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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
f0488831-1ac4-4f13-b424-d4fcba8a9ed5 | chance-constrained-trajectory-optimization | 2005.04374 | null | https://arxiv.org/abs/2005.04374v3 | https://arxiv.org/pdf/2005.04374v3.pdf | Chance-Constrained Trajectory Optimization for Safe Exploration and Learning of Nonlinear Systems | Learning-based control algorithms require data collection with abundant supervision for training. Safe exploration algorithms ensure the safety of this data collection process even when only partial knowledge is available. We present a new approach for optimal motion planning with safe exploration that integrates chance-constrained stochastic optimal control with dynamics learning and feedback control. We derive an iterative convex optimization algorithm that solves an \underline{Info}rmation-cost \underline{S}tochastic \underline{N}onlinear \underline{O}ptimal \underline{C}ontrol problem (Info-SNOC). The optimization objective encodes control cost for performance and exploration cost for learning, and the safety is incorporated as distributionally robust chance constraints. The dynamics are predicted from a robust regression model that is learned from data. The Info-SNOC algorithm is used to compute a sub-optimal pool of safe motion plans that aid in exploration for learning unknown residual dynamics under safety constraints. A stable feedback controller is used to execute the motion plan and collect data for model learning. We prove the safety of rollout from our exploration method and reduction in uncertainty over epochs, thereby guaranteeing the consistency of our learning method. We validate the effectiveness of Info-SNOC by designing and implementing a pool of safe trajectories for a planar robot. We demonstrate that our approach has higher success rate in ensuring safety when compared to a deterministic trajectory optimization approach. | ['Soon-Jo Chung', 'Yisong Yue', 'Anima Anandkumar', 'Anqi Liu', 'Yashwanth Kumar Nakka', 'Guanya Shi'] | 2020-05-09 | null | null | null | null | ['optimal-motion-planning', 'safe-exploration'] | ['robots', 'robots'] | [ 4.84659262e-02 8.16603303e-01 -8.20884764e-01 1.31940812e-01
-1.19174874e+00 -5.89782476e-01 3.74899656e-01 1.70270249e-01
-7.90819466e-01 1.15946603e+00 -1.24557897e-01 -5.18317759e-01
-8.53774071e-01 -3.95099849e-01 -1.26741171e+00 -9.73418653e-01
-6.40711010e-01 4.56020057e-01 -8.80234167e-02 -5.60247293e-03
1.98184982e-01 4.24773753e-01 -1.36069775e+00 -4.59048450e-01
1.10934961e+00 1.20635033e+00 5.23873568e-01 6.41088903e-01
6.63498044e-01 6.12977326e-01 -2.56481264e-02 4.21052039e-01
6.18905962e-01 -3.34387720e-02 -7.04443336e-01 1.28220424e-01
-1.59819618e-01 -1.46556661e-01 -2.65841305e-01 1.07105410e+00
2.94522166e-01 7.51691818e-01 5.41188300e-01 -1.54113865e+00
5.88646457e-02 4.00767654e-01 -1.42257899e-01 -2.08840728e-01
-2.58266442e-02 5.40932596e-01 8.65960181e-01 -6.18327379e-01
7.66129196e-01 1.00128222e+00 2.06318766e-01 7.62689948e-01
-1.13248122e+00 -3.01211774e-01 5.24120331e-01 -4.68130037e-02
-1.31651688e+00 -1.12746075e-01 2.65771508e-01 -4.87485915e-01
7.92470992e-01 3.66574645e-01 7.53907859e-01 8.38267267e-01
7.83541083e-01 9.03444409e-01 8.20917487e-01 -1.91529170e-01
7.80701518e-01 1.93716869e-01 -1.91924170e-01 9.04457808e-01
3.28630626e-01 1.02148330e+00 -3.92931759e-01 -2.17753481e-02
6.04898214e-01 -2.59623587e-01 -2.61766970e-01 -8.33350062e-01
-1.07930017e+00 9.47715282e-01 3.35744858e-01 -2.83921540e-01
-3.71883065e-01 3.80145520e-01 2.03914836e-01 4.96649206e-01
-3.04815453e-02 8.66202414e-01 -6.30085945e-01 -1.20186463e-01
-5.66409349e-01 6.96212232e-01 7.60123014e-01 1.24258852e+00
4.48279649e-01 4.39391285e-01 -2.74658334e-02 1.97633058e-01
3.38070095e-01 7.53501058e-01 1.40207708e-01 -1.56747079e+00
7.51296222e-01 3.60494167e-01 8.29931378e-01 -5.24223149e-01
-6.15366876e-01 -2.97290623e-01 -5.47747970e-01 9.48416948e-01
3.58138412e-01 -6.33686483e-01 -6.90108538e-01 1.97433412e+00
4.80065137e-01 -6.11273833e-02 2.59015441e-01 9.68419552e-01
-4.96614248e-01 6.95197523e-01 -2.10803717e-01 -6.70091331e-01
4.37987328e-01 -6.61025047e-01 -6.96991622e-01 -9.39058438e-02
8.54218423e-01 3.08384866e-01 1.26058531e+00 8.51832569e-01
-1.09302318e+00 -8.41599926e-02 -1.31875014e+00 4.71198887e-01
8.73731896e-02 -4.22140136e-02 3.46053317e-02 3.33902717e-01
-7.99507320e-01 1.08716524e+00 -1.20811820e+00 8.65390226e-02
2.49454156e-01 7.82959580e-01 -2.26588905e-01 4.52205569e-01
-8.19959402e-01 1.23433471e+00 8.85528803e-01 3.85446809e-02
-1.74455583e+00 -6.46787763e-01 -9.27895069e-01 -5.02020538e-01
1.11199367e+00 -2.40094215e-01 1.14511061e+00 -4.42713797e-01
-1.65714574e+00 3.43367159e-02 2.67314315e-01 -8.47671866e-01
7.00139940e-01 -6.80756330e-01 1.32586315e-01 -6.50409088e-02
1.02697439e-01 5.95155239e-01 9.30912018e-01 -1.26710773e+00
-8.39857757e-01 -1.65280551e-01 -1.15340225e-01 3.84276569e-01
-3.49998325e-01 -4.39674735e-01 -1.20398872e-01 -4.80336547e-01
-1.94259167e-01 -1.37373197e+00 -7.24519014e-01 1.57553643e-01
-4.91374463e-01 -2.62734219e-02 7.66681254e-01 -5.06287873e-01
1.44742787e+00 -1.64642286e+00 7.50856280e-01 5.73806465e-01
-3.20901692e-01 -1.43754363e-01 -5.51675446e-02 2.92472363e-01
3.17892730e-01 9.53518450e-02 -5.53285718e-01 -2.10550278e-01
-5.90081364e-02 5.09520710e-01 -7.61722624e-01 7.03743339e-01
-2.68683359e-02 4.87714320e-01 -8.50230634e-01 -2.65483499e-01
1.89238846e-01 -1.82177708e-01 -6.13739789e-01 3.57876092e-01
-8.37979496e-01 6.72551036e-01 -6.91024065e-01 3.86840522e-01
-7.93154612e-02 2.06039160e-01 -4.85765524e-02 7.00382054e-01
-4.26871449e-01 -1.91962242e-01 -1.48505449e+00 1.53984606e+00
-5.54778516e-01 1.75475374e-01 6.10549629e-01 -8.18514824e-01
7.49076366e-01 1.12966128e-01 6.87197089e-01 -1.52076319e-01
2.27955431e-01 4.39429760e-01 -2.88708866e-01 -4.89155710e-01
3.94767672e-01 -4.13842015e-02 -3.27250212e-01 2.61240095e-01
-1.68870002e-01 -5.44741631e-01 -2.32915059e-01 -1.60079718e-01
9.94988620e-01 4.40813929e-01 1.44773468e-01 -6.02730513e-01
5.47795892e-01 3.61838073e-01 7.78473675e-01 7.11011767e-01
-1.21639840e-01 -1.11583434e-01 6.21114850e-01 -1.20467022e-01
-1.01307929e+00 -8.60074520e-01 6.65611178e-02 7.31852412e-01
3.23908627e-01 -4.54861075e-02 -6.43668115e-01 -7.35824943e-01
1.76976800e-01 1.21209300e+00 -7.97763646e-01 -3.69770706e-01
-7.89133310e-01 -2.71427989e-01 8.25835466e-02 4.43578660e-01
-7.63986409e-02 -8.07402730e-01 -1.24889708e+00 1.19481370e-01
3.99492592e-01 -5.54984331e-01 -7.26080179e-01 5.37554622e-01
-9.25594509e-01 -1.12970996e+00 -1.54092729e-01 -3.62563044e-01
7.03316092e-01 -5.52938819e-01 2.30739191e-01 -3.60344976e-01
-1.13750190e-01 4.27506924e-01 1.50537983e-01 -7.41398335e-01
-3.57063711e-01 -8.32378715e-02 6.55371785e-01 -2.30796978e-01
-6.57878757e-01 -2.29283974e-01 -2.42541179e-01 3.84366989e-01
-4.70997304e-01 -1.90729976e-01 3.83291990e-01 9.02062178e-01
1.08064282e+00 1.95875466e-01 4.94411767e-01 -2.21591815e-01
5.21975994e-01 -4.96160001e-01 -1.41647303e+00 6.69873208e-02
-1.03651273e+00 6.34555399e-01 7.46670127e-01 -6.28391206e-01
-7.78394103e-01 4.60238367e-01 3.93798232e-01 -7.90618062e-01
2.93651938e-01 4.73840147e-01 -3.20883691e-01 -9.09093674e-03
6.62955582e-01 3.23190987e-02 3.51785153e-01 -1.51191443e-01
4.20181930e-01 3.15261930e-01 5.42383552e-01 -9.41196382e-01
7.78525293e-01 4.22541946e-01 4.32442188e-01 -6.54145539e-01
-7.83993244e-01 -1.71738312e-01 -5.73159218e-01 -3.48660022e-01
8.11557591e-01 -5.89334846e-01 -1.05154347e+00 6.29687821e-03
-7.18847036e-01 -9.92416263e-01 -7.53410518e-01 6.35749578e-01
-1.43782449e+00 -5.44414259e-02 2.23077033e-02 -1.52645385e+00
-6.63024932e-02 -1.04683208e+00 7.63832688e-01 -4.44006622e-02
-1.98657036e-01 -9.09888625e-01 1.56910807e-01 -1.42774984e-01
-1.33352086e-01 7.67315626e-01 6.18350267e-01 -4.24917668e-01
-8.19594443e-01 -1.29825398e-01 5.40044188e-01 1.80756286e-01
-3.88303429e-01 -4.09251541e-01 -4.31262285e-01 -7.37635851e-01
3.13176960e-01 -5.57589591e-01 6.55385971e-01 5.13971269e-01
1.14600456e+00 -1.13425243e+00 -5.58435321e-01 5.62492847e-01
1.55664194e+00 5.60107648e-01 1.22430408e-02 6.46789789e-01
4.59338635e-01 9.42047119e-01 1.44673193e+00 6.27498150e-01
1.28624275e-01 6.46315575e-01 1.01013076e+00 7.17630804e-01
7.22107530e-01 -4.41899806e-01 8.11638653e-01 8.91338289e-02
1.47153094e-01 3.99754271e-02 -9.20709014e-01 8.82348895e-01
-2.32291794e+00 -8.38178277e-01 2.18658641e-01 2.86966491e+00
1.04265308e+00 1.86719745e-01 2.05849007e-01 -6.08456805e-02
3.32682848e-01 -3.04803193e-01 -1.14542592e+00 -4.79835033e-01
2.58817554e-01 -2.59424120e-01 1.03456378e+00 1.15852201e+00
-1.08057868e+00 5.37814558e-01 5.26237440e+00 8.72897327e-01
-8.36355329e-01 -8.77921060e-02 3.65009069e-01 -5.63358366e-01
-2.11109862e-01 -1.14051839e-02 -1.00877130e+00 4.25356239e-01
1.17573094e+00 -4.28307235e-01 6.50321722e-01 1.31786466e+00
7.06929862e-01 -3.59560102e-01 -1.23555553e+00 4.18129593e-01
-4.21221614e-01 -1.21993232e+00 -5.06909251e-01 2.43810162e-01
9.73607838e-01 -5.77914976e-02 1.73859328e-01 1.69379041e-01
6.83417261e-01 -1.18768203e+00 1.25183392e+00 6.49664879e-01
7.22193599e-01 -1.19013870e+00 2.70904154e-01 1.01973236e+00
-1.01479292e+00 -8.45870674e-01 2.08021374e-03 1.09284416e-01
4.42432731e-01 2.35852882e-01 -6.38140857e-01 5.77534616e-01
3.48107666e-01 4.28757101e-01 2.28231847e-01 9.16563988e-01
-4.49498743e-01 1.98729798e-01 -5.32327831e-01 -3.79480600e-01
4.84226137e-01 -3.18873107e-01 1.17640233e+00 6.34296894e-01
2.98442572e-01 5.31235768e-04 6.98997080e-01 8.74178231e-01
5.76611102e-01 -2.43260890e-01 -8.40658009e-01 8.66767839e-02
5.56550324e-01 6.84726477e-01 -2.80633420e-01 4.09030318e-02
3.31438839e-01 4.67828393e-01 2.77351171e-01 1.85984105e-01
-7.32892632e-01 -4.57850367e-01 5.87429106e-01 -2.38618791e-01
1.13181747e-01 -5.43293774e-01 -6.02024436e-01 -5.92220902e-01
4.19383198e-02 -5.80801904e-01 5.39593160e-01 -1.81546330e-01
-6.35931492e-01 2.24445745e-01 4.72683936e-01 -1.19332945e+00
-6.07597351e-01 -6.19883418e-01 -4.76619154e-01 6.14793241e-01
-1.08545995e+00 -4.95353431e-01 3.28661233e-01 5.27919412e-01
6.44631028e-01 -2.70828366e-01 4.31665421e-01 -4.94869202e-01
-7.65331984e-01 1.93242326e-01 2.78360426e-01 -6.56314135e-01
1.01602435e-01 -1.51712477e+00 -3.20695370e-01 8.28256547e-01
-5.35340130e-01 3.22726458e-01 1.08020258e+00 -1.03041673e+00
-1.78650773e+00 -1.60653389e+00 3.15864503e-01 -6.30924582e-01
9.66641724e-01 -1.72926143e-01 -6.72968507e-01 6.64685547e-01
-2.61752963e-01 -1.32027879e-01 -1.99715123e-01 -4.75379795e-01
3.03556263e-01 4.28065993e-02 -1.14135396e+00 7.87046075e-01
7.59347677e-01 -1.64073997e-03 -4.22190100e-01 2.89144933e-01
1.19157183e+00 -5.80979764e-01 -9.28657174e-01 4.93163377e-01
3.20958167e-01 -1.11236475e-01 5.79351366e-01 -7.34270632e-01
-5.33024482e-02 -3.26532334e-01 -1.07518144e-01 -1.17656279e+00
4.18401062e-01 -1.47981632e+00 -7.41501451e-01 4.53193188e-01
6.68506622e-01 -4.33442056e-01 9.52961206e-01 8.22794378e-01
-4.18199509e-01 -1.30160320e+00 -1.44477189e+00 -1.41209161e+00
5.23347437e-01 -6.03163481e-01 3.44943404e-02 4.29956943e-01
3.50167066e-01 -3.28775167e-01 -6.51762307e-01 4.46241587e-01
9.59442556e-01 -1.24697842e-01 5.33982098e-01 -6.91697359e-01
-4.21184808e-01 -2.49269232e-01 3.27067316e-01 -6.21519208e-01
3.89376283e-01 -6.65768683e-01 6.86614215e-01 -1.22892511e+00
-3.01542163e-01 -5.98723590e-01 -2.37864163e-02 5.18313944e-01
1.67888165e-01 -8.94783735e-01 2.30163679e-01 2.74770539e-02
-6.30210638e-01 9.25341845e-01 1.14564013e+00 7.98473060e-02
-9.22805011e-01 2.94427484e-01 -2.10819066e-01 7.22771287e-01
7.71951675e-01 -3.46762121e-01 -9.73683715e-01 -1.33792460e-01
1.34205922e-01 5.80400527e-01 2.39295110e-01 -9.56929684e-01
2.44867504e-01 -8.89207661e-01 -3.21081191e-01 -5.53825557e-01
3.04277390e-01 -9.87674534e-01 1.10736422e-01 1.08288622e+00
-7.83232987e-01 -2.91021559e-02 1.70485064e-01 1.08568203e+00
3.13660502e-01 -2.57084966e-01 8.47922802e-01 1.22847207e-01
-6.17614567e-01 5.05844355e-01 -4.71971512e-01 1.53873801e-01
1.45935059e+00 4.33646776e-02 1.39407471e-01 -3.53632897e-01
-9.67835546e-01 1.14364457e+00 2.88584322e-01 3.77845079e-01
6.27475619e-01 -1.12616479e+00 -2.78489023e-01 -7.83570558e-02
-2.66324133e-01 2.03793913e-01 -1.85847208e-01 8.76235545e-01
7.83147588e-02 7.19415009e-01 2.68894490e-02 -4.64538425e-01
-7.51304805e-01 8.43032002e-01 4.62351978e-01 -2.82514244e-01
-5.72480202e-01 6.99057877e-01 -2.82610208e-01 -5.12259364e-01
6.29825056e-01 -3.97926509e-01 1.33056164e-01 -1.70597270e-01
2.28336230e-01 8.96100581e-01 -2.31547862e-01 -5.71425185e-02
-2.00068146e-01 3.60194862e-01 2.80205935e-01 -8.17638636e-01
1.24321592e+00 -2.37932220e-01 3.45296144e-01 5.85677207e-01
8.94600630e-01 -3.89635861e-01 -2.07542372e+00 2.90868104e-01
5.27691901e-01 -2.21732289e-01 -2.71869265e-02 -8.15846562e-01
-5.21258354e-01 5.00422895e-01 4.01198477e-01 -3.10802013e-01
6.52519643e-01 -2.74417371e-01 3.79770368e-01 7.70661056e-01
9.12315726e-01 -1.69381285e+00 1.70369089e-01 6.20713353e-01
1.34562194e+00 -9.67709005e-01 -6.49174899e-02 -3.77176628e-02
-9.01369333e-01 8.69133890e-01 8.93483520e-01 -3.66637617e-01
6.86685860e-01 4.65442568e-01 -4.41455990e-01 1.69798866e-01
-1.12201929e+00 -1.11911587e-01 2.58182347e-01 4.93818045e-01
-5.68515658e-01 -4.41488512e-02 -4.46718663e-01 6.37495518e-01
-1.34771585e-01 -1.76527381e-01 4.39373732e-01 1.18118179e+00
-1.02746499e+00 -8.96837115e-01 -4.12063539e-01 2.54054576e-01
-1.42260818e-02 5.94850779e-01 -1.10160075e-02 1.06530488e+00
-6.82399347e-02 9.15510476e-01 -2.50603616e-01 -3.20499182e-01
2.86372334e-01 -9.02155489e-02 2.90596962e-01 -5.78654826e-01
-1.31629501e-02 6.86189160e-03 2.66623348e-01 -1.21581089e+00
1.83209747e-01 -8.85865629e-01 -1.78599477e+00 3.67929965e-01
-4.21949208e-01 3.16531688e-01 6.92183912e-01 9.79234040e-01
1.80642009e-01 2.26697609e-01 8.65597367e-01 -7.73277223e-01
-1.52250087e+00 -2.63296127e-01 -4.49739099e-01 -3.87318254e-01
7.54413962e-01 -7.82867849e-01 -4.16261464e-01 -2.17894539e-01] | [4.685691833496094, 2.160020112991333] |
f1a73b32-2bf7-4ad6-9fae-f0021c3431fa | voxsim-a-visual-platform-for-modeling-motion | null | null | https://aclanthology.org/C16-2012 | https://aclanthology.org/C16-2012.pdf | VoxSim: A Visual Platform for Modeling Motion Language | Much existing work in text-to-scene generation focuses on generating static scenes. By introducing a focus on motion verbs, we integrate dynamic semantics into a rich formal model of events to generate animations in real time that correlate with human conceptions of the event described. This paper presents a working system that generates these animated scenes over a test set, discussing challenges encountered and describing the solutions implemented. | ['James Pustejovsky', 'Nikhil Krishnaswamy'] | 2016-12-01 | voxsim-a-visual-platform-for-modeling-motion-1 | https://aclanthology.org/C16-2012 | https://aclanthology.org/C16-2012.pdf | coling-2016-12 | ['scene-generation'] | ['computer-vision'] | [ 3.09714317e-01 5.05248725e-01 2.24883556e-01 -3.58234018e-01
-4.05482829e-01 -6.25732541e-01 1.44090891e+00 9.36783776e-02
-8.08799416e-02 8.46908927e-01 6.55451834e-01 -1.37339428e-01
3.60194504e-01 -1.02169824e+00 -2.23964453e-01 -8.94985422e-02
-2.32753545e-01 5.61596334e-01 6.67203248e-01 -6.82855964e-01
1.23383343e-01 5.77118695e-01 -1.66677403e+00 7.34268367e-01
1.62321374e-01 -9.16425362e-02 2.58532286e-01 1.37358880e+00
-5.94047189e-01 1.40659678e+00 -1.32535577e+00 -1.87166303e-01
-1.35528252e-01 -1.10370207e+00 -1.09110498e+00 2.96078116e-01
1.11500822e-01 -1.44413188e-01 -1.58217967e-01 6.47693992e-01
4.28552687e-01 4.37133044e-01 5.61704159e-01 -1.56223822e+00
-9.22218636e-02 8.24660420e-01 4.65175090e-03 1.32642061e-01
1.46779788e+00 4.53987926e-01 4.72439140e-01 -3.07367772e-01
1.55526543e+00 1.57110083e+00 4.63378817e-01 1.07509601e+00
-1.23842728e+00 -1.32659376e-01 -8.02870467e-02 9.43267066e-03
-1.12372875e+00 -3.95921588e-01 7.92548776e-01 -3.73315543e-01
1.53429723e+00 7.01227188e-01 1.38640988e+00 1.22456694e+00
2.96734273e-01 6.03977382e-01 1.01617622e+00 -7.74696171e-01
2.44497776e-01 2.86180321e-02 -9.89069045e-02 4.19885606e-01
-1.94118306e-01 9.77463573e-02 -4.08982217e-01 -2.19662279e-01
1.11381054e+00 -6.86268508e-01 9.62746739e-02 -4.08904720e-03
-1.54641116e+00 7.56095886e-01 -5.24951257e-02 5.30334234e-01
-4.68982279e-01 6.96732581e-01 7.44064033e-01 -9.61706787e-02
8.90371501e-02 5.43856859e-01 -8.84630531e-03 -7.05174327e-01
-8.98318350e-01 1.16097856e+00 9.45338488e-01 1.12773931e+00
2.82258183e-01 5.02190590e-01 -4.70558435e-01 -8.16651061e-03
3.50718498e-01 3.57851774e-01 5.42979002e-01 -1.20290363e+00
1.22628644e-01 4.28866506e-01 4.70304221e-01 -1.05574560e+00
-3.64838153e-01 4.85099852e-01 -5.84465005e-02 5.65791488e-01
5.44257499e-02 -2.57548779e-01 -8.61421168e-01 1.44019341e+00
5.11102378e-01 2.71816015e-01 2.08503440e-01 5.60297012e-01
1.10078824e+00 1.14579165e+00 7.05990195e-01 -3.59259605e-01
1.28772712e+00 -5.94948351e-01 -1.25633788e+00 6.45502508e-02
7.72138357e-01 -8.89367759e-01 1.13033545e+00 1.27567902e-01
-1.48777544e+00 -5.66034734e-01 -1.01511204e+00 8.92324559e-03
-4.30682570e-01 -5.67763925e-01 7.91167438e-01 5.31531811e-01
-1.29298663e+00 5.65321207e-01 -8.42647314e-01 -5.91654480e-01
-1.80668578e-01 2.50899885e-02 1.64883886e-03 6.84324920e-01
-1.25413597e+00 1.02242124e+00 7.82108903e-01 -5.08735955e-01
-7.88857520e-01 -2.88613051e-01 -1.13118017e+00 -4.51524436e-01
-5.61387204e-02 -9.96505201e-01 1.92841721e+00 -1.02246809e+00
-1.81530786e+00 8.75910997e-01 -2.35140011e-01 -4.70617235e-01
5.95732272e-01 -3.94576937e-02 -6.45964921e-01 2.96603233e-01
1.55576169e-01 1.11801469e+00 3.94210637e-01 -1.57004631e+00
-5.27903795e-01 3.77176791e-01 3.52410197e-01 2.43451178e-01
3.94541413e-01 6.07930064e-01 -2.51301318e-01 -9.00928378e-01
-5.09749889e-01 -6.76557362e-01 -6.89441562e-01 -3.62717181e-01
-3.83065075e-01 -1.74230292e-01 1.17182112e+00 -2.50858307e-01
1.41085076e+00 -1.75314367e+00 7.14882910e-02 4.70772833e-02
-1.07446350e-01 1.51843891e-01 -1.38353422e-01 1.08978224e+00
-3.54735017e-01 3.09700072e-01 1.36918813e-01 -2.83655286e-01
6.80757612e-02 2.63639271e-01 -6.10383451e-01 -9.47906002e-02
1.99981481e-01 8.95520806e-01 -1.18804955e+00 -9.98379469e-01
7.50206649e-01 5.49383998e-01 -4.14528549e-01 3.74175072e-01
-7.88821757e-01 4.25815731e-01 -5.03313899e-01 1.80733979e-01
7.57804289e-02 1.68312505e-01 2.24724293e-01 1.36335328e-01
-3.42231125e-01 3.91380399e-01 -1.41443229e+00 1.83963180e+00
-3.36261809e-01 8.59327257e-01 -5.67090154e-01 -1.67219728e-01
7.17718780e-01 8.28371942e-01 3.95040065e-01 -5.44712663e-01
7.17995763e-02 -2.20217288e-01 -3.37049663e-01 -7.65409410e-01
8.33477914e-01 -4.42172378e-01 -5.85362017e-01 6.76206529e-01
-2.43604213e-01 -1.05506432e+00 6.09686852e-01 5.30382276e-01
9.08965230e-01 9.97222781e-01 8.87978435e-01 3.95607436e-03
4.62296754e-01 8.54977071e-01 1.82289332e-02 6.71084821e-01
9.98976454e-03 5.69056451e-01 3.89513165e-01 -7.02763796e-01
-1.24271548e+00 -9.97333109e-01 4.73917067e-01 7.67051697e-01
1.84648484e-01 -1.20515394e+00 -9.86611664e-01 -3.69229615e-01
-8.75156522e-01 1.28206909e+00 -5.76290548e-01 2.88881987e-01
-1.09059834e+00 -3.32781076e-01 6.74540222e-01 4.52111036e-01
1.68376788e-01 -1.89003086e+00 -1.73220491e+00 5.51945150e-01
-1.41220108e-01 -1.08931124e+00 -1.05284691e-01 -4.50403512e-01
-7.35856235e-01 -8.57627034e-01 -5.26108265e-01 -6.13008559e-01
2.83592701e-01 4.42210846e-02 1.55988026e+00 1.13051116e-01
-6.45083010e-01 5.50797820e-01 -6.45966947e-01 -8.61533284e-01
-1.20196140e+00 -5.22500336e-01 -5.90314984e-01 -7.25145340e-01
-1.18295833e-01 -2.65570253e-01 -1.72905520e-01 -1.41401496e-02
-1.25867891e+00 9.55216408e-01 -3.33798081e-01 1.39781043e-01
4.03673619e-01 -8.73138290e-03 -1.08834980e-02 -9.94708478e-01
8.92699242e-01 -1.39747426e-01 -3.17545295e-01 7.46659040e-02
2.12078661e-01 -7.21498486e-03 6.06857181e-01 -6.42397940e-01
-1.51905394e+00 2.63725519e-01 -1.75472915e-01 1.22876652e-01
-5.88760018e-01 1.87921718e-01 8.71049240e-02 4.75798279e-01
9.00015533e-01 -4.65204790e-02 -4.75702018e-01 1.73064634e-01
8.39977384e-01 -3.38017084e-02 5.93405426e-01 -5.98433077e-01
7.62294888e-01 5.84017456e-01 8.53450894e-02 -7.80573666e-01
-2.22649142e-01 3.57449837e-02 -5.29676080e-01 -8.83213580e-01
1.02646613e+00 -5.24412811e-01 -5.17239630e-01 2.85572678e-01
-1.59259868e+00 -8.73624086e-01 -1.08586872e+00 2.13813588e-01
-1.14327800e+00 1.19670749e-01 -3.83663177e-01 -8.10366035e-01
-1.34283707e-01 -9.69771028e-01 1.25710011e+00 3.48690510e-01
-1.31728816e+00 -1.17991745e+00 6.68778241e-01 -3.67953211e-01
3.38256508e-01 1.14757740e+00 7.45630562e-01 -9.74759161e-02
-3.32677722e-01 -1.58160508e-01 4.22809988e-01 -6.25133574e-01
9.72221270e-02 7.86668301e-01 -5.98700523e-01 3.75461757e-01
-3.63996029e-01 -1.92940861e-01 -2.44509399e-01 2.75686204e-01
5.23696601e-01 -2.90956378e-01 -4.94228840e-01 1.55524984e-01
1.39872062e+00 7.22638845e-01 9.63835001e-01 5.93612731e-01
5.51318586e-01 6.64444327e-01 7.13851511e-01 6.93277538e-01
4.10273373e-01 1.04108858e+00 5.24148867e-02 -1.77349284e-01
-4.73561406e-01 -6.03379548e-01 4.63864923e-01 3.93877983e-01
-1.74971759e-01 -5.11573970e-01 -9.63143051e-01 6.14408851e-01
-1.87162495e+00 -1.67804694e+00 -8.22385609e-01 1.29413855e+00
8.08873892e-01 1.38547555e-01 3.85487139e-01 1.91440806e-01
6.31260276e-01 6.00220859e-01 2.16302276e-01 -1.05460846e+00
-3.96553725e-02 5.44826269e-01 -1.92831069e-01 7.44303584e-01
-8.89122307e-01 1.31831825e+00 8.06471443e+00 6.39129758e-01
-1.04232728e+00 -3.04969680e-02 -8.31245109e-02 1.20482646e-01
-6.59449637e-01 4.01809961e-01 -5.35223782e-01 -8.61093029e-03
1.25125313e+00 -7.74226427e-01 1.43853836e-02 6.38183892e-01
8.78151953e-01 -3.14228684e-01 -1.04775345e+00 4.54351634e-01
1.04080625e-01 -1.79178762e+00 6.90702021e-01 -4.92205441e-01
6.54121399e-01 -7.95745075e-01 -5.93075573e-01 8.96512344e-02
7.63929963e-01 -1.11765635e+00 1.47130871e+00 6.17971659e-01
7.79441059e-01 -6.04887247e-01 1.38444960e-01 2.46712714e-01
-1.42007780e+00 7.28272259e-01 2.44674683e-01 -3.49888057e-01
9.54240561e-01 -1.87788010e-01 -1.22335660e+00 4.61793274e-01
3.58437181e-01 3.71137559e-01 -6.06951118e-01 7.84916699e-01
-4.01533633e-01 4.06744689e-01 -9.87104699e-02 -4.03171211e-01
1.56477123e-01 1.38826221e-01 8.53725195e-01 1.92269182e+00
1.50204271e-01 2.31954321e-01 1.63806453e-01 7.18963802e-01
7.97840238e-01 1.44325569e-01 -1.06395853e+00 -2.59373579e-02
1.21721715e-01 1.00373220e+00 -1.13482821e+00 -7.79054701e-01
-1.81380585e-02 9.72218454e-01 -3.97947341e-01 1.59653693e-01
-1.21733356e+00 -5.92403650e-01 1.01887785e-01 3.98435891e-01
-2.25044161e-01 -4.05701816e-01 -7.73129910e-02 -9.09406781e-01
-4.67214316e-01 -9.26048398e-01 3.09364051e-01 -1.53278971e+00
-3.69233519e-01 8.73840153e-01 9.89835739e-01 -1.16129375e+00
-1.14689815e+00 -4.08660211e-02 -1.29659522e+00 6.12930417e-01
-5.44744372e-01 -1.28165996e+00 -4.46394414e-01 6.71689987e-01
9.62772906e-01 2.97096312e-01 1.18306172e+00 -1.12069845e-01
-5.50736263e-02 -2.11123407e-01 -1.00695741e+00 -6.57755435e-02
2.21013576e-01 -1.11390984e+00 9.65847552e-01 9.78190184e-01
2.04440549e-01 5.23588598e-01 1.51044917e+00 -8.59936476e-01
-1.02005661e+00 -8.10598135e-01 1.03771782e+00 -4.46241558e-01
5.30755997e-01 -2.37037912e-01 -6.87487960e-01 7.88898885e-01
8.44669223e-01 -4.16213512e-01 4.22860146e-01 -8.42514396e-01
1.71002924e-01 8.29279006e-01 -1.12623680e+00 1.15266263e+00
1.16679871e+00 -2.13274285e-01 -9.50889409e-01 3.50094974e-01
9.16987300e-01 -8.88236344e-01 -4.24093306e-01 1.14266030e-01
3.00262511e-01 -9.87153530e-01 9.86362219e-01 -8.93321455e-01
5.83594441e-01 -6.69997036e-01 2.85575420e-01 -1.05271208e+00
8.77636224e-02 -1.21972013e+00 6.24732189e-02 1.22744679e+00
7.10082874e-02 -8.84477943e-02 6.09926701e-01 4.79925275e-01
-9.08285901e-02 -1.06747568e-01 -6.74717307e-01 -4.35913980e-01
-3.96375269e-01 -6.83794498e-01 5.68954289e-01 8.41857553e-01
1.99187577e-01 2.82202035e-01 -1.65848106e-01 -3.14849585e-01
1.31496608e-01 -6.19223788e-02 1.11406040e+00 -8.74323368e-01
-2.23031014e-01 -4.16123182e-01 -5.92011511e-01 -3.80768716e-01
1.90829530e-01 -5.49891353e-01 1.31057963e-01 -1.95175433e+00
-1.13410078e-01 1.95288152e-01 8.43994617e-01 1.99041694e-01
-5.45583256e-02 1.26770377e-01 6.33889973e-01 -1.38677377e-02
-6.75939977e-01 1.72344670e-01 1.38353932e+00 2.29140475e-01
-4.99352962e-01 -4.68214333e-01 -2.07386300e-01 9.94554520e-01
8.17624331e-01 -6.24428928e-01 -5.94037116e-01 -9.04746875e-02
4.44173187e-01 4.99517083e-01 4.75177050e-01 -1.29444098e+00
3.64921540e-02 -7.73093998e-01 2.51980871e-01 -5.93654573e-01
4.78398412e-01 -4.77714866e-01 1.06029999e+00 7.58142233e-01
-5.29262781e-01 6.64035916e-01 5.73021114e-01 1.08440682e-01
-2.87474126e-01 -3.46414834e-01 6.32733285e-01 -4.42728013e-01
-1.11909163e+00 -3.22899163e-01 -1.10806346e+00 5.47503531e-02
1.67513847e+00 -5.07533014e-01 -1.59228370e-01 -7.48392105e-01
-9.97500658e-01 -8.51206481e-02 6.48423910e-01 4.78826463e-01
6.56172216e-01 -1.38579798e+00 -6.71379924e-01 -1.47661254e-01
-1.28653020e-01 -2.32502788e-01 -6.67246208e-02 -1.75581023e-01
-1.32050169e+00 2.46215016e-01 -4.66528058e-01 -3.67779315e-01
-1.39551783e+00 3.62033606e-01 2.28280663e-01 -2.94972092e-01
-9.46264088e-01 3.50116402e-01 -1.48240566e-01 1.49094731e-01
-2.26915434e-01 -9.89256278e-02 -5.33513665e-01 -5.52551299e-02
6.79445922e-01 3.30449998e-01 -4.82783556e-01 -9.12122548e-01
-2.54052609e-01 5.66258550e-01 4.77911860e-01 -1.13876331e+00
1.02260005e+00 1.57827307e-02 3.52913514e-02 7.86483943e-01
5.81350803e-01 1.79328278e-01 -8.82466257e-01 4.34197843e-01
1.84832826e-01 -3.07888538e-01 -6.50557876e-01 -4.62355763e-01
-3.19384724e-01 5.67252576e-01 2.79656947e-01 6.26021624e-01
8.81606519e-01 2.17088670e-01 6.75104797e-01 1.76023945e-01
5.03576398e-01 -8.27896237e-01 2.95316070e-01 4.79902208e-01
1.30860877e+00 -4.35200065e-01 1.19246013e-01 -5.34094691e-01
-1.00421727e+00 1.39446759e+00 6.83328986e-01 -3.31497043e-01
2.49767214e-01 7.25237072e-01 3.67132366e-01 -5.46293139e-01
-9.52250540e-01 -9.96554792e-02 -2.21603647e-01 1.01457632e+00
6.45278394e-01 1.05036750e-01 -6.65760577e-01 2.44004782e-02
-6.06879294e-01 2.98205912e-01 1.15641201e+00 1.50701749e+00
-3.34462285e-01 -1.23975480e+00 -5.95351815e-01 -2.10658371e-01
-4.04971451e-01 1.86620548e-01 -7.62045860e-01 1.36334884e+00
1.06844634e-01 9.77405488e-01 3.10984850e-01 -7.85161182e-02
6.95340872e-01 3.64904888e-02 6.94592893e-01 -1.12910914e+00
-1.10633290e+00 2.68991739e-02 6.31989539e-01 -7.19473779e-01
-9.75679576e-01 -6.74016774e-01 -1.93439174e+00 -2.80062407e-01
2.25469679e-01 1.67265430e-01 4.24095631e-01 5.91679335e-01
6.61659241e-02 6.70523405e-01 3.21042806e-01 -1.31291068e+00
3.18022877e-01 -6.44031167e-01 3.54013965e-02 8.36434424e-01
8.88577849e-03 -4.61911671e-02 -5.90018444e-02 8.53947461e-01] | [11.250494003295898, 0.7991821765899658] |
cd661e39-9d72-4db1-afb9-0a91803062c2 | membership-inference-attack-in-face-of-data | null | null | https://openreview.net/forum?id=z_gX7gZe2cV | https://openreview.net/pdf?id=z_gX7gZe2cV | Membership Inference Attack in Face of Data Transformations | Membership inference attacks (MIAs) on machine learning models, which try to infer whether an example is in the training dataset of a target model, are widely studied in recent years as data privacy attracts increasing attention. One unignorable problem in the traditional MIA threat model is that it assumes the attacker can obtain exactly the same example as in the training dataset. In reality, however, the attacker is more likely to collect only a transformed version of the original example. For instance, the attacker may download a down-scaled image from a website, while the smaller image has the same content as the original image used for model training. Generally, after transformations that would not affect its semantics, a transformed training member should still be treated the same as the original one regarding privacy leakage. In this paper, we propose extending the concept of MIAs into more realistic scenarios by considering data transformations and derive two MIAs for transformed examples: one follows the efficient loss-thresholding ideas, and the other tries to approximately reverse the transformations. We demonstrated the effectiveness of our attacks by extensive evaluations on multiple common data transformations and comparison with other state-of-the-art attacks. Moreover, we also studied the coverage difference between our two attacks to show their limitations and advantages. | ['Hao Chen', 'Yiwen Guo', 'Jiyu Chen'] | 2021-09-29 | null | null | null | null | ['membership-inference-attack'] | ['computer-vision'] | [ 4.55649495e-01 2.21275032e-01 -1.00076526e-01 -2.45816275e-01
-6.74035430e-01 -9.11502600e-01 3.12186301e-01 4.25226718e-01
-5.36275804e-01 7.20320225e-01 -4.65012699e-01 -4.85925466e-01
5.13418987e-02 -1.10776019e+00 -1.19585752e+00 -8.87972951e-01
6.21110201e-02 2.33039021e-01 2.23924801e-01 1.33402571e-01
3.39642465e-02 5.21773577e-01 -1.07574546e+00 3.00172538e-01
7.50618279e-01 1.08031929e+00 -2.36802444e-01 2.20693752e-01
1.53337270e-01 4.36829567e-01 -8.45452011e-01 -1.16427493e+00
7.95606613e-01 -3.39400172e-01 -6.94755614e-01 1.09274397e-02
7.81560600e-01 -3.05296898e-01 -6.22046649e-01 1.67866611e+00
3.49426046e-02 -1.56261057e-01 4.22206968e-01 -1.67058635e+00
-4.64567214e-01 7.01283634e-01 -7.25614667e-01 -1.17194779e-01
1.43493982e-02 1.03257246e-01 8.52429032e-01 -3.34353387e-01
4.85027611e-01 1.12887216e+00 2.56240785e-01 5.48624694e-01
-1.28672945e+00 -1.18876672e+00 5.31989969e-02 4.32533383e-01
-1.47924578e+00 -1.44585714e-01 5.03725111e-01 -1.99628845e-01
4.72940318e-02 8.33518684e-01 1.47293985e-01 9.55449283e-01
-1.60370916e-02 5.88834763e-01 1.34354889e+00 -3.61478120e-01
2.05834642e-01 8.41606200e-01 2.64563560e-01 4.10681814e-01
7.64971435e-01 1.69508383e-01 -1.35707542e-01 -7.45175242e-01
2.52824754e-01 1.23953380e-01 -6.36801004e-01 -6.24861419e-01
-8.10268104e-01 8.73441160e-01 4.38354254e-01 1.71466351e-01
-7.08880872e-02 -4.73283418e-02 3.82117361e-01 3.98285329e-01
6.24894723e-02 2.09859326e-01 -3.37123185e-01 3.25869858e-01
-7.44269073e-01 4.45650369e-01 8.56907845e-01 8.02982450e-01
8.68313670e-01 -2.93953657e-01 4.16732989e-02 1.80235952e-01
1.28409222e-01 3.86947840e-01 4.27744895e-01 -5.46776593e-01
8.02569926e-01 2.93539494e-01 9.94517729e-02 -1.37423396e+00
3.50124389e-01 -3.86848062e-01 -8.82556081e-01 3.12669098e-01
8.62086713e-01 -2.32427623e-02 -3.80122274e-01 2.10448551e+00
4.07456249e-01 2.60592878e-01 1.46015823e-01 6.80124342e-01
8.26483965e-02 6.86896384e-01 4.01586369e-02 -2.70085007e-01
1.42513072e+00 -5.56123495e-01 -5.14518559e-01 -2.11540144e-02
6.23099804e-01 -5.32295823e-01 8.89435649e-01 5.59588790e-01
-8.73743355e-01 -3.58528852e-01 -1.30614376e+00 1.73278898e-01
-3.78168255e-01 -3.91474441e-02 3.98509413e-01 1.20301807e+00
-4.87697721e-01 6.71720564e-01 -7.06236720e-01 -2.19536543e-01
7.54484892e-01 2.94995159e-01 -7.14651227e-01 -2.44014580e-02
-1.42558157e+00 4.89568114e-01 5.37573695e-01 -3.27912450e-01
-9.45532024e-01 -8.13651979e-01 -6.54796898e-01 2.70086527e-01
5.03318191e-01 -3.44848633e-01 9.43502486e-01 -1.02039611e+00
-8.46269369e-01 1.05792928e+00 -3.79878879e-02 -1.00877440e+00
1.03743446e+00 -1.13196559e-02 -4.36938941e-01 9.07413810e-02
-8.30636248e-02 1.84777081e-01 1.10823572e+00 -1.36731553e+00
-7.61494219e-01 -8.20715487e-01 4.10272628e-01 -1.51235327e-01
-7.63327360e-01 -1.52518630e-01 -3.18363428e-01 -7.10487366e-01
-1.23648427e-01 -9.76782143e-01 -2.22216472e-01 4.36704129e-01
-7.72182822e-01 3.77621889e-01 1.23557913e+00 -6.52414262e-01
1.39556265e+00 -2.56292725e+00 -2.50799000e-01 6.39522672e-01
2.12433428e-01 5.67453325e-01 1.75253212e-01 2.35357970e-01
-3.60706568e-01 5.00884295e-01 -5.05035281e-01 -2.87816614e-01
3.84508632e-02 4.59058993e-02 -8.85636210e-01 7.47174263e-01
-3.69904101e-01 3.51234198e-01 -5.23312807e-01 -3.36134851e-01
-2.48508044e-02 1.70469597e-01 -4.82366174e-01 1.89796656e-01
-2.24713653e-01 2.43875697e-01 -2.38967597e-01 4.46986295e-02
1.36772883e+00 -2.95788851e-02 2.75566280e-01 -3.45572889e-01
3.75377774e-01 8.20487589e-02 -1.42002177e+00 9.98418033e-01
-3.50918055e-01 5.04798710e-01 3.22591811e-02 -9.55428600e-01
6.76091254e-01 2.62793005e-01 1.11658446e-01 -3.09376389e-01
1.06855288e-01 1.14470929e-01 1.33472264e-01 -1.89283431e-01
1.60032168e-01 -9.54678729e-02 -1.06866278e-01 5.91248691e-01
-4.75648552e-01 2.93170750e-01 -1.07824437e-01 1.38207331e-01
6.93247139e-01 -2.97123879e-01 4.07242298e-01 7.69748073e-03
8.09936523e-01 -3.48263830e-01 5.63811421e-01 8.74233842e-01
-1.75281376e-01 3.90280694e-01 6.79836392e-01 -1.81102931e-01
-8.52258742e-01 -1.07326376e+00 -1.98134229e-01 4.99217063e-01
3.91363651e-01 -5.68258584e-01 -9.97108877e-01 -1.25767720e+00
2.05144346e-01 9.32429016e-01 -6.17973864e-01 -3.83368492e-01
-2.88003325e-01 -5.88521957e-01 8.76727939e-01 3.86055559e-02
9.53168988e-01 -5.57753444e-01 -3.34115684e-01 -2.31581628e-01
-1.70639217e-01 -1.03215694e+00 -6.17871881e-01 -3.01177025e-01
-6.10169768e-01 -1.32090259e+00 -2.33877376e-01 -4.99887019e-01
8.84323239e-01 3.65363359e-01 4.26330626e-01 1.13620348e-01
-8.53023678e-02 -3.30840275e-02 2.69559342e-02 -4.24062431e-01
-9.11284149e-01 -8.42370838e-02 1.38241826e-02 7.15528131e-01
3.28052610e-01 -5.38438022e-01 -3.43115032e-01 3.11014682e-01
-1.47082722e+00 -1.97342604e-01 3.82661998e-01 3.77221346e-01
4.79577184e-01 6.12322092e-01 2.96386749e-01 -1.47890270e+00
3.06992888e-01 -4.86641467e-01 -7.10342824e-01 3.81672859e-01
-5.33834994e-01 -5.99080473e-02 1.10804570e+00 -6.45300925e-01
-8.15919638e-01 -1.94115900e-02 -2.73726061e-02 -7.34966040e-01
-2.69881457e-01 1.63430929e-01 -8.42789114e-01 -1.56138107e-01
4.82946366e-01 3.77684355e-01 5.81906475e-02 -3.92152607e-01
3.72986764e-01 7.84305990e-01 5.32093346e-01 -4.16491121e-01
1.35827363e+00 6.77631319e-01 1.89200699e-01 -7.14380324e-01
-4.26750779e-01 9.27174687e-02 -2.28464842e-01 2.14255467e-01
4.14353371e-01 -5.36987960e-01 -8.46005917e-01 7.10231006e-01
-1.01314378e+00 2.01044932e-01 -3.11520904e-01 3.75130326e-01
-3.13127607e-01 9.00558949e-01 -3.32418412e-01 -6.24808967e-01
-1.41613901e-01 -1.25961697e+00 4.07627016e-01 1.43198267e-01
-5.63061833e-02 -8.02622080e-01 -3.30646127e-01 3.91880870e-01
9.92119759e-02 4.08961236e-01 1.29054129e+00 -1.27754009e+00
-6.32813692e-01 -6.31359220e-01 -1.04022875e-01 5.97206056e-01
3.55096102e-01 -1.17331810e-01 -9.08871889e-01 -6.74953401e-01
5.37952125e-01 -7.86002353e-02 6.63095355e-01 -5.67656048e-02
1.58756196e+00 -9.54878569e-01 -3.00830573e-01 4.47143376e-01
1.45725811e+00 1.63651481e-01 8.30048978e-01 2.78217226e-01
5.45666158e-01 3.91994864e-01 7.15698123e-01 4.70605940e-01
-1.06415294e-01 7.33570993e-01 7.16010153e-01 1.64155550e-02
4.33274269e-01 -4.40706342e-01 3.02311242e-01 7.10167363e-02
5.03097773e-01 -2.04927489e-01 -1.87742069e-01 2.79648989e-01
-1.62240493e+00 -1.03613925e+00 2.43343133e-02 2.99778795e+00
9.60389376e-01 2.55156577e-01 7.36771617e-04 1.69759974e-01
9.62803066e-01 3.41412216e-01 -5.78353405e-01 -2.90948689e-01
-2.37227324e-02 5.20970523e-02 7.18403637e-01 3.83003712e-01
-1.38013518e+00 6.47150576e-01 4.75139761e+00 1.20508289e+00
-1.11243689e+00 1.14966907e-01 9.67376590e-01 -3.10501270e-02
-3.04294229e-01 3.34952712e-01 -7.74507284e-01 6.89541221e-01
7.56170809e-01 -7.34214067e-01 2.59234607e-01 1.02601254e+00
-1.87307550e-03 1.96824625e-01 -1.38266397e+00 7.24953592e-01
1.04798183e-01 -1.11254358e+00 3.77378076e-01 4.44382012e-01
4.03491825e-01 -6.30654097e-01 3.63762200e-01 1.74932078e-01
1.38777658e-01 -7.41130769e-01 6.14585578e-01 1.67611793e-01
8.08163404e-01 -1.05574250e+00 5.06628096e-01 5.59297860e-01
-9.28212464e-01 -9.53921117e-03 -5.52051306e-01 2.27964565e-01
-3.31236541e-01 4.03954089e-01 -7.25566208e-01 8.17429364e-01
5.15212476e-01 2.92317808e-01 -6.70306563e-01 1.01783943e+00
-1.83257192e-01 7.27364063e-01 -4.83445138e-01 2.88742244e-01
9.85710993e-02 -3.37394834e-01 7.95738399e-01 9.37902689e-01
1.86320603e-01 -2.57600874e-01 2.77784839e-02 8.70995939e-01
-6.48887157e-01 2.59694904e-01 -6.72563672e-01 8.13252479e-02
6.60036206e-01 9.97661352e-01 -2.13567480e-01 -3.07154328e-01
-4.51866716e-01 1.08725619e+00 1.42817736e-01 1.75549150e-01
-1.13588703e+00 -3.83752227e-01 9.56145883e-01 3.26789051e-01
2.80451745e-01 1.96101248e-01 6.14988431e-03 -1.17802179e+00
3.17248046e-01 -1.23687971e+00 7.23335445e-01 -2.63750821e-01
-1.28315675e+00 4.61706877e-01 6.61154687e-02 -1.46739924e+00
1.33472711e-01 -1.53656274e-01 -5.83814263e-01 9.21444237e-01
-1.24940145e+00 -1.00828874e+00 -5.85417077e-02 7.53603816e-01
6.06866367e-02 4.65639792e-02 8.39168608e-01 4.16437298e-01
-5.69486141e-01 1.24368501e+00 2.11256787e-01 3.55074018e-01
6.46702886e-01 -9.84479487e-01 2.84911394e-01 1.29983544e+00
4.83116329e-01 8.66298139e-01 6.41378284e-01 -6.16738617e-01
-1.11801767e+00 -1.34593928e+00 7.21021652e-01 5.52168451e-02
5.78749180e-01 -3.93345267e-01 -1.29422808e+00 1.05490184e+00
3.58363241e-02 -9.59287807e-02 6.24120057e-01 -6.10346198e-01
-7.20425010e-01 -3.85885894e-01 -1.64670229e+00 7.54002512e-01
3.28406632e-01 -4.91150141e-01 -2.63198316e-01 2.75963455e-01
6.47763073e-01 -2.47640386e-01 -5.19032836e-01 2.59845257e-01
3.52447987e-01 -9.14207280e-01 8.44610393e-01 -8.35389376e-01
3.20046805e-02 -4.22429889e-01 -2.73676068e-01 -9.54116940e-01
1.19311191e-01 -6.56518281e-01 -1.92815319e-01 1.43361247e+00
2.82314092e-01 -9.22527671e-01 9.07689393e-01 7.30108380e-01
7.31718123e-01 -3.61628622e-01 -1.18498075e+00 -9.62909222e-01
2.19370082e-01 -2.87973553e-01 9.13208067e-01 1.09202898e+00
-1.95451275e-01 -2.77441233e-01 -4.75304276e-01 8.98915350e-01
8.80034447e-01 1.00101769e-01 1.05953372e+00 -9.83916938e-01
-4.90682095e-01 -1.79518238e-01 -6.06219053e-01 -9.31619346e-01
2.23766804e-01 -9.84433651e-01 -5.19597948e-01 -7.13838458e-01
2.92158604e-01 -3.69148731e-01 -3.22402477e-01 4.07634288e-01
-1.93373844e-01 2.11353675e-01 4.82035160e-01 2.71925628e-01
3.55755277e-02 2.44283900e-01 8.05566728e-01 -2.63367653e-01
9.68382061e-02 7.14298666e-01 -8.59577596e-01 6.03766263e-01
8.18650723e-01 -8.62266362e-01 -6.23598456e-01 2.34458864e-01
-5.50249629e-02 -5.92050180e-02 4.34322268e-01 -1.06141555e+00
1.88062117e-01 -6.74693361e-02 -2.43583322e-02 -2.60388166e-01
1.33274019e-01 -1.57473588e+00 5.08197427e-01 6.99977219e-01
-5.41036546e-01 -2.74744421e-01 2.06874445e-01 7.51801968e-01
-1.38361573e-01 -5.73106408e-01 1.12246799e+00 4.12075669e-02
-1.36832371e-01 5.90302706e-01 8.45858362e-03 -1.35261416e-01
1.44744325e+00 -2.07699418e-01 -3.61512661e-01 -4.92982388e-01
-7.11423755e-01 -1.60637930e-01 8.14175904e-01 1.98448241e-01
4.85267878e-01 -1.09456468e+00 -6.27200425e-01 2.76645750e-01
6.77848980e-02 -2.22852319e-01 1.63004056e-01 5.49451172e-01
-4.57252979e-01 2.46032905e-02 -3.78621928e-02 -1.97534606e-01
-1.78424990e+00 1.06812298e+00 3.87371868e-01 -3.76310736e-01
-4.34989274e-01 2.54752010e-01 6.95444047e-01 -1.82502985e-01
4.08957064e-01 -1.28884956e-01 1.31534338e-01 -2.02634186e-01
7.15589404e-01 2.84547150e-01 5.75204976e-02 -6.01702392e-01
-1.34557828e-01 4.53184322e-02 -5.02318025e-01 8.41459911e-03
7.90707231e-01 -1.45715728e-01 -2.29963735e-01 6.08192496e-02
1.56499612e+00 3.15186799e-01 -9.30702984e-01 -5.16396582e-01
-2.03488708e-01 -1.10971069e+00 -3.28689545e-01 -4.06318426e-01
-1.31700754e+00 9.19045746e-01 6.64573967e-01 4.58370328e-01
1.21784365e+00 -2.24193156e-01 8.25045228e-01 3.22048485e-01
4.48170394e-01 -5.39937675e-01 -3.66638899e-01 -1.94842607e-01
5.85832715e-01 -9.07766402e-01 1.44823328e-01 -6.56168997e-01
-7.13069022e-01 8.95224094e-01 5.99782050e-01 -6.17732015e-03
6.15763068e-01 -1.60962213e-02 -1.91027388e-01 3.39185447e-01
-3.43351424e-01 5.08414924e-01 -1.09656528e-01 4.61956978e-01
-3.71856332e-01 2.31866464e-02 -3.29608083e-01 6.81444347e-01
-2.33459875e-01 -2.74664044e-01 7.00342000e-01 6.94989383e-01
8.18868726e-02 -1.29035532e+00 -5.73990285e-01 3.36943924e-01
-8.93916190e-01 -1.21025415e-02 -2.55212814e-01 8.81030917e-01
2.73846872e-02 6.41680598e-01 -1.21241309e-01 -3.39774489e-01
2.98682094e-01 3.32098007e-02 2.06460431e-01 -4.60618705e-01
-6.36035144e-01 -2.28013977e-01 -2.47950375e-01 -4.10182148e-01
-3.85835432e-02 -6.57555938e-01 -1.03009367e+00 -6.15631700e-01
-4.08390850e-01 2.74958849e-01 5.63582718e-01 6.25308335e-01
2.52318859e-01 -1.61227118e-02 1.26070368e+00 -4.09641229e-02
-9.45556223e-01 -5.97796321e-01 -8.14581096e-01 7.97054529e-01
2.25415349e-01 -1.59261242e-01 -7.51546323e-01 1.92244127e-01] | [5.87234354019165, 7.1700310707092285] |
3de3cd6f-98c6-4ef7-9ad8-91422892be17 | comments-on-fast-and-scalable-search-of-whole | 2304.08297 | null | https://arxiv.org/abs/2304.08297v4 | https://arxiv.org/pdf/2304.08297v4.pdf | Comments on 'Fast and scalable search of whole-slide images via self-supervised deep learning' | Chen et al. [Chen2022] recently published the article 'Fast and scalable search of whole-slide images via self-supervised deep learning' in Nature Biomedical Engineering. The authors call their method 'self-supervised image search for histology', short SISH. We express our concerns that SISH is an incremental modification of Yottixel, has used MinMax binarization but does not cite the original works, and is based on a misnomer 'self-supervised image search'. As well, we point to several other concerns regarding experiments and comparisons performed by Chen et al. | ['H. R. Tizhoosh', 'Shivam Kalra', 'Abubakr Shafique', 'Mehdi Afshari', 'Milad Sikaroudi'] | 2023-04-07 | null | null | null | null | ['whole-slide-images'] | ['computer-vision'] | [ 2.96280473e-01 4.06691879e-02 -3.96527082e-01 -6.03838146e-01
-1.16453767e+00 -3.72151524e-01 3.09163600e-01 6.30776942e-01
-7.99183130e-01 9.74635422e-01 1.47316992e-01 -1.88768461e-01
-3.00282240e-01 -3.77559602e-01 -5.91171384e-01 -1.12568474e+00
-4.47380207e-02 5.80250025e-01 3.97140831e-01 -1.03241131e-01
5.96975386e-01 4.76473093e-01 -1.26278114e+00 4.01765347e-01
2.97356695e-01 9.20556962e-01 2.69052863e-01 9.91386116e-01
1.12096347e-01 5.82774818e-01 -5.89376926e-01 -1.00987129e-01
-7.51529783e-02 -4.12350148e-01 -1.27896392e+00 -3.26921284e-01
2.52714187e-01 9.91074443e-02 -1.68418974e-01 1.12535906e+00
1.01739085e+00 -3.12633574e-01 7.52228379e-01 -8.51965010e-01
-4.79198545e-01 5.56576490e-01 -5.57309985e-01 7.70965159e-01
1.86338741e-02 1.07312456e-01 7.80730665e-01 -5.44670165e-01
1.01907873e+00 6.59730196e-01 1.07154143e+00 6.86578512e-01
-1.16683114e+00 -6.86882496e-01 -6.11928821e-01 4.02177364e-01
-1.63786900e+00 -4.30902034e-01 3.76523942e-01 -2.58128196e-01
9.66564655e-01 6.39864683e-01 4.96590197e-01 1.08038545e+00
4.62647349e-01 7.45543122e-01 1.45667398e+00 -6.99964941e-01
5.66320419e-01 4.10916880e-02 2.72305030e-03 8.88368547e-01
1.43127367e-01 3.65693688e-01 -5.98536789e-01 -3.71318340e-01
6.33003771e-01 -1.76012203e-01 1.28702000e-01 -5.64086847e-02
-1.42529035e+00 7.70864606e-01 5.54754615e-01 7.97503531e-01
8.21810588e-02 2.13636234e-01 7.21178651e-01 2.84801513e-01
5.36013663e-01 5.46344221e-01 -3.33706319e-01 -5.29701263e-03
-1.24606240e+00 -1.98589146e-01 5.72436094e-01 7.65742958e-01
4.42760676e-01 -2.71382213e-01 1.64170906e-01 6.08742058e-01
3.91185611e-01 2.00827286e-01 1.00474405e+00 -9.96105134e-01
-6.55527890e-01 4.07213271e-01 -4.50682402e-01 -8.95984471e-01
-7.34303534e-01 -4.20582831e-01 -1.23840761e+00 2.04641938e-01
1.74118295e-01 -9.81463194e-02 -9.95033801e-01 1.05162656e+00
-4.15682904e-02 -2.87122697e-01 -1.74050987e-01 8.17367435e-01
1.34072626e+00 1.26282230e-01 4.36509028e-02 -2.40666270e-01
1.18253577e+00 -8.37006330e-01 -8.89783859e-01 1.72182620e-01
7.71726847e-01 -9.55107272e-01 5.73473215e-01 6.08230114e-01
-9.49347854e-01 -2.86053360e-01 -1.18696284e+00 -1.36423871e-01
-9.72408354e-01 1.37936503e-01 8.23765039e-01 4.91878718e-01
-1.54612911e+00 6.62872195e-01 -6.70354724e-01 -9.45255578e-01
7.30601549e-01 6.37743592e-01 -5.10346115e-01 5.19655347e-01
-9.64351773e-01 1.07751775e+00 2.13120162e-01 -1.91304520e-01
-7.81240225e-01 -4.42068040e-01 -4.32672054e-01 -3.09494704e-01
1.71311602e-01 -3.79374802e-01 1.19062746e+00 -4.50031012e-01
-1.49059844e+00 2.13013411e+00 -1.16232798e-01 -6.91628397e-01
4.13384169e-01 2.20546737e-01 -5.10945082e-01 3.71584743e-01
9.08783525e-02 9.17576015e-01 2.37265348e-01 -8.80630910e-01
-4.87876326e-01 -2.82057226e-01 -4.80652660e-01 -8.97453539e-03
-1.60270512e-01 3.25736880e-01 -3.02052319e-01 -7.10837126e-01
1.66713938e-01 -8.16070676e-01 -4.95529950e-01 2.92323649e-01
-7.99026608e-01 -3.31065595e-01 5.28410017e-01 -1.55759275e-01
1.23889053e+00 -2.33168793e+00 -4.45808768e-01 3.91347677e-01
5.96241713e-01 1.75517261e-01 6.26196414e-02 4.67895508e-01
-5.21818340e-01 5.19870162e-01 -1.16907693e-01 -5.03365621e-02
-1.66589692e-01 1.80164967e-02 2.00657859e-01 1.01637793e+00
-9.62168798e-02 1.07138121e+00 -1.02255678e+00 -1.22657466e+00
5.70511408e-02 1.80037752e-01 5.61172776e-02 -1.88985616e-01
2.02995017e-01 3.68441999e-01 -7.43119493e-02 8.99788201e-01
3.68554026e-01 -6.80325270e-01 1.98258348e-02 -4.20837641e-01
-1.66599244e-01 -1.19132668e-01 -4.96570438e-01 1.85586596e+00
3.23873937e-01 9.92990792e-01 -3.87872458e-02 -1.34042239e+00
9.11814928e-01 2.33902305e-01 9.42731917e-01 -4.90184665e-01
4.60541010e-01 4.46975976e-01 -2.36161262e-01 -6.35205269e-01
-3.40977639e-01 -2.67012984e-01 5.31788096e-02 3.47381532e-01
2.04214796e-01 -1.63524345e-01 -7.09785521e-02 1.68008357e-01
1.37350094e+00 -1.89090297e-01 6.13870919e-01 -8.25930119e-01
4.58888441e-01 3.25476617e-01 -8.62531438e-02 1.04203069e+00
-6.83556318e-01 1.00749290e+00 2.67913759e-01 -4.64355260e-01
-9.07793760e-01 -9.78499949e-01 -5.11981726e-01 9.25694644e-01
2.28319019e-02 -2.99194455e-01 -5.45226574e-01 -6.51131213e-01
-1.50472373e-01 2.97408160e-02 -1.14011037e+00 2.20853537e-02
-1.48915485e-01 -8.67188394e-01 8.05582285e-01 3.53547603e-01
3.20963413e-01 -9.75989223e-01 -9.26689208e-01 -6.42115697e-02
1.08022466e-01 -5.98598003e-01 -4.89315450e-01 9.72445428e-01
-7.07021415e-01 -1.10821950e+00 -1.10084152e+00 -1.33238089e+00
8.10544133e-01 -8.49348232e-02 1.03405356e+00 3.06666613e-01
-1.01727319e+00 -5.20786038e-03 -5.42542100e-01 -5.76013505e-01
-4.53936189e-01 1.63780659e-01 -1.43369988e-01 -4.77452844e-01
5.23128390e-01 -4.29317951e-01 -1.01442242e+00 -9.48060956e-03
-1.03869081e+00 -1.58371061e-01 7.42278814e-01 1.33505666e+00
8.67614627e-01 1.24775246e-02 4.64821666e-01 -1.15298212e+00
3.93230528e-01 -1.33686036e-01 -2.14741871e-01 2.40439534e-01
-1.08260489e+00 2.23147962e-02 3.07854861e-01 -1.21179512e-02
-4.18393731e-01 1.71860352e-01 -5.48279107e-01 -5.90609796e-02
-4.34032410e-01 3.61027479e-01 4.78333741e-01 -7.99895108e-01
1.14447761e+00 5.22528291e-01 4.93603766e-01 -2.54262656e-01
-1.79330587e-01 7.67603159e-01 7.97804892e-01 1.35082573e-01
3.62505019e-01 8.64619255e-01 2.40943208e-01 -5.51071405e-01
-5.39015055e-01 -1.04035437e+00 -3.83276969e-01 -1.23954885e-01
8.94337356e-01 -5.48249006e-01 -1.02088642e+00 4.63706553e-01
-8.44530106e-01 -3.02006721e-01 -4.60857064e-01 3.44586760e-01
-6.11321509e-01 5.95122576e-01 -8.32355380e-01 -3.85928333e-01
-6.08588755e-01 -9.77716565e-01 1.16627729e+00 2.73713261e-01
-4.92057830e-01 -1.25056195e+00 4.62094307e-01 -1.14802681e-02
6.50309205e-01 2.95394063e-01 6.85786128e-01 -1.04084182e+00
1.36434332e-01 -5.31647563e-01 -4.75812882e-01 5.73928319e-02
3.24326277e-01 -8.85030553e-02 -9.49970603e-01 -4.64011937e-01
-2.74753645e-02 -5.78808427e-01 6.31649196e-01 6.72125041e-01
1.53952384e+00 -4.48909044e-01 -8.22233021e-01 8.26054275e-01
1.81793523e+00 3.39352310e-01 7.98158169e-01 6.64955974e-01
9.27162692e-02 4.06747997e-01 4.25823927e-01 1.39925063e-01
-1.06877580e-01 2.87849247e-01 2.20744967e-01 -6.73663199e-01
-7.40495836e-03 2.78326094e-01 -3.45288098e-01 7.87469864e-01
1.17364250e-01 2.45554838e-02 -9.95311677e-01 5.81133604e-01
-1.46353722e+00 -5.54949641e-01 5.61395809e-02 1.72015905e+00
1.15010071e+00 1.29085720e-01 6.54653013e-02 9.22888443e-02
6.49175584e-01 -1.12648629e-01 -5.50297081e-01 -2.84410894e-01
-4.95983541e-01 5.45028806e-01 8.06928456e-01 4.24889982e-01
-1.37629580e+00 7.60167062e-01 7.81716299e+00 1.21363366e+00
-1.33858073e+00 1.62624955e-01 1.24531412e+00 7.57863298e-02
7.05622882e-02 -1.99199855e-01 -6.24558389e-01 5.46708524e-01
9.48093891e-01 -3.16442668e-01 -1.39841571e-01 5.14004171e-01
-6.88549578e-02 -2.77226299e-01 -8.57561886e-01 1.16299665e+00
3.73160869e-01 -1.74283874e+00 -3.56094718e-01 -1.52567029e-01
4.83149081e-01 1.86871871e-01 3.60583402e-02 -2.66096085e-01
2.39158452e-01 -1.31311011e+00 -1.96426615e-01 5.45529842e-01
1.05686045e+00 -5.60903013e-01 9.75899577e-01 3.60622257e-02
-4.50807303e-01 2.00749263e-01 -2.46687844e-01 4.33106929e-01
-3.50459725e-01 7.76019096e-01 -1.25611615e+00 -2.66560391e-02
9.69085455e-01 4.33748126e-01 -8.81169558e-01 1.42669582e+00
5.10552108e-01 7.18387723e-01 -4.16215092e-01 -5.13517141e-01
1.29973263e-01 5.02649426e-01 3.79109621e-01 1.65490806e+00
5.44847175e-02 -2.39686385e-01 -2.28734925e-01 6.29552841e-01
6.01311326e-01 3.69703203e-01 -3.21471721e-01 -2.83153832e-01
1.73065156e-01 1.20462358e+00 -1.42950523e+00 -4.95290607e-01
3.79226059e-01 9.60500479e-01 -1.46067217e-01 -8.81802756e-03
-5.27663171e-01 -5.84251463e-01 -3.91651362e-01 -1.54210553e-02
1.29410073e-01 4.92086947e-01 -7.18761921e-01 -5.78728557e-01
-7.11808383e-01 -7.16477752e-01 5.48458517e-01 -8.23976994e-01
-1.43374157e+00 6.22235417e-01 -9.53515545e-02 -1.25626194e+00
2.32347511e-02 -9.14509296e-01 -1.67866513e-01 3.12518656e-01
-1.26298654e+00 -9.96829391e-01 6.74294382e-02 3.56218159e-01
1.59859404e-01 -2.24254623e-01 1.31232083e+00 2.12854464e-02
-1.96297601e-01 8.33727837e-01 5.60818553e-01 7.27932975e-02
1.05368149e+00 -1.50153112e+00 -1.47736212e-02 9.09815505e-02
-1.97456703e-01 6.06998086e-01 9.02026176e-01 -3.91280681e-01
-1.27542365e+00 -5.89684963e-01 1.19022691e+00 -1.52272239e-01
4.91656601e-01 4.04005349e-02 -5.84009051e-01 3.55284959e-01
7.01286137e-01 7.93656036e-02 1.13145113e+00 -4.30169523e-01
2.29635745e-01 -3.36276919e-01 -1.38237703e+00 5.68955362e-01
6.56851888e-01 -3.51488858e-01 -3.36345822e-01 1.16378856e+00
2.57051498e-01 -4.89452690e-01 -1.04793513e+00 6.79741263e-01
5.13758421e-01 -1.04468679e+00 1.03334844e+00 -2.20127165e-01
3.90454739e-01 1.06096886e-01 3.84146899e-01 -1.00932360e+00
-5.25565028e-01 -6.32770836e-01 1.59467012e-01 6.14251018e-01
2.43723184e-01 -4.89919722e-01 1.24347234e+00 -2.48623013e-01
-1.04680797e-02 -1.30142117e+00 -1.26395750e+00 -6.21410191e-01
1.48793027e-01 8.25557634e-02 3.01846653e-01 1.13776052e+00
7.56051660e-01 1.39142200e-01 -2.37523057e-02 -4.29660022e-01
6.87375069e-01 -1.65280044e-01 5.83013110e-02 -9.57644105e-01
-2.02240840e-01 -8.99436057e-01 -5.52777290e-01 -4.14147943e-01
-2.19586506e-01 -9.11595821e-01 6.20608509e-01 -1.13314939e+00
5.55776834e-01 -1.33005589e-01 -7.69186020e-01 5.68662405e-01
1.67745873e-01 8.83269548e-01 -4.93358016e-01 5.04437149e-01
-1.00222194e+00 -4.76655722e-01 1.21584761e+00 -2.95563608e-01
2.86600500e-01 -1.96771786e-01 -9.20370996e-01 2.86561042e-01
9.71007645e-01 -4.37533110e-01 -5.79721574e-03 -3.09542082e-02
2.37909004e-01 -1.76742196e-01 4.37778234e-01 -1.00465012e+00
8.43891919e-01 1.02091418e-03 5.80176413e-01 -6.11637473e-01
-4.70895395e-02 -6.29603505e-01 1.32106468e-01 9.65714514e-01
-7.99098909e-01 -2.31646281e-02 2.50121176e-01 2.17096999e-01
-1.85229629e-01 -3.65638971e-01 1.11328733e+00 -6.07760012e-01
-6.78121150e-01 -1.76735640e-01 -8.29857707e-01 -4.94836390e-01
8.80930960e-01 -6.62144721e-01 -3.62365216e-01 -3.98949325e-01
-7.94791520e-01 3.74102257e-02 3.40322316e-01 -3.40842217e-01
8.94800603e-01 -1.04143298e+00 -6.39677107e-01 -2.07949970e-02
2.56151110e-01 -4.99146171e-02 3.78607780e-01 1.34840477e+00
-9.46459770e-01 7.73308098e-01 -1.47232085e-01 -7.47360706e-01
-1.35690928e+00 6.04748130e-01 1.60768077e-01 -3.29047292e-01
-3.76828820e-01 1.17345750e+00 -2.92981446e-01 -3.90809476e-01
3.33751649e-01 3.74086089e-02 -2.88893551e-01 -1.62702322e-01
2.88003951e-01 2.14822188e-01 1.70363888e-01 -3.91759783e-01
-7.34815896e-01 3.83458674e-01 -2.47714639e-01 -3.81118581e-02
1.24660110e+00 1.32978763e-02 -2.69206852e-01 7.05350578e-01
1.82923973e+00 -2.69630432e-01 -3.58867854e-01 -1.08944187e-02
8.23053941e-02 3.64708453e-02 2.94797152e-01 -1.14374518e+00
-9.61913586e-01 6.77023768e-01 1.31799650e+00 8.65692198e-02
1.31518519e+00 -7.87087530e-03 5.81285954e-01 4.17239606e-01
6.63796216e-02 -1.05247796e+00 1.17835470e-01 3.01063687e-01
8.23563099e-01 -1.30658650e+00 5.36295593e-01 -1.63213640e-01
-1.81643724e-01 1.01744854e+00 4.96773869e-01 -7.78953582e-02
8.55699599e-01 6.09553039e-01 4.37691838e-01 -4.53696460e-01
-7.01252103e-01 -9.09969211e-02 1.43456534e-01 5.23320854e-01
5.73426485e-01 -3.73282641e-01 -8.11126351e-01 2.29805321e-01
-2.52810955e-01 6.72423363e-01 -1.68710742e-02 1.27532673e+00
-3.50024521e-01 -9.13395286e-01 -3.26791167e-01 7.71520078e-01
-6.90147281e-01 -2.55726337e-01 -6.18675649e-01 5.33443272e-01
3.19572568e-01 1.96950495e-01 -6.49474040e-02 -4.78778958e-01
-4.42354828e-01 -2.25683361e-01 3.76041889e-01 -5.64761162e-01
-5.94221532e-01 3.22671831e-01 -3.56758714e-01 -1.58534393e-01
-7.20181882e-01 -5.23453772e-01 -1.12897468e+00 1.66190472e-02
-4.60026890e-01 4.74031478e-01 6.10022187e-01 9.64733601e-01
3.35498840e-01 4.14198637e-01 5.72670639e-01 -4.16269928e-01
-3.79237115e-01 -1.00912082e+00 -7.84933209e-01 2.12667197e-01
4.64474887e-01 -1.95068732e-01 -5.66410005e-01 2.51139373e-01] | [15.090270042419434, -2.9483370780944824] |
c336e9f4-d6d4-4471-b784-4b1049be2c92 | flowgrad-controlling-the-output-of-generative | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Liu_FlowGrad_Controlling_the_Output_of_Generative_ODEs_With_Gradients_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Liu_FlowGrad_Controlling_the_Output_of_Generative_ODEs_With_Gradients_CVPR_2023_paper.pdf | FlowGrad: Controlling the Output of Generative ODEs With Gradients | Generative modeling with ordinary differential equations (ODEs) has achieved fantastic results on a variety of applications. Yet, few works have focused on controlling the generated content of a pre-trained ODE-based generative model. In this paper, we propose to optimize the output of ODE models according to a guidance function to achieve controllable generation. We point out that, the gradients can be efficiently back-propagated from the output to any intermediate time steps on the ODE trajectory, by decomposing the back-propagation and computing vector-Jacobian products. To further accelerate the computation of the back-propagation, we propose to use a non-uniform discretization to approximate the ODE trajectory, where we measure how straight the trajectory is and gather the straight parts into one discretization step. This allows us to save 90% of the back-propagation time with ignorable error. Our framework, named FlowGrad, outperforms the state-of-the-art baselines on text-guided image manipulation. Moreover, FlowGrad enables us to find global semantic directions in frozen ODE-based generative models that can be used to manipulate new images without extra optimization. | ['Qiang Liu', 'Wei Ping', 'Chengyue Gong', 'Shujian Zhang', 'Lemeng Wu', 'Xingchao Liu'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['image-manipulation'] | ['computer-vision'] | [-8.57489035e-02 1.28214210e-01 2.88457811e-01 -1.67521253e-01
-4.07018006e-01 -6.94620371e-01 7.98825026e-01 -3.09792757e-01
-3.54748935e-01 5.22725940e-01 4.32664081e-02 -1.69135571e-01
1.22975074e-01 -9.50304508e-01 -9.64156747e-01 -7.34227121e-01
3.27682912e-01 5.54279029e-01 6.73702583e-02 -2.78569788e-01
2.66305417e-01 5.97054958e-01 -1.13513994e+00 -6.90985564e-03
1.19453549e+00 7.61763632e-01 2.59934217e-01 8.15249860e-01
-9.47913751e-02 9.05721962e-01 -4.45503563e-01 -5.41155756e-01
1.82692021e-01 -7.82025814e-01 -6.87499166e-01 1.42581850e-01
1.89169034e-01 -5.69858253e-01 -4.98764187e-01 1.10222638e+00
5.38088679e-01 3.75250250e-01 8.96588027e-01 -1.19243598e+00
-9.85521376e-01 4.62482989e-01 -2.79099166e-01 -1.06988475e-01
1.09003291e-01 4.51574355e-01 5.89941621e-01 -9.35648620e-01
1.24551785e+00 1.24462044e+00 4.59977955e-01 5.48479021e-01
-1.34030938e+00 -3.46119940e-01 2.47502580e-01 5.10894470e-02
-1.32718396e+00 -2.41599262e-01 8.62411797e-01 -5.84938347e-01
5.91527343e-01 2.27180347e-01 9.51157808e-01 9.71502900e-01
3.25930893e-01 6.97103381e-01 8.40593517e-01 -1.87430739e-01
3.71344090e-01 1.07894965e-01 -4.62759405e-01 1.02475011e+00
-2.51771927e-01 -4.01378982e-02 -2.69984037e-01 4.82623428e-02
1.18739724e+00 -8.98057744e-02 -3.05803269e-01 -3.47954720e-01
-1.10374165e+00 1.05438113e+00 6.45058453e-01 1.06511861e-01
-4.06536788e-01 4.28837150e-01 1.59281448e-01 1.73364207e-01
5.75821221e-01 7.14505315e-01 -9.70447809e-02 -3.37010980e-01
-1.04853833e+00 6.18954003e-01 8.51873755e-01 1.02414215e+00
6.60310745e-01 4.86739315e-02 -5.59516430e-01 6.88627005e-01
3.52997154e-01 3.76024902e-01 3.27818811e-01 -1.21727431e+00
2.44990602e-01 4.64521438e-01 1.42300621e-01 -1.13179052e+00
-2.95596588e-02 -3.99291039e-01 -8.44089270e-01 2.00738013e-01
5.44085383e-01 -3.40266168e-01 -1.20797253e+00 1.60410762e+00
5.65411925e-01 5.50210997e-02 -3.49001884e-01 1.13237000e+00
4.61994350e-01 1.02978909e+00 -1.38571471e-01 6.25752881e-02
1.01306045e+00 -1.27167034e+00 -5.95069051e-01 3.32378811e-04
5.14481962e-01 -7.77154922e-01 1.24664783e+00 2.02819213e-01
-1.44278944e+00 -3.18543434e-01 -8.40698838e-01 -4.53715354e-01
-1.68626845e-01 1.99707270e-01 4.26361561e-01 7.46146962e-02
-1.08336043e+00 9.96444046e-01 -1.23695064e+00 1.23650879e-01
4.96925175e-01 5.34658991e-02 1.86993126e-02 4.00077254e-02
-1.05490804e+00 7.51775503e-01 7.26980045e-02 2.52913594e-01
-1.12496090e+00 -1.03164649e+00 -7.99011230e-01 2.00964157e-02
3.53953719e-01 -1.25633848e+00 1.00789607e+00 -7.29174197e-01
-2.03762293e+00 6.26914442e-01 -2.03027859e-01 -2.21361741e-01
1.16047227e+00 -1.53626367e-01 2.72771776e-01 5.99397309e-02
-1.05905436e-01 1.01053476e+00 1.02855134e+00 -1.18264616e+00
-5.11583388e-02 -6.44565672e-02 3.09365302e-01 4.15205926e-01
-4.02660251e-01 -2.04054773e-01 -7.91976869e-01 -8.97259533e-01
-2.48114951e-02 -1.35082352e+00 -4.69566762e-01 5.32848895e-01
-6.30172729e-01 2.48058625e-02 7.12005436e-01 -7.45938838e-01
1.26392746e+00 -1.93030179e+00 7.75117159e-01 -6.14039563e-02
2.20223829e-01 2.59268045e-01 -7.16480836e-02 4.81822848e-01
4.62705970e-01 7.93205947e-02 -4.97318983e-01 -6.00636601e-01
2.27966815e-01 1.74522594e-01 -4.96939808e-01 2.66570300e-01
7.42841437e-02 1.16586578e+00 -1.00900590e+00 -4.10031527e-01
2.68120408e-01 9.95495617e-01 -9.41597879e-01 2.35554159e-01
-6.69109285e-01 7.00923979e-01 -4.03848052e-01 4.31193225e-03
7.23964453e-01 -2.76345402e-01 -2.08273217e-01 -1.95291132e-01
-2.18436852e-01 2.72822678e-01 -9.95337725e-01 1.90501082e+00
-6.87185168e-01 6.51950300e-01 -6.01015128e-02 -6.94842815e-01
7.44054317e-01 -8.28781649e-02 2.38477811e-01 -2.87425905e-01
2.66860247e-01 3.15045565e-02 -2.23877713e-01 -3.92675579e-01
5.28448105e-01 1.14275888e-01 2.16849074e-01 4.08226550e-01
-9.59535688e-03 -6.37364984e-01 4.34914470e-01 2.59311080e-01
6.28306806e-01 5.27476311e-01 -3.84457082e-01 -9.79687124e-02
3.82566601e-01 1.45001531e-01 1.79974914e-01 4.81579512e-01
2.52384663e-01 8.05898368e-01 6.24677181e-01 -2.15381324e-01
-1.14279354e+00 -9.14325237e-01 1.31587505e-01 8.32825124e-01
1.52447522e-01 -3.89784873e-01 -1.25456357e+00 -4.70798910e-01
-1.36554867e-01 9.07283366e-01 -6.22751057e-01 -3.53956968e-01
-6.91142797e-01 -5.52897513e-01 3.34094852e-01 4.68461573e-01
6.41877413e-01 -1.02676582e+00 -4.33026224e-01 3.22511673e-01
-1.17422044e-01 -8.71902406e-01 -8.57568145e-01 -2.03662887e-01
-8.59653175e-01 -5.26019394e-01 -1.16346478e+00 -6.73508584e-01
1.14414966e+00 -1.10795066e-01 8.83138478e-01 -8.20271820e-02
-1.52422458e-01 7.25208372e-02 2.64385082e-02 -1.02837859e-02
-5.12441218e-01 1.85406104e-01 -5.68371117e-01 -6.26292871e-03
-3.18957001e-01 -5.70566893e-01 -8.16819787e-01 3.02730650e-01
-9.07132685e-01 4.29947197e-01 1.86742350e-01 7.38669813e-01
8.23107660e-01 -1.95810273e-01 1.43182545e-03 -6.73129976e-01
7.81830907e-01 -2.32809350e-01 -7.29682326e-01 9.41236615e-02
-4.56448138e-01 5.65103710e-01 9.04429793e-01 -6.68434680e-01
-1.12045097e+00 -2.27043908e-02 -9.82789770e-02 -7.77211964e-01
3.05447668e-01 4.57128972e-01 1.18978113e-01 -1.72562599e-01
5.17793655e-01 3.42757225e-01 -2.84271985e-02 -3.36373538e-01
8.93240571e-01 1.76829457e-01 4.85585064e-01 -6.24230266e-01
8.29504251e-01 4.98905301e-01 6.20042682e-02 -5.98866463e-01
-6.97536647e-01 1.73166677e-01 -4.36686277e-01 -2.64465302e-01
9.96890306e-01 -6.20054245e-01 -6.25186503e-01 7.95637488e-01
-1.35962236e+00 -8.30342591e-01 -2.68578768e-01 3.06178421e-01
-7.08755136e-01 3.61542478e-02 -8.91849160e-01 -5.01611173e-01
-4.40419912e-01 -1.27939391e+00 1.09827745e+00 1.50407001e-01
-2.98479736e-01 -1.20645106e+00 1.39150098e-01 3.33362222e-02
6.43302500e-01 2.69824386e-01 7.08204567e-01 1.34885833e-01
-8.67892325e-01 -4.22686003e-02 -4.45960611e-02 2.37518445e-01
-1.30102918e-01 3.37982118e-01 -5.80022573e-01 -1.72778383e-01
1.62783116e-02 -1.23134099e-01 9.40063953e-01 4.39317673e-01
1.28747535e+00 -6.35054588e-01 -2.38453612e-01 1.15128982e+00
1.09585595e+00 7.63975680e-02 7.30286479e-01 1.78284422e-01
9.69485939e-01 2.91492522e-01 3.46966267e-01 3.86803895e-01
5.78174889e-01 6.58710957e-01 2.53282547e-01 3.59927639e-02
-3.00084770e-01 -6.19308949e-01 3.22350591e-01 7.98486114e-01
-2.71202058e-01 -3.37690026e-01 -5.97635567e-01 3.53677571e-01
-1.77512717e+00 -7.57184267e-01 -1.42517453e-02 2.01815748e+00
1.05928230e+00 2.58355197e-02 -2.17204049e-01 -2.74457157e-01
4.46871072e-01 2.95422912e-01 -7.65528679e-01 -4.05810684e-01
3.05102885e-01 7.55480453e-02 2.85216331e-01 8.11565399e-01
-8.96914065e-01 1.21363378e+00 6.21879721e+00 8.80925477e-01
-1.49435973e+00 -4.16389108e-02 5.29330850e-01 -1.98435783e-01
-6.46322489e-01 8.78019556e-02 -6.95444405e-01 6.09522521e-01
5.38419783e-01 -2.28200823e-01 7.44617403e-01 8.07054877e-01
4.48869914e-01 6.19906187e-02 -1.00233459e+00 7.14316547e-01
1.85479317e-02 -1.43355119e+00 1.75443947e-01 -9.87354442e-02
1.09278274e+00 -3.16152900e-01 2.84128815e-01 1.53386995e-01
5.23226738e-01 -8.56121302e-01 9.74440932e-01 7.61188030e-01
5.35338163e-01 -6.22865617e-01 2.15771366e-02 4.24899280e-01
-9.20270443e-01 4.53549802e-01 -4.12371188e-01 1.33096367e-01
4.95099157e-01 7.80131161e-01 -7.38114536e-01 1.76924542e-01
2.90042400e-01 5.12925625e-01 -2.64774591e-01 7.40771472e-01
-6.39970958e-01 3.92613590e-01 -3.10493648e-01 -1.78310305e-01
4.01120812e-01 -7.45661497e-01 6.31772339e-01 9.17962909e-01
5.05322814e-01 -7.39352778e-02 -6.57801330e-02 1.67089951e+00
-2.31600478e-01 -1.21801026e-01 -3.39277178e-01 -2.52145559e-01
3.48375529e-01 1.12591147e+00 -6.25929475e-01 -4.09236640e-01
1.23941615e-01 1.48648977e+00 4.32406902e-01 5.37141681e-01
-1.17627716e+00 -6.13608241e-01 5.63279748e-01 1.53165311e-01
4.13092136e-01 -5.54824412e-01 -2.87788004e-01 -1.24066532e+00
1.08598225e-01 -4.97277349e-01 -1.75877377e-01 -1.04429400e+00
-6.88868523e-01 6.70915127e-01 -4.51823398e-02 -8.48554492e-01
-5.20537853e-01 -2.65679300e-01 -5.79357982e-01 9.76143181e-01
-1.27592337e+00 -1.02214432e+00 -4.17343169e-01 4.14570957e-01
3.66940349e-01 3.69637489e-01 5.30964017e-01 3.21058810e-01
-4.92168546e-01 5.82307935e-01 7.97586218e-02 1.57113057e-02
5.39072514e-01 -1.05383098e+00 6.48049176e-01 7.99969614e-01
-5.92313427e-03 7.48858213e-01 7.42740273e-01 -6.00577533e-01
-1.33049166e+00 -1.21881223e+00 7.94114590e-01 -2.46244997e-01
5.38162291e-01 -4.03473645e-01 -8.91845226e-01 7.70747185e-01
1.68353438e-01 1.54302865e-01 -1.68015704e-01 -5.64372241e-01
-1.98424682e-01 6.43007010e-02 -9.94697750e-01 9.38778639e-01
1.19701445e+00 -4.52615291e-01 1.17353918e-02 3.52683991e-01
7.11569667e-01 -9.24146473e-01 -8.68317187e-01 1.08074229e-02
5.00913322e-01 -8.16868782e-01 9.70979273e-01 -3.83725494e-01
1.00302958e+00 -3.87711972e-01 4.10952359e-01 -1.81535447e+00
-2.39852443e-01 -1.05019855e+00 -3.24014038e-01 1.05361569e+00
4.34639364e-01 -5.30686140e-01 5.55189669e-01 6.83211088e-01
-3.00980300e-01 -1.09806156e+00 -3.11517060e-01 -6.64352894e-01
3.24952692e-01 -1.82634279e-01 5.28382003e-01 6.30180418e-01
-3.06534350e-01 -4.31451239e-02 -3.74498457e-01 -1.37210429e-01
5.02137303e-01 1.12835668e-01 8.74516189e-01 -7.03407228e-01
-5.08236408e-01 -5.78899205e-01 -2.03503564e-01 -1.62970400e+00
-1.43163409e-02 -9.31878448e-01 7.19091818e-02 -1.76791465e+00
6.49038255e-02 -4.05350268e-01 3.64787251e-01 2.87686795e-01
-3.27987492e-01 1.51962146e-01 4.27833170e-01 2.44824961e-01
-2.08473787e-01 8.78143191e-01 2.13137937e+00 -1.06717207e-01
-3.77632201e-01 -2.65803903e-01 -5.42912841e-01 7.39902735e-01
4.87588286e-01 -3.85782242e-01 -5.12486577e-01 -7.26275444e-01
2.94133842e-01 -2.09292658e-02 4.09558296e-01 -7.32141018e-01
3.53043675e-01 -3.04760128e-01 1.94514364e-01 -3.66828203e-01
4.07038629e-01 -2.55930811e-01 3.75869513e-01 4.74820822e-01
-6.00126803e-01 -1.46038949e-01 -6.72133267e-02 2.68043965e-01
-1.67398646e-01 -3.64857614e-01 8.82792115e-01 -1.36372253e-01
-9.81248841e-02 4.87129718e-01 -2.90403575e-01 2.16089219e-01
9.92161751e-01 1.62419841e-01 -1.81324095e-01 -5.88934600e-01
-8.46960723e-01 1.09364390e-01 6.76062107e-01 3.14076841e-01
4.11632150e-01 -1.44076335e+00 -4.86761421e-01 8.56498405e-02
-4.95881170e-01 3.09676856e-01 1.88475132e-01 8.66082489e-01
-8.79757106e-01 2.01103732e-01 5.65917753e-02 -5.00953734e-01
-7.44601905e-01 4.89837229e-01 5.39691746e-01 -3.51901799e-01
-7.07736492e-01 1.02988875e+00 1.63146108e-01 -3.73616546e-01
1.32701829e-01 -4.73991901e-01 2.41068155e-01 2.22640834e-03
3.60012144e-01 5.81668019e-01 -1.37286708e-01 -3.88123572e-01
-3.27991024e-02 7.37736702e-01 -1.02735963e-02 -4.57548171e-01
1.26133740e+00 -1.64561085e-02 -1.34920239e-01 2.41135210e-01
1.52099037e+00 -1.71554327e-01 -1.84509218e+00 2.19650403e-01
-7.12628961e-01 -3.76269221e-01 2.06430107e-01 -6.81866229e-01
-1.45830393e+00 1.01439965e+00 1.33636028e-01 2.01006830e-02
7.36027181e-01 -1.83555767e-01 1.07404530e+00 1.68966472e-01
1.20849259e-01 -9.74060535e-01 2.07971424e-01 5.90390742e-01
1.14772189e+00 -8.24346900e-01 -2.29752675e-01 -5.85560560e-01
-7.43333220e-01 1.10600102e+00 3.92079592e-01 -4.53614950e-01
5.49792588e-01 3.18350285e-01 -1.12005666e-01 -8.19990709e-02
-5.60189724e-01 2.30358958e-01 3.49766374e-01 3.56215835e-01
3.51341903e-01 -1.21927150e-01 -4.09315765e-01 1.02021143e-01
-5.27304411e-01 2.53541917e-01 4.51267034e-01 7.23698318e-01
-1.42096952e-01 -9.55149651e-01 -1.56214043e-01 1.93345383e-01
-1.66450873e-01 -1.90481201e-01 -3.07822376e-01 5.39329469e-01
-1.32362321e-01 5.44348419e-01 1.40436575e-01 -1.77569557e-02
2.97157586e-01 6.36293218e-02 8.23113322e-01 -2.61738211e-01
-2.89812297e-01 5.84203862e-02 -2.36437678e-01 -6.47589684e-01
-6.77354634e-02 -5.77624798e-01 -1.41132665e+00 -4.96063828e-01
-1.97773337e-01 -3.41906548e-02 7.09518492e-01 7.64379561e-01
6.39597416e-01 7.05372155e-01 6.12861812e-01 -1.14445508e+00
-6.88763559e-01 -7.46149659e-01 -1.60817698e-01 4.82345104e-01
1.19431809e-01 -5.90243816e-01 -5.12713790e-01 2.96379328e-01] | [11.442907333374023, -0.49640223383903503] |
12788411-d167-410a-8c37-f76fe2def43b | robust-text-line-detection-in-historical | 2203.12346 | null | https://arxiv.org/abs/2203.12346v2 | https://arxiv.org/pdf/2203.12346v2.pdf | Robust Text Line Detection in Historical Documents: Learning and Evaluation Methods | Text line segmentation is one of the key steps in historical document understanding. It is challenging due to the variety of fonts, contents, writing styles and the quality of documents that have degraded through the years. In this paper, we address the limitations that currently prevent people from building line segmentation models with a high generalization capacity. We present a study conducted using three state-of-the-art systems Doc-UFCN, dhSegment and ARU-Net and show that it is possible to build generic models trained on a wide variety of historical document datasets that can correctly segment diverse unseen pages. This paper also highlights the importance of the annotations used during training: each existing dataset is annotated differently. We present a unification of the annotations and show its positive impact on the final text recognition results. In this end, we present a complete evaluation strategy using standard pixel-level metrics, object-level ones and introducing goal-oriented metrics. | ['Thierry Paquet', 'Christopher Kermorvant', 'Mélodie Boillet'] | 2022-03-23 | null | null | null | null | ['line-detection'] | ['computer-vision'] | [ 2.26644412e-01 -2.29782715e-01 -2.88320575e-02 -3.09299082e-01
-5.47388911e-01 -9.94759381e-01 7.70967662e-01 2.46963829e-01
-3.44049603e-01 7.60597587e-01 4.40866686e-02 -2.80164331e-01
-2.06494167e-01 -5.73545992e-01 -4.94849056e-01 -3.12762111e-01
3.04206699e-01 6.84361219e-01 5.29576838e-01 -2.76500940e-01
7.04283059e-01 8.14972222e-01 -1.44853771e+00 6.15549088e-01
9.38928425e-01 7.39055157e-01 1.75934061e-01 1.14403033e+00
-6.01675630e-01 5.47570705e-01 -1.08124864e+00 -6.47935927e-01
1.80063009e-01 -4.69550520e-01 -1.00017619e+00 5.52880883e-01
9.70185935e-01 -3.65300745e-01 -1.97987288e-01 9.16966736e-01
4.00073647e-01 -5.67087531e-02 7.99812257e-01 -8.68863702e-01
-7.17198789e-01 7.53890336e-01 -4.56042349e-01 1.68175593e-01
3.02533388e-01 -2.15211898e-01 7.97547042e-01 -5.48719227e-01
9.49832082e-01 8.30860376e-01 6.64504588e-01 4.68136579e-01
-9.48694110e-01 1.22161373e-01 2.65200257e-01 6.48514628e-02
-9.16018426e-01 -1.02852836e-01 5.76819956e-01 -6.40303254e-01
8.41795206e-01 5.41029811e-01 3.59309286e-01 1.02965868e+00
-8.31286833e-02 1.21282494e+00 1.27880144e+00 -1.00165057e+00
5.75967059e-02 3.12064409e-01 6.84239328e-01 5.16833365e-01
2.93383181e-01 -4.92296010e-01 -3.76406670e-01 3.58232856e-01
6.76704466e-01 -4.49864507e-01 -4.81303215e-01 -4.61545348e-01
-1.00523233e+00 5.46963811e-01 9.09578055e-02 7.84957230e-01
1.68402895e-01 -2.00992882e-01 5.34956813e-01 -6.41215667e-02
1.62677720e-01 6.80169463e-01 -3.45979989e-01 -3.31339002e-01
-1.62952983e+00 1.06294699e-01 9.75258470e-01 1.06611395e+00
1.24168813e-01 -2.67043244e-02 -3.33223522e-01 9.24809694e-01
-3.75971012e-02 2.87786901e-01 6.01673484e-01 -6.06658399e-01
5.78783929e-01 4.97933805e-01 1.54369310e-01 -8.12932253e-01
-4.33315486e-01 -5.91081023e-01 -3.43037277e-01 2.90592909e-01
7.22193420e-01 -4.28754203e-02 -1.30732131e+00 8.76248658e-01
-1.67356819e-01 -7.94186413e-01 -1.35258242e-01 5.21656156e-01
4.60928738e-01 5.82039475e-01 -1.67006895e-01 8.47991630e-02
1.04327631e+00 -1.50183046e+00 -9.33351219e-01 -2.50114352e-01
6.82782054e-01 -1.18833780e+00 1.21987665e+00 9.59431469e-01
-9.74209011e-01 -7.25961030e-01 -1.34366357e+00 -1.18318863e-01
-9.57149267e-01 6.50965273e-01 4.82743114e-01 1.24896872e+00
-7.22459257e-01 6.93225861e-01 -5.28582156e-01 -7.62339830e-01
2.44771242e-01 2.14775845e-01 -1.64538831e-01 -8.07706043e-02
-6.17790043e-01 8.32315862e-01 7.63901949e-01 9.95068625e-02
-2.89657652e-01 -2.22939104e-01 -2.20500559e-01 9.16348919e-02
3.84976178e-01 -1.98524594e-01 1.25210905e+00 -9.13358271e-01
-1.27181160e+00 7.41599083e-01 2.11418271e-01 -4.48464781e-01
1.27648842e+00 -7.12793767e-01 -6.46648824e-01 1.88746542e-01
-2.23494023e-01 7.22230434e-01 4.86428648e-01 -1.64501131e+00
-9.19555843e-01 -3.70800287e-01 -1.83092147e-01 1.65966675e-01
-5.55590510e-01 -2.60193288e-01 -8.80220056e-01 -8.15311313e-01
-5.66994713e-04 -7.55802691e-01 2.45374948e-01 -2.06434235e-01
-7.08854496e-01 1.33893922e-01 1.07289314e+00 -1.01669657e+00
1.37679839e+00 -1.82800794e+00 -1.10109098e-01 2.14070678e-01
-3.08732808e-01 3.93830806e-01 9.32158008e-02 4.11544830e-01
2.73870945e-01 4.31593567e-01 -3.54038864e-01 -1.96300387e-01
1.32580260e-02 6.65125400e-02 -3.16625357e-01 3.83028090e-02
-2.72424281e-01 6.26403511e-01 -5.19041598e-01 -5.60880184e-01
3.73439252e-01 2.97650009e-01 1.62838265e-01 -2.17719704e-01
-5.04926145e-01 -9.04492140e-02 -1.61200762e-01 7.35489964e-01
6.37141466e-01 8.80502537e-02 1.22868516e-01 -2.65240759e-01
-4.06204790e-01 -1.87147960e-01 -1.29698682e+00 1.68820930e+00
-2.05952078e-01 1.20541596e+00 -4.22514498e-01 -5.19303739e-01
1.02912390e+00 5.50170280e-02 1.41927063e-01 -6.68198347e-01
2.19243348e-01 3.61486375e-01 -1.55264765e-01 -3.55218321e-01
1.17974329e+00 6.82593465e-01 1.73818722e-01 1.13232322e-01
2.08999519e-03 -3.06986660e-01 8.77164960e-01 8.17580447e-02
5.37230492e-01 4.24161881e-01 -9.06995311e-02 -2.62413710e-01
7.14734614e-01 5.31253994e-01 -6.85512722e-02 1.02506483e+00
-1.49672136e-01 1.08875787e+00 5.77125490e-01 -2.72959799e-01
-1.42139506e+00 -8.35603774e-01 -3.20414156e-01 8.80839765e-01
-2.65195891e-02 -2.99844950e-01 -1.18131042e+00 -7.30215311e-01
-2.98914015e-01 1.04583192e+00 -6.45820200e-01 4.04233187e-01
-6.21789098e-01 -8.58973205e-01 6.44975722e-01 7.31484294e-01
7.99088895e-01 -7.42146134e-01 -4.92991120e-01 1.02365501e-01
1.04895130e-01 -1.19586587e+00 -2.87084609e-01 3.27170968e-01
-1.06825781e+00 -8.82576346e-01 -9.82585430e-01 -7.87084699e-01
6.67601824e-01 1.47429779e-01 1.05410290e+00 -8.38402808e-02
-4.62868005e-01 4.74257767e-01 -5.74466467e-01 -4.18944657e-01
-4.46502119e-01 6.17956519e-01 -6.92417920e-01 -2.69685745e-01
2.85251588e-01 4.27735656e-01 -1.95389807e-01 3.16082716e-01
-1.13494158e+00 1.61844611e-01 5.39004087e-01 6.71954334e-01
4.39083874e-01 2.70901203e-01 -8.32284465e-02 -1.36289561e+00
9.16707456e-01 1.69495553e-01 -6.55724883e-01 8.16768110e-01
-7.64302075e-01 -1.75087538e-03 4.62196350e-01 -1.35121226e-01
-1.27826369e+00 1.51663810e-01 -1.40720364e-02 3.41375858e-01
-5.23619950e-01 2.27763504e-01 -2.19488636e-01 4.90119085e-02
9.01867330e-01 2.04959884e-01 -5.34739852e-01 -6.57607377e-01
5.22762418e-01 8.72401655e-01 7.87958562e-01 -5.08922100e-01
5.42481363e-01 3.99923652e-01 -2.73888171e-01 -1.20907569e+00
-7.99300671e-01 -5.51204145e-01 -1.29494536e+00 -4.76899177e-01
9.94780838e-01 -2.03592286e-01 8.33839923e-02 9.27560091e-01
-1.01209962e+00 -3.74293178e-01 -2.77223706e-01 1.64232567e-01
-2.87443787e-01 6.70017004e-01 -5.80990076e-01 -6.95013762e-01
-3.18069577e-01 -1.01676047e+00 7.28994608e-01 6.56828463e-01
-8.69943500e-02 -1.11583745e+00 2.58492306e-02 6.04408324e-01
1.52094364e-01 2.57885873e-01 9.52005863e-01 -7.11101413e-01
-4.48292524e-01 -4.50986922e-01 -3.08484495e-01 4.83922124e-01
3.88244204e-02 5.99152088e-01 -9.83100533e-01 -4.50177788e-04
-4.49984998e-01 -3.82201597e-02 9.08835530e-01 3.14511180e-01
1.20685375e+00 2.14541033e-01 -3.24994922e-01 2.27637351e-01
1.71029627e+00 4.46314991e-01 9.21082914e-01 9.94456470e-01
7.97086120e-01 6.38496399e-01 6.17494166e-01 1.05231009e-01
-2.67360032e-01 4.09986943e-01 1.00287631e-01 -1.77721664e-01
-3.29070240e-01 -7.77539015e-02 5.94885983e-02 4.02188182e-01
-5.17058894e-02 -8.79287183e-01 -1.11991906e+00 5.37365377e-01
-1.63706076e+00 -8.06820154e-01 -6.96014225e-01 2.08036518e+00
5.97968459e-01 7.40642905e-01 2.13445872e-01 5.44391215e-01
7.64597416e-01 1.44266326e-03 -2.01841563e-01 -8.23762059e-01
-6.67323470e-01 -5.74761666e-02 9.22728121e-01 5.42086720e-01
-1.22935653e+00 1.09049404e+00 6.85655642e+00 7.13481426e-01
-1.10568178e+00 -3.38071108e-01 8.02110612e-01 2.62353778e-01
-2.75715813e-03 -1.66560277e-01 -1.13715637e+00 2.67069250e-01
7.33506322e-01 1.71542153e-01 1.94557384e-02 9.13787663e-01
-7.01828003e-02 -4.17053759e-01 -9.49553072e-01 6.52279913e-01
5.86837709e-01 -1.30902112e+00 1.33356199e-01 -1.07230172e-01
1.02040565e+00 -2.17510492e-01 9.81842726e-03 -1.94366813e-01
6.86440170e-02 -9.42708194e-01 9.32933509e-01 6.05131567e-01
3.99962664e-01 -6.01279795e-01 6.14774525e-01 -5.26255518e-02
-5.80502272e-01 4.83596846e-02 -2.21679911e-01 5.51658094e-01
1.30846068e-01 5.37185252e-01 -8.21858227e-01 6.89218819e-01
5.64625025e-01 2.52712101e-01 -1.22510374e+00 1.49710774e+00
-2.97919631e-01 5.48529387e-01 -2.84441143e-01 -1.70371547e-01
4.38519269e-01 -2.64681041e-01 1.71008438e-01 1.72657120e+00
2.23877683e-01 -5.15790880e-01 8.71337019e-03 5.69604933e-01
4.61934395e-02 3.50266546e-01 -1.06623545e-01 -2.30313018e-01
7.61874691e-02 1.26091015e+00 -1.44022286e+00 -4.13676828e-01
-5.48724353e-01 1.44093573e+00 -8.68157670e-02 2.84080237e-01
-6.15779638e-01 -6.89464092e-01 -1.95172057e-01 -1.39637722e-03
3.34521443e-01 -4.60683554e-01 -7.18311608e-01 -9.04345334e-01
1.27284393e-01 -8.41882765e-01 3.88398230e-01 -9.19862151e-01
-9.23009157e-01 6.69045925e-01 -1.07560255e-01 -8.49005163e-01
5.11147082e-02 -1.14929366e+00 -4.37134504e-01 6.62880182e-01
-1.27434874e+00 -1.27463162e+00 -5.17512679e-01 1.00218937e-01
1.01676881e+00 -1.89987704e-01 5.28674245e-01 1.27290770e-01
-7.02510595e-01 5.26074946e-01 7.88749754e-01 4.13719207e-01
7.64561832e-01 -1.69120646e+00 4.36096668e-01 1.15411747e+00
4.45769697e-01 2.92916298e-01 8.52228045e-01 -6.68161631e-01
-7.34213471e-01 -6.45888150e-01 6.27836049e-01 -4.75586385e-01
4.41452682e-01 -4.05790627e-01 -9.06017780e-01 4.96063471e-01
2.39794075e-01 -5.80071926e-01 6.04245186e-01 -5.48905060e-02
-2.04561055e-01 6.10056482e-02 -8.93508911e-01 6.68420553e-01
6.54840350e-01 -2.36474380e-01 -5.85649014e-01 4.58808362e-01
4.86883745e-02 -4.33007538e-01 -7.47730076e-01 -1.36220187e-01
6.65014505e-01 -1.26434696e+00 6.49917305e-01 -4.08912808e-01
4.16850567e-01 -1.61560863e-01 1.42795444e-01 -9.80925441e-01
-6.42067418e-02 -2.25541621e-01 2.96605509e-02 1.63174355e+00
7.89019048e-01 -1.76094621e-01 1.06720471e+00 6.09180450e-01
-1.74235255e-01 -3.39647532e-01 -2.85755098e-01 -8.32411408e-01
2.95006931e-01 -1.82563573e-01 3.65223140e-01 8.50358248e-01
-2.19171226e-01 1.63556561e-01 -2.48031259e-01 -2.88780648e-02
3.25303376e-01 5.54423779e-02 7.22978771e-01 -1.27417123e+00
-7.12635219e-02 -7.09453821e-01 -3.03215474e-01 -1.03756821e+00
-4.44912612e-01 -4.75289971e-01 -1.10308468e-01 -2.00320101e+00
-1.14453182e-01 -1.20250553e-01 2.05045506e-01 1.41975611e-01
8.19931459e-03 1.06462047e-01 3.97868156e-01 3.01546723e-01
-3.25977474e-01 -1.84679300e-01 1.14634359e+00 -3.26452017e-01
-3.06712925e-01 -2.56630301e-01 -2.32215807e-01 6.97319090e-01
9.66823995e-01 -9.82745364e-02 -2.69472361e-01 -7.28394687e-01
1.03758588e-01 -5.29564261e-01 -4.77928296e-02 -1.43166614e+00
3.00803781e-01 2.07998529e-01 1.10506594e+00 -1.07303822e+00
2.11611707e-02 -7.52358317e-01 -3.42361361e-01 4.85521883e-01
-4.29510921e-01 1.08762607e-01 2.96416581e-01 2.62933105e-01
-9.21441466e-02 -9.89065170e-01 7.01284349e-01 -2.09633678e-01
-1.07159817e+00 -2.92454809e-01 -3.15067261e-01 -6.03694469e-02
9.96966839e-01 -7.80196607e-01 -5.42740166e-01 -2.89139301e-02
-6.80187583e-01 6.52570799e-02 6.90659404e-01 5.67434669e-01
2.34164391e-03 -6.40223324e-01 -4.79122818e-01 -2.72624880e-01
5.17588519e-02 -2.83042759e-01 1.08889729e-01 3.38515878e-01
-1.37450349e+00 7.76602745e-01 -6.06943250e-01 -4.36634690e-01
-1.35799313e+00 4.77154851e-01 4.12901938e-01 -3.21813017e-01
-4.57868516e-01 5.32773912e-01 -4.64354813e-01 -2.05325726e-02
3.48112792e-01 -3.12834322e-01 -4.30819571e-01 3.14267427e-01
3.36425662e-01 7.85100877e-01 5.12516856e-01 -3.95476252e-01
1.90310761e-01 6.34617209e-01 -5.15582681e-01 -3.30062449e-01
1.05029964e+00 -4.27464843e-02 2.26518497e-01 5.44195533e-01
7.01379180e-01 1.45429820e-01 -1.20476758e+00 2.67831266e-01
4.38971132e-01 -5.65047503e-01 1.44353732e-01 -1.33982491e+00
-8.65109086e-01 9.47389960e-01 9.22158837e-01 5.02321422e-01
9.28637683e-01 -4.17719752e-01 5.51296890e-01 4.48298544e-01
-1.38988659e-01 -1.95086753e+00 -7.34406859e-02 3.27875495e-01
6.47123158e-01 -1.09474266e+00 2.09362432e-01 -6.95306599e-01
-5.08923709e-01 1.82646024e+00 6.05377376e-01 3.66995364e-01
1.62768140e-02 2.44695365e-01 1.95972264e-01 3.47664580e-02
7.99534619e-02 -2.28803620e-01 3.51152182e-01 6.60948098e-01
7.78937757e-01 -1.67354584e-01 -7.09457934e-01 1.45915121e-01
-3.47013772e-01 -1.22245125e-01 8.60557139e-01 1.26176369e+00
-6.07126236e-01 -1.43192303e+00 -7.28281677e-01 5.14128983e-01
-4.32132870e-01 1.35843828e-01 -1.00730634e+00 1.18751764e+00
-2.16806121e-02 7.83139348e-01 3.05689704e-02 8.41745809e-02
4.11180675e-01 3.75683904e-01 6.57148361e-01 -2.52048403e-01
-5.14509022e-01 6.69106469e-02 2.61735648e-01 1.71397701e-01
-2.50194490e-01 -8.13543737e-01 -1.18524981e+00 -1.51723683e-01
-5.50697744e-01 1.23709202e-01 1.06211066e+00 7.38861382e-01
-1.11897849e-01 6.83454871e-01 4.67129014e-02 -4.02232021e-01
-4.61602300e-01 -9.13358331e-01 -5.14954746e-01 3.72742355e-01
-1.49831042e-01 -1.17058948e-01 -7.75218681e-02 6.12719357e-01] | [11.787986755371094, 2.6287829875946045] |
7912fbd5-775a-4fa0-a564-f1891e6b3d03 | privacy-preserving-tensor-factorization-for | 1908.09888 | null | https://arxiv.org/abs/1908.09888v2 | https://arxiv.org/pdf/1908.09888v2.pdf | Privacy-Preserving Tensor Factorization for Collaborative Health Data Analysis | Tensor factorization has been demonstrated as an efficient approach for computational phenotyping, where massive electronic health records (EHRs) are converted to concise and meaningful clinical concepts. While distributing the tensor factorization tasks to local sites can avoid direct data sharing, it still requires the exchange of intermediary results which could reveal sensitive patient information. Therefore, the challenge is how to jointly decompose the tensor under rigorous and principled privacy constraints, while still support the model's interpretability. We propose DPFact, a privacy-preserving collaborative tensor factorization method for computational phenotyping using EHR. It embeds advanced privacy-preserving mechanisms with collaborative learning. Hospitals can keep their EHR database private but also collaboratively learn meaningful clinical concepts by sharing differentially private intermediary results. Moreover, DPFact solves the heterogeneous patient population using a structured sparsity term. In our framework, each hospital decomposes its local tensors, and sends the updated intermediary results with output perturbation every several iterations to a semi-trusted server which generates the phenotypes. The evaluation on both real-world and synthetic datasets demonstrated that under strict privacy constraints, our method is more accurate and communication-efficient than state-of-the-art baseline methods. | ['Jian Lou', 'Xiaoqian Jiang', 'Jing Ma', 'Qiuchen Zhang', 'Li Xiong', 'Joyce C. Ho'] | 2019-08-26 | null | null | null | null | ['computational-phenotyping'] | ['medical'] | [-1.98141038e-01 2.24216923e-01 -1.39525384e-01 -4.36043620e-01
-8.13785851e-01 -9.39716935e-01 -3.71839076e-01 5.30265510e-01
-2.00983271e-01 7.13335872e-01 4.54531699e-01 -4.94272470e-01
-2.76169032e-01 -4.33979809e-01 -5.30313075e-01 -8.41813385e-01
-4.69050080e-01 6.24062061e-01 -6.27925217e-01 2.56552666e-01
-4.75977749e-01 2.36948416e-01 -5.64544141e-01 8.79961371e-01
7.06510365e-01 9.07187819e-01 -6.60221159e-01 5.04855931e-01
5.14015496e-01 9.95371222e-01 -2.48867139e-01 -1.01977253e+00
7.40152001e-01 1.21920124e-01 -9.41295028e-01 -1.87256128e-01
-1.59418613e-01 -5.23692429e-01 -4.07480419e-01 9.70385969e-01
5.40738881e-01 -4.18315262e-01 -2.05644071e-01 -1.54342103e+00
-7.94482648e-01 7.66872287e-01 -3.19087744e-01 -3.59775633e-01
3.99876535e-01 2.42219076e-01 1.02389443e+00 -3.81443232e-01
7.58859634e-01 7.90340662e-01 1.00223720e+00 5.45907855e-01
-1.61369479e+00 -6.72605872e-01 -4.65629846e-02 -2.40116388e-01
-1.41942704e+00 -1.73420057e-01 4.52205747e-01 -5.79197884e-01
4.57813114e-01 1.08616650e+00 7.28579581e-01 9.92332935e-01
2.19331622e-01 8.82543206e-01 8.97712231e-01 3.08000296e-01
4.37074572e-01 1.56369194e-01 1.03351310e-01 4.10988778e-01
6.51475191e-01 -6.56656250e-02 -6.44569695e-01 -1.68561125e+00
1.65299088e-01 8.50371182e-01 -6.58271313e-01 -4.67487544e-01
-1.71619642e+00 6.00436866e-01 3.47083807e-02 -1.19229481e-01
-6.22668624e-01 1.30476683e-01 6.90066338e-01 4.66578186e-01
5.72079182e-01 3.15142602e-01 -9.63046551e-01 2.49163732e-01
-6.80268109e-01 4.35597599e-01 9.56502259e-01 9.63235557e-01
5.00866830e-01 -6.69225991e-01 -4.92803812e-01 3.15254666e-02
3.46637428e-01 3.67512852e-01 1.50529444e-01 -8.59432459e-01
5.96946955e-01 7.27169216e-01 3.71470243e-01 -1.32223177e+00
-9.66885686e-02 -3.18569779e-01 -1.35758328e+00 -6.14884973e-01
2.54280001e-01 -5.60138643e-01 -1.44606292e-01 1.65356851e+00
9.31915700e-01 3.26932788e-01 1.82524621e-01 8.75581980e-01
3.88837099e-01 3.02732378e-01 8.58387500e-02 -4.59305525e-01
1.77951825e+00 -5.03450751e-01 -9.10966516e-01 6.32326841e-01
9.53729689e-01 -4.77062970e-01 3.53202224e-01 4.96290267e-01
-8.86988521e-01 4.16680545e-01 -4.53533411e-01 -2.10907713e-01
8.85673314e-02 1.40685931e-01 1.33933568e+00 8.55051517e-01
-1.04703069e+00 4.56600845e-01 -1.18740904e+00 8.54972601e-02
9.07737911e-01 4.74482536e-01 -8.87845516e-01 -2.54377186e-01
-9.62259531e-01 -2.60881364e-01 8.20888802e-02 1.43320814e-01
-9.01427388e-01 -1.44774342e+00 -5.43876529e-01 4.82988209e-02
2.96458960e-01 -1.53266120e+00 8.89206588e-01 -2.40089193e-01
-1.11596513e+00 6.74973428e-01 1.03251286e-01 -4.57273692e-01
9.02024090e-01 -5.08773439e-02 -2.48532832e-01 2.55604565e-01
1.26341328e-01 -1.76690891e-01 5.93458891e-01 -1.02250040e+00
-3.13667089e-01 -9.51217711e-01 -1.98039070e-01 -3.06082398e-01
-4.85693961e-01 3.60938236e-02 -2.40481287e-01 -7.32582033e-01
2.27151901e-01 -1.05240643e+00 -6.98497891e-01 3.72936994e-01
-8.93114805e-01 1.44155189e-01 6.86233521e-01 -9.85510349e-01
1.11001563e+00 -2.27127099e+00 1.27984777e-01 4.07335043e-01
1.08746743e+00 -1.02694243e-01 1.72139004e-01 6.57898486e-01
-7.72456229e-02 2.63987243e-01 -1.98440552e-01 -6.96872354e-01
-2.13830441e-01 3.66194099e-02 -5.55020034e-01 7.76681662e-01
-3.47120374e-01 1.01591992e+00 -1.01206517e+00 -1.71289220e-01
-4.38006669e-01 6.38094783e-01 -1.08780110e+00 1.89400181e-01
-9.19870213e-02 7.10213661e-01 -8.89432490e-01 8.96442115e-01
1.17168617e+00 -7.04484463e-01 1.07002687e+00 -2.34338015e-01
4.35022652e-01 -2.13172272e-01 -9.62388515e-01 1.87434423e+00
1.75615326e-01 -2.55951852e-01 4.44011897e-01 -5.43116212e-01
3.65434468e-01 7.68752813e-01 1.14278567e+00 4.85133559e-01
1.44058960e-02 2.66502704e-02 -4.32565242e-01 -4.63237405e-01
1.63193241e-01 -1.50932511e-02 -1.38202116e-01 9.20333862e-01
-4.22403753e-01 6.41534507e-01 -7.58373916e-01 5.80450475e-01
1.40675497e+00 -4.52639133e-01 1.23353146e-01 -2.14948311e-01
1.95544511e-01 -2.15032235e-01 1.11403739e+00 4.73233253e-01
-1.12792410e-01 3.74002278e-01 7.74004459e-01 -1.11553717e+00
-5.96905231e-01 -8.16792428e-01 2.90471558e-02 6.13696754e-01
-2.60479599e-01 -1.07452929e+00 -5.48364997e-01 -1.11356926e+00
7.91022003e-01 1.04913898e-02 -8.77352417e-01 1.28157064e-01
-6.88960403e-02 -1.17686987e+00 7.40348697e-01 1.13050267e-01
1.30682215e-01 -3.24997231e-02 -3.92054826e-01 1.72225595e-01
-5.26699126e-01 -9.21420634e-01 -8.42141986e-01 -3.03707540e-01
-9.32710409e-01 -1.31555068e+00 -3.72880101e-01 1.09654767e-02
1.04497004e+00 2.49575779e-01 7.39941478e-01 4.69723111e-03
-4.62503374e-01 3.48858833e-01 -1.62570104e-01 -2.66014367e-01
-7.25043640e-02 -3.09018850e-01 1.59970373e-01 8.14938009e-01
3.04809511e-01 -6.22226119e-01 -1.15681219e+00 8.50252360e-02
-1.20051134e+00 -9.61198956e-02 -2.51840930e-02 8.99030626e-01
7.68456757e-01 2.20432244e-02 8.02621618e-02 -1.50600863e+00
6.81606829e-01 -7.21223712e-01 -4.94411737e-01 5.02877355e-01
-7.37139046e-01 -6.68913946e-02 7.45447159e-01 -8.03082064e-02
-7.17140913e-01 2.60761380e-01 4.58303660e-01 -6.77521825e-01
3.62289459e-01 5.21553457e-01 -2.59181738e-01 -4.63088639e-02
3.83116513e-01 1.74535334e-01 1.57054678e-01 -6.71203077e-01
3.81215870e-01 7.06672490e-01 4.34934087e-02 -7.55214810e-01
5.94772995e-01 9.51913118e-01 9.74242836e-02 -9.88356024e-02
-4.27808642e-01 -4.62177992e-01 -1.91957787e-01 4.95535493e-01
5.84522903e-01 -1.35665035e+00 -1.67892146e+00 4.44421977e-01
-1.12491238e+00 2.67386705e-01 -3.67797583e-01 4.79078859e-01
-1.85532823e-01 8.49224746e-01 -9.15423810e-01 -6.30205870e-01
-7.71150231e-01 -8.99858475e-01 1.13090575e+00 -6.69237196e-01
-6.47932366e-02 -7.35980928e-01 2.83382267e-01 8.98075581e-01
3.41475189e-01 7.02274024e-01 9.08522725e-01 -6.86456859e-01
-8.90902877e-01 -4.76165086e-01 7.14321956e-02 8.06677788e-02
2.77523398e-01 -2.40022495e-01 -1.00434899e+00 -8.12919736e-01
2.64290690e-01 -2.18440101e-01 3.83082330e-01 -2.80500064e-03
1.72007191e+00 -1.21605015e+00 -6.02179646e-01 1.28486598e+00
1.29799271e+00 -2.79019743e-01 3.77702117e-01 -4.12667334e-01
8.54016185e-01 4.78138864e-01 4.07606721e-01 1.30327427e+00
7.53852844e-01 3.76823992e-01 1.76160127e-01 -1.25467986e-01
7.95282662e-01 -4.27812785e-01 -2.20595337e-02 8.60018134e-01
-8.05473402e-02 3.48989703e-02 -8.19477737e-01 3.41444194e-01
-2.31094646e+00 -5.63273787e-01 -1.25474483e-01 2.29552221e+00
1.14417195e+00 -1.05133986e+00 1.61467075e-01 -1.80291161e-01
3.29892844e-01 -1.35475039e-01 -6.45186901e-01 6.02824651e-02
4.68678363e-02 -1.67527825e-01 8.81340265e-01 -3.36171947e-02
-9.13035154e-01 3.41288328e-01 5.67311001e+00 2.35831395e-01
-1.08373559e+00 6.10838175e-01 1.15202332e+00 -5.34412324e-01
-8.17088306e-01 -7.52381375e-03 -1.88239723e-01 6.00321710e-01
8.16951811e-01 -4.97606814e-01 4.52568173e-01 1.04160213e+00
1.04362644e-01 4.26002294e-01 -1.39094925e+00 1.30900741e+00
-3.76983941e-01 -1.79161501e+00 -4.73495713e-03 4.53538477e-01
9.08410966e-01 4.39010188e-02 2.70699292e-01 -3.73741210e-01
7.18304992e-01 -7.68759906e-01 2.64585108e-01 6.15721345e-01
1.05141604e+00 -6.83041692e-01 7.10754633e-01 2.21780807e-01
-8.55256796e-01 -2.90908933e-01 -5.08605778e-01 2.32939497e-01
-8.45912546e-02 1.13220108e+00 -1.04883111e+00 1.13919508e+00
8.47415864e-01 6.49921775e-01 -2.80188203e-01 5.66216946e-01
1.90243751e-01 4.56877977e-01 -4.10341382e-01 3.79802018e-01
-2.98976004e-01 -2.49912530e-01 2.65339255e-01 8.54216576e-01
2.71736950e-01 4.25874501e-01 4.22917455e-01 7.11470366e-01
-4.82389182e-01 3.81853372e-01 -6.62492812e-01 -3.39766085e-01
5.39015234e-01 1.13576496e+00 3.86425517e-02 -2.34614283e-01
-2.98545621e-02 1.19976175e+00 2.25222394e-01 3.57247710e-01
-5.25407314e-01 1.48236491e-02 1.38166964e+00 2.05712900e-01
-3.26863211e-03 1.59197330e-01 -4.21172261e-01 -1.84558976e+00
4.47398752e-01 -1.25251079e+00 1.02626967e+00 -2.02566132e-01
-1.51891077e+00 4.55213994e-01 -6.22358263e-01 -1.17295122e+00
1.18360259e-01 -1.31050050e-01 -6.94371760e-02 7.67077744e-01
-1.23635650e+00 -1.47607887e+00 9.73186642e-02 1.26710916e+00
-8.35741997e-01 -7.48640150e-02 1.39666986e+00 3.87483329e-01
-8.41209888e-01 1.16915190e+00 4.27175492e-01 1.81768104e-01
6.00407183e-01 -9.58976924e-01 8.50743353e-02 7.41596639e-01
-2.99620658e-01 1.30125141e+00 3.40986192e-01 -8.57978642e-01
-2.28082705e+00 -1.54240513e+00 1.08121574e+00 -6.96087539e-01
5.50010860e-01 -6.63275719e-01 -7.07564771e-01 1.09714627e+00
-1.57301918e-01 7.73425519e-01 1.67935288e+00 1.28838480e-01
-8.40053976e-01 -4.94596452e-01 -1.84900081e+00 2.55566746e-01
8.75227988e-01 -9.33512330e-01 1.23978078e-01 9.59174812e-01
1.10549748e+00 -1.74269825e-01 -1.48827708e+00 2.08113883e-02
4.58703220e-01 -6.98148668e-01 8.32142234e-01 -1.31082308e+00
-1.84612197e-03 -5.96548498e-01 -1.75061330e-01 -9.10358667e-01
-4.28732991e-01 -1.59363580e+00 -2.88372457e-01 9.17317629e-01
3.33107054e-01 -1.05120265e+00 1.08436632e+00 1.39645505e+00
6.17425144e-01 -7.29717731e-01 -1.11432660e+00 -3.42025280e-01
-3.74793947e-01 -2.81798482e-01 1.28313994e+00 1.70045996e+00
4.11977708e-01 -3.47373366e-01 -7.91783571e-01 6.39916003e-01
9.04678583e-01 2.09187865e-01 9.40290034e-01 -1.00027227e+00
-7.50292778e-01 6.37027860e-01 -5.14268339e-01 -3.08967233e-01
-1.65674284e-01 -1.17832053e+00 -7.21502125e-01 -9.04014528e-01
5.31973660e-01 -7.75154054e-01 -4.46551621e-01 9.44785953e-01
-2.69425064e-01 -9.86191928e-02 7.59929791e-02 4.49269533e-01
-6.13059044e-01 3.00754577e-01 1.04442132e+00 -1.38618246e-01
-7.93638639e-03 3.42866741e-02 -1.30591738e+00 9.55454633e-03
5.13484061e-01 -6.87203288e-01 -6.01696253e-01 -4.02214527e-01
7.31672347e-01 5.26667237e-01 3.42672437e-01 -2.39956126e-01
5.03962517e-01 -2.78684437e-01 5.99592403e-02 -3.72753829e-01
6.93396607e-04 -1.17100883e+00 1.05188239e+00 7.68316031e-01
-2.97541767e-01 7.12999776e-02 -2.25592554e-01 1.00054765e+00
-1.71361715e-02 8.07776570e-01 1.19344316e-01 -1.10663392e-01
4.90456253e-01 9.86993849e-01 8.70718434e-02 -9.55071673e-02
1.41195500e+00 3.14924777e-01 -2.08120704e-01 -1.78440556e-01
-8.83940935e-01 5.18174171e-01 6.84564710e-01 -1.12878568e-01
6.69314384e-01 -1.12705481e+00 -1.01830685e+00 5.11778712e-01
3.05037081e-01 3.12524378e-01 7.96821058e-01 1.15314651e+00
-4.45760876e-01 3.99199516e-01 2.05888405e-01 -4.72935677e-01
-1.33693194e+00 1.13831508e+00 3.44645418e-02 -4.15130883e-01
-7.23681629e-01 9.31594968e-01 2.80398905e-01 -6.87837899e-01
1.64570242e-01 -4.86107498e-01 6.26653075e-01 -6.11081831e-02
1.00288355e+00 2.73188591e-01 2.66448289e-01 -6.13184161e-02
-5.35282969e-01 -1.28526121e-01 -3.34861547e-01 3.99986386e-01
1.64264882e+00 -4.65217531e-02 -8.42594922e-01 -2.20556617e-01
1.35521889e+00 2.71542430e-01 -7.89497077e-01 -1.85106263e-01
-2.39253893e-01 -1.01746356e+00 -1.42989591e-01 -8.26476097e-01
-1.51364446e+00 2.76662946e-01 1.55640155e-01 -5.68033420e-02
1.16568363e+00 -3.63823742e-01 9.74803448e-01 4.53987956e-01
8.39111924e-01 -3.76900434e-01 -7.21063077e-01 -3.33005518e-01
6.14885092e-01 -8.44135046e-01 1.02836773e-01 -8.09292793e-01
-8.90215099e-01 6.83162451e-01 -2.14616537e-01 3.82975280e-01
1.09170020e+00 3.45647246e-01 8.13999698e-02 -2.79713154e-01
-1.31769252e+00 1.06169367e+00 -3.52740109e-01 6.66690350e-01
8.94469321e-02 8.37087035e-01 -2.09030673e-01 1.10405767e+00
-1.50170242e-02 5.61012566e-01 3.33157480e-01 9.19617474e-01
8.25850129e-01 -1.41527402e+00 -3.89813423e-01 5.36159277e-01
-9.62873518e-01 -1.80273160e-01 -3.28326344e-01 -1.86874032e-01
1.71814322e-01 6.61302984e-01 -4.45487052e-01 -4.90007102e-01
4.96498495e-02 -1.87707897e-02 -7.34241828e-02 -4.71669197e-01
-1.19961047e+00 -2.09003314e-01 -6.11430109e-02 -1.11595130e+00
1.13782965e-01 -8.94262969e-01 -8.65647912e-01 -7.62178719e-01
1.07827894e-01 5.24122119e-01 4.26047623e-01 3.05482090e-01
1.35064507e+00 8.25179159e-04 1.11834157e+00 2.61529416e-01
-9.72310543e-01 -1.30876914e-01 -8.66225958e-01 4.89734501e-01
4.83765692e-01 2.36762181e-01 -2.20392987e-01 1.39882922e-01] | [6.152403831481934, 6.431461334228516] |
cf05e287-158d-4836-a5c0-237ffc13a613 | s-2-contact-graph-based-network-for-3d-hand | 2208.00874 | null | https://arxiv.org/abs/2208.00874v1 | https://arxiv.org/pdf/2208.00874v1.pdf | S$^2$Contact: Graph-based Network for 3D Hand-Object Contact Estimation with Semi-Supervised Learning | Despite the recent efforts in accurate 3D annotations in hand and object datasets, there still exist gaps in 3D hand and object reconstructions. Existing works leverage contact maps to refine inaccurate hand-object pose estimations and generate grasps given object models. However, they require explicit 3D supervision which is seldom available and therefore, are limited to constrained settings, e.g., where thermal cameras observe residual heat left on manipulated objects. In this paper, we propose a novel semi-supervised framework that allows us to learn contact from monocular images. Specifically, we leverage visual and geometric consistency constraints in large-scale datasets for generating pseudo-labels in semi-supervised learning and propose an efficient graph-based network to infer contact. Our semi-supervised learning framework achieves a favourable improvement over the existing supervised learning methods trained on data with `limited' annotations. Notably, our proposed model is able to achieve superior results with less than half the network parameters and memory access cost when compared with the commonly-used PointNet-based approach. We show benefits from using a contact map that rules hand-object interactions to produce more accurate reconstructions. We further demonstrate that training with pseudo-labels can extend contact map estimations to out-of-domain objects and generalise better across multiple datasets. | ['Hyung Jin Chang', 'Feng Zheng', 'Ales Leonardis', 'Kwang In Kim', 'Zhongqun Zhang', 'Tze Ho Elden Tse'] | 2022-08-01 | null | null | null | null | ['hand-object-pose'] | ['computer-vision'] | [ 1.36355445e-01 2.03896105e-01 -3.79008561e-01 -3.40584964e-01
-5.57206571e-01 -6.65480614e-01 3.89517933e-01 -2.75765806e-01
1.38903568e-02 5.28519869e-01 -1.14233501e-01 7.48645933e-03
-1.37155890e-01 -4.67995018e-01 -1.27123570e+00 -3.80460948e-01
2.82638252e-01 1.07767260e+00 3.04828405e-01 4.38168123e-02
1.42059878e-01 7.73875237e-01 -1.34687889e+00 -3.26091168e-03
6.83104336e-01 9.02973354e-01 5.96342802e-01 4.13327396e-01
8.85839909e-02 5.29419065e-01 -2.43598521e-01 -2.31462389e-01
5.16967297e-01 -8.25360045e-02 -8.98273408e-01 3.35882545e-01
6.59052610e-01 -7.39355564e-01 -4.30616707e-01 7.20615029e-01
4.74976033e-01 -1.52769789e-01 6.86107457e-01 -1.35909116e+00
-6.32866144e-01 4.22344059e-01 -6.26814306e-01 -5.65828621e-01
5.56371510e-01 2.52334177e-01 8.21870685e-01 -9.35725689e-01
1.11464393e+00 1.32894957e+00 6.51842952e-01 7.25551188e-01
-1.61517859e+00 -4.37339127e-01 1.91968575e-01 1.65815681e-01
-1.30489218e+00 -2.22804427e-01 1.05725884e+00 -4.23937380e-01
1.00620389e+00 1.36571247e-02 6.57573700e-01 1.45805478e+00
-7.94809461e-02 1.01054883e+00 1.13281155e+00 -5.38582921e-01
1.75890580e-01 -1.92624070e-02 -3.36215317e-01 7.53800035e-01
1.23831928e-01 1.26519978e-01 -5.68237543e-01 -8.04348215e-02
1.47156072e+00 2.20131904e-01 -2.67198026e-01 -1.27375889e+00
-1.40386569e+00 4.38671768e-01 7.24125266e-01 -7.80332536e-02
-4.20498967e-01 4.35545146e-01 8.37689713e-02 -8.61880034e-02
5.54606020e-01 3.43543977e-01 -5.26443839e-01 -1.80381741e-02
-8.29482853e-01 3.23070079e-01 9.11232412e-01 1.66879117e+00
7.27560580e-01 -1.87349394e-01 1.42583013e-01 6.55965269e-01
3.83140534e-01 6.22970998e-01 -2.72299945e-01 -1.27136850e+00
6.62335217e-01 6.21705055e-01 4.14165974e-01 -8.24643135e-01
-3.50528479e-01 -8.20375904e-02 -6.43445671e-01 5.61825991e-01
5.16946197e-01 2.83016413e-01 -1.12947786e+00 1.46690702e+00
4.23813432e-01 1.71390455e-02 -3.37399364e-01 1.20421863e+00
5.19857109e-01 1.24009043e-01 -1.68186247e-01 1.08681634e-01
8.35023761e-01 -1.20091987e+00 -6.99117482e-01 -1.64885670e-01
3.29444677e-01 -7.28019536e-01 1.09405661e+00 5.20196021e-01
-1.18951845e+00 -4.98398691e-01 -9.37359571e-01 -2.14308411e-01
-1.61283612e-01 2.57418931e-01 7.17807412e-01 1.49599746e-01
-9.08250153e-01 9.43297029e-01 -1.27985144e+00 -3.40045094e-01
6.42028809e-01 4.89509553e-01 -4.81865793e-01 -2.11312771e-01
-4.19785976e-01 1.10105956e+00 4.27369833e-01 3.36251348e-01
-1.02876878e+00 -6.11812830e-01 -8.14263582e-01 -4.62069452e-01
8.39938641e-01 -8.05471599e-01 1.10236025e+00 -4.41607714e-01
-1.70431876e+00 8.53170455e-01 6.23008832e-02 -1.39730036e-01
9.51565623e-01 -5.17031670e-01 5.35863876e-01 3.28031033e-01
-1.75350577e-01 9.59666193e-01 9.16652441e-01 -1.81377065e+00
-5.35768224e-03 -6.04492009e-01 3.10872823e-01 2.20440149e-01
2.99632121e-02 -2.84692734e-01 -5.49585581e-01 -5.70861161e-01
4.19186115e-01 -1.19553614e+00 -2.24306375e-01 7.11319923e-01
-6.48624301e-01 -3.29810649e-01 1.03449893e+00 -7.41411865e-01
2.33217180e-01 -1.68115151e+00 6.59223855e-01 1.23611256e-01
2.31532201e-01 2.69141763e-01 -3.14493366e-02 4.00186390e-01
3.14990520e-01 -1.87378511e-01 -1.55764326e-01 -6.60216331e-01
2.73439199e-01 5.02090216e-01 -2.51456976e-01 5.54660022e-01
2.15424120e-01 1.17540729e+00 -8.78004670e-01 -6.62323833e-01
6.43906534e-01 7.07116127e-01 -5.90285540e-01 5.14634788e-01
-7.37461448e-01 8.46827269e-01 -4.10043448e-01 8.45892072e-01
7.71303773e-01 -3.30389053e-01 1.35803521e-01 -4.60479796e-01
2.03964740e-01 2.36146823e-02 -1.12997949e+00 2.48848367e+00
-5.53648472e-01 1.46196008e-01 3.40605259e-01 -8.90284061e-01
9.94542420e-01 3.52946490e-01 4.48523909e-01 -6.31099790e-02
2.42121011e-01 3.09491277e-01 -4.82426971e-01 -4.49300975e-01
1.23825252e-01 4.57275286e-02 3.29660147e-01 5.19091368e-01
3.11955839e-01 -7.86861837e-01 -2.91212261e-01 1.99065283e-02
7.41301775e-01 1.29765880e+00 -1.22777104e-01 3.29889134e-02
7.27906898e-02 8.18197206e-02 1.06344476e-01 4.78886992e-01
1.14659667e-01 9.24206138e-01 1.67558894e-01 -3.88133615e-01
-1.34801602e+00 -1.03356886e+00 -8.63936469e-02 8.83789718e-01
5.00015914e-01 -2.23404646e-01 -6.51973248e-01 -7.79946864e-01
3.50358754e-01 4.99493122e-01 -3.86585563e-01 1.33460313e-01
-7.65937865e-01 -1.14824578e-01 1.66359633e-01 1.09483063e+00
2.82055467e-01 -1.08540082e+00 -6.35074139e-01 3.81902382e-02
-2.17953958e-02 -1.39979947e+00 -2.63373941e-01 4.23985943e-02
-1.12273574e+00 -1.13594532e+00 -9.75757062e-01 -7.30833709e-01
1.01642036e+00 5.97934499e-02 9.29440856e-01 -7.22414628e-02
-4.91735071e-01 6.31168604e-01 -3.81888330e-01 -3.44012022e-01
-2.67789662e-01 2.18763068e-01 1.45088524e-01 -4.25785333e-01
6.53240159e-02 -8.13382268e-01 -6.74611270e-01 4.54590917e-01
-6.08341873e-01 3.18818629e-01 6.74867988e-01 7.39178479e-01
5.79606235e-01 -6.21989906e-01 2.13105634e-01 -5.31563759e-01
1.61494389e-01 -1.54614272e-02 -5.41655004e-01 2.72759289e-01
-5.34275413e-01 1.87879533e-01 3.56117666e-01 -5.44505298e-01
-1.20954800e+00 6.22255981e-01 2.02242866e-01 -1.04863012e+00
-3.48076314e-01 -5.15022986e-02 -1.55885085e-01 -3.76675159e-01
6.36178315e-01 -1.19343638e-01 1.84286252e-01 -8.40349376e-01
5.90551496e-01 5.40556550e-01 5.78732610e-01 -9.00385678e-01
9.47909594e-01 6.40570283e-01 2.51053214e-01 -3.15161437e-01
-7.98372328e-01 -3.10129255e-01 -1.43496621e+00 -1.78996995e-01
7.71854162e-01 -6.82378948e-01 -9.42221701e-01 4.25749302e-01
-1.41517901e+00 -6.11052036e-01 -2.18411043e-01 5.48534930e-01
-1.13147140e+00 5.07386088e-01 -6.89327180e-01 -9.51990426e-01
-3.12230766e-01 -1.09045863e+00 1.57886791e+00 -2.56044179e-01
-1.20376624e-01 -8.90213788e-01 -2.94561058e-01 4.62767690e-01
9.54665318e-02 5.02756178e-01 4.94578689e-01 -1.95268050e-01
-1.03139770e+00 -1.57862023e-01 -3.35801214e-01 3.78951132e-01
2.84153730e-01 -3.21048200e-01 -8.55298281e-01 -4.52035338e-01
-2.27023467e-01 -8.13334823e-01 5.15861332e-01 1.34115353e-01
1.30205381e+00 -1.26733139e-01 -5.85488141e-01 5.06022274e-01
1.24328756e+00 -3.26209396e-01 2.79326141e-01 7.26949647e-02
1.23921597e+00 7.39368677e-01 8.28126669e-01 3.48204106e-01
2.57393032e-01 1.06876624e+00 7.50157237e-01 4.40329779e-03
-3.09163600e-01 -5.61712027e-01 -1.54705215e-02 6.43258393e-01
-7.13511288e-01 1.08987525e-01 -9.04352427e-01 5.06411552e-01
-2.01086879e+00 -3.79184008e-01 2.10186448e-02 2.12160516e+00
1.07909143e+00 1.36914721e-03 6.21958785e-02 3.80521342e-02
5.09753227e-01 -1.50691541e-02 -8.29046726e-01 2.68902451e-01
1.73399404e-01 2.58280337e-01 4.83888656e-01 4.82620716e-01
-8.44744265e-01 1.10925221e+00 5.92270994e+00 4.34077680e-01
-7.90713310e-01 1.40995279e-01 -7.94908330e-02 -1.36243001e-01
-4.95990664e-02 1.18549755e-02 -5.70777357e-01 -9.68044400e-02
1.31929070e-01 4.23069865e-01 6.52547956e-01 1.02041519e+00
-3.49025615e-02 3.63841956e-03 -1.72943819e+00 1.13138938e+00
3.42904717e-01 -9.49666202e-01 -1.01828642e-01 8.31106752e-02
7.81163216e-01 -1.94863141e-01 -2.29220316e-01 -2.19200984e-01
4.82518137e-01 -8.65284145e-01 8.85425746e-01 6.28444910e-01
9.12109971e-01 -3.75744373e-01 4.71042573e-01 6.28196478e-01
-9.95672226e-01 2.27850348e-01 -3.60865444e-01 1.72005929e-02
4.33547199e-01 2.29747966e-01 -1.04202378e+00 7.43079960e-01
7.78384864e-01 7.53617465e-01 -2.71187216e-01 7.72687972e-01
-4.14124489e-01 1.16373606e-01 -5.54874182e-01 1.39215142e-01
-4.69216295e-02 -1.27450123e-01 4.87791657e-01 7.78399944e-01
5.57903275e-02 2.17183325e-02 4.09025192e-01 1.34076512e+00
5.99231534e-02 -2.78388530e-01 -6.35614395e-01 1.53679371e-01
2.66421795e-01 1.09179711e+00 -7.60513604e-01 -2.38016531e-01
-1.85949281e-02 1.29982781e+00 6.17023766e-01 3.53129447e-01
-6.50813162e-01 -5.43923303e-02 1.12650774e-01 3.08498532e-01
4.21576768e-01 -7.33296454e-01 -3.33634317e-01 -1.16134584e+00
4.38530713e-01 -3.67589980e-01 -2.54543751e-01 -1.09501445e+00
-1.36689782e+00 2.74675936e-01 4.93084431e-01 -1.04213262e+00
-3.62861007e-01 -9.37182069e-01 9.45619866e-02 7.63173997e-01
-1.40698719e+00 -1.72705030e+00 -6.64285898e-01 5.77934980e-01
5.44280887e-01 2.58855790e-01 7.68832564e-01 -1.21110059e-01
5.49712405e-02 3.19480807e-01 -2.81566709e-01 -8.40943400e-03
7.06186056e-01 -1.38772762e+00 3.74793142e-01 2.38785088e-01
2.38141105e-01 5.57096839e-01 4.72871721e-01 -8.91396463e-01
-1.82093108e+00 -8.79259169e-01 2.76641399e-01 -8.96181941e-01
2.80228674e-01 -7.69683659e-01 -7.29206026e-01 1.08950865e+00
-1.17155761e-01 3.91360551e-01 -8.70780796e-02 -7.77104348e-02
-3.99805456e-01 1.37820199e-01 -1.24403799e+00 4.96986538e-01
1.85772765e+00 -5.92530608e-01 -6.26030266e-01 6.61192238e-01
6.07922852e-01 -1.05683780e+00 -1.02863550e+00 5.48491478e-01
7.18096197e-01 -6.79167688e-01 1.19125426e+00 -5.70199311e-01
4.30416673e-01 -1.53803870e-01 -3.07644568e-02 -1.17576671e+00
-4.14848104e-02 -7.20953166e-01 -5.53312659e-01 1.01540196e+00
-6.56351307e-03 -3.89070868e-01 1.01496494e+00 7.81158566e-01
-9.24199522e-02 -7.69169450e-01 -6.89276218e-01 -9.30741906e-01
-9.70477387e-02 -3.78592640e-01 4.69720453e-01 6.65999234e-01
-9.25981402e-02 -3.27768922e-02 -4.89721149e-01 1.51053101e-01
1.01757050e+00 3.01862538e-01 1.06189477e+00 -1.42504025e+00
-2.18100071e-01 6.34383708e-02 -3.40468943e-01 -1.45266116e+00
4.70364273e-01 -9.03922379e-01 3.97289008e-01 -1.69459915e+00
2.49547169e-01 -7.50651419e-01 2.19411835e-01 7.43138015e-01
1.78298682e-01 4.78318602e-01 3.75935465e-01 5.22773385e-01
-5.01975358e-01 6.14942491e-01 1.77323234e+00 2.26469710e-02
-1.35445654e-01 -6.15505911e-02 6.63207993e-02 8.25382829e-01
4.93172497e-01 -3.10059160e-01 -3.44859064e-01 -7.22785652e-01
-9.02447999e-02 -2.04235897e-03 9.44077492e-01 -8.11295271e-01
1.71464756e-01 -2.38421917e-01 4.26984370e-01 -7.11562812e-01
6.42598331e-01 -1.20068097e+00 2.50835687e-01 1.97919577e-01
-2.25277767e-01 -2.93479592e-01 2.13314798e-02 7.51218140e-01
3.65196735e-01 -1.97193995e-01 2.11077943e-01 -4.24394131e-01
-2.85486549e-01 4.53606337e-01 3.71630877e-01 -3.64915907e-01
1.07955158e+00 -3.51817667e-01 4.76232395e-02 -3.23124200e-01
-8.45073044e-01 3.33372280e-02 8.16744268e-01 6.90546870e-01
6.88161314e-01 -1.43727362e+00 -4.63240057e-01 1.42511621e-01
2.22125817e-02 7.99099922e-01 -1.05346143e-01 6.66756988e-01
-6.19739354e-01 5.10008335e-01 -1.83137834e-01 -1.16725838e+00
-1.10681438e+00 5.72614253e-01 6.60084262e-02 1.10053353e-01
-8.94585609e-01 5.97297609e-01 -3.02509759e-02 -8.82775903e-01
5.19025028e-01 -5.45073390e-01 4.13540900e-01 -5.79870820e-01
-7.57501870e-02 5.24878860e-01 -2.20050616e-03 -6.27627969e-01
-1.96032792e-01 9.58987236e-01 1.07679535e-02 -1.22953348e-01
1.57170427e+00 1.49158880e-01 -8.70329812e-02 4.29740816e-01
9.76416171e-01 -3.67411405e-01 -1.94923532e+00 -3.90617192e-01
-1.52776316e-01 -6.51155829e-01 -1.35475755e-01 -8.94720912e-01
-1.00478446e+00 9.69046772e-01 3.21413726e-01 -3.40985596e-01
6.54527545e-01 5.52632272e-01 6.53046846e-01 6.22401834e-01
1.08622658e+00 -1.00171804e+00 4.14649934e-01 1.01589516e-01
1.41422904e+00 -1.27056909e+00 2.25227594e-01 -1.02415991e+00
-4.23619419e-01 1.10263932e+00 8.04211915e-01 -2.95020580e-01
4.16572988e-01 3.31463814e-01 -1.15426958e-01 -3.85594636e-01
-2.58204252e-01 1.06542639e-01 4.01503563e-01 8.87255371e-01
4.53732064e-04 -1.28916474e-02 1.77057952e-01 1.29701868e-01
-1.50106475e-01 3.67097646e-01 -8.64872430e-03 1.33317530e+00
8.62512141e-02 -1.15776670e+00 -3.50115657e-01 2.41837531e-01
6.59035146e-03 4.00424570e-01 -4.91226017e-01 7.40173519e-01
-7.58668780e-02 5.98740697e-01 -1.25934854e-02 -1.71782792e-01
4.99297172e-01 3.28972675e-02 1.34642243e+00 -7.07292974e-01
-1.26584396e-01 -3.00661544e-03 -1.58783495e-01 -8.88926625e-01
-9.16945815e-01 -4.54370230e-01 -1.15582728e+00 -8.55914056e-02
-7.47183204e-01 -4.75344598e-01 7.82144070e-01 1.04857206e+00
4.60244298e-01 2.17207700e-01 2.56715894e-01 -1.84722304e+00
-9.19427931e-01 -1.11561072e+00 -4.09015298e-01 6.37036622e-01
1.55605420e-01 -1.23888874e+00 -1.76553145e-01 2.89130181e-01] | [6.208086967468262, -1.0374020338058472] |
5ccb2f7a-c982-42dc-a612-d2c8e87df092 | industrial-anomaly-detection-with-domain | 2304.02216 | null | https://arxiv.org/abs/2304.02216v1 | https://arxiv.org/pdf/2304.02216v1.pdf | Industrial Anomaly Detection with Domain Shift: A Real-world Dataset and Masked Multi-scale Reconstruction | Industrial anomaly detection (IAD) is crucial for automating industrial quality inspection. The diversity of the datasets is the foundation for developing comprehensive IAD algorithms. Existing IAD datasets focus on the diversity of data categories, overlooking the diversity of domains within the same data category. In this paper, to bridge this gap, we propose the Aero-engine Blade Anomaly Detection (AeBAD) dataset, consisting of two sub-datasets: the single-blade dataset and the video anomaly detection dataset of blades. Compared to existing datasets, AeBAD has the following two characteristics: 1.) The target samples are not aligned and at different scales. 2.) There is a domain shift between the distribution of normal samples in the test set and the training set, where the domain shifts are mainly caused by the changes in illumination and view. Based on this dataset, we observe that current state-of-the-art (SOTA) IAD methods exhibit limitations when the domain of normal samples in the test set undergoes a shift. To address this issue, we propose a novel method called masked multi-scale reconstruction (MMR), which enhances the model's capacity to deduce causality among patches in normal samples by a masked reconstruction task. MMR achieves superior performance compared to SOTA methods on the AeBAD dataset. Furthermore, MMR achieves competitive performance with SOTA methods to detect the anomalies of different types on the MVTec AD dataset. Code and dataset are available at https://github.com/zhangzilongc/MMR. | ['Xuefeng Chen', 'Chuang Sun', 'Xingwu Zhang', 'Zhibin Zhao', 'Zilong Zhang'] | 2023-04-05 | null | null | null | null | ['video-anomaly-detection'] | ['computer-vision'] | [-7.87906498e-02 -5.97172737e-01 3.89718294e-01 -7.57457092e-02
-4.07465756e-01 -5.74433744e-01 4.44384515e-01 -4.52947281e-02
4.26389515e-01 2.16543525e-01 -2.26333112e-01 -3.51956367e-01
-2.28461683e-01 -7.49371529e-01 -3.59880894e-01 -1.06481504e+00
4.68361042e-02 8.08088407e-02 3.63331735e-01 -3.09308052e-01
1.75759986e-01 6.80787325e-01 -1.66551256e+00 4.19554353e-01
8.10015261e-01 1.27355707e+00 1.59451604e-01 4.36062664e-01
-1.03231380e-03 3.84356111e-01 -1.04143536e+00 9.54249054e-02
6.01393998e-01 -6.30306721e-01 -3.98054481e-01 4.63467270e-01
3.95795316e-01 -2.14541972e-01 -4.27742720e-01 1.18568540e+00
4.76653516e-01 1.09649694e-03 6.29410982e-01 -1.55170250e+00
-5.00454068e-01 -1.68169782e-01 -9.08405781e-01 5.76273918e-01
2.20203891e-01 2.38284886e-01 7.66187668e-01 -1.10540831e+00
4.07185882e-01 9.45908904e-01 5.09677768e-01 2.73794740e-01
-9.31606114e-01 -4.49153632e-01 3.59843522e-01 5.69891274e-01
-1.20287097e+00 9.67105031e-02 1.28470850e+00 -6.58546209e-01
5.79269350e-01 3.80403996e-01 5.33126473e-01 1.30772126e+00
3.12839419e-01 5.47775328e-01 1.06319630e+00 -2.31465146e-01
3.59684885e-01 -2.88907886e-01 -1.80923939e-02 3.98913503e-01
4.56144005e-01 8.78371671e-02 -5.41719973e-01 -2.20393002e-01
7.85789490e-01 1.50328398e-01 -3.00389230e-01 -1.84455261e-01
-1.00831044e+00 6.33107781e-01 -1.14419675e-02 2.40230069e-01
-6.05850101e-01 -7.57039011e-01 6.34459555e-01 6.31714761e-01
5.28995693e-01 4.61867839e-01 -7.15441108e-01 3.90767294e-04
-4.07651424e-01 2.83437371e-01 6.05056524e-01 7.29433477e-01
5.49952447e-01 4.17357087e-01 8.29742700e-02 8.34099054e-01
2.37880155e-01 5.35385966e-01 6.26407862e-01 -5.92394412e-01
4.10293102e-01 7.85674274e-01 -1.16225213e-01 -1.22131073e+00
-3.58249664e-01 -4.06584710e-01 -9.30806100e-01 6.08608305e-01
6.92244351e-01 -5.93353882e-02 -6.24743819e-01 1.10494041e+00
6.11754954e-01 2.83437937e-01 1.47101745e-01 1.02153301e+00
6.95502222e-01 6.02417886e-01 -5.50094843e-01 -2.30472773e-01
1.33277547e+00 -7.62543380e-01 -8.33902121e-01 -9.73507091e-02
5.35506666e-01 -1.01339388e+00 1.16546202e+00 8.41209650e-01
-5.77210963e-01 -7.72511721e-01 -1.12330306e+00 6.44848228e-01
-3.97274047e-01 1.45763561e-01 1.11216471e-01 2.85804808e-01
-4.18569505e-01 3.62769663e-01 -6.36962235e-01 -3.27888817e-01
2.30375022e-01 -2.69546062e-01 -3.37642968e-01 -2.42683485e-01
-9.14624631e-01 5.38928628e-01 1.10699363e-01 2.67505169e-01
-9.77798998e-01 -7.59824693e-01 -7.01075375e-01 -3.92670393e-01
5.33046484e-01 -2.09106147e-01 9.98901069e-01 -9.66909528e-01
-1.13167334e+00 6.45837188e-01 1.15872219e-01 -1.28737137e-01
4.30465549e-01 -3.73750448e-01 -1.09268034e+00 1.68104082e-01
7.58257583e-02 -4.22001332e-01 1.13688612e+00 -1.30344474e+00
-8.04193616e-01 -6.03234887e-01 -2.64478445e-01 -2.48477057e-01
-1.31817028e-01 1.00299669e-02 -1.13512479e-01 -1.14189649e+00
2.35280216e-01 -8.11255634e-01 5.22820540e-02 -2.08809406e-01
-6.11693144e-01 -4.29525226e-01 1.63110006e+00 -7.86327839e-01
1.25315928e+00 -2.65836167e+00 -6.22498579e-02 2.12597221e-01
7.78976753e-02 2.76118249e-01 -1.26059562e-01 4.12450105e-01
-4.70923632e-01 -3.83264720e-01 -3.10854226e-01 4.09669168e-02
-6.30919710e-02 3.92544478e-01 -6.14861190e-01 6.12123549e-01
4.53112394e-01 1.19012550e-01 -7.28437424e-01 -5.18803559e-02
3.35730135e-01 -1.28102422e-01 -1.93897456e-01 4.71243590e-01
-9.29255225e-03 7.34352052e-01 -4.79944468e-01 1.11193895e+00
9.26200330e-01 1.73212662e-01 -1.37834713e-01 -4.18373913e-01
-2.38006979e-01 -6.11734316e-02 -1.46852732e+00 1.41676104e+00
1.02723632e-02 4.51948971e-01 1.34579495e-01 -1.13016379e+00
1.20344162e+00 4.56129879e-01 7.86919475e-01 -6.35937393e-01
-1.20173015e-01 4.32471097e-01 8.93129706e-02 -6.90990269e-01
1.80420846e-01 9.13449451e-02 3.36271636e-02 6.21983074e-02
-1.11244224e-01 -5.40038571e-02 2.04743788e-01 -1.24384224e-01
1.37826121e+00 1.56224556e-02 1.69122949e-01 -3.78248930e-01
5.79717994e-01 1.01162486e-01 1.08098960e+00 4.32663023e-01
-3.67390662e-01 7.78029561e-01 6.96781576e-01 -7.53457129e-01
-9.56089079e-01 -1.27900493e+00 -2.78006613e-01 4.17319149e-01
2.94600427e-01 -2.20538452e-01 -3.53243262e-01 -1.08891869e+00
2.14895248e-01 6.35871112e-01 -5.00989258e-01 -1.99685723e-01
-5.94230711e-01 -9.40645516e-01 2.99414605e-01 4.64093208e-01
6.35870337e-01 -9.09372091e-01 -4.79670048e-01 -9.71454568e-03
-5.05684912e-02 -1.21554303e+00 -1.30604342e-01 7.81434253e-02
-8.65478098e-01 -1.51177394e+00 -1.84292555e-01 -5.55479288e-01
7.17677712e-01 4.05831873e-01 1.27228057e+00 -2.01936275e-01
-5.02454758e-01 4.15510863e-01 -6.32808149e-01 -7.89417744e-01
-4.78792429e-01 -3.66226494e-01 3.53656769e-01 2.21105278e-01
4.08343703e-01 -5.64806759e-01 -5.84553480e-01 7.71551609e-01
-9.29306686e-01 -5.11672795e-01 6.07441425e-01 7.52777040e-01
8.01266789e-01 5.07065058e-01 7.62185454e-01 -6.74395919e-01
4.01111603e-01 -9.10720944e-01 -6.64223254e-01 -2.29172602e-01
-6.52827203e-01 -5.37241817e-01 8.31646800e-01 -3.63436550e-01
-1.01018357e+00 -2.05254346e-01 -8.43916833e-02 -8.45392644e-01
-7.19313145e-01 3.23239475e-01 -2.96174347e-01 3.72160584e-01
5.87759078e-01 5.49801551e-02 1.90863296e-01 -9.38971221e-01
-3.75668198e-01 6.25627398e-01 6.67916238e-01 -4.72955763e-01
1.07244432e+00 5.10114908e-01 1.36377990e-01 -1.14266551e+00
-7.74717569e-01 -6.47258043e-01 -7.27979958e-01 -3.75554144e-01
6.87761605e-01 -8.86731327e-01 -1.29309043e-01 7.88344979e-01
-8.66133213e-01 -2.06037417e-01 -7.53478587e-01 5.52457094e-01
-1.53353974e-01 6.38095260e-01 -4.32041228e-01 -5.63992381e-01
2.35918295e-02 -9.03642952e-01 1.03486967e+00 7.34335743e-03
-1.16324745e-01 -9.09492612e-01 1.41653731e-01 3.43748368e-02
3.56528945e-02 5.56745410e-01 1.05383170e+00 -9.40174282e-01
-2.27822676e-01 -2.83594757e-01 1.45270988e-01 7.99284518e-01
6.55316114e-01 2.69262880e-01 -9.53176856e-01 -4.46901262e-01
4.54007626e-01 1.38010696e-01 4.56664741e-01 3.00379038e-01
1.44234633e+00 -7.07226321e-02 1.73251834e-02 4.41402078e-01
1.13594043e+00 4.03465390e-01 6.17608726e-01 6.01116538e-01
6.53569221e-01 5.30535579e-01 1.12831533e+00 8.05717587e-01
-1.52357697e-01 8.04314256e-01 9.21632051e-01 -2.43321016e-01
-1.35069519e-01 2.98264086e-01 6.27088785e-01 1.05104125e+00
-1.49396271e-01 -1.31888151e-01 -7.84486949e-01 6.04346275e-01
-1.71532929e+00 -8.77150178e-01 -5.66466987e-01 2.11396551e+00
3.17465484e-01 4.60576750e-02 3.30993801e-01 5.64596713e-01
6.67461455e-01 2.96218812e-01 -6.96149945e-01 -3.26329440e-01
-2.93679833e-01 -8.39914680e-02 2.66274586e-02 -1.98744312e-01
-1.23491859e+00 1.67261362e-01 5.58533239e+00 7.63071001e-01
-9.70843136e-01 -6.99217170e-02 2.66222090e-01 1.83746189e-01
1.39063537e-01 -4.23575073e-01 -6.06250405e-01 6.93160474e-01
6.46623075e-01 2.01353207e-01 8.53254497e-02 8.88968706e-01
2.09318548e-01 1.65378734e-01 -8.26649010e-01 8.36144030e-01
9.63284373e-02 -7.74795830e-01 -8.07883516e-02 1.29652649e-01
6.61804020e-01 -4.35923636e-02 -4.34868447e-02 8.18085819e-02
-2.95965672e-02 -5.45565248e-01 5.96674144e-01 2.50316858e-01
5.41007161e-01 -6.01094723e-01 1.04902363e+00 1.51143342e-01
-1.16879225e+00 -3.15410078e-01 -4.25065577e-01 -1.07746378e-01
-1.59384191e-01 1.07060289e+00 -6.23020053e-01 1.18731487e+00
1.08815706e+00 9.25109267e-01 -6.03611946e-01 8.42518628e-01
-2.00085178e-01 9.19596374e-01 -9.30019468e-02 6.71738923e-01
-9.10338536e-02 -5.82172453e-01 1.06169319e+00 8.19777906e-01
7.45401263e-01 -4.48915184e-01 3.04594249e-01 6.82203412e-01
3.92086029e-01 -1.34429976e-01 -9.01327252e-01 2.98652142e-01
3.81846458e-01 1.32717347e+00 -4.92473245e-01 8.65834951e-02
-9.15512800e-01 8.59267831e-01 -4.13032711e-01 4.72621381e-01
-8.49456012e-01 -2.97682732e-01 1.18822598e+00 1.13644987e-01
5.13504863e-01 -3.02666128e-01 -2.78680801e-01 -1.01426899e+00
2.77947962e-01 -1.14428711e+00 9.59639847e-01 -5.13462424e-01
-2.01703763e+00 5.87385595e-01 -6.09168261e-02 -1.98937023e+00
1.73995066e-02 -9.71937180e-01 -9.12667632e-01 6.36585712e-01
-1.36656129e+00 -9.55025911e-01 -7.88342595e-01 7.23901153e-01
9.43430305e-01 -6.09075785e-01 7.56316483e-01 3.79604340e-01
-1.00081301e+00 1.52685210e-01 1.75702140e-01 1.88736141e-01
8.43588769e-01 -1.36516809e+00 4.23637182e-01 1.28559840e+00
2.36757211e-02 3.77852321e-02 7.56159604e-01 -6.17266059e-01
-1.31406188e+00 -1.29111421e+00 5.77092990e-02 -4.56671447e-01
7.29052544e-01 -2.89886799e-02 -1.23864567e+00 5.04854143e-01
1.27978891e-01 4.37596917e-01 7.58357525e-01 -7.03250319e-02
-3.00103486e-01 -3.28245312e-01 -1.06964135e+00 2.97730863e-01
8.02346826e-01 -3.43266606e-01 -6.24855280e-01 2.65079468e-01
3.84088248e-01 -3.84282529e-01 -9.98106718e-01 7.34850228e-01
5.90021573e-02 -1.01470113e+00 9.82308805e-01 -4.23315436e-01
4.47746724e-01 -8.18628907e-01 -3.32064420e-01 -1.52290452e+00
-3.16828400e-01 -3.65876108e-01 -4.76907194e-01 1.35066378e+00
-9.43896100e-02 -8.80015731e-01 3.77500921e-01 -1.52860507e-01
-6.83509946e-01 -8.28359246e-01 -1.02965915e+00 -1.18884873e+00
-1.49198264e-01 -4.06218767e-01 7.42818415e-01 1.11690795e+00
-4.48102862e-01 -1.77709647e-02 -6.64810911e-02 7.65142858e-01
4.48089659e-01 4.19063807e-01 8.03320467e-01 -1.40665400e+00
-1.45809546e-01 -1.24737427e-01 -5.58007598e-01 -6.11876965e-01
-2.08295554e-01 -5.54619253e-01 -8.62101763e-02 -1.09094477e+00
-3.80645901e-01 -3.27612549e-01 -5.20724714e-01 3.27488720e-01
-1.48280263e-01 1.92511618e-01 -1.52101442e-01 4.37592328e-01
-2.23082617e-01 5.14311850e-01 1.20524752e+00 2.53400747e-02
-6.32057041e-02 2.21642554e-01 -3.12071323e-01 9.17721033e-01
1.06179178e+00 -2.89808959e-01 -2.64602065e-01 -1.34642869e-01
-4.43874896e-02 -3.31428409e-01 6.44639313e-01 -1.17902207e+00
-2.69603342e-01 -1.80600151e-01 4.60517436e-01 -7.74859130e-01
-8.27798247e-02 -1.00746560e+00 -2.67688688e-02 5.04620910e-01
4.25492734e-01 7.05803573e-01 2.80380845e-01 7.63435304e-01
-5.48838794e-01 -1.04668885e-01 5.81236422e-01 1.28313631e-01
-1.01576579e+00 2.29121864e-01 -5.92699587e-01 1.55316576e-01
1.26821351e+00 -2.22650707e-01 -5.43361843e-01 6.52037859e-02
-3.07313919e-01 -4.58859503e-02 4.49825466e-01 6.85239911e-01
6.04182482e-01 -1.40686738e+00 -7.31449425e-01 7.70978868e-01
2.47741073e-01 4.74497676e-01 4.93933529e-01 1.06800282e+00
-4.44301873e-01 -1.17861554e-01 -3.55236679e-01 -8.03344309e-01
-1.29084647e+00 5.25900543e-01 3.09346616e-01 -8.99510011e-02
-8.62260818e-01 4.49261487e-01 3.87432992e-01 -3.59300703e-01
-1.81230441e-01 -2.70276785e-01 -2.37724915e-01 8.82287547e-02
3.96670938e-01 7.85110116e-01 3.05504680e-01 -4.59329814e-01
-2.70617515e-01 4.61668670e-01 1.37989014e-01 6.70239091e-01
1.24659467e+00 -3.41128232e-03 -1.98016688e-01 7.31150091e-01
7.60299027e-01 4.42839503e-01 -1.34408081e+00 -8.05326402e-02
6.44890144e-02 -8.90748024e-01 -2.46429250e-01 -6.00569367e-01
-1.26009119e+00 6.65598094e-01 7.94051588e-01 6.94962084e-01
1.60079539e+00 -9.02488176e-03 7.67582297e-01 -3.47380601e-02
-1.47634037e-02 -1.11753201e+00 4.99389738e-01 4.43379521e-01
9.75364268e-01 -1.21719491e+00 4.80370484e-02 -6.08143568e-01
-7.57945359e-01 1.27920198e+00 9.36305523e-01 -7.89826438e-02
5.90798795e-01 3.25568050e-01 3.79494548e-01 -5.54268301e-01
-5.10592341e-01 2.05449946e-02 2.47510329e-01 6.10620379e-01
-7.89593533e-02 -7.25047737e-02 1.40093267e-01 6.70986891e-01
8.17172080e-02 -5.29188693e-01 3.90345752e-01 9.33139086e-01
-1.35307923e-01 -1.02864337e+00 -6.11154795e-01 4.59009767e-01
-4.92763638e-01 5.40742576e-01 -2.15138584e-01 8.71274889e-01
2.73709685e-01 9.79321122e-01 2.85425276e-01 -5.42420089e-01
7.61322320e-01 1.45643950e-01 2.03618035e-02 -3.54910195e-01
-3.05272609e-01 -7.83430934e-02 -5.40777668e-02 -6.46647155e-01
-2.10916400e-01 -9.40047204e-01 -9.97741818e-01 9.22030304e-03
-2.54196763e-01 9.06756446e-02 3.75328004e-01 8.29113185e-01
5.13472438e-01 1.09372592e+00 1.07643878e+00 -5.57861328e-01
-6.59893274e-01 -9.38525677e-01 -1.08327556e+00 8.63965631e-01
4.01760399e-01 -8.74256134e-01 -7.96747625e-01 7.64032155e-02] | [7.570939064025879, 2.0306143760681152] |
023eded4-dbd3-496e-b37b-137791c359e5 | heart-beat-characterization-from | 1605.04634 | null | http://arxiv.org/abs/1605.04634v1 | http://arxiv.org/pdf/1605.04634v1.pdf | Heart Beat Characterization from Ballistocardiogram Signals using Extended Functions of Multiple Instances | A multiple instance learning (MIL) method, extended Function of Multiple
Instances ($e$FUMI), is applied to ballistocardiogram (BCG) signals produced by
a hydraulic bed sensor. The goal of this approach is to learn a personalized
heartbeat "concept" for an individual. This heartbeat concept is a prototype
(or "signature") that characterizes the heartbeat pattern for an individual in
ballistocardiogram data. The $e$FUMI method models the problem of learning a
heartbeat concept from a BCG signal as a MIL problem. This approach elegantly
addresses the uncertainty inherent in a BCG signal e. g., misalignment between
training data and ground truth, mis-collection of heartbeat by some
transducers, etc. Given a BCG training signal coupled with a ground truth
signal (e.g., a pulse finger sensor), training "bags" labeled with only binary
labels denoting if a training bag contains a heartbeat signal or not can be
generated. Then, using these bags, $e$FUMI learns a personalized concept of
heartbeat for a subject as well as several non-heartbeat background concepts.
