-01 -7.11900234e-01 -2.03338161e-01 2.76470661e-01
-2.51774132e-01 6.70428216e-01 1.03610861e+00 -1.50619543e+00
-3.33945304e-01 -1.90149650e-01 -2.25671902e-01 2.01421797e-01
4.68730122e-01 -4.21263576e-01 -6.59092844e-01 -6.70393825e-01
5.74282587e-01 -1.57409465e+00 -5.79075277e-01 1.12732530e-01
-4.26081896e-01 7.93087602e-01 8.18067491e-01 -6.68703914e-01
7.10723639e-01 -1.75664830e+00 -2.90140986e-01 4.67260361e-01
1.16281576e-01 -8.13114867e-02 2.17312202e-01 -6.10579387e-04
-1.07894868e-01 -4.61740702e-01 1.61010697e-01 -1.80804566e-01
-2.21078783e-01 4.93334010e-02 -2.41113797e-01 9.14011121e-01
-2.09793136e-01 6.01226270e-01 -9.83651876e-01 -5.72579026e-01
7.99391627e-01 5.36245823e-01 -4.87318456e-01 2.25601837e-01
6.77409694e-02 7.26058781e-01 -2.12346345e-01 6.26820683e-01
4.61098939e-01 1.53036729e-01 1.19637989e-01 -4.46710289e-01
-4.92033549e-02 -1.13316495e-02 -1.56809878e+00 2.10833192e+00
-4.94333237e-01 6.00312531e-01 1.80520877e-01 -1.10595092e-01
1.13866532e+00 3.08863431e-01 7.19751120e-01 -8.76005888e-02
5.63352644e-01 4.36587870e-01 -2.42467567e-01 -1.26808574e-02
6.68767869e-01 2.81561688e-02 -2.00960666e-01 1.44558296e-01
2.06199586e-01 -8.52187097e-01 -3.07068378e-01 1.03287213e-01
8.05970848e-01 9.91601586e-01 5.22385418e-01 -8.85864347e-02
2.10739046e-01 4.02323753e-01 5.43166697e-01 4.38755035e-01
-2.64831871e-01 1.05881715e+00 -1.32168131e-02 -4.14759576e-01
-1.30959487e+00 -1.17583883e+00 -1.92709774e-01 2.68004447e-01
5.28638899e-01 -5.43361723e-01 -8.59068811e-01 -4.17656273e-01
3.13362449e-01 3.73201162e-01 -2.73311913e-01 -1.67659760e-01
-9.72083211e-01 -2.03604758e-01 5.25849640e-01 5.90355456e-01
5.62029555e-02 1.20126437e-02 -8.82304788e-01 4.89671379e-01
4.02114362e-01 -1.16525698e+00 -5.62723160e-01 3.80868733e-01
-1.27846539e+00 -1.48589611e+00 -7.12027848e-01 -4.81491864e-01
8.06189477e-01 -8.44365433e-02 9.15126383e-01 -3.53139997e-01
-2.16225892e-01 8.49032819e-01 1.80269070e-02 -4.73663241e-01
-4.32449132e-01 -2.51641095e-01 4.11097556e-01 -7.10058987e-01
4.80185822e-02 -1.38456598e-01 -5.10189295e-01 7.29992449e-01
-5.63273765e-02 -2.67939836e-01 4.22915310e-01 7.66585767e-01
8.70324373e-01 -9.31487203e-01 -1.88115731e-01 -6.87138200e-01
3.90425652e-01 1.46761805e-01 -1.23799849e+00 -1.56398378e-02
-7.27096796e-01 1.98030815e-01 5.12384236e-01 -8.10311198e-01
-8.09766293e-01 8.95534456e-01 2.22703397e-01 -9.63446915e-01
-8.71528015e-02 2.73515508e-02 8.32529515e-02 -7.38974154e-01
1.21939373e+00 -4.31086868e-01 1.47034079e-01 -1.88328788e-01
5.25356352e-01 3.79083425e-01 1.11713386e+00 -9.65930700e-01
9.34069812e-01 3.20813447e-01 8.71797651e-02 -4.82408136e-01
-2.79277414e-01 -3.56881022e-01 -1.05082917e+00 -2.36341774e-01
5.15679359e-01 -1.12492001e+00 -9.28715587e-01 1.99983463e-01
-1.03526092e+00 -4.54696268e-01 -3.03899497e-01 1.15838778e+00
-1.01297593e+00 5.86432099e-01 -4.91004169e-01 -5.60511768e-01
-1.91569224e-01 -1.28867853e+00 1.30761111e+00 4.25117075e-01
-5.09202898e-01 -1.05387425e+00 4.77748394e-01 -1.83783218e-01
-1.27786785e-01 6.96568072e-01 -1.93355560e-01 -3.70235056e-01
-7.97048271e-01 -6.91740692e-01 5.43953955e-01 -5.56981675e-02
-1.27869502e-01 2.73057729e-01 -8.05111885e-01 -5.74662030e-01
1.43441603e-01 -7.98135474e-02 1.30213752e-01 5.30803502e-01
6.13075018e-01 9.67934541e-03 -7.04900444e-01 1.06958902e+00
1.50044727e+00 2.11340502e-01 5.73598623e-01 6.66226625e-01
7.26237893e-01 1.36967907e-02 1.22003877e+00 4.84419763e-01
2.53475904e-01 8.79911900e-01 3.17664415e-01 -2.04747766e-02
-4.38042404e-03 -6.04801536e-01 3.41198087e-01 5.42164922e-01
-1.15101747e-01 3.48220468e-01 -9.88907516e-01 2.29905114e-01
-1.53810990e+00 -4.28128719e-01 -3.44723642e-01 2.73685598e+00
7.19591796e-01 3.09112966e-01 6.03373945e-02 -1.36508152e-01
8.01743507e-01 -4.85814512e-01 -6.37577593e-01 -2.66747028e-01
4.31932420e-01 3.33966136e-01 1.35198641e+00 8.50891888e-01
-1.11465263e+00 1.00072408e+00 7.28303337e+00 1.20203294e-01
-1.31149578e+00 -2.30478734e-01 1.18966632e-01 3.41112554e-01
1.31921411e-01 4.22241628e-01 -1.06709743e+00 2.35185534e-01
1.12294793e+00 5.03775524e-03 1.68209374e-01 1.25935471e+00
1.93303302e-02 -4.50051636e-01 -1.19638240e+00 1.26232791e+00
8.03338811e-02 -9.96559560e-01 -6.92106903e-01 1.28192373e-03
7.72288740e-01 4.94595245e-03 -2.01686367e-01 -3.72954458e-02
5.97609043e-01 -5.65283477e-01 5.57769477e-01 5.66798270e-01
8.47613454e-01 -6.31023943e-01 4.13710713e-01 3.15960526e-01
-1.05275524e+00 3.18537831e-01 -4.30559337e-01 3.37008595e-01
1.43891990e-01 1.57111377e-01 -1.39900208e+00 5.92543483e-01
8.79901797e-02 5.53127110e-01 -5.72481513e-01 1.42254436e+00
-4.61927533e-01 2.65635461e-01 -7.62535453e-01 2.19584197e-01
-1.43072516e-01 -2.25456044e-01 6.23826981e-01 9.77351487e-01
4.93530601e-01 -9.77755934e-02 2.22910926e-01 4.10344809e-01
3.74120235e-01 -3.03542227e-01 -5.81093848e-01 5.66113234e-01
5.74282944e-01 1.29126179e+00 -5.33510923e-01 -1.98764294e-01
2.90644113e-02 1.05582154e+00 -2.40578100e-01 1.82673305e-01
-1.11670709e+00 -6.74017429e-01 4.48388636e-01 9.93811190e-02
9.67854932e-02 -8.24723899e-01 -3.19214195e-01 -1.29748762e+00
-2.45032519e-01 -6.69978678e-01 1.76164180e-01 -8.66204858e-01
-4.89169091e-01 1.59961313e-01 1.76266029e-01 -1.91364336e+00
-9.70614791e-01 -7.81260490e-01 -4.51975644e-01 6.08140588e-01
-1.00923228e+00 -9.49883878e-01 -4.81120944e-01 5.44812262e-01
1.50453329e-01 9.68917459e-02 7.72299528e-01 4.49150391e-02
-4.05218042e-02 7.82680273e-01 1.19223915e-01 2.76630431e-01
1.29753280e+00 -1.40256774e+00 3.73549134e-01 6.84174359e-01
1.24661542e-01 9.16594267e-01 9.95230198e-01 -7.82077134e-01
-1.67635381e+00 -8.13151062e-01 2.32092455e-01 -1.08011389e+00
5.38178921e-01 -1.62083268e-01 -4.17985708e-01 1.08690214e+00
-3.55019003e-01 4.15911436e-01 2.27275595e-01 -2.61437148e-01
7.67374709e-02 -1.22202665e-01 -1.29264545e+00 5.54891050e-01
5.96768618e-01 -3.32695216e-01 -7.51189649e-01 1.40066445e-01
4.90531415e-01 -1.58921063e+00 -1.11156940e+00 2.92180181e-01
8.57222676e-01 -5.43522477e-01 1.18153310e+00 -3.59551348e-02
-3.20545614e-01 -6.88401937e-01 1.12957783e-01 -1.33223975e+00
1.57506391e-01 -1.10644329e+00 -2.06995919e-01 7.12340295e-01
2.13900626e-01 -4.50891793e-01 1.23263025e+00 8.68450582e-01
-3.02109033e-01 -3.32869351e-01 -8.55937004e-01 -8.64739478e-01
-2.06440523e-01 -3.94330204e-01 1.10588901e-01 8.08435559e-01
-2.54483595e-02 1.49172291e-01 -3.98252010e-01 4.01775151e-01
1.02172744e+00 -1.14526831e-01 1.46105790e+00 -9.16405380e-01
-4.37977672e-01 2.76558906e-01 -7.91318417e-01 -1.14993095e+00
-1.69244900e-01 -2.13555321e-01 2.20293432e-01 -9.48963821e-01
-4.13423836e-01 -4.18877661e-01 3.47354233e-01 9.45158675e-02
3.12318113e-02 1.47407323e-01 1.17087848e-01 1.81969851e-01
-4.49884802e-01 -3.51012242e-03 1.03202367e+00 3.83820683e-01
-4.59602058e-01 1.44240022e-01 -6.71696588e-02 9.82667089e-01
7.18296647e-01 -5.80035210e-01 -2.96843022e-01 -3.18480045e-01
-7.44781047e-02 4.64263231e-01 5.41704953e-01 -1.49985588e+00
5.45923173e-01 -3.87779600e-03 1.04590440e+00 -6.44012809e-01
5.52836955e-01 -1.07396054e+00 3.79806340e-01 7.38048911e-01
6.85880482e-02 3.83191168e-01 2.77888238e-01 6.09262049e-01
1.00210935e-01 -2.82425553e-01 8.25136065e-01 -7.18884319e-02
-5.04033506e-01 7.82329738e-02 5.43779647e-03 -1.51612237e-01
1.27731538e+00 -8.30099642e-01 -3.75910960e-02 -3.78880918e-01
-9.02338982e-01 -1.81987047e-01 1.16578388e+00 1.36926159e-01
6.07804775e-01 -1.31906188e+00 -1.65006414e-01 2.55364388e-01
7.24330992e-02 2.13367254e-01 -4.80072498e-01 9.01252389e-01
-1.12598872e+00 3.18761706e-01 4.36067581e-03 -1.15249681e+00
-1.16525078e+00 3.49692196e-01 6.32669389e-01 1.93618611e-01
-2.71638453e-01 5.26439130e-01 -5.50302386e-01 -8.57710540e-01
1.08382180e-01 -6.51996255e-01 5.61077654e-01 -6.13726318e-01
-2.05610543e-02 5.06437957e-01 5.08060679e-02 -5.88097155e-01
-3.60712588e-01 1.16801560e+00 2.50894457e-01 -4.74455684e-01
9.48154628e-01 9.29656811e-03 6.56221509e-01 2.48674437e-01
1.07757902e+00 3.53359550e-01 -1.86831117e+00 2.63532341e-01
1.46228135e-01 -8.52889836e-01 -2.32174009e-01 -6.16484582e-01
-5.58430552e-01 6.69078231e-01 9.37672615e-01 -7.65338361e-01
4.44533974e-01 -3.12957466e-01 4.98006105e-01 4.39743042e-01
7.47849464e-01 -1.42210650e+00 1.52829677e-01 4.20633227e-01
5.10934353e-01 -1.15562356e+00 5.18396497e-01 -4.85414684e-01
-4.18768615e-01 1.52894080e+00 7.56989956e-01 -6.20489597e-01
4.22973841e-01 6.87043309e-01 4.31541026e-01 2.39267394e-01
-1.43781602e-01 4.09697384e-01 1.98271990e-01 7.81519413e-01
3.41336608e-01 -5.17982095e-02 -1.33622825e-01 2.01004967e-01
-3.19066465e-01 1.41871855e-01 6.61593854e-01 1.25827980e+00
-4.36886966e-01 -1.50119293e+00 -9.60495353e-01 -3.74706984e-02
-2.53033876e-01 3.57374460e-01 -2.40788057e-01 1.31686854e+00
-3.26042354e-01 4.48125869e-01 -1.72598734e-01 -5.46429038e-01
4.12740707e-01 -2.97570795e-01 9.73643064e-01 -4.03070539e-01
-5.85865676e-01 5.37677109e-01 2.48712376e-02 -8.22231889e-01
-1.70383960e-01 -6.56591177e-01 -1.62384307e+00 -1.57705590e-01
-8.02755713e-01 3.89671884e-02 1.09867167e+00 4.19010997e-01
3.86229694e-01 6.21418096e-02 4.65911567e-01 -1.17868519e+00
-8.95798802e-01 -6.48660362e-01 -2.52723575e-01 9.44187939e-02
4.10195172e-01 -6.09471262e-01 -3.00881565e-01 2.49254435e-01] | [7.260104179382324, -1.4198358058929443] |
9b3bf274-3ac7-4036-a75d-4bfbd6eecfa0 | first-insight-into-quality-adaptive-dialogue | null | null | https://aclanthology.org/L14-1092 | https://aclanthology.org/L14-1092.pdf | First Insight into Quality-Adaptive Dialogue | While Spoken Dialogue Systems have gained in importance in recent years, most systems applied in the real world are still static and error-prone. To overcome this, the user is put into the focus of dialogue management. Hence, an approach for adapting the course of the dialogue to Interaction Quality, an objective variant of user satisfaction, is presented in this work. In general, rendering the dialogue adaptive to user satisfaction enables the dialogue system to improve the course of the dialogue and to handle problematic situations better. In this contribution, we present a pilot study of quality-adaptive dialogue. By selecting the confirmation strategy based on the current IQ value, the course of the dialogue is adapted in order to improve the overall user experience. In a user experiment comparing three different confirmation strategies in a train booking domain, the adaptive strategy performs successful and is among the two best rated strategies based on the overall user experience. | ['H{\\"u}seyin Dikme', 'Stefan Ultes', 'Wolfgang Minker'] | 2014-05-01 | null | null | null | lrec-2014-5 | ['dialogue-management'] | ['natural-language-processing'] | [ 1.29703134e-02 5.47652066e-01 2.36997768e-01 -7.11884081e-01
-2.40807086e-01 -6.34212792e-01 4.91956264e-01 4.03407305e-01
-6.86190665e-01 8.55055749e-01 3.27813685e-01 -9.17325243e-02
-4.44368422e-01 -6.84958160e-01 3.26477170e-01 -2.21282646e-01
3.84858757e-01 6.70807362e-01 3.98932010e-01 -8.30088019e-01
7.99059033e-01 3.65039259e-01 -1.57755220e+00 2.73616642e-01
1.07617295e+00 5.28718412e-01 4.24798727e-01 7.57472277e-01
-5.31015396e-01 4.89861101e-01 -9.58061278e-01 -1.65978730e-01
-7.09421816e-04 -1.03712487e+00 -1.27416289e+00 2.37350747e-01
-6.87317327e-02 -2.95286715e-01 1.42248526e-01 7.02248096e-01
8.78164709e-01 5.56120038e-01 3.10595602e-01 -8.98225605e-01
3.56276840e-01 3.64354789e-01 5.11318803e-01 9.64826047e-02
9.94107544e-01 -1.51370272e-01 6.00289047e-01 -4.55230623e-01
4.86705333e-01 1.40914953e+00 9.31932330e-02 6.66276455e-01
-1.20585859e+00 -1.61205575e-01 -1.01622537e-01 7.64059052e-02
-9.96507347e-01 -4.71185148e-01 4.76861537e-01 -4.16145027e-01
7.32350290e-01 3.57615113e-01 6.75283194e-01 3.52723360e-01
1.04280464e-01 3.60343546e-01 1.14364994e+00 -7.26064682e-01
5.08543909e-01 9.67567503e-01 2.83708662e-01 3.06795329e-01
-3.10232460e-01 -3.30220222e-01 -4.47152674e-01 -1.66427985e-01
5.77115953e-01 -4.54841942e-01 -4.04527634e-01 -3.00014824e-01
-6.23446465e-01 9.73720431e-01 -1.94722284e-02 9.01407421e-01
-4.25110757e-01 -7.99135029e-01 5.28229594e-01 6.14014506e-01
2.72652388e-01 1.01264584e+00 -4.78412896e-01 -9.38624442e-01
-4.51007515e-01 5.23525476e-01 1.46783853e+00 2.93132365e-01
3.87133300e-01 -2.93597460e-01 -4.09616560e-01 1.27654302e+00
2.12160289e-01 6.58992156e-02 4.67130601e-01 -8.55512500e-01
1.42135611e-02 1.01752770e+00 2.76794314e-01 -6.25478566e-01
-6.97213292e-01 -4.50739525e-02 -2.43330538e-01 4.06825244e-01
6.69360638e-01 -5.50761461e-01 -3.07037443e-01 1.31334531e+00
3.76530528e-01 -8.71845782e-01 1.70954496e-01 7.55663216e-01
8.23388278e-01 5.81828594e-01 3.31216902e-02 -5.89517415e-01
1.27396178e+00 -6.11912847e-01 -1.33661675e+00 6.18160777e-02
6.83895946e-01 -9.09485459e-01 1.08833766e+00 5.25246739e-01
-1.15209091e+00 -6.44977868e-01 -7.66930759e-01 5.09163857e-01
-2.25067183e-01 -1.46569595e-01 3.86138191e-03 8.64182293e-01
-9.97626364e-01 5.91050684e-01 -1.67483523e-01 -7.82380998e-01
-6.85115993e-01 4.08540279e-01 -1.59504533e-01 2.89057165e-01
-1.40357542e+00 1.38086462e+00 3.69179249e-01 -7.43179470e-02
1.09756164e-01 -1.09409831e-01 -5.73072016e-01 1.90606564e-02
2.42403492e-01 -4.42772418e-01 1.57753611e+00 -1.01327980e+00
-2.26335454e+00 5.86205661e-01 2.26337865e-01 -1.51923999e-01
6.88027263e-01 -1.09126396e-01 -1.43179402e-01 6.77503049e-02
-8.44580308e-02 1.69389337e-01 2.09002540e-01 -1.17851281e+00
-9.29376841e-01 -2.68490911e-01 4.27752703e-01 8.31179678e-01
-1.42656311e-01 6.46754354e-02 -3.59382242e-01 1.01119526e-01
-1.69240236e-01 -8.35651934e-01 -1.34890288e-01 -5.81957281e-01
1.88868567e-01 -3.88714701e-01 5.85205913e-01 -4.96930391e-01
1.62479126e+00 -2.04764891e+00 2.68764675e-01 2.68221378e-01
-1.18972734e-01 5.51412165e-01 2.95682400e-01 7.71505713e-01
2.53245145e-01 -1.70360133e-01 7.25017115e-02 -2.29361728e-01
-1.99511871e-02 4.20398831e-01 3.34628195e-01 -1.64559148e-02
-1.80253550e-01 7.13388994e-02 -8.55681181e-01 -4.26015615e-01
4.27097291e-01 1.47578627e-01 -4.65127945e-01 9.80190694e-01
1.59809105e-02 7.84317851e-01 -3.73613715e-01 -1.10381380e-01
3.06553781e-01 3.50303531e-01 2.28738695e-01 1.15756884e-01
-4.64195579e-01 4.01228756e-01 -1.15233541e+00 1.23639929e+00
-7.63065159e-01 3.72384399e-01 1.11762561e-01 -6.64763689e-01
1.13568950e+00 7.25145757e-01 4.00705814e-01 -1.02749646e+00
1.39814883e-01 5.85346818e-02 4.17710841e-01 -9.61968124e-01
9.02405977e-01 1.17895855e-02 2.13409401e-03 4.82614011e-01
-1.20803170e-01 -4.33325499e-01 3.78662288e-01 2.69921869e-01
6.77105963e-01 3.28917988e-02 6.84989572e-01 -1.54438645e-01
7.26491213e-01 -1.47114635e-01 -3.58177386e-02 5.62380910e-01
-2.92296946e-01 1.80276200e-01 6.80245876e-01 -2.86742318e-02
-6.79589152e-01 -6.53126299e-01 -5.78273050e-02 1.20498109e+00
4.22222130e-02 -2.91655719e-01 -1.06742382e+00 -5.96123993e-01
-5.15069485e-01 1.09515417e+00 -1.44785807e-01 -1.42764449e-01
-3.49077493e-01 -2.78379798e-01 1.38025716e-01 -2.27267414e-01
5.96119523e-01 -1.32756710e+00 -8.42132211e-01 5.67323625e-01
-4.84553635e-01 -5.95415652e-01 -1.89350277e-01 1.03305638e-01
-7.35619485e-01 -7.01740265e-01 -6.64561152e-01 -4.94489759e-01
3.72116178e-01 -3.21331657e-02 9.03640926e-01 2.55766124e-01
1.75829932e-01 5.70828080e-01 -8.05413723e-01 -2.58719981e-01
-1.06229675e+00 2.30593204e-01 -1.41568527e-01 -3.11586261e-01
3.13361138e-02 -1.23832589e-02 -4.34962004e-01 4.71401066e-01
-8.05225730e-01 -2.56023053e-02 2.59622544e-01 8.67568016e-01
-1.63672447e-01 1.32779688e-01 9.32633877e-01 -9.97586548e-01
1.43454838e+00 -1.18143693e-01 -2.94508666e-01 8.26801136e-02
-9.20371294e-01 9.18422267e-02 5.12272060e-01 -1.40609890e-01
-1.26219785e+00 -2.59141743e-01 -6.60141170e-01 6.71474099e-01
-5.33781290e-01 5.22277832e-01 -2.49359280e-01 -7.82267898e-02
6.60800636e-01 -8.82542878e-02 5.00262916e-01 -1.79800272e-01
-9.34078097e-02 8.60349178e-01 -1.84952579e-02 -2.71731853e-01
1.66181996e-01 -5.07708013e-01 -4.00291413e-01 -9.63087022e-01
-3.65691215e-01 -7.02182293e-01 -7.61291683e-01 -8.24909031e-01
7.37046540e-01 -1.87075227e-01 -7.76797235e-01 2.56623983e-01
-9.78049934e-01 -4.05747831e-01 -1.43056497e-01 3.39817375e-01
-4.76041138e-01 2.76421696e-01 -2.56058156e-01 -1.30745089e+00
-1.91935197e-01 -1.23663843e+00 4.85809833e-01 5.69153190e-01
-6.89664423e-01 -1.20388222e+00 2.10717097e-01 3.77839595e-01
5.82603514e-01 -1.01755708e-01 8.29454660e-01 -1.00347507e+00
1.68523476e-01 -4.41262960e-01 1.99961782e-01 3.14313084e-01
4.14457262e-01 -2.03603089e-01 -7.24734664e-01 -1.60445526e-01
9.29562449e-02 -1.73940435e-01 -6.43504634e-02 -1.51873156e-02
3.32204729e-01 -1.02171287e-01 7.32490495e-02 -4.15123612e-01
8.66409242e-01 8.16343009e-01 7.10487604e-01 5.44589579e-01
-2.25746911e-02 1.08882451e+00 1.31117046e+00 7.28159487e-01
2.86644846e-01 1.06043518e+00 1.25044733e-01 -2.04532042e-01
4.08655167e-01 8.94768983e-02 3.30809087e-01 6.51234746e-01
-4.63350192e-02 -2.36226812e-01 -6.57931864e-01 1.80965528e-01
-1.78330708e+00 -8.68346632e-01 7.65798539e-02 2.48626375e+00
8.32163870e-01 3.78256857e-01 5.24291456e-01 3.20813507e-01
5.08423090e-01 -1.71550453e-01 1.91130221e-01 -1.06008422e+00
4.78663445e-01 4.16436009e-02 8.14548060e-02 8.65992785e-01
-4.46839303e-01 5.74553311e-01 5.82766533e+00 1.71278387e-01
-1.05308509e+00 -2.97792882e-01 4.19076115e-01 2.33436793e-01
-1.73893590e-02 1.81710329e-02 -4.94437695e-01 3.63894492e-01
9.81231868e-01 -2.96131462e-01 2.85864681e-01 4.67840016e-01
6.73435807e-01 -7.15563715e-01 -9.64148343e-01 4.82943535e-01
-1.18403524e-01 -5.58012724e-01 -3.08343261e-01 -1.18857296e-02
1.14966154e-01 -8.41619313e-01 -2.60333955e-01 4.33052570e-01
-2.35558540e-01 -8.32592607e-01 1.80372015e-01 6.11217201e-01
3.15253347e-01 -1.07325923e+00 1.38023067e+00 6.48017049e-01
-5.36962450e-01 1.50115922e-01 1.49925247e-01 -2.48515174e-01
4.33108360e-01 -3.18242945e-02 -1.46497846e+00 4.58302408e-01
4.94217843e-01 -3.84928137e-01 -4.14051831e-01 1.05133212e+00
1.18150443e-01 3.43839347e-01 -7.39938999e-03 -4.92911190e-01
9.34919193e-02 -3.62493664e-01 6.10032320e-01 1.30772305e+00
9.00320560e-02 4.27807599e-01 2.85830498e-01 2.52735317e-01
6.24836922e-01 7.60367215e-01 -5.39842844e-01 5.09273484e-02
3.75258684e-01 1.15956223e+00 -5.43566883e-01 -3.31354886e-01
5.35853393e-02 1.00507832e+00 -1.12104915e-01 2.28501167e-02
-3.87047529e-01 -6.83811784e-01 2.37769768e-01 9.26603824e-02
-2.29776695e-01 1.11464441e-01 -1.21269353e-01 -3.53442430e-01
-3.15249532e-01 -1.03773391e+00 5.72843030e-02 -3.77316445e-01
-7.12842226e-01 8.32696497e-01 1.78025678e-01 -1.03641939e+00
-6.29142821e-01 -2.64710695e-01 -4.76845860e-01 1.09152806e+00
-9.23504531e-01 -2.57835209e-01 -3.48703980e-01 1.45512506e-01
8.50872874e-01 -1.15289569e-01 1.18584108e+00 3.32306743e-01
-1.97605371e-01 5.33120155e-01 -1.32639661e-01 -4.09409881e-01
1.11448479e+00 -1.55164695e+00 -2.76935041e-01 2.06184000e-01
-5.88029325e-01 4.94002014e-01 1.31366372e+00 -4.51457709e-01
-8.76578748e-01 -2.34037772e-01 1.05897450e+00 1.49209678e-01
3.33471388e-01 5.33756800e-02 -1.11106849e+00 -3.39154482e-01
5.62836945e-01 -1.10588706e+00 9.01038289e-01 1.81300163e-01
4.84294057e-01 -5.82166500e-02 -1.36035180e+00 5.37951469e-01
2.53140807e-01 -3.50996614e-01 -7.22977102e-01 -1.97755452e-02
3.30167383e-01 -6.26768291e-01 -1.07190228e+00 1.33595318e-01
5.32539666e-01 -1.20459211e+00 3.50794375e-01 -1.43819407e-01
9.01563391e-02 4.44027521e-02 2.63016284e-01 -1.73129082e+00
-1.30878866e-01 -5.64136803e-01 5.83708346e-01 1.21562648e+00
5.57082951e-01 -6.41198456e-01 4.39188093e-01 1.05948174e+00
-6.12913743e-02 -6.74809873e-01 -6.74668133e-01 5.55434823e-02
-2.58976758e-01 2.38921538e-01 2.06282198e-01 5.31813502e-01
7.00625896e-01 7.84923434e-01 -3.83043736e-01 -1.28962263e-01
-5.15799271e-03 -2.28365898e-01 1.02812243e+00 -1.36844897e+00
-2.95536160e-01 -5.03151357e-01 -2.99140811e-01 -9.49336410e-01
-3.47588390e-01 -6.41820505e-02 5.34009516e-01 -1.77905118e+00
-2.91933149e-01 -8.61893371e-02 1.54363573e-01 -1.35566108e-02
-3.96101534e-01 -2.72234261e-01 2.89624423e-01 -1.67210191e-01
-4.80012655e-01 2.68279821e-01 1.24701321e+00 2.78302312e-01
-8.95646334e-01 7.06477344e-01 -4.18011367e-01 4.15569186e-01
9.33212876e-01 -6.42106608e-02 -6.11355782e-01 2.74569690e-01
-6.55665621e-02 7.19471276e-01 -4.16356653e-01 -6.92514360e-01
2.27309793e-01 -1.97368562e-01 -1.54187888e-01 -3.85619342e-01
3.83828372e-01 -9.90929544e-01 -1.44291194e-02 5.07436991e-01
-6.52478576e-01 9.98934284e-02 2.64695942e-01 7.38714412e-02
-3.42478842e-01 -7.50272512e-01 9.46768641e-01 2.47608591e-02
-5.72102487e-01 -3.74923974e-01 -9.95782793e-01 -2.49623030e-01
8.23482215e-01 -5.28617501e-01 2.04898983e-01 -8.83192956e-01
-7.47117519e-01 4.09428686e-01 3.06616783e-01 3.40234905e-01
4.72104192e-01 -6.53071165e-01 -3.92721027e-01 -6.51533157e-02
2.18730986e-01 -4.18091595e-01 2.92453825e-01 7.21444011e-01
-4.56772417e-01 4.33267444e-01 -4.36672837e-01 -3.85092765e-01
-1.87203467e+00 -9.22611356e-02 4.93130207e-01 -3.25362235e-01
-2.23930821e-01 3.48476440e-01 -2.64819711e-01 -4.38616395e-01
5.47824740e-01 8.18654150e-02 -1.12991107e+00 4.29297596e-01
4.75492299e-01 3.64889562e-01 2.25719139e-01 -5.63378930e-01
8.04380998e-02 5.61878011e-02 -9.06410441e-02 -4.73396689e-01
8.66464734e-01 -5.67619860e-01 -1.21979997e-01 1.00278986e+00
7.23727286e-01 2.22946033e-01 -7.23926365e-01 -3.65731828e-02
2.61835456e-01 -6.32813454e-01 -2.43867546e-01 -1.18773711e+00
-3.81900012e-01 3.58582914e-01 9.06034648e-01 9.45782900e-01
1.08852684e+00 -3.36876988e-01 2.66241491e-01 5.31793773e-01
3.17729890e-01 -1.45889962e+00 1.82656139e-01 8.33875716e-01
8.17618847e-01 -1.12157643e+00 -2.53149360e-01 -2.84606487e-01
-1.20835578e+00 1.32453632e+00 8.11936557e-01 4.19760078e-01
4.17690307e-01 -5.55241555e-02 5.53520083e-01 -1.80731237e-01
-6.00168943e-01 -9.43953767e-02 6.15287013e-02 3.39913726e-01
1.13037419e+00 -5.48237227e-02 -1.11212921e+00 2.82148600e-01
-2.12221503e-01 -7.53880059e-03 6.13648057e-01 8.50812852e-01
-8.27965498e-01 -1.43531787e+00 -4.55172747e-01 3.46623868e-01
-3.27431411e-01 3.93029422e-01 -7.07181990e-01 5.52484035e-01
-1.10702589e-01 1.42698526e+00 -2.13650391e-01 -1.88193053e-01
1.18771493e+00 2.91947514e-01 2.93817312e-01 -7.92309105e-01
-1.21324265e+00 1.48297340e-01 6.96983278e-01 -2.03879759e-01
-3.73676240e-01 -4.87376451e-01 -1.30409169e+00 -2.09284782e-01
-7.39260852e-01 9.41211998e-01 8.73976707e-01 1.10062528e+00
-4.58173342e-02 6.38976753e-01 9.36051786e-01 -5.41767418e-01
-7.68175066e-01 -1.36113143e+00 -5.12353837e-01 5.26539981e-01
8.26744270e-03 -6.07591152e-01 -1.39655113e-01 -4.31147218e-01] | [13.053155899047852, 7.8845133781433105] |
89a0ec2d-e602-4eb8-b062-3e0ccbac46cd | modeling-irregularly-sampled-clinical-time | 1812.00531 | null | http://arxiv.org/abs/1812.00531v1 | http://arxiv.org/pdf/1812.00531v1.pdf | Modeling Irregularly Sampled Clinical Time Series | While the volume of electronic health records (EHR) data continues to grow,
it remains rare for hospital systems to capture dense physiological data
streams, even in the data-rich intensive care unit setting. Instead, typical
EHR records consist of sparse and irregularly observed multivariate time
series, which are well understood to present particularly challenging problems
for machine learning methods. In this paper, we present a new deep learning
architecture for addressing this problem based on the use of a semi-parametric
interpolation network followed by the application of a prediction network. The
interpolation network allows for information to be shared across multiple
dimensions during the interpolation stage, while any standard deep learning
model can be used for the prediction network. We investigate the performance of
this architecture on the problems of mortality and length of stay prediction. | ['Satya Narayan Shukla', 'Benjamin M. Marlin'] | 2018-12-03 | null | null | null | null | ['length-of-stay-prediction'] | ['medical'] | [ 1.21514350e-01 1.20181151e-01 -1.69861868e-01 -3.99977535e-01
-6.00336850e-01 -1.20159201e-01 1.24490172e-01 6.16003096e-01
-4.08046722e-01 8.68672550e-01 2.44435459e-01 -4.36056316e-01
-1.34844571e-01 -6.61816597e-01 -6.89738691e-01 -6.44573808e-01
-5.14326453e-01 6.26123965e-01 -3.82921994e-01 2.08140120e-01
-3.40657979e-01 3.87811661e-01 -1.01349175e+00 3.94843191e-01
5.95294416e-01 1.15271997e+00 -1.06952704e-01 6.28938317e-01
1.75008029e-01 8.61835718e-01 -3.03027481e-01 7.54760653e-02
3.02909702e-01 -4.72558826e-01 -5.64779282e-01 -3.26290429e-01
-4.92943749e-02 -5.64701915e-01 -5.04453361e-01 5.52350104e-01
8.45782876e-01 4.92849834e-02 4.03866410e-01 -1.08066726e+00
-2.75644809e-01 5.47128677e-01 -6.35624900e-02 2.03424305e-01
9.69925672e-02 1.68654650e-01 6.26610160e-01 -4.72731948e-01
2.70836383e-01 8.30219090e-01 1.03712165e+00 4.75009441e-01
-1.54160357e+00 -5.35163820e-01 -2.25619972e-01 -3.24866064e-02
-1.25583410e+00 -4.78683323e-01 7.18054771e-01 -6.90483749e-01
6.06364787e-01 5.28662279e-02 9.68944907e-01 1.45411384e+00
4.64639783e-01 5.59703827e-01 6.35399997e-01 1.40880272e-01
3.30532938e-01 -3.88021134e-02 3.08637396e-02 2.83962309e-01
1.66279793e-01 1.33846730e-01 -2.35346094e-01 -5.22751808e-01
1.06814158e+00 8.26684356e-01 -2.69823879e-01 -1.56298295e-01
-1.47705710e+00 7.28461146e-01 3.91235262e-01 1.92088857e-01
-9.20087457e-01 9.28278863e-02 6.82261825e-01 3.51507038e-01
4.55349743e-01 5.92508256e-01 -7.81836212e-01 -1.30858660e-01
-1.08822501e+00 4.44101930e-01 9.41310167e-01 7.06281126e-01
2.99640924e-01 -9.05845240e-02 -4.01714206e-01 6.19880140e-01
2.07911685e-01 1.29231095e-01 5.36711574e-01 -9.73123670e-01
2.11471945e-01 6.33309126e-01 1.84124291e-01 -7.83037722e-01
-9.06071424e-01 -3.27640444e-01 -1.51837850e+00 -2.86146015e-01
6.43920422e-01 -4.81465042e-01 -5.63123822e-01 1.67358828e+00
2.02835113e-01 6.70286894e-01 2.36686375e-02 8.14709127e-01
6.84623480e-01 4.66697782e-01 2.38024190e-01 -4.01262939e-01
1.18050492e+00 -4.62453544e-01 -8.77859175e-01 1.67976439e-01
8.90204728e-01 -8.46089795e-02 7.29577661e-01 2.95958549e-01
-1.20211947e+00 -3.60421747e-01 -5.99995852e-01 -2.85012364e-01
-6.83740154e-02 -1.93257913e-01 4.78869736e-01 3.18389945e-02
-7.80956149e-01 7.64000773e-01 -1.25832987e+00 -1.31411910e-01
7.57563710e-01 4.60327238e-01 -1.67187244e-01 -1.46997347e-01
-1.17980969e+00 6.94931686e-01 2.63157248e-01 3.05302471e-01
-7.73100138e-01 -1.20907211e+00 -7.69033372e-01 3.28620404e-01
-1.40989244e-01 -1.00169230e+00 9.42739367e-01 -7.78637886e-01
-1.13588786e+00 6.60777867e-01 -1.89112023e-01 -6.97943389e-01
5.97223639e-01 -1.82409272e-01 -2.52849460e-01 -6.51771482e-03
-4.19274509e-01 2.55436540e-01 5.12185276e-01 -7.19619036e-01
-2.08347321e-01 -4.91654336e-01 -3.03934097e-01 -1.94901690e-01
-2.25107476e-01 -2.02376783e-01 3.28679718e-02 -6.61995888e-01
-2.39331990e-01 -9.11017418e-01 -6.87033474e-01 6.32857382e-02
-3.51268202e-01 3.43325734e-02 5.64013660e-01 -8.70421469e-01
1.34252560e+00 -2.45728230e+00 8.67568031e-02 3.78605276e-02
6.79724991e-01 1.46653056e-01 9.52405483e-02 3.89450133e-01
-2.87186712e-01 -3.13793719e-02 -4.67492670e-01 -4.01173115e-01
-3.27805072e-01 2.55916595e-01 -1.84763983e-01 5.26897073e-01
2.04841018e-01 1.10014534e+00 -9.62846935e-01 -3.61591756e-01
7.47621357e-02 7.47807145e-01 -6.29852414e-01 7.10124075e-01
-4.80362922e-02 1.13430023e+00 -4.89016414e-01 2.93152303e-01
3.02223146e-01 -6.72356308e-01 1.24151491e-01 -3.99411097e-02
5.42663708e-02 3.54817539e-01 -6.54477835e-01 1.81026232e+00
-2.25783169e-01 4.52796638e-01 -7.48976171e-02 -1.26649106e+00
9.87467289e-01 8.52970064e-01 1.26793230e+00 -5.53506792e-01
2.32101247e-01 5.68329096e-02 1.12581447e-01 -7.02652454e-01
-1.01817809e-01 -4.09677744e-01 -3.66219804e-02 3.93812090e-01
-3.11428845e-01 4.17006910e-01 -1.98899835e-01 -3.03247273e-01
1.35593736e+00 -3.68469089e-01 2.52123117e-01 -1.95889130e-01
2.48404399e-01 -1.55561967e-02 8.08295012e-01 5.15517294e-01
-2.13727593e-01 7.68668711e-01 6.23257220e-01 -1.07492602e+00
-1.17270410e+00 -9.51463640e-01 -5.81984282e-01 6.93960428e-01
-2.56588191e-01 -1.28879040e-01 -4.77811754e-01 -3.25199395e-01
2.79817492e-01 2.48455510e-01 -6.30262017e-01 -3.47946703e-01
-5.95701516e-01 -1.12225282e+00 6.33103907e-01 6.42717540e-01
-3.14782821e-02 -1.23076558e+00 -9.29823339e-01 6.03181005e-01
-2.61215717e-01 -9.86754775e-01 -1.71749756e-01 3.41851771e-01
-1.28482533e+00 -9.08676445e-01 -6.86051667e-01 -6.11708581e-01
3.48478943e-01 -4.12126809e-01 1.22062087e+00 1.28293157e-01
-4.42577243e-01 1.41883165e-01 5.19226938e-02 -7.67192662e-01
-3.82656991e-01 2.46332452e-01 1.04346849e-01 2.22747386e-01
5.87386370e-01 -7.72810102e-01 -1.07473922e+00 -1.46516711e-01
-9.34836507e-01 6.08135611e-02 3.24926615e-01 9.54929054e-01
7.30967045e-01 -5.38233638e-01 1.10375905e+00 -9.42851901e-01
7.12029755e-01 -1.09321570e+00 -3.49756509e-01 -8.04451331e-02
-6.14602089e-01 -3.60365361e-02 8.90082777e-01 -4.94774222e-01
-3.50887805e-01 1.60306334e-01 -2.21495107e-01 -6.45228922e-01
-3.01525921e-01 7.34845757e-01 2.35092655e-01 4.40681010e-01
5.08677125e-01 6.84192926e-02 3.19856077e-01 -6.33411050e-01
-1.81662768e-01 6.80477142e-01 4.97305989e-01 -1.98026076e-01
2.36804888e-01 2.66116291e-01 1.44054919e-01 -7.10820377e-01
-7.96100259e-01 -2.94894069e-01 -8.37083578e-01 1.68715075e-01
1.14910865e+00 -1.16188991e+00 -8.62630129e-01 2.08756372e-01
-1.04827285e+00 -6.95431530e-01 -6.31638885e-01 6.32879555e-01
-6.42714977e-01 3.70611325e-02 -8.61830652e-01 -6.28367841e-01
-5.42344630e-01 -9.87312734e-01 9.41893041e-01 -1.71716973e-01
-4.20570016e-01 -1.25486612e+00 4.00039732e-01 -1.74431399e-01
5.98185360e-01 8.07584882e-01 1.11682415e+00 -9.02545393e-01
-3.87682617e-01 -4.57977146e-01 -9.20768902e-02 2.77942598e-01
1.91750571e-01 -3.99132699e-01 -8.83595109e-01 -3.10359985e-01
2.35486880e-01 -2.05437526e-01 5.92776120e-01 8.01401019e-01
1.85666263e+00 -4.12330925e-01 -2.51682848e-01 1.14135313e+00
1.17850828e+00 1.04059577e-01 4.93797660e-01 -6.42306432e-02
8.29692721e-01 4.31268424e-01 -9.56395566e-02 8.60641301e-01
6.86430037e-01 2.49316186e-01 3.56167674e-01 -5.14062941e-01
4.66687024e-01 -8.45234022e-02 -1.69992715e-01 9.97789741e-01
-1.10657230e-01 1.28177047e-01 -1.21003449e+00 4.69486892e-01
-2.06945133e+00 -8.52118492e-01 -2.48928636e-01 2.43733811e+00
9.10840034e-01 -3.20185602e-01 1.93171337e-01 9.86689031e-02
3.25850815e-01 -1.10275462e-01 -9.27321315e-01 -2.66501606e-01
1.05296411e-01 1.73370853e-01 3.46057475e-01 2.28589047e-02
-1.07690012e+00 1.52644724e-01 7.15879202e+00 -2.75788367e-01
-1.25564575e+00 7.05814734e-02 9.97207046e-01 -2.80639142e-01
5.45521528e-02 -4.81258750e-01 -3.14271301e-01 7.69355357e-01
1.61630404e+00 -1.60468176e-01 4.31501925e-01 6.01218998e-01
5.94673395e-01 2.26125404e-01 -1.59122264e+00 1.28071845e+00
-5.31255543e-01 -1.32786059e+00 -2.84027725e-01 1.06509991e-01
5.49638331e-01 3.45133871e-01 -2.70897895e-02 2.53373325e-01
1.01759858e-01 -1.44300056e+00 -4.27769125e-02 7.53112376e-01
8.15340102e-01 -5.34745932e-01 7.69462943e-01 6.73223972e-01
-6.65953398e-01 -3.24348658e-01 -2.67070413e-01 -2.08649546e-01
2.13355035e-01 7.41933227e-01 -7.30056763e-01 1.51728481e-01
7.55832016e-01 1.01696503e+00 -1.30057335e-01 1.23633969e+00
4.41226304e-01 7.20745742e-01 -2.63650894e-01 5.57163775e-01
-9.94777586e-03 -2.13101685e-01 1.25790551e-01 1.16401362e+00
4.21996564e-01 1.72978267e-01 4.04443741e-01 8.75362277e-01
-3.14486742e-01 9.15219560e-02 -8.01525652e-01 2.99720895e-02
1.42161980e-01 1.03258789e+00 -4.59835231e-02 -4.62865025e-01
-5.42626023e-01 6.05288506e-01 3.86415541e-01 4.70825195e-01
-7.13026106e-01 5.32431863e-02 7.98034549e-01 4.32880461e-01
-8.35091770e-02 -2.18217820e-01 -6.54343367e-01 -1.37110209e+00
-1.65021986e-01 -8.51176977e-01 5.66318750e-01 -2.27340311e-01
-1.59464467e+00 4.82802957e-01 -3.23446363e-01 -1.10442805e+00
-5.41762114e-01 -2.55704671e-01 -5.88324368e-01 1.16913438e+00
-1.64893699e+00 -6.32956803e-01 -3.87454212e-01 8.78051460e-01
1.62221864e-01 1.66007876e-01 1.27557003e+00 6.42068386e-01
-7.97307193e-01 4.31814909e-01 4.24838752e-01 3.28659683e-01
6.26734078e-01 -1.11769962e+00 1.86776459e-01 3.05536270e-01
-4.65356052e-01 4.71721232e-01 4.49803323e-01 -3.68184626e-01
-1.52278221e+00 -1.40784454e+00 8.48973036e-01 -4.64288324e-01
5.49764395e-01 -4.69199300e-01 -1.50402689e+00 9.03907061e-01
-1.22187778e-01 6.16267145e-01 1.02564323e+00 4.35215756e-02
-8.86583626e-02 -2.38882095e-01 -1.05203617e+00 1.49654537e-01
6.45488739e-01 -4.52542961e-01 -3.60669374e-01 3.24845105e-01
5.99215388e-01 -3.04914773e-01 -1.57579863e+00 5.30959904e-01
6.85621321e-01 -5.59759796e-01 9.58528876e-01 -9.68823850e-01
6.27017796e-01 8.63633454e-02 1.84884682e-01 -1.24560952e+00
-3.03165615e-01 -6.68993473e-01 -6.18535221e-01 5.85012138e-01
3.30768228e-01 -6.44729018e-01 6.55942678e-01 1.09072363e+00
5.56263700e-02 -9.57971096e-01 -9.41426933e-01 -3.26867461e-01
1.98028266e-01 -1.71247736e-01 7.42687225e-01 1.25528800e+00
1.97322875e-01 3.25557411e-01 -5.06220341e-01 1.33608822e-02
5.03729582e-01 1.56499550e-01 5.41232884e-01 -1.61946499e+00
-4.70149636e-01 -2.31990635e-01 -2.22000122e-01 -8.69890690e-01
-1.45995036e-01 -9.60289359e-01 7.55279465e-03 -1.53966975e+00
1.21508740e-01 -7.69208908e-01 -7.78043568e-01 3.83763611e-01
-2.97041059e-01 -5.40948920e-02 -2.84382328e-02 1.83578640e-01
-3.27702552e-01 5.75031161e-01 1.15268254e+00 9.87930074e-02
-6.76892877e-01 1.94775835e-01 -4.55892265e-01 5.11838555e-01
8.71884942e-01 -5.44223785e-01 -2.74560601e-01 -5.91362000e-01
1.39089838e-01 7.17123270e-01 3.33051831e-01 -8.24619293e-01
4.10548747e-02 2.97327358e-02 7.28252113e-01 -3.73390883e-01
1.46099925e-01 -1.05911851e+00 3.18344623e-01 5.32582760e-01
-5.28766572e-01 2.79305190e-01 1.36952505e-01 5.48160493e-01
-2.94334352e-01 4.82504070e-01 6.86727285e-01 -1.19956627e-01
1.52954459e-01 8.32515597e-01 -3.22761148e-01 1.53234750e-01
9.14026201e-01 1.09535761e-01 9.64359343e-02 -1.67036086e-01
-1.07555783e+00 3.66501540e-01 2.00601399e-01 3.10942173e-01
5.92994869e-01 -1.23334110e+00 -8.37830365e-01 3.90768319e-01
3.66496593e-02 6.28154039e-01 3.03330898e-01 1.30565023e+00
-3.52795660e-01 8.94881561e-02 -2.57927030e-01 -7.36439884e-01
-5.81273437e-01 8.68006349e-01 4.27402705e-01 -5.02233505e-01
-1.18459439e+00 1.99776933e-01 4.40966994e-01 -3.55910361e-01
2.50015259e-01 -6.28672898e-01 -1.67650312e-01 -6.57271743e-02
6.08696461e-01 2.48143569e-01 -6.82904273e-02 -1.28036454e-01
-9.49253961e-02 -1.27125382e-01 2.65963614e-01 4.63991910e-01
1.80734730e+00 2.80669145e-02 -1.99820802e-01 1.00303197e+00
1.37084496e+00 -8.22245359e-01 -1.40800428e+00 -2.55510300e-01
1.36063606e-01 2.79301926e-02 -3.27088907e-02 -5.50035417e-01
-1.00411844e+00 1.17447340e+00 6.04081750e-01 2.52909064e-01
1.17815709e+00 -2.94160217e-01 1.05324078e+00 3.50914925e-01
8.70814919e-02 -5.22879601e-01 -3.49661648e-01 4.05328959e-01
6.94382429e-01 -1.33343136e+00 -3.98413539e-01 1.50729373e-01
-3.59342635e-01 1.02983356e+00 1.76858529e-01 -3.42999548e-01
1.10985601e+00 4.89104301e-01 2.03937870e-02 -6.90082237e-02
-1.04716694e+00 4.06977117e-01 3.25702988e-02 3.15711826e-01
6.55469477e-01 8.79221857e-02 -1.26859277e-01 9.03126776e-01
5.28538600e-02 4.71635669e-01 4.88081992e-01 7.90976346e-01
2.58471575e-02 -7.66399562e-01 -1.04143217e-01 9.04869020e-01
-9.01302397e-01 -2.32513085e-01 1.90373644e-01 2.89149612e-01
-7.92068392e-02 4.73955572e-01 3.49124402e-01 3.06753442e-02
3.61158252e-01 4.71307278e-01 3.97351123e-02 -5.47932208e-01
-8.36699843e-01 -2.90469266e-03 -3.25462937e-01 -7.02333927e-01
-1.58786178e-01 -6.75932586e-01 -1.21233869e+00 -2.17180759e-01
4.51713622e-01 -1.46725342e-01 6.08648002e-01 8.49573731e-01
7.44019806e-01 8.04900110e-01 5.27660072e-01 -6.73606157e-01
-4.77583945e-01 -1.01964200e+00 -6.85975432e-01 6.65192127e-01
9.48940933e-01 -2.24818572e-01 -6.94796741e-02 3.97449046e-01] | [7.944121837615967, 6.1911187171936035] |
84047ce8-9c1a-47bd-a2ae-3a237ab6d5e9 | motion-scenario-decoupling-for-rat-aware | 2305.1831 | null | https://arxiv.org/abs/2305.18310v1 | https://arxiv.org/pdf/2305.18310v1.pdf | Motion-Scenario Decoupling for Rat-Aware Video Position Prediction: Strategy and Benchmark | Recently significant progress has been made in human action recognition and behavior prediction using deep learning techniques, leading to improved vision-based semantic understanding. However, there is still a lack of high-quality motion datasets for small bio-robotics, which presents more challenging scenarios for long-term movement prediction and behavior control based on third-person observation. In this study, we introduce RatPose, a bio-robot motion prediction dataset constructed by considering the influence factors of individuals and environments based on predefined annotation rules. To enhance the robustness of motion prediction against these factors, we propose a Dual-stream Motion-Scenario Decoupling (\textit{DMSD}) framework that effectively separates scenario-oriented and motion-oriented features and designs a scenario contrast loss and motion clustering loss for overall training. With such distinctive architecture, the dual-branch feature flow information is interacted and compensated in a decomposition-then-fusion manner. Moreover, we demonstrate significant performance improvements of the proposed \textit{DMSD} framework on different difficulty-level tasks. We also implement long-term discretized trajectory prediction tasks to verify the generalization ability of the proposed dataset. | ['Nenggan Zheng', 'Risheng Liu', 'Yaohua Liu', 'Jiaxin Gao', 'Xiaofeng Liu'] | 2023-05-17 | null | null | null | null | ['motion-prediction', 'trajectory-prediction', 'action-recognition-in-videos', 'action-recognition'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 6.84313476e-02 -5.10484576e-01 -2.46344060e-01 -2.63320744e-01
-2.45018229e-01 -2.03561425e-01 5.48507869e-01 -1.03123151e-01
-4.57677275e-01 5.07128954e-01 3.91749859e-01 1.44444644e-01
-6.43288255e-01 -4.47334290e-01 -4.87331271e-01 -1.00129688e+00
-1.62977368e-01 2.34803155e-01 4.25580531e-01 -1.80297598e-01
5.76998405e-02 4.41740006e-01 -1.61761618e+00 4.02086288e-01
8.63425195e-01 8.54644239e-01 5.48874855e-01 4.73953485e-01
3.15770000e-01 7.56415367e-01 -2.24675626e-01 -1.09629668e-01
3.11153859e-01 -2.54878640e-01 -6.17280006e-01 2.38819882e-01
1.03575066e-01 -3.94453913e-01 -7.84056962e-01 5.99026799e-01
7.30034649e-01 3.51736128e-01 5.50473630e-01 -1.48445618e+00
-1.75730392e-01 3.63368034e-01 -4.75230873e-01 4.94334213e-02
1.57510072e-01 7.05619693e-01 6.16725743e-01 -4.47885454e-01
7.63216019e-01 1.24772060e+00 6.38665378e-01 7.02252150e-01
-9.07525301e-01 -4.30769086e-01 4.64180619e-01 8.70033979e-01
-1.02327180e+00 -2.31348485e-01 9.08900440e-01 -6.97526097e-01
8.77893567e-01 9.01803002e-02 7.40937114e-01 1.45406044e+00
2.25713894e-01 1.11289573e+00 4.12598968e-01 7.19192550e-02
2.31886610e-01 -3.65067422e-01 6.17450178e-02 6.61654711e-01
1.99389577e-01 -4.01073135e-03 -4.63659585e-01 3.18396747e-01
5.38283587e-01 2.66129941e-01 -2.93204010e-01 -7.00934172e-01
-1.67484820e+00 2.34942809e-01 5.07203758e-01 2.25582510e-01
-4.31880802e-01 3.21849227e-01 5.76317847e-01 -2.06124708e-01
-3.35641541e-02 1.37947733e-03 -5.11444092e-01 -3.33474487e-01
-7.03196526e-01 5.56523919e-01 3.65565777e-01 1.01991606e+00
3.55231673e-01 9.61289927e-02 -4.64306861e-01 6.95645213e-01
1.29837126e-01 4.83109772e-01 7.41419196e-01 -1.04100049e+00
8.09215009e-01 8.53596985e-01 2.97259629e-01 -1.44975996e+00
-9.81078386e-01 -3.91157180e-01 -9.40993249e-01 -6.32077977e-02
4.07016635e-01 -4.60636988e-02 -6.97576582e-01 1.73293340e+00
5.79921544e-01 2.85481066e-01 9.31999311e-02 1.20007050e+00
6.44577205e-01 6.08983636e-01 3.02106678e-01 -9.34459865e-02
1.21594656e+00 -1.21175015e+00 -5.48821807e-01 -3.32611147e-03
1.02288604e+00 -8.55395719e-02 8.80075991e-01 4.26569074e-01
-6.08323157e-01 -7.98471212e-01 -9.79157567e-01 -1.01393446e-01
-1.90091833e-01 6.95088923e-01 6.85172856e-01 4.32218313e-01
-4.57567453e-01 7.93096304e-01 -1.09898174e+00 -5.16535103e-01
5.18950105e-01 3.74426961e-01 -4.36138749e-01 -9.14228186e-02
-9.91798878e-01 6.73589110e-01 3.65650058e-01 4.58909482e-01
-1.10031903e+00 -5.62664390e-01 -7.85833955e-01 -7.76379406e-02
2.61537194e-01 -7.60928273e-01 7.02546775e-01 -4.68681306e-01
-1.44598269e+00 3.35409343e-01 -1.91326707e-03 -3.98923576e-01
8.45329165e-01 -3.33700091e-01 -4.47835296e-01 2.67655879e-01
9.23196375e-02 9.28361475e-01 5.75657248e-01 -1.04535210e+00
-7.96615481e-01 -6.15201056e-01 -2.47358624e-02 4.26579148e-01
-5.33760548e-01 -4.21824306e-01 -6.06189430e-01 -7.03879356e-01
9.07500740e-03 -1.10026085e+00 -4.03232992e-01 1.82951331e-01
-1.32218167e-01 -1.97667494e-01 8.29000652e-01 -7.45858908e-01
1.27409863e+00 -2.05836391e+00 5.60679615e-01 -2.46191636e-01
-5.34217209e-02 4.71900374e-01 -2.97158688e-01 3.52948129e-01
2.73977548e-01 -3.15533757e-01 -2.56054133e-01 -3.27370673e-01
1.81328692e-02 1.76018938e-01 -1.30179241e-01 4.65702146e-01
6.56145215e-02 8.36067796e-01 -9.33628440e-01 -5.29908597e-01
5.81576169e-01 2.93513149e-01 -7.06796646e-01 -2.82418597e-02
-2.67586678e-01 8.46575379e-01 -7.32503355e-01 6.92173541e-01
4.55608219e-01 -4.97753546e-03 5.62371984e-02 -3.69529814e-01
-1.31021380e-01 -3.88930917e-01 -9.87803996e-01 2.13738346e+00
-3.26585084e-01 4.15778667e-01 -1.66408852e-01 -1.26448107e+00
8.15152287e-01 3.94299366e-02 1.15901780e+00 -5.82135022e-01
1.74647838e-01 3.39990743e-02 1.40644759e-02 -1.02189755e+00
3.51773560e-01 2.14369863e-01 -2.48165756e-01 -1.73464522e-01
-8.79131258e-02 3.93128812e-01 1.28448516e-01 -1.79965496e-01
1.23237514e+00 8.26929629e-01 -6.29336983e-02 -1.14603184e-01
8.55603218e-01 4.25306737e-01 8.61286163e-01 4.47848350e-01
-8.31930816e-01 5.57998478e-01 9.73883495e-02 -5.96359968e-01
-9.22832847e-01 -7.36103058e-01 1.24080136e-01 9.21465576e-01
7.83056438e-01 -2.32729420e-01 -7.09039986e-01 -6.32780194e-01
2.53217574e-02 4.71012533e-01 -3.78747195e-01 -4.23842341e-01
-8.78045738e-01 -9.41689014e-01 6.57325506e-01 7.17819571e-01
8.12193751e-01 -1.11067891e+00 -9.63920832e-01 4.31759685e-01
-4.56238270e-01 -1.23983443e+00 -2.16216072e-01 -3.17739546e-01
-7.45113313e-01 -1.05905199e+00 -8.64076197e-01 -8.35989475e-01
4.28078294e-01 4.08579409e-01 2.17029676e-01 -1.66358039e-01
-4.60536778e-01 3.83160859e-01 -5.29139757e-01 3.83969545e-02
9.80975106e-02 -4.09619249e-02 3.38948905e-01 2.55348802e-01
2.42222264e-01 -5.82946062e-01 -1.00452709e+00 6.75124168e-01
-6.22105241e-01 2.19618335e-01 6.48835123e-01 7.69921780e-01
4.33295846e-01 1.79376811e-01 6.01171017e-01 9.74820182e-03
1.20757088e-01 -5.11633158e-01 -2.52539843e-01 3.70461076e-01
-1.55677825e-01 -3.69280837e-02 7.31616318e-01 -6.56985760e-01
-1.25291145e+00 4.35939372e-01 1.83027573e-02 -6.54452145e-01
-4.29732710e-01 1.65411860e-01 -6.63242936e-01 1.65328965e-01
3.26500535e-01 5.09862781e-01 -4.28995350e-03 -3.69554222e-01
4.40305769e-01 6.30635679e-01 6.21202230e-01 -5.45590699e-01
4.48456883e-01 8.21641684e-01 2.47963488e-01 -8.25641632e-01
-3.91393840e-01 -4.50405031e-01 -8.10896277e-01 -6.40871882e-01
1.18696678e+00 -9.81837749e-01 -9.70312297e-01 8.10346425e-01
-1.01692653e+00 -3.95413846e-01 3.58882397e-02 7.83626914e-01
-8.49301100e-01 5.50224066e-01 -3.35422069e-01 -7.56041229e-01
-9.17368010e-02 -1.34059191e+00 1.12948275e+00 -3.97738162e-03
-6.35090023e-02 -6.88264906e-01 -2.17736866e-02 6.56082153e-01
7.46287927e-02 4.78116244e-01 5.49544334e-01 -3.46191913e-01
-5.53444624e-01 -1.72450259e-01 3.65682021e-02 -1.16992462e-02
7.12264329e-02 -2.81826764e-01 -4.98403519e-01 -1.97010398e-01
-8.24968889e-02 -3.19712311e-01 9.22108173e-01 4.69580978e-01
1.29408884e+00 -1.24513827e-01 -7.28709221e-01 6.14533067e-01
1.01523054e+00 3.01138163e-01 6.46720648e-01 4.61866736e-01
1.20420110e+00 1.04224813e+00 9.42765296e-01 7.10004210e-01
6.17980838e-01 9.89026427e-01 2.99740821e-01 3.29230368e-01
-1.44665390e-01 -2.90661275e-01 5.05314708e-01 5.27376354e-01
-1.69759527e-01 -2.35002935e-01 -7.18945861e-01 6.62011862e-01
-2.37380672e+00 -1.23863673e+00 -2.74896026e-01 1.83157670e+00
2.21505761e-01 -1.01888113e-01 2.62892097e-01 1.82156041e-01
7.29336619e-01 1.56058058e-01 -8.04457247e-01 2.79330283e-01
-3.82213434e-03 -6.81868076e-01 3.74198914e-01 -1.36528863e-02
-1.47302866e+00 8.14455688e-01 4.64498281e+00 1.13476586e+00
-1.11201990e+00 -1.43860072e-01 3.22610050e-01 -2.23235413e-01
-2.03170180e-02 -1.50713608e-01 -7.84118354e-01 7.29703367e-01
4.62697446e-01 1.95303768e-01 3.62145118e-02 8.88938963e-01
7.11363137e-01 3.30796614e-02 -1.05343640e+00 1.10343099e+00
-8.86853114e-02 -9.95894790e-01 2.77280718e-01 3.45981009e-02
5.35146832e-01 -3.83934587e-01 -2.47170217e-03 4.31525081e-01
-1.31714627e-01 -6.80867553e-01 9.18051064e-01 6.03482664e-01
2.55380362e-01 -5.32973289e-01 4.56379831e-01 7.14315057e-01
-1.47989798e+00 -6.52292252e-01 -3.51236880e-01 -6.10087514e-02
3.87463599e-01 2.53108531e-01 -3.96771371e-01 1.09907651e+00
6.65502727e-01 1.21363151e+00 -4.69725013e-01 1.09224486e+00
2.35127687e-01 3.71520579e-01 -1.74172267e-01 -1.49813205e-01
3.65876555e-01 -2.00494915e-01 7.22240984e-01 9.79420781e-01
4.84042525e-01 1.80396825e-01 3.19115311e-01 5.32664537e-01
5.50403953e-01 -1.17016628e-01 -4.89193261e-01 1.75536096e-01
2.69395798e-01 1.17984259e+00 -7.68891752e-01 -9.47129503e-02
-2.37822145e-01 1.05052245e+00 1.62115827e-01 2.35817701e-01
-1.13002694e+00 -2.91732159e-02 6.85242653e-01 -2.39914004e-02
4.76369768e-01 -4.51260865e-01 -1.58151820e-01 -1.29765499e+00
2.55654544e-01 -4.79442388e-01 3.00042182e-01 -4.95592356e-01
-9.93452370e-01 1.43041402e-01 2.32220650e-01 -1.93418443e+00
-3.74852307e-02 -7.82658517e-01 -4.40636009e-01 2.38184839e-01
-1.04342723e+00 -1.46854889e+00 -5.14170051e-01 5.91582596e-01
8.18762481e-01 -2.60283172e-01 3.67564023e-01 5.12165308e-01
-9.14964020e-01 3.98104519e-01 1.56875893e-01 -9.31409821e-02
3.83182704e-01 -6.78804278e-01 -4.33729701e-02 7.52113223e-01
-1.86359286e-01 3.05428088e-01 5.20099878e-01 -7.91786790e-01
-1.56619442e+00 -1.52297258e+00 2.32111216e-01 -4.77279902e-01
4.06671137e-01 -1.73235491e-01 -6.53669059e-01 4.33593422e-01
-3.49058717e-01 -3.37456428e-02 3.45212102e-01 -5.27756929e-01
1.09616876e-01 -3.42838585e-01 -1.04286683e+00 8.19068670e-01
1.63252056e+00 2.03488991e-01 -4.61256027e-01 2.88760543e-01
6.27358854e-01 -1.44448295e-01 -6.80508316e-01 7.41866708e-01
8.53978097e-01 -7.08907664e-01 1.08804762e+00 -6.83633029e-01
5.47964215e-01 -6.62378669e-01 -1.81788787e-01 -9.38515544e-01
-4.47734475e-01 -3.45267206e-01 -1.36104435e-01 1.03723061e+00
-2.39003431e-02 -3.70221399e-02 1.00646675e+00 4.55190182e-01
-3.49307030e-01 -8.04186940e-01 -9.84552085e-01 -9.76788223e-01
-4.90271449e-02 -4.37848210e-01 4.91989344e-01 7.33214796e-01
1.61767617e-01 -8.68747234e-02 -7.83971012e-01 1.66065663e-01
5.08656323e-01 6.98420703e-02 9.93162930e-01 -8.43385160e-01
-1.78254381e-01 -4.52595234e-01 -7.97626555e-01 -1.38174450e+00
1.20766059e-01 -6.78191483e-01 2.76787817e-01 -1.68424153e+00
5.51327690e-02 -2.26032779e-01 -3.00197005e-01 2.90516257e-01
-1.54276788e-01 -1.44929245e-01 1.92354128e-01 4.16640699e-01
-1.01465023e+00 1.16167188e+00 1.33466327e+00 -2.61411816e-01
-3.85001302e-01 -4.61402237e-02 -2.84849666e-02 7.15568364e-01
5.96307576e-01 -7.52588883e-02 -7.63495743e-01 -4.81778890e-01
-3.52591962e-01 2.56330371e-01 6.45412207e-01 -1.68137801e+00
4.42988455e-01 -3.57032478e-01 4.25682157e-01 -8.87888074e-01
4.51801062e-01 -7.83898294e-01 2.42328674e-01 7.60855138e-01
-3.83583337e-01 -2.35800698e-01 1.24807686e-01 9.99944329e-01
1.49223972e-02 2.26982459e-01 4.21473175e-01 -4.18999046e-03
-1.27792025e+00 4.55119252e-01 -4.50116068e-01 -3.52381170e-01
1.48377836e+00 -4.76944298e-01 -1.83390751e-01 -1.86799765e-02
-7.23327100e-01 7.02326357e-01 2.60967880e-01 6.62795007e-01
5.89677989e-01 -1.43248498e+00 -5.04858315e-01 -1.30924299e-01
3.47636849e-01 6.64928257e-02 1.00950003e+00 1.11695683e+00
-5.22699952e-01 3.77385795e-01 -5.97995102e-01 -7.07132757e-01
-9.71986115e-01 6.58609986e-01 1.94032833e-01 -2.01208517e-01
-8.84301841e-01 5.42585313e-01 3.43106002e-01 -5.51768780e-01
1.99831188e-01 -2.62708396e-01 -6.33373380e-01 -1.79088816e-01
4.83144879e-01 8.08815300e-01 -3.27779233e-01 -9.46781754e-01
-7.35045195e-01 7.66441047e-01 3.39560211e-01 -1.92796942e-02
1.26305723e+00 -3.75490963e-01 4.07861114e-01 1.56627029e-01
9.21961248e-01 -5.33602953e-01 -1.66303515e+00 2.20087156e-01
8.94454196e-02 -2.79370993e-01 -4.09836650e-01 -6.09685004e-01
-9.96067405e-01 1.01263714e+00 6.81940973e-01 -4.28016096e-01
1.04073632e+00 -3.98518562e-01 9.47380126e-01 6.13523066e-01
7.03185499e-01 -1.17172956e+00 2.88036257e-01 1.54610217e-01
8.70348334e-01 -1.04334402e+00 1.17392298e-02 -3.72339457e-01
-9.81717110e-01 8.65045786e-01 9.13260281e-01 -2.07413156e-02
4.87741292e-01 -2.98999608e-01 -2.35736430e-01 8.84763971e-02
-5.18674135e-01 -1.56453848e-01 2.81287938e-01 8.82613420e-01
-1.91841736e-01 3.52581814e-02 -4.36868578e-01 1.10806894e+00
4.26813699e-02 2.91651934e-01 3.00158530e-01 8.39065015e-01
-4.53645140e-01 -6.99777842e-01 -1.99611172e-01 1.50974497e-01
1.22545131e-01 5.01421392e-01 -2.93660257e-02 7.23958850e-01
5.55257440e-01 9.52656806e-01 -1.55367181e-01 -8.93124223e-01
3.97282660e-01 -2.39741892e-01 5.23355544e-01 -1.19039610e-01
-1.87799603e-01 -6.10916726e-02 7.02331737e-02 -8.31757605e-01
-5.83348215e-01 -7.82784402e-01 -1.51426542e+00 2.46881023e-02
3.08580827e-02 -1.82115585e-01 4.10944074e-01 9.79880452e-01
7.09786057e-01 6.35564327e-01 4.03735459e-01 -1.21248531e+00
-6.27236247e-01 -8.42893362e-01 -3.83971363e-01 7.16571152e-01
8.28429759e-02 -1.09756160e+00 -6.83776662e-02 1.77031398e-01] | [7.561258316040039, -0.05420202016830444] |
c0fb91e4-a464-4246-9218-f763de2e8ead | building-multimodal-ai-chatbots | 2305.03512 | null | https://arxiv.org/abs/2305.03512v1 | https://arxiv.org/pdf/2305.03512v1.pdf | Building Multimodal AI Chatbots | This work aims to create a multimodal AI system that chats with humans and shares relevant photos. While earlier works were limited to dialogues about specific objects or scenes within images, recent works have incorporated images into open-domain dialogues. However, their response generators are unimodal, accepting text input but no image input, thus prone to generating responses contradictory to the images shared in the dialogue. Therefore, this work proposes a complete chatbot system using two multimodal deep learning models: an image retriever that understands texts and a response generator that understands images. The image retriever, implemented by ViT and BERT, selects the most relevant image given the dialogue history and a database of images. The response generator, implemented by ViT and GPT-2/DialoGPT, generates an appropriate response given the dialogue history and the most recently retrieved image. The two models are trained and evaluated on PhotoChat, an open-domain dialogue dataset in which a photo is shared in each session. In automatic evaluation, the proposed image retriever outperforms existing baselines VSE++ and SCAN with Recall@1/5/10 of 0.1/0.3/0.4 and MRR of 0.2 when ranking 1,000 images. The proposed response generator also surpasses the baseline Divter with PPL of 16.9, BLEU-1/2 of 0.13/0.03, and Distinct-1/2 of 0.97/0.86, showing a significant improvement in PPL by -42.8 and BLEU-1/2 by +0.07/0.02. In human evaluation with a Likert scale of 1-5, the complete multimodal chatbot system receives higher image-groundedness of 4.3 and engagingness of 4.3, along with competitive fluency of 4.1, coherence of 3.9, and humanness of 3.1, when compared to other chatbot variants. The source code is available at: https://github.com/minniie/multimodal_chat.git. | ['Min Young Lee'] | 2023-04-21 | null | null | null | null | ['multimodal-deep-learning'] | ['natural-language-processing'] | [ 7.80508965e-02 3.70671600e-01 2.72618979e-01 -4.42264646e-01
-1.16925764e+00 -8.23667705e-01 8.41534436e-01 -2.29451597e-01
-6.62479162e-01 8.94868731e-01 3.28024924e-02 1.00052670e-01
4.12660390e-01 -6.29534185e-01 -5.39561212e-01 -8.00148129e-01
4.72590536e-01 7.48545945e-01 3.68459910e-01 -5.28453231e-01
3.61379653e-01 -2.14701787e-01 -1.28172016e+00 7.16626287e-01
6.58723712e-01 1.05009627e+00 5.51347613e-01 1.04529107e+00
-8.54502618e-02 9.76129770e-01 -9.15592015e-01 -7.84593046e-01
6.37882948e-02 -9.11462545e-01 -1.20541847e+00 9.07678753e-02
3.84790421e-01 -6.53656244e-01 -3.22981924e-01 7.85531104e-01
7.25953877e-01 1.97220564e-01 6.57290637e-01 -1.35614502e+00
-8.58907402e-01 4.12480265e-01 -5.79072833e-01 -1.54412985e-01
5.96181154e-01 6.95904851e-01 9.23108757e-01 -8.49129260e-01
7.97346771e-01 1.32353604e+00 1.50480181e-01 9.30983186e-01
-1.08804619e+00 -4.79711562e-01 -3.60796750e-01 2.10026100e-01
-1.31530333e+00 -3.48535359e-01 3.64250034e-01 -1.76082239e-01
8.74023855e-01 3.23237449e-01 2.95809925e-01 1.32600570e+00
8.85932148e-02 9.88254011e-01 1.21498668e+00 -4.75742787e-01
1.16461702e-02 5.28793216e-01 -3.82281065e-01 6.26131535e-01
-5.02561927e-01 -4.38430548e-01 -5.65416634e-01 1.42422050e-01
7.35896111e-01 -2.14351416e-01 -1.94084764e-01 1.74221143e-01
-1.31430924e+00 9.65459585e-01 5.88933587e-01 1.43309042e-01
-4.24055189e-01 1.09829940e-01 3.88574213e-01 5.31072855e-01
2.01463133e-01 4.36449111e-01 1.79091334e-01 -2.32349679e-01
-6.25034988e-01 2.53071547e-01 9.96809602e-01 9.16516483e-01
7.55808055e-01 -2.93701053e-01 -5.09267449e-01 1.26681006e+00
1.84296906e-01 8.12287033e-01 3.49527806e-01 -1.26258159e+00
3.53728741e-01 6.95989966e-01 2.56310582e-01 -8.59944284e-01
-7.08347932e-02 1.47111297e-01 -8.22801888e-01 1.26707971e-01
5.39855599e-01 -2.83707410e-01 -8.77947807e-01 1.67618763e+00
1.33501619e-01 -7.78608680e-01 3.92691284e-01 1.23887002e+00
1.26810682e+00 1.11207426e+00 2.75064707e-01 7.46274740e-03
1.59256577e+00 -1.26546383e+00 -6.89344406e-01 -2.80681759e-01
3.03222716e-01 -1.17615736e+00 1.41885734e+00 3.53172332e-01
-1.34895945e+00 -5.11165261e-01 -6.47990942e-01 -1.74570262e-01
-1.83199331e-01 3.54728013e-01 5.63917346e-02 3.82522941e-01
-1.43273497e+00 -7.21828490e-02 6.43049367e-03 -8.16353559e-01
4.41310257e-02 1.64056078e-01 -4.14227664e-01 -2.50725716e-01
-1.18995488e+00 1.05562294e+00 2.71553785e-01 -2.74236500e-01
-1.35104775e+00 -1.58076108e-01 -3.52091581e-01 -1.70405075e-01
4.18710649e-01 -6.98225558e-01 1.71866179e+00 -1.22434270e+00
-1.69461167e+00 1.25788033e+00 1.50914371e-01 -4.46452409e-01
8.42274964e-01 -7.96076879e-02 -1.48728311e-01 7.22257853e-01
2.75756240e-01 1.44135320e+00 5.75584829e-01 -1.42352915e+00
-7.63120532e-01 -1.35733739e-01 5.27605534e-01 6.95383251e-01
-7.78074414e-02 -3.97001393e-02 -8.32585454e-01 -9.21651199e-02
-4.71960366e-01 -1.06680524e+00 -8.47002789e-02 -1.85908079e-01
-3.17382962e-01 -2.91867435e-01 6.36919260e-01 -7.19487906e-01
6.42446935e-01 -1.96835577e+00 -1.11434937e-01 -2.85643935e-01
1.30018711e-01 1.89321429e-01 -4.71705616e-01 8.52952421e-01
4.47471410e-01 9.66381002e-03 -2.11793065e-01 -2.50903934e-01
-3.83004583e-02 1.37627557e-01 -5.37367798e-02 1.86279148e-01
2.43153870e-01 9.70886827e-01 -1.03396142e+00 -5.30914783e-01
2.25156114e-01 3.53104621e-01 -1.54140145e-01 6.90892577e-01
-2.89574653e-01 4.08915520e-01 -2.00503722e-01 3.63760471e-01
5.22971392e-01 -3.47929120e-01 9.44387466e-02 8.28565657e-03
-6.73587546e-02 -6.47780299e-02 -7.84441650e-01 1.66973639e+00
-4.93841201e-01 8.11905682e-01 1.59186780e-01 -5.75332224e-01
1.04140139e+00 6.29321516e-01 2.04200149e-01 -1.09235835e+00
2.33061627e-01 1.68145791e-01 -1.03505306e-01 -7.14838386e-01
7.31947720e-01 -7.80807883e-02 -4.14912313e-01 7.24813163e-01
3.19706291e-01 -4.25410151e-01 4.19486582e-01 7.11649179e-01
1.01411474e+00 -5.66706434e-02 2.12448146e-02 8.04940239e-02
5.21676719e-01 3.90198916e-01 -2.99888826e-03 9.53710914e-01
-1.09704860e-01 8.86912823e-01 7.46868551e-01 -2.23167449e-01
-1.23996925e+00 -9.38662767e-01 3.53640527e-01 1.23822796e+00
4.60037917e-01 -4.49156240e-02 -1.02746975e+00 -5.88727117e-01
-5.37019491e-01 7.03050911e-01 -4.77963179e-01 1.96998287e-02
-3.55459064e-01 -2.85223663e-01 7.77663887e-01 1.43221304e-01
1.02354276e+00 -1.52219701e+00 -7.15892851e-01 4.39833896e-03
-1.03466797e+00 -1.14899170e+00 -5.73571146e-01 -3.40828687e-01
-2.58338779e-01 -1.04068696e+00 -1.14885700e+00 -8.80150497e-01
6.30000591e-01 4.60906178e-01 1.26633966e+00 3.25861931e-01
-1.52300343e-01 6.32532835e-01 -7.75396526e-01 -2.46552974e-01
-7.39437759e-01 4.18276675e-02 -4.73894775e-01 9.25320461e-02
6.42900467e-02 1.40008360e-01 -9.53600705e-01 7.64832616e-01
-1.02535629e+00 4.35312450e-01 7.79076159e-01 1.08286619e+00
1.95791155e-01 -6.96810961e-01 5.62274873e-01 -5.46142817e-01
9.27097857e-01 -4.81169105e-01 -2.14362249e-01 4.88402516e-01
-3.62179607e-01 -3.55449587e-01 4.16563511e-01 -4.82762992e-01
-1.38159347e+00 1.52200907e-01 -5.45555772e-03 -1.12869762e-01
-2.84068078e-01 6.38604462e-02 1.18954614e-01 1.65939674e-01
9.30112183e-01 2.94190794e-01 1.85092077e-01 4.66042273e-02
5.41210294e-01 1.03989148e+00 6.97320938e-01 -2.92970300e-01
3.59974682e-01 2.86771804e-01 -7.77792275e-01 -7.12483644e-01
-4.48947132e-01 -5.45207262e-01 -2.59271532e-01 -7.41739690e-01
1.12040329e+00 -9.32443500e-01 -9.40379441e-01 4.91152138e-01
-1.34952438e+00 -6.11296296e-01 -3.29748616e-02 1.43314108e-01
-7.23710060e-01 2.85056323e-01 -7.23483324e-01 -9.46054161e-01
-6.06327415e-01 -1.27313638e+00 9.85867500e-01 5.15374482e-01
-4.13039327e-01 -5.75284660e-01 -2.81772390e-02 9.30712759e-01
4.76256818e-01 9.52946916e-02 2.26720601e-01 -6.22018456e-01
-6.32915974e-01 -1.56934738e-01 -5.20436585e-01 2.99818307e-01
-1.67161912e-01 -2.52862364e-01 -1.04520345e+00 -7.56528974e-02
-1.63545921e-01 -1.03751791e+00 7.17037976e-01 1.72467157e-02
4.27121341e-01 -3.67130965e-01 -1.36582379e-03 -1.99334219e-01
1.23067379e+00 4.75833237e-01 8.12522709e-01 1.41149551e-01
2.01969385e-01 8.79717708e-01 7.92879581e-01 3.86489123e-01
5.44745803e-01 5.61411560e-01 6.73295617e-01 -2.50295520e-01
-6.65364414e-02 -2.12990165e-01 5.29332042e-01 3.61957312e-01
-4.44957316e-02 -6.61414862e-01 -9.04405534e-01 9.01222587e-01
-1.99765384e+00 -9.07479167e-01 -3.69663417e-01 1.90843296e+00
9.22680676e-01 -2.62102365e-01 3.31950605e-01 -4.25793827e-01
8.43586981e-01 1.89209431e-02 -4.29429919e-01 -6.63368583e-01
-2.02602714e-01 -2.12547258e-02 2.70446569e-01 4.25291210e-01
-6.07184708e-01 1.22237074e+00 4.90199804e+00 6.16468549e-01
-9.70352829e-01 2.66440272e-01 9.54942346e-01 -6.95049614e-02
-5.63837774e-02 -7.08269179e-02 -3.39069813e-01 3.74349296e-01
1.05723011e+00 -5.36482669e-02 5.97915053e-01 5.55188358e-01
1.57680884e-01 -6.27634645e-01 -8.51126134e-01 9.86838996e-01
4.11190480e-01 -1.20952344e+00 1.16071455e-01 -1.52055651e-01
6.64744854e-01 -4.66838852e-03 9.74542797e-02 4.73050177e-01
4.36841935e-01 -1.16373956e+00 7.39286840e-01 4.56482917e-01
8.80697668e-01 -4.52452213e-01 1.05498922e+00 4.28439498e-01
-6.88629210e-01 1.51236236e-01 -2.77708262e-01 1.49367794e-01
2.97255099e-01 -1.13420643e-01 -1.44561493e+00 4.70014721e-01
8.87625098e-01 6.46068826e-02 -4.25690740e-01 7.84264863e-01
-4.34518427e-01 3.28401953e-01 -7.72126019e-02 -4.47869241e-01
5.56249797e-01 -9.09411982e-02 2.96445161e-01 1.37698448e+00
1.63557082e-01 4.05856788e-01 5.91886975e-03 8.04762542e-01
-3.26234668e-01 2.42322221e-01 -4.10602152e-01 1.05026588e-01
3.42870742e-01 1.47773492e+00 -6.45728946e-01 -4.64208812e-01
-1.56541139e-01 1.39498794e+00 2.83224303e-02 2.82560498e-01
-1.01112211e+00 -5.39640009e-01 -4.11678925e-02 -1.82938606e-01
1.29650474e-01 2.88718760e-01 9.33602154e-02 -6.19687796e-01
-6.36098608e-02 -1.12176776e+00 5.94963610e-01 -1.33415997e+00
-1.20946348e+00 9.93367970e-01 -4.30661477e-02 -1.09267712e+00
-4.67734993e-01 -2.74802566e-01 -5.34213483e-01 8.83707941e-01
-1.02296031e+00 -1.26134396e+00 -6.19163513e-01 7.02519238e-01
1.05418396e+00 -1.02948777e-01 7.90889740e-01 -2.35131122e-02
-1.84018508e-01 4.96290505e-01 -3.12401831e-01 1.76802978e-01
1.13183129e+00 -1.15768373e+00 1.28945634e-01 3.25661182e-01
-8.92732516e-02 2.35597581e-01 6.17724299e-01 -3.48101526e-01
-1.26233053e+00 -7.19013810e-01 8.95416141e-01 -5.35453498e-01
4.00853664e-01 -2.36758500e-01 -8.14426005e-01 2.69670695e-01
1.17348719e+00 -5.38012743e-01 4.09493834e-01 -4.97051090e-01
-1.32958323e-01 -8.54189694e-02 -1.46705437e+00 6.65347993e-01
5.70569754e-01 -3.58676761e-01 -3.68350118e-01 4.59268361e-01
5.07009268e-01 -5.15635073e-01 -7.95905054e-01 -6.40782863e-02
5.72404385e-01 -1.04828680e+00 7.64607131e-01 -1.26507476e-01
7.00764179e-01 -8.24449360e-02 -1.81245893e-01 -1.08561313e+00
-4.93866056e-02 -6.87367618e-01 6.74063027e-01 1.21521401e+00
6.71052694e-01 -3.01957726e-01 4.07727182e-01 5.95265567e-01
2.35940013e-02 -2.84857243e-01 -8.11594069e-01 -2.82013744e-01
-5.96692711e-02 1.77584644e-02 1.40057251e-01 7.68935323e-01
3.38260308e-02 8.43822360e-01 -5.88765740e-01 -2.11513296e-01
2.55578458e-01 -4.02354896e-02 1.18060720e+00 -7.63459265e-01
-3.21747512e-02 -4.25995886e-01 1.17610976e-01 -1.09625673e+00
-2.65459597e-01 -5.40074587e-01 3.93712670e-01 -1.90877581e+00
5.49995005e-01 1.63070392e-02 3.36940982e-03 6.85762107e-01
2.06729118e-02 6.28331244e-01 4.67950165e-01 4.71634567e-01
-9.83179271e-01 4.08808231e-01 1.37511563e+00 -3.26852441e-01
-2.07296327e-01 -2.87208587e-01 -6.66485250e-01 4.08445537e-01
1.29288578e+00 -2.45483190e-01 -3.63081127e-01 -4.10968155e-01
-7.55939586e-03 4.14129287e-01 5.54594636e-01 -7.92667389e-01
2.39587754e-01 -1.60277814e-01 1.84216753e-01 -4.81505126e-01
6.03956640e-01 -4.60056007e-01 1.23470187e-01 3.76365542e-01
-5.69753766e-01 1.63950145e-01 8.53467956e-02 4.32211488e-01
-3.17490846e-01 -3.26491445e-01 7.76091218e-01 -7.08293378e-01
-9.23696160e-01 -2.41038412e-01 -5.37828028e-01 1.34268746e-01
1.10408962e+00 -2.79157370e-01 -6.07851863e-01 -1.12175667e+00
-4.85298008e-01 5.29983938e-01 4.46352214e-01 6.19573951e-01
8.60628784e-01 -1.09656525e+00 -1.06165349e+00 -2.85850346e-01
3.43890101e-01 -1.87319800e-01 5.06929636e-01 9.38578188e-01
-6.81938410e-01 2.63699979e-01 -4.15028870e-01 -7.88744986e-01
-1.43212914e+00 -9.41021293e-02 1.77056849e-01 -1.81456700e-01
-2.87955344e-01 7.65157461e-01 4.09244895e-01 -3.93286049e-01
1.61472693e-01 4.29632187e-01 -1.02053292e-01 4.26541530e-02
5.16912103e-01 1.15621969e-01 -1.82978079e-01 -7.28681445e-01
-1.86261788e-01 1.27858981e-01 -6.43961057e-02 -7.61308730e-01
1.16879916e+00 -2.75347471e-01 -1.10195518e-01 3.20494086e-01
9.47445571e-01 -3.30343604e-01 -1.19618189e+00 -4.23996635e-02
-3.12988251e-01 -3.82997036e-01 -4.95595783e-01 -1.48573136e+00
-7.59877384e-01 7.40538538e-01 6.11120045e-01 3.94925475e-01
1.22900271e+00 2.70817697e-01 8.29182327e-01 3.57045025e-01
1.54817089e-01 -1.18010223e+00 7.54182875e-01 6.22620940e-01
1.38117480e+00 -1.43164396e+00 -3.69859666e-01 1.11818470e-01
-1.58491683e+00 8.67371738e-01 9.29765463e-01 1.11155540e-01
-3.52007419e-01 -2.43580714e-01 4.75237161e-01 -1.83645949e-01
-1.03694940e+00 -2.46982947e-01 8.23621824e-03 5.33360183e-01
4.78590727e-01 5.29464595e-02 -2.86511272e-01 1.99230060e-01
-1.87410131e-01 -2.44669542e-01 6.38465881e-01 7.28675187e-01
-5.04559040e-01 -9.77565289e-01 -4.87083405e-01 -7.46459588e-02
-2.78585136e-01 7.05808774e-02 -9.45703983e-01 9.87054348e-01
-3.33386004e-01 1.49303341e+00 1.73080415e-01 -3.45408171e-01
4.24365491e-01 8.24607611e-02 2.60156751e-01 -4.24501240e-01
-9.18682337e-01 1.37468427e-01 4.15963411e-01 -4.26269591e-01
-5.08092523e-01 -2.43580773e-01 -1.16948807e+00 -3.51151496e-01
-2.79408932e-01 3.19519401e-01 6.79590225e-01 4.63706136e-01
2.92029917e-01 4.92894426e-02 6.36842608e-01 -6.30865574e-01
-3.11915755e-01 -1.22232854e+00 -1.53778583e-01 6.81769788e-01
3.68024297e-02 -9.60665271e-02 -5.15484251e-02 3.08959603e-01] | [10.97941780090332, 1.3535040616989136] |
b25b0df0-030a-4fce-9b6c-c1f5165bd615 | doc3-deep-one-class-classification-using | 2105.07636 | null | https://arxiv.org/abs/2105.07636v2 | https://arxiv.org/pdf/2105.07636v2.pdf | DOC3-Deep One Class Classification using Contradictions | This paper introduces the notion of learning from contradictions (a.k.a Universum learning) for deep one class classification problems. We formalize this notion for the widely adopted one class large-margin loss, and propose the Deep One Class Classification using Contradictions (DOC3) algorithm. We show that learning from contradictions incurs lower generalization error by comparing the Empirical Rademacher Complexity (ERC) of DOC3 against its traditional inductive learning counterpart. Our empirical results demonstrate the efficacy of DOC3 compared to popular baseline algorithms on several real-life data sets. | ['Bernardo Gonzalez Torres', 'Sauptik Dhar'] | 2021-05-17 | null | null | null | null | ['one-class-classification'] | ['miscellaneous'] | [-2.81498700e-01 3.71208131e-01 -3.85152221e-01 -5.85374475e-01
-1.25144160e+00 -3.80411565e-01 3.77545863e-01 5.02157629e-01
-7.08354175e-01 1.07626235e+00 -2.51269221e-01 -8.02811444e-01
-5.75233161e-01 -8.88440788e-01 -1.02688611e+00 -7.58242428e-01
-4.94890392e-01 4.81509358e-01 -8.93990919e-02 -1.63916305e-01
1.90668553e-01 2.73564488e-01 -1.44856274e+00 4.94067013e-01
9.65033770e-01 1.43035424e+00 -8.86939526e-01 4.95754719e-01
3.40062007e-02 1.09026361e+00 -6.88833237e-01 -7.55707324e-01
4.13232654e-01 -1.59678206e-01 -1.26052475e+00 -3.87925953e-01
6.56095922e-01 -3.75802457e-01 1.29175603e-01 9.08178329e-01
3.61287355e-01 -8.22291747e-02 1.18938947e+00 -1.89153242e+00
-3.60300362e-01 8.83422971e-01 -4.96034473e-01 4.66520458e-01
-9.02365148e-02 -4.62890327e-01 1.52830732e+00 -1.02959526e+00
3.16644132e-01 1.43814480e+00 1.14783943e+00 4.50432122e-01
-1.22712064e+00 -9.40256238e-01 6.27167597e-02 5.35599530e-01
-1.30218136e+00 -4.46945289e-03 6.42460763e-01 -3.86119276e-01
9.11179900e-01 -4.85474057e-02 2.10468262e-01 1.31631887e+00
3.86913925e-01 1.28775418e+00 1.20980537e+00 -6.00211561e-01
5.16470671e-01 1.84525579e-01 6.77721620e-01 9.72225130e-01
7.56474555e-01 1.17169149e-01 -3.08531553e-01 -1.77756250e-01
-3.68142314e-02 -2.29379132e-01 2.97693517e-02 -3.80858272e-01
-3.47342908e-01 1.22415268e+00 5.75847566e-01 1.56211868e-01
1.09802090e-01 3.77937853e-01 6.59879804e-01 9.50392425e-01
5.16137958e-01 1.00708887e-01 -8.77044439e-01 2.59783059e-01
-5.68412721e-01 2.99408704e-01 9.69174981e-01 9.30827916e-01
5.94872832e-01 -2.51858860e-01 3.51627648e-01 4.43594545e-01
1.56240821e-01 2.45779917e-01 6.87401891e-01 -7.22675264e-01
6.66784406e-01 4.89074826e-01 -1.06293140e-02 -4.54619706e-01
-6.64419055e-01 -6.34464085e-01 -8.39152098e-01 3.82050663e-01
5.01188040e-01 -1.40077636e-01 -5.26954412e-01 1.90742540e+00
2.52637774e-01 -8.43362212e-02 4.44235653e-01 3.88182491e-01
4.60674703e-01 2.72182614e-01 4.05601300e-02 -1.84945956e-01
8.09465528e-01 -7.13507771e-01 -4.68318224e-01 1.86605573e-01
1.26968372e+00 -1.52520388e-01 1.10789478e+00 9.70465720e-01
-8.10415924e-01 -2.75392592e-01 -1.36251271e+00 -3.85408849e-01
-8.24952245e-01 -2.89185226e-01 7.57332444e-01 9.69458699e-01
-6.19067550e-01 7.95971751e-01 -4.69222695e-01 1.49654776e-01
9.58221018e-01 3.93780589e-01 -3.28515142e-01 3.93326171e-02
-1.55748618e+00 9.79506135e-01 5.32820404e-01 -1.10048935e-01
-8.59952211e-01 -5.61365902e-01 -9.62151825e-01 -1.90242492e-02
2.75719225e-01 -5.15945733e-01 1.36605918e+00 -8.71255815e-01
-1.03469074e+00 1.14816999e+00 2.91006744e-01 -9.38627183e-01
1.06247330e+00 -5.26355803e-01 -1.60258042e-03 -1.22474119e-01
-8.32646899e-03 2.54677743e-01 6.17462277e-01 -1.31288803e+00
-9.14171815e-01 -3.61478686e-01 2.63794422e-01 -3.03991139e-01
-5.74784398e-01 -5.62892675e-01 7.31568575e-01 -5.27628124e-01
1.57888919e-01 -6.44303143e-01 1.75624326e-01 1.75293595e-01
-3.25362593e-01 -1.12063468e+00 7.10071623e-01 3.93238245e-03
1.04501617e+00 -1.93628347e+00 -1.99508429e-01 2.08780885e-01
3.11002105e-01 1.51614830e-01 5.42240813e-02 3.37451175e-02
-4.93412584e-01 3.86597157e-01 -3.71794492e-01 -3.07634503e-01
4.03933734e-01 3.82616431e-01 -4.31575030e-01 6.28548503e-01
1.17265783e-01 8.53352010e-01 -8.28892231e-01 -6.97490394e-01
-2.32743979e-01 -1.07885629e-03 -6.14511669e-01 -2.29968488e-01
2.06640810e-02 -3.82453710e-01 -1.79319575e-01 6.85181916e-01
7.60316491e-01 -4.02290553e-01 1.66968971e-01 1.12286687e-01
4.50093895e-01 3.64734113e-01 -1.05783951e+00 1.12845874e+00
-5.50965011e-01 6.63141549e-01 -6.22576118e-01 -1.47936833e+00
6.35224164e-01 1.77538842e-01 6.01384304e-02 -5.89063883e-01
4.18736190e-01 5.81954718e-01 -4.08364944e-02 -4.75319386e-01
6.58856332e-02 -6.37236476e-01 -3.55359465e-01 2.67979980e-01
1.27404347e-01 1.94532514e-01 2.14751847e-02 2.05076188e-01
9.31374431e-01 3.67013402e-02 4.76309419e-01 -6.00780129e-01
4.69705135e-01 -5.76108098e-02 5.64899623e-01 9.86086309e-01
-5.30923367e-01 -2.30225604e-02 9.49817777e-01 -7.68151939e-01
-7.43956208e-01 -1.35182250e+00 -7.01983631e-01 1.00004041e+00
1.18743032e-01 -2.86085367e-01 -5.86718440e-01 -1.47594547e+00
5.20445406e-01 7.20000327e-01 -9.78638530e-01 -4.20804799e-01
-4.00903225e-01 -8.92695546e-01 8.63003433e-01 7.63284266e-01
4.27642882e-01 -6.13786042e-01 -5.60231149e-01 7.39751733e-04
-2.19844759e-01 -9.02953207e-01 4.36779350e-01 9.09542203e-01
-1.07991326e+00 -1.53489280e+00 -2.68971384e-01 -9.28755760e-01
4.74806577e-01 -3.51653218e-01 1.27970874e+00 2.06692666e-02
-2.46858582e-01 1.16584599e-01 -2.16341197e-01 -6.45935237e-01
-3.41596097e-01 8.53881091e-02 5.74936688e-01 -3.81034821e-01
6.33929133e-01 -3.95959973e-01 -4.21543092e-01 -6.27598017e-02
-8.32314312e-01 -3.92332792e-01 4.50867683e-01 9.68362153e-01
3.91536802e-01 5.38519979e-01 1.06686318e+00 -1.03010261e+00
5.99413097e-01 -8.30668867e-01 -7.09642351e-01 2.88288116e-01
-1.15384614e+00 2.85199165e-01 8.42752337e-01 -3.67652357e-01
-6.67128861e-01 -5.03633201e-01 3.21744755e-02 -2.43452460e-01
2.61306375e-01 2.59011477e-01 -6.21674284e-02 1.06064394e-01
7.97822714e-01 -1.68705314e-01 -5.41801035e-01 -3.56549799e-01
2.35616699e-01 7.55476773e-01 4.34035987e-01 -8.54304433e-01
5.95872998e-01 5.66658676e-01 2.76993036e-01 -3.37225288e-01
-1.40594220e+00 -1.12424888e-01 -7.57755458e-01 -2.71252114e-02
3.71033818e-01 -8.13059986e-01 -1.13028908e+00 3.37941885e-01
-9.98064220e-01 -3.56582910e-01 -4.10841972e-01 3.44820976e-01
-6.11330271e-01 2.53415078e-01 -6.86696768e-01 -9.33456004e-01
-3.87523979e-01 -5.08258700e-01 7.02471614e-01 -1.86316371e-01
-1.74378619e-01 -1.25745153e+00 -3.71197350e-02 2.69870520e-01
-3.10182035e-01 7.26899683e-01 1.35943925e+00 -1.07258070e+00
-9.21649635e-02 -2.88804919e-01 -1.69367656e-01 8.08504045e-01
-1.77128837e-01 -3.40596318e-01 -1.06627822e+00 -2.10644528e-01
1.23723544e-01 -1.11558545e+00 1.27475917e+00 2.47039288e-01
1.45412064e+00 -4.14050519e-01 8.07985291e-02 4.38625574e-01
1.76965964e+00 -1.24472328e-01 5.20333767e-01 7.57528067e-01
4.15200740e-01 5.81810772e-01 7.03046083e-01 6.55905724e-01
5.34342349e-01 -1.30822659e-01 4.68130708e-01 3.26758139e-02
2.77793944e-01 -1.12350017e-01 2.72874534e-01 3.84447873e-01
3.71679008e-01 -1.80330768e-01 -1.07437539e+00 6.29164875e-01
-1.79009497e+00 -8.26210976e-01 -2.84996957e-01 1.92791724e+00
1.14681304e+00 7.84345031e-01 4.38810885e-02 1.06821847e+00
2.83991277e-01 -4.38611895e-01 -5.96772790e-01 -7.02512026e-01
-3.49692851e-01 3.48634541e-01 3.65578830e-01 5.75916767e-01
-1.28010750e+00 7.86386371e-01 6.66135359e+00 9.11501169e-01
-6.92238510e-01 1.22978061e-01 9.33197677e-01 -2.01430488e-02
-7.06546381e-02 -2.48641804e-01 -8.48116934e-01 2.21889436e-01
8.95798147e-01 -2.26852059e-01 -1.62902966e-01 1.17750192e+00
-3.62469107e-01 -6.17649741e-02 -1.71903694e+00 8.60447168e-01
-3.47918272e-02 -9.78638470e-01 6.37943000e-02 -1.84909180e-02
7.99962521e-01 -1.55054599e-01 2.78269351e-01 7.11585581e-01
4.35106456e-01 -1.02254450e+00 8.58374178e-01 2.71028113e-02
6.46308601e-01 -1.08143878e+00 1.03709710e+00 3.96709919e-01
-6.06903017e-01 -5.33818543e-01 -3.59625131e-01 -1.37162000e-01
-5.94212353e-01 8.40525270e-01 -7.59888291e-01 5.68642795e-01
9.43167925e-01 3.94292116e-01 -4.46965367e-01 6.57444060e-01
-2.30457336e-01 5.80798090e-01 -5.47751606e-01 -5.95965981e-02
5.29909968e-01 3.66072148e-01 1.81861445e-02 1.30812919e+00
-3.10832560e-01 4.67360355e-02 -1.51961759e-01 4.90790546e-01
-6.79985642e-01 -9.15625468e-02 -6.16638422e-01 3.93230796e-01
2.98259258e-01 8.35841000e-01 -5.71053684e-01 -6.44725025e-01
-1.34844288e-01 7.13323593e-01 5.43098390e-01 -3.83035205e-02
-1.04546535e+00 -7.29851067e-01 3.30453545e-01 -2.57001609e-01
3.13495666e-01 2.18849197e-01 -5.60319662e-01 -1.10687387e+00
2.29318812e-01 -7.81253338e-01 1.12282920e+00 -2.30461866e-01
-1.74528980e+00 3.59738469e-01 1.34880558e-01 -1.17262900e+00
5.68527170e-02 -1.19771814e+00 -2.92637885e-01 2.16945425e-01
-2.29692173e+00 -8.35453987e-01 -1.06922641e-01 6.84350014e-01
3.84453863e-01 -9.85846892e-02 6.94452643e-01 2.05626100e-01
-5.33815086e-01 1.07737494e+00 4.95706290e-01 2.20139608e-01
5.31831026e-01 -1.71213913e+00 4.76280460e-03 4.24218655e-01
-2.01153774e-02 3.99997115e-01 5.52895308e-01 -1.57567129e-01
-8.93458962e-01 -1.00492549e+00 1.16982245e+00 -6.74821556e-01
9.16043818e-01 -3.23970497e-01 -8.38874757e-01 8.62098455e-01
-3.07941377e-01 3.29262972e-01 9.58628118e-01 3.30486357e-01
-1.06254029e+00 -3.92575473e-01 -1.52318954e+00 2.83292145e-01
9.81301308e-01 -4.27491874e-01 -1.07650542e+00 5.16095340e-01
4.11826044e-01 -6.15764260e-02 -8.95050585e-01 6.36822641e-01
8.21892381e-01 -1.03532147e+00 8.73434186e-01 -1.23495793e+00
8.15531671e-01 2.55159974e-01 -3.83698285e-01 -8.66882563e-01
2.18540225e-02 -5.36231339e-01 -5.04710197e-01 8.87681663e-01
3.65635902e-01 -8.15038681e-01 6.40418887e-01 3.10763836e-01
2.13299125e-01 -1.08770645e+00 -1.27819765e+00 -1.40995646e+00
1.20128238e+00 -5.66592753e-01 3.00360352e-01 1.23992026e+00
1.29185766e-01 4.70389068e-01 1.40272658e-02 1.33777469e-01
1.06425321e+00 2.62769312e-01 4.12531614e-01 -1.68689430e+00
1.81714445e-02 -4.35170561e-01 -6.03827953e-01 -4.85821575e-01
5.56665599e-01 -1.30874062e+00 -1.79222777e-01 -1.13830781e+00
1.91694185e-01 -5.75373471e-01 -7.78618336e-01 6.72745645e-01
-5.56786098e-02 1.70055524e-01 -2.35702381e-01 -3.03847063e-02
-1.00358546e+00 6.99733377e-01 7.01836765e-01 -1.24211796e-01
2.20576659e-01 -1.38617322e-01 -7.87195385e-01 1.17445314e+00
9.54447627e-01 -9.75098848e-01 -2.78033465e-01 -2.74748385e-01
5.38990319e-01 -3.43832672e-01 4.14490461e-01 -1.03286994e+00
3.36475708e-02 1.52158216e-01 5.79225481e-01 -6.77370667e-01
-2.43932083e-01 -8.06424439e-01 -8.46992493e-01 1.07419193e+00
-6.91012025e-01 -1.75726980e-01 2.57068723e-01 6.98167145e-01
1.81053236e-01 -2.87291318e-01 1.25024593e+00 2.46308312e-01
-3.72928828e-01 -1.26034319e-02 -1.79866813e-02 7.12627053e-01
1.04706705e+00 -8.58308375e-02 -4.05761957e-01 2.55058240e-02
-7.00936317e-01 4.56662297e-01 -1.80734590e-01 1.92990094e-01
6.93967879e-01 -1.27602077e+00 -6.97575748e-01 1.46189062e-02
2.65990347e-01 9.92690325e-02 -2.83457875e-01 7.46015787e-01
-4.21630919e-01 4.03179914e-01 -9.06710979e-04 -4.52624857e-01
-1.33776045e+00 5.82598448e-01 6.52521312e-01 -4.84336585e-01
-7.21672058e-01 1.11278093e+00 9.48414579e-02 -4.94507551e-01
6.29551172e-01 -5.24389625e-01 1.52666822e-01 1.53486952e-01
3.02515090e-01 6.72280252e-01 3.06578845e-01 4.68911463e-03
-5.30366600e-01 2.56246090e-01 -3.74663144e-01 -1.09407231e-02
1.32205617e+00 2.09068775e-01 8.87009799e-02 7.17703640e-01
1.72245097e+00 -3.19134533e-01 -1.01968002e+00 -4.03382719e-01
6.23086274e-01 -1.25192851e-01 -9.57414359e-02 -9.13260818e-01
-6.49977744e-01 9.92944777e-01 7.77039468e-01 2.66462237e-01
9.44658220e-01 1.35921864e-02 7.59630442e-01 1.22865438e+00
4.36664820e-01 -1.30099249e+00 5.68054132e-02 5.98724544e-01
7.16625392e-01 -1.51226890e+00 1.93252087e-01 -1.18124550e-02
-3.07956845e-01 1.26180100e+00 5.86786747e-01 -2.94799536e-01
1.01142156e+00 3.33147973e-01 2.25748122e-02 -4.03439216e-02
-8.62630010e-01 -1.21654965e-01 -1.31014615e-01 3.08237374e-01
1.16817050e-01 -1.31811380e-01 -3.29428047e-01 9.92609262e-01
-3.71852726e-01 1.13247652e-02 3.90955985e-01 9.78376091e-01
-3.49210888e-01 -7.91819394e-01 -6.60098791e-02 2.44252130e-01
-6.85849071e-01 -9.46848616e-02 -5.08277297e-01 1.17747879e+00
3.00087571e-01 8.58454764e-01 4.63330187e-02 -2.33596280e-01
1.81457326e-01 2.72836179e-01 6.96113527e-01 -1.67025670e-01
-3.82560462e-01 -7.00865507e-01 -9.78019312e-02 -2.45496213e-01
-3.39417547e-01 -4.66549337e-01 -1.49800801e+00 -5.19089520e-01
-5.42922795e-01 1.29910544e-01 4.73653078e-01 1.18194449e+00
-8.64769220e-02 3.20126057e-01 1.01415277e+00 -1.49623379e-01
-1.18042421e+00 -5.72907567e-01 -6.18988276e-01 4.58595395e-01
7.84533560e-01 -7.88696885e-01 -1.07269955e+00 -1.63474381e-01] | [8.80726432800293, 4.035182952880859] |
833e366a-7fac-4648-8584-d4a65695990e | k-core-based-temporal-graph-convolutional | 2003.09902 | null | https://arxiv.org/abs/2003.09902v4 | https://arxiv.org/pdf/2003.09902v4.pdf | K-Core based Temporal Graph Convolutional Network for Dynamic Graphs | Graph representation learning is a fundamental task in various applications that strives to learn low-dimensional embeddings for nodes that can preserve graph topology information. However, many existing methods focus on static graphs while ignoring evolving graph patterns. Inspired by the success of graph convolutional networks(GCNs) in static graph embedding, we propose a novel k-core based temporal graph convolutional network, the CTGCN, to learn node representations for dynamic graphs. In contrast to previous dynamic graph embedding methods, CTGCN can preserve both local connective proximity and global structural similarity while simultaneously capturing graph dynamics. In the proposed framework, the traditional graph convolution is generalized into two phases, feature transformation and feature aggregation, which gives the CTGCN more flexibility and enables the CTGCN to learn connective and structural information under the same framework. Experimental results on 7 real-world graphs demonstrate that the CTGCN outperforms existing state-of-the-art graph embedding methods in several tasks, including link prediction and structural role classification. The source code of this work can be obtained from \url{https://github.com/jhljx/CTGCN}. | ['You Song', 'Jingxin Liu', 'Chang Yin', 'Chang Xu', 'Weiqiang Wu'] | 2020-03-22 | null | null | null | null | ['dynamic-graph-embedding'] | ['graphs'] | [-2.82775432e-01 3.08102190e-01 -5.57087779e-01 -2.36730743e-02
2.72736222e-01 -5.33512235e-01 6.26637161e-01 3.50311399e-01
1.35653719e-01 2.69033700e-01 3.34062189e-01 -4.36925948e-01
-3.02854478e-01 -1.15770161e+00 -2.12499157e-01 -7.56241262e-01
-5.41682959e-01 2.22869962e-01 3.49265933e-01 -3.55259866e-01
-1.02088399e-01 4.79380190e-01 -9.20929313e-01 -1.01049930e-01
6.67253017e-01 6.43604517e-01 -7.74721801e-02 5.87349057e-01
-1.94615841e-01 7.18285441e-01 -1.24969214e-01 -4.27297592e-01
1.08514555e-01 -2.36834392e-01 -6.64294004e-01 -2.70415217e-01
2.95296162e-01 -1.13703378e-01 -1.40594614e+00 1.11441755e+00
4.05774295e-01 2.38304004e-01 3.73476326e-01 -1.59816432e+00
-1.29528975e+00 7.46010661e-01 -3.02333742e-01 5.66784024e-01
3.05196017e-01 7.38429353e-02 1.53236842e+00 -5.15893459e-01
7.07981765e-01 1.35950065e+00 7.21117496e-01 4.56732988e-01
-1.25632024e+00 -6.15208328e-01 5.07999539e-01 3.47127736e-01
-1.38013244e+00 2.46343054e-02 1.18787909e+00 -3.45989227e-01
9.04432118e-01 2.52942950e-01 1.06818402e+00 1.08574069e+00
3.04192483e-01 6.15626931e-01 4.70071167e-01 -1.30767629e-01
-2.26162881e-01 -4.92536753e-01 2.18742833e-01 1.07695508e+00
4.89386052e-01 1.59727633e-01 -2.72636503e-01 -1.58714667e-01
8.97979617e-01 4.71292406e-01 -4.45425242e-01 -7.33316839e-01
-1.13565171e+00 9.37876880e-01 1.17786992e+00 7.07563579e-01
-8.24027956e-02 7.48266399e-01 7.04789042e-01 6.29487395e-01
6.70201719e-01 1.19846798e-01 -1.00324102e-01 8.78406242e-02
-1.74550965e-01 -3.02732270e-02 7.58611858e-01 8.22244883e-01
7.87563980e-01 7.35645294e-02 -1.22106105e-01 6.00823998e-01
3.74328852e-01 1.07975453e-01 4.25641030e-01 -3.26094806e-01
3.39126796e-01 1.24741590e+00 -5.84847927e-01 -1.67513883e+00
-2.91124076e-01 -4.55463380e-01 -9.73053038e-01 -3.34043682e-01
-8.55740681e-02 2.74644077e-01 -7.85949230e-01 1.65517926e+00
3.87686938e-01 5.73632181e-01 -2.33165368e-01 6.76046431e-01
1.15716922e+00 5.31971157e-01 -9.22070518e-02 1.59779236e-01
9.74265158e-01 -9.58214223e-01 -7.81332076e-01 -5.74836172e-02
8.40250492e-01 -2.36713871e-01 9.52389002e-01 -3.09556037e-01
-5.83860576e-01 -4.22957778e-01 -1.03517234e+00 -1.68653771e-01
-6.08636796e-01 -3.36374819e-01 1.18417847e+00 3.08725297e-01
-1.40436292e+00 6.95050359e-01 -1.07385588e+00 -7.26318419e-01
4.79214638e-01 3.20905149e-01 -5.76244533e-01 -3.54862809e-01
-1.31361890e+00 3.58578384e-01 5.96175373e-01 1.78202733e-01
-8.11505020e-01 -6.24212205e-01 -1.16574371e+00 3.83068085e-01
4.80767012e-01 -5.08652210e-01 6.63964093e-01 -5.67341030e-01
-1.10479605e+00 6.47059858e-01 1.87149644e-01 -3.01312774e-01
9.16917808e-03 1.38003126e-01 -6.91122711e-01 2.78239071e-01
-5.33742420e-02 2.56555021e-01 5.92828691e-01 -8.80538583e-01
-2.85236314e-02 -3.79420519e-01 2.98868090e-01 4.77011204e-02
-9.29380596e-01 -3.74123961e-01 -6.16665304e-01 -8.91855061e-01
1.49200574e-01 -8.89578521e-01 -1.32799402e-01 2.40664154e-01
-3.67885828e-01 -5.17936349e-01 1.19676173e+00 -4.97432649e-01
1.71681321e+00 -2.08444262e+00 5.68361878e-01 2.78891563e-01
9.57727909e-01 2.83860147e-01 -2.96489924e-01 9.83335614e-01
-3.83052796e-01 1.50111392e-01 -8.63416269e-02 -4.97990809e-02
1.13384463e-01 3.20563495e-01 -7.90119022e-02 6.07010007e-01
6.14367947e-02 1.49784112e+00 -1.24025249e+00 -3.82170439e-01
3.20062339e-01 6.06737196e-01 -4.80912000e-01 7.06740171e-02
-2.38144085e-01 1.86591044e-01 -6.20399356e-01 6.78677261e-01
5.17745495e-01 -6.57097638e-01 6.95351660e-01 -1.75891832e-01
3.16300958e-01 7.12873638e-02 -8.73633981e-01 1.79184711e+00
-9.06711593e-02 6.00413382e-01 -1.51950726e-02 -1.39387786e+00
9.76177454e-01 2.62411535e-01 6.88059390e-01 -5.85778296e-01
8.71043056e-02 4.71371040e-02 2.35859603e-01 -1.89543128e-01
2.06495002e-01 3.10990751e-01 -7.51337707e-02 5.49782991e-01
2.76498884e-01 3.57419014e-01 2.91643560e-01 8.42867494e-01
1.54806495e+00 -1.49528280e-01 2.43100628e-01 -3.61784309e-01
5.63353837e-01 -4.40097272e-01 6.49926424e-01 1.41221717e-01
-3.55469435e-01 1.10588714e-01 8.35661113e-01 -8.45957041e-01
-7.63473153e-01 -1.14868891e+00 3.22446823e-01 9.81399298e-01
3.57664883e-01 -1.03595901e+00 -2.44180515e-01 -8.77170980e-01
4.68377471e-01 -3.33572971e-03 -9.04954731e-01 -6.42039537e-01
-6.15260184e-01 -4.96455789e-01 3.49929512e-01 4.98915732e-01
3.46889138e-01 -8.77457917e-01 2.05416158e-01 2.30915397e-01
8.71437266e-02 -8.46229494e-01 -8.35346997e-01 -1.86103314e-01
-8.64637554e-01 -1.40452433e+00 -3.17711532e-01 -1.00731611e+00
7.66785562e-01 5.98895907e-01 1.07925248e+00 9.00386393e-01
-4.24041390e-01 6.91432238e-01 -5.90140224e-01 2.74015695e-01
-2.22061455e-01 3.66014302e-01 -9.36427265e-02 -2.97985785e-02
3.81271809e-01 -1.20450032e+00 -7.39855468e-01 6.93927985e-03
-9.08183932e-01 -1.19659016e-02 4.52438712e-01 9.43652570e-01
4.27032530e-01 1.57433867e-01 4.78958786e-01 -1.12842822e+00
7.43377805e-01 -6.31807685e-01 -4.00587916e-01 2.00999528e-01
-9.48006213e-01 -2.31856713e-03 6.47666574e-01 -3.91449511e-01
-4.19993579e-01 -4.87076521e-01 2.31687024e-01 -8.89245808e-01
4.04510915e-01 8.21301222e-01 -2.07077876e-01 -3.28839928e-01
3.34721148e-01 3.72344434e-01 2.13582322e-01 -4.17067200e-01
5.15257716e-01 1.92763973e-02 1.61698014e-01 -2.71964192e-01
1.14041221e+00 3.93662870e-01 9.18069631e-02 -6.32391334e-01
-4.14622128e-01 -4.61531878e-01 -6.96039975e-01 -2.60895073e-01
5.50410688e-01 -7.70316184e-01 -6.69775367e-01 2.38083303e-01
-7.68966794e-01 -3.70228857e-01 -9.98715013e-02 1.67098612e-01
-1.94025531e-01 6.60817504e-01 -8.00410986e-01 -2.78549850e-01
-4.06189382e-01 -6.04952872e-01 8.69905949e-01 7.99857825e-02
1.88524216e-01 -1.69900787e+00 3.32806557e-01 -3.64474021e-02
4.75662321e-01 6.56223178e-01 1.19272661e+00 -4.57690597e-01
-7.48283803e-01 -3.88266444e-01 -2.10152775e-01 1.02184704e-02
5.03190219e-01 2.24177409e-02 -4.23492700e-01 -9.33222830e-01
-7.09921479e-01 -1.02225512e-01 1.08949280e+00 1.27550870e-01
1.09215915e+00 -4.02928263e-01 -7.91555405e-01 8.71383429e-01
1.60833085e+00 -2.12199941e-01 5.74440122e-01 5.49197495e-02
1.29140568e+00 1.68942094e-01 1.90508962e-01 1.50395527e-01
5.80800295e-01 5.48256636e-01 7.68166125e-01 -6.96792081e-02
-3.18608701e-01 -6.00199342e-01 3.18244517e-01 1.22925270e+00
-1.58884488e-02 -2.95158416e-01 -9.26513910e-01 7.35833943e-01
-2.08997440e+00 -9.41778541e-01 1.97915230e-02 1.62821198e+00
3.53347093e-01 1.13072500e-01 1.12066537e-01 3.41867432e-02
8.05965483e-01 8.93571675e-01 -6.17215931e-01 -2.98492700e-01
1.38321221e-01 5.00579216e-02 3.78170103e-01 4.11359727e-01
-1.06340551e+00 1.11042893e+00 4.81445122e+00 5.83939373e-01
-1.07672238e+00 1.67583480e-01 2.50453919e-01 1.09452434e-01
-6.94250226e-01 2.64376611e-01 -1.75419971e-01 4.29538429e-01
7.88517118e-01 -6.47445798e-01 6.31668866e-01 7.94088781e-01
-1.42718449e-01 6.54519856e-01 -9.52761412e-01 8.72065187e-01
1.19111743e-02 -1.50923502e+00 2.41477281e-01 1.88670680e-01
5.89663088e-01 5.76693267e-02 5.18647395e-02 5.13107538e-01
4.71047491e-01 -1.18545830e+00 9.81627554e-02 4.26779211e-01
9.27365005e-01 -7.23383784e-01 5.03320336e-01 -1.08895443e-01
-2.16047430e+00 -2.41236746e-01 -4.02468860e-01 -3.33318040e-02
-7.81013742e-02 3.87767911e-01 -6.52099967e-01 1.04045379e+00
6.67276859e-01 1.54557383e+00 -7.93247461e-01 7.41420984e-01
-3.75061423e-01 5.43429255e-01 1.82808921e-01 5.21727232e-03
2.75661916e-01 -4.54004049e-01 7.12551177e-01 9.30312276e-01
9.79125723e-02 -2.09517088e-02 4.87638474e-01 8.64891768e-01
-3.67624283e-01 7.63787469e-03 -1.01021922e+00 -7.17138469e-01
6.81275249e-01 1.49440014e+00 -8.30961943e-01 1.04857415e-01
-5.70088744e-01 1.09624767e+00 7.77653158e-01 3.71363461e-01
-6.71958804e-01 -6.35836065e-01 9.23085809e-01 2.09905267e-01
4.24090117e-01 -4.52649623e-01 4.33236301e-01 -1.32347453e+00
3.01145855e-02 -6.16769075e-01 7.57109821e-01 -3.49184960e-01
-1.33352423e+00 6.86468720e-01 -1.30732864e-01 -1.11452734e+00
1.27221733e-01 -7.28672922e-01 -9.98922110e-01 5.44902027e-01
-1.44115758e+00 -1.61705697e+00 -5.58079958e-01 7.75978625e-01
6.03967384e-02 -1.68532759e-01 8.53798807e-01 3.95655602e-01
-6.18500650e-01 7.17195570e-01 1.19641781e-01 5.72208643e-01
3.13328177e-01 -1.44033563e+00 6.30150378e-01 8.41984153e-01
2.74816304e-01 8.46111059e-01 2.37849221e-01 -7.41001010e-01
-1.88275909e+00 -1.40538239e+00 5.62983215e-01 -2.02454761e-01
1.08243299e+00 -6.48034990e-01 -1.02661026e+00 9.87600684e-01
4.96006571e-02 6.54005647e-01 6.01861417e-01 2.38354370e-01
-6.38403058e-01 -2.32843071e-01 -8.28381658e-01 6.23160303e-01
1.63514376e+00 -8.17772746e-01 -1.40995339e-01 3.86795759e-01
1.18684804e+00 -1.89687252e-01 -1.29052198e+00 2.20056340e-01
4.20276582e-01 -5.40565908e-01 9.52601969e-01 -8.17822278e-01
9.38025936e-02 -2.53203481e-01 2.59126741e-02 -1.39037323e+00
-7.61829257e-01 -7.14093804e-01 -7.28918672e-01 9.96438503e-01
1.51232379e-02 -1.03855991e+00 8.16326559e-01 1.40290130e-02
-2.13075012e-01 -9.03714657e-01 -9.75845218e-01 -7.99081266e-01
1.69228995e-03 1.35495648e-01 7.23303735e-01 1.33138621e+00
7.07987845e-02 3.50675464e-01 -2.18044057e-01 2.33014703e-01
5.23582935e-01 4.13401186e-01 6.34801984e-01 -1.56368840e+00
-1.02700688e-01 -5.83526075e-01 -1.13091791e+00 -7.88523495e-01
4.54310119e-01 -1.68596804e+00 -6.93241239e-01 -1.80294514e+00
2.78464735e-01 -3.45698774e-01 -5.31732202e-01 5.52246034e-01
-1.32629737e-01 -3.84238968e-03 9.67794806e-02 8.54009092e-02
-7.23349154e-01 9.45204914e-01 1.45516503e+00 -3.78730446e-01
-3.37515511e-02 -3.91448647e-01 -8.49303961e-01 1.47477090e-01
6.87328577e-01 -1.83659330e-01 -8.64169717e-01 -3.71165961e-01
1.66408569e-01 -2.13331580e-01 4.72512215e-01 -8.22331369e-01
2.33643636e-01 3.44881453e-02 7.01019689e-02 -2.57890493e-01
5.09480648e-02 -7.71267354e-01 3.79806817e-01 7.59252310e-01
-8.33699945e-03 2.83012569e-01 -8.59806256e-04 1.17650306e+00
-1.96605325e-01 3.93456727e-01 4.92285430e-01 -1.01019479e-01
-8.45108807e-01 1.11477482e+00 1.85753882e-01 -3.65826190e-02
1.05593538e+00 -2.52358168e-01 -5.68619311e-01 -3.96050870e-01
-6.96252406e-01 5.19454241e-01 5.51928461e-01 7.88488269e-01
8.94736707e-01 -1.85487735e+00 -5.22447169e-01 1.18465804e-01
3.65023434e-01 -1.40889466e-01 2.91200310e-01 8.78343046e-01
-4.60541308e-01 3.87110084e-01 -8.38201120e-02 -3.72109920e-01
-1.17117918e+00 8.10776830e-01 4.88058358e-01 -5.04674554e-01
-1.07818151e+00 7.52533078e-01 1.51244417e-01 -6.93702817e-01
7.02845901e-02 -2.94127882e-01 -2.54204452e-01 -1.04090594e-01
1.30882531e-01 1.47522867e-01 -1.69636101e-01 -5.54105699e-01
-3.89763474e-01 4.10861641e-01 -2.61969209e-01 5.74278235e-01
1.58268428e+00 4.12937813e-02 -4.97138768e-01 3.59021544e-01
1.42831504e+00 -3.03385943e-01 -9.45500851e-01 -4.51408088e-01
6.79411665e-02 -6.83991373e-01 7.62460753e-02 -1.54822126e-01
-1.48688483e+00 7.21452534e-01 4.47593302e-01 4.18612182e-01
9.41611826e-01 1.90124482e-01 8.96352291e-01 2.69378752e-01
3.48603576e-01 -5.65429032e-01 5.91799259e-01 4.41688180e-01
9.47589099e-01 -1.08302963e+00 4.02156487e-02 -5.85255742e-01
-2.58984029e-01 1.10814786e+00 9.10503149e-01 -4.99460667e-01
1.07500565e+00 -3.04914534e-01 -5.10857522e-01 -7.79196560e-01
-8.86082828e-01 -6.86881468e-02 3.00579250e-01 5.52167952e-01
4.16995943e-01 2.56865501e-01 -3.35697770e-01 3.15989375e-01
9.36578885e-02 -4.62967843e-01 3.53957921e-01 8.81372213e-01
-1.50665760e-01 -1.35532582e+00 5.13100505e-01 7.16375768e-01
-5.13926633e-02 -1.15351975e-01 -5.58424890e-01 8.82712722e-01
-2.73844898e-01 4.59448516e-01 3.79017591e-02 -7.45197535e-01
9.40266177e-02 -2.24569947e-01 2.40752056e-01 -8.54292095e-01
-3.17897916e-01 -3.04085076e-01 -4.27495465e-02 -7.83060908e-01
-2.65025884e-01 -3.80848527e-01 -1.16387892e+00 -6.42520130e-01
-2.48926118e-01 2.29400620e-01 -6.82269037e-02 2.95782536e-01
6.20253026e-01 7.98288524e-01 6.48140848e-01 -6.45105124e-01
-1.80507168e-01 -1.00108051e+00 -8.44479263e-01 5.16051352e-01
3.38464290e-01 -8.56593192e-01 -4.63738829e-01 -3.95628899e-01] | [7.205883502960205, 6.166987419128418] |
fc5ec8c3-d196-49e0-a307-0ec5d957cb4c | zero-shot-pose-transfer-for-unrigged-stylized-1 | 2306.002 | null | https://arxiv.org/abs/2306.00200v1 | https://arxiv.org/pdf/2306.00200v1.pdf | Zero-shot Pose Transfer for Unrigged Stylized 3D Characters | Transferring the pose of a reference avatar to stylized 3D characters of various shapes is a fundamental task in computer graphics. Existing methods either require the stylized characters to be rigged, or they use the stylized character in the desired pose as ground truth at training. We present a zero-shot approach that requires only the widely available deformed non-stylized avatars in training, and deforms stylized characters of significantly different shapes at inference. Classical methods achieve strong generalization by deforming the mesh at the triangle level, but this requires labelled correspondences. We leverage the power of local deformation, but without requiring explicit correspondence labels. We introduce a semi-supervised shape-understanding module to bypass the need for explicit correspondences at test time, and an implicit pose deformation module that deforms individual surface points to match the target pose. Furthermore, to encourage realistic and accurate deformation of stylized characters, we introduce an efficient volume-based test-time training procedure. Because it does not need rigging, nor the deformed stylized character at training time, our model generalizes to categories with scarce annotation, such as stylized quadrupeds. Extensive experiments demonstrate the effectiveness of the proposed method compared to the state-of-the-art approaches trained with comparable or more supervision. Our project page is available at https://jiashunwang.github.io/ZPT | ['Jan Kautz', 'Xiaolong Wang', 'Orazio Gallo', 'Shalini De Mello', 'Sifei Liu', 'Xueting Li', 'Jiashun Wang'] | 2023-05-31 | zero-shot-pose-transfer-for-unrigged-stylized | http://openaccess.thecvf.com//content/CVPR2023/html/Wang_Zero-Shot_Pose_Transfer_for_Unrigged_Stylized_3D_Characters_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Wang_Zero-Shot_Pose_Transfer_for_Unrigged_Stylized_3D_Characters_CVPR_2023_paper.pdf | cvpr-2023-1 | ['pose-transfer'] | ['computer-vision'] | [-4.36574444e-02 1.72995508e-01 6.79295743e-03 -1.61719248e-01
-8.87099147e-01 -9.68869150e-01 4.58409071e-01 -4.13850158e-01
-9.09870937e-02 4.99740601e-01 -1.80634096e-01 -4.35106410e-03
5.36895394e-01 -9.28309500e-01 -1.03044868e+00 -4.40480500e-01
4.36353475e-01 1.08254683e+00 6.07397079e-01 -3.78329307e-01
1.76536992e-01 6.79970741e-01 -1.26484632e+00 -9.07029435e-02
1.12420905e+00 8.61673534e-01 -1.54652625e-01 5.20360112e-01
-2.05633938e-02 -6.82918206e-02 -4.44196582e-01 -9.24218953e-01
6.14032149e-01 -3.43663514e-01 -7.77370155e-01 3.85880262e-01
8.93598855e-01 -5.31345785e-01 8.43559504e-02 9.59577739e-01
4.73609477e-01 1.92662865e-01 7.75353730e-01 -9.84289885e-01
-6.48251772e-01 1.10319346e-01 -8.17635894e-01 -5.67529023e-01
4.73651081e-01 7.88627714e-02 9.83669817e-01 -1.20059860e+00
8.89993072e-01 1.13153136e+00 7.55517244e-01 7.78352082e-01
-1.28568089e+00 -7.19186366e-01 1.04890473e-01 -4.09653127e-01
-1.22538626e+00 -2.81861871e-01 1.02195525e+00 -7.28140175e-01
1.39595330e-01 1.40048921e-01 8.00666034e-01 1.08919418e+00
-2.86408067e-01 6.91903889e-01 7.67656028e-01 -2.14059010e-01
2.02806309e-01 -1.30802482e-01 -1.53466895e-01 1.02103221e+00
1.83950633e-01 -1.52276233e-01 -2.52883136e-01 -2.04667196e-01
1.39159393e+00 -8.60750601e-02 -1.87001795e-01 -8.92162859e-01
-1.04067004e+00 6.56219065e-01 3.19991201e-01 -5.17974533e-02
-1.85529783e-01 2.62494534e-01 3.61446589e-01 -2.76091285e-02
5.63335180e-01 3.73911291e-01 -3.46482694e-01 -2.52563626e-01
-9.61526930e-01 2.27448463e-01 7.57280946e-01 1.21516526e+00
1.02541745e+00 1.49018645e-01 4.59499843e-02 7.13560641e-01
1.22951292e-01 4.85552967e-01 -1.30984141e-02 -1.05031323e+00
4.78668272e-01 6.80894375e-01 1.99036241e-01 -7.74274290e-01
8.41403157e-02 4.57181185e-02 -6.58891857e-01 4.84488994e-01
7.62450337e-01 -1.34099752e-01 -1.09935510e+00 1.42407608e+00
7.65783072e-01 4.37673628e-01 -3.00056607e-01 9.97862697e-01
7.92275727e-01 3.71362895e-01 -2.89004058e-01 2.64031947e-01
1.09164691e+00 -1.01226151e+00 -2.80009955e-01 1.06758893e-01
5.00916302e-01 -7.86571622e-01 1.58344400e+00 2.44119346e-01
-1.26210833e+00 -5.00073671e-01 -8.16090345e-01 -2.65618145e-01
1.67259537e-02 1.17937058e-01 3.89216721e-01 7.36378312e-01
-7.38058627e-01 8.89335811e-01 -8.99219513e-01 -1.29550904e-01
5.68506300e-01 3.19271803e-01 -3.17375720e-01 1.74239576e-01
-8.73728275e-01 6.36532068e-01 -7.59233907e-02 1.25143617e-01
-8.47954750e-01 -8.71409059e-01 -1.07363570e+00 -3.64514649e-01
4.23544407e-01 -7.08513558e-01 1.16028357e+00 -1.04410231e+00
-2.00414085e+00 1.17814028e+00 2.49733217e-02 2.87111044e-01
1.12542236e+00 -4.81677294e-01 2.34188005e-01 7.71707222e-02
6.04014248e-02 4.80220258e-01 8.72573793e-01 -1.59471834e+00
-1.04428818e-02 -2.81885654e-01 2.13682637e-01 3.05137187e-01
2.84524281e-02 -2.98851490e-01 -1.04294026e+00 -7.99985647e-01
1.57555982e-01 -1.11647177e+00 -1.00556158e-01 4.31020677e-01
-6.79026604e-01 -1.45183578e-02 8.10731530e-01 -7.56909192e-01
6.05152369e-01 -1.96390724e+00 3.48794401e-01 4.59214360e-01
9.54726711e-02 1.66319922e-01 -3.85148600e-02 1.98491141e-01
1.67043194e-01 8.25079158e-02 -2.99786240e-01 -6.38122857e-01
3.54857370e-03 2.21215978e-01 -2.43556976e-01 4.61196572e-01
1.77789748e-01 1.00693417e+00 -9.17883277e-01 -6.60029590e-01
7.87251294e-02 5.36260903e-01 -8.06457937e-01 2.15775385e-01
-3.94489944e-01 1.05953264e+00 -6.63591981e-01 8.28551054e-01
6.10548198e-01 -2.23519951e-01 -1.47224134e-02 -3.50020111e-01
1.12262100e-01 1.05880059e-01 -1.38304329e+00 2.08848739e+00
-2.81665474e-01 3.14098690e-03 3.08155548e-03 -6.44451022e-01
1.09331858e+00 3.71287376e-01 4.05200034e-01 -1.81502476e-01
2.33086288e-01 3.75678867e-01 -3.81428927e-01 -9.33028460e-02
3.18824381e-01 -1.94571078e-01 -1.48530647e-01 5.39396405e-01
-1.96191952e-01 -6.78206325e-01 -6.96962252e-02 3.63451336e-03
5.51823318e-01 9.94395137e-01 -4.43156436e-02 -9.84193683e-02
3.93312097e-01 -2.01869849e-02 7.77876914e-01 1.70099899e-01
2.91039288e-01 1.04297447e+00 4.77314979e-01 -4.69567209e-01
-1.45062208e+00 -1.10008824e+00 -9.98459384e-02 1.02595711e+00
3.93103659e-01 -1.64979592e-01 -1.00088966e+00 -6.63466871e-01
-3.15568037e-02 4.73485798e-01 -8.03767383e-01 3.13560694e-01
-9.31140780e-01 -1.53248399e-01 4.69856113e-01 7.53434002e-01
4.30032849e-01 -8.99001539e-01 -4.34935838e-01 -1.34404868e-01
8.25024769e-02 -9.38061416e-01 -9.37789023e-01 -4.28561032e-01
-9.99372125e-01 -9.70152199e-01 -1.13870215e+00 -8.32207561e-01
1.10553730e+00 -2.53872067e-01 1.17684281e+00 2.78933495e-01
3.06902714e-02 2.76219994e-01 -2.59276092e-01 -8.16921592e-02
-4.16763276e-01 9.72279608e-02 -4.65934910e-02 8.37978572e-02
-4.12805736e-01 -7.15532303e-01 -5.86484611e-01 6.96015418e-01
-6.13302112e-01 4.05612677e-01 2.30852857e-01 7.29452491e-01
9.52565193e-01 -5.61693132e-01 1.70466125e-01 -1.26523507e+00
8.40750337e-02 1.77471235e-01 -7.22203493e-01 1.62106439e-01
-1.31660625e-01 2.73354147e-02 6.33480132e-01 -7.18103111e-01
-9.43226933e-01 2.39469588e-01 -2.09824324e-01 -8.98184776e-01
-7.48481601e-02 9.01665837e-02 -5.58330417e-01 -1.78845331e-01
6.01724029e-01 -5.74231073e-02 4.10499424e-02 -7.11243212e-01
4.91003633e-01 1.22682542e-01 6.25576019e-01 -1.21997046e+00
1.30121255e+00 5.47273934e-01 6.88696876e-02 -4.04699564e-01
-6.47443712e-01 -4.67628688e-02 -1.22979164e+00 -2.47587129e-01
7.62784481e-01 -6.93674088e-01 -8.05707991e-01 4.33396906e-01
-1.20466089e+00 -6.74973190e-01 -5.17833829e-01 1.01313278e-01
-6.97470427e-01 4.81221765e-01 -6.45771682e-01 -4.22148049e-01
-4.22685474e-01 -1.19760346e+00 1.46753418e+00 3.79229672e-02
-2.31319994e-01 -1.03396535e+00 1.85126767e-01 4.68598783e-01
-1.13562778e-01 7.70906389e-01 6.32795334e-01 -3.91297966e-01
-5.65962195e-01 -3.49237382e-01 1.08574599e-01 1.58639461e-01
2.81967640e-01 2.36151397e-01 -7.63144791e-01 -3.21925312e-01
-5.54614186e-01 -4.82250005e-01 2.55664676e-01 2.32890751e-02
1.13608289e+00 -1.27513424e-01 -1.25933319e-01 8.87127221e-01
1.09420133e+00 -1.60940990e-01 5.96983910e-01 3.25074680e-02
1.32442427e+00 5.57224452e-01 5.59045732e-01 3.59765857e-01
3.80783737e-01 8.28498960e-01 3.59072208e-01 -1.92710936e-01
-1.90339223e-01 -5.25643110e-01 2.01359376e-01 9.37244713e-01
-7.17621446e-01 4.73031104e-02 -9.52332199e-01 2.49715805e-01
-1.64418209e+00 -6.59095228e-01 -1.57970622e-01 2.51788330e+00
1.12917900e+00 4.75302488e-02 2.56266713e-01 -5.39772324e-02
7.21619785e-01 -1.01499997e-01 -8.28596592e-01 -3.50926578e-01
2.68937290e-01 4.00621623e-01 2.12472618e-01 7.46425569e-01
-8.45345616e-01 1.25175917e+00 5.43785334e+00 6.88955665e-01
-1.16627419e+00 5.98758422e-02 5.19872844e-01 5.58982529e-02
-5.22712946e-01 4.73756455e-02 -5.79611838e-01 2.50458688e-01
2.11940005e-01 -2.27790307e-02 2.50447422e-01 9.12936032e-01
1.02037631e-01 1.51026666e-01 -1.32265544e+00 8.49942029e-01
2.91844029e-02 -1.30255103e+00 1.69552162e-01 1.80900358e-02
8.67094338e-01 -4.98259395e-01 -6.21753149e-02 1.13358811e-01
4.04890984e-01 -9.50618446e-01 8.24746311e-01 7.26178765e-01
1.31790519e+00 -6.77869737e-01 4.11737919e-01 2.32119039e-01
-1.44411373e+00 8.14694405e-01 -2.16249079e-01 2.15192765e-01
1.76686615e-01 1.77550107e-01 -6.31302655e-01 7.34034836e-01
3.58924806e-01 6.45991743e-01 -4.80393887e-01 6.85436189e-01
-3.84834260e-01 5.44797421e-01 -3.29052806e-01 1.95669010e-01
-5.02133742e-02 -6.31561518e-01 4.66559708e-01 7.96868026e-01
3.22857141e-01 3.43412459e-02 4.72556263e-01 9.79738712e-01
-2.06155181e-01 2.52888501e-01 -4.42462504e-01 2.72179306e-01
4.14241076e-01 1.22245622e+00 -7.27995515e-01 -3.38223368e-01
-2.63296008e-01 1.41986108e+00 2.29868203e-01 2.87295938e-01
-8.82649660e-01 -1.16853140e-01 3.77848297e-01 5.23787677e-01
2.32494682e-01 -3.22300225e-01 -5.48447251e-01 -1.25159454e+00
2.19662502e-01 -7.23038495e-01 5.59437945e-02 -7.58913755e-01
-1.27514100e+00 5.39945126e-01 -4.88888696e-02 -1.55760741e+00
-2.50233740e-01 -4.30844218e-01 -8.98673058e-01 1.06196511e+00
-1.03805876e+00 -1.72546566e+00 -5.14850497e-01 6.33109331e-01
3.95000845e-01 1.83804974e-01 7.79132187e-01 1.85892761e-01
-4.95005131e-01 8.76155257e-01 -3.39721531e-01 4.79712248e-01
8.16969991e-01 -1.31723845e+00 6.62103713e-01 6.05871499e-01
2.86534987e-02 5.29742062e-01 6.49710536e-01 -9.51290369e-01
-1.33124864e+00 -1.17110360e+00 3.28812212e-01 -7.29030490e-01
5.07673323e-01 -5.31569958e-01 -1.17352641e+00 8.42518151e-01
-1.96825922e-01 1.98456690e-01 4.97043610e-01 1.37272822e-02
-3.12957197e-01 2.16990173e-01 -1.04258049e+00 8.71714473e-01
1.17956352e+00 -3.64775568e-01 -4.98170376e-01 2.61438489e-01
4.88412380e-01 -1.04284453e+00 -1.02137256e+00 3.10643673e-01
6.76920176e-01 -6.35016620e-01 9.40158188e-01 -5.91359735e-01
4.54120994e-01 -4.63500500e-01 1.09935507e-01 -1.19954050e+00
-9.28239338e-03 -8.84772480e-01 4.26150039e-02 1.16107810e+00
3.47844839e-01 -3.70589346e-01 1.31309438e+00 9.21046019e-01
-2.66076773e-01 -9.60911214e-01 -8.01156402e-01 -8.65117729e-01
4.45901990e-01 -1.59782425e-01 5.57476342e-01 1.08238077e+00
-3.44729245e-01 1.14503101e-01 -4.62481081e-01 2.03718573e-01
6.54937446e-01 3.75884026e-01 1.35594046e+00 -1.32980490e+00
-2.74658442e-01 -2.32864454e-01 -3.67226094e-01 -1.23086095e+00
2.48313576e-01 -8.56995523e-01 -1.18625201e-02 -1.41299939e+00
-2.28447057e-02 -6.54767573e-01 5.81743419e-01 6.13758028e-01
-2.88153082e-01 5.56698084e-01 4.76071760e-02 4.08833206e-01
-2.95747846e-01 8.05405438e-01 2.03264713e+00 2.34777883e-01
-3.79101425e-01 1.38026401e-01 -2.79967457e-01 1.07142580e+00
6.65404677e-01 -2.36746818e-01 -2.92104572e-01 -4.23250139e-01
1.43576384e-01 2.26774514e-02 5.47706962e-01 -6.04836702e-01
-1.39988195e-02 -2.71229714e-01 3.63462508e-01 -4.87829208e-01
5.95271051e-01 -6.70439303e-01 3.31442893e-01 2.51573443e-01
4.11907136e-02 8.46303701e-02 1.12676129e-01 2.69515932e-01
2.63967991e-01 -2.99092770e-01 9.37033176e-01 -4.54290248e-02
-2.52502024e-01 7.67847240e-01 3.46590042e-01 3.04299802e-01
1.01182115e+00 -6.73954129e-01 2.15881050e-01 -2.61785954e-01
-8.89225364e-01 9.16081518e-02 1.24215043e+00 2.13292554e-01
5.55874825e-01 -1.45728171e+00 -6.18971050e-01 5.97118512e-02
-5.89754507e-02 5.96630871e-01 6.19085021e-02 6.39641225e-01
-8.41496050e-01 -3.78081083e-01 -1.86942145e-01 -6.80815697e-01
-1.05477774e+00 3.13772649e-01 4.23398763e-01 7.31836557e-02
-8.69438052e-01 7.34775782e-01 3.51605654e-01 -7.33301759e-01
4.90089618e-02 -2.12945893e-01 2.17882916e-01 -8.44830051e-02
8.04128498e-03 2.89242595e-01 -2.29217745e-02 -7.39462912e-01
-8.94590914e-02 1.16790748e+00 9.82497036e-02 -2.49763325e-01
1.19227338e+00 2.75885969e-01 1.34101436e-01 6.68448985e-01
8.99066210e-01 4.76160437e-01 -1.72990549e+00 -2.77530342e-01
-2.90495455e-01 -6.15243554e-01 -5.28429329e-01 -3.92990679e-01
-1.17035353e+00 9.29357231e-01 1.04828998e-01 -3.51862252e-01
6.26741707e-01 6.81043193e-02 9.24027562e-01 2.04148218e-01
5.66673279e-01 -9.16428685e-01 3.98544759e-01 4.78477687e-01
1.17725766e+00 -1.04244757e+00 1.16132759e-02 -8.19425523e-01
-8.28622103e-01 1.08911836e+00 7.90541708e-01 -4.87466156e-01
5.33034503e-01 3.20319951e-01 1.80139139e-01 -2.00091720e-01
-1.04271710e-01 1.01604290e-01 6.14374220e-01 5.08713067e-01
3.45868409e-01 1.57239456e-02 4.07826155e-02 3.72500896e-01
-5.12594283e-01 -2.54949331e-01 3.06587577e-01 9.21492636e-01
-1.11148536e-01 -1.34929395e+00 -4.82450336e-01 2.39886850e-01
-1.94064796e-01 -3.78064089e-03 -4.74421203e-01 7.88319170e-01
7.21249580e-02 2.63263851e-01 1.73731729e-01 -2.91479230e-01
6.38309777e-01 6.77816123e-02 6.52319968e-01 -9.31883216e-01
-6.28037274e-01 2.51611501e-01 -1.39183432e-01 -5.14061213e-01
-2.44738504e-01 -7.51694918e-01 -1.42532849e+00 -4.60481524e-01
-3.50649744e-01 3.74467559e-02 1.96149275e-01 7.61482894e-01
2.47294664e-01 2.08774015e-01 6.43058717e-01 -1.50263166e+00
-2.49924719e-01 -7.93859422e-01 -4.13641632e-01 7.75051653e-01
-5.94772622e-02 -9.42549586e-01 -3.75375152e-02 4.59144384e-01] | [7.361363410949707, -1.63212251663208] |
5c500bb5-ea62-42f1-96ba-75738336409a | a-compressive-multi-kernel-method-for-privacy | 2106.10671 | null | https://arxiv.org/abs/2106.10671v1 | https://arxiv.org/pdf/2106.10671v1.pdf | A compressive multi-kernel method for privacy-preserving machine learning | As the analytic tools become more powerful, and more data are generated on a daily basis, the issue of data privacy arises. This leads to the study of the design of privacy-preserving machine learning algorithms. Given two objectives, namely, utility maximization and privacy-loss minimization, this work is based on two previously non-intersecting regimes -- Compressive Privacy and multi-kernel method. Compressive Privacy is a privacy framework that employs utility-preserving lossy-encoding scheme to protect the privacy of the data, while multi-kernel method is a kernel based machine learning regime that explores the idea of using multiple kernels for building better predictors. The compressive multi-kernel method proposed consists of two stages -- the compression stage and the multi-kernel stage. The compression stage follows the Compressive Privacy paradigm to provide the desired privacy protection. Each kernel matrix is compressed with a lossy projection matrix derived from the Discriminant Component Analysis (DCA). The multi-kernel stage uses the signal-to-noise ratio (SNR) score of each kernel to non-uniformly combine multiple compressive kernels. The proposed method is evaluated on two mobile-sensing datasets -- MHEALTH and HAR -- where activity recognition is defined as utility and person identification is defined as privacy. The results show that the compression regime is successful in privacy preservation as the privacy classification accuracies are almost at the random-guess level in all experiments. On the other hand, the novel SNR-based multi-kernel shows utility classification accuracy improvement upon the state-of-the-art in both datasets. These results indicate a promising direction for research in privacy-preserving machine learning. | ['S. Y. Kung', 'J. Morris Chang', 'Thee Chanyaswad'] | 2021-06-20 | null | null | null | null | ['person-identification'] | ['computer-vision'] | [ 6.60536885e-01 -7.24405348e-02 -2.84973264e-01 -3.70317042e-01
-1.04921544e+00 -3.18999141e-01 2.96298385e-01 1.87339321e-01
-4.85738069e-01 8.25766563e-01 4.59060848e-01 -2.18785837e-01
-4.59832340e-01 -6.53525591e-01 -4.95484143e-01 -1.10933983e+00
-1.45717561e-01 -1.62096977e-01 -2.92688042e-01 2.34373599e-01
3.32005657e-02 4.90475446e-01 -1.24826169e+00 5.48790991e-01
7.60321617e-01 1.28865469e+00 -3.99931014e-01 3.88002425e-01
5.32246649e-01 6.05099440e-01 1.53087407e-01 -8.09846759e-01
9.17981207e-01 -4.78882730e-01 -3.52226764e-01 -3.19706798e-01
-8.31203312e-02 -2.63668746e-01 -5.30789554e-01 1.13916290e+00
4.93402332e-01 3.01395613e-03 5.04225731e-01 -1.39383936e+00
-4.70745355e-01 1.26119643e-01 -6.05293632e-01 -1.62684232e-01
2.70226210e-01 -3.77806932e-01 5.68648577e-01 -5.66653192e-01
2.07808793e-01 8.20066869e-01 1.04883659e+00 3.12128991e-01
-1.10187364e+00 -6.30415916e-01 -6.92154050e-01 2.84393787e-01
-1.80792570e+00 -5.74261367e-01 7.04101026e-01 -2.58494109e-01
5.30156493e-01 1.02265310e+00 5.57112277e-01 6.99397743e-01
3.69600177e-01 5.57626605e-01 1.41697931e+00 -4.03243273e-01
5.20954669e-01 6.86835706e-01 2.52721459e-02 5.29838145e-01
7.46515572e-01 3.19226235e-01 -5.95428169e-01 -8.33352804e-01
4.66252029e-01 4.80144441e-01 -4.24173921e-01 -7.01214969e-01
-6.53222919e-01 8.30029905e-01 -1.61619186e-01 1.75967783e-01
-7.39219964e-01 -1.60728455e-01 4.04724866e-01 5.04265130e-01
3.54503125e-01 2.63652275e-03 -2.06027493e-01 6.36066794e-02
-1.41134048e+00 3.89862031e-01 1.08324885e+00 7.42202222e-01
4.81520295e-01 -2.62329251e-01 -3.64159226e-01 6.55621290e-01
3.21589112e-01 2.87266076e-01 6.06982529e-01 -7.95416951e-01
6.08202517e-01 3.85915399e-01 2.75463294e-02 -1.26003969e+00
1.48754865e-01 -6.21806383e-01 -1.10764456e+00 1.94105059e-01
2.92035550e-01 -3.23509157e-01 -2.19503164e-01 1.72754872e+00
3.03110570e-01 2.32239470e-01 4.88547921e-01 7.16938317e-01
9.86022130e-02 5.51645577e-01 6.23931028e-02 -6.61410272e-01
1.49534559e+00 -5.01853824e-01 -8.49831164e-01 1.83160767e-01
4.88112926e-01 -4.44406271e-01 5.12525856e-01 4.77635652e-01
-1.03030252e+00 -1.71838589e-02 -1.25420952e+00 1.72436193e-01
-3.65480244e-01 1.84333459e-01 6.30482554e-01 1.35729492e+00
-9.70241666e-01 4.83284652e-01 -6.63000047e-01 -4.63731259e-01
7.98316777e-01 3.61217916e-01 -8.84234846e-01 -4.13022012e-01
-1.05680215e+00 5.16989768e-01 1.88949600e-01 -3.45956713e-01
-5.10547757e-01 -7.13167906e-01 -6.03950441e-01 3.10686707e-01
-5.67459427e-02 -7.43417263e-01 6.33556247e-01 -9.79210436e-01
-1.10580719e+00 9.64011312e-01 -6.70960471e-02 -8.96055818e-01
8.17990005e-01 -2.84133572e-02 -6.18427336e-01 2.60824591e-01
2.27222610e-02 7.03683645e-02 9.63559985e-01 -9.93722379e-01
-6.91948950e-01 -9.90341544e-01 -7.11601734e-01 3.40504378e-01
-6.27122581e-01 -1.26275301e-01 -1.69749781e-01 -8.36847186e-01
2.37587228e-01 -9.00462091e-01 -1.38946831e-01 2.14483976e-01
-3.12736660e-01 5.41104972e-01 1.05909002e+00 -1.28060007e+00
1.42613280e+00 -2.47837424e+00 -1.70297667e-01 5.03957808e-01
-3.13022290e-03 1.92961067e-01 4.04535532e-01 6.04408920e-01
-1.85648799e-01 -4.70308587e-02 -6.35941207e-01 -4.78355318e-01
-1.33978218e-01 6.23070337e-02 -1.95509419e-01 9.88101423e-01
-5.35217285e-01 6.72151268e-01 -3.65529120e-01 -2.90258825e-01
8.10017437e-02 6.67443812e-01 -4.12369430e-01 8.98786262e-02
5.78849733e-01 2.04351261e-01 -6.17014349e-01 8.51285040e-01
1.13519037e+00 9.34338421e-02 1.51035100e-01 -7.30935633e-02
8.22703764e-02 -6.01885378e-01 -1.26025188e+00 1.66001666e+00
1.90373704e-01 2.40496337e-01 3.36634040e-01 -1.15161955e+00
7.19853401e-01 4.81049687e-01 7.16126680e-01 -3.96944851e-01
4.76769693e-02 1.75631762e-01 -4.16921973e-01 -6.05197191e-01
3.44075233e-01 -2.26603106e-01 -8.07889998e-02 3.40595394e-01
-3.64856035e-01 6.30529523e-01 -6.92808032e-01 -2.68992875e-02
1.19314563e+00 -3.20387900e-01 7.80227780e-01 -4.79608208e-01
4.83796924e-01 -2.98501968e-01 6.86027408e-01 6.88961744e-01
-4.74891484e-01 4.74761903e-01 1.57693133e-01 -1.96554556e-01
-8.39099765e-01 -9.72753763e-01 -2.80847192e-01 6.56475782e-01
-1.96731519e-02 -1.90971866e-01 -8.94621193e-01 -5.56446612e-01
4.81731266e-01 7.94580579e-01 -6.47043169e-01 -2.28710309e-01
8.09907988e-02 -8.40472519e-01 9.62445319e-01 1.59141794e-01
9.21706498e-01 -4.91223812e-01 -8.23567629e-01 -1.58420771e-01
-6.59403997e-03 -7.84213006e-01 -3.18267435e-01 4.78618667e-02
-1.04154015e+00 -8.50569367e-01 -7.05800533e-01 -4.57456440e-01
4.83136117e-01 3.25502366e-01 1.31219789e-01 -3.48953933e-01
-2.39860654e-01 7.65956819e-01 -4.41330880e-01 -4.54222560e-01
-9.54829976e-02 -3.04363489e-01 2.85008475e-02 8.73883665e-01
5.86105466e-01 -9.68620658e-01 -6.39721394e-01 -1.51546970e-01
-1.01128614e+00 -2.39617884e-01 8.38705659e-01 5.56603789e-01
6.81221306e-01 3.60244572e-01 4.96509492e-01 -8.63755763e-01
8.64943147e-01 -7.39910781e-01 -1.04608059e-01 4.23607051e-01
-9.92012560e-01 -7.27042332e-02 5.47053814e-01 -2.58854657e-01
-9.59425926e-01 3.81809145e-01 1.90449804e-01 -4.26201880e-01
-1.12437956e-01 3.81407470e-01 -5.45363665e-01 -3.29202503e-01
5.71724474e-01 7.60226727e-01 1.98386863e-01 -4.30619866e-01
2.12128848e-01 1.14632511e+00 5.38063347e-01 -2.99167991e-01
6.16418302e-01 6.46220505e-01 2.83197582e-01 -1.12027633e+00
-4.58713854e-03 -6.81512654e-01 -8.03117752e-02 1.06054284e-01
7.73071587e-01 -9.54270244e-01 -3.29751134e-01 5.21506906e-01
-5.59195101e-01 4.94176924e-01 -4.32984561e-01 6.98034406e-01
-6.17576718e-01 7.05959558e-01 -3.19382399e-01 -1.29318893e+00
-7.66189694e-01 -8.09053719e-01 7.00610280e-01 1.98683292e-01
-2.44162492e-02 -6.92695916e-01 2.72350516e-02 6.32233024e-01
5.09138763e-01 6.89088464e-01 6.87347472e-01 -1.01476455e+00
-2.44650587e-01 -8.44424009e-01 -1.70856323e-02 4.76044804e-01
1.08598985e-01 -1.08403146e+00 -1.18803978e+00 -6.96048141e-01
5.82018971e-01 -1.83867708e-01 5.17421186e-01 3.37441713e-01
1.32541883e+00 -8.20371032e-01 -2.72854745e-01 9.63718235e-01
1.65294647e+00 2.91557550e-01 1.05433333e+00 1.60544261e-01
2.97459900e-01 6.06316447e-01 3.44688922e-01 9.39104795e-01
1.84639439e-01 4.54954892e-01 1.85921595e-01 1.12711661e-01
4.77426469e-01 -3.86673898e-01 2.67673403e-01 2.87052214e-01
4.70340028e-02 -4.72012274e-02 -4.06544298e-01 3.72829348e-01
-2.13086581e+00 -1.33818340e+00 1.19142994e-01 2.82065082e+00
5.44495761e-01 -5.33467412e-01 1.17287345e-01 3.57403934e-01
4.17661160e-01 6.80386275e-02 -4.75596964e-01 -3.09338421e-01
-1.35830045e-01 -1.70066524e-02 9.75461006e-01 3.09565276e-01
-1.13245714e+00 3.52486968e-01 5.11287498e+00 9.13805962e-01
-8.59798551e-01 3.39452058e-01 8.65747392e-01 -5.52863143e-02
-2.15532020e-01 1.16819032e-01 -4.22811449e-01 5.22479832e-01
8.39783847e-01 -1.73438475e-01 6.73991442e-01 8.69759142e-01
2.14276657e-01 -2.38388941e-01 -8.82957637e-01 1.54992950e+00
3.21372926e-01 -1.15754330e+00 -7.78929889e-02 5.99468529e-01
3.06522131e-01 -5.43981910e-01 2.31248900e-01 -3.81985940e-02
-2.66513735e-01 -8.33531797e-01 3.31545979e-01 9.75811839e-01
9.99737501e-01 -9.44397509e-01 6.85353398e-01 6.47662580e-01
-1.04291487e+00 -3.35005999e-01 -4.22662735e-01 7.70806447e-02
3.24915573e-02 6.53472841e-01 -5.08054793e-01 8.48070741e-01
7.81771302e-01 3.13261777e-01 -3.34256291e-01 9.74849582e-01
4.41766590e-01 7.10827291e-01 -3.21609825e-01 1.22404948e-01
-1.85466155e-01 -4.91925716e-01 7.15752661e-01 1.33694816e+00
6.29978001e-01 4.28193122e-01 -1.23920374e-01 5.69233418e-01
-2.16174554e-02 4.24251646e-01 -9.08226013e-01 -5.06457947e-02
6.56867743e-01 1.02910674e+00 -1.68606058e-01 1.99552476e-02
-4.37983543e-01 1.51117122e+00 3.54891503e-03 2.05305398e-01
-5.63110411e-01 -4.48750198e-01 6.28519833e-01 3.38400662e-01
1.96801975e-01 9.32810158e-02 -5.86462736e-01 -1.07376003e+00
1.67528063e-01 -1.00943267e+00 7.27786362e-01 -2.81562269e-01
-1.28299665e+00 5.83834052e-02 1.21102266e-01 -1.28813517e+00
-5.23133799e-02 -8.11025128e-02 -2.60139346e-01 8.68245780e-01
-1.17512643e+00 -1.43561721e+00 -3.91393900e-03 9.59798753e-01
-8.06860924e-02 -5.78887522e-01 1.06453192e+00 3.09367061e-01
-2.50327379e-01 1.10008693e+00 6.68062449e-01 -1.28037527e-01
3.50208759e-01 -6.57164991e-01 -5.38988233e-01 1.03436470e+00
-1.47802442e-01 6.88363492e-01 3.93267095e-01 -8.22525561e-01
-1.74032521e+00 -1.10144997e+00 1.08299434e+00 2.99584810e-02
-1.38548568e-01 -2.49991208e-01 -7.81872630e-01 5.87062359e-01
1.05292804e-03 1.36454374e-01 1.44019806e+00 -3.49152058e-01
-2.30062813e-01 -4.91834402e-01 -2.14114380e+00 1.74035192e-01
5.50470412e-01 -5.90608835e-01 -3.93421650e-01 1.01922989e-01
3.25392991e-01 1.13762550e-01 -9.52149928e-01 2.02371970e-01
9.61024940e-01 -1.09894526e+00 9.74951029e-01 -5.19661129e-01
-1.05191618e-01 -9.59753394e-02 -9.75184262e-01 -5.70474446e-01
-3.19930345e-01 -7.51077235e-01 -3.41306657e-01 1.28180456e+00
1.87962547e-01 -7.72595406e-01 1.08926666e+00 1.08019197e+00
4.87698257e-01 -6.82707965e-01 -1.31464958e+00 -6.59189463e-01
-3.31887454e-01 -2.62111932e-01 4.76040065e-01 1.26545715e+00
1.12071574e-01 -2.47787893e-01 -9.34964418e-01 2.86293685e-01
1.11114478e+00 -3.32399040e-01 6.58688426e-01 -9.83399153e-01
-5.79821408e-01 2.06941515e-01 -5.89627743e-01 -5.08452654e-01
-3.46966803e-01 -8.96840155e-01 -6.55743301e-01 -9.09833193e-01
4.65051591e-01 -1.58295035e-01 -5.73168337e-01 4.59949225e-01
1.59860536e-01 -1.32014707e-01 1.07790411e-01 2.92481601e-01
-2.11938947e-01 4.95617181e-01 5.32331765e-01 1.86548725e-01
-3.78020942e-01 4.35984313e-01 -1.10725939e+00 3.43991607e-01
8.15203428e-01 -5.88852763e-01 -7.88052082e-01 1.58175588e-01
-2.75723964e-01 5.31191409e-01 3.89701039e-01 -1.39379561e+00
4.58546340e-01 -1.26981243e-01 5.22321463e-01 -3.76159728e-01
4.96208280e-01 -1.38089645e+00 6.74614906e-01 6.29046082e-01
-4.17089880e-01 -3.17330807e-02 -7.10506588e-02 1.07335865e+00
3.35709304e-02 -2.33476982e-02 8.88120711e-01 -4.07977588e-02
-2.74354845e-01 3.16028565e-01 -8.80570412e-02 -3.45272839e-01
1.37963367e+00 -7.70733535e-01 2.18586579e-01 -5.45771539e-01
-6.01948321e-01 -2.44274318e-01 4.27528888e-01 -7.42497528e-03
8.73566806e-01 -1.27022827e+00 -8.03562522e-01 5.90247631e-01
4.48922254e-02 -6.59085810e-01 3.73733878e-01 8.41099918e-01
-3.26057762e-01 3.31358641e-01 -2.11658299e-01 -1.14821829e-01
-1.48579633e+00 4.63409781e-01 1.74808711e-01 -3.71879160e-01
-5.44314623e-01 5.37051320e-01 -1.23760074e-01 -1.48235977e-01
3.46244991e-01 2.62628376e-01 -3.25043537e-02 -2.31272146e-01
7.78447747e-01 8.62923503e-01 -2.52846163e-02 -6.44511223e-01
-3.47525895e-01 2.18657240e-01 1.49367407e-01 -4.30184990e-01
1.33432913e+00 -2.36982971e-01 -2.01851591e-01 -4.42978255e-02
1.54794943e+00 2.66651720e-01 -9.94307280e-01 -1.99743226e-01
-2.91737691e-02 -8.44665170e-01 1.77993104e-01 -8.57242405e-01
-7.86747277e-01 4.28283513e-01 1.22841787e+00 4.36855741e-02
1.53156376e+00 -5.48954129e-01 8.65787029e-01 9.73888859e-02
3.91157627e-01 -1.00848258e+00 -6.26260698e-01 -1.77066728e-01
7.12212145e-01 -9.83294070e-01 2.68299520e-01 -2.84730554e-01
-8.58063281e-01 7.19070375e-01 -1.50678381e-01 5.78930564e-02
1.07316697e+00 1.57691404e-01 -4.28978622e-01 2.18372256e-01
-2.61394531e-01 5.38942218e-01 2.73467600e-01 8.72128367e-01
-1.27766803e-01 4.73102838e-01 -7.17434347e-01 1.30970550e+00
-1.41663581e-01 3.73090684e-01 -7.82125071e-02 1.11560738e+00
-3.16047609e-01 -1.15375829e+00 -5.75858772e-01 7.83727527e-01
-6.11315131e-01 8.98663849e-02 -3.42159003e-01 4.18937176e-01
2.10252270e-01 9.84010279e-01 -6.11496627e-01 -6.90525055e-01
1.52908489e-01 1.93453625e-01 1.63016230e-01 1.04040504e-02
-5.91830313e-01 -1.35727257e-01 -8.44501182e-02 -6.44917250e-01
-2.10761324e-01 -1.08811677e+00 -8.60344887e-01 -5.61872900e-01
9.52715054e-02 2.97420472e-01 7.55493045e-01 7.31732547e-01
5.93031943e-01 -3.17595333e-01 6.93158209e-01 -1.06068119e-01
-9.95786369e-01 -4.24780220e-01 -1.14367115e+00 4.71961468e-01
3.85116905e-01 -4.24126722e-02 -2.99702168e-01 1.10463994e-02] | [5.954610347747803, 6.624238014221191] |
2412269a-7102-4f76-a221-89b6fbf678bc | mixed-nondeterministic-probabilistic-automata | 2201.07474 | null | https://arxiv.org/abs/2201.07474v1 | https://arxiv.org/pdf/2201.07474v1.pdf | Mixed Nondeterministic-Probabilistic Automata: Blending graphical probabilistic models with nondeterminism | Graphical models in probability and statistics are a core concept in the area of probabilistic reasoning and probabilistic programming-graphical models include Bayesian networks and factor graphs. In this paper we develop a new model of mixed (nondeterministic/probabilistic) automata that subsumes both nondeterministic automata and graphical probabilistic models. Mixed Automata are equipped with parallel composition, simulation relation, and support message passing algorithms inherited from graphical probabilistic models. Segala's Probabilistic Automatacan be mapped to Mixed Automata. | ['Jean-Baptiste Raclet', 'Albert Benveniste'] | 2022-01-19 | null | null | null | null | ['probabilistic-programming'] | ['methodology'] | [-2.02491656e-01 3.59036505e-01 -1.22415133e-01 -4.86528993e-01
-5.39797843e-01 -7.23231077e-01 1.27430487e+00 -2.67659854e-02
-1.67234063e-01 7.65058994e-01 1.31743446e-01 -8.02872956e-01
-6.53053522e-01 -1.21031928e+00 -3.12379032e-01 -6.26778960e-01
-4.78521287e-01 1.13146758e+00 7.65591562e-01 -6.98536709e-02
-1.69836521e-01 9.66568291e-01 -1.32307482e+00 -1.20477453e-01
3.53808284e-01 3.37824924e-03 -4.10449237e-01 1.40364969e+00
-6.59700871e-01 8.62654984e-01 -1.28531396e-01 -4.99104857e-01
-3.65908384e-01 -3.43676209e-01 -4.95058447e-01 -5.19632101e-01
-2.10356295e-01 -1.36120379e-01 -7.18489528e-01 1.10635412e+00
-1.07103415e-01 6.74362481e-02 1.04324841e+00 -1.92700481e+00
-1.32166475e-01 1.49771369e+00 -3.09143215e-01 2.51846090e-02
6.44762278e-01 -2.61221945e-01 8.05615306e-01 -3.92504334e-01
2.42130592e-01 2.02196026e+00 4.39950228e-01 2.26557225e-01
-1.60540390e+00 -4.37524140e-01 5.45252636e-02 -1.84417903e-01
-1.51903510e+00 1.83528855e-01 3.14940125e-01 -4.13962245e-01
6.44651115e-01 4.67656255e-01 3.66362900e-01 7.22601831e-01
5.21787822e-01 8.75378966e-01 1.47785592e+00 -7.24567950e-01
4.12798136e-01 2.71348413e-02 1.08881855e+00 8.46322298e-01
5.38537860e-01 8.60866606e-01 -2.28336170e-01 -8.92917573e-01
6.86424673e-01 1.68689162e-01 3.86533171e-01 -3.65652263e-01
-8.18314493e-01 7.76882589e-01 -4.89911437e-01 2.75722116e-01
-2.99846809e-02 7.20208764e-01 9.64683364e-04 8.18972215e-02
-4.01138842e-01 -4.20357376e-01 -1.10806867e-01 -1.25968695e-01
-4.99756008e-01 7.41487503e-01 1.23551047e+00 1.04689574e+00
6.78571701e-01 1.34477079e-01 -6.35575056e-02 2.00199679e-01
1.15382707e+00 1.37645388e+00 -1.80993736e-01 -7.76362479e-01
-7.51959682e-02 2.01531827e-01 3.98599595e-01 -6.70194030e-01
-4.52808708e-01 2.80697316e-01 -6.22033060e-01 3.86331260e-01
6.13806903e-01 7.19847903e-02 -1.06816852e+00 1.61191332e+00
1.84359953e-01 5.62592149e-02 1.01281414e-02 3.16071600e-01
4.67035264e-01 8.84195089e-01 3.66606355e-01 -1.32312672e-02
1.36542344e+00 -2.19250121e-03 -1.03617728e+00 2.81581253e-01
4.43246037e-01 -5.19995749e-01 3.82215351e-01 5.17606735e-01
-8.34604800e-01 4.36842106e-02 -8.76570344e-01 4.63845491e-01
-5.21766901e-01 -4.75769430e-01 8.70570362e-01 1.41658974e+00
-1.07988155e+00 3.00713331e-01 -1.21708465e+00 -8.25602747e-03
-3.14305782e-01 3.62079710e-01 -2.94283599e-01 -9.24083441e-02
-1.28137517e+00 9.46765244e-01 5.86762369e-01 -1.20891906e-01
-1.05425143e+00 1.57657027e-01 -9.05599713e-01 -1.50227666e-01
2.42508128e-01 -4.37903166e-01 1.40880752e+00 -5.57795763e-02
-1.76740873e+00 1.64708450e-01 -1.49970189e-01 -4.98276949e-01
1.61732078e-01 -4.27866131e-02 -9.20394719e-01 -2.09800556e-01
-3.27291220e-01 7.99823646e-03 2.94974327e-01 -1.14442062e+00
-5.09361625e-01 -5.05325496e-01 -7.40850270e-02 -2.80656368e-01