After learning the heartbeat concept, heartbeat detection and heart rate
estimation can be applied to test data. Experimental results show that the
estimated heartbeat concept found by $e$FUMI is more representative and a more
discriminative prototype of the heartbeat signals than those found by
comparison MIL methods in the literature. | ['Marjorie Skubic', 'Licet Rosales', 'Alina Zare', 'Princess Lyons', 'Changzhe Jiao'] | 2016-05-16 | null | null | null | null | ['heart-rate-estimation'] | ['medical'] | [ 5.55480838e-01 3.72483075e-01 -2.33953092e-02 -4.74660069e-01
-8.23201418e-01 -1.96124494e-01 -8.32818449e-02 7.70381764e-02
1.69412598e-01 9.27240193e-01 -3.55216414e-01 1.18143521e-01
-3.88958216e-01 -6.46414042e-01 -6.79882228e-01 -9.43885446e-01
-3.37993294e-01 6.40595853e-01 -4.45119530e-01 2.40626380e-01
8.36195499e-02 1.03538550e-01 -1.33672345e+00 2.28649244e-01
2.10830197e-01 1.25085568e+00 5.47952391e-02 1.00609553e+00
1.41762137e-01 4.29865628e-01 -1.16046643e+00 2.64538556e-01
1.24343716e-01 -8.72057438e-01 -4.63258713e-01 4.81023453e-02
4.70393479e-01 -7.21780304e-03 3.14131081e-02 6.54192328e-01
6.78718507e-01 -1.38353035e-01 7.71218956e-01 -1.34062314e+00
2.42241010e-01 7.73202181e-01 -1.59074053e-01 1.54138014e-01
2.67435282e-01 -3.18275020e-02 7.55162120e-01 -5.68604648e-01
1.02687038e-01 7.61896968e-01 1.24968159e+00 5.57884455e-01
-1.51376021e+00 -7.17555463e-01 -3.97298425e-01 -1.14687383e-01
-1.47229397e+00 -4.91672792e-02 1.05402732e+00 -5.28589547e-01
2.70940304e-01 6.98262632e-01 1.00007761e+00 9.54312146e-01
5.14621735e-01 7.40354240e-01 1.17221129e+00 -4.76182580e-01
3.55561078e-01 3.68432730e-01 3.89977753e-01 5.80587149e-01
1.45991191e-01 2.50045717e-01 -6.52371824e-01 -2.67382473e-01
6.56674206e-01 -3.02529112e-02 -3.76040727e-01 5.96821308e-02
-1.11648440e+00 5.30381203e-01 -4.61637974e-02 4.73961651e-01
-3.75029027e-01 4.30478841e-01 4.55461502e-01 2.77160734e-01
2.28127614e-01 7.62327254e-01 -6.48827195e-01 8.65232386e-03
-1.16841388e+00 4.24196154e-01 1.13044095e+00 7.13068783e-01
9.62972343e-01 2.32424989e-01 -4.18156266e-01 4.64468449e-01
3.78806561e-01 9.13634002e-01 4.53736812e-01 -8.90570998e-01
1.14128448e-01 2.81581998e-01 1.65946186e-01 -9.82908905e-01
-7.31067657e-01 -4.91653234e-01 -9.14638162e-01 -2.78864145e-01
5.78999937e-01 -4.25229341e-01 -5.13215661e-01 1.58127534e+00
2.82725364e-01 9.34336424e-01 5.78516759e-02 8.82914424e-01
1.03830624e+00 6.68030679e-01 -1.65592343e-01 -7.03262627e-01
1.30869377e+00 -9.30181295e-02 -8.01637411e-01 -4.34838325e-01
1.31385654e-01 -2.77966201e-01 6.23190343e-01 6.36956751e-01
-4.59341288e-01 -1.06446004e+00 -1.09220457e+00 8.70619237e-01
-5.07612675e-02 -1.08891735e-02 5.16335607e-01 1.00889301e+00
-2.48722538e-01 8.89288664e-01 -5.30241132e-01 8.21407437e-02
2.87209526e-02 8.77973884e-02 1.06877647e-01 1.64731324e-01
-1.47759318e+00 5.31120598e-01 5.40231049e-01 3.00232917e-01
-1.12992203e+00 -1.06740403e+00 -8.23446929e-01 -3.14863354e-01
1.22401617e-01 -3.10909599e-01 8.46598804e-01 -7.40136921e-01
-1.26077461e+00 8.39371741e-01 1.22945316e-01 -6.35576546e-01
3.80978018e-01 -1.24610975e-01 -6.56541228e-01 1.81377098e-01
-8.71210843e-02 4.97389399e-02 1.73489451e+00 -1.30505025e+00
-2.87372798e-01 -1.93853974e-01 -7.04014659e-01 -2.80234426e-01
2.52924740e-01 -4.90428388e-01 1.38231575e-01 -5.59611082e-01
4.29600894e-01 -8.82893860e-01 1.91214964e-01 -6.53368175e-01
-7.63557494e-01 -9.88474414e-02 9.76014614e-01 -7.59956241e-01
1.30725968e+00 -2.10017395e+00 -2.17710763e-01 6.04722857e-01
1.39532074e-01 1.15337744e-02 3.76653790e-01 1.43527389e-01
-1.79819927e-01 -6.78967535e-02 -3.89994979e-01 -5.69188967e-02
-6.14420837e-03 3.70712399e-01 -3.21946025e-01 6.87430382e-01
6.32418320e-02 6.09958768e-01 -1.04147279e+00 -5.42613506e-01
3.42171401e-01 3.87711048e-01 9.62354168e-02 5.42043447e-01
-1.83994696e-01 7.66196072e-01 -1.87001199e-01 7.13272095e-01
2.73191959e-01 -8.36523101e-02 3.14469278e-01 -7.44290173e-01
2.77021557e-01 -3.15854847e-01 -1.50250494e+00 1.51383579e+00
-1.93880588e-01 6.23793125e-01 -3.03623229e-01 -1.50666082e+00
1.64705098e+00 7.76924312e-01 8.68806303e-01 -3.25123906e-01
1.23471618e-01 2.15061694e-01 -5.89745641e-02 -9.45141137e-01
-2.10902393e-02 -6.18880272e-01 -3.75985563e-01 4.80244964e-01
2.13228643e-01 -5.50844789e-01 -1.26042321e-01 -2.92207330e-01
9.19859588e-01 1.68881625e-01 1.87532559e-01 -4.05490905e-01
4.11303788e-01 -3.62302810e-01 8.27743173e-01 1.30252564e+00
-3.08487266e-01 6.80801570e-01 1.38983294e-01 -7.13686705e-01
-8.18036854e-01 -1.04998684e+00 -5.00891626e-01 5.57932317e-01
-2.70900205e-02 -2.97797441e-01 -4.99232352e-01 -2.03689188e-01
1.88975424e-01 3.30047578e-01 -5.66032052e-01 -2.48165727e-01
-8.46855223e-01 -1.15150940e+00 7.94597924e-01 4.76811200e-01
4.09793198e-01 -1.10731912e+00 -1.31354415e+00 6.74057424e-01
-6.75453782e-01 -8.62615287e-01 -1.38317078e-01 4.73321766e-01
-1.00157487e+00 -1.18900681e+00 -3.59672606e-01 -4.48514968e-01
2.24020585e-01 -6.88617408e-01 1.30526459e+00 -5.39776124e-02
-1.22651970e+00 7.62965202e-01 -2.61515588e-01 -9.31752920e-01
-4.04441118e-01 -6.52941465e-01 1.42808691e-01 4.45885688e-01
1.85496494e-01 -3.86937529e-01 -6.48117483e-01 4.19589058e-02
-2.98226446e-01 -3.02095085e-01 2.34746784e-01 8.95479918e-01
8.15974772e-01 3.36015493e-01 1.05319178e+00 -9.91269052e-01
4.54602569e-01 -3.61196637e-01 -9.34508443e-02 1.01847671e-01
-7.96562672e-01 -1.87766507e-01 4.38196361e-01 -6.19403780e-01
-4.92344230e-01 4.55278270e-02 3.27054113e-01 -5.28023362e-01
-3.78900319e-01 5.19914925e-01 3.65945930e-03 4.21699405e-01
9.07342613e-01 2.70033956e-01 4.16475721e-02 -3.46022695e-01
1.12599395e-01 6.67810202e-01 1.05911314e+00 -9.49027538e-01
2.46965006e-01 6.12038001e-03 4.07607049e-01 -9.91827309e-01
-8.14186811e-01 -6.37606740e-01 -4.43532526e-01 -6.71130300e-01
8.50917518e-01 -7.83998728e-01 -9.41728652e-01 6.74338758e-01
-7.87227154e-01 -3.45411748e-01 -5.78436494e-01 7.23545611e-01
-8.90079856e-01 2.03459486e-01 -5.62254965e-01 -1.48028195e+00
-6.26814961e-01 -3.24959487e-01 1.03270280e+00 2.59491891e-01
-6.23757303e-01 -1.00623286e+00 8.49568248e-02 4.51815963e-01
1.76252887e-01 1.05464041e+00 6.51302516e-01 -7.38299847e-01
-1.29217699e-01 -7.26589978e-01 3.64669800e-01 5.49238086e-01
2.80225217e-01 -2.80482948e-01 -1.50605059e+00 -4.22337860e-01
5.26753843e-01 -3.27119350e-01 3.92900735e-01 6.67279065e-01
9.99252498e-01 -3.58144581e-01 -3.25829536e-01 2.70266086e-01
1.56769407e+00 3.98628324e-01 5.58959842e-01 -1.97698891e-01
5.34081578e-01 4.47285026e-01 6.77463531e-01 8.06899130e-01
-7.93155730e-02 4.40073133e-01 2.15604201e-01 7.47794658e-02
1.31148860e-01 -1.50896339e-02 2.25989357e-01 6.07180953e-01
-1.06628224e-01 1.79247066e-01 -8.21995616e-01 3.70956719e-01
-1.70648801e+00 -1.18259323e+00 -1.62531167e-01 2.49398375e+00
9.83115554e-01 5.71481884e-02 2.38490298e-01 9.10924196e-01
9.78495717e-01 -1.33534268e-01 -8.10100794e-01 -1.23211116e-01
1.35748863e-01 7.52762020e-01 1.79865047e-01 3.20416182e-01
-1.26463282e+00 -4.51998860e-02 6.41883230e+00 2.55019426e-01
-1.23991895e+00 -7.55048320e-02 7.74662018e-01 3.40647250e-01
1.88011020e-01 -3.20646286e-01 -8.28580797e-01 6.01168394e-01
1.27342439e+00 -7.52920583e-02 2.35344142e-01 5.03773391e-01
3.54468040e-02 2.21635159e-02 -1.55864990e+00 1.41396010e+00
4.15567249e-01 -1.09037650e+00 -7.35943079e-01 -4.87075180e-01
2.96365440e-01 -5.13468504e-01 -3.70263070e-01 5.40591836e-01
-4.88990098e-01 -1.18237841e+00 5.48970461e-01 1.10826492e+00
8.72510314e-01 -3.46950322e-01 7.92956293e-01 5.43025672e-01
-1.22587371e+00 -3.67834382e-02 -1.83886569e-03 1.11413427e-01
-8.54589120e-02 1.09528327e+00 -1.15501142e+00 6.26143694e-01
6.17094100e-01 6.00641131e-01 -1.80449918e-01 9.56563413e-01
-6.78869113e-02 1.09480035e+00 -3.95106435e-01 5.09492345e-02
-3.62699002e-01 -4.78787534e-02 5.69960535e-01 1.33568192e+00
2.76939094e-01 9.33598652e-02 3.68981451e-01 1.18825150e+00
4.68821168e-01 -2.08399042e-01 -5.54025531e-01 1.56424977e-02
5.76088190e-01 1.28779459e+00 -5.40590405e-01 -5.36321640e-01
5.11174509e-03 5.59881806e-01 -5.29426336e-01 3.26051503e-01
-8.34667802e-01 -5.26753366e-01 1.53354416e-02 1.53270796e-01
-9.85009819e-02 4.88010287e-01 -2.73378640e-01 -7.28744686e-01
-1.56944022e-01 -7.02521205e-01 5.20020425e-01 -7.72977948e-01
-1.42849672e+00 3.27980340e-01 2.00496852e-01 -1.36148429e+00
-6.90094829e-01 -5.14920056e-01 -7.34202445e-01 1.05095851e+00
-8.27191889e-01 -7.03926742e-01 -5.78019738e-01 6.93747580e-01
1.74207360e-01 2.31386274e-02 1.14342022e+00 2.30833247e-01
-2.94042617e-01 3.68715584e-01 -3.25530767e-01 3.11937720e-01
5.86928785e-01 -1.54323971e+00 -2.57266492e-01 3.48302603e-01
8.92155766e-02 3.29880804e-01 1.10329926e+00 -5.42320430e-01
-1.43674576e+00 -1.29341781e+00 7.12046504e-01 -5.52022338e-01
2.21122012e-01 -1.41965076e-01 -9.85150993e-01 5.20786822e-01
-3.81182909e-01 3.95282269e-01 8.88230741e-01 2.36665793e-02
-1.40258241e-02 -7.21923828e-01 -1.44258952e+00 -2.69772470e-01
2.50391692e-01 -5.65785408e-01 -8.84623826e-01 3.27475935e-01
1.89286441e-01 -6.27263904e-01 -1.44953620e+00 6.75330997e-01
7.55210280e-01 -6.36596859e-01 1.04667103e+00 -3.86487007e-01
-2.26297125e-01 -3.24064642e-01 -1.61366642e-01 -1.21319926e+00
-6.31395634e-03 -7.43561089e-01 -5.07439256e-01 1.08519018e+00
9.60254446e-02 -4.63532597e-01 8.33136022e-01 6.13180995e-01
-1.44121215e-01 -5.79425037e-01 -9.99996901e-01 -9.86484528e-01
-3.40521425e-01 -7.13256717e-01 4.19251770e-01 1.00825036e+00
3.27125043e-01 2.78509706e-01 -9.09710646e-01 1.87758461e-01
1.14383519e+00 4.88528192e-01 6.42067552e-01 -1.50744808e+00
-6.92162275e-01 2.04143628e-01 -5.69738686e-01 -3.54507595e-01
-2.69524336e-01 -8.07534039e-01 5.85718155e-01 -9.93881345e-01
-6.08701259e-02 -8.03222120e-01 -5.99487901e-01 3.31561267e-01
-1.55578390e-01 4.29757088e-01 1.09218895e-01 -3.01691405e-02
-1.44712895e-01 -2.23347366e-01 9.67192590e-01 -2.95646578e-01
-4.38250065e-01 3.83156389e-01 -1.44779503e-01 5.48712254e-01
9.52334940e-01 -4.51427549e-01 -4.09712493e-01 6.05654538e-01
9.34328586e-02 8.41676831e-01 6.71633899e-01 -1.34913921e+00
-8.71828496e-02 1.57698944e-01 6.16981328e-01 -4.96809453e-01
4.49707717e-01 -6.40877306e-01 7.54923940e-01 8.40813160e-01
-1.15504906e-01 -4.71934170e-01 9.24550928e-03 5.44183135e-01
-1.42936036e-01 -5.06907284e-01 6.82703018e-01 -4.58420098e-01
-3.36149037e-01 1.10008769e-01 -2.16064811e-01 2.79554993e-01
7.02195108e-01 -2.51306027e-01 -9.56261605e-02 -8.51520076e-02
-1.17883027e+00 -8.24338794e-02 -6.95407689e-01 1.80154964e-01
7.59024918e-01 -1.26941311e+00 -8.71111810e-01 6.17558360e-01
1.72477201e-01 5.95067218e-02 1.81844264e-01 9.01531398e-01
-8.53392407e-02 2.05882668e-01 8.86118636e-02 -1.12533772e+00
-9.96639311e-01 2.52207994e-01 7.25150704e-01 -3.97303253e-02
-8.75130177e-01 6.30068541e-01 -3.86554956e-01 -1.03527628e-01
3.47107589e-01 -6.17722392e-01 -1.25800624e-01 1.37692392e-01
5.80295563e-01 4.89500642e-01 -2.77865306e-02 -2.83786952e-01
-3.16322386e-01 6.02091014e-01 5.89249671e-01 1.02915406e-01
8.93379867e-01 5.59583381e-02 4.60751615e-02 1.32779050e+00
8.42147291e-01 -3.56533229e-01 -9.39447284e-01 -1.01347037e-01
-8.47581774e-02 -2.83542275e-01 -3.87021065e-01 -9.74636137e-01
-9.84768212e-01 6.21947408e-01 1.14130318e+00 5.06875515e-01
1.05428720e+00 -4.20529284e-02 5.18908322e-01 3.38701278e-01
4.68819231e-01 -1.21635818e+00 3.49117070e-01 -3.98593657e-02
7.54462242e-01 -1.09557712e+00 -7.92317316e-02 -2.93155666e-02
-3.67102176e-01 1.29966378e+00 4.19123054e-01 -1.74300373e-01
9.98617589e-01 3.27957988e-01 1.62810832e-01 -2.41334096e-01
-2.91036993e-01 1.99924231e-01 2.88648844e-01 7.88297772e-01
3.49168688e-01 5.65620542e-01 -2.67031252e-01 8.90317321e-01
-1.01633333e-01 3.41059864e-01 3.13953698e-01 6.75213456e-01
-6.46158338e-01 -6.48620009e-01 -8.01112294e-01 8.01981926e-01
-3.91240329e-01 1.38422117e-01 2.05873132e-01 6.13824964e-01
6.93749309e-01 1.04771125e+00 2.03340650e-01 -5.39326489e-01
3.37319046e-01 7.39858329e-01 6.66738451e-01 -6.72887266e-01
-5.94912171e-01 2.59495676e-01 7.30374530e-02 -2.17004225e-01
-5.87201953e-01 -6.37846470e-01 -1.45354152e+00 1.86669692e-01
-2.20802799e-01 2.81792194e-01 4.42449868e-01 1.09972477e+00
-2.83527464e-01 5.84451556e-01 6.79393291e-01 -6.69228494e-01
-4.72467840e-01 -1.18862486e+00 -1.27836323e+00 7.00752854e-01
5.21024048e-01 -2.58019805e-01 -7.33003080e-01 5.58044255e-01] | [14.199908256530762, 3.18831729888916] |
2529ac94-8cf6-4087-83e8-47c690ed03ff | a-comparative-analysis-of-techniques-and | 2305.13941 | null | https://arxiv.org/abs/2305.13941v2 | https://arxiv.org/pdf/2305.13941v2.pdf | A Comparative Analysis of Techniques and Algorithms for Recognising Sign Language | Sign language is a visual language that enhances communication between people and is frequently used as the primary form of communication by people with hearing loss. Even so, not many people with hearing loss use sign language, and they frequently experience social isolation. Therefore, it is necessary to create human-computer interface systems that can offer hearing-impaired people a social platform. Most commercial sign language translation systems now on the market are sensor-based, pricey, and challenging to use. Although vision-based systems are desperately needed, they must first overcome several challenges. Earlier continuous sign language recognition techniques used hidden Markov models, which have a limited ability to include temporal information. To get over these restrictions, several machine learning approaches are being applied to transform hand and sign language motions into spoken or written language. In this study, we compare various deep learning techniques for recognising sign language. Our survey aims to provide a comprehensive overview of the most recent approaches and challenges in this field. | ['S. K Singh', 'Ashutosh Bajpai', 'Ayush Sinha', 'Rupesh Kumar'] | 2023-05-05 | null | null | null | null | ['sign-language-recognition', 'sign-language-translation'] | ['computer-vision', 'computer-vision'] | [-1.60826460e-01 -2.27124274e-01 -4.24473166e-01 -1.17190003e-01
-6.02377057e-01 -1.60581514e-01 2.81861246e-01 -6.13607764e-01
-7.56809950e-01 7.20354497e-01 5.16197860e-01 -2.57016748e-01
2.49772325e-01 -4.20887679e-01 7.13621229e-02 -6.74607575e-01
1.79747581e-01 2.14840919e-01 4.56978887e-01 -3.48006189e-01
6.20338582e-02 6.09086990e-01 -1.65862691e+00 2.25029200e-01
7.16758072e-01 6.07152998e-01 2.99642622e-01 7.51095176e-01
-6.21895492e-01 7.07188368e-01 -5.40065050e-01 -9.54098592e-04
-6.65948018e-02 -8.13572645e-01 -5.22037923e-01 -1.36374325e-01
2.43341401e-01 -7.48667181e-01 -4.06139910e-01 6.94729865e-01
1.12789273e+00 -3.35468948e-01 4.43452567e-01 -1.12361217e+00
-4.85023081e-01 1.43495694e-01 -2.65056640e-01 -3.38303387e-01
7.85203815e-01 2.24898279e-01 6.76670194e-01 -7.45864451e-01
5.68120420e-01 1.08325088e+00 6.30576253e-01 1.15041351e+00
-6.92215502e-01 -6.60944283e-01 2.32869074e-01 4.99669850e-01
-1.14044917e+00 -6.29051924e-01 7.84267962e-01 -5.02942681e-01
1.13721395e+00 1.78621069e-01 1.30654812e+00 9.89918351e-01
-1.83154672e-01 1.17709875e+00 1.16156590e+00 -9.68026161e-01
-3.20988372e-02 -1.93122000e-01 2.36586154e-01 5.65295935e-01
2.15605438e-01 8.83075893e-02 -8.53298724e-01 -7.93880075e-02
6.40985608e-01 -2.90957442e-03 -6.03986621e-01 -3.45680118e-01
-1.09614706e+00 5.52288532e-01 2.42900461e-01 6.79627240e-01
-3.04695070e-01 2.06885308e-01 1.43909439e-01 5.31581938e-01
-2.80419588e-01 -1.72107279e-01 -2.03644615e-02 -7.89771616e-01
-7.36120164e-01 -2.04685200e-02 7.02362239e-01 5.15868843e-01
-2.81917565e-02 2.32219696e-02 -9.11675394e-02 1.01657236e+00
8.50338221e-01 8.25148821e-01 6.27445221e-01 -6.87074780e-01
-1.80387888e-02 4.97732610e-01 8.28251839e-02 -3.44294578e-01
-4.82315451e-01 2.61254638e-01 -7.02871084e-01 1.04324901e+00
6.03218794e-01 -1.13365933e-01 -1.43722427e+00 1.32350826e+00
-1.00207277e-01 -3.36983204e-01 -2.85817543e-03 1.01672614e+00
9.80115235e-01 -5.26517583e-03 8.15323666e-02 -1.55374736e-01
1.06837142e+00 -8.20642412e-01 -1.04373074e+00 -1.71015993e-01
3.61371398e-01 -9.29727316e-01 1.11366653e+00 4.64462936e-01
-9.09611344e-01 -9.76877734e-02 -7.93972075e-01 -1.75544485e-01
-3.18203926e-01 6.18416034e-02 6.86781406e-01 9.29767489e-01
-1.15890801e+00 1.20689340e-01 -1.06167257e+00 -8.71117830e-01
2.76390970e-01 3.90944153e-01 -3.43209475e-01 1.17599465e-01
-9.19849575e-01 1.13299286e+00 -1.84050247e-01 1.71977416e-01
8.58366266e-02 3.13023366e-02 -7.09569931e-01 -4.71597672e-01
-3.11134607e-01 -6.99812412e-01 1.54938436e+00 -9.42057073e-01
-2.04665661e+00 1.20391512e+00 -4.90783900e-01 -6.96313567e-03
9.73656654e-01 -3.53595525e-01 -5.67764938e-01 -4.37259907e-03
-3.81730288e-01 4.55786139e-01 8.69802773e-01 -9.79903758e-01
-8.95230651e-01 -2.18096375e-01 -3.63633156e-01 9.03875306e-02
-1.96371093e-01 6.30651712e-01 -4.73530233e-01 -5.04951179e-01
3.95915568e-01 -7.30507553e-01 1.64354682e-01 8.14613402e-01
-2.39481449e-01 -2.84939349e-01 9.14522946e-01 -9.49707329e-01
1.10476351e+00 -2.15250683e+00 -1.03178918e-02 1.74116090e-01
2.49494597e-01 7.41772771e-01 1.59082144e-01 4.08420116e-01
3.85884374e-01 -2.87082970e-01 -2.81722307e-01 -1.98305488e-01
-3.10717952e-02 3.32107782e-01 -1.69938654e-01 3.34686518e-01
-4.71298814e-01 8.79623175e-01 -7.11076736e-01 -3.64501804e-01
5.12669504e-01 8.44359100e-01 -1.97523579e-01 2.91811954e-03
1.94583476e-01 5.69649577e-01 -1.17709056e-01 1.08228505e+00
3.63157928e-01 1.32982373e-01 1.07875168e-01 3.27645928e-01
-2.72999674e-01 8.72974098e-02 -1.06514180e+00 1.37733245e+00
-2.93027520e-01 9.47808146e-01 3.34760040e-01 -5.77974439e-01
6.67672992e-01 4.79762048e-01 4.57776815e-01 -6.37832642e-01
2.18430817e-01 8.16188395e-01 1.01634294e-01 -8.29106867e-01
-1.50524929e-01 -4.27444845e-01 3.27235371e-01 3.81671697e-01
-4.56186056e-01 -2.18168437e-01 -1.01592526e-01 -3.87913167e-01
9.77813661e-01 -2.54211333e-02 4.68406707e-01 5.87400377e-01
4.38669652e-01 -2.39831150e-01 3.93080264e-01 5.72919846e-01
-5.85578918e-01 8.25395405e-01 -8.91071633e-02 -3.44169021e-01
-5.68129241e-01 -1.21905839e+00 2.27613285e-01 5.97039402e-01
-7.84020945e-02 6.55712634e-02 -2.94103414e-01 -1.26173154e-01
-2.34640464e-02 2.79560983e-01 1.41588926e-01 1.28429577e-01
-5.35621047e-01 -6.13251179e-02 6.26191437e-01 6.79402888e-01
8.06118011e-01 -1.50300336e+00 -8.57549667e-01 2.63029337e-01
-2.28355199e-01 -8.94347668e-01 -3.57710779e-01 -4.84388620e-01
-4.78097916e-01 -9.89524007e-01 -1.72350216e+00 -1.46254325e+00
4.83160317e-01 2.53082752e-01 3.14199001e-01 1.89218313e-01
-4.17223573e-01 6.84433222e-01 -5.88579237e-01 -7.22785830e-01
-3.88402313e-01 -3.11251253e-01 2.71984130e-01 -1.13499254e-01
7.31345296e-01 -7.19723821e-01 -4.36928689e-01 5.28931506e-02
-3.42561543e-01 4.66104411e-02 7.67036736e-01 9.17926192e-01
1.49040595e-01 -6.26990855e-01 1.39217526e-01 2.38732532e-01
8.19119513e-01 5.08245468e-01 -3.81243587e-01 4.48704183e-01
-3.84580314e-01 -1.11394130e-01 -1.20406769e-01 -7.24397421e-01
-8.63144696e-01 2.55433738e-01 -3.98178339e-01 9.62658823e-02
-1.81849688e-01 3.25374782e-01 -1.02883875e-01 -4.61984932e-01
4.00472701e-01 3.50688457e-01 5.85751653e-01 -6.58950746e-01
2.03957960e-01 1.48084807e+00 6.45068049e-01 1.80452392e-01
5.40706098e-01 7.28373349e-01 -3.27881813e-01 -1.40938973e+00
-8.52253810e-02 -7.26946414e-01 -6.47152483e-01 -5.87857842e-01
8.38276267e-01 -5.11668563e-01 -8.42028916e-01 1.47032547e+00
-1.30702567e+00 -5.07094502e-01 -1.97244167e-01 9.72822011e-01
-6.19831443e-01 5.50834656e-01 -3.61330241e-01 -1.17361557e+00
-3.81297290e-01 -8.05468917e-01 8.27687204e-01 3.48075002e-01
-4.89044517e-01 -5.01772344e-01 3.33622158e-01 3.02787513e-01
6.81909859e-01 1.79515537e-02 6.89708352e-01 1.60288081e-01
-3.48600030e-01 -5.56538701e-01 -2.48749360e-01 4.58554715e-01
4.73221242e-01 -7.98595622e-02 -9.60487485e-01 -2.54511744e-01
-5.22516787e-01 -1.90702036e-01 8.31653237e-01 6.71408236e-01
4.86654669e-01 3.18992101e-02 -4.61992741e-01 3.81027967e-01
7.18879044e-01 5.95905006e-01 8.43111396e-01 3.29538971e-01
4.48068082e-01 5.40648878e-01 7.02647120e-02 7.53326640e-02
3.71528268e-01 9.19124186e-01 -2.86327183e-01 -3.63847286e-01
-7.09992468e-01 -2.28226274e-01 5.80062628e-01 8.30546200e-01
-6.36919796e-01 1.37160972e-01 -1.13413811e+00 4.92557794e-01
-1.68157315e+00 -1.06167567e+00 -2.53195670e-02 2.08108568e+00
7.76556015e-01 -1.17082886e-01 2.88394272e-01 6.80413961e-01
4.90521014e-01 -3.52929205e-01 -4.94856566e-01 -2.24841505e-01
-3.42731029e-01 3.76511216e-01 3.27072829e-01 6.71004951e-01
-8.01570117e-01 9.46505010e-01 6.97117901e+00 1.38759673e-01
-1.70074916e+00 -6.26431927e-02 -4.82820094e-01 -5.18827289e-02
8.01337287e-02 -2.58751005e-01 -5.49246848e-01 2.23267466e-01
2.15872794e-01 3.32239107e-03 1.61476120e-01 5.96405029e-01
6.01426482e-01 -1.75100029e-01 -7.94277072e-01 1.63406181e+00
1.68386847e-01 -8.19640458e-01 -2.25152045e-01 1.42219560e-02
1.15970790e-01 1.00975260e-01 -1.40158907e-01 1.33070827e-01
6.46638945e-02 -1.00992632e+00 6.72077358e-01 7.53974795e-01
1.14126492e+00 -1.59794092e-01 6.02947712e-01 1.51512459e-01
-9.82641459e-01 -1.79612249e-01 3.50602478e-01 -4.88467962e-01
7.99329579e-01 3.02742887e-02 -5.15194237e-01 -2.42517948e-01
1.00442111e+00 3.05213898e-01 -8.45271870e-02 1.70996749e+00
-6.74725056e-01 3.86445761e-01 -5.77551186e-01 -5.44613659e-01
-2.15047434e-01 7.74220675e-02 7.59772837e-01 8.00275207e-01
6.15175545e-01 -4.30090204e-02 9.10188705e-02 1.35617122e-01
5.06268442e-01 1.94388792e-01 -7.46685565e-01 -1.54238477e-01
4.77461778e-02 3.45313758e-01 -3.14800709e-01 -1.30009223e-02
-9.04356360e-01 1.41157413e+00 -3.21482152e-01 5.18517494e-01
-2.20691949e-01 -6.96628988e-01 9.62026715e-01 3.35547656e-01
-8.80585685e-02 -6.60259724e-01 -2.95351863e-01 -1.00429642e+00
4.49886143e-01 -8.10646892e-01 1.99026600e-01 -8.30152929e-01
-9.96653378e-01 2.01373681e-01 -3.49871695e-01 -1.56549573e+00
-5.61430216e-01 -1.00726855e+00 -4.85150576e-01 8.51205230e-01
-1.40114069e+00 -1.43131244e+00 -5.05052984e-01 6.36029661e-01
2.95236707e-01 -4.45005566e-01 1.08880520e+00 3.07712406e-01
2.65401155e-02 8.31913233e-01 2.58664433e-02 4.40076530e-01
7.38516152e-01 -7.00407624e-01 9.89942476e-02 6.86236084e-01
1.09107845e-01 4.74109828e-01 5.34869552e-01 -5.88204384e-01
-1.00006187e+00 -1.22285508e-01 1.45606685e+00 -1.76804915e-01
3.34339291e-01 -5.47045432e-02 -5.03465474e-01 4.57983524e-01
-1.70571799e-03 -3.65366697e-01 5.95377147e-01 -1.89237580e-01
-2.13692725e-01 3.72884646e-02 -1.12140012e+00 9.97573555e-01
1.35873163e+00 -7.70289183e-01 -7.30849385e-01 8.75319466e-02
2.58914530e-01 -1.46559179e-01 -9.95211452e-02 3.08327705e-01
1.32836223e+00 -7.66715288e-01 8.15019190e-01 -3.28932434e-01
-4.52552170e-01 -2.93776244e-01 1.05955843e-02 -1.00424516e+00
8.83258358e-02 -8.77194464e-01 -6.37494922e-02 9.27009463e-01
2.94434845e-01 -1.04539323e+00 8.64458621e-01 8.23389053e-01
2.37015322e-01 -1.35574654e-01 -1.31734812e+00 -1.11510026e+00
6.91624209e-02 -5.63164115e-01 2.46518418e-01 4.98852223e-01
4.94747102e-01 -1.23296589e-01 -4.80501473e-01 -3.66159737e-01
5.10610342e-01 -1.06110021e-01 8.64355147e-01 -1.33670330e+00
-1.26019478e-01 -9.51867998e-01 -9.03552711e-01 -1.07213223e+00
-1.69483632e-01 -5.61495066e-01 5.16818166e-02 -2.19770479e+00
-1.95530191e-01 1.10955499e-01 3.76371061e-03 7.81036794e-01
4.83474404e-01 -1.20217735e-02 1.89178303e-01 3.00618172e-01
1.96390420e-01 3.92778903e-01 1.31922555e+00 -3.88532251e-01
-6.98724508e-01 4.78744179e-01 -2.68839806e-01 9.13148940e-01
9.88672972e-01 1.15946375e-01 -5.33122979e-02 -4.88121301e-01
-1.34211123e-01 -2.13523746e-01 4.76611733e-01 -1.02742791e+00
4.90381867e-01 -2.01287299e-01 -1.47728063e-03 -6.32483900e-01
4.55501467e-01 -5.72133243e-01 -2.32183605e-01 8.82023513e-01
1.98510081e-01 -5.08220911e-01 -1.82979167e-01 3.65928560e-02
-4.60778385e-01 1.55887470e-01 1.01430130e+00 -9.53348577e-02
-1.11922479e+00 2.51805671e-02 -1.09069395e+00 -3.94450724e-01
8.30756843e-01 -6.03911698e-01 1.53148159e-01 -9.49826121e-01
-8.98635626e-01 -1.29828090e-02 3.52287799e-01 6.42248571e-01
9.64629471e-01 -1.27779245e+00 -5.24065554e-01 4.55842704e-01
3.61385554e-01 -3.45373183e-01 -8.05777404e-03 9.20836687e-01
-8.87570322e-01 5.69547236e-01 -3.77210945e-01 -3.01224291e-01
-1.87512708e+00 -1.23659641e-01 5.10051847e-01 3.70381445e-01
-1.18389380e+00 9.73117292e-01 -5.90961456e-01 -3.51124763e-01
8.81403983e-01 -4.18168753e-01 -2.99119800e-01 -4.04138938e-02
8.26875746e-01 3.52220148e-01 -2.75215268e-01 -6.60021365e-01
-5.14029503e-01 9.46300745e-01 2.53086895e-01 -7.76134491e-01
9.36604500e-01 -1.10013872e-01 -1.84025127e-03 4.12618458e-01
7.60848165e-01 9.92199057e-04 -7.75439441e-01 -3.08238298e-01
7.44015053e-02 -3.86486530e-01 -1.55151151e-02 -8.68379414e-01
-7.04429090e-01 1.15169656e+00 1.22302735e+00 -2.53420979e-01
1.24600601e+00 8.29718933e-02 1.01359951e+00 4.81086522e-01
7.81472206e-01 -1.40421677e+00 -7.79377818e-02 5.33268929e-01
1.16974378e+00 -1.19314790e+00 -3.89828563e-01 -2.44142741e-01
-3.93967271e-01 1.10660648e+00 2.33084425e-01 4.59443301e-01
1.08370912e+00 4.12105531e-01 9.62872922e-01 1.51720539e-01
1.19923607e-01 -8.91236305e-01 1.85548007e-01 1.14139533e+00
5.36247134e-01 2.21074164e-01 -8.04964662e-01 3.67307216e-01
-3.92268151e-01 5.78513384e-01 2.37717539e-01 1.35860658e+00
-5.80082893e-01 -1.48096335e+00 -4.96531606e-01 2.70226538e-01
-7.17052370e-02 3.05476189e-01 -6.65837228e-01 5.42158246e-01
-1.43063739e-02 1.08545113e+00 -3.79133165e-01 -2.22636610e-01
5.84042251e-01 4.02830571e-01 7.06698358e-01 -2.17112735e-01
2.18530625e-01 -4.44618650e-02 6.57039583e-02 -4.44045693e-01
-4.87782001e-01 -9.14518237e-01 -1.53165591e+00 -9.22092721e-02
1.76783875e-01 -6.16017044e-01 8.02719593e-01 9.78985488e-01
-1.14313059e-01 7.22087100e-02 -1.07081346e-01 -6.16774499e-01
-3.06156695e-01 -1.01113331e+00 -7.24661350e-01 8.34778845e-02
6.50573194e-01 -6.85231507e-01 -4.00270313e-01 -1.15414359e-01] | [9.07785415649414, -6.384265422821045] |
f6fe5a4a-63bd-4495-8dd3-d9c038c4dd6e | a-probabilistic-time-evolving-approach-to | 2204.09404 | null | https://arxiv.org/abs/2204.09404v1 | https://arxiv.org/pdf/2204.09404v1.pdf | A Probabilistic Time-Evolving Approach to Scanpath Prediction | Human visual attention is a complex phenomenon that has been studied for decades. Within it, the particular problem of scanpath prediction poses a challenge, particularly due to the inter- and intra-observer variability, among other reasons. Besides, most existing approaches to scanpath prediction have focused on optimizing the prediction of a gaze point given the previous ones. In this work, we present a probabilistic time-evolving approach to scanpath prediction, based on Bayesian deep learning. We optimize our model using a novel spatio-temporal loss function based on a combination of Kullback-Leibler divergence and dynamic time warping, jointly considering the spatial and temporal dimensions of scanpaths. Our scanpath prediction framework yields results that outperform those of current state-of-the-art approaches, and are almost on par with the human baseline, suggesting that our model is able to generate scanpaths whose behavior closely resembles those of the real ones. | ['Belen Masia', 'Diego Gutierrez', 'Daniel Martin'] | 2022-04-20 | null | null | null | null | ['scanpath-prediction'] | ['computer-vision'] | [ 7.26030171e-02 -1.64770961e-01 -3.62493157e-01 -5.14453173e-01
-5.42660117e-01 -4.08832759e-01 5.84258318e-01 -3.76718752e-02
-5.74356794e-01 4.25062925e-01 1.90299377e-01 -9.73591655e-02
-4.11316663e-01 -2.23114505e-01 -7.34302163e-01 -4.69433188e-01
-1.48447677e-01 5.55618882e-01 7.01043189e-01 -9.39528272e-02
5.89417100e-01 5.22113264e-01 -2.03628087e+00 -5.77762425e-02
6.48286641e-01 1.00377023e+00 1.89607129e-01 5.95506489e-01
2.48428717e-01 5.21131098e-01 -1.45485565e-01 -4.49476421e-01
1.20837756e-01 -3.57310385e-01 -8.38615835e-01 -4.34567183e-01
7.32577980e-01 -1.03594698e-01 -3.75361204e-01 7.27614939e-01
3.42671186e-01 3.30736369e-01 6.74245059e-01 -1.17295873e+00
-5.07580698e-01 1.60245180e-01 -8.93388808e-01 4.22137529e-01
4.84261960e-01 3.81118864e-01 1.41281390e+00 -4.77539062e-01
6.81207120e-01 1.24033308e+00 5.49784660e-01 4.46611375e-01
-1.58496368e+00 -6.30780876e-01 3.62786293e-01 1.12783289e+00
-1.34847999e+00 -9.76154953e-02 4.14010972e-01 -6.89634919e-01
7.50079095e-01 2.06562355e-02 9.50760543e-01 1.39322460e+00
6.29754961e-01 8.85890007e-01 1.14428616e+00 -3.01017880e-01
6.37208670e-02 -1.67817682e-01 -7.36228079e-02 4.07652169e-01
-2.90172875e-01 4.69629437e-01 -7.63379633e-01 1.98846593e-01
4.30114865e-01 -2.42822960e-01 -2.25862950e-01 -6.51007056e-01
-1.20593929e+00 6.59637630e-01 5.09014189e-01 1.56002373e-01
-6.10754013e-01 1.92881569e-01 -1.10421479e-01 -1.09519817e-01
2.83631444e-01 4.04906988e-01 -3.62541199e-01 -4.89350796e-01
-1.25317967e+00 5.30837119e-01 6.91820800e-01 7.44631708e-01
6.84531510e-01 -4.44463581e-01 -6.02163434e-01 3.78051579e-01
5.26108742e-01 6.24880731e-01 3.49279225e-01 -8.56655657e-01
2.44033262e-01 3.02398145e-01 3.04904401e-01 -9.55940247e-01
-5.48999012e-01 -2.15700030e-01 -2.23416805e-01 4.62949723e-01
5.68613946e-01 2.19540179e-01 -8.16611886e-01 1.84578192e+00
2.38827363e-01 3.68149221e-01 -6.75792456e-01 1.14081931e+00
2.35914066e-01 4.80653256e-01 2.76377976e-01 -2.54242215e-02
1.11194670e+00 -1.05039763e+00 -6.99393392e-01 -2.10799187e-01
-1.23193957e-01 -6.95995152e-01 1.39273167e+00 4.20692533e-01
-1.03800917e+00 -5.88388085e-01 -7.06155121e-01 2.94877624e-04
-1.09182678e-01 7.99517632e-02 3.57911050e-01 3.13316464e-01
-1.29779935e+00 5.79290390e-01 -9.61787224e-01 -5.53582430e-01
3.31662387e-01 3.02907109e-01 -2.78918684e-01 2.20469445e-01
-8.72077584e-01 1.12563610e+00 1.98384151e-01 -8.98502674e-03
-7.96750605e-01 -7.69296408e-01 -5.19421101e-01 1.52481362e-01
4.09638226e-01 -5.41258514e-01 1.59975183e+00 -7.97945917e-01
-1.68206584e+00 7.71058202e-01 -6.27773225e-01 -6.07563734e-01
6.29167080e-01 -3.82825464e-01 -3.07656407e-01 -7.04124123e-02
2.07089745e-02 9.73256409e-01 1.14194679e+00 -8.13244045e-01
-7.76365519e-01 -3.39617550e-01 -6.07614741e-02 1.48143500e-01
4.96990457e-02 -1.39125101e-02 -6.22770965e-01 -3.94787133e-01
-2.63628244e-01 -1.13132775e+00 -1.50431603e-01 3.23324442e-01
-5.11032939e-01 -5.09660184e-01 6.22918308e-01 -4.75115567e-01
1.48959017e+00 -1.96192002e+00 4.45429921e-01 9.32728350e-02
1.43168077e-01 2.33651444e-01 -7.89398700e-02 2.88656265e-01
-1.55311495e-01 -8.32880810e-02 6.43304512e-02 -5.02267540e-01
1.04729809e-01 -3.64239030e-02 -2.82114595e-01 5.76699674e-01
3.36800814e-02 6.60895288e-01 -8.37392986e-01 -3.82624477e-01
1.74277589e-01 4.02826726e-01 -4.88729089e-01 3.23285699e-01
-5.85398972e-01 6.89154685e-01 -2.62234390e-01 3.29815000e-01
3.91789913e-01 -4.17932898e-01 -1.59798965e-01 -4.34836298e-02
-4.93345320e-01 5.96559942e-02 -8.42791796e-01 1.71672666e+00
-3.97762954e-01 1.06571054e+00 -5.24474919e-01 -3.55555058e-01
5.08318543e-01 -7.00577870e-02 3.22128087e-01 -1.03294647e+00
1.79538071e-01 -9.25040916e-02 1.08380228e-01 -7.17258513e-01
5.80678284e-01 2.65836239e-01 2.40062058e-01 7.40105808e-01
-1.54374028e-02 2.46108919e-01 1.46385074e-01 -2.33190119e-01
1.00846434e+00 5.16624451e-01 2.74321228e-01 4.99296598e-02
5.98255336e-01 -6.38047308e-02 1.93390384e-01 7.29945779e-01
-4.20576364e-01 6.04197204e-01 6.39838338e-01 -4.82983530e-01
-9.93168652e-01 -1.06288505e+00 -3.08535211e-02 1.27797759e+00
4.49779093e-01 -1.43909141e-01 -8.87659788e-01 -6.25342071e-01
-1.15133122e-01 1.03000760e+00 -9.98601258e-01 -1.61361426e-01
-5.70039511e-01 -2.32304052e-01 3.37739676e-01 4.06005770e-01
2.21294656e-01 -1.20131123e+00 -1.13812912e+00 1.01203263e-01
-3.43902124e-04 -1.00052547e+00 -6.23751998e-01 -3.27816904e-01
-4.94298995e-01 -1.25681794e+00 -7.65265703e-01 -1.31340176e-01
6.26866072e-02 1.17789157e-01 1.19550025e+00 -2.40001306e-01
-4.56905544e-01 6.85230136e-01 -2.57213801e-01 -6.29568934e-01
7.38353003e-03 4.30571854e-01 -1.75290182e-01 2.67132282e-01
6.54937804e-01 -5.16934574e-01 -1.02546215e+00 5.13070285e-01
-3.93619746e-01 7.02842101e-02 7.52563119e-01 6.24041498e-01
8.84725451e-01 -2.87423074e-01 8.41194205e-03 -5.39639652e-01
3.40900511e-01 -5.97486556e-01 -1.09252465e+00 2.09344313e-01
-9.83762503e-01 1.77761525e-01 1.77666947e-01 -4.72994149e-01
-1.20682979e+00 1.33816749e-01 8.33322778e-02 -5.37463307e-01
-3.89813900e-01 1.62654042e-01 3.59562457e-01 -1.23504497e-01
6.65261865e-01 2.29165271e-01 3.63283455e-02 -4.98582572e-01
3.86261493e-01 1.63984209e-01 3.03387791e-01 -6.14275932e-02
6.10083461e-01 5.70258260e-01 3.23748559e-01 -5.64198375e-01
-1.01830089e+00 -5.12560666e-01 -9.25787807e-01 -4.77960676e-01
1.22554708e+00 -5.52866042e-01 -1.10408854e+00 4.37528044e-01
-1.34380043e+00 -6.04941368e-01 -7.53978966e-03 6.47888422e-01
-8.37563455e-01 4.92232442e-02 8.27352405e-02 -6.00619256e-01
3.81795838e-02 -1.23827267e+00 1.18200588e+00 4.97726083e-01
-4.54910517e-01 -1.05880916e+00 5.29030621e-01 2.58436978e-01
5.29154241e-01 1.88110575e-01 9.25991595e-01 -4.32827622e-01
-1.01840639e+00 -1.15059450e-01 -4.06905502e-01 -7.98617229e-02
-3.02440524e-01 9.60433111e-02 -9.25610125e-01 -8.40737522e-02
-3.79328996e-01 7.63466209e-02 7.28435040e-01 7.48895764e-01
1.38972354e+00 2.75141653e-02 -5.62701583e-01 6.29081368e-01
1.25788224e+00 -1.68050174e-02 4.86633629e-01 5.05019665e-01
6.64787352e-01 8.31649303e-01 9.58324730e-01 4.70229775e-01
7.84801185e-01 1.14999402e+00 9.06002402e-01 2.16287017e-01
-9.02880207e-02 -2.42953435e-01 4.26486395e-02 -2.21172452e-01
-3.80851120e-01 -4.84296769e-01 -1.15668523e+00 6.47458732e-01
-2.07562685e+00 -1.06593382e+00 -4.93218124e-01 2.32426286e+00
5.10988653e-01 -9.12469327e-02 4.20924455e-01 -2.29631484e-01
6.47972584e-01 2.73683250e-01 -8.15611780e-01 -3.09728235e-01
2.15706289e-01 2.67627865e-01 5.59546471e-01 3.00652325e-01
-1.13624775e+00 9.67586875e-01 7.27298880e+00 4.03645515e-01
-1.34955740e+00 9.57735926e-02 1.81087069e-02 -2.96423376e-01
1.15995975e-02 -6.54261932e-02 -8.23554933e-01 6.32953644e-01
1.08968854e+00 -1.95050880e-01 6.69946253e-01 6.97937369e-01
5.22770643e-01 -3.89770806e-01 -1.37261820e+00 1.03200626e+00
2.17121780e-01 -8.95091057e-01 -3.91277879e-01 2.31476650e-01
6.49464965e-01 3.38155001e-01 4.28071946e-01 4.46370132e-02
8.04103762e-02 -1.24571979e+00 1.09378910e+00 9.77569759e-01
4.66607869e-01 -4.55860585e-01 4.61657166e-01 3.20960671e-01
-9.98675585e-01 -1.85782075e-01 -7.33693838e-02 1.63102180e-01
4.11405981e-01 2.20137075e-01 -9.02954578e-01 6.90520853e-02
9.53346312e-01 9.62523580e-01 -1.02181530e+00 1.69724047e+00
-5.49959242e-01 5.99970996e-01 -8.82466808e-02 -1.57430828e-01
2.44235948e-01 1.51668545e-02 7.92088091e-01 1.00000834e+00
4.22596604e-01 -2.96332121e-01 -3.16587031e-01 1.10598707e+00
2.04439014e-01 1.49670213e-01 -4.15140063e-01 1.76997051e-01
4.03420597e-01 9.20706332e-01 -3.79163027e-01 1.85165361e-01
-4.33671117e-01 7.38725483e-01 4.23741579e-01 3.37534338e-01
-1.08628440e+00 -1.90316319e-01 6.16352737e-01 3.96180302e-01
5.23576319e-01 5.07234037e-02 -2.16622144e-01 -8.08431983e-01
1.57228306e-01 -4.86209095e-01 4.23780024e-01 -8.56981039e-01
-1.16146553e+00 6.39971018e-01 2.94056475e-01 -1.50871265e+00
-4.56791282e-01 -5.36814809e-01 -7.51780331e-01 9.77200150e-01
-1.80382383e+00 -1.21908367e+00 -6.07493639e-01 6.71733320e-01
5.00988185e-01 -1.09962270e-01 5.90011120e-01 6.28025830e-02
-3.43884021e-01 5.84293306e-01 -2.43130893e-01 -5.98984063e-01
9.63755369e-01 -1.30218577e+00 3.16927344e-01 8.25102627e-01
1.72826990e-01 3.14857841e-01 1.07264078e+00 -2.57275581e-01
-1.05590785e+00 -8.29200506e-01 9.19223070e-01 -7.07178771e-01
7.68559456e-01 -5.87617718e-02 -1.08729219e+00 8.21143389e-01
3.68616074e-01 -1.64321363e-01 4.98451591e-01 2.93477595e-01
-3.45969468e-01 -1.93763420e-01 -7.22014368e-01 7.45207608e-01
1.04037249e+00 -3.20287079e-01 -5.26930392e-01 1.44172177e-01
2.95672089e-01 -5.14622211e-01 -5.38852155e-01 1.20359421e-01
9.87752140e-01 -1.49369860e+00 7.80831754e-01 -4.51758921e-01
3.05217236e-01 -2.35149100e-01 2.92303562e-01 -1.32964730e+00
-5.41442215e-01 -5.51949978e-01 -1.36178657e-01 7.55865514e-01
3.54091614e-01 -4.35069412e-01 8.21105242e-01 4.76323694e-01
1.38848454e-01 -6.64062202e-01 -1.01573348e+00 -6.18681133e-01
-2.47089341e-01 -4.96520162e-01 5.96146345e-01 1.87798679e-01
-3.83289635e-01 2.46197253e-01 -5.05005896e-01 1.45776704e-01
7.66889572e-01 1.43906996e-01 9.71940577e-01 -1.58695114e+00
-3.85919303e-01 -8.13779116e-01 -4.86931175e-01 -1.04422295e+00
5.80897033e-01 -3.94743323e-01 3.60301942e-01 -1.21577930e+00
8.86027515e-02 -8.25411156e-02 -3.40703845e-01 3.66770357e-01
-1.72744840e-01 1.62945539e-01 2.62531579e-01 3.51767302e-01
-8.03197742e-01 5.47731042e-01 1.29854941e+00 1.30928680e-01
-2.02073187e-01 4.09855396e-01 -3.40017617e-01 7.27938116e-01
5.33632159e-01 -5.55663228e-01 -4.80258167e-01 -5.46181500e-01
3.84664804e-01 -6.51336694e-03 5.81176221e-01 -1.02005172e+00
5.13604641e-01 -4.18112963e-01 1.85491319e-03 -8.60342681e-01
5.55412114e-01 -8.39781046e-01 9.63236690e-02 2.01114014e-01
-3.68329287e-01 1.26719743e-01 1.77079186e-01 1.04438484e+00
-1.79065913e-01 -6.06112219e-02 9.09524381e-01 1.84520110e-01
-1.02220213e+00 5.43432176e-01 -4.11371201e-01 1.41219413e-02
1.25118375e+00 -2.52670079e-01 -1.38119489e-01 -4.84092802e-01
-6.83469892e-01 2.76356488e-01 5.10758162e-01 7.27944136e-01
3.29918444e-01 -1.06694293e+00 -4.35947001e-01 1.35305271e-01
4.44116563e-01 -4.07274872e-01 5.00384808e-01 1.35215616e+00
-3.39601457e-01 6.26952350e-01 -4.03055429e-01 -9.87669408e-01
-1.38967514e+00 4.80745286e-01 2.03086317e-01 -2.80589253e-01
-4.36165184e-01 7.39910007e-01 1.19792201e-01 1.27184555e-01
3.16343904e-01 -7.61462003e-02 -6.69854879e-01 1.09842852e-01
5.42618811e-01 5.62190831e-01 -1.57348663e-01 -8.29033256e-01
-2.52790749e-01 9.08835053e-01 -2.39095800e-02 -1.82784170e-01
1.35017121e+00 -4.10471767e-01 3.25597450e-02 6.38418317e-01
7.62726963e-01 -2.04150006e-01 -1.59231472e+00 -2.02483684e-01
1.42030343e-01 -7.36359000e-01 6.16159216e-02 -8.72913063e-01
-6.73365653e-01 1.13670576e+00 9.17890966e-01 5.72945848e-02
1.01476705e+00 1.34676099e-01 5.67742944e-01 1.72362834e-01
4.43350315e-01 -9.20069158e-01 2.14260265e-01 5.32711983e-01
8.49420011e-01 -1.39324081e+00 -1.29094347e-01 -6.79225177e-02
-7.87860334e-01 1.02972305e+00 7.06406295e-01 -1.54247507e-01
6.90387249e-01 -3.80476683e-01 5.56055754e-02 -1.66682437e-01
-1.03665709e+00 -3.82732809e-01 8.10564041e-01 6.88178122e-01
2.79636949e-01 -9.77100953e-02 -1.42020509e-01 3.06898028e-01
-2.50431240e-01 8.04892629e-02 4.23302889e-01 4.25071299e-01
-1.12551697e-01 -1.18602645e+00 -2.31713325e-01 2.15587825e-01
-3.74874741e-01 1.58431798e-01 8.13198334e-04 8.30214143e-01
1.30822867e-01 6.25773549e-01 2.73682177e-01 -3.46495837e-01
5.43051302e-01 -1.17340006e-01 5.37179589e-01 -5.95842004e-01
-2.35352382e-01 -1.28251314e-01 -2.67764091e-01 -1.14713538e+00
-5.02820551e-01 -9.88515556e-01 -7.47245729e-01 -2.45047361e-01
-4.69957218e-02 -3.84409130e-01 6.95386350e-01 8.34189653e-01
4.28928584e-01 3.75691742e-01 4.21298474e-01 -1.23643589e+00
-4.40370381e-01 -8.08639765e-01 -5.82647443e-01 4.65097576e-01
4.91539329e-01 -1.12384832e+00 -2.31690064e-01 1.31884396e-01] | [10.062260627746582, 1.2241708040237427] |
ed81a457-eaea-4573-921b-12adf3d0874c | life-is-a-circus-and-we-are-the-clowns | 2210.12197 | null | https://arxiv.org/abs/2210.12197v2 | https://arxiv.org/pdf/2210.12197v2.pdf | Life is a Circus and We are the Clowns: Automatically Finding Analogies between Situations and Processes | Analogy-making gives rise to reasoning, abstraction, flexible categorization and counterfactual inference -- abilities lacking in even the best AI systems today. Much research has suggested that analogies are key to non-brittle systems that can adapt to new domains. Despite their importance, analogies received little attention in the NLP community, with most research focusing on simple word analogies. Work that tackled more complex analogies relied heavily on manually constructed, hard-to-scale input representations. In this work, we explore a more realistic, challenging setup: our input is a pair of natural language procedural texts, describing a situation or a process (e.g., how the heart works/how a pump works). Our goal is to automatically extract entities and their relations from the text and find a mapping between the different domains based on relational similarity (e.g., blood is mapped to water). We develop an interpretable, scalable algorithm and demonstrate that it identifies the correct mappings 87% of the time for procedural texts and 94% for stories from cognitive-psychology literature. We show it can extract analogies from a large dataset of procedural texts, achieving 79% precision (analogy prevalence in data: 3%). Lastly, we demonstrate that our algorithm is robust to paraphrasing the input texts. | ['Dafna Shahaf', 'Oren Sultan'] | 2022-10-21 | null | null | null | null | ['textual-analogy-parsing', 'analogical-similarity'] | ['natural-language-processing', 'reasoning'] | [ 2.19679043e-01 2.49047279e-01 -9.71853361e-02 -2.84314364e-01
-1.32173747e-01 -8.49884391e-01 9.94369030e-01 8.04353058e-01
-2.31557459e-01 6.19276762e-01 6.37200713e-01 -6.86253428e-01
-3.41455728e-01 -7.77451456e-01 -7.40750492e-01 1.21169902e-01
1.07769221e-01 8.29107285e-01 -5.64599261e-02 -4.30270135e-01
7.24551558e-01 5.05296528e-01 -1.35979152e+00 4.54040051e-01
7.22499907e-01 5.72296977e-01 2.12177262e-02 4.89032149e-01
-2.94608027e-01 1.11932302e+00 -4.61749554e-01 -8.65743101e-01
2.01441944e-01 -6.63134456e-01 -1.30817270e+00 -5.83479941e-01
4.65893388e-01 -8.42335373e-02 -4.10571516e-01 6.72866762e-01
2.24251077e-01 2.05471992e-01 1.04236019e+00 -1.25912404e+00
-1.32558441e+00 9.17003334e-01 -3.09276342e-01 4.38789189e-01
9.99873340e-01 1.39605001e-01 1.26052487e+00 -5.13525546e-01
6.98572934e-01 1.38310504e+00 8.70489836e-01 4.86558497e-01
-1.74329698e+00 -4.56847101e-01 -1.71584681e-01 6.21350586e-01
-1.09838104e+00 -2.19765544e-01 6.42188370e-01 -3.89663130e-01
1.14266658e+00 4.60860819e-01 9.44072723e-01 1.28792834e+00
2.27906764e-01 3.86474907e-01 1.35348880e+00 -5.46408057e-01
4.56751883e-01 1.11294642e-01 1.09477043e-01 4.21265543e-01
3.61006141e-01 -9.10812989e-02 -6.18424952e-01 -2.88792878e-01
8.09804797e-01 4.79354411e-02 -2.45066077e-01 -1.06593342e-02
-1.23516250e+00 6.47432625e-01 5.16500473e-01 6.01797938e-01
-3.45756739e-01 1.11273251e-01 3.39136392e-01 5.01150608e-01
3.61132436e-04 1.35597122e+00 -3.26175213e-01 -5.59153557e-02
-6.38513207e-01 8.43858063e-01 1.18764567e+00 1.07685745e+00
5.03902793e-01 -4.63353604e-01 1.49347290e-01 5.77130437e-01
-5.92737785e-03 1.81525901e-01 6.73457503e-01 -1.26033413e+00
3.40262741e-01 6.12426996e-01 1.34519681e-01 -1.43713927e+00
-5.76062202e-01 -5.60361370e-02 -4.57504839e-01 -1.67062432e-01
5.52308500e-01 3.57640952e-01 -2.32765809e-01 1.89888155e+00
1.72851920e-01 1.92466110e-01 3.25670749e-01 8.28096271e-01
6.63811684e-01 4.96039599e-01 3.04673046e-01 3.38041969e-03
1.81148839e+00 -5.12494743e-01 -5.82207084e-01 -6.12905204e-01
6.47841930e-01 -6.83034897e-01 1.53779733e+00 1.04368933e-01
-1.21071196e+00 -2.27025315e-01 -1.08978152e+00 -4.84796166e-01
-5.34961522e-01 -6.73700213e-01 8.95184398e-01 3.39981765e-01
-7.20368803e-01 9.35449839e-01 -2.87710547e-01 -7.81342626e-01
3.38934541e-01 -1.72580943e-01 -4.27917063e-01 -1.64633095e-01
-1.40235710e+00 1.68394232e+00 4.09055829e-01 -4.26474214e-01
-1.28502786e-01 -1.06811750e+00 -8.15227866e-01 1.25595778e-01
5.21165311e-01 -1.18417966e+00 1.34529126e+00 -8.84384394e-01
-1.26486528e+00 1.11319959e+00 -3.15514877e-02 -6.25394702e-01
3.92572343e-01 -1.31045982e-01 -3.87721002e-01 -1.74668413e-02
3.40782553e-01 3.31749529e-01 1.91352561e-01 -9.68879163e-01
-9.31436941e-02 -4.24370378e-01 4.31229949e-01 2.55991936e-01
-5.46218343e-02 6.63676262e-02 2.81514287e-01 -8.79430890e-01
2.92478800e-01 -9.91242170e-01 -5.17004961e-03 3.86127800e-01
-2.71491200e-01 -2.83100903e-01 7.64140859e-02 -7.35414088e-01
1.08857596e+00 -1.93542933e+00 2.47278661e-01 -1.34588378e-02
3.01308811e-01 -3.44207525e-01 1.76307857e-01 7.92382658e-01
-3.19906175e-01 3.46347809e-01 -4.28818136e-01 8.36478919e-02
5.28005183e-01 3.50221336e-01 -7.14306235e-01 2.41908774e-01
1.59635339e-02 1.08277285e+00 -1.18050206e+00 -6.21165395e-01
8.30469429e-02 -8.19759537e-03 -6.08146012e-01 1.60526246e-01
-1.11858442e-01 -6.80636475e-03 -5.87376170e-02 2.90925920e-01
1.94686487e-01 -1.69582158e-01 2.87510753e-01 -2.84151942e-01
1.47043735e-01 7.17541218e-01 -9.74460125e-01 1.91995370e+00
-6.93341672e-01 8.50159943e-01 -8.10161471e-01 -8.59414220e-01
7.73613155e-01 2.83819348e-01 -5.30687645e-02 -6.37122333e-01
5.24840467e-02 3.18019450e-01 3.41466635e-01 -6.97841227e-01
4.72071081e-01 -8.62304688e-01 -2.92384446e-01 8.63527656e-01
-2.51628697e-01 -1.01825953e+00 4.07608300e-02 3.69444817e-01
1.17034769e+00 9.57111940e-02 1.04532433e+00 -4.45486784e-01
2.01841608e-01 2.03330681e-01 1.01659246e-01 7.56817639e-01
6.12643808e-02 3.61554503e-01 4.53341454e-01 -6.46935523e-01
-1.29271019e+00 -1.33676553e+00 -1.18719786e-01 8.70269120e-01
1.99521944e-01 -5.76926410e-01 -5.40011287e-01 -2.93080173e-02
2.24093765e-01 1.56827366e+00 -8.46531332e-01 -4.19359654e-01
-6.13177538e-01 -2.33765274e-01 7.45873690e-01 6.99683070e-01
1.86216429e-01 -1.02108300e+00 -9.52881038e-01 2.77903616e-01
-3.82105142e-01 -1.06620395e+00 -1.86597094e-01 -1.37228265e-01
-8.72857332e-01 -1.06226361e+00 -9.34472680e-02 -4.15000051e-01
2.05838248e-01 -5.81651693e-05 1.49910128e+00 -3.61077227e-02
-3.79371196e-01 2.88564861e-01 -9.71624851e-02 -6.49937868e-01
-8.26725781e-01 -4.58309084e-01 2.12241292e-01 -5.45803964e-01
4.99806106e-01 -1.09848261e+00 -3.71007025e-01 1.73409656e-01
-8.69117260e-01 3.14021587e-01 3.85117948e-01 6.36268914e-01
2.50182062e-01 -3.69095922e-01 4.46996987e-01 -7.95652032e-01
1.04730380e+00 -8.35903764e-01 4.88019213e-02 3.35684687e-01
-5.63090861e-01 3.30798596e-01 9.59875941e-01 -7.11061776e-01
-8.28841925e-01 -4.89151984e-01 3.40851992e-01 -1.38092816e-01
-1.81865349e-01 5.61049640e-01 8.62255916e-02 5.16899884e-01
1.22394311e+00 5.54949529e-02 -3.25345933e-01 -1.58634812e-01
9.57672536e-01 5.34716606e-01 9.63459671e-01 -1.18194056e+00
5.59300661e-01 4.22414780e-01 1.49707854e-01 -5.07727802e-01
-8.28433573e-01 1.77797563e-02 -5.13764918e-01 1.87317029e-01
6.72459424e-01 -4.24778312e-01 -9.28214729e-01 -2.91311532e-01
-1.37488389e+00 -1.50388822e-01 -6.66216433e-01 4.68835473e-01
-9.53969419e-01 2.88414091e-01 -4.48863566e-01 -6.05888784e-01
-1.68820754e-01 -4.66741294e-01 5.99906087e-01 7.44972005e-02
-1.40820312e+00 -1.10230136e+00 1.57800317e-01 1.78384021e-01
4.85577703e-01 3.83279026e-01 1.42996001e+00 -9.10205007e-01
-1.55228019e-01 -9.07837451e-02 -6.85540140e-02 -3.87843549e-01
3.00833136e-01 -1.68828085e-01 -7.60757208e-01 4.14108157e-01
3.56830299e-01 -4.04679447e-01 2.40143120e-01 -1.45684257e-01
1.07977736e+00 -6.69645727e-01 -3.46453428e-01 2.39227578e-01
1.03423154e+00 8.14363733e-02 6.76834345e-01 3.40316981e-01
3.24434429e-01 9.97332811e-01 2.46146873e-01 2.60757446e-01
6.79560483e-01 5.90292931e-01 -1.53038710e-01 3.83475661e-01
6.81073293e-02 -4.58149225e-01 9.89447348e-03 4.82548594e-01
-2.11669177e-01 -2.07609199e-02 -1.20564759e+00 6.29957974e-01
-1.78054047e+00 -1.33614349e+00 -5.34580126e-02 1.86676002e+00
1.28804386e+00 1.73432782e-01 -1.66705415e-01 2.01652259e-01
4.15466398e-01 -2.11090192e-01 -6.32127583e-01 -7.76008308e-01
-1.28311515e-01 4.87320870e-01 -1.33542851e-01 3.93553674e-01
-5.62614083e-01 7.88937807e-01 6.27785969e+00 3.80613804e-01
-6.42172813e-01 -2.01686710e-01 3.45030159e-01 -6.40360191e-02
-6.21561408e-01 1.79518312e-01 8.02267268e-02 2.14401275e-01
1.11115479e+00 -7.62549937e-01 7.55350232e-01 4.96650517e-01
1.42210834e-02 -5.56112044e-02 -2.05139685e+00 1.12978435e+00
2.37112314e-01 -1.41591465e+00 3.87491971e-01 -5.89931250e-01
3.42146516e-01 -4.73074883e-01 -1.63354129e-01 1.24213628e-01
5.36404073e-01 -1.30998611e+00 1.01618767e+00 6.11990631e-01
4.26822126e-01 -2.44389176e-01 3.84364605e-01 3.27692628e-01
-9.59890068e-01 -8.38601440e-02 -3.15865010e-01 -7.42027700e-01
6.44940808e-02 1.31045580e-01 -9.20807719e-01 4.86829132e-01
5.92760265e-01 7.39710867e-01 -4.49912250e-01 6.24548376e-01
-4.95192498e-01 3.00807327e-01 -3.47385734e-01 -3.24324787e-01
-1.89084813e-01 -3.98946032e-02 4.73025262e-01 1.18998122e+00
2.00919762e-01 6.43377662e-01 -3.50272745e-01 1.56560135e+00
-1.53227393e-02 1.10131532e-01 -1.01183414e+00 -1.62171260e-01
8.84104669e-01 7.45269418e-01 -6.06091917e-01 -5.37432194e-01
-2.72428006e-01 1.04035747e+00 5.71165979e-01 1.16172120e-01
-8.74994397e-01 -2.49786347e-01 8.11876059e-01 8.84423256e-02
-3.09248447e-01 -1.91688426e-02 -7.13705480e-01 -1.03663969e+00
1.52570382e-01 -9.16749299e-01 4.81631160e-01 -1.26417553e+00
-1.95627844e+00 2.81552523e-01 3.00352901e-01 -8.24192047e-01
-4.53390628e-01 -5.59300482e-01 -6.32240772e-01 1.03756905e+00
-8.21509123e-01 -8.10516119e-01 -2.85097867e-01 3.68668646e-01
4.49787140e-01 3.06397974e-01 1.19090188e+00 -2.07126364e-01
-2.03436520e-02 4.29207295e-01 -3.46304715e-01 -1.24586821e-01
7.23786771e-01 -1.32825732e+00 9.09256756e-01 2.89078623e-01
2.71990240e-01 1.22028959e+00 1.16371512e+00 -5.07363796e-01
-1.50696146e+00 -5.29387891e-01 1.05821538e+00 -9.11613345e-01
1.20488262e+00 -2.41407320e-01 -1.23819339e+00 1.05621815e+00
1.90358669e-01 -2.52374470e-01 9.53159392e-01 4.67120633e-02
-9.76756513e-01 2.20662773e-01 -1.22284114e+00 1.20909572e+00
1.50581896e+00 -9.91434455e-01 -1.70603633e+00 4.09789950e-01
8.68942857e-01 -5.01842976e-01 -9.21212912e-01 -1.85778081e-01
8.12998652e-01 -7.56142318e-01 1.04302478e+00 -1.23896062e+00
1.02009273e+00 -2.28127554e-01 -3.71968210e-01 -1.44087458e+00
-3.32914680e-01 -5.34270763e-01 9.40497220e-03 8.71495068e-01
3.98231357e-01 -8.64219904e-01 1.78632975e-01 1.33105886e+00
-5.72954193e-02 -7.17884243e-01 -8.50216031e-01 -8.38268280e-01
3.06766182e-01 -4.83179778e-01 8.18086445e-01 1.50700986e+00
7.18111873e-01 6.18290782e-01 8.02564695e-02 -7.52512589e-02
2.90394217e-01 5.53636730e-01 5.92256486e-01 -1.36167824e+00
-5.86024165e-01 -7.36216784e-01 -5.26357472e-01 -6.89261913e-01
2.85858512e-01 -1.27563083e+00 -2.96176493e-01 -1.72150242e+00
3.22666287e-01 -3.49076480e-01 2.04714552e-01 4.97823030e-01
-2.05369785e-01 -1.36761844e-01 3.40898424e-01 2.42433220e-01
-8.40908382e-03 2.94683218e-01 9.63299751e-01 -2.27579564e-01
3.69578302e-02 -5.26922703e-01 -1.19291949e+00 8.81599247e-01
7.18528152e-01 -6.33678079e-01 -5.24223089e-01 -3.32950890e-01
5.52533686e-01 8.39476362e-02 6.98660135e-01 -7.02131212e-01
2.81750321e-01 -5.40101588e-01 2.45682791e-01 2.14292675e-01
3.29656899e-01 -8.74553144e-01 5.92750490e-01 5.74464142e-01
-6.37964368e-01 5.96756518e-01 4.54441279e-01 5.99788666e-01
2.68480927e-01 -2.93641359e-01 4.80057895e-01 -2.79014438e-01
-5.07341862e-01 -4.91288155e-01 -1.90229818e-01 5.15609264e-01
9.86349940e-01 -2.08851412e-01 -6.11966074e-01 -3.65491867e-01
-5.18745840e-01 -2.53879353e-02 6.83553159e-01 4.67024654e-01
5.63929737e-01 -1.22654045e+00 -9.04006362e-01 -4.82445151e-01
2.31788754e-01 -3.12216640e-01 4.09352127e-03 6.07626855e-01
-6.95997179e-01 1.96588770e-01 -5.01253009e-01 -1.28281161e-01
-8.51637542e-01 8.64168227e-01 4.13880259e-01 1.74405470e-01
-6.86360776e-01 5.74558437e-01 1.66126415e-01 -4.01095510e-01
-3.28369915e-01 -6.88907504e-01 4.54569720e-02 -2.08137572e-01
5.11053801e-01 4.76847529e-01 -3.33803743e-01 -2.13777989e-01
-5.60580373e-01 4.74068999e-01 -6.63692877e-03 -2.67773002e-01
1.29892957e+00 9.49107036e-02 -4.88637924e-01 1.08130705e+00
8.04417849e-01 -1.49485797e-01 -5.01609981e-01 -2.78754801e-01
1.88411668e-01 -5.13306439e-01 -6.33345962e-01 -9.01176155e-01
-3.23028192e-02 7.20385790e-01 -1.88915774e-01 4.00418341e-01
7.45613873e-01 4.48143572e-01 5.71646035e-01 8.74042153e-01
3.53614300e-01 -6.63375139e-01 1.21797964e-01 3.00220430e-01
1.34735632e+00 -7.74728537e-01 4.65065867e-01 -2.63419330e-01
-5.88872313e-01 1.12392771e+00 4.07559454e-01 -6.26662970e-02
2.32595742e-01 1.09211639e-01 -3.23234797e-01 -3.40789407e-01
-7.52196729e-01 4.73689586e-01 1.96409062e-01 5.09118140e-01
6.68799400e-01 2.75571287e-01 -3.80761117e-01 8.42712760e-01
-9.61034536e-01 -2.72356659e-01 8.00848067e-01 7.58403599e-01
-3.95032503e-02 -6.52308524e-01 -3.08085382e-01 5.12730420e-01
3.40174213e-02 -4.81360018e-01 -8.56081367e-01 9.06645596e-01
-1.31069958e-01 7.68356621e-01 2.27764368e-01 -1.08520880e-01
5.92197955e-01 2.24666834e-01 9.17108238e-01 -5.33447444e-01
-6.28057063e-01 -9.74441171e-01 1.99761093e-01 -5.55625916e-01
-4.49902415e-01 -7.51547992e-01 -1.59362841e+00 -8.34618151e-01
1.46071061e-01 8.26609135e-02 4.28804278e-01 1.13624883e+00
3.48942220e-01 2.85280794e-01 -7.04094991e-02 -6.11192346e-01
-4.70312893e-01 -8.81904304e-01 -3.35834593e-01 8.43844056e-01
-3.79250683e-02 -7.26701558e-01 -3.23946208e-01 1.68901712e-01] | [10.5869722366333, 2.4113900661468506] |
4ff7685d-a47b-417a-bd53-c3048bd6ff81 | textual-representations-for-crosslingual | null | null | https://aclanthology.org/2021.ecnlp-1.14 | https://aclanthology.org/2021.ecnlp-1.14.pdf | Textual Representations for Crosslingual Information Retrieval | In this paper, we explored different levels of textual representations for cross-lingual information retrieval. Beyond the traditional token level representation, we adopted the subword and character level representations for information retrieval that had shown to improve neural machine translation by reducing the out-of-vocabulary issues in machine translation. We found that crosslingual information retrieval performance can be improved by combining search results from subwords and token level representation.Additionally, we improved the search performance by combining and re-ranking the result sets from the different text representations for German, French and Japanese. | ['Liling Tan', 'Hang Zhang'] | null | null | null | null | acl-ecnlp-2021-8 | ['cross-lingual-information-retrieval'] | ['natural-language-processing'] | [ 1.56347185e-01 -2.19840378e-01 -6.80361032e-01 -1.32608684e-02
-1.80351186e+00 -8.04864109e-01 9.45495963e-01 4.73251522e-01
-6.77283764e-01 8.88673961e-01 6.63778186e-01 -5.39277673e-01
-2.34600171e-01 -5.62802613e-01 -4.95040417e-01 -9.50442627e-02
5.15038252e-01 6.64332151e-01 -1.89700529e-01 -6.28325999e-01
6.70203328e-01 3.78042400e-01 -1.34580600e+00 8.96825194e-01
1.05721414e+00 7.09123790e-01 1.68030053e-01 1.55179456e-01
-5.47725797e-01 7.77659938e-02 -7.53113687e-01 -3.27283978e-01
2.24535108e-01 -3.56847316e-01 -9.82998610e-01 -6.61905348e-01
1.01239669e+00 1.12988792e-01 -4.18838739e-01 9.13523018e-01
7.70496309e-01 1.83819085e-01 9.11823094e-01 -4.43200767e-01
-1.44573891e+00 8.04697633e-01 -4.96182442e-01 5.84981322e-01
5.98629415e-01 -5.03840089e-01 1.19152999e+00 -1.26720405e+00
9.25682068e-01 1.71361554e+00 5.41028202e-01 5.63110173e-01
-1.03287494e+00 -4.75907803e-01 -9.91494879e-02 2.44954303e-01
-1.36835229e+00 -4.04390395e-01 2.85482645e-01 -1.64806858e-01
1.93422139e+00 2.18235537e-01 8.27309564e-02 1.05494857e+00
5.95178545e-01 7.17582881e-01 1.02784145e+00 -1.07661605e+00
-5.63475788e-01 1.17710151e-01 3.48503917e-01 5.22919059e-01
4.72700596e-01 9.73861292e-02 -5.37883043e-01 -2.49617890e-01
5.27470887e-01 -2.75106132e-01 5.86220510e-02 2.99192846e-01
-1.36453187e+00 1.09539568e+00 3.65136206e-01 9.48210716e-01
-4.50720400e-01 3.18531543e-01 9.06271756e-01 6.36495948e-01
6.74897075e-01 1.03511882e+00 -8.12277675e-01 4.41728719e-02
-8.86711180e-01 -1.17756076e-01 5.12208939e-01 9.40177739e-01
6.13574326e-01 6.77341893e-02 -7.12866426e-01 1.30102909e+00
2.41355728e-02 6.86629176e-01 1.22978306e+00 -4.64530528e-01
1.04345882e+00 4.98758376e-01 -2.60218382e-01 -8.35086405e-01
-3.11399009e-02 -3.66790920e-01 -4.36849028e-01 -6.21403217e-01
-1.01833992e-01 2.51801368e-02 -1.10617507e+00 1.40843391e+00
-4.59014148e-01 -7.84441531e-01 4.22093689e-01 6.07090473e-01
9.32145774e-01 9.62866306e-01 6.95842057e-02 -2.03718737e-01
1.42847800e+00 -1.02962506e+00 -1.10670280e+00 -4.16532815e-01
1.15194702e+00 -1.20729244e+00 8.53124440e-01 -9.59469602e-02
-1.13951850e+00 -7.49032378e-01 -9.78836119e-01 -4.94130045e-01
-1.06916976e+00 1.81440711e-01 6.02523685e-01 3.03061754e-01
-1.27100825e+00 4.63311970e-01 -2.13652700e-01 -6.24184251e-01
-7.91465491e-02 4.13595229e-01 -4.32318598e-01 -3.13441277e-01
-1.83300614e+00 1.66558695e+00 5.62095821e-01 -2.79450625e-01
-2.91903228e-01 -4.62428540e-01 -9.82488930e-01 7.85991270e-03
-5.87198660e-02 -6.49679542e-01 9.58185494e-01 -7.38538444e-01
-1.11304617e+00 1.03229547e+00 -4.43408728e-01 -3.71726304e-01
-3.00729841e-01 -4.00945902e-01 -4.22702014e-01 9.40003321e-02
3.80435526e-01 8.91719878e-01 5.06962001e-01 -8.25807333e-01
-2.52950191e-01 -3.97757471e-01 -2.78577000e-01 5.59060872e-01
-2.29689375e-01 5.42769969e-01 -5.99808395e-01 -1.00469995e+00
-6.65276721e-02 -9.39081132e-01 -3.44227329e-02 -7.11155534e-01
1.90776616e-01 -7.53661692e-01 4.29367989e-01 -1.07867897e+00
1.18352842e+00 -1.72041190e+00 3.83984387e-01 -7.99636990e-02
-4.55349892e-01 4.93862242e-01 -5.58338821e-01 9.64128315e-01
7.68024996e-02 8.61185253e-01 3.26296538e-01 2.37465665e-01
-1.56558990e-01 2.17912272e-01 -3.07826042e-01 -5.47486432e-02
2.85850585e-01 1.42426729e+00 -7.51226246e-01 -6.90193832e-01
2.73818467e-02 5.29225349e-01 -2.54559278e-01 -3.28472823e-01
6.85162023e-02 -2.39199519e-01 -6.11627758e-01 5.94064713e-01
1.51763812e-01 2.00423330e-01 1.35859191e-01 1.04948580e-01
4.62437756e-02 1.13846242e+00 -2.70312160e-01 2.15168142e+00
-8.15255761e-01 1.00541604e+00 -5.57154715e-01 -7.94029951e-01
9.90214527e-01 6.20629251e-01 3.00403565e-01 -1.49887729e+00
4.26602550e-02 5.61612546e-01 -2.15242952e-02 -2.41369426e-01
9.67222512e-01 3.82231660e-02 -5.18356025e-01 5.61963141e-01
3.49145651e-01 -2.54203141e-01 2.47087374e-01 1.63775589e-02
6.02829337e-01 3.55396807e-01 4.41688031e-01 -4.57342178e-01
4.05651391e-01 3.79178554e-01 1.00837005e-02 7.06226051e-01
2.46310160e-01 3.64495754e-01 -9.36469212e-02 -2.59093672e-01
-9.77301121e-01 -8.61105204e-01 -2.68586636e-01 1.35786927e+00
-2.02591151e-01 -6.21899903e-01 -6.44350350e-01 -5.31213760e-01
8.69918615e-02 7.24740684e-01 -4.97486025e-01 -4.40493554e-01
-8.42071414e-01 -5.23720980e-01 9.01601553e-01 3.13978553e-01
4.93672974e-02 -9.63705301e-01 1.70591250e-02 2.91856706e-01
-4.54580247e-01 -8.92496407e-01 -7.58238792e-01 5.20138144e-01
-1.29887247e+00 -3.23031604e-01 -7.89213419e-01 -1.23300493e+00
3.27194065e-01 3.47689390e-01 1.30418122e+00 2.63278812e-01
-4.16103184e-01 2.76555091e-01 -6.57592177e-01 -2.55591691e-01
-4.72368032e-01 5.10248423e-01 1.89084589e-01 -7.86583304e-01
7.57091105e-01 3.32862109e-01 -3.17731142e-01 -1.08437277e-01
-9.30945635e-01 -5.14221668e-01 8.57909739e-01 9.87491846e-01
5.42304873e-01 -5.32616615e-01 7.27313936e-01 -5.52238703e-01
1.48491728e+00 -2.85785168e-01 -9.12485868e-02 6.36102438e-01
-8.59952390e-01 6.38197660e-01 2.74926782e-01 -5.53389311e-01
-7.51293540e-01 -4.94647682e-01 1.59032166e-01 -8.24323893e-02
1.87310621e-01 9.91278887e-01 5.50821602e-01 -8.55496824e-02
7.82881141e-01 3.87050450e-01 -1.38517678e-01 -7.40958571e-01
6.42741382e-01 8.35018456e-01 -2.20766351e-01 -7.81745017e-01
3.56527269e-01 -4.92025405e-01 -4.46314782e-01 -6.86231673e-01
-4.99830991e-01 -4.69245762e-01 -9.40220475e-01 1.06808245e-01
9.04073536e-01 -8.49407136e-01 7.57632330e-02 -4.34517898e-02
-1.45873046e+00 3.93233180e-01 -1.81350753e-01 6.38362646e-01
-1.87344268e-01 4.28548247e-01 -8.64933610e-01 -2.97735572e-01
-6.12418175e-01 -1.09555519e+00 1.33706939e+00 -2.14039981e-01
-4.17224646e-01 -1.22072661e+00 3.29668760e-01 3.99291128e-01
6.05944872e-01 -3.28844666e-01 1.45324433e+00 -9.12636399e-01
-3.90285477e-02 -3.44527215e-01 -1.10720612e-01 2.49151722e-01
3.85418981e-01 -3.49922687e-01 -7.15756536e-01 -3.68757993e-01
-1.88471496e-01 -4.98747230e-01 1.02844870e+00 2.26393670e-01
5.55769265e-01 -4.30247068e-01 -4.08139586e-01 5.93302958e-02
1.43204868e+00 3.93933058e-01 6.73770189e-01 6.13814354e-01
4.64126289e-01 7.76170135e-01 5.58449686e-01 -4.73111987e-01
-8.23031068e-02 1.09636199e+00 -3.73494804e-01 -1.22886993e-01
-5.48610389e-01 -3.40583563e-01 5.53940356e-01 1.43840647e+00
8.04896057e-02 -2.71116912e-01 -8.89923394e-01 6.66219771e-01
-1.74416506e+00 -7.94086099e-01 5.10956831e-02 1.94324732e+00
1.22365201e+00 -1.69601962e-01 -3.87183070e-01 -4.19469297e-01
6.68529391e-01 -6.31345510e-02 -2.26914749e-01 -1.29453599e+00
-2.11575091e-01 6.23315990e-01 7.21657217e-01 7.05056310e-01
-6.10275388e-01 1.67368567e+00 7.69904804e+00 1.20686162e+00
-9.19203222e-01 3.14389110e-01 2.61442095e-01 -1.92099586e-02
-4.50982094e-01 -2.63312697e-01 -9.62018132e-01 7.92615414e-02
1.22402227e+00 -5.44564724e-01 3.76384705e-01 4.48764056e-01
-2.25709677e-01 2.63928175e-01 -1.27303326e+00 8.77804220e-01
5.06760418e-01 -1.38771355e+00 1.06212449e+00 1.49174035e-01
8.80823910e-01 3.05105418e-01 6.84110150e-02 7.04969704e-01
2.84983814e-01 -1.18731880e+00 1.96018055e-01 4.03025031e-01
8.68278682e-01 -6.69862151e-01 8.53612959e-01 1.22953124e-01
-9.97221291e-01 3.32237512e-01 -6.99421167e-01 7.17741847e-02
-2.13743836e-01 -1.67782873e-01 -8.49538028e-01 9.31783617e-01
3.35027814e-01 5.93819559e-01 -8.28661561e-01 6.75988853e-01
3.85022722e-02 6.48324192e-02 -3.72148380e-02 -2.39804566e-01
6.79538012e-01 -1.56341210e-01 2.40905225e-01 1.63360512e+00
3.69168013e-01 -4.31857765e-01 1.57287847e-02 4.36595887e-01
-2.57466555e-01 6.44255221e-01 -1.23013139e+00 -4.02435750e-01
5.10561943e-01 6.99997246e-01 -3.52116942e-01 -5.34177780e-01
-4.65626597e-01 1.18906903e+00 4.60173815e-01 4.60015357e-01
-2.04117030e-01 -5.76905012e-01 7.76015818e-01 -3.19624603e-01
1.25581369e-01 -3.55265647e-01 -2.65015572e-01 -1.29166996e+00
-1.19988464e-01 -1.13394177e+00 5.59911013e-01 -7.06928611e-01
-1.32577574e+00 8.99164975e-01 1.63376927e-01 -1.16385484e+00
-7.35905886e-01 -8.20231616e-01 3.33090276e-02 1.29125631e+00
-1.38458180e+00 -1.08204198e+00 8.82607877e-01 4.45112348e-01
1.06697857e+00 -5.76370776e-01 1.29480243e+00 4.07576293e-01
-6.96150213e-02 9.29636598e-01 4.16524172e-01 2.09111527e-01
8.84333670e-01 -7.85091221e-01 6.34624839e-01 6.20773494e-01
6.18875742e-01 1.13791215e+00 9.73686427e-02 -7.80111194e-01
-1.62396181e+00 -6.37743950e-01 1.60465097e+00 -6.01950884e-01
7.23158598e-01 -1.44459799e-01 -8.51494193e-01 5.69419742e-01
8.62576783e-01 -7.21522331e-01 6.44930124e-01 3.29976082e-01
-6.28673077e-01 1.30209848e-01 -8.79331350e-01 7.71152973e-01
7.04710007e-01 -1.22846699e+00 -1.36627769e+00 5.12369394e-01
1.15728676e+00 -3.06280833e-02 -1.00034249e+00 3.35829109e-01
5.10078490e-01 2.93961853e-01 1.21774888e+00 -1.10447836e+00
2.63989687e-01 2.89029002e-01 -3.44494045e-01 -1.55395401e+00
-6.59558713e-01 -2.31745869e-01 3.61396223e-01 1.06733668e+00
9.51501667e-01 -7.28045166e-01 -5.29725067e-02 -5.01605161e-02
-2.19612122e-01 -6.00033939e-01 -1.26442540e+00 -9.01928186e-01
9.53681886e-01 -1.02260709e-01 1.99229792e-01 9.14448857e-01
3.32528323e-01 9.43988144e-01 -1.85367793e-01 -4.72777992e-01
3.19305770e-02 -3.68840583e-02 -5.02023334e-03 -1.20813966e+00
1.89843893e-01 -7.52341747e-01 -3.12125832e-01 -9.52344596e-01
4.29792434e-01 -1.77959156e+00 1.14673175e-01 -1.97047246e+00
2.61651814e-01 7.02691451e-02 -7.00732708e-01 5.86013138e-01
-2.26034850e-01 5.49848616e-01 1.26921400e-01 4.83194739e-01
-2.56065041e-01 3.02686155e-01 1.30570710e+00 -4.67849404e-01
1.69957787e-01 -8.09373438e-01 -8.44005287e-01 7.66878575e-02
8.29836965e-01 -7.70571589e-01 -1.49775386e-01 -1.16271162e+00
4.34462428e-01 -1.34952944e-02 -4.02981579e-01 -4.33329284e-01
5.28648049e-02 -4.34229933e-02 4.20731574e-01 -5.81650734e-01
4.46376830e-01 -6.31042778e-01 -4.94334221e-01 5.73314548e-01
-9.14919913e-01 7.57355154e-01 5.65725386e-01 2.51836061e-01
-4.61346686e-01 -5.30414760e-01 3.74131262e-01 -4.92588758e-01
-5.60357034e-01 -1.15037911e-01 -1.02518487e+00 1.56292856e-01
4.22742307e-01 -1.14079982e-01 -4.86166924e-01 -1.56380817e-01
-3.54208618e-01 7.89660737e-02 2.13316694e-01 1.18439245e+00
7.32771993e-01 -1.67253792e+00 -9.17724192e-01 -1.74481533e-02
4.39423382e-01 -9.40954268e-01 -6.65547073e-01 4.05992061e-01
-3.05751055e-01 1.39039958e+00 -5.07817388e-01 -1.73814014e-01
-1.24557126e+00 4.62470174e-01 1.25847250e-01 -5.58569849e-01
-2.35793993e-01 5.61783016e-01 -2.07434356e-01 -7.29959190e-01
1.79711118e-01 -3.68773192e-01 -4.96492237e-01 5.10209858e-01
3.04774880e-01 1.55169293e-01 4.78085667e-01 -7.65007377e-01
-5.48516273e-01 9.05046999e-01 -6.48367703e-01 -4.19379324e-01
9.87928689e-01 -1.16117485e-01 -4.19551551e-01 6.90265417e-01
1.48654842e+00 7.22363684e-03 1.64957702e-01 -4.89053369e-01
5.49763560e-01 -2.25208759e-01 2.34179735e-01 -1.11747718e+00
-6.06691420e-01 9.61126328e-01 6.23217225e-01 -7.73025900e-02
6.80835545e-01 7.51131773e-02 9.43190396e-01 1.09907269e+00
5.31778574e-01 -1.53819799e+00 -1.11876652e-01 1.02759683e+00
1.09914815e+00 -1.11649823e+00 -1.18434280e-02 1.12156145e-01
-4.81457263e-01 1.34400928e+00 2.36752778e-01 -1.48975238e-01
1.47333920e-01 -5.02739027e-02 2.85510927e-01 -2.27513880e-01
-9.78213131e-01 -2.43117347e-01 1.00961626e+00 4.67883825e-01
1.11943257e+00 -1.22459903e-02 -1.07820964e+00 4.49148789e-02
-9.87133682e-02 -4.08344448e-01 3.54124457e-02 9.02942598e-01
-4.74522978e-01 -1.56605649e+00 -2.55655855e-01 5.17539322e-01
-9.68244910e-01 -1.04207647e+00 -1.13073015e+00 7.03570485e-01
-3.74216914e-01 1.04570282e+00 2.00459156e-02 -3.29787463e-01
1.14546314e-01 6.12858951e-01 6.02321267e-01 -9.95337427e-01
-1.04029977e+00 2.00603288e-02 4.63774115e-01 -3.75604749e-01
-2.09528625e-01 -4.08148497e-01 -8.41271341e-01 1.83701426e-01
-4.55520630e-01 6.17220402e-01 1.04726064e+00 8.87315333e-01
5.76451421e-01 4.07017171e-01 3.21496546e-01 -6.70517385e-01
-4.57579225e-01 -1.54693675e+00 1.37214363e-01 1.23015888e-01
1.35083035e-01 -4.29177672e-01 -2.02582374e-01 -2.40853250e-01] | [11.417068481445312, 9.879083633422852] |
0de7ad92-5c87-4a08-aed6-e94f9cb01538 | lung-nodule-segmentation-via-level-set | 1910.03191 | null | https://arxiv.org/abs/1910.03191v3 | https://arxiv.org/pdf/1910.03191v3.pdf | Level set image segmentation with velocity term learned from data with applications to lung nodule segmentation | Purpose: Lung nodule segmentation, i.e., the algorithmic delineation of the lung nodule surface, is a fundamental component of computational nodule analysis pipelines. We propose a new method for segmentation that is a machine learning based extension of current approaches, using labeled image examples to improve its accuracy. Approach: We introduce an extension of the standard level set image segmentation method where the velocity function is learned from data via machine learning regression methods, rather than a priori designed. Instead, the method employs a set of features to learn a velocity function that guides the level set evolution from initialization. Results: We apply the method to image volumes of lung nodules from CT scans in the publicly available LIDC dataset, obtaining an average intersection over union score of 0.7185($\pm$0.1114), which is competitive with other methods. We analyze segmentation performance by anatomical and appearance-based categories of the nodules, finding that the method performs better for isolated nodules with well-defined margins. We find that the segmentation performance for nodules in more complex surroundings and having more complex CT appearance is improved with the addition of combined global-local features. Conclusions: The level set machine learning segmentation approach proposed herein is competitive with current methods. It provides accurate lung nodule segmentation results in a variety of anatomical contexts. | ['Matthew C Hancock', 'Jerry F Magnan'] | 2019-10-08 | null | null | null | null | ['lung-nodule-segmentation'] | ['medical'] | [ 2.30136827e-01 3.95032972e-01 -3.69089842e-01 -1.92882776e-01
-1.00312471e+00 -7.49264061e-01 4.74196643e-01 1.34066314e-01
-3.13869268e-01 1.85697302e-01 -2.33244207e-02 -5.76636016e-01
-1.70501366e-01 -4.95482862e-01 -3.76733750e-01 -6.77132905e-01
1.32334686e-03 9.59068418e-01 8.09682488e-01 1.13349408e-02
8.70620906e-02 7.36165524e-01 -9.27946746e-01 2.77722597e-01
7.33291030e-01 7.95834839e-01 1.44400761e-01 9.60818529e-01
-1.37076080e-01 5.52161396e-01 -2.09939450e-01 -2.15090532e-02
4.55664903e-01 -7.40003467e-01 -1.09658587e+00 5.22326767e-01
3.55039656e-01 -4.49039228e-02 1.41693756e-01 6.60547972e-01
2.54421562e-01 -3.60088795e-01 1.19263744e+00 -6.49818122e-01
3.19005139e-02 2.96263725e-01 -6.55246854e-01 3.92334521e-01
-2.28727758e-01 3.82107973e-01 1.01759517e+00 -6.78059459e-01
6.85872555e-01 6.70202374e-01 1.00376368e+00 4.59475428e-01
-1.22949088e+00 -8.91049132e-02 -3.26754361e-01 -3.26196164e-01
-1.20234835e+00 2.72441149e-01 1.20219022e-01 -9.34526384e-01
5.93137622e-01 4.23842162e-01 9.33001041e-01 1.27682418e-01
2.81748146e-01 5.51858902e-01 9.13597643e-01 -5.30578017e-01
1.12767346e-01 1.86604306e-01 -9.87416506e-02 1.24415886e+00
5.35590589e-01 6.77214265e-02 3.86336982e-01 -2.40408331e-01
1.17491114e+00 -1.21831343e-01 -1.14869207e-01 -7.37483978e-01
-1.40504074e+00 9.22147393e-01 6.97307646e-01 6.34129524e-01
-3.97691727e-01 4.53494072e-01 1.41095698e-01 -3.07040334e-01
2.62958705e-01 6.25032485e-01 -3.08587641e-01 1.40923008e-01
-1.26129472e+00 -3.47712822e-02 9.80913937e-01 4.88999814e-01
6.49485409e-01 -3.34841222e-01 -5.68193018e-01 5.26785910e-01
5.30961275e-01 1.45411491e-01 5.49594104e-01 -1.09081960e+00
-2.93863297e-01 9.15177941e-01 -7.19523281e-02 -3.13751817e-01
-6.56884193e-01 -4.41761643e-01 -3.85740578e-01 4.20284837e-01
7.47631013e-01 -3.49622290e-03 -1.42369652e+00 9.52635229e-01
3.57220322e-01 9.35849100e-02 -2.74980843e-01 6.12936616e-01
9.77895856e-01 6.91031590e-02 3.09369769e-02 -1.67691186e-01
1.51846385e+00 -1.07675278e+00 -2.72221237e-01 -6.07526079e-02
9.63954806e-01 -7.00895607e-01 1.07682550e+00 6.86036944e-02
-1.01814020e+00 -2.02655524e-01 -7.75294244e-01 3.22172105e-01
3.41772131e-04 2.77808219e-01 4.87636179e-01 1.00564063e+00
-1.29074132e+00 4.80936348e-01 -1.19322002e+00 -5.66016138e-01
5.98736107e-01 7.19586909e-01 5.12297079e-02 2.98933029e-01
-4.18714046e-01 9.23758686e-01 3.48962367e-01 -4.62515265e-01
-5.61777830e-01 -1.02906358e+00 -5.34216523e-01 -1.70268521e-01
3.94402474e-01 -6.50807142e-01 1.74049222e+00 -7.37155199e-01
-1.06639767e+00 1.15995169e+00 -5.32771982e-02 -4.72386032e-01
7.13787138e-01 3.15218091e-01 2.65511960e-01 3.80321532e-01
1.17348447e-01 9.18179929e-01 7.00408518e-01 -1.33741331e+00
-7.13915050e-01 1.91344842e-01 -4.13834244e-01 1.19934313e-01
-1.29678063e-02 -2.44613945e-01 -7.01194346e-01 -2.74179846e-01
2.17879102e-01 -1.05880940e+00 -8.58787119e-01 1.90743893e-01
-2.79497564e-01 -1.69883937e-01 8.15167964e-01 -5.50854862e-01
1.28837991e+00 -1.59365368e+00 -3.62954706e-01 6.41020894e-01
5.75553656e-01 5.90173863e-02 4.55893099e-01 -2.59787917e-01
8.14330354e-02 5.83415389e-01 -6.71684921e-01 3.28958631e-01
-3.08279663e-01 1.43534794e-01 6.52128935e-01 4.51126218e-01
2.30108902e-01 1.37177277e+00 -8.28341663e-01 -1.11676884e+00
3.09447378e-01 1.54820472e-01 -7.52618134e-01 -6.86264597e-03
-3.02170247e-01 5.45766532e-01 -3.82086158e-01 6.15387321e-01
3.01483035e-01 -8.03080678e-01 1.24103546e-01 -1.79028347e-01
-1.77601397e-01 -1.94074959e-01 -8.98163676e-01 1.32859051e+00
-1.21715240e-01 4.09792542e-01 -6.63736090e-02 -4.91035283e-01
7.12788641e-01 4.36976016e-01 1.25660717e+00 1.20178081e-01
3.27419937e-01 4.64119613e-01 7.19651997e-01 -4.72240359e-01
-2.34358177e-01 -3.39620292e-01 1.43563911e-01 5.02273262e-01
-7.67892078e-02 -9.02732491e-01 4.23511237e-01 1.04083546e-01
1.53615499e+00 -1.22074038e-01 7.34198451e-01 -6.38936281e-01
5.74465334e-01 6.29358649e-01 7.74689615e-02 7.82585800e-01
-4.43267763e-01 9.82680678e-01 6.79010391e-01 -1.63930818e-01
-1.11852252e+00 -9.47034180e-01 -4.99157459e-01 5.67753494e-01
-5.90474159e-02 -2.01654300e-01 -9.86225724e-01 -1.07823956e+00
3.44497971e-02 2.87942350e-01 -8.88137281e-01 1.95010290e-01
-8.16849232e-01 -8.44672620e-01 4.06526744e-01 5.04242301e-01
2.79496342e-01 -1.06142795e+00 -8.62257302e-01 2.29729399e-01
3.08151133e-02 -9.15795028e-01 -5.10165989e-01 2.68666834e-01
-1.19991815e+00 -1.38273120e+00 -8.11951280e-01 -9.03230250e-01
9.27944481e-01 -9.89085902e-03 1.36640155e+00 5.46780884e-01
-9.67787743e-01 6.79755986e-01 -1.11065231e-01 -3.32704335e-01
-8.02447200e-01 1.90276310e-01 -5.40679872e-01 -3.92222852e-01
-3.05951424e-02 1.43738583e-01 -7.47344911e-01 4.57245231e-01
-9.19552743e-01 -5.90055622e-02 1.02606392e+00 8.39302599e-01
9.83450830e-01 1.38672581e-02 5.45645542e-02 -1.23079252e+00
3.43792766e-01 -4.73132312e-01 -5.08069575e-01 1.60996154e-01
-5.91355205e-01 1.29567817e-01 2.17133269e-01 -2.88706899e-01
-8.76278639e-01 4.21866238e-01 -3.86086106e-02 -2.65606344e-01
-1.36129573e-01 3.07841063e-01 4.15279776e-01 -4.80767161e-01
1.01764333e+00 -1.05462372e-01 4.40689862e-01 -4.37674411e-02
2.16482520e-01 2.14572415e-01 3.13276798e-01 -1.65038690e-01
1.05479538e+00 7.52889454e-01 2.21073598e-01 -6.93375826e-01
-7.39867210e-01 -9.08884883e-01 -1.08055615e+00 -3.89209628e-01
1.01346755e+00 -2.25539669e-01 -3.48370343e-01 -1.83649659e-01
-5.51925063e-01 -7.23522067e-01 -8.01314592e-01 6.45649850e-01
-7.18296707e-01 1.46356061e-01 -5.52862823e-01 -3.15654099e-01
-2.85064667e-01 -1.32154000e+00 1.08966100e+00 3.38173538e-01
-3.74894589e-01 -1.32338715e+00 1.17359042e-01 2.81979322e-01
5.14246702e-01 4.66047287e-01 8.93343568e-01 -1.00628662e+00
-7.88004518e-01 -2.43895695e-01 -2.14023188e-01 5.18183373e-02
5.25108874e-01 4.27192122e-01 -6.16825819e-01 -1.60696477e-01
3.40825655e-02 4.84550279e-03 8.29921007e-01 9.43155825e-01
1.21541870e+00 8.25414248e-03 -8.26713741e-01 5.37074447e-01
1.41977227e+00 1.90305561e-01 4.10242736e-01 2.65984684e-01
5.25280416e-01 4.96256590e-01 5.82823813e-01 1.29831135e-01
-1.41024575e-01 1.62292361e-01 5.45214593e-01 -5.91912925e-01
-4.47904229e-01 2.10162625e-01 -1.71200603e-01 3.02351564e-01
-2.14750826e-01 1.17783464e-01 -1.51149368e+00 3.89515072e-01
-1.39283597e+00 -5.88441253e-01 -5.23922443e-01 1.99245894e+00
6.94352925e-01 4.20089066e-01 3.50839585e-01 -1.53322682e-01
4.85859364e-01 -3.11782211e-01 -5.54925740e-01 -1.02855064e-01
4.75651324e-01 3.93156141e-01 7.75165677e-01 5.45543551e-01
-1.29971504e+00 8.12173307e-01 7.22896242e+00 6.63080454e-01
-1.01134861e+00 7.45759010e-02 8.83867383e-01 1.16165258e-01
-2.09457517e-01 -2.02551838e-02 -6.07107818e-01 -7.76706114e-02
6.41290009e-01 -1.38783082e-01 -7.66686425e-02 6.40405059e-01
1.48789182e-01 -2.58007914e-01 -1.10528207e+00 5.41266501e-01
-8.21674317e-02 -1.63358223e+00 -3.31947654e-01 2.86317825e-01
8.42009246e-01 1.68348819e-01 -4.22501849e-04 2.04093471e-01
3.76404375e-01 -1.30756521e+00 1.54268190e-01 3.73186290e-01
8.61999750e-01 -2.49698982e-01 9.45380986e-01 2.65042126e-01
-1.31742406e+00 2.16387227e-01 5.52565195e-02 7.23150909e-01
-2.26009697e-01 4.41930979e-01 -2.04417229e+00 2.64805466e-01
5.78633189e-01 4.50514674e-01 -1.16879320e+00 1.48144281e+00
1.29973948e-01 9.51025963e-01 -4.77940053e-01 -9.86769572e-02
3.85627002e-01 -1.29182458e-01 4.08219159e-01 1.45249856e+00
1.82100520e-01 1.00485235e-01 3.03338140e-01 9.67408419e-01
1.41790718e-01 4.96275216e-01 -3.91511351e-01 3.63475755e-02
1.50218636e-01 1.75485790e+00 -1.59293652e+00 -2.36606002e-01
-4.50657338e-01 4.73322839e-01 -1.43176690e-01 -4.05824669e-02
-9.28064942e-01 1.39982566e-01 -1.83291569e-01 5.92807829e-01
3.29168886e-01 1.19809248e-01 -4.22922134e-01 -4.49282080e-01
-3.52801502e-01 -4.72562999e-01 5.90642333e-01 -3.76747906e-01
-9.81042862e-01 4.63546127e-01 1.09018266e-01 -1.40636373e+00
-3.51052731e-01 -7.91243017e-01 -8.96570086e-01 5.04305243e-01
-1.25817335e+00 -1.06286156e+00 -6.16987050e-01 1.31307647e-01
4.40827638e-01 5.17826602e-02 4.87438351e-01 -2.07306221e-01
-1.60319090e-01 3.06895882e-01 2.99973600e-02 2.38833666e-01
5.85904777e-01 -1.74139774e+00 2.10779607e-02 4.84258652e-01
-5.86376786e-02 4.08670641e-02 3.57519746e-01 -8.42122614e-01
-8.97790253e-01 -1.24792361e+00 2.25743681e-01 -6.76119924e-01
7.35134363e-01 1.68721676e-01 -8.58590305e-01 5.97334623e-01
5.57149164e-02 4.26978707e-01 6.48974955e-01 -4.37437594e-01
3.96887988e-01 2.30946109e-01 -1.35113859e+00 5.13223529e-01
6.03195131e-01 1.87426820e-01 -2.67750293e-01 6.24034941e-01
3.56991380e-01 -5.33266664e-01 -1.16859925e+00 7.59042501e-01
3.09730291e-01 -9.22919452e-01 9.76820946e-01 -2.89156854e-01
1.91930205e-01 -3.23941588e-01 3.49227399e-01 -8.32624376e-01
-6.10242546e-01 -3.45431149e-01 1.86095312e-01 5.49058259e-01
8.29971433e-01 -3.29959780e-01 1.38953698e+00 4.57191080e-01
-3.18375677e-01 -1.20253849e+00 -7.69199610e-01 -5.61054826e-01
3.62824947e-01 1.59242132e-03 -2.47553270e-02 6.08400762e-01
-3.97441208e-01 -9.96241122e-02 7.69945145e-01 -1.18945189e-01
6.01477325e-01 -1.83234941e-02 4.16462719e-01 -1.39345145e+00
-2.95919299e-01 -9.52034473e-01 -3.18696052e-01 -6.12582386e-01
-3.01861256e-01 -1.33083439e+00 2.06644446e-01 -2.00927234e+00
4.56583858e-01 -7.16994762e-01 -4.73376811e-02 4.36825246e-01
-2.41768986e-01 3.20326716e-01 -1.40432984e-01 5.82068920e-01
-3.83982658e-01 -2.89171010e-01 1.84884763e+00 5.57945669e-02
-3.80888820e-01 4.63724375e-01 -5.92431068e-01 1.06581998e+00
9.45194244e-01 -4.04465973e-01 -1.78455204e-01 3.07571590e-01
-4.02909547e-01 -6.32038563e-02 3.68383676e-01 -1.09906709e+00
1.90358654e-01 -2.25932330e-01 5.83539784e-01 -8.56071472e-01
3.75028290e-02 -8.71522605e-01 1.09898089e-03 1.21836531e+00
-1.60693765e-01 -3.06897432e-01 1.07656084e-01 4.02791500e-01
1.79200500e-01 -6.24668837e-01 1.28433549e+00 -5.81297576e-01
-4.54070836e-01 4.19306368e-01 -4.51624811e-01 1.80133328e-01
1.46494293e+00 -6.85991108e-01 7.60797858e-02 -1.91202417e-01
-9.77571428e-01 1.25229701e-01 5.27299583e-01 -2.54480720e-01
3.15756232e-01 -1.21967721e+00 -9.12908375e-01 9.00379047e-02
5.07415161e-02 3.93697560e-01 -3.82336169e-01 1.41148877e+00
-1.12491906e+00 5.51910639e-01 2.15091705e-01 -1.25940812e+00
-1.32831836e+00 9.87621918e-02 1.10478675e+00 -7.30864227e-01
-5.46579123e-01 1.04914641e+00 1.81994721e-01 -5.24898052e-01
2.85832006e-02 -7.03376591e-01 -2.11704344e-01 -1.13818556e-01
-1.56748593e-01 2.88092881e-01 2.79421240e-01 -4.45282400e-01
-3.73770624e-01 4.75063771e-01 -3.22245271e-03 -1.10431902e-01
8.99667084e-01 2.82543391e-01 -1.23018742e-01 2.78051436e-01
9.58045781e-01 -1.26022594e-02 -1.20321274e+00 -1.34335533e-01
5.57592630e-01 -1.79554179e-01 3.07821240e-02 -7.76619375e-01
-1.01429904e+00 4.65837896e-01 9.24959242e-01 2.90419400e-01
8.63610804e-01 6.02073967e-01 2.95435905e-01 1.15547031e-01
-1.75735787e-01 -6.99027956e-01 5.35998583e-01 3.85952890e-01
5.86393535e-01 -1.43838239e+00 2.39827603e-01 -8.09838116e-01
-6.42186046e-01 1.23050904e+00 8.33165169e-01 -1.72332615e-01
8.46652985e-01 5.25045872e-01 1.62044212e-01 -3.92465949e-01
-4.79073197e-01 -5.66287339e-01 7.10300386e-01 3.40836048e-01
7.24612057e-01 5.47182225e-02 -2.87920088e-01 3.37046653e-01
-1.51103437e-01 -3.04365605e-01 4.64694589e-01 1.17105985e+00
-1.00197732e+00 -1.16732097e+00 -6.26507699e-01 1.23243761e+00
-5.15537739e-01 -3.99914309e-02 -5.40609837e-01 1.34396410e+00
1.47303715e-01 4.54975992e-01 1.86125845e-01 2.00076774e-01
2.78416872e-01 3.12211961e-02 5.94574928e-01 -1.07844043e+00
-9.19980645e-01 6.23480976e-01 -2.24149927e-01 -2.79016286e-01
-3.96797776e-01 -8.36642623e-01 -1.82607114e+00 3.36637646e-01
-6.19392633e-01 9.07849744e-02 4.47995663e-01 7.30332673e-01
-2.41847783e-01 8.78765523e-01 5.85578203e-01 -7.76562929e-01
-3.83598208e-01 -7.40695298e-01 -3.93180102e-01 4.07819897e-01
1.49952203e-01 -5.16156733e-01 -5.15900135e-01 3.28996658e-01] | [15.362509727478027, -2.1535327434539795] |
378906b6-1947-4f29-9ddf-ccbc6de8f047 | oriented-r-cnn-for-object-detection | 2108.05699 | null | https://arxiv.org/abs/2108.05699v1 | https://arxiv.org/pdf/2108.05699v1.pdf | Oriented R-CNN for Object Detection | Current state-of-the-art two-stage detectors generate oriented proposals through time-consuming schemes. This diminishes the detectors' speed, thereby becoming the computational bottleneck in advanced oriented object detection systems. This work proposes an effective and simple oriented object detection framework, termed Oriented R-CNN, which is a general two-stage oriented detector with promising accuracy and efficiency. To be specific, in the first stage, we propose an oriented Region Proposal Network (oriented RPN) that directly generates high-quality oriented proposals in a nearly cost-free manner. The second stage is oriented R-CNN head for refining oriented Regions of Interest (oriented RoIs) and recognizing them. Without tricks, oriented R-CNN with ResNet50 achieves state-of-the-art detection accuracy on two commonly-used datasets for oriented object detection including DOTA (75.87% mAP) and HRSC2016 (96.50% mAP), while having a speed of 15.1 FPS with the image size of 1024$\times$1024 on a single RTX 2080Ti. We hope our work could inspire rethinking the design of oriented detectors and serve as a baseline for oriented object detection. Code is available at https://github.com/jbwang1997/OBBDetection. | ['Junwei Han', 'Xiwen Yao', 'Jiabao Wang', 'Gong Cheng', 'Xingxing Xie'] | 2021-08-12 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Xie_Oriented_R-CNN_for_Object_Detection_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Xie_Oriented_R-CNN_for_Object_Detection_ICCV_2021_paper.pdf | iccv-2021-1 | ['object-detection-in-aerial-images'] | ['computer-vision'] | [-2.65039295e-01 1.86767772e-01 -2.49485269e-01 -4.83679287e-02
-7.68734932e-01 -3.14343154e-01 2.02873364e-01 -2.55461007e-01
-6.08567119e-01 1.59860831e-02 -3.26244831e-01 -3.69064420e-01
5.58210075e-01 -9.60798919e-01 -6.01688921e-01 -4.18332487e-01
-1.47497490e-01 1.43962681e-01 1.04535246e+00 -3.43850136e-01
3.44832420e-01 7.69767225e-01 -1.49214208e+00 4.55352336e-01
4.24309105e-01 1.03064835e+00 3.38679790e-01 9.74240422e-01
1.61275640e-01 7.98077703e-01 -6.35746121e-01 -4.23625678e-01
4.78686303e-01 1.00930661e-01 -5.41605175e-01 -3.13747860e-02
4.90236282e-01 -7.09850848e-01 -7.27834761e-01 9.75328207e-01
7.83286989e-01 -2.56591558e-01 3.40317726e-01 -1.08776164e+00
-8.26419950e-01 6.54575288e-01 -8.49945605e-01 7.23216176e-01
-2.10675865e-01 5.46460390e-01 8.77233207e-01 -1.47325957e+00
6.09095812e-01 1.30122411e+00 4.95496958e-01 7.51282334e-01
-6.86954856e-01 -7.52247930e-01 3.31548840e-01 2.29347467e-01
-1.80727053e+00 -2.58087277e-01 3.83097053e-01 -2.34176889e-01
1.08270872e+00 9.23306271e-02 7.86214411e-01 7.31185913e-01
1.71212077e-01 1.13977408e+00 7.07509160e-01 -2.73274004e-01
2.52765808e-02 -7.62378797e-02 4.63543296e-01 9.68849480e-01
7.01992869e-01 3.87700230e-01 -4.92390662e-01 3.12804699e-01
8.78706932e-01 -8.63001719e-02 9.47533846e-02 -2.69881468e-02
-1.23442745e+00 6.53276205e-01 1.17158794e+00 1.39677942e-01
-2.67098159e-01 2.83055216e-01 3.86169940e-01 -2.31299222e-01
5.75505197e-01 1.72608420e-01 -1.25456691e-01 4.25766945e-01
-7.00517297e-01 3.76415342e-01 1.78235650e-01 1.28333759e+00
5.43140054e-01 2.87947625e-01 -4.33563858e-01 6.23754203e-01
5.21700501e-01 8.06827605e-01 2.41418511e-01 -6.29298091e-01
4.66537148e-01 9.30794239e-01 5.41274175e-02 -8.09999585e-01
-6.29361629e-01 -7.10509598e-01 -6.82979822e-01 3.09013784e-01
3.44387054e-01 4.11277711e-02 -1.19257295e+00 1.01209998e+00
4.18261260e-01 -3.17146063e-01 -1.83451056e-01 1.34900784e+00
1.37995636e+00 6.71515465e-01 1.17392726e-01 4.95906562e-01
1.83230746e+00 -1.10659170e+00 -7.56256133e-02 -3.35410625e-01
5.57873070e-01 -7.07632244e-01 8.46560955e-01 3.65739107e-01
-1.27814960e+00 -8.13061416e-01 -1.13805938e+00 -2.75211126e-01
-4.24801201e-01 4.85532343e-01 4.97077197e-01 5.87529778e-01
-1.12385261e+00 1.04434766e-01 -8.74047935e-01 -4.32378232e-01
9.30690944e-01 4.54556286e-01 9.36947390e-03 -4.42269519e-02
-6.59261107e-01 6.51768982e-01 5.68804026e-01 2.00480148e-01
-9.60063577e-01 -6.89104915e-01 -4.72897202e-01 -4.72845859e-05
4.43463594e-01 -5.59132099e-01 1.29625511e+00 -3.75299662e-01
-1.00976396e+00 9.61135507e-01 -2.50173453e-02 -6.06085658e-01
6.20091259e-01 -4.29289192e-01 -4.35218424e-01 2.95619100e-01
2.58605003e-01 1.32120585e+00 9.75221515e-01 -9.12578762e-01
-1.10674775e+00 -3.58072847e-01 -3.15904841e-02 3.04896291e-02
-3.14169824e-01 4.49584067e-01 -9.59953010e-01 -5.79344749e-01
2.31186330e-01 -7.74002492e-01 -3.72784466e-01 4.65336978e-01
-7.42718995e-01 -6.48427963e-01 9.01390195e-01 -2.78530896e-01
1.05167985e+00 -2.02703547e+00 -4.38993752e-01 -1.89970911e-01
6.04703605e-01 8.51126790e-01 -3.29703689e-01 -1.84859946e-01
4.18206602e-02 2.72075236e-01 1.13654368e-01 -1.32847711e-01
-3.51431847e-01 -5.40646672e-01 -1.99186131e-01 5.97277582e-01
5.17407358e-01 1.31850064e+00 -7.75380433e-01 -6.40152216e-01
3.45259815e-01 4.82389718e-01 -5.70755124e-01 1.01595998e-01
2.48408169e-01 -1.50766477e-01 -4.61594373e-01 1.20415044e+00
8.29091966e-01 -4.13365185e-01 -3.32881153e-01 -4.01792407e-01
-5.04449487e-01 2.27635354e-01 -1.07814574e+00 1.18627679e+00
8.85776356e-02 1.01842773e+00 -2.10161611e-01 -7.49822557e-01
1.20539367e+00 -2.84647271e-02 1.84954613e-01 -9.95146573e-01
2.61652797e-01 2.68495739e-01 2.48219609e-01 -1.83449209e-01
9.43778574e-01 2.85354048e-01 9.26530659e-02 2.94188082e-01
-8.07712749e-02 -5.47202155e-02 3.10861558e-01 3.00889492e-01
1.04522002e+00 -9.50057954e-02 4.38170403e-01 -2.69333273e-01
3.17189753e-01 1.92815915e-01 4.91751552e-01 1.16015947e+00
-4.98383403e-01 1.04467416e+00 7.61949196e-02 -8.43741417e-01
-1.24168682e+00 -1.04503655e+00 -2.21919984e-01 1.01619852e+00
4.27529752e-01 -3.31528038e-01 -5.97571015e-01 -9.30814326e-01
-1.09484792e-01 2.28365049e-01 -3.57393116e-01 -1.22989621e-02
-1.02900159e+00 -9.08248365e-01 7.43750453e-01 9.43593860e-01
8.86896491e-01 -1.36169350e+00 -9.38650489e-01 1.68476820e-01
4.95648421e-02 -1.02379501e+00 -4.90320176e-01 9.27471071e-02
-1.00665963e+00 -1.09156108e+00 -9.95505095e-01 -8.61927986e-01
9.72831905e-01 1.01084638e+00 1.09905744e+00 1.97361723e-01
-6.68328166e-01 -5.87578826e-02 -4.67849314e-01 -7.01003075e-01
-2.20133569e-02 3.35475057e-01 -5.24178799e-03 -4.09829766e-01
5.37731826e-01 2.39399776e-01 -1.11739182e+00 7.20797777e-01
-7.70602584e-01 4.83360216e-02 8.06916893e-01 5.69817901e-01
5.17056227e-01 -3.77909958e-01 4.95051801e-01 -5.77509224e-01
1.19234808e-01 -7.80027583e-02 -1.07077587e+00 -6.90265894e-02
-4.34497416e-01 -2.74019867e-01 4.45116967e-01 -4.69681740e-01
-8.09123039e-01 1.09377988e-01 -2.72849351e-01 -3.06299210e-01
-8.90086144e-02 -3.13466758e-01 3.12611789e-01 -1.47471145e-01
1.03359878e+00 2.62938738e-01 -4.21480238e-01 -3.62344235e-01
4.44852144e-01 9.17566955e-01 4.25339460e-01 -9.54420194e-02
1.00981116e+00 7.72369623e-01 -3.70768070e-01 -9.57427323e-01
-6.59802139e-01 -8.83315921e-01 -4.84839082e-01 -3.48905772e-01
6.39362037e-01 -1.17262352e+00 -8.47495079e-01 7.10342944e-01
-1.33561623e+00 -3.33287418e-01 -3.69826257e-01 1.45698667e-01
-1.21476576e-01 1.89008847e-01 -7.51979113e-01 -7.61247993e-01
-8.45112801e-01 -1.09154391e+00 1.34296966e+00 5.07882118e-01
1.12572141e-01 -2.77900577e-01 -4.59015071e-01 2.86504835e-01
5.64614475e-01 -4.20988083e-01 1.60371959e-01 -3.30684185e-01
-1.13002026e+00 -5.08308053e-01 -1.05494881e+00 1.70510799e-01
-2.57263452e-01 7.84712508e-02 -9.48150635e-01 -3.39253902e-01
-6.29518211e-01 -1.02542937e-01 1.06768346e+00 3.60104859e-01
1.39859927e+00 -7.38415271e-02 -3.83339673e-01 5.81298232e-01
1.16588902e+00 2.90395558e-01 8.91076982e-01 5.51486135e-01
8.87842357e-01 2.54977643e-01 9.18659151e-01 3.18148345e-01
3.38631272e-01 8.18580091e-01 4.48431820e-01 -4.42768902e-01
-6.08475149e-01 -1.99159130e-01 2.84918427e-01 3.77119660e-01
-3.56848866e-01 -3.52875501e-01 -8.80532146e-01 5.95362306e-01
-1.50063539e+00 -9.29182827e-01 -3.40300620e-01 1.98191094e+00
3.53504479e-01 4.20471579e-01 5.06053686e-01 3.88482492e-03
9.06383693e-01 2.13889599e-01 -6.13914549e-01 -1.15939409e-01
-1.24401515e-04 1.20516032e-01 9.04914975e-01 2.47238334e-02
-1.39460468e+00 1.23317802e+00 6.02522993e+00 7.38516450e-01
-1.34803939e+00 4.38249968e-02 8.26538980e-01 -1.16975764e-02
5.17270148e-01 -2.96708494e-01 -1.58892345e+00 2.46944144e-01
5.70468724e-01 -2.28417013e-03 -3.41498330e-02 1.58772695e+00
2.87441999e-01 8.38770047e-02 -6.67858481e-01 8.24297786e-01
-2.63626524e-03 -1.40475142e+00 -4.98971641e-02 2.13653535e-01
4.39508915e-01 5.01446784e-01 1.44481689e-01 4.75202888e-01
1.17655449e-01 -7.12386906e-01 1.12750113e+00 -1.62184294e-02
8.45901370e-01 -3.93801481e-01 6.18043602e-01 2.43453100e-01
-1.65109992e+00 -2.83149719e-01 -8.15019190e-01 2.61768639e-01
2.10823685e-01 5.91130793e-01 -8.98206353e-01 1.29032746e-01
1.19496059e+00 7.79680967e-01 -8.70127499e-01 1.49092031e+00
-1.70940310e-01 5.68374157e-01 -3.24360937e-01 -2.76435673e-01
1.70927346e-01 4.23577756e-01 7.34754503e-01 1.37295067e+00
1.70743480e-01 2.68895984e-01 -1.00305289e-01 1.07793546e+00
-2.89783895e-01 -4.45409566e-02 -4.57502276e-01 2.50321776e-01
4.21471477e-01 1.65889955e+00 -1.29047394e+00 -5.83311856e-01
-3.25798839e-01 6.56700850e-01 4.20277536e-01 5.38017265e-02
-1.07784581e+00 -4.94655311e-01 2.87481755e-01 2.76702493e-01
6.42517328e-01 -2.53393710e-01 -2.45826945e-01 -1.15231431e+00
-8.61975178e-03 -7.59859681e-01 2.54489809e-01 -6.49923623e-01
-1.14641452e+00 6.75071001e-01 -1.09919399e-01 -1.62552178e+00
4.13059115e-01 -1.05330980e+00 -5.85591197e-01 4.51250553e-01
-1.67160368e+00 -1.11247134e+00 -6.85183406e-01 1.22037023e-01
9.54996884e-01 -1.22813158e-01 6.45133406e-02 4.07601029e-01
-9.94499922e-01 6.58055067e-01 -4.86869067e-01 6.15470469e-01
4.34972197e-01 -9.41073537e-01 1.24576867e+00 1.11093271e+00
3.54322866e-02 5.74061275e-01 1.49682730e-01 -5.61099827e-01
-1.46647954e+00 -1.68733859e+00 6.66458368e-01 -4.29394543e-01
3.34437490e-01 -4.64862376e-01 -7.80033350e-01 4.90889996e-01
-2.53840238e-01 7.65062094e-01 -1.62180349e-01 -3.34511369e-01
-4.03541505e-01 -2.06077665e-01 -9.98216510e-01 8.72366130e-01
1.26407349e+00 -1.14736706e-01 -1.88232005e-01 3.69686037e-01
6.91767156e-01 -5.09286165e-01 -4.23788577e-01 5.27177453e-01
6.49446249e-01 -1.14038312e+00 1.38397944e+00 -2.03325346e-01
3.82583141e-02 -5.53756595e-01 6.54747412e-02 -4.91457403e-01
-4.75043058e-01 -2.63410896e-01 -2.03865215e-01 7.92929292e-01
4.34801072e-01 -7.48197854e-01 1.05890965e+00 1.30499333e-01
-3.10351521e-01 -8.39365184e-01 -6.66684508e-01 -9.65488911e-01
-2.77831078e-01 -6.32812738e-01 4.13078457e-01 1.72834828e-01
-3.69899720e-01 1.06820300e-01 1.25471368e-01 4.07498896e-01
7.22220182e-01 -1.58379331e-01 8.68866444e-01 -8.42787743e-01
-3.35645899e-02 -6.41190529e-01 -6.76320851e-01 -1.47452867e+00
-5.50768733e-01 -5.00151396e-01 2.28540331e-01 -1.32833874e+00
2.36971080e-01 -7.66397536e-01 -1.07645072e-01 4.64617819e-01
-3.25469911e-01 1.06143034e+00 4.82911468e-01 3.90285850e-01
-1.05129254e+00 3.30311716e-01 1.14646566e+00 -2.73867518e-01
-4.17379826e-01 4.97946143e-02 -6.38769746e-01 8.20124269e-01
9.10648346e-01 -5.42990685e-01 8.85266066e-02 -3.16022038e-01
-2.59884357e-01 -4.15281922e-01 5.51659226e-01 -1.28654587e+00
1.78622633e-01 1.00010172e-01 4.08337981e-01 -1.21399653e+00
2.15045869e-01 -2.85269886e-01 -5.48100770e-01 8.66340816e-01
-1.57599986e-01 9.25294384e-02 3.23694408e-01 2.06335425e-01
1.11576036e-01 -7.23522576e-03 1.06353939e+00 -1.35310680e-01
-1.16552818e+00 4.70701337e-01 -3.18262964e-01 3.52920815e-02
1.15551531e+00 -3.53837788e-01 -8.62677217e-01 1.33358508e-01
-2.49407426e-01 9.52544212e-02 3.21697444e-01 6.91987097e-01
9.72423196e-01 -9.98570442e-01 -7.89300561e-01 3.47105786e-02
4.10149425e-01 3.72838676e-01 2.06042364e-01 6.42708242e-01
-1.01458061e+00 8.08815956e-01 -5.72834015e-02 -9.05581832e-01
-1.28737640e+00 4.26683307e-01 4.63702738e-01 -5.69819137e-02
-1.04819942e+00 1.00918376e+00 4.40582603e-01 -1.36797488e-01
2.52930254e-01 -3.19820195e-01 -1.86507970e-01 -1.91128388e-01
8.44779551e-01 6.88627481e-01 1.95985526e-01 -6.89502418e-01
-6.24221683e-01 5.70078075e-01 -5.10042131e-01 2.43058160e-01
9.69984293e-01 2.14018047e-01 3.77407581e-01 -1.90346718e-01
8.75858009e-01 -4.64850724e-01 -1.21529174e+00 -4.79744114e-02
-1.73469484e-01 -3.76596153e-01 1.61046579e-01 -3.95415366e-01
-1.26596344e+00 8.06382179e-01 8.65578890e-01 1.46633312e-01
8.42036366e-01 3.11085492e-01 6.65818870e-01 4.54998732e-01
4.10055071e-01 -7.54004598e-01 4.55809861e-01 3.10685575e-01
8.28126907e-01 -1.48831630e+00 2.75167167e-01 -5.39960504e-01
-3.56831044e-01 1.02564311e+00 1.10942996e+00 -4.47664171e-01
3.18886608e-01 2.19541818e-01 1.86827421e-01 -3.80756408e-01
-3.56648445e-01 -5.06707847e-01 5.08940518e-01 6.89708710e-01
4.79611158e-01 1.23426817e-01 -2.29300037e-01 8.56524855e-02
4.54326980e-02 -1.95075810e-01 4.78476077e-01 9.36380148e-01
-7.13616431e-01 -6.72735035e-01 -5.64707041e-01 4.51753557e-01
-3.25190693e-01 -1.00951895e-01 -1.48606762e-01 1.11580896e+00
2.87983213e-02 9.19853628e-01 3.15254480e-01 -2.01888710e-01
4.66239244e-01 -5.19081295e-01 -1.28802867e-03 -8.91286135e-01
-5.33962786e-01 -1.06802851e-01 -1.64660811e-01 -8.35661888e-01
-1.42285123e-01 -3.64997864e-01 -1.36182678e+00 -2.98341870e-01
-6.98518455e-01 -4.36613262e-01 7.01049268e-01 5.02481937e-01
6.56744361e-01 6.20990455e-01 5.23805737e-01 -1.30399370e+00
-3.46098006e-01 -9.45791721e-01 -3.26029956e-01 -9.14101377e-02
1.74402103e-01 -4.79691118e-01 -4.22179699e-02 -2.58354247e-01] | [8.742850303649902, -0.24156330525875092] |
666a2199-e816-4496-bc64-1654b7a95565 | automatic-features-for-essay-scoring-a-an | null | null | https://aclanthology.org/D16-1115 | https://aclanthology.org/D16-1115.pdf | Automatic Features for Essay Scoring -- An Empirical Study | null | ['Yue Zhang', 'Fei Dong'] | 2016-11-01 | null | null | null | emnlp-2016-11 | ['automated-essay-scoring'] | ['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] | [-7.477737903594971, 3.8463637828826904] |
ebb021fa-0750-43eb-8ab6-f1ff1ba33e0a | multilingual-natural-language-processing | null | null | https://aclanthology.org/W12-0506 | https://aclanthology.org/W12-0506.pdf | Multilingual Natural Language Processing | null | ['Rada Mihalcea'] | 2012-04-01 | null | null | null | ws-2012-4 | ['subjectivity-analysis'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.368608474731445, 3.664233922958374] |
e87f35a9-1c7d-476c-b29e-f5f8b8862418 | deep-autoencoder-like-nonnegative-matrix | null | null | https://dl.acm.org/citation.cfm?id=3271697 | https://smartyfh.com/Documents/18DANMF.pdf | Deep Autoencoder-like Nonnegative Matrix Factorization for Community Detection | Community structure is ubiquitous in real-world complex networks. The task of community detection over these networks is of paramount importance in a variety of applications. Recently, nonnegative matrix factorization (NMF) has been widely adopted for community detection due to its great interpretability and its natural fitness for capturing the community membership of nodes. However, the existing NMF-based community detection approaches are shallow methods. They learn the community assignment by mapping the original network to the community membership space directly. Considering the complicated and diversified topology structures of real-world networks, it is highly possible that the mapping between the original network and the community membership space contains rather complex hierarchical information, which cannot be interpreted by classic shallow NMF-based approaches. Inspired by the unique feature representation learning capability of deep autoencoder, we propose a novel model, named Deep Autoencoder-like NMF (DANMF), for community detection. Similar to deep autoencoder, DANMF consists of an encoder component and a decoder component. This architecture empowers DANMF to learn the hierarchical mappings between the original network and the final community assignment with implicit low-to-high level hidden attributes of the original network learnt in the intermediate layers. Thus, DANMF should be better suited to the community detection task. Extensive experiments on benchmark datasets demonstrate that DANMF can achieve better performance than the state-of-the-art NMF-based community detection approaches. | ['Zibin Zheng', 'Fanghua Ye', 'Chuan Chen'] | 2018-10-22 | null | null | null | cikm-2018-10 | ['network-community-partition', 'local-community-detection'] | ['graphs', 'graphs'] | [-2.15764150e-01 -1.82220951e-01 9.99342576e-02 -2.81183925e-02
2.66536474e-01 -3.89186174e-01 6.07341766e-01 2.66501755e-01
-1.91545933e-01 4.76488113e-01 2.13383570e-01 -1.45064667e-01
-1.79356873e-01 -1.12829185e+00 -4.41702515e-01 -7.63909042e-01
-5.67454457e-01 7.26391792e-01 -3.40367062e-03 -1.77699625e-01
-9.58456993e-02 8.71034116e-02 -1.31507528e+00 6.04580343e-01
9.76788223e-01 5.51718533e-01 3.69046837e-01 5.24159551e-01
-2.77649403e-01 8.68326902e-01 -5.20400345e-01 -3.95363599e-01
-1.44596636e-01 -2.51514316e-01 -8.14671814e-01 3.33546370e-01
-8.29300210e-02 -1.50432408e-01 -6.60827935e-01 1.18924069e+00
5.59904315e-02 -1.47767782e-01 7.26554334e-01 -1.39015794e+00
-9.69717264e-01 9.54269171e-01 -7.17585862e-01 1.91408798e-01
3.04911017e-01 -2.38832951e-01 1.31746411e+00 -1.02962244e+00
6.80019915e-01 1.36876631e+00 8.09606910e-01 3.91400933e-01
-1.48526430e+00 -7.10678816e-01 4.47989881e-01 3.27596843e-01
-1.67583990e+00 1.41260833e-01 7.39171982e-01 -1.01912880e+00
7.88915634e-01 -1.88251898e-01 1.25243235e+00 1.20345020e+00
-3.34641784e-02 8.83791924e-01 7.07359433e-01 -1.78407609e-01
1.98430747e-01 -7.39641413e-02 -2.35556886e-02 7.06202090e-01
4.13824826e-01 -2.28652745e-01 -1.65916204e-01 -3.55790406e-01
1.05644369e+00 4.83299524e-01 -3.57630759e-01 -6.22585058e-01
-1.43961334e+00 1.25033438e+00 1.05538416e+00 5.89559138e-01
-5.09417355e-01 7.02739656e-02 4.03839290e-01 2.67285377e-01
2.07370073e-01 1.76998913e-01 1.34731218e-01 4.54278052e-01
-7.77953863e-01 -8.70500654e-02 8.01396489e-01 7.31746674e-01
6.87999845e-01 -2.26113424e-02 1.89835742e-01 5.74824154e-01
6.18157029e-01 -1.29924975e-02 4.10932869e-01 -6.38082922e-01
2.60875642e-01 1.18029892e+00 -3.65417957e-01 -1.38081598e+00
-3.10197741e-01 -8.20144653e-01 -1.74433994e+00 -1.42490119e-01
4.28813249e-02 -3.34586762e-02 -4.01396811e-01 1.63796043e+00
3.13273013e-01 4.49071258e-01 1.75025627e-01 9.58486140e-01
9.38836813e-01 5.78846812e-01 -2.33809158e-01 1.54625941e-02
1.38203835e+00 -8.31775248e-01 -5.71749449e-01 -6.45231456e-02
2.41920844e-01 -2.81000406e-01 5.69002271e-01 5.96225709e-02
-3.43099892e-01 -3.59897763e-01 -1.16570616e+00 2.90591866e-01
-2.36753359e-01 1.08533859e-01 1.13135850e+00 2.27166951e-01
-1.10534835e+00 5.11817217e-01 -8.19131434e-01 -4.64774489e-01
4.66281682e-01 4.71398771e-01 -7.58026659e-01 -1.84597671e-01
-1.35655999e+00 2.20244005e-01 7.06928909e-01 5.41447401e-01
-1.05528569e+00 -1.79308176e-01 -7.88728476e-01 3.89084458e-01
3.53728622e-01 -9.81355429e-01 5.67364275e-01 -1.06349707e+00
-8.48418474e-01 4.86387849e-01 1.25649720e-01 -3.89102280e-01
2.27015287e-01 3.73462498e-01 -4.92819428e-01 2.60241508e-01
2.25170106e-01 6.25415564e-01 1.03464890e+00 -1.21312082e+00
-4.64454502e-01 -1.71514690e-01 2.15866148e-01 9.16853845e-02
-9.52186823e-01 -1.10699564e-01 -1.74090996e-01 -6.28846824e-01
3.47636580e-01 -7.49233603e-01 -4.02678907e-01 -2.91298460e-02
-5.48526645e-01 -3.67178887e-01 8.71725321e-01 -3.52974772e-01
1.45093453e+00 -1.91784537e+00 4.20765072e-01 3.78144592e-01
1.07102334e+00 1.58021465e-01 -3.68051410e-01 7.24863291e-01
-2.88473397e-01 2.71814108e-01 -2.65236974e-01 -2.96758175e-01
-1.49249047e-01 2.97525525e-01 3.24440747e-02 4.35299844e-01
4.72436309e-01 7.62360215e-01 -1.14315355e+00 -4.35665965e-01
-7.72022754e-02 8.59437227e-01 -7.20481813e-01 1.45037577e-01
1.96728125e-01 1.64619341e-01 -2.64238000e-01 4.69745427e-01
4.88581806e-01 -8.62760305e-01 7.56508827e-01 -5.99145815e-02
1.50027275e-01 -1.53253376e-01 -1.47064912e+00 1.32964730e+00
7.29324296e-02 8.30151260e-01 4.60369706e-01 -1.15845978e+00
8.37308645e-01 5.79513013e-01 4.53563690e-01 3.52160782e-02
-2.12553982e-02 8.25597122e-02 5.86767256e-01 -2.49413550e-01
2.75285929e-01 -1.10334158e-01 2.32000664e-01 6.73655689e-01
2.89124221e-01 6.32884681e-01 5.07673979e-01 6.92992449e-01
1.10260916e+00 -7.12732434e-01 2.70199507e-01 -4.64107186e-01
8.29650164e-01 -3.69063973e-01 8.01514328e-01 4.88230586e-01
-2.32522011e-01 4.48998779e-01 5.74172258e-01 -5.99765599e-01
-9.33739126e-01 -1.03761816e+00 6.30591065e-02 9.03380811e-01
-9.81181040e-02 -7.41374135e-01 -7.36097336e-01 -5.77065766e-01
2.15346709e-01 -5.27324438e-01 -9.63284194e-01 -1.83513239e-01
-4.50275958e-01 -8.85198176e-01 4.16240752e-01 6.48230731e-01
3.57979923e-01 -9.73514855e-01 2.57631093e-01 3.86504918e-01
-6.21188641e-01 -9.95723426e-01 -3.63767684e-01 1.29138947e-01
-1.02533627e+00 -1.31567872e+00 -4.41993088e-01 -1.12714279e+00
9.20419693e-01 5.60017824e-01 1.15261221e+00 6.56536102e-01
-7.03945011e-02 3.17462045e-03 -4.00708258e-01 1.10232562e-01
-5.33423722e-01 2.45300591e-01 3.18442464e-01 4.98980850e-01
4.54735011e-01 -9.04148519e-01 -6.03323936e-01 7.36104622e-02
-9.13061440e-01 3.81009839e-02 6.78873003e-01 1.17980433e+00
2.19033271e-01 5.12619019e-01 6.51921391e-01 -6.90306187e-01
9.00889516e-01 -6.68146551e-01 -2.02782735e-01 8.87690261e-02
-3.54089409e-01 -2.67403517e-02 7.94845998e-01 -5.28104663e-01
-5.40646791e-01 -6.88221604e-02 1.91460475e-01 -5.66274881e-01
8.34993050e-02 1.09727585e+00 -2.47078449e-01 1.46301270e-01
4.97593760e-01 2.52335668e-01 1.49948016e-01 -4.93500084e-01
7.54706264e-02 7.39171267e-01 4.87534940e-01 -5.09324789e-01
8.82817745e-01 6.46835029e-01 -2.71589875e-01 -7.74839342e-01
-5.16065836e-01 -8.13678801e-01 -1.11366820e+00 -3.35732013e-01
7.81753182e-01 -1.36649477e+00 -8.10028136e-01 2.24286288e-01
-1.25739419e+00 1.18731976e-01 5.86430021e-02 3.64829898e-01
-9.39169079e-02 6.86868191e-01 -9.83461142e-01 -6.36731863e-01
-3.87387007e-01 -9.72223163e-01 7.62192130e-01 2.79705152e-02
-9.44350436e-02 -1.30170119e+00 9.37808007e-02 2.64798015e-01
1.21198535e-01 2.53229082e-01 1.08736670e+00 -5.12516320e-01
-5.95989227e-01 -2.39534721e-01 -3.78968179e-01 2.45431349e-01
1.48128271e-01 2.62486607e-01 -7.46367991e-01 -7.88542032e-01
-3.46442074e-01 1.45707661e-02 1.12595725e+00 1.80523634e-01
7.04683304e-01 -2.78635532e-01 -4.01992410e-01 3.85689586e-01
1.36411417e+00 -2.93735236e-01 4.72345561e-01 3.15373063e-01
1.06293356e+00 5.29000580e-01 -3.83987948e-02 4.91242081e-01
5.33876657e-01 2.18547896e-01 7.45378017e-01 -1.31122828e-01
1.66974634e-01 -2.44736433e-01 4.04502869e-01 1.37816560e+00
-2.55671859e-01 -1.03654630e-01 -1.05116296e+00 5.77512622e-01
-2.22048235e+00 -1.32414055e+00 -4.02381659e-01 1.76730824e+00
7.34748483e-01 1.22280940e-01 3.40816528e-01 4.06055719e-01
1.23258495e+00 4.04249877e-02 -2.68040836e-01 -8.82778987e-02
-3.91956210e-01 -4.51856613e-01 -4.58910048e-01 2.67864972e-01
-1.40031385e+00 5.60863256e-01 5.81183434e+00 5.63274145e-01
-8.95992696e-01 5.80557995e-02 1.82872981e-01 4.29777414e-01
-2.45192215e-01 3.67240049e-02 -3.49177480e-01 3.53529125e-01
6.38693810e-01 -9.80546102e-02 5.06493509e-01 7.94106603e-01
-1.86485611e-02 5.16339719e-01 -1.14125276e+00 7.70216405e-01
-1.49905741e-01 -1.47193193e+00 2.99991488e-01 4.29339439e-01
7.40079105e-01 7.57784694e-02 -1.84894219e-01 3.99665982e-01
4.53919828e-01 -1.10870218e+00 4.61486489e-01 3.81305754e-01
5.62620521e-01 -8.26759219e-01 9.67269242e-01 6.27293110e-01
-1.70207381e+00 -4.99386370e-01 -6.66256189e-01 -3.96817476e-01
-1.28887594e-01 9.39761281e-01 -1.01224530e+00 4.49054509e-01
7.25002170e-01 1.31431890e+00 -7.16540158e-01 1.00400293e+00
-1.36444256e-01 5.53162038e-01 -2.37350598e-01 1.09192990e-02
2.97141165e-01 -3.40658873e-01 5.84629059e-01 1.28052199e+00
1.37539208e-02 -3.14252496e-01 5.40983975e-01 1.10198486e+00