8.22881639e-01 5.52339315e-01 -1.97240874e-01 -2.81074166e-01
4.13027823e-01 -9.82660532e-01 -4.97834116e-01 -6.44177079e-01
-7.48929858e-01 -2.98121423e-01 3.74111772e-01 -1.85927898e-02
1.26903796e+00 -1.80436575e+00 5.93854338e-02 9.00195181e-01
6.51671588e-02 -4.99945581e-01 1.76633403e-01 5.30426919e-01
4.34942916e-02 1.43705055e-01 -4.80076224e-01 2.28272945e-01
6.49634123e-01 5.77080548e-01 -4.53772187e-01 4.97148842e-01
-5.50460955e-03 6.56051099e-01 -8.22031260e-01 -7.75299370e-01
3.46673638e-01 1.03245452e-01 -9.30602252e-02 1.29367441e-01
-4.90928173e-01 -2.27768794e-01 -7.51401782e-01 6.03521228e-01
8.48828316e-01 3.66879046e-01 7.60942280e-01 7.58711696e-01
-1.95938975e-01 1.12311631e-01 -1.96452582e+00 1.08447790e+00
-9.81946066e-02 3.74784358e-02 -6.98904246e-02 7.14702308e-02
7.73679137e-01 4.24466997e-01 -2.18360543e-01 2.40580112e-01
4.50183362e-01 -3.35364938e-02 -1.13283731e-01 -8.62946436e-02
5.94051242e-01 -3.92080396e-01 -8.59903634e-01 1.10813296e+00
2.00045481e-01 -7.91696131e-01 2.50467002e-01 4.67627347e-01
9.95992303e-01 3.38314712e-01 3.44605207e-01 -4.24845457e-01
2.22111598e-01 -1.05847910e-01 2.79234380e-01 1.34792256e+00
1.86866209e-01 1.33172020e-01 8.68969858e-01 -2.78313130e-01
-7.32021987e-01 -2.03014255e+00 -1.06772646e-01 1.12522817e+00
-2.40371048e-01 -6.98913634e-01 -4.35379475e-01 -4.82411265e-01
-2.16099575e-01 9.87056136e-01 -4.24812078e-01 1.13415465e-01
-1.53516471e-01 -9.12411869e-01 9.51433957e-01 5.93210995e-01
-2.03226894e-01 -4.41196680e-01 -4.85689379e-03 3.13420743e-01
4.45046574e-01 -7.04507232e-01 -1.46169007e-01 3.44409049e-01
-9.63456154e-01 -9.42530036e-01 -6.04229458e-02 6.25636876e-02
1.27981439e-01 -1.03564732e-01 9.31425571e-01 -4.21858102e-01
3.12425226e-01 4.32985038e-01 3.35459337e-02 -7.06814706e-01
-1.13543880e+00 -4.36840028e-01 4.17077653e-02 -3.51223022e-01
3.90391916e-01 -6.04406714e-01 5.65430999e-01 4.30994689e-01
-1.19023287e+00 -2.15817600e-01 4.64118868e-01 6.53346241e-01
3.62619430e-01 2.56385773e-01 -3.95926714e-01 -1.10133958e+00
3.77027184e-01 -4.65565801e-01 -1.26807618e+00 2.85434395e-01
-3.58409762e-01 6.89209402e-01 -1.92601141e-02 -1.78334936e-01
-1.13882506e+00 2.40646914e-01 8.00357759e-02 2.28920579e-01
-5.13557255e-01 5.82799852e-01 -8.42397869e-01 3.08605701e-01
5.29230237e-01 -4.22622561e-02 -4.69654426e-02 -1.67241558e-01
7.42322326e-01 3.86679769e-01 3.36559683e-01 -1.05994010e+00
5.79826593e-01 5.07905960e-01 6.72482193e-01 -4.53410596e-01
-2.01350376e-01 -9.41211060e-02 -4.22602654e-01 -1.10576533e-01
5.98801196e-01 -4.53708380e-01 -7.84108102e-01 7.82810807e-01
-1.11567008e+00 2.98457146e-02 1.82881597e-02 8.42988372e-01
-5.36972702e-01 2.61779249e-01 -5.96953154e-01 -1.61911547e+00
4.72736388e-01 -1.09077251e+00 5.95585048e-01 2.84445196e-01
-4.21646595e-01 -1.04182196e+00 7.01221883e-01 -5.58041513e-01
-6.72095828e-03 1.04423873e-02 9.60666239e-01 -9.98208284e-01
-4.52676445e-01 -7.44118631e-01 -1.44170463e-01 -2.11988345e-01
-6.51804745e-01 8.27624738e-01 -7.24274635e-01 6.32914364e-01
-5.85606575e-01 3.03315699e-01 3.83148909e-01 2.90744394e-01
4.97984737e-01 -8.28450993e-02 -6.73801243e-01 -3.83893669e-01
1.31997931e+00 2.44073853e-01 6.72833443e-01 -3.30535352e-01
4.42711383e-01 3.41657490e-01 -3.10753789e-02 1.99097231e-01
7.67218292e-01 3.83014709e-01 4.37960438e-02 6.96971059e-01
4.44761217e-01 -5.02373159e-01 4.83337402e-01 5.99510431e-01
-1.94134876e-01 -2.21327484e-01 -1.55775023e+00 3.62580061e-01
-1.81266797e+00 -1.42128420e+00 -9.95841980e-01 2.28354788e+00
7.48158097e-01 5.56891143e-01 4.47231561e-01 3.41095120e-01
1.15637124e+00 -1.36714548e-01 5.58934033e-01 -8.13416779e-01
-1.95810273e-01 3.50239277e-01 5.66113889e-01 1.09543228e+00
-8.23389292e-01 5.25874913e-01 8.30431843e+00 7.82118201e-01
3.66539545e-02 2.81304598e-01 9.56198797e-02 4.41029340e-01
-9.17298317e-01 5.48173189e-01 -9.60048378e-01 2.19955429e-01
1.75137818e+00 -6.28208881e-03 2.13511243e-01 7.18100607e-01
-3.58276963e-01 -6.59832954e-01 -9.73834634e-01 4.47441965e-01
-3.70155603e-01 -9.30249214e-01 1.26187265e-01 2.40969900e-02
4.84716505e-01 -8.19955692e-02 -3.50622743e-01 3.91695231e-01
1.96478403e+00 -9.88808393e-01 9.52543139e-01 9.26388144e-01
1.73521861e-01 -8.88025820e-01 9.56451595e-01 2.14336231e-01
-9.82731640e-01 1.54526442e-01 1.23488985e-01 -6.96472228e-02
4.10367519e-01 8.31169188e-01 -5.14394641e-01 6.42114401e-01
3.26114357e-01 -2.75786757e-01 3.62259708e-02 9.02088702e-01
-5.76105237e-01 8.29024494e-01 -8.73838127e-01 -3.62437993e-01
7.06388205e-02 -3.34517956e-01 6.43020391e-01 1.34166145e+00
1.65913686e-01 2.72332728e-01 2.78626502e-01 8.27605426e-01
6.01532102e-01 -3.58990043e-01 -7.07633257e-01 -2.01438949e-01
6.54721498e-01 7.28012323e-01 -9.75834489e-01 -5.00951052e-01
-3.50999624e-01 -1.05052553e-01 -3.97030234e-01 1.97920576e-01
-7.40822256e-01 -4.54511791e-01 7.71239102e-02 -2.53363755e-02
-4.03920919e-01 -5.64821124e-01 -4.07197565e-01 -7.54371703e-01
-8.82596672e-01 -3.90296429e-01 8.79490793e-01 -8.22692454e-01
-1.05886853e+00 5.79825819e-01 9.69321191e-01 -5.33319712e-01
-5.28719366e-01 -7.08457887e-01 -6.02557540e-01 1.13100660e+00
-5.99045455e-01 -1.10759270e+00 3.49006414e-01 2.84017235e-01
-5.32895446e-01 7.81250969e-02 1.25227213e+00 -5.33574224e-01
-3.47301990e-01 5.91473393e-02 -9.26913880e-03 6.81990460e-02
-6.52038455e-02 -1.61698222e+00 5.19038856e-01 1.00376451e+00
1.50106624e-01 6.23420656e-01 1.21543038e+00 -6.62614822e-01
-1.57219231e+00 -5.89434743e-01 9.77054238e-01 -7.30243981e-01
9.98211443e-01 -3.94064426e-01 -6.29955173e-01 9.24509406e-01
-1.24011571e-02 -1.08214051e-01 6.00123882e-01 4.68967319e-01
-8.17406118e-01 2.50669569e-01 -1.22403479e+00 6.91199481e-01
5.15790820e-01 -7.60620773e-01 -9.35968280e-01 -2.46746942e-01
4.83905941e-01 -1.25640020e-01 -7.80090511e-01 3.31452489e-02
9.43808794e-01 -8.14619482e-01 5.84764421e-01 -4.99519467e-01
-4.35337573e-01 -6.70317531e-01 -4.29246128e-01 -7.45326936e-01
-1.99871510e-01 -9.01577652e-01 -1.70649230e-01 1.02984595e+00
5.65404415e-01 -9.00606930e-01 5.59560299e-01 9.70460415e-01
2.78894484e-01 1.87187746e-01 -9.31344271e-01 -7.22795010e-01
8.58270377e-02 -1.26119626e+00 1.17386723e+00 3.99414212e-01
3.64683419e-01 -1.91283792e-01 1.68557987e-01 5.39481938e-01
7.24833369e-01 6.23763353e-02 5.59222877e-01 -1.42453015e+00
-4.84045058e-01 -6.70240283e-01 -8.43078315e-01 -3.55802059e-01
3.02907169e-01 -5.47519803e-01 3.54133606e-01 -1.29279709e+00
2.52039015e-01 -5.46887934e-01 -1.95371062e-01 4.99484062e-01
5.68008907e-02 -2.38969058e-01 -2.53486544e-01 -3.34879935e-01
-4.83225793e-01 1.73470110e-01 3.77732784e-01 2.26037875e-02
4.22495417e-02 4.38204259e-01 -2.50909124e-02 7.82127917e-01
7.79344678e-01 -9.61718857e-01 -2.74815559e-01 1.99884519e-01
6.41217470e-01 7.33023703e-01 4.69873160e-01 -6.55580044e-01
1.38437539e-01 -6.52266979e-01 -2.04045027e-01 -8.41055632e-01
9.86632034e-02 -5.98205507e-01 8.84627223e-01 4.63587612e-01
-1.95992574e-01 2.62581587e-01 2.01757863e-01 8.21916580e-01
-7.53796846e-02 -3.77974838e-01 5.83801448e-01 2.53812056e-02
-3.29754144e-01 -1.33362622e-03 -1.43810081e+00 -3.98860335e-01
8.25927436e-01 8.23864564e-02 -3.21669161e-01 -3.27743888e-01
-1.34192002e+00 -1.42505288e-01 3.88105184e-01 -9.30851623e-02
4.84573156e-01 -1.13045168e+00 -5.01279533e-01 -2.14660466e-01
9.43081360e-03 -3.82181227e-01 2.67478228e-01 6.27293229e-01
-4.50310111e-01 4.68211234e-01 -8.06053802e-02 -3.87651891e-01
-8.31200898e-01 5.25614262e-01 2.28862524e-01 -1.82689980e-01
2.85429507e-02 6.88440204e-01 -3.78331065e-01 -8.48276854e-01
-2.01963335e-02 -5.37229478e-01 1.70830153e-02 -4.23676729e-01
8.04908693e-01 7.37824798e-01 -3.09144557e-01 -3.20614487e-01
-5.63331544e-01 1.11132348e-02 7.35617504e-02 -1.07899666e+00
8.83165479e-01 3.31888385e-02 -6.93267584e-01 1.23371255e+00
2.47552574e-01 1.82064012e-01 -4.57274139e-01 -1.28977850e-01
4.28869724e-01 -5.67405820e-02 -2.57153660e-02 -7.58778751e-01
-2.16276184e-01 7.13421464e-01 5.22295952e-01 6.04110897e-01
3.99131536e-01 1.98099941e-01 -1.29489169e-01 1.24524795e-01
7.73636937e-01 -5.40374935e-01 -9.35546637e-01 5.72040617e-01
4.72238243e-01 -5.72912216e-01 7.07738660e-03 -4.47840899e-01
-4.19532031e-01 1.17779183e+00 6.82134107e-02 -1.59641027e-01
1.31751931e+00 1.04204476e+00 -2.96583027e-01 -3.00793469e-01
-8.55906487e-01 -3.11171263e-01 1.18757680e-01 7.47406304e-01
1.67477131e-01 9.29648459e-01 -1.51833907e-01 1.34214699e+00
-6.87100142e-02 1.19974375e-01 9.43481624e-01 1.32108200e+00
-4.57565397e-01 -1.29264331e+00 -9.57241416e-01 3.46700698e-01
-1.53195590e-01 2.11502984e-02 -1.12793811e-01 7.58589506e-01
-3.32911462e-01 1.08182955e+00 -8.21351353e-03 -4.21627164e-01
5.04812561e-02 3.34369123e-01 6.79276824e-01 -6.48245394e-01
-1.47310168e-01 -1.02993049e-01 3.68795693e-01 -2.61616588e-01
-2.04873800e-01 -9.66750622e-01 -1.27667665e+00 -7.25769997e-01
-6.68331683e-01 6.73290133e-01 9.52048779e-01 1.10449398e+00
-1.60863429e-01 2.44343475e-01 7.93659836e-02 -2.24585369e-01
-4.69698012e-01 -9.38196898e-01 -1.16475344e+00 -8.03997815e-01
-2.12868989e-01 -5.94966292e-01 -2.42371604e-01 -2.31043518e-01] | [8.368379592895508, 6.2983832359313965] |
9f715cbe-de44-432f-b6c6-192a5f9134b1 | visual-slam-with-graph-cut-optimized-multi | 2108.04281 | null | https://arxiv.org/abs/2108.04281v2 | https://arxiv.org/pdf/2108.04281v2.pdf | Visual SLAM with Graph-Cut Optimized Multi-Plane Reconstruction | This paper presents a semantic planar SLAM system that improves pose estimation and mapping using cues from an instance planar segmentation network. While the mainstream approaches are using RGB-D sensors, employing a monocular camera with such a system still faces challenges such as robust data association and precise geometric model fitting. In the majority of existing work, geometric model estimation problems such as homography estimation and piece-wise planar reconstruction (PPR) are usually solved by standard (greedy) RANSAC separately and sequentially. However, setting the inlier-outlier threshold is difficult in absence of information about the scene (i.e. the scale). In this work, we revisit these problems and argue that two mentioned geometric models (homographies/3D planes) can be solved by minimizing an energy function that exploits the spatial coherence, i.e. with graph-cut optimization, which also tackles the practical issue when the output of a trained CNN is inaccurate. Moreover, we propose an adaptive parameter setting strategy based on our experiments, and report a comprehensive evaluation on various open-source datasets. | ['Didier Stricker', 'Alain Pagani', 'Jason Rambach', 'Yaxu Xie', 'Fangwen Shu'] | 2021-08-09 | null | null | null | null | ['homography-estimation'] | ['computer-vision'] | [ 2.93102562e-01 1.02757521e-01 -5.26332892e-02 -3.78262758e-01
-6.77612603e-01 -6.78574681e-01 3.43851715e-01 4.37633209e-02
-4.12006885e-01 4.34554100e-01 -3.21796656e-01 -3.61282714e-02
-4.26080436e-01 -5.92160463e-01 -1.07583213e+00 -5.07636368e-01
4.32481170e-01 9.93830442e-01 1.54155537e-01 -3.41361687e-02
6.75973475e-01 8.06979120e-01 -1.27408314e+00 -5.44956803e-01
9.41926956e-01 1.23039651e+00 1.62123591e-01 4.20626372e-01
6.55362234e-02 2.42139146e-01 -1.47973359e-01 -3.04238737e-01
6.25374913e-01 1.58247277e-02 -4.75904673e-01 3.88461798e-01
9.02463615e-01 -3.01537722e-01 -2.05967054e-01 1.25733113e+00
3.30217510e-01 1.67912886e-01 4.20177311e-01 -1.36850524e+00
-1.14485763e-01 3.26449797e-03 -8.25113237e-01 -5.36003232e-01
4.09213126e-01 -1.78151026e-01 8.75820339e-01 -9.02982831e-01
7.90526986e-01 1.14102113e+00 8.57110262e-01 -9.49599501e-03
-1.06914616e+00 -4.61620599e-01 2.55592912e-01 9.53496993e-02
-1.57221580e+00 -2.67566532e-01 1.21955943e+00 -3.60817283e-01
6.25347972e-01 2.60753781e-01 6.81354523e-01 7.45095611e-01
-1.27584785e-01 5.56517065e-01 9.13964450e-01 -2.60306895e-01
2.02153563e-01 -1.22586951e-01 -5.55295497e-02 6.30134046e-01
5.07420242e-01 -1.66087568e-01 -5.92782497e-01 -3.54679413e-02
1.09866560e+00 8.16824734e-02 -3.36475253e-01 -1.23946571e+00
-1.18605328e+00 6.93344355e-01 6.66806221e-01 4.27794382e-02
-2.60787725e-01 2.81156957e-01 -2.51078531e-02 -1.48090333e-01
3.66528869e-01 6.64057016e-01 -3.95211488e-01 1.42334774e-01
-1.02567160e+00 1.30387008e-01 6.95636749e-01 1.34704578e+00
1.26443219e+00 -2.03100055e-01 6.48537278e-01 5.92234969e-01
4.51711029e-01 6.23064280e-01 -3.55857089e-02 -1.14140749e+00
6.52870655e-01 6.79714799e-01 2.36878023e-01 -1.67510164e+00
-7.84747660e-01 -3.72425556e-01 -7.32470512e-01 -8.92932639e-02
6.47654891e-01 4.68506739e-02 -7.01152205e-01 1.42761981e+00
5.47906876e-01 2.89871871e-01 -3.03177267e-01 1.12707746e+00
5.28145850e-01 7.46520758e-02 -6.05940282e-01 -5.15793785e-02
1.03752458e+00 -1.00780773e+00 -4.92726386e-01 -4.60568666e-01
4.67812538e-01 -9.23396826e-01 5.89339554e-01 5.26964307e-01
-7.76602507e-01 -1.72203511e-01 -1.15101576e+00 -5.57917893e-01
-1.70985878e-01 3.12940061e-01 5.93646884e-01 2.82210976e-01
-9.47872519e-01 5.26076198e-01 -8.30172181e-01 -7.09089994e-01
-2.89237350e-02 5.09015620e-01 -6.02672994e-01 -2.25755617e-01
-5.15672505e-01 7.48952627e-01 2.79652774e-01 4.68021959e-01
-4.40268785e-01 -5.51280200e-01 -9.63436902e-01 -2.75126457e-01
8.52706194e-01 -9.24460471e-01 7.85723567e-01 -6.31160796e-01
-1.51784396e+00 9.48195696e-01 -9.68138054e-02 -4.31059152e-01
9.49206889e-01 -5.86657882e-01 4.09889251e-01 1.79334596e-01
-4.40252125e-02 5.75703740e-01 7.51610756e-01 -1.56804359e+00
-2.73993313e-01 -8.64013672e-01 9.63726640e-02 5.36669135e-01
1.89734772e-01 -4.52288568e-01 -8.49091470e-01 -1.86536849e-01
1.19920647e+00 -1.13291872e+00 -4.98731881e-01 1.81972474e-01
-7.85704017e-01 1.78998590e-01 6.49208665e-01 -7.39336610e-01
7.43884206e-01 -1.92178333e+00 2.47847214e-01 5.89055181e-01
1.52272508e-01 -2.58725733e-01 1.62216708e-01 2.71903604e-01
3.24027896e-01 -1.21511295e-01 -2.55207360e-01 -6.89623952e-01
-5.64425774e-02 2.84543574e-01 -5.31416237e-02 1.12890387e+00
-1.72397912e-01 6.18764937e-01 -6.93304777e-01 -4.69229162e-01
6.96745753e-01 5.88969529e-01 -5.78173995e-01 2.28510126e-02
-1.52018160e-01 7.53998637e-01 -2.77968913e-01 7.10107505e-01
1.12663651e+00 -9.72310528e-02 6.26692921e-02 -5.47492206e-01
-3.26629370e-01 7.56202936e-02 -1.69012845e+00 2.29598522e+00
-4.03492332e-01 3.83512318e-01 2.12863326e-01 -8.88875902e-01
1.03022468e+00 -2.29776323e-01 6.49597824e-01 -5.05373478e-01
3.51931214e-01 4.53015655e-01 -4.25462306e-01 -2.18462795e-01
6.95617616e-01 3.55070561e-01 1.54514790e-01 -2.02516586e-01
-1.26982465e-01 -4.99129355e-01 -2.92224556e-01 -8.16102847e-02
7.77863443e-01 6.33160591e-01 3.70852441e-01 -1.66296765e-01
4.09324229e-01 3.09706748e-01 6.12454057e-01 5.69217801e-01
5.70375286e-02 1.05781662e+00 2.92460799e-01 -3.84985238e-01
-1.08272851e+00 -5.74813306e-01 -1.02154143e-01 2.00574756e-01
8.27044368e-01 -2.95318455e-01 -6.63112283e-01 -2.91730702e-01
7.99256936e-02 3.63925129e-01 -3.58846635e-01 1.40566334e-01
-7.88512051e-01 -5.14277160e-01 7.66359791e-02 3.52307528e-01
5.49244285e-01 -3.06653023e-01 -7.95257866e-01 8.08666721e-02
-2.42487490e-01 -1.56560075e+00 -2.34238401e-01 2.05806047e-01
-9.63510752e-01 -1.29901540e+00 -4.90117311e-01 -4.01961863e-01
8.73911142e-01 5.78448117e-01 8.51336777e-01 4.57844585e-02
-4.84211817e-02 3.16194415e-01 -1.54477537e-01 4.97434475e-02
3.30336958e-01 2.60712177e-01 4.32802737e-02 1.27520308e-01
-5.98264597e-02 -6.05924189e-01 -7.16039956e-01 5.35237253e-01
-4.91685629e-01 3.41906518e-01 5.89251459e-01 4.49587762e-01
1.10480416e+00 -2.01059774e-01 -1.87120005e-01 -8.32972586e-01
-1.68898880e-01 -2.05388635e-01 -1.07893503e+00 1.90201864e-01
-5.38270354e-01 -6.39129728e-02 3.09239298e-01 -1.24982379e-01
-8.43982577e-01 7.25425124e-01 4.63820063e-02 -7.63794005e-01
-1.79174379e-01 2.51168519e-01 -4.25335735e-01 -7.06774652e-01
2.81345040e-01 7.37701654e-02 -2.02357143e-01 -5.34781516e-01
5.08136332e-01 2.59091079e-01 6.30110562e-01 -3.66721213e-01
1.00686550e+00 9.07325864e-01 6.24702334e-01 -7.62047887e-01
-8.32190514e-01 -8.71801794e-01 -1.08538663e+00 -2.36185431e-01
8.75856280e-01 -8.61214995e-01 -8.00732493e-01 2.59782791e-01
-1.28549707e+00 -2.04610173e-02 1.80655405e-01 5.83453059e-01
-8.60343814e-01 6.94956541e-01 -2.09595382e-01 -9.24126029e-01
4.78837304e-02 -1.22528970e+00 1.54482234e+00 4.71457951e-02
-4.51906100e-02 -8.19971383e-01 -6.41830871e-03 6.25583470e-01
3.63639742e-02 6.21256351e-01 3.54191124e-01 -2.30805576e-01
-1.21714914e+00 -1.85074717e-01 -3.33135098e-01 -2.12596357e-01
-2.28846326e-01 -6.28704950e-02 -9.61109340e-01 -1.28957093e-01
1.69923946e-01 5.69795072e-02 5.05343556e-01 4.52953666e-01
9.35116291e-01 -1.01210579e-01 -3.21630120e-01 1.25042415e+00
1.71437967e+00 -8.15008506e-02 4.04568523e-01 6.65720224e-01
1.29822052e+00 6.14867449e-01 8.19879115e-01 3.38706315e-01
6.73908114e-01 9.76853788e-01 1.06135380e+00 1.90638863e-02
6.49523661e-02 -4.47127491e-01 -2.39922166e-01 7.22804427e-01
-7.38009736e-02 -2.57734582e-02 -9.89718080e-01 3.23764294e-01
-2.12961602e+00 -2.20416382e-01 -4.77555126e-01 2.46923232e+00
2.90138632e-01 -1.23321623e-01 -2.39427522e-01 -1.59784034e-02
5.69181025e-01 -5.56446053e-02 -6.91762030e-01 1.03624180e-01
-3.18408370e-01 -2.73099661e-01 1.12811780e+00 7.99374104e-01
-1.05650342e+00 1.09884799e+00 4.91870213e+00 5.49645603e-01
-1.15330398e+00 -1.55208647e-01 2.20368370e-01 1.86596557e-01
-1.50623009e-01 3.92524928e-01 -8.37328374e-01 1.65985059e-02
3.13757271e-01 4.06392336e-01 5.48677266e-01 9.07521427e-01
5.83210588e-02 -4.67525333e-01 -1.20788836e+00 1.46625841e+00
4.34920043e-01 -9.85700130e-01 -1.30666718e-01 2.07582176e-01
7.37147272e-01 1.14340208e-01 -1.36803836e-01 -2.82539785e-01
-1.13969631e-01 -7.82369971e-01 8.47261906e-01 5.26353538e-01
4.86567080e-01 -6.84823394e-01 6.66302264e-01 6.11009181e-01
-1.21949911e+00 1.67846397e-01 -4.97125596e-01 4.81270216e-02
2.40833551e-01 7.47667909e-01 -6.40866101e-01 1.13626027e+00
5.80572426e-01 7.88739324e-01 -6.60114467e-01 1.38907361e+00
-2.76251644e-01 -1.33729592e-01 -7.62026310e-01 3.56745988e-01
1.37769356e-01 -6.68828130e-01 6.04485631e-01 7.53779233e-01
4.75175828e-01 -4.43663681e-03 2.20399246e-01 8.70555937e-01
7.16164187e-02 2.62981385e-01 -6.05444551e-01 5.08131742e-01
3.84411305e-01 1.36468732e+00 -1.03343189e+00 2.39231531e-02
-2.52451539e-01 9.57911015e-01 3.16001654e-01 2.10555196e-01
-6.16037071e-01 3.24612856e-02 2.69508600e-01 2.05551594e-01
6.81171985e-03 -6.69554293e-01 -6.99606299e-01 -1.37325358e+00
2.34454110e-01 -4.59034830e-01 6.16283864e-02 -9.57323074e-01
-8.22733045e-01 3.57650280e-01 -4.10779566e-02 -1.22264552e+00
-2.12723538e-02 -6.46546066e-01 -1.13218859e-01 5.35400450e-01
-1.68568730e+00 -1.13618994e+00 -7.64485657e-01 4.43267226e-01
3.55587363e-01 5.42787552e-01 2.76431859e-01 2.84515440e-01
-4.60087061e-01 2.32509598e-01 -5.01804203e-02 -8.64296332e-02
7.24226594e-01 -1.14765751e+00 1.13557681e-01 9.20357347e-01
1.41345888e-01 6.30068898e-01 7.50230908e-01 -7.08068788e-01
-1.88647461e+00 -9.30384338e-01 7.59279370e-01 -3.76635760e-01
5.34085393e-01 -5.12728214e-01 -6.56779408e-01 8.13370228e-01
-2.42922261e-01 2.57205479e-02 1.11036971e-02 -2.25455612e-01
-1.20208755e-01 -7.48999976e-03 -1.12031865e+00 4.46211785e-01
1.28831470e+00 -2.84758449e-01 5.55288419e-03 4.49390024e-01
6.05224967e-01 -1.05741072e+00 -6.50611579e-01 5.77505291e-01
4.13376123e-01 -1.09480739e+00 1.08295941e+00 -6.66246116e-02
-3.53473499e-02 -6.39779627e-01 -4.58232731e-01 -1.01088130e+00
1.80749729e-01 -6.45673692e-01 6.48444518e-02 1.02166831e+00
-1.16767116e-01 -5.36838770e-01 9.09418404e-01 6.41475797e-01
-2.53457427e-01 -6.13791525e-01 -1.11145914e+00 -7.69951820e-01
-4.35594231e-01 -6.72015131e-01 5.26910603e-01 9.95035052e-01
-5.40685534e-01 1.80328399e-01 -4.95083928e-01 6.82735920e-01
9.69386876e-01 2.47711629e-01 1.22657478e+00 -1.34495950e+00
-4.42638844e-02 -2.45530427e-01 -6.37227356e-01 -1.48194492e+00
6.60088658e-02 -5.71180463e-01 2.91760087e-01 -1.47515845e+00
-1.94935650e-01 -5.82815647e-01 3.71450633e-01 6.75778612e-02
2.97941238e-01 2.56745756e-01 2.03179136e-01 3.82759541e-01
-7.31282413e-01 4.71540302e-01 1.07893240e+00 1.40951470e-01
-5.74142784e-02 -1.16890848e-01 -2.96004653e-01 1.20079482e+00
5.19029677e-01 -3.09419870e-01 -2.38640949e-01 -8.10620129e-01
6.73183680e-01 1.86760336e-01 5.53060234e-01 -1.11226285e+00
5.91963828e-01 -2.22467273e-01 2.57807404e-01 -9.72563446e-01
6.84035122e-01 -1.28417945e+00 4.01117951e-01 1.14445321e-01
3.16714235e-02 1.06995113e-01 -1.63021699e-01 6.49185836e-01
-8.43204409e-02 -2.36520931e-01 6.84778273e-01 -1.81506827e-01
-6.24939501e-01 4.87571329e-01 4.44070905e-01 -2.13212833e-01
9.56688404e-01 -4.36651558e-01 -1.19214773e-01 -3.73320907e-01
-6.31828845e-01 2.67817885e-01 1.03902078e+00 1.60211265e-01
6.49291158e-01 -1.22224295e+00 -2.62698799e-01 1.53405190e-01
2.07355633e-01 7.52169669e-01 1.77761480e-01 1.21768260e+00
-8.97032619e-01 4.29551810e-01 6.73643947e-02 -1.05313623e+00
-1.05031097e+00 4.21600997e-01 5.00778556e-01 -1.49458311e-02
-6.23909950e-01 7.05063224e-01 2.11988375e-01 -6.08432829e-01
3.84681523e-01 -4.83413875e-01 -1.63448241e-03 -9.28063765e-02
-1.38277307e-01 4.78823602e-01 2.10364789e-01 -1.00778246e+00
-5.26003718e-01 1.31648469e+00 3.47496450e-01 -6.41864398e-03
1.27250826e+00 -5.66810012e-01 -3.08882713e-01 3.71893555e-01
1.15624213e+00 -7.98522383e-02 -1.30761695e+00 -2.82160193e-01
4.16277573e-02 -7.11071014e-01 1.51400559e-03 -1.88038483e-01
-1.08223176e+00 9.38664556e-01 3.77281517e-01 -1.27725586e-01
7.62476802e-01 -1.49256811e-01 7.19583869e-01 4.83070582e-01
7.49383092e-01 -1.06692660e+00 -1.28962293e-01 6.02706850e-01
8.74587119e-01 -1.27929437e+00 2.34521911e-01 -8.83702695e-01
-1.94940969e-01 1.28259468e+00 6.15230501e-01 -4.20175374e-01
4.70572829e-01 -8.79220665e-02 -4.42349911e-02 -2.62653261e-01
1.25114292e-01 -1.46656841e-01 4.13182706e-01 2.94428021e-01
-6.27254173e-02 1.16711250e-02 -1.65409327e-01 1.10007405e-01
-5.60405731e-01 -2.57275313e-01 4.01099414e-01 7.61573970e-01
-3.39240164e-01 -8.40139449e-01 -6.94102883e-01 -7.43874684e-02
1.82802957e-02 1.88153103e-01 -4.82837141e-01 8.25844526e-01
2.26550534e-01 7.58746564e-01 2.27721054e-02 -2.85948843e-01
3.96633476e-01 -2.39770308e-01 6.61670029e-01 -4.50282961e-01
-1.03313535e-01 3.89715850e-01 -1.48018822e-01 -9.48850155e-01
-6.76694810e-01 -7.78827906e-01 -1.12889302e+00 -6.23014905e-02
-5.63259423e-01 -2.84103930e-01 1.17777371e+00 1.05191088e+00
2.14027718e-01 5.61955944e-02 4.12466884e-01 -1.11840343e+00
-3.43460917e-01 -6.70822382e-01 -5.34434915e-01 2.44602203e-01
3.18602353e-01 -8.35084081e-01 -4.18644190e-01 -3.23031008e-01] | [7.7036614418029785, -2.5118346214294434] |
35e65d86-b312-4c62-a631-c637b4f4ad8a | modeling-3d-surface-manifolds-with-a-locally | 2102.05984 | null | https://arxiv.org/abs/2102.05984v1 | https://arxiv.org/pdf/2102.05984v1.pdf | Modeling 3D Surface Manifolds with a Locally Conditioned Atlas | Recently proposed 3D object reconstruction methods represent a mesh with an atlas - a set of planar patches approximating the surface. However, their application in a real-world scenario is limited since the surfaces of reconstructed objects contain discontinuities, which degrades the quality of the final mesh. This is mainly caused by independent processing of individual patches, and in this work, we postulate to mitigate this limitation by preserving local consistency around patch vertices. To that end, we introduce a Locally Conditioned Atlas (LoCondA), a framework for representing a 3D object hierarchically in a generative model. Firstly, the model maps a point cloud of an object into a sphere. Secondly, by leveraging a spherical prior, we enforce the mapping to be locally consistent on the sphere and on the target object. This way, we can sample a mesh quad on that sphere and project it back onto the object's manifold. With LoCondA, we can produce topologically diverse objects while maintaining quads to be stitched together. We show that the proposed approach provides structurally coherent reconstructions while producing meshes of quality comparable to the competitors. | ['Kacper Kania', 'Tomasz Trzciński', 'Maciej Zięba', 'Sebastian Winczowski', 'Przemysław Spurek'] | 2021-02-11 | null | null | null | null | ['3d-object-reconstruction'] | ['computer-vision'] | [ 2.13951260e-01 4.81019199e-01 4.11119074e-01 -1.60624143e-02
-7.48495460e-01 -4.65888768e-01 5.86063087e-01 1.28129184e-01
2.62297630e-01 3.03305715e-01 -3.56862396e-02 3.98036659e-01
2.07319595e-02 -1.05672765e+00 -1.12835431e+00 -7.21315503e-01
3.80089015e-01 9.92954552e-01 5.68452358e-01 1.24768704e-01
1.71093911e-01 9.78546917e-01 -1.50709856e+00 2.17164323e-01
8.87887180e-01 9.14738953e-01 2.55254418e-01 1.75770432e-01
-1.23805307e-01 1.03485607e-01 -2.82950521e-01 -5.50230384e-01
3.69481832e-01 -4.03388679e-01 -6.04580820e-01 5.81780195e-01
5.89078367e-01 1.90006569e-02 -9.04558375e-02 1.13786352e+00
-6.55989349e-02 -2.20768645e-01 1.07073104e+00 -8.92246425e-01
-2.82545507e-01 -1.08466344e-03 -6.19597912e-01 -7.41227031e-01
3.22024643e-01 -2.17565820e-01 7.83154488e-01 -1.27404213e+00
9.71522987e-01 1.17852378e+00 8.03003430e-01 3.55816662e-01
-1.75659859e+00 -1.83748320e-01 -3.11772116e-02 -4.30864334e-01
-1.52264929e+00 -4.61158156e-01 1.24637687e+00 -6.82429016e-01
2.72778749e-01 5.18132150e-01 8.14742506e-01 7.05158412e-01
4.51837063e-01 4.00584459e-01 9.15346563e-01 -2.84591913e-01
5.34016371e-01 1.65700182e-01 -2.25151405e-01 5.44431567e-01
1.39280677e-01 -4.01013166e-01 -4.39549863e-01 -4.59138095e-01
1.28386247e+00 6.31937310e-02 -3.75576615e-01 -1.08870828e+00
-1.16575503e+00 4.88680243e-01 5.04042327e-01 2.05070332e-01
-5.63574672e-01 -7.51287118e-02 -5.21216020e-02 5.72964288e-02
7.66904712e-01 2.11864471e-01 1.73245043e-01 4.96241957e-01
-1.06728423e+00 3.67136359e-01 7.47824371e-01 1.00017238e+00
9.51760232e-01 -2.49926984e-01 2.44680673e-01 6.73283160e-01
5.74813366e-01 4.28384483e-01 -3.70560765e-01 -1.12051821e+00
1.02492742e-01 6.91065907e-01 2.30709508e-01 -1.38928139e+00
-1.32700186e-02 -3.27188462e-01 -1.08291221e+00 4.96658057e-01
1.38580009e-01 7.05275655e-01 -8.26931059e-01 1.41579914e+00
7.38442481e-01 3.79400551e-01 -2.34333500e-01 8.96168292e-01
4.87901479e-01 6.76622629e-01 -3.84531140e-01 -2.90706784e-01
1.12700999e+00 -5.79133987e-01 -4.42634284e-01 9.42377746e-02
-5.95356198e-03 -8.08910310e-01 8.41323972e-01 5.61200857e-01
-1.49785483e+00 -5.59648931e-01 -9.67291713e-01 -1.16761662e-02
4.03216720e-01 -2.13913664e-01 -1.55929312e-01 2.09206179e-01
-9.48792458e-01 7.20116198e-01 -1.06255352e+00 -1.04452088e-01
4.48986322e-01 -7.62916915e-03 -5.03039896e-01 -1.51105061e-01
-4.92700309e-01 6.68099523e-01 -1.69779342e-02 7.29654077e-03
-7.78897107e-01 -8.28781247e-01 -6.92691982e-01 -6.13757521e-02
5.40175810e-02 -9.86461341e-01 6.85470402e-01 -7.47457027e-01
-1.41080523e+00 1.07854497e+00 -2.39657134e-01 -5.12276180e-02
7.89106905e-01 2.18173981e-01 6.51204512e-02 1.93325952e-01
1.73994094e-01 3.58420253e-01 1.19889915e+00 -2.10013294e+00
-1.31219298e-01 -5.68267882e-01 -2.64130533e-01 5.81846163e-02
1.02525517e-01 -4.48474169e-01 -6.97470129e-01 -7.88000584e-01
7.56962895e-01 -8.23598981e-01 -2.76285112e-01 4.06059235e-01
-4.43195075e-01 -1.87838208e-02 8.72350335e-01 -6.60411358e-01
8.09585035e-01 -2.56813526e+00 7.69360483e-01 7.44438708e-01
4.39853400e-01 -2.90222019e-01 7.19804689e-02 4.48342770e-01
1.59852594e-01 -8.84202495e-02 -7.10631788e-01 -9.36739683e-01
-1.66079536e-01 2.99410433e-01 -1.71382919e-01 8.40585768e-01
1.46176472e-01 5.98976731e-01 -7.08133996e-01 -4.53908682e-01
1.88837811e-01 7.17861414e-01 -8.93754900e-01 2.63628840e-01
-4.57061380e-01 8.18358362e-01 -5.69400132e-01 4.75605309e-01
1.06229758e+00 -1.53520778e-01 4.89034168e-02 -3.87447327e-01
-8.60669538e-02 3.24917920e-02 -1.48271716e+00 1.90250587e+00
-3.76088977e-01 -1.90763567e-02 5.69007635e-01 -9.49430883e-01
1.15843177e+00 3.52028638e-01 6.05090201e-01 -1.98325977e-01
7.09483447e-03 4.83144760e-01 -6.64326966e-01 3.65577526e-02
1.06471494e-01 -3.23289275e-01 3.06941211e-01 1.28485233e-01
-3.19894016e-01 -6.41615570e-01 -4.01103199e-01 -1.84604828e-03
8.67115021e-01 2.09205866e-01 -1.48655787e-01 -4.86191213e-01
4.16595787e-01 -4.68287803e-02 4.43109691e-01 1.98369384e-01
5.17501175e-01 1.25622010e+00 3.04284334e-01 -2.67447352e-01
-1.36905575e+00 -1.53900254e+00 -5.79075158e-01 4.21224050e-02
4.00004297e-01 -3.27871829e-01 -9.38923299e-01 -3.83157521e-01
1.70900494e-01 5.01472592e-01 -6.43257320e-01 -1.04143888e-01
-6.60351813e-01 -1.17411755e-01 -1.59263924e-01 1.04131848e-01
8.78720731e-02 -7.67344117e-01 -5.00927866e-01 3.84090930e-01
-2.28841364e-01 -8.02851081e-01 -3.40377539e-01 -2.05580264e-01
-1.11298013e+00 -8.76106977e-01 -8.98385763e-01 -9.64596212e-01
1.13515580e+00 1.97400287e-01 1.19597924e+00 9.34095234e-02
6.89583868e-02 3.33692461e-01 -9.20143127e-02 2.44351506e-01
-7.39598989e-01 -3.48915547e-01 -3.97922583e-02 6.13539457e-01
-5.52091599e-01 -1.10072148e+00 -5.22054076e-01 4.95430499e-01
-1.06036901e+00 4.15052235e-01 3.16861331e-01 5.37172616e-01
1.23026431e+00 1.66085124e-01 1.57384336e-01 -7.16912508e-01
-2.65566856e-02 -5.68256617e-01 -5.34943044e-01 7.85925388e-02
-8.02653134e-02 -1.27889693e-01 3.83142233e-01 -3.28249931e-01
-8.56232285e-01 3.60985577e-01 -1.96264744e-01 -9.74268258e-01
-9.28660482e-02 2.47646451e-01 -4.92129236e-01 -9.74837169e-02
4.49208498e-01 3.39278907e-01 2.25779861e-01 -9.90291715e-01
1.47504851e-01 3.69734287e-01 5.65159738e-01 -6.39105678e-01
9.47687387e-01 9.48914587e-01 2.94844270e-01 -9.62741256e-01
-4.52344000e-01 -1.73538357e-01 -7.62772620e-01 -3.32661897e-01
8.17564011e-01 -6.60156071e-01 -3.14195365e-01 3.96843404e-01
-1.27259529e+00 -1.84965432e-01 -5.33097982e-01 1.58857629e-01
-9.38070059e-01 4.44786996e-01 -3.19096237e-01 -6.34662390e-01
1.60904564e-02 -1.01770771e+00 1.38676941e+00 -2.52878815e-01
-2.23999634e-01 -8.18006933e-01 2.99458295e-01 2.38661930e-01
1.00424454e-01 5.91773927e-01 9.06839073e-01 9.05504450e-02
-8.03259194e-01 -1.54098064e-01 3.68609689e-02 3.71861398e-01
1.81629330e-01 1.07182801e-01 -7.00977802e-01 -3.05161476e-01
5.04329979e-01 1.30354896e-01 3.95321965e-01 3.85242403e-01
9.68205214e-01 -2.54779607e-01 -5.34905791e-01 5.25700688e-01
1.46439707e+00 -2.33748883e-01 7.20994830e-01 3.55791226e-02
7.21061826e-01 7.94549584e-01 2.66375780e-01 3.51125538e-01
2.07314149e-01 8.98632288e-01 6.16843879e-01 -2.68105604e-02
-2.50690579e-01 -3.30712616e-01 9.75841954e-02 1.00756621e+00
-1.72535911e-01 -2.80810036e-02 -8.18453193e-01 5.71918070e-01
-1.74947882e+00 -4.98225063e-01 -5.39390445e-01 2.42778778e+00
7.78853059e-01 1.54182985e-01 4.14377041e-02 1.60240263e-01
7.33518004e-01 -1.76404506e-01 -2.92251348e-01 5.88702112e-02
-4.41847593e-02 1.46616265e-01 -1.48515165e-01 7.45700359e-01
-5.81043482e-01 5.24417698e-01 5.45032358e+00 7.93787539e-01
-9.91077900e-01 1.27692819e-01 3.82253379e-01 2.29447365e-01
-7.68461883e-01 1.21653363e-01 -4.07091409e-01 5.32956362e-01
4.41757292e-01 -8.69169179e-03 1.59935787e-01 6.81308746e-01
4.51886728e-02 1.11281067e-01 -1.22129989e+00 7.87712038e-01
4.69144696e-04 -1.40797746e+00 4.01981443e-01 3.17329139e-01
9.64873195e-01 -2.61226743e-01 -1.35989100e-01 -4.15067762e-01
-1.81228980e-01 -8.95862281e-01 1.17407501e+00 9.45171535e-01
7.32852519e-01 -7.60678113e-01 3.08967143e-01 5.69452763e-01
-1.18362665e+00 7.20451057e-01 -3.98160994e-01 4.09006745e-01
3.82506102e-01 9.92748916e-01 -4.30114478e-01 8.60403001e-01
6.28341734e-01 5.76950431e-01 -2.02799782e-01 1.09947693e+00
2.03923523e-01 3.11529666e-01 -5.00556469e-01 6.64219797e-01
-8.01565349e-02 -7.03489304e-01 8.88502657e-01 7.21781492e-01
4.58230317e-01 2.06187293e-01 2.97173262e-01 1.22633970e+00
-1.66584589e-02 2.57958412e-01 -6.67558610e-01 4.53576982e-01
3.36323380e-01 1.00450885e+00 -9.04878378e-01 -2.70981967e-01
-2.91474462e-01 1.17876983e+00 3.87560934e-01 1.54081821e-01
-6.66677356e-01 2.73869455e-01 4.10635531e-01 7.09643364e-01
3.10585916e-01 -2.93266535e-01 -5.79731405e-01 -1.05057836e+00
4.74695683e-01 -6.53635859e-01 -2.24822775e-01 -7.93690264e-01
-1.33792746e+00 6.03369534e-01 -1.66780561e-01 -1.44641328e+00
4.54950273e-01 -1.93253517e-01 -4.88288224e-01 8.83209527e-01
-9.59566832e-01 -1.09436750e+00 -2.70530999e-01 5.37931561e-01
3.21332246e-01 3.63532335e-01 6.75045669e-01 2.09053263e-01
-6.62575886e-02 3.73861082e-02 1.62590936e-01 -3.38938951e-01
3.84660125e-01 -9.96737063e-01 3.34822536e-01 5.18175006e-01
1.77253619e-01 5.52825570e-01 6.84286356e-01 -9.22327995e-01
-1.42798185e+00 -1.15696037e+00 6.76143765e-01 -5.10778487e-01
1.55888110e-01 -4.48279023e-01 -1.47301793e+00 5.64708292e-01
-1.53946474e-01 1.13708481e-01 7.04856515e-02 -3.50151509e-01
-2.53472745e-01 1.16904356e-01 -1.40712309e+00 6.17808580e-01
1.04574311e+00 -3.23783278e-01 -5.59126198e-01 2.98810333e-01
3.18786174e-01 -5.21487474e-01 -1.21218598e+00 4.74512339e-01
3.57779503e-01 -1.05397928e+00 1.01632130e+00 -1.20347030e-01
4.38523650e-01 -6.20685279e-01 -2.30114609e-01 -1.27763784e+00
-4.49864268e-01 -4.83718246e-01 -1.58777446e-01 1.13758862e+00
-1.96496341e-02 -5.84305167e-01 7.05594897e-01 3.28530490e-01
-5.55442274e-01 -9.05971527e-01 -1.26767707e+00 -7.38131285e-01
1.80206805e-01 -6.70775920e-02 4.94446337e-01 9.23027933e-01
-3.99330884e-01 -3.16302255e-02 -1.11471251e-01 5.73440015e-01
1.10130405e+00 2.82924861e-01 7.80981958e-01 -1.53661013e+00
-1.21633574e-01 -3.92017722e-01 -4.64690268e-01 -1.19839633e+00
-4.68207486e-02 -9.55424070e-01 1.31556451e-01 -1.49674511e+00
1.21927664e-01 -8.88798058e-01 2.94436663e-01 -9.00077000e-02
1.51631176e-01 5.39671779e-01 -1.23870812e-01 6.12402081e-01
-1.23841695e-01 8.80177975e-01 1.55919385e+00 8.52612928e-02
-1.03935555e-01 -6.81250170e-02 -2.11203918e-01 8.58882248e-01
3.40890914e-01 -4.58689809e-01 -1.00880049e-01 -3.85319710e-01
1.63311958e-01 2.52381623e-01 5.67878127e-01 -8.71690392e-01
6.56958148e-02 -5.08745164e-02 2.67698228e-01 -7.23643839e-01
6.54598296e-01 -1.21123564e+00 1.07433069e+00 1.90379992e-01
2.08839342e-01 -3.61476809e-01 -4.90612909e-02 7.16411293e-01
-1.81007951e-01 -1.33808494e-01 1.18582189e+00 5.89382509e-03
2.79355407e-01 3.50735456e-01 2.67106225e-03 -1.60427868e-01
1.15770733e+00 -3.08147639e-01 3.73572469e-01 3.52942124e-02
-8.94765079e-01 -9.17152092e-02 1.37816393e+00 5.03859334e-02
8.21433187e-01 -1.57916105e+00 -7.63537705e-01 6.33700907e-01
1.96051206e-02 7.98718393e-01 2.19190419e-01 8.33184063e-01
-7.22600818e-01 -1.38380200e-01 -7.29702413e-02 -1.11643016e+00
-1.02289438e+00 4.51506734e-01 3.84077221e-01 2.26238742e-01
-1.20125449e+00 5.93561709e-01 5.53143978e-01 -4.95841265e-01
1.55920193e-01 -3.54023010e-01 2.57387847e-01 -1.72859818e-01
1.15033343e-01 2.64948666e-01 2.39899650e-01 -7.38054693e-01
-2.62251914e-01 1.02502847e+00 2.16347635e-01 -1.50278643e-01
1.45818818e+00 -2.51234844e-02 -3.74929667e-01 5.21193504e-01
1.06681049e+00 4.67101723e-01 -1.33486772e+00 -2.15740904e-01
-3.19406629e-01 -7.10769057e-01 5.55793568e-02 -8.88528898e-02
-1.00587773e+00 4.86640841e-01 5.59466965e-02 4.08409446e-01
7.92632103e-01 4.71242100e-01 6.92008257e-01 -2.08696142e-01
7.02683091e-01 -4.68806505e-01 4.41043973e-02 1.49230883e-02
1.45853317e+00 -6.04164004e-01 2.86729448e-03 -9.16360855e-01
-1.60276741e-01 8.98236752e-01 6.69726506e-02 -6.32494092e-01
8.55544865e-01 5.93057871e-02 -3.25382859e-01 -5.36199570e-01
-3.42395902e-01 4.07280207e-01 4.79295045e-01 4.38821107e-01
-4.94409502e-02 -3.08934133e-02 -2.34429706e-02 2.62333483e-01
-1.07854269e-01 -2.89454550e-01 2.26426050e-01 9.09970939e-01
-3.74301374e-01 -9.85198975e-01 -7.18892336e-01 1.53945759e-01
-8.18713605e-02 3.28779936e-01 -3.02007198e-01 4.18552548e-01
6.79225177e-02 4.61982042e-01 4.05928731e-01 3.86699103e-03
7.24834442e-01 -8.27805400e-02 7.68561423e-01 -7.85336852e-01
-1.14859551e-01 4.50700969e-01 -2.66734928e-01 -5.21369576e-01
-2.88118780e-01 -8.54392827e-01 -1.14345527e+00 -2.48943776e-01
-1.86460659e-01 1.77987307e-01 5.32917917e-01 5.74714959e-01
4.67707127e-01 1.88545004e-01 9.87212658e-01 -1.27978873e+00
-2.55992562e-01 -5.18826842e-01 -8.46315265e-01 4.98099029e-01
1.98720381e-01 -9.72780287e-01 -6.05066180e-01 2.32335061e-01] | [8.705268859863281, -3.3548901081085205] |
26f07b36-b7d3-425c-9ba7-e5e5a212066b | influence-of-lossy-speech-codecs-on-hearing | 2306.02344 | null | https://arxiv.org/abs/2306.02344v1 | https://arxiv.org/pdf/2306.02344v1.pdf | Influence of Lossy Speech Codecs on Hearing-aid, Binaural Sound Source Localisation using DNNs | Hearing aids are typically equipped with multiple microphones to exploit spatial information for source localisation and speech enhancement. Especially for hearing aids, a good source localisation is important: it not only guides source separation methods but can also be used to enhance spatial cues, increasing user-awareness of important events in their surroundings. We use a state-of-the-art deep neural network (DNN) to perform binaural direction-of-arrival (DoA) estimation, where the DNN uses information from all microphones at both ears. However, hearing aids have limited bandwidth to exchange this data. Bluetooth low-energy (BLE) is emerging as an attractive option to facilitate such data exchange, with the LC3plus codec offering several bitrate and latency trade-off possibilities. In this paper, we investigate the effect of such lossy codecs on localisation accuracy. Specifically, we consider two conditions: processing at one ear vs processing at a central point, which influences the number of channels that need to be encoded. Performance is benchmarked against a baseline that allows full audio-exchange - yielding valuable insights into the usage of DNNs under lossy encoding. We also extend the Pyroomacoustics library to include hearing-device and head-related transfer functions (HD-HRTFs) to suitably train the networks. This can also benefit other researchers in the field. | ['Alexander Bohlender. Nilesh Madhu', 'Jasper Maes', 'Stijn Kindt', 'Siyuan Song'] | 2023-06-04 | null | null | null | null | ['speech-enhancement'] | ['speech'] | [ 1.01407029e-01 -4.19067770e-01 3.04550767e-01 1.73894912e-02
-1.16276586e+00 -3.55696261e-01 1.31999344e-01 2.38848343e-01
-5.03248274e-01 5.31290054e-01 6.25817060e-01 -2.47206926e-01
-2.22011030e-01 -7.73464501e-01 -4.55336988e-01 -1.03413248e+00
-3.41748297e-01 -1.22596987e-01 3.15345615e-01 -1.85547128e-01
-1.09967910e-01 5.56583107e-01 -1.94438970e+00 1.18648104e-01
5.90544760e-01 1.24927866e+00 4.30124968e-01 8.60903382e-01
1.65705875e-01 3.17954838e-01 -1.02581608e+00 -1.01853088e-01
1.09166615e-01 -2.12452084e-01 -7.83936027e-03 -7.57560372e-01
1.02610953e-01 -4.50793356e-01 -2.23091677e-01 8.26628387e-01
1.69002604e+00 1.33022830e-01 4.90361273e-01 -8.74335527e-01
2.60654032e-01 3.05500090e-01 -1.83411360e-01 3.57009023e-01
4.40609634e-01 9.26443934e-02 5.67419171e-01 -6.51828051e-01
-4.09564041e-02 9.37911987e-01 8.45866442e-01 3.64100009e-01
-9.32146013e-01 -9.79630053e-01 -7.15574324e-01 4.77942050e-01
-1.40175951e+00 -1.11485243e+00 9.51070011e-01 -1.00549310e-01
1.00753367e+00 2.73156196e-01 6.30494714e-01 1.10016835e+00
-7.71222860e-02 3.96252722e-01 1.07190955e+00 -7.08398640e-01
4.71416324e-01 8.60029180e-03 -3.73589605e-01 -1.24950849e-01
-3.90292108e-02 5.21251738e-01 -1.17035079e+00 -8.99401456e-02
5.83368480e-01 -4.88584131e-01 -6.79372966e-01 7.63068274e-02
-7.46013224e-01 4.39655632e-01 4.02672350e-01 5.60508132e-01
-5.62230527e-01 2.03645095e-01 2.01529682e-01 3.60178888e-01
4.00044948e-01 3.16415161e-01 -2.78826714e-01 -5.91733396e-01
-1.08677745e+00 -7.81380385e-02 9.08998013e-01 4.06743109e-01
3.50398868e-01 2.67086983e-01 -6.96198717e-02 1.31839335e+00
5.35506189e-01 5.98941684e-01 3.68341625e-01 -8.82163942e-01
3.53336781e-01 -4.06439483e-01 -3.43790762e-02 -8.75898719e-01
-4.05806154e-01 -7.28915453e-01 -9.19471025e-01 3.93725067e-01
3.45756084e-01 -4.83674467e-01 -7.04665124e-01 1.74897873e+00
2.93537498e-01 3.15548867e-01 -3.98972668e-02 9.08008397e-01
5.32165289e-01 5.90156853e-01 -1.70219779e-01 -1.92133635e-01
1.40388966e+00 -4.34383541e-01 -8.81004810e-01 -2.11222455e-01
1.82801753e-01 -9.89299655e-01 7.15398192e-01 7.57968366e-01
-1.30060649e+00 -5.53857625e-01 -9.69457388e-01 1.32053450e-01
-3.89480799e-01 -2.17873976e-01 3.22166950e-01 1.21045518e+00
-1.49542272e+00 3.44805002e-01 -8.10010314e-01 -3.63983512e-02
1.50780492e-02 5.17911971e-01 -1.22543722e-01 9.34650227e-02
-1.41138160e+00 7.00477123e-01 -1.52537614e-01 1.16642132e-01
-7.52065539e-01 -8.06708038e-01 -6.47081017e-01 2.13994056e-01
-2.08602712e-01 -5.10406852e-01 1.42309976e+00 -5.87375581e-01
-1.75510085e+00 3.49412978e-01 -3.83324325e-01 -5.03877342e-01
2.01716766e-01 -1.46507949e-01 -8.72405767e-01 1.47533223e-01
-1.71426266e-01 5.33921838e-01 9.38984156e-01 -1.02127290e+00
-6.45642400e-01 -2.43862092e-01 -2.80327201e-01 1.88813850e-01
-4.89554524e-01 2.13776588e-01 -3.29943061e-01 -8.23320448e-01
1.78445578e-01 -3.53766531e-01 5.53607792e-02 4.66506407e-02
-3.41320515e-01 9.08051804e-02 5.64143419e-01 -1.04429758e+00
1.24910808e+00 -2.39875245e+00 -3.28021705e-01 3.35418671e-01
-4.03365195e-02 4.82768923e-01 2.95386836e-02 3.95312428e-01
-6.38400670e-03 -2.30180338e-01 -8.41230750e-02 -5.11551797e-01
-2.48849001e-02 -2.30780736e-01 9.36326534e-02 3.12846243e-01
-2.33007878e-01 1.97960749e-01 -4.27809387e-01 -1.86128855e-01
3.19353670e-01 1.17223108e+00 -6.68218672e-01 2.85202235e-01
4.88027006e-01 4.62850571e-01 1.40589133e-01 6.90141499e-01
9.06535745e-01 5.31590939e-01 -3.01433086e-01 -2.59913653e-01
-2.09482759e-01 7.76195705e-01 -1.39957070e+00 1.51271403e+00
-1.09384620e+00 9.20265019e-01 8.24709475e-01 -5.67499518e-01
8.58575284e-01 5.99692941e-01 1.78779870e-01 -9.77164507e-01
4.45924699e-02 6.46761000e-01 1.20532423e-01 -3.43002319e-01
2.98454970e-01 -1.90625936e-01 3.49621594e-01 1.91064343e-01
1.11465991e-01 -6.73827678e-02 -2.76018143e-01 -3.08422834e-01
1.13489974e+00 -6.04930282e-01 1.81994900e-01 -5.92194274e-02
3.99926633e-01 -9.59145308e-01 2.03597039e-01 5.83769441e-01
-2.33071908e-01 5.41931808e-01 -1.24480680e-01 3.90768915e-01
-5.36569953e-01 -1.16539145e+00 -2.71337509e-01 1.16245103e+00
-1.13953188e-01 -1.24784663e-01 -8.15920293e-01 1.99094623e-01
-2.03960538e-01 8.13625753e-01 1.30457163e-01 -9.19010937e-02
-3.11915487e-01 -2.25121632e-01 8.15949798e-01 4.55319405e-01
6.50283039e-01 -9.51315463e-01 -5.93900263e-01 4.55512822e-01
-3.29854578e-01 -6.77969813e-01 -2.49607801e-01 6.22196555e-01
-5.48462689e-01 -4.85228181e-01 -1.03671181e+00 -6.60919964e-01
6.14028163e-02 2.36424237e-01 8.23968112e-01 -3.50578487e-01
7.19772056e-02 6.58755064e-01 -2.17351615e-01 -7.89746404e-01
-3.43733042e-01 -1.10955328e-01 3.72207880e-01 -1.11258693e-01
3.02886635e-01 -1.26491201e+00 -9.13789868e-01 4.18201327e-01
-8.49264979e-01 -5.16265988e-01 5.25544107e-01 3.29705536e-01
2.19489276e-01 5.26941180e-01 7.04682171e-01 2.08970428e-01
9.63988483e-01 -3.21049958e-01 -3.48472953e-01 -2.71018803e-01
-1.28358632e-01 -3.39923024e-01 4.21755284e-01 -3.09265912e-01
-1.21017790e+00 -3.03489894e-01 -1.00929952e+00 6.28058240e-02
-4.35491383e-01 3.08170021e-01 -7.09037066e-01 -2.40857437e-01
6.62668765e-01 1.65053800e-01 -1.61902964e-01 -7.45854795e-01
-1.48427524e-02 1.40111315e+00 6.32417858e-01 -2.02388301e-01
3.78717065e-01 3.35337162e-01 -1.25629723e-01 -1.27291441e+00
1.16923928e-01 -8.14312398e-01 -9.25450101e-02 -1.98819712e-01
4.79075909e-01 -9.45003152e-01 -7.51823485e-01 8.16650629e-01
-1.29611027e+00 -2.86202341e-01 -2.41343677e-02 7.45099485e-01
-4.13600117e-01 -2.21479349e-02 -5.57193875e-01 -1.24933767e+00
-3.31017852e-01 -1.05273712e+00 1.00641608e+00 2.65887737e-01
-2.03975290e-01 -8.20592761e-01 7.43677020e-02 1.67955220e-01
9.93859828e-01 -3.07554603e-01 5.03281713e-01 -4.07488585e-01
-1.58841327e-01 -6.53516948e-02 1.66114971e-01 4.77878541e-01
1.99865684e-01 -7.00768888e-01 -1.70064163e+00 -1.90770775e-01
3.60057324e-01 5.03593497e-02 7.15438187e-01 9.76243079e-01
9.13161218e-01 -3.85193288e-01 -3.46006989e-01 5.60245633e-01
1.10964024e+00 3.48902076e-01 1.04591978e+00 2.27105662e-01
1.68198183e-01 6.25910878e-01 1.66748703e-01 4.01796132e-01
-2.13441849e-02 9.09941971e-01 4.67127740e-01 -3.00310045e-01
-5.44624269e-01 -5.39709665e-02 4.86844510e-01 8.35525870e-01
2.10815202e-02 -4.65953976e-01 -7.45772243e-01 6.44272387e-01
-8.39143515e-01 -6.68825090e-01 -5.67482077e-02 2.52479911e+00
1.11239123e+00 1.49379838e-02 2.16133252e-01 7.59778798e-01
7.21106291e-01 8.17136392e-02 -2.18009695e-01 -4.04927641e-01
-7.21280053e-02 7.71171927e-01 5.35516620e-01 7.32672691e-01
-7.07337141e-01 2.26594344e-01 5.62115002e+00 1.22026098e+00
-1.38312924e+00 4.25195992e-01 3.19857657e-01 -2.47822821e-01
-2.64931023e-01 -6.67756855e-01 -6.30167603e-01 6.31024718e-01
1.37640095e+00 4.50426877e-01 5.13701022e-01 6.65910900e-01
5.52597404e-01 -4.45409864e-01 -9.13496852e-01 1.19235957e+00
-2.49065489e-01 -8.11398864e-01 -6.70096099e-01 2.05580741e-01
2.79984083e-02 6.25549033e-02 2.57903516e-01 -8.97755548e-02
-2.02031061e-01 -8.74798536e-01 6.42724931e-01 3.52523416e-01
9.72257972e-01 -8.68720472e-01 7.70073771e-01 2.35671729e-01
-1.34593868e+00 -9.18045491e-02 -7.00849220e-02 -3.65912206e-02
4.18091863e-01 1.11998308e+00 -1.07976282e+00 2.32857823e-01
1.01992619e+00 -7.22502470e-02 -1.38281971e-01 1.66501606e+00
-3.70382398e-01 9.31320965e-01 -7.78634489e-01 7.40437023e-03
-2.86650628e-01 3.63182694e-01 8.91267776e-01 1.35251772e+00
8.52384150e-01 -3.18112150e-02 -7.19465435e-01 4.75077122e-01
3.46095562e-02 -1.50399029e-01 -6.22144997e-01 5.87698102e-01
8.63655567e-01 7.87162423e-01 -3.86891901e-01 3.00381720e-01
-8.36966932e-02 9.63552117e-01 -5.16915798e-01 5.70789993e-01
-5.15482187e-01 -7.85441101e-01 9.54535604e-01 3.72128904e-01
3.58051330e-01 -3.40003788e-01 -7.76521266e-02 -2.93385565e-01
1.09045766e-02 -8.24932337e-01 -6.24418110e-02 -1.04725134e+00
-9.69479442e-01 4.22506094e-01 -2.72875428e-01 -1.21595836e+00
-3.03757548e-01 -6.70802176e-01 -5.39598107e-01 1.17697120e+00
-1.63491035e+00 -6.13045752e-01 1.55956624e-02 7.28204489e-01
3.80499631e-01 1.28524736e-01 8.51361692e-01 1.00577307e+00
4.48960587e-02 8.88245702e-01 2.67595738e-01 2.62809563e-02
7.96787620e-01 -9.97396111e-01 9.89450514e-02 7.57120311e-01
-6.32303506e-02 5.96204937e-01 8.80754948e-01 -3.54893178e-01
-1.02322590e+00 -7.80990601e-01 9.40343618e-01 -5.85861541e-02
2.30800077e-01 -5.15352786e-01 -6.53605402e-01 -2.87381429e-02
3.99834126e-01 -2.69372463e-01 9.74834621e-01 -8.63045733e-03
-2.87468340e-02 -5.60781658e-01 -1.16588676e+00 3.89282942e-01
7.97179043e-01 -8.57759774e-01 -1.55156866e-01 -2.00111806e-01
6.57902896e-01 -3.54455680e-01 -4.32563603e-01 1.80118054e-01
6.48015678e-01 -1.36418664e+00 1.24694514e+00 4.24812108e-01
-6.98783249e-02 -1.82269230e-01 -3.63548994e-01 -1.66224241e+00
1.05719984e-01 -8.86931956e-01 -1.64491326e-01 1.64619720e+00
3.23600322e-01 -1.00775671e+00 5.04133821e-01 2.01062351e-01
-2.05072775e-01 -3.22886944e-01 -1.40715623e+00 -9.14578438e-01
-3.15362841e-01 -1.13862514e+00 5.53011775e-01 4.20382589e-01
-7.61144236e-02 7.58622810e-02 -1.97583675e-01 4.94006038e-01
4.63041455e-01 -8.41920912e-01 3.88378859e-01 -1.01572943e+00
-4.93456721e-01 -3.90084803e-01 -3.56503278e-01 -1.39295697e+00
-3.54658812e-01 -3.47714365e-01 3.60557944e-01 -1.35740435e+00
-7.55583405e-01 -5.52089989e-01 -4.07158375e-01 1.49613097e-01
2.07671538e-01 2.19333351e-01 -2.85523199e-03 -2.39673316e-01
-7.03983456e-02 4.60977674e-01 6.19946718e-01 -9.33363885e-02
-4.97515231e-01 7.33684480e-01 -5.15091658e-01 7.65011251e-01
7.12271214e-01 -5.07155657e-01 -2.31843337e-01 -4.33677346e-01
9.34699476e-02 3.50588024e-01 5.77711582e-01 -1.68713582e+00
6.73287511e-01 5.26745677e-01 1.18375011e-01 -5.22121489e-01
8.08697283e-01 -1.17511070e+00 3.03010613e-01 3.20800930e-01
-1.31484210e-01 -4.53975499e-01 4.60994303e-01 5.00719249e-01
-4.35185343e-01 -1.84505448e-01 6.47358477e-01 1.76525474e-01
-3.92962158e-01 -2.33490974e-01 -7.47646868e-01 -3.17743421e-01
3.69346738e-01 -3.50955009e-01 9.63896662e-02 -1.01751316e+00
-4.29632992e-01 -4.36351031e-01 -2.07638498e-02 -8.22610259e-02
5.96239388e-01 -1.29979479e+00 -3.90413791e-01 3.69264543e-01
-3.69526535e-01 -8.05811062e-02 2.91420877e-01 8.25546682e-01
-2.21197769e-01 5.89301705e-01 -3.90937328e-02 -7.12569833e-01
-1.28667307e+00 -6.14173524e-02 6.53511107e-01 9.26184207e-02
-4.87204231e-02 1.26111305e+00 -3.82344723e-02 -1.10340342e-01
7.48240829e-01 -2.58036107e-01 -2.01841787e-01 2.26629660e-01
7.59732604e-01 8.45693767e-01 6.62899375e-01 -3.53638679e-01
-4.34236526e-01 6.23939872e-01 6.36945546e-01 -7.75047600e-01
1.19455838e+00 -4.32488978e-01 -6.41388744e-02 2.29971394e-01
1.38687372e+00 6.66430116e-01 -9.68897521e-01 -1.31545709e-02
-3.32457989e-01 -4.68463808e-01 5.49569905e-01 -1.07778978e+00
-9.39070761e-01 1.36528087e+00 1.28744912e+00 4.15529877e-01
1.86798012e+00 -9.64032412e-02 9.41334665e-01 3.33255948e-03
6.00786626e-01 -9.67883885e-01 9.25693382e-03 2.34135583e-01
8.59760702e-01 -7.79538691e-01 -6.31005883e-01 -1.90145731e-01
-1.44647688e-01 9.61883962e-01 4.84258235e-02 6.57519937e-01
9.33125913e-01 6.65783823e-01 2.26624995e-01 8.69043395e-02
3.98262357e-03 -3.39621007e-01 1.53086603e-01 1.12694776e+00
5.18492818e-01 9.28553659e-03 7.73573294e-02 7.51767755e-01
-5.08330941e-01 -7.47874454e-02 5.86301945e-02 7.09990382e-01
-5.50372422e-01 -1.06866384e+00 -9.53729868e-01 3.06379020e-01
-6.26656592e-01 -5.45542002e-01 -1.03611246e-01 2.55280524e-01
3.09448212e-01 1.55360842e+00 4.58282642e-02 -3.89312923e-01
5.78071415e-01 6.05440401e-02 1.36220977e-01 -2.16153011e-01
-5.72240233e-01 5.50334454e-01 2.03474715e-01 -4.59105939e-01
-3.83025289e-01 -5.64507127e-01 -1.00045919e+00 -4.53229904e-01
-5.60279131e-01 8.42604637e-02 1.08271587e+00 5.52382410e-01
4.64570522e-01 8.69962454e-01 5.43310940e-01 -1.22546339e+00
-1.79189473e-01 -1.00860023e+00 -9.30895567e-01 -3.03999633e-01
8.23173404e-01 -5.71948588e-01 -6.67877972e-01 -2.26304710e-01] | [15.033156394958496, 5.841567039489746] |
fc31fc39-d45c-4be1-957a-36e79eb92c9f | n-stage-latent-dirichlet-allocation-a-novel | 2110.08591 | null | https://arxiv.org/abs/2110.08591v2 | https://arxiv.org/pdf/2110.08591v2.pdf | n-stage Latent Dirichlet Allocation: A Novel Approach for LDA | Nowadays, data analysis has become a problem as the amount of data is constantly increasing. In order to overcome this problem in textual data, many models and methods are used in natural language processing. The topic modeling field is one of these methods. Topic modeling allows determining the semantic structure of a text document. Latent Dirichlet Allocation (LDA) is the most common method among topic modeling methods. In this article, the proposed n-stage LDA method, which can enable the LDA method to be used more effectively, is explained in detail. The positive effect of the method has been demonstrated by the applied English and Turkish studies. Since the method focuses on reducing the word count in the dictionary, it can be used language-independently. You can access the open-source code of the method and the example: https://github.com/anil1055/n-stage_LDA | ['Tolgahan Cakaloglu', 'Banu Diri', 'Zekeriya Anil Guven'] | 2021-10-16 | null | null | null | null | ['twitter-sentiment-analysis'] | ['natural-language-processing'] | [-5.05089402e-01 -2.65220851e-01 -5.34047425e-01 -2.32607797e-01
-4.62173641e-01 -2.82980800e-01 6.70272470e-01 3.78180265e-01
-3.81711930e-01 3.40823710e-01 5.21342278e-01 -1.83964431e-01
8.52932706e-02 -9.54249740e-01 1.68655008e-01 -7.04388499e-01
4.51460361e-01 5.19008636e-01 1.53021008e-01 -8.51979777e-02
7.23199010e-01 7.55444095e-02 -1.66147220e+00 1.26181617e-01
9.95024681e-01 3.37567061e-01 7.17378914e-01 -7.85001293e-02
-9.53484654e-01 2.01363206e-01 -3.64444226e-01 -2.22584873e-01
-5.76942451e-02 -3.93366456e-01 -6.32096171e-01 1.23955332e-01
-5.38636625e-01 -1.45028770e-01 1.02836765e-01 1.10577106e+00
5.23588479e-01 2.44761229e-01 8.22884440e-01 -1.03850067e+00
-3.44893217e-01 6.78638995e-01 -6.87240601e-01 8.47514123e-02
2.86721051e-01 -5.11509895e-01 7.29229808e-01 -1.22277939e+00
4.52725738e-01 1.46399057e+00 8.62808675e-02 1.53157339e-01
-7.78876901e-01 -9.99423921e-01 -1.53954789e-01 4.27054673e-01
-1.44780600e+00 -9.87358857e-04 1.10430694e+00 -7.19676197e-01
5.17467499e-01 -1.17608413e-01 7.54103065e-01 8.06514204e-01
3.74043018e-01 8.76967132e-01 1.24337924e+00 -7.90907145e-01
2.55896032e-01 4.14950907e-01 5.56940317e-01 1.71289518e-01
4.67599839e-01 -6.41113043e-01 -3.63347024e-01 -2.97643840e-01
4.05788332e-01 3.32755506e-01 -2.05176929e-03 -1.86543003e-01
-7.71050155e-01 1.33129561e+00 -1.99102655e-01 8.66183281e-01
-4.64002222e-01 -4.22863990e-01 6.05125785e-01 -9.81600657e-02
1.01014185e+00 1.65283144e-01 -1.45647511e-01 -5.37216365e-01
-1.01063550e+00 1.38230115e-01 7.69366264e-01 6.77931249e-01
6.91343546e-01 -2.71917015e-01 1.03473708e-01 1.09810877e+00
8.28242242e-01 4.07335013e-01 8.10514331e-01 -4.14230764e-01
1.31973401e-01 9.81409907e-01 -1.02366090e-01 -1.12398160e+00
-1.76402152e-01 4.01146822e-02 -5.66921771e-01 -2.30343267e-01
1.98565125e-01 -1.47812426e-01 -7.29650617e-01 1.18150616e+00
6.26146257e-01 -2.08784863e-01 -9.95171443e-02 5.74409544e-01
9.99956429e-01 1.16872966e+00 4.16144401e-01 -5.30652642e-01
1.78238225e+00 -5.29331446e-01 -1.30522025e+00 2.23619677e-02
6.52382672e-01 -1.27848613e+00 1.20065033e+00 5.66622734e-01
-6.63050354e-01 -2.92102486e-01 -7.17999756e-01 -2.66963989e-01
-7.49083102e-01 3.29594970e-01 5.74052453e-01 6.22968137e-01
-5.71990430e-01 -9.67874229e-02 -8.87514353e-01 -8.85660231e-01
3.39136310e-02 -1.59262810e-02 -2.47147512e-02 -1.50771230e-01
-1.51628923e+00 8.35892558e-01 6.75550401e-01 -1.53878137e-01
-4.23076957e-01 -3.84704560e-01 -6.90436304e-01 4.01330143e-02
3.88674766e-01 -5.36502153e-02 1.01963603e+00 -4.93272305e-01
-1.47194183e+00 6.88587964e-01 -4.99657303e-01 7.57344961e-02
4.89305705e-02 -4.23665881e-01 -2.41015330e-01 3.55568603e-02
1.98665023e-01 2.95636356e-01 6.57751918e-01 -1.04470432e+00
-5.88238835e-01 -4.85840172e-01 -3.64084572e-01 2.08314985e-01
-7.48548388e-01 4.52670634e-01 -5.57140708e-01 -7.72606134e-01
2.98036039e-01 -8.91773403e-01 -1.99374445e-02 -5.22230804e-01
-1.62385866e-01 -8.91362607e-01 9.78941441e-01 -9.91562247e-01
1.67633259e+00 -2.27383351e+00 -4.67418991e-02 7.01558590e-02
-7.96695501e-02 2.61465609e-01 5.01285970e-01 9.23137009e-01
1.12480588e-01 2.52193034e-01 -1.55335411e-01 -2.68855751e-01
9.18375887e-03 1.38379056e-02 -2.96320975e-01 3.37002426e-01
-4.12007451e-01 2.16015607e-01 -6.50738597e-01 -6.79422677e-01
3.25223088e-01 7.08505392e-01 -3.28010291e-01 2.64525991e-02
-2.05086935e-02 4.26842958e-01 -6.85957670e-01 2.60807157e-01
7.55574048e-01 3.75601985e-02 3.52313608e-01 1.67143457e-02
-5.80985010e-01 3.48715723e-01 -1.23440373e+00 1.56081283e+00
-3.30442101e-01 5.75854421e-01 -2.83153504e-01 -8.02818835e-01
1.22508931e+00 5.87916493e-01 6.56772852e-01 -3.74424160e-01
4.10494864e-01 1.75813884e-01 -4.80666198e-02 -7.49218583e-01
6.08005822e-01 -9.10903066e-02 -8.86744112e-02 6.04419172e-01
-1.54957056e-01 -2.06528306e-01 5.65454781e-01 2.00272650e-01
3.00668597e-01 -1.12071119e-01 6.79381907e-01 -5.57532191e-01
5.36161661e-01 2.34623894e-01 5.71754277e-01 1.78640574e-01
1.12672627e-01 3.94113474e-02 4.86025572e-01 -2.44266301e-01
-1.22145808e+00 -3.99491370e-01 -4.91576135e-01 8.96356642e-01
-6.47822842e-02 -6.59322619e-01 -6.90626502e-01 -1.60499737e-01
-1.73482955e-01 1.13587523e+00 -2.91676998e-01 8.73074755e-02
-2.31616229e-01 -7.23500490e-01 2.85776835e-02 8.61997604e-02
6.84480429e-01 -9.49953914e-01 -3.64817977e-01 1.46387219e-01
-5.49063087e-01 -6.02716982e-01 1.88665222e-02 -2.82527506e-01
-9.92933214e-01 -6.63891256e-01 -8.13249052e-01 -6.05475426e-01
6.29332304e-01 4.80132729e-01 5.43045223e-01 -1.24258148e-02
-1.64871830e-02 1.19974010e-01 -8.47966850e-01 -7.35545337e-01
-3.42731684e-01 4.07978594e-01 8.35505128e-02 -1.74312443e-01
1.19856811e+00 -3.97243589e-01 -4.09951985e-01 1.58132523e-01
-1.09189284e+00 1.34488076e-01 3.94476116e-01 4.50217694e-01
2.64297366e-01 3.95788580e-01 5.23480237e-01 -1.03364182e+00
9.40480292e-01 -7.85266817e-01 -7.11434424e-01 -9.82310176e-02
-9.25083578e-01 -1.70632824e-01 8.12503397e-02 -2.04532921e-01
-1.28491843e+00 -1.99307635e-01 -2.26011872e-01 -9.70546342e-03
-4.16270435e-01 1.06680620e+00 -1.89609826e-01 5.05033255e-01
1.45299956e-01 2.43225038e-01 7.63968155e-02 -8.80536675e-01
1.06831044e-01 1.21491408e+00 -5.85926414e-01 -3.04162860e-01
2.83806473e-01 4.58403766e-01 -3.45302671e-01 -1.08489835e+00
-6.21354759e-01 -1.02164733e+00 -7.24837244e-01 -3.73752624e-01
9.78910387e-01 -9.13893700e-01 -4.04495865e-01 4.90107566e-01
-1.14948368e+00 -6.88319746e-03 2.42473617e-01 9.96653855e-01
-1.94248129e-02 4.78441626e-01 -3.28595877e-01 -8.66703987e-01
-3.01471025e-01 -1.14458251e+00 5.87770820e-01 2.45609820e-01
-2.89473265e-01 -1.22231531e+00 2.47161701e-01 2.52298087e-01
2.72404313e-01 -2.28232950e-01 1.15353966e+00 -1.08443058e+00
2.55109947e-02 -3.42498153e-01 6.46663681e-02 1.92818999e-01
4.29229647e-01 2.30044305e-01 -6.53563201e-01 -2.68623263e-01
4.49359864e-01 2.04837486e-01 5.30232906e-01 5.18369019e-01
9.30354834e-01 -1.42730683e-01 -4.07347441e-01 -1.52578548e-01
1.47556317e+00 4.96215910e-01 8.14228714e-01 5.40782988e-01
4.78538543e-01 7.39580989e-01 9.71541166e-01 8.00359607e-01
4.61258620e-01 7.03978300e-01 -4.68858927e-02 3.69494595e-02
2.25314245e-01 -3.56902815e-02 2.37009093e-01 1.37281430e+00
2.73869801e-02 -1.99677721e-01 -1.46530068e+00 4.57574725e-01
-1.74231923e+00 -7.66334713e-01 -5.64363778e-01 2.10498834e+00
8.15326452e-01 -5.22005968e-02 6.11480623e-02 3.86166900e-01
8.10642302e-01 1.21741720e-01 5.80428131e-02 -4.89637911e-01
2.82822877e-01 -2.42789015e-01 1.61227480e-01 5.17789602e-01
-9.84802783e-01 1.03219903e+00 5.38207865e+00 1.13933527e+00
-1.23131680e+00 3.97861183e-01 4.00975794e-01 2.10463285e-01
-2.60893703e-01 1.94228709e-01 -1.14796245e+00 7.62316883e-01
8.53788018e-01 -5.95092118e-01 -1.65711790e-01 1.00798106e+00
8.15696061e-01 -6.90015912e-01 -3.91419053e-01 9.03910756e-01
1.52559966e-01 -8.21104586e-01 2.37085178e-01 3.58533978e-01
6.40951633e-01 -3.97299141e-01 2.19019353e-02 3.58389705e-01
-8.52341726e-02 -3.47518235e-01 1.81860939e-01 3.23335737e-01
4.76847291e-01 -8.11855912e-01 9.56674814e-01 4.66077775e-01
-9.57340300e-01 7.96310417e-03 -6.50983930e-01 -1.51545271e-01
3.38097632e-01 1.14946544e+00 -8.89049768e-01 4.67967153e-01
7.07511127e-01 8.29898000e-01 -3.55356604e-01 1.28235400e+00
-4.76444930e-01 1.00695348e+00 -2.85600692e-01 -2.99659938e-01
1.00206248e-01 -7.03056753e-01 7.04902887e-01 1.13971698e+00
6.03452563e-01 1.47468477e-01 2.11122096e-01 6.15560830e-01
3.67600083e-01 8.60112607e-01 -7.32845247e-01 -1.78060830e-01
6.43253148e-01 1.25731826e+00 -1.17150807e+00 -4.76909041e-01
-4.92666066e-01 4.16191697e-01 -2.99305707e-01 1.27282247e-01
-6.41033769e-01 -3.68903399e-01 9.61458609e-02 2.31456399e-01
1.80102661e-01 -5.17822921e-01 -3.79264504e-01 -1.08832788e+00
-2.79629230e-01 -7.94628918e-01 3.83931488e-01 -5.86525559e-01
-1.03081119e+00 3.77113432e-01 5.75707972e-01 -1.24661291e+00
-1.16610043e-01 -4.18906868e-01 -3.91777009e-01 8.87297213e-01
-1.12021804e+00 -9.46324229e-01 -1.80521607e-01 4.52977836e-01
1.13111222e+00 -1.59154430e-01 8.42091680e-01 6.39289856e-01
-5.76753676e-01 -2.97539476e-02 5.33993959e-01 4.97045517e-02
8.76104891e-01 -9.01048779e-01 -4.32864018e-02 8.92039835e-01
-1.17031768e-01 8.50139201e-01 8.55603218e-01 -1.10549021e+00
-9.41238403e-01 -4.90698755e-01 1.25453568e+00 -2.39990652e-02
7.89006233e-01 -4.96239007e-01 -9.90024984e-01 4.63809043e-01
4.42768931e-01 -1.01561654e+00 1.15040398e+00 3.81214350e-01
8.31867680e-02 1.19198218e-01 -8.28188479e-01 5.74557662e-01
-5.80568686e-02 -1.97254032e-01 -7.25634992e-01 3.22831035e-01
3.76735866e-01 -8.40994865e-02 -8.75869095e-01 -4.87795882e-02
3.81715059e-01 -6.68213129e-01 4.48243886e-01 -7.76754990e-02
4.00299072e-01 -1.51634008e-01 5.62325679e-02 -1.19844306e+00
-2.33556047e-01 -1.29819736e-01 1.27185300e-01 1.63485634e+00
2.88344473e-01 -8.16478968e-01 4.91824836e-01 7.05064058e-01
1.21001340e-01 -4.90167737e-01 -7.03504682e-01 -5.07745981e-01
1.32286161e-01 -3.58225971e-01 3.66694778e-01 1.06425059e+00
3.84032071e-01 2.68082440e-01 -2.48002917e-01 -2.19913900e-01
4.82373387e-01 -3.29247303e-02 6.73792839e-01 -1.58477581e+00
3.76138985e-01 -2.72248238e-01 -1.59978151e-01 -8.13446999e-01
2.00914666e-02 -6.48747325e-01 -1.90453082e-01 -1.79801190e+00
4.40615505e-01 -4.16554302e-01 1.69906288e-01 4.17875499e-01
-1.93358362e-01 -1.57044172e-01 1.28598765e-01 8.24531734e-01
-1.62685350e-01 6.66400313e-01 1.09391546e+00 2.11844757e-01
-5.18313766e-01 4.53767739e-02 -5.05682647e-01 7.90961385e-01
1.12318587e+00 -8.25363517e-01 -4.42104667e-01 -1.67803705e-01
3.49915266e-01 -4.65935677e-01 -2.05621466e-01 -6.45604968e-01
3.20598334e-01 -9.39909816e-02 -1.09875046e-01 -8.59634817e-01
3.06270212e-01 -9.86712873e-01 2.47294873e-01 4.70742404e-01
-1.97304383e-01 7.77923390e-02 2.49363512e-01 2.31183708e-01
-5.17841876e-01 -6.61849976e-01 4.85659122e-01 -1.63388550e-01
-7.55030036e-01 5.66953830e-02 -8.73724222e-01 -3.56794327e-01
1.04789495e+00 -4.99390177e-02 -9.41320434e-02 -4.36105609e-01
-6.35315776e-01 4.32426780e-02 2.78844565e-01 3.94484192e-01
4.18098271e-01 -1.29306340e+00 -6.69844031e-01 -1.47803754e-01
3.22665945e-02 -1.66296050e-01 4.77760434e-01 9.87765610e-01
-5.11091769e-01 7.54042983e-01 -7.49557093e-02 -2.92362094e-01
-1.29851854e+00 5.88247001e-01 -3.16648364e-01 -2.38469660e-01
-5.33443272e-01 2.56105244e-01 1.71371669e-01 6.74964860e-03
-9.47285385e-04 9.91985351e-02 -9.76925135e-01 4.98126656e-01
4.86102462e-01 5.89836717e-01 -1.09889179e-01 -7.02990711e-01
-2.09497362e-01 5.26181400e-01 -2.40903646e-01 -3.82944822e-01
1.37906861e+00 -4.97078627e-01 -6.25396132e-01 1.03873754e+00
9.74357188e-01 1.53412476e-01 -4.02251869e-01 -1.33450449e-01
2.38585055e-01 -5.96349597e-01 3.93236697e-01 -3.50044459e-01
-6.20372772e-01 1.01034307e+00 5.19330859e-01 6.75038218e-01
8.90033960e-01 -3.14219147e-02 5.67992628e-01 6.52317926e-02
1.57979310e-01 -1.23957896e+00 -2.00322658e-01 5.65412283e-01
8.44634175e-01 -1.34320605e+00 2.40787312e-01 -6.19503975e-01
-6.34120405e-01 1.10740948e+00 2.60410935e-01 8.83565545e-02
1.20512223e+00 -2.28537316e-03 9.58221778e-02 -3.85557592e-01
-4.15014297e-01 -5.33967428e-02 1.31599352e-01 1.37899816e-01
1.04915178e+00 6.63871644e-03 -1.34317458e+00 7.30114758e-01
-1.29219562e-01 1.96437892e-02 4.68408376e-01 8.29877913e-01
-6.02904916e-01 -1.54826427e+00 -6.45176470e-01 3.30722660e-01
-6.96118176e-01 -2.05827847e-01 -8.32244530e-02 6.98963702e-01
-1.60712153e-01 1.13962233e+00 1.95457619e-02 -7.19611570e-02
-2.73996025e-01 3.23038697e-01 -2.62185782e-01 -8.18317473e-01
-1.32946923e-01 5.17382801e-01 -1.45866543e-01 -9.27713811e-02
-7.29732096e-01 -9.30905163e-01 -1.21629369e+00 -4.50612307e-01
-4.60282147e-01 7.10815787e-01 1.27622461e+00 9.70226526e-01
3.58899981e-01 1.51909500e-01 5.72898805e-01 -5.43311119e-01
7.18420744e-02 -1.33770871e+00 -7.09355533e-01 1.37058526e-01
-4.81380135e-01 -9.15471137e-01 -4.30130303e-01 1.87412828e-01] | [10.431411743164062, 7.095047950744629] |
07c5c766-9eeb-4a6a-916d-4afd6c43f5d2 | human-to-human-interaction-detection | 2307.00464 | null | https://arxiv.org/abs/2307.00464v1 | https://arxiv.org/pdf/2307.00464v1.pdf | Human-to-Human Interaction Detection | A comprehensive understanding of interested human-to-human interactions in video streams, such as queuing, handshaking, fighting and chasing, is of immense importance to the surveillance of public security in regions like campuses, squares and parks. Different from conventional human interaction recognition, which uses choreographed videos as inputs, neglects concurrent interactive groups, and performs detection and recognition in separate stages, we introduce a new task named human-to-human interaction detection (HID). HID devotes to detecting subjects, recognizing person-wise actions, and grouping people according to their interactive relations, in one model. First, based on the popular AVA dataset created for action detection, we establish a new HID benchmark, termed AVA-Interaction (AVA-I), by adding annotations on interactive relations in a frame-by-frame manner. AVA-I consists of 85,254 frames and 86,338 interactive groups, and each image includes up to 4 concurrent interactive groups. Second, we present a novel baseline approach SaMFormer for HID, containing a visual feature extractor, a split stage which leverages a Transformer-based model to decode action instances and interactive groups, and a merging stage which reconstructs the relationship between instances and groups. All SaMFormer components are jointly trained in an end-to-end manner. Extensive experiments on AVA-I validate the superiority of SaMFormer over representative methods. The dataset and code will be made public to encourage more follow-up studies. | ['Cong Bai', 'Jifeng Ning', 'Jiajun Meng', 'Kaining Ying', 'Zhenhua Wang'] | 2023-07-02 | null | null | null | null | ['action-detection', 'human-interaction-recognition'] | ['computer-vision', 'computer-vision'] | [ 2.38366604e-01 -3.03873479e-01 -5.99151962e-02 -2.90777385e-01
-4.59095210e-01 -5.56112111e-01 9.70351160e-01 -1.26370460e-01
-4.00163919e-01 9.23560858e-02 6.66820168e-01 -1.55223399e-01
5.19226491e-02 -4.54961121e-01 -4.56381172e-01 -5.54551661e-01
-5.35961449e-01 2.88092613e-01 4.22612429e-01 -1.39352769e-01
-1.54072002e-01 4.48820442e-01 -1.53964198e+00 7.85813093e-01
1.87906563e-01 1.20452678e+00 -3.58247697e-01 1.04620504e+00
4.59867299e-01 1.34203660e+00 -7.99780786e-01 -4.74663466e-01
3.72466505e-01 -6.31483674e-01 -7.98337281e-01 4.62700248e-01
5.12000024e-01 -7.50982523e-01 -8.15453589e-01 5.16043007e-01
4.37788785e-01 4.37880486e-01 5.56910515e-01 -1.67952621e+00
4.98100594e-02 2.80113280e-01 -6.76214695e-01 4.31205988e-01
1.16521287e+00 5.33375204e-01 9.59581971e-01 -9.00585592e-01
5.81548035e-01 1.74514294e+00 5.75769007e-01 3.29762936e-01
-1.00867152e+00 -7.77964830e-01 3.17614555e-01 2.82651424e-01
-1.38512921e+00 -5.67655623e-01 3.79488498e-01 -6.22466743e-01
1.03786993e+00 4.48990077e-01 9.32057381e-01 1.40872550e+00
-7.04782009e-02 1.31732166e+00 5.97209930e-01 -2.20637187e-01
-4.42228280e-02 -3.67688924e-01 2.26950884e-01 7.00095475e-01
-4.89214854e-03 -8.09752569e-02 -9.26073730e-01 -1.40721098e-01
8.38388026e-01 2.27034777e-01 -3.60543400e-01 -2.96004832e-01
-1.66652822e+00 4.78949755e-01 2.94809099e-02 -4.67105433e-02
-3.33726853e-01 -8.52793753e-02 7.18665004e-01 1.76273853e-01
1.87247515e-01 7.90321901e-02 2.93364711e-02 -5.26887298e-01
-8.20114017e-01 6.34229958e-01 8.38963270e-01 9.17106330e-01
2.92386204e-01 -4.97366607e-01 -7.19984949e-01 6.04985297e-01
1.59458980e-01 4.10970867e-01 7.47132599e-02 -1.03553855e+00
7.07333207e-01 6.99358881e-01 2.23110318e-01 -1.39101148e+00
-5.50171375e-01 2.04682305e-01 -9.21121180e-01 -8.94822478e-02
4.41195905e-01 -8.53419453e-02 -5.67991555e-01 1.45811915e+00
6.05436146e-01 4.55907404e-01 -1.81358904e-01 1.08556759e+00
1.04721928e+00 5.59629500e-01 6.84760213e-02 -1.56970114e-01
1.69704485e+00 -1.28790903e+00 -7.25562036e-01 -4.53169011e-02
5.94813287e-01 -4.67863590e-01 7.68216550e-01 3.83421183e-01
-9.95519459e-01 -8.05806816e-01 -5.81747472e-01 -1.47575485e-02
-1.13195069e-01 2.03524753e-02 5.49848557e-01 5.00131786e-01
-7.41171122e-01 2.02678859e-01 -9.81216431e-01 -6.08525932e-01
6.73500061e-01 1.71289191e-01 -6.13868713e-01 1.69915199e-01
-1.01267111e+00 3.50007623e-01 1.60010353e-01 1.64940104e-01
-1.04977441e+00 -4.53368902e-01 -1.16070139e+00 -6.60872385e-02
8.15950155e-01 -5.88811874e-01 1.31861663e+00 -6.60422981e-01
-1.22803330e+00 9.44140136e-01 -3.28752726e-01 -5.28330505e-01
8.57992589e-01 -5.27300179e-01 -5.80276072e-01 4.26526070e-01
2.11249456e-01 5.34013689e-01 8.05884600e-01 -1.00534379e+00
-9.61846411e-01 -2.00318411e-01 3.83469403e-01 2.25562274e-01
-1.05640396e-01 6.88773632e-01 -9.28341329e-01 -7.70446002e-01
-2.06035793e-01 -9.72752035e-01 3.25482385e-03 4.81819548e-02
-6.12523973e-01 -3.43624383e-01 9.45367575e-01 -6.51980579e-01
1.62674332e+00 -2.36490321e+00 1.47757560e-01 1.96339756e-01
6.59458220e-01 3.02566469e-01 -1.31564304e-01 4.88760114e-01
-1.45902306e-01 -2.92114049e-01 3.55601273e-02 -5.29794812e-01
2.39308298e-01 9.90256667e-03 -9.20394510e-02 6.19126379e-01
2.86842044e-02 9.79612052e-01 -9.66950476e-01 -6.13801420e-01
4.27569211e-01 3.30944419e-01 -6.51283383e-01 5.08536398e-01
2.72956997e-01 5.40452003e-01 -3.58838528e-01 9.24064636e-01
5.70268929e-01 -3.12371343e-01 1.70477703e-01 -2.50272632e-01
-4.87327836e-02 5.60503639e-02 -1.31886327e+00 1.38288736e+00
5.01318984e-02 8.37445557e-01 9.26818252e-02 -8.71524394e-01
3.98340672e-01 3.63435537e-01 8.69838774e-01 -5.63042164e-01
1.96317405e-01 -4.16833907e-01 -1.67316556e-01 -7.48322904e-01
3.83480638e-01 5.68482339e-01 -4.92230147e-01 4.21713501e-01
9.37073827e-02 6.09363019e-01 4.74817067e-01 6.40009224e-01
1.69993258e+00 4.27541472e-02 5.64013779e-01 2.48778358e-01
6.55914247e-01 -2.79462844e-01 5.04228055e-01 9.41558421e-01
-6.03740752e-01 6.05022371e-01 8.02549899e-01 -6.80551410e-01
-4.90504563e-01 -9.27852392e-01 3.18954468e-01 1.34673524e+00
3.00502956e-01 -8.62643361e-01 -7.10045278e-01 -1.11485481e+00
-1.18268896e-02 3.24107826e-01 -7.20588267e-01 3.04034464e-02
-7.91252017e-01 -4.31181878e-01 5.83008826e-01 5.14412344e-01
7.88397193e-01 -1.18004513e+00 -9.80737984e-01 -6.30863979e-02
-5.34920633e-01 -1.65021837e+00 -8.79591584e-01 -2.74716526e-01
8.08757395e-02 -1.56829417e+00 -5.99561930e-01 -5.13589442e-01
4.49159503e-01 7.04142809e-01 1.24928045e+00 6.61447225e-03
-3.94950807e-01 8.76729965e-01 -6.88394248e-01 -2.27953285e-01
-4.11964096e-02 -3.12563807e-01 1.24233566e-01 5.45207143e-01
4.32135105e-01 -2.74256259e-01 -6.67829812e-01 8.38192403e-01
-4.94598001e-01 2.59791642e-01 4.06621158e-01 5.14629483e-01
9.38541368e-02 5.01443446e-03 -1.67788565e-01 -8.43712151e-01
2.43468300e-01 -4.55558836e-01 -2.24745065e-01 1.21695332e-01
3.03620189e-01 -6.30578578e-01 3.38771492e-01 -4.42485362e-01
-9.44922686e-01 1.61344633e-01 4.39868234e-02 -6.35209620e-01
-5.55525839e-01 4.61737849e-02 -4.24794585e-01 1.78499669e-01
4.38672602e-01 1.25953943e-01 -1.19578660e-01 -1.41342342e-01
4.24810164e-02 5.82554817e-01 8.50521028e-01 -2.40780622e-01
8.79182696e-01 5.28665245e-01 -1.90507248e-01 -1.05620027e+00
-7.54221559e-01 -1.06780052e+00 -8.74400377e-01 -7.28770256e-01
1.09420240e+00 -1.14280260e+00 -1.35143781e+00 8.83043766e-01
-1.12995005e+00 -4.84955281e-01 -3.45452465e-02 4.52449888e-01
-4.18007702e-01 6.30257308e-01 -5.88646412e-01 -9.56250489e-01
6.18211254e-02 -9.06268418e-01 1.48248792e+00 8.47354624e-03
-5.78004599e-01 -5.84557056e-01 -1.49964616e-01 7.65436769e-01
-1.27622873e-01 4.76411700e-01 1.90374553e-01 -7.31044650e-01
-4.97002631e-01 -3.77975017e-01 -2.06702337e-01 8.92604291e-02
2.50658428e-04 -1.65993944e-01 -6.50987327e-01 -2.13540390e-01
-3.99954945e-01 -3.15303326e-01 7.73592412e-01 2.57782489e-01
1.13485801e+00 -4.10738558e-01 -4.92885232e-01 4.98690754e-01
5.62618375e-01 3.79249424e-01 8.05615485e-01 1.30037054e-01
9.83149529e-01 6.41675472e-01 7.90826380e-01 9.13906336e-01
5.56846917e-01 1.05044222e+00 2.81938553e-01 -2.45285735e-01
-3.91751379e-02 3.26480414e-03 7.19980597e-01 3.45708519e-01
-5.07215858e-01 -5.68756640e-01 -9.08638895e-01 2.75876343e-01
-2.18885469e+00 -1.56488717e+00 -1.98583260e-01 2.08653116e+00
2.88675487e-01 1.07034273e-01 8.98878038e-01 2.70597309e-01
6.18785143e-01 5.21000445e-01 -2.10253388e-01 2.61127323e-01
1.60927519e-01 -2.86656544e-02 2.03891009e-01 1.99043930e-01
-1.67077196e+00 7.94295132e-01 5.93296480e+00 6.99503303e-01
-5.95484436e-01 -9.98330191e-02 5.04314542e-01 -2.20364019e-01
6.26444340e-01 -2.94290483e-01 -9.31324959e-01 5.37082493e-01
6.72573209e-01 3.17315310e-01 3.76220852e-01 5.31954765e-01
6.18010879e-01 -2.60072708e-01 -1.28601086e+00 1.30816305e+00
3.30039114e-01 -1.00176251e+00 -2.26170775e-02 -7.88918510e-02
2.34053373e-01 -5.82296252e-01 -3.29291284e-01 5.25003612e-01
3.15683514e-01 -8.13255787e-01 8.94623578e-01 5.45656979e-01
5.18628478e-01 -6.02113724e-01 6.70910895e-01 2.83095121e-01
-1.68717420e+00 -2.37158045e-01 3.03431213e-01 -2.63990372e-01
5.16405404e-01 1.61272317e-01 -4.60402757e-01 5.35148025e-01
1.01052570e+00 1.10219371e+00 -6.44423187e-01 8.72476876e-01
-4.35252786e-02 6.11254215e-01 -4.67673168e-02 2.23602638e-01
1.63038358e-01 -5.92310652e-02 5.97070456e-01 1.50072277e+00
2.24633124e-02 5.80467105e-01 7.00022399e-01 2.99431205e-01
8.24861601e-02 -1.87446177e-01 -5.06814122e-01 -7.79765472e-02
2.86220580e-01 1.23048794e+00 -8.05127203e-01 -6.14943862e-01
-8.28553736e-01 1.25905311e+00 -9.31382403e-02 3.71821553e-01
-1.37927961e+00 -3.17664593e-01 8.25972021e-01 3.54977906e-01
4.16314274e-01 -2.90884227e-01 4.14422482e-01 -1.27352095e+00
2.63231061e-03 -1.28893387e+00 6.97908282e-01 -4.57137018e-01
-9.79057014e-01 3.67269307e-01 4.68758285e-01 -1.43324304e+00
-3.82260084e-01 -4.51005757e-01 -5.45521438e-01 3.04620534e-01
-7.39840329e-01 -1.32470036e+00 -9.09508228e-01 8.88480663e-01
8.80841136e-01 -2.05143243e-01 4.61760789e-01 5.41900516e-01
-1.04118264e+00 6.74963892e-01 -6.80690110e-01 7.14571595e-01
7.50852466e-01 -9.09905076e-01 5.47842801e-01 9.36329246e-01
2.03058571e-01 2.94840842e-01 5.65648973e-01 -4.52586323e-01
-1.37107003e+00 -1.22214282e+00 8.43364060e-01 -7.51920044e-01
5.59585750e-01 -8.23751271e-01 -6.31107271e-01 9.46099997e-01
1.15974955e-01 3.05689603e-01 6.97065353e-01 4.60903123e-02
-1.79505497e-01 6.77613467e-02 -6.78266525e-01 5.60039163e-01
1.73632026e+00 -4.53073829e-01 -3.74162495e-01 4.92839128e-01
3.74481112e-01 -6.01501167e-01 -5.69782913e-01 2.97414333e-01
8.11192811e-01 -1.34233987e+00 1.22381246e+00 -6.17336690e-01
3.02494317e-01 -3.91637444e-01 -4.62644249e-02 -6.44290209e-01
-4.68214840e-01 -8.81110430e-01 -5.58257759e-01 1.14146388e+00
-2.65233487e-01 -2.46801645e-01 6.60575688e-01 5.49172163e-01
-1.52945533e-01 -5.49617112e-01 -6.26601458e-01 -7.98495829e-01
-9.32807624e-01 -4.42027986e-01 3.68189424e-01 7.51838684e-01
-7.69216046e-02 3.64508033e-01 -8.27843308e-01 1.57122120e-01
5.14430881e-01 -1.87066689e-01 1.56702864e+00 -1.03726697e+00
-5.89866817e-01 -3.20319325e-01 -7.67030358e-01 -1.40916979e+00
1.76428519e-02 -3.69801253e-01 8.81953612e-02 -1.21099222e+00
4.36705589e-01 1.24979861e-01 -3.17523926e-02 6.13525510e-01
-1.64898768e-01 4.53288227e-01 3.57435793e-01 3.30464661e-01
-1.51391995e+00 3.97141308e-01 9.92987096e-01 -1.57372192e-01
-3.24950784e-01 1.34120136e-01 -3.12981695e-01 9.63838875e-01
1.20780520e-01 -6.58810586e-02 -1.97503969e-01 -7.41764233e-02
-2.11635217e-01 2.73581535e-01 7.67203450e-01 -1.39860511e+00
3.48619759e-01 -2.05932125e-01 4.01782751e-01 -6.77418113e-01
3.98558170e-01 -7.28769243e-01 1.51540518e-01 3.99849683e-01
-3.62577468e-01 2.40668878e-02 -1.93955362e-01 7.03289568e-01
-4.33885068e-01 3.97240937e-01 4.23806787e-01 -5.33784628e-02
-1.08343148e+00 4.56917405e-01 -7.23199844e-01 1.60552502e-01
1.51688647e+00 -3.30030590e-01 -1.33599535e-01 -7.30558217e-01
-6.77543819e-01 4.74603832e-01 9.53156874e-02 5.72124600e-01
4.35487181e-01 -1.41740417e+00 -7.18087971e-01 2.38746837e-01
2.53740966e-01 -1.21378452e-01 5.38491488e-01 1.14247632e+00
-4.76674020e-01 2.71302968e-01 -4.92773205e-02 -7.22841263e-01
-1.89672887e+00 5.26413500e-01 5.12346067e-02 -3.22764277e-01
-7.96786010e-01 7.96553552e-01 5.16430140e-01 -1.37553820e-02
6.18511736e-01 -1.96009189e-01 -4.44742233e-01 2.25197822e-01
9.67203021e-01 6.26648903e-01 -3.03818852e-01 -1.18514800e+00
-5.73082447e-01 1.42442852e-01 5.93880564e-03 1.42373443e-01
1.06592929e+00 -1.25958622e-01 1.41806811e-01 1.91550136e-01
1.09058118e+00 -7.36207590e-02 -1.34430099e+00 -1.95945710e-01
-2.64950514e-01 -8.90030980e-01 -4.45754617e-01 -3.54293466e-01
-9.79820549e-01 6.04205966e-01 3.35876256e-01 3.20309669e-01
1.17854679e+00 1.13133408e-01 8.97572815e-01 2.38042459e-01
3.37542385e-01 -8.18154693e-01 5.93564987e-01 5.80174029e-01
9.23078239e-01 -1.31956518e+00 -9.09875892e-03 -7.05984652e-01
-8.43464255e-01 8.51935267e-01 7.33447015e-01 1.15564249e-01
5.31719625e-01 2.66607761e-01 -1.64326340e-01 -2.59358823e-01
-6.97319448e-01 -2.72078633e-01 3.39841992e-01 4.91456777e-01
2.07924232e-01 -3.56671028e-03 1.22521341e-01 7.27449059e-01
4.04663123e-02 -1.35692075e-01 1.75253958e-01 1.07094526e+00
-9.82009768e-02 -6.81901753e-01 -5.80186188e-01 4.36786741e-01
-2.16996104e-01 2.43583575e-01 -4.07746226e-01 7.91772246e-01
3.87347281e-01 1.12547541e+00 3.59508604e-01 -6.91368222e-01
6.72675014e-01 -1.76759467e-01 1.74718380e-01 -4.51204568e-01
-7.02960968e-01 4.95125465e-02 3.04074228e-01 -1.25429857e+00
-6.38359666e-01 -9.99514222e-01 -9.93286133e-01 -4.59744930e-01
1.37062550e-01 -1.88295305e-01 -1.78942859e-01 9.63286161e-01
3.78555179e-01 5.80712259e-01 6.09715104e-01 -1.32205582e+00
1.96421761e-02 -7.37817049e-01 -4.74348873e-01 8.95810008e-01
3.86668950e-01 -8.80775332e-01 -2.40677178e-01 2.45512277e-01] | [8.233217239379883, 0.5577858686447144] |
fc7f3193-0952-4194-914c-8ab82aec0c64 | deep-color-mismatch-correction-in | null | null | https://ieeexplore.ieee.org/document/9506036 | https://v-sense.scss.tcd.ie/wp-content/uploads/2021/06/ICIP_2021_compressed.pdf | Deep Color Mismatch Correction In Stereoscopic 3D Images | Color mismatch in stereoscopic 3D (S3D) images can create visual discomfort and affect the performance of S3D image processing algorithms, e.g., for depth estimation. In this paper, we propose a new deep learning-based solution for the problem of color mismatch correction. The proposed solution consists of a multi-task convolutional neural network, where color correction is the primary task and correspondence estimation is the secondary task. For the training and evaluation of the proposed network, a new S3D image dataset with color mismatch was created. Based on this dataset, experiments were conducted showing the effectiveness of our solution. | ['Aljosa Smolic', 'Sebastian Knorr', 'Roman Dudek', 'Emin Zerman', 'Cagri Ozcinar', 'Simone Croci'] | 2021-06-01 | null | null | null | ieee-international-conference-on-image-10 | ['color-mismatch-correction'] | ['computer-vision'] | [ 3.08446735e-02 -2.84350663e-01 2.60153413e-01 -3.93424630e-01
-3.54396701e-01 -1.25168622e-01 1.63922459e-01 -1.44671127e-01
-3.94520909e-01 4.20883179e-01 1.89147413e-01 -3.45628560e-01
2.85183340e-01 -5.01510262e-01 -5.48637390e-01 -4.95246500e-01
4.29247737e-01 -2.46953979e-01 3.19953173e-01 6.08379953e-02
7.54202962e-01 7.23234773e-01 -1.61231899e+00 1.62173122e-01
8.60810697e-01 1.38327074e+00 2.99481839e-01 5.17530203e-01
-1.64544135e-01 5.81089437e-01 -5.13698101e-01 3.26698795e-02
5.98298788e-01 -1.78779811e-01 -5.69218278e-01 2.60138869e-01
6.33162022e-01 -7.97758698e-01 -2.79714108e-01 9.45428729e-01
1.05865586e+00 1.94745928e-01 2.98577845e-01 -1.21280050e+00
-3.47287416e-01 -3.98903698e-01 -7.40693688e-01 1.26281664e-01
2.98789531e-01 7.67841041e-02 2.26740062e-01 -1.05532932e+00
4.86623228e-01 1.22982275e+00 4.52754349e-01 5.82122922e-01
-7.61677265e-01 -5.97389817e-01 2.14213692e-02 3.44177365e-01
-1.20532131e+00 -3.10332954e-01 1.08539557e+00 -4.28418994e-01
8.67692769e-01 -3.37470360e-02 7.89143085e-01 7.88434267e-01
3.25239539e-01 6.86806560e-01 1.35937858e+00 -4.43768501e-01
1.97017714e-01 4.69009764e-03 -5.31720072e-02 5.78544259e-01
2.69679248e-01 1.97918102e-01 -4.09552485e-01 3.45921963e-01
1.02363420e+00 -7.10191131e-02 -2.02046156e-01 -5.25964916e-01
-7.28210866e-01 4.26226437e-01 5.86776137e-01 8.75745267e-02
2.18527671e-02 9.11003575e-02 2.18659744e-01 -9.94383730e-03
7.78945267e-01 1.23527959e-01 -2.33851388e-01 -1.31091848e-01
-5.52461982e-01 1.50819570e-01 1.73329860e-01 7.41919994e-01
6.33553743e-01 -2.50327647e-01 -6.85128570e-02 8.65926147e-01
2.42620841e-01 3.59072149e-01 4.15731579e-01 -9.73186493e-01
4.03630555e-01 7.03174233e-01 4.13240552e-01 -1.08181763e+00
-7.19203949e-01 -7.24945366e-02 -7.93392122e-01 8.15332055e-01
2.88474321e-01 -2.21749216e-01 -1.15360212e+00 1.54028165e+00
4.16759729e-01 -5.37203625e-02 -1.39795080e-01 1.48495030e+00
1.15723157e+00 4.99699354e-01 -3.01902950e-01 2.03033909e-02
6.73602462e-01 -8.81634116e-01 -7.52341390e-01 -3.00599009e-01
3.12286556e-01 -8.86028528e-01 9.94086623e-01 3.91388774e-01
-1.37525499e+00 -4.95869040e-01 -1.13877308e+00 -5.14176488e-01
-3.28038692e-01 1.78769693e-01 6.07499599e-01 3.91519070e-01
-1.21603274e+00 1.97393745e-01 -3.53334427e-01 -3.13870430e-01
3.33960295e-01 9.09057185e-02 -3.24445844e-01 -2.83331215e-01
-9.47554588e-01 1.13821876e+00 1.28386706e-01 3.55220556e-01
-6.29972339e-01 -2.38168046e-01 -7.49449611e-01 -1.43387303e-01
1.05870716e-01 -6.27867758e-01 1.16214848e+00 -8.74560833e-01
-1.48798633e+00 1.31257057e+00 -3.50925118e-01 1.18301526e-01
5.54320931e-01 -2.84388036e-01 -2.03732237e-01 1.05886996e-01
-1.78767014e-02 5.98454118e-01 8.44665527e-01 -1.52678084e+00
-7.25248814e-01 -5.20660639e-01 1.54985994e-01 5.09330332e-01
-4.12292928e-02 -1.51547760e-01 -8.28639567e-01 -4.29029822e-01
5.04643619e-01 -5.92775464e-01 4.24272381e-02 4.06045854e-01
-4.71951962e-01 -1.03927933e-01 6.79763734e-01 -6.01057768e-01
9.02256846e-01 -2.35722375e+00 -6.30836934e-03 1.39863372e-01
4.47815537e-01 3.41001332e-01 -1.45697489e-01 -5.18688560e-02
-1.78673774e-01 -6.09043837e-02 6.34060800e-03 -4.82788265e-01
-3.14233840e-01 -1.85938045e-01 1.91178083e-01 3.35722953e-01
2.60367215e-01 5.08817971e-01 -5.34697294e-01 -3.54235828e-01
5.68516493e-01 4.77219820e-01 -5.29235065e-01 5.52358270e-01
1.34838492e-01 5.80533981e-01 -1.19919434e-01 6.54249549e-01
1.23538184e+00 7.95177072e-02 -3.24261695e-01 -4.43238407e-01
-4.12654996e-01 1.00644961e-01 -1.02503538e+00 1.84684920e+00
-6.61808491e-01 8.49782526e-01 -8.30840021e-02 -6.33331895e-01
1.16872215e+00 -5.84687665e-02 5.59968650e-01 -1.28121281e+00
4.72066998e-01 3.94492388e-01 -1.48913816e-01 -9.21845675e-01
4.87544745e-01 8.32253918e-02 2.74221420e-01 4.16083664e-01
-3.55192542e-01 -3.90098572e-01 -2.78325826e-01 -1.89959079e-01
6.16021633e-01 8.34662914e-02 9.84985605e-02 1.33067846e-01
6.29892349e-01 -2.56057084e-01 4.77028817e-01 2.66449839e-01
-5.19432187e-01 7.33072877e-01 4.54559565e-01 -7.45405793e-01
-1.16947603e+00 -1.01887608e+00 3.62416916e-02 3.91377330e-01
9.16952133e-01 1.77050099e-01 -5.19757867e-01 -4.56134617e-01
2.16670912e-02 2.88795352e-01 -5.81095099e-01 -2.84351557e-01
-3.16519648e-01 -4.36938286e-01 -5.22936955e-02 4.46594626e-01
9.22192097e-01 -7.02876091e-01 -8.74217987e-01 -1.71685874e-01
-1.94313765e-01 -1.18745863e+00 -3.84086162e-01 6.77871238e-03
-7.91798472e-01 -1.32399082e+00 -8.61869216e-01 -1.00442982e+00
7.97033072e-01 8.06738317e-01 8.12105715e-01 -1.26808044e-02
-3.48968416e-01 1.27176821e-01 -1.69434726e-01 -5.21583796e-01
8.05062652e-02 -3.34458560e-01 -2.24794552e-01 -4.98347655e-02
2.74044216e-01 -4.77284431e-01 -1.06291699e+00 3.90686274e-01
-7.25561500e-01 3.95685405e-01 7.30524421e-01 6.68806434e-01
4.04297709e-01 -1.88148975e-01 1.77343205e-01 -4.57546353e-01
6.28322542e-01 7.08905756e-02 -7.47678399e-01 -4.31316122e-02
-4.73238736e-01 -1.37464434e-01 1.84722751e-01 -2.50375271e-01
-1.24293554e+00 -1.43456571e-02 -1.60129040e-01 -5.58882535e-01
-1.51950002e-01 2.24610701e-01 -3.15448612e-01 -4.83527243e-01
4.53861207e-01 9.83898044e-02 9.44369361e-02 -5.09480059e-01
-4.03749608e-02 9.42194819e-01 5.13515711e-01 4.03220952e-02
5.75485349e-01 4.79920626e-01 2.33125642e-01 -6.30031526e-01
-7.31458545e-01 -3.55063140e-01 -5.31844735e-01 -6.46509230e-01
9.38605070e-01 -1.10353160e+00 -8.92401755e-01 1.08551288e+00
-1.35726833e+00 -3.64259362e-01 2.94270933e-01 6.15973890e-01
-3.70206743e-01 1.56748623e-01 -2.29330346e-01 -7.16675758e-01
-2.18941629e-01 -1.07963765e+00 9.92419243e-01 3.44263971e-01
1.62264854e-01 -7.97339559e-01 9.14180875e-02 5.24171591e-01
4.38610315e-01 4.38943088e-01 1.14166331e+00 1.39804661e-01
-7.18820691e-01 -1.49360135e-01 -7.71792650e-01 3.66862118e-01
3.02870363e-01 -1.86139151e-01 -1.22249711e+00 -4.97751087e-02
9.33768451e-02 -3.74996960e-01 7.59722114e-01 6.91960156e-01
1.53720534e+00 2.63414741e-01 5.24136722e-02 1.02490282e+00
1.66102552e+00 4.84368593e-01 9.00284529e-01 5.82602739e-01
8.77496243e-01 6.69015765e-01 7.60413527e-01 5.30516565e-01
4.14317787e-01 7.84718513e-01 6.76998019e-01 -6.66489601e-01
-3.65351677e-01 -2.16178522e-02 1.35938786e-02 5.49400270e-01
1.95886381e-02 -9.63957608e-02 -8.33089709e-01 3.04079592e-01
-1.51929688e+00 -4.56332058e-01 -1.63687721e-01 2.17183542e+00
4.84847784e-01 1.26401514e-01 -2.03202382e-01 1.57309502e-01
7.51679599e-01 3.03653516e-02 -8.15303445e-01 -6.25181794e-01
-9.55892950e-02 1.59891501e-01 3.40451539e-01 5.37276745e-01
-1.09110188e+00 7.26070046e-01 5.93278265e+00 3.42292070e-01
-1.58393645e+00 -1.48575783e-01 5.84473729e-01 -2.34907359e-01
-2.04089209e-01 -4.50857222e-01 -4.19514835e-01 5.15175462e-01
-7.84552097e-03 5.06076310e-03 2.37158775e-01 5.22410929e-01
6.42383456e-01 -7.34367371e-01 -9.49499249e-01 1.69745409e+00
3.85662496e-01 -1.01953411e+00 4.37927209e-02 -2.78529495e-01
7.54975915e-01 -1.63535804e-01 3.20590436e-01 -3.05841386e-01
-3.68611872e-01 -7.77839601e-01 5.78411400e-01 2.95358658e-01
9.31712210e-01 -8.80861044e-01 8.11097264e-01 -4.78777662e-03
-7.42526770e-01 -9.98623073e-02 -2.34343067e-01 -7.66200647e-02
-6.07421249e-02 5.57978392e-01 -3.85141313e-01 2.05991685e-01
9.25687313e-01 8.29356074e-01 -6.42714977e-01 1.53967130e+00
-2.76296228e-01 -2.92080224e-01 3.65092345e-02 -4.17193398e-04
1.48386702e-01 -1.46247625e-01 3.73909175e-01 7.08323836e-01
4.89068449e-01 1.26702249e-01 -3.10035646e-01 7.85233259e-01
1.32518355e-02 -8.50118324e-02 -7.06790984e-01 4.11897272e-01
1.48411423e-01 1.02479649e+00 -4.53936756e-01 -2.11501494e-02
-4.94439304e-01 1.07296705e+00 2.41591275e-01 4.48921263e-01
-6.84736013e-01 -6.65023685e-01 8.28563809e-01 -6.65183738e-03
1.05039299e-01 -3.20730567e-01 -6.67760611e-01 -1.07850838e+00
2.29049712e-01 -4.43976939e-01 -2.21807808e-02 -1.29676390e+00
-1.06516659e+00 3.25119525e-01 -2.73615539e-01 -1.50330400e+00
1.18347049e-01 -8.13731790e-01 -7.97371149e-01 1.14887524e+00
-1.92012215e+00 -8.97173285e-01 -1.10828686e+00 8.99780035e-01
3.17602515e-01 -5.87629899e-03 5.52192688e-01 4.60095108e-01
-6.43945992e-01 4.84086663e-01 3.28084594e-03 -1.69883639e-01
1.01013827e+00 -1.01658511e+00 3.13984722e-01 9.87040877e-01
-7.85527110e-01 6.85534328e-02 3.81312490e-01 -5.30371070e-01
-1.24064541e+00 -9.92747188e-01 7.23161042e-01 6.76352531e-02
-2.44115144e-02 -3.38099688e-01 -5.36646128e-01 1.00327522e-01
-6.34820536e-02 -1.51209921e-01 4.77906376e-01 -3.36181998e-01
-6.01167157e-02 -2.86505103e-01 -1.27081048e+00 5.53770065e-01
1.04154968e+00 -5.72274327e-01 -2.11796716e-01 2.63229385e-02
5.52145720e-01 -8.66334975e-01 -5.17913818e-01 3.50039303e-01
6.47559166e-01 -1.38550925e+00 8.88869286e-01 5.79819307e-02
7.86464870e-01 -3.36412966e-01 -1.05242226e-02 -1.37147176e+00
-4.14710231e-02 -1.76275790e-01 1.24116376e-01 7.17724562e-01
1.17557421e-01 -4.01422054e-01 8.06619465e-01 6.09245718e-01
-4.02187407e-01 -5.92617571e-01 -8.88798356e-01 -5.27008772e-01
-2.05318525e-01 -3.40838313e-01 4.27820802e-01 6.72302961e-01
-2.57316411e-01 1.07708663e-01 -3.95389885e-01 1.42437160e-01
6.39903188e-01 3.24031204e-01 8.19741309e-01 -1.03746653e+00
1.67201996e-01 -3.90010625e-01 -5.55864573e-01 -9.81019258e-01
-1.30403966e-01 -4.80511308e-01 -6.69451896e-04 -1.77420592e+00
4.13966402e-02 -2.60607779e-01 -1.52643353e-01 1.85222015e-01
-2.49587938e-01 4.62724894e-01 1.87665094e-02 -1.20773263e-01
-2.09208056e-01 6.40944839e-01 1.68852961e+00 -1.64426878e-01
-5.57242692e-01 -8.99628252e-02 -4.48740959e-01 5.89426875e-01
9.02947962e-01 3.33339162e-02 -5.37186503e-01 -8.69810939e-01
1.76172242e-01 -7.82936718e-03 4.09402043e-01 -9.34357524e-01
1.43422678e-01 -3.47276330e-02 8.29403639e-01 -9.49347496e-01
5.92391729e-01 -9.25307631e-01 -3.92586350e-01 6.18787467e-01
-1.71803862e-01 -1.61363378e-01 4.13025737e-01 1.02234200e-01
-3.58148992e-01 1.93139434e-01 9.86436903e-01 6.61646761e-03
-1.12669194e+00 3.73108387e-01 -1.06378004e-01 -3.99597362e-02
1.08431876e+00 -4.62302625e-01 -4.49316591e-01 -3.23625624e-01
-3.90367568e-01 2.15808332e-01 6.02221489e-01 4.24556613e-01
1.41702104e+00 -1.49987507e+00 -3.30073744e-01 5.31750560e-01
4.07298148e-01 1.25728309e-01 4.12842542e-01 7.07285047e-01
-8.37434471e-01 2.99436212e-01 -8.43052447e-01 -5.65271020e-01
-1.23260069e+00 3.61250818e-01 5.80740511e-01 4.23943579e-01
-5.66893995e-01 9.05192077e-01 2.04154611e-01 -1.08308449e-01
5.76917171e-01 -5.02322674e-01 -3.58512163e-01 -2.74547070e-01
3.41326535e-01 3.94263893e-01 1.49554253e-01 -2.42542297e-01
-3.99955958e-01 9.80625033e-01 1.26264542e-01 -6.42495304e-02
1.36780119e+00 -5.87223291e-01 -1.14655882e-01 3.86211485e-01
1.35281456e+00 -3.10630620e-01 -1.36865330e+00 -1.77104622e-02
-4.93774742e-01 -1.10812771e+00 3.81146848e-01 -7.81249046e-01
-1.18910301e+00 1.13345730e+00 1.13561082e+00 -3.27479392e-01
1.51832747e+00 -4.52572823e-01 9.39847827e-01 1.23542115e-01
2.81962514e-01 -1.13772166e+00 2.88864970e-01 4.06898081e-01
7.35682964e-01 -1.65440810e+00 -2.16508448e-01 -3.60896885e-01
-4.21801388e-01 1.18212807e+00 1.10210395e+00 1.49558615e-02
5.36965668e-01 -5.57039827e-02 3.03925365e-01 -1.61964238e-01
-3.15709352e-01 -2.40726426e-01 3.73526067e-01 7.99071193e-01
2.82334179e-01 -3.43525320e-01 -2.71390885e-01 2.07128942e-01
1.37578398e-01 1.33015305e-01 6.80428743e-01 7.69360662e-01
-3.75608593e-01 -7.74860740e-01 -3.18956941e-01 1.76380545e-01
-4.12556194e-02 7.38672018e-02 -5.89126706e-01 6.31333709e-01
4.12764966e-01 1.05082345e+00 2.82509297e-01 -5.27062476e-01
4.98782575e-01 -2.58726418e-01 6.36802256e-01 -3.19741994e-01
-1.29488468e-01 -8.25725198e-02 -2.87117884e-02 -1.04520869e+00
-6.33129597e-01 -1.19101554e-01 -9.03412104e-01 -4.92235303e-01
9.02273282e-02 -4.78174746e-01 1.08203912e+00 6.90517008e-01
2.52862394e-01 4.42211390e-01 1.10199332e+00 -1.06394744e+00
1.38738379e-01 -7.29239821e-01 -7.14327931e-01 5.94203949e-01
6.00573778e-01 -7.44542003e-01 -3.93305779e-01 -3.00308347e-01] | [9.117269515991211, -2.4518380165100098] |
37224062-2457-4477-929f-b2763ecd4e27 | implicit-transfer-operator-learning-multiple | 2305.18046 | null | https://arxiv.org/abs/2305.18046v1 | https://arxiv.org/pdf/2305.18046v1.pdf | Implicit Transfer Operator Learning: Multiple Time-Resolution Surrogates for Molecular Dynamics | Computing properties of molecular systems rely on estimating expectations of the (unnormalized) Boltzmann distribution. Molecular dynamics (MD) is a broadly adopted technique to approximate such quantities. However, stable simulations rely on very small integration time-steps ($10^{-15}\,\mathrm{s}$), whereas convergence of some moments, e.g. binding free energy or rates, might rely on sampling processes on time-scales as long as $10^{-1}\, \mathrm{s}$, and these simulations must be repeated for every molecular system independently. Here, we present Implict Transfer Operator (ITO) Learning, a framework to learn surrogates of the simulation process with multiple time-resolutions. We implement ITO with denoising diffusion probabilistic models with a new SE(3) equivariant architecture and show the resulting models can generate self-consistent stochastic dynamics across multiple time-scales, even when the system is only partially observed. Finally, we present a coarse-grained CG-SE3-ITO model which can quantitatively model all-atom molecular dynamics using only coarse molecular representations. As such, ITO provides an important step towards multiple time- and space-resolution acceleration of MD. | ['Simon Olsson', 'Ole Winther', 'Mathias Schreiner'] | 2023-05-29 | null | null | null | null | ['operator-learning'] | ['miscellaneous'] | [ 2.03843698e-01 -5.19778490e-01 1.08431034e-01 -1.78076208e-01
-1.11951649e+00 -5.47761738e-01 6.61470890e-01 2.57267714e-01
-8.53107095e-01 1.19500446e+00 -2.56437838e-01 -4.08274084e-01
-1.15484774e-01 -8.30536366e-01 -9.23974991e-01 -1.46939623e+00
-2.96632499e-01 8.36249471e-01 2.38038749e-01 -2.26169050e-01
1.60679176e-01 6.04673207e-01 -1.24275875e+00 6.58962578e-02
9.43755507e-01 6.96864545e-01 1.66711267e-02 9.69019115e-01
7.53552373e-03 5.31360388e-01 -2.05552801e-01 -7.91753307e-02
-2.57524461e-01 -6.35552466e-01 -4.03297603e-01 -5.55381477e-01
1.75324380e-01 9.90094244e-02 -5.27571201e-01 1.36307347e+00
4.75667983e-01 6.47361517e-01 1.30881107e+00 -4.03917998e-01
-5.77214539e-01 4.02949125e-01 -3.96321982e-01 2.65406817e-01
-3.55130769e-02 7.20892966e-01 7.27403343e-01 -7.41228044e-01
9.29387510e-01 1.06713438e+00 6.10212743e-01 5.37005544e-01
-1.88575518e+00 -5.51441193e-01 -3.75356525e-02 -1.39810532e-01
-1.49930465e+00 -2.22518355e-01 5.45473337e-01 -4.79674876e-01
1.16836298e+00 9.91672426e-02 5.62187433e-01 1.31429994e+00
8.56715322e-01 -4.51348536e-02 1.47280467e+00 -2.27953076e-01
8.66195560e-01 -3.23107898e-01 3.79673153e-01 6.68592751e-01
4.13628966e-01 3.12532037e-01 -3.21475416e-01 -4.83670831e-01
7.98024058e-01 7.12455660e-02 -2.19593585e-01 -1.32700160e-01
-1.04530931e+00 8.88643265e-01 2.06041172e-01 7.07155094e-02
-6.43997490e-01 7.90043890e-01 3.78998965e-01 1.03753693e-01
4.37288880e-01 2.49833554e-01 -2.83755839e-01 -4.44224507e-01
-1.11430395e+00 5.90730608e-01 6.67457879e-01 5.52776575e-01
7.70341814e-01 2.51056105e-01 -6.81591406e-02 1.31518707e-01
2.84052163e-01 8.61833394e-01 1.00519225e-01 -1.21222866e+00
4.79372703e-02 -1.85645998e-01 4.98417139e-01 -2.71787554e-01
-3.15974444e-01 -2.27683887e-01 -1.29837728e+00 5.14359295e-01
6.60179257e-01 -1.18383698e-01 -1.02730465e+00 1.78585315e+00
3.40108514e-01 4.98660542e-02 -1.02519639e-01 7.70598352e-01
2.55173981e-01 9.63722706e-01 4.35876608e-01 -6.40135467e-01
1.09422922e+00 -3.57017606e-01 -5.90369463e-01 1.53422147e-01
4.63810027e-01 -5.00131667e-01 8.89361262e-01 4.78833169e-01
-1.49732530e+00 -4.46437657e-01 -9.50502396e-01 1.25206960e-02
-3.72285157e-01 -3.74748915e-01 6.45788610e-01 6.78354740e-01
-9.29187775e-01 1.43155420e+00 -1.39267945e+00 -6.30365089e-02
1.91642135e-01 5.70859790e-01 -1.59562781e-01 1.40577063e-01
-1.07464850e+00 8.06206167e-01 5.27413525e-02 7.37315863e-02
-1.19048309e+00 -7.01556802e-01 -3.66866440e-01 -2.51630962e-01
-1.58068255e-01 -9.78719056e-01 9.01315629e-01 -4.31137294e-01
-1.69643116e+00 5.49043477e-01 -5.81820607e-01 -6.05335116e-01
5.28564215e-01 3.19467187e-01 -3.04917008e-01 1.47822380e-01
-1.55422688e-01 4.99494553e-01 8.00537348e-01 -9.52798247e-01
3.00066680e-01 -4.29591715e-01 -2.97787964e-01 -1.98858067e-01
2.96373218e-01 -4.82352749e-02 1.71631634e-01 -4.56660300e-01
8.33250582e-02 -1.09500754e+00 -6.24093592e-01 -2.33859703e-01
-1.47306412e-01 2.06778273e-01 6.69018775e-02 -5.46279311e-01
1.06912351e+00 -1.70599151e+00 6.62031353e-01 3.53260100e-01
3.32593381e-01 5.28022237e-02 1.57940000e-01 6.43044531e-01
-1.66449636e-01 2.35991195e-01 -7.22654045e-01 -2.07841724e-01
1.17467530e-02 6.81593269e-02 -2.81608820e-01 6.60755038e-01
-4.29809913e-02 8.13789845e-01 -8.52194905e-01 -1.11887179e-01
1.63656056e-01 9.05086637e-01 -5.86078763e-01 -1.65692255e-01
-5.76185882e-01 8.64913404e-01 -4.66650367e-01 2.67398447e-01
7.19411194e-01 -4.76371080e-01 1.53085560e-01 -2.49411464e-01
-3.93183261e-01 2.35960752e-01 -1.12399173e+00 1.60606873e+00
-1.95241183e-01 -4.25194316e-02 2.15340406e-01 -8.13414454e-01
5.21657169e-01 2.21334234e-01 5.09292006e-01 -4.77880299e-01
1.15554318e-01 4.12591815e-01 2.71607995e-01 2.53511909e-02
3.75988871e-01 -8.11212957e-01 -6.94189519e-02 6.53267503e-01
-3.74096520e-02 -4.86796916e-01 1.63457349e-01 1.48342609e-01
1.12133968e+00 3.17731977e-01 2.03547269e-01 -4.89126086e-01
3.38573962e-01 2.79341242e-03 1.77596211e-01 1.03201640e+00
-2.62903571e-01 3.89247805e-01 4.92288828e-01 -3.84499162e-01
-1.57401395e+00 -1.34484792e+00 -3.82596791e-01 9.02635932e-01
2.77795922e-02 -5.07686913e-01 -1.03337121e+00 -1.94123521e-01
1.39544718e-02 7.24019825e-01 -7.02842236e-01 -3.95549089e-01
-5.24895847e-01 -1.52629316e+00 5.30057132e-01 2.41667628e-01
-2.24383231e-02 -1.12452424e+00 -2.43250683e-01 7.16458976e-01
3.37679088e-01 -4.46595073e-01 -4.96837080e-01 5.76603174e-01
-9.69036460e-01 -6.52488232e-01 -5.55817664e-01 -4.95238788e-02
5.69760621e-01 -5.58827892e-02 9.39730227e-01 -3.28619897e-01
-5.38193226e-01 2.00484216e-01 1.94497108e-01 1.19594857e-01
-6.99526608e-01 -1.53437093e-01 5.55453062e-01 -3.69093925e-01
2.39666849e-01 -9.78460729e-01 -9.36528444e-01 -1.50083601e-01
-8.87870550e-01 -1.90808967e-01 3.17480505e-01 7.53361344e-01
1.07332206e+00 -1.23280257e-01 2.62548447e-01 -7.15738237e-01
6.44478858e-01 -3.79114777e-01 -6.42015994e-01 -6.42502084e-02
-4.07069504e-01 5.98148823e-01 8.03694427e-01 -6.18370533e-01
-9.38883245e-01 -2.64014065e-01 -4.72798437e-01 -4.28002864e-01
-1.82394937e-01 4.00393367e-01 1.85455516e-01 -1.43266827e-01
7.71113217e-01 6.17393196e-01 -6.51692413e-03 -3.92053902e-01
4.85201776e-01 1.13687016e-01 1.70275763e-01 -1.21996808e+00
4.67954814e-01 7.84596682e-01 3.93543005e-01 -9.83919859e-01
-3.07365716e-01 5.35306381e-03 -6.09094679e-01 4.99595590e-02
1.04870117e+00 -7.93216825e-01 -1.30468035e+00 3.27944905e-01
-1.04986942e+00 -5.74753284e-01 -5.20927310e-01 6.39744878e-01
-7.10033417e-01 2.86553442e-01 -1.04553568e+00 -1.09561646e+00
-3.65510017e-01 -1.42330778e+00 1.03397059e+00 3.66198197e-02
-5.25228679e-01 -1.03379142e+00 3.71815890e-01 -4.92250547e-03
5.73176801e-01 3.27744812e-01 1.10089397e+00 -1.46275669e-01
-7.41642118e-01 -1.59359515e-01 1.01888701e-01 1.07570685e-01
-1.49571136e-01 2.90581256e-01 -6.73891366e-01 -4.84945536e-01
8.10694098e-02 -2.32602686e-01 1.25889897e+00 7.68042147e-01
1.17228949e+00 -2.13656023e-01 -4.21100378e-01 5.27382791e-01
1.17134607e+00 2.70362049e-01 5.14020085e-01 -1.90763772e-01
7.14026690e-01 1.61160260e-01 1.51094824e-01 5.13959527e-01
7.77225643e-02 3.90180588e-01 -1.56112676e-02 2.20861524e-01
1.03436328e-01 -2.72564858e-01 8.77335846e-01 1.12136233e+00
-5.52596927e-01 -1.02617919e-01 -7.23498464e-01 -4.84285168e-02
-1.43309903e+00 -1.31220889e+00 -3.59898895e-01 2.40895772e+00
1.06171203e+00 2.72083730e-01 1.12010099e-01 -2.30588913e-01
5.17166257e-01 3.31056267e-01 -1.02637911e+00 -3.10004324e-01
-1.59070104e-01 8.32434773e-01 5.92014730e-01 9.71541107e-01
-7.67255366e-01 9.76546645e-01 6.51794052e+00 1.05685031e+00
-1.28691447e+00 4.00151908e-01 8.02600026e-01 -3.27270150e-01
-5.39863110e-01 1.11736193e-01 -8.78220081e-01 8.48623276e-01
1.44158936e+00 -1.86295390e-01 5.97559392e-01 5.85432053e-01
6.00157976e-01 -1.52030796e-01 -9.53902721e-01 8.34608495e-01
-5.52508175e-01 -1.66136229e+00 9.31599960e-02 3.03243607e-01
7.52327859e-01 2.05770656e-01 3.17372829e-01 2.26329386e-01
6.73921049e-01 -1.24520946e+00 7.15947330e-01 7.61702418e-01
1.03027260e+00 -7.52574861e-01 3.50176364e-01 4.55337018e-01
-1.05879116e+00 5.75890064e-01 -6.11238658e-01 2.66324095e-02
3.19130957e-01 8.90451968e-01 -6.84961528e-02 1.54001027e-01
3.97903115e-01 4.56630349e-01 1.46273240e-01 4.40243632e-01
1.79157928e-01 6.94126606e-01 -5.84803522e-01 -2.19259009e-01
2.87807882e-01 -1.14945781e+00 4.73393589e-01 1.15894306e+00
5.67559302e-01 4.08047020e-01 -1.74345579e-02 1.27824199e+00
-1.23908997e-01 -3.56213421e-01 -4.13308978e-01 -3.53749663e-01
3.03396583e-01 9.22008038e-01 -7.55189478e-01 -2.82774717e-01
7.35710338e-02 1.10536635e+00 3.39436501e-01 6.80264533e-01
-1.09749520e+00 -1.78570300e-01 1.09849942e+00 3.31210673e-01
4.02469218e-01 -8.54877591e-01 1.27175868e-01 -1.13486624e+00
-1.52893245e-01 -6.57406747e-01 -9.24406424e-02 -4.85671133e-01
-1.36965358e+00 2.92510867e-01 -2.64897883e-01 -6.39597654e-01
-4.46913354e-02 -7.93599784e-01 -5.67526519e-01 1.16898906e+00
-1.19796240e+00 -6.08359456e-01 3.83085907e-01 3.43476027e-01
-1.08176842e-01 2.72002667e-01 9.71639454e-01 9.74763259e-02
-5.30503571e-01 1.56931072e-01 9.33622837e-01 -4.50400442e-01
5.83261430e-01 -1.23635006e+00 5.27742922e-01 3.84898484e-01
-9.27358717e-02 1.12170589e+00 1.11781538e+00 -9.02391911e-01
-1.83925343e+00 -1.06657362e+00 3.76962721e-01 -6.03056967e-01
8.76597226e-01 -5.50422788e-01 -1.15783858e+00 4.12740469e-01
-1.56803384e-01 3.98259372e-01 6.29438818e-01 -2.83202410e-01
-2.44048849e-01 1.51315674e-01 -1.27005160e+00 6.19853318e-01
1.14462948e+00 -9.23091054e-01 -1.40406355e-01 4.41665411e-01
4.79664773e-01 -3.53720278e-01 -1.25340676e+00 7.37550184e-02
5.41004419e-01 -1.07177305e+00 1.25870430e+00 -8.18700850e-01
9.44175199e-02 -3.92270267e-01 -2.88760215e-01 -1.20128214e+00
-2.68174350e-01 -9.54767287e-01 -3.23397279e-01 5.94673872e-01
3.87129247e-01 -8.10063004e-01 7.86387265e-01 5.50105929e-01
1.19226731e-01 -5.52274704e-01 -1.30343831e+00 -7.70981431e-01
7.51450062e-01 -5.51205397e-01 2.43461236e-01 5.83418310e-01
-1.90239340e-01 -5.60998209e-02 -1.93960294e-01 1.47811040e-01
9.49312150e-01 1.19494814e-02 4.73665744e-02 -8.33763123e-01
-5.00824809e-01 -4.14329112e-01 1.42603628e-02 -9.85340714e-01
1.71724811e-01 -8.51407111e-01 -7.94376880e-02 -9.80593920e-01
3.16200137e-01 -3.69276136e-01 -4.04343516e-01 -3.12170088e-01
-1.19892307e-01 4.63683456e-02 -2.10100442e-01 2.37689659e-01
-7.17398942e-01 9.22840238e-01 1.21790564e+00 -7.23353997e-02
-5.17538004e-02 -3.41998458e-01 -9.20377895e-02 5.75358450e-01
5.77150106e-01 -6.56326890e-01 -1.43419594e-01 1.06442712e-01
4.52077061e-01 3.36964071e-01 4.57401961e-01 -1.06374609e+00
2.37379223e-01 -3.08068156e-01 3.57754171e-01 -3.80852908e-01
6.84710383e-01 -1.98797718e-01 5.50908387e-01 6.03214502e-01
-3.88839722e-01 2.84012556e-02 6.49674460e-02 8.47301722e-01
1.72230005e-02 -2.24726386e-02 1.08047009e+00 -4.72892225e-01
8.72785524e-02 6.94546342e-01 -7.87828386e-01 1.12616807e-01
7.61387169e-01 2.95051597e-02 -1.42548531e-01 -2.33730465e-01
-9.13149774e-01 -3.62908810e-01 9.38218236e-01 -4.99473959e-01
3.45182300e-01 -1.03637064e+00 -4.45087731e-01 -7.70404562e-02
-2.77323335e-01 -2.96105910e-02 5.99427640e-01 8.61166835e-01
-5.92559218e-01 4.57414299e-01 1.03169695e-01 -4.93792593e-01
-6.50038362e-01 4.17707771e-01 4.34833318e-01 -4.38796192e-01
-2.87994206e-01 8.09529364e-01 4.28964198e-02 -3.72412652e-01
-4.08637881e-01 -3.89356196e-01 5.85103154e-01 -4.40687127e-02
4.62692887e-01 4.96170044e-01 6.11617276e-03 -5.78083515e-01
-2.07431242e-01 7.89357483e-01 -2.72796690e-01 -3.35686594e-01
1.27983475e+00 7.05766603e-02 -2.15230763e-01 7.53038883e-01
1.04556894e+00 -2.21813377e-02 -1.73572898e+00 1.39560148e-01
-1.95462972e-01 8.41806233e-02 -3.56295751e-03 -3.87468845e-01
-3.32781851e-01 1.28154659e+00 5.17188191e-01 -1.52991656e-02
2.04917639e-01 -1.07928477e-01 7.09307849e-01 4.83197063e-01
6.74577892e-01 -9.23249185e-01 -1.62715167e-01 6.80964649e-01
4.58538622e-01 -1.01717520e+00 5.53751960e-02 2.99566761e-02
-3.30334067e-01 1.17510617e+00 6.15210384e-02 -2.57216305e-01
6.06492698e-01 1.97638169e-01 -5.33349574e-01 -1.96545452e-01
-7.16892004e-01 8.91255587e-02 -1.30552053e-02 3.25850993e-01
6.90291524e-01 3.03757459e-01 -5.32461107e-02 5.02981424e-01
8.26213732e-02 -1.12077571e-01 4.26115781e-01 9.46533859e-01
-4.87558484e-01 -1.29634845e+00 -2.61948973e-01 3.21323037e-01
-2.73418009e-01 -3.16129953e-01 1.20654613e-01 5.73714495e-01
-1.08473741e-01 3.37148279e-01 8.75758938e-03 1.27177879e-01
-6.62782714e-02 4.64215249e-01 9.00599182e-01 -3.58900011e-01
-3.46508592e-01 -5.69639765e-02 -1.58613279e-01 -6.05038822e-01
-2.16109276e-01 -8.34018052e-01 -1.50234294e+00 -9.36069965e-01
3.22547518e-02 4.01800454e-01 5.42587340e-01 8.16341758e-01
4.87714529e-01 4.59443748e-01 1.03536606e-01 -1.19618225e+00
-7.65597284e-01 -7.34585524e-01 -8.32318664e-01 3.77488285e-01
4.40205932e-01 -4.69962060e-01 -5.85730135e-01 -8.87443945e-02] | [5.137246131896973, 5.1673479080200195] |
a484fb3d-3c28-461f-b3cb-2ce7414dd691 | abstractive-text-summarization-using-sequence | 1602.06023 | null | http://arxiv.org/abs/1602.06023v5 | http://arxiv.org/pdf/1602.06023v5.pdf | Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyond | In this work, we model abstractive text summarization using Attentional
Encoder-Decoder Recurrent Neural Networks, and show that they achieve
state-of-the-art performance on two different corpora. We propose several novel
models that address critical problems in summarization that are not adequately
modeled by the basic architecture, such as modeling key-words, capturing the
hierarchy of sentence-to-word structure, and emitting words that are rare or
unseen at training time. Our work shows that many of our proposed models
contribute to further improvement in performance. We also propose a new dataset
consisting of multi-sentence summaries, and establish performance benchmarks
for further research. | ['Bo-Wen Zhou', 'Cicero Nogueira dos santos', 'Bing Xiang', 'Ramesh Nallapati', 'Caglar Gulcehre'] | 2016-02-19 | abstractive-text-summarization-using-sequence-1 | https://aclanthology.org/K16-1028 | https://aclanthology.org/K16-1028.pdf | conll-2016-8 | ['summarization', 'abstractive-sentence-summarization'] | ['natural-language-processing', 'natural-language-processing'] | [ 5.16824245e-01 1.11921817e-01 -4.50768173e-01 -2.00888649e-01
-1.06317556e+00 -1.17165700e-01 3.40568811e-01 4.74483609e-01
-3.86374474e-01 9.06700730e-01 1.19966149e+00 -2.41618052e-01
2.13162228e-01 -5.27547419e-01 -6.20580256e-01 -2.92263836e-01
1.42895386e-01 2.85688579e-01 1.34175435e-01 -5.60202241e-01
7.41934001e-01 1.27768651e-01 -1.42940640e+00 7.05689967e-01
1.07860291e+00 3.49826008e-01 2.50098765e-01 1.06389856e+00
-4.28699076e-01 1.04986227e+00 -1.19866502e+00 -1.90227807e-01
-4.42805290e-01 -8.46918941e-01 -9.41579819e-01 -2.69087791e-01
7.56770253e-01 -3.55538994e-01 -5.66213310e-01 9.60864604e-01
8.12102079e-01 3.74680519e-01 6.91666424e-01 -3.59719306e-01
-1.17077053e+00 1.19380605e+00 -4.42446828e-01 7.14735746e-01
4.82823431e-01 -2.25493968e-01 1.26988614e+00 -7.52168417e-01
5.83707094e-01 1.09968412e+00 6.02286994e-01 9.08310354e-01
-8.93597305e-01 -4.26477045e-01 4.37633485e-01 1.98708996e-01
-8.02165508e-01 -8.87163699e-01 8.71918023e-01 4.55173627e-02
1.63574672e+00 5.02581120e-01 4.99942362e-01 1.46257949e+00
7.13736951e-01 1.26949787e+00 1.47809789e-01 -5.72586536e-01
1.59999486e-02 -2.74177074e-01 7.54691005e-01 4.43324804e-01
5.73115528e-01 -5.55292904e-01 -7.68512428e-01 -2.95474753e-02
2.36313328e-01 -8.82556848e-03 -4.75135922e-01 3.24595720e-01
-9.54265773e-01 9.66296256e-01 1.59307092e-01 4.34043378e-01
-6.55942261e-01 7.65413865e-02 7.38810599e-01 9.04269144e-02
8.97180319e-01 7.34937191e-01 -3.21688324e-01 -3.10687363e-01
-1.43532813e+00 1.57657385e-01 8.50484788e-01 1.11399043e+00
3.28693211e-01 5.65670192e-01 -7.64213681e-01 8.94292653e-01
-1.99604154e-01 3.21233988e-01 1.06891227e+00 -7.55383670e-01
7.27798760e-01 4.42838967e-01 -2.70895734e-02 -6.98619485e-01
-4.31651831e-01 -6.48594320e-01 -1.12794399e+00 -8.13460827e-01
-5.53943455e-01 -2.58571118e-01 -8.79734695e-01 1.44597006e+00
-5.47411859e-01 3.06202155e-02 4.43764061e-01 2.37567544e-01
1.43895030e+00 1.17901075e+00 -6.32794350e-02 -7.03996956e-01
1.01417816e+00 -1.36450827e+00 -1.21119988e+00 -5.86827934e-01
6.38591647e-01 -5.33116996e-01 9.94922578e-01 1.64566431e-02
-1.63502467e+00 -3.99583369e-01 -1.09851992e+00 -3.86252612e-01
-2.06306383e-01 1.86079323e-01 4.26747769e-01 2.25200012e-01
-1.30356681e+00 7.06318259e-01 -7.18677640e-01 -6.17960036e-01
3.49822611e-01 6.33149073e-02 2.33035579e-01 2.96149611e-01
-1.23024940e+00 1.01007926e+00 6.39784694e-01 -1.65248252e-02
-6.99866176e-01 -4.36715722e-01 -9.62084115e-01 5.60867548e-01
1.21932805e-01 -1.11586463e+00 1.58064437e+00 -5.15289187e-01
-1.45988238e+00 3.81543547e-01 -7.04491615e-01 -7.97297001e-01
-1.66881353e-01 -6.44124985e-01 -4.61793542e-01 1.85607716e-01
1.33829609e-01 4.64841276e-01 5.39043844e-01 -1.05822444e+00
-4.92993683e-01 -1.61017552e-01 -1.60422102e-01 3.56008559e-01
-7.97201276e-01 1.47798538e-01 -1.80968180e-01 -8.94996881e-01
-3.04348469e-01 -4.12429869e-01 -3.11857790e-01 -1.10871255e+00
-8.18300307e-01 -4.08791006e-01 5.33548772e-01 -8.59450579e-01
1.96585572e+00 -1.63553417e+00 3.74171257e-01 -6.58327758e-01
6.47289157e-02 5.45618236e-01 -3.57034594e-01 9.07782316e-01
2.86476295e-02 3.95635515e-01 -2.76564300e-01 -6.44148469e-01
-2.34910011e-01 -7.91969895e-03 -8.23995411e-01 -1.56184733e-01
1.82940453e-01 1.26312685e+00 -9.41055179e-01 -6.25247419e-01
-6.93316385e-02 7.33330101e-02 -2.93804944e-01 2.26065502e-01
-2.81616628e-01 -1.60745442e-01 -5.40347755e-01 4.51408803e-01
6.89105317e-02 -1.82782099e-01 -1.72532737e-01 9.52709168e-02
-1.15703940e-01 8.66007447e-01 -3.75415385e-01 1.80770123e+00
-3.72271240e-01 9.13254082e-01 -3.71456891e-01 -9.13345754e-01
7.95289874e-01 2.57576168e-01 1.04468623e-02 -5.12247384e-01
1.20287701e-01 -1.77567434e-02 -1.81274712e-01 -5.53712249e-01
1.58612907e+00 5.02526872e-02 -2.01372504e-01 5.57329178e-01
2.23114103e-01 -1.01893529e-01 5.56203067e-01 5.40674210e-01
1.10858417e+00 -3.89160424e-01 5.94708383e-01 -2.11265534e-01
3.61922622e-01 6.75047340e-04 3.68064582e-01 1.27124453e+00
1.43144056e-01 7.18414068e-01 4.68071818e-01 -1.59982890e-01
-9.41541135e-01 -8.31287920e-01 1.49589688e-01 1.31473923e+00
-1.05711617e-01 -7.98154593e-01 -8.75810742e-01 -5.86794436e-01
-4.03760910e-01 1.45572841e+00 -6.13896668e-01 -4.26379681e-01
-7.83220410e-01 -6.90539420e-01 7.18857527e-01 7.77071416e-01
2.54624277e-01 -1.50726485e+00 -6.25655770e-01 4.10466820e-01
-4.22334284e-01 -7.41937757e-01 -7.52016425e-01 2.03902960e-01
-1.30383253e+00 -5.34321904e-01 -8.32546294e-01 -9.72077966e-01
4.10222113e-01 6.07556283e-01 1.48934209e+00 5.53614013e-02
-7.79011175e-02 2.32355639e-01 -5.04270017e-01 -6.47725403e-01
-6.01329863e-01 9.23744202e-01 -1.47655502e-01 -5.01331747e-01
2.12216794e-01 -4.82970536e-01 -3.86345237e-01 -4.58761275e-01
-1.06887531e+00 1.09703317e-01 7.18306839e-01 9.30100322e-01
2.31078118e-01 -3.99861902e-01 1.13478303e+00 -1.06817329e+00
1.61432517e+00 -3.54620188e-01 2.32934773e-01 5.04163623e-01
-3.74825865e-01 2.34556451e-01 8.03526044e-01 -3.35400939e-01
-1.20164895e+00 -6.04775190e-01 -4.43610489e-01 -4.58320752e-02
2.09045950e-02 8.44087958e-01 2.59591520e-01 6.14220262e-01
6.09620333e-01 7.65114903e-01 -3.20796251e-01 -5.25364578e-01
4.45682794e-01 9.78513181e-01 5.11485100e-01 -2.15780318e-01
1.51577324e-01 1.46936134e-01 -5.28228045e-01 -1.08511865e+00
-1.34329188e+00 -5.39385676e-01 -4.35446799e-01 1.08723283e-01
5.92412829e-01 -7.83357739e-01 -1.98257282e-01 1.87844142e-01
-1.62049723e+00 4.56843823e-02 -6.61456764e-01 2.55441397e-01
-4.08779383e-01 6.91923976e-01 -9.00133669e-01 -8.39523196e-01
-1.32975864e+00 -8.12967479e-01 1.10785902e+00 4.83626097e-01
-5.68601251e-01 -1.08061850e+00 3.52108538e-01 3.10456418e-02
7.39716411e-01 -2.34714851e-01 1.05436075e+00 -1.01643443e+00
-6.92101792e-02 -2.05273792e-01 1.31747991e-01 4.02416289e-01
1.41135931e-01 1.55863119e-02 -6.21570289e-01 -3.24746668e-01
2.72579372e-01 -4.40876186e-01 1.74604201e+00 7.92818069e-01
1.27371609e+00 -5.73152184e-01 -3.02164286e-01 2.79894590e-01
8.96319091e-01 8.28316286e-02 7.34513760e-01 1.98437452e-01
5.76785922e-01 4.15674299e-01 2.59683609e-01 3.55107576e-01
4.26649600e-01 7.84371868e-02 1.45670041e-01 1.02142639e-01
-2.00617865e-01 -4.59211975e-01 5.97232401e-01 1.73830688e+00
7.72852227e-02 -8.54910374e-01 -4.80055004e-01 9.21115756e-01
-2.10243964e+00 -1.29952884e+00 -1.75907388e-01 1.77559566e+00
9.46268618e-01 3.69548827e-01 -1.05145946e-01 -3.06589931e-01
7.93295085e-01 8.04110646e-01 -5.58194637e-01 -9.54815686e-01
-3.68705511e-01 3.31646562e-01 2.18455046e-01 5.29761016e-01
-9.14622009e-01 1.26477265e+00 7.84074402e+00 8.16835821e-01
-9.37499642e-01 -9.15357545e-02 5.85398674e-01 -3.67666632e-01
-5.17351151e-01 -2.75034726e-01 -1.02753043e+00 2.43081108e-01
1.36074698e+00 -7.01345861e-01 1.30937412e-01 8.12559962e-01
2.23931983e-01 1.38092101e-01 -1.06845748e+00 5.20756423e-01
8.46709967e-01 -1.75714493e+00 7.27746248e-01 -3.01817954e-01
9.58054185e-01 1.47963211e-01 -8.62212852e-02 6.90366983e-01
2.56646246e-01 -9.80937660e-01 3.16126943e-01 5.60881972e-01
5.31576514e-01 -6.68216467e-01 9.26771104e-01 5.14267921e-01
-7.40139365e-01 1.62500292e-01 -7.09185004e-01 -1.68022394e-01
3.39470029e-01 3.43473047e-01 -6.61535263e-01 7.80109107e-01
2.20394686e-01 1.18106592e+00 -6.94033682e-01 8.60248506e-01
-3.54751050e-01 9.53431010e-01 1.95186213e-01 -6.22927129e-01
3.01325172e-01 1.23232886e-01 8.25138986e-01 1.71144176e+00
3.67486209e-01 -1.10263117e-01 -1.07446715e-01 6.34561598e-01
-4.07025307e-01 2.56266445e-01 -6.67163134e-01 -8.00449401e-02
3.81838948e-01 8.30978155e-01 -3.17018718e-01 -6.82828128e-01
-1.65576547e-01 9.84038472e-01 5.60537159e-01 4.79222417e-01
-5.65350711e-01 -7.46013343e-01 3.56613845e-01 -2.65378386e-01
4.07302082e-01 -1.64672419e-01 -3.52498621e-01 -1.52254713e+00
7.33373463e-02 -6.92662418e-01 3.72329742e-01 -6.93398714e-01
-1.19267583e+00 7.25105345e-01 9.42891166e-02 -8.41831386e-01
-5.68422198e-01 -7.68607482e-02 -1.12004936e+00 5.71966946e-01
-1.65816128e+00 -7.87355423e-01 8.74530971e-02 9.38415080e-02
1.47899055e+00 -2.81659931e-01 9.68977094e-01 -1.50007486e-01
-9.75863338e-01 5.25727272e-01 4.61698383e-01 -1.35215491e-01
6.13605201e-01 -1.08808243e+00 9.36616778e-01 9.15641069e-01
3.09331149e-01 8.49352419e-01 8.27188730e-01 -7.30012476e-01
-1.09119189e+00 -1.19685578e+00 1.39486921e+00 -3.44910443e-01
5.04965484e-01 -2.00638026e-01 -1.03963196e+00 9.39027011e-01
9.69717860e-01 -8.47030044e-01 7.88738728e-01 3.53002369e-01
7.20021650e-02 2.36709684e-01 -5.28331339e-01 8.02375436e-01
1.07796299e+00 -2.91132182e-01 -1.41965258e+00 3.86885881e-01
1.15488517e+00 -3.53154182e-01 -3.28964561e-01 2.01187044e-01
2.89170414e-01 -7.29921222e-01 6.64682150e-01 -1.11152232e+00
1.04228139e+00 3.48274231e-01 1.17121130e-01 -1.77061284e+00
-4.52287585e-01 -7.79712856e-01 -6.91167653e-01 1.40690744e+00
5.14824629e-01 -4.28899497e-01 6.05398178e-01 6.31620409e-03
-6.96453989e-01 -7.79215515e-01 -8.35272908e-01 -5.89373291e-01
3.84524345e-01 -6.99530318e-02 3.88581216e-01 5.71297646e-01
2.67458051e-01 1.11329579e+00 -4.68385667e-01 -5.11601567e-01
1.77291945e-01 1.47182390e-01 4.05398607e-01 -1.00606120e+00
1.89778611e-01 -7.52366543e-01 1.51303917e-01 -1.54274547e+00
5.96877217e-01 -8.28651428e-01 1.08199716e-01 -2.18181109e+00
6.07850671e-01 6.13588393e-01 -3.75235349e-01 1.74660459e-01
-5.99828303e-01 -1.97659507e-01 7.21560121e-02 1.98161632e-01
-1.06596816e+00 1.14026916e+00 9.88775373e-01 -4.45706695e-01
-2.35422745e-01 -1.67807862e-02 -1.22245467e+00 5.59926748e-01
9.23098445e-01 -5.83340943e-01 -3.14676195e-01 -8.01582098e-01
3.58009711e-02 1.72409117e-01 -2.56477535e-01 -7.51799345e-01
4.88080531e-01 7.17170164e-03 2.17780665e-01 -1.16695178e+00
2.83507526e-01 -8.15214962e-02 -7.12825716e-01 4.13590372e-01
-1.09480381e+00 3.93547118e-01 2.73014039e-01 6.78703845e-01
-2.76797205e-01 -7.79781282e-01 2.09991142e-01 -3.04328024e-01
-1.74966663e-01 -1.85418472e-01 -6.90146506e-01 4.67620641e-01
5.33821225e-01 -7.38436803e-02 -6.34568214e-01 -8.05550575e-01
-2.44164601e-01 3.81948948e-01 2.05151618e-01 5.67683995e-01
8.27928603e-01 -8.80324066e-01 -1.27306569e+00 -1.54264644e-01
2.65314858e-02 -1.44154578e-01 2.27510437e-01 1.85182273e-01
-3.08536470e-01 9.13122892e-01 4.77920324e-02 -2.99033135e-01
-1.28790700e+00 3.66153330e-01 -6.38170764e-02 -7.35276818e-01
-7.47232437e-01 6.73822284e-01 1.00625791e-02 -6.75906837e-02
3.42374951e-01 -4.28361386e-01 -6.02568686e-01 2.35616952e-01
7.21733153e-01 4.85443443e-01 9.88650545e-02 -4.22154099e-01
-1.03432246e-01 1.96804374e-01 -7.05837131e-01 1.74754873e-01
1.55004275e+00 -1.92470044e-01 -2.39385083e-01 8.05205643e-01
9.78974640e-01 5.30171348e-03 -5.59671164e-01 -2.51279593e-01
8.84967521e-02 5.85682206e-02 2.62320172e-02 -6.48117721e-01
-6.63059652e-01 9.94511425e-01 -1.62987754e-01 4.22972858e-01
1.02878046e+00 -7.73847848e-02 1.19667947e+00 8.20229292e-01
-2.53420830e-01 -1.18481922e+00 2.16334090e-01 1.11047018e+00
1.01191187e+00 -9.17676628e-01 2.60115743e-01 6.34107590e-02
-7.06026852e-01 1.25214434e+00 5.36986351e-01 -2.30991334e-01
-1.16047658e-01 3.05258241e-02 -2.45729029e-01 -2.23327130e-01
-1.31915617e+00 -8.92200917e-02 2.41694912e-01 2.18796715e-01
7.65459120e-01 -1.26864478e-01 -5.56625485e-01 7.13402152e-01
-3.85428607e-01 -2.01161757e-01 1.01835299e+00 9.11079526e-01
-7.96138704e-01 -7.74547160e-01 -3.71159166e-02 8.61854434e-01
-6.60799921e-01 -6.42083049e-01 -7.50004590e-01 5.09987652e-01
-8.01190794e-01 1.14332604e+00 1.41225323e-01 -1.57032639e-01
4.84649599e-01 3.53346199e-01 2.44958699e-01 -1.07672453e+00
-9.94444728e-01 -4.07922119e-02 4.27587181e-01 7.22520724e-02
-1.93599582e-01 -6.17905080e-01 -8.64718795e-01 -1.09431162e-01
-5.36069989e-01 4.85480934e-01 4.45691854e-01 7.08239973e-01
7.31894433e-01 1.20105755e+00 5.16258359e-01 -8.79405141e-01
-8.05934548e-01 -1.57562304e+00 -2.51344532e-01 1.39267549e-01
6.14428520e-01 1.10931978e-01 -3.42562914e-01 -4.71650809e-02] | [12.467828750610352, 9.447436332702637] |
9da19f07-bf11-43d4-8ce2-084974ebf9f3 | improving-vision-and-language-navigation-by | 2304.04907 | null | https://arxiv.org/abs/2304.04907v1 | https://arxiv.org/pdf/2304.04907v1.pdf | Improving Vision-and-Language Navigation by Generating Future-View Image Semantics | Vision-and-Language Navigation (VLN) is the task that requires an agent to navigate through the environment based on natural language instructions. At each step, the agent takes the next action by selecting from a set of navigable locations. In this paper, we aim to take one step further and explore whether the agent can benefit from generating the potential future view during navigation. Intuitively, humans will have an expectation of how the future environment will look like, based on the natural language instructions and surrounding views, which will aid correct navigation. Hence, to equip the agent with this ability to generate the semantics of future navigation views, we first propose three proxy tasks during the agent's in-domain pre-training: Masked Panorama Modeling (MPM), Masked Trajectory Modeling (MTM), and Action Prediction with Image Generation (APIG). These three objectives teach the model to predict missing views in a panorama (MPM), predict missing steps in the full trajectory (MTM), and generate the next view based on the full instruction and navigation history (APIG), respectively. We then fine-tune the agent on the VLN task with an auxiliary loss that minimizes the difference between the view semantics generated by the agent and the ground truth view semantics of the next step. Empirically, our VLN-SIG achieves the new state-of-the-art on both the Room-to-Room dataset and the CVDN dataset. We further show that our agent learns to fill in missing patches in future views qualitatively, which brings more interpretability over agents' predicted actions. Lastly, we demonstrate that learning to predict future view semantics also enables the agent to have better performance on longer paths. | ['Mohit Bansal', 'Jialu Li'] | 2023-04-11 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Li_Improving_Vision-and-Language_Navigation_by_Generating_Future-View_Image_Semantics_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Li_Improving_Vision-and-Language_Navigation_by_Generating_Future-View_Image_Semantics_CVPR_2023_paper.pdf | cvpr-2023-1 | ['vision-and-language-navigation', 'trajectory-modeling'] | ['robots', 'time-series'] | [ 3.13825548e-01 4.11971956e-01 -3.83715797e-03 -8.22795272e-01
-5.59727967e-01 -5.73042572e-01 9.92995858e-01 -3.27765405e-01
-3.42663020e-01 6.06128752e-01 5.46601057e-01 -4.54273999e-01
1.35567456e-01 -1.02109957e+00 -1.12059617e+00 -4.40246701e-01
-1.26958072e-01 6.85357988e-01 2.28936132e-02 -3.16633105e-01
2.20904291e-01 2.28635550e-01 -1.72144413e+00 4.33238506e-01
9.04713273e-01 6.77915812e-01 8.17584753e-01 7.77357578e-01
2.21574470e-01 1.28473186e+00 1.00931428e-01 7.63656273e-02
3.59793097e-01 -4.31260854e-01 -8.71907890e-01 2.73273498e-01
4.73456711e-01 -7.74465621e-01 -5.35862386e-01 7.63527513e-01
1.05038635e-01 6.76402986e-01 6.70760751e-01 -1.28167403e+00
-3.20138246e-01 5.17171204e-01 -1.49789482e-01 -3.02077532e-01
6.59984529e-01 8.18104386e-01 8.33783805e-01 -8.49005580e-01
1.07803679e+00 1.47989666e+00 3.56046669e-02 7.65177608e-01
-9.98498082e-01 -2.59509772e-01 7.88905680e-01 3.40377569e-01
-6.83728874e-01 -3.85181904e-01 6.37157202e-01 -4.10251707e-01
1.03270626e+00 -5.13770208e-02 5.80599844e-01 1.31355321e+00
2.95682311e-01 8.92465353e-01 1.04037225e+00 -2.58634776e-01
4.32801008e-01 -9.26332176e-02 -2.93826133e-01 1.13666034e+00
-3.76276553e-01 5.90968668e-01 -6.49816692e-01 2.29383513e-01
6.48076534e-01 1.70567498e-01 -4.91156489e-01 -8.08022320e-01
-1.44836605e+00 8.00389349e-01 6.27919495e-01 -2.45462433e-01
-7.65625894e-01 1.47448182e-01 4.02563810e-02 1.25632584e-01
-1.79952547e-01 4.99754012e-01 -5.25751472e-01 -1.59131780e-01
-4.39498901e-01 4.48400289e-01 5.68401575e-01 9.55328166e-01
8.64348114e-01 1.59641821e-02 -1.91898104e-02 3.23957801e-01
4.89179462e-01 6.85630739e-01 3.30007434e-01 -1.71317005e+00
8.29707026e-01 5.84899783e-01 3.70438755e-01 -5.75863361e-01
-4.48274761e-01 1.02262519e-01 -4.76869076e-01 8.04926455e-01
5.56177616e-01 -2.83648342e-01 -1.28674281e+00 2.13699174e+00
4.82786149e-01 -1.41326776e-02 4.55465287e-01 9.46254313e-01
4.49846685e-01 8.90465558e-01 -3.98440734e-02 1.28829941e-01
1.05467916e+00 -1.47386205e+00 -3.28645676e-01 -8.50990832e-01
7.89789498e-01 -3.37468982e-01 1.38488281e+00 2.97317058e-01
-8.60359430e-01 -7.47631192e-01 -8.13932240e-01 -3.06928288e-02
-9.51439068e-02 1.06184594e-01 6.16111577e-01 -2.72357017e-02
-1.07964158e+00 3.65392804e-01 -1.05887198e+00 -4.64213312e-01
1.44082919e-01 4.47711386e-02 -6.65647864e-01 -3.22496712e-01
-9.26226318e-01 1.01897967e+00 1.89369902e-01 5.25296964e-02
-1.61744535e+00 -5.28938055e-01 -1.18894923e+00 -2.21492410e-01
6.62958145e-01 -1.02056968e+00 1.31191266e+00 -7.53965199e-01
-1.40410459e+00 6.32664502e-01 -5.15468597e-01 -7.46240139e-01
7.20052779e-01 -1.63470328e-01 -1.65222492e-02 -1.25933573e-01
3.58995736e-01 1.28393483e+00 5.57646930e-01 -1.59297490e+00
-1.17995727e+00 -5.49049795e-01 6.02541447e-01 7.38114834e-01
5.38138390e-01 -9.87195134e-01 -4.48139578e-01 -1.41398773e-01
2.33670577e-01 -1.29985225e+00 -4.93303627e-01 9.83057097e-02
-4.61534441e-01 -3.92528772e-02 4.95610237e-01 -7.25699484e-01
4.63928223e-01 -1.93620884e+00 4.57158357e-01 7.20477924e-02
-2.36661788e-02 -3.55370760e-01 -3.51391554e-01 3.91023219e-01
1.87632307e-01 -3.19096327e-01 -2.01196238e-01 -4.77934659e-01
-9.28307511e-03 4.43464607e-01 -7.14654863e-01 5.21542616e-02
-4.55384016e-01 8.04994166e-01 -9.24146831e-01 6.06459333e-03
4.78656381e-01 2.34074995e-01 -7.42486537e-01 3.91123980e-01
-7.65824795e-01 8.85032594e-01 -5.71619928e-01 2.20721498e-01
3.34839731e-01 -2.57720709e-01 1.88251033e-01 1.07212901e-01
-1.97848126e-01 3.75702858e-01 -8.77681077e-01 1.97857356e+00
-9.88221467e-01 5.32386065e-01 -2.27921139e-02 -4.54479963e-01
6.43683732e-01 -1.01618133e-01 1.41392365e-01 -9.97260094e-01
-3.70280653e-01 -1.48363225e-02 -6.44948035e-02 -6.43443406e-01
5.71761668e-01 9.86962467e-02 -6.47657216e-02 6.81372285e-01
-3.02367955e-01 7.38568753e-02 -5.60297705e-02 2.80524760e-01
1.03666723e+00 6.03443861e-01 1.76837996e-01 2.84107596e-01
4.63585079e-01 3.58616441e-01 2.52045155e-01 1.00347745e+00
-7.08087608e-02 4.15551543e-01 2.23164946e-01 -6.76019847e-01
-9.93639708e-01 -1.20545542e+00 6.11877203e-01 1.13319206e+00
3.52548063e-01 -5.52525260e-02 -7.44093001e-01 -1.01746082e+00
-3.30335647e-01 1.51026535e+00 -7.29797363e-01 -1.61223188e-01
-9.16939855e-01 -1.03631735e-01 -9.23033059e-02 6.28986180e-01
6.11638725e-01 -1.54123068e+00 -1.21525550e+00 -2.08797045e-02
-6.79625988e-01 -1.04186714e+00 -6.75098836e-01 6.50095642e-02
-8.04876387e-01 -1.05683517e+00 -1.86041340e-01 -7.94984460e-01
8.58632863e-01 3.30333233e-01 9.99527216e-01 -1.68042362e-01
2.54565328e-01 6.29821062e-01 -2.54235953e-01 6.27225339e-02
-6.36378348e-01 -2.03883812e-01 -1.04999440e-02 -2.11731553e-01
-1.00014443e-02 -5.45157135e-01 -7.32841313e-01 2.30638176e-01
-3.99246931e-01 7.72909343e-01 6.30355000e-01 7.08172381e-01
7.57466733e-01 -5.43334410e-02 -7.36653665e-03 -7.88201928e-01
2.18797088e-01 -2.38722041e-01 -7.86765635e-01 2.00886652e-01
-6.98851764e-01 4.97912973e-01 7.87718832e-01 -1.33150190e-01
-1.40531290e+00 3.62182409e-01 -2.92142667e-02 -2.09691778e-01
-4.47259754e-01 2.96625584e-01 -3.65255743e-01 4.31616783e-01
5.33454776e-01 5.90087950e-01 1.59843150e-03 -1.28760532e-01
7.44734585e-01 2.15575956e-02 5.70602417e-01 -4.82733488e-01
5.74189603e-01 7.49374926e-01 -5.53532839e-02 -3.64295185e-01
-8.49952757e-01 -3.14536989e-02 -5.60361981e-01 -2.10234448e-01
1.05557990e+00 -1.03081560e+00 -7.08875716e-01 2.05873862e-01
-1.16620815e+00 -1.00585628e+00 -3.03022593e-01 4.11740750e-01
-1.10877502e+00 3.19014676e-02 3.21204737e-02 -7.28635073e-01
3.39373574e-02 -1.49575067e+00 9.78012323e-01 2.07311779e-01
-2.23628953e-01 -9.75806057e-01 1.55930351e-02 5.50579667e-01
3.00688716e-03 1.59366280e-01 1.01827121e+00 -3.87100607e-01
-1.16833496e+00 2.53014594e-01 1.21442258e-01 -6.88152537e-02
-2.44066082e-02 -5.62774897e-01 -7.61143625e-01 -2.25817904e-01
3.97160724e-02 -2.61703193e-01 1.02338588e+00 5.76349616e-01
9.52987731e-01 -5.65965235e-01 -5.03514886e-01 6.87115848e-01
1.30543387e+00 7.61372209e-01 6.44850910e-01 6.37276053e-01
7.36684620e-01 9.36437488e-01 1.08856213e+00 3.30995888e-01
9.88142252e-01 6.99823558e-01 9.80656326e-01 2.88902819e-01
-2.72461791e-02 -8.42026830e-01 6.85121655e-01 -2.02879682e-02
2.25226954e-02 -4.83694941e-01 -8.20163667e-01 5.12646794e-01
-2.12395024e+00 -1.06354201e+00 2.80076474e-01 2.30571604e+00
3.19262415e-01 1.45200327e-01 -1.42074943e-01 -5.93211710e-01
2.50512213e-01 4.36357051e-01 -9.79431570e-01 -1.62902683e-01
2.53410369e-01 -3.55866104e-01 3.24491382e-01 1.08032465e+00
-9.33482051e-01 1.11584783e+00 4.94035959e+00 2.00535432e-01
-8.14014852e-01 -1.52881905e-01 6.64727569e-01 -2.36639325e-02
-4.70103741e-01 2.61771709e-01 -9.65443850e-01 2.43424147e-01
7.40313649e-01 2.54438251e-01 9.37654018e-01 1.04664314e+00
6.53006732e-01 -3.40309143e-01 -1.53745484e+00 6.52066052e-01
1.50877004e-02 -1.33590329e+00 2.95669347e-01 2.11839169e-01
7.72143126e-01 1.46432564e-01 2.33537659e-01 4.85032260e-01
7.11647809e-01 -1.04263461e+00 9.42275286e-01 6.97697103e-01
4.03487533e-01 -5.86675882e-01 3.92756313e-01 8.46486568e-01
-1.03776991e+00 -4.00700271e-01 -4.45261970e-02 -1.16016977e-01
6.33033633e-01 -1.64028853e-01 -1.18298793e+00 2.80190855e-01
3.93712193e-01 5.21081924e-01 -3.76004726e-01 4.52945679e-01
-5.11700511e-01 3.01509593e-02 -1.01056635e-01 5.02661392e-02
5.54761887e-01 -3.93792123e-01 6.55750692e-01 3.91979247e-01
4.43301290e-01 1.46494433e-01 4.38203216e-01 9.98445332e-01
2.13661477e-01 -4.79478687e-01 -8.92103195e-01 3.49666297e-01
4.68835294e-01 8.76020789e-01 -4.51868296e-01 -3.31317604e-01
-2.69253373e-01 1.09488845e+00 2.99294323e-01 5.78823686e-01
-8.23670924e-01 2.25340828e-01 7.25926220e-01 1.95942029e-01
3.23169500e-01 -1.65519968e-01 -1.53679550e-01 -9.37743604e-01
1.25668645e-01 -9.51701820e-01 1.96023762e-01 -1.12095046e+00
-6.78005815e-01 6.35705113e-01 -1.69222727e-01 -1.21078777e+00
-9.27120507e-01 -4.77459222e-01 -5.45649886e-01 7.90401399e-01
-1.39971185e+00 -1.32178116e+00 -5.47545433e-01 4.05139804e-01
1.27454722e+00 -1.90809384e-01 7.41226614e-01 -3.36797476e-01
3.59461606e-02 4.74267714e-02 -2.68283069e-01 -9.28093567e-02
4.65492696e-01 -1.35184896e+00 6.16834819e-01 8.46432745e-01
2.07232088e-01 4.80875075e-01 7.88755655e-01 -7.28384256e-01
-1.30851138e+00 -1.25358629e+00 7.15370357e-01 -7.45552957e-01
1.94801748e-01 -3.61678749e-02 -5.39242029e-01 1.21091187e+00
-4.41339165e-02 -3.13596845e-01 1.07193226e-02 -2.07356393e-01
-1.39004752e-01 1.45173371e-01 -9.18862641e-01 1.08265626e+00
1.20659399e+00 -3.32892597e-01 -5.57431102e-01 3.13635111e-01
8.17858398e-01 -6.27100348e-01 -9.53435823e-02 2.28918672e-01
5.97296476e-01 -1.31966352e+00 1.06106591e+00 -5.79651356e-01
7.48068094e-01 -4.62127596e-01 -4.85346138e-01 -1.48957479e+00
-1.18591309e-01 -3.61385912e-01 1.18785619e-03 7.03876972e-01
6.98524654e-01 -4.43096727e-01 9.49499130e-01 7.93213844e-01
-2.63749599e-01 -7.41517842e-01 -7.32523322e-01 -2.92614281e-01
-2.54961550e-01 -4.37836468e-01 6.65699005e-01 3.44490618e-01
-3.33092749e-01 3.79614413e-01 -4.65379119e-01 5.27303398e-01
5.68050146e-01 4.68901277e-01 1.19788814e+00 -7.44340479e-01
-3.35092843e-01 -7.04062581e-02 1.88024715e-01 -1.71466947e+00
3.53689224e-01 -7.38489509e-01 4.25972909e-01 -2.06650400e+00
1.33035392e-01 -3.10669810e-01 2.95435011e-01 4.44541901e-01
-5.05046174e-02 -5.87552726e-01 4.06081051e-01 1.26805529e-01
-7.28688061e-01 7.24587560e-01 1.62728322e+00 -9.78592709e-02
-6.11117125e-01 2.52998054e-01 -4.22219098e-01 1.04289293e+00
4.88607824e-01 -1.61634222e-01 -9.97231305e-01 -6.20980382e-01
1.08817093e-01 6.15444899e-01 6.60919428e-01 -9.37669337e-01
2.55932629e-01 -4.82070744e-01 3.53406310e-01 -8.38138819e-01
7.41359115e-01 -8.94026041e-01 9.36366022e-02 5.71833789e-01
-5.80464602e-01 2.27898940e-01 -1.47109300e-01 8.85498583e-01
8.91085640e-02 -1.29632838e-03 5.09253144e-01 -5.65313876e-01
-1.31966078e+00 3.41863602e-01 -4.36466873e-01 7.07536563e-02
9.45881486e-01 -3.08814883e-01 -3.21224868e-01 -8.76901925e-01
-9.02312338e-01 7.65287876e-01 7.58513153e-01 5.97472310e-01
9.52147365e-01 -1.22654510e+00 -2.36024573e-01 3.88146162e-01
1.53431967e-01 3.38363200e-01 5.68005860e-01 5.34791291e-01
-4.52285111e-01 5.50818622e-01 -1.67139441e-01 -5.30705631e-01
-1.02962446e+00 6.04905128e-01 4.06882584e-01 -4.15954858e-01
-8.28819215e-01 7.76200056e-01 8.99333894e-01 -9.93320525e-01
2.00551882e-01 -4.04648125e-01 -3.85641783e-01 -3.78422350e-01
4.48179662e-01 2.98785806e-01 -4.60366309e-01 -6.75517201e-01
-8.32234174e-02 4.57007408e-01 -7.73442835e-02 -6.20122910e-01
1.27690089e+00 -5.03115177e-01 2.51195043e-01 3.82153124e-01
8.13471675e-01 -1.21192694e-01 -2.13916135e+00 -9.03511569e-02
-2.87202060e-01 -4.01650429e-01 -1.32472262e-01 -1.18153965e+00
-7.41607726e-01 8.42469811e-01 5.80171585e-01 -3.54563117e-01
8.11934352e-01 2.60001980e-02 8.12718689e-01 6.95599973e-01
7.56344438e-01 -9.20149982e-01 3.02764952e-01 6.75783873e-01
9.66074884e-01 -1.40103745e+00 -3.41766208e-01 -1.16718806e-01
-1.18766272e+00 9.30194020e-01 1.09314907e+00 2.20139578e-01
2.76214242e-01 -2.21630231e-01 7.18555525e-02 -2.48503417e-01
-1.03369367e+00 -4.84890565e-02 1.48110986e-01 8.82754803e-01
-1.52383983e-01 2.45385051e-01 6.36379540e-01 1.64778009e-01
-5.98151743e-01 -2.72793114e-01 6.33553445e-01 7.55037308e-01
-6.84028268e-01 -8.04645658e-01 -1.71444848e-01 2.45787278e-01
3.75603378e-01 7.78756440e-02 -2.78921664e-01 7.70868957e-01
1.36150032e-01 8.64298880e-01 -6.95100660e-03 -3.21545303e-01
4.72402364e-01 4.14011674e-03 4.61967558e-01 -6.17781043e-01
1.31321386e-01 -2.06449568e-01 2.30923876e-01 -1.05446744e+00
-1.82675287e-01 -7.55327582e-01 -1.73021078e+00 -2.00817585e-01
2.55057752e-01 -1.65002763e-01 4.16017145e-01 1.10665178e+00
3.75902563e-01 6.55572295e-01 4.41887438e-01 -1.08141756e+00
-5.05730629e-01 -6.29457653e-01 -2.82611651e-03 3.30380350e-01
5.63710034e-01 -5.77324808e-01 -2.48880804e-01 2.14639395e-01] | [4.488414287567139, 0.5935385227203369] |
1536c4f8-1448-4d6a-95a4-211109271d1e | on-classification-thresholds-for-graph | 2210.10014 | null | https://arxiv.org/abs/2210.10014v1 | https://arxiv.org/pdf/2210.10014v1.pdf | On Classification Thresholds for Graph Attention with Edge Features | The recent years we have seen the rise of graph neural networks for prediction tasks on graphs. One of the dominant architectures is graph attention due to its ability to make predictions using weighted edge features and not only node features. In this paper we analyze, theoretically and empirically, graph attention networks and their ability of correctly labelling nodes in a classic classification task. More specifically, we study the performance of graph attention on the classic contextual stochastic block model (CSBM). In CSBM the nodes and edge features are obtained from a mixture of Gaussians and the edges from a stochastic block model. We consider a general graph attention mechanism that takes random edge features as input to determine the attention coefficients. We study two cases, in the first one, when the edge features are noisy, we prove that the majority of the attention coefficients are up to a constant uniform. This allows us to prove that graph attention with edge features is not better than simple graph convolution for achieving perfect node classification. Second, we prove that when the edge features are clean graph attention can distinguish intra- from inter-edges and this makes graph attention better than classic graph convolution. | ['Shenghao Yang', 'Anton Tsitsulin', 'Bryan Perozzi', 'Silvio Lattanzi', 'Dake He', 'Kimon Fountoulakis'] | 2022-10-18 | null | null | null | null | ['stochastic-block-model'] | ['graphs'] | [ 2.29609355e-01 6.95518136e-01 -1.39859229e-01 -1.04239278e-01
-1.27526313e-01 -3.46636653e-01 5.76416075e-01 5.47512889e-01
-9.58662331e-02 3.47078562e-01 6.25916198e-03 -5.86268246e-01
-1.80065170e-01 -1.03582740e+00 -8.75600696e-01 -8.17582607e-01
-4.49362129e-01 5.35948753e-01 2.70598233e-01 -1.57608598e-01
-9.04438086e-03 6.25398099e-01 -1.29306877e+00 6.03817217e-03
6.02361858e-01 8.82148921e-01 1.71567082e-01 1.24715889e+00
-2.06738845e-01 8.52544546e-01 -4.23794866e-01 -5.40196538e-01
1.43538669e-01 -3.93476814e-01 -1.07865143e+00 -9.76796523e-02
4.11193699e-01 2.92501450e-01 -4.83233958e-01 1.26818252e+00
1.78122059e-01 1.66792810e-01 7.72650361e-01 -1.49925387e+00
-1.12505567e+00 8.32264662e-01 -3.60797375e-01 2.94013619e-01
2.25290075e-01 -1.05183288e-01 1.57812583e+00 -3.57607067e-01
5.39109886e-01 1.25312483e+00 8.45789313e-01 4.52386767e-01
-1.35371053e+00 -2.39117935e-01 3.93476725e-01 5.52692533e-01
-1.20689321e+00 1.38680011e-01 7.23491371e-01 -4.74425852e-01
1.02108538e+00 4.28855896e-01 5.72499394e-01 8.23408484e-01
4.07098442e-01 5.83914995e-01 6.95957899e-01 -6.22390389e-01
-5.22607416e-02 -2.95313317e-02 6.03987634e-01 9.86632288e-01
3.31919104e-01 -6.48559853e-02 5.75109795e-02 1.67709906e-02
4.28727269e-01 5.96786439e-02 -4.81774360e-01 -3.38152498e-01
-8.04266930e-01 9.76147056e-01 1.01509452e+00 5.66588461e-01
-4.14949715e-01 6.29288673e-01 1.70667440e-01 4.47670937e-01
4.17387486e-01 2.83151507e-01 -2.52328783e-01 3.88285220e-01
-3.39761406e-01 -1.86234564e-01 8.87943983e-01 7.68191993e-01
9.01812315e-01 -1.04044437e-01 -3.91688734e-01 3.42751175e-01
2.18161330e-01 3.69349301e-01 2.85023659e-01 -4.38544601e-01
2.04344928e-01 5.77646792e-01 -3.46163034e-01 -1.29830170e+00
-6.20407104e-01 -6.26976311e-01 -1.13316333e+00 3.85277830e-02
6.14384472e-01 7.29929749e-03 -1.19369090e+00 1.95162332e+00
-8.82821977e-02 2.04111338e-01 -2.22668007e-01 5.95846832e-01
8.34990084e-01 4.80540216e-01 2.10299283e-01 1.16251804e-01
1.17826247e+00 -7.69442737e-01 -5.00726581e-01 -2.72739351e-01
8.30401540e-01 -2.66332805e-01 8.86885524e-01 1.08566083e-01
-8.30258608e-01 -5.54429531e-01 -9.62599039e-01 -1.06038652e-01
-7.13422060e-01 -2.38079190e-01 6.82615697e-01 6.51903868e-01
-1.44605470e+00 1.04776323e+00 -6.53436899e-01 -5.69412827e-01
3.30830812e-01 5.26142418e-01 -5.41549385e-01 -1.53459355e-01
-1.16787708e+00 9.12379861e-01 1.70655042e-01 1.67243659e-01
-5.68326056e-01 -2.75619090e-01 -1.06017208e+00 7.29711294e-01
1.79148138e-01 -5.48180521e-01 8.14174414e-01 -1.30253863e+00
-9.31530774e-01 7.45452225e-01 -8.21315274e-02 -6.30561769e-01
2.16014981e-01 3.11792642e-01 -1.42142072e-01 1.07316561e-01
-3.78672406e-02 2.87508279e-01 6.86673522e-01 -1.01464677e+00
-3.32774788e-01 -3.85523260e-01 3.00938249e-01 -1.78883076e-01
-1.36555344e-01 -3.30647349e-01 -2.23044336e-01 -4.29836452e-01
-9.94278193e-02 -1.04908013e+00 -4.14394498e-01 -2.34778225e-01
-6.64672554e-01 -2.58747131e-01 4.32542175e-01 -5.94981313e-01
9.37598348e-01 -2.15904522e+00 2.12727532e-01 5.68487465e-01
7.24604547e-01 8.35025162e-02 -2.45557606e-01 4.87549901e-01
-4.95106190e-01 6.03758574e-01 -5.63170649e-02 -3.26273441e-01
1.55328587e-01 3.08671534e-01 3.16118747e-02 6.36573851e-01
3.81877214e-01 1.23678386e+00 -8.77195656e-01 -2.36495391e-01
9.71907005e-02 6.50941670e-01 -4.47250038e-01 -1.10754464e-02
-7.46202692e-02 1.11467421e-01 -2.40068331e-01 5.96824214e-02
4.11953211e-01 -7.68716931e-01 2.78612018e-01 5.12511805e-02
5.78393936e-01 1.60790399e-01 -8.60441387e-01 9.63202834e-01
-3.90423715e-01 1.02680635e+00 1.02960140e-01 -1.27683973e+00
7.11389065e-01 1.83738098e-01 2.51255333e-01 -4.65225428e-01
2.33737692e-01 -1.87820509e-01 4.45583701e-01 -8.00142884e-02
1.25234574e-01 -7.54493698e-02 9.64590833e-02 4.01848078e-01
1.76634833e-01 3.13626260e-01 2.22206458e-01 4.99374747e-01
1.60240996e+00 -4.67300951e-01 1.84321776e-01 -3.45335245e-01
5.37120044e-01 -3.17752451e-01 1.73160955e-01 9.34239686e-01
-7.39882067e-02 5.70654511e-01 1.14376843e+00 -4.45059121e-01
-7.91238010e-01 -7.37413287e-01 3.69982034e-01 1.18360209e+00
-1.81726255e-02 -5.26653886e-01 -8.10208499e-01 -9.10583198e-01
1.32509291e-01 6.01022303e-01 -1.19522810e+00 -5.18927634e-01
-2.90622115e-01 -5.45198500e-01 8.08547959e-02 6.39298916e-01
7.99869224e-02 -1.14237511e+00 -1.56980947e-01 -1.66447490e-01
1.99498266e-01 -9.32318866e-01 -4.43004727e-01 5.16787231e-01
-4.94923830e-01 -1.34970701e+00 -5.36003351e-01 -9.37771022e-01
7.73245811e-01 1.85654595e-01 1.31187737e+00 6.13068104e-01
-1.88858941e-01 4.39446092e-01 -3.95946354e-01 -2.12978646e-01
-4.45195854e-01 3.93481493e-01 -4.97368872e-01 2.55929172e-01
2.66302049e-01 -6.81573212e-01 -2.96306998e-01 -2.26511266e-02
-6.53897464e-01 1.01189598e-01 7.02067018e-01 8.72855783e-01
2.36454830e-01 -1.34682804e-02 2.57929295e-01 -1.25544465e+00
6.11099303e-01 -6.12547040e-01 -4.72982407e-01 3.23593706e-01
-6.10742986e-01 4.56574380e-01 8.88267219e-01 -2.68374741e-01
-2.70819396e-01 -2.68348847e-02 -1.56821713e-01 -3.28350097e-01
-2.96605434e-02 5.91394663e-01 -5.68514317e-02 -3.04163873e-01
4.86468941e-01 4.12240885e-02 -9.42049399e-02 -3.01562369e-01
3.61462444e-01 3.10631096e-01 1.32482901e-01 -2.37608850e-01
5.57897687e-01 2.73606688e-01 3.42243314e-01 -8.31948459e-01
-5.90262949e-01 -1.83984041e-01 -5.73886335e-01 -2.38234669e-01
1.04081523e+00 -1.80256113e-01 -9.23796475e-01 1.37033328e-01
-1.06264627e+00 -6.39919996e-01 -2.86693633e-01 1.71703890e-01
-4.10175145e-01 4.25042242e-01 -9.41421211e-01 -8.61971140e-01
-2.34096438e-01 -1.11896086e+00 8.84717166e-01 -2.40744036e-02
3.53242606e-02 -1.36937284e+00 -2.14661360e-01 -3.32686961e-01
3.62051070e-01 2.80908555e-01 1.33647716e+00 -9.60421264e-01
-4.60910916e-01 -4.16050315e-01 -5.06335855e-01 1.72517344e-01
-9.28382277e-02 8.52066725e-02 -7.41822898e-01 -2.85606682e-01
-4.20162171e-01 3.30778211e-01 1.01405287e+00 5.21266878e-01
1.28416598e+00 -1.91578418e-01 -5.21465123e-01 6.24890983e-01
1.41377246e+00 -1.50164679e-01 5.82484365e-01 -1.10220760e-01
1.01331508e+00 4.52113360e-01 -1.25824183e-01 -1.36689529e-01
2.65896201e-01 4.83545482e-01 7.55539298e-01 -2.35220835e-01
-1.19002432e-01 -2.22435489e-01 3.98458652e-02 5.82612038e-01
-1.54559478e-01 -6.07420087e-01 -9.92260993e-01 6.40025854e-01
-1.81611764e+00 -7.22296894e-01 -5.66841424e-01 2.19623709e+00
1.13767847e-01 3.03996205e-01 6.40978068e-02 2.38322079e-01
1.09954298e+00 9.70005244e-02 -9.21974406e-02 -7.04034865e-01
1.89661589e-02 5.02201498e-01 8.71107459e-01 8.27658117e-01
-9.08843696e-01 6.30468667e-01 6.25295115e+00 5.61878383e-01
-9.54866707e-01 1.97431996e-01 8.13127160e-01 4.36141431e-01
-3.95539939e-01 -5.48810586e-02 -3.93920600e-01 4.38626558e-01
9.99783695e-01 -1.17748417e-01 4.44857895e-01 8.13299775e-01
-4.45817858e-01 1.58793002e-01 -1.19545448e+00 7.48496294e-01
-1.24890320e-02 -1.10744441e+00 -4.40809084e-03 3.36927295e-01
5.00087559e-01 8.56549889e-02 -1.21738389e-01 3.42403650e-01
7.06272066e-01 -1.43942893e+00 4.07914996e-01 4.81963813e-01
5.31656384e-01 -6.77402377e-01 1.02072144e+00 2.05350906e-01
-1.28686643e+00 -7.07947016e-02 -3.12580079e-01 -1.61584899e-01
-1.50272250e-01 4.06577617e-01 -8.26443911e-01 4.69110191e-01
3.62538397e-01 6.13367736e-01 -6.07246220e-01 1.00741029e+00
-3.07523191e-01 6.96376204e-01 -2.70005047e-01 -2.50108242e-01
4.55156028e-01 -1.31495237e-01 4.59488094e-01 1.36267090e+00
1.07839659e-01 -1.35229811e-01 1.14219047e-01 8.14200222e-01
-2.04362169e-01 5.60708903e-02 -5.93899369e-01 -2.02184767e-01
-7.37798885e-02 1.22631514e+00 -1.03020394e+00 -3.18205446e-01
-4.68181759e-01 1.03032565e+00 8.47231567e-01 3.76609981e-01
-7.60264814e-01 -5.18626153e-01 6.07516587e-01 1.73051476e-01
6.97965801e-01 -7.23377317e-02 -3.53934877e-02 -7.24360704e-01
-1.76425099e-01 -2.24927977e-01 4.78514224e-01 -7.25696564e-01
-1.35442913e+00 7.18103886e-01 -3.29735696e-01 -3.74885052e-01
-4.67305146e-02 -9.40988421e-01 -9.04035866e-01 9.40013885e-01
-1.25894022e+00 -1.13554406e+00 -3.99233252e-01 4.18746203e-01
-3.08076888e-02 3.12769771e-01 8.04786801e-01 1.52361825e-01
-5.00350654e-01 5.92442989e-01 -4.32590321e-02 5.08407652e-01
9.94886458e-02 -1.74145997e+00 9.39104617e-01 7.25605428e-01
5.26217461e-01 3.26729715e-01 7.21440196e-01 -6.60138249e-01
-1.20157182e+00 -9.70160961e-01 1.07678318e+00 -2.91450739e-01
7.21973181e-01 -6.37095273e-01 -1.08609068e+00 1.02368975e+00
8.07028636e-02 5.14417648e-01 3.17670375e-01 3.24325532e-01
-4.29894656e-01 1.37940273e-01 -9.02838469e-01 4.00283009e-01
1.05632007e+00 -4.58791047e-01 -6.55915961e-02 4.02810186e-01
7.19107151e-01 -1.01460747e-01 -7.96063840e-01 2.38531351e-01
3.35817099e-01 -9.57378507e-01 6.56829000e-01 -1.07308412e+00
2.91241199e-01 1.59722436e-02 1.95660554e-02 -1.52965319e+00
-8.05221617e-01 -5.63211501e-01 -5.29972352e-02 8.76002491e-01
6.07949197e-01 -9.07959461e-01 8.76020432e-01 3.24431896e-01
5.69298193e-02 -7.87705421e-01 -7.03641653e-01 -6.00421131e-01
7.48911407e-03 -3.87889564e-01 4.58291829e-01 7.85714328e-01
2.87707914e-02 6.02762103e-01 -1.10899568e-01 1.28640845e-01
3.43283504e-01 1.24293581e-01 4.38749880e-01 -1.44975078e+00
-5.36037803e-01 -9.13528383e-01 -1.03055429e+00 -9.16512907e-01
2.69743055e-01 -1.21086657e+00 -9.57512204e-03 -1.74945879e+00
1.47224918e-01 -4.62517828e-01 -3.92589062e-01 3.79372478e-01
-3.14719856e-01 1.93259418e-01 2.06257209e-01 -3.84126246e-01
-5.72034240e-01 1.24748804e-01 1.01748645e+00 -2.49315813e-01
8.00539628e-02 1.35921896e-01 -6.98442340e-01 4.52515543e-01
6.93840504e-01 -3.98864537e-01 -1.69590533e-01 -2.10531831e-01
4.87843752e-01 -1.25943765e-01 4.52022731e-01 -9.47640181e-01
1.98047996e-01 2.73182750e-01 1.91107899e-01 -3.52287591e-02
1.54503107e-01 -8.91591489e-01 1.72216579e-01 6.04482353e-01
-4.86417800e-01 3.25176306e-02 -2.67573707e-02 9.03297544e-01
1.10517427e-01 -2.37932459e-01 6.14946425e-01 -1.13969356e-01
-3.91266674e-01 3.70314956e-01 -3.09042752e-01 -1.89545415e-02
8.54809523e-01 2.14780904e-02 -3.19232076e-01 -7.90529370e-01
-1.25916517e+00 1.20733507e-01 3.10167074e-01 1.00823916e-01
1.96888581e-01 -1.28173292e+00 -6.76023901e-01 2.74862617e-01
-9.62782204e-02 -2.01977268e-01 5.18498197e-02 1.03050482e+00
-5.26396513e-01 3.70774925e-01 8.50541890e-03 -5.56052387e-01
-1.40003371e+00 1.02663422e+00 4.61334169e-01 -4.39931601e-01
-4.23010409e-01 9.62858915e-01 4.45334971e-01 -5.96275926e-02
2.41141260e-01 -5.65083981e-01 -1.61578864e-01 -3.01314175e-01
2.78002739e-01 1.45448148e-01 1.62933916e-01 -6.76012039e-01
-1.60081878e-01 4.45716560e-01 9.92782339e-02 4.63908046e-01
1.32468987e+00 1.48803681e-01 -9.65923890e-02 4.48272437e-01
1.22318387e+00 3.29295769e-02 -8.28957915e-01 -6.20111935e-02
-5.16150799e-03 -1.35860696e-01 -2.37298431e-04 -4.44493353e-01
-1.36424112e+00 1.15531433e+00 1.89584330e-01 1.24854720e+00
9.12687600e-01 2.99561918e-01 4.69282389e-01 1.21854506e-01
5.13682440e-02 -4.17541265e-01 -3.47979516e-01 5.09384215e-01
6.81086063e-01 -1.10490358e+00 -2.85180867e-01 -7.01149106e-01
-2.03542858e-01 1.03202105e+00 3.38580102e-01 -5.59395790e-01
8.94102812e-01 1.44692898e-01 -5.67953289e-01 -4.20658678e-01
-7.09199965e-01 -7.11730957e-01 4.94317144e-01 7.50784338e-01
3.56074482e-01 2.73743808e-01 9.44662653e-03 6.14585042e-01
-8.59144628e-02 -6.01409376e-01 3.65946084e-01 3.04876328e-01
-4.93528515e-01 -8.36833417e-01 4.35284041e-02 7.97234416e-01
-6.18274152e-01 -2.98442721e-01 -8.35413992e-01 9.12104011e-01
-1.66257411e-01 6.61246002e-01 2.17625052e-01 -5.50713539e-01
8.48389864e-02 1.21284902e-01 4.89231855e-01 -6.86105609e-01
-6.37151778e-01 -5.50489783e-01 6.62186518e-02 -4.09673542e-01
2.89368145e-02 -2.27684066e-01 -1.07189655e+00 -5.86049199e-01
-6.28164589e-01 4.07865584e-01 6.27788246e-01 8.83000195e-01
3.81709427e-01 8.01154077e-01 3.08885604e-01 -8.45478117e-01
-2.51616716e-01 -1.05511892e+00 -5.83697617e-01 5.31620622e-01
4.37710464e-01 -5.84200799e-01 -6.07527316e-01 -3.67618233e-01] | [6.893809795379639, 6.120396137237549] |
0c5c7702-bcb5-4ac4-8987-e2fab0815fa3 | if-you-want-to-go-far-go-together | null | null | https://aclanthology.org/2021.naacl-main.363 | https://aclanthology.org/2021.naacl-main.363.pdf | If You Want to Go Far Go Together: Unsupervised Joint Candidate Evidence Retrieval for Multi-hop Question Answering | Multi-hop reasoning requires aggregation and inference from multiple facts. To retrieve such facts, we propose a simple approach that retrieves and reranks set of evidence facts jointly. Our approach first generates unsupervised clusters of sentences as candidate evidence by accounting links between sentences and coverage with the given query. Then, a RoBERTa-based reranker is trained to bring the most representative evidence cluster to the top. We specifically emphasize on the importance of retrieving evidence jointly by showing several comparative analyses to other methods that retrieve and rerank evidence sentences individually. First, we introduce several attention- and embedding-based analyses, which indicate that jointly retrieving and reranking approaches can learn compositional knowledge required for multi-hop reasoning. Second, our experiments show that jointly retrieving candidate evidence leads to substantially higher evidence retrieval performance when fed to the same supervised reranker. In particular, our joint retrieval and then reranking approach achieves new state-of-the-art evidence retrieval performance on two multi-hop question answering (QA) datasets: 30.5 Recall@2 on QASC, and 67.6{\%} F1 on MultiRC. When the evidence text from our joint retrieval approach is fed to a RoBERTa-based answer selection classifier, we achieve new state-of-the-art QA performance on MultiRC and second best result on QASC. | ['Mihai Surdeanu', 'Steven Bethard', 'Vikas Yadav'] | 2021-06-01 | null | null | null | naacl-2021-4 | ['multi-hop-question-answering', 'answer-selection'] | ['knowledge-base', 'natural-language-processing'] | [ 6.26124218e-02 3.90949488e-01 -6.17852688e-01 -2.26725131e-01
-1.97881353e+00 -5.12255967e-01 8.51496100e-01 7.32300222e-01
-4.90577728e-01 8.88894200e-01 6.98441088e-01 -2.05699101e-01
-8.25686455e-01 -8.35967839e-01 -9.35775697e-01 -2.67052174e-01
1.45472195e-02 8.57798100e-01 8.63428175e-01 -4.35308278e-01
6.19848847e-01 7.97555521e-02 -1.80366564e+00 8.79985690e-01
1.07776141e+00 1.05924666e+00 -7.40272086e-03 9.13438380e-01
-3.28195095e-01 1.32485259e+00 -8.11859190e-01 -7.77208686e-01
-2.77758539e-01 -2.07053483e-01 -1.35412908e+00 -3.46288741e-01
9.87239242e-01 -2.12793514e-01 -3.33514750e-01 8.81887257e-01
3.30186039e-01 -2.59007644e-02 8.18452775e-01 -8.24012935e-01
-7.99720705e-01 1.09465063e+00 -1.75450400e-01 6.26705885e-01
8.38457406e-01 -3.26689929e-01 1.80664444e+00 -9.61677015e-01
7.94767857e-01 1.10710227e+00 1.90483108e-01 1.97974473e-01
-7.70901442e-01 -1.72728181e-01 1.30695790e-01 6.53189600e-01
-1.12259316e+00 -2.19464689e-01 6.34075463e-01 -4.50588465e-02
1.01284838e+00 5.92796564e-01 3.36205781e-01 7.11580575e-01
-2.11505294e-01 1.21910799e+00 8.76854599e-01 -6.79685473e-01
2.04031825e-01 1.08012743e-01 6.48779213e-01 9.90707576e-01
4.64916945e-01 -4.94048059e-01 -7.42846847e-01 -5.14679253e-01
-6.68895096e-02 -2.58445382e-01 -2.74018139e-01 2.11602196e-01
-1.23995829e+00 9.26838517e-01 3.63033026e-01 4.61938441e-01
-5.15683055e-01 3.43794763e-01 2.99925745e-01 4.46565509e-01
2.90444434e-01 7.18655825e-01 -4.47840244e-01 2.11967766e-01
-1.07157421e+00 3.04208070e-01 8.53918910e-01 8.80003095e-01
5.89655697e-01 -7.14496732e-01 -8.23322952e-01 1.07630754e+00
2.27902755e-01 6.57176554e-01 4.80205953e-01 -1.26980960e+00
8.74014020e-01 8.63311529e-01 5.70110865e-02 -1.09815216e+00
-1.47860333e-01 -5.98887205e-01 -2.81766474e-01 -7.44765162e-01
5.18707968e-02 2.66975284e-01 -6.14160717e-01 1.49067187e+00
2.67236769e-01 -1.53452054e-01 5.44142962e-01 7.23856986e-01
1.15043688e+00 6.67006910e-01 -1.19220495e-01 -4.94217537e-02
1.76991916e+00 -1.20470679e+00 -5.50006747e-01 -5.13553526e-03
7.22801387e-01 -6.58949614e-01 1.03633535e+00 3.54264230e-01
-1.18258584e+00 -3.54858667e-01 -1.15770888e+00 -2.22924545e-01
-4.22091305e-01 1.48892388e-01 3.73803735e-01 1.84285074e-01
-8.85737538e-01 3.54566962e-01 -1.64990425e-02 -1.87383592e-01
1.72881544e-01 7.88782164e-02 -1.24630876e-01 -6.41923904e-01
-1.67126620e+00 1.04902101e+00 4.17730778e-01 -4.46484894e-01
-8.00964355e-01 -5.98729491e-01 -6.18638337e-01 1.99737236e-01
7.74725258e-01 -9.35530424e-01 1.22093737e+00 -1.28913656e-01
-9.26074088e-01 7.25704908e-01 -3.84743333e-01 -6.44214749e-01
-4.31339666e-02 -4.08137321e-01 -8.44336629e-01 1.05251527e+00
5.65619230e-01 5.44462442e-01 5.91181159e-01 -1.41423965e+00
-9.59845066e-01 3.26111168e-02 4.23695326e-01 2.64811754e-01
-4.77622896e-01 -5.03925756e-02 -8.55429411e-01 -4.41131860e-01
1.60093591e-01 -4.65475976e-01 3.90530974e-01 -5.82310557e-01
-6.01974070e-01 -8.88456225e-01 5.09591520e-01 -3.61534864e-01
1.52552998e+00 -1.71894801e+00 1.00042775e-01 4.33802873e-01
2.20084697e-01 4.77094110e-03 -2.97641039e-01 6.64897859e-01
2.88021505e-01 2.56523520e-01 -1.73423350e-01 1.24557782e-02
1.90672472e-01 2.63052195e-01 -7.16774762e-01 -3.02663911e-02
3.15276384e-01 9.49174225e-01 -1.28348672e+00 -1.11008596e+00
-3.14645946e-01 -7.25481212e-02 -5.87958872e-01 -5.20729460e-02
-5.78326523e-01 -2.30399698e-01 -6.64345860e-01 9.08158898e-01
-2.79291291e-02 -7.29241788e-01 1.35930493e-01 -3.86575371e-01
4.72363174e-01 6.89129949e-01 -8.02880228e-01 1.37317467e+00
-3.95170003e-01 4.74620193e-01 -5.06706655e-01 -8.86793315e-01
6.41283214e-01 4.67077523e-01 3.47062558e-01 -8.10729384e-01
-2.35488445e-01 6.67374253e-01 -1.73963979e-01 -8.44592333e-01
8.46341372e-01 9.35338363e-02 -3.03216010e-01 5.51609635e-01
1.91931844e-01 -3.19335669e-01 7.74256587e-01 9.67285573e-01
1.47996426e+00 -5.33153296e-01 1.01512810e-02 -2.16687888e-01
8.46639335e-01 4.75048214e-01 -3.64901461e-02 1.12275827e+00
3.43304664e-01 3.48632038e-01 3.28733027e-01 -5.12034744e-02
-6.68516219e-01 -1.08473587e+00 -1.45486251e-01 1.18355179e+00
2.54170030e-01 -6.40756488e-01 -2.22534075e-01 -1.18502140e+00
2.88595825e-01 9.57101703e-01 -5.20530581e-01 -2.54506260e-01
-6.28255308e-01 -4.97808516e-01 7.31762052e-01 3.46258968e-01
4.99968916e-01 -8.67430508e-01 -2.27670044e-01 1.71347067e-01
-8.19147587e-01 -1.04370916e+00 -1.69610456e-01 -7.18353838e-02
-7.75071084e-01 -1.42704499e+00 -5.97857654e-01 -7.07984984e-01
4.59877104e-01 2.78019190e-01 1.50632644e+00 8.45183611e-01
6.12362660e-02 9.47101414e-01 -6.63620532e-01 -7.20865503e-02
-3.29381227e-01 2.80317545e-01 -9.85173061e-02 -3.14521164e-01
2.22306520e-01 -1.06540181e-01 -7.28342950e-01 2.68212348e-01
-1.07820153e+00 -5.44812322e-01 7.39852011e-01 8.90686154e-01
5.91748893e-01 1.82245255e-01 8.61942887e-01 -8.08694422e-01
8.76790822e-01 -7.21787035e-01 -1.67330191e-01 8.34946752e-01
-7.46962786e-01 4.52964544e-01 4.41840708e-01 -1.28477290e-01
-9.53040242e-01 -7.46204257e-01 -9.56974849e-02 7.61967599e-02
2.67085224e-01 8.45197737e-01 4.31498796e-01 4.14032578e-01
7.91288614e-01 7.43917599e-02 -6.54779375e-01 -2.07396552e-01
6.70473695e-01 7.00128734e-01 5.37447751e-01 -1.08398366e+00
6.38468683e-01 4.08563375e-01 -1.83987394e-01 -4.89138633e-01
-1.56817007e+00 -8.16138625e-01 -2.78162718e-01 -3.51922244e-01
7.06437469e-01 -8.04917872e-01 -5.71001232e-01 -4.01308417e-01
-1.14562082e+00 7.50515163e-02 -4.15866971e-01 3.96845460e-01
-2.79372275e-01 4.08950210e-01 -6.81308866e-01 -6.17303491e-01
-3.84442240e-01 -8.56662929e-01 1.28582215e+00 -1.11649491e-01
-2.31486484e-01 -7.32905626e-01 1.71903655e-01 9.90100324e-01
6.32148087e-02 -3.49132717e-01 1.26840293e+00 -1.03863657e+00
-8.23650539e-01 -4.83999461e-01 -2.57391125e-01 -8.53410736e-03
-1.96112301e-02 -1.23744287e-01 -8.41447771e-01 9.36916471e-02
-3.38088185e-01 -7.96558380e-01 1.30140018e+00 -1.06094919e-01
9.73603725e-01 -4.63092864e-01 -4.00927722e-01 -4.53258216e-01
1.27917969e+00 -1.54368937e-01 3.84717733e-01 4.64841038e-01
2.47133821e-01 8.03651035e-01 9.10030603e-01 -1.01883449e-02
7.59079635e-01 5.30157387e-01 2.01003835e-01 3.19634616e-01
-4.76830930e-01 -3.53896320e-01 2.98332810e-01 1.34315825e+00
1.34332225e-01 -3.97616923e-01 -7.93940067e-01 9.15558398e-01
-1.82884061e+00 -1.28594041e+00 -2.03794435e-01 1.78447235e+00
1.24792838e+00 2.49409840e-01 -1.63623076e-02 2.35200837e-01
5.62308788e-01 3.21316682e-02 -3.13830584e-01 3.06272674e-02
-3.58048499e-01 1.36051297e-01 1.02158681e-01 6.86351001e-01
-7.48960912e-01 8.86429191e-01 6.33826351e+00 1.24184132e+00
-4.05809313e-01 1.58497065e-01 4.18943703e-01 -1.38862133e-01
-8.65714133e-01 9.67269465e-02 -9.73790824e-01 3.46469253e-01
8.50676298e-01 -5.75005487e-02 -3.00594070e-03 5.75412512e-01
-5.53884268e-01 -4.58765864e-01 -1.00205016e+00 4.36489254e-01
4.28195745e-01 -1.73571599e+00 4.98014033e-01 -1.98538631e-01
8.95253003e-01 -2.05679983e-03 -1.54616028e-01 5.72264791e-01
6.41440868e-01 -6.68910801e-01 7.18252480e-01 7.01970577e-01
4.04967636e-01 -5.91402888e-01 1.09500504e+00 3.45947921e-01
-1.05536819e+00 -2.32166037e-01 -2.70249367e-01 4.25494820e-01
3.18093926e-01 9.19400275e-01 -8.25169206e-01 9.65123177e-01
7.25489974e-01 2.16453105e-01 -7.75138140e-01 9.09699380e-01
-6.13650739e-01 7.61901855e-01 -3.19884479e-01 -6.11531794e-01
2.74035126e-01 4.91880327e-01 5.20644605e-01 1.42879391e+00
1.61038727e-01 2.54623741e-01 -3.15161496e-02 4.66225147e-01
-5.55582583e-01 9.19163004e-02 -3.68551642e-01 8.68049189e-02
8.80534530e-01 9.43130016e-01 -5.24534225e-01 -1.08075285e+00
-7.51868784e-02 6.59795702e-01 6.80732310e-01 1.46102041e-01
-6.07264578e-01 -4.56677675e-01 -1.70052320e-01 -3.69117320e-01
4.46201175e-01 2.04318032e-01 2.05389768e-01 -1.18768263e+00
3.03762048e-01 -8.24001133e-01 1.07052851e+00 -9.41858709e-01
-1.55333328e+00 5.70137858e-01 4.20548677e-01 -1.03715515e+00
-3.29433411e-01 -3.66838604e-01 -3.84364069e-01 2.77702451e-01
-2.02607632e+00 -6.76804304e-01 5.00953943e-02 4.89175260e-01
5.85236132e-01 -2.02045143e-01 6.55024052e-01 4.10856545e-01
-2.05677629e-01 4.41534579e-01 -1.26058981e-01 -4.72944044e-02
6.92629695e-01 -1.48116720e+00 -4.03821707e-01 6.89288259e-01
6.41373396e-01 8.69418859e-01 4.11984354e-01 -6.50240839e-01
-1.39777374e+00 -7.93654025e-01 1.36322141e+00 -8.02129984e-01
8.66426349e-01 2.80883968e-01 -8.77960622e-01 3.75729263e-01
4.56942976e-01 -2.83082366e-01 6.57105148e-01 3.42763811e-01
-5.10516465e-01 -2.70482332e-01 -1.03260684e+00 7.36123860e-01
7.82861650e-01 -8.95580351e-01 -1.47253418e+00 7.01892436e-01
1.11349547e+00 -7.08920136e-03 -1.08273041e+00 5.92683852e-01
3.45416009e-01 -6.77256525e-01 1.20903480e+00 -5.82994699e-01
7.65234351e-01 -4.61870283e-01 -6.80997729e-01 -8.87596548e-01
-1.00673318e-01 -5.96839488e-02 -6.03958726e-01 1.13251150e+00
1.07240546e+00 -4.28263634e-01 4.19713110e-01 1.90597638e-01
-1.91877082e-01 -9.94204104e-01 -9.52756107e-01 -7.58632481e-01
-4.09202836e-02 -4.75199759e-01 5.76923192e-01 8.07024181e-01
3.55781198e-01 4.96649504e-01 7.38849565e-02 3.59683275e-01
5.02574444e-01 5.95981419e-01 4.18268234e-01 -1.08287752e+00
-2.64402270e-01 -3.31791222e-01 1.31505176e-01 -1.28563654e+00
9.85617340e-02 -1.21288967e+00 2.06552356e-01 -1.98292661e+00
4.68471348e-01 -3.83086562e-01 -5.18669307e-01 2.66156495e-01
-5.43462396e-01 2.08491266e-01 4.01141681e-02 4.58006322e-01
-1.31366050e+00 5.50245702e-01 1.12596476e+00 -4.97878700e-01
3.28419805e-01 -3.76148671e-01 -9.23579216e-01 5.11026025e-01
6.31319344e-01 -6.31428659e-01 -4.36782569e-01 -5.72587550e-01
8.71501327e-01 1.23748250e-01 4.91248786e-01 -9.13200140e-01
7.44306862e-01 2.10904881e-01 3.24980076e-03 -8.98370802e-01
5.32705486e-01 -5.44342041e-01 -6.53699815e-01 4.05529112e-01
-7.58896291e-01 2.09323511e-01 -6.54712468e-02 8.14622045e-01
-6.14709318e-01 -6.87082171e-01 5.86243570e-02 -1.53301165e-01
-5.65957129e-01 -1.98413774e-01 -2.86413610e-01 5.22735655e-01
2.31434509e-01 2.54483759e-01 -1.03247428e+00 -4.69416827e-01
-3.96466106e-01 5.38562655e-01 -2.98324347e-01 2.70365059e-01
8.98756921e-01 -1.45350587e+00 -8.99959743e-01 -5.54260969e-01
5.03921926e-01 -2.12040663e-01 6.91058487e-02 9.24912095e-01
-1.19294293e-01 7.65914321e-01 7.18218625e-01 -5.05142808e-01
-1.17371202e+00 4.37567413e-01 -2.35454544e-01 -6.70367539e-01
-4.48927104e-01 9.31305707e-01 -6.41930401e-01 -4.42295790e-01
1.59251437e-01 -4.15164232e-01 -4.57704842e-01 2.87516356e-01
4.01895076e-01 4.40806776e-01 3.07430178e-01 -1.46374047e-01
-2.70242304e-01 6.25436842e-01 -1.41382366e-01 -4.95037615e-01
1.12067747e+00 1.11756045e-02 -3.82952392e-01 3.76799554e-01
1.16683233e+00 5.91480553e-01 -3.84238571e-01 -4.77206379e-01
5.05429268e-01 -1.82082832e-01 -7.34297708e-02 -1.06201518e+00
-5.79794049e-01 4.15914804e-01 -3.75045463e-02 5.95997095e-01
8.43042195e-01 7.40578055e-01 8.36657763e-01 1.05189586e+00
3.40364546e-01 -1.21925318e+00 7.06461489e-01 5.68071067e-01
1.16653526e+00 -1.15076971e+00 2.41106763e-01 -4.17556971e-01
-5.29743791e-01 9.00723279e-01 3.31572771e-01 1.22366017e-02
3.52952212e-01 -3.56510669e-01 -3.56567949e-01 -1.08229470e+00
-1.03646934e+00 -4.05566901e-01 8.81077766e-01 -1.45376280e-01
2.06874758e-01 -1.48386568e-01 -3.64348084e-01 4.30798709e-01
-3.09064358e-01 -4.20820832e-01 6.81714863e-02 8.97560656e-01
-9.93136644e-01 -8.87957811e-01 -4.02811795e-01 7.87298501e-01
-4.35686082e-01 -4.41439211e-01 -4.58863199e-01 7.90739775e-01
-1.16438292e-01 1.44273508e+00 -1.76110014e-01 -2.88328677e-01
2.72056460e-01 3.32050771e-01 4.97302353e-01 -6.27014101e-01
-5.61402977e-01 -3.08562428e-01 7.53247738e-01 -3.16420317e-01
-6.96514547e-01 -4.26250786e-01 -1.25233066e+00 5.28022014e-02
-6.11009836e-01 7.07860768e-01 3.02111924e-01 1.23896551e+00
4.55448538e-01 5.72609305e-01 4.37375724e-01 1.07842781e-01
-5.27794540e-01 -1.00877142e+00 -1.41309619e-01 4.71509784e-01
3.34909767e-01 -5.82555473e-01 -4.72851336e-01 -3.02513540e-01] | [10.773762702941895, 7.869656085968018] |
86a6806b-3526-423d-a0cc-7cc1b9a062f0 | modeling-the-effects-of-multiple-myeloma-on | 1602.03214 | null | http://arxiv.org/abs/1602.03214v2 | http://arxiv.org/pdf/1602.03214v2.pdf | Modeling the Effects of Multiple Myeloma on Kidney Function | Multiple myeloma (MM), a plasma cell cancer, is associated with many health
challenges, including damage to the kidney by tubulointerstitial fibrosis. We
develop a mathematical model which captures the qualitative behavior of the
cell and protein populations involved. Specifically, we model the interaction
between cells in the proximal tubule of the kidney, free light chains, renal
fibroblasts, and myeloma cells. We analyze the model for steady-state solutions
to find a mathematically and biologically relevant stable steady-state
solution. This foundational model provides a representation of dynamics between
key populations in tubulointerstitial fibrosis that demonstrates how these
populations interact to affect patient prognosis in patients with MM and renal
impairment. | [] | 2018-07-17 | null | null | null | null | ['kidney-function'] | ['medical'] | [-4.59091179e-02 -1.30438171e-02 -2.19988391e-01 2.87889779e-01
-9.20380875e-02 -5.03511190e-01 2.68749923e-01 3.99776369e-01
2.36892458e-02 7.13114142e-01 4.69608784e-01 -1.05641596e-01
-1.25573993e-01 -1.06570005e+00 -3.64112973e-01 -9.51094270e-01
-7.58734494e-02 1.20135927e+00 -6.77462667e-02 -2.00399254e-02
-4.36021127e-02 1.02951741e+00 -4.44138408e-01 2.23853648e-01
1.00122237e+00 1.98785201e-01 6.03892319e-02 9.36592937e-01
-7.41530359e-01 1.14763808e+00 -7.50810206e-02 -8.32542777e-02
2.18451366e-01 -5.91437578e-01 -4.36119556e-01 2.67772704e-01
2.79840142e-01 8.85406211e-02 -5.24854422e-01 7.20217526e-01
4.21301365e-01 -6.89723432e-01 1.04510927e+00 -1.04587901e+00
-4.98052776e-01 4.07397419e-01 -7.27499247e-01 3.15282911e-01
1.01713762e-01 3.25078458e-01 -3.90005074e-02 -9.17054653e-01
9.31620598e-01 1.41916156e+00 7.85501897e-01 7.55825937e-01
-1.78551054e+00 -4.09884274e-01 -2.99100220e-01 -1.43474609e-01
-1.09000862e+00 -1.56860575e-01 -1.39280528e-01 -9.34959233e-01
6.63611472e-01 4.27995443e-01 1.06949675e+00 3.39096785e-01
9.20395732e-01 4.67898279e-01 1.16282773e+00 -6.07247293e-01
2.68532753e-01 -1.81871489e-01 8.99617374e-01 3.44472408e-01
5.58774769e-01 7.14828372e-02 -3.46884817e-01 -8.61226439e-01
1.26036942e+00 4.56157088e-01 -4.40503359e-01 -6.97345316e-01
-7.81980157e-01 6.52945518e-01 -8.25995803e-02 3.65857869e-01
-5.88444173e-01 6.74033403e-01 -1.42135769e-01 3.57187957e-01
-5.61385512e-01 -2.89540708e-01 -2.99794860e-02 2.64636159e-01
-2.04969451e-01 4.84752953e-01 1.04791331e+00 2.32266918e-01
4.80075598e-01 -6.27050042e-01 -2.27597535e-01 8.12171280e-01
8.67880881e-01 1.03314245e+00 -1.47557436e-02 -8.39454830e-01
-1.37202621e-01 1.08454680e+00 5.78774437e-02 -4.75271456e-02
-3.13849241e-01 -3.71652484e-01 -9.67840016e-01 4.44492012e-01
9.80611980e-01 5.96649200e-02 -8.22259486e-01 1.44162142e+00
3.64604235e-01 1.57978386e-01 2.28135467e-01 7.51115084e-01
1.78834677e-01 1.94414631e-01 5.50049305e-01 -6.29723787e-01
1.43794525e+00 -4.71490562e-01 -6.31956995e-01 -3.86609547e-02
6.54688716e-01 -6.84453607e-01 3.41049045e-01 -3.71701151e-01
-1.40166676e+00 5.54934859e-01 -5.08276045e-01 2.03625217e-01
2.75118649e-01 -6.03394866e-01 4.04677123e-01 4.79237348e-01
-9.13652718e-01 6.42668426e-01 -1.12079680e+00 -9.38965261e-01
3.43926132e-01 3.26221347e-01 -1.25596657e-01 -4.73730057e-01
-8.14699173e-01 1.42706788e+00 -6.95027709e-01 -2.57574730e-02
3.18683833e-02 -1.37842715e+00 -2.32935712e-01 1.00147620e-01
-7.54319131e-01 -1.88004649e+00 1.18018210e+00 -1.42989337e-01
-1.11390662e+00 8.73661697e-01 -5.88356674e-01 -2.55464971e-01
7.29869962e-01 3.53458554e-01 -1.95147380e-01 4.19195592e-01
-2.75310725e-01 -3.79604697e-01 -4.34946001e-01 -1.28472185e+00
-3.66168499e-01 -1.02003229e+00 -7.52964795e-01 1.21092997e-01
4.54288810e-01 3.01829726e-01 -2.00858533e-01 -3.52125198e-01
6.16895020e-01 -9.64079022e-01 -8.91595662e-01 3.97284329e-01
5.07758232e-03 -2.12111343e-02 4.29951042e-01 -1.58824712e-01
6.67809129e-01 -1.78382862e+00 3.03517193e-01 4.07920599e-01
5.42597413e-01 -3.64477158e-01 3.44354570e-01 7.78745770e-01
7.96152577e-02 2.92733252e-01 8.57400373e-02 1.91099405e-01
-1.23098440e-01 3.38160753e-01 9.09771677e-03 8.09570312e-01
-1.43313006e-01 1.38330948e+00 -6.74093843e-01 -3.58234733e-01
1.97510123e-01 8.07889521e-01 5.04588112e-02 -2.85485387e-01
-1.49444386e-01 6.42540634e-01 -6.37649655e-01 1.09058154e+00
7.17882037e-01 -6.60302281e-01 7.24003017e-01 2.20821276e-01
5.43133244e-02 -3.00337970e-01 -8.49152029e-01 8.27184737e-01
2.33836934e-01 -1.55221522e-01 6.51838064e-01 -1.25330418e-01
3.52045566e-01 3.24865967e-01 8.64715815e-01 -6.07652903e-01
2.54235398e-02 4.95822370e-01 -2.23072320e-01 -3.67228776e-01
-6.71166480e-01 -1.13054562e+00 4.12125707e-01 5.25093734e-01
-3.49631995e-01 1.79870382e-01 3.73358220e-01 3.71009499e-01
1.61286998e+00 -7.71579504e-01 -5.74924499e-02 -6.58646762e-01
5.86326182e-01 1.68051511e-01 5.21507740e-01 5.61254144e-01
-5.24106860e-01 2.76631624e-01 7.70113349e-01 -3.85811746e-01
-1.10982168e+00 -1.91765773e+00 -7.04114318e-01 -2.29385763e-01
5.52266598e-01 1.12154204e-02 -2.62057874e-02 3.60006064e-01
1.20606756e+00 1.74150229e-01 -6.30276799e-01 4.69578803e-02
-7.76411176e-01 -1.21518302e+00 5.01613736e-01 2.23124906e-01
-1.45999134e-01 -6.00626111e-01 1.90328853e-03 3.12391967e-01
2.32853398e-01 -1.43298909e-01 -1.23796210e-01 3.30363214e-03
-1.37664175e+00 -1.65965283e+00 -9.00594056e-01 -8.02662909e-01
8.62009227e-01 -9.79323685e-02 1.13613141e+00 4.45292354e-01
-8.42020273e-01 6.16827190e-01 7.78739631e-01 5.81098609e-02
-8.12815726e-01 -7.92974889e-01 2.16682907e-02 -3.59315544e-01
4.35558110e-01 -1.10122144e+00 -6.87465489e-01 5.79318218e-02
-5.67540526e-01 1.13044433e-01 8.76883090e-01 2.27415726e-01
1.03760242e+00 -3.67506027e-01 5.60761392e-01 -1.04512250e+00
4.62998122e-01 -8.32314014e-01 -1.93191558e-01 5.48422456e-01
-7.94650137e-01 -1.39074758e-01 -5.78636490e-03 -1.06476240e-01
-9.08334970e-01 -1.17937312e-03 4.90609884e-01 -2.18517736e-01
3.77978012e-02 2.46082410e-01 1.89480424e-01 -5.31135738e-01
6.27547920e-01 5.34465492e-01 5.41303217e-01 -5.08130789e-01
2.67691851e-01 4.07841086e-01 6.68183029e-01 -6.17216170e-01
7.08470047e-01 7.95597076e-01 4.87180293e-01 -6.96896791e-01
-2.92094685e-02 -6.79163039e-01 -1.68091908e-01 -2.68227786e-01
2.94978350e-01 -8.12666953e-01 -1.23839617e+00 8.19152296e-01
-9.00940597e-01 -6.84219360e-01 -6.71139061e-01 6.34028316e-01
-9.66868162e-01 4.63632882e-01 -1.67837727e+00 -8.63497198e-01
-3.87206465e-01 -5.57039320e-01 3.79855067e-01 6.03218615e-01
-2.95832872e-01 -9.88569260e-01 1.06611335e+00 4.91940230e-01
7.29407370e-01 5.44818699e-01 1.70454800e+00 1.06498636e-01
-1.15706468e+00 3.35379511e-01 -6.85142428e-02 -7.35410452e-01
-2.75451392e-01 1.97944835e-01 -6.09560348e-02 5.01556247e-02
1.93741575e-01 2.56834656e-01 8.48909795e-01 7.92130828e-01
-2.10423872e-01 -2.79768646e-01 -1.05017972e+00 5.38569927e-01
1.96553028e+00 2.27661192e-01 8.54965627e-01 3.25698018e-01
1.13676421e-01 5.62256217e-01 -1.85572878e-01 3.23468506e-01
3.50063443e-01 -1.80202678e-01 -1.29136175e-01 -6.75932094e-02
-4.63893086e-01 3.59161317e-01 -8.63573849e-02 3.84941339e-01
1.51211902e-01 5.74926734e-02 -1.30636299e+00 7.18152523e-01
-2.03229403e+00 -8.10270071e-01 -7.64340818e-01 1.79494798e+00
1.12488282e+00 -8.06534514e-02 -2.23399214e-02 -8.36430907e-01
8.65637958e-01 -1.04667819e+00 -9.16358531e-01 1.64446086e-01
-3.14521223e-01 1.86518371e-01 7.12949634e-01 7.82623053e-01
-5.98585457e-02 -4.69786786e-02 8.15099144e+00 -9.70660821e-02
-7.61686325e-01 -1.37799397e-01 4.81134355e-01 -4.23433572e-01
-6.37149692e-01 4.61200118e-01 -5.41078687e-01 3.11113268e-01
8.35781336e-01 -1.03674757e+00 3.19764405e-01 1.47586837e-01
4.86414224e-01 -5.19879460e-01 -9.95352745e-01 5.26351929e-01
-7.39217103e-01 -1.79709792e+00 -9.04764906e-02 6.14739180e-01
6.51836038e-01 3.27284604e-01 -5.04552901e-01 9.43247750e-02
7.07558155e-01 -1.12776291e+00 6.16345182e-03 1.31373608e+00
3.67183238e-01 -3.19539934e-01 4.82758194e-01 4.37714189e-01
-6.50135458e-01 -2.83225253e-02 -1.63053527e-01 -1.58917427e-01
4.90683347e-01 1.24276268e+00 -6.88759446e-01 -3.61780673e-02
3.29018503e-01 4.26241308e-02 -2.60131627e-01 1.56051791e+00
5.32328606e-01 2.97996223e-01 -4.28425163e-01 4.18757856e-01
-9.21787202e-01 -5.16062856e-01 4.84513551e-01 7.01438844e-01
6.87037944e-04 6.26707733e-01 -6.14590757e-02 1.29908991e+00
-4.19228189e-02 4.34091240e-01 -1.85433239e-01 -2.77493782e-02
2.85380691e-01 1.15762258e+00 -5.58813095e-01 -3.46366316e-01
-4.34305906e-01 2.06219882e-01 3.71265888e-01 5.35304725e-01
-1.45159075e-02 6.00617766e-01 1.04047430e+00 7.59247184e-01
-4.23181117e-01 1.22772381e-01 -7.97926962e-01 -1.04647696e+00
-3.30579758e-01 -1.91862181e-01 4.55497205e-01 -3.42189789e-01
-1.71744406e+00 -2.15610951e-01 -6.53391898e-01 -3.87865663e-01
1.75354749e-01 -2.03993440e-01 -7.81834781e-01 1.39624476e+00
-1.18186760e+00 -1.23104775e+00 -1.93048328e-01 2.26961032e-01
-3.96665573e-01 4.46573853e-01 8.49684417e-01 4.60934192e-02
-5.72369814e-01 -1.20407594e-02 5.12241781e-01 -1.60310373e-01
5.84707320e-01 -1.50553942e+00 -1.56908602e-01 2.86314666e-01
-1.12205160e+00 1.09483695e+00 9.73799527e-01 -1.37090373e+00
-2.01563692e+00 -8.96825314e-01 1.18233585e+00 -6.45113230e-01
6.29862785e-01 3.55190933e-02 -6.19766653e-01 9.45049524e-01
-2.36183226e-01 1.03648836e-02 1.03268349e+00 -3.75999995e-02
7.25372881e-02 1.30456582e-01 -1.49103773e+00 7.95850217e-01
4.90938336e-01 -1.24401830e-01 -3.83586198e-01 7.04534709e-01
-2.29221582e-02 -6.71616495e-01 -1.65847862e+00 4.19571996e-01
7.14543819e-01 -7.70879388e-01 8.73220086e-01 -4.50073242e-01
8.03362057e-02 -6.49757326e-01 2.48159036e-01 -7.85649538e-01
-6.13274634e-01 -5.63892841e-01 -7.82461703e-01 9.10763144e-01
1.05942413e-02 -7.28088796e-01 1.21682608e+00 1.26731837e+00
3.28242689e-01 -8.92896295e-01 -9.75625038e-01 -4.99621719e-01
9.52488363e-01 5.89275599e-01 7.76667371e-02 9.69860017e-01
7.23908067e-01 1.52860880e-01 5.96623003e-01 -2.12582499e-02
1.03542757e+00 4.15001929e-01 5.87022185e-01 -1.39362371e+00
-1.32588148e-01 -7.19046354e-01 -3.05865705e-01 -2.42253408e-01
-5.00061810e-01 -8.89391363e-01 -5.50975084e-01 -2.01181507e+00
1.01019788e+00 -8.62594724e-01 -1.72549501e-01 -4.48044948e-02
4.35689092e-02 -9.70762875e-03 2.27502927e-01 8.85791183e-01
9.83234346e-02 -2.76864674e-02 1.47948325e+00 2.27836296e-02
-1.21686138e-01 2.17477195e-02 -8.20460975e-01 2.34045625e-01
4.73357499e-01 -5.19419909e-01 3.47943716e-02 4.74972129e-01
2.55273819e-01 8.14711094e-01 4.36643481e-01 -4.29028541e-01
6.23469114e-01 -8.44883800e-01 7.80278862e-01 -7.72912920e-01
-1.15552150e-01 -6.21248186e-01 1.16785920e+00 1.46186745e+00
-2.38181740e-01 -6.04693174e-01 -1.66140363e-01 6.60400689e-01
3.41185302e-01 -8.92281309e-02 1.25762260e+00 -5.72296023e-01
1.83507875e-01 1.37599826e-01 -1.13102961e+00 -2.44356468e-01
1.42820930e+00 -2.37246275e-01 -9.28029358e-01 1.20093964e-01
-1.25271189e+00 8.05514932e-01 1.11116803e+00 -4.75893617e-01
-4.74975742e-02 -1.35148621e+00 -9.68385100e-01 -2.37122193e-01
-1.29950166e-01 -1.07389838e-01 5.49152255e-01 1.19954491e+00
-1.00800169e+00 4.50327426e-01 -1.05373114e-01 -1.03031671e+00
-8.88463259e-01 4.16429609e-01 1.24875808e+00 -6.66122079e-01
-7.38414586e-01 4.06366616e-01 7.39661157e-02 -3.75145674e-01
-1.96903273e-01 -1.63475990e-01 2.42231220e-01 -5.56779861e-01
5.79017341e-01 3.44562411e-01 -3.94625008e-01 -2.30935350e-01
-3.54104489e-01 3.73363733e-01 5.00702970e-02 3.64009768e-01
1.16429174e+00 -4.97893631e-01 -1.07121980e+00 4.83800620e-01
3.37752968e-01 4.49842066e-02 -6.06964350e-01 -4.03028667e-01
-4.08563279e-02 -1.60305649e-01 -1.62328556e-01 -9.15669322e-01
-9.19200838e-01 -1.35785058e-01 6.55253470e-01 3.54922898e-02
6.53417766e-01 3.58485788e-01 9.88721848e-01 -5.17301410e-02
1.75594866e-01 -4.88435835e-01 -5.05098760e-01 1.90240547e-01
6.93332255e-01 -2.40480542e-01 9.86746475e-02 -8.31512868e-01
1.62812710e-01 1.11760986e+00 7.76933014e-01 -3.68810385e-01
9.56084549e-01 9.18734789e-01 7.79443800e-01 -4.44922358e-01
-1.39180493e+00 1.67200536e-01 -6.35358691e-01 4.62517828e-01
6.17140889e-01 8.58910680e-02 -9.87321079e-01 4.99837726e-01
3.45213771e-01 4.31922406e-01 9.13319468e-01 1.12025023e+00
-8.78195465e-01 -1.30976260e+00 -7.52856374e-01 4.99985188e-01
-5.17487109e-01 4.64505821e-01 -5.27862787e-01 6.75849736e-01
-2.79171348e-01 3.18946302e-01 2.49884948e-02 6.36128843e-01
2.66237289e-01 4.91929382e-01 8.62841845e-01 -3.27076882e-01
-6.16798639e-01 5.22992790e-01 -3.84480953e-01 -2.04800189e-01
-2.19152972e-01 -9.00619626e-01 -2.04163218e+00 -8.32701504e-01
-2.29783639e-01 -2.38941610e-01 2.03230888e-01 9.85263884e-01
7.74312690e-02 1.55882195e-01 4.24896806e-01 1.17642134e-01
-4.78836626e-01 -7.10773110e-01 -1.42130256e+00 3.61768782e-01
3.91584754e-01 -2.74144590e-01 -5.43503046e-01 4.94427606e-02] | [13.641148567199707, -2.9381842613220215] |
7405b9ff-6f14-4e74-a9e4-d504472e83e0 | improving-hyperparameter-learning-under | 2306.04201 | null | https://arxiv.org/abs/2306.04201v1 | https://arxiv.org/pdf/2306.04201v1.pdf | Improving Hyperparameter Learning under Approximate Inference in Gaussian Process Models | Approximate inference in Gaussian process (GP) models with non-conjugate likelihoods gets entangled with the learning of the model hyperparameters. We improve hyperparameter learning in GP models and focus on the interplay between variational inference (VI) and the learning target. While VI's lower bound to the marginal likelihood is a suitable objective for inferring the approximate posterior, we show that a direct approximation of the marginal likelihood as in Expectation Propagation (EP) is a better learning objective for hyperparameter optimization. We design a hybrid training procedure to bring the best of both worlds: it leverages conjugate-computation VI for inference and uses an EP-like marginal likelihood approximation for hyperparameter learning. We compare VI, EP, Laplace approximation, and our proposed training procedure and empirically demonstrate the effectiveness of our proposal across a wide range of data sets. | ['Arno Solin', 'ST John', 'Rui Li'] | 2023-06-07 | null | null | null | null | ['hyperparameter-optimization'] | ['methodology'] | [-9.14331451e-02 4.56968009e-01 -2.28940591e-01 -2.03480914e-01
-1.09446716e+00 -4.34922636e-01 1.09726155e+00 -1.03482164e-01
-3.20539057e-01 8.48313749e-01 1.80504024e-01 -4.85358864e-01
-2.39633620e-01 -8.01562905e-01 -9.14244413e-01 -8.90493274e-01
1.59749866e-01 9.28574681e-01 8.22645128e-02 2.61822402e-01
3.40741515e-01 3.73046815e-01 -9.38458443e-01 -2.59439856e-01
7.59250283e-01 8.10118973e-01 -1.33529633e-01 7.17314959e-01
-1.61348790e-01 4.46594685e-01 -2.60564476e-01 -6.22406304e-01
1.59149900e-01 -2.35115021e-01 -5.23115993e-01 -2.21393153e-01
2.49231219e-01 -2.63475120e-01 -1.09450355e-01 1.04817486e+00
2.92385548e-01 1.34147823e-01 1.31629074e+00 -1.14111483e+00
-3.77198905e-01 7.00640559e-01 -4.57739085e-01 -7.06654415e-02
-3.25977802e-02 4.09181148e-01 1.00398445e+00 -9.84092116e-01
3.86424720e-01 1.50692189e+00 1.02625167e+00 1.11902498e-01
-1.64123917e+00 -2.18209028e-01 -1.70385733e-01 -3.50920677e-01
-1.65037203e+00 -3.10459465e-01 3.22725266e-01 -4.89052504e-01
7.73695409e-01 -1.88768625e-01 5.53402901e-01 1.39728117e+00
2.76269376e-01 7.88767099e-01 8.65853488e-01 -4.70915735e-01
7.58005917e-01 4.78202403e-01 1.36826085e-02 6.49832070e-01
3.21965933e-01 2.30610877e-01 -4.04204309e-01 -8.35011363e-01
1.09467399e+00 -2.72837788e-01 -3.58281702e-01 -6.23939991e-01
-8.52360189e-01 1.09774649e+00 -6.72379658e-02 -2.12601483e-01
-4.09033209e-01 7.68271506e-01 -6.88755810e-02 -1.17113777e-01
4.68571335e-01 3.84568363e-01 -5.31296968e-01 -3.00502032e-01
-9.64127302e-01 6.47028208e-01 1.45922399e+00 9.52693760e-01
8.24629247e-01 -3.36911231e-02 -5.61400175e-01 6.39243126e-01
9.86049056e-01 8.51557851e-01 -2.59545505e-01 -1.38077593e+00
2.71674186e-01 -1.68585062e-01 6.54045582e-01 -4.00166869e-01
5.51909320e-02 -5.19043803e-01 -4.19198811e-01 1.63776621e-01
9.04445827e-01 -4.03368384e-01 -7.88333356e-01 1.86689019e+00
4.72748071e-01 4.63943392e-01 -1.97168663e-01 5.42475402e-01
-6.38718829e-02 8.43278527e-01 2.96787322e-01 -3.22410733e-01
1.13621008e+00 -6.64778113e-01 -5.59682727e-01 -8.96448866e-02
2.02134550e-01 -4.77116197e-01 9.83312488e-01 2.89548397e-01
-1.35723603e+00 -1.42498732e-01 -9.77783561e-01 -2.30602939e-02
7.04236999e-02 -1.95312187e-01 4.51569527e-01 8.94323826e-01
-1.08211064e+00 1.03143573e+00 -1.19549263e+00 -1.24251157e-01
4.52118844e-01 2.06128359e-01 2.58519441e-01 3.24381739e-01
-7.91163981e-01 1.00914490e+00 3.30949932e-01 -2.87293959e-02
-9.28171694e-01 -1.00637603e+00 -5.27375400e-01 2.01924875e-01
3.60442877e-01 -1.13852632e+00 1.52013433e+00 -1.37960047e-01
-2.07211971e+00 2.98363477e-01 -3.15386206e-01 -4.84606445e-01
8.55096161e-01 -2.83062637e-01 2.22695589e-01 4.33105007e-02
-4.04489845e-01 5.74091554e-01 1.42098963e+00 -1.30609703e+00
-3.27990592e-01 -1.30627751e-01 -2.62205601e-01 -2.98633631e-02
1.75614744e-01 -4.77159232e-01 -4.24436271e-01 -1.46319777e-01
1.27447024e-01 -9.44830239e-01 -9.85192657e-02 9.31301489e-02
-3.58365119e-01 -3.30243677e-01 2.94614077e-01 -3.88399601e-01
1.08983350e+00 -1.85383725e+00 1.24614805e-01 6.08338892e-01
1.48441210e-01 1.68714169e-02 2.28954926e-01 5.20766973e-01
5.44423938e-01 3.62518072e-01 -5.61248183e-01 -9.09158468e-01
7.57514298e-01 5.20890355e-01 -6.12471044e-01 6.06702626e-01
2.80952781e-01 1.06997979e+00 -8.35200489e-01 -6.75357282e-01
9.67998654e-02 7.66198635e-01 -7.99816132e-01 1.38873652e-01
-4.93683964e-01 3.25229257e-01 -5.09721160e-01 2.66112745e-01
7.40114570e-01 -4.49892521e-01 2.14138001e-01 -8.32525939e-02
6.85910359e-02 4.89780530e-02 -1.27593696e+00 1.28565526e+00
-2.55034804e-01 3.45193684e-01 1.67414203e-01 -5.54499805e-01
6.43083513e-01 2.69220620e-01 -4.14170288e-02 1.46699831e-01
1.08670786e-01 2.11383656e-01 -4.22658145e-01 -2.98585504e-01
4.09776241e-01 -4.60096359e-01 1.91761196e-01 4.60301369e-01
4.20110643e-01 -5.49548864e-01 -1.20499860e-02 4.28861044e-02
6.35253072e-01 8.52826774e-01 3.29900235e-01 -4.04725134e-01
1.51666641e-01 -4.40658718e-01 3.14971894e-01 1.36693954e+00
-8.52759704e-02 5.33178806e-01 9.67903912e-01 6.72880411e-02
-1.32087386e+00 -1.69833064e+00 -6.51913047e-01 8.97256792e-01
-3.64296556e-01 -2.92955637e-01 -7.52193630e-01 -4.44950223e-01
1.86697006e-01 1.19349098e+00 -4.49503452e-01 -5.02259564e-03
-3.89959186e-01 -9.91545439e-01 5.51764369e-01 1.84832767e-01
7.26745054e-02 -4.85853195e-01 -1.51783943e-01 9.24017131e-02
1.66499659e-01 -8.65966439e-01 -2.82689393e-01 1.77184209e-01
-8.78820002e-01 -5.51658213e-01 -6.15174949e-01 -2.60951612e-02
3.11650902e-01 -7.23499417e-01 1.16655815e+00 -5.29270411e-01
4.38976169e-01 6.19602263e-01 1.91511050e-01 -4.59609926e-01
-7.48901248e-01 6.94729090e-02 -3.73406895e-02 -1.30408645e-01
2.08455890e-01 -8.82008791e-01 -4.53560024e-01 -6.56418279e-02
-6.08838499e-01 -3.01220417e-01 6.64791882e-01 7.16552019e-01
5.79306901e-01 -3.97899657e-01 2.86334977e-02 -8.15044463e-01
7.26058662e-01 -6.60545945e-01 -1.01680684e+00 3.39564562e-01
-7.33401716e-01 7.07919896e-01 2.40811676e-01 -5.51929653e-01
-1.38162959e+00 -2.86052823e-01 -1.74074382e-01 -5.33339679e-01
1.18597247e-01 4.26661372e-01 7.75691271e-02 1.55787412e-02
5.96106887e-01 -8.31391364e-02 -7.02402070e-02 -4.88638222e-01
6.30075574e-01 2.13611856e-01 5.97481608e-01 -1.14402997e+00
7.46237218e-01 5.19363880e-01 4.85414982e-01 -6.75484121e-01
-8.68926466e-01 7.13066086e-02 -3.50217134e-01 2.04485163e-01
8.11165929e-01 -7.20843554e-01 -1.01503885e+00 1.33768782e-01
-1.21765471e+00 -4.34707224e-01 -6.01206958e-01 7.64849186e-01
-8.26048255e-01 4.28122640e-01 -5.94691277e-01 -1.26725960e+00
-1.31148145e-01 -1.03244150e+00 1.26896381e+00 1.98168859e-01
-2.21472070e-01 -1.54393733e+00 4.55543935e-01 -2.91216880e-01
4.34161335e-01 2.45427229e-02 9.85695481e-01 -6.92445159e-01
-8.08631003e-01 -1.73110794e-02 -2.31783241e-01 1.81901217e-01
-5.04663587e-01 3.45294505e-01 -1.15864897e+00 -3.59320715e-02
4.27950293e-01 -1.18260846e-01 8.44707727e-01 9.10799682e-01
6.15897834e-01 -3.67862463e-01 -3.63272995e-01 8.26504290e-01
1.63412571e+00 -3.40600282e-01 6.25799119e-01 -3.66915502e-02
4.64858562e-01 2.20237821e-01 2.16257423e-01 7.75301635e-01
2.63181180e-01 4.27549720e-01 5.05997315e-02 6.09767437e-01
3.32016498e-01 -7.61550784e-01 3.92603338e-01 4.85077083e-01
2.46580131e-03 -1.90590799e-01 -1.00686622e+00 1.30213752e-01
-1.93341470e+00 -8.08668852e-01 -1.79406218e-02 2.37261224e+00
1.05847514e+00 3.65117848e-01 -1.48854116e-02 -5.44523299e-01
4.00424987e-01 -6.59680292e-02 -5.41181564e-01 -2.30935812e-01
2.39927098e-01 4.47102070e-01 5.30396581e-01 1.08884430e+00
-9.29172695e-01 7.07924783e-01 7.84391689e+00 1.06848657e+00
-5.05844831e-01 3.34476769e-01 3.90592963e-01 -1.09090488e-02
-6.14763856e-01 4.71533298e-01 -1.10604262e+00 5.90920210e-01
1.13800502e+00 1.39209349e-02 5.06676733e-01 7.55410016e-01
2.42674366e-01 -2.68942177e-01 -1.39076447e+00 7.94779956e-01
-5.06946325e-01 -1.30340743e+00 1.04672909e-01 4.29683149e-01
7.75544107e-01 1.69802755e-01 2.14328378e-01 5.86657226e-01
6.67904496e-01 -1.06355464e+00 8.25763345e-01 1.06143463e+00
8.52818266e-02 -6.65949583e-01 6.64130867e-01 4.80057985e-01
-6.74404562e-01 6.16543666e-02 -4.66946989e-01 2.87655890e-02
5.17402768e-01 6.89255893e-01 -9.85691071e-01 3.50980088e-02
5.05219474e-02 1.43987581e-01 -1.60817429e-01 9.81909931e-01
-4.96330589e-01 8.72640848e-01 -8.70655835e-01 -1.65792882e-01
2.83867836e-01 -8.34984601e-01 8.97282600e-01 1.23588407e+00
3.11036855e-01 -1.41293958e-01 -1.39358528e-02 1.62049508e+00
1.74734086e-01 -2.14647189e-01 -2.32198611e-01 -1.83600858e-01
7.60324061e-01 1.02375531e+00 -4.16487604e-01 -3.12639058e-01
-9.65514407e-02 4.49065059e-01 4.32023704e-01 8.36798966e-01
-1.01420510e+00 7.99208283e-02 6.93362236e-01 -2.03218356e-01
6.98553503e-01 -4.35066819e-01 -2.83548713e-01 -9.63648617e-01
-1.37005314e-01 -4.76595670e-01 7.64783770e-02 -7.60182023e-01
-1.43057990e+00 -1.75881878e-01 5.36680579e-01 -7.83585608e-01
-8.30955982e-01 -7.29647815e-01 -6.80974364e-01 1.08969855e+00
-1.35137451e+00 -1.07340312e+00 4.35227185e-01 9.72591564e-02
-7.05473348e-02 3.04018825e-01 6.02679491e-01 -3.54952097e-01
-3.92117798e-01 3.75545651e-01 4.39815670e-01 -4.59774941e-01
3.93726110e-01 -1.55108953e+00 4.16313946e-01 5.02300441e-01
3.90322469e-02 8.81122589e-01 1.20849979e+00 -5.57705998e-01
-1.56725049e+00 -5.48518956e-01 5.47278523e-01 -8.39635968e-01
9.46255267e-01 -2.38553345e-01 -8.81688654e-01 9.28691566e-01
-4.94857319e-03 -2.97842562e-01 3.95922452e-01 3.13891023e-01
-3.22647184e-01 2.63698161e-01 -1.17582858e+00 7.33627379e-01
5.88155866e-01 -6.49030685e-01 -5.53503036e-01 2.32762456e-01
7.61751294e-01 -3.24165344e-01 -1.21735966e+00 8.30996856e-02
6.95226550e-01 -7.38643050e-01 1.09822237e+00 -3.57533723e-01
1.28637373e-01 -2.13113800e-01 -3.42728704e-01 -1.19124281e+00
-1.95104957e-01 -1.04800010e+00 -6.08871579e-01 1.20530307e+00
4.22936171e-01 -7.73735702e-01 7.53588080e-01 9.44046497e-01
1.14716180e-01 -7.86800921e-01 -8.30358446e-01 -5.83058000e-01
4.90099847e-01 -8.54282975e-01 4.50626850e-01 2.98988223e-01
-7.21517578e-02 4.55051392e-01 -2.23236457e-01 4.12044138e-01
1.24383223e+00 -2.52797425e-01 7.70427704e-01 -1.19935155e+00
-8.84707510e-01 -5.22443235e-01 2.34530251e-02 -1.29220164e+00
2.26649940e-01 -8.12804163e-01 5.39433695e-02 -1.21600246e+00
2.91514218e-01 -1.78579405e-01 2.13693947e-01 -2.32934151e-02
-2.37257853e-01 -2.75799304e-01 5.46036242e-03 2.39824623e-01
-3.86344850e-01 6.68606877e-01 1.09116602e+00 2.53017366e-01
-2.29537278e-01 2.15964630e-01 -4.18050975e-01 9.05709803e-01
4.79025781e-01 -5.77798784e-01 -4.59612101e-01 -1.10030383e-01
6.53901994e-01 9.96607915e-02 5.16981542e-01 -6.32299006e-01
4.13229793e-01 -1.38712242e-01 3.47472697e-01 -5.73697209e-01
5.66199839e-01 -3.55273187e-01 2.20777825e-01 1.25738546e-01
-3.19728136e-01 -6.22279525e-01 -5.94643354e-02 8.33451331e-01
3.29601616e-01 -7.40873396e-01 8.82083654e-01 -1.31258085e-01
-3.70173901e-02 2.92915940e-01 -2.98676312e-01 3.81793141e-01
2.06636369e-01 1.18945412e-01 -1.23044908e-01 -5.39515257e-01
-8.12800944e-01 1.90794498e-01 4.11511630e-01 -4.49063063e-01
3.09009314e-01 -1.00426340e+00 -7.08251059e-01 5.28159216e-02
-3.19596082e-01 -8.34448040e-02 -2.28084862e-01 1.06298220e+00
-3.59223932e-01 2.79236436e-01 4.80667293e-01 -8.76292229e-01
-3.98054391e-01 1.78461045e-01 5.13873160e-01 -5.02618909e-01
-3.28386635e-01 7.87414730e-01 7.15512503e-03 -5.58750570e-01
3.07296157e-01 -4.65604961e-01 3.81053358e-01 -1.95302561e-01
3.82620245e-01 6.62717342e-01 -5.28496444e-01 -6.75287172e-02
1.36187255e-01 3.96958262e-01 1.37312680e-01 -8.26242208e-01
1.01717794e+00 -3.88790667e-01 -1.14374481e-01 8.23568940e-01
1.17175257e+00 -7.96179920e-02 -1.91764700e+00 -2.41801158e-01
5.50197400e-02 -1.60574719e-01 1.71549752e-01 -3.49836469e-01
-2.21706644e-01 8.75396907e-01 4.35417771e-01 1.64136454e-01
1.48079902e-01 9.12167877e-02 3.23050320e-01 6.10149503e-01
2.49751776e-01 -1.05014086e+00 -1.33651614e-01 5.24454296e-01
4.65126991e-01 -1.01867461e+00 3.07490230e-01 -2.50060499e-01
-5.47345042e-01 9.65245426e-01 -5.61688580e-02 -1.94829792e-01
1.11546588e+00 3.26753408e-01 -3.76303673e-01 -1.71740845e-01
-8.11274111e-01 -4.20658477e-02 2.38007277e-01 6.26792490e-01
3.50807942e-02 -3.82737592e-02 1.00716287e-02 3.34001690e-01
-2.87446707e-01 9.33246017e-02 1.20339558e-01 6.87973320e-01
-5.78336596e-01 -1.03069520e+00 -3.01111311e-01 2.56886512e-01
-5.54002523e-01 -2.60748208e-01 1.65778011e-01 8.67282450e-01
-1.70313656e-01 6.95102334e-01 1.39037862e-01 3.56491894e-01
-1.94008559e-01 5.71446657e-01 8.40346992e-01 -1.90786034e-01
-1.20396771e-01 2.65116364e-01 2.37924512e-02 -3.65434736e-01
-1.04486935e-01 -8.35082293e-01 -9.28620577e-01 -5.37576437e-01
-2.42144078e-01 1.24464355e-01 8.74607682e-01 1.28542435e+00
1.02111772e-01 -1.11060150e-01 3.29643667e-01 -1.05489767e+00
-1.50259900e+00 -1.01816893e+00 -7.00660944e-01 -1.22511983e-02
2.59173125e-01 -5.06506741e-01 -7.08260655e-01 -2.22862631e-01] | [6.8966498374938965, 3.9221391677856445] |
33819a4a-8d6e-4a97-89e6-ff99918a0a99 | uncertainty-estimation-of-transformer | null | null | https://aclanthology.org/2022.acl-long.566 | https://aclanthology.org/2022.acl-long.566.pdf | Uncertainty Estimation of Transformer Predictions for Misclassification Detection | Uncertainty estimation (UE) of model predictions is a crucial step for a variety of tasks such as active learning, misclassification detection, adversarial attack detection, out-of-distribution detection, etc. Most of the works on modeling the uncertainty of deep neural networks evaluate these methods on image classification tasks. Little attention has been paid to UE in natural language processing. To fill this gap, we perform a vast empirical investigation of state-of-the-art UE methods for Transformer models on misclassification detection in named entity recognition and text classification tasks and propose two computationally efficient modifications, one of which approaches or even outperforms computationally intensive methods. | ['Leonid Zhukov', 'Manvel Avetisian', 'Mikhail Burtsev', 'Gleb Gusev', 'Alexander Panchenko', 'Maxim Panov', 'Kirill Fedyanin', 'Evgenii Tsymbalov', 'Akim Tsvigun', 'Artem Shelmanov', 'Gleb Kuzmin', 'Artem Vazhentsev'] | null | null | null | null | acl-2022-5 | ['adversarial-attack-detection', 'adversarial-attack-detection'] | ['computer-vision', 'knowledge-base'] | [ 1.40412405e-01 4.18861270e-01 2.20645647e-02 -7.42979825e-01
-8.28566790e-01 -4.36932892e-01 8.00722063e-01 4.43406880e-01
-7.40848780e-01 1.03518200e+00 -2.41076022e-01 -6.06451750e-01
-9.57100242e-02 -7.46381044e-01 -8.87054741e-01 -3.45073521e-01
-1.64848268e-02 4.81408238e-01 2.59447038e-01 4.05935198e-01
3.14661235e-01 5.40719807e-01 -1.20379722e+00 3.60002518e-01
7.04145193e-01 1.36835182e+00 -5.35893083e-01 6.25691175e-01
-4.30889606e-01 1.09316337e+00 -9.96186197e-01 -1.28579068e+00
-2.06720382e-01 2.03605071e-01 -8.23480964e-01 -2.44335309e-01
1.68200940e-01 -3.12517405e-01 -4.23583925e-01 1.45711732e+00
6.10656381e-01 3.38764399e-01 1.06356382e+00 -1.51709318e+00
-5.70081353e-01 1.17822516e+00 5.68037480e-02 5.57946801e-01
-2.12476164e-01 -3.48798811e-01 6.91202044e-01 -1.12069774e+00
4.34769064e-01 1.13151455e+00 7.45552897e-01 8.02779675e-01
-8.56460810e-01 -6.98277354e-01 2.18083143e-01 4.50204670e-01
-1.27988672e+00 -5.91079831e-01 4.59925175e-01 -3.31866533e-01
1.15800405e+00 2.14690924e-01 -1.98952749e-01 1.60182333e+00
3.33295912e-01 1.25633788e+00 7.66600132e-01 -3.91886115e-01
7.54747987e-01 5.05637467e-01 4.18595552e-01 5.38466096e-01
2.49756515e-01 1.88630864e-01 -6.15777791e-01 -5.80060124e-01
2.20641345e-01 -2.02186987e-01 -2.11375684e-01 -9.32221860e-02
-7.72726059e-01 9.29207325e-01 7.99353570e-02 8.10506120e-02
-1.27474397e-01 2.58691311e-01 7.39463389e-01 2.30557203e-01
1.10160041e+00 2.65074998e-01 -8.19033980e-01 -1.96070656e-01
-1.05895352e+00 3.51997726e-02 8.89821231e-01 9.97636318e-01
1.03447981e-01 3.02979261e-01 -2.49762326e-01 6.20413065e-01
5.14209032e-01 1.84397176e-01 5.08394361e-01 -2.65999377e-01
6.84088230e-01 2.96472788e-01 1.23737521e-01 -6.48680270e-01
-3.85451049e-01 -4.86909598e-01 -1.05088949e+00 3.08300883e-01
8.21191221e-02 -4.70014215e-01 -9.30441499e-01 1.41382301e+00
1.06786266e-01 5.32864690e-01 -4.25403491e-02 2.96602130e-01
8.94641638e-01 4.39599454e-01 4.20007616e-01 -1.90209210e-01
9.63817835e-01 -5.96880853e-01 -8.53970349e-01 -2.08490416e-01
5.28940916e-01 -5.12431860e-01 4.13125157e-01 5.04843771e-01
-8.69773686e-01 -2.94812351e-01 -1.13760233e+00 2.64869571e-01
-9.82313991e-01 2.88313150e-01 7.99362838e-01 1.11408186e+00
-3.83120000e-01 8.43515277e-01 -1.17557108e+00 3.33397448e-01
9.06409144e-01 4.26967174e-01 -2.26287052e-01 4.25079525e-01
-1.73935795e+00 1.11478639e+00 7.95803428e-01 2.97618419e-01
-1.08809364e+00 -6.54024363e-01 -9.33170557e-01 3.06308329e-01
3.35147053e-01 -4.20654148e-01 1.49516237e+00 -9.78638053e-01
-1.40923238e+00 7.94195056e-01 1.50275171e-01 -1.02951717e+00
1.07960296e+00 -4.28209990e-01 -7.65933216e-01 -6.03634000e-01
-5.92719316e-01 2.07141027e-01 1.14533460e+00 -8.59730005e-01
-4.84319448e-01 -4.95001316e-01 -1.83033794e-01 -2.25449085e-01
-2.92479873e-01 4.19397265e-01 2.18843132e-01 -8.43081415e-01
-2.03157544e-01 -5.94598532e-01 -8.66323411e-02 5.57763539e-02
-7.84307659e-01 -4.20608163e-01 8.97224784e-01 -4.86978650e-01
1.41503310e+00 -2.05576038e+00 -4.96325046e-02 1.06009744e-01
1.87433749e-01 6.44394338e-01 3.74640077e-01 2.08192885e-01
-1.46580532e-01 3.70429695e-01 -4.21361685e-01 -8.75013471e-01
6.41401470e-01 -1.85459107e-02 -9.37562585e-01 4.78858978e-01
5.88522792e-01 8.58458042e-01 -6.09900713e-01 -3.02426040e-01
1.65447474e-01 5.26637912e-01 9.63160992e-02 1.72407970e-01
-3.43614310e-01 -5.65531552e-02 -1.43839031e-01 5.56806803e-01
7.70054579e-01 -1.23689108e-01 -1.43724382e-01 5.95650710e-02
3.11023414e-01 4.25667793e-01 -1.44249272e+00 1.18889403e+00
-4.23837274e-01 9.31252003e-01 -3.47674370e-01 -1.16453302e+00
6.54051483e-01 5.53851008e-01 -1.50151193e-01 1.10885100e-02
5.47417998e-01 1.52451187e-01 -1.78859800e-01 -2.62145221e-01
5.54453075e-01 1.49298698e-01 -1.75670385e-02 3.11063588e-01
4.73348051e-01 3.77590537e-01 2.14665476e-02 1.58947319e-01
1.03470922e+00 -8.93454626e-02 5.14964044e-01 2.07000878e-02
4.98738974e-01 -3.66303772e-01 3.56640160e-01 1.06476521e+00
-6.41434193e-01 4.40292597e-01 4.19856787e-01 -3.47399384e-01
-9.05916214e-01 -1.18773842e+00 -4.49070990e-01 1.04714727e+00
-3.36979091e-01 -1.69484213e-01 -1.00948048e+00 -1.36686194e+00
-4.91362251e-02 1.04463613e+00 -7.57437468e-01 -4.26171958e-01
-9.16140005e-02 -1.07099104e+00 1.12435257e+00 8.36871147e-01
7.71081567e-01 -1.13503814e+00 -3.22477132e-01 1.38996691e-01
1.57153189e-01 -9.31771755e-01 9.84986871e-02 5.53877473e-01
-7.19125390e-01 -9.64878321e-01 -5.18151700e-01 -3.50730687e-01
5.26425362e-01 -7.27074027e-01 1.29515219e+00 -3.24478537e-01
-2.30952084e-01 3.21523994e-02 -1.40907854e-01 -1.14580870e+00
-6.20623946e-01 2.22119186e-02 3.53603959e-01 -6.97798431e-02
7.88834035e-01 -2.20870823e-02 -2.26883218e-02 2.84299374e-01
-1.18256521e+00 -5.90528250e-01 3.33223820e-01 1.03999305e+00
3.52251351e-01 3.18220973e-01 6.17682219e-01 -1.33713567e+00
9.01206672e-01 -5.42107940e-01 -7.89009213e-01 6.28337204e-01
-4.65547621e-01 2.57642925e-01 4.51316893e-01 -8.76165032e-01
-1.28920650e+00 3.15952525e-02 -5.01999378e-01 -4.43262070e-01
-4.04408813e-01 4.78031695e-01 -3.45169663e-01 5.22178933e-02
8.19590449e-01 -4.06395756e-02 -6.48305893e-01 -2.33650833e-01
2.35647187e-01 9.60866332e-01 3.53278577e-01 -2.44378388e-01
3.61658543e-01 1.99674129e-01 -2.60444641e-01 -5.66091120e-01
-1.09769237e+00 -8.22513998e-02 -5.73257685e-01 8.13309196e-03
5.89130878e-01 -5.66992402e-01 -6.18179560e-01 8.06120038e-01
-1.36379290e+00 -1.59372419e-01 -1.16096370e-01 7.01877177e-01
-4.41350430e-01 2.83514351e-01 -6.09859765e-01 -1.25413060e+00
-5.67470312e-01 -1.06891787e+00 8.48549485e-01 1.14784194e-02
-1.15590148e-01 -1.22099769e+00 4.92699407e-02 2.70680748e-02
5.91307402e-01 -8.98220688e-02 7.61307478e-01 -1.27869689e+00
-4.05241191e-01 -4.43116993e-01 -7.24770799e-02 8.23002279e-01
-4.49479014e-01 1.15032785e-01 -1.47242153e+00 2.45833583e-02
1.19674370e-01 -4.19537693e-01 1.08714294e+00 3.96399349e-01
1.50264204e+00 -1.34934902e-01 -2.58365214e-01 4.09544349e-01
1.22776628e+00 1.97388530e-01 9.42401052e-01 1.34335190e-01
1.76453307e-01 2.95718133e-01 4.64571387e-01 5.37292182e-01
-3.41433823e-01 2.23290011e-01 6.47458971e-01 4.36100423e-01
4.06221390e-01 6.22424930e-02 2.79136062e-01 2.46899635e-01
2.93780208e-01 -7.85439253e-01 -1.12161374e+00 1.93071142e-01
-1.95386434e+00 -1.06257391e+00 2.36551985e-01 2.13898373e+00
6.82763577e-01 7.74529099e-01 -5.16471624e-01 3.94714504e-01
1.00822425e+00 -7.22331554e-02 -7.59979486e-01 -4.74137902e-01
8.98618847e-02 5.85991517e-02 4.48507428e-01 2.58577108e-01
-1.75819254e+00 7.50486970e-01 6.63946915e+00 1.16669226e+00
-9.39655304e-01 2.42491961e-01 1.08810854e+00 2.85228729e-01
6.99107945e-02 -3.73804390e-01 -1.05163348e+00 5.11189699e-01
1.08836222e+00 1.80762321e-01 -1.47653282e-01 1.34128666e+00
-4.76394594e-01 5.71996123e-02 -1.28875363e+00 1.05540085e+00
2.07378134e-01 -1.31698704e+00 -1.44762769e-01 -3.56448531e-01
6.79818213e-01 4.94571552e-02 1.89692825e-01 6.74577117e-01
3.90653551e-01 -1.07187319e+00 8.44365239e-01 8.21339130e-01
4.71648782e-01 -1.06827545e+00 1.20488787e+00 6.39157891e-01
-5.64416051e-01 -8.89126658e-02 -6.52662039e-01 2.27855951e-01
3.87139246e-02 8.96405756e-01 -8.45524669e-01 1.57801673e-01
4.76824582e-01 6.51068985e-02 -4.08027112e-01 1.07384789e+00
-2.62723923e-01 8.94812286e-01 -2.69289613e-01 -3.62307906e-01
-1.19165704e-02 4.56630081e-01 6.13740802e-01 1.27493882e+00
-4.04731333e-02 -2.48813242e-01 -3.51410031e-01 1.11340427e+00
-6.72146559e-01 -2.52562046e-01 -5.47660947e-01 -3.98946434e-01
4.59095389e-01 8.97076905e-01 -6.89376950e-01 -4.38957363e-01
-1.70946553e-01 1.11884546e+00 4.94024456e-01 1.78792506e-01
-1.20070314e+00 -7.06388772e-01 4.87796426e-01 -4.84311312e-01
2.35942438e-01 -1.82184517e-01 -3.24147820e-01 -1.44272065e+00
-1.19085684e-01 -5.11983752e-01 3.82325560e-01 -4.96233702e-01
-1.61153340e+00 8.54763329e-01 -4.08519357e-02 -1.09436774e+00
-4.00315344e-01 -1.03525436e+00 -6.64145648e-01 7.02255011e-01
-1.14611745e+00 -8.62693787e-01 1.59222096e-01 4.10944581e-01
5.64238012e-01 -7.18456924e-01 1.03056121e+00 3.61307442e-01
-8.51891696e-01 9.41932380e-01 5.23326337e-01 6.23121142e-01
5.79847574e-01 -1.28128219e+00 7.49188781e-01 1.12835073e+00
4.83785987e-01 5.57342649e-01 7.02669740e-01 -7.13616788e-01
-6.02881372e-01 -1.14777720e+00 1.03923893e+00 -8.13763976e-01
6.91487551e-01 -6.43587530e-01 -8.34371865e-01 8.20081532e-01
-1.41497910e-01 2.69031346e-01 7.27795422e-01 8.27052444e-02
-4.32351708e-01 2.71421909e-01 -1.46753502e+00 3.17395627e-01
6.19374633e-01 -7.17777133e-01 -6.30354166e-01 3.42526376e-01
6.13047481e-01 -6.46724403e-01 -6.79803729e-01 4.22029048e-01
3.14697862e-01 -8.98622513e-01 8.54633570e-01 -1.26469791e+00
3.84207815e-01 2.12828815e-01 -5.26717901e-02 -1.32066226e+00
4.43764850e-02 -3.83797377e-01 -7.39940941e-01 1.48671579e+00
7.41095304e-01 -6.35434091e-01 8.76040220e-01 9.13755536e-01
-2.02989131e-01 -7.38095760e-01 -1.15739501e+00 -4.91410851e-01
8.36798102e-02 -1.02142692e+00 5.28390348e-01 9.23694611e-01
-1.98394999e-01 -4.30572256e-02 -2.70270795e-01 4.09155816e-01
6.37617350e-01 -5.58971286e-01 1.20837376e-01 -1.38666999e+00
-1.76526308e-01 -2.74829388e-01 -8.48744035e-01 -3.84465456e-01
5.71044445e-01 -4.31370556e-01 9.79865268e-02 -1.13184917e+00
-1.12895444e-01 -1.69826120e-01 -3.48241627e-01 3.49127650e-01
-1.07627541e-01 2.87140198e-02 -2.20022708e-01 -1.12365775e-01
-7.61944532e-01 5.89824498e-01 1.50565818e-01 -5.20362258e-01
3.72850955e-01 6.70665622e-01 -1.15736544e-01 9.96749163e-01
8.60843062e-01 -8.48542571e-01 -4.45061356e-01 -2.99201488e-01
6.55880868e-01 -9.32146013e-02 3.47034752e-01 -9.53528762e-01
3.12294424e-01 1.66314930e-01 6.19191051e-01 -6.92809284e-01
1.31483704e-01 -8.93567622e-01 -2.04637960e-01 2.89124846e-01
-8.45887065e-01 -8.86137560e-02 2.59643316e-01 8.62146497e-01
-2.99051344e-01 -8.79077911e-01 8.39776635e-01 -2.44242176e-01
-6.99431539e-01 4.78021652e-01 -5.49547613e-01 1.36089414e-01
9.57024097e-01 3.52800250e-01 -5.42551339e-01 -8.52415562e-02
-1.13158154e+00 -2.04103306e-01 -4.20359224e-01 5.36322713e-01
6.84719443e-01 -1.05640745e+00 -6.99519098e-01 8.55799243e-02
1.21599764e-01 -5.30631319e-02 1.62387937e-01 4.65792954e-01
-3.35826755e-01 6.72621489e-01 9.48602408e-02 -3.33039165e-01
-1.32717180e+00 5.59601128e-01 6.19185269e-01 -5.06374598e-01
6.16544709e-02 1.28575015e+00 -2.07010359e-01 -6.64912343e-01
9.85133708e-01 -2.08377630e-01 -2.06811830e-01 8.58299993e-03
8.51583898e-01 6.64386034e-01 6.02547526e-01 -3.88723910e-01
-3.01195145e-01 -5.27134538e-01 -4.30807471e-01 5.83559498e-02
8.83598924e-01 -2.95907594e-02 1.47698790e-01 7.47198999e-01
1.07971954e+00 -4.55651224e-01 -9.39712167e-01 -3.61815542e-01
3.29931945e-01 -1.23187184e-01 3.66466939e-01 -1.02345514e+00
-8.03369164e-01 1.28001380e+00 1.05258453e+00 3.53713214e-01
7.19022036e-01 -2.40118414e-01 3.01907778e-01 9.01421547e-01
3.65258932e-01 -1.29269409e+00 -1.83495253e-01 6.89837039e-01
7.41188467e-01 -1.57974792e+00 -1.16739914e-01 -1.81340575e-01
-6.90221906e-01 1.18974292e+00 4.66880947e-01 1.39680803e-01
1.26589346e+00 5.84402621e-01 -1.79967418e-01 1.41644388e-01
-7.48206854e-01 1.78689450e-01 4.30825770e-01 6.32320225e-01
4.63096023e-01 -9.91448760e-04 -1.38818681e-01 8.30661774e-01
2.00433612e-01 -5.19742258e-03 3.97958338e-01 9.54947114e-01
-2.26829723e-01 -7.87351847e-01 -1.73133433e-01 5.77296615e-01
-9.32252407e-01 -5.06721318e-01 -4.55160081e-01 4.92942989e-01
-9.45791882e-03 7.83284187e-01 1.31007925e-01 -2.74935812e-01
1.43242776e-01 5.55899382e-01 1.49825603e-01 -4.24284935e-01
-6.40133619e-01 -7.32226610e-01 3.30176443e-01 -2.35477075e-01
-2.72085905e-01 -4.40633297e-01 -1.15703535e+00 -1.56795919e-01
-9.76411581e-01 2.71156263e-02 1.04595625e+00 1.19219923e+00
3.10991973e-01 5.78926027e-01 2.64674664e-01 -5.17000914e-01
-1.22823310e+00 -1.35460746e+00 -8.43814611e-01 3.59744012e-01
1.48788810e-01 -7.83696175e-01 -7.26832032e-01 -2.33280137e-01] | [7.603456497192383, 3.7725448608398438] |
c880411f-318c-4426-b1f9-0f7918431c8a | predicting-potential-drug-targets-using | 2105.10578 | null | https://arxiv.org/abs/2105.10578v3 | https://arxiv.org/pdf/2105.10578v3.pdf | A Knowledge Graph-Enhanced Tensor Factorisation Model for Discovering Drug Targets | The drug discovery and development process is a long and expensive one, costing over 1 billion USD on average per drug and taking 10-15 years. To reduce the high levels of attrition throughout the process, there has been a growing interest in applying machine learning methodologies to various stages of drug discovery and development in the recent decade, especially at the earliest stage identification of druggable disease genes. In this paper, we have developed a new tensor factorisation model to predict potential drug targets (genes or proteins) for treating diseases. We created a three dimensional data tensor consisting of 1,048 gene targets, 860 diseases and 230,011 evidence attributes and clinical outcomes connecting them, using data extracted from the Open Targets and PharmaProjects databases. We enriched the data with gene target representations learned from a drug discovery oriented knowledge graph and applied our proposed method to predict the clinical outcomes for unseen gene target and disease pairs. We designed three evaluation strategies to measure the prediction performance and benchmarked several commonly used machine learning classifiers together with Bayesian matrix and tensor factorisation methods. The result shows that incorporating knowledge graph embeddings significantly improves the prediction accuracy and that training tensor factorisation alongside a dense neural network outperforms all other baselines. In summary, our framework combines two actively studied machine learning approaches to disease target identification, namely tensor factorisation and knowledge graph representation learning, which could be a promising avenue for further exploration in data driven drug discovery. | ['Ian Barrett', 'Stephen Bonner', 'Rowan Swiers', 'Cheng Ye'] | 2021-05-20 | null | null | null | null | ['knowledge-graph-embeddings', 'knowledge-graph-embeddings'] | ['graphs', 'methodology'] | [ 1.86108187e-01 1.10008985e-01 -7.18978107e-01 -5.93948364e-02
-3.97011369e-01 -4.60849226e-01 4.75710750e-01 5.21937907e-01
-6.52750358e-02 7.60476410e-01 4.98610377e-01 -6.13433123e-01
-9.22802627e-01 -6.65567935e-01 -4.00980741e-01 -7.82352567e-01
-4.87018526e-01 9.53419805e-01 -1.89179093e-01 9.42550004e-02
1.07429996e-01 7.03528762e-01 -7.41356969e-01 4.58792478e-01
6.79901123e-01 8.99733424e-01 -2.84666896e-01 4.68302846e-01
1.36877775e-01 5.86108088e-01 -2.85124511e-01 -6.01408005e-01
-5.52774668e-02 -1.59463156e-02 -9.61418808e-01 -1.10666849e-01
1.70292750e-01 1.67719543e-01 -5.37466884e-01 8.75376284e-01
7.46993005e-01 -3.41901422e-01 7.28102744e-01 -9.35615838e-01
-1.02887774e+00 4.47883487e-01 -3.60470414e-01 3.56767178e-01
3.51686060e-01 7.20187724e-02 1.27452457e+00 -1.08266687e+00
8.54302824e-01 1.27204251e+00 5.95815182e-01 4.05985564e-01
-1.50287628e+00 -4.25111711e-01 -4.57956642e-01 3.98443997e-01
-1.38401937e+00 -1.57269925e-01 4.27468181e-01 -9.61825907e-01
1.25587702e+00 2.45379314e-01 6.71943426e-01 1.24976873e+00
6.03317022e-01 4.06589389e-01 9.26430702e-01 -1.84031069e-01
2.02761337e-01 -2.21372582e-02 3.14364612e-01 8.95964742e-01
2.93736130e-01 4.10624176e-01 -4.64034289e-01 -1.10020208e+00
5.65458894e-01 2.94569194e-01 -2.88106680e-01 -4.15005326e-01
-1.65746009e+00 1.15212309e+00 3.31444800e-01 4.25505966e-01
-6.47319794e-01 9.99429896e-02 5.91882169e-01 2.48505175e-01
6.62813544e-01 7.42398679e-01 -8.84977579e-01 -5.96137047e-02
-4.20214713e-01 1.06105015e-01 7.02208579e-01 3.13382894e-01
2.50448376e-01 -1.39320478e-01 -2.63967782e-01 8.30644965e-01
3.74992996e-01 -1.52943302e-02 7.72849917e-01 -2.64773607e-01
7.99074546e-02 9.56728399e-01 -3.39744449e-01 -1.25306380e+00
-5.38704157e-01 -4.85591471e-01 -9.11698878e-01 -2.66097337e-01
2.85674036e-01 -9.14739668e-02 -1.03638732e+00 1.15142381e+00
6.89076424e-01 4.78103578e-01 -8.03724863e-03 5.53497851e-01
7.96735823e-01 4.08709884e-01 3.41355205e-01 -6.26832619e-02
1.63914824e+00 -4.92333084e-01 -7.25674391e-01 5.88039637e-01
1.43639290e+00 -8.73879731e-01 4.27342743e-01 5.28446615e-01
-4.11523163e-01 4.20804471e-02 -7.03768194e-01 1.10742718e-01
-6.99596405e-01 2.31539249e-01 1.30821526e+00 6.78580582e-01
-6.81030691e-01 1.06941462e+00 -8.28663111e-01 -4.16722655e-01
9.54392314e-01 7.38656044e-01 -8.14725161e-01 -5.53359866e-01
-1.19127715e+00 8.85844767e-01 4.00189131e-01 -4.22971137e-02
-1.03423750e+00 -1.30625606e+00 -5.35031974e-01 -1.53610066e-01
-4.45555672e-02 -1.02661479e+00 3.47980857e-01 -6.39269277e-02
-1.28804183e+00 5.06202400e-01 3.33768785e-01 -1.75250366e-01
-2.49209777e-01 -6.57905042e-02 -6.02007449e-01 -1.29009962e-01
3.17290053e-03 3.78202677e-01 3.62739563e-01 -1.69611350e-01
-2.12354004e-01 -8.48164797e-01 -3.48737031e-01 -6.72208890e-02
-6.92956209e-01 1.32784471e-01 -6.42451122e-02 -6.45081520e-01
-1.76447049e-01 -1.06957412e+00 -4.68457341e-01 -2.02789143e-01
-7.00420737e-01 -6.41806364e-01 7.00920939e-01 -6.30136192e-01
1.10188067e+00 -1.84771812e+00 7.10759223e-01 3.64282668e-01
7.17691958e-01 3.63211632e-01 -4.40191150e-01 5.26021957e-01
-8.33635092e-01 2.39405364e-01 3.00346613e-01 3.73909920e-01
-4.63542402e-01 1.88556150e-01 5.97077794e-02 6.89627767e-01
3.46899927e-01 1.10889828e+00 -1.07377601e+00 -1.11574568e-01
-3.71050574e-02 8.86026680e-01 -5.34800053e-01 1.79011934e-02
-1.55168906e-01 3.83197099e-01 -8.49457324e-01 1.10763025e+00
4.30051804e-01 -6.33723795e-01 5.04241705e-01 -5.69626391e-01
4.01722014e-01 3.49918038e-01 -6.55993223e-01 1.67561555e+00
1.87953621e-01 1.61782712e-01 -6.50979757e-01 -1.30178344e+00
7.59281456e-01 7.13088751e-01 1.16931725e+00 -2.52651393e-01
9.35280994e-02 1.65777147e-01 5.17642498e-01 -6.56394899e-01
-2.60070980e-01 -9.15778354e-02 2.76014924e-01 2.06994295e-01
4.28782195e-01 3.60483319e-01 -3.00793387e-02 2.02159479e-01
1.79201806e+00 -1.28824309e-01 3.38457882e-01 -1.45179480e-01
2.11041972e-01 3.24341357e-01 6.05611026e-01 -1.11674257e-01
-1.27601564e-01 -5.31078726e-02 7.88359821e-01 -9.57266092e-01
-8.76372933e-01 -6.52530015e-01 -5.98531067e-01 8.70037615e-01
-8.23763430e-01 -8.05309415e-01 -1.32895157e-01 -9.85509038e-01
3.76890451e-01 8.14616382e-02 -1.03020704e+00 -4.02138442e-01
-4.40585800e-02 -1.52283669e+00 7.58193493e-01 1.16165228e-01
-3.58894557e-01 -4.56503063e-01 6.43151581e-01 3.39389920e-01
2.98460007e-01 -6.80265367e-01 -3.57822746e-01 3.34946930e-01
-1.28292847e+00 -1.50319815e+00 -7.63369858e-01 -6.50608122e-01
5.90440929e-01 -1.02306113e-01 8.41744423e-01 -3.27079713e-01
-7.49397993e-01 4.75085946e-03 -4.15248096e-01 -4.31868196e-01
-3.11024278e-01 -1.42796934e-01 4.29971755e-01 2.51154825e-02
8.03310037e-01 -4.41394091e-01 -6.25249624e-01 2.46938974e-01
-8.23823869e-01 -3.34810406e-01 8.53697479e-01 1.10077989e+00
6.71571076e-01 5.28105572e-02 5.40883064e-01 -1.07001853e+00
9.13494349e-01 -8.57954323e-01 -3.67004693e-01 9.82898623e-02
-9.11856949e-01 2.29020357e-01 2.08207041e-01 -5.64289331e-01
-2.42341757e-01 2.66339749e-01 -9.83339846e-02 -4.24004763e-01
3.10523678e-02 1.13691950e+00 2.05865934e-01 -4.57674444e-01
9.87541735e-01 -1.66923493e-01 2.61470973e-01 -7.59202957e-01
3.95926148e-01 6.31373405e-01 -4.17154640e-01 -3.04695785e-01
4.44127291e-01 1.16020411e-01 5.00503004e-01 -6.02485299e-01
-5.41542172e-01 -5.91385067e-01 -4.79186594e-01 3.09211433e-01
7.13347673e-01 -8.27064991e-01 -9.43354905e-01 -2.15062592e-02
-1.11123681e+00 2.49327108e-01 5.20701036e-02 1.02199030e+00
-4.91884872e-02 4.52605128e-01 -8.22419882e-01 -1.66430641e-02
-4.50349331e-01 -1.20465481e+00 8.36429238e-01 -4.87823546e-01
-3.18136007e-01 -1.30062914e+00 7.75315404e-01 4.23484415e-01
2.15598851e-01 4.34087336e-01 1.54236436e+00 -9.84615564e-01
-2.06648290e-01 -3.87069494e-01 -2.58749396e-01 2.17408627e-01
5.03025234e-01 -3.88088152e-02 -5.90883136e-01 -2.78726250e-01
-4.31806326e-01 -3.33174676e-01 7.42149174e-01 4.89797384e-01
1.05401242e+00 -4.60100502e-01 -8.53717983e-01 3.64735186e-01
1.32001793e+00 2.55967587e-01 6.41378641e-01 -4.01313640e-02
1.13244677e+00 4.96363103e-01 2.52616048e-01 3.77979159e-01
8.51725861e-02 9.02997971e-01 3.58992845e-01 -1.79807618e-01
-4.43785340e-02 1.07751898e-01 2.48665020e-01 7.84674466e-01
-3.90777737e-01 3.78104337e-02 -1.04415846e+00 5.77685773e-01
-1.65963304e+00 -5.64696312e-01 -4.11912918e-01 2.07994843e+00
1.23662710e+00 -4.28902805e-01 1.24895029e-01 -5.75434975e-03
4.06418651e-01 -5.15374184e-01 -4.43128854e-01 -4.80078638e-01
5.36466762e-02 6.26124918e-01 4.82405037e-01 1.56092003e-01
-1.03517747e+00 6.49772644e-01 6.73373842e+00 9.77373779e-01
-1.04932952e+00 1.02904700e-02 7.21760094e-01 9.63653848e-02
-1.03106230e-01 -1.93503663e-01 -6.45275831e-01 1.80048764e-01
1.40319633e+00 -1.76615745e-01 1.47661820e-01 7.86541820e-01
3.09166759e-01 4.87552583e-01 -1.34717476e+00 8.07203531e-01
-1.84761167e-01 -1.90858352e+00 3.31801742e-01 6.55116737e-01
5.97443759e-01 3.90501738e-01 1.52782127e-01 -6.04537129e-02
6.55876160e-01 -1.43577850e+00 -7.08167195e-01 4.98127937e-01
7.08170593e-01 -5.10708451e-01 7.59391665e-01 -1.62698328e-01
-6.56709433e-01 1.47454292e-02 -4.85529572e-01 1.63753688e-01
-2.98166752e-01 1.20699596e+00 -1.59248340e+00 8.04012597e-01
4.92388308e-01 1.30067945e+00 -5.29189408e-01 1.22053742e+00
2.01062024e-01 7.14010119e-01 -8.65754411e-02 -8.62865802e-03
1.95203066e-01 -4.98982333e-02 3.29562634e-01 9.64516759e-01
3.09217930e-01 -1.44205973e-01 1.60802186e-01 5.40179908e-01
-1.20413385e-01 4.06212538e-01 -8.74879599e-01 -8.70178103e-01
-3.11459303e-02 1.41518533e+00 -3.27241838e-01 -2.10560769e-01
-4.02064800e-01 7.33116031e-01 1.75185308e-01 2.42009908e-01
-6.05106533e-01 -3.10984224e-01 9.75543737e-01 5.01672812e-02
1.29396990e-01 3.09962891e-02 -4.77949455e-02 -1.02631021e+00
-5.95952451e-01 -1.03000426e+00 6.85104549e-01 -2.22999513e-01
-1.70646644e+00 5.70041180e-01 -4.05408978e-01 -1.13100350e+00
7.12637380e-02 -1.15729165e+00 3.16113024e-03 8.64683747e-01
-1.35747719e+00 -9.94517028e-01 5.09073377e-01 5.89134634e-01
-5.37734777e-02 -4.58366066e-01 1.41312361e+00 8.14585686e-01
-9.61894274e-01 3.52316052e-01 4.06837106e-01 1.79554820e-01
6.92922354e-01 -1.07066870e+00 2.15244330e-02 -1.00251533e-01
2.05914155e-01 9.31256533e-01 3.33627015e-01 -8.74896944e-01
-1.79496253e+00 -1.40833545e+00 1.01784384e+00 -6.96802616e-01
1.05184627e+00 -1.24464065e-01 -8.45407069e-01 5.05460203e-01
-4.28178906e-02 3.87621999e-01 1.49851441e+00 4.53446329e-01
-4.51170385e-01 1.54519096e-01 -1.01184738e+00 2.02774972e-01
7.07809031e-01 -4.40468192e-01 -2.64297903e-01 1.26279926e+00
5.77591062e-01 4.95273210e-02 -1.84687424e+00 4.31547433e-01
3.96281332e-01 -1.87740669e-01 1.16948807e+00 -1.54254889e+00
2.65285581e-01 -2.82996774e-01 7.68792927e-02 -1.52713990e+00
-9.87183273e-01 -3.64944011e-01 -2.12255687e-01 6.71889901e-01
8.10750246e-01 -5.45882344e-01 1.15963638e+00 3.69883686e-01
2.94523910e-02 -1.40230536e+00 -1.17028475e+00 -5.77227890e-01
2.31045082e-01 -8.37850869e-02 1.63442552e-01 1.43460739e+00
3.92668873e-01 8.08841169e-01 -4.58603263e-01 3.54864746e-02
3.93566698e-01 -2.58865237e-01 3.06382835e-01 -1.49642479e+00
-4.61965919e-01 -3.30893129e-01 -1.06280160e+00 -3.24129581e-01
-1.79851830e-01 -1.44139147e+00 -8.52291405e-01 -1.43706846e+00
4.65715080e-01 -4.44451123e-01 -5.47637522e-01 8.74913037e-01
-1.38864405e-02 1.89700261e-01 -7.24538386e-01 1.92238495e-01
-6.49881512e-02 4.46075231e-01 1.36199296e+00 -4.96745229e-01
-1.16646953e-01 -2.76470244e-01 -8.51223648e-01 3.29290867e-01
4.66507286e-01 -6.76125765e-01 -3.17742854e-01 -1.15707412e-01
3.76940578e-01 -1.71618447e-01 1.88629076e-01 -3.79648209e-01
5.64425737e-02 -2.49509305e-01 5.60082018e-01 -5.46461195e-02
2.64421105e-01 -6.07939363e-01 5.05477309e-01 8.49991262e-01
-1.80925101e-01 -6.82767481e-02 1.74909696e-01 8.92006338e-01
-1.05121493e-01 3.00726563e-01 3.22902292e-01 1.67021528e-02
-2.95335293e-01 8.23023677e-01 -2.33369663e-01 -4.68866348e-01
1.14466584e+00 -8.37658271e-02 -2.76712924e-01 5.36231339e-01
-1.00963759e+00 -1.53290465e-01 -2.59768575e-01 4.12106007e-01
8.21167529e-01 -1.54768920e+00 -8.28601718e-01 -2.86671892e-02
4.85935271e-01 -6.88285768e-01 2.89061829e-03 1.21303606e+00
-3.80029798e-01 8.77600551e-01 -3.80326957e-02 -5.10664403e-01
-1.34482694e+00 8.51710796e-01 2.53079742e-01 -6.01315618e-01
-4.35180038e-01 9.40560758e-01 7.86261857e-02 -4.53255862e-01
7.12810904e-02 -1.79945886e-01 -6.10491872e-01 1.58540934e-01
4.14103776e-01 3.54480386e-01 4.18173164e-01 -5.44085681e-01
-5.78091204e-01 6.12929225e-01 -5.50433934e-01 6.06433570e-01
1.70517755e+00 7.40280926e-01 -5.25789022e-01 -2.56713498e-02
1.54323471e+00 -4.32748824e-01 -1.06519736e-01 -3.28554332e-01
3.43916774e-01 -4.39539641e-01 5.12610674e-01 -1.02514946e+00
-1.09870005e+00 6.11282110e-01 9.50273752e-01 -4.30556498e-02
6.10260844e-01 1.66821480e-01 4.53010708e-01 4.21337634e-01
3.46581824e-02 -6.68241918e-01 1.50592417e-01 3.16621602e-01
8.33923221e-01 -1.10031426e+00 2.66137302e-01 -6.87907040e-01
-3.66929144e-01 1.21006858e+00 -7.06952438e-02 -1.28567452e-02
9.69181299e-01 -1.57572329e-01 -2.84698576e-01 -1.03534520e+00
-1.03348589e+00 2.23570347e-01 6.93697274e-01 6.97221696e-01
9.94601369e-01 3.14494580e-01 -4.63006437e-01 2.33444586e-01
4.14031595e-01 3.65303487e-01 1.82204619e-01 6.12116635e-01
7.67385401e-03 -1.78378427e+00 -2.78284788e-01 9.68698919e-01
-8.69122624e-01 -4.84342247e-01 -6.96811676e-01 2.95108199e-01
1.43602759e-01 7.17280805e-01 -5.95903158e-01 -7.00984776e-01
2.41443217e-01 1.29599407e-01 4.48697448e-01 -9.45261002e-01
-4.20114726e-01 9.06708017e-02 3.10323775e-01 -6.59429967e-01
-3.63581300e-01 -4.93337631e-01 -8.43345702e-01 -3.33661675e-01
-6.61469221e-01 1.81111977e-01 6.86271310e-01 7.27158129e-01
1.15068114e+00 7.29223967e-01 5.06866336e-01 -2.72707582e-01
-4.22304273e-01 -1.00620520e+00 -6.86215818e-01 1.67012572e-01
-2.99968589e-02 -9.15945053e-01 -6.50492609e-02 1.08574843e-02] | [5.463380336761475, 5.8961029052734375] |
14f4c3f0-65d9-473e-ab21-6b3379bdaf90 | enhanced-chart-understanding-in-vision-and | 2305.18641 | null | https://arxiv.org/abs/2305.18641v1 | https://arxiv.org/pdf/2305.18641v1.pdf | Enhanced Chart Understanding in Vision and Language Task via Cross-modal Pre-training on Plot Table Pairs | Building cross-model intelligence that can understand charts and communicate the salient information hidden behind them is an appealing challenge in the vision and language(V+L) community. The capability to uncover the underlined table data of chart figures is a critical key to automatic chart understanding. We introduce ChartT5, a V+L model that learns how to interpret table information from chart images via cross-modal pre-training on plot table pairs. Specifically, we propose two novel pre-training objectives: Masked Header Prediction (MHP) and Masked Value Prediction (MVP) to facilitate the model with different skills to interpret the table information. We have conducted extensive experiments on chart question answering and chart summarization to verify the effectiveness of the proposed pre-training strategies. In particular, on the ChartQA benchmark, our ChartT5 outperforms the state-of-the-art non-pretraining methods by over 8% performance gains. | ['Shih-Fu Chang', 'Heng Ji', 'Christopher Thomas', 'Long Chen', 'Yi R. Fung', 'Mingyang Zhou'] | 2023-05-29 | null | null | null | null | ['value-prediction', 'chart-question-answering', 'chart-question-answering'] | ['computer-code', 'computer-code', 'computer-vision'] | [ 5.04632831e-01 2.49860287e-01 -3.82838845e-01 -3.43616784e-01
-1.03338695e+00 -8.03596258e-01 3.43723178e-01 5.73738575e-01
3.92944515e-01 2.64625847e-01 4.90271330e-01 -8.19127560e-01
2.81097703e-02 -5.49148381e-01 -1.15568745e+00 -1.01039402e-01
-7.42022991e-02 4.40080553e-01 -5.85062541e-02 -2.86434740e-02
3.59072685e-01 5.34873381e-02 -1.33055890e+00 1.32041728e+00
1.09291470e+00 1.35681772e+00 -1.05642565e-01 9.83092904e-01
-6.25944555e-01 1.64119756e+00 -6.95831954e-01 -6.66095853e-01
-5.77145182e-02 -3.78765672e-01 -7.30763078e-01 3.95056129e-01
7.86451280e-01 -4.28826034e-01 1.42132446e-01 8.52059484e-01
9.64433700e-02 -3.99528056e-01 6.76367342e-01 -1.38013268e+00
-1.07619667e+00 9.27405834e-01 -9.63702619e-01 3.05707425e-01
6.09999120e-01 2.75587469e-01 1.49868298e+00 -8.30517828e-01
5.00466585e-01 1.38779306e+00 4.81432676e-01 2.73999840e-01
-1.24863136e+00 -4.28728670e-01 4.31005120e-01 5.46848238e-01
-9.12157595e-01 -1.89758942e-01 9.29999709e-01 -6.02934182e-01
8.03773284e-01 6.89070821e-01 4.30402845e-01 8.59527588e-01
-8.21662135e-03 1.43729150e+00 1.05167758e+00 -4.78638858e-01
1.69327691e-01 4.57548320e-01 3.04599881e-01 1.11584544e+00
2.72821993e-01 -5.33100903e-01 -1.00099659e+00 3.06970388e-01
5.73778808e-01 -2.29905829e-01 -2.47909054e-01 -6.15910649e-01
-1.41669977e+00 5.93755066e-01 5.72696030e-01 -3.01783651e-01
-2.63501883e-01 4.98550832e-02 5.67442834e-01 3.25231463e-01
-5.99732772e-02 5.86579382e-01 -4.73000884e-01 1.14142343e-01
-4.98616546e-01 2.52565797e-02 7.22229660e-01 1.25456882e+00
6.82082534e-01 -8.53686556e-02 -7.41350591e-01 3.03791910e-01
3.45589697e-01 4.58875030e-01 -4.99646142e-02 -6.62071705e-01
1.35048497e+00 1.24452436e+00 -9.61792246e-02 -9.71408784e-01
-8.68118480e-02 -3.13231647e-01 -8.89061213e-01 -8.45250562e-02
4.75519359e-01 -1.68908760e-01 -1.07442760e+00 1.07772708e+00
9.36672464e-02 -2.91710109e-01 3.08983803e-01 6.64417744e-01
1.27208328e+00 8.47860873e-01 4.43166606e-02 6.68926015e-02
1.67315280e+00 -1.12213564e+00 -8.28683496e-01 -3.46374333e-01
7.46626675e-01 -7.58121014e-01 1.53250873e+00 1.87077597e-01
-8.19360197e-01 -6.99335158e-01 -1.14569581e+00 -4.02000755e-01
-4.55933124e-01 6.11813068e-01 6.42792165e-01 5.60335159e-01
-5.58635533e-01 -1.32514909e-01 -4.57693666e-01 -7.07846582e-02
8.14794302e-01 -5.44644445e-02 -8.65538269e-02 6.21898053e-03
-6.48142993e-01 4.18072283e-01 4.35986459e-01 7.95341283e-02
-7.65202522e-01 -1.25185275e+00 -1.01009595e+00 4.37698334e-01
8.30471098e-01 -6.11937881e-01 1.25796151e+00 -8.31131637e-01
-9.53508258e-01 8.47419977e-01 -3.82480830e-01 -5.90460122e-01
5.51457465e-01 -3.92156780e-01 -8.08445588e-02 3.32741499e-01
9.20944214e-02 8.10953140e-01 8.23250294e-01 -1.63329911e+00
-4.86257166e-01 -3.75900149e-01 2.34228194e-01 1.28236189e-01
-1.18667461e-01 -2.83214867e-01 -5.82844913e-01 -5.54218888e-01
-1.64086223e-02 -4.65270668e-01 2.47119755e-01 2.20324155e-02
-8.94720733e-01 -2.38349691e-01 8.97422850e-01 -9.95573223e-01
1.34551644e+00 -1.75917006e+00 -1.33209661e-01 8.58649388e-02
4.90405589e-01 9.21088383e-02 -2.38962043e-02 5.15479565e-01
-9.41208899e-02 2.26703823e-01 -2.36331895e-01 -1.49081843e-02
1.86565027e-01 -1.94275063e-02 -9.07443285e-01 -4.31911796e-01
6.50137722e-01 1.19855595e+00 -5.35512805e-01 -7.13934302e-01
2.76402712e-01 1.94433313e-02 -3.92100424e-01 3.26044112e-01
-8.54368865e-01 3.03495675e-01 -5.18521309e-01 1.12863195e+00
4.92570043e-01 -1.01236880e+00 2.61402816e-01 -3.22878927e-01
1.83502119e-02 -3.24309804e-02 -8.14263463e-01 1.46067989e+00
-3.07156771e-01 9.94252861e-01 -3.75430405e-01 -7.10530937e-01
9.88886774e-01 4.24159281e-02 -1.58908740e-01 -9.51202989e-01
7.52196684e-02 -1.77731767e-01 -1.96230575e-01 -7.30078816e-01
5.19760966e-01 4.93437201e-01 -9.20224115e-02 1.46046355e-01
-2.64107257e-01 7.05414414e-02 3.58535856e-01 5.00046611e-01
7.71630585e-01 -9.25561413e-02 3.15123171e-01 4.67745252e-02
9.53447759e-01 5.33534169e-01 1.40982270e-01 7.93169916e-01
-8.44379738e-02 4.41969097e-01 1.35580873e+00 -4.67218548e-01
-8.31863344e-01 -9.69550431e-01 4.36926156e-01 1.25054061e+00
1.08473256e-01 -7.05810606e-01 -7.24176884e-01 -9.74969625e-01
5.12545407e-02 1.08644235e+00 -8.91637921e-01 2.05928639e-01
-4.67205822e-01 -2.50022322e-01 3.05856526e-01 1.12382150e+00
8.35722625e-01 -9.94715214e-01 -8.12668681e-01 -4.01828319e-01
-2.27176249e-01 -1.42600131e+00 -4.40838814e-01 2.35145390e-01
-6.97097778e-01 -1.14778793e+00 -3.57008606e-01 -1.06950021e+00
9.05175745e-01 1.96726173e-01 1.38572598e+00 6.70144781e-02
-4.77281548e-02 5.06987214e-01 -3.71417224e-01 -9.42724228e-01
-4.59946454e-01 5.96986488e-02 -9.68964338e-01 1.93516225e-01
4.53698188e-01 -9.29482132e-02 -3.15964639e-01 4.28142957e-02
-7.88319588e-01 7.86036074e-01 1.06750846e+00 5.90596676e-01
6.45574808e-01 -1.81307778e-01 3.39927495e-01 -1.18108869e+00
6.20169222e-01 -8.33975747e-02 -8.82948995e-01 9.34494376e-01
-6.15306139e-01 5.05547583e-01 6.53076112e-01 -7.82414302e-02
-1.07779396e+00 3.83055955e-02 3.61863673e-01 -5.10614634e-01
-1.04239419e-01 6.39262974e-01 -5.63034892e-01 4.02934730e-01
3.55783761e-01 1.90628335e-01 -4.23861027e-01 -3.44135255e-01
6.86752737e-01 2.97153503e-01 7.50639856e-01 -3.90525401e-01
1.04300094e+00 4.20623899e-01 5.84728941e-02 -5.39429903e-01
-1.10002518e+00 -4.55946475e-01 -7.60777950e-01 -2.92339623e-01
1.34136450e+00 -9.64026153e-01 -9.79535162e-01 -3.82829346e-02
-1.24715745e+00 -1.66603193e-01 -8.92276168e-02 -1.78550228e-01
-4.53150272e-01 1.34161571e-02 -8.26906189e-02 -8.32908630e-01
-4.21329051e-01 -1.20091581e+00 1.14008009e+00 5.33272266e-01
-7.98767731e-02 -9.87741470e-01 -3.35825354e-01 1.01371169e+00
-6.02057688e-02 5.80293000e-01 1.67269719e+00 -7.17212975e-01
-1.24805522e+00 -6.26900885e-03 -8.42587888e-01 -1.82147305e-02
-5.36007918e-02 -6.89733922e-02 -1.01869190e+00 3.26031417e-01
-5.62981248e-01 -7.00163960e-01 1.07187688e+00 1.01151660e-01
1.38787186e+00 -4.56276178e-01 1.22604705e-02 5.49183905e-01
1.35958588e+00 2.95731485e-01 4.84948844e-01 4.49039280e-01
1.19712234e+00 6.35346472e-01 5.50717652e-01 3.35808724e-01
8.04404795e-01 4.10814323e-02 5.46749473e-01 -1.96950004e-01
-6.14917465e-02 -8.23964536e-01 1.40519455e-01 5.36450982e-01
1.53476954e-01 -2.80759633e-01 -1.15085173e+00 6.79664969e-01
-2.03520870e+00 -7.39469409e-01 -3.03395540e-01 1.64272153e+00
6.70727789e-01 4.46728766e-01 -1.05253004e-01 1.77912906e-01
4.62433189e-01 4.99171525e-01 -7.52452374e-01 -5.65831006e-01
-2.69724429e-02 -1.91550657e-01 4.22867149e-01 1.10908486e-01
-1.23574090e+00 8.47633839e-01 5.63012218e+00 4.75464076e-01
-8.37244809e-01 -4.84352976e-01 1.09790742e+00 4.04903054e-01
-4.45165724e-01 -1.55435195e-02 -6.14745140e-01 -1.28546715e-01
6.14017546e-01 -3.42208356e-01 9.32118297e-02 1.10666120e+00
-1.17172942e-01 -1.27758563e-01 -1.42727256e+00 1.16987789e+00
4.03184623e-01 -1.98779953e+00 7.22323835e-01 -3.56176615e-01
8.22501898e-01 -7.04448998e-01 4.66362357e-01 4.85191673e-01
1.06231749e-01 -1.21987891e+00 7.99033105e-01 4.12265033e-01
6.69484675e-01 -6.11283839e-01 6.42770767e-01 9.48620960e-02
-1.35486591e+00 -1.00118943e-01 3.70823145e-02 -1.68966781e-02
-2.27956876e-01 6.01406619e-02 -1.25018907e+00 7.39310384e-01
6.73877358e-01 7.20282614e-01 -1.31296968e+00 6.14712238e-01
-4.93073732e-01 8.13024879e-01 3.75157267e-01 -2.47143701e-01
2.70538568e-01 -5.99369826e-03 3.31617087e-01 1.01554561e+00
-1.55616999e-01 -1.04131825e-01 -1.27602341e-02 1.05604815e+00
-4.89588171e-01 1.12421773e-02 -4.09050405e-01 -3.50990206e-01
-2.74812039e-02 9.53363895e-01 -7.36030519e-01 -6.48576498e-01
-7.75200784e-01 5.61058640e-01 2.42026180e-01 3.96402597e-01
-8.59547019e-01 -3.33941460e-01 3.18349421e-01 -9.14994925e-02
8.19915116e-01 -9.02868249e-03 -1.08208346e+00 -1.12284422e+00
3.98334324e-01 -1.24138677e+00 6.94489837e-01 -1.17222738e+00
-9.30723786e-01 4.45973516e-01 1.16902299e-01 -1.21659112e+00
-1.58455566e-01 -7.92369008e-01 -5.94559550e-01 5.58737576e-01
-1.70336449e+00 -1.33810091e+00 -7.37305820e-01 4.30642009e-01
1.02759659e+00 -1.69210985e-01 4.01734680e-01 -2.57663757e-01
-6.14799261e-01 5.64467251e-01 -2.75005847e-01 6.20054126e-01
4.23550427e-01 -1.67589343e+00 5.59901357e-01 1.16256797e+00
4.68394697e-01 4.43872958e-01 7.65022755e-01 -5.69044113e-01
-1.89574993e+00 -1.32366598e+00 5.88071644e-01 -8.92847955e-01
8.53165567e-01 -5.02402306e-01 -1.09971488e+00 1.03461087e+00
6.65785611e-01 -4.10593927e-01 6.96879327e-01 5.55736274e-02
-8.63013804e-01 -3.77002835e-01 -5.20359814e-01 6.18714750e-01
6.39825046e-01 -6.03896499e-01 -9.09625471e-01 2.86809266e-01
1.01595426e+00 -5.56897998e-01 -6.22286916e-01 2.84065783e-01
5.55968106e-01 -1.16438591e+00 9.55856144e-01 -6.12668931e-01
9.33781266e-01 -3.65831554e-01 -1.45276353e-01 -9.50207591e-01
1.46072477e-01 -6.36081874e-01 -4.28472430e-01 1.31193066e+00
6.98141217e-01 -9.22918394e-02 9.15940225e-01 5.45252919e-01
-3.91952135e-02 -6.39143288e-01 -3.08678776e-01 -3.58885318e-01
-3.93600643e-01 -4.86268193e-01 7.29039669e-01 7.00642109e-01
1.79149639e-02 6.96892679e-01 -3.35832059e-01 4.19931710e-01
5.46851099e-01 5.79083025e-01 1.11794090e+00 -9.31546092e-01
2.97212205e-03 -4.06205624e-01 -5.74492961e-02 -1.26830590e+00
-7.88377225e-02 -6.77738309e-01 -1.41163900e-01 -1.91620648e+00
1.69908360e-01 3.70199353e-01 -6.13239147e-02 3.34228039e-01
-4.24914151e-01 -2.55365968e-01 6.53782606e-01 3.47501636e-02
-9.52054739e-01 4.03629214e-01 1.46647561e+00 -5.79355538e-01
-3.67376238e-01 -1.12660684e-01 -1.05934930e+00 6.53936327e-01
5.68081737e-01 -8.90104622e-02 -7.93207943e-01 -6.98054433e-01
3.09485018e-01 3.33684534e-01 3.77168149e-01 -9.45715487e-01
2.90308386e-01 -1.72596842e-01 7.67918229e-01 -1.50967550e+00
-7.91637599e-02 -7.49778688e-01 -5.76056480e-01 2.75293201e-01
-8.64133298e-01 5.35555780e-01 5.06963074e-01 7.22651005e-01
-3.40282172e-01 3.74844223e-01 2.09277287e-01 -8.87986049e-02
-1.04349482e+00 -1.50517717e-01 -6.88212365e-02 2.27202103e-01
8.54521871e-01 3.23270224e-02 -8.40979636e-01 -4.12670821e-01
-3.42895001e-01 8.05071771e-01 1.35687562e-02 5.63762248e-01
7.95616984e-01 -1.00329256e+00 -4.82197762e-01 2.90947318e-01
7.63249695e-01 2.71889418e-01 1.29949510e-01 6.71373844e-01
-9.35548663e-01 9.67697144e-01 -2.52055496e-01 -8.16456378e-01
-1.46681619e+00 8.03184032e-01 -2.21672550e-01 -7.13358819e-01
-4.80817825e-01 8.31200242e-01 4.65379030e-01 -4.40731682e-02
6.16654396e-01 -1.10767782e+00 -4.85286951e-01 1.59435108e-01
7.47777045e-01 2.26341724e-01 -3.87109108e-02 -1.80727884e-01
-3.24211299e-01 3.82870704e-01 -4.12311018e-01 7.96431080e-02
9.82740641e-01 -2.42123324e-02 1.00031450e-01 5.70064723e-01
8.93507123e-01 -4.61231135e-02 -1.30054426e+00 -9.98082086e-02
4.45956230e-01 -3.71159256e-01 -1.78307980e-01 -1.15715253e+00
-5.80384374e-01 1.05989802e+00 1.65921256e-01 2.79579163e-01
1.32412124e+00 6.34114444e-02 6.12177789e-01 7.40580738e-01
-3.45072210e-01 -7.12694705e-01 4.37615722e-01 2.53186852e-01
1.22230136e+00 -1.59701943e+00 5.62690906e-02 -6.87351644e-01
-1.41060102e+00 1.22165143e+00 1.00806093e+00 1.91852853e-01
1.44196615e-01 -1.17642641e-01 5.76732457e-01 -3.09786797e-01
-1.25582528e+00 -4.19644505e-01 9.71075892e-01 6.63980484e-01
3.23193163e-01 -3.06101460e-02 4.33522612e-01 3.40884358e-01
-3.86040717e-01 -2.55911261e-01 6.18567348e-01 8.56737792e-01
-2.38901615e-01 -5.36675096e-01 -6.86199009e-01 5.64732432e-01
8.55517760e-02 -1.57681510e-01 -8.88037264e-01 1.07947242e+00
-1.94392115e-01 1.00308478e+00 7.11190403e-02 -4.22856122e-01
4.45868939e-01 2.32441857e-01 2.10377336e-01 -3.24287295e-01
-4.79660690e-01 -7.89687037e-02 1.32053167e-01 -4.62055743e-01
-3.10299993e-01 -5.05116761e-01 -1.26346207e+00 5.31684011e-02
3.19946915e-01 -7.08908066e-02 3.69864345e-01 8.55267406e-01
3.21711123e-01 1.08035088e+00 2.99557418e-01 -7.24627753e-04
-3.06969076e-01 -6.81891978e-01 7.24244714e-02 3.37663889e-01
7.44145036e-01 -2.95617789e-01 -5.45296259e-02 6.50963902e-01] | [11.21567440032959, 2.057987689971924] |
b98e8911-b742-4b74-9f68-3527686ef0e9 | a-multi-task-learning-framework-for-carotid | 2307.00583 | null | https://arxiv.org/abs/2307.00583v1 | https://arxiv.org/pdf/2307.00583v1.pdf | A multi-task learning framework for carotid plaque segmentation and classification from ultrasound images | Carotid plaque segmentation and classification play important roles in the treatment of atherosclerosis and assessment for risk of stroke. Although deep learning methods have been used for carotid plaque segmentation and classification, most focused on a single task and ignored the relationship between the segmentation and classification of carotid plaques. Therefore, we propose a multi-task learning framework for ultrasound carotid plaque segmentation and classification, which utilizes a region-weight module (RWM) and a sample-weight module (SWM) to exploit the correlation between these two tasks. The RWM provides a plaque regional prior knowledge to the classification task, while the SWM is designed to learn the categorical sample weight for the segmentation task. A total of 1270 2D ultrasound images of carotid plaques were collected from Zhongnan Hospital (Wuhan, China) for our experiments. The results of the experiments showed that the proposed method can significantly improve the performance compared to existing networks trained for a single task, with an accuracy of 85.82% for classification and a Dice similarity coefficient of 84.92% for segmentation. In the ablation study, the results demonstrated that both the designed RWM and SWM were beneficial in improving the network's performance. Therefore, we believe that the proposed method could be useful for carotid plaque analysis in clinical trials and practice. | ['Aaron Fenster', 'Xiaoyan Wu', 'Xinyao Cheng', 'Furong Wang', 'Yanghan Ou', 'Ran Zhou', 'Haitao Gan'] | 2023-07-02 | null | null | null | null | ['classification-1', 'multi-task-learning'] | ['methodology', 'methodology'] | [-1.79298028e-01 -3.09211522e-01 -1.70744732e-01 -5.09173751e-01
-1.09788918e+00 -2.33747065e-01 5.65416589e-02 -1.63551137e-01
-1.77000776e-01 3.86047781e-01 2.02835843e-01 -7.49288797e-01
-7.28084072e-02 -7.45206475e-01 -1.65893883e-01 -1.12642598e+00
-4.01359290e-01 3.40758622e-01 6.20307922e-01 2.37550408e-01
2.74136335e-01 3.58077168e-01 -1.04179251e+00 4.19668257e-01
1.02473533e+00 1.17044270e+00 4.89437640e-01 3.24445069e-01
-1.48559630e-01 2.96923846e-01 -5.82266927e-01 -1.57116592e-01
3.94108415e-01 -3.05055350e-01 -4.27669555e-01 -8.83385092e-02
1.73879489e-02 -2.68412411e-01 -1.29069328e-01 9.63736773e-01
1.03478765e+00 -1.26901597e-01 8.76228392e-01 -6.24974549e-01
-5.56957245e-01 5.81064403e-01 -9.09662187e-01 5.63324153e-01
-6.75918758e-01 3.63499783e-02 6.00002110e-01 -7.16137886e-01
1.19750008e-01 1.19595754e+00 7.06225455e-01 2.19980776e-01
-6.20091617e-01 -1.03219938e+00 1.06908627e-01 2.78773785e-01
-1.23417068e+00 -5.99459670e-02 8.05472970e-01 -7.26795673e-01
-1.07960068e-01 5.47595099e-02 8.87315452e-01 7.04045475e-01
4.94521976e-01 6.65622532e-01 1.10418391e+00 1.83745623e-02
1.76330537e-01 -1.02081895e-02 3.93312782e-01 6.73839092e-01
6.30417407e-01 1.13085352e-01 5.08435547e-01 -2.58082718e-01
1.13846421e+00 -1.72761027e-02 2.71487553e-02 -1.66683421e-01
-8.85384440e-01 1.13362980e+00 7.59475231e-01 3.42369705e-01
-2.66186804e-01 -2.49157816e-01 6.75699949e-01 -3.86528261e-02
5.82971156e-01 1.85667612e-02 -1.83761016e-01 4.19900715e-01
-8.43934655e-01 -1.22286826e-01 4.67210919e-01 2.99413115e-01
1.90331608e-01 -2.53351368e-02 -3.53496552e-01 1.01229417e+00
7.24917114e-01 5.03189504e-01 7.92559385e-01 -9.40995991e-01
1.52026892e-01 4.67641443e-01 -1.23204388e-01 -1.20271444e+00
-6.33898199e-01 -8.39276612e-01 -1.10792565e+00 4.74036515e-01
6.41755104e-01 -5.24762452e-01 -1.20946074e+00 1.35645902e+00
2.98906446e-01 5.64823806e-01 -2.24060401e-01 1.20014620e+00
1.02934504e+00 3.49311143e-01 4.70580995e-01 -2.12120473e-01
1.90648699e+00 -7.96934724e-01 -6.19539142e-01 -2.37315401e-01
7.17793703e-01 -7.67622113e-01 9.66375649e-01 8.85765180e-02
-9.55642998e-01 -6.52893543e-01 -1.06270552e+00 2.33159527e-01
3.57152820e-02 4.32814956e-01 6.02420747e-01 1.00109982e+00
-7.15802491e-01 9.44550112e-02 -1.01002789e+00 1.83397532e-01
1.13190150e+00 5.17436885e-04 2.79372394e-01 9.86051783e-02
-1.30374777e+00 9.60043311e-01 1.09244078e-01 5.01534343e-01
-8.36081505e-01 -8.32543790e-01 -5.10446966e-01 -3.66751738e-02
5.63257898e-04 -4.63257641e-01 9.98153150e-01 -5.55711687e-01
-1.43466306e+00 7.68302917e-01 -1.15218498e-01 -2.31380135e-01
4.79326338e-01 -1.72627389e-01 -4.05584365e-01 4.19412434e-01
3.49987596e-01 2.67226726e-01 5.76573789e-01 -1.36742699e+00
-5.98484576e-01 -4.27155763e-01 -3.41547430e-01 -6.62524402e-02
-2.09359914e-01 -2.52323188e-02 -2.74890095e-01 -8.40595305e-01
3.30429107e-01 -7.32870221e-01 -5.49194932e-01 4.04003024e-01
-9.21808779e-02 -2.69735813e-01 8.90407622e-01 -1.24597585e+00
1.07173061e+00 -2.02095175e+00 -2.60410160e-01 6.49276376e-01
5.35752118e-01 8.09120834e-01 -1.88895494e-01 -5.93778253e-01
-7.78774992e-02 4.15440559e-01 -3.22369754e-01 -6.42687008e-02
-5.65819502e-01 -2.07885280e-02 1.07424572e-01 5.31235635e-01
7.75502622e-02 9.35352027e-01 -7.28722870e-01 -8.97432327e-01
9.73023698e-02 6.18815005e-01 -1.67734474e-01 1.49831865e-02
1.19262211e-01 6.66192889e-01 -1.00726044e+00 4.83626902e-01
9.05291617e-01 -4.00898516e-01 1.74113750e-01 -5.09563625e-01
-3.61770988e-02 -1.64163485e-01 -1.18822432e+00 1.40135467e+00
-3.38864863e-01 2.34503105e-01 1.64200127e-01 -1.27848852e+00
1.10461748e+00 6.99135959e-01 7.53815413e-01 -6.64565980e-01
6.15229845e-01 4.03978199e-01 5.39883912e-01 -8.68876696e-01
-6.07722819e-01 -1.37917206e-01 4.55033004e-01 6.10688269e-01
-6.21043682e-01 2.48865187e-01 1.87388718e-01 -6.94527030e-02
5.93360066e-01 -1.13070384e-01 1.86143648e-02 -5.70920825e-01
7.07539141e-01 -1.35708287e-01 7.87845135e-01 6.15982354e-01
-7.32693970e-01 5.74291229e-01 5.29170334e-01 -5.50428391e-01
-8.07033122e-01 -1.21291602e+00 -4.51202989e-01 6.39610708e-01
3.82571310e-01 9.25990790e-02 -5.72791278e-01 -6.02939665e-01
-2.01940075e-01 8.17617178e-02 -6.24225914e-01 -3.06042939e-01
-1.17457867e+00 -1.43728399e+00 4.79001850e-01 7.29839683e-01
8.58754396e-01 -7.35546470e-01 -4.62069690e-01 2.83607394e-01
-2.75552690e-01 -8.52520704e-01 -5.54620326e-01 -4.80836958e-01
-1.19436872e+00 -1.34506178e+00 -1.23448801e+00 -1.43999505e+00
4.92140532e-01 4.26514685e-01 9.34727609e-01 2.98216373e-01
-5.91406107e-01 -3.32688719e-01 -1.94199830e-01 -5.63908756e-01
-3.45829725e-01 3.14840376e-02 -5.72209716e-01 1.15602233e-01
1.91884264e-01 -6.32566690e-01 -1.32298136e+00 5.56342900e-01
-4.63780761e-01 -2.54636109e-01 1.07761848e+00 4.91426378e-01
3.99740756e-01 -1.25003383e-01 1.17374635e+00 -7.29723990e-01
7.23074973e-01 -5.14520824e-01 -3.58630091e-01 2.22495362e-01
-3.08345854e-01 -4.22412843e-01 1.86145842e-01 -5.57228208e-01
-1.26983905e+00 -8.58412832e-02 -3.37081969e-01 6.14674985e-02
1.91666484e-01 5.96060693e-01 -6.48012981e-02 -8.54830742e-02
6.16445482e-01 9.20576304e-02 7.61376202e-01 -5.17864704e-01
1.73338950e-01 8.99850845e-01 9.32046026e-02 -6.31029785e-01
4.54042196e-01 5.30928433e-01 2.11334154e-01 -4.57503855e-01
-8.14368606e-01 -4.66143817e-01 -5.36473989e-01 -3.11536640e-01
1.25306487e+00 -8.85229409e-01 -6.73600256e-01 4.55701083e-01
-7.43178666e-01 -3.95833030e-02 1.44724578e-01 7.28128731e-01
-1.00276001e-01 8.09846520e-01 -7.50372529e-01 -5.58101535e-01
-8.05891693e-01 -1.64434505e+00 7.42826223e-01 2.95545489e-01
2.18331993e-01 -1.12055528e+00 -1.14632227e-01 2.38693610e-01
6.97989464e-01 1.84241131e-01 1.31044602e+00 -5.22882402e-01
-2.53262967e-01 -2.61706352e-01 -7.20355749e-01 3.30299348e-01
5.03605425e-01 -3.24352533e-01 -6.37583971e-01 -1.92998722e-01
2.61519700e-01 -2.92413123e-02 9.38339293e-01 1.10783088e+00
1.42568159e+00 3.35082054e-01 -5.85603476e-01 4.18587297e-01
1.14992213e+00 7.09609866e-01 8.78560901e-01 3.28314036e-01
4.44255382e-01 3.13436389e-01 1.89059198e-01 5.25989942e-02
2.13180676e-01 3.82894367e-01 2.64401823e-01 -5.04470289e-01
-6.17272437e-01 6.68679178e-01 -8.28674957e-02 7.17892408e-01
-3.44625711e-01 3.79964262e-01 -9.11090970e-01 3.20913345e-01
-1.55660641e+00 -7.02039063e-01 -5.14924526e-01 1.94382858e+00
9.61339414e-01 1.24996349e-01 2.17691451e-01 -1.47979990e-01
9.91030633e-01 -5.77170178e-02 -5.59133112e-01 2.54515469e-01
7.20294341e-02 3.45666260e-01 8.89508128e-02 4.27141428e-01
-1.46121275e+00 3.77580583e-01 6.50165081e+00 7.07597077e-01
-1.12882996e+00 3.76003534e-01 8.98666739e-01 2.80378312e-01
1.78271934e-01 -3.88129264e-01 -5.47280610e-01 6.39139891e-01
6.52871013e-01 1.72971308e-01 -4.00103003e-01 5.70811868e-01
6.34426773e-01 2.53934741e-01 -2.13767126e-01 5.17071307e-01
-2.31209710e-01 -1.27653873e+00 1.18791729e-01 -1.12084955e-01
4.14854705e-01 -8.53049755e-02 1.19120449e-01 2.53550828e-01
6.83490932e-02 -8.40543270e-01 -4.06242255e-03 5.29661655e-01
3.64453584e-01 -2.78224170e-01 1.04498005e+00 -6.98182955e-02
-1.36567819e+00 7.40511809e-03 -2.08329961e-01 2.09769443e-01
5.34805894e-01 1.06296206e+00 -3.96752715e-01 4.28270936e-01
6.10398054e-01 7.34243035e-01 -6.82708263e-01 1.52381837e+00
2.91522406e-03 8.72712374e-01 -9.42727402e-02 1.35833859e-01
2.24013135e-01 -4.77260292e-01 3.98068398e-01 1.07749510e+00
3.35281909e-01 -5.48981875e-02 6.65964067e-01 5.97579658e-01
4.26463008e-01 3.62431109e-01 3.19567949e-01 5.33281565e-01
3.40085357e-01 1.26575875e+00 -1.02764809e+00 -6.29297912e-01
-5.54512382e-01 3.43310973e-03 -2.25133091e-01 3.84429872e-01
-9.40487206e-01 -4.54198003e-01 4.57545072e-01 2.20562786e-01
2.48752728e-01 -2.64768302e-01 -1.06561136e+00 -8.67183506e-01
-9.62863639e-02 -5.10990679e-01 6.37448192e-01 -3.39062124e-01
-1.48518562e+00 4.68756199e-01 1.06378542e-02 -1.21804917e+00
4.57429260e-01 -8.50963473e-01 -1.10589087e+00 1.27052546e+00
-1.80030525e+00 -1.00047672e+00 -4.54873443e-01 2.66816080e-01
4.89805222e-01 -3.64165932e-01 6.56511486e-01 6.46430016e-01
-7.54922032e-01 3.75147104e-01 -5.73881995e-03 5.80806971e-01
6.96245968e-01 -1.04801774e+00 -3.95002775e-02 6.17383659e-01
-6.37927055e-01 7.75766253e-01 2.44210646e-01 -8.56307685e-01
-5.92425048e-01 -1.30487359e+00 2.19298750e-01 1.58920348e-01
4.90602583e-01 2.43380845e-01 -1.05330181e+00 4.37866151e-01
-5.49971312e-02 1.36117026e-01 9.26268578e-01 -1.86050251e-01
-1.91704005e-01 -2.85107456e-02 -1.05222976e+00 6.60211623e-01
6.78620815e-01 1.81004882e-01 -5.32110393e-01 6.06147468e-01
4.39929217e-01 -3.12059969e-01 -1.19894588e+00 5.43969333e-01
8.00992846e-01 -4.23292071e-01 1.46197855e+00 -5.41119754e-01
3.54907721e-01 -3.64431560e-01 2.13515446e-01 -1.04083133e+00
-7.40070820e-01 2.36970276e-01 -9.56863612e-02 7.11419404e-01
3.77350241e-01 -7.42368579e-01 8.57056618e-01 9.62535590e-02
-6.11656070e-01 -8.90314460e-01 -9.49895024e-01 -5.02275288e-01
7.00066388e-01 -8.56475383e-02 2.81677365e-01 4.15584415e-01
-6.45447552e-01 1.86769649e-01 -5.48853986e-02 3.51825893e-01
9.71089840e-01 2.24987790e-01 2.71966998e-02 -1.64551485e+00
1.30687803e-01 -7.95825005e-01 -2.09855750e-01 -8.05434763e-01
-5.44500351e-03 -1.25009620e+00 -1.33533493e-01 -1.84667730e+00
2.97973812e-01 -6.73237681e-01 -4.58390087e-01 2.06079677e-01
-6.61103487e-01 3.95653158e-01 -2.54390925e-01 4.51012909e-01
8.42709094e-02 1.99384242e-01 2.08528757e+00 -2.49844253e-01
-4.12520885e-01 5.47257006e-01 -1.15405202e+00 7.52659976e-01
1.18173134e+00 -2.89089739e-01 -4.66146201e-01 -4.61355984e-01
-5.59795678e-01 -3.00554782e-01 4.51599389e-01 -8.32050502e-01
5.00187576e-02 8.11325163e-02 5.86832464e-01 -4.64980721e-01
1.47900939e-01 -5.56957722e-01 -3.90050650e-01 1.02620745e+00
-2.05482483e-01 -4.53750432e-01 1.98297109e-03 3.92172754e-01
1.79756135e-01 -2.31855974e-01 9.86172736e-01 -4.30201352e-01
-4.96310323e-01 4.08038646e-01 -6.42419696e-01 4.15388197e-02
1.15121174e+00 1.79776121e-02 -3.33449870e-01 -6.52935496e-03
-1.10280943e+00 3.57010484e-01 -7.52074718e-01 3.00782681e-01
7.04684675e-01 -1.28239930e+00 -1.18650436e+00 2.33600304e-01
-2.70474225e-01 -5.27994744e-02 3.18541706e-01 1.19852161e+00
-6.00138724e-01 4.89652194e-02 -2.18842477e-01 -8.02861392e-01
-1.05791795e+00 1.32582456e-01 7.12457120e-01 -2.95973778e-01
-1.14926422e+00 5.01403689e-01 1.82272688e-01 -1.68572590e-01
3.13446164e-01 -3.24542612e-01 -6.14669621e-01 1.31129101e-01
7.99834669e-01 4.85594898e-01 1.05935305e-01 -3.68647903e-01
-1.62424073e-01 9.67362702e-01 -3.18355143e-01 1.65629074e-01
1.12960541e+00 -7.73748308e-02 -1.45294577e-01 4.35677953e-02
1.13641667e+00 -5.20587564e-01 -9.48761702e-01 -3.37467194e-01
-1.43716618e-01 -2.44328573e-01 3.63136202e-01 -8.89075994e-01
-1.56476033e+00 1.06806040e+00 1.20640981e+00 -8.09335038e-02
9.88476515e-01 -6.79008439e-02 1.10755599e+00 -5.57722747e-02
3.03605590e-02 -5.35800517e-01 1.37399018e-01 1.73464030e-01
6.60879612e-01 -1.27144778e+00 5.57214506e-02 -8.36366892e-01
-4.39059377e-01 1.34523499e+00 4.65447932e-01 -3.98430467e-01
1.46055603e+00 2.36867249e-01 6.26994133e-01 -1.56180799e-01
2.74145026e-02 -9.01202336e-02 3.95270556e-01 7.31383085e-01
4.60850626e-01 9.29262638e-02 -7.79204369e-01 8.40136111e-01
2.12430760e-01 1.27299517e-01 7.73201063e-02 8.51551056e-01
-9.72370028e-01 -9.72561061e-01 -4.19985592e-01 9.48496819e-01
-4.77762818e-01 1.00640401e-01 4.06794816e-01 5.76217175e-01
1.15196213e-01 9.35752332e-01 1.40309976e-02 -7.30783641e-02
2.12884486e-01 -3.14908773e-02 5.16845398e-02 -3.97994161e-01
-4.49611247e-01 4.78322148e-01 -1.87202305e-01 -1.93171531e-01
-5.44322968e-01 -5.94397068e-01 -1.37858677e+00 4.28898573e-01
-2.25996479e-01 3.18723768e-01 5.28286219e-01 1.15772939e+00
1.95215613e-01 1.03171825e+00 5.62400341e-01 -7.44440794e-01
-5.82689106e-01 -1.11493409e+00 -6.09870970e-01 3.23544890e-01
2.07293444e-04 -8.72076690e-01 -2.60886699e-01 2.17140704e-01] | [14.546273231506348, -2.388399124145508] |
17d81c88-3304-4fb6-b297-47ec9a10bbc7 | the-leaky-integrator-that-could-or-recursive | 2206.04284 | null | https://arxiv.org/abs/2206.04284v3 | https://arxiv.org/pdf/2206.04284v3.pdf | The leaky integrator that could: Or recursive polynomial regression for online signal analysis | Fitting a local polynomial model to a noisy sequence of uniformly sampled observations or measurements (i.e. regressing) by minimizing the sum of weighted squared errors (i.e. residuals) may be used to design digital filters for a diverse range of signal-analysis problems, such as detection, classification and tracking, in biomedical, financial, and aerospace applications, for instance. Furthermore, the recursive realization of such filters, using a network of so-called leaky integrators, yields simple digital components with a low computational complexity and an infinite impulse response (IIR) that are ideal in embedded online sensing systems with high data rates. Target tracking, pulse-edge detection, peak detection and anomaly/change detection are considered in this tutorial as illustrative examples. Erlang-weighted polynomial regression provides a design framework within which the various design trade-offs of state estimators (e.g. bias errors vs. random errors) and IIR smoothers (e.g. frequency isolation vs. time localization) may be intuitively balanced. Erlang weights are configured using a smoothing parameter which determines the decay rate of the exponential tail; and a shape parameter which may be used to discount more recent data, so that a greater relative emphasis is placed on a past time interval. In Morrison's 1969 treatise on sequential smoothing and prediction, the exponential weight (i.e. the zero shape-parameter case) and the Laguerre polynomials that are orthogonal with respect to this weight, are described in detail; however, more general Erlang weights and the resulting associated Laguerre polynomials are not considered there, nor have they been covered in detail elsewhere since. Thus, one of the purposes of this tutorial is to explain how Erlang weights may be used to shape and improve the response of recursive regression filters. | ['Hugh L Kennedy'] | 2022-06-09 | null | null | null | null | ['edge-detection'] | ['computer-vision'] | [ 3.27712834e-01 -1.83014676e-01 6.40361458e-02 -5.16755134e-02
-4.59833771e-01 -3.26275438e-01 3.08467537e-01 1.04645997e-01
-4.79681015e-01 7.05985844e-01 -2.65758187e-01 -3.84573251e-01
-4.38953131e-01 -3.84713054e-01 -8.90971422e-02 -9.43893492e-01
-3.77088100e-01 -1.93326071e-01 3.10335070e-01 -4.71636616e-02
1.01041533e-01 7.85468578e-01 -1.23909402e+00 -4.61471438e-01
6.00269198e-01 1.19107068e+00 4.02167886e-02 8.95278454e-01
4.54881608e-01 3.44071358e-01 -9.24760163e-01 4.76779453e-02
-5.79344593e-02 -2.13780120e-01 -6.24767318e-02 -1.72352910e-01
-3.03680301e-01 -7.11105466e-02 -2.59163052e-01 9.22977507e-01
5.83448827e-01 3.37370366e-01 8.07446063e-01 -7.59878516e-01
-1.72594041e-01 1.34600565e-01 -2.64915347e-01 4.82319325e-01
2.44556531e-01 5.92462253e-04 3.37515175e-01 -7.31552362e-01
-3.69182676e-02 1.04136753e+00 1.11653674e+00 3.18977535e-01
-1.69692516e+00 -4.03321415e-01 2.82232789e-03 -1.35187164e-01
-1.43494844e+00 -5.16972184e-01 6.79367840e-01 -5.31588197e-01
7.29130387e-01 6.06214345e-01 4.60263580e-01 8.67084384e-01
6.92035377e-01 4.27420139e-01 9.50340509e-01 -3.43794405e-01
2.95735240e-01 2.63442963e-01 2.99549788e-01 3.50028694e-01
2.37577707e-01 4.51585561e-01 -9.84405801e-02 -7.65210211e-01
1.01934028e+00 8.08641687e-02 -4.93978769e-01 1.27267301e-01
-8.28908145e-01 7.58992255e-01 6.69254065e-02 3.79272640e-01
-5.10033906e-01 3.46119463e-01 4.96021718e-01 6.14616752e-01
6.40587926e-01 2.00152442e-01 -4.15653080e-01 1.05458349e-01
-9.00936246e-01 2.67729223e-01 1.01256382e+00 5.88104069e-01
2.80095339e-01 7.96112835e-01 -4.31585386e-02 8.81882668e-01
5.03961146e-01 7.00643122e-01 4.19305712e-01 -7.26383924e-01
-1.92052364e-01 -2.92491019e-01 3.57793659e-01 -6.06006265e-01
-8.39552879e-01 -6.31693661e-01 -9.32406247e-01 2.61596739e-01
5.18033922e-01 -4.55934137e-01 -5.31724989e-01 1.49887872e+00
1.98982358e-01 3.08704197e-01 1.27051398e-01 6.47206008e-01
2.18466237e-01 7.29522347e-01 -3.76956537e-02 -8.00328732e-01
1.38711143e+00 -1.87094525e-01 -9.42638934e-01 -1.46134123e-01
5.74450307e-02 -1.00397980e+00 5.36719382e-01 6.59016132e-01
-1.13310778e+00 -4.96782631e-01 -9.87206578e-01 4.34082836e-01
-1.43726036e-01 9.80404392e-02 3.11227679e-01 7.53399670e-01
-8.48902225e-01 1.02279830e+00 -1.11216867e+00 -1.73679188e-01