-2.83648074e-01 -7.89131746e-02 -5.95338583e-01 -5.63651621e-01
5.57390392e-01 1.55548918e+00 -1.00357950e+00 -2.48650923e-01
-3.56737077e-01 8.69051397e-01 6.36479080e-01 4.04567093e-01
-4.50654030e-01 -2.33418360e-01 8.48510206e-01 2.30017424e-01
4.96591985e-01 -1.64238289e-01 1.95864126e-01 -1.56848466e+00
-7.92138577e-02 -9.33784187e-01 6.02178037e-01 -3.57053787e-01
-1.73583519e+00 8.52780998e-01 -5.04452169e-01 -1.34521294e+00
-1.58784881e-01 -4.66885358e-01 -7.96088398e-01 6.62651777e-01
-1.10448754e+00 -1.35689688e+00 -4.09059435e-01 7.02999353e-01
1.01405315e-01 -4.71560150e-01 1.12770951e+00 4.74635541e-01
-9.50981379e-01 3.95718992e-01 4.36286390e-01 9.79867101e-01
2.05690697e-01 -1.47925723e+00 4.46634173e-01 9.65534508e-01
2.67799377e-01 1.22974241e+00 2.31899843e-01 -7.14925230e-01
-1.20051312e+00 -1.31197631e+00 7.64476776e-01 -1.82623759e-01
1.01518786e+00 -8.28925669e-01 -1.06407118e+00 4.62954104e-01
1.92944601e-01 1.28070697e-01 1.01589644e+00 4.20045316e-01
-6.83687925e-01 7.08359033e-02 -6.42615974e-01 4.64169174e-01
9.98247802e-01 -9.92321432e-01 -4.14692700e-01 1.28450200e-01
5.78417540e-01 4.00820076e-01 -1.24124551e+00 3.07122439e-01
5.93151450e-01 -9.67522919e-01 1.00146544e+00 -7.42793918e-01
6.58986092e-01 -7.26612747e-01 -7.21357614e-02 -1.34392822e+00
-1.06677723e+00 -2.07735911e-01 -6.93996787e-01 1.12475443e+00
6.68945014e-02 -4.45003778e-01 9.83916998e-01 -2.70367265e-01
1.97285593e-01 -7.34097898e-01 -8.13432455e-01 -5.23310900e-01
-4.45529073e-02 1.31747633e-01 6.37255549e-01 1.41442454e+00
6.26487136e-02 8.27958941e-01 -1.54693812e-01 1.47608578e-01
7.75184751e-01 5.14309525e-01 4.92592663e-01 -2.08567691e+00
-3.96085948e-01 -5.99054992e-01 -7.09013820e-01 -6.93487287e-01
2.12894693e-01 -1.19613206e+00 -2.46765435e-01 -1.68285811e+00
8.14259589e-01 -1.96760833e-01 -4.32284892e-01 3.76467496e-01
-3.03038239e-01 3.10423702e-01 2.52471805e-01 5.05982459e-01
-7.87048340e-01 6.57969773e-01 1.25386834e+00 -5.46467900e-01
-2.84433626e-02 -1.30679056e-01 -8.32518697e-01 6.56298399e-01
4.18759942e-01 -4.12387162e-01 -3.16862285e-01 -3.79493028e-01
5.80533028e-01 -1.53533652e-01 3.06392968e-01 -9.43329275e-01
5.36431551e-01 1.42976254e-01 5.56735575e-01 -5.87948680e-01
9.48781371e-02 -8.36307943e-01 3.66705686e-01 6.57047808e-01
5.06993458e-02 1.49372205e-01 -3.55829448e-01 1.06817722e+00
-6.27728462e-01 -3.68716568e-02 3.79546016e-01 -3.24011475e-01
-6.40500128e-01 5.27021170e-01 -8.15485954e-01 -2.39365816e-01
5.85438192e-01 -4.71101344e-01 -2.19379574e-01 -4.27280962e-01
-9.47901428e-01 5.01483083e-01 5.17158210e-01 4.45619166e-01
8.49276602e-01 -1.58815849e+00 -9.93186295e-01 3.25734466e-01
2.25980446e-01 2.10195601e-01 1.95164755e-01 9.03316081e-01
-4.32416975e-01 1.50556907e-01 -2.45530263e-01 -8.33526373e-01
-1.15391886e+00 7.10610926e-01 3.41598749e-01 -4.76815104e-01
-5.94618142e-01 7.58980930e-01 4.60742742e-01 -6.52643502e-01
2.06477731e-01 1.13553680e-01 -6.03200078e-01 4.18021262e-01
7.10174024e-01 9.88366082e-02 -1.05202898e-01 -7.54449129e-01
-4.09704208e-01 1.81800425e-01 -1.95188344e-01 4.01308507e-01
1.55133760e+00 -5.93575910e-02 -7.23997653e-01 2.69598752e-01
1.21657324e+00 -4.28639889e-01 -9.72886741e-01 -3.62775505e-01
2.21300602e-01 -9.64070186e-02 -9.28887501e-02 -2.33012259e-01
-1.29843485e+00 1.12729943e+00 2.20543519e-01 4.25522178e-01
9.96209264e-01 -1.08148031e-01 4.46479648e-01 6.07684493e-01
1.93804055e-01 -8.79873037e-01 4.48974609e-01 7.86405683e-01
7.82301366e-01 -1.29183233e+00 4.52104211e-02 -4.82315063e-01
-3.02002311e-01 1.28469658e+00 5.99695086e-01 -1.24900445e-01
9.00215030e-01 1.27607286e-01 -2.13939250e-01 -3.92600298e-01
-9.27604258e-01 -1.70754731e-01 1.83692738e-01 8.86040032e-01
4.26791251e-01 3.00237775e-01 2.69407511e-01 8.54182184e-01
6.38718382e-02 -3.80493999e-01 7.18105257e-01 5.07339418e-01
-3.68551433e-01 -1.04233742e+00 -3.53044152e-01 5.34123421e-01
-5.65176941e-02 -1.05544358e-01 -7.97265172e-01 5.69772065e-01
1.65353373e-01 9.33747411e-01 2.77633548e-01 -8.01000237e-01
-1.82699651e-01 -2.29893520e-01 9.46555287e-02 -9.08602715e-01
-7.24733353e-01 2.05145717e-01 -2.13801116e-01 -1.43622652e-01
-6.44615769e-01 -6.40182257e-01 -1.23691213e+00 -5.82655609e-01
-6.34301901e-01 4.54213411e-01 7.66249076e-02 7.50492394e-01
3.57720971e-01 5.39667130e-01 7.07815647e-01 -7.92933226e-01
-2.54096299e-01 -1.11480498e+00 -6.95354044e-01 4.52826768e-01
4.15517926e-01 -7.99693346e-01 -2.70683557e-01 -1.29934713e-01] | [7.282864093780518, 5.950841903686523] |
daf1366a-5ee8-445d-b8e4-72ace7d35c63 | multi-view-redescription-mining-using-tree | 2006.12227 | null | https://arxiv.org/abs/2006.12227v2 | https://arxiv.org/pdf/2006.12227v2.pdf | Approaches For Multi-View Redescription Mining | The task of redescription mining explores ways to re-describe different subsets of entities contained in a dataset and to reveal non-trivial associations between different subsets of attributes, called views. This interesting and challenging task is encountered in different scientific fields, and is addressed by a number of approaches that obtain redescriptions and allow for the exploration and analyses of attribute associations. The main limitation of existing approaches to this task is their inability to use more than two views. Our work alleviates this drawback. We present a memory efficient, extensible multi-view redescription mining framework that can be used to relate multiple, i.e. more than two views, disjoint sets of attributes describing one set of entities. The framework can use any multi-target regression or multi-label classification algorithm, with models that can be represented as sets of rules, to generate redescriptions. Multi-view redescriptions are built using incremental view-extending heuristic from initially created two-view redescriptions. In this work, we use different types of Predictive Clustering trees algorithms (regular, extra, with random output selection) and the Random Forest thereof in order to improve the quality of final redescription sets and/or execution time needed to generate them. We provide multiple performance analyses of the proposed framework and compare it against the naive approach to multi-view redescription mining. We demonstrate the usefulness of the proposed multi-view extension on several datasets, including a use-case on understanding of machine learning models - a topic of growing importance in machine learning and artificial intelligence in general. | ['Tomislav Šmuc', 'Matej Mihelčić'] | 2020-06-22 | null | null | null | null | ['multi-target-regression'] | ['miscellaneous'] | [ 4.44002867e-01 2.20519990e-01 -2.22108930e-01 -6.11750662e-01
-3.39014262e-01 -7.45982289e-01 5.74765861e-01 4.67376232e-01
-4.74434486e-03 8.90545070e-01 -1.12584442e-01 -3.08653921e-01
-7.72442341e-01 -1.20810604e+00 -3.51216972e-01 -6.06632054e-01
-1.67710900e-01 1.07986951e+00 2.97951698e-01 4.66573536e-02
3.61902654e-01 5.26033521e-01 -2.36650157e+00 8.90057921e-01
8.44159305e-01 7.08760738e-01 1.01151317e-01 4.40738857e-01
-2.24774733e-01 1.09770909e-01 -2.54822791e-01 -3.79228771e-01
3.86979282e-01 -2.36560389e-01 -8.30910385e-01 2.15397775e-01
1.93975165e-01 2.40577668e-01 8.61421645e-01 7.97932267e-01
2.15210304e-01 -1.20302511e-03 9.71691966e-01 -1.41554952e+00
-7.43228644e-02 5.68371356e-01 -8.39994073e-01 -5.19907102e-02
3.41147304e-01 -4.08508897e-01 9.42614853e-01 -8.51583242e-01
1.02056158e+00 1.32937253e+00 4.86564070e-01 2.42381543e-01
-1.50498307e+00 -8.20795119e-01 3.36246550e-01 3.61540288e-01
-1.51971817e+00 -2.40412951e-01 4.03044611e-01 -6.04962945e-01
9.97217357e-01 7.29526997e-01 4.96148646e-01 7.19887555e-01
-8.23053066e-03 3.09554398e-01 1.68745744e+00 -5.98177791e-01
3.26821923e-01 7.97563374e-01 3.66727322e-01 5.01306355e-01
8.35029006e-01 -1.39039889e-01 -3.56457591e-01 -5.80862343e-01
1.74832940e-02 9.27448645e-02 3.65371592e-02 -7.94359565e-01
-8.01341057e-01 8.54148209e-01 -2.77707517e-01 3.07778001e-01
-3.39638710e-01 -6.70729518e-01 4.63518679e-01 2.48503178e-01
2.49525875e-01 4.09692943e-01 -8.25643241e-01 3.61849397e-01
-9.11622167e-01 1.81104884e-01 8.94809544e-01 1.09322846e+00
9.72773373e-01 -4.56663221e-01 2.79939026e-01 6.92949057e-01
2.48754755e-01 2.09680960e-01 4.38444883e-01 -4.81347144e-01
3.25267822e-01 1.22610247e+00 -1.54687077e-01 -8.42519522e-01
-5.82816839e-01 -3.05637300e-01 -8.22968423e-01 4.35415745e-01
2.89096776e-03 4.47552763e-02 -7.32155740e-01 1.51371717e+00
8.16780865e-01 -2.67040372e-01 3.09538811e-01 2.31610477e-01
7.90449560e-01 2.25280866e-01 1.05920574e-02 -6.82458699e-01
1.56795764e+00 -5.13184726e-01 -4.02928054e-01 3.44260693e-01
6.80980384e-01 -5.67158759e-01 4.50532168e-01 9.69828725e-01
-7.01877654e-01 -5.00948071e-01 -1.17463529e+00 5.53405643e-01
-7.85673320e-01 5.91317676e-02 6.51549697e-01 8.14727068e-01
-5.32617748e-01 6.98510051e-01 -4.81819034e-01 -6.83914542e-01
2.88683444e-01 7.14294255e-01 -7.58279502e-01 6.24366291e-02
-8.36822331e-01 8.01685810e-01 9.05537665e-01 -3.83861303e-01
-3.26260537e-01 -5.29229641e-01 -4.89398599e-01 1.24407284e-01
8.41703653e-01 -7.86556423e-01 4.70940500e-01 -8.54136825e-01
-6.86099470e-01 9.07694757e-01 -2.16790438e-01 -4.78495300e-01
2.61729538e-01 2.02273019e-02 -8.21100891e-01 -1.45551831e-01
1.94693670e-01 1.92992195e-01 5.15570283e-01 -1.58011913e+00
-9.70452368e-01 -8.28114450e-01 -1.40966371e-01 9.87960622e-02
-1.89365685e-01 -1.99005291e-01 -1.38933465e-01 -3.18122298e-01
2.73240954e-01 -1.08457315e+00 -2.48094022e-01 -6.21023834e-01
-7.13633120e-01 -1.68603137e-01 1.05322385e+00 -2.30608270e-01
1.41838467e+00 -1.57534468e+00 1.59260437e-01 6.92130566e-01
1.59179449e-01 1.64960504e-01 4.65418220e-01 7.28014290e-01
-4.59627628e-01 5.63720465e-01 -3.56100142e-01 1.76111404e-02
-5.20165861e-01 4.03913349e-01 -5.36123514e-02 2.03596130e-01
-6.80057183e-02 9.79772359e-02 -4.27012086e-01 -7.16963232e-01
5.37355803e-02 -6.49234056e-02 -2.29791999e-01 -1.00734625e-02
-1.60317421e-01 9.00227576e-03 -3.86463583e-01 5.93240678e-01
8.74294877e-01 -1.04417548e-01 7.31295526e-01 -3.77533525e-01
-2.73125827e-01 -1.02150515e-01 -1.90449476e+00 1.10920489e+00
-3.21124107e-01 3.26147452e-02 -3.57688099e-01 -1.14476871e+00
1.14886558e+00 3.78409326e-01 6.16604745e-01 4.42788983e-03
-1.44653156e-01 2.21727028e-01 5.80960959e-02 -6.19290948e-01
7.38206282e-02 -2.99864411e-01 1.15446560e-01 6.92106128e-01
1.39444843e-02 5.51255584e-01 5.51029027e-01 -2.53166836e-02
8.75366986e-01 2.06417799e-01 1.30171669e+00 -3.23002249e-01
9.48615730e-01 2.70319551e-01 7.71133363e-01 4.59152132e-01
4.56503808e-01 2.82422394e-01 5.10579348e-01 -7.25763202e-01
-1.04250050e+00 -8.69417012e-01 -3.60176444e-01 7.19743609e-01
-8.24979767e-02 -6.00959182e-01 -2.46466935e-01 -1.16963041e+00
9.21639428e-02 8.07593405e-01 -7.98138916e-01 2.59212375e-01
-3.56016368e-01 -1.07861030e+00 2.58891255e-01 1.65836036e-01
2.31849939e-01 -8.22439969e-01 -9.24205422e-01 1.02561317e-01
3.83812785e-02 -8.30078423e-01 4.85319823e-01 5.53119063e-01
-1.13601494e+00 -1.52769208e+00 2.15170011e-01 -4.66467559e-01
6.41608119e-01 7.81584308e-02 1.09868622e+00 -1.25980228e-02
-3.10618013e-01 2.01247483e-01 -5.38459897e-01 -7.73466229e-01
-5.49155951e-01 1.81803286e-01 1.18073314e-01 1.98931515e-01
6.35643721e-01 -9.30664062e-01 -8.30076411e-02 5.13720751e-01
-9.81832087e-01 2.81513840e-01 7.87783265e-01 7.46106505e-01
9.36194003e-01 3.44391912e-01 8.74274313e-01 -1.84620893e+00
4.42645103e-01 -7.36507356e-01 -4.48916256e-01 7.70359218e-01
-1.36268950e+00 5.09758115e-01 6.62115633e-01 -2.54235059e-01
-1.13238704e+00 4.79777873e-01 3.06573093e-01 -3.29537839e-01
-6.27925456e-01 5.69362044e-01 -3.66187513e-01 3.84771347e-01
6.67554736e-01 1.52686462e-01 -1.21192761e-01 -5.48638463e-01
3.61070007e-01 7.26874888e-01 1.07792251e-01 -4.46029752e-01
7.02181518e-01 6.09333754e-01 4.75424230e-01 -6.21151447e-01
-3.64569128e-01 -7.25852430e-01 -1.12308359e+00 -1.13334559e-01
6.27110124e-01 -5.77960014e-01 -5.44211388e-01 -3.41740608e-01
-8.78608763e-01 6.88593805e-01 -4.71176915e-02 3.14077497e-01
-4.80972230e-01 3.78229529e-01 3.08126628e-01 -9.53190327e-01
-5.47732890e-01 -8.95967901e-01 6.92788243e-01 -6.94336668e-02
-4.14218009e-01 -6.63570583e-01 2.39935592e-01 2.48946026e-01
-1.71273306e-01 7.84759104e-01 1.35656381e+00 -1.42909181e+00
-3.33263218e-01 4.89414595e-02 2.57631838e-01 -1.24574006e-01
3.15850705e-01 1.78307220e-01 -1.02860272e+00 -1.55336514e-01
-8.74096155e-02 -9.21258852e-02 7.46232390e-01 8.61066803e-02
1.00523126e+00 -3.58625442e-01 -9.43053424e-01 3.64051342e-01
1.71728516e+00 7.15926766e-01 4.21393454e-01 5.08119702e-01
2.04614908e-01 1.05642712e+00 9.17630434e-01 7.05912054e-01
2.05214679e-01 8.39689910e-01 3.14441800e-01 1.04495525e-01
4.97299641e-01 9.63311717e-02 -1.76071882e-01 4.27765518e-01
-4.43921417e-01 -2.20441684e-01 -8.52771521e-01 4.76619244e-01
-2.02093244e+00 -1.30078650e+00 -4.25419390e-01 2.47118449e+00
4.90181148e-01 1.44211814e-01 4.36039537e-01 5.43297589e-01
8.92460167e-01 -4.08095449e-01 -3.71514678e-01 -8.31695139e-01
1.70608200e-02 1.73877880e-01 3.19316864e-01 2.70771414e-01
-1.16138792e+00 5.41613281e-01 5.04936886e+00 5.75007379e-01
-7.34111667e-01 -1.37186795e-01 3.95161986e-01 6.01536185e-02
-4.87198889e-01 3.17341954e-01 -1.09488106e+00 3.27859595e-02
8.55668247e-01 -3.17059368e-01 -3.94318439e-02 1.03846097e+00
4.87564392e-02 -3.14718097e-01 -1.32215106e+00 6.19832993e-01
6.06809855e-02 -1.26140463e+00 2.89015740e-01 3.39822143e-01
6.34511650e-01 -5.29348433e-01 -3.99945498e-01 4.99483570e-02
2.22762644e-01 -8.88151586e-01 1.91030428e-01 3.69842738e-01
6.87443554e-01 -8.95820081e-01 7.27176011e-01 5.24161518e-01
-1.45822418e+00 -1.73098013e-01 -1.08536206e-01 1.86794072e-01
-6.82413653e-02 5.89922488e-01 -1.31364679e+00 1.20475447e+00
9.04851258e-01 4.43376988e-01 -7.62535870e-01 8.56917202e-01
3.01543862e-01 1.92467779e-01 -3.45068306e-01 -1.73928272e-02
-2.09234461e-01 -2.28524432e-01 4.96449590e-01 1.20979810e+00
3.79167020e-01 1.17915876e-01 5.93117476e-02 6.06607020e-01
4.58670527e-01 4.33011979e-01 -1.04188585e+00 4.38130140e-01
5.71708500e-01 1.37677443e+00 -9.82565284e-01 -4.69056159e-01
-5.28324008e-01 3.87886763e-01 2.06450194e-01 -3.64420786e-02
-4.73075897e-01 -2.67529517e-01 2.80209988e-01 4.53065246e-01
4.62226391e-01 2.26166740e-01 -4.64322507e-01 -8.64615262e-01
-9.76155847e-02 -1.05500829e+00 1.04739690e+00 -4.11611289e-01
-1.22528911e+00 7.81336427e-01 7.45165527e-01 -1.45012701e+00
-5.78298867e-01 -3.27892035e-01 -3.44645917e-01 6.58578575e-01
-8.72594178e-01 -1.26400006e+00 -1.90673813e-01 5.17563522e-01
6.73563123e-01 -3.78742784e-01 1.12649393e+00 3.23851639e-03
-3.04448396e-01 3.22265238e-01 -3.30249104e-03 -4.64028358e-01
6.75003171e-01 -1.31301498e+00 -4.00730856e-02 6.02216899e-01
3.22461069e-01 7.35662520e-01 6.65948212e-01 -9.20629323e-01
-9.28890169e-01 -1.05354440e+00 8.51094604e-01 -3.05546075e-01
7.05664232e-02 -4.37665254e-01 -8.95367801e-01 6.58049047e-01
2.07846791e-01 -2.75051266e-01 1.28457463e+00 5.03313363e-01
-2.78924853e-01 -1.57366529e-01 -1.40037262e+00 3.47607642e-01
9.40231204e-01 1.63820431e-01 -6.38151824e-01 -3.43641713e-02
1.17546618e-01 2.03787953e-01 -1.09464335e+00 7.27159321e-01
8.64544451e-01 -1.35167360e+00 9.29527581e-01 -8.40000331e-01
2.75486529e-01 -5.54449141e-01 -2.52913415e-01 -1.07980156e+00
-2.52558023e-01 -2.16646969e-01 -1.05251774e-01 1.49386191e+00
8.42341721e-01 -4.63573426e-01 7.89139211e-01 3.99141431e-01
2.54070759e-01 -1.10582614e+00 -7.26558268e-01 -6.47775114e-01
-4.26567584e-01 -2.15747818e-01 7.08699107e-01 1.11242604e+00
7.27875764e-03 7.26173401e-01 -3.67753893e-01 3.04207355e-01
6.51316226e-01 8.51313114e-01 1.00595224e+00 -1.86984849e+00
-4.60129470e-01 -1.26501517e-02 -5.40749848e-01 6.09237701e-02
-2.83162415e-01 -9.70777333e-01 -6.66209280e-01 -1.51424766e+00
4.06891942e-01 -8.00235152e-01 -3.44636180e-02 4.31724370e-01
1.08029068e-01 -2.50708669e-01 8.28232467e-02 2.38490731e-01
-2.92125225e-01 3.55450716e-03 7.27551281e-01 3.19446996e-02
-4.86105055e-01 3.74911427e-01 -6.94717944e-01 8.03718328e-01
7.28668988e-01 -8.81504714e-01 -7.87218750e-01 4.94394749e-01
3.15468639e-01 1.35215178e-01 -7.70056844e-02 -9.27911162e-01
6.66805133e-02 -3.84332955e-01 4.87065762e-01 -1.05351949e+00
3.03071719e-02 -1.14559281e+00 1.10743511e+00 5.65269947e-01
-2.26109654e-01 4.42129463e-01 3.79827153e-03 7.60705054e-01
-6.44145533e-02 -5.62191546e-01 6.79594755e-01 -5.27402818e-01
-6.62547827e-01 -6.53424859e-02 -2.93542624e-01 -2.18980387e-01
1.60002291e+00 -5.90383291e-01 -1.36945918e-01 1.29856810e-01
-1.11791694e+00 4.48575020e-02 3.21652263e-01 4.14286941e-01
4.80671853e-01 -1.15500534e+00 -5.68894207e-01 1.27630726e-01
3.81330013e-01 4.82453033e-02 5.29444590e-02 6.20238841e-01
-2.76599050e-01 3.03429365e-01 -5.19483387e-01 -5.79451323e-01
-2.02488565e+00 1.06824493e+00 -5.53654730e-02 -6.45537674e-01
-4.83219028e-01 2.01926544e-01 2.40737140e-01 -3.69344413e-01
-1.26177028e-01 -8.77290517e-02 -9.07348454e-01 3.95166814e-01
2.80878633e-01 6.20327473e-01 2.93292642e-01 -4.30222780e-01
-5.54997981e-01 6.39649212e-01 -2.80202478e-01 1.05394334e-01
1.59418082e+00 -8.61560777e-02 -1.32712543e-01 6.69868588e-01
8.79375398e-01 1.76811606e-01 -3.75783235e-01 -1.73601955e-02
5.02624512e-01 -2.86598295e-01 -6.21733189e-01 -9.23287690e-01
-7.10833788e-01 4.81063366e-01 6.80175662e-01 4.59802836e-01
1.32058358e+00 1.91123150e-02 3.76863629e-02 4.35304731e-01
4.42409277e-01 -1.04724431e+00 -4.87153262e-01 -1.78836539e-01
7.83625424e-01 -1.20383406e+00 6.08515561e-01 -8.25599253e-01
-7.91420043e-01 1.44540620e+00 8.23871851e-01 2.92534292e-01
7.58743227e-01 4.77192938e-01 -3.31583679e-01 -4.74232674e-01
-1.26511455e+00 -2.00079411e-01 2.18984097e-01 7.52744555e-01
3.25175375e-01 1.63469449e-01 -8.45443189e-01 7.19221890e-01
-9.27683637e-02 -1.22518286e-01 5.59271395e-01 9.91880596e-01
-5.34627438e-01 -1.58846390e+00 -4.63545352e-01 9.59840238e-01
-4.29995894e-01 1.06251046e-01 -6.88490093e-01 1.07712054e+00
8.59267771e-01 8.95252585e-01 -2.03740269e-01 -6.84551835e-01
3.95363331e-01 2.78110772e-01 2.29307622e-01 -7.59205937e-01
-7.46834874e-01 1.84356794e-02 5.47143042e-01 -1.33412570e-01
-8.46765995e-01 -1.07368731e+00 -1.12476432e+00 1.51932724e-02
-6.54173553e-01 2.65113354e-01 5.42786658e-01 8.92030954e-01
4.21170801e-01 2.19964191e-01 6.85562670e-01 -1.39562607e-01
-1.83117345e-01 -8.33423734e-01 -5.37186861e-01 4.99977231e-01
-2.46134385e-01 -1.02307153e+00 2.01801658e-02 2.45803192e-01] | [8.03699016571045, 4.815268516540527] |
c21170a9-b580-46ae-805a-3584e25a6566 | its-not-you-its-me-detecting-flirting-and-its | null | null | https://web.stanford.edu/~jurafsky/emnlp09.pdf | https://web.stanford.edu/~jurafsky/emnlp09.pdf | It’s Not You, it’s Me: Detecting Flirting and its Misperception in Speed-Dates | Automatically detecting human social intentions from spoken conversation is an
important task for dialogue understanding. Since the social intentions of the
speaker may differ from what is perceived
by the hearer, systems that analyze human
conversations need to be able to extract
both the perceived and the intended social
meaning. We investigate this difference
between intention and perception by using
a spoken corpus of speed-dates in which
both the speaker and the listener rated the
speaker on flirtatiousness. Our flirtationdetection system uses prosodic, dialogue,
and lexical features to detect a speaker’s
intent to flirt with up to 71.5% accuracy,
significantly outperforming the baseline,
but also outperforming the human interlocuters. Our system addresses lexical feature sparsity given the small amount of
training data by using an autoencoder network to map sparse lexical feature vectors
into 30 compressed features. Our analysis shows that humans are very poor perceivers of intended flirtatiousness, instead
often projecting their own intended behavior onto their interlocutors. | ['a'] | 2009-03-01 | null | null | null | emnlp09-2009-3 | ['dialogue-understanding'] | ['natural-language-processing'] | [ 1.68714002e-01 6.72502697e-01 -4.52854410e-02 -7.88732350e-01
-4.02202368e-01 -5.38367212e-01 7.85511553e-01 1.96857184e-01
-3.43511194e-01 3.32479775e-01 1.19900453e+00 5.18487170e-02
5.25703788e-01 -3.21824163e-01 -1.03812076e-01 -3.55214804e-01
2.58068480e-02 5.65188825e-01 -3.70740592e-01 -3.75779480e-01
2.01243445e-01 -2.28826255e-01 -1.52454078e+00 6.66253805e-01
2.89733469e-01 9.81450081e-01 1.26815364e-01 1.03772569e+00
-2.50952303e-01 1.26277030e+00 -8.25386882e-01 2.47558649e-03
-1.78605750e-01 -7.13086903e-01 -9.65207100e-01 3.31629127e-01
3.00785393e-01 -3.44723582e-01 -7.35101327e-02 5.97968817e-01
3.34764570e-01 1.00111112e-01 6.03177667e-01 -1.06122255e+00
-4.97008294e-01 1.25480390e+00 -3.79988737e-03 1.60142004e-01
1.24563599e+00 -1.26692250e-01 1.43208194e+00 -6.58188641e-01
4.28415030e-01 1.57555389e+00 6.69867098e-01 5.56189775e-01
-1.12506819e+00 -5.64041495e-01 5.74526377e-03 -1.32422075e-01
-1.12984204e+00 -1.13282704e+00 9.13281739e-01 -6.87856734e-01
1.16762161e+00 4.97039348e-01 5.77631235e-01 1.26053715e+00
4.57383096e-02 8.91829610e-01 9.21100557e-01 -4.08729672e-01
2.01373830e-01 5.58770120e-01 5.60988247e-01 2.76031733e-01
-8.37753832e-01 -1.34430826e-01 -9.17750180e-01 -4.70304579e-01
1.72458425e-01 -2.55490124e-01 -6.02383137e-01 3.07493687e-01
-1.50247967e+00 1.19144690e+00 2.52271503e-01 7.31905639e-01
-5.15108228e-01 -3.38544399e-01 4.90936577e-01 8.08743060e-01
5.30444324e-01 6.38019025e-01 -4.67068553e-01 -7.01690733e-01
-5.55284500e-01 4.46663685e-02 1.61897433e+00 5.17996609e-01
6.60076976e-01 -9.99550745e-02 6.45780191e-02 1.15801847e+00
3.03406745e-01 4.93572980e-01 9.83156681e-01 -9.23688531e-01
1.53363526e-01 6.39683008e-01 5.79431467e-02 -1.34420943e+00
-6.38243198e-01 2.13947713e-01 -5.10740757e-01 -5.43878734e-01
3.76059651e-01 -5.30696154e-01 -2.16089115e-01 1.73581207e+00
1.56170964e-01 -3.19916993e-01 4.40729558e-01 7.28780389e-01
1.08153009e+00 1.03574896e+00 -1.04221486e-01 -5.94148099e-01
1.47589111e+00 -6.09642506e-01 -1.01778984e+00 -5.88160992e-01
8.20895910e-01 -8.61277401e-01 9.82795358e-01 2.43373662e-01
-8.48799765e-01 -5.89273036e-01 -7.83634424e-01 3.31506059e-02
-6.66669756e-03 -1.46915495e-01 4.76262599e-01 7.45064676e-01
-9.65431392e-01 7.30138794e-02 -3.50616455e-01 -5.67407370e-01
-2.78596044e-01 2.84800440e-01 -3.59554052e-01 4.34718043e-01
-1.32177925e+00 8.39407206e-01 3.89915496e-01 -2.26254344e-01
-2.51769781e-01 -3.89948756e-01 -1.24157798e+00 7.69809857e-02
8.91184434e-02 5.28960042e-02 1.61076295e+00 -1.44605863e+00
-1.67865384e+00 8.42183828e-01 -4.90390331e-01 -6.37284219e-01
5.47067150e-02 -8.45115073e-03 -4.69477773e-01 4.99961600e-02
-2.38902848e-02 7.47649550e-01 9.38282371e-01 -7.16110706e-01
-5.83998680e-01 -2.19706386e-01 -2.38124654e-01 5.33213019e-01
-4.24822092e-01 2.01892853e-01 2.31466562e-01 -3.31700712e-01
2.73153424e-01 -9.14297819e-01 5.05522430e-01 -6.18611932e-01
-4.82811809e-01 -7.68265367e-01 1.09981978e+00 -6.71798885e-01
1.10344100e+00 -2.31602287e+00 1.81742743e-01 -6.63694693e-03
4.30721045e-01 -8.31002966e-02 6.86629713e-02 7.18147755e-01
6.40230626e-02 -2.30264708e-01 1.95035577e-01 -4.09595042e-01
1.68068424e-01 1.14926361e-01 -4.62316781e-01 5.67409158e-01
-3.00834715e-01 7.07626045e-01 -7.33253002e-01 -2.25252673e-01
1.58924058e-01 3.32190424e-01 -5.97254753e-01 7.03976333e-01
-1.48006439e-01 3.94837260e-01 -2.04153478e-01 6.47175536e-02
-1.34962350e-02 -7.34790787e-02 3.91859055e-01 -9.25990865e-02
-9.20037106e-02 9.27588344e-01 -7.99179673e-01 1.10008991e+00
-7.11839616e-01 1.03201222e+00 5.44495225e-01 -6.95270836e-01
1.20555699e+00 7.67125607e-01 2.96740949e-01 -4.67298627e-01
2.90007412e-01 -3.30399387e-02 3.36885720e-01 -4.37043577e-01
4.74967480e-01 -5.60388744e-01 -5.86053729e-01 1.12726247e+00
-7.72973523e-02 -2.26526722e-01 -2.87114739e-01 9.16612297e-02
8.24420393e-01 -8.01622570e-01 7.53791332e-01 -2.98700869e-01
5.81984282e-01 -3.08639705e-01 3.53214800e-01 5.61773479e-01
-4.16339576e-01 2.31161863e-01 6.60500169e-01 -5.73177874e-01
-5.30647457e-01 -6.09151304e-01 1.14723742e-01 1.72616792e+00
-2.83691704e-01 -3.38731050e-01 -7.83447564e-01 -4.38778639e-01
-1.43766910e-01 1.14133203e+00 -4.60016966e-01 -6.20008186e-02
-4.86939013e-01 -1.67017877e-01 4.87386167e-01 1.59504652e-01
2.06161350e-01 -1.44488394e+00 -5.80293417e-01 4.35587347e-01
-6.53619051e-01 -1.23280907e+00 -8.02052319e-01 2.16363277e-02
3.46598849e-02 -6.67848647e-01 -1.52268320e-01 -9.40802693e-01
9.74843800e-02 1.27448529e-01 1.04631162e+00 6.75732270e-02
2.34593078e-01 6.11480236e-01 -5.23047447e-01 -3.52767259e-01
-1.18957472e+00 9.74728391e-02 3.09038907e-01 1.96379662e-01
8.89653087e-01 -4.34412718e-01 -1.82150036e-01 5.32387912e-01
-4.24689680e-01 9.61548313e-02 5.43212779e-02 7.36352086e-01
-4.16058272e-01 -2.28525922e-01 8.10265481e-01 -5.82221448e-01
1.08503783e+00 -4.79714632e-01 1.52605593e-01 -2.20754847e-01
4.22810204e-03 -1.56435505e-01 4.52890784e-01 -4.93854046e-01
-9.56635296e-01 6.94576651e-03 -3.62319201e-01 2.69120038e-01
-6.65356934e-01 3.56681436e-01 7.44562447e-02 5.07705390e-01
6.06464624e-01 2.66435176e-01 3.88365865e-01 -2.05185875e-01
2.51637280e-01 1.58996344e+00 2.05295518e-01 -6.49250224e-02
8.22191313e-02 1.39787555e-01 -1.07299590e+00 -1.73103142e+00
-1.01015317e+00 -8.39388907e-01 -4.13845181e-01 -4.01712596e-01
9.97154295e-01 -9.15937245e-01 -1.19207704e+00 4.48493481e-01
-1.30026531e+00 -2.37096041e-01 -5.51451407e-02 4.93899584e-01
-6.77731216e-01 3.33729506e-01 -8.85885179e-01 -9.03393090e-01
-3.08381915e-01 -1.08800232e+00 9.85190570e-01 -2.74044216e-01
-1.42847192e+00 -1.03936732e+00 1.64981171e-01 8.10298979e-01
4.65323180e-01 -2.51232386e-01 4.93291825e-01 -1.53888273e+00
3.02474111e-01 -1.85827330e-01 -3.16392519e-02 3.65175873e-01
5.94275177e-01 -5.18254161e-01 -1.21464646e+00 -2.34872475e-01
6.61153138e-01 -7.69193888e-01 4.52463239e-01 3.56712341e-01
2.83202201e-01 -8.55419636e-01 -8.69296305e-03 -1.70655116e-01
3.57021511e-01 2.69161522e-01 8.92609581e-02 -3.52407277e-01
5.43021977e-01 1.14777017e+00 1.72025234e-01 5.34087181e-01
6.90088153e-01 5.59972286e-01 -1.51141033e-01 3.55823517e-01
5.77771664e-02 -2.39285827e-01 9.02647257e-01 1.34395969e+00
5.92892170e-01 -3.34396273e-01 -1.00284839e+00 4.18392271e-01
-1.59837973e+00 -1.05851078e+00 1.79841354e-01 1.67212451e+00
9.44450974e-01 1.16229035e-01 2.05228299e-01 1.98662102e-01
6.21505916e-01 5.94798446e-01 -3.46130639e-01 -8.60189259e-01
2.86854208e-01 -5.84794044e-01 -1.09456360e-01 1.08652985e+00
-9.89619553e-01 9.21813011e-01 6.39443731e+00 1.88057840e-01
-1.22131538e+00 1.89434141e-02 7.38354981e-01 2.00969577e-01
-1.64180338e-01 -3.20464373e-01 -7.33918607e-01 2.96354145e-01
1.41505098e+00 -2.62750357e-01 5.59122026e-01 7.00332820e-01
4.24330235e-01 -8.50584954e-02 -1.47439516e+00 9.53432143e-01
6.43710792e-01 -8.24581802e-01 -4.63746846e-01 -1.86175648e-02
1.28668904e-01 -1.03646763e-01 -2.11851195e-01 4.41240937e-01
2.68558770e-01 -1.15619063e+00 5.25266528e-01 3.35949630e-01
2.00539857e-01 -4.67078567e-01 8.17655385e-01 8.22792232e-01
-8.45695972e-01 -5.46020381e-02 8.71813744e-02 -9.11772907e-01
2.68235385e-01 2.51233846e-01 -1.75445259e+00 -5.37361741e-01
3.16017032e-01 8.38407040e-01 -8.09022561e-02 -1.20673448e-01
5.76988980e-02 8.81057262e-01 -4.93148208e-01 -5.46890736e-01
1.32014588e-01 7.88249522e-02 6.29026532e-01 1.30949187e+00
-3.56676988e-03 1.39625967e-01 6.52409434e-01 6.19827092e-01
-2.93830000e-02 3.02644730e-01 -8.14886630e-01 -4.84402537e-01
4.14799482e-01 8.15860987e-01 -3.29479665e-01 -3.08996290e-01
-3.65624309e-01 9.46528614e-01 2.83614039e-01 8.13117251e-03
-2.85529792e-01 -1.97642714e-01 7.11486995e-01 -9.50783715e-02
2.60688402e-02 -1.39341041e-01 -1.08245611e-02 -9.84121561e-01
-2.65040070e-01 -1.21232820e+00 1.43910572e-01 -6.17891848e-01
-1.25620389e+00 7.54523873e-01 -1.68419376e-01 -6.91568255e-01
-9.64285076e-01 -2.20476449e-01 -5.45495808e-01 7.40483284e-01
-8.01069975e-01 -8.37307930e-01 -1.08878985e-01 3.01521778e-01
1.07182407e+00 -2.53161371e-01 1.13207638e+00 -3.85155767e-01
-1.85379401e-01 2.70072341e-01 -6.20572507e-01 3.79110217e-01
7.57021129e-01 -1.08133173e+00 2.19507143e-01 -9.25247651e-03
2.38212273e-01 7.09621966e-01 1.13451040e+00 -3.41141224e-01
-1.31296730e+00 -4.50449854e-01 1.52966726e+00 -4.89796430e-01
1.07573140e+00 -5.80247760e-01 -9.10776615e-01 9.85288739e-01
7.77355850e-01 -8.14944267e-01 1.21171570e+00 5.49724758e-01
-1.96043208e-01 2.54187495e-01 -1.01140749e+00 4.84542966e-01
6.02851331e-01 -9.42402124e-01 -1.39265323e+00 4.77044642e-01
9.25303757e-01 -1.06167279e-01 -8.37016106e-01 -2.11373195e-01
6.27395153e-01 -1.30972457e+00 6.06407821e-01 -9.33487564e-02
2.11825520e-01 2.46143863e-01 -4.03777033e-01 -1.37919915e+00
1.34027675e-01 -7.44517624e-01 2.78124481e-01 1.12570179e+00
3.28101277e-01 -6.85577035e-01 4.00548846e-01 7.23982096e-01
1.25972629e-01 -6.25476986e-02 -9.10019577e-01 -6.85368925e-02
-1.67474627e-01 -5.93658626e-01 4.04408544e-01 1.09433138e+00
9.85595345e-01 1.13025284e+00 -5.67698598e-01 -1.24908805e-01
1.46861151e-01 1.35086745e-01 8.51666152e-01 -1.30127442e+00
-2.08667010e-01 -3.48989516e-01 -3.91165704e-01 -1.52871335e+00
7.05132484e-01 -7.58109868e-01 4.35643226e-01 -9.15676773e-01
-6.02938160e-02 6.24391437e-02 2.53326803e-01 4.91059065e-01
2.16699675e-01 -1.21749021e-01 1.18095845e-01 1.50018096e-01
-5.04705071e-01 6.80253923e-01 8.40849042e-01 -2.44905859e-01
-7.11035192e-01 1.05019197e-01 -7.47993827e-01 1.15857649e+00
6.29389584e-01 -9.16419402e-02 -2.65089184e-01 6.61893785e-02
1.44264460e-01 5.38646042e-01 3.74685787e-02 -8.08978379e-01
2.49372259e-01 1.89771801e-02 -1.30742282e-01 -4.44401681e-01
6.47310913e-01 -6.34445548e-01 -3.95079076e-01 3.22842628e-01
-1.01819813e+00 -1.17912233e-01 -7.93422088e-02 4.27169055e-01
-5.50732493e-01 -7.88067188e-03 6.11264944e-01 -1.33759156e-01
-7.01345980e-01 -4.54656571e-01 -1.17621410e+00 1.04622170e-02
7.91862488e-01 -1.03727847e-01 -4.12718505e-02 -1.32134235e+00
-9.42149580e-01 6.16505481e-02 3.01104307e-01 6.04391217e-01
6.72790468e-01 -8.41513813e-01 -8.02683830e-01 5.12098074e-01
8.05569291e-02 -6.59762681e-01 1.03653513e-01 6.72280610e-01
1.34986565e-02 5.66903234e-01 1.52039170e-01 -7.36323833e-01
-1.48323834e+00 1.57813817e-01 1.80097520e-01 3.26302648e-01
-5.51497757e-01 9.24617827e-01 2.69189268e-01 -8.36922646e-01
3.92798781e-01 -4.38892901e-01 -5.03568232e-01 6.22355163e-01
8.32586467e-01 7.75707364e-02 -2.69930422e-01 -1.34534442e+00
-2.53930658e-01 6.91869929e-02 -1.97242022e-01 -1.87902868e-01
1.07002294e+00 -5.05044520e-01 -1.58988580e-01 1.18625176e+00
1.64097786e+00 1.70908481e-01 -7.02307463e-01 -3.55499446e-01
2.49658842e-02 -1.09616511e-01 2.03610957e-01 -4.28683817e-01
-3.59027535e-01 6.01628065e-01 8.73356983e-02 1.07243419e+00
5.78890026e-01 3.60060900e-01 1.03181660e+00 7.79433131e-01
1.49715468e-01 -1.03357697e+00 1.73784301e-01 8.75508606e-01
1.27188075e+00 -1.60364234e+00 -2.86529809e-01 -2.43376970e-01
-1.30420697e+00 9.81227040e-01 1.28288805e-01 -2.58554574e-02
9.92068112e-01 1.41398488e-02 6.43030047e-01 -2.66631395e-01
-1.03326988e+00 2.01190665e-01 2.44041249e-01 2.66980916e-01
8.34058225e-01 4.06371027e-01 -4.63925898e-02 3.85343373e-01
-1.00242209e+00 -6.03672922e-01 6.00093901e-01 6.21263564e-01
-7.22579777e-01 -5.56483865e-01 -3.80841136e-01 4.61497217e-01
-2.17101544e-01 1.44559229e-02 -9.02505219e-01 3.75888735e-01
-3.69426817e-01 1.67802298e+00 3.90843272e-01 -5.35723090e-01
8.89395177e-02 5.11336148e-01 -3.57321352e-01 -9.38620746e-01
-8.06916356e-01 2.49153733e-01 5.64284205e-01 -5.24760187e-01
-6.96859896e-01 -9.46103811e-01 -1.11821079e+00 -2.51300782e-01
-1.81180805e-01 3.76083225e-01 4.07541364e-01 1.29855883e+00
7.38841891e-02 2.58251075e-02 9.08701003e-01 -8.17011893e-01
-4.40558016e-01 -1.48461282e+00 -5.84986687e-01 4.74425882e-01
7.69534707e-01 -2.05193967e-01 -8.93470287e-01 1.16226021e-02] | [12.75313663482666, 7.73417854309082] |
19519152-3d49-4ba9-91fb-1f222280df6a | biconditional-generative-adversarial-networks | 1911.01861 | null | https://arxiv.org/abs/1911.01861v2 | https://arxiv.org/pdf/1911.01861v2.pdf | Biconditional Generative Adversarial Networks for Multiview Learning with Missing Views | In this paper, we present a conditional GAN with two generators and a common discriminator for multiview learning problems where observations have two views, but one of them may be missing for some of the training samples. This is for example the case for multilingual collections where documents are not available in all languages. Some studies tackled this problem by assuming the existence of view generation functions to approximately complete the missing views; for example Machine Translation to translate documents into the missing languages. These functions generally require an external resource to be set and their quality has a direct impact on the performance of the learned multiview classifier over the completed training set. Our proposed approach addresses this problem by jointly learning the missing views and the multiview classifier using a tripartite game with two generators and a discriminator. Each of the generators is associated to one of the views and tries to fool the discriminator by generating the other missing view conditionally on the corresponding observed view. The discriminator then tries to identify if for an observation, one of its views is completed by one of the generators or if both views are completed along with its class. Our results on a subset of Reuters RCV1/RCV2 collections show that the discriminator achieves significant classification performance; and that the generators learn the missing views with high quality without the need of any consequent external resource. | ['Massih-Reza Amini', 'Anastasiia Doinychko'] | 2019-11-05 | null | null | null | null | ['multiview-learning'] | ['computer-vision'] | [ 1.07391529e-01 4.12398905e-01 -4.10360605e-01 -4.68683869e-01
-9.68201578e-01 -8.21881771e-01 8.45651925e-01 -4.17097248e-02
-6.64064735e-02 9.60759103e-01 -1.50858894e-01 -1.58441765e-03
1.87058821e-01 -8.09552133e-01 -9.72799659e-01 -9.24441159e-01
1.59039199e-01 1.12024844e+00 -1.23734362e-01 1.14331864e-01
-2.91180491e-01 2.19288636e-02 -1.70490646e+00 5.73334157e-01
8.02448511e-01 8.30358267e-01 5.12134790e-01 7.38672018e-01
-1.39832899e-01 6.63529336e-01 -7.92578459e-01 -4.57670122e-01
3.85043472e-01 -6.02423906e-01 -4.88177568e-01 3.54662299e-01
3.06346208e-01 -1.08569957e-01 3.70966524e-01 8.25118005e-01
1.72205284e-01 -2.76237845e-01 9.62288797e-01 -1.62747586e+00
-1.36734247e-01 7.27082610e-01 -4.35879856e-01 -2.47349039e-01
1.55731276e-01 -1.65269345e-01 1.16677511e+00 -7.46263981e-01
9.97551858e-01 8.96081388e-01 8.04001838e-02 5.00214219e-01
-1.23879313e+00 -5.75949192e-01 2.73939461e-01 -1.49683163e-01
-1.01084232e+00 -3.80027920e-01 6.91737890e-01 -5.68583429e-01
5.36593258e-01 1.27386987e-01 4.02176529e-01 1.27386689e+00
1.52745590e-01 6.50854766e-01 1.32305288e+00 -5.72658837e-01
3.81192923e-01 8.27835441e-01 -1.53691679e-01 4.48413402e-01
2.04759106e-01 -1.46135464e-01 -4.01991338e-01 -1.80929229e-02
2.32566267e-01 -1.22327268e-01 -1.38126627e-01 -4.93887633e-01
-9.29290593e-01 9.56389904e-01 7.27559924e-02 4.07061607e-01
-3.70705426e-01 -3.19548577e-01 1.83315098e-01 5.35984039e-01
6.09420419e-01 3.16031367e-01 -3.21100354e-01 3.56099635e-01
-1.00866795e+00 -4.20771092e-02 8.54999304e-01 1.00579321e+00
1.01422691e+00 -1.70256391e-01 6.05153255e-02 6.77436888e-01
-5.07286303e-02 4.71470207e-01 4.22465920e-01 -6.25166059e-01
8.82192492e-01 8.93961906e-01 -1.41689321e-02 -4.61268336e-01
-1.23114288e-02 -4.91488069e-01 -8.67674232e-01 4.76553649e-01
5.42504489e-01 -1.84377551e-01 -6.92212880e-01 1.95925927e+00
4.71915126e-01 -2.56287009e-01 4.19384658e-01 5.12257099e-01
6.27662480e-01 7.64430165e-01 -2.99777865e-01 -2.09413320e-01
1.13649416e+00 -8.81469250e-01 -4.52413231e-01 -2.99480379e-01
6.00483835e-01 -5.97715497e-01 8.05509031e-01 4.73147571e-01
-1.03709006e+00 -7.00946569e-01 -1.26571620e+00 1.94903508e-01
-2.98601031e-01 8.39883327e-01 2.48459280e-01 4.44344640e-01
-8.23358893e-01 2.06489116e-01 -5.81678808e-01 -9.41596851e-02
1.56742215e-01 2.65626937e-01 -8.33765030e-01 -3.54298204e-01
-9.52442408e-01 6.71250761e-01 3.55947882e-01 -2.68373817e-01
-1.29582012e+00 -2.52755612e-01 -6.88939273e-01 5.08946665e-02
4.77536887e-01 -6.60779178e-01 8.20611835e-01 -1.56307828e+00
-9.91270840e-01 8.48216057e-01 -8.63479152e-02 -2.87666589e-01
1.08920395e+00 2.77381748e-01 -2.87941158e-01 -6.72642738e-02
5.09218037e-01 5.50330579e-01 1.13224053e+00 -1.46034777e+00
-8.36701870e-01 -6.59808040e-01 3.43156427e-01 4.19053018e-01
-1.17451735e-01 -5.07625222e-01 -4.90046114e-01 -3.62455159e-01
2.25065202e-02 -9.89343107e-01 2.85488635e-01 -1.63009778e-01
-5.60958564e-01 -2.21589029e-01 9.14470434e-01 -7.10107565e-01
5.22763431e-01 -1.93885171e+00 6.42640650e-01 9.88821089e-02
1.28439497e-02 -7.93660954e-02 -1.09110445e-01 5.67838013e-01
-1.65657043e-01 1.75261497e-02 -1.70513853e-01 -6.92939460e-01
-3.35542500e-01 3.30977350e-01 -3.42267543e-01 5.10235846e-01
-7.31774494e-02 1.23370856e-01 -4.63301659e-01 -1.98863223e-01
-5.81085607e-02 3.95659566e-01 -2.91239679e-01 5.35120845e-01
-4.03177261e-01 4.81715143e-01 -2.08273217e-01 1.37937516e-01
5.96452951e-01 -2.07781151e-01 7.52663851e-01 1.60554014e-02
8.79119262e-02 7.38630295e-02 -1.66751027e+00 1.23345959e+00
-7.32642055e-01 3.46860647e-01 2.97841758e-01 -1.19033158e+00
9.32029188e-01 6.40108645e-01 2.42312178e-01 -4.01057959e-01
-2.52370059e-01 2.82703012e-01 -1.77305579e-01 -3.55014086e-01
2.42615506e-01 -3.01187932e-01 -2.67020259e-02 8.12618017e-01
5.84238350e-01 -9.66463890e-03 1.62753895e-01 2.61339456e-01
5.70702910e-01 -7.69050326e-03 3.07424098e-01 -1.84631012e-02
7.05063343e-01 -3.14363867e-01 7.79759049e-01 6.61364794e-01
3.85475785e-01 6.45296216e-01 8.58504772e-01 -1.94651023e-01
-1.23455334e+00 -8.23958635e-01 3.56697708e-01 9.10938382e-01
-2.21841425e-01 -2.38822654e-01 -6.27135217e-01 -1.17909825e+00
-2.64639348e-01 8.11308563e-01 -7.73073256e-01 7.76958093e-02
-1.14322416e-01 -3.14510643e-01 2.75704324e-01 3.06932449e-01
3.28720450e-01 -7.78990209e-01 -5.04473627e-01 -1.05936378e-01
-5.29119194e-01 -1.05281401e+00 -1.90469384e-01 4.46414262e-01
-6.34436488e-01 -1.34578693e+00 -6.52643085e-01 -6.82178676e-01
1.02685595e+00 6.41007796e-02 9.87911701e-01 -8.11677277e-02
2.09122792e-01 2.13234231e-01 -3.98771405e-01 -4.36334819e-01
-8.73703718e-01 2.48050302e-01 -4.38295975e-02 3.21412653e-01
-2.52416849e-01 -3.39424223e-01 -4.70787566e-03 2.59071320e-01
-9.43753183e-01 3.71590018e-01 2.67457813e-01 9.72499192e-01
5.97590625e-01 2.14500785e-01 6.77822590e-01 -1.35139573e+00
1.59716919e-01 -5.33749402e-01 -7.65288055e-01 4.93168712e-01
-3.36584270e-01 2.76881009e-01 9.39900458e-01 -3.12913001e-01
-1.05040514e+00 1.66467220e-01 1.95379794e-01 -4.03131396e-01
-1.67161629e-01 5.52743673e-01 -6.85561478e-01 6.93510413e-01
5.05630910e-01 2.56087035e-01 3.24793756e-02 -3.37379813e-01
2.15757221e-01 6.22172058e-01 1.79862633e-01 -4.55303699e-01
6.08693779e-01 4.15369481e-01 -1.03676975e-01 -6.60392880e-01
-4.72175717e-01 -2.60972917e-01 -7.09040344e-01 -2.96750307e-01
7.26699531e-01 -1.16820979e+00 -3.88841301e-01 4.19266433e-01
-1.07026041e+00 -9.67005491e-02 -4.29100037e-01 5.95637858e-01
-5.54993212e-01 -1.02919266e-02 -2.91338954e-02 -8.44188690e-01
-1.53802276e-01 -1.62766063e+00 8.82774293e-01 -1.71832368e-02
5.56191653e-02 -8.71660531e-01 9.68630612e-03 4.08836931e-01
-9.79394466e-03 1.87425151e-01 1.29340327e+00 -1.11914349e+00
-4.93336171e-01 -3.96584302e-01 4.39646065e-01 6.47610843e-01
1.61768317e-01 -8.45472515e-02 -1.07269132e+00 -5.44827759e-01
2.37465635e-01 -5.55257261e-01 7.89723992e-01 3.26770395e-01
8.03190589e-01 -4.19819772e-01 -3.26860785e-01 2.97527671e-01
1.56688869e+00 2.90638596e-01 2.43497118e-01 2.46199053e-02
4.94487196e-01 9.35913146e-01 3.84403795e-01 1.81129485e-01
3.89121950e-01 7.01837420e-01 5.44732034e-01 -7.92337358e-02
2.32754964e-02 -3.44039440e-01 5.44943094e-01 6.67513788e-01
-1.55370729e-02 -5.52161813e-01 -5.67612231e-01 5.21892369e-01
-1.47705078e+00 -9.59922254e-01 1.79775298e-01 2.61249566e+00
5.81241369e-01 -5.12524657e-02 1.06604248e-01 1.37982681e-01
8.37447941e-01 2.83641160e-01 -6.07489169e-01 -2.07221150e-01
-3.19056988e-01 -4.74625677e-02 5.49479723e-02 6.50542796e-01
-8.66488755e-01 5.04422307e-01 5.15318251e+00 5.89868069e-01
-1.06942773e+00 1.33745402e-01 7.19817519e-01 1.71447843e-02
-6.51099920e-01 3.57061237e-01 -9.41115856e-01 4.10804361e-01
6.52595282e-01 6.81905672e-02 4.04406488e-01 8.34087610e-01
-2.06526473e-01 -4.06097919e-01 -1.45119929e+00 4.47604746e-01
3.65101129e-01 -1.16486835e+00 1.89976349e-01 4.01199996e-01
7.75919497e-01 -1.55304700e-01 -3.62188332e-02 2.29855597e-01
3.76813024e-01 -8.79920185e-01 8.73322904e-01 3.78364086e-01
9.62660849e-01 -8.63677323e-01 7.02733517e-01 1.06519222e+00
-9.32312608e-01 -9.75672603e-02 -2.93737322e-01 2.24790826e-01
-1.66733131e-01 6.42275155e-01 -1.17629647e+00 1.00513387e+00
2.48802975e-01 4.77623314e-01 -5.34415841e-01 4.44334745e-01
-4.31940198e-01 4.29485202e-01 -3.35588306e-02 4.05871451e-01
-5.11534978e-04 -4.30685163e-01 7.98703372e-01 5.63022614e-01
4.99705583e-01 -3.64050359e-01 3.49887669e-01 8.22248936e-01
-2.85818666e-01 5.67187667e-02 -9.38860118e-01 -1.73281401e-03
3.48992914e-01 1.20015585e+00 -5.78429103e-01 -6.79400444e-01
-5.80961704e-01 8.85070801e-01 4.71529961e-01 4.07192945e-01
-4.80380446e-01 6.22230470e-02 2.65019536e-01 7.92209432e-02
4.29546803e-01 1.97483927e-01 -1.84916053e-02 -1.46716273e+00
1.55048966e-01 -1.13714540e+00 5.86156666e-01 -7.24055767e-01
-1.18728459e+00 6.35474920e-01 -8.50485489e-02 -1.30765736e+00
-6.90746486e-01 -2.08446383e-01 -4.11696166e-01 1.21404552e+00
-1.29377043e+00 -1.55089867e+00 -7.00018555e-02 6.69311464e-01
6.48865521e-01 -7.12451875e-01 9.76766944e-01 -2.76035070e-02
-3.27556431e-01 4.77860123e-01 5.68995886e-02 4.92831580e-02
7.53116965e-01 -1.29886758e+00 -1.20471999e-01 7.51327753e-01
4.85189676e-01 1.65330276e-01 6.81558132e-01 -7.04729736e-01
-1.20615077e+00 -9.95098352e-01 1.12617242e+00 -1.41426802e-01
-4.00330089e-02 -6.18084431e-01 -7.11071491e-01 1.04856396e+00
3.71218681e-01 -2.36248583e-01 7.26729870e-01 6.70768395e-02
-5.24107635e-01 -1.64830074e-01 -1.21865916e+00 1.49653241e-01
2.66406327e-01 -4.43126619e-01 -3.98075491e-01 3.20847034e-01
2.74977744e-01 -1.69863030e-01 -4.82716054e-01 3.23565245e-01
5.48680127e-01 -1.00106311e+00 5.47328830e-01 -4.84260231e-01
6.79269016e-01 -2.41631001e-01 -3.94470543e-01 -1.69586992e+00
2.13387042e-01 1.66920662e-01 1.88208744e-01 1.58236563e+00
6.22099817e-01 -6.26172006e-01 5.81640720e-01 9.25561786e-02
2.35243320e-01 -4.43057150e-01 -9.47142661e-01 -6.64669454e-01
8.31041783e-02 -2.53848165e-01 6.49500132e-01 9.44778085e-01
-4.03997540e-01 6.66834891e-01 -6.31216288e-01 9.12569836e-02
4.31509018e-01 5.30134976e-01 1.04920065e+00 -1.17899275e+00
-7.06938982e-01 1.84755892e-01 -1.44073367e-01 -6.08664334e-01
3.59850556e-01 -1.30229318e+00 -2.60058016e-01 -1.51642668e+00
4.34123427e-01 -5.73194027e-01 2.85617381e-01 4.05228168e-01
4.39401949e-03 -2.57396698e-01 3.05204779e-01 2.72362322e-01
-1.39054865e-01 3.32203835e-01 9.88202512e-01 -3.53122592e-01
-1.99990958e-01 4.27166611e-01 -5.39330125e-01 5.64029157e-01
5.65309286e-01 -5.22398710e-01 -7.73306787e-01 -3.00837934e-01
2.34814703e-01 6.43619418e-01 2.27255955e-01 -8.94964159e-01
2.35664025e-01 3.89423482e-02 6.98374629e-01 -5.55187285e-01
4.58652616e-01 -1.00511801e+00 5.74157476e-01 3.56663704e-01
-4.74056184e-01 3.14377725e-01 -2.46412337e-01 6.96509957e-01
-3.16415876e-01 -4.02850151e-01 6.62193775e-01 -4.30417776e-01
-1.36449888e-01 1.08124875e-01 -2.97888190e-01 7.20881019e-03
1.19159281e+00 -5.00163883e-02 -2.22978443e-01 -6.64098620e-01
-8.98325324e-01 2.85504013e-01 5.18211544e-01 4.27982807e-01
3.51421624e-01 -1.15861487e+00 -7.99283445e-01 4.97938484e-01
1.70066699e-01 -7.96336867e-03 2.05989584e-01 4.90312815e-01
4.94061671e-02 1.99541226e-01 -3.01071376e-01 -7.84819603e-01
-1.41181159e+00 4.57365870e-01 3.57273489e-01 -7.62811720e-01
-4.35438752e-01 4.07540977e-01 3.97357881e-01 -6.64113343e-01
1.47977814e-01 2.72133723e-02 -4.63097185e-01 7.66504049e-01
3.33617717e-01 1.47191226e-01 2.30696082e-01 -8.53448510e-01
5.30678546e-04 2.48421282e-01 -1.12685353e-01 -6.09114945e-01
1.43251395e+00 4.24924977e-02 -2.53329799e-02 9.27154422e-01
1.14009798e+00 2.59136278e-02 -9.98230994e-01 -2.29490161e-01
-5.15333772e-01 -3.81930619e-01 -3.89503211e-01 -8.15760195e-01
-1.27123654e+00 1.00912476e+00 3.05830389e-01 1.24276891e-01
1.13251770e+00 2.13207260e-01 1.01697400e-01 -7.22127855e-02
6.32569611e-01 -9.97796476e-01 3.99470367e-02 2.28817910e-01
9.31538999e-01 -1.19365358e+00 -1.26775160e-01 -2.91445762e-01
-9.83593643e-01 1.01991987e+00 5.96808732e-01 9.48715508e-02
3.99981439e-01 1.69323638e-01 -1.38000026e-01 -1.31996959e-01
-1.07283783e+00 1.46484256e-01 2.18864888e-01 5.90519190e-01
2.26768628e-01 4.21336323e-01 -2.02243134e-01 5.25318921e-01
-1.70382127e-01 -2.33844608e-01 6.31227255e-01 6.41365945e-01
-1.88139915e-01 -1.34041977e+00 -2.00730100e-01 5.37624598e-01
-3.32989067e-01 2.10992813e-01 -4.61787522e-01 8.02395880e-01
3.41808856e-01 8.01643491e-01 2.57813901e-01 -1.47727907e-01
2.36441955e-01 4.17448312e-01 3.81308377e-01 -8.70423734e-01
-5.98599911e-01 3.22589248e-01 1.02973685e-01 -8.38443637e-02
-4.81480330e-01 -9.78600979e-01 -7.24918008e-01 1.81085393e-01
-4.27495271e-01 2.53170341e-01 7.08156943e-01 1.03812885e+00
1.04456469e-01 2.62758613e-01 9.35753167e-01 -4.88351256e-01
-6.80449784e-01 -8.87947083e-01 -7.80290902e-01 3.30810100e-01
4.37746108e-01 -5.83360434e-01 -4.03456181e-01 2.28150979e-01] | [9.876120567321777, 2.9388506412506104] |
eeab7a78-59fb-4d99-bdce-204dc4513b1b | an-analysis-of-categorical-distributional | 1802.08163 | null | http://arxiv.org/abs/1802.08163v1 | http://arxiv.org/pdf/1802.08163v1.pdf | An Analysis of Categorical Distributional Reinforcement Learning | Distributional approaches to value-based reinforcement learning model the
entire distribution of returns, rather than just their expected values, and
have recently been shown to yield state-of-the-art empirical performance. This
was demonstrated by the recently proposed C51 algorithm, based on categorical
distributional reinforcement learning (CDRL) [Bellemare et al., 2017]. However,
the theoretical properties of CDRL algorithms are not yet well understood. In
this paper, we introduce a framework to analyse CDRL algorithms, establish the
importance of the projected distributional Bellman operator in distributional
RL, draw fundamental connections between CDRL and the Cram\'er distance, and
give a proof of convergence for sample-based categorical distributional
reinforcement learning algorithms. | ['Rémi Munos', 'Will Dabney', 'Mark Rowland', 'Yee Whye Teh', 'Marc G. Bellemare'] | 2018-02-22 | null | null | null | null | ['distributional-reinforcement-learning'] | ['methodology'] | [-2.69386977e-01 -3.33455615e-02 -4.64788109e-01 -3.24659675e-01
-1.06171799e+00 -9.32000220e-01 8.07766557e-01 2.81620085e-01
-6.77723110e-01 1.02703643e+00 1.31077647e-01 -6.87402129e-01
-7.24179447e-01 -8.59788299e-01 -6.50814354e-01 -8.42257559e-01
-4.49900925e-01 6.39603198e-01 -2.20681593e-01 -1.04061402e-01
4.50612754e-01 4.54814941e-01 -1.59212303e+00 8.40994194e-02
5.74175239e-01 9.96596575e-01 -2.12489471e-01 8.85375559e-01
-2.73448706e-01 1.04751384e+00 -6.53967023e-01 -4.69875723e-01
2.16701567e-01 -4.83955115e-01 -4.95885223e-01 -5.08264959e-01
5.95374666e-02 -3.93570244e-01 -8.81837085e-02 1.05762684e+00
5.63023806e-01 2.03796625e-01 1.30172002e+00 -1.34643662e+00
-9.21086550e-01 1.00405943e+00 -6.97532058e-01 1.82858542e-01
1.59811378e-01 9.56448168e-02 1.47045767e+00 -6.70858562e-01
4.94409770e-01 1.25318313e+00 7.08006322e-01 3.31516176e-01
-1.52388251e+00 -4.95345771e-01 -8.34051147e-02 1.26134887e-01
-9.82198894e-01 1.69198498e-01 6.64617836e-01 -4.45390135e-01
7.42906451e-01 -1.10860735e-01 6.03196740e-01 1.04462790e+00
1.38937682e-01 1.01776361e+00 1.53258193e+00 -5.89415073e-01
5.66384792e-01 -4.00691777e-02 -1.07069626e-01 2.13715404e-01
3.82952183e-01 7.27874935e-01 -1.38350293e-01 -5.69446981e-01
7.98016727e-01 2.42561800e-03 3.12224984e-01 -9.69187498e-01
-6.59074128e-01 1.50742900e+00 1.35371402e-01 2.18059778e-01
-4.67966706e-01 6.41624272e-01 6.62896931e-01 3.91229242e-01
4.20969486e-01 1.58777758e-01 -5.47334135e-01 -5.26487470e-01
-9.14920032e-01 7.33898580e-01 6.94339931e-01 8.42838824e-01
5.03131747e-01 4.96537425e-02 -3.20951521e-01 6.87168479e-01
4.41308051e-01 7.29306996e-01 2.86487818e-01 -1.13403034e+00
1.85704678e-01 3.18187661e-02 3.36893350e-01 -3.10280949e-01
-3.17047834e-01 -2.92556494e-01 -5.95922470e-01 2.84088224e-01
8.40045154e-01 -3.60202730e-01 -2.37400398e-01 1.80488801e+00
8.07434097e-02 2.48863678e-02 2.35450923e-01 3.59248221e-01
7.97071010e-02 4.37087506e-01 2.98043549e-01 -3.56799126e-01
6.43229246e-01 -2.47857198e-01 -2.98199028e-01 5.57082117e-01
8.00834179e-01 -4.04126853e-01 1.34118068e+00 5.02096355e-01
-1.07204700e+00 -3.52757931e-01 -7.05708802e-01 3.93248111e-01
-2.94732749e-01 -2.55313724e-01 6.50104880e-01 8.93469751e-01
-1.08316159e+00 7.40669072e-01 -4.76607442e-01 2.42269337e-02
5.53438902e-01 1.49815485e-01 2.79703796e-01 9.92637873e-03
-1.20939541e+00 6.68325067e-01 3.10760289e-01 -4.56362456e-01
-1.09124029e+00 -8.60116065e-01 -6.47031784e-01 7.86266848e-03
3.52097243e-01 -3.67381334e-01 1.70264733e+00 -6.91659033e-01
-1.51059687e+00 4.94817317e-01 3.13670099e-01 -9.53048825e-01
6.77420676e-01 -1.23362429e-01 -1.78989485e-01 1.19651869e-01
-6.30183443e-02 2.89058983e-01 6.88953638e-01 -1.08814752e+00
-8.39197814e-01 -2.30762303e-01 1.31033659e-01 -8.16609412e-02
-5.80630004e-02 -2.63250560e-01 6.25606418e-01 -6.27394795e-01
-7.44835198e-01 -6.06036365e-01 -1.31372645e-01 -4.21669990e-01
3.06986123e-01 -8.05655420e-01 1.13328740e-01 -1.33709848e-01
1.15343249e+00 -2.31873465e+00 -2.73888350e-01 4.17927027e-01
-9.17607024e-02 -2.00041067e-02 -2.38777235e-01 7.22144604e-01
1.64054483e-02 8.30561742e-02 -4.02118236e-01 1.64131239e-01
7.28094578e-01 2.70426393e-01 -5.64897418e-01 7.04214513e-01
-3.03394208e-03 9.99648511e-01 -1.24763095e+00 -2.65066087e-01
3.35184932e-01 1.35803506e-01 -7.86095858e-01 2.25769117e-01
-4.69799578e-01 -1.34972677e-01 -1.20817065e-01 3.47267240e-01
5.81481934e-01 2.23572999e-01 2.52648503e-01 4.64503795e-01
-1.68991864e-01 1.94060519e-01 -1.07292914e+00 1.22887707e+00
-4.30602789e-01 3.49507332e-01 -2.44465932e-01 -1.21413338e+00
9.24425006e-01 3.43425311e-02 6.43605828e-01 -7.80946553e-01
8.57266411e-02 3.05881798e-01 3.73925082e-02 -1.37398720e-01
5.04872262e-01 -7.55579054e-01 -3.00064474e-01 7.45072126e-01
1.23252429e-01 -3.30544233e-01 2.67148763e-01 -5.75839262e-03
9.14209664e-01 5.16228497e-01 3.56364965e-01 -6.40387893e-01
3.33567023e-01 -1.23304002e-01 7.83485770e-02 1.12283754e+00
-3.48714739e-01 7.57479221e-02 1.02184272e+00 -4.27293554e-02
-1.11087203e+00 -1.67189455e+00 -3.58037531e-01 1.52819896e+00
-4.49297845e-01 -1.39242232e-01 -7.43898690e-01 -8.92580092e-01
6.16400301e-01 1.25550687e+00 -8.64486873e-01 -1.57197699e-01
-7.52166882e-02 -6.09651625e-01 8.89007449e-01 7.97884166e-01
-1.17046483e-01 -1.17662621e+00 -7.57145584e-01 1.67727336e-01
5.00778079e-01 -2.60015339e-01 -1.44704968e-01 5.80369413e-01
-9.24642563e-01 -1.11059690e+00 -8.11940432e-01 -3.69336903e-01
6.80433586e-02 -3.47572833e-01 1.35858369e+00 -5.69095433e-01
-4.77085896e-02 8.46265078e-01 -3.39534670e-01 -6.36725366e-01
-3.85766327e-01 -1.17781460e-01 6.20017424e-02 -3.15536410e-01
5.48700035e-01 -4.55925554e-01 -4.46635872e-01 -1.04219519e-01
-8.98980618e-01 -1.03750336e+00 6.05817080e-01 1.12020433e+00
7.34172583e-01 6.76233545e-02 1.01106787e+00 -1.12242460e+00
1.06682205e+00 -6.70457900e-01 -7.71721482e-01 1.55584261e-01
-9.20910478e-01 4.46076214e-01 7.59491801e-01 -3.38766962e-01
-1.04273605e+00 -4.11239803e-01 -3.32309008e-01 -3.01874101e-01
-1.01605363e-01 6.12463713e-01 1.81039512e-01 6.06985927e-01
5.46108603e-01 3.80523890e-01 9.13384631e-02 -2.14935154e-01
9.18748915e-01 6.74153566e-01 4.17628884e-01 -1.16026759e+00
5.40131032e-01 2.28231281e-01 1.69668034e-01 -6.96738780e-01
-8.81626666e-01 -3.31108987e-01 -5.58052301e-01 -3.72079909e-02
5.41317046e-01 -6.14169061e-01 -1.03305364e+00 2.50686314e-02
-5.29123604e-01 -7.51593888e-01 -1.05102313e+00 5.08988023e-01
-1.33538091e+00 1.57549113e-01 -4.55741107e-01 -1.32130849e+00
-5.76776788e-02 -7.91104257e-01 6.33604288e-01 -1.59401119e-01
-1.08265363e-01 -1.39674330e+00 5.06958127e-01 -3.95753920e-01
4.01900530e-01 2.31213987e-01 1.38944018e+00 -8.86911690e-01
1.42014921e-01 1.94286197e-01 -2.78626949e-01 5.88689744e-01
-2.61353105e-01 3.87003385e-02 -7.28164852e-01 -3.75688612e-01
5.06464858e-03 -5.06767273e-01 7.70674884e-01 9.06467557e-01
1.28520501e+00 -1.13796808e-01 4.53158438e-01 3.91833931e-01
1.62789583e+00 1.40520170e-01 5.19156992e-01 3.85780334e-01
1.49219871e-01 6.12090290e-01 1.00115132e+00 9.84011292e-01
2.28629306e-01 -1.11691505e-01 4.33021516e-01 4.16253477e-01
3.07276040e-01 -6.06171131e-01 5.94935298e-01 5.12651145e-01
3.33634838e-02 2.01988459e-01 -6.82829499e-01 5.43641627e-01
-1.76107812e+00 -1.24626875e+00 -9.73361060e-02 2.40960789e+00
9.22950923e-01 3.83481145e-01 8.56332481e-01 8.14009011e-02
4.81634140e-01 6.00811392e-02 -6.54177845e-01 -9.52851892e-01
8.98777216e-05 7.93337226e-01 8.52059841e-01 4.42922175e-01
-7.07585514e-01 7.06831455e-01 7.48964977e+00 1.15451503e+00
-5.97284913e-01 7.25722089e-02 4.76569414e-01 -7.99545720e-02
-7.45731771e-01 2.44293530e-02 -5.34762561e-01 3.84134531e-01
1.32392550e+00 -5.20535648e-01 5.36943853e-01 1.29697514e+00
-9.65123400e-02 -1.20357670e-01 -1.17866814e+00 9.54727471e-01
-1.42121151e-01 -9.01492596e-01 -6.45950064e-02 3.40723097e-01
7.77148724e-01 2.14735493e-02 3.68036866e-01 9.63106930e-01
1.00584662e+00 -1.30437207e+00 7.69626856e-01 6.40968323e-01
8.67100239e-01 -1.56857502e+00 7.90437222e-01 3.66586953e-01
-8.71972740e-01 -3.79666418e-01 -8.31453621e-01 -3.23336393e-01
-4.12422895e-01 6.42209172e-01 -6.16518438e-01 4.47690278e-01
4.18187976e-01 7.44680464e-01 -3.93598348e-01 6.74416006e-01
-2.45634928e-01 1.02637863e+00 -8.79423767e-02 -4.35380697e-01
5.72095513e-01 -3.46983105e-01 7.81719536e-02 1.22564089e+00
3.60296428e-01 -2.62673616e-01 -6.35388494e-02 9.34418321e-01
-9.66564342e-02 2.32446402e-01 -8.11648548e-01 -2.79626489e-01
4.88773584e-01 9.22747135e-01 -3.80290240e-01 -1.54464945e-01
-4.69407797e-01 1.94952518e-01 3.85026097e-01 2.09201843e-01
-9.81154799e-01 -4.77401793e-01 6.83347762e-01 1.26736999e-01
6.24020040e-01 -6.22465312e-02 -1.44309446e-01 -6.02995872e-01
-3.36165130e-01 -9.35874283e-01 7.59783387e-01 -2.96331108e-01
-1.87706816e+00 -1.10440575e-01 2.50150770e-01 -1.09586787e+00
-6.02652729e-01 -9.38449383e-01 -3.19112748e-01 7.04643846e-01
-1.56056166e+00 -6.11645937e-01 5.65593302e-01 8.12881291e-01
1.44983143e-01 -3.24933231e-01 7.21281826e-01 -1.17959902e-01
6.27944320e-02 6.24579966e-01 9.87672746e-01 -7.33561488e-03
6.92069471e-01 -1.89890850e+00 -2.03945674e-02 -4.70783608e-03
1.75775796e-01 4.72257018e-01 6.46818161e-01 -3.50692183e-01
-1.56530118e+00 -9.59931433e-01 2.92824060e-01 -5.61594307e-01
1.05594790e+00 1.35349785e-03 -5.23280501e-01 5.55976689e-01
1.75315753e-01 -9.33793336e-02 8.81800771e-01 3.13175440e-01
-5.99343240e-01 -2.12470174e-01 -1.23981762e+00 3.90343815e-01
6.18281662e-01 -6.92190230e-01 -8.48936141e-01 -1.09098524e-01
3.53792518e-01 1.05989166e-01 -1.07320631e+00 1.52443513e-01
6.79101646e-01 -1.35689068e+00 9.03144896e-01 -6.30584300e-01
5.15866399e-01 9.64081436e-02 -4.87002522e-01 -1.49505055e+00
-2.12664261e-01 -4.65360105e-01 -1.64500996e-01 1.19113743e+00
4.75999974e-02 -6.90198004e-01 4.21767771e-01 -4.40441445e-02
1.79411188e-01 -9.32217538e-01 -9.28103089e-01 -1.20975602e+00
1.12516582e+00 -8.34982395e-01 7.06721485e-01 6.69552922e-01
2.69922674e-01 1.08756684e-02 -7.01018274e-02 -5.16381085e-01
9.49137568e-01 2.06079602e-01 5.35244048e-01 -1.25723231e+00
-5.52423298e-01 -9.55728829e-01 -1.26113862e-01 -7.73826420e-01
4.68951672e-01 -1.13850820e+00 -3.25002484e-02 -1.12545764e+00
1.60246745e-01 -4.49705362e-01 -7.16084003e-01 3.60926148e-03
9.57052112e-02 -1.72204804e-02 1.11956172e-01 -1.10369392e-01
-6.62637830e-01 7.19309568e-01 8.88458252e-01 4.14056070e-02
-5.55918105e-02 -3.50779085e-03 -7.85515428e-01 6.58705235e-01
9.60625708e-01 -4.86193866e-01 -7.03580797e-01 1.91655457e-01
3.96869957e-01 1.87294364e-01 2.70707786e-01 -5.27508914e-01
-2.35879809e-01 -3.83745581e-01 5.81892133e-01 -7.02835083e-01
-3.90291721e-01 -5.39020538e-01 -3.56237054e-01 6.33741617e-01
-8.37602913e-01 3.37739080e-01 5.64582013e-02 7.04467595e-01
-5.74981160e-02 -4.89358455e-01 8.17695022e-01 -5.58728315e-02
-2.45793015e-01 7.23969042e-02 -3.92751098e-01 6.49591267e-01
1.01614285e+00 2.10751846e-01 8.43183976e-03 -4.21424419e-01
-4.38020676e-01 5.16089275e-02 3.13021451e-01 -1.01183467e-01
4.64117706e-01 -1.36564410e+00 -8.43428135e-01 -3.57531682e-02
4.97231521e-02 -3.40629846e-01 -2.37121191e-02 6.94233716e-01
-2.73882598e-01 3.23611319e-01 -1.90075725e-01 -5.06156802e-01
-5.43182552e-01 9.27280962e-01 6.15129657e-02 -4.79657710e-01
-2.93503344e-01 4.71384585e-01 -2.09236890e-02 -6.41890109e-01
2.84339756e-01 -4.10425901e-01 8.04901645e-02 3.64709675e-01
4.62404102e-01 5.70967972e-01 -3.15914363e-01 -1.10808574e-01
-1.39999807e-01 2.94529706e-01 1.81326076e-01 -5.21246850e-01
1.41482913e+00 4.88802642e-02 5.97428950e-03 9.77671206e-01
1.06383669e+00 -6.79584667e-02 -1.47194111e+00 -1.04666837e-01
3.57578307e-01 -3.92853081e-01 -2.67279804e-01 -8.46478164e-01
-6.31705999e-01 1.11819613e+00 5.80288410e-01 4.74804789e-01
8.99275780e-01 -1.17768124e-01 3.84391755e-01 3.01021188e-01
4.51645613e-01 -1.38252997e+00 5.71884550e-02 5.92247605e-01
5.82813501e-01 -8.55062723e-01 1.91642363e-02 5.60230792e-01
-8.18305671e-01 1.09522116e+00 -2.74612214e-02 -5.24007738e-01
8.12213838e-01 3.79585981e-01 -3.70161206e-01 1.90424278e-01
-7.56834924e-01 -4.73327696e-01 -1.25065044e-01 9.85360920e-01
5.40580273e-01 3.85151416e-01 -4.12817061e-01 7.28046179e-01
-3.35046768e-01 -4.69469652e-02 4.62721318e-01 7.91545510e-01
-4.52039033e-01 -1.28013563e+00 -2.29447290e-01 5.99729240e-01
-5.89515924e-01 -2.06120923e-01 -1.55838430e-01 9.59949493e-01
-2.21791685e-01 7.37847805e-01 1.68174624e-01 3.63517888e-02
2.55154908e-01 3.79463822e-01 8.28598082e-01 -3.69688302e-01
-4.56633478e-01 8.42191577e-02 -4.09180075e-01 -1.05487488e-01
-3.04094225e-01 -1.12999368e+00 -1.21869767e+00 -5.33614933e-01
1.12009823e-01 4.79394913e-01 4.49159384e-01 7.86954165e-01
4.84732725e-02 2.41335467e-01 8.52166116e-01 -5.07387400e-01
-1.58759820e+00 -9.14258480e-01 -1.27562547e+00 4.47472811e-01
1.99417084e-01 -6.97214544e-01 -5.34872174e-01 -2.97334671e-01] | [4.053205490112305, 2.5638010501861572] |
3114461c-e1cd-434c-ada6-102b228f4dc9 | field-of-junctions | 2011.13866 | null | https://arxiv.org/abs/2011.13866v3 | https://arxiv.org/pdf/2011.13866v3.pdf | Field of Junctions: Extracting Boundary Structure at Low SNR | We introduce a bottom-up model for simultaneously finding many boundary elements in an image, including contours, corners and junctions. The model explains boundary shape in each small patch using a 'generalized M-junction' comprising M angles and a freely-moving vertex. Images are analyzed using non-convex optimization to cooperatively find M+2 junction values at every location, with spatial consistency being enforced by a novel regularizer that reduces curvature while preserving corners and junctions. The resulting 'field of junctions' is simultaneously a contour detector, corner/junction detector, and boundary-aware smoothing of regional appearance. Notably, its unified analysis of contours, corners, junctions and uniform regions allows it to succeed at high noise levels, where other methods for segmentation and boundary detection fail. | ['Todd Zickler', 'Dor Verbin'] | 2020-11-27 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Verbin_Field_of_Junctions_Extracting_Boundary_Structure_at_Low_SNR_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Verbin_Field_of_Junctions_Extracting_Boundary_Structure_at_Low_SNR_ICCV_2021_paper.pdf | iccv-2021-1 | ['junction-detection', 'image-smoothing'] | ['computer-vision', 'computer-vision'] | [ 2.21042454e-01 3.18418056e-01 -1.75921395e-01 8.69163200e-02
-8.61566663e-01 -7.83422351e-01 2.82593071e-01 4.22882915e-01
-1.45383835e-01 2.57120639e-01 -1.48501098e-01 4.89316396e-02
1.55931637e-01 -5.87310970e-01 -5.29764414e-01 -7.63118923e-01
-1.85744345e-01 1.48345843e-01 6.91023409e-01 -1.24740824e-01
4.78003502e-01 6.93138599e-01 -1.05116308e+00 4.68767807e-02
8.25019062e-01 8.55409324e-01 -9.14885998e-02 9.50025558e-01
-5.91764152e-02 -5.63170165e-02 -1.13941990e-01 -4.76841360e-01
4.67649281e-01 -4.50764224e-02 -5.95173895e-01 4.28899616e-01
7.54986703e-01 6.10066317e-02 2.05583617e-01 1.25052619e+00
1.72495246e-01 6.72904029e-02 7.75766969e-01 -9.48672116e-01
-7.86463141e-01 -1.49565160e-01 -1.35258937e+00 -1.06453940e-01
1.71093032e-01 -1.18273914e-01 1.07052886e+00 -1.08298516e+00
9.74178612e-01 9.93814528e-01 1.20664096e+00 1.67787209e-01
-1.45225585e+00 1.77785605e-01 3.34176868e-01 -4.62598711e-01
-1.62065947e+00 -2.45320767e-01 9.75551605e-01 -4.47277397e-01
6.05148137e-01 6.47211313e-01 6.73213363e-01 2.47566968e-01
5.31028986e-01 6.83163285e-01 6.29285455e-01 -5.53510904e-01
3.45849395e-01 -2.61842191e-01 1.77742928e-01 8.48756671e-01
2.08150849e-01 -3.85604203e-01 -4.59946357e-02 -4.46038425e-01
1.39646459e+00 -6.13124445e-02 -2.79384702e-01 -5.86139917e-01
-1.10442901e+00 5.65628290e-01 4.93000567e-01 3.29772711e-01
-2.99444199e-01 6.53046146e-02 -9.80466604e-02 2.44136155e-02
6.17757440e-01 1.07338428e-01 -4.25194293e-01 5.32200456e-01
-1.04711604e+00 1.47443891e-01 5.99239886e-01 8.71709108e-01
1.13211286e+00 -3.15833151e-01 3.46917659e-02 9.26702142e-01
4.70043719e-01 3.23725253e-01 -2.27027044e-01 -1.42618334e+00
1.73942745e-01 1.03488052e+00 2.22397670e-01 -1.38098538e+00
-6.35700524e-01 -1.26092032e-01 -1.05647683e+00 6.99877024e-01
5.57893515e-01 -2.08281338e-01 -1.20002115e+00 1.34521878e+00
8.25636625e-01 3.28140646e-01 -4.84619945e-01 8.01189840e-01
5.88900924e-01 5.34943461e-01 -4.35586795e-02 -2.32108325e-01
1.51015794e+00 -9.06045079e-01 -5.62640965e-01 -1.70455083e-01
4.31886077e-01 -9.87912357e-01 5.42913377e-01 2.43489578e-01
-1.39535856e+00 -2.85848975e-01 -5.89669645e-01 -4.67761904e-01
-2.03203678e-01 1.77811339e-01 3.95272106e-01 2.57313102e-01
-1.47245634e+00 4.98231381e-01 -5.78780413e-01 -6.65502921e-02
3.87041688e-01 2.54333824e-01 -4.26550776e-01 4.25796747e-01
-5.07397473e-01 6.74333632e-01 -1.42947897e-01 4.70149130e-01
-9.66141671e-02 -7.36402929e-01 -1.03577220e+00 -1.92332149e-01
2.02771261e-01 -8.75827372e-01 7.44638681e-01 -1.03106773e+00
-1.28319836e+00 1.46952510e+00 -6.44975245e-01 4.52950858e-02
6.77532911e-01 1.11816458e-01 -1.80265769e-01 3.40816021e-01
1.94112465e-01 7.69184768e-01 9.35721219e-01 -1.85149074e+00
-7.68550098e-01 -3.59215140e-01 -3.91810149e-01 2.22821698e-01
4.64869022e-01 -1.75417528e-01 -1.08846903e+00 -9.60360467e-01
8.34186554e-01 -7.79779255e-01 -8.60206246e-01 5.81239283e-01
-7.86037087e-01 -1.19291402e-01 8.11111867e-01 -9.60360646e-01
1.33467674e+00 -2.18448663e+00 8.97121876e-02 9.14405882e-01
5.96611679e-01 -1.41930476e-01 -6.35386780e-02 3.37550789e-01
6.03084452e-02 4.10754621e-01 -6.55624270e-01 -6.66080534e-01
-3.16182256e-01 5.85476577e-04 2.73802727e-01 9.04562354e-01
2.58064359e-01 1.02622175e+00 -6.30942464e-01 -7.90393591e-01
3.38706970e-01 6.83274567e-01 -5.19425154e-01 -2.78171748e-01
-6.27369881e-02 3.23343724e-01 -5.37284136e-01 9.92728114e-01
1.09262443e+00 -2.38968268e-01 1.21223673e-01 -1.98154181e-01
-6.06505096e-01 -6.65421128e-01 -1.71489263e+00 1.33537996e+00
3.59867210e-03 5.66397369e-01 1.01691616e+00 -6.23814225e-01
1.06207216e+00 1.65410206e-01 7.04624414e-01 -4.09567744e-01
1.02860980e-01 6.40245080e-02 -4.91350502e-01 -2.35077858e-01
4.27369773e-01 1.72825754e-01 2.49325827e-01 7.00387359e-02
-5.22590816e-01 -1.28844723e-01 -2.40923371e-02 6.49536178e-02
7.82580197e-01 -3.17868926e-02 2.35865369e-01 -6.15709901e-01
5.26464224e-01 1.53836936e-01 7.93413222e-01 6.11338913e-01
-3.38590056e-01 1.15756202e+00 4.33978945e-01 -5.56835473e-01
-1.10872042e+00 -1.21076083e+00 -5.41087925e-01 8.34577978e-01
7.80176759e-01 -1.41256943e-01 -1.03918576e+00 -1.93169177e-01
2.52206922e-01 4.35103737e-02 -8.97616625e-01 3.85983646e-01
-9.15785074e-01 -4.07825291e-01 -6.51701540e-02 3.18026721e-01
3.70484531e-01 -7.59629607e-01 -4.28926021e-01 2.34764397e-01
-8.86934921e-02 -6.72761381e-01 -1.13301551e+00 -7.47855892e-03
-1.07815123e+00 -1.24147439e+00 -9.33937550e-01 -1.30240488e+00
1.36138403e+00 3.74258906e-01 1.05347180e+00 5.17872453e-01
-4.71991360e-01 1.82881474e-01 -3.45299244e-02 1.93245217e-01
-1.37387231e-01 -3.74740839e-01 -5.03363431e-01 3.05879027e-01
-4.82917398e-01 -4.75211203e-01 -1.02951825e+00 5.19686580e-01
-5.80227792e-01 -1.28734872e-01 3.02289337e-01 4.99737710e-01
1.04108202e+00 -2.02041402e-01 5.29703982e-02 -7.20268726e-01
2.44580433e-01 -3.35287064e-01 -6.77011728e-01 4.51317489e-01
-1.43331856e-01 -4.44479644e-01 3.65940869e-01 -1.74716130e-01
-8.80445123e-01 2.21817613e-01 -3.75606827e-02 -1.92367032e-01
-7.96620101e-02 2.20462382e-02 -1.04867198e-01 -5.67653835e-01
4.83028948e-01 4.91540879e-02 6.53819069e-02 -6.09545052e-01
6.75327897e-01 2.25577489e-01 6.43728435e-01 -3.23545516e-01
7.29994059e-01 1.02660739e+00 1.03344709e-01 -1.20030653e+00
-3.96708310e-01 -7.31502414e-01 -1.09313607e+00 -3.38284016e-01
9.28815961e-01 -5.72254956e-01 -6.75614476e-01 5.16137302e-01
-1.30338061e+00 -2.73746371e-01 -4.44776654e-01 -3.00738178e-02
-3.66626501e-01 7.29306817e-01 -6.99701130e-01 -8.07417035e-01
-3.05815756e-01 -8.77729714e-01 1.26617503e+00 5.80688834e-01
-1.81217790e-01 -1.57415771e+00 2.68327862e-01 1.65828079e-01
1.69733793e-01 6.84871972e-01 7.66290724e-01 -7.08222985e-02
-7.11598754e-01 -2.97371566e-01 -2.67457366e-01 9.57375839e-02
6.57505170e-02 6.86344028e-01 -6.71200573e-01 -1.78054944e-01
-2.22364083e-01 3.98197711e-01 7.40736902e-01 1.16380835e+00
5.48387587e-01 -2.11974800e-01 -7.35463917e-01 8.92585218e-01
1.57197261e+00 2.10969016e-01 5.75206399e-01 2.08624899e-01
7.03456521e-01 4.88762856e-01 1.60435215e-01 1.82969630e-01
3.75109553e-01 4.22984213e-01 3.55699658e-01 -9.42502677e-01
-1.53160483e-01 8.17387700e-02 -1.29361317e-01 5.39976001e-01
-2.56904513e-01 3.76793183e-02 -8.79812002e-01 7.86780655e-01
-1.87800145e+00 -6.32399023e-01 -8.22504103e-01 2.08881593e+00
6.61814749e-01 -5.98961525e-02 1.43865034e-01 -1.86707765e-01
1.22031295e+00 7.31802434e-02 -4.92972642e-01 -4.61898893e-01
-2.96010733e-01 -1.53408751e-01 5.93409121e-01 1.07930624e+00
-1.39960229e+00 8.69627714e-01 7.93553591e+00 6.64793670e-01
-7.38547087e-01 -3.28089237e-01 9.02570009e-01 5.45145631e-01
-6.01211727e-01 2.11878806e-01 -7.83382952e-01 3.17311525e-01
-9.41577330e-02 2.58489162e-01 1.28822774e-01 7.32582510e-01
3.62240374e-01 -5.65610349e-01 -4.88889217e-01 7.36220181e-01
-2.12367192e-01 -1.73437381e+00 8.32467806e-03 -2.58829333e-02
1.12321973e+00 -1.87734321e-01 -9.56437886e-02 -5.96929312e-01
3.43210161e-01 -6.94481075e-01 7.13328838e-01 5.71358740e-01
6.02421582e-01 -4.34805781e-01 1.57599151e-01 2.40377769e-01
-1.63341033e+00 1.33157015e-01 -1.72721669e-01 1.66974977e-01
3.34577531e-01 8.12247276e-01 -2.03728721e-01 3.44002694e-01
4.55230504e-01 4.58814591e-01 -2.74730086e-01 1.25512397e+00
1.08542563e-02 1.45373747e-01 -7.50577986e-01 4.36382562e-01
2.22942576e-01 -1.00753748e+00 7.14804530e-01 1.38635969e+00
-7.87783489e-02 2.05480739e-01 3.28371853e-01 8.05269122e-01
7.25353658e-02 3.53975743e-01 -5.20122766e-01 1.03055668e+00
5.62894881e-01 1.51778364e+00 -1.48288655e+00 -2.57620782e-01
-3.81100535e-01 9.18501318e-01 4.19127166e-01 6.54502094e-01
-5.71676016e-01 -1.26284555e-01 7.00045705e-01 2.92484969e-01
1.53686598e-01 -3.95823985e-01 -9.38625872e-01 -9.21174705e-01
2.08293810e-01 -2.75394261e-01 3.52216810e-01 -4.82466340e-01
-1.04757440e+00 1.01402797e-01 -3.71333450e-01 -9.56631720e-01
6.55327618e-01 -3.66536707e-01 -1.17775118e+00 7.54813015e-01
-1.33603764e+00 -1.25674963e+00 -3.17941755e-02 4.84207153e-01
3.12733501e-01 6.60719335e-01 3.51152152e-01 6.00015372e-02
-8.24634612e-01 3.73282284e-01 3.28202605e-01 2.95483172e-01
3.45746666e-01 -1.49295509e+00 5.51564217e-01 9.55851316e-01
-5.87982610e-02 7.36530304e-01 3.43442500e-01 -1.02685034e+00
-1.16570151e+00 -1.08203351e+00 8.79028320e-01 -4.56373811e-01
4.39280987e-01 -2.83346921e-01 -1.06162202e+00 6.96084142e-01
8.34872276e-02 1.98060304e-01 1.72600046e-01 -1.31292984e-01
9.99616534e-02 2.35360831e-01 -1.23400438e+00 1.06025314e+00
9.55137789e-01 -1.68914601e-01 -4.26643521e-01 2.81332433e-01
2.03008100e-01 -3.93990308e-01 -9.14196730e-01 2.43908197e-01
4.53224629e-01 -9.63922620e-01 1.29314625e+00 -1.25659466e-01
2.27940781e-03 -4.51188743e-01 4.09321010e-01 -1.06856251e+00
-5.49545288e-01 -1.22833908e+00 2.89642274e-01 1.17147851e+00
5.33148348e-01 -5.99499345e-01 1.03269839e+00 7.07973182e-01
-3.06231529e-01 -9.96054471e-01 -1.31151700e+00 -3.96478653e-01
9.02988389e-02 -2.27467686e-01 5.80439856e-03 1.06234348e+00
-4.44693957e-03 -1.25973195e-01 2.13095434e-02 3.35020721e-01
9.07247424e-01 1.91433027e-01 5.93482316e-01 -1.22727084e+00
3.94529462e-01 -8.65697324e-01 -8.75903964e-02 -1.27873349e+00
-2.54809380e-01 -9.04976249e-01 1.10772364e-02 -1.65764737e+00
-3.85479033e-02 -6.87462628e-01 2.87778407e-01 3.00155401e-01
-4.44840223e-01 4.81764764e-01 -6.35190755e-02 4.00561541e-01
-4.22219574e-01 3.11062261e-02 1.77645659e+00 1.59452260e-01
-6.11310363e-01 -1.10312484e-01 -5.83187699e-01 1.20637286e+00
6.08269095e-01 -1.96992725e-01 1.38806835e-01 -2.25932255e-01
1.81925878e-01 9.02268142e-02 3.97305578e-01 -3.77862632e-01
4.75209862e-01 -2.43055210e-01 4.80538547e-01 -5.67531407e-01
2.45339364e-01 -7.07702518e-01 3.03973883e-01 1.12113133e-01
-6.09183637e-03 8.20508599e-02 6.57627881e-02 6.28910065e-01
-7.59507492e-02 -1.87607676e-01 1.23457229e+00 -2.10194200e-01
-7.04686284e-01 1.83923215e-01 -3.32431674e-01 9.56655815e-02
1.55833483e+00 -8.83994520e-01 -1.59250826e-01 -1.29002228e-01
-1.17817211e+00 4.42326277e-01 9.36464608e-01 -1.73091635e-01
6.07195497e-01 -1.12109184e+00 -7.14081109e-01 4.44365203e-01
-4.66898501e-01 4.95251626e-01 4.20601934e-01 1.04666924e+00
-9.62975502e-01 -2.74338245e-01 2.18493313e-01 -8.08412611e-01
-1.21477675e+00 3.09663117e-01 6.11348629e-01 -7.57013261e-02
-1.05616450e+00 1.06906033e+00 2.62345076e-01 -1.46619260e-01
-1.82216410e-02 -5.83057940e-01 5.02313152e-02 1.32009670e-01
1.76030710e-01 6.65409684e-01 -2.02322304e-01 -8.68544757e-01
-5.00160694e-01 1.58525491e+00 2.07743078e-01 1.05844811e-01
9.37003136e-01 -4.67322648e-01 -3.58591318e-01 1.75252408e-01
1.16036105e+00 4.13516998e-01 -1.58086002e+00 1.40127212e-01
9.63222757e-02 -4.09755230e-01 6.64380267e-02 -2.96395481e-01
-1.20841587e+00 3.72882754e-01 1.75933212e-01 4.53039020e-01
8.16549361e-01 -1.37008093e-02 7.60403395e-01 -2.53786474e-01
-2.17595603e-03 -1.28523946e+00 -2.52711236e-01 4.21211153e-01
8.10932755e-01 -1.14933550e+00 8.02635401e-02 -1.03908801e+00
-5.26619434e-01 1.16355193e+00 2.28245154e-01 -4.42851633e-01
1.05233908e+00 6.01938665e-01 4.47409242e-01 -3.65030944e-01
-2.60413885e-01 -2.09325209e-01 5.07491410e-01 6.51208222e-01
1.46691367e-01 3.43781188e-02 -1.95999682e-01 -2.52466891e-02
3.14355075e-01 -4.50881749e-01 2.22571909e-01 9.04894590e-01
-6.86402142e-01 -8.15261066e-01 -7.77392328e-01 3.13158453e-01
-2.66444236e-01 5.35083413e-02 -3.51380587e-01 9.50806081e-01
2.15427518e-01 8.88303459e-01 4.78089631e-01 1.58470750e-01
4.95300502e-01 -2.13282987e-01 8.27765986e-02 -2.98527330e-01
-6.35179102e-01 6.13736451e-01 -1.60265639e-01 -5.36357820e-01
-1.69991016e-01 -8.00198436e-01 -1.51274812e+00 -1.77968860e-01
-4.23814267e-01 -2.35380381e-02 4.25440580e-01 6.25957668e-01
5.26661813e-01 -6.35169297e-02 8.33041906e-01 -1.02407074e+00
-3.25028710e-02 -3.87370795e-01 -8.35610092e-01 4.72172350e-01
5.80270290e-01 -3.97193760e-01 -5.52022994e-01 5.52932918e-01] | [11.565929412841797, -2.7298026084899902] |
1b5db403-0a1c-4714-92e7-37983f2ec32f | bright-channel-prior-attention-for | 2305.12845 | null | https://arxiv.org/abs/2305.12845v1 | https://arxiv.org/pdf/2305.12845v1.pdf | Bright Channel Prior Attention for Multispectral Pedestrian Detection | Multispectral methods have gained considerable attention due to their promising performance across various fields. However, most existing methods cannot effectively utilize information from two modalities while optimizing time efficiency. These methods often prioritize accuracy or time efficiency, leaving room for improvement in their performance. To this end, we propose a new method bright channel prior attention for enhancing pedestrian detection in low-light conditions by integrating image enhancement and detection within a unified framework. The method uses the V-channel of the HSV image of the thermal image as an attention map to trigger the unsupervised auto-encoder for visible light images, which gradually emphasizes pedestrian features across layers. Moreover, we utilize unsupervised bright channel prior algorithms to address light compensation in low light images. The proposed method includes a self-attention enhancement module and a detection module, which work together to improve object detection. An initial illumination map is estimated using the BCP, guiding the learning of the self-attention map from the enhancement network to obtain more informative representation focused on pedestrians. The extensive experiments show effectiveness of the proposed method is demonstrated through. | ['Yechenhao Yang', 'Jinyu Xie', 'Chenhang Cui'] | 2023-05-22 | null | null | null | null | ['pedestrian-detection', 'image-enhancement'] | ['computer-vision', 'computer-vision'] | [ 6.05807006e-01 -5.49999356e-01 2.02114984e-01 -4.70279753e-01
-4.15229648e-01 -1.25981763e-01 4.36202079e-01 -3.58444542e-01
-6.22606575e-01 6.46513581e-01 1.13606647e-01 1.85211509e-01
3.98184150e-01 -8.95445287e-01 -6.05684936e-01 -1.13405406e+00
6.46852314e-01 -5.38303673e-01 4.19274390e-01 -9.84528735e-02
2.41375059e-01 1.36591524e-01 -1.76288676e+00 3.17703843e-01
1.19442368e+00 8.76081705e-01 6.73766851e-01 6.64694905e-01
1.55378819e-01 8.49725723e-01 -2.67826825e-01 -3.49907160e-01
2.34699056e-01 -2.70433158e-01 -2.82754630e-01 5.26696086e-01
5.05346835e-01 -8.10013175e-01 -4.73418266e-01 1.31495559e+00
8.14725757e-01 3.15228671e-01 4.55524862e-01 -1.23184574e+00
-7.85782099e-01 1.25662506e-01 -8.22539985e-01 5.54133594e-01
7.21325725e-02 5.69908440e-01 6.74245894e-01 -9.97708440e-01
1.88940421e-01 1.08848155e+00 4.69768941e-01 3.05664212e-01
-9.66219425e-01 -5.63284039e-01 1.46986008e-01 7.05464840e-01
-1.27126229e+00 -4.45170581e-01 1.06106484e+00 -1.12181716e-01
7.88397729e-01 9.76732671e-02 9.02501345e-01 7.38160193e-01
1.17307827e-01 9.71162975e-01 1.43602753e+00 -3.87631267e-01
-1.18938982e-02 4.39589262e-01 1.21565789e-01 7.30544984e-01
5.12065530e-01 5.37672639e-01 -5.00583827e-01 4.93678719e-01
6.07895970e-01 2.26432160e-01 -1.19178340e-01 6.50389865e-02
-6.76857531e-01 4.83396649e-01 7.56824315e-01 1.91596776e-01
-5.85347295e-01 9.88064259e-02 5.89293391e-02 -3.62807989e-01
3.70563835e-01 -1.72934428e-01 -1.24192908e-01 3.88891369e-01
-7.92198956e-01 -1.68808460e-01 -9.00686756e-02 6.53566897e-01
1.00609767e+00 6.06416054e-02 -4.39958453e-01 9.82455134e-01
6.38542235e-01 7.82692909e-01 2.51951348e-03 -7.27522075e-01
3.72329950e-01 5.72347820e-01 1.91816553e-01 -1.02570605e+00
-2.27840453e-01 -6.78716242e-01 -9.00410712e-01 3.06706756e-01
1.59944370e-01 -1.77494779e-01 -8.17033529e-01 1.52231801e+00
4.78445321e-01 1.38652116e-01 4.22446653e-02 1.31251585e+00
6.73955202e-01 8.14847052e-01 3.99580538e-01 -2.80910701e-01
1.41020870e+00 -1.06561220e+00 -6.24399602e-01 -4.90470946e-01
3.75945307e-02 -8.64478767e-01 8.41570616e-01 2.70694405e-01
-1.12496376e+00 -1.04045403e+00 -1.11688423e+00 -3.52714598e-01
-4.32347178e-01 7.21695900e-01 5.73076785e-01 8.95532191e-01
-1.06431091e+00 1.02636434e-01 -6.02898955e-01 -4.10517514e-01
4.74428505e-01 1.56032264e-01 1.05607457e-01 -3.97968620e-01
-1.26180243e+00 8.96682978e-01 4.69253302e-01 5.94793856e-01
-1.08065081e+00 -2.79935688e-01 -7.59217381e-01 3.58726531e-02
1.43051848e-01 -7.47514307e-01 6.99151576e-01 -9.58151996e-01
-1.46127880e+00 5.63126206e-01 -2.92554945e-01 -1.97488025e-01
2.99916625e-01 -2.74919391e-01 -3.09253961e-01 2.88208365e-01
7.18958825e-02 7.78134763e-01 1.03046298e+00 -1.47400987e+00
-1.10192955e+00 -2.90475190e-01 9.17834342e-02 6.28472865e-01
-6.69536173e-01 1.33607522e-01 -6.79287553e-01 -3.51656824e-01
-8.67196321e-02 -5.84064186e-01 -1.60270676e-01 -1.86055928e-01
-3.09859365e-01 5.88140450e-02 8.88547242e-01 -9.28010583e-01
1.08218122e+00 -1.90495312e+00 -2.91096747e-01 -4.51828763e-02
-6.37967587e-02 4.59980428e-01 -2.56007135e-01 -1.23190865e-01
-7.40250293e-03 -4.47728544e-01 -2.04389513e-01 -2.71379620e-01
-2.05523804e-01 -2.84967534e-02 1.95695199e-02 5.93922853e-01
4.81546313e-01 7.83668816e-01 -9.47064400e-01 -7.74575055e-01
7.45594025e-01 9.83115196e-01 -4.25013661e-01 1.94920108e-01
2.69993335e-01 3.67814124e-01 -3.17988425e-01 7.19290853e-01
1.16984010e+00 -7.44274929e-02 -1.93834722e-01 -7.52179503e-01
-5.37064135e-01 -2.48527244e-01 -1.22828281e+00 1.27004373e+00
-4.31331843e-01 7.46225834e-01 8.28415975e-02 -7.36849189e-01
7.41290867e-01 -6.38021082e-02 3.84568483e-01 -9.73198414e-01
3.62976551e-01 -6.09510913e-02 -1.64741039e-01 -7.04449296e-01
7.41247773e-01 2.19906583e-01 5.35629869e-01 3.00050229e-01
-7.53818378e-02 1.27020836e-01 2.76402295e-01 3.48680466e-02
4.47634190e-01 2.69034475e-01 -6.03010133e-03 1.02779746e-01
9.84570146e-01 -2.66217291e-01 4.94952023e-01 5.82404196e-01
-4.93866801e-01 2.74000823e-01 -3.77499998e-01 -2.65285999e-01
-1.17387187e+00 -8.74655426e-01 -1.37009263e-01 1.35155165e+00
6.42724633e-01 -6.31146953e-02 -7.85846591e-01 -3.55181068e-01
-3.57421011e-01 6.94021821e-01 -6.10427856e-01 -2.98695832e-01
-4.54337180e-01 -1.31111848e+00 1.75780445e-01 5.69720507e-01
1.39034164e+00 -8.27461720e-01 -8.86877418e-01 9.56239402e-02
-5.43445230e-01 -1.06788480e+00 -5.03073871e-01 9.12684128e-02
-4.54687834e-01 -9.68160152e-01 -8.30338717e-01 -7.77477980e-01
8.43727350e-01 9.36620355e-01 5.56149364e-01 -5.94420172e-02
-6.69139326e-01 5.76301754e-01 -1.05443627e-01 -4.02801275e-01
5.15274927e-02 -2.16671556e-01 -2.11610720e-01 5.23629785e-01
4.91581053e-01 -1.74440950e-01 -1.07833743e+00 3.12662035e-01
-8.44521582e-01 3.03915471e-01 1.03327155e+00 9.05335665e-01
4.69332129e-01 3.07850540e-01 1.87145606e-01 -2.51043379e-01
2.82438666e-01 -1.37878852e-02 -6.59809172e-01 3.91451925e-01
-5.46278238e-01 -1.94188610e-01 3.51552904e-01 -1.74893945e-01
-1.90070713e+00 2.21273631e-01 -1.09517291e-01 2.09886897e-02
-2.78821200e-01 -1.12630904e-01 -4.28286403e-01 -3.38449478e-01
4.23085749e-01 7.28223622e-01 -4.03572142e-01 -1.27921984e-01
4.91124302e-01 7.94455945e-01 5.70368826e-01 -1.89225838e-01
7.88906395e-01 7.05215693e-01 -3.01443152e-02 -8.98217857e-01
-8.87054145e-01 -5.45730710e-01 -7.17670739e-01 -7.34193742e-01
1.02574360e+00 -1.03620875e+00 -8.30712020e-01 6.79991543e-01
-1.22224391e+00 -5.35637029e-02 1.60771817e-01 5.11354685e-01
-2.65913427e-01 4.23837185e-01 -6.45028830e-01 -1.27400970e+00
-4.61912841e-01 -1.18713319e+00 1.13522124e+00 7.40751445e-01
7.60809600e-01 -7.30589271e-01 -4.09272820e-01 4.72245008e-01
5.86141109e-01 -2.21720695e-01 3.50898504e-01 2.05342636e-01
-8.53430688e-01 5.16062602e-02 -9.27276850e-01 5.33841193e-01
5.01179211e-02 -1.10167921e-01 -1.38040674e+00 -1.74919039e-01
-1.76320001e-01 -6.52763769e-02 1.30774748e+00 7.35585153e-01
1.02873969e+00 2.56582964e-02 -2.91774690e-01 6.38144493e-01
1.66656208e+00 2.74905354e-01 8.15276802e-01 4.55839217e-01
7.13617980e-01 7.24598706e-01 7.40724921e-01 5.32175958e-01
4.67343569e-01 5.31329572e-01 5.19307077e-01 -6.15840554e-01
-5.16344070e-01 8.84920284e-02 4.96100903e-01 3.08346748e-01
-3.70372891e-01 6.49924874e-02 -5.44778824e-01 4.46700186e-01
-1.65657020e+00 -1.33244443e+00 -2.55873740e-01 2.08582163e+00
7.35701084e-01 -8.60376582e-02 6.32634386e-02 1.40247680e-02
1.09966314e+00 2.60639712e-02 -6.10504448e-01 1.20262682e-01
-4.69790071e-01 -1.93448022e-01 5.54067135e-01 5.07057250e-01
-1.33793199e+00 9.02122378e-01 5.94621038e+00 6.84694946e-01
-1.03590763e+00 1.70703471e-01 8.45043361e-01 3.80105264e-02
8.52167383e-02 -1.40027508e-01 -1.08673751e+00 5.91539204e-01
3.58658254e-01 1.48945361e-01 5.13763130e-01 6.04161501e-01
5.39085805e-01 -4.35623527e-01 -5.18542290e-01 1.16713536e+00
4.01815385e-01 -7.19344676e-01 -1.56020284e-01 1.75148211e-02
8.66281450e-01 -2.34209552e-01 5.17860353e-01 4.41200100e-02
6.95899725e-02 -5.98882973e-01 5.64392030e-01 7.19750702e-01
5.36731422e-01 -8.44267845e-01 6.37738883e-01 1.02690503e-01
-1.47432339e+00 -4.24040109e-01 -6.59934044e-01 1.37772024e-01
1.81894884e-01 6.25167906e-01 -5.85736752e-01 4.80034292e-01
6.72514081e-01 7.77798235e-01 -7.85643518e-01 1.24393249e+00
-4.25180286e-01 4.54486609e-01 -1.83753818e-01 3.92239429e-02
1.81822434e-01 -2.66728789e-01 4.10062224e-01 1.39529574e+00
1.00468256e-01 2.51024127e-01 1.52273402e-01 8.66229296e-01
3.17825973e-01 -1.69470571e-02 -3.55580181e-01 2.99887925e-01
1.98140785e-01 1.59375310e+00 -6.85086668e-01 -5.06600618e-01
-5.73671341e-01 1.09777939e+00 -6.93555251e-02 7.05704808e-01
-1.11178076e+00 -5.05064130e-01 1.84684932e-01 -6.64689839e-02
2.86926240e-01 -3.40681933e-02 -3.30693513e-01 -1.10309410e+00
-1.66181713e-01 -5.36971688e-01 3.50057036e-01 -1.33257186e+00
-1.05013692e+00 4.76467282e-01 -2.39817917e-01 -1.18645442e+00
3.05963069e-01 -6.84499264e-01 -3.85535508e-01 1.03349316e+00
-2.09072566e+00 -1.49167168e+00 -8.69318426e-01 8.09872031e-01
6.17527783e-01 2.99052708e-02 5.01604490e-02 6.52459145e-01
-9.28379238e-01 4.45472062e-01 -3.07389386e-02 -1.83005352e-02
8.04323852e-01 -9.80712652e-01 -1.39337599e-01 1.57371497e+00
-3.48107964e-01 4.86001253e-01 5.56942165e-01 -6.94821477e-01
-1.20798361e+00 -1.39055896e+00 4.64287847e-01 -7.32247531e-02
1.21722154e-01 1.27006359e-02 -6.16136432e-01 3.45819235e-01
5.01156688e-01 -6.13478497e-02 3.62884581e-01 -2.27840766e-01
-4.12821807e-02 -4.56721038e-01 -9.05862510e-01 5.57320595e-01
8.13605666e-01 -3.44037831e-01 -3.14192712e-01 2.48372152e-01
2.74073929e-01 -8.37123469e-02 -3.53924394e-01 2.15304881e-01
5.13887048e-01 -9.89090085e-01 1.28023958e+00 -2.98428209e-03
4.14783597e-01 -7.39406943e-01 -2.28616238e-01 -1.17882717e+00
-5.89398444e-01 -1.31873041e-01 5.63674346e-02 1.39614880e+00
-2.48834901e-02 -4.89090890e-01 5.98684788e-01 5.50396144e-01
-1.48908719e-01 -3.38719964e-01 -4.48967457e-01 -2.47523203e-01
-6.13203585e-01 -3.34785402e-01 4.50800091e-01 5.99795759e-01
-4.25379068e-01 2.93694407e-01 -6.34491980e-01 4.91434366e-01
1.27072239e+00 6.04429021e-02 6.30535305e-01 -9.13864076e-01
-1.77367836e-01 -3.92047495e-01 -1.51729316e-01 -1.00544775e+00
-2.67343938e-01 -5.85440099e-01 3.85886103e-01 -1.76618290e+00
7.60292649e-01 5.16449921e-02 -4.76658344e-01 4.04702425e-01
-7.93940604e-01 6.40034854e-01 1.52426496e-01 -6.75367415e-02
-8.54273975e-01 6.95346177e-01 1.33911431e+00 -4.52174723e-01
-1.50890633e-01 -2.91358709e-01 -7.02648461e-01 7.22896397e-01
6.84342265e-01 6.59286696e-03 -3.07805151e-01 -6.29430652e-01
-8.53776857e-02 -3.66327196e-01 7.81797886e-01 -1.18774259e+00
5.67745268e-01 -1.97209269e-01 1.15237951e+00 -9.31475282e-01
4.67849731e-01 -8.84637475e-01 -3.36489320e-01 5.21277606e-01
-2.18829364e-01 -3.69532406e-01 1.41902208e-01 5.92075646e-01
2.53270827e-02 -6.49548844e-02 1.09098375e+00 -8.63146037e-03
-1.19570398e+00 3.20100009e-01 -2.97572255e-01 -4.70622718e-01
1.06499517e+00 -4.09465164e-01 -5.46873391e-01 -5.31634428e-02
-4.85467225e-01 2.65496671e-01 1.63014501e-01 1.40609801e-01
9.57500339e-01 -1.22694016e+00 -7.84659803e-01 4.75701630e-01
1.00905932e-02 -5.27216375e-01 8.69836509e-01 9.53451753e-01
-2.27257937e-01 3.73736590e-01 -5.24968922e-01 -6.91663742e-01
-1.38034141e+00 6.31126881e-01 4.39832240e-01 1.23148642e-01
-3.39738011e-01 7.80112028e-01 4.38202322e-01 1.99781939e-01
1.11721843e-01 -1.63749587e-02 -4.90831196e-01 -2.74772551e-02
8.93976331e-01 7.00180888e-01 -1.91667423e-01 -8.35446358e-01
-2.73626924e-01 6.01383865e-01 -1.71125736e-02 -1.63112730e-01
1.30078959e+00 -7.96605229e-01 -5.47574759e-02 7.08937040e-03
1.04208302e+00 -2.57477075e-01 -1.65162051e+00 -2.88644105e-01
-6.85247242e-01 -8.84543717e-01 5.06648958e-01 -9.03406560e-01
-1.17297971e+00 1.13751888e+00 1.16371059e+00 -4.98716906e-02
1.75684857e+00 -4.38374668e-01 7.13987768e-01 1.12189256e-01
1.33462399e-01 -1.40359187e+00 3.99160922e-01 2.85561651e-01
4.65491354e-01 -1.43786478e+00 1.21537428e-02 -4.94477600e-01
-6.46731377e-01 1.04916954e+00 9.67076600e-01 8.60132352e-02
2.00506181e-01 1.07651636e-01 -1.15881644e-01 2.16494039e-01
-4.68503565e-01 -8.15929770e-01 3.14706564e-01 9.68165994e-01
2.88034528e-01 -3.00103188e-01 -1.60081714e-01 4.09825325e-01
3.88405055e-01 -8.91772136e-02 3.34114969e-01 5.21411657e-01
-8.77730608e-01 -6.17687345e-01 -7.61207521e-01 1.80125460e-01
-2.01972634e-01 -2.75664687e-01 -1.03771687e-01 3.37062925e-01
4.77353185e-01 1.32280946e+00 -1.33391649e-01 -4.77095276e-01
1.40751645e-01 -2.84616023e-01 5.32667577e-01 -1.25044852e-01
-3.13242823e-01 3.14995646e-01 -2.55760312e-01 -4.14354235e-01
-7.42198229e-01 -6.99597597e-01 -6.93568707e-01 -1.11303143e-01
-4.75824058e-01 -1.97469324e-01 6.97148085e-01 7.72392154e-01
1.06793329e-01 7.83720613e-01 8.36087823e-01 -1.08023942e+00
-3.03835440e-02 -8.84421706e-01 -5.56434929e-01 3.96283060e-01
1.74776658e-01 -7.88452268e-01 -1.01400584e-01 3.23414117e-01] | [10.0680570602417, -1.5439214706420898] |
16ef7822-1386-4484-bc4b-3a7617b769eb | fast-likelihood-based-change-point-detection | 2301.08892 | null | https://arxiv.org/abs/2301.08892v1 | https://arxiv.org/pdf/2301.08892v1.pdf | Fast likelihood-based change point detection | Change point detection plays a fundamental role in many real-world applications, where the goal is to analyze and monitor the behaviour of a data stream. In this paper, we study change detection in binary streams. To this end, we use a likelihood ratio between two models as a measure for indicating change. The first model is a single bernoulli variable while the second model divides the stored data in two segments, and models each segment with its own bernoulli variable. Finding the optimal split can be done in $O(n)$ time, where $n$ is the number of entries since the last change point. This is too expensive for large $n$. To combat this we propose an approximation scheme that yields $(1 - \epsilon)$ approximation in $O(\epsilon^{-1} \log^2 n)$ time. The speed-up consists of several steps: First we reduce the number of possible candidates by adopting a known result from segmentation problems. We then show that for fixed bernoulli parameters we can find the optimal change point in logarithmic time. Finally, we show how to construct a candidate list of size $O(\epsilon^{-1} \log n)$ for model parameters. We demonstrate empirically the approximation quality and the running time of our algorithm, showing that we can gain a significant speed-up with a minimal average loss in optimality. | ['Nikolaj Tatti'] | 2023-01-21 | null | null | null | null | ['change-detection'] | ['computer-vision'] | [ 2.49355167e-01 -2.62729049e-01 -6.00327738e-02 -7.26574212e-02
-6.57591701e-01 -7.52793789e-01 -9.50647071e-02 9.58006680e-01
-9.29999530e-01 4.68708664e-01 -5.96421182e-01 -5.42875707e-01
-1.38219342e-01 -1.13461399e+00 -8.26388776e-01 -5.87076664e-01
-5.83399773e-01 6.04809403e-01 8.42414737e-01 -6.38026297e-02
4.07727569e-01 5.32299101e-01 -1.22237706e+00 -3.52609217e-01
5.01810670e-01 1.44713557e+00 -1.77632838e-01 1.20728493e+00
-2.21080497e-01 2.21264631e-01 -6.81938887e-01 -3.65707248e-01
5.15050530e-01 -7.71105528e-01 -7.13377118e-01 6.09492995e-02
-1.56894058e-01 2.23861653e-02 -1.37369961e-01 1.19686413e+00
2.47918859e-01 8.58848169e-02 3.43683243e-01 -1.39156413e+00
4.86306727e-01 7.49021411e-01 -8.92425239e-01 5.60808778e-01
9.82147604e-02 -9.16267484e-02 7.53261983e-01 -7.74936453e-02
2.64396429e-01 8.84735107e-01 5.95110595e-01 -2.09964905e-02
-1.28267980e+00 -8.33770096e-01 1.47308379e-01 8.39820281e-02
-1.36315978e+00 -1.37953982e-01 4.35670584e-01 -2.35291690e-01
4.60390598e-01 5.54745913e-01 6.35622084e-01 -8.33909214e-02
-4.08737808e-02 5.09807467e-01 6.94271088e-01 -5.99562466e-01
5.03708303e-01 -1.42527819e-01 6.47310987e-02 6.90237999e-01
4.94184226e-01 -3.38910162e-01 -2.41043612e-01 -3.32256585e-01
6.26009345e-01 1.05407976e-01 -3.88882548e-01 -1.22242514e-02
-7.89484203e-01 8.28735709e-01 1.70903027e-01 2.62839913e-01
-5.63538224e-02 6.86943412e-01 2.58802325e-01 4.96060967e-01
2.88016170e-01 1.52353093e-01 -4.80124384e-01 -6.47458911e-01
-1.02109075e+00 2.90315729e-02 1.00370896e+00 9.09247398e-01
6.70818985e-01 -2.50747412e-01 3.33305478e-01 6.82727158e-01
7.14264959e-02 7.68813729e-01 2.72954613e-01 -1.01619482e+00
6.11225188e-01 2.96778858e-01 1.67323783e-01 -1.02202332e+00
-2.00032949e-01 -1.82514310e-01 -7.66911387e-01 1.65930673e-01
7.95949101e-01 -2.52045214e-01 -5.06994963e-01 1.78064764e+00
2.36041695e-01 4.16742936e-02 -4.50055301e-01 2.53294379e-01
-3.73447567e-01 8.29438150e-01 -3.92690897e-01 -8.55397701e-01
1.37364733e+00 -4.88844931e-01 -3.34675670e-01 -2.43487582e-02
4.16328520e-01 -6.95002437e-01 6.83301687e-01 5.75787723e-01
-1.36235666e+00 -7.04999864e-02 -9.26720858e-01 6.31244957e-01
-3.39104561e-03 -5.02367437e-01 2.79397339e-01 7.86543131e-01
-7.96085477e-01 5.87341070e-01 -1.16020858e+00 -7.67086633e-03
3.05876046e-01 4.82525617e-01 1.51663437e-01 5.09516224e-02
-5.85513830e-01 -1.95328170e-03 2.08911777e-01 -1.47065580e-01
-4.84634131e-01 -4.10475880e-01 -6.46105409e-01 2.82066137e-01
7.41756022e-01 -1.86154500e-01 1.28484094e+00 -6.30783021e-01
-1.05220997e+00 4.05716151e-01 -6.57572687e-01 -7.04413712e-01
6.86349690e-01 1.44682959e-01 4.80745770e-02 3.97722751e-01
-3.44617255e-02 4.92292941e-02 5.83713293e-01 -1.04618490e+00
-1.16812718e+00 -4.27126914e-01 -5.01225740e-02 -1.76974520e-01
-1.41458854e-01 2.27944523e-01 -7.79980838e-01 -4.71008927e-01
5.15389860e-01 -9.81097460e-01 -5.43374240e-01 1.14703737e-01
4.41705063e-02 -5.01004085e-02 2.85337001e-01 -3.43223125e-01
1.42947543e+00 -2.22146678e+00 -4.62816715e-01 5.91298401e-01
1.97600037e-01 2.30768323e-01 3.72344911e-01 3.06450367e-01
1.84917510e-01 4.25664246e-01 -5.73768795e-01 -3.58675361e-01
-2.26364657e-01 1.08858429e-01 -2.12966517e-01 5.42440236e-01
-2.25974247e-01 3.04048538e-01 -9.33241069e-01 -3.16279620e-01
-1.60541922e-01 -8.98725018e-02 -5.20283163e-01 -3.84623967e-02
-2.53478467e-01 -1.87898964e-01 -1.31040767e-01 2.68891186e-01
8.89651716e-01 -2.40965292e-01 2.80697465e-01 2.48485506e-01
-1.34353667e-01 2.02892497e-01 -1.82699156e+00 9.43609834e-01
-4.24675852e-01 6.41356289e-01 2.72341996e-01 -1.07084346e+00
7.66087651e-01 7.15387017e-02 7.99175560e-01 -4.69618469e-01
2.59687483e-01 4.54235137e-01 -1.87030882e-01 -2.11090028e-01
4.03746873e-01 -5.38891017e-01 -2.27062076e-01 7.40502179e-01
-6.01922989e-01 5.78500554e-02 4.41286474e-01 3.72552536e-02
1.61752534e+00 -5.29693961e-01 2.63741493e-01 2.34664097e-01
1.91332817e-01 -1.17688335e-01 7.90000021e-01 9.10862565e-01
-2.73121774e-01 2.81067461e-01 1.24124432e+00 -4.55832668e-02
-6.53477728e-01 -7.79374838e-01 6.49017245e-02 7.63471544e-01
5.02591431e-01 -4.20885384e-01 -6.69668555e-01 -4.98191059e-01
-1.18333153e-01 5.94668269e-01 -4.57190484e-01 6.33816570e-02
-9.45196867e-01 -8.19568038e-01 5.26214063e-01 5.04379988e-01
4.29955512e-01 -9.15886283e-01 -1.20021689e+00 3.98146123e-01
-6.19968548e-02 -9.22717452e-01 -4.80409890e-01 4.90170807e-01
-1.07883084e+00 -1.03444386e+00 -3.67340922e-01 -5.24668634e-01
7.18556702e-01 3.66969764e-01 6.93956852e-01 6.29529580e-02
-1.41133070e-02 2.80260146e-01 -4.88260448e-01 -4.72188443e-01
-3.79910082e-01 -1.51097715e-01 -3.13612163e-01 -1.09475888e-01
3.29597563e-01 -4.98772532e-01 -5.77042043e-01 3.18315655e-01
-1.24343550e+00 -5.40752172e-01 2.41384968e-01 4.51816976e-01
8.30826163e-01 7.25770831e-01 1.47733748e-01 -7.85787284e-01
4.16918010e-01 -4.89299268e-01 -1.13229740e+00 -2.16700211e-02
-7.10408866e-01 7.80592933e-02 9.21948910e-01 -3.31025034e-01
-3.20562363e-01 9.05396715e-02 -9.48046073e-02 -3.16447258e-01
-5.27833886e-02 4.56287742e-01 1.96243194e-03 1.09707758e-01
1.95540115e-01 3.44494373e-01 6.98076114e-02 -4.85503405e-01
5.06511657e-03 5.48150003e-01 6.55877531e-01 -3.38216543e-01
4.86671150e-01 6.60610735e-01 3.06696802e-01 -8.44118416e-01
-2.71644205e-01 -6.46685243e-01 -1.79501325e-01 7.67669873e-04
2.43437618e-01 -3.22422713e-01 -1.21146083e+00 3.26938301e-01
-8.14995289e-01 -3.82614106e-01 -3.08465898e-01 2.25776270e-01
-5.14097333e-01 6.16584957e-01 -4.29690391e-01 -1.20016444e+00
-1.58003896e-01 -9.30295765e-01 5.57187438e-01 1.35121778e-01
-7.63400197e-02 -6.71843767e-01 -2.01593246e-02 -1.62110493e-01
2.92540878e-01 2.71239609e-01 8.07317615e-01 -7.24806726e-01
-6.87000871e-01 -5.24941206e-01 -1.15611926e-01 8.10494050e-02
4.90125641e-02 -1.34367213e-01 -1.61053792e-01 -3.88517737e-01
2.71608770e-01 4.59922194e-01 8.27254593e-01 2.56276965e-01
1.27932417e+00 -5.37066519e-01 -3.63350689e-01 4.88102525e-01
1.66483140e+00 6.92743301e-01 5.13233840e-01 2.55418926e-01
2.52723128e-01 1.50995314e-01 7.28709280e-01 9.69769537e-01
1.74255356e-01 4.98101056e-01 5.02432108e-01 7.39194080e-02
2.90479690e-01 2.75533553e-02 1.97982371e-01 4.89427865e-01
3.68727893e-01 -5.68316936e-01 -9.16551411e-01 8.71588349e-01
-1.71047676e+00 -8.02152336e-01 -2.60173321e-01 2.80333447e+00
9.05537426e-01 6.95084333e-01 4.76340741e-01 8.97246897e-01
7.05168426e-01 -1.31619200e-01 -3.83809984e-01 -5.38420737e-01
2.14377299e-01 6.46981001e-01 9.65934575e-01 5.94965219e-01
-8.07049692e-01 5.55993676e-01 5.23584747e+00 8.93695652e-01
-1.20517933e+00 -5.95954061e-02 5.47946811e-01 -2.87658572e-01
-1.20719120e-01 2.82721311e-01 -6.33933961e-01 1.00516748e+00
1.12176919e+00 -3.25479060e-01 2.85170406e-01 5.75932264e-01
2.11687550e-01 -7.49919534e-01 -1.03119135e+00 1.04003298e+00
1.25447838e-02 -9.24751461e-01 -3.83636534e-01 3.67073089e-01
3.07690114e-01 -3.57501745e-01 -3.08081001e-01 -2.58195162e-01
1.80912882e-01 -6.30374610e-01 7.73165643e-01 4.00553197e-02
5.79160035e-01 -1.03358078e+00 7.11592019e-01 4.24590409e-01
-1.40145814e+00 -2.42450222e-01 -1.93844363e-01 7.52201453e-02
5.47374487e-01 8.58759820e-01 -9.31319177e-01 3.16741318e-01
7.13048875e-01 -3.91591154e-02 -7.34310597e-02 1.57485974e+00
1.90574855e-01 8.55069697e-01 -1.13466215e+00 -2.30258197e-01
9.79023948e-02 -1.84664801e-01 6.81730330e-01 1.27829528e+00
4.53887820e-01 2.89710134e-01 1.12032734e-01 2.28827566e-01
-2.08990246e-01 7.83984959e-02 -6.92938343e-02 2.31887534e-01
7.16011047e-01 6.66014850e-01 -1.34808838e+00 -4.17130411e-01
-9.49956998e-02 9.36910689e-01 -2.01422423e-01 -7.26288324e-03
-8.86081994e-01 -1.02461803e+00 3.32405478e-01 4.48488116e-01
9.72157598e-01 -5.40108740e-01 -4.65147346e-01 -6.45524800e-01
4.31392103e-01 -6.64397240e-01 5.73432922e-01 1.09651983e-01
-6.68650270e-01 4.84751195e-01 1.04702085e-01 -9.87873018e-01
-2.19442964e-01 -2.31470808e-01 -2.73276091e-01 5.70139825e-01
-1.23532367e+00 -1.26606211e-01 -2.35489890e-01 2.82553166e-01
3.51354271e-01 4.69746560e-01 3.99390250e-01 3.19898665e-01
-4.88142908e-01 7.84889221e-01 2.89063185e-01 8.25617388e-02
4.12273884e-01 -1.11897743e+00 3.04476887e-01 1.16494679e+00
6.78649917e-02 2.83270836e-01 9.29441690e-01 -5.23183763e-01
-9.84147131e-01 -7.03882635e-01 1.00563169e+00 2.38873839e-01
4.91236120e-01 -8.33260342e-02 -8.11348736e-01 3.82918507e-01
-2.81547368e-01 -6.24398282e-03 5.93608081e-01 -2.36408114e-01
-7.35480785e-02 -5.35538018e-01 -1.23749447e+00 4.26437557e-01
6.78192794e-01 -5.05556464e-02 -7.16458634e-02 -1.30444124e-01
5.04242599e-01 -2.19948456e-01 -8.12317014e-01 2.31998399e-01
3.76882732e-01 -9.95843887e-01 3.76922667e-01 -2.24261358e-01
-4.26331609e-02 -5.58105886e-01 -2.24073157e-01 -8.36677551e-01
2.00444654e-01 -1.10147953e+00 -1.93932787e-01 1.00651252e+00
4.90444183e-01 -7.45975792e-01 7.87767529e-01 6.59465849e-01
3.31149906e-01 -1.00081015e+00 -1.17690229e+00 -8.20985138e-01
-8.26959312e-02 -8.64093184e-01 5.12917936e-01 4.43847984e-01
4.12559956e-02 -3.31243314e-02 -3.47120548e-03 1.16873927e-01
4.05281633e-01 1.88153595e-01 7.42979467e-01 -1.03458035e+00
-7.37367630e-01 -6.41201377e-01 -4.60548490e-01 -1.54408872e+00
-5.70351601e-01 -2.92827934e-01 2.93226928e-01 -1.12442648e+00
1.92837507e-01 -1.10426152e+00 -3.02503735e-01 4.30534899e-01
-1.17156319e-01 5.52063175e-02 3.62186909e-01 1.35511816e-01
-6.94702387e-01 1.12210520e-01 5.36458552e-01 2.09144264e-01
-5.26103616e-01 5.14859021e-01 -5.50957620e-01 6.99231803e-01
8.05158913e-01 -8.77403796e-01 -1.91655770e-01 -3.44795674e-01
4.85845327e-01 4.61807668e-01 5.70738269e-03 -9.99046862e-01
2.93489516e-01 -1.34461209e-01 -1.47201791e-01 -6.20610952e-01
1.86595008e-01 -7.75777936e-01 5.59746884e-02 6.79652154e-01
-2.34302714e-01 3.61919641e-01 2.08119959e-01 8.30311239e-01
-1.13580674e-01 -8.29752862e-01 9.98153210e-01 8.16414505e-02
-5.57516627e-02 3.59070003e-01 -4.24470484e-01 -8.22632853e-03
1.26189387e+00 -2.45654464e-01 -1.02229137e-02 -5.41360736e-01
-5.39412737e-01 3.23462456e-01 5.22125840e-01 -9.58388895e-02
3.69163245e-01 -9.28318202e-01 -3.63397419e-01 1.51630849e-01
-3.90991181e-01 3.03530753e-01 -5.32497168e-02 9.23812151e-01
-8.97976458e-01 1.32120371e-01 3.94015580e-01 -6.87240243e-01
-1.44440079e+00 6.00840330e-01 2.07717419e-01 -4.57038581e-01
-2.52125204e-01 8.99632633e-01 -4.05347377e-01 5.44059336e-01
3.97936165e-01 -4.18134868e-01 3.75597358e-01 9.45063233e-02
6.14755929e-01 7.78047144e-01 3.79118696e-02 -1.69503555e-01
-4.00701851e-01 4.43845332e-01 -2.39494052e-02 -7.24186659e-01
1.28933358e+00 -2.96875834e-01 -2.51155853e-01 5.15655875e-01
1.39248836e+00 2.34575823e-01 -1.18227422e+00 -2.03105465e-01
1.04365617e-01 -6.03813410e-01 -3.62906128e-01 -2.29593769e-01
-1.00481844e+00 7.57481992e-01 7.14044213e-01 8.52336884e-01
1.56788778e+00 2.51591559e-02 1.11766672e+00 1.27294555e-01
5.17253935e-01 -8.68530095e-01 -1.37079805e-01 3.92348677e-01
2.86322892e-01 -7.39478767e-01 -1.55967191e-01 -5.58672249e-01
-2.73152739e-01 1.11824775e+00 1.95253089e-01 -7.49729872e-02
9.55795527e-01 2.78437883e-01 -2.59941667e-01 8.89869705e-02
-5.68796217e-01 -1.66051164e-01 -4.78671491e-01 -8.99464078e-03
3.67767513e-02 1.09118506e-01 -7.12373614e-01 5.36728144e-01
-2.41742149e-01 -3.64016034e-02 7.04645932e-01 1.35952580e+00
-8.15698624e-01 -1.42114484e+00 -4.51728582e-01 5.42011499e-01
-8.51120472e-01 1.30677447e-01 2.57157180e-02 4.33389246e-01
1.07099302e-01 1.21357322e+00 4.16421473e-01 -1.15454994e-01
3.36757421e-01 2.24126466e-02 3.74235660e-01 -1.84143931e-01
-3.97686005e-01 4.11013007e-01 -2.42767975e-01 -4.19009000e-01
-4.19794358e-02 -7.22934604e-01 -1.46485007e+00 -5.03638685e-01
-2.33255088e-01 3.77225041e-01 7.81021118e-01 8.07226241e-01
2.37646475e-01 2.32303768e-01 1.03440809e+00 -2.51949698e-01
-4.37873393e-01 -6.77170932e-01 -5.39858818e-01 4.42527272e-02
3.21993232e-01 -2.66516924e-01 -5.34441531e-01 1.73164472e-01] | [6.711023807525635, 4.784191608428955] |
7b0b7286-b766-4091-aad9-9683fd7e8deb | attributes-and-categories-for-generic | null | null | http://openaccess.thecvf.com/content_cvpr_2015/html/Tao_Attributes_and_Categories_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Tao_Attributes_and_Categories_2015_CVPR_paper.pdf | Attributes and Categories for Generic Instance Search From One Example | This paper aims for generic instance search from one example where the instance can be an arbitrary 3D object like shoes, not just near-planar and one-sided instances like buildings and logos. Firstly, we evaluate state-of-the-art instance search methods on this problem. We observe that what works for buildings loses its generality on shoes. Secondly, we propose to use automatically learned category-specific attributes to address the large appearance variations present in generic instance search. On the problem of searching among instances from the same category as the query, the category-specific attributes outperform existing approaches by a large margin. On a shoe dataset containing 6624 shoe images recorded from all viewing angles, we improve the performance from 36.73 to 56.56 using category-specific attributes. Thirdly, we extend our methods to search objects without restricting to the specifically known category. We show the combination of category-level information and the category-specific attributes is superior to combining category-level information with low-level features such as Fisher vector. | ['Arnold W. M. Smeulders', 'Shih-Fu Chang', 'Ran Tao'] | 2015-06-01 | null | null | null | cvpr-2015-6 | ['instance-search'] | ['computer-vision'] | [ 1.86313421e-01 -1.84057772e-01 -3.04345608e-01 -4.56904143e-01
-8.98102939e-01 -6.64567649e-01 4.85295057e-01 -1.79251414e-02
-2.34953269e-01 4.64671850e-01 -1.30813316e-01 1.05617367e-01
-8.75861466e-01 -6.26663148e-01 -7.40900695e-01 -5.35423815e-01
-1.46215498e-01 1.02443063e+00 5.47486007e-01 -2.97844470e-01
4.37216818e-01 6.53192461e-01 -2.15632296e+00 1.99349061e-01
6.59880340e-01 1.22485721e+00 3.36740255e-01 3.14003289e-01
-1.37949595e-02 -4.48051207e-02 -6.21334076e-01 -4.44719642e-01
5.30828238e-01 8.14796761e-02 -7.74398208e-01 4.52725857e-01
1.23673177e+00 2.55818330e-02 -1.09336279e-01 7.83734620e-01
5.53022623e-01 2.71606386e-01 8.60144436e-01 -1.28310442e+00
-6.08435929e-01 -1.24595374e-01 -4.28238779e-01 1.74699798e-01
4.28571880e-01 -9.05244052e-02 1.39111161e+00 -1.07750082e+00
7.00430810e-01 1.24972475e+00 5.53685665e-01 2.98972368e-01
-1.24087501e+00 -3.74143928e-01 3.27386320e-01 4.97305393e-01
-1.62788248e+00 -1.88729748e-01 7.36036897e-01 -4.17034209e-01
8.86467576e-01 5.07045686e-01 4.63101864e-01 6.91331923e-01
-1.96524307e-01 8.20119917e-01 1.20368135e+00 -2.35569939e-01
3.48062143e-02 8.82172137e-02 2.52279907e-01 6.44472241e-01
3.88299704e-01 1.41505733e-01 -3.72331798e-01 -3.83933455e-01
6.17798209e-01 1.63742051e-01 -8.76738131e-03 -8.35761070e-01
-1.38342226e+00 7.24502683e-01 5.26742876e-01 3.20382528e-02
-4.05740917e-01 9.56540927e-02 7.10793734e-02 1.12678573e-01
3.26752067e-01 7.92046368e-01 -7.82531142e-01 1.56676471e-01
-9.39895988e-01 4.59238052e-01 8.71173203e-01 1.31563234e+00
8.19764018e-01 -4.63425517e-01 -1.73730180e-01 1.28557169e+00
3.42009068e-02 7.62391269e-01 -4.08441946e-02 -9.47953403e-01
2.33102828e-01 5.06724238e-01 2.02908725e-01 -9.04558063e-01
-3.78219545e-01 -7.33806372e-01 -3.05467367e-01 1.23115564e-02
4.30987954e-01 4.82508093e-01 -1.30617440e+00 1.45021462e+00
4.92423534e-01 -1.06579185e-01 -1.83397755e-01 1.15714014e+00
1.04621041e+00 2.56349534e-01 -1.71761855e-01 1.11930273e-01
1.83829653e+00 -1.08472514e+00 -2.87766993e-01 -4.26172376e-01
3.28139096e-01 -8.70535076e-01 1.05138493e+00 5.13727069e-01
-8.42456162e-01 -4.94276077e-01 -8.57884467e-01 8.07712898e-02
-6.63657546e-01 2.29317293e-01 6.65680945e-01 5.50527215e-01
-7.73007274e-01 3.07731420e-01 -1.62573844e-01 -5.72835982e-01
1.63289875e-01 7.46275544e-01 -5.45405209e-01 -3.64788055e-01
-1.07491016e+00 7.13086963e-01 2.64764220e-01 -3.15835744e-01
-6.42879784e-01 -5.31950295e-01 -6.86464608e-01 -2.12793555e-02
1.00048053e+00 -8.16194713e-01 9.64236140e-01 -4.80372310e-01
-8.80292356e-01 1.04137242e+00 -2.54389167e-01 1.70812365e-02
1.85624093e-01 -2.55804300e-01 -5.44900596e-01 1.35240749e-01
3.48278552e-01 4.80105549e-01 7.90815890e-01 -1.59875464e+00
-7.71111250e-01 -5.12441874e-01 3.72541815e-01 4.10887152e-01