-2.76291579e-01 2.21849382e-01 3.65778476e-01 5.95820546e-01
-1.37675750e+00 1.04046535e+00 -1.90225989e-01 4.05628175e-01
-4.74509001e-01 -8.14245403e-01 -5.74453294e-01 -1.36543393e-01
-2.04885751e-02 -4.13121164e-01 1.32687914e+00 -8.98812056e-01
-1.60975134e+00 3.19844067e-01 -1.50816023e-01 -7.00718641e-01
3.99948597e-01 -9.21086445e-02 -9.41878855e-01 5.12462780e-02
-4.03097332e-01 -3.78775746e-01 1.47811878e+00 -6.23600721e-01
-4.74405736e-01 -1.67108670e-01 -5.35859048e-01 -1.97449833e-01
1.15608118e-01 1.95419297e-01 3.57055277e-01 -8.51356149e-01
1.27085149e-01 -6.73632145e-01 -6.04660392e-01 -2.05512829e-02
7.85317570e-02 -8.94439220e-02 6.42083585e-01 -7.45394349e-01
1.54707241e+00 -2.34533143e+00 -2.81327695e-01 5.62218368e-01
-9.24669653e-02 9.41022485e-02 2.61517107e-01 6.11451864e-01
-2.58321613e-01 -4.05343592e-01 -3.52403432e-01 1.21983737e-01
-2.08153948e-01 -4.06995751e-02 -5.06262779e-01 9.85035658e-01
2.28249505e-01 2.39722192e-01 -8.52939844e-01 1.62975505e-01
3.03991795e-01 7.40235984e-01 -1.67634729e-02 -7.40374923e-02
2.97252506e-01 3.83011967e-01 -5.01725554e-01 3.75343114e-01
2.56670415e-01 -3.14414175e-03 -2.91957885e-01 -3.58846933e-01
-5.65081716e-01 1.52984977e-01 -1.34187758e+00 7.66199589e-01
-6.23753905e-01 8.68333161e-01 3.98060739e-01 -9.93517816e-01
1.32193291e+00 7.94446468e-01 6.12902224e-01 -2.51704276e-01
9.89596099e-02 4.69699234e-01 -3.07325441e-02 -3.18526536e-01
4.67493266e-01 -4.19698387e-01 8.38713869e-02 8.65441933e-02
-2.87785143e-01 -1.64053723e-01 -1.97972089e-01 -2.62315392e-01
9.65845585e-01 -2.41860181e-01 7.17889965e-01 -5.45730174e-01
6.16317630e-01 -2.55981594e-01 5.09579957e-01 6.10520363e-01
-2.59818822e-01 1.46847934e-01 2.52365708e-01 -1.35626376e-01
-1.00459945e+00 -8.77050102e-01 -6.84247851e-01 6.36374354e-01
-1.97399184e-01 7.96087980e-02 -3.04036260e-01 5.14752865e-01
1.72223210e-01 7.28452325e-01 -1.05388083e-01 -3.82882208e-01
-6.04059875e-01 -6.93628490e-01 5.64432800e-01 4.18108433e-01
-4.24293801e-02 -6.09847009e-01 -9.87556815e-01 8.00314546e-01
2.86027968e-01 -8.68667006e-01 -3.41630280e-01 4.98450100e-01
-1.18481541e+00 -7.67531931e-01 -9.56928730e-01 -4.15135324e-01
4.57313150e-01 -7.09784403e-03 6.86443865e-01 -3.39948237e-01
-2.85247594e-01 1.04118681e+00 -8.94826502e-02 -5.46694279e-01
-3.44847381e-01 -7.39919186e-01 4.11094487e-01 7.38581866e-02
5.55823790e-03 -6.42092228e-01 -6.32675409e-01 4.00717467e-01
-7.45063543e-01 -6.97418809e-01 2.48441666e-01 9.32514608e-01
5.02625346e-01 1.30111754e-01 8.52670014e-01 -6.45031095e-01
1.02812064e+00 -3.85892957e-01 -1.08550298e+00 -1.15902133e-01
-6.17504776e-01 -2.49471396e-01 8.70004535e-01 -9.92690444e-01
-7.94464290e-01 -2.24979669e-01 -1.96390420e-01 -4.57726419e-01
2.73030609e-01 6.61828697e-01 3.67418230e-01 -4.85244066e-01
8.95075321e-01 3.15914929e-01 1.38279229e-01 -3.03726673e-01
1.56180561e-01 7.46370435e-01 6.05448544e-01 -3.72521073e-01
5.09317219e-01 2.24758670e-01 2.63913512e-01 -1.36369991e+00
-1.67074203e-01 -6.23646796e-01 6.68833265e-04 -2.10579008e-01
3.59100878e-01 -7.56344259e-01 -9.17852819e-01 4.91591662e-01
-9.93663967e-01 -2.38543153e-01 -5.72449088e-01 7.94304967e-01
-9.56814766e-01 2.59488255e-01 -8.29578400e-01 -1.67097521e+00
-3.88956606e-01 -7.15413809e-01 5.91556132e-01 3.65475178e-01
-4.15743083e-01 -1.28965545e+00 -1.15240529e-01 -5.90434492e-01
6.80214703e-01 3.47322792e-01 6.26552820e-01 -5.87035775e-01
-1.79731637e-01 -6.37699962e-01 2.69620448e-01 5.81005812e-01
-1.14170298e-01 1.56812117e-01 -9.26305592e-01 -4.63999450e-01
7.26399302e-01 4.68068779e-01 4.53262091e-01 9.31191027e-01
7.52980888e-01 -4.25946832e-01 -4.07368183e-01 2.87494063e-01
1.43312097e+00 4.49882120e-01 4.46601391e-01 -1.03147969e-01
-2.37507299e-02 3.66550416e-01 6.56441808e-01 7.47063398e-01
-4.25864875e-01 4.13701445e-01 1.66911203e-02 1.89292934e-02
3.03669721e-01 3.78797740e-01 4.54345375e-01 7.93392062e-01
-2.10335031e-01 1.03108749e-01 -3.30313921e-01 3.30214798e-01
-1.58100009e+00 -9.22305763e-01 -4.42448407e-01 2.79935884e+00
6.53749228e-01 2.58621126e-01 1.15093417e-01 3.85885030e-01
7.65774190e-01 -1.93716027e-02 -5.99370122e-01 -5.65363944e-01
1.95785001e-01 4.07543302e-01 1.15390205e+00 6.81261182e-01
-9.18398678e-01 6.02835380e-02 6.79332066e+00 8.37838352e-01
-1.09531987e+00 -6.21583499e-03 1.87190309e-01 2.44735658e-01
5.71658462e-02 -1.57068763e-02 -8.13610613e-01 5.61902940e-01
1.47954512e+00 -3.95255536e-01 5.23650169e-01 6.89256370e-01
7.38085330e-01 -2.79026568e-01 -8.52135777e-01 9.71196949e-01
-6.00112975e-01 -8.24827909e-01 -8.42475176e-01 -6.30835965e-02
3.69512349e-01 -2.04709515e-01 8.01123083e-02 1.40398368e-01
-7.03621730e-02 -6.96066141e-01 7.08145499e-01 1.14995313e+00
6.56165838e-01 -6.77696228e-01 4.40849781e-01 5.32249570e-01
-1.30076098e+00 -1.78497404e-01 -4.54435349e-01 -1.65312383e-02
4.90837067e-01 1.23783743e+00 -4.23147768e-01 5.07085145e-01
2.20183149e-01 3.45359772e-01 3.37470084e-01 1.47851610e+00
1.33520439e-01 7.57391334e-01 -7.65982091e-01 -4.07179862e-01
-8.98902584e-03 -3.31723928e-01 1.08506703e+00 1.30755687e+00
4.39706951e-01 3.01642537e-01 3.63031365e-02 7.45140612e-01
6.68379903e-01 -1.89293876e-01 -1.76688537e-01 2.31773078e-01
6.56694412e-01 1.11726153e+00 -4.70592022e-01 -2.18035698e-01
-5.80368876e-01 4.32564616e-01 -6.40583754e-01 6.79216146e-01
-7.36667514e-01 -8.00425410e-01 7.06849933e-01 3.22634161e-01
6.47019744e-02 -3.59662145e-01 -1.64962366e-01 -6.04014933e-01
-2.68861085e-01 -5.74271321e-01 2.93505043e-01 -5.36435187e-01
-1.36808550e+00 4.39817011e-01 1.13476291e-01 -1.37202656e+00
-4.04872656e-01 -5.26278257e-01 -7.01504171e-01 1.31470633e+00
-1.23007727e+00 -2.92901933e-01 2.29929358e-01 5.86879611e-01
3.76221091e-01 1.17584504e-01 7.88271844e-01 3.40236813e-01
-2.52621323e-01 3.36088270e-01 5.60742855e-01 -3.47952485e-01
4.97928888e-01 -1.09152591e+00 1.22776918e-01 6.39408529e-01
-5.18877327e-01 7.30534077e-01 1.41171467e+00 -5.84490120e-01
-1.18734705e+00 -8.50670397e-01 7.02384531e-01 3.82833369e-02
1.06665003e+00 1.33203313e-01 -1.11564672e+00 4.13763046e-01
-3.49344969e-01 1.50272250e-02 3.69881988e-01 -2.16211036e-01
1.56897038e-01 -3.81923229e-01 -1.27777135e+00 5.09799123e-01
1.96972668e-01 -4.34542328e-01 -3.85745972e-01 4.06626880e-01
3.48845840e-01 -5.03615618e-01 -1.17849422e+00 4.13391948e-01
6.67398751e-01 -6.48420632e-01 1.02253902e+00 -1.88642815e-01
-5.82322955e-01 -3.73503208e-01 1.39748007e-02 -1.26974201e+00
-5.44961691e-01 -1.22152317e+00 -2.90233493e-01 9.08343434e-01
1.77209109e-01 -1.33168435e+00 1.81164950e-01 4.19218063e-01
-2.07579851e-01 -7.51978576e-01 -1.00078356e+00 -1.04427981e+00
-3.61752599e-01 -3.28676939e-01 -1.06690088e-02 3.99252117e-01
1.31268576e-01 5.82244880e-02 -2.60463506e-01 3.99629027e-01
7.72606134e-01 -3.60725880e-01 1.83747187e-01 -1.28036547e+00
-4.26485419e-01 -5.46421766e-01 -5.44624686e-01 -1.24951077e+00
-4.98775095e-01 -3.24462503e-01 1.34865671e-01 -8.66788507e-01
-7.12480903e-01 -4.59133208e-01 -2.87955463e-01 -8.98056328e-02
3.80422287e-02 -1.35424972e-01 -2.57260293e-01 2.43107110e-01
3.95463556e-01 1.41106918e-01 8.05640399e-01 2.15632111e-01
-4.86678690e-01 1.07808971e+00 -3.19823563e-01 8.92044425e-01
7.05291688e-01 -4.16660547e-01 -3.76751512e-01 3.64920050e-01
6.03859611e-02 7.65168488e-01 6.55160904e-01 -1.05323362e+00
4.43982959e-01 -2.24058535e-02 3.47103447e-01 -5.32503426e-01
4.98117089e-01 -9.76343453e-01 6.19981706e-01 7.75244415e-01
-1.85950920e-01 -5.22007085e-02 1.17955372e-01 8.82527709e-01
-3.20497863e-02 -5.82276940e-01 1.20050693e+00 1.11780539e-01
-2.84167200e-01 -9.09894705e-03 -8.13901782e-01 -3.98468435e-01
7.96407104e-01 -4.17038947e-01 4.79397215e-02 -5.25613248e-01
-1.11382616e+00 4.68826331e-02 -2.26403177e-02 -1.08977973e-01
4.96824443e-01 -1.11568165e+00 -5.81992745e-01 1.36049464e-01
-3.53782356e-01 -4.48889673e-01 2.79499233e-01 1.31320667e+00
-6.58791140e-02 4.01078910e-01 2.53226310e-01 -6.23941481e-01
-1.16033173e+00 2.54314303e-01 6.50523543e-01 -1.63473323e-01
-5.98910987e-01 5.50943673e-01 -9.43761021e-02 2.83834815e-01
3.41184318e-01 -5.94214201e-01 -1.99361313e-02 -1.81674771e-02
7.77663946e-01 7.18714952e-01 1.51101336e-01 -1.43614516e-01
-3.79336119e-01 4.79341358e-01 3.20580363e-01 -7.26535842e-02
1.02193511e+00 -1.53396338e-01 -7.91809261e-02 9.43322122e-01
8.73236299e-01 7.49383494e-02 -1.40984809e+00 -1.97379217e-01
3.07060294e-02 -1.22319154e-01 1.91034265e-02 -4.95644361e-01
-6.08457267e-01 5.53475440e-01 6.67559147e-01 8.71932924e-01
1.31958711e+00 -3.79058838e-01 5.20203233e-01 4.27419804e-02
2.73150027e-01 -9.83941674e-01 -4.67627883e-01 5.26631296e-01
7.77884662e-01 -2.39610285e-01 1.15351200e-01 -2.99868912e-01
-1.79648340e-01 1.49314129e+00 -2.60349751e-01 -6.87262774e-01
8.90626371e-01 5.19025743e-01 -1.28265545e-01 2.96672702e-01
-5.88752508e-01 -5.58924079e-02 3.07441831e-01 7.17224538e-01
5.37669539e-01 2.20442876e-01 -7.48342454e-01 6.36706889e-01
9.59021673e-02 -6.20183945e-02 5.88850975e-01 7.08052993e-01
-9.24795926e-01 -6.93420172e-01 -9.98893440e-01 6.95375562e-01
-5.79232395e-01 6.69466853e-02 5.06749690e-01 5.59069276e-01
-4.87708300e-01 1.05142856e+00 -4.82023582e-02 1.30578905e-01
6.44385099e-01 -1.55158923e-03 4.39412862e-01 -1.39492840e-01
-5.86927950e-01 6.96063459e-01 1.72921255e-01 -3.98924649e-01
-1.82372704e-01 -7.28885710e-01 -1.14622366e+00 -3.68496269e-01
-6.12513602e-01 1.27338663e-01 8.41393828e-01 7.76093543e-01
-2.73309618e-01 7.33511925e-01 6.22957706e-01 -7.77532756e-01
-1.19776523e+00 -7.91476488e-01 -1.10322630e+00 -3.83676946e-01
7.63303638e-01 -4.71266359e-01 -6.61692083e-01 -7.13406205e-02] | [6.517111301422119, 3.5630974769592285] |
a5c449ac-1f19-4faf-ada3-9ff9108009f1 | physics-guided-generative-adversarial-1 | 2203.14352 | null | https://arxiv.org/abs/2203.14352v3 | https://arxiv.org/pdf/2203.14352v3.pdf | Physics Guided Deep Learning for Generative Design of Crystal Materials with Symmetry Constraints | Discovering new materials is a challenging task in materials science crucial to the progress of human society. Conventional approaches based on experiments and simulations are labor-intensive or costly with success heavily depending on experts' heuristic knowledge. Here, we propose a deep learning based Physics Guided Crystal Generative Model (PGCGM) for efficient crystal material design with high structural diversity and symmetry. Our model increases the generation validity by more than 700\% compared to FTCP, one of the latest structure generators and by more than 45\% compared to our previous CubicGAN model. Density Functional Theory (DFT) calculations are used to validate the generated structures with 1,869 materials out of 2,000 are successfully optimized and deposited into the Carolina Materials Database \url{www.carolinamatdb.org}, of which 39.6\% have negative formation energy and 5.3\% have energy-above-hull less than 0.25 eV/atom, indicating their thermodynamic stability and potential synthesizability. | ['Mohammed Al-Fahdi', 'Nihang Fu', 'Jianjun Hu', 'Ming Hu', 'Zhenyao Wu', 'Edirisuriya M. Dilanga Siriwardane', 'Yong Zhao'] | 2022-03-27 | null | null | null | null | ['formation-energy'] | ['miscellaneous'] | [-1.10729426e-01 1.60430118e-01 -1.44479483e-01 1.85362641e-02
-6.49042010e-01 -2.61284858e-01 4.68828738e-01 7.03693107e-02
-6.88901916e-02 1.48097539e+00 2.11006224e-01 -1.70193344e-01
-3.21058333e-02 -1.23031604e+00 -9.70835924e-01 -1.17335081e+00
1.37703672e-01 7.41692960e-01 1.12730321e-02 -3.56119037e-01
4.83398855e-01 3.93898308e-01 -1.72129476e+00 1.15516089e-01
1.31962454e+00 7.72090673e-01 5.28912127e-01 2.32786596e-01
-1.22966291e-02 1.79603621e-01 -3.07631671e-01 -2.29457721e-01
1.23430528e-01 -4.79307085e-01 -6.55998170e-01 -4.65366393e-01
-1.68506689e-02 -1.34012271e-02 -1.96482748e-01 9.00707901e-01
6.51479185e-01 5.09967506e-02 1.10349393e+00 -5.63438833e-01
-1.12092888e+00 7.48949230e-01 -4.14034367e-01 -2.41603255e-02
2.85172522e-01 4.19522017e-01 1.06520367e+00 -1.06629264e+00
9.31799412e-01 6.47514462e-01 4.35278058e-01 7.54978418e-01
-1.08648431e+00 -1.07716954e+00 -3.65080029e-01 2.84986615e-01
-1.55029392e+00 -3.98709238e-01 7.80432642e-01 -6.09899163e-01
1.31586492e+00 8.37720232e-04 8.80680561e-01 9.06952083e-01
5.12688994e-01 3.58117744e-02 9.81040478e-01 -4.67300653e-01
4.44347829e-01 -3.13331708e-02 -1.90446272e-01 7.16838598e-01
8.94496083e-01 2.25306079e-01 -6.69719458e-01 4.75931028e-03
6.41084135e-01 -3.22675854e-01 -8.31456408e-02 -2.22028971e-01
-8.34029555e-01 8.50317121e-01 8.06433439e-01 3.22082222e-01
-5.81724346e-01 3.88021946e-01 -2.01150849e-01 -3.14508051e-01
7.57584497e-02 8.89969468e-01 -2.51242489e-01 -1.14214703e-01
-7.11279929e-01 7.67904818e-01 4.13241446e-01 8.80208611e-01
5.95023274e-01 3.66130859e-01 9.90993157e-02 5.18126011e-01
3.14667165e-01 5.90472341e-01 1.10433064e-01 -5.44484913e-01
6.17096536e-02 5.76110005e-01 4.67287004e-02 -6.45436525e-01
-3.49995971e-01 -5.95678866e-01 -1.05997765e+00 1.12974465e-01
-7.07557052e-02 -1.18030518e-01 -1.00956845e+00 1.52788603e+00
1.90825999e-01 -4.18427825e-01 3.29939499e-02 5.94923556e-01
1.27619541e+00 1.07421958e+00 8.14811364e-02 -4.58488047e-01
8.77878368e-01 -5.74341655e-01 -3.49794984e-01 3.56927216e-01
3.51542532e-01 -5.78428388e-01 8.71391237e-01 5.10078251e-01
-1.36401057e+00 -4.15569603e-01 -1.29690409e+00 2.46483997e-01
-3.09318155e-01 -2.27069050e-01 9.23113644e-01 8.44458520e-01
-7.90580571e-01 8.65780294e-01 -6.71876252e-01 6.33472428e-02
6.94968581e-01 6.55244350e-01 -3.96116264e-02 -1.91451088e-02
-9.57143366e-01 6.71099246e-01 6.34408057e-01 -2.51076519e-01
-8.03635478e-01 -8.69324207e-01 -2.57211506e-01 -1.06942818e-01
2.27652565e-01 -8.97604108e-01 8.87799501e-01 -9.58119407e-02
-1.50383353e+00 5.92407763e-01 -2.43968979e-01 -2.70098239e-01
6.81044534e-02 2.39939280e-02 -4.23159271e-01 -4.22552824e-02
1.58326715e-01 6.34368420e-01 2.93297976e-01 -1.40446007e+00
-1.14535376e-01 -2.64109075e-02 -3.16404462e-01 -1.11039579e-02
-1.63876310e-01 -4.48714912e-01 3.46598923e-02 -6.22866273e-01
1.72478452e-01 -9.07525897e-01 -4.34629321e-01 -4.61964637e-01
-6.64466977e-01 -3.27079117e-01 2.97642052e-01 -5.83305001e-01
1.08868384e+00 -1.33371675e+00 2.44941399e-01 4.90958661e-01
3.97327840e-01 1.14943631e-01 3.83012712e-01 8.14343035e-01
-1.46236941e-01 2.60866523e-01 -3.27243656e-01 4.86260295e-01
-1.12485692e-01 -2.65024394e-01 1.97658747e-01 2.74304986e-01
-9.83877704e-02 1.03193343e+00 -7.29128838e-01 -1.28407389e-01
1.47056401e-01 2.81095296e-01 -8.99309337e-01 -1.64783195e-01
-6.28310800e-01 7.13710308e-01 -6.10625446e-01 8.75943720e-01
7.19396770e-01 -3.88255119e-01 1.54118016e-01 -3.65180150e-02
-4.35211211e-01 5.25541067e-01 -8.47393274e-01 1.47710717e+00
-1.03805318e-01 5.31201139e-02 -3.26245993e-01 -6.95560455e-01
9.72672760e-01 1.11461639e-01 8.15834165e-01 -9.81014609e-01
5.42906076e-02 4.19178784e-01 3.97553831e-01 -2.47834668e-01
5.63585699e-01 -4.63279307e-01 -5.46484180e-02 5.28476596e-01
-6.94458485e-02 -3.71863544e-01 3.32601577e-01 1.37513146e-01
7.86699295e-01 1.05131984e-01 1.13717690e-01 -7.19232082e-01
3.77142370e-01 1.92696765e-01 4.61801738e-01 5.71797192e-01
3.37088823e-01 3.56564879e-01 3.01866676e-04 -6.15311682e-01
-1.54400992e+00 -1.19587588e+00 -3.45090538e-01 5.03512144e-01
1.20990291e-01 -6.10008061e-01 -7.34417617e-01 3.93913798e-02
-2.02133521e-01 8.53012025e-01 -1.07797422e-01 -4.22805786e-01
-5.67624569e-01 -1.24893129e+00 -9.88142863e-02 3.89207304e-01
6.27096355e-01 -1.31872869e+00 -1.73208758e-01 3.57686639e-01
9.23337489e-02 -6.56369328e-01 2.93349940e-02 7.26271570e-02
-6.47002876e-01 -8.54234874e-01 -5.64031541e-01 -6.99658275e-01
6.43085182e-01 -1.28140122e-01 1.14361954e+00 1.45278305e-01
-7.53002822e-01 -3.33413363e-01 -2.62696832e-01 -4.14008200e-01
-6.84067011e-01 3.17438930e-01 3.93131822e-01 -7.41859317e-01
7.29250759e-02 -9.33165252e-01 -8.73865306e-01 -9.40558389e-02
-4.31513727e-01 3.86316389e-01 8.09365273e-01 7.07050860e-01
8.45405281e-01 5.98783851e-01 6.99012399e-01 -6.58145487e-01
5.04995167e-01 -4.68980998e-01 -6.60573244e-01 3.10571305e-02
-8.99227440e-01 3.57599765e-01 6.54180229e-01 8.76120850e-02
-1.07453394e+00 6.17807470e-02 -4.11609799e-01 1.75481111e-01
-5.07003181e-02 3.18977237e-01 -3.14509064e-01 -1.02803044e-01
7.45552242e-01 3.63633931e-01 -5.06140709e-01 -2.46972591e-01
2.44470179e-01 3.04757923e-01 4.10058320e-01 -9.65066969e-01
8.76819611e-01 1.15730591e-01 4.07069683e-01 -1.00779116e+00
-4.62499022e-01 1.60885453e-01 -5.80701828e-01 -2.66146600e-01
9.94547486e-01 -7.40970910e-01 -1.06069028e+00 3.44712913e-01
-6.93011343e-01 -4.15374711e-02 -1.09537721e-01 2.95876622e-01
-3.49857390e-01 4.78621237e-02 -4.47384477e-01 -7.72052705e-01
-8.18889081e-01 -1.20383847e+00 8.49415481e-01 3.42299074e-01
-1.96656004e-01 -6.82811797e-01 -1.23866452e-02 6.27229989e-01
5.76684952e-01 7.24272847e-01 1.35510325e+00 9.62268189e-03
-1.02801800e+00 4.57313657e-02 1.79091394e-02 -1.55107751e-01
2.88108110e-01 5.70725417e-03 -6.64978087e-01 -2.44523510e-01
-2.27127150e-01 -1.03645302e-01 9.65767860e-01 6.95859194e-01
1.33570480e+00 -2.82719672e-01 -5.29349923e-01 3.62226933e-01
1.63561440e+00 6.82666540e-01 7.62843668e-01 1.52644187e-01
9.14354026e-01 2.25613162e-01 3.60970907e-02 6.07710898e-01
1.45226806e-01 5.05060256e-01 4.64412093e-01 2.50605434e-01
-2.55347583e-02 -2.72795647e-01 9.59498882e-02 9.03555870e-01
-7.07019150e-01 -3.44172746e-01 -1.11607194e+00 1.37986526e-01
-1.28630209e+00 -1.16842389e+00 -4.87100571e-01 2.22635412e+00
8.75730813e-01 4.86177385e-01 9.70210060e-02 3.08305491e-02
5.92382967e-01 -1.32919714e-01 -9.70453262e-01 -2.51086235e-01
-1.18692577e-01 8.59226406e-01 5.94117463e-01 1.53546676e-01
-5.99179089e-01 1.02399147e+00 6.71461678e+00 9.64738071e-01
-9.56839502e-01 -2.13637009e-01 7.00708270e-01 -3.29914153e-01
-6.14832163e-01 7.63849169e-02 -8.97100568e-01 7.04519451e-01
9.12899017e-01 -4.51673985e-01 3.30483913e-01 6.38940573e-01
7.40059558e-03 -1.56959876e-01 -7.48744845e-01 1.06342804e+00
-4.38188434e-01 -2.15678096e+00 2.17120916e-01 5.12053728e-01
1.06504107e+00 -2.12659940e-01 5.43379039e-02 7.42139742e-02
3.86904627e-01 -1.36417842e+00 6.85459316e-01 6.39943182e-01
9.26078916e-01 -1.16137528e+00 3.10560971e-01 1.55294761e-01
-1.12849784e+00 2.39873767e-01 -2.58156389e-01 3.00936438e-02
2.98646957e-01 9.24183428e-01 -8.61995041e-01 6.45089328e-01
7.45664299e-01 4.65615392e-01 -1.70081943e-01 7.37776339e-01
-7.18482137e-02 6.87406659e-01 -3.11623663e-01 -4.49581742e-01
1.01604499e-01 -5.68168819e-01 4.58967656e-01 4.79853094e-01
5.05371153e-01 2.33411342e-01 9.94293839e-02 1.33899629e+00
-3.75983387e-01 -2.92568114e-02 -2.80294180e-01 -2.52994984e-01
5.80965579e-01 7.92984843e-01 -8.16775322e-01 -2.26761863e-01
-2.70686205e-03 5.41795611e-01 -8.55376292e-03 1.06867664e-02
-8.49072337e-01 -1.80056930e-01 4.10046577e-01 4.15985733e-01
3.22471738e-01 -3.68522555e-01 -2.59774208e-01 -6.71229362e-01
1.43084943e-01 -4.48928744e-01 -1.61842540e-01 -6.45383537e-01
-1.32162654e+00 2.98343331e-01 2.38744188e-02 -6.99649036e-01
2.38541998e-02 -6.28428400e-01 -6.02042019e-01 8.22538912e-01
-9.91320133e-01 -9.10019815e-01 -1.97476372e-01 5.79652935e-02
2.85378933e-01 -2.02952355e-01 6.95158839e-01 1.98418066e-01
-6.63281918e-01 4.06516135e-01 4.84357506e-01 -5.39826751e-01
1.29288167e-01 -1.10598302e+00 4.70693767e-01 4.65440542e-01
-3.56909364e-01 5.02754569e-01 9.27955151e-01 -8.98398101e-01
-1.78270185e+00 -8.92923117e-01 7.35261381e-01 -2.38849327e-01
3.93400341e-01 -6.30162001e-01 -7.71125972e-01 -9.68614873e-03
2.64888585e-01 -7.21474648e-01 8.10845912e-01 -1.03266366e-01
1.90362677e-01 1.36185005e-01 -1.22249925e+00 6.68616593e-01
1.48184431e+00 6.33242214e-03 -1.62078440e-01 7.40429342e-01
7.66000330e-01 -3.42791140e-01 -1.02889121e+00 6.12576485e-01
4.97929722e-01 -9.94615972e-01 1.04678762e+00 -2.77810991e-01
6.12089634e-01 -5.33005111e-02 -2.47463137e-01 -9.04096305e-01
-6.68714046e-01 -6.02977276e-01 -5.51578477e-02 1.10615301e+00
6.21261001e-01 -4.78689849e-01 1.15604377e+00 5.14319658e-01
-5.65448105e-01 -1.02627516e+00 -8.20726454e-01 -8.44094574e-01
5.24171054e-01 -1.76624894e-01 9.66325223e-01 6.43269837e-01
-3.32090825e-01 3.53053331e-01 -1.95456326e-01 -1.34976313e-01
6.35694146e-01 2.78921425e-01 4.14575368e-01 -1.47564757e+00
-2.17948109e-01 -5.84001720e-01 -3.77870239e-02 -4.18309212e-01
7.11975852e-03 -1.14285076e+00 -3.29518527e-01 -1.88928771e+00
4.04593259e-01 -6.23377025e-01 -4.56099398e-02 3.49772304e-01
3.24444950e-01 6.03785105e-02 -3.29768896e-01 2.94147730e-01
-1.38857633e-01 1.11447990e+00 1.40025318e+00 -4.78257146e-03
-3.26705098e-01 -2.37197146e-01 -9.48377669e-01 3.63921732e-01
9.85269964e-01 -3.52243096e-01 -2.17871934e-01 9.49242264e-02
7.63070166e-01 -2.39301398e-01 -2.82893237e-02 -1.30180001e+00
-2.48769850e-01 -2.83611864e-01 6.01051152e-01 -9.37025368e-01
3.62291425e-01 -2.48728484e-01 6.72929704e-01 8.28567445e-01
9.15120468e-02 -1.02659866e-01 8.18553790e-02 3.71977776e-01
2.84902483e-01 -2.75546640e-01 7.71916807e-01 -4.52109188e-01
-4.64869589e-01 5.96318424e-01 -5.01515985e-01 -2.58963048e-01
1.13685668e+00 -3.51657659e-01 -1.89206764e-01 1.16307316e-02
-6.16143167e-01 2.18929839e-03 6.24137998e-01 1.02055155e-01
5.98084271e-01 -1.47157669e+00 -5.37293613e-01 4.08941060e-02
2.48691328e-02 2.33220354e-01 4.40856695e-01 3.29784542e-01
-9.79241848e-01 6.02069020e-01 -2.64308095e-01 -3.76716882e-01
-8.13935578e-01 4.52406615e-01 1.11694068e-01 -3.67377363e-02
-6.29683852e-01 8.82217348e-01 8.54791328e-02 -1.69102564e-01
-3.23586404e-01 -2.27584854e-01 8.99092406e-02 -3.76940906e-01
5.09906337e-02 4.01171535e-01 3.10723186e-01 -6.51179016e-01
-3.36562037e-01 7.00716436e-01 -2.77024478e-01 1.31085804e-02
1.92035389e+00 4.79780465e-01 -1.82593748e-01 -7.31415898e-02
8.01799595e-01 -3.77322137e-02 -9.66418803e-01 9.59951878e-02
-1.82055831e-01 -1.85766406e-02 6.98859710e-03 -6.58907175e-01
-9.40376461e-01 3.62696201e-01 6.08198762e-01 -1.22049913e-01
8.38990390e-01 2.25475058e-01 1.07501781e+00 5.24942040e-01
5.52651048e-01 -1.16272414e+00 1.69884205e-01 2.60724843e-01
8.22802961e-01 -9.73988533e-01 3.73411328e-01 -4.26592290e-01
-1.61683723e-01 9.16684568e-01 6.90055490e-01 -1.82938203e-01
6.07990801e-01 -1.00750057e-02 -7.38736272e-01 -6.07187569e-01
-6.33130729e-01 6.70580417e-02 1.26890555e-01 5.35487831e-01
6.53600395e-01 1.63482666e-01 -4.26053673e-01 3.80484343e-01
-8.42565119e-01 -2.88734466e-01 2.75160164e-01 1.18978143e+00
-7.15869367e-01 -1.43642759e+00 -1.44534722e-01 5.65709472e-01
-1.93661213e-01 -3.27677786e-01 -4.28012788e-01 5.97851336e-01
3.47860724e-01 6.51430190e-01 -7.41340369e-02 -1.72708213e-01
-5.04302867e-02 -5.03623672e-02 8.76410306e-01 -5.17247856e-01
-2.55385816e-01 -8.07223767e-02 4.10179831e-02 -2.06845522e-01
-3.52023840e-01 -6.57944977e-01 -1.70660651e+00 -5.76252341e-01
-4.41413641e-01 2.55639136e-01 7.86030829e-01 7.60943294e-01
5.22043228e-01 5.00712395e-01 5.58242142e-01 -1.00100994e+00
2.01434761e-01 -6.32802010e-01 -5.27765572e-01 6.09426498e-02
-2.63988018e-01 -9.35115635e-01 8.86313841e-02 -1.03565946e-01] | [5.179972171783447, 5.355531215667725] |
7e706213-5af6-4b8c-96d9-2864e0b3f313 | weakly-supervised-hoi-detection-from | 2303.05546 | null | https://arxiv.org/abs/2303.05546v1 | https://arxiv.org/pdf/2303.05546v1.pdf | Weakly-Supervised HOI Detection from Interaction Labels Only and Language/Vision-Language Priors | Human-object interaction (HOI) detection aims to extract interacting human-object pairs and their interaction categories from a given natural image. Even though the labeling effort required for building HOI detection datasets is inherently more extensive than for many other computer vision tasks, weakly-supervised directions in this area have not been sufficiently explored due to the difficulty of learning human-object interactions with weak supervision, rooted in the combinatorial nature of interactions over the object and predicate space. In this paper, we tackle HOI detection with the weakest supervision setting in the literature, using only image-level interaction labels, with the help of a pretrained vision-language model (VLM) and a large language model (LLM). We first propose an approach to prune non-interacting human and object proposals to increase the quality of positive pairs within the bag, exploiting the grounding capability of the vision-language model. Second, we use a large language model to query which interactions are possible between a human and a given object category, in order to force the model not to put emphasis on unlikely interactions. Lastly, we use an auxiliary weakly-supervised preposition prediction task to make our model explicitly reason about space. Extensive experiments and ablations show that all of our contributions increase HOI detection performance. | ['Adriana Kovashka', 'Mesut Erhan Unal'] | 2023-03-09 | null | null | null | null | ['human-object-interaction-detection'] | ['computer-vision'] | [ 4.38412070e-01 5.66464007e-01 1.73882656e-02 -3.53749037e-01
-4.86599505e-01 -4.36393589e-01 8.60474944e-01 2.26020366e-01
-5.08969963e-01 4.14213270e-01 8.47071186e-02 -1.10764727e-01
2.95879934e-02 -4.90001261e-01 -9.38112378e-01 -5.36076725e-01
-4.60989438e-02 8.30124557e-01 4.82320666e-01 8.28274637e-02
-2.96902433e-02 2.51418591e-01 -1.80717885e+00 4.29022521e-01
6.23455644e-01 8.29382122e-01 2.39840567e-01 4.17575896e-01
3.09740990e-01 7.96389818e-01 -2.99927860e-01 -3.70444059e-01
3.02498013e-01 -5.84255993e-01 -8.16812575e-01 4.24132049e-01
5.20154119e-01 -1.56552434e-01 1.00943036e-01 9.03309703e-01
1.08188480e-01 -6.56998679e-02 7.62444496e-01 -1.46998155e+00
-3.85288030e-01 5.60775220e-01 -6.90719187e-01 -1.02794757e-02
3.29330087e-01 4.08075273e-01 1.33015060e+00 -1.18115127e+00
8.32773089e-01 1.35018718e+00 2.85386980e-01 4.09922868e-01
-1.31503487e+00 -3.47080022e-01 2.69285351e-01 1.70176849e-01
-1.23547328e+00 -1.68690398e-01 7.01745510e-01 -4.70887810e-01
1.04875553e+00 3.02633643e-01 8.14244211e-01 9.88584518e-01
-2.50063151e-01 1.28465152e+00 1.01389706e+00 -8.50071728e-01
-5.41228950e-02 7.08001673e-01 5.26819289e-01 9.05933499e-01
2.42699698e-01 1.33350089e-01 -7.23152697e-01 -6.54086843e-02
4.50739175e-01 -2.84245431e-01 -7.95091614e-02 -8.05726767e-01
-1.49784923e+00 7.29809046e-01 5.74709296e-01 4.16220933e-01
-2.56669968e-01 -1.36043325e-01 1.25572965e-01 -9.28833634e-02
2.87198812e-01 4.63224620e-01 -1.31625339e-01 4.72233534e-01
-5.82884073e-01 4.78334934e-01 8.24245632e-01 9.31986034e-01
7.82296181e-01 -7.46022344e-01 -3.56658936e-01 6.79783285e-01
4.59768027e-01 2.14717910e-01 -8.23465958e-02 -7.04793692e-01
4.43857193e-01 8.24879050e-01 1.02244884e-01 -8.90736222e-01
-3.13747704e-01 -2.16106072e-01 -5.53795636e-01 3.38092774e-01
5.63114226e-01 2.65654802e-01 -7.38067031e-01 1.67562079e+00
6.27303302e-01 -2.07685053e-01 -1.17005825e-01 1.05066431e+00
6.83043301e-01 3.03890556e-01 2.73162603e-01 -6.61618542e-03
1.60692334e+00 -1.19689643e+00 -3.01699936e-01 -6.80235922e-01
8.11585426e-01 -4.89457160e-01 1.16869485e+00 2.28966996e-01
-1.16076279e+00 -6.41294122e-01 -9.07135963e-01 -2.88435340e-01
-3.89734149e-01 2.75407672e-01 7.74845481e-01 1.32932201e-01
-7.41746604e-01 -5.65643199e-02 -5.16001284e-01 -6.02710545e-01
4.95328039e-01 3.97266567e-01 -4.45256531e-01 -1.57625690e-01
-1.02633440e+00 1.12417734e+00 5.87610006e-01 1.48703694e-01
-7.44242430e-01 -3.60378414e-01 -9.78958786e-01 -9.75475386e-02
7.90945172e-01 -9.59029913e-01 9.37812150e-01 -1.09939396e+00
-6.79353297e-01 1.24031019e+00 -2.13372573e-01 -6.10987723e-01
5.89173317e-01 -1.34366930e-01 1.81892693e-01 1.56130001e-01
1.98346987e-01 1.21686840e+00 8.96932483e-01 -1.61974728e+00
-8.45725060e-01 -5.05547345e-01 2.37134486e-01 4.32078987e-01
-1.06730573e-01 1.97768375e-01 -6.99485540e-01 -3.29021215e-01
2.26478785e-01 -1.25829792e+00 -1.41892344e-01 1.77728102e-01
-5.02121091e-01 -6.61083102e-01 7.81206548e-01 -4.35982466e-01
7.38125980e-01 -2.14932561e+00 1.92761421e-01 1.29629076e-01
2.78569907e-01 1.71367243e-01 -2.25010604e-01 8.98888782e-02
-1.33627299e-02 -1.66760758e-01 -3.54005605e-01 -5.34413457e-01
2.45534386e-02 3.67348641e-01 -2.97724068e-01 4.49849784e-01
4.51755255e-01 1.25679493e+00 -9.60648477e-01 -8.07433605e-01
3.59756500e-01 2.83135384e-01 -7.23793864e-01 3.83426100e-01
-5.62627792e-01 4.59004670e-01 -1.26017094e-01 5.93547404e-01
3.85127306e-01 -3.41823280e-01 8.96937475e-02 -2.30626285e-01
-1.65992752e-02 2.40347490e-01 -1.07144654e+00 1.22203577e+00
-3.08631390e-01 3.21459800e-01 3.15833502e-02 -1.16301644e+00
5.07204115e-01 1.17417872e-01 4.27981406e-01 -4.48759615e-01
1.91360787e-02 8.59047472e-02 2.84957290e-01 -5.52796721e-01
4.51567098e-02 2.50096936e-02 2.68940907e-02 3.04584116e-01
1.44079223e-01 -1.09990552e-01 3.24761987e-01 4.53732044e-01
7.98218012e-01 1.89736411e-01 3.23644519e-01 -7.92142823e-02
5.37648082e-01 3.13883796e-02 3.62217486e-01 1.18220174e+00
-2.83275515e-01 4.80501622e-01 3.67011696e-01 -2.69322097e-01
-8.46762180e-01 -8.74660552e-01 -6.20073229e-02 1.36955190e+00
3.33330184e-01 -2.93665797e-01 -5.20243287e-01 -1.07041752e+00
-7.32273087e-02 6.45071626e-01 -6.18842661e-01 7.40376338e-02
-5.96314251e-01 -6.15855455e-01 7.87233412e-02 4.63559479e-01
3.36121410e-01 -1.56080067e+00 -7.72191226e-01 -4.46531326e-02
-1.93879485e-01 -1.27672803e+00 -3.18877846e-01 4.27505016e-01
-3.22711796e-01 -1.03545654e+00 -4.08156335e-01 -1.10014641e+00
7.82193780e-01 2.76366174e-01 1.20975256e+00 1.89027250e-01
-6.65323794e-01 5.36625922e-01 -1.55399367e-01 -6.94396257e-01
-3.43066484e-01 -2.78595448e-01 -1.48747101e-01 1.66277355e-03
6.60620034e-01 -1.57069579e-01 -3.31149578e-01 2.78669029e-01
-7.47501194e-01 4.15398717e-01 8.09847534e-01 9.78526354e-01
4.25316721e-01 -6.08758302e-03 -8.55858847e-02 -7.17032909e-01
-2.75183786e-02 -3.07027876e-01 -5.16267657e-01 4.51743513e-01
-2.46654078e-01 -1.00135095e-01 1.29921198e-01 -6.01828218e-01
-1.05469894e+00 4.79107499e-01 3.09035778e-01 -2.07994178e-01
-4.72151667e-01 3.31472874e-01 -4.68569070e-01 -6.52191741e-03
6.21539533e-01 9.02486071e-02 -1.09599493e-01 -4.60397638e-02
4.92152274e-01 3.60169858e-01 6.18830025e-01 -4.44305688e-01
7.98516810e-01 5.90744317e-01 -1.70745403e-01 -9.41390574e-01
-1.34691775e+00 -8.40651751e-01 -8.81378889e-01 -1.45298839e-01
1.03691566e+00 -9.44670618e-01 -7.05018401e-01 1.39262363e-01
-1.42734551e+00 -2.26134360e-01 -2.86545694e-01 4.65932757e-01
-5.64133108e-01 4.52784896e-01 -3.75845760e-01 -1.13385391e+00
-3.71925980e-02 -1.09306359e+00 1.36378837e+00 -1.04294606e-01
-3.51601779e-01 -5.86805701e-01 -2.70537108e-01 7.57437706e-01
-1.82444349e-01 -5.12153693e-02 8.51257920e-01 -7.17347920e-01
-9.15044606e-01 -1.80282503e-01 -4.83602166e-01 3.05893123e-01
-3.27888370e-01 -2.48231843e-01 -9.83641505e-01 -8.95677134e-02
1.69265717e-01 -6.26696169e-01 1.05176413e+00 1.87838346e-01
1.02025354e+00 -3.60133111e-01 -5.53873360e-01 1.65986434e-01
9.94995475e-01 -1.95022434e-01 2.69802779e-01 1.41243219e-01
9.67291832e-01 1.21772063e+00 7.72786379e-01 2.21217334e-01
5.23527741e-01 8.09406877e-01 4.68437880e-01 -4.22275215e-01
-1.47143945e-01 -2.50845641e-01 2.28456557e-01 1.23968430e-01
1.04613185e-01 -2.99417228e-01 -1.11861265e+00 7.10527003e-01
-2.17873406e+00 -8.25644970e-01 -2.55590439e-01 2.05253363e+00
6.58948362e-01 4.14857835e-01 2.95360416e-01 3.13820764e-02
4.46901739e-01 -1.50585636e-01 -3.17586958e-01 7.52449334e-02
-1.33237064e-01 -1.83800161e-01 2.71146923e-01 4.98213828e-01
-1.32929993e+00 1.04526758e+00 5.20719528e+00 3.61317992e-01
-6.23029113e-01 6.52405918e-02 5.06817222e-01 5.86394332e-02
-1.19697817e-01 2.86409587e-01 -1.14707649e+00 8.06539208e-02
4.33641404e-01 3.26416016e-01 2.84436256e-01 9.01819885e-01
-8.29573944e-02 -4.76038575e-01 -1.63970351e+00 9.70089257e-01
3.92538249e-01 -8.80685568e-01 2.57028248e-02 2.54450083e-01
3.01403999e-01 -2.73112990e-02 1.08207567e-02 3.47659141e-01
2.03401893e-01 -8.41588080e-01 9.81338322e-01 4.01603431e-01
1.10095635e-01 -3.63567948e-01 6.87465608e-01 7.87152410e-01
-9.68323827e-01 -1.90667704e-01 -1.29818380e-01 -2.03963563e-01
1.01979904e-01 4.67825294e-01 -1.01952279e+00 2.14961603e-01
7.43335247e-01 5.60892165e-01 -7.28926778e-01 7.55245268e-01
-3.71394426e-01 4.68135238e-01 -6.46776259e-01 -5.87858558e-02
3.13281357e-01 -1.55439839e-01 6.45221472e-01 1.18613172e+00
-2.99916536e-01 2.11476505e-01 6.66664422e-01 1.16251612e+00
2.32219592e-01 9.29455534e-02 -7.71812618e-01 -4.04443108e-02
1.21104844e-01 1.22537470e+00 -7.46991575e-01 -4.03773397e-01
-6.21586502e-01 1.15138412e+00 4.95375991e-01 1.97876215e-01
-7.70889699e-01 4.28400151e-02 2.18926668e-01 2.04795390e-01
2.43795216e-01 -2.40532070e-01 -2.77674973e-01 -1.07961082e+00
4.68479007e-01 -7.18291223e-01 5.27208149e-01 -6.46326244e-01
-1.46626091e+00 1.66075066e-01 1.45532683e-01 -8.62672508e-01
-2.61330903e-01 -5.95187187e-01 -3.70734215e-01 7.84842253e-01
-1.35770583e+00 -1.69723380e+00 -3.26464206e-01 4.58623439e-01
6.72044098e-01 2.56761402e-01 6.39513731e-01 -7.34203681e-02
-3.26701701e-01 3.39577109e-01 -8.22640955e-01 1.73932254e-01
5.27332664e-01 -1.21427917e+00 6.90317154e-02 8.02513123e-01
6.25856221e-01 8.44258606e-01 8.04280281e-01 -5.98902166e-01
-1.24776876e+00 -8.89018893e-01 1.32405901e+00 -1.05176091e+00
5.72639048e-01 -8.95204246e-01 -9.68937933e-01 6.50628209e-01
1.24339484e-01 4.11935747e-01 4.24347818e-01 2.67533720e-01
-4.02144104e-01 2.95960248e-01 -7.54106343e-01 6.37634873e-01
1.20481062e+00 -6.68595076e-01 -7.72569478e-01 6.62713051e-01
5.76488495e-01 -5.46135083e-02 -2.87396640e-01 6.13155425e-01
3.85534883e-01 -9.06436861e-01 1.13603556e+00 -4.56562251e-01
1.88792199e-01 -4.90001738e-01 -3.64908529e-03 -6.18007600e-01
-2.72691488e-01 -1.54761404e-01 -2.39921346e-01 1.13684011e+00
3.36683422e-01 -1.64860606e-01 6.35715902e-01 7.49543309e-01
1.59283027e-01 -6.26918852e-01 -5.87605000e-01 -7.56210864e-01
-3.91597271e-01 -3.88476700e-01 -2.84147207e-02 6.86360240e-01
1.86190903e-01 9.93905067e-01 -3.45025450e-01 4.34285820e-01
8.28795552e-01 2.22216204e-01 9.78235900e-01 -1.29289079e+00
-7.34920502e-01 -2.08815262e-01 -2.80889720e-01 -9.76357341e-01
4.02368009e-01 -9.40792859e-01 5.02291858e-01 -1.33742690e+00
7.86739171e-01 -4.44011271e-01 -7.38299042e-02 7.48756230e-01
-2.61870444e-01 4.66950715e-01 3.07535827e-01 4.24412549e-01
-9.62724984e-01 4.11656797e-01 8.47661555e-01 -2.73850888e-01
-2.73332894e-01 -1.52661353e-01 -5.06236076e-01 1.08590877e+00
2.64436841e-01 -3.12770307e-01 -4.42470640e-01 -2.55165160e-01
4.42673787e-02 -2.01701954e-01 1.03627312e+00 -8.67153704e-01
2.10937381e-01 -1.43633243e-02 2.99816161e-01 -7.14738309e-01
2.52642334e-01 -8.14897776e-01 -2.70551771e-01 2.62945116e-01
-5.28945804e-01 -5.52553356e-01 -1.12404920e-01 5.99089503e-01
-1.42874435e-01 -1.62477791e-01 7.42824674e-01 -3.07443976e-01
-8.26041937e-01 1.64790303e-01 -1.23225898e-01 -1.02172174e-01
1.11218750e+00 -9.44563076e-02 1.21939950e-01 6.25240207e-02
-8.64469111e-01 2.42444724e-01 3.92130792e-01 4.36937928e-01
4.31131303e-01 -7.37456083e-01 -4.71855491e-01 2.69161165e-01
7.03930497e-01 8.74224901e-02 -1.40357196e-01 8.80239606e-01
2.18702883e-01 5.37076890e-01 1.64782815e-02 -8.95280480e-01
-1.47499406e+00 8.81229579e-01 4.90109958e-02 -2.77624011e-01
-7.47162461e-01 1.01914513e+00 8.37471604e-01 -4.15541559e-01
7.37511933e-01 -1.07589357e-01 -1.66976869e-01 -3.25646996e-02
4.56775874e-01 -2.26833493e-01 -1.17130853e-01 -7.79056370e-01
-4.63724941e-01 1.50168911e-01 -1.94970086e-01 -2.30322227e-01
1.05790591e+00 -1.66829914e-01 -2.32816011e-01 4.49164093e-01
1.02012062e+00 -3.75463307e-01 -1.13980830e+00 -4.06522363e-01
3.52346838e-01 -2.21403256e-01 -1.24607719e-01 -7.99994767e-01
-4.42742199e-01 8.81727517e-01 4.55760002e-01 4.38882560e-02
7.01133490e-01 6.71176374e-01 4.06040847e-01 4.38363254e-01
4.12357390e-01 -8.85459006e-01 3.70235503e-01 4.46418613e-01
8.32744539e-01 -1.71123135e+00 -2.44345084e-01 -7.34492838e-01
-7.60532439e-01 6.20796740e-01 9.54275608e-01 1.58731997e-01
4.58929360e-01 1.54593661e-01 -2.66297609e-01 -3.58679295e-01
-6.25496924e-01 -6.90088689e-01 5.65647602e-01 5.26507378e-01
4.38341439e-01 -1.29636258e-01 -3.86987656e-01 3.92842174e-01
1.47527918e-01 -9.93955582e-02 3.84771079e-02 8.72335076e-01
-4.47450370e-01 -8.44940126e-01 -2.95898795e-01 2.65956700e-01
2.97430176e-02 -7.48688504e-02 -6.56408727e-01 6.91312790e-01
5.43370664e-01 8.17178130e-01 -9.25968494e-03 1.26243532e-01
2.41030753e-01 6.89086616e-02 6.32105350e-01 -9.44209278e-01
-4.12627280e-01 8.16343129e-02 -7.64414761e-03 -6.58572853e-01
-5.66928089e-01 -5.86939394e-01 -1.17649806e+00 5.23990333e-01
-4.59233075e-01 -2.06895191e-02 5.22211730e-01 1.22787452e+00
2.65383963e-02 7.11383000e-02 1.39523581e-01 -1.00168169e+00
-3.64210397e-01 -7.39360452e-01 -4.80651617e-01 8.35369349e-01
2.11977303e-01 -7.37097681e-01 -3.44994873e-01 2.13936418e-01] | [9.934235572814941, 1.4837478399276733] |
2cf643c7-e321-4282-b328-b38e4b2dc776 | towards-process-oriented-modular-and | 2205.00355 | null | https://arxiv.org/abs/2205.00355v1 | https://arxiv.org/pdf/2205.00355v1.pdf | Towards Process-Oriented, Modular, and Versatile Question Generation that Meets Educational Needs | NLP-powered automatic question generation (QG) techniques carry great pedagogical potential of saving educators' time and benefiting student learning. Yet, QG systems have not been widely adopted in classrooms to date. In this work, we aim to pinpoint key impediments and investigate how to improve the usability of automatic QG techniques for educational purposes by understanding how instructors construct questions and identifying touch points to enhance the underlying NLP models. We perform an in-depth need finding study with 11 instructors across 7 different universities, and summarize their thought processes and needs when creating questions. While instructors show great interests in using NLP systems to support question design, none of them has used such tools in practice. They resort to multiple sources of information, ranging from domain knowledge to students' misconceptions, all of which missing from today's QG systems. We argue that building effective human-NLP collaborative QG systems that emphasize instructor control and explainability is imperative for real-world adoption. We call for QG systems to provide process-oriented support, use modular design, and handle diverse sources of input. | ['Lu Wang', 'Jessica Houghton', 'Simin Fan', 'Xu Wang'] | 2022-04-30 | null | https://aclanthology.org/2022.naacl-main.22 | https://aclanthology.org/2022.naacl-main.22.pdf | naacl-2022-7 | ['misconceptions'] | ['miscellaneous'] | [-2.11287737e-01 4.58170027e-01 -5.84296398e-02 -3.54548991e-01
-1.00539148e+00 -1.11231565e+00 7.07094073e-02 5.50565064e-01
-7.26932436e-02 6.28447294e-01 4.24826473e-01 -1.28767669e+00
-5.08812845e-01 -7.15680659e-01 -3.95432860e-01 1.20954014e-01
7.09128201e-01 1.41900316e-01 2.85008609e-01 -6.02570236e-01
9.53501046e-01 6.08813524e-01 -1.81142342e+00 2.87178576e-01
1.58336961e+00 1.38705954e-01 5.99145964e-02 7.78494775e-01
-9.07444894e-01 1.46445441e+00 -1.08534157e+00 -4.17694241e-01
-1.82975948e-01 -9.23073530e-01 -1.15490830e+00 -1.34425625e-01
8.24913323e-01 -3.91579717e-01 1.42887622e-01 7.88305461e-01
6.47837222e-01 4.68085289e-01 1.53208464e-01 -1.37492204e+00
-1.12713790e+00 4.63950127e-01 1.20671853e-01 2.74837375e-01
8.49863589e-01 2.17355505e-01 9.12836969e-01 -4.70446438e-01
5.45138121e-01 1.08407283e+00 6.18887722e-01 6.27395809e-01
-9.10185397e-01 -9.17891026e-01 -1.05358645e-01 2.57250458e-01
-8.14576089e-01 -3.85499597e-01 6.30886316e-01 -7.14195609e-01
8.62767577e-01 2.63130307e-01 1.18934965e+00 5.18580139e-01
5.92705786e-01 5.23670137e-01 1.30005765e+00 -7.33856797e-01
2.33740315e-01 9.29292858e-01 3.52580398e-01 8.70862365e-01
4.52650428e-01 -5.48125207e-01 -7.73171842e-01 -2.17639402e-01
6.63982928e-01 -2.68867731e-01 -1.95253715e-01 2.70009339e-02
-7.44526327e-01 8.80469263e-01 -1.54827490e-01 4.87217367e-01
1.25346743e-02 -1.59198597e-01 -6.13764375e-02 6.27201140e-01
-6.10085540e-02 1.38788283e+00 -7.27866709e-01 -8.09501886e-01
-6.86734378e-01 5.71446955e-01 1.46929550e+00 1.19442284e+00
3.90107870e-01 -1.73058122e-01 -1.33823842e-01 7.05685914e-01
6.10458255e-01 2.95927405e-01 5.95790684e-01 -1.57353163e+00
1.56387046e-01 1.02392542e+00 2.56494284e-01 -1.00077391e+00
1.19725242e-01 -2.14093626e-01 3.83764327e-01 1.64245293e-01
7.78332949e-01 -4.43912953e-01 -5.38686395e-01 1.28154981e+00
3.04821432e-01 -4.29049164e-01 -9.06862468e-02 3.86822969e-01
1.27185273e+00 4.87838805e-01 4.75108147e-01 8.57234076e-02
1.72840643e+00 -7.03253090e-01 -1.16534197e+00 -1.05738990e-01
1.03217065e+00 -1.30667257e+00 1.50541842e+00 6.93473756e-01
-1.34011543e+00 -6.86692894e-01 -7.62734234e-01 -4.32721019e-01
-3.37844044e-01 -1.42180016e-02 2.38696650e-01 1.15286219e+00
-1.06854653e+00 4.68128711e-01 -2.72227824e-01 -6.31373286e-01
2.46371210e-01 1.48562016e-02 5.55761307e-02 -1.29493490e-01
-1.11658502e+00 7.81455040e-01 -2.99220175e-01 -2.77417362e-01
-3.62596542e-01 -1.11204183e+00 -5.94896376e-01 2.12279171e-01
5.10994613e-01 -6.91102445e-01 1.98521972e+00 -5.68105578e-01
-1.88711858e+00 4.53827977e-01 2.19807476e-02 5.16256452e-01
1.28834546e-01 -2.85688639e-01 1.11065824e-02 2.62110293e-01
2.89887071e-01 6.13461614e-01 2.58659005e-01 -9.72730637e-01
-5.67831933e-01 -4.87655774e-03 3.72613311e-01 3.83523613e-01
-4.69849706e-01 1.00446224e-01 1.13704614e-01 -2.23452121e-01
3.28532755e-01 -6.62346900e-01 -1.45796895e-01 1.34799257e-01
1.86949670e-01 -6.74073160e-01 5.38505435e-01 -7.16488898e-01
1.32153499e+00 -1.65500724e+00 -9.19511855e-01 8.35546777e-02
4.50600296e-01 4.69547361e-01 4.35730182e-02 1.20569062e+00
2.01379225e-01 6.20495617e-01 5.53207457e-01 4.33897942e-01
3.95606756e-01 -4.46452424e-02 -2.47781321e-01 -1.92597941e-01
-1.00758001e-01 8.73929739e-01 -1.43299055e+00 -7.64118075e-01
1.65960029e-01 1.17969513e-01 -7.01981127e-01 6.29170775e-01
-2.25820839e-01 3.32238041e-02 -7.50322461e-01 4.33130533e-01
2.10028529e-01 -3.30906481e-01 1.93689257e-01 5.87184906e-01
-4.84652668e-01 1.05113733e+00 -1.13370574e+00 1.44577146e+00
-7.11295903e-01 7.32447803e-01 -3.06718536e-02 -2.69977003e-01
1.12327611e+00 4.71119106e-01 2.55801648e-01 -6.22631669e-01
-4.03457060e-02 1.99939087e-01 1.78521618e-01 -1.09771037e+00
6.19426191e-01 -2.10725069e-01 3.47990125e-01 9.73022461e-01
2.14382723e-01 -7.33205378e-01 3.32593173e-01 5.29009402e-01
9.96827006e-01 4.29440290e-01 3.39651108e-01 -3.95289540e-01
1.51943117e-01 2.46869296e-01 2.41112590e-01 9.81407523e-01
-5.70859313e-01 1.02736108e-01 5.40651858e-01 6.62325472e-02
-4.43710864e-01 -6.46160185e-01 3.35919946e-01 1.30807531e+00
-3.14072967e-01 -7.55955100e-01 -7.10256279e-01 -7.27673888e-01
-2.55689561e-01 1.28463519e+00 7.82424137e-02 2.04336941e-02
-1.22392714e-01 1.80050820e-01 3.29435825e-01 3.66363466e-01
1.92187905e-01 -1.36217356e+00 -8.57499361e-01 7.03702509e-01
-3.50914568e-01 -7.67344594e-01 -3.82470280e-01 -1.20133519e-01
-9.23658252e-01 -9.72103596e-01 -1.30758643e-01 -8.81082952e-01
7.95234025e-01 6.66024625e-01 1.38265681e+00 4.16758239e-01
-1.78992152e-01 1.01691103e+00 -5.37136555e-01 -6.85325861e-01
-6.60094082e-01 -1.21221788e-01 -6.19212270e-01 -1.12226796e+00
8.23340535e-01 -4.84342724e-01 -5.70307553e-01 3.28688532e-01
-9.37891483e-01 1.28671095e-01 6.85497582e-01 3.16877514e-01
7.03011006e-02 -2.33734082e-02 1.10588741e+00 -1.02449155e+00
1.39198947e+00 -3.71237904e-01 -4.13738668e-01 2.82523394e-01
-9.30904150e-01 -1.03546627e-01 5.86816609e-01 -1.88663572e-01
-1.13985002e+00 -5.54970264e-01 -5.12474120e-01 2.56658047e-01
-4.54166800e-01 5.53306401e-01 -1.92494199e-01 -5.67579508e-01
9.38395560e-01 -1.67121083e-01 1.49958000e-01 -1.20759822e-01
3.36082697e-01 5.27669907e-01 -1.46224871e-01 -9.59773064e-01
6.92596972e-01 -4.41176504e-01 -4.04928952e-01 -9.93484676e-01
-9.62596178e-01 -6.44269526e-01 -2.23842263e-01 -5.75291574e-01
5.08561134e-01 -6.55354917e-01 -1.07757604e+00 -3.70977998e-01
-9.44907963e-01 -5.70747316e-01 -7.01216280e-01 6.25619829e-01
-2.95328230e-01 2.38920644e-01 -6.77161276e-01 -8.48941088e-01
-1.20243207e-01 -9.84466672e-01 2.05221742e-01 1.03746116e+00
-7.00790048e-01 -1.06620777e+00 1.54885203e-01 1.37457728e+00
5.62030733e-01 -3.94703597e-01 1.03647327e+00 -8.15503955e-01
-5.61983645e-01 -1.73219908e-02 7.65039250e-02 1.95440605e-01
2.62007982e-01 3.68995488e-01 -8.89968514e-01 2.05086708e-01
1.30858094e-01 -4.56313998e-01 -2.39566043e-01 3.36342417e-02
8.57568145e-01 -8.30476999e-01 1.98394526e-02 -2.35010654e-01
1.27754676e+00 4.92237985e-01 2.77606100e-01 4.15976912e-01
4.01442915e-01 1.21973085e+00 6.95474684e-01 1.47040546e-01
7.19393849e-01 2.95210481e-02 -2.30199695e-01 4.61810350e-01
-9.86958668e-02 -7.33988464e-01 4.43080574e-01 1.26349068e+00
3.77262950e-01 -1.90736920e-01 -1.11950600e+00 8.48913670e-01
-1.48500288e+00 -7.59642005e-01 -3.29048604e-01 1.82478368e+00
9.73848164e-01 3.89427543e-02 -1.48145735e-01 -1.22032568e-01
-4.01262864e-02 -3.70755166e-01 6.29128665e-02 -7.79468954e-01
8.03535521e-01 5.33837020e-01 -1.26892356e-02 6.35843635e-01
-4.10049707e-02 7.87477076e-01 5.94370556e+00 3.23170274e-01
-6.12661064e-01 -1.36451229e-01 2.68356889e-01 3.53398800e-01
-1.00440359e+00 2.44781420e-01 -8.97452652e-01 -5.07397503e-02
1.27810001e+00 -6.54393554e-01 -2.10350901e-02 9.05655921e-01
5.75234711e-01 -1.82418283e-02 -7.78730512e-01 4.30670172e-01
1.04487509e-01 -1.17304170e+00 -9.82935503e-02 4.15304117e-02
8.81716371e-01 -7.46462405e-01 -3.30972195e-01 5.41931987e-01
8.12689960e-01 -1.01165771e+00 4.24646467e-01 3.20147544e-01
-9.13323760e-02 -6.27092063e-01 5.45055747e-01 3.07794034e-01
-7.64023364e-01 2.59643365e-02 -2.39258498e-01 -4.29463267e-01
-1.14270940e-01 1.18563630e-01 -1.48037064e+00 1.87222496e-01
5.22844732e-01 4.38217968e-02 -7.37910926e-01 1.18430436e+00
-6.99605942e-01 9.69516575e-01 5.91519400e-02 -6.95660710e-01
1.44556060e-01 -3.26165766e-01 7.19991997e-02 1.00496900e+00
4.44014132e-01 6.27835572e-01 2.57363588e-01 7.97913492e-01
1.65454566e-01 7.12405145e-01 -6.40598595e-01 -4.55661476e-01
8.93565893e-01 1.48600233e+00 -7.72698939e-01 -1.31318629e-01
-5.15252411e-01 3.00652534e-01 -1.96421042e-01 4.28454131e-01
-3.04859549e-01 -7.87200272e-01 6.38641715e-01 7.18332291e-01
-2.08480075e-01 -4.07161295e-01 -3.08103085e-01 -6.98025763e-01
-2.92325586e-01 -1.45827830e+00 -8.50156546e-02 -1.02310538e+00
-1.21654582e+00 -2.63552275e-02 -6.46529123e-02 -9.16181684e-01
-3.57752591e-01 -5.50541937e-01 -7.06536651e-01 1.06458282e+00
-1.35269880e+00 -7.40083575e-01 -5.46247661e-01 1.15736298e-01
4.78255600e-01 3.73539090e-01 7.11783409e-01 1.57583967e-01
-1.59928456e-01 3.90406191e-01 -3.53578568e-01 -2.47980803e-01
1.25278449e+00 -1.43872797e+00 1.65078089e-01 5.63718200e-01
1.61684990e-01 1.26321054e+00 7.82162070e-01 -6.05816424e-01
-1.77512634e+00 -5.22535264e-01 1.55823481e+00 -6.52896941e-01
6.03324711e-01 4.41051871e-02 -1.17468905e+00 4.88936037e-01
5.22899210e-01 -5.80825984e-01 1.63162327e+00 -5.70183396e-02
2.65532956e-02 6.20486885e-02 -1.09134245e+00 7.71498084e-01
7.44319379e-01 -6.20242238e-01 -1.16598201e+00 3.70116830e-01
6.67141259e-01 -3.00920069e-01 -1.14930272e+00 -2.94022173e-01
6.45373285e-01 -7.93198526e-01 5.81051350e-01 -6.09775364e-01
5.25855780e-01 -2.33618796e-01 3.94009531e-01 -1.04741061e+00
-8.95286277e-02 -9.48533714e-01 3.07892978e-01 1.59595585e+00
2.98597962e-01 -6.45642817e-01 9.99923706e-01 1.27876186e+00
-5.05671322e-01 -6.74537778e-01 -1.53465807e-01 -2.38667935e-01
2.37284616e-01 -4.75676179e-01 4.76166874e-01 1.29180717e+00
3.92860234e-01 4.26590502e-01 4.29774493e-01 -1.08958304e-01
3.43443781e-01 2.39379127e-02 9.67555940e-01 -1.36996293e+00
-2.22776812e-02 -5.21850169e-01 7.82010555e-02 -1.07216358e+00
-1.44670695e-01 -5.07133543e-01 1.52305707e-01 -2.17146659e+00
-2.82181531e-01 -3.71619672e-01 2.24757537e-01 6.01778686e-01
-3.55208486e-01 -2.52503783e-01 1.53741166e-01 -1.37599126e-01
-4.90869403e-01 2.13249028e-01 1.84697115e+00 4.42443639e-01
-6.68105721e-01 -1.15301579e-01 -1.65377474e+00 6.82212353e-01
9.06787217e-01 -4.37903792e-01 -1.09791052e+00 -1.52763605e-01
6.87934637e-01 1.41894475e-01 -6.02009930e-02 -9.19079483e-01
6.09751582e-01 -8.60744119e-01 3.52495611e-01 -3.02783757e-01
-4.14488196e-01 -6.71776175e-01 -2.22504362e-01 1.07998177e-01
-4.76499915e-01 2.06253454e-01 5.98767877e-01 -1.55569568e-01
-3.42578381e-01 -8.83616269e-01 2.79120743e-01 -3.63890111e-01
-3.53586555e-01 -2.71147311e-01 -9.67132032e-01 2.57258445e-01
9.45624709e-01 -5.45219600e-01 -5.23205400e-01 -8.31879497e-01
-2.01732755e-01 4.53833520e-01 3.17611754e-01 3.58989567e-01
6.91839099e-01 -8.28955233e-01 -3.06508392e-01 1.04249358e-01
7.34102130e-02 -8.08896646e-02 2.87027985e-01 2.71893770e-01
-8.63455534e-01 7.98091650e-01 -4.38038826e-01 -5.85138351e-02
-1.17211497e+00 -3.86835076e-02 1.40658274e-01 -2.10886732e-01
-2.69149274e-01 1.00248826e+00 -3.10453892e-01 -7.52371907e-01
1.61360819e-02 -5.69106638e-01 -6.97876036e-01 4.15893048e-01
6.87692225e-01 4.71477121e-01 1.43753812e-01 2.34509334e-01
2.58306682e-01 3.04653406e-01 -5.86953610e-02 -4.44327027e-01
1.00943530e+00 -1.86811283e-01 1.95457637e-01 3.60883206e-01
5.42397141e-01 5.29628277e-01 -7.62448668e-01 2.41543904e-01
2.55132288e-01 -5.50035894e-01 -1.73562363e-01 -1.18757594e+00
-3.33481908e-01 8.59905481e-01 1.04623854e-01 5.37770629e-01
7.77780294e-01 -2.27996737e-01 5.88352144e-01 5.43918550e-01
1.10726900e-01 -1.05658078e+00 4.76662815e-01 3.42249751e-01
6.96412086e-01 -9.27160501e-01 -2.66302656e-02 -4.55691874e-01
-4.08254385e-01 1.29988611e+00 1.10774636e+00 5.32390893e-01
5.00315666e-01 -1.83311298e-01 4.92860407e-01 -3.85362357e-01
-1.08186281e+00 8.77437443e-02 1.32309929e-01 4.61199522e-01
1.32044590e+00 -1.79303318e-01 -7.78095722e-01 6.25281096e-01
-6.16335273e-01 5.53187132e-01 1.16438568e+00 1.61468744e+00
-9.41596627e-01 -1.41887820e+00 -6.19400501e-01 6.03331804e-01
-6.54682040e-01 -1.74531907e-01 -7.39464462e-01 7.08865404e-01
1.92862824e-02 1.43848681e+00 -4.35548127e-01 -6.50578588e-02
6.12580895e-01 6.40803337e-01 4.55959707e-01 -1.16427863e+00
-1.33883631e+00 -3.59751552e-01 8.32226947e-02 -2.71504968e-01
-4.05980349e-02 -4.91933018e-01 -1.04517114e+00 -3.44580829e-01
-3.93686384e-01 8.27237189e-01 8.00725996e-01 7.30808735e-01
5.25721490e-01 5.83068192e-01 7.13741258e-02 1.71024576e-01
-8.07467461e-01 -1.03090239e+00 -1.97743028e-01 -6.78396448e-02
-1.12898201e-02 -2.25484416e-01 -2.62961417e-01 1.02876551e-01] | [10.654154777526855, 7.633048057556152] |
198997fa-b0c4-4a4d-b1da-c9c5b5066460 | hyperminer-topic-taxonomy-mining-with | 2210.10625 | null | https://arxiv.org/abs/2210.10625v1 | https://arxiv.org/pdf/2210.10625v1.pdf | HyperMiner: Topic Taxonomy Mining with Hyperbolic Embedding | Embedded topic models are able to learn interpretable topics even with large and heavy-tailed vocabularies. However, they generally hold the Euclidean embedding space assumption, leading to a basic limitation in capturing hierarchical relations. To this end, we present a novel framework that introduces hyperbolic embeddings to represent words and topics. With the tree-likeness property of hyperbolic space, the underlying semantic hierarchy among words and topics can be better exploited to mine more interpretable topics. Furthermore, due to the superiority of hyperbolic geometry in representing hierarchical data, tree-structure knowledge can also be naturally injected to guide the learning of a topic hierarchy. Therefore, we further develop a regularization term based on the idea of contrastive learning to inject prior structural knowledge efficiently. Experiments on both topic taxonomy discovery and document representation demonstrate that the proposed framework achieves improved performance against existing embedded topic models. | ['Mingyuan Zhou', 'Zhibin Duan', 'Ruiying Lu', 'Bo Chen', 'Dongsheng Wang', 'Yishi Xu'] | 2022-10-16 | null | null | null | null | ['graph-structure-learning', 'topic-models'] | ['graphs', 'natural-language-processing'] | [-3.95000339e-01 6.00528419e-01 -4.56914246e-01 -4.23033595e-01
-3.48609477e-01 -4.63962466e-01 4.80918229e-01 2.45540023e-01
1.45379156e-01 2.92032540e-01 6.83677733e-01 -2.24858195e-01
-5.31103492e-01 -1.01664305e+00 -2.82159597e-01 -7.61499465e-01
-1.75271526e-01 4.42062467e-01 1.87455028e-01 -1.83810830e-01
1.38652265e-01 -1.02396555e-01 -1.25606358e+00 -7.20577314e-02
1.13262439e+00 8.33975971e-01 2.43770063e-01 1.58284307e-02
-4.96602982e-01 7.28394985e-01 -2.46251464e-01 -4.14046079e-01
-3.87560166e-02 -9.81844068e-02 -6.76991761e-01 3.11838359e-01
3.75333093e-02 -3.47745240e-01 -6.89914286e-01 9.07545507e-01
4.06616107e-02 1.37092963e-01 8.58420372e-01 -1.47446263e+00
-1.00853634e+00 1.02372849e+00 -3.69042516e-01 -6.72938898e-02
9.01340097e-02 -4.52903122e-01 1.82194149e+00 -1.14087427e+00
4.42478716e-01 1.34546959e+00 4.63345051e-01 8.93671811e-02
-1.12704182e+00 -7.01512516e-01 6.42915845e-01 3.14903468e-01
-1.66032600e+00 2.31345639e-01 1.19006920e+00 -3.43241602e-01
3.14838588e-01 8.51083100e-02 8.54601324e-01 9.70989823e-01
1.60826519e-01 9.79395568e-01 8.94007325e-01 -5.72727323e-02
2.95998365e-01 5.97804606e-01 6.34380162e-01 7.80813098e-01
6.06215000e-01 -3.39787513e-01 -5.78964293e-01 -3.95416915e-01
6.11366630e-01 2.16703951e-01 -3.80869746e-01 -8.83879125e-01
-9.16908860e-01 1.44390762e+00 7.56142735e-01 9.12589729e-02
-2.58042395e-01 -4.39434759e-02 3.43955517e-01 -8.09607729e-02
6.40170813e-01 4.50300395e-01 -2.08823010e-01 4.53975230e-01
-7.32843995e-01 3.40624034e-01 7.18585968e-01 1.23289084e+00
1.03625345e+00 -1.15739800e-01 3.06533370e-03 6.87513351e-01
9.35043275e-01 9.02001113e-02 5.37288070e-01 -5.56746125e-01
3.41427654e-01 1.03398526e+00 -1.43480852e-01 -1.45223773e+00
-4.25986439e-01 -4.06467736e-01 -8.03400338e-01 -4.22981799e-01
-1.37511492e-01 1.14083953e-01 -7.13276863e-01 1.47415745e+00
4.82613176e-01 1.33969143e-01 1.86045632e-01 8.13577890e-01
7.70777583e-01 9.78644013e-01 5.68864942e-02 -8.34317058e-02
1.87972903e+00 -7.78589487e-01 -1.08245444e+00 1.26320437e-01
5.88224351e-01 -1.52218714e-01 1.17698205e+00 1.03225395e-01
-5.79330206e-01 -2.24492982e-01 -1.12840140e+00 -5.81775784e-01
-5.71213186e-01 -2.67713755e-01 9.61863697e-01 6.09789133e-01
-6.94979310e-01 -1.33135647e-01 -8.17441881e-01 -9.65547487e-02
3.51073086e-01 1.50945038e-01 -1.30935786e-02 -7.79318139e-02
-1.45087457e+00 3.29806030e-01 9.09101903e-01 -1.06935591e-01
-7.13558614e-01 -9.17608559e-01 -1.08775246e+00 2.96490461e-01
4.16808546e-01 -6.33647323e-01 9.03434753e-01 -1.32177293e-01
-1.26739824e+00 3.58183950e-01 -1.81401879e-01 -4.95681047e-01
1.19475976e-01 -2.41463766e-01 -1.13790505e-01 3.84874433e-01
1.81266829e-01 7.74104118e-01 8.82027745e-01 -1.31917000e+00
-6.61187291e-01 -5.99686146e-01 3.49577874e-01 3.34982038e-01
-1.16905761e+00 -2.96055317e-01 -3.25996101e-01 -8.86010289e-01
5.50553679e-01 -9.07409847e-01 -1.09625660e-01 8.00939873e-02
-5.08993208e-01 -8.87898088e-01 1.28469050e+00 -4.59153205e-01
1.41876006e+00 -2.09082770e+00 3.98597717e-01 1.14093855e-01
7.12225318e-01 -3.70651960e-01 2.41715252e-01 4.79503363e-01
1.87929198e-01 3.29898715e-01 -1.36103049e-01 -2.52634406e-01
2.83177465e-01 3.48013014e-01 -9.32942152e-01 3.47284019e-01
2.54063364e-02 7.74246395e-01 -8.60084295e-01 -6.27371430e-01
1.21887950e-02 5.08034885e-01 -7.56633759e-01 1.45736203e-01
-3.67526889e-01 -4.81905453e-02 -9.38374162e-01 3.25953960e-01
6.23213291e-01 -5.56953609e-01 1.61112204e-01 -1.35375068e-01
7.10896775e-02 5.24001539e-01 -1.27357376e+00 1.59403872e+00
-6.67609572e-01 7.33670652e-01 -1.52760863e-01 -9.47599292e-01
1.10080945e+00 5.79082906e-01 4.11820650e-01 3.07311844e-02
-5.10338061e-02 -1.18556254e-01 -2.28732929e-01 -1.92549899e-01
7.91428864e-01 -1.58200845e-01 -5.00056706e-02 5.62309206e-01
-3.19850817e-02 -2.47756019e-01 -1.95136160e-01 6.29120946e-01
5.74574292e-01 -4.20971453e-01 2.88642555e-01 -6.13784134e-01
3.17221075e-01 -1.02503210e-01 5.68042934e-01 3.02580625e-01
9.64150503e-02 3.50062549e-01 6.34350598e-01 -4.34558868e-01
-1.02370644e+00 -1.16067803e+00 -5.77894151e-01 1.23186135e+00
4.91566688e-01 -8.92718017e-01 -2.91910261e-01 -6.91231549e-01
-1.20663844e-01 8.67547691e-01 -8.43443573e-01 -2.17830673e-01
-2.51946658e-01 -5.96758068e-01 1.90580174e-01 7.25029290e-01
3.85744929e-01 -5.82075298e-01 -3.32694739e-01 9.63124260e-02
-2.51245141e-01 -1.00978744e+00 -5.02304494e-01 1.18845619e-01
-1.12245047e+00 -8.38195980e-01 -6.70142710e-01 -8.90827775e-01
6.71204805e-01 5.35540223e-01 7.04421997e-01 -1.51143491e-01
-4.98364121e-02 2.41126701e-01 -5.00767946e-01 -4.87711072e-01
3.25374007e-02 4.23822612e-01 4.54748124e-02 -1.89709336e-01
5.74171364e-01 -5.89685082e-01 -6.96595907e-01 2.90809184e-01
-1.18740118e+00 1.33812129e-01 3.91765058e-01 8.71414363e-01
2.01153159e-01 6.20518684e-01 5.12975454e-01 -8.53193581e-01
7.56150901e-01 -7.25812972e-01 -6.14237964e-01 1.74059510e-01
-8.27598214e-01 3.10149491e-01 6.00807190e-01 -4.58725959e-01
-1.16965747e+00 -4.19278532e-01 3.21881413e-01 -3.67912948e-01
5.05660921e-02 6.17937744e-01 -2.96981603e-01 4.28675532e-01
1.76253855e-01 3.33554596e-01 -2.69095898e-01 -4.12418544e-01
8.55992675e-01 8.20875764e-01 -2.01801360e-02 -5.91888964e-01
9.75776672e-01 8.88957977e-01 -2.54346430e-01 -1.00016987e+00
-1.10971642e+00 -7.52557278e-01 -5.41590154e-01 3.60088199e-01
9.69445050e-01 -1.18379056e+00 -3.35871786e-01 -1.21540703e-01
-1.19895053e+00 3.04221541e-01 -1.82516947e-01 5.58735847e-01
-2.91651756e-01 4.15343851e-01 -6.04772985e-01 -7.26665616e-01
-3.14648658e-01 -9.72677350e-01 1.31574404e+00 1.72571540e-01
-2.25844666e-01 -1.59826994e+00 1.32647842e-01 2.65022039e-01
1.00092962e-01 -1.71768412e-01 1.33166003e+00 -7.70915747e-01
-8.57569396e-01 -4.82851639e-02 -3.41672003e-01 -4.64087985e-02
3.12066108e-01 -2.27127969e-01 -8.71340573e-01 -2.56114483e-01
2.15067670e-01 -2.13510782e-01 8.20060968e-01 1.36600748e-01
1.10500872e+00 -6.35792732e-01 -5.06146193e-01 5.85966706e-01
1.25715959e+00 -2.12436900e-01 4.09449279e-01 3.80940676e-01
7.32614160e-01 7.83392906e-01 5.37531495e-01 6.46421254e-01
8.36009383e-01 6.17578626e-01 1.67176053e-01 1.30325228e-01
3.10329765e-01 -6.24497712e-01 1.75727844e-01 1.22599423e+00
3.15905958e-01 -1.36664256e-01 -8.88325274e-01 6.97984636e-01
-1.58434415e+00 -6.65542781e-01 -3.86885814e-02 1.66892612e+00
8.56259346e-01 2.50866767e-02 3.49679440e-02 9.15127993e-02
5.44743180e-01 3.89506489e-01 -3.60798538e-01 -4.92315441e-02
2.69972440e-02 -3.96269023e-01 5.05154170e-02 3.55184704e-01
-1.11181653e+00 9.57215965e-01 5.98418760e+00 8.40411782e-01
-8.48825872e-01 -3.30023980e-03 2.23803416e-01 4.38132703e-01
-8.30303967e-01 1.57246441e-01 -1.09663820e+00 3.25845927e-01
5.13695419e-01 -7.73881495e-01 -2.22558469e-01 1.21208239e+00
8.44892263e-02 5.20996988e-01 -9.08256352e-01 6.15224600e-01
4.23884057e-02 -1.11431515e+00 5.13568521e-01 4.13850814e-01
8.51040483e-01 -5.23023129e-01 3.71933073e-01 4.19727594e-01
4.42220241e-01 -8.75542581e-01 3.97123724e-01 -3.29505019e-02
3.39297265e-01 -7.18701959e-01 6.44021273e-01 2.96516806e-01
-1.54511881e+00 -1.70843646e-01 -1.01305747e+00 6.98337331e-02
1.27494320e-01 4.75448161e-01 -9.64135647e-01 6.64293706e-01
6.00034654e-01 8.51258934e-01 -4.05888230e-01 8.49417746e-01
-5.07297456e-01 6.95718348e-01 -3.05506080e-01 -1.12712972e-01
5.13167202e-01 -3.87917131e-01 5.71976006e-01 9.18179452e-01
1.44072622e-01 3.81650656e-01 3.21682841e-01 9.46041644e-01
-1.38095185e-01 4.56931084e-01 -7.29039371e-01 -3.01946234e-02
6.11953318e-01 1.31074440e+00 -7.75790572e-01 -3.14562619e-01
-6.33793354e-01 5.91285050e-01 4.16254342e-01 4.33019519e-01
-7.42487729e-01 -4.16132808e-01 6.96439445e-01 7.07440153e-02
4.53900933e-01 -5.25681674e-01 -3.55425507e-01 -1.25100768e+00
-7.08550811e-02 -6.23281062e-01 4.71128732e-01 -4.27917212e-01
-1.27755547e+00 5.67972660e-01 4.37969565e-01 -1.04449689e+00
5.05565060e-03 -5.62762260e-01 -6.01826966e-01 4.76930082e-01
-1.46039915e+00 -1.31353927e+00 -3.45139384e-01 4.57458138e-01
8.02907348e-01 7.15526864e-02 7.07034469e-01 -2.51716644e-01
-3.80389154e-01 3.82010132e-01 2.40610436e-01 1.11280993e-01
3.80209267e-01 -1.57635677e+00 1.14434017e-02 3.11110467e-01
4.10660952e-01 1.03391147e+00 7.07364738e-01 -5.44050992e-01
-1.25762331e+00 -1.27543569e+00 6.16194546e-01 -4.74423796e-01
1.11000824e+00 -7.89866328e-01 -1.31389737e+00 9.21847522e-01
1.91103563e-01 -4.81207013e-01 1.16380310e+00 6.75553083e-01
-6.65822029e-01 -2.15993430e-02 -5.20226300e-01 7.15828836e-01
7.35454738e-01 -7.30663478e-01 -1.22076774e+00 4.89096135e-01
1.42269611e+00 5.43066859e-02 -1.03742099e+00 2.63885677e-01
3.89655977e-01 -4.90509659e-01 8.66085291e-01 -5.90173006e-01
4.08125401e-01 -9.52508152e-02 -2.20427930e-01 -1.21151376e+00
-3.67877036e-01 -4.60322499e-01 -4.99365836e-01 1.34350145e+00
2.99779594e-01 -5.58401287e-01 1.01086926e+00 5.77493668e-01
9.58204195e-02 -7.29314446e-01 -7.34321535e-01 -7.95667410e-01
3.89768124e-01 -3.23239267e-01 5.64185500e-01 9.70546722e-01
3.57569695e-01 6.81286812e-01 -2.52809227e-01 5.85583866e-01
8.09695244e-01 3.82364482e-01 5.52139521e-01 -1.56886840e+00
1.20069750e-01 -1.72252193e-01 -5.82985520e-01 -1.49239051e+00
5.12418866e-01 -8.27189267e-01 -7.54650384e-02 -1.48342383e+00
3.64294440e-01 -7.91356623e-01 -2.07996503e-01 1.69633403e-01
-4.98084217e-01 -9.29454640e-02 2.27568913e-02 4.67990339e-01
-6.62991643e-01 1.34008706e+00 1.09536731e+00 -9.10471380e-02
-2.62072980e-01 -9.53637734e-02 -8.71954083e-01 7.90063262e-01
6.61636472e-01 -4.78753120e-01 -9.41614270e-01 -4.10236180e-01
2.25729957e-01 -2.20769078e-01 2.00108707e-01 -5.46603322e-01
4.48276490e-01 5.22042662e-02 -1.37927011e-01 -8.63352120e-01
4.42992002e-01 -9.88657594e-01 -3.76979172e-01 1.14463553e-01
-5.91128588e-01 -1.84532434e-01 -8.35033208e-02 1.13595366e+00
-3.84792686e-01 -1.04832239e-01 3.39627147e-01 2.78273374e-01
-5.21880507e-01 5.55088043e-01 -3.52128118e-01 1.36635840e-01
1.00629687e+00 -7.39674494e-02 -4.38441858e-02 -5.63225687e-01
-5.13890862e-01 5.47171116e-01 3.78069907e-01 6.28926814e-01
7.43520677e-01 -1.44920743e+00 -5.37351072e-01 -5.31065427e-02
3.62222224e-01 3.25327426e-01 1.38122812e-01 3.88112903e-01
-1.09954566e-01 9.85124826e-01 3.40300590e-01 -6.88492417e-01
-9.34076905e-01 9.67491686e-01 -1.86777920e-01 -2.07532659e-01
-1.00462091e+00 5.83210051e-01 1.08878148e+00 -4.21662092e-01
4.04040039e-01 -5.40990770e-01 -3.84822041e-01 2.35782042e-01
5.30931532e-01 3.16733330e-01 -3.94921869e-01 -4.57990170e-01
-1.23716742e-01 5.37815154e-01 -4.30146068e-01 -8.84809941e-02
1.08929944e+00 -4.21129197e-01 -1.65035397e-01 7.10500479e-01
1.37470937e+00 -8.13131779e-02 -9.61884081e-01 -5.74344039e-01
3.38494688e-01 -3.22633147e-01 2.70200759e-01 1.01808645e-01
-6.51494741e-01 1.10731590e+00 2.04676300e-01 6.82505429e-01
7.79298484e-01 4.51709211e-01 6.21943891e-01 5.29010296e-01
8.77280161e-02 -7.85792887e-01 3.65778267e-01 3.09713453e-01
8.11839461e-01 -1.05169761e+00 1.82823420e-01 -9.60371315e-01
-5.19090176e-01 1.13951337e+00 5.41136384e-01 -2.85036117e-02
9.60071743e-01 -1.26103535e-01 -1.66510329e-01 -5.91685891e-01
-8.43363464e-01 3.51013914e-02 5.13405204e-01 4.95406926e-01
4.63995725e-01 1.15383603e-02 -1.10497832e-01 7.76403904e-01
-5.05420327e-01 -5.57595730e-01 4.63597149e-01 3.98454964e-01
-6.80203915e-01 -7.41897404e-01 -2.87944674e-01 1.32775888e-01
-2.00798497e-01 -1.30492568e-01 -2.35134944e-01 8.88667643e-01
-2.04041854e-01 8.53146851e-01 1.99252367e-01 5.97346528e-03
-1.37965173e-01 1.31430207e-02 -2.23500192e-01 -8.12126160e-01
1.69839129e-01 9.10016075e-02 -4.55164492e-01 -1.91312283e-01
-8.94925669e-02 -5.02088428e-01 -1.23463559e+00 4.91321785e-04
-7.99847543e-01 7.43304968e-01 5.52590311e-01 8.74505758e-01
2.56720066e-01 4.59754944e-01 7.15057194e-01 -1.67108014e-01
-6.28873706e-01 -9.80999053e-01 -9.22215998e-01 1.96871385e-01
1.81696177e-01 -9.29567754e-01 -5.31358659e-01 -5.44334501e-02] | [10.37403678894043, 6.942378520965576] |
9a7774fb-95de-46c5-9bbb-bed077384321 | cate-computation-aware-neural-architecture | 2102.07108 | null | https://arxiv.org/abs/2102.07108v2 | https://arxiv.org/pdf/2102.07108v2.pdf | CATE: Computation-aware Neural Architecture Encoding with Transformers | Recent works (White et al., 2020a; Yan et al., 2020) demonstrate the importance of architecture encodings in Neural Architecture Search (NAS). These encodings encode either structure or computation information of the neural architectures. Compared to structure-aware encodings, computation-aware encodings map architectures with similar accuracies to the same region, which improves the downstream architecture search performance (Zhang et al., 2019; White et al., 2020a). In this work, we introduce a Computation-Aware Transformer-based Encoding method called CATE. Different from existing computation-aware encodings based on fixed transformation (e.g. path encoding), CATE employs a pairwise pre-training scheme to learn computation-aware encodings using Transformers with cross-attention. Such learned encodings contain dense and contextualized computation information of neural architectures. We compare CATE with eleven encodings under three major encoding-dependent NAS subroutines in both small and large search spaces. Our experiments show that CATE is beneficial to the downstream search, especially in the large search space. Moreover, the outside search space experiment demonstrates its superior generalization ability beyond the search space on which it was trained. Our code is available at: https://github.com/MSU-MLSys-Lab/CATE. | ['Mi Zhang', 'Fei Liu', 'Kaiqiang Song', 'Shen Yan'] | 2021-02-14 | null | null | null | null | ['unsupervised-pre-training'] | ['methodology'] | [ 8.34903028e-03 -1.93304658e-01 -1.22439183e-01 -2.47881889e-01
-6.21783733e-01 -8.77938211e-01 5.95551312e-01 -1.13939382e-01
-4.63830531e-01 2.27083519e-01 2.63491631e-01 -5.18990099e-01
-1.84089720e-01 -1.03050315e+00 -9.04950202e-01 -5.00096440e-01
1.25081673e-01 3.97181392e-01 3.01931649e-01 -4.74561363e-01
5.10010123e-01 4.76712763e-01 -1.56743646e+00 5.72316587e-01
7.37919450e-01 1.20445955e+00 6.19544566e-01 5.46112299e-01
-1.58446684e-01 5.52618861e-01 -2.90100306e-01 -3.24540794e-01
4.79765356e-01 -4.09424216e-01 -1.13985777e+00 -7.41533220e-01
5.00592411e-01 -7.21790716e-02 -5.58160841e-01 9.39526737e-01
4.20639813e-01 3.07922572e-01 4.54825222e-01 -9.74098623e-01
-1.13082731e+00 7.88649082e-01 -5.31693287e-02 5.16755283e-01
-9.85550135e-02 5.24194002e-01 1.39775848e+00 -1.20895302e+00
4.41128105e-01 1.13247550e+00 5.96762359e-01 6.12406015e-01
-1.22601724e+00 -9.00701463e-01 3.11777264e-01 4.96497095e-01
-1.61211169e+00 -4.76409346e-01 7.47715235e-01 -3.10465902e-01
1.55435216e+00 2.25435048e-01 7.95103431e-01 9.16111529e-01
1.60621360e-01 1.10378230e+00 7.04245269e-01 -3.21730405e-01
2.90840864e-01 -4.41255063e-01 3.17367673e-01 8.46313179e-01
-7.05399588e-02 4.43430305e-01 -6.45011365e-01 -1.35866394e-02
9.08665717e-01 -2.87687592e-02 -2.46802703e-01 -2.41976574e-01
-1.18487990e+00 7.21629560e-01 1.09530973e+00 4.43894297e-01
-1.88467085e-01 5.74512661e-01 3.81215453e-01 4.93528813e-01
1.20101618e-02 9.34775472e-01 -5.75775385e-01 -1.09139040e-01
-8.19456577e-01 2.27558732e-01 3.55089426e-01 1.18424618e+00
7.81901181e-01 2.27772161e-01 -3.48639101e-01 7.60257125e-01
2.59692699e-01 1.33002639e-01 1.05487931e+00 -7.67486930e-01
6.67980254e-01 7.57744431e-01 -4.66798633e-01 -7.87677407e-01
-3.88359964e-01 -8.50546062e-01 -7.96563804e-01 -2.35243857e-01
1.26347288e-01 2.77910471e-01 -9.93971646e-01 2.11631441e+00
-1.52946681e-01 6.55961186e-02 2.35981867e-02 9.41077232e-01
7.24576056e-01 7.31009662e-01 -2.56676953e-02 1.83081880e-01
1.33637369e+00 -1.67937636e+00 -3.96293014e-01 -4.90497619e-01
9.71723914e-01 -4.77862954e-01 1.44903743e+00 1.31334454e-01
-1.32633138e+00 -7.86251962e-01 -1.00414920e+00 -3.71066421e-01
-5.23391902e-01 1.80661619e-01 7.74422050e-01 4.53124076e-01
-1.65324318e+00 5.07023275e-01 -8.85378063e-01 -2.88337469e-01
4.55734700e-01 4.25753057e-01 -2.64687333e-02 4.38017249e-02
-1.42989779e+00 8.32649291e-01 6.30755067e-01 2.54521612e-03
-1.08278799e+00 -6.94124460e-01 -7.89818108e-01 4.91126567e-01
2.00064600e-01 -8.45521331e-01 1.32901347e+00 -8.83624196e-01
-1.32678390e+00 6.61019325e-01 -3.70267600e-01 -5.28569639e-01
-2.37752929e-01 -2.06448138e-01 -1.56800091e-01 -9.44095105e-02
-1.77337840e-01 8.72910559e-01 5.12348413e-01 -7.76080430e-01
-1.60310104e-01 -5.23690283e-01 2.08976448e-01 3.53239328e-01
-4.78012413e-01 -4.43057343e-02 -6.62426114e-01 -9.52448368e-01
2.98008651e-01 -1.01682353e+00 -1.19494908e-02 -8.25786144e-02
-2.97528535e-01 -2.91415483e-01 3.23253274e-01 -3.96685302e-01
1.62894952e+00 -2.07381558e+00 5.28583765e-01 7.30754510e-02
2.21290946e-01 2.45293662e-01 -5.98519981e-01 5.66185594e-01
-8.57447311e-02 2.71593094e-01 -4.87423182e-01 -1.89160436e-01
1.11563556e-01 7.81615898e-02 -4.24250990e-01 1.94947254e-02
1.96214706e-01 1.48909819e+00 -7.40085542e-01 -1.79953158e-01
-3.08165938e-01 4.20006774e-02 -8.74273479e-01 6.40430897e-02
-2.08118856e-01 1.71111703e-01 -3.60372484e-01 7.67343760e-01
3.14846307e-01 -5.55676103e-01 8.35404843e-02 -2.55438596e-01
-1.47160172e-01 6.49203658e-01 -6.80043042e-01 2.14160872e+00
-7.46309161e-01 6.52708709e-01 -2.14931533e-01 -1.00079477e+00
8.52400720e-01 1.14204578e-01 -2.57363450e-02 -1.06832778e+00
5.17042987e-02 4.10839707e-01 4.43326056e-01 4.44841012e-02
4.82373267e-01 3.38963151e-01 -9.33329314e-02 7.52551496e-01
2.63130635e-01 -1.37471501e-02 1.39077216e-01 2.94460863e-01
1.31288004e+00 1.44132987e-01 2.86050498e-01 -5.37457407e-01
4.72325355e-01 -1.72259837e-01 3.20098281e-01 6.78844750e-01
-1.97313488e-01 5.27646244e-01 9.52083021e-02 -6.17026985e-01
-8.92799377e-01 -9.29527581e-01 -1.60935462e-01 1.49264240e+00
1.45532250e-01 -8.55985463e-01 -7.76829541e-01 -3.57292354e-01
-2.12650165e-01 6.05448604e-01 -8.69448483e-01 -6.70635164e-01
-9.75158691e-01 -4.46903080e-01 9.01228309e-01 8.14462900e-01
7.32131183e-01 -1.15627766e+00 -8.48796785e-01 4.21236008e-02
-9.92258415e-02 -5.29803216e-01 -8.22679520e-01 4.64570254e-01
-1.10441494e+00 -7.79942155e-01 -6.12225413e-01 -9.66338456e-01
5.85708797e-01 1.02686033e-01 1.18279433e+00 3.65019262e-01
-1.62460908e-01 1.36844903e-01 -2.71349967e-01 1.16911575e-01
1.19701885e-01 5.78988075e-01 -1.40316606e-01 -3.20438892e-01
2.31995568e-01 -6.22416615e-01 -7.65099645e-01 3.84547949e-01
-6.68208718e-01 2.41527364e-01 7.88658679e-01 1.02686906e+00
7.27541506e-01 -2.98779130e-01 2.39452809e-01 -5.66821337e-01
7.52667487e-01 -5.30063391e-01 -6.55967653e-01 3.99688452e-01
-1.04434526e+00 6.21897876e-01 6.55215502e-01 -3.46444190e-01
-7.52744794e-01 -8.79082680e-02 -1.72807872e-01 -4.62613225e-01
5.59560321e-02 5.93164802e-01 -3.86366248e-02 -1.42062634e-01
6.73474252e-01 7.73026049e-01 -3.52311581e-01 -6.78689659e-01
2.42341593e-01 1.00708872e-01 2.45661095e-01 -9.37525153e-01
5.41703522e-01 -1.91347271e-01 -1.75888747e-01 -3.75381857e-02
-5.11039317e-01 -2.09796280e-01 -5.44514120e-01 2.59738714e-01
6.81418300e-01 -6.40895545e-01 -4.49918717e-01 3.54954571e-01
-1.21030378e+00 -7.32237518e-01 -1.10612534e-01 2.91502714e-01
-6.37581229e-01 -1.14643514e-01 -8.60810161e-01 -3.48936647e-01
-4.88492191e-01 -1.65310085e+00 9.57998812e-01 4.16814350e-03
-2.26686999e-01 -8.77945602e-01 4.91036177e-02 6.63973987e-02
7.67194211e-01 -3.45394671e-01 1.20542955e+00 -5.99500716e-01
-9.07837391e-01 4.11285907e-01 -3.01474631e-01 -1.06520951e-01
-2.49829754e-01 -3.88272375e-01 -9.18507993e-01 -2.66228169e-01
3.59191634e-02 -4.74980026e-01 1.20330667e+00 2.01403618e-01
1.76813281e+00 -5.53872883e-01 -3.61142337e-01 1.21235311e+00
1.37390542e+00 4.35586512e-01 7.18447387e-01 7.01100409e-01
4.72370565e-01 2.82480776e-01 2.80788898e-01 1.86435487e-02
2.85274208e-01 9.30792272e-01 5.30267000e-01 4.20755029e-01
-3.01578879e-01 -3.63923460e-01 4.00106937e-01 1.09697235e+00
-7.54984245e-02 -2.69622535e-01 -1.27229679e+00 6.33001387e-01
-1.88932204e+00 -7.96948314e-01 2.19812408e-01 1.85934627e+00
1.20421040e+00 -1.08447723e-01 -2.29485750e-01 -1.35086417e-01
4.47596192e-01 1.34825587e-01 -8.20135355e-01 -6.05705202e-01
-9.97255668e-02 4.62374687e-01 3.44704300e-01 5.13743103e-01
-7.85837710e-01 1.25779784e+00 5.76226664e+00 9.96797442e-01
-9.11620617e-01 3.07375073e-01 5.70771575e-01 -2.92904109e-01
-6.01286948e-01 1.42274201e-02 -8.54562104e-01 5.39586842e-01
1.18080533e+00 -6.22559302e-02 8.05042803e-01 7.92920828e-01
-6.43054962e-01 3.28020692e-01 -1.37456298e+00 1.03541660e+00
-4.71859276e-02 -1.69832397e+00 2.93398291e-01 6.74655661e-03
7.34441817e-01 4.04570967e-01 3.84616375e-01 6.54586792e-01
2.89154232e-01 -1.30431366e+00 9.07714427e-01 3.46948117e-01
7.91337073e-01 -5.22296071e-01 4.60464060e-01 2.98085809e-01
-1.53020990e+00 -3.34376931e-01 -3.41308475e-01 -1.48825645e-01
-2.26359237e-02 3.95460911e-02 -3.59474123e-01 1.27909586e-01
7.10934281e-01 6.24458432e-01 -9.01737571e-01 7.98247635e-01
-4.23565686e-01 5.66806972e-01 -1.58706471e-01 -1.67743191e-01
5.01395881e-01 -2.72068586e-02 2.51840621e-01 1.12870240e+00
4.65291113e-01 2.71335870e-01 -1.47186041e-01 1.24080992e+00
-2.83198386e-01 -9.19815153e-02 -4.53612506e-01 -1.42533496e-01
7.98936188e-01 8.26839268e-01 -5.43208003e-01 -2.01741025e-01
-1.02052741e-01 1.00975263e+00 7.82465339e-01 5.26618838e-01
-8.26535881e-01 -3.73696327e-01 6.33510709e-01 -6.75367489e-02
3.92195314e-01 -1.70247972e-01 -6.36965454e-01 -1.03910768e+00
-4.54621501e-02 -8.55067253e-01 2.60615438e-01 -7.55412757e-01
-7.93593228e-01 1.07804596e+00 3.00050974e-02 -9.23221767e-01
-3.22802901e-01 -6.65035188e-01 -6.11562550e-01 8.85642588e-01
-1.19311571e+00 -1.10339737e+00 -4.55718078e-02 6.29327595e-01
8.44835162e-01 -6.06472790e-01 9.85958695e-01 2.17893451e-01
-5.44224560e-01 1.11768603e+00 -3.15284468e-02 1.46692231e-01
2.62059301e-01 -1.02193344e+00 9.49409664e-01 7.50849009e-01
4.09988552e-01 1.37135375e+00 6.26121461e-03 -3.83485913e-01
-1.69241416e+00 -1.02237892e+00 7.86946535e-01 -5.00664711e-01
6.39403820e-01 -4.98820633e-01 -1.00786340e+00 9.28764760e-01
4.13376570e-01 1.42099038e-01 6.13348246e-01 3.59111160e-01
-7.92975307e-01 -4.95721102e-02 -6.54126108e-01 6.43173873e-01
1.59329915e+00 -7.46096075e-01 -6.12117887e-01 8.87493566e-02
1.14884949e+00 -4.23984945e-01 -7.94956028e-01 4.43653077e-01
6.16927326e-01 -8.85310054e-01 1.07465875e+00 -6.17112398e-01
4.12091792e-01 -1.07564367e-01 -3.80923152e-01 -1.40399778e+00
-7.36955464e-01 -2.57314682e-01 -4.44238245e-01 7.59365141e-01
7.49947786e-01 -8.20479095e-01 4.43928689e-01 4.12125260e-01
-6.89694345e-01 -1.27110684e+00 -8.41282904e-01 -9.81335521e-01
4.25764591e-01 -3.95500392e-01 9.44432855e-01 8.92725050e-01
-1.37732297e-01 3.99046212e-01 2.40869284e-01 -1.06664546e-01
1.99799448e-01 4.78815019e-01 2.80823052e-01 -9.01303172e-01
-6.03807211e-01 -1.14349592e+00 -1.30438775e-01 -1.31242979e+00
2.69963056e-01 -1.42118359e+00 -1.33722678e-01 -1.42869425e+00
1.74129933e-01 -7.14492977e-01 -6.58719122e-01 8.02875400e-01
5.10006621e-02 1.31492153e-01 2.71391213e-01 6.99266195e-01
-4.24673975e-01 7.60121226e-01 1.22472727e+00 -2.58786052e-01
6.29552901e-02 -2.92085230e-01 -7.24823236e-01 3.83082747e-01
1.21843374e+00 -1.99843347e-01 -7.83614755e-01 -1.14305556e+00
4.03330445e-01 -3.83821368e-01 3.07464898e-01 -1.06802261e+00
6.23117507e-01 -2.86398567e-02 1.36240035e-01 -5.02801597e-01
3.90135646e-01 -5.31786680e-01 5.96373603e-02 7.72041500e-01
-9.00045097e-01 5.71296155e-01 6.11736178e-01 3.45708698e-01
-3.66583705e-01 -3.71089965e-01 5.09850144e-01 -3.74525815e-01
-1.04531097e+00 3.84691119e-01 -1.19834833e-01 1.86335742e-01
7.18078792e-01 -2.55888313e-01 -6.47853076e-01 -3.15105394e-02
-5.92467904e-01 2.08246976e-01 3.73674482e-01 3.43979686e-01
7.37715006e-01 -1.53529012e+00 -3.19588363e-01 4.45743710e-01
3.72337282e-01 9.72624347e-02 1.09411925e-01 7.80193627e-01
-4.22983378e-01 9.28241491e-01 -3.02180082e-01 -2.87110478e-01
-7.80875504e-01 7.22032964e-01 2.67714620e-01 -2.96758413e-01
-1.80838004e-01 1.45593691e+00 6.74384236e-01 -6.25211358e-01
3.14556688e-01 -5.11019647e-01 1.00690521e-01 -3.09292346e-01
5.86575389e-01 1.03948608e-01 1.93037391e-01 -3.82565141e-01
-4.73862141e-01 6.37801766e-01 -1.95436791e-01 -2.59452686e-02
1.15860426e+00 1.02373950e-01 -2.08668113e-01 3.38887602e-01
1.35590124e+00 -4.48542625e-01 -9.27831590e-01 -4.08394814e-01
3.40467766e-02 -2.31894702e-01 3.03332925e-01 -9.74828005e-01
-1.29875684e+00 1.29781997e+00 5.55515528e-01 -1.53889790e-01
1.42046225e+00 7.23441094e-02 8.27791750e-01 6.51205897e-01
5.16956866e-01 -8.92388940e-01 9.71090198e-02 9.53923345e-01
1.31320572e+00 -9.24311996e-01 -4.24350858e-01 -9.52719245e-03
-4.54923451e-01 1.03421152e+00 1.17923105e+00 8.87039490e-03
5.19853890e-01 2.00187191e-01 -5.67722738e-01 -3.04268777e-01
-1.26514530e+00 -2.90044069e-01 5.10818839e-01 3.21366042e-01
4.74883139e-01 -1.04887687e-01 -2.21009791e-01 7.30103970e-01
-3.19569677e-01 -3.45379472e-01 -1.46555901e-01 8.48208249e-01
-4.24819559e-01 -1.05545115e+00 -6.03734031e-02 4.18067217e-01
1.81952730e-01 -8.62489164e-01 -5.50955474e-01 5.23985744e-01
-2.82126721e-02 2.99919575e-01 2.54732788e-01 -6.38780355e-01
2.96594873e-02 4.28954124e-01 4.55370605e-01 -7.77119637e-01
-9.93899822e-01 -1.97012022e-01 -5.99871241e-02 -1.00129640e+00
1.14009120e-01 -4.72223222e-01 -1.35613704e+00 -1.28982410e-01
-2.11805642e-01 6.25587851e-02 3.91466022e-01 5.99345922e-01
9.62763667e-01 6.66651011e-01 2.22149640e-01 -5.72087944e-01
-6.83589399e-01 -9.32190597e-01 -8.95825997e-02 6.66554645e-02
1.87536091e-01 -6.90207958e-01 -9.36987400e-02 -1.47845864e-01] | [8.778410911560059, 3.3667538166046143] |
b659532f-e2f7-4950-82f4-b99797e5a522 | 190600852 | 1906.00852 | null | https://arxiv.org/abs/1906.00852v1 | https://arxiv.org/pdf/1906.00852v1.pdf | Hierarchical Auxiliary Learning | Conventional application of convolutional neural networks (CNNs) for image classification and recognition is based on the assumption that all target classes are equal(i.e., no hierarchy) and exclusive of one another (i.e., no overlap). CNN-based image classifiers built on this assumption, therefore, cannot take into account an innate hierarchy among target classes (e.g., cats and dogs in animal image classification) or additional information that can be easily derived from the data (e.g.,numbers larger than five in the recognition of handwritten digits), thereby resulting in scalability issues when the number of target classes is large. Combining two related but slightly different ideas of hierarchical classification and logical learning by auxiliary inputs, we propose a new learning framework called hierarchical auxiliary learning, which not only address the scalability issues with a large number of classes but also could further reduce the classification/recognition errors with a reasonable number of classes. In the hierarchical auxiliary learning, target classes are semantically or non-semantically grouped into superclasses, which turns the original problem of mapping between an image and its target class into a new problem of mapping between a pair of an image and its superclass and the target class. To take the advantage of superclasses, we introduce an auxiliary block into a neural network, which generates auxiliary scores used as additional information for final classification/recognition; in this paper, we add the auxiliary block between the last residual block and the fully-connected output layer of the ResNet. Experimental results demonstrate that the proposed hierarchical auxiliary learning can reduce classification errors up to 0.56, 1.6 and 3.56 percent with MNIST, SVHN and CIFAR-10 datasets, respectively. | ['Jaehoon Cha', 'Sanghyuk Lee', 'Kyeong Soo Kim'] | 2019-06-03 | null | null | null | null | ['auxiliary-learning'] | ['methodology'] | [ 2.67627865e-01 2.78157387e-02 -1.15276359e-01 -5.50049245e-01
9.63458046e-02 -3.44908476e-01 4.17660743e-01 2.27206275e-01
-5.65923572e-01 6.44795001e-01 -1.91984281e-01 -2.72649944e-01
-7.48209134e-02 -1.18393159e+00 -6.67449117e-01 -6.44488633e-01
2.41587535e-01 1.50339380e-01 5.50528288e-01 -6.88920021e-02
7.83058330e-02 5.61500490e-01 -1.81523740e+00 6.81412578e-01
8.38279068e-01 1.47654021e+00 2.80039459e-01 2.23777965e-01
-1.98009863e-01 1.03702605e+00 -7.03917921e-01 -1.59596786e-01
2.40153342e-01 -4.15489167e-01 -8.25083375e-01 8.89870673e-02
2.83906341e-01 -2.77614683e-01 -2.73224056e-01 1.04157925e+00
7.98956156e-02 1.33495793e-01 6.07304692e-01 -1.50655866e+00
-7.99298346e-01 4.82589543e-01 -2.40274265e-01 1.01666138e-01
-2.32759625e-01 -1.06509335e-01 8.73105526e-01 -7.70551741e-01
9.25305337e-02 1.20674813e+00 6.86484814e-01 3.81536216e-01
-1.07928312e+00 -1.09631908e+00 2.82570809e-01 4.06431645e-01
-1.50319779e+00 -6.10623546e-02 5.87130487e-01 -5.30030727e-01
6.22910082e-01 3.01930249e-01 4.97922629e-01 7.63426483e-01
-5.41418865e-02 5.58363557e-01 1.26791465e+00 -2.48351783e-01
4.25991341e-02 5.24576426e-01 6.06915891e-01 4.40147609e-01
3.24445307e-01 9.10456255e-02 -4.47848998e-03 8.69970173e-02
7.63635635e-01 2.63212651e-01 -1.22614615e-01 -2.43855298e-01
-1.01091290e+00 7.70046711e-01 9.37356055e-01 3.56467932e-01
-1.49305180e-01 -3.19583803e-01 2.62792498e-01 2.72274971e-01
-6.96184561e-02 2.21341088e-01 -5.05887866e-01 5.91899216e-01
-5.41856945e-01 -1.85623735e-01 5.24493575e-01 1.02526379e+00
1.06784236e+00 1.75346851e-01 2.83314218e-03 9.31677818e-01
1.42928839e-01 2.37262577e-01 6.90571487e-01 -4.58817065e-01
4.04599875e-01 1.13365006e+00 -2.13560551e-01 -1.07603979e+00
-4.74365771e-01 -8.57758522e-01 -1.31603074e+00 3.10308725e-01
3.01079869e-01 2.78703779e-01 -1.14344895e+00 1.76809144e+00
2.49398783e-01 3.01253885e-01 3.33714217e-01 7.36445010e-01
1.32278967e+00 6.98080242e-01 9.30103585e-02 1.26994550e-02
1.35372996e+00 -1.06922424e+00 -3.19154620e-01 -2.05175236e-01
4.21583533e-01 -5.49058616e-01 1.14796329e+00 2.59141862e-01
-5.54140329e-01 -1.03387046e+00 -1.24031639e+00 -1.51623525e-02
-7.31937468e-01 3.27275395e-01 5.49877584e-01 3.75354260e-01
-7.09367931e-01 5.49881339e-01 -5.36715806e-01 -1.40389249e-01
4.22456324e-01 6.44452929e-01 -5.46250820e-01 7.50819501e-03
-1.17717040e+00 6.88895345e-01 9.40167010e-01 1.36016369e-01
-6.79226995e-01 -4.34354603e-01 -7.21208513e-01 3.57009560e-01
1.94610506e-01 -1.80268809e-01 8.71041179e-01 -1.53262758e+00
-1.07470512e+00 6.39010429e-01 2.17669696e-01 -3.59041512e-01
2.61291027e-01 1.07692167e-01 -5.29793382e-01 -8.62389132e-02
1.59430355e-02 7.66068757e-01 6.32155240e-01 -1.32858014e+00
-9.36438084e-01 -3.61702859e-01 3.83128524e-01 6.42044842e-02
-8.19439292e-01 -2.05389872e-01 -2.51770854e-01 -7.68988788e-01
3.49194050e-01 -8.75970006e-01 -2.34137084e-02 -1.51922464e-01
-3.71761858e-01 -3.26084077e-01 9.93494451e-01 -5.40507436e-01
1.00682771e+00 -2.08819079e+00 -2.74654198e-02 3.16658854e-01
1.50398359e-01 4.73306030e-01 -3.38802993e-01 -1.16506144e-01
-4.45204467e-01 1.48744330e-01 -2.10766241e-01 2.49420032e-01
-4.02005732e-01 3.84835303e-01 -1.91676587e-01 6.61940053e-02
1.96460128e-01 7.38819778e-01 -6.60190165e-01 -4.04436767e-01
2.55922705e-01 3.26763391e-01 -4.21932995e-01 2.33206660e-01
1.69521138e-01 5.06631672e-01 -3.00976098e-01 5.22922814e-01
7.79925406e-01 -4.12027389e-01 5.86640239e-02 -3.92010987e-01
-4.77773789e-03 2.32441321e-01 -1.31308675e+00 8.95554185e-01
-4.40537483e-01 1.95297241e-01 -3.00517619e-01 -1.43961644e+00
1.17047131e+00 2.62078732e-01 2.22533301e-01 -6.73008144e-01
7.53712505e-02 2.83413053e-01 5.09024501e-01 -1.35267541e-01
9.01369005e-02 -4.76930812e-02 6.62060454e-02 1.65623724e-01
1.96889937e-01 3.53190064e-01 5.81960902e-02 -2.40280196e-01
7.45482504e-01 -2.66521096e-01 4.57333595e-01 -2.53679693e-01
9.06290233e-01 -1.21523358e-01 9.16390300e-01 7.24796295e-01
7.28292614e-02 3.93723011e-01 3.59316379e-01 -8.68542433e-01
-1.04150510e+00 -8.75848591e-01 -3.24527770e-01 9.26535368e-01
2.83824027e-01 -1.90792024e-01 -5.15659571e-01 -7.86878645e-01
5.93511900e-03 2.83904314e-01 -7.26740241e-01 -3.43623132e-01
-4.58492219e-01 -9.08714056e-01 5.07597566e-01 8.08173895e-01
1.01261675e+00 -1.20442402e+00 -3.78710449e-01 6.48352504e-02
-1.98917426e-02 -1.19871914e+00 6.11390136e-02 3.66673499e-01
-9.74418223e-01 -1.19598544e+00 -4.80544239e-01 -1.04148042e+00
8.81728530e-01 3.08511674e-01 9.00415599e-01 4.00718570e-01
-8.51504877e-02 -1.68335199e-01 -4.62589651e-01 -2.87829489e-01
-3.62022310e-01 1.04155950e-01 8.63380134e-02 2.92683542e-01
3.86138648e-01 -6.67342603e-01 -3.44719559e-01 7.00335264e-01
-1.00947213e+00 3.40578347e-01 7.07334816e-01 1.12730932e+00
4.78617817e-01 3.23309034e-01 8.36622655e-01 -7.04204381e-01
1.65695131e-01 -4.51152176e-01 -6.85610712e-01 3.74599218e-01
-4.11226749e-01 -2.94202399e-02 1.04593349e+00 -6.74719334e-01
-7.79013753e-01 1.22340955e-01 -1.75666183e-01 -3.93530995e-01
-2.99319685e-01 6.02190316e-01 -4.84439135e-01 -1.79003567e-01
5.59389234e-01 3.65111828e-01 -1.96311072e-01 -5.19447267e-01
-3.41120586e-02 8.78160536e-01 4.93234903e-01 -2.93865830e-01
8.12956870e-01 1.52658656e-01 7.48077258e-02 -6.28945351e-01
-9.30641830e-01 -2.73222387e-01 -8.77960205e-01 6.09846488e-02
7.12753415e-01 -9.18565631e-01 -7.79275358e-01 7.03662336e-01
-1.10345912e+00 -2.07612291e-01 -1.30503282e-01 6.37398958e-01
-9.32783410e-02 2.97436297e-01 -5.21325648e-01 -4.58591074e-01
-8.38183835e-02 -1.20970845e+00 4.78221834e-01 3.47988904e-01
5.28631657e-02 -5.90781033e-01 -6.69886470e-01 3.24954093e-01
3.21354151e-01 1.25073388e-01 1.26871717e+00 -1.03991687e+00
-6.75412893e-01 -3.06150198e-01 -5.69678187e-01 8.94488811e-01
4.08922940e-01 -3.39992642e-01 -9.58813727e-01 -3.32355976e-01
-5.62839136e-02 -5.31117141e-01 8.90855849e-01 -8.96153785e-03
1.40217304e+00 -4.42862898e-01 -1.77404612e-01 5.36476076e-01
1.34741735e+00 5.64518988e-01 7.89962411e-01 4.68799651e-01
7.72790074e-01 6.98040843e-01 4.62618977e-01 1.85696155e-01
1.57525986e-01 6.46329165e-01 4.03630346e-01 -2.71392077e-01
-1.13066636e-01 -3.01582664e-02 -3.03387810e-02 9.46611106e-01
-1.71975657e-01 -1.30110085e-01 -9.35798943e-01 3.65505338e-01
-1.69282842e+00 -6.91178381e-01 -8.64313021e-02 2.46749210e+00
7.45859861e-01 2.65850484e-01 -1.73443928e-01 3.01354438e-01
1.06975269e+00 -2.77442187e-01 -7.02415228e-01 -1.26730308e-01
-1.77344009e-01 2.46786729e-01 3.72657865e-01 2.29352236e-01
-1.33283973e+00 7.49131680e-01 5.11894989e+00 9.12341356e-01
-1.27488017e+00 -2.50102457e-04 7.96397805e-01 4.82350111e-01
1.79108992e-01 -7.14154243e-02 -9.78044271e-01 4.32089746e-01
5.38870633e-01 8.35996345e-02 1.91171348e-01 9.77092326e-01
-2.58535832e-01 2.03006044e-01 -1.24747932e+00 9.29894507e-01
-8.06113631e-02 -1.12849212e+00 4.51434702e-01 1.28811551e-02
6.09730184e-01 -1.68693393e-01 -3.30939554e-02 5.51989913e-01
1.61992490e-01 -1.17637694e+00 7.39651084e-01 1.08519867e-01
6.01197720e-01 -7.50915170e-01 1.12033844e+00 5.74648738e-01
-1.38098311e+00 -3.53399962e-01 -7.06454754e-01 -1.68146148e-01
-5.42460382e-01 2.65698016e-01 -3.41704935e-01 5.83875000e-01
9.38110590e-01 7.11780846e-01 -7.21548021e-01 1.08712780e+00
-2.58840621e-01 3.24639916e-01 -2.63487577e-01 1.74715310e-01
2.80073434e-01 2.73088715e-03 7.03376234e-02 8.50666881e-01
2.76523292e-01 3.40259254e-01 3.93279403e-01 7.82740593e-01
-2.34811813e-01 1.96835726e-01 -3.63016665e-01 2.09126711e-01
3.91640007e-01 1.25630140e+00 -9.00428355e-01 -6.20880067e-01
-5.65884471e-01 6.57919109e-01 4.48384672e-01 3.51411045e-01
-7.97989309e-01 -6.81081951e-01 3.48511964e-01 -5.06159924e-02
3.68687063e-01 9.62614790e-02 -2.92433977e-01 -1.12843192e+00
1.93225488e-01 -8.95600796e-01 4.90906566e-01 -5.28474510e-01
-1.23189354e+00 9.31380630e-01 -1.37965187e-01 -1.45080721e+00
1.11866482e-01 -8.69929910e-01 -5.70056915e-01 9.24875855e-01
-1.46734953e+00 -1.23896801e+00 -7.07122862e-01 7.27503717e-01
2.18007043e-01 -3.48968625e-01 8.77940476e-01 4.63161439e-01
-4.69148129e-01 8.41714561e-01 1.17953077e-01 6.34820819e-01
3.87647539e-01 -8.87876987e-01 -8.99014771e-02 6.70835078e-01
-1.30068138e-01 6.29861891e-01 1.03905663e-01 -4.06736791e-01
-7.36696661e-01 -1.44232237e+00 8.34371567e-01 -4.68998402e-03
3.98560375e-01 -4.98130232e-01 -1.12785268e+00 7.39939094e-01
-1.18024968e-01 2.64336556e-01 6.67310774e-01 4.33514006e-02
-7.06470907e-01 -5.07047534e-01 -9.93602931e-01 3.17906529e-01
7.48218596e-01 -3.58654648e-01 -7.17311978e-01 2.10959509e-01
7.71450520e-01 -1.40085638e-01 -8.83652866e-01 9.76396441e-01
6.48233771e-01 -9.02143538e-01 1.05115080e+00 -8.47308397e-01
5.78261077e-01 -4.90522623e-01 -3.16431254e-01 -1.05158782e+00
-4.83655512e-01 3.66376787e-01 2.61201084e-01 1.18679631e+00
3.55208993e-01 -8.21201384e-01 7.14144111e-01 3.01953882e-01
-2.15202421e-01 -7.57496715e-01 -8.43495190e-01 -1.17494071e+00
1.22488156e-01 -1.10212825e-01 7.82875180e-01 1.15101027e+00
-4.02002424e-01 5.03943741e-01 -2.60239631e-01 2.73162633e-01
3.27091008e-01 3.48290622e-01 6.50471091e-01 -1.61747944e+00
-2.24703833e-01 -3.37543368e-01 -8.40334654e-01 -9.85734999e-01
1.40176669e-01 -9.88735318e-01 -6.29974976e-02 -1.33795607e+00
3.27846169e-01 -9.63413835e-01 -8.39312315e-01 9.32014763e-01
8.63223523e-03 5.00002861e-01 2.59056002e-01 4.07352984e-01
-2.79341847e-01 5.56784928e-01 1.08618462e+00 -3.38805258e-01
-1.82631701e-01 1.16305202e-01 -6.98313355e-01 8.79809260e-01
8.50370526e-01 -4.09468114e-01 -4.55416918e-01 -3.08739096e-01
-1.23010747e-01 -2.06192940e-01 6.17279470e-01 -1.29749477e+00
2.17503458e-01 -1.34671211e-01 6.89473450e-01 -5.29555440e-01
2.60633230e-01 -1.06305873e+00 1.11012951e-01 6.87571883e-01
-3.85820508e-01 -1.14172831e-01 3.65535438e-01 3.24304760e-01
-5.43555140e-01 -2.99834877e-01 1.02048099e+00 -2.41125777e-01
-8.13801646e-01 3.08851153e-01 -1.52380735e-01 -2.55940497e-01
1.07912123e+00 -4.32577491e-01 -4.57184285e-01 -9.72703695e-02
-9.73696291e-01 1.08535521e-01 8.12164843e-02 6.12836301e-01
6.35810912e-01 -1.69621158e+00 -6.36390865e-01 3.17728698e-01
3.16574365e-01 8.21445659e-02 1.83396369e-01 6.45316899e-01
-3.21789593e-01 5.33157289e-01 -6.70915723e-01 -6.60958827e-01
-1.43538225e+00 7.30685294e-01 3.28489214e-01 -2.99132228e-01
-3.02269727e-01 7.26044655e-01 8.81725252e-01 -8.40431809e-01
3.26979935e-01 -4.16889608e-01 -5.81903338e-01 -4.77044471e-02
4.41275001e-01 9.62846652e-02 1.27874911e-01 -9.01693881e-01
-3.11248392e-01 5.37862122e-01 -2.45825842e-01 5.63187003e-01
1.34259999e+00 2.67076343e-01 -3.95194024e-01 3.48649919e-01
1.38793206e+00 -4.34198707e-01 -7.04024792e-01 -5.46151161e-01
-4.57565263e-02 -2.33195767e-01 -1.54742777e-01 -7.41153240e-01
-1.26548636e+00 1.07673526e+00 7.82890916e-01 1.54370978e-01
1.37749839e+00 -1.39848247e-01 5.98976552e-01 5.85015297e-01
2.17064694e-01 -9.30244923e-01 6.67402595e-02 6.43184125e-01
8.48129213e-01 -1.19558501e+00 -1.01343021e-01 -5.55736005e-01
-2.74966240e-01 1.25736034e+00 1.15516031e+00 -1.36512266e-02
6.76615715e-01 -7.09329918e-02 -1.02798939e-01 1.09766856e-01
-4.43747371e-01 -3.12376544e-02 4.89100456e-01 4.27697718e-01
2.67807007e-01 1.24354459e-01 -1.05465852e-01 9.18394923e-01
-1.24732114e-01 -3.51815894e-02 2.77762055e-01 8.07631671e-01
-6.17085814e-01 -1.01785207e+00 -3.47864151e-01 6.51142180e-01
-2.47904569e-01 -2.25548759e-01 -2.24397451e-01 9.21545863e-01
7.18685031e-01 7.78135955e-01 3.02413642e-01 -7.37859011e-01
4.15102869e-01 -1.83097780e-01 1.01096652e-01 -7.52185822e-01
-5.03096640e-01 -2.13187948e-01 -2.70933598e-01 -2.46866256e-01
-4.47691381e-01 -1.29697010e-01 -1.12765169e+00 -7.69417286e-02
-4.75785047e-01 1.63389161e-01 3.44310552e-01 9.56081927e-01
1.08182587e-01 5.80734074e-01 6.21608317e-01 -4.30866212e-01
-5.34255207e-01 -1.03086030e+00 -4.89054650e-01 4.24064368e-01
1.61972046e-01 -7.26540565e-01 -3.01773250e-01 2.20528115e-02] | [9.411824226379395, 2.407099723815918] |
d130d3cd-da12-41e1-81d3-e5f771e22d7a | a-systematic-literature-review-about-idea | 2202.12826 | null | https://arxiv.org/abs/2202.12826v1 | https://arxiv.org/pdf/2202.12826v1.pdf | A Systematic Literature Review about Idea Mining: The Use of Machine-driven Analytics to Generate Ideas | Idea generation is the core activity of innovation. Digital data sources, which are sources of innovation, such as patents, publications, social media, websites, etc., are increasingly growing at unprecedented volume. Manual idea generation is time-consuming and is affected by the subjectivity of the individuals involved. Therefore, the use machine-driven data analytics techniques to analyze data to generate ideas and support idea generation by serving users is useful. The objective of this study is to study state-of the-art machine-driven analytics for idea generation and data sources, hence the result of this study will generally server as a guideline for choosing techniques and data sources. A systematic literature review is conducted to identify relevant scholarly literature from IEEE, Scopus, Web of Science and Google Scholar. We selected a total of 71 articles and analyzed them thematically. The results of this study indicate that idea generation through machine-driven analytics applies text mining, information retrieval (IR), artificial intelligence (AI), deep learning, machine learning, statistical techniques, natural language processing (NLP), NLP-based morphological analysis, network analysis, and bibliometric to support idea generation. The results include a list of techniques and procedures in idea generation through machine-driven idea analytics. Additionally, characterization and heuristics used in idea generation are summarized. For the future, tools designed to generate ideas could be explored. | ['Gustaf Juell-Skielse', 'Workneh Y. Ayele'] | 2022-01-30 | null | null | null | null | ['morphological-analysis'] | ['natural-language-processing'] | [-1.82377800e-01 -1.85786076e-02 -4.95700091e-01 4.49961215e-01
-2.15161547e-01 -6.84444606e-01 6.04769945e-01 6.82279527e-01
-8.92237797e-02 8.58459115e-01 3.32484514e-01 -9.84479487e-01
-5.87076962e-01 -1.34182501e+00 -3.46313059e-01 -2.78322935e-01
-3.30887511e-02 4.19683427e-01 -4.94353801e-01 -2.15450197e-01
1.60756123e+00 6.04772449e-01 -1.70954692e+00 2.23069370e-01
1.25113249e+00 1.22232127e+00 2.49670833e-01 1.36185572e-01
-1.20587575e+00 8.83245766e-01 -6.61128283e-01 -3.31801474e-01
4.10936236e-01 -3.74899507e-01 -7.78447330e-01 -3.01308542e-01
-3.22146595e-01 1.33456364e-01 2.59500086e-01 9.13943887e-01
5.83232522e-01 -1.68342263e-01 7.16310680e-01 -1.39400899e+00
-1.27231872e+00 1.04539239e+00 -3.88015866e-01 3.06861341e-01
7.01388121e-01 -5.82819805e-02 7.76086509e-01 -1.10763514e+00
1.05743194e+00 1.06554246e+00 5.56767523e-01 -2.05546781e-01
-4.74790990e-01 -9.72722769e-01 -2.94296801e-01 1.80061668e-01
-1.06406403e+00 1.85980238e-02 1.41749561e+00 -9.64029789e-01
9.01204407e-01 -3.03424329e-01 1.27722204e+00 7.48735964e-01
6.70529604e-01 5.29238224e-01 1.40853107e+00 -8.95922840e-01
7.77352870e-01 2.93789595e-01 2.86194235e-01 4.52312440e-01
8.84706676e-01 -1.26909941e-01 -6.71274722e-01 -1.85823053e-01
6.43090189e-01 3.71150553e-01 4.23509121e-01 2.35643983e-01
-1.19063008e+00 1.24508560e+00 2.94864804e-01 8.10111046e-01
-1.14386070e+00 -3.72576565e-01 3.66930544e-01 5.83482444e-01
4.63121474e-01 1.06268573e+00 -3.97528112e-01 -4.86864984e-01
-1.11391723e+00 2.52721012e-01 1.06147969e+00 9.54137325e-01
5.89827597e-01 2.79241592e-01 1.26482293e-01 2.65568614e-01
6.09280765e-01 2.10175961e-01 1.20918667e+00 -8.23301196e-01
3.56812596e-01 1.55872393e+00 -4.51100677e-01 -1.41642249e+00
-1.32451341e-01 -2.76498705e-01 -8.29852343e-01 1.08545452e-01
-1.26597807e-01 -4.41546530e-01 -1.00628924e+00 7.51851976e-01
-6.23668432e-02 -6.61419153e-01 -4.93064187e-02 3.33674252e-01
1.18865085e+00 8.00011992e-01 1.16400711e-01 -4.30905342e-01
1.22117472e+00 -5.36649823e-01 -8.36057544e-01 3.78868103e-01
5.77362537e-01 -8.96725953e-01 5.64568698e-01 9.23133492e-01
-1.32547295e+00 -6.42347574e-01 -6.42011285e-01 1.74364626e-01
-1.32345068e+00 -1.37027800e-01 9.97484803e-01 4.52189595e-01
-1.00980914e+00 4.21140820e-01 -2.06416130e-01 -4.52545464e-01
8.54981482e-01 4.29732442e-01 -1.59859195e-01 2.32664868e-01
-1.06786644e+00 8.22709620e-01 7.67900586e-01 -3.37089539e-01
-6.36960194e-02 -7.49688327e-01 -3.63162428e-01 -9.34116095e-02
1.43319860e-01 -8.62863064e-01 5.51451445e-01 -1.29549730e+00
-1.37115788e+00 5.75342357e-01 1.30893469e-01 -7.33378589e-01
-5.24414442e-02 1.42547384e-01 -3.99833858e-01 1.96122691e-01
2.15730712e-01 5.11648059e-01 4.07552540e-01 -1.08814478e+00
-7.32916057e-01 -4.53191906e-01 -2.96759158e-01 -2.87223458e-01
-6.62891328e-01 1.13566846e-01 9.30815861e-02 -8.96130025e-01
1.39983281e-01 -4.22403008e-01 -4.13256973e-01 -4.67916578e-01
-3.01917791e-01 -6.73808575e-01 6.70783401e-01 -8.87259722e-01
1.44111693e+00 -1.67489195e+00 -1.89999342e-01 6.74412251e-01
4.28596675e-01 1.50389001e-01 4.24960315e-01 1.02732146e+00
1.66634340e-02 9.31352377e-01 2.39540651e-01 5.80912769e-01
-1.34447694e-01 -2.47931331e-01 -2.36552060e-01 -1.76447749e-01
-2.25321099e-01 9.34657574e-01 -6.48774087e-01 -5.95996976e-01
3.95358711e-01 5.96715882e-02 -3.85436475e-01 -1.40256971e-01
3.15629840e-02 -4.16886322e-02 -8.54080975e-01 1.17409384e+00
3.09743881e-01 -2.62456924e-01 -2.30498746e-01 -1.02309078e-01
-7.20233023e-01 3.08629066e-01 -9.30723011e-01 1.67278886e+00
-3.93770188e-01 9.53249156e-01 -5.00657558e-01 -1.08829570e+00
1.55426157e+00 2.85880208e-01 6.81424737e-01 -8.09539557e-01
3.78728628e-01 4.99700695e-01 2.11706653e-01 -7.29368687e-01
5.00698626e-01 9.02354941e-02 2.89568901e-01 4.97565627e-01
-4.07771878e-02 1.05773486e-01 5.13627827e-01 3.61647516e-01
9.82469738e-01 -3.25158745e-01 7.43021727e-01 -2.35944077e-01
2.62699276e-01 5.28279066e-01 2.93010861e-01 6.29237533e-01
2.34214753e-01 6.38397709e-02 3.33884031e-01 -6.46463692e-01
-1.18062305e+00 -6.80710435e-01 1.15112357e-01 4.98976618e-01
-3.72263253e-01 -3.13727736e-01 -4.20008212e-01 -1.03927404e-01
1.32147476e-01 8.88299763e-01 -1.93924934e-01 2.57386118e-01
4.31921929e-02 -3.26684266e-01 -1.81682575e-02 1.72392502e-01
5.28261542e-01 -1.61451554e+00 -3.82813722e-01 4.83203083e-01
3.08565587e-01 -2.11814031e-01 7.47751057e-01 -1.89658627e-01
-1.29847491e+00 -9.13967848e-01 -6.90091372e-01 -7.09024727e-01
8.34929049e-01 4.52117771e-02 1.00107467e+00 2.58439500e-02
-3.95862132e-01 1.45001918e-01 -8.40546966e-01 -1.01387691e+00
-3.46872300e-01 1.78672567e-01 -1.14458308e-01 -6.52741015e-01
7.50092864e-01 -7.90455580e-01 -5.22292674e-01 -5.38936019e-01
-9.13509011e-01 -8.76203477e-02 1.19585347e+00 3.84317756e-01
4.83070254e-01 6.91899657e-01 1.14346206e+00 -5.64636171e-01
1.39752603e+00 -1.02611327e+00 -4.87469673e-01 -1.20544970e-01
-1.30743313e+00 -1.54074281e-01 4.97768372e-01 -1.63699090e-01
-9.34882760e-01 -3.97882313e-01 2.12707385e-01 -6.83134496e-02
-3.01774651e-01 1.37920606e+00 1.21899806e-01 -1.42999023e-01
6.30326569e-01 2.07419634e-01 2.74684340e-01 -5.35215795e-01
3.73426318e-01 1.09193075e+00 -2.77523011e-01 -3.17000479e-01
6.19635344e-01 2.24768266e-01 1.50685655e-02 -7.41052508e-01
-1.15351215e-01 -3.37091535e-01 -4.46729809e-01 -6.29475236e-01
3.49649221e-01 -6.08269036e-01 -5.81933081e-01 4.60037887e-02
-1.16843021e+00 4.27588016e-01 -5.29063702e-01 6.11433804e-01
-9.23077315e-02 -5.41753992e-02 -2.86450237e-01 -8.72108519e-01
-1.12581420e+00 -8.37756038e-01 2.96708167e-01 7.00509310e-01
-3.81820470e-01 -1.09964895e+00 -4.01935875e-02 6.71135426e-01
7.39300787e-01 4.86908734e-01 9.74439681e-01 -1.26022732e+00
-5.27704597e-01 -4.34633434e-01 -7.88273066e-02 9.92733911e-02
3.09441656e-01 4.08483237e-01 -1.78725421e-01 3.25244993e-01
-1.50155514e-01 -4.67881896e-02 4.11687851e-01 5.95245719e-01
1.27479303e+00 -8.43190730e-01 -3.97161365e-01 -1.01349726e-01
1.59458697e+00 9.72909272e-01 7.84490883e-01 9.34128046e-01
4.30445433e-01 8.05364788e-01 5.35601795e-01 7.62006938e-01
2.41208822e-01 -2.24352941e-01 1.08750805e-01 3.91483724e-01
2.73105204e-01 -5.57558276e-02 1.38970554e-01 1.05564165e+00
-5.18187940e-01 -2.06332162e-01 -1.25571513e+00 6.79046154e-01
-1.54526937e+00 -1.16220438e+00 -4.38519716e-01 1.72365570e+00
6.15985096e-01 4.58564669e-01 3.32522482e-01 4.22484159e-01
4.57628757e-01 -3.76961291e-01 -2.23270819e-01 -7.65342772e-01
1.95357986e-02 5.55430114e-01 5.27625680e-01 -3.18310082e-01
-2.64596760e-01 7.95840502e-01 5.80038452e+00 7.88378298e-01
-1.20192683e+00 -2.66743273e-01 4.98035461e-01 1.27987340e-01
-8.17849934e-01 2.10613459e-01 -3.05970699e-01 7.59595454e-01
8.65601003e-01 -7.63715982e-01 3.52849156e-01 1.13816476e+00
4.78307605e-01 -7.62149543e-02 -3.29889357e-01 9.61525559e-01
2.96697896e-02 -2.26113963e+00 6.11723363e-01 5.39445460e-01
8.93844664e-01 -1.65570959e-01 4.93773855e-02 1.08434007e-01
2.22671941e-01 -7.46527910e-01 4.03852224e-01 9.61763918e-01
-5.88240400e-02 -9.10089195e-01 9.15649712e-01 -2.72262301e-02
-8.52191627e-01 -4.47960407e-01 -2.17711031e-01 -6.40553415e-01
8.85649249e-02 1.10935771e+00 -8.02035332e-01 8.31912994e-01
6.11379087e-01 6.82516873e-01 -6.40485764e-01 1.31725359e+00
4.52226549e-02 7.75364339e-01 5.35668842e-02 -5.18340647e-01
1.03297263e-01 -3.84333998e-01 4.97277677e-01 7.98900306e-01
7.86731422e-01 -9.57110338e-03 -2.01053530e-01 1.04830122e+00
-1.43688112e-01 6.25406325e-01 -1.02389324e+00 -9.92309988e-01
8.65650952e-01 1.11288941e+00 -1.14430034e+00 -3.90537232e-01
-2.92579830e-01 1.11838445e-01 -2.76789010e-01 1.35235265e-01
5.10657281e-02 -8.54275584e-01 -2.45721880e-02 4.03910190e-01
-6.69117435e-04 -3.28200817e-01 -1.04445601e+00 -5.98178804e-01
-1.78807497e-01 -8.91894579e-01 2.16170266e-01 -8.86349797e-01
-1.61240804e+00 2.43850619e-01 -1.93568114e-02 -1.19572723e+00
-3.34632128e-01 -8.19972634e-01 -1.13056386e+00 6.74967170e-01
-8.59454453e-01 -8.73450518e-01 -1.43995434e-01 2.98334688e-01
6.27796352e-01 -1.16660857e+00 3.55201542e-01 1.76428437e-01
-3.32320392e-01 -1.51107356e-01 1.23764776e-01 -6.83268532e-02
1.68750063e-02 -1.09577298e+00 1.18342504e-01 5.69371581e-01
-1.24786325e-01 8.92989814e-01 2.36909166e-01 -1.11205971e+00
-1.69350123e+00 -4.55265015e-01 1.16202211e+00 1.27604008e-01
9.69053447e-01 2.00166881e-01 -5.38512170e-01 2.07879603e-01
7.11218476e-01 -8.66307855e-01 1.17212367e+00 -3.75548676e-02
3.61378670e-01 -2.10493207e-01 -1.13789809e+00 6.17076516e-01
5.80445766e-01 6.65414333e-02 -1.00915968e+00 1.88288972e-01
6.02408290e-01 3.99712175e-01 -1.10274196e+00 -7.40378648e-02
5.29815912e-01 -6.68802261e-01 9.21152651e-01 -5.81034243e-01
7.31114745e-01 1.01315543e-01 4.91250187e-01 -9.32369113e-01
-4.12844390e-01 -6.24365926e-01 -1.62357330e-01 1.75489795e+00
5.35230160e-01 -7.24633157e-01 7.19265699e-01 7.49134600e-01
-7.46759921e-02 -7.30895340e-01 -6.34139299e-01 -4.24616754e-01
3.04514123e-03 -4.73678708e-01 5.61765194e-01 1.19287217e+00
4.45130825e-01 2.36678451e-01 1.71831504e-01 -7.60828435e-01
4.88526374e-01 1.69700399e-01 7.75588036e-01 -1.79594159e+00
6.07352197e-01 -9.70771313e-01 -6.29550397e-01 -7.07987100e-02
-1.44782200e-01 -8.99509370e-01 -7.10100830e-01 -2.41693997e+00
-1.32156953e-01 -3.66652131e-01 -2.67059535e-01 3.00664395e-01
3.65623266e-01 -4.23081040e-01 1.68541208e-01 5.25401235e-01
-6.13794550e-02 3.11630249e-01 1.23509526e+00 3.13764922e-02
-6.74280643e-01 -1.77756771e-01 -1.40992570e+00 6.87069714e-01
1.16750109e+00 -5.05443275e-01 -3.29517394e-01 2.03519836e-01
1.06215727e+00 -3.70156646e-01 -4.91132848e-02 -6.66198909e-01
5.82842052e-01 -3.17314506e-01 5.09419680e-01 -9.44103241e-01
-4.59421933e-01 -8.70641053e-01 1.79696128e-01 7.04110563e-01
-2.49673799e-01 3.99843574e-01 1.64333090e-01 2.08152264e-01
-4.49119300e-01 -4.24061686e-01 -7.11734965e-02 -3.05906773e-01
-6.93977594e-01 1.05756745e-01 -8.64124179e-01 -2.30589747e-01
1.03303516e+00 -6.71105027e-01 -2.09236011e-01 -2.84553587e-01
-1.32889926e-01 -7.67565966e-02 1.71518072e-01 4.54087257e-01
8.97998273e-01 -1.44354296e+00 -4.43604857e-01 -1.32465392e-01
-2.26647854e-01 1.41558172e-02 -1.51768714e-01 6.28638208e-01
-9.51709211e-01 4.91592079e-01 -8.60579550e-01 2.47236401e-01
-7.15200305e-01 8.35903585e-01 -6.76897049e-01 -7.19053447e-02
-2.87063211e-01 5.63483655e-01 -6.83728337e-01 -4.25537489e-02
6.05366044e-02 -2.67396152e-01 -7.13150144e-01 6.62541449e-01
3.16829503e-01 9.75675762e-01 -2.29763851e-01 -3.23272705e-01
-1.67040899e-01 4.57726419e-01 1.18047744e-01 -1.92644164e-01
1.46941900e+00 9.97758359e-02 -6.97251201e-01 4.06196058e-01
9.34145689e-01 -1.40291005e-01 2.10015222e-01 2.54641891e-01
3.08106393e-01 -3.21711093e-01 2.48689666e-01 -8.32838535e-01
-1.00362015e+00 6.31138265e-01 4.32100177e-01 8.56749833e-01
1.16190147e+00 -2.05285370e-01 7.68266380e-01 4.58680362e-01
1.70309260e-01 -1.60975134e+00 8.04886445e-02 2.26550385e-01
1.02076924e+00 -1.12279141e+00 1.20064907e-01 6.36640713e-02
-3.38431746e-01 1.64412332e+00 2.75615007e-01 -9.35945436e-02
1.14245653e+00 4.79258224e-02 -2.94702500e-01 -6.29079342e-01
-3.80929202e-01 7.09490851e-02 4.05761510e-01 4.39494699e-01
7.28692591e-01 -1.78543255e-01 -1.17000246e+00 7.89381087e-01
-6.16554141e-01 6.16617382e-01 2.87633419e-01 1.31352305e+00
-7.88235009e-01 -1.24826169e+00 -8.42659235e-01 1.05163252e+00
-7.14497983e-01 -3.15634370e-01 -1.00806129e+00 7.53452539e-01
4.13442224e-01 1.17683280e+00 1.02087721e-01 -3.17609102e-01
-1.31652504e-01 -9.88261029e-02 -1.93214610e-01 -3.32868397e-01
-7.75203884e-01 -1.55729890e-01 -2.75539644e-02 4.09360230e-02
-5.77108383e-01 -4.69437957e-01 -1.28674364e+00 -5.04283488e-01
-2.76564896e-01 3.76789063e-01 1.46887231e+00 1.07271183e+00
8.95811141e-01 6.43862784e-01 3.73250991e-01 -5.17448485e-01
3.80497485e-01 -1.05209470e+00 -3.53465378e-01 -2.43098978e-02
-5.29129148e-01 -4.91481751e-01 -3.45919281e-01 1.46024108e-01] | [9.493489265441895, 8.064936637878418] |
42cbd79e-615a-4347-9f7a-5e0207fa2122 | do-not-fire-the-linguist-grammatical-profiles | 2204.05717 | null | https://arxiv.org/abs/2204.05717v1 | https://arxiv.org/pdf/2204.05717v1.pdf | Do Not Fire the Linguist: Grammatical Profiles Help Language Models Detect Semantic Change | Morphological and syntactic changes in word usage (as captured, e.g., by grammatical profiles) have been shown to be good predictors of a word's meaning change. In this work, we explore whether large pre-trained contextualised language models, a common tool for lexical semantic change detection, are sensitive to such morphosyntactic changes. To this end, we first compare the performance of grammatical profiles against that of a multilingual neural language model (XLM-R) on 10 datasets, covering 7 languages, and then combine the two approaches in ensembles to assess their complementarity. Our results show that ensembling grammatical profiles with XLM-R improves semantic change detection performance for most datasets and languages. This indicates that language models do not fully cover the fine-grained morphological and syntactic signals that are explicitly represented in grammatical profiles. An interesting exception are the test sets where the time spans under analysis are much longer than the time gap between them (for example, century-long spans with a one-year gap between them). Morphosyntactic change is slow so grammatical profiles do not detect in such cases. In contrast, language models, thanks to their access to lexical information, are able to detect fast topical changes. | ['Lidia Pivovarova', 'Andrey Kutuzov', 'Mario Giulianelli'] | 2022-04-12 | null | https://aclanthology.org/2022.lchange-1.6 | https://aclanthology.org/2022.lchange-1.6.pdf | lchange-acl-2022-5 | ['xlm-r'] | ['natural-language-processing'] | [ 1.19557530e-01 -1.53955877e-01 -1.98894575e-01 -5.00051677e-01
-4.79051828e-01 -9.21346605e-01 9.86298859e-01 6.31516278e-01
-8.89985085e-01 5.73264420e-01 4.82167780e-01 -3.83745313e-01
-5.48152253e-03 -9.31841314e-01 -7.44615555e-01 -3.65500391e-01
-1.84188653e-02 2.96461076e-01 2.22502246e-01 -6.05601549e-01
2.79327333e-01 9.36061144e-02 -1.63510072e+00 4.22938883e-01
9.99886394e-01 2.49504820e-01 3.68673116e-01 1.80216506e-01
-4.82607394e-01 1.89985424e-01 -3.43166977e-01 -1.75601155e-01
-5.99415042e-02 -3.96651983e-01 -7.67591715e-01 -3.85311514e-01
6.57215774e-01 3.61939073e-01 1.39584869e-01 1.14540398e+00
3.49968910e-01 2.26681642e-02 6.00115836e-01 -3.30047667e-01
-6.70341313e-01 1.12278783e+00 -3.27742279e-01 6.25303328e-01
5.12283921e-01 1.51077807e-01 1.31006742e+00 -9.24235880e-01
1.07417035e+00 1.77780211e+00 8.98831546e-01 1.76030055e-01
-1.42122865e+00 -2.08767653e-01 5.87376356e-01 1.00959010e-01
-1.00764561e+00 -4.65236604e-01 8.06966782e-01 -6.68122709e-01
1.50285065e+00 4.17380780e-02 4.48949993e-01 1.14888084e+00
4.43311006e-01 4.55368668e-01 1.39893830e+00 -7.37134457e-01
7.89857730e-02 1.59106329e-01 3.87474120e-01 3.43736231e-01
3.13606083e-01 2.16452539e-01 -4.96744186e-01 -1.21905349e-01
9.29340441e-03 -3.57172400e-01 -8.48596096e-02 1.48759130e-02
-1.21298814e+00 8.40034723e-01 2.00493813e-01 9.90660787e-01
-3.13628286e-01 4.66688275e-02 7.51207471e-01 7.51419604e-01
7.54142940e-01 7.55449772e-01 -8.85150254e-01 -1.67752653e-01
-9.18717325e-01 1.49049476e-01 5.12361765e-01 3.53070587e-01
1.03241599e+00 9.22869667e-02 -2.13676295e-03 1.15544260e+00
1.15880534e-01 4.46809441e-01 1.02404177e+00 -2.92035341e-01
3.48098040e-01 6.10738397e-01 -2.15990916e-01 -9.17063117e-01
-6.08392179e-01 -2.17761204e-01 -3.78763117e-02 -2.51341194e-01
5.01708508e-01 7.06384704e-02 -6.02578759e-01 2.27820539e+00
2.59951174e-01 -3.90573800e-01 -6.89350814e-02 3.97830993e-01
5.47617435e-01 4.65384930e-01 5.39776623e-01 -1.85993299e-01
1.59195220e+00 -1.19693033e-01 -2.96112180e-01 -6.03711486e-01
9.87219632e-01 -7.60027111e-01 1.67800820e+00 -9.57837030e-02
-8.43812287e-01 -6.72596633e-01 -6.81079686e-01 1.08534873e-01
-8.27521503e-01 -1.88193724e-01 5.96137226e-01 5.81759751e-01
-8.70918751e-01 8.33810449e-01 -6.75020874e-01 -9.73726630e-01
5.61734959e-02 -4.39214520e-02 -2.73739874e-01 9.03694853e-02
-1.57614541e+00 1.19773948e+00 7.28171647e-01 -2.42398903e-01
-3.82965326e-01 -5.83779991e-01 -8.87239099e-01 -3.32418382e-01
2.55757779e-01 -3.66313517e-01 1.01760483e+00 -1.21765316e+00
-1.14678860e+00 1.44597399e+00 -1.18027747e-01 -3.17939043e-01
1.93952024e-01 6.52749389e-02 -4.96561736e-01 -4.64322746e-01
2.86915034e-01 5.37925601e-01 5.49864590e-01 -8.33469033e-01
-6.35351837e-01 -5.79707742e-01 -1.13226086e-01 6.85396791e-02
-2.98215121e-01 5.15335977e-01 2.34075218e-01 -7.22478867e-01
-1.49092034e-01 -7.73016691e-01 -1.00739256e-01 -7.79472530e-01
-6.83377089e-04 -6.20932817e-01 2.95604587e-01 -8.25681686e-01
1.38119423e+00 -2.07912946e+00 -3.41293775e-02 7.90155381e-02
-4.24811214e-01 1.54428959e-01 -2.83306271e-01 7.55096674e-01
-1.22978978e-01 2.91801721e-01 -3.89250785e-01 -1.61812529e-01
1.72351420e-01 4.86451507e-01 -2.23117724e-01 4.99106944e-01
1.20093703e-01 1.11703074e+00 -9.43002045e-01 8.68460461e-02
8.18101466e-02 1.31825522e-01 -4.97794032e-01 -5.70098519e-01
-4.80510801e-01 3.23319405e-01 -2.51142383e-02 4.35357094e-01
2.46149868e-01 3.60669672e-01 3.74724209e-01 2.66112596e-01
-4.82012838e-01 1.11114001e+00 -6.65488422e-01 1.66904306e+00
-8.05848300e-01 6.28247142e-01 -3.64470482e-01 -8.81692529e-01
9.74462450e-01 2.79749627e-03 7.82210287e-03 -1.20193052e+00
-5.78040965e-02 6.21085465e-01 5.05591094e-01 -3.50683123e-01
5.59129357e-01 -4.52801734e-01 -4.78173971e-01 3.17001045e-01
-1.81950182e-02 -5.81334047e-02 4.35359985e-01 -2.67827511e-01
1.11075926e+00 7.21869692e-02 7.30311990e-01 -7.03529477e-01
5.43440342e-01 -6.36139810e-02 8.87708366e-01 6.07055128e-01
1.47702351e-01 1.31545231e-01 4.80923384e-01 -3.78354788e-01
-9.65319991e-01 -9.60469425e-01 -3.76885831e-01 1.48752582e+00
-3.37612808e-01 -5.57193875e-01 -5.30206561e-01 -7.51026213e-01
2.33037040e-01 1.22199428e+00 -6.27866924e-01 -2.26662353e-01
-9.44298565e-01 -9.49384093e-01 5.07036030e-01 3.01778227e-01
-7.01508820e-02 -1.62770271e+00 -6.22284412e-01 5.53581536e-01
8.79120305e-02 -8.11263859e-01 -1.48480460e-01 2.05821142e-01
-9.33212698e-01 -8.89718413e-01 -9.44166705e-02 -6.32598996e-01
3.53007972e-01 -2.43808210e-01 1.42985928e+00 -2.08563283e-01
-6.44920394e-02 3.73977959e-01 -4.86327231e-01 -4.47888047e-01
-7.96694875e-01 3.89796585e-01 1.24386333e-01 -3.87719750e-01
7.41128325e-01 -8.27884376e-01 -2.13035226e-01 2.64456533e-02
-8.87852311e-01 -6.55786097e-01 5.04415751e-01 6.40610933e-01
4.31746989e-01 -2.53808379e-01 8.88743579e-01 -1.36113489e+00
7.00419366e-01 -7.23188877e-01 -4.80272502e-01 1.22529857e-01
-7.55564928e-01 2.53482610e-01 5.82574069e-01 -3.93016160e-01
-9.16534185e-01 -3.42386514e-01 -4.24442768e-01 4.88287389e-01
-4.87120360e-01 8.36625516e-01 -2.36076266e-01 3.42981398e-01
8.85985494e-01 8.61645117e-02 -2.26519570e-01 -8.10978055e-01
4.22296464e-01 5.29899240e-01 3.51789922e-01 -7.08067179e-01
5.25401533e-01 2.87109941e-01 -5.33270717e-01 -8.95299315e-01
-6.53881252e-01 -3.96402001e-01 -7.50287652e-01 1.27288505e-01
5.50399125e-01 -7.12466955e-01 4.21766378e-02 4.37344074e-01
-1.02529895e+00 -4.40416485e-01 -4.93495822e-01 5.18031836e-01
-4.87696022e-01 3.82655680e-01 -5.27382255e-01 -3.38446617e-01
-7.43689686e-02 -7.98378348e-01 9.42727804e-01 -2.56428421e-01
-7.92821467e-01 -1.49955857e+00 5.41908622e-01 -2.29418412e-01
4.85345602e-01 4.12150383e-01 1.50124002e+00 -8.48156393e-01
1.65227652e-01 1.57256708e-01 2.26778731e-01 2.77532369e-01
4.39422607e-01 -7.47564137e-02 -7.99946487e-01 -3.80244315e-01
-1.01139739e-01 1.12576284e-01 1.23993170e+00 3.84393722e-01
6.36994243e-01 -1.70097485e-01 -2.87493765e-01 2.36600637e-01
1.35448468e+00 2.20070496e-01 3.72973830e-01 5.81375659e-01
4.03922886e-01 8.65831017e-01 4.39655602e-01 2.14911383e-02
4.46283877e-01 8.64235818e-01 3.97996753e-02 2.60007352e-01
-2.95193225e-01 -4.29184049e-01 1.01593661e+00 1.04886925e+00
3.03463399e-01 -1.14961984e-02 -1.19701076e+00 9.03568447e-01
-1.62652540e+00 -6.70899391e-01 -3.23678672e-01 2.34692121e+00
1.05410314e+00 3.93736482e-01 1.31441042e-01 -8.22430849e-02
6.24860764e-01 5.29168427e-01 -2.99757123e-01 -9.65246379e-01
-4.56029624e-01 4.17213708e-01 4.27686155e-01 5.83424270e-01
-8.21216345e-01 1.34595728e+00 6.43486547e+00 6.66152835e-01
-1.06673729e+00 2.55628973e-01 1.21651985e-01 4.38961387e-02
-8.13584566e-01 1.07184574e-01 -7.88298190e-01 8.30047846e-01
1.38304341e+00 -2.15857178e-02 3.99898708e-01 5.31803608e-01
3.44794452e-01 -1.36462286e-01 -1.26408291e+00 6.09058142e-01
7.43586347e-02 -8.85445118e-01 6.13433421e-02 2.05231104e-02
6.42690480e-01 3.77554238e-01 -5.79156987e-02 4.76649940e-01
4.08875376e-01 -9.56448376e-01 9.53996539e-01 3.42749596e-01
7.41759777e-01 -5.95275760e-01 6.07845962e-01 3.52851450e-01
-8.94822955e-01 7.55759776e-02 -5.98344207e-01 -2.52206594e-01
1.47397295e-01 8.86970043e-01 -7.39119351e-01 3.77128989e-01
6.17363751e-01 8.84261608e-01 -9.44624603e-01 6.20276034e-01
-4.23111707e-01 1.01892984e+00 -3.64425629e-01 7.74178281e-02
3.82485211e-01 -1.49841666e-01 8.32165718e-01 1.66170275e+00
2.75272310e-01 -5.71893513e-01 -3.91964987e-02 7.36270428e-01
2.66389847e-02 5.24327636e-01 -8.31997097e-01 -2.22512320e-01
5.43905139e-01 8.35460246e-01 -6.50247395e-01 -3.93365696e-02
-7.08531737e-01 7.19901621e-01 5.22917867e-01 3.56308445e-02
-4.08190250e-01 -1.99958473e-01 8.24473441e-01 2.17973888e-01
2.64802128e-01 -1.95849001e-01 -1.35743871e-01 -1.07203650e+00
1.28004089e-01 -9.49523449e-01 4.81094956e-01 -3.69445860e-01
-1.44404984e+00 4.35728252e-01 -1.33636534e-01 -6.24219775e-01
-5.19729078e-01 -7.81146526e-01 -6.56543434e-01 8.96537960e-01
-1.61538517e+00 -1.05095947e+00 2.51972944e-01 2.64146805e-01
5.98766506e-01 -1.41930044e-01 7.86975801e-01 1.80383861e-01
-3.17909062e-01 5.02148509e-01 1.66904151e-01 -9.74074900e-02
8.14040720e-01 -1.43961668e+00 9.33099151e-01 8.58589947e-01
6.31571949e-01 7.66581953e-01 7.02609360e-01 -8.33997011e-01
-7.30399966e-01 -1.00249672e+00 1.39741409e+00 -6.61301911e-01
8.43710840e-01 -5.31227469e-01 -1.23612726e+00 7.71082520e-01
2.34336019e-01 -5.13842285e-01 6.35311246e-01 7.07978845e-01
-5.73231399e-01 1.28832757e-01 -8.46504331e-01 4.35006231e-01
1.38208151e+00 -7.96914399e-01 -1.16056883e+00 1.22167960e-01
6.03129804e-01 4.92086895e-02 -7.09715307e-01 4.20146972e-01
2.61908114e-01 -9.11007941e-01 5.80397069e-01 -8.44774961e-01
1.49982497e-01 -1.58709273e-01 -2.83214509e-01 -1.84932923e+00
-3.64369631e-01 -4.29409772e-01 5.87885618e-01 1.45334208e+00
6.21815264e-01 -1.04918182e+00 2.08986521e-01 -8.35228711e-02
-1.83705837e-01 -4.32059795e-01 -1.00999844e+00 -1.01153255e+00
6.30799890e-01 -6.68523431e-01 6.97065353e-01 1.19826031e+00
-2.67961062e-02 3.92628044e-01 2.46982723e-01 -2.30098277e-01
-1.39327049e-02 1.22017547e-01 2.55893558e-01 -1.43318903e+00
-2.48707831e-01 -7.95245826e-01 -5.16946554e-01 -4.57200766e-01
6.39584780e-01 -1.48307586e+00 -8.78666043e-02 -1.25095010e+00
1.09503798e-01 -4.75913823e-01 -3.75228196e-01 4.98319209e-01
-3.05424511e-01 -1.12775534e-01 7.96474144e-02 6.95128366e-02
-2.02916101e-01 3.12111050e-01 4.92604136e-01 6.11040369e-02
-3.30270290e-01 -2.44215518e-01 -7.60567188e-01 9.83025372e-01
8.96998763e-01 -5.81835151e-01 -1.10131703e-01 -4.01950300e-01
7.63636172e-01 -5.59346616e-01 1.07796453e-01 -5.53657591e-01
-4.03374434e-01 -1.60756767e-01 -4.33822311e-02 -1.60032302e-01
-2.37392843e-01 -3.70346129e-01 1.35033518e-01 4.72837627e-01
-3.41345608e-01 4.01511848e-01 3.99131268e-01 3.04576933e-01
-3.83430690e-01 -4.85910267e-01 6.94174051e-01 -3.51475358e-01
-1.08106148e+00 -9.54513028e-02 -6.68523431e-01 2.69913137e-01
6.05592072e-01 -2.49454841e-01 -2.37299293e-01 7.97205567e-02
-7.49840677e-01 -1.26543820e-01 7.41383851e-01 8.34294796e-01
-7.86408558e-02 -1.25227106e+00 -7.83603191e-01 9.29168612e-02
4.09003764e-01 -5.18659353e-01 -4.62379456e-02 7.20738232e-01
-1.31457984e-01 4.86360788e-01 -2.05139369e-02 -4.26393121e-01
-1.12920308e+00 5.46309948e-01 4.51061845e-01 -3.86413097e-01
-4.11230445e-01 6.10685885e-01 2.07335085e-01 -1.00087059e+00
-3.29990804e-01 -8.04238081e-01 -2.37555787e-01 6.52578950e-01
1.17733866e-01 1.38109982e-01 3.54178369e-01 -7.50806451e-01
-6.30778015e-01 6.98718071e-01 -6.18916303e-02 -1.63854621e-02
1.46639264e+00 -3.50642383e-01 -4.19459015e-01 1.06814194e+00
1.13580883e+00 5.09130061e-01 -7.45206475e-01 -4.60496664e-01
7.51250744e-01 -2.09413588e-01 -2.59583414e-01 -9.13147271e-01
-4.91144329e-01 7.35066831e-01 3.74419600e-01 2.15304479e-01
7.66938210e-01 2.65926689e-01 5.78506589e-01 2.10040346e-01
4.39152479e-01 -1.38423097e+00 -4.33626354e-01 8.85042429e-01
8.16458106e-01 -1.08334553e+00 -4.14499074e-01 -1.56412169e-01
-3.97650659e-01 8.56413305e-01 2.99895316e-01 -6.87687397e-02
3.18223387e-01 -1.54236943e-01 -3.41709331e-02 -2.69559443e-01
-8.84686232e-01 -5.03805757e-01 1.84863761e-01 3.73900503e-01
6.14278018e-01 3.98162156e-01 -9.07035112e-01 3.32471520e-01
-6.32524252e-01 -6.36153638e-01 3.80629480e-01 5.90294361e-01
-5.21517098e-01 -1.48119819e+00 -2.18512371e-01 4.99438703e-01
-5.48837841e-01 -3.49230200e-01 -6.91320240e-01 9.97235119e-01
2.96142638e-01 6.79232538e-01 3.94187897e-01 -8.56879652e-02
5.23822427e-01 6.28537595e-01 5.06614089e-01 -1.04276192e+00
-8.22927237e-01 -2.49797747e-01 4.03901845e-01 -6.26484096e-01
-3.11441004e-01 -1.17347884e+00 -1.07833087e+00 -5.66222109e-02
4.81089726e-02 -5.68113141e-02 5.50677955e-01 9.28051233e-01
1.43331155e-01 3.65906328e-01 3.75088423e-01 -3.67143929e-01
-4.76671815e-01 -1.26576376e+00 -5.53934813e-01 7.01916814e-01
-1.66723840e-02 -5.79453707e-01 -2.56373644e-01 -1.94374710e-01] | [10.42525577545166, 9.34833812713623] |
d4104b21-1b87-4b39-bf68-c2b3484914e1 | collision-aware-in-hand-6d-object-pose | 2301.13667 | null | https://arxiv.org/abs/2301.13667v1 | https://arxiv.org/pdf/2301.13667v1.pdf | Collision-aware In-hand 6D Object Pose Estimation using Multiple Vision-based Tactile Sensors | In this paper, we address the problem of estimating the in-hand 6D pose of an object in contact with multiple vision-based tactile sensors. We reason on the possible spatial configurations of the sensors along the object surface. Specifically, we filter contact hypotheses using geometric reasoning and a Convolutional Neural Network (CNN), trained on simulated object-agnostic images, to promote those that better comply with the actual tactile images from the sensors. We use the selected sensors configurations to optimize over the space of 6D poses using a Gradient Descent-based approach. We finally rank the obtained poses by penalizing those that are in collision with the sensors. We carry out experiments in simulation using the DIGIT vision-based sensor with several objects, from the standard YCB model set. The results demonstrate that our approach estimates object poses that are compatible with actual object-sensor contacts in $87.5\%$ of cases while reaching an average positional error in the order of $2$ centimeters. Our analysis also includes qualitative results of experiments with a real DIGIT sensor. | ['Lorenzo Natale', 'Fabrizio Bottarel', 'Nicola A. Piga', 'Gabriele M. Caddeo'] | 2023-01-31 | null | null | null | null | ['6d-pose-estimation'] | ['computer-vision'] | [ 4.16824102e-01 2.06216201e-01 2.43523687e-01 -2.88201779e-01
-7.76991010e-01 -4.85740989e-01 1.97495982e-01 1.72027871e-01
-6.30884171e-01 4.21267778e-01 -1.96259692e-01 1.04448147e-01
-4.75598335e-01 -4.53059405e-01 -1.00800240e+00 -1.76891774e-01
-1.34494258e-02 1.01204431e+00 3.41267556e-01 -2.24641506e-02
7.81637371e-01 8.14821780e-01 -1.72062778e+00 2.86612995e-02
5.73577285e-01 1.50740004e+00 4.80628133e-01 5.62840879e-01
3.23011011e-01 -2.71212980e-02 -4.77120340e-01 -1.64990097e-01
5.24268091e-01 2.45324656e-01 -2.99833953e-01 2.15949565e-01
5.57595253e-01 -4.15213287e-01 4.31298986e-02 9.76327598e-01
3.78383487e-01 -1.98020026e-01 8.36493552e-01 -1.03416431e+00
-5.16502708e-02 2.55892664e-01 -4.14301366e-01 -6.20852172e-01
8.06351423e-01 3.63922380e-02 7.08188891e-01 -1.05831921e+00
7.47131765e-01 1.26105821e+00 7.65585601e-01 4.33007538e-01
-1.01400387e+00 -1.31637573e-01 -4.42294627e-02 -3.61335725e-02
-1.48894668e+00 -3.30055058e-01 1.11521137e+00 -4.51928675e-01
7.78636634e-01 3.71418834e-01 5.20315111e-01 9.09923196e-01
3.34838480e-01 3.47287387e-01 9.48483169e-01 -6.35473788e-01
3.78320247e-01 1.43346675e-02 -2.16876119e-01 6.05867743e-01
4.05988812e-01 -9.17598307e-02 -5.00965655e-01 -1.66283906e-01
1.15631044e+00 -9.12688896e-02 -6.22011255e-03 -9.73796427e-01
-1.23268926e+00 4.29830968e-01 5.80219150e-01 7.49040991e-02
-6.06914878e-01 3.59261036e-01 -1.92108616e-01 -1.74591720e-01
3.36110815e-02 7.70667195e-01 -4.83275652e-01 -1.45921245e-01
-2.47796997e-01 3.79094034e-01 1.02763176e+00 9.83304262e-01
5.58168471e-01 -4.59047079e-01 1.89893588e-01 6.14950776e-01
5.02252340e-01 6.06651187e-01 -2.37556666e-01 -1.32069814e+00
6.29228950e-01 7.17411935e-01 7.22330153e-01 -1.12591672e+00
-3.36540550e-01 6.71702027e-02 -4.06000495e-01 7.80383408e-01
5.18409073e-01 -2.59416133e-01 -9.43004787e-01 1.10921395e+00
1.98953226e-01 -3.86261880e-01 -2.56848037e-01 1.23007822e+00
2.14534417e-01 1.87529922e-02 -3.66334021e-01 1.88528970e-01
1.05680239e+00 -4.00198936e-01 -3.80225062e-01 -3.69323194e-01
-1.16435155e-01 -8.87225091e-01 9.18670416e-01 6.89172149e-01
-1.09383929e+00 -6.25977993e-01 -1.30485511e+00 2.31534481e-01
-3.20087671e-01 5.84931612e-01 3.79057765e-01 2.52118576e-02
-7.81204104e-01 7.84256637e-01 -7.88203120e-01 -2.74386495e-01
1.23879850e-01 5.28560936e-01 -9.17480439e-02 1.65217653e-01
-6.15804732e-01 1.23887897e+00 1.24497168e-01 4.10464287e-01
-5.99094033e-01 -1.87928289e-01 -5.47261894e-01 -2.41712689e-01
5.57192326e-01 -2.71587849e-01 1.15441382e+00 -4.84786063e-01
-1.51594996e+00 9.01842117e-01 1.11919589e-01 -2.11902454e-01
9.23869312e-01 -6.98644638e-01 6.53348863e-02 3.95343214e-01
-1.45054758e-02 7.03107297e-01 7.48941958e-01 -1.92059493e+00
-3.44079614e-01 -5.37942708e-01 -1.43281799e-02 2.77198970e-01
1.05040334e-01 -6.12632155e-01 -6.07079685e-01 -1.41713142e-01
7.14858294e-01 -1.07198370e+00 -3.41499478e-01 6.76449478e-01
-8.70686352e-01 -1.83504269e-01 7.16832995e-01 -3.33004951e-01
4.41107154e-01 -1.82069528e+00 1.69654608e-01 7.87745237e-01
3.61254849e-02 -6.27172217e-02 1.09763116e-01 4.77709860e-01
4.69343901e-01 -2.77079403e-01 4.69110683e-02 -4.51834977e-01
4.78129648e-03 -2.35311198e-03 -1.37862405e-02 3.77153903e-01
5.53845707e-03 6.09400094e-01 -5.43723881e-01 -3.16804618e-01
4.16438311e-01 3.60026300e-01 -2.17267960e-01 2.62013793e-01
-5.63866258e-01 1.19918518e-01 -7.38494098e-01 8.25037539e-01
7.12440729e-01 -5.61893824e-03 2.37741277e-01 -6.41325474e-01
-2.38967359e-01 -1.08722150e-01 -1.51222980e+00 1.81123376e+00
-2.48909071e-01 2.63121426e-01 3.60723406e-01 -6.65619850e-01
1.33155942e+00 -9.95572731e-02 7.27617264e-01 -4.53916460e-01
2.59593278e-01 5.42857468e-01 -1.30665421e-01 -5.97072423e-01
4.35501665e-01 3.79894644e-01 -1.65336996e-01 1.85267523e-01
-3.41104507e-01 -6.04488492e-01 -2.86479115e-01 -3.92264992e-01
7.85217166e-01 2.78440446e-01 3.22315879e-02 -2.57528663e-01
-5.91484383e-02 4.10050154e-02 -9.07454342e-02 9.37430203e-01
1.27165705e-01 8.42604101e-01 1.66032314e-01 -1.97606921e-01
-1.18139374e+00 -1.41198933e+00 -5.39855585e-02 3.39533478e-01
8.36844385e-01 1.21408150e-01 -6.10341966e-01 -2.28763252e-01
5.90426326e-01 3.23458940e-01 -6.99840784e-01 2.90113315e-02
-6.64156675e-01 2.32145175e-01 4.41029705e-02 7.70855963e-01
3.56211752e-01 -9.50002134e-01 -1.65330803e+00 2.20400870e-01
1.83678642e-01 -9.99244928e-01 8.21227431e-02 5.57821691e-01
-7.37402856e-01 -1.30006087e+00 -7.19576955e-01 -6.62059784e-01
8.50163877e-01 -1.71191216e-01 8.13648522e-01 -2.72755265e-01
-6.10603750e-01 6.11059725e-01 -2.43366480e-01 -5.16162157e-01
3.11671980e-02 -1.92700028e-01 1.64593965e-01 -3.20474386e-01
2.48793915e-01 -2.36519501e-01 -7.24149525e-01 4.57939684e-01
-4.17414427e-01 -1.21532427e-02 6.17706180e-01 5.47565162e-01
7.58297324e-01 -2.36884519e-01 4.97948937e-03 -7.51865208e-02
5.98857760e-01 7.98386931e-02 -7.09172785e-01 1.91426024e-01
-2.16872543e-01 6.04093634e-02 1.58918515e-01 -7.30837405e-01
-7.12071896e-01 5.57600737e-01 1.30694687e-01 -5.90889573e-01
-1.97062358e-01 4.17506576e-01 -1.76927552e-01 -1.43986523e-01
7.11222351e-01 -2.17798218e-01 1.05571300e-01 -6.06106102e-01
5.37870005e-02 5.64856112e-01 4.92120415e-01 -7.13782966e-01
5.00691116e-01 5.31758249e-01 1.98316768e-01 -7.05755115e-01
-2.99800634e-01 -4.51159216e-02 -7.20276475e-01 -5.81166685e-01
8.13245833e-01 -2.88839430e-01 -1.34497666e+00 5.60130119e-01
-1.30907023e+00 -2.04403311e-01 -1.63975731e-01 6.18632972e-01
-9.14698482e-01 -5.79786208e-03 -3.41180563e-01 -1.21246624e+00
-3.59891988e-02 -1.34005499e+00 1.52994215e+00 2.00258389e-01
-6.34802520e-01 -2.66753584e-01 -3.50337654e-01 -6.47077512e-04
1.33059546e-01 4.82953221e-01 6.29739821e-01 -3.14818203e-01
-7.95395195e-01 -5.74046373e-01 -1.05348580e-01 -6.66109324e-02
2.28693143e-01 -8.52310434e-02 -8.24441969e-01 -1.87736586e-01
-2.07146153e-01 -4.49324727e-01 5.93109190e-01 5.92247128e-01
1.29163945e+00 5.90400845e-02 -7.40818918e-01 1.20324999e-01
1.61901402e+00 5.52949309e-01 5.26922405e-01 1.65036380e-01
6.60991788e-01 8.00112963e-01 1.00712752e+00 4.56561357e-01
-1.27405405e-01 9.24858868e-01 1.05892992e+00 1.29105866e-01
3.54893476e-01 -3.13620269e-01 -2.14804679e-01 4.81752567e-02
-9.53867584e-02 -2.15498134e-01 -9.29953694e-01 3.91684741e-01
-1.72529376e+00 -2.34952644e-01 1.69162393e-01 2.22018170e+00
4.71531242e-01 6.35449350e-01 -8.52760226e-02 1.38137072e-01
6.66576564e-01 -3.31893981e-01 -9.39898849e-01 -2.11793467e-01
2.92600274e-01 2.59769291e-01 6.77724898e-01 7.00249732e-01
-9.55615282e-01 7.29612410e-01 6.19390631e+00 2.20443264e-01
-1.19093418e+00 -6.52587891e-01 2.35790163e-01 2.54609399e-02
-9.68137905e-02 -2.57926881e-01 -6.31931484e-01 2.56271243e-01
1.40637487e-01 2.72340208e-01 2.76188850e-01 9.08163846e-01
-1.62804082e-01 -6.05532408e-01 -1.48648334e+00 1.01672494e+00
4.84396331e-02 -1.18130481e+00 -1.72277346e-01 -2.53406093e-02
2.79657125e-01 -1.82999521e-01 1.56946823e-01 -4.93352026e-01
4.20524701e-02 -9.95649397e-01 1.15925717e+00 1.00640142e+00
9.66547132e-01 -4.38990623e-01 5.42276621e-01 4.72672999e-01
-9.93237019e-01 2.40682764e-03 -2.58519024e-01 -1.11660108e-01
7.72794783e-02 4.40930426e-01 -1.13241363e+00 2.51826644e-01
9.05538380e-01 1.77087918e-01 -2.12490603e-01 9.54992652e-01
1.48090035e-01 -1.77717894e-01 -7.59207010e-01 -7.19198585e-01
-1.07983604e-01 -1.00170314e-01 5.61776757e-01 7.47472763e-01
5.88793874e-01 2.93169660e-03 1.06913336e-01 1.05468535e+00
2.09562242e-01 -4.02068555e-01 -7.15305507e-01 2.51227975e-01
8.92897904e-01 8.62182617e-01 -8.37963581e-01 -7.93600604e-02
3.01597953e-01 1.03660929e+00 7.83783272e-02 2.97316641e-01
-4.49152350e-01 -6.11261487e-01 4.56930876e-01 2.27854148e-01
5.16980648e-01 -4.12867665e-01 -7.06487834e-01 -6.68013215e-01
6.98278606e-01 -4.71582413e-01 -3.64389330e-01 -1.24239779e+00
-1.14193571e+00 3.69807363e-01 9.27383900e-02 -1.38878310e+00
-2.36785799e-01 -1.07144213e+00 -2.21165076e-01 9.64021444e-01
-6.79170966e-01 -8.25630665e-01 -5.55437744e-01 2.31986091e-01
3.02617222e-01 3.49192262e-01 8.19387019e-01 -2.68761933e-01
2.96009809e-01 3.08901161e-01 -2.06006587e-01 -8.77722427e-02
4.59939599e-01 -1.05258429e+00 1.21275008e-01 1.36715755e-01
-2.47973084e-01 6.97633505e-01 8.32000434e-01 -8.87386143e-01
-1.69030106e+00 -5.09683251e-01 6.47208869e-01 -4.14669812e-01
3.05398285e-01 -5.53228557e-01 -5.83984792e-01 4.35315222e-01
-1.54574722e-01 -3.25243801e-01 -2.13605836e-01 -1.33390188e-01
-4.77506295e-02 -2.77398258e-01 -1.47577024e+00 8.22024643e-01
1.26038432e+00 -3.97276759e-01 -6.24273300e-01 6.70999959e-02
2.66755700e-01 -7.90548146e-01 -7.18837559e-01 6.71887696e-01
1.10596597e+00 -8.16302955e-01 1.13039613e+00 -6.23619892e-02
3.52542132e-01 -2.08587259e-01 -4.18735713e-01 -1.23272192e+00
-1.59914084e-02 -2.65066266e-01 1.82830796e-01 5.91104507e-01
2.66634226e-01 -4.80260342e-01 1.19433725e+00 4.96383369e-01
-6.53840303e-02 -9.72344935e-01 -1.30611730e+00 -4.99221325e-01
-3.01949382e-01 -3.51271838e-01 2.91159749e-01 2.53544599e-01
5.79078645e-02 -1.68109447e-01 -1.01737522e-01 1.97595581e-01
8.23450089e-01 1.59451485e-01 6.47840917e-01 -1.48930395e+00
-1.85249761e-01 -4.01928812e-01 -5.36737084e-01 -1.26214886e+00
-1.72013819e-01 -5.74519373e-02 7.16529727e-01 -1.58828533e+00
-1.62800521e-01 -5.42166114e-01 7.02112466e-02 2.17696935e-01
2.51732886e-01 1.07683931e-02 5.37426285e-02 2.15957034e-02
-3.89458954e-01 5.51398993e-02 1.21965051e+00 -7.80433714e-02
-5.96971922e-02 -1.63116325e-02 -2.98086137e-01 9.31402147e-01
7.48377740e-01 3.20558771e-02 -2.56633759e-02 -5.70528090e-01
1.65547833e-01 3.60330641e-01 6.40414238e-01 -1.31126714e+00
2.59819984e-01 -1.10321037e-01 8.53578866e-01 -9.58574474e-01
1.04524922e+00 -1.37442350e+00 1.82617530e-01 7.81822801e-01
-5.80416620e-01 -1.33565962e-01 5.76874651e-02 2.64521718e-01
2.26331517e-01 -2.48571262e-01 5.44979990e-01 -1.61047816e-01
-4.67298090e-01 7.17655243e-03 -2.64450222e-01 -5.83553970e-01
9.73056912e-01 -7.08490074e-01 6.46927953e-02 -4.73386228e-01
-8.20873380e-01 8.98514241e-02 6.93505287e-01 2.81384230e-01
1.11705136e+00 -1.19967389e+00 -2.89207608e-01 3.28543305e-01
2.58369476e-01 8.40579271e-02 -2.88296312e-01 2.57130474e-01
-5.69659054e-01 2.82059431e-01 -2.12553695e-01 -9.95070338e-01
-1.11313212e+00 1.94430321e-01 4.48965013e-01 3.74666631e-01
-2.24362567e-01 8.48872602e-01 -5.46620011e-01 -3.65946084e-01
5.52379906e-01 -7.42097676e-01 6.64584264e-02 -3.97809207e-01
-2.17177361e-01 5.05881786e-01 6.26638010e-02 -2.05439359e-01
-5.12250185e-01 1.07489169e+00 3.11613351e-01 -4.41642165e-01
1.21075571e+00 1.77226722e-01 2.45697945e-01 4.41964805e-01
1.19609237e+00 1.06075287e-01 -1.92950940e+00 5.76999411e-02
1.19372390e-01 -4.12538201e-01 -5.72659433e-01 -1.08469045e+00
-6.07952297e-01 6.96712375e-01 8.63408983e-01 1.58982903e-01
7.33391702e-01 3.29587698e-01 3.49990010e-01 6.89419985e-01
8.01368594e-01 -1.36389732e+00 4.55771625e-01 4.07482475e-01
1.32583380e+00 -1.11813331e+00 -8.94018561e-02 -6.22292042e-01
-2.34532103e-01 1.32243323e+00 8.16546082e-01 -5.81340671e-01
6.52779043e-01 5.91482937e-01 2.35755309e-01 -4.05988097e-01
-1.03648029e-01 2.09120199e-01 4.50793356e-01 5.70214331e-01
7.45601160e-03 2.76569098e-01 -1.41054571e-01 4.13386337e-02
-1.14826426e-01 -6.48953095e-02 2.87842564e-02 1.41762686e+00
-7.60434449e-01 -6.52466655e-01 -6.14268243e-01 5.66849709e-01
-1.96091477e-02 5.04284978e-01 -7.75979578e-01 7.44927764e-01
2.08216146e-01 8.91129613e-01 4.32408750e-01 -5.85512459e-01
8.10973048e-01 -2.18189135e-01 1.09299171e+00 -3.55171412e-01
-8.89611915e-02 1.93751201e-01 1.17646322e-01 -7.76986361e-01
-1.48660794e-01 -6.72164679e-01 -1.30231416e+00 3.11663717e-01
-2.15019494e-01 -1.81877524e-01 1.05174172e+00 7.41874993e-01
2.81014234e-01 1.80257142e-01 4.56882715e-01 -1.52243280e+00
-7.85305142e-01 -1.00339544e+00 -4.91716951e-01 4.45863098e-01
2.10787207e-01 -1.04073834e+00 -2.56548762e-01 -2.20351964e-01] | [5.923403263092041, -0.9123615622520447] |
f4a00d43-1289-41cf-b1dc-293b4c4b7649 | towards-generalisable-video-moment-retrieval | 2303.0004 | null | https://arxiv.org/abs/2303.00040v2 | https://arxiv.org/pdf/2303.00040v2.pdf | Towards Generalisable Video Moment Retrieval: Visual-Dynamic Injection to Image-Text Pre-Training | The correlation between the vision and text is essential for video moment retrieval (VMR), however, existing methods heavily rely on separate pre-training feature extractors for visual and textual understanding. Without sufficient temporal boundary annotations, it is non-trivial to learn universal video-text alignments. In this work, we explore multi-modal correlations derived from large-scale image-text data to facilitate generalisable VMR. To address the limitations of image-text pre-training models on capturing the video changes, we propose a generic method, referred to as Visual-Dynamic Injection (VDI), to empower the model's understanding of video moments. Whilst existing VMR methods are focusing on building temporal-aware video features, being aware of the text descriptions about the temporal changes is also critical but originally overlooked in pre-training by matching static images with sentences. Therefore, we extract visual context and spatial dynamic information from video frames and explicitly enforce their alignments with the phrases describing video changes (e.g. verb). By doing so, the potentially relevant visual and motion patterns in videos are encoded in the corresponding text embeddings (injected) so to enable more accurate video-text alignments. We conduct extensive experiments on two VMR benchmark datasets (Charades-STA and ActivityNet-Captions) and achieve state-of-the-art performances. Especially, VDI yields notable advantages when being tested on the out-of-distribution splits where the testing samples involve novel scenes and vocabulary. | ['Yang Liu', 'Hailin Jin', 'Shaogang Gong', 'Jiabo Huang', 'Dezhao Luo'] | 2023-02-28 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Luo_Towards_Generalisable_Video_Moment_Retrieval_Visual-Dynamic_Injection_to_Image-Text_Pre-Training_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Luo_Towards_Generalisable_Video_Moment_Retrieval_Visual-Dynamic_Injection_to_Image-Text_Pre-Training_CVPR_2023_paper.pdf | cvpr-2023-1 | ['moment-retrieval'] | ['computer-vision'] | [ 2.92468905e-01 -5.53443670e-01 -4.71176237e-01 -3.52343529e-01
-6.10051334e-01 -6.57351196e-01 1.01799464e+00 -2.10079730e-01
-4.27606940e-01 2.84520209e-01 4.12029743e-01 -1.12327538e-01
9.24844593e-02 -2.34282255e-01 -8.37932229e-01 -5.79454839e-01
-9.15467292e-02 1.83383286e-01 1.43633112e-01 -1.00699149e-01
1.82040095e-01 3.04797322e-01 -1.54152393e+00 6.11154914e-01
5.18203378e-01 9.88492966e-01 5.20820618e-01 8.00820053e-01
-1.84218466e-01 9.18638229e-01 -3.00883532e-01 -2.97253162e-01
2.64908254e-01 -3.42224896e-01 -5.76396644e-01 3.37186545e-01
6.52633786e-01 -5.24374485e-01 -8.52427959e-01 8.33117187e-01
2.31155291e-01 2.79218644e-01 8.09899926e-01 -1.19730628e+00
-7.99254775e-01 2.92147517e-01 -7.35714138e-01 6.57365322e-01
6.22672081e-01 4.10891563e-01 1.15397847e+00 -1.05367649e+00
8.72606456e-01 1.16261470e+00 3.43592376e-01 5.81657648e-01
-1.13010585e+00 -6.07872963e-01 4.64224160e-01 7.46661007e-01
-1.30337834e+00 -6.42146111e-01 1.01424801e+00 -6.50070548e-01
1.17693150e+00 3.76651496e-01 7.81686246e-01 1.70860052e+00
1.33682683e-01 1.11335111e+00 7.37529695e-01 -1.40914962e-01
-1.56174630e-01 1.18227825e-02 -2.84227580e-01 3.93766731e-01
-2.34939829e-01 3.00793257e-02 -9.18853223e-01 3.96458000e-01
7.07466662e-01 1.49822965e-01 -5.87989330e-01 -4.33446854e-01
-1.49483860e+00 5.48680425e-01 9.16851312e-02 4.33493972e-01
-3.45622987e-01 1.65957600e-01 6.16409063e-01 2.46759459e-01
5.05827963e-01 8.42035189e-02 -4.31566209e-01 -4.67411786e-01
-1.10156083e+00 1.02690794e-02 3.15958798e-01 8.61235023e-01
6.85781956e-01 1.50224254e-01 -3.91797483e-01 7.15306878e-01
2.29562655e-01 7.68157184e-01 6.15452230e-01 -7.15578377e-01
8.44726384e-01 3.15139860e-01 -5.09765595e-02 -1.25641012e+00
8.63539577e-02 -1.08889356e-01 -6.54855967e-01 -4.71605927e-01
1.68591216e-01 4.59653229e-01 -1.19153893e+00 1.66127133e+00
9.07576904e-02 5.77308118e-01 1.42440259e-01 1.00857377e+00
8.01274419e-01 8.86053562e-01 2.35417619e-01 -3.47270787e-01
1.26530910e+00 -9.37208295e-01 -8.02770793e-01 -4.02945369e-01
3.87842774e-01 -7.78688788e-01 1.24297452e+00 6.73438013e-02
-8.28391552e-01 -6.86899781e-01 -7.54799485e-01 -1.90003991e-01
-2.84269571e-01 2.07663979e-02 3.50135028e-01 1.30706489e-01
-9.02735233e-01 2.92568326e-01 -1.00738955e+00 -4.68775690e-01
2.90237844e-01 1.60674468e-01 -6.23434007e-01 -2.88797975e-01
-1.10980082e+00 7.73323119e-01 2.48294532e-01 1.51537538e-01
-1.23062146e+00 -7.24368691e-01 -1.04394281e+00 -1.53379858e-01
4.04539198e-01 -6.13582790e-01 9.11655068e-01 -1.36239779e+00
-1.39073515e+00 8.88066709e-01 -4.22615230e-01 -3.37383121e-01
6.55064940e-01 -2.21495539e-01 -3.61999929e-01 6.46891534e-01
-1.40599571e-02 9.47436035e-01 1.40465462e+00 -1.25877666e+00
-5.77788115e-01 -1.11438893e-01 2.03780815e-01 3.81969929e-01
-6.28997326e-01 1.10942267e-01 -1.07762516e+00 -9.29110408e-01
-1.11720338e-01 -8.33883166e-01 1.99344605e-01 -7.51285255e-02
-1.56146616e-01 -1.91920679e-02 1.13663459e+00 -1.05786204e+00
1.13191843e+00 -2.16192341e+00 4.56185222e-01 -1.00998893e-01
7.40005672e-02 2.53968030e-01 -5.28114557e-01 2.89836228e-01
-1.27778560e-01 -5.66566437e-02 1.29670918e-01 -3.57989192e-01
-1.16802000e-01 3.69956434e-01 -6.71683431e-01 4.86027002e-01
4.25845832e-01 1.19348919e+00 -9.09561336e-01 -7.01006532e-01
6.07559502e-01 6.38957739e-01 -6.33902311e-01 3.28599930e-01
-4.36070651e-01 7.33426571e-01 -4.40645784e-01 7.70305753e-01
2.28424758e-01 -2.56637216e-01 2.54722219e-02 -6.32385373e-01
1.77700490e-01 4.04478535e-02 -6.47685111e-01 1.77990806e+00
-4.78546351e-01 1.09824622e+00 -3.28934819e-01 -1.16862822e+00
3.98730874e-01 4.21687752e-01 8.56053770e-01 -9.64292586e-01
7.49553442e-02 -1.90007433e-01 -1.66548535e-01 -9.22325552e-01
5.99956751e-01 2.50377476e-01 1.46162450e-01 1.44110680e-01
1.91757038e-01 2.38206699e-01 2.64028698e-01 3.26792717e-01
1.02239466e+00 4.19538170e-01 2.10279971e-01 2.70704448e-01
5.97626567e-01 -1.95992693e-01 4.48697627e-01 4.53284889e-01
-3.18826377e-01 7.09512472e-01 2.21194446e-01 -3.78663212e-01
-1.18979168e+00 -1.12643766e+00 3.84044908e-02 1.02004623e+00
2.21482724e-01 -6.80319846e-01 -3.36139500e-01 -6.87747896e-01
-1.31099433e-01 5.06361842e-01 -7.45691776e-01 1.80165633e-03
-5.51906347e-01 -3.36060345e-01 3.19610029e-01 5.38857639e-01
4.13747013e-01 -9.20138836e-01 -4.61703420e-01 3.09368912e-02
-7.22184002e-01 -1.77536750e+00 -7.26231039e-01 -1.97523087e-01
-6.24565482e-01 -9.44602668e-01 -6.62682533e-01 -5.76179445e-01
5.83687186e-01 6.05994225e-01 9.49696362e-01 -1.09087974e-01
-3.21883082e-01 8.98941815e-01 -6.64796770e-01 1.69473395e-01
-2.48590618e-01 -2.93783635e-01 5.07326350e-02 2.28106484e-01
3.20178509e-01 -5.95412493e-01 -6.91210687e-01 3.78367841e-01
-1.02850771e+00 3.66824478e-01 6.02894127e-01 8.06323528e-01
7.00436831e-01 -1.99539974e-01 7.01807588e-02 -4.07389104e-01
-2.08781753e-02 -4.68020558e-01 -2.52008438e-01 4.32361245e-01
-1.58039361e-01 -9.98724252e-02 6.45300865e-01 -9.91692722e-01
-9.18482959e-01 4.89796773e-02 1.42577499e-01 -1.24499679e+00
-7.64885917e-03 3.70940000e-01 -2.71295071e-01 7.43862838e-02
2.15567246e-01 6.81447506e-01 -3.47929984e-01 -1.33336723e-01
5.52305222e-01 5.65560758e-01 6.40647531e-01 -6.30510628e-01
1.02406669e+00 5.92232645e-01 -2.54185468e-01 -1.07010889e+00
-6.96035922e-01 -7.62154400e-01 -6.53176546e-01 -4.50724304e-01
1.17235684e+00 -1.12440324e+00 -3.26990336e-01 1.75046682e-01
-1.11550391e+00 -2.82887697e-01 5.82571886e-02 5.33663750e-01
-7.78596401e-01 6.41524017e-01 -4.72440898e-01 -4.49121565e-01
-1.88713834e-01 -1.20165789e+00 1.25993466e+00 -1.07124135e-01
-1.75386444e-01 -1.07712197e+00 -5.67908539e-03 5.53812444e-01
1.74098790e-01 3.64790112e-02 7.74107754e-01 -5.51832616e-01
-9.29986775e-01 -7.12191910e-02 -2.88905412e-01 3.05985093e-01
1.95701674e-01 1.76684365e-01 -1.01741028e+00 -4.21675831e-01
-4.57879454e-02 -2.16126248e-01 9.01901066e-01 2.76371032e-01
1.27750945e+00 -3.70030761e-01 -3.46057415e-01 8.97789717e-01
1.28096890e+00 8.26742798e-02 6.80245817e-01 4.56355751e-01
1.08223522e+00 4.94962990e-01 8.58802378e-01 3.49971086e-01
4.70386744e-01 9.61047828e-01 3.00708562e-01 1.54577479e-01
-1.69188738e-01 -3.66375685e-01 8.34704101e-01 1.11205304e+00
-1.84882224e-01 -3.69786441e-01 -8.41065764e-01 5.59736133e-01
-1.87374234e+00 -1.20670319e+00 2.97239780e-01 1.85492408e+00
9.14892435e-01 -9.73527133e-02 -2.38107726e-01 -1.17438458e-01
4.60135639e-01 7.26795733e-01 -5.44517100e-01 5.13403676e-02
-1.51326090e-01 -7.32205436e-02 3.08392942e-01 2.04674155e-01
-1.24288404e+00 1.14209557e+00 5.10646343e+00 9.79479790e-01
-1.29497075e+00 2.37642512e-01 3.87341201e-01 -4.03518826e-01
-3.14507186e-01 5.74892247e-03 -6.69179916e-01 4.65905577e-01
9.47160184e-01 3.08425119e-03 6.37208521e-01 5.03248394e-01
4.64823991e-01 7.09244534e-02 -1.43603897e+00 1.33683336e+00
4.06224251e-01 -1.23125935e+00 5.38298845e-01 -1.20147124e-01
7.51185298e-01 2.79163793e-02 2.32070923e-01 2.95615822e-01
-2.92078912e-01 -9.08744156e-01 9.44854021e-01 5.41300654e-01
1.03491521e+00 -3.78922284e-01 3.98432434e-01 2.94888988e-02
-1.48103738e+00 -5.80120273e-02 -2.64698863e-01 3.50664169e-01
2.70251781e-01 1.98546216e-01 -6.22275054e-01 7.49060452e-01
8.15544724e-01 1.44809532e+00 -8.20203722e-01 5.39363086e-01
-1.22219846e-01 5.21584749e-01 -1.15285270e-01 3.64185154e-01
2.68178999e-01 -1.11337654e-01 6.66557491e-01 1.24662912e+00
2.34725982e-01 -1.00142471e-01 2.42541641e-01 6.89335167e-01
-1.08780051e-02 9.00503770e-02 -7.49329448e-01 -5.19290686e-01
1.77461013e-01 1.17764342e+00 -6.61713004e-01 -2.40372762e-01
-7.90009022e-01 1.19405472e+00 2.09146246e-01 6.88883245e-01
-1.14794743e+00 1.32677525e-01 8.52251232e-01 -9.63085368e-02
5.80479681e-01 -4.48953718e-01 2.37738162e-01 -1.56645274e+00
1.97633058e-01 -1.15794241e+00 2.11182863e-01 -9.50113416e-01
-1.12832201e+00 4.55152988e-01 1.96394816e-01 -1.51913095e+00
-4.57576931e-01 -5.65834105e-01 -3.31718683e-01 4.33838338e-01
-1.60066509e+00 -1.55879593e+00 -4.64246273e-01 9.14106309e-01
9.82465684e-01 -6.72417358e-02 3.39796096e-01 5.09733975e-01
-4.93057907e-01 5.87102652e-01 -7.41768675e-03 2.51124769e-01
8.34282279e-01 -8.26782048e-01 2.30420917e-01 1.06203580e+00
6.74605072e-01 5.38120329e-01 7.29860663e-01 -6.62498891e-01
-1.81889677e+00 -1.27119029e+00 4.30179745e-01 -6.94100320e-01
9.51744497e-01 -3.99154007e-01 -9.46399689e-01 8.81418526e-01
1.69885471e-01 9.10985172e-02 4.57747817e-01 -3.25330615e-01
-5.13494730e-01 -8.03920478e-02 -3.89581621e-01 8.64256263e-01
1.19214475e+00 -1.18217027e+00 -6.68744683e-01 2.89797336e-01
7.27370381e-01 -3.90056908e-01 -7.55319476e-01 4.27985907e-01
5.40406764e-01 -6.41560912e-01 1.28149056e+00 -6.82169795e-01
7.29435802e-01 -3.30423057e-01 -4.73564923e-01 -9.58388925e-01
1.14752762e-01 -4.96966690e-01 -3.51959676e-01 1.36364675e+00
-5.37170470e-02 -1.25211105e-01 5.02560258e-01 3.00471514e-01
2.08175648e-02 -5.88029146e-01 -7.86857605e-01 -7.19273984e-01
-4.09861833e-01 -7.79595554e-01 8.78276750e-02 1.20041955e+00
-2.33770937e-01 1.97452664e-01 -6.80766046e-01 1.80644140e-01
3.89969468e-01 4.55327295e-02 9.31111813e-01 -6.03273392e-01
-3.35882872e-01 -3.23368996e-01 -6.64240658e-01 -1.29680824e+00
5.47240794e-01 -9.37451124e-01 1.04873836e-01 -1.25476110e+00
5.52545369e-01 7.22313076e-02 -3.41588974e-01 3.87906164e-01
-2.05951154e-01 2.83012062e-01 3.03959966e-01 4.44423735e-01
-9.04433131e-01 7.84992099e-01 1.33749700e+00 -4.33840632e-01
-7.44749680e-02 -4.41989630e-01 7.77034974e-03 5.91184616e-01
3.80713284e-01 -2.76390404e-01 -6.26567841e-01 -5.32275021e-01
4.35699634e-02 1.75452173e-01 5.84260583e-01 -8.57208550e-01
1.17432073e-01 -3.39781642e-01 4.18143630e-01 -7.87646115e-01
5.11285007e-01 -8.05699348e-01 2.50276685e-01 4.01143625e-04
-3.55035394e-01 1.74378201e-01 2.06682160e-01 9.30781245e-01
-4.58864272e-01 8.84779543e-02 4.72909629e-01 6.33461922e-02
-1.15621936e+00 6.71289861e-01 -2.43322983e-01 1.83339491e-01
1.05249536e+00 -3.87998134e-01 -3.56061876e-01 -5.23753643e-01
-4.96022165e-01 2.34104320e-01 7.40773857e-01 9.11742985e-01
8.93280745e-01 -1.20471072e+00 -3.40162814e-01 2.26436883e-01
4.59417343e-01 -3.20521533e-01 4.98517007e-01 1.02724719e+00
-3.75463784e-01 6.29977584e-01 -2.56677508e-01 -1.00905585e+00
-1.41352677e+00 7.75272548e-01 2.19405010e-01 -7.67455250e-02
-8.59010935e-01 5.13366163e-01 6.41325712e-01 2.81438112e-01
3.33836347e-01 -4.16369617e-01 -2.47542933e-01 1.33508340e-01
5.46329796e-01 -1.57561749e-01 -2.82638371e-01 -1.06934774e+00
-3.10365438e-01 8.64682615e-01 -1.74585849e-01 -1.30796477e-01
1.15841150e+00 -4.31439161e-01 2.10499689e-02 5.23280561e-01
1.46259594e+00 -5.61099127e-02 -1.55178785e+00 -4.02787715e-01
-9.34219267e-03 -7.81630933e-01 5.22514922e-04 -3.49336952e-01
-1.15220881e+00 1.09379542e+00 5.89751422e-01 -3.22651595e-01
1.18179858e+00 1.70842856e-02 8.32412481e-01 4.04830277e-01
1.76723897e-01 -9.37440395e-01 6.27230585e-01 4.40607518e-01
8.45821679e-01 -1.34032881e+00 -6.07149452e-02 -1.68968290e-01
-7.72881806e-01 1.11336660e+00 6.11642420e-01 2.44672611e-01
3.20250005e-01 -1.48052722e-01 5.16748540e-02 5.45814559e-02
-9.82981920e-01 -3.51973563e-01 7.56113350e-01 6.07994914e-01
8.99292678e-02 -3.00823778e-01 1.40530095e-01 3.53949249e-01
1.24836586e-01 -1.63999960e-01 2.36541897e-01 7.53688157e-01
-6.14147373e-02 -8.94780457e-01 -8.32967311e-02 2.68973351e-01
-5.23750663e-01 -1.92380264e-01 -2.13497147e-01 9.58555937e-01
-1.75432965e-01 6.84726477e-01 2.02300146e-01 -6.46900237e-01
-1.81112252e-02 -2.51787566e-02 5.73645592e-01 -4.37364459e-01
-1.30583823e-01 1.63134769e-01 -3.32252420e-02 -9.12286401e-01
-7.89285064e-01 -6.19599879e-01 -1.08549917e+00 -1.45485908e-01
5.56724779e-02 -2.57310480e-01 5.24195254e-01 1.14826536e+00
1.79954141e-01 6.12642884e-01 5.52710652e-01 -1.14057994e+00
-2.37768784e-01 -8.26274455e-01 -1.52743906e-01 9.17101264e-01
4.77402389e-01 -8.32331359e-01 -4.29465652e-01 5.26750982e-01] | [10.080577850341797, 0.8318478465080261] |
ce0e07e0-4279-448b-a55f-302bac300b4b | minimum-efforts-to-build-an-end-to-end | 2206.03064 | null | https://arxiv.org/abs/2206.03064v2 | https://arxiv.org/pdf/2206.03064v2.pdf | A Simple and Efficient Pipeline to Build an End-to-End Spatial-Temporal Action Detector | Spatial-temporal action detection is a vital part of video understanding. Current spatial-temporal action detection methods mostly use an object detector to obtain person candidates and classify these person candidates into different action categories. So-called two-stage methods are heavy and hard to apply in real-world applications. Some existing methods build one-stage pipelines, But a large performance drop exists with the vanilla one-stage pipeline and extra classification modules are needed to achieve comparable performance. In this paper, we explore a simple and effective pipeline to build a strong one-stage spatial-temporal action detector. The pipeline is composed by two parts: one is a simple end-to-end spatial-temporal action detector. The proposed end-to-end detector has minor architecture changes to current proposal-based detectors and does not add extra action classification modules. The other part is a novel labeling strategy to utilize unlabeled frames in sparse annotated data. We named our model as SE-STAD. The proposed SE-STAD achieves around 2% mAP boost and around 80% FLOPs reduction. Our code will be released at https://github.com/4paradigm-CV/SE-STAD. | ['Feng Han', 'Lixin Gu', 'Chen-Lin Zhang', 'Lin Sui'] | 2022-06-07 | null | null | null | null | ['action-classification'] | ['computer-vision'] | [ 1.05424426e-01 -2.15410382e-01 -2.33576223e-01 -3.90395373e-01
-6.25321925e-01 -2.94156432e-01 5.18179774e-01 -1.62415013e-01
-6.06031299e-01 3.01617980e-01 4.19402719e-01 1.42813474e-01
4.82587725e-01 -4.42395180e-01 -3.55881304e-01 -4.21620786e-01
1.71541512e-01 2.18888476e-01 1.24338424e+00 -1.30832912e-02
1.03266843e-01 2.16347054e-01 -1.46234477e+00 6.60706103e-01
4.94066983e-01 7.28309929e-01 1.32290855e-01 8.44016433e-01
4.55359779e-02 1.00941682e+00 -4.00010467e-01 -2.63963163e-01
4.71814692e-01 -7.01143384e-01 -8.49402308e-01 2.63556421e-01
4.21728671e-01 -6.77862823e-01 -6.02928400e-01 8.63175154e-01
5.80389738e-01 9.85591635e-02 5.66421561e-02 -1.38255310e+00
-1.45184889e-01 4.13670927e-01 -8.72964025e-01 5.30185640e-01
5.83446801e-01 4.93518859e-01 5.90519845e-01 -9.67349052e-01
3.96370053e-01 1.45855641e+00 7.85678744e-01 6.67420626e-01
-7.84433722e-01 -5.09391725e-01 3.81847799e-01 4.21868533e-01
-1.45439208e+00 -6.27219200e-01 4.23201889e-01 -5.32051027e-01
8.91658962e-01 9.42330137e-02 8.56943846e-01 9.25548613e-01
-1.22853093e-01 1.40804696e+00 9.51944828e-01 -6.51925430e-02
1.20335482e-01 -1.27213299e-01 2.35064268e-01 8.39729905e-01
1.54337704e-01 -2.14551389e-01 -5.62443972e-01 1.81982201e-02
8.84297371e-01 3.54233295e-01 -1.41705545e-02 -1.82758510e-01
-1.33034241e+00 5.06806433e-01 5.31836510e-01 2.66016692e-01
-3.16025585e-01 2.34964028e-01 5.67462087e-01 -1.56097159e-01
4.28254813e-01 -1.01008102e-01 -3.69781882e-01 -4.56661463e-01
-1.13242590e+00 2.83227772e-01 3.93147022e-01 9.10255969e-01
4.67299312e-01 -2.64127702e-01 -6.58463359e-01 5.41697204e-01
2.90928274e-01 3.53711188e-01 3.88688833e-01 -1.00868857e+00
3.84484828e-01 7.84783781e-01 2.94059187e-01 -7.46038735e-01
-5.37325859e-01 -1.46448553e-01 -5.74793279e-01 1.25482082e-01
4.49413598e-01 -1.31720439e-01 -1.01120651e+00 1.34813702e+00
6.12476408e-01 5.90472817e-01 -3.47421855e-01 1.25503886e+00
8.64575744e-01 4.60041940e-01 3.03762406e-01 2.89150309e-02
1.59355962e+00 -1.65639830e+00 -5.95515609e-01 -6.23158753e-01
7.35168695e-01 -6.06334150e-01 8.63740027e-01 7.98045471e-02
-1.15131950e+00 -7.52203941e-01 -9.48104560e-01 -2.68544883e-01
-1.70635954e-01 6.33027315e-01 6.24153733e-01 5.87878466e-01
-1.04361534e+00 3.32070827e-01 -1.14696181e+00 -6.31371140e-01
6.52246952e-01 1.79192722e-01 -4.65684235e-01 -8.71725082e-02
-8.79119158e-01 6.60873055e-01 4.11780685e-01 3.40302959e-02
-8.12302768e-01 -2.48622492e-01 -8.36157799e-01 -1.51935518e-01
7.06137598e-01 -6.69400334e-01 1.55921650e+00 -1.04414451e+00
-1.51185572e+00 8.79428566e-01 -6.43540382e-01 -4.17917669e-01
7.83400595e-01 -4.59371984e-01 -2.85366356e-01 3.70118976e-01
3.82414579e-01 7.61959612e-01 6.68111086e-01 -5.10362446e-01
-1.11942363e+00 -2.89742023e-01 3.82917235e-03 3.26279223e-01
-1.98193848e-01 5.09924233e-01 -1.16803491e+00 -6.83889925e-01
9.39153135e-02 -1.01893556e+00 -3.56030196e-01 3.61845016e-01
-1.48241818e-01 -4.77128208e-01 8.21016014e-01 -6.53264761e-01
1.28803360e+00 -2.10287023e+00 -1.79232389e-01 -3.98554623e-01
2.54034668e-01 6.45290911e-01 -8.77091140e-02 2.44354680e-01
1.59925282e-01 -1.77026466e-01 -1.41131252e-01 -5.31338274e-01
-1.09017052e-01 -2.20371932e-01 2.74249583e-01 5.82375586e-01
3.91414016e-01 8.77292395e-01 -1.00301695e+00 -5.96781731e-01
4.16596949e-01 3.29861552e-01 -5.09077013e-01 1.38347879e-01
6.89693540e-02 4.18236226e-01 -5.77758491e-01 8.72278452e-01
6.39667928e-01 -2.90853143e-01 1.05188847e-01 -6.68068528e-02
-3.53792161e-01 4.13933963e-01 -1.41199386e+00 1.89768708e+00
3.01696032e-01 5.85820436e-01 7.93998130e-03 -9.39662755e-01
5.01600564e-01 2.24405542e-01 6.67644620e-01 -4.89022851e-01
1.58683777e-01 1.73422229e-02 6.03255108e-02 -5.92496276e-01
3.68228853e-01 1.92659155e-01 -7.69173503e-02 2.67457306e-01
-3.92491892e-02 7.37108588e-01 5.97134471e-01 3.08573425e-01
1.51836002e+00 5.93483984e-01 4.58809137e-01 -1.07997268e-01
6.83056712e-01 1.33815169e-01 1.17525589e+00 6.05070531e-01
-8.67995203e-01 7.66906202e-01 4.12760824e-01 -7.35730767e-01
-8.53484094e-01 -6.63914800e-01 3.32734257e-01 1.22184348e+00
3.31974179e-01 -7.42922425e-01 -7.64351130e-01 -9.15998578e-01
-2.23476499e-01 2.05878049e-01 -4.16980028e-01 1.17329754e-01
-7.99769878e-01 -6.21074557e-01 6.64581120e-01 9.70342278e-01
1.11397243e+00 -1.00518835e+00 -9.19845343e-01 1.83041587e-01
-3.05485427e-01 -1.29228032e+00 -7.18406558e-01 -3.92998248e-01
-6.07094944e-01 -1.22523677e+00 -8.72819304e-01 -6.99269235e-01
5.25782228e-01 7.26892829e-01 9.36865032e-01 -5.39681176e-03
-3.26262772e-01 2.01535195e-01 -5.86462617e-01 -3.35745394e-01
1.78199098e-01 -4.48999293e-02 1.34447217e-01 2.90416837e-01
9.35769379e-01 -9.97351184e-02 -1.19601285e+00 7.00056612e-01
-3.73120874e-01 3.13223422e-01 7.88092196e-01 2.74154425e-01
6.51975930e-01 -5.11995330e-02 1.54399246e-01 -4.68418539e-01
-2.28552651e-02 -3.13760757e-01 -3.20825398e-01 1.37970671e-01
-7.81738013e-02 -1.42047629e-01 2.40661889e-01 -4.43284005e-01
-1.07787538e+00 7.73720324e-01 -2.44924173e-01 -3.91886473e-01
-3.87636602e-01 -2.94810999e-02 -2.09791481e-01 1.93801776e-01
5.07858813e-01 1.44000158e-01 -1.46744281e-01 -7.12340951e-01
1.61755666e-01 6.60950601e-01 4.43604290e-01 -5.13902418e-02
5.62062025e-01 7.63048530e-01 -3.52976501e-01 -6.63283169e-01
-9.46851909e-01 -9.92710829e-01 -9.15536582e-01 -4.13395464e-01
1.13983452e+00 -1.23310280e+00 -5.94704926e-01 8.19268405e-01
-9.68642592e-01 -6.13914847e-01 -1.10292539e-01 5.98148584e-01
-3.28092992e-01 4.59059238e-01 -6.51149631e-01 -8.18204343e-01
-3.39552730e-01 -1.08707094e+00 1.36025572e+00 4.54684943e-01
-1.31630763e-01 -3.95706266e-01 2.98430137e-02 5.82606852e-01
2.32496604e-01 5.27588166e-02 -3.27144742e-01 -4.22454894e-01
-5.20181894e-01 -2.46567458e-01 -2.74300873e-01 1.08715251e-01
-3.69340368e-02 -2.11735845e-01 -8.36819470e-01 -2.70536155e-01
-2.09343895e-01 -1.84981719e-01 1.16160989e+00 5.38110077e-01
1.04022658e+00 2.40865219e-02 -6.36952758e-01 4.81163472e-01
9.38054442e-01 1.30145341e-01 7.97060609e-01 2.97921717e-01
7.65190363e-01 4.10695404e-01 9.85280871e-01 3.66698533e-01
4.16674137e-01 1.01421082e+00 -1.23332277e-01 -2.44265109e-01
-4.48872387e-01 -2.73794055e-01 8.82525861e-01 2.61369795e-01
-3.02961767e-01 -1.83430776e-01 -9.78660583e-01 5.54733694e-01
-2.29883385e+00 -1.36298954e+00 -5.06302595e-01 2.01042271e+00
5.43100893e-01 2.83955634e-01 7.99918771e-01 1.05217129e-01
8.66797805e-01 1.50207549e-01 -4.93782908e-01 2.01898962e-01
1.36079267e-01 -1.20017536e-01 4.54153359e-01 2.86952496e-01
-1.76200640e+00 1.18147135e+00 5.72188282e+00 6.88404799e-01
-7.70247817e-01 3.42060715e-01 4.71098661e-01 -4.11548883e-01
6.67877972e-01 9.14210547e-03 -1.15900719e+00 4.96409625e-01
7.35732317e-01 1.59144685e-01 -3.19924712e-01 1.00603724e+00
6.19642437e-01 -4.40047532e-01 -1.02166998e+00 1.20505881e+00
1.57220960e-01 -1.06299448e+00 -3.20336401e-01 -1.78902954e-01
2.62971967e-01 4.87808138e-02 -5.21079123e-01 3.71921420e-01
1.37779236e-01 -5.47527790e-01 8.06309164e-01 3.81465793e-01
7.08415508e-01 -4.37359899e-01 7.80646384e-01 3.35299581e-01
-1.81926394e+00 -2.59493440e-01 -2.82021850e-01 -3.52882892e-01
5.96425831e-01 3.38369071e-01 -4.15210783e-01 2.46586218e-01
1.02672255e+00 1.14763546e+00 -8.65871489e-01 1.33077550e+00
-3.57827485e-01 6.32441640e-01 -2.02420309e-01 8.81106332e-02
2.74113059e-01 -2.96553317e-02 4.55862761e-01 1.49770141e+00
1.98955327e-01 4.62515146e-01 7.31753588e-01 2.52894014e-01
2.23735034e-01 -2.19919980e-01 -2.08103865e-01 1.55838847e-01
2.35318363e-01 1.30526161e+00 -1.02307940e+00 -7.96223879e-01
-8.12733173e-01 1.44889605e+00 1.56390816e-01 -5.95370829e-02
-1.21234465e+00 -1.94041044e-01 7.04127848e-01 5.16536415e-01
3.38430971e-01 -1.59640700e-01 -7.81576112e-02 -1.33322787e+00
1.58147186e-01 -7.90909171e-01 5.97631335e-01 -5.39591134e-01
-1.00125360e+00 3.42504770e-01 2.09478028e-02 -1.42584658e+00
2.17596367e-01 -5.90203822e-01 -5.18585682e-01 4.97629702e-01
-1.09432566e+00 -1.35010493e+00 -6.89376712e-01 7.49386668e-01
1.08764136e+00 -6.66834116e-02 5.10377944e-01 6.05041802e-01
-9.63004053e-01 5.25405884e-01 -5.56363702e-01 6.30766928e-01
8.04549634e-01 -1.11102617e+00 5.99303603e-01 1.43874824e+00
-5.52651547e-02 3.21195871e-01 4.03042465e-01 -7.86654472e-01
-1.08895528e+00 -1.30667567e+00 8.70140254e-01 -4.96930689e-01
3.71244818e-01 -4.28519487e-01 -6.69949532e-01 7.41609454e-01
6.93451092e-02 3.07664067e-01 6.27513587e-01 -1.17140740e-01
-1.65581837e-01 -6.63448274e-02 -9.47005808e-01 5.81746817e-01
1.49928796e+00 -7.94683993e-02 -5.36130846e-01 4.35887724e-01
3.61513227e-01 -2.82491684e-01 -3.26094747e-01 2.60639012e-01
6.06585920e-01 -1.11623657e+00 8.12072039e-01 -4.02897060e-01
1.63214743e-01 -8.47707033e-01 2.44343176e-01 -6.50333226e-01
-8.52515340e-01 -6.41266346e-01 -3.76843482e-01 8.73042345e-01
7.55507350e-02 -3.16784978e-01 9.65927720e-01 5.64888239e-01
-2.19124526e-01 -5.43790996e-01 -6.90411925e-01 -8.65436971e-01
-5.17422080e-01 -2.34174356e-01 5.28239049e-02 6.60481453e-01
3.64484638e-02 4.16874945e-01 -5.63181400e-01 1.53962269e-01
6.37457013e-01 -1.59284875e-01 1.02198684e+00 -9.27603066e-01
-4.62918341e-01 -4.17297751e-01 -7.11468279e-01 -1.50667596e+00
-4.14200068e-01 -5.42627692e-01 2.14298174e-01 -1.75000370e+00
6.30960107e-01 -1.06589593e-01 -1.90904737e-01 8.14823508e-01
-4.51593667e-01 5.23719311e-01 1.88053712e-01 2.54115641e-01
-1.20508325e+00 4.61784989e-01 8.17172229e-01 -2.75690518e-02
-3.53200644e-01 1.16443276e-01 -3.73468250e-01 1.01663721e+00
1.04365778e+00 -4.48514491e-01 -4.25303727e-01 -4.33107287e-01
-2.91900426e-01 -3.36210608e-01 6.08029127e-01 -1.48675060e+00
3.44615221e-01 -2.11654812e-01 5.56524634e-01 -7.83357859e-01
4.45421875e-01 -4.41040009e-01 -6.67114854e-02 7.59073853e-01
-3.74559574e-02 3.38691063e-02 -3.20788324e-02 5.21397054e-01
-1.26192942e-01 1.35196537e-01 9.48614419e-01 -3.51103872e-01
-1.28026104e+00 3.95730913e-01 -5.52703261e-01 -1.67148054e-01
1.54039419e+00 -3.75289977e-01 -1.71222389e-01 -1.50238946e-01
-6.63119316e-01 4.20641333e-01 3.64689261e-01 5.62629640e-01
5.93586922e-01 -1.26984656e+00 -9.46501791e-01 -8.08498412e-02
1.42859504e-01 -1.73208937e-01 4.26724762e-01 1.20326936e+00
-7.24279940e-01 4.22608554e-01 -2.43723437e-01 -6.67324364e-01
-1.65296221e+00 5.55007041e-01 3.49076420e-01 -2.65198618e-01
-9.29943025e-01 1.01346064e+00 2.00752288e-01 3.91204059e-02
1.74750879e-01 -6.12331517e-02 -1.82666078e-01 -7.47671276e-02
1.07768047e+00 5.66076577e-01 -3.90160739e-01 -8.15130830e-01
-8.53392303e-01 3.29290688e-01 5.67696691e-02 -1.52199164e-01
1.15330982e+00 2.28026853e-04 1.20833039e-01 2.26902574e-01
7.92149186e-01 -3.51013869e-01 -1.53846240e+00 -1.79008022e-01
5.47922328e-02 -6.50960743e-01 -6.38170615e-02 -5.07002115e-01
-1.08465374e+00 7.29752362e-01 7.68293262e-01 -5.44339307e-02
1.27243376e+00 9.53629464e-02 8.43797266e-01 9.28698853e-02
3.83346081e-01 -1.34262347e+00 2.32266203e-01 4.53557968e-01
7.39205778e-01 -1.36185229e+00 2.99318105e-01 -6.81354105e-01
-7.56761611e-01 7.92387962e-01 1.04368687e+00 -8.73538107e-02
4.53182727e-01 2.09875017e-01 -1.29967436e-01 -2.01744542e-01
-5.24730861e-01 -6.99967563e-01 3.10285121e-01 3.41580808e-01
4.73235339e-01 -3.89947370e-02 -6.71183586e-01 4.87315685e-01
3.46783429e-01 2.48846978e-01 1.86289370e-01 1.26100361e+00
-6.11932695e-01 -1.07171655e+00 -2.88543075e-01 2.27930844e-01
-4.80840862e-01 2.44643569e-01 -4.54012424e-01 6.57060742e-01
4.43832636e-01 1.11803496e+00 1.53627992e-01 -5.32116890e-01
3.32805365e-01 7.97862560e-02 1.90090433e-01 -6.59910619e-01
-4.80124235e-01 2.11698458e-01 1.86499342e-01 -1.19018781e+00
-8.77826691e-01 -9.28936720e-01 -1.33748996e+00 -2.58249223e-01
1.61263789e-03 -2.45979458e-01 1.79971114e-01 8.48990619e-01
6.55604839e-01 3.80380422e-01 2.16301322e-01 -9.38402236e-01
6.70885071e-02 -1.13499415e+00 -1.42806917e-01 4.08753574e-01
-4.71159257e-02 -7.29402721e-01 1.26473293e-01 3.30853969e-01] | [8.331151962280273, 0.4585148096084595] |
43375f68-96f7-4df5-acf0-00b55c6515bb | graph-transformation-policy-network-for | 1812.09441 | null | http://arxiv.org/abs/1812.09441v1 | http://arxiv.org/pdf/1812.09441v1.pdf | Graph Transformation Policy Network for Chemical Reaction Prediction | We address a fundamental problem in chemistry known as chemical reaction
product prediction. Our main insight is that the input reactant and reagent
molecules can be jointly represented as a graph, and the process of generating
product molecules from reactant molecules can be formulated as a sequence of
graph transformations. To this end, we propose Graph Transformation Policy
Network (GTPN) -- a novel generic method that combines the strengths of graph
neural networks and reinforcement learning to learn the reactions directly from
data with minimal chemical knowledge. Compared to previous methods, GTPN has
some appealing properties such as: end-to-end learning, and making no
assumption about the length or the order of graph transformations. In order to
guide model search through the complex discrete space of sets of bond changes
effectively, we extend the standard policy gradient loss by adding useful
constraints. Evaluation results show that GTPN improves the top-1 accuracy over
the current state-of-the-art method by about 3% on the large USPTO dataset. Our
model's performances and prediction errors are also analyzed carefully in the
paper. | ['Truyen Tran', 'Svetha Venkatesh', 'Kien Do'] | 2018-12-22 | null | https://openreview.net/forum?id=r1f78iAcFm | https://openreview.net/pdf?id=r1f78iAcFm | null | ['chemical-reaction-prediction'] | ['medical'] | [ 4.77197975e-01 2.99556226e-01 -6.38848901e-01 -1.41148180e-01
-5.02530515e-01 -6.56971633e-01 7.90326357e-01 3.59580606e-01
-2.26028755e-01 1.05105436e+00 -1.10486872e-01 -7.61070549e-01
1.19096719e-01 -8.58276010e-01 -1.16241848e+00 -9.82137561e-01
2.54971418e-03 5.93430996e-01 1.36086434e-01 -1.58314332e-01
3.26686174e-01 8.66994023e-01 -7.69722164e-01 1.67338196e-02
1.06140518e+00 7.57844567e-01 9.48213711e-02 4.15739894e-01
-1.27735272e-01 1.01271522e+00 -1.35813490e-01 -4.53096688e-01
2.39501581e-01 -5.77117383e-01 -8.49670172e-01 -3.81271780e-01
2.56986432e-02 1.51019365e-01 -4.91756648e-01 1.02941501e+00
5.17157435e-01 5.20618439e-01 1.11028361e+00 -8.66848111e-01
-6.76224113e-01 6.60618424e-01 -2.94895083e-01 -2.19596833e-01
2.38909155e-01 2.98628271e-01 1.25308096e+00 -7.30006993e-01
8.90093684e-01 1.31651092e+00 2.54606903e-01 6.53346956e-01
-1.40137470e+00 -7.00851202e-01 5.13321280e-01 2.27341637e-01
-1.00267446e+00 -2.65586644e-01 9.38491523e-01 -5.20169437e-01
1.38212204e+00 5.74024692e-02 6.34976685e-01 1.10009217e+00
4.85016495e-01 5.92676580e-01 6.69433534e-01 -3.38084757e-01
3.97066832e-01 -2.45314509e-01 -1.79509148e-01 1.01667774e+00
1.87914595e-01 3.83608013e-01 -2.88317353e-01 -9.54803079e-02
6.70891643e-01 1.19185872e-01 -4.12925705e-02 -6.26633406e-01
-9.04759169e-01 1.14988148e+00 7.55214512e-01 -1.30467966e-01
-3.42577308e-01 2.81118244e-01 2.38161802e-01 6.71533868e-02
4.90785599e-01 1.03410947e+00 -4.09575999e-01 2.70459473e-01
-3.63126665e-01 3.66467506e-01 9.40331161e-01 8.03170145e-01
9.25730705e-01 7.38812834e-02 -1.21243797e-01 3.56142879e-01
4.16792393e-01 6.90259784e-02 5.72477654e-02 -6.22986615e-01
6.59564614e-01 5.09898067e-01 2.58810282e-01 -5.05132616e-01
-4.97315437e-01 -2.47162715e-01 -8.29654098e-01 1.08551778e-01
4.23758298e-01 -3.66668701e-01 -1.11360300e+00 1.66378689e+00
4.55981374e-01 1.96914226e-02 1.00830741e-01 5.23550868e-01
5.18169820e-01 1.15333092e+00 3.71165514e-01 -6.26694083e-01
7.26457953e-01 -1.33344781e+00 -7.17626512e-01 -2.22840503e-01
7.87241638e-01 -4.73352194e-01 7.20542490e-01 4.34151113e-01
-1.18541527e+00 -3.84101361e-01 -1.19365203e+00 -2.69257486e-01
-7.98663735e-01 -2.42460519e-01 1.11180663e+00 2.96885103e-01
-5.54616332e-01 1.33188236e+00 -7.96784222e-01 -9.01388936e-03
2.73340613e-01 6.38114810e-01 -1.88006505e-01 -6.36107177e-02
-1.26954806e+00 8.84866238e-01 7.62370408e-01 8.49670842e-02
-1.11088502e+00 -9.79446650e-01 -8.22234511e-01 -7.96138588e-03
8.51934731e-01 -8.44594300e-01 1.35979056e+00 -4.95207042e-01
-2.00666833e+00 2.47504622e-01 -2.01608032e-01 -3.64272088e-01
6.33082926e-01 -8.97335336e-02 -1.94231704e-01 -8.27518031e-02
-2.36018509e-01 4.83844429e-01 5.08024037e-01 -9.10347819e-01
-4.60081667e-01 -2.00862259e-01 4.64511029e-02 2.34470099e-01
2.65595585e-01 -3.09847653e-01 -2.67405421e-01 -3.76053780e-01
-1.16025381e-01 -8.90907466e-01 -7.07915545e-01 -1.74134031e-01
-6.94504857e-01 -5.89246690e-01 4.34993148e-01 -2.18509987e-01
1.11135650e+00 -1.62679088e+00 4.45909798e-01 4.05761898e-01
3.05261731e-01 2.79356211e-01 -1.01176687e-01 7.65438080e-01
-4.85520661e-01 2.93007523e-01 -1.21720992e-01 1.49641568e-02
3.19665596e-02 1.68656744e-02 -2.41463810e-01 3.62961411e-01
3.01466227e-01 1.14577103e+00 -1.23346663e+00 -1.60581637e-02
1.24825232e-01 3.26467335e-01 -4.71990883e-01 2.32803419e-01
-8.88240516e-01 4.07853305e-01 -7.16484964e-01 4.38201189e-01
2.99193889e-01 -4.62345511e-01 4.70342994e-01 -3.35686713e-01
-1.85206741e-01 5.77599287e-01 -9.01029110e-01 1.76629388e+00
-7.32251629e-02 -3.14078517e-02 -4.88620132e-01 -1.16014946e+00
9.53454614e-01 3.24386746e-01 4.97095615e-01 -6.20398164e-01
7.60315657e-02 1.35386661e-01 1.55740783e-01 -1.31191686e-01
6.13966919e-02 -2.87515700e-01 1.82102293e-01 2.06825331e-01
-1.37546249e-02 -2.56179363e-01 5.55620670e-01 1.46112377e-02
7.43741512e-01 2.90067941e-01 5.68278372e-01 -4.46977913e-02
5.42413592e-01 -1.16611719e-01 5.90126157e-01 7.40929902e-01
1.69489875e-01 -1.64497755e-02 7.90007532e-01 -7.05142260e-01
-1.12583828e+00 -8.22319567e-01 3.86638492e-01 9.02416289e-01
-4.72586267e-02 -4.52214926e-01 -6.21477604e-01 -9.04067755e-01
1.54434636e-01 6.89548492e-01 -5.88318348e-01 -4.71688092e-01
-4.70653236e-01 -8.01693678e-01 2.94794086e-02 5.06359220e-01
1.77054763e-01 -1.11739731e+00 4.22376990e-01 7.00748086e-01
3.85894150e-01 -8.06847394e-01 -7.61113167e-01 5.22895157e-01
-8.13922286e-01 -1.28216922e+00 -3.22675347e-01 -8.82646561e-01
5.74518085e-01 -4.12736041e-03 8.72934163e-01 -4.05466348e-01
-3.51958632e-01 -2.68363267e-01 1.46876201e-02 -5.68655729e-01
-7.31116414e-01 2.56922573e-01 -1.58212990e-01 -6.31006733e-02
9.07330215e-02 -5.69667220e-01 -6.57285392e-01 -4.08561006e-02
-6.10552847e-01 1.53502241e-01 5.95585465e-01 7.25300908e-01
8.21157277e-01 -5.47850020e-02 4.67900842e-01 -1.38985336e+00
5.76348424e-01 -1.74885973e-01 -1.02573800e+00 4.76744890e-01
-1.10387003e+00 7.00805187e-01 1.02224159e+00 -5.17429650e-01
-1.01058686e+00 4.53202188e-01 6.61975518e-02 -3.22232544e-01
1.84840530e-01 5.54137409e-01 -4.06481832e-01 -1.39834598e-01
6.01252973e-01 1.91701114e-01 -1.15342818e-01 -4.16913122e-01
6.58922911e-01 -1.75495837e-02 2.14884594e-01 -8.28867614e-01
7.02015579e-01 8.83027017e-02 6.68732941e-01 -4.54925746e-01
-8.70217144e-01 -3.28552693e-01 -4.83232588e-01 2.29508340e-01
9.14702296e-01 -6.95376098e-01 -1.41171718e+00 1.42888993e-01
-1.27025390e+00 -5.59229910e-01 -4.51910757e-02 4.59755898e-01
-6.45698309e-01 5.55537462e-01 -9.05180514e-01 -6.85860455e-01
-5.23055315e-01 -1.30499220e+00 5.75165808e-01 9.61273760e-02
1.44444078e-01 -1.17944169e+00 1.76506609e-01 -2.52482388e-03
-8.47676545e-02 5.02623796e-01 1.31455290e+00 -6.40027583e-01
-6.74852788e-01 -6.26370460e-02 -4.55650873e-03 1.28983513e-01
3.10858876e-01 1.79014161e-01 -6.15516245e-01 -3.75043839e-01
-4.53228623e-01 -1.87709689e-01 1.12971938e+00 4.32452768e-01
1.18321311e+00 -4.41465795e-01 -5.41839123e-01 6.32352769e-01
1.32577717e+00 9.37066197e-01 4.99743462e-01 1.16514571e-01
1.12295282e+00 4.15946066e-01 3.82887065e-01 2.39039794e-01
-1.16226748e-01 3.50702256e-01 5.60726345e-01 -2.37046108e-01
1.61760956e-01 -8.44527721e-01 3.41197997e-01 3.95850152e-01
-2.90953338e-01 -5.85654557e-01 -5.15585840e-01 -1.98041946e-01
-2.06893373e+00 -8.85554969e-01 1.74210981e-01 2.05660033e+00
9.15032208e-01 4.30877984e-01 3.52714509e-01 -3.17749977e-02
6.19091988e-01 2.25179061e-01 -1.06924272e+00 -4.25143152e-01
2.46938512e-01 3.17405522e-01 9.17077899e-01 8.43299329e-01
-1.06084394e+00 1.35101974e+00 7.15059614e+00 8.61576676e-01
-1.31553853e+00 -4.13685620e-01 7.46182024e-01 1.29999042e-01
-1.71394244e-01 5.38240671e-02 -1.00910342e+00 2.69750774e-01
8.24855030e-01 -2.87743688e-01 7.50981688e-01 7.29770720e-01
4.41513449e-01 5.64353526e-01 -1.52106011e+00 7.40037978e-01
-4.90598977e-01 -1.89388788e+00 3.56407195e-01 1.15996838e-01
7.15334177e-01 -1.99501663e-01 3.59350964e-02 2.62614876e-01
7.59298921e-01 -1.07797015e+00 5.03538966e-01 4.37971294e-01
6.09101892e-01 -9.38879669e-01 9.08952430e-02 1.20006405e-01
-1.15353739e+00 5.45681007e-02 -4.62024033e-01 1.10993274e-02
-8.67900439e-03 4.42346543e-01 -1.21698892e+00 8.08121085e-01
-1.78277642e-01 1.07745397e+00 -1.58360861e-02 7.44652450e-01
-4.84082133e-01 6.15740001e-01 -1.81425259e-01 -5.05666852e-01
5.27891457e-01 -7.93390751e-01 1.06138755e-02 1.06929290e+00
1.40328631e-01 1.33725017e-01 4.43066478e-01 1.12513387e+00
-4.22047168e-01 2.19450563e-01 -7.46553779e-01 -5.38966060e-01
2.98590094e-01 9.39890027e-01 -4.64893132e-01 -2.16367036e-01
-3.59572738e-01 7.93479323e-01 5.99451125e-01 7.28019416e-01
-7.78287113e-01 -3.82134050e-01 5.40711999e-01 -4.85747233e-02
5.07492483e-01 -1.04081042e-01 2.35725507e-01 -8.27390015e-01
-1.74696654e-01 -6.51969850e-01 3.46997350e-01 -4.00883317e-01
-1.19502342e+00 1.61249056e-01 -3.61709893e-01 -7.41589487e-01
-1.61190972e-01 -1.04015982e+00 -5.45952260e-01 7.49601007e-01
-1.72924161e+00 -8.76382530e-01 4.52525228e-01 3.97520930e-01
4.47013319e-01 -8.36177766e-02 7.28216827e-01 -6.77881315e-02
-9.65487480e-01 5.25153279e-01 4.18510258e-01 -1.71377957e-01
5.94359398e-01 -1.51105547e+00 7.49909997e-01 6.00217938e-01
-8.01190510e-02 6.17531002e-01 6.28614545e-01 -7.45778799e-01
-1.65691364e+00 -1.25788391e+00 7.10343540e-01 -1.52691990e-01
8.11763823e-01 -5.60334444e-01 -8.15611541e-01 5.44513226e-01
-6.72117397e-02 1.00123934e-01 4.67238188e-01 5.87195717e-02
-5.15074134e-01 -2.19935939e-01 -9.76155102e-01 8.89763117e-01
1.16002786e+00 -3.43948752e-01 -9.16544572e-02 9.24716413e-01
1.14517415e+00 -4.76700127e-01 -1.06061363e+00 1.93517044e-01
2.37956330e-01 -3.76163632e-01 1.14502621e+00 -1.36573970e+00
3.95845145e-01 -2.31176347e-01 2.10891247e-01 -1.32346225e+00
-7.55497038e-01 -1.35886574e+00 -3.37724298e-01 5.95609426e-01
9.00679410e-01 -7.66759098e-01 9.64955270e-01 6.73041046e-01
-2.67891616e-01 -8.73886228e-01 -4.69141036e-01 -8.64434719e-01
2.60047376e-01 1.33184552e-01 5.69779634e-01 8.53262126e-01
2.43448496e-01 1.05036175e+00 -4.49447930e-01 1.08291328e-01
4.91497457e-01 -1.18464250e-02 3.89056146e-01 -1.14254594e+00
-4.30085480e-01 -5.55599749e-01 5.77969439e-02 -1.13023746e+00
3.35169196e-01 -1.08054471e+00 -1.59079507e-01 -1.54316235e+00
1.04341721e-02 -5.49516305e-02 -5.63327432e-01 5.20319045e-01
-3.01534176e-01 -5.71578145e-01 1.44991308e-01 -1.50182247e-01
-6.69862270e-01 7.27658570e-01 1.77501988e+00 -4.22235191e-01
-3.81212503e-01 6.02282137e-02 -7.17787147e-01 1.85684234e-01
7.46646166e-01 -4.35718685e-01 -6.34558439e-01 2.51821250e-01
4.22836721e-01 2.94829845e-01 -3.15833926e-01 -6.11644685e-01
2.53693879e-01 -6.74065471e-01 2.43832871e-01 -5.02374113e-01
3.00409079e-01 -4.92393047e-01 2.78982759e-01 8.29214275e-01
-6.08267784e-01 -1.52802050e-01 2.12655459e-02 8.05098832e-01
2.12926671e-01 -1.02889980e-03 6.82471454e-01 -3.13563913e-01
-2.85078049e-01 9.00986552e-01 -2.70382017e-01 -2.54123718e-01
1.01653552e+00 1.19063944e-01 -2.88159996e-01 -1.73314840e-01
-7.47653961e-01 1.88313857e-01 2.52974957e-01 1.49036348e-01
2.75556237e-01 -1.22112954e+00 -3.29928815e-01 -2.89607272e-02
-5.50063513e-02 1.94399640e-01 -3.05111915e-01 2.67428398e-01
-5.39958596e-01 7.33472407e-01 2.33676150e-01 1.01931253e-02
-8.71411204e-01 1.16902006e+00 5.69055915e-01 -6.58517540e-01
-2.87751675e-01 6.70329988e-01 5.35894573e-01 -3.70065123e-01
2.70686805e-01 -6.10691249e-01 6.45737872e-02 -2.40554452e-01
2.37777397e-01 2.41610363e-01 4.84567918e-02 1.94486007e-01
2.04205960e-02 4.63180125e-01 -7.43107021e-01 3.82939160e-01
1.42374182e+00 6.15910590e-01 1.00337997e-01 2.21542940e-01
1.29086888e+00 -4.16521519e-01 -1.56296074e+00 -2.42491215e-01
-3.18411998e-02 8.08061659e-02 -1.04502983e-01 -8.11881542e-01
-8.56022179e-01 7.91891336e-01 2.71737963e-01 1.18179426e-01
5.03872395e-01 -1.52711064e-01 5.87530971e-01 1.05348563e+00
-8.06261301e-02 -1.04714012e+00 2.86002606e-01 5.30430079e-01
6.66984856e-01 -1.11378574e+00 1.47071630e-01 -6.89698994e-01
-3.41331899e-01 1.25921750e+00 3.80921602e-01 -6.58289716e-02
4.95368749e-01 -1.88831508e-01 -2.82829136e-01 -2.04015538e-01
-7.64447391e-01 4.98171970e-02 2.78628498e-01 4.54058707e-01
5.08797944e-01 5.83791062e-02 -3.89672875e-01 2.42187187e-01
5.00211492e-02 -1.26310423e-01 1.92947924e-01 8.94167721e-01
-4.73163933e-01 -1.49056649e+00 9.81916264e-02 2.42352948e-01
-3.81244808e-01 -1.44917116e-01 -6.59933746e-01 8.27185750e-01
-3.15718830e-01 7.70890653e-01 -5.23298085e-01 -1.63450360e-01
4.85420376e-01 9.30991098e-02 6.84247553e-01 -5.28863907e-01
-3.36964995e-01 1.33865073e-01 5.09068519e-02 -6.28026366e-01
-2.60747522e-01 -4.45941240e-01 -1.31442261e+00 -3.70002359e-01
-3.92738223e-01 4.68020350e-01 5.38047612e-01 8.57836187e-01
3.50910693e-01 6.78160429e-01 8.54925811e-01 -8.16289723e-01
-7.55629003e-01 -7.06147730e-01 -6.43681705e-01 2.81716079e-01
2.77074516e-01 -6.18223548e-01 -2.24451423e-01 -9.28421915e-02] | [4.528529167175293, 6.081355094909668] |
0e1633f9-2493-468a-81de-ffef290d54a9 | differentiable-spline-approximations | 2110.01532 | null | https://arxiv.org/abs/2110.01532v1 | https://arxiv.org/pdf/2110.01532v1.pdf | Differentiable Spline Approximations | The paradigm of differentiable programming has significantly enhanced the scope of machine learning via the judicious use of gradient-based optimization. However, standard differentiable programming methods (such as autodiff) typically require that the machine learning models be differentiable, limiting their applicability. Our goal in this paper is to use a new, principled approach to extend gradient-based optimization to functions well modeled by splines, which encompass a large family of piecewise polynomial models. We derive the form of the (weak) Jacobian of such functions and show that it exhibits a block-sparse structure that can be computed implicitly and efficiently. Overall, we show that leveraging this redesigned Jacobian in the form of a differentiable "layer" in predictive models leads to improved performance in diverse applications such as image segmentation, 3D point cloud reconstruction, and finite element analysis. | ['Chinmay Hegde', 'Adarsh Krishnamurthy', 'Baskar Ganapathysubramanian', 'Soumik Sarkar', 'Biswajit Khara', 'Anjana Deva Prasad', 'Ameya Joshi', 'Aditya Balu', 'Minsu Cho'] | 2021-10-04 | null | http://proceedings.neurips.cc/paper/2021/hash/a952ddeda0b7e2c20744e52e728e5594-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/a952ddeda0b7e2c20744e52e728e5594-Paper.pdf | neurips-2021-12 | ['3d-point-cloud-reconstruction', 'point-cloud-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 7.52316264e-04 2.03145608e-01 -3.32329810e-01 -3.48708183e-01
-8.15985560e-01 -4.47426617e-01 5.39898992e-01 -1.18631519e-01
4.58250828e-02 7.36722529e-01 -1.49459571e-01 -4.52823192e-01
-1.22884624e-01 -5.59770644e-01 -1.17682755e+00 -7.15480387e-01
-4.90052216e-02 3.73736382e-01 -1.19887084e-01 -3.47765386e-01
3.30475509e-01 9.08466280e-01 -1.06651723e+00 -2.40074337e-01
1.08249342e+00 8.73352587e-01 -3.32449287e-01 4.81290221e-01
-9.06878337e-02 3.68556738e-01 8.96981210e-02 -1.38616189e-01
3.95805418e-01 -4.66094278e-02 -7.66847372e-01 5.85288592e-02
6.76841676e-01 -2.94946373e-01 -2.08359629e-01 6.85526252e-01
4.11228426e-02 4.43282843e-01 8.24949503e-01 -1.12704837e+00
-5.82849443e-01 1.41858533e-01 -5.11076629e-01 -2.48350456e-01
1.46494120e-01 1.68318719e-01 1.03405905e+00 -1.17321801e+00
4.80040401e-01 1.06292033e+00 1.44508004e+00 3.38290215e-01
-1.80073869e+00 -1.67554513e-01 2.73197711e-01 -2.73395151e-01
-1.16087663e+00 -2.20670283e-01 1.03707874e+00 -7.74779797e-01
8.51784527e-01 4.46183443e-01 8.33035052e-01 4.03397799e-01
1.35330155e-01 7.12556422e-01 1.13862514e+00 -4.69106466e-01
1.06672935e-01 4.17567603e-02 2.03910708e-01 9.20760572e-01
-1.69407323e-01 2.38302201e-01 -2.49250218e-01 -5.19496262e-01
1.13648355e+00 6.71361908e-02 -3.13385367e-01 -7.78037429e-01
-9.25210893e-01 1.08372819e+00 5.57489753e-01 -1.34369601e-02
-3.80601764e-01 5.88038743e-01 2.40693733e-01 8.06763172e-02
9.20860410e-01 5.85106671e-01 -6.53470993e-01 -1.49114624e-01
-1.08223319e+00 3.39332908e-01 1.01517415e+00 9.26038742e-01
1.11492300e+00 4.03614879e-01 1.15289621e-01 9.72387254e-01
4.78024274e-01 4.13281381e-01 -9.16099027e-02 -1.35854375e+00
7.97892809e-02 3.41006994e-01 2.62745738e-01 -1.17084658e+00
-2.10591823e-01 -5.22720635e-01 -6.51882172e-01 4.79715943e-01
3.94241750e-01 -1.19997405e-01 -7.37469733e-01 1.37987018e+00
6.38559401e-01 3.83962572e-01 -4.53900367e-01 7.25912094e-01
3.52066666e-01 7.00334907e-01 -2.63780896e-02 -8.46944526e-02
4.73373860e-01 -1.12967777e+00 -1.33312374e-01 6.96427673e-02
5.07661641e-01 -5.57562828e-01 1.18733060e+00 3.40495199e-01
-1.38053596e+00 -3.47113192e-01 -9.31848228e-01 -3.03651631e-01
3.26423831e-02 -1.25151783e-01 9.79541361e-01 4.20872748e-01
-1.20626831e+00 1.24500632e+00 -9.54223096e-01 2.31421828e-01
6.94688499e-01 4.20700610e-01 -1.34104177e-01 7.64325410e-02
-5.29825985e-01 9.14565206e-01 -2.45549351e-01 2.41974905e-01
-7.07497358e-01 -1.47702634e+00 -7.74759173e-01 5.62691726e-02
6.68773279e-02 -9.35572743e-01 1.24653089e+00 -1.16358900e+00
-1.68191350e+00 7.82843769e-01 -4.13071722e-01 -3.14588010e-01
7.54340172e-01 -1.70809582e-01 3.16666156e-01 9.67509076e-02
-2.61738211e-01 4.63094831e-01 1.25227892e+00 -1.43181038e+00
-2.66206831e-01 -2.07103357e-01 -4.39831689e-02 -5.80210872e-02
-1.08620510e-01 -2.07992479e-01 -2.33804733e-01 -6.53072596e-01
1.43836215e-01 -9.92267728e-01 -6.51530445e-01 5.64284921e-01
-1.66768789e-01 -2.01020241e-01 6.74106538e-01 -8.75974298e-01
8.67809713e-01 -2.04246092e+00 2.76446164e-01 5.40525615e-01
2.21386760e-01 -6.71081915e-02 2.03491852e-01 3.60065669e-01
3.00606620e-02 1.81884244e-01 -6.99123442e-01 -5.68589747e-01
9.04572681e-02 4.55884278e-01 -3.53219658e-01 7.30817080e-01
3.37207407e-01 9.72696006e-01 -8.89821291e-01 -4.36359286e-01
2.41826251e-01 7.94601083e-01 -8.42184186e-01 4.87050675e-02
-4.78346825e-01 5.85430562e-01 -6.55384123e-01 5.56155264e-01
7.22012639e-01 -3.03485572e-01 -2.02593550e-01 -1.17538661e-01
-4.12879527e-01 1.55221432e-01 -8.49574029e-01 1.65036106e+00
-6.17604792e-01 6.65229201e-01 5.56088567e-01 -1.27819180e+00
7.84355640e-01 1.07885689e-01 9.55140412e-01 4.81752641e-02
-3.11465472e-01 3.77928078e-01 -4.87596601e-01 -2.32145235e-01
2.32117206e-01 -2.32579038e-01 4.45563942e-01 -7.46528851e-03
-1.56144634e-01 -5.23532987e-01 -1.83078632e-01 -9.36931521e-02
7.44843423e-01 6.02871120e-01 -6.49043471e-02 -6.01633251e-01
4.27969664e-01 5.24661303e-01 2.98054069e-01 6.05623007e-01
1.96804121e-01 5.55246711e-01 4.36775357e-01 -3.58498812e-01
-1.29149127e+00 -8.87031436e-01 -6.08775675e-01 9.33458865e-01
-3.11364532e-01 -3.23314294e-02 -6.08513832e-01 -4.13178325e-01
5.85628748e-01 5.55158675e-01 -4.78024513e-01 4.13445756e-02
-9.18280125e-01 -4.62455571e-01 3.56300324e-01 5.95811427e-01
8.46959949e-02 -5.23279130e-01 -1.40837863e-01 3.28431427e-01
5.01575351e-01 -7.22652316e-01 -6.42684460e-01 1.46098822e-01
-1.49179959e+00 -8.88498724e-01 -8.04688573e-01 -7.52971530e-01
7.06137240e-01 -6.72606677e-02 1.22553658e+00 2.16955081e-01
-6.22289404e-02 8.72973025e-01 3.00381452e-01 -9.76004601e-02
-4.22415912e-01 -1.07010871e-01 -3.34994972e-01 1.84619706e-02
-4.24197227e-01 -8.98850679e-01 -5.07746160e-01 5.87084778e-02
-5.17286539e-01 3.19681838e-02 2.77951151e-01 1.02907908e+00
1.06442308e+00 -5.23301065e-01 3.67265910e-01 -9.22700226e-01
5.47169924e-01 -5.72866917e-01 -9.36659217e-01 1.22023985e-01
-9.01401937e-01 9.46931541e-02 7.69004703e-01 -5.02071321e-01
-9.68077362e-01 1.91674054e-01 -3.54932606e-01 -6.48376167e-01
2.12040469e-01 7.82834709e-01 3.05400550e-01 -9.62747693e-01
4.89971548e-01 1.40746951e-01 4.01747435e-01 -6.45084500e-01
4.74830717e-01 5.60028180e-02 5.30278563e-01 -9.81249034e-01
7.30080247e-01 6.58510923e-01 5.55883527e-01 -1.12020898e+00
-5.11025548e-01 -4.63905096e-01 -4.18102145e-01 -1.43670827e-01
3.54598224e-01 -5.41876793e-01 -6.58399463e-01 2.65199631e-01
-1.11469936e+00 -6.93908393e-01 -5.48491478e-01 1.46899536e-01
-9.03799891e-01 4.48760837e-01 -5.62064826e-01 -7.35976636e-01
-3.56693238e-01 -9.91718709e-01 1.00821650e+00 -3.39712165e-02
-6.63933009e-02 -1.47736311e+00 1.11336939e-01 -2.29006186e-02
5.28133631e-01 5.18172681e-01 1.09285033e+00 -1.86062023e-01
-7.16263533e-01 -1.98256031e-01 -5.96806267e-03 6.74280226e-01
-1.36818752e-01 5.36159396e-01 -7.26815403e-01 -2.43518911e-02
1.26484260e-01 8.56020376e-02 6.81074977e-01 7.61192799e-01
1.42519462e+00 -4.37256783e-01 -4.13232774e-01 1.27199435e+00
1.60926056e+00 -3.32160950e-01 4.21949983e-01 2.61101890e-02
9.37509477e-01 4.47780758e-01 2.53826261e-01 2.01155320e-01
2.53468931e-01 6.34649158e-01 2.58917153e-01 -2.33680502e-01
6.55295253e-02 -2.40407914e-01 4.89952937e-02 8.43990684e-01
-3.46248180e-01 7.37275481e-01 -9.15980875e-01 4.06818867e-01
-1.91771662e+00 -6.27330184e-01 -5.28790712e-01 1.99245906e+00
1.16517770e+00 -1.64707661e-01 7.99502730e-02 -1.36765957e-01
3.05621713e-01 -9.16443169e-02 -7.07300782e-01 -6.85759962e-01
1.80016980e-01 4.99577016e-01 7.21486866e-01 7.73378849e-01
-1.00752842e+00 5.73798060e-01 7.81480932e+00 8.08869421e-01
-1.34522033e+00 3.79583798e-04 4.87761915e-01 1.14531092e-01
-7.01537967e-01 1.90771043e-01 -6.50508106e-01 4.08669114e-01
6.82820916e-01 -1.41098678e-01 6.62537634e-01 1.21490943e+00
3.93453985e-01 2.05598816e-01 -1.12489223e+00 6.22784615e-01
-3.67017448e-01 -1.78395367e+00 -3.60362947e-01 1.36956172e-02
1.05054855e+00 1.33622557e-01 1.35176450e-01 2.31755927e-01
1.98870152e-01 -1.33817363e+00 5.37952662e-01 8.65046859e-01
6.18361771e-01 -4.92560714e-01 -7.43917823e-02 3.51198703e-01
-6.34266138e-01 2.24748105e-01 -2.01182380e-01 2.85438430e-02
1.56044677e-01 8.12434196e-01 -6.28435969e-01 2.07145125e-01
4.21303838e-01 9.20727551e-01 -1.96078625e-02 1.25188792e+00
6.26595467e-02 8.78874063e-01 -6.43889546e-01 5.58525175e-02
2.83698767e-01 -7.39277482e-01 8.36056352e-01 1.09547424e+00
5.09860255e-02 -4.89418507e-02 1.84101135e-01 1.26006591e+00
-3.55475470e-02 9.73695442e-02 -4.61141437e-01 1.43725529e-01
1.00410186e-01 1.17916453e+00 -1.96002394e-01 -9.94358435e-02
-5.59993327e-01 3.68185401e-01 4.15512919e-01 6.77744985e-01
-6.40913725e-01 -1.33908605e-02 7.11627722e-01 3.18807453e-01
6.03538573e-01 -8.31962347e-01 -9.50351536e-01 -8.89001727e-01
2.02316880e-01 -7.60033548e-01 -7.36142471e-02 -6.78080857e-01
-1.45243430e+00 1.02385633e-01 1.46136791e-01 -7.79520988e-01
-2.32809424e-01 -7.79898465e-01 -7.08999157e-01 1.11376417e+00
-1.63357675e+00 -1.25349689e+00 -2.60708574e-02 5.09912848e-01
2.59089977e-01 3.12080622e-01 6.97187662e-01 2.12311208e-01
-2.56466478e-01 3.82908434e-01 6.23744786e-01 -2.17464566e-01
2.10448280e-01 -1.42535543e+00 1.92318678e-01 4.27806020e-01
-1.15089282e-01 6.91350222e-01 7.42915869e-01 -5.87873101e-01
-1.87532842e+00 -8.34647179e-01 4.15733010e-01 -3.72497976e-01
7.03206718e-01 1.03062131e-01 -1.13464224e+00 8.74676943e-01
-2.43645310e-01 3.59469235e-01 2.18586490e-01 2.27360979e-01
-8.13623667e-02 -1.15853839e-01 -1.25785875e+00 5.47771871e-01
7.53035247e-01 -4.80695635e-01 -1.83529541e-01 3.87505502e-01
6.43737435e-01 -6.04535341e-01 -1.20313108e+00 4.09826815e-01
5.16671121e-01 -5.70599854e-01 1.36026275e+00 -9.97042537e-01
5.86665869e-01 2.40519829e-02 2.02110154e-03 -1.37292695e+00
-1.62168875e-01 -1.06016803e+00 -5.83852232e-01 7.76583672e-01
3.14315051e-01 -9.36121941e-01 8.01902294e-01 1.19823134e+00
-7.04408944e-01 -1.37164712e+00 -8.97339404e-01 -7.06200957e-01
7.03584254e-01 -2.34055474e-01 3.28547269e-01 9.44152653e-01
-1.83789760e-01 -2.79535979e-01 -1.38149917e-01 -3.70451994e-02
7.16679633e-01 1.27223894e-01 6.95279956e-01 -1.26478612e+00
-4.55528319e-01 -5.96091032e-01 -1.25580087e-01 -1.54268885e+00
3.38126719e-01 -1.04985178e+00 6.85069412e-02 -1.30742288e+00
-4.00957108e-01 -1.16757643e+00 1.95975795e-01 2.09073320e-01
-4.23627719e-02 -9.37186480e-02 -1.91740796e-01 4.29461896e-01
2.20741868e-01 5.56564331e-01 1.49369955e+00 -5.40746599e-02
-1.69113755e-01 3.13786864e-01 -5.20782948e-01 9.23948467e-01
5.48795938e-01 -3.35541636e-01 -2.45966874e-02 -5.18415272e-01
1.57573536e-01 1.25881836e-01 7.17853308e-01 -4.88475144e-01
1.42396465e-01 -4.94763047e-01 2.55687833e-01 -2.87493318e-01
5.53198338e-01 -8.58416319e-01 1.31182760e-01 1.22421421e-01
-3.70922625e-01 -1.02493443e-01 2.65940547e-01 4.86051947e-01
-1.11633211e-01 -5.09076893e-01 7.81168342e-01 -1.06034778e-01
-3.62062871e-01 4.90006715e-01 8.26148465e-02 9.08768252e-02
7.42316782e-01 -2.05336109e-01 1.92494467e-01 -2.07697615e-01
-6.94508731e-01 5.80510162e-02 6.62188232e-01 -3.99197757e-01
5.23887277e-01 -1.22447741e+00 -6.66013896e-01 1.40529022e-01
-6.05208457e-01 3.15855503e-01 -7.40092471e-02 1.13927090e+00
-6.91835344e-01 2.03282908e-01 2.33915478e-01 -8.56888354e-01
-7.50435174e-01 2.42018968e-01 7.33201623e-01 -1.32033780e-01
-7.73704290e-01 7.68322766e-01 -2.65302062e-01 -4.12426144e-01
5.99843450e-02 -5.39957762e-01 4.44246769e-01 -4.75828767e-01
-2.06837595e-01 6.14548981e-01 8.55679661e-02 -4.02205586e-01
-7.34372810e-02 6.18146896e-01 3.95852238e-01 6.95025325e-02
1.58679163e+00 2.18766063e-01 -3.42899919e-01 3.43336284e-01
1.47573304e+00 1.90409511e-01 -1.80737138e+00 6.24051914e-02
-1.20542586e-01 -5.75702846e-01 2.11627200e-01 -3.56991738e-01
-9.86814141e-01 7.86146164e-01 -6.70437589e-02 3.53172809e-01
6.43234074e-01 -9.05796364e-02 7.88469493e-01 2.99893916e-01
2.21314162e-01 -9.14023101e-01 -2.51142204e-01 6.28254116e-01
1.12583613e+00 -1.12675214e+00 1.79839894e-01 -7.04120576e-01
-1.72500685e-01 1.29305971e+00 7.87191689e-02 -7.84003913e-01
1.05938196e+00 2.28284597e-01 -3.20499122e-01 -6.87344968e-02
-3.70455533e-01 2.76630282e-01 7.71681905e-01 6.96833670e-01
3.65160972e-01 -1.37393447e-02 -2.90511340e-01 3.04276824e-01
-6.55975193e-02 1.06434196e-01 1.13921113e-01 8.72862518e-01
-2.23199576e-01 -9.98916090e-01 -3.18251938e-01 4.77654725e-01
-3.75914603e-01 -1.72880039e-01 3.05819605e-02 7.99243212e-01
-4.30709660e-01 4.03588653e-01 -1.95784181e-01 3.18931758e-01
2.35763177e-01 1.20953679e-01 7.41571009e-01 -3.87938529e-01
-6.71213567e-01 -7.51127750e-02 3.13084945e-02 -4.63610768e-01
-3.13021123e-01 -7.26806641e-01 -1.23137534e+00 -4.22555894e-01
-2.77146816e-01 1.07285818e-02 9.81051683e-01 8.87999117e-01
4.26169813e-01 1.87770292e-01 6.49605274e-01 -1.15005004e+00
-1.16142297e+00 -4.33230191e-01 -3.16504747e-01 2.38605797e-01
5.93548357e-01 -7.00892210e-01 -4.25591350e-01 8.64436477e-02] | [6.678960800170898, 3.6051948070526123] |
46404c32-d857-4c85-82bf-6ef873d6dad3 | robust-single-image-dehazing-based-on | 2203.15325 | null | https://arxiv.org/abs/2203.15325v1 | https://arxiv.org/pdf/2203.15325v1.pdf | Robust Single Image Dehazing Based on Consistent and Contrast-Assisted Reconstruction | Single image dehazing as a fundamental low-level vision task, is essential for the development of robust intelligent surveillance system. In this paper, we make an early effort to consider dehazing robustness under variational haze density, which is a realistic while under-studied problem in the research filed of singe image dehazing. To properly address this problem, we propose a novel density-variational learning framework to improve the robustness of the image dehzing model assisted by a variety of negative hazy images, to better deal with various complex hazy scenarios. Specifically, the dehazing network is optimized under the consistency-regularized framework with the proposed Contrast-Assisted Reconstruction Loss (CARL). The CARL can fully exploit the negative information to facilitate the traditional positive-orient dehazing objective function, by squeezing the dehazed image to its clean target from different directions. Meanwhile, the consistency regularization keeps consistent outputs given multi-level hazy images, thus improving the model robustness. Extensive experimental results on two synthetic and three real-world datasets demonstrate that our method significantly surpasses the state-of-the-art approaches. | ['Jiande Sun', 'Xinbo Gao', 'Nannan Wang', 'Dingwen Zhang', 'Yan Li', 'De Cheng'] | 2022-03-29 | null | null | null | null | ['image-dehazing'] | ['computer-vision'] | [ 3.54730904e-01 -2.58753121e-01 2.32527062e-01 -1.12123139e-01
-4.32915300e-01 7.69385556e-03 4.23810452e-01 -6.02363825e-01
-3.05490136e-01 4.76154357e-01 1.14186853e-01 1.64586809e-02
-2.25892946e-01 -7.07389712e-01 -8.30472171e-01 -1.53287053e+00
3.70384753e-01 -1.68469191e-01 5.65575242e-01 -2.92800456e-01
6.98494911e-02 -6.37832209e-02 -1.50850999e+00 -6.53996840e-02
1.38442063e+00 1.06123471e+00 4.30173427e-01 4.20651078e-01
4.79332924e-01 1.03364205e+00 -5.30016303e-01 -4.84936029e-01
4.20549780e-01 -3.49161923e-01 2.35968173e-01 4.48743045e-01
6.35878801e-01 -7.37280607e-01 -4.76040363e-01 1.85253930e+00
2.77897328e-01 3.04709136e-01 6.87281549e-01 -1.22802436e+00
-1.18446040e+00 -1.22131810e-01 -9.06192362e-01 4.37157363e-01
-4.15546775e-01 4.59205717e-01 3.98912638e-01 -8.42497468e-01
1.19128950e-01 1.35166323e+00 4.83557314e-01 5.52995205e-01
-8.29000592e-01 -7.12939560e-01 2.39625946e-01 2.68993050e-01
-1.53749931e+00 -3.39483052e-01 9.81163144e-01 -4.83653218e-01
1.79107279e-01 6.08302429e-02 4.71080720e-01 6.25685036e-01
3.46254349e-01 6.77614391e-01 9.47069705e-01 -8.79707336e-02
2.62407139e-02 2.82178551e-01 -8.69653970e-02 7.46571660e-01
6.56757951e-01 2.62312114e-01 -1.76334500e-01 8.97471011e-02
6.74106836e-01 4.05160457e-01 -7.88450062e-01 -3.56621921e-01
-7.94413984e-01 6.62713051e-01 4.79068071e-01 3.00680865e-02
-3.73972356e-01 -1.68071449e-01 -3.02401096e-01 6.64092898e-02
8.52429867e-01 7.60689005e-02 5.92819341e-02 5.63774168e-01
-9.99622226e-01 3.15805525e-01 2.26403683e-01 8.47734690e-01
8.39338303e-01 5.21141112e-01 -1.78935170e-01 7.97791421e-01
6.56604171e-01 8.40653598e-01 3.84992771e-02 -9.93052483e-01
4.27615404e-01 3.67406905e-01 3.77019346e-01 -1.24208117e+00
1.67602375e-01 -5.83837271e-01 -1.43714380e+00 5.22881508e-01
1.68536365e-01 -1.58198997e-01 -1.17852926e+00 1.50314081e+00
4.82803345e-01 5.59885025e-01 1.68389350e-01 1.19595504e+00
6.30268633e-01 1.30562830e+00 -1.37204275e-01 -6.39293253e-01
9.94960666e-01 -9.03677225e-01 -1.17106521e+00 -2.72071809e-01
-7.53184929e-02 -6.28992975e-01 5.88031054e-01 6.74955904e-01
-1.22801542e+00 -5.22006452e-01 -1.32625175e+00 -8.18345174e-02
-9.96818542e-02 -2.44295709e-02 6.18326180e-02 6.40203774e-01
-1.05058742e+00 4.85120453e-02 -6.52313948e-01 5.65495752e-02
5.63684046e-01 -7.93584660e-02 -1.03942856e-01 -6.85775876e-01
-1.34462786e+00 9.22009110e-01 5.12520015e-01 6.31944895e-01
-1.37146604e+00 -7.56730258e-01 -1.01558924e+00 -2.30899110e-01
6.87670052e-01 -7.05179214e-01 6.41727686e-01 -7.22326458e-01
-1.27995706e+00 6.11592174e-01 2.65416410e-02 -2.84829825e-01
5.44130445e-01 -2.23433092e-01 -5.60794890e-01 2.44128734e-01
2.80442331e-02 3.56150627e-01 1.52949548e+00 -1.56587768e+00
-5.34569085e-01 -4.97732341e-01 -1.31642014e-01 3.29965442e-01
-3.20940793e-01 -2.05706492e-01 -4.75573182e-01 -8.78581941e-01
-5.73099814e-02 -5.56932807e-01 -1.27738565e-02 4.04195487e-01
-1.55180797e-01 1.70820996e-01 1.01568627e+00 -1.08259499e+00
1.11614549e+00 -2.40456200e+00 5.15532017e-01 -1.51846856e-01
4.24535394e-01 6.68528914e-01 1.03958078e-01 -1.63988799e-01
2.28488043e-01 -5.96071854e-02 -8.70234370e-01 -3.23724508e-01
-4.16067988e-01 2.35477149e-01 -2.84746557e-01 9.09648061e-01
5.36453366e-01 5.14801145e-01 -8.05370390e-01 -5.81753194e-01
5.46443224e-01 8.42822015e-01 -5.58086336e-01 7.49997616e-01
-1.59484223e-01 4.41353649e-01 -4.15443331e-01 5.52110612e-01
1.35052073e+00 -4.35684063e-02 -6.06959343e-01 -2.49469399e-01
-3.16947639e-01 -6.38861060e-01 -1.01667690e+00 1.19989097e+00
-1.80733334e-02 4.04723346e-01 6.93488181e-01 -1.08046746e+00
7.97401547e-01 4.03479971e-02 1.60111874e-01 -4.33716804e-01
1.42860934e-01 1.71209918e-03 -2.02874959e-01 -8.60691667e-01
3.48123759e-01 -4.52125698e-01 5.53218424e-01 -2.91182011e-01
-5.21943644e-02 -4.28639650e-01 -2.02371448e-01 1.02172211e-01
2.38987133e-01 -1.35463744e-01 -3.02506555e-02 -3.61065060e-01
7.15258956e-01 -2.74094015e-01 9.07502532e-01 5.79901814e-01
-5.35234988e-01 9.20084655e-01 1.30438238e-01 -1.13038763e-01
-8.86919796e-01 -1.05860245e+00 -1.73027858e-01 5.97863615e-01
5.91265440e-01 2.45375425e-01 -9.82869267e-01 -1.71898112e-01
-3.79239976e-01 6.85004592e-01 -6.88031137e-01 -6.03676975e-01
-4.25317556e-01 -1.19953942e+00 1.86367735e-01 -1.15917437e-01
1.30154395e+00 -4.89558250e-01 -2.62798052e-02 -1.48429528e-01
-4.14934039e-01 -1.01136160e+00 -6.06937587e-01 -4.28868026e-01
-4.66713428e-01 -9.89147246e-01 -1.24604321e+00 -8.84179235e-01
5.14364779e-01 9.88402128e-01 6.18788838e-01 1.82744011e-01
5.16628884e-02 2.21726805e-01 -3.67266238e-01 -5.97243071e-01
-2.36086175e-01 -5.11107326e-01 -1.19562475e-02 7.03470826e-01
2.40955390e-02 -4.63042229e-01 -7.46975183e-01 3.83559912e-01
-1.50843072e+00 -4.20505926e-02 7.35504806e-01 1.03891182e+00
4.46228296e-01 8.34401667e-01 1.23466477e-01 -3.79230767e-01
2.09059685e-01 -8.00192237e-01 -8.14472377e-01 3.41128051e-01
-5.83529115e-01 -2.01503873e-01 4.00021076e-01 -5.62737405e-01
-1.45897102e+00 -4.04583663e-01 -1.77523512e-02 -8.40828180e-01
1.81515604e-01 3.59240562e-01 -6.14542425e-01 -2.92676479e-01
3.08436900e-01 8.85477185e-01 3.20830524e-01 -1.61848634e-01
2.42201880e-01 5.59020758e-01 8.10395837e-01 -1.85545743e-01
1.40673876e+00 8.82100642e-01 -8.36646706e-02 -1.13653457e+00
-1.00464272e+00 -3.60889375e-01 -1.27380818e-01 -4.15668547e-01
1.31559181e+00 -1.12077284e+00 -4.38516289e-01 1.07991624e+00
-1.17303288e+00 -9.67343673e-02 2.52402931e-01 3.70924264e-01
-4.33601290e-02 7.81253159e-01 -2.52623975e-01 -1.25265336e+00
-2.53123701e-01 -1.08437824e+00 9.92514074e-01 6.10441923e-01
1.08392572e+00 -1.05185091e+00 -1.59972355e-01 5.77147543e-01
4.65996265e-01 3.48730624e-01 6.63388550e-01 2.18179762e-01
-1.03130198e+00 1.22533917e-01 -2.89945155e-01 1.03333080e+00
8.17330629e-02 3.46146710e-02 -9.41329896e-01 -4.27563816e-01
6.88609362e-01 -3.03508211e-02 1.20588255e+00 6.47532046e-01
1.03801572e+00 -3.08039010e-01 1.18254319e-01 1.13207197e+00
1.30178797e+00 2.03539446e-01 8.61324966e-01 3.33922774e-01
7.90787876e-01 8.39347899e-01 7.77972937e-01 3.34854931e-01
3.53221595e-01 3.95316184e-01 9.96690214e-01 -3.12580705e-01
-1.51680894e-02 -1.33997917e-01 4.07154858e-01 5.35650253e-01
-9.60713327e-02 -5.76050043e-01 -4.20825064e-01 6.12813473e-01
-1.80846882e+00 -1.00411594e+00 -8.05455297e-02 2.03952312e+00
6.24198377e-01 -1.03457317e-01 -2.82018244e-01 -1.37314171e-01
1.02968049e+00 7.81033933e-01 -5.41469812e-01 4.81684566e-01
-4.24604714e-01 -5.01882911e-01 5.16941786e-01 5.95260739e-01
-1.24129498e+00 7.98632860e-01 5.21165943e+00 9.71010804e-01
-8.60212982e-01 1.49373904e-01 6.50917828e-01 5.46055920e-02
-2.52695501e-01 -2.92622983e-01 -6.26120925e-01 9.36957419e-01
4.15510565e-01 2.85918806e-02 5.81995130e-01 6.76683545e-01
4.43291992e-01 -8.36465508e-02 -3.28946739e-01 1.15036261e+00
5.23401439e-01 -1.21005189e+00 2.12611184e-01 -2.17354950e-02
9.07957137e-01 -2.91863501e-01 6.52020812e-01 1.10044159e-01
7.84305334e-02 -6.89496100e-01 7.86838949e-01 6.92878127e-01
6.71061218e-01 -6.34689271e-01 9.01631355e-01 5.05493283e-01
-1.00530493e+00 -5.77738024e-02 -7.75573134e-01 8.54547694e-02
2.71748215e-01 6.60387576e-01 2.82592932e-03 7.54980147e-01
1.06887341e+00 1.12433100e+00 -4.35274720e-01 9.33148026e-01
-2.08050475e-01 5.70161343e-01 3.11313476e-02 4.58979458e-01
3.26540679e-01 -5.55345476e-01 8.45463336e-01 7.78210402e-01
2.74128556e-01 6.15716457e-01 -4.03092578e-02 9.23932493e-01
2.22063780e-01 -4.31010574e-01 -5.44847369e-01 2.53576040e-01
2.05713421e-01 9.32103157e-01 2.95045556e-05 -3.35764736e-01
-2.76495486e-01 9.07106340e-01 -7.59803876e-02 6.64067149e-01
-1.01328123e+00 -3.31181943e-01 7.42721498e-01 9.69101712e-02
3.21584284e-01 -2.11718112e-01 6.69476762e-02 -1.48388779e+00
2.09561035e-01 -7.60646522e-01 1.95549175e-01 -1.07257175e+00
-1.47725499e+00 3.08524698e-01 2.26060465e-01 -1.35009027e+00
2.26674423e-01 -7.09391832e-01 -9.09707785e-01 9.67397511e-01
-2.24140167e+00 -1.19318712e+00 -7.62690067e-01 9.38022077e-01
5.81298590e-01 -9.37121138e-02 -8.60499293e-02 4.42683458e-01
-9.27259505e-01 3.39379162e-01 3.51328909e-01 -1.45559818e-01
7.75993764e-01 -8.43986332e-01 -2.16684535e-01 1.55187511e+00
-5.91613591e-01 5.04729092e-01 9.38287079e-01 -6.11493528e-01
-1.30293751e+00 -1.55163491e+00 3.98507006e-02 -3.08941901e-01
5.27727783e-01 -2.52894592e-02 -1.35823023e+00 4.21484113e-01
4.61195797e-01 1.74086675e-01 2.19436824e-01 -7.90761292e-01
-2.46794969e-01 -3.40339571e-01 -1.23537588e+00 4.26125824e-01
7.29561448e-01 -1.26868650e-01 -8.18360150e-01 2.65125841e-01
1.08863842e+00 -3.03863138e-01 -5.33916175e-01 5.44508040e-01
1.73968077e-01 -1.08295250e+00 1.12286258e+00 -8.89204964e-02
3.27823550e-01 -7.09320009e-01 -1.58397123e-01 -1.35985911e+00
-4.81139809e-01 -8.34017098e-01 -3.04204732e-01 1.21672273e+00
-1.80304393e-01 -7.95067549e-01 4.11764652e-01 4.10145640e-01
-3.65181237e-01 -3.94966483e-01 -7.94645786e-01 -7.11353242e-01
8.83500353e-02 -2.08917931e-02 3.21642637e-01 8.93559873e-01
-8.09954226e-01 -1.60789847e-01 -1.24457860e+00 9.38919902e-01
1.36802506e+00 -5.34799874e-01 6.61217809e-01 -9.51602042e-01
-1.01403467e-01 -1.64974570e-01 -3.47811073e-01 -1.19390655e+00
-1.59515440e-02 -1.79226413e-01 3.90050858e-01 -1.31241333e+00
1.49698600e-01 4.40964177e-02 -3.54168862e-01 -1.86100662e-01
-7.85102725e-01 2.88216412e-01 9.52718481e-02 4.33390379e-01
-4.28132623e-01 1.09389746e+00 1.50499713e+00 -6.53321087e-01
1.46811426e-01 -2.60016434e-02 -7.36943007e-01 6.30945921e-01
5.13434291e-01 -4.14007396e-01 -7.04783320e-01 -7.55397677e-01
-2.68272936e-01 -4.89421561e-02 5.96592546e-01 -9.38238323e-01
3.80859286e-01 -6.21021450e-01 3.00653100e-01 -5.37675083e-01
6.02760136e-01 -9.24625635e-01 -1.10616073e-01 3.15948099e-01
2.14350950e-02 -2.93450713e-01 -4.08924185e-02 1.16371846e+00
-4.96310979e-01 -1.19918749e-01 1.31481230e+00 -5.74365333e-02
-7.77589083e-01 6.85643256e-01 -1.73687384e-01 1.55294597e-01
9.94868338e-01 -2.36564279e-01 -5.84435046e-01 -4.09264922e-01
-1.48146868e-01 4.31556016e-01 5.73029220e-01 2.22316265e-01
1.09971023e+00 -9.20128107e-01 -9.95947897e-01 4.51955497e-01
1.14227571e-01 4.14872795e-01 7.88735986e-01 8.20086539e-01
-5.86078048e-01 -1.35315448e-01 -1.06206149e-01 -7.27902651e-01
-1.04501629e+00 9.16690826e-01 6.07540190e-01 1.49922386e-01
-6.28227770e-01 8.75209093e-01 8.90624583e-01 8.38983338e-03
2.06265524e-01 -1.44460902e-01 -3.59408081e-01 -4.16451752e-01
9.93734837e-01 4.24015164e-01 -3.39692205e-01 -7.62413740e-01
-5.56981415e-02 7.38856733e-01 -6.00627363e-02 -2.46807598e-02
1.24089551e+00 -4.56482530e-01 -3.34530085e-01 6.70583248e-02
1.00694740e+00 -2.24300385e-01 -1.91125190e+00 -3.10937673e-01
-6.02508307e-01 -1.00346625e+00 5.22763014e-01 -3.14816654e-01
-1.29377091e+00 9.93188381e-01 7.66038120e-01 2.79450715e-01
1.31542087e+00 -3.72373134e-01 7.92783618e-01 2.33550146e-01
9.27300751e-02 -8.01226199e-01 3.66698086e-01 8.35841596e-02
8.58282208e-01 -1.55505455e+00 1.43718407e-01 -3.80973458e-01
-7.49690652e-01 6.23238683e-01 7.63138235e-01 -3.38072419e-01
7.89981604e-01 -2.95551598e-01 -1.34316029e-03 -1.79717228e-01
-4.18234497e-01 -1.29864309e-02 3.44832271e-01 9.27825987e-01
-2.99654663e-01 -2.61047512e-01 1.40397489e-01 2.31059685e-01
4.65011328e-01 -3.13645840e-01 6.67942941e-01 6.26448691e-01
-7.72349477e-01 -2.64832169e-01 -7.55868912e-01 1.10202126e-01
-2.94733405e-01 -2.04707325e-01 -2.47494187e-02 7.41654813e-01
3.44136745e-01 1.37950540e+00 -2.00414777e-01 -3.09681147e-01
1.52411595e-01 -5.06834447e-01 2.87576973e-01 -3.64382386e-01
7.87755027e-02 1.73430040e-01 -6.73016191e-01 -1.07665755e-01
-7.38100767e-01 -3.72886360e-01 -5.11242032e-01 -2.50377744e-01
-5.07288814e-01 1.25526562e-01 2.09604546e-01 8.54900718e-01
7.32460320e-02 5.23180127e-01 8.14364314e-01 -9.66234982e-01
-6.10603392e-01 -9.48865056e-01 -8.86930525e-01 2.30422258e-01
9.68220055e-01 -1.03782880e+00 -9.13829267e-01 2.55147755e-01] | [10.957984924316406, -3.1757986545562744] |
310ab483-e26c-4cb0-9ce3-b152fcee59b2 | unconditional-audio-generation-with | 2005.08526 | null | https://arxiv.org/abs/2005.08526v1 | https://arxiv.org/pdf/2005.08526v1.pdf | Unconditional Audio Generation with Generative Adversarial Networks and Cycle Regularization | In a recent paper, we have presented a generative adversarial network (GAN)-based model for unconditional generation of the mel-spectrograms of singing voices. As the generator of the model is designed to take a variable-length sequence of noise vectors as input, it can generate mel-spectrograms of variable length. However, our previous listening test shows that the quality of the generated audio leaves room for improvement. The present paper extends and expands that previous work in the following aspects. First, we employ a hierarchical architecture in the generator to induce some structure in the temporal dimension. Second, we introduce a cycle regularization mechanism to the generator to avoid mode collapse. Third, we evaluate the performance of the new model not only for generating singing voices, but also for generating speech voices. Evaluation result shows that new model outperforms the prior one both objectively and subjectively. We also employ the model to unconditionally generate sequences of piano and violin music and find the result promising. Audio examples, as well as the code for implementing our model, will be publicly available online upon paper publication. | ['Yi-Hsuan Yang', 'Yin-Cheng Yeh', 'Yu-Hua Chen', 'Jen-Yu Liu'] | 2020-05-18 | null | null | null | null | ['audio-generation'] | ['audio'] | [ 2.78520852e-01 2.54436791e-01 3.41075331e-01 1.33469567e-01
-8.95732820e-01 -8.65436971e-01 3.83951157e-01 -7.40363538e-01
1.48946077e-01 8.31526279e-01 3.23492467e-01 -1.11274876e-01
4.41766828e-02 -6.87690020e-01 -5.05295813e-01 -7.99507439e-01
-1.92992166e-01 6.61483034e-02 1.65248305e-01 -3.02505702e-01
-8.70146528e-02 1.67906314e-01 -1.55499625e+00 3.49188596e-01
6.94272578e-01 7.14417994e-01 1.39757335e-01 1.13613176e+00
5.54539919e-01 6.99352145e-01 -9.96583700e-01 -3.51154208e-01
3.67665768e-01 -1.15909433e+00 -5.88655472e-01 -6.56180829e-02
1.53682351e-01 -1.63876995e-01 -1.23301752e-01 9.49656487e-01
9.24508095e-01 2.53287464e-01 6.48835599e-01 -1.04930949e+00
-5.21895349e-01 9.22458172e-01 9.89365205e-02 -8.69759843e-02
5.24801612e-01 4.19932038e-01 1.00127923e+00 -7.05579221e-01
4.76619154e-01 1.04454410e+00 6.99836850e-01 8.30979288e-01
-1.25113153e+00 -8.45137239e-01 -4.64017361e-01 1.14793973e-02
-1.41582263e+00 -5.79835653e-01 1.11929166e+00 -3.96092266e-01
6.15374506e-01 7.03918815e-01 8.33585739e-01 1.38860857e+00
1.06782041e-01 4.87433136e-01 1.01864600e+00 -7.19538808e-01
2.14750424e-01 -1.82576887e-02 -5.46739101e-01 3.60658258e-01
-3.90593797e-01 6.87472820e-01 -5.31426489e-01 -9.95664969e-02
8.57517600e-01 -7.75914192e-01 -2.37815440e-01 2.10535347e-01
-1.11538708e+00 7.33537674e-01 8.92439038e-02 5.38350701e-01
-4.26123291e-01 3.70130867e-01 2.31475413e-01 3.15131038e-01
3.48028183e-01 7.05702364e-01 1.08717300e-01 -3.54925603e-01
-1.26930642e+00 5.40573597e-01 9.15599167e-01 5.97960472e-01
2.00605378e-01 8.14615488e-01 -5.16795516e-01 8.02718520e-01
2.10811436e-01 4.74414408e-01 6.96405053e-01 -1.26282597e+00
2.96715528e-01 -2.06796363e-01 -5.38924560e-02 -8.59952033e-01
-5.79343326e-02 -5.35860479e-01 -8.83540750e-01 3.78308266e-01
2.35612467e-01 -5.40343702e-01 -8.18508506e-01 1.88343930e+00
8.19439292e-02 5.20992100e-01 4.62580547e-02 8.17250609e-01
8.63002419e-01 7.90900528e-01 -3.13704759e-01 -4.84137863e-01
8.75971317e-01 -9.41743076e-01 -1.13685811e+00 2.46814623e-01
-1.13233820e-01 -1.15231645e+00 1.13994646e+00 5.98097980e-01
-1.45172989e+00 -9.71177876e-01 -1.10495949e+00 3.64352643e-01
4.95503731e-02 1.49028733e-01 4.18413430e-01 1.05890000e+00
-1.10768974e+00 7.51543939e-01 -5.85638821e-01 4.10149172e-02
-1.68062583e-01 2.42957890e-01 5.79556935e-02 6.52709544e-01
-1.39674342e+00 2.64206856e-01 4.16645378e-01 5.90225421e-02
-1.13649452e+00 -4.74176526e-01 -7.06033170e-01 6.63366541e-02
-2.04013549e-02 -6.90164268e-01 1.53888404e+00 -1.02421725e+00
-2.17736769e+00 3.65159422e-01 -2.77939532e-02 -5.32420635e-01
3.63313884e-01 4.03398722e-02 -7.45689452e-01 5.39666899e-02
-1.84467450e-01 6.98019564e-01 1.10585177e+00 -1.23087156e+00
-3.27893287e-01 3.47787172e-01 -4.75983182e-03 4.44184914e-02
-2.43170813e-01 -1.61346868e-01 -2.38118082e-01 -1.29135334e+00
-7.87473843e-02 -1.04976642e+00 -1.91337258e-01 -8.43759894e-01
-7.21605957e-01 6.49662390e-02 6.06302500e-01 -5.90798795e-01
1.58450425e+00 -2.32181811e+00 1.41530812e-01 2.15105772e-01
-2.34567687e-01 3.01961690e-01 -1.70028865e-01 6.61254883e-01
-3.68606150e-01 2.04144135e-01 -3.13091218e-01 -3.80399853e-01
5.58601730e-02 2.04150509e-02 -6.79778159e-01 -1.42272050e-02
4.55022380e-02 8.56231391e-01 -7.05831885e-01 -1.98165357e-01
7.47909304e-03 5.01596451e-01 -9.36386526e-01 4.87547934e-01
-2.09364951e-01 8.39736581e-01 1.19757839e-01 3.33568096e-01
3.59254539e-01 3.55084658e-01 -9.19095799e-02 -1.94383547e-01
-6.24056607e-02 4.32288945e-01 -1.27667701e+00 1.59961772e+00
-4.16017443e-01 5.08832276e-01 -8.99646655e-02 -5.63134909e-01
9.68915105e-01 9.20852840e-01 2.84725934e-01 -2.59616137e-01
4.56774347e-02 1.72404110e-01 4.06641871e-01 -4.76899594e-01
4.81337547e-01 -4.88111138e-01 -1.03683941e-01 4.57527816e-01
1.63151026e-01 -8.21532905e-01 3.48688662e-01 -2.54797991e-02
8.03366244e-01 1.86987653e-01 7.55666271e-02 1.01633355e-01
7.06072509e-01 -3.66751969e-01 5.31529188e-01 5.90737045e-01
1.47735253e-01 9.59373116e-01 4.02625024e-01 6.23688176e-02
-1.13529062e+00 -1.13432944e+00 1.42567739e-01 7.81164646e-01
-3.42441946e-01 -5.13472915e-01 -9.97697651e-01 -1.23211309e-01
-4.50964630e-01 8.87434185e-01 -4.20451790e-01 -6.27621785e-02
-6.48875117e-01 -4.58897024e-01 9.67282057e-01 4.85181540e-01
2.37225845e-01 -1.59643555e+00 -2.05259398e-01 4.35294092e-01
-4.16487277e-01 -8.63346279e-01 -8.82892311e-01 6.45655915e-02
-7.37364054e-01 -6.02904141e-01 -7.35377133e-01 -8.99065554e-01
1.72660798e-01 -4.06582505e-01 9.13851917e-01 -2.15862840e-01
-2.10103780e-01 3.58031482e-01 -5.34191728e-01 -5.47872305e-01
-1.01451075e+00 2.57901903e-02 2.36148223e-01 1.49939805e-01
-2.57822067e-01 -1.03054786e+00 -4.94834810e-01 3.22121024e-01
-1.14944029e+00 1.02686808e-01 1.85435548e-01 7.94690013e-01
4.51328754e-01 3.72377664e-01 9.56997991e-01 -6.58169568e-01
1.22164106e+00 -1.20204590e-01 -3.82931530e-01 -4.30752695e-01
-3.11610878e-01 -1.65273651e-01 9.52887893e-01 -6.75490379e-01
-9.24402237e-01 7.98663199e-02 -6.81659937e-01 -4.02978480e-01
-1.50178194e-01 2.91565835e-01 -1.22838065e-01 2.75699914e-01
8.02232027e-01 3.00209880e-01 -1.52419820e-01 -4.71227795e-01
4.87513214e-01 7.06648350e-01 9.36314106e-01 -4.43577975e-01
1.21165144e+00 3.47260870e-02 -9.28479731e-02 -7.46335149e-01
-4.67351645e-01 6.83138380e-03 -4.55102712e-01 -3.14736933e-01
7.90010154e-01 -7.58312881e-01 -7.84675658e-01 4.37474191e-01
-1.11908889e+00 -4.89199817e-01 -6.67838097e-01 5.91155708e-01
-1.00871253e+00 2.25904912e-01 -7.22986400e-01 -1.06402504e+00
-3.68227988e-01 -9.18441236e-01 8.11411023e-01 1.15419954e-01
-4.95685637e-01 -8.60133111e-01 4.09187704e-01 2.14908615e-01
4.67189193e-01 3.84665787e-01 7.37562239e-01 -3.59383732e-01
-4.79254752e-01 -1.61772847e-01 6.76778913e-01 8.37376595e-01
2.81799018e-01 2.14998759e-02 -1.17978251e+00 -3.15743536e-01
2.44748667e-01 -3.05568665e-01 5.76603115e-01 3.26739222e-01
1.02361381e+00 -5.33141255e-01 2.47956842e-01 6.35879040e-01
9.44514096e-01 5.36467016e-01 9.13139164e-01 -1.65130377e-01
4.03360009e-01 4.41554338e-01 3.99873167e-01 4.91688728e-01
-1.69615731e-01 7.49586999e-01 3.60757619e-01 -9.78769138e-02
-5.83559752e-01 -6.23026133e-01 5.80889940e-01 1.43517303e+00
-4.64089096e-01 -3.45019907e-01 -3.23652834e-01 5.71378589e-01
-1.41560352e+00 -1.44215024e+00 4.87011708e-02 2.12588549e+00
1.01750898e+00 2.77193729e-02 4.72380787e-01 7.44869173e-01
5.77364981e-01 2.49924675e-01 -2.61059374e-01 -6.30298853e-01
-5.34938499e-02 7.13985026e-01 -8.26157257e-02 7.11875796e-01
-9.12111402e-01 8.26513529e-01 7.32531166e+00 9.73903298e-01
-1.23367023e+00 -4.50778715e-02 3.70644122e-01 -2.51023144e-01
-4.55863446e-01 -9.57397521e-02 -4.63144869e-01 5.41028738e-01
1.06603372e+00 -3.78348887e-01 8.90488625e-01 5.00326455e-01
5.62774122e-01 3.80009949e-01 -9.86563802e-01 8.45639586e-01
-1.67383086e-02 -1.16284597e+00 2.76699718e-02 -2.90941559e-02
8.52879465e-01 -6.53959274e-01 3.83377999e-01 3.71005177e-01
1.06599122e-01 -1.34899211e+00 9.18098330e-01 5.41942120e-01
9.33540165e-01 -1.02102160e+00 5.33472657e-01 3.89172286e-01
-1.17855906e+00 5.35721965e-02 -6.42128382e-03 -9.15367007e-02
3.59567642e-01 4.10131842e-01 -1.11783803e+00 5.68709016e-01
2.83214092e-01 2.10591421e-01 -3.22805077e-01 1.04585838e+00
-4.74971861e-01 1.29929399e+00 -3.78796495e-02 1.39038652e-01
1.03666484e-01 -1.47304833e-01 8.48335624e-01 1.21373713e+00
6.83580339e-01 3.35061587e-02 -8.46480429e-02 1.10333145e+00
1.48391668e-02 1.73184186e-01 -6.85894191e-01 -2.62080520e-01
3.79032552e-01 1.15033221e+00 -2.53454626e-01 -3.03967372e-02
1.18703336e-01 1.04269397e+00 -3.59315544e-01 5.37945211e-01
-9.53797758e-01 -7.04541385e-01 2.37764239e-01 1.22791991e-01
4.11521643e-01 -2.96969295e-01 -2.40417406e-01 -7.73142695e-01
-8.08878914e-02 -1.28739917e+00 4.59496342e-02 -8.87878060e-01
-1.03250873e+00 1.05024683e+00 -2.38952518e-01 -1.54477024e+00
-1.05765998e+00 -1.27618849e-01 -9.15866375e-01 1.10381854e+00
-9.08927679e-01 -8.81703138e-01 4.63706674e-03 4.97887015e-01
6.04251504e-01 -4.52913523e-01 1.07215416e+00 3.47031474e-01
-2.30785832e-01 7.69188523e-01 -1.52885154e-01 -1.43548083e-02
6.13124371e-01 -1.15382504e+00 5.49887836e-01 8.80893469e-01
6.14647090e-01 5.01697242e-01 1.06709445e+00 -4.17484879e-01
-9.45606828e-01 -9.61224616e-01 8.32722247e-01 -3.27917397e-01
4.45347667e-01 -4.20262635e-01 -7.42701352e-01 4.46711123e-01
6.81112170e-01 -4.49909806e-01 8.77079368e-01 -2.60435283e-01
5.28003313e-02 -6.18578233e-02 -1.06267333e+00 5.83816707e-01
7.51991808e-01 -4.55676198e-01 -6.47014618e-01 2.92260908e-02
9.68588293e-01 -5.79542220e-01 -7.40958214e-01 4.89424318e-01
5.30779660e-01 -1.20741498e+00 8.52257073e-01 -3.05088729e-01
4.55398262e-01 -5.72942674e-01 -1.49362177e-01 -1.57933617e+00
-4.29193854e-01 -1.37422585e+00 -1.03079677e-01 1.49839246e+00
5.41404188e-01 -2.63589144e-01 5.53700566e-01 -7.36495331e-02
-1.38417870e-01 -4.16805327e-01 -9.05883253e-01 -1.13807213e+00
1.26736462e-01 -6.87946737e-01 6.21060908e-01 5.70199668e-01
-6.28803894e-02 5.89463055e-01 -8.74687493e-01 5.28251976e-02
4.40571725e-01 1.53990701e-01 8.84871960e-01 -8.36776018e-01
-9.56451833e-01 -3.31077665e-01 -1.00396872e-01 -9.25637007e-01
-2.60545556e-02 -9.17722583e-01 2.55262464e-01 -1.08785117e+00
-2.76956648e-01 -1.11562632e-01 -1.94841653e-01 1.33768544e-01
-1.10314198e-01 7.30396807e-01 5.90306401e-01 -1.19931772e-01
3.02088428e-02 6.99985683e-01 1.50433886e+00 7.47459233e-02
-5.14772892e-01 5.25662065e-01 -5.79974353e-01 6.71760678e-01
8.58665109e-01 -5.01605630e-01 -5.64163744e-01 8.27714726e-02
-6.43794015e-02 2.83354670e-01 2.95218706e-01 -1.36808920e+00
-5.59078380e-02 9.80055258e-02 2.10692510e-01 -5.87056577e-01
6.18154883e-01 -5.31668425e-01 5.10891020e-01 4.74524319e-01
-4.37885821e-01 -1.01779938e-01 2.71196514e-01 3.34298790e-01
-5.12088299e-01 -3.59223068e-01 8.26561153e-01 -4.41342453e-03
-1.16300695e-02 5.58497347e-02 -5.10228574e-01 6.25469759e-02
6.35120153e-01 -3.33316289e-02 2.34232694e-01 -9.31430459e-01
-9.62551713e-01 -4.24067229e-01 5.02543896e-02 4.04071987e-01
4.41244036e-01 -1.65796685e+00 -8.11185956e-01 4.21090037e-01
-3.80683154e-01 -4.08181787e-01 1.30280539e-01 4.16144639e-01
-3.16418022e-01 4.31118160e-01 -9.74604264e-02 -3.80685806e-01
-1.24443960e+00 5.21504343e-01 2.37663135e-01 -1.76840082e-01
-3.54638487e-01 6.94810390e-01 1.07372515e-01 -3.75783205e-01
3.98534775e-01 -2.15509608e-01 -1.92937002e-01 -7.87250623e-02
4.85854357e-01 3.08124930e-01 -2.22373754e-01 -4.88846838e-01
-1.26309749e-02 4.29987431e-01 5.91939509e-01 -7.49407470e-01
1.09555733e+00 1.53755188e-01 -4.15633619e-02 7.59300113e-01
8.95470023e-01 6.54849648e-01 -1.00814652e+00 1.83589533e-01
-3.87042761e-01 -3.52295518e-01 -2.90463895e-01 -7.56893337e-01
-9.20034647e-01 7.97624886e-01 5.58894455e-01 5.92641354e-01
1.40568757e+00 -3.22677463e-01 9.01333511e-01 -1.20438837e-01
1.85169622e-01 -8.67384017e-01 3.80574524e-01 6.13466501e-01
1.16368735e+00 -5.04544973e-01 -4.74690765e-01 -3.17756802e-01
-6.31995142e-01 1.02389383e+00 4.11567092e-01 -2.20354870e-01
4.71583843e-01 4.03766990e-01 2.96743542e-01 2.86785156e-01
-7.97928333e-01 -1.85243458e-01 4.75377619e-01 7.48543620e-01
7.65707731e-01 4.60491441e-02 -4.90526706e-01 7.74477363e-01
-1.02493060e+00 -2.70955525e-02 6.48805201e-01 3.12113345e-01
-2.74646699e-01 -1.52516770e+00 -6.40057385e-01 -1.24690466e-01
-7.28176177e-01 -3.08185488e-01 -4.42288429e-01 4.49834049e-01
3.20492089e-01 1.22692490e+00 -1.04546584e-01 -6.56795859e-01
3.95133734e-01 3.40474814e-01 4.19922501e-01 -5.51848531e-01
-7.79455066e-01 5.27763963e-01 1.75124452e-01 -2.31221810e-01
-3.01996410e-01 -3.83809835e-01 -9.53036964e-01 -1.11794136e-01
-1.82291597e-01 4.46170509e-01 4.84912813e-01 4.96299267e-01
3.96663845e-02 1.10902607e+00 1.09454095e+00 -9.14338350e-01
-5.97050667e-01 -1.26717043e+00 -8.35765302e-01 4.71092969e-01
3.32728446e-01 -3.76223959e-02 -5.82055748e-01 4.62791532e-01] | [15.556055068969727, 5.987834930419922] |
1980188f-4aad-43c1-8cef-8e0df68590a7 | theoretical-analysis-of-deep-neural-networks | 2202.09954 | null | https://arxiv.org/abs/2202.09954v2 | https://arxiv.org/pdf/2202.09954v2.pdf | Theoretical Analysis of Deep Neural Networks in Physical Layer Communication | Recently, deep neural network (DNN)-based physical layer communication techniques have attracted considerable interest. Although their potential to enhance communication systems and superb performance have been validated by simulation experiments, little attention has been paid to the theoretical analysis. Specifically, most studies in the physical layer have tended to focus on the application of DNN models to wireless communication problems but not to theoretically understand how does a DNN work in a communication system. In this paper, we aim to quantitatively analyze why DNNs can achieve comparable performance in the physical layer comparing with traditional techniques, and also drive their cost in terms of computational complexity. To achieve this goal, we first analyze the encoding performance of a DNN-based transmitter and compare it to a traditional one. And then, we theoretically analyze the performance of DNN-based estimator and compare it with traditional estimators. Third, we investigate and validate how information is flown in a DNN-based communication system under the information theoretic concepts. Our analysis develops a concise way to open the "black box" of DNNs in physical layer communication, which can be applied to support the design of DNN-based intelligent communication techniques and help to provide explainable performance assessment. | ['Jibo Wei', 'Kai Mei', 'Dongtang Ma', 'Haitao Zhao', 'Jun Liu'] | 2022-02-21 | null | null | null | null | ['intelligent-communication'] | ['time-series'] | [ 2.82545179e-01 3.22829813e-01 -4.94412214e-01 -2.69645929e-01
-1.15042068e-01 2.72976086e-02 3.73698980e-01 -1.40885994e-01
-4.25747991e-01 8.90650094e-01 -3.90645415e-02 -7.50273049e-01
-5.50449550e-01 -8.35989356e-01 -3.01163733e-01 -8.85810733e-01
-3.72276127e-01 -2.21803263e-01 2.68557593e-02 -1.21689402e-01
-5.84695823e-02 5.90206265e-01 -1.09148741e+00 -3.46651763e-01
3.53943527e-01 1.33352721e+00 4.68172669e-01 6.84796870e-01
-1.59295663e-01 8.41084361e-01 -9.87150490e-01 -3.74122709e-01
1.60691768e-01 -7.18946040e-01 -4.88490254e-01 -2.90240020e-01
4.37871665e-02 -7.44094908e-01 -9.24866855e-01 1.19410634e+00
7.52880454e-01 -4.23026353e-01 5.05995095e-01 -1.21636641e+00
-3.27209592e-01 1.27064848e+00 -2.03495413e-01 4.21496660e-01
-2.28754684e-01 -4.44398195e-01 8.58445823e-01 -8.16810727e-02
2.17130065e-01 1.28795457e+00 6.55298054e-01 8.74718368e-01
-9.99213874e-01 -1.04179871e+00 -6.61891401e-02 2.79522419e-01
-1.09165919e+00 -6.38973653e-01 6.64537728e-01 1.93777680e-01
4.26642329e-01 2.42890865e-02 8.10295820e-01 1.09899676e+00
3.22460979e-01 9.71095204e-01 4.59126741e-01 -6.80451989e-01
4.69101340e-01 1.21933833e-01 2.29220480e-01 4.79040325e-01
5.69644094e-01 4.36031640e-01 -5.33004522e-01 2.25543678e-01
8.35956573e-01 -3.44752818e-01 -4.16679174e-01 -1.38642475e-01
-7.60105312e-01 6.95856690e-01 5.29929519e-01 5.74965715e-01
-3.57086748e-01 9.07019854e-01 2.41481647e-01 4.47422534e-01
3.04112345e-01 8.62076357e-02 -2.59843439e-01 -2.15354592e-01
-9.90365207e-01 3.68692055e-02 1.03083706e+00 1.19168830e+00
2.61049420e-01 3.96320641e-01 8.39601681e-02 7.89405644e-01
7.20027208e-01 7.27574766e-01 1.92785606e-01 -1.32392991e+00
2.75551319e-01 -8.05091783e-02 -5.46780944e-01 -7.15891182e-01
-5.41957080e-01 -8.85500312e-01 -1.24531019e+00 2.87137628e-01
3.35800558e-01 -6.51842594e-01 -3.44887763e-01 1.72414517e+00
-2.51133174e-01 -1.49213776e-01 6.12667501e-01 7.05504715e-01
5.57422817e-01 7.64239430e-01 -3.54732215e-01 -3.75486910e-01
1.10389864e+00 -6.37354195e-01 -9.98219907e-01 3.24277468e-02
6.69779718e-01 -4.71307009e-01 -1.03032783e-01 3.58237594e-01
-1.11509728e+00 -3.41897458e-01 -1.36590958e+00 3.43206793e-01
5.65660140e-03 1.32003754e-01 5.78481317e-01 1.33398688e+00
-1.30140495e+00 5.16965687e-01 -7.49524176e-01 -5.76245606e-01
5.30758560e-01 5.23055732e-01 2.42146149e-01 9.66013074e-02
-1.33720791e+00 8.50945055e-01 5.14075696e-01 3.14992636e-01
-8.91737282e-01 -5.98223507e-01 -5.73098540e-01 3.19952309e-01
1.23330027e-01 -6.10276461e-01 1.57604849e+00 -7.50566244e-01
-1.66209865e+00 -7.70060048e-02 5.86812058e-03 -1.06225431e+00
2.40932092e-01 1.83128014e-01 -5.25595188e-01 3.30070466e-01
-4.45762515e-01 5.11043251e-01 2.06471205e-01 -1.08527362e+00
-7.05981255e-01 -1.24326028e-01 3.67073148e-01 -2.40404055e-01
-5.05599439e-01 -7.61672258e-02 -1.68446705e-01 -6.28842890e-01
2.95798779e-01 -4.72383708e-01 -3.45413089e-01 5.77320337e-01
-3.62260669e-01 3.91274355e-02 1.04634583e+00 -3.39435488e-02
1.15484345e+00 -2.10074568e+00 -3.28156710e-01 3.54545951e-01
5.55226922e-01 5.42654276e-01 -1.55933918e-02 6.44577980e-01
4.12132829e-01 8.31035897e-02 1.22279584e-01 -1.41693532e-01
-2.18700394e-02 4.73885298e-01 -2.01046243e-01 4.86713111e-01
-3.38160187e-01 3.83687854e-01 -7.73226380e-01 -3.57993543e-01
2.82196850e-01 7.08188236e-01 -5.57094932e-01 -3.63560766e-02
1.08211055e-01 2.26454958e-02 -6.90924227e-01 3.31190854e-01
9.49846685e-01 -4.52503748e-02 3.58149081e-01 -3.04926485e-01
-3.14607084e-01 3.35240901e-01 -8.04480553e-01 1.12218487e+00
-6.83281422e-01 1.50402641e+00 4.26970392e-01 -1.44748735e+00
8.41930866e-01 4.50191617e-01 4.24659699e-01 -1.02204359e+00
4.63008553e-01 3.22339624e-01 4.86115426e-01 -9.24140736e-02
2.90282011e-01 -1.48791313e-01 2.96266019e-01 3.58590215e-01
4.50886160e-01 2.82435060e-01 3.32712047e-02 1.51577458e-01
1.17005885e+00 -5.95111132e-01 2.08205104e-01 -5.00561535e-01
4.80663866e-01 -5.86006641e-01 3.04372460e-01 1.16487598e+00
-3.46322626e-01 -7.86399543e-02 6.30100727e-01 -1.94876175e-02
-9.74273205e-01 -1.13323486e+00 -5.71864486e-01 5.73614061e-01
6.54767692e-01 -1.94807708e-01 -1.00690746e+00 -9.97673348e-02
-2.94609696e-01 6.82669461e-01 -2.21343398e-01 -2.58731157e-01
4.11686599e-02 -9.90096629e-01 1.14489949e+00 3.01366538e-01
1.30913579e+00 -5.32078862e-01 -6.29983366e-01 4.92205590e-01
2.56449819e-01 -1.33087313e+00 2.49508917e-01 6.44689381e-01
-9.19956446e-01 -6.64625049e-01 -9.36288416e-01 -5.75054765e-01
3.71769160e-01 2.11603969e-01 5.27386189e-01 1.60368934e-01
9.14072916e-02 3.74295264e-01 -5.17289340e-01 -6.07425809e-01
-6.40836120e-01 2.82881171e-01 1.60010129e-01 -2.92682886e-01
3.48773539e-01 -7.56987631e-01 -4.88810718e-01 2.15064958e-01
-6.98386729e-01 -9.85752940e-02 8.52554381e-01 4.71045226e-01
-3.35188419e-01 4.61152703e-01 6.22800350e-01 -4.32215244e-01
8.21221292e-01 -3.86482596e-01 -7.78231025e-01 1.41243845e-01
-6.61872149e-01 4.32761461e-01 6.30571783e-01 -2.73872074e-02
-1.00869918e+00 -4.84884858e-01 -7.24868357e-01 3.34995657e-01
1.75819993e-01 3.06822300e-01 -2.35543147e-01 -4.21157360e-01
3.56674522e-01 2.92481840e-01 1.20767109e-01 -2.72578955e-01
1.07305385e-02 1.03450191e+00 1.52638271e-01 -4.08491462e-01
7.36119747e-01 4.20777947e-01 3.17316294e-01 -1.39884508e+00
-7.75181174e-01 -4.43116166e-02 -2.98785508e-01 -4.37714875e-01
6.64560676e-01 -7.20046937e-01 -1.05218828e+00 4.25196826e-01
-1.35089910e+00 -4.62282002e-01 -4.87698615e-02 1.03035069e+00
-4.62383002e-01 2.84000754e-01 -7.11622834e-01 -1.38477123e+00
-1.99661911e-01 -9.89484847e-01 4.79446232e-01 3.77471507e-01
2.18922406e-01 -1.28364861e+00 -2.96314895e-01 -7.91091323e-02
7.61073112e-01 -3.16546768e-01 6.74087822e-01 -3.48205119e-01
-7.57959545e-01 1.00542270e-01 -7.76282310e-01 6.38526797e-01
-2.39418596e-01 -2.87605911e-01 -1.16776121e+00 -1.56576306e-01
7.33450726e-02 8.00678805e-02 8.52091432e-01 6.78472698e-01
1.16491234e+00 -2.32364446e-01 -5.34046292e-01 7.76531041e-01
1.47575510e+00 4.67602789e-01 5.69765210e-01 7.25689232e-02
-1.42584117e-02 3.41861188e-01 -1.31041080e-01 5.34937441e-01
5.98726720e-02 5.92037499e-01 6.89230978e-01 -5.22171557e-02
-3.18250358e-01 9.23852772e-02 2.32964978e-01 1.04429722e+00
2.58144494e-02 -1.06541395e+00 -5.25466025e-01 7.31135756e-02
-1.67410898e+00 -9.45827067e-01 -2.41437942e-01 1.73159873e+00
4.93131191e-01 4.31539744e-01 -1.63532659e-01 4.49868351e-01
6.47666276e-01 2.19499558e-01 -3.34436059e-01 -4.31730360e-01
-1.54762954e-01 1.47881871e-03 1.08961606e+00 4.16484475e-01
-5.44179738e-01 3.95309210e-01 7.05578423e+00 1.10428512e+00
-9.90812063e-01 7.53383189e-02 2.10876822e-01 2.32672259e-01
-2.12860957e-01 -1.69702441e-01 -9.69155014e-01 5.88117182e-01
1.32177353e+00 -2.99218833e-01 2.51741976e-01 6.18968010e-01
4.79756951e-01 -3.21730763e-01 -1.20613837e+00 1.02383387e+00
-2.90090293e-01 -1.36617339e+00 2.91667163e-01 6.43496811e-01
2.22381160e-01 -2.37900317e-02 -1.76984057e-01 2.81567901e-01
1.63538635e-01 -7.62293220e-01 6.35665476e-01 4.77518082e-01
3.11346918e-01 -7.64127135e-01 1.16103458e+00 4.79261220e-01
-8.67223501e-01 -2.74615794e-01 -6.78928316e-01 -4.17335331e-01
3.54900330e-01 9.18793976e-01 -4.88067389e-01 5.35947978e-01
3.65301490e-01 3.03154916e-01 2.69368857e-01 1.40284252e+00
5.02573401e-02 7.85790563e-01 -4.38007653e-01 -8.24339926e-01
4.69077617e-01 1.13940440e-01 4.28201556e-01 1.23325121e+00
4.51410353e-01 1.94624498e-01 -4.40284222e-01 9.74171519e-01
-3.67455304e-01 -4.20500457e-01 -4.94650006e-01 1.04486771e-01
6.93892539e-01 8.11676502e-01 -5.50336182e-01 -1.90215871e-01
-5.26157260e-01 6.25668406e-01 -1.51463926e-01 4.72346693e-01
-6.17087007e-01 -7.73632765e-01 9.63072717e-01 -3.38898040e-02
1.75176650e-01 -4.63770360e-01 -2.87361681e-01 -6.91271603e-01
-1.62265956e-01 -2.77438819e-01 -3.28811198e-01 -4.38758314e-01
-9.13331807e-01 3.69875818e-01 -1.65207399e-04 -1.08454418e+00
-1.19672157e-02 -8.99969161e-01 -3.18564266e-01 5.44595242e-01
-1.69088304e+00 -2.49892697e-01 -2.57964581e-01 2.17697039e-01
3.02270383e-01 -3.46944362e-01 6.76643014e-01 8.23007941e-01
-5.74922800e-01 9.23575103e-01 7.30990410e-01 4.99307632e-01
-3.59803624e-02 -7.11287379e-01 6.67766556e-02 8.35258901e-01
-8.32820982e-02 3.98390204e-01 8.33973825e-01 1.07875317e-02
-1.41266477e+00 -7.47212887e-01 5.98501623e-01 2.71485627e-01
5.13029993e-01 -4.13891703e-01 -3.04961622e-01 1.98827535e-01
3.59241635e-01 -1.28536463e-01 5.47963440e-01 -2.69966274e-02
1.88239962e-01 -5.47850013e-01 -1.09092999e+00 6.57233059e-01
1.05421567e+00 -2.68746763e-01 -2.46328443e-01 -1.26722634e-01
5.01525164e-01 -7.07153678e-02 -6.56917691e-01 1.12558194e-01
8.89465332e-01 -1.00124443e+00 8.21188211e-01 7.43335038e-02
1.42297238e-01 1.09969467e-01 -4.22327995e-01 -1.23583925e+00
9.59992260e-02 -5.14255226e-01 -1.54158294e-01 1.11165059e+00
3.49452168e-01 -6.78490996e-01 9.41562593e-01 4.21508811e-02
-1.05807111e-01 -6.87324703e-01 -1.11138749e+00 -1.29186487e+00
1.67858656e-02 -8.79242420e-01 2.70272374e-01 2.67816603e-01
2.02403605e-01 2.09127232e-01 -2.34750062e-01 2.75904834e-01
9.44407046e-01 -5.92169464e-01 4.73933935e-01 -1.30905652e+00
3.61106247e-02 -8.66682827e-01 -8.07866752e-01 -1.87462199e+00
3.60771008e-02 -7.49442279e-01 3.42508964e-02 -1.52819848e+00
-6.31988347e-02 -4.97133732e-01 -2.93298692e-01 -1.89311728e-01
7.20706403e-01 9.05293524e-02 2.30911419e-01 -7.58808777e-02
-4.43478614e-01 5.93191385e-01 8.90573442e-01 6.00856356e-02
2.46385202e-01 3.56161237e-01 -7.45833278e-01 7.35918522e-01
9.12975252e-01 -3.53766859e-01 -5.96491039e-01 -6.19816363e-01
6.90082163e-02 2.77597815e-01 5.00012755e-01 -1.56566226e+00
5.21370709e-01 2.97809899e-01 2.17147395e-01 -4.65045571e-01
4.27323222e-01 -1.27897751e+00 -3.73745263e-01 7.78366506e-01
-5.02910674e-01 -6.29400849e-01 -1.40653700e-01 6.34103298e-01
-1.85116515e-01 -6.88803852e-01 8.55930328e-01 2.49150947e-01
-5.97434580e-01 2.37938732e-01 -1.06727493e+00 -6.90899789e-02
9.40972090e-01 -3.76627237e-01 -2.49011636e-01 -7.39015222e-01
-3.85189742e-01 2.77305245e-01 -2.32337669e-01 -1.10137612e-01
5.61655045e-01 -1.15689552e+00 -2.40205050e-01 7.28153959e-02
-2.40415648e-01 -4.62711066e-01 -1.49422009e-02 6.14078581e-01
-7.52315700e-01 8.74200165e-01 -1.77313000e-01 -5.26532114e-01
-8.88586879e-01 4.79342937e-02 7.42231905e-01 1.01710126e-01
-4.10329998e-01 8.64186466e-01 1.92821268e-02 5.56134656e-02
9.85858798e-01 -5.60103297e-01 1.02744822e-03 -1.99083716e-01
4.69355226e-01 5.12802958e-01 -2.34716073e-01 2.03393530e-02
-3.52706045e-01 2.90876359e-01 2.00908408e-01 -2.00572476e-01
1.11036384e+00 -5.79578042e-01 1.38216212e-01 3.10331970e-01
1.36924386e+00 -3.33156765e-01 -9.99325931e-01 -3.83026928e-01
-8.89200717e-04 -1.49372026e-01 7.06808865e-01 -3.43334496e-01
-1.38514388e+00 1.09688437e+00 7.70998478e-01 7.56626606e-01
1.17406023e+00 1.28104892e-02 8.58296335e-01 8.61980677e-01
5.75249732e-01 -1.06924033e+00 -1.61259457e-01 5.41696310e-01
1.81520507e-01 -8.70712817e-01 -1.11425839e-01 -3.43871117e-01
3.02809745e-01 1.66521668e+00 1.58288434e-01 1.52436242e-01
1.14785099e+00 7.74073482e-01 -1.22493394e-01 -1.02807716e-01
-7.17380702e-01 -2.59244531e-01 -4.22860265e-01 7.09127784e-01
6.14071846e-01 -1.67806745e-01 -3.72682780e-01 2.39382342e-01
-4.19694364e-01 5.73186837e-02 7.59204984e-01 5.50897539e-01
-8.56847286e-01 -1.24019170e+00 -2.14543119e-02 3.25926930e-01
-5.72020352e-01 -2.15727001e-01 -2.80849040e-02 9.46562886e-01
1.01804063e-01 1.12293935e+00 1.89264342e-01 -4.28315967e-01
5.42041957e-02 -5.94269156e-01 3.36986601e-01 -1.78239882e-01
-3.63630219e-03 -3.05217892e-01 3.52758914e-01 -1.68309659e-01
-7.20023990e-01 -3.22630167e-01 -1.04232657e+00 -8.20768595e-01
-4.49841112e-01 3.95616025e-01 9.23986673e-01 1.20761573e+00
-2.95307450e-02 8.92072499e-01 5.12007892e-01 -4.50649440e-01
-3.57365668e-01 -6.77424371e-01 -1.04321814e+00 -8.73161912e-01
8.33575070e-01 -4.21379864e-01 -3.90317500e-01 -5.16558349e-01] | [6.292558670043945, 1.5346989631652832] |
c1ab5db4-382f-48ef-ae96-e5ed794ca8ff | speech2properties2gestures-gesture-property | 2106.14736 | null | https://arxiv.org/abs/2106.14736v2 | https://arxiv.org/pdf/2106.14736v2.pdf | Speech2Properties2Gestures: Gesture-Property Prediction as a Tool for Generating Representational Gestures from Speech | We propose a new framework for gesture generation, aiming to allow data-driven approaches to produce more semantically rich gestures. Our approach first predicts whether to gesture, followed by a prediction of the gesture properties. Those properties are then used as conditioning for a modern probabilistic gesture-generation model capable of high-quality output. This empowers the approach to generate gestures that are both diverse and representational. Follow-ups and more information can be found on the project page: https://svito-zar.github.io/speech2properties2gestures/ . | ['Gustav Eje Henter', 'Hedvig Kjellström', 'Michael Neff', 'Patrik Jonell', 'Rajmund Nagy', 'Taras Kucherenko'] | 2021-06-28 | null | null | null | null | ['gesture-generation'] | ['robots'] | [ 7.81067088e-02 2.69569188e-01 -1.62390381e-01 -4.86608833e-01
-1.03021622e+00 -7.78271019e-01 1.07815123e+00 -3.73527199e-01
-1.46530822e-01 5.42696714e-01 8.32785726e-01 -2.29969844e-01
1.28417060e-01 -9.19889331e-01 -5.98385036e-01 -5.04061937e-01
-2.18655746e-02 7.15859830e-01 1.41442627e-01 -2.53643543e-01
1.17014199e-01 4.41519141e-01 -1.83452356e+00 7.39865482e-01
2.16977254e-01 5.37202120e-01 3.55765313e-01 1.09045208e+00
-1.53044090e-01 3.01456839e-01 -3.72172177e-01 -3.71307552e-01
1.60849154e-01 -5.74009717e-01 -9.70382273e-01 -5.47864258e-01
-6.58779964e-02 -5.13195932e-01 -2.19269902e-01 7.41841257e-01
7.28061557e-01 4.39259470e-01 9.04267251e-01 -1.21540093e+00
-4.13672060e-01 1.15723658e+00 3.29732448e-01 -3.81477088e-01
1.03516567e+00 4.78217274e-01 1.13142121e+00 -8.05756211e-01
1.04659915e+00 1.41116250e+00 8.62915888e-02 1.33396828e+00
-1.01081765e+00 -6.11918151e-01 5.13478667e-02 -1.86279044e-01
-1.30751646e+00 -6.59820676e-01 5.47977090e-01 -2.86406487e-01
7.54861176e-01 7.53427744e-01 7.90206313e-01 1.91542935e+00
-5.14072478e-01 1.25394332e+00 8.37161720e-01 -6.02205217e-01
2.56424457e-01 -2.65227914e-01 -9.40096602e-02 5.27652621e-01
-1.14660710e-01 6.76571965e-01 -8.83332849e-01 -2.16255393e-02
8.68714988e-01 -3.33003730e-01 -2.40952909e-01 -2.12382507e-02
-1.37880135e+00 5.31247020e-01 1.56153753e-01 4.83627379e-01
-5.27907014e-01 5.28514981e-01 8.16823989e-02 1.46474183e-01
1.16188072e-01 4.46235061e-01 -4.69526589e-01 -7.73321271e-01
-7.85413444e-01 9.18104708e-01 9.97715533e-01 1.13395584e+00
3.28941554e-01 -2.23352224e-01 -3.91823977e-01 4.90207016e-01
8.41593325e-01 6.91240907e-01 5.27718961e-01 -8.76589775e-01
4.36961949e-01 3.34958851e-01 1.87154263e-01 -2.03588679e-01
-2.03412071e-01 4.12698567e-01 -2.05282554e-01 4.08132046e-01
6.83057129e-01 -3.15844953e-01 -9.62382555e-01 1.63402152e+00
3.29157323e-01 1.01952360e-03 9.72733349e-02 1.13678956e+00
1.04595482e+00 6.00146234e-01 4.50052768e-01 3.09312314e-01
1.32486475e+00 -4.63505715e-01 -5.22826195e-01 4.52648044e-01
6.12919867e-01 -7.56525815e-01 1.48606193e+00 3.23161066e-01
-1.01375258e+00 -3.11756968e-01 -7.10987151e-01 1.80899873e-02
-3.20927829e-01 -3.02102193e-02 8.32364082e-01 8.00033331e-01
-1.16784012e+00 7.51366258e-01 -1.31881201e+00 -5.51665723e-01
3.47300470e-01 2.70933330e-01 -1.18401542e-01 2.33778939e-01
-9.56061721e-01 7.32507288e-01 6.40450060e-01 -7.57124871e-02
-8.73652458e-01 -3.32912892e-01 -7.89155602e-01 -3.76388520e-01
7.60410503e-02 -7.42533922e-01 1.70357835e+00 -7.70910621e-01
-2.10878825e+00 5.34405708e-01 -2.15955347e-01 -5.20239547e-02
8.35700214e-01 -4.37474728e-01 -1.41129911e-01 3.64405126e-03
-2.78331667e-01 1.01883388e+00 4.42366302e-01 -1.29507744e+00
-5.54315627e-01 -2.74057448e-01 3.94061431e-02 4.31873709e-01
1.12786360e-01 2.36022204e-01 -4.67870325e-01 -7.00983286e-01
-7.03504086e-02 -9.37966704e-01 -2.29928792e-01 -2.46434018e-01
-7.67185330e-01 -3.76710773e-01 2.91040927e-01 -5.69565773e-01
9.49932337e-01 -1.85530484e+00 2.82669604e-01 3.08242738e-01
-3.22197855e-01 -5.35577722e-02 -4.21158701e-01 7.15249836e-01
2.52741072e-02 3.77880841e-01 -2.12867297e-02 -4.27086830e-01
4.90401238e-01 1.12390026e-01 -4.18313295e-01 -8.91681686e-02
2.88302481e-01 1.20206320e+00 -1.10254538e+00 -4.70564902e-01
4.81244951e-01 6.70391083e-01 -4.72005218e-01 5.76816559e-01
-7.22059250e-01 9.39065576e-01 -8.20484161e-01 6.18476093e-01
4.72737700e-02 1.61939397e-01 2.62454569e-01 2.16018841e-01
-1.53893977e-01 7.34370530e-01 -1.26483226e+00 2.02724385e+00
-3.42353702e-01 2.48550832e-01 -3.98894489e-01 -1.31163463e-01
8.35503638e-01 5.74444115e-01 2.66081333e-01 -2.36663103e-01
3.09177756e-01 1.62422657e-01 -3.03382933e-01 -7.06266761e-01
5.05345702e-01 -3.21049064e-01 -1.63350806e-01 7.13228941e-01
2.53529493e-02 -5.63028872e-01 2.88269192e-01 5.40174954e-02
9.41614807e-01 1.15878248e+00 1.12699112e-02 4.63850833e-02
9.21731442e-02 6.45373836e-02 3.70151014e-03 6.43636703e-01
1.69079140e-01 7.38110960e-01 5.61980344e-02 -1.94423482e-01
-9.22417581e-01 -1.32840741e+00 1.21005967e-01 1.30770516e+00
-7.55065605e-02 -5.99740148e-01 -8.53329360e-01 -7.00064540e-01
-3.02382350e-01 1.13143337e+00 -4.38724071e-01 2.22069964e-01
-5.81478000e-01 -1.92892775e-01 9.30212796e-01 6.41446352e-01
-1.87269270e-01 -1.76646495e+00 -7.83665478e-01 -8.52934942e-02
-3.57520580e-01 -4.62853551e-01 -3.01932663e-01 1.34222701e-01
-9.69080091e-01 -8.11946630e-01 -9.18356955e-01 -6.08819664e-01
3.97045910e-01 -6.97496951e-01 1.01983285e+00 3.42634529e-01
-4.72464375e-02 4.07636017e-01 -8.60264122e-01 -4.14682716e-01
-7.36203492e-01 2.19420299e-01 -6.83898106e-02 -3.30856204e-01
3.79313827e-01 -5.46692431e-01 -3.73642832e-01 -3.81271467e-02
-1.02786386e+00 4.10344303e-01 4.67661917e-01 5.33852696e-01
4.70428616e-01 -7.03431427e-01 7.95985833e-02 -5.69326580e-01
6.98350787e-01 -2.65717268e-01 -5.16897619e-01 7.98540115e-02
-1.86015517e-01 5.72459459e-01 2.42399633e-01 -6.22173429e-01
-1.17337739e+00 6.44359410e-01 -6.74453199e-01 -8.25911202e-03
-1.00986707e+00 2.95786917e-01 -4.15000975e-01 6.36513948e-01
6.42564118e-01 3.50201875e-01 -1.95398331e-01 -7.08567023e-01
9.68322337e-01 5.28340399e-01 3.79904926e-01 -1.01011670e+00
7.81775892e-01 2.30186090e-01 -3.42865258e-01 -7.96710134e-01
-1.12049319e-02 -1.04089633e-01 -9.06448603e-01 -3.30879211e-01
7.72219777e-01 -4.35685426e-01 -7.96062648e-01 3.90890628e-01
-1.20613217e+00 -1.08358395e+00 -5.35288632e-01 7.51782715e-01
-9.61157858e-01 4.81944717e-02 -5.42436957e-01 -1.15355957e+00
-2.71354705e-01 -1.06388497e+00 1.36463702e+00 2.12948799e-01
-9.47344482e-01 -7.57205606e-01 2.80272365e-01 3.85504737e-02
1.52284533e-01 1.04291849e-01 6.33539736e-01 -8.19510162e-01
-5.45471430e-01 -3.83817181e-02 2.41316751e-01 -1.12825513e-01
-1.12074479e-01 3.73361379e-01 -9.09334004e-01 2.94257551e-01
-7.75945902e-01 -4.08828467e-01 5.09201288e-01 2.73637027e-01
7.77071834e-01 -3.38840753e-01 -5.00571311e-01 5.43384016e-01
9.79879618e-01 1.86732575e-01 7.05457091e-01 1.06923386e-01
5.81021607e-01 7.63787389e-01 5.65337896e-01 4.80967134e-01
4.87733305e-01 1.00641227e+00 2.98705220e-01 4.28909719e-01
-4.62026209e-01 -8.20468128e-01 5.08434236e-01 5.61715305e-01
-5.56303740e-01 -4.74568099e-01 -1.06765699e+00 5.48875213e-01
-1.78687894e+00 -1.20384872e+00 -3.37695867e-01 2.03364277e+00
1.13581002e+00 -3.62423986e-01 4.95050251e-01 1.61778163e-02
4.60181683e-01 2.17378005e-01 -9.79355797e-02 -2.75353521e-01
2.36366421e-01 5.91534197e-01 1.53137073e-01 7.92664289e-01
-9.13264573e-01 1.41801703e+00 6.27172136e+00 6.35753393e-01
-9.66501892e-01 -1.09574124e-01 2.63840348e-01 -3.90951067e-01
-6.83875442e-01 7.68444687e-02 -7.77877569e-01 4.59684491e-01
1.08036065e+00 6.23649806e-02 6.44080102e-01 6.28225684e-01
4.29589391e-01 9.59298387e-02 -1.19519556e+00 7.87669599e-01
-1.73162654e-01 -1.18635070e+00 4.92460251e-01 4.27380018e-03
3.09605211e-01 -5.42910434e-02 -4.01368737e-02 2.54811376e-01
8.09177518e-01 -1.08269274e+00 1.26474321e+00 7.12555766e-01
1.04128706e+00 -4.20731992e-01 2.83797532e-01 4.34839934e-01
-1.12674403e+00 2.43411571e-01 2.30595589e-01 -1.08152151e-01
4.89029884e-01 -9.35800746e-02 -1.05396152e+00 5.14820516e-01
4.94851947e-01 2.47439802e-01 -1.65190145e-01 8.95945966e-01
-8.75425100e-01 6.91383481e-01 -4.50163305e-01 -6.54022217e-01
-6.37054369e-02 5.92135359e-03 7.27808654e-01 1.36413622e+00
5.08453131e-01 3.46146196e-01 6.30979240e-02 8.95227551e-01
1.41336888e-01 2.90179312e-01 -7.60544956e-01 -2.23470926e-01
6.76835597e-01 8.51500094e-01 -6.67213321e-01 -4.12362874e-01
1.94091782e-01 1.13427150e+00 2.27651354e-02 4.09610391e-01
-4.88214493e-01 -9.52328593e-02 6.53886497e-01 2.40035951e-02
8.54377821e-02 -4.55049962e-01 -4.18794960e-01 -1.17730641e+00
-9.43834633e-02 -9.38380957e-01 3.26668471e-01 -8.53885591e-01
-8.15354764e-01 7.33591437e-01 3.63477379e-01 -1.11109412e+00
-8.74352276e-01 -6.55402780e-01 -5.68723023e-01 9.25057769e-01
-8.94585431e-01 -1.53849411e+00 -1.96687520e-01 6.16566718e-01
6.59767687e-01 5.74465767e-02 1.43865252e+00 -9.49122757e-02
-2.79088621e-03 4.18077767e-01 -5.39615631e-01 3.96120489e-01
4.03667748e-01 -1.44162512e+00 6.62457883e-01 7.34684348e-01
7.22968161e-01 5.53307891e-01 8.44465017e-01 -9.37575996e-01
-1.35847020e+00 -7.93571472e-01 1.06280184e+00 -8.77205431e-01
4.91206169e-01 -3.16255391e-01 -3.52259248e-01 6.97188973e-01
-2.86575145e-04 -4.87450659e-01 6.32342041e-01 1.51573971e-01
-2.95888156e-01 6.52123392e-01 -8.35659683e-01 1.05115163e+00
1.42790806e+00 -3.53966594e-01 -7.71444559e-01 9.98711735e-02
4.45785522e-01 -6.04501963e-01 -7.35280871e-01 9.87773389e-02
1.06881177e+00 -6.05850756e-01 6.56898916e-01 -7.27832794e-01
4.95653093e-01 -1.51163369e-01 -2.96877354e-01 -1.20175552e+00
1.34088248e-01 -1.02840400e+00 -2.35762089e-01 1.09807253e+00
8.45026255e-01 -3.68255645e-01 9.69172299e-01 6.70961738e-01
-8.15936774e-02 -3.42975110e-01 -8.04522932e-01 -7.65094459e-01
1.16082780e-01 -1.03995657e+00 9.41085160e-01 5.45386016e-01
3.36305797e-01 1.33432848e-02 -5.73679693e-02 2.01648310e-01
2.51480550e-01 6.23138398e-02 1.03148222e+00 -1.01025116e+00
-4.81259078e-01 -7.76798010e-01 -2.45098114e-01 -1.27954876e+00
-8.29664543e-02 -1.19380939e+00 4.84260291e-01 -1.69080245e+00
-3.02077770e-01 -5.47906280e-01 4.64709774e-02 8.08616936e-01
2.17489712e-02 3.59989166e-01 4.43615049e-01 2.51272261e-01
-4.43358481e-01 5.46026170e-01 1.19661295e+00 1.97041035e-01
-6.41280830e-01 2.18650118e-01 -3.51480752e-01 5.26593089e-01
9.89763021e-01 -5.63945711e-01 -1.10123806e-01 -4.37767476e-01
-1.62939683e-01 9.22326520e-02 3.70942235e-01 -9.31224287e-01
-5.86845055e-02 -3.55046004e-01 3.16758037e-01 -4.48420256e-01
6.71615481e-01 -4.83758122e-01 4.13277537e-01 4.74005044e-01
-7.00698137e-01 -5.26140220e-02 -2.08647270e-02 1.50167376e-01
7.27552250e-02 -3.09259146e-01 1.43064246e-01 -1.51568472e-01
-6.02051318e-01 2.28729442e-01 -5.59439778e-01 -2.92682081e-01
8.50178778e-01 -2.25525841e-01 -5.51110506e-02 -6.31827295e-01
-1.15028739e+00 -1.53900668e-01 4.72612411e-01 8.01510870e-01
6.51262045e-01 -1.35684252e+00 -9.58163083e-01 8.13964680e-02
1.44959688e-01 -1.75294116e-01 -2.04793423e-01 4.49689835e-01
-6.21956229e-01 2.43292242e-01 -1.73702553e-01 -3.42368424e-01
-1.40862989e+00 6.14814423e-02 4.69532870e-02 1.26554770e-02
-7.87015736e-01 9.89657462e-01 -4.75711048e-01 -6.27421379e-01
2.65110821e-01 -4.39330488e-01 -1.32719114e-01 -2.33721405e-01
6.56897604e-01 8.50150362e-02 -3.57909650e-01 -7.12332308e-01
-3.45222980e-01 2.04973027e-01 5.41333556e-01 -1.13147950e+00
1.39597547e+00 2.98871726e-01 3.25381845e-01 6.08749509e-01
5.85299432e-01 1.47492141e-01 -1.35785460e+00 2.45086029e-01
3.82463455e-01 -4.29159284e-01 -4.10144001e-01 -1.25006783e+00
-5.43516159e-01 6.86767459e-01 4.59385097e-01 2.46275831e-02
7.74922311e-01 3.37395668e-01 6.29449904e-01 2.60944188e-01
6.02417648e-01 -7.30368435e-01 -1.19402029e-01 6.32165909e-01
1.15028715e+00 -7.59651840e-01 -4.47453529e-01 -2.01194197e-01
-7.96864867e-01 1.25922501e+00 2.42097661e-01 -1.44729272e-01
4.82854307e-01 5.45014799e-01 4.41793323e-01 -1.48516029e-01
-7.37360418e-01 -6.29106641e-01 4.74012434e-01 1.17929268e+00
7.88849235e-01 5.88581443e-01 -4.39103186e-01 7.03989804e-01
-7.35745370e-01 9.85459611e-02 8.01762417e-02 8.41668189e-01
-4.36168425e-02 -1.70272255e+00 -3.46281677e-01 1.81767270e-01
-1.35248214e-01 -3.87616664e-01 -6.30961359e-01 7.54777014e-01
-3.17741111e-02 8.89264643e-01 -1.31792128e-01 -6.71866357e-01
3.97372842e-01 6.39619052e-01 7.26162195e-01 -9.15834665e-01
-6.56961679e-01 -3.47760543e-02 4.66966450e-01 -7.84741700e-01
-1.53367430e-01 -9.03835058e-01 -1.48135865e+00 -7.17600361e-02
1.78630985e-02 1.12946041e-01 8.76597345e-01 7.45437622e-01
2.16485277e-01 2.61496186e-01 2.66824290e-02 -1.41699779e+00
-4.50015098e-01 -1.18628895e+00 -2.05298707e-01 6.22558832e-01
-1.59581870e-01 -4.42686498e-01 -1.28813341e-01 5.20495355e-01] | [5.601586818695068, -0.11921820044517517] |
ad924e17-3246-460f-9c1b-8b8a06507afd | gpurir-a-python-library-for-room-impulse | 1810.11359 | null | http://arxiv.org/abs/1810.11359v1 | http://arxiv.org/pdf/1810.11359v1.pdf | gpuRIR: A python library for Room Impulse Response simulation with GPU acceleration | The Image Source Method (ISM) is one of the most employed techniques to
calculate acoustic Room Impulse Responses (RIRs), however, its computational
complexity grows fast with the reverberation time of the room and its
computation time can be prohibitive for some applications where a huge number
of RIRs are needed. In this paper, we present a new implementation that
dramatically improves the computation speed of the ISM by using Graphic
Processing Units (GPUs) to parallelize both the simulation of multiple RIRs and
the computation of the images inside each RIR. We provide a Python library
under GNU license that can be easily used without any knowledge about GPU
programming and we show that it is about 100 times faster than other state of
the art CPU libraries. | ['David Diaz-Guerra', 'Jose R. Beltran', 'Antonio Miguel'] | 2018-10-26 | null | null | null | null | ['room-impulse-response'] | ['audio'] | [ 2.68542111e-01 -6.63198411e-01 1.26182461e+00 2.24185344e-02
-6.89634383e-01 -6.05282426e-01 1.95417121e-01 -1.29778525e-02
-5.19226730e-01 2.49613911e-01 1.37806116e-02 -7.76303649e-01
1.46094874e-01 -9.91538167e-01 -4.04481620e-01 -8.44545484e-01
1.65105894e-01 1.96452618e-01 5.65526187e-01 -2.10909739e-01
1.18929781e-01 4.18834835e-01 -1.67373037e+00 6.97972998e-02
5.66804588e-01 8.60526562e-01 3.96222949e-01 1.35372889e+00
6.26671165e-02 6.59104884e-01 -5.90254664e-01 1.93359047e-01
8.74999464e-02 -4.25282001e-01 -5.76350570e-01 -4.27990943e-01
-1.30066022e-01 -1.04738198e-01 -2.11739898e-01 7.70683706e-01
9.93389785e-01 6.22145534e-01 3.25985223e-01 -6.36022210e-01
4.20967899e-02 1.73848569e-01 -4.17235792e-01 2.76055276e-01
8.07345033e-01 -2.35097870e-01 1.65322125e-01 -8.05993378e-01
1.24362111e-01 9.92927790e-01 6.00542665e-01 2.40813985e-01
-9.19756055e-01 -6.35508835e-01 -5.16640782e-01 5.63387461e-02
-1.37130237e+00 -4.31860059e-01 7.01753914e-01 -3.66674438e-02
1.00249767e+00 8.82012963e-01 4.06134248e-01 7.18752325e-01
-2.11229324e-01 1.02855787e-01 1.38727486e+00 -7.62912750e-01
5.29571772e-01 4.87952679e-03 3.11845958e-01 5.49830079e-01
-1.66384563e-01 3.75555083e-02 -1.88985452e-01 -5.17956972e-01
9.56623793e-01 -2.46131599e-01 -3.84902835e-01 2.67852217e-01
-1.06484520e+00 5.67010760e-01 2.81900823e-01 4.90419716e-01
-3.00180241e-02 2.78640449e-01 3.89443606e-01 1.00965224e-01
1.56712830e-01 -2.76195128e-02 1.15157701e-02 -3.86633009e-01
-6.54446006e-01 2.80696124e-01 1.11161864e+00 5.63442051e-01
4.98136252e-01 6.81127384e-02 3.75611424e-01 1.00203788e+00
3.27794880e-01 7.37803698e-01 2.71609694e-01 -7.52572775e-01
1.04759410e-01 5.05205663e-03 2.94136107e-01 -9.72945869e-01