-3.48859318e-02 -2.74966031e-01 -1.03961182e+00 -5.79696298e-01
5.55453479e-01 -1.27638447e+00 -2.02548444e-01 4.24616635e-02
-4.05387521e-01 -4.14363921e-01 6.18162632e-01 -2.84328282e-01
1.14488888e+00 -2.09513116e+00 7.78937712e-02 4.95496958e-01
-5.21464571e-02 -1.25067815e-01 -1.16895095e-01 3.24146122e-01
1.07880130e-01 7.19711483e-02 -6.40523992e-03 -4.40281667e-02
2.27610901e-01 5.51660836e-01 8.74461234e-02 3.43202055e-01
1.09381318e-01 6.90004587e-01 -9.13723946e-01 -5.93610108e-01
5.49766608e-02 1.67086616e-01 -6.52340591e-01 -2.25836128e-01
-2.79468983e-01 -6.94063455e-02 -6.95973217e-01 1.09422779e+00
6.61559939e-01 -5.91569781e-01 -5.89790270e-02 -3.09293330e-01
3.86702210e-01 1.55830458e-01 -1.50617754e+00 1.55137062e+00
-4.59876865e-01 2.89293945e-01 -7.74030760e-02 -8.94182146e-01
9.56069708e-01 -5.32214008e-02 5.10541856e-01 -6.35727882e-01
-6.11739635e-01 4.70777214e-01 -1.00829639e-01 -4.72731531e-01
5.63147187e-01 9.33564082e-02 -3.92920077e-01 8.05008039e-02
6.69735596e-02 -2.73512155e-01 1.51777416e-01 4.81804796e-02
9.49425340e-01 2.19837166e-02 1.88995510e-01 -3.93211484e-01
3.63812387e-01 1.80513307e-01 1.26030937e-01 1.12743890e+00
2.53518403e-01 8.09495628e-01 -6.09552674e-02 -5.85450709e-01
-9.26500022e-01 -1.08498406e+00 -5.88095188e-01 1.22711265e+00
6.31817698e-01 -4.39239651e-01 -3.13415706e-01 -7.09765971e-01
5.03559172e-01 4.53579187e-01 -5.69221854e-01 -1.32633289e-02
-2.85057276e-01 -6.71636403e-01 -1.55226123e-02 3.86033207e-01
4.44949597e-01 -7.02708960e-01 -4.91701365e-01 2.68256273e-02
-1.74111202e-01 -1.00329053e+00 -6.03494108e-01 2.63813406e-01
-7.38299072e-01 -1.08762908e+00 -6.52235627e-01 -6.47652686e-01
6.19989455e-01 3.17833245e-01 1.53485799e+00 2.18158931e-01
-5.82779348e-01 7.65489399e-01 -3.38344216e-01 -2.72311270e-01
3.77895534e-01 9.87458676e-02 1.75599217e-01 -9.23305936e-03
5.51620185e-01 -2.51061916e-01 -7.43599236e-01 1.00368798e+00
-6.10557556e-01 -4.27723289e-01 7.32942760e-01 9.51408982e-01
9.57610726e-01 1.45141765e-01 2.79057264e-01 -7.89424896e-01
2.11733505e-01 -4.73849446e-01 -3.30157697e-01 3.78990322e-01
-8.15151215e-01 1.00524470e-01 3.31458189e-02 -7.79188573e-01
-4.71056819e-01 2.20961154e-01 1.43585712e-01 -4.72771913e-01
-2.91916609e-01 2.88308531e-01 -1.09026141e-01 -8.23754370e-02
6.86795771e-01 2.58390099e-01 -3.77961785e-01 -6.82545125e-01
1.58000395e-01 4.81822848e-01 4.64849830e-01 -8.33489537e-01
8.28480840e-01 3.86620998e-01 2.51025379e-01 -9.44845021e-01
-1.00580585e+00 -9.36485887e-01 -5.37398338e-01 -1.01519767e-02
7.78566182e-01 -7.56071985e-01 -6.75435185e-01 6.45968542e-02
-7.50768960e-01 -1.08931959e-01 -2.60363013e-01 4.55348670e-01
-6.51259124e-01 1.31762385e-01 -3.97457890e-02 -7.58232772e-01
8.27384964e-02 -1.09570813e+00 1.83559144e+00 5.93373403e-02
8.86306763e-02 -6.56489074e-01 -3.47173721e-01 5.76507449e-01
1.46137327e-01 1.11491472e-01 8.51671457e-01 -1.09597552e+00
-9.02154088e-01 -3.45149636e-01 -1.34788424e-01 -8.46166983e-02
1.71318367e-01 -2.37074658e-01 -7.75024891e-01 -2.39936292e-01
-3.47706407e-01 -6.75460622e-02 8.75537932e-01 3.42914462e-01
1.25651872e+00 -2.13579461e-01 -5.83290815e-01 5.76363981e-01
1.51710236e+00 1.22350380e-01 5.04354417e-01 4.50028569e-01
5.05673885e-01 5.54172158e-01 1.21773851e+00 4.21693116e-01
4.59067374e-01 1.27191496e+00 7.37025797e-01 -1.12587586e-01
-8.14167932e-02 2.49763839e-02 -3.42656255e-01 4.84380424e-02
-3.96164358e-01 -2.43071139e-01 -8.63974094e-01 7.78166592e-01
-1.81065321e+00 -1.02491605e+00 -5.37210666e-02 2.42496324e+00
5.19405782e-01 1.56314135e-01 3.71203154e-01 -4.72741611e-02
7.44877934e-01 -9.70300287e-02 -5.08368850e-01 -7.13556409e-02
-5.31250387e-02 -3.27011719e-02 6.58119619e-01 2.79573560e-01
-1.38822770e+00 7.17656553e-01 6.54335451e+00 1.14204800e+00
-6.58080757e-01 -1.20963149e-01 2.23431706e-01 -2.55929291e-01
-2.32905179e-01 -1.45242065e-01 -1.05872655e+00 3.73720378e-01
3.11061621e-01 2.17226893e-01 4.10298973e-01 1.24194825e+00
-5.01481295e-01 -8.07065591e-02 -1.24221826e+00 1.13256812e+00
2.92995661e-01 -1.00307560e+00 -1.43060744e-01 4.97248322e-01
6.23595178e-01 -2.22192168e-01 2.44998530e-01 4.58922714e-01
1.51808798e-01 -1.04276812e+00 6.17163658e-01 6.62347913e-01
8.43057513e-01 -5.96721351e-01 5.50729096e-01 2.08959252e-01
-1.22492576e+00 -1.01533592e-01 -4.75749671e-01 3.51703018e-01
-6.65358976e-02 3.75756055e-01 -8.54315579e-01 5.86653054e-01
1.01024354e+00 4.37686175e-01 -7.12281644e-01 1.33477581e+00
1.35008022e-01 2.57043332e-01 -8.98583770e-01 -2.75763087e-02
3.02105963e-01 -8.01038221e-02 6.55294895e-01 1.07715237e+00
2.83918440e-01 2.23154321e-01 4.51299816e-01 4.91298109e-01
2.38888189e-01 3.70167233e-02 -6.79086506e-01 3.77549648e-01
4.17731315e-01 1.07785296e+00 -6.78324401e-01 -3.17389816e-01
-4.11655843e-01 1.07306445e+00 3.80753130e-02 3.63300949e-01
-7.50143647e-01 -3.34932506e-01 7.01966166e-01 4.53416646e-01
7.80824184e-01 -9.46387872e-02 1.51289091e-01 -8.10926914e-01
3.73690665e-01 -7.67284036e-01 5.37982941e-01 -8.20944011e-01
-1.55485618e+00 4.87688959e-01 5.48419833e-01 -1.30151582e+00
-2.54185379e-01 -8.00788999e-01 -4.99054082e-02 6.96732402e-01
-1.47119498e+00 -1.34655392e+00 -4.02054727e-01 6.09798074e-01
8.13398480e-01 -1.07452057e-01 8.48097026e-01 5.05120277e-01
1.09811403e-01 5.85288346e-01 1.00396797e-01 -1.02010943e-01
8.53655696e-01 -1.34127259e+00 1.00632787e-01 1.40757531e-01
5.29593647e-01 8.09309900e-01 7.42331028e-01 -5.48010826e-01
-1.42555082e+00 -6.63349569e-01 9.25503075e-01 -9.09006894e-01
3.81713122e-01 -3.29554498e-01 -7.14847624e-01 3.67571384e-01
-4.51436311e-01 5.38451374e-01 4.67454076e-01 5.99822521e-01
-5.92086196e-01 -1.70931205e-01 -1.21564603e+00 3.43340933e-01
1.58669090e+00 -5.16377091e-01 -6.34999633e-01 6.84362471e-01
4.60971862e-01 -4.05274004e-01 -1.15700173e+00 6.69895828e-01
8.59533608e-01 -7.94966221e-01 1.62353611e+00 -6.18591607e-01
1.95935801e-01 -4.53496605e-01 -6.57041430e-01 -9.47687685e-01
-5.11439502e-01 -1.97837368e-01 -9.22800303e-02 7.23989367e-01
4.66313213e-01 -4.94935423e-01 8.04319441e-01 5.67773581e-01
-1.65581346e-01 -1.02557468e+00 -9.39267337e-01 -1.31503868e+00
-2.84220248e-01 -7.30302855e-02 7.41940141e-01 6.51137233e-01
-6.23268604e-01 3.02162409e-01 -3.77560824e-01 4.43410933e-01
7.65291452e-01 6.55950010e-01 7.73749709e-01 -1.65065694e+00
-4.89477456e-01 -2.25683123e-01 -7.51622558e-01 -1.38622499e+00
-1.29653811e-01 -6.61313593e-01 3.77646029e-06 -1.58696759e+00
3.91659200e-01 -1.01295435e+00 -3.32858473e-01 5.37647307e-01
-2.25232437e-01 5.38161159e-01 2.47604340e-01 3.87083262e-01
-8.59153032e-01 1.23255745e-01 1.40111887e+00 -2.50855476e-01
3.37781496e-02 3.35052818e-01 -5.11716723e-01 4.48027521e-01
4.68132198e-01 -4.39483136e-01 -3.75560105e-01 -1.98769808e-01
1.54844224e-01 -3.48568447e-02 6.70874059e-01 -9.84008491e-01
1.16086096e-01 -3.03410679e-01 3.83622587e-01 -8.80623519e-01
8.85360420e-01 -1.25795949e+00 4.27611589e-01 1.96811184e-01
-1.68437243e-01 -2.77052283e-01 4.07299511e-02 8.06174219e-01
-1.29242525e-01 -2.47870788e-01 4.27877665e-01 -2.29463756e-01
-1.17558551e+00 3.80300879e-01 1.73495665e-01 8.70262757e-02
9.66756701e-01 -7.43257105e-01 -3.48971307e-01 -3.37914735e-01
-1.10451388e+00 2.23911256e-01 6.51121259e-01 5.33402801e-01
5.20235896e-01 -1.37534189e+00 -5.77135324e-01 1.58752561e-01
6.77033782e-01 -7.99271092e-02 -2.37616524e-02 6.49915576e-01
-2.67339256e-02 5.02944648e-01 9.85590816e-02 -9.74769473e-01
-1.49412870e+00 8.80636871e-01 1.61949426e-01 -7.78871924e-02
-2.46505216e-01 8.47387373e-01 4.43317264e-01 -6.46628618e-01
1.54720306e-01 2.77198479e-03 -4.99474518e-02 1.13380685e-01
1.01392847e-02 1.78744227e-01 1.64635494e-01 -7.53264785e-01
-7.84675479e-01 1.24397194e+00 1.24863699e-01 1.59529537e-01
1.23783255e+00 -4.09430936e-02 2.20830053e-01 2.77765810e-01
1.08490968e+00 5.92154451e-02 -1.01287961e+00 -3.81405205e-01
5.11671677e-02 -8.09481502e-01 -2.20497251e-01 -1.08637273e+00
-1.00047958e+00 5.68317771e-01 7.18568504e-01 1.80341095e-01
1.08248270e+00 6.71085000e-01 3.27323109e-01 6.49252355e-01
7.90138006e-01 -1.02425086e+00 1.75010473e-01 3.90858978e-01
8.42318833e-01 -1.54014587e+00 3.01755518e-01 -6.75144076e-01
-6.35128736e-01 8.53697419e-01 5.56037128e-01 -2.24316269e-01
6.01750016e-01 -1.97902068e-01 -3.65844786e-01 -6.48505449e-01
-5.99278569e-01 -7.17740476e-01 1.02012062e+00 6.49413109e-01
-4.19187583e-02 5.76785952e-02 -1.91063970e-01 4.10602450e-01
-5.62563017e-02 -4.23937768e-01 -7.97531083e-02 8.09202015e-01
-5.02742171e-01 -1.09228253e+00 -4.95415777e-01 9.28459823e-01
-4.05130774e-01 -2.40718827e-01 -2.71709114e-01 1.12111068e+00
2.13947147e-01 7.95981586e-01 1.53204143e-01 -4.26159263e-01
6.51555777e-01 8.05905983e-02 6.22180700e-01 -7.58289397e-01
-4.11652178e-01 1.23270519e-01 3.80295217e-01 -5.52268565e-01
-5.11401653e-01 -7.03679502e-01 -8.01423192e-01 1.42430022e-01
-6.66917503e-01 5.07931001e-02 7.82417715e-01 5.80162227e-01
4.59444165e-01 5.24307974e-02 3.68534744e-01 -8.30177784e-01
-5.77696264e-01 -6.15548134e-01 -7.59132802e-01 7.39075482e-01
4.36609596e-01 -1.31710875e+00 -5.24581373e-01 -7.26880729e-02] | [7.892367362976074, -2.363372325897217] |
c2885a0b-5f12-4b85-9c2a-efcc1df21b14 | classifications-of-skull-fractures-using-ct | 2203.10786 | null | https://arxiv.org/abs/2203.10786v1 | https://arxiv.org/pdf/2203.10786v1.pdf | Classifications of Skull Fractures using CT Scan Images via CNN with Lazy Learning Approach | Classification of skull fracture is a challenging task for both radiologists and researchers. Skull fractures result in broken pieces of bone, which can cut into the brain and cause bleeding and other injury types. So it is vital to detect and classify the fracture very early. In real world, often fractures occur at multiple sites. This makes it harder to detect the fracture type where many fracture types might summarize a skull fracture. Unfortunately, manual detection of skull fracture and the classification process is time-consuming, threatening a patient's life. Because of the emergence of deep learning, this process could be automated. Convolutional Neural Networks (CNNs) are the most widely used deep learning models for image categorization because they deliver high accuracy and outstanding outcomes compared to other models. We propose a new model called SkullNetV1 comprising a novel CNN by taking advantage of CNN for feature extraction and lazy learning approach which acts as a classifier for classification of skull fractures from brain CT images to classify five fracture types. Our suggested model achieved a subset accuracy of 88%, an F1 score of 93%, the Area Under the Curve (AUC) of 0.89 to 0.98, a Hamming score of 92% and a Hamming loss of 0.04 for this seven-class multi-labeled classification. | ['Moqsadur Rahman', 'Tareque Rahman Ornob', 'Md Moniruzzaman Emon'] | 2022-03-21 | null | null | null | null | ['image-categorization'] | ['computer-vision'] | [-1.19863041e-01 -1.38705075e-01 1.98619500e-01 -2.78246492e-01
-5.61868429e-01 -1.74947202e-01 6.08923212e-02 7.43389428e-01
-6.27892852e-01 5.30350506e-01 -1.74854174e-01 -3.73023242e-01
-2.13172287e-01 -1.09469724e+00 -2.62715608e-01 -7.71441519e-01
-2.32578650e-01 5.64250946e-01 5.95828712e-01 -2.11506218e-01
1.66095763e-01 7.13256121e-01 -1.53667223e+00 5.28055072e-01
4.66604203e-01 1.16177106e+00 1.98517926e-02 3.70786190e-01
-3.81558319e-03 8.58186364e-01 -5.84482431e-01 -5.19731462e-01
2.40624979e-01 9.99231562e-02 -8.78932118e-01 -3.05309147e-01
-2.52245426e-01 -2.12855890e-01 -3.68935049e-01 9.32733893e-01
6.94317877e-01 -9.45624411e-02 8.75038087e-01 -1.10550272e+00
-2.30697617e-01 5.53859353e-01 -7.52567172e-01 5.53293049e-01
2.02793837e-01 5.91290705e-02 6.03508234e-01 -8.86678100e-01
2.19419867e-01 7.39314020e-01 9.34034288e-01 5.28277099e-01
-6.36543632e-01 -9.61253583e-01 -7.13876426e-01 4.18991208e-01
-1.63617992e+00 -1.19140603e-01 2.68953979e-01 -6.48876011e-01
6.29687726e-01 1.43547967e-01 8.15492690e-01 8.58164132e-01
9.82320845e-01 2.80055523e-01 1.01694918e+00 -2.22969204e-01
3.21015239e-01 -1.81381285e-01 1.99545771e-01 1.06120455e+00
2.61269093e-01 1.34198770e-01 -2.01006442e-01 -1.07842728e-01
4.61626977e-01 6.04150772e-01 -3.07136118e-01 9.87201631e-02
-8.99137139e-01 1.03686583e+00 7.44538546e-01 1.60912395e-01
-5.93428791e-01 -1.81211531e-02 7.14119971e-01 2.98755467e-01
-3.08845509e-02 -5.63574210e-02 -4.24888358e-03 1.12850398e-01
-8.92402947e-01 1.91470250e-01 5.51616251e-01 4.10030752e-01
7.52615556e-02 -2.31268615e-01 -1.90649219e-02 8.14819217e-01
1.56780809e-01 3.16326439e-01 9.28617239e-01 6.23339787e-03
1.48396894e-01 7.41334319e-01 -4.76102173e-01 -9.12576258e-01
-9.43258464e-01 -6.39472485e-01 -1.22892666e+00 5.78743756e-01
2.61301786e-01 1.27816468e-01 -9.72557545e-01 1.18494439e+00
9.96167064e-02 -1.77287191e-01 -1.78163469e-01 9.23209429e-01
1.12958336e+00 3.09919119e-01 -2.02270687e-01 -5.62253632e-02
1.70683146e+00 -7.02919245e-01 -3.90265912e-01 8.92990828e-02
4.19137269e-01 -6.93940163e-01 7.32324660e-01 5.97518325e-01
-9.22300220e-01 -1.58012748e-01 -1.24827063e+00 2.97219872e-01
-2.67773420e-01 -2.59365559e-01 5.07476628e-01 7.70890772e-01
-8.85878980e-01 5.08934736e-01 -9.45640028e-01 -2.29488343e-01
8.60647917e-01 6.06561840e-01 -5.66979527e-01 -1.67278543e-01
-1.00276577e+00 1.16907942e+00 3.07464778e-01 -7.13985264e-02
-8.54381084e-01 -3.17761779e-01 -5.02091050e-01 6.69353679e-02
-3.46845500e-02 -6.23388588e-01 1.12931967e+00 -4.64869648e-01
-9.26275671e-01 9.12067235e-01 2.84099013e-01 -7.19706357e-01
6.65572524e-01 1.49651706e-01 -3.45713466e-01 1.67726040e-01
1.52741998e-01 1.99701890e-01 7.55607307e-01 -6.88928843e-01
-7.77126312e-01 -5.99408031e-01 -2.04198524e-01 -2.78223962e-01
-8.27256367e-02 1.54596448e-01 2.66060024e-01 -4.27530199e-01
4.41523790e-01 -7.96763539e-01 -1.81656569e-01 -9.10947919e-02
-1.97640941e-01 -2.61610150e-01 5.90698540e-01 -6.01619601e-01
9.97881234e-01 -2.08567381e+00 -2.28372663e-01 3.38778734e-01
6.54336154e-01 2.18410179e-01 3.61169010e-01 2.54813373e-01
-4.19892490e-01 -8.48893821e-02 -3.99942726e-01 9.09826383e-02
-4.55433339e-01 -6.84168711e-02 1.37538165e-01 7.20294476e-01
-1.10179245e-01 6.52349830e-01 -7.29513347e-01 -3.39443594e-01
3.42614651e-02 5.03498137e-01 -3.30828726e-01 -3.13686207e-02
5.17756522e-01 1.06338173e-01 -2.57061392e-01 7.43550122e-01
6.42554700e-01 2.73462702e-02 -4.11469907e-01 -4.27412987e-02
1.99987426e-01 -1.72787920e-01 -1.20453811e+00 1.32146180e+00
-2.75630951e-01 5.39624512e-01 -2.83853501e-01 -9.74064827e-01
1.09553039e+00 5.43005586e-01 5.96170247e-01 -5.37063420e-01
6.84978127e-01 4.75226343e-01 2.90269077e-01 -7.87263513e-01
-2.26515234e-01 -6.84603572e-01 1.77037582e-01 4.80005890e-01
-1.27692312e-01 -7.65201300e-02 -3.99677753e-02 -1.86955437e-01
1.48522925e+00 -5.90779722e-01 5.35204887e-01 -2.48699918e-01
3.87843221e-01 -2.79862553e-01 3.93134385e-01 3.42568487e-01
-3.29372525e-01 8.75974774e-01 -2.49616727e-02 -1.07099605e+00
-6.65609241e-01 -1.02949572e+00 -4.47886795e-01 4.83142287e-01
-7.18190223e-02 -2.74123877e-01 -7.67795563e-01 -4.38078523e-01
-1.18337329e-02 4.05965626e-01 -7.03130484e-01 -5.93686879e-01
-4.19252515e-01 -9.52450752e-01 7.57965922e-01 5.52761674e-01
5.01745582e-01 -9.48800147e-01 -1.27593207e+00 2.70556897e-01
-1.25230864e-01 -6.79913342e-01 -1.36461362e-01 5.64689338e-01
-7.72989869e-01 -1.53931582e+00 -6.61369860e-01 -9.18119192e-01
4.86829758e-01 1.31875917e-01 7.39920199e-01 4.53406781e-01
-7.68610716e-01 -2.35447690e-01 -4.12250787e-01 -7.21269548e-01
-2.62489170e-01 -6.62052035e-02 1.83082730e-01 3.78769748e-02
4.27460700e-01 -6.60380483e-01 -8.18342566e-01 6.77076876e-02
-8.54055882e-01 -1.65720850e-01 6.54872060e-01 9.42486823e-01
3.06247234e-01 3.00629199e-01 6.14256859e-01 -5.46898484e-01
6.75756335e-01 -7.06580520e-01 -1.34854004e-01 2.64359191e-02
-6.19845450e-01 -2.75586277e-01 5.75106680e-01 -6.73310533e-02
-3.13257009e-01 1.63314342e-01 -4.37987804e-01 -2.36002773e-01
-1.08043246e-01 4.78082657e-01 3.92096013e-01 -1.34222895e-01
9.08726156e-01 2.55665362e-01 2.27205962e-01 -1.52320415e-01
-3.93788069e-01 1.03401446e+00 5.14834523e-01 -1.07037798e-01
4.36583847e-01 4.01211351e-01 1.99952960e-01 -7.73254871e-01
-6.08047307e-01 -5.64230382e-01 -6.77927196e-01 -3.06790709e-01
1.03535271e+00 -6.36156559e-01 -7.07901776e-01 3.89238596e-01
-1.06156600e+00 2.42748514e-01 4.15966511e-02 5.70205688e-01
-2.26889625e-01 3.08482140e-01 -5.35444915e-01 -6.61069274e-01
-8.80201280e-01 -1.52540982e+00 5.65976501e-01 4.46202978e-02
-2.66366452e-01 -5.11828423e-01 -2.06050351e-01 3.33847970e-01
5.58748484e-01 5.65923810e-01 1.00532734e+00 -8.24603260e-01
-1.44471228e-01 -6.34305716e-01 -1.62318990e-01 1.63177758e-01
1.37945995e-01 -7.61939660e-02 -9.30013061e-01 -3.91911119e-01
3.05833399e-01 -1.61416441e-01 9.91309762e-01 4.58740652e-01
1.23888540e+00 1.27951950e-01 -3.35090309e-01 6.70614183e-01
1.46600032e+00 7.35297441e-01 6.33217037e-01 6.82858288e-01
2.85804272e-01 2.07224950e-01 2.74165515e-02 5.74457288e-01
2.98173040e-01 3.22572470e-01 7.88759053e-01 1.50121838e-01
-7.36911967e-02 4.50076699e-01 -1.50859937e-01 7.45685875e-01
-3.14051718e-01 1.78410893e-03 -1.51722848e+00 3.89700770e-01
-1.35220814e+00 -9.52037632e-01 -4.44743901e-01 2.27958775e+00
6.19430065e-01 3.97393137e-01 1.10056937e-01 8.31593096e-01
8.05633962e-01 -4.79823947e-01 -2.20012203e-01 -2.91066557e-01
1.57241017e-01 7.41656959e-01 3.06295902e-01 9.19216722e-02
-1.16240597e+00 2.96392113e-01 5.91428709e+00 4.65425670e-01
-1.48207939e+00 2.99013436e-01 4.86962825e-01 -1.36832684e-01
3.28748941e-01 -5.14596999e-01 -3.59281361e-01 6.26149476e-01
9.56271350e-01 -8.95182267e-02 -3.32374647e-02 6.02531672e-01
-9.37649608e-03 -2.77492851e-02 -9.49927330e-01 1.30066586e+00
9.68497843e-02 -1.24240661e+00 6.44635856e-02 -1.31694928e-01
9.59351957e-02 2.22228497e-01 -2.31225237e-01 1.43608525e-01
-7.67271593e-02 -1.40050948e+00 8.17655087e-01 1.60673603e-01
7.86571503e-01 -1.08542323e+00 1.20009506e+00 3.01403344e-01
-1.08786714e+00 -5.12418509e-01 -4.60741162e-01 -2.23169520e-01
1.16799206e-01 6.15747154e-01 -8.85860324e-01 3.19966406e-01
1.03180730e+00 2.49646977e-01 -2.97845334e-01 1.72287261e+00
4.32001352e-02 3.44475597e-01 -3.21539670e-01 -4.32867892e-02
2.54458189e-01 1.27649456e-01 2.63133496e-01 9.96006250e-01
4.85457122e-01 3.39421839e-01 1.56692475e-01 1.95542663e-01
2.73842122e-02 4.57248330e-01 -6.21583343e-01 6.77523255e-01
4.76194918e-01 9.54206645e-01 -1.23978400e+00 -2.47205254e-02
-3.26393813e-01 6.77244425e-01 1.57980189e-01 -4.61896420e-01
-9.31663930e-01 -5.70771039e-01 2.25337207e-01 2.37846762e-01
-7.50916004e-02 1.76517829e-01 -5.58446586e-01 -7.84581959e-01
1.16010286e-01 -6.55338526e-01 6.32975042e-01 -4.44765627e-01
-9.98696148e-01 9.62394238e-01 -4.73841816e-01 -1.30216897e+00
2.84969866e-01 -4.80675846e-01 -7.25014746e-01 5.01485586e-01
-1.24367714e+00 -8.04079413e-01 -5.94067872e-01 5.92921972e-01
6.12898886e-01 -4.22619045e-01 9.91427481e-01 4.37849581e-01
-4.97326285e-01 6.78819597e-01 1.68445762e-02 4.71469045e-01
5.27787626e-01 -9.94550049e-01 -1.07214279e-01 5.45817792e-01
-7.82163069e-02 4.11596745e-01 4.81517404e-01 -2.82276094e-01
-1.03238106e+00 -1.11010695e+00 7.01352715e-01 -2.66772918e-02
3.65082681e-01 1.37047783e-01 -8.30362976e-01 3.58759463e-01
3.63573991e-02 7.86096007e-02 1.02500796e+00 -2.95920640e-01
-3.88199627e-01 -1.42306909e-01 -1.45179832e+00 2.45882228e-01
6.40069246e-01 -1.53783470e-01 -8.29632819e-01 5.48378885e-01
1.52271226e-01 -4.51720238e-01 -9.39469755e-01 4.16564316e-01
7.60818660e-01 -1.17500448e+00 9.79800940e-01 -4.42774326e-01
7.13760555e-01 -2.37667877e-02 -1.10832900e-01 -1.31769645e+00
-4.69177783e-01 1.71573013e-02 3.21162432e-01 4.57726449e-01
5.14291823e-01 -5.82284808e-01 5.45402408e-01 3.50028485e-01
-3.85057926e-01 -1.12487316e+00 -1.22416127e+00 -6.72854304e-01
2.35558867e-01 -4.07361329e-01 7.59163618e-01 8.58991921e-01
1.60693631e-01 1.73088431e-01 1.35180190e-01 1.84215605e-01
6.00601852e-01 -7.33677447e-02 1.88554958e-01 -1.67769527e+00
8.58329460e-02 -7.25917220e-01 -1.10151935e+00 -1.51822681e-03
-2.63896763e-01 -1.32789230e+00 -2.15628698e-01 -1.65363824e+00
4.64472324e-01 -4.31563228e-01 -6.11779213e-01 6.91570580e-01
1.75284073e-01 3.71722996e-01 -1.44769624e-01 2.62822866e-01
-3.20151448e-02 2.74392486e-01 9.75104749e-01 -4.19349521e-01
1.86180130e-01 2.21973866e-01 -3.82713765e-01 8.35972071e-01
9.18601990e-01 -7.62754321e-01 -2.54951566e-01 -4.49844778e-01
4.41758782e-02 2.59096205e-01 1.85267791e-01 -1.46125829e+00
4.07493472e-01 2.91086771e-02 3.89360219e-01 -5.58637440e-01
1.36009783e-01 -8.22104812e-01 2.44970798e-01 1.24239123e+00
-7.16047511e-02 3.13679427e-01 3.72453518e-02 3.37704182e-01
-3.86653960e-01 -4.10246789e-01 1.08760691e+00 -3.04466903e-01
-4.00996149e-01 5.72640240e-01 -5.63619614e-01 -2.90845901e-01
1.45791912e+00 -2.91720390e-01 -2.07170974e-02 5.22015803e-02
-8.67901683e-01 -9.49345306e-02 6.37318790e-02 3.18703800e-01
1.12375844e+00 -1.24465859e+00 -9.64335501e-01 3.32290530e-01
2.82719463e-01 1.81609482e-01 1.13148816e-01 1.01690102e+00
-1.13045406e+00 2.50605457e-02 -5.54377913e-01 -6.73326910e-01
-1.48455048e+00 3.15859795e-01 5.83670795e-01 1.82036489e-01
-1.19282353e+00 1.14819384e+00 -1.69867486e-01 -1.40001968e-01
2.94768751e-01 -3.89641136e-01 -3.89637500e-01 1.86159566e-01
8.62115324e-01 2.90997207e-01 7.61307061e-01 -5.50801396e-01
-6.21973634e-01 3.99106145e-01 -2.90422440e-01 2.71882236e-01
1.68152225e+00 4.32459921e-01 -3.48009109e-01 1.94327042e-01
1.25310028e+00 -4.66090679e-01 -3.62955540e-01 -5.62265776e-02
-1.51186988e-01 -2.67561078e-01 2.34344482e-01 -7.26489365e-01
-1.39898014e+00 1.04411864e+00 1.03835177e+00 1.09942712e-01
1.02301395e+00 -1.53816417e-01 1.16758358e+00 3.38854879e-01
6.99810684e-01 -7.63073087e-01 -9.23028216e-02 4.91284698e-01
7.91561007e-01 -1.23673487e+00 -1.63822398e-01 -9.73198116e-02
-2.28760689e-01 1.61052847e+00 2.61177540e-01 -2.10767105e-01
1.18374622e+00 4.86055672e-01 1.66355416e-01 -6.50312006e-01
-5.50793946e-01 1.10215463e-01 -1.52531609e-01 4.51167405e-01
4.88371491e-01 2.59566754e-01 -5.83424568e-01 8.48530650e-01
-5.83264649e-01 1.81634218e-01 6.11310840e-01 1.20541596e+00
-8.45764101e-01 -6.62508547e-01 -4.42272276e-01 9.04705346e-01
-7.76197314e-01 9.70650911e-02 7.49896169e-02 5.48886895e-01
2.36762062e-01 1.01062012e+00 -9.06541124e-02 -7.11133003e-01
4.02040124e-01 1.69055566e-01 3.22952151e-01 -5.30175209e-01
-7.31346548e-01 -4.85608697e-01 -3.58032912e-01 -2.78579950e-01
-6.28909990e-02 -5.69227874e-01 -1.57318878e+00 -3.75768155e-01
-4.41413581e-01 1.73230574e-01 7.56007850e-01 1.13032770e+00
-8.42577443e-02 7.39896715e-01 6.55085087e-01 -4.48191136e-01
-6.15484595e-01 -9.14080441e-01 -7.60557473e-01 4.19515848e-01
3.62798244e-01 -1.02697301e+00 -9.94810984e-02 -6.68731406e-02] | [14.832159042358398, -2.445688486099243] |
c66bc924-9d15-43e0-a99b-8d745e6a2344 | understanding-membership-inferences-on-well | 1802.04889 | null | http://arxiv.org/abs/1802.04889v1 | http://arxiv.org/pdf/1802.04889v1.pdf | Understanding Membership Inferences on Well-Generalized Learning Models | Membership Inference Attack (MIA) determines the presence of a record in a
machine learning model's training data by querying the model. Prior work has
shown that the attack is feasible when the model is overfitted to its training
data or when the adversary controls the training algorithm. However, when the
model is not overfitted and the adversary does not control the training
algorithm, the threat is not well understood. In this paper, we report a study
that discovers overfitting to be a sufficient but not a necessary condition for
an MIA to succeed. More specifically, we demonstrate that even a
well-generalized model contains vulnerable instances subject to a new
generalized MIA (GMIA). In GMIA, we use novel techniques for selecting
vulnerable instances and detecting their subtle influences ignored by
overfitting metrics. Specifically, we successfully identify individual records
with high precision in real-world datasets by querying black-box machine
learning models. Further we show that a vulnerable record can even be
indirectly attacked by querying other related records and existing
generalization techniques are found to be less effective in protecting the
vulnerable instances. Our findings sharpen the understanding of the fundamental
cause of the problem: the unique influences the training instance may have on
the model. | ['Xiao-Feng Wang', 'Vincent Bindschaedler', 'Kai Chen', 'Haixu Tang', 'Carl A. Gunter', 'Yunhui Long', 'Diyue Bu', 'Lei Wang'] | 2018-02-13 | null | null | null | null | ['membership-inference-attack'] | ['computer-vision'] | [ 5.64332187e-01 2.10565016e-01 -3.76614571e-01 -2.53468454e-01
-9.17185068e-01 -1.05905092e+00 3.99824440e-01 3.99377435e-01
2.37737503e-02 6.17015541e-01 -4.58456397e-01 -7.63295829e-01
-4.06234175e-01 -1.10264122e+00 -1.37816894e+00 -7.39434659e-01
-2.83876419e-01 6.42654955e-01 1.56536043e-01 1.61632270e-01
2.87938207e-01 7.36199737e-01 -1.48456359e+00 3.78685653e-01
4.40718323e-01 4.16798413e-01 -7.36765146e-01 7.02655137e-01
2.96425432e-01 6.84067130e-01 -1.04280031e+00 -5.28519332e-01
5.28956890e-01 -2.52318591e-01 -9.58866239e-01 -3.42329293e-01
9.62204039e-01 -1.73641786e-01 -2.64918596e-01 1.40304101e+00
-1.25343889e-01 -3.28139096e-01 4.03853446e-01 -1.87140334e+00
-2.01176614e-01 9.51785743e-01 -2.09310323e-01 1.24010578e-01
2.57835984e-01 8.69985074e-02 8.71403873e-01 5.15429713e-02
4.40485328e-01 1.15390861e+00 6.20154500e-01 6.11944258e-01
-1.59173679e+00 -1.14201677e+00 -1.19214430e-01 -1.24047205e-01
-1.61434722e+00 -3.88002276e-01 4.60253686e-01 -3.02803725e-01
5.07546723e-01 1.06531477e+00 -2.15323061e-01 9.11398768e-01
2.55278319e-01 1.34196430e-01 9.86536562e-01 -3.79466742e-01
4.18318778e-01 4.24890399e-01 5.79638422e-01 4.24884796e-01
8.94495189e-01 5.88110864e-01 -1.18052617e-01 -1.21656871e+00
7.10529685e-02 5.38264997e-02 -2.37785831e-01 -1.34520397e-01
-5.93839824e-01 8.22325408e-01 7.89661855e-02 1.82244733e-01
2.33307518e-02 1.85074434e-01 2.74538040e-01 6.84575498e-01
-5.34221670e-03 8.26334536e-01 -6.28716350e-01 2.52444088e-01
-6.48225605e-01 3.44315022e-01 1.14989567e+00 7.54292309e-01
7.35598147e-01 -3.17345470e-01 1.84835494e-01 -1.80007070e-01
9.51621905e-02 4.23098117e-01 1.49737587e-02 -8.28594744e-01
2.16391832e-01 7.34930098e-01 1.15356646e-01 -1.12400699e+00
4.82554222e-03 -3.38101178e-01 -3.84924382e-01 3.23144346e-01
6.84263527e-01 6.82846829e-02 -6.36196971e-01 2.19769001e+00
3.97212476e-01 4.96631980e-01 2.34404087e-01 2.84037262e-01
1.18393630e-01 5.06633937e-01 3.32391024e-01 -2.22916082e-01
1.22960103e+00 2.56386157e-02 -4.88448590e-01 3.38861011e-02
9.34075475e-01 -2.86202461e-01 8.38881731e-01 4.18434560e-01
-8.84049356e-01 -2.86816746e-01 -1.41833353e+00 4.33462888e-01
-5.82582295e-01 -6.99252725e-01 5.23060083e-01 1.13412762e+00
-4.93988395e-01 5.23631930e-01 -7.75482893e-01 -2.01097935e-01
4.56827670e-01 6.82615161e-01 -4.75245178e-01 -1.42221168e-01
-1.45052564e+00 7.61769116e-01 5.27802527e-01 -4.62767452e-01
-1.09523380e+00 -8.70020807e-01 -5.78834355e-01 1.27806529e-01
6.89148426e-01 -4.30834085e-01 8.83418381e-01 -6.08272612e-01
-4.58186120e-01 7.69637942e-01 -7.16843978e-02 -7.07133114e-01
5.46533823e-01 -1.61524210e-02 -7.58223951e-01 2.49132380e-01
-2.30211511e-01 -1.05175711e-01 8.58235657e-01 -1.53471231e+00
-4.76855248e-01 -7.50846326e-01 3.43026429e-01 -5.60892463e-01
-4.67211276e-01 1.67858124e-01 1.17647432e-01 -4.73576814e-01
-1.13301791e-01 -8.80201042e-01 -2.22954690e-01 -2.22716838e-01
-8.10827136e-01 3.55113088e-03 1.20396817e+00 -9.78469253e-02
1.49695110e+00 -2.22547603e+00 -6.07557833e-01 9.65150297e-01
2.14590877e-01 6.03017390e-01 2.03745708e-01 4.93439764e-01
-9.31282789e-02 8.01089704e-01 -4.19316888e-01 2.72970516e-02
-4.81715649e-02 5.39131045e-01 -1.02241170e+00 6.78663731e-01
-4.71253619e-02 5.02561629e-01 -4.15779620e-01 -1.76310197e-01
-1.89737171e-01 1.85604528e-01 -5.92102706e-01 2.58966237e-01
-4.52381879e-01 7.99031183e-02 -5.60430348e-01 5.19626498e-01
8.28675151e-01 -2.74628133e-01 4.22133356e-01 4.66480944e-03
3.42328459e-01 3.14803690e-01 -1.30162859e+00 6.13520861e-01
9.49875787e-02 1.32754549e-01 -1.09880172e-01 -7.74621069e-01
8.02872360e-01 3.85962516e-01 -2.89802458e-02 3.82471532e-02
2.32446575e-04 1.70981899e-01 -8.93870089e-03 -5.27937531e-01
1.97580829e-01 -2.21780658e-01 -1.33945018e-01 8.19830537e-01
-5.67675829e-01 3.84414226e-01 -2.71645635e-01 3.22608858e-01
1.43715107e+00 -6.02421939e-01 1.77502289e-01 -3.75460595e-01
4.38265264e-01 2.49232262e-01 6.27671480e-01 1.35275233e+00
-4.41398062e-02 2.63897806e-01 5.28532922e-01 -2.56578177e-01
-8.45709205e-01 -1.14325452e+00 -5.93693912e-01 8.04740906e-01
2.05064803e-01 -5.17554104e-01 -7.82413363e-01 -1.06311297e+00
5.39556146e-01 8.63332570e-01 -8.48924816e-01 -6.92397058e-01
-4.64829981e-01 -7.57597387e-01 1.32730770e+00 4.27999675e-01
3.94809783e-01 -5.52970529e-01 -4.01587337e-01 -1.76058024e-01
6.72620237e-02 -9.15332675e-01 -3.86879146e-01 1.81512430e-01
-8.71730804e-01 -1.60404921e+00 5.53879321e-01 -4.84742075e-01
1.02889621e+00 -3.04225720e-02 9.30297852e-01 6.98544562e-01
-2.39258274e-01 2.12894380e-01 2.02004552e-01 -5.28439045e-01
-1.05416226e+00 -9.38675075e-04 2.13922322e-01 -3.35202888e-02
8.68709803e-01 -4.82261479e-01 -2.14096457e-02 4.76429909e-01
-1.35326600e+00 -7.67012715e-01 1.49244219e-01 2.96165228e-01
3.59600961e-01 6.09506667e-01 7.13054299e-01 -1.74323368e+00
3.60324621e-01 -7.26644933e-01 -7.19783425e-01 6.15327656e-01
-7.50278711e-01 -4.69447337e-02 6.84076965e-01 -8.35341811e-01
-5.83378315e-01 -1.24699557e-02 1.99475497e-01 -5.07368267e-01
-4.67121869e-01 4.52501386e-01 -7.21216261e-01 -1.29235670e-01
9.02495623e-01 -5.51544428e-02 2.83007063e-02 -3.98677528e-01
-5.59652597e-02 5.10281384e-01 6.33119047e-01 -9.07008886e-01
1.32832241e+00 6.30620658e-01 2.82947123e-01 -5.12842894e-01
-6.01988554e-01 -1.14639632e-01 -2.13495359e-01 1.91943392e-01
2.38813117e-01 -4.56159085e-01 -1.07902145e+00 2.33714849e-01
-6.57364726e-01 -5.31315804e-02 -8.08063224e-02 2.28665739e-01
-2.26942264e-02 3.95220160e-01 -3.31313223e-01 -8.77671778e-01
3.31965201e-02 -9.38486457e-01 3.31584215e-01 -5.19755855e-03
-6.28644347e-01 -9.45686638e-01 -2.52491292e-02 4.00740862e-01
1.71985462e-01 5.21976709e-01 1.34370697e+00 -1.52523315e+00
-5.18079638e-01 -7.52871037e-01 2.48545215e-01 2.16270462e-01
1.37567058e-01 2.90806890e-01 -1.32903385e+00 -6.31343961e-01
3.66967231e-01 -2.26080641e-01 4.42923129e-01 -9.70294550e-02
1.23383725e+00 -8.95477414e-01 -6.32815242e-01 5.68856776e-01
1.37242961e+00 4.04584587e-01 6.94708645e-01 4.04902667e-01
3.79428595e-01 4.72461879e-01 3.89532924e-01 5.05118184e-02
-1.45997807e-01 4.18107003e-01 4.52913582e-01 2.59546824e-02
7.09103644e-01 -4.67541963e-01 2.34675527e-01 -2.37299174e-01
5.37257493e-01 -1.51360422e-01 -7.26350725e-01 2.71776110e-01
-1.43646955e+00 -1.05394840e+00 -1.98295310e-01 2.68601680e+00
1.14239645e+00 4.63332564e-01 7.88440257e-02 4.58435982e-01
7.03380346e-01 -3.10117722e-01 -6.38487458e-01 -4.55798477e-01
-9.91182327e-02 2.69454628e-01 7.41896868e-01 6.87801063e-01
-1.15215600e+00 4.56591040e-01 6.77746058e+00 5.37364662e-01
-8.60580385e-01 -2.77418971e-01 7.11479902e-01 2.67038532e-02
-4.96642888e-01 2.96282768e-01 -8.84516656e-01 3.45774382e-01
1.23379147e+00 -4.88414705e-01 1.89336777e-01 9.77696598e-01
-3.09963942e-01 1.23612948e-01 -1.64483106e+00 1.57024905e-01
1.00720532e-01 -1.06744850e+00 1.54510945e-01 5.19102156e-01
4.37754005e-01 -5.37764907e-01 3.20475340e-01 4.25390929e-01
5.91861248e-01 -1.23221040e+00 1.98961943e-01 3.41645360e-01
5.69559515e-01 -1.20577753e+00 5.67701042e-01 5.62688291e-01
-4.91053432e-01 -2.65529841e-01 -2.79473275e-01 2.72051431e-02
-5.24140835e-01 2.79334664e-01 -1.01301336e+00 3.29769045e-01
5.06289959e-01 -1.20807067e-01 -7.55921543e-01 7.92561471e-01
1.39242977e-01 1.17046797e+00 -6.61199033e-01 5.77749491e-01
-5.18471561e-02 3.00911754e-01 7.28212655e-01 9.01845753e-01
-8.31662491e-02 2.76776135e-01 2.24688753e-01 9.79347825e-01
-2.51847237e-01 -4.29074496e-01 -8.23808372e-01 -2.50931859e-01
8.60665500e-01 6.52902067e-01 -2.94562846e-01 -3.66548419e-01
-5.88673279e-02 2.39152700e-01 -3.96272130e-02 4.21592027e-01
-7.35985637e-01 -3.59537780e-01 9.51364160e-01 2.79853582e-01
5.93935773e-02 3.07944745e-01 -3.96347672e-01 -9.11889434e-01
-6.99728653e-02 -1.27250612e+00 9.55102503e-01 -2.29331046e-01
-1.34780824e+00 2.65347570e-01 4.97551896e-02 -7.91819394e-01
-2.49681249e-01 -2.61278510e-01 -7.64611423e-01 9.16297019e-01
-9.48909044e-01 -9.80821788e-01 3.37481529e-01 7.73738742e-01
-4.23954010e-01 4.11352031e-02 1.23162925e+00 1.46110103e-01
-7.50408590e-01 1.30001342e+00 -1.15626052e-01 2.47276619e-01
5.38093090e-01 -1.02668667e+00 2.61010438e-01 1.20219362e+00
1.11801498e-01 1.35795295e+00 9.56812501e-01 -9.12227690e-01
-1.42688191e+00 -1.29808331e+00 9.89509404e-01 -7.47447729e-01
4.83938903e-01 -3.57575744e-01 -1.62001073e+00 1.18959439e+00
-4.49350864e-01 -1.61705865e-03 9.82753813e-01 1.73455536e-01
-9.87551630e-01 -1.59223229e-01 -1.74456441e+00 3.59926522e-01
4.81075168e-01 -8.99337292e-01 -7.15533853e-01 -2.08518989e-02
6.99163020e-01 -4.71240394e-02 -8.54785085e-01 6.86985433e-01
4.16340262e-01 -6.37062907e-01 1.01467085e+00 -1.31518114e+00
-1.55394375e-01 -5.08122921e-01 -4.23383772e-01 -4.74274784e-01
6.13978773e-04 -7.51372874e-01 -5.74329555e-01 1.29687297e+00
4.78588343e-01 -9.50501204e-01 8.06366861e-01 1.22953892e+00
5.27758420e-01 -2.68242955e-01 -1.01543546e+00 -9.23368633e-01
2.82852203e-01 -4.50564682e-01 1.12005079e+00 1.52282774e+00
1.17134586e-01 -2.70584792e-01 -2.15886727e-01 1.19629705e+00
8.81822526e-01 1.45893499e-01 9.11679924e-01 -1.41586375e+00
-3.06394279e-01 -1.27113879e-01 -4.60698217e-01 -3.21065545e-01
-3.72273736e-02 -7.75805950e-01 -2.93749720e-01 -5.40224791e-01
2.28193432e-01 -6.71305895e-01 -5.09366035e-01 6.90428197e-01
-2.43767679e-01 2.12551668e-01 9.97603759e-02 3.38773042e-01
-5.11916690e-02 -4.53942955e-01 5.00229895e-01 -1.20193481e-01
3.41637768e-02 5.35181344e-01 -9.27976906e-01 4.47718889e-01
9.42963243e-01 -9.83567536e-01 -4.93456393e-01 3.90760839e-01
4.48084027e-01 -5.50542250e-02 6.39194846e-01 -8.61196637e-01
3.99894446e-01 -2.59260714e-01 8.08444098e-02 -3.98515165e-01
1.49124460e-02 -1.22238648e+00 6.71784580e-01 8.52130771e-01
-8.64572644e-01 -1.37329653e-01 3.71919602e-01 6.83273137e-01
2.05593139e-01 -3.78354043e-01 7.06539035e-01 2.42491186e-01
-2.23483052e-02 2.07007140e-01 -2.87244737e-01 7.23776668e-02
1.25220239e+00 -1.06543824e-01 -7.49662578e-01 -1.17992215e-01
-6.52131677e-01 1.38349399e-01 8.10394287e-01 1.15618803e-01
5.11186719e-01 -8.92996907e-01 -5.40755510e-01 7.02579916e-01
2.53745049e-01 -2.50173420e-01 -1.46667749e-01 4.42542940e-01
-1.03168868e-01 -3.03648636e-02 2.02438414e-01 -3.63575786e-01
-1.82152486e+00 1.06226230e+00 4.08168912e-01 8.37190226e-02
-5.18331587e-01 5.35296023e-01 1.09577104e-01 -2.84701943e-01
5.94627321e-01 -2.00623482e-01 3.27361703e-01 -3.22465628e-01
8.80867839e-01 3.07427645e-01 5.79861067e-02 -4.35572267e-01
-3.11767966e-01 7.00631142e-02 -5.27613103e-01 3.03017169e-01
9.18214977e-01 2.02764317e-01 -2.13843808e-01 3.40020955e-01
1.24935257e+00 2.24177331e-01 -7.01757073e-01 -3.42673481e-01
7.81635493e-02 -7.09037304e-01 -1.76355004e-01 -8.60141754e-01
-9.56591427e-01 6.09097064e-01 3.45425159e-01 7.08282590e-01
9.49008524e-01 -6.28145710e-02 5.50014973e-01 6.85297906e-01
3.52683842e-01 -4.85643506e-01 -5.16265512e-01 -4.56222035e-02
4.05871660e-01 -9.78007078e-01 9.91065279e-02 -3.15669268e-01
-3.31493109e-01 7.57667780e-01 5.40864170e-01 4.02297918e-03
8.05098951e-01 6.36956573e-01 1.30899355e-01 -1.59977883e-01
-8.91707897e-01 6.13695502e-01 -9.30147767e-02 3.98721337e-01
-4.56080854e-01 9.40257385e-02 3.09786260e-01 6.70310855e-01
-2.26997659e-01 -4.72527780e-02 5.67175448e-01 1.07982981e+00
-4.00453568e-01 -9.91282642e-01 -8.00166070e-01 4.90633577e-01
-9.31996107e-01 1.13012545e-01 -6.42277479e-01 9.11068201e-01
-3.48755158e-02 1.02084017e+00 4.85045686e-02 -6.28511965e-01
1.62821919e-01 3.43175471e-01 1.23999109e-02 -4.97443050e-01
-9.72054958e-01 -4.06341493e-01 3.35688926e-02 -3.65105450e-01
-1.78317856e-02 -5.42698622e-01 -1.19047928e+00 -8.45127165e-01
-5.07944226e-01 4.83837664e-01 2.40252137e-01 8.42173159e-01
4.69794214e-01 1.09174050e-01 9.23000693e-01 3.03655744e-01
-9.79514360e-01 -5.13653398e-01 -6.44019961e-01 5.22440016e-01
5.49834311e-01 -2.96683758e-01 -1.01191044e+00 1.74408983e-02] | [5.8367085456848145, 7.3339667320251465] |
d9410ffc-5b7d-426e-a791-87632daf2807 | denoising-autoencoder-based-defensive | 2303.15901 | null | https://arxiv.org/abs/2303.15901v1 | https://arxiv.org/pdf/2303.15901v1.pdf | Denoising Autoencoder-based Defensive Distillation as an Adversarial Robustness Algorithm | Adversarial attacks significantly threaten the robustness of deep neural networks (DNNs). Despite the multiple defensive methods employed, they are nevertheless vulnerable to poison attacks, where attackers meddle with the initial training data. In order to defend DNNs against such adversarial attacks, this work proposes a novel method that combines the defensive distillation mechanism with a denoising autoencoder (DAE). This technique tries to lower the sensitivity of the distilled model to poison attacks by spotting and reconstructing poisonous adversarial inputs in the training data. We added carefully created adversarial samples to the initial training data to assess the proposed method's performance. Our experimental findings demonstrate that our method successfully identified and reconstructed the poisonous inputs while also considering enhancing the DNN's resilience. The proposed approach provides a potent and robust defense mechanism for DNNs in various applications where data poisoning attacks are a concern. Thus, the defensive distillation technique's limitation posed by poisonous adversarial attacks is overcome. | ['António Casimiro', 'José Cecílio', 'Bakary Badjie'] | 2023-03-28 | null | null | null | null | ['data-poisoning'] | ['adversarial'] | [ 1.12043284e-01 -1.17988847e-01 6.55888438e-01 -2.92658191e-02
-3.83440435e-01 -1.22004116e+00 6.54891372e-01 -3.10561378e-02
-7.12426901e-01 7.51561344e-01 3.40531766e-02 -2.81282276e-01
2.64111906e-01 -1.07002544e+00 -9.19572353e-01 -1.24584460e+00
9.11295712e-02 -2.60183007e-01 2.70329595e-01 -4.01606888e-01
2.88000405e-01 1.13102162e+00 -1.38030553e+00 2.17766374e-01
8.28124225e-01 9.23872232e-01 -1.08705811e-01 7.26256430e-01
1.32643819e-01 1.04420245e+00 -1.29290128e+00 -7.99092591e-01
7.20269620e-01 -1.70164898e-01 -3.27357829e-01 -4.44997519e-01
4.91070181e-01 -8.25403929e-01 -6.24397755e-01 1.64013314e+00
6.36852443e-01 -8.52487236e-02 4.35179383e-01 -1.46722627e+00
-6.87693298e-01 5.25812328e-01 -1.84136510e-01 4.66053098e-01
-1.79086268e-01 5.08368611e-01 2.94651240e-01 -5.03487110e-01
1.96463525e-01 1.24348319e+00 4.63588148e-01 1.09013462e+00
-9.22422469e-01 -1.24009371e+00 -3.95287313e-02 -2.50108689e-01
-1.13456416e+00 -4.66633439e-01 9.09217298e-01 -1.73082575e-01
4.19413775e-01 3.19691002e-01 1.44371316e-01 1.60854995e+00
5.50026774e-01 4.15071249e-01 9.97001350e-01 -4.92433831e-03
5.09099603e-01 2.35675171e-01 -1.13598108e-02 4.28923577e-01
4.42407876e-01 3.72573078e-01 -1.14874437e-01 -5.25153220e-01
6.25357747e-01 1.16128735e-01 -2.29295075e-01 4.41604853e-01
-5.02991259e-01 8.65001917e-01 6.69778585e-01 2.63493598e-01
-4.95627165e-01 7.67336711e-02 7.74626493e-01 4.01816130e-01
4.97329570e-02 4.85456944e-01 -3.31263512e-01 4.19671386e-01
-3.45438331e-01 1.82263795e-02 5.94117820e-01 5.45809150e-01
1.53369516e-01 7.81802475e-01 2.09191456e-01 5.23543835e-01
3.77840132e-01 4.98094231e-01 3.55090648e-01 -4.08348620e-01
5.09737372e-01 4.82683301e-01 -9.95064154e-02 -1.07493138e+00
1.74111426e-01 -4.53573704e-01 -8.53812099e-01 8.21676016e-01
3.48096877e-01 -6.02492511e-01 -9.95976329e-01 1.82799888e+00
4.33599442e-01 2.26347491e-01 7.61872709e-01 6.69900954e-01
8.00677061e-01 7.51574039e-01 5.70239186e-01 1.85271055e-01
1.16216886e+00 -3.28323036e-01 -7.09823072e-01 7.07665682e-02
1.80703148e-01 -5.24667859e-01 7.91217148e-01 6.94607317e-01
-8.45034361e-01 -5.31665921e-01 -1.49346471e+00 3.62265408e-01
-5.96265912e-01 -3.87651771e-01 2.54124761e-01 1.29867864e+00
-6.55078411e-01 6.49417937e-01 -7.10214078e-01 1.46353524e-02
7.07204640e-01 6.20174050e-01 -5.64964235e-01 6.43626824e-02
-1.54348814e+00 8.06682825e-01 4.17307764e-01 1.97383761e-01
-1.58687294e+00 -7.40538478e-01 -5.12003243e-01 7.50409625e-03
-9.35751721e-02 -1.50911182e-01 8.08245182e-01 -8.84469032e-01
-1.21628237e+00 5.39631188e-01 8.21617126e-01 -8.65063310e-01
7.48243868e-01 -3.96528512e-01 -6.04484618e-01 4.77146685e-01
-4.54866529e-01 2.29274064e-01 1.10567021e+00 -1.49223053e+00
-3.31190139e-01 -4.38684493e-01 1.75484076e-01 -1.31869148e-02
-9.39720154e-01 1.90311581e-01 4.99896735e-01 -1.03951311e+00
-1.61870018e-01 -4.22016233e-01 -2.00124726e-01 5.60787767e-02
-4.53613281e-01 1.57659307e-01 1.27373016e+00 -3.64476681e-01
8.41902912e-01 -2.32323265e+00 -3.87715727e-01 1.72240674e-01
2.49715060e-01 1.04615426e+00 -1.24060668e-01 6.47250891e-01
-1.72785848e-01 4.37798440e-01 -2.20055968e-01 -5.74027039e-02
-1.75835997e-01 3.82954389e-01 -8.43151391e-01 7.74421155e-01
3.17481101e-01 3.35882753e-01 -5.90256393e-01 -1.28185209e-02
1.56275913e-01 7.86211491e-01 -3.53407025e-01 4.71575260e-01
-4.64603119e-03 2.72147059e-01 -5.78270614e-01 6.93115532e-01
1.04060149e+00 8.79093409e-01 -2.58771390e-01 -3.32204133e-01
2.57836103e-01 -3.11200678e-01 -1.15031803e+00 5.65734804e-01
-2.82103550e-02 3.03273618e-01 3.25924784e-01 -7.50697613e-01
1.27637374e+00 4.67186421e-01 1.22603662e-01 -4.18211699e-01
7.47380555e-01 1.70322284e-01 1.41893327e-01 -5.15756369e-01
4.85961474e-02 -5.42779267e-01 -3.88055369e-02 2.33750284e-01
1.14755124e-01 2.18308166e-01 -4.60443318e-01 3.24296385e-01
1.31756461e+00 -3.79844099e-01 -9.03252140e-02 -3.78607631e-01
9.00362253e-01 -2.97232658e-01 6.48203731e-01 6.51864171e-01
-5.31472623e-01 5.40493056e-02 4.02057648e-01 -5.04529476e-01
-1.00241601e+00 -1.34491742e+00 -1.61752030e-01 8.18502665e-01
1.18972389e-02 1.44041896e-01 -1.12442827e+00 -9.02162015e-01
-3.23454030e-02 6.70816898e-01 -7.12566674e-01 -8.32128644e-01
-4.53403622e-01 -7.00305939e-01 1.66598153e+00 6.19333744e-01
1.04461837e+00 -9.63601112e-01 -4.91290748e-01 7.18569160e-02
3.72409075e-01 -9.28979337e-01 -3.31435055e-02 5.59845924e-01
-9.46217537e-01 -1.07116520e+00 -3.95741194e-01 -6.52848184e-01
6.79944277e-01 -1.45929242e-02 4.85487610e-01 3.23966891e-01
-2.21617565e-01 -6.92614615e-02 -3.75473410e-01 -7.59349167e-01
-1.00978661e+00 -5.00073731e-01 5.38091481e-01 2.84150206e-02
3.62236261e-01 -7.02247322e-01 -5.55716038e-01 3.06582570e-01
-1.49508417e+00 -1.02822363e+00 3.55194926e-01 6.10454381e-01
1.75464481e-01 7.17896342e-01 9.90806580e-01 -8.95174980e-01
8.43440354e-01 -6.42729223e-01 -6.25572562e-01 -1.31841842e-02
-1.33613929e-01 3.01483199e-02 1.44618511e+00 -8.20381522e-01
-1.05764449e+00 -1.31192217e-02 -7.43775010e-01 -6.33702874e-01
-5.04968584e-01 6.32096827e-02 -8.07004571e-01 -7.24563122e-01
9.28014398e-01 1.71362713e-01 -8.51349011e-02 -5.65040290e-01
9.71038863e-02 4.58486736e-01 8.20954740e-01 -4.63335305e-01
1.25054383e+00 3.96191329e-01 1.28978297e-01 -7.22992420e-01
-3.96128863e-01 2.31081724e-01 -3.25193137e-01 -2.71947324e-01
8.87799084e-01 -7.68813908e-01 -8.92659128e-01 1.03851926e+00
-1.08110857e+00 9.56801772e-02 2.09390208e-01 9.95192453e-02
1.34918913e-01 5.72835803e-01 -8.58122587e-01 -9.85269070e-01
-5.77828050e-01 -1.03229797e+00 1.49374411e-01 2.72834867e-01
3.93560588e-01 -8.51537883e-01 -4.07660194e-02 3.68968338e-01
3.71621460e-01 9.88384843e-01 9.98859465e-01 -1.20816171e+00
-2.43315041e-01 -5.78436911e-01 1.23812921e-01 1.02169907e+00
1.09913751e-01 2.68734843e-01 -1.31710303e+00 -4.26404983e-01
8.28625262e-01 -5.10599971e-01 6.07128561e-01 -2.51820683e-01
8.77752244e-01 -7.96959221e-01 1.78477824e-01 6.98164403e-01
1.79309535e+00 6.55237079e-01 1.04210114e+00 5.12289762e-01
6.50713325e-01 5.04862130e-01 2.65048832e-01 2.88507611e-01
-5.25537968e-01 -2.05329031e-01 1.31782162e+00 -8.37267488e-02
3.00085783e-01 -2.50852108e-01 5.95848024e-01 4.11660910e-01
1.61236420e-01 -5.92450023e-01 -6.16227508e-01 1.30543202e-01
-1.12779880e+00 -1.00168884e+00 -1.98922381e-01 1.89926517e+00
7.82425761e-01 4.38838571e-01 -7.89589435e-02 5.70407212e-01
7.23178387e-01 4.88983877e-02 -7.83267736e-01 -6.06237292e-01
-3.73613238e-01 3.11706811e-01 7.70610571e-01 3.42539161e-01
-1.22775447e+00 8.94165814e-01 6.20081806e+00 6.22747898e-01
-1.22078097e+00 -4.68427949e-02 4.18717563e-01 1.05288932e-02
1.41496360e-01 -6.22590482e-01 -7.31879950e-01 4.89358872e-01
1.04494703e+00 1.31571263e-01 1.47208259e-01 9.18394804e-01
-1.22402720e-02 3.09545547e-01 -8.06920946e-01 4.99795139e-01
-8.52873102e-02 -1.04720020e+00 4.40188378e-01 5.66188954e-02
4.21418667e-01 -9.89724696e-02 3.14555526e-01 2.02836871e-01
6.59292042e-01 -1.08949232e+00 6.74761772e-01 1.80717587e-01
2.63226449e-01 -1.44931746e+00 9.41146672e-01 4.31919992e-01
-7.83364236e-01 -4.46975082e-01 -5.80107093e-01 3.45578790e-02
-2.97336608e-01 1.60817459e-01 -5.86799324e-01 1.61908314e-01
6.75833464e-01 -4.49000634e-02 -6.70584857e-01 5.33577263e-01
-2.90336013e-01 7.70732820e-01 -1.17617548e-01 1.61246136e-01
4.17335659e-01 -2.35088426e-03 6.54255867e-01 9.17920649e-01
-6.23681583e-02 2.30612695e-01 -2.90830374e-01 8.12994182e-01
-3.56034547e-01 -1.26851559e-01 -1.02135420e+00 -1.87864915e-01
7.44299471e-01 1.02254295e+00 -6.31564140e-01 -1.56755716e-01
2.64746882e-02 8.61491263e-01 5.23452759e-02 3.20690602e-01
-9.75710928e-01 -6.60782695e-01 1.00645924e+00 -2.99891889e-01
1.22675598e-01 -1.41603947e-01 -3.84600848e-01 -7.71279216e-01
-1.54706359e-01 -1.25745523e+00 4.91860658e-01 -4.20637429e-01
-1.57438111e+00 9.97277200e-01 -3.01221073e-01 -1.07703650e+00
3.66749704e-01 -7.71524549e-01 -8.96963596e-01 8.22311461e-01
-9.97523308e-01 -1.00049114e+00 -7.24365115e-02 1.09111822e+00
3.88506323e-01 -6.15159035e-01 7.30907440e-01 2.74305046e-01
-8.63992155e-01 8.94952536e-01 5.86570017e-02 5.89546740e-01
4.16738480e-01 -9.05850887e-01 2.94001877e-01 1.50066912e+00
-2.54834831e-01 1.07189310e+00 1.03534496e+00 -8.55438173e-01
-1.44406033e+00 -1.24275219e+00 4.93060052e-02 -2.57593721e-01
7.40863800e-01 -6.88968897e-01 -1.22252023e+00 4.52048361e-01
3.34335566e-01 -2.09054090e-02 8.81501317e-01 -7.46316373e-01
-6.92419112e-01 -2.26546481e-01 -1.98597956e+00 7.73668647e-01
3.51139605e-01 -6.25565410e-01 -7.66731501e-01 -1.34775799e-03
8.08977485e-01 1.17588326e-01 -9.57793295e-01 1.99692994e-01
2.86706239e-01 -8.63348484e-01 1.10239339e+00 -8.46567214e-01
4.09684509e-01 -4.24691707e-01 -4.24507737e-01 -1.06464398e+00
-6.25544786e-02 -4.50375468e-01 -5.82914539e-02 1.57429898e+00
-5.83803505e-02 -8.38820517e-01 7.27058411e-01 6.73060238e-01
1.45882536e-02 -3.63205522e-01 -8.59255552e-01 -8.18522930e-01
5.17848492e-01 7.10414797e-02 6.25527024e-01 1.03022707e+00
-5.98025262e-01 -1.73914328e-01 -3.96588683e-01 8.31794143e-01
1.00259638e+00 -1.07777441e+00 5.06481528e-01 -1.03556097e+00
3.69033106e-02 -4.45819274e-02 -5.90171278e-01 -1.50826052e-01
-1.53227616e-02 -5.19384623e-01 -1.20847644e-02 -7.12528884e-01
-4.28631842e-01 -7.87608027e-02 -6.70108199e-01 5.28708220e-01
-2.86940664e-01 4.98999357e-01 2.36738577e-01 5.04840389e-02
1.88862875e-01 4.18111235e-01 9.10492241e-01 -7.58289695e-02
1.64705008e-01 -5.12711667e-02 -8.90004039e-01 7.38135755e-01
1.09217513e+00 -9.79170382e-01 -6.97941959e-01 -3.96518320e-01
-2.23473743e-01 -3.64669263e-01 6.53204978e-01 -1.19768345e+00
2.75699049e-01 -7.03178570e-02 4.56083685e-01 -4.23585862e-01
6.57349452e-02 -1.26380587e+00 1.26585856e-01 9.86440301e-01
-2.24162623e-01 1.38372421e-01 6.13866866e-01 6.75843179e-01
-6.75776079e-02 -4.69689488e-01 1.37657642e+00 -2.13676602e-01
-4.30369645e-01 1.27834961e-01 -7.43732274e-01 -2.43626282e-01
1.34918082e+00 -3.94809693e-01 -5.64453661e-01 1.24233238e-01
-4.43348944e-01 -3.95284072e-02 3.86711657e-01 1.03685163e-01
8.28651667e-01 -1.14810777e+00 -4.43148077e-01 3.22142363e-01
-3.09408367e-01 -2.70147443e-01 3.09356421e-01 8.09006244e-02
-8.40358496e-01 -2.10625187e-01 -9.12835956e-01 9.75989699e-02
-1.55877388e+00 1.19334912e+00 4.65535641e-01 2.50877380e-01
-6.29012227e-01 1.01381612e+00 1.54836267e-01 -1.17076315e-01
6.89017773e-01 7.19034150e-02 -3.37794989e-01 -1.43984571e-01
6.55123532e-01 4.44469005e-01 6.30501360e-02 -4.75195616e-01
-4.22320575e-01 -3.38157192e-02 -3.71439457e-01 1.54279843e-01
1.42097056e+00 1.28360406e-01 -9.73246992e-02 -1.68811992e-01
1.23638630e+00 7.20567033e-02 -1.39350295e+00 1.67394295e-01
-2.83375949e-01 -3.62387151e-01 1.82444111e-01 -9.29830134e-01
-1.30539072e+00 9.19628084e-01 6.89336061e-01 3.27683389e-01
1.28069878e+00 -6.37202859e-01 9.86131728e-01 5.14111400e-01
8.45083371e-02 -7.41151094e-01 3.72849703e-01 1.90426141e-01
7.34007359e-01 -8.73648703e-01 -2.45416343e-01 -1.75780505e-01
-5.75440526e-01 1.20728314e+00 8.41860890e-01 -6.86067224e-01
4.81420070e-01 6.68983102e-01 5.80735505e-01 -1.88559413e-01
-6.14775002e-01 6.47724628e-01 -2.42375240e-01 9.78516221e-01
-2.46447951e-01 -4.48270470e-01 -1.50209395e-02 7.37768233e-01
-1.58625141e-01 -6.16042197e-01 6.74905896e-01 1.08817458e+00
-4.88216877e-01 -9.78734314e-01 -8.34824204e-01 1.98691133e-02
-9.29592788e-01 -6.67692162e-03 -7.53118992e-01 6.67893827e-01
3.01002473e-01 1.10414267e+00 -4.79337811e-01 -8.42091680e-01
3.21168035e-01 -5.77932112e-02 -4.87131923e-02 -1.55800760e-01
-1.42723560e+00 -1.41151443e-01 -2.15591952e-01 -2.69210994e-01
3.86888422e-02 -2.93248624e-01 -1.06572497e+00 -4.90922600e-01
-2.22510800e-01 9.00210664e-02 7.49359369e-01 6.84559822e-01
-9.66237206e-03 5.57232141e-01 1.07626545e+00 -3.75445634e-01
-1.07899523e+00 -6.51561379e-01 -6.52428150e-01 3.93274426e-01
6.28610075e-01 -5.28709888e-01 -7.96000779e-01 5.52365445e-02] | [5.517302513122559, 7.9446258544921875] |
5a491b7d-54db-486b-8e29-c66d9e41fcbd | marnet-backdoor-attacks-against-value | null | null | https://openreview.net/forum?id=-VsGCG_AQ69 | https://openreview.net/pdf?id=-VsGCG_AQ69 | MARNET: Backdoor Attacks against Value-Decomposition Multi-Agent Reinforcement Learning | Recent works have revealed that backdoor attacks against Deep Reinforcement Learning (DRL) could lead to abnormal action selection of the agent, which may result in failure or even catastrophe in crucial decision processes. However, existing attacks only consider single-agent RL systems, in which the only agent can observe the global state and have full control of the decision process. In this paper, we explore a new backdoor attack paradigm in cooperative multi-agent reinforcement learning (CMARL) scenarios, where a group of agents coordinate with each other to achieve a common goal, while each agent can only observe the local state, e.g., StarCraft II (Vinyals et al. (2017)). In the proposed MARNet attack framework, we carefully design a pipeline of trigger design, action poisoning and reward hacking modules to accommodate the cooperative multi-agent momentums. In particular, as only a subset of agents can observe the triggers in their local observations, we maneuver their actions to the worst actions suggested by an expert policy model. Since the global reward in CMARL is aggregated by individual rewards from all agents, we propose to modify the reward in a way that boosts the bad actions of poisoned agents (agents who observe the triggers) but mitigates the influence on non-poisoned agents. We conduct extensive experiments on two classical MARL algorithms VDN (Sunehag et al. (2018)) and QMIX (Rashid et al. (2018)), in two popular CMARL games Predator Prey (Boehmer et al. (2020)) and SMAC (Samvelyan et al. (2019)). The results show that MARNet outperforms baselines extended from single-agent DRL backdoor attacks TrojDRL (Kiourti et al. (2020)) and Multitasking learning (Ashcraft & Karra (2021)) by reducing the utility under attack by as much as 100%. We apply fine-tuning as a defense against MARNet, and demonstrate that fine-tuning cannot entirely eliminate the effect of the attack. | ['Xueluan Gong', 'Zhicong Zheng', 'Yanjiao Chen'] | 2021-09-29 | null | null | null | null | ['smac-1', 'smac'] | ['playing-games', 'playing-games'] | [-4.42129105e-01 2.78298020e-01 -1.79036722e-01 4.69911873e-01
-4.99140024e-01 -1.10814381e+00 9.21044230e-01 5.37482537e-02
-8.86754990e-01 9.31440651e-01 -7.59270862e-02 -3.09153855e-01
-1.33889332e-01 -7.59457707e-01 -7.52098978e-01 -1.00667143e+00
-4.62575257e-01 5.94945431e-01 3.01180810e-01 -4.72313643e-01
1.47207584e-02 3.82695735e-01 -9.15507734e-01 -1.89968780e-01
6.61727428e-01 4.58278775e-01 -1.00719035e-01 7.29580641e-01
6.86677635e-01 1.34650183e+00 -9.57505465e-01 -1.80618316e-01
4.45651084e-01 -3.43866229e-01 -5.02378106e-01 -3.65768880e-01
-1.73312917e-01 -7.15995193e-01 -5.17444015e-01 1.13341403e+00
6.55512452e-01 5.12135737e-02 4.04318273e-01 -1.64345360e+00
-2.06336185e-01 1.15863907e+00 -8.18748653e-01 8.88967216e-02
-1.26666781e-02 9.43115354e-01 9.46561337e-01 -3.35440040e-02
3.51868659e-01 1.44128096e+00 2.09966362e-01 1.02112269e+00
-1.23782837e+00 -9.37012792e-01 6.90751016e-01 9.13768262e-02
-8.17984521e-01 -1.59192681e-01 4.70401376e-01 -1.04246542e-01
9.27209437e-01 8.91260579e-02 6.75317109e-01 1.76764667e+00
5.10260046e-01 9.87931907e-01 1.26696038e+00 5.51447980e-02
7.16371059e-01 -3.86011451e-01 -1.16511136e-01 5.49382627e-01
4.87236649e-01 7.44809508e-01 -4.12760019e-01 -7.79442310e-01
7.51574159e-01 -6.05925806e-02 7.21680969e-02 -2.53691614e-01
-1.07579970e+00 1.06603944e+00 4.15350437e-01 -4.41219732e-02
-8.32165480e-01 6.26126528e-01 5.35735905e-01 6.63021922e-01
1.57257449e-02 9.35511589e-01 -4.98654544e-01 -2.78286338e-01
-1.24318644e-01 7.14980960e-01 8.45686495e-01 2.65863061e-01
4.81555760e-01 4.18019652e-01 -3.15324754e-01 3.87194343e-02
2.44483024e-01 5.75947583e-01 3.91925961e-01 -1.29353762e+00
2.45643258e-01 2.00928450e-01 4.90738213e-01 -6.37404442e-01
-6.86844826e-01 -6.08755827e-01 -4.73899543e-01 7.78332472e-01
4.74873990e-01 -8.28543305e-01 -5.61573148e-01 2.28895354e+00
3.72566462e-01 2.76207298e-01 5.32057464e-01 9.57658529e-01
1.53079957e-01 5.61962187e-01 2.75834888e-01 -3.45442295e-01
1.09774876e+00 -9.04424489e-01 -3.38529170e-01 -3.24446112e-01
7.05544710e-01 -1.73073575e-01 9.21120405e-01 7.48717546e-01
-1.03339744e+00 8.56202841e-02 -8.74052167e-01 8.02702963e-01
5.66768795e-02 -4.74598795e-01 5.78444898e-01 3.38679999e-01
-7.59971559e-01 4.90753800e-01 -1.27202368e+00 -8.45008716e-02
4.38698709e-01 3.98256093e-01 1.83403641e-01 4.17462498e-01
-1.40616310e+00 8.91360939e-01 2.18085513e-01 -2.08401158e-01
-2.10723686e+00 -5.08624196e-01 -3.76079381e-01 2.93351244e-03
1.21635079e+00 -5.36205292e-01 1.42473781e+00 -7.82867491e-01
-1.80036604e+00 2.30974644e-01 8.06298375e-01 -9.50488210e-01
7.37897277e-01 -4.99234021e-01 1.76861063e-01 4.98205386e-02
3.78113873e-02 5.96465290e-01 9.60457206e-01 -1.20258069e+00
-6.48916006e-01 -1.88716844e-01 6.79872334e-01 5.42401969e-01
-1.18079036e-01 -6.69298768e-02 3.49746555e-01 -5.34129560e-01
-7.02607274e-01 -1.18488896e+00 -5.12380064e-01 -6.67432606e-01
-5.66216111e-01 -3.18605065e-01 7.15726972e-01 -4.21006382e-02
7.65529573e-01 -1.93375063e+00 4.46612000e-01 -3.54180601e-03
5.50983608e-01 2.38928616e-01 -5.73279738e-01 6.84834778e-01
2.49266863e-01 1.02570459e-01 9.75782499e-02 -2.20592484e-01
2.34212577e-01 3.90566975e-01 -7.40343273e-01 7.41540492e-01
-2.11073026e-01 9.47165847e-01 -1.07662046e+00 9.94342342e-02
3.12727131e-02 1.03686959e-01 -6.42117977e-01 2.56340146e-01
-6.70026958e-01 7.10704267e-01 -7.55648732e-01 4.72441554e-01
1.86159804e-01 2.07435831e-01 3.00080240e-01 2.91823000e-01
-2.74057209e-01 4.01019186e-01 -1.09515011e+00 9.90741789e-01
-1.90071553e-01 7.35905245e-02 3.49019766e-01 -6.25110447e-01
3.38355690e-01 4.62612033e-01 5.76255381e-01 -5.64284742e-01
2.45652184e-01 -8.27565789e-03 5.18477380e-01 7.75295077e-03
1.96260050e-01 1.13453194e-02 -3.84186298e-01 9.20645535e-01
-2.10027680e-01 -2.83220247e-03 1.91489458e-01 5.59473753e-01
1.51287234e+00 5.09460643e-02 4.36373085e-01 2.39495952e-02
1.99428171e-01 2.34047845e-01 7.14697599e-01 1.49721503e+00
-5.83827257e-01 -3.93895626e-01 1.19012833e+00 -2.88356960e-01
-7.49331594e-01 -9.48167443e-01 5.65394580e-01 1.53525722e+00
1.18919976e-01 -4.09950703e-01 -8.16879928e-01 -1.00619972e+00
2.05973655e-01 8.80337596e-01 -9.26185846e-01 -5.65220296e-01
-6.86084509e-01 -7.68931389e-01 1.08122742e+00 5.32672815e-02
4.23243791e-01 -1.61618292e+00 -1.17645395e+00 3.04862827e-01
1.79279551e-01 -6.59879565e-01 -4.71650809e-01 4.01546955e-01
-3.37326169e-01 -1.19903696e+00 -1.51205480e-01 2.72552781e-02
2.13896707e-01 7.90419430e-02 7.05359519e-01 -4.79931049e-02
7.77485147e-02 5.99497139e-01 -4.00058061e-01 -3.69112879e-01
-6.92104518e-01 9.02389809e-02 6.93102658e-01 -3.20497513e-01
-8.27285368e-03 -4.86435294e-01 -4.40135062e-01 2.61037141e-01
-7.20422745e-01 -3.63701284e-01 5.23509026e-01 7.99441874e-01
1.64144337e-01 7.38333017e-02 8.55707943e-01 -5.66341937e-01
1.05035198e+00 -5.84343195e-01 -1.21718073e+00 -2.50499925e-05
-4.46034342e-01 8.38390812e-02 9.83769596e-01 -9.91483808e-01
-5.62395215e-01 -2.98060894e-01 3.48149925e-01 -7.16834545e-01
-4.64129597e-02 1.51598990e-01 -8.54367316e-02 -7.15735629e-02
7.24737942e-01 2.52283216e-01 1.88148394e-02 -2.35173643e-01
3.57412606e-01 2.94731290e-04 2.16086894e-01 -9.08511996e-01
1.09087777e+00 2.21032187e-01 2.19484363e-02 -1.85973898e-01
-6.26958489e-01 3.09154630e-01 9.34211016e-02 -3.06924701e-01
7.08436430e-01 -1.00121260e+00 -1.51475406e+00 6.10467136e-01
-9.96306002e-01 -9.10636723e-01 -4.10048634e-01 4.39022452e-01
-6.28357768e-01 1.62639111e-01 -7.89557993e-01 -9.25110102e-01
-1.13342851e-01 -1.28611457e+00 4.42199498e-01 3.26508343e-01
-1.17220702e-02 -6.75087392e-01 5.35540819e-01 -5.98069914e-02
3.77192438e-01 2.56741822e-01 6.12793565e-01 -1.02491522e+00
-5.70764780e-01 3.58915329e-01 4.70121741e-01 7.87029639e-02
-7.51460791e-02 -1.45677045e-01 -7.24016130e-01 -7.37556875e-01
2.46049926e-01 -9.04014766e-01 7.37462819e-01 3.82698476e-01
6.18731081e-01 -9.24588859e-01 -1.25138760e-01 2.05452681e-01
9.30314422e-01 4.83847201e-01 1.91514760e-01 8.34233880e-01
4.93133724e-01 4.71533895e-01 7.26954937e-01 7.27875531e-01
3.04335862e-01 4.12083387e-01 1.24668252e+00 2.61032879e-01
4.34295148e-01 -4.07549292e-01 1.22148252e+00 -3.54153179e-02
-5.23257162e-03 -3.72954339e-01 -4.57022130e-01 1.24435965e-02
-2.17287111e+00 -1.26563561e+00 3.29497963e-01 2.21506238e+00
8.95132303e-01 3.78533274e-01 6.34339213e-01 -4.11108285e-01
4.50260907e-01 3.10439378e-01 -1.20935798e+00 -4.26816970e-01
-1.00982524e-01 -1.52491659e-01 8.99423122e-01 7.28948593e-01
-1.01012075e+00 1.37099743e+00 5.16168785e+00 7.33382285e-01
-9.43325996e-01 2.80788392e-01 4.27491665e-01 -6.87083900e-01
8.22770372e-02 1.10456176e-01 -7.18112171e-01 3.73260468e-01
8.31661224e-01 -2.57270932e-01 1.09718466e+00 7.63633192e-01
4.54806089e-01 -1.66960537e-01 -9.77144539e-01 4.48224604e-01
-2.73531348e-01 -9.53008711e-01 -4.20677662e-01 4.70505625e-01
6.29408121e-01 5.06497324e-01 2.81355262e-01 7.76034653e-01
1.57314122e+00 -1.06167650e+00 8.39093149e-01 6.37152344e-02
9.98416264e-03 -1.13693070e+00 3.72750401e-01 7.79439211e-01
-7.10489213e-01 -5.79355359e-01 -1.68694049e-01 -2.83744365e-01
-5.07907458e-02 2.22749207e-02 -6.72529936e-01 1.04592592e-01
5.42129099e-01 3.18559229e-01 -2.42875025e-01 3.48931134e-01
-7.24984884e-01 9.27443981e-01 -4.93405223e-01 -1.46519199e-01
8.49706829e-01 -1.90212369e-01 1.08116436e+00 4.34457511e-01
-2.63530850e-01 1.04994014e-01 7.23200738e-01 8.65582168e-01
-1.28412977e-01 -3.25197726e-01 -5.24152517e-01 -1.50666565e-01
5.91765940e-01 1.35844290e+00 -3.85542274e-01 -2.36890718e-01
2.57164724e-02 6.05059266e-01 4.84338790e-01 5.18438756e-01
-1.09095955e+00 2.01718748e-01 1.10275984e+00 -4.08906102e-01
1.91014424e-01 -1.66566998e-01 8.65302756e-02 -9.78685439e-01
-3.94072145e-01 -1.54694855e+00 4.02528703e-01 -2.41716638e-01
-1.29637945e+00 4.21600282e-01 -1.51806086e-01 -8.99243712e-01
-4.87859875e-01 -1.60206303e-01 -7.65732706e-01 4.73656207e-01
-1.23916769e+00 -8.18299472e-01 4.57664728e-01 7.14168549e-01
3.17140877e-01 -4.08314764e-01 4.78879064e-01 -3.05758089e-01
-9.84528005e-01 4.02794659e-01 5.41594699e-02 -2.88068317e-02
6.79847538e-01 -1.28649139e+00 3.25942039e-01 7.45462656e-01
-3.20386626e-02 4.80596423e-01 9.99172568e-01 -8.67418885e-01
-1.61235261e+00 -1.04499626e+00 -2.65358955e-01 -4.57989931e-01
1.09013939e+00 -3.05349886e-01 -7.44574368e-01 7.85559773e-01
3.78980160e-01 -2.14466989e-01 1.48274317e-01 -1.65312499e-01
-3.61735970e-01 1.69437617e-01 -1.12417161e+00 1.11855972e+00
6.23599052e-01 -1.31556720e-01 -2.79181182e-01 3.43509883e-01
9.14157033e-01 -2.62629569e-01 -3.84624660e-01 1.09758362e-01
2.48591796e-01 -9.32833433e-01 6.60679996e-01 -1.04801679e+00
9.01628658e-02 -3.39290529e-01 -2.06692982e-02 -1.70743275e+00
-1.33360416e-01 -1.07094789e+00 -3.93415540e-01 7.67849445e-01
6.55545443e-02 -9.86611426e-01 5.01268208e-01 1.25755385e-01
1.16710335e-01 -5.08367419e-01 -1.10525465e+00 -9.51447904e-01
4.62985456e-01 -5.97649403e-02 3.14226538e-01 7.86612153e-01
1.02297992e-01 2.27013990e-01 -7.36861289e-01 5.88808000e-01
8.80332768e-01 -3.10748219e-01 9.51756537e-01 -7.61229038e-01
-8.75921249e-01 -5.73821425e-01 2.85390913e-01 -6.27111852e-01
3.91152024e-01 -5.48012972e-01 -5.69632351e-02 -1.05781460e+00
2.69126654e-01 -3.63220960e-01 -6.11037970e-01 9.57938731e-01
-1.20520078e-01 -2.68758684e-01 6.45025194e-01 3.46337736e-01
-9.13178265e-01 6.76147223e-01 1.11380851e+00 6.13322761e-03
-4.66273874e-01 -1.13648260e-02 -8.11706960e-01 7.98557997e-01
1.02241373e+00 -7.76535690e-01 -4.19013500e-01 -4.51905243e-02
3.38323593e-01 1.77352637e-01 6.53149009e-01 -7.20590711e-01
2.45364338e-01 -7.29237258e-01 -2.04925284e-01 -1.08057506e-01
1.40987545e-01 -4.35668319e-01 -1.84074417e-01 1.24195361e+00
-6.39512360e-01 2.47835234e-01 5.07216081e-02 6.95526063e-01
4.95728523e-01 -2.70236237e-03 9.31624353e-01 -1.17730349e-01
-1.46308601e-01 1.74172729e-01 -8.86974633e-01 2.03718513e-01
1.31191516e+00 5.59036136e-01 -8.61490130e-01 -6.77963436e-01
-4.18254286e-01 8.28910649e-01 4.59311008e-01 2.01379761e-01
2.77773350e-01 -8.30754340e-01 -8.45780611e-01 -1.90260828e-01
-4.14724052e-01 -1.34209082e-01 8.23890194e-02 9.11877155e-01
-1.26825301e-02 1.04343563e-01 -3.45376402e-01 -6.42935708e-02
-9.22149718e-01 7.53382504e-01 7.42692888e-01 -7.46642232e-01
-5.92333615e-01 4.69972253e-01 4.91976738e-01 -4.24394578e-01
3.43670338e-01 -1.07535077e-02 -1.78623497e-01 3.90757136e-02
5.00754595e-01 5.08857191e-01 -5.46908796e-01 -1.62612289e-01
-4.85703528e-01 -1.18390955e-01 -5.31703115e-01 -4.99051243e-01
1.25730360e+00 1.92083985e-01 1.03787966e-01 2.55820364e-01
3.60983044e-01 1.84060007e-01 -1.77536547e+00 -1.55940771e-01
-2.39740476e-01 3.27366628e-02 2.56099110e-03 -1.14265811e+00
-9.77672458e-01 3.43939960e-01 2.27417782e-01 4.82120156e-01
8.28255653e-01 1.66720506e-02 5.00755906e-01 7.09280908e-01
6.67646468e-01 -1.02778208e+00 4.84917015e-01 7.80555367e-01
7.09643841e-01 -8.99808586e-01 -7.50795975e-02 4.90879625e-01
-1.00597286e+00 7.27915347e-01 9.96061862e-01 -5.15686035e-01
-1.32732661e-02 4.09752667e-01 -1.68093177e-03 -2.14611188e-01
-1.46054161e+00 -2.08991602e-01 -6.38494015e-01 5.05751193e-01
-4.55885679e-01 2.60406077e-01 -7.33062327e-02 5.73877811e-01
4.78605479e-02 -4.87614572e-01 9.90599930e-01 9.13891852e-01
-7.39874661e-01 -1.08841324e+00 -7.09249735e-01 2.81810760e-01
-4.40978587e-01 5.60631268e-02 -6.88182175e-01 8.62166166e-01
-1.56078547e-01 1.03357220e+00 -1.74912214e-01 -2.57491589e-01
1.98442131e-01 -2.53863037e-01 1.14153281e-01 -4.64306831e-01
-1.03425801e+00 3.10415238e-01 -9.79017019e-02 -8.03929925e-01
3.09793949e-02 -6.67230546e-01 -1.36961162e+00 -4.50099796e-01
1.29368471e-03 2.18765974e-01 2.54022896e-01 9.50101733e-01
1.64705917e-01 6.97927713e-01 9.33861911e-01 -6.52063668e-01
-1.31486440e+00 -8.89223814e-01 -6.34641647e-01 -1.16895922e-01
6.02292061e-01 -9.17054057e-01 -8.17316055e-01 -7.76794553e-01] | [3.986900806427002, 2.304965019226074] |
f7ba7079-d682-4610-81ff-2c0f230e5fd9 | synthtiger-synthetic-text-image-generator | 2107.09313 | null | https://arxiv.org/abs/2107.09313v1 | https://arxiv.org/pdf/2107.09313v1.pdf | SynthTIGER: Synthetic Text Image GEneratoR Towards Better Text Recognition Models | For successful scene text recognition (STR) models, synthetic text image generators have alleviated the lack of annotated text images from the real world. Specifically, they generate multiple text images with diverse backgrounds, font styles, and text shapes and enable STR models to learn visual patterns that might not be accessible from manually annotated data. In this paper, we introduce a new synthetic text image generator, SynthTIGER, by analyzing techniques used for text image synthesis and integrating effective ones under a single algorithm. Moreover, we propose two techniques that alleviate the long-tail problem in length and character distributions of training data. In our experiments, SynthTIGER achieves better STR performance than the combination of synthetic datasets, MJSynth (MJ) and SynthText (ST). Our ablation study demonstrates the benefits of using sub-components of SynthTIGER and the guideline on generating synthetic text images for STR models. Our implementation is publicly available at https://github.com/clovaai/synthtiger. | ['Sungrae Park', 'Han-Cheol Cho', 'Yoonsik Kim', 'Moonbin Yim'] | 2021-07-20 | null | null | null | null | ['scene-text-recognition'] | ['computer-vision'] | [ 5.13612032e-01 -3.45830396e-02 2.18920350e-01 -1.71122581e-01
-7.21861601e-01 -8.13950837e-01 8.53095591e-01 -3.55858296e-01
-1.22704118e-01 5.53825438e-01 1.38689429e-01 -4.03971732e-01
3.90474230e-01 -4.88627881e-01 -8.63081455e-01 -3.41309547e-01
7.91251719e-01 4.09919769e-01 3.14803988e-01 6.32372946e-02
3.21669847e-01 3.23512912e-01 -1.60143983e+00 7.18058348e-01
1.02507055e+00 5.22296131e-01 4.28293586e-01 1.04515719e+00
-3.86029184e-01 9.49590981e-01 -9.37162042e-01 -5.89882135e-01
3.58587563e-01 -6.54337466e-01 -3.95962238e-01 4.64411080e-01
6.65912449e-01 -4.76410925e-01 -4.39667314e-01 7.43093729e-01
3.80221784e-01 -2.43638933e-01 8.61584663e-01 -1.41685510e+00
-7.94813335e-01 8.18961680e-01 -7.58857965e-01 -2.73669600e-01
2.33016014e-01 5.14140964e-01 7.08469033e-01 -1.03430820e+00
8.67145002e-01 1.22377193e+00 3.64849538e-01 6.06010616e-01
-1.44642222e+00 -5.54783702e-01 -6.12634383e-02 -2.76647419e-01
-1.33815527e+00 -6.73448861e-01 6.62262797e-01 -5.67544281e-01
7.14154899e-01 4.77701783e-01 6.00158334e-01 1.64863980e+00
-1.45201042e-01 1.34509635e+00 1.07169271e+00 -7.65106499e-01
1.04854308e-01 3.05584699e-01 -2.28688017e-01 5.72990596e-01
4.26170528e-01 -1.95671901e-01 -6.62671447e-01 4.78560962e-02
1.00351822e+00 -2.57952332e-01 -1.74973711e-01 -5.44266701e-01
-1.41118598e+00 5.35505176e-01 -3.46231014e-01 1.43472508e-01
1.67554751e-01 1.73795834e-01 2.14382097e-01 2.66223639e-01
3.68709683e-01 5.14997125e-01 -1.43243611e-01 -1.71934083e-01
-1.06926835e+00 3.15990180e-01 7.66470194e-01 1.36980283e+00
5.53561747e-01 4.12418753e-01 -4.28360730e-01 9.97204661e-01
-3.40505950e-02 8.46939027e-01 5.16746044e-01 -8.28501284e-01
6.36603773e-01 5.56372464e-01 2.05844581e-01 -7.18440711e-01
7.48425536e-03 4.70222868e-02 -5.25780678e-01 1.45090193e-01
7.55036652e-01 -2.84285188e-01 -1.15930712e+00 1.27414560e+00
-4.99828458e-02 -2.32924238e-01 -5.47379702e-02 5.10017037e-01
6.48517907e-01 6.34299338e-01 -1.38056412e-01 3.15221578e-01
1.02134526e+00 -1.17932117e+00 -4.17506695e-01 -4.13472325e-01
8.54590595e-01 -1.15611529e+00 1.65908003e+00 4.08293843e-01
-1.02309597e+00 -3.80208552e-01 -9.26930189e-01 3.83827370e-03
-2.97094822e-01 8.10649574e-01 3.80573303e-01 7.79693067e-01
-1.06704903e+00 2.22781822e-01 -7.02214122e-01 -3.48910183e-01
3.78036320e-01 -9.77409258e-02 -6.37596920e-02 2.25464115e-04
-6.41995728e-01 5.30055821e-01 3.12970251e-01 -1.40940085e-01
-7.53826439e-01 -5.20990252e-01 -7.88647413e-01 -1.30975291e-01
5.30524731e-01 -6.82303190e-01 1.57631707e+00 -1.34043896e+00
-1.58388901e+00 8.89637589e-01 -1.73641562e-01 -2.17834681e-01
1.02123523e+00 -1.94638610e-01 -1.36538774e-01 2.38588661e-01
-1.84352882e-02 9.93435919e-01 1.28063631e+00 -1.39615369e+00
-2.33057559e-01 2.55220700e-02 -4.33185697e-01 2.11890310e-01
-2.26290271e-01 -3.98749821e-02 -7.30509520e-01 -1.10566270e+00
-2.32995167e-01 -7.22708702e-01 -9.66337249e-02 1.09584622e-01
-8.21633041e-01 3.04078549e-01 8.98027837e-01 -5.24208009e-01
9.14017975e-01 -2.12695169e+00 -1.45278364e-01 6.05990216e-02
1.44941568e-01 2.99371272e-01 -4.64066088e-01 7.58430421e-01
3.82722286e-03 4.71900165e-01 -1.08250119e-01 -4.90034997e-01
9.78431106e-02 3.02138403e-02 -5.96139908e-01 3.03219110e-02
2.08463773e-01 1.03967738e+00 -5.45215309e-01 -6.57926798e-01
3.53799909e-01 2.45465040e-01 -4.33612078e-01 -2.59950627e-02
-8.12796891e-01 3.15195709e-01 -5.22104383e-01 5.89534760e-01
5.83900690e-01 -3.97483498e-01 9.19628069e-02 8.02360997e-02
-9.56653729e-02 8.99750292e-02 -1.19127727e+00 1.57219255e+00
-2.89456278e-01 9.86180663e-01 -2.29828328e-01 -3.52483064e-01
9.50571239e-01 -2.17199326e-02 1.81503028e-01 -6.89998925e-01
1.71891674e-01 6.08199164e-02 -3.08189362e-01 -4.28138971e-01
8.99299145e-01 3.04263562e-01 9.95839536e-02 9.21319962e-01
-2.34651297e-01 -5.68748355e-01 6.76970899e-01 4.78723913e-01
8.88104379e-01 4.34713781e-01 -3.39641534e-02 1.25496209e-01
5.04681617e-02 1.90680176e-01 -2.23886082e-03 1.13609445e+00
2.34343886e-01 1.16969788e+00 7.47162163e-01 -2.82596618e-01
-1.55609190e+00 -1.06524646e+00 5.78650385e-02 8.09650898e-01
-1.05607599e-01 -5.33556879e-01 -9.63269234e-01 -5.35950601e-01
-6.89407140e-02 9.14761126e-01 -5.26491046e-01 2.63082623e-01
-4.41756427e-01 -6.03438973e-01 9.37337339e-01 4.35118735e-01
5.00626624e-01 -1.00906754e+00 -6.85063124e-01 -2.56663233e-01
-1.66958094e-01 -1.28280890e+00 -8.29602301e-01 -2.02748761e-01
-5.57458162e-01 -8.02596986e-01 -9.45132673e-01 -5.79061568e-01
8.75296175e-01 3.98810208e-01 1.09611690e+00 1.63092285e-01
-5.80555320e-01 4.55979586e-01 -6.03228033e-01 -4.75683272e-01
-8.48634601e-01 3.87974381e-02 -4.58047569e-01 1.35881796e-01
1.06007012e-03 -1.98528871e-01 -5.08229494e-01 4.57040727e-01
-1.19781220e+00 9.38729107e-01 6.29275441e-01 7.11142838e-01
4.06688750e-01 -2.36053184e-01 1.96981087e-01 -1.04267097e+00
6.78534567e-01 1.00489743e-01 -6.78354979e-01 3.82115036e-01
-4.04016584e-01 9.97429490e-02 5.99596322e-01 -7.36120522e-01
-1.21489978e+00 2.15513751e-01 1.90581113e-01 -4.67439711e-01
-4.10938382e-01 9.84067097e-03 1.50591195e-01 2.15723336e-01
8.71640503e-01 6.31286800e-01 3.93577851e-03 -3.66931021e-01
5.20684838e-01 8.26939940e-01 3.36513728e-01 -8.01508009e-01
9.18121696e-01 3.81734073e-01 -4.67363477e-01 -1.06222689e+00
-4.29856360e-01 7.34610334e-02 -5.24221599e-01 -3.88174020e-02
5.65335453e-01 -8.06200147e-01 -2.25008458e-01 8.63172114e-01
-1.04300570e+00 -9.34308290e-01 -4.97974485e-01 1.42331466e-01
-6.62736058e-01 6.66040003e-01 -4.67379630e-01 -8.23069751e-01
-2.07551852e-01 -1.13215256e+00 1.46227217e+00 1.66348755e-01
-2.42570147e-01 -7.53621280e-01 -1.77620068e-01 2.75773317e-01
3.32520485e-01 1.76655322e-01 7.74882138e-01 -5.14960051e-01
-8.67982090e-01 -6.51709437e-02 -3.85736376e-01 1.38711914e-01
4.45042439e-02 5.05592883e-01 -8.84295404e-01 -1.33781269e-01
-4.16193098e-01 -4.87640858e-01 7.29163051e-01 3.18635046e-01
1.29406297e+00 -3.88533413e-01 -1.23539217e-01 5.81793666e-01
1.21752477e+00 1.87347189e-01 7.40398288e-01 3.32150906e-01
7.83840060e-01 4.49516594e-01 2.94966400e-01 4.78984892e-01
2.59550810e-01 6.29346073e-01 -9.62718129e-02 -3.42064649e-01
-4.18430954e-01 -6.23310924e-01 4.28371936e-01 4.80138332e-01
4.03357208e-01 -6.92328632e-01 -1.06893647e+00 4.44103986e-01
-1.72855294e+00 -9.46282327e-01 -1.87334031e-01 2.14352512e+00
9.17592466e-01 1.51856810e-01 1.34813160e-01 5.39979041e-02
7.07465887e-01 6.57172352e-02 -5.85744321e-01 -3.19760501e-01
-5.16101420e-01 3.33065465e-02 4.88059729e-01 2.80565113e-01
-8.74969304e-01 1.19603062e+00 6.52743959e+00 1.01510036e+00
-1.06036973e+00 -3.53984088e-01 7.84173667e-01 -1.22659080e-01
-5.67773819e-01 -1.01730390e-03 -8.04929852e-01 4.60140318e-01
5.14067113e-01 -3.43398303e-01 4.78028983e-01 7.17143536e-01
3.77871364e-01 -3.46963435e-01 -1.10740483e+00 1.08333814e+00
2.75361717e-01 -1.46262884e+00 6.14650309e-01 -1.86998963e-01
8.98604870e-01 -1.49959370e-01 1.72895908e-01 -5.77256233e-02
5.58082402e-01 -1.02843320e+00 1.03306842e+00 3.56213689e-01
1.16092205e+00 -2.93351442e-01 1.09527290e-01 2.88696259e-01
-8.30839992e-01 1.77795291e-01 -2.66627133e-01 2.72327602e-01
-9.78522748e-02 4.09424424e-01 -1.22532976e+00 4.16260868e-01
4.00458336e-01 7.03741789e-01 -1.23428559e+00 8.35472286e-01
-2.71478504e-01 6.42383575e-01 -2.88067967e-01 -2.55060911e-01
-3.32149789e-02 -2.27687627e-01 4.06777382e-01 1.31918836e+00
3.42656821e-01 -4.00457710e-01 4.21401672e-02 1.23929620e+00
-1.49879148e-02 2.54479945e-01 -9.13341522e-01 -4.36005920e-01
5.21988273e-01 1.09705186e+00 -1.02581143e+00 -3.75715286e-01
-3.37218225e-01 1.05790758e+00 -7.34736770e-02 6.23280823e-01
-6.46144986e-01 -4.57541168e-01 4.45460714e-02 2.31474549e-01
3.16436529e-01 -3.00100595e-01 -8.24808955e-01 -1.46150064e+00
1.58405647e-01 -1.44434512e+00 1.36008412e-01 -1.44063175e+00
-8.82592857e-01 5.40478945e-01 1.09597519e-01 -1.29356861e+00
-2.95956135e-01 -6.80795312e-01 -5.56741118e-01 7.64432013e-01
-9.92475510e-01 -1.35071707e+00 -3.13040048e-01 5.29587567e-01
1.01366901e+00 -1.80383921e-01 5.26795566e-01 -1.92619309e-01
-6.88521266e-01 7.87140250e-01 2.70699412e-01 2.65579045e-01
7.23142743e-01 -1.19008172e+00 1.01758349e+00 1.05752778e+00
3.32074761e-01 4.30632651e-01 6.61219418e-01 -7.14266479e-01
-1.47110260e+00 -9.15113330e-01 3.67432773e-01 -6.32132709e-01
5.78224838e-01 -7.38417327e-01 -6.39482558e-01 7.22534418e-01
3.05084974e-01 -6.30787373e-01 2.85493106e-01 -4.14511025e-01
-4.90589112e-01 1.87027574e-01 -8.55777025e-01 1.40161204e+00
9.26498711e-01 -2.86218613e-01 -1.70507237e-01 2.65546858e-01
5.49392641e-01 -5.36488891e-01 -3.89829576e-01 -4.15345132e-02
5.50411463e-01 -1.10530496e+00 7.35549092e-01 -1.12200715e-01
7.11441875e-01 -3.80589426e-01 6.50545806e-02 -1.06129205e+00
3.23224813e-01 -8.77118289e-01 1.52794883e-01 1.08955038e+00
7.00714409e-01 -5.54161370e-01 7.80764997e-01 5.67140698e-01
8.08096863e-03 -2.65436143e-01 -3.48520517e-01 -7.68971860e-01
1.02844335e-01 -3.73747289e-01 5.26246190e-01 7.98893392e-01
-1.59609184e-01 2.42649212e-01 -5.43522894e-01 -3.19148690e-01
5.45052052e-01 1.52812049e-01 1.35711551e+00 -7.43169487e-01
-5.10631382e-01 -5.11368394e-01 3.56148626e-03 -1.27092218e+00
-2.24373922e-01 -8.00628483e-01 1.53048402e-02 -1.53988469e+00
3.62036884e-01 -3.45910221e-01 5.45091808e-01 6.18049741e-01
-9.52045396e-02 3.74058306e-01 4.34836924e-01 3.27862859e-01
-4.37253088e-01 4.68546063e-01 1.64628446e+00 -1.70645490e-01
-1.96207270e-01 -2.75558144e-01 -6.96553111e-01 6.04219675e-01
9.88556087e-01 -4.50504184e-01 -5.81481159e-01 -8.69590998e-01
2.60601193e-01 3.37531120e-02 3.80995601e-01 -8.43474388e-01
4.17050608e-02 -2.31593654e-01 6.76614106e-01 -7.34400272e-01
2.39613786e-01 -2.89390802e-01 2.83580482e-01 1.37265965e-01
-4.92596924e-01 1.65427670e-01 2.84774303e-01 2.46144280e-01
1.08952731e-01 -3.47221941e-01 7.12113202e-01 -1.92122355e-01
-2.62277752e-01 -4.60893363e-02 -6.08855426e-01 2.20526770e-01
7.67794311e-01 -5.67040563e-01 -8.08899522e-01 -5.36230087e-01
-3.01027685e-01 1.48223966e-01 1.02034056e+00 5.59707165e-01
7.28673100e-01 -9.53582942e-01 -7.48088300e-01 3.72393012e-01
1.85779884e-01 6.04533311e-03 8.30894262e-02 5.55679500e-01
-8.37835312e-01 1.53165728e-01 -1.78153425e-01 -6.74964488e-01
-1.25651765e+00 4.07540977e-01 2.22675160e-01 -1.39551163e-01
-9.39474463e-01 3.36312026e-01 5.10557175e-01 -3.57733637e-01
2.30144814e-01 -2.81013370e-01 2.47005552e-01 -2.93758124e-01
5.32919526e-01 1.33055389e-01 -5.33330254e-02 -2.74015993e-01
2.87657559e-01 4.99596953e-01 -2.87348956e-01 -4.37784880e-01
9.89549816e-01 -1.83835149e-01 2.92109728e-01 3.54656547e-01
4.99768347e-01 2.20097259e-01 -1.43811202e+00 -8.04749224e-03
-2.44387779e-02 -6.19285524e-01 -4.43422556e-01 -9.51692343e-01
-7.80618429e-01 7.68974245e-01 2.31449828e-01 1.46519274e-01
1.03055954e+00 -1.84301242e-01 5.60184360e-01 5.81330538e-01
6.81923330e-02 -1.05561781e+00 4.68016624e-01 4.35850322e-01
9.92858589e-01 -8.35862398e-01 -1.27518564e-01 -4.87258881e-01
-1.05832434e+00 1.14289379e+00 6.68755889e-01 2.85329223e-01
8.79917759e-03 6.25597715e-01 3.12703460e-01 6.39040843e-02
-9.07072306e-01 -1.41664729e-01 -6.77676722e-02 6.66102648e-01
3.44891340e-01 -1.31302044e-01 -5.48450574e-02 1.01410270e-01
-4.30075496e-01 5.68153476e-03 1.05086195e+00 1.04878962e+00
-2.07063586e-01 -1.27982044e+00 -5.37538409e-01 5.37371159e-01
-3.72164279e-01 -2.85196066e-01 -8.72040451e-01 8.18353415e-01
-4.28015918e-01 8.21793556e-01 -6.73992932e-03 -2.67086804e-01
2.17028052e-01 1.74405456e-01 5.49137056e-01 -5.55879891e-01
-2.59412467e-01 3.55488032e-01 2.93499589e-01 -6.23549372e-02
2.26489380e-02 -6.72290385e-01 -9.01893556e-01 -1.91944256e-01
-8.68745446e-02 -2.11100832e-01 6.44973457e-01 5.65034211e-01
5.00931084e-01 3.14799458e-01 6.25798762e-01 -8.52730870e-01
-4.24797207e-01 -9.01776254e-01 -3.83991569e-01 4.48130608e-01
-1.35478098e-03 -2.63796508e-01 -2.31587231e-01 6.96771622e-01] | [11.467917442321777, 0.023084156215190887] |
4fdd17d0-40ea-4406-aa1e-b5c4e69cd589 | dish-detection-in-food-platters-a-framework | 2305.07552 | null | https://arxiv.org/abs/2305.07552v1 | https://arxiv.org/pdf/2305.07552v1.pdf | Dish detection in food platters: A framework for automated diet logging and nutrition management | Diet is central to the epidemic of lifestyle disorders. Accurate and effortless diet logging is one of the significant bottlenecks for effective diet management and calorie restriction. Dish detection from food platters is a challenging problem due to a visually complex food layout. We present an end-to-end computational framework for diet management, from data compilation, annotation, and state-of-the-art model identification to its mobile app implementation. As a case study, we implement the framework in the context of Indian food platters known for their complex presentation that poses a challenge for the automated detection of dishes. Starting with the 61 most popular Indian dishes, we identify the state-of-the-art model through a comparative analysis of deep-learning-based object detection architectures. Rooted in a meticulous compilation of 68,005 platter images with 134,814 manual dish annotations, we first compare ten architectures for multi-label classification to identify ResNet152 (mAP=84.51%) as the best model. YOLOv8x (mAP=87.70%) emerged as the best model architecture for dish detection among the eight deep-learning models implemented after a thorough performance evaluation. By comparing with the state-of-the-art model for the IndianFood10 dataset, we demonstrate the superior object detection performance of YOLOv8x for this subset and establish Resnet152 as the best architecture for multi-label classification. The models thus trained on richly annotated data can be extended to include dishes from across global cuisines. The proposed framework is demonstrated through a proof-of-concept mobile application with diverse applications for diet logging, food recommendation systems, nutritional interventions, and mitigation of lifestyle disorders. | ['Ganesh Bagler', 'Kirti Vashishtha', 'Abhuday Tiwari', 'Nikhilesh Verhwani', 'Nitesh Narwade', 'Samiksha Garg', 'Astha Jain', 'Meenal Jain', 'Ronak Chhajed', 'Ekta Gambhir', 'Hareesh Amuru', 'Nikhila Vishnumolakala', 'Anushka Gupta', 'Pratik Chauhan', 'Nidhi Verma', 'Shounak Ghatak', 'Shashank Dargar', 'Mansi Goel'] | 2023-05-12 | null | null | null | null | ['food-recommendation'] | ['miscellaneous'] | [ 1.34142801e-01 -2.54499286e-01 -2.61762589e-01 5.13290428e-02
-5.94509363e-01 -5.77983081e-01 -1.42329544e-01 9.62465286e-01
-3.24534953e-01 9.10040811e-02 2.04268530e-01 -2.19592210e-02
4.38915677e-02 -7.88568556e-01 -8.76733661e-01 -4.18764293e-01
-4.29112047e-01 5.53918958e-01 -2.25466101e-05 -2.49732465e-01
-2.88993925e-01 -1.48073032e-01 -1.55293465e+00 5.07954001e-01
7.63726771e-01 1.05645752e+00 6.17878258e-01 8.32935154e-01
3.72551605e-02 7.67977715e-01 -3.86371195e-01 -3.35641384e-01
1.97047353e-01 -2.25328609e-01 -1.84114337e-01 3.12850960e-02
7.86583662e-01 -4.49733168e-01 2.93230414e-01 1.33940780e+00
7.26555586e-01 -4.21152622e-01 5.56202710e-01 -9.35299575e-01
-9.76034760e-01 9.15648878e-01 -4.49003786e-01 -2.68211700e-02
3.19959491e-01 1.59219697e-01 8.94473195e-01 -6.90601766e-01
4.89280641e-01 1.04335606e+00 1.32611132e+00 2.58040041e-01
-1.25361645e+00 -7.44328082e-01 1.96857557e-01 2.43605956e-01
-1.43840849e+00 -2.46423930e-01 4.13098454e-01 -6.63923025e-01
9.22716558e-01 2.18010843e-01 1.19355047e+00 9.41940725e-01
2.48743352e-02 8.00728083e-01 1.10355818e+00 -5.45309007e-01
3.37132901e-01 2.48947144e-01 3.08827043e-01 1.01777017e+00
8.37189257e-01 -1.06386602e-01 -1.56064078e-01 1.35364324e-01
3.66833806e-01 -3.49809825e-02 1.09288521e-01 -4.15559620e-01
-9.83955562e-01 9.84670222e-01 6.54481113e-01 1.54425785e-01
-4.39805597e-01 9.12484080e-02 5.66316068e-01 5.88572435e-02
6.27801180e-01 3.65719020e-01 -5.71188390e-01 6.01253033e-01
-9.18858051e-01 1.92002103e-01 9.08659577e-01 9.45140839e-01
2.75722593e-01 9.10255983e-02 -1.26821518e-01 7.79486358e-01
6.05657637e-01 9.52829361e-01 1.76414531e-02 -3.98472130e-01
9.19491351e-02 8.91228735e-01 2.38036111e-01 -1.05141807e+00
-1.11620343e+00 -5.99698424e-01 -8.53010535e-01 4.37179118e-01
7.08276331e-01 -1.69489175e-01 -8.87370288e-01 1.26735806e+00
5.64057171e-01 -4.56403941e-01 -4.97471541e-02 7.48980522e-01
1.29960632e+00 2.73084849e-01 7.78540313e-01 1.96353883e-01
2.05440259e+00 -1.09010494e+00 -5.62871218e-01 -1.91472009e-01
8.81211698e-01 -8.29490960e-01 9.45947349e-01 7.06997037e-01
-7.00296581e-01 -6.98795497e-01 -1.46174121e+00 -1.19769447e-01
-8.49275708e-01 7.82497644e-01 4.93266314e-01 9.23525214e-01
-9.06473696e-01 3.00563872e-01 -5.98519206e-01 -8.93781245e-01
3.13400239e-01 3.94441098e-01 -1.05625942e-01 -1.41955927e-01
-7.30223894e-01 8.16745341e-01 4.08968210e-01 -6.96045533e-02
-8.23353350e-01 -8.90246034e-01 -8.10607016e-01 4.40961197e-02
4.55934376e-01 -6.49082184e-01 1.19044983e+00 -1.06202376e+00
-1.20091021e+00 1.12296152e+00 6.98103487e-01 -6.36910319e-01
4.20866340e-01 -2.83297092e-01 -7.39579439e-01 2.12305591e-01
2.37611160e-01 6.91827059e-01 5.07927418e-01 -1.13704860e+00
-8.57278109e-01 -3.02267432e-01 1.30878583e-01 4.53156531e-02
-1.05337597e-01 2.04621851e-01 1.22642823e-01 -5.81445754e-01
-2.69883543e-01 -1.01450574e+00 -2.09676728e-01 3.25657964e-01
-5.25021374e-01 4.11128625e-02 3.10589671e-01 -9.39641118e-01
1.05916488e+00 -2.01502275e+00 -3.24878514e-01 -9.59825739e-02
3.63714397e-01 3.39439183e-01 1.25326872e-01 9.28563252e-02
2.94345487e-02 -8.56652707e-02 2.39028618e-01 -3.10737312e-01
3.62331301e-01 -2.11697638e-01 6.43915057e-01 8.28206837e-01
-1.91499129e-01 6.92584276e-01 -9.80168223e-01 -4.45838571e-01
6.96870565e-01 5.93636930e-01 -5.99137723e-01 4.50260341e-02
-7.55962372e-01 1.08269222e-01 -1.16530389e-01 1.00460386e+00
7.68343866e-01 -5.27137041e-01 7.02197254e-01 -8.78446460e-01
-3.73082817e-01 -2.82736659e-01 -1.34344220e+00 1.83541656e+00
-1.50752902e-01 -2.84817703e-02 3.55847865e-01 -5.71127355e-01
7.07836330e-01 1.84987128e-01 7.02105165e-01 -8.07565093e-01
4.29104954e-01 1.37891650e-01 1.59733966e-02 -8.18786383e-01
4.51208591e-01 2.20089555e-01 -1.86195001e-01 5.96170500e-03
1.06132932e-01 6.62425041e-01 6.19822264e-01 -1.20309532e-01
8.49771261e-01 3.36088419e-01 8.60112607e-01 -8.47309291e-01
1.50131330e-01 6.37248814e-01 1.36725873e-01 6.66930556e-01
-5.18494427e-01 3.17358315e-01 -1.72778606e-01 -8.83092046e-01
-1.06929481e+00 -8.31752479e-01 6.31345510e-02 1.58169055e+00
-1.56979412e-01 -4.21433181e-01 -8.34389746e-01 -8.00082326e-01
1.39871582e-01 4.75577801e-01 -7.85557866e-01 2.52755880e-01
-5.05736530e-01 -9.70914543e-01 4.81369764e-01 4.91299331e-01
8.01668465e-01 -1.00170600e+00 -9.59389925e-01 5.27470827e-01
-4.10085656e-02 -9.82428491e-01 -5.18770933e-01 4.44197655e-01
-4.40413430e-02 -1.63431132e+00 -7.99137294e-01 -1.09365857e+00
8.04008991e-02 3.84698212e-01 1.56952393e+00 2.92600989e-02
-6.11113071e-01 3.71568382e-01 -5.14038622e-01 -6.03989780e-01
-8.72333169e-01 1.37072623e-01 -4.41421196e-02 -4.08943236e-01
7.42385447e-01 -9.32544395e-02 -1.30478704e+00 1.68609291e-01
-5.48858881e-01 1.16732188e-01 3.23265076e-01 1.68857679e-01
7.75242686e-01 -1.14968300e-01 4.06275719e-01 -8.57537150e-01
-1.37883067e-01 -7.42675960e-01 -7.61457145e-01 3.35501373e-01
-5.84315538e-01 -4.40512925e-01 5.97884119e-01 -6.49373114e-01
-5.70966661e-01 6.04247987e-01 -5.02494216e-01 6.49167076e-02
-6.98918998e-01 1.67618781e-01 -8.88181850e-02 -2.90890317e-02
9.06509101e-01 -2.56506860e-01 -2.76215762e-01 -9.26008701e-01
5.48871934e-01 3.60732108e-01 2.87473172e-01 -1.22585319e-01
9.47813224e-03 3.49910676e-01 -1.99567795e-01 -1.12652886e+00
-9.65010226e-01 -7.07519472e-01 -4.53777581e-01 -2.71185160e-01
1.50765812e+00 -1.36216962e+00 -9.94692385e-01 5.90764105e-01
-5.75629950e-01 -6.21720433e-01 4.78467420e-02 3.56861174e-01
-2.24199221e-01 1.55270353e-01 -6.08720481e-01 -7.71481395e-01
-9.56575334e-01 -1.02525401e+00 1.09052300e+00 -4.05939519e-02
-3.31731647e-01 -8.69825244e-01 1.78745866e-01 1.89292058e-01
3.21407139e-01 5.39236009e-01 1.06378353e+00 -6.19273603e-01
-1.45398071e-02 7.64316469e-02 -4.18191552e-01 3.26444879e-02
8.54275003e-02 -2.58917212e-01 -9.50556517e-01 -2.95705527e-01
-3.90622824e-01 -1.81435883e-01 8.00230980e-01 8.15382838e-01
5.04923820e-01 -1.20069101e-01 -3.16158801e-01 5.69166064e-01
1.88813782e+00 1.96295425e-01 2.38151774e-01 5.68943858e-01
1.02617097e+00 9.48986486e-02 2.95944959e-01 3.18463743e-01
9.27656174e-01 8.16542864e-01 8.00537467e-01 -7.01227486e-01
-9.45737064e-01 -1.76895812e-01 4.82705027e-01 5.12179375e-01
2.98347145e-01 -3.08150113e-01 -4.74999368e-01 4.07585114e-01
-1.59539843e+00 -6.50245368e-01 -5.14844120e-01 1.94963253e+00
4.05733466e-01 -1.84663773e-01 9.77363646e-01 1.67648464e-01
5.88516653e-01 -2.56209284e-01 -4.87020284e-01 -1.55011835e-02
-9.99165848e-02 3.96053791e-02 9.83063698e-01 -2.92119514e-02
-1.73711383e+00 6.19136333e-01 6.26152563e+00 4.26345438e-01
-7.48602986e-01 4.85854715e-01 3.33767831e-01 -4.37094644e-03
6.81674480e-01 -9.89803791e-01 -1.11150885e+00 5.33497632e-01
8.90096486e-01 8.08511019e-01 4.15069133e-01 1.01030099e+00
1.30590469e-01 -1.28857821e-01 -1.08566928e+00 8.38883102e-01
2.56258845e-01 -1.18088913e+00 -2.99951166e-01 -1.49466157e-01
5.68802297e-01 3.36583734e-01 -2.87487395e-02 3.66031498e-01
7.98943996e-01 -6.70115471e-01 1.30730486e+00 -4.06342521e-02
8.95283103e-01 -5.07079542e-01 3.48597914e-01 -2.21305713e-01
-1.71993017e+00 -3.61373872e-01 -5.09830832e-01 6.56492040e-02
-1.34082451e-01 5.54996014e-01 -7.41229653e-01 3.10260266e-01
9.57042277e-01 6.49605513e-01 -7.22692609e-01 1.30435157e+00
2.20652774e-01 5.08012354e-01 -6.26790285e-01 -1.09316044e-01
1.32687747e-01 1.25667796e-01 1.55025512e-01 1.79708171e+00
4.92731035e-01 -1.34470433e-01 9.44020092e-01 3.65831137e-01
-8.56526196e-02 7.35405505e-01 -4.25918937e-01 2.20851123e-01
-1.78734750e-01 1.35145080e+00 -1.23848426e+00 -3.54100406e-01
-6.67429864e-01 8.79995465e-01 2.12139145e-01 -6.37785792e-02
-9.98762429e-01 4.25001442e-01 6.46899283e-01 2.72332877e-01
6.60810709e-01 6.51089922e-02 -1.74204092e-02 -7.66239107e-01
-4.68612969e-01 -1.17249751e+00 8.07841301e-01 -6.02158844e-01
-1.24245274e+00 3.12256724e-01 -1.89366624e-01 -9.17336524e-01
4.28439170e-01 -8.74209702e-01 -3.54306176e-02 2.93254822e-01
-1.48190665e+00 -1.83906651e+00 -5.58125496e-01 4.03388739e-01
7.85174966e-01 1.85066462e-01 1.23379660e+00 9.17451084e-01
-6.55063748e-01 6.24783218e-01 2.97878087e-01 7.02733500e-03
4.16130692e-01 -1.56978118e+00 2.33783796e-01 4.04703051e-01
2.66254693e-02 -1.67824671e-01 7.71202743e-01 -1.00758898e+00
-1.53084707e+00 -1.65565050e+00 3.73087704e-01 -2.26052001e-01
4.82557356e-01 -5.79314947e-01 -3.90743315e-01 4.72696006e-01
2.13906258e-01 -3.58286768e-01 7.97340631e-01 -1.76961627e-02
-4.15532142e-01 -1.59232020e-01 -1.39314520e+00 3.51906180e-01
1.01155210e+00 9.83134508e-02 1.53378949e-01 8.00438285e-01
6.01493180e-01 -3.62361610e-01 -9.64056909e-01 6.51888996e-02
1.00920308e+00 -8.12273622e-01 1.25907111e+00 -1.15160652e-01
2.32135504e-01 -3.02898943e-01 -4.89316404e-01 -1.19233608e+00
-7.36713767e-01 -6.47087991e-02 -1.56820975e-02 1.01854908e+00
3.25652808e-01 -6.78925682e-03 7.19204605e-01 -1.37764946e-01
-3.70670706e-01 -2.50181019e-01 -3.72625411e-01 -4.77260560e-01
-1.47128580e-02 -1.70975417e-01 5.74925482e-01 9.33187902e-01
-2.46495247e-01 2.89682984e-01 -6.57533765e-01 4.60597098e-01
7.17453182e-01 1.49907127e-01 6.25420749e-01 -1.13846064e+00
-4.06925589e-01 -3.37190002e-01 -2.25680292e-01 -6.44181669e-01
-5.91881692e-01 -1.02371442e+00 3.30696613e-01 -1.79159939e+00
3.41083944e-01 -4.52843219e-01 -2.61831105e-01 4.89692152e-01
2.56397203e-02 7.32163191e-01 4.04849857e-01 -1.49080962e-01
-8.75083506e-01 -3.11076969e-01 8.69332671e-01 -5.54553211e-01
-3.17162007e-01 -2.30729550e-01 -7.09690928e-01 9.83470082e-01
8.07447016e-01 -5.58445811e-01 -3.62593494e-02 -3.33137870e-01
4.03205335e-01 -5.77875376e-01 4.13927138e-01 -1.42711234e+00
-2.04835027e-01 2.85841256e-01 6.07521653e-01 -5.57804227e-01
1.93281725e-01 -1.23003888e+00 6.49331510e-01 1.02491331e+00
-1.47317946e-01 -5.82645983e-02 3.42123598e-01 4.26151931e-01
8.28214347e-01 4.79037501e-03 7.38521636e-01 -5.50056279e-01
-9.23230112e-01 -3.40125374e-02 -4.32570219e-01 -2.82418519e-01
8.31170440e-01 3.91772240e-02 -4.99519020e-01 3.49461704e-01
-1.12054586e+00 -5.01445075e-03 3.65511805e-01 2.71734864e-01
6.27516061e-02 -1.14164007e+00 -9.29873407e-01 2.06593156e-01
3.59494895e-01 -6.78072274e-01 4.38332111e-01 6.31805837e-01
-8.00903380e-01 2.06234857e-01 -2.48138487e-01 -6.62655294e-01
-1.52522278e+00 1.03183532e+00 4.02789533e-01 -5.46030581e-01
-1.05858505e+00 3.63692492e-01 4.02430087e-01 -1.71673790e-01
5.66068888e-01 -1.07329214e+00 -5.26603639e-01 1.90499291e-01
8.25499237e-01 6.91612363e-01 2.79824495e-01 -5.15678585e-01
-2.62373358e-01 5.72997391e-01 2.01454952e-01 9.42955017e-01
1.16934299e+00 -3.34247142e-01 4.51428175e-01 2.05162153e-01
7.60732591e-01 -3.19786429e-01 -1.01920497e+00 1.34849131e-01
-2.43088856e-01 1.63909271e-01 8.51463303e-02 -1.47149670e+00
-1.16844702e+00 4.47956562e-01 1.63347387e+00 4.75161314e-01
9.73448813e-01 -1.13007925e-01 7.71222055e-01 3.04787923e-02
3.04308772e-01 -8.19180727e-01 -1.28408626e-01 2.51728892e-02
5.79763830e-01 -1.51269257e+00 3.08780782e-02 -5.43588042e-01
-2.20588043e-01 8.07603002e-01 3.18567455e-01 -1.78798497e-01
7.97052503e-01 3.57684463e-01 3.49664837e-01 -6.15928113e-01
1.14160404e-01 -5.39681315e-01 2.02912465e-01 9.81402576e-01
3.21303129e-01 5.66012979e-01 -7.67572597e-02 7.93925703e-01
8.54827017e-02 3.52520496e-01 1.70248318e-02 6.17250085e-01
-6.69574201e-01 -5.93008041e-01 -4.30291563e-01 4.58987296e-01
-5.39583564e-01 -3.51153404e-01 1.31334335e-01 1.00872028e+00
9.36847508e-01 1.20273006e+00 -1.49065703e-01 -8.89444053e-02
6.24103844e-01 6.57707639e-03 5.85534453e-01 -4.79293525e-01
-1.30388057e+00 3.69009316e-01 5.63082695e-01 -3.09541494e-01
-6.72065556e-01 -5.23481369e-01 -8.44186783e-01 -4.38267529e-01
-1.72002420e-01 -4.96156961e-01 8.21123898e-01 7.81916916e-01
9.28801820e-02 7.24231422e-01 -1.00366674e-01 -8.78624499e-01
-3.57767522e-01 -1.00858307e+00 -8.81016612e-01 4.54294801e-01
2.00156137e-01 -7.37546921e-01 1.92804575e-01 3.72891486e-01] | [11.563957214355469, 4.393401145935059] |
4f3f4388-ebff-4f86-bda6-4e6e401f6a87 | variational-monte-carlo-approach-to-partial | 2206.01927 | null | https://arxiv.org/abs/2206.01927v2 | https://arxiv.org/pdf/2206.01927v2.pdf | Variational Monte Carlo Approach to Partial Differential Equations with Neural Networks | The accurate numerical solution of partial differential equations is a central task in numerical analysis allowing to model a wide range of natural phenomena by employing specialized solvers depending on the scenario of application. Here, we develop a variational approach for solving partial differential equations governing the evolution of high dimensional probability distributions. Our approach naturally works on the unbounded continuous domain and encodes the full probability density function through its variational parameters, which are adapted dynamically during the evolution to optimally reflect the dynamics of the density. For the considered benchmark cases we observe excellent agreement with numerical solutions as well as analytical solutions in regimes inaccessible to traditional computational approaches. | ['Martin Gärttner', 'Moritz Reh'] | 2022-06-04 | null | null | null | null | ['variational-monte-carlo'] | ['miscellaneous'] | [-1.46592408e-01 -2.13982955e-01 3.79018515e-01 2.94667065e-01
-4.24139321e-01 -4.23047036e-01 6.11632586e-01 3.33395064e-01
-6.71082079e-01 1.11232007e+00 -4.16544318e-01 1.77583862e-02
-4.10065800e-01 -7.73970544e-01 -2.68016964e-01 -1.03494775e+00
-2.47487985e-02 9.75527465e-01 2.30381668e-01 -2.75026590e-01
3.29968110e-02 8.03885460e-01 -1.48638916e+00 -3.91477436e-01
8.60300481e-01 9.52879250e-01 -4.29843664e-02 6.12344623e-01
-1.87074795e-01 2.43079141e-01 -4.98237431e-01 -1.71682075e-01
8.27892795e-02 -4.08063918e-01 -4.40500706e-01 2.01307178e-01
-4.27260786e-01 2.79641002e-01 -1.51373953e-01 1.04721880e+00
2.91910797e-01 6.01336837e-01 1.34238255e+00 -7.82285869e-01
-1.34795710e-01 1.58981279e-01 -2.76608407e-01 4.02916580e-01
1.46563649e-01 2.19111875e-01 6.68781459e-01 -5.15435457e-01
5.52254081e-01 7.46586621e-01 5.87100625e-01 2.96309084e-01
-1.41010022e+00 1.64268419e-01 -5.04337907e-01 -6.06006570e-02
-1.66920185e+00 -8.81987624e-03 4.82494891e-01 -7.11332858e-01
4.48576093e-01 1.16521761e-01 6.90647900e-01 5.58933020e-01
4.75711465e-01 4.03387621e-02 9.77189004e-01 -2.51419395e-01
7.54783571e-01 2.63999134e-01 -5.49078509e-02 3.67653638e-01
4.06459600e-01 1.18320413e-01 8.23079422e-02 -6.11434698e-01
6.28809512e-01 -3.06560993e-01 -3.57864320e-01 -7.38266587e-01
-8.64051759e-01 7.46845663e-01 -2.48662550e-02 3.95999849e-01
-9.02539372e-01 -9.21892673e-02 9.21775997e-02 -2.78403670e-01
3.03079128e-01 3.27283442e-01 -1.66990593e-01 -3.55630279e-01
-8.10785294e-01 6.85934722e-01 1.03283763e+00 2.54504651e-01
6.69853568e-01 1.10408656e-01 -1.90937966e-01 6.29694104e-01
2.07886286e-02 7.65663326e-01 3.21521550e-01 -1.06998873e+00
-2.14012250e-01 2.94302672e-01 6.59080327e-01 -6.04033053e-01
-1.93082228e-01 -5.31360745e-01 -1.05973721e+00 3.01290125e-01
5.78032553e-01 -3.15773815e-01 -4.63655680e-01 1.39530361e+00
6.09570622e-01 -1.26377180e-01 1.46307513e-01 6.60604477e-01
2.10786052e-03 8.69431019e-01 1.21535935e-01 -5.27582824e-01
9.42027271e-01 -2.90960609e-03 -4.15125400e-01 3.24748099e-01
3.35486352e-01 -5.67285955e-01 5.35525978e-01 4.28220451e-01
-1.10661995e+00 -5.12682125e-02 -7.09537208e-01 5.52525580e-01
-2.42930442e-01 -1.20746056e-02 -2.42714472e-02 5.22167325e-01
-8.68253887e-01 9.90944624e-01 -7.45606720e-01 -3.82022411e-02
1.86078429e-01 2.62053072e-01 -1.25336424e-01 2.56103009e-01
-1.03593862e+00 8.81602347e-01 3.53667438e-01 2.28007838e-01
-4.88627106e-01 -7.80940592e-01 -6.63637221e-01 4.28791374e-01
-3.49495448e-02 -6.50447428e-01 1.07261169e+00 -3.98545116e-01
-1.59362745e+00 6.30768418e-01 -8.61513838e-02 -3.76089990e-01
9.08477366e-01 3.78567755e-01 -6.37202412e-02 7.94825107e-02
-1.44050285e-01 -1.64349296e-03 7.08912909e-01 -9.86392796e-01
-1.91925168e-02 -1.20521501e-01 -3.13264966e-01 2.01578155e-01
-2.61881948e-01 -4.53324616e-01 7.08584636e-02 -3.54473829e-01
-2.59231478e-01 -7.76449203e-01 -5.26192546e-01 1.15238592e-01
7.89117366e-02 1.19178593e-01 6.60658538e-01 -4.09461051e-01
9.30995762e-01 -1.98884761e+00 8.33739102e-01 3.21244478e-01
1.08417105e-02 3.65840107e-01 4.05078143e-01 5.59016466e-01
1.48024738e-01 -2.98059195e-01 -8.44343781e-01 -3.15638214e-01
1.10032715e-01 1.93133295e-01 -1.42741114e-01 8.72897148e-01
7.93687031e-02 3.66308451e-01 -5.27275980e-01 -4.70782459e-01
3.70676428e-01 8.52747500e-01 -6.17472589e-01 3.04841012e-01
-5.16980886e-01 6.07855201e-01 -5.62298894e-01 -1.15715295e-01
7.76641428e-01 -2.40567997e-01 6.63844496e-02 1.71524599e-01
-4.33513314e-01 -3.04919362e-01 -1.38581967e+00 9.39741969e-01
-5.74068785e-01 1.96418986e-01 6.26407325e-01 -1.23751283e+00
8.71455729e-01 3.93040329e-01 6.17189467e-01 -1.71142176e-01
5.05224168e-01 3.33980203e-01 3.10988780e-02 -2.59972662e-01
4.25333440e-01 -9.15219247e-01 -9.86155216e-03 3.67439508e-01
-5.37663624e-02 -4.31546718e-01 3.64992291e-01 6.20446727e-02
9.54773307e-01 -2.65215009e-01 5.79494298e-01 -8.13273132e-01
8.53338778e-01 5.71273565e-02 1.35690734e-01 3.11501384e-01
-5.13030812e-02 4.50450331e-01 8.51699471e-01 -1.15827419e-01
-1.06186998e+00 -8.56906652e-01 -7.87516415e-01 1.73604295e-01
5.02864234e-02 1.77458525e-01 -8.60424280e-01 1.57011107e-01
6.19352497e-02 5.28674722e-01 -7.80107081e-01 -1.37628824e-01
-4.90230024e-01 -1.07718742e+00 1.09053031e-01 -9.06749070e-02
4.13763613e-01 -1.07165182e+00 -6.45775497e-01 3.20749909e-01
2.73643553e-01 -1.13867188e+00 8.92994627e-02 4.83050607e-02
-8.16264451e-01 -9.20995712e-01 -1.13766158e+00 -3.70961726e-02
2.05294579e-01 -5.90847254e-01 8.45394492e-01 -6.49374500e-02
-6.54098451e-01 4.69210178e-01 2.13418320e-01 1.98211119e-01
-8.44226599e-01 9.35376659e-02 -7.08940327e-02 1.30301625e-01
-2.76590317e-01 -7.12741256e-01 -3.75538945e-01 1.55965880e-01
-9.00587678e-01 -5.69200695e-01 1.88651487e-01 4.25700635e-01
5.47411978e-01 4.28533018e-01 2.04767868e-01 -6.07561827e-01
7.16626942e-01 -7.65150487e-01 -1.17383361e+00 -5.29618934e-02
-8.47783312e-02 4.18066889e-01 9.25518334e-01 -3.26490551e-01
-1.13037682e+00 -1.11196257e-01 -1.32957086e-01 -3.63273531e-01
-2.36739963e-01 2.57251024e-01 2.13009752e-02 -2.65657932e-01
4.39930171e-01 3.22262466e-01 1.16362922e-01 -5.73050916e-01
-8.84192362e-02 3.00232530e-01 5.75107992e-01 -8.91371310e-01
6.14119470e-01 5.12627840e-01 7.76250005e-01 -1.04495025e+00
-4.42046613e-01 -2.33706191e-01 -5.70236027e-01 -1.63890749e-01
5.46316862e-01 -1.40888646e-01 -1.05443490e+00 5.67503750e-01
-9.86475050e-01 -5.78991294e-01 -7.45724261e-01 3.18326503e-01
-6.74445570e-01 3.95339638e-01 -3.66829783e-01 -1.14925778e+00
3.45603675e-02 -9.08750236e-01 7.24885464e-01 2.85761893e-01
-1.29599810e-01 -1.51514149e+00 6.10850036e-01 -4.81071144e-01
6.55716419e-01 5.23522019e-01 1.03970623e+00 -4.84191924e-02
-1.88119248e-01 -2.90976912e-01 7.07166269e-02 2.05026656e-01
-1.95167318e-01 3.00562173e-01 -4.59382415e-01 2.25566160e-02
1.93061545e-01 2.17299446e-01 6.04189515e-01 5.64362168e-01
8.26008976e-01 1.30296409e-01 -3.42491686e-01 3.90979499e-01
1.51140583e+00 -2.44107485e-01 2.74739951e-01 -2.45148256e-01
1.73062742e-01 6.00471318e-01 2.42092624e-01 9.69580650e-01
-6.58677220e-02 7.17696428e-01 2.47651011e-01 4.48913634e-01
3.14077854e-01 2.34220117e-01 -2.51933411e-02 8.78139287e-02
-2.61866245e-02 -3.82434130e-01 -9.94758189e-01 5.59406817e-01
-1.60404539e+00 -8.64285350e-01 -3.98876458e-01 2.38735700e+00
6.24848604e-01 1.39617309e-01 2.99174488e-01 3.07016641e-01
8.34498048e-01 4.14174385e-02 -3.11691254e-01 -4.63283867e-01
7.25776283e-03 4.52220082e-01 2.70470440e-01 8.11057568e-01
-6.93287373e-01 2.52409041e-01 7.21982765e+00 6.42298996e-01
-9.79219496e-01 1.69429049e-01 2.42502049e-01 -9.60976351e-03
-1.61350325e-01 -1.77679390e-01 -5.61053634e-01 6.43346131e-01
9.45586145e-01 -5.43734550e-01 3.84415448e-01 4.08342719e-01
1.09912574e-01 -5.17219365e-01 -5.65015197e-01 5.40824473e-01
-3.81584436e-01 -1.18934870e+00 -2.39922404e-01 2.97223836e-01
8.00149560e-01 -2.88908839e-01 2.28979602e-01 -5.43056801e-02
7.98757523e-02 -9.26437378e-01 4.86405402e-01 6.16734207e-01
6.04573965e-01 -7.17138171e-01 6.28720284e-01 6.21998370e-01
-8.86187911e-01 5.65158762e-02 -2.75337100e-01 -3.47015440e-01
7.84306765e-01 9.57931101e-01 -1.78234160e-01 1.37315214e-01
3.18410218e-01 3.67813975e-01 2.99773719e-02 1.25710523e+00
3.27132672e-01 4.81645405e-01 -7.65003383e-01 -1.02298744e-01
1.91320091e-01 -6.34179115e-01 8.59119594e-01 8.27651978e-01
6.27871871e-01 3.89804214e-01 -3.00552398e-01 1.16873419e+00
1.68096393e-01 1.30575061e-01 -3.03272277e-01 -1.05192140e-01
4.02733572e-02 9.52222884e-01 -8.01201463e-01 -5.27222045e-02
9.85079929e-02 4.87851530e-01 3.51344734e-01 4.56812799e-01
-6.37309849e-01 -1.54744282e-01 7.88190901e-01 4.39758599e-01
5.95444620e-01 -5.33336878e-01 -9.73432884e-02 -1.04639220e+00
-1.42819494e-01 -8.33488777e-02 3.65616471e-01 -2.84490347e-01
-1.00644112e+00 6.32298648e-01 4.07246739e-01 -7.35486150e-01
-3.80840272e-01 -8.20853889e-01 -8.37519705e-01 1.06010795e+00
-1.06480300e+00 -1.34765834e-01 -3.83692347e-02 4.61983174e-01
-2.38642573e-01 3.40219662e-02 6.67129874e-01 1.91168964e-01
-6.20391428e-01 -1.79150686e-01 5.25902331e-01 -4.54783469e-01
-6.00722320e-02 -1.08813107e+00 -2.48379618e-01 6.22348666e-01
-4.11270678e-01 2.28438124e-01 1.34270144e+00 -4.53105330e-01
-9.67067420e-01 -3.86428893e-01 6.21059000e-01 1.07779913e-01
7.28741348e-01 -3.18564028e-02 -1.27675450e+00 1.74733728e-01
1.21425673e-01 4.83881831e-01 2.35353023e-01 -3.48532081e-01
3.10102344e-01 2.29778662e-01 -1.29521537e+00 1.50032729e-01
4.13606793e-01 -2.95060705e-02 -3.77810746e-02 1.10714786e-01
-7.15732500e-02 -5.12150586e-01 -7.61699617e-01 3.48213315e-01
1.98263630e-01 -9.27184701e-01 7.65140057e-01 -1.92939386e-01
1.88577026e-01 -5.45171909e-02 2.12890096e-02 -1.17239761e+00
3.52534205e-02 -8.52353334e-01 -1.81456059e-01 9.64064062e-01
-8.77619833e-02 -7.49512732e-01 5.65875053e-01 5.81149518e-01
5.13513088e-01 -7.66230762e-01 -1.19743085e+00 -5.01206756e-01
3.67306083e-01 -2.44457155e-01 4.15366441e-01 3.64410281e-01
-1.31860062e-01 -3.23257387e-01 -1.65920462e-02 2.09180757e-01
8.80696356e-01 7.49527887e-02 1.92975625e-01 -1.37204719e+00
-5.11280715e-01 -6.94396317e-01 -5.18443167e-01 -5.39485276e-01
4.85606939e-01 -5.22089541e-01 1.67481363e-01 -1.10649478e+00
3.98202837e-02 -4.79467779e-01 -7.60250166e-03 -3.41976017e-01
-6.60259798e-02 3.65436345e-01 -3.39879572e-01 2.00824551e-02
-2.76147157e-01 8.41678739e-01 9.07738686e-01 3.81100714e-01
-1.00612372e-01 2.76702017e-01 -3.63365263e-02 6.35196805e-01
6.23466432e-01 -5.60863078e-01 -1.70647904e-01 1.02017768e-01
2.51177907e-01 3.15679222e-01 5.47078848e-01 -1.09519088e+00
1.17399819e-01 -2.96318561e-01 -8.21268037e-02 -1.73386782e-01
3.94823730e-01 -5.13068497e-01 3.19956273e-01 4.30038780e-01
-1.18219323e-01 -7.96449557e-02 5.08723497e-01 4.66355950e-01
-3.30158055e-01 -6.89278841e-01 1.32406652e+00 7.66483247e-02
-2.34666049e-01 3.23586971e-01 -9.05756116e-01 4.45788443e-01
1.02331662e+00 2.60906339e-01 1.16182417e-01 -5.37793636e-01
-1.13737202e+00 -7.48419389e-02 7.22078145e-01 -5.87296128e-01
1.74085394e-01 -1.04773331e+00 -4.14279282e-01 7.11199418e-02
-3.73005271e-01 -8.96879211e-02 5.82986236e-01 8.14006925e-01
-8.39576781e-01 1.44508243e-01 -2.20768124e-01 -3.41193110e-01
-6.14622414e-01 5.00400245e-01 8.99176121e-01 -4.08374995e-01
-4.45190996e-01 4.28106219e-01 6.34253398e-02 -1.60887927e-01
-1.84225440e-01 -2.45319456e-02 -1.21130161e-01 1.17494620e-03
2.21279770e-01 5.95410109e-01 -8.60801712e-02 -9.51966047e-01
-2.07289651e-01 6.44974172e-01 4.95449036e-01 -4.54564482e-01
1.15403521e+00 -5.59789538e-02 -2.00645193e-01 4.65223670e-01
1.14702523e+00 -1.40757993e-01 -1.40913057e+00 -1.06390513e-01