-3.62216413e-01 -1.60448059e-01 -8.21616948e-01 3.79694641e-01
4.88420337e-01 -2.46928111e-01 -6.35857701e-01 8.46212626e-01
5.36008239e-01 3.24759930e-01 1.85543820e-02 8.63399088e-01
1.11094677e+00 1.29481435e+00 -2.08498418e-01 -5.89832403e-02
1.54672372e+00 -1.02055407e+00 -5.34883499e-01 -2.78906345e-01
3.83220106e-01 -1.36544299e+00 1.02351058e+00 3.21937561e-01
-1.11604643e+00 -6.11865401e-01 -9.24093187e-01 1.36861265e-01
-9.90029797e-02 -6.97634667e-02 3.92495334e-01 1.11516309e+00
-1.05949628e+00 1.77001089e-01 -8.87990355e-01 -6.92437738e-02
-3.92080605e-01 3.34913164e-01 -9.80619341e-02 -1.56832024e-01
-7.33955562e-01 5.51444471e-01 -2.33932093e-01 2.66030371e-01
-7.77211249e-01 -6.84022665e-01 -7.03744054e-01 1.37988046e-01
1.55892044e-01 -3.92199904e-01 1.30553675e+00 -5.69024384e-01
-1.81125140e+00 4.39460516e-01 -1.90139815e-01 1.22053966e-01
5.14540553e-01 -1.58597142e-01 -3.08403254e-01 6.09217621e-02
-5.73433042e-01 -2.89603382e-01 7.58372128e-01 -1.29980338e+00
-1.13842152e-02 1.02943815e-01 -1.02366090e-01 3.37727189e-01
3.50819826e-02 4.52424675e-01 -3.60021442e-01 -3.09791028e-01
1.18649073e-01 -1.04635882e+00 -5.11237919e-01 -6.06341541e-01
-1.65567130e-01 7.64817819e-02 5.75132072e-01 -6.13793015e-01
1.03310347e+00 -2.43488002e+00 -2.42468014e-01 4.15176988e-01
-1.58430412e-01 3.96769255e-01 2.98634470e-01 6.44606054e-01
1.52087342e-02 -3.43532443e-01 -3.49744290e-01 -2.55030245e-01
-2.02004403e-01 4.90364805e-02 -3.02516371e-01 5.74022114e-01
-9.37326789e-01 1.24129087e-01 -7.54886806e-01 -1.90785572e-01
4.92701948e-01 8.74575973e-01 -3.72059017e-01 4.73782569e-01
5.45512736e-01 7.54314184e-01 -4.20741647e-01 7.51276836e-02
1.08743203e+00 1.00069590e-01 6.00112528e-02 -1.57903358e-02
-6.62903845e-01 4.65408415e-01 -1.57970059e+00 1.18783617e+00
-1.08090234e+00 3.89351785e-01 4.06267464e-01 -7.10283875e-01
8.55831802e-01 5.95031738e-01 6.18486032e-02 -5.66240549e-01
2.09263548e-01 4.50731546e-01 -2.02368245e-01 -3.42107922e-01
2.73573101e-01 -2.57625669e-01 1.43671706e-01 7.96112955e-01
-6.43295467e-01 -4.86005813e-01 -9.87425521e-02 -1.78789929e-01
1.21123898e+00 -1.59155324e-01 9.24627259e-02 -3.07273209e-01
8.29659879e-01 -1.63429976e-01 -6.24478683e-02 7.30757833e-01
1.45468876e-01 7.20760584e-01 -6.71463087e-02 -5.33534884e-01
-1.10062075e+00 -1.07574368e+00 -9.84884501e-02 1.11831319e+00
1.10041574e-02 -2.74357021e-01 -9.78791475e-01 1.78739667e-01
-6.39617920e-01 5.59269965e-01 -4.69548292e-02 4.07410890e-01
-9.46517766e-01 -1.00070226e+00 5.21267295e-01 3.68748546e-01
6.74089134e-01 -1.14423239e+00 -1.02738261e+00 3.08766961e-01
-3.40125531e-01 -1.32790732e+00 -3.44298899e-01 2.00861648e-01
-7.00319707e-01 -5.82314730e-01 -7.57461071e-01 -9.03130412e-01
8.82867396e-01 6.68875873e-01 9.51891363e-01 4.17151481e-01
-6.80380642e-01 6.49908781e-01 -4.55548733e-01 -5.00772238e-01
-5.64722240e-01 -4.24401551e-01 -1.10145569e-01 -1.92892611e-01
-1.59041584e-01 -7.33380020e-01 -7.19035089e-01 3.63450497e-01
-8.62130463e-01 1.63705766e-01 -2.19732560e-02 2.70172030e-01
3.35118681e-01 3.86890769e-01 2.19461992e-02 -6.43461168e-01
5.84908545e-01 6.72997758e-02 -8.90180588e-01 -9.24905539e-02
2.35496201e-02 -1.30818769e-01 7.94546247e-01 -2.46182874e-01
-1.32606232e+00 1.52131960e-01 -7.65947461e-01 1.86781347e-01
-3.12395364e-01 1.19732030e-01 2.46143922e-01 -5.45168519e-01
5.09595156e-01 3.24830443e-01 -4.34049994e-01 -6.42476320e-01
-1.31510198e-01 7.32356369e-01 4.66769934e-01 -4.88525301e-01
8.74806106e-01 5.01815975e-01 9.18310657e-02 -1.23225963e+00
-4.50110465e-01 -7.15078831e-01 -1.96127862e-01 -3.27040195e-01
8.44924629e-01 -7.52270520e-01 -1.03431225e+00 6.82505906e-01
-1.20351613e+00 -4.79312211e-01 2.53011346e-01 7.09328890e-01
-2.56575495e-01 4.83293355e-01 -6.63573802e-01 -1.13335621e+00
-5.59696138e-01 -1.15285027e+00 7.39145219e-01 3.18189949e-01
-9.90367383e-02 -9.19195950e-01 3.93547922e-01 3.96308154e-01
7.80839801e-01 1.29235998e-01 6.76499009e-01 8.72724205e-02
-4.56926107e-01 -3.63111794e-01 -1.21326752e-01 1.54614672e-02
-1.06233865e-01 -5.31702489e-02 -1.36529243e+00 -2.90838093e-01
7.20844924e-01 1.68781713e-01 4.25785393e-01 4.07459706e-01
1.18928838e+00 -1.08350284e-01 -2.33893082e-01 4.68273133e-01
1.73208964e+00 3.70859832e-01 8.81338358e-01 3.36457610e-01
5.86647809e-01 5.04107416e-01 2.86507100e-01 6.49031758e-01
5.51146939e-02 5.79378486e-01 2.85952985e-01 -2.99162835e-01
8.17775503e-02 2.05777377e-01 3.44638258e-01 1.17467964e+00
-7.27350712e-01 -2.19934851e-01 -1.00541270e+00 1.96113557e-01
-1.36375225e+00 -7.77552426e-01 -6.65660322e-01 2.65284371e+00
4.98954117e-01 -3.53739560e-01 -1.55324191e-01 5.61942518e-01
5.12723267e-01 8.59829634e-02 3.66883695e-01 -9.64473486e-01
4.71327990e-01 6.32754266e-01 3.29451799e-01 9.07013774e-01
-8.04611504e-01 4.35804546e-01 7.20679617e+00 4.86568540e-01
-8.89845312e-01 3.66725802e-01 4.17724818e-01 1.20830990e-01
-1.67253211e-01 -2.17493579e-01 -5.36880016e-01 1.90699473e-01
1.21491146e+00 4.95818965e-02 6.89407170e-01 8.44491959e-01
3.82209420e-01 -5.03786623e-01 -5.34927964e-01 1.22393084e+00
-5.24010696e-02 -7.60990322e-01 -6.37126982e-01 -1.31625772e-01
3.32536668e-01 -9.12228450e-02 4.16329429e-02 -1.82680905e-01
2.16716737e-01 -1.01014674e+00 3.78000021e-01 -8.19239691e-02
4.48925406e-01 -1.00743639e+00 6.97386742e-01 3.77283663e-01
-1.37286901e+00 2.70588785e-01 -6.98414207e-01 -3.12838525e-01
3.92905116e-01 7.30471373e-01 -6.79942846e-01 3.36623579e-01
9.77896392e-01 -5.88195980e-01 -3.39236975e-01 1.19847882e+00
-1.25192881e-01 8.06687117e-01 -7.10618854e-01 -1.61011919e-01
2.84941256e-01 -3.75725150e-01 4.29127306e-01 1.26564372e+00
7.58003354e-01 2.81006157e-01 -1.66989919e-02 4.56322074e-01
4.04032558e-01 3.16945523e-01 -3.96404743e-01 6.22227073e-01
8.70888606e-02 1.50260806e+00 -1.03887415e+00 -2.73721784e-01
-4.40503657e-01 1.15692174e+00 -2.50762522e-01 2.59872377e-01
-1.11895490e+00 -6.80545032e-01 2.76455194e-01 1.81075260e-01
1.39332741e-01 -7.66018152e-01 -3.16015393e-01 -6.51043713e-01
-2.73571879e-01 -5.74926078e-01 6.57612160e-02 -9.30359483e-01
-6.56120479e-01 9.89495158e-01 -2.60088928e-02 -1.06688023e+00
2.75005568e-02 -5.24721980e-01 -8.12161863e-01 1.06786501e+00
-1.24769282e+00 -6.73450112e-01 -7.01742589e-01 8.23672652e-01
3.41352135e-01 4.98002976e-01 1.22974336e+00 5.54563224e-01
-1.80058286e-01 3.38120848e-01 3.50199729e-01 -6.44286498e-02
4.12378132e-01 -9.96499360e-01 4.65299606e-01 7.89498866e-01
-1.40449122e-01 6.52320027e-01 1.21097851e+00 -1.87777162e-01
-1.71545088e+00 -5.66244841e-01 8.29262257e-01 -8.49636197e-02
4.20272380e-01 -5.77165604e-01 -8.68978918e-01 3.05711299e-01
2.15067729e-01 2.08833903e-01 1.02734661e+00 -8.78935456e-02
-1.79617703e-01 -1.16558380e-01 -9.35783863e-01 4.74031419e-01
4.92594898e-01 -4.59195584e-01 -8.84730890e-02 5.35964131e-01
2.79167145e-01 -8.46750438e-01 -5.01041234e-01 -1.61967337e-01
3.75945866e-01 -1.23673713e+00 1.29275477e+00 7.34518290e-01
-8.28139558e-02 -5.81814826e-01 -1.50109395e-01 -1.00829506e+00
-2.12598637e-01 -6.57744884e-01 3.03092211e-01 1.07621777e+00
7.51831979e-02 -7.67240822e-01 3.11807215e-01 7.03090847e-01
-1.25795797e-01 1.46240070e-02 -1.03325403e+00 -6.55184925e-01
-3.87266606e-01 -8.41612756e-01 4.48477834e-01 2.42760465e-01
-1.20554239e-01 1.94792166e-01 -4.49739456e-01 5.61165035e-01
6.77768290e-01 2.72237420e-01 6.14236653e-01 -8.92988265e-01
-5.12333930e-01 2.15667978e-01 -1.31522164e-01 -8.92618060e-01
-3.57831895e-01 -4.03021872e-01 3.72143835e-01 -1.55632854e+00
-5.71883246e-02 -6.40787363e-01 5.95593229e-02 1.89220145e-01
-5.70708141e-02 5.74957430e-01 1.41107859e-02 -4.77632955e-02
-1.51640803e-01 -3.28323580e-02 1.18740928e+00 3.74909878e-01
-4.18596238e-01 3.27688009e-01 -1.15385003e-01 8.96265805e-01
6.90494835e-01 -6.54394329e-01 -2.21944600e-01 -5.27939141e-01
2.91012734e-01 -1.51646724e-02 4.52502012e-01 -1.46114135e+00
2.92813838e-01 2.12776616e-01 3.26079994e-01 -5.32797396e-01
3.37424338e-01 -7.95022309e-01 3.96337897e-01 6.36327505e-01
1.01227805e-01 1.83714762e-01 3.97278279e-01 2.82004565e-01
-6.19326159e-02 -6.39868259e-01 1.03751338e+00 -4.34966236e-01
-1.60659984e-01 -8.17977637e-02 -9.88613725e-01 -4.70942616e-01
8.05078745e-01 4.17177826e-02 4.93710041e-02 -4.35893685e-01
-3.98213506e-01 -6.19572043e-01 3.85784447e-01 -1.14590742e-01
5.66177964e-01 -9.66849029e-01 -4.35378104e-01 1.99861512e-01
-4.50933397e-01 7.98692480e-02 7.64413238e-01 6.66855574e-01
-1.39423048e+00 1.96456373e-01 1.21156476e-01 -4.16830689e-01
-1.85438669e+00 6.33224010e-01 2.69528031e-01 -2.46719912e-01
-1.08162856e+00 9.14519250e-01 5.03964782e-01 -2.98028082e-01
-1.34944757e-02 -4.09899279e-02 -2.65183896e-01 -4.47994500e-01
1.14448214e+00 8.67098153e-01 2.63438106e-01 -6.44035518e-01
-4.92933065e-01 9.38151062e-01 4.21778917e-01 -4.46751386e-01
1.33712673e+00 -2.34217152e-01 -4.69165921e-01 3.17660600e-01
1.17127430e+00 4.24894005e-01 -8.13264847e-01 1.46051541e-01
-5.68375230e-01 -5.56098521e-01 2.57598281e-01 -4.38923299e-01
-6.97894275e-01 1.14817798e+00 8.49629402e-01 2.96177953e-01
1.44701159e+00 -2.59952515e-01 9.90796626e-01 2.00860158e-01
5.41699767e-01 -8.74245763e-01 -1.37337357e-01 6.62633777e-01
6.79238796e-01 -7.61229753e-01 1.49105368e-02 -8.80861223e-01
-2.74126828e-01 1.43073034e+00 6.99246675e-02 -2.04432234e-01
8.17781508e-01 8.99346232e-01 2.25588873e-01 8.42505395e-02
-1.27122954e-01 1.59669742e-01 2.99343597e-02 5.22491157e-01
7.57427216e-01 1.07032172e-01 -3.30287278e-01 3.41121912e-01
-3.48779500e-01 -2.21457273e-01 6.86662614e-01 1.09578335e+00
-5.59351027e-01 -1.13806653e+00 -1.12064826e+00 -2.01062977e-01
-8.83618414e-01 -2.79232353e-01 3.92144173e-01 7.46815503e-02
-2.23965853e-01 1.28219450e+00 -2.12996379e-01 -1.09924428e-01
2.64425844e-01 -1.63535565e-01 6.22846901e-01 -3.97824228e-01
-8.64128292e-01 4.12113011e-01 1.79914963e-02 -3.67079675e-01
-3.10648173e-01 -4.03932899e-01 -1.60618496e+00 -6.73793972e-01
-6.50468171e-02 2.91606754e-01 9.81556654e-01 7.11246192e-01
1.69528406e-02 7.39198267e-01 6.87244058e-01 -1.05501211e+00
6.00529723e-02 -6.96335733e-01 -5.75947523e-01 -1.46975974e-02
1.23479530e-01 -2.56295919e-01 -5.50337732e-01 2.32952107e-02] | [15.263585090637207, 5.6859612464904785] |
6699f504-3976-41d7-b7c2-62579fff9242 | end-to-end-robust-joint-unsupervised-image | null | null | http://openaccess.thecvf.com//content/ICCV2021/html/Zeng_End-to-End_Robust_Joint_Unsupervised_Image_Alignment_and_Clustering_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Zeng_End-to-End_Robust_Joint_Unsupervised_Image_Alignment_and_Clustering_ICCV_2021_paper.pdf | End-to-End Robust Joint Unsupervised Image Alignment and Clustering | Computing dense pixel-to-pixel image correspondences is a fundamental task of computer vision. Often, the objective is to align image pairs from the same semantic category for manipulation or segmentation purposes. Despite achieving superior performance, existing deep learning alignment methods cannot cluster images; consequently, clustering and pairing images needed to be a separate laborious and expensive step. Given a dataset with diverse semantic categories, we propose a multi-task model, Jim-Net, that can directly learn to cluster and align images without any pixel-level or image-level annotations. We design a pair-matching alignment unsupervised training algorithm that selectively matches and aligns image pairs from the clustering branch. Our unsupervised Jim-Net achieves comparable accuracy with state-of-the-art supervised methods on benchmark 2D image alignment dataset PF-PASCAL. Specifically, we apply Jim-Net to cryo-electron tomography, a revolutionary 3D microscopy imaging technique of native subcellular structures. After extensive evaluation on seven datasets, we demonstrate that Jim-Net enables systematic discovery and recovery of representative macromolecular structures in situ, which is essential for revealing molecular mechanisms underlying cellular functions. To our knowledge, Jim-Net is the first end-to-end model that can simultaneously align and cluster images, which significantly improves the performance as compared to performing each task alone. | ['Min Xu', 'Gregory Howe', 'Xiangrui Zeng'] | 2021-01-01 | null | null | null | iccv-2021-1 | ['electron-tomography'] | ['medical'] | [ 3.90457362e-01 -4.02205139e-01 5.50208725e-02 -5.60789764e-01
-1.08255970e+00 -6.41229212e-01 2.73889065e-01 4.01058525e-01
-6.53043330e-01 4.69895989e-01 -4.95573908e-01 -1.16300955e-01
1.88095868e-01 -2.37132952e-01 -9.16564226e-01 -9.03001130e-01
2.29437858e-01 1.04787946e+00 2.08117351e-01 4.14496064e-01
3.82327855e-01 6.27567172e-01 -1.19442475e+00 3.44815850e-01
5.50592005e-01 6.61366105e-01 7.70567894e-01 5.64864695e-01
-2.53145814e-01 3.50210667e-01 -1.85573623e-01 -9.06844065e-02
3.66348535e-01 -4.32502389e-01 -1.09640789e+00 2.01633096e-01
9.81216729e-01 -8.91453251e-02 1.44206613e-01 1.22169673e+00
6.11438036e-01 -2.32663900e-01 6.41969800e-01 -1.13004041e+00
-5.71555972e-01 8.52120593e-02 -6.57257199e-01 3.06402236e-01
1.07945437e-02 1.02160625e-01 9.53545928e-01 -8.25355113e-01
1.05393219e+00 8.52956235e-01 8.13152552e-01 5.32879293e-01
-1.94118392e+00 -5.00809908e-01 -2.48947367e-02 1.70948356e-01
-1.24747753e+00 -3.69308054e-01 6.48497283e-01 -6.11621141e-01
1.11528945e+00 -7.25210831e-02 5.90206623e-01 5.57179272e-01
1.40315056e-01 6.77533031e-01 1.31827629e+00 -1.77798912e-01
9.20720473e-02 -6.12574875e-01 6.60302863e-02 7.62507975e-01
-5.84532097e-02 -4.02911425e-01 -3.32806885e-01 -3.97601984e-02
7.72420347e-01 3.09707463e-01 -1.64336741e-01 -7.98336804e-01
-1.76954150e+00 1.99508816e-01 4.15812224e-01 2.97477454e-01
-3.43443274e-01 1.36878937e-01 3.48668396e-01 2.05396995e-01
2.96965897e-01 6.08928680e-01 -5.64849913e-01 5.56028038e-02
-1.08523774e+00 2.57902205e-01 4.04408664e-01 7.36540139e-01
1.00416660e+00 -7.36039460e-01 2.28116557e-01 7.27394700e-01
2.86870878e-02 2.82297462e-01 3.45144480e-01 -1.30216718e+00
1.12741590e-02 6.32064521e-01 -3.02735530e-02 -9.79547799e-01
-6.47961617e-01 -5.39490208e-02 -8.84777665e-01 2.64093250e-01
7.43128479e-01 3.76230121e-01 -9.40175295e-01 1.82417572e+00
4.86111462e-01 2.36409724e-01 -1.85235262e-01 1.05519938e+00
6.67602539e-01 3.64945203e-01 3.03754420e-03 -2.18922511e-01
1.20505488e+00 -9.36636448e-01 -4.65190470e-01 -8.67490098e-02
7.08590686e-01 -1.05777705e+00 9.90657210e-01 4.87649739e-02
-9.98663962e-01 -4.36454266e-01 -8.68686497e-01 -4.41109270e-01
-3.51275414e-01 1.30995885e-02 5.98967314e-01 -1.69152722e-01
-9.87335801e-01 7.60773659e-01 -1.00842917e+00 -5.05221963e-01
7.19792008e-01 4.85473573e-01 -9.01076853e-01 -3.27124409e-02
-2.94130862e-01 8.60166609e-01 2.41169944e-01 -1.83378443e-01
-8.76254320e-01 -9.30295765e-01 -5.33284664e-01 -6.94464669e-02
-1.16362214e-01 -8.82631242e-01 1.12596309e+00 -7.71586537e-01
-1.19763112e+00 1.77998281e+00 -4.62914020e-01 -3.98265988e-01
2.80784726e-01 -1.56741403e-02 2.20279679e-01 2.97375768e-01
4.50637847e-01 1.07183850e+00 3.90664428e-01 -1.41443360e+00
-5.17578602e-01 -6.61307573e-01 -4.88904357e-01 1.66027039e-01
6.24825358e-02 3.13566849e-02 -6.44342780e-01 -3.27378780e-01
4.33730602e-01 -8.64922345e-01 -3.06785196e-01 4.08875048e-01
-5.02548873e-01 -1.35187088e-02 9.96140599e-01 -5.37383556e-01
4.03320879e-01 -1.97393990e+00 5.77721059e-01 -2.66628005e-02
5.30035794e-01 1.61601916e-01 -1.28214940e-01 6.77467510e-02
-2.33855650e-01 -1.58272624e-01 -4.55548465e-01 -8.28314960e-01
-1.54693291e-01 1.72566399e-01 7.62637779e-02 8.87267828e-01
-5.49844354e-02 1.05290914e+00 -8.96499932e-01 -7.71858454e-01
4.36568052e-01 1.94617823e-01 -4.43312496e-01 4.67412204e-01
-2.61813074e-01 8.14540744e-01 -1.97841274e-03 7.91439533e-01
7.01642513e-01 -8.91235530e-01 4.34308589e-01 -7.15209603e-01
-4.21900526e-02 7.14774802e-02 -6.31955266e-01 2.18873668e+00
-6.07254170e-02 6.03949368e-01 2.95990735e-01 -1.50820637e+00
8.01127613e-01 -2.21643560e-02 9.23464477e-01 -7.07996964e-01
1.08168967e-01 3.69098365e-01 -7.95114040e-02 -2.35094860e-01
-5.59563516e-03 8.11477378e-02 7.03971237e-02 5.54463327e-01
2.85249591e-01 -4.53349173e-01 1.55061156e-01 1.54063180e-01
1.14118707e+00 3.01270157e-01 1.84021145e-01 -4.08008724e-01
3.40604663e-01 3.39944541e-01 7.91829109e-01 4.06143665e-01
-4.22270000e-01 8.10736597e-01 2.28773654e-01 -7.96595871e-01
-1.65972888e+00 -1.20724380e+00 -2.49667034e-01 8.70258510e-01
4.52795982e-01 -1.42290637e-01 -8.43425691e-01 -5.30061662e-01
2.49793772e-02 -2.58333802e-01 -2.75216401e-01 2.53931761e-01
-5.85836172e-01 -9.22497213e-01 3.46859366e-01 2.54351467e-01
3.73864532e-01 -8.89022231e-01 -3.14294964e-01 2.03590423e-01
-3.44735444e-01 -1.44318318e+00 -5.60962260e-01 5.39204478e-01
-8.02356303e-01 -1.39696240e+00 -6.12843573e-01 -1.33490217e+00
1.21037769e+00 4.74426150e-01 1.24027729e+00 9.29835737e-02
-5.96270740e-01 6.20822534e-02 9.21553746e-02 7.05440938e-02
-1.65451109e-01 7.25066885e-02 8.47822055e-02 -3.80097300e-01
5.32779574e-01 -8.58044863e-01 -7.01491714e-01 5.61228812e-01
-5.98185718e-01 5.09539187e-01 5.22292793e-01 8.87382507e-01
1.56041372e+00 -3.48165810e-01 2.10644230e-01 -9.85766053e-01
1.50982320e-01 -3.54365371e-02 -7.85705030e-01 2.62651056e-01
-2.93845862e-01 -1.42617747e-01 7.69463181e-01 -1.21691875e-01
-3.81467640e-01 7.71526158e-01 -7.34211579e-02 -6.80350900e-01
-4.25295591e-01 1.77509949e-01 -1.31526858e-01 -4.10026103e-01
3.30033064e-01 3.30487251e-01 4.16218251e-01 -3.93287808e-01
3.99070710e-01 3.53258967e-01 1.26546764e+00 -7.27203906e-01
5.88026524e-01 8.32114339e-01 2.83486098e-01 -6.23830795e-01
-9.34985936e-01 -9.73461568e-01 -1.19915557e+00 -5.58481254e-02
1.20894516e+00 -8.72348428e-01 -9.06553864e-01 7.65812874e-01
-1.17852068e+00 -6.31787300e-01 9.74290147e-02 1.37823388e-01
-9.92415845e-01 6.45862162e-01 -8.35179269e-01 8.36659744e-02
-4.77841228e-01 -1.43303442e+00 1.50745976e+00 3.58924344e-02
-2.23351821e-01 -8.84112716e-01 1.95141658e-01 5.70520699e-01
1.14699498e-01 3.47836673e-01 9.64032531e-01 -5.52754462e-01
-7.37189114e-01 -1.24599691e-02 -4.89971906e-01 1.24207720e-01
4.12903368e-01 6.23701774e-02 -7.11597681e-01 -3.80068094e-01
-3.42528939e-01 -5.16749322e-01 7.62321174e-01 4.54708546e-01
1.60842252e+00 1.52336910e-01 -6.18433774e-01 1.03724813e+00
1.38236070e+00 -1.26440570e-01 4.20338571e-01 4.85778242e-01
1.00964248e+00 4.22746211e-01 4.92055327e-01 8.87031108e-03
3.46909195e-01 8.88193369e-01 4.96690303e-01 -5.64060211e-01
-6.52134717e-02 8.70853439e-02 -1.60432264e-01 9.74904358e-01
2.06098318e-01 1.13607444e-01 -9.38484073e-01 5.98949432e-01
-2.03510141e+00 -9.80631888e-01 -1.14939541e-01 1.98244667e+00
1.07641768e+00 -1.77287400e-01 3.52373756e-02 -2.86871612e-01
9.51055586e-01 -1.51726842e-01 -7.36643076e-01 1.86156169e-01
-1.94311231e-01 1.96872145e-01 5.07774234e-01 3.19548219e-01
-1.40541077e+00 1.04204297e+00 6.49608040e+00 7.17531264e-01
-1.02666938e+00 2.39803310e-04 8.48594189e-01 1.83365829e-02
5.44556454e-02 -2.92667821e-02 -8.57105434e-01 6.65945947e-01
4.06375885e-01 1.52002305e-01 4.22654480e-01 6.28276885e-01
1.85932055e-01 1.07085459e-01 -1.46206224e+00 1.20468128e+00
-6.67537078e-02 -1.82967138e+00 2.97568068e-02 1.83483660e-01
8.13580453e-01 3.85565072e-01 -1.07197128e-01 -4.75788951e-01
4.59216505e-01 -1.02485907e+00 3.33980232e-01 4.99378622e-01
6.68523073e-01 -5.07917523e-01 7.13262081e-01 1.95457444e-01
-1.04093254e+00 4.95386958e-01 -6.09334290e-01 3.70774209e-01
3.88199031e-01 8.93708408e-01 -7.62255371e-01 2.56744534e-01
9.34103489e-01 1.00754511e+00 -3.17408413e-01 9.87595022e-01
2.09898561e-01 7.06246644e-02 -3.54822904e-01 5.19405723e-01
1.35454327e-01 -5.19828320e-01 2.35306010e-01 1.18074369e+00
1.97817028e-01 -7.33093470e-02 6.19816184e-01 1.03842783e+00
-4.82413858e-01 -7.07217827e-02 -5.37752926e-01 -1.04036957e-01
6.04464233e-01 1.65561819e+00 -1.13991678e+00 -2.58631974e-01
-2.91731477e-01 1.32626283e+00 8.05380940e-01 3.79006080e-02
-6.88339591e-01 -1.37339398e-01 1.05432892e+00 -5.16682258e-03
2.50018269e-01 -4.73793477e-01 -3.95366490e-01 -9.64099109e-01
-7.16946572e-02 -6.82271719e-01 1.89444304e-01 -7.75862217e-01
-1.55111146e+00 1.68867141e-01 -4.63828385e-01 -1.02645838e+00
2.13186145e-01 -6.91945732e-01 -5.06867349e-01 6.98440433e-01
-1.41516888e+00 -1.27933025e+00 -5.50278485e-01 4.30850595e-01
2.61646658e-01 -8.27428177e-02 9.21940207e-01 4.95704204e-01
-5.26190102e-01 3.21769118e-01 4.97378856e-01 1.71696216e-01
9.68377650e-01 -1.55066383e+00 4.96656746e-01 6.92100942e-01
1.78215265e-01 7.29474843e-01 4.64991868e-01 -3.82791728e-01
-1.29980743e+00 -1.26694238e+00 7.53647327e-01 -5.05845964e-01
5.29785156e-01 -3.78662556e-01 -1.07098877e+00 6.26370013e-01
7.28154331e-02 3.53678733e-01 8.59198868e-01 -2.07692206e-01
-2.72263646e-01 -2.38199020e-03 -1.17083073e+00 5.64679742e-01
1.05498576e+00 -8.06503594e-01 -4.24130350e-01 7.07042515e-01
7.28079915e-01 -5.20748675e-01 -1.22679007e+00 4.09805983e-01
4.28931117e-01 -9.34950054e-01 1.35899901e+00 -4.61992890e-01
4.67731446e-01 -8.75194371e-01 -2.59147882e-01 -1.15776777e+00
-3.79837334e-01 -3.83278221e-01 2.43836924e-01 1.14364862e+00
2.66737461e-01 -2.05684707e-01 9.62635338e-01 2.74106294e-01
-4.76937830e-01 -6.06508493e-01 -9.83105302e-01 -5.68541765e-01
8.99128318e-02 1.28601477e-01 3.68816018e-01 1.16381741e+00
-1.70058683e-01 2.39049509e-01 -1.18653387e-01 1.33778155e-01
1.03365004e+00 5.42606950e-01 1.12154746e+00 -1.13454187e+00
-1.45631686e-01 -5.93750834e-01 -6.50288284e-01 -1.19439590e+00
4.83847469e-01 -1.06758165e+00 4.01689082e-01 -1.43844867e+00
7.25872278e-01 -4.72534180e-01 -1.54871538e-01 6.11109078e-01
-2.11371392e-01 6.54791474e-01 -3.42003815e-02 7.22482800e-01
-1.02113318e+00 2.45848686e-01 1.31912136e+00 -2.49078840e-01
1.04192644e-01 -4.82540518e-01 -3.31589162e-01 6.73356593e-01
7.49399424e-01 -3.02104414e-01 5.11776879e-02 -4.56479430e-01
-2.10223515e-02 -3.18549722e-01 4.50109839e-01 -1.03476727e+00
5.33930063e-01 -9.33240950e-02 5.02595901e-01 -8.18628013e-01
1.84286430e-01 -8.61514688e-01 2.20317841e-01 2.69706219e-01
-2.99016804e-01 1.20459132e-01 -4.35402282e-02 4.34907168e-01
-2.44978696e-01 2.00341016e-01 1.22048497e+00 -3.36239755e-01
-8.49786341e-01 5.75314820e-01 -2.33239502e-01 -6.08303547e-02
1.09993386e+00 -1.48649618e-01 -5.52080512e-01 2.76620448e-01
-7.71296144e-01 2.86716312e-01 1.29087782e+00 -6.77949712e-02
5.41258395e-01 -1.16005874e+00 -2.89341927e-01 3.12978886e-02
2.51949817e-01 5.68862140e-01 7.23983021e-03 1.03071201e+00
-9.55534399e-01 1.77722916e-01 -5.97565711e-01 -1.19576049e+00
-1.57454538e+00 5.08452296e-01 4.88777548e-01 -2.15106532e-01
-5.88214159e-01 7.89699435e-01 3.42698038e-01 -9.65181053e-01
-4.47667949e-03 -1.84999466e-01 1.51003867e-01 -3.97558898e-01
2.51418710e-01 -5.46563864e-02 1.03134602e-01 -7.02951312e-01
-3.83831173e-01 9.93363440e-01 -3.72146577e-01 4.64210540e-01
1.34976876e+00 -3.13677013e-01 -7.56741226e-01 2.40184918e-01
1.47799206e+00 -7.25105226e-01 -1.29389894e+00 -2.73120254e-01
-3.32902819e-02 -3.86324078e-01 -1.32216185e-01 -4.57287759e-01
-1.03181195e+00 9.11492407e-01 6.35463059e-01 -3.14846426e-01
9.65771139e-01 2.51201510e-01 9.61909473e-01 6.01215780e-01
4.40245956e-01 -1.00489163e+00 1.05440117e-01 4.46205288e-01
1.81715503e-01 -1.50390291e+00 9.47778001e-02 -4.19164538e-01
-1.54374421e-01 9.27756667e-01 8.33488405e-01 -1.34234786e-01
3.06981117e-01 3.63962263e-01 2.04553381e-01 -4.97316778e-01
-5.74776351e-01 -2.60935187e-01 -3.71272564e-02 7.42889047e-01
4.91194278e-01 -1.53041303e-01 2.81535424e-02 -5.79960011e-02
1.38988107e-01 -1.97801799e-01 1.56104015e-02 8.64558935e-01
-5.48713207e-01 -1.22327232e+00 -1.28326133e-01 4.33863074e-01
-5.85780025e-01 5.77316098e-02 -5.17053425e-01 4.00842130e-01
1.20545551e-02 4.07867193e-01 4.42102224e-01 -2.47321352e-01
-5.35159633e-02 -4.11066301e-02 6.07704937e-01 -5.40754557e-01
-2.09666014e-01 4.38833274e-02 -3.38299334e-01 -7.14800000e-01
-9.29878235e-01 -6.85516298e-01 -1.65179694e+00 -4.34089333e-01
-7.92834759e-02 4.63453401e-03 5.66536605e-01 1.10479176e+00
6.40000165e-01 3.23868275e-01 4.30609733e-01 -1.30862916e+00
7.64613152e-02 -5.84954500e-01 -3.62865418e-01 9.02745008e-01
1.13568559e-01 -4.53915596e-01 -1.05507262e-01 5.83684623e-01] | [13.969453811645508, -3.0978147983551025] |
2a94bce3-7372-438e-866e-dd65a029aba3 | semattnet-towards-attention-based-semantic | 2204.13635 | null | https://arxiv.org/abs/2204.13635v1 | https://arxiv.org/pdf/2204.13635v1.pdf | SemAttNet: Towards Attention-based Semantic Aware Guided Depth Completion | Depth completion involves recovering a dense depth map from a sparse map and an RGB image. Recent approaches focus on utilizing color images as guidance images to recover depth at invalid pixels. However, color images alone are not enough to provide the necessary semantic understanding of the scene. Consequently, the depth completion task suffers from sudden illumination changes in RGB images (e.g., shadows). In this paper, we propose a novel three-branch backbone comprising color-guided, semantic-guided, and depth-guided branches. Specifically, the color-guided branch takes a sparse depth map and RGB image as an input and generates color depth which includes color cues (e.g., object boundaries) of the scene. The predicted dense depth map of color-guided branch along-with semantic image and sparse depth map is passed as input to semantic-guided branch for estimating semantic depth. The depth-guided branch takes sparse, color, and semantic depths to generate the dense depth map. The color depth, semantic depth, and guided depth are adaptively fused to produce the output of our proposed three-branch backbone. In addition, we also propose to apply semantic-aware multi-modal attention-based fusion block (SAMMAFB) to fuse features between all three branches. We further use CSPN++ with Atrous convolutions to refine the dense depth map produced by our three-branch backbone. Extensive experiments show that our model achieves state-of-the-art performance in the KITTI depth completion benchmark at the time of submission. | ['Muhammad Zeshan Afzal', 'Didier Stricker', 'Marcus Liwicki', 'Danish Nazir'] | 2022-04-28 | null | null | null | null | ['depth-completion'] | ['computer-vision'] | [ 5.43908656e-01 -9.53072868e-03 7.58123994e-02 -5.54313660e-01
-8.71376157e-01 -3.43356192e-01 2.51393557e-01 -7.78051466e-02
-1.76010981e-01 4.94092375e-01 1.80610403e-01 -6.95185596e-03
2.92015851e-01 -1.14927363e+00 -7.37746596e-01 -8.27825367e-01
5.84730268e-01 2.48296812e-01 4.83972222e-01 -3.92800272e-02
5.67261755e-01 3.63060564e-01 -1.69314539e+00 7.05886364e-01
1.08743000e+00 1.38228607e+00 6.49690390e-01 7.32868612e-01
-6.58144772e-01 7.49168992e-01 -2.37613425e-01 1.65697217e-01
3.71010333e-01 -3.29619825e-01 -7.14734554e-01 2.24518806e-01
4.81583685e-01 -8.93873215e-01 -4.31378752e-01 1.12012744e+00
1.75196290e-01 2.99322367e-01 1.69588119e-01 -1.22791219e+00
-3.31795663e-01 4.41289581e-02 -8.91226828e-01 -4.59647998e-02
4.52664077e-01 2.15486944e-01 5.95550418e-01 -1.02434516e+00
4.24037874e-01 1.25205791e+00 1.86014194e-02 5.47360420e-01
-7.89050102e-01 -7.24834323e-01 6.27235055e-01 2.29999989e-01
-1.01207507e+00 -1.89186528e-01 1.10581088e+00 -7.18187466e-02
7.51070499e-01 -3.51376049e-02 8.67519617e-01 7.24224329e-01
-3.12701277e-02 9.34516191e-01 1.33822536e+00 -2.55290508e-01
3.61039311e-01 -1.67771474e-01 -1.44849673e-01 9.88919318e-01
-4.67911037e-03 1.60825208e-01 -8.41155827e-01 3.10432494e-01
9.71846759e-01 2.53397793e-01 -5.17586529e-01 -1.32835060e-01
-1.01922381e+00 5.04088581e-01 7.78040588e-01 -1.18616819e-01
-4.27167624e-01 3.05934399e-01 -2.05626413e-01 -1.72289923e-01
5.57434976e-01 -1.84074253e-01 -4.01576430e-01 6.34706020e-02
-8.94952834e-01 1.29377142e-01 3.71180058e-01 8.23338687e-01
1.57002962e+00 -1.07318789e-01 9.23898593e-02 5.82307220e-01
4.67609763e-01 6.69360101e-01 2.14408129e-01 -1.28585315e+00
7.87862480e-01 8.25924039e-01 9.68153104e-02 -8.68420482e-01
-3.79979938e-01 2.30457205e-02 -6.92571938e-01 3.66154552e-01
4.25266504e-01 7.58254901e-02 -1.37879765e+00 1.44970918e+00
5.95232069e-01 6.13089144e-01 1.33179694e-01 1.22151840e+00
1.03489673e+00 8.22558939e-01 -2.64062732e-01 1.61090747e-01
9.83227193e-01 -1.28492844e+00 -3.67912143e-01 -7.74404526e-01
3.55928093e-01 -5.38083971e-01 9.49244440e-01 5.83878100e-01
-1.08233976e+00 -5.39070904e-01 -1.03031778e+00 -4.94540572e-01
-1.72379881e-01 -4.41043451e-02 7.64594734e-01 2.71313548e-01
-1.29822791e+00 2.34860286e-01 -9.46322441e-01 -8.91376957e-02
3.58804345e-01 2.45881770e-02 -3.34802121e-01 -9.76673067e-01
-8.57455790e-01 3.25935721e-01 3.97510171e-01 2.37808481e-01
-1.14588416e+00 -7.05938816e-01 -9.81092155e-01 -2.18009755e-01
1.79823756e-01 -7.21123636e-01 1.02584052e+00 -9.04044032e-01
-1.22037554e+00 6.42847657e-01 -6.95125222e-01 1.94352016e-01
2.06928805e-01 -7.31183365e-02 1.93388864e-01 5.96804202e-01
3.57642591e-01 1.14552212e+00 7.87165761e-01 -1.71715748e+00
-1.14926350e+00 -6.51936591e-01 3.51244152e-01 6.57629430e-01
-1.68853495e-02 -6.96845233e-01 -9.88324463e-01 -2.47745112e-01
9.85823691e-01 -5.99069774e-01 -2.48811051e-01 2.45920509e-01
-5.12943685e-01 2.66600281e-01 9.05292928e-01 -7.01540351e-01
8.53469849e-01 -1.99373889e+00 2.22551614e-01 1.59680963e-01
2.50121862e-01 -3.64102840e-01 -3.80325109e-01 4.42706468e-03
7.50942826e-02 -2.52580911e-01 -3.87127906e-01 -7.32610464e-01
-4.61923480e-01 3.94697964e-01 -2.15440080e-01 3.11736166e-01
1.21580288e-01 7.10530519e-01 -1.05963266e+00 -2.99823672e-01
6.15898252e-01 6.73536539e-01 -8.05697083e-01 3.92664522e-01
-4.74311620e-01 7.20889330e-01 -6.24792278e-01 1.10819042e+00
1.00342548e+00 -1.39602140e-01 -1.87464967e-01 -3.96139264e-01
-1.33221298e-01 1.56878650e-01 -9.55846429e-01 2.38326979e+00
-7.08952248e-01 4.07365710e-01 1.51824355e-01 -6.33754313e-01
8.31137598e-01 -2.15339065e-01 4.83168155e-01 -9.18503582e-01
1.13998808e-01 2.25790352e-01 -4.72718269e-01 -1.12853795e-01
6.69500947e-01 -5.09503148e-02 4.56186011e-02 5.17456174e-01
-2.77626097e-01 -6.11048162e-01 -2.20525265e-01 3.75163674e-01
9.79280651e-01 4.46210206e-01 -4.17909503e-01 3.49630237e-01
4.99090225e-01 3.87768932e-02 7.14212477e-01 4.23003048e-01
-1.07420288e-01 1.11124575e+00 4.02450830e-01 -2.90267050e-01
-7.49017894e-01 -1.09257233e+00 3.37422937e-01 8.09120059e-01
8.12396049e-01 2.05594134e-02 -7.43844867e-01 -4.67312068e-01
-4.94208410e-02 6.20194435e-01 -7.48938859e-01 -1.54849976e-01
-3.60970378e-01 -4.92088825e-01 -1.15617933e-02 6.20784044e-01
9.90015626e-01 -1.01067579e+00 -6.69349253e-01 1.73082724e-01
-6.76422536e-01 -1.43342328e+00 -3.41097742e-01 3.00383985e-01
-9.96632755e-01 -1.29060602e+00 -5.91527998e-01 -6.01469219e-01
8.03483248e-01 8.48752916e-01 9.00258362e-01 5.24869077e-02
-1.86108679e-01 4.13855761e-01 -3.79921645e-01 -7.67278448e-02
1.73174158e-01 -3.80392134e-01 -6.08330548e-01 1.77793324e-01
2.65392333e-01 -5.33146381e-01 -1.15268326e+00 1.78279042e-01
-1.01865280e+00 5.28153956e-01 5.52763879e-01 5.33530295e-01
9.09047306e-01 9.49030370e-02 1.38860404e-01 -7.13883102e-01
-2.01656505e-01 -4.87369984e-01 -6.70592606e-01 -2.75443252e-02
-3.81434619e-01 -1.06230028e-01 2.36961171e-01 7.30293691e-02
-1.46017134e+00 2.84149051e-01 -9.03950557e-02 -6.67054713e-01
-2.25072190e-01 1.63441762e-01 -5.00244141e-01 -7.25292638e-02
1.74087897e-01 3.77228737e-01 -2.15315238e-01 -3.16713899e-01
3.25986505e-01 5.29424846e-01 7.34535933e-01 -6.13134384e-01
5.69391251e-01 1.07586658e+00 -1.78729277e-02 -5.24864197e-01
-1.14884317e+00 -6.43257141e-01 -4.38325673e-01 -2.55449712e-01
1.04245007e+00 -1.36391592e+00 -4.26188022e-01 9.64050174e-01
-1.27965665e+00 -7.85118103e-01 9.68288332e-02 1.94709629e-01
-5.22829950e-01 3.96936387e-01 -6.49568737e-01 -6.12044454e-01
-2.51792699e-01 -1.23105180e+00 1.65132964e+00 5.27028024e-01
5.18462837e-01 -9.01305139e-01 -2.95319825e-01 7.86246836e-01
6.20027073e-02 3.00157189e-01 7.52592623e-01 4.18588609e-01
-1.26527441e+00 1.30627885e-01 -7.49122858e-01 3.32168788e-01
1.44428954e-01 -3.42713237e-01 -1.33670616e+00 6.43825764e-03
-1.94197834e-01 -2.56893456e-01 1.13130391e+00 5.15205801e-01
1.43478024e+00 2.51482487e-01 -1.82250187e-01 1.16222405e+00
1.82669568e+00 1.88875079e-01 7.95451283e-01 2.90754378e-01
1.24578357e+00 6.91912055e-01 1.00987506e+00 6.41716421e-01
9.03288603e-01 3.52374345e-01 8.89178991e-01 -2.76988089e-01
-3.89318973e-01 -3.14215422e-01 2.47180089e-01 3.65697831e-01
2.37167433e-01 -2.62446254e-01 -8.27271819e-01 4.74385440e-01
-1.69432008e+00 -4.72339481e-01 -1.27462566e-01 1.95206463e+00
6.64010882e-01 -1.45788444e-02 -3.47462356e-01 8.74301493e-02
3.65127623e-01 2.62250990e-01 -9.31182146e-01 -5.05951904e-02
-3.09585631e-01 2.18812227e-01 4.68763083e-01 7.54254520e-01
-7.61285126e-01 1.08134556e+00 4.37730503e+00 5.03038049e-01
-1.08072019e+00 6.22435808e-02 7.07705319e-01 -2.49568149e-01
-8.85802507e-01 1.21105097e-01 -7.57786989e-01 3.91974449e-01
3.58688384e-01 2.67368555e-01 6.01276517e-01 6.57075107e-01
3.43354970e-01 -9.96186316e-01 -9.22055900e-01 1.36147630e+00
3.05098951e-01 -1.06935227e+00 8.08641613e-02 -3.11699808e-02
1.10475183e+00 1.79181084e-01 3.94548811e-02 -1.40994102e-01
3.35924864e-01 -7.70798147e-01 8.47536147e-01 4.33280259e-01
1.05018234e+00 -7.62612462e-01 6.04581833e-01 2.19402790e-01
-1.41326272e+00 -2.11857438e-01 -3.03540379e-01 2.16385592e-02
2.51335800e-01 9.24681008e-01 -2.35768259e-01 6.39897466e-01
8.22676778e-01 1.05334890e+00 -4.72488791e-01 7.83405185e-01
-5.73229790e-01 4.24407348e-02 -2.53876120e-01 4.98638302e-01
5.07144690e-01 -3.22182894e-01 -3.04869302e-02 5.92980146e-01
4.06514347e-01 4.57745135e-01 2.25692183e-01 9.76164758e-01
1.40097365e-01 -5.09128630e-01 -2.87016600e-01 1.66347474e-01
4.56413120e-01 1.21130967e+00 -8.08671713e-01 -3.90783578e-01
-6.33540273e-01 1.42087960e+00 2.48265162e-01 7.30860770e-01
-6.95216477e-01 -2.11877346e-01 8.75469565e-01 7.45684579e-02
1.17325909e-01 -3.84039134e-02 -7.01256275e-01 -1.20769227e+00
-8.38131085e-02 -3.64479572e-01 1.98174000e-01 -1.31002402e+00
-7.90923595e-01 4.34970021e-01 -1.82587594e-01 -9.50407326e-01
4.11698036e-02 -4.83030200e-01 -4.27590221e-01 1.17149723e+00
-2.11412954e+00 -1.08142829e+00 -1.30497885e+00 9.27610099e-01
7.31631100e-01 4.15643752e-01 3.99628073e-01 2.22674772e-01
-4.24809307e-01 3.92281227e-02 -2.87685156e-01 -1.37072116e-01
4.41910177e-01 -1.20208240e+00 1.94183454e-01 8.32629144e-01
-3.45502645e-01 8.41784552e-02 2.23914832e-01 -7.14540958e-01
-1.42253733e+00 -1.32509995e+00 3.51840317e-01 -1.43156022e-01
1.46341950e-01 -1.46386683e-01 -7.25520849e-01 3.92107815e-01
-4.34973627e-01 2.40220100e-01 1.05960868e-01 -4.17967945e-01
-3.92734587e-01 -3.04627836e-01 -1.14935088e+00 3.19951534e-01
1.11157548e+00 -7.47669637e-01 -1.12295970e-01 1.69520766e-01
8.22146654e-01 -7.19565988e-01 -3.07151586e-01 2.27465153e-01
5.31703055e-01 -1.56458771e+00 1.02273333e+00 3.62537593e-01
9.05704081e-01 -5.51898718e-01 -5.05977690e-01 -1.10922742e+00
1.86378524e-01 1.28269568e-01 1.09908625e-01 8.00643027e-01
6.85794875e-02 -4.39478338e-01 1.26150274e+00 6.01634681e-01
-6.03055239e-01 -7.83587575e-01 -6.33158505e-01 -8.29742923e-02
-1.76584706e-01 -7.16873765e-01 5.77779889e-01 6.15645349e-01
-4.02926415e-01 -9.49532166e-02 1.83063615e-02 3.49753976e-01
8.76172364e-01 5.45753777e-01 7.74052083e-01 -8.69221389e-01
-1.92725956e-01 -2.16655388e-01 -1.31885037e-01 -1.45647681e+00
1.20135643e-01 -6.32812977e-01 3.21353287e-01 -2.06574011e+00
3.33881319e-01 -6.54085159e-01 -3.23460966e-01 6.21595085e-01
-3.34782153e-01 4.85079676e-01 8.25503692e-02 -8.94711912e-02
-4.95126158e-01 6.50845647e-01 1.47453928e+00 -1.35244295e-01
-3.38529110e-01 -4.23562676e-01 -6.74353361e-01 6.90928221e-01
6.71083450e-01 -2.10087463e-01 -7.54779875e-01 -7.04518318e-01
4.77962941e-02 3.66845757e-01 6.01461351e-01 -9.85641778e-01
2.33238965e-01 -3.16333175e-01 5.97315907e-01 -9.94661987e-01
7.75711536e-01 -7.21374571e-01 -1.36751771e-01 2.64870793e-01
1.92042604e-01 -3.28115672e-01 7.02872314e-03 7.74767280e-01
-4.48290229e-01 2.28187367e-01 6.85596347e-01 -4.11966890e-01
-1.24118733e+00 6.14888370e-01 2.67694741e-01 -1.07579194e-01
8.54602575e-01 -6.46906376e-01 -4.40019399e-01 -3.76060635e-01
-4.63706583e-01 4.70391005e-01 7.83081830e-01 3.75474334e-01
1.24547839e+00 -1.06767929e+00 -3.27089965e-01 6.07494235e-01
1.37724817e-01 8.52435648e-01 6.83753848e-01 7.14625061e-01
-7.10716367e-01 2.01672494e-01 -1.34532407e-01 -7.70294309e-01
-9.21422958e-01 1.72905818e-01 3.01757783e-01 8.32955837e-02
-6.53364658e-01 1.20504582e+00 8.86072338e-01 -2.13587075e-01
2.79642224e-01 -6.04258657e-01 2.39098012e-01 -1.29698724e-01
6.48065686e-01 2.69305736e-01 1.20917531e-02 -4.96774554e-01
-3.96008193e-01 8.65901589e-01 1.73444301e-01 -4.08820421e-01
1.11759889e+00 -5.50829232e-01 -2.29925230e-01 2.82610893e-01
1.30065584e+00 -3.30985427e-01 -1.81513560e+00 -2.62547672e-01
-6.30674660e-01 -9.24833000e-01 4.84454155e-01 -8.83587420e-01
-1.60207200e+00 1.17545021e+00 4.26040709e-01 -7.23684907e-01
1.64789319e+00 -3.59788053e-02 1.14794481e+00 -1.65984258e-01
7.43099988e-01 -9.42011774e-01 4.63931233e-01 3.91408414e-01
5.56641281e-01 -1.31879795e+00 -7.07329437e-02 -6.20517373e-01
-4.08125222e-01 1.01869690e+00 1.06058729e+00 1.21157810e-01
4.34812069e-01 2.23891273e-01 1.48632511e-01 -7.62088150e-02
-6.54702723e-01 -4.38458085e-01 1.53737798e-01 7.94889450e-01
-9.49803293e-02 -1.59313083e-01 3.36590111e-01 3.15142602e-01
9.89805833e-02 2.92960089e-03 5.84251702e-01 8.34813476e-01
-7.47724593e-01 -7.88995743e-01 -5.55266380e-01 2.48945355e-01
1.64079487e-01 -2.75056601e-01 -1.16847768e-01 3.61488372e-01
3.91068250e-01 1.23503900e+00 2.20508158e-01 -5.37547529e-01
7.38847628e-02 -3.26773047e-01 6.30059242e-01 -8.34486902e-01
-1.01873174e-01 1.34258166e-01 -2.16977358e-01 -1.35426545e+00
-3.56461108e-01 -2.70805299e-01 -1.71911359e+00 -2.72042811e-01
2.98089702e-02 -2.98498780e-01 9.63587999e-01 8.25495839e-01
3.93035859e-02 5.37473381e-01 6.34198785e-01 -1.21915758e+00
4.38356519e-01 -7.08534122e-01 -7.06412911e-01 2.52135128e-01
5.58192492e-01 -5.70744276e-01 -5.95377922e-01 -2.00072110e-01] | [8.949834823608398, -2.533717632293701] |
36231d20-57c9-4f94-8db5-54989cdc8c26 | sparse-high-dimensional-linear-regression-1 | 2209.08139 | null | https://arxiv.org/abs/2209.08139v4 | https://arxiv.org/pdf/2209.08139v4.pdf | Sparse high-dimensional linear regression with a partitioned empirical Bayes ECM algorithm | Bayesian variable selection methods are powerful techniques for fitting and inferring on sparse high-dimensional linear regression models. However, many are computationally intensive or require restrictive prior distributions on model parameters. In this paper, we proposed a computationally efficient and powerful Bayesian approach for sparse high-dimensional linear regression. Minimal prior assumptions on the parameters are used through the use of plug-in empirical Bayes estimates of hyperparameters. Efficient maximum a posteriori (MAP) estimation is completed through a Parameter-Expanded Expectation-Conditional-Maximization (PX-ECM) algorithm. The PX-ECM results in a robust computationally efficient coordinate-wise optimization, which adjusts for the impact of other predictor variables. The completion of the E-step uses an approach motivated by the popular two-groups approach to multiple testing. The result is a PaRtitiOned empirical Bayes Ecm (PROBE) algorithm applied to sparse high-dimensional linear regression, which can be completed using one-at-a-time or all-at-once type optimization. We compare the empirical properties of PROBE to comparable approaches with numerous simulation studies and an analysis of cancer cell lines drug response study. The proposed approach is implemented in the R package probe. | ['Howard Bondell', 'Anja Zgodic', 'Alexander C. McLain'] | 2022-09-16 | null | null | null | null | ['variable-selection', 'prediction-intervals'] | ['methodology', 'miscellaneous'] | [ 2.69326568e-01 -2.36676827e-01 -1.57322139e-01 -5.06854057e-01
-1.34921169e+00 -1.09906025e-01 1.82349950e-01 3.00800294e-01
-4.50145781e-01 1.12099504e+00 1.06559120e-01 -3.31414729e-01
-5.98526776e-01 -6.59455240e-01 -7.32343137e-01 -1.07167017e+00
1.35142507e-03 9.27111447e-01 -6.48862645e-02 3.63761663e-01
3.16498518e-01 4.98945951e-01 -1.21863031e+00 -1.87259793e-01
7.25420654e-01 5.06855249e-01 1.61013380e-01 6.24462366e-01
2.33749077e-01 9.98606980e-02 -7.79874697e-02 -4.70058359e-02
2.85419263e-02 -2.82182813e-01 -1.34927273e-01 1.64861849e-05
1.02894612e-01 -9.46743488e-02 1.83626264e-01 7.06220269e-01
8.39283705e-01 2.20372200e-01 1.05466664e+00 -1.00281179e+00
8.94356295e-02 3.08243722e-01 -9.81809676e-01 -1.29121810e-01
3.31864655e-01 -4.17596921e-02 6.20019495e-01 -1.27673209e+00
4.69730169e-01 1.32443535e+00 8.20615530e-01 -1.11956991e-01
-1.85180247e+00 -6.48529053e-01 -2.57463127e-01 -1.28526434e-01
-1.81538630e+00 -5.37422359e-01 4.73331511e-01 -7.44251966e-01
6.75626874e-01 3.76676619e-01 3.32238793e-01 7.84773827e-01
3.09200406e-01 3.56025785e-01 1.28356755e+00 -5.54963052e-01
5.98488450e-01 3.02775621e-01 1.98520616e-01 6.68916047e-01
5.73122978e-01 1.68736398e-01 -3.99840117e-01 -1.01085675e+00
7.24930704e-01 -6.87669814e-02 -2.98273582e-02 -4.15784448e-01
-9.58203912e-01 1.17285633e+00 -3.11046302e-01 4.59534973e-02
-7.97503233e-01 4.69087251e-02 2.19223976e-01 -1.74329355e-01
5.66782296e-01 1.23837143e-01 -5.20086288e-01 1.50052260e-03
-1.27066565e+00 3.27052683e-01 8.55787337e-01 6.33533537e-01
7.68382013e-01 -1.65611029e-01 -2.98878998e-01 1.01983988e+00
7.45399237e-01 4.14531291e-01 1.54180512e-01 -7.36880779e-01
1.98201820e-01 1.23856425e-01 9.51247513e-02 -1.14639115e+00
-6.43319726e-01 -4.41427141e-01 -1.17640889e+00 -3.78301740e-02
2.62482435e-01 -3.93532604e-01 -6.51108623e-01 1.67429125e+00
7.81687021e-01 5.86516261e-01 -2.64482558e-01 6.52990043e-01
4.64054585e-01 5.09115398e-01 3.04708064e-01 -8.94215047e-01
1.21311748e+00 -3.29536080e-01 -7.20195472e-01 1.50277197e-01
6.76608384e-01 -5.92343569e-01 5.59492767e-01 5.69138408e-01
-1.00813377e+00 -1.23666592e-01 -6.93804562e-01 2.66954631e-01
3.26258987e-02 3.48648459e-01 6.31066144e-01 8.89392138e-01
-8.17626834e-01 3.72450799e-01 -9.28611636e-01 -5.20511679e-02
4.32361007e-01 7.19967425e-01 -4.13114548e-01 -2.39403501e-01
-7.58187532e-01 6.84210539e-01 1.66389358e-03 2.31910586e-01
-7.54743934e-01 -8.65912914e-01 -9.35910821e-01 -1.06858732e-02
3.86331111e-01 -9.20100570e-01 6.40030861e-01 -4.06201363e-01
-1.60487998e+00 7.41547525e-01 -7.29816318e-01 -7.56035969e-02
4.43277836e-01 2.78050210e-02 1.04113124e-01 -1.20964445e-01
1.38263360e-01 1.83448464e-01 7.28212357e-01 -1.01242054e+00
-4.71167117e-02 -4.88273650e-01 -9.65960741e-01 4.15349081e-02
1.11809790e-01 3.49002153e-01 -6.53125405e-01 -6.36840701e-01
4.46741074e-01 -1.00789440e+00 -7.15248585e-01 -1.74569562e-01
-5.04611611e-01 1.17861107e-01 2.51522273e-01 -9.09027576e-01
1.51466203e+00 -2.11356473e+00 4.71782088e-01 8.30888093e-01
1.63765848e-02 -3.84193808e-01 7.91151449e-03 4.83249962e-01
-2.75770068e-01 1.54663753e-02 -4.62398678e-01 -4.51830417e-01
-1.54438227e-01 -7.14935660e-02 7.42885917e-02 1.08572602e+00
8.77861604e-02 3.33363354e-01 -3.36228400e-01 -6.65516615e-01
1.97370648e-01 6.80846930e-01 -9.34576392e-01 6.16339557e-02
-4.20140624e-02 5.21803021e-01 -6.42525971e-01 6.21702790e-01
9.49537933e-01 -3.66742462e-01 3.74358535e-01 -1.61877602e-01
-1.58330813e-01 -1.31549999e-01 -1.77594519e+00 1.38801432e+00
-1.17231637e-01 2.20446378e-01 2.52186149e-01 -1.10464358e+00
1.01604664e+00 3.88096869e-01 6.99715137e-01 -6.89190552e-02
8.53506923e-02 3.05611074e-01 -2.03620151e-01 -3.93123031e-01
7.87733775e-03 -2.77797610e-01 2.84582865e-03 1.02910958e-01
-2.25693211e-01 -1.41903400e-01 1.10108882e-01 1.29249021e-02
9.60791647e-01 1.66669324e-01 8.93478394e-01 -4.94415075e-01
4.79901463e-01 -9.85266715e-02 7.32111335e-01 8.16052854e-01
3.73724222e-01 4.71474051e-01 7.68706083e-01 5.46758026e-02
-7.62157857e-01 -7.56392300e-01 -8.04997027e-01 6.97584033e-01
-4.92724627e-01 -3.02255839e-01 -3.62992257e-01 -3.77539694e-02
2.60914385e-01 7.41783261e-01 -6.22106254e-01 1.88641027e-01
-3.86642158e-01 -1.46278560e+00 2.04665750e-01 8.60594958e-02
-1.77639499e-01 -2.85402566e-01 -1.74224138e-01 4.45758313e-01
2.11044088e-01 -6.98395073e-01 8.83225072e-03 4.70226705e-01
-1.03431404e+00 -8.28458965e-01 -7.24862397e-01 -3.42110753e-01
8.51869702e-01 -2.50473946e-01 7.48056233e-01 -2.70167798e-01
-4.97065544e-01 2.88475513e-01 -3.36337425e-02 -1.64299473e-01
-6.14434406e-02 -3.67840052e-01 3.02192837e-01 1.36065245e-01
4.32441413e-01 -5.43396592e-01 -3.84798497e-01 5.00882685e-01
-5.79888523e-01 -2.02229112e-01 7.99426556e-01 1.08689964e+00
1.16300523e+00 2.28969231e-02 5.04501522e-01 -1.02452803e+00
4.69621062e-01 -9.38322425e-01 -9.69148278e-01 9.68053937e-02
-4.57251549e-01 8.96213502e-02 1.10706940e-01 -5.77340186e-01
-8.83760035e-01 4.15280074e-01 -1.03786536e-01 -4.38143164e-01
-1.73063084e-01 1.06924164e+00 -2.37147793e-01 1.94412712e-02
5.23362279e-01 6.33577406e-02 -3.28369252e-02 -5.00441253e-01
-9.29766819e-02 7.13441491e-01 -3.45364155e-04 -6.87848687e-01
4.40401107e-01 1.81547403e-01 7.04469204e-01 -1.16470659e+00
-3.29594195e-01 -5.76102197e-01 -5.37202120e-01 5.71286269e-02
7.56439328e-01 -9.38514173e-01 -5.63007414e-01 3.17501575e-01
-6.81902885e-01 -2.41516754e-01 1.33223131e-01 1.04461479e+00
-5.63278854e-01 3.18127811e-01 -7.51268864e-02 -9.01264250e-01
7.61906058e-02 -1.34275901e+00 1.02323592e+00 -1.15907043e-01
-3.65830928e-01 -9.48444009e-01 5.89017987e-01 2.13169113e-01
1.45818859e-01 2.48902872e-01 9.28524792e-01 -8.05926323e-01
-3.56765419e-01 -4.10404563e-01 -3.86381336e-02 -1.42632738e-01
-1.03153400e-01 1.62706316e-01 -6.37185276e-01 -4.44650650e-01
5.14071546e-02 -4.43417057e-02 5.99659026e-01 1.34321916e+00
1.31397104e+00 -2.76331693e-01 -6.86574399e-01 8.00057709e-01
1.69787776e+00 -3.94279957e-02 5.80070257e-01 -9.96370986e-02
3.73424411e-01 5.30773282e-01 7.90079355e-01 1.01204336e+00
7.86297992e-02 9.52061594e-01 1.18330801e-02 -1.38410330e-01
5.24529874e-01 1.02417290e-01 5.83494194e-02 5.33759773e-01
1.62110656e-01 -9.71334502e-02 -8.92312169e-01 2.78503686e-01
-1.78859329e+00 -7.83560693e-01 -5.26048601e-01 2.57770562e+00
1.05699921e+00 -2.91045040e-01 1.01770468e-01 -1.38819501e-01
7.65430212e-01 -3.65721792e-01 -5.30822158e-01 -2.15710208e-01
-1.64843291e-01 1.66066974e-01 6.05750978e-01 7.19365060e-01
-8.60737562e-01 4.35594022e-01 6.64831972e+00 1.01886630e+00
-6.97515368e-01 4.41936478e-02 6.89409494e-01 -1.35722429e-01
-1.40828893e-01 2.07467303e-01 -1.18174005e+00 4.26079750e-01
1.05715597e+00 6.39880076e-02 1.24550313e-01 6.01414740e-01
7.72680581e-01 -7.08543897e-01 -1.00608265e+00 9.48122919e-01
-8.42720792e-02 -1.10636830e+00 -4.30659503e-01 3.87088090e-01
6.22265220e-01 -1.75265789e-01 -1.16648167e-01 -2.85192002e-02
2.25796551e-01 -1.00440514e+00 1.02550931e-01 8.56305897e-01
8.75070214e-01 -6.25069261e-01 6.59684956e-01 3.76620740e-01
-7.37664104e-01 -6.98080799e-03 -5.37891626e-01 2.12104142e-01
2.18037531e-01 1.25700057e+00 -9.25177693e-01 2.42432356e-01
5.69595039e-01 5.17592251e-01 -1.61057517e-01 1.35216844e+00
2.09669292e-01 7.86930859e-01 -7.13756025e-01 -1.33640170e-01
-2.63310999e-01 -6.43076360e-01 7.11517870e-01 1.38390207e+00
4.80977505e-01 2.26267800e-01 9.12875086e-02 7.76220620e-01
4.55038965e-01 4.71496254e-01 -3.41464728e-01 1.10526271e-01
5.64989030e-01 1.01467764e+00 -5.30047238e-01 -1.87527135e-01
-5.23339748e-01 3.30696404e-01 5.46370856e-02 4.33488309e-01
-5.74125171e-01 -1.03770401e-02 2.76226759e-01 1.18202895e-01
5.47869742e-01 -1.56225607e-01 -4.57270026e-01 -8.53618979e-01
-3.36025536e-01 -1.05241621e+00 6.14223599e-01 -5.80863595e-01
-1.27821338e+00 -1.44057050e-01 4.90987420e-01 -8.02247167e-01
-4.35735643e-01 -5.28834581e-01 -2.95363992e-01 1.26161003e+00
-1.14535105e+00 -6.95613742e-01 1.62173957e-02 6.00977838e-01
1.78212270e-01 -1.58404335e-01 9.62333322e-01 3.08733106e-01
-9.72328663e-01 4.48876828e-01 6.31257296e-01 -4.65039760e-01
8.34371924e-01 -8.33303332e-01 -5.95287263e-01 5.27345002e-01
-4.81755465e-01 6.86278880e-01 1.08360779e+00 -1.05129504e+00
-1.47491026e+00 -6.82336092e-01 7.77041614e-01 3.62829044e-02
5.35585046e-01 -2.03947470e-01 -9.41234827e-01 5.76616645e-01
-3.77634555e-01 -2.58420676e-01 1.22566998e+00 3.87643933e-01
1.43064022e-01 7.30124190e-02 -1.30674148e+00 3.46912682e-01
1.34515911e-01 -1.67774372e-02 -8.82307962e-02 5.14353156e-01
1.47223547e-01 -1.08634435e-01 -1.16890228e+00 4.71344471e-01
5.18336058e-01 -6.04765475e-01 9.91804957e-01 -5.18340409e-01
1.23983636e-01 -4.31465834e-01 -4.79526907e-01 -9.83836472e-01
-4.49706286e-01 -6.64571464e-01 1.28413811e-01 1.06493902e+00
5.27009368e-01 -5.94528258e-01 6.44180775e-01 7.70846307e-01
1.77713037e-01 -7.63515472e-01 -1.18115413e+00 -3.06786478e-01
-5.79157583e-02 -3.96287262e-01 1.35018930e-01 9.15099680e-01
-1.53252333e-01 4.40777205e-02 -5.25617123e-01 3.85403156e-01
1.04856241e+00 -2.75341868e-01 8.72545421e-01 -1.34196782e+00
-8.45359504e-01 -7.64366910e-02 -3.44514012e-01 -6.35710120e-01
1.99286863e-01 -5.84161878e-01 -1.78791471e-02 -1.15036500e+00
7.14146078e-01 -5.08204401e-01 3.08201425e-02 3.56742114e-01
-2.27769092e-01 -1.79079533e-01 -5.18891573e-01 1.10303864e-01
-8.26910287e-02 3.48997027e-01 7.85558105e-01 2.81849384e-01
-5.07812560e-01 3.52027535e-01 -4.15554196e-01 6.09763741e-01
6.22966170e-01 -1.00117922e+00 -1.12795576e-01 2.21232593e-01
5.10595143e-01 6.68862641e-01 3.34309101e-01 -3.83578628e-01
3.45511079e-01 -3.81639510e-01 4.85761523e-01 -8.01912010e-01
5.44071138e-01 -7.15830505e-01 7.35679388e-01 4.43229824e-01
-7.37161264e-02 -1.03981271e-01 3.04621726e-01 6.36431634e-01
-5.39405979e-02 -5.05904078e-01 9.05333221e-01 5.71761988e-02
4.86217886e-02 8.90472978e-02 -5.50041497e-01 -3.47125739e-01
9.75097716e-01 -2.23315105e-01 1.81120977e-01 -3.29037130e-01
-1.05108476e+00 1.31152824e-01 4.88872267e-02 -4.93323624e-01
6.48716986e-01 -1.05048752e+00 -1.25482440e+00 3.00318390e-01
-2.71470249e-01 -1.31965816e-01 3.28536928e-01 1.39621663e+00
-3.16314906e-01 3.75331312e-01 8.66751447e-02 -7.88010120e-01
-1.51370096e+00 3.83673340e-01 2.46319473e-01 -3.93457830e-01
-1.94079980e-01 7.03319609e-01 1.29533559e-02 -4.09714878e-01
3.25720645e-02 1.90374106e-01 -1.95853576e-01 1.83993787e-01
2.73758948e-01 7.42860377e-01 -1.68060362e-01 -4.80734438e-01
-4.39766347e-01 7.68694043e-01 8.52775052e-02 -3.62053156e-01
1.76642787e+00 -2.02578634e-01 -5.43608367e-01 4.51470464e-01
1.20603240e+00 2.64085948e-01 -9.68650997e-01 -7.54149705e-02
-8.67362097e-02 -5.72416008e-01 5.23171425e-01 -5.90221822e-01
-5.80542445e-01 3.96679848e-01 4.83310759e-01 -4.16442603e-01
9.79960442e-01 -1.83899373e-01 -2.99396366e-01 3.25042665e-01
1.04358248e-01 -9.53953803e-01 -4.80356604e-01 1.22927837e-01
8.83326232e-01 -1.21467173e+00 7.42885947e-01 -6.14311278e-01
-3.93837571e-01 8.13630879e-01 1.26715824e-01 -8.32154080e-02
1.07818711e+00 4.86375302e-01 -4.69958156e-01 -8.93105417e-02
-8.07433128e-01 3.43959957e-01 4.29748178e-01 3.94756198e-01
4.63981330e-01 -4.93373349e-02 -7.06927180e-01 6.71144128e-01
2.58082747e-01 1.59294426e-01 1.84375450e-01 6.82624400e-01
-2.63643235e-01 -1.09860086e+00 -8.16032946e-01 7.54675090e-01
-4.67655152e-01 -2.47282222e-01 -4.30323407e-02 7.18262494e-01
-4.69826758e-01 8.45829546e-01 -7.74844643e-03 2.14602068e-01
-9.73137189e-03 4.59670760e-02 4.31932509e-01 -4.76286530e-01
-3.12953919e-01 7.56411076e-01 2.19668940e-01 -4.85451281e-01
-2.34324858e-01 -1.39376402e+00 -8.04729402e-01 -5.95801435e-02
-6.85336828e-01 1.92660108e-01 9.49515581e-01 8.34347188e-01
4.03217852e-01 3.36031824e-01 6.29283905e-01 -9.60806966e-01
-6.14250720e-01 -8.81950438e-01 -8.18762362e-01 -2.39232793e-01
1.16443403e-01 -7.60717154e-01 -6.55364811e-01 -1.79574728e-01] | [7.40571403503418, 4.701920986175537] |
7cf6896b-10f1-4f2d-956a-2164fece3b6b | irfl-image-recognition-of-figurative-language | 2303.15445 | null | https://arxiv.org/abs/2303.15445v1 | https://arxiv.org/pdf/2303.15445v1.pdf | IRFL: Image Recognition of Figurative Language | Figures of speech such as metaphors, similes, and idioms allow language to be expressive, invoke emotion, and communicate abstract ideas that might otherwise be difficult to visualize. These figurative forms are often conveyed through multiple modes, such as text and images, and frequently appear in advertising, news, social media, etc. Understanding multimodal figurative language is an essential component of human communication, and it plays a significant role in our daily interactions. While humans can intuitively understand multimodal figurative language, this poses a challenging task for machines that requires the cognitive ability to map between domains, abstraction, commonsense, and profound language and cultural knowledge. In this work, we propose the Image Recognition of Figurative Language dataset to examine vision and language models' understanding of figurative language. We leverage human annotation and an automatic pipeline we created to generate a multimodal dataset and introduce two novel tasks as a benchmark for multimodal figurative understanding. We experiment with several baseline models and find that all perform substantially worse than humans. We hope our dataset and benchmark will drive the development of models that will better understand figurative language. | ['Dafna Shahaf', 'Yonatan Bitton', 'Ron Yosef'] | 2023-03-27 | null | null | null | null | ['visual-reasoning', 'visual-reasoning'] | ['computer-vision', 'reasoning'] | [-2.92124301e-02 4.38568145e-02 1.44763365e-01 -4.64949906e-01
-2.31970653e-01 -1.06883073e+00 1.40100312e+00 2.89550304e-01
3.47193629e-02 3.61151844e-01 7.03157365e-01 -3.86371255e-01
2.51706511e-01 -5.89404285e-01 -6.10567868e-01 -9.13104191e-02
1.20857112e-01 4.59958822e-01 -1.50130525e-01 -8.22716475e-01
2.14016885e-01 3.65122795e-01 -1.30619693e+00 1.05169022e+00
4.57864195e-01 8.33424866e-01 1.70006797e-01 8.30509603e-01
-6.17560804e-01 1.27831769e+00 -8.06918383e-01 -6.54517531e-01
-3.24157238e-01 -4.62508678e-01 -9.26692009e-01 1.57439858e-01
4.75498229e-01 -1.72761902e-01 -1.15680844e-01 7.77675211e-01
-6.66709570e-03 -5.60776927e-02 7.59738564e-01 -1.15477598e+00
-1.16729259e+00 7.16751754e-01 -4.75514829e-01 -2.71704197e-01
8.62872303e-01 1.48204923e-01 8.47652912e-01 -5.91472268e-01
9.18359518e-01 2.16249347e+00 2.89951622e-01 5.95881999e-01
-1.31613481e+00 -4.71325010e-01 9.43249911e-02 -1.99616507e-01
-8.31314266e-01 -3.46607506e-01 8.37226689e-01 -5.16478240e-01
6.28054142e-01 5.73381245e-01 1.01477671e+00 1.22486150e+00
4.26255576e-02 1.15760291e+00 1.18368375e+00 -3.94945472e-01
-1.19133353e-01 4.44023132e-01 -3.60345021e-02 9.63579237e-01
-5.04977703e-01 -3.98165792e-01 -6.59017563e-01 -4.63779457e-02
7.40580559e-01 1.43808827e-01 -4.82278794e-01 -1.24615513e-01
-1.68984473e+00 9.11136568e-01 7.71329105e-01 4.07961756e-01
-1.29635766e-01 5.20454168e-01 5.57782233e-01 4.32789028e-01
3.17530453e-01 7.86972702e-01 9.13247913e-02 -4.53137487e-01
-2.97046423e-01 5.48475049e-02 7.87786305e-01 7.09989488e-01
4.26777929e-01 -4.14087176e-01 1.94186002e-01 1.11934018e+00
2.52950072e-01 7.88408577e-01 2.94954360e-01 -9.46761906e-01
2.14270115e-01 9.85700548e-01 1.27794862e-01 -1.39078665e+00
-3.96198213e-01 6.77720085e-02 -6.69040203e-01 -9.09087211e-02
3.67899030e-01 7.28455633e-02 -7.22269475e-01 1.68750560e+00
-2.13875007e-02 -4.67831194e-01 6.73272163e-02 1.20954537e+00
1.38126695e+00 1.10917342e+00 1.91188514e-01 2.38318697e-01
1.57567585e+00 -6.49447918e-01 -7.30189204e-01 -2.37774521e-01
7.74910808e-01 -1.04305589e+00 1.60263538e+00 1.74231946e-01
-1.14392865e+00 -3.15567881e-01 -8.70568573e-01 -4.80165392e-01
-8.01364243e-01 7.30920061e-02 8.19216967e-01 1.68383092e-01
-9.42195833e-01 2.37249359e-02 -3.99040103e-01 -8.69471133e-01
3.56691182e-01 -1.89229831e-01 -5.30513585e-01 1.19017556e-01
-9.59042549e-01 1.02129138e+00 2.46082276e-01 -1.96975004e-02
-4.25708473e-01 -3.79924804e-01 -1.02841926e+00 -1.93511337e-01
2.51239181e-01 -8.01907480e-01 1.30113268e+00 -1.40446472e+00
-1.17451680e+00 1.29322720e+00 -1.08551458e-01 -2.95262605e-01
2.30570868e-01 -1.73613176e-01 -3.37750852e-01 5.13051629e-01
-1.44688800e-01 1.33226609e+00 7.90919960e-01 -1.70487726e+00
1.20176934e-02 -1.81899279e-01 6.24356210e-01 1.61722023e-02
-2.09495306e-01 2.03022793e-01 -3.37665349e-01 -6.84159279e-01
1.29752398e-01 -9.06878471e-01 3.36527348e-01 3.40440303e-01
-5.98697305e-01 -1.17555007e-01 1.12238586e+00 -6.25893354e-01
9.43648636e-01 -2.40994191e+00 4.85002667e-01 5.17303729e-03
5.99816382e-01 -2.73211032e-01 -1.02015369e-01 8.12213600e-01
3.07950497e-01 4.53195453e-01 -1.97646320e-01 -2.65852660e-01
4.01705623e-01 2.29863882e-01 -5.21396160e-01 -1.19680472e-01
1.43100575e-01 1.34043217e+00 -1.07190788e+00 -5.63674390e-01
1.85492381e-01 5.07411480e-01 -3.68920326e-01 2.18308538e-01
-5.64143717e-01 3.80380660e-01 -4.19433028e-01 6.98120654e-01
2.64651656e-01 -6.63457692e-01 2.56026357e-01 -3.44649553e-01
-5.75989597e-02 5.66805936e-02 -3.89082134e-01 1.70872843e+00
-7.44686067e-01 1.27978849e+00 9.10370871e-02 -5.55900633e-01
7.94796944e-01 1.14159890e-01 -1.26006559e-01 -6.64145231e-01
2.53961742e-01 -1.28153801e-01 6.49329200e-02 -7.01641560e-01
5.23767412e-01 -4.45975840e-01 -5.57666302e-01 8.15598488e-01
-5.34196496e-01 -8.90210927e-01 -1.62711069e-01 8.38116586e-01
6.00779235e-01 -1.01365171e-01 1.78247109e-01 -1.19615182e-01
2.85213709e-01 1.89619347e-01 -3.36235851e-01 4.08230960e-01
5.74780628e-03 4.67240721e-01 7.62373924e-01 -8.20263088e-01
-6.98481023e-01 -1.27300334e+00 9.29511860e-02 1.20556855e+00
3.31286728e-01 -4.49836403e-01 -5.43008208e-01 -3.26267570e-01
-5.19768596e-02 8.47706139e-01 -7.10620403e-01 3.78044434e-02
-3.11691582e-01 -3.12801003e-01 4.60024744e-01 2.08408520e-01
6.04800045e-01 -1.13990903e+00 -9.68574643e-01 -1.69363305e-01
-2.56829679e-01 -1.17114699e+00 -3.38638097e-01 -3.41051519e-01
-3.52865875e-01 -9.13176954e-01 -5.23827016e-01 -8.56563509e-01
5.91391683e-01 4.20944214e-01 1.61474216e+00 3.60534847e-01
-2.79276401e-01 8.99081171e-01 -5.18064857e-01 -7.12833703e-01
-8.73509109e-01 -4.22958165e-01 -3.84959280e-01 -3.28051066e-03
2.11766317e-01 -2.79284805e-01 -4.23478276e-01 1.52823672e-01
-1.22580278e+00 8.28091621e-01 2.66371250e-01 8.07616711e-01
-8.86415914e-02 -5.04056871e-01 2.93612272e-01 -7.70509720e-01
9.76539731e-01 -4.69967157e-01 8.88287723e-02 3.74630243e-01
5.10037065e-01 4.30035368e-02 5.66432416e-01 -6.37418687e-01
-9.42782640e-01 -2.73461699e-01 4.10270870e-01 -9.31985974e-02
-2.11197689e-01 4.95779395e-01 3.06777880e-02 6.55915961e-02
6.10019743e-01 1.12103567e-01 3.18984330e-01 -2.63296187e-01
1.02160215e+00 7.35476971e-01 4.08788949e-01 -6.86213374e-01
5.88245928e-01 8.18681240e-01 -1.42048284e-01 -1.22242951e+00
-5.43860376e-01 -1.46076828e-01 -4.78302389e-01 -5.34007549e-01
1.05789351e+00 -7.84839690e-01 -9.75959122e-01 -3.36254425e-02
-1.50251997e+00 -2.11597621e-01 2.20831335e-01 -5.96754253e-02
-4.12093520e-01 3.55469257e-01 -5.39466619e-01 -6.35319293e-01
-6.61506429e-02 -1.07860315e+00 1.29755902e+00 8.72501060e-02
-8.41039121e-01 -1.29101062e+00 -3.55870932e-01 5.47508776e-01
5.20423710e-01 6.61731362e-01 1.45594645e+00 -2.93028057e-01
-6.78788304e-01 3.32967080e-02 -6.39161408e-01 -1.89500228e-01
1.28405049e-01 9.24994797e-03 -7.77804077e-01 1.44217774e-01
-4.84182715e-01 -1.09625614e+00 7.46953070e-01 -1.14086591e-01
1.00967538e+00 -4.94831443e-01 -2.50876158e-01 5.75917214e-02
9.12816703e-01 2.78452098e-01 5.18400729e-01 1.90821514e-01
6.64329588e-01 1.11855829e+00 2.82861978e-01 2.02032641e-01
6.93555236e-01 5.61951458e-01 3.97609413e-01 -3.55664968e-01
1.17660865e-01 -2.60398984e-01 2.35303208e-01 8.23080420e-01
9.52679142e-02 -2.19814017e-01 -1.11368525e+00 2.96606123e-01
-1.80553877e+00 -1.09596324e+00 -1.00183576e-01 1.37209702e+00
8.16888213e-01 1.04396837e-02 1.23218007e-01 -3.04907352e-01
2.07448274e-01 2.83237100e-01 -1.69228494e-01 -1.00541615e+00
-2.85885334e-01 -3.35076511e-01 -4.03592885e-01 4.04558122e-01
-9.27731693e-01 1.13161254e+00 6.01388597e+00 3.24784845e-01
-1.55194950e+00 -2.31653705e-01 7.71025002e-01 -2.67494116e-02
-7.00299740e-01 -2.32521474e-01 9.08029824e-02 9.90552157e-02
5.32257855e-01 1.78956211e-01 6.27859473e-01 3.26142818e-01
3.14454794e-01 -2.13161871e-01 -1.47248495e+00 1.39884353e+00
4.04192209e-01 -1.56915045e+00 6.57033503e-01 -3.05276752e-01
2.93733418e-01 -4.50034022e-01 2.89744437e-01 1.07095033e-01
-5.84528781e-02 -1.31703174e+00 7.87693381e-01 6.90918565e-01
5.59017003e-01 -3.55546892e-01 2.97981232e-01 1.81146219e-01
-8.76627803e-01 1.22726776e-01 4.69521880e-02 -3.37468684e-01
2.25165442e-01 9.11109522e-03 -8.21138799e-01 3.12252603e-02
4.90106940e-01 7.97280371e-01 -7.42896557e-01 4.05713797e-01
-5.65597862e-02 1.20409407e-01 -1.16079874e-01 -7.87646234e-01
4.32030052e-01 -1.09846726e-01 6.55141413e-01 1.47722614e+00
-3.56378146e-02 1.87782258e-01 1.89158246e-01 1.02538788e+00
-1.95373625e-01 2.11963832e-01 -1.00736034e+00 -8.06754053e-01
3.10981780e-01 1.06631327e+00 -8.96925926e-01 -4.34249610e-01
-5.12359619e-01 1.06757128e+00 2.01981679e-01 3.59220982e-01
-7.80096412e-01 -8.06504861e-02 4.91130292e-01 1.77217484e-01
-6.19051576e-01 -4.58092213e-01 -1.74722582e-01 -1.20388114e+00
-3.39107126e-01 -1.14942002e+00 5.42299189e-02 -1.48936224e+00
-1.27271128e+00 5.80399275e-01 1.59950346e-01 -1.00677204e+00
-3.12268496e-01 -7.88434148e-01 -6.21367753e-01 4.48084325e-01
-7.66762912e-01 -1.64448583e+00 -4.36280727e-01 2.45427340e-01
7.24460840e-01 -9.63186324e-02 8.86484206e-01 -2.51638174e-01
-1.02388069e-01 2.45913699e-01 -3.70677263e-01 1.16691351e-01
7.78321266e-01 -1.18080580e+00 2.22464025e-01 1.30474523e-01
5.57037115e-01 9.14627731e-01 7.86302507e-01 -2.36044854e-01
-1.88626456e+00 -4.18489695e-01 4.83467668e-01 -8.03767741e-01
1.03195584e+00 -6.72093093e-01 -7.20297396e-01 5.90199888e-01
7.52866924e-01 -6.52586639e-01 8.79055381e-01 1.56955928e-01
-6.89835131e-01 3.33323061e-01 -8.98641527e-01 1.18751216e+00
9.03236628e-01 -8.02903056e-01 -9.67172444e-01 4.59049851e-01
9.55170751e-01 -1.80921838e-01 -5.51305830e-01 -6.47067428e-02
8.98952186e-01 -8.56849551e-01 1.04307437e+00 -8.36370826e-01
8.56639683e-01 -1.47649646e-01 -2.71242976e-01 -1.30546153e+00
2.07203880e-01 -7.72429764e-01 2.53905326e-01 1.12037003e+00
6.75023735e-01 -5.23496389e-01 1.06277168e-01 7.34455287e-01
3.41572613e-02 -4.08415347e-01 -4.59691495e-01 -3.51747662e-01
-1.35499090e-01 -4.68597442e-01 4.43933517e-01 1.24757254e+00
4.53925103e-01 7.79402614e-01 -1.74713016e-01 -2.83066273e-01
1.13139011e-01 4.85905737e-01 1.12050831e+00 -1.09257936e+00
-1.73932984e-01 -7.32323825e-01 -4.01153803e-01 -1.16822934e+00
2.76729524e-01 -9.81764317e-01 -2.41770372e-01 -1.68430841e+00
4.41204697e-01 -1.57132357e-01 3.36851358e-01 5.31132281e-01
3.09976459e-01 4.45952535e-01 7.04408169e-01 2.30868861e-01
-8.13266277e-01 4.63858306e-01 1.83180678e+00 -5.73141694e-01
-2.17649147e-01 -9.27284837e-01 -8.60139728e-01 1.00034583e+00
6.17957354e-01 2.38854766e-01 -4.24799353e-01 -5.92760980e-01
6.22761607e-01 -1.01217870e-02 7.73482740e-01 -3.99787366e-01
-1.13612093e-01 -2.83906907e-01 3.86259764e-01 -4.32821095e-01
8.01297784e-01 -8.04177165e-01 -7.35044405e-02 3.72164905e-01
-4.69647229e-01 2.57070273e-01 3.94784093e-01 3.94584715e-01
-3.24489981e-01 3.98647666e-01 5.73414147e-01 -2.54869550e-01
-7.37686694e-01 -5.43704748e-01 -5.05079269e-01 7.95087740e-02
7.05164194e-01 -9.45888981e-02 -6.68764710e-01 -1.13986313e+00
-5.84832966e-01 2.72754431e-01 8.31878901e-01 8.33576620e-01
9.03838336e-01 -1.27599955e+00 -5.26846409e-01 -2.65185773e-01
4.55044627e-01 -3.27379644e-01 -8.58000368e-02 5.77472210e-01
-7.93797314e-01 3.07831794e-01 -1.88089103e-01 -7.79013753e-01
-1.26399398e+00 3.62092406e-01 1.35660112e-01 4.98110354e-01
-7.08764553e-01 8.49633634e-01 8.55644703e-01 -3.92223924e-01
-5.62797440e-03 -6.23279810e-01 -2.54083008e-01 1.04647875e-01
6.17379487e-01 -1.59203663e-01 -9.86143708e-01 -9.65490341e-01
-2.96536386e-01 6.06870890e-01 1.22202642e-01 -2.12200001e-01
9.72368538e-01 -2.32417956e-01 -6.00192130e-01 9.03203547e-01
1.32432330e+00 -4.58127819e-02 -5.07779539e-01 2.17370898e-01
-7.14405731e-04 -4.01430398e-01 -3.58529270e-01 -1.20438898e+00
-3.82087439e-01 1.23541045e+00 1.44016430e-01 7.55157113e-01
1.03591835e+00 6.68300390e-01 1.00760806e+00 6.45459294e-01
1.19661123e-01 -6.66263103e-01 6.77729428e-01 5.38747966e-01
1.67250133e+00 -1.47867239e+00 -1.58144265e-01 -5.03305435e-01
-8.09069157e-01 1.34880447e+00 4.63697582e-01 4.23076451e-01
2.84493148e-01 -3.42712775e-02 4.40297961e-01 -6.57095075e-01
-9.54774022e-01 -4.85450923e-02 5.79086542e-01 2.92980433e-01
8.38350713e-01 4.24406111e-01 -7.21644610e-02 3.78930867e-01
-5.60290754e-01 -4.36485559e-01 5.28034508e-01 8.04415226e-01
-4.18145955e-01 -6.03715599e-01 -5.97884059e-01 1.29309997e-01
-1.19360477e-01 -7.72755221e-02 -1.41634357e+00 9.78877962e-01
-4.67530489e-02 1.04121995e+00 2.18439341e-01 -2.79017717e-01
1.09253392e-01 1.03853747e-01 6.02932990e-01 -3.76614898e-01
-5.01665473e-01 -1.03333093e-01 5.60905516e-01 -3.85989040e-01
-6.00260854e-01 -2.17681691e-01 -1.61977530e+00 -5.06573737e-01
4.48508352e-01 -6.43222257e-02 7.54567564e-01 1.03520429e+00
2.59214759e-01 3.10280919e-01 1.90958366e-01 -8.10883105e-01
2.72293001e-01 -8.97857726e-01 -8.67853761e-02 8.32315385e-01
3.71737629e-01 -5.42674363e-01 -2.34035149e-01 2.92773724e-01] | [10.877326011657715, 1.537501335144043] |
4e6c6853-6db7-4653-bf0c-939df6532f51 | attentionxml-extreme-multi-label-text | 1811.01727 | null | https://arxiv.org/abs/1811.01727v3 | https://arxiv.org/pdf/1811.01727v3.pdf | AttentionXML: Label Tree-based Attention-Aware Deep Model for High-Performance Extreme Multi-Label Text Classification | Extreme multi-label text classification (XMTC) is an important problem in the era of big data, for tagging a given text with the most relevant multiple labels from an extremely large-scale label set. XMTC can be found in many applications, such as item categorization, web page tagging, and news annotation. Traditionally most methods used bag-of-words (BOW) as inputs, ignoring word context as well as deep semantic information. Recent attempts to overcome the problems of BOW by deep learning still suffer from 1) failing to capture the important subtext for each label and 2) lack of scalability against the huge number of labels. We propose a new label tree-based deep learning model for XMTC, called AttentionXML, with two unique features: 1) a multi-label attention mechanism with raw text as input, which allows to capture the most relevant part of text to each label; and 2) a shallow and wide probabilistic label tree (PLT), which allows to handle millions of labels, especially for "tail labels". We empirically compared the performance of AttentionXML with those of eight state-of-the-art methods over six benchmark datasets, including Amazon-3M with around 3 million labels. AttentionXML outperformed all competing methods under all experimental settings. Experimental results also show that AttentionXML achieved the best performance against tail labels among label tree-based methods. The code and datasets are available at http://github.com/yourh/AttentionXML . | ['Hiroshi Mamitsuka', 'Suyang Dai', 'Shanfeng Zhu', 'Ziye Wang', 'Zihan Zhang', 'Ronghui You'] | 2018-11-01 | attentionxml-label-tree-based-attention-aware | http://papers.nips.cc/paper/8817-attentionxml-label-tree-based-attention-aware-deep-model-for-high-performance-extreme-multi-label-text-classification | http://papers.nips.cc/paper/8817-attentionxml-label-tree-based-attention-aware-deep-model-for-high-performance-extreme-multi-label-text-classification.pdf | neurips-2019-12 | ['product-categorization', 'web-page-tagging', 'news-annotation'] | ['miscellaneous', 'natural-language-processing', 'natural-language-processing'] | [ 1.44264917e-03 -2.04292923e-01 -3.20894450e-01 -5.32139778e-01
-1.16163445e+00 -4.94717300e-01 3.53514671e-01 3.85498881e-01
-5.03353536e-01 5.26680648e-01 3.19587469e-01 -2.15477139e-01
4.16765139e-02 -5.63714921e-01 -4.67567891e-01 -7.39822447e-01
4.25876766e-01 9.33366537e-01 1.64393872e-01 1.49285393e-02
1.83757111e-01 -1.18969657e-01 -1.54933763e+00 6.18353426e-01
4.59192008e-01 1.51025140e+00 1.04619980e-01 3.62010986e-01
-7.55312622e-01 1.04805708e+00 -4.63043720e-01 -5.51499307e-01
-2.46480126e-02 -1.10534497e-01 -1.11645055e+00 -2.14433894e-01
5.73891878e-01 -8.38923305e-02 6.43837154e-02 1.00396347e+00
5.47158062e-01 2.11304262e-01 7.92614400e-01 -1.33582222e+00
-7.53569543e-01 8.57268512e-01 -9.61334825e-01 -4.77393307e-02
9.90518704e-02 -3.27931851e-01 1.50107086e+00 -9.56580222e-01
2.49654949e-01 1.56351376e+00 9.52898502e-01 5.55287123e-01
-1.03985870e+00 -1.02591801e+00 1.68294206e-01 1.35836706e-01
-1.31458580e+00 -5.88223524e-03 4.38962221e-01 -5.32633901e-01
8.87285531e-01 1.69346541e-01 8.38669669e-03 1.22113299e+00
1.88079059e-01 8.70652914e-01 1.19602001e+00 -6.53210998e-01
1.79667100e-01 1.77639484e-01 9.70260799e-01 6.30270422e-01
2.04561725e-02 -3.82709712e-01 -2.57567465e-01 -5.61166167e-01
1.97430402e-02 2.56198972e-01 1.54335111e-01 2.69230474e-02
-8.58672917e-01 1.20790315e+00 5.18580496e-01 4.46981609e-01
-2.22835734e-01 5.81464529e-01 7.42848098e-01 8.81242752e-02
9.42285299e-01 3.09850842e-01 -8.30805600e-01 1.65779382e-01
-5.70513070e-01 -2.34624464e-02 5.15292704e-01 1.07500291e+00
8.37854087e-01 -4.61384922e-01 -5.72232008e-01 1.21381187e+00
3.24202091e-01 2.86221832e-01 8.74414802e-01 -7.62231767e-01
5.10219693e-01 7.89399564e-01 -3.85160185e-02 -7.52704918e-01
-7.45440662e-01 -5.12686670e-01 -9.20510352e-01 -2.96372354e-01
1.88149244e-01 -7.67293423e-02 -1.06909597e+00 1.66010034e+00
3.82155836e-01 2.17986410e-03 -2.74464279e-01 6.16894186e-01
1.13769078e+00 8.63855839e-01 6.02549374e-01 4.12578955e-02
1.84908223e+00 -1.36627197e+00 -7.88004994e-01 -4.91236508e-01
1.19722056e+00 -6.05480909e-01 1.21276796e+00 1.35741219e-01
-5.58089793e-01 -6.03736579e-01 -4.88931328e-01 -4.27320510e-01
-7.36658275e-01 7.20166415e-02 6.90853000e-01 4.66716826e-01
-9.01230276e-01 1.63870081e-01 -2.19261274e-01 -3.98971856e-01
4.61929381e-01 1.96248263e-01 -3.33647639e-01 -3.43206942e-01
-1.42323864e+00 7.37835228e-01 4.64468718e-01 -2.97996849e-01
-5.60849607e-01 -5.75236440e-01 -7.51217961e-01 4.55904961e-01
4.25205678e-01 -4.14439350e-01 1.61197639e+00 -9.14896786e-01
-9.02155876e-01 1.01364195e+00 -8.59660506e-02 -1.42305359e-01
1.02973752e-01 -3.82987887e-01 -1.68024093e-01 -5.01846522e-02
5.60475945e-01 9.22857702e-01 6.51220322e-01 -1.23725629e+00
-9.47464108e-01 -3.98095131e-01 -1.79489955e-01 8.30204561e-02
-6.87064528e-01 3.90463650e-01 -4.09836948e-01 -6.47356033e-01
-7.63907507e-02 -9.91724968e-01 -1.42882705e-01 -4.73863870e-01
-4.24163878e-01 -9.34674263e-01 7.62611091e-01 -5.10723114e-01
1.37346387e+00 -2.16165352e+00 -1.19645797e-01 -2.52772272e-01
3.71964604e-01 1.53807178e-01 -1.96251765e-01 4.38784003e-01
-8.15829411e-02 3.24391484e-01 1.80515140e-01 -7.99146414e-01
2.93676198e-01 -3.04871667e-02 -4.44343567e-01 2.13129222e-01
-2.66139090e-01 1.18999887e+00 -9.51730311e-01 -8.47297966e-01
-9.16112661e-02 3.29565406e-01 -1.77421525e-01 2.03563735e-01
-4.61331159e-01 1.75235033e-01 -5.35013020e-01 6.66339099e-01
4.46489602e-01 -8.36535692e-01 1.63538717e-02 -6.41737506e-02
3.10897857e-01 8.45982060e-02 -8.61072958e-01 1.46809590e+00
-6.81990504e-01 3.26473325e-01 -2.17284113e-01 -6.62304878e-01
8.28565300e-01 5.83218575e-01 5.62457621e-01 -6.09842360e-01
6.28290892e-01 1.57485053e-01 -5.17587066e-01 -2.80358762e-01
4.31994498e-01 -2.94409603e-01 -6.62645698e-01 7.95627654e-01
1.91935286e-01 2.98904657e-01 -5.43381495e-04 3.22603524e-01
1.03208208e+00 -3.06901753e-01 3.16102117e-01 -2.29502648e-01
2.37554431e-01 -1.76666304e-01 5.60059249e-01 8.60771894e-01
-1.22037441e-01 4.16908354e-01 4.00102168e-01 -7.24348247e-01
-7.98142552e-01 -2.95703322e-01 -1.82681859e-01 1.97684073e+00
-4.09223437e-02 -5.13647497e-01 -5.25761902e-01 -1.09875727e+00
2.16141373e-01 8.30075383e-01 -8.63143325e-01 -1.28706023e-01
-3.01829398e-01 -7.49038815e-01 4.71223265e-01 7.63052881e-01
2.98600942e-01 -1.32515335e+00 -2.68793911e-01 2.69326925e-01
-4.87499893e-01 -1.09795511e+00 -6.44372940e-01 6.47310913e-01
-4.73374754e-01 -7.41135299e-01 -6.33253336e-01 -8.85405123e-01
3.42453241e-01 4.21238750e-01 1.23951042e+00 2.33250648e-01
-9.81411412e-02 3.77107225e-02 -7.05034912e-01 -4.15372133e-01
-3.11269581e-01 3.89902771e-01 -9.40808654e-02 1.03328161e-01
7.79514253e-01 -1.10096052e-01 -2.89174587e-01 4.61383373e-01
-7.96901584e-01 1.65301949e-01 4.67089951e-01 1.03259718e+00
5.18425286e-01 1.93663687e-01 7.71939158e-01 -1.29479706e+00
5.08051872e-01 -8.02791834e-01 -2.79020250e-01 3.24297190e-01
-6.55178010e-01 -1.75495204e-02 7.26397634e-01 -5.38543403e-01
-8.08614850e-01 -9.56187248e-02 -4.11397994e-01 -1.66726366e-01
-2.70456672e-01 4.24394488e-01 -3.95025648e-02 2.15784341e-01
4.22078073e-01 -2.08221331e-01 -4.19995487e-01 -9.55085039e-01
4.59975213e-01 1.17616940e+00 7.67844170e-02 -6.33710861e-01
2.17103641e-02 1.92109883e-01 -2.11364791e-01 -1.09583519e-01
-1.91490710e+00 -1.08447421e+00 -5.80288768e-01 -9.55326483e-02
9.45791543e-01 -8.83260965e-01 -5.88386953e-01 4.80612874e-01
-9.62926507e-01 -3.33795607e-01 -1.09184831e-01 1.47311226e-01
-3.61295372e-01 2.69013286e-01 -1.18567109e+00 -5.49266577e-01
-8.48633409e-01 -1.08252668e+00 1.51813006e+00 1.21829316e-01
-1.32324487e-01 -1.05294406e+00 6.11387119e-02 6.20501161e-01
3.17458123e-01 -7.02456310e-02 1.25769114e+00 -1.05107570e+00
1.26530468e-01 -5.40888250e-01 -6.17705941e-01 2.51693338e-01
-1.53425783e-01 -5.33941329e-01 -1.20067906e+00 -5.40112674e-01
-2.80153096e-01 -9.31256950e-01 1.13111782e+00 3.57757926e-01
1.33472490e+00 -8.83738771e-02 -7.33959138e-01 3.59874040e-01
1.44004893e+00 -2.10469887e-02 1.53473958e-01 4.32413965e-01
1.06134963e+00 5.52971005e-01 6.61686778e-01 3.26158464e-01
4.33771610e-01 8.98862183e-01 4.96419847e-01 -1.01708427e-01
-1.35885939e-01 -2.04562843e-01 -1.85316261e-02 1.03983319e+00
5.76547801e-01 -6.52045310e-01 -1.09322917e+00 4.07572269e-01
-2.00433064e+00 -5.66791892e-01 -2.66229659e-01 1.81490314e+00
8.89567196e-01 1.28272772e-01 -1.19666895e-02 3.20038721e-02
1.00453758e+00 -1.37943570e-02 -4.72129136e-01 -3.26066941e-01
9.22451764e-02 -1.23979775e-02 4.40959185e-01 2.66321987e-01
-1.49459743e+00 1.13259315e+00 5.50314379e+00 1.13595665e+00
-7.49111593e-01 7.37621307e-01 7.01745272e-01 1.38392925e-01
5.50505295e-02 -1.17384702e-01 -1.41900337e+00 6.32926524e-01
1.15502012e+00 2.49439582e-01 -1.14915513e-01 9.96563792e-01
-4.80235249e-01 1.88419670e-01 -8.45936954e-01 8.12065244e-01
1.10239930e-01 -8.95602167e-01 -3.42953168e-02 1.24998294e-01
8.04794908e-01 1.73502892e-01 -4.46713530e-02 8.72198045e-01
6.52108848e-01 -8.07304800e-01 7.50320911e-01 2.31214896e-01
1.05219233e+00 -6.63333654e-01 1.10319710e+00 5.19820094e-01
-1.22754610e+00 -5.59027016e-01 -5.92380822e-01 8.78303647e-02
7.76612759e-02 6.92815065e-01 -4.80963022e-01 2.22013697e-01
8.78758311e-01 6.42120600e-01 -7.40547419e-01 8.15838754e-01
-1.57746732e-01 8.09826434e-01 -1.76828787e-01 -2.25278139e-01
6.45635605e-01 3.70006263e-01 -5.04426956e-02 1.50129759e+00
2.67021090e-01 -8.68513063e-02 5.25447726e-01 3.10560703e-01
-4.32092458e-01 5.49568713e-01 -2.87545532e-01 2.17427686e-01
5.86875498e-01 1.50988793e+00 -8.96368027e-01 -5.81679761e-01
-5.91797531e-01 7.25257754e-01 6.88610733e-01 8.75524059e-02
-7.99666643e-01 -2.44389787e-01 3.86104256e-01 -1.92122683e-01
3.10308874e-01 3.03867996e-01 -2.31795564e-01 -8.72516215e-01
-3.18942159e-01 -6.69280887e-01 8.99491668e-01 -9.41149056e-01
-1.61706364e+00 8.00392032e-01 -8.94734785e-02 -8.85579467e-01
6.85714781e-02 -6.60159945e-01 -1.55287802e-01 7.00332224e-01
-1.30917120e+00 -1.55740047e+00 -4.14482027e-01 2.20823333e-01
8.39506447e-01 -5.03880717e-02 1.02131808e+00 4.98517215e-01
-6.05925441e-01 8.24286580e-01 5.72303474e-01 2.05537930e-01
1.04898036e+00 -1.43493259e+00 5.12201250e-01 1.44868106e-01
2.15171322e-01 1.19474970e-01 2.81610787e-01 -6.20752871e-01
-7.50893414e-01 -1.42903519e+00 1.27344310e+00 -7.90448189e-01
5.80320835e-01 -6.05179965e-01 -9.02552426e-01 9.15771961e-01
1.26788244e-01 8.87591690e-02 1.00338876e+00 5.23184717e-01
-7.72885263e-01 -7.87565112e-03 -8.11880171e-01 1.69939891e-01
6.65799797e-01 -3.02922130e-01 -4.16899800e-01 7.81280935e-01
1.13643694e+00 -2.27056239e-02 -8.26558471e-01 2.45929673e-01
5.05630016e-01 -5.65835655e-01 7.48188138e-01 -6.73875511e-01
4.05716091e-01 6.67909980e-02 -3.03130895e-01 -1.26893198e+00
-1.00878406e+00 5.11771366e-02 -1.90658048e-01 1.51118731e+00
3.23986411e-01 -5.60150266e-01 5.03651619e-01 3.82046640e-01
-1.47302955e-01 -9.15477395e-01 -7.29600370e-01 -5.93025506e-01
3.35792929e-01 -4.54376608e-01 7.49105513e-01 1.17024624e+00
-1.02858908e-01 9.23466384e-01 -6.40977561e-01 -1.75428048e-01
5.27998567e-01 3.24685305e-01 3.92049521e-01 -1.62153304e+00
-4.56439443e-02 -4.43070441e-01 -1.77689254e-01 -9.30252671e-01
5.74498534e-01 -1.22710621e+00 1.29647836e-01 -1.75763214e+00
7.23618269e-01 -7.88005710e-01 -7.44896650e-01 1.02583241e+00
-5.43929338e-01 4.58738208e-01 1.45517722e-01 3.93035680e-01
-1.11777008e+00 3.71885031e-01 9.02012825e-01 -2.99610674e-01
2.72668779e-01 -7.09302798e-02 -9.54242587e-01 6.32682204e-01
6.93516552e-01 -1.09295034e+00 -1.10251032e-01 -5.11900187e-01
4.36451316e-01 -1.63575977e-01 -8.66849422e-02 -7.88611054e-01
2.25267589e-01 1.05167992e-01 1.64225161e-01 -5.36561131e-01
1.70842901e-01 -8.87835741e-01 -2.29908340e-02 1.02291666e-01
-7.44297087e-01 2.19813026e-02 -2.27567442e-02 5.11381686e-01
-7.88110588e-03 -5.74499369e-01 8.76442134e-01 -1.64364159e-01
-8.24080527e-01 3.27022493e-01 -2.03568205e-01 2.28903636e-01
9.03065979e-01 3.80310655e-01 -5.48413396e-01 -2.39824966e-01
-4.70085442e-01 4.41471398e-01 1.70848280e-01 6.05304956e-01
1.00417182e-01 -1.53494203e+00 -7.45632887e-01 -5.44661842e-02
4.78304565e-01 -4.67642024e-02 2.43564308e-01 4.51221466e-01
-1.60644263e-01 6.25538886e-01 1.19640402e-01 -4.41503555e-01
-1.18496025e+00 1.07058692e+00 8.40906799e-02 -6.72769248e-01
-7.56024718e-01 1.12800300e+00 5.72542846e-01 -5.70156097e-01
5.62919974e-01 -1.65473327e-01 -3.64048809e-01 2.87982792e-01
7.21797228e-01 1.39130056e-01 1.16936699e-01 -8.55900168e-01
-2.59945750e-01 8.51696372e-01 -4.23889965e-01 3.38777661e-01
1.01003551e+00 -2.37048849e-01 -1.82612956e-01 6.80501938e-01
1.51468825e+00 -5.04442871e-01 -8.49974871e-01 -6.87082767e-01
2.91829854e-01 -2.31392756e-01 3.76218617e-01 -9.26709950e-01
-9.95971382e-01 1.03636003e+00 4.87784833e-01 4.08711731e-01
9.42609727e-01 2.87679493e-01 1.22170579e+00 1.87609270e-01
4.61750120e-01 -1.06429815e+00 2.66732365e-01 6.83983743e-01
5.82192242e-01 -1.45586145e+00 -2.20712632e-01 -1.56889915e-01
-7.13110864e-01 8.05325329e-01 7.42734730e-01 4.72767234e-01
8.92532527e-01 1.80417255e-01 3.50705653e-01 -3.55450481e-01
-1.06819701e+00 -2.90939838e-01 2.61440217e-01 6.58828914e-02
5.39105713e-01 1.31237581e-01 -1.68850318e-01 8.56517434e-01
2.62944818e-01 -2.39508793e-01 -5.22839688e-02 8.80943716e-01
-7.71785438e-01 -9.97517765e-01 -3.46497327e-01 7.79374480e-01
-7.87987709e-01 -2.01634109e-01 -2.73299336e-01 4.22853976e-01
3.39323312e-01 9.80494797e-01 -4.82446188e-03 -4.02347416e-01
5.73412664e-02 4.80266929e-01 -1.75812319e-01 -8.73545647e-01
-8.18823099e-01 1.53859422e-01 -6.54484928e-02 -1.72452077e-01
-6.89836740e-02 -3.93121839e-01 -1.12069082e+00 -2.15867966e-01
-7.47313261e-01 3.73364896e-01 7.08473027e-01 8.79153788e-01
3.65737140e-01 4.42728817e-01 5.46344519e-01 -6.30120516e-01
-6.50835991e-01 -1.49638402e+00 -7.41184473e-01 7.00820625e-01
3.30242962e-02 -8.98228824e-01 -3.75630528e-01 -9.92193148e-02] | [9.658628463745117, 4.485793113708496] |
d6f9dde2-9ea7-46d2-adc6-f43ac67291e8 | temporal-topic-modeling-to-assess | 1606.00411 | null | http://arxiv.org/abs/1606.00411v1 | http://arxiv.org/pdf/1606.00411v1.pdf | Temporal Topic Modeling to Assess Associations between News Trends and Infectious Disease Outbreaks | In retrospective assessments, internet news reports have been shown to
capture early reports of unknown infectious disease transmission prior to
official laboratory confirmation. In general, media interest and reporting
peaks and wanes during the course of an outbreak. In this study, we quantify
the extent to which media interest during infectious disease outbreaks is
indicative of trends of reported incidence. We introduce an approach that uses
supervised temporal topic models to transform large corpora of news articles
into temporal topic trends. The key advantages of this approach include,
applicability to a wide range of diseases, and ability to capture disease
dynamics - including seasonality, abrupt peaks and troughs. We evaluated the
method using data from multiple infectious disease outbreaks reported in the
United States of America (U.S.), China and India. We noted that temporal topic
trends extracted from disease-related news reports successfully captured the
dynamics of multiple outbreaks such as whooping cough in U.S. (2012), dengue
outbreaks in India (2013) and China (2014). Our observations also suggest that
efficient modeling of temporal topic trends using time-series regression
techniques can estimate disease case counts with increased precision before
official reports by health organizations. | ['Naren Ramakrishnan', 'Elaine O. Nsoesie', 'Saurav Ghosh', 'John S. Brownstein', 'Emily Cohn', 'Sumiko R. Mekaru', 'Prithwish Chakraborty'] | 2016-06-01 | null | null | null | null | ['time-series-regression'] | ['time-series'] | [-5.94283715e-02 -1.78734586e-01 -5.42098284e-01 1.08386530e-03
-8.09171498e-01 -5.37532032e-01 1.05728734e+00 1.02726674e+00
-1.28938720e-01 8.11257303e-01 6.60711348e-01 -3.76199841e-01
-1.76715419e-01 -8.77911389e-01 -6.24319911e-01 -5.49100101e-01
-9.36288238e-01 6.29214883e-01 4.36261520e-02 3.55927125e-02
6.74973354e-02 1.17501929e-01 -8.24578822e-01 2.30867639e-01
6.72718585e-01 4.13469523e-01 3.52639377e-01 6.84091389e-01
-4.94837403e-01 6.03889585e-01 -1.02594101e+00 2.36691028e-01
-2.41976887e-01 -8.23425129e-02 -3.27603042e-01 2.79647168e-02
-3.73955041e-01 -5.92833579e-01 -5.36291599e-02 3.70996684e-01
-8.05500224e-02 -4.97528404e-01 7.61781871e-01 -1.43663156e+00
-3.83410424e-01 4.86027747e-01 -9.91479397e-01 8.75763834e-01
4.82515395e-01 -4.13088724e-02 4.98908371e-01 -2.88492709e-01
1.24032593e+00 9.63260114e-01 1.13194835e+00 -1.10237680e-01
-1.24218798e+00 -7.61872888e-01 2.51968503e-01 -5.40492654e-01
-1.24568045e+00 4.87766415e-02 4.82785732e-01 -1.09896219e+00
1.09553516e+00 -3.75028700e-02 7.69634128e-01 1.16064405e+00
8.15806031e-01 3.00266176e-01 9.97027338e-01 2.01866731e-01
1.49697131e-02 1.58343479e-01 2.51777530e-01 1.58580720e-01
5.75035632e-01 -2.25425571e-01 -4.22529936e-01 -1.03883231e+00
4.04010385e-01 8.63237023e-01 -9.41569880e-02 5.46381176e-01
-1.27318740e+00 1.24633396e+00 -4.07102741e-02 3.44200075e-01
-1.13362157e+00 -3.32245320e-01 5.35761476e-01 3.45452577e-01
1.72721267e+00 6.65268600e-02 -8.85172009e-01 -3.82120647e-02
-1.03721583e+00 4.13854867e-01 7.50302553e-01 5.34486353e-01
2.79037029e-01 -1.23743109e-01 1.58250645e-01 4.81484115e-01
7.19645843e-02 1.04040122e+00 -9.63559672e-02 -1.19882487e-01
4.71470058e-01 3.25156152e-01 4.28757459e-01 -1.37264764e+00
-7.79095352e-01 -3.27069819e-01 -6.44058168e-01 -8.12423944e-01
9.81376693e-02 -5.97282887e-01 -9.06866133e-01 1.72646213e+00
4.07051206e-01 4.38225955e-01 1.10019535e-01 2.82647491e-01
4.57187057e-01 1.63672364e+00 6.59237325e-01 -1.08710146e+00
1.59351230e+00 -6.43727854e-02 -1.23979163e+00 2.65965939e-01
3.19943070e-01 -8.66587579e-01 2.35524282e-01 9.04889777e-02
-7.64054775e-01 1.08838901e-01 -2.22479850e-01 8.50442708e-01
-5.37989259e-01 -3.76949251e-01 4.61825490e-01 -1.67638510e-02
-9.61951077e-01 2.38036718e-02 -1.19997823e+00 -1.21699131e+00
2.17593729e-01 3.64894420e-02 -9.13113058e-02 1.94019467e-01
-1.11099136e+00 6.27340913e-01 3.61052275e-01 -3.19308400e-01
-7.26612866e-01 -1.19420278e+00 -6.27212882e-01 -3.51709634e-01
2.08540171e-01 -3.79421353e-01 1.02367640e+00 -2.20372856e-01
-6.47402525e-01 5.32932222e-01 -5.97447455e-01 -7.08269954e-01
3.74568217e-02 -1.59210995e-01 -8.53991151e-01 4.27518487e-01
3.97075534e-01 9.92219672e-02 3.47263336e-01 -8.58167768e-01
-9.38226044e-01 -4.31153059e-01 -4.02610391e-01 -3.31139654e-01
-3.07759970e-01 8.27882349e-01 1.23541757e-01 -1.09836602e+00
-2.19895482e-01 -7.08388507e-01 -2.13338837e-01 -6.24414921e-01
-7.51956478e-02 -4.26819742e-01 1.16563368e+00 -1.03877461e+00
1.40142941e+00 -1.67927969e+00 -4.46522415e-01 -1.53712168e-01
2.16023698e-01 -3.11818779e-01 2.15010524e-01 1.19140506e+00
3.32150877e-01 3.91761243e-01 -1.25971377e-01 2.07067356e-01
-7.16197073e-01 1.81049868e-01 -1.00534272e+00 8.43392313e-01
6.90927327e-01 4.25700158e-01 -9.35059249e-01 -2.76851922e-01
-6.94428459e-02 6.52899921e-01 -4.07225788e-01 2.32657269e-01
-5.11616766e-01 4.64366078e-01 -5.61401606e-01 4.57509071e-01
6.87670887e-01 -7.53540516e-01 2.44372681e-01 1.79149002e-01
-6.57798350e-01 6.60737097e-01 -4.33984160e-01 7.33402431e-01
-5.93597889e-02 9.03734386e-01 7.82029554e-02 -6.11281037e-01
6.42648220e-01 9.25194979e-01 1.15083861e+00 -2.47089669e-01
-2.03566223e-01 -2.99646407e-01 -5.00028491e-01 -4.12553161e-01
5.06994367e-01 -1.01668738e-01 -4.45714667e-02 1.13676417e+00
-7.48824775e-02 8.29733834e-02 1.30337119e-01 3.56575400e-01
8.68394375e-01 -4.83349979e-01 5.14549792e-01 -4.59106833e-01
-3.70106995e-01 9.10531521e-01 4.59152937e-01 4.11150634e-01
8.29423815e-02 1.61794469e-01 4.88549739e-01 -7.26976514e-01
-1.19642878e+00 -1.02872884e+00 -5.62427044e-01 1.09611630e+00
-4.76455629e-01 -5.02188504e-01 -3.01426589e-01 -2.39543077e-02
-1.59515426e-01 5.33714414e-01 -9.36163485e-01 5.90481281e-01
-6.12436533e-01 -1.60817266e+00 9.40525010e-02 -5.81847057e-02
8.78923107e-03 -7.65806556e-01 -8.26939523e-01 6.38966382e-01
-4.75076169e-01 -9.67632711e-01 -3.30200016e-01 3.82426120e-02
-1.00957656e+00 -1.15131533e+00 -1.06800067e+00 -3.98699671e-01
7.67011762e-01 2.76768118e-01 8.71307969e-01 -4.86121058e-01
-1.86580390e-01 5.94150960e-01 -1.58023536e-01 -1.04629493e+00
-4.93110985e-01 -1.17484324e-01 3.23727965e-01 -5.29513180e-01
5.87098718e-01 -9.09013674e-02 -4.55806345e-01 7.52737820e-02
-1.07299662e+00 -2.18058065e-01 7.29126111e-02 3.74469101e-01
3.78010809e-01 2.62379766e-01 8.85258138e-01 -9.37779903e-01
7.42025197e-01 -1.61473668e+00 -5.76881289e-01 1.02578193e-01
-4.96038377e-01 -5.33486843e-01 7.36285299e-02 -7.54854560e-01
-1.00378215e+00 -5.83109379e-01 5.06485343e-01 2.44106278e-01
-3.27918321e-01 1.19389069e+00 1.22888136e+00 9.46692228e-01
3.40349913e-01 3.88707727e-01 2.05105186e-01 -2.83176631e-01
-3.74520987e-01 7.22530246e-01 1.83572039e-01 -3.41840722e-02
4.75012690e-01 7.44155109e-01 -6.49694920e-01 -1.33443093e+00
-5.22356451e-01 -1.02102184e+00 -2.14473531e-02 -1.63966432e-01
8.08459342e-01 -1.32453811e+00 -4.11304206e-01 5.84079504e-01
-1.60788310e+00 -2.10438855e-02 3.38999450e-01 8.65042329e-01
2.15435818e-01 -3.16764563e-01 -1.13867450e+00 -9.42104161e-01
-5.44348836e-01 -5.46673656e-01 8.60711992e-01 9.34156962e-03
-6.80058658e-01 -1.61439788e+00 8.26150298e-01 -3.27736825e-01
8.31649899e-01 7.21573651e-01 7.51644969e-01 -9.48435009e-01
2.69871429e-02 -7.50331432e-02 -1.29990384e-01 -6.34645641e-01
9.05768156e-01 5.18718779e-01 -5.42895794e-01 -4.73412305e-01
2.15697274e-01 1.86878636e-01 4.50953186e-01 1.04816067e+00
-2.29382347e-02 -1.16003597e+00 -8.36714029e-01 5.69113232e-02
1.26443326e+00 7.37944067e-01 -1.06842987e-01 2.60335684e-01
-5.09149656e-02 9.72791731e-01 3.59080613e-01 8.31307113e-01
4.51467603e-01 4.33812320e-01 -2.87283868e-01 -2.77353972e-01
7.35394061e-01 -3.84876435e-03 4.25353229e-01 1.01580179e+00
-7.25716427e-02 -4.84252721e-01 -1.41004920e+00 1.06680369e+00
-1.16056061e+00 -1.13845730e+00 -2.84843862e-01 1.90562022e+00
8.86684060e-01 -6.56346455e-02 5.46260476e-01 -6.97532773e-01
6.83250189e-01 3.89761627e-01 -8.86030123e-02 -2.49746516e-01
-8.37780535e-02 -4.39345360e-01 5.14108121e-01 2.34069303e-01
-1.27363372e+00 5.27458429e-01 7.32630777e+00 3.39522250e-02
-1.40167820e+00 4.21897799e-01 8.09319556e-01 8.01770091e-02
-1.98719308e-01 -4.49771106e-01 -6.30904078e-01 4.30074006e-01
1.47021580e+00 -6.36357665e-01 -2.41816416e-01 5.29468119e-01
8.11477005e-01 -8.48140940e-02 -3.12633127e-01 2.17859998e-01
1.70250814e-02 -1.59896564e+00 1.64457690e-03 1.80152133e-01
1.15527475e+00 4.50648010e-01 2.03143075e-01 -3.00005376e-01
5.18995523e-01 -4.78339911e-01 6.58453181e-02 2.90918350e-01
5.02723575e-01 -7.25313783e-01 8.81749690e-01 2.80422807e-01
-1.04054999e+00 4.40543056e-01 -2.02256888e-01 5.46519607e-02
6.48124874e-01 8.19781840e-01 -1.44663775e+00 1.64206892e-01
8.26526523e-01 8.20735037e-01 2.34619752e-01 6.10148966e-01
2.62304753e-01 1.13442636e+00 -6.27868116e-01 1.38685852e-01
5.63303351e-01 7.73876533e-02 7.82607794e-01 1.46441197e+00
4.42130923e-01 5.56342840e-01 1.99547037e-01 6.33407831e-01
3.45491111e-01 2.30863437e-01 -1.08232236e+00 -5.55860162e-01
2.76382506e-01 4.65834916e-01 -1.06328118e+00 -4.42080081e-01
-3.58200431e-01 2.94607550e-01 -3.40268523e-01 5.98670781e-01
-6.11130357e-01 -1.05238304e-01 6.44119799e-01 4.92294699e-01
1.55814275e-01 -3.84242356e-01 2.06024036e-01 -9.11189795e-01
-3.88464838e-01 -4.58405793e-01 5.47518492e-01 -3.59630138e-01
-1.26158011e+00 8.42771471e-01 5.99381924e-01 -1.01400483e+00
-5.37498534e-01 1.06107071e-01 -9.98968959e-01 6.13890588e-01
-1.12762535e+00 -8.27922881e-01 2.93366462e-01 3.17195833e-01
7.55041063e-01 1.26408920e-01 6.65053964e-01 -2.55834788e-01
-4.26612914e-01 -3.44583422e-01 4.85455006e-01 -5.44480048e-02
4.55563158e-01 -6.54838085e-01 5.93722761e-01 3.59874815e-01
-3.79140228e-01 1.08707309e+00 1.08718324e+00 -1.46951163e+00
-9.06989694e-01 -1.41345632e+00 1.27767479e+00 -2.51198471e-01
1.25260317e+00 -2.48744115e-01 -1.00542915e+00 9.05108213e-01
6.38512492e-01 -8.09045017e-01 9.10308421e-01 2.75445729e-01
-1.15432173e-01 2.68801033e-01 -1.06687200e+00 2.59238809e-01
-2.92625632e-02 -3.64304721e-01 -8.82771313e-01 1.02117074e+00
1.03381431e+00 -2.80902721e-03 -1.08272207e+00 1.48927122e-01
4.53476369e-01 -6.28074557e-02 6.28314435e-01 -9.82838333e-01
4.48357999e-01 1.85812920e-01 2.30764240e-01 -1.41096544e+00
-1.04657836e-01 -7.62051344e-01 2.56486237e-01 1.15006423e+00
5.74128449e-01 -9.18245375e-01 2.33821839e-01 2.28668854e-01
4.10379291e-01 -4.01903421e-01 -5.22682250e-01 -5.97926021e-01
-2.35977128e-01 -3.99168611e-01 4.36935693e-01 1.32967687e+00
7.67916963e-02 -1.25227883e-01 -5.42462766e-01 6.94814503e-01
5.16515076e-01 2.22669989e-01 5.27049005e-01 -1.27786016e+00
3.84472430e-01 1.16573840e-01 -5.61357252e-02 -4.14602816e-01
-4.13532317e-01 2.21430678e-02 -3.37696016e-01 -1.61995447e+00
5.35879791e-01 -1.99254215e-01 -1.47319362e-01 2.31236726e-01
2.40476001e-02 -6.02369532e-02 -2.23890245e-01 7.09745347e-01
-3.12473118e-01 -2.48104334e-02 4.79411364e-01 -2.60546714e-01
-6.83353961e-01 2.13954076e-01 -2.35494316e-01 7.18770802e-01
1.10596550e+00 -9.64379787e-01 -2.44413942e-01 -2.46523947e-01
2.91392744e-01 3.30250770e-01 2.17938885e-01 -3.98140252e-01
2.82329023e-01 -6.21435940e-01 7.51625886e-03 -1.18491518e+00
-1.26593068e-01 -7.65892267e-01 7.36323118e-01 9.75685239e-01
-2.58605421e-01 6.66694760e-01 6.46609068e-01 8.14803898e-01
-3.50468457e-01 7.21091211e-01 1.38879791e-01 1.75508797e-01
-1.73243031e-01 3.48813385e-01 -1.21235263e+00 6.12106957e-02
1.14296710e+00 2.96255171e-01 -9.04710293e-01 -5.83865047e-01
-4.97070521e-01 3.75267416e-01 1.38564721e-01 4.69371587e-01
4.44636047e-01 -7.09951460e-01 -1.23100650e+00 -1.83280170e-01
-1.70125030e-02 -5.08107007e-01 3.05322349e-01 1.22121716e+00
-6.28796756e-01 1.13302970e+00 3.16426978e-02 -7.40588367e-01
-9.15339112e-01 7.51911819e-01 -3.99428934e-01 -3.96265149e-01
-6.13019347e-01 2.10131481e-01 3.51275355e-01 -7.76239708e-02
1.09622963e-01 -6.21476948e-01 -2.91501224e-01 5.65673590e-01
9.31876481e-01 2.52281129e-01 -4.57093358e-01 -6.79691732e-01
-5.08503377e-01 3.15748602e-01 -4.28540260e-01 -6.46415651e-02
1.88724804e+00 -2.24765986e-01 -2.30818108e-01 9.94106770e-01
1.22718143e+00 1.66338531e-03 -9.56243694e-01 -1.46675736e-01
3.14042330e-01 2.86618322e-01 -3.27103250e-02 -6.56325579e-01
-6.64041638e-01 2.82354444e-01 3.07544500e-01 7.82593012e-01
9.37562168e-01 4.65120375e-01 8.49847198e-01 1.49053991e-01
2.90978193e-01 -6.19282186e-01 -5.19119024e-01 2.55039275e-01
9.52195048e-01 -1.18839419e+00 1.48307413e-01 -2.14138508e-01
-5.19730091e-01 9.02734697e-01 -2.56435424e-01 1.50055617e-01
1.12230372e+00 4.67525750e-01 9.04472917e-02 -7.42653310e-01
-1.20219338e+00 3.77333403e-01 1.67631462e-01 5.40766358e-01
4.32792664e-01 2.71228373e-01 -3.78038526e-01 2.76710510e-01
2.98135757e-01 -5.16753681e-02 5.30318737e-01 9.88810062e-01
-4.87740904e-01 -1.45658717e-01 -8.04431796e-01 8.49298954e-01
-1.25139844e+00 -3.28997493e-01 -3.03782791e-01 1.11050677e+00
-3.02921921e-01 1.01961410e+00 7.45069861e-01 6.25436474e-03
-1.53875053e-01 -3.74976359e-02 -4.46807653e-01 -4.44323242e-01
-5.01368046e-01 5.62744141e-01 -2.81549562e-02 -1.81205109e-01
-8.11878562e-01 -9.44427073e-01 -1.02310336e+00 -5.52097797e-01
-2.78529882e-01 5.12116134e-01 8.11483562e-01 8.06069493e-01
4.16331619e-01 2.65684903e-01 6.96039379e-01 -2.89456666e-01
2.61595368e-01 -1.26125431e+00 -5.14459014e-01 7.20453262e-02
7.46214390e-01 -3.44356537e-01 -4.52514738e-01 4.59914237e-01] | [5.976090431213379, 4.388387680053711] |
2aeb5ec1-3376-4b20-8c76-1cc07517074a | generalized-zero-shot-learning-for-medical | 2204.01728 | null | https://arxiv.org/abs/2204.01728v2 | https://arxiv.org/pdf/2204.01728v2.pdf | Interpretable Saliency Maps And Self-Supervised Learning For Generalized Zero Shot Medical Image Classification | In many real world medical image classification settings we do not have access to samples of all possible disease classes, while a robust system is expected to give high performance in recognizing novel test data. We propose a generalized zero shot learning (GZSL) method that uses self supervised learning (SSL) for: 1) selecting anchor vectors of different disease classes; and 2) training a feature generator. Our approach does not require class attribute vectors which are available for natural images but not for medical images. SSL ensures that the anchor vectors are representative of each class. SSL is also used to generate synthetic features of unseen classes. Using a simpler architecture, our method matches a state of the art SSL based GZSL method for natural images and outperforms all methods for medical images. Our method is adaptable enough to accommodate class attribute vectors when they are available for natural images. | ['Dwarikanath Mahapatra'] | 2022-04-04 | null | null | null | null | ['generalized-zero-shot-learning', 'generalized-zero-shot-learning'] | ['computer-vision', 'methodology'] | [ 7.41338909e-01 2.26240486e-01 -1.77769676e-01 -4.76325989e-01
-9.89086866e-01 -2.31586605e-01 7.29897678e-01 3.34790796e-01
-4.92365628e-01 7.70005047e-01 -1.00615852e-01 4.58419733e-02
-1.24932215e-01 -9.94962990e-01 -4.46024448e-01 -7.41169631e-01
-1.35596588e-01 8.52220595e-01 5.35828412e-01 -1.47239625e-01
-7.68347783e-03 5.40667057e-01 -1.95820260e+00 4.42984432e-01
6.51684105e-01 6.85315013e-01 2.59616971e-01 9.38969553e-01
-8.47318396e-02 5.51035404e-01 -6.73258126e-01 1.79015502e-01
3.56580555e-01 -7.63323605e-01 -8.38941395e-01 3.81311655e-01
3.63952726e-01 -1.14843890e-01 3.53902765e-02 9.17483926e-01
7.27508545e-01 2.63399959e-01 1.08474624e+00 -1.31798291e+00
-4.37791675e-01 2.08396092e-01 -2.27784619e-01 9.54618379e-02
4.36717957e-01 1.75540656e-01 6.98135138e-01 -8.37846935e-01
1.06521714e+00 1.08674741e+00 3.79050523e-01 1.02425826e+00
-1.27388144e+00 -3.41141105e-01 -3.36771548e-01 4.39758986e-01
-1.29026473e+00 -5.95791519e-01 6.80151761e-01 -4.94760543e-01
6.36673808e-01 3.35414469e-01 4.92534250e-01 1.24383962e+00
2.38782555e-01 8.71002138e-01 1.09329641e+00 -7.56672382e-01
8.47517788e-01 5.02198339e-01 5.18205874e-02 7.80928433e-01
1.41530588e-01 1.12181805e-01 -3.42666715e-01 -3.91256899e-01
7.02792764e-01 2.16533363e-01 -2.12635830e-01 -6.92739308e-01
-1.40066063e+00 9.66313303e-01 2.42069036e-01 4.37927365e-01
-3.57775599e-01 -2.81865448e-01 3.57877910e-01 4.70064789e-01
2.00339735e-01 5.93410432e-01 -3.17698896e-01 2.64016360e-01
-1.01558352e+00 2.71451734e-02 6.71665132e-01 1.02418864e+00
7.12257385e-01 8.24099481e-02 -4.29402888e-01 9.03210580e-01
-1.89903855e-01 4.84378397e-01 1.11996627e+00 -8.74931514e-01
-2.49017373e-01 5.24972618e-01 -3.97913195e-02 -6.77537203e-01
-3.05119991e-01 -2.64542222e-01 -8.13001812e-01 2.70222276e-01
2.55514979e-01 1.39918849e-01 -1.37234032e+00 1.63470626e+00
3.46858114e-01 3.85618657e-01 5.55923700e-01 3.21953863e-01
1.06314838e+00 6.27416134e-01 -2.41943747e-01 -3.63216102e-01
1.15532732e+00 -5.73970854e-01 -4.53843445e-01 -2.36589134e-01
6.79942310e-01 -5.17738163e-01 1.12939990e+00 2.91393310e-01
-6.13700151e-01 -4.68892515e-01 -1.16234481e+00 4.30646062e-01
-6.19944990e-01 -9.63073075e-02 4.82334435e-01 7.52101779e-01
-8.32292259e-01 5.55181265e-01 -8.57254863e-01 -6.64581537e-01
5.93514323e-01 3.43854547e-01 -5.93596995e-01 -3.10350984e-01
-1.05689538e+00 6.63813233e-01 3.60012621e-01 -4.78329569e-01
-1.06550956e+00 -4.08630639e-01 -1.18609989e+00 -1.23951450e-01
2.95745462e-01 -5.07017910e-01 1.06807697e+00 -1.10152090e+00
-1.08191764e+00 1.20604837e+00 5.11718318e-02 -5.13831079e-01
4.52517807e-01 5.08244395e-01 -5.25737226e-01 4.71055478e-01
2.61982679e-01 8.11253905e-01 1.05844080e+00 -1.25348294e+00
-3.92588586e-01 -1.38947427e-01 -4.62054044e-01 -3.39911878e-02
-3.60787004e-01 -3.65091592e-01 -9.03778151e-02 -6.81422114e-01
-2.79291836e-03 -8.50505531e-01 -5.81284940e-01 3.22458148e-01
-6.00694299e-01 1.24532497e-02 8.00891757e-01 2.14684941e-03
7.57804036e-01 -2.08126378e+00 -4.08804983e-01 3.75096858e-01
1.02454022e-01 4.61437017e-01 -4.28769410e-01 3.52761596e-01
-2.25674406e-01 -3.18074226e-01 -4.14120764e-01 1.27718613e-01
-3.97658259e-01 5.62807858e-01 -1.40759975e-01 3.46480310e-01
3.32625747e-01 6.70659125e-01 -1.12006819e+00 -8.32338333e-01
3.32014471e-01 2.52701461e-01 -4.49771613e-01 1.94579437e-01
-2.10054163e-02 1.42736152e-01 -3.69835973e-01 6.34702086e-01
3.47731113e-01 -4.73332345e-01 7.04474282e-03 1.70800276e-02
5.55909216e-01 -3.53539467e-01 -1.29437065e+00 1.38656878e+00
-1.95405826e-01 3.46719325e-01 -5.95646739e-01 -1.04001176e+00
1.06929648e+00 5.36337435e-01 6.27062082e-01 -2.23069400e-01
5.06989546e-02 1.13042146e-01 8.49606246e-02 -6.15791082e-01
-2.96387911e-01 -3.42187166e-01 -9.08875316e-02 3.75742525e-01
5.98487675e-01 -2.91764498e-01 1.54903546e-01 3.68945688e-01
1.35871005e+00 -4.20286119e-01 7.01377392e-01 -1.94259197e-01
5.14273167e-01 -4.55765091e-02 7.44318306e-01 1.08382559e+00
-3.10103714e-01 8.06383312e-01 2.36760512e-01 -6.71142101e-01
-1.12448943e+00 -1.46059346e+00 -3.39402586e-01 6.66615069e-01
-1.72402170e-02 -5.73017299e-02 -7.50861168e-01 -8.18240106e-01
-2.27171347e-01 7.75203764e-01 -9.25220370e-01 -4.23201054e-01
9.26180650e-03 -7.37884164e-01 2.52681941e-01 2.79797971e-01
1.51709586e-01 -1.40406120e+00 -8.93185675e-01 2.82569349e-01
1.54907167e-01 -8.51355076e-01 -3.69544029e-01 2.46357307e-01
-6.14771962e-01 -1.26208997e+00 -8.44865799e-01 -1.11976802e+00
1.06742334e+00 4.08820249e-02 1.06748605e+00 -9.07849967e-02
-1.08954334e+00 5.23797691e-01 -4.64515001e-01 -5.97704828e-01
-8.19882154e-01 -1.82764634e-01 1.80127233e-01 4.11656410e-01
5.59751987e-01 -2.47531712e-01 -4.92754042e-01 2.65981853e-01
-1.18084180e+00 -1.62672680e-02 6.46222770e-01 1.36712861e+00
7.72438645e-01 -2.62250798e-03 8.24885666e-01 -1.46844530e+00
3.03619683e-01 -3.82038414e-01 -1.55184537e-01 4.36581224e-01
-6.62164748e-01 1.07834555e-01 7.55116522e-01 -7.51246870e-01
-6.37052357e-01 4.09235567e-01 -6.66311011e-02 -5.06346107e-01
-6.02966905e-01 1.96074229e-03 1.09680690e-01 -5.68805262e-03
1.26087236e+00 3.16366762e-01 4.14731562e-01 -1.96350038e-01
1.60784319e-01 8.82323384e-01 5.73869884e-01 -1.16941601e-01
6.69415832e-01 5.39345145e-01 9.67647061e-02 -9.93267417e-01
-7.84541011e-01 -6.01880372e-01 -6.89170003e-01 -7.97950923e-02
7.56467938e-01 -6.36122108e-01 -4.47464585e-02 3.23090851e-01
-4.33325320e-01 -8.11785609e-02 -8.59989285e-01 5.71972787e-01
-9.21089232e-01 1.54641703e-01 -5.26081622e-01 -6.17898405e-01
-3.81542712e-01 -1.28393817e+00 1.08298242e+00 2.86863893e-02
-3.57323736e-01 -8.83910418e-01 1.86448961e-01 3.79466154e-02
2.68482327e-01 4.84366030e-01 9.44416881e-01 -1.17830455e+00
-2.23274231e-01 -6.44860566e-01 2.39380285e-01 3.14366192e-01
6.05790734e-01 -1.71593428e-01 -1.00035095e+00 -5.75784564e-01
-1.82517078e-02 -7.08079159e-01 9.08126593e-01 3.67026001e-01
1.07548618e+00 -3.43880028e-01 -5.65590143e-01 3.82735848e-01
1.35293591e+00 3.75508338e-01 6.53334618e-01 -3.92068624e-02
3.17449450e-01 6.85164511e-01 6.77611589e-01 3.90896022e-01
-8.10318291e-02 3.93501788e-01 -2.72404291e-02 -5.10585129e-01
-1.03850421e-02 7.01872557e-02 -3.63598205e-02 4.61633623e-01
4.25303638e-01 -1.00518622e-01 -9.99486506e-01 8.17602277e-01
-1.65677929e+00 -1.10045552e+00 2.81835288e-01 2.42490458e+00
9.76548016e-01 1.61110252e-01 -1.55595783e-02 2.19222888e-01
7.21123159e-01 -2.67969072e-01 -7.88649738e-01 -1.00311741e-01
-7.63314515e-02 6.17977500e-01 2.56831050e-01 2.87523001e-01
-1.26239693e+00 6.63275301e-01 6.68980408e+00 9.30071712e-01
-8.48325908e-01 1.20428363e-02 8.48918080e-01 -1.16870016e-01
-5.43102212e-02 -2.28056535e-01 -5.09926975e-01 2.75911301e-01
9.09487963e-01 -2.91975886e-01 -3.74216698e-02 9.82590556e-01
-1.48386821e-01 -9.14838724e-03 -1.15815699e+00 1.22201657e+00
4.09287274e-01 -1.39300680e+00 2.97101557e-01 -2.52526522e-01
7.68425584e-01 -1.22072481e-01 -7.53618181e-02 1.58871830e-01
3.89388651e-01 -1.00573409e+00 -7.17512369e-02 5.45007467e-01
1.10496008e+00 -7.05248177e-01 7.72654593e-01 3.84324849e-01
-6.78384960e-01 6.64249733e-02 -5.09834349e-01 5.91786563e-01
-2.03596219e-01 3.70828182e-01 -1.29630601e+00 8.58962461e-02
4.28137839e-01 6.60818875e-01 -8.29071879e-01 1.09595799e+00
-2.12869998e-02 5.10064363e-01 -1.50010288e-01 4.98715639e-02
2.44546399e-01 3.79683584e-01 4.33790535e-01 1.16553211e+00
2.76961774e-01 -1.03272296e-01 5.33648729e-01 3.85416299e-01
2.39439338e-01 3.54183316e-01 -9.88712430e-01 2.77995560e-02
4.36999083e-01 1.10656297e+00 -1.05261266e+00 -7.06002414e-01
-2.99099237e-01 1.25307810e+00 -1.96589410e-01 1.20059021e-01
-1.60238847e-01 -7.81369388e-01 3.12555283e-01 7.00136796e-02
2.84000367e-01 4.48308200e-01 3.29158723e-01 -1.23674142e+00
-4.38036382e-01 -9.68784273e-01 7.32205868e-01 -6.91854060e-01
-1.57331836e+00 9.52493072e-01 -1.16597205e-01 -1.84347534e+00
-7.43408561e-01 -5.86359620e-01 -4.46683049e-01 4.94876385e-01
-1.14666474e+00 -1.07347488e+00 -2.04508528e-01 8.82853985e-01
7.46350586e-01 -7.28017032e-01 1.44297874e+00 -1.35977060e-01
-1.05319001e-01 5.85327506e-01 3.89235467e-01 2.35816270e-01
1.02228570e+00 -1.33615696e+00 3.11405987e-01 5.12511075e-01
3.56228143e-01 2.80680895e-01 7.21329927e-01 -6.10552728e-01
-9.87492919e-01 -1.19836497e+00 8.01172674e-01 -2.83228487e-01
2.34634146e-01 -3.98274958e-01 -1.11335087e+00 4.13080722e-01
-2.10730612e-01 4.07373101e-01 1.14625156e+00 -3.89191091e-01
-2.21707329e-01 -2.35863164e-01 -1.62001753e+00 4.34696555e-01
5.67578733e-01 -3.04673910e-01 -6.41207337e-01 7.33127892e-01
3.98533374e-01 -6.89094737e-02 -6.81109071e-01 3.33875000e-01
3.75060767e-01 -6.64363027e-01 1.06855202e+00 -7.88151443e-01
1.27960593e-01 -3.54512185e-01 -1.29376754e-01 -1.39378166e+00
-3.25778306e-01 -2.36125275e-01 2.35860378e-01 7.21873105e-01
5.28434813e-01 -6.95963979e-01 8.60640287e-01 5.35559356e-01
2.74023682e-01 -8.56229961e-01 -9.61489975e-01 -1.03854525e+00
-2.61797220e-01 -1.00143902e-01 2.77786136e-01 1.26804352e+00
-3.51256728e-02 3.31344575e-01 -3.78689855e-01 -1.10158645e-01
9.56315219e-01 1.22036144e-01 5.59772730e-01 -1.43247211e+00
-5.96659601e-01 -1.09668262e-02 -1.12576413e+00 -5.49517386e-02
-5.70504442e-02 -9.78077233e-01 1.58256575e-01 -1.37599075e+00
3.21587861e-01 -4.73053008e-01 -4.25072014e-01 6.93825781e-01
-5.14874086e-02 6.00639999e-01 -1.19969808e-01 1.67358115e-01
-4.73453432e-01 2.43077308e-01 9.90201116e-01 -3.35377604e-01
-1.87813535e-01 1.28127694e-01 -4.75342214e-01 6.48064494e-01
6.83966398e-01 -5.39099574e-01 -7.43891537e-01 3.50741863e-01
-4.03600574e-01 1.82022989e-01 1.58889383e-01 -1.17731190e+00
1.07131466e-01 -2.16449052e-01 8.09954166e-01 -4.59236562e-01
2.58051276e-01 -6.82968855e-01 1.07768916e-01 7.23616958e-01
-5.26358783e-01 -4.20275062e-01 -2.58652568e-01 6.65333390e-01
-4.41859700e-02 -6.40337586e-01 1.33465624e+00 -6.20977938e-01
-7.50807166e-01 4.34773088e-01 -5.40016353e-01 -4.83478867e-02
1.38002157e+00 -3.97528917e-01 -2.23775650e-03 -3.40018958e-01
-1.10092485e+00 1.34807736e-01 5.74258208e-01 3.39226156e-01
1.03001142e+00 -1.35214305e+00 -8.46340716e-01 7.30851591e-01
6.96320355e-01 -2.43215546e-01 1.56010807e-01 2.27523372e-01
-4.75833684e-01 -4.40617912e-02 -3.08839798e-01 -6.74834967e-01
-1.36802936e+00 9.61557269e-01 2.37975866e-01 -1.73709437e-01
-7.22971797e-01 7.37693727e-01 1.93101183e-01 -4.13768947e-01
1.19780444e-01 2.16837764e-01 -2.94540733e-01 9.07557085e-02
9.59342182e-01 -3.89733464e-02 1.82526603e-01 -5.22529423e-01
-2.67969638e-01 3.22265148e-01 -4.22068894e-01 -1.45388171e-01
1.49091327e+00 3.07628989e-01 3.31258059e-01 8.81315827e-01
1.20500267e+00 -4.84007597e-01 -8.32735360e-01 -5.66549897e-01
-2.43028123e-02 -4.53836679e-01 -2.05696583e-01 -6.59044027e-01
-8.02641988e-01 7.51105666e-01 9.46121991e-01 2.29274631e-02
1.29213715e+00 1.38340279e-01 5.33216119e-01 6.76708162e-01
6.10117972e-01 -1.12978530e+00 2.57854760e-01 -3.46708298e-02
5.25222600e-01 -1.42206597e+00 -6.35720193e-02 -3.04030508e-01
-7.36516893e-01 1.13026249e+00 4.31042105e-01 -2.47195497e-01
7.16701448e-01 1.77020773e-01 2.63161093e-01 -2.24712417e-01
-9.38269258e-01 -3.03617299e-01 3.24392140e-01 9.11131263e-01
1.69023037e-01 3.06220427e-02 -8.21144953e-02 4.41214181e-02
1.33204944e-02 -1.19040541e-01 7.31186688e-01 1.27744436e+00
-5.20722568e-01 -1.25542819e+00 -3.81482065e-01 1.13645613e+00
-2.61618763e-01 -4.88657467e-02 -2.65484452e-01 4.44992065e-01
2.95014065e-02 6.61266863e-01 2.71394968e-01 -1.36799484e-01
1.33976743e-01 1.85080662e-01 5.14190316e-01 -1.13269424e+00
-7.92102888e-02 1.05687603e-02 -1.82238936e-01 -3.61798733e-01
-2.73716748e-01 -6.51765347e-01 -1.13663375e+00 4.02583718e-01
-4.30224478e-01 1.63715988e-01 2.41262719e-01 7.45301008e-01
1.76031977e-01 2.35828191e-01 9.84048367e-01 -6.11769199e-01
-2.95634627e-01 -6.96940422e-01 -8.84233475e-01 8.24096501e-01
4.31578994e-01 -6.60627902e-01 -1.68417931e-01 4.89693373e-01] | [9.927934646606445, 2.996230125427246] |
d6d94564-1452-4659-9dd9-abf23bc49305 | causal-discovery-with-missing-data-in-a | 2305.1005 | null | https://arxiv.org/abs/2305.10050v1 | https://arxiv.org/pdf/2305.10050v1.pdf | Causal Discovery with Missing Data in a Multicentric Clinical Study | Causal inference for testing clinical hypotheses from observational data presents many difficulties because the underlying data-generating model and the associated causal graph are not usually available. Furthermore, observational data may contain missing values, which impact the recovery of the causal graph by causal discovery algorithms: a crucial issue often ignored in clinical studies. In this work, we use data from a multi-centric study on endometrial cancer to analyze the impact of different missingness mechanisms on the recovered causal graph. This is achieved by extending state-of-the-art causal discovery algorithms to exploit expert knowledge without sacrificing theoretical soundness. We validate the recovered graph with expert physicians, showing that our approach finds clinically-relevant solutions. Finally, we discuss the goodness of fit of our graph and its consistency from a clinical decision-making perspective using graphical separation to validate causal pathways. | ['Fabio Stella', 'Marco Scutari', 'Casper Reijnen', 'Hanny Pijnenborg', 'Peter J. F. Lucas', 'Alice Bernasconi', 'Alessio Zanga'] | 2023-05-17 | null | null | null | null | ['causal-inference', 'causal-discovery', 'causal-inference'] | ['knowledge-base', 'knowledge-base', 'miscellaneous'] | [ 4.69461441e-01 6.55839384e-01 -8.27881932e-01 -8.68571177e-02
-3.60615879e-01 -6.13261104e-01 2.53487885e-01 7.06853986e-01
1.48306163e-02 1.09190965e+00 5.88128865e-01 -1.08990920e+00
-1.14036047e+00 -8.12973559e-01 -9.70108330e-01 -4.02660638e-01
-5.69869339e-01 5.49747884e-01 -1.48432940e-01 2.51093447e-01
-8.20777379e-03 2.72329092e-01 -8.35610151e-01 3.66265297e-01
1.10008299e+00 1.05194978e-01 -2.43182003e-01 3.95549834e-01
2.65267640e-01 7.54202247e-01 -2.77110338e-01 -4.37284321e-01
-1.24894939e-01 -6.85551822e-01 -6.30178094e-01 -2.49437466e-01
-2.61473861e-02 -1.86116099e-01 -2.19512030e-01 7.39484966e-01
3.76548439e-01 -5.49037635e-01 7.81461060e-01 -1.47874701e+00
-3.88089001e-01 1.08281398e+00 -6.85209751e-01 1.87698051e-01
3.60751152e-01 -6.48270994e-02 1.13653207e+00 -3.58024329e-01
9.64120686e-01 1.31713510e+00 6.07440889e-01 6.02266230e-02
-1.58701038e+00 -8.41671467e-01 1.02321491e-01 1.44221947e-01
-1.16052759e+00 -2.97226965e-01 4.10249889e-01 -6.21164322e-01
2.83152670e-01 4.27639067e-01 6.36226952e-01 1.33137453e+00
3.56020987e-01 3.04506361e-01 1.34510267e+00 -3.92378896e-01
3.60547870e-01 -2.24379927e-01 1.69237003e-01 7.68668950e-01
9.32245910e-01 5.51621854e-01 -3.98366660e-01 -9.01732683e-01
8.84745359e-01 9.65000093e-02 -4.03052181e-01 -2.79512286e-01
-1.21274781e+00 1.05045485e+00 2.94280589e-01 1.40460327e-01
-4.97162491e-01 1.31451398e-01 5.00661582e-02 2.51991540e-01
2.60383308e-01 3.44570160e-01 -4.51072335e-01 3.22094142e-01
-9.02748644e-01 2.12738857e-01 7.67057776e-01 7.30802953e-01
-5.36817610e-02 -4.52459693e-01 -3.57878745e-01 2.70660996e-01
4.14726436e-01 4.34677541e-01 -4.71696556e-02 -6.72818959e-01
6.93551004e-02 7.12804854e-01 1.32615268e-01 -9.04030383e-01
-6.30708218e-01 -5.58519721e-01 -9.17129695e-01 -1.23531044e-01
6.70215130e-01 -2.94294864e-01 -8.10065329e-01 1.65763402e+00
5.08221030e-01 4.81830418e-01 -1.55312940e-01 7.88732052e-01
5.61434090e-01 -1.38225675e-01 3.88957530e-01 -6.21111810e-01
1.44545496e+00 -2.21089870e-01 -8.88567269e-01 7.26752430e-02
6.88488722e-01 -3.49804729e-01 6.52724564e-01 4.23341811e-01
-8.31713736e-01 2.24361077e-01 -7.46099949e-01 2.95290321e-01
1.03552118e-01 -5.08695766e-02 8.41687262e-01 5.08891046e-01
-5.38352430e-01 7.44766533e-01 -7.06799626e-01 -3.45649391e-01
5.25654197e-01 2.41318524e-01 -4.04674351e-01 -3.16482782e-01
-1.35268354e+00 6.54419363e-01 3.72101367e-01 2.94595566e-02
-9.14678395e-01 -1.42578721e+00 -4.67583120e-01 2.28410453e-01
9.02115762e-01 -1.27130556e+00 8.11164677e-01 -4.38238502e-01
-5.90414405e-01 3.75736624e-01 -2.73116499e-01 -2.95571506e-01
7.96523511e-01 2.56575584e-01 -4.44564641e-01 2.74083391e-02
2.57656306e-01 -2.00638935e-01 3.92571777e-01 -1.22546864e+00
-5.79126656e-01 -6.75913751e-01 -3.15751880e-01 -3.57258439e-01
1.33768290e-01 -1.19932830e-01 -7.90199935e-02 -6.25152767e-01
1.05449185e-01 -8.50122154e-01 -6.48280740e-01 -1.51244238e-01
-7.41948485e-01 5.74279055e-02 3.08928847e-01 -6.19029939e-01
1.57313824e+00 -1.99501240e+00 8.57285857e-02 4.64954287e-01
6.46875858e-01 -4.22584176e-01 5.32192886e-02 6.65588021e-01
-6.24534905e-01 4.44530487e-01 -4.00149435e-01 3.28481823e-01
-4.07391995e-01 1.97602779e-01 -2.98930943e-01 6.48078501e-01
2.62284875e-01 1.02257764e+00 -1.08211839e+00 -5.61123788e-01
-1.26257211e-01 1.91934317e-01 -7.33901262e-01 -2.90525351e-02
-1.01107657e-01 5.96536338e-01 -6.28775716e-01 4.90166605e-01
4.38209444e-01 -6.11626744e-01 1.18621409e+00 7.24006668e-02
-4.49667349e-02 2.83749133e-01 -1.12140524e+00 1.32470644e+00
-7.81221613e-02 1.85509607e-01 -9.00795162e-02 -1.02541709e+00
4.69302326e-01 3.64381075e-01 6.69340611e-01 -4.41375285e-01
-6.05913922e-02 1.01357460e-01 4.87344414e-01 -5.82898378e-01
-2.86540121e-01 -4.80082840e-01 1.18889011e-01 4.75622356e-01
-4.53182667e-01 5.25991440e-01 7.25917146e-02 3.15198720e-01
1.57796228e+00 -3.20906699e-01 7.05868959e-01 -4.78352785e-01
-1.64166376e-01 4.50187862e-01 8.55547786e-01 8.52756500e-01
3.39107513e-01 3.86774331e-01 1.36343575e+00 -1.26690477e-01
-7.14205682e-01 -1.02866399e+00 -5.32821178e-01 3.05945456e-01
-4.36890960e-01 -2.43291587e-01 -2.68263370e-01 -7.90883482e-01
4.97302920e-01 8.96128953e-01 -1.14998078e+00 -2.89545745e-01
-1.85335711e-01 -1.30959344e+00 5.52827716e-01 3.60902637e-01
-4.12807554e-01 -5.21020055e-01 -5.30868173e-01 3.61990720e-01
-2.55683839e-01 -8.03102672e-01 1.10941678e-02 7.91275650e-02
-1.11017275e+00 -1.88308549e+00 -3.91810179e-01 6.52300715e-02
8.83850276e-01 -6.30759150e-02 1.17710972e+00 1.53102636e-01
-2.98172325e-01 -1.20281922e-02 -1.30120367e-01 -6.09475076e-01
-6.36659503e-01 -4.13132071e-01 -8.99081230e-02 -2.07555473e-01
3.46463248e-02 -7.27901161e-01 -6.89518452e-01 3.16342086e-01
-8.49787831e-01 -2.04272568e-02 9.05248821e-01 9.21693265e-01
4.47552621e-01 1.54538617e-01 7.41244912e-01 -1.39795864e+00
6.52738750e-01 -9.46100891e-01 -7.99150765e-01 3.21839660e-01
-1.10726011e+00 1.43468976e-01 1.54739976e-01 -3.58873785e-01
-1.00471771e+00 -2.71243602e-01 4.19625372e-01 -1.45465851e-01
3.18962336e-03 1.12950516e+00 -9.21303853e-02 4.36406910e-01
7.10543513e-01 -5.37112296e-01 9.22599733e-02 -4.14492756e-01
3.20743501e-01 2.10516587e-01 1.77621394e-01 -3.15480828e-01
3.98316145e-01 6.00144029e-01 6.34295166e-01 -3.02317083e-01
-7.41006315e-01 -3.19353640e-01 -4.95545894e-01 2.36980796e-01
4.84172106e-01 -7.87396908e-01 -1.33612692e+00 -6.02142513e-01
-1.04964685e+00 -2.77438968e-01 -1.72593951e-01 8.74349058e-01
-2.14298815e-01 8.60282779e-02 -3.06777328e-01 -7.87334621e-01
3.45165767e-02 -9.30552006e-01 8.34862649e-01 -3.69798720e-01
-5.00835955e-01 -1.22225118e+00 4.72458214e-01 2.73692459e-01
-2.17158407e-01 6.50983453e-01 1.54890406e+00 -5.69034755e-01
-3.87831748e-01 -4.62994091e-02 -5.40199399e-01 -8.63512337e-01
2.21710682e-01 1.41483337e-01 -6.26537442e-01 -1.92873925e-01
-3.85208935e-01 3.75430018e-01 8.01201046e-01 1.06033492e+00
9.64443684e-01 -5.87083638e-01 -8.94551516e-01 2.36796603e-01
1.23778796e+00 -3.86498310e-03 4.56722528e-01 -9.85787287e-02
6.62693262e-01 1.11428797e+00 4.05302733e-01 5.97891927e-01
3.02701175e-01 6.53649032e-01 4.73331004e-01 -4.09717292e-01
3.57247028e-03 -6.30944371e-01 -2.84181356e-01 3.09866108e-02
-2.24623993e-01 -1.95365816e-01 -1.15834415e+00 5.27642429e-01
-2.29606819e+00 -7.35950887e-01 -1.00349987e+00 2.39259529e+00
9.39822972e-01 4.18199860e-02 3.06278467e-01 9.62101743e-02
5.05306125e-01 -6.11758113e-01 -3.09106320e-01 -6.31816033e-03
-1.69080291e-02 -3.65939029e-02 8.62771094e-01 4.25085485e-01
-4.63601470e-01 2.54558712e-01 7.11548996e+00 3.68392229e-01
-7.41864800e-01 1.84449077e-01 5.53355515e-01 -8.11671764e-02
-7.22158551e-01 5.50630152e-01 -2.77806878e-01 2.39563540e-01
1.11388469e+00 -4.25614089e-01 -1.22931309e-01 2.34897748e-01
1.02440023e+00 -3.05591464e-01 -1.28119445e+00 3.77959758e-01
-5.55436909e-01 -1.54806924e+00 -1.78928405e-01 8.16435575e-01
7.12338567e-01 -4.52676415e-01 -3.02521884e-01 -3.33578348e-01
1.00326824e+00 -1.32349813e+00 3.88807654e-01 6.82068527e-01
9.92657781e-01 -4.85017836e-01 8.53418350e-01 7.43368492e-02
-3.51695478e-01 -2.80236721e-01 -1.83836669e-01 3.80215086e-02
1.66861087e-01 1.42961323e+00 -1.46287847e+00 1.10117555e+00
4.71830368e-01 6.20466352e-01 -3.86802584e-01 9.79405165e-01
-4.27525789e-01 1.15721655e+00 5.11384681e-02 4.04069483e-01
-1.32174417e-01 -1.26269937e-01 4.37781960e-01 8.87622952e-01
3.57987285e-01 3.48172963e-01 -2.80513585e-01 1.17396033e+00
-4.16055396e-02 3.35923210e-02 -8.19313705e-01 -1.54258117e-01
2.93039680e-01 9.55401659e-01 -7.00828493e-01 -1.02523938e-01
-3.44889283e-01 3.01616937e-01 6.44680187e-02 2.98258960e-01
-6.68333352e-01 4.80263025e-01 3.34741503e-01 5.45610607e-01
-1.18391402e-01 1.42685071e-01 -5.42169392e-01 -8.08675468e-01
-3.60771775e-01 -9.61153150e-01 1.06249940e+00 -4.30565506e-01
-1.19538748e+00 -1.84188366e-01 4.00761276e-01 -8.95202756e-01
-3.20745379e-01 -2.45684028e-01 -3.81193101e-01 8.47010851e-01
-1.24353850e+00 -9.80559587e-01 -7.62662068e-02 4.03888881e-01
-1.25840262e-01 3.71959686e-01 6.72472417e-01 1.73641041e-01
-7.32865214e-01 2.94508666e-01 -2.42364630e-02 -1.83153316e-01
8.08227718e-01 -1.22293234e+00 -2.49774009e-02 6.58341825e-01
-2.25485489e-01 8.02810967e-01 8.72547507e-01 -1.28159261e+00
-1.41258132e+00 -8.25639844e-01 9.88509059e-01 -3.72367859e-01
1.16996562e+00 -8.68943334e-02 -8.80178928e-01 7.60976434e-01
-2.22331360e-01 -1.53102845e-01 8.68442535e-01 7.55659461e-01
-2.99450099e-01 2.14753062e-01 -9.81336296e-01 7.38451183e-01
1.24238753e+00 5.84140494e-02 -4.09236670e-01 2.25252762e-01
6.24340892e-01 2.00683415e-01 -9.47379947e-01 4.90299731e-01
7.28522897e-01 -7.60941982e-01 8.19279373e-01 -1.14421475e+00
6.89260185e-01 -2.39437968e-01 2.54190862e-01 -1.27272475e+00
-4.59572285e-01 -5.16743422e-01 2.28260875e-01 8.46849740e-01
7.44014382e-01 -4.95732635e-01 6.13057613e-01 5.01551509e-01
2.18846723e-01 -3.31672817e-01 -8.30943704e-01 -4.61750805e-01
1.09560765e-01 -3.00590932e-01 4.65772122e-01 1.45927525e+00
3.18079323e-01 5.16102672e-01 -2.73519337e-01 4.22577709e-01
9.64452088e-01 4.02594715e-01 5.27211607e-01 -1.63462472e+00
-5.13719559e-01 -1.37390882e-01 -2.81198509e-02 1.60359107e-02
-6.96661845e-02 -7.98316777e-01 -4.89732921e-01 -1.62701201e+00
7.09980011e-01 -6.67048693e-01 -1.27078697e-01 7.24131107e-01
-3.78199071e-01 -3.88952076e-01 -3.61099869e-01 1.36034325e-01
1.09880581e-01 2.06095651e-01 1.21690810e+00 -1.15340576e-01
-3.35549593e-01 -1.88956205e-02 -9.06748831e-01 6.74453795e-01
4.75859195e-01 -1.09441352e+00 -6.60898030e-01 -2.06430256e-03
6.64467573e-01 7.47097909e-01 8.91305625e-01 7.56850243e-02
6.30861595e-02 -4.63355482e-01 1.60375908e-01 -3.29603672e-01
-4.87038374e-01 -8.72858882e-01 9.43838239e-01 9.89077806e-01
-4.80103254e-01 2.59059388e-02 2.68572897e-01 9.58215177e-01
9.06572863e-03 -1.21654034e-01 6.86466619e-02 3.53394151e-02
2.59356555e-02 -2.50539407e-02 -1.73137784e-02 -1.08040296e-01
7.90465057e-01 3.26043457e-01 -5.33046186e-01 -2.60560453e-01
-8.44870508e-01 2.38618344e-01 1.04757547e-01 1.11325070e-01
4.39332902e-01 -1.05482078e+00 -1.26697576e+00 -2.67611176e-01
2.26434439e-01 -3.53061706e-01 4.28925186e-01 1.54496658e+00
3.05406228e-02 4.58795726e-01 8.65070447e-02 -3.55177641e-01
-1.34754157e+00 9.00742471e-01 -4.55078073e-02 -5.62877655e-01
-5.79977751e-01 2.85657439e-02 5.79909563e-01 -3.79531793e-02
-2.19532460e-01 -1.64712042e-01 -8.18397030e-02 2.13451654e-01
3.22481096e-01 6.09050155e-01 -1.94281414e-02 1.76114693e-01
-6.14851534e-01 -4.06971276e-02 3.00327927e-01 -1.62035972e-01
1.37602985e+00 -7.58205215e-03 -3.64374310e-01 2.58745044e-01
6.80098951e-01 4.63493973e-01 -7.94451118e-01 2.06259992e-02
2.65639395e-01 -3.64493430e-01 4.82479520e-02 -9.33473468e-01
-7.38444448e-01 4.05045152e-01 2.23225042e-01 1.72777727e-01
8.29007924e-01 7.06295446e-02 -3.02246243e-01 -1.80160448e-01
9.18573961e-02 -3.41597199e-01 -4.73521471e-01 -4.42291170e-01
9.56083655e-01 -1.29267716e+00 4.68654871e-01 -9.74974751e-01
-2.86723137e-01 7.58306146e-01 -1.61631465e-01 3.02451670e-01
6.99826121e-01 2.54371047e-01 -3.12194824e-01 -6.04468286e-01
-1.07072425e+00 -9.09805000e-02 4.09086406e-01 5.80778003e-01
5.56467354e-01 5.28449714e-01 -9.02457833e-01 8.76858234e-01
2.80334670e-02 2.96338320e-01 8.09663117e-01 5.99028945e-01
3.59313071e-01 -1.10935152e+00 -6.06112063e-01 6.22529507e-01
-7.59170592e-01 -3.59644413e-01 -5.51335990e-01 1.07041204e+00
-9.69396457e-02 1.23719811e+00 -1.44851223e-01 1.90411806e-01
5.66330433e-01 -2.95037508e-01 2.72091717e-01 -4.01090086e-01
-1.17850825e-01 4.89971429e-01 4.28180546e-01 -5.97576559e-01
-4.11733717e-01 -7.93452024e-01 -1.12601125e+00 -4.24834758e-01
-2.85145402e-01 2.54537821e-01 2.90546626e-01 8.77181888e-01
6.55001462e-01 9.52278852e-01 3.88947129e-01 3.82509053e-01
-2.93927878e-01 -6.39753401e-01 -5.36720812e-01 2.93178380e-01
2.51407772e-01 -7.58762300e-01 -2.20924646e-01 8.77170339e-02] | [7.901843070983887, 5.38322114944458] |
65467737-416c-4a66-8508-41600d3d96d2 | canonical-saliency-maps-decoding-deep-face | 2105.01386 | null | https://arxiv.org/abs/2105.01386v2 | https://arxiv.org/pdf/2105.01386v2.pdf | Canonical Saliency Maps: Decoding Deep Face Models | As Deep Neural Network models for face processing tasks approach human-like performance, their deployment in critical applications such as law enforcement and access control has seen an upswing, where any failure may have far-reaching consequences. We need methods to build trust in deployed systems by making their working as transparent as possible. Existing visualization algorithms are designed for object recognition and do not give insightful results when applied to the face domain. In this work, we present 'Canonical Saliency Maps', a new method that highlights relevant facial areas by projecting saliency maps onto a canonical face model. We present two kinds of Canonical Saliency Maps: image-level maps and model-level maps. Image-level maps highlight facial features responsible for the decision made by a deep face model on a given image, thus helping to understand how a DNN made a prediction on the image. Model-level maps provide an understanding of what the entire DNN model focuses on in each task and thus can be used to detect biases in the model. Our qualitative and quantitative results show the usefulness of the proposed canonical saliency maps, which can be used on any deep face model regardless of the architecture. | ['C V Jawahar', 'Vineeth N Balasubramanian', 'Thrupthi Ann John'] | 2021-05-04 | null | null | null | null | ['face-model'] | ['computer-vision'] | [-1.15206661e-02 3.99488837e-01 9.00314376e-02 -6.56302154e-01
3.23102951e-01 -1.71445325e-01 5.20691037e-01 -9.71666202e-02
-6.90651610e-02 1.19048938e-01 1.77972406e-01 -1.82325378e-01
-1.23080961e-01 -5.32958388e-01 -4.70374167e-01 -5.04567087e-01
-8.31308514e-02 1.16607085e-01 2.24366784e-01 -3.55921328e-01