-2.74486989e-01 -1.60556719e-01 2.07715183e-01 -2.11587459e-01
-7.64702201e-01 1.00675988e+00 1.58557236e-01 5.53900838e-01
7.37118602e-01 3.32664773e-02 6.30426109e-01 4.83184718e-02
2.16116741e-01 -7.50916183e-01 -2.64363945e-01 4.11780357e-01
6.57371163e-01 -7.26353884e-01 1.20922320e-01 -4.62535769e-01
-4.08704966e-01 1.08412850e+00 7.05337375e-02 -4.02121812e-01
9.59398746e-01 5.62118888e-01 -3.13395679e-01 -8.02598521e-02
-6.12938821e-01 -1.08602576e-01 1.48849592e-01 3.61155033e-01
2.32385963e-01 -1.53792929e-02 -5.36025286e-01 2.17727497e-01
-4.41115210e-03 -3.91013324e-02 6.23788476e-01 6.70579851e-01
-1.74863860e-01 -1.18578041e+00 -3.74923110e-01 1.68003693e-01
-4.26258624e-01 3.45996976e-01 5.56656905e-03 7.78022587e-01
-1.85310185e-01 3.14021468e-01 2.10711330e-01 6.59395039e-01
2.04240218e-01 2.73962498e-01 6.27942026e-01 -4.50533003e-01
-2.21385360e-01 -2.50192821e-01 -3.99127871e-01 -1.70963868e-01
-2.34786838e-01 -1.15690005e+00 -1.33451283e+00 -4.36312169e-01
1.25317127e-01 5.95977902e-01 4.61740792e-01 1.06082237e+00
9.94579345e-02 3.36750537e-01 1.87557966e-01 -1.03439307e+00
-6.62047088e-01 -6.35849237e-01 -1.04069877e+00 1.42339826e-01
3.53854865e-01 -9.11044121e-01 -6.18118763e-01 -4.29411292e-01] | [6.388342380523682, 3.709676504135132] |
545c48f1-3382-4e99-8307-19d3d45583d5 | 11k-hands-gender-recognition-and-biometric | 1711.04322 | null | http://arxiv.org/abs/1711.04322v9 | http://arxiv.org/pdf/1711.04322v9.pdf | 11K Hands: Gender recognition and biometric identification using a large dataset of hand images | The human hand possesses distinctive features which can reveal gender
information. In addition, the hand is considered one of the primary biometric
traits used to identify a person. In this work, we propose a large dataset of
human hand images (dorsal and palmar sides) with detailed ground-truth
information for gender recognition and biometric identification. Using this
dataset, a convolutional neural network (CNN) can be trained effectively for
the gender recognition task. Based on this, we design a two-stream CNN to
tackle the gender recognition problem. This trained model is then used as a
feature extractor to feed a set of support vector machine classifiers for the
biometric identification task. We show that the dorsal side of hand images,
captured by a regular digital camera, convey effective distinctive features
similar to, if not better, those available in the palmar hand images. To
facilitate access to the proposed dataset and replication of our experiments,
the dataset, trained CNN models, and Matlab source code are available at
(https://goo.gl/rQJndd). | ['Mahmoud Afifi'] | 2017-11-12 | null | null | null | null | ['animal-pose-estimation'] | ['computer-vision'] | [ 1.94520384e-01 -2.22778141e-01 -2.19652161e-01 -5.36966324e-01
-2.66750157e-01 -7.36709595e-01 4.08484876e-01 -2.25951388e-01
-3.28429371e-01 4.35445428e-01 -1.92713514e-01 -2.33851090e-01
4.34771404e-02 -8.21117282e-01 -3.45215350e-01 -1.03220201e+00
1.23746127e-01 3.33390236e-01 -5.53980708e-01 9.00563374e-02
4.65441883e-01 8.64456952e-01 -1.82901525e+00 4.66679260e-02
3.01359326e-01 1.40446281e+00 -2.75136828e-01 4.92282391e-01
2.67703950e-01 8.33333954e-02 -3.92381370e-01 -8.21180046e-01
4.24784988e-01 -1.25163138e-01 -6.72940493e-01 -1.72740713e-01
4.93798822e-01 -7.47024536e-01 -4.28604573e-01 9.83698368e-01
1.01159990e+00 -1.86320528e-01 5.76798022e-01 -1.23155105e+00
-3.51561546e-01 3.96876305e-01 -7.34391630e-01 -1.05372153e-01
2.15542912e-01 1.62124142e-01 8.62472534e-01 -9.01674092e-01
3.75496954e-01 1.08259523e+00 5.50610065e-01 8.70595276e-01
-8.00230265e-01 -1.03798962e+00 -4.74748820e-01 4.76738252e-02
-1.61812782e+00 -6.53703868e-01 9.96601999e-01 -1.93207368e-01
2.06635788e-01 3.90025258e-01 7.27061093e-01 1.40183687e+00
1.61564481e-02 7.81600833e-01 1.23869312e+00 -2.11208165e-01
-8.85235220e-02 1.36696532e-01 1.50305601e-02 7.61164904e-01
2.83513069e-01 -6.41947729e-04 -6.24486089e-01 -1.36991277e-01
8.38394940e-01 2.20221385e-01 -3.89433764e-02 9.06872675e-02
-8.61955166e-01 4.57528025e-01 2.10166946e-01 2.09554031e-01
-4.89557564e-01 8.33024681e-02 4.13794130e-01 1.38708010e-01
2.02697262e-01 1.19477406e-01 -3.51783931e-01 -2.51505375e-01
-9.61821139e-01 2.54859656e-01 8.35943341e-01 7.03385592e-01
5.90233147e-01 -1.23470634e-01 -2.62653708e-01 8.55487704e-01
2.95303881e-01 9.25322831e-01 1.95711106e-01 -6.88179135e-01
2.71656752e-01 7.75572002e-01 -1.33052677e-01 -1.17574322e+00
-3.57682377e-01 -3.35943758e-01 -1.01345170e+00 -7.91466683e-02
8.50579381e-01 1.32842148e-02 -7.51022041e-01 1.47656190e+00
2.15954512e-01 -3.61237288e-01 -2.89501578e-01 9.56687987e-01
9.50198948e-01 1.13477401e-01 -2.36798570e-01 1.91422313e-01
1.78421378e+00 -2.15087831e-01 -4.82612729e-01 6.24798276e-02
4.34123864e-03 -7.52499580e-01 7.68816113e-01 3.83717448e-01
-7.76105762e-01 -4.16094214e-01 -8.11606765e-01 3.61469649e-02
-5.14823139e-01 7.62292683e-01 5.82874775e-01 1.05493772e+00
-6.62284195e-01 5.30869782e-01 -7.55592644e-01 -4.20306891e-01
8.97607148e-01 7.66315997e-01 -3.94267380e-01 -1.16390355e-01
-8.30557406e-01 3.86779070e-01 6.14306331e-02 6.29247010e-01
-7.37546146e-01 -3.41437340e-01 -8.03354144e-01 -1.81215227e-01
-1.49830272e-02 -2.30975449e-01 9.63776529e-01 -7.76962578e-01
-1.29316068e+00 1.23028207e+00 -3.85618478e-01 1.48542896e-01
6.40524030e-01 7.77633116e-02 -1.06885657e-01 1.91338822e-01
-1.16725024e-02 3.97804976e-01 1.20024586e+00 -1.06298769e+00
-3.83568257e-01 -1.17072058e+00 -2.46178120e-01 -3.06063771e-01
-5.49021900e-01 3.42132300e-01 -4.42288965e-01 -5.17922163e-01
6.89539686e-02 -9.13946271e-01 4.19223458e-01 -3.05685431e-01
-7.53963530e-01 -2.20095649e-01 7.89186478e-01 -1.08651936e+00
9.53992128e-01 -2.14645028e+00 -3.28933564e-03 7.28450656e-01
4.09606725e-01 1.42409384e-01 -3.45436409e-02 3.16897750e-01
-1.29320756e-01 -1.09648190e-01 -1.26589507e-01 -4.19902831e-01
6.13010041e-02 -6.64831549e-02 -4.55829240e-02 6.70214236e-01
1.76865786e-01 9.75315928e-01 -4.68975753e-01 -4.24952567e-01
-2.82561705e-02 7.02971876e-01 6.98906854e-02 2.83414453e-01
3.32065374e-01 7.62530804e-01 -3.63694400e-01 1.18543005e+00
1.06627822e+00 1.42567873e-01 2.22636014e-01 -4.38218713e-01
1.06004246e-01 -2.85717696e-01 -1.13144529e+00 1.12499928e+00
-2.41395429e-01 6.09164476e-01 7.60622397e-02 -6.39475703e-01
1.29639196e+00 2.23838449e-01 4.16019112e-01 -7.00888515e-01
6.01673663e-01 3.96940976e-01 1.26215473e-01 -5.46696544e-01
3.44379783e-01 8.97787735e-02 1.14483662e-01 8.11058819e-01
1.91303641e-01 3.32370788e-01 1.63689166e-01 -2.73059070e-01
4.81594861e-01 -2.42028534e-01 -1.37132555e-01 -9.80730578e-02
6.76226795e-01 -4.90914285e-01 3.46227288e-01 5.01191139e-01
-1.73412412e-01 6.59245253e-01 6.38688087e-01 -7.99323380e-01
-1.21114266e+00 -6.73120677e-01 -3.57580721e-01 1.14078355e+00
-8.27919543e-02 -1.67098656e-01 -9.49173331e-01 -4.61266875e-01
2.57327586e-01 -1.43918440e-01 -8.71170580e-01 1.00633509e-01
-5.74649751e-01 -6.13126278e-01 1.06481481e+00 7.56751597e-01
8.87573242e-01 -9.93995607e-01 -7.09461689e-01 -3.03553730e-01
-1.19800553e-01 -7.48017967e-01 -2.85153717e-01 -3.64229232e-02
-5.63749194e-01 -1.47848272e+00 -1.14433908e+00 -6.78673685e-01
7.70410180e-01 -2.13179380e-01 6.39227152e-01 2.33838290e-01
-6.47233844e-01 3.51776779e-01 -1.62201785e-02 -5.99377811e-01
2.21496392e-02 5.17292559e-01 2.77917951e-01 4.98594910e-01
6.03016555e-01 -5.41990221e-01 -8.98108900e-01 2.22959459e-01
-6.40016317e-01 -2.31498748e-01 6.27547026e-01 8.79455268e-01
3.16705078e-01 -2.25115195e-02 4.48227733e-01 -5.63487947e-01
6.32346153e-01 -3.89416367e-02 -4.96835619e-01 8.44133645e-02
-5.00599980e-01 -2.48106197e-01 5.27248561e-01 -1.51942194e-01
-9.13768232e-01 1.09434217e-01 -2.67650843e-01 -2.60402590e-01
-3.61946315e-01 3.81082684e-01 -3.78159434e-01 -2.04712242e-01
2.10934296e-01 5.31760454e-01 3.95677537e-01 -5.92023432e-01
9.72997025e-02 1.26686347e+00 6.15141392e-01 -6.44276202e-01
7.03349948e-01 4.19992715e-01 1.89346701e-01 -8.33873451e-01
-4.59105253e-01 -3.07293117e-01 -1.01390624e+00 -4.31181103e-01
5.34209430e-01 -5.99022686e-01 -1.29188550e+00 1.22255480e+00
-8.12909365e-01 3.51817161e-02 2.68437713e-01 6.03886917e-02
-3.59546602e-01 1.52223721e-01 -6.07295334e-01 -1.14445579e+00
-7.35375524e-01 -8.66673589e-01 1.36667204e+00 5.02205074e-01
-1.97253704e-01 -7.48690844e-01 -3.11109602e-01 3.77030730e-01
4.15212870e-01 2.26636067e-01 8.08189988e-01 -7.52378404e-01
-2.18191981e-01 -7.23532259e-01 -4.63055879e-01 2.23300412e-01
3.56978863e-01 2.90029913e-01 -1.39622152e+00 -5.21645129e-01
-1.95457563e-01 -3.30412656e-01 8.66550684e-01 3.11787009e-01
1.45607948e+00 -3.72126251e-01 -1.05729155e-01 7.30195463e-01
1.08535635e+00 6.07614703e-02 6.48546755e-01 1.67077571e-01
8.38680565e-01 7.77934790e-01 3.00287396e-01 8.44246924e-01
4.53733593e-01 4.54933047e-01 2.61384428e-01 -1.52041167e-01
1.47316203e-01 -6.57337978e-02 -2.26430044e-01 2.52266765e-01
-9.08328414e-01 1.87986359e-01 -1.18075407e+00 3.73990893e-01
-1.43299162e+00 -9.05206680e-01 1.01421013e-01 2.13736081e+00
6.76813304e-01 -5.46596289e-01 6.08326554e-01 5.76820910e-01
8.58974993e-01 8.45709667e-02 -5.11623859e-01 -1.37454107e-01
-6.42242581e-02 4.79504108e-01 4.78953928e-01 -1.51536688e-01
-1.32262778e+00 5.88600814e-01 5.45757627e+00 4.65230972e-01
-1.58209372e+00 -3.72566342e-01 7.34221220e-01 3.36697847e-02
1.87396020e-01 -6.35328174e-01 -6.16496205e-01 4.79663581e-01
5.86560249e-01 -9.35078319e-03 8.05474401e-01 5.57024300e-01
-6.57745153e-02 9.76792872e-02 -1.19701111e+00 1.41299033e+00
4.32777405e-02 -8.69300604e-01 -5.31986766e-02 3.58263731e-01
2.71951318e-01 -6.10005796e-01 6.95427179e-01 -1.84261397e-01
-3.35046470e-01 -1.46170259e+00 5.35512745e-01 4.35512245e-01
1.26021230e+00 -1.12594879e+00 1.17192352e+00 3.78758721e-02
-1.13546360e+00 -3.57607245e-01 -1.27227768e-01 1.08334012e-01
-3.89566213e-01 2.62087673e-01 -5.40339053e-01 5.74704707e-01
8.55289757e-01 6.80251956e-01 -8.17927778e-01 7.20234036e-01
-7.62912855e-02 4.70861077e-01 -9.16848704e-02 -5.83579578e-03
-3.21849644e-01 2.54248213e-02 1.87550038e-01 1.01932979e+00
3.24124753e-01 4.75606285e-02 -4.08949852e-01 1.01261592e+00
-2.00175151e-01 2.70277774e-03 -5.01376808e-01 -5.18004954e-01
4.58147407e-01 1.48275447e+00 -7.56556630e-01 -7.80859292e-02
-1.60744730e-02 8.67034018e-01 -1.63988546e-01 3.86147201e-01
-3.65101546e-01 -7.50301659e-01 6.51574969e-01 -1.66216847e-02
1.56197026e-01 -3.70045267e-02 -5.86653888e-01 -1.08701837e+00
1.78338289e-01 -1.03253186e+00 2.66810685e-01 -2.67263860e-01
-1.26849771e+00 4.59673554e-01 -2.28749514e-01 -9.92732465e-01
-2.66002804e-01 -1.01441407e+00 -6.40527070e-01 1.20053387e+00
-1.17144358e+00 -1.69898450e+00 -9.00225401e-01 7.98930347e-01
-3.50233763e-02 -6.24357104e-01 9.96639311e-01 3.54946941e-01
-7.45233655e-01 1.02707553e+00 3.24103534e-02 9.29057539e-01
6.64761961e-01 -1.02849388e+00 3.22414696e-01 3.63170594e-01
-3.17481369e-01 1.08098364e+00 3.92651796e-01 -5.62874556e-01
-1.84394729e+00 -9.05737042e-01 5.81191957e-01 -4.27140146e-01
1.87500179e-01 -5.22267163e-01 -4.66755420e-01 5.15652478e-01
-1.32982269e-01 4.84719966e-03 9.96985853e-01 1.23724855e-01
-4.99258786e-01 -2.88778424e-01 -1.40399897e+00 2.08068371e-01
7.61883676e-01 -7.83265889e-01 -2.87107006e-02 -1.51299745e-01
-2.63928056e-01 -4.88358706e-01 -9.77907717e-01 2.20263287e-01
1.60095000e+00 -8.01338255e-01 8.59538198e-01 -5.98726928e-01
8.16140473e-01 5.27329221e-02 -2.45071903e-01 -8.49721134e-01
1.24120384e-01 -5.99151924e-02 -1.58293694e-01 1.38610351e+00
6.98733479e-02 -7.22780347e-01 1.02625847e+00 5.60938537e-01
5.89630783e-01 -8.35979164e-01 -8.81271362e-01 -7.17518806e-01
1.08246379e-01 -1.47965878e-01 9.33203995e-01 6.52237058e-01
-1.90912008e-01 -2.20874101e-01 -2.64730096e-01 3.44058336e-03
1.09782720e+00 3.75420481e-01 8.16783190e-01 -1.36360741e+00
1.64042443e-01 -4.21925128e-01 -6.49880469e-01 -4.54976320e-01
2.40511000e-01 -7.76157796e-01 -3.28114152e-01 -9.80525672e-01
5.06451905e-01 -3.29914361e-01 -3.32497358e-01 6.69898093e-01
7.50100613e-02 8.35233927e-01 2.00242832e-01 2.38734707e-01
1.73328817e-02 2.78328210e-01 1.15234137e+00 -4.38700587e-01
-8.73913839e-02 4.00376171e-01 -7.61777043e-01 3.91875565e-01
9.97924387e-01 -1.58751220e-01 1.35414511e-01 -1.18470944e-01
-6.68087900e-02 9.74505267e-04 6.93698406e-01 -6.24205828e-01
1.57729596e-01 6.20085858e-02 1.01781523e+00 -6.25260472e-01
3.37623000e-01 -6.28824115e-01 -7.93583468e-02 5.02650440e-01
-2.38164052e-01 -1.96998924e-01 1.61272615e-01 -1.05602920e-01
-1.91858321e-01 1.07679730e-02 5.89258313e-01 -4.77683730e-02
-2.88624704e-01 5.40362775e-01 -1.28365472e-01 -5.51427007e-01
7.12707222e-01 -4.08637404e-01 -3.10972661e-01 -3.02659929e-01
-5.97983003e-01 -7.42846653e-02 3.88237387e-01 2.95634001e-01
7.12271631e-01 -1.44016874e+00 -7.15842664e-01 6.39313638e-01
2.22261980e-01 -3.05218726e-01 1.84680134e-01 8.83488297e-01
-4.58274335e-01 4.31597531e-01 -8.46637666e-01 -5.84473073e-01
-1.61153829e+00 2.00497836e-01 4.29179847e-01 4.44326669e-01
-1.68873653e-01 8.87194514e-01 -2.70887077e-01 -6.14610374e-01
4.56123799e-01 1.03051506e-01 -4.23156619e-01 3.70542616e-01
7.75705695e-01 5.32918394e-01 1.01422109e-01 -7.67060935e-01
-6.00937366e-01 7.40485072e-01 -1.07739475e-02 -8.99715628e-03
1.59760773e+00 9.54691544e-02 -5.87738752e-01 1.73358038e-01
1.14313781e+00 1.59872457e-01 -8.73406708e-01 -5.13923280e-02
-2.86276132e-01 -9.27949905e-01 -3.86603206e-01 -7.71111190e-01
-1.53935122e+00 1.17857444e+00 9.74969327e-01 1.44094834e-02
1.31713724e+00 -9.07550287e-03 6.39184773e-01 4.26719278e-01
4.44520801e-01 -1.00702536e+00 -2.23678082e-01 2.95387864e-01
8.59026909e-01 -1.21310914e+00 -2.36009747e-01 -1.57843038e-01
-2.25102574e-01 1.58021331e+00 5.37846208e-01 2.04501495e-01
4.88833547e-01 1.82532027e-01 2.48332202e-01 -2.48390749e-01
1.69162676e-02 -1.32569537e-01 2.48069108e-01 8.13419282e-01
5.05673289e-01 1.28847107e-01 -1.21222660e-01 9.87233937e-01
-5.38993955e-01 5.63127398e-02 1.70133904e-01 6.68218493e-01
6.82532117e-02 -1.06612575e+00 -5.03441393e-01 7.04295695e-01
-6.05651319e-01 7.09977522e-02 -6.55167222e-01 3.81005794e-01
1.71821401e-01 7.37944305e-01 6.51051477e-02 -7.75802195e-01
1.41679361e-01 2.15316221e-01 8.35030258e-01 -8.99758562e-02
-4.46087688e-01 -3.35400730e-01 -1.92706048e-01 -4.43177611e-01
-1.73741773e-01 -6.58956468e-01 -7.29229510e-01 -8.48739624e-01
5.55062518e-02 -2.93022752e-01 8.31373215e-01 7.91031659e-01
8.51688385e-02 1.02231279e-01 8.73836696e-01 -9.95222390e-01
-6.01943612e-01 -1.05767739e+00 -1.11263645e+00 3.99431795e-01
2.85535067e-01 -6.65300131e-01 -3.21365222e-02 1.36784196e-01] | [13.329047203063965, 0.9680781364440918] |
5b13a600-7b1b-4e4c-82e9-d48e372660c4 | fingerflex-inferring-finger-trajectories-from | 2211.01960 | null | https://arxiv.org/abs/2211.01960v2 | https://arxiv.org/pdf/2211.01960v2.pdf | FingerFlex: Inferring Finger Trajectories from ECoG signals | Motor brain-computer interface (BCI) development relies critically on neural time series decoding algorithms. Recent advances in deep learning architectures allow for automatic feature selection to approximate higher-order dependencies in data. This article presents the FingerFlex model - a convolutional encoder-decoder architecture adapted for finger movement regression on electrocorticographic (ECoG) brain data. State-of-the-art performance was achieved on a publicly available BCI competition IV dataset 4 with a correlation coefficient between true and predicted trajectories up to 0.74. The presented method provides the opportunity for developing fully-functional high-precision cortical motor brain-computer interfaces. | ['Alexey Timchenko', 'Alexander Kovalev', 'Vladislav Lomtev'] | 2022-10-23 | null | null | null | null | ['brain-decoding', 'brain-decoding'] | ['medical', 'miscellaneous'] | [ 1.70937806e-01 -2.81759530e-01 -1.43114120e-01 -1.07709244e-01
-5.27629077e-01 -4.65594083e-02 7.31544435e-01 -6.40858889e-01
-7.31604517e-01 8.44362736e-01 2.89570689e-01 -8.05202127e-02
-3.99673104e-01 -6.46748245e-02 -5.96807241e-01 -3.92103374e-01
-8.33914876e-01 2.93925047e-01 2.97417164e-01 -1.53903246e-01
3.53555143e-01 4.82129991e-01 -1.30067110e+00 7.37469673e-01
6.71695054e-01 1.01262200e+00 3.79049420e-01 6.72186613e-01
4.30682600e-01 5.99312246e-01 -4.39509600e-01 -1.93287447e-01
-3.65305901e-01 -5.86966515e-01 -6.74542904e-01 -7.24694610e-01
-5.71039803e-02 -3.34073126e-01 -9.65390444e-01 8.10510933e-01
6.65220022e-01 -3.50763530e-01 1.11849010e+00 -9.84074116e-01
-5.52217185e-01 5.45792222e-01 3.34058292e-02 7.47116208e-01
6.47583306e-02 4.93507624e-01 5.31911790e-01 -6.91276014e-01
7.79718637e-01 7.04824448e-01 7.49764085e-01 7.75322556e-01
-1.20475268e+00 -8.13052595e-01 -2.76932299e-01 9.95651484e-01
-1.27786601e+00 -3.84605318e-01 5.77444255e-01 -7.05812454e-01
1.84255576e+00 -7.99036622e-02 1.23157191e+00 1.73035097e+00
1.24157679e+00 7.34172463e-01 1.01231718e+00 -1.38280094e-01
3.89243034e-03 -2.87240267e-01 3.04668576e-01 1.11727610e-01
-3.96122932e-02 5.85453629e-01 -1.00980306e+00 1.86502248e-01
9.48051453e-01 -3.93352896e-01 -2.90975064e-01 2.16537669e-01
-1.33192647e+00 5.21286488e-01 4.50417995e-01 6.21673167e-01
-1.11459482e+00 5.42739213e-01 5.98665237e-01 4.96152043e-01
1.41472280e-01 2.63305694e-01 -3.87399316e-01 -8.80125165e-01
-1.14317584e+00 2.19763130e-01 5.16708195e-01 8.74753356e-01
-1.82459936e-01 1.55058309e-01 -4.22687680e-01 7.14782774e-01
1.82796404e-01 3.41095924e-01 9.84556079e-01 -7.19015896e-01
5.78340411e-01 4.66692522e-02 -8.26105475e-02 -3.58668923e-01
-8.68594408e-01 -1.17766604e-01 -8.41344118e-01 5.13281584e-01
4.27339375e-01 -1.16127275e-01 -7.87211716e-01 1.52479136e+00
-6.23313904e-01 -1.78836405e-01 -4.08416629e-01 8.96802902e-01
2.11031526e-01 -8.16075131e-02 2.44766608e-01 -1.27543002e-01
1.02102387e+00 -3.68492723e-01 -8.42968106e-01 -2.55440265e-01
3.58958900e-01 1.05531104e-01 7.35271037e-01 7.69130945e-01
-1.03142095e+00 -2.72562563e-01 -1.10158587e+00 1.89606607e-01
-2.48054191e-01 2.67773390e-01 5.46609461e-01 3.17739576e-01
-1.02447534e+00 1.13133705e+00 -1.49298990e+00 -3.04069161e-01
8.64376128e-01 9.92715955e-01 -7.65208006e-01 6.04772270e-01
-1.33324826e+00 1.53751051e+00 1.10097066e-01 9.73851979e-02
-9.02415633e-01 -5.46363473e-01 -4.73371055e-03 -2.75347650e-01
-6.76434696e-01 -4.71119046e-01 1.13555729e+00 -7.70794570e-01
-1.73730993e+00 9.00891006e-01 -2.12437153e-01 -7.17648268e-01
4.48996961e-01 -4.80775863e-01 -6.73748374e-01 -8.05896968e-02
-2.39699692e-01 6.38723135e-01 7.81144261e-01 -2.84058541e-01
-4.70735967e-01 -7.90960431e-01 -9.22803223e-01 -3.15405130e-01
-2.51623899e-01 1.99425131e-01 -3.20253819e-02 -7.10127890e-01
-2.07868204e-01 -9.64231253e-01 1.19007096e-01 -2.48597443e-01
-1.72762722e-01 -3.58833909e-01 1.60745934e-01 -1.16910815e+00
1.39746165e+00 -1.76042879e+00 4.45568085e-01 1.47826344e-01
2.18020231e-01 2.04011187e-01 -1.66471899e-01 3.97360265e-01
-5.34712672e-01 -4.13789660e-01 -1.03755087e-01 4.04726088e-01
-9.98577103e-02 -2.91021615e-01 -1.02800652e-01 7.53402352e-01
4.39732790e-01 1.32871473e+00 -6.85090423e-01 2.31805816e-01
1.31348416e-01 5.11466146e-01 -4.86652941e-01 1.32095367e-01
1.17884368e-01 7.53122211e-01 1.08974501e-01 2.66954005e-01
-1.94187090e-02 1.35895908e-01 5.31108156e-02 -1.40147746e-01
-2.72663325e-01 5.21815062e-01 -2.57627875e-01 2.06156921e+00
-2.18796179e-01 1.47460532e+00 -4.66087013e-01 -1.00025535e+00
6.94568396e-01 5.85730195e-01 8.62039745e-01 -1.27428734e+00
6.35747790e-01 4.56231058e-01 7.06812382e-01 -7.14263320e-01
-3.84955257e-01 8.75827819e-02 3.01734388e-01 2.59348929e-01
4.07645077e-01 2.96083182e-01 1.02284729e-01 -4.16747421e-01
1.36899531e+00 3.30943912e-01 5.75078800e-02 -4.07122165e-01
2.25109398e-01 -3.96369305e-03 -1.31033808e-01 6.08125925e-01
-2.21516132e-01 5.97593248e-01 4.57522035e-01 -2.81640172e-01
-1.07866144e+00 -1.13489866e+00 -2.42762759e-01 7.54471064e-01
-3.89329970e-01 -2.30254248e-01 -1.04343176e+00 -3.58207300e-02
-1.30191725e-02 9.22475100e-01 -6.90027177e-01 -6.19853854e-01
-8.01593065e-01 -3.20862651e-01 6.71491802e-01 9.11438346e-01
1.13470346e-01 -1.71308637e+00 -1.02018154e+00 9.70896840e-01
-3.04908417e-02 -9.39585805e-01 6.75447285e-02 7.19166815e-01
-1.02167106e+00 -1.15722477e+00 -9.40248966e-01 -9.59842920e-01
-9.40562692e-03 -6.60527825e-01 5.27467728e-01 -7.04528093e-01
-5.34640908e-01 2.15479702e-01 -1.57368168e-01 -5.33575416e-01
-1.17178082e-01 1.50488228e-01 2.83143282e-01 -3.67895424e-01
1.13986564e+00 -1.01729631e+00 -5.48634946e-01 -8.10360759e-02
-2.58462995e-01 8.75855684e-02 6.79713488e-01 9.19165194e-01
4.82893914e-01 -5.54685235e-01 9.51336741e-01 -2.30205804e-02
1.15428138e+00 -4.17631745e-01 -3.68809700e-01 2.28497952e-01
-7.96811938e-01 3.73419791e-01 6.43779159e-01 -8.49178076e-01
-7.03992963e-01 1.10304326e-01 -2.07237676e-01 -3.25889707e-01
-5.26769608e-02 3.08989763e-01 2.90350676e-01 6.47345558e-04
9.01986063e-01 5.47775745e-01 1.58889033e-02 -4.53312993e-01
1.25617892e-01 1.05893803e+00 1.13252115e+00 3.67152318e-03
-4.00981158e-02 1.42374873e-01 -2.15327591e-01 -8.51154685e-01
2.42306888e-01 -4.80195805e-02 -1.20310116e+00 -4.62280989e-01
9.53732550e-01 -8.12899411e-01 -7.36790061e-01 9.66013551e-01
-1.50077534e+00 -8.05387259e-01 2.01669335e-01 1.11332405e+00
-1.20565152e+00 -2.35869721e-01 -6.08518958e-01 -5.94117165e-01
-6.52019799e-01 -1.00286996e+00 5.19134045e-01 -1.93469003e-01
-8.53647828e-01 -4.10047829e-01 4.68970537e-01 -5.68855882e-01
5.31926632e-01 1.81240980e-02 7.26740777e-01 -5.41608572e-01
1.65471554e-01 -4.70607191e-01 1.07543543e-03 3.26258987e-01
-2.58789450e-01 -4.44319248e-01 -6.88078046e-01 -6.26594722e-02
-1.30130231e-01 -2.02910841e-01 6.27164721e-01 6.87941670e-01
1.36856449e+00 -4.07788157e-03 -4.88976091e-01 5.59592903e-01
1.03490055e+00 4.62637603e-01 1.11694527e+00 3.26857477e-01
4.35611248e-01 2.82012403e-01 -7.88583308e-02 4.50268179e-01
5.05734459e-02 8.19020987e-01 7.36020971e-03 6.37348533e-01
-2.47548968e-01 -1.33441746e-01 5.47466934e-01 6.15809083e-01
-6.08707666e-01 2.09020942e-01 -1.01886678e+00 6.52662754e-01
-1.79341590e+00 -1.08539295e+00 -5.29129863e-01 2.09633327e+00
9.13606167e-01 1.58890620e-01 1.38403550e-01 2.99936354e-01
3.81603092e-01 -5.66056848e-01 -1.02112889e+00 -4.80796635e-01
1.24297857e-01 8.67986083e-01 5.69265544e-01 -5.53956144e-02
-7.03190923e-01 1.09292364e+00 6.73892403e+00 5.83116829e-01
-1.36631632e+00 4.28600341e-01 -1.18099108e-01 -6.88366413e-01
4.10150975e-01 -6.95130944e-01 -4.32173699e-01 8.20469379e-01
1.71088398e+00 -3.14113319e-01 1.09537530e+00 4.59120154e-01
4.49650884e-01 -4.29864042e-04 -1.32354498e+00 1.58344841e+00
-2.97115207e-01 -1.29292953e+00 -4.51950639e-01 8.94495845e-02
3.23063135e-01 8.80791783e-01 -9.24426913e-02 1.56570822e-01
-2.89945841e-01 -1.43545270e+00 1.03967071e+00 1.07306314e+00
1.60660219e+00 -6.20656848e-01 2.23042741e-01 2.90791810e-01
-9.95981038e-01 -3.39837998e-01 -7.56557733e-02 -6.62064105e-02
1.68418124e-01 -3.05594176e-01 -2.46851057e-01 -2.58242607e-01
8.04386377e-01 9.65290785e-01 -2.99818873e-01 1.22501707e+00
-2.32745379e-01 7.53650844e-01 -1.77970082e-01 -4.63616580e-01
5.37101030e-02 2.56645113e-01 3.93202126e-01 1.49787664e+00
5.61743498e-01 2.73841113e-01 -1.09961390e+00 8.89300108e-01
2.72036791e-01 -3.41657519e-01 -6.47081554e-01 -3.62634718e-01
2.49969691e-01 6.36947215e-01 -2.47491568e-01 1.75977677e-01
-4.11376238e-01 1.50739503e+00 5.65528035e-01 3.92268568e-01
-8.65379930e-01 -7.65571892e-01 6.76136553e-01 1.42326146e-01
1.78835034e-01 -5.82352400e-01 -7.61086524e-01 -9.13302720e-01
2.28289738e-01 -7.70037413e-01 -1.18031397e-01 -8.22492301e-01
-9.86357629e-01 7.84847558e-01 -1.69119574e-02 -1.18570507e+00
-8.34201157e-01 -1.26346397e+00 -5.74582934e-01 1.15609980e+00
-9.27451968e-01 -5.59438825e-01 -8.83009583e-02 1.05622506e+00
2.57902145e-01 -4.92011726e-01 1.11489534e+00 1.57590106e-01
-2.76360691e-01 5.39078057e-01 5.32900751e-01 8.71400014e-02
3.50947201e-01 -7.81727910e-01 5.13117433e-01 5.12965143e-01
1.04357675e-01 3.33126664e-01 3.23394269e-01 -4.70250189e-01
-1.34564042e+00 -7.47564912e-01 1.05978584e+00 -5.09002209e-01
8.62473547e-01 -2.65868962e-01 -7.40959704e-01 5.49387515e-01
3.13405395e-01 -4.80317295e-01 4.25492823e-01 -2.45288223e-01
-1.75762344e-02 -8.22441466e-03 -8.65199566e-01 7.62227416e-01
1.37202168e+00 -7.87953019e-01 -8.70353401e-01 2.15393320e-01
-3.05408120e-01 8.23559612e-02 -9.36214566e-01 2.14583740e-01
1.16152668e+00 -7.43362188e-01 7.27856934e-01 -8.90157998e-01
4.04193431e-01 5.84957719e-01 2.94142723e-01 -1.64010966e+00
-7.66251326e-01 -7.96361446e-01 -5.69676876e-01 1.98821828e-01
3.17866325e-01 -2.69741148e-01 5.03151357e-01 7.10077703e-01
-1.67111784e-01 -9.17156100e-01 -1.21500397e+00 -1.01985252e+00
4.50170696e-01 -9.95339036e-01 1.93180233e-01 -2.46641301e-02
1.08457923e+00 2.56108463e-01 -2.66153753e-01 -4.70242471e-01
2.46502995e-01 -6.17882729e-01 -5.40813878e-02 -1.29126275e+00
1.78956196e-01 -1.10334063e+00 -7.20804214e-01 -8.21343780e-01
3.31489503e-01 -1.31508446e+00 1.66089207e-01 -1.76567614e+00
4.32429723e-02 4.20196772e-01 -6.85713649e-01 5.14504075e-01
3.89674217e-01 1.68646470e-01 -2.14109033e-01 3.00318331e-01
-1.29452348e-01 6.73395514e-01 8.13682199e-01 -9.38984305e-02
-3.12626660e-01 -1.06915489e-01 -4.02625091e-02 3.01854968e-01
1.27150595e+00 -6.18340373e-01 -1.16063431e-01 -4.78820980e-01
-2.98071325e-01 2.82591343e-01 4.16142166e-01 -1.53967798e+00
4.87318635e-01 3.98994327e-01 8.36494088e-01 -4.78411883e-01
1.70368940e-01 -4.57350105e-01 1.00546837e-01 9.02469516e-01
-5.46369672e-01 -5.39677218e-02 3.06191653e-01 4.34844077e-01
-1.40005313e-02 1.07149750e-01 8.44011426e-01 7.86257237e-02
-7.91316986e-01 5.22090197e-01 -1.22307754e+00 8.38140100e-02
8.34618926e-01 -4.62738752e-01 2.57071294e-02 1.41216815e-02
-8.80513310e-01 -4.01002765e-01 -3.26998472e-01 9.15247381e-01
7.04343677e-01 -1.64312816e+00 -9.23283994e-01 5.62258601e-01
1.31054625e-01 -1.19457793e+00 -1.55843258e-01 1.21520030e+00
-3.40741873e-01 1.10875571e+00 -1.17098451e+00 -3.77296031e-01
-8.04675162e-01 -1.16887065e-02 6.15641594e-01 2.31568329e-03
-1.00889230e+00 1.00118470e+00 -5.00992179e-01 3.51125211e-01
5.10657012e-01 -2.57985711e-01 -4.19047862e-01 -9.01321787e-03
5.91541648e-01 3.45148385e-01 2.41174132e-01 -5.27518332e-01
-6.93693280e-01 3.33955772e-02 7.61919320e-02 -6.17451072e-01
1.67156041e+00 3.73152673e-01 9.44324359e-02 5.77518225e-01
1.30805743e+00 -1.04271770e+00 -1.49831772e+00 1.95403174e-01
3.73357415e-01 1.30793437e-01 4.22227859e-01 -1.26627254e+00
-9.07760620e-01 1.27954626e+00 1.16675329e+00 -4.79740024e-01
9.63902116e-01 -4.01089191e-01 9.21027422e-01 5.86080432e-01
8.10887933e-01 -1.46639466e+00 -1.40215054e-01 6.74639702e-01
1.36425972e+00 -1.04057181e+00 -3.74259651e-01 7.91252375e-01
-6.75030708e-01 1.72882831e+00 4.70952988e-01 -6.49286211e-01
1.09669113e+00 2.72512138e-01 -3.31675708e-01 -3.97868901e-01
-5.60281336e-01 -1.59547746e-01 6.76154613e-01 1.02430546e+00
8.41893911e-01 3.42474282e-01 -8.80213678e-01 1.42544603e+00
-2.41091967e-01 6.48316920e-01 1.40047790e-02 5.97938240e-01
-1.44964486e-01 -7.74596930e-01 1.55532956e-01 9.76692557e-01
-3.93590927e-01 -2.22360194e-01 -3.45332593e-01 7.49145865e-01
-3.04724634e-01 8.40262949e-01 1.28070161e-01 -8.73242021e-01
4.20135409e-01 4.20939803e-01 1.15691543e+00 -2.18164742e-01
-8.02982092e-01 -2.04629183e-01 -6.86945543e-02 -1.07408738e+00
1.21845111e-01 -7.70313621e-01 -1.48977184e+00 -5.76074980e-02
1.03760799e-02 -5.52020073e-01 9.75943208e-01 9.60519314e-01
5.38560808e-01 7.08933711e-01 9.80278254e-02 -1.23264813e+00
-6.85280681e-01 -1.48257339e+00 -6.22221231e-01 1.69829935e-01
2.15330809e-01 -7.43650436e-01 7.41190091e-02 2.95594633e-02] | [12.982634544372559, 3.407255172729492] |
b83ab35d-13a3-445b-aa72-4213d225376a | malafide-a-novel-adversarial-convolutive | 2306.07655 | null | https://arxiv.org/abs/2306.07655v1 | https://arxiv.org/pdf/2306.07655v1.pdf | Malafide: a novel adversarial convolutive noise attack against deepfake and spoofing detection systems | We present Malafide, a universal adversarial attack against automatic speaker verification (ASV) spoofing countermeasures (CMs). By introducing convolutional noise using an optimised linear time-invariant filter, Malafide attacks can be used to compromise CM reliability while preserving other speech attributes such as quality and the speaker's voice. In contrast to other adversarial attacks proposed recently, Malafide filters are optimised independently of the input utterance and duration, are tuned instead to the underlying spoofing attack, and require the optimisation of only a small number of filter coefficients. Even so, they degrade CM performance estimates by an order of magnitude, even in black-box settings, and can also be configured to overcome integrated CM and ASV subsystems. Integrated solutions that use self-supervised learning CMs, however, are more robust, under both black-box and white-box settings. | ['Nicholas Evans', 'Massimiliano Todisco', 'Hemlata Tak', 'Wanying Ge', 'Michele Panariello'] | 2023-06-13 | null | null | null | null | ['adversarial-attack', 'face-swapping', 'speaker-verification'] | ['adversarial', 'computer-vision', 'speech'] | [ 3.27479631e-01 2.98970819e-01 -1.54862218e-02 -5.61112054e-02
-8.73747587e-01 -1.22190499e+00 6.74319208e-01 -1.52558843e-02
-3.67424726e-01 2.16483146e-01 6.91892058e-02 -9.35182631e-01
6.86253160e-02 -2.06346542e-01 -4.68315333e-01 -7.85733879e-01
-4.30076629e-01 -9.01743844e-02 2.11322114e-01 -4.45522487e-01
-1.16776146e-01 7.79232025e-01 -1.24526799e+00 1.60198152e-01
4.60898459e-01 7.47987270e-01 -3.81239861e-01 1.36476481e+00
2.99015760e-01 5.40649414e-01 -1.28590977e+00 -4.90178555e-01
2.82476306e-01 -2.86201209e-01 -6.95095181e-01 -3.56376141e-01
4.71027702e-01 -1.89554185e-01 -6.93149984e-01 1.04561186e+00
9.71813381e-01 -1.56622201e-01 1.28895372e-01 -1.29101288e+00
-2.56220371e-01 9.04125810e-01 2.81386524e-01 3.88283819e-01
5.57346642e-01 5.50449789e-01 5.37181199e-01 -2.02812612e-01
2.24373966e-01 1.45381665e+00 7.87301242e-01 9.78133202e-01
-1.28160918e+00 -1.04351020e+00 -2.12472584e-02 -1.17591275e-02
-1.23624218e+00 -1.11197519e+00 6.46702945e-01 -7.60723427e-02
9.92050409e-01 5.93405366e-01 2.56013989e-01 1.52133334e+00
1.68679565e-01 5.01671791e-01 8.68030131e-01 -3.99413556e-01
7.77815953e-02 2.29899794e-01 -2.39153996e-01 4.42468762e-01
-5.39304782e-03 5.89230001e-01 -2.81442225e-01 -5.98375857e-01
2.19128892e-01 -5.79759300e-01 -4.83116835e-01 -7.46718273e-02
-1.01997936e+00 9.96734798e-01 1.12388255e-02 3.44813347e-01
8.70641172e-02 1.64929062e-01 7.86035836e-01 7.57590175e-01
1.35208026e-01 6.75234616e-01 -6.55357182e-01 -4.51060086e-02
-9.06506777e-01 1.55516714e-01 9.81827855e-01 5.22132099e-01
3.18261385e-01 7.56882668e-01 -1.85118362e-01 4.58135873e-01
2.91632563e-01 9.72193539e-01 4.72861856e-01 -6.40612304e-01
4.66207683e-01 -3.56802672e-01 -2.99630463e-02 -8.19328189e-01
-6.32872283e-01 -6.21603906e-01 -4.50492740e-01 4.36598301e-01
3.38085711e-01 -6.36111677e-01 -9.64124620e-01 1.80086172e+00
1.73118383e-01 2.09798649e-01 1.87618867e-01 5.87097824e-01
4.55402195e-01 3.57592911e-01 -6.99213892e-02 -2.20515847e-01
1.36508882e+00 -5.51727772e-01 -9.87174332e-01 -3.96541208e-01
5.54153204e-01 -1.17456353e+00 7.32681036e-01 4.83758032e-01
-9.53759193e-01 -3.28225672e-01 -1.55431759e+00 5.23152769e-01
-4.30664092e-01 -4.17312801e-01 3.12910110e-01 1.87107897e+00
-1.15864003e+00 5.75531185e-01 -5.89086592e-01 1.98143825e-01
1.22194029e-01 5.91424823e-01 -2.69660622e-01 2.98252225e-01
-1.75037384e+00 1.00110912e+00 2.00096428e-01 -9.81307700e-02
-1.32517910e+00 -6.67871356e-01 -9.80645955e-01 6.28673136e-02
1.53219374e-02 -3.10193211e-01 1.25785851e+00 -6.77607656e-01
-2.04641366e+00 6.60080850e-01 2.35871390e-01 -9.12567198e-01
6.61311805e-01 -1.51139587e-01 -1.37094617e+00 2.73706019e-01
-5.43086767e-01 1.54326379e-01 1.73827934e+00 -9.39525485e-01
-4.16262522e-02 3.55724432e-02 -1.46460189e-02 -5.16574979e-01
-5.29393911e-01 5.11313438e-01 1.98286071e-01 -8.05602252e-01
-5.01435436e-02 -1.00692987e+00 -1.19396366e-01 -4.35627013e-01
-4.30773824e-01 3.27976853e-01 1.37299132e+00 -1.05493152e+00
1.25298882e+00 -2.36398983e+00 -2.68919259e-01 2.03345627e-01
-1.59934372e-01 1.01929426e+00 -3.06709379e-01 3.62347841e-01
-3.77177447e-01 4.68779027e-01 -1.37208283e-01 -4.43035156e-01
1.92748811e-02 -3.38558704e-02 -3.98272276e-01 9.40099478e-01
1.87195092e-01 5.43816566e-01 -6.62883699e-01 -1.44170374e-01
3.98270071e-01 7.34669387e-01 -5.19266605e-01 2.02642128e-01
3.65192443e-02 5.37392378e-01 1.50758162e-01 2.50796586e-01
9.67015922e-01 5.88932276e-01 7.94633925e-02 1.93994209e-01
2.10921869e-01 7.38668203e-01 -1.31599069e+00 1.13793564e+00
-6.81668758e-01 8.76471758e-01 7.55855322e-01 -7.60676444e-01
8.92872691e-01 8.74846399e-01 2.69983619e-01 -3.37855101e-01
3.53793234e-01 1.01124696e-01 1.23358943e-01 -4.13059980e-01
7.16898516e-02 -1.03723481e-01 -1.97138354e-01 5.01533568e-01
2.18307734e-01 -2.72750974e-01 -4.53354716e-01 1.72386453e-01
1.33259881e+00 -5.76728284e-01 1.60820797e-01 -1.54526860e-01
7.84391880e-01 -7.24402487e-01 4.98016953e-01 8.88694108e-01
-6.41809046e-01 4.69531149e-01 1.75032288e-01 1.24378745e-02
-9.55647528e-01 -9.97595012e-01 -2.93850213e-01 9.27200973e-01
-2.06010386e-01 -4.29025292e-01 -1.00041544e+00 -8.21524620e-01
6.78746030e-03 6.45178437e-01 -2.99693376e-01 -6.06037378e-01
-7.59151042e-01 -2.84296304e-01 1.62939620e+00 -1.42708770e-03
3.52016300e-01 -7.85723448e-01 -2.40815088e-01 3.43920618e-01
-5.39771877e-02 -1.51364219e+00 -7.59316802e-01 2.13498279e-01
-3.39162469e-01 -6.98409200e-01 -3.06042016e-01 -4.37137425e-01
5.70696741e-02 1.64772525e-01 7.72290647e-01 2.99057037e-01
-1.96352541e-01 2.65604973e-01 -3.21701258e-01 -5.66317916e-01
-1.27026987e+00 -8.97787511e-02 5.53405404e-01 9.23849940e-02
-1.38062509e-02 -4.66037095e-01 -3.25701922e-01 4.73168224e-01
-8.04476917e-01 -8.69553208e-01 1.28181115e-01 8.24826658e-01
-5.55375755e-01 3.72149974e-01 7.11425960e-01 -4.88335848e-01
5.26220739e-01 -1.96218625e-01 -6.53960824e-01 -3.99371237e-02
-5.69917500e-01 1.98425129e-01 7.14801073e-01 -9.13390458e-01
-7.43537843e-01 -2.39857033e-01 -7.07105398e-01 -4.84271735e-01
-2.96257228e-01 -1.28653169e-01 -5.18913567e-01 -5.33224523e-01
9.66033995e-01 8.32567066e-02 2.14672804e-01 -2.81636387e-01
3.65814209e-01 9.79354382e-01 6.68542504e-01 -2.04457920e-02
1.25657642e+00 3.29185605e-01 -3.89783144e-01 -9.35060799e-01
-2.43845984e-01 -3.48938435e-01 -3.24225694e-01 1.77767389e-02
5.91001630e-01 -8.34386349e-01 -1.00897455e+00 7.81334817e-01
-1.16999030e+00 -1.43194169e-01 1.14038818e-01 4.27800119e-01
-3.75805736e-01 6.71175241e-01 -8.26825380e-01 -1.12761080e+00
-4.37313646e-01 -9.28490579e-01 6.75939143e-01 -1.77343041e-01
-2.86343306e-01 -1.02144659e+00 -8.86126906e-02 4.51269001e-01
7.97583222e-01 2.00907644e-02 4.36009854e-01 -9.90679681e-01
6.04306068e-03 -7.19082177e-01 3.71052444e-01 7.32679904e-01
1.28837705e-01 3.84592339e-02 -1.57211256e+00 -8.89160812e-01
3.89315933e-01 2.62361504e-02 6.67958736e-01 1.69846952e-01
6.86432540e-01 -8.62335265e-01 -8.04267824e-02 7.09081769e-01
8.79848599e-01 2.04158425e-01 7.48354137e-01 1.83637425e-01
3.78321767e-01 4.84082043e-01 3.27748246e-02 3.63934517e-01
-2.89101481e-01 9.04083252e-01 5.48957884e-01 1.80260465e-02
-2.14468479e-01 -1.96845487e-01 1.07411814e+00 5.56340277e-01
4.74122941e-01 -3.21410179e-01 -6.51795506e-01 2.91115403e-01
-9.29359376e-01 -1.35164845e+00 1.32041201e-01 2.44039631e+00
6.81784809e-01 4.78176594e-01 3.01087976e-01 8.39011192e-01
9.16588902e-01 5.55074573e-01 -2.90415436e-01 -1.01338601e+00
-3.19003910e-01 2.55255044e-01 9.07484472e-01 8.73871505e-01
-1.22134542e+00 9.95390356e-01 7.42151642e+00 7.33422279e-01
-1.50606084e+00 5.05912602e-01 -1.56500209e-02 -2.96291914e-02
-2.05536604e-01 -4.23320755e-02 -5.91058910e-01 4.73874390e-01
1.52232802e+00 1.96343631e-01 6.61701322e-01 5.42295098e-01
1.68441579e-01 6.84796333e-01 -7.35440671e-01 8.18000674e-01
1.36496574e-01 -1.06313765e+00 -5.24913788e-01 9.60099027e-02
2.81137526e-01 1.65877678e-02 4.43635225e-01 2.22441524e-01
5.37865818e-01 -1.12658274e+00 7.36592650e-01 -3.74769300e-01
8.07503223e-01 -9.13943887e-01 6.56340361e-01 2.70756245e-01
-1.01787591e+00 -4.80150521e-01 -7.80146196e-02 1.74222469e-01
3.25971037e-01 5.33516467e-01 -9.93742883e-01 3.14305991e-01
5.15450239e-01 -1.56295970e-01 -3.35008919e-01 6.25911653e-01
-4.87837851e-01 1.06320977e+00 -1.60139427e-01 1.80080220e-01
5.16086668e-02 7.20710635e-01 1.01841176e+00 1.46593463e+00
-1.17008965e-02 -2.93916464e-01 -8.73326231e-03 3.77924889e-01
1.59196541e-01 -3.05811286e-01 -5.44585168e-01 -1.69582054e-01
7.54272044e-01 8.86956811e-01 -4.99224439e-02 -6.38284981e-02
1.71788380e-01 1.06743109e+00 -1.13404281e-01 2.95239657e-01
-8.89125764e-01 -6.16111815e-01 1.13665903e+00 -1.10048220e-01
4.61596191e-01 -3.61418128e-01 -8.75192434e-02 -9.26070750e-01
-2.60154545e-01 -1.48793054e+00 3.31559062e-01 2.12734453e-02
-9.29695845e-01 7.08854616e-01 -5.11942863e-01 -1.28405523e+00
-3.88101280e-01 -5.19019306e-01 -7.46711254e-01 9.07083273e-01
-1.35772669e+00 -8.94352376e-01 4.27372873e-01 7.79230118e-01
2.19542429e-01 -3.76550257e-01 1.14141762e+00 3.02585334e-01
-5.14552891e-01 1.31212783e+00 -1.22390591e-01 2.87266523e-01
6.23625994e-01 -9.20176387e-01 9.33354020e-01 1.17230403e+00
1.54054165e-01 6.12214088e-01 1.30339956e+00 -3.38992804e-01
-1.29170620e+00 -8.14666271e-01 7.70515323e-01 -5.30134737e-01
6.67236030e-01 -7.19977081e-01 -8.82049918e-01 3.24374467e-01
3.21590483e-01 2.12178260e-01 6.68321192e-01 -1.83259919e-02
-1.04613686e+00 1.04491547e-01 -1.53416586e+00 3.59676898e-01
7.60902882e-01 -1.14779317e+00 -2.89354980e-01 3.68592143e-01
1.08384967e+00 -3.15753728e-01 -7.90913939e-01 2.33965948e-01
4.08389181e-01 -9.06066120e-01 1.21552885e+00 -7.38665521e-01
-5.34468472e-01 -1.71776965e-01 -8.96760002e-02 -1.41645885e+00
-2.54755855e-01 -1.66431749e+00 -2.69787610e-01 1.33767116e+00
5.39907396e-01 -9.79082465e-01 4.27517921e-01 -2.98624020e-03
-6.74895421e-02 1.53109059e-01 -1.28308260e+00 -1.30660152e+00
2.01937452e-01 -7.92805731e-01 8.88091683e-01 1.01457393e+00
2.21653298e-01 4.87263687e-03 -8.84303570e-01 9.49518025e-01
6.18990839e-01 -7.78665543e-01 6.49864972e-01 -8.09800446e-01
-5.19789934e-01 -5.18225670e-01 -6.29157305e-01 -6.55903995e-01
3.51665765e-01 -5.79163313e-01 1.42961785e-01 -3.96423817e-01
-7.04842091e-01 -4.13499534e-01 -4.39536661e-01 2.46763751e-01
-1.80522829e-01 6.17051721e-02 2.98907667e-01 -2.65100598e-01
1.83762982e-01 1.04067154e-01 5.91365337e-01 -5.20593524e-01
-1.55913427e-01 4.16991472e-01 -3.45967919e-01 4.60625708e-01
9.27855968e-01 -5.76903820e-01 -2.64247894e-01 -1.39183015e-01
-2.34034166e-01 1.93233296e-01 3.27006936e-01 -1.28805673e+00
1.28279969e-01 3.82679887e-02 3.90347801e-02 2.73059569e-02
3.52588087e-01 -7.15457559e-01 -4.70237955e-02 8.87555540e-01
-4.25738186e-01 -1.12016827e-01 5.22333741e-01 3.90454799e-01
8.62004980e-03 -1.53222054e-01 1.31387603e+00 3.27023029e-01
-1.00937858e-01 2.28734594e-02 -6.94120347e-01 -2.49326646e-01
7.22815156e-01 7.93287605e-02 -5.76110840e-01 -7.56935179e-01
-5.98181486e-01 -2.40900621e-01 2.58349001e-01 6.50575399e-01
4.94425237e-01 -9.51301336e-01 -8.49419713e-01 6.43019497e-01
-8.34742188e-02 -7.54413724e-01 4.40803647e-01 5.13722718e-01
-2.52850920e-01 6.39371455e-01 5.94603196e-02 -4.45884913e-01
-1.57982147e+00 1.15006340e+00 5.60644269e-01 -4.37288806e-02
-5.52467406e-01 9.41055536e-01 -3.25917244e-01 -6.06670976e-01
4.81654227e-01 2.48658329e-01 8.29268340e-03 -2.37843767e-01
1.01734948e+00 1.68422535e-01 4.66856658e-01 -8.77050102e-01
-6.30096674e-01 1.22078471e-01 1.65660337e-01 -2.65939683e-01
7.10513353e-01 -1.88813716e-01 3.12984169e-01 -1.03882998e-01
1.50017655e+00 6.70800328e-01 -1.02614105e+00 -2.36324400e-01
-7.17390701e-02 -4.46607083e-01 2.02062532e-01 -6.87838137e-01
-9.43244040e-01 8.58026087e-01 9.97316539e-01 5.09722292e-01
8.88009071e-01 -2.16442004e-01 8.62671554e-01 -1.38292275e-02
1.41044289e-01 -8.39582920e-01 1.81184560e-02 4.07124311e-01
7.47698188e-01 -9.86256778e-01 -3.34569216e-01 -4.32437509e-01
-4.85185057e-01 8.92763734e-01 6.63351789e-02 2.48807713e-01
7.58667946e-01 7.16046929e-01 6.21366918e-01 2.80606061e-01
-5.80292523e-01 4.60021496e-02 1.13538876e-01 1.03522551e+00
1.41664088e-01 2.44774923e-01 1.73783273e-01 1.66349784e-01
-5.53068995e-01 -8.34501803e-01 5.53711712e-01 7.61710465e-01
-3.08661938e-01 -1.24901474e+00 -9.91217852e-01 -1.51728883e-01
-8.47976923e-01 -2.26892792e-02 -2.61456043e-01 2.33333990e-01
-7.98231214e-02 1.75806475e+00 -3.01245153e-01 -8.84270012e-01
1.48050070e-01 1.38517514e-01 1.06454648e-01 -2.02900276e-01
-8.73878598e-01 -1.68645009e-01 3.63756895e-01 -4.82337922e-01
6.68871158e-04 -6.55784309e-01 -9.41776991e-01 -7.87817538e-01
-7.58224547e-01 7.39712715e-02 1.08980274e+00 1.10882092e+00
2.85715193e-01 4.70119625e-01 1.32936609e+00 -6.14768744e-01
-1.16301763e+00 -1.00173450e+00 -3.43230575e-01 3.78892839e-01
1.07804286e+00 -3.69514704e-01 -9.64950383e-01 -3.08490098e-01] | [14.063446998596191, 5.870436668395996] |
3673859d-3260-416e-8a43-d6371d455215 | german-abusive-language-dataset-with-focus-on | null | null | https://aclanthology.org/2021.konvens-1.26 | https://aclanthology.org/2021.konvens-1.26.pdf | German Abusive Language Dataset with Focus on COVID-19 | null | ['Georg Groh', 'Svenja Räther', 'Maximilian Wich'] | null | null | null | null | konvens-ws-2021-9 | ['abusive-language'] | ['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] | [-7.287423133850098, 3.77508807182312] |
04fa1ec8-b11d-4c99-9584-1608bd67ed25 | adaptive-noisy-matrix-completion | 2203.08340 | null | https://arxiv.org/abs/2203.08340v1 | https://arxiv.org/pdf/2203.08340v1.pdf | Adaptive Noisy Matrix Completion | Low-rank matrix completion has been studied extensively under various type of categories. The problem could be categorized as noisy completion or exact completion, also active or passive completion algorithms. In this paper we focus on adaptive matrix completion with bounded type of noise. We assume that the matrix $\mathbf{M}$ we target to recover is composed as low-rank matrix with addition of bounded small noise. The problem has been previously studied by \cite{nina}, in a fixed sampling model. Here, we study this problem in adaptive setting that, we continuously estimate an upper bound for the angle with the underlying low-rank subspace and noise-added subspace. Moreover, the method suggested here, could be shown requires much smaller observation than aforementioned method. | ['Ilqar Ramazanli'] | 2022-03-16 | null | null | null | null | ['low-rank-matrix-completion'] | ['methodology'] | [ 5.90643823e-01 3.33861321e-01 3.45243901e-01 2.29189500e-01
-1.17413616e+00 -8.13490570e-01 2.36759603e-01 -1.43967852e-01
-4.84977365e-01 7.18698859e-01 2.81057507e-01 -4.10387740e-02
-6.27222002e-01 -2.00926423e-01 -8.25434864e-01 -9.21414733e-01
-3.24200422e-01 2.05219045e-01 -3.92872542e-01 -2.60061562e-01
9.03169364e-02 1.95714429e-01 -1.07067287e+00 -3.09898525e-01
8.26998889e-01 7.50191271e-01 2.13552743e-01 7.92104602e-01
4.53442335e-01 2.58258432e-01 -3.24348241e-01 -1.69910178e-01
8.35652709e-01 -6.41252026e-02 -3.18889230e-01 3.90551090e-01
4.75173593e-01 -1.38570905e-01 -6.65353894e-01 1.36991215e+00
6.26247644e-01 3.93357515e-01 5.45049369e-01 -8.02287102e-01
-4.04593758e-02 5.74952364e-01 -8.93405318e-01 2.39830315e-01
5.86242557e-01 -3.07983965e-01 5.52439213e-01 -1.58590436e+00
6.66945040e-01 9.21190023e-01 6.82420254e-01 2.94010609e-01
-1.64341640e+00 -6.03034675e-01 -7.14373812e-02 -6.39484376e-02
-1.81514323e+00 -8.54427457e-01 8.55139315e-01 -4.49743927e-01
3.18283550e-02 4.45209473e-01 2.33673289e-01 8.91302764e-01
-4.13432837e-01 4.89613354e-01 1.28367066e+00 -4.91337001e-01
3.01749378e-01 -8.43245164e-02 3.02678287e-01 3.23769301e-01
9.49981272e-01 -1.54246435e-01 -3.82977545e-01 -4.90792930e-01
5.79543352e-01 1.94432475e-02 -6.14406884e-01 -3.74955326e-01
-1.33153152e+00 6.06495798e-01 -1.71363339e-01 2.65720665e-01
-4.01142955e-01 1.45755515e-01 7.36473426e-02 4.85671341e-01
1.89201951e-01 1.01236165e-01 -1.38808787e-01 1.52491152e-01
-1.05860245e+00 2.68656373e-01 9.75052357e-01 1.44534075e+00
7.29046285e-01 4.10627276e-01 -7.11429715e-02 8.32022190e-01
1.61967576e-01 9.36976194e-01 -6.71738386e-02 -1.09885979e+00
1.02406037e+00 -2.33764991e-01 5.54314137e-01 -1.27004123e+00
-4.07786578e-01 -8.03330064e-01 -1.39455473e+00 1.12276390e-01
6.65697217e-01 -5.14832020e-01 -2.73087382e-01 1.82292783e+00
4.49176073e-01 5.34593165e-01 1.61651269e-01 1.00964689e+00
1.65061578e-01 4.16053265e-01 -5.52674532e-01 -8.36652756e-01
8.94747794e-01 -4.51229930e-01 -1.06608617e+00 -5.04728444e-02
3.03629994e-01 -1.12312031e+00 6.33218884e-01 9.59057987e-01
-1.19778395e+00 -3.26253206e-01 -1.34583783e+00 2.47852147e-01
2.43191391e-01 6.06845438e-01 2.82956541e-01 6.36777937e-01
-1.10038877e+00 5.67390621e-01 -6.49548888e-01 -2.91670799e-01
-1.94227338e-01 4.24985409e-01 -5.64490557e-01 -4.57187027e-01
-9.06444788e-01 2.19496608e-01 5.86193427e-02 5.60942292e-01
-1.05288780e+00 -2.50620812e-01 -5.57391882e-01 -2.66451508e-01
5.88448346e-01 -6.00615263e-01 7.21886992e-01 -5.81561148e-01
-1.14300740e+00 5.47832370e-01 -9.17288437e-02 -2.93304980e-01
6.19055450e-01 -5.63972235e-01 -3.34921032e-01 1.94306061e-01
1.15704618e-01 -2.80300230e-01 1.34944975e+00 -1.29679346e+00
-1.21322036e-01 -7.36021280e-01 9.61799640e-03 3.58519763e-01
-1.95377320e-01 -3.07633519e-01 -4.20773983e-01 -9.00299966e-01
8.89222860e-01 -1.11966467e+00 -6.42400205e-01 -1.52192861e-01
-3.99321139e-01 3.27541560e-01 6.31977677e-01 -7.04778075e-01
1.33743680e+00 -2.36166644e+00 4.02476400e-01 4.39089060e-01
3.64754617e-01 -7.11715445e-02 1.85187105e-02 9.31417823e-01
-2.04984874e-01 2.18341872e-03 -3.95874560e-01 -4.78283346e-01
-1.96406052e-01 -1.34341642e-01 -5.36882758e-01 1.24295759e+00
-4.08552527e-01 -1.39506876e-01 -7.86580324e-01 -2.45782912e-01
-2.65194979e-02 3.95372719e-01 -4.84019846e-01 5.40130474e-02
4.56905633e-01 5.59998572e-01 -5.11363804e-01 3.92564923e-01
1.11671019e+00 4.14874144e-02 3.73561233e-01 -4.75171715e-01
-1.71304181e-01 -5.92752576e-01 -2.16391730e+00 2.02084327e+00
-3.77770364e-01 4.91283000e-01 1.18334579e+00 -1.17458165e+00
7.04102933e-01 5.64072788e-01 5.91322362e-01 6.76774085e-02
-1.97541099e-02 3.63929093e-01 -7.12231025e-02 -3.64206344e-01
6.10337615e-01 6.26557618e-02 1.13180645e-01 1.39479518e-01
-2.67004043e-01 3.94056469e-01 3.18300545e-01 6.08609557e-01
1.43189549e+00 -2.09575165e-02 4.04837489e-01 -5.69422781e-01
5.62107980e-01 -2.27852494e-01 4.94338483e-01 1.11478186e+00
-5.45861498e-02 7.49095142e-01 6.58003986e-02 5.20142853e-01
-1.10429418e+00 -9.81792688e-01 -2.85849661e-01 7.84665048e-01
3.01472425e-01 -5.03461540e-01 -5.47465205e-01 6.20557889e-02
-2.68463612e-01 6.72687367e-02 -4.42552149e-01 2.75434017e-01
-7.10893393e-01 -7.95027912e-01 4.82236832e-01 -5.87919205e-02
4.37410057e-01 1.95040274e-02 1.56482175e-01 2.18360871e-01
-1.60954550e-01 -1.11837506e+00 -5.85944951e-01 7.51619088e-03
-1.13741767e+00 -1.21785963e+00 -9.19458926e-01 -6.24746561e-01
9.84580040e-01 7.71212876e-01 6.38051629e-01 -2.60069966e-01
-3.87532152e-02 8.47612381e-01 -3.92882258e-01 1.02176160e-01
3.55660439e-01 -3.10054570e-01 6.51431561e-01 5.62448919e-01
-2.83107728e-01 -1.01280189e+00 -9.02323723e-01 2.73516923e-01
-1.17582417e+00 -2.68684536e-01 6.51595354e-01 8.16973329e-01
6.54475212e-01 5.54891899e-02 4.71479625e-01 -1.10930371e+00
6.34713173e-01 -6.63057625e-01 -4.79129255e-01 -1.01629406e-01
-4.36122805e-01 -5.78681976e-02 7.72176445e-01 -4.11622077e-01
-8.97894084e-01 3.64544958e-01 2.28073567e-01 -7.23761559e-01
1.43638164e-01 5.65510392e-01 -4.11280632e-01 -9.75634456e-02
9.24459398e-01 2.59752363e-01 -3.38302732e-01 -8.08209240e-01
5.56799233e-01 4.89133894e-01 5.81679642e-01 -7.96596467e-01
1.63303506e+00 8.46363366e-01 5.29416561e-01 -1.16060901e+00
-6.44390523e-01 -8.86771381e-01 -6.62513673e-01 -3.85303684e-02
2.47876689e-01 -1.18190455e+00 -6.29719377e-01 -9.11874175e-02
-9.47879374e-01 5.63181788e-02 -1.69304892e-01 8.22363377e-01
-6.54053688e-01 9.08708751e-01 -4.13727015e-01 -1.29917717e+00
-2.71561533e-01 -6.81818128e-01 8.19176078e-01 -3.14647257e-01
6.50926381e-02 -8.26972306e-01 3.38895023e-01 2.44267389e-01
2.67462105e-01 4.34786320e-01 1.85660869e-01 -5.33974528e-01
-6.98231816e-01 -3.73170853e-01 -2.47398540e-02 2.78297335e-01
-1.15665324e-01 -7.54936755e-01 -7.17669785e-01 -9.41035867e-01
4.28148866e-01 -1.04775675e-01 6.53277099e-01 3.11167687e-01
9.52672720e-01 -6.12612963e-01 -2.37394288e-01 7.29857624e-01
1.68201566e+00 -2.50591666e-01 4.99435127e-01 -6.25837371e-02
5.67538857e-01 4.28098500e-01 8.71423066e-01 8.82683575e-01
-3.00130635e-01 6.39036775e-01 1.57644063e-01 1.86451182e-01
1.42003849e-01 6.13870323e-02 3.96490365e-01 1.34559202e+00
-2.66670704e-01 -6.12686295e-03 -6.28645360e-01 3.48506957e-01
-2.07617974e+00 -1.00045252e+00 -4.21981603e-01 2.57156229e+00
5.02126932e-01 -2.38227546e-01 -8.97769257e-02 4.86659169e-01
9.66076314e-01 1.19668871e-01 -3.26568574e-01 3.22213918e-01
-4.98001337e-01 6.86889812e-02 9.35416818e-01 8.04693878e-01
-9.13304150e-01 3.14901710e-01 6.14181852e+00 1.10108197e+00
-5.70995986e-01 3.16163808e-01 7.94266462e-02 -5.43229841e-02
-2.27121115e-01 4.27268028e-01 -7.04576135e-01 -2.06392119e-03
5.11784077e-01 -1.79401994e-01 8.02605093e-01 7.86701500e-01
5.65444469e-01 -7.60059431e-02 -1.09079814e+00 1.47149694e+00
3.28075022e-01 -7.62773871e-01 -1.55092791e-01 2.49827713e-01
8.76683354e-01 -4.01925176e-01 1.30140036e-01 2.51294613e-01
1.01794682e-01 -8.43239605e-01 3.59308332e-01 7.79381633e-01
1.05949903e+00 -4.02142256e-01 3.63275290e-01 4.66601193e-01
-1.41791892e+00 -4.93017584e-02 -4.79668498e-01 -2.23012939e-01
2.52465278e-01 9.77851689e-01 -3.25162679e-01 8.79243910e-01
1.62663355e-01 6.26746356e-01 -1.93183422e-01 1.19377661e+00
2.06308514e-01 8.53150189e-01 -5.97472966e-01 4.97526467e-01
-1.82277754e-01 -8.12762499e-01 1.21535826e+00 1.07497060e+00
7.66915619e-01 5.99309981e-01 6.61161542e-01 2.06311703e-01
7.24177510e-02 3.90544593e-01 -8.96204054e-01 2.71633863e-01
4.40008461e-01 1.36506641e+00 -4.29894388e-01 -3.12123984e-01
-2.78122276e-01 9.45042729e-01 2.21234322e-01 8.29019964e-01
-4.04270738e-01 -4.80288237e-01 1.87303334e-01 3.43558729e-01
3.65814269e-02 -7.63860464e-01 -2.50211895e-01 -1.50345302e+00
1.98341444e-01 -8.23871374e-01 3.27035993e-01 -4.39820439e-01
-1.07575715e+00 3.33693713e-01 -6.97397068e-02 -1.64664900e+00
5.89458607e-02 -3.54001552e-01 -2.81303465e-01 8.57079625e-01
-6.63172483e-01 -6.78892851e-01 -3.61859888e-01 9.29133594e-01
4.17187035e-01 -1.18138991e-01 4.60409313e-01 7.11072564e-01
-6.60680532e-01 3.88396263e-01 8.79264355e-01 -4.44914214e-02
8.14527333e-01 -1.11530161e+00 -5.65756619e-01 1.18306398e+00
-1.04785748e-01 1.05233848e+00 1.27660048e+00 -7.78469205e-01
-2.16370630e+00 -1.06768346e+00 1.66513756e-01 -2.92578459e-01
8.19891334e-01 -4.72542018e-01 -5.04623771e-01 5.84103942e-01
8.83992985e-02 8.91154706e-02 5.69369674e-01 1.47565857e-01
-2.13162914e-01 -3.01975310e-01 -1.07572854e+00 4.21210885e-01
1.32965255e+00 -4.66420949e-01 -1.58668712e-01 6.19981349e-01
3.62842649e-01 -3.82721603e-01 -9.06131327e-01 5.31810999e-01
2.05887571e-01 -5.31152010e-01 1.07176173e+00 -4.33783740e-01
-1.86452210e-01 -7.29471028e-01 -8.60199094e-01 -1.01127219e+00
-4.51854944e-01 -1.25970352e+00 -2.81603247e-01 1.13398397e+00
1.48049802e-01 -2.53186852e-01 9.54541802e-01 3.01649183e-01
-1.58012763e-01 -3.13582331e-01 -1.16982996e+00 -8.23358715e-01
-3.26250732e-01 -4.68433350e-01 -2.15294138e-01 8.52191508e-01
-7.55661130e-02 4.99580383e-01 -9.79674280e-01 5.56082606e-01
1.29694700e+00 -3.31751496e-01 1.09445572e+00 -9.23696697e-01
-7.22847521e-01 1.13961726e-01 -4.99426834e-02 -1.47805929e+00
-2.00180396e-01 -8.25090706e-01 -2.98624653e-02 -1.20479214e+00
1.44971490e-01 -5.78737736e-01 -4.01554480e-02 -2.49181062e-01
-1.15751706e-01 3.78692329e-01 1.36622116e-01 4.94848222e-01
-7.14856207e-01 4.96948034e-01 1.06945419e+00 -8.29857364e-02
-1.23931870e-01 1.01190910e-01 -4.71463025e-01 5.80512762e-01
4.66067880e-01 -3.28249574e-01 -6.18100047e-01 -2.01980591e-01
5.28405607e-01 7.30994403e-01 2.62385607e-02 -1.11269736e+00
4.67164278e-01 -2.32809363e-03 1.92192614e-01 -5.97115576e-01
6.66488945e-01 -8.94224882e-01 5.73880434e-01 1.35889441e-01
-3.11468124e-01 7.33908638e-03 -3.31056237e-01 1.42837560e+00
-4.62564752e-02 -5.38873434e-01 6.45190895e-01 -2.11389754e-02
-1.16646759e-01 4.51278329e-01 -4.47269469e-01 1.30716816e-01
7.17709541e-01 -3.26817989e-01 -6.64177118e-03 -7.55841911e-01
-1.44117093e+00 -1.22904457e-01 2.75908589e-01 -2.66649723e-01
5.20849705e-01 -1.28243399e+00 -8.70097816e-01 -2.69672573e-01
-1.74040496e-02 -1.48554653e-01 5.74501634e-01 1.18555188e+00
-2.53122061e-01 3.12554240e-01 5.26793659e-01 -4.67896640e-01
-1.05749834e+00 7.05891728e-01 -1.16456896e-01 -1.99614152e-01
-3.69555444e-01 2.18732297e-01 1.11629687e-01 -8.29745829e-02
3.31582904e-01 3.35970223e-01 -1.33483514e-01 -6.68318644e-02
4.79572177e-01 9.16734278e-01 -6.45494387e-02 -8.24243724e-01
1.07045613e-01 6.85073256e-01 1.59696132e-01 -5.61209857e-01
9.45400953e-01 -6.53767049e-01 -1.97329298e-01 4.52851027e-01
1.26516819e+00 8.32135677e-01 -9.54355061e-01 -3.82849514e-01
-4.60597500e-02 -6.74475431e-01 -1.78769752e-01 -9.14422572e-02
-7.40297556e-01 4.95577306e-01 9.36476767e-01 1.13936506e-01
1.07753980e+00 -5.49202800e-01 2.70573825e-01 7.69981682e-01
7.79719174e-01 -1.05989945e+00 6.02945685e-02 3.22360128e-01
1.02601826e+00 -1.17398524e+00 1.51744083e-01 -7.14929223e-01
1.07885124e-02 9.05491829e-01 4.08153594e-01 -6.23615682e-01
8.30006897e-01 2.54326295e-02 -3.24545205e-01 -7.46419560e-03
-3.60811979e-01 -3.32034439e-01 1.07263237e-01 6.74117327e-01
4.33866113e-01 4.55710180e-02 -8.84635448e-01 5.11086166e-01
4.02322933e-02 -3.04857999e-01 9.10781741e-01 7.63622105e-01
-4.87629801e-01 -1.12869000e+00 -1.08761060e+00 4.48698848e-01
-4.81502712e-01 -2.21757874e-01 1.57489590e-02 5.28040707e-01
-1.05419219e-01 1.17129970e+00 -6.86001420e-01 -2.85293847e-01
3.66952568e-01 -3.89762074e-01 4.07476038e-01 -6.14599466e-01
-1.55099168e-01 5.13083756e-01 2.16803402e-01 -2.65557230e-01
-3.40467006e-01 -8.79782259e-01 -6.68566346e-01 -4.64398190e-02
-2.04269528e-01 5.76411068e-01 4.90112901e-01 5.46702743e-01
1.40268117e-01 -1.35034829e-01 8.97146285e-01 -6.71168625e-01
-8.17909420e-01 -1.23796964e+00 -1.13918447e+00 5.22184968e-01
3.17749530e-01 -3.87251019e-01 -7.37006783e-01 -1.89737916e-01] | [6.976259231567383, 4.638597011566162] |
9acdf2b1-50d9-451d-98f6-ae84f2033595 | mmformer-multimodal-medical-transformer-for | 2206.02425 | null | https://arxiv.org/abs/2206.02425v2 | https://arxiv.org/pdf/2206.02425v2.pdf | mmFormer: Multimodal Medical Transformer for Incomplete Multimodal Learning of Brain Tumor Segmentation | Accurate brain tumor segmentation from Magnetic Resonance Imaging (MRI) is desirable to joint learning of multimodal images. However, in clinical practice, it is not always possible to acquire a complete set of MRIs, and the problem of missing modalities causes severe performance degradation in existing multimodal segmentation methods. In this work, we present the first attempt to exploit the Transformer for multimodal brain tumor segmentation that is robust to any combinatorial subset of available modalities. Concretely, we propose a novel multimodal Medical Transformer (mmFormer) for incomplete multimodal learning with three main components: the hybrid modality-specific encoders that bridge a convolutional encoder and an intra-modal Transformer for both local and global context modeling within each modality; an inter-modal Transformer to build and align the long-range correlations across modalities for modality-invariant features with global semantics corresponding to tumor region; a decoder that performs a progressive up-sampling and fusion with the modality-invariant features to generate robust segmentation. Besides, auxiliary regularizers are introduced in both encoder and decoder to further enhance the model's robustness to incomplete modalities. We conduct extensive experiments on the public BraTS $2018$ dataset for brain tumor segmentation. The results demonstrate that the proposed mmFormer outperforms the state-of-the-art methods for incomplete multimodal brain tumor segmentation on almost all subsets of incomplete modalities, especially by an average 19.07% improvement of Dice on tumor segmentation with only one available modality. The code is available at https://github.com/YaoZhang93/mmFormer. | ['Yefeng Zheng', 'Zhiqiang He', 'Yang Zhang', 'Yawen Huang', 'Dong Wei', 'Yuexiang Li', 'Jiawei Yang', 'Nanjun He', 'Yao Zhang'] | 2022-06-06 | null | null | null | null | ['brain-tumor-segmentation'] | ['medical'] | [ 3.90851796e-01 6.76367506e-02 -3.30044150e-01 -4.93678778e-01
-1.64995623e+00 -3.56564760e-01 3.39003354e-01 -4.51987460e-02
-5.09699523e-01 6.10905051e-01 3.93040448e-01 -3.08928132e-01
-1.07870966e-01 -3.06732565e-01 -5.64092755e-01 -1.07326937e+00
1.45894855e-01 5.17960072e-01 2.69265343e-02 5.71869835e-02
-1.74376264e-01 2.35566482e-01 -9.05665219e-01 5.36315203e-01
1.08307362e+00 1.16427040e+00 6.12880230e-01 4.77781504e-01
-1.48075417e-01 7.49319553e-01 1.27048651e-02 -2.63883621e-01
4.93265502e-03 -3.00889611e-01 -9.37514424e-01 4.26535577e-01
2.95308322e-01 -4.23890412e-01 -4.81503516e-01 1.15970337e+00
5.46887100e-01 -2.52991229e-01 6.64884031e-01 -1.04904211e+00
-1.46702990e-01 8.31631720e-01 -8.77591491e-01 1.24909364e-01
-5.50931320e-02 2.85836697e-01 7.85803199e-01 -8.23649764e-01
6.10741436e-01 6.57916129e-01 5.09432614e-01 5.34763873e-01
-9.62754369e-01 -6.11521065e-01 -4.31315787e-02 1.08203679e-01
-1.27915907e+00 -2.69857138e-01 7.85541475e-01 -4.62102145e-01
5.13487279e-01 3.09948713e-01 4.71593291e-01 1.10619414e+00
4.31428701e-01 1.02840185e+00 1.28905499e+00 1.36683893e-03
-1.31702408e-01 -2.50833422e-01 2.52271503e-01 9.62276101e-01
-1.30233198e-01 7.80206174e-04 -4.20719892e-01 -1.11716874e-01
6.04346037e-01 1.16704963e-01 -4.07938659e-01 -2.74248958e-01
-1.65515745e+00 6.40225708e-01 6.94193482e-01 5.60663402e-01
-4.90478307e-01 1.10110538e-02 4.13640827e-01 -1.23935100e-03
2.48052299e-01 -4.79055904e-02 -2.52330035e-01 1.39604792e-01
-1.16414297e+00 -2.98820734e-01 4.33681220e-01 7.67848194e-01
4.60982829e-01 -2.84662902e-01 -2.11225227e-01 9.79607105e-01
4.69121039e-01 7.86032021e-01 7.17722178e-01 -7.73818135e-01
5.59153080e-01 4.61647004e-01 -5.44952869e-01 -4.29515630e-01
-7.50514805e-01 -6.24033391e-01 -1.21338332e+00 -3.17049921e-01
5.29073775e-01 -1.71808586e-01 -1.25215471e+00 1.87413394e+00
2.22122341e-01 3.02131146e-01 -1.49321795e-01 9.40077007e-01
1.21021068e+00 1.25271559e-01 1.93286553e-01 -1.81763142e-01
1.60670507e+00 -1.11010778e+00 -7.40177691e-01 -1.43998131e-01
7.41076767e-01 -8.34851980e-01 6.83257699e-01 4.83099706e-02
-1.12931526e+00 -1.55944854e-01 -6.70443952e-01 -8.81980988e-04
-6.34137020e-02 1.48980781e-01 6.25945747e-01 4.85196143e-01
-1.00413930e+00 4.48578037e-02 -1.27494073e+00 -5.07309241e-03
7.00718284e-01 5.95829904e-01 -7.04077601e-01 -4.29140955e-01
-1.01057792e+00 8.88294339e-01 3.17740738e-01 1.96260139e-01
-1.22174847e+00 -1.00911450e+00 -9.58639860e-01 -3.04495245e-01
3.41302454e-01 -8.65484655e-01 1.11196649e+00 -9.09840941e-01
-1.13814259e+00 8.12927663e-01 -4.63445097e-01 -1.26411244e-01
5.70257306e-01 2.88829684e-01 -2.90861756e-01 5.25473952e-01
1.37102887e-01 9.52404499e-01 7.95195222e-01 -1.33646619e+00
-5.34385204e-01 -6.38828397e-01 -5.00370920e-01 4.48752731e-01
-2.25116685e-01 -2.90131360e-01 -6.28644824e-01 -5.74743152e-01
4.28505480e-01 -1.04292142e+00 -3.13979805e-01 -2.54118592e-01
-7.36916065e-01 1.84551254e-01 6.33855879e-01 -1.12349284e+00
9.95391846e-01 -1.97969353e+00 4.94298279e-01 4.31536049e-01
4.50031579e-01 -2.39317805e-01 -3.76557082e-01 -3.50741409e-02
-3.06430072e-01 -1.17613129e-01 -7.07406819e-01 -6.84680283e-01
-1.47494763e-01 2.99031198e-01 2.65194535e-01 6.25257432e-01
-1.36418611e-01 1.19399750e+00 -6.85352027e-01 -7.67042756e-01
2.80716985e-01 5.36420941e-01 -2.88840473e-01 2.66216286e-02
2.63251662e-01 1.02595985e+00 -4.85422432e-01 1.17896008e+00
7.43630052e-01 -3.92178982e-01 1.51725426e-01 -5.51747918e-01
2.20463753e-01 -1.46894753e-01 -6.67923450e-01 2.25068688e+00
-3.02278310e-01 1.48831412e-01 3.80410612e-01 -1.15679753e+00
3.51801634e-01 6.67224526e-01 1.21385884e+00 -7.30952859e-01
3.72191876e-01 6.32903159e-01 1.21190742e-01 -5.57068467e-01
5.72395772e-02 -4.31325614e-01 -2.04534587e-02 1.83128953e-01
2.16341347e-01 -2.72895247e-01 1.35205090e-01 1.96550429e-01
1.16166389e+00 -3.50423008e-01 -1.72097325e-01 -4.62313481e-02
7.24627495e-01 -1.55253068e-01 4.76135314e-01 5.26573956e-01
-3.80573601e-01 8.82970750e-01 3.32509100e-01 2.92877611e-02
-7.74241805e-01 -1.24170637e+00 -4.80680525e-01 6.70512438e-01
1.54024050e-01 7.74242654e-02 -6.89365149e-01 -9.58342433e-01
-2.56480873e-01 4.55890328e-01 -6.35162234e-01 7.91405514e-03
-4.75169420e-01 -1.15331876e+00 5.44236302e-01 5.49219966e-01
5.50158024e-01 -6.35763168e-01 -2.98760027e-01 -4.56212759e-02
-7.96415091e-01 -1.27832651e+00 -7.42544472e-01 3.19043636e-01
-1.14499331e+00 -1.12653387e+00 -1.12895954e+00 -6.77044630e-01
9.99666870e-01 9.89719182e-02 8.47637951e-01 -3.73770036e-02
-2.36277968e-01 5.03142536e-01 -2.12990135e-01 6.96317479e-02
-4.37693298e-01 3.12323004e-01 -5.09852827e-01 6.17076047e-02
-1.51934717e-02 -3.69058251e-01 -6.18516147e-01 2.80045897e-01
-1.21733332e+00 4.41886336e-01 1.00110161e+00 1.24597943e+00
7.40479290e-01 -1.99953035e-01 3.57659817e-01 -6.97776616e-01
1.74982965e-01 -6.50858700e-01 -1.02320284e-01 3.83987755e-01
-1.39163554e-01 -4.27274667e-02 1.23293452e-01 -1.78732827e-01
-1.02456331e+00 3.95221323e-01 -4.33676243e-01 -2.28404641e-01
-2.75774598e-01 8.96452069e-01 -1.26351222e-01 -2.25244701e-01
1.04007870e-01 3.72215182e-01 2.95786589e-01 -2.46937662e-01
2.92312503e-01 4.51431811e-01 5.46062648e-01 -5.61856329e-01
5.34143925e-01 6.04006588e-01 1.94622710e-01 -6.13787115e-01
-7.45990098e-01 -5.94375372e-01 -7.73861468e-01 -2.54246503e-01
9.88093555e-01 -9.17589009e-01 -3.09104353e-01 6.91919386e-01
-7.49203265e-01 -3.71306419e-01 -8.89626071e-02 7.94065714e-01
-6.54034734e-01 5.42894602e-01 -8.74114513e-01 3.27918567e-02
-4.71144289e-01 -2.03067183e+00 1.30258965e+00 1.21570028e-01
8.55602548e-02 -1.17850137e+00 -1.45666659e-01 9.50412989e-01
4.59950089e-01 2.15086147e-01 8.71361554e-01 -6.48466349e-01
-4.99863327e-01 -1.07308604e-01 -2.78202385e-01 3.08093101e-01
2.91014045e-01 -4.36152607e-01 -8.41375411e-01 -4.59762573e-01
-1.29319683e-01 -3.87106240e-01 1.15325773e+00 8.35035086e-01
1.19968069e+00 2.91560829e-01 -4.40221488e-01 7.84649074e-01
1.24512434e+00 3.62857990e-02 5.12358069e-01 -7.39891222e-03
9.05583382e-01 3.56699586e-01 1.72038257e-01 2.75658399e-01
6.66526675e-01 4.48883861e-01 6.60774946e-01 -3.77067894e-01
-3.81425589e-01 8.01593438e-02 1.26180515e-01 1.15068877e+00
6.14261739e-02 -1.66592968e-03 -1.14533842e+00 6.15936100e-01
-1.64831102e+00 -6.46529973e-01 -3.43964063e-02 2.09326124e+00
9.25588906e-01 -3.43673378e-01 -1.79422557e-01 -1.59372926e-01
5.57984591e-01 4.78772260e-02 -5.96209884e-01 3.71570110e-01
-2.14664519e-01 1.17464885e-02 6.09091520e-01 5.91540515e-01
-1.27439284e+00 5.33730149e-01 5.42481232e+00 1.13212550e+00
-1.19222188e+00 5.18681049e-01 8.57512355e-01 -3.47213037e-02
-5.18842340e-01 -2.08845928e-01 -3.67902368e-01 4.63293225e-01
6.57096624e-01 3.41197461e-01 1.95206359e-01 2.04549775e-01
9.55950841e-02 -3.21104527e-01 -9.27454650e-01 1.04790986e+00
1.05174661e-01 -1.19688308e+00 -3.66625302e-02 3.21384281e-01
8.41334045e-01 4.21198934e-01 3.07546616e-01 -3.06750592e-02
-1.13943443e-01 -1.14319849e+00 4.98927712e-01 8.05241525e-01
7.00589538e-01 -6.65384710e-01 1.07935297e+00 3.29303116e-01
-1.10451794e+00 2.27517132e-02 2.21572533e-01 9.05167699e-01
3.43289256e-01 7.17474520e-01 -7.00345814e-01 9.42510426e-01
3.09959859e-01 8.74071360e-01 -6.96172416e-01 1.09563589e+00
2.77686790e-02 5.60929716e-01 -2.95888305e-01 6.77244604e-01
3.20536733e-01 2.47384608e-02 6.29864991e-01 1.07955635e+00
5.42314351e-01 2.70732418e-02 3.37905645e-01 5.03826082e-01
-1.07426248e-01 -2.39263400e-02 -2.14308456e-01 7.09501281e-02
1.96665481e-01 1.53282523e+00 -7.06479669e-01 -3.37694138e-01
-6.44103289e-01 7.68356562e-01 2.67003458e-02 4.13923919e-01
-1.05310965e+00 2.39200205e-01 9.68590528e-02 -1.35266855e-01
2.85885371e-02 -1.11278966e-01 -6.68287098e-01 -1.35099757e+00
-1.65267348e-01 -9.90029812e-01 7.27525771e-01 -5.46082497e-01
-1.45741117e+00 7.01560974e-01 2.40343045e-02 -1.14714766e+00
-4.58956063e-02 -4.40525800e-01 -3.16884756e-01 8.50153029e-01
-1.65833950e+00 -1.72593093e+00 -3.38056773e-01 9.78801906e-01
3.46801251e-01 -1.28014773e-01 6.80042505e-01 4.93607551e-01
-5.44038177e-01 6.59077466e-01 2.22877972e-02 1.87867895e-01
7.90468454e-01 -1.17344022e+00 -6.87569201e-01 6.56525970e-01
-2.28222832e-01 4.04932141e-01 2.44421527e-01 -6.80890620e-01
-1.55566478e+00 -7.66789138e-01 5.30877829e-01 -1.33759707e-01
7.66026199e-01 2.43333504e-01 -7.57297218e-01 7.26863861e-01
4.50534970e-01 1.96108460e-01 9.70649660e-01 -2.56195217e-01
-7.11746439e-02 -6.65659681e-02 -1.19733715e+00 4.25172240e-01
6.41918063e-01 -5.19656718e-01 -3.77830178e-01 3.92587960e-01
3.16140175e-01 -7.81088650e-01 -1.37693918e+00 8.54217410e-01
4.19373602e-01 -6.54574990e-01 9.44521248e-01 -1.10680744e-01
5.18146217e-01 -2.54989237e-01 -3.73297453e-01 -1.27033901e+00
4.47772769e-03 -1.25696138e-01 1.96507514e-01 7.95611978e-01
6.31684959e-01 -5.90324759e-01 5.91214240e-01 7.01408684e-01
-6.42220140e-01 -9.82872009e-01 -1.32633865e+00 -2.64187336e-01
3.41146946e-01 -6.76026046e-01 2.81961858e-01 9.52174962e-01
1.31799802e-01 -9.65620503e-02 -1.40695706e-01 2.43906006e-01
8.53087962e-01 2.99974587e-02 6.24181516e-02 -6.36754394e-01
-1.73079625e-01 -6.85914278e-01 -1.24968648e-01 -8.89325380e-01
2.43703634e-01 -1.37152934e+00 2.23163124e-02 -1.62925422e+00
8.98365140e-01 -3.82438362e-01 -6.49925292e-01 7.26911485e-01
-1.63837850e-01 3.19538921e-01 8.91542584e-02 1.50737152e-01
-5.68175256e-01 6.15006745e-01 1.93585980e+00 -5.22462785e-01
1.98589742e-01 -1.39205456e-01 -5.49646020e-01 6.83462501e-01
6.81013823e-01 -3.12202513e-01 -3.52805823e-01 -5.11256933e-01
-4.00240511e-01 5.47786653e-01 5.20505011e-01 -8.75911832e-01
4.12867188e-01 5.55700297e-03 4.62469637e-01 -7.27509856e-01
2.90987849e-01 -9.64628637e-01 1.31087705e-01 5.10331690e-01
-4.08322603e-01 -9.79857296e-02 1.98661491e-01 2.87390500e-01
-5.47179699e-01 -1.52181610e-01 9.73541975e-01 -1.52711824e-01
-4.30752575e-01 7.15798259e-01 -2.59015411e-01 5.19690395e-04
8.83629799e-01 1.86563089e-01 -3.96895617e-01 -1.97086602e-01
-1.19743860e+00 4.39112663e-01 1.98802054e-01 2.04303607e-01
7.97475994e-01 -1.34041595e+00 -8.06304812e-01 5.32422625e-02
-7.35779777e-02 -5.68785407e-02 7.73799837e-01 1.95406973e+00
-2.77776510e-01 3.82373780e-01 -1.85656980e-01 -8.69346499e-01
-1.27374983e+00 1.15623511e-01 6.27365410e-01 -6.79647863e-01
-4.84914869e-01 8.13084364e-01 4.10848647e-01 -6.02527559e-01
1.66716039e-01 -3.83084267e-01 -1.49062870e-03 -2.65898854e-02
2.85318315e-01 -5.97388344e-03 1.63450047e-01 -1.10169232e+00
-4.79684651e-01 5.09132504e-01 -1.67961776e-01 -3.62614900e-01
1.15763199e+00 -2.29613394e-01 -3.83076310e-01 1.27695218e-01
1.33595562e+00 -2.91209280e-01 -1.20067286e+00 -4.83059525e-01
-2.29997590e-01 -1.66575477e-01 3.77310932e-01 -1.12073290e+00
-1.69249654e+00 9.89154220e-01 7.63031125e-01 -2.76594490e-01
1.46363103e+00 1.81610376e-01 1.14393067e+00 -6.90760687e-02
2.39842966e-01 -7.91707218e-01 6.13728389e-02 4.67094809e-01
8.38241935e-01 -1.51976788e+00 -1.78425282e-01 -4.03825611e-01
-8.94829869e-01 1.08736479e+00 4.05299753e-01 4.19475704e-01
8.48207653e-01 4.64858949e-01 1.54755846e-01 -3.56473655e-01
-5.02598226e-01 -2.38108680e-01 6.73341513e-01 3.33363950e-01
5.21407008e-01 3.90796691e-01 -2.27813914e-01 7.09313095e-01
1.33055180e-01 -2.13512387e-02 1.48125216e-01 7.65744805e-01
-2.15750948e-01 -1.22084153e+00 -3.51364821e-01 5.03322065e-01
-4.80686396e-01 -1.85533717e-01 6.56288043e-02 6.60267889e-01
2.18945056e-01 7.93478727e-01 -2.65502632e-01 -2.24467665e-01
-2.67128367e-03 5.03503270e-02 6.01816416e-01 -3.08245718e-01
-4.04894501e-01 6.20391250e-01 -9.28086489e-02 -5.62595487e-01
-5.65390527e-01 -9.93497491e-01 -1.47150552e+00 5.05596176e-02
-1.14156045e-01 -1.56121165e-01 6.54354036e-01 1.17616296e+00
2.20654607e-02 7.87288606e-01 6.19632065e-01 -7.23154783e-01
-3.38453054e-01 -9.02270019e-01 -5.53945541e-01 4.29831624e-01
4.62266207e-01 -6.01282716e-01 -1.75375178e-01 2.97766011e-02] | [14.505309104919434, -2.310826063156128] |
efd55d4e-22fb-4b4b-8fab-cfda2acb23b1 | justices-for-information-bottleneck-theory | 2305.11387 | null | https://arxiv.org/abs/2305.11387v1 | https://arxiv.org/pdf/2305.11387v1.pdf | Justices for Information Bottleneck Theory | This study comes as a timely response to mounting criticism of the information bottleneck (IB) theory, injecting fresh perspectives to rectify misconceptions and reaffirm its validity. Firstly, we introduce an auxiliary function to reinterpret the maximal coding rate reduction method as a special yet local optimal case of IB theory. Through this auxiliary function, we clarify the paradox of decreasing mutual information during the application of ReLU activation in deep learning (DL) networks. Secondly, we challenge the doubts about IB theory's applicability by demonstrating its capacity to explain the absence of a compression phase with linear activation functions in hidden layers, when viewed through the lens of the auxiliary function. Lastly, by taking a novel theoretical stance, we provide a new way to interpret the inner organizations of DL networks by using IB theory, aligning them with recent experimental evidence. Thus, this paper serves as an act of justice for IB theory, potentially reinvigorating its standing and application in DL and other fields such as communications and biomedical research. | ['Zhijing Yang', 'Adil Mehmood Khan', 'Yongqiang Cheng', 'Faxian Cao'] | 2023-05-19 | null | null | null | null | ['misconceptions'] | ['miscellaneous'] | [ 4.78188545e-01 7.76965678e-01 -4.16582346e-01 -1.73161477e-01
1.55993477e-01 -2.48399258e-01 6.20541513e-01 3.11561137e-01
-7.63105690e-01 7.16392577e-01 5.25533617e-01 -9.61665630e-01
-6.53883755e-01 -4.88809049e-01 -7.27090240e-01 -8.09830666e-01
-8.04984346e-02 -2.72978749e-02 -1.63956285e-02 -3.20699275e-01
4.64557439e-01 6.51966929e-01 -1.30619717e+00 2.79510707e-01
4.71448600e-01 7.68071413e-01 2.79340565e-01 3.85787964e-01
-7.65938535e-02 1.13503075e+00 -3.98456782e-01 -5.45167923e-01
1.05784126e-01 -6.81983709e-01 -1.16827679e+00 -2.16848597e-01
-6.75871409e-03 -8.22712407e-02 -6.75336063e-01 7.66599178e-01
4.48282957e-01 -1.89731598e-01 4.13620293e-01 -1.00287569e+00
-5.36713719e-01 7.62983739e-01 -2.90600151e-01 7.89768696e-01
-2.13242173e-01 1.12350183e-02 1.22333348e+00 -4.98384416e-01
7.40622580e-01 1.24471557e+00 5.76603293e-01 6.94378197e-01
-1.17271876e+00 -3.91824841e-01 2.44861320e-01 3.44663024e-01
-1.07049835e+00 -7.04676449e-01 4.69357520e-01 -2.61718631e-01
9.24351335e-01 2.44542047e-01 8.81484926e-01 1.02701986e+00
3.23924482e-01 8.60178411e-01 1.02666056e+00 -7.00755298e-01
1.84314266e-01 -9.30953845e-02 1.26071557e-01 5.61903596e-01
6.44086361e-01 1.63126290e-01 -7.34831214e-01 1.90598309e-01
8.01953733e-01 -2.13777035e-01 -4.28626180e-01 3.53909098e-02
-1.07353592e+00 9.47648287e-01 4.16902155e-01 6.47212207e-01
-8.93470198e-02 1.96333617e-01 5.94885647e-01 4.32717055e-01
4.67835069e-01 4.75253940e-01 -4.01703238e-01 -1.47043630e-01
-7.34536290e-01 -1.23489618e-01 6.62013113e-01 4.93181020e-01
4.53845948e-01 -8.64067078e-02 2.68503606e-01 5.49355865e-01
4.19349790e-01 8.69225040e-02 4.81955916e-01 -1.09844100e+00
1.86885148e-01 4.65695441e-01 -5.18096268e-01 -1.01800394e+00
-6.65722966e-01 -6.64023817e-01 -1.10182238e+00 -5.71779199e-02
4.78313237e-01 1.10473819e-01 -2.18412906e-01 1.87363970e+00
-2.10769802e-01 -1.38330743e-01 -1.30384667e-02 8.35951805e-01
4.99325156e-01 4.32038397e-01 -1.40356675e-01 -4.87080067e-01
1.06979537e+00 -5.32946885e-01 -7.12776542e-01 -1.19471386e-01
9.95814800e-01 -3.76045406e-01 7.85528481e-01 4.51217830e-01
-1.26071680e+00 -2.19789550e-01 -1.17903578e+00 -1.75051004e-01
-1.94885939e-01 -4.32667106e-01 1.02827871e+00 6.60587430e-01
-1.38147032e+00 6.79453373e-01 -8.65749776e-01 -5.94174862e-01
6.31298840e-01 3.73692304e-01 -3.40057433e-01 -3.90169933e-03
-1.15876210e+00 9.74497378e-01 3.82952034e-01 3.76305968e-01
-4.67228502e-01 -5.79714477e-01 -5.78004003e-01 1.05304986e-01
3.00802380e-01 -7.83754706e-01 8.06306422e-01 -9.72501814e-01
-1.22154713e+00 1.05988467e+00 1.61904339e-02 -7.28679955e-01
3.35315138e-01 6.86626360e-02 -1.12589590e-01 3.07910562e-01
-4.21064198e-01 5.81521749e-01 3.91704500e-01 -1.09035635e+00
-2.03688204e-01 -6.15924239e-01 8.22974145e-02 9.49464887e-02
-3.86893630e-01 -2.12106377e-01 5.84048126e-03 -4.03238595e-01
3.32497418e-01 -6.75571918e-01 -1.69465989e-01 -9.18678939e-03
-2.15275198e-01 -3.77004631e-02 3.55050862e-01 -1.38083071e-01
1.39093661e+00 -2.11715245e+00 2.98998415e-01 1.20642945e-01
9.03216541e-01 2.75911510e-01 -8.54190886e-02 7.02886820e-01
-3.27241987e-01 4.80097979e-01 -1.36124343e-01 -3.12121622e-02
-1.33634642e-01 4.91927981e-01 -1.89205244e-01 8.02116752e-01
1.28249571e-01 1.01651454e+00 -7.78896630e-01 -3.04794163e-01
3.82880643e-02 6.21085763e-01 -6.46027565e-01 -3.53205591e-01
3.18391740e-01 3.14080358e-01 -1.97260499e-01 2.49604031e-01
4.12075728e-01 -4.85106021e-01 6.44222736e-01 -1.75397336e-01
-4.78613675e-01 4.85416502e-01 -6.05800569e-01 1.37827897e+00
-5.12395166e-02 1.26973879e+00 2.56955057e-01 -1.63154840e+00
7.29289353e-01 7.90094286e-02 4.73929584e-01 -9.10983980e-01
4.86382216e-01 1.39388770e-01 5.09706438e-01 -4.25507516e-01
1.42383248e-01 -2.97013313e-01 4.36509222e-01 6.50622964e-01
2.77699351e-01 3.55553746e-01 -2.32781265e-02 3.42238039e-01
1.20734191e+00 -2.84792572e-01 2.24362582e-01 -7.32082665e-01
3.20396692e-01 -5.04181504e-01 3.34516689e-02 9.50860262e-01
-2.74509102e-01 1.86301172e-01 9.86389756e-01 -6.91633999e-01
-1.36414719e+00 -8.33239734e-01 -5.13369381e-01 8.72511029e-01
-1.84637502e-01 -5.30387104e-01 -7.30983734e-01 -3.51808339e-01
-1.12519413e-01 3.24545294e-01 -9.06235993e-01 -3.89444411e-01
-3.87014121e-01 -1.17153704e+00 8.47486734e-01 9.53876376e-02
5.17115295e-01 -8.34388196e-01 -1.02841854e+00 -9.90249515e-02
-2.20745027e-01 -9.49126363e-01 3.33927602e-01 6.46808445e-01
-1.12420321e+00 -1.07497668e+00 -5.71856022e-01 -5.23767889e-01
6.57790899e-01 5.22189915e-01 8.88529360e-01 7.04803169e-01
-2.93279231e-01 3.22749674e-01 -2.67989218e-01 -2.77210653e-01
-5.92889965e-01 4.05904949e-01 1.18573289e-02 -5.90879202e-01
4.62214589e-01 -6.67084157e-01 -7.40427077e-01 -5.63717075e-02
-1.36043882e+00 1.99308142e-01 7.93191493e-01 7.94886649e-01
-1.77094057e-01 -1.44786298e-01 7.15108395e-01 -7.85582781e-01
8.33455324e-01 -6.16696417e-01 -2.53349662e-01 -4.05971706e-02
-1.17584217e+00 3.19226116e-01 3.05204272e-01 7.28087649e-02
-9.47687149e-01 -7.50265300e-01 -2.59668797e-01 2.68714130e-01
1.40242934e-01 5.77196598e-01 1.44870669e-01 -6.69815615e-02
7.23026156e-01 1.63837656e-01 5.24106920e-01 -2.96680123e-01
2.45332986e-01 3.96509856e-01 2.32400790e-01 -3.35152507e-01
2.78082967e-01 7.30540454e-01 2.87307858e-01 -9.98152077e-01
-8.55935633e-01 -1.85775936e-01 -8.71263266e-01 -3.51329446e-01
6.66850150e-01 -5.42631388e-01 -1.16281557e+00 1.69152781e-01
-1.13262677e+00 -3.28042448e-01 -3.03436249e-01 5.28445899e-01
-7.58298635e-01 4.57682967e-01 -7.64613807e-01 -7.67003834e-01
6.19375929e-02 -9.24179733e-01 5.02200723e-01 -7.30384365e-02
3.52390446e-02 -1.34330618e+00 -1.70698926e-01 3.13047111e-01
4.10807520e-01 -1.26086190e-01 1.23783469e+00 -6.28327310e-01
-4.02739197e-01 1.44180268e-01 -3.71050447e-01 2.58839101e-01
-1.15179308e-01 -1.46567032e-01 -1.08379209e+00 -1.75530791e-01
2.01419219e-01 -2.07168207e-01 1.17135704e+00 3.51279169e-01
9.46775854e-01 -5.03305137e-01 -1.19601592e-01 6.42555296e-01
1.55163324e+00 7.06807226e-02 6.54287159e-01 4.90105361e-01
1.39545575e-01 9.49758351e-01 -1.79272860e-01 5.85315943e-01
1.87913015e-01 3.64356965e-01 5.43803573e-01 -2.92518944e-01
-1.61154926e-01 -2.07734272e-01 9.05271992e-02 1.06201792e+00
-1.86393514e-01 -2.74172008e-01 -8.93184066e-01 2.19291270e-01
-1.75684822e+00 -1.01855588e+00 -2.49276534e-01 2.23455215e+00
7.73616493e-01 4.14561182e-01 1.09656662e-01 2.75054634e-01
4.78968292e-01 2.52982497e-01 -3.64085019e-01 -5.09655178e-01
-3.36821526e-01 -5.40307350e-02 6.05205059e-01 5.08278191e-01
-3.87908876e-01 7.70743847e-01 7.78491783e+00 5.63138902e-01
-1.03390121e+00 1.16973229e-01 8.82078886e-01 8.35741833e-02
-4.56465334e-01 -6.18125610e-02 -4.85332161e-01 2.14143246e-01
1.25703418e+00 -8.53046253e-02 5.99808395e-01 1.80440813e-01
4.49039102e-01 -2.79250920e-01 -1.03786564e+00 7.01644242e-01
-1.12971319e-02 -1.61241543e+00 -1.24411598e-01 4.72013861e-01
2.59978384e-01 2.23677918e-01 1.49627343e-01 -8.13850611e-02
1.70837224e-01 -1.26541841e+00 5.39816976e-01 3.92956376e-01
5.04232824e-01 -5.56209564e-01 7.86205053e-01 4.01395649e-01
-4.87219781e-01 -1.75361305e-01 -6.26588225e-01 -5.78149438e-01
-2.36065183e-02 5.65621495e-01 -7.88259804e-01 2.53320664e-01
4.10979152e-01 4.38049734e-01 -4.76923853e-01 7.66266763e-01
-4.55362163e-02 6.43224239e-01 -8.66811275e-02 1.29037440e-01
2.71498114e-01 8.45831782e-02 4.25469816e-01 1.22969413e+00
-8.41831714e-02 -1.09372232e-02 -7.81815708e-01 8.27625930e-01
5.06832227e-02 2.91283000e-02 -9.36461508e-01 -3.08753759e-01
2.80384153e-01 8.47344875e-01 -1.08437002e+00 -2.70615127e-02
-4.60337400e-01 6.69107914e-01 4.31476802e-01 3.18599880e-01
-5.16424596e-01 3.31950113e-02 2.67649770e-01 1.51842028e-01
-5.97050600e-02 -3.88806254e-01 -4.92192924e-01 -1.02701914e+00
-2.37185791e-01 -5.38499117e-01 3.44716050e-02 -5.13651133e-01
-7.96723008e-01 3.13992351e-01 1.98500440e-01 -6.68848217e-01
8.42764676e-02 -7.19200313e-01 2.38545286e-03 5.82175136e-01
-1.57671630e+00 -5.85078359e-01 8.57348219e-02 2.13542774e-01
2.14129597e-01 4.95690182e-02 6.51690483e-01 3.42875302e-01
-5.96842945e-01 4.76601541e-01 3.48451406e-01 -1.24052517e-01
2.42785141e-01 -8.35620701e-01 2.78150916e-01 5.39177477e-01
6.00178838e-02 9.41727698e-01 8.47507417e-01 -1.12624511e-01
-1.32604098e+00 -4.51438695e-01 1.05828214e+00 -2.72537738e-01
7.45322764e-01 -4.32542890e-01 -7.77086496e-01 6.59865975e-01
2.56549418e-01 -2.29614511e-01 8.80329907e-01 2.54917234e-01
-2.79870808e-01 -1.92386983e-03 -9.10405874e-01 5.75490952e-01
1.28708065e+00 -3.03852141e-01 -5.63010991e-01 3.37558240e-01
6.71353698e-01 1.49557158e-01 -7.98750460e-01 2.83552080e-01
7.04655886e-01 -1.30169439e+00 8.32795382e-01 -7.45069027e-01
5.76934636e-01 3.91264290e-01 -1.41093493e-01 -9.33216989e-01
-3.74877125e-01 -6.55556619e-01 1.50142103e-01 7.32895136e-01
3.12210649e-01 -8.73162508e-01 5.89320779e-01 3.09464306e-01
-2.69933909e-01 -1.02357864e+00 -1.20713472e+00 -6.42774224e-01
3.32797825e-01 -5.61963022e-01 5.03741612e-04 1.03502560e+00
6.17558062e-01 3.63230973e-01 -8.25503320e-02 -5.65660357e-01
3.31966072e-01 -7.52368510e-01 3.81659895e-01 -1.39030325e+00
-1.07659943e-01 -7.58009732e-01 -4.68632489e-01 -1.23268330e+00
1.03299201e-01 -1.16428626e+00 -2.94609189e-01 -1.26395786e+00
3.36348861e-01 -3.39250267e-01 -3.69154334e-01 2.91762352e-01
4.72267121e-01 3.06055695e-01 4.22265649e-01 5.32303333e-01
-5.50191462e-01 2.85768509e-01 1.24644005e+00 2.38216341e-01
9.23267677e-02 -2.34007627e-01 -1.35922468e+00 6.60573781e-01
8.33358049e-01 -3.97540987e-01 -5.29687107e-01 -3.99298936e-01
1.07614863e+00 -1.15557432e-01 4.58157748e-01 -6.98991418e-01
2.62679607e-01 -4.52926233e-02 2.28433475e-01 2.01056916e-02
3.47288474e-02 -7.62989581e-01 -3.34090680e-01 7.54924953e-01
-8.26397657e-01 1.84372619e-01 3.03151369e-01 4.60554332e-01
1.88436314e-01 -3.69199514e-01 8.02177846e-01 -3.75469118e-01
-4.41627026e-01 -4.22703708e-03 -6.55914307e-01 1.29934028e-01
7.74006248e-01 -3.28316510e-01 -7.27884293e-01 -3.87754649e-01
-7.40223467e-01 -3.83013934e-02 3.06060344e-01 1.49192587e-01
6.57075346e-01 -9.24462438e-01 -4.67137814e-01 3.81655186e-01
-1.70878977e-01 -2.92524517e-01 2.12734833e-01 1.37998748e+00
-5.83666325e-01 1.07227492e+00 -2.54901677e-01 -4.17968959e-01
-8.33887815e-01 6.24406159e-01 3.55398804e-01 -3.43003310e-02
-8.08371603e-01 8.29498351e-01 2.25388646e-01 9.86949131e-02
2.57614017e-01 -1.46567658e-01 -8.17690119e-02 1.40137479e-01
4.70966876e-01 4.38874334e-01 1.16832681e-01 -3.87296140e-01
-3.30344170e-01 -2.90898932e-03 -9.82335061e-02 -2.32098728e-01
1.36750901e+00 -5.27644873e-01 -3.35566610e-01 6.70842469e-01
1.17670095e+00 -1.94844633e-01 -1.09380841e+00 -7.92849064e-02
1.17781244e-01 -1.66905627e-01 2.05861315e-01 -5.97095490e-01
-8.58823299e-01 1.11460519e+00 2.70391464e-01 6.41458690e-01
1.08239353e+00 2.69358546e-01 2.03893244e-01 4.06458318e-01
6.44158795e-02 -9.89391506e-01 1.33115709e-01 4.34098721e-01
7.15170085e-01 -8.36470187e-01 1.64101534e-02 -1.27221271e-02
-2.99051046e-01 1.34752679e+00 1.68265894e-01 1.68829724e-01
5.99606812e-01 3.00835311e-01 -2.21241862e-01 -3.48532379e-01
-1.28334761e+00 -5.91296963e-02 -3.02991092e-01 5.79851687e-01
9.39916074e-01 -2.51516074e-01 -9.72749472e-01 2.08743289e-01
-3.20587963e-01 8.52221996e-03 6.80551887e-01 7.28451788e-01
-7.59831011e-01 -1.12026238e+00 -8.37757811e-02 3.40369403e-01
-7.00061321e-01 -4.15948391e-01 -3.20420653e-01 1.14889288e+00
-5.05720228e-02 8.53624642e-01 1.69967949e-01 -3.06275368e-01
-3.10322851e-01 7.65933096e-02 6.16141140e-01 -2.63409257e-01
-4.08999175e-01 -2.80582085e-02 -2.09712341e-01 -3.78625095e-01
-5.70863307e-01 -4.90819991e-01 -9.25625145e-01 -7.46110797e-01
-2.27484226e-01 1.27899349e-01 7.28285074e-01 1.18892384e+00
3.30130398e-01 6.80891454e-01 3.15848976e-01 -4.09917116e-01
-4.21016634e-01 -7.99218714e-01 -4.69594479e-01 3.36390547e-02
6.35359168e-01 -3.27859342e-01 -6.07156932e-01 -2.33055592e-01] | [8.04636001586914, 3.531985282897949] |
7bdb9255-76a5-4556-9329-8045acb2bb35 | semantic-aware-transmission-for-robust-point | 2306.13296 | null | https://arxiv.org/abs/2306.13296v1 | https://arxiv.org/pdf/2306.13296v1.pdf | Semantic-aware Transmission for Robust Point Cloud Classification | As three-dimensional (3D) data acquisition devices become increasingly prevalent, the demand for 3D point cloud transmission is growing. In this study, we introduce a semantic-aware communication system for robust point cloud classification that capitalizes on the advantages of pre-trained Point-BERT models. Our proposed method comprises four main components: the semantic encoder, channel encoder, channel decoder, and semantic decoder. By employing a two-stage training strategy, our system facilitates efficient and adaptable learning tailored to the specific classification tasks. The results show that the proposed system achieves classification accuracy of over 89\% when SNR is higher than 10 dB and still maintains accuracy above 66.6\% even at SNR of 4 dB. Compared to the existing method, our approach performs at 0.8\% to 48\% better across different SNR values, demonstrating robustness to channel noise. Our system also achieves a balance between accuracy and speed, being computationally efficient while maintaining high classification performance under noisy channel conditions. This adaptable and resilient approach holds considerable promise for a wide array of 3D scene understanding applications, effectively addressing the challenges posed by channel noise. | ['Zhiguo Shi', 'Qianqian Yang', 'Kaiyi Chi', 'Tianxiao Han'] | 2023-06-23 | null | null | null | null | ['point-cloud-classification', 'scene-understanding', 'classification-1'] | ['computer-vision', 'computer-vision', 'methodology'] | [ 3.00773978e-01 -1.46044672e-01 -1.03210472e-01 -9.85875055e-02
-1.19138432e+00 -2.92356461e-01 2.69969910e-01 2.23080844e-01
-1.08101323e-01 2.17207164e-01 -1.29009262e-01 -3.46646309e-01
-9.22623649e-02 -7.39078164e-01 -6.42444313e-01 -5.99619210e-01
-3.93380553e-01 1.02107473e-01 3.18889856e-01 7.64763951e-02
3.20800602e-01 8.83566201e-01 -1.54091799e+00 -1.31869450e-01
7.15956688e-01 1.65918684e+00 4.01796162e-01 8.87217522e-01
4.24664281e-02 3.84339631e-01 -8.23469639e-01 -6.88197166e-02
2.91627824e-01 3.72869335e-02 -2.82894045e-01 1.84292704e-01
-1.75261684e-02 -3.03899914e-01 -5.27351618e-01 8.44659865e-01
8.71441662e-01 -4.04892683e-01 4.10778284e-01 -1.06558359e+00
-3.95943552e-01 -1.67193655e-02 -4.82957989e-01 1.18083144e-02
4.71715480e-01 6.60591340e-03 7.44020164e-01 -7.57210255e-01
1.17917955e-01 9.11924303e-01 7.90784955e-01 3.84641677e-01
-8.79318655e-01 -8.71128619e-01 -1.71927691e-01 2.22090214e-01
-1.66548395e+00 -4.87693518e-01 8.44554484e-01 -3.61765087e-01
9.45610464e-01 5.39302155e-02 5.96940815e-01 8.14158976e-01
1.43437788e-01 6.21355534e-01 9.21488881e-01 -2.81026155e-01
4.45179462e-01 7.65288919e-02 -3.14319521e-01 4.07250047e-01
7.61647075e-02 1.34476528e-01 -7.54763246e-01 -1.71553150e-01
6.07839286e-01 -3.11747938e-01 -2.95977712e-01 -3.51545900e-01
-1.02173996e+00 4.19488847e-01 5.90025663e-01 6.49068132e-02
-3.11557829e-01 3.98331732e-01 2.14949891e-01 3.29134613e-01
3.70117098e-01 6.68885335e-02 -2.32527494e-01 -3.22460145e-01
-7.19942451e-01 -4.84746359e-02 6.03017390e-01 1.57658529e+00
2.75304973e-01 1.59431309e-01 8.98552686e-02 8.72533739e-01
6.45119727e-01 1.00998020e+00 -1.72758341e-01 -9.09607112e-01
6.40192986e-01 2.61934906e-01 -8.07408616e-02 -1.05363393e+00
-4.56652403e-01 -9.44656610e-01 -9.65045691e-01 1.22380830e-01
-1.68176591e-01 1.83846742e-01 -7.80307353e-01 1.53510630e+00
1.69083506e-01 2.99916029e-01 2.50228345e-01 8.58243465e-01
6.87619090e-01 6.54877901e-01 -7.46763572e-02 -1.38684079e-01
1.05061305e+00 -1.46315411e-01 -4.89949673e-01 -6.92810165e-03
3.13410729e-01 -8.11467052e-01 7.82832921e-01 4.11175340e-01
-9.16643441e-01 -4.88785654e-01 -1.48234415e+00 2.62322277e-01
1.88551560e-01 -1.41622052e-01 3.45612645e-01 9.59718525e-01
-1.04581118e+00 6.62249327e-02 -7.19114959e-01 -3.39465469e-01
7.70122647e-01 4.42755669e-01 -1.45298569e-02 -1.99965984e-01
-8.52073789e-01 4.76802468e-01 -9.49831456e-02 -1.22313865e-01
-7.51138151e-01 -6.35311604e-01 -5.75311661e-01 -1.02170696e-02
8.80584866e-02 -6.98166847e-01 1.22430313e+00 -2.77750075e-01
-1.82061923e+00 4.74786580e-01 -1.98924407e-01 -4.92796123e-01
2.73016900e-01 -2.38709047e-01 -5.87340951e-01 5.58508456e-01
7.04299808e-02 2.69810617e-01 9.00197983e-01 -1.57013142e+00
-6.70200586e-01 -5.49284220e-01 -1.95462167e-01 3.54824454e-01
-9.34828222e-02 -8.05180147e-02 -8.87510061e-01 -4.94824499e-01
7.15861559e-01 -9.30708408e-01 -4.47626971e-02 4.95054305e-01
-2.69716173e-01 2.89007813e-01 1.01794064e+00 -1.47288039e-01
7.87212849e-01 -2.32789421e+00 -3.32927346e-01 5.82084000e-01
2.28865683e-01 1.84206724e-01 1.21266142e-01 4.16356176e-01
4.51097220e-01 5.06773032e-03 -2.92115927e-01 -2.98267126e-01
-1.24049336e-01 9.82864499e-02 -1.35982394e-01 6.20468557e-01
3.61277051e-02 5.21635532e-01 -6.67804301e-01 -1.15111724e-01
5.28613269e-01 6.84589922e-01 -6.09107018e-01 8.75602588e-02
1.99268401e-01 6.25485122e-01 -6.59372866e-01 9.49333787e-01
8.15841258e-01 -4.04016703e-01 -1.14583209e-01 5.15739769e-02
2.26873592e-01 2.84649730e-01 -1.27538371e+00 1.79235160e+00
-7.03877151e-01 7.67365813e-01 3.55479389e-01 -9.73928452e-01
1.18251932e+00 2.79286802e-01 6.44203365e-01 -1.01963544e+00
1.09088413e-01 5.13226330e-01 -2.49904811e-01 -3.42313021e-01
2.84203142e-01 -1.57798350e-01 -2.42530271e-01 -4.64001521e-02
-1.93198144e-01 -6.07316911e-01 -7.80472040e-01 1.71640247e-01
1.36281931e+00 -3.66276652e-01 8.19544122e-02 -6.40792474e-02
5.59089720e-01 -2.84530312e-01 2.98367500e-01 8.31638634e-01
-1.99899986e-01 4.62937951e-01 -9.85743478e-03 1.36479348e-01
-9.45815265e-01 -1.21309066e+00 -4.19232637e-01 1.83684841e-01
9.02698278e-01 -3.56583893e-01 -3.04315150e-01 -1.17377788e-01
2.49917075e-01 5.17511308e-01 1.69918612e-01 -3.34532291e-01
-1.85923174e-01 -5.76019228e-01 7.47253537e-01 4.55559492e-01
9.58048165e-01 -6.21564798e-02 -4.42298949e-01 1.92969009e-01
-1.44879576e-02 -1.65697098e+00 2.23409757e-01 6.72511905e-02
-8.28496516e-01 -8.00680697e-01 -3.64986658e-01 -5.14647782e-01
3.75408381e-01 6.93266511e-01 7.89555430e-01 -3.41999605e-02
-2.47748420e-02 6.62234485e-01 -5.86148620e-01 -5.75120866e-01
-3.08540463e-01 -1.42537981e-01 1.20986059e-01 -9.79813635e-02
1.47497028e-01 -7.43352532e-01 -6.52962446e-01 4.43231642e-01
-4.73559231e-01 -2.40144387e-01 6.19020820e-01 5.12697995e-01
7.79768586e-01 3.21178734e-01 6.27299309e-01 -3.12723070e-01
2.56601244e-01 -5.43771505e-01 -6.63247168e-01 -3.87922615e-01
-6.56235278e-01 -2.13161826e-01 5.59944510e-01 1.70166045e-01
-7.03031480e-01 9.68416333e-02 -5.50375342e-01 -3.42764825e-01
-2.25276366e-01 -1.06012717e-01 -5.23736417e-01 -6.98434591e-01
5.52164435e-01 2.65933126e-01 -4.65204418e-02 -3.83342743e-01
2.67322421e-01 1.26564682e+00 5.10281324e-01 -2.91772276e-01
1.22476971e+00 7.91894138e-01 2.88796753e-01 -1.30783176e+00
-3.81360203e-01 -5.63859701e-01 -3.93837333e-01 -4.78087246e-01
5.61559081e-01 -1.33097720e+00 -9.23327386e-01 5.22006631e-01
-9.18336272e-01 -2.60748081e-02 -1.76146328e-01 6.90477610e-01
-6.39368117e-01 4.96302128e-01 -3.07249904e-01 -8.16821754e-01
-3.37234169e-01 -1.32147849e+00 1.48237002e+00 -2.26355642e-01
1.00981489e-01 -5.64332783e-01 -6.09445751e-01 4.39871758e-01
6.16112590e-01 2.27211326e-01 9.07846868e-01 -2.00618908e-01
-1.13932443e+00 -4.82601196e-01 -4.07669127e-01 2.86865801e-01
-1.41490875e-02 -5.61607838e-01 -1.17917156e+00 -3.07681173e-01
8.96990076e-02 -5.22574969e-02 4.39455599e-01 1.98155180e-01
1.44863355e+00 -1.04058661e-01 -3.78116637e-01 9.59538877e-01
1.50150824e+00 3.03535402e-01 5.07501364e-01 1.99619442e-01
5.25656462e-01 -5.00493050e-02 4.58966523e-01 5.79861581e-01
4.50048894e-01 8.48769963e-01 8.39518547e-01 1.06393258e-04
-2.66827852e-01 -2.15691671e-01 -2.64290161e-03 7.86899745e-01
3.14533919e-01 -5.79830468e-01 -9.90241587e-01 3.18709195e-01
-1.36357868e+00 -6.86484456e-01 -9.47968513e-02 2.25025296e+00
2.87314117e-01 4.87144738e-01 -2.88503021e-01 6.22651815e-01
5.09460032e-01 1.27393454e-01 -6.15134835e-01 -1.42336905e-01
-1.06843598e-01 2.07554623e-01 9.18042302e-01 3.22040677e-01
-9.22624588e-01 5.03651261e-01 6.31150627e+00 8.08367789e-01
-1.24470377e+00 -5.74634299e-02 1.55109972e-01 -5.21178283e-02
-1.61111161e-01 -2.87358493e-01 -5.11201799e-01 5.51130950e-01
8.94093812e-01 3.02772280e-02 7.22360387e-02 8.12028587e-01
2.93245763e-01 -1.68249547e-01 -7.97048688e-01 1.44308341e+00
1.52159348e-01 -1.28237820e+00 -1.23486081e-02 2.59538531e-01
2.86044955e-01 3.27460647e-01 1.52991280e-01 -2.68271506e-01
-8.31856504e-02 -7.65434742e-01 9.06985402e-01 8.82756338e-02
1.03887665e+00 -8.74435842e-01 5.72980344e-01 6.01606011e-01
-1.20475936e+00 -2.24275798e-01 -1.10649914e-01 8.76826718e-02
2.64424384e-01 8.63684356e-01 -9.13663805e-01 7.45852947e-01
9.00846064e-01 7.36503959e-01 -1.04124948e-01 1.28893328e+00
-1.31682202e-01 5.54521263e-01 -7.33925343e-01 -6.56971196e-03
3.18090394e-02 3.69635999e-01 1.00433803e+00 1.06976306e+00
6.32802010e-01 3.45716715e-01 9.33943540e-02 2.25011840e-01
-1.27304047e-02 -9.30881798e-02 -7.98894882e-01 5.47367394e-01
9.28282261e-01 5.68329513e-01 -4.04016465e-01 3.38626988e-02
-5.40342391e-01 9.65113878e-01 -1.45323440e-01 2.16790393e-01
-7.11756110e-01 -4.71398383e-01 8.82668495e-01 5.33605143e-02
4.43971306e-01 -7.72111058e-01 -6.87159240e-01 -6.75429881e-01
2.79765129e-01 -4.87564713e-01 -8.83078426e-02 -6.47232533e-01
-9.34488356e-01 5.11186659e-01 -3.22607696e-01 -1.77909386e+00
2.00343654e-01 -5.00839770e-01 -6.26787022e-02 5.50383985e-01
-1.86597919e+00 -9.19557989e-01 -5.49278200e-01 5.76184750e-01
3.98666590e-01 -1.06932998e-01 6.95262074e-01 6.76471829e-01
-2.20239818e-01 6.12030149e-01 3.27601701e-01 -1.25535160e-01
1.72023863e-01 -8.25965703e-01 4.08358932e-01 8.01551163e-01
-5.93562871e-02 6.55476078e-02 5.12587965e-01 -3.80879223e-01
-1.87217104e+00 -1.21038473e+00 5.54686964e-01 -1.56450957e-01
2.09087551e-01 -6.87805593e-01 -6.46952927e-01 -2.34367065e-02
-2.68860728e-01 -3.58693786e-02 7.89801419e-01 -2.65184313e-01
-4.04337287e-01 -3.09343159e-01 -1.28532898e+00 3.43842745e-01
1.40076721e+00 -7.21842766e-01 -5.70354462e-02 3.60226452e-01
6.84183717e-01 -7.92231679e-01 -9.66829181e-01 7.37294853e-01
2.81780243e-01 -8.39561820e-01 1.10073698e+00 3.40940714e-01
1.58394985e-02 -4.80217636e-01 -7.37691343e-01 -1.00321877e+00
-1.69736877e-01 -7.67082930e-01 -2.37510219e-01 7.76006520e-01
4.35447425e-01 -5.61138868e-01 8.80838573e-01 1.61342084e-01
-3.57901275e-01 -5.56635201e-01 -1.34707928e+00 -9.65926588e-01
-4.47403014e-01 -1.30802715e+00 6.70174956e-01 5.23862243e-01
-1.26038566e-01 2.14903161e-01 -2.15175763e-01 9.49476421e-01
1.05906105e+00 -8.95055085e-02 9.76099968e-01 -1.06231117e+00
-2.12424144e-01 -2.12468103e-01 -1.10790181e+00 -1.71173787e+00
-3.14823836e-02 -9.68390405e-01 -4.89926450e-02 -1.43851268e+00
-2.30716392e-01 -1.03788376e+00 -4.99670813e-03 1.05224214e-01
2.35103846e-01 5.03312647e-01 1.47720665e-01 4.20589238e-01
-5.00901103e-01 7.48959899e-01 1.00280833e+00 -1.56657621e-01
-1.33675739e-01 3.68529737e-01 -7.46178031e-01 4.94272858e-01
8.38322818e-01 -2.01895043e-01 -4.28088158e-01 -7.68422186e-01
-1.28178060e-01 9.10749659e-02 5.90203643e-01 -1.55021417e+00
3.18466812e-01 2.70267099e-01 4.10802156e-01 -6.06042206e-01
8.08234513e-01 -1.33049369e+00 1.49319276e-01 4.18003947e-01
1.92565448e-03 -4.28356498e-01 4.39048111e-01 1.04760027e+00
-1.81165442e-01 5.38160861e-01 9.03249085e-01 2.61935920e-01
-8.61539900e-01 3.23875308e-01 -3.25903505e-01 -2.55991459e-01
1.19967496e+00 -6.26674473e-01 -1.36929929e-01 -7.00783253e-01
-4.42097664e-01 3.02574873e-01 3.76682371e-01 4.61899519e-01
9.51994658e-01 -1.23237753e+00 -4.86007750e-01 5.47384501e-01
4.56553251e-01 1.14652470e-01 1.76763713e-01 5.27362347e-01
-5.55754066e-01 4.18420821e-01 2.55233556e-01 -1.30465245e+00
-1.03219378e+00 4.73405793e-03 3.32105011e-01 5.62413335e-01
-9.72407460e-01 9.21006680e-01 -3.59981000e-01 -8.84819701e-02
5.12332976e-01 -1.79572716e-01 3.03162545e-01 -5.97412527e-01
2.13258684e-01 3.20656329e-01 3.96753162e-01 -7.97803640e-01
-4.67477679e-01 9.70061064e-01 3.33063960e-01 5.02917357e-03
1.21939540e+00 -4.38183069e-01 5.47267377e-01 1.31758317e-01
1.36446583e+00 5.65370992e-02 -1.19251978e+00 -4.71329391e-01
-1.48322091e-01 -8.38641167e-01 4.20172155e-01 -6.00441158e-01
-8.90869677e-01 8.08529735e-01 8.66765440e-01 2.37336025e-01
1.40132523e+00 3.64877194e-01 9.06781137e-01 3.21300328e-01
8.84029448e-01 -8.43388855e-01 -5.89970686e-02 3.73234212e-01
5.14262915e-01 -1.08664227e+00 1.54210534e-02 -8.25312316e-01
-2.72680849e-01 7.33013630e-01 7.84753039e-02 8.81802291e-03
8.61035287e-01 4.53167021e-01 1.72548637e-01 -3.55638832e-01
-5.34154534e-01 -2.49351636e-01 -3.98825854e-02 1.03095388e+00
-2.46762082e-01 -3.09939384e-02 3.11629921e-01 1.75747618e-01
-2.95481592e-01 -2.68798709e-01 2.07478538e-01 9.75889742e-01
-7.04687893e-01 -8.80523145e-01 -4.10357475e-01 3.64496857e-01
-3.40164989e-01 1.38692111e-01 -1.76702410e-01 7.71081269e-01
-1.24879703e-01 1.16310287e+00 6.69195428e-02 -6.88473523e-01
6.05183959e-01 -3.90903383e-01 3.08497429e-01 -4.38474923e-01
6.65739831e-03 8.79829079e-02 2.25016959e-02 -8.67014945e-01
-5.35122216e-01 -5.93819380e-01 -1.32305419e+00 -2.35274360e-01
-4.50291783e-01 -9.10152793e-02 1.15680993e+00 7.05162585e-01
7.95055687e-01 4.79287654e-01 1.25972843e+00 -6.65251911e-01
-3.88010830e-01 -2.20376283e-01 -5.76428235e-01 -5.94909899e-02
6.61264420e-01 -6.29620552e-01 -3.04725647e-01 -2.92380095e-01] | [7.834747791290283, -2.8643531799316406] |
915e4601-1734-45c2-85cb-6c0ac2327e6f | effectiveness-of-random-deep-feature | 1910.12392 | null | https://arxiv.org/abs/1910.12392v2 | https://arxiv.org/pdf/1910.12392v2.pdf | Effectiveness of random deep feature selection for securing image manipulation detectors against adversarial examples | We investigate if the random feature selection approach proposed in [1] to improve the robustness of forensic detectors to targeted attacks, can be extended to detectors based on deep learning features. In particular, we study the transferability of adversarial examples targeting an original CNN image manipulation detector to other detectors (a fully connected neural network and a linear SVM) that rely on a random subset of the features extracted from the flatten layer of the original network. The results we got by considering three image manipulation detection tasks (resizing, median filtering and adaptive histogram equalization), two original network architectures and three classes of attacks, show that feature randomization helps to hinder attack transferability, even if, in some cases, simply changing the architecture of the detector, or even retraining the detector is enough to prevent the transferability of the attacks. | ['Bo-Wen Zhang', 'Ehsan Nowroozi', 'Benedetta Tondi', 'Mauro Barni'] | 2019-10-25 | null | null | null | null | ['image-manipulation-detection'] | ['computer-vision'] | [ 4.50125128e-01 2.74823934e-01 3.03122938e-01 1.84499070e-01
-2.40562931e-02 -1.15498567e+00 7.52959728e-01 4.63571772e-02
-4.88254040e-01 2.94448972e-01 -1.20480113e-01 -4.03181881e-01
-1.41952232e-01 -8.83210778e-01 -9.04409945e-01 -7.05034673e-01
-2.90735006e-01 7.09352270e-03 5.88920891e-01 -2.71876991e-01
4.87997711e-01 1.16821373e+00 -1.39061964e+00 5.06032109e-01
8.17462876e-02 7.94080257e-01 -5.88622212e-01 8.54458630e-01
3.59762162e-01 7.83810437e-01 -8.93168807e-01 -5.81818759e-01
8.23291481e-01 -6.21091090e-02 -6.62436306e-01 -1.57670885e-01
7.23110616e-01 -5.42861879e-01 -8.23196590e-01 1.26295769e+00
3.76053542e-01 -2.13275030e-01 6.00213230e-01 -1.32573938e+00
-5.70053577e-01 5.16614020e-01 -1.55971646e-01 4.47992802e-01
3.76217991e-01 7.05387652e-01 3.13445240e-01 -2.82705754e-01
6.69609964e-01 1.48386586e+00 6.70817256e-01 6.46409392e-01
-1.12960088e+00 -7.96589494e-01 -2.00106844e-01 1.83246151e-01
-9.82637346e-01 -3.53815973e-01 6.64220512e-01 -4.02749568e-01
5.97307682e-01 2.40798220e-01 2.33360246e-01 1.49607706e+00
2.27471590e-01 2.67268479e-01 1.09275675e+00 -3.20727378e-01
1.94833115e-01 4.16265100e-01 4.24203575e-01 4.71679866e-01
6.57941580e-01 6.84796214e-01 -8.51081163e-02 -4.59657371e-01
5.22624314e-01 -1.14166468e-01 -4.30331528e-01 -3.52528542e-01
-8.42644572e-01 9.09978271e-01 6.23287797e-01 6.01113498e-01
-3.19938898e-01 1.27981827e-01 8.19357634e-01 7.74492383e-01
-7.67052546e-02 9.44193840e-01 -4.23171848e-01 3.16623628e-01
-6.29468560e-01 7.30859712e-02 8.62269580e-01 5.03654599e-01
7.30619609e-01 3.65910292e-01 -4.24362540e-01 -4.17425185e-01
-2.92049587e-01 4.55944926e-01 5.05999207e-01 -5.11431813e-01
3.62279207e-01 6.50222659e-01 -1.67254567e-01 -1.09638798e+00
-4.38862175e-01 -5.94588257e-02 -5.13443112e-01 5.65047503e-01
8.40594530e-01 -3.71907324e-01 -7.86625862e-01 1.41901624e+00
7.95816481e-02 1.30248144e-01 -1.07960537e-01 5.65020561e-01
3.36447716e-01 1.48495600e-01 6.70503601e-02 2.99440861e-01
1.01393151e+00 -5.23071945e-01 -2.15524033e-01 7.90663436e-03
6.54378951e-01 -4.77149397e-01 8.72751832e-01 3.88022006e-01
-7.51673818e-01 -6.49613619e-01 -1.49279690e+00 3.35328519e-01
-9.72735286e-01 -2.50203729e-01 2.86672831e-01 1.40804672e+00
-8.24910462e-01 1.14758456e+00 -5.49286187e-01 -1.79079935e-01
4.52870339e-01 6.98551714e-01 -7.88841069e-01 3.43015492e-02
-1.42729032e+00 9.92464006e-01 5.50257802e-01 -1.98531598e-01
-1.05508113e+00 -3.98648351e-01 -5.95835507e-01 4.57819730e-01
1.57092109e-01 -2.25793496e-01 5.81979990e-01 -1.33745086e+00
-1.36361587e+00 8.44887137e-01 5.49027503e-01 -7.65407979e-01
8.07839096e-01 -6.56061620e-02 -2.28939980e-01 4.24726844e-01
-4.57893968e-01 2.73881257e-01 1.46741951e+00 -9.64918673e-01
-4.60594185e-02 -4.64800447e-01 4.31867123e-01 -6.82082593e-01
-8.01532030e-01 3.71905684e-01 4.20313627e-01 -4.84746903e-01
-7.31489956e-01 -1.02596104e+00 -1.17366306e-01 -3.01010441e-02
-7.07293391e-01 4.46626872e-01 1.28951859e+00 -7.48482704e-01
7.38462150e-01 -2.50411034e+00 -9.72071569e-03 5.47056615e-01
-6.34947196e-02 8.72498274e-01 -4.22693819e-01 2.58019745e-01
-7.17288554e-01 2.67869383e-01 -2.15677992e-01 1.46487564e-01
-1.69845559e-02 -2.44173691e-01 -6.94002748e-01 1.08623445e+00
3.71691406e-01 7.55649149e-01 -4.80699927e-01 1.25533089e-01
4.38015580e-01 2.98206896e-01 -3.37559909e-01 6.00921661e-02
1.01628661e-01 4.08188820e-01 -3.36704493e-01 2.58231580e-01
7.68868804e-01 3.87863725e-01 -2.30811894e-01 -7.52843618e-02
2.44286776e-01 -5.70278578e-02 -1.12547278e+00 8.70261252e-01
-1.37531504e-01 8.36017370e-01 -1.19126607e-02 -1.01171815e+00
7.53660679e-01 3.60347837e-01 5.58711663e-02 -6.61915541e-01
5.42577922e-01 -1.27070487e-01 4.20382857e-01 -4.43436861e-01
1.00103945e-01 3.19067746e-01 -1.48055300e-01 5.72462261e-01
1.39364332e-01 2.64396846e-01 3.87208611e-02 6.08119667e-02
1.56578672e+00 -3.23108763e-01 7.68199330e-03 -8.77276510e-02
7.10003376e-01 -9.40427408e-02 3.17273252e-02 1.14980268e+00
-1.29371017e-01 4.28683072e-01 8.29664409e-01 -5.39153755e-01
-1.29977393e+00 -9.33402658e-01 7.97700733e-02 8.28972578e-01
-2.10752100e-01 -1.11353807e-01 -1.14511585e+00 -1.19571602e+00
1.78014249e-01 6.87521219e-01 -1.02725410e+00 -8.75138164e-01
-9.82168198e-01 -6.47601902e-01 1.23193693e+00 3.77668351e-01
5.85588396e-01 -1.20145464e+00 -7.51370549e-01 -9.75618958e-02
6.53368652e-01 -1.27640080e+00 -4.94869351e-02 3.86866480e-01
-4.91597563e-01 -1.49779832e+00 -2.26266563e-01 -2.74428904e-01
5.92368603e-01 -7.69234300e-02 4.96185720e-01 5.34172356e-01
-4.59202409e-01 4.49929565e-01 -3.88771474e-01 -4.19409782e-01
-8.31338704e-01 1.50311485e-01 1.88309867e-02 6.67142719e-02
1.53973117e-01 -4.50250238e-01 -1.65427580e-01 9.03531238e-02
-1.35724378e+00 -8.39097977e-01 5.02668142e-01 4.98799056e-01
-5.19620240e-01 2.12725282e-01 2.54335284e-01 -1.02564144e+00
6.04757845e-01 -2.64104545e-01 -5.19147098e-01 2.16672391e-01
-2.30767906e-01 2.10970223e-01 9.96085107e-01 -1.03012383e+00
-4.35867399e-01 1.20753080e-01 -4.06633504e-02 -7.61503935e-01
-4.70368266e-01 -1.53558359e-01 -4.85636801e-01 -7.25496352e-01
1.17051399e+00 1.13878570e-01 -1.77097891e-03 -2.82694727e-01
4.91411865e-01 2.64197797e-01 3.74384671e-01 -3.00065637e-01
1.59854305e+00 5.57096183e-01 3.41215789e-01 -5.91725290e-01
-1.20257884e-01 7.67835677e-02 -7.97117293e-01 -7.60072246e-02
6.90905452e-01 -3.15294713e-01 -8.32904875e-01 7.88293302e-01
-1.26512170e+00 -2.85664976e-01 -4.34880167e-01 5.50526716e-02
-3.77863377e-01 5.14776349e-01 -7.19837904e-01 -3.92310739e-01
-2.77700990e-01 -1.19663656e+00 6.74600661e-01 2.08203346e-01
5.27464785e-02 -8.92045796e-01 -1.34187236e-01 -1.67848170e-01
6.91797435e-01 5.69049776e-01 1.17581546e+00 -1.43140590e+00
-3.36068064e-01 -7.82010078e-01 -7.47895017e-02 6.44132376e-01
-8.11389983e-02 1.14877239e-01 -1.29190969e+00 -5.37997663e-01
3.40367466e-01 -2.97295481e-01 9.50672448e-01 -1.11343063e-01
1.04602921e+00 -4.69288141e-01 -1.78412095e-01 4.31927592e-01
1.27225232e+00 -2.20218375e-02 1.02800751e+00 7.33518600e-01
6.05850875e-01 5.09528875e-01 -6.49926662e-02 1.55132294e-01
-6.49276435e-01 6.94110811e-01 7.78914273e-01 -6.29565790e-02
-7.63512552e-02 5.34569137e-02 7.44906187e-01 -3.72844309e-01
1.17400065e-01 -6.55830204e-02 -6.42296970e-01 1.26950249e-01
-1.30228519e+00 -1.05928195e+00 1.09725550e-01 2.32435560e+00
2.51923203e-01 5.32245219e-01 3.33942473e-01 6.05393112e-01
1.10140049e+00 5.89667521e-02 -4.29244816e-01 -8.09157431e-01
-1.87142357e-01 6.22230828e-01 1.03438711e+00 6.58628270e-02
-1.23198688e+00 8.29430640e-01 6.46958447e+00 8.46425593e-01
-1.35848546e+00 -2.61814799e-02 2.82447428e-01 2.33069688e-01
3.09140682e-01 5.13828918e-02 -4.60913241e-01 5.23565114e-01
1.28921199e+00 1.80185735e-01 4.45262969e-01 9.39724088e-01
-1.34683982e-01 2.75008380e-01 -1.22259820e+00 2.72193164e-01
2.05275133e-01 -1.25716352e+00 2.59486049e-01 3.38927239e-01
1.89580113e-01 -4.76015478e-01 3.72068137e-01 3.97729486e-01
3.20067346e-01 -1.01854694e+00 6.02861822e-01 3.17343175e-01
3.00909579e-01 -8.66616607e-01 7.84168363e-01 3.54082912e-01
-6.22222364e-01 -4.97212142e-01 -2.48226762e-01 -3.10704336e-02
-3.30439270e-01 1.71767503e-01 -7.24431157e-01 3.03613812e-01
4.72164303e-01 2.94434968e-02 -1.06128526e+00 9.79643285e-01
-4.88179505e-01 5.44139206e-01 -7.80418664e-02 2.19216108e-01
2.73777872e-01 4.07669246e-01 7.23707378e-01 1.24166059e+00
-1.94846079e-01 -3.38240892e-01 -3.29555310e-02 5.63576221e-01
-8.25819597e-02 -1.94884703e-01 -8.94865274e-01 -3.33070233e-02
3.28918397e-01 1.23542881e+00 -8.18521619e-01 -2.55749132e-02
-1.83114335e-01 9.50536907e-01 2.86113292e-01 4.05196659e-02
-9.70794499e-01 -5.23927331e-01 3.92211258e-01 3.08429778e-01
5.80834627e-01 -7.83499796e-03 -7.55864829e-02 -9.87312794e-01
-8.29492062e-02 -1.04057240e+00 2.48444915e-01 -4.48199302e-01
-1.13591576e+00 4.64860678e-01 -2.92957313e-02 -1.01553893e+00
-1.90262750e-01 -1.03812563e+00 -1.08127713e+00 5.07465124e-01
-1.12895572e+00 -1.17091811e+00 8.69374871e-02 8.89926016e-01
-4.14207391e-02 -4.41413373e-01 7.10703075e-01 2.43742660e-01
-5.57460308e-01 8.84832203e-01 -1.58626363e-01 7.71653950e-01
5.66359997e-01 -9.51344669e-01 6.27024472e-01 1.05095470e+00
7.56145343e-02 6.67692900e-01 6.87740028e-01 -4.77988005e-01
-1.33072615e+00 -1.10458899e+00 2.80095309e-01 -3.70525241e-01
8.57065976e-01 -5.47557652e-01 -1.02194321e+00 7.81644166e-01
2.42660612e-01 8.92441869e-02 2.63057172e-01 -3.56306106e-01
-7.95092463e-01 4.06902552e-01 -1.70039833e+00 4.99013841e-01
6.21851742e-01 -5.77004969e-01 -5.46612620e-01 2.90146321e-01
5.17425418e-01 -7.10370317e-02 -5.65931499e-01 4.20715392e-01
4.30647045e-01 -9.85398829e-01 1.29245973e+00 -1.09841311e+00
1.15292661e-01 -1.72903519e-02 1.51893431e-02 -1.06208062e+00
-3.17463487e-01 -4.48656231e-01 -1.59017608e-01 1.09769762e+00
1.21384382e-01 -7.30129957e-01 6.61765337e-01 5.18010259e-01
2.09878758e-01 -1.83656320e-01 -1.10520399e+00 -1.05249727e+00
4.38809484e-01 -7.37056285e-02 4.97512400e-01 1.03318036e+00
-2.69337654e-01 -8.34311545e-02 -3.21130991e-01 4.85956848e-01
4.00750756e-01 -3.16430569e-01 1.04831076e+00 -1.08001363e+00
-4.00224894e-01 -5.83412826e-01 -1.19489539e+00 -1.80696875e-01
5.30885160e-01 -6.68031454e-01 -4.68633860e-01 -3.60547781e-01
1.89543515e-02 -1.12937577e-02 -1.86229095e-01 7.06258357e-01
-1.12929679e-01 3.89824778e-01 6.90129459e-01 1.03087485e-01
-2.03649774e-02 -4.66530286e-02 7.04400718e-01 -4.03211206e-01
1.48420036e-01 7.56269619e-02 -4.64719653e-01 8.03470314e-01
6.91881120e-01 -8.55431259e-01 6.44446239e-02 -1.98248085e-02
5.74800409e-02 -2.72527456e-01 7.44728744e-01 -1.34768653e+00
2.81014383e-01 1.16674513e-01 5.93049467e-01 1.88859060e-01
1.12643033e-01 -9.62594092e-01 -9.45213363e-02 9.94249463e-01
-3.78654718e-01 -4.55682538e-03 3.95366251e-01 3.87199849e-01
2.92537045e-02 -7.96018124e-01 1.07154369e+00 -2.40382269e-01
-4.98693436e-01 4.78173830e-02 -6.52195573e-01 -2.72594213e-01
1.13265347e+00 -3.63787591e-01 -7.25108147e-01 -6.87019080e-02
-8.59049261e-01 -4.17848974e-01 6.49988949e-01 3.12254071e-01
3.22533816e-01 -8.63541543e-01 -5.49462259e-01 2.26576477e-01
-3.20961982e-01 -8.24075341e-01 5.57964947e-03 4.84086007e-01
-3.87471765e-01 1.93625018e-01 -7.22139835e-01 -1.13902919e-01
-1.28476119e+00 1.24553287e+00 4.83949959e-01 -3.83402318e-01
-4.96256828e-01 3.87272447e-01 -7.54889473e-02 -2.15409592e-01
1.72725976e-01 9.78576839e-02 -2.45734289e-01 -1.49136171e-01
5.20435810e-01 3.82688463e-01 3.62748712e-01 -5.55121243e-01
-2.64113694e-01 3.27762693e-01 -1.89719602e-01 2.31706485e-01
1.13700235e+00 5.68279207e-01 1.65379733e-01 -1.10390879e-01
1.23027802e+00 1.56328648e-01 -9.09389555e-01 5.48569076e-02
7.31271058e-02 -5.27995944e-01 -6.25498816e-02 -4.36127543e-01
-1.26420498e+00 8.59122932e-01 8.00768912e-01 5.98048270e-01
9.56315935e-01 -4.20107096e-01 4.09024537e-01 6.63095772e-01
2.59981602e-01 -5.72814763e-01 2.48358294e-01 4.96148705e-01
5.90365469e-01 -8.12262654e-01 6.57939911e-02 -2.26263180e-01
-2.85691589e-01 1.50573528e+00 5.20433843e-01 -8.49629879e-01
6.54178202e-01 3.92675191e-01 -4.09081787e-01 -1.27635568e-01
-2.36356854e-01 -1.28255948e-01 -7.48012736e-02 7.62546539e-01
-2.28532702e-01 -4.00494128e-01 6.49999604e-02 1.80031702e-01
-1.00095391e-01 -3.73677909e-01 8.38250935e-01 9.30128515e-01
-4.02496338e-01 -1.05014956e+00 -7.10972786e-01 2.98752010e-01
-4.48457658e-01 2.26755291e-02 -9.44526553e-01 1.13703120e+00
3.98952633e-01 6.70169950e-01 -4.50376496e-02 -4.38397110e-01
3.92026663e-01 1.04503490e-01 5.62346518e-01 -4.11987036e-01
-1.42656934e+00 -7.04199195e-01 -1.16451979e-01 -5.68312466e-01
-1.38593152e-01 -7.80814767e-01 -6.04861617e-01 -5.40577829e-01
-5.48755169e-01 -2.72441953e-01 6.77250206e-01 1.11246073e+00
1.84044689e-01 3.58444571e-01 8.68430972e-01 -9.49383557e-01
-9.67586398e-01 -9.92368758e-01 -5.18467247e-01 7.26873636e-01
3.97685349e-01 -7.62148559e-01 -8.54259253e-01 -6.39672130e-02] | [5.583834171295166, 7.829930782318115] |
9e6663f9-23ab-4ba1-90fb-eb92019b978e | natural-language-person-search-using-deep | 1809.00365 | null | http://arxiv.org/abs/1809.00365v1 | http://arxiv.org/pdf/1809.00365v1.pdf | Natural Language Person Search Using Deep Reinforcement Learning | Recent success in deep reinforcement learning is having an agent learn how to
play Go and beat the world champion without any prior knowledge of the game. In
that task, the agent has to make a decision on what action to take based on the
positions of the pieces. Person Search is recently explored using natural
language based text description of images for video surveillance applications
(S.Li et.al). We see (Fu.et al) provides an end to end approach for
object-based retrieval using deep reinforcement learning without constraints
placed on which objects are being detected. However, we believe for real-world
applications such as person search defining specific constraints which identify
a person as opposed to starting with a general object detection will have
benefits in terms of performance and computational resources required. In our
task, Deep reinforcement learning would localize the person in an image by
reshaping the sizes of the bounding boxes. Deep Reinforcement learning with
appropriate constraints would look only for the relevant person in the image as
opposed to an unconstrained approach where each individual objects in the image
are ranked. For person search, the agent is trying to form a tight bounding box
around the person in the image who matches the description. The bounding box is
initialized to the full image and at each time step, the agent makes a decision
on how to change the current bounding box so that it has a tighter bound around
the person based on the description of the person and the pixel values of the
current bounding box. After the agent takes an action, it will be given a
reward based on the Intersection over Union (IoU) of the current bounding box
and the ground truth box. Once the agent believes that the bounding box is
covering the person, it will indicate that the person is found. | ['Ankit Shah', 'Tyler Vuong'] | 2018-09-02 | null | null | null | null | ['person-search'] | ['computer-vision'] | [ 1.49152696e-01 1.23682737e-01 2.40283608e-02 -1.87962666e-01
-6.25447690e-01 -5.04374743e-01 3.10343564e-01 4.16822642e-01
-9.26461041e-01 5.37589252e-01 1.44634277e-01 1.56122267e-01
-2.02071965e-01 -8.70792270e-01 -5.93501151e-01 -6.18284762e-01
-1.16969058e-02 1.09534729e+00 4.25026804e-01 -1.83399841e-01
2.70257562e-01 3.43541712e-01 -1.17426729e+00 2.15492249e-01
1.14580899e-01 8.32492948e-01 3.14568758e-01 8.02439928e-01
3.04336756e-01 7.58257747e-01 -8.99594605e-01 -1.18639991e-01
7.68504500e-01 -6.00805044e-01 -9.96359766e-01 4.77856308e-01
4.13114965e-01 -7.03511477e-01 -4.12193984e-01 1.11737967e+00
3.64860475e-01 5.16719282e-01 4.81452644e-01 -1.07410169e+00
-6.70809299e-02 2.43510947e-01 -5.68628550e-01 3.59011441e-01
6.71056271e-01 3.47242326e-01 9.16861653e-01 -4.43697125e-01
7.90234089e-01 1.17656660e+00 3.38208526e-01 6.66893661e-01
-9.79304373e-01 -4.53940541e-01 4.74442631e-01 1.95843250e-01
-1.17780054e+00 -1.29047781e-01 4.11665738e-01 -3.07130247e-01
9.02673423e-01 -1.37882140e-02 9.00301099e-01 4.99587238e-01
-9.33779031e-03 6.66160703e-01 7.18990803e-01 -7.27186382e-01
5.22240400e-01 8.90614316e-02 -1.94143206e-01 9.58666801e-01
7.51539990e-02 2.10538954e-01 -5.02085865e-01 -1.84191331e-01
7.30826318e-01 4.84621264e-02 1.11294456e-01 -5.97303510e-01
-1.05967212e+00 8.31275463e-01 7.55608857e-01 1.78718343e-01
-7.13818491e-01 5.09980619e-01 4.74624515e-01 -4.14165147e-02
-5.02769044e-03 5.87677598e-01 -1.01981744e-01 -1.13950865e-02
-9.69958663e-01 7.51651645e-01 5.53792775e-01 5.05230844e-01
7.25374162e-01 -3.32819134e-01 -2.68268079e-01 4.43103969e-01
2.06192341e-02 3.07334602e-01 1.22829273e-01 -1.26730120e+00
3.05382818e-01 7.54099309e-01 4.45547402e-01 -7.84823775e-01
-2.03777209e-01 -3.67912985e-02 -2.63100713e-01 7.37594128e-01
7.14943290e-01 -3.92521411e-01 -9.56833303e-01 1.59569895e+00
4.97527510e-01 -2.78791897e-02 1.33043528e-01 1.32483804e+00
4.25582439e-01 6.21535003e-01 1.62151739e-01 6.18012473e-02
1.55662453e+00 -7.82641828e-01 -1.40953317e-01 -7.24029303e-01
3.00762028e-01 -3.65770012e-01 6.12324655e-01 3.61818463e-01
-1.05675733e+00 -6.00072324e-01 -1.05445611e+00 2.10436478e-01
-1.49333313e-01 2.39929231e-03 3.44117224e-01 3.19039643e-01
-8.84130597e-01 3.91947806e-01 -6.79561675e-01 -6.93893552e-01
2.99482197e-01 4.14049029e-01 -4.07940716e-01 -1.48293376e-01
-1.10272777e+00 8.93077672e-01 6.13190114e-01 1.39121842e-02
-1.17293930e+00 3.85134146e-02 -8.84796917e-01 2.78060794e-01
9.42611933e-01 -7.26621032e-01 1.22180915e+00 -1.31168151e+00
-1.06564343e+00 8.33238304e-01 -3.63285579e-02 -8.03083301e-01
6.33761585e-01 -1.52602019e-02 1.80132136e-01 3.67448866e-01
4.64984804e-01 1.16981375e+00 8.02795768e-01 -9.81151938e-01
-1.19839680e+00 -5.88993430e-01 6.51735067e-01 8.11363399e-01
4.93226536e-02 2.98958808e-01 -7.33187675e-01 -2.66208351e-01
8.51557851e-02 -1.03074765e+00 -2.90654659e-01 1.80863276e-01
-5.81038445e-02 -4.59212333e-01 6.76537693e-01 -5.14037669e-01
6.57130361e-01 -2.00281620e+00 -1.63875967e-01 2.46973291e-01
1.25559881e-01 1.14894681e-01 -1.25246584e-01 2.61947542e-01
2.44215384e-01 -1.57562926e-01 9.21658874e-02 -1.38567969e-01
-7.07810447e-02 1.56092331e-01 -5.82131930e-02 4.77985770e-01
1.58532150e-02 6.50323629e-01 -8.97199392e-01 -4.18146789e-01
1.99408129e-01 7.01660514e-02 -6.50978088e-01 2.54385382e-01
-3.88649583e-01 1.16010703e-01 -5.47087729e-01 2.18224078e-01
2.67049581e-01 -8.89091119e-02 9.87387684e-05 2.50773191e-01
-6.38104603e-02 2.00931981e-01 -1.56501305e+00 1.23523355e+00
1.05323367e-01 5.90448260e-01 1.40368044e-01 -8.69290411e-01
7.62211502e-01 2.22076029e-01 5.71529806e-01 -5.96911430e-01
-4.69116867e-02 -1.26346007e-01 2.09317118e-01 -2.88270980e-01
4.46177125e-01 -2.01341227e-01 -2.05688849e-01 6.14990652e-01
-2.93611050e-01 7.78697059e-03 4.73006457e-01 2.97644615e-01
1.14329600e+00 2.68962979e-01 2.45244980e-01 1.23420455e-01
3.49825054e-01 4.67847973e-01 4.72448438e-01 1.20740163e+00
-4.80284363e-01 4.21716630e-01 4.40918595e-01 -8.89289200e-01
-1.06533206e+00 -7.54241228e-01 3.19878727e-01 1.21823227e+00
3.54701161e-01 -2.26129785e-01 -9.93710220e-01 -6.66745961e-01
-1.28341079e-01 2.85371214e-01 -8.02384615e-01 -1.08501926e-01
-7.30199277e-01 -2.18567401e-01 2.42030472e-01 4.38889652e-01
8.88610303e-01 -1.53511310e+00 -1.53551471e+00 1.45359114e-01
-5.20122163e-02 -9.46822822e-01 -7.35630155e-01 2.34455705e-01
-4.85729098e-01 -1.16965711e+00 -6.84715748e-01 -7.27506876e-01
9.04666066e-01 -2.73685325e-02 6.59852326e-01 1.43199489e-01
-4.04564887e-01 4.53999877e-01 -2.46501490e-01 -3.12619895e-01
-1.40675873e-01 -3.71808171e-01 5.72607527e-03 -2.30949551e-01
4.27367210e-01 5.02294302e-01 -8.02060187e-01 2.35685259e-01
-5.62472343e-01 -8.66141543e-02 3.48648816e-01 6.46707952e-01
5.09449184e-01 4.90264207e-01 1.89877644e-01 -5.58639407e-01
6.68962598e-01 -5.41615188e-02 -7.62400210e-01 1.77433163e-01
-2.36254603e-01 1.28830418e-01 2.46799827e-01 -3.53993118e-01
-6.76524282e-01 2.74728030e-01 3.97945762e-01 -3.24648827e-01
-2.46919736e-01 3.71650249e-01 1.73929766e-01 4.37664211e-01
6.71595752e-01 1.10614002e-01 -1.79645061e-01 8.34405124e-02
-8.05986673e-02 2.83516228e-01 6.91055894e-01 -4.66950506e-01
3.85745108e-01 4.78766412e-01 -1.50641412e-01 -1.74331978e-01
-6.86402738e-01 -6.03089094e-01 -5.64138532e-01 -4.23156768e-01
1.13112867e+00 -7.89186299e-01 -1.10792768e+00 -3.33687365e-02
-1.05932963e+00 -2.93293893e-01 -2.97832429e-01 4.54223096e-01
-6.06576085e-01 4.39226367e-02 -2.37100616e-01 -1.07567048e+00
-1.88436925e-01 -1.12275696e+00 9.04905856e-01 4.82597053e-01
-4.11795944e-01 -4.77684975e-01 -3.63511480e-02 4.81026888e-01
-8.51122197e-04 9.45965797e-02 7.02735722e-01 -7.65208066e-01
-6.91636741e-01 -5.96188426e-01 -7.70596638e-02 7.29852691e-02
-7.39931613e-02 -4.38238442e-01 -5.56762695e-01 -4.48011518e-01
-2.72465974e-01 -2.44157016e-01 6.98418677e-01 5.78862607e-01
6.90513253e-01 -2.71978587e-01 -4.87161815e-01 -2.56540459e-02
1.30530798e+00 6.51631117e-01 3.74561310e-01 7.87317753e-01
1.35252550e-01 7.12750077e-01 6.54896140e-01 4.87803280e-01
1.10327557e-01 7.95266151e-01 5.22955239e-01 -6.85569495e-02
-3.95222977e-02 -1.40010864e-01 4.57994968e-01 -7.89822042e-01
9.83079150e-02 -3.99574697e-01 -7.84648895e-01 5.98242044e-01
-1.99736321e+00 -1.30304730e+00 6.17131650e-01 2.51548815e+00
4.83450115e-01 4.46938157e-01 4.82978493e-01 -1.26251221e-01
7.58160770e-01 -3.79723273e-02 -8.37217867e-01 -5.39379060e-01
2.84945935e-01 -1.89152271e-01 4.72008377e-01 6.15495980e-01
-1.08190536e+00 1.13730776e+00 5.50557470e+00 2.60453224e-01
-8.57453525e-01 -2.56591022e-01 6.72647238e-01 -3.81256670e-01
5.29648602e-01 8.59770030e-02 -9.74942088e-01 2.28544608e-01
4.25249904e-01 6.30857572e-02 6.44843936e-01 6.32257223e-01
4.58861619e-01 -8.81006300e-01 -1.38627863e+00 7.68547595e-01
2.43370697e-01 -9.70670223e-01 -3.04620385e-01 1.60105899e-01
6.09809458e-01 3.35904793e-03 -1.51161045e-01 2.95124561e-01
5.33616900e-01 -9.60420728e-01 8.09028864e-01 4.41300362e-01
2.73608893e-01 -9.49787974e-01 7.80135274e-01 6.72979891e-01
-9.29802954e-01 -3.12334001e-01 -3.84689063e-01 -3.09569955e-01
-2.42665242e-02 -1.22633010e-01 -1.37622738e+00 -1.63491890e-01
9.37265754e-01 1.28772836e-02 -2.27640063e-01 9.69790995e-01
-1.45248353e-01 1.12064719e-01 -3.74546647e-01 -1.65388122e-01
5.21433949e-01 -8.94979611e-02 6.11924887e-01 9.89830852e-01
-9.36489627e-02 4.25675184e-01 8.72072399e-01 8.83054912e-01
1.54334143e-01 -1.16444610e-01 -3.80354077e-01 3.47378343e-01
2.69231945e-01 9.37730432e-01 -9.56320345e-01 -6.12378776e-01
-2.07491204e-01 1.01879191e+00 2.48236105e-01 3.45675975e-01
-5.25709093e-01 -2.57331252e-01 4.46509302e-01 4.71024305e-01
4.75756794e-01 6.13676347e-02 1.63343474e-02 -4.94113356e-01
1.53868506e-02 -8.96675289e-01 7.96183765e-01 -1.08224475e+00
-6.83672905e-01 7.08826542e-01 8.41651633e-02 -9.81058300e-01
-6.14840031e-01 -5.28688431e-01 -5.21006823e-01 1.04446232e+00
-8.92482042e-01 -7.83325374e-01 -2.51842290e-01 4.94624972e-01
6.95594192e-01 -3.09343129e-01 7.02709794e-01 -2.22968131e-01
-2.42286310e-01 1.01727910e-01 -3.46255809e-01 6.35894775e-01
4.62765068e-01 -1.15036988e+00 5.28883887e-03 7.87306726e-01
2.53572077e-01 5.95418155e-01 8.17833662e-01 -9.01826262e-01
-7.42672205e-01 -4.79782581e-01 7.65737891e-01 -4.23898309e-01
2.28863642e-01 -2.38600105e-01 -5.63156545e-01 6.41371489e-01
2.24708319e-01 6.48834631e-02 1.29713833e-01 -1.43863708e-01
-4.18275855e-02 -8.00462887e-02 -1.33582699e+00 6.93602860e-01
4.21827793e-01 -2.69306868e-01 -4.47064757e-01 5.33074856e-01
1.06757134e-01 -4.87009317e-01 -1.35661647e-01 1.34652868e-01
4.53704298e-01 -8.73620152e-01 7.60210991e-01 -7.91335940e-01
1.42049044e-01 -6.69791579e-01 1.52726322e-02 -1.18863428e+00
-3.94768953e-01 -3.26455653e-01 3.39083791e-01 5.68892837e-01
2.31490001e-01 6.13530865e-03 1.23988628e+00 1.10823560e+00
3.25180084e-01 -7.06363082e-01 -9.98435795e-01 -3.40278298e-01
-2.28090823e-01 -1.05728783e-01 2.77921885e-01 2.70927370e-01
2.61241317e-01 2.33484790e-01 -2.67777324e-01 3.35089236e-01
3.35694492e-01 1.14981405e-01 8.68882120e-01 -9.65328276e-01
-4.08723146e-01 -2.65177995e-01 -3.95481855e-01 -9.74788606e-01
-7.57760778e-02 -5.08083642e-01 3.09371322e-01 -1.86753798e+00
6.66529536e-01 -2.81525940e-01 -1.99127018e-01 7.58265615e-01
-9.15467590e-02 2.77193058e-02 5.49730301e-01 1.98905393e-01
-9.04995978e-01 -8.54961202e-02 1.09983265e+00 -3.73093933e-01
-2.83041269e-01 2.22207934e-01 -4.27920371e-01 7.85042942e-01
7.72187650e-01 -4.61932629e-01 -1.07688293e-01 -3.10878992e-01
1.17689922e-01 5.07120490e-01 5.88258982e-01 -1.12266243e+00
5.98441899e-01 -2.19326943e-01 8.84598076e-01 -4.74353373e-01
5.64770222e-01 -1.00000596e+00 -1.38079643e-01 9.19084549e-01
-6.87217057e-01 2.68347681e-01 1.94205567e-01 4.27919805e-01
1.81662202e-01 -6.93305552e-01 7.60235190e-01 -5.87804258e-01
-9.53628778e-01 1.10043913e-01 -5.32450736e-01 -2.31693909e-01
1.11778331e+00 -6.22933388e-01 7.36208558e-02 -7.68133581e-01
-9.87244427e-01 5.17523348e-01 3.17264408e-01 2.16615468e-01
7.17883825e-01 -8.76069784e-01 -7.51183391e-01 7.26157874e-02
5.07315807e-02 -3.78099456e-02 -2.13388614e-02 9.06407163e-02
-4.53076243e-01 4.08816785e-01 -2.12123141e-01 -3.59747529e-01
-1.45118940e+00 7.31023192e-01 7.96709657e-01 -3.81289601e-01
-6.23511016e-01 7.84809172e-01 4.03638601e-01 3.83743607e-02
4.94927794e-01 -1.23463362e-01 -2.68279761e-01 -1.55061752e-01
6.02663696e-01 1.95726886e-01 -1.58353195e-01 -6.93032384e-01
-4.55218464e-01 3.54780018e-01 -4.07164425e-01 -5.43735027e-01
1.15722537e+00 5.37494682e-02 2.14234106e-02 -6.22555539e-02
7.44886100e-01 -1.15244783e-01 -1.71651876e+00 -2.43595615e-01
-1.57343775e-01 -5.70639014e-01 3.79478745e-02 -1.15124631e+00
-8.33007216e-01 3.85269731e-01 8.33723068e-01 -4.46999297e-02
9.40895915e-01 2.04114065e-01 4.52237666e-01 5.58522582e-01
5.01694739e-01 -1.40338969e+00 5.18059134e-01 4.45434451e-01
8.21211755e-01 -1.22527242e+00 3.61506879e-01 1.02740489e-01
-9.62910235e-01 1.03895152e+00 8.91240954e-01 -3.84965271e-01
5.49611673e-02 1.25403568e-01 -3.16182664e-03 -3.99400234e-01
-5.94654918e-01 -3.39003414e-01 1.90015420e-01 5.10423362e-01
-1.42512560e-01 3.75214070e-02 1.10875741e-01 2.02651843e-02
-1.62936136e-01 -9.16431472e-02 4.57903892e-01 8.66786838e-01
-9.11620557e-01 -7.87007570e-01 -8.32653582e-01 2.67198503e-01
-3.59954447e-01 1.78692669e-01 -5.99566936e-01 6.00353956e-01
5.41528285e-01 9.41393495e-01 4.06036824e-01 1.49578452e-01
4.35371459e-01 -1.31318629e-01 5.50941944e-01 -8.25580835e-01
-6.14870846e-01 2.28053659e-01 -1.27398521e-01 -1.70373097e-01
-2.15419784e-01 -8.39344800e-01 -1.80252838e+00 -2.03495715e-02
-2.15002507e-01 1.60902932e-01 4.40936804e-01 9.60755706e-01
-3.73471268e-02 1.88396469e-01 2.14474529e-01 -6.50886059e-01
-3.67344379e-01 -7.07609355e-01 -3.15655470e-01 4.33214158e-01
4.38078910e-01 -5.21142840e-01 1.54046923e-01 4.00315151e-02] | [9.22881031036377, 0.5381669402122498] |
573c1492-ce75-4e42-aca0-f6263a858fef | keypoint-based-weakly-supervised-human | 1809.05285 | null | http://arxiv.org/abs/1809.05285v1 | http://arxiv.org/pdf/1809.05285v1.pdf | Keypoint Based Weakly Supervised Human Parsing | Fully convolutional networks (FCN) have achieved great success in human
parsing in recent years. In conventional human parsing tasks, pixel-level
labeling is required for guiding the training, which usually involves enormous
human labeling efforts. To ease the labeling efforts, we propose a novel weakly
supervised human parsing method which only requires simple object keypoint
annotations for learning. We develop an iterative learning method to generate
pseudo part segmentation masks from keypoint labels. With these pseudo masks,
we train an FCN network to output pixel-level human parsing predictions.
Furthermore, we develop a correlation network to perform joint prediction of
part and object segmentation masks and improve the segmentation performance.
The experiment results show that our weakly supervised method is able to
achieve very competitive human parsing results. Despite our method only uses
simple keypoint annotations for learning, we are able to achieve comparable
performance with fully supervised methods which use the expensive pixel-level
annotations. | ['Guosheng Lin', 'Zhonghua Wu', 'Jianfei Cai'] | 2018-09-14 | null | null | null | null | ['human-parsing'] | ['computer-vision'] | [ 4.68404323e-01 6.07377946e-01 -1.85099587e-01 -6.77822351e-01
-8.92775059e-01 -5.26699007e-01 3.72990787e-01 5.86120337e-02
-6.57347739e-01 5.16746998e-01 -1.96041092e-01 -3.35668921e-01
6.11972809e-01 -7.91455507e-01 -9.22578096e-01 -3.31768423e-01
4.37058091e-01 5.15534401e-01 7.61655927e-01 7.97319710e-02
-5.65307438e-02 2.87374228e-01 -1.13777542e+00 2.91378528e-01
6.67123914e-01 8.54453146e-01 5.30954301e-01 8.02999079e-01
-6.18327558e-01 8.78694534e-01 -5.45537114e-01 -4.51073706e-01
2.73633629e-01 -5.60143411e-01 -1.24021208e+00 2.69604355e-01
3.51414740e-01 -2.22660586e-01 1.31172597e-01 1.24948502e+00
2.90522397e-01 -1.56949773e-01 6.70013204e-02 -8.87061477e-01
-2.40475103e-01 8.37034464e-01 -4.94562268e-01 -1.92194551e-01
1.28904864e-01 4.19019945e-02 9.84027803e-01 -6.37433231e-01
5.38046598e-01 1.20123649e+00 8.32536042e-01 9.11039650e-01
-9.80450153e-01 -5.16857207e-01 3.04820061e-01 -2.04289570e-01
-8.25622201e-01 1.05926283e-02 8.58159840e-01 -1.95831954e-01
6.44480526e-01 -6.82239234e-02 5.91203451e-01 7.11569965e-01
-2.64146209e-01 1.21722722e+00 1.24867332e+00 -5.32082498e-01
2.28928074e-01 -8.48408490e-02 4.21224564e-01 1.16525996e+00
1.89801566e-02 -3.24590117e-01 -2.28351802e-01 1.59387112e-01
1.06874382e+00 -1.04353331e-01 1.04127549e-01 -2.83343464e-01
-1.12192976e+00 7.44703174e-01 6.90638840e-01 3.28559756e-01
-1.72700167e-01 3.52125823e-01 2.91747034e-01 -3.21509629e-01
3.44999999e-01 3.21546316e-01 -7.83697188e-01 1.52054518e-01
-1.00102675e+00 -4.32224311e-02 7.06053674e-01 9.78589416e-01
9.96938825e-01 -3.72994155e-01 -2.30939463e-01 8.55058134e-01
1.45621985e-01 2.55349785e-01 3.46506298e-01 -1.36149549e+00
6.03848994e-01 9.05761898e-01 1.00730091e-01 -5.44746757e-01
-7.03083158e-01 -2.91920394e-01 -7.46281147e-01 2.33335763e-01
7.05257952e-01 -2.63586581e-01 -1.46268928e+00 1.45709062e+00
3.74966294e-01 -3.02615948e-02 -7.48663694e-02 8.17330480e-01
8.60854506e-01 4.99959677e-01 5.54686844e-01 5.69907613e-02
1.43672955e+00 -1.40486801e+00 -6.01895273e-01 -4.06496644e-01
8.04146409e-01 -5.73477268e-01 1.22003305e+00 1.80393338e-01
-1.21608210e+00 -8.89418483e-01 -8.94399285e-01 -2.68004358e-01
-1.94697410e-01 4.55307245e-01 9.83400106e-01 6.09478891e-01
-9.09744263e-01 7.24470973e-01 -1.28622985e+00 -2.80993521e-01
8.20228159e-01 5.21648824e-01 -5.15021503e-01 -1.62342191e-02
-8.25216949e-01 5.62981665e-01 7.26537347e-01 2.91922778e-01
-5.30703008e-01 -3.34192842e-01 -9.67941046e-01 1.89306196e-02
5.25602400e-01 -5.90575159e-01 1.63783336e+00 -1.17925072e+00
-1.64226770e+00 1.12301028e+00 -7.89462253e-02 -4.77249026e-01
6.94512844e-01 -4.26899165e-01 2.27177396e-01 3.93115461e-01
3.20939362e-01 1.44648957e+00 5.24208963e-01 -1.25594604e+00
-9.53122497e-01 -9.51062217e-02 1.30597785e-01 2.56761983e-02
-5.15825935e-02 8.15833360e-02 -1.07435513e+00 -5.23817182e-01
5.42509496e-01 -8.25500846e-01 -7.48281062e-01 3.26774642e-02
-7.34749317e-01 -5.41356504e-01 7.75184274e-01 -5.70792794e-01
7.21423626e-01 -1.91728115e+00 -2.03657269e-01 -8.92471299e-02
2.11738691e-01 4.35509682e-01 -1.71426982e-01 -2.71728426e-01
1.81204498e-01 1.46724328e-01 -6.57603264e-01 -7.64364719e-01
-1.00706123e-01 4.34801131e-01 1.64150268e-01 6.79092705e-02
4.71244127e-01 1.24461985e+00 -8.47920239e-01 -8.06294143e-01
2.51282364e-01 2.08174303e-01 -5.93017578e-01 5.05866051e-01
-5.65002918e-01 5.55353761e-01 -5.55866897e-01 7.87575841e-01
5.92411280e-01 -4.62381482e-01 1.91342264e-01 -1.65845007e-01
3.48361186e-03 3.47312976e-04 -8.08672488e-01 1.89640629e+00
-2.92166412e-01 2.60293156e-01 2.53531158e-01 -1.21468782e+00
7.46244013e-01 1.74468040e-01 2.43218392e-01 -5.11287808e-01
3.09638143e-01 1.88112155e-01 -1.45434827e-01 -2.69623548e-01
2.25156769e-01 -1.83044851e-01 -3.66900504e-01 1.26760304e-01
2.40859970e-01 -3.27534199e-01 2.44539782e-01 9.92303193e-02
1.24143529e+00 5.33127785e-01 6.34448677e-02 -1.52497977e-01
6.91358328e-01 3.42950016e-01 9.88193870e-01 7.24809825e-01
-3.61904353e-01 9.78667200e-01 7.12146521e-01 -7.04861343e-01
-1.07989883e+00 -8.20209444e-01 1.26749605e-01 1.16007447e+00
2.17743590e-01 -2.21621081e-01 -1.33759165e+00 -1.22349024e+00
-4.89545643e-01 1.96676612e-01 -5.18158078e-01 4.08171088e-01
-1.02014434e+00 -7.09031343e-01 5.52364826e-01 1.05322874e+00
1.00606704e+00 -1.46907294e+00 -6.44736409e-01 3.06274742e-01
-6.83507323e-02 -1.63924599e+00 -2.64928967e-01 5.62803984e-01
-1.01317918e+00 -1.09710038e+00 -9.87375617e-01 -1.27577162e+00
1.04379940e+00 -3.49912122e-02 1.08185267e+00 5.09979606e-01
-3.79156083e-01 9.14879367e-02 -3.71302128e-01 -1.45107374e-01
-3.38257879e-01 2.31741473e-01 -4.48470235e-01 -4.82367754e-01
2.93662220e-01 -1.51834443e-01 -5.96628666e-01 3.58307779e-01
-8.32466900e-01 4.57482189e-01 8.84497941e-01 7.81725228e-01
9.13917363e-01 6.98055625e-02 2.75758892e-01 -1.53537118e+00
-7.01439276e-04 2.24973902e-01 -8.48639369e-01 2.84915745e-01
-2.15741634e-01 1.40849337e-01 6.52051568e-01 -1.74563900e-01
-1.25995028e+00 8.91495585e-01 -5.20374537e-01 -5.25446348e-02
-6.06464863e-01 4.41811234e-02 -4.37129080e-01 -6.49019107e-02
2.96603739e-01 -2.18996242e-01 -3.53635222e-01 -7.40750253e-01
4.93757218e-01 4.58987564e-01 8.49319041e-01 -6.93456292e-01
6.33363187e-01 3.74070376e-01 -8.58483240e-02 -3.84874105e-01
-1.26672661e+00 -5.01220107e-01 -1.20125747e+00 1.49772331e-01
1.61648405e+00 -9.80967700e-01 -4.50522959e-01 4.91591722e-01
-1.33777583e+00 -9.11864400e-01 -1.83695525e-01 2.47796431e-01
-5.24036765e-01 4.40762162e-01 -9.45256412e-01 -4.60225165e-01
-2.82789767e-01 -1.02638423e+00 1.32008731e+00 3.65354687e-01
-1.34391626e-02 -8.68442893e-01 -2.40237832e-01 6.79579496e-01
-4.95829061e-02 2.18555823e-01 8.23926747e-01 -6.57513618e-01
-8.00688863e-01 -3.31645161e-01 -5.77485502e-01 4.47379321e-01
2.39523053e-02 -3.26161802e-01 -9.13399518e-01 1.37748763e-01
-3.90527338e-01 -4.04531807e-01 9.53869581e-01 3.86032522e-01
1.50354099e+00 1.83266476e-01 -5.59709013e-01 6.09330595e-01
1.20945668e+00 -6.30001277e-02 3.69593769e-01 2.17500120e-01
1.14371562e+00 8.17421556e-01 7.18369961e-01 -4.24505100e-02
3.57748568e-01 3.30546737e-01 1.14166133e-01 -6.28504694e-01
-4.30189699e-01 -5.02040744e-01 1.49291684e-03 4.97324169e-01
-1.65129915e-01 -1.29106015e-01 -9.19734240e-01 5.01385629e-01
-1.98452353e+00 -4.36410576e-01 -2.33205810e-01 1.68122947e+00
9.63001788e-01 6.18147552e-01 5.20166680e-02 -2.58935709e-02
9.23296154e-01 -1.17495403e-01 -4.34578866e-01 -1.52318195e-01
1.76901087e-01 6.41012192e-01 6.46355689e-01 5.02725720e-01
-1.53543079e+00 1.67658949e+00 6.57766151e+00 4.99472857e-01
-8.29111040e-01 2.25866258e-01 8.51335287e-01 3.91821653e-01
2.42038984e-02 -3.22243683e-02 -1.05113757e+00 1.36625648e-01
6.18231833e-01 7.74526596e-01 -2.71747708e-01 1.11943269e+00
5.91391400e-02 -1.32907748e-01 -1.02229583e+00 7.46668696e-01
-3.36761981e-01 -1.13791227e+00 -9.52797681e-02 -2.22181335e-01
8.32849681e-01 -1.77937806e-01 -4.78169709e-01 3.25193226e-01
5.94586194e-01 -9.78103995e-01 5.87375164e-01 1.96432799e-01
4.74715590e-01 -6.74674928e-01 9.43437040e-01 5.52591860e-01
-1.22380018e+00 6.26220554e-02 -5.26295602e-01 -7.47408941e-02
4.40828174e-01 6.76822126e-01 -6.38864160e-01 6.87977746e-02
8.14439893e-01 3.41093510e-01 -5.90212345e-01 7.45423257e-01
-8.32316816e-01 7.29329228e-01 -2.56839335e-01 2.77787913e-02
4.10906106e-01 5.52241616e-02 -1.28629163e-01 1.35615468e+00
-3.34682874e-02 3.21302861e-01 7.39970624e-01 8.42719615e-01
-3.66891831e-01 -5.32771088e-03 -7.06335455e-02 3.76172969e-03
1.45858571e-01 1.57403159e+00 -1.53471613e+00 -5.96741199e-01
-4.89087611e-01 1.41944838e+00 5.33447027e-01 5.79862073e-02
-6.79110229e-01 -2.93018371e-01 7.03198239e-02 -1.40062636e-02
5.05020976e-01 -2.73823082e-01 -5.19883275e-01 -9.22493577e-01
2.13325545e-02 -3.86852145e-01 2.77140111e-01 -6.90973163e-01
-1.02461731e+00 6.64061964e-01 -9.03548747e-02 -6.64695203e-01
-6.74897358e-02 -9.03301716e-01 -7.29303062e-01 6.93474650e-01
-1.52854681e+00 -1.52211654e+00 -3.96580964e-01 3.88088524e-01
7.06941545e-01 2.62545943e-01 7.21962392e-01 1.15859427e-01
-5.95630527e-01 4.67499018e-01 -8.01563025e-01 7.67666996e-01
3.56608748e-01 -1.77749598e+00 8.03294063e-01 9.47779894e-01
1.32498845e-01 4.97040808e-01 3.73852879e-01 -8.28394234e-01
-7.52352953e-01 -1.09401751e+00 8.92332792e-01 -4.35541183e-01
2.47658089e-01 -5.77269256e-01 -7.90192127e-01 6.09606802e-01
9.82497633e-02 4.78999048e-01 5.21405339e-01 -1.12428772e-03
-1.25549093e-01 2.46969730e-01 -1.10645473e+00 3.23704123e-01
1.18894684e+00 -2.27930889e-01 -6.06051028e-01 4.25963461e-01
1.03415012e+00 -4.62903708e-01 -5.07481217e-01 7.57389784e-01
2.91393101e-01 -8.55461061e-01 8.57159376e-01 -4.16550279e-01
3.54959518e-01 -4.92571890e-01 1.55201942e-01 -7.11902082e-01
-3.09966952e-01 -3.02474380e-01 1.83731735e-01 1.35885215e+00
6.12949610e-01 -1.51113272e-01 1.52116847e+00 8.41906726e-01
-2.19218910e-01 -6.19950473e-01 -5.11147022e-01 -5.24494886e-01
9.52244848e-02 -3.28054398e-01 1.84527874e-01 5.76442003e-01
-3.16874385e-01 3.21028084e-01 -1.22621141e-01 3.30017418e-01
7.17975974e-01 2.34869257e-01 7.13224947e-01 -1.33939922e+00
-3.82418692e-01 -2.31649846e-01 -2.45865658e-01 -1.32659090e+00
3.74924660e-01 -6.51590347e-01 7.70527542e-01 -1.78417790e+00
2.52186239e-01 -5.58243215e-01 -2.28413463e-01 9.01199698e-01
-6.41831160e-01 6.21547222e-01 2.00577557e-01 4.86193188e-02
-8.69435966e-01 -2.08085906e-02 1.40511131e+00 -2.55482690e-03
-2.54887372e-01 1.83661491e-01 -5.09494901e-01 1.19434905e+00
7.93550253e-01 -3.85054797e-01 -1.14880219e-01 -4.66359943e-01
-7.62982666e-02 -7.27986991e-02 3.75752211e-01 -1.13292921e+00
2.10558236e-01 1.44080013e-01 6.14243031e-01 -7.30117738e-01
5.94787374e-02 -7.07475126e-01 -4.25864041e-01 5.79131365e-01
-3.29703271e-01 -1.80119693e-01 9.37891006e-03 4.79349405e-01
-2.45319635e-01 -4.54057515e-01 8.44352663e-01 -7.03460693e-01
-1.03081954e+00 3.41727525e-01 -6.39420375e-02 -1.09028898e-01
6.98990524e-01 -6.72843978e-02 1.43852353e-01 6.05724342e-02
-9.92222190e-01 3.42303544e-01 4.27269071e-01 1.65548399e-01
2.28894144e-01 -9.58022058e-01 -3.63504767e-01 -2.93799816e-03
-9.57257599e-02 6.02641940e-01 -5.10200448e-02 4.39282835e-01
-8.97891045e-01 3.67512286e-01 -9.40482318e-03 -6.73019707e-01
-1.27075803e+00 4.27866578e-01 1.40216827e-01 -4.20991659e-01
-8.31989050e-01 1.17490184e+00 3.33291680e-01 -8.89051318e-01
3.40621799e-01 -5.92887342e-01 -1.33501559e-01 -3.54451448e-01
3.56018603e-01 -2.10476704e-02 -1.75044164e-01 -3.78392488e-01
-2.42964312e-01 8.12848866e-01 -7.12073117e-04 -1.10665314e-01
1.35028315e+00 1.41747802e-01 -4.71378565e-02 4.44496125e-02
1.07447100e+00 -1.29419968e-01 -1.69737792e+00 -1.97064102e-01
2.99158692e-01 -2.76009403e-02 -5.86508438e-02 -9.15160477e-01
-1.07838356e+00 1.11941886e+00 5.22370577e-01 2.78858189e-02
1.00587881e+00 2.79867232e-01 1.20941377e+00 4.69985366e-01
4.18431073e-01 -1.22818363e+00 6.36954382e-02 4.65137094e-01
2.01374516e-01 -1.53177428e+00 -1.57558277e-01 -9.16855693e-01
-4.07572985e-01 1.14883947e+00 9.39830899e-01 -1.66467324e-01
3.37228090e-01 3.48463446e-01 4.44278181e-01 -8.87482837e-02
-1.43407941e-01 -5.81903994e-01 2.66485810e-01 4.69558388e-01
5.83731472e-01 9.61450860e-02 -4.27579433e-01 7.46338248e-01
-1.20525301e-01 -1.82517752e-01 7.99427330e-02 9.55738127e-01
-8.30745637e-01 -1.69162691e+00 -2.13582844e-01 -7.74902152e-03
-8.03193510e-01 -2.36767828e-02 -5.58252454e-01 7.73877978e-01
3.95772338e-01 7.41303563e-01 -1.83324760e-03 -1.56420749e-02
2.47869581e-01 2.58911699e-01 4.63875979e-01 -1.02625084e+00
-5.15561402e-01 2.91099012e-01 -4.14118264e-03 -6.99406385e-01
-6.65743589e-01 -2.91567802e-01 -1.93351030e+00 4.16198730e-01
-3.04802150e-01 2.04023987e-01 6.50338888e-01 1.17780674e+00
-1.33545354e-01 6.26227856e-01 2.04608873e-01 -9.24251497e-01
-5.25560938e-02 -9.11631167e-01 -3.13483238e-01 4.76225346e-01
4.84412462e-02 -2.60465801e-01 -5.49365906e-03 3.34644079e-01] | [9.410160064697266, 0.4702279567718506] |
59134d5b-5edd-4ff9-9c8a-7354a61b5b12 | fault-detection-via-occupation-kernel | 2303.11138 | null | https://arxiv.org/abs/2303.11138v2 | https://arxiv.org/pdf/2303.11138v2.pdf | Fault Detection via Occupation Kernel Principal Component Analysis | The reliable operation of automatic systems is heavily dependent on the ability to detect faults in the underlying dynamical system. While traditional model-based methods have been widely used for fault detection, data-driven approaches have garnered increasing attention due to their ease of deployment and minimal need for expert knowledge. In this paper, we present a novel principal component analysis (PCA) method that uses occupation kernels. Occupation kernels result in feature maps that are tailored to the measured data, have inherent noise-robustness due to the use of integration, and can utilize irregularly sampled system trajectories of variable lengths for PCA. The occupation kernel PCA method is used to develop a reconstruction error approach to fault detection and its efficacy is validated using numerical simulations. | ['Rushikesh Kamalapurkar', 'Yingzhao Lian', 'Benjamin P. Russo', 'Zachary Morrison'] | 2023-03-20 | null | null | null | null | ['fault-detection'] | ['miscellaneous'] | [-2.35542469e-02 -3.90289903e-01 2.24728018e-01 1.31013989e-01
-5.41113377e-01 -2.84001857e-01 4.24516201e-01 3.27759743e-01
1.03268616e-01 5.60476303e-01 -4.27203685e-01 -2.77394056e-01
-7.95866787e-01 -5.31289697e-01 -2.96771228e-01 -9.64652002e-01
-2.47624069e-01 2.29541048e-01 3.13803613e-01 -2.77937859e-01
2.24419907e-01 8.28425884e-01 -1.39490354e+00 -6.18213356e-01
7.76487708e-01 9.79843676e-01 -1.20488361e-01 5.11654437e-01
4.08950925e-01 5.53014755e-01 -5.63495278e-01 3.34172308e-01
1.32541850e-01 -4.95545805e-01 -3.68845403e-01 4.70037520e-01
-5.95514834e-01 -2.71277428e-02 -6.00997984e-01 8.65604699e-01
2.26455972e-01 3.18776846e-01 7.08167613e-01 -1.29020286e+00
-2.46711429e-02 -1.19247176e-01 -2.34744295e-01 2.09551528e-01
2.88194001e-01 -1.43277564e-03 3.90849859e-01 -9.61325586e-01
9.49481577e-02 7.44775832e-01 7.17404246e-01 1.67431042e-01
-1.52342105e+00 -4.72427607e-02 -3.04080814e-01 2.46437073e-01
-1.47539401e+00 -2.50349581e-01 1.08292067e+00 -6.57013774e-01
1.00533116e+00 2.94881761e-01 6.37615442e-01 5.55585504e-01
8.07635069e-01 2.17803627e-01 1.16202986e+00 -4.60563332e-01
6.61702454e-01 -1.18615471e-01 2.33249664e-01 6.37402236e-01
6.25167787e-01 2.19300687e-01 -1.02489507e-02 -4.43581045e-01
8.25007677e-01 3.13356459e-01 -4.91714746e-01 -6.32155538e-01
-7.42891848e-01 6.78476274e-01 7.10482001e-02 2.90753454e-01
-6.13614202e-01 1.46443918e-01 4.17810261e-01 1.16372287e-01
3.58705938e-01 4.02474016e-01 -2.15490162e-01 -4.75411475e-01
-8.08558941e-01 2.37246409e-01 6.88743591e-01 4.67709452e-01
5.19706964e-01 5.56167364e-01 4.56273109e-01 4.59341109e-01
2.36671448e-01 7.41892397e-01 4.95081395e-01 -8.10127497e-01
-2.14993462e-01 9.30871785e-01 1.51837513e-01 -1.06731343e+00
-4.20168698e-01 -2.58450598e-01 -8.65522504e-01 3.77084106e-01
-1.31182477e-01 -5.80409728e-02 -7.21131742e-01 1.12243247e+00
2.55181134e-01 4.38102260e-02 1.62279427e-01 6.35054588e-01
-4.52037692e-01 6.02234840e-01 -5.87572515e-01 -3.88767362e-01
8.78981590e-01 -3.16605777e-01 -5.72976351e-01 9.11745280e-02
2.15639725e-01 -4.93775994e-01 7.67459571e-01 4.34468538e-01
-6.70845926e-01 -2.72508115e-01 -1.11492038e+00 7.44692981e-01
-1.83252748e-02 1.10727668e-01 1.77317709e-01 6.69616044e-01
-8.38344991e-01 6.84802651e-01 -1.35894239e+00 -2.38341659e-01
-7.24744350e-02 3.05789560e-01 -4.26178068e-01 1.86384227e-02
-6.40394568e-01 1.02284181e+00 9.75750163e-02 2.57950246e-01
-7.45942593e-01 -3.54349762e-01 -7.00015545e-01 1.60196230e-01
3.33593011e-01 -2.23407924e-01 1.01976442e+00 -6.46459818e-01
-1.50732636e+00 -2.52033770e-01 -2.03298107e-01 -3.49743217e-01
3.17548513e-01 -5.79968877e-02 -5.40855050e-01 5.44510543e-01
-8.84117261e-02 -5.44796109e-01 1.20841050e+00 -1.07113051e+00
-1.45657137e-01 -2.10176408e-02 -4.95789737e-01 -3.02742183e-01
-2.59854734e-01 -2.82759160e-01 -1.99594838e-03 -5.07217646e-01
4.35874194e-01 -1.01442790e+00 -3.68755132e-01 -4.74085212e-01
-2.09665701e-01 2.32056320e-01 1.23576832e+00 -6.04703784e-01
1.23972404e+00 -2.12911153e+00 9.69946086e-02 6.15929306e-01
-2.71463711e-02 2.80831963e-01 5.63760698e-01 1.07869470e+00
-1.36520281e-01 -3.74755651e-01 -6.78305268e-01 -1.97475459e-02
-3.00984353e-01 2.18863934e-01 -3.24578315e-01 8.05369198e-01
6.59444332e-01 1.88545808e-01 -5.76749921e-01 -1.25582576e-01
8.33617210e-01 4.98796225e-01 -2.46567741e-01 2.20426649e-01
3.09061557e-01 1.98077023e-01 -4.56207156e-01 5.82911551e-01
1.14556566e-01 -1.60265923e-01 -9.56543256e-03 -1.27352357e-01
-1.75507993e-01 -1.70414388e-01 -1.25536442e+00 6.83245122e-01
-3.64737421e-01 6.24288738e-01 2.58248597e-01 -1.17447531e+00
1.05129600e+00 5.75867295e-01 8.39311898e-01 -3.75325471e-01
2.22904131e-01 2.28563979e-01 5.64306192e-02 -3.82049918e-01
2.03860998e-01 -1.57910049e-01 -1.05933182e-01 3.45303804e-01
-2.01171592e-01 -3.05633217e-01 8.81654769e-02 -1.96942613e-02
1.52505624e+00 -9.19417888e-02 3.84279341e-01 -5.20570338e-01
5.64891100e-01 3.06143820e-01 5.84955156e-01 1.09397769e-01
-1.31067172e-01 2.84378111e-01 5.33959210e-01 -3.15973043e-01
-1.07164562e+00 -7.78651118e-01 -3.78417492e-01 -4.18645777e-02
1.11531779e-01 -2.92001367e-01 -7.82054305e-01 -3.74142915e-01
1.10968150e-01 5.83008528e-01 -5.23962736e-01 -6.63529992e-01
-4.16593879e-01 -6.82448208e-01 2.46247709e-01 5.79712927e-01
1.64502248e-01 -6.10908687e-01 -1.07631779e+00 6.25346839e-01
3.71615857e-01 -6.46696270e-01 6.42877221e-02 4.96393144e-01
-1.05594552e+00 -1.37166262e+00 -4.42051709e-01 -3.65256488e-01
1.03157306e+00 1.89506382e-01 4.62678581e-01 -5.65234199e-02
-7.11180091e-01 9.91236806e-01 -3.11504871e-01 -5.59698641e-02
-5.51566720e-01 -4.46745574e-01 5.49751759e-01 6.21016063e-02
-2.87296884e-02 -4.35718358e-01 -2.27584109e-01 5.24337649e-01
-9.80724871e-01 -4.02089089e-01 6.69580758e-01 1.11406243e+00
5.23079693e-01 7.63243377e-01 7.19129562e-01 -5.54603457e-01
8.65666449e-01 -4.55671370e-01 -9.31741834e-01 7.44262561e-02
-1.16616893e+00 8.72804224e-02 8.40747774e-01 -3.84833723e-01
-8.17297339e-01 1.96225315e-01 1.57908782e-01 -8.05019319e-01
-1.96015209e-01 8.82435203e-01 -2.79631745e-02 -3.24520975e-01
4.70991462e-01 3.62015128e-01 5.35878956e-01 -3.47274214e-01
-1.44961551e-01 3.80017400e-01 4.28484172e-01 -5.44715285e-01
7.00091243e-01 1.61565840e-01 4.28672642e-01 -1.21485150e+00
1.00909673e-01 -4.49625582e-01 -4.16024506e-01 -4.50539529e-01
2.24547759e-01 -5.47469437e-01 -6.38033569e-01 4.73183244e-01
-5.31008601e-01 -8.09921101e-02 -2.99122155e-01 4.55844432e-01
-5.77388763e-01 4.67299193e-01 -6.04058087e-01 -1.23537052e+00
-1.80070579e-01 -1.13322413e+00 6.59373343e-01 1.70441136e-01
-2.19941959e-01 -1.03394854e+00 2.05578357e-01 -3.54934961e-01
6.11062050e-01 4.98561263e-01 9.48743284e-01 -4.58570719e-01
-1.99104398e-01 -1.02125478e+00 1.80945039e-01 5.43221056e-01
4.78787452e-01 4.47688550e-01 -6.54950202e-01 -4.67427313e-01
5.30573666e-01 2.86369592e-01 1.95333838e-01 3.55539709e-01
6.17392004e-01 -1.69503227e-01 -4.10622239e-01 -1.63879320e-01
1.63514864e+00 2.50295281e-01 3.15634966e-01 2.01557085e-01
4.58427578e-01 5.17712295e-01 5.83519161e-01 6.22846007e-01
-1.03024594e-01 4.28860307e-01 4.21697438e-01 1.72906265e-01
3.99397731e-01 1.82008803e-01 3.46374571e-01 1.03474474e+00
4.78386469e-02 2.59650022e-01 -1.14867377e+00 6.05999529e-01
-1.92040384e+00 -6.88589513e-01 -2.96313167e-01 2.25454617e+00
3.11952531e-01 3.60651225e-01 1.15467437e-01 8.59912753e-01
6.09031320e-01 -4.25746828e-01 -4.57994550e-01 -2.34070033e-01
2.15469778e-01 7.08846822e-02 5.68056703e-01 2.40439057e-01
-7.82576680e-01 8.67418796e-02 6.50279140e+00 5.58290720e-01
-1.20089197e+00 -2.71805823e-01 1.80760726e-01 3.40694726e-01
1.96936086e-01 8.55462775e-02 -3.08939070e-01 5.03328323e-01
1.22518349e+00 -2.96292901e-01 2.11951107e-01 1.06255710e+00
5.34920752e-01 -4.85478938e-01 -6.19487524e-01 7.00522423e-01
-2.48516053e-01 -8.17568183e-01 -6.15327597e-01 1.45814940e-01
5.47135949e-01 -4.24557358e-01 -1.21837668e-01 -5.66382259e-02
-1.29651308e-01 -7.86602259e-01 6.08885169e-01 7.91650474e-01
3.97905111e-01 -1.02971089e+00 9.84608293e-01 2.87573189e-01
-1.20676184e+00 -3.21621239e-01 -2.12033302e-01 -2.94136256e-01
4.44152296e-01 9.37794209e-01 -7.80466735e-01 5.69874287e-01
4.34326500e-01 4.08944935e-01 -4.50425774e-01 1.25124466e+00
1.02442116e-01 1.04404342e+00 -4.16890591e-01 -1.24335978e-02
-1.04599304e-01 -2.48598933e-01 5.73471367e-01 5.99325418e-01
4.24461305e-01 -1.40201151e-01 2.85173148e-01 5.29945195e-01
8.45566511e-01 -1.88731313e-01 -6.34722829e-01 -3.45983297e-01
5.31582654e-01 1.13065875e+00 -9.70663369e-01 5.59875369e-02
-5.27996480e-01 6.86976075e-01 -1.26312718e-01 2.64285892e-01
-5.89941382e-01 -6.24246836e-01 6.15196466e-01 4.19769615e-01
2.69290864e-01 -6.36333466e-01 -1.49509594e-01 -7.03523159e-01
9.42647085e-02 -7.54223049e-01 -8.72643013e-03 -2.90281743e-01
-1.19097233e+00 5.71656525e-01 2.70066589e-01 -1.42862964e+00
-6.00156188e-01 -6.65009797e-01 -6.75943434e-01 9.04093564e-01
-7.63313115e-01 -6.17206752e-01 -1.16380125e-01 5.69821894e-01
2.33404249e-01 -2.36136809e-01 1.00941062e+00 1.95339367e-01
-9.30118978e-01 -1.49494186e-01 5.37776887e-01 -1.33968964e-01
2.29770586e-01 -1.10163951e+00 1.83617473e-01 1.19797611e+00
-2.99248248e-01 5.71817160e-01 1.09103787e+00 -8.69078457e-01
-2.00496793e+00 -8.97384107e-01 2.09964573e-01 -6.56998828e-02
7.41501510e-01 2.04997122e-01 -1.02041304e+00 3.76128256e-01
-3.00770313e-01 2.72453427e-01 3.61810982e-01 -3.57187718e-01
1.61422670e-01 -2.24170402e-01 -1.19832158e+00 1.99842468e-01
-8.88069049e-02 -6.00497782e-01 -4.88604635e-01 -2.94400305e-01
2.73052454e-02 1.21249654e-03 -9.24643040e-01 4.35773164e-01
2.05727920e-01 -8.41684878e-01 6.06615186e-01 -3.13828550e-02
-2.50242472e-01 -7.04769135e-01 -3.90080288e-02 -1.33241403e+00
-3.83605808e-01 -4.00085688e-01 -3.61089975e-01 1.04001975e+00
5.12626916e-02 -1.00750268e+00 5.83228171e-01 7.77609169e-01
-1.87016651e-01 -8.16268623e-01 -1.04038775e+00 -1.06907845e+00
-4.07015890e-01 -3.61095339e-01 2.08091632e-01 6.28460824e-01
4.39231306e-01 -4.40501831e-02 -1.43107891e-01 5.08881629e-01
6.43667996e-01 -5.39116971e-02 4.14703041e-01 -1.26728296e+00
-4.44940180e-01 -3.47230941e-01 -7.66791403e-01 -1.39276505e-01
-1.77664921e-01 -1.85884923e-01 5.20276189e-01 -1.26756084e+00
-3.22193712e-01 -4.17710304e-01 -9.42770168e-02 2.40811735e-01
-2.01062888e-01 2.98250262e-02 -3.38707983e-01 3.19118053e-01
-6.65581599e-02 7.36741781e-01 1.95669577e-01 3.27832162e-01
-1.43914446e-01 1.43891901e-01 8.56773481e-02 4.87636596e-01
8.72568607e-01 -4.59470540e-01 -5.53726494e-01 1.59961909e-01
-4.46676612e-02 3.02153021e-01 3.92126322e-01 -1.58247411e+00
2.72480380e-02 -6.95990995e-02 3.24583650e-01 -3.24969262e-01
2.92187244e-01 -1.15560532e+00 6.29776239e-01 8.25869858e-01
2.83442080e-01 6.17436111e-01 3.01793307e-01 8.22513819e-01
-4.69333202e-01 -2.55669951e-01 6.48886204e-01 2.80006379e-01
-5.68613172e-01 -8.89507383e-02 -8.93893301e-01 -6.15552902e-01
1.30386794e+00 -7.79401213e-02 -1.62831601e-02 -2.19643891e-01
-4.37665641e-01 -1.93470493e-01 7.60944366e-01 -1.56347360e-02
7.54651487e-01 -9.90846932e-01 -1.58500329e-01 3.57237130e-01
-3.19283083e-03 -2.31002972e-01 6.91559091e-02 1.01476383e+00
-5.79168260e-01 4.16148424e-01 -6.67375252e-02 -6.48381472e-01
-8.02154958e-01 7.20456541e-01 1.91439152e-01 1.08084105e-01
-8.39366615e-01 2.46987864e-01 -4.95110035e-01 7.35534430e-02
-2.59608001e-01 -1.28767848e-01 3.03214461e-01 -3.09350222e-01
3.50801975e-01 7.99843669e-01 3.44501555e-01 -6.07767463e-01
-3.89793009e-01 3.40025544e-01 3.11352223e-01 -2.30463326e-01
1.14690411e+00 -1.04383297e-01 -1.38629138e-01 7.25882709e-01
8.10602427e-01 -1.52002618e-01 -1.55223775e+00 2.07450718e-01
3.60498101e-01 -2.61456519e-01 3.50735366e-01 -3.31972480e-01
-8.92597675e-01 5.92972159e-01 6.42154217e-01 9.92434144e-01
1.19394016e+00 -4.06666487e-01 4.14470553e-01 8.30003023e-02
5.42474687e-01 -9.15542066e-01 -5.71106337e-02 3.54913473e-01
5.20861924e-01 -6.80781066e-01 6.21278770e-03 -3.38191718e-01
-3.67357790e-01 1.36771762e+00 2.95433372e-01 -5.87663412e-01
8.32494974e-01 5.98140836e-01 -9.69910771e-02 -1.57192960e-01
-6.60240769e-01 -1.49275307e-02 1.17204309e-01 6.78403318e-01
8.89647380e-02 6.49442077e-02 -1.69008479e-01 5.15715420e-01
3.16536486e-01 -1.64957330e-01 5.83091140e-01 1.55163622e+00
-4.75506485e-01 -1.06445181e+00 -6.27220511e-01 6.78092360e-01
-4.96206358e-02 4.74876970e-01 -1.65695287e-02 7.61903882e-01
-5.71016252e-01 9.45867300e-01 -1.45802587e-01 -3.05597663e-01
4.23021853e-01 2.90704876e-01 3.40011477e-01 -2.85350025e-01
-1.38201699e-01 2.09167764e-01 -1.01200409e-01 -4.81269836e-01
5.18343598e-02 -9.00663197e-01 -1.25136471e+00 -1.69970199e-01
-5.82983494e-01 4.59527999e-01 8.38341415e-01 1.07377768e+00
5.74235678e-01 6.73667252e-01 8.56656075e-01 -7.70356774e-01
-5.80495596e-01 -8.27190280e-01 -8.65398109e-01 -5.46188373e-03
3.68896842e-01 -1.05648828e+00 -4.20294076e-01 -7.58595765e-02] | [6.520080089569092, 2.532090187072754] |
88c93277-277d-4570-bc3b-d153698409d3 | a-novel-approach-for-enhancing-sentiment | null | null | https://www.researchgate.net/publication/357150745_A_Novel_Approach_for_Enhancing_Sentiment_Classification_of_Persian_Reviews_Using_Convolutional_Neural_Network_and_Majority_Voting_Classifier | https://www.researchgate.net/publication/357150745_A_Novel_Approach_for_Enhancing_Sentiment_Classification_of_Persian_Reviews_Using_Convolutional_Neural_Network_and_Majority_Voting_Classifier | A Novel Approach for Enhancing Sentiment Classification of Persian Reviews Using Convolutional Neural Network and Majority Voting Classifier | Due to the rapid development of Internet-based applications such as social media, sentiment analysis has become one of the most widely used research areas of natural language processing and an important tool for extracting opinions from texts. Because a huge number of comments and reviews are generated today through social media, and these comments are very significant. However, manually analyzing and summarizing them requires a lot of time and money. Therefore, sentiment analysis has entered the field to organize and analyze them by creating an automated system. In recent years, the use of deep learning in sentiment analysis has shown powerful results. However, creating a model alone may not provide the best predictions and lead to errors such as bias and high variance. To reduce these errors and improve the efficiency of model predictions, combining several models known as ensemble learning may provide better results. Therefore, the main purpose of this article is to create a model based on ensemble learning using Several convolutional neural networks and majority voting classifiers to enhance sentiment analysis in Persian reviews. The proposed model was evaluated on two datasets of reviews of electronic products and movies by 5-fold and 10-fold cross-validation. The results indicate that this new approach increases the efficiency of the sentiment analysis model in Persian language. | ['Milad Vazan'] | 2021-12-17 | null | null | null | digital-transformation-and-intelligent | ['persian-sentiment-anlysis'] | ['natural-language-processing'] | [-1.92647457e-01 -2.32724741e-01 -1.85974233e-03 -6.38788700e-01
-1.11037180e-01 -3.99863690e-01 5.69501162e-01 5.26624620e-01
-5.90254188e-01 8.02025616e-01 1.36479348e-01 -2.28934437e-01
2.09323674e-01 -8.86415720e-01 -7.05634281e-02 -5.26316524e-01
3.47943395e-01 6.55076578e-02 -5.09014539e-03 -6.01967275e-01
7.21245348e-01 2.09345371e-01 -1.74218476e+00 5.35424829e-01
1.16699433e+00 1.14465463e+00 8.44240114e-02 2.50069588e-01
-7.60381341e-01 8.60460699e-01 -6.57986462e-01 -8.07825327e-01
9.81911570e-02 -3.93468052e-01 -4.73103166e-01 5.92418807e-03
-2.97749519e-01 1.14264525e-02 3.18953514e-01 9.40183043e-01
5.75517416e-01 2.07738072e-01 4.73009229e-01 -9.70689893e-01
-6.33702636e-01 7.41884530e-01 -7.75639176e-01 4.91197361e-03
1.14917099e-01 -5.03771782e-01 8.06464076e-01 -9.56931114e-01
3.21081489e-01 1.00842583e+00 5.32170892e-01 1.53517187e-01
-5.34514844e-01 -8.01830411e-01 4.27702293e-02 3.10000271e-01
-7.91609764e-01 -1.20487452e-01 8.85338604e-01 -3.83504450e-01
8.51875067e-01 -1.22174211e-01 6.52066052e-01 7.10043073e-01
7.57642746e-01 7.01273799e-01 1.32830691e+00 -5.97794533e-01
2.52888471e-01 8.75029087e-01 4.66388941e-01 3.46380353e-01
3.38582933e-01 -4.66728002e-01 -3.89282972e-01 6.97905570e-02
-1.54422730e-01 2.53252208e-01 1.71047762e-01 6.49598241e-02
-5.99835753e-01 1.16849411e+00 2.65867949e-01 5.06018519e-01
-4.91823852e-01 -4.85838205e-01 6.91921711e-01 1.78797498e-01
9.83850479e-01 4.87955749e-01 -5.67440093e-01 -1.81325704e-01
-8.42436612e-01 1.18528448e-01 6.89322531e-01 2.37860098e-01
5.77089906e-01 -7.72250770e-03 5.43094695e-01 1.12485719e+00
3.69752914e-01 5.58163166e-01 1.01844680e+00 -1.44547224e-01
3.60993236e-01 1.27172697e+00 1.12239309e-01 -1.67237508e+00
-6.56934083e-01 -5.30839026e-01 -1.11553204e+00 1.88568354e-01
-2.74903551e-02 -5.97158134e-01 -7.57156372e-01 1.13845217e+00
2.50274330e-01 -4.51718360e-01 1.71071023e-01 7.24007070e-01
9.16624963e-01 9.93180454e-01 1.71845108e-01 -2.37776682e-01
1.32630587e+00 -8.54051232e-01 -8.82330894e-01 -2.34181866e-01
6.55918539e-01 -1.08065391e+00 6.55398667e-01 7.08233178e-01
-6.06542170e-01 -8.02440286e-01 -1.11975801e+00 2.00170934e-01
-7.87586093e-01 3.82368416e-01 7.30453670e-01 6.85695112e-01
-6.01349652e-01 5.69825113e-01 -5.14011383e-01 -3.82997245e-01
4.12392378e-01 4.18856204e-01 -4.85217810e-01 1.69623956e-01
-1.41159308e+00 1.10538650e+00 3.25835645e-01 4.16314274e-01
1.98233128e-01 -5.55925444e-02 -6.82245135e-01 -1.71348788e-02
5.71238846e-02 -2.08922580e-01 9.45074856e-01 -1.65630400e+00
-1.27476180e+00 5.02270997e-01 -2.03498587e-01 -2.92005152e-01
8.78733099e-02 -3.36386979e-01 -7.58461356e-01 -3.45858753e-01
1.99022680e-03 2.56070465e-01 4.96736288e-01 -9.35817063e-01
-9.52905655e-01 -6.34798169e-01 9.15897917e-03 1.01729646e-01
-7.70271838e-01 3.55848849e-01 -1.27535692e-04 -3.71692032e-01
-3.45713943e-02 -1.05495954e+00 -2.44677424e-01 -7.84021258e-01
-1.66244730e-01 -3.89616311e-01 8.57720852e-01 -8.10460687e-01
1.31291127e+00 -2.03985143e+00 -1.97829947e-01 2.32533693e-01
-2.53160857e-02 6.27876759e-01 2.11008817e-01 5.00825167e-01
-1.31074771e-01 3.50966632e-01 -1.43679202e-01 -2.70983815e-01
-3.79930973e-01 -9.94129255e-02 -3.30268890e-01 -9.60010197e-03
2.28526220e-01 4.21462774e-01 -4.91218597e-01 -4.38973248e-01
2.81335771e-01 5.57999194e-01 -3.09320301e-01 1.22473732e-01
-6.99310154e-02 3.83999258e-01 -5.92142522e-01 3.23240191e-01
6.61505699e-01 -1.62100911e-01 1.17377646e-01 -1.51337117e-01
-3.59483123e-01 1.48697183e-01 -1.17185080e+00 1.05173290e+00
-6.90032303e-01 5.48125863e-01 -4.12384689e-01 -1.15897870e+00
1.30739343e+00 2.00029507e-01 3.95998150e-01 -6.26358390e-01
7.32024848e-01 2.77777642e-01 2.11577758e-01 -6.16633832e-01
8.79553497e-01 -1.74735293e-01 -1.55202612e-01 4.69421595e-01
-1.73144594e-01 7.19101578e-02 4.93999302e-01 1.57192469e-01
4.17275012e-01 -8.01607221e-02 4.19528931e-01 -9.92939845e-02
9.30663705e-01 1.12602204e-01 5.49011052e-01 -8.18510875e-02
7.07549080e-02 2.98476994e-01 4.10905629e-01 -7.79857278e-01
-7.55111039e-01 -6.43168166e-02 -8.43875334e-02 8.76700640e-01
-4.09095138e-02 -3.96519244e-01 -5.51736355e-01 -7.46788442e-01
-2.10554957e-01 6.28180563e-01 -4.98462439e-01 1.29059795e-02
-2.13167176e-01 -1.05066729e+00 -8.27303976e-02 4.89344835e-01
8.19755316e-01 -1.48347902e+00 -4.36927646e-01 3.12411368e-01
-2.18794137e-01 -9.68595445e-01 2.26962715e-01 2.17811614e-01
-8.48210096e-01 -9.62512255e-01 -6.21856689e-01 -6.47951186e-01
6.39913797e-01 2.83339709e-01 7.89832413e-01 8.77221897e-02
1.51126906e-01 -3.21001083e-01 -9.58153665e-01 -1.09266579e+00
-3.32331806e-01 3.72693002e-01 7.82293677e-02 2.66338140e-01
7.60970116e-01 -1.70760036e-01 -4.79063243e-01 1.54680042e-02
-8.58436346e-01 -2.65226252e-02 5.08951426e-01 6.74820721e-01
1.68517187e-01 4.35907036e-01 1.06422675e+00 -1.10917342e+00
9.75347281e-01 -5.78325391e-01 -5.89549422e-01 2.49943435e-02
-8.61186743e-01 -1.64199471e-02 1.06311977e+00 -8.39955267e-03
-1.40575171e+00 -1.77437320e-01 -2.96872377e-01 3.92049879e-01
-1.64957729e-03 1.15747857e+00 6.79551736e-02 3.80996943e-01
5.19556284e-01 -5.56878448e-02 1.10615894e-01 -2.60855854e-01
-8.65830854e-02 1.28074300e+00 -3.18366170e-01 7.45372549e-02
2.56616682e-01 1.40400231e-01 -2.34841302e-01 -8.25113118e-01
-1.01058435e+00 -4.62577969e-01 -4.79240328e-01 -3.61894071e-01
8.60146761e-01 -7.67574370e-01 -4.20960248e-01 7.89239764e-01
-1.04379439e+00 3.84339303e-01 1.63374752e-01 5.89827478e-01
2.41751313e-01 3.72090846e-01 -2.37216741e-01 -9.85890925e-01
-7.55355597e-01 -1.28031135e+00 5.23972988e-01 6.56457782e-01
-4.20330256e-01 -8.67962718e-01 1.89665720e-01 4.57255065e-01
6.33607805e-01 -5.32414801e-02 7.43997812e-01 -9.34492707e-01
1.11347325e-01 -6.40138090e-01 -8.42801332e-02 9.05186832e-01
2.29192361e-01 3.75538379e-01 -8.97938132e-01 9.05113146e-02
3.19698788e-02 -3.34779590e-01 7.84515262e-01 3.57752800e-01
1.02749681e+00 6.19418919e-02 -1.98330119e-01 -5.29403538e-02
1.31299126e+00 6.79685116e-01 5.87722480e-01 5.07267058e-01
4.17670429e-01 9.12569761e-01 7.74439335e-01 5.52047551e-01
5.32813609e-01 1.46395266e-01 3.65902275e-01 2.67506968e-02
5.75499237e-01 9.12751406e-02 4.04728293e-01 1.16911125e+00
-2.23752856e-01 -4.43731815e-01 -7.94398189e-01 4.03880656e-01
-1.80661416e+00 -9.34028089e-01 -2.30283841e-01 1.91241181e+00
4.57059413e-01 1.85187757e-01 -2.32920311e-02 5.19472897e-01
6.88872218e-01 1.08073570e-01 -3.13441485e-01 -9.50018167e-01
-6.34454340e-02 3.67994457e-01 1.50857881e-01 1.10986508e-01
-1.14407146e+00 8.34178507e-01 4.88187504e+00 4.42488134e-01
-1.45522213e+00 -1.94329724e-01 9.52756405e-01 3.08834612e-01
-1.22162960e-01 -2.65697300e-01 -9.28207695e-01 6.97676539e-01
1.01313365e+00 2.26650387e-02 -5.89126013e-02 1.08464265e+00
2.95363873e-01 -3.62514049e-01 -3.41208458e-01 7.51620591e-01
3.22863996e-01 -1.09191144e+00 2.35041557e-03 -2.96395123e-02
1.03373313e+00 -1.23618282e-01 -1.10147037e-01 4.07830298e-01
2.01041058e-01 -7.89002120e-01 1.57509595e-01 2.94817597e-01
1.16492212e-01 -1.15010047e+00 1.69799805e+00 4.16289836e-01
-8.52537453e-01 -1.50239065e-01 -5.10505974e-01 -3.80499035e-01
1.38059005e-01 9.70734119e-01 -5.82793415e-01 5.28035045e-01
8.99992168e-01 8.75665009e-01 -5.94303608e-01 7.44128942e-01
-1.90617278e-01 5.85703909e-01 -6.89146891e-02 -6.85756922e-01
1.30979821e-01 -4.94024575e-01 -9.82138142e-02 9.75543737e-01
4.76324022e-01 8.72600675e-02 -4.57666628e-02 1.49555504e-01
-2.07868367e-02 9.04809177e-01 -5.44391334e-01 -3.22264105e-01
-5.08512221e-02 1.70213485e+00 -1.08472538e+00 -3.15832257e-01
-5.74133813e-01 5.09971380e-01 1.04800738e-01 -1.87433630e-01
-7.19039261e-01 -6.12703204e-01 2.56347328e-01 -1.31669780e-02
1.33899420e-01 -3.58534120e-02 -5.25540948e-01 -1.12907720e+00
-1.42684489e-01 -1.10455751e+00 1.15533911e-01 -6.84281766e-01
-1.10876131e+00 9.77245092e-01 -3.97112906e-01 -1.25239027e+00
-1.88044026e-01 -8.11959803e-01 -6.25348210e-01 8.23844016e-01
-1.31856430e+00 -8.51405799e-01 -2.60891587e-01 3.21180791e-01
6.82428718e-01 -5.66978335e-01 8.45903397e-01 3.51471722e-01
-5.99925756e-01 2.98645020e-01 1.70628577e-01 1.34679258e-01
8.98258388e-01 -1.04645014e+00 8.01818632e-03 7.55030513e-01
-1.10367976e-01 5.76688409e-01 6.91665947e-01 -5.14419317e-01
-8.90845120e-01 -7.52536118e-01 1.27430880e+00 -1.22019630e-02
5.51285326e-01 3.74660194e-02 -6.78043187e-01 2.86836475e-01
5.01011312e-01 -4.03917074e-01 1.03929579e+00 2.58072376e-01
1.54866844e-01 -5.36036789e-01 -9.15339589e-01 4.44288850e-01
-3.51847447e-02 -5.04473001e-02 -5.70276022e-01 5.71504943e-02
2.35105589e-01 2.12736893e-02 -5.70236146e-01 5.12411833e-01
7.09160864e-01 -1.10346258e+00 1.85828879e-01 -3.53046536e-01
1.04806316e+00 -2.41204187e-01 2.94470023e-02 -1.79219246e+00
-4.85935695e-02 2.38446474e-01 4.51860815e-01 1.40945876e+00
7.95337915e-01 -7.64049411e-01 7.42955863e-01 6.33614779e-01
1.24266595e-02 -8.49818051e-01 -1.78323522e-01 1.64099246e-01
-8.10872838e-02 -4.51509327e-01 4.94416863e-01 9.06539977e-01
3.01799744e-01 7.18087554e-01 -3.76592219e-01 -3.78849149e-01
3.02146710e-02 1.54055446e-01 8.21390569e-01 -1.57706785e+00
1.60645127e-01 -3.22382659e-01 -3.76000851e-01 -4.48615730e-01
2.09744930e-01 -5.93870699e-01 -2.06659198e-01 -1.73673165e+00
8.70945826e-02 -3.47244203e-01 -4.26487058e-01 3.30181748e-01
-2.56481111e-01 2.79356480e-01 5.45917563e-02 2.95780189e-02
-4.28663462e-01 5.00639379e-01 1.11705673e+00 -4.95923273e-02
-1.69537574e-01 2.86154479e-01 -1.13028562e+00 1.02473605e+00
1.15916955e+00 -3.54276270e-01 -4.76144046e-01 -1.22211702e-01
9.82088208e-01 -1.92872018e-01 -4.96308595e-01 -7.54533052e-01
1.55036330e-01 3.45922746e-02 6.14954174e-01 -8.58501852e-01
1.80755928e-02 -9.36158061e-01 -1.01809643e-01 2.51137555e-01
-2.40056440e-01 3.50564271e-01 2.47404784e-01 2.34212443e-01
-7.69981861e-01 -3.91262770e-01 5.03090858e-01 -1.62774786e-01
-8.43841434e-01 -8.38095620e-02 -4.93510693e-01 -6.07946873e-01
1.07104421e+00 -6.24526851e-02 -2.45868251e-01 -4.28011924e-01
-4.48874146e-01 1.64959759e-01 6.16788641e-02 7.27978647e-01
4.38949883e-01 -7.95673370e-01 -6.51027620e-01 2.10199296e-01
1.29007190e-01 -1.19776368e-01 4.98200476e-01 6.61011159e-01
-5.24595678e-01 4.06179518e-01 -2.74024934e-01 -2.51913518e-01
-1.45549786e+00 2.05909193e-01 4.79611894e-03 -6.52885973e-01
2.35947557e-02 6.62510276e-01 -1.97843567e-01 -5.93646765e-01
-1.56528398e-01 -9.04180557e-02 -1.22153890e+00 5.65571368e-01
6.46331012e-01 3.21414113e-01 4.63511765e-01 -8.06821108e-01
-2.93614149e-01 5.23207664e-01 -4.43448663e-01 1.39691055e-01
1.41903710e+00 -6.55042306e-02 -4.58263129e-01 5.70235908e-01
8.62660587e-01 3.50150168e-01 -3.82045090e-01 2.66232520e-01
-3.30335125e-02 -1.07641332e-01 2.04213172e-01 -9.77011919e-01
-1.23540866e+00 9.81420636e-01 4.61440235e-01 4.97546077e-01
1.13188493e+00 -5.88973343e-01 7.94096649e-01 4.77002650e-01
3.43378603e-01 -1.30243516e+00 -2.53818989e-01 6.04860365e-01
6.22685373e-01 -1.74810684e+00 1.75321743e-01 -7.66850039e-02
-1.04785705e+00 1.19621336e+00 6.43988907e-01 -1.90317124e-01
9.98663962e-01 1.30537271e-01 4.13224906e-01 2.54011396e-02
-5.54622054e-01 1.57750636e-01 2.57174700e-01 1.69459537e-01
1.08917165e+00 1.27095617e-02 -9.23970401e-01 9.62735891e-01
-3.99960071e-01 1.36309043e-01 5.56153595e-01 8.12125802e-01
-5.70238650e-01 -1.40188122e+00 -3.89764041e-01 7.67899156e-01
-1.06545830e+00 2.02650148e-02 -4.66731042e-01 6.48003936e-01
9.36754197e-02 1.54646146e+00 -7.28934556e-02 -7.65583932e-01
1.68898135e-01 1.37992367e-01 -1.71724588e-01 -5.19347250e-01
-8.25476468e-01 -1.87365979e-01 1.59774184e-01 -4.61913198e-02
-9.62721825e-01 -2.94457495e-01 -1.21443164e+00 -4.31629539e-01
-8.43616962e-01 5.78946948e-01 1.18228793e+00 1.21866667e+00
4.36255217e-01 4.68809217e-01 1.09286833e+00 -5.68317235e-01
-3.33269656e-01 -1.34887099e+00 -4.65538502e-01 4.36253190e-01
-2.19850600e-01 -5.72142422e-01 -3.28693300e-01 3.33568938e-02] | [11.112783432006836, 6.903971195220947] |
f03a3bd2-807c-4e81-b699-1fe4a6a8628f | coach-a-coarse-to-fine-approach-for-cross | 2004.11727 | null | https://arxiv.org/abs/2004.11727v1 | https://arxiv.org/pdf/2004.11727v1.pdf | Coach: A Coarse-to-Fine Approach for Cross-domain Slot Filling | As an essential task in task-oriented dialog systems, slot filling requires extensive training data in a certain domain. However, such data are not always available. Hence, cross-domain slot filling has naturally arisen to cope with this data scarcity problem. In this paper, we propose a Coarse-to-fine approach (Coach) for cross-domain slot filling. Our model first learns the general pattern of slot entities by detecting whether the tokens are slot entities or not. It then predicts the specific types for the slot entities. In addition, we propose a template regularization approach to improve the adaptation robustness by regularizing the representation of utterances based on utterance templates. Experimental results show that our model significantly outperforms state-of-the-art approaches in slot filling. Furthermore, our model can also be applied to the cross-domain named entity recognition task, and it achieves better adaptation performance than other existing baselines. The code is available at https://github.com/zliucr/coach. | ['Pascale Fung', 'Zihan Liu', 'Peng Xu', 'Genta Indra Winata'] | 2020-04-24 | coach-a-coarse-to-fine-approach-for-cross-1 | https://aclanthology.org/2020.acl-main.3 | https://aclanthology.org/2020.acl-main.3.pdf | acl-2020-6 | ['cross-domain-named-entity-recognition'] | ['natural-language-processing'] | [-0.02446181 0.47969723 -0.45328107 -0.5970683 -0.84991425 -0.35838184
0.38560805 0.03354181 -0.46712124 0.96790195 0.34809935 -0.4277723
0.27856183 -0.75461984 -0.3415331 -0.22989385 0.44381 1.0093409
0.3766774 -0.42785665 -0.05487952 -0.15360434 -0.9975701 0.3158929
1.2895814 0.8791895 0.51067245 -0.01030416 -0.81695074 0.47858623
-0.5353538 -0.29112512 0.11853372 -0.31802282 -1.183529 0.36113614
-0.1440231 -0.13159089 -0.24086381 0.8830408 0.41108465 0.41145656
0.31508833 -1.0274351 -0.47243753 0.79000455 -0.33973843 -0.13898262
0.31132278 -0.29381743 1.1612434 -1.1300932 0.5831785 1.3263713
0.39615214 0.9728558 -1.1144239 -0.610685 0.41291392 -0.04980117
-1.3153718 -0.6074355 0.75275666 -0.19544527 0.8917066 0.02491261
0.08939142 0.8884229 -0.63858175 1.2052721 1.007846 -0.5900429
0.16895391 0.33362898 0.23370753 0.5242962 -0.11286885 -0.35346407
-0.43622664 -0.3310645 0.78198284 -0.07150584 -0.1272783 -0.3932408
-0.98158664 0.97809273 0.14885749 0.44436586 -0.31570643 -0.5568406
0.6413758 0.1724129 0.66649395 0.45975175 -0.92310846 -0.16979463
-0.69962096 0.31862366 1.0112783 1.1114396 0.8596547 -0.20301522
-0.17742468 1.5879008 0.39605635 0.20859338 0.6592243 -0.72952694
0.80970365 0.7973987 0.25668097 -0.31138346 -0.5001847 0.04694277
-0.6057497 -0.497233 0.47319824 -0.32685298 -0.8478004 1.8600024
0.6847665 0.10440256 0.35881212 0.7996976 1.0923504 0.5742358
0.51609445 -0.19110513 1.764913 -1.2131462 -1.0208364 -0.63909394
0.640138 -0.77061623 1.3068713 -0.20475325 -0.78342766 -0.49653915
-0.63325864 -0.18855491 -0.3656645 0.35333258 0.768085 0.5529055
-0.4983066 0.17029619 -0.77490056 -0.41923523 0.1243905 0.23984869
-0.3528119 -0.09312697 -1.6010224 0.7225806 0.7539295 -0.10255547
-0.23253107 -0.41270307 -1.2634249 -0.0619419 0.7039284 -0.3794467
1.7015257 -0.4764751 -1.5346259 0.8565279 -0.54892755 -0.43077442
0.25611913 -0.26117554 -0.40741116 -0.31577316 0.33480418 0.73950285
0.41971081 -1.1336814 -0.753799 -0.19060683 0.22467467 0.4834294
-0.260406 -0.08089764 -0.7686191 -0.43107045 0.39287892 -0.9066766
-0.37304592 -0.4680041 -0.56166035 -0.73272145 0.74506474 -0.6640462
1.4765794 -2.2546544 -0.25677767 -0.05848286 -0.14102042 0.5030998
-0.0463283 0.5854754 0.22916055 -0.04586035 -0.49815777 -0.63462216
0.31313425 0.52080876 -0.4800698 -0.07938927 0.22208121 0.7409504
-0.7060785 -0.66016567 0.14847043 0.17481165 -0.53332835 0.52071035
-0.55652404 0.56264347 -0.78230155 0.55345875 0.84424794 -0.47000387
0.68734103 0.03601235 0.02354274 0.94411063 -1.1725428 1.8652617
-0.5646282 0.08610589 0.0669231 -1.1382922 1.0845547 0.5400827
0.37882397 -0.76175314 -0.06393536 0.46082973 -0.21592493 -0.22005332
0.9052457 -0.1905303 -0.5438178 0.32400087 0.14205828 -0.05492483
0.334141 0.05937509 0.74031854 0.14191052 0.6709741 0.05193445
0.626487 0.08290933 1.1270945 0.49786162 -0.29959306 0.32831204
0.26211125 -0.20443581 -0.8616082 -0.8664463 -0.4027542 1.3800242
0.23052219 -0.50974435 -0.81337637 -0.8992621 0.06782147 0.5092748
-0.3211659 0.11541756 -0.69430685 -0.4063936 0.45556217 0.45697507
0.7749723 -1.4572933 -0.07120772 0.48321104 -0.66480285 -1.3977675
-0.5035059 0.27464333 -0.7548597 -0.89727527 -0.55310404 -1.076521
0.54215115 0.24603322 1.2459124 0.1953598 0.161952 -0.00742322
-0.6038132 -0.18810816 -0.31403384 0.43336442 0.07547427 -0.3487973
0.96753347 -0.29567114 -0.3318285 0.6530123 -0.9161789 0.06966278
0.41205764 1.2579814 0.622777 -0.05241159 0.84480655 -1.4691082
0.74295294 -0.5475682 -0.39493665 0.33605683 -0.4193617 0.14763373
0.38656142 -0.35143447 -1.5099876 0.23360254 -0.45130467 0.0917894
-0.50346744 0.62216276 -0.50470257 0.38792032 0.42966715 0.04317984
-0.36752123 -0.8526807 0.3459543 0.935991 0.2907335 -0.8895142
0.61278576 0.08054683 -0.8253421 -0.7703028 -1.09674 -0.873825
-0.6079145 0.40468812 0.7036598 -1.1027567 -0.25832707 0.22982658
-1.127997 -0.5774569 -0.2489052 0.28259945 -0.48184144 0.44295368
-0.76016873 -0.850407 -0.2938004 -0.94768804 0.98883575 0.5837258
-0.32065248 -1.0975913 0.12853684 0.65168846 0.29997337 -0.50557077
0.74441504 -1.2520913 -0.30270436 0.14652368 -0.2528429 -0.02075075
0.33033523 -0.6914877 -0.92111456 -0.03152676 -0.20828843 -0.5752596
0.7089678 0.12526298 0.9092428 -0.2613592 -0.48290345 0.12652363
0.8755478 0.2251258 0.52569526 0.43819702 0.34632036 0.7588674
1.2914227 0.6450499 0.8373435 0.989317 -0.04272639 -0.24711198
0.10625737 -0.59555644 -0.04764496 0.75056005 0.5615967 -0.1426072
-0.9490263 0.8334001 -2.165269 -0.55792505 0.23174125 1.9188917
1.283937 0.07595471 0.0798843 -0.14026403 0.9029586 0.27801508
-0.6591813 -0.09920287 0.04698535 0.260181 0.14609073 0.53120667
-1.4367921 1.6714234 5.2885113 0.95716166 -0.7772996 0.24299195
0.48550963 0.5948317 -0.18057269 0.10725594 -1.1768879 0.35068592
0.6343282 -0.06820192 0.16403984 1.1807292 -0.1372431 -0.03105314
-0.62576354 0.6642985 -0.3186255 -0.9350179 -0.3636664 -0.02394977
0.53968936 -0.1385513 -0.24458374 0.8602662 0.55211383 -0.631322
0.23676322 -0.05638622 0.6986505 -0.41763616 0.7448868 0.4923986
-1.3847786 0.17080534 -0.67055106 0.09403115 0.46636888 0.4918553
-1.2241501 0.5176925 0.38044235 0.28536144 -0.1692706 0.89406395
-0.3682481 0.5971351 -0.23788966 0.18535873 0.06798306 -0.10693432
0.18092436 1.2612239 0.1634304 0.30895528 0.68508846 0.5259079
-0.13704312 0.40344235 -0.32479763 -0.15452887 1.0146929 1.2092686
-0.4842713 -0.3542251 -0.57607365 1.0482495 0.6081972 0.29755726
-0.4370429 -0.37335333 0.94614446 -0.3475546 0.42314747 -0.16895758
-0.2787521 -1.4112146 0.15656403 -0.8972935 0.64019537 -0.25166002
-1.4118156 0.79546213 -0.16932815 -1.1991119 -0.51809734 -0.5721685
-0.4195761 0.80868155 -1.7190721 -1.0461725 -0.06638478 0.5350116
0.85057855 -0.19770369 1.2008425 0.39221865 -0.4991522 0.729298
-0.14662497 0.5783903 0.8885784 -1.1203114 0.8100171 0.5810332
0.17168818 0.7739095 0.5525413 -0.71833956 -0.70821893 -0.9183688
1.2889354 -0.07696281 0.5620347 -0.47991988 -1.3633527 0.7081628
0.1177428 -0.04350767 1.0030085 0.59956646 -0.43544784 0.3340822
-1.0308611 0.42470047 0.95595616 -0.66477627 -0.8633332 0.38124272
0.97692055 -0.87630945 -0.78723574 0.36472246 0.17519067 -0.5335302
0.840419 -0.69406646 -0.01363612 -0.25047794 -0.19281158 -1.332292
-0.01015595 -0.3963106 -0.08500712 1.5729918 0.81425655 -0.65522456
0.9838074 0.9547129 -0.23161423 -0.55658436 -1.1090037 -0.6267628
0.04707623 -0.21517764 0.8136523 1.1404244 0.47018266 0.5106718
-0.5622587 0.11229786 0.17015733 0.34498185 0.7996279 -1.2521358
-0.37344804 -0.04946565 0.1012499 -1.6216801 0.6175694 -0.5649805
0.32021093 -1.5101923 -0.01562283 -1.1561126 -0.2632931 0.79864746
-0.5236315 -0.15581182 0.23941043 0.23362206 -0.88422567 0.7074111
0.9496211 0.04665256 -0.53544956 0.22063638 -0.6998284 0.46177918
1.1220549 -0.3396367 -0.17165312 -0.27285406 -0.27491522 0.35356143
-0.29578227 -0.7248387 0.18778786 -0.27350482 -0.10799625 -0.5180001
0.6783154 -0.76410955 -0.20910767 0.06193384 -0.32936445 -0.30766216
0.30233145 0.3576193 -0.5156671 -0.40832046 0.59106386 -0.23243138
-0.9840341 0.3115944 -0.12779768 0.42778212 0.54165643 0.22381502
-0.42346677 -0.27968228 -0.6863984 0.5908567 0.4029595 0.6312387
0.36326924 -1.4182224 -0.34793523 0.2218691 0.44324276 0.13110049
0.374928 0.3980361 0.13926215 0.7367109 -0.03308022 -0.3276399
-1.1442053 0.29690295 0.16120212 -0.7340021 -0.38125616 0.8599118
0.41431472 -1.2728926 0.2544629 0.02965091 -0.42393345 -0.01657777
0.54294825 -0.33859348 -0.06068641 -0.6773012 -0.32799575 0.13185082
-0.45547175 -0.15186872 1.1786283 -0.39826947 0.0846466 0.4070582
0.6997471 0.04621805 -1.2799317 -1.0628064 0.5335165 -0.46401033
-0.47311068 -0.808385 -0.67874116 0.7150597 0.24469298 0.2722583
0.93696994 0.2298295 1.0626026 0.53259814 0.51514137 -1.4224573
-0.16559222 1.0314878 0.5509809 -1.6326149 -0.38658327 -0.81088203
-1.012913 0.75956076 1.0240765 0.27138424 0.47550052 -0.05092201
0.52611876 -0.01538807 -0.6420759 -0.55929536 0.18246327 0.6214945
0.8381699 0.2844214 -0.65144384 0.8005599 -0.16692293 -0.25918257
0.15730771 0.93177634 -0.6398978 -1.8670533 -0.17588829 0.21014656
-0.44505125 -0.29776028 -0.30573392 0.74107516 -0.13704535 1.0536567
-0.10877633 -0.1422049 0.49625182 0.56303805 -0.11144935 -1.0148704
-0.28481337 0.19429094 0.460547 -0.31653124 -0.35799533 -0.6493641
-1.5937396 0.02400681 -0.53994745 0.7621409 0.5196214 0.99044925
0.48245725 0.2693843 0.44658026 -0.51899743 -0.45524657 -1.1657823
-0.6476113 0.5207517 0.10086674 -1.0104411 0.03693974 -0.2016526 ] | [12.588020324707031, 7.406780242919922] |
6da4fdc6-c3a8-4b25-b1ad-61fed2d87473 | optimal-application-of-trajectory | 2303.00549 | null | https://arxiv.org/abs/2303.00549v1 | https://arxiv.org/pdf/2303.00549v1.pdf | Optimal Application of Trajectory Optimization by Travel Profile for Electric Train in an Electrical Transportation System | Increasing greenhouse gas (GHG) emission is one of the most important concerns of world decision-makers. The considerable part of it belongs to the transportation systems. As a solution, the world is going to use urban electrical transportation systems, and thus designing optimal methods and novel strategies is essential nowadays. Indeed, reducing the amount of consumed electrical energy and having optimum and efficient energy management for electrical means of transport requires several vital elements. Designing a travel profile is one of the most crucial parts of these methods. A travel profile for the fully electric train ET245 with three stations is designed and analyzed in Milan, Italy, as a case study in this paper. As the first step, route data are acquired by the Google Maps tool. Then, the train data is collected from the manufacturer's catalogues and datasheets. Finally, computations were performed by Excel spreadsheets. The results indicate that the designed path in the proposed travel profile is covered by a commercial speed and requires reasonable electrical energy, which is considered economical and advantageous. | ['Armin Mosavi'] | 2023-03-01 | null | null | null | null | ['energy-management'] | ['time-series'] | [-1.63739040e-01 -4.70764861e-02 -1.71342045e-01 -1.06286459e-01
-2.68078387e-01 -4.09166634e-01 4.01213914e-01 3.00057769e-01
-5.20253837e-01 9.35054064e-01 -3.28831732e-01 -4.86176193e-01
-7.24958837e-01 -1.42955792e+00 -2.60776103e-01 -8.68841290e-01
2.64015198e-01 4.22171533e-01 -1.10528560e-03 -2.71285743e-01
4.94078100e-01 9.21819568e-01 -1.49936819e+00 -6.75459802e-01
1.23813140e+00 9.57898080e-01 5.50110042e-01 1.40934765e-01
-4.08791870e-01 5.39115407e-02 -3.28683853e-01 -1.22519732e-01
-2.69004926e-02 -5.21288931e-01 -5.67888677e-01 5.27704172e-02
-8.30600083e-01 -9.57120359e-02 3.41376305e-01 8.47742200e-01
2.31424838e-01 5.03677726e-01 6.87598646e-01 -1.22964299e+00
3.41692001e-01 1.83955491e-01 -1.28537826e-02 -3.41618866e-01
-2.04913706e-01 6.79659471e-02 4.74823803e-01 -5.52372217e-01
1.09618239e-01 6.26602769e-01 -4.05556196e-03 -1.03691459e-01
-8.50862622e-01 -4.20806885e-01 -4.36080426e-01 4.75157201e-01
-1.51984549e+00 -2.81932931e-02 8.58739853e-01 -3.73894721e-01
8.10441375e-01 5.92964172e-01 1.12202764e+00 2.67924219e-01
3.50487262e-01 7.61537477e-02 1.06480646e+00 -5.22901297e-01
4.90216881e-01 5.91795087e-01 -1.68204471e-01 7.99223930e-02
4.80894238e-01 -1.54339388e-01 3.02866429e-01 4.38695878e-01
2.45457664e-01 -1.12730317e-01 -1.43808573e-01 -9.18890908e-02
-5.51535428e-01 5.59811175e-01 4.11136329e-01 8.30332398e-01
-5.89444697e-01 -4.61854376e-02 8.57827626e-03 -1.58462375e-01
2.37526745e-01 3.28119636e-01 -1.05457030e-01 -5.71891665e-01
-6.97295785e-01 -1.04730442e-01 5.38887262e-01 7.23821938e-01
7.37178564e-01 -8.57791975e-02 5.34833610e-01 5.33248544e-01
3.46866608e-01 6.53939366e-01 -9.86505579e-03 -6.05891109e-01
4.64363456e-01 8.18225145e-01 2.82488197e-01 -1.02090287e+00
-6.31610513e-01 -4.90575045e-01 -7.40513742e-01 -3.96238901e-02
4.14823025e-01 -2.27258712e-01 -3.72707665e-01 1.11549759e+00
3.36483151e-01 -5.46353698e-01 -1.97695568e-01 5.62277138e-01
3.63323808e-01 9.86852944e-01 2.46524796e-01 -3.54142308e-01
1.24465859e+00 -4.61655557e-01 -1.02909136e+00 6.92164451e-02
7.48873353e-01 -4.34737325e-01 8.64336729e-01 4.24708575e-01
-9.74581361e-01 -2.74766445e-01 -8.38359237e-01 3.57036144e-01
-8.38637888e-01 3.16427439e-01 3.52260262e-01 8.62547815e-01
-6.48594916e-01 5.66169620e-01 -5.47912300e-01 -5.89580178e-01
1.21191517e-01 6.04821704e-02 -6.22064900e-03 -3.57969068e-02
-1.30876660e+00 1.14190125e+00 3.91565114e-01 4.78771955e-01
-2.38426551e-01 -4.22208011e-01 -4.29783791e-01 2.36605749e-01
2.24822506e-01 -1.38172239e-01 7.68662214e-01 -9.36199501e-02
-1.40949035e+00 1.22713953e-01 7.11960271e-02 1.06137201e-01
6.62556231e-01 1.73095241e-01 -7.32620180e-01 7.92191476e-02
-4.88274731e-02 -9.08373594e-02 1.46559596e-01 -1.02682209e+00
-8.99651170e-01 -1.94429249e-01 -9.75602046e-02 -3.73647362e-02
-4.28348482e-01 -1.66276902e-01 -2.21168831e-01 6.20374605e-02
-1.73223883e-01 -1.11678565e+00 -2.60078639e-01 -5.02593577e-01
-4.33411658e-01 -4.52718794e-01 4.97099131e-01 -7.58951485e-01
1.46998847e+00 -1.94019818e+00 -4.44393419e-02 1.04382145e+00
-4.78882253e-01 -7.51237106e-03 4.95194376e-01 6.92377985e-01
1.06751442e-01 9.79547054e-02 -3.57762277e-01 8.08608010e-02
1.33496627e-01 1.35569304e-01 3.98209035e-01 4.44631726e-01
-1.14378415e-01 3.63114983e-01 -6.96428239e-01 -3.62663329e-01
6.80766940e-01 5.72877467e-01 -3.84798273e-02 -1.17296338e-01
1.23817911e-02 5.26608527e-01 -7.99331665e-01 4.20291007e-01
6.02129817e-01 3.75456452e-01 2.44744271e-01 -1.90128908e-01
-9.65977550e-01 2.14060947e-01 -1.10314226e+00 1.38138664e+00
-8.60825598e-01 4.51287568e-01 -9.77460817e-02 -9.21322405e-01
1.22384334e+00 3.15444946e-01 8.00576925e-01 -1.29732060e+00
4.15418684e-01 6.92388833e-01 -1.51405185e-01 -6.72676384e-01
6.70156121e-01 6.92422464e-02 -6.28420413e-02 2.89015770e-01
-5.93593299e-01 -5.11551380e-01 4.91717845e-01 -1.81121454e-01
4.72975671e-01 4.55327556e-02 9.16123018e-02 -5.09922624e-01
7.43889749e-01 2.90571153e-01 3.47631305e-01 -3.26548159e-01
3.11697006e-01 -2.42161572e-01 3.75967681e-01 -5.66315986e-02
-9.59480405e-01 -4.72180486e-01 -2.87507355e-01 1.76742405e-01
3.60172719e-01 4.36568297e-02 -9.33914185e-01 -8.04272890e-02
-3.48127037e-01 1.43067753e+00 -3.24499868e-02 -3.46766561e-02
-4.13745075e-01 -5.73325753e-01 -9.98676196e-02 1.92818090e-01
7.74071991e-01 -8.10575783e-01 -1.07754540e+00 3.34248662e-01
-4.45155293e-01 -7.26355791e-01 9.63127539e-02 1.95977435e-01
-7.77585745e-01 -1.01566339e+00 -5.69816768e-01 -2.66336173e-01
7.15340078e-01 2.54767060e-01 8.19855571e-01 1.91860065e-01
-2.00572461e-01 -1.37043089e-01 -2.94104040e-01 -5.60567260e-01
-3.38580072e-01 4.96864676e-01 -2.61673272e-01 -1.82948589e-01
3.50440830e-01 -3.45508575e-01 -6.29124284e-01 7.00678766e-01
-6.24432206e-01 5.00495508e-02 4.51302528e-01 -1.14340233e-02
6.19647741e-01 9.91199195e-01 8.54463875e-01 -5.28310418e-01
4.76340413e-01 -6.27075911e-01 -1.02814829e+00 1.23313800e-01
-8.66667032e-01 -2.21502483e-01 5.77501714e-01 6.11105323e-01
-1.27475977e+00 -3.86723317e-02 -4.13523883e-01 4.95835602e-01
-3.81237328e-01 3.89927238e-01 -8.48386049e-01 5.20266704e-02
4.08641174e-02 6.49386495e-02 -1.17941059e-01 -5.71510077e-01
6.64547877e-03 5.88400245e-01 5.64208813e-02 -2.97665715e-01
9.50438142e-01 1.79083720e-01 5.40184081e-01 -1.11401737e+00
2.03730091e-01 -2.85040557e-01 -5.19625962e-01 -9.52787459e-01
9.49097514e-01 -2.18844146e-01 -9.92904603e-01 2.74084717e-01
-7.00404525e-01 -1.49899736e-01 -1.50520727e-01 7.54100204e-01
-4.48361546e-01 -1.00268565e-01 1.62702158e-01 -1.02481472e+00
-8.36898834e-02 -1.30575347e+00 1.02916852e-01 4.44066644e-01
-2.28419155e-01 -9.70601976e-01 -1.27500623e-01 1.36957914e-01
6.86550200e-01 4.41780299e-01 1.03677952e+00 -2.47215126e-02
-8.09632182e-01 -3.45285714e-01 -8.74305610e-03 3.19348186e-01
4.69669849e-01 2.56839097e-01 -6.23560965e-01 -8.71367976e-02
-7.47210756e-02 4.66179013e-01 2.47925758e-01 4.20006424e-01
1.09011960e+00 -1.29588813e-01 -4.27496254e-01 1.94026291e-01
1.92441893e+00 9.51816559e-01 1.03471696e+00 6.31594718e-01
2.32269764e-01 1.05952406e+00 1.19343424e+00 4.57348853e-01
4.01188314e-01 5.84974706e-01 7.76575506e-01 -1.80979460e-01
4.54991341e-01 -1.00270145e-01 -6.04161024e-02 9.20803964e-01
-5.52129209e-01 -4.47671235e-01 -8.27496588e-01 7.67013788e-01
-1.31113231e+00 -9.38407779e-01 -6.75878227e-01 2.37952948e+00
1.46451697e-01 5.09472899e-02 2.02953115e-01 7.71443307e-01
6.21535003e-01 -3.69139999e-01 -2.41534691e-02 -6.49240375e-01
2.80588806e-01 -3.77497710e-02 7.73219347e-01 2.46008977e-01
-3.62282097e-01 -1.00813070e-02 4.34386683e+00 8.44872534e-01
-7.44324923e-01 -7.80087635e-02 2.90246338e-01 1.11939562e-02
-6.71543956e-01 6.34551495e-02 -3.86162102e-01 7.86870718e-01
1.23685980e+00 -4.36831623e-01 3.44449401e-01 4.73519295e-01
9.32847142e-01 -8.27997088e-01 -5.38729310e-01 5.53801298e-01
-6.33888543e-01 -1.13358259e+00 -5.80099106e-01 5.18808782e-01
4.02204514e-01 -4.90173936e-01 -2.61124581e-01 -1.63869008e-01
-2.87596881e-01 -6.30204916e-01 7.67332792e-01 7.72182882e-01
5.90192437e-01 -1.61498022e+00 8.92671466e-01 3.33978802e-01
-1.40106082e+00 1.61193078e-03 -4.66364585e-02 3.22087079e-01
7.70359397e-01 9.09305274e-01 -5.80045819e-01 9.67074752e-01
7.35378802e-01 1.90420657e-01 -1.27392620e-01 1.17645204e+00
-3.17641050e-01 5.05416870e-01 -5.43825150e-01 -5.51474631e-01
1.90334946e-01 -1.05162323e+00 2.25338936e-01 7.81456709e-01
8.99567544e-01 -9.84414946e-03 -5.39930224e-01 7.63030112e-01
2.78640538e-01 3.80008906e-01 -6.44176483e-01 -1.25533402e-01
6.70408130e-01 1.25203061e+00 -8.90488982e-01 3.60896066e-02
-2.87293315e-01 2.07789645e-01 -7.08721101e-01 2.89520353e-01
-1.16485405e+00 -1.01976025e+00 3.42353463e-01 4.90865707e-01
-2.09460095e-01 -2.86589116e-01 -2.12565646e-01 -6.57361895e-02
-3.12936455e-02 1.38824051e-02 -1.10731654e-01 -5.56941152e-01
-4.43659425e-01 2.67348319e-01 2.23393545e-01 -1.30387664e+00
-1.98581666e-01 -3.52214009e-01 -5.11228621e-01 1.05253243e+00
-1.59200728e+00 -5.58679879e-01 -6.14019573e-01 3.93635452e-01
4.63241220e-01 3.16686183e-01 6.55249238e-01 9.14222836e-01
-7.82785475e-01 1.12351426e-03 5.78877449e-01 -4.08924848e-01
-4.80071409e-03 -8.31119776e-01 5.18223979e-02 6.14170313e-01
-6.83747649e-01 1.83509022e-01 7.11387336e-01 -4.87663925e-01
-1.39020681e+00 -8.66623521e-01 1.23664975e+00 3.75596046e-01
4.91804421e-01 -3.64474729e-02 -7.06982672e-01 2.28395462e-01
3.10864598e-01 -8.34477663e-01 5.13429105e-01 -4.03713018e-01
1.05137861e+00 -2.26563931e-01 -1.32269681e+00 4.91527766e-01
6.07353568e-01 -1.65411249e-01 1.90583631e-01 2.49914959e-01
1.63560003e-01 5.92587553e-02 -1.04372346e+00 -4.64713946e-02
5.15284121e-01 -8.32589805e-01 3.77022117e-01 1.82685122e-01
1.87732093e-02 -4.04702336e-01 1.14427373e-01 -1.57387924e+00
2.23620366e-02 -3.58053952e-01 6.23026013e-01 1.62756729e+00
5.83153546e-01 -9.10191894e-01 4.59366024e-01 9.30354714e-01
-4.27023083e-01 -5.65852106e-01 -9.16136980e-01 -7.48401940e-01
-3.14428777e-01 -6.79918408e-01 1.07216048e+00 6.17554307e-01
-2.56758891e-02 -2.54183382e-01 6.58304766e-02 -3.03207543e-02
7.90273786e-01 -1.33928791e-01 6.11600280e-01 -1.23397517e+00
4.58853930e-01 -3.23484778e-01 -5.40789999e-02 -3.18042397e-01
-1.95363760e-01 -6.01653457e-01 -1.49924438e-02 -2.34032798e+00
-4.08408999e-01 -7.55683005e-01 -1.29338645e-03 9.71614867e-02
7.12362289e-01 -2.49232799e-01 5.75857013e-02 -7.04997331e-02
4.03712004e-01 7.40116239e-01 1.11345971e+00 -1.30612165e-01
-1.20893128e-01 4.54498142e-01 -2.12707147e-01 5.89773536e-01
1.10917044e+00 -5.26139617e-01 -5.46024680e-01 -1.63200065e-01
6.02450967e-01 -2.00922620e-02 1.18149467e-01 -9.34843659e-01
2.04393879e-01 -4.46115553e-01 -2.08303601e-01 -1.07627177e+00
1.82899222e-01 -1.78418100e+00 9.49829698e-01 6.46435380e-01
2.81086028e-01 1.22338742e-01 5.22211343e-02 1.75843865e-01
-1.14910521e-01 -3.61692160e-01 6.26333773e-01 2.08943576e-01
-7.00058103e-01 -1.03922233e-01 -7.13536501e-01 -6.87308252e-01
1.50750029e+00 -6.32701993e-01 -3.69209081e-01 -1.44542068e-01
-4.14279729e-01 4.26642865e-01 3.67412895e-01 9.49918702e-02
2.23014578e-01 -1.23926926e+00 -2.40934581e-01 5.50290849e-03
-1.19783752e-01 -7.36125633e-02 7.50913203e-01 1.16780710e+00
-7.51663089e-01 7.74250925e-01 -3.76970410e-01 -2.90587306e-01
-8.32523882e-01 2.53817081e-01 2.40336925e-01 4.38549183e-02
-3.36238652e-01 1.26025423e-01 -4.45059925e-01 1.46462973e-02
-2.49750391e-01 -4.26702470e-01 -5.28501272e-01 4.91096556e-01
1.18001111e-01 1.20338917e+00 4.74794328e-01 -6.04052424e-01
-5.24608314e-01 8.84587526e-01 8.63932669e-01 -7.10873008e-02
1.32913089e+00 -3.95023286e-01 -1.34991556e-01 3.51225555e-01
9.46023226e-01 8.76419172e-02 -7.25776613e-01 4.25582260e-01
7.49720186e-02 -5.85588992e-01 8.63648579e-02 -6.31684542e-01
-1.08868003e+00 6.86745703e-01 2.94606477e-01 6.21536136e-01
1.52106881e+00 -2.93987781e-01 6.50621712e-01 9.95978564e-02
7.47867703e-01 -1.59847403e+00 -6.30477011e-01 3.33510600e-02
5.65415263e-01 -5.93309939e-01 -8.29173326e-02 -4.47093755e-01
-3.19828033e-01 1.04423857e+00 1.36725396e-01 3.42813194e-01
6.77520633e-01 7.18711689e-02 -3.88251156e-01 -1.42164409e-01
2.27161460e-02 -1.94590971e-01 -2.17059046e-01 2.32782766e-01
2.61735588e-01 2.33503357e-01 -9.19108629e-01 5.23327947e-01
-1.29427701e-01 1.30691200e-01 4.30430681e-01 8.72575104e-01
-5.95603526e-01 -1.00812340e+00 -5.29244840e-01 3.57826412e-01
-1.15747884e-01 4.50680733e-01 1.50307834e-01 1.29099476e+00
2.31766149e-01 1.29431546e+00 1.29189581e-01 -3.12060952e-01
9.09221232e-01 -1.97053283e-01 1.17240056e-01 1.09285355e-01
-4.27791327e-01 -2.45093346e-01 3.96477431e-01 -3.41472566e-01
-4.43981022e-01 -4.67368484e-01 -1.70541024e+00 -7.26599157e-01
-5.09649813e-01 8.18014562e-01 1.67124677e+00 9.98413503e-01
2.09684507e-03 9.53807175e-01 1.07520700e+00 -5.67462564e-01
1.41310230e-01 -7.00373352e-01 -1.01558352e+00 -1.07967891e-01
-2.77407140e-01 -7.74037004e-01 -4.60481793e-01 -5.39381623e-01] | [5.594106674194336, 2.1510982513427734] |
707e1179-884c-4dba-8a6f-482a4c79e372 | long-history-short-term-memory-for-long-term | null | null | https://openreview.net/forum?id=HklmoRVYvr | https://openreview.net/pdf?id=HklmoRVYvr | Long History Short-Term Memory for Long-Term Video Prediction | While video prediction approaches have advanced considerably in recent years, learning to predict long-term future is challenging — ambiguous future or error propagation over time yield blurry predictions. To address this challenge, existing algorithms rely on extra supervision (e.g., action or object pose), motion flow learning, or adversarial training. In this paper, we propose a new recurrent unit, Long History Short-Term Memory (LH-STM). LH-STM incorporates long history states into a recurrent unit to learn longer range dependencies. To capture spatio-temporal dynamics in videos, we combined LH-STM with the Context-aware Video Prediction model (ContextVP). Our experiments on the KTH human actions and BAIR robot pushing datasets demonstrate that our approach produces not only sharper near-future predictions, but also farther into the future compared to the state-of-the-art methods. | ['Jan Kautz', 'Wonmin Byeon'] | 2019-09-25 | null | null | null | null | ['video-prediction'] | ['computer-vision'] | [ 2.71845579e-01 -9.74611193e-03 -5.71991682e-01 -4.81267542e-01
-3.65402699e-01 -2.23518476e-01 8.23948085e-01 -4.59995389e-01
-2.06889227e-01 8.60841215e-01 6.26652062e-01 -1.93458214e-01
1.70402870e-01 -5.73512316e-01 -1.04725063e+00 -3.28142136e-01
-3.64811659e-01 1.17343210e-01 6.91281438e-01 -6.57113045e-02
3.42827469e-01 2.80042887e-01 -1.31847441e+00 6.71936691e-01
4.74044919e-01 8.96164536e-01 4.87318277e-01 9.20382559e-01
1.37811020e-01 1.57793081e+00 -1.65471941e-01 -2.41191775e-01
2.36896664e-01 -4.46414888e-01 -9.51538861e-01 5.24778478e-02
3.72741371e-01 -7.78389215e-01 -9.43632841e-01 6.48176670e-01
-1.28827719e-02 3.83316219e-01 3.46022695e-01 -1.24226546e+00
-7.59753883e-01 4.11785185e-01 -3.91394079e-01 3.57942134e-01
5.24575174e-01 5.88455021e-01 8.05173397e-01 -7.72750020e-01
1.12370670e+00 1.30552769e+00 7.72148788e-01 8.32592249e-01
-9.84651625e-01 -5.87003529e-01 7.58671761e-01 6.63072526e-01
-9.40254569e-01 -3.04985791e-01 8.02014053e-01 -4.48002398e-01
1.09944201e+00 -6.60693124e-02 7.03729570e-01 1.73759508e+00
6.63724184e-01 1.16499376e+00 6.68450177e-01 1.68016642e-01
7.27930740e-02 -4.86273319e-01 -2.15010956e-01 8.17431808e-01
-4.07232523e-01 4.92262930e-01 -6.54238641e-01 8.90066549e-02
9.36681509e-01 3.83907884e-01 -3.68424952e-01 -4.22503173e-01
-1.37296569e+00 5.09284437e-01 4.57908154e-01 5.61650395e-02
-4.91476566e-01 4.55131829e-01 4.52945262e-01 2.07481965e-01
4.51171786e-01 2.37629220e-01 -4.58711773e-01 -4.33581740e-01
-9.41014707e-01 3.52851003e-01 7.21186817e-01 9.16103005e-01
4.80288744e-01 1.10708117e-01 -3.61612409e-01 2.08015993e-01
1.53912887e-01 3.07615072e-01 5.95585167e-01 -1.31932950e+00
8.38410616e-01 1.23951353e-01 4.93366510e-01 -1.06198013e+00
-3.25129211e-01 -1.54207334e-01 -6.71756685e-01 1.57642901e-01
3.88727188e-01 -1.84687823e-01 -1.21046162e+00 1.82269776e+00
6.80545270e-02 9.82304752e-01 1.31959677e-01 1.13843763e+00
3.98322761e-01 1.11506045e+00 3.39431405e-01 -1.98841453e-01
5.21066844e-01 -1.49198496e+00 -6.39071763e-01 -5.17380774e-01
5.75670421e-01 -3.61128300e-01 7.52543509e-01 1.80893540e-01
-1.04675758e+00 -8.44305396e-01 -9.26367402e-01 -5.56990653e-02
-1.48589283e-01 -1.26773596e-01 4.59149361e-01 -1.29678518e-01
-9.26994979e-01 9.94514763e-01 -1.39339340e+00 -2.18689695e-01
3.67029637e-01 5.55079654e-02 -5.07905960e-01 -2.32056111e-01
-1.10469556e+00 9.25373793e-01 2.28223071e-01 8.60534832e-02
-1.18507564e+00 -6.26113832e-01 -1.14455223e+00 -2.95848353e-03
5.94870985e-01 -6.91661358e-01 1.35036194e+00 -1.01315892e+00
-1.58203959e+00 2.89179534e-01 -3.54033440e-01 -9.50296760e-01
8.18374574e-01 -7.27415144e-01 -3.19503307e-01 2.42707148e-01
-6.91574588e-02 1.03633881e+00 1.02077603e+00 -9.92861450e-01
-7.00047851e-01 -1.64291710e-01 1.68824688e-01 2.21116692e-01
-1.24443714e-02 -2.85319626e-01 -3.60912561e-01 -8.75328779e-01
-1.03976354e-01 -1.08366942e+00 -4.43032980e-01 1.18129648e-01
-9.00193229e-02 -1.20492265e-01 1.35438168e+00 -8.78898680e-01
1.15918517e+00 -1.93631959e+00 4.15310651e-01 -4.79039103e-01
-3.67701203e-02 3.97086233e-01 -3.77147317e-01 2.89059192e-01
-8.27801321e-03 -7.25854486e-02 -1.03994466e-01 -4.67202753e-01
-9.22588781e-02 5.25143802e-01 -8.46027076e-01 3.32779795e-01
4.23605978e-01 1.24111056e+00 -1.20670283e+00 -4.83351052e-01
4.35588390e-01 5.79453766e-01 -6.18085146e-01 3.34587902e-01
-5.98969996e-01 8.11281502e-01 -4.78527337e-01 5.47399938e-01
2.89001942e-01 -4.07856554e-01 6.33041635e-02 1.33174345e-01
-1.55855909e-01 1.56561971e-01 -6.13649249e-01 1.99856722e+00
-3.14112276e-01 6.55488431e-01 -3.23847353e-01 -1.01997089e+00
7.21222341e-01 4.02101487e-01 6.09996319e-01 -5.73311806e-01
-2.81665564e-01 -2.31855959e-02 -3.27001184e-01 -7.90100276e-01
7.80577779e-01 -7.79285803e-02 9.37446952e-02 2.71090299e-01
-7.27725495e-03 3.45268071e-01 -2.27104761e-02 1.77957669e-01
1.26776671e+00 1.00594747e+00 6.24229424e-02 3.03927243e-01
4.46574718e-01 6.00284263e-02 1.02653778e+00 7.53004193e-01
-6.26093209e-01 8.33019555e-01 3.96919042e-01 -8.21420670e-01
-1.16140723e+00 -8.65638673e-01 4.33284432e-01 1.04865384e+00
4.57154244e-01 -4.25482929e-01 -3.01439583e-01 -9.48677123e-01
-1.18271910e-01 7.17361629e-01 -6.21445119e-01 -3.27135772e-01
-1.26472366e+00 -1.34231463e-01 2.08423674e-01 1.02462840e+00
5.63544869e-01 -1.39490044e+00 -9.41097856e-01 5.30075133e-01
-2.38398284e-01 -1.47179592e+00 -5.78383982e-01 -3.42793971e-01
-1.07425511e+00 -8.19429278e-01 -9.43393290e-01 -6.23454332e-01
2.23514199e-01 5.87159097e-02 1.03316391e+00 -2.80051291e-01
-8.53406787e-02 4.50193703e-01 -4.95241076e-01 1.14880227e-01
-3.62707824e-01 -2.05610216e-01 -8.90772417e-02 -1.16704091e-01
7.70084038e-02 -6.63253784e-01 -8.12857151e-01 3.72465283e-01
-6.79688454e-01 3.47171217e-01 5.86414337e-01 1.06362045e+00
6.15191400e-01 -3.59180599e-01 5.83387733e-01 -6.45441175e-01
1.03235692e-01 -5.59720516e-01 -4.08853233e-01 3.76570135e-01
-4.36792761e-01 5.90446889e-02 9.28040802e-01 -5.92958331e-01
-1.49264157e+00 3.30673099e-01 -3.64460796e-02 -9.85671639e-01
-1.72069311e-01 3.53339642e-01 2.02926770e-01 2.60077208e-01
2.66568810e-01 3.42191905e-01 -1.19171284e-01 -3.07392001e-01
3.13436508e-01 8.39642063e-02 8.88127148e-01 -1.93915769e-01
4.15362686e-01 5.94726205e-01 -3.10375802e-02 -3.55155855e-01
-9.06655788e-01 -1.92156702e-01 -7.38058090e-01 -3.75027955e-01
1.07078207e+00 -9.12837207e-01 -3.92331511e-01 4.25256222e-01
-1.27470696e+00 -7.54055262e-01 -2.64984012e-01 5.40674210e-01
-8.75846565e-01 3.25184286e-01 -1.02110946e+00 -7.38300920e-01
-2.13966012e-01 -1.08228433e+00 9.46978331e-01 1.59323663e-01
-2.07667902e-01 -9.99032378e-01 9.63331684e-02 2.69114077e-01
2.92391568e-01 3.92942190e-01 3.86309624e-01 -1.99583292e-01
-1.01033366e+00 -2.12556079e-01 -4.55278754e-02 1.61198616e-01
-1.59434974e-01 -1.34737387e-01 -5.90026319e-01 -2.47430682e-01
-1.48925737e-01 -3.47267270e-01 1.25488853e+00 4.74032104e-01
1.33970988e+00 -3.62182140e-01 -5.88889003e-01 6.86244369e-01
9.88221407e-01 2.44312048e-01 7.88985610e-01 2.33075812e-01
8.31279993e-01 4.44516480e-01 1.20112205e+00 4.93426204e-01
5.04175901e-01 6.17576301e-01 5.22322953e-01 3.38917106e-01
-2.72824377e-01 -7.80435801e-01 7.44059801e-01 3.65462661e-01
-2.91078120e-01 -3.41110617e-01 -8.33312333e-01 5.21566272e-01
-2.40474772e+00 -1.42703211e+00 2.72029251e-01 1.82439721e+00
3.50418597e-01 3.52456748e-01 -7.97763690e-02 -2.56869018e-01
6.45098150e-01 6.00567460e-01 -8.97111714e-01 -1.55583724e-01
1.65277928e-01 -2.23257273e-01 3.01444769e-01 5.15883148e-01
-1.28944147e+00 1.16859090e+00 6.40159845e+00 3.93350095e-01
-1.33409643e+00 3.57484780e-02 6.18873954e-01 -1.17483616e-01
-5.11009060e-02 1.21291026e-01 -6.83554351e-01 3.80286276e-01
7.95829892e-01 1.33061036e-01 1.31663814e-01 9.22529876e-01
2.50264436e-01 4.61184643e-02 -1.15711832e+00 7.65917540e-01
1.70648560e-01 -1.48229814e+00 5.03089838e-02 -1.41295880e-01
7.85360992e-01 1.23231217e-01 1.23990983e-01 6.06576502e-01
1.92032933e-01 -9.55748320e-01 8.06967616e-01 8.83410752e-01
5.21454811e-01 -5.82120538e-01 5.86145878e-01 6.25189126e-01
-1.36374760e+00 -4.55270529e-01 -2.91627109e-01 -3.70851368e-01
7.79902935e-01 1.79447401e-02 -5.70979714e-01 4.04945195e-01
6.76087797e-01 1.44758594e+00 -4.19457704e-01 7.50695169e-01
-2.51607805e-01 5.71277797e-01 1.11667570e-02 2.42979124e-01
5.20836413e-01 -5.37815727e-02 8.52595389e-01 1.08754301e+00
3.23877186e-01 2.63715386e-01 5.17906785e-01 6.86883271e-01
1.88638046e-02 -5.87154567e-01 -7.29465425e-01 -4.77148183e-02
9.06062350e-02 7.83825099e-01 -3.60889554e-01 -5.67210078e-01
-5.94009817e-01 1.32113183e+00 3.98209035e-01 6.29513562e-01
-1.19244623e+00 3.69246192e-02 6.65647209e-01 2.63555143e-02
6.76712990e-01 -3.74227762e-01 3.67039964e-02 -1.44821835e+00
1.83551937e-01 -4.03314203e-01 3.97741169e-01 -8.99780095e-01
-1.11742854e+00 5.13499260e-01 -1.20269969e-01 -1.40978003e+00
-7.39301622e-01 -4.45130914e-01 -6.75059140e-01 4.75882679e-01
-1.55769646e+00 -1.22689462e+00 -1.39963061e-01 4.31683242e-01
1.02424991e+00 1.66667923e-02 6.50241196e-01 6.73746616e-02
-4.18860227e-01 1.95400387e-01 -2.01918364e-01 1.61642730e-01
5.73396325e-01 -8.84746730e-01 7.04830766e-01 9.48259234e-01
1.12558611e-01 2.24860802e-01 7.78667629e-01 -9.51651096e-01
-1.26552618e+00 -1.37888753e+00 6.86897397e-01 -4.14387286e-01
8.58379185e-01 -6.75466284e-02 -1.00739634e+00 1.23385143e+00
6.34322166e-02 6.17616713e-01 1.04166530e-01 -3.24667394e-01
-3.28385711e-01 1.21741697e-01 -7.93735802e-01 6.67721570e-01
1.41657162e+00 -5.34413278e-01 -7.27206886e-01 -6.14275364e-03
9.24312472e-01 -6.51365936e-01 -6.82926297e-01 7.57089257e-01
7.69526899e-01 -1.01451421e+00 1.07227826e+00 -8.50155771e-01
8.84062171e-01 -1.05665304e-01 4.38274555e-02 -1.05053103e+00
-4.19423014e-01 -5.53334236e-01 -8.47109079e-01 7.04554021e-01
3.10043454e-01 -1.97857410e-01 1.08381522e+00 7.87087739e-01
-3.26264352e-01 -1.02270257e+00 -7.84359455e-01 -8.37626517e-01
2.44889650e-02 -4.59564835e-01 3.03500712e-01 7.65714407e-01
7.42404312e-02 1.92234486e-01 -1.09076095e+00 9.06752273e-02
2.75150478e-01 4.80062485e-01 7.43172407e-01 -6.58348143e-01
-3.85768175e-01 -8.88962299e-02 -6.69778407e-01 -1.72974277e+00
5.20149887e-01 -4.31720853e-01 2.11645156e-01 -1.43721867e+00
2.67906904e-01 -9.50403512e-02 -2.69398898e-01 3.85022730e-01
-3.12137812e-01 2.18322258e-02 4.93787616e-01 2.07146347e-01
-1.09915352e+00 8.69619250e-01 1.60001910e+00 -1.21106364e-01
-2.57599384e-01 7.70573225e-03 1.33342475e-01 9.04914200e-01
6.32877946e-01 -3.04660082e-01 -4.49231058e-01 -5.71430147e-01
-2.74661869e-01 6.00346386e-01 6.87841058e-01 -1.24181402e+00
3.65031749e-01 -2.66299099e-01 5.50069273e-01 -7.95992196e-01
7.93315530e-01 -5.67904770e-01 3.04979738e-02 6.76567972e-01
-5.90914488e-01 1.64733946e-01 -6.14019632e-02 1.16334581e+00
-3.91895741e-01 1.84862301e-01 3.93326908e-01 -2.59800524e-01
-1.35827553e+00 6.34377778e-01 -3.82327378e-01 -1.10126153e-01
1.16689897e+00 -1.15049757e-01 -3.33596736e-01 -6.19572878e-01
-9.99955475e-01 4.44351703e-01 3.62004876e-01 9.15446162e-01
9.31123018e-01 -1.18037319e+00 -5.13666391e-01 -9.70074981e-02
-6.78451881e-02 -9.05143395e-02 4.52772677e-01 7.28432775e-01
-3.50957721e-01 5.39472282e-01 -2.44343445e-01 -6.69256091e-01
-1.09502804e+00 7.51543760e-01 1.52665347e-01 -4.69474316e-01
-1.10416245e+00 8.68531466e-01 2.77087241e-01 -2.86144555e-01
3.49926472e-01 -2.65419096e-01 -2.54206359e-01 -4.81820166e-01
5.77989221e-01 3.50799888e-01 -7.56412625e-01 -7.30939984e-01
-6.26525655e-02 3.48906249e-01 -1.30011246e-01 -6.42156973e-02
1.43238330e+00 -3.34754974e-01 3.65322381e-01 6.44482076e-01
1.09591568e+00 -6.44415855e-01 -2.19111824e+00 -1.79239199e-01
1.82551056e-01 -6.25452578e-01 -1.86718687e-01 -5.40201426e-01
-1.02634370e+00 9.20098186e-01 3.64359200e-01 -2.53800750e-01
8.70929539e-01 -1.12143062e-01 1.28891790e+00 4.49456662e-01
6.15875304e-01 -9.56455708e-01 4.14228171e-01 8.23132932e-01
9.02046680e-01 -1.45288968e+00 -1.23819150e-01 -2.68860102e-01
-8.84310067e-01 1.16840994e+00 1.08637667e+00 -3.25343460e-01
6.02461517e-01 -5.97708300e-02 -1.33705959e-01 2.28266239e-01
-1.30859733e+00 6.14204071e-02 2.28808448e-01 3.90350431e-01
2.04873383e-01 -1.65414110e-01 4.65115644e-02 3.89304608e-01
1.35447159e-01 4.74219769e-01 4.19550806e-01 1.21070099e+00
-2.54600525e-01 -8.49991083e-01 -1.03457756e-01 3.23878467e-01
-3.99021417e-01 1.99915752e-01 -8.93925317e-03 6.47374272e-01
-4.48690436e-04 6.86862051e-01 2.54595697e-01 -6.05078220e-01
1.27115279e-01 1.02242529e-01 5.11707962e-01 -2.64892757e-01
-1.98384121e-01 -1.16681233e-01 1.05227187e-01 -1.19987142e+00
-5.42415619e-01 -8.01506519e-01 -1.36945045e+00 -1.78234845e-01
-1.26686946e-01 -2.71373272e-01 1.26910344e-01 1.05848444e+00
5.52664220e-01 5.46748817e-01 2.27592483e-01 -1.20226467e+00
-5.01518309e-01 -9.76224959e-01 -1.70526579e-01 5.92766941e-01
5.70950627e-01 -6.55586898e-01 -1.84561580e-01 4.03883100e-01] | [8.09397029876709, 0.30766355991363525] |
a7e1be60-a3c7-4630-b7f1-c2f3f32d060c | a-cnn-regression-model-to-estimate-buildings | 2307.01378 | null | https://arxiv.org/abs/2307.01378v1 | https://arxiv.org/pdf/2307.01378v1.pdf | A CNN regression model to estimate buildings height maps using Sentinel-1 SAR and Sentinel-2 MSI time series | Accurate estimation of building heights is essential for urban planning, infrastructure management, and environmental analysis. In this study, we propose a supervised Multimodal Building Height Regression Network (MBHR-Net) for estimating building heights at 10m spatial resolution using Sentinel-1 (S1) and Sentinel-2 (S2) satellite time series. S1 provides Synthetic Aperture Radar (SAR) data that offers valuable information on building structures, while S2 provides multispectral data that is sensitive to different land cover types, vegetation phenology, and building shadows. Our MBHR-Net aims to extract meaningful features from the S1 and S2 images to learn complex spatio-temporal relationships between image patterns and building heights. The model is trained and tested in 10 cities in the Netherlands. Root Mean Squared Error (RMSE), Intersection over Union (IOU), and R-squared (R2) score metrics are used to evaluate the performance of the model. The preliminary results (3.73m RMSE, 0.95 IoU, 0.61 R2) demonstrate the effectiveness of our deep learning model in accurately estimating building heights, showcasing its potential for urban planning, environmental impact analysis, and other related applications. | ['Yifang Ban', 'Andrea Nascetti', 'Ritu Yadav'] | 2023-07-03 | null | null | null | null | ['management'] | ['miscellaneous'] | [ 2.19127417e-01 -4.12697569e-02 6.99196383e-03 -3.78333390e-01
-1.06672025e+00 -5.65314107e-02 5.68509877e-01 3.30234617e-01
-2.79182941e-01 8.73006582e-01 2.52406627e-01 -6.64042234e-01
-3.00291628e-01 -1.59431982e+00 -4.09830511e-01 -8.19593549e-01
-7.60807633e-01 3.95356379e-02 -1.30290940e-01 -6.67667091e-01
-4.28871423e-01 7.32537210e-01 -1.39937401e+00 -6.04349189e-02
9.09449637e-01 1.19894564e+00 3.05614442e-01 5.72467625e-01
3.70027304e-01 4.33066308e-01 -2.40735844e-01 4.13807809e-01
5.17003953e-01 1.05811939e-01 -4.52785969e-01 1.00543514e-01
3.77477795e-01 -3.91124904e-01 -1.54252961e-01 6.91555142e-01
4.47325975e-01 -9.67566818e-02 4.97474134e-01 -1.05231452e+00
-3.44494045e-01 3.59811127e-01 -5.27277112e-01 1.53423548e-01
-9.09648314e-02 2.22228259e-01 1.01787424e+00 -5.17166436e-01
5.90121225e-02 7.86091328e-01 9.94474471e-01 -2.38613501e-01
-1.10174012e+00 -6.66070759e-01 -6.96649775e-02 -1.83201745e-01
-2.06911874e+00 -2.05863923e-01 3.04810673e-01 -7.36199737e-01
7.75753796e-01 6.50252521e-01 9.56064403e-01 3.04000795e-01
1.59628242e-01 4.17769879e-01 9.57410455e-01 -3.08830798e-01
1.35102779e-01 -1.05405524e-01 -2.59432614e-01 7.25985527e-01
3.45968395e-01 4.87882793e-01 1.86012134e-01 2.66514421e-01
6.93897545e-01 2.40412369e-01 -7.95617700e-03 1.33064941e-01
-1.00214779e+00 1.01709259e+00 1.38405454e+00 5.09603858e-01
-7.11794376e-01 2.66863972e-01 -2.69269258e-01 1.51999712e-01
6.94593310e-01 7.09365830e-02 -2.55603820e-01 4.72350150e-01
-1.30468357e+00 3.11523438e-01 -1.14861175e-01 5.63752949e-01
8.50959897e-01 3.33090365e-01 6.06177486e-02 8.71900916e-01
5.63430011e-01 1.21224141e+00 -9.07379016e-02 -6.36836827e-01
5.09012699e-01 8.40225637e-01 -1.26695903e-02 -1.16531527e+00
-8.92518640e-01 -7.79294014e-01 -1.37611222e+00 5.09588160e-02
-1.39217734e-01 -1.51983336e-01 -8.43905926e-01 1.40944970e+00
9.45771709e-02 -2.24014580e-01 3.62703167e-02 6.72416449e-01
1.02671552e+00 7.90220797e-01 1.54566735e-01 2.37606885e-03
8.35937202e-01 -4.15404469e-01 -1.13763824e-01 -6.31429553e-01
6.47519827e-01 -2.62248069e-01 6.76358998e-01 -3.80317420e-01
-5.20025194e-01 -5.27974248e-01 -9.31484818e-01 5.10756493e-01
-7.02019572e-01 3.27654421e-01 5.50193906e-01 6.76646471e-01
-1.18988967e+00 3.38859051e-01 -6.62227392e-01 -4.68466878e-01
5.43055892e-01 2.13616818e-01 -7.28260577e-02 -2.96866130e-02
-1.36638319e+00 6.00940287e-01 3.20312500e-01 5.13576031e-01
-5.02105653e-01 -6.13350928e-01 -1.01994908e+00 3.92110944e-02
-2.09560290e-01 -3.90860051e-01 6.48113072e-01 -5.45554936e-01
-4.76845175e-01 8.11406910e-01 3.87686044e-01 -6.67627156e-01
2.02563405e-01 1.46945789e-01 -8.21406543e-01 -2.16988534e-01
5.81942499e-01 8.62195492e-01 1.87286898e-01 -1.26422024e+00
-9.22453582e-01 -6.89690113e-01 -2.71625310e-01 2.57417522e-02
4.46624346e-02 -3.44308168e-01 2.43106872e-01 -4.52273726e-01
5.48842967e-01 -7.88548350e-01 -6.05488896e-01 -3.76183212e-01
-2.34922454e-01 3.35543782e-01 4.60847408e-01 -1.00887489e+00
1.45586693e+00 -1.97902834e+00 -5.69134533e-01 6.69791996e-01
-6.14229366e-02 1.30722821e-01 -2.12189049e-01 5.30503988e-01
1.15555607e-01 1.88062280e-01 -6.53688550e-01 3.63030881e-01
-1.78939596e-01 2.18410775e-01 -6.33752346e-02 3.95416796e-01
2.96040773e-02 1.21509159e+00 -6.82868123e-01 -2.93396235e-01
6.81682408e-01 6.23074770e-01 -3.13941482e-03 -3.09614778e-01
1.42524704e-01 3.81293625e-01 -4.66614753e-01 7.32197344e-01
8.57981801e-01 -1.34858817e-01 -8.02920908e-02 1.77394096e-02
-7.12213337e-01 8.44366997e-02 -1.06378376e+00 8.82485688e-01
-7.53629446e-01 8.91830981e-01 -6.86943680e-02 -6.93614364e-01
1.13083184e+00 1.01182334e-01 5.69427907e-01 -1.29017782e+00
-1.83643281e-01 1.00940853e-01 -2.97298908e-01 -3.65117133e-01
4.00677383e-01 -2.12332100e-01 -1.31768256e-01 -1.97251961e-01
-7.93929994e-01 -3.37113976e-01 -1.80595726e-01 -2.13258252e-01
7.19209969e-01 -4.94093478e-01 5.02948582e-01 -2.09576353e-01
4.74108279e-01 -6.73569888e-02 1.49789929e-01 5.10498941e-01
1.02623235e-02 4.09998536e-01 -9.54364836e-02 -8.88088346e-01
-1.12317181e+00 -1.08994460e+00 -5.22481680e-01 7.46985555e-01
-2.90879160e-01 1.55427113e-01 -2.78694242e-01 -6.60867169e-02
1.72859073e-01 9.39545095e-01 -4.24780816e-01 2.37853453e-01
-2.75656253e-01 -1.14199173e+00 6.04876161e-01 6.60809696e-01
9.06084239e-01 -7.22228646e-01 -8.48979592e-01 2.33037829e-01
-5.50815344e-01 -1.00453138e+00 3.91841531e-01 -1.71390459e-01
-8.80247295e-01 -9.59042251e-01 -4.79470938e-01 -2.77635127e-01
3.32627565e-01 3.87584656e-01 1.12972200e+00 1.10167496e-01
-3.09830070e-01 3.61497343e-01 -1.34903967e-01 -2.65226632e-01
-1.37396947e-01 2.45981827e-01 -2.19375357e-01 -1.29697978e-01
7.21406341e-02 -7.41749883e-01 -5.53195953e-01 3.21542293e-01
-6.95107877e-01 2.12621897e-01 7.81902730e-01 2.15587437e-01
5.12374997e-01 5.00850737e-01 4.39860225e-01 -1.18614025e-01
1.30354643e-01 -4.66912627e-01 -1.07969224e+00 1.97623849e-01
-6.98232055e-01 -4.02532041e-01 -3.88406701e-02 4.49601203e-01
-6.18974447e-01 1.14469923e-01 -4.20908362e-01 3.76088798e-01
-4.94776964e-01 9.05700982e-01 -5.29975258e-02 5.65015674e-02
7.52135217e-01 2.24772513e-01 -4.43727016e-01 -2.44719476e-01
7.01145530e-02 9.48256910e-01 3.95764977e-01 -6.08866364e-02
1.08350492e+00 6.19319022e-01 2.42454574e-01 -1.60618508e+00
-1.07808018e+00 -8.13643277e-01 -7.75725424e-01 -4.15702671e-01