4.93774176e-01 8.68669927e-01 -1.87738717e+00 6.45267069e-01
3.08045566e-01 1.08001471e+00 2.35800311e-01 2.76945978e-01
-1.37035787e-01 8.82441223e-01 -8.08673859e-01 -4.65968490e-01
1.41417533e-01 -2.08926499e-01 -6.40421689e-01 -2.23706350e-01
8.14834714e-01 -3.40559602e-01 -1.76412277e-02 1.45448601e+00
3.98652285e-01 -3.30881000e-01 5.13980687e-01 -1.63335681e+00
-7.77402699e-01 2.73677528e-01 -5.10632038e-01 3.81320506e-01
2.10714459e-01 6.58113435e-02 8.18988919e-01 -1.32851243e+00
6.80428445e-01 1.81180954e+00 3.12657714e-01 8.09816837e-01
-1.25720382e+00 -7.19941437e-01 3.33937138e-01 3.51614982e-01
-1.01235986e+00 -7.13362515e-01 8.68849277e-01 -5.78813612e-01
7.27959573e-01 2.57508725e-01 6.25357807e-01 8.95993710e-01
1.93502530e-01 7.41729796e-01 1.21515810e+00 -3.05000365e-01
3.74578536e-01 5.82106113e-01 1.56031877e-01 6.53504491e-01
3.40917021e-01 2.14082688e-01 -9.79848027e-01 2.31587123e-02
8.15316141e-01 1.51863560e-01 -1.19613074e-01 -4.51382607e-01
-8.86940897e-01 6.63772345e-01 9.24864411e-01 3.31134796e-01
-5.68411708e-01 2.20084950e-01 1.95536301e-01 2.30762511e-01
3.85868579e-01 2.83597022e-01 -1.50615200e-01 4.13581312e-01
-1.29913414e+00 2.56746501e-01 3.86756212e-01 5.87435246e-01
7.76460886e-01 2.74607152e-01 -2.59236932e-01 6.39137030e-01
6.05855465e-01 3.48176628e-01 3.44391257e-01 -9.65386748e-01
-1.98945671e-01 7.56900966e-01 -3.82157564e-02 -1.46422470e+00
-4.81846362e-01 -2.35002786e-01 -4.83179718e-01 1.10077310e+00
1.75458938e-01 1.38396993e-01 -1.01429725e+00 1.80593026e+00
2.38708213e-01 -1.00308180e-01 -2.15753898e-01 1.17039227e+00
9.80655253e-01 2.97919095e-01 3.71490687e-01 2.52533823e-01
1.54016912e+00 -7.87615418e-01 -6.41148686e-01 -4.20984179e-01
2.07484692e-01 -6.32483423e-01 1.15502930e+00 2.12345749e-01
-9.94514048e-01 -6.75299287e-01 -1.05624616e+00 -1.19855469e-02
-4.96703714e-01 2.09436476e-01 3.36510211e-01 7.42892802e-01
-1.63822567e+00 5.92666388e-01 -5.03688395e-01 -6.95115566e-01
1.01252770e+00 3.77900302e-01 -2.91111857e-01 1.63787544e-01
-7.23320425e-01 1.29828465e+00 -5.53506985e-02 2.22390324e-01
-1.26868868e+00 -6.48502231e-01 -6.49997532e-01 4.93695021e-01
6.00693859e-02 -3.71608943e-01 1.28085780e+00 -1.51633823e+00
-9.89221811e-01 1.15454400e+00 -5.04963517e-01 -1.95905879e-01
2.69141376e-01 -1.73587017e-02 -4.34305519e-01 2.76740998e-01
8.24670270e-02 1.20697284e+00 1.14485919e+00 -1.53470814e+00
-5.77599227e-01 -7.05445826e-01 1.21398747e-01 -2.74850167e-02
-4.78005916e-01 5.88023901e-01 -1.62314698e-02 -5.96815050e-01
2.28714291e-02 -6.72218323e-01 -8.56255814e-02 6.60245240e-01
-2.69028902e-01 -1.76586941e-01 1.23475397e+00 -6.56345606e-01
1.12612307e+00 -2.05076647e+00 -5.18743992e-02 2.54470944e-01
5.97804189e-01 5.30967295e-01 -2.13936463e-01 1.56555563e-01
-3.10579240e-01 3.83311570e-01 8.21256172e-03 -2.38486066e-01
-2.94462200e-02 -2.69835442e-01 -2.44639128e-01 2.22609639e-01
5.81290722e-01 6.71956956e-01 -6.54979706e-01 -4.63473290e-01
1.44568250e-01 6.74500942e-01 -4.90721941e-01 1.02589749e-01
-2.29281764e-02 1.08252637e-01 -2.00140640e-01 7.38078296e-01
8.41295302e-01 -1.40286639e-01 1.40202910e-01 -2.33601511e-01
-2.13109050e-02 1.06678747e-01 -6.71228051e-01 1.21533859e+00
-9.72293764e-02 1.05648005e+00 3.67188454e-01 -5.35798192e-01
1.00199616e+00 1.90035343e-01 -5.24512976e-02 -9.76869404e-01
1.75598592e-01 -6.29490390e-02 2.50054032e-01 -3.30347449e-01
2.53350765e-01 -1.99942574e-01 5.06526172e-01 5.96646845e-01
2.06567660e-01 3.20315063e-01 -2.54482716e-01 2.87505805e-01
6.32764339e-01 1.21055404e-02 1.23962447e-01 -7.74495900e-01
4.51950252e-01 -1.77991122e-01 3.80347550e-01 2.97604501e-01
-5.82615256e-01 5.06903291e-01 9.91068065e-01 -8.72029424e-01
-8.27372611e-01 -7.74484158e-01 -5.06079681e-02 1.39447856e+00
6.27307668e-02 -2.12546602e-01 -1.21399486e+00 -6.78790927e-01
-3.17777954e-02 7.88539171e-01 -1.09253132e+00 -3.21090579e-01
-1.73386395e-01 -3.71793211e-01 6.69735447e-02 4.17689741e-01
3.61473769e-01 -1.60283482e+00 -1.22275960e+00 -2.69300848e-01
4.70496148e-01 -7.50866294e-01 -1.43431097e-01 -1.82014734e-01
-8.38890016e-01 -1.14428079e+00 -6.11208618e-01 -8.98760319e-01
1.05231607e+00 5.85764945e-01 1.25019908e+00 5.18297732e-01
-3.80029649e-01 1.74807221e-01 6.89989254e-02 -8.76342893e-01
-1.84456855e-01 -2.70749658e-01 2.71226019e-01 2.12187871e-01
6.90442562e-01 -3.73863548e-01 -8.28196645e-01 4.75825310e-01
-7.41131067e-01 1.09910637e-01 3.90401840e-01 5.05357087e-01
-7.64369369e-02 -4.30331647e-01 5.62458932e-01 -9.42570150e-01
8.80434871e-01 -2.77002841e-01 -5.43629169e-01 3.97590101e-01
-7.17735827e-01 -1.82644755e-01 2.24328414e-01 8.37920308e-02
-1.11162806e+00 -1.34550765e-01 1.61914721e-01 -6.47891223e-01
-2.44167954e-01 1.86654270e-01 -1.58676445e-01 -1.66193441e-01
8.88650179e-01 -8.55379775e-02 3.43864799e-01 -3.88502389e-01
1.66865453e-01 4.09981549e-01 2.96277553e-01 -2.46958897e-01
6.31679654e-01 5.52401304e-01 -5.39145879e-02 -7.98205256e-01
-5.96065521e-01 8.65773633e-02 -8.95313799e-01 -7.25808561e-01
6.49600744e-01 -5.95179260e-01 -1.02702701e+00 2.38063216e-01
-1.25242484e+00 -1.67938292e-01 -2.50045899e-02 -1.49448529e-01
-2.38813937e-01 -1.38705611e-01 -1.19032450e-01 -8.62502277e-01
-2.76561916e-01 -1.17475045e+00 9.78527844e-01 6.51044130e-01
-4.25801843e-01 -8.96105945e-01 -1.95529848e-01 -4.28815782e-02
7.01868713e-01 -9.95809510e-02 9.94292319e-01 -5.84504545e-01
-3.75351131e-01 9.74938273e-02 -6.84958994e-01 2.87173867e-01
-7.21625332e-03 3.08840662e-01 -1.65928102e+00 -2.59947002e-01
1.37880072e-01 -1.53846577e-01 6.25671625e-01 3.90398294e-01
1.35947967e+00 -4.60415810e-01 -4.51208562e-01 2.23572582e-01
1.14901710e+00 2.07559615e-01 7.36087024e-01 1.59770682e-01
4.79622751e-01 1.30930054e+00 5.88766694e-01 1.02316931e-01
3.88844684e-02 6.95930600e-01 7.68197656e-01 -5.30109882e-01
-1.97664648e-01 -3.02601188e-01 3.65527898e-01 -1.50601327e-01
-4.99089509e-02 2.21511513e-01 -1.07841897e+00 4.64276493e-01
-1.85532093e+00 -9.53864813e-01 1.09769374e-01 2.14660621e+00
2.77367949e-01 7.13639110e-02 2.14490861e-01 -1.56757921e-01
8.59004736e-01 1.06315345e-01 -5.76888323e-01 -6.62263751e-01
6.77429810e-02 1.25070572e-01 -1.53663740e-01 3.26824218e-01
-8.82664502e-01 1.17857230e+00 7.15783072e+00 4.76395428e-01
-1.55283856e+00 1.87283292e-01 1.10537326e+00 -3.49991173e-01
-3.07386130e-01 -1.39580831e-01 -6.31208420e-01 4.04983461e-01
9.90388393e-01 -1.29278451e-01 3.50965470e-01 1.24493957e+00
4.67690498e-01 -1.24108866e-01 -1.22888136e+00 1.10137558e+00
1.22940354e-01 -1.45765162e+00 2.82967180e-01 7.33441785e-02
3.13352317e-01 -2.60376841e-01 5.30556083e-01 -6.66824877e-02
1.48763433e-01 -1.37411213e+00 1.02826297e+00 5.85618973e-01
5.95538080e-01 -6.78244472e-01 3.88714492e-01 3.39937583e-02
-6.69615448e-01 -1.34049997e-01 -6.38616025e-01 -1.23335183e-01
-1.64166614e-01 2.24856153e-01 -1.09569538e+00 -4.07069117e-01
1.08773649e+00 4.29637909e-01 -9.64411616e-01 8.38852584e-01
-2.44366169e-01 3.28408122e-01 4.35401089e-02 -7.92590976e-02
1.93574019e-02 1.21763892e-01 3.19404602e-01 1.06808352e+00
1.91662431e-01 -2.07004875e-01 -4.49944615e-01 1.34849906e+00
1.00962110e-02 6.89006597e-02 -9.21859324e-01 9.43494737e-02
4.32581723e-01 1.57133532e+00 -8.99209201e-01 -2.93746650e-01
-3.18985105e-01 8.24296474e-01 3.81698221e-01 3.63706619e-01
-5.06224751e-01 -5.18317707e-02 1.20400822e+00 4.72585559e-01
9.08332169e-02 1.39050663e-01 -6.04480207e-01 -6.25838697e-01
-1.00534126e-01 -8.15719604e-01 7.73935169e-02 -1.11873031e+00
-8.90081644e-01 9.59068060e-01 -1.28958404e-01 -8.95019114e-01
-6.45572785e-03 -9.96220946e-01 -8.84746432e-01 9.58253026e-01
-1.41118717e+00 -1.09767294e+00 -3.76853496e-01 5.04160762e-01
3.16659331e-01 -4.77693558e-01 7.65322328e-01 6.40802970e-03
-6.73583865e-01 4.95696098e-01 -3.29908013e-01 7.04789683e-02
5.30201435e-01 -8.83827925e-01 4.19695884e-01 7.88956821e-01
2.67596245e-01 8.45598340e-01 7.82044113e-01 -3.54461700e-01
-8.92264724e-01 -8.12746286e-01 8.25757265e-01 -6.71701670e-01
1.98917285e-01 -5.63379943e-01 -9.67396855e-01 5.21736503e-01
3.37683797e-01 6.00180700e-02 5.53752780e-01 2.07103997e-01
-4.26708996e-01 -3.58627081e-01 -1.52885950e+00 7.66986609e-01
9.36451077e-01 -6.83091223e-01 -3.75351816e-01 1.91419825e-01
4.01820898e-01 1.39856219e-01 -2.02985600e-01 1.89967915e-01
6.51740253e-01 -1.57889640e+00 8.71357083e-01 -9.37871575e-01
4.52905923e-01 -3.18010449e-01 9.61395279e-02 -1.28483057e+00
-6.06468439e-01 -2.62502849e-01 7.66045377e-02 1.11345124e+00
4.51144606e-01 -5.70203304e-01 8.34406316e-01 9.75697041e-01
1.38896495e-01 -6.72721088e-01 -8.61307502e-01 -2.13809833e-01
-2.64551371e-01 -2.30596974e-01 7.41579413e-01 7.60365725e-01
2.80392002e-02 1.29319996e-01 -6.05847687e-02 1.99356064e-01
4.34268206e-01 -1.75609440e-01 5.93395948e-01 -1.42995417e+00
5.12832820e-01 -8.34454238e-01 -7.13770986e-01 -2.52578616e-01
5.84926344e-02 -6.25841737e-01 -2.49986455e-01 -1.24262416e+00
2.53200620e-01 -2.08601490e-01 -4.88864779e-01 7.68186092e-01
6.19192943e-02 4.37849641e-01 5.58483660e-01 2.60920912e-01
-2.55171269e-01 3.14448029e-01 9.70736384e-01 -4.76831980e-02
2.28631973e-01 -2.36498296e-01 -1.17095661e+00 9.12842929e-01
6.51114523e-01 -4.37078744e-01 -3.84323299e-01 -5.13957202e-01
1.11916654e-01 -5.26274860e-01 7.77265549e-01 -1.14740622e+00
2.26320952e-01 -5.42878546e-02 7.68537343e-01 -6.73215389e-02
2.95543730e-01 -9.60235715e-01 -6.01036921e-02 6.05642378e-01
-2.81526297e-01 1.15807429e-01 4.57044929e-01 2.97583401e-01
-2.93230593e-01 -1.06374510e-01 1.19289446e+00 -1.17450170e-01
-9.14535403e-01 1.03594951e-01 -3.42292011e-01 -5.62575579e-01
1.20007551e+00 -4.00864244e-01 -3.72674435e-01 -5.57277918e-01
-8.58455718e-01 7.40216523e-02 6.10559940e-01 6.89087570e-01
9.21431839e-01 -1.32707489e+00 -5.35673678e-01 5.54443955e-01
3.36298272e-02 -4.87871796e-01 2.45953560e-01 5.57658911e-01
-5.47468841e-01 5.35525262e-01 -8.36293757e-01 -7.47827053e-01
-1.55148745e+00 6.53636217e-01 5.03274143e-01 4.18291628e-01
-1.77456394e-01 1.02785075e+00 9.27523851e-01 -1.90981016e-01
2.19370231e-01 -2.05291837e-01 -5.66977680e-01 2.87730485e-01
9.46218014e-01 2.03228414e-01 1.79393232e-01 -9.55043554e-01
-6.30902231e-01 4.42302793e-01 -2.77133346e-01 2.28781961e-02
1.36796963e+00 1.10804819e-01 -3.85622144e-01 1.25814110e-01
7.89185464e-01 -5.02692997e-01 -1.35802555e+00 -6.88710138e-02
2.99638957e-02 -7.41298854e-01 1.77048519e-01 -1.01310623e+00
-1.45092845e+00 1.46675754e+00 1.00898254e+00 2.67455727e-01
1.25566888e+00 -1.70750003e-02 -1.08058505e-01 6.52195886e-02
3.70140523e-01 -9.82104182e-01 1.95406318e-01 1.74930081e-01
1.36895788e+00 -1.27745354e+00 -1.83358327e-01 -2.58830428e-01
-8.74345183e-01 1.18601847e+00 9.16419864e-01 1.49793352e-03
9.28347468e-01 -9.09277797e-02 3.42350036e-01 -7.83109069e-01
-8.43869925e-01 -5.59311686e-03 4.59596187e-01 7.48225272e-01
5.30044317e-01 8.30369815e-02 2.81183779e-01 4.92189020e-01
-2.06003070e-01 -7.10538700e-02 2.64125645e-01 7.18715906e-01
-7.17858255e-01 -8.09440136e-01 -5.07902682e-01 4.58467484e-01
-1.90975323e-01 1.88325159e-02 -8.17737639e-01 5.15080273e-01
3.97339016e-02 6.79169774e-01 2.39477664e-01 -3.73335332e-01
2.96889693e-01 2.22758546e-01 2.63338268e-01 -7.90013909e-01
-5.42562425e-01 -2.61069983e-01 -2.30082482e-01 -8.71626258e-01
-2.00432613e-01 -5.91941059e-01 -1.03546631e+00 -5.13713777e-01
1.50533184e-01 -1.81584790e-01 8.47989380e-01 5.93435884e-01
7.51416445e-01 3.66901129e-01 4.01420534e-01 -1.18746185e+00
-1.14976235e-01 -1.05115724e+00 -4.86970693e-01 2.88924187e-01
3.52171332e-01 -8.71789396e-01 -5.25526851e-02 -4.56400663e-02] | [10.25788688659668, 2.1486356258392334] |
2739dad8-9dea-4366-9c93-661f9c2e0d85 | neudf-leaning-neural-unsigned-distance-fields | 2304.1008 | null | https://arxiv.org/abs/2304.10080v1 | https://arxiv.org/pdf/2304.10080v1.pdf | NeUDF: Leaning Neural Unsigned Distance Fields with Volume Rendering | Multi-view shape reconstruction has achieved impressive progresses thanks to the latest advances in neural implicit surface rendering. However, existing methods based on signed distance function (SDF) are limited to closed surfaces, failing to reconstruct a wide range of real-world objects that contain open-surface structures. In this work, we introduce a new neural rendering framework, coded NeUDF, that can reconstruct surfaces with arbitrary topologies solely from multi-view supervision. To gain the flexibility of representing arbitrary surfaces, NeUDF leverages the unsigned distance function (UDF) as surface representation. While a naive extension of an SDF-based neural renderer cannot scale to UDF, we propose two new formulations of weight function specially tailored for UDF-based volume rendering. Furthermore, to cope with open surface rendering, where the in/out test is no longer valid, we present a dedicated normal regularization strategy to resolve the surface orientation ambiguity. We extensively evaluate our method over a number of challenging datasets, including DTU}, MGN, and Deep Fashion 3D. Experimental results demonstrate that nEudf can significantly outperform the state-of-the-art method in the task of multi-view surface reconstruction, especially for complex shapes with open boundaries. | ['Lin Gao', 'Bo Yang', 'Xiaoxu Meng', 'Weikai Chen', 'Jie Yang', 'Li Wang', 'Yu-Tao Liu'] | 2023-04-20 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Liu_NeUDF_Leaning_Neural_Unsigned_Distance_Fields_With_Volume_Rendering_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Liu_NeUDF_Leaning_Neural_Unsigned_Distance_Fields_With_Volume_Rendering_CVPR_2023_paper.pdf | cvpr-2023-1 | ['neural-rendering'] | ['computer-vision'] | [ 4.32744712e-01 4.73607294e-02 3.92838687e-01 -3.13813508e-01
-6.51257455e-01 -5.04439175e-01 5.60959756e-01 -2.89982319e-01
1.98947072e-01 6.12773836e-01 -6.99599162e-02 -2.17507020e-01
3.25569101e-02 -1.09769094e+00 -9.24771547e-01 -5.28389156e-01
7.50631914e-02 6.10148013e-01 1.88157186e-01 -3.53320867e-01
4.68184464e-02 7.65534401e-01 -1.73189425e+00 4.76187766e-01
8.31178546e-01 1.21643841e+00 -1.09083187e-02 3.73435467e-01
-3.06282073e-01 7.60050118e-02 -1.50074631e-01 -3.28545153e-01
5.19385040e-01 3.41503299e-03 -5.32204986e-01 -3.62422168e-02
1.02721202e+00 -5.56943595e-01 -9.62493271e-02 7.15008140e-01
6.60785198e-01 1.65664449e-01 7.62401938e-01 -8.77674162e-01
-7.74863601e-01 -2.55107462e-01 -5.79643548e-01 -2.14569375e-01
2.97934741e-01 -2.21438989e-01 9.37039375e-01 -1.31178045e+00
8.79060268e-01 1.32278264e+00 8.88000846e-01 6.47274494e-01
-1.35775971e+00 -4.68076617e-01 3.70399535e-01 -4.45711017e-01
-9.96101797e-01 -1.98145047e-01 1.13036692e+00 -4.29802060e-01
9.63533044e-01 4.76059109e-01 5.98228037e-01 1.06399095e+00
-2.89850831e-02 6.13758147e-01 1.28441787e+00 -3.10059600e-02
8.71192068e-02 -4.39096153e-01 -1.65640786e-01 7.32925177e-01
1.82142913e-01 1.07250720e-01 -4.01697248e-01 -5.27355909e-01
1.35780549e+00 -4.73652259e-02 -4.55596983e-01 -9.01639163e-01
-1.09662998e+00 8.41081917e-01 2.48372242e-01 1.51518977e-03
-1.72489524e-01 4.98793386e-02 4.10012901e-01 1.19255483e-01
1.14791584e+00 1.33610144e-01 -6.25208855e-01 6.03783503e-02
-7.03577995e-01 4.40203458e-01 8.65288615e-01 6.14161432e-01
7.33937442e-01 2.02054530e-01 2.79149152e-02 1.05080557e+00
2.67941922e-01 4.38414961e-01 -1.68436155e-01 -1.21171296e+00
4.72359657e-01 6.59485698e-01 7.82884061e-02 -9.50922012e-01
-4.08049315e-01 -4.39157814e-01 -1.02091849e+00 6.72428548e-01
2.84672052e-01 2.18443453e-01 -9.47906852e-01 1.32907176e+00
6.93352222e-01 2.82021672e-01 -2.06045881e-01 8.95263910e-01
1.11688471e+00 4.65111881e-01 -5.61359465e-01 6.88784346e-02
1.07664824e+00 -7.83100188e-01 -3.71951044e-01 2.48882659e-02
3.49740267e-01 -4.81401026e-01 1.18111336e+00 6.24994874e-01
-1.27226520e+00 -2.74306625e-01 -9.62050676e-01 -3.86149496e-01
-6.37121350e-02 -2.00587422e-01 8.54861438e-01 3.43070567e-01
-1.14396405e+00 8.81359875e-01 -8.86058748e-01 1.32544354e-01
7.14908540e-01 2.20694602e-01 -5.01676142e-01 -1.33334458e-01
-7.84630239e-01 2.95602977e-01 -2.13255137e-01 2.00945407e-01
-5.00207365e-01 -9.89709318e-01 -8.98494840e-01 -7.08330423e-02
2.77686507e-01 -1.17334628e+00 7.63604999e-01 -6.42130196e-01
-1.52850032e+00 1.16202497e+00 -1.11343004e-01 1.35228008e-01
7.99290359e-01 -8.70899931e-02 3.29934210e-02 5.03115803e-02
-4.41895500e-02 3.20579082e-01 8.73455822e-01 -1.85772657e+00
-7.98854232e-03 -7.17320979e-01 1.59923583e-01 2.40171298e-01
-2.39637308e-02 -4.56383735e-01 -2.09024534e-01 -7.83082664e-01
5.33419132e-01 -6.52302384e-01 -9.77400318e-02 6.58421159e-01
-3.46043229e-01 -2.46416166e-01 1.01657403e+00 -6.01903498e-01
8.33838046e-01 -2.14608097e+00 3.63055855e-01 2.79735476e-01
5.07131994e-01 1.32336959e-01 -7.11396486e-02 2.28931084e-01
6.20197654e-02 7.37922639e-02 -7.02698469e-01 -6.87729776e-01
5.16497530e-02 3.79782945e-01 -2.97058463e-01 4.79642123e-01
2.73433351e-03 7.98654199e-01 -6.39388800e-01 -9.68682393e-02
4.52629514e-02 1.03417003e+00 -8.41447532e-01 1.67269975e-01
-3.53478163e-01 7.27457166e-01 -5.31027794e-01 6.62720263e-01
1.04878247e+00 -5.21338463e-01 -5.27025200e-02 -1.56454563e-01
-1.17027096e-01 -3.00301961e-03 -1.10160446e+00 2.09262538e+00
-5.85796952e-01 2.49255031e-01 5.66481471e-01 -7.44670570e-01
1.04570699e+00 1.78918079e-01 5.63674331e-01 -4.99506116e-01
-7.16609210e-02 4.32399780e-01 -4.13056970e-01 -2.34176874e-01
2.74906635e-01 -3.33753526e-01 4.47309196e-01 2.88126379e-01
-3.70150506e-01 -2.95749605e-01 -2.59641230e-01 -1.36788100e-01
9.87075806e-01 6.86121523e-01 -7.77303949e-02 -1.98427871e-01
4.02860731e-01 -6.48101926e-01 4.90348816e-01 2.41298676e-01
2.08402470e-01 1.25517404e+00 4.15813506e-01 -7.02390432e-01
-1.02853000e+00 -1.33842874e+00 -5.29295146e-01 9.59856689e-01
1.20833321e-02 -1.82835221e-01 -6.00409567e-01 -5.91177583e-01
3.99575025e-01 3.53375137e-01 -7.41836190e-01 4.11244810e-01
-9.19546962e-01 -6.71723425e-01 2.90560097e-01 5.17456770e-01
2.34732121e-01 -9.54653263e-01 -6.54247344e-01 1.18696906e-01
3.84639092e-02 -1.11288357e+00 -3.03881347e-01 -1.66882381e-01
-1.18169737e+00 -1.09311950e+00 -1.01749194e+00 -5.43574572e-01
6.81318343e-01 2.74744987e-01 1.29499543e+00 2.77207464e-01
-1.23090960e-01 5.01673222e-01 7.73375258e-02 -2.75708595e-03
-3.03794771e-01 -1.73656806e-01 -2.89686382e-01 8.64066631e-02
-4.09022301e-01 -1.21573591e+00 -8.01285088e-01 3.03224325e-01
-9.95168388e-01 2.71383017e-01 7.52472952e-02 7.80931056e-01
1.01180422e+00 -5.84487498e-01 4.60286081e-01 -1.04052746e+00
3.35624814e-01 -3.78775179e-01 -4.31259006e-01 7.45167583e-02
-4.28508103e-01 -1.13383541e-03 5.55528343e-01 -2.32155964e-01
-1.00753379e+00 -2.51162201e-01 -4.97489721e-01 -7.93807626e-01
-1.28880352e-01 2.73852468e-01 -1.22332431e-01 -3.46652120e-01
4.26344633e-01 1.77440017e-01 1.82717144e-01 -8.59685063e-01
1.69864014e-01 2.11983204e-01 2.00266302e-01 -7.91303217e-01
5.37341475e-01 1.10189939e+00 4.06461149e-01 -9.47250605e-01
-8.84266555e-01 -1.22467197e-01 -6.08713686e-01 -6.02201857e-02
6.59078836e-01 -7.92543828e-01 -7.96519220e-01 6.37812078e-01
-1.39605677e+00 -4.19130117e-01 -3.26231986e-01 -8.53496939e-02
-8.93604398e-01 6.32655382e-01 -6.02902651e-01 -6.66143417e-01
-4.47785467e-01 -1.15323651e+00 1.60589218e+00 -3.29949677e-01
1.28152639e-01 -1.18658674e+00 3.53789292e-02 3.42259467e-01
5.10933995e-01 9.55272853e-01 1.02630877e+00 -1.67125016e-01
-6.85965896e-01 1.78338587e-01 -3.83464724e-01 3.64033848e-01
1.31374016e-01 -7.46975243e-02 -1.13685822e+00 -4.85185951e-01
1.87229291e-01 -2.61014104e-01 9.36596632e-01 3.61216098e-01
1.49634814e+00 -3.25637423e-02 -1.34484544e-01 1.29578900e+00
1.35487723e+00 -1.45556405e-01 5.80135703e-01 2.81851351e-01
1.17064619e+00 5.68922400e-01 4.27215211e-02 5.11825800e-01
5.71943760e-01 7.35328496e-01 9.03004766e-01 -1.78308651e-01
-2.11759716e-01 -1.69671088e-01 5.68322930e-03 8.13325226e-01
-5.24957776e-01 -1.85099110e-01 -7.90249586e-01 3.71948332e-02
-1.57010257e+00 -4.89679217e-01 -4.45323408e-01 2.17540050e+00
4.76501703e-01 2.76198778e-02 -1.33168906e-01 -1.91833242e-03
2.36559421e-01 4.83299196e-01 -7.21540511e-01 -3.15419257e-01
-3.19622576e-01 3.88646990e-01 5.26094101e-02 4.62972790e-01
-9.74236965e-01 6.31142735e-01 5.67463541e+00 7.58444846e-01
-1.28668904e+00 2.39630565e-01 4.69640523e-01 4.81007397e-02
-9.83517885e-01 -3.44193310e-01 -5.58886886e-01 5.37795648e-02
2.80021369e-01 4.39091325e-01 5.65022647e-01 5.71604431e-01
-8.37505981e-02 2.79405653e-01 -1.06612074e+00 1.07484579e+00
1.96457297e-01 -1.40026307e+00 2.74706841e-01 1.86766341e-01
7.30198562e-01 2.25573525e-01 1.19392283e-01 3.83294672e-02
-1.41658708e-01 -1.04840922e+00 6.58937752e-01 5.66379189e-01
1.29451776e+00 -4.33385819e-01 8.16499516e-02 4.19003308e-01
-1.26699078e+00 4.02215540e-01 -2.74318039e-01 -5.99362142e-02
3.56229126e-01 6.52873099e-01 -1.91166803e-01 8.30207884e-01
5.75351417e-01 9.01256621e-01 -1.64728299e-01 6.00449860e-01
2.14791745e-01 1.64870590e-01 -4.72392559e-01 6.00075305e-01
1.25919610e-01 -4.83492792e-01 7.75640726e-01 7.05866873e-01
4.20498103e-01 1.16864003e-01 1.03703588e-01 1.06579864e+00
-3.47802490e-02 1.34114087e-01 -8.83160472e-01 3.06157857e-01
-8.68400261e-02 1.10245287e+00 -7.23396480e-01 -1.52090071e-02
-4.36264008e-01 1.06145227e+00 6.91340148e-01 5.54600537e-01
-6.27665579e-01 2.61795044e-01 6.78454399e-01 5.79481959e-01
3.78888607e-01 -4.77296084e-01 -5.77964187e-01 -1.32695818e+00
4.70730722e-01 -5.76092482e-01 -1.73443109e-02 -7.19467580e-01
-1.56087840e+00 6.86260819e-01 -1.10257044e-01 -1.14048731e+00
1.89840645e-01 -7.81990469e-01 -4.24985439e-01 9.28349912e-01
-1.77887177e+00 -1.42692435e+00 -3.22220653e-01 3.96048129e-01
4.30590123e-01 1.17029734e-01 9.34995234e-01 4.17379618e-01
-6.58544675e-02 2.50308633e-01 6.12087138e-02 -2.04061866e-01
3.93816978e-01 -1.19652629e+00 7.37274647e-01 2.04152927e-01
-1.03773370e-01 5.67265391e-01 2.81929433e-01 -7.54859269e-01
-1.63160396e+00 -1.02567327e+00 2.35887364e-01 -5.05132914e-01
2.07667187e-01 -5.63232243e-01 -1.40170729e+00 5.85975528e-01
-2.89662778e-01 6.74900234e-01 5.67278743e-01 -1.54895745e-02
-4.85704660e-01 2.10663557e-01 -1.28580928e+00 2.53480285e-01
1.58606696e+00 -3.92512560e-01 -1.02593742e-01 2.07788140e-01
6.93587899e-01 -9.05989110e-01 -1.10025799e+00 8.61881077e-01
7.82530010e-01 -1.36331081e+00 1.45768416e+00 -5.84133089e-01
5.88956535e-01 -9.90619287e-02 -4.42460835e-01 -1.19929969e+00
-1.80271968e-01 -5.03374815e-01 -4.08189595e-01 8.27254415e-01
-1.20490545e-03 -9.29412186e-01 8.70095730e-01 3.41351807e-01
-5.79327404e-01 -1.49186277e+00 -1.29681814e+00 -5.81039548e-01
3.18784326e-01 -4.43204612e-01 6.41139686e-01 9.54319656e-01
-8.57776701e-01 -9.33315307e-02 -1.47907734e-01 1.08169846e-01
8.10050845e-01 3.96488607e-01 6.69536173e-01 -1.70645881e+00
-3.80447030e-01 -3.23295087e-01 -1.26618877e-01 -1.25446761e+00
3.43290806e-01 -1.12357116e+00 -2.91160315e-01 -1.81203234e+00
-1.93848416e-01 -8.14772666e-01 1.21413343e-01 1.72212765e-01
-1.36313690e-02 4.26498562e-01 7.90778399e-02 8.88941064e-02
-1.45286590e-01 9.10475552e-01 2.01292229e+00 1.36150450e-01
-6.80660037e-03 4.08869721e-02 -3.35777462e-01 1.10777795e+00
3.39369923e-01 -1.08038783e-01 -2.78860152e-01 -7.47005701e-01
5.50963819e-01 3.10172349e-01 7.13154197e-01 -5.57127297e-01
-2.33280420e-01 -8.10098369e-03 2.90135205e-01 -6.17714524e-01
7.77764678e-01 -8.25022817e-01 2.88879484e-01 -5.33053614e-02
1.85127035e-01 -2.76716746e-04 1.19158976e-01 6.77347183e-01
2.57047713e-02 2.68009901e-01 7.01187611e-01 -2.89394885e-01
-2.15431556e-01 7.51470089e-01 4.28315848e-01 2.86070406e-01
6.64590836e-01 -4.75159168e-01 -1.16432734e-01 -5.76766580e-02
-8.78474534e-01 -3.93229611e-02 8.16833973e-01 3.87386739e-01
9.15908098e-01 -1.46182048e+00 -6.56710505e-01 5.90688944e-01
-9.31660384e-02 7.67548680e-01 4.08161759e-01 7.21856236e-01
-5.77972233e-01 -9.26072672e-02 -4.02333541e-03 -7.75858819e-01
-8.96680355e-01 1.34918094e-01 4.52090889e-01 -3.27164412e-01
-1.23662758e+00 9.14802969e-01 8.06206048e-01 -8.03393483e-01
3.64582650e-02 -4.90523279e-01 3.60606168e-03 -6.80608526e-02
1.29664615e-01 4.86138254e-01 3.54792207e-01 -6.07876897e-01
-7.55913034e-02 9.59829569e-01 2.71084517e-01 1.72503561e-01
1.72497296e+00 1.90850183e-01 -2.52104640e-01 6.69130445e-01
1.22975326e+00 3.73689383e-02 -1.68450427e+00 6.32082447e-02
-5.00077248e-01 -5.67790329e-01 2.92727314e-02 -6.17109478e-01
-1.23653281e+00 1.09506774e+00 3.59099597e-01 1.06899649e-01
7.23806083e-01 -1.96707711e-01 1.23659706e+00 2.11356282e-02
7.56137908e-01 -5.45478284e-01 -5.85761853e-02 6.65749252e-01
1.39112926e+00 -1.14682829e+00 -4.06381227e-02 -8.48960698e-01
-1.38082519e-01 1.25132608e+00 4.48174447e-01 -4.03296769e-01
8.90852511e-01 4.20651019e-01 -2.19292894e-01 -6.04928792e-01
-3.89278144e-01 2.37700686e-01 5.27183950e-01 4.09567147e-01
5.10146081e-01 -1.89099740e-02 -2.23787827e-03 2.99591482e-01
-6.54397607e-02 -6.96315542e-02 2.68667191e-01 9.12099242e-01
-1.21813364e-01 -1.05571294e+00 -2.37179518e-01 6.61828339e-01
-4.50391024e-01 1.74196865e-02 -1.21145606e-01 6.87953174e-01
-5.07138018e-03 1.39729455e-01 2.83474207e-01 5.84462956e-02
6.25436485e-01 -7.95453638e-02 7.48311758e-01 -6.67082727e-01
-4.94988233e-01 1.32249936e-01 2.47777645e-02 -8.08426857e-01
-5.04571140e-01 -6.85077310e-01 -1.30812418e+00 -1.44664049e-01
-1.15083747e-01 -4.34417218e-01 5.99035084e-01 7.59386182e-01
4.83354151e-01 5.73402524e-01 4.73710805e-01 -1.46328962e+00
-5.33988595e-01 -5.08637965e-01 -5.90644896e-01 5.04116237e-01
5.62803507e-01 -1.08955073e+00 -4.62227851e-01 -3.40856165e-01] | [8.996253967285156, -3.3135502338409424] |
e5e69b6f-c575-4e3e-b6ef-4a191187b5ac | tanet-transformer-based-asymmetric-network | 2207.01172 | null | https://arxiv.org/abs/2207.01172v1 | https://arxiv.org/pdf/2207.01172v1.pdf | TANet: Transformer-based Asymmetric Network for RGB-D Salient Object Detection | Existing RGB-D SOD methods mainly rely on a symmetric two-stream CNN-based network to extract RGB and depth channel features separately. However, there are two problems with the symmetric conventional network structure: first, the ability of CNN in learning global contexts is limited; second, the symmetric two-stream structure ignores the inherent differences between modalities. In this paper, we propose a Transformer-based asymmetric network (TANet) to tackle the issues mentioned above. We employ the powerful feature extraction capability of Transformer (PVTv2) to extract global semantic information from RGB data and design a lightweight CNN backbone (LWDepthNet) to extract spatial structure information from depth data without pre-training. The asymmetric hybrid encoder (AHE) effectively reduces the number of parameters in the model while increasing speed without sacrificing performance. Then, we design a cross-modal feature fusion module (CMFFM), which enhances and fuses RGB and depth features with each other. Finally, we add edge prediction as an auxiliary task and propose an edge enhancement module (EEM) to generate sharper contours. Extensive experiments demonstrate that our method achieves superior performance over 14 state-of-the-art RGB-D methods on six public datasets. Our code will be released at https://github.com/lc012463/TANet. | ['Yutao Wang', 'Yunhua Zhang', 'Hangxu Wang', 'Shuo Wang', 'Gang Yang', 'Chang Liu'] | 2022-07-04 | null | null | null | null | ['rgb-d-salient-object-detection'] | ['computer-vision'] | [ 6.48019537e-02 -1.20179709e-02 -8.84661525e-02 -4.05360311e-01
-6.52305126e-01 -1.26208216e-01 2.91912884e-01 -3.97005916e-01
-4.42446291e-01 3.93093318e-01 2.32513681e-01 -2.79772133e-01
1.79637492e-01 -1.05505192e+00 -4.56186533e-01 -7.66931653e-01
3.22319895e-01 -3.96762967e-01 6.24988019e-01 -2.19766632e-01
9.13847163e-02 6.17695570e-01 -1.48459411e+00 3.31088156e-01
8.83182824e-01 1.72078252e+00 2.11289182e-01 4.20789838e-01
-3.63597631e-01 8.09904575e-01 -9.42426547e-02 -4.37540829e-01
5.26680887e-01 -3.82214159e-01 -5.94421148e-01 6.50291145e-02
2.60305971e-01 -7.76880622e-01 -7.84103096e-01 9.05917108e-01
9.06043291e-01 -8.03562030e-02 2.48476699e-01 -1.29787362e+00
-5.01931369e-01 9.49079692e-02 -8.38551044e-01 3.94982658e-02
2.70677060e-02 1.92799523e-01 5.67145109e-01 -1.07097363e+00
4.79597658e-01 1.03354788e+00 8.92128229e-01 5.66444337e-01
-6.38609409e-01 -6.49316728e-01 5.57302721e-02 2.29664207e-01
-1.34290457e+00 -3.75369012e-01 1.17195106e+00 1.07759111e-01
8.32398474e-01 6.00997433e-02 9.01663423e-01 7.58103669e-01
-2.61054002e-03 1.04961395e+00 1.05633879e+00 -8.59558731e-02
8.18324387e-02 -3.17834139e-01 -3.52357537e-01 1.00141168e+00
-1.43459514e-01 2.25203037e-01 -6.70963943e-01 3.51608098e-01
1.11514223e+00 2.49747440e-01 -5.90438843e-01 -3.90337706e-01
-9.13498998e-01 6.73884034e-01 8.94728065e-01 2.62153059e-01
-3.61155719e-01 2.19603792e-01 2.90983588e-01 1.44252613e-01
5.52656531e-01 -2.84509778e-01 -5.60450673e-01 -1.04551591e-01
-8.08729529e-01 -7.16109425e-02 3.13125551e-01 9.78101373e-01
1.06369877e+00 -1.64775506e-01 4.48058918e-02 7.55831480e-01
3.79783124e-01 4.07592475e-01 4.39871818e-01 -9.95462894e-01
5.58822751e-01 8.55929554e-01 -2.94259548e-01 -8.79863262e-01
-5.82031250e-01 -2.72571981e-01 -9.79151785e-01 2.32956395e-01
2.04420790e-01 -1.80082500e-01 -1.12944436e+00 1.40075409e+00
4.18544084e-01 2.39354447e-01 6.43237308e-02 1.16106880e+00
1.24926162e+00 5.55312634e-01 -1.28546387e-01 2.15711832e-01
1.22339153e+00 -1.12942600e+00 -5.27223945e-01 -2.27286935e-01
4.73821908e-01 -7.19878674e-01 1.05786419e+00 1.96621388e-01
-1.19551039e+00 -4.72223967e-01 -1.04554188e+00 -6.67051315e-01
-4.58615482e-01 2.41518244e-01 7.78875709e-01 3.37850302e-01
-1.06285203e+00 3.69252652e-01 -9.32985425e-01 -2.70911418e-02
6.07683182e-01 4.57665294e-01 -3.72301191e-01 -2.63260663e-01
-1.16931844e+00 5.64431250e-01 3.88907820e-01 4.64074135e-01
-3.91278535e-01 -7.78126180e-01 -1.13518250e+00 -4.62568142e-02
2.60247946e-01 -7.67443359e-01 1.20786011e+00 -9.27490890e-01
-1.70426965e+00 6.04987681e-01 -1.08077794e-01 1.17935658e-01
5.49523890e-01 -2.19995417e-02 -2.76182771e-01 4.53680754e-01
-1.05107851e-01 9.28124011e-01 6.05163097e-01 -1.26297843e+00
-9.54431415e-01 -4.23111200e-01 4.80838642e-02 3.29638481e-01
-4.30267096e-01 -2.50328392e-01 -9.27872598e-01 -7.66948283e-01
5.70121288e-01 -5.65512359e-01 -4.93747145e-02 5.37833214e-01
-4.53799486e-01 -9.14545357e-02 1.18142331e+00 -5.59438348e-01
1.12563777e+00 -2.26386714e+00 2.68169884e-02 1.88651815e-01
4.70834851e-01 2.65993506e-01 -1.95172444e-01 1.11166798e-01
1.79725140e-01 -9.18126926e-02 -4.50780183e-01 -6.60261571e-01
-1.30407169e-01 1.29880577e-01 5.86938001e-02 3.19622129e-01
2.68546164e-01 1.19117427e+00 -7.93239772e-01 -6.96093202e-01
4.52994108e-01 8.57931554e-01 -5.21452010e-01 1.63703218e-01
1.32733941e-01 2.79186159e-01 -5.19798458e-01 1.00011194e+00
1.08216631e+00 -2.26825818e-01 -1.88409597e-01 -7.53765821e-01
-3.46293122e-01 2.04738870e-01 -1.07892644e+00 2.14839602e+00
-5.79361916e-01 3.56438965e-01 -2.77046524e-02 -7.60714948e-01
8.46630335e-01 1.33540273e-01 6.12863541e-01 -1.00169981e+00
4.77122068e-01 3.44480127e-01 -3.12896609e-01 -5.15074968e-01
2.72738904e-01 -1.61050990e-01 5.47492243e-02 2.40855157e-01
1.25428170e-01 -1.13466084e-01 -1.83572322e-01 -6.93925284e-03
9.44070458e-01 2.85605311e-01 -7.00410008e-02 2.04203442e-01
5.28900027e-01 -2.78992832e-01 9.09803212e-01 1.50718674e-01
-3.04593533e-01 8.78949881e-01 3.16285908e-01 -4.47619647e-01
-8.06430578e-01 -1.15241981e+00 4.73216176e-02 5.69538653e-01
6.26913786e-01 -4.64908272e-01 -3.76573175e-01 -8.15551400e-01
-8.67872611e-02 1.02693208e-01 -6.15415454e-01 -1.88978314e-01
-5.84855795e-01 -5.58050931e-01 5.38572669e-01 9.40811038e-01
1.34506178e+00 -7.84682512e-01 -8.17834735e-01 6.68621957e-02
-3.58825862e-01 -1.23659313e+00 -5.53580761e-01 1.30740806e-01
-9.89556909e-01 -9.07351553e-01 -8.91177773e-01 -7.88130701e-01
5.78456759e-01 5.41039765e-01 6.66546106e-01 1.37811795e-01
-1.53127268e-01 1.16168961e-01 -4.68351364e-01 -2.28209391e-01
2.74466127e-01 1.74328282e-01 -5.20150900e-01 -2.48198994e-02
3.23171109e-01 -5.96945286e-01 -1.12141645e+00 1.36682898e-01
-1.08275676e+00 5.42447925e-01 8.53911877e-01 7.42481411e-01
5.82514763e-01 6.77918792e-02 2.03249156e-01 -4.21747833e-01
1.92014575e-01 -2.68131524e-01 -2.96123415e-01 1.71797588e-01
-4.41485018e-01 -8.73502418e-02 4.26206201e-01 -1.63564563e-01
-1.31770611e+00 2.16675282e-01 -3.77904415e-01 -5.15427351e-01
6.44561574e-02 4.24102336e-01 -3.65911275e-01 -2.54319996e-01
-1.99553166e-02 3.92372370e-01 4.04378772e-02 -5.49246490e-01
2.41864547e-01 6.86912537e-01 5.48804760e-01 -2.69694477e-01
7.36715436e-01 6.82173431e-01 5.24052456e-02 -5.47530472e-01
-8.83192599e-01 -3.15858513e-01 -5.75361907e-01 -2.57247895e-01
8.44762683e-01 -9.86455262e-01 -7.92791069e-01 9.29843545e-01
-1.00434351e+00 -5.60090244e-01 -1.03763565e-01 4.63304758e-01
-4.51287061e-01 3.71687770e-01 -8.16676915e-01 -4.15268213e-01
-5.23135006e-01 -1.11859238e+00 1.28700340e+00 6.32297158e-01
5.03149807e-01 -8.78662765e-01 -2.75786042e-01 3.16441983e-01
6.50705457e-01 3.63896489e-01 4.84470546e-01 1.83218360e-01
-7.52582908e-01 5.42983934e-02 -7.65279293e-01 4.50376242e-01
2.66387969e-01 -1.50279149e-01 -1.09977651e+00 -5.67525290e-02
-4.86243106e-02 -3.62475634e-01 1.01557612e+00 2.86282212e-01
1.52403975e+00 -8.83931573e-03 -1.67397335e-01 1.17307413e+00
1.64649999e+00 4.08607461e-02 8.56263518e-01 4.61273193e-01
9.85884190e-01 2.98210502e-01 4.48920816e-01 4.35067564e-01
7.69260705e-01 4.77056831e-01 5.17901897e-01 -5.96555471e-01
-4.71437573e-01 -3.67397040e-01 2.77395755e-01 7.69254327e-01
-2.10324943e-01 -1.41123146e-01 -7.79646039e-01 3.21234941e-01
-1.68143559e+00 -5.67052364e-01 -2.10659076e-02 1.71176398e+00
8.58743250e-01 9.76101384e-02 -5.04421666e-02 2.68799245e-01
3.75355989e-01 4.31470685e-02 -6.37602746e-01 3.64381634e-02
-3.05548996e-01 3.57979387e-01 7.31445074e-01 2.79817194e-01
-1.11453569e+00 8.51833582e-01 4.91304159e+00 8.50831151e-01
-1.45905018e+00 1.13350749e-01 6.00399554e-01 -1.04103841e-01
-3.66849363e-01 -8.67772922e-02 -4.94529456e-01 4.56573784e-01
3.52345437e-01 2.46776670e-01 7.13443160e-02 5.18096030e-01
6.73996210e-02 -2.21629828e-01 -6.50920331e-01 1.18092251e+00
3.15063335e-02 -1.28711748e+00 -1.28815621e-01 -3.14927734e-02
6.05902135e-01 2.04236314e-01 -3.43409143e-02 -1.46315610e-02
-1.55593455e-01 -7.17880011e-01 8.37881744e-01 5.33802330e-01
9.51381922e-01 -8.49129856e-01 9.10825670e-01 2.95183454e-02
-1.55160475e+00 -1.07603863e-01 -2.98909783e-01 2.03092664e-01
1.81560203e-01 7.35993683e-01 -1.23599067e-01 7.50670910e-01
1.01917934e+00 9.40070748e-01 -5.24734855e-01 1.06676412e+00
-3.96184206e-01 1.05706371e-01 -5.21174431e-01 3.14178824e-01
4.65330005e-01 7.88929239e-02 1.09289102e-01 1.00130951e+00
4.88885105e-01 3.22914302e-01 -6.61112815e-02 7.36662686e-01
-2.06101298e-01 -1.34116173e-01 -3.49169999e-01 3.93771321e-01
5.19177854e-01 1.46576083e+00 -7.91527271e-01 -1.61713734e-01
-8.33371341e-01 1.33700407e+00 2.04400688e-01 2.57354409e-01
-8.34599793e-01 -7.12945402e-01 4.99949455e-01 -1.99891441e-03
5.10707617e-01 -2.03076124e-01 -3.27560633e-01 -1.32036746e+00
2.02434480e-01 -5.27831197e-01 3.22744131e-01 -9.34031188e-01
-1.22646606e+00 5.85924149e-01 -3.34977120e-01 -1.46692955e+00
3.78097415e-01 -7.24913001e-01 -6.48384750e-01 7.60282636e-01
-2.10133052e+00 -1.54419553e+00 -9.42914069e-01 9.90942478e-01
3.41841906e-01 3.75351012e-01 3.23066682e-01 6.06091976e-01
-6.77365899e-01 7.15892315e-01 -3.66434276e-01 4.47953105e-01
6.34183943e-01 -1.14424717e+00 2.23632529e-01 7.55406439e-01
-5.80177844e-01 2.76425630e-01 -3.70652452e-02 -4.29883450e-01
-1.49335527e+00 -1.19041824e+00 7.24724412e-01 1.65462077e-01
2.90009618e-01 -2.82234430e-01 -7.23095596e-01 4.55785990e-01
-1.26445964e-02 4.95347261e-01 5.03362775e-01 -4.90496874e-01
-3.19430411e-01 -3.08977008e-01 -1.20013964e+00 4.76147711e-01
1.23617375e+00 -5.39930105e-01 -2.59072989e-01 -2.44280979e-01
7.93746054e-01 -6.41173780e-01 -1.18910503e+00 7.09306240e-01
6.01648808e-01 -1.42557192e+00 1.02145410e+00 1.31307647e-01
7.85068989e-01 -5.61264157e-01 -2.51938313e-01 -1.11742890e+00
-1.82756055e-02 -3.05192888e-01 -2.89360553e-01 1.23981893e+00
1.56909809e-01 -7.30373323e-01 9.43328381e-01 5.62931240e-01
-4.66709316e-01 -1.29249465e+00 -8.70064437e-01 -5.40562212e-01
-1.08943656e-01 -4.99765456e-01 8.36882234e-01 8.71617496e-01
-2.67714471e-01 6.08955212e-02 -1.17632352e-01 2.46111993e-02
5.84054053e-01 2.03279227e-01 5.85588038e-01 -8.14357281e-01
-6.76969662e-02 -5.46850860e-01 -3.61431837e-01 -1.34743631e+00
-2.76315868e-01 -7.58138239e-01 7.08156750e-02 -1.66140521e+00
4.78479527e-02 -7.43104637e-01 -3.41516167e-01 7.52129257e-01
-1.24679454e-01 5.90597987e-01 2.64813572e-01 1.71690974e-02
-4.28936154e-01 1.00453115e+00 1.62751245e+00 1.21405043e-01
-1.97355047e-01 -2.39686742e-01 -7.22009361e-01 8.53943348e-01
9.45186257e-01 -2.06663668e-01 -4.80181068e-01 -6.95921183e-01
4.77703437e-02 4.67720218e-02 6.37242615e-01 -1.07089710e+00
4.18814600e-01 4.08986956e-02 7.90758789e-01 -7.97352612e-01
3.80417585e-01 -8.28130484e-01 -3.01125348e-01 3.40591609e-01
1.46625519e-01 1.95362687e-01 2.22960904e-01 2.83411145e-01
-4.44061846e-01 2.37510294e-01 7.18166709e-01 -3.02312616e-02
-1.07652998e+00 8.35286856e-01 8.64829943e-02 -1.05236180e-01
8.84742022e-01 -3.70464534e-01 -4.26119417e-01 -2.30721176e-01
-4.21523273e-01 2.89214611e-01 6.63686335e-01 3.46802384e-01
1.03738856e+00 -1.54063261e+00 -3.15216482e-01 3.63747835e-01
1.55426413e-01 5.48618257e-01 5.93356073e-01 1.06683528e+00
-7.02615142e-01 1.70304105e-02 -2.97929078e-01 -4.48581010e-01
-8.59117508e-01 1.33019865e-01 5.35593808e-01 -4.63739410e-02
-8.73952270e-01 9.80493009e-01 8.68819281e-02 -3.68934304e-01
2.44214728e-01 -2.94014812e-01 1.76753014e-01 -8.97745341e-02
3.74174744e-01 3.20489436e-01 1.72089159e-01 -4.60446358e-01
-4.02674705e-01 8.32808316e-01 9.45592374e-02 6.34882739e-03
1.48162150e+00 -2.83750474e-01 -1.26852676e-01 1.05087705e-01
1.63480914e+00 -2.94529021e-01 -1.64186418e+00 -4.35929030e-01
-4.52855200e-01 -6.32851124e-01 6.49820685e-01 -6.75540268e-01
-1.83305287e+00 1.03435409e+00 7.46380389e-01 -2.45181337e-01
1.86708081e+00 -1.34450691e-02 1.47111464e+00 -8.98943655e-03
1.00952484e-01 -1.02717936e+00 1.98540568e-01 3.06756109e-01
6.04432881e-01 -1.23327231e+00 2.06364095e-02 -6.07743204e-01
-5.06388068e-01 1.31654942e+00 8.87491107e-01 -2.34922040e-02
8.98920715e-01 4.11797553e-01 1.14127077e-01 -3.29068631e-01
-3.43081534e-01 -4.32700545e-01 2.76689410e-01 4.25943196e-01
3.23253363e-01 -2.39748791e-01 -9.56530422e-02 5.71257412e-01
-4.11997456e-03 2.87604153e-01 2.73562998e-01 1.31491721e+00
-2.40759507e-01 -1.10899448e+00 -2.51111053e-02 2.74940819e-01
-2.31308997e-01 -1.92973927e-01 -9.99318957e-02 9.22480106e-01
5.66855133e-01 7.26509511e-01 1.73241556e-01 -6.97885811e-01
3.06601077e-01 -2.63622850e-01 5.40401459e-01 -1.15980372e-01
-3.17607045e-01 1.18483573e-01 -2.17952281e-01 -9.08342123e-01
-7.04765439e-01 -4.34872448e-01 -1.45652211e+00 -5.25245786e-01
-2.09714413e-01 -3.94206941e-01 6.72862768e-01 7.58466482e-01
5.23209751e-01 6.11203432e-01 7.21290410e-01 -1.02315962e+00
-1.12399988e-01 -7.21302927e-01 -4.64792490e-01 1.46982029e-01
4.38595355e-01 -7.96084762e-01 -1.39736339e-01 -1.32984281e-01] | [9.584208488464355, -0.9389311671257019] |
09ce1cd1-bfaf-4ccd-ab32-a25450bd9420 | lingvo-a-modular-and-scalable-framework-for | 1902.08295 | null | http://arxiv.org/abs/1902.08295v1 | http://arxiv.org/pdf/1902.08295v1.pdf | Lingvo: a Modular and Scalable Framework for Sequence-to-Sequence Modeling | Lingvo is a Tensorflow framework offering a complete solution for
collaborative deep learning research, with a particular focus towards
sequence-to-sequence models. Lingvo models are composed of modular building
blocks that are flexible and easily extensible, and experiment configurations
are centralized and highly customizable. Distributed training and quantized
inference are supported directly within the framework, and it contains existing
implementations of a large number of utilities, helper functions, and the
newest research ideas. Lingvo has been used in collaboration by dozens of
researchers in more than 20 papers over the last two years. This document
outlines the underlying design of Lingvo and serves as an introduction to the
various pieces of the framework, while also offering examples of advanced
features that showcase the capabilities of the framework. | ['Pat Rondon', 'William Chan', 'Ian McGraw', 'Semih Yavuz', 'Rob Suderman', 'David Rybach', 'Colin Raffel', 'Antoine Bruguier', 'Ben Vanik', 'Kuan-Chieh Wang', 'Wei-Ning Hsu', 'Rohan Anil', 'Sébastien Jean', 'Ciprian Chelba', 'Rohit Prabhavalkar', 'Jan Chorowski', 'Chung-Cheng Chiu', 'Anjuli Kannan', 'Yonghui Wu', 'Jonathan Shen', 'Zelin Wu', 'Ye Jia', 'Xiaobing Liu', 'Uri Alon', 'Todd Wang', 'Stella Laurenzo', 'Shankar Kumar', 'Sara Sabour', 'Ron J. Weiss', 'Naveen Ari', 'Michiel A. U. Bacchiani', 'Llion Jones', 'Kazuki Irie', 'Katrin Tomanek', 'John Richardson', 'Isaac Caswell', 'Ian Williams', 'Ekaterina Gonina', 'Bowen Liang', 'Bo Li', 'Ankur Bapna', 'Zongheng Yang', 'Youlong Cheng', 'Ye Tian', 'Thomas B. Jablin', 'Suyog Gupta', 'Qi Ge', 'Navdeep Jaitly', 'Mike Schuster', 'Max Galkin', 'Matthew Murray', 'Kanishka Rao', 'Justin Carlson', 'Golan Pundak', 'Dehao Chen', 'Baohua Liao', 'Tara Sainath', 'Shuyuan Zhang', 'Shubham Toshniwal', 'Ruoming Pang', 'Rajat Tibrewal', 'Qiao Liang', 'Patrick Nguyen', 'Parisa Haghani', 'Orhan Firat', 'Melvin Johnson', 'Deepti Bhatia', 'Colin Cherry', 'Benoit Jacob', 'Zhifeng Chen', 'Yu Zhang', 'Yuan Cao', 'Yanzhang He', 'Yanping Huang', 'Wolfgang Macherey', 'Vijayaditya Peddinti', 'Smit Hinsu', 'Raziel Alvarez', 'Otavio Good', 'Michael Nirschl', 'Mia X. Chen', 'Maxim Krikun', 'Klaus Macherey', 'James Qin', 'HyoukJoong Lee', 'Heiga Zen', 'George Foster', 'Dmitry Lepikhin', 'Chad Whipkey', 'Benjamin Lee', 'Akiko Eriguchi'] | 2019-02-21 | null | null | null | null | ['sequence-to-sequence-speech-recognition'] | ['speech'] | [-7.00434268e-01 -3.00814509e-01 -4.95095104e-01 -7.47082889e-01
-6.48712933e-01 -7.14799047e-01 6.30936146e-01 -3.49632055e-01
-1.98635668e-01 7.53234863e-01 5.03938556e-01 -4.62112039e-01
-1.56410411e-01 -4.81084198e-01 -2.84282327e-01 -5.74803948e-01
-4.12629217e-01 4.88098830e-01 3.77429016e-02 -3.23022813e-01
1.80824772e-01 4.70872879e-01 -1.60182691e+00 6.37543857e-01
2.26352513e-01 8.93864930e-01 2.01575845e-01 1.18895829e+00
-3.52222532e-01 1.30777836e+00 -8.52455676e-01 -6.61016107e-01
2.36275308e-02 -6.79941475e-02 -1.03796911e+00 -1.33951485e-01
5.35533607e-01 -6.31154776e-01 -2.91940391e-01 6.45761907e-01
8.88427317e-01 2.10985273e-01 1.57146826e-01 -1.52653766e+00
-7.26859212e-01 8.55095446e-01 1.69698149e-01 6.04902208e-01
1.71522394e-01 5.02428591e-01 1.63631833e+00 -7.69896030e-01
7.67858684e-01 1.09415436e+00 1.02142155e+00 5.90575337e-01
-1.00501704e+00 -4.04271722e-01 1.49631232e-01 4.60189462e-01
-1.17955649e+00 -6.93566501e-01 3.27821374e-01 -5.58917880e-01
1.68746090e+00 4.60488677e-01 6.47520185e-01 1.44129443e+00
2.99643308e-01 1.40999866e+00 5.73630393e-01 -1.71765804e-01
4.25121725e-01 1.04829622e-02 2.78351426e-01 8.18313062e-01
-1.85332701e-01 -7.88270459e-02 -5.23959219e-01 -4.61310893e-01
7.60594904e-01 1.25671700e-01 7.65677020e-02 -2.02943742e-01
-1.07939553e+00 8.46109807e-01 2.65448689e-01 5.62889874e-01
-1.42084703e-01 4.03597474e-01 9.39119518e-01 3.39834064e-01
4.88919169e-01 3.28344852e-01 -7.71603823e-01 -8.46362174e-01
-1.01252544e+00 8.16881418e-01 1.03576779e+00 1.08565509e+00
4.99758691e-01 3.42448562e-01 -3.91827852e-01 8.72153401e-01
3.30365300e-01 -2.33390108e-01 6.91504061e-01 -1.47154284e+00
1.55880928e-01 2.19736621e-02 1.12982169e-01 -5.54615974e-01
-3.06473881e-01 -5.46758413e-01 -5.49124360e-01 6.76451325e-02
1.41298445e-02 -4.10204381e-01 -4.18370366e-01 1.23100376e+00
2.73578674e-01 2.63976306e-01 -2.19233215e-01 9.47763681e-01
1.24853659e+00 4.81001109e-01 8.06898400e-02 2.23724365e-01
1.04267299e+00 -1.32553196e+00 -8.17467749e-01 -5.38663678e-02
8.59794080e-01 -8.02339733e-01 1.01962960e+00 5.20709753e-01
-1.45033133e+00 -5.26435614e-01 -8.64409685e-01 -5.82532048e-01
-3.23738009e-01 -1.33633152e-01 1.42899871e+00 7.20179021e-01
-1.49104643e+00 8.24132800e-01 -1.04673076e+00 -1.81490749e-01
4.03849155e-01 8.82280245e-03 9.63997189e-03 1.34201720e-01
-1.19846094e+00 6.35586500e-01 2.71328270e-01 -3.38039221e-03
-1.08024585e+00 -7.11400747e-01 -5.03059208e-01 2.51610637e-01
8.48005563e-02 -9.14427340e-01 1.92938828e+00 -5.41796923e-01
-1.70825315e+00 5.56438625e-01 -2.35910460e-01 -4.51597631e-01
6.92753136e-01 -3.32770318e-01 -3.60817015e-01 -1.30090758e-01
-1.30890477e-02 3.82632643e-01 3.00186068e-01 -5.23875058e-01
-6.73735797e-01 -1.38278559e-01 2.34386563e-01 9.59395170e-02
-3.01987827e-01 4.44188237e-01 -7.10891545e-01 -7.29918957e-01
-6.07863069e-01 -4.34987485e-01 -4.81210649e-01 -9.86938849e-02
-1.80214927e-01 -6.14524484e-01 6.93650961e-01 -4.62583065e-01
1.37983358e+00 -1.87837648e+00 -1.06913261e-01 -3.93417060e-01
4.92059737e-01 2.82554984e-01 -1.50342822e-01 8.95367563e-01
1.34968430e-01 3.37141603e-01 2.24506527e-01 -6.39267385e-01
4.55998540e-01 4.24259365e-01 -2.49914035e-01 2.98751473e-01
-2.12310538e-01 1.22604096e+00 -1.08491302e+00 -3.28436643e-01
2.48338640e-01 3.15174967e-01 -5.63714802e-01 4.68652308e-01
-4.32272106e-01 5.12615107e-02 -3.01518559e-01 6.80046856e-01
4.58316892e-01 -4.97168452e-01 4.50613886e-01 1.76974803e-01
-3.15763652e-01 7.20609188e-01 -1.09886539e+00 1.69135356e+00
-3.14021170e-01 9.70134258e-01 3.32270175e-01 -8.44321549e-01
7.31117964e-01 6.65547550e-01 4.99556094e-01 6.10507384e-04
6.79863840e-02 7.23231286e-02 -2.32486799e-01 -7.98992395e-01
6.31640077e-01 1.54601261e-01 3.05322170e-01 7.18792856e-01
6.68391526e-01 3.70874293e-02 3.97258848e-01 3.44761759e-01
1.11410367e+00 1.72492713e-01 4.26478267e-01 2.04920042e-02
1.83090165e-01 -2.18596220e-01 5.32959342e-01 1.01960897e+00
-5.33924997e-01 9.67840403e-02 4.72577423e-01 -8.29789937e-01
-1.38748133e+00 -1.03257728e+00 -9.28095281e-02 1.67990017e+00
-6.51910007e-01 -9.92835939e-01 -5.17255068e-01 -6.40920818e-01
7.68465921e-02 4.50249255e-01 -5.81699610e-01 4.62686419e-01
-1.17277928e-01 -6.79496408e-01 8.42495680e-01 1.03467321e+00
4.87637997e-01 -1.00913429e+00 -2.50166804e-01 1.70053199e-01
-1.49858162e-01 -9.91826594e-01 -2.04639539e-01 1.38971403e-01
-1.01445961e+00 -7.14565277e-01 -4.65394855e-01 -6.73940599e-01
-3.59747082e-01 1.55635878e-01 1.60367811e+00 1.14677489e-01
-3.58159721e-01 2.91186601e-01 -2.90776521e-01 -2.94861197e-01
-2.60582089e-01 2.86901116e-01 -1.16847456e-01 -5.00595510e-01
5.67657113e-01 -4.13729668e-01 -3.12551558e-01 -3.52572836e-02
-6.20347142e-01 -3.20318669e-01 2.47553855e-01 9.06238854e-01
1.32079661e-01 -3.07554394e-01 4.83632267e-01 -7.26964831e-01
9.77994919e-01 -6.31898165e-01 -4.70058918e-01 1.35971271e-02
-5.24414062e-01 -2.73757249e-01 5.74730575e-01 6.38420209e-02
-1.04133821e+00 -4.12802905e-01 -8.22879255e-01 -4.26390618e-01
-3.28633040e-01 6.24018133e-01 -7.75173157e-02 1.34655133e-01
6.04327619e-01 1.28834739e-01 -1.29525766e-01 -8.23239565e-01
7.01231122e-01 9.70011413e-01 3.52634192e-01 -5.45102358e-01
3.14744748e-02 3.00085664e-01 -4.16081518e-01 -8.43928933e-01
-6.68713748e-01 -5.83848178e-01 -6.32227957e-01 -1.91664696e-01
4.54096466e-01 -8.95729840e-01 -1.17452478e+00 4.55031365e-01
-1.12810683e+00 -6.38105452e-01 -1.25235811e-01 3.48708451e-01
-4.90807176e-01 4.03020561e-01 -1.30614495e+00 -6.28354967e-01
-6.26835823e-01 -1.32824838e+00 8.06127489e-01 5.22919334e-02
-2.49271616e-01 -1.62944806e+00 2.14092776e-01 4.11478490e-01
5.45672357e-01 -2.01440498e-01 5.19763947e-01 -9.04610217e-01
-6.14210427e-01 -1.10426158e-01 7.09916055e-02 6.25505030e-01
-2.12160468e-01 5.79364300e-01 -1.11638844e+00 -4.11997885e-01
-2.32392401e-01 -5.46996236e-01 7.42004871e-01 4.92223263e-01
1.32583439e+00 -3.11509192e-01 -3.14579576e-01 8.72157454e-01
1.12383056e+00 -1.37231991e-01 3.50699574e-01 2.94560254e-01
5.25012434e-01 1.33269385e-01 8.77152830e-02 7.68853784e-01
5.85265458e-01 4.66757387e-01 3.50055367e-01 1.89191297e-01
1.80335954e-01 8.41249302e-02 3.82665992e-01 1.07152653e+00
-3.03680122e-01 -3.81190836e-01 -7.92008936e-01 5.12248397e-01
-1.87349093e+00 -1.31562901e+00 -4.60513145e-01 1.49383235e+00
9.88084078e-01 -1.21060476e-01 2.09450886e-01 -1.70826614e-01
2.61541069e-01 5.19575834e-01 -5.10298193e-01 -9.50329721e-01
2.25283951e-02 3.20715338e-01 1.90733448e-01 4.77233499e-01
-1.06096578e+00 1.02822232e+00 9.19556808e+00 7.86167860e-01
-1.01708221e+00 3.06561649e-01 3.28529000e-01 -3.71576250e-01
-1.93257794e-01 -5.06800972e-02 -8.89131606e-01 3.56760323e-01
1.18359876e+00 -3.33343118e-01 4.61136162e-01 1.23323071e+00
1.93319887e-01 4.06096250e-01 -1.05688310e+00 7.14142561e-01
-1.47748336e-01 -2.16534209e+00 -1.76061824e-01 -1.49585381e-01
7.18338430e-01 8.98438692e-01 6.21621422e-02 4.36513066e-01
1.07008684e+00 -8.86714935e-01 6.95090234e-01 4.51352775e-01
4.29712445e-01 -6.64069355e-01 8.00721586e-01 3.19656074e-01
-9.27203238e-01 -2.60073632e-01 -4.86172706e-01 -6.02852225e-01
8.41148421e-02 7.28834391e-01 -9.91914928e-01 4.02942032e-01
1.22879243e+00 1.06463575e+00 -3.88028711e-01 1.24220562e+00
-3.63390893e-01 9.12971258e-01 7.73557499e-02 -2.02134103e-01
5.97689152e-01 6.06421679e-02 3.84894907e-01 1.56917667e+00
-3.28013957e-01 -3.85904312e-01 4.05791432e-01 8.24894011e-01
-1.05615251e-01 2.21679509e-02 -3.80057663e-01 -1.52682081e-01
7.88089514e-01 1.29812407e+00 -1.71836004e-01 -6.75855577e-01
-6.18217409e-01 8.25235128e-01 5.38795710e-01 4.53807741e-01
-7.53628671e-01 -5.59392512e-01 1.20830822e+00 -4.46684390e-01
5.25121272e-01 -5.17638266e-01 -3.48446280e-01 -1.18274939e+00
-2.24850729e-01 -9.51405168e-01 4.53658164e-01 -7.67994165e-01
-1.53894496e+00 4.67606276e-01 -3.41949373e-01 -6.29081905e-01
-5.34268558e-01 -7.05818653e-01 -6.07850790e-01 1.00850248e+00
-1.13148034e+00 -9.30345595e-01 -2.21552536e-01 4.87713099e-01
6.26508594e-01 -2.77499914e-01 1.11975753e+00 3.99344087e-01
-7.71213233e-01 7.36343384e-01 6.83631420e-01 4.31210876e-01
5.68834662e-01 -1.41425943e+00 1.12010956e+00 5.17937303e-01
2.19181165e-01 1.08023500e+00 6.16649866e-01 -4.01815772e-01
-1.34100401e+00 -1.03300524e+00 1.02878070e+00 -6.57819629e-01
1.05413032e+00 -5.97244442e-01 -7.46299386e-01 1.35674000e+00
5.19014955e-01 4.11349960e-04 1.06292522e+00 6.21112466e-01
-3.56975824e-01 1.17616452e-01 -7.88730443e-01 6.29094243e-01
9.33853328e-01 -6.83676541e-01 -4.02189046e-01 5.35830259e-01
5.59010625e-01 -5.67443728e-01 -1.20198548e+00 -1.43027067e-01
7.11839259e-01 -1.34832525e+00 8.99595559e-01 -1.07136202e+00
3.16545069e-01 6.22482188e-02 -9.92690474e-02 -1.20990443e+00
-7.65428841e-01 -1.08737159e+00 -5.99553823e-01 1.05426586e+00
1.87543109e-01 -4.11876947e-01 8.36601496e-01 3.08028936e-01
-5.14805377e-01 -8.79648268e-01 -6.76217318e-01 -8.16647410e-01
2.14424208e-01 -8.52462590e-01 8.93479288e-01 1.02323627e+00
1.78358138e-01 3.98662031e-01 -2.68816710e-01 -2.70995498e-01
3.57136637e-01 -8.65862072e-02 9.74966824e-01 -1.01473844e+00
-6.02664530e-01 -6.96572363e-01 -4.38423693e-01 -1.40231907e+00
1.42215807e-02 -1.16879034e+00 -4.21173155e-01 -1.62650251e+00
1.92477688e-01 -3.88007730e-01 -1.90301403e-01 7.45588541e-01
-9.19146165e-02 1.83629468e-01 2.01513603e-01 2.34944209e-01
-7.63650835e-01 2.98910201e-01 1.03977942e+00 -1.03466131e-01
-1.91503540e-02 1.20922714e-01 -6.98272526e-01 5.34189284e-01
7.60066569e-01 6.90337047e-02 -3.29598725e-01 -8.30888927e-01
2.11725518e-01 -3.49724531e-01 3.60656887e-01 -7.19791830e-01
9.30652097e-02 -1.20242402e-01 3.69631350e-01 -5.96578836e-01
3.03745896e-01 -1.86294064e-01 -3.37041274e-04 8.69840086e-02
-5.71369350e-01 1.21046752e-01 3.38128395e-02 1.64629430e-01
-1.16087943e-01 -2.05095246e-01 5.04944980e-01 -4.63849992e-01
-1.08303738e+00 3.18399608e-01 -4.59632933e-01 5.83846532e-02
9.10859644e-01 1.33274436e-01 -2.66277075e-01 -4.52062905e-01
-8.22395682e-01 2.23900214e-01 2.43788674e-01 6.06343865e-01
2.51691014e-01 -1.11504686e+00 -6.68832958e-01 1.17758505e-01
-2.15251632e-02 -3.86795133e-01 2.67393887e-01 6.65094316e-01
-6.13595545e-01 7.33470857e-01 -9.53712463e-02 -5.78990042e-01
-1.12262106e+00 5.00458896e-01 5.04396439e-01 -2.03360811e-01
-7.09562302e-01 1.24169588e+00 -4.87643063e-01 -6.56830966e-01
4.45826501e-01 -4.60466176e-01 -4.00354788e-02 -1.02761038e-01
9.62664425e-01 6.39137030e-01 4.52764183e-01 -2.62384266e-01
-2.04705209e-01 -2.71157295e-01 -3.24966311e-01 1.22537315e-01
1.45610631e+00 -1.25632763e-01 -2.36907735e-01 7.52072573e-01
1.26712060e+00 -3.18033755e-01 -1.22427404e+00 -2.89981991e-01
-8.00201595e-02 -2.60819733e-01 3.89987707e-01 -9.36306536e-01
-1.06743407e+00 1.07758617e+00 2.00154558e-01 1.46116570e-01
5.49615204e-01 5.44319600e-02 8.64848912e-01 3.91353428e-01
3.45726013e-01 -1.04064655e+00 -1.36857256e-01 9.20475900e-01
4.86171246e-01 -7.95317113e-01 -1.29678594e-02 1.51066482e-01
-7.61208415e-01 1.19892108e+00 3.55449349e-01 7.59783015e-02
7.44541407e-01 5.96463382e-01 1.39507815e-01 -2.40373775e-01
-1.50355065e+00 -1.42422318e-02 -1.41765222e-01 7.72367060e-01
1.11014819e+00 1.70935988e-01 -1.85168743e-01 5.54808080e-01
-4.35295939e-01 4.98501956e-01 3.79558235e-01 1.00698328e+00
-3.56619954e-01 -1.16907561e+00 -2.13851660e-01 4.59727079e-01
-5.74265182e-01 -2.74294227e-01 -2.18068808e-01 4.59774345e-01
7.26534575e-02 8.85081589e-01 2.63212353e-01 -6.05829298e-01
-4.49495614e-02 1.63980827e-01 2.57492214e-01 -4.97806817e-01
-1.05910444e+00 -4.54937071e-02 2.84721732e-01 -8.36128831e-01
-1.62588879e-02 -5.97331345e-01 -9.89116371e-01 -9.59652364e-01
-5.51808290e-02 2.27176368e-01 6.50639296e-01 8.89515996e-01
6.73296154e-01 5.75889826e-01 4.57853019e-01 -9.70223248e-01
-6.43938959e-01 -1.11853147e+00 -7.54124165e-01 -1.44914761e-01
3.86617988e-01 -4.17667001e-01 -1.36227906e-01 -6.41293600e-02] | [8.69691276550293, 3.203683614730835] |
5918ec82-fc51-423d-9865-0754859c9b22 | learning-sentence-ordering-for-opinion | null | null | https://aclanthology.org/W15-0512 | https://aclanthology.org/W15-0512.pdf | Learning Sentence Ordering for Opinion Generation of Debate | null | ['Makoto Iwayama', 'Toshinori Miyoshi', 'Paul Reisert', 'Yoshiki Niwa', 'Misa Sato', 'Kohsuke Yanai', 'Kentaro Inui', 'Toshihiko Yanase'] | 2015-06-01 | null | null | null | ws-2015-6 | ['sentence-ordering'] | ['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] | [-7.2973103523254395, 3.6287550926208496] |
72f646b5-7d21-4db3-80b2-f78211d88c50 | slot-transferability-for-cross-domain-slot | null | null | https://aclanthology.org/2021.findings-acl.440 | https://aclanthology.org/2021.findings-acl.440.pdf | Slot Transferability for Cross-domain Slot Filling | null | ['Wei Wu', 'Huixing Jiang', 'Shuyu Lei', 'Xiaojie Wang', 'Caixia Yuan', 'Zhuoxin Han', 'Hengtong Lu'] | null | null | null | null | findings-acl-2021-8 | ['slot-filling'] | ['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] | [-7.393661975860596, 3.6534528732299805] |
3e4bcc29-a2fd-4540-88a0-ad2fe21d9dbd | camul-calibrated-and-accurate-multi-view-time | 2109.07438 | null | https://arxiv.org/abs/2109.07438v3 | https://arxiv.org/pdf/2109.07438v3.pdf | CAMul: Calibrated and Accurate Multi-view Time-Series Forecasting | Probabilistic time-series forecasting enables reliable decision making across many domains. Most forecasting problems have diverse sources of data containing multiple modalities and structures. Leveraging information as well as uncertainty from these data sources for well-calibrated and accurate forecasts is an important challenging problem. Most previous work on multi-modal learning and forecasting simply aggregate intermediate representations from each data view by simple methods of summation or concatenation and do not explicitly model uncertainty for each data-view. We propose a general probabilistic multi-view forecasting framework CAMul, that can learn representations and uncertainty from diverse data sources. It integrates the knowledge and uncertainty from each data view in a dynamic context-specific manner assigning more importance to useful views to model a well-calibrated forecast distribution. We use CAMul for multiple domains with varied sources and modalities and show that CAMul outperforms other state-of-art probabilistic forecasting models by over 25\% in accuracy and calibration. | ['B. Aditya Prakash', 'Chao Zhang', 'Alexander Rodríguez', 'Lingkai Kong', 'Harshavardhan Kamarthi'] | 2021-09-15 | null | null | null | null | ['probabilistic-time-series-forecasting'] | ['time-series'] | [-3.63238156e-01 -1.48066700e-01 -4.90665734e-01 -1.09715617e+00
-1.49757361e+00 -1.00442517e+00 1.27692342e+00 8.94123688e-02
4.03625309e-01 8.16227794e-01 8.06809902e-01 -1.38203382e-01
-9.54077467e-02 -7.27849722e-01 -8.16033006e-01 -6.80452883e-01
1.39598802e-01 9.59152162e-01 -9.13030207e-02 5.48785552e-03
2.96526104e-02 9.89234149e-02 -1.67784846e+00 5.75251281e-01
5.77643573e-01 1.30141318e+00 -9.24678370e-02 5.33244729e-01
-6.93834901e-01 8.85962009e-01 -5.01838923e-01 -4.00759369e-01
-2.46020518e-02 3.38700593e-01 -2.89997488e-01 -3.54734547e-02
4.78441238e-01 -1.62690789e-01 1.84803575e-01 6.09665811e-01
3.34940910e-01 2.16561288e-01 1.13989007e+00 -1.48937416e+00
-9.04498100e-01 1.12244499e+00 -7.34501421e-01 1.32402673e-01
2.39016175e-01 -4.65001166e-01 6.43961191e-01 -1.09105515e+00
3.64255816e-01 1.82873988e+00 8.23632598e-01 3.00641328e-01
-1.35599267e+00 -4.91788179e-01 7.28274643e-01 -1.14898132e-02
-1.06965482e+00 -4.15746599e-01 8.11055243e-01 -7.12835491e-01
5.80137670e-01 7.98450857e-02 -2.23562703e-01 1.41984153e+00
5.06480753e-01 7.11651921e-01 1.31937420e+00 -9.64888260e-02
4.46059555e-01 2.97395110e-01 1.94581062e-01 -1.70455813e-01
-4.09733392e-02 3.36256772e-01 -9.07873094e-01 -4.42767113e-01
1.26505807e-01 4.27721024e-01 2.14363918e-01 -2.16189418e-02
-1.44596922e+00 7.87052095e-01 1.62345424e-01 1.15154289e-01
-5.45086145e-01 3.39127392e-01 2.51989216e-01 -1.51112014e-02
1.14966965e+00 -1.03256278e-01 -9.07123148e-01 -6.92656487e-02
-1.14064908e+00 1.81882426e-01 8.56084287e-01 9.19930816e-01
4.86852884e-01 3.35918874e-01 -2.29935050e-01 7.43094385e-01
7.40273058e-01 1.32046795e+00 3.76692116e-02 -1.10498977e+00
5.85560024e-01 1.25099882e-01 5.81619680e-01 -1.01432884e+00
-4.39933479e-01 -1.89317003e-01 -7.96662688e-01 1.74795106e-01
1.73154950e-01 -4.97181207e-01 -1.16612279e+00 1.95454049e+00
3.93196821e-01 4.58074361e-01 4.05978024e-01 5.59293747e-01
1.14146733e+00 1.19273889e+00 5.52610993e-01 -6.18051589e-02
1.26008868e+00 -6.67926967e-01 -7.29633629e-01 -2.96812356e-01
5.90976961e-02 -8.02128017e-01 4.29149538e-01 2.70032793e-01
-6.04400873e-01 -4.75934714e-01 -7.05353856e-01 2.04882801e-01
-7.77456045e-01 4.19033952e-02 5.33514380e-01 5.22025168e-01
-8.83758962e-01 2.29716480e-01 -8.33809853e-01 2.25503728e-01
-7.26498663e-02 -2.80955821e-01 -3.04293096e-01 -5.11959910e-01
-1.42052484e+00 1.17360795e+00 4.24903005e-01 -1.08113617e-01
-9.61339056e-01 -1.08996844e+00 -9.98914123e-01 -1.85299236e-02
1.49668306e-01 -6.89694762e-01 1.24850905e+00 -6.70732737e-01
-1.36541164e+00 8.38836879e-02 -5.56428015e-01 -2.01533765e-01
2.03533545e-01 -1.84547961e-01 -1.06230187e+00 -3.53166759e-01
-8.74488091e-04 5.25133729e-01 1.16882443e+00 -1.92901874e+00
-6.83590233e-01 -4.61793333e-01 -1.29520506e-01 2.07279354e-01
3.29949230e-01 -9.21440721e-02 -1.14247091e-01 -7.38079131e-01
2.45328054e-01 -7.40670383e-01 -4.19908464e-02 -3.97151500e-01
-2.13212013e-01 -3.50765318e-01 9.15353894e-01 -7.22453654e-01
9.11226511e-01 -1.75367939e+00 2.82308638e-01 1.38610259e-01
6.72381073e-02 -5.25456011e-01 -4.32786904e-02 6.03358269e-01
8.33017975e-02 5.92219383e-02 -7.17607811e-02 -7.71399498e-01
2.42361814e-01 6.42381251e-01 -1.17330968e+00 1.63419619e-01
-1.34032071e-02 7.61060774e-01 -7.72680402e-01 -3.03636849e-01
3.91245484e-01 8.89608562e-01 2.29614243e-01 1.90145195e-01
-6.87127888e-01 5.45082211e-01 -3.36689442e-01 9.30837393e-01
9.17255402e-01 -4.79061812e-01 1.13865711e-01 -2.75183797e-01
1.54689580e-01 -1.47182897e-01 -1.36851978e+00 1.43655181e+00
-7.81024635e-01 4.29349720e-01 -3.08761090e-01 -5.36049366e-01
1.00053918e+00 5.12019753e-01 5.43333948e-01 -6.82175979e-02
-1.02755792e-01 1.20491795e-02 -9.69489872e-01 5.01371697e-02
5.59438407e-01 -5.05436122e-01 -5.53158522e-01 7.00338423e-01
3.77975553e-01 -3.69418114e-01 -2.95920402e-01 2.46098176e-01
1.45437047e-01 3.45461160e-01 -3.47145051e-02 -8.15953389e-02
1.54845208e-01 -1.65652812e-01 5.25902808e-01 7.18067110e-01
2.62309313e-01 7.12752700e-01 2.62204766e-01 -6.80787086e-01
-7.92961419e-01 -1.29894614e+00 -3.91375482e-01 1.69971871e+00
-1.30533576e-01 -2.23247126e-01 5.20965457e-02 -5.69881022e-01
4.86376196e-01 1.45473278e+00 -8.47601533e-01 3.47800255e-01
2.40404263e-01 -7.59882689e-01 9.88400131e-02 9.93063748e-01
-1.82070181e-01 -4.77078378e-01 3.08374576e-02 2.92079329e-01
-2.61950105e-01 -1.06336486e+00 -1.85414299e-01 -6.00368381e-02
-8.11003566e-01 -6.16623282e-01 -8.70284557e-01 1.54229373e-01
3.50162059e-01 1.11743033e-01 1.70987189e+00 -1.04030085e+00
6.41180515e-01 9.23641443e-01 -8.30948800e-02 -9.45398271e-01
-4.09029692e-01 -2.85028458e-01 3.19174528e-01 1.71982311e-02
4.68930423e-01 -6.12596273e-01 1.67791767e-03 3.71647269e-01
-8.20789635e-01 -3.47196497e-02 1.21928416e-01 6.23637199e-01
7.58481562e-01 -2.95277685e-01 9.02933896e-01 -9.53428984e-01
4.19200212e-01 -1.25991082e+00 -9.30880666e-01 8.08542907e-01
-8.32594335e-01 3.53186280e-01 2.26331323e-01 -3.31728160e-01
-1.69551325e+00 -7.75496513e-02 4.36711699e-01 -6.83178365e-01
-2.70640403e-01 9.65823472e-01 9.11583602e-02 4.91457194e-01
6.90074921e-01 1.37776658e-01 -2.99591005e-01 -5.57898223e-01
1.15713739e+00 5.04354894e-01 6.46943569e-01 -1.06248891e+00
4.82539773e-01 4.99184281e-01 4.43207733e-02 -1.36846781e-01
-1.37566984e+00 -2.49135092e-01 -4.71725702e-01 -5.26795447e-01
4.60494190e-01 -1.54991221e+00 -4.76550579e-01 4.30755436e-01
-1.05454648e+00 3.10107589e-01 -1.85689583e-01 6.03065789e-01
-3.30713004e-01 1.71073806e-02 -1.50713563e-01 -1.34416819e+00
-2.50718385e-01 -8.26456189e-01 1.53518796e+00 2.87442893e-01
1.87877655e-01 -1.43629515e+00 3.66232544e-01 1.59485742e-01
8.16937208e-01 5.42326927e-01 5.62234819e-01 -6.32684767e-01
-4.40938562e-01 -2.37550408e-01 -1.70600876e-01 1.37702569e-01
-2.42449343e-02 2.51090556e-01 -1.42707050e+00 8.25940222e-02
-1.17576919e-01 -4.50646281e-01 1.21396780e+00 7.51411796e-01
1.01216745e+00 -3.19574744e-01 -3.28610659e-01 2.92427838e-01
1.35919511e+00 -1.12988189e-01 8.20807219e-02 -1.89733371e-01
6.33713961e-01 7.76014805e-01 3.43606114e-01 9.01836276e-01
1.15466154e+00 1.61610901e-01 3.10791254e-01 5.00775218e-01
3.87784481e-01 -3.16395462e-01 4.10397083e-01 9.05166566e-01
-1.53469339e-01 -5.75376153e-01 -1.17282116e+00 5.82663476e-01
-2.08348703e+00 -1.18533218e+00 2.93441176e-01 1.91020095e+00
7.17944682e-01 -2.70233124e-01 -5.03022671e-02 -4.48849112e-01
5.89150250e-01 5.81723511e-01 -8.45202029e-01 -1.17224462e-01
-4.90877867e-01 -6.22232974e-01 3.65389138e-01 6.46448851e-01
-1.25592029e+00 4.51124519e-01 7.52656126e+00 6.97968006e-01
-9.15376186e-01 1.99049443e-01 9.96054232e-01 8.40064511e-03
-1.21324813e+00 -1.49210291e-02 -1.16044962e+00 5.71624815e-01
1.50631249e+00 -2.84428924e-01 2.18780369e-01 1.12102056e+00
-3.19313437e-01 -2.95893610e-01 -1.12569368e+00 1.01441610e+00
2.08165273e-01 -1.82239854e+00 6.89050257e-02 -2.66779512e-01
1.20002663e+00 3.30066234e-01 2.73769528e-01 4.03995991e-01
1.08582592e+00 -1.04369915e+00 8.21613252e-01 1.52182388e+00
6.71976030e-01 -8.13669682e-01 8.86164367e-01 4.20203716e-01
-1.31002462e+00 -5.24300113e-02 -1.69291303e-01 2.76076138e-01
5.51976502e-01 1.07177913e+00 -3.36436629e-01 8.50294411e-01
6.54093325e-01 9.73284602e-01 -2.30781719e-01 4.17059362e-01
1.37582406e-01 6.86474502e-01 -6.22086167e-01 5.24608493e-01
-7.71122277e-02 -4.14795568e-03 3.73612374e-01 9.83948410e-01
7.94961393e-01 8.20530057e-02 3.40864807e-01 6.84895098e-01
6.48991913e-02 -4.52599168e-01 -6.30066752e-01 -4.18874025e-02
7.78908670e-01 9.14944828e-01 1.88394394e-02 -5.82099974e-01
-4.81528610e-01 9.74148512e-02 2.27197055e-02 5.16426563e-01
-6.91746175e-01 4.99073535e-01 7.06959128e-01 -5.57569385e-01
2.10341617e-01 -3.59478891e-01 -6.55289173e-01 -1.43384957e+00
-3.21328789e-01 -3.42737913e-01 8.09791625e-01 -1.14061904e+00
-2.35337162e+00 7.53464997e-01 5.37550449e-01 -1.42266440e+00
-1.01120651e+00 -3.57141167e-01 -2.85849899e-01 1.38076997e+00
-1.64723849e+00 -1.85326779e+00 2.14754473e-02 6.38391614e-01
4.57694918e-01 -4.81392235e-01 1.06000459e+00 -3.16867441e-01
-9.76412222e-02 5.96135147e-02 6.79858029e-01 -4.99014735e-01
8.76766860e-01 -1.42880988e+00 3.09517413e-01 6.23938084e-01
1.57833770e-01 3.72745603e-01 7.94838250e-01 -6.09938622e-01
-1.32067180e+00 -1.06902504e+00 1.01331842e+00 -1.18450773e+00
6.22797787e-01 2.14747787e-02 -8.71012032e-01 7.84097135e-01
3.82878691e-01 1.10392213e-01 1.00418961e+00 6.14971280e-01
-1.12156415e+00 -2.84658641e-01 -1.26156151e+00 -9.39438641e-02
1.16079338e-01 -6.48483515e-01 -8.09511125e-01 3.77424181e-01
8.60192299e-01 -6.25354707e-01 -1.30456722e+00 2.96553731e-01
5.15611470e-01 -6.02463901e-01 1.17395365e+00 -6.53354645e-01
6.08864009e-01 -4.25729901e-01 -8.68672550e-01 -1.69678938e+00
-3.79344225e-01 -1.21594250e-01 -7.83385217e-01 1.57291031e+00
6.35435402e-01 -6.49621367e-01 -4.06098180e-02 1.25150657e+00
1.97245345e-01 -3.02861899e-01 -8.74201357e-01 -4.28381413e-01
2.94113398e-01 -1.01681769e+00 1.18189049e+00 9.64179456e-01
-2.89041251e-01 4.46147248e-02 -9.44179595e-01 6.05340242e-01
9.22509372e-01 7.66260862e-01 3.19612771e-01 -1.67192769e+00
1.17036141e-01 -4.09485660e-02 7.40119964e-02 -4.91378367e-01
4.61808473e-01 -4.08545166e-01 -9.03510489e-03 -1.57653272e+00
7.77572813e-03 -5.96275508e-01 -4.90825355e-01 5.11280537e-01
-1.39861554e-01 -3.28772634e-01 3.06399316e-01 4.36068922e-01
-4.64578271e-01 6.59360468e-01 6.99056327e-01 -2.07228616e-01
1.19500220e-01 2.53054321e-01 -8.44249845e-01 7.06138074e-01
5.74633479e-01 -2.64928192e-01 -6.53161049e-01 -6.74477398e-01
4.22249556e-01 5.53832352e-01 1.16895162e-01 -6.91600442e-01
1.88311696e-01 -7.52327859e-01 8.37996185e-01 -1.21169198e+00
7.09196985e-01 -9.40208972e-01 5.76644003e-01 -6.74363196e-01
-4.55614179e-01 2.31301025e-01 4.48632717e-01 1.21906495e+00
-3.98915678e-01 4.13114458e-01 3.05658281e-01 -5.00353463e-02
-1.05696952e+00 2.41844133e-01 -1.70244962e-01 -2.00765189e-02
7.88973689e-01 5.07833123e-01 -6.61030412e-01 -8.66777420e-01
-9.99133348e-01 4.57804114e-01 7.02636316e-02 8.46430600e-01
3.32582980e-01 -1.61880326e+00 -9.66803551e-01 -9.82053056e-02
3.64178270e-01 -2.40117878e-01 7.63856888e-01 1.38644412e-01
5.00486791e-01 5.26886642e-01 -5.64757995e-02 -8.42010677e-01
-5.39211512e-01 7.27209985e-01 1.24756321e-01 -2.20580050e-03
-8.02216232e-02 7.93866515e-01 -6.33000433e-02 -8.53331685e-01
9.69380215e-02 -4.02989775e-01 -4.73669678e-01 6.17392302e-01
8.32096457e-01 4.04795408e-01 -1.05979145e-01 -9.47481155e-01
-4.98410821e-01 4.38111603e-01 4.29301858e-01 -5.12765348e-01
1.47696340e+00 -6.02848947e-01 3.95495556e-02 1.26496410e+00
6.03479624e-01 -2.79212624e-01 -1.59414554e+00 -8.00770402e-01
1.38721047e-02 -5.04843891e-01 3.08570772e-01 -1.47310412e+00
-9.87975717e-01 8.57911825e-01 5.34786105e-01 9.93546546e-02
8.42483342e-01 3.70570034e-01 2.95456350e-01 1.44654095e-01
6.13691509e-01 -8.97831380e-01 -4.53003973e-01 6.80694759e-01
1.36365628e+00 -1.81662643e+00 9.25837178e-03 1.36810482e-01
-1.36085773e+00 1.21215105e+00 3.58794093e-01 3.11270863e-01
1.37597036e+00 4.29347008e-01 4.69121218e-01 -3.14833522e-02
-1.31114316e+00 1.09915189e-01 1.01191449e+00 8.33964050e-01
4.01397884e-01 7.08307624e-01 6.07524514e-01 1.03540361e+00
2.03175366e-01 8.48495886e-02 1.41719162e-01 7.37333596e-01
-3.33493382e-01 -9.61787164e-01 -8.45706642e-01 5.52949548e-01
-1.21020794e-01 2.55008731e-02 2.18751565e-01 5.04313447e-02
3.69318202e-03 1.16601312e+00 1.15073696e-02 -4.07772034e-01
-1.10900588e-01 5.44813633e-01 -1.54235587e-01 -2.73616582e-01
-6.55640811e-02 6.25253916e-02 1.28153145e-01 -3.25198978e-01
-7.86039174e-01 -8.82292688e-01 -7.10084856e-01 -5.69568276e-01
2.86814477e-02 -1.25888094e-01 9.78915691e-01 1.03246367e+00
8.53742063e-01 4.17477369e-01 7.04141319e-01 -1.42982709e+00
-7.83804536e-01 -1.05635130e+00 -6.90112233e-01 7.08904816e-03
5.40158570e-01 -1.10744143e+00 -3.30612212e-01 3.08783352e-01] | [6.98618221282959, 3.2742764949798584] |
dc66dc8a-ec94-483b-bdcf-1343286987f4 | synthvsr-scaling-up-visual-speech-recognition | 2303.172 | null | https://arxiv.org/abs/2303.17200v2 | https://arxiv.org/pdf/2303.17200v2.pdf | SynthVSR: Scaling Up Visual Speech Recognition With Synthetic Supervision | Recently reported state-of-the-art results in visual speech recognition (VSR) often rely on increasingly large amounts of video data, while the publicly available transcribed video datasets are limited in size. In this paper, for the first time, we study the potential of leveraging synthetic visual data for VSR. Our method, termed SynthVSR, substantially improves the performance of VSR systems with synthetic lip movements. The key idea behind SynthVSR is to leverage a speech-driven lip animation model that generates lip movements conditioned on the input speech. The speech-driven lip animation model is trained on an unlabeled audio-visual dataset and could be further optimized towards a pre-trained VSR model when labeled videos are available. As plenty of transcribed acoustic data and face images are available, we are able to generate large-scale synthetic data using the proposed lip animation model for semi-supervised VSR training. We evaluate the performance of our approach on the largest public VSR benchmark - Lip Reading Sentences 3 (LRS3). SynthVSR achieves a WER of 43.3% with only 30 hours of real labeled data, outperforming off-the-shelf approaches using thousands of hours of video. The WER is further reduced to 27.9% when using all 438 hours of labeled data from LRS3, which is on par with the state-of-the-art self-supervised AV-HuBERT method. Furthermore, when combined with large-scale pseudo-labeled audio-visual data SynthVSR yields a new state-of-the-art VSR WER of 16.9% using publicly available data only, surpassing the recent state-of-the-art approaches trained with 29 times more non-public machine-transcribed video data (90,000 hours). Finally, we perform extensive ablation studies to understand the effect of each component in our proposed method. | ['Christian Fuegen', 'Maja Pantic', 'Stavros Petridis', 'Jáchym Kolář', 'Niko Moritz', 'Morrie Doulaty', 'Ruiming Xie', 'Honglie Chen', 'Pingchuan Ma', 'Konstantinos Vougioukas', 'Egor Lakomkin', 'Xubo Liu'] | 2023-03-30 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Liu_SynthVSR_Scaling_Up_Visual_Speech_Recognition_With_Synthetic_Supervision_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Liu_SynthVSR_Scaling_Up_Visual_Speech_Recognition_With_Synthetic_Supervision_CVPR_2023_paper.pdf | cvpr-2023-1 | ['visual-speech-recognition'] | ['speech'] | [ 3.53599519e-01 3.81493628e-01 -2.84809858e-01 -2.31962368e-01
-1.52975917e+00 -4.13482279e-01 6.72582150e-01 -5.60591400e-01
-2.94681787e-01 4.75086898e-01 4.49414790e-01 -2.40566790e-01
7.80281484e-01 -5.41317910e-02 -9.66093838e-01 -4.91151214e-01
3.87301832e-01 3.91243428e-01 1.28812894e-01 -1.31549761e-01
-8.68397132e-02 3.82198274e-01 -2.02858567e+00 5.13921738e-01
6.08777642e-01 9.24752176e-01 3.14528555e-01 1.05519092e+00
-9.96996462e-02 4.92696971e-01 -6.90358043e-01 -2.54179776e-01
1.14160724e-01 -5.37309766e-01 -4.79137450e-01 2.99897015e-01
7.00428367e-01 -3.43274325e-01 -5.25319457e-01 6.42605543e-01
8.78216207e-01 1.41932070e-02 4.99248832e-01 -1.38216579e+00
-5.85863471e-01 5.67448676e-01 -4.71697628e-01 -3.64835501e-01
5.98876178e-01 7.90686309e-01 8.97418737e-01 -1.20629382e+00
8.19037378e-01 1.29818714e+00 3.12233448e-01 1.25014889e+00
-1.19002676e+00 -8.18645000e-01 -1.54277742e-01 1.39179945e-01
-1.59100115e+00 -1.39265966e+00 8.32285404e-01 -2.76362538e-01
1.01410723e+00 2.38237351e-01 6.83180094e-01 1.70286727e+00
-7.02355564e-01 1.19736266e+00 1.08679247e+00 -5.22134781e-01
9.06438455e-02 1.94493622e-01 -5.53154647e-01 5.02855420e-01
-3.26663941e-01 2.39616379e-01 -1.05737603e+00 1.39639825e-01
2.05942616e-01 -6.72396481e-01 -5.13704598e-01 -1.49966031e-01
-1.30169487e+00 6.31221771e-01 -2.51738936e-01 -5.77636547e-02
5.30533865e-02 1.80495173e-01 6.63607001e-01 2.14582250e-01
5.71041346e-01 -3.79302502e-02 -4.53668952e-01 -5.16362727e-01
-1.39719558e+00 1.32853910e-02 5.63201249e-01 1.06551719e+00
4.69530523e-01 6.37103438e-01 -3.50493193e-01 1.19526637e+00
5.34557819e-01 1.01257181e+00 6.30505025e-01 -9.51418042e-01
7.49186635e-01 5.01843132e-02 -1.90320864e-01 -1.58416212e-01
1.15903579e-01 1.52946264e-01 -4.44580197e-01 2.57768393e-01
4.45094317e-01 -5.06080613e-02 -1.27359092e+00 1.90739453e+00
1.34254992e-01 4.31336671e-01 6.05501294e-01 6.95645511e-01
1.24740231e+00 9.56026435e-01 -6.50296509e-02 -4.09846008e-01
9.57106292e-01 -1.22672033e+00 -7.91896224e-01 -6.15095533e-02
5.14123619e-01 -8.66795182e-01 1.53623390e+00 2.15589419e-01
-1.10656774e+00 -5.70229232e-01 -7.40114033e-01 8.55768658e-03
2.54591014e-02 6.15699589e-01 1.38741046e-01 8.01695406e-01
-1.54518986e+00 9.51075554e-02 -6.25675976e-01 -3.42282534e-01
4.44505244e-01 2.68903077e-01 -5.98357201e-01 -1.51704282e-01
-9.17404711e-01 4.55309302e-01 -1.83541119e-01 -1.52317332e-02
-1.40109336e+00 -8.48618686e-01 -1.20105863e+00 -1.52045578e-01
5.20904124e-01 -8.49675685e-02 1.37544751e+00 -1.06215942e+00
-2.22827148e+00 1.01462579e+00 -6.12588465e-01 -3.47202748e-01
8.94787908e-01 1.06949605e-01 -5.76276839e-01 3.83063644e-01
-2.05656394e-01 1.24024332e+00 1.13051844e+00 -1.41347194e+00
-1.88955382e-01 9.49072465e-02 -4.83569771e-01 1.39509678e-01
-1.35942563e-01 1.30858779e-01 -8.23097646e-01 -6.44204617e-01
-5.69874883e-01 -1.16668785e+00 3.46824288e-01 1.23633884e-01
-5.04056215e-01 -3.14451873e-01 1.00021696e+00 -7.60878265e-01
7.29066193e-01 -2.35283375e+00 -2.67239120e-02 -4.00550812e-01
-9.86845568e-02 7.23216474e-01 -6.16521299e-01 3.19964230e-01
-1.51288450e-01 3.02857459e-01 -1.54910564e-01 -8.71000469e-01
-8.54422450e-02 -1.05304934e-01 -3.80257726e-01 3.28184783e-01
2.66526878e-01 9.70698714e-01 -7.05735028e-01 -5.56998193e-01
4.54505950e-01 8.19797575e-01 -5.23869038e-01 5.05270302e-01
-3.08304518e-01 4.69190359e-01 1.83795586e-01 8.46419990e-01
4.28666025e-01 6.93255141e-02 -3.74573134e-02 -2.02780098e-01
2.99557787e-03 2.74103999e-01 -8.70163500e-01 1.86100733e+00
-6.65163636e-01 1.30632973e+00 3.97033133e-02 -7.05222011e-01
9.69828784e-01 7.79668808e-01 3.75351578e-01 -5.69195390e-01
-1.37068838e-01 1.81572005e-01 -2.53743023e-01 -5.87187827e-01
3.00487190e-01 1.57465264e-01 1.58724025e-01 2.79430836e-01
3.71125340e-01 -4.01865363e-01 1.33586183e-01 2.30974108e-01
9.62014616e-01 2.40751907e-01 -8.67056847e-02 1.85873419e-01
6.57288253e-01 -3.82480294e-01 3.01230878e-01 4.03971374e-01
-3.57025385e-01 8.99739265e-01 4.02690828e-01 1.95364371e-01
-1.23462045e+00 -1.15214753e+00 -5.17399199e-02 8.73016357e-01
-3.69666606e-01 -5.61300278e-01 -1.12126195e+00 -6.47778153e-01
-2.00960323e-01 7.92013049e-01 -5.19402444e-01 1.02952190e-01
-3.56841475e-01 -3.07451598e-02 1.10424280e+00 3.34188640e-01
3.40084344e-01 -1.40799332e+00 -1.07936762e-01 -1.57259971e-01
-3.31298143e-01 -1.65557575e+00 -6.88349426e-01 -4.62559611e-01
-1.54029429e-01 -7.93897510e-01 -1.13686872e+00 -5.83893955e-01
4.82116580e-01 1.29623830e-01 8.39000463e-01 -1.15067080e-01
-2.86595196e-01 5.41685879e-01 -3.40961605e-01 -2.47773245e-01
-1.14978957e+00 -5.29168174e-02 3.31263244e-01 2.99977779e-01
-5.85164465e-02 -1.74018964e-01 -3.39845479e-01 3.16431850e-01
-5.32589197e-01 3.39860737e-01 4.61678386e-01 8.10827255e-01
6.02400362e-01 -7.34845877e-01 8.58317137e-01 -4.29805934e-01
2.22752079e-01 -7.01944754e-02 -6.08878195e-01 2.47530147e-01
-3.95582646e-01 8.36988795e-04 6.49798512e-01 -8.74584436e-01
-1.07765818e+00 1.43832594e-01 -5.25035322e-01 -1.05234289e+00
-3.09172094e-01 -2.17457607e-01 -3.66696090e-01 1.53450780e-02
3.25521290e-01 4.77373630e-01 2.73843974e-01 -3.25821996e-01
6.16304278e-01 1.27808440e+00 5.87239861e-01 -3.43256980e-01
6.54800475e-01 3.04936379e-01 -2.55786926e-01 -1.32052672e+00
-3.24484289e-01 -2.40597978e-01 -1.30588055e-01 -4.60888088e-01
8.21945906e-01 -1.15195727e+00 -9.33168113e-01 7.57148206e-01
-1.01655483e+00 -9.15192902e-01 -4.28452879e-01 4.15057570e-01
-9.96019185e-01 2.57976860e-01 -3.56218100e-01 -1.00989497e+00
-2.43885830e-01 -1.64424920e+00 1.51343155e+00 -5.11828177e-02
-8.57425183e-02 -4.52964753e-01 2.76560113e-02 8.66394341e-01
2.18455181e-01 -1.25830054e-01 1.81481421e-01 -6.88460350e-01
-4.21702832e-01 4.79635969e-02 -9.25236642e-02 7.39606500e-01
7.25332350e-02 3.71135145e-01 -1.39904118e+00 -3.39701563e-01
-6.82765961e-01 -9.68165278e-01 8.86749566e-01 2.56441325e-01
1.18509448e+00 -3.38719904e-01 -1.68293417e-01 6.00299299e-01
8.92248988e-01 -7.91136734e-03 5.90396941e-01 -3.16815972e-01
7.97555268e-01 6.20761633e-01 6.56269312e-01 2.35407576e-01
3.62945557e-01 9.69242811e-01 1.55557051e-01 -1.67965159e-01
-9.01321590e-01 -7.07563818e-01 8.77744079e-01 1.01682806e+00
1.31229147e-01 -5.92830062e-01 -9.12968636e-01 6.97693110e-01
-1.35336757e+00 -9.13302481e-01 2.97557265e-01 2.23422837e+00
1.04222393e+00 -1.40195981e-01 1.79830641e-01 2.06669256e-01
7.65225768e-01 4.30586845e-01 -5.36283016e-01 -2.97485769e-01
-1.90052330e-01 1.54649451e-01 3.09294403e-01 6.66087329e-01
-7.38623917e-01 1.44224179e+00 6.11466646e+00 1.09722579e+00
-1.46034551e+00 1.35282308e-01 6.20263100e-01 -4.79829788e-01
-3.55134040e-01 -2.52868384e-01 -9.26615477e-01 4.91586387e-01
1.40068126e+00 3.29265408e-02 7.41107047e-01 7.86280513e-01
6.00819707e-01 7.04003721e-02 -1.04540050e+00 1.38712072e+00
5.66132367e-01 -1.47684836e+00 9.65579823e-02 5.12556545e-02
6.33220851e-01 4.37080592e-01 2.74160743e-01 3.00673038e-01
1.12634897e-01 -1.18146706e+00 8.72980058e-01 2.10762337e-01
1.80945122e+00 -5.91450930e-01 1.59094572e-01 1.48384333e-01
-1.36171520e+00 2.00448185e-01 -4.10903338e-03 5.98033130e-01
3.29122633e-01 -5.43817952e-02 -1.21042013e+00 1.32067040e-01
5.93712986e-01 6.11697435e-01 -5.85417032e-01 5.27785838e-01
-3.40258062e-01 1.03238225e+00 -4.03115690e-01 -1.24391476e-02
-9.53402836e-03 3.30592304e-01 7.27178633e-01 1.24475288e+00
2.24631771e-01 -2.41580337e-01 -1.35928020e-01 6.91816270e-01
-4.66562212e-01 2.55940527e-01 -7.79550791e-01 -3.21625143e-01
6.00778401e-01 1.03367567e+00 -2.82443374e-01 -4.53864127e-01
-4.92365927e-01 7.94529915e-01 1.86769530e-01 5.87258160e-01
-8.28914344e-01 -8.58994648e-02 7.11931705e-01 1.18215077e-01
2.98360974e-01 -9.29265395e-02 2.30316728e-01 -1.17869449e+00
-2.45810896e-02 -1.32425201e+00 -2.20689386e-01 -1.08531857e+00
-6.87603593e-01 8.39536071e-01 -7.46818632e-02 -1.16593289e+00
-6.74103260e-01 -4.78779227e-01 -3.16819668e-01 6.96645081e-01
-1.62109721e+00 -1.35745764e+00 -2.58216381e-01 7.08739340e-01
1.23462641e+00 -6.41581059e-01 8.61206353e-01 -2.65434291e-02
-4.70116496e-01 1.09010303e+00 -2.09374493e-03 2.19397306e-01
1.01578164e+00 -7.68578827e-01 8.04841518e-01 7.56405056e-01
4.03008699e-01 1.80311652e-03 5.90394974e-01 -4.44282979e-01
-1.58896899e+00 -1.09876561e+00 7.02065289e-01 -2.90698618e-01
5.66023648e-01 -8.91281843e-01 -7.46990383e-01 5.01749992e-01
2.34578073e-01 5.06649375e-01 5.79435408e-01 -4.20431584e-01
-5.28778017e-01 -1.61961854e-01 -1.11328959e+00 8.40490222e-01
1.24758387e+00 -8.46811295e-01 -3.50439608e-01 1.65958211e-01
1.06799209e+00 -3.34935665e-01 -5.33412874e-01 5.44900477e-01
6.18702888e-01 -7.25739777e-01 8.24761808e-01 -3.48451644e-01
2.62337655e-01 -1.95055202e-01 -3.51974964e-01 -1.31290221e+00
7.30595231e-01 -1.20307744e+00 -7.38783777e-02 1.67401421e+00
6.60014331e-01 -4.28205580e-01 7.65528858e-01 2.82094657e-01
-1.23527899e-01 -6.71936214e-01 -1.08524954e+00 -8.63336563e-01
-2.29674667e-01 -8.40494037e-01 3.66563082e-01 5.89536309e-01
-2.35197946e-01 3.38195711e-01 -6.47076726e-01 -7.04349503e-02
5.70971608e-01 -3.52463990e-01 1.20439243e+00 -7.55933464e-01
-3.10129255e-01 -3.98542136e-02 -3.82940829e-01 -1.12869251e+00
9.67050135e-01 -9.03486729e-01 3.11863333e-01 -1.29569399e+00
4.13447134e-02 -2.88588345e-01 2.15771064e-01 6.60763502e-01
1.22796379e-01 4.98438478e-01 5.16685188e-01 6.11382015e-02
-4.05302346e-01 8.94577742e-01 1.12669206e+00 -2.76685774e-01
-3.75328630e-01 -2.25949019e-01 -1.97916016e-01 6.07596755e-01
5.64533412e-01 -2.66993612e-01 -5.75199008e-01 -2.38989279e-01
-4.13977146e-01 4.31130677e-01 1.93705544e-01 -8.28386664e-01
1.30623654e-01 -4.71161157e-02 -2.21747234e-02 -5.26009619e-01
8.96374643e-01 -3.28612924e-01 -1.05632365e-01 -4.53052064e-03
-5.66312432e-01 -3.33467662e-01 3.95483464e-01 4.78237063e-01
-1.74413294e-01 1.70119092e-01 9.72199500e-01 3.23858976e-01
-5.53274930e-01 3.19831431e-01 -4.29978997e-01 4.73567337e-01
9.18755412e-01 -9.89208147e-02 -3.06183875e-01 -7.32006788e-01
-5.80129981e-01 -1.23623081e-01 5.77219009e-01 7.64867425e-01
9.31233287e-01 -1.22144091e+00 -9.52938437e-01 4.12681848e-01
3.22489381e-01 -1.19890153e-01 2.63156682e-01 4.85593021e-01
-2.15945885e-01 3.83424044e-01 1.50549948e-01 -9.62992311e-01
-1.70875347e+00 4.00525302e-01 1.06887668e-01 2.53580391e-01
-7.86627412e-01 6.86027169e-01 9.10942405e-02 -3.19297463e-01
3.64982873e-01 -1.21547006e-01 4.42244932e-02 -8.95841941e-02
5.49188972e-01 1.35895178e-01 -4.94724028e-02 -1.11448944e+00
-3.54343355e-01 6.45430148e-01 2.45643854e-01 -6.73370183e-01
1.06524825e+00 -2.21707150e-02 5.91690779e-01 5.81732869e-01
1.37434030e+00 4.93885666e-01 -1.64605474e+00 1.61862850e-01
-6.55449688e-01 -3.88893038e-01 -4.42938544e-02 -7.29099810e-01
-1.10393167e+00 1.02112269e+00 5.75123131e-01 -4.52284575e-01
1.00189030e+00 3.09825599e-01 7.40783870e-01 2.52750009e-01
1.10868394e-01 -9.80373859e-01 3.25821370e-01 1.48986399e-01
1.09663928e+00 -1.41703784e+00 -5.12588799e-01 -4.65169430e-01
-1.20427716e+00 8.11703861e-01 3.90778095e-01 2.68954724e-01
2.14684248e-01 5.87479055e-01 4.20473337e-01 4.79033202e-01
-1.03561997e+00 -2.77093589e-01 2.88226545e-01 8.91784847e-01
2.75745183e-01 -8.46045034e-04 3.28612030e-01 2.67650545e-01
-3.88132125e-01 9.29801986e-02 5.18890560e-01 3.50810379e-01
-2.19901334e-02 -1.02257872e+00 -1.37795970e-01 -2.71436740e-02
-4.13873553e-01 -3.43836814e-01 -3.78226876e-01 8.45900953e-01
-4.02319640e-01 1.20198727e+00 6.49108365e-03 -3.09018075e-01
1.88827619e-01 2.59475410e-01 3.89620215e-01 -4.90421981e-01
-1.10296756e-01 2.22056180e-01 3.34842771e-01 -6.87964857e-01
-2.75670171e-01 -7.50177860e-01 -1.15518045e+00 -9.44285989e-02
-2.15120628e-01 -1.60531536e-01 9.94230449e-01 8.50292623e-01
4.80527818e-01 2.75848031e-01 7.56419420e-01 -1.06302261e+00
-4.32081789e-01 -1.02034974e+00 -1.11916192e-01 3.49708289e-01
6.16195261e-01 -6.32269859e-01 -7.22638071e-01 3.15011352e-01] | [14.353209495544434, 5.082540988922119] |
0b086c54-6b33-4c1d-a752-1246ab0067b8 | yiriyou-smm4h22-stance-and-premise | null | null | https://aclanthology.org/2022.smm4h-1.7 | https://aclanthology.org/2022.smm4h-1.7.pdf | yiriyou@SMM4H’22: Stance and Premise Classification in Domain Specific Tweets with Dual-View Attention Neural Networks | The paper introduces the methodology proposed for the shared Task 2 of the Social Media Mining for Health Application (SMM4H) in 2022. Task 2 consists of two subtasks: Stance Detection and Premise Classification, named Subtask 2a and Subtask 2b, respectively. Our proposed system is based on dual-view attention neural networks and achieves an F1 score of 0.618 for Subtask 2a (0.068 more than the median) and an F1 score of 0.630 for Subtask 2b (0.017 less than the median). Further experiments show that the domain-specific pre-trained model, cross-validation, and pseudo-label techniques contribute to the improvement of system performance. | ['Yanru Zhang', 'Zhongjian Zhang', 'Huabin Yang'] | null | null | null | null | smm4h-coling-2022-10 | ['stance-detection'] | ['natural-language-processing'] | [ 1.91672221e-01 8.58284891e-01 -4.02633548e-01 -3.51437926e-01
-7.66858757e-01 4.00073081e-02 5.85981190e-01 3.26168001e-01
-4.48228925e-01 9.74593520e-01 3.19699049e-01 -3.65413755e-01
9.93055850e-02 -5.65764666e-01 -4.97395575e-01 -3.49215895e-01
8.75346661e-02 4.73752171e-01 3.04713070e-01 -4.52529848e-01
1.94767073e-01 -3.60971212e-01 -1.33107150e+00 8.00588310e-01
5.46962738e-01 1.14873600e+00 -2.73722317e-02 7.02283978e-01
1.12372033e-01 1.02262402e+00 -7.54449964e-01 -3.14088851e-01
-1.40079767e-01 -3.65243703e-01 -1.21566808e+00 -2.28978589e-01
-2.84266174e-01 2.27105636e-02 1.88379660e-01 8.46421003e-01
6.48548067e-01 -1.07275084e-01 7.33639598e-01 -1.15783787e+00
-5.76049089e-01 7.18459487e-01 -8.72990668e-01 3.13940138e-01
3.25656384e-01 -3.81348968e-01 1.01653361e+00 -8.40440750e-01
5.41719794e-01 8.22489679e-01 8.12175930e-01 7.07620800e-01
-7.05530405e-01 -8.01291943e-01 6.00494444e-02 4.69578981e-01
-1.05350912e+00 -2.37603992e-01 5.39202869e-01 -5.89276135e-01
9.24330056e-01 2.78526038e-01 5.56294620e-01 1.20301700e+00
4.22069043e-01 7.69285858e-01 1.45383728e+00 -4.98897046e-01
-1.69203747e-02 5.89770675e-01 7.09861338e-01 2.95035481e-01
1.87611207e-01 -3.31923008e-01 -2.57154405e-01 -1.08906820e-01
2.30108455e-01 -2.03175560e-01 -1.66447107e-02 4.48075652e-01
-1.09832525e+00 1.05046868e+00 2.66148835e-01 3.59603971e-01
-7.96719909e-01 -3.09617609e-01 7.64810681e-01 1.41616821e-01
8.22789609e-01 3.23287994e-01 -5.13145447e-01 1.45312533e-01
-8.39652121e-01 1.76413968e-01 5.86031079e-01 7.31627345e-01
1.65693983e-02 -2.85783738e-01 -4.82082158e-01 8.50995898e-01
2.64122397e-01 2.29339004e-01 7.88457096e-01 -7.60152638e-01
2.48173341e-01 6.96184039e-01 2.46228322e-01 -8.60115230e-01
-8.31764817e-01 -7.62897432e-01 -6.81115568e-01 -1.48394048e-01
1.08902127e-01 -4.17491198e-01 -8.26886177e-01 1.69471765e+00
2.26591691e-01 -3.62302065e-01 1.85998470e-01 4.77375180e-01
1.41997242e+00 4.06187236e-01 2.87163973e-01 -4.32861686e-01
1.60548747e+00 -1.35301435e+00 -9.93092358e-01 -1.53323963e-01
5.01506209e-01 -8.24897170e-01 8.36013198e-01 3.68210554e-01
-1.28298724e+00 -7.12052345e-01 -1.06343675e+00 1.94983408e-01
-2.92246372e-01 -5.25973625e-02 -6.52322620e-02 5.49284577e-01
-7.03236520e-01 1.79836258e-01 -4.76932734e-01 -4.45666879e-01
2.47451827e-01 3.00730050e-01 -1.34967953e-01 4.07822520e-01
-1.52698839e+00 9.14727747e-01 4.49131399e-01 -2.80505717e-01
-6.35520458e-01 -6.02565408e-01 -4.24364656e-01 1.62626566e-05
1.48448393e-01 -4.23348188e-01 1.35300457e+00 -8.22128356e-01
-1.12926102e+00 1.40685129e+00 8.32943395e-02 -7.06604421e-01
5.41885078e-01 -4.60988134e-01 -6.26850247e-01 1.78699028e-02
4.97484714e-01 4.61565524e-01 4.21136558e-01 -9.26395118e-01
-8.09915721e-01 -2.11948946e-01 1.39511153e-01 1.98537409e-01
-2.68746138e-01 3.66382629e-01 -1.61200732e-01 -4.71431524e-01
-2.73354709e-01 -8.31622601e-01 -7.41572157e-02 -8.56039345e-01
-5.10420561e-01 -4.81206536e-01 6.86299384e-01 -1.06155038e+00
1.53700745e+00 -1.74145985e+00 -1.65160000e-01 6.95200497e-03
4.29870695e-01 3.17984283e-01 3.28824610e-01 2.84214914e-01
-4.55952853e-01 4.81589325e-02 1.20758861e-01 -1.12999916e-01
-3.77450794e-01 -2.53653735e-01 3.41683686e-01 2.21913114e-01
-2.72671357e-02 8.01517904e-01 -8.82889569e-01 -4.43433702e-01
-2.41324976e-01 -6.25375733e-02 -3.85837555e-01 1.01470910e-01
-1.61326993e-02 3.12133849e-01 -2.75850505e-01 4.16010648e-01
2.43371010e-01 -6.67657316e-01 3.21811765e-01 -3.11285436e-01
-1.17497519e-01 3.16552877e-01 -6.12951875e-01 1.01673222e+00
-3.72767329e-01 5.02830505e-01 -2.95045003e-02 -1.16854990e+00
8.82619083e-01 7.63940096e-01 8.96435916e-01 -7.06171751e-01
3.89024168e-01 5.37420250e-02 2.71052390e-01 -9.08089280e-01
2.22062811e-01 -2.32276484e-01 -1.13532089e-01 5.60074568e-01
-8.82471949e-02 8.56004834e-01 1.33166030e-01 8.12284350e-02
9.42685783e-01 -6.63998872e-02 7.11448789e-01 -5.76604784e-01
6.99760497e-01 -1.85627386e-01 6.47057056e-01 6.03073895e-01
-5.23989141e-01 4.70499337e-01 4.87027377e-01 -4.06636268e-01
-1.04469252e+00 -2.21158072e-01 -1.32609248e-01 1.24950814e+00
-1.40911475e-01 -2.41574585e-01 -9.40489292e-01 -1.12601697e+00
-2.35849231e-01 6.81807876e-01 -1.12913370e+00 -7.38101546e-04
-4.96708363e-01 -1.01221430e+00 4.17650431e-01 6.99193954e-01
7.73457050e-01 -1.29197812e+00 -8.40055466e-01 3.20046753e-01
-7.47175694e-01 -1.13508511e+00 -3.29493225e-01 2.13474780e-01
-3.74595344e-01 -1.34239340e+00 -9.98804390e-01 -7.23264396e-01
1.72147751e-01 5.49771674e-02 1.13984287e+00 3.63427103e-02
2.17438698e-01 -1.25202164e-01 -5.07099748e-01 -1.04720211e+00
-4.05805498e-01 2.72740304e-01 -2.57410351e-02 1.29932046e-01
6.99539542e-01 -1.77487269e-01 -5.98971009e-01 2.39525512e-01
-4.00299311e-01 3.75912219e-01 4.72520918e-01 1.02564776e+00
4.19002324e-01 -3.78267854e-01 1.31274354e+00 -1.37627149e+00
6.98610485e-01 -7.87403524e-01 1.06381979e-02 2.85122663e-01
-9.10068929e-01 -5.01886070e-01 2.03461573e-01 -2.39992648e-01
-9.06460404e-01 -3.03019762e-01 -3.32360834e-01 -2.89959274e-03
-8.93420950e-02 8.29093695e-01 1.22221448e-01 5.53504765e-01
9.34387982e-01 -1.49222225e-01 2.25432172e-01 -4.26369816e-01
-1.46908998e-01 1.21257782e+00 2.93378115e-01 1.54118896e-01
7.28174224e-02 1.92934155e-01 -3.85150641e-01 -6.68875337e-01
-1.45855594e+00 -7.59820223e-01 -1.40233368e-01 -3.81370842e-01
1.37866533e+00 -1.07815909e+00 -7.02808440e-01 3.78753662e-01
-9.31026638e-01 -1.77472487e-01 2.76331673e-03 5.24611533e-01
-4.63076741e-01 1.02298446e-01 -5.72208703e-01 -8.56098354e-01
-9.24744546e-01 -8.69069576e-01 5.86133599e-01 -2.79325875e-03
-8.24691057e-01 -8.00771117e-01 1.27532363e-01 1.17710555e+00
2.28083640e-01 4.90349978e-01 9.97161567e-01 -1.41035295e+00
4.19200510e-01 -2.31505141e-01 -1.55895606e-01 3.46951157e-01
9.89080891e-02 -4.89377469e-01 -1.08637905e+00 -2.17639193e-01
1.66424900e-01 -4.87475336e-01 6.25931621e-01 7.21462727e-01
1.10425317e+00 -1.77770719e-01 -3.91896814e-01 -1.79205298e-01
9.46357906e-01 5.22658706e-01 6.39825702e-01 6.02898836e-01
3.76539856e-01 6.64656997e-01 7.07036674e-01 2.54543602e-01
4.73271072e-01 6.83884799e-01 3.47972631e-01 -3.42456728e-01
-1.28540322e-01 1.38479084e-01 2.42629483e-01 9.74659979e-01
-2.55397052e-01 -6.14379197e-02 -1.05653536e+00 7.01854825e-01
-2.03979969e+00 -9.81872559e-01 -4.93325889e-01 1.93971920e+00
7.89289713e-01 6.16429985e-01 7.37981260e-01 5.19810736e-01
9.05074894e-01 -1.18045151e-01 -2.70797521e-01 -6.73199832e-01
2.73013502e-01 -3.37387389e-03 2.27021217e-01 4.42094654e-01
-1.36575997e+00 3.74439090e-01 6.21472549e+00 6.43762767e-01
-7.87780583e-01 4.30406332e-01 9.25550342e-01 4.02062200e-03
-5.69413193e-02 -6.11111999e-01 -7.45712459e-01 7.48352528e-01
1.39603925e+00 -1.29798919e-01 -2.35525832e-01 6.54184639e-01
1.49333879e-01 6.90066218e-02 -6.38925493e-01 4.48210478e-01
3.37361932e-01 -1.29729640e+00 -1.41462654e-01 1.66343480e-01
8.62300277e-01 1.90314963e-01 -1.92828700e-01 8.07097793e-01
3.52951093e-03 -8.84507239e-01 6.11186564e-01 2.98612982e-01
7.57403016e-01 -6.65649116e-01 1.30624020e+00 4.83442217e-01
-9.01836276e-01 -1.90276325e-01 2.21315339e-01 -5.24963364e-02
1.14687867e-01 6.66650355e-01 -1.09151053e+00 8.17923069e-01
8.79727185e-01 4.47979361e-01 -7.56377429e-02 6.61403179e-01
-7.66749308e-02 6.51180565e-01 6.04378283e-02 -2.20663592e-01
9.46134105e-02 4.00757998e-01 5.19971848e-01 1.04678118e+00
-3.13147157e-02 6.31630421e-03 1.50975555e-01 2.08920911e-01
-2.29809359e-01 3.72741640e-01 -3.10193956e-01 1.00214221e-01
3.22507739e-01 1.08477986e+00 -5.64158678e-01 -6.65948391e-01
-4.78534639e-01 3.72626811e-01 2.45065447e-02 1.93335816e-01
-1.14382136e+00 -5.07002890e-01 -3.80737446e-02 2.22618356e-01
1.62850708e-01 5.86011708e-01 -6.29488707e-01 -7.22421527e-01
-2.51292288e-01 -1.08954334e+00 6.36356115e-01 -5.86144567e-01
-1.13324511e+00 7.60794342e-01 1.06102526e-02 -8.66048396e-01
-2.52673954e-01 -4.23048824e-01 -6.03727639e-01 8.20758402e-01
-1.12647152e+00 -1.28300190e+00 -3.36064130e-01 3.06386232e-01
6.89761043e-01 -3.56939614e-01 1.06731641e+00 6.43338144e-01
-5.71173072e-01 7.44611025e-01 -7.64591917e-02 2.63925605e-02
6.74782097e-01 -1.16877508e+00 1.28578812e-01 1.42780513e-01
-4.99787629e-01 3.10891062e-01 6.33514106e-01 -7.71390319e-01
-1.50598958e-01 -1.21559346e+00 1.61518407e+00 -3.58935803e-01
2.77456641e-01 9.56959501e-02 -6.02428615e-01 6.90831006e-01
4.19793218e-01 -6.02537274e-01 1.22289157e+00 5.98675966e-01
5.90114892e-02 1.59386411e-01 -1.17538488e+00 1.87707573e-01
4.62205499e-01 -2.56788641e-01 -6.66287482e-01 6.38236940e-01
5.63528419e-01 -3.69902313e-01 -1.15596569e+00 7.79098630e-01
7.68790424e-01 -7.84255505e-01 8.46943200e-01 -8.88154089e-01
7.96654642e-01 1.17955711e-02 -3.38350087e-01 -6.97526336e-01
-5.37587345e-01 -2.61211336e-01 -4.09419149e-01 8.37110996e-01
9.10242081e-01 -3.20836306e-01 8.53285313e-01 1.26630589e-01
-1.22588947e-01 -1.41141605e+00 -7.08300769e-01 -3.82894963e-01
9.73992944e-02 -8.09603557e-02 2.79437572e-01 1.14691198e+00
4.10304852e-02 9.39018428e-01 -6.40267611e-01 -4.89346869e-02
3.92439544e-01 2.48017460e-01 3.57082605e-01 -1.43488657e+00
-3.73361379e-01 -2.13106558e-01 -5.73538095e-02 -5.38665652e-01
-2.66507864e-01 -7.81866431e-01 -3.52236032e-01 -1.64323843e+00
7.24043667e-01 -3.24318670e-02 -8.07778716e-01 4.92627740e-01
-3.25150818e-01 4.25509840e-01 -2.32549861e-01 1.48907855e-01
-5.23235142e-01 2.04382733e-01 7.52010942e-01 -2.85892427e-01
-1.30039215e-01 4.15778041e-01 -9.97408688e-01 9.10525084e-01
1.12154973e+00 -5.62572658e-01 -3.14773262e-01 -6.67647943e-02
1.15944766e-01 -1.39275258e-02 1.57909561e-02 -8.22050750e-01
-1.36757150e-01 -1.51704162e-01 2.53128260e-01 -9.57809031e-01
1.90163344e-01 -2.89001077e-01 -8.39711875e-02 7.60544360e-01
-6.23894930e-01 6.96911588e-02 3.64912152e-02 4.92311180e-01
-5.23822894e-03 -3.90258163e-01 7.83916056e-01 -2.41415724e-01
-1.75021082e-01 -2.44127378e-01 -6.01038158e-01 1.45256847e-01
1.18539631e+00 -2.61391532e-02 -1.13853641e-01 -4.55024511e-01
-1.27674592e+00 3.50650221e-01 -2.88482279e-01 3.66038024e-01
2.29605794e-01 -1.20742893e+00 -9.83283222e-01 -1.36044072e-02
9.28239301e-02 -2.81453073e-01 2.57974863e-01 1.13044202e+00
-3.18579406e-01 5.45833886e-01 -1.70810118e-01 -3.51545036e-01
-1.63043594e+00 3.23124945e-01 2.24367723e-01 -9.06920195e-01
-3.44629109e-01 9.16217506e-01 2.05565736e-01 -2.50712395e-01
3.44568104e-01 7.44870603e-02 -8.92687440e-01 2.33467862e-01
6.14124477e-01 6.49681866e-01 2.81990916e-01 -5.51656425e-01
-4.64321494e-01 1.62413090e-01 -5.25310516e-01 -3.00806630e-02
1.36201060e+00 -6.12170063e-02 1.35777444e-01 5.26641309e-01
9.76565003e-01 -3.62216234e-01 -3.28037024e-01 -8.19867402e-02
2.27843180e-01 4.42222387e-01 1.07356928e-01 -1.18733180e+00
-8.07869315e-01 6.83710873e-01 6.65607154e-01 5.39248228e-01
9.02069986e-01 -9.42262709e-02 9.35238600e-01 1.40043646e-01
-3.72065306e-02 -1.35533535e+00 2.09154904e-01 6.72879994e-01
9.29076731e-01 -1.28832483e+00 4.31495011e-02 -9.44301561e-02
-7.90802479e-01 5.85718989e-01 6.26606882e-01 1.90622523e-01
7.66890705e-01 -1.85568005e-01 3.20381075e-01 -4.72164512e-01
-8.41358900e-01 2.46458203e-01 4.47111070e-01 3.27304542e-01
8.34283650e-01 1.83204249e-01 -9.74765956e-01 1.18707669e+00
2.33541243e-02 1.18147947e-01 2.09758818e-01 9.41206098e-01
-4.64153945e-01 -9.42946255e-01 -1.49719790e-01 6.16340697e-01
-1.16099894e+00 1.55118078e-01 -4.62427527e-01 6.77431822e-01
3.51620018e-01 1.11899889e+00 -2.35412672e-01 -5.71874499e-01
2.78301507e-01 4.00919616e-01 1.48010790e-01 -6.94661200e-01
-9.12835181e-01 2.55866468e-01 6.94614410e-01 -2.38216102e-01
-7.62271583e-01 -2.83528417e-01 -1.13724804e+00 -1.10968865e-01
-5.10126114e-01 2.85751760e-01 4.36577052e-01 9.30835545e-01
3.80884826e-01 9.59505796e-01 5.90115905e-01 -1.87830642e-01
-4.26385969e-01 -1.45340884e+00 -2.85986811e-01 4.68707234e-01
2.26235747e-01 -6.75749302e-01 1.02578141e-01 1.96996734e-01] | [8.53195858001709, 8.929500579833984] |
c76cfafc-6ba4-4c9a-a821-d191b9c3964d | uncertainty-aware-null-space-networks-for | 2304.06955 | null | https://arxiv.org/abs/2304.06955v1 | https://arxiv.org/pdf/2304.06955v1.pdf | Uncertainty-Aware Null Space Networks for Data-Consistent Image Reconstruction | Reconstructing an image from noisy and incomplete measurements is a central task in several image processing applications. In recent years, state-of-the-art reconstruction methods have been developed based on recent advances in deep learning. Especially for highly underdetermined problems, maintaining data consistency is a key goal. This can be achieved either by iterative network architectures or by a subsequent projection of the network reconstruction. However, for such approaches to be used in safety-critical domains such as medical imaging, the network reconstruction should not only provide the user with a reconstructed image, but also with some level of confidence in the reconstruction. In order to meet these two key requirements, this paper combines deep null-space networks with uncertainty quantification. Evaluation of the proposed method includes image reconstruction from undersampled Radon measurements on a toy CT dataset and accelerated MRI reconstruction on the fastMRI dataset. This work is the first approach to solving inverse problems that additionally models data-dependent uncertainty by estimating an input-dependent scale map, providing a robust assessment of reconstruction quality. | ['Markus Haltmeier', 'Simon Göppel', 'Christoph Angermann'] | 2023-04-14 | null | null | null | null | ['image-reconstruction', 'mri-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 4.02554333e-01 9.06542093e-02 2.95158058e-01 -7.05267012e-01
-8.95896971e-01 2.08996702e-02 3.07992131e-01 2.09582508e-01
-7.28698671e-01 8.76352966e-01 1.78049371e-01 -2.19691455e-01
-7.57759452e-01 -6.27163172e-01 -6.23526275e-01 -7.81798899e-01
-1.63589373e-01 7.52952337e-01 8.68519321e-02 6.56215250e-02
7.42076784e-02 6.21964633e-01 -1.17026997e+00 1.20409064e-01
6.59901917e-01 1.14777327e+00 3.78489822e-01 3.35325509e-01
6.68258741e-02 7.67611921e-01 -4.43539657e-02 7.70231113e-02
2.86359817e-01 -4.18795764e-01 -5.37245095e-01 -1.04086921e-01
-9.37361196e-02 -5.44354439e-01 -1.66230112e-01 1.26205802e+00
7.15755582e-01 1.68166876e-01 5.75389802e-01 -5.95816553e-01
6.28049448e-02 7.34194696e-01 -2.82187849e-01 2.23129019e-01
3.35391164e-02 -1.60179138e-01 4.78257924e-01 -6.83739364e-01
4.82550830e-01 8.31580222e-01 6.89505994e-01 1.55521944e-01
-1.35234880e+00 -4.60941672e-01 -5.16255975e-01 2.57744431e-01
-1.13021660e+00 -5.49678683e-01 8.96312892e-01 -3.82589757e-01
3.15053403e-01 1.65318057e-01 3.85677814e-01 8.93066645e-01
2.59726375e-01 2.00385407e-01 1.48145676e+00 -4.03370768e-01
4.45628554e-01 1.29134089e-01 -9.75934342e-02 3.62228841e-01
3.44896138e-01 2.14786202e-01 -1.76263899e-01 3.02200690e-02
1.03484857e+00 -1.18991114e-01 -6.32799149e-01 -6.20376587e-01
-1.17928684e+00 7.70272136e-01 6.31475270e-01 6.70300305e-01
-7.35829771e-01 1.00498773e-01 2.81919539e-01 9.39573199e-02
4.76874948e-01 3.64333361e-01 -1.21178091e-01 -1.98681921e-01
-1.26453006e+00 -4.96895565e-03 8.32701981e-01 3.10566843e-01
3.56044918e-01 7.89906383e-02 8.84597078e-02 6.47040486e-01
3.80656034e-01 2.27414295e-01 4.78782952e-01 -8.19006681e-01
1.61329746e-01 6.65729642e-02 6.59266710e-02 -9.00486708e-01
-7.86440969e-01 -9.17731702e-01 -1.32019818e+00 5.80450833e-01
5.22055387e-01 -1.24960408e-01 -8.58293116e-01 1.76124620e+00
4.72300053e-01 3.79107565e-01 -6.96672350e-02 1.15544546e+00
6.28652930e-01 3.41429740e-01 -2.49408394e-01 -3.91886920e-01
1.11044204e+00 -2.62527406e-01 -9.49232459e-01 -4.77841571e-02
2.27453589e-01 -7.85556018e-01 4.34214175e-01 7.34363198e-01
-1.13553822e+00 -4.92200911e-01 -1.25057065e+00 2.80841321e-01
2.30225608e-01 5.81349619e-02 5.01144566e-02 5.33234894e-01
-8.81844699e-01 1.07882547e+00 -9.70808923e-01 -5.71835339e-02
2.24391565e-01 4.42238718e-01 -4.56282914e-01 -3.03934574e-01
-1.28749430e+00 1.23606956e+00 4.17970687e-01 4.41984296e-01
-8.59865606e-01 -6.54631257e-01 -9.01752830e-01 -2.34390832e-02
4.04148817e-01 -5.26396871e-01 1.08018875e+00 -5.94026744e-01
-1.41028929e+00 4.72126573e-01 3.63902837e-01 -6.43383145e-01
1.03072524e+00 -1.72537282e-01 -2.62035161e-01 2.35212907e-01
-3.37660648e-02 2.74059474e-02 9.88723338e-01 -1.32530379e+00
-2.45710816e-02 -5.10509074e-01 -2.15675801e-01 1.17504805e-01
1.08333118e-01 -2.09063739e-01 -4.15293761e-02 -3.15197617e-01
7.39112139e-01 -8.40718985e-01 -4.79098797e-01 4.37284037e-02
-3.25806856e-01 5.05217016e-01 4.35753524e-01 -1.06982577e+00
6.86615407e-01 -1.99948478e+00 2.97496647e-01 4.59365904e-01
1.71661705e-01 -6.81885472e-03 1.61282450e-01 7.48544037e-02
-3.60202312e-01 -3.68392795e-01 -6.50322199e-01 -5.71473658e-01
-3.71674836e-01 3.20168994e-02 1.94497526e-01 1.05666876e+00
-5.13788462e-02 4.92723584e-01 -7.40929484e-01 -4.83314455e-01
6.65791631e-01 7.91245103e-01 -3.29311520e-01 3.41138601e-01
4.46683057e-02 1.27282536e+00 -3.21273655e-01 5.53801507e-02
7.51631439e-01 -1.98521584e-01 2.25955322e-01 -6.37807071e-01
-1.62516803e-01 -1.45835891e-01 -1.53549922e+00 2.07107902e+00
-7.62253344e-01 2.64929682e-01 5.58039665e-01 -1.48483574e+00
7.41380572e-01 6.27699316e-01 9.74602878e-01 -8.65614951e-01
4.00758386e-01 5.76361179e-01 2.03834087e-01 -6.78944886e-01
2.43765965e-01 -7.73861349e-01 2.75315672e-01 4.35028970e-01
3.91957052e-02 -4.25766855e-01 -1.34825096e-01 -1.25400573e-01
1.05980396e+00 2.59552777e-01 3.56343746e-01 -4.00319815e-01
5.99683583e-01 -3.76811326e-01 3.59452486e-01 5.98034382e-01
-1.16551392e-01 9.95780706e-01 2.50106364e-01 -3.59979957e-01
-1.24512827e+00 -9.59341347e-01 -7.04038918e-01 -2.37757955e-02
-2.27435589e-01 4.25411642e-01 -7.03016043e-01 -3.98335397e-01
-3.11671913e-01 6.95406735e-01 -5.08165956e-01 2.11268589e-02
-6.08235359e-01 -6.91567004e-01 2.22161800e-01 1.69436112e-01
5.02351165e-01 -1.08938742e+00 -8.40051293e-01 4.56474185e-01
-4.48397636e-01 -1.15526140e+00 2.11451858e-01 4.81078625e-01
-1.02217638e+00 -1.15413892e+00 -9.83416080e-01 -2.83987254e-01
6.17466688e-01 -1.70005172e-01 1.10111332e+00 -1.17066473e-01
-3.06114018e-01 2.06678376e-01 -2.05178067e-01 1.23494538e-02
-6.80490911e-01 -2.82243460e-01 2.48910546e-01 1.78778656e-02
-4.47433919e-01 -8.87292802e-01 -5.54937541e-01 2.36204863e-01
-1.28708351e+00 -9.29083899e-02 8.19452941e-01 9.82637703e-01
7.40147710e-01 4.40406233e-01 6.44500613e-01 -7.74484158e-01
5.50794423e-01 -4.55531180e-01 -8.71404231e-01 1.71115905e-01
-3.18751007e-01 3.59629422e-01 5.68917513e-01 -1.06164970e-01
-1.29665160e+00 2.56411850e-01 -8.13850284e-01 -3.16584110e-01
-2.30546668e-01 8.51316154e-01 5.81575138e-03 -1.95899054e-01
8.17200899e-01 1.44559830e-01 3.35831702e-01 -4.34478313e-01
1.62245482e-01 3.97146165e-01 5.61467886e-01 -4.83329922e-01
4.89052743e-01 6.36401296e-01 4.85800534e-01 -1.00290656e+00
-6.69020355e-01 -4.66289371e-01 -6.97518528e-01 -3.67062837e-01
8.21227789e-01 -5.87506056e-01 -5.13993263e-01 3.15722793e-01
-1.00319767e+00 -1.07897528e-01 -4.67209190e-01 1.03885937e+00
-7.56653130e-01 6.03757560e-01 -4.39425677e-01 -7.20382154e-01
-3.07921499e-01 -1.60346854e+00 7.03430891e-01 -1.10063367e-01
6.20006621e-02 -8.87000740e-01 2.08779812e-01 2.12144881e-01
7.52638519e-01 4.99460310e-01 6.85413241e-01 -4.66596186e-01
-4.32620853e-01 -2.37029344e-01 -1.61719158e-01 6.92294002e-01
7.87751377e-02 -7.50056684e-01 -9.00793076e-01 -2.54125267e-01
9.72113729e-01 -4.19988990e-01 5.95004857e-01 9.65122104e-01
1.10235703e+00 1.82080969e-01 1.55936614e-01 6.05918884e-01
1.67362010e+00 -1.33891245e-02 6.95595741e-01 1.91769272e-01
2.69093961e-01 7.24769950e-01 4.32344526e-01 4.59536433e-01
-1.32438287e-01 5.36680162e-01 8.66956115e-01 1.82198528e-02
-1.00584231e-01 1.73006147e-01 -3.74271989e-01 7.95930743e-01
1.10357804e-02 1.78257391e-01 -8.12905967e-01 3.64784956e-01
-1.76660299e+00 -7.04018474e-01 -2.50837982e-01 2.49140120e+00
5.71606338e-01 1.46751657e-01 -4.28028107e-01 4.20213640e-01
5.73806465e-01 -7.50310197e-02 -5.45825899e-01 2.22089127e-01
1.94262460e-01 1.92216590e-01 3.18317086e-01 5.35138607e-01
-8.35034728e-01 4.12008427e-02 5.59974909e+00 5.99511743e-01
-1.19403565e+00 3.19916636e-01 7.27288127e-01 2.22813636e-01
-2.83410013e-01 -1.80443779e-01 -5.63314632e-02 2.90614843e-01
7.60797739e-01 3.31681073e-01 4.12268639e-01 5.69436371e-01
3.78264487e-01 -5.20314872e-01 -9.80207860e-01 1.13435459e+00
5.26614545e-04 -1.17151999e+00 -5.36447883e-01 1.05117366e-01
5.79028666e-01 7.12108612e-02 -5.11923768e-02 -6.54482609e-03
-1.29740745e-01 -1.13992965e+00 6.19464159e-01 6.92381024e-01
7.50068367e-01 -8.09289157e-01 1.05745161e+00 6.93153799e-01
-6.63035095e-01 1.14427991e-01 -2.63952315e-01 2.02781215e-01
6.72668397e-01 1.32413328e+00 -6.82079136e-01 8.94721627e-01
6.66880548e-01 3.99412513e-01 3.81576009e-02 1.25532758e+00
-2.69317657e-01 3.29544067e-01 -5.04751205e-01 3.77312958e-01
2.06278805e-02 -4.74658221e-01 5.98024905e-01 7.98016906e-01
7.11789906e-01 2.24189207e-01 -1.16919465e-01 9.16393220e-01
4.96899225e-02 -3.33646722e-02 -5.88592827e-01 3.03507537e-01
-1.34295404e-01 1.43656254e+00 -9.33810830e-01 -5.39843775e-02
3.26509611e-03 7.11539567e-01 2.19805196e-01 1.45209372e-01
-7.96457767e-01 1.03876427e-01 1.68425918e-01 3.47481728e-01
-1.06187962e-01 -3.53763700e-01 -2.00759992e-01 -9.21309114e-01
6.58913180e-02 -7.82244027e-01 4.92725298e-02 -7.11417317e-01
-1.04035079e+00 9.16409969e-01 1.39664054e-01 -1.16320610e+00
-4.46898758e-01 -4.57014203e-01 -2.43004218e-01 1.00638390e+00
-1.60241008e+00 -7.80363858e-01 -3.67808938e-01 4.99739736e-01
1.91684902e-01 1.35834947e-01 7.93049634e-01 6.50901496e-01
-1.80465188e-02 -2.65350074e-01 2.40237549e-01 -1.12622350e-01
5.57787836e-01 -1.09135103e+00 -1.62129849e-01 8.63488495e-01
3.56263593e-02 4.57359374e-01 1.08724344e+00 -6.64625466e-01
-1.18436825e+00 -6.92823410e-01 3.12124193e-01 1.92474514e-01
4.56956059e-01 -1.16898436e-02 -1.01846993e+00 3.64831716e-01
1.46761909e-02 6.30816281e-01 2.85718083e-01 -1.44774765e-01
3.97305459e-01 -1.17171407e-01 -1.53417778e+00 2.48151571e-02
4.49741930e-01 -4.11518693e-01 -4.46889937e-01 2.84635544e-01
2.01024175e-01 -7.14108527e-01 -9.59703386e-01 7.09577858e-01
5.66954851e-01 -1.31967843e+00 9.57796395e-01 -1.67218953e-01
4.08134907e-01 -3.09015036e-01 -1.65896744e-01 -1.63013864e+00
-1.45807356e-01 -1.14683453e-02 2.62763239e-02 7.58207202e-01
1.07074291e-01 -3.90847892e-01 7.64033556e-01 4.42144215e-01
-1.26441121e-01 -6.02572262e-01 -1.39457750e+00 -6.19752705e-01
-1.42075211e-01 -9.07619655e-01 2.73002326e-01 6.68241024e-01
-3.99757355e-01 -2.30708174e-04 -6.60474300e-01 3.83930713e-01
1.15807366e+00 -2.96186209e-01 2.37598166e-01 -1.21077180e+00
-5.61271966e-01 -3.05841416e-02 -3.98669481e-01 -5.97434998e-01
-4.97321710e-02 -6.93272293e-01 4.06647652e-01 -1.81946695e+00
7.36943632e-02 -7.16931164e-01 -3.41939121e-01 -1.82052612e-01
3.46675247e-01 2.49795079e-01 -8.43297169e-02 9.15206224e-02
-2.29810953e-01 5.42976797e-01 1.25603032e+00 5.35729825e-02
1.79451808e-01 2.10186779e-01 -1.40634522e-01 7.63711572e-01
6.61323786e-01 -6.50425971e-01 -5.00185847e-01 -2.39534453e-01
5.26181996e-01 5.93309104e-01 4.62479234e-01 -1.33223808e+00
3.09841037e-01 2.57278264e-01 4.67164338e-01 -5.21465719e-01
5.88028193e-01 -1.41878104e+00 6.03205264e-01 6.85023844e-01
-2.62527436e-01 -1.78200245e-01 6.23909645e-02 5.39988339e-01
-3.81674707e-01 -6.81864917e-01 1.16658914e+00 -3.93695354e-01
-3.75054806e-01 1.57994390e-01 3.17702368e-02 -2.75786251e-01
7.68138409e-01 1.95414290e-01 1.90391064e-01 -6.73618495e-01
-1.10000563e+00 -2.23109454e-01 1.44239023e-01 -4.69558872e-02
9.02817667e-01 -1.14563465e+00 -8.39797080e-01 1.59963384e-01
-1.44319862e-01 7.23662823e-02 7.09591746e-01 1.17664969e+00
-6.68437481e-01 3.71691793e-01 -2.88246423e-01 -8.48509431e-01
-9.41764593e-01 4.49865848e-01 6.58255696e-01 -6.48465216e-01
-6.64651275e-01 5.49792528e-01 -1.29637524e-01 -5.86686075e-01
1.26731962e-01 -3.59020501e-01 -2.96018273e-01 -1.00370608e-01
4.24069405e-01 3.46923858e-01 5.34562767e-01 -6.98554099e-01
-1.13975488e-01 3.89213979e-01 3.75718236e-01 -5.30431211e-01
1.58834887e+00 -1.34531289e-01 -1.69587478e-01 4.21160400e-01
1.19050217e+00 -1.86463043e-01 -1.15406120e+00 -2.55321801e-01
-8.64042267e-02 -2.97535300e-01 5.92579007e-01 -8.24886441e-01
-1.22411001e+00 9.20380533e-01 9.32762146e-01 3.19322571e-03
9.93804514e-01 -2.82045096e-01 3.47461581e-01 2.30576247e-01
6.48260891e-01 -1.01479542e+00 6.92243204e-02 9.68196020e-02
1.19117165e+00 -1.46138573e+00 2.78630197e-01 -1.13136940e-01
-2.65349239e-01 1.13287032e+00 1.71028990e-02 -1.61188021e-01
1.13443673e+00 3.92380536e-01 1.79792140e-02 -1.77697048e-01
1.14443459e-01 -3.42191570e-02 3.36563140e-01 2.11845875e-01
5.34524024e-01 -1.51196003e-01 -4.73204464e-01 2.68153518e-01
1.62374631e-01 2.41738126e-01 4.01134610e-01 8.94728899e-01
-3.18605155e-01 -1.04815269e+00 -6.81084335e-01 4.51075315e-01
-5.91327429e-01 1.21122286e-01 4.80951071e-01 6.78659260e-01
1.87740233e-02 6.91378057e-01 -3.02202404e-01 1.45722821e-03
2.64507622e-01 -2.10179642e-01 6.54964030e-01 -4.59129542e-01
-3.31604749e-01 1.99118048e-01 -5.09119267e-03 -5.36958575e-01
-6.11800134e-01 -6.55079067e-01 -1.24470842e+00 1.65812120e-01
-3.93242240e-01 1.62700757e-01 1.49190784e+00 1.07153463e+00
-2.66666412e-01 1.01829779e+00 3.44934404e-01 -9.24990296e-01
-8.96636963e-01 -1.06438112e+00 -9.39003289e-01 3.58595550e-01
4.11294132e-01 -6.15807235e-01 -3.57877940e-01 -4.78876740e-01] | [13.304848670959473, -2.486924648284912] |
1ed23a60-fc53-46b9-9e8b-ca1f94af32c1 | pp-yoloe-an-evolved-version-of-yolo | 2203.1625 | null | https://arxiv.org/abs/2203.16250v3 | https://arxiv.org/pdf/2203.16250v3.pdf | PP-YOLOE: An evolved version of YOLO | In this report, we present PP-YOLOE, an industrial state-of-the-art object detector with high performance and friendly deployment. We optimize on the basis of the previous PP-YOLOv2, using anchor-free paradigm, more powerful backbone and neck equipped with CSPRepResStage, ET-head and dynamic label assignment algorithm TAL. We provide s/m/l/x models for different practice scenarios. As a result, PP-YOLOE-l achieves 51.4 mAP on COCO test-dev and 78.1 FPS on Tesla V100, yielding a remarkable improvement of (+1.9 AP, +13.35% speed up) and (+1.3 AP, +24.96% speed up), compared to the previous state-of-the-art industrial models PP-YOLOv2 and YOLOX respectively. Further, PP-YOLOE inference speed achieves 149.2 FPS with TensorRT and FP16-precision. We also conduct extensive experiments to verify the effectiveness of our designs. Source code and pre-trained models are available at https://github.com/PaddlePaddle/PaddleDetection. | ['Baohua Lai', 'Yuning Du', 'Shengyu Wei', 'Qingqing Dang', 'Guanzhong Wang', 'Kaipeng Deng', 'Cheng Cui', 'Qinyao Chang', 'Wenyu Lv', 'Xinxin Wang', 'Shangliang Xu'] | 2022-03-30 | null | null | null | null | ['dense-object-detection', 'online-multi-object-tracking', 'real-time-object-detection'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-4.58010852e-01 -2.91952640e-01 -2.89056033e-01 -9.10315737e-02
-9.81636524e-01 -6.42088354e-01 8.39203894e-02 -1.35252625e-01
-3.12508255e-01 4.91765052e-01 -8.33020627e-01 -1.91721559e-01
1.94225982e-01 -4.37827826e-01 -1.09233165e+00 -5.47465026e-01
-1.22698106e-01 5.89925528e-01 8.41690183e-01 2.29067937e-01
1.71630338e-01 4.66454774e-01 -1.42483068e+00 1.45831108e-01
3.77572566e-01 1.59055245e+00 3.92315060e-01 1.05512488e+00
3.51273447e-01 7.27071404e-01 -8.00458670e-01 -7.58801639e-01
3.51128012e-01 4.97501194e-01 -3.94334823e-01 -1.39823079e-01
8.57284665e-01 -3.20482403e-01 -4.57020938e-01 1.00063384e+00
6.01098299e-01 -1.99935004e-01 3.45045954e-01 -1.54379570e+00
-7.64414430e-01 4.48771536e-01 -1.01756752e+00 4.21440959e-01
-8.88389051e-02 4.36866641e-01 7.81293452e-01 -1.40968525e+00
6.86352551e-01 1.21656370e+00 7.18601108e-01 2.95285195e-01
-8.63062859e-01 -1.06180632e+00 2.04548419e-01 4.39863384e-01
-1.62345612e+00 -3.20028752e-01 2.17129618e-01 -1.04367338e-01
1.02159107e+00 -2.76413094e-03 4.00360763e-01 1.14606798e+00
5.26464224e-01 1.13459539e+00 7.56700993e-01 -1.73510745e-01
8.00267905e-02 2.51632005e-01 3.54009658e-01 1.02552950e+00
3.83641094e-01 -1.23589039e-01 -6.76663041e-01 2.32910320e-01
7.19642401e-01 -1.39099225e-01 2.43097574e-01 -2.13613659e-01
-1.24435711e+00 4.31594491e-01 4.43911225e-01 -1.32884979e-01
-2.62673169e-01 5.56939185e-01 8.54142547e-01 -8.59878957e-02
2.90286303e-01 3.37741882e-01 -6.89821541e-01 -3.53518039e-01
-8.53428245e-01 3.47516954e-01 6.20913267e-01 1.90168297e+00
2.33813509e-01 1.50549889e-01 -2.52881140e-01 5.47797561e-01
4.43816841e-01 1.00250208e+00 6.70536831e-02 -1.37323177e+00
5.12131453e-01 3.72865722e-02 1.83810443e-01 -6.63115799e-01
-2.40757301e-01 -7.57749379e-01 -4.55011696e-01 9.91167724e-02
3.64623398e-01 6.51474902e-03 -1.01680946e+00 1.29401648e+00
3.32672745e-01 3.09445530e-01 -3.39138024e-02 8.84757638e-01
8.48054588e-01 8.37202549e-01 2.52317965e-01 9.47287381e-02
1.77998006e+00 -1.73971617e+00 -6.19085491e-01 -2.44868055e-01
7.34681964e-01 -9.95574057e-01 1.02654767e+00 7.97641635e-01
-1.11137116e+00 -9.04498518e-01 -1.20082545e+00 -8.22136775e-02
-3.84055108e-01 8.49259138e-01 6.15664423e-01 4.74495709e-01
-6.22967958e-01 4.88245308e-01 -1.01899052e+00 -3.09563071e-01
6.18903577e-01 3.01604122e-01 -2.55511794e-02 -1.12841919e-01
-7.28731871e-01 6.10389531e-01 3.34978163e-01 1.20625064e-01
-1.24102318e+00 -9.05133963e-01 -3.16466957e-01 -1.56654585e-02
8.70045781e-01 -3.44546080e-01 1.71440434e+00 3.09383143e-02
-1.37169600e+00 7.81853199e-01 1.25018731e-02 -5.83741963e-01
5.11121869e-01 -8.34820330e-01 -5.69880784e-01 1.54256508e-01
2.99404353e-01 8.50150526e-01 4.91218388e-01 -1.04664254e+00
-9.70062554e-01 -4.19788510e-01 -1.35674730e-01 -6.98706284e-02
3.73294540e-02 1.75942346e-01 -1.15291929e+00 -4.37916845e-01
-3.04996282e-01 -1.17661786e+00 6.80973753e-02 2.12434843e-01
-6.63219929e-01 -4.70726669e-01 1.12807751e+00 -5.24841905e-01
1.03412712e+00 -2.20605612e+00 -4.20636326e-01 -2.23824203e-01
3.92345160e-01 7.86367595e-01 5.97614795e-02 5.74386492e-02
4.36536223e-01 -1.31968260e-01 2.24108577e-01 -6.93670630e-01
3.07246983e-01 -6.81051612e-02 -1.30445033e-01 6.15774453e-01
4.62579988e-02 8.27316880e-01 -8.12876821e-01 -8.36706519e-01
5.50713778e-01 3.05922985e-01 -3.54221851e-01 6.05753846e-02
-9.72452685e-02 7.30942115e-02 -6.38600767e-01 1.30283618e+00
8.12964559e-01 -4.36862916e-01 -1.46343574e-01 -3.62715602e-01
-1.87534809e-01 9.66234952e-02 -1.19535446e+00 1.83886468e+00
-4.04927105e-01 8.30172300e-01 2.16791064e-01 -5.23192286e-01
8.77295852e-01 2.04906046e-01 2.62861848e-01 -5.81572115e-01
4.89673704e-01 2.50563025e-01 -3.77654225e-01 -2.41308421e-01
6.58980906e-01 6.14469767e-01 -1.24277603e-02 -2.03750372e-01
4.24820453e-01 2.71646649e-01 5.24790883e-01 4.08425450e-01
1.23924899e+00 4.65736181e-01 -1.05775036e-01 -2.88221180e-01
2.98217267e-01 -1.00793973e-01 6.30851448e-01 8.35955918e-01
-6.34568632e-01 5.09041667e-01 4.31295961e-01 -2.43126437e-01
-1.03092289e+00 -1.27853584e+00 -4.63888168e-01 1.18412435e+00
5.07373035e-01 -7.44369864e-01 -8.25405598e-01 -7.26149976e-01
1.26992092e-01 7.70621002e-01 -2.10096136e-01 1.91652641e-01
-7.73157775e-01 -6.78167939e-01 6.90916419e-01 9.59394932e-01
7.67337322e-01 -9.79820013e-01 -6.75229192e-01 1.90697923e-01
2.12422952e-01 -1.83089101e+00 -3.68671596e-01 1.37668997e-01
-6.17873728e-01 -8.78371000e-01 -5.43298662e-01 -7.53336966e-01
1.45983264e-01 2.72948891e-01 1.39155829e+00 -3.21240187e-01
-5.36393583e-01 3.01777683e-02 -2.87262201e-01 -6.78720355e-01
7.56232664e-02 3.39207500e-01 1.83493227e-01 -5.16747236e-01
4.61867660e-01 -2.61090308e-01 -7.77684391e-01 6.44085526e-01
-1.03181638e-01 -1.48140699e-01 7.24254012e-01 3.91971409e-01
7.70350635e-01 -4.47858244e-01 2.48716429e-01 -8.49400103e-01
-1.66285098e-01 -3.29965323e-01 -1.14531088e+00 2.43638068e-01
-8.61153662e-01 -2.69268960e-01 4.94279325e-01 -4.22432542e-01
-8.33918869e-01 8.02585259e-02 -2.55432248e-01 -1.09585226e+00
-6.13705106e-02 -3.34333599e-01 -6.43192977e-02 -7.70567311e-03
4.33333308e-01 1.85277394e-03 -5.31471670e-01 -5.84872961e-01
4.96511608e-01 8.18447828e-01 9.58620250e-01 -7.33243704e-01
5.20037532e-01 2.41418034e-01 -1.76630929e-01 -6.47988319e-01
-8.57396543e-01 -7.17842638e-01 -4.15398657e-01 -3.36691916e-01
9.96832907e-01 -1.20927322e+00 -1.34690630e+00 4.67930853e-01
-1.13460600e+00 -1.70685589e-01 -1.09217778e-01 4.73005295e-01
-4.78812009e-01 7.36572519e-02 -1.00058043e+00 -6.83883607e-01
-7.07418382e-01 -1.44426417e+00 1.56804693e+00 2.46111602e-01
1.92829207e-01 -3.31820577e-01 -2.88372517e-01 4.87062216e-01
1.73854709e-01 4.26698513e-02 -4.32486124e-02 -3.99760902e-01
-8.54616344e-01 -3.42170775e-01 -7.02454925e-01 3.88517469e-01
-6.14964664e-01 3.04953456e-01 -1.01136458e+00 -3.85650903e-01
-4.55558091e-01 -5.01436353e-01 6.87395453e-01 4.16314095e-01
1.41022372e+00 1.61981136e-01 -8.21014524e-01 7.76384950e-01
1.42481160e+00 3.01266044e-01 4.73026603e-01 4.66901958e-01
8.68536711e-01 -1.26848027e-01 1.15801620e+00 4.80902016e-01
3.22744399e-01 9.52346027e-01 5.81557333e-01 2.05199793e-01
-3.24830085e-01 -8.71093348e-02 3.91762912e-01 8.08284521e-01
-1.85994446e-01 -6.27955854e-01 -7.30808854e-01 3.86038363e-01
-1.85727513e+00 -5.70868850e-01 -3.85394633e-01 1.87396145e+00
4.46372390e-01 7.08488822e-01 1.96798354e-01 -1.35303572e-01
7.69159198e-01 7.94338286e-02 -6.81856215e-01 -2.25361556e-01
1.86587960e-01 1.22960925e-01 1.16143346e+00 4.40751016e-02
-1.33615172e+00 1.28018892e+00 5.30153322e+00 1.20048738e+00
-7.39027023e-01 5.01654804e-01 5.80096781e-01 -3.52744490e-01
9.91735041e-01 -2.71245599e-01 -1.64045203e+00 7.09876359e-01
1.26607823e+00 1.58059180e-01 4.00611088e-02 1.57603776e+00
-2.04066932e-01 -8.12458172e-02 -1.08748364e+00 1.15901089e+00
3.20665888e-03 -1.34988415e+00 -5.97044706e-01 -7.06272107e-03
3.88315320e-01 2.88483441e-01 1.05853073e-01 7.32788563e-01
5.82903683e-01 -5.01699030e-01 1.15294838e+00 7.63560086e-02
9.13581371e-01 -6.78767622e-01 9.39903438e-01 6.87166080e-02
-1.59203780e+00 1.13873281e-01 -5.79756260e-01 4.42822039e-01
3.74001682e-01 4.79057968e-01 -7.64950752e-01 5.63103735e-01
1.23889530e+00 6.51633441e-01 -6.42302394e-01 8.59919906e-01
-3.57908726e-01 7.25021839e-01 -5.36471605e-01 -1.47372842e-01
3.32686186e-01 3.09817374e-01 4.54607934e-01 1.59996700e+00
2.73046851e-01 -1.17109768e-01 3.77414525e-01 5.83665729e-01
-3.41958702e-01 -2.96686411e-01 -8.41304138e-02 1.97172374e-01
9.58884895e-01 1.55318463e+00 -8.51353467e-01 -6.30006969e-01
-3.35370958e-01 9.56676781e-01 2.00744852e-01 -2.93125007e-02
-1.71684873e+00 -4.01728511e-01 6.40427053e-01 -7.71124214e-02
7.84403920e-01 -3.54527831e-01 -1.63575038e-01 -8.82984877e-01
1.29157349e-01 -5.31039059e-01 1.31748825e-01 -9.78002667e-01
-1.06524086e+00 5.80716312e-01 -1.30097613e-01 -1.09708226e+00
3.03116411e-01 -1.08227921e+00 -2.09588736e-01 2.30967715e-01
-1.27853036e+00 -1.42950559e+00 -4.79818910e-01 1.17863879e-01
1.03438783e+00 -1.98799804e-01 5.06886482e-01 6.70512974e-01
-9.25489187e-01 9.01095688e-01 2.13333249e-01 1.92445397e-01
8.05111408e-01 -1.24350226e+00 7.30128407e-01 7.95082867e-01
8.42939094e-02 2.94511735e-01 6.67733490e-01 -4.33575392e-01
-1.63418186e+00 -1.38263381e+00 4.17922497e-01 -6.99573696e-01
8.19521010e-01 -7.16467619e-01 -5.65244436e-01 1.01115417e+00
1.46435544e-01 4.64912891e-01 1.81006715e-01 1.54002185e-03
-2.99952894e-01 -3.63892764e-01 -9.79834795e-01 3.25161248e-01
1.22227025e+00 2.00193822e-02 -1.42383099e-01 7.28771567e-01
9.52101707e-01 -8.75322104e-01 -1.28383517e+00 3.96415979e-01
7.15705216e-01 -8.02588642e-01 1.09626591e+00 6.04527397e-03
2.57686347e-01 -3.33137423e-01 -3.77387285e-01 -6.81731284e-01
-4.52251822e-01 -5.09133697e-01 -4.52548474e-01 1.43284225e+00
3.86347622e-01 -3.77432317e-01 9.18403685e-01 3.52978781e-02
-5.74354291e-01 -8.73344541e-01 -8.29374790e-01 -1.17971110e+00
-2.37877384e-01 -6.23640776e-01 2.67245382e-01 4.07166213e-01
-5.42964816e-01 3.97334367e-01 -3.83470953e-01 3.94234180e-01
8.98525178e-01 -7.56624043e-02 9.29647326e-01 -8.30479920e-01
-3.98390502e-01 -8.33908543e-02 -3.22807640e-01 -1.33701134e+00
-1.33653432e-01 -5.81906855e-01 1.27105653e-01 -1.15270507e+00
2.56269891e-02 -5.24763525e-01 -3.74569595e-01 4.91661102e-01
8.28327164e-02 4.75905210e-01 6.98457360e-01 3.91662896e-01
-1.30674148e+00 4.65168893e-01 1.04841030e+00 -6.08840361e-02
1.41341537e-01 -7.93519020e-02 -1.78904966e-01 6.93769991e-01
1.02358246e+00 -7.17276037e-01 -3.54053662e-03 -4.48640257e-01
-2.70151168e-01 -1.64366722e-01 3.41446787e-01 -1.48739433e+00
2.04578415e-01 3.66661549e-01 4.62482780e-01 -9.31785882e-01
7.44555354e-01 -5.16060889e-01 -2.27744669e-01 5.42272151e-01
1.35084704e-01 4.67167109e-01 3.68402064e-01 4.95094478e-01
1.68404281e-01 -1.56718418e-01 7.38502920e-01 -1.99270081e-02
-1.18234754e+00 2.46910945e-01 -1.25148848e-01 -2.72968654e-02
1.34883249e+00 -6.04101419e-02 -8.14489543e-01 3.21347862e-01
-5.97218752e-01 5.71289778e-01 1.86405137e-01 4.75637943e-01
3.38275641e-01 -1.19280148e+00 -5.46752214e-01 -1.49775028e-01
1.82941258e-01 8.18597618e-03 2.37885416e-01 9.28911090e-01
-9.43981469e-01 7.96991646e-01 -1.27626166e-01 -8.27073753e-01
-1.32533646e+00 7.55509675e-01 -4.56216671e-02 -4.66294706e-01
-7.89291441e-01 1.06309378e+00 9.16489661e-02 -2.16121480e-01
5.17293930e-01 -2.93398649e-01 2.69903570e-01 -4.08775717e-01
4.45614576e-01 9.40497160e-01 2.15900019e-01 -2.79275507e-01
-7.30491102e-01 5.04884005e-01 -2.51151949e-01 2.30588138e-01
1.01775193e+00 1.91081703e-01 3.81558836e-01 3.28647435e-01
1.03199935e+00 -1.90942377e-01 -1.37116420e+00 5.93902990e-02
-1.52135402e-01 -4.31555450e-01 1.73771515e-01 -7.99687326e-01
-1.16149628e+00 7.77120888e-01 9.72985029e-01 -1.42330796e-01
8.17497551e-01 3.25648338e-01 9.95254695e-01 4.05001372e-01
5.27948201e-01 -1.11161602e+00 7.29090497e-02 2.88814753e-01
5.20398021e-01 -1.31492507e+00 2.30715707e-01 -8.83816063e-01
-4.23546880e-01 8.50185990e-01 1.08034480e+00 -4.09804404e-01
4.82975930e-01 7.90038347e-01 -2.36971661e-01 -3.81671518e-01
-7.53483534e-01 7.57574365e-02 -1.63113639e-01 2.45842591e-01
2.71667361e-01 2.47244269e-01 -9.40909758e-02 3.48416239e-01