9.01250064e-01 -1.31218958e+00 -3.06288362e-01 3.28433514e-01
-6.68964148e-01 -4.47838038e-01 -1.86500639e-01 4.50228095e-01
-2.57734120e-01 -2.59640813e-01 7.41207898e-02 -1.17992914e+00
-4.94077861e-01 -7.08882391e-01 1.15333056e+00 -5.91125190e-02
1.63338095e-01 -8.98771465e-01 1.70709342e-01 5.61875045e-01
8.82099092e-01 1.02839208e+00 6.20948493e-01 2.04491019e-02
-8.88982654e-01 -3.82089764e-01 -6.60306394e-01 2.61725008e-01
2.40151227e-01 -2.32984096e-01 -8.83968711e-01 -1.80278942e-01
-5.97731590e-01 8.99706408e-02 1.02519119e+00 8.94174874e-01
7.26608217e-01 -5.13182163e-01 -1.03325248e-01 4.90969568e-01
1.70972550e+00 6.37587830e-02 9.91318166e-01 7.82152772e-01
5.18037379e-01 4.44283515e-01 4.54885989e-01 5.72825074e-01
5.56822777e-01 6.08948231e-01 1.01620543e+00 -6.10353649e-01
1.56674653e-01 -4.61605154e-02 1.08844891e-01 3.24832439e-01
-4.56490725e-01 2.65819430e-01 -1.59442759e+00 6.55707240e-01
-1.57259798e+00 -1.15057766e+00 -5.96831799e-01 2.30596900e+00
-4.92204987e-02 -4.57442440e-02 1.72289997e-01 1.66573808e-01
4.29578334e-01 6.30003333e-01 -1.54419944e-01 2.74667740e-01
-3.37768555e-01 1.46013618e-01 1.10998356e+00 6.12510979e-01
-1.44056416e+00 7.45056033e-01 5.79844189e+00 3.62813264e-01
-1.09807611e+00 -1.18878447e-02 6.75381064e-01 3.78690772e-02
-2.79153556e-01 -3.77202153e-01 -5.80054760e-01 -1.44420210e-02
1.16482949e+00 1.94290042e-01 3.51128817e-01 7.91299820e-01
6.21348202e-01 -2.70961523e-01 -1.61974758e-01 6.52902424e-01
-4.88548487e-01 -1.44996059e+00 -2.27383271e-01 5.06433129e-01
8.19768608e-01 8.57622266e-01 1.26315281e-01 3.62966388e-01
3.14823180e-01 -1.03662777e+00 3.26309770e-01 4.82300252e-01
1.02353132e+00 -9.17705059e-01 9.85359371e-01 3.61159593e-01
-1.79240954e+00 -3.20948958e-01 -2.08338067e-01 -1.46697968e-01
1.03976503e-01 8.78620446e-01 -8.05677772e-01 7.13802814e-01
7.22834647e-01 5.95504701e-01 -5.55529952e-01 8.94144416e-01
-2.08073542e-01 5.46583891e-01 -4.77537841e-01 3.47694725e-01
8.09924364e-01 -2.87184447e-01 1.91286191e-01 1.01088083e+00
7.03097582e-01 3.59938234e-01 2.78291345e-01 7.03995824e-01
2.33263731e-01 2.12480426e-01 -8.35044980e-01 1.67201027e-01
2.63824672e-01 1.11795950e+00 -5.22242367e-01 -1.31426491e-02
-2.34600455e-01 1.43176645e-01 -2.85351455e-01 2.44439736e-01
-7.40192235e-01 -6.54021800e-02 6.94337785e-01 7.89711475e-01
2.93302029e-01 -5.17516732e-01 -8.79482776e-02 -5.56433558e-01
-3.40462863e-01 -4.54747826e-01 1.82717696e-01 -8.64533722e-01
-6.51282310e-01 3.67108911e-01 1.04852244e-01 -1.17528188e+00
3.43127772e-02 -2.31968805e-01 -5.54786861e-01 7.10598826e-01
-1.82157385e+00 -1.55238199e+00 -7.25660443e-01 1.71042129e-01
2.68885702e-01 -9.85804647e-02 7.54547596e-01 3.23307186e-01
-4.68820453e-01 -7.14517236e-02 3.93637240e-01 2.06291303e-01
-2.61989743e-01 -9.33970571e-01 6.48563385e-01 5.70288420e-01
-7.77412727e-02 -8.90890360e-02 4.40023124e-01 -6.21613026e-01
-5.15400171e-01 -1.86111534e+00 1.21792328e+00 1.26852870e-01
4.77584124e-01 2.04775169e-01 -4.42242593e-01 7.36928344e-01
-3.85687530e-01 -1.29126862e-01 7.47967422e-01 3.23491305e-01
-7.95876682e-02 -6.08566999e-01 -1.19928300e+00 2.69362092e-01
6.22533977e-01 -3.38410646e-01 7.31380954e-02 6.46561205e-01
5.89035809e-01 1.68065652e-02 -9.80300605e-01 6.80724084e-01
6.27670109e-01 -1.19537771e+00 1.25434411e+00 -1.56533957e-01
4.04666036e-01 -1.10025220e-01 -1.04891694e+00 -1.04663193e+00
-6.40312314e-01 3.16274077e-01 5.15983522e-01 6.29044354e-01
5.73895276e-01 -5.20094395e-01 4.77217138e-01 3.73063147e-01
-3.81100625e-02 -5.41305184e-01 -1.17618728e+00 -9.32269752e-01
-3.25910002e-02 -9.08836305e-01 1.07757640e+00 7.20585465e-01
-8.83543074e-01 1.48363397e-01 -1.38989821e-01 1.10251415e+00
6.46083891e-01 1.39424279e-01 7.13320076e-01 -1.69651806e+00
4.20950145e-01 -4.35280859e-01 -4.61026222e-01 -5.18638551e-01
-1.72018275e-01 -6.31526291e-01 -2.85510719e-01 -2.00345159e+00
-3.36028002e-02 -7.66099036e-01 -4.30799901e-01 6.98300123e-01
3.34099710e-01 2.89853007e-01 1.17295962e-02 5.93602993e-02
-7.89024234e-02 7.35813916e-01 5.60644329e-01 -4.47743058e-01
-3.80232900e-01 5.77418566e-01 -1.65909365e-01 6.37039840e-01
1.13610423e+00 -2.64853746e-01 -5.05015329e-02 -3.73974711e-01
4.33624953e-01 1.75967023e-01 8.77070189e-01 -1.58599150e+00
-3.16826582e-01 -3.53261799e-01 4.78530824e-01 -1.31709266e+00
2.58028090e-01 -1.10231221e+00 3.75074565e-01 8.45787227e-01
3.58232446e-02 -3.79378498e-02 3.30607623e-01 2.69533157e-01
-7.55333528e-02 9.77155566e-02 7.12881088e-01 -1.52953580e-01
-7.39128828e-01 5.31239152e-01 -4.69361842e-01 -4.38049018e-01
7.98474133e-01 -2.19494224e-01 1.58730149e-02 -6.64642096e-01
-7.02592611e-01 4.41310257e-01 2.91111648e-01 1.72132298e-01
7.00202703e-01 -1.55237830e+00 -1.00530255e+00 5.65385699e-01
3.97730857e-01 -3.76778767e-02 4.27415252e-01 6.67719305e-01
-6.19241953e-01 8.13957751e-01 1.20225167e-02 -7.63891935e-01
-1.23466539e+00 -1.58375874e-02 6.19942844e-01 -3.51519495e-01
-4.53640610e-01 5.91713130e-01 -8.45434070e-02 -1.04080677e+00
-2.83465117e-01 -4.98549879e-01 -1.75245821e-01 2.88989872e-01
3.47213566e-01 6.12962842e-01 1.23389922e-01 -1.01201797e+00
-4.41288561e-01 5.94880581e-01 6.68529212e-01 -8.59731436e-02
1.76829231e+00 -3.12023520e-01 1.71848591e-02 5.57301104e-01
9.24629867e-01 -2.98922211e-01 -8.37603450e-01 -3.39723974e-01
7.05338418e-02 -3.41445029e-01 7.00095356e-01 -6.76392317e-01
-1.12381601e+00 7.02096164e-01 1.08758461e+00 4.47068870e-01
1.26029396e+00 -4.66154777e-02 7.16876090e-01 5.75145066e-01
2.44193301e-01 -1.04760277e+00 -5.61052024e-01 3.76700848e-01
1.15072644e+00 -1.50863922e+00 3.38013500e-01 -9.27067772e-02
-4.13441330e-01 8.40524614e-01 5.39745130e-02 4.87927258e-01
8.17214310e-01 -1.54887706e-01 1.07193395e-01 -3.75610799e-01
-1.69081956e-01 -8.98696005e-01 3.15703124e-01 5.20885468e-01
2.92578489e-01 7.66583800e-01 2.78875321e-01 -1.47202522e-01
-5.61293125e-01 -2.27956444e-01 3.28516662e-02 6.30124092e-01
-1.16636157e+00 -4.22509581e-01 -6.22307301e-01 5.51865697e-01
3.63753617e-01 -2.78729588e-01 -1.58627778e-02 1.17697918e+00
2.12369546e-01 8.91850531e-01 3.19987148e-01 -2.79320002e-01
4.29472387e-01 -3.29642087e-01 1.14113413e-01 -2.83374369e-01
-2.54503280e-01 -1.78465862e-02 9.75535288e-02 -4.70554113e-01
-5.03853321e-01 -6.40196741e-01 -9.50380802e-01 -4.66779083e-01
-1.66910395e-01 -1.77494958e-01 1.03343391e+00 8.60608816e-01
1.05792612e-01 5.91867149e-01 9.91600633e-01 -7.02493966e-01
-1.77218940e-03 -9.83376801e-01 -8.13351274e-01 -2.17608705e-01
6.45979166e-01 -3.52697939e-01 -7.09929690e-02 -4.53645766e-01] | [9.383211135864258, -1.4673593044281006] |
99460561-361d-4781-af20-36f3e26575a9 | transferring-models-trained-on-natural-images | 2303.01491 | null | https://arxiv.org/abs/2303.01491v1 | https://arxiv.org/pdf/2303.01491v1.pdf | Transferring Models Trained on Natural Images to 3D MRI via Position Encoded Slice Models | Transfer learning has remarkably improved computer vision. These advances also promise improvements in neuroimaging, where training set sizes are often small. However, various difficulties arise in directly applying models pretrained on natural images to radiologic images, such as MRIs. In particular, a mismatch in the input space (2D images vs. 3D MRIs) restricts the direct transfer of models, often forcing us to consider only a few MRI slices as input. To this end, we leverage the 2D-Slice-CNN architecture of Gupta et al. (2021), which embeds all the MRI slices with 2D encoders (neural networks that take 2D image input) and combines them via permutation-invariant layers. With the insight that the pretrained model can serve as the 2D encoder, we initialize the 2D encoder with ImageNet pretrained weights that outperform those initialized and trained from scratch on two neuroimaging tasks -- brain age prediction on the UK Biobank dataset and Alzheimer's disease detection on the ADNI dataset. Further, we improve the modeling capabilities of 2D-Slice models by incorporating spatial information through position embeddings, which can improve the performance in some cases. | ["the Alzheimer's Disease Neuroimaging Initiative", 'Greg Ver Steeg', 'Paul M. Thompson', 'Nikhil Dhinagar', 'Tamoghna Chattopadhyay', 'Umang Gupta'] | 2023-03-02 | null | null | null | null | ['alzheimer-s-disease-detection'] | ['medical'] | [ 2.60189146e-01 5.00216186e-01 -5.79199456e-02 -7.44146109e-01
-5.42248905e-01 -3.22662473e-01 5.90446413e-01 -1.35521665e-01
-8.69349778e-01 6.15957737e-01 4.84663874e-01 -4.55656052e-01
8.41639638e-02 -7.20138013e-01 -9.77009416e-01 -4.54492867e-01
-5.68808973e-01 6.54768348e-01 1.69339538e-01 1.29263699e-01
-1.58956066e-01 5.19676983e-01 -8.61697257e-01 2.19598010e-01
6.07179284e-01 1.03070080e+00 3.23497891e-01 6.45003319e-01
1.88425928e-01 5.18245876e-01 -1.44883990e-01 -3.87678117e-01
4.21806753e-01 -7.25139454e-02 -9.41148162e-01 1.23142846e-01
7.16522634e-01 -8.65163147e-01 -7.98852623e-01 9.87596452e-01
6.42046273e-01 -1.99831113e-01 8.17519963e-01 -9.20096755e-01
-1.21456897e+00 6.44263387e-01 -6.25486493e-01 7.39085555e-01
-3.15114826e-01 2.01045841e-01 9.05273080e-01 -6.50347054e-01
6.57976985e-01 1.17583561e+00 7.96633601e-01 7.88363159e-01
-1.24158144e+00 -4.95029092e-01 6.59764111e-02 4.66546595e-01
-9.63234305e-01 -3.06175649e-01 5.49183190e-01 -6.13719881e-01
1.02797890e+00 -1.13028556e-01 9.02942479e-01 1.16599667e+00
3.98037076e-01 7.57867932e-01 1.18213189e+00 -1.69356719e-01
9.15272161e-02 -3.47885460e-01 1.92893744e-01 6.10121489e-01
1.39434129e-01 4.29288708e-02 2.80509088e-02 -2.21656053e-03
1.25651383e+00 1.44595280e-01 -1.20990619e-01 -5.33987105e-01
-1.41142595e+00 8.98816228e-01 1.01485741e+00 2.32960492e-01
-4.85451400e-01 6.48184791e-02 4.84812468e-01 2.48734325e-01
5.27196646e-01 5.31995237e-01 -4.76874709e-01 4.71436173e-01
-9.77948070e-01 2.08070040e-01 2.60719150e-01 5.46812713e-01
3.42070490e-01 -1.41127482e-01 -4.35542548e-03 8.73092771e-01
7.29140043e-02 2.56177485e-01 8.97841990e-01 -7.96141446e-01
4.11236078e-01 3.36342841e-01 -3.46778572e-01 -4.32227343e-01
-8.04007173e-01 -5.91502309e-01 -9.20703173e-01 1.74098700e-01
5.49628735e-01 -2.28785723e-01 -1.44516611e+00 1.88293791e+00
5.66529781e-02 1.46903127e-01 -2.23641947e-01 1.05248249e+00
5.02566397e-01 1.99855879e-01 1.05107225e-01 3.59106719e-01
1.43111253e+00 -1.16014433e+00 -2.97506362e-01 -6.04508758e-01
7.26108253e-01 -1.48129478e-01 8.12304318e-01 1.34713948e-01
-1.28174973e+00 -4.23895985e-01 -1.13315880e+00 -3.87086779e-01
-4.10610497e-01 3.88214272e-03 7.56964266e-01 4.44504797e-01
-1.43545544e+00 4.80100870e-01 -1.25680590e+00 -3.59146565e-01
9.34509516e-01 4.99173015e-01 -6.45119786e-01 -4.32968020e-01
-1.24966598e+00 1.29536879e+00 2.98387975e-01 -7.87150934e-02
-9.58777547e-01 -1.27872479e+00 -9.13915873e-01 1.85805392e-02
-4.02796734e-03 -9.78185117e-01 1.15886867e+00 -6.74268782e-01
-9.96501088e-01 1.21471012e+00 2.29146093e-01 -9.55445588e-01
5.13147354e-01 -4.52727154e-02 -3.58579814e-01 4.56869483e-01
1.67406827e-01 1.31416070e+00 7.10583746e-01 -6.40203774e-01
-1.31183863e-01 -8.27479899e-01 1.20528035e-01 2.36041561e-01
-3.14757079e-01 -2.80225798e-02 -1.48092970e-01 -4.84968632e-01
6.13885969e-02 -9.32346225e-01 -5.71941614e-01 4.60649818e-01
-2.41617665e-01 1.17111966e-01 4.65339988e-01 -1.03554094e+00
5.46497405e-01 -2.12735200e+00 1.90931976e-01 6.36873841e-02
6.60115480e-01 9.84933823e-02 -2.56282300e-01 -2.63900489e-01
-6.32004082e-01 1.52770072e-01 -3.31938297e-01 -3.10505152e-01
-1.63916156e-01 3.13912809e-01 2.37015963e-01 4.44413453e-01
4.59139735e-01 1.24389148e+00 -9.03603017e-01 -3.73191714e-01
2.26533003e-02 4.87954468e-01 -8.47508132e-01 9.21639726e-02
2.52854347e-01 5.22183359e-01 -2.20812827e-01 1.87288582e-01
6.36110961e-01 -4.55779225e-01 7.30045792e-03 -3.43211472e-01
1.92539141e-01 3.95474970e-01 -4.65761840e-01 1.98488426e+00
-3.81569266e-01 4.76561248e-01 -2.68314686e-02 -1.37171817e+00
4.23388422e-01 2.33282641e-01 6.89476252e-01 -9.30396438e-01
-1.01909108e-01 1.77172393e-01 6.80075645e-01 -5.79595447e-01
-1.79883882e-01 -2.93085396e-01 1.95047930e-01 5.73818803e-01
4.09677237e-01 -2.42158547e-02 4.88704666e-02 6.74290508e-02
1.31127882e+00 -2.55193532e-01 9.73598957e-02 -2.43491322e-01
1.23709373e-01 -2.57449836e-01 3.47937346e-01 6.09198689e-01
-4.08422858e-01 9.95380580e-01 6.14020348e-01 -6.32602155e-01
-1.66003728e+00 -1.37789845e+00 -4.32728380e-01 7.11263597e-01
-5.67908525e-01 -1.49007924e-02 -8.49783003e-01 -8.04424345e-01
6.10303693e-02 3.33028257e-01 -1.18362260e+00 -3.14510137e-01
-7.61698484e-01 -1.07472968e+00 5.65337121e-01 1.03172100e+00
3.62922221e-01 -6.81497872e-01 -6.77539468e-01 1.15234621e-01
2.14364946e-01 -1.17208362e+00 -6.50700510e-01 4.50920254e-01
-1.22553337e+00 -1.06849730e+00 -1.26713884e+00 -8.50585043e-01
9.83101547e-01 -1.18263096e-01 1.09002316e+00 -9.82071236e-02
-6.22377515e-01 2.71507293e-01 -3.22369933e-02 -4.63240057e-01
-7.97826648e-02 3.01635474e-01 1.46016136e-01 -2.88010508e-01
2.63127655e-01 -7.61811972e-01 -9.93242085e-01 -7.23857433e-02
-9.20071006e-01 1.28128067e-01 9.10375893e-01 1.05253792e+00
4.03453588e-01 -4.52578545e-01 7.36642480e-01 -9.86804128e-01
4.51840192e-01 -5.74934006e-01 2.89791655e-02 9.80210528e-02
-4.99737501e-01 2.16702029e-01 3.70022327e-01 -4.98503745e-01
-6.27174497e-01 -1.06558606e-01 -2.97287554e-01 -4.92471606e-01
-8.73172060e-02 5.67004681e-01 9.39454436e-02 4.11278978e-02
6.43641770e-01 9.32421014e-02 4.27896053e-01 -6.22600555e-01
3.64110529e-01 4.01058495e-01 6.09137237e-01 -2.95232534e-01
3.84970963e-01 4.57174510e-01 1.38632627e-02 -5.85079730e-01
-8.42382967e-01 1.32842883e-02 -9.94656622e-01 2.33437389e-01
1.18144500e+00 -9.76957738e-01 -1.82279367e-02 4.51784104e-01
-8.22191715e-01 -5.60239553e-01 -3.09086531e-01 7.85536230e-01
-5.17824471e-01 1.85895085e-01 -9.34148431e-01 -9.43300575e-02
-2.93625414e-01 -1.42030263e+00 9.10492957e-01 -8.73635858e-02
-1.41377151e-01 -1.20627725e+00 -1.59221798e-01 2.11446345e-01
5.36629319e-01 4.46727090e-02 1.48205030e+00 -8.87862027e-01
-1.43322527e-01 -8.95806327e-02 -4.64730173e-01 5.44843733e-01
7.39032924e-02 -7.21412003e-01 -8.24563146e-01 -3.64150882e-01
-6.81534468e-04 -5.15022218e-01 9.47671175e-01 7.53686011e-01
1.45715117e+00 -4.66358140e-02 -2.59393036e-01 7.32780695e-01
9.13280427e-01 5.40591776e-02 6.24013364e-01 3.91542643e-01
6.84017718e-01 3.14660072e-01 -3.50000918e-01 -5.11299036e-02
5.75735688e-01 3.54398727e-01 4.40981627e-01 -4.17571127e-01
-5.19256890e-01 -4.68805619e-02 -5.47451116e-02 9.33843315e-01
-1.09071760e-02 3.71061713e-01 -9.79771137e-01 6.27824962e-01
-1.43094420e+00 -5.23900092e-01 1.93559110e-01 2.00308013e+00
1.11874783e+00 2.60472506e-01 7.33206198e-02 -1.13862284e-01
5.11362672e-01 1.85494199e-01 -8.42764735e-01 -1.58276126e-01
-3.40243131e-02 4.83907461e-01 7.61395931e-01 3.67234737e-01
-1.14595652e+00 4.31187004e-01 6.91656113e+00 9.97537300e-02
-1.33484316e+00 5.05123258e-01 1.03263426e+00 -3.65483761e-01
-1.71168789e-01 -4.50167924e-01 -3.70468736e-01 4.92696464e-01
1.12683940e+00 2.56650113e-02 3.46092373e-01 6.34471714e-01
3.21945325e-02 1.48804024e-01 -1.48311675e+00 9.56257045e-01
1.19952731e-01 -1.20672131e+00 8.46723318e-02 1.92148715e-01
6.76652849e-01 5.19416094e-01 4.09691781e-01 3.13909948e-01
1.55423149e-01 -1.40667212e+00 4.74323541e-01 5.69180727e-01
9.41539407e-01 -4.38191235e-01 6.66650534e-01 7.42918923e-02
-5.13027012e-01 -2.71501355e-02 -4.76193100e-01 2.23178472e-02
1.31568179e-01 6.34265959e-01 -1.15096486e+00 1.70507714e-01
7.97699392e-01 7.28176415e-01 -8.28297138e-01 1.28879011e+00
-8.15921575e-02 5.86911619e-01 -3.45261097e-01 4.32369858e-01
4.58687186e-01 -9.94760543e-03 1.16441280e-01 9.52874243e-01
2.22952545e-01 1.67799313e-02 3.04632671e-02 9.97930527e-01
-2.66883999e-01 -2.27017611e-01 -5.54195464e-01 1.43797085e-01
1.05162330e-01 1.09628737e+00 -5.51070988e-01 -3.61648679e-01
-6.21814191e-01 8.59156370e-01 4.84625310e-01 3.06192547e-01
-6.43150628e-01 -1.15210414e-01 6.09216809e-01 2.80298471e-01
3.73443216e-01 -4.10425425e-01 -5.27748346e-01 -1.10016119e+00
-1.96026400e-01 -7.42294669e-01 1.18761607e-01 -8.48007739e-01
-1.53398478e+00 6.02044225e-01 -3.43038253e-02 -5.52969396e-01
-2.39739165e-01 -1.01474893e+00 -5.61537385e-01 1.07346976e+00
-1.55279028e+00 -1.19116187e+00 1.83771461e-01 4.82072085e-01
3.49354714e-01 2.80764955e-03 6.93991363e-01 6.08728051e-01
-4.14686441e-01 6.18685305e-01 -9.85893458e-02 5.30796826e-01
8.66127849e-01 -1.41373384e+00 7.30673492e-01 5.68731606e-01
-1.18186504e-01 7.13909566e-01 2.21124113e-01 -5.39565742e-01
-9.68791246e-01 -1.08774400e+00 7.32821703e-01 -4.80117470e-01
8.79874825e-01 -4.83929634e-01 -9.79687810e-01 1.01405215e+00
2.04410881e-01 4.64154541e-01 6.78612292e-01 3.28079872e-02
-3.59211296e-01 1.33220822e-01 -1.18352890e+00 6.25924468e-01
1.21208513e+00 -6.40769660e-01 -9.14550900e-01 3.42323005e-01
7.08216071e-01 -4.77814168e-01 -1.27333486e+00 2.10633725e-01
5.69508433e-01 -5.05592108e-01 1.33370483e+00 -1.13210130e+00
6.44236267e-01 3.14417869e-01 4.32632491e-02 -1.68081248e+00
-5.00080705e-01 1.64879397e-01 -6.58131838e-02 4.70312715e-01
5.08395076e-01 -6.02478921e-01 7.05614865e-01 7.99851954e-01
-3.02722692e-01 -1.00681698e+00 -9.33071733e-01 -6.58495665e-01
6.97330117e-01 -1.69122189e-01 6.08139992e-01 8.41462195e-01
-1.80543914e-01 2.43439645e-01 2.37479177e-03 5.78698926e-02
5.82183778e-01 -4.86039996e-01 -4.53850254e-02 -1.28662956e+00
-1.93148851e-01 -4.51425940e-01 -6.30978882e-01 -9.47856188e-01
1.58001423e-01 -1.18878579e+00 -2.26177216e-01 -1.45245647e+00
4.07768577e-01 -4.62750107e-01 -6.89343393e-01 6.96547806e-01
-2.17531025e-01 4.92480427e-01 2.02051327e-01 3.41860987e-02
-1.63746387e-01 3.07203799e-01 1.64534664e+00 -3.78428459e-01
2.74943709e-01 -2.13409200e-01 -7.24164128e-01 6.03065908e-01
7.69790232e-01 -3.02756160e-01 -4.82605934e-01 -9.87060189e-01
-9.42417681e-02 -2.05765396e-01 5.70791483e-01 -1.00375104e+00
1.86826080e-01 3.66956860e-01 9.94812250e-01 -1.07041128e-01
2.41273999e-01 -6.47137642e-01 -2.95164555e-01 4.64592069e-01
-6.55797362e-01 3.27998370e-01 6.69442117e-02 2.74415076e-01
9.29439440e-02 -3.17712724e-01 8.14057469e-01 -5.83490491e-01
-3.99039298e-01 8.20604563e-01 -2.45108962e-01 1.58776000e-01
6.89113975e-01 -7.40963891e-02 -1.34967327e-01 -1.17568411e-01
-1.23889351e+00 2.20364273e-01 3.09751749e-01 2.91294336e-01
4.78616863e-01 -1.40633786e+00 -6.91089809e-01 2.99839318e-01
-1.51408195e-01 1.78165510e-01 4.00145411e-01 1.18576777e+00
-4.40247416e-01 5.39838791e-01 -8.48877966e-01 -5.07053435e-01
-6.72658384e-01 5.85787952e-01 5.03404200e-01 -4.50034112e-01
-8.59690249e-01 8.42162013e-01 4.66465086e-01 -5.57342529e-01
7.89023489e-02 -6.86336398e-01 -4.91402447e-02 -5.80577627e-02
6.95154011e-01 -8.77542198e-02 3.52382064e-01 -2.44786158e-01
-2.71273315e-01 2.07671687e-01 -6.96072221e-01 -2.06060976e-01
1.88531744e+00 2.57702954e-02 8.37048218e-02 2.20200986e-01
1.56861007e+00 -5.86570501e-01 -1.47296119e+00 -4.77690369e-01
-8.74992553e-03 -6.94506317e-02 3.62742066e-01 -9.14008081e-01
-1.44068670e+00 1.25072420e+00 7.98343241e-01 -1.04383990e-01
7.16684401e-01 1.04498476e-01 1.00394881e+00 2.81546116e-01
1.73721761e-01 -8.73609304e-01 7.71031827e-02 4.74503517e-01
8.38017344e-01 -1.20001554e+00 -1.36494815e-01 1.10019840e-01
-4.95542318e-01 1.18646264e+00 5.84759116e-01 -4.38158482e-01
7.82842338e-01 3.04219425e-01 -9.22043398e-02 -2.36010700e-01
-4.79378402e-01 2.99225245e-02 3.16734731e-01 9.20642912e-01
3.98550510e-01 1.24737613e-01 -2.16586158e-01 7.07254887e-01
-2.73908913e-01 2.13586926e-01 4.86713350e-01 7.71557152e-01
-1.52224079e-01 -1.05703580e+00 -5.37700057e-02 1.05840492e+00
-4.80776906e-01 -3.31812978e-01 -1.04810953e-01 6.73507810e-01
1.50554985e-01 6.78377971e-02 4.79141742e-01 -1.39713302e-01
1.34344831e-01 4.18455511e-01 8.28116834e-01 -8.76487195e-01
-2.88714141e-01 -2.58945227e-01 -9.03291479e-02 -3.34481537e-01
-2.36530378e-01 -7.44468987e-01 -1.11126840e+00 7.62566105e-02
2.11677969e-01 -2.78796822e-01 6.71757400e-01 9.78328347e-01
4.40205693e-01 8.13890338e-01 2.61868626e-01 -1.00326705e+00
-8.46694291e-01 -1.15215671e+00 -6.56546474e-01 4.02165860e-01
4.37633216e-01 -7.08976209e-01 1.26969248e-01 1.45597622e-01] | [14.421257972717285, -1.9827381372451782] |
a610eea9-bb15-4469-b604-b54c9e2f9500 | multi-agent-cross-translated-diversification | 2006.02163 | null | https://arxiv.org/abs/2006.02163v4 | https://arxiv.org/pdf/2006.02163v4.pdf | Cross-model Back-translated Distillation for Unsupervised Machine Translation | Recent unsupervised machine translation (UMT) systems usually employ three main principles: initialization, language modeling and iterative back-translation, though they may apply them differently. Crucially, iterative back-translation and denoising auto-encoding for language modeling provide data diversity to train the UMT systems. However, the gains from these diversification processes has seemed to plateau. We introduce a novel component to the standard UMT framework called Cross-model Back-translated Distillation (CBD), that is aimed to induce another level of data diversification that existing principles lack. CBD is applicable to all previous UMT approaches. In our experiments, CBD achieves the state of the art in the WMT'14 English-French, WMT'16 English-German and English-Romanian bilingual unsupervised translation tasks, with 38.2, 30.1, and 36.3 BLEU respectively. It also yields 1.5-3.3 BLEU improvements in IWSLT English-French and English-German tasks. Through extensive experimental analyses, we show that CBD is effective because it embraces data diversity while other similar variants do not. | ['Ai Ti Aw', 'Thanh-Tung Nguyen', 'Xuan-Phi Nguyen', 'Wu Kui', 'Shafiq Joty'] | 2020-06-03 | cross-model-back-translated-distillation-for | https://openreview.net/forum?id=K5a_QFEUzA1 | https://openreview.net/pdf?id=K5a_QFEUzA1 | null | ['unsupervised-machine-translation'] | ['natural-language-processing'] | [ 1.62803009e-01 -2.04426393e-01 -5.15401602e-01 -2.56268531e-01
-1.27701044e+00 -6.73536718e-01 9.02302682e-01 -1.76350310e-01
-4.67897624e-01 1.01113188e+00 3.30012947e-01 -8.65715683e-01
3.24720889e-01 -2.66939282e-01 -7.33868361e-01 -5.97090840e-01
3.79393280e-01 9.45945084e-01 -2.41191000e-01 -6.15210593e-01
8.48797485e-02 1.50719196e-01 -7.60689437e-01 5.21838248e-01
1.20083523e+00 3.25309485e-01 2.42952257e-01 3.58159900e-01
-4.08112168e-01 4.89036977e-01 -3.80794138e-01 -7.58097589e-01
4.90216643e-01 -8.73234093e-01 -9.54023898e-01 -1.66273322e-02
8.19404274e-02 4.25245129e-02 -1.11074680e-02 7.63267934e-01
6.92088723e-01 -1.61368594e-01 7.05389917e-01 -9.07283902e-01
-9.13453102e-01 1.04153907e+00 -7.02876747e-01 -3.06249894e-02
2.12176234e-01 -7.97071382e-02 8.46350849e-01 -1.32946634e+00
7.87090600e-01 1.35904467e+00 7.51211524e-01 8.35241914e-01
-1.37656868e+00 -5.59071481e-01 -9.47375596e-02 -1.94848880e-01
-1.17757440e+00 -5.49943447e-01 3.22523236e-01 -2.26323798e-01
1.27633333e+00 3.88395816e-01 2.64614910e-01 1.31274223e+00
4.32524592e-01 9.55761492e-01 1.53591073e+00 -9.30146575e-01
-2.10382804e-01 4.76857424e-01 -9.81291682e-02 4.82353896e-01
1.41709790e-01 5.19291349e-02 -6.30365372e-01 -8.38276073e-02
5.10422885e-01 -3.41031671e-01 8.86189938e-02 -7.70081282e-02
-1.64485276e+00 7.13270605e-01 3.58564332e-02 5.56177437e-01
-2.05875456e-01 -2.40817577e-01 4.85498935e-01 9.81030166e-01
8.13191891e-01 5.52508593e-01 -7.15137780e-01 -2.36513019e-01
-1.12546027e+00 -6.67480379e-02 7.27304280e-01 1.33545744e+00
7.44701982e-01 1.96857780e-01 -1.92572296e-01 1.05727506e+00
1.64417312e-01 7.37302482e-01 7.79012382e-01 -6.38986588e-01
8.88783634e-01 3.33289802e-01 -2.01475233e-01 -3.48743051e-01
-1.42811105e-01 -6.05918348e-01 -9.47917044e-01 -2.47335553e-01
8.46253112e-02 -3.27441603e-01 -1.19234359e+00 1.78866041e+00
-9.28976759e-02 -4.54974622e-01 3.92848164e-01 5.77664435e-01
4.58503902e-01 7.63582408e-01 -1.69794679e-01 -4.87620205e-01
9.87685442e-01 -1.32702374e+00 -9.65374053e-01 -4.01368797e-01
8.78572822e-01 -1.37792861e+00 1.21530175e+00 1.30246922e-01
-1.31673014e+00 -5.60305655e-01 -7.80934691e-01 -1.97638825e-01
-2.99710214e-01 2.73356229e-01 4.96866345e-01 6.65025830e-01
-1.26539266e+00 4.25707996e-01 -8.21233332e-01 -7.11121380e-01
-3.62268575e-02 5.15437007e-01 -3.81576061e-01 -6.66989312e-02
-1.21394205e+00 1.21149576e+00 3.61825526e-01 -5.66768497e-02
-7.46162295e-01 -3.31584364e-01 -5.34268022e-01 -4.37152416e-01
7.89229125e-02 -9.57847416e-01 1.36198783e+00 -1.22658885e+00
-1.87537527e+00 9.39782083e-01 -5.50174057e-01 -5.29952943e-01
7.57225931e-01 -3.02406222e-01 -3.78392905e-01 -3.75038415e-01
3.27330410e-01 7.75932431e-01 5.86525023e-01 -1.18944252e+00
-5.65072894e-01 -2.00851411e-01 -4.72425461e-01 2.97213167e-01
-3.11483532e-01 4.72498238e-01 -5.42414010e-01 -9.96451259e-01
3.86439770e-01 -1.19462144e+00 -2.60273516e-01 -7.55283058e-01
-3.37264031e-01 -4.24728841e-02 5.56787014e-01 -9.00566339e-01
1.37332189e+00 -1.82128298e+00 5.32666385e-01 -1.98398277e-01
-2.12245882e-01 1.81091100e-01 -3.33150327e-01 7.84409583e-01
9.65206511e-03 2.81621695e-01 -3.95204842e-01 -7.97191143e-01
1.24691967e-02 5.43455720e-01 -2.83829391e-01 1.47921130e-01
3.67373973e-01 1.17758381e+00 -7.21625507e-01 -4.11140621e-01
-7.87943751e-02 1.95081294e-01 -5.42502344e-01 -5.08985370e-02
-1.87494904e-01 6.55735970e-01 -5.92363775e-02 8.13189507e-01
6.20901287e-01 1.61101595e-01 3.41438621e-01 2.24094674e-01
-3.54466528e-01 7.16304481e-01 -6.23956084e-01 2.02767301e+00
-3.92093748e-01 6.07092440e-01 -1.38125420e-01 -8.05071056e-01
8.64868402e-01 4.41041589e-01 2.38984704e-01 -7.34435678e-01
1.32634297e-01 9.94355857e-01 1.38099626e-01 -2.10045308e-01
6.07883990e-01 -3.93580765e-01 -2.36578509e-01 4.69640017e-01
2.87458420e-01 -1.09760486e-01 3.33254069e-01 5.32562882e-02
6.64023101e-01 4.69470859e-01 5.68196252e-02 -7.26088345e-01
4.16545123e-01 2.94375002e-01 7.81253695e-01 6.41558766e-01
1.14828102e-01 6.87117040e-01 5.19124977e-02 -2.68123120e-01
-1.33453655e+00 -1.12003398e+00 1.59073714e-02 1.09974349e+00
-1.82408929e-01 -5.41175723e-01 -7.79795587e-01 -6.58226907e-01
-4.14249480e-01 8.49493265e-01 -3.37866992e-01 -2.12645177e-02
-9.57695305e-01 -1.12041306e+00 6.83147848e-01 2.12622479e-01
6.37436748e-01 -7.26487637e-01 1.18399791e-01 2.85503447e-01
-7.40665853e-01 -1.06227887e+00 -7.00553477e-01 4.52824712e-01
-1.18964803e+00 -1.58567980e-01 -9.40126121e-01 -1.09772611e+00
5.55259407e-01 3.54870886e-01 1.33403885e+00 -4.07216519e-01
4.25064355e-01 -1.44952103e-01 -4.93946433e-01 -3.37470680e-01
-1.00118554e+00 6.62301242e-01 5.55895925e-01 -1.80075899e-01
6.20853722e-01 -5.49708664e-01 -2.72789001e-02 3.97404253e-01
-7.94305980e-01 3.70438874e-01 9.61751819e-01 1.04330945e+00
6.38426721e-01 -3.88816327e-01 3.95973235e-01 -8.68831396e-01
7.95711875e-01 -2.96282351e-01 -1.31303519e-01 5.05338371e-01
-1.00885260e+00 3.74707580e-01 6.06466532e-01 -6.27771616e-01
-1.05654716e+00 -2.64021844e-01 -1.29026592e-01 -2.59971827e-01
2.43992865e-01 4.94865537e-01 -1.41459167e-01 9.51026678e-02
7.44874775e-01 5.58704615e-01 -1.58832833e-01 -7.54359186e-01
3.76333833e-01 9.37879801e-01 2.52058744e-01 -7.99656808e-01
8.92347038e-01 -1.58863049e-02 -4.61991966e-01 -4.70828027e-01
-3.66950452e-01 -2.02036679e-01 -8.16909254e-01 3.72790933e-01
6.86904192e-01 -1.09602344e+00 2.63329238e-01 4.33556944e-01
-1.09295738e+00 -3.71354491e-01 -1.79677591e-01 6.23482168e-01
-4.25586969e-01 4.51224059e-01 -9.41642284e-01 -4.52887923e-01
-7.70180583e-01 -1.38125312e+00 1.05463004e+00 -2.98254043e-01
-3.14810902e-01 -9.17693555e-01 1.75154835e-01 3.14571887e-01
5.56633532e-01 -2.09808812e-01 1.02073359e+00 -5.29497743e-01
-2.45785818e-01 1.82713211e-01 8.39245617e-02 6.00891232e-01
3.90557081e-01 -2.43924886e-01 -5.46863317e-01 -6.27067387e-01
3.09711024e-02 -2.24394053e-01 7.71372139e-01 1.01348162e-01
2.38529578e-01 -3.10963213e-01 -1.12665989e-01 7.08675146e-01
1.26325202e+00 2.69152552e-01 5.31618774e-01 5.18343627e-01
4.83189523e-01 4.29533780e-01 4.79547292e-01 -5.24311475e-02
3.66996944e-01 7.93932378e-01 -2.96163201e-01 -4.02591377e-01
-3.38344216e-01 -3.22922230e-01 9.90760326e-01 1.85878134e+00
-1.98027194e-01 -1.17784970e-01 -9.47248995e-01 4.74609435e-01
-1.68446863e+00 -6.61994040e-01 -2.10777223e-01 2.15010476e+00
1.27202201e+00 2.30587512e-01 1.30221620e-01 -8.24831203e-02
4.92900401e-01 -3.44486773e-01 -2.11802170e-01 -9.87516344e-01
-5.91546118e-01 3.17061067e-01 6.26389861e-01 5.58984280e-01
-6.94188416e-01 1.51843071e+00 6.57734919e+00 9.44838226e-01
-1.15965664e+00 4.39884275e-01 6.63515091e-01 4.68127020e-02
-5.28684020e-01 1.22655578e-01 -8.00812423e-01 3.37101460e-01
1.24314785e+00 -1.35854006e-01 6.65313184e-01 3.03822577e-01
1.43467247e-01 2.31205955e-01 -9.92636621e-01 8.27469468e-01
1.21863289e-02 -1.00477326e+00 4.91805553e-01 1.06214665e-01
1.21937966e+00 4.05874342e-01 1.56026289e-01 6.44998431e-01
2.25900725e-01 -7.93645561e-01 7.66339123e-01 6.00102730e-02
9.77229595e-01 -7.68234491e-01 5.96995175e-01 5.52110255e-01
-8.11884999e-01 2.58853912e-01 -4.14578289e-01 5.04696406e-02
7.56614581e-02 4.72762853e-01 -8.78682017e-01 1.11639690e+00
3.57766479e-01 4.35226351e-01 -3.98883909e-01 3.62545699e-01
-2.06647038e-01 7.79413402e-01 -3.19517642e-01 1.41830340e-01
5.61693907e-01 -5.01210392e-01 6.58637524e-01 1.62843311e+00
5.29646993e-01 -4.81854856e-01 6.77235946e-02 4.78472829e-01
-2.31620684e-01 3.58162433e-01 -5.09480178e-01 -1.92126364e-01
2.47341901e-01 6.64050400e-01 -4.27111238e-01 -4.91258115e-01
-3.47180784e-01 1.50231361e+00 2.68803656e-01 4.79493707e-01
-8.63411486e-01 -1.38477460e-01 7.23118126e-01 -8.02265331e-02
8.13883450e-03 -5.60397565e-01 -4.48545456e-01 -1.56103063e+00
1.07790463e-01 -1.48802519e+00 -3.79977711e-02 -5.13188481e-01
-1.17014337e+00 1.17445624e+00 -1.42408624e-01 -1.37452805e+00
-5.99205256e-01 -6.19649649e-01 -3.33263800e-02 1.16778851e+00
-1.31845999e+00 -1.37838578e+00 4.24759448e-01 4.90151107e-01
9.83427227e-01 -5.06631076e-01 9.88142967e-01 4.82517302e-01
-6.15276456e-01 9.56985950e-01 6.86027050e-01 6.10648654e-02
1.06015754e+00 -1.19642913e+00 9.18742001e-01 1.13385534e+00
4.42817122e-01 9.37185705e-01 6.63757741e-01 -6.59592688e-01
-1.64830482e+00 -9.98806119e-01 1.65847933e+00 -7.40069211e-01
5.86350143e-01 -6.10083103e-01 -5.93816638e-01 9.01363909e-01
7.64856398e-01 -6.12586319e-01 6.21027648e-01 1.19487382e-01
-3.62695724e-01 8.01044032e-02 -8.74101639e-01 9.66884613e-01
1.13703084e+00 -4.79620546e-01 -7.28315175e-01 3.67976248e-01
8.89122128e-01 -3.79056036e-01 -7.84364462e-01 5.62881470e-01
4.66783792e-01 -6.36833668e-01 7.96118438e-01 -8.34347308e-01
4.61395383e-01 -1.30679756e-01 -4.74104047e-01 -1.43851984e+00
-4.09187406e-01 -1.20621848e+00 1.20711930e-01 1.32528961e+00
9.94193077e-01 -7.74336100e-01 3.06528211e-01 9.51122865e-02
-2.75010854e-01 -6.15064561e-01 -1.09679818e+00 -1.04460800e+00
7.26758838e-01 -3.08101982e-01 4.73955750e-01 1.26444101e+00
-1.09557569e-01 8.34775031e-01 -7.76657045e-01 -3.05830479e-01
5.29090047e-01 1.19840845e-01 7.88350642e-01 -6.94874942e-01
-4.69319791e-01 -6.08585119e-01 2.73234189e-01 -1.31410491e+00
-8.53243396e-02 -1.32976711e+00 -1.65673003e-01 -1.56956923e+00
2.94957548e-01 -2.40705773e-01 -2.94661492e-01 4.72086191e-01
-2.81020999e-01 3.59748721e-01 1.48811907e-01 5.75942039e-01
-1.80937335e-01 4.69876051e-01 1.33977103e+00 -1.35031328e-01
-3.66150439e-01 -3.04577742e-02 -6.85142159e-01 1.69546545e-01
1.06540108e+00 -5.59065998e-01 -2.69778997e-01 -1.20800447e+00
8.79267696e-03 -1.90864399e-01 -3.82557094e-01 -7.29617178e-01
9.91390459e-03 -3.36232632e-02 2.28797555e-01 -4.63037729e-01
1.73779145e-01 -5.94416082e-01 3.34702194e-01 4.41692829e-01
-1.99731141e-01 6.17286801e-01 3.53385359e-01 4.06723395e-02
-2.68815279e-01 7.20623583e-02 6.48487329e-01 -2.82716423e-01
-3.48824888e-01 -2.53125909e-03 -5.15840650e-01 -1.64362974e-02
4.76399481e-01 -2.27257952e-01 -3.43423113e-02 -1.44967645e-01
-4.11128253e-01 1.64443702e-01 3.81088346e-01 6.31229341e-01
2.75829464e-01 -1.58925271e+00 -1.40364647e+00 5.24499297e-01
1.67086385e-02 -3.72604787e-01 -3.47881049e-01 1.21099699e+00
-3.97888809e-01 6.08013153e-01 -9.34173390e-02 -6.59043789e-01
-1.10743940e+00 4.38544750e-01 1.54537708e-01 -5.86064577e-01
-3.41556400e-01 7.41490483e-01 -4.11088057e-02 -9.48512375e-01
-7.14784339e-02 -3.15178305e-01 4.29452509e-01 -9.26548094e-02
1.17939070e-01 1.91579431e-01 2.81238943e-01 -6.56825781e-01
-2.43484244e-01 4.82378483e-01 -3.61944377e-01 -5.25901735e-01
1.05548131e+00 -3.84574264e-01 -4.70256120e-01 6.08654320e-01
1.09454572e+00 3.14499378e-01 -4.76127386e-01 -5.36390126e-01
3.77181143e-01 -1.54808074e-01 -2.01771557e-01 -1.18051565e+00
-6.08972371e-01 9.27898586e-01 4.09956396e-01 -1.78448901e-01
1.36096632e+00 -2.99118131e-01 9.69422460e-01 3.95470679e-01
6.87989354e-01 -1.03340209e+00 -4.10701066e-01 8.84328485e-01
7.50317216e-01 -1.17377317e+00 -3.42315435e-01 -1.09082654e-01
-5.88742197e-01 1.00878167e+00 2.90326506e-01 3.13196212e-01
7.10196123e-02 4.17000532e-01 4.25559431e-01 4.64493126e-01
-8.56308103e-01 -1.20505795e-01 2.99324542e-01 1.42806336e-01
7.90284634e-01 1.59117490e-01 -9.04176891e-01 4.18317258e-01
-2.50838339e-01 -8.01608041e-02 1.09309062e-01 9.60684478e-01
-2.83631712e-01 -1.87234664e+00 -4.31713998e-01 -1.79181341e-02
-6.27011061e-01 -6.09282613e-01 -7.01251268e-01 8.62619877e-01
-1.51992880e-03 9.79090035e-01 -3.69239628e-01 -6.37069523e-01
2.14539036e-01 4.07377303e-01 5.80889165e-01 -4.89768356e-01
-8.80064189e-01 6.65003240e-01 1.40020296e-01 -6.61566854e-02
-4.52655762e-01 -5.98209977e-01 -8.59581769e-01 -5.01652420e-01
-2.69048959e-01 5.05019665e-01 7.32370496e-01 8.46877396e-01
4.61469173e-01 3.51130545e-01 7.59925365e-01 -4.59928304e-01
-7.40462482e-01 -1.42556775e+00 -3.83672267e-02 1.99982643e-01
1.35233715e-01 -1.27859876e-01 -2.48357803e-01 9.07071456e-02] | [11.601953506469727, 10.297959327697754] |
5b1fab77-1aa3-4e7b-90d8-60ed17a647c3 | ntnu-domain-semi-independent-short-message | null | null | https://aclanthology.org/S13-2071 | https://aclanthology.org/S13-2071.pdf | NTNU: Domain Semi-Independent Short Message Sentiment Classification | null | ['Bj{\\"o}rn Gamb{\\"a}ck', '{\\O}yvind Selmer', 'Lars Bungum', 'Mikael Brevik'] | 2013-06-01 | null | null | null | semeval-2013-6 | ['twitter-sentiment-analysis'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.346166133880615, 3.647867441177368] |
7f3371fe-b76f-4d26-9873-2498118b3cc1 | ecdans-efficient-temporal-causal-discovery | 2303.02833 | null | https://arxiv.org/abs/2303.02833v1 | https://arxiv.org/pdf/2303.02833v1.pdf | eCDANs: Efficient Temporal Causal Discovery from Autocorrelated and Non-stationary Data (Student Abstract) | Conventional temporal causal discovery (CD) methods suffer from high dimensionality, fail to identify lagged causal relationships, and often ignore dynamics in relations. In this study, we present a novel constraint-based CD approach for autocorrelated and non-stationary time series data (eCDANs) capable of detecting lagged and contemporaneous causal relationships along with temporal changes. eCDANs addresses high dimensionality by optimizing the conditioning sets while conducting conditional independence (CI) tests and identifies the changes in causal relations by introducing a surrogate variable to represent time dependency. Experiments on synthetic and real-world data show that eCDANs can identify time influence and outperform the baselines. | ['Md Osman Gani', 'Uzma Hasan', 'Muhammad Hasan Ferdous'] | 2023-03-06 | null | null | null | null | ['causal-discovery'] | ['knowledge-base'] | [-1.89650711e-02 -6.38365686e-01 -5.35922110e-01 -1.88690662e-01
-1.04336239e-01 -6.59570456e-01 1.10616851e+00 1.32898152e-01
8.29252750e-02 1.10161340e+00 6.55485809e-01 -6.64160013e-01
-1.00661147e+00 -7.09715426e-01 -2.62562633e-01 -4.72603381e-01
-1.04776382e+00 3.23992580e-01 2.53832668e-01 1.88932061e-01
1.60037354e-01 3.23783636e-01 -1.00763154e+00 -6.82805032e-02
1.02076650e+00 4.79624301e-01 -1.13857307e-01 3.81454796e-01
3.37798953e-01 8.42166483e-01 -5.70594430e-01 1.31112993e-01
-8.70524906e-04 -6.11917913e-01 -4.84226018e-01 -5.97657979e-01
-2.86752105e-01 -1.11892506e-01 -5.12643516e-01 4.87072170e-01
1.44450933e-01 6.18251637e-02 7.84259498e-01 -1.73022819e+00
-7.40067124e-01 6.87090933e-01 -7.33325541e-01 1.03003895e+00
4.49073374e-01 -1.28369212e-01 1.05463850e+00 -8.15725923e-01
5.73372126e-01 1.69236636e+00 6.08356833e-01 8.08221698e-02
-1.51186717e+00 -8.50218892e-01 5.58692336e-01 3.16310614e-01
-1.25373816e+00 2.22355705e-02 7.52002418e-01 -6.16279960e-01
7.74047732e-01 3.74044001e-01 5.83165526e-01 1.34053552e+00
4.62937832e-01 6.13153338e-01 1.22117078e+00 -1.52139992e-01
2.88994849e-01 -8.79940212e-01 2.65067458e-01 2.67991245e-01
3.35265130e-01 8.96649361e-01 -7.59088039e-01 -7.17667401e-01
8.66800308e-01 5.09995818e-02 -1.35977745e-01 1.39431208e-01
-1.48445010e+00 9.20300901e-01 1.65783256e-01 4.60948229e-01
-6.46092534e-01 1.51845738e-01 3.73148918e-01 5.02962649e-01
5.44848979e-01 3.47541094e-01 -7.20030308e-01 -2.34090030e-01
-4.45861638e-01 3.91844004e-01 6.23980165e-01 7.47616470e-01
4.19843681e-02 3.85036319e-02 -5.05239844e-01 2.15875477e-01
4.18777019e-02 6.29190087e-01 2.56724358e-01 -6.49901748e-01
3.10900480e-01 6.01683855e-01 2.41942734e-01 -1.29827356e+00
-5.81779242e-01 -2.51605451e-01 -8.95691276e-01 -3.74296874e-01
2.65568852e-01 -3.80800366e-01 -6.28204584e-01 1.85148346e+00
4.13672984e-01 8.03075910e-01 -8.74970928e-02 8.55773211e-01
2.99775809e-01 6.75727725e-01 3.10200721e-01 -9.55855131e-01
8.96118402e-01 -1.45448714e-01 -1.10020065e+00 2.15016544e-01
1.77067235e-01 -5.01540661e-01 6.54174924e-01 9.25582722e-02
-6.03562474e-01 -2.22945258e-01 -5.71078360e-01 5.76679170e-01
-1.27664894e-01 -2.15232104e-01 1.08723688e+00 1.19651087e-01
-5.27145982e-01 5.09897649e-01 -1.06029248e+00 -1.02624089e-01
6.58491701e-02 2.35348806e-01 2.26348359e-03 1.02932818e-01
-1.77663398e+00 5.34727693e-01 2.72064686e-01 5.89160398e-02
-1.15844834e+00 -1.30921471e+00 -4.60269928e-01 -1.52756587e-01
5.26374638e-01 -5.71746945e-01 8.71831238e-01 -4.10565704e-01
-9.13835943e-01 -3.42129581e-02 -2.73946673e-01 -4.55177128e-01
5.47717273e-01 -2.98829645e-01 -1.10894608e+00 -1.83693573e-01
4.03490752e-01 -1.64565593e-01 6.14795327e-01 -8.99907827e-01
-7.35723615e-01 -2.25191236e-01 -1.81128412e-01 -2.77757376e-01
-1.26566842e-01 3.11665148e-01 -4.28255163e-02 -1.09492683e+00
8.50902945e-02 -7.75901258e-01 -3.24482262e-01 -5.12798250e-01
-5.20573258e-01 -5.95579505e-01 1.19431365e+00 -4.75316912e-01
2.03588653e+00 -2.04656911e+00 2.07316965e-01 2.36111492e-01
2.61714339e-01 -9.50935185e-02 -5.55367582e-02 8.32035422e-01
-4.64325190e-01 2.07914397e-01 -2.68602014e-01 2.32970431e-01
-3.26682836e-01 5.69053590e-01 -5.42851686e-01 5.47881961e-01
3.98594886e-01 7.72421360e-01 -1.34651256e+00 -4.18803066e-01
4.11976576e-02 2.24505037e-01 -2.40562409e-01 1.83262110e-01
-3.32108051e-01 8.57196271e-01 -6.26183629e-01 6.47330165e-01
3.85760218e-01 -2.33877569e-01 6.72450125e-01 1.87995002e-01
-6.30825937e-01 3.94475132e-01 -1.21760941e+00 8.99316430e-01
-4.60865237e-02 7.00094938e-01 -4.90014702e-01 -1.01761198e+00
8.90303910e-01 5.98780096e-01 9.72526610e-01 -9.49898064e-01
-1.41299382e-01 3.57983299e-02 1.55612931e-01 -5.95598102e-01
-1.41134903e-01 -6.43982086e-03 -1.49586052e-01 5.13199925e-01
-4.03676063e-01 7.08839953e-01 2.76346594e-01 2.55364954e-01
1.63151598e+00 -1.30932942e-01 4.43831861e-01 -4.06576961e-01
2.29790047e-01 1.33418605e-01 1.23931682e+00 4.56873357e-01
-2.63483520e-03 -2.12317538e-02 9.29646134e-01 -6.19119942e-01
-7.82827735e-01 -1.28169215e+00 -2.34355450e-01 5.68724394e-01
4.34920713e-02 -4.85106707e-01 2.18226656e-01 -6.74516439e-01
3.95762801e-01 8.43335092e-01 -8.69015634e-01 -1.51903793e-01
-8.87450695e-01 -1.18647408e+00 2.61195540e-01 7.42362201e-01
9.50605702e-03 -5.53009391e-01 -5.87046683e-01 5.84567964e-01
-2.63887852e-01 -7.02009082e-01 -2.69776225e-01 2.61604279e-01
-9.21427786e-01 -1.46011376e+00 -1.47868097e-01 -1.30738571e-01
3.75876784e-01 1.08832136e-01 1.01726985e+00 -4.11009043e-01
-3.41435254e-01 1.31715357e-01 -3.05686116e-01 -4.38015282e-01
-9.74018946e-02 -3.96284640e-01 3.66852164e-01 -1.54080585e-01
4.00838464e-01 -9.47197080e-01 -6.70992315e-01 5.47304928e-01
-5.47404408e-01 -3.33288461e-01 3.44845533e-01 9.67093050e-01
4.21039224e-01 4.20123339e-01 1.03761387e+00 -6.41906142e-01
8.31114233e-01 -1.09819376e+00 -9.50836539e-01 2.61411697e-01
-1.23064995e+00 -3.27226194e-03 3.90616208e-01 -9.07291770e-01
-1.17996502e+00 -2.72451997e-01 8.84494781e-01 -5.06461203e-01
1.65613353e-01 1.10418630e+00 1.78350300e-01 7.98891783e-01
5.42792082e-01 -2.53412098e-01 -4.93695021e-01 -4.98201758e-01
2.36202851e-01 1.33605972e-01 5.74045837e-01 -6.72330856e-01
7.86288798e-01 6.60665989e-01 3.28105390e-01 -2.05029652e-01
-5.89731634e-01 -4.77357626e-01 -8.18283856e-01 -2.22705811e-01
6.00404263e-01 -8.20962727e-01 -7.41743326e-01 -5.44016659e-02
-1.18168378e+00 -1.72750130e-01 -6.14296235e-02 8.62600565e-01
-7.97435790e-02 -2.49022201e-01 -5.14040351e-01 -1.04321909e+00
7.98025280e-02 -2.17851162e-01 6.01671576e-01 -1.30790770e-01
-7.11397886e-01 -1.43792117e+00 5.75928330e-01 -4.64559138e-01
4.06767651e-02 9.20091867e-01 1.15672898e+00 -3.31812650e-01
-3.30860734e-01 -1.50903076e-01 -2.38819495e-01 -5.17503142e-01
6.05680645e-01 4.14398193e-01 -3.16518813e-01 6.04640581e-02
-3.12046677e-01 5.94623268e-01 5.75094163e-01 7.06634283e-01
6.48749530e-01 -5.43376148e-01 -8.18922639e-01 1.54063568e-01
1.29809093e+00 8.66578102e-01 3.46194059e-01 2.45064408e-01
6.28781378e-01 6.45291865e-01 1.02239025e+00 9.22684371e-01
3.86265129e-01 4.64702547e-01 3.26487660e-01 -8.53737965e-02
1.68765947e-01 -3.79211307e-01 1.86621085e-01 6.59117222e-01
-2.46204674e-01 -3.53469998e-01 -1.09718955e+00 9.91371989e-01
-2.28590465e+00 -1.13063920e+00 -1.20775175e+00 1.85059810e+00
8.94479096e-01 -2.64358036e-02 4.39423382e-01 1.58849895e-01
7.01908529e-01 -1.64191052e-01 -6.73668981e-01 2.17854977e-02
-4.27733839e-01 3.49502265e-02 5.60917437e-01 3.62326652e-01
-1.20636213e+00 5.36977589e-01 7.65671396e+00 1.21822312e-01
-8.76615167e-01 2.06929401e-01 3.29166651e-01 -2.07892597e-01
-3.13203007e-01 3.17450076e-01 -3.83217961e-01 4.61287439e-01
1.13608098e+00 -6.81120992e-01 3.18854712e-02 2.31489912e-01
8.84148240e-01 -2.67689079e-02 -1.11372662e+00 5.28198659e-01
-7.51748681e-01 -1.12247849e+00 -2.29885474e-01 3.74398269e-02
1.14983749e+00 -1.83653668e-01 -1.70406431e-01 -2.20997810e-01
1.04360652e+00 -8.41340423e-01 5.16056180e-01 9.89138782e-01
5.31086504e-01 -6.48008049e-01 4.46322978e-01 3.85513417e-02
-1.33553398e+00 -5.03957272e-01 9.67458710e-02 -5.16709566e-01
3.61867994e-01 9.86857593e-01 -5.92306852e-01 8.58871222e-01
7.51314819e-01 1.23166192e+00 -1.76999494e-01 8.15483212e-01
-3.92204016e-01 1.16390312e+00 -2.90511072e-01 -5.89825884e-02
1.22170255e-01 -1.15641765e-01 8.09992075e-01 9.99137759e-01
2.14546770e-01 5.29246271e-01 1.11038968e-01 8.29330683e-01
4.95420724e-01 -4.71281558e-01 -4.66193318e-01 -2.45820954e-01
9.28577363e-01 4.29762512e-01 -5.50867617e-01 -1.71754777e-01
-3.86754960e-01 5.30484438e-01 -1.37555599e-01 6.12335622e-01
-1.02459908e+00 2.26632684e-01 7.56864548e-01 8.55974257e-02
3.03728376e-02 -6.59421921e-01 -3.59818995e-01 -9.63794231e-01
-1.49059817e-01 -6.53717995e-01 1.12513244e+00 -2.91131169e-01
-1.59355795e+00 1.17235579e-01 5.12275100e-01 -1.37745714e+00
-2.54425585e-01 -3.31339091e-02 -7.72371709e-01 9.22435462e-01
-1.21132410e+00 -8.26398551e-01 3.84855680e-02 9.36020553e-01
3.14328521e-01 1.27044424e-01 6.23765588e-01 5.19102454e-01
-8.00665379e-01 -1.01433285e-01 1.11718401e-01 -1.20005682e-01
5.96397758e-01 -1.25553346e+00 2.85588354e-01 9.79025841e-01
-2.87065536e-01 8.22156608e-01 9.14861500e-01 -1.25648093e+00
-1.47205508e+00 -1.16189241e+00 1.14548433e+00 -2.72034883e-01
1.38237238e+00 -1.64876416e-01 -8.53821754e-01 5.87562859e-01
7.62447864e-02 -8.41056556e-02 6.19892120e-01 5.25686264e-01
-3.78975898e-01 -3.12053096e-02 -6.85043395e-01 5.66038370e-01
1.39768422e+00 -2.76702940e-01 -5.91121912e-01 3.19023132e-01
8.25946569e-01 2.97059089e-01 -1.02801275e+00 6.10981107e-01
4.31124508e-01 -3.73632789e-01 8.57602894e-01 -9.24650908e-01
5.60230553e-01 -3.94485593e-01 2.96040904e-02 -1.22770548e+00
-7.61422157e-01 -8.40269089e-01 -3.82227272e-01 1.37375081e+00
2.05564752e-01 -7.03818083e-01 1.99279934e-01 4.43026602e-01
2.49390587e-01 -2.44958982e-01 -1.04779506e+00 -1.25862789e+00
-1.48956835e-01 -3.71622741e-01 7.63059735e-01 1.76230419e+00
9.37674418e-02 4.89974916e-01 -5.18067718e-01 5.03389597e-01
6.94139361e-01 3.58992189e-01 3.78140330e-01 -1.64303088e+00
-8.39400291e-03 -3.47897619e-01 -2.04726562e-01 -3.59335363e-01
2.09289733e-02 -3.74930620e-01 -3.60579997e-01 -1.33691025e+00
-7.80327152e-03 -7.36176610e-01 -4.42549318e-01 4.91682291e-01
-4.54345584e-01 -5.38500369e-01 -4.60471898e-01 5.50053358e-01
-1.44479722e-01 6.46817505e-01 9.32804167e-01 -1.41158581e-01
-4.58818227e-01 -2.33514793e-02 -2.87410408e-01 3.99038821e-01
7.03945220e-01 -8.25326324e-01 -7.79790819e-01 -1.08026341e-01
3.05563420e-01 5.48570335e-01 6.88783169e-01 -4.26644921e-01
1.90044254e-01 -8.31701577e-01 9.44114476e-02 -9.56216872e-01
-3.77864718e-01 -7.71626234e-01 8.12239289e-01 7.27253854e-01
-3.03471595e-01 5.87563097e-01 2.37614900e-01 1.09179735e+00
-3.20881844e-01 5.38521111e-01 7.07927197e-02 3.26736867e-01
-6.66890264e-01 2.28126183e-01 -4.93494004e-01 -2.31075659e-01
1.12565148e+00 3.23008537e-01 -4.76332128e-01 -1.92570388e-01
-8.10688794e-01 6.43166244e-01 -4.26559538e-01 6.39196336e-01
4.97437686e-01 -1.57751417e+00 -8.79539847e-01 -3.69491756e-01
-1.12336725e-01 -4.24325794e-01 1.40237354e-03 1.16369045e+00
1.64911613e-01 7.41414189e-01 2.71800578e-01 -4.74413037e-01
-1.25960767e+00 9.80668783e-01 7.19978809e-02 -3.93351406e-01
-6.72508478e-01 4.78181273e-01 2.13561878e-01 3.40389609e-02
-8.20238888e-02 -2.51419723e-01 -2.54220515e-01 4.02216971e-01
4.73179042e-01 7.90184736e-01 -4.42837179e-01 -1.28182247e-01
-6.48743033e-01 5.54533228e-02 3.95203799e-01 -2.41403133e-01
1.57834542e+00 -1.82808861e-01 -3.84202003e-01 8.92960131e-01
9.93647754e-01 -1.67513177e-01 -1.31241763e+00 -3.61003548e-01
7.74255276e-01 -6.62158787e-01 -4.48978804e-02 -8.79891396e-01
-8.03533018e-01 2.61300236e-01 4.60872769e-01 5.36520541e-01
1.37470174e+00 6.56205714e-02 3.38782966e-01 -2.60037452e-01
3.00639063e-01 -6.98308229e-01 8.39045830e-03 3.94514173e-01
1.07057118e+00 -9.54008996e-01 1.88307494e-01 -6.43718541e-01
-2.89036036e-01 1.00046396e+00 3.43692571e-01 -9.35956538e-02
1.04865837e+00 4.79875356e-01 -3.04154158e-01 -5.19710004e-01
-1.42737246e+00 -1.93946376e-01 2.74016142e-01 5.09124517e-01
3.59970689e-01 5.09903610e-01 -9.37082708e-01 3.79430592e-01
1.00030489e-01 -6.71158731e-02 -1.05151761e-04 9.76813555e-01
3.79133373e-01 -1.03837121e+00 -5.42199612e-01 3.03202271e-01
-3.60369533e-01 -1.71249360e-01 -6.15895092e-01 1.04096866e+00
4.10527736e-02 1.31297565e+00 3.41050327e-01 -4.35922235e-01
5.20836055e-01 -2.25860879e-01 -4.40920331e-02 -1.73554853e-01
-3.41159672e-01 4.69068319e-01 3.55505198e-01 -6.26577735e-01
-5.77664196e-01 -1.39505422e+00 -1.31584454e+00 -5.85431278e-01
-3.41083795e-01 1.64849624e-01 2.52433479e-01 8.96993995e-01
5.01106083e-01 9.20076311e-01 1.00567341e+00 1.43090814e-01
3.41441780e-02 -8.81447554e-01 -2.95779645e-01 2.06314236e-01
4.24722582e-01 -1.17943859e+00 -4.17965353e-01 2.63209075e-01] | [7.700697898864746, 5.20880126953125] |
65e41d7d-94e1-4f58-b283-46b7106d8936 | holo-dex-teaching-dexterity-with-immersive | 2210.06463 | null | https://arxiv.org/abs/2210.06463v1 | https://arxiv.org/pdf/2210.06463v1.pdf | Holo-Dex: Teaching Dexterity with Immersive Mixed Reality | A fundamental challenge in teaching robots is to provide an effective interface for human teachers to demonstrate useful skills to a robot. This challenge is exacerbated in dexterous manipulation, where teaching high-dimensional, contact-rich behaviors often require esoteric teleoperation tools. In this work, we present Holo-Dex, a framework for dexterous manipulation that places a teacher in an immersive mixed reality through commodity VR headsets. The high-fidelity hand pose estimator onboard the headset is used to teleoperate the robot and collect demonstrations for a variety of general-purpose dexterous tasks. Given these demonstrations, we use powerful feature learning combined with non-parametric imitation to train dexterous skills. Our experiments on six common dexterous tasks, including in-hand rotation, spinning, and bottle opening, indicate that Holo-Dex can both collect high-quality demonstration data and train skills in a matter of hours. Finally, we find that our trained skills can exhibit generalization on objects not seen in training. Videos of Holo-Dex are available at https://holo-dex.github.io. | ['Lerrel Pinto', 'Soumith Chintala', 'Irmak Güzey', 'Sridhar Pandian Arunachalam'] | 2022-10-12 | null | null | null | null | ['mixed-reality'] | ['computer-vision'] | [-4.73853767e-01 1.42706975e-01 -1.29710473e-02 -8.51243176e-03
-2.68831253e-01 -1.03756726e+00 1.89162493e-01 -5.74709892e-01
-7.96229169e-02 5.23953617e-01 -3.98481369e-01 -4.58863020e-01
-3.06504309e-01 -1.88417703e-01 -1.10234523e+00 -3.52926761e-01
-2.71071911e-01 6.83017969e-01 2.28063270e-01 -6.91244781e-01
1.55178428e-01 1.07651246e+00 -1.87090933e+00 -1.74575835e-01
1.11453700e+00 3.58584851e-01 9.81009841e-01 1.05024266e+00
5.60658693e-01 1.00825536e+00 -5.29990315e-01 2.56342381e-01
3.40132236e-01 1.54102564e-01 -6.31537378e-01 -4.95406464e-02
5.34638226e-01 -1.06242287e+00 -7.74309993e-01 8.09848428e-01
3.81441891e-01 5.65856397e-01 6.46951854e-01 -1.56882167e+00
-3.93832088e-01 5.17279148e-01 -1.74374625e-01 -1.87341720e-01
1.00889313e+00 4.33371931e-01 6.72515154e-01 -7.90361524e-01
9.82689083e-01 1.33499575e+00 3.66982132e-01 7.82756567e-01
-7.70156562e-01 -9.64101791e-01 -5.55063933e-02 3.97835486e-02
-8.57602239e-01 -5.73793016e-02 2.85874039e-01 -6.13433957e-01
8.62128198e-01 -2.55904421e-02 9.76856411e-01 1.51263535e+00
4.79994625e-01 1.01330221e+00 9.11962569e-01 -7.40534067e-02
7.17806350e-03 1.35366112e-01 -1.17843993e-01 8.76103461e-01
1.08653054e-01 4.43516165e-01 -3.71553391e-01 3.86255324e-01
1.49545515e+00 3.72862846e-01 -4.44066852e-01 -1.12540948e+00
-1.25704587e+00 4.78847325e-01 5.07119894e-01 -1.00369588e-01
-3.55533212e-01 2.54346550e-01 -5.50342984e-02 8.23240101e-01
-4.86760825e-01 9.89372790e-01 -8.02296996e-01 -9.45031285e-01
3.03595871e-01 4.92663711e-01 1.23106503e+00 1.67883801e+00
3.14504534e-01 -2.68890150e-02 5.57875454e-01 4.74871367e-01
2.06561074e-01 5.46721578e-01 2.27825120e-01 -1.49197674e+00
4.09715474e-01 3.20591420e-01 3.95145595e-01 -3.11960757e-01
-4.77855951e-01 3.34330231e-01 9.27481428e-02 1.06764805e+00
3.64531428e-01 -3.09072673e-01 -9.49107766e-01 1.23086810e+00
9.07647908e-01 3.74061347e-04 7.37583339e-02 1.25991213e+00
1.05695546e+00 2.93862522e-01 -1.65422067e-01 4.11181390e-01
1.10239315e+00 -1.10404372e+00 -6.16939008e-01 1.48205698e-01
8.79186511e-01 -7.66981542e-01 1.56612778e+00 8.87717187e-01
-7.92740822e-01 -5.29629290e-01 -1.02093530e+00 -1.70520186e-01
-1.66080862e-01 -2.23513739e-03 1.02530789e+00 6.12546736e-03
-5.33972561e-01 1.09433484e+00 -1.37109506e+00 -1.94987774e-01
8.77323467e-03 6.06087863e-01 -8.27517271e-01 -4.45022732e-02
-5.42841852e-01 1.43547082e+00 3.59336771e-02 -2.19752729e-01
-1.70574999e+00 -7.98408151e-01 -1.06536102e+00 -2.92777777e-01
6.63571775e-01 -1.51647970e-01 1.91584706e+00 -5.52357221e-03
-2.20401311e+00 4.36677277e-01 7.93518186e-01 3.38910013e-01
8.12578321e-01 -8.51082444e-01 1.12542853e-01 3.88213247e-01
2.97093764e-02 6.94116592e-01 6.89556062e-01 -1.23347199e+00
-4.74369109e-01 -2.06182882e-01 3.30662698e-01 7.23708630e-01
1.65149659e-01 -2.51398683e-01 -1.02134189e-02 -3.31947953e-01
3.23221207e-01 -1.46427965e+00 -1.23496860e-01 4.86876011e-01
-8.74166042e-02 -5.30531824e-01 1.44060278e+00 -4.13358688e-01
-1.30352646e-01 -2.09314561e+00 7.36351848e-01 -2.20131606e-01
4.66020674e-01 3.94765176e-02 1.16447769e-01 4.60471958e-01
6.14683256e-02 -5.89124799e-01 5.27509809e-01 -5.67365699e-02
2.38872573e-01 4.48120385e-01 -1.22470662e-01 5.44911206e-01
-1.99330181e-01 7.53587604e-01 -1.26074183e+00 -3.57754290e-01
7.36310244e-01 5.63371897e-01 -5.14914274e-01 9.30136502e-01
-2.05269650e-01 1.13856816e+00 -5.48989892e-01 8.03625166e-01
2.59591907e-01 1.40474550e-02 2.31039822e-01 3.65212709e-01
-9.55771804e-02 4.95461732e-01 -1.19768655e+00 1.97690797e+00
-9.04454589e-01 7.04357624e-01 6.91894650e-01 -4.77462023e-01
6.55445218e-01 2.96685368e-01 4.38165098e-01 -2.17721194e-01
4.44615602e-01 3.65153551e-01 2.09469154e-01 -1.05848348e+00
6.45348489e-01 2.05240563e-01 4.26446134e-03 3.86938751e-01
4.98346269e-01 -1.49496198e+00 -9.76003855e-02 1.23881511e-01
1.20806110e+00 7.48544812e-01 -8.27613026e-02 -9.75129977e-02
-3.45248222e-01 1.32752150e-01 -5.54071926e-02 4.62864369e-01
-2.07719550e-01 1.82730585e-01 5.07277157e-03 -7.05199018e-02
-1.06652641e+00 -1.40384030e+00 5.32336906e-02 1.34615171e+00
1.98777094e-01 1.01822846e-01 -4.01731163e-01 -5.95179200e-01
5.10721385e-01 5.49715996e-01 -3.47693324e-01 -2.72901338e-02
-5.66903055e-01 8.08931053e-01 1.51602879e-01 7.41746306e-01
4.75944355e-02 -1.41675901e+00 -1.13919234e+00 -5.20610549e-02
4.77216721e-01 -1.15236402e+00 -3.17407101e-01 4.93189722e-01
-9.10855770e-01 -1.22318590e+00 -7.47941434e-01 -1.35186470e+00
6.92386329e-01 6.66839182e-01 6.23136044e-01 -1.50257349e-01
-6.06598616e-01 9.47316945e-01 -6.41280651e-01 -4.96803045e-01
-5.30723393e-01 -9.77129117e-02 4.74233955e-01 -1.33714807e+00
-7.54071474e-02 -1.04621363e+00 -2.21955597e-01 4.84704137e-01
-2.65063107e-01 1.61304623e-02 7.85673976e-01 8.64044189e-01
5.90690710e-02 -4.33154523e-01 2.66962331e-02 -2.72632182e-01
5.14611602e-01 -2.78841078e-01 -8.37280691e-01 -1.76805511e-01
-1.51225477e-01 -1.50471538e-01 5.90861320e-01 -1.15441632e+00
-7.80852497e-01 3.30625236e-01 -1.02259265e-02 -1.05034053e+00
-3.50138426e-01 4.64870036e-02 -6.06789887e-02 -5.51984549e-01
6.61913157e-01 -1.55954748e-01 3.95841330e-01 -4.36142355e-01
4.65329647e-01 7.84979284e-01 6.97952926e-01 -1.02746332e+00
8.77198696e-01 -3.52487247e-03 -1.37366131e-01 -9.22499299e-01
-2.47443780e-01 -2.83457011e-01 -9.26038623e-01 -3.16113591e-01
4.67226624e-01 -1.08941007e+00 -1.78907537e+00 3.77724826e-01
-7.70490944e-01 -1.35253358e+00 -3.89043987e-01 1.20744801e+00
-1.02316356e+00 -8.71824250e-02 -9.01364684e-01 -6.22985721e-01
2.60591269e-01 -1.74472737e+00 1.10548747e+00 3.99511963e-01
-2.57537365e-01 -5.29707372e-01 -7.15014488e-02 3.95634651e-01
-9.65720341e-02 1.08122930e-01 1.60157785e-01 -4.39455867e-01
-5.66084027e-01 -1.01407021e-01 2.31443450e-01 5.93105815e-02
9.71970782e-02 8.56949612e-02 -5.86034656e-01 -6.12205029e-01
-1.82332844e-01 -1.21559560e+00 -2.32221723e-01 -9.75807384e-02
1.09409523e+00 2.51199901e-02 -2.02885121e-01 5.03573060e-01
6.76312387e-01 5.48047125e-01 2.14908481e-01 1.62705004e-01
1.07782805e+00 6.04879618e-01 1.22819293e+00 2.09810525e-01
4.29400086e-01 6.86021507e-01 5.93302488e-01 3.19410533e-01
6.74772682e-03 -4.45714653e-01 4.37763840e-01 9.49749351e-01
-2.30502635e-01 1.84242576e-01 -8.21201980e-01 3.61066967e-01
-1.41386688e+00 -5.11880457e-01 -1.00862579e-02 1.75336039e+00
7.85359263e-01 -2.14783922e-01 1.99283436e-02 -8.44961107e-02
1.54604867e-01 -5.37184298e-01 -8.33385408e-01 -4.70427036e-01
8.01297605e-01 2.32807323e-01 4.03740644e-01 4.82275456e-01
-7.62521505e-01 9.97043371e-01 5.61234522e+00 1.16305374e-01
-1.16451204e+00 -1.29612759e-01 -6.08574331e-01 -5.44493347e-02
2.26222590e-01 -2.60542989e-01 -4.19465750e-01 1.67597115e-01
2.76788384e-01 2.50466168e-01 6.74151242e-01 1.56403279e+00
-6.85260370e-02 -2.13320822e-01 -1.58177590e+00 9.88981605e-01
-2.21651211e-01 -5.66206634e-01 -6.10697389e-01 1.55247182e-01
8.04792881e-01 4.61995043e-02 2.44274154e-01 8.14762473e-01
1.14237225e+00 -1.13701761e+00 5.24254441e-01 -1.07657574e-01
9.26202416e-01 -5.06718695e-01 8.80474597e-02 6.15406215e-01
-9.03513312e-01 -3.53078425e-01 -1.42111331e-01 -4.70045030e-01
-1.04917541e-01 -3.33566189e-01 -1.37569642e+00 2.28230059e-02
9.04140592e-01 6.68266058e-01 1.41071483e-01 8.43867242e-01
-7.32136488e-01 -2.69868057e-02 -3.49238873e-01 -5.16169786e-01
1.92284182e-01 -2.07224637e-01 6.94282353e-01 4.57778066e-01
2.18894333e-01 5.63887656e-01 3.30635548e-01 6.62126482e-01
-1.18314363e-01 -4.08639073e-01 -1.06651688e+00 -7.36228377e-02
5.38036466e-01 1.28182709e+00 -4.49143380e-01 -1.89790521e-02
1.11602180e-01 1.10244226e+00 4.23697144e-01 5.09327427e-02
-9.99074399e-01 -7.75771260e-01 9.51910079e-01 -3.42166215e-01
3.36600959e-01 -1.06031847e+00 2.94495940e-01 -1.12124276e+00
4.81681619e-03 -1.19205689e+00 -4.22320098e-01 -1.14830089e+00
-6.40611649e-01 5.15160114e-02 2.04519451e-01 -1.65388167e+00
-3.78797978e-01 -9.72224951e-01 -4.63025212e-01 3.41709644e-01
-8.83575320e-01 -7.80454099e-01 -9.95502889e-01 5.43125808e-01
9.11469400e-01 -2.57341266e-02 6.77820325e-01 2.50837319e-02
-2.76396006e-01 3.27223212e-01 -5.98048605e-02 -8.33860338e-02
7.30286539e-01 -1.46120131e+00 2.74348289e-01 2.50657424e-02
-1.20920025e-01 7.36780405e-01 1.06932616e+00 -5.54862201e-01
-2.40292215e+00 -3.00154269e-01 -2.79738009e-01 -8.49902153e-01
8.46091628e-01 -5.40634632e-01 -7.21256196e-01 1.25196278e+00
-1.77925766e-01 1.24811523e-01 -5.15077030e-03 -1.23299705e-02
-1.46136552e-01 3.79448205e-01 -1.26544881e+00 7.53209710e-01
1.35916281e+00 -4.87301826e-01 -1.09866393e+00 6.43534124e-01
9.29320574e-01 -1.43183744e+00 -1.13528574e+00 2.07030952e-01
1.11934614e+00 -3.15716654e-01 8.00248563e-01 -6.17097557e-01
7.03639328e-01 -1.21477515e-01 2.33617183e-02 -1.77481878e+00
3.43944207e-02 -9.26572502e-01 -2.44044647e-01 3.47288370e-01
-2.92098075e-02 -7.13086307e-01 7.85114884e-01 3.88668746e-01
-5.67750871e-01 -7.12595642e-01 -4.41699207e-01 -1.11591280e+00
1.56089306e-01 -1.28536776e-01 4.29870009e-01 1.10150087e+00
8.34653676e-01 4.58961464e-02 3.71121876e-02 4.46256936e-01
4.57253575e-01 7.07880631e-02 1.52006876e+00 -1.35953152e+00
-4.42687571e-01 -1.02517091e-01 -3.14388245e-01 -1.52882147e+00
4.30117548e-01 -5.14301956e-01 5.40446639e-01 -1.33972573e+00
-2.43122444e-01 -7.96301425e-01 7.35204577e-01 3.93304110e-01
1.84957549e-01 -3.62911165e-01 3.11682254e-01 2.24643499e-01
-4.71219182e-01 6.28115952e-01 2.14628291e+00 2.56500274e-01
-6.16958678e-01 3.29765528e-02 4.91364785e-02 7.31895804e-01
7.55084038e-01 -2.87235886e-01 -6.00547194e-01 -4.23550963e-01
-1.95345849e-01 5.79955935e-01 3.37203503e-01 -1.11207867e+00
3.00910205e-01 -5.17898560e-01 3.59250814e-01 -2.96210617e-01
7.47751772e-01 -1.09972656e+00 5.94971813e-02 7.89318025e-01
-1.84193432e-01 3.99244390e-02 2.78438121e-01 7.43177384e-02
2.88985670e-01 -3.63098048e-02 4.64652032e-01 -3.03624809e-01
-7.91946590e-01 8.01297128e-02 -4.32143420e-01 -1.13678716e-01
1.54333901e+00 -2.24952325e-01 -4.40312594e-01 -5.64772606e-01
-8.39589059e-01 6.25139892e-01 6.12034738e-01 6.38608634e-01
8.29265833e-01 -7.92650104e-01 -1.01385377e-01 2.83727467e-01
-7.47735649e-02 6.52203918e-01 1.92575634e-01 7.29504526e-01
-1.00947082e+00 1.34495780e-01 -6.84180737e-01 -8.35760593e-01
-1.36138785e+00 4.65649575e-01 2.95836985e-01 4.80768889e-01
-1.06417978e+00 1.02502823e+00 3.41580026e-02 -1.32585049e+00
6.49239004e-01 -7.38641202e-01 1.36472628e-01 -5.49192309e-01
1.10499941e-01 5.81735611e-01 -2.97532886e-01 -1.25418097e-01
1.27574811e-02 4.32784259e-01 1.65089536e-02 -1.45656109e-01
1.50393224e+00 3.18782330e-01 3.86984169e-01 5.52143991e-01
9.53916728e-01 -8.87170807e-02 -1.85335481e+00 4.87185031e-01
-5.79220355e-01 -6.09293282e-01 -3.91669929e-01 -7.60242403e-01
-6.36613488e-01 8.48386168e-01 3.48232359e-01 -2.62754589e-01
3.14589590e-01 1.92412451e-01 5.01583815e-01 1.23010540e+00
1.04307902e+00 -1.06538701e+00 7.40990937e-01 7.63117731e-01
1.29471493e+00 -1.38058627e+00 4.44942601e-02 -6.19959235e-01
-7.05524385e-01 1.12575531e+00 1.36313426e+00 -4.62892592e-01
5.07849813e-01 5.99896491e-01 1.59011841e-01 -4.52337176e-01
-4.99474198e-01 2.41152272e-01 -9.57273990e-02 8.43642890e-01
9.66119692e-02 4.34201151e-01 3.87985438e-01 3.23524401e-02
-7.39968419e-01 -7.40735903e-02 1.00734651e+00 1.58417940e+00
-6.07342243e-01 -5.56368053e-01 -3.60042840e-01 1.98064283e-01
1.37583598e-01 6.63846254e-01 -1.98578209e-01 1.26159084e+00
-3.96454573e-01 7.33308196e-01 -1.02302827e-01 -6.07441783e-01
7.78059185e-01 -3.70536953e-01 1.21916175e+00 -9.98471081e-01
-6.29855156e-01 -4.32149678e-01 -1.89241931e-01 -9.89875853e-01
3.89718294e-01 -5.86236775e-01 -1.63033938e+00 -4.49600250e-01
-5.00696659e-01 -1.46854566e-02 1.05138624e+00 5.58029890e-01
2.43592504e-02 7.35908985e-01 6.77467048e-01 -1.72493768e+00
-1.20880783e+00 -1.19294918e+00 -7.31704533e-01 2.37050101e-01
6.16062701e-01 -1.49763703e+00 -4.60936964e-01 -5.24286866e-01] | [4.710915565490723, 0.6470305323600769] |
144a26c1-267f-4b5b-ab66-ad9c78b521d2 | modularity-based-linkage-model-for | 2306.01227 | null | https://arxiv.org/abs/2306.01227v1 | https://arxiv.org/pdf/2306.01227v1.pdf | Modularity based linkage model for neuroevolution | Crossover between neural networks is considered disruptive due to the strong functional dependency between connection weights. We propose a modularity-based linkage model at the weight level to preserve functionally dependent communities (building blocks) in neural networks during mixing. A proximity matrix is built by estimating the dependency between weights, then a community detection algorithm maximizing modularity is run on the graph described by such matrix. The resulting communities/groups of parameters are considered to be mutually independent and used as crossover masks in an optimal mixing EA. A variant is tested with an operator that neutralizes the permutation problem of neural networks to a degree. Experiments were performed on 8 and 10-bit parity problems as the intrinsic hierarchical nature of the dependencies in these problems are challenging to learn. The results show that our algorithm finds better, more functionally dependent linkage which leads to more successful crossover and better performance. | ['Marcus Gallagher', 'Yukai Qiao'] | 2023-06-02 | null | null | null | null | ['community-detection'] | ['graphs'] | [ 5.64751148e-01 4.20334935e-01 -1.50146961e-01 -4.76908386e-02
3.09552193e-01 -7.01135635e-01 2.83548623e-01 2.17753559e-01
-3.35249782e-01 9.26802516e-01 -8.94749351e-03 -2.64036655e-01
-7.52549827e-01 -8.48879814e-01 -9.83758211e-01 -9.42004323e-01
-7.80954897e-01 4.40744102e-01 3.86847377e-01 -2.83641964e-01
3.99895698e-01 4.32252765e-01 -1.74351299e+00 3.81174594e-01
1.20134902e+00 8.75000805e-02 1.84292048e-01 7.13879526e-01
5.81239117e-03 3.71694952e-01 -6.44453049e-01 -4.89283472e-01
4.33401406e-01 -6.44482255e-01 -6.27979577e-01 -8.58003497e-02
2.78703779e-01 6.22485459e-01 -8.67332704e-03 1.48753071e+00
3.29338014e-01 -2.25751519e-01 7.19009995e-01 -1.69470608e+00
-4.54714745e-01 1.33286810e+00 -7.49587357e-01 2.19572872e-01
-1.44938529e-01 -2.36082152e-01 1.19872558e+00 -2.58160442e-01
7.60933220e-01 1.21536434e+00 6.70562804e-01 5.69390953e-01
-2.01247263e+00 -8.25027168e-01 5.99250570e-02 5.96528351e-02
-1.41909122e+00 -1.35666296e-01 8.16620708e-01 -3.94513249e-01
8.81236017e-01 4.39517021e-01 1.17452121e+00 7.78126061e-01
5.51310897e-01 1.57474041e-01 7.32950628e-01 -4.27212000e-01
3.86235625e-01 -5.53595759e-02 2.06257757e-02 7.10056186e-01
9.75294232e-01 -1.49688840e-01 -6.68687105e-01 -3.37581158e-01
7.62538493e-01 -3.55284214e-01 -4.18457448e-01 -1.04043198e+00
-1.13596869e+00 7.81938612e-01 7.39215374e-01 5.64536273e-01
-1.46201387e-01 1.68990701e-01 1.67672232e-01 8.92999291e-01
-8.64446089e-02 6.11041367e-01 -2.32260734e-01 3.52837116e-01
-7.59782672e-01 -7.57202134e-02 9.44929242e-01 7.23175585e-01
6.98420644e-01 -3.40814084e-01 1.40499428e-01 7.00279057e-01
3.95093143e-01 -1.72049090e-01 2.57062227e-01 -9.46561754e-01
3.65669519e-01 9.38116372e-01 -7.03254759e-01 -1.26657355e+00
-3.88876706e-01 -7.77788579e-01 -1.21693623e+00 4.29124147e-01
3.05660427e-01 -6.27018511e-02 -5.82752883e-01 2.34362268e+00
3.25239688e-01 1.61881641e-01 1.84488371e-01 2.94650495e-01
2.99430609e-01 3.75280082e-01 -3.30522686e-01 -3.21137249e-01
1.25111663e+00 -5.26607633e-01 -4.51014936e-01 -1.04873613e-01
3.91286790e-01 -2.92374641e-01 2.15203002e-01 2.21786633e-01
-1.17436242e+00 -3.07530463e-01 -1.48627126e+00 4.61869627e-01
-2.88497984e-01 -2.11448789e-01 6.23124123e-01 8.39269042e-01
-1.65925395e+00 8.41805160e-01 -7.25785553e-01 -5.00533581e-01
3.38105649e-01 9.82265532e-01 -5.16240060e-01 2.65836656e-01
-1.03313482e+00 6.67318285e-01 8.38448942e-01 7.73568153e-02
-5.41485965e-01 -5.35037577e-01 -4.62368429e-01 2.30062515e-01
-1.40318394e-01 -7.27153897e-01 9.27750170e-02 -1.23776889e+00
-8.80769372e-01 8.63385737e-01 4.12367612e-01 -4.71563905e-01
4.26940739e-01 6.82567775e-01 -1.44356176e-01 1.46879420e-01
-2.08347499e-01 1.36309147e+00 6.26631975e-01 -1.43716621e+00
-6.02442026e-01 -1.51819572e-01 -1.95845030e-02 1.33067206e-01
-5.15090227e-01 -1.72454953e-01 -9.00677368e-02 -3.55233252e-01
6.95139825e-01 -1.13662291e+00 -6.66205212e-02 -1.85265075e-02
-7.29461133e-01 9.39267948e-02 2.83577055e-01 -3.80065292e-01
1.12788594e+00 -2.00449657e+00 8.54028165e-01 8.13747823e-01
5.33808827e-01 -1.27342448e-01 -4.48523045e-01 5.91352105e-01
-7.54793167e-01 2.59922653e-01 -5.12448132e-01 2.53487080e-01
-3.88913870e-01 2.73085207e-01 2.53267437e-01 5.08816779e-01
3.57075602e-01 4.03925985e-01 -6.17665350e-01 -4.53769177e-01
-5.74593544e-01 4.50457931e-01 -1.01484644e+00 -2.13944942e-01
8.79700854e-02 1.73670158e-01 2.36837212e-02 3.15004945e-01
7.69454956e-01 -2.83499081e-02 8.89454126e-01 -8.07912871e-02
-3.38245966e-02 -1.30459577e-01 -1.35837412e+00 1.59235156e+00
4.23657447e-01 7.05393553e-01 1.74325556e-01 -1.28954697e+00
9.59511459e-01 1.96690187e-01 4.71558422e-01 2.51829084e-02
1.67432487e-01 -1.22290805e-01 9.84344125e-01 -2.33673990e-01
5.75880799e-03 2.90717751e-01 3.11009347e-01 2.36555070e-01
2.00311080e-01 2.46615916e-01 6.88934445e-01 4.78318214e-01
1.51639497e+00 -3.19841713e-01 1.18282445e-01 -9.63822484e-01
4.03632611e-01 -1.64857671e-01 8.94546211e-01 6.56397998e-01
-1.59170493e-01 4.58290815e-01 1.36254191e+00 6.39984086e-02
-1.16925395e+00 -1.10394490e+00 -2.11399049e-01 6.52870715e-01
-9.46134254e-02 -2.38241196e-01 -1.04517281e+00 -3.36439073e-01
4.27768305e-02 2.13025838e-01 -9.63564336e-01 -6.98357224e-01
-4.59883928e-01 -1.07440352e+00 7.37320423e-01 3.80841605e-02
3.03435832e-01 -8.90863657e-01 -4.65906173e-01 3.66661549e-02
-4.88134511e-02 -2.17983022e-01 -1.77946001e-01 9.09418821e-01
-1.46314895e+00 -1.15009904e+00 -4.14884239e-01 -1.33137119e+00
1.24680352e+00 -5.60017303e-02 8.50417376e-01 4.73283082e-01
-2.35598892e-01 -2.90062487e-01 8.95562768e-02 6.97166324e-02
-6.04921162e-01 2.25788608e-01 7.12834001e-02 -5.80145698e-03
-1.32803038e-01 -1.28246009e+00 -2.82546461e-01 3.09296191e-01
-1.15069115e+00 6.95083439e-02 7.03345358e-01 9.56979632e-01
4.43852991e-02 5.37789404e-01 6.45134807e-01 -7.03778088e-01
9.15007532e-01 -5.01429677e-01 -5.48065484e-01 5.50735354e-01
-5.96301198e-01 5.68664908e-01 1.86075240e-01 -7.07951367e-01
-7.01792657e-01 7.03996196e-02 4.20990944e-01 1.23325698e-02
1.65255487e-01 6.07481003e-01 -5.04437447e-01 -2.83518076e-01
8.70643497e-01 -1.75362155e-01 1.87429070e-01 -1.45194039e-01
1.40419185e-01 2.13606939e-01 2.89079398e-01 -5.84431648e-01
8.50577116e-01 2.04406559e-01 3.09932262e-01 -5.91978788e-01
2.35737205e-01 1.21457063e-01 -8.93777370e-01 -2.18676135e-01
6.33690953e-01 -5.90790868e-01 -6.84700489e-01 1.88255981e-01
-1.01758158e+00 9.83581543e-02 -9.87677947e-02 3.43168467e-01
-2.39700049e-01 2.77179390e-01 -6.46398008e-01 -7.79334009e-01
-1.38701610e-02 -8.87373805e-01 2.82222852e-02 2.01991946e-01
-2.03407824e-01 -8.21502864e-01 4.80658770e-01 -5.03177522e-03
3.13847438e-02 2.55089134e-01 1.56569982e+00 -6.16251171e-01
-6.55272961e-01 1.40843093e-01 2.63045654e-02 -8.74689023e-04
-1.34165483e-02 6.96699083e-01 -5.95010161e-01 -5.32806098e-01
-3.31665665e-01 1.25856131e-01 1.09559977e+00 4.42117214e-01
4.08714235e-01 -2.05458790e-01 -6.93931937e-01 3.39594722e-01
1.47486472e+00 4.22217965e-01 7.44896233e-01 2.84426510e-01
5.15969336e-01 9.93920743e-01 -1.95028797e-01 -1.45693451e-01
-2.04513192e-01 2.46428460e-01 5.90024948e-01 5.12180217e-02
9.49300528e-02 -4.51747179e-02 3.31683874e-01 1.17391062e+00
-2.04777732e-01 -2.54968882e-01 -8.38159204e-01 5.06917059e-01
-1.99933505e+00 -1.06279480e+00 -2.47066453e-01 2.19692802e+00
1.30827689e+00 4.25382107e-01 1.08889207e-01 4.85129088e-01
1.37453008e+00 -3.36211652e-01 -3.41166824e-01 -2.85015196e-01
-3.71311575e-01 3.12540606e-02 3.48658830e-01 3.53239715e-01
-7.93398499e-01 5.23477972e-01 6.49279213e+00 5.66813111e-01
-6.19841576e-01 -9.13653076e-02 4.53618586e-01 -2.58616537e-01
-4.54914957e-01 2.12620676e-01 -4.15073246e-01 3.43006283e-01
6.32926464e-01 1.31186962e-01 6.53668165e-01 2.62186646e-01
-3.18196177e-01 -2.11347863e-02 -1.07900083e+00 5.18270135e-01
-1.28948495e-01 -1.13705301e+00 -3.24190147e-02 3.04918617e-01
9.26730335e-01 -2.79598664e-02 -7.74563625e-02 -2.42668465e-01
6.23162389e-01 -7.10933328e-01 6.64687157e-01 4.44530338e-01
3.49610209e-01 -1.30687654e+00 3.10334682e-01 3.00721169e-01
-1.08479154e+00 -3.57985705e-01 -8.11825752e-01 1.08214885e-01
-3.44565123e-01 8.18208218e-01 -5.71379066e-01 2.54293084e-01
6.05081081e-01 5.02802432e-01 -7.54934788e-01 1.52691162e+00
-1.93442360e-01 2.32396290e-01 -2.14220837e-01 -9.24469307e-02
-3.07572037e-01 -6.63559139e-01 7.74558544e-01 8.42774510e-01
2.85467148e-01 -4.29757297e-01 -3.32272559e-01 1.10876811e+00
-2.56719917e-01 -1.51470497e-01 -8.63267422e-01 -1.16382107e-01
5.03622591e-01 1.38880360e+00 -1.57585609e+00 9.50436220e-02
1.53989613e-01 8.29829216e-01 7.15830386e-01 2.47541517e-01
-8.36148143e-01 -3.92347038e-01 9.04593408e-01 -3.14259797e-01
4.32794690e-01 -4.02362980e-02 -2.26148903e-01 -9.91304100e-01
-1.76218241e-01 -7.47972369e-01 3.85800004e-01 -4.43697214e-01
-1.11680615e+00 5.39476573e-01 -2.86363997e-02 -9.30400014e-01
3.67034227e-01 -3.05833042e-01 -7.62456179e-01 4.38471347e-01
-8.26529980e-01 -5.29204130e-01 1.86554134e-01 3.63422394e-01
-2.14757919e-01 -2.51411438e-01 5.82331836e-01 3.59132618e-01
-8.81397188e-01 4.54766393e-01 1.72298163e-01 -3.86725599e-03
3.79885674e-01 -1.22381437e+00 2.81696588e-01 9.85267758e-01
3.20099145e-01 1.20764339e+00 7.31184781e-01 -1.10339916e+00
-1.09621775e+00 -7.57561982e-01 1.00889921e+00 -1.05494916e-01
6.06411815e-01 -7.33348429e-01 -8.71974826e-01 4.09878492e-01
6.46001041e-01 -5.03715038e-01 7.90689111e-01 6.42731637e-02
-3.34855825e-01 -2.22562909e-01 -1.19658887e+00 7.55668342e-01
1.57107818e+00 -1.59840912e-01 -5.28794169e-01 1.07688285e-01
6.40420079e-01 3.15105498e-01 -7.61142671e-01 2.18768314e-01
5.71117282e-01 -1.12299609e+00 1.06073952e+00 -6.48697555e-01
4.82454598e-01 -5.02981901e-01 2.60371745e-01 -1.50948930e+00
-7.28695750e-01 -5.11883616e-01 3.18935752e-01 1.49717081e+00
8.77133191e-01 -7.50931561e-01 8.44765425e-01 -2.86268312e-02
2.80732512e-01 -2.97333390e-01 -1.13880396e+00 -7.38638341e-01
-4.28485163e-02 3.98259699e-01 2.99018174e-01 1.39994431e+00
2.40193829e-01 4.62459415e-01 -1.18297609e-02 1.02568999e-01
9.07412350e-01 -2.27322429e-01 1.34700745e-01 -1.82009542e+00
-3.17888319e-01 -7.84292877e-01 -7.52712488e-01 -1.29471883e-01
1.20215476e-01 -1.34736133e+00 -2.83484180e-02 -1.17283010e+00
4.51999068e-01 -3.92202944e-01 -4.50774640e-01 5.45282006e-01
1.51563570e-01 2.48192593e-01 -4.61916700e-02 1.95129052e-01
-1.84693381e-01 2.05130741e-01 7.45321751e-01 -3.14289600e-01
-3.07547092e-01 -1.66988760e-01 -7.85436273e-01 4.30939615e-01
9.35421050e-01 -9.78827715e-01 -6.72799528e-01 -1.87537804e-01
7.60551989e-01 -5.29320948e-02 1.93930864e-02 -1.40619349e+00
5.36424160e-01 1.33179173e-01 5.04690886e-01 -3.25560749e-01
4.20287177e-02 -1.07758415e+00 8.94015491e-01 1.13650334e+00
-7.90134668e-01 3.97988081e-01 -2.45045796e-02 6.36612058e-01
1.64269567e-01 -4.22748238e-01 7.99066186e-01 1.92301095e-01
5.86967468e-02 -1.26528680e-01 -4.25220400e-01 -2.14532062e-01
1.11285496e+00 -5.07450402e-01 -3.98839176e-01 7.07476586e-02
-7.30236471e-01 8.47731680e-02 4.60755020e-01 2.35166669e-01
4.07477796e-01 -1.31834757e+00 -7.12161481e-01 4.77656722e-01
-6.80285245e-02 -4.34356332e-01 -1.20761335e-01 8.53959680e-01
-5.99743009e-01 -5.97619563e-02 -9.88269389e-01 -6.21677518e-01
-1.73747122e+00 5.41985393e-01 3.42960864e-01 -4.65784110e-02
-1.98623970e-01 1.16440797e+00 -2.66347751e-02 -4.33892697e-01
5.00154674e-01 -4.02265757e-01 -5.72600186e-01 4.21789199e-01
1.18131386e-02 2.51471847e-01 3.04973274e-02 -2.66175896e-01
-3.41544241e-01 3.38392496e-01 5.24287904e-03 -1.72607437e-01
1.48378313e+00 8.78209472e-02 -8.07416499e-01 1.64796337e-01
1.00895894e+00 -3.59032340e-02 -1.17345631e+00 1.86034188e-01
2.53912002e-01 -2.02470124e-01 -3.48372638e-01 -4.62883711e-01
-1.49082553e+00 4.57026035e-01 6.14749372e-01 5.83248973e-01
1.06583643e+00 -1.57999724e-01 -2.04531282e-01 5.34656048e-01
3.55235897e-02 -8.55105937e-01 -5.34939356e-02 2.76020139e-01
6.61964715e-01 -4.17890251e-01 -1.07194528e-01 -4.63802248e-01
-1.06573664e-01 1.06936121e+00 6.07674301e-01 -1.60722777e-01
5.51086545e-01 5.42663276e-01 -4.37893271e-01 -1.97079152e-01
-9.50296402e-01 6.30776361e-02 1.33263096e-01 8.10080469e-01
2.56874621e-01 -2.00371087e-01 -5.24524629e-01 -7.62434560e-04
-1.69714496e-01 -3.95888418e-01 5.08395851e-01 1.03791499e+00
-6.09406352e-01 -1.42559624e+00 -3.44231695e-01 2.67892987e-01
-2.13767663e-02 -1.00416057e-01 -9.20920491e-01 5.30016184e-01
4.09851164e-01 8.85587633e-01 1.70760900e-01 -5.13224959e-01
1.95800755e-02 1.16057068e-01 6.25479400e-01 -3.36272985e-01
-9.48513985e-01 -2.14855112e-02 9.80926156e-02 -2.09623903e-01
-4.32840496e-01 -6.91093743e-01 -1.09792650e+00 -5.22431314e-01
-6.30275249e-01 4.29787546e-01 6.40771031e-01 4.05599385e-01
4.26016688e-01 8.92507374e-01 4.94733125e-01 -3.26495498e-01
-9.49363336e-02 -6.38768077e-01 -7.87087321e-01 1.23226404e-01
-2.59684026e-02 -6.17750764e-01 -4.84590679e-01 -6.98158070e-02] | [7.016350746154785, 5.528370380401611] |
64f76aa9-afcd-49dc-93c0-51d42f9d59fc | human-semantic-segmentation-using-millimeter | 2304.14132 | null | https://arxiv.org/abs/2304.14132v2 | https://arxiv.org/pdf/2304.14132v2.pdf | Human Semantic Segmentation using Millimeter-Wave Radar Sparse Point Clouds | This paper presents a framework for semantic segmentation on sparse sequential point clouds of millimeter-wave radar. Compared with cameras and lidars, millimeter-wave radars have the advantage of not revealing privacy, having a strong anti-interference ability, and having long detection distance. The sparsity and capturing temporal-topological features of mmWave data is still a problem. However, the issue of capturing the temporal-topological coupling features under the human semantic segmentation task prevents previous advanced segmentation methods (e.g PointNet, PointCNN, Point Transformer) from being well utilized in practical scenarios. To address the challenge caused by the sparsity and temporal-topological feature of the data, we (i) introduce graph structure and topological features to the point cloud, (ii) propose a semantic segmentation framework including a global feature-extracting module and a sequential feature-extracting module. In addition, we design an efficient and more fitting loss function for a better training process and segmentation results based on graph clustering. Experimentally, we deploy representative semantic segmentation algorithms (Transformer, GCNN, etc.) on a custom dataset. Experimental results indicate that our model achieves mean accuracy on the custom dataset by $\mathbf{82.31}\%$ and outperforms the state-of-the-art algorithms. Moreover, to validate the model's robustness, we deploy our model on the well-known S3DIS dataset. On the S3DIS dataset, our model achieves mean accuracy by $\mathbf{92.6}\%$, outperforming baseline algorithms. | ['Han Cheng', 'Luoyu MEI', 'Pengfei Song'] | 2023-04-27 | null | null | null | null | ['graph-clustering'] | ['graphs'] | [ 7.97470436e-02 -4.68648337e-02 1.63424745e-01 -4.02202606e-01
-6.97922707e-01 -4.11006629e-01 1.81947455e-01 -1.82674780e-01
-2.88584560e-01 4.89964306e-01 -3.74016970e-01 -3.16560000e-01
-4.39762414e-01 -1.11053610e+00 -5.66027641e-01 -7.53239095e-01
-3.86172861e-01 6.09921753e-01 2.91961819e-01 2.04489022e-01
2.12006181e-01 7.24040806e-01 -1.21696115e+00 -3.36940914e-01
1.02501166e+00 1.44155967e+00 7.75814876e-02 5.45248017e-02
-1.87388867e-01 -7.90578872e-02 -5.33491850e-01 -4.64643240e-01
7.60565639e-01 -1.96141317e-01 -2.80660361e-01 1.77764252e-01
5.16252220e-01 -8.90141726e-02 -3.81972134e-01 1.32411766e+00
4.17389214e-01 -4.97249290e-02 4.89980906e-01 -1.48174369e+00
-2.23588839e-01 1.39703780e-01 -8.63746345e-01 -1.24218576e-01
1.80500090e-01 -8.79708081e-02 9.18814063e-01 -6.17157459e-01
5.44360638e-01 1.03745246e+00 7.80673206e-01 2.01485038e-01
-8.74231458e-01 -1.26345599e+00 1.85405359e-01 5.58094233e-02
-1.69858360e+00 -4.32958305e-02 8.57961774e-01 -3.02552640e-01
3.89203787e-01 3.65675658e-01 8.80818367e-01 6.31035030e-01
6.81058988e-02 5.37389874e-01 9.69299972e-01 4.83122990e-02
4.20620561e-01 -1.32078201e-01 1.33941874e-01 9.16338086e-01
4.73905772e-01 2.18671754e-01 -1.86918080e-01 -2.75573403e-01
7.76352167e-01 3.69651169e-01 -3.26604843e-01 -5.50394475e-01
-9.63973880e-01 7.67551541e-01 7.33775675e-01 2.43802294e-01
-2.61515975e-01 3.26829463e-01 -1.73511654e-01 2.02599525e-01
4.16086346e-01 2.51873642e-01 -3.42079222e-01 1.67313442e-01
-1.29422820e+00 1.88020140e-01 7.08584428e-01 1.51977706e+00
1.07871485e+00 -1.68147877e-01 4.90563624e-02 3.50684792e-01
5.25286376e-01 8.98230314e-01 -1.59960940e-01 -6.62483037e-01
6.14543915e-01 8.33545089e-01 -1.19876042e-01 -1.34506345e+00
-6.55537724e-01 -8.54469121e-01 -1.08800471e+00 -1.00847952e-01
1.49713814e-01 -1.40806735e-01 -1.15936291e+00 1.44043446e+00
2.48274967e-01 4.80148077e-01 -2.31011547e-02 1.00236177e+00
8.34416986e-01 4.63552624e-01 -1.71710864e-01 -3.23810786e-01
1.37308407e+00 -5.00304282e-01 -4.48162317e-01 -2.26707265e-01
4.14663702e-01 -4.73090827e-01 6.18407249e-01 3.26036602e-01
-7.18722820e-01 -2.04102144e-01 -1.07626534e+00 3.91488731e-01
-2.11492538e-01 3.67791904e-03 7.55306423e-01 7.89700091e-01
-8.33715200e-01 2.71641523e-01 -9.17964816e-01 -4.87480462e-01
8.90233576e-01 4.55344707e-01 2.73044910e-02 -2.49135554e-01
-8.54271114e-01 1.62005797e-01 8.02711472e-02 4.74735461e-02
-5.25981486e-01 -8.76613319e-01 -6.45907640e-01 9.36267674e-02
3.25805604e-01 -7.42824316e-01 6.38234138e-01 -3.50141287e-01
-1.01315117e+00 5.57752848e-01 -7.99498931e-02 -4.96586919e-01
5.44848979e-01 2.03999840e-02 -3.93548906e-01 5.15500724e-01
1.52816504e-01 5.50329983e-01 7.08133101e-01 -1.43271935e+00
-7.34968781e-01 -6.71742141e-01 -5.78108691e-02 4.02538758e-03
-1.23152390e-01 -3.83373111e-01 -5.89738607e-01 -5.91539741e-01
8.35868537e-01 -1.06871712e+00 -5.54647267e-01 6.73799291e-02
-6.34213150e-01 1.46941826e-01 1.08020937e+00 -3.57877970e-01
9.96887028e-01 -2.34250689e+00 -2.04358339e-01 8.12880039e-01
4.21257615e-01 1.14239175e-02 2.08087772e-01 2.68116534e-01
2.65466511e-01 1.31344348e-01 -6.51936352e-01 -2.38243252e-01
3.23260948e-02 4.25500423e-02 -9.23278481e-02 5.51849782e-01
-3.40585023e-01 9.27772343e-01 -6.50514603e-01 -4.67490703e-01
2.70974278e-01 1.08341411e-01 -6.17275357e-01 -1.20186903e-01
-2.17557028e-01 4.56353307e-01 -9.29442763e-01 1.11940515e+00
1.19443250e+00 -3.19269866e-01 -1.25983715e-01 -4.24055845e-01
-5.24077415e-02 -3.88520509e-01 -1.18459237e+00 1.75881374e+00
-8.22822601e-02 1.95494354e-01 2.18630642e-01 -1.14141321e+00
1.33038735e+00 1.27073422e-01 9.40793157e-01 -8.18179965e-01
2.21042335e-01 1.45097300e-01 -2.43859634e-01 -2.80859053e-01
9.32083875e-02 -1.55413866e-01 -3.02043766e-01 3.28738779e-01
-2.10742369e-01 -3.24796647e-01 -2.17525095e-01 2.12481722e-01
1.44030082e+00 -3.88976723e-01 -9.01132226e-02 -3.30170572e-01
3.70400369e-01 1.63320169e-01 6.56908154e-01 6.77586496e-01
-1.71400025e-01 5.77571452e-01 4.23783958e-02 -2.42906004e-01
-4.08650786e-01 -1.26203465e+00 -1.59135342e-01 1.93074793e-01
6.97921991e-01 -3.63992780e-01 -5.09928703e-01 -7.80247152e-01
1.69015720e-01 5.10194659e-01 7.60836154e-02 -5.60923219e-02
-3.51723671e-01 -8.75928640e-01 4.22604084e-01 2.65514553e-01
1.03023899e+00 -5.15267074e-01 -5.76751292e-01 5.63620478e-02
-2.62258917e-01 -1.50690103e+00 -7.53266513e-02 -1.27910718e-01
-1.03388119e+00 -1.01095247e+00 -2.92772502e-01 -6.79293633e-01
7.87811279e-01 6.12848699e-01 8.09860468e-01 9.90394577e-02
-3.56786937e-01 4.44985420e-01 -3.57082188e-01 -4.30982441e-01
5.74556470e-01 6.92764148e-02 -1.33595809e-01 2.43503563e-02
4.30425406e-01 -1.07828796e+00 -7.72327304e-01 4.04686451e-01
-6.58881664e-01 -9.55751985e-02 8.26285660e-01 3.29208404e-01
7.25865126e-01 3.67949247e-01 3.01453918e-01 -8.52856338e-01
5.93074225e-02 -3.58067006e-01 -1.00273740e+00 -7.02692568e-02
-6.62052572e-01 -3.92106265e-01 4.03533280e-01 1.03019051e-01
-5.76777995e-01 2.62955070e-01 7.59729743e-02 -5.74890316e-01
-1.05486661e-02 5.45814991e-01 -4.76964623e-01 -6.17016852e-01
1.52982876e-01 1.99552715e-01 -1.61397263e-01 -3.77811402e-01
1.87596217e-01 4.35838252e-01 4.10910398e-01 -3.09678137e-01
1.50457406e+00 9.27162468e-01 4.26597536e-01 -1.06277156e+00
-6.13600433e-01 -6.47312045e-01 -3.77998203e-01 -1.00722864e-01
7.58568704e-01 -1.10832846e+00 -7.46096075e-01 1.95435598e-01
-9.24662292e-01 2.44542912e-01 -1.19467661e-01 5.83572090e-01
-5.20043194e-01 4.87904966e-01 -7.91331530e-02 -7.78502941e-01
-3.60179156e-01 -9.41898227e-01 1.18759835e+00 2.14579850e-01
1.46494225e-01 -6.66072130e-01 -3.93945098e-01 5.55736423e-01
3.08306754e-01 5.85693896e-01 8.15016210e-01 -5.81476152e-01
-1.47523379e+00 -1.91556275e-01 -6.15253329e-01 5.87916002e-02
-8.36587977e-03 -4.79001999e-01 -7.05754340e-01 -3.91301244e-01
1.31842330e-01 3.17193657e-01 7.72814751e-01 5.29204369e-01
1.33720708e+00 -2.38598451e-01 -7.73729980e-01 1.16934609e+00
1.66419041e+00 3.57155234e-01 7.60011613e-01 1.30253226e-01
6.73930287e-01 3.11141223e-01 7.82718897e-01 4.27577883e-01
4.61479485e-01 3.90550017e-01 7.64889896e-01 -1.81140691e-01
1.20062508e-01 -1.06631376e-01 -2.92156655e-02 6.95605457e-01
7.09486380e-02 -1.67337254e-01 -9.27364230e-01 4.59287971e-01
-1.89173424e+00 -7.95143604e-01 -1.80988580e-01 1.96322048e+00
7.92856589e-02 1.79210559e-01 -1.47927210e-01 -1.41879702e-02
6.30431712e-01 1.77481279e-01 -3.97241771e-01 2.65641540e-01
6.85503632e-02 4.59343851e-01 9.67109025e-01 1.91863701e-01
-1.21856785e+00 9.30829048e-01 4.78502083e+00 8.61006320e-01
-1.07397485e+00 9.83160920e-03 1.79796219e-01 -2.36888472e-02
-3.58681589e-01 1.39368981e-01 -7.23510444e-01 5.69982409e-01
3.83564591e-01 -2.78974921e-01 2.07192123e-01 7.21714675e-01
3.76215912e-02 2.49721855e-02 -8.59018564e-01 1.35890746e+00
-7.18766963e-03 -1.28105688e+00 -3.43426168e-02 4.30872560e-01
5.56821704e-01 8.55912790e-02 -8.72732699e-02 1.15713611e-01
1.79948583e-01 -8.98266912e-01 4.47145760e-01 4.28321987e-01
6.99197650e-01 -7.64344633e-01 7.39762604e-01 3.94464970e-01
-1.39734781e+00 -5.71443290e-02 -3.77594829e-01 1.66678905e-01
1.47315398e-01 1.17771232e+00 -6.31620824e-01 1.19671953e+00
8.46025884e-01 8.40433955e-01 -4.07460153e-01 1.40020692e+00
-3.60589288e-02 5.41937649e-01 -6.87696576e-01 -3.07464395e-02
2.67388523e-01 -6.76268637e-01 8.15050185e-01 1.01511407e+00
7.48862326e-01 4.61331338e-01 5.70553601e-01 8.47806811e-01
-1.68877207e-02 -6.90997168e-02 -6.83956087e-01 4.91752438e-02
7.88764715e-01 1.32123113e+00 -9.05637026e-01 1.85822636e-01
-3.73624057e-01 6.12442315e-01 -3.06276530e-01 4.66341347e-01
-8.36982548e-01 -5.04324675e-01 7.47193456e-01 2.06067160e-01
5.37596107e-01 -6.06728077e-01 -5.66186965e-01 -8.85303378e-01
1.87897697e-01 -4.52701956e-01 3.22592795e-01 -4.69351739e-01
-1.13841605e+00 4.52454984e-01 -8.61759037e-02 -1.55648553e+00
3.57064188e-01 -2.48580486e-01 -5.61211228e-01 4.37464505e-01
-1.58741128e+00 -1.40229332e+00 -7.43643463e-01 7.06309199e-01
4.92579602e-02 -2.06057057e-01 5.27137339e-01 4.71946806e-01
-2.01255798e-01 4.51776296e-01 -1.91143081e-01 2.16930538e-01
1.11346737e-01 -7.86382616e-01 3.25409919e-01 8.91347945e-01
7.20104203e-02 4.32223052e-01 4.20821011e-01 -7.33017087e-01
-1.61895943e+00 -1.39202452e+00 4.42415774e-01 -1.22952238e-01
5.07609367e-01 -6.47559345e-01 -5.16849101e-01 6.90797687e-01
-3.98202866e-01 6.09333813e-02 4.57946479e-01 -1.25409126e-01
2.12754104e-02 -3.70085567e-01 -1.53351295e+00 4.53999162e-01
1.61446571e+00 -1.62901253e-01 -3.74989569e-01 4.61102694e-01
8.31454635e-01 -3.35296929e-01 -6.63961291e-01 7.87991464e-01
2.58848876e-01 -9.01652813e-01 1.18795419e+00 1.50957748e-01
-8.85764360e-02 -4.72177058e-01 -4.05924797e-01 -9.65793490e-01
-4.45130914e-01 -5.88792801e-01 1.05172746e-01 1.20347285e+00
1.99669480e-01 -8.78493786e-01 1.31069565e+00 1.39945075e-01
-1.86229169e-01 -6.38972163e-01 -1.31718397e+00 -1.06028700e+00
-3.80407840e-01 -5.82921684e-01 8.33879411e-01 8.47509325e-01
-3.93663466e-01 1.13099635e-01 1.43006248e-02 8.10132563e-01
1.19735444e+00 5.58179200e-01 9.32676971e-01 -1.66831076e+00
1.84139132e-01 -2.33043134e-01 -9.20127988e-01 -1.12815177e+00
1.22455973e-02 -1.04559767e+00 -1.96066990e-01 -1.78533876e+00
-1.30088419e-01 -9.46751237e-01 -6.89208657e-02 3.46845865e-01
4.00287002e-01 2.40993008e-01 1.85309336e-01 1.20352417e-01
-6.13844872e-01 6.71184003e-01 1.24598122e+00 -3.00740063e-01
-9.10444334e-02 3.52542728e-01 -7.09661841e-01 7.02136517e-01
7.82380104e-01 -5.32368124e-01 -3.63296300e-01 -3.34495634e-01
-6.01389110e-02 7.25675300e-02 6.47086501e-01 -1.50646150e+00
5.09910285e-01 -2.47550651e-01 2.38909751e-01 -1.06751764e+00
4.94440407e-01 -1.48474813e+00 3.38782758e-01 5.44925869e-01
6.23138249e-01 -2.33844727e-01 9.50004254e-03 7.37285078e-01
-2.03541934e-01 1.52316436e-01 6.99868917e-01 -1.15593985e-01
-5.98934174e-01 8.75788867e-01 9.27693695e-02 4.11090665e-02
1.29951966e+00 -5.88853538e-01 -2.47682080e-01 -3.91000301e-01
-4.05119479e-01 6.51726186e-01 4.64795679e-01 1.58562347e-01
7.58877993e-01 -1.23101354e+00 -4.96515185e-01 4.29464281e-01
2.16508284e-01 2.85374314e-01 1.31184354e-01 8.15075040e-01
-5.26362121e-01 3.35678041e-01 3.72540206e-02 -8.96728933e-01
-1.03854406e+00 2.52680719e-01 1.32181391e-01 2.40374431e-02
-8.68823051e-01 9.41542387e-01 3.68714690e-01 -4.40432221e-01
2.32787296e-01 -4.22921121e-01 2.02888191e-01 -9.11735147e-02
-6.46857843e-02 3.38506669e-01 1.52186304e-01 -3.72307360e-01
-7.46135712e-01 1.00391197e+00 2.29819864e-01 -3.28143314e-02
1.42126870e+00 -5.70498928e-02 -1.69836417e-01 -1.79674119e-01
1.09081364e+00 1.49167165e-01 -8.97114575e-01 -2.28791356e-01
1.62367970e-02 -6.09862566e-01 3.71104293e-02 -6.63307130e-01
-1.51883125e+00 7.43106961e-01 6.62111282e-01 2.08225980e-01
1.28575003e+00 4.23283651e-02 1.16738260e+00 3.11421424e-01
1.00420070e+00 -8.26175988e-01 -1.37607247e-01 4.37255383e-01
5.00362694e-01 -8.61865997e-01 2.56045729e-01 -9.64841485e-01
-1.23641282e-01 7.44534850e-01 4.65581059e-01 -2.88072377e-01
1.16056001e+00 2.39006773e-01 4.49618325e-02 -7.95206964e-01
-1.10037804e-01 -2.87450671e-01 1.68208346e-01 6.98735774e-01
-4.32549685e-01 1.23184487e-01 -1.82933733e-01 7.13354230e-01
-7.33512461e-01 -6.93294480e-02 9.97108892e-02 9.08931434e-01
-6.15875602e-01 -8.04546237e-01 -3.35248858e-01 7.46999860e-01
-6.48971200e-02 1.05646722e-01 -3.07488114e-01 9.43062901e-01
3.01004499e-01 9.89861727e-01 5.69637232e-02 -5.05271494e-01
4.41490382e-01 -2.88781822e-01 3.28026593e-01 -4.86751318e-01
-2.62639552e-01 3.80144119e-02 -2.16758832e-01 -7.46216774e-01
-3.10501307e-01 -6.84978783e-01 -1.40299225e+00 -1.56896561e-01
-2.42261544e-01 3.04204196e-01 7.78803110e-01 9.14647043e-01
4.09759045e-01 4.49768841e-01 8.06184173e-01 -5.15156567e-01
-2.45689496e-01 -4.78034884e-01 -8.46517444e-01 1.41787291e-01
8.25780109e-02 -7.57983208e-01 -4.93930191e-01 -4.20892268e-01] | [7.868849277496338, -3.1256461143493652] |
35108f9d-4ea5-41b8-9aab-45767f6d802b | equivariant-few-shot-learning-from-pretrained | 2305.09900 | null | https://arxiv.org/abs/2305.09900v1 | https://arxiv.org/pdf/2305.09900v1.pdf | Equivariant Few-Shot Learning from Pretrained Models | Efficient transfer learning algorithms are key to the success of foundation models on diverse downstream tasks even with limited data. Recent works of \cite{basu2022equi} and \cite{kaba2022equivariance} propose group averaging (\textit{equitune}) and optimization-based methods, respectively, over features from group-transformed inputs to obtain equivariant outputs from non-equivariant neural networks. While \cite{kaba2022equivariance} are only concerned with training from scratch, we find that equitune performs poorly on equivariant zero-shot tasks despite good finetuning results. We hypothesize that this is because pretrained models provide better quality features for certain transformations than others and simply averaging them is deleterious. Hence, we propose $\lambda$-\textit{equitune} that averages the features using \textit{importance weights}, $\lambda$s. These weights are learned directly from the data using a small neural network, leading to excellent zero-shot and finetuned results that outperform equitune. Further, we prove that $\lambda$-equitune is equivariant and a universal approximator of equivariant functions. Additionally, we show that the method of \cite{kaba2022equivariance} used with appropriate loss functions, which we call \textit{equizero}, also gives excellent zero-shot and finetuned performance. Both equitune and equizero are special cases of $\lambda$-equitune. To show the simplicity and generality of our method, we validate on a wide range of diverse applications and models such as 1) image classification using CLIP, 2) deep Q-learning, 3) fairness in natural language generation (NLG), 4) compositional generalization in languages, and 5) image classification using pretrained CNNs such as Resnet and Alexnet. | ['Lav R. Varshney', 'Payel Das', 'Katherine Driggs-Campbell', 'Vijil Chenthamarakshan', 'Prasanna Sattigeri', 'Pulkit Katdare', 'Sourya Basu'] | 2023-05-17 | null | null | null | null | ['q-learning'] | ['methodology'] | [ 2.00262129e-01 1.06646277e-01 -9.51367766e-02 -3.60784113e-01
-8.30195665e-01 -5.70491493e-01 6.34675205e-01 -2.83482462e-01
-5.63221574e-01 1.08630013e+00 8.64344314e-02 -3.21171761e-01
-3.79531294e-01 -1.10164261e+00 -1.20519090e+00 -5.40538251e-01
3.79050151e-02 3.02447736e-01 -1.74913079e-01 -7.27893233e-01
-5.68635538e-02 2.55745947e-01 -1.30391121e+00 1.71969786e-01
1.05543947e+00 1.11302388e+00 -1.72275603e-02 7.20746756e-01
2.92695500e-02 6.91089809e-01 -4.60406214e-01 -7.13159859e-01
6.06289268e-01 -7.20452070e-01 -8.28736663e-01 -4.64117914e-01
6.40006125e-01 -1.73563674e-01 -2.41106778e-01 1.26475120e+00
5.31657696e-01 4.51011807e-01 8.50425363e-01 -1.30847442e+00
-1.05174398e+00 7.27507830e-01 -3.36664498e-01 1.77289590e-01
-2.31860191e-01 3.37954968e-01 1.44289434e+00 -9.71450210e-01
5.01899719e-01 1.29697120e+00 7.84457862e-01 7.79686689e-01
-1.34095991e+00 -1.09066391e+00 5.15079983e-02 5.19885309e-02
-1.13090396e+00 -3.28778803e-01 3.61182630e-01 -4.16484952e-01
9.78892565e-01 5.14154494e-01 2.86349624e-01 1.05026138e+00
3.42910290e-01 7.04530180e-01 1.15711153e+00 -4.64893490e-01
1.73663214e-01 9.14316550e-02 1.41016185e-01 9.04618084e-01
-7.88948908e-02 2.93078542e-01 -3.14466774e-01 3.14342193e-02
9.81605291e-01 2.74939854e-02 -1.30725533e-01 -5.74168004e-02
-1.23284209e+00 1.15812981e+00 7.65496433e-01 2.24836022e-01
-1.64137423e-01 5.34946859e-01 3.31131786e-01 7.88904369e-01
4.89336252e-01 8.80353689e-01 -6.09225869e-01 2.35659376e-01
-6.89173043e-01 3.40869397e-01 4.04618770e-01 1.04369366e+00
1.14209712e+00 4.74434912e-01 -4.73912179e-01 1.10898995e+00
-1.80971056e-01 5.49520016e-01 7.32962966e-01 -1.31098938e+00
3.83674324e-01 2.54528970e-01 -1.95576981e-01 -5.39277136e-01
-2.65332341e-01 -3.90738100e-01 -1.06489754e+00 4.37606543e-01
3.76758605e-01 -2.92011231e-01 -1.05157173e+00 2.43965816e+00
-2.64761537e-01 1.07936397e-01 1.19259633e-01 6.64323151e-01
7.19138741e-01 8.13346088e-01 2.52622366e-01 -9.11196694e-02
1.13139343e+00 -9.78016853e-01 -2.12285519e-01 4.99403104e-02
5.05911291e-01 -6.87391341e-01 1.64046776e+00 2.73497701e-01
-1.09053802e+00 -6.00695908e-01 -1.02820122e+00 -9.84231979e-02
-5.03446460e-01 -2.74546355e-01 7.03151405e-01 6.07999325e-01
-1.26118767e+00 8.74236763e-01 -4.90687966e-01 -2.87323356e-01
5.16414523e-01 5.63806295e-01 -3.23676258e-01 3.20285670e-02
-1.49815059e+00 9.08335626e-01 2.96997488e-01 -3.78658116e-01
-1.04324567e+00 -1.17353809e+00 -8.48965883e-01 1.72479182e-01
1.90385148e-01 -8.02261114e-01 1.13413811e+00 -1.42307973e+00
-1.77668452e+00 7.35076606e-01 1.39066443e-01 -6.40085757e-01
5.21528482e-01 -5.30534312e-02 -3.27297032e-01 -2.01838046e-01
2.22836390e-01 9.98534858e-01 9.29656386e-01 -8.10131490e-01
-5.77031791e-01 -9.99461114e-02 2.84749508e-01 6.11811839e-02
-6.50316358e-01 4.25616093e-02 9.02506188e-02 -9.34338331e-01
-3.23263884e-01 -8.82436275e-01 -4.57669571e-02 -1.10428318e-01
-3.50584477e-01 -3.88807207e-01 6.72714055e-01 -5.11414587e-01
9.62223113e-01 -1.88018656e+00 2.46118605e-01 1.63032264e-01
2.72251397e-01 3.55644703e-01 -5.46347618e-01 6.51568100e-02
-4.44292009e-01 4.32249874e-01 -4.56269771e-01 1.90537691e-01
1.01991378e-01 2.01831907e-01 -5.02513170e-01 9.20689702e-02
2.80846119e-01 1.25919616e+00 -8.04615974e-01 -9.88469571e-02
1.01426229e-01 1.09719642e-01 -8.93402159e-01 1.19032308e-01
-3.40342969e-01 2.14831531e-01 -1.92141458e-01 2.92620182e-01
3.89477551e-01 -1.89586818e-01 -3.30639005e-01 -2.61347920e-01
-8.23956132e-02 -1.19660251e-01 -7.42086530e-01 1.49184620e+00
-7.79544115e-01 4.09163594e-01 -2.33257681e-01 -1.22417724e+00
7.50003099e-01 2.62621552e-01 4.96493638e-01 -9.37780082e-01
1.38739854e-01 4.67063606e-01 3.23454710e-03 -1.63315132e-01
3.18438113e-01 -5.90709031e-01 -2.62049943e-01 5.45455635e-01
5.70248127e-01 -4.39645320e-01 2.31103763e-01 1.49683714e-01
1.00952506e+00 1.23494983e-01 3.70398313e-01 -6.16844416e-01
4.66673136e-01 -3.75124007e-01 5.20538568e-01 8.98782909e-01
-2.80038416e-01 6.46758556e-01 4.31113899e-01 -5.35756826e-01
-1.37108839e+00 -1.28262067e+00 3.09851822e-02 1.58793151e+00
-1.02790333e-01 -1.44258142e-01 -8.81339729e-01 -6.93158686e-01
-1.53181285e-01 1.01820779e+00 -7.40044355e-01 -5.85847914e-01
-6.11024618e-01 -1.05052245e+00 8.70359302e-01 5.16026556e-01
7.26286530e-01 -1.19615662e+00 -3.42551023e-01 4.74961177e-02
-1.70684502e-01 -6.25226617e-01 -8.65544319e-01 2.06265658e-01
-6.05745614e-01 -6.51988804e-01 -9.02888834e-01 -6.23453915e-01
4.66747940e-01 -1.20138682e-01 1.13080347e+00 -2.39130348e-01
-2.05769762e-01 3.00306618e-01 -1.66082978e-01 -6.58061028e-01
-4.69570309e-01 7.92600811e-02 1.63362905e-01 -1.31071731e-02
7.20977336e-02 -7.13682175e-01 -5.39911687e-01 3.23307306e-01
-1.05164695e+00 -3.81161235e-02 5.52333891e-01 1.21220326e+00
5.48127890e-01 -3.00905585e-01 8.70360076e-01 -9.08956587e-01
7.92104244e-01 -3.80064815e-01 -4.83753473e-01 2.56764233e-01
-8.03353667e-01 3.19893390e-01 1.05972302e+00 -6.29598856e-01
-8.99766922e-01 -4.77125973e-01 -1.38052896e-01 -7.45946229e-01
2.66686738e-01 1.45533279e-01 1.94127075e-02 7.29672611e-02
1.11633039e+00 6.03463203e-02 9.33448076e-02 -1.52732581e-01
6.14249587e-01 2.58624017e-01 3.72365475e-01 -7.43106425e-01
7.34560132e-01 3.07992399e-01 -1.18978456e-01 -5.87148845e-01
-8.50063980e-01 2.02216849e-01 -1.48718625e-01 1.85391940e-02
9.81671274e-01 -7.53917336e-01 -7.32416093e-01 3.42169762e-01
-9.82107759e-01 -6.28773570e-01 -7.11869895e-01 4.61688191e-01
-9.87318873e-01 -9.92562473e-02 -5.50502837e-01 -4.26956743e-01
-6.66716695e-01 -1.17444181e+00 8.26932609e-01 1.48441106e-01
-3.52463394e-01 -1.02083945e+00 -1.32130802e-01 1.76711917e-01
7.17988729e-01 1.34301260e-01 1.37150002e+00 -7.53205717e-01
-4.36747164e-01 7.22275972e-02 -3.26912820e-01 8.25265467e-01
1.41265959e-01 -2.10756719e-01 -9.72770154e-01 -3.45592648e-01
-2.65390575e-01 -6.22074842e-01 9.35783625e-01 5.11331737e-01
1.35634148e+00 -5.51183760e-01 2.96133071e-01 7.74106741e-01
1.31525826e+00 1.50758699e-01 7.41631150e-01 1.84861854e-01
6.62527084e-01 3.60428631e-01 7.90562183e-02 2.39073008e-01
7.67391548e-02 5.30586004e-01 3.30740720e-01 -1.90333441e-01
-2.76176751e-01 -2.57672697e-01 6.06799960e-01 8.60149503e-01
-3.67038071e-01 1.81994941e-02 -5.21084249e-01 4.27864164e-01
-1.66687393e+00 -1.05426180e+00 4.22333956e-01 2.16454315e+00
1.12005556e+00 -4.31948826e-02 8.18657316e-03 -1.51075274e-01
7.31951058e-01 1.33975549e-02 -7.05249190e-01 -8.88209343e-01
-2.38102898e-01 9.84446049e-01 5.92114747e-01 5.88616431e-01
-1.03098834e+00 1.36090147e+00 5.53572083e+00 1.33964264e+00
-1.36620128e+00 3.75211149e-01 9.89313126e-01 -2.57403791e-01
-5.60860515e-01 -7.32257292e-02 -4.87738878e-01 3.19320768e-01
9.01416957e-01 -4.41891134e-01 8.28401804e-01 7.03535736e-01
5.98813929e-02 5.16717911e-01 -1.17163146e+00 7.24500716e-01
2.56352802e-03 -1.43593681e+00 4.20892507e-01 -1.56194061e-01
1.10811388e+00 1.51995242e-01 4.41697478e-01 7.79057443e-01
8.45300078e-01 -1.28366053e+00 8.74853253e-01 4.49703068e-01
1.19435763e+00 -7.98250139e-01 3.86080414e-01 1.46500245e-01
-7.62920022e-01 -1.49367422e-01 -4.91206020e-01 1.33048028e-01
-2.01567143e-01 3.60779375e-01 -5.57025552e-01 4.86381114e-01
7.43842721e-01 4.32933122e-01 -1.96933359e-01 5.05354822e-01
-1.22632489e-01 6.05659485e-01 -1.01269059e-01 -1.01354539e-01
4.16426510e-01 -2.47345686e-01 5.69787860e-01 1.07298744e+00
6.06318414e-01 6.24860525e-02 -6.39311001e-02 1.18628180e+00
-5.64517558e-01 3.44686180e-01 -6.65745139e-01 1.05439112e-01
1.04146980e-01 1.17708266e+00 -2.14307860e-01 -4.12133366e-01
-1.93087488e-01 9.04859781e-01 3.90437067e-01 5.97781181e-01
-9.24878955e-01 -7.74560869e-01 9.47510540e-01 -1.15016960e-01
2.37881288e-01 5.53481802e-02 -2.52878398e-01 -1.32948077e+00
-2.68856525e-01 -1.10090303e+00 4.35910493e-01 -6.30479038e-01
-1.38531494e+00 6.74837410e-01 1.26173869e-01 -1.11169708e+00
-3.10917586e-01 -7.76914060e-01 -6.48972273e-01 9.99834597e-01
-1.37375700e+00 -1.10818481e+00 4.06954698e-02 8.59918177e-01
6.43642902e-01 -5.32923937e-01 9.77429569e-01 1.89888328e-01
-4.84790295e-01 1.01032960e+00 1.62578523e-01 1.02924153e-01
7.28722453e-01 -1.17054975e+00 2.50520110e-01 6.94929242e-01
1.42409667e-01 7.52559185e-01 6.04246795e-01 -3.01312417e-01
-9.80482876e-01 -1.44610846e+00 5.27471304e-01 -3.06497931e-01
6.95547402e-01 -2.83426255e-01 -6.66567385e-01 8.37209225e-01
2.99060166e-01 1.29382998e-01 4.98375058e-01 -1.69357345e-01
-6.12484396e-01 -5.38095713e-01 -1.19553065e+00 8.76548052e-01
1.07154536e+00 -4.71784502e-01 -4.69712585e-01 3.48380387e-01
1.13464022e+00 4.38343547e-02 -9.01188850e-01 5.66781580e-01
4.83835846e-01 -7.59062350e-01 8.44566107e-01 -1.08088458e+00
8.08881044e-01 1.61071226e-01 -4.13216412e-01 -1.57901311e+00
-4.21803474e-01 -8.31449628e-01 3.35111141e-01 8.96027684e-01
7.49187410e-01 -8.60215008e-01 4.87986892e-01 5.19917190e-01
-2.34833539e-01 -6.05375588e-01 -9.87739384e-01 -9.21509564e-01
8.84966433e-01 -4.24550295e-01 5.61398923e-01 1.06204236e+00
-2.55474448e-01 2.70340770e-01 -4.71582174e-01 -5.24077833e-01
4.87283558e-01 -2.09461063e-01 6.02854073e-01 -9.62632298e-01
-2.79022783e-01 -7.56269574e-01 -1.92475751e-01 -5.94214857e-01
4.12767887e-01 -1.41065383e+00 -1.47326067e-01 -9.87368047e-01
2.85637289e-01 -5.71875513e-01 -6.38136327e-01 8.61274183e-01
-1.26842976e-01 4.40625638e-01 2.31294379e-01 5.44018224e-02
-2.30185762e-01 7.27642953e-01 1.26574612e+00 -3.80973399e-01
-1.07899442e-01 -1.15081862e-01 -9.86711621e-01 5.22218049e-01
7.79945612e-01 -2.22247571e-01 -5.37231624e-01 -4.22800273e-01
1.73281685e-01 -2.79350787e-01 4.47808713e-01 -7.52488196e-01
-7.57399648e-02 -3.06297511e-01 2.02651516e-01 3.88188452e-01
2.28781685e-01 -2.60480255e-01 -1.19047165e-01 4.19382125e-01
-6.17964327e-01 3.84952426e-02 1.06894068e-01 2.45232269e-01
-1.10983066e-01 -2.27772743e-01 1.08890963e+00 -4.36020374e-01
-5.48813105e-01 6.42803848e-01 -1.66256487e-01 3.33887011e-01
9.13059592e-01 1.01098530e-01 -3.48834366e-01 -5.41093290e-01
-6.59544528e-01 8.13936964e-02 3.69192846e-02 4.71749812e-01
4.78038400e-01 -1.53129995e+00 -9.77301717e-01 2.93719888e-01
5.67821553e-03 -1.97005570e-01 1.03724360e-01 6.83139980e-01
-3.33606243e-01 3.54810625e-01 -4.48500961e-01 -2.61770159e-01
-7.62962520e-01 3.97026837e-01 5.06551504e-01 -3.70177537e-01
-2.17928723e-01 1.00681329e+00 7.06273317e-01 -8.00172150e-01
-2.25746423e-01 -2.27920845e-01 1.31788626e-01 -1.20615296e-01
2.43950337e-01 4.06464428e-01 2.42857710e-02 -4.31745976e-01
1.13185483e-03 4.83460695e-01 -1.67909175e-01 -3.58830065e-01
1.34818804e+00 2.92994440e-01 -2.80243397e-01 2.57070631e-01
1.53052282e+00 -2.54469991e-01 -1.13489413e+00 -2.73308069e-01
-3.00176144e-01 -4.55975980e-02 -2.56612986e-01 -7.97369123e-01
-1.13224733e+00 9.97086346e-01 5.94516039e-01 -4.09605727e-02
1.05421257e+00 -1.76663563e-01 7.97701418e-01 5.05195856e-01
3.94792497e-01 -1.18105614e+00 2.89374769e-01 6.51225209e-01
1.18813491e+00 -1.13571024e+00 -2.84217536e-01 7.10407495e-02
-6.99385285e-01 9.32531238e-01 7.14685082e-01 -2.61421770e-01
6.00938797e-01 -4.37093303e-02 -8.31644461e-02 2.92114496e-01
-7.63139606e-01 -9.04419646e-02 1.63344830e-01 3.35915774e-01
3.30942959e-01 1.73141733e-01 -4.70362872e-01 6.83254659e-01
-6.28501296e-01 -7.33770356e-02 2.02569172e-01 5.29459536e-01
-2.93623000e-01 -1.12892675e+00 -1.39263824e-01 6.34075284e-01
-4.80769843e-01 -4.07745451e-01 -1.82262465e-01 8.49475265e-01
2.71044999e-01 5.24945319e-01 4.82791997e-02 -4.48743403e-01
2.85206586e-01 2.16074169e-01 4.68713850e-01 -4.94992763e-01
-7.18465507e-01 -1.48889989e-01 -2.15506643e-01 -4.22325909e-01
-1.17906965e-01 -3.47346693e-01 -1.25711071e+00 -3.14396888e-01
-9.88288522e-02 7.13674650e-02 4.09295201e-01 8.37374747e-01
2.89057404e-01 5.81964433e-01 7.24825025e-01 -6.15490973e-01
-9.33755338e-01 -1.07473350e+00 -4.39639926e-01 6.98992014e-01
1.53190061e-01 -4.57391232e-01 -3.79063457e-01 9.17383656e-02] | [10.55920124053955, 8.170063972473145] |
ba243020-c114-4271-a6a3-a6cb1ecdb160 | segeqa-video-segmentation-based-visual | null | null | http://openaccess.thecvf.com/content_ICCV_2019/html/Luo_SegEQA_Video_Segmentation_Based_Visual_Attention_for_Embodied_Question_Answering_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Luo_SegEQA_Video_Segmentation_Based_Visual_Attention_for_Embodied_Question_Answering_ICCV_2019_paper.pdf | SegEQA: Video Segmentation Based Visual Attention for Embodied Question Answering | Embodied Question Answering (EQA) is a newly defined research area where an agent is required to answer the user's questions by exploring the real world environment. It has attracted increasing research interests due to its broad applications in automatic driving system, in-home robots, and personal assistants. Most of the existing methods perform poorly in terms of answering and navigation accuracy due to the absence of local details and vulnerability to the ambiguity caused by complicated vision conditions. To tackle these problems, we propose a segmentation based visual attention mechanism for Embodied Question Answering. Firstly, We extract the local semantic features by introducing a novel high-speed video segmentation framework. Then by the guide of extracted semantic features, a bottom-up visual attention mechanism is proposed for the Visual Question Answering (VQA) sub-task. Further, a feature fusion strategy is proposed to guide the training of the navigator without much additional computational cost. The ablation experiments show that our method boosts the performance of VQA module by 4.2% (68.99% vs 64.73%) and leads to 3.6% (48.59% vs 44.98%) overall improvement in EQA accuracy.