-2.18692660e-01 5.05211875e-02 1.74465716e-01 1.02749097e+00
-3.89453769e-01 -6.76155448e-01 -3.97476643e-01 2.37547472e-01
-6.10052168e-01 6.36582449e-02 1.58213854e-01 1.20027578e+00
1.87463671e-01 8.75829697e-01 2.26839632e-01 -5.03534675e-01
5.11067152e-01 -1.63620844e-01 5.21660447e-01 -4.61133391e-01
-4.78318363e-01 2.73869812e-01 1.81574911e-01 -9.27080333e-01
7.43196085e-02 -6.12774432e-01 -1.33894646e+00 -6.05835497e-01
-6.26083732e-01 -1.09783098e-01 9.22162771e-01 5.67349553e-01
6.67395055e-01 9.85195756e-01 1.45937502e-01 -9.57198858e-01
-5.87188780e-01 -1.16705072e+00 -6.89240813e-01 -1.48701161e-01
-1.87629640e-01 -9.80280042e-01 -1.32188620e-02 1.00127079e-01] | [8.654861450195312, -0.2544531226158142] |
6c8275cf-aab0-4b70-b00e-0b4f80e0d54e | rumble-data-independence-for-large-messy-data | 1910.11582 | null | https://arxiv.org/abs/1910.11582v2 | https://arxiv.org/pdf/1910.11582v2.pdf | Rumble: Data Independence for Large Messy Data Sets | This paper introduces Rumble, an engine that executes JSONiq queries on large, heterogeneous and nested collections of JSON objects, leveraging the parallel capabilities of Spark so as to provide a high degree of data independence. The design is based on two key insights: (i) how to map JSONiq expressions to Spark transformations on RDDs and (ii) how to map JSONiq FLWOR clauses to Spark SQL on DataFrames. We have developed a working implementation of these mappings showing that JSONiq can efficiently run on Spark to query billions of objects into, at least, the TB range. The JSONiq code is concise in comparison to Spark's host languages while seamlessly supporting the nested, heterogeneous data sets that Spark SQL does not. The ability to process this kind of input, commonly found, is paramount for data cleaning and curation. The experimental analysis indicates that there is no excessive performance loss, occasionally even a gain, over Spark SQL for structured data, and a performance gain over PySpark. This demonstrates that a language such as JSONiq is a simple and viable approach to large-scale querying of denormalized, heterogeneous, arborescent data sets, in the same way as SQL can be leveraged for structured data sets. The results also illustrate that Codd's concept of data independence makes as much sense for heterogeneous, nested data sets as it does on highly structured tables. | ['Gustavo Alonso', 'Ghislain Fourny', 'Ingo Müller', 'Can Berker Cikis', 'Stefan Irimescu'] | 2019-10-25 | null | null | null | null | ['jsoniq-query-execution'] | ['miscellaneous'] | [-7.34423339e-01 -4.34239917e-02 8.08962584e-02 -6.04157031e-01
-6.40042782e-01 -7.07700968e-01 4.45617348e-01 7.20612228e-01
-1.39423251e-01 3.53473663e-01 3.13762367e-01 -4.68602538e-01
-3.18295151e-01 -1.27093542e+00 -5.37215412e-01 -5.08434117e-01
-3.17202181e-01 9.72336590e-01 5.26933253e-01 -4.81716305e-01
2.90604234e-02 7.41111815e-01 -2.09609795e+00 3.59168380e-01
6.11657202e-01 9.10749793e-01 -1.98210165e-01 6.75735772e-01
-4.46533769e-01 1.27224195e+00 -7.09493518e-01 -1.16330609e-01
4.91103560e-01 4.12980288e-01 -6.83938622e-01 -7.57279992e-01
2.87410110e-01 -2.44097367e-01 2.06741076e-02 5.37982762e-01
4.02340025e-01 -2.18801647e-01 -3.03725183e-01 -1.52752566e+00
-2.00617105e-01 7.43882000e-01 -3.11528563e-01 2.52734989e-01
4.39607859e-01 3.98518831e-01 8.07969987e-01 -8.37545991e-01
8.64015877e-01 1.02229464e+00 9.55054343e-01 -3.47884297e-01
-1.29343295e+00 -6.45437598e-01 -3.82529795e-01 -2.43606091e-01
-1.60882747e+00 -6.58081353e-01 -3.11782416e-02 -5.57614267e-01
1.20276618e+00 1.18024945e+00 6.02554798e-01 -1.44578114e-01
2.64447778e-01 3.31884682e-01 1.06857383e+00 -1.08737685e-01
5.16748786e-01 2.76871830e-01 5.30693471e-01 1.50716767e-01
5.23420870e-01 1.08915977e-02 -7.19419599e-01 -7.10418046e-01
1.61355197e-01 5.11533879e-02 2.32338667e-01 -4.17939872e-01
-1.12075901e+00 5.87092161e-01 1.64816290e-01 2.38280252e-01
-6.22390389e-01 2.76529491e-01 1.15637600e+00 4.41905111e-01
1.94666833e-01 5.35231233e-01 -3.24342340e-01 -4.70298737e-01
-9.56875920e-01 5.78214884e-01 1.47522593e+00 1.18050444e+00
9.79557216e-01 -6.91901371e-02 -8.79310369e-02 3.76904309e-01
6.63056523e-02 8.14733505e-01 -1.64162338e-01 -9.43737984e-01
3.52702707e-01 7.70954907e-01 7.66223520e-02 -8.45958769e-01
-4.23485726e-01 -5.49101597e-03 -6.59439027e-01 6.40339851e-01
-1.04593284e-01 3.49867821e-01 -6.66342318e-01 1.26451027e+00
7.90856242e-01 -7.23578036e-01 6.75405264e-02 5.72904885e-01
8.37812126e-01 5.49121499e-01 4.41766828e-02 5.59726022e-02
1.54362106e+00 -1.39879689e-01 -7.05034494e-01 3.68138313e-01
6.44981682e-01 -9.27753687e-01 1.21872306e+00 4.63156879e-01
-1.40956020e+00 -3.14599723e-01 -9.82797921e-01 -4.00328010e-01
-6.25744343e-01 -8.13304782e-01 1.12472260e+00 5.21992505e-01
-1.20633411e+00 -1.34738453e-03 -1.15006578e+00 -4.90139157e-01
7.81583264e-02 1.34610340e-01 -2.23242298e-01 -1.42948246e-02
-8.67791355e-01 7.51776218e-01 5.77549279e-01 -2.72335172e-01
-7.84191370e-01 -1.23990571e+00 -3.17156643e-01 2.10599139e-01
4.13353413e-01 -6.00906730e-01 1.02501082e+00 1.10940393e-02
-5.62738478e-01 1.03348327e+00 4.42665927e-02 -4.60361332e-01
5.56516171e-01 1.41639973e-03 -5.34999847e-01 -1.02736868e-01
6.45482838e-01 -1.22139633e-01 1.93397570e-02 -1.14871287e+00
-6.29048765e-01 -8.01026225e-01 -7.05676600e-02 -1.21717647e-01
-4.58805710e-02 4.87537086e-01 -4.11542356e-01 -2.43016317e-01
-1.00234598e-01 -5.66806734e-01 -1.08049951e-01 -1.18992507e-01
-5.72684348e-01 -4.99917716e-01 5.91964364e-01 -4.70615864e-01
1.43662727e+00 -2.24120212e+00 -4.27482516e-01 6.62556827e-01
6.01972103e-01 1.73800841e-01 4.05683428e-01 1.26721990e+00
1.23903736e-01 2.79828250e-01 -7.07640797e-02 1.70201957e-01
1.93012461e-01 4.69033509e-01 -6.80495322e-01 1.62978545e-01
-4.21575189e-01 5.61929166e-01 -6.52420104e-01 -4.31266665e-01
2.21762457e-03 9.68618020e-02 -2.75263876e-01 3.54139417e-01
-4.17458177e-01 -3.01520437e-01 -3.36540699e-01 6.91712737e-01
1.06570041e+00 -2.39656895e-01 3.57736379e-01 -1.29221335e-01
-8.06178570e-01 3.45170110e-01 -1.43513191e+00 1.38903832e+00
-4.43687528e-01 1.61548823e-01 6.96676552e-01 -5.82613051e-01
1.23610687e+00 2.70388454e-01 7.34414339e-01 -7.92630613e-01
-6.14967585e-01 4.01587009e-01 -6.10612154e-01 -7.75064409e-01
6.88055754e-01 -1.62553221e-01 -3.55945200e-01 1.03965628e+00
-7.84682572e-01 -5.99066794e-01 6.90958321e-01 7.62823164e-01
1.66001821e+00 -3.10748011e-01 1.07817650e-01 -7.09122419e-01
4.88110892e-02 7.24024713e-01 5.90454400e-01 9.00007606e-01
4.93916214e-01 2.66985297e-01 6.39581084e-01 -7.83169329e-01
-1.26645398e+00 -1.57579458e+00 -3.88034821e-01 1.29445720e+00
-4.29837927e-02 -1.11635339e+00 -2.40630612e-01 -3.90899777e-02
7.13455677e-01 9.65318501e-01 -4.48877126e-01 4.33115810e-01
-6.30749702e-01 -6.25879049e-01 7.59458303e-01 5.13156116e-01
4.79602605e-01 -9.57811475e-01 -1.04241121e+00 2.32641920e-01
3.57486367e-01 -6.72013402e-01 -3.88312005e-02 2.83273250e-01
-6.74585700e-01 -1.27752709e+00 5.08600414e-01 -1.89474180e-01
4.36085165e-02 3.00667286e-01 1.76392925e+00 3.43629181e-01
-6.15513444e-01 4.10246223e-01 -2.50427604e-01 -7.65867412e-01
-6.85707867e-01 -5.70964478e-02 -2.47279316e-01 -5.40853262e-01
8.19579840e-01 -9.39182997e-01 -2.43250787e-01 4.03782964e-01
-1.47977173e+00 -5.80958650e-02 -1.23560531e-02 2.12421224e-01
7.32476056e-01 2.86482185e-01 5.21652281e-01 -1.04017198e+00
7.34546900e-01 -7.17493713e-01 -8.82768691e-01 3.40142548e-01
-1.14276946e+00 5.40518994e-03 8.23582590e-01 5.32984793e-01
-8.82226050e-01 -2.70902142e-02 -1.67467549e-01 -1.15434848e-01
1.31550012e-02 7.11287558e-01 -1.05434351e-01 5.86876571e-01
9.64501441e-01 -1.01086445e-01 6.68562949e-01 -8.23845088e-01
6.17832839e-01 8.58824432e-01 8.61069322e-01 -8.24257910e-01
6.55629814e-01 1.13030064e+00 -2.09485874e-01 -6.45774424e-01
-1.74166799e-01 -8.20303559e-01 -1.47226095e-01 4.22787845e-01
7.00009465e-01 -8.52859676e-01 -1.14628112e+00 8.11439455e-02
-8.05446327e-01 -2.40824953e-01 -9.89083529e-01 -5.95637038e-02
-4.95870471e-01 -5.26936539e-02 -5.96432209e-01 -6.24473393e-01
-6.43063784e-01 -6.84463561e-01 8.53079021e-01 -2.79531986e-01
-2.66127676e-01 -4.74291712e-01 4.96359169e-01 2.95341820e-01
8.85147274e-01 5.27718961e-01 7.31247723e-01 -8.82650256e-01
-1.02531219e+00 -3.59970629e-01 -3.33053052e-01 -3.70385982e-02
-1.50027871e-01 3.91979426e-01 -6.33606017e-01 -3.12765211e-01
6.24344833e-02 -2.16324285e-01 2.00942650e-01 -2.52666086e-01
8.66564751e-01 -2.01393455e-01 -4.08915192e-01 6.82806790e-01
1.55189168e+00 5.81968240e-02 7.95210719e-01 6.19365156e-01
5.14261484e-01 4.17669803e-01 7.63902307e-01 7.40905464e-01
6.96757495e-01 5.75552940e-01 5.03919244e-01 -2.59419769e-01
5.18422686e-02 8.57472494e-02 -1.28009602e-01 8.47436309e-01
3.08023602e-01 3.04758400e-01 -1.43560290e+00 6.67317331e-01
-1.85450685e+00 -1.07819104e+00 -7.41328478e-01 2.22790408e+00
1.03114796e+00 2.71645933e-03 2.40604892e-01 8.26170743e-02
4.35622156e-01 -2.23712507e-03 -5.84912479e-01 -7.87117720e-01
-1.79769173e-01 3.18167359e-01 6.11271203e-01 2.27052376e-01
-5.71335495e-01 4.74290580e-01 6.83398485e+00 1.49903774e-01
-1.00565338e+00 2.24712878e-01 6.06012121e-02 -3.44840705e-01
-8.92678261e-01 1.76919296e-01 -7.99439132e-01 4.92959559e-01
1.15282643e+00 -6.60254657e-01 4.57799971e-01 9.38216984e-01
3.49988073e-01 -2.65295535e-01 -1.15025532e+00 7.33187854e-01
-4.86826092e-01 -1.74646497e+00 -4.09857810e-01 1.07747972e-01
4.29385841e-01 6.47108972e-01 -5.93107224e-01 7.73930624e-02
1.16948509e+00 -1.00243747e+00 7.43140936e-01 6.84745312e-01
6.68505192e-01 -5.68799913e-01 5.56821406e-01 1.68687433e-01
-9.89489377e-01 4.25246693e-02 -1.76542491e-01 -1.11175224e-01
1.29208878e-01 1.40108931e+00 -1.07850063e+00 6.62057042e-01
1.50895083e+00 1.95992693e-01 -6.17485285e-01 8.60582888e-01
4.29311007e-01 4.64319289e-01 -9.83089983e-01 2.12375000e-01
-2.94413656e-01 -7.59507567e-02 2.59946436e-01 1.31195259e+00
8.26896951e-02 -1.09927714e-01 2.20368639e-01 8.26567709e-01
3.30373555e-01 1.31224558e-01 -6.71724916e-01 6.28612936e-02
1.05508685e+00 1.00307739e+00 -4.71238703e-01 -9.30827022e-01
-2.77768701e-01 -1.29306078e-01 5.53413332e-02 2.10189581e-01
-5.69966793e-01 -6.27412617e-01 8.64530385e-01 7.24860311e-01
3.44880998e-01 -6.51071072e-02 -9.47790384e-01 -7.59252369e-01
4.02880996e-01 -1.22387779e+00 6.59241438e-01 -7.75191605e-01
-1.31181991e+00 6.64051294e-01 2.58927137e-01 -5.50127566e-01
-3.72547597e-01 2.50334918e-01 -4.21716601e-01 9.68757153e-01
-7.35887885e-01 -8.67817283e-01 -7.15395272e-01 5.71964025e-01
-3.31880897e-01 1.23530984e-01 8.00643682e-01 4.39870507e-01
-2.37519965e-01 1.12119064e-01 4.30134982e-01 -3.53683352e-01
5.46455383e-01 -1.29304302e+00 6.63197756e-01 6.87326252e-01
-2.37169638e-01 1.20089424e+00 1.02370262e+00 -7.46620655e-01
-1.93749285e+00 -7.94759750e-01 6.14086390e-01 -7.09198415e-01
6.87172592e-01 -5.62213838e-01 -1.30066836e+00 8.52565646e-01
8.74456018e-02 2.57473916e-01 4.49524909e-01 3.20544749e-01
-7.23478258e-01 -1.03263378e+00 -1.20368671e+00 -1.29027188e-01
7.25334883e-01 -6.89135492e-01 -5.71144640e-01 6.74941540e-01
6.49955511e-01 -2.49709815e-01 -1.63320434e+00 3.67932022e-01
4.67222184e-01 -1.57210565e+00 1.04276597e+00 -5.34322977e-01
1.55408075e-02 -7.03708649e-01 -4.44007367e-01 -7.76274681e-01
-4.33644243e-02 -7.20520675e-01 1.76843777e-01 1.51682329e+00
-1.42081613e-02 -9.16045189e-01 7.22520769e-01 1.03015137e+00
-2.32251987e-01 -4.48353499e-01 -7.22415447e-01 -9.96657133e-01
-2.36299187e-01 -4.79640424e-01 1.41700029e+00 9.56533551e-01
-1.19152099e-01 5.25586633e-03 2.89436907e-01 5.62286228e-02
5.06430328e-01 5.14522493e-01 1.69848418e+00 -1.36821687e+00
-3.25799048e-01 6.59879521e-02 -3.48987311e-01 -2.77729362e-01
-6.09272659e-01 -1.03825998e+00 -1.39658317e-01 -1.57120037e+00
-9.77159757e-03 -1.56399143e+00 -2.98160799e-02 6.66510403e-01
1.76016629e-01 -4.15256210e-02 2.91116536e-01 6.78960383e-01
-4.63870019e-01 -1.85877606e-01 5.86001635e-01 3.13512310e-02
-3.40555996e-01 -1.08135924e-01 -7.98094571e-01 2.42961496e-01
4.47457373e-01 -6.47453368e-01 -4.30215418e-01 -4.12042618e-01
8.39926779e-01 1.34888843e-01 4.66690272e-01 -9.80837464e-01
5.20914376e-01 -2.01853201e-01 -2.53727406e-01 -9.18794274e-01
5.06288409e-02 -7.40830898e-01 1.29620898e+00 3.47112209e-01
-1.26041681e-01 5.95327139e-01 2.13049948e-01 2.03254968e-01
-4.09121424e-01 1.58648059e-01 3.55278492e-01 -3.42061847e-01
-3.99509430e-01 1.27968506e-03 -1.18667008e-02 4.69469249e-01
1.12389612e+00 -3.83806154e-02 -6.71580791e-01 -8.42892081e-02
-5.60597301e-01 4.82560515e-01 1.06545722e+00 1.88388795e-01
-2.01744232e-02 -1.05741298e+00 -8.08187187e-01 4.52878803e-01
2.66562819e-01 4.77249831e-01 -1.23689145e-01 8.76869202e-01
-1.05803502e+00 2.43808359e-01 -9.87208784e-02 -8.57028246e-01
-1.19749427e+00 9.90448534e-01 -3.81825045e-02 -2.73275584e-01
-9.56884205e-01 4.36747968e-01 -1.45254374e-01 -6.76552713e-01
-4.21997189e-04 -2.98537433e-01 9.17950273e-01 1.40544400e-01
6.86707616e-01 5.74685454e-01 6.57581151e-01 1.20676674e-01
-5.69085717e-01 1.41537383e-01 1.82428546e-02 1.22000530e-01
1.71509182e+00 -3.55231203e-02 -1.01368833e+00 5.34606278e-01
9.11359966e-01 6.36430442e-01 -7.36321390e-01 1.73762262e-01
1.15280531e-01 -6.05541646e-01 -4.30343956e-01 -9.52997029e-01
-8.44438791e-01 4.30438221e-01 1.83741495e-01 1.14673746e+00
1.03819144e+00 2.69714475e-01 6.33836269e-01 2.88917214e-01
5.87210059e-01 -1.06905794e+00 -7.43525565e-01 -1.23323634e-01
1.00297606e+00 -6.46796048e-01 3.96532863e-01 -3.83978307e-01
-3.19822043e-01 9.74541724e-01 3.16593915e-01 1.16175085e-01
8.33617091e-01 1.09064651e+00 3.25260580e-01 -9.13098335e-01
-1.04470384e+00 2.21442789e-01 -8.52610707e-01 4.94876921e-01
1.58546105e-01 3.46218377e-01 -3.26799124e-01 -1.12598747e-01
-6.40075862e-01 5.33649862e-01 5.24309158e-01 1.42993045e+00
-3.84323806e-01 -1.06264925e+00 -8.97115469e-01 7.54103005e-01
-2.95006126e-01 -3.84727796e-03 1.02078333e-01 1.14544189e+00
-1.79096479e-02 7.59062171e-01 6.11241400e-01 -1.50718704e-01
5.76191783e-01 -1.40642956e-01 -2.85584688e-01 -4.43486273e-01
-1.22447073e+00 -4.11584288e-01 4.05948907e-01 -8.30674767e-01
1.08180158e-01 -6.28747344e-01 -1.56289005e+00 -9.79974151e-01
3.55377048e-01 6.92722619e-01 1.00707316e+00 2.71003723e-01
9.56346154e-01 2.16517210e-01 5.95600545e-01 2.29048150e-04
-8.61186683e-01 -4.33034092e-01 -8.49750102e-01 6.99121535e-01
3.61292139e-02 -2.29042158e-01 -2.61287689e-01 -1.56342909e-01] | [9.123960494995117, 7.608835220336914] |
30085dee-aee7-4452-b671-ca71a279662f | classifying-dialogue-acts-in-multi-party-live | null | null | https://aclanthology.org/Y12-1050 | https://aclanthology.org/Y12-1050.pdf | Classifying Dialogue Acts in Multi-party Live Chats | null | ['Timothy Baldwin', 'Su Nam Kim', 'Lawrence Cavedon'] | 2012-11-01 | classifying-dialogue-acts-in-multi-party-live-1 | https://aclanthology.org/Y12-1050 | https://aclanthology.org/Y12-1050.pdf | paclic-2012-11 | ['dialogue-act-classification'] | ['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] | [-7.220917224884033, 3.8009259700775146] |
75a8588b-4b7b-4b65-ab5c-613198ed8f3b | sipmask-spatial-information-preservation-for | 2007.14772 | null | https://arxiv.org/abs/2007.14772v1 | https://arxiv.org/pdf/2007.14772v1.pdf | SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation | Single-stage instance segmentation approaches have recently gained popularity due to their speed and simplicity, but are still lagging behind in accuracy, compared to two-stage methods. We propose a fast single-stage instance segmentation method, called SipMask, that preserves instance-specific spatial information by separating mask prediction of an instance to different sub-regions of a detected bounding-box. Our main contribution is a novel light-weight spatial preservation (SP) module that generates a separate set of spatial coefficients for each sub-region within a bounding-box, leading to improved mask predictions. It also enables accurate delineation of spatially adjacent instances. Further, we introduce a mask alignment weighting loss and a feature alignment scheme to better correlate mask prediction with object detection. On COCO test-dev, our SipMask outperforms the existing single-stage methods. Compared to the state-of-the-art single-stage TensorMask, SipMask obtains an absolute gain of 1.0% (mask AP), while providing a four-fold speedup. In terms of real-time capabilities, SipMask outperforms YOLACT with an absolute gain of 3.0% (mask AP) under similar settings, while operating at comparable speed on a Titan Xp. We also evaluate our SipMask for real-time video instance segmentation, achieving promising results on YouTube-VIS dataset. The source code is available at https://github.com/JialeCao001/SipMask. | ['Ling Shao', 'Fahad Shahbaz Khan', 'Jiale Cao', 'Hisham Cholakkal', 'Rao Muhammad Anwer', 'Yanwei Pang'] | 2020-07-29 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/2057_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123590001.pdf | eccv-2020-8 | ['real-time-instance-segmentation', 'video-instance-segmentation'] | ['computer-vision', 'computer-vision'] | [ 2.13376343e-01 -1.82317328e-02 -4.21765625e-01 -2.00651288e-01
-1.04597175e+00 -6.09253824e-01 3.55388761e-01 8.73115659e-02
-3.36892188e-01 2.70295650e-01 -2.21291035e-01 -1.96125552e-01
1.22908615e-01 -4.78335679e-01 -7.62160480e-01 -5.28447509e-01
-9.81090814e-02 3.85577440e-01 9.50228214e-01 2.93579310e-01
3.27654779e-01 2.72579402e-01 -1.38346744e+00 6.58711970e-01
7.49628365e-01 1.20081949e+00 1.97097257e-01 6.27419472e-01
7.01019261e-03 4.95434463e-01 -5.10855496e-01 -2.85316527e-01
4.00302351e-01 -1.08957745e-01 -9.35635924e-01 1.83659226e-01
8.05888653e-01 -3.99637580e-01 -4.02426839e-01 6.54970765e-01
2.08928913e-01 -4.64319475e-02 3.00695807e-01 -1.20501244e+00
1.50422296e-02 4.53208774e-01 -1.08092260e+00 4.91349369e-01
-7.15144947e-02 1.58337072e-01 1.05000484e+00 -9.22459066e-01
5.23631513e-01 7.97789335e-01 6.65436208e-01 3.47193331e-01
-1.41415703e+00 -6.44791961e-01 3.08732510e-01 2.85966456e-01
-1.58932519e+00 -4.07289177e-01 6.02780521e-01 -4.20277596e-01
9.28985178e-01 5.15040100e-01 6.12518907e-01 6.46738172e-01
6.25267625e-02 1.28004801e+00 1.06059766e+00 -1.14549108e-01
2.28068128e-01 -6.14997223e-02 1.42435446e-01 7.61371017e-01
-2.96360124e-02 -1.77287117e-01 -6.80744112e-01 5.68622909e-02
8.73399258e-01 -1.28995217e-02 -3.43768626e-01 -4.00079548e-01
-1.39828968e+00 4.19210464e-01 4.52758402e-01 3.62112880e-01
-3.00619155e-01 3.19449663e-01 4.53648418e-01 -1.55805781e-01
6.53866291e-01 2.46598899e-01 -5.33772945e-01 -4.17425096e-01
-1.61556244e+00 1.79357812e-01 5.01937389e-01 9.52959359e-01
6.76836133e-01 -8.50747898e-02 -4.75807101e-01 8.20735991e-01
7.54751191e-02 1.54296294e-01 4.84481245e-01 -1.06784344e+00
5.73383749e-01 7.51461685e-01 -1.19112268e-01 -9.47455764e-01
-2.81191885e-01 -5.45828164e-01 -4.90550995e-01 1.22061051e-01
5.42770267e-01 1.83620840e-01 -1.05579257e+00 1.27221441e+00
6.79458261e-01 7.83485532e-01 -3.51632208e-01 1.02634549e+00
6.21931255e-01 8.19293022e-01 -2.13113382e-01 -4.72709397e-03
1.55799937e+00 -1.41022527e+00 -2.73222774e-01 -3.13823313e-01
6.33579850e-01 -6.90838456e-01 1.09051549e+00 6.05432391e-01
-1.02903020e+00 -4.45317179e-01 -9.59938347e-01 1.70929674e-02
-1.75352007e-01 3.18832755e-01 5.43466628e-01 5.84613025e-01
-9.94432330e-01 5.88185251e-01 -1.09640086e+00 -1.59950927e-01
7.26634085e-01 4.08668756e-01 -2.98980385e-01 5.51266409e-02
-5.60313463e-01 2.75493860e-01 2.55534261e-01 -5.26487455e-02
-8.52957964e-01 -1.23748255e+00 -8.23822796e-01 5.08918948e-02
7.69133985e-01 -1.48460791e-01 1.37597370e+00 -8.66651833e-01
-1.32633746e+00 7.17662632e-01 -3.07221860e-01 -5.90834856e-01
5.35970807e-01 -2.38518491e-01 -1.88978747e-01 4.40990090e-01
2.99498647e-01 9.25048232e-01 8.87183845e-01 -1.20369625e+00
-9.61964428e-01 -2.66622156e-01 1.41278775e-02 1.45295098e-01
-4.08717215e-01 -1.41131254e-02 -1.26058435e+00 -9.26450312e-01
9.00306776e-02 -9.02395725e-01 -2.37154260e-01 -1.38232755e-02
-5.73790312e-01 -1.21494094e-02 1.03743076e+00 -7.43468106e-01
1.60323429e+00 -2.20766449e+00 5.52426502e-02 1.16631933e-01
2.99972177e-01 5.98927021e-01 -9.81242508e-02 1.10839657e-01
6.61452040e-02 -2.06839107e-03 -6.06476665e-01 -6.73619986e-01
-2.03927219e-01 9.25799757e-02 -2.73058325e-01 4.16940987e-01
2.73340166e-01 9.60982680e-01 -7.10927844e-01 -5.23003757e-01
5.12852907e-01 4.35787201e-01 -5.77255547e-01 -8.55082199e-02
-2.06226438e-01 2.76017249e-01 -1.74596950e-01 9.25675213e-01
7.73943007e-01 -2.67894506e-01 1.58184692e-01 -2.64165252e-01
-1.83647364e-01 1.38556823e-01 -1.33049369e+00 1.75358403e+00
-3.18377107e-01 7.93958485e-01 4.27477248e-02 -1.00603330e+00
5.46750307e-01 7.23378360e-02 5.96577525e-01 -7.99591660e-01
-9.01464596e-02 2.25473702e-01 -1.92884624e-01 -6.49313107e-02
6.32252514e-01 3.93376023e-01 -5.66175543e-02 2.75625676e-01
-1.28713608e-01 1.84155941e-01 4.77246642e-01 3.73257220e-01
1.23095119e+00 3.47028941e-01 3.39776762e-02 -3.75283837e-01
3.31221431e-01 1.31236404e-01 7.20179379e-01 5.37584901e-01
-2.89366275e-01 1.02185524e+00 5.45728385e-01 -4.27131087e-01
-8.38565230e-01 -8.94122183e-01 -3.03172588e-01 8.90335858e-01
4.10158932e-01 -6.85297191e-01 -1.06312478e+00 -1.00947225e+00
9.29369405e-02 4.34017092e-01 -6.17598414e-01 3.63058746e-01
-7.49628246e-01 -7.73068905e-01 5.58816195e-01 6.49461567e-01
6.07788324e-01 -8.74223232e-01 -7.44664013e-01 3.30277711e-01
-7.48577416e-02 -1.38981009e+00 -7.48644948e-01 -1.08013101e-01
-9.06119883e-01 -1.03563201e+00 -8.88151586e-01 -5.47918379e-01
6.69297874e-01 4.73628491e-01 1.06226695e+00 1.35097712e-01
-5.73899031e-01 2.02802852e-01 -4.10810411e-01 3.47223170e-02
-9.92870890e-03 2.61368096e-01 -2.35003173e-01 1.74825743e-01
7.80250877e-02 -3.58463883e-01 -9.71212387e-01 7.41259873e-01
-1.06438351e+00 2.92406082e-01 4.99158144e-01 5.87400317e-01
1.03207660e+00 -4.91030477e-02 2.85482883e-01 -8.30743313e-01
-9.69749913e-02 -3.24112535e-01 -7.34313011e-01 1.34299085e-01
-6.15536332e-01 -2.96033502e-01 3.88726592e-01 -3.71647477e-01
-8.34648371e-01 2.39063159e-01 2.30211932e-02 -4.75324631e-01
-1.46545291e-01 1.58598408e-01 5.70547394e-02 -4.23168093e-02
2.59814113e-01 2.54072160e-01 2.05086097e-02 -5.91534376e-01
2.39745185e-01 5.23747861e-01 4.61706758e-01 -4.74524140e-01
6.63313985e-01 7.18159974e-01 -2.46098369e-01 -6.85611606e-01
-7.38150537e-01 -7.75946915e-01 -6.16486490e-01 -2.62289494e-01
6.44317806e-01 -8.81886303e-01 -4.51743186e-01 6.20178044e-01
-7.48514056e-01 -7.97735631e-01 -1.96613565e-01 1.31151184e-01
-2.59752661e-01 4.34885323e-01 -7.22071409e-01 -3.44678074e-01
-3.28741521e-01 -1.28746581e+00 1.37291360e+00 2.18164846e-01
-1.66730911e-01 -6.80685759e-01 -1.11670822e-01 7.09795475e-01
1.86087236e-01 1.30542323e-01 3.00497144e-01 -4.22158599e-01
-8.40066791e-01 -2.02919692e-01 -3.43136460e-01 3.37518126e-01
2.79550087e-02 1.95123658e-01 -8.48589420e-01 -4.53785330e-01
-4.83519495e-01 1.36645839e-01 1.05405784e+00 5.57881653e-01
1.42894971e+00 -2.15326205e-01 -5.94423711e-01 6.74790442e-01
1.33102369e+00 1.46032140e-01 6.92023456e-01 4.37564701e-01
8.50804806e-01 3.44887257e-01 1.06167161e+00 3.35764647e-01
3.19543272e-01 1.15856934e+00 4.93962348e-01 -3.75228077e-01
-3.97257805e-01 5.92846498e-02 2.82562494e-01 5.57862759e-01
2.62229685e-02 -1.59057796e-01 -9.52716410e-01 8.46510410e-01
-2.03946877e+00 -7.00683594e-01 -2.95604646e-01 2.26885581e+00
6.92229390e-01 3.25526476e-01 5.38601398e-01 2.80130386e-01
6.85798407e-01 3.59016031e-01 -5.36791623e-01 -1.31809592e-01
1.36761189e-01 3.11494946e-01 5.84486783e-01 5.46009839e-01
-1.30242288e+00 1.11463928e+00 5.22446251e+00 1.37701273e+00
-1.11221099e+00 2.23812461e-01 9.86570835e-01 -4.47534770e-01
1.25628442e-01 4.30294732e-03 -8.40674400e-01 7.66389847e-01
6.36723220e-01 2.42780566e-01 1.40579998e-01 8.93895388e-01
1.91069558e-01 -4.30174470e-01 -8.51037681e-01 7.73133218e-01
8.28271806e-02 -1.66325343e+00 -9.94390473e-02 1.40170768e-01
6.86079323e-01 9.21344534e-02 7.68312812e-02 2.28205711e-01
-3.44163090e-01 -8.28274071e-01 9.44436729e-01 -6.66271672e-02
8.55481803e-01 -7.57964194e-01 5.63783765e-01 1.01179659e-01
-1.56983805e+00 6.70941770e-02 -1.21803030e-01 3.65876630e-02
1.77770495e-01 7.47531116e-01 -8.42160583e-01 5.83238423e-01
9.20587420e-01 6.69876814e-01 -5.43988824e-01 1.28105581e+00
-1.47016078e-01 9.48521376e-01 -3.48427773e-01 4.17070031e-01
4.09021944e-01 -4.71481346e-02 5.99267244e-01 1.55334032e+00
1.40810102e-01 1.02287270e-01 1.46353945e-01 7.16442883e-01
-2.19341274e-02 2.91409623e-02 -2.93037985e-02 2.39528701e-01
6.00579798e-01 1.49294686e+00 -1.19776678e+00 -5.85866153e-01
-3.42119902e-01 1.25978935e+00 1.08924940e-01 1.51295081e-01
-1.24769437e+00 -2.97929853e-01 8.71960580e-01 4.34625536e-01
8.38635683e-01 -1.22474305e-01 -5.59096754e-01 -9.18119252e-01
3.70660484e-01 -7.76748240e-01 4.55051422e-01 -3.54700178e-01
-6.90784335e-01 6.10408306e-01 6.71220152e-03 -1.36120200e+00
1.88393638e-01 -6.28900826e-01 -5.60484707e-01 4.53378171e-01
-1.43681598e+00 -1.20679510e+00 -4.25517321e-01 2.36986890e-01
9.01894033e-01 1.70877427e-01 5.51225364e-01 5.57905078e-01
-1.05090070e+00 8.44985783e-01 5.57674654e-02 1.95191443e-01
5.44612706e-01 -1.17919958e+00 6.11313105e-01 1.02923954e+00
4.19228017e-01 3.80710363e-01 4.22018498e-01 -5.24414539e-01
-1.06698453e+00 -1.30111551e+00 4.48677957e-01 -4.21027243e-01
4.44904089e-01 -3.71270269e-01 -9.37653601e-01 6.42451048e-01
1.40618589e-02 2.91754782e-01 5.15698254e-01 -1.17540300e-01
-4.32843566e-01 -3.97890449e-01 -1.12307119e+00 5.82942784e-01
9.44204807e-01 -1.09521002e-01 -1.29296407e-01 2.51544565e-01
7.62386739e-01 -7.96011508e-01 -8.95133376e-01 5.79245508e-01
5.50474882e-01 -1.20260704e+00 9.99669254e-01 -5.98283894e-02
5.16569972e-01 -6.77022457e-01 7.32802376e-02 -9.40008879e-01
-2.93151945e-01 -6.92660213e-01 -3.48729789e-01 1.30317068e+00
4.74209845e-01 -4.19874430e-01 1.15729845e+00 4.73292708e-01
-3.03033441e-01 -1.33803856e+00 -1.09644401e+00 -8.61409843e-01
-2.92455047e-01 -7.59148657e-01 5.31503439e-01 8.75058651e-01
-7.99629092e-02 -1.20947659e-01 -2.59247392e-01 2.90076643e-01
5.28801143e-01 2.88765758e-01 7.56659150e-01 -6.63338125e-01
-4.44524050e-01 -5.48591495e-01 -5.22550583e-01 -1.34010100e+00
-2.10877135e-01 -6.52766764e-01 -6.37187138e-02 -1.56897068e+00
2.75302798e-01 -5.24467170e-01 -4.07030970e-01 7.88702428e-01
-2.29088381e-01 1.08525491e+00 3.31025690e-01 2.57119715e-01
-8.28798473e-01 1.66527867e-01 9.13287938e-01 -1.21928789e-02
-2.92544931e-01 -7.20197260e-02 -5.50354600e-01 6.58103943e-01
8.60119760e-01 -4.37706709e-01 -2.46059045e-01 -4.09445524e-01
-3.45265001e-01 -1.72157168e-01 3.69610995e-01 -1.22245884e+00
4.41241125e-03 1.97565407e-02 2.67570108e-01 -7.83863783e-01
4.84025359e-01 -6.15295529e-01 1.11538388e-01 4.81147379e-01
9.75451693e-02 -5.59161417e-02 5.84291875e-01 4.09468561e-01
-2.04750970e-01 7.08602443e-02 7.04811573e-01 1.83451369e-01
-1.12301576e+00 3.63343179e-01 -2.60405362e-01 -2.75134202e-02
1.42897582e+00 -5.93730628e-01 -3.46766621e-01 1.18495993e-01
-4.47151244e-01 3.98609161e-01 5.88743091e-01 3.73156548e-01
5.81790268e-01 -1.00076377e+00 -6.15573704e-01 1.13861322e-01
9.56121981e-02 2.17204168e-01 5.35587490e-01 1.11152124e+00
-7.76463270e-01 3.11880618e-01 1.06325753e-01 -9.43254173e-01
-1.62911642e+00 2.21884727e-01 1.18074015e-01 -3.07591200e-01
-8.58166039e-01 1.07790065e+00 4.37693238e-01 -1.40571669e-01
2.20993325e-01 -4.80351299e-01 1.34460598e-01 -3.35947908e-02
7.01711953e-01 5.43269575e-01 2.77761608e-01 -6.51333392e-01
-6.57256186e-01 6.18981898e-01 -3.34016412e-01 -3.33189629e-02
1.22443974e+00 1.00423560e-01 7.84876943e-02 6.99437410e-02
9.63594317e-01 7.00876042e-02 -1.62051892e+00 -2.08320379e-01
-1.43320495e-02 -8.09501946e-01 5.31328544e-02 -8.68883550e-01
-1.44110990e+00 5.54809690e-01 4.37107474e-01 1.23197831e-01
1.17142665e+00 1.96754351e-01 1.02543724e+00 -3.76528025e-01
3.81374806e-01 -9.30851400e-01 1.95931539e-01 2.03037664e-01
6.30099773e-01 -1.09611309e+00 6.61491379e-02 -8.55567753e-01
-7.67424107e-01 8.11247885e-01 6.78157866e-01 -1.50047332e-01
2.39823893e-01 3.79753470e-01 5.69794104e-02 -1.73520297e-01
-5.64083695e-01 3.87217253e-02 5.73558331e-01 2.90144175e-01
3.00799012e-01 2.17556044e-01 -2.40288302e-01 4.58556622e-01
5.92933446e-02 -2.80371964e-01 3.55052829e-01 8.19326818e-01
-3.46518368e-01 -1.01105082e+00 -3.96685332e-01 5.05256176e-01
-6.32878900e-01 -1.95413068e-01 -1.22726500e-01 8.23143363e-01
1.10552825e-01 8.49273860e-01 3.46859217e-01 -4.89997357e-01
3.11831713e-01 -2.19262376e-01 3.15846264e-01 -5.53780198e-01
-6.46858275e-01 2.43354246e-01 1.27299912e-02 -1.10028136e+00
-2.38211885e-01 -7.66351521e-01 -1.40740156e+00 -2.71031559e-01
-3.23115051e-01 1.52305206e-02 4.90312129e-01 6.82628334e-01
7.31097996e-01 6.26755655e-01 3.86449456e-01 -1.11022043e+00
2.33576577e-02 -6.98557258e-01 -3.92024726e-01 1.72369763e-01
1.31902337e-01 -5.24621010e-01 -1.68458685e-01 2.31655221e-02] | [9.325258255004883, -0.0025354502722620964] |
92de431b-49b3-463a-81c2-4914db6f4d30 | actup-analyzing-and-consolidating-tsne-and | 2305.0732 | null | https://arxiv.org/abs/2305.07320v1 | https://arxiv.org/pdf/2305.07320v1.pdf | ActUp: Analyzing and Consolidating tSNE and UMAP | tSNE and UMAP are popular dimensionality reduction algorithms due to their speed and interpretable low-dimensional embeddings. Despite their popularity, however, little work has been done to study their full span of differences. We theoretically and experimentally evaluate the space of parameters in both tSNE and UMAP and observe that a single one -- the normalization -- is responsible for switching between them. This, in turn, implies that a majority of the algorithmic differences can be toggled without affecting the embeddings. We discuss the implications this has on several theoretic claims behind UMAP, as well as how to reconcile them with existing tSNE interpretations. Based on our analysis, we provide a method (\ourmethod) that combines previously incompatible techniques from tSNE and UMAP and can replicate the results of either algorithm. This allows our method to incorporate further improvements, such as an acceleration that obtains either method's outputs faster than UMAP. We release improved versions of tSNE, UMAP, and \ourmethod that are fully plug-and-play with the traditional libraries at https://github.com/Andrew-Draganov/GiDR-DUN | ['Cigdem Aslay', 'Tyrus Berry', 'Ira Assent', 'Davide Mottin', 'Katrine Scheel Nellemann', 'Jakob Rødsgaard Jørgensen', 'Andrew Draganov'] | 2023-05-12 | null | null | null | null | ['dimensionality-reduction'] | ['methodology'] | [-1.58111677e-01 -7.79808313e-03 -2.15494975e-01 -1.39769435e-01
-4.58389759e-01 -9.54275191e-01 6.80407166e-01 6.89725056e-02
-1.36846229e-01 4.62629527e-01 3.40069115e-01 -6.83223784e-01
-4.82283771e-01 -5.78851342e-01 -3.07574838e-01 -6.34845793e-01
-1.22384012e-01 5.00593960e-01 3.68472606e-01 -2.54266769e-01
4.54644859e-01 3.97462994e-01 -1.60503256e+00 -6.82782307e-02
5.87010682e-01 5.85916579e-01 -4.26179357e-02 8.58901441e-01
7.44369850e-02 4.74226683e-01 -6.50484264e-02 -4.39123929e-01
6.68648601e-01 -3.96796197e-01 -9.82038736e-01 -1.74926400e-01
2.15637326e-01 -3.07721883e-01 -6.31069541e-01 8.79109085e-01
3.29547286e-01 1.81497321e-01 8.80144238e-01 -1.65931308e+00
-9.01407599e-01 4.83929306e-01 -4.58214581e-01 2.04627931e-01
2.73175418e-01 2.92520486e-02 1.26819611e+00 -1.00329328e+00
5.99230170e-01 1.22193372e+00 8.69138300e-01 4.41041231e-01
-1.29884243e+00 -4.63314116e-01 -1.82803005e-01 1.73018500e-01
-1.47048461e+00 -6.95928574e-01 6.78516030e-01 -4.11279231e-01
9.06462193e-01 4.10198063e-01 4.24987137e-01 7.90881455e-01
2.28507239e-02 7.97792137e-01 9.44979429e-01 -6.12181962e-01
2.59843796e-01 2.91392446e-01 3.11805248e-01 5.76381862e-01
6.27784491e-01 1.17092654e-01 -7.13812485e-02 -4.17720705e-01
6.01544797e-01 -2.33767763e-01 -1.99456170e-01 -9.20713365e-01
-1.12069666e+00 9.17933583e-01 -2.27738600e-02 2.41902083e-01
8.95453915e-02 1.92910761e-01 4.23400760e-01 4.14317936e-01
3.25905681e-01 6.85270011e-01 -4.81577784e-01 -6.82638049e-01
-6.28374159e-01 5.24967432e-01 9.29172456e-01 8.97075713e-01
5.92492402e-01 -3.87174696e-01 5.68317473e-01 5.92276335e-01
2.02811778e-01 9.11285430e-02 4.47386593e-01 -1.14265013e+00
2.03370929e-01 4.42496568e-01 2.95628965e-01 -1.08373988e+00
-4.22044605e-01 2.03642040e-01 -3.42567742e-01 1.03035495e-01
3.95467907e-01 4.88795014e-03 -5.16725302e-01 1.61945415e+00
3.62690300e-01 1.77625746e-01 7.07001165e-02 7.13412523e-01
2.92416960e-01 5.02323806e-01 -3.37843627e-01 8.17530230e-03
1.33831489e+00 -7.46102273e-01 -5.51079154e-01 2.09981631e-02
1.21352541e+00 -8.71433616e-01 1.25016141e+00 4.11620408e-01
-9.31810737e-01 -1.13450371e-01 -1.27035403e+00 -4.23032463e-01
-5.34122288e-01 -2.06803739e-01 8.25067461e-01 8.45705092e-01
-1.16292775e+00 8.95266354e-01 -1.14880705e+00 -7.67594516e-01
1.09709471e-01 3.82003337e-01 -2.33961925e-01 2.67102093e-01
-1.09468102e+00 9.75245059e-01 3.87353361e-01 -1.67423829e-01
-9.55362692e-02 -5.67065954e-01 -5.92351973e-01 -1.81297556e-01
2.84749746e-01 -4.95512545e-01 1.36410749e+00 -2.91852444e-01
-1.24285150e+00 6.14062965e-01 -2.16697127e-01 -3.29986364e-01
5.41064858e-01 -1.83125019e-01 -2.00271308e-01 1.49812549e-01
-1.35936588e-02 5.10140717e-01 2.45787695e-01 -1.23058069e+00
-5.18241405e-01 -4.20210183e-01 1.00003287e-01 2.16983750e-01
-6.49318635e-01 8.07708874e-02 -4.58626032e-01 -5.80168366e-01
8.74936506e-02 -1.28800797e+00 -8.83834362e-02 1.23829640e-01
-3.65945041e-01 -3.12796801e-01 8.78861308e-01 -2.72902280e-01
1.53388965e+00 -2.29555035e+00 2.10271642e-01 1.31459534e-01
4.85879302e-01 3.57231826e-01 -1.54238224e-01 8.96917582e-01
-3.55037153e-01 6.35370612e-01 -2.55561680e-01 -5.01556695e-01
3.44667405e-01 4.10034090e-01 -3.98580551e-01 7.03704536e-01
-1.76214743e-02 6.36946023e-01 -8.52592826e-01 -2.49635085e-01
5.46534777e-01 3.71233881e-01 -6.49423778e-01 -1.45979434e-01
7.49624521e-02 5.72825596e-03 -1.56158417e-01 3.55395079e-01
7.28644133e-01 -3.39621282e-03 4.38303113e-01 -5.54109216e-02
-4.07775015e-01 5.42131007e-01 -1.45937932e+00 1.32964921e+00
-9.74230990e-02 7.81114578e-01 -1.42508999e-01 -9.49590981e-01
6.59237742e-01 1.10908382e-01 3.45112830e-01 -3.30369979e-01
2.35704511e-01 3.51553291e-01 -1.93915237e-02 -5.31652153e-01
7.77981281e-01 -2.03484073e-01 -3.96028385e-02 7.97753870e-01
-2.02966273e-01 -1.35857552e-01 3.07922572e-01 3.21316570e-01
1.09184086e+00 8.90530124e-02 5.11458933e-01 -4.48963493e-01
-6.25078566e-03 -3.84645090e-02 4.60377604e-01 7.52590179e-01
-4.71285641e-01 7.71564901e-01 8.13092113e-01 -2.44569108e-01
-1.49538362e+00 -1.02689409e+00 -4.16663975e-01 1.06657338e+00
2.25629717e-01 -7.98953712e-01 -5.73072612e-01 -4.25301075e-01
7.86889270e-02 9.77354944e-01 -7.54016876e-01 -1.81060925e-01
-3.98694873e-01 -8.85354698e-01 8.35677505e-01 7.10180044e-01
-4.94279666e-03 -5.06411374e-01 -4.94679540e-01 -1.79654315e-01
1.63193002e-01 -7.81196773e-01 -2.43133128e-01 2.89938897e-01
-8.75662625e-01 -1.23538685e+00 -4.84153032e-02 -6.41590834e-01
4.61682260e-01 4.89514738e-01 6.25238776e-01 7.83113297e-03
-1.34034768e-01 3.40480834e-01 -5.49590886e-01 -3.45490277e-01
-4.91074681e-01 2.27517247e-01 5.16444623e-01 -5.04414141e-01
7.61222959e-01 -7.99095452e-01 -4.20214385e-01 4.83511835e-01
-1.08593142e+00 -1.45491779e-01 2.97502756e-01 5.92056453e-01
1.62544772e-01 2.50543535e-01 2.25356922e-01 -8.34052026e-01
9.54685509e-01 -6.04586661e-01 -5.41072726e-01 4.67796773e-02
-9.03027594e-01 3.49831581e-01 5.97286165e-01 -2.93409109e-01
-4.08193380e-01 -2.20316395e-01 -2.29014516e-01 -5.57147026e-01
-1.69130303e-02 2.43246943e-01 -1.20722033e-01 4.17173393e-02
6.06631041e-01 -9.83465277e-03 2.04163864e-01 -7.00174093e-01
7.17782795e-01 9.28060591e-01 4.14721400e-01 -6.88747764e-01
9.78331625e-01 5.06277919e-01 -3.12174112e-01 -8.53749275e-01
-4.68006551e-01 -4.82676238e-01 -7.93012619e-01 4.79037344e-01
5.08177817e-01 -4.59043264e-01 -4.24313486e-01 3.58723015e-01
-9.48865354e-01 -2.02946171e-01 -3.95595163e-01 4.57069963e-01
-5.23435950e-01 6.96665049e-01 -4.72609550e-01 -8.41642201e-01
2.20733315e-01 -1.19085312e+00 6.26318514e-01 -2.31645047e-03
-5.73016644e-01 -1.32309711e+00 4.10162807e-01 -2.02539423e-03
2.28904769e-01 -8.40321481e-02 1.11325693e+00 -1.16023731e+00
3.92258987e-02 -1.89334586e-01 -2.99991399e-01 4.28786963e-01
2.24404186e-01 3.87810498e-01 -9.73035395e-01 -3.89882475e-01
8.99046212e-02 -3.92897055e-02 6.73917353e-01 1.24286398e-01
1.01192951e+00 -3.35471064e-01 -3.98063242e-01 6.27227366e-01
1.34440196e+00 1.62893921e-01 7.06881881e-01 5.41834712e-01
4.83982921e-01 4.43286240e-01 3.78203273e-01 4.29213107e-01
4.68590111e-01 7.02269733e-01 2.56023556e-01 1.20134734e-01
4.56761010e-03 -2.64769346e-01 4.25301254e-01 1.15604472e+00
2.93616727e-02 -2.89952666e-01 -1.02539885e+00 7.86824822e-01
-1.94803309e+00 -9.78612483e-01 -1.99434400e-01 2.38458037e+00
5.29965937e-01 1.06898658e-01 4.27759796e-01 3.88678104e-01
6.11101270e-01 2.01101646e-01 -5.14720619e-01 -9.07127917e-01
1.17508225e-01 1.61875367e-01 7.08264828e-01 7.00352848e-01
-1.03179967e+00 7.91363895e-01 7.11418486e+00 5.17292619e-01
-7.94395149e-01 -7.67429546e-02 2.32811213e-01 -2.48042077e-01
-4.46717650e-01 2.83915609e-01 -7.11213410e-01 3.41936767e-01
1.06098831e+00 -5.51065147e-01 5.87101996e-01 8.58859718e-01
2.84576803e-01 1.42844841e-01 -1.41313136e+00 7.93127239e-01
5.42293712e-02 -1.08270907e+00 3.59719875e-03 3.57077628e-01
3.92619878e-01 6.66661412e-02 1.93159997e-01 2.59566724e-01
5.28867960e-01 -8.67509902e-01 5.64962447e-01 -1.80580363e-01
5.24638653e-01 -7.33339369e-01 6.50137842e-01 1.09997112e-02
-1.14897120e+00 -2.11768612e-01 -5.87871075e-01 -4.17164326e-01
-1.02134012e-01 4.35232937e-01 -9.22736585e-01 5.92677236e-01
3.58700931e-01 6.11961842e-01 -5.60938001e-01 8.33638668e-01
-1.73236117e-01 6.44506276e-01 -5.32818973e-01 4.40392569e-02
1.55044124e-01 -3.45249504e-01 6.34479105e-01 1.11639345e+00
3.26953113e-01 1.13415480e-01 2.27836370e-02 7.52234817e-01
1.30549550e-01 -1.55420080e-01 -7.85818040e-01 -4.20421571e-01
1.01108754e+00 1.03766489e+00 -6.74587190e-01 -1.64669320e-01
-4.52712297e-01 7.45250583e-01 2.36413121e-01 5.79499686e-03
-9.33044553e-01 -7.83552170e-01 1.22682762e+00 2.40534097e-01
4.99935418e-01 -4.48809028e-01 -5.31783760e-01 -1.27006423e+00
9.54040736e-02 -8.05414438e-01 3.32073718e-01 -5.04749000e-01
-1.36819828e+00 2.68712401e-01 2.33679622e-01 -1.18847454e+00
-2.58822620e-01 -8.38886797e-01 -4.46890861e-01 5.12750387e-01
-1.18776107e+00 -7.13828743e-01 1.58921912e-01 2.74316788e-01
2.94445932e-01 2.79489398e-01 8.37198913e-01 2.32565701e-01
-7.66021132e-01 6.20679080e-01 5.31391203e-01 9.84632447e-02
7.80016303e-01 -1.32540596e+00 6.37058854e-01 9.35417354e-01
2.05057431e-02 9.54260647e-01 1.03755736e+00 -3.65740389e-01
-1.54926956e+00 -7.99152017e-01 7.34616637e-01 -9.67225790e-01
1.25963986e+00 -4.90385264e-01 -8.00095975e-01 1.01133406e+00
2.26016060e-01 -2.48460323e-01 8.76575232e-01 5.39664090e-01
-4.34577852e-01 2.63371140e-01 -9.38765883e-01 8.07827592e-01
1.08615708e+00 -5.70353806e-01 -1.02820075e+00 9.17306319e-02
5.86501837e-01 -1.68664649e-01 -9.57974076e-01 9.11691636e-02
6.93472028e-01 -1.07803309e+00 9.14285719e-01 -7.59812832e-01
3.26585919e-01 -2.88806289e-01 -4.39301550e-01 -1.24089444e+00
-4.10538703e-01 -6.82080090e-01 -9.61909816e-02 1.20333362e+00
3.34007472e-01 -1.04472852e+00 5.00881076e-01 8.01334739e-01
-4.96764407e-02 -7.81259596e-01 -7.37604618e-01 -1.06405771e+00
6.08875692e-01 -6.89282238e-01 6.62686884e-01 1.13152206e+00
3.78453434e-01 1.00488298e-01 -2.40630329e-01 1.17303200e-01
4.39960212e-01 -1.01215869e-01 9.18934166e-01 -1.32758331e+00
-1.37294829e-01 -4.66487795e-01 -6.55779779e-01 -1.01354492e+00
-1.52815968e-01 -8.99709523e-01 -2.88135350e-01 -1.26859653e+00
1.85723647e-01 -5.74832320e-01 -2.25081995e-01 7.74365783e-01
1.56307265e-01 2.08273157e-01 3.29929292e-01 5.36190629e-01
-4.96686190e-01 3.75456274e-01 7.70731688e-01 4.24659461e-01
-4.17737544e-01 -4.97598469e-01 -1.17511606e+00 8.23698699e-01
8.42513740e-01 -4.97923493e-01 -5.68095446e-01 -3.70548576e-01
3.02175581e-01 -5.68022668e-01 3.01591665e-01 -9.10898685e-01
6.55853823e-02 -1.02791071e-01 1.06379425e-03 -2.43265748e-01
2.43267253e-01 -3.93967330e-01 -2.90386956e-02 2.22871080e-01
-3.29781532e-01 3.18829268e-01 2.60376364e-01 3.06856990e-01
2.53142267e-01 -3.03954154e-01 6.87449038e-01 2.21073683e-02
-6.81358159e-01 -1.49884000e-02 -3.41429949e-01 2.79608428e-01
9.94350851e-01 -2.88371176e-01 -5.96010923e-01 -2.12716937e-01
-5.71831703e-01 5.20886853e-02 1.05229497e+00 6.25258625e-01
3.10240716e-01 -1.34255838e+00 -4.81352448e-01 1.71795517e-01
6.39594942e-02 -1.26744583e-01 -1.78888202e-01 1.13517058e+00
-5.85233390e-01 4.77002472e-01 1.89807061e-02 -1.72570407e-01
-1.22328222e+00 7.25184679e-01 1.95760489e-01 1.93590950e-02
-7.26439655e-01 5.42508662e-01 1.07716031e-01 -6.79469049e-01
-1.13190554e-01 -4.09507900e-01 3.13035369e-01 1.74275324e-01
6.61720932e-01 5.67047894e-01 -7.21449479e-02 -3.49140316e-01
-5.47433913e-01 3.64409357e-01 -2.53961951e-01 -1.83288679e-01
1.49088669e+00 -4.05988008e-01 -2.15328977e-01 7.28036046e-01
1.38700747e+00 1.91190660e-01 -8.60470414e-01 1.59609109e-01
-9.07908306e-02 -6.04668498e-01 -9.68510583e-02 -3.54831010e-01
-5.20157099e-01 7.95277655e-01 5.17851651e-01 6.53604567e-01
9.31067407e-01 1.37404814e-01 7.22249925e-01 4.48173106e-01
1.43760204e-01 -1.02554238e+00 -3.68633717e-01 3.56517971e-01
5.19823968e-01 -8.12123835e-01 1.74131975e-01 -2.95334280e-01
-6.08832955e-01 1.07724917e+00 2.51310825e-01 -2.53232062e-01
4.99578238e-01 3.50571603e-01 -1.01108076e-02 -7.47579560e-02
-9.36796784e-01 5.89532554e-02 -2.93754101e-01 4.91917133e-01
3.42598647e-01 2.75167581e-02 -5.10197997e-01 2.88353682e-01
-6.00680351e-01 7.63833895e-02 8.00377369e-01 1.06820774e+00
-3.36370915e-01 -1.42371750e+00 -3.86670232e-01 3.75844181e-01
-7.84716234e-02 -1.03268176e-01 -6.29592001e-01 1.19426894e+00
-8.73998459e-03 8.51651847e-01 1.43962428e-01 -7.91415811e-01
8.28296468e-02 2.15728164e-01 3.29608887e-01 -3.98217261e-01
-6.74754009e-02 -1.10339649e-01 1.77062482e-01 -4.71296012e-01
-1.99042737e-01 -9.56005156e-01 -1.26800668e+00 -9.15354490e-01
-3.49034637e-01 3.28929961e-01 5.99445999e-01 8.26755583e-01
6.75280213e-01 -2.60714022e-03 6.18561447e-01 -7.86110878e-01
-7.41429567e-01 -7.02514112e-01 -6.66251481e-01 4.07843947e-01
2.37515569e-01 -8.86109054e-01 -8.28004539e-01 -2.75216848e-01] | [8.41382122039795, 4.050363063812256] |
62099829-7dbf-4093-96bb-5ddfd19a09ff | mixed-td-efficient-neural-network-accelerator | 2306.05021 | null | https://arxiv.org/abs/2306.05021v2 | https://arxiv.org/pdf/2306.05021v2.pdf | Mixed-TD: Efficient Neural Network Accelerator with Layer-Specific Tensor Decomposition | Neural Network designs are quite diverse, from VGG-style to ResNet-style, and from Convolutional Neural Networks to Transformers. Towards the design of efficient accelerators, many works have adopted a dataflow-based, inter-layer pipelined architecture, with a customised hardware towards each layer, achieving ultra high throughput and low latency. The deployment of neural networks to such dataflow architecture accelerators is usually hindered by the available on-chip memory as it is desirable to preload the weights of neural networks on-chip to maximise the system performance. To address this, networks are usually compressed before the deployment through methods such as pruning, quantization and tensor decomposition. In this paper, a framework for mapping CNNs onto FPGAs based on a novel tensor decomposition method called Mixed-TD is proposed. The proposed method applies layer-specific Singular Value Decomposition (SVD) and Canonical Polyadic Decomposition (CPD) in a mixed manner, achieving 1.73x to 10.29x throughput per DSP to state-of-the-art CNNs. Our work is open-sourced: https://github.com/Yu-Zhewen/Mixed-TD | ['Christos-Savvas Bouganis', 'Zhewen Yu'] | 2023-06-08 | null | null | null | null | ['quantization'] | ['methodology'] | [-1.95761383e-01 -8.49208608e-03 -2.96071153e-02 -3.50919485e-01
3.97714257e-01 -4.00988489e-01 3.86793017e-01 3.04394867e-02
-5.55974007e-01 3.07349443e-01 3.88370641e-02 -6.86266184e-01
-1.97709948e-01 -9.36685145e-01 -4.96802688e-01 -6.03763044e-01
-2.59093456e-02 -1.46937855e-02 3.37601066e-01 -2.56324023e-01
1.62480071e-01 8.42031598e-01 -1.73101461e+00 4.02449042e-01
4.70626622e-01 1.41599715e+00 2.74538338e-01 6.70705855e-01
-3.40526462e-01 7.24143267e-01 -4.59230632e-01 -6.62259996e-01
5.46456337e-01 5.90652563e-02 -3.54635268e-01 -2.17770055e-01
3.55965197e-01 -4.21432048e-01 -4.29716915e-01 1.16870773e+00
3.46828520e-01 -2.30962694e-01 2.76116222e-01 -1.14822173e+00
-1.49519488e-01 7.29531288e-01 -3.26080948e-01 2.66213477e-01
-3.94504488e-01 1.55504137e-01 7.28174984e-01 -8.49181354e-01
3.54231596e-01 1.19653380e+00 5.60835183e-01 3.15067112e-01
-9.70769405e-01 -8.03599656e-01 -6.16472512e-02 2.54900068e-01
-1.04205024e+00 -4.85263050e-01 6.88345373e-01 -5.19916952e-01
1.17334962e+00 1.16927668e-01 9.17114437e-01 8.19247901e-01
4.90453422e-01 3.21170330e-01 8.58927011e-01 -1.67246357e-01
5.34606993e-01 -2.48293951e-02 1.82521671e-01 7.54201531e-01
5.48969090e-01 -9.14380699e-02 -6.02211595e-01 2.13837579e-01
9.21616793e-01 9.56077278e-02 1.04129173e-01 1.51321128e-01
-1.15857768e+00 6.02415860e-01 7.50128746e-01 3.82034063e-01
-5.66675186e-01 3.89711738e-01 9.81686652e-01 2.56137639e-01
2.28501603e-01 1.24047987e-01 -4.50307369e-01 -1.18303306e-01
-1.16477275e+00 2.86460429e-01 8.43804896e-01 1.11180127e+00
5.41675508e-01 4.20082450e-01 1.11990748e-03 4.47611660e-01
5.45688629e-01 2.89465636e-02 7.31973350e-01 -7.32495308e-01
6.07159317e-01 7.69722342e-01 -5.47378600e-01 -8.47480595e-01
-6.31272972e-01 -6.92203164e-01 -1.37025368e+00 3.92774075e-01
1.83672056e-01 -1.67430222e-01 -8.27827990e-01 1.20660245e+00
4.03695881e-01 8.99965987e-02 1.23373531e-01 9.40515161e-01
7.83491075e-01 6.85877085e-01 5.11513539e-02 3.06289911e-01
1.71971798e+00 -1.05977046e+00 -5.27425468e-01 -1.25227183e-01
7.16927469e-01 -9.00780499e-01 7.79987574e-01 5.95809877e-01
-1.05753326e+00 -7.81757891e-01 -1.49898207e+00 -3.16420525e-01
-4.89131272e-01 7.68400133e-01 5.87964535e-01 8.31414104e-01
-1.19336545e+00 7.91038096e-01 -1.13576901e+00 -1.43577829e-01
5.24643779e-01 5.54630756e-01 -2.45826051e-01 1.87018797e-01
-9.17768776e-01 7.05314517e-01 6.84203029e-01 5.31237364e-01
-5.65440178e-01 -8.75540435e-01 -3.88674021e-01 3.14907283e-01
-2.25428060e-01 -7.56787300e-01 1.03273487e+00 -8.46407831e-01
-1.75035560e+00 3.05457920e-01 1.53442010e-01 -9.52927887e-01
3.10028464e-01 3.06962319e-02 -2.82735378e-01 2.39480045e-02
-4.34425294e-01 7.19144523e-01 1.02129805e+00 -3.03480029e-01
-6.21300697e-01 -2.93991834e-01 2.10298434e-01 -2.25086108e-01
-8.21411550e-01 2.06664391e-02 4.33836989e-02 -5.67878067e-01
1.71767309e-01 -8.04500580e-01 -3.11788470e-01 3.08142066e-01
-3.18385601e-01 -1.32370621e-01 8.90515208e-01 -5.03103554e-01
1.10103011e+00 -2.17216873e+00 2.76880506e-02 5.06875068e-02
4.23781782e-01 7.96538234e-01 2.65439749e-01 1.97117880e-01
-5.69041073e-02 -3.98882985e-01 2.74700552e-01 -4.56774384e-01
-1.07083336e-01 3.60670567e-01 -4.47587043e-01 4.67934191e-01
3.81326914e-01 4.10586894e-01 -6.81813955e-01 -2.92238712e-01
3.46397728e-01 7.57396698e-01 -6.51193023e-01 -3.15218195e-02
-3.70668545e-02 3.86463515e-02 -3.28195274e-01 5.88836372e-01
1.00886548e+00 -2.18578316e-02 1.91822335e-01 -8.17707479e-01
-6.85429215e-01 3.59591305e-01 -1.11308014e+00 1.56694925e+00
-5.11805773e-01 7.10030735e-01 3.41069400e-01 -9.98303413e-01
1.28830421e+00 3.77678514e-01 2.07420379e-01 -4.10886288e-01
5.32341599e-01 5.45756221e-01 2.61029243e-01 -2.15100557e-01
7.78909981e-01 2.04993442e-01 3.31891775e-01 -1.15396291e-01
3.72102857e-01 3.06907088e-01 4.91114438e-01 -4.80775423e-02
1.09179783e+00 2.14467287e-01 -1.85083840e-02 -6.30923927e-01
7.07704067e-01 6.26275837e-02 5.18057525e-01 1.32452950e-01
7.76710510e-02 4.33425233e-02 7.21096814e-01 -6.27860844e-01
-1.39569294e+00 -7.47190475e-01 -3.26020479e-01 7.80106187e-01
-4.86394346e-01 -6.97449505e-01 -8.77785206e-01 2.76827198e-02
-2.75157541e-01 2.22227544e-01 -5.21143042e-02 3.08060553e-04
-5.73193550e-01 -6.23385489e-01 9.28141236e-01 5.04765868e-01
9.01750267e-01 -5.89086890e-01 -1.13806236e+00 5.66066861e-01
7.07972348e-01 -1.21460843e+00 1.59481108e-01 5.43083847e-01
-1.28465986e+00 -5.28610289e-01 -5.64907014e-01 -6.10340655e-01
7.28617072e-01 1.05723720e-02 7.33218014e-01 -2.05847621e-01
-2.95729488e-01 -3.61756414e-01 -3.54919344e-01 -3.90150845e-01
-1.82853162e-01 3.93846929e-01 1.91678062e-01 1.26738459e-01
2.33206153e-01 -8.10047805e-01 -8.24119449e-01 5.01205623e-02
-8.64715338e-01 4.44182724e-01 9.04631317e-01 5.93293190e-01
4.91653234e-01 -4.61187512e-02 2.55798362e-02 -5.37577510e-01
5.01623511e-01 -4.01774168e-01 -1.13884234e+00 -1.81500480e-01
-5.26208699e-01 2.06922725e-01 1.28955007e+00 -3.54454011e-01
-6.53485000e-01 2.93256819e-01 -4.61845100e-01 -7.61303961e-01
-6.59115659e-03 4.08059210e-01 2.33667940e-02 -3.57355118e-01
6.51432931e-01 -5.61627001e-02 1.01009021e-02 -5.65582514e-01
1.71029627e-01 6.03511572e-01 4.22763735e-01 -3.19272488e-01
5.17563462e-01 4.96945262e-01 4.90460306e-01 -9.53159332e-01
-3.61714900e-01 -1.51011080e-01 -5.55872142e-01 -3.99867326e-01
8.16885114e-01 -1.05646634e+00 -7.54375160e-01 5.03863633e-01
-1.40629327e+00 -2.82025874e-01 -2.30750404e-02 7.41734266e-01
-4.32362035e-02 -5.64426892e-02 -6.71608090e-01 -4.73612279e-01
-7.13631809e-01 -1.45288086e+00 6.71162844e-01 3.99845809e-01
1.10211521e-01 -6.81231081e-01 -4.24131989e-01 -1.05374746e-01
8.00027847e-01 1.36535808e-01 6.95717633e-01 -2.34771639e-01
-7.46832728e-01 2.87056062e-03 -5.88975430e-01 6.67706907e-01
-3.14590514e-01 2.61529118e-01 -8.96859467e-01 -3.49289149e-01
1.72802657e-01 1.13910094e-01 6.56737328e-01 2.59106010e-01
1.14456367e+00 -3.01053703e-01 -1.39171422e-01 1.03881550e+00
1.59434688e+00 -6.36397004e-02 5.89682698e-01 3.60223234e-01
7.99700677e-01 3.02116424e-01 2.81909168e-01 6.67280138e-01
2.70630211e-01 5.34405351e-01 7.56983340e-01 4.28361334e-02
-1.92063421e-01 1.05854318e-01 3.28103930e-01 1.24805880e+00
-2.08686456e-01 4.14991304e-02 -8.10775161e-01 3.51470381e-01
-1.54663932e+00 -5.07244468e-01 -5.15648305e-01 1.92782915e+00
5.01803279e-01 4.89413559e-01 -1.15257278e-01 4.85467702e-01
4.62862879e-01 9.88613367e-02 -3.15981001e-01 -9.36525464e-01
2.88650632e-01 3.92162144e-01 1.08891845e+00 4.66672145e-02
-8.93007517e-01 6.89889610e-01 4.82185030e+00 8.30662668e-01
-1.69021034e+00 3.10786843e-01 3.90642971e-01 -1.70707867e-01
2.26229131e-01 1.51543021e-02 -1.23131526e+00 5.22058785e-01
1.45467198e+00 1.31142229e-01 3.47485214e-01 1.18321133e+00
1.60081923e-01 2.40012065e-01 -7.48907924e-01 1.04815578e+00
-5.29706061e-01 -1.46377611e+00 -9.45108682e-02 2.11293772e-01
3.91470134e-01 3.46117675e-01 -1.60059202e-02 2.17799824e-02
-6.96336180e-02 -6.40752733e-01 9.96103883e-01 3.53150427e-01
6.85752273e-01 -8.35805535e-01 6.48918211e-01 1.03843465e-01
-1.46202624e+00 -2.48278454e-01 -9.30578053e-01 -3.33859086e-01
9.12307575e-03 1.14367628e+00 -7.56732464e-01 4.02405024e-01
7.88437068e-01 6.63259566e-01 -5.18176615e-01 9.73616958e-01
-5.30025400e-02 6.37462914e-01 -3.81202072e-01 -2.41847456e-01
5.61807692e-01 -2.47498393e-01 3.39231670e-01 1.26646972e+00
5.52293837e-01 -2.64585882e-01 -3.86465043e-01 7.40765333e-01
8.48724917e-02 -7.58371428e-02 -3.32043260e-01 -7.58132413e-02
4.02482212e-01 1.79181695e+00 -9.29799438e-01 -3.65552545e-01
-4.45854336e-01 5.87170005e-01 1.28926173e-01 -1.06347077e-01
-8.15004826e-01 -8.31543326e-01 9.56827760e-01 4.03854363e-02
4.83511955e-01 -6.06257081e-01 -5.14913797e-01 -9.70523655e-01
1.12383485e-01 -5.08658707e-01 -3.80407572e-02 -4.71319407e-01
-8.15209985e-01 1.01435709e+00 -2.09316403e-01 -1.39276505e+00
-4.93945293e-02 -1.20312035e+00 -4.38396782e-01 8.96515548e-01
-1.42896676e+00 -9.83869374e-01 -4.50706959e-01 4.72673476e-01
3.66599411e-01 -5.06296992e-01 6.51378810e-01 8.27928424e-01
-6.56625569e-01 5.34278750e-01 7.47529566e-02 8.14106986e-02
1.90930158e-01 -9.10698354e-01 5.81246912e-01 1.03158200e+00
-2.54511178e-01 6.32257640e-01 6.18883193e-01 -2.38934904e-01
-1.99330842e+00 -1.34425581e+00 6.64529860e-01 4.54909712e-01
8.00240934e-01 -6.13122702e-01 -7.18343496e-01 4.03855026e-01
2.99539983e-01 3.06410283e-01 4.22036141e-01 -3.19444329e-01
-2.86586791e-01 -7.01912880e-01 -1.19347203e+00 6.59173250e-01
7.84799457e-01 -1.74093679e-01 -2.36917175e-02 2.55276382e-01
6.26749516e-01 -6.21392190e-01 -1.07608986e+00 6.01730403e-03
4.88721400e-01 -1.15725493e+00 7.72101939e-01 -9.47248563e-02
6.61075294e-01 -5.24886787e-01 -1.25920475e-01 -1.00904822e+00
-2.56618559e-01 -4.98323441e-01 -3.44969004e-01 1.08672965e+00
8.29607844e-02 -8.11053395e-01 8.31020474e-01 2.80376583e-01
-5.71562231e-01 -8.80491734e-01 -1.11048746e+00 -6.63475335e-01
-2.64012128e-01 -3.98315758e-01 9.13709939e-01 4.67235833e-01
-1.71687737e-01 1.35864198e-01 1.63648389e-02 2.70648599e-01
5.62345684e-01 -4.31757003e-01 5.54999888e-01 -1.08408332e+00
-3.24794114e-01 -6.92891061e-01 -1.02507007e+00 -9.51022446e-01
-2.11775988e-01 -9.75672960e-01 -4.22775835e-01 -1.29145420e+00
-7.33201981e-01 -5.29034138e-01 -5.65511994e-02 3.39608043e-01
7.26826370e-01 2.74671495e-01 2.58334994e-01 -1.19761461e-02
-2.33636182e-02 2.30663598e-01 1.09531391e+00 7.77490661e-02
6.15392663e-02 -1.60989746e-01 -2.98182964e-01 5.48657298e-01
9.69465196e-01 -4.03398961e-01 -5.00805974e-01 -8.40177953e-01
3.01450104e-01 -9.25417915e-02 3.47022831e-01 -1.65511739e+00
6.77645206e-01 3.50144178e-01 3.37718666e-01 -7.18334675e-01
3.84284556e-01 -8.82739544e-01 3.11081350e-01 8.00298035e-01
1.69736687e-02 4.37704772e-01 3.17752779e-01 4.77792416e-03
-2.97501296e-01 -4.59535718e-01 8.05965781e-01 -4.46670167e-02
-6.79282069e-01 2.91494727e-01 -4.86074805e-01 -4.99172926e-01
7.48722017e-01 -1.04459688e-01 -4.02516484e-01 5.23175657e-01
-3.80154938e-01 -1.13754921e-01 -9.99928042e-02 1.22435480e-01
5.28715611e-01 -1.26164949e+00 -4.43680197e-01 4.26762879e-01
-3.30784321e-01 1.98453054e-01 3.77592683e-01 8.63313675e-01
-1.13670099e+00 7.75956035e-01 -8.51050735e-01 -6.05493903e-01
-1.11060512e+00 3.23146015e-01 1.62354589e-01 -1.98467493e-01
-6.59922183e-01 8.31502378e-01 -3.69697571e-01 -1.09334096e-01
2.11073890e-01 -8.99551630e-01 -1.52418599e-01 1.19327061e-01
6.92155719e-01 5.41114092e-01 6.25475764e-01 -4.12121028e-01
-2.43844792e-01 3.54175538e-01 -3.80180590e-02 7.55791068e-02
1.27903605e+00 3.62346977e-01 -4.94805187e-01 8.05991050e-03
1.44933510e+00 -6.00813925e-01 -1.16818273e+00 -3.10957292e-03
-1.10836834e-01 -3.09403658e-01 6.38779283e-01 -5.83713427e-02
-1.54966104e+00 1.10922194e+00 6.83667302e-01 1.40438586e-01
1.34729648e+00 -5.59381247e-01 7.17567027e-01 4.35157835e-01
4.18643862e-01 -9.10401404e-01 -2.65608162e-01 7.29956150e-01
5.47338188e-01 -4.74125654e-01 1.04536720e-01 -2.22767249e-01
-1.52235836e-01 1.70273411e+00 5.71155608e-01 -5.43326557e-01
8.72297108e-01 7.75101900e-01 -3.67958248e-01 5.52416872e-03
-9.52976823e-01 2.02268958e-02 -9.81224179e-02 2.28369504e-01
4.48977023e-01 1.52485728e-01 -4.68541592e-01 4.60313082e-01
-6.59239471e-01 3.03823352e-01 5.16238153e-01 8.83789122e-01
-2.09275082e-01 -1.14460289e+00 -2.96692640e-01 6.16620660e-01
-4.49640363e-01 -1.46069407e-01 3.70748371e-01 2.97603577e-01
4.71555352e-01 4.79767203e-01 3.94752383e-01 -7.31133044e-01
2.76457191e-01 -2.31864452e-01 3.96578014e-01 -2.72109538e-01
-1.01167178e+00 -9.28258300e-02 -3.75390835e-02 -5.76003015e-01
-1.75837517e-01 -4.29666132e-01 -1.04877186e+00 -6.31818533e-01
7.00137988e-02 -2.35763058e-01 1.56145501e+00 5.64133525e-01
5.14991462e-01 9.87700045e-01 4.16547328e-01 -1.16800857e+00
-4.98449504e-01 -8.06333065e-01 -4.56332594e-01 -4.31281567e-01
2.50507474e-01 -5.22324383e-01 -8.38455483e-02 -3.68372574e-02] | [8.415566444396973, 2.862128734588623] |
e5ea284e-f76a-4ec0-984f-452a8f76346b | the-role-of-emotions-in-native-language | null | null | https://aclanthology.org/W18-6218 | https://aclanthology.org/W18-6218.pdf | The Role of Emotions in Native Language Identification | We explore the hypothesis that emotion is one of the dimensions of language that surfaces from the native language into a second language. To check the role of emotions in native language identification (NLI), we model emotion information through polarity and emotion load features, and use document representations using these features to classify the native language of the author. The results indicate that emotion is relevant for NLI, even for high proficiency levels and across topics. | ['Carlo Strapparava', 'Vivi Nastase', 'Ilia Markov', 'Grigori Sidorov'] | 2018-10-01 | null | null | null | ws-2018-10 | ['deception-detection', 'native-language-identification'] | ['miscellaneous', 'natural-language-processing'] | [-3.25900108e-01 -9.86254308e-03 -8.65749896e-01 -2.50236779e-01
-1.56980366e-01 -8.62566769e-01 8.01080883e-01 3.26865226e-01
-4.30025369e-01 2.67082810e-01 6.66958451e-01 -5.01520216e-01
8.36246759e-02 -4.84382629e-01 -7.26320297e-02 1.71573323e-04
7.84193203e-02 -2.66782232e-02 -9.29024518e-01 -2.38761440e-01
6.56620085e-01 6.35980308e-01 -1.38210535e+00 1.38957828e-01
1.03885531e+00 7.72443593e-01 -2.32886419e-01 3.95374894e-01
-8.52948606e-01 9.13336217e-01 -9.82884824e-01 -4.99839753e-01
-9.43484828e-02 -5.08823812e-01 -9.44371164e-01 -2.38639131e-01
3.84854913e-01 2.59947985e-01 -1.26777049e-02 1.17678308e+00
1.55386761e-01 -6.18044334e-03 9.96987283e-01 -1.02601922e+00
-1.08699763e+00 7.97968090e-01 -2.25450471e-01 2.26287752e-01
1.24091637e+00 -2.40444273e-01 8.92861605e-01 -1.12567902e+00
1.05658078e+00 1.69936442e+00 5.44549525e-01 4.37477887e-01
-1.17428732e+00 -9.80435550e-01 3.88688207e-01 -2.61151284e-01
-1.33444786e+00 -4.62414771e-01 9.58274424e-01 -1.04326618e+00
9.33047056e-01 3.08737326e-02 8.36890459e-01 1.30184364e+00
5.51006913e-01 3.98682714e-01 2.01523352e+00 -7.15200663e-01
-1.85561121e-01 1.02278817e+00 8.46007884e-01 5.96557677e-01
2.47561172e-01 6.04559556e-02 -1.10735500e+00 -2.62742117e-02
1.46902412e-01 -5.86803198e-01 -2.25037396e-01 5.60645163e-01
-1.38876104e+00 1.01330304e+00 -6.34863377e-02 7.98333764e-01
-2.62552738e-01 -6.33782327e-01 4.91038412e-01 9.96157229e-01
6.16790593e-01 1.16431892e+00 -4.26664650e-01 -7.61946082e-01
-7.84454703e-01 -2.42484346e-01 1.03538120e+00 3.98356050e-01
6.29005432e-01 9.29483473e-02 -9.18482766e-02 1.04428816e+00
2.75933385e-01 4.63916957e-01 1.23541057e+00 -4.47337419e-01
8.48934054e-03 9.71368074e-01 -3.28649998e-01 -1.17723858e+00
-3.58245522e-01 -2.53209293e-01 -4.09830093e-01 1.53783232e-01
1.62875757e-01 -2.76093692e-01 -3.73797417e-01 1.87075353e+00
-3.00511628e-01 -7.09925115e-01 5.00526667e-01 7.36761093e-01
1.04828238e+00 8.40031624e-01 3.09434414e-01 -5.77233672e-01
1.40581989e+00 -5.76716304e-01 -8.66952658e-01 -4.62503821e-01
1.02862263e+00 -8.75777185e-01 1.25596237e+00 5.26514411e-01
-9.99217570e-01 -9.36646521e-01 -8.93756151e-01 -9.14931744e-02
-1.01096201e+00 4.44041342e-01 1.09229076e+00 1.02045441e+00
-1.17523086e+00 2.95349360e-01 7.46933147e-02 -5.55419862e-01
-2.66473413e-01 1.68310106e-01 -7.14187026e-01 1.90933168e-01
-1.58997381e+00 9.49902952e-01 1.94645017e-01 -3.10417593e-01
-1.15203941e-02 -4.97716039e-01 -1.05430615e+00 -2.72398949e-01
-5.95107913e-01 1.63541794e-01 4.92482871e-01 -1.81139874e+00
-1.80731785e+00 1.56309509e+00 -4.63276893e-01 4.21854615e-01
-1.91484585e-01 -1.29579142e-01 -8.53956044e-01 -6.69308305e-02
4.06689763e-01 5.23640394e-01 6.41201079e-01 -9.22765672e-01
-4.31368738e-01 -4.91176128e-01 -1.00951672e-01 3.41106534e-01
-8.27698469e-01 5.89099348e-01 -8.60886183e-03 -6.56607270e-01
-3.71615449e-03 -7.04384446e-01 3.92728031e-01 -5.60883105e-01
-1.13456316e-01 -9.39797461e-01 2.64148325e-01 -1.13205850e+00
1.35004842e+00 -2.56594515e+00 2.25861460e-01 6.78864419e-01
2.60754842e-02 -4.03036118e-01 -1.47702545e-01 2.65264004e-01
-2.44474113e-01 8.61932993e-01 7.17910230e-01 5.87732233e-02
1.15920350e-01 -1.69675499e-01 -2.51263268e-02 3.12388867e-01
1.59303561e-01 8.80232394e-01 -5.26820421e-01 -3.16651523e-01
-5.25243998e-01 3.81097764e-01 -1.97160661e-01 2.23779455e-01
4.41449195e-01 5.17457545e-01 1.78978257e-02 6.70394421e-01
3.02634597e-01 2.37938970e-01 1.09567992e-01 2.74927914e-01
-6.60779536e-01 7.73886323e-01 -7.59272337e-01 1.22863853e+00
-6.39535367e-01 1.09709191e+00 2.61867851e-01 -6.99781477e-01
1.32222509e+00 2.32552707e-01 7.61911795e-02 -9.44370270e-01
1.92459285e-01 2.74702251e-01 4.35092747e-01 -2.00942516e-01
5.39737940e-01 -4.88708876e-02 -6.91760123e-01 4.93797094e-01
5.70921339e-02 -4.21814770e-02 2.46057600e-01 1.75760016e-01
3.57793689e-01 -4.12235856e-01 4.43495750e-01 -8.54197443e-01
6.17544889e-01 -1.14939660e-01 4.98162627e-01 5.94120920e-01
-1.35854423e-01 -5.51565170e-01 7.72870421e-01 8.56306106e-02
-2.91265368e-01 -6.72292352e-01 -1.98526204e-01 1.55642426e+00
-1.24127105e-01 -2.31104910e-01 -4.25857455e-01 -3.96281898e-01
-1.60438064e-02 8.75375688e-01 -5.49775839e-01 -3.56213570e-01
1.28461093e-01 4.23327368e-03 2.20525026e-01 2.40364984e-01
-2.15916056e-02 -9.76450920e-01 -4.12164479e-02 -3.52887332e-01
-3.99443321e-02 -9.84823048e-01 -2.87694961e-01 2.02104717e-01
-3.49780858e-01 -3.54443014e-01 -4.28100944e-01 -1.12061429e+00
5.81341803e-01 -1.61668673e-01 1.20662081e+00 2.70035833e-01
-8.33034664e-02 9.58368897e-01 -4.17381376e-01 -4.82005209e-01
-7.08045900e-01 2.61073947e-01 3.83863956e-01 -4.04328167e-01
7.11003840e-01 -6.38238946e-03 1.58497214e-01 -3.01853925e-01
-3.96397322e-01 -2.64961898e-01 2.44820639e-01 4.79863942e-01
-2.66723111e-02 2.56201595e-01 2.88021356e-01 -5.15435994e-01
1.30068874e+00 -5.26616991e-01 -1.93147242e-01 2.32971907e-01
-8.72537076e-01 -2.42586002e-01 4.60453123e-01 -8.42337549e-01
-1.00124753e+00 -2.74547100e-01 -7.31835365e-02 2.18238264e-01
-4.01631653e-01 1.11871541e+00 -1.00660868e-01 -4.32492614e-01
2.05607891e-01 5.06870099e-04 -3.72622237e-02 -1.72587633e-01
4.49934565e-02 9.14665401e-01 1.88532442e-01 -7.25405276e-01
5.30851007e-01 -2.82799244e-01 -4.96404052e-01 -1.22730422e+00
-7.34366238e-01 -3.80718142e-01 -7.43709505e-01 -4.57808316e-01
8.15457702e-01 -1.24393368e+00 -9.00068998e-01 3.00339252e-01
-1.08624125e+00 -2.32583568e-01 1.16400957e-01 8.99322271e-01
9.71751958e-02 2.32992396e-02 -7.15780795e-01 -9.64006066e-01
-3.13303292e-01 -1.06912601e+00 5.77403486e-01 3.36324692e-01
-9.94709015e-01 -1.27059436e+00 1.38533682e-01 8.98991153e-02
1.70119286e-01 -2.45096818e-01 1.30468357e+00 -1.00990689e+00
4.17666167e-01 -1.71221867e-01 -1.46839162e-02 2.43545815e-01
-6.86415955e-02 1.51263133e-01 -7.10795939e-01 -1.34705931e-01
-9.92815476e-03 -7.44450033e-01 5.51741362e-01 3.18223648e-02
7.44868696e-01 -4.42371249e-01 -1.31677091e-01 4.52239215e-01
1.32335770e+00 2.56273925e-01 7.90194944e-02 2.41578728e-01
5.92540443e-01 9.34825718e-01 4.59017187e-01 1.41849607e-01
5.06634176e-01 3.28969091e-01 -7.32493997e-01 -1.12300701e-02
3.12200874e-01 -2.75478542e-01 1.12721181e+00 1.29938495e+00
3.28486532e-01 1.27320006e-01 -1.17334950e+00 4.11771685e-01
-7.93053865e-01 -6.26316965e-01 6.13200106e-02 1.94102275e+00
9.66458857e-01 -4.70976233e-02 -7.17183668e-03 1.18676685e-01
1.83845192e-01 -1.30637665e-03 -2.72104025e-01 -1.42862272e+00
-4.00597751e-01 2.36862615e-01 3.05374652e-01 8.58762026e-01
-6.64223731e-01 1.36980402e+00 7.72469139e+00 3.71355623e-01
-1.72299683e+00 -2.68908948e-01 8.96710455e-01 3.84759903e-01
-5.59291840e-01 -2.31851399e-01 -9.33418632e-01 1.73545703e-01
1.13685870e+00 -5.80056608e-01 5.83331823e-01 9.32055891e-01
5.27382344e-02 -2.30473906e-01 -1.17075622e+00 9.13230717e-01
4.37031001e-01 -4.37232852e-01 1.21357270e-01 1.63769469e-01
4.37804222e-01 -4.65224743e-01 2.89840400e-01 7.33513236e-01
6.79640695e-02 -1.21436024e+00 7.41589606e-01 6.04807675e-01
1.03423607e+00 -5.78921080e-01 4.58504975e-01 1.76569164e-01
-7.51022160e-01 -3.29664126e-02 -1.00343369e-01 -9.25684392e-01
-4.41972554e-01 2.26640359e-01 -6.80226684e-01 -1.95347399e-01
5.77333510e-01 5.83153188e-01 -8.95590782e-01 -3.47931944e-02
-1.30802676e-01 8.62943470e-01 4.80828919e-02 -2.50689417e-01
-7.65423253e-02 -2.61238486e-01 3.88132960e-01 1.45786166e+00
4.57221299e-01 -8.12136233e-02 4.08833146e-01 6.11861050e-01
-8.12771991e-02 9.48643267e-01 -1.02153361e+00 -1.01672161e+00
5.93019545e-01 1.18366551e+00 -8.03642452e-01 -4.45106506e-01
-6.22619271e-01 1.41428673e+00 4.40747857e-01 4.24824595e-01
1.03033789e-01 -4.36548918e-01 1.00493610e+00 -3.38793397e-01
-6.40682638e-01 -3.33590716e-01 -6.52391255e-01 -1.12224162e+00
-2.80775577e-01 -1.04128063e+00 8.60231817e-02 -4.72253531e-01
-1.44423914e+00 3.97505939e-01 -5.33331871e-01 -4.75942969e-01
-3.82285386e-01 -1.23567438e+00 -4.83713031e-01 1.26399350e+00
-1.15706611e+00 -8.52978885e-01 1.44207589e-02 4.89756018e-01
2.14201704e-01 -5.88390112e-01 1.21275818e+00 -1.52538389e-01
-5.52661121e-01 8.85546088e-01 -1.55184731e-01 4.00003076e-01
8.88024271e-01 -1.22246242e+00 -1.10312961e-01 6.16585195e-01
1.45728186e-01 1.13440943e+00 2.75361896e-01 -7.53884912e-01
-1.29059649e+00 -3.02074432e-01 1.69317138e+00 -2.34828219e-01
9.07240629e-01 -5.33022523e-01 -6.90627873e-01 3.55954349e-01
6.98563397e-01 -8.52277040e-01 1.26630557e+00 6.29454851e-01
-4.70056981e-01 4.05095458e-01 -7.41054893e-01 6.42823219e-01
8.92424166e-01 -1.11832154e+00 -5.05608499e-01 1.78517058e-01
8.32279027e-01 1.53311491e-01 -1.53371501e+00 8.00712109e-02
6.53821230e-01 -7.25602865e-01 7.12210953e-01 -4.65995520e-01
5.03015757e-01 5.45707464e-01 1.67045534e-01 -1.64711964e+00
-5.94456077e-01 -4.73710716e-01 5.41624665e-01 1.58149517e+00
6.81582570e-01 -9.55726862e-01 7.29552731e-02 8.36951673e-01
3.05202544e-01 -4.59422082e-01 -7.43834734e-01 -4.87616152e-01
6.80419683e-01 -3.55645925e-01 4.41913992e-01 1.60665476e+00
7.10920870e-01 6.49826407e-01 2.34746397e-01 -1.36996165e-01
-2.57104039e-01 3.15443501e-02 4.17683631e-01 -1.83181763e+00
2.14818195e-01 -1.14131939e+00 -3.82380575e-01 -4.46418494e-01
1.39402211e+00 -1.20909929e+00 -3.65634412e-01 -1.34687817e+00
-9.28183123e-02 -5.86181134e-02 -1.24616258e-01 2.97769248e-01
-1.86391532e-01 -3.96233425e-02 2.52538323e-01 1.21035315e-01
-1.09305859e-01 1.16203830e-01 7.11262882e-01 -1.59850657e-01
-4.93539900e-01 -4.72393900e-01 -1.16301107e+00 8.40784609e-01
9.79365766e-01 4.26688641e-02 -2.71667808e-01 -7.58347195e-03
6.29442155e-01 2.57524531e-02 -2.06455141e-01 -7.18340695e-01
-1.09694235e-01 -2.64005929e-01 8.09885621e-01 2.07547415e-02
1.62709221e-01 -6.92931414e-01 -4.88866031e-01 6.71523139e-02
-7.49615431e-01 6.76030099e-01 4.14094269e-01 -3.98150921e-01
-5.91968954e-01 -4.01902974e-01 4.67866331e-01 7.11949868e-03
-7.23581493e-01 -2.71761239e-01 -1.08390725e+00 1.95555836e-01
6.65383041e-01 -3.99004489e-01 8.58113915e-02 -4.12361890e-01
-4.90685940e-01 -2.30930105e-01 7.55677164e-01 1.08132350e+00
3.48242879e-01 -1.25194371e+00 -3.24204952e-01 6.14364684e-01
1.79625720e-01 -1.21129537e+00 -3.29553097e-01 6.32573128e-01
-4.09422964e-02 6.41385734e-01 -3.24018508e-01 8.03894326e-02
-1.36987793e+00 3.00469279e-01 4.98580597e-02 -1.00014798e-01
5.53737395e-02 8.78551662e-01 8.56634676e-02 -6.35596037e-01
3.07677150e-01 8.45779702e-02 -8.07509243e-01 6.61619127e-01
3.94757867e-01 6.15577064e-02 -4.93594736e-01 -1.36893332e+00
-4.05082613e-01 6.25078917e-01 5.13171516e-02 -5.81313074e-01
6.23584330e-01 -2.28244394e-01 -7.88646638e-01 1.24094474e+00
1.37805676e+00 9.43461657e-01 2.50204295e-01 -1.85239743e-02
2.31271684e-01 -1.73548266e-01 9.08024311e-02 -1.04761207e+00
-4.72960442e-01 6.68740332e-01 7.75674820e-01 -3.29422355e-02
7.12262750e-01 -1.62207320e-01 5.70086390e-02 2.24994361e-01
2.93490756e-02 -1.49922025e+00 -2.60332793e-01 1.11087406e+00
8.12333107e-01 -1.21250951e+00 -3.26870024e-01 -2.30361596e-01
-8.89532447e-01 1.12585974e+00 9.03096437e-01 3.12658489e-01
9.87813950e-01 1.44643486e-01 7.21454322e-01 -1.39269531e-01
-5.55144250e-01 1.66508276e-02 5.88609278e-01 4.51520920e-01
1.53028631e+00 4.24654573e-01 -1.00944757e+00 8.15765738e-01
-9.79163408e-01 -3.11487883e-01 2.75238156e-01 4.11198020e-01
-3.26182358e-02 -1.11280942e+00 -6.12578630e-01 5.28012037e-01
-6.76009178e-01 -3.76558661e-01 -1.25101709e+00 5.85671186e-01
4.87697870e-02 9.88432109e-01 3.37591529e-01 -5.69055676e-01
-5.50864302e-02 8.29506457e-01 -7.92516861e-03 -4.25313681e-01
-9.88628924e-01 -2.16546431e-01 2.26716816e-01 -3.44534189e-01
-4.14240003e-01 -6.66202903e-01 -1.17392588e+00 -1.02197476e-01
5.31781353e-02 6.49513185e-01 7.97594607e-01 9.34362829e-01
2.81964749e-01 1.17188402e-01 7.01987863e-01 -4.85136122e-01
7.64597952e-02 -1.26864326e+00 -6.75757766e-01 4.75710094e-01
-1.81314293e-02 -4.94340301e-01 -7.14158535e-01 -2.36822471e-01] | [10.530211448669434, 10.261173248291016] |
ed780f96-9f40-4047-b97b-57590b5279bc | learning-to-pronounce-as-measuring-cross | 2202.00794 | null | https://arxiv.org/abs/2202.00794v2 | https://arxiv.org/pdf/2202.00794v2.pdf | Learning to pronounce as measuring cross-lingual joint orthography-phonology complexity | Machine learning models allow us to compare languages by showing how hard a task in each language might be to learn and perform well on. Following this line of investigation, we explore what makes a language "hard to pronounce" by modelling the task of grapheme-to-phoneme (g2p) transliteration. By training a character-level transformer model on this task across 22 languages and measuring the model's proficiency against its grapheme and phoneme inventories, we show that certain characteristics emerge that separate easier and harder languages with respect to learning to pronounce. Namely the complexity of a language's pronunciation from its orthography is due to the expressive or simplicity of its grapheme-to-phoneme mapping. Further discussion illustrates how future studies should consider relative data sparsity per language to design fairer cross-lingual comparison tasks. | ['Domenic Rosati'] | 2022-01-29 | null | null | null | null | ['transliteration'] | ['natural-language-processing'] | [ 3.81399877e-02 7.70185888e-02 -3.65541101e-01 -3.62790316e-01
-6.44189298e-01 -8.81351292e-01 8.55900824e-01 1.29828587e-01
-6.89546943e-01 3.84111226e-01 6.44228041e-01 -9.77056324e-01
4.80536222e-02 -5.94652891e-01 -7.44241774e-01 -1.56518281e-01
2.70312339e-01 6.06887639e-01 -5.19774377e-01 -2.42748782e-01
8.42903331e-02 3.72914404e-01 -1.07091188e+00 2.13580474e-01
1.13267291e+00 1.94017932e-01 5.50279021e-01 3.61235887e-01
-3.88670653e-01 5.87313294e-01 -4.36650366e-01 -8.04300904e-01
2.92186677e-01 -3.49346638e-01 -8.73543143e-01 -1.55275881e-01
9.41434920e-01 1.29066110e-02 -9.22476500e-02 1.29337764e+00
-6.41095042e-02 -2.08121222e-02 1.04819381e+00 -5.56924403e-01
-1.09094536e+00 8.81603599e-01 2.88257226e-02 4.02006656e-01
4.40188766e-01 2.32076645e-01 1.23112798e+00 -8.84116769e-01
5.54227829e-01 1.53500712e+00 6.52622283e-01 4.46506411e-01
-1.38516605e+00 -7.06374109e-01 2.50024766e-01 2.25556660e-02
-1.52830744e+00 -5.97654343e-01 5.65423012e-01 -6.55400991e-01
1.34521592e+00 2.73830503e-01 6.64203644e-01 1.08424139e+00
3.30655575e-01 5.41930735e-01 1.62715614e+00 -6.71304762e-01
-3.13082844e-01 3.80709141e-01 1.55462608e-01 8.17665994e-01
4.47328120e-01 2.42342845e-01 -7.11718202e-01 2.99314260e-01
7.81400859e-01 -5.49758732e-01 -2.94079661e-01 2.03244597e-01
-1.07622838e+00 7.70491838e-01 1.47704169e-01 5.88831365e-01
5.83194979e-02 -1.66203752e-01 2.10014865e-01 6.16625071e-01
7.52236396e-02 8.83564889e-01 -6.26201272e-01 -2.31247127e-01
-7.41958082e-01 -1.01834171e-01 8.56684268e-01 8.76991808e-01
6.23907089e-01 4.45438504e-01 2.13860124e-01 1.00533760e+00
2.07053766e-01 5.53714335e-01 7.73648083e-01 -6.41874850e-01
6.31234467e-01 3.70185345e-01 -2.69250840e-01 -7.22972870e-01
-2.89867669e-01 -4.11804557e-01 -5.44408619e-01 -1.07682317e-01
8.72246802e-01 -5.06798029e-02 -5.63771009e-01 1.85469639e+00
-4.52382445e-01 -4.76154923e-01 4.89350930e-02 5.89371741e-01
4.14173514e-01 8.18449736e-01 5.29485941e-01 -1.21040039e-01
1.57534528e+00 -6.98903799e-01 -3.81241232e-01 -7.89940774e-01
1.06698692e+00 -6.48557425e-01 1.89317667e+00 4.23256010e-01
-1.21604717e+00 -7.76036263e-01 -9.92433190e-01 -4.81198490e-01
-6.10426188e-01 8.88479948e-02 7.94080138e-01 8.15750480e-01
-9.18922961e-01 5.15356302e-01 -7.41104841e-01 -4.72321719e-01
-2.72308826e-01 1.82119235e-01 -5.03139675e-01 1.12252876e-01
-1.26219177e+00 1.51126027e+00 6.10270262e-01 -4.27823514e-01
-6.44986749e-01 -5.89469969e-01 -1.05234802e+00 1.96004778e-01
-2.66599655e-01 -5.53800046e-01 1.14252579e+00 -1.40224981e+00
-1.48707616e+00 1.28492308e+00 -2.23616064e-01 -1.80020347e-01
8.25522244e-02 -1.36108503e-01 -6.46551371e-01 -5.46590269e-01
7.58517683e-02 4.45365936e-01 5.36126137e-01 -7.95344412e-01
-8.37025404e-01 -5.16957581e-01 -8.92384127e-02 5.04674613e-01
-5.16755760e-01 4.85709786e-01 -2.64603496e-01 -7.16062844e-01
1.85755081e-02 -7.88297892e-01 3.03214341e-01 -5.02121508e-01
-2.60739893e-01 -4.99764353e-01 -8.87663811e-02 -1.08401334e+00
1.23223412e+00 -1.98291051e+00 2.69775450e-01 1.90456197e-01
-6.16669059e-02 3.94210443e-02 -3.02153796e-01 4.07480329e-01
-5.79097830e-02 5.04794478e-01 1.20573670e-01 -3.49906236e-01
3.35561216e-01 2.43242323e-01 -1.43262357e-01 4.86494511e-01
2.61748582e-01 1.00849080e+00 -9.16064501e-01 -1.84838474e-01
-8.11839700e-02 4.62279826e-01 -6.21424615e-01 -7.16486946e-02
1.61442220e-01 3.48158479e-01 2.25411236e-01 4.96795952e-01
2.65678167e-01 2.16674805e-01 6.85159266e-01 2.14553967e-01
-3.52176577e-01 1.09757411e+00 -7.17650175e-01 1.25507152e+00
-9.32395577e-01 7.43690133e-01 1.86044089e-02 -6.89432323e-01
5.75950384e-01 4.54622507e-02 -5.45523584e-01 -1.04571617e+00
3.34632024e-02 7.34816492e-01 7.74525881e-01 -4.08576094e-02
6.18177176e-01 -3.88107300e-01 -2.50103742e-01 1.97588399e-01
2.73301035e-01 -3.18203062e-01 5.23931719e-02 -3.77590209e-01
5.21450937e-01 -4.36752811e-02 4.18340117e-01 -1.12080085e+00
4.06074882e-01 -1.62426084e-01 6.47341192e-01 6.24686003e-01
1.78378716e-01 1.18751107e-02 2.96484023e-01 -2.67902225e-01
-1.02420115e+00 -1.48878336e+00 -3.65136057e-01 1.38160145e+00
-3.98607522e-01 -3.97514224e-01 -8.10774982e-01 -3.13523889e-01
1.14396572e-01 1.35779202e+00 -3.07795972e-01 -1.57155663e-01
-8.72841775e-01 -3.79782766e-01 7.05658972e-01 3.53861123e-01
7.99880102e-02 -9.37143922e-01 -9.04156920e-03 1.76404007e-02
9.78099406e-02 -8.50472689e-01 -6.13117397e-01 1.63871989e-01
-6.93417668e-01 -4.66912717e-01 -5.04750729e-01 -1.28034377e+00
5.54805219e-01 -1.76510140e-01 1.42773068e+00 8.45646635e-02
3.44307452e-01 1.62922233e-01 2.06998512e-02 -6.56474829e-01
-8.42615306e-01 4.95476842e-01 4.50844675e-01 -4.09568220e-01
7.47072637e-01 -3.37726593e-01 -1.17346950e-01 -1.45578280e-01
-5.53701460e-01 1.48868129e-01 3.74606729e-01 5.42621851e-01
4.49155360e-01 -4.23347950e-02 -3.48840258e-03 -1.07310617e+00
8.28145027e-01 -5.20101309e-01 -8.87064338e-01 2.82065958e-01
-7.02421784e-01 1.68554723e-01 1.12302709e+00 -4.47250456e-01
-9.14996564e-01 -2.11367905e-01 -1.48166075e-01 2.10350811e-01
7.46106133e-02 6.51641488e-01 -2.51745909e-01 2.78652809e-03
5.12081504e-01 4.06117529e-01 -2.32727960e-01 -7.16339171e-01
5.04600525e-01 5.79302609e-01 7.20804393e-01 -9.12684679e-01
1.08070946e+00 -1.50977552e-01 -5.54395854e-01 -1.12913954e+00
-6.86832070e-01 -1.36942700e-01 -5.23395479e-01 2.10785046e-01
9.51384306e-01 -1.13298261e+00 -5.94935060e-01 4.73837078e-01
-1.22314978e+00 -5.71331799e-01 -2.07390174e-01 7.26566553e-01
-3.36206526e-01 2.66632050e-01 -7.61881888e-01 -4.05884922e-01
-1.04076646e-01 -1.03885806e+00 7.84630835e-01 2.20764801e-02
-6.97543502e-01 -1.50141525e+00 5.61060347e-02 1.87628463e-01
3.88848573e-01 -5.04227519e-01 1.55924022e+00 -8.06708694e-01
-4.61782098e-01 2.33746439e-01 -1.09895684e-01 6.57897770e-01
1.08232208e-01 -2.12581858e-01 -6.88606799e-01 -3.75029743e-01
2.55458057e-01 -1.82632059e-01 5.72529495e-01 9.04942825e-02
8.65378737e-01 -7.17186630e-01 3.97267193e-02 1.05936193e+00
1.49163961e+00 -1.04982611e-02 5.18800020e-01 3.51525724e-01
8.93537343e-01 6.97740316e-01 3.36012384e-03 -3.36006403e-01
8.87999475e-01 5.22730827e-01 -5.76847494e-01 -1.52086034e-01
-4.02080834e-01 -6.40203536e-01 8.99266183e-01 1.78501236e+00
1.38947815e-01 -1.51417315e-01 -1.30596542e+00 7.33759105e-01
-8.83736253e-01 -3.87289882e-01 -1.37009881e-02 2.42724657e+00
1.13530338e+00 2.56714016e-01 -9.49805230e-02 -1.47744387e-01
3.10477018e-01 2.48152129e-02 -1.97020415e-02 -1.16609299e+00
-4.48293000e-01 6.02008164e-01 6.31881773e-01 1.03855836e+00
-5.85665166e-01 1.60128474e+00 6.72329521e+00 8.65326822e-01
-1.31818068e+00 6.43837526e-02 6.33006215e-01 4.79982108e-01
-7.11559296e-01 4.74118739e-02 -9.71357763e-01 4.14125532e-01
1.26304150e+00 -3.15616310e-01 1.01935661e+00 4.59576607e-01
-6.37656003e-02 1.13550164e-01 -1.75759542e+00 8.09334219e-01
1.05469987e-01 -8.83683801e-01 3.97626698e-01 -1.04772002e-01
6.25643373e-01 2.47449443e-01 2.47890085e-01 6.82835400e-01
5.12810886e-01 -1.45280159e+00 9.21415687e-01 3.34615231e-01
1.19372571e+00 -4.83900726e-01 5.49033135e-02 3.87805551e-01
-9.59644675e-01 5.77144958e-02 -3.97798836e-01 -5.88779986e-01
-1.75820097e-01 -1.12091415e-01 -8.41343462e-01 -6.66114315e-02
2.70443082e-01 1.63023815e-01 -7.94311583e-01 5.22579730e-01
-4.23823804e-01 8.99096429e-01 -2.35287666e-01 -1.07105166e-01
1.71748281e-01 -4.12041068e-01 4.22096759e-01 1.62230563e+00
4.97031420e-01 -7.41246492e-02 9.67063680e-02 9.82719362e-01
-9.65894759e-02 7.14779854e-01 -5.99471033e-01 -3.67848486e-01
6.22372687e-01 7.66531050e-01 -4.58990574e-01 -2.47576490e-01
-8.05411577e-01 1.15221679e+00 7.72414982e-01 3.17083627e-01
-6.49846792e-02 -2.92954266e-01 7.71423161e-01 2.72388935e-01
-1.85215250e-01 -6.65617704e-01 -4.60681409e-01 -1.32909286e+00
-1.71688482e-01 -1.20422947e+00 1.19115703e-01 -5.73304653e-01
-1.43938851e+00 3.65564734e-01 -2.20443636e-01 -4.62823004e-01
-3.00780833e-01 -1.17802036e+00 -4.88450825e-01 1.39164209e+00
-1.40348625e+00 -1.14867926e+00 3.99743348e-01 3.74211103e-01
5.11881590e-01 -3.67400289e-01 1.00027037e+00 1.02038950e-01
-3.87060553e-01 1.08168149e+00 1.43441707e-01 4.55173969e-01
4.45707321e-01 -1.38405395e+00 9.21338141e-01 8.64302874e-01
6.61368728e-01 9.76753771e-01 4.15299594e-01 -5.71050167e-01
-1.44727719e+00 -7.87367225e-01 1.84390831e+00 -8.46786439e-01
1.06423306e+00 -7.05007970e-01 -8.75232160e-01 1.13714731e+00
3.01208943e-01 -5.56442797e-01 5.97826898e-01 6.75768197e-01
-6.11776769e-01 3.18945013e-02 -6.98426604e-01 1.04506409e+00
1.16306484e+00 -1.27853358e+00 -9.76596415e-01 3.19491684e-01
6.74226761e-01 -1.31206095e-01 -8.93076479e-01 -1.90233998e-02
4.73850816e-01 -6.90374613e-01 5.67454755e-01 -7.42228091e-01
2.05838859e-01 -4.23009545e-02 -2.08446935e-01 -1.66674459e+00
-7.89721429e-01 -6.52640343e-01 4.91409510e-01 1.24495423e+00
5.92662394e-01 -8.02430689e-01 3.23598653e-01 2.88045049e-01
-1.81515396e-01 -2.89355785e-01 -9.33437347e-01 -1.01554561e+00
1.00919771e+00 -4.36751068e-01 4.97906983e-01 1.32189989e+00
7.50108808e-02 5.99101484e-01 1.87780246e-01 5.07411212e-02
1.41666085e-01 -3.43646824e-01 4.39526677e-01 -1.29304194e+00
-5.21673441e-01 -9.01679814e-01 -3.19314122e-01 -8.87242317e-01
6.76679015e-01 -1.64813375e+00 -2.76140898e-01 -9.67181802e-01
1.39752448e-01 -3.57439727e-01 -2.17695013e-01 2.15880916e-01
-1.71392307e-01 2.02349667e-02 5.07539213e-01 1.48197375e-02
-1.33607328e-01 2.85367910e-02 1.04811168e+00 -3.65937054e-02
-2.07076460e-01 -2.13086277e-01 -8.46779406e-01 9.51380730e-01
6.73017502e-01 -2.31011167e-01 -2.04648554e-01 -1.06363189e+00
4.73487973e-01 -1.11882254e-01 -1.10200502e-01 -8.06298852e-01
1.60993431e-02 -2.03124091e-01 3.16494823e-01 2.77517796e-01
9.00645033e-02 -5.16028583e-01 -1.96519401e-02 3.97086591e-01
-2.57049292e-01 6.63214743e-01 4.03014958e-01 -2.31356919e-01
-1.24710843e-01 -2.75402278e-01 6.59728944e-01 -3.04074585e-01
-7.45053232e-01 3.62024158e-02 -7.72403359e-01 5.74881017e-01
3.50550324e-01 -2.74803907e-01 -9.99013186e-02 -1.52991146e-01
-3.71836096e-01 -1.04757480e-01 8.27562869e-01 5.91727078e-01
-4.65659238e-02 -1.33196437e+00 -9.90708113e-01 4.07949895e-01
1.02527685e-01 -7.36808956e-01 -3.94585878e-01 4.22454655e-01
-6.38405740e-01 7.10157931e-01 -3.30305457e-01 -8.50814432e-02
-9.01258826e-01 3.90836358e-01 5.84379792e-01 -1.92431360e-01
-1.73506469e-01 9.82954025e-01 4.97427255e-01 -7.89019823e-01
1.23466246e-01 -4.59427148e-01 5.21141849e-02 1.45109110e-02
1.52522922e-01 2.62904823e-01 2.61798017e-02 -9.98818576e-01
-3.72102648e-01 5.98529100e-01 -1.65941462e-01 -1.85394675e-01
8.28293800e-01 -4.09924313e-02 -1.99086338e-01 7.36053884e-01
1.18210483e+00 7.84118652e-01 -7.85339773e-01 -1.19603008e-01
2.74375319e-01 -3.55783142e-02 -1.42852113e-01 -7.85393953e-01
-4.36364770e-01 9.24531877e-01 2.76174188e-01 -1.26681075e-01
6.85050726e-01 -2.01172441e-01 5.98397851e-01 4.85686570e-01
2.15618566e-01 -1.54602230e+00 -5.84882498e-01 1.20145833e+00
7.70045102e-01 -1.04806387e+00 -3.54826778e-01 -2.49281496e-01
-5.45424283e-01 8.16971719e-01 3.91482562e-01 -1.91471085e-01
5.46497107e-01 1.79593280e-01 1.38214916e-01 -3.11230589e-02
-4.50435996e-01 -1.43635526e-01 6.88868761e-01 6.69573069e-01
9.75062788e-01 6.21281028e-01 -6.87506735e-01 6.91670001e-01
-1.23160172e+00 -5.74170768e-01 3.03290755e-01 2.59611636e-01
-2.07642809e-01 -1.24593985e+00 -2.63505191e-01 4.13784117e-01
-6.14351988e-01 -8.81416321e-01 -5.56368291e-01 1.09723020e+00
2.70229608e-01 6.14690483e-01 5.06305933e-01 -2.84053147e-01
1.63952053e-01 4.43081409e-01 7.61417985e-01 -9.96342480e-01
-7.10606158e-01 -3.07578743e-01 3.01281482e-01 -2.23700270e-01
2.51430869e-01 -7.48297334e-01 -1.01294529e+00 -4.06214684e-01
8.12934041e-02 -3.67254876e-02 5.99112928e-01 9.81087446e-01
-1.08715504e-01 1.42247871e-01 8.23489428e-02 -4.36889559e-01
-7.63968825e-01 -8.62744153e-01 -6.96133852e-01 4.74386901e-01
1.15210027e-01 -1.24468260e-01 -4.70410436e-01 -1.42350301e-01] | [10.869118690490723, 9.914188385009766] |
2a4c18f4-5cb2-4141-b04b-9f954ff41d2f | unitail-detecting-reading-and-matching-in | 2204.00298 | null | https://arxiv.org/abs/2204.00298v4 | https://arxiv.org/pdf/2204.00298v4.pdf | Unitail: Detecting, Reading, and Matching in Retail Scene | To make full use of computer vision technology in stores, it is required to consider the actual needs that fit the characteristics of the retail scene. Pursuing this goal, we introduce the United Retail Datasets (Unitail), a large-scale benchmark of basic visual tasks on products that challenges algorithms for detecting, reading, and matching. With 1.8M quadrilateral-shaped instances annotated, the Unitail offers a detection dataset to align product appearance better. Furthermore, it provides a gallery-style OCR dataset containing 1454 product categories, 30k text regions, and 21k transcriptions to enable robust reading on products and motivate enhanced product matching. Besides benchmarking the datasets using various state-of-the-arts, we customize a new detector for product detection and provide a simple OCR-based matching solution that verifies its effectiveness. | ['Marios Savvides', 'Chenchen Zhu', 'Uzair Ahmed', 'Yongxin Zhang', 'Hao Chen', 'Shentong Mo', 'Jiachen Dou', 'Zaiwang Li', 'Han Zhang', 'Fangyi Chen'] | 2022-04-01 | null | null | null | null | ['dense-object-detection'] | ['computer-vision'] | [ 5.16480446e-01 -1.89693540e-01 -2.85876602e-01 -5.33538163e-01
-9.33014989e-01 -1.17392147e+00 5.70009291e-01 3.10706943e-01
6.39896393e-02 -3.72755110e-01 2.05657288e-01 -3.37454885e-01
2.07037777e-01 -5.27137637e-01 -7.58008003e-01 -2.50907987e-01
1.64835513e-01 3.88439417e-01 -1.43762752e-02 -4.44138736e-01
4.28877711e-01 6.02648675e-01 -1.59616518e+00 7.40643740e-01
5.35935104e-01 1.28914809e+00 1.63713336e-01 8.03585351e-01
8.42678770e-02 2.91985869e-01 -4.12302613e-01 -1.08908033e+00
9.35982466e-01 4.67616990e-02 -5.23742020e-01 7.76823878e-01
1.28977668e+00 -2.42300615e-01 -2.83744127e-01 1.01672125e+00
4.97881681e-01 -4.47650775e-02 7.75680661e-01 -1.26298463e+00
-1.46790695e+00 4.45426047e-01 -9.76920605e-01 1.99036136e-01
6.92194700e-01 6.47975087e-01 1.28196776e+00 -1.03072703e+00
7.43617237e-01 1.32900691e+00 7.17641652e-01 -1.37801200e-01
-1.44790220e+00 -4.57851350e-01 4.71001975e-02 6.67716116e-02
-1.25528300e+00 -5.55068612e-01 6.41594827e-01 -3.77679259e-01
1.04282832e+00 3.30443650e-01 6.94060087e-01 1.02929306e+00
1.19012073e-01 9.63764846e-01 1.08308935e+00 -4.06827331e-01
3.87672335e-02 9.21242684e-02 2.01506943e-01 4.57646549e-01
3.06583881e-01 2.26746812e-01 -4.20884252e-01 3.10710758e-01
5.32452285e-01 -2.95737069e-02 -1.42189324e-01 -5.71169019e-01
-9.51951146e-01 8.24980795e-01 4.35906529e-01 -1.56569555e-01
1.37089530e-03 -6.37795031e-02 4.61918950e-01 2.00403735e-01
3.67440879e-01 5.28308272e-01 -3.22930008e-01 1.52148038e-01
-1.04768550e+00 3.15122932e-01 4.58182037e-01 1.51620877e+00
6.05813265e-01 -5.58843538e-02 -2.53524452e-01 9.82973635e-01
4.83111471e-01 8.62760723e-01 2.66545080e-02 -9.01208341e-01
7.00879514e-01 6.99481308e-01 1.13474905e-01 -1.34856331e+00
-3.45512480e-01 -4.70827222e-01 -4.62820351e-01 1.82372615e-01
5.94927132e-01 5.82323253e-01 -1.18770194e+00 8.69848788e-01
1.69801503e-01 -3.03488165e-01 -2.20516518e-01 9.31754053e-01
7.95182168e-01 5.96170783e-01 -1.92169905e-01 3.10322672e-01
1.83058262e+00 -1.31751966e+00 -5.08151531e-01 -7.22768307e-01
6.24607921e-01 -1.44438910e+00 1.36373222e+00 5.51662624e-01
-1.12097812e+00 -9.02985156e-01 -1.41379023e+00 -3.98019135e-01
-7.70230353e-01 2.76816934e-01 6.53670430e-01 8.68274808e-01
-8.13788533e-01 2.16368198e-01 -1.50154278e-01 -5.38478732e-01
2.04917699e-01 -1.89648792e-01 -1.67556629e-01 -5.04548848e-01
-6.42320335e-01 1.00279522e+00 3.81349713e-01 2.55544245e-01
-6.83573246e-01 -7.86339998e-01 -1.24994326e+00 -1.15380503e-01
1.08477131e-01 -2.02462718e-01 1.43018150e+00 -9.43261802e-01
-1.03090501e+00 1.29699838e+00 2.14339361e-01 -3.75388741e-01
4.05656397e-01 -1.11355763e-02 -9.05113339e-01 1.39119267e-01
1.95045039e-01 8.43230665e-01 1.11371660e+00 -1.41440463e+00
-7.84220099e-01 -5.34191191e-01 -4.93889526e-02 3.46017294e-02
1.00488700e-01 2.03552544e-01 -9.10181224e-01 -8.01473081e-01
-2.53248047e-02 -8.48954320e-01 1.08064428e-01 5.71388789e-02
-5.21210790e-01 2.25868806e-01 6.28080487e-01 -9.47812080e-01
1.02953064e+00 -2.55546451e+00 -4.36168820e-01 3.18011463e-01
4.49772961e-02 1.22634903e-01 -6.40184045e-01 2.83649802e-01
-1.87651739e-01 -2.29721796e-02 1.83265090e-01 -5.57357550e-01
5.03324091e-01 -1.54057527e-02 -2.87927210e-01 5.68958521e-01
1.00544691e-01 1.49004090e+00 -5.21566212e-01 -6.09163463e-01
4.56731141e-01 1.49456367e-01 -2.61637419e-01 -2.17130318e-01
-1.07865095e-01 -4.68080401e-01 5.39425835e-02 1.31485951e+00
1.19404256e+00 -9.47454646e-02 -1.23074360e-01 -6.35071099e-01
5.97451478e-02 -2.67324865e-01 -1.36767900e+00 1.51537383e+00
-3.30444515e-01 8.27213764e-01 2.25977242e-01 -4.51766461e-01
1.04468477e+00 -4.01477724e-01 1.65818065e-01 -1.42667806e+00
1.37845308e-01 1.17872037e-01 -2.70939559e-01 -2.32127249e-01
1.20864248e+00 6.48776829e-01 -6.40664026e-02 2.20493495e-01
-1.91440195e-01 -3.87211978e-01 5.09207428e-01 2.20088422e-01
7.92793930e-01 -1.83658041e-02 -9.82495695e-02 -2.22766027e-01
3.26587260e-02 3.14504415e-01 5.82338907e-02 7.89157271e-01
-3.39133054e-01 1.14795089e+00 -6.52660951e-02 -6.70006990e-01
-1.54941261e+00 -1.21097267e+00 -3.04087371e-01 1.21843171e+00
3.43736976e-01 -4.78235453e-01 -6.33224428e-01 -4.13533747e-01
4.18086410e-01 6.36128724e-01 -6.91542387e-01 2.02772561e-02
-2.29811147e-01 -5.95653236e-01 4.23998177e-01 6.02346003e-01
6.56075478e-01 -9.37118530e-01 -4.27386373e-01 2.21644100e-02
6.26932010e-02 -1.45437908e+00 -1.05369198e+00 1.36017188e-01
-3.46381724e-01 -1.19447935e+00 -6.45851851e-01 -1.09598124e+00
3.92843187e-01 7.15924680e-01 1.58238685e+00 -4.19390410e-01
-8.07265520e-01 4.45366442e-01 -4.40036446e-01 -3.08745652e-01
-4.57009614e-01 1.39434665e-01 -4.14688110e-01 5.50839379e-02
6.42107189e-01 1.06487364e-01 -8.05425406e-01 6.33484304e-01
-7.82540739e-01 -6.91797435e-02 6.96712255e-01 3.21341157e-01
5.73431015e-01 -7.04544485e-02 -2.21965332e-02 -4.09165323e-01
6.40552402e-01 1.93372518e-01 -7.04464078e-01 5.12252629e-01
-8.30750763e-01 -2.53725171e-01 2.42800251e-01 -5.37311316e-01
-8.57974231e-01 4.45156187e-01 -3.94503353e-03 -1.82715967e-01
-1.55514389e-01 -1.06428750e-01 -2.02463437e-02 3.89638282e-02
7.57118523e-01 3.38622391e-01 -9.86770391e-02 -6.09593987e-01
9.68881965e-01 9.76721168e-01 9.97572780e-01 -8.03668946e-02
8.47488761e-01 2.30532587e-01 -5.89421511e-01 -7.04264104e-01
-7.54771233e-01 -8.72719467e-01 -6.92071676e-01 -1.42819598e-01
9.56562698e-01 -9.48038995e-01 -6.58309460e-01 4.20246065e-01
-1.02493048e+00 -2.22144201e-01 -3.76679778e-01 -1.06395431e-01
-3.98779839e-01 3.93246233e-01 -8.48732114e-01 -5.25728762e-01
-5.06742179e-01 -1.15945876e+00 1.69573414e+00 1.57826096e-01
-1.07429922e-01 -4.29104000e-01 -1.94900960e-01 7.84542501e-01
2.17574492e-01 8.15230682e-02 6.66024089e-01 -3.65703583e-01
-4.59105194e-01 -2.73550957e-01 -6.39561713e-01 1.98953971e-01
-2.06109434e-02 4.04776722e-01 -9.74779367e-01 -3.59902292e-01
-6.19562030e-01 -1.72885224e-01 8.46541703e-01 2.93360889e-01
7.81929493e-01 3.26837860e-02 -1.71679199e-01 6.66442692e-01
1.46595740e+00 1.99499309e-01 8.66575539e-01 7.42779970e-01
6.81534171e-01 5.12805700e-01 1.04145026e+00 4.71985310e-01
3.82759899e-01 9.21237111e-01 6.23460233e-01 -6.87277734e-01
-2.92031318e-01 -3.22992921e-01 2.87355483e-01 2.19947234e-01
2.85611957e-01 -6.73883185e-02 -5.29978395e-01 5.52484393e-01
-1.60784888e+00 -8.58412802e-01 -4.41968501e-01 1.83205366e+00
5.71631670e-01 2.88649917e-01 3.92145753e-01 -2.69294903e-02
6.56227767e-01 3.15808624e-01 -5.82497835e-01 -9.06348169e-01
-4.76982534e-01 7.11281225e-02 1.19726026e+00 1.91041052e-01
-1.36075675e+00 8.13544512e-01 7.59066153e+00 9.67362165e-01
-7.04456925e-01 -7.35746101e-02 8.11566174e-01 1.53743654e-01
-2.29885235e-01 -4.81480628e-01 -1.05203080e+00 5.16246796e-01
5.56871951e-01 2.78681159e-01 4.79540706e-01 1.12381375e+00
1.95512235e-01 -1.47765830e-01 -1.16302800e+00 1.33016610e+00
3.76401752e-01 -1.35290289e+00 -1.37781039e-01 2.11627468e-01
8.82906914e-01 -1.89032078e-01 5.48411906e-01 3.24365288e-01
5.08666873e-01 -1.04967225e+00 1.15539658e+00 -8.66809487e-02
7.44447947e-01 -6.61935329e-01 5.94508171e-01 -3.44027340e-01
-1.63762391e+00 -1.82997465e-01 -4.94808882e-01 3.57305288e-01
2.55613446e-01 3.73073339e-01 -8.03848088e-01 3.59648883e-01
9.66522396e-01 7.78579354e-01 -1.41007042e+00 1.06880355e+00
1.49108678e-01 6.23653419e-02 -3.53671759e-01 2.25193113e-01
2.81854779e-01 -2.94243515e-01 -1.28390834e-01 1.58930695e+00
1.33389413e-01 -4.29995716e-01 2.72536099e-01 8.03400576e-01
2.38012653e-02 2.47194454e-01 -5.84107637e-01 -1.20752223e-01
2.91977733e-01 1.62201571e+00 -9.23064351e-01 -3.80231738e-02
-4.61727500e-01 1.21156287e+00 -1.42213389e-01 2.96704412e-01
-1.09485257e+00 -2.87838876e-01 7.92236328e-01 1.56460553e-01
6.41805649e-01 -3.33267272e-01 -4.29099351e-01 -7.57310152e-01
1.44043267e-01 -1.19347548e+00 2.57706195e-01 -1.13601184e+00
-1.35949600e+00 2.50964850e-01 -1.18705817e-01 -1.21889734e+00
9.88750085e-02 -1.08572674e+00 -2.21017867e-01 4.35261369e-01
-1.33055365e+00 -1.64624166e+00 -5.51757693e-01 4.66805100e-01
9.38619792e-01 -2.14504544e-02 2.88808703e-01 6.01768911e-01
-5.43330312e-01 9.14917111e-01 3.93948495e-01 2.13354900e-01
8.83708298e-01 -1.29247308e+00 1.09136891e+00 7.51916528e-01
4.33614522e-01 4.13234949e-01 8.42390180e-01 -3.50130737e-01
-1.73897922e+00 -1.10011995e+00 5.03983736e-01 -6.83967888e-01
8.35896313e-01 -7.51924753e-01 -4.18644100e-01 7.34159291e-01
3.25371295e-01 -1.69669986e-01 3.11222792e-01 -9.22576115e-02
-8.07532609e-01 -3.42006624e-01 -1.05623257e+00 6.40575528e-01
1.13250947e+00 -5.45593739e-01 -4.91985917e-01 5.15423954e-01
6.19155228e-01 -5.03780961e-01 -9.86695766e-01 1.04002893e-01
7.25105941e-01 -8.83823097e-01 1.35847747e+00 -2.07854554e-01
4.75614041e-01 -3.81331533e-01 -3.87580425e-01 -1.19085324e+00
-4.45606381e-01 -3.08254153e-01 2.72740096e-01 1.45191646e+00
6.95471346e-01 -2.62200356e-01 8.23301136e-01 7.23613143e-01
-2.45433539e-01 -4.39388603e-01 -3.32826227e-01 -7.59993613e-01
-3.23990285e-01 -4.94836688e-01 6.95510685e-01 6.32675588e-01
-3.61230791e-01 1.94289535e-01 -2.86075890e-01 6.86175525e-02
5.52646577e-01 4.85174030e-01 8.70259464e-01 -8.95677805e-01
-4.04008389e-01 -4.98652995e-01 -4.99185979e-01 -1.26895249e+00
-2.96299607e-01 -7.02997744e-01 6.50634170e-02 -1.49755073e+00
2.72466987e-01 -1.75572485e-01 1.58182845e-01 3.97680402e-01
1.68091521e-01 9.81362104e-01 5.87458730e-01 2.16332600e-02
-8.11709404e-01 3.43431048e-02 1.13519239e+00 -8.17265570e-01
-4.02352400e-02 -2.19105482e-01 -8.26120555e-01 4.43060130e-01
6.57702386e-01 8.66265446e-02 -1.60255402e-01 -4.14254040e-01
2.35251978e-01 -7.36908317e-01 2.07393348e-01 -4.79792774e-01
-1.67320650e-02 1.35752156e-01 8.13408911e-01 -1.10739422e+00
3.83658737e-01 -8.47866476e-01 1.71489984e-01 1.74407646e-01
-3.45417231e-01 5.49759984e-01 1.05470806e-01 4.58789408e-01
-1.40155470e-02 -2.11108401e-01 7.75086284e-01 -8.69757459e-02
-1.25218666e+00 -2.91189807e-03 -1.07509725e-01 7.54548609e-02
1.14502275e+00 -9.55820262e-01 -6.65272772e-01 -1.23420112e-01
-5.30359924e-01 2.96868354e-01 8.84304881e-01 8.48982394e-01
5.19615412e-01 -1.30802393e+00 -7.21195459e-01 2.59786040e-01
7.47345507e-01 -3.91913921e-01 8.38159323e-02 3.76432627e-01
-9.84375358e-01 4.74366218e-01 -1.47459909e-01 -7.51136482e-01
-1.31655681e+00 1.12285233e+00 1.93609416e-01 -8.57945085e-02
-5.44965148e-01 5.62829971e-01 3.65780629e-02 -1.98833153e-01
1.15715161e-01 -4.36858147e-01 4.65516038e-02 3.28073949e-01
6.51466072e-01 3.33785117e-01 6.51855767e-01 -8.26191366e-01
-1.93253040e-01 8.38692904e-01 -2.64657140e-01 1.11067280e-01
8.88112962e-01 -4.17889148e-01 4.90230322e-01 -1.46348640e-01
1.34176743e+00 -4.05809171e-02 -1.35529494e+00 -1.68493658e-01
9.12559927e-02 -7.40029335e-01 -6.36923760e-02 -8.80456209e-01
-1.26369023e+00 5.44840574e-01 1.11152077e+00 4.31954861e-01
1.15176439e+00 1.00944735e-01 7.89095461e-01 -3.03535294e-02
3.03010017e-01 -1.46849167e+00 1.87396124e-01 1.43233672e-01
8.48739326e-01 -1.45769012e+00 3.91750410e-02 -7.82277107e-01
-9.29424405e-01 1.00072598e+00 4.33415055e-01 6.43087924e-02
1.19274341e-01 4.34617400e-01 2.98039258e-01 -2.82356262e-01
-1.99500024e-01 -4.96232063e-01 4.24719691e-01 9.59197640e-01
8.56840461e-02 5.75010031e-02 1.02249295e-01 3.24517898e-02
-4.12346393e-01 -5.87912500e-01 2.24339187e-01 7.11387157e-01
-3.69376600e-01 -8.51614296e-01 -5.77169418e-01 3.70785654e-01
-2.04894856e-01 -2.00643286e-01 -4.78048563e-01 9.18493807e-01
2.59593017e-02 1.25895834e+00 2.52449811e-01 -3.00972432e-01
7.06565738e-01 -3.46050858e-01 5.14558077e-01 -3.99259418e-01
-8.02289724e-01 1.87490255e-01 1.98271543e-01 -7.96543241e-01
-8.85301009e-02 -6.46983266e-01 -6.51475787e-01 -5.34103811e-01
-2.86928743e-01 -4.89750236e-01 7.93617129e-01 4.62397039e-01
5.30246854e-01 1.84244424e-01 4.90663826e-01 -1.05079114e+00
-6.71529710e-01 -7.94402063e-01 -1.00889099e+00 9.35506046e-01
9.33178961e-02 -4.49657947e-01 -2.49687955e-01 3.44020098e-01] | [10.921278953552246, 1.4068013429641724] |
a6ab5b97-bc31-4363-b403-53332846b94a | viewnet-unsupervised-viewpoint-estimation-1 | 2212.00435 | null | https://arxiv.org/abs/2212.00435v1 | https://arxiv.org/pdf/2212.00435v1.pdf | ViewNet: Unsupervised Viewpoint Estimation from Conditional Generation | Understanding the 3D world without supervision is currently a major challenge in computer vision as the annotations required to supervise deep networks for tasks in this domain are expensive to obtain on a large scale. In this paper, we address the problem of unsupervised viewpoint estimation. We formulate this as a self-supervised learning task, where image reconstruction provides the supervision needed to predict the camera viewpoint. Specifically, we make use of pairs of images of the same object at training time, from unknown viewpoints, to self-supervise training by combining the viewpoint information from one image with the appearance information from the other. We demonstrate that using a perspective spatial transformer allows efficient viewpoint learning, outperforming existing unsupervised approaches on synthetic data, and obtains competitive results on the challenging PASCAL3D+ dataset. | ['Hakan Bilen', 'Oisin Mac Aodha', 'Octave Mariotti'] | 2022-12-01 | viewnet-unsupervised-viewpoint-estimation | http://openaccess.thecvf.com//content/ICCV2021/html/Mariotti_ViewNet_Unsupervised_Viewpoint_Estimation_From_Conditional_Generation_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Mariotti_ViewNet_Unsupervised_Viewpoint_Estimation_From_Conditional_Generation_ICCV_2021_paper.pdf | iccv-2021-1 | ['viewpoint-estimation'] | ['computer-vision'] | [ 2.82303602e-01 1.29985601e-01 6.73704520e-02 -8.14613819e-01
-4.70136017e-01 -6.97428703e-01 5.29223323e-01 -1.34016007e-01
-4.78809744e-01 2.67459482e-01 -1.83297753e-01 -6.70599565e-02
4.63541359e-01 -4.01607454e-01 -1.18620276e+00 -5.43897510e-01
4.12189871e-01 8.28081489e-01 3.75999838e-01 1.48599759e-01
1.17652059e-01 4.96032923e-01 -1.39114940e+00 2.28205532e-01
3.55154216e-01 9.06191528e-01 5.24576724e-01 6.57017887e-01
4.03808266e-01 9.12024498e-01 -2.70269752e-01 -1.96732610e-01
5.38741767e-01 -1.08690470e-01 -5.66600263e-01 7.34233081e-01
8.44019890e-01 -8.44232321e-01 -3.79955083e-01 8.93565655e-01
1.91080779e-01 2.04684749e-01 6.67025328e-01 -1.10385263e+00
-2.27315709e-01 -1.47088826e-01 -4.97096568e-01 3.65295522e-02
1.56555176e-01 -1.24437232e-02 8.36053431e-01 -8.49532008e-01
8.55554461e-01 1.02902377e+00 3.02917838e-01 4.86456305e-01
-1.37820792e+00 -1.07838288e-01 4.11064625e-01 1.93099335e-01
-1.15036714e+00 -5.88160813e-01 1.08159244e+00 -6.09822154e-01
9.64281142e-01 -3.83028418e-01 6.05132103e-01 9.00633991e-01
-6.97384700e-02 7.82971740e-01 9.80769515e-01 -4.33526844e-01
3.70325983e-01 2.83952922e-01 -2.32377082e-01 8.20149958e-01
5.24514774e-03 1.50234386e-01 -5.50475538e-01 3.45911920e-01
1.00225163e+00 1.40100524e-01 -1.59697235e-01 -1.17615235e+00
-9.15557444e-01 6.10612035e-01 5.82980156e-01 -1.96066946e-01
-2.97590673e-01 -3.55148055e-02 -7.72935674e-02 3.32328409e-01
6.86895013e-01 2.65525371e-01 -7.29311109e-01 1.48825794e-01
-6.11596346e-01 -1.54038947e-02 5.56562841e-01 1.04370511e+00
7.40293622e-01 5.87038249e-02 5.78811169e-01 5.83021700e-01
3.73015195e-01 5.39089978e-01 -8.63920897e-03 -1.44655097e+00
5.02962053e-01 5.82387865e-01 3.44443619e-01 -7.23728836e-01
-3.67402822e-01 -3.82682711e-01 -5.32816172e-01 6.26372695e-01
7.45548248e-01 1.07293934e-01 -9.03165102e-01 1.67711234e+00
6.21020734e-01 -4.48657349e-02 1.22457504e-01 8.92959595e-01
4.60091799e-01 5.18257022e-01 -4.76956785e-01 -5.19717112e-02
9.81244624e-01 -1.10304153e+00 -2.58010954e-01 -6.29966319e-01
3.91158849e-01 -7.59088874e-01 6.72014654e-01 5.07816315e-01
-1.16109824e+00 -6.88409686e-01 -1.06644785e+00 -3.11031342e-01
-6.76501263e-03 4.62367773e-01 3.40971470e-01 1.17008671e-01
-1.02779937e+00 4.18858200e-01 -1.18748677e+00 -3.16365451e-01
4.09214526e-01 3.63427222e-01 -7.34344184e-01 -4.11487490e-01
-3.76838088e-01 1.07177472e+00 1.47398591e-01 6.86414465e-02
-1.36723876e+00 -2.89721519e-01 -1.07641137e+00 -1.61723644e-01
4.18253690e-01 -6.10546529e-01 1.35399783e+00 -1.23267639e+00
-1.38928437e+00 1.23888075e+00 -1.63457856e-01 -4.45292652e-01
6.84208512e-01 -2.76914150e-01 1.68150321e-01 3.28255177e-01
8.21831524e-02 9.44233358e-01 1.02805614e+00 -1.53485572e+00
-4.57611531e-01 -8.55398715e-01 3.32122892e-01 5.12672305e-01
-8.43188353e-03 -3.11203480e-01 -6.80706680e-01 -9.19962674e-02
4.05307472e-01 -1.22250724e+00 -2.71078706e-01 3.42143804e-01
-2.85603255e-01 -1.79358311e-02 8.64734769e-01 -6.25850558e-01
-9.79566053e-02 -2.18587780e+00 3.90490562e-01 -2.10111454e-01
2.46477395e-01 4.46694419e-02 4.30722125e-02 4.12207842e-02
-5.23554683e-02 -6.73421621e-01 -9.26889181e-02 -7.54151881e-01
-3.21009070e-01 4.30053949e-01 -1.03755437e-01 8.21319759e-01
1.39198646e-01 6.93026245e-01 -8.62146020e-01 -2.37305567e-01
5.48302591e-01 4.24850464e-01 -6.45577431e-01 6.20930135e-01
-4.60148782e-01 8.81774724e-01 -7.64923841e-02 2.26476371e-01
4.84167248e-01 -5.89031875e-01 3.56419206e-01 -3.30736309e-01
3.66959907e-02 3.40810448e-01 -1.00377357e+00 2.10280180e+00
-7.40130246e-01 7.34518290e-01 2.17205927e-01 -1.26428401e+00
7.62253284e-01 2.83268809e-01 3.35856169e-01 -4.42842036e-01
2.51647472e-01 -3.43091115e-02 -9.16040093e-02 -5.39217532e-01
8.10108893e-03 -1.74038053e-01 2.30342478e-01 4.25583273e-01
4.09987420e-01 -5.38799882e-01 -5.20237461e-02 1.67791948e-01
9.03635085e-01 4.17096257e-01 1.77905127e-01 2.42871851e-01
2.54748017e-01 2.11309679e-02 4.50124115e-01 3.28666240e-01
-1.34190340e-02 7.90669978e-01 2.50614405e-01 -7.34478652e-01
-1.29895020e+00 -1.15218842e+00 1.70702383e-01 6.63905203e-01
2.82553732e-01 1.33696869e-02 -5.12391925e-01 -9.46683764e-01
-9.97543186e-02 6.44608498e-01 -5.59066653e-01 9.74218920e-02
-3.78434271e-01 -1.06641099e-01 -2.74415970e-01 7.04146683e-01
4.51050282e-01 -6.88481629e-01 -7.39796102e-01 -7.77282193e-02
-1.69730425e-01 -1.76539302e+00 -4.11690056e-01 3.76781225e-01
-9.36024010e-01 -1.15743816e+00 -5.07035792e-01 -8.63469720e-01
1.26762331e+00 7.60504007e-01 1.22350585e+00 -1.55101791e-01
-1.96909562e-01 6.06149554e-01 4.97333407e-02 -3.71776670e-01
-3.74542147e-01 -2.65280873e-01 3.73745970e-02 1.83731228e-01
9.77414921e-02 -8.79736364e-01 -5.59225917e-01 4.21435535e-01
-6.54232323e-01 3.29318345e-01 6.88461185e-01 6.72437251e-01
6.90703392e-01 -1.17117502e-01 3.71070355e-02 -9.85918999e-01
-4.21530366e-01 -7.42188841e-02 -1.03886747e+00 3.96514982e-02
-2.34447792e-01 2.17111185e-02 7.30943322e-01 -1.67051375e-01
-1.16703582e+00 8.81214797e-01 1.18145756e-01 -8.44891071e-01
-4.65797514e-01 -4.67608236e-02 -4.00421768e-01 -1.29716799e-01
4.91516411e-01 3.38172913e-01 1.92140549e-01 -5.02133369e-01
3.09596270e-01 3.44796836e-01 5.38091540e-01 -2.75595129e-01
9.31417525e-01 9.50331867e-01 4.20654081e-02 -7.63268411e-01
-1.38278413e+00 -5.87056458e-01 -1.29242671e+00 -1.99778184e-01
1.03865421e+00 -1.39284050e+00 -3.90065402e-01 4.23528135e-01
-1.31399238e+00 -4.88989562e-01 -3.30554575e-01 5.54704130e-01
-7.88627803e-01 2.56046146e-01 -2.90616065e-01 -5.51932633e-01
1.73286855e-01 -1.08217096e+00 1.38270867e+00 6.10780977e-02
2.08350405e-01 -1.05307639e+00 1.18751815e-02 8.90839577e-01
-7.87340105e-02 1.37128755e-01 6.34111404e-01 -4.89860088e-01
-9.93961930e-01 -3.54641646e-01 -1.92494035e-01 7.67172575e-01
1.08708747e-01 -2.40477413e-01 -1.14129543e+00 -3.11223924e-01
4.38975811e-01 -6.71523750e-01 6.71746612e-01 3.95140976e-01
1.05842841e+00 2.96722613e-02 -6.44350499e-02 7.67956257e-01
1.28877902e+00 -1.11141354e-02 9.24118459e-02 1.15880067e-03
9.08551753e-01 7.85740077e-01 5.52397013e-01 1.95409432e-01
6.63556933e-01 8.19416523e-01 7.07951069e-01 -6.84765130e-02
-4.07775529e-02 -3.88652682e-01 1.42558053e-01 4.95510846e-01
-8.50547105e-02 -1.12067126e-01 -6.97594583e-01 5.45000613e-01
-1.73734713e+00 -7.58184850e-01 1.52918637e-01 2.43095636e+00
3.86588156e-01 2.66990036e-01 -7.11212382e-02 -6.83286265e-02
4.31435168e-01 1.42171323e-01 -8.32641602e-01 1.27445191e-01
8.43820944e-02 -3.51303786e-01 3.83325249e-01 6.39417291e-01
-1.27594185e+00 8.88633549e-01 5.85699940e+00 -8.12879875e-02
-1.12630332e+00 1.22373752e-01 6.86990976e-01 -1.55632719e-01
6.60230219e-02 1.23370901e-01 -6.70370519e-01 1.01078279e-01
5.14396906e-01 3.48574847e-01 5.22888720e-01 1.11318719e+00
1.56874731e-01 -2.10542783e-01 -1.62333763e+00 1.01220620e+00
5.77185035e-01 -9.62820470e-01 -1.65366337e-01 1.17639173e-02
1.00166023e+00 4.75835174e-01 -8.42809454e-02 -9.86317024e-02
5.03681779e-01 -6.70657218e-01 6.26747072e-01 1.16355710e-01
5.33342600e-01 -3.22444111e-01 4.51980203e-01 8.44601572e-01
-7.61065841e-01 -3.54500301e-03 -4.27733690e-01 -2.25501433e-01
3.46566707e-01 4.78005528e-01 -1.04098213e+00 3.70846927e-01
5.99199951e-01 1.19490623e+00 -6.34792447e-01 8.82834196e-01
-6.01648986e-01 1.60549581e-01 -4.77918684e-01 4.36305612e-01
1.29146166e-02 -3.54286104e-01 4.28600788e-01 3.51180792e-01
5.17688170e-02 -2.70792525e-02 2.39873782e-01 7.31737316e-01
-2.53296107e-01 -3.41555268e-01 -9.32275414e-01 3.37783515e-01
2.81470623e-02 1.16254616e+00 -7.33083487e-01 -2.98759133e-01
-5.27703583e-01 1.22076523e+00 6.08187139e-01 2.26623386e-01
-5.54075360e-01 2.87823170e-01 2.10384205e-01 1.21974751e-01
5.88791013e-01 -5.78145146e-01 2.07013376e-02 -1.41484642e+00
3.09973598e-01 -5.05847931e-01 1.36265248e-01 -1.18936491e+00
-1.04427195e+00 4.77325827e-01 -4.63262834e-02 -1.53140378e+00
-2.90898442e-01 -8.80069315e-01 -4.75678235e-01 4.88991559e-01
-1.52175653e+00 -1.25492477e+00 -5.47441363e-01 5.10179162e-01
8.80007625e-01 1.76335759e-02 5.43121397e-01 5.92027381e-02
-1.41597286e-01 1.40619770e-01 5.24674244e-02 9.64584872e-02
6.48661435e-01 -1.42070305e+00 3.92287791e-01 7.15639770e-01
7.39008307e-01 2.06691131e-01 6.05371118e-01 -2.00150117e-01
-1.51669526e+00 -1.03373325e+00 6.81720078e-01 -8.98198843e-01
2.36598626e-01 -5.36737442e-01 -6.21481061e-01 9.12873864e-01
-2.76513601e-04 5.83244145e-01 4.10266131e-01 -1.73646864e-02
-5.34941614e-01 -2.93395311e-01 -8.09298098e-01 2.40635753e-01
1.05028105e+00 -6.75290942e-01 -5.56952178e-01 7.35297859e-01
5.41950345e-01 -5.81354260e-01 -5.83913386e-01 1.93112135e-01
4.31177884e-01 -1.09453154e+00 1.02598357e+00 -5.51624417e-01
5.02903581e-01 -3.47160965e-01 -3.28158408e-01 -1.36507177e+00
2.24659309e-01 -1.31056637e-01 1.77811205e-01 7.89980233e-01
3.24259669e-01 -3.97029966e-01 1.00432825e+00 5.39403915e-01
-3.77198309e-02 -5.35616875e-01 -8.07505846e-01 -6.93925738e-01
-2.30233312e-01 -1.35675877e-01 -9.26942006e-02 7.57506371e-01
-4.61287379e-01 7.39717722e-01 -4.69769418e-01 3.83654267e-01
9.91925180e-01 3.64679009e-01 1.21446204e+00 -1.21858025e+00
-6.98620737e-01 2.18855739e-01 -6.45144403e-01 -1.42299497e+00
2.70753413e-01 -7.18849599e-01 2.34104648e-01 -1.35865068e+00
2.74318546e-01 -1.67380244e-01 8.29327255e-02 2.95464247e-01
1.40778542e-01 4.38515723e-01 1.56676605e-01 3.00115019e-01
-9.32170570e-01 6.93024755e-01 1.18636346e+00 -1.33700639e-01
-9.44723561e-03 3.56180519e-01 -2.46193767e-01 1.17225194e+00
5.14533699e-01 -4.43122566e-01 -7.37959802e-01 -1.10201132e+00
1.26643673e-01 3.43183786e-01 6.06050253e-01 -8.75583768e-01
2.93215036e-01 2.23525483e-02 6.97085559e-01 -7.47486830e-01
9.01217461e-01 -1.17837369e+00 -6.60498664e-02 5.00989854e-02
-4.02895451e-01 5.01187257e-02 -9.03980732e-02 8.94570887e-01
-1.89529315e-01 -1.30243361e-01 8.23382556e-01 -2.81404078e-01
-5.48374236e-01 4.20230389e-01 6.44139424e-02 1.36684477e-01
1.05242348e+00 2.48600282e-02 -1.01986550e-01 -4.74445343e-01
-8.00013542e-01 1.17114902e-01 1.06680655e+00 3.61920446e-01
7.72274613e-01 -9.91183460e-01 -3.79641056e-01 3.63552868e-01
2.55984813e-01 6.93251252e-01 2.46747926e-01 6.75898552e-01
-6.20199621e-01 2.60601312e-01 -2.37911940e-01 -1.07684112e+00
-1.39821780e+00 6.45214081e-01 4.36891407e-01 -1.38586417e-01
-6.38436675e-01 9.41411316e-01 6.89539671e-01 -7.82864571e-01
2.48877153e-01 -5.94027825e-02 -1.03605248e-03 -4.20544893e-01
3.22418272e-01 -1.74823597e-01 2.28018925e-01 -7.41452754e-01
-1.33406475e-01 6.85634732e-01 -2.78641492e-01 -2.35731319e-01
1.58649588e+00 -2.35481486e-01 4.06437404e-02 6.58197284e-01
1.51747382e+00 -3.02062958e-01 -1.94195747e+00 -5.86795628e-01
-3.45807940e-01 -7.27452755e-01 1.46470875e-01 -6.74546361e-01
-1.13276029e+00 1.23583496e+00 4.07524318e-01 -2.53737181e-01
7.88611948e-01 1.89168826e-01 3.87689918e-01 7.62649596e-01
3.28592211e-01 -1.02396274e+00 3.25883836e-01 3.97413880e-01
7.46668935e-01 -1.85244393e+00 1.76878437e-01 -4.05648172e-01
-7.32155561e-01 9.85690475e-01 8.15479755e-01 -3.91951025e-01
6.91280782e-01 -7.17028901e-02 1.82169303e-01 -2.61323124e-01
-9.21846330e-01 1.86901707e-02 2.83696204e-01 5.85239112e-01
-1.30245453e-02 -1.37080044e-01 6.91289485e-01 -6.62125796e-02
-5.17056556e-04 -3.08543295e-01 5.74249864e-01 9.67967689e-01
-2.49814942e-01 -1.08294094e+00 -7.93728679e-02 1.84481040e-01
-1.77098438e-01 2.23989591e-01 -6.35291398e-01 5.22083879e-01
1.98881757e-02 7.81147242e-01 1.53564900e-01 -9.36178565e-02
2.14696243e-01 -4.83165458e-02 9.25713956e-01 -9.14652109e-01
1.02116525e-01 1.67665377e-01 -1.59813967e-02 -4.84852374e-01
-6.66373312e-01 -8.08356762e-01 -8.43353450e-01 2.94647157e-01
-2.63286293e-01 -6.75780326e-02 9.51540530e-01 1.24071753e+00
2.35896274e-01 3.67459983e-01 9.97435153e-01 -1.30936396e+00
-5.70918441e-01 -6.92268670e-01 -4.97427136e-01 4.79383409e-01
4.96476531e-01 -8.56884122e-01 -5.94935060e-01 5.07582068e-01] | [8.346981048583984, -2.74723219871521] |
3907b556-0e02-4b1b-8e66-6eeb8ac303fd | towards-understanding-iterative-magnitude | 2106.06955 | null | https://arxiv.org/abs/2106.06955v1 | https://arxiv.org/pdf/2106.06955v1.pdf | Towards Understanding Iterative Magnitude Pruning: Why Lottery Tickets Win | The lottery ticket hypothesis states that sparse subnetworks exist in randomly initialized dense networks that can be trained to the same accuracy as the dense network they reside in. However, the subsequent work has failed to replicate this on large-scale models and required rewinding to an early stable state instead of initialization. We show that by using a training method that is stable with respect to linear mode connectivity, large networks can also be entirely rewound to initialization. Our subsequent experiments on common vision tasks give strong credence to the hypothesis in Evci et al. (2020b) that lottery tickets simply retrain to the same regions (although not necessarily to the same basin). These results imply that existing lottery tickets could not have been found without the preceding dense training by iterative magnitude pruning, raising doubts about the use of the lottery ticket hypothesis. | ['Marie-Francine Moens', 'Mingxiao Li', 'Jaron Maene'] | 2021-06-13 | null | null | null | null | ['linear-mode-connectivity'] | ['knowledge-base'] | [ 1.11690767e-01 8.49297047e-01 -6.99502975e-02 -3.96467835e-01
2.17173532e-01 -5.02766013e-01 8.21771979e-01 -4.94391888e-01
-6.09122872e-01 1.31687391e+00 1.13566287e-01 -4.03474420e-01
-3.31220269e-01 -1.01784921e+00 -8.88536155e-01 -5.80147266e-01
-3.16391200e-01 9.81330156e-01 7.34241664e-01 -1.45992175e-01
1.40937254e-01 7.19796777e-01 -1.27530468e+00 1.30928770e-01
3.11114222e-01 1.28261000e-01 1.39827445e-01 4.50212955e-01
2.00315341e-01 7.41228998e-01 -5.94370425e-01 -1.78021520e-01
8.28380823e-01 -5.00805140e-01 -1.02143693e+00 2.14905925e-02
8.91685247e-01 -3.63603979e-01 -5.89383125e-01 1.20969117e+00
-4.05406281e-02 2.19505653e-01 4.47956026e-01 -9.43892002e-01
-1.60478026e-01 1.20465040e+00 -5.49384236e-01 5.03908992e-01
-4.29009646e-01 3.33628058e-01 8.43652546e-01 -7.28565693e-01
8.43505085e-01 1.24869537e+00 1.15551996e+00 5.20541012e-01
-1.47105300e+00 -9.76660371e-01 4.87571329e-01 -2.88807690e-01
-1.56509650e+00 -7.78893232e-01 2.41082340e-01 -2.77100980e-01
1.32588184e+00 2.12862659e-02 1.24916279e+00 8.25071514e-01
1.55779690e-01 1.83939800e-01 1.14892995e+00 -2.37594098e-01
1.92191169e-01 1.33684590e-01 8.82748608e-03 9.26965117e-01
1.03039837e+00 3.48708481e-01 -3.33363980e-01 4.52964380e-03
1.37897050e+00 -2.95700461e-01 -1.46496803e-01 -4.27386016e-01
-1.10171258e+00 1.02308011e+00 9.40061629e-01 5.05373120e-01
-2.95391351e-01 2.12213397e-01 8.26083794e-02 2.72808164e-01
1.95135191e-01 9.37695801e-01 -4.44395870e-01 3.61675113e-01
-1.42952192e+00 -8.09462890e-02 1.11617851e+00 7.47603714e-01
1.06954730e+00 2.75217146e-01 5.89208007e-01 2.73753256e-01
3.88382047e-01 1.13445066e-01 3.61678302e-01 -1.14045739e+00
4.18152541e-01 4.07185167e-01 -3.01457286e-01 -8.55589986e-01
-3.90228838e-01 -7.53597319e-01 -1.14882505e+00 5.25914371e-01
5.91136098e-01 -5.61312437e-01 -1.10516214e+00 1.69129717e+00
-3.90214473e-02 5.66200376e-01 4.87049632e-02 5.51897943e-01
5.30133367e-01 4.63954151e-01 -2.16351211e-01 2.14100227e-01
7.50283837e-01 -5.29331326e-01 6.44670874e-02 -6.25639737e-01
3.95396173e-01 -3.40785176e-01 6.71332479e-01 3.88194591e-01
-1.02215731e+00 -4.84390259e-01 -1.49324620e+00 4.88709480e-01
-2.97218800e-01 -4.75262552e-01 1.17047155e+00 8.35892498e-01
-1.56270385e+00 9.46178734e-01 -1.03750563e+00 -7.05007911e-01
6.03302777e-01 7.55311251e-01 -4.59514678e-01 2.51495615e-02
-1.01164925e+00 1.15678370e+00 8.14768851e-01 1.97056562e-01
-1.10418892e+00 -3.68266523e-01 -6.49819851e-01 -1.09082935e-02
1.31660149e-01 -9.36312854e-01 7.88380027e-01 -1.17942560e+00
-1.25575542e+00 8.57696950e-01 -1.65266357e-02 -1.17687857e+00
1.57503590e-01 2.85954088e-01 6.43771216e-02 2.61810333e-01
2.15771571e-01 1.21291077e+00 8.22953880e-01 -1.23335040e+00
-5.95488548e-01 -2.32369661e-01 2.47496024e-01 3.08530360e-01
5.36466436e-03 -4.32901740e-01 -1.91606507e-01 -2.76225984e-01
3.94042462e-01 -8.62574339e-01 -4.08709168e-01 -1.79662988e-01
-5.05720139e-01 1.20557480e-01 4.79249418e-01 1.18988985e-03
7.02869117e-01 -1.77435064e+00 -1.92164585e-01 7.52810657e-01
7.76772082e-01 5.75113073e-02 -1.12085462e-01 -2.02173710e-01
-3.07345867e-01 4.37337697e-01 -1.95363373e-01 -6.61792830e-02
-2.85035282e-01 4.75887716e-01 -3.27856302e-01 8.14091921e-01
-1.50187612e-02 8.73663008e-01 -6.78177416e-01 -4.91837144e-01
1.40805572e-01 -8.09412003e-02 -6.11812770e-01 -5.26532292e-01
1.45980582e-01 -1.74305350e-01 7.45333498e-03 1.82664797e-01
5.69378316e-01 -4.09818202e-01 3.78849864e-01 -4.23536897e-02
-1.75431922e-01 3.32229316e-01 -1.02392983e+00 1.11917078e+00
-1.10078633e-01 9.48830545e-01 4.18414652e-01 -1.03847682e+00
6.26892626e-01 1.31316349e-01 2.79744506e-01 -3.31966609e-01
1.41283303e-01 1.58085048e-01 6.34144366e-01 3.01562101e-01
2.84693837e-01 -6.34525061e-01 2.33894244e-01 4.52583998e-01
3.82310063e-01 -2.69926608e-01 1.54751107e-01 4.20046002e-01
1.37818730e+00 -3.15097004e-01 -4.84281965e-03 -6.74233913e-01
-1.72301561e-01 3.33153367e-01 4.14093435e-01 1.35485232e+00
-1.72049746e-01 6.76516473e-01 4.24388170e-01 -5.08578837e-01
-1.39786410e+00 -1.07855392e+00 -3.97009909e-01 6.32294834e-01
9.99125019e-02 -1.12791173e-03 -6.45802438e-01 -6.59928322e-01
-4.70745042e-02 3.21036279e-01 -8.73081982e-01 -1.19205840e-01
-6.03408456e-01 -9.52577233e-01 8.96005452e-01 3.69766563e-01
7.84963250e-01 -1.10063434e+00 -3.31200659e-01 3.28921348e-01
3.85724366e-01 -7.06511617e-01 9.59599912e-02 8.73199284e-01
-1.43520439e+00 -1.12564516e+00 -7.65798211e-01 -1.02609015e+00
1.09852743e+00 2.35008851e-01 1.34165144e+00 3.71444106e-01
-4.82369103e-02 3.59032229e-02 1.47817373e-01 -9.50380266e-02
-2.88606256e-01 6.49975002e-01 8.62289444e-02 -7.07632840e-01
4.55561876e-01 -8.54196489e-01 -4.55485344e-01 2.77244300e-01
-5.28298259e-01 -4.41898778e-02 8.44637036e-01 9.41148460e-01
3.04548204e-01 4.50160712e-01 7.26808250e-01 -1.21868694e+00
4.86852258e-01 -5.98890603e-01 -7.15885878e-01 -1.58531755e-01
-5.96024692e-01 2.55079806e-01 5.02420545e-01 -3.41499060e-01
-7.97156751e-01 1.81469560e-01 -2.40377441e-01 -3.13643426e-01
-3.29591006e-01 2.83327639e-01 9.74415541e-02 -4.16330099e-01
7.90868819e-01 -5.64203598e-02 -3.38945724e-02 -1.32509604e-01
3.49739164e-01 -1.97130367e-01 4.48517144e-01 -3.07573855e-01
1.09923375e+00 7.02908039e-01 -1.01602435e-01 -8.71820271e-01
-8.43173385e-01 -3.26066375e-01 -8.99118006e-01 5.88831864e-02
6.64004803e-01 -9.54463661e-01 -2.42261261e-01 1.21261738e-01
-1.01935792e+00 -6.32053614e-01 -7.58192003e-01 4.87181753e-01
-3.11969936e-01 1.49150908e-01 -5.35595953e-01 -1.79254800e-01
1.60991568e-02 -6.20060742e-01 2.81750858e-01 2.12916240e-01
-5.04061043e-01 -1.13394439e+00 1.13487534e-01 -1.09669104e-01
2.87535161e-01 -2.47173652e-01 7.51844108e-01 -5.49223006e-01
-5.32284021e-01 -3.67264338e-02 -3.60709429e-01 1.73701167e-01
3.20216306e-02 -2.13658251e-02 -8.90095353e-01 -4.12564576e-01
2.31484007e-02 -3.46613437e-01 1.47298586e+00 6.56420410e-01
5.10861456e-01 -4.42506254e-01 -5.84114075e-01 7.08019137e-01
1.50070155e+00 -6.31663874e-02 8.99586916e-01 6.56588018e-01
4.79340404e-01 1.96672082e-01 -1.74164519e-01 -2.32838839e-02
-1.67185992e-01 -6.28855526e-02 6.15521669e-01 -4.90389258e-01
-2.93454915e-01 -3.52510601e-01 1.89676598e-01 3.49843234e-01
-6.96877018e-02 -1.01407193e-01 -7.50321865e-01 9.24463212e-01
-1.59754372e+00 -1.46890581e+00 1.77400950e-02 2.10722280e+00
1.08194816e+00 1.06891167e+00 -5.83448671e-02 -2.05567047e-01
8.00235331e-01 2.23196298e-01 -6.79648399e-01 -1.85713425e-01
-2.36219227e-01 3.88518423e-01 1.00086772e+00 7.53737807e-01
-1.06973970e+00 1.32424974e+00 7.79862547e+00 6.60051405e-01
-9.05213416e-01 3.64088118e-02 7.12499022e-01 -2.20108077e-01
-2.70839155e-01 3.12128663e-01 -1.35629523e+00 7.03475997e-02
9.17846620e-01 3.69187184e-02 3.59042257e-01 6.96786106e-01
-6.09461330e-02 -2.58320928e-01 -8.69534612e-01 3.60581309e-01
-1.66775852e-01 -1.68565023e+00 -3.17897834e-02 1.99478611e-01
9.82014596e-01 7.12020874e-01 -1.21144980e-01 4.23828632e-01
1.19285154e+00 -1.31502473e+00 6.14505589e-01 2.78547227e-01
6.89408302e-01 -6.17061377e-01 7.62899995e-01 5.61510801e-01
-1.17829216e+00 2.79301524e-01 -9.36468720e-01 -1.77378058e-01
5.87119833e-02 5.48097432e-01 -1.41188991e+00 -1.72756806e-01
5.02827764e-01 5.96804678e-01 -7.36204505e-01 1.38487232e+00
-3.08993220e-01 1.00964081e+00 -1.04070580e+00 1.06217429e-01
6.64120793e-01 1.15505140e-03 4.48273122e-01 9.81796622e-01
1.04040712e-01 -1.63239017e-01 -4.47750501e-02 9.56745327e-01
-1.31007537e-01 -5.74391007e-01 -1.09954703e+00 1.01347663e-01
5.34136713e-01 9.64087129e-01 -1.44536531e+00 -4.68770623e-01
-3.06273401e-01 6.27142787e-01 2.47980207e-01 2.33445808e-01
-5.24607062e-01 -5.21608412e-01 3.11874479e-01 5.95896423e-01
6.14755094e-01 -1.61982045e-01 -5.75483203e-01 -1.16618299e+00
-5.33371627e-01 -5.96855700e-01 5.02806418e-02 -5.19795537e-01
-1.30154586e+00 8.54908884e-01 4.65076230e-02 -7.66355038e-01
-2.48905540e-01 -4.52478766e-01 -9.28684533e-01 7.88345933e-01
-1.06514704e+00 -6.50144219e-01 9.67567489e-02 8.51230919e-01
4.30312246e-01 -2.59212315e-01 6.29043400e-01 -1.26843611e-02
-4.44172174e-01 2.91944087e-01 9.49102119e-02 3.72794092e-01
2.63128012e-01 -1.15686500e+00 6.09570444e-01 1.06700754e+00
5.11002600e-01 1.07900393e+00 5.91248393e-01 -9.58839297e-01
-6.24131858e-01 -9.93747711e-01 6.14181519e-01 -4.55611914e-01
7.38212883e-01 -3.45683038e-01 -8.12825382e-01 1.11019194e+00
5.43721199e-01 -3.34169835e-01 9.90489423e-02 3.75573695e-01
-2.75417548e-02 2.37778630e-02 -1.21631777e+00 3.98851484e-01
1.01666498e+00 -2.01719269e-01 -9.29868758e-01 1.47251531e-01
3.95328760e-01 1.12032652e-01 -4.27345574e-01 2.12323830e-01
2.68603623e-01 -8.76593411e-01 1.06262803e+00 -6.24519289e-01
-1.30995512e-01 -2.63186485e-01 1.73353404e-01 -1.10657084e+00
-5.52542031e-01 -2.68467814e-01 3.11790437e-01 8.35916340e-01
6.97324455e-01 -8.02377820e-01 1.31724095e+00 3.95716488e-01
-4.68517169e-02 -3.94349337e-01 -1.27392316e+00 -9.39402819e-01
3.49044204e-01 -4.08564508e-02 4.03338134e-01 9.85621095e-01
-2.17743516e-01 7.38202155e-01 -2.83065010e-02 1.96042493e-01
8.20336521e-01 3.19356956e-02 4.25630152e-01 -1.72640038e+00
-2.64198929e-01 -6.70345724e-01 -3.21493298e-01 -1.19950020e+00
2.26087600e-01 -9.99384880e-01 2.85727739e-01 -1.72590744e+00
1.46983296e-01 -8.42879295e-01 1.75584629e-02 8.80882204e-01
3.59567314e-01 6.89383924e-01 -3.44245993e-02 3.87068778e-01
-5.02616644e-01 1.81904688e-01 1.01345837e+00 -2.31516600e-01
-1.18863367e-01 1.68675765e-01 -1.02902889e+00 1.28509212e+00
1.05625892e+00 -7.53118336e-01 -5.59560776e-01 -3.50860208e-01
3.83475751e-01 -4.58649457e-01 4.39351588e-01 -1.33800840e+00
2.74843186e-01 3.59381884e-02 7.77871609e-01 -4.58018482e-01
2.92208374e-01 -7.46589363e-01 7.46337175e-02 7.27747917e-01
-6.32255673e-02 -5.79655766e-02 5.74682832e-01 4.45721447e-01
6.64018616e-02 -7.02957511e-01 1.17536795e+00 -7.32237577e-01
-8.89375329e-01 2.17318878e-01 -7.86380947e-01 8.80477875e-02
8.30222726e-01 -7.43447423e-01 -4.17861253e-01 -1.79124251e-01
-1.00403786e+00 4.16829288e-02 7.31789708e-01 -2.42717952e-01
6.23812377e-01 -9.14040148e-01 -5.91146529e-01 4.49877262e-01
-8.60642612e-01 2.43711308e-01 -3.83956611e-01 5.29688895e-01
-6.92324579e-01 3.99614424e-01 -3.26278657e-01 -4.32159692e-01
-8.38507175e-01 2.18893856e-01 7.49501586e-01 -3.48491609e-01
-9.44705486e-01 1.23439360e+00 1.84361383e-01 -3.52589965e-01
2.17673659e-01 -1.87530473e-01 -1.42084928e-02 6.53522462e-02
1.39674455e-01 -3.80965285e-02 9.52284690e-03 -2.47362345e-01
-4.46444124e-01 2.81144440e-01 -3.65978986e-01 -4.22212481e-01
1.65730500e+00 -2.10913923e-03 -1.85455345e-02 4.55459133e-02
9.02255833e-01 -1.29759222e-01 -1.37969148e+00 4.35310826e-02
-1.70785427e-01 -2.34692663e-01 1.86096698e-01 -6.25358462e-01
-1.33622849e+00 7.70165980e-01 2.66147405e-01 4.05505806e-01
6.50506377e-01 2.89785802e-01 3.93194526e-01 8.94727468e-01
5.88330269e-01 -9.05530870e-01 -1.68638244e-01 9.65059936e-01
7.56719351e-01 -9.87816513e-01 3.40380609e-01 3.56816165e-02
-2.77599156e-01 9.26621139e-01 9.45547104e-01 -7.71618187e-01
8.77489150e-01 5.14301896e-01 -2.69552380e-01 -4.50775445e-01
-8.13663244e-01 -2.16244146e-01 -1.94595739e-01 8.37063491e-01
3.07743344e-02 -2.98371851e-01 3.32343191e-01 6.58981726e-02
-5.72506845e-01 -1.78090051e-01 6.65973067e-01 7.69911945e-01
-1.11342227e+00 -5.60725808e-01 -3.50709409e-01 9.19880807e-01
5.31104319e-02 -5.19180954e-01 -6.44368649e-01 1.16939962e+00
2.90885895e-01 5.19839585e-01 4.29465950e-01 -2.33061254e-01
-4.32954810e-04 -8.32924470e-02 7.04059839e-01 -1.02965903e+00
-6.71696484e-01 1.03275225e-01 1.66232333e-01 -2.48248860e-01
-2.56597072e-01 -4.83631492e-01 -1.26105261e+00 -6.97301865e-01
-6.59738064e-01 3.12466621e-01 2.72783283e-02 8.21056843e-01
2.07785740e-02 3.77550691e-01 1.99164093e-01 -1.05602837e+00
-3.92515987e-01 -1.08287716e+00 -8.02316666e-01 -3.23047817e-01
1.53961495e-01 -4.19719517e-01 -7.38880515e-01 -1.14288323e-01] | [8.540655136108398, 3.308889389038086] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.