| [' Yazhou Yao', ' Zhenmin Tang', ' Fayao Liu', ' Zichuan Liu', ' Guosheng Lin', 'Haonan Luo'] | 2019-10-01 | null | null | null | iccv-2019-10 | ['embodied-question-answering'] | ['computer-vision'] | [ 5.44877015e-02 1.42840207e-01 2.64000833e-01 -3.91225249e-01
-7.31333315e-01 -4.89885300e-01 5.27592123e-01 -5.13528399e-02
-7.31767356e-01 3.69580567e-01 2.98358407e-03 -3.56477171e-01
-3.70605253e-02 -8.32697868e-01 -6.40011191e-01 -4.68348593e-01
2.96844721e-01 2.84280390e-01 4.75539595e-01 -5.89711130e-01
3.00398976e-01 1.28886431e-01 -1.66792691e+00 6.03085048e-02
1.23300958e+00 1.04961050e+00 6.69512749e-01 5.56699276e-01
-3.23545396e-01 7.56784976e-01 -5.49923301e-01 -1.42752990e-01
-1.02631889e-01 -4.37869251e-01 -1.09101427e+00 -2.87476033e-02
1.90152258e-01 -3.50888401e-01 -4.38838303e-01 1.05197370e+00
4.16220307e-01 4.15889144e-01 4.13198829e-01 -1.48038566e+00
-6.97539747e-01 -8.64072964e-02 -5.25663793e-01 1.37814596e-01
6.40025139e-01 2.55582184e-01 7.33093619e-01 -8.87292147e-01
5.06627023e-01 1.32863677e+00 2.68154383e-01 4.86101717e-01
-5.42132199e-01 -3.73588830e-01 3.93699288e-01 6.62903488e-01
-1.18079209e+00 -3.69765192e-01 7.90350318e-01 -2.74393409e-01
8.03749144e-01 2.03227386e-01 4.84229773e-01 7.96676815e-01
1.31853640e-01 9.78987932e-01 8.88692856e-01 -2.12284818e-01
2.87215292e-01 6.45659566e-02 2.24890411e-01 1.14584470e+00
-1.02812704e-02 -1.71827674e-01 -4.10773039e-01 2.19456345e-01
4.88155633e-01 -6.11207858e-02 -1.92767605e-01 -5.35118401e-01
-1.16479564e+00 8.87941301e-01 8.81989896e-01 1.80321127e-01
-4.52113807e-01 2.06198439e-01 3.28401387e-01 3.07353737e-04
-1.30479112e-01 3.92194301e-01 -2.57369399e-01 -1.91211820e-01
-3.51706415e-01 3.34635288e-01 5.43661535e-01 9.76257682e-01
8.24291646e-01 -2.46749036e-02 -2.93196976e-01 8.21188509e-01
6.11026764e-01 5.85525393e-01 3.96749794e-01 -1.23419440e+00
5.23985028e-01 8.57697070e-01 3.20848197e-01 -1.26988852e+00
-7.49613047e-01 -2.89408743e-01 -4.84783500e-01 1.67545527e-01
5.28846562e-01 5.12441853e-03 -1.06695092e+00 1.65666890e+00
5.57725370e-01 -1.84648752e-01 1.94442719e-01 1.23738110e+00
1.04698467e+00 7.18846560e-01 2.54955858e-01 2.84808844e-01
1.93176603e+00 -1.44992268e+00 -8.60677719e-01 -5.83042979e-01
4.61078376e-01 -4.30509925e-01 1.29570043e+00 -2.92641781e-02
-8.15057695e-01 -7.66062200e-01 -1.14508080e+00 -3.92402887e-01
-5.59802353e-01 1.52966395e-01 4.58352029e-01 4.19367045e-01
-1.07433581e+00 -2.35064089e-01 -4.84052002e-01 -6.33518994e-01
5.86005151e-01 4.25121903e-01 -3.21372986e-01 -3.60815585e-01
-1.17873895e+00 7.55590200e-01 2.73730278e-01 3.02753299e-01
-8.19804132e-01 -2.01196522e-01 -1.11905587e+00 7.56255984e-02
6.49685383e-01 -9.72610176e-01 1.26884651e+00 -5.59328020e-01
-1.48562181e+00 6.47666395e-01 -5.08570015e-01 -1.95245609e-01
4.48422462e-01 -3.93455803e-01 -2.09443763e-01 5.58273911e-01
4.87535894e-01 9.97437298e-01 7.00677693e-01 -1.18823266e+00
-7.71500111e-01 -7.22181737e-01 4.74291503e-01 5.75521529e-01
-1.26453787e-02 -3.98285151e-01 -7.90879846e-01 -3.05035353e-01
2.33947679e-01 -8.16745818e-01 -2.47519806e-01 4.77426238e-02
1.55157363e-02 -4.61597234e-01 1.22410619e+00 -9.18194234e-01
7.75197446e-01 -2.01259613e+00 1.24289207e-01 -4.14504930e-02
3.52464318e-01 2.94894904e-01 -2.51992464e-01 1.02847360e-01
5.82036674e-01 -1.71716228e-01 -3.76193434e-01 -2.02768996e-01
5.69567271e-02 1.97272897e-01 -1.20219253e-01 1.66599885e-01
2.26770461e-01 1.25056326e+00 -1.01825213e+00 -5.25289357e-01
4.01687235e-01 4.88045156e-01 -4.80036944e-01 3.02467704e-01
-1.98502511e-01 5.96580923e-01 -7.45047092e-01 7.43663251e-01
5.85220039e-01 -1.04725078e-01 -2.72655964e-01 -2.68983804e-02
-1.61932949e-02 -1.63791418e-01 -8.11381817e-01 2.18302608e+00
-5.45191944e-01 6.76038325e-01 2.92689204e-01 -9.85480607e-01
9.62204099e-01 5.63227572e-02 7.08258450e-02 -1.39532185e+00
4.49657917e-01 7.88374022e-02 -6.92350715e-02 -9.69854712e-01
5.79227626e-01 2.04102278e-01 -2.63948023e-01 1.21642925e-01
-2.15012152e-02 8.03159252e-02 -3.14679854e-02 2.83081293e-01
9.88026559e-01 2.78643608e-01 4.96641956e-02 -2.49211043e-01
8.73312950e-01 2.56772220e-01 2.50058651e-01 7.15502501e-01
-7.81387627e-01 4.40325677e-01 9.09809172e-02 -3.14405859e-01
-7.32298672e-01 -7.99818099e-01 2.98492610e-01 1.18503761e+00
8.92688334e-01 -2.16953903e-01 -1.02436888e+00 -7.48092532e-01
-2.76470482e-01 6.08772755e-01 -6.54278696e-01 -3.39858264e-01
-4.44581777e-01 -3.29939067e-01 2.47164682e-01 6.29093826e-01
1.25236440e+00 -1.32904100e+00 -1.11647570e+00 1.01899661e-01
-5.92074811e-01 -1.36324906e+00 -2.85057902e-01 -3.40907127e-01
-5.95739067e-01 -1.02736676e+00 -8.44450116e-01 -1.15911674e+00
6.12278759e-01 7.21119702e-01 6.67921305e-01 1.25494495e-01
-2.34253660e-01 7.10160375e-01 -3.94571096e-01 -2.44176567e-01
1.14881389e-01 2.17468679e-01 -2.57182211e-01 2.62315292e-02
4.57533062e-01 -1.28672495e-01 -8.39303792e-01 4.34140503e-01
-5.88424444e-01 1.79233536e-01 5.67346036e-01 7.09493637e-01
4.62321579e-01 -1.61586940e-01 8.71114492e-01 -3.59158278e-01
7.77562916e-01 -4.69382256e-01 -3.59279513e-01 2.14268193e-01
-2.36007780e-01 8.85421038e-02 4.35368568e-01 -1.46067813e-01
-1.24378955e+00 1.44259203e-02 -2.07011297e-01 -1.21075837e-02
-4.10911322e-01 2.28747934e-01 -4.19018507e-01 -5.62243797e-02
3.12709123e-01 2.63174266e-01 7.33652264e-02 -1.23169042e-01
4.73471075e-01 6.78082168e-01 7.74708986e-01 -3.92824203e-01
5.82599640e-01 6.09257221e-01 -2.61636823e-01 -7.92934358e-01
-6.75900280e-01 -5.72206438e-01 -4.25891340e-01 -4.63635266e-01
1.38619030e+00 -8.27532649e-01 -1.14188540e+00 4.55077410e-01
-1.14243495e+00 -1.22951001e-01 1.71360984e-01 3.25554997e-01
-6.68289423e-01 3.73893559e-01 -1.76391736e-01 -8.75298083e-01
-5.21856785e-01 -1.44442225e+00 1.14481091e+00 7.11947262e-01
-1.28670946e-01 -5.36018074e-01 -3.28614831e-01 9.39310312e-01
5.17942965e-01 3.54337916e-02 7.46462226e-01 -3.47463220e-01
-7.55116522e-01 5.55764139e-02 -6.42455101e-01 -1.65133208e-01
-1.87864024e-02 -6.52052701e-01 -9.67098176e-01 -7.86568075e-02
1.46351466e-02 -1.41513363e-01 7.67568052e-01 1.02791816e-01
9.41531718e-01 1.22555971e-01 -3.11319143e-01 3.67472291e-01
1.18346679e+00 7.73042083e-01 5.56090832e-01 6.44744158e-01
8.63929212e-01 8.20943475e-01 8.84732902e-01 2.43893474e-01
9.54504490e-01 6.21526599e-01 8.28368127e-01 -2.47210693e-02
9.09409206e-03 -3.16314042e-01 2.35075712e-01 4.61400658e-01
1.69204995e-01 -1.15029298e-01 -1.05763710e+00 8.67023349e-01
-2.11438751e+00 -6.60903037e-01 -2.53338277e-01 1.71487701e+00
1.53403461e-01 5.64302206e-02 4.73555066e-02 3.38211954e-02
7.01833069e-01 1.52157947e-01 -8.01008403e-01 -3.95980000e-01
2.37510413e-01 -1.51620895e-01 1.17818333e-01 5.53154647e-01
-1.09495914e+00 1.10233927e+00 5.27016115e+00 6.10127866e-01
-6.76198721e-01 3.05101182e-02 3.80565792e-01 4.14958835e-01
-1.59401804e-01 -1.64319232e-01 -4.75655645e-01 3.62749904e-01
5.17088890e-01 2.43814960e-01 2.91072726e-01 7.60107815e-01
1.62844613e-01 -5.15733600e-01 -5.54733217e-01 1.27641797e+00
2.58968443e-01 -9.10734117e-01 3.88123132e-02 -2.49672174e-01
3.75424325e-01 -3.03647637e-01 1.14562646e-01 5.47637820e-01
-5.75133190e-02 -9.80733335e-01 8.46992314e-01 5.98972976e-01
3.66897166e-01 -9.72977936e-01 8.95736873e-01 4.52112377e-01
-1.30574048e+00 -2.50513673e-01 -1.67758927e-01 -2.30969906e-01
3.26857865e-01 -5.03865592e-02 -6.26268923e-01 5.27696908e-01
9.74325180e-01 3.13469857e-01 -6.02820516e-01 8.88574004e-01
-2.98336357e-01 3.01421016e-01 -2.70300180e-01 -3.23141366e-01
5.72858572e-01 -2.12258548e-01 4.31281209e-01 7.48877406e-01
1.10176414e-01 2.90911168e-01 -3.00512463e-02 8.39364350e-01
3.08756363e-02 7.17344061e-02 -4.35429871e-01 2.17282549e-01
2.20839009e-01 1.17973530e+00 -7.92997956e-01 -2.51737952e-01
-4.26139891e-01 1.37281334e+00 3.54808718e-01 6.35355115e-01
-1.07753134e+00 -9.45102632e-01 3.97011787e-01 -8.11132863e-02
3.75740230e-01 -3.88981760e-01 -1.90379947e-01 -9.53855753e-01
2.27039546e-01 -6.94908440e-01 1.61655739e-01 -1.18346262e+00
-6.61184311e-01 7.19010532e-01 -3.41016144e-01 -7.49511898e-01
-2.31807251e-02 -6.06386423e-01 -4.89579290e-01 7.52358913e-01
-1.68843174e+00 -1.06449544e+00 -9.94700313e-01 5.41349947e-01
7.95677602e-01 -1.29141331e-01 5.79507232e-01 2.52283961e-01
-5.37659764e-01 4.39822108e-01 -1.87500238e-01 9.32167172e-02
4.09985214e-01 -1.07752013e+00 4.06112939e-01 6.44904375e-01
-1.90039173e-01 4.47522372e-01 5.08592665e-01 -2.65371859e-01
-1.50974381e+00 -9.20510352e-01 8.20084691e-01 -4.08415705e-01
2.73643970e-01 -2.96976030e-01 -9.30556297e-01 4.00588304e-01
2.83864230e-01 -1.28447965e-01 3.31050754e-01 -2.98868984e-01
-1.58865049e-01 -1.72017030e-02 -1.28708494e+00 7.56250620e-01
1.02283156e+00 -6.26552165e-01 -7.11237729e-01 -2.04594150e-01
8.28885972e-01 -3.62122923e-01 -4.21518952e-01 2.51326680e-01
5.68080485e-01 -8.87367547e-01 8.77384126e-01 -2.71753162e-01
1.77673683e-01 -7.33147740e-01 -2.05460057e-01 -7.92927861e-01
-2.06303164e-01 -3.19201440e-01 -5.22535592e-02 1.04086649e+00
1.78368777e-01 -6.42890453e-01 5.48867106e-01 6.35186315e-01
-2.28387624e-01 -8.50787878e-01 -8.18926811e-01 -2.58080184e-01
-3.98236632e-01 -3.92999291e-01 6.37320876e-01 6.82110906e-01
-1.38814673e-01 7.06020951e-01 -5.67204021e-02 3.46589923e-01
3.59789252e-01 1.63002685e-01 8.23688090e-01 -9.55670953e-01
1.87279552e-01 -3.34142745e-01 -4.35027003e-01 -1.50438654e+00
-1.24583142e-02 -6.79275155e-01 3.77029687e-01 -2.22751069e+00
3.28908376e-02 -1.59593508e-01 -1.95771292e-01 3.18132669e-01
-3.42168301e-01 1.69671863e-01 3.08697790e-01 -2.15613693e-02
-1.06720030e+00 8.90295923e-01 1.41214621e+00 -1.91216186e-01
-1.79186821e-01 -3.04140449e-01 -7.73881674e-01 7.45804548e-01
1.05819559e+00 -8.80265236e-02 -5.80617666e-01 -7.18430698e-01
-2.72948854e-02 7.33161122e-02 6.40427232e-01 -1.08903205e+00
2.71506339e-01 1.69275790e-01 2.74663985e-01 -7.71251321e-01
4.38635945e-01 -7.98968375e-01 -4.91288364e-01 4.89301801e-01
-1.12908585e-02 1.10066973e-01 3.08863878e-01 7.18859076e-01
-3.25486571e-01 -1.30625784e-01 5.51150084e-01 -3.72954756e-02
-1.36698020e+00 6.43405616e-02 -5.75491309e-01 1.02206685e-01
1.25041246e+00 -3.36189121e-01 -2.96304435e-01 -5.46989977e-01
-5.00665665e-01 8.08642149e-01 1.10469215e-01 7.78492093e-01
7.75446892e-01 -1.31182241e+00 -2.99859554e-01 1.33699492e-01
2.89690763e-01 3.29896629e-01 5.53536713e-01 6.89079881e-01
-5.97382128e-01 6.58787072e-01 -2.79528916e-01 -5.87091744e-01
-1.05178297e+00 4.11095202e-01 3.79332602e-01 8.76413956e-02
-5.47777534e-01 8.43824327e-01 4.57245052e-01 -4.74707127e-01
2.34458476e-01 -2.09385365e-01 -6.06557190e-01 -2.47789230e-02
4.70237672e-01 4.97468233e-01 -1.78069547e-01 -9.35848355e-01
-5.37454844e-01 8.06356311e-01 -2.35527847e-02 -7.81377926e-02
8.26677024e-01 -4.97557461e-01 2.94187456e-01 1.04256742e-01
1.18819666e+00 -2.84156978e-01 -1.25606298e+00 -1.10132098e-01
-1.47977382e-01 -2.10090458e-01 -5.11169024e-02 -9.18212593e-01
-8.97680461e-01 1.20431101e+00 7.23135114e-01 2.50837337e-02
1.02449262e+00 1.19436644e-01 1.02217662e+00 4.46207374e-01
4.06449586e-01 -9.82042551e-01 2.92316645e-01 5.49974382e-01
8.97540152e-01 -1.49549198e+00 -4.63908166e-01 -2.64419854e-01
-7.47091949e-01 7.22675800e-01 9.32087183e-01 5.84344845e-03
1.09920986e-01 -4.11623538e-01 2.93516845e-01 -3.76788735e-01
-2.11684614e-01 -3.94237161e-01 2.91616321e-01 6.83910906e-01
-1.72687441e-01 -1.67958602e-01 -2.87374467e-01 9.09909189e-01
-1.84236571e-01 -3.29643875e-01 1.83823377e-01 1.05529487e+00
-7.45014310e-01 -4.95939553e-01 -2.71366179e-01 1.33425370e-01
-1.04186095e-01 3.02150041e-01 -2.44758189e-01 7.73397446e-01
1.87978014e-01 1.35702038e+00 -1.69447064e-02 -4.26743776e-01
5.89226246e-01 1.11651301e-01 2.40727112e-01 -2.06305847e-01
-2.22362682e-01 -2.10726336e-01 -9.25963894e-02 -7.12872326e-01
-5.30810535e-01 -4.50677484e-01 -1.91319358e+00 7.44005963e-02
-2.32684493e-01 1.74903527e-01 7.35612094e-01 1.10050535e+00
6.12022221e-01 8.08929384e-01 2.76456863e-01 -5.44627666e-01
2.61099804e-02 -6.42998755e-01 -5.81714623e-02 2.47567698e-01
4.17780966e-01 -8.92862260e-01 -2.09037364e-01 -1.27789006e-01] | [4.49120569229126, 0.38993972539901733] |
2d532c63-8a88-42e4-a6df-a0a8e8ae7817 | gradient-coherent-strong-regularization-for | 1811.08056 | null | https://arxiv.org/abs/1811.08056v2 | https://arxiv.org/pdf/1811.08056v2.pdf | Gradient-Coherent Strong Regularization for Deep Neural Networks | Regularization plays an important role in generalization of deep neural networks, which are often prone to overfitting with their numerous parameters. L1 and L2 regularizers are common regularization tools in machine learning with their simplicity and effectiveness. However, we observe that imposing strong L1 or L2 regularization with stochastic gradient descent on deep neural networks easily fails, which limits the generalization ability of the underlying neural networks. To understand this phenomenon, we first investigate how and why learning fails when strong regularization is imposed on deep neural networks. We then propose a novel method, gradient-coherent strong regularization, which imposes regularization only when the gradients are kept coherent in the presence of strong regularization. Experiments are performed with multiple deep architectures on three benchmark data sets for image recognition. Experimental results show that our proposed approach indeed endures strong regularization and significantly improves both accuracy and compression (up to 9.9x), which could not be achieved otherwise. | ['Dae Hoon Park', 'Yi Chang', 'Huaqing Zhang', 'Chiu Man Ho'] | 2018-11-20 | null | null | null | null | ['l2-regularization'] | ['methodology'] | [ 1.81731030e-01 -1.91005930e-01 -1.52445197e-01 -3.64961416e-01
-2.49148965e-01 -2.65356004e-01 3.07003200e-01 -6.62361756e-02
-7.20062256e-01 7.90874720e-01 1.60199590e-02 -3.07368726e-01
-1.65059060e-01 -6.06517375e-01 -9.08401251e-01 -9.56293523e-01
2.52478927e-01 -1.46674395e-01 1.42826498e-01 -4.16619191e-03
1.63521186e-01 5.85142553e-01 -1.30489826e+00 3.22000384e-02
8.03140163e-01 1.08135653e+00 7.56793097e-02 3.53629708e-01
-1.55229822e-01 1.10161412e+00 -4.78615999e-01 -3.17208499e-01
3.42218131e-01 -1.06899150e-01 -4.66271549e-01 1.35880709e-01
5.09542823e-01 -1.62406474e-01 -2.32786894e-01 1.19594395e+00
2.97130853e-01 1.26448184e-01 4.99950677e-01 -6.36616170e-01
-5.71510911e-01 4.69848961e-01 -8.06849241e-01 3.23655754e-01
-3.83781075e-01 7.32279569e-02 8.83694053e-01 -1.02839613e+00
6.82599768e-02 1.21826696e+00 9.46605623e-01 6.71987653e-01
-1.27901876e+00 -6.84579492e-01 3.40904206e-01 -2.81496316e-01
-1.30821562e+00 -4.60267723e-01 8.73039961e-01 -1.37291238e-01
7.45383739e-01 2.76616842e-01 4.18458246e-02 8.48145008e-01
-8.33091810e-02 8.91468942e-01 8.54739487e-01 -3.87598366e-01
1.38886660e-01 2.81644493e-01 5.31677008e-01 8.21624279e-01
4.28009868e-01 -3.62304114e-02 -2.65237510e-01 5.71595542e-02
6.94682539e-01 2.93612406e-02 -3.71404976e-01 6.88732266e-02
-7.39925265e-01 1.01731610e+00 5.76268137e-01 3.85946989e-01
-3.49412501e-01 2.47751683e-01 5.34359574e-01 4.00456041e-01
5.54930568e-01 1.01377763e-01 -5.06929159e-01 1.69471413e-01
-6.46236002e-01 -1.05493844e-01 6.00057662e-01 6.51852608e-01
6.61455989e-01 5.32384753e-01 2.79835109e-02 1.30515218e+00
9.96036008e-02 1.60026029e-01 6.54169440e-01 -6.85019433e-01
5.68729401e-01 5.88636756e-01 -3.31613064e-01 -1.15016186e+00
-3.91660661e-01 -9.57332373e-01 -1.47868919e+00 3.42825592e-01
4.89804566e-01 -3.15296143e-01 -8.38908434e-01 1.82437789e+00
-1.22719057e-01 3.26661229e-01 2.13667482e-01 8.36290538e-01
9.55962896e-01 5.35548627e-01 1.56647876e-01 -2.82811046e-01
8.37352633e-01 -1.01476336e+00 -7.22287655e-01 -5.45162916e-01
8.94028723e-01 -6.07671320e-01 1.40550303e+00 3.09866607e-01
-1.06073236e+00 -6.19611442e-01 -1.17359340e+00 1.36358127e-01
-1.08259380e-01 4.20037925e-01 6.01080239e-01 7.29128540e-01
-8.54028463e-01 7.47158587e-01 -8.10764313e-01 6.27287775e-02
6.86950684e-01 4.50575143e-01 -2.79861569e-01 4.23962250e-02
-9.36743081e-01 6.39299452e-01 4.81246203e-01 3.23644638e-01
-3.90495330e-01 -6.77117109e-01 -5.81848323e-01 2.54289567e-01
2.71573275e-01 -1.61995471e-01 9.54794884e-01 -1.12939548e+00
-1.40402269e+00 7.47011065e-01 9.49840620e-02 -6.48471415e-01
6.29784822e-01 -6.30078197e-01 -7.67035484e-02 -2.10808590e-01
-5.19417763e-01 2.87806213e-01 8.80299509e-01 -1.06594896e+00
-3.23995918e-01 -3.63318086e-01 -1.62870526e-01 -5.26168430e-03
-9.36705649e-01 -9.68090296e-02 -5.32186091e-01 -7.40794718e-01
3.20892096e-01 -9.40277576e-01 -4.40581620e-01 5.60718030e-02
-5.08906662e-01 -1.38242275e-01 8.73472393e-01 -3.27928096e-01
1.29223025e+00 -2.28437209e+00 -8.97129402e-02 2.39649042e-01
2.65924037e-01 7.63674676e-01 -1.76111624e-01 -2.19535545e-01
-1.71118632e-01 2.39249468e-01 -4.34685946e-01 -4.06395733e-01
-1.58762157e-01 4.47156906e-01 -2.88364500e-01 6.17187321e-01
2.45515317e-01 6.67432904e-01 -3.88823271e-01 -7.18362257e-02
1.29878253e-01 8.01912189e-01 -7.13268876e-01 4.27302681e-02
-4.08139266e-02 3.71848285e-01 -4.42259759e-01 3.01267684e-01
9.88289475e-01 -4.46586072e-01 2.06409886e-01 -2.82265425e-01
5.32059371e-02 8.98645520e-02 -1.17045641e+00 1.19717503e+00
-4.38921243e-01 6.95293009e-01 1.62678719e-01 -1.56428719e+00
1.08320773e+00 -4.10331078e-02 2.68420279e-01 -5.54925144e-01
2.40508825e-01 3.39955360e-01 -8.12804103e-02 -3.56177837e-01
7.04791993e-02 -9.98610184e-02 2.88280576e-01 2.20673829e-01
-1.25502959e-01 2.16562837e-01 2.34130412e-01 -1.22101746e-01
8.63252699e-01 -4.17075574e-01 -1.74666718e-02 -5.08477926e-01
7.99577057e-01 -7.01679051e-01 7.94172704e-01 7.68762231e-01
1.40937017e-02 6.04829967e-01 6.05698884e-01 -7.30060637e-01
-8.10423136e-01 -5.67848742e-01 -3.31291914e-01 1.12148142e+00
-2.09526852e-01 -3.41142327e-01 -6.26248598e-01 -6.38161242e-01
-9.16142613e-02 4.77481604e-01 -6.49330497e-01 -4.34528857e-01
-7.53033698e-01 -1.39047122e+00 6.48867130e-01 6.26768768e-01
8.91892493e-01 -7.92628050e-01 -3.71299833e-01 8.65718946e-02
2.31647387e-01 -1.22873068e+00 -1.68472871e-01 4.32646096e-01
-1.27736831e+00 -8.54466975e-01 -8.48200262e-01 -8.87261450e-01
9.41891730e-01 1.66641921e-01 1.04861915e+00 6.48520768e-01
-1.29324406e-01 -1.91144913e-01 1.89506523e-02 -1.40925333e-01
-2.85849243e-01 1.99414149e-01 1.88727200e-01 -1.20834690e-02
3.91894132e-01 -9.05226231e-01 -4.68910754e-01 2.64318556e-01
-1.01289892e+00 -2.04543814e-01 7.45652139e-01 1.11265635e+00
4.44616199e-01 9.73732024e-02 4.90574449e-01 -1.33031917e+00
8.45797956e-01 -2.25974649e-01 -7.89110363e-01 1.11210920e-01
-6.11064672e-01 3.42694759e-01 8.32400322e-01 -5.26108205e-01
-9.76834357e-01 2.69750897e-02 -4.36790019e-01 -4.77530062e-01
6.77643577e-03 6.94303453e-01 1.03396475e-01 -4.69452649e-01
9.79049087e-01 1.81156859e-01 -1.57864809e-01 -8.08299959e-01
6.38291091e-02 3.40856522e-01 3.05421352e-01 -5.92421234e-01
6.01298988e-01 2.39662126e-01 1.27084151e-01 -1.03168416e+00
-1.15055394e+00 -9.79902148e-02 -4.85073864e-01 1.09566428e-01
4.07425582e-01 -7.32850134e-01 -7.68863440e-01 5.50271928e-01
-9.49012399e-01 -2.78391004e-01 -1.16197556e-01 6.21082842e-01
-8.00179988e-02 3.73261213e-01 -8.03865910e-01 -6.67849541e-01
-4.10894394e-01 -1.15108621e+00 1.83464751e-01 1.57296926e-01
1.60917982e-01 -1.20316982e+00 -3.24702948e-01 6.24435432e-02
5.89845538e-01 2.43897200e-01 9.14418161e-01 -6.87463760e-01
-2.76255488e-01 -2.62268186e-01 -4.92109299e-01 9.78954911e-01
1.34628609e-01 2.86241267e-02 -1.01484311e+00 -4.47144717e-01
3.01956296e-01 -6.00884378e-01 1.28146398e+00 4.55533743e-01
1.52647638e+00 -3.92481625e-01 7.08418638e-02 9.92162108e-01
1.63858449e+00 -2.61764318e-01 6.18553519e-01 3.07569832e-01
8.09749782e-01 4.23974365e-01 -1.01163104e-01 3.44330251e-01
-3.27002496e-01 3.47777575e-01 4.60935324e-01 -3.31069648e-01
-7.31675550e-02 1.84591487e-01 1.89617112e-01 1.06246293e+00
-2.53794044e-01 9.23022926e-02 -7.90371120e-01 2.81250179e-01
-1.72304475e+00 -6.31915092e-01 -3.10195416e-01 2.15631509e+00
1.02238643e+00 6.25529110e-01 -2.28706345e-01 4.52889770e-01
5.16299546e-01 2.80097514e-01 -6.36908114e-01 -4.43879485e-01
-3.53704005e-01 1.42403126e-01 6.19897664e-01 4.97504890e-01
-1.06452179e+00 8.65629971e-01 6.14695263e+00 8.82154644e-01
-1.39924514e+00 1.06258597e-02 9.08644080e-01 -2.32305080e-01
-3.33904661e-02 -5.04586637e-01 -9.56138492e-01 3.97415817e-01
7.48492956e-01 1.19624414e-01 1.62192017e-01 7.88389623e-01
-3.59186120e-02 1.90850377e-01 -9.46199834e-01 1.10510540e+00
-1.92747578e-01 -1.44836926e+00 1.00641429e-01 -1.73077121e-01
8.95973802e-01 1.65086702e-01 4.71865624e-01 3.52172017e-01
1.75242141e-01 -1.09065986e+00 2.94695497e-01 1.32514298e-01
5.71813822e-01 -1.01333296e+00 6.83101416e-01 5.65347195e-01
-7.86017418e-01 -2.41829291e-01 -8.30689967e-01 -1.49622276e-01
-2.13121697e-01 9.58110392e-01 -4.48925376e-01 1.34674031e-02
5.95850945e-01 7.62477398e-01 -4.95594919e-01 8.02758813e-01
-2.59693533e-01 8.71744990e-01 -5.15082121e-01 1.91663299e-02
4.93989348e-01 -3.28268826e-01 3.97000790e-01 1.45423150e+00
1.79077625e-01 5.84429614e-02 8.11098367e-02 6.96504354e-01
-6.46006286e-01 1.82431653e-01 -5.95341623e-01 8.13660324e-02
7.89591745e-02 9.40654218e-01 -6.80148125e-01 -1.90861434e-01
-5.59468329e-01 6.95268393e-01 6.01952374e-01 5.81961334e-01
-6.97798789e-01 -2.23697573e-01 5.06054103e-01 -1.20733187e-01
2.05350205e-01 -1.44187868e-01 -6.65851772e-01 -1.13370216e+00
3.04522485e-01 -7.40164757e-01 3.46529245e-01 -9.63423997e-02
-1.18099189e+00 8.12131464e-01 -5.48035800e-01 -9.71460938e-01
1.14805512e-01 -8.12538743e-01 -5.05301893e-01 7.46494591e-01
-1.81427658e+00 -6.01631880e-01 -2.17461452e-01 5.66318989e-01
6.41153216e-01 -4.09475297e-01 7.07976639e-01 6.30396843e-01
-9.14152980e-01 9.27712262e-01 4.75742728e-01 2.05419555e-01
3.19552928e-01 -8.26702535e-01 9.42564905e-02 7.08034217e-01
6.00448437e-02 6.65787101e-01 7.05362499e-01 -1.94833502e-01
-1.11843240e+00 -1.30604291e+00 5.86755872e-01 2.17727333e-01
5.22298753e-01 -2.15831533e-01 -1.39404035e+00 5.18710256e-01
-2.28623357e-02 4.35780704e-01 6.84528589e-01 1.70962602e-01
-3.95894200e-01 -4.14227635e-01 -1.01443624e+00 5.65750718e-01
7.53429532e-01 -3.85924399e-01 -3.29972208e-01 6.69341385e-01
5.77526331e-01 -1.77781820e-01 -4.78287905e-01 7.47190297e-01
3.48527044e-01 -1.06134152e+00 1.12930202e+00 -6.99012637e-01
4.68578249e-01 1.61493823e-01 -2.63435930e-01 -1.05895686e+00
-2.17613980e-01 -5.25238574e-01 -1.16379984e-01 1.02661002e+00
5.28222263e-01 -6.32249355e-01 1.08664906e+00 7.21180677e-01
1.22889541e-01 -9.40415740e-01 -6.49814069e-01 -1.06287467e+00
4.17218834e-01 -4.20474082e-01 1.72310472e-01 1.07941628e+00
-5.11653244e-01 2.00914904e-01 -5.49019814e-01 7.12715089e-02
7.61045814e-01 -1.86556429e-01 4.10493374e-01 -1.34034050e+00
-9.04251933e-02 -6.80356264e-01 -2.41264388e-01 -1.26164532e+00
4.04638410e-01 -7.94328988e-01 -2.72230878e-02 -9.52871442e-01
-4.35945904e-03 -6.29907727e-01 -6.90315425e-01 3.62583995e-01
-2.24484876e-01 5.26875675e-01 1.09645100e-02 3.06841344e-01
-2.84899026e-01 7.02072203e-01 1.05312479e+00 -9.98241082e-02
-3.39952976e-01 3.47324848e-01 -6.50880039e-01 1.23837078e+00
1.06566668e+00 -4.65474874e-01 -4.21767533e-01 -8.52782607e-01
1.32240310e-01 -4.55908358e-01 1.14633650e-01 -1.19766545e+00
2.62410313e-01 2.41015442e-02 4.17358637e-01 -1.67342424e-01
2.09619775e-01 -8.28167498e-01 -2.52821773e-01 6.08392358e-01
-7.70169616e-01 -1.11245275e-01 5.10980904e-01 4.62092787e-01
-3.56800318e-01 -5.07736564e-01 1.24785376e+00 -5.06990775e-02
-5.02185404e-01 3.45135033e-01 -1.83683902e-01 9.52999443e-02
5.59211910e-01 1.23400681e-01 -8.15121178e-03 -1.89181730e-01
-6.37278795e-01 -8.11425596e-02 4.32493761e-02 1.07453011e-01
8.48479569e-01 -1.40883851e+00 -8.98509562e-01 4.46343035e-01
-2.75343418e-01 -3.88053879e-02 -6.15793988e-02 8.12729120e-01
-5.05964041e-01 2.69257903e-01 -8.55466723e-02 -6.32007182e-01
-1.36681867e+00 4.79538471e-01 3.07573110e-01 -3.30130041e-01
-8.23348641e-01 1.31498086e+00 3.86587739e-01 -3.46292019e-01
7.33486712e-01 -3.85254532e-01 -2.60800987e-01 -9.95614678e-02
5.85103631e-01 1.33620232e-01 2.14858919e-01 -5.00925541e-01
-2.99880803e-01 6.86900616e-01 -4.56330091e-01 4.41979825e-01
1.67265558e+00 2.14613359e-02 -6.34899065e-02 3.16780508e-01
1.59849489e+00 -4.26192917e-02 -1.40245247e+00 -6.96861625e-01
2.33445808e-01 -4.16918576e-01 3.69817615e-01 -2.83992887e-01
-1.77712929e+00 1.28483319e+00 4.33033377e-01 2.83751458e-01
1.19464743e+00 -4.00615722e-01 8.30256104e-01 9.07357395e-01
-2.10123822e-01 -1.07085443e+00 1.80376396e-01 8.24552953e-01
9.15085673e-01 -1.34015846e+00 1.38648883e-01 -2.74908781e-01
-3.28911692e-01 1.22667348e+00 6.60263717e-01 -4.51812834e-01
7.55428553e-01 3.13999265e-01 2.59574428e-02 -5.59454933e-02
-7.11425424e-01 1.31026074e-01 2.30715558e-01 4.48970236e-02
6.29412830e-01 -3.45362335e-01 -4.20098871e-01 5.22337496e-01
1.11473896e-01 -5.07121868e-02 2.51530409e-01 8.78801823e-01
-5.25930524e-01 -8.80326211e-01 -1.03179000e-01 3.89036059e-01
-8.36231291e-01 -1.89236239e-01 -1.25613704e-01 6.49304390e-01
-1.20686673e-01 6.50253892e-01 -2.02209696e-01 -2.92372823e-01
3.35131586e-01 -2.21946269e-01 2.09763899e-01 -3.57402354e-01
-5.35644352e-01 9.04052407e-02 -1.04367517e-01 -4.07283843e-01
-4.08318102e-01 -1.53397530e-01 -1.19268584e+00 -3.89436543e-01
-3.79799932e-01 1.47630006e-01 6.20164514e-01 9.26690161e-01
7.14196339e-02 6.03103876e-01 6.25026584e-01 -5.30558586e-01
-9.37221885e-01 -8.78588855e-01 -5.16816556e-01 5.09071529e-01
6.44614041e-01 -4.37393278e-01 -8.13440740e-01 1.22354940e-01] | [8.543329238891602, 3.557657480239868] |
034422e9-429c-471c-83f5-57fd5f4546ec | can-we-trust-explainable-ai-methods-on-asr-an | 2305.18011 | null | https://arxiv.org/abs/2305.18011v1 | https://arxiv.org/pdf/2305.18011v1.pdf | Can We Trust Explainable AI Methods on ASR? An Evaluation on Phoneme Recognition | Explainable AI (XAI) techniques have been widely used to help explain and understand the output of deep learning models in fields such as image classification and Natural Language Processing. Interest in using XAI techniques to explain deep learning-based automatic speech recognition (ASR) is emerging. but there is not enough evidence on whether these explanations can be trusted. To address this, we adapt a state-of-the-art XAI technique from the image classification domain, Local Interpretable Model-Agnostic Explanations (LIME), to a model trained for a TIMIT-based phoneme recognition task. This simple task provides a controlled setting for evaluation while also providing expert annotated ground truth to assess the quality of explanations. We find a variant of LIME based on time partitioned audio segments, that we propose in this paper, produces the most reliable explanations, containing the ground truth 96% of the time in its top three audio segments. | ['Ajitha Rajan', 'Peter Bell', 'Xiaoliang Wu'] | 2023-05-29 | null | null | null | null | ['automatic-speech-recognition'] | ['speech'] | [ 4.21362132e-01 7.99847424e-01 -3.43684882e-01 -8.06695998e-01
-9.71941888e-01 -3.16335857e-01 9.17275071e-01 -7.17447996e-02
2.29574457e-01 6.47486210e-01 4.84219909e-01 -6.25046968e-01
-5.19158781e-01 -3.42652090e-02 -9.91760552e-01 -4.88172412e-01
3.18109877e-02 7.40776837e-01 -4.13716555e-01 9.09370109e-02
2.40062818e-01 2.80292124e-01 -1.59471655e+00 9.76374209e-01
5.82099199e-01 1.07031941e+00 -2.10546702e-01 8.90289247e-01
-1.76997632e-01 1.01751769e+00 -8.50265920e-01 -2.62850583e-01
-1.47398636e-01 -6.94442272e-01 -1.26819646e+00 6.67889938e-02
4.33795720e-01 -7.34689012e-02 -1.06137976e-01 5.50880671e-01
-1.37870625e-01 -1.35889128e-01 6.95269406e-01 -1.80846715e+00
-8.29649448e-01 1.14273322e+00 1.21007100e-01 1.81792289e-01
1.45060316e-01 6.74468949e-02 1.17535710e+00 -8.29526603e-01
3.65597755e-01 1.54497707e+00 4.39929008e-01 9.76392448e-01
-1.27912807e+00 -6.17196441e-01 3.55672598e-01 5.74012816e-01
-9.11072254e-01 -3.73776376e-01 5.46800017e-01 -1.35007098e-01
1.19063771e+00 8.15031528e-01 3.42846125e-01 1.54365277e+00
2.22621053e-01 9.49677348e-01 1.09393871e+00 -6.26245379e-01
4.21030760e-01 2.55065084e-01 4.81912225e-01 5.43739736e-01
7.13034347e-02 2.42652968e-01 -1.29539680e+00 -7.64778778e-02
2.76016146e-01 -2.27970466e-01 -5.19736707e-01 -1.19366916e-03
-1.47143102e+00 9.51726675e-01 5.09322643e-01 3.56111705e-01
-5.82978010e-01 5.98915935e-01 2.84369320e-01 5.20165980e-01
6.93982065e-01 8.25257182e-01 -9.22984123e-01 -2.00761646e-01
-8.76300156e-01 2.48959452e-01 6.56700134e-01 6.04501784e-01
5.23143530e-01 5.21792591e-01 7.55044399e-03 4.79897320e-01
6.59655035e-01 2.97840476e-01 7.09963024e-01 -1.27377594e+00
1.40311122e-01 4.20667171e-01 -2.52337396e-01 -5.91508389e-01
-1.78168178e-01 -4.41673845e-01 -7.17147589e-01 4.79455501e-01
2.19770119e-01 1.53834313e-01 -1.16887140e+00 1.75556731e+00
-2.65307337e-01 3.69878531e-01 2.55493581e-01 1.15672541e+00
8.27354372e-01 7.99257159e-01 -2.97677889e-02 -1.11173633e-02
1.17422879e+00 -1.05815792e+00 -7.11246610e-01 -4.18916970e-01
4.24687415e-01 -4.51695263e-01 1.25025725e+00 8.25949669e-01
-8.56739342e-01 -6.49868965e-01 -1.18323267e+00 -2.04041656e-02
-2.92216778e-01 1.00063205e-01 9.49756861e-01 4.67285961e-01
-1.21305513e+00 6.59428656e-01 -8.63857448e-01 -2.44351536e-01
2.89972693e-01 5.97351313e-01 -5.25985003e-01 1.51687130e-01
-9.91038978e-01 9.21843529e-01 1.22091435e-01 -7.88876191e-02
-1.27611053e+00 -6.06312513e-01 -7.59120584e-01 4.26719874e-01
1.32784873e-01 -7.25536823e-01 1.59195209e+00 -1.60755563e+00
-1.35825038e+00 5.90098381e-01 -5.62428832e-01 -1.08604789e+00
7.32745230e-02 -4.03965175e-01 -3.14551055e-01 1.06391057e-01
1.92094892e-02 1.24561310e+00 9.58596826e-01 -1.47575271e+00
-2.38783091e-01 -1.26426741e-01 6.40891716e-02 -3.31649512e-01
4.85895164e-02 -2.38536894e-01 1.61598250e-01 -4.90042299e-01
3.46091211e-01 -1.07414591e+00 9.29132625e-02 -6.29287399e-03
-6.30655408e-01 -2.60622025e-01 7.72141218e-01 -6.01210117e-01
6.68669939e-01 -2.00605083e+00 1.47907123e-01 1.13870136e-01
4.37973142e-01 5.55649102e-02 -2.42236704e-01 5.01570627e-02
-7.15629876e-01 5.44493496e-01 -1.78472728e-01 -6.55220926e-01
2.48908788e-01 6.80463135e-01 -1.03983271e+00 1.05589166e-01
4.61443067e-01 7.41461992e-01 -7.03711867e-01 -2.79353932e-02
3.48726571e-01 6.01300418e-01 -4.81866628e-01 1.14909988e-02
-5.18340349e-01 4.50267732e-01 -1.28985524e-01 2.94884413e-01
-9.82002616e-02 -3.00997645e-01 -3.89897257e-01 -1.09624192e-02
7.74190053e-02 7.19050169e-01 -5.56837559e-01 1.70155978e+00
-4.23096985e-01 1.56284904e+00 -4.38109457e-01 -1.12140787e+00
8.39991331e-01 9.28976238e-01 1.62237152e-01 -3.77108008e-01
2.89263390e-02 3.07338744e-01 3.14401209e-01 -3.24491650e-01
1.72215536e-01 -1.02468297e-01 2.27823898e-01 5.79011798e-01
2.11330816e-01 2.60749869e-02 -5.28712869e-01 4.16288197e-01
1.23760879e+00 -6.01802133e-02 2.02847272e-01 -2.17507139e-01
4.64873344e-01 2.98406601e-01 1.87521055e-01 1.03109717e+00
6.02812739e-03 1.06453383e+00 2.35648751e-01 -7.90296435e-01
-1.02079916e+00 -8.08254480e-01 1.87632307e-01 6.88712120e-01
-1.29928827e-01 -4.57855552e-01 -1.00080788e+00 -5.82264185e-01
-3.75650972e-01 1.36658621e+00 -7.63171673e-01 -4.00590450e-01
-7.82233179e-02 6.98649371e-03 4.25190270e-01 4.42619205e-01
2.00755581e-01 -1.34054971e+00 -6.29902482e-01 2.73071742e-03
-4.20615792e-01 -8.59886885e-01 -4.64986674e-02 6.34677529e-01
-8.04870427e-01 -7.22906232e-01 -2.24278018e-01 -6.45575225e-02
4.17012542e-01 3.02766502e-01 1.28184962e+00 3.64387721e-01
-9.34209302e-02 6.67174220e-01 -3.85587245e-01 -9.42183375e-01
-8.19409013e-01 -4.34358930e-03 3.28800917e-01 -2.20693164e-02
3.83310795e-01 -4.27820951e-01 -2.40682214e-01 2.39620328e-01
-9.36256289e-01 3.30907077e-01 5.50854981e-01 8.12639654e-01
5.65647364e-01 -2.16904938e-01 4.96861249e-01 -7.89707243e-01
4.99393433e-01 -3.97281706e-01 9.18678269e-02 1.55472323e-01
-8.38423371e-01 6.43303752e-01 5.17947018e-01 -5.98616242e-01
-8.49939644e-01 -8.39785486e-02 -1.79697365e-01 -5.72025657e-01
-7.10758507e-01 6.91800952e-01 -1.93871796e-01 2.08361372e-01
9.31387544e-01 -8.69574100e-02 -3.10148783e-02 -3.37873220e-01
4.85331446e-01 6.81414425e-01 9.62744355e-01 -3.27918589e-01
6.88809335e-01 4.01371747e-01 -1.16328001e-01 -5.30012727e-01
-1.15515804e+00 -2.16105301e-02 -2.88580775e-01 -2.52194762e-01
8.79955173e-01 -6.48329139e-01 -6.47425234e-01 -3.47902358e-01
-1.77905643e+00 -2.55416900e-01 -4.38780367e-01 8.16969872e-01
-1.01183140e+00 -1.92139968e-01 -2.81232119e-01 -8.93720031e-01
-3.87021273e-01 -1.20843530e+00 1.30169821e+00 -1.33094370e-01
-1.18133330e+00 -7.99751401e-01 -1.42340764e-01 6.41244590e-01
3.47171962e-01 1.41975120e-01 1.14436436e+00 -1.26265192e+00
-7.97818482e-01 -9.41275060e-02 1.97593905e-02 2.33999386e-01
-1.82739664e-02 -2.26615230e-03 -1.60697353e+00 3.17250729e-01
2.27493152e-01 -3.43370348e-01 9.95934248e-01 5.90566456e-01
1.38104486e+00 -6.89168632e-01 -1.29012644e-01 1.68626592e-01
8.23264897e-01 1.40063494e-01 5.56083858e-01 4.71248478e-01
4.66775566e-01 6.71279013e-01 5.58451474e-01 1.94073897e-02
7.53934234e-02 7.92903662e-01 7.94116616e-01 -1.93513736e-01
-1.56226829e-01 -2.04774812e-01 5.99443913e-01 6.36731088e-01
1.93803698e-01 -5.42612851e-01 -9.45135713e-01 6.23102248e-01
-2.01348805e+00 -1.18049860e+00 -2.99615890e-01 1.95404053e+00
5.39344072e-01 3.28557074e-01 -3.42330456e-01 7.71413565e-01
1.73690617e-01 -3.22690904e-01 -5.10450304e-01 -9.72915471e-01
-2.09437087e-02 1.31886885e-01 -7.51234964e-02 9.02365804e-01
-6.33398056e-01 5.99314094e-01 6.32951641e+00 4.91329104e-01
-1.29018128e+00 1.00260928e-01 9.14114356e-01 7.31223971e-02
-8.37986588e-01 2.47510448e-02 -1.21462561e-01 2.70289928e-02
1.62122750e+00 3.81180756e-02 5.62610388e-01 1.14003706e+00
5.19879580e-01 2.00924307e-01 -1.67656827e+00 1.05550575e+00
1.46702528e-01 -1.65691006e+00 3.58433723e-01 7.90863410e-02
3.37614030e-01 -1.52922422e-02 3.11816841e-01 9.27899480e-02
4.32459414e-02 -1.49582016e+00 1.18054509e+00 5.26976466e-01
4.24619168e-01 -5.82483947e-01 7.02659786e-01 4.07627046e-01
-5.43639839e-01 1.09013721e-01 -1.68165490e-01 -3.35703522e-01
-9.44985151e-02 4.47747797e-01 -1.33812416e+00 2.99618244e-01
6.11495554e-01 4.75814402e-01 -4.24768507e-01 6.35793149e-01
-4.72742945e-01 1.28499866e+00 -1.21357769e-01 2.15206891e-01
3.39556009e-01 4.15531546e-01 6.91631198e-01 9.43339705e-01
3.29774946e-01 5.47369905e-02 -4.22287971e-01 1.24731863e+00
1.67033225e-01 -5.69605768e-01 -7.28657782e-01 -1.47994384e-01
2.51572669e-01 8.26914012e-01 -5.75307786e-01 -5.38733482e-01
-7.17918128e-02 1.08813715e+00 -4.75987531e-02 2.75691718e-01
-8.79931509e-01 2.26016656e-01 8.36686015e-01 -1.36828482e-01
-1.68947130e-01 -9.15364996e-02 -6.83324397e-01 -1.03333795e+00
-1.51234373e-01 -1.40319824e+00 6.40676590e-03 -1.45412290e+00
-8.67934704e-01 1.10771525e+00 3.20334852e-01 -1.06702268e+00
-9.85849142e-01 -7.64051497e-01 -6.49709523e-01 7.70564914e-01
-1.21317482e+00 -9.40717280e-01 -3.68190378e-01 2.91555256e-01
1.22453761e+00 -4.24561828e-01 1.12318790e+00 -1.79891825e-01
-1.39961377e-01 3.27277392e-01 -3.25801492e-01 -4.37484473e-01
3.52685958e-01 -1.38337409e+00 6.79639757e-01 8.20687771e-01
1.22763741e+00 8.37385297e-01 1.43086267e+00 -1.61422685e-01
-1.10266840e+00 -8.22103977e-01 1.34830296e+00 -8.07674050e-01
2.42632270e-01 -1.22746401e-01 -1.20361018e+00 8.67975295e-01
4.24154788e-01 -3.95249009e-01 6.89772666e-01 2.45801061e-01
-4.88410801e-01 -6.37968481e-02 -8.56619596e-01 5.22366703e-01
7.41331041e-01 -6.29022896e-01 -5.78608096e-01 4.21163946e-01
1.06973290e+00 -2.64190346e-01 -2.95169324e-01 2.18148082e-01
4.45910007e-01 -1.12624800e+00 8.42383504e-01 -7.92634606e-01
6.98185921e-01 -2.70368010e-01 -2.70327091e-01 -1.47380352e+00
7.96824321e-02 -6.39577389e-01 8.94496311e-03 1.21240127e+00
7.99580753e-01 -4.54956710e-01 8.20544064e-01 9.86782908e-01
-5.76200426e-01 -6.25121772e-01 -1.02260685e+00 -5.58275044e-01
-4.44076300e-01 -1.16675270e+00 6.40747249e-01 6.46953940e-01
5.28629348e-02 5.07531881e-01 -3.67712677e-01 3.74039501e-01
4.09986585e-01 -1.14343815e-01 8.82583141e-01 -1.39005506e+00
-3.22293371e-01 -2.21918941e-01 -6.14969611e-01 -8.19809496e-01
5.70046604e-01 -6.84980750e-01 3.27453852e-01 -1.50431252e+00
3.47384103e-02 7.12821409e-02 -1.00663260e-01 8.13505411e-01
1.43767342e-01 1.81151941e-01 2.38617733e-01 4.23313528e-01
-4.15946275e-01 4.91233408e-01 5.27711332e-01 -5.28166533e-01
2.19068080e-01 -4.34427001e-02 -8.74223173e-01 8.92636716e-01
8.51869106e-01 -7.06866205e-01 -7.90922105e-01 -5.91309726e-01
-2.25507487e-02 -1.31347193e-03 7.50463247e-01 -1.06453025e+00
-3.49702947e-02 3.10050622e-02 1.89991578e-01 -2.59909868e-01
5.76538801e-01 -9.20038283e-01 4.08862650e-01 3.72384816e-01
-1.09688938e+00 8.48363340e-02 3.71623695e-01 4.66543436e-01
-5.42325556e-01 -2.44075224e-01 2.62151301e-01 6.50258809e-02
-5.18228292e-01 2.08695829e-02 -4.61512089e-01 -4.66252327e-01
2.56103605e-01 -2.46710807e-01 -3.15677732e-01 -1.17445815e+00
-6.87327385e-01 -3.00541431e-01 -7.93966800e-02 6.97922647e-01
9.38922107e-01 -1.21725285e+00 -6.45145893e-01 4.38984558e-02
1.68366313e-01 -3.24102521e-01 -2.55319059e-01 5.20541966e-01
-1.77586228e-01 1.10510159e+00 5.66314906e-02 -8.10001254e-01
-1.50687599e+00 3.89914930e-01 3.69038731e-01 1.78576931e-01
-5.64139068e-01 8.63581598e-01 3.99121791e-01 -2.92199850e-01
5.24302959e-01 -8.14196706e-01 1.03627399e-01 -6.06139064e-01
6.09937668e-01 -9.96403545e-02 7.64432922e-02 -4.87751037e-01
-3.32483470e-01 6.31471677e-03 1.21726684e-01 -6.44448519e-01
1.27139771e+00 -1.71310633e-01 2.03266278e-01 9.57948864e-01
9.14640725e-01 -4.18331563e-01 -1.02111602e+00 1.38020352e-01
1.65486068e-01 -2.07409188e-01 1.24598064e-01 -1.04456353e+00
-7.47651935e-01 1.31658196e+00 6.05519295e-01 4.24316496e-01
8.42549264e-01 2.93937653e-01 4.07511562e-01 5.35284996e-01
1.52302176e-01 -3.13315570e-01 2.46612310e-01 1.84810907e-01
1.32101774e+00 -1.30589402e+00 -2.40268245e-01 -1.20064333e-01
-9.04005647e-01 1.19619858e+00 4.32857692e-01 2.77025014e-01
1.76249430e-01 -1.53464712e-02 5.19550562e-01 -3.57182741e-01
-1.34621501e+00 9.80646685e-02 6.32271886e-01 7.52312779e-01
5.46967864e-01 -7.31155872e-02 2.36048505e-01 7.34429836e-01
-4.70225990e-01 -2.00657755e-01 4.67824131e-01 7.58570135e-02
-4.09694374e-01 -9.72672224e-01 -5.03679216e-01 1.90351397e-01
-2.61033267e-01 -9.94863659e-02 -1.07074344e+00 7.94844210e-01
-1.46182448e-01 1.30621243e+00 -1.18944034e-01 -4.89603549e-01
-1.35232091e-01 3.02076966e-01 -3.64800766e-02 -5.44051051e-01
-5.59696794e-01 -3.15282315e-01 1.51639804e-01 -6.45747840e-01
-5.30175567e-01 -3.94160271e-01 -1.66586614e+00 -2.55718172e-01
-3.75120491e-01 4.05721992e-01 1.04704118e+00 1.59992361e+00
3.95747930e-01 6.90472662e-01 2.79022008e-01 -9.49356973e-01
-2.88397372e-01 -8.69352818e-01 -6.69232607e-02 5.06383181e-01
7.72512913e-01 -4.77831572e-01 -7.67337441e-01 3.29360992e-01] | [8.969198226928711, 5.62040376663208] |
8b1eca46-fc86-4808-927b-1dd60cbc3b4b | uncertainty-driven-action-quality-assessment | 2207.14513 | null | https://arxiv.org/abs/2207.14513v1 | https://arxiv.org/pdf/2207.14513v1.pdf | Uncertainty-Driven Action Quality Assessment | Automatic action quality assessment (AQA) has attracted more interests due to its wide applications. However, existing AQA methods usually employ the multi-branch models to generate multiple scores, which is not flexible for dealing with a variable number of judges. In this paper, we propose a novel Uncertainty-Driven AQA (UD-AQA) model to generate multiple predictions only using one single branch. Specifically, we design a CVAE (Conditional Variational Auto-Encoder) based module to encode the uncertainty, where multiple scores can be produced by sampling from the learned latent space multiple times. Moreover, we output the estimation of uncertainty and utilize the predicted uncertainty to re-weight AQA regression loss, which can reduce the contributions of uncertain samples for training. We further design an uncertainty-guided training strategy to dynamically adjust the learning order of the samples from low uncertainty to high uncertainty. The experiments show that our proposed method achieves new state-of-the-art results on the Olympic events MTL-AQA and surgical skill JIGSAWS datasets. | ['Yaping Huang', 'Caixia Zhou'] | 2022-07-29 | null | null | null | null | ['action-quality-assessment'] | ['computer-vision'] | [ 1.24880143e-01 2.81573176e-01 -2.83892065e-01 -5.27230918e-01
-1.39586544e+00 -5.12929037e-02 3.72764885e-01 1.04461730e-01
-4.06748325e-01 9.77919459e-01 5.77831507e-01 -4.99542169e-02
-4.10333514e-01 -7.42648304e-01 -5.47488391e-01 -8.56174946e-01
4.07849222e-01 5.73378384e-01 2.46061340e-01 -5.20251468e-02
2.66428322e-01 -2.70209372e-01 -1.11317229e+00 3.27571869e-01
1.48111808e+00 8.93785655e-01 4.56990264e-02 4.10248369e-01
-1.67379284e-03 6.91538453e-01 -5.83794415e-01 -6.76065564e-01
1.04696415e-01 -6.19397998e-01 -7.20738113e-01 -1.01810187e-01
-4.27062660e-02 -3.81136328e-01 -7.85702020e-02 9.78035688e-01
5.72124243e-01 2.13455692e-01 9.01835561e-01 -1.06618249e+00
-4.95267868e-01 7.78746724e-01 -5.16095281e-01 1.37160331e-01
6.42805174e-02 3.00007731e-01 1.07689166e+00 -5.65363348e-01
2.75520444e-01 1.34084797e+00 4.13149446e-01 7.90619373e-01
-1.07388282e+00 -6.30552411e-01 2.10552648e-01 3.11617821e-01
-1.19101083e+00 -1.66235045e-01 6.79411530e-01 -5.04941285e-01
1.80120677e-01 -9.67715830e-02 7.50459611e-01 1.07177877e+00
7.19161391e-01 1.12565458e+00 1.11078179e+00 -2.21441954e-01
5.40462971e-01 -1.72532663e-01 -2.64057130e-01 7.99525082e-01
9.78044197e-02 1.03030972e-01 -5.86921275e-01 -1.51875436e-01
7.03947484e-01 3.04902047e-02 -1.39259838e-03 -3.73306632e-01
-1.08131850e+00 9.48605061e-01 3.10123801e-01 -2.44491220e-01
-4.77096558e-01 2.09230348e-01 4.20023799e-01 3.19947116e-02
2.76363432e-01 4.29363757e-01 -2.58920461e-01 -2.64484107e-01
-9.02320027e-01 3.66330773e-01 8.12818259e-02 5.44827759e-01
3.53599727e-01 -3.09933200e-02 -9.83358443e-01 8.96305561e-01
6.87974513e-01 4.09644753e-01 6.78932309e-01 -1.16983712e+00
4.85416889e-01 5.49725711e-01 3.17619622e-01 -7.84023941e-01
-2.58298635e-01 -4.59916323e-01 -8.51196587e-01 3.00910860e-01
1.53390780e-01 -1.76082104e-01 -1.03202403e+00 1.81280196e+00
3.59188288e-01 1.77647337e-01 7.58168399e-02 9.13697422e-01
4.94870514e-01 6.08245254e-01 1.56523496e-01 -4.85060841e-01
1.20745087e+00 -1.06954443e+00 -9.80453014e-01 -1.00425214e-01
4.25266266e-01 -3.54619503e-01 1.20663583e+00 7.65061438e-01
-1.10233223e+00 -4.42079723e-01 -1.12888765e+00 1.62177533e-01
2.93320358e-01 2.35324875e-01 4.04637784e-01 4.70898390e-01
-4.28186536e-01 7.15348423e-01 -9.78747666e-01 4.79687035e-01
6.91298962e-01 1.86550647e-01 -2.25218106e-02 -1.75070599e-01
-1.65895963e+00 8.84482026e-01 3.09211791e-01 2.86065787e-01
-1.23733950e+00 -5.55336177e-01 -8.56880546e-01 -1.35985002e-01
6.07051790e-01 -7.64787078e-01 1.35444808e+00 -4.60591584e-01
-1.91909385e+00 2.46409133e-01 5.50837591e-02 -3.38817894e-01
6.80113137e-01 -2.91552484e-01 -2.37604544e-01 -6.52194172e-02
2.18724191e-01 6.24916315e-01 7.18057871e-01 -9.38764453e-01
-5.65368712e-01 -3.54243577e-01 6.61602896e-03 4.63150918e-01
-9.29537788e-02 -3.56513977e-01 -3.07922304e-01 -6.88296497e-01
-6.62237853e-02 -8.24059546e-01 -6.53513610e-01 -5.23375198e-02
-2.70111471e-01 -4.51189369e-01 -2.73611963e-01 -5.16400516e-01
1.60589516e+00 -2.02695155e+00 4.71146792e-01 2.23149672e-01
4.22312096e-02 1.27145126e-01 -2.89097670e-02 -1.36391004e-03
4.12121981e-01 9.94265378e-02 -5.21915138e-01 -4.21882063e-01
2.04538945e-02 3.69758278e-01 -6.01795763e-02 5.95262647e-02
2.75508225e-01 5.01993120e-01 -1.16295373e+00 -9.36519563e-01
7.43140131e-02 5.15524112e-02 -8.01843464e-01 3.42533618e-01
-3.76198739e-01 6.64726973e-01 -8.83542061e-01 4.82122421e-01
5.03283024e-01 -2.16549933e-01 -1.02539873e-02 -1.17512003e-01
5.55443168e-02 1.40539082e-02 -1.25327778e+00 2.05982256e+00
-4.90242213e-01 -1.18662320e-01 -3.44474077e-01 -5.36269724e-01
8.87637377e-01 3.56943071e-01 4.35068607e-01 -4.74372864e-01
2.08651170e-01 2.53149301e-01 1.01767644e-01 -4.29334611e-01
3.77915144e-01 -4.43334609e-01 -2.84522742e-01 1.02593593e-01
-6.53175125e-03 -1.72566220e-01 1.02513351e-01 1.65736079e-01
8.81595135e-01 2.95485437e-01 2.74167448e-01 -1.00922570e-01
6.71182036e-01 -1.78137794e-01 1.18329358e+00 6.04919016e-01
-4.52588201e-01 7.36672580e-01 7.60295868e-01 -2.67767936e-01
-8.18725884e-01 -1.20837009e+00 -3.20281416e-01 6.22160196e-01
1.29518405e-01 -4.39244866e-01 -6.25776708e-01 -8.72493327e-01
-1.62136719e-01 8.77315164e-01 -7.79502153e-01 -6.48280859e-01
-2.62509078e-01 -7.96001077e-01 2.22024634e-01 5.40892959e-01
3.21258068e-01 -1.10702741e+00 -3.63515109e-01 4.66289759e-01
-4.76390183e-01 -5.32459915e-01 -6.28365159e-01 -1.47787198e-01
-8.54895115e-01 -9.29454565e-01 -9.22400177e-01 -1.56508356e-01
5.25960684e-01 -4.31653082e-01 8.79068792e-01 -3.44757795e-01
-7.65765384e-02 1.73175186e-02 -2.83325493e-01 -5.13351440e-01
-4.52941746e-01 -9.74198207e-02 1.52917221e-01 1.20697655e-01
1.14124514e-01 -2.21192420e-01 -9.65825081e-01 3.21094543e-01
-9.46748435e-01 1.71520993e-01 8.99708867e-01 1.09820390e+00
1.03264451e+00 -1.37512758e-01 8.44603658e-01 -9.00629342e-01
8.76220703e-01 -5.13430953e-01 -4.21603739e-01 4.22022164e-01
-1.11581612e+00 5.29588223e-01 4.88001794e-01 -4.74371076e-01
-1.28464806e+00 -6.98675811e-02 -1.44363508e-01 -5.47318161e-01
2.94647068e-01 6.57890916e-01 -1.02402888e-01 5.63095450e-01
5.97895324e-01 7.26185590e-02 -1.57277565e-02 -2.41723761e-01
4.15774107e-01 7.02218115e-01 2.90472031e-01 -7.69701481e-01
3.01351070e-01 1.35866135e-01 1.30731627e-01 2.01671600e-01
-1.28912032e+00 -2.04843804e-01 -1.26956150e-01 -4.44262415e-01
1.10366940e+00 -1.04382479e+00 -5.21514058e-01 3.32094878e-01
-8.35871398e-01 -1.48132637e-01 -4.12826478e-01 8.70792627e-01
-7.90139258e-01 3.28940272e-01 -6.10772967e-01 -7.89542735e-01
-5.32515049e-01 -1.67650461e+00 1.07770705e+00 5.14299691e-01
-6.94939047e-02 -4.59722996e-01 4.21899825e-01 5.51042855e-01
1.16340496e-01 3.19026947e-01 7.80152202e-01 -2.82681882e-01
-4.17769819e-01 4.06870618e-02 2.01883480e-01 5.45958042e-01
6.38996959e-02 -1.90858305e-01 -5.50314009e-01 -6.97856024e-02
-2.42979228e-01 -7.13492393e-01 8.34718645e-01 5.35503745e-01
1.43643534e+00 -1.65360764e-01 -5.85035644e-02 1.84289575e-01
1.11827314e+00 1.04568280e-01 7.68836379e-01 3.04378837e-01
3.60955000e-01 2.31527269e-01 1.15652812e+00 8.14439476e-01
3.28965247e-01 5.87737739e-01 4.54479516e-01 3.76142830e-01
2.89833993e-01 -3.82032514e-01 3.17519337e-01 9.16917205e-01
-1.86289623e-01 -4.85150591e-02 -6.60491526e-01 6.02411032e-01
-2.09251308e+00 -6.90867960e-01 1.06041268e-01 2.22940016e+00
1.29902923e+00 4.72133070e-01 -8.90256390e-02 -1.18465155e-01
5.00093460e-01 6.16518855e-02 -6.81042373e-01 -2.71144480e-01
4.11504418e-01 7.03611672e-02 2.24812254e-01 5.17458439e-01
-8.06107879e-01 6.31686687e-01 5.76645899e+00 1.29081440e+00
-4.68874514e-01 1.14037290e-01 6.50569677e-01 -2.47868299e-01
-7.35792220e-01 -1.29504949e-01 -5.74532211e-01 7.76790679e-01
9.32745457e-01 -8.36016908e-02 9.53422338e-02 7.83930361e-01
4.43539947e-01 -1.23493724e-01 -8.67861032e-01 6.50396466e-01
-2.41885602e-01 -1.20226371e+00 2.63421357e-01 -8.13918561e-02
8.37881267e-01 -4.31272745e-01 1.18638866e-01 5.86660683e-01
5.12726009e-01 -9.33392048e-01 6.22608066e-01 1.21631742e+00
6.54487312e-01 -9.35735047e-01 1.03069174e+00 3.47593337e-01
-7.31167257e-01 -1.77588463e-01 -5.80892622e-01 4.46544439e-01
2.99152613e-01 6.44963384e-01 -5.56231797e-01 7.82524109e-01
4.14939672e-01 3.80570471e-01 -3.93281400e-01 1.07229018e+00
-4.39418048e-01 7.44803309e-01 3.61335874e-02 -3.43128815e-02
2.14763746e-01 -2.37014130e-01 3.19070190e-01 5.45773864e-01
4.70399916e-01 2.88010210e-01 1.36869058e-01 8.61662328e-01
-3.90025927e-03 1.46732152e-01 3.79572921e-02 6.15480132e-02
4.38890457e-01 9.50099289e-01 -1.55521661e-01 -1.31565735e-01
-2.41421983e-01 7.86497176e-01 3.52669388e-01 -1.69510141e-01
-9.25351322e-01 -2.07019478e-01 2.98176199e-01 -5.44173867e-02
-4.13319394e-02 1.40845552e-01 -5.45247644e-02 -1.01882875e+00
-1.36797950e-01 -8.50722671e-01 7.54171789e-01 -9.77487087e-01
-1.39624488e+00 3.71555626e-01 1.24984078e-01 -1.67487609e+00
-3.04843158e-01 -2.43725687e-01 -6.01913869e-01 8.85029256e-01
-1.34678054e+00 -7.65051425e-01 -1.00409769e-01 3.33591968e-01
6.99792683e-01 -1.67049617e-01 7.53459096e-01 1.74052119e-01
-6.38621151e-01 6.84875369e-01 2.05702066e-01 -1.41386196e-01
1.04539168e+00 -1.38086474e+00 -2.70538330e-01 5.33311546e-01
-3.55702996e-01 4.13809657e-01 6.30313098e-01 -6.88775241e-01
-7.57435858e-01 -9.03413951e-01 4.99520361e-01 -3.53488564e-01
3.01753789e-01 1.63839430e-01 -8.06127310e-01 3.59598190e-01
-3.84449698e-02 -4.79112491e-02 7.85151482e-01 1.66924119e-01
2.87723984e-03 -3.19933444e-01 -1.16142595e+00 6.70037031e-01
6.05827928e-01 -7.14110062e-02 -7.60592461e-01 1.37854934e-01
9.70334053e-01 -5.14543295e-01 -1.19493270e+00 6.51510119e-01
5.02764702e-01 -8.26875865e-01 6.90307379e-01 -6.33783102e-01
8.24853301e-01 -5.12683392e-01 -9.44161974e-03 -1.57939839e+00
-4.59277272e-01 -2.96179026e-01 -2.59680510e-01 1.00419593e+00
5.15048325e-01 -1.51682019e-01 7.60454595e-01 7.66027391e-01
-3.80975902e-01 -1.25578904e+00 -1.13721216e+00 -4.68849540e-01
1.63398460e-01 -2.08994150e-01 5.89904368e-01 5.73564947e-01
3.36294547e-02 2.70824522e-01 -7.56067395e-01 2.45893896e-02
6.07230127e-01 2.66676582e-03 3.15509588e-01 -1.18252003e+00
-6.22742474e-01 -2.28379950e-01 -2.50992090e-01 -7.99259841e-01
-2.58017272e-01 -6.08225048e-01 4.22332764e-01 -1.71612740e+00
3.08087438e-01 -3.68926197e-01 -7.48407185e-01 3.52663696e-01
-6.87004209e-01 -2.18824103e-01 -9.95283276e-02 1.70199811e-01
-9.01293933e-01 1.12740052e+00 1.72763836e+00 -9.95229855e-02
-1.36533827e-01 2.87916750e-01 -6.79248035e-01 5.01402497e-01
6.59501851e-01 -6.33780420e-01 -7.25086689e-01 -2.61842370e-01
4.85397339e-01 5.00424623e-01 -1.20138027e-01 -1.00363827e+00
-9.32716355e-02 -5.97712636e-01 3.70104879e-01 -5.94839931e-01
1.18719190e-01 -4.34807718e-01 -3.07579897e-02 4.79274809e-01
-7.57151484e-01 -1.81829199e-01 -1.26539916e-01 1.07253456e+00
-1.83750048e-01 -4.94345635e-01 7.77411342e-01 -2.53612399e-01
-3.05447370e-01 5.89556992e-01 -3.43351990e-01 2.81833708e-01
8.60098839e-01 2.65784115e-01 3.55958454e-02 -1.87306240e-01
-8.96439314e-01 7.83083498e-01 -6.30403459e-02 3.24746013e-01
6.75463498e-01 -1.59211123e+00 -9.56222832e-01 -1.77690044e-01
2.95727313e-01 5.62948585e-01 7.39400208e-01 7.04855263e-01
-2.01687396e-01 1.08624354e-01 -2.55697429e-01 -4.93331790e-01
-8.15251172e-01 3.57489884e-01 3.02079976e-01 -7.93430805e-01
-4.29798335e-01 6.75393879e-01 -6.52088318e-04 -3.96495968e-01
1.62003770e-01 -1.01137862e-01 -5.00713587e-01 -9.58142336e-03
5.00220776e-01 3.94216239e-01 -1.13880053e-01 -8.17389190e-02
-1.40725985e-01 2.67238468e-01 -1.88467979e-01 -3.20037454e-01
1.10203719e+00 -1.01219773e-01 2.91716725e-01 6.74120009e-01
6.97173297e-01 -3.85899618e-02 -1.57911420e+00 -3.21782678e-01
-2.92355567e-01 -4.02853817e-01 2.13900372e-01 -9.40200210e-01
-7.95066595e-01 9.26640213e-01 8.16540003e-01 -3.69846374e-01
1.00992942e+00 -8.53644684e-02 7.64373362e-01 1.11659251e-01
3.94011557e-01 -1.35532916e+00 5.39869010e-01 7.92680085e-02
8.86663318e-01 -1.24370623e+00 1.06232241e-01 -2.50116996e-02
-1.17292786e+00 8.67118239e-01 8.72628689e-01 1.23578925e-02
4.12728101e-01 -2.84853518e-01 4.92552631e-02 -2.59259064e-02
-7.92523861e-01 1.79786503e-01 4.19327855e-01 2.83978015e-01
1.78009465e-01 3.48771304e-01 -7.78128564e-01 1.16833758e+00
7.19889253e-02 2.09552318e-01 6.99656367e-01 6.54574335e-01
-4.21476424e-01 -1.26907468e+00 -1.71896424e-02 7.22329497e-01
-4.58657116e-01 7.02677015e-03 3.29911977e-01 2.93392003e-01
1.94764286e-01 7.55543530e-01 -2.35160112e-01 -4.54207510e-01
3.56833130e-01 2.12634608e-01 3.68351340e-01 -4.80060160e-01
-1.93952143e-01 5.47041818e-02 -2.00594738e-02 -6.21849358e-01
-4.34258938e-01 -8.89650464e-01 -1.49242699e+00 2.62456000e-01
-3.78514200e-01 4.68427122e-01 3.85708332e-01 9.71500456e-01
2.54683554e-01 9.35473382e-01 7.26729751e-01 -1.37205094e-01
-1.05189323e+00 -1.08522022e+00 -5.77440679e-01 2.75102645e-01
1.11914419e-01 -9.70468104e-01 -2.25267127e-01 -3.30112636e-01] | [8.485690116882324, 0.8467597365379333] |
175717d5-0543-44b5-a92f-4b0ede9aa333 | incentive-allocation-in-vertical-federated | 2307.03515 | null | https://arxiv.org/abs/2307.03515v1 | https://arxiv.org/pdf/2307.03515v1.pdf | Incentive Allocation in Vertical Federated Learning Based on Bankruptcy Problem | Vertical federated learning (VFL) is a promising approach for collaboratively training machine learning models using private data partitioned vertically across different parties. Ideally in a VFL setting, the active party (party possessing features of samples with labels) benefits by improving its machine learning model through collaboration with some passive parties (parties possessing additional features of the same samples without labels) in a privacy preserving manner. However, motivating passive parties to participate in VFL can be challenging. In this paper, we focus on the problem of allocating incentives to the passive parties by the active party based on their contributions to the VFL process. We formulate this problem as a variant of the Nucleolus game theory concept, known as the Bankruptcy Problem, and solve it using the Talmud's division rule. We evaluate our proposed method on synthetic and real-world datasets and show that it ensures fairness and stability in incentive allocation among passive parties who contribute their data to the federated model. Additionally, we compare our method to the existing solution of calculating Shapley values and show that our approach provides a more efficient solution with fewer computations. | ['Anna Wilbik', 'Frank Thuijsman', 'Marijn ten Thij', 'Afsana Khan'] | 2023-07-07 | null | null | null | null | ['fairness', 'federated-learning', 'fairness'] | ['computer-vision', 'methodology', 'miscellaneous'] | [ 5.27530350e-02 6.06671751e-01 -3.47054005e-01 -4.04327482e-01
-8.52164030e-01 -1.21141899e+00 2.00396881e-01 3.46451283e-01
-6.08006299e-01 1.10761988e+00 1.59870118e-01 -3.93517077e-01
-4.34060156e-01 -1.02838171e+00 -6.61235988e-01 -1.10300183e+00
5.45176305e-02 5.11865079e-01 -2.43420288e-01 2.18502477e-01
9.71287780e-04 3.74227881e-01 -1.33873379e+00 4.06664431e-01
7.51167297e-01 9.20215487e-01 -5.76324463e-01 2.81779915e-01
-1.89322904e-01 1.06144071e+00 -4.31145877e-01 -1.13952506e+00
1.28786910e+00 -3.17447871e-01 -1.05769944e+00 -9.21922550e-02
1.83586121e-01 -1.25392258e-01 -3.68029177e-02 1.14508355e+00
3.18878919e-01 1.93157330e-01 3.69663984e-01 -1.99753070e+00
-5.64469159e-01 1.01447487e+00 -7.66490817e-01 -2.21357584e-01
-1.20720506e-01 -1.38585076e-01 1.31299472e+00 -2.16846704e-01
8.73871684e-01 8.10297310e-01 5.76291800e-01 1.03588057e+00
-1.44299722e+00 -7.14296877e-01 -5.83372526e-02 -1.87042400e-01
-8.68415356e-01 -4.36607778e-01 8.47667634e-01 -4.11115110e-01
3.82091463e-01 8.45498979e-01 3.36910486e-01 4.19054568e-01
-2.13490412e-01 9.97519314e-01 1.34679592e+00 -3.24539602e-01
6.06581807e-01 4.85232770e-01 1.56969666e-01 1.97506726e-01
7.47387230e-01 7.85450041e-02 -6.70911789e-01 -8.68629038e-01
9.86259803e-03 1.66991010e-01 -2.21792236e-01 -1.16132128e+00
-6.64373815e-01 1.11499166e+00 4.09143418e-01 -1.31612331e-01
-4.79124486e-01 3.49367922e-03 3.59515160e-01 6.56261086e-01
5.36149502e-01 2.30428144e-01 -7.95866787e-01 4.28435534e-01
-7.39298880e-01 4.42234695e-01 1.01293600e+00 8.44272792e-01
7.33807743e-01 -6.13029003e-01 -2.80005455e-01 2.20581427e-01
3.42376292e-01 -1.07355483e-01 1.20525733e-01 -1.50100219e+00
8.41472447e-01 7.40567744e-01 5.40878892e-01 -6.65397525e-01
1.38474435e-01 -3.20815295e-01 -8.11554909e-01 5.06743789e-01
7.08816051e-01 -5.39264381e-01 3.26951221e-02 2.09478569e+00
8.09626043e-01 -4.75093663e-01 4.66712087e-01 6.91853464e-01
2.90053666e-01 2.74591446e-01 2.93456540e-02 -6.48544729e-01
1.02658343e+00 -1.01202774e+00 -6.31826162e-01 2.96350926e-01
6.18508995e-01 -2.49077454e-01 3.09232056e-01 3.38566184e-01
-1.20376801e+00 2.16527015e-01 -4.99827981e-01 3.14356908e-02
-2.49238238e-01 -4.48989153e-01 8.94894540e-01 1.24901676e+00
-1.12365556e+00 8.86801243e-01 -3.44620109e-01 -1.66454688e-01
1.18152046e+00 7.49516606e-01 -8.15348506e-01 1.10712610e-01
-1.01341271e+00 6.79593682e-02 6.44389614e-02 -3.51943105e-01
-6.91075265e-01 -7.67400682e-01 -5.73355913e-01 3.15202266e-01
4.60387230e-01 -8.51512015e-01 1.22872579e+00 -1.09100378e+00
-9.91084814e-01 1.21877241e+00 1.67224482e-01 -7.50713110e-01
1.19924521e+00 4.36816335e-01 3.95730257e-01 -2.12132767e-01
5.67051619e-02 3.89482349e-01 3.49435389e-01 -1.41726112e+00
-1.15307951e+00 -9.11805391e-01 5.28980494e-01 1.35852292e-01
-4.55730468e-01 6.57118931e-02 2.30044380e-01 -3.02252859e-01
-4.65885252e-02 -7.23281443e-01 -4.57430124e-01 4.49780762e-01
-6.07106462e-02 -4.13777858e-01 9.33879435e-01 -2.46737391e-01
7.63000786e-01 -1.90620136e+00 -1.93664312e-01 2.28737950e-01
6.26685202e-01 -5.24504436e-03 2.44930089e-01 4.80602622e-01
3.09296846e-01 4.94968981e-01 -2.72325486e-01 -5.73366821e-01
3.06749970e-01 1.89706326e-01 -5.24122529e-02 7.21136153e-01
-6.18782222e-01 7.71835685e-01 -8.07455301e-01 -5.85061610e-01
-3.12461346e-01 -2.04414874e-01 -8.93713713e-01 3.50361206e-02
9.79646295e-02 3.20274025e-01 -3.87487084e-01 3.22019994e-01
1.20831025e+00 -1.07360803e-01 5.03262162e-01 3.21022838e-01
-9.04409681e-03 -2.18437493e-01 -1.31691265e+00 1.27998173e+00
-1.61639303e-01 1.99684680e-01 7.36096323e-01 -8.87256205e-01
6.68901265e-01 5.66541851e-01 6.76052451e-01 -1.16135262e-01
4.33511108e-01 1.78429484e-02 -2.97674000e-01 -1.95487425e-01
2.78281987e-01 -5.74898660e-01 -1.44849315e-01 1.14211500e+00
5.00554964e-02 4.49507803e-01 -1.21315233e-01 3.56537074e-01
9.66013134e-01 -2.41510436e-01 5.97468376e-01 -2.87048310e-01
5.56925654e-01 -1.88200489e-01 1.03536963e+00 9.41631079e-01
-8.25978518e-01 3.31533030e-02 8.14983428e-01 -4.05325115e-01
-1.02012205e+00 -9.02800858e-01 2.88087010e-01 1.07382381e+00
3.70745314e-03 -6.47249147e-02 -8.57544780e-01 -1.34308755e+00
5.38014352e-01 3.43711853e-01 -8.07634830e-01 1.74784690e-01
1.47050336e-01 -4.92996573e-01 3.54721129e-01 1.77949920e-01
7.20885098e-01 -8.20737541e-01 -6.31072581e-01 -1.22331485e-01
-1.73638776e-01 -4.69072521e-01 -5.18259645e-01 3.77045780e-01
-8.73106718e-01 -1.25329125e+00 -4.55620825e-01 -5.25985837e-01
8.70729148e-01 2.94716954e-01 6.75402284e-01 -1.00634806e-01
8.58492311e-03 1.09995946e-01 -2.04252645e-01 -4.72799182e-01
-2.95037687e-01 -9.77593511e-02 4.52433154e-02 4.95523632e-01
4.89551544e-01 -3.88552248e-01 -7.79296398e-01 2.11456835e-01
-9.35973585e-01 -1.21559449e-01 -1.87144771e-01 8.49399686e-01
3.91958863e-01 2.96676636e-01 7.14726686e-01 -1.70159590e+00
7.70344436e-01 -5.63267052e-01 -8.82661104e-01 3.71746987e-01
-8.05341601e-01 5.23718856e-02 8.85122120e-01 1.45661738e-02
-1.29158843e+00 3.63742799e-01 4.86685336e-01 -2.59416252e-01
1.72049463e-01 -4.18010354e-02 -7.47362256e-01 -4.07867104e-01
4.99176621e-01 -1.32257000e-01 3.95910032e-02 -5.24725139e-01
5.04737556e-01 1.03583503e+00 4.82136309e-02 -7.04320729e-01
6.98704183e-01 7.58876622e-01 1.88458025e-01 3.85306448e-01
-6.59657001e-01 -3.70897233e-01 -3.86987746e-01 -1.33818194e-01
4.49761093e-01 -6.28265917e-01 -1.56402683e+00 3.99539351e-01
-7.93186426e-01 1.97718099e-01 -9.87597048e-01 1.92718938e-01
-6.86366081e-01 2.58855343e-01 -4.60657865e-01 -1.37033617e+00
-5.26315987e-01 -6.76388144e-01 3.72580439e-01 4.13958639e-01
1.80554405e-01 -8.14069211e-01 1.83151245e-01 1.16481459e+00
3.44158143e-01 4.89047170e-01 6.92202568e-01 -9.45752323e-01
-5.52898467e-01 -3.72577429e-01 2.80901249e-02 3.12425017e-01
2.49667838e-01 -3.88214320e-01 -1.27586079e+00 -6.80684566e-01
4.45033997e-01 -4.16005433e-01 5.55484116e-01 2.02933233e-03
1.10026634e+00 -9.46618259e-01 -1.47710547e-01 4.35645461e-01
1.68885028e+00 9.37151462e-02 5.54851182e-02 1.80599004e-01
2.06840783e-01 1.13530982e+00 5.35136044e-01 9.39391255e-01
3.97067249e-01 1.23337388e-01 5.83845675e-01 1.47241607e-01
5.80919862e-01 -4.24071193e-01 -1.49735212e-01 7.14334324e-02
-1.96412485e-03 -2.35123187e-01 -5.04874825e-01 5.13618290e-01
-2.12511849e+00 -1.19010198e+00 -7.11345747e-02 2.38529396e+00
8.74975562e-01 -3.53469253e-01 2.44299829e-01 1.94699585e-01
8.82021189e-01 -4.62389365e-02 -6.38576627e-01 -7.04484105e-01
-2.44330809e-01 7.46836290e-02 9.35895145e-01 5.17971873e-01
-1.03717065e+00 5.74760139e-01 5.44055176e+00 6.63250923e-01
-4.49556679e-01 5.82409203e-01 1.11029875e+00 -4.97128993e-01
-7.49886692e-01 3.20127368e-01 -4.01817530e-01 2.77518690e-01
4.12984312e-01 -1.01637447e+00 7.00439334e-01 9.07853603e-01
9.37882587e-02 4.16097306e-02 -1.20606565e+00 7.52063572e-01
-2.87662566e-01 -1.51571035e+00 -1.44974440e-01 4.55140024e-01
1.25315857e+00 -2.52716660e-01 -7.58659244e-02 9.65768397e-02
9.58176672e-01 -6.66366637e-01 7.34591305e-01 2.76818007e-01
4.16973025e-01 -1.13639092e+00 6.87988400e-01 8.72760653e-01
-8.70815694e-01 -6.59699380e-01 -3.15767139e-01 -2.54643440e-01
-3.10542464e-01 2.82385558e-01 -3.18596572e-01 8.23396325e-01
6.58770323e-01 -3.73373218e-02 -1.95454925e-01 9.87835407e-01
-3.61993164e-02 3.19732249e-01 -3.02781314e-02 9.50311124e-02
1.48017304e-02 -4.79722291e-01 1.90193474e-01 4.59508568e-01
-4.72892411e-02 4.00798559e-01 -2.38833413e-01 8.53792727e-01
-8.72700453e-01 4.29677367e-01 -8.01809251e-01 1.71159312e-01
5.10079443e-01 1.36974120e+00 -5.42624414e-01 -6.00748807e-02
-3.87206614e-01 8.03194284e-01 5.21161795e-01 8.13063085e-02
-3.33942980e-01 -2.26952821e-01 8.05423677e-01 -1.10615402e-01
1.44042060e-01 5.54292738e-01 -4.02392030e-01 -1.09370255e+00
5.71659319e-02 -7.86732197e-01 9.80186105e-01 -4.21165615e-01
-1.66401374e+00 2.82664508e-01 -6.10340238e-01 -1.09944141e+00
1.73629999e-01 3.15099880e-02 -5.61666965e-01 1.04056728e+00
-1.26083457e+00 -1.14096594e+00 -5.11811255e-03 1.03189337e+00
-9.09172148e-02 -1.66091740e-01 7.93085933e-01 7.96961561e-02
-2.29999766e-01 8.47189546e-01 5.92505276e-01 -2.95543466e-02
4.26779360e-01 -1.24418902e+00 1.14708152e-02 9.20317113e-01
8.74799490e-02 6.03759825e-01 4.72553939e-01 -5.45290589e-01
-1.18006372e+00 -1.26401865e+00 1.34187543e+00 -4.27529067e-01
3.92901212e-01 -5.88267088e-01 -4.70088989e-01 7.52528191e-01
2.50344127e-01 4.13802415e-01 1.27863276e+00 -8.44456926e-02
-2.22408697e-01 -4.78109807e-01 -2.24595189e+00 2.83303380e-01
1.10508466e+00 -4.05394495e-01 -3.02205801e-01 2.98669636e-01
7.13153124e-01 7.54277315e-03 -6.25357389e-01 -4.99107055e-02
4.93978322e-01 -1.00475132e+00 3.46442580e-01 -1.19365609e+00
2.08147570e-01 -1.54810473e-01 -3.01230520e-01 -9.70349729e-01
-1.94196358e-01 -1.06071770e+00 1.14371836e-01 1.63751352e+00
2.53245056e-01 -1.03453374e+00 1.64215624e+00 1.31519043e+00
8.30480516e-01 -6.29790366e-01 -1.40839636e+00 -5.93187869e-01
4.85416651e-01 -5.14613204e-02 1.06060517e+00 1.24144721e+00
3.07113111e-01 -3.68733048e-01 -4.45975512e-01 2.31058169e-02
1.36405599e+00 5.49556494e-01 8.04604709e-01 -1.44718921e+00
-3.86418849e-01 -5.98768033e-02 -4.74003032e-02 -3.24750036e-01
2.71523088e-01 -1.31502163e+00 -3.17480564e-01 -1.32147849e+00
6.84831917e-01 -8.06135535e-01 -5.10270059e-01 7.98641264e-01
1.07918866e-01 3.81954573e-02 6.74219787e-01 1.98066205e-01
-6.70429587e-01 4.09895241e-01 1.03171861e+00 -2.53119141e-01
-6.32052273e-02 5.55612266e-01 -1.39431751e+00 3.72690171e-01
8.15435469e-01 -9.11791921e-01 -4.63043898e-01 1.74155682e-02
2.54644096e-01 4.34200734e-01 1.81878522e-01 -4.91834193e-01
6.67752445e-01 -3.56015027e-01 5.01835868e-02 -3.35251093e-01
-2.72518873e-01 -1.23384821e+00 5.60481131e-01 6.26196921e-01
-7.60166109e-01 -4.14287806e-01 -3.66247654e-01 7.04991996e-01
7.41720200e-04 -3.91633838e-01 7.61939108e-01 -2.73847371e-01
2.60602176e-01 5.67864954e-01 -1.36767611e-01 1.24445111e-01
1.68433082e+00 -8.70231241e-02 -3.89603108e-01 -3.98895323e-01
-7.79870570e-01 6.79951906e-01 8.22170556e-01 4.55794558e-02
8.56546909e-02 -1.26991570e+00 -6.46266282e-01 1.82326630e-01
4.28242944e-02 -1.42538309e-01 3.39154661e-01 5.10227501e-01
-2.25616679e-01 3.92873973e-01 -3.62598717e-01 9.18265507e-02
-1.78732812e+00 8.59181404e-01 3.87924284e-01 -5.71910203e-01
-2.68165544e-02 9.16636169e-01 1.61801532e-01 -7.40199924e-01
3.49374980e-01 4.35335547e-01 -7.33436048e-02 3.03700536e-01
1.09014466e-01 5.59851527e-01 -1.71018131e-02 -4.71612036e-01
-3.22104305e-01 -9.61804911e-02 6.19892310e-03 -2.24875122e-01
1.51718223e+00 -2.18803540e-01 -4.38295215e-01 -7.24175870e-02
1.09683728e+00 3.42434168e-01 -1.17665100e+00 -4.24332589e-01
-6.25458211e-02 -9.30411220e-01 -2.12173358e-01 -6.81900978e-01
-1.46092010e+00 4.44199562e-01 4.52853799e-01 3.87195379e-01
1.16995668e+00 3.99932750e-02 3.85390759e-01 3.74222808e-02
1.00695014e+00 -9.66341555e-01 -7.42152572e-01 -2.01454043e-01
3.11077863e-01 -1.18312550e+00 -1.60356805e-01 -3.04365546e-01
-7.27025390e-01 5.96731782e-01 5.01911998e-01 1.33928806e-01
6.52552783e-01 1.90879285e-01 1.66265994e-01 1.36571079e-01
-9.96510208e-01 2.84634829e-01 -4.63034302e-01 6.30765676e-01
-4.90675047e-02 4.12922591e-01 -7.94706285e-01 1.07613623e+00
-9.14011747e-02 2.49454036e-01 5.84472299e-01 1.11342740e+00
2.76402533e-02 -1.62507200e+00 -3.27433765e-01 6.12615407e-01
-8.08857858e-01 1.74775973e-01 -7.33069479e-01 3.77856731e-01
5.75212300e-01 1.09126949e+00 -1.36093751e-01 -7.56108165e-02
5.66253848e-02 7.56835490e-02 2.83098370e-01 -4.04303968e-01
-1.21714342e+00 -3.71525854e-01 -2.02508882e-01 -3.81301820e-01
-5.83218336e-01 -8.32929373e-01 -9.82593119e-01 -7.14900911e-01
-4.87866551e-01 7.93164432e-01 4.70966578e-01 5.18640041e-01
2.89654434e-01 -7.16670156e-02 1.41205013e+00 1.03261918e-02
-9.28385913e-01 -2.64819145e-01 -1.21990621e+00 6.63607061e-01
1.68178201e-01 3.70401191e-04 -7.75080740e-01 -1.62066057e-01] | [5.8807902336120605, 6.516333103179932] |
4e258202-2f99-4d6c-bafa-376014c6be6b | a-novel-self-knowledge-distillation-approach | 2209.01311 | null | https://arxiv.org/abs/2209.01311v1 | https://arxiv.org/pdf/2209.01311v1.pdf | A Novel Self-Knowledge Distillation Approach with Siamese Representation Learning for Action Recognition | Knowledge distillation is an effective transfer of knowledge from a heavy network (teacher) to a small network (student) to boost students' performance. Self-knowledge distillation, the special case of knowledge distillation, has been proposed to remove the large teacher network training process while preserving the student's performance. This paper introduces a novel Self-knowledge distillation approach via Siamese representation learning, which minimizes the difference between two representation vectors of the two different views from a given sample. Our proposed method, SKD-SRL, utilizes both soft label distillation and the similarity of representation vectors. Therefore, SKD-SRL can generate more consistent predictions and representations in various views of the same data point. Our benchmark has been evaluated on various standard datasets. The experimental results have shown that SKD-SRL significantly improves the accuracy compared to existing supervised learning and knowledge distillation methods regardless of the networks. | ['Jia-Ching Wang', 'Trang Phung', 'Duc-Quang Vu'] | 2022-09-03 | null | null | null | null | ['self-knowledge-distillation'] | ['computer-vision'] | [ 2.12926537e-01 3.28520179e-01 -4.78884071e-01 -2.96184003e-01
-3.23784471e-01 -4.25378859e-01 4.87868547e-01 5.83274886e-02
-2.88634449e-01 1.17699099e+00 1.40560910e-01 -9.77245569e-02
-3.90435934e-01 -8.95218790e-01 -6.29369676e-01 -8.15344036e-01
4.72966015e-01 3.83501709e-01 5.21560252e-01 -2.12308347e-01
1.16769664e-01 5.73659956e-01 -1.56856847e+00 1.29921138e-01
1.31747282e+00 8.67967069e-01 -7.47394701e-03 2.29352728e-01
-3.25033575e-01 1.03531802e+00 -5.73443770e-01 -4.37122852e-01
2.25238234e-01 -4.22703117e-01 -7.51655400e-01 -3.01755935e-01
9.24528420e-01 4.61425185e-02 -4.09145445e-01 1.12767017e+00
5.79548359e-01 6.14140987e-01 8.68186057e-01 -1.24823058e+00
-1.13919330e+00 6.92181230e-01 -7.56689370e-01 3.99570376e-01
2.27167457e-01 -2.14581221e-01 5.93574405e-01 -7.52057135e-01
5.79666376e-01 1.23370457e+00 5.94063938e-01 5.39901435e-01
-1.38597703e+00 -1.16888869e+00 8.48239660e-02 4.72670019e-01
-1.53569400e+00 1.33533135e-01 1.02183425e+00 -5.50547302e-01
5.07141531e-01 -3.96431945e-02 7.38523781e-01 9.10537004e-01
-3.00442725e-01 1.02261293e+00 1.33364666e+00 -4.80310619e-01
5.30373603e-02 6.54689848e-01 3.54290694e-01 8.52570534e-01
5.37317395e-01 9.35099348e-02 -6.15056813e-01 -3.79365236e-02
5.82058012e-01 2.09808826e-01 -3.26508492e-01 -1.00512445e+00
-9.05754685e-01 7.67669380e-01 6.55766129e-01 9.82944220e-02
-2.86497194e-02 -2.46830150e-01 3.18720549e-01 5.63577890e-01
3.60569805e-01 4.70317811e-01 -5.91528952e-01 -6.87650666e-02
-9.18649256e-01 -2.31015813e-02 6.87276483e-01 9.05176103e-01
8.75674427e-01 4.41235155e-01 -2.65557855e-01 9.10799861e-01
1.54968128e-01 5.23014307e-01 1.07911313e+00 -6.38668656e-01
4.47879493e-01 1.17979753e+00 -4.14206386e-01 -1.05998349e+00
1.48038462e-01 -5.85459828e-01 -8.50890815e-01 3.53657901e-01
-6.89728633e-02 -4.16768342e-01 -9.58377779e-01 1.72901952e+00
6.64259613e-01 9.77849662e-01 5.14964461e-01 4.04038996e-01
1.35669744e+00 5.17144859e-01 -7.92604536e-02 -1.86926335e-01
7.84858346e-01 -1.11943936e+00 -5.68872929e-01 1.64595395e-01
7.05247819e-01 -6.38819098e-01 6.87712193e-01 5.10154605e-01
-8.98829937e-01 -7.21064210e-01 -1.07839715e+00 1.80608779e-01
-7.83767760e-01 6.27966896e-02 4.34891462e-01 5.16815603e-01
-7.28467345e-01 8.09205115e-01 -2.94781268e-01 -1.25491628e-02
8.21851134e-01 4.67335075e-01 -6.22499108e-01 -1.87094569e-01
-1.21152294e+00 9.80266452e-01 7.40654409e-01 -5.04816890e-01
-7.76561975e-01 -1.08559561e+00 -8.01696479e-01 2.79276609e-01
4.91958857e-01 -4.24301416e-01 1.03738248e+00 -1.06221259e+00
-1.66110754e+00 7.67364442e-01 5.26995480e-01 -3.06973904e-01
3.92932445e-01 -2.23744556e-01 -2.02940449e-01 -5.03649898e-02
-1.74949244e-01 5.39264679e-01 7.57646501e-01 -1.16898823e+00
-4.34834421e-01 -2.96968699e-01 -4.23240423e-01 6.94425404e-01
-5.30129731e-01 -6.26596034e-01 -3.09710186e-02 -7.41750121e-01
-3.83800231e-02 -8.18437397e-01 -5.15716383e-03 -3.40251505e-01
-1.44972742e-01 -6.40992641e-01 1.30698657e+00 -2.77976722e-01
1.07486153e+00 -1.94558513e+00 2.03390777e-01 4.13490742e-01
1.75396949e-01 1.07222199e+00 -2.48766080e-01 -5.42817377e-02
-6.10748947e-01 -1.38821483e-01 1.10918835e-01 8.60250369e-02
-4.25805926e-01 4.67470735e-01 -2.44888499e-01 2.60295600e-01
-1.87403128e-01 7.46048033e-01 -1.24113035e+00 -6.78116262e-01
1.78176403e-01 4.58157837e-01 -4.10097122e-01 3.69542181e-01
1.90519929e-01 1.67563275e-01 -4.91419435e-01 2.95248151e-01
7.08333194e-01 -1.14029832e-01 8.81082639e-02 -3.84470597e-02
2.52035111e-01 -4.54756292e-03 -1.32003689e+00 1.65626466e+00
-3.34578216e-01 4.64406520e-01 -4.11204636e-01 -1.29345930e+00
1.35246778e+00 2.59655267e-01 4.37532276e-01 -4.79652673e-01
1.23471014e-01 -3.16747651e-02 7.43368566e-02 -3.68241757e-01
3.79638672e-01 -1.23471029e-01 2.23795667e-01 5.01054525e-01
4.67731029e-01 -2.92350203e-01 2.99766976e-02 3.97075117e-01
6.42015755e-01 3.88438374e-01 4.34030861e-01 -8.49872231e-02
6.65210426e-01 -2.19080165e-01 7.01992989e-01 5.10620177e-01
-3.83398086e-01 1.74193531e-01 2.22716406e-01 -2.80333221e-01
-6.05340660e-01 -1.28308427e+00 9.96659398e-02 1.13293815e+00
1.57954827e-01 -1.19296886e-01 -5.24854541e-01 -1.38715065e+00
4.16791826e-01 7.39156246e-01 -7.08993316e-01 -8.01873326e-01
-3.77536565e-01 -1.53005823e-01 3.98501903e-01 5.65746307e-01
6.53015494e-01 -7.72079229e-01 5.56885228e-02 -1.60704926e-01
2.76592880e-01 -6.81243956e-01 -3.66511792e-01 6.14715703e-02
-9.86291230e-01 -1.14106524e+00 -7.89899111e-01 -9.37823415e-01
1.02129865e+00 6.51325822e-01 8.70160580e-01 -2.66313016e-01
-1.49155006e-01 4.08115327e-01 -2.50598013e-01 -3.42070729e-01
-9.37408730e-02 1.10068515e-01 3.33893836e-01 -6.25720769e-02
6.86690390e-01 -9.37400281e-01 -3.54909927e-01 1.57749534e-01
-7.67898023e-01 4.07056361e-02 7.98178375e-01 9.85713601e-01
8.55718136e-01 2.24903062e-01 8.42473447e-01 -1.25098729e+00
7.75719106e-01 -5.88460445e-01 -4.86700207e-01 5.73877156e-01
-1.15741026e+00 3.50189835e-01 8.38125825e-01 -8.56802404e-01
-1.36585784e+00 4.15362678e-02 5.07134378e-01 -9.92364824e-01
-1.13681853e-01 3.69085252e-01 8.11237395e-02 -4.83026326e-01
8.61652613e-01 5.21154344e-01 5.26035652e-02 -5.87743938e-01
6.96711063e-01 5.22411585e-01 6.41730487e-01 -7.10571706e-01
1.01888394e+00 9.52538922e-02 8.70947540e-02 -4.67752337e-01
-1.24429524e+00 -4.65449393e-01 -8.07702720e-01 -2.08468596e-03
4.99040574e-01 -1.05431402e+00 -6.74264014e-01 2.69563556e-01
-6.31507814e-01 -3.48356850e-02 -8.88248563e-01 7.90237904e-01
-2.63050228e-01 3.39011550e-01 2.04031751e-03 -3.47917140e-01
-3.24981958e-01 -9.34180796e-01 2.16402501e-01 8.89132023e-01
1.49036914e-01 -1.20845401e+00 5.38850188e-01 7.02774286e-01
3.20258379e-01 1.25727564e-01 6.46480918e-01 -1.25510204e+00
-2.56938517e-01 -2.52709538e-01 -2.52330691e-01 8.51497054e-01
2.51179755e-01 -4.88196611e-02 -8.63921165e-01 -3.25920463e-01
-2.90956825e-01 -8.26070845e-01 9.77235794e-01 -8.27310905e-02
1.24304020e+00 -4.12625551e-01 -2.69808203e-01 4.92002636e-01
1.26538944e+00 1.54415742e-01 4.16371018e-01 1.68500215e-01
1.00296474e+00 2.55738527e-01 4.50176209e-01 1.44772567e-02
2.26568297e-01 3.90290380e-01 1.15262598e-01 3.85484509e-02
-5.24297476e-01 -5.74566483e-01 3.11212838e-01 1.28592288e+00
-2.02820748e-01 2.59222388e-01 -7.49764144e-01 6.67981863e-01
-1.69608450e+00 -9.23074126e-01 1.96149439e-01 2.14490104e+00
1.27779913e+00 -1.03233136e-01 -1.99698269e-01 -5.97330295e-02
7.14887023e-01 7.51040429e-02 -6.62092090e-01 -5.76244891e-01
7.43288845e-02 5.43771744e-01 4.16698068e-01 4.39745724e-01
-7.68787801e-01 1.01202846e+00 6.06358671e+00 1.23543739e+00
-1.06686306e+00 7.52669349e-02 3.23786646e-01 -7.76889920e-02
-1.62537947e-01 -2.40499958e-01 -1.07857502e+00 3.93452495e-01
7.32692063e-01 -4.09123033e-01 6.14062287e-02 1.27441764e+00
-2.07199335e-01 -1.12705648e-01 -9.63684499e-01 1.00159383e+00
4.44487363e-01 -1.21561968e+00 4.68069971e-01 -3.58544350e-01
1.56434810e+00 -3.15206110e-01 2.44714707e-01 8.45668435e-01
8.17891061e-01 -1.02221596e+00 -1.85751408e-01 7.40656972e-01
5.53623080e-01 -9.94913578e-01 6.22894108e-01 3.35281163e-01
-9.51279402e-01 1.54494584e-01 -3.79746199e-01 1.10272035e-01
-5.54455817e-01 5.13842821e-01 -1.28733110e+00 6.20253861e-01
3.91829014e-01 9.44308221e-01 -8.09858978e-01 9.25976574e-01
-6.62711561e-01 7.64428973e-01 2.33678170e-03 -5.51279308e-03
1.20024264e-01 -2.98236877e-01 3.26990664e-01 9.08139706e-01
3.11929226e-01 4.91396636e-02 4.27396774e-01 5.84001601e-01
-4.11255032e-01 1.19380392e-01 -8.81696582e-01 9.09731612e-02
6.52147591e-01 1.20072353e+00 -2.23114848e-01 -6.95370197e-01
-3.62579912e-01 6.56685531e-01 7.08324015e-01 1.85315058e-01
-4.93884891e-01 -6.61589563e-01 4.90577608e-01 -9.88140106e-02
4.18521106e-01 2.08379298e-01 2.26039235e-02 -1.11402810e+00
-3.89883578e-01 -6.54263020e-01 5.62144756e-01 -9.77143288e-01
-1.31317925e+00 1.91120636e-02 6.88952282e-02 -1.27015090e+00
-7.23550245e-02 -4.02546346e-01 -7.76936769e-01 9.20218766e-01
-1.89388394e+00 -8.77038300e-01 -4.70563918e-01 9.55264330e-01
1.55595586e-01 -8.30012083e-01 7.61484385e-01 1.79286793e-01
-6.60238683e-01 9.67629313e-01 4.93214339e-01 1.56288221e-01
9.12751913e-01 -1.40206444e+00 -5.07789314e-01 4.81024951e-01
5.78479879e-02 4.48817313e-01 2.98835009e-01 -6.80063963e-01
-8.90951097e-01 -1.19623685e+00 5.49777687e-01 -1.84171081e-01
4.64737505e-01 3.48845482e-01 -1.13711262e+00 7.68239021e-01
3.33946735e-01 2.20353067e-01 1.19601607e+00 7.65578747e-02
-6.88456893e-01 -4.99473214e-01 -1.29715896e+00 4.55832928e-01
4.60919201e-01 -3.59001309e-01 -1.30267048e+00 2.80613899e-01
7.31357217e-01 -3.41417253e-01 -1.17981720e+00 4.39265996e-01
3.74758214e-01 -8.52423728e-01 1.11154592e+00 -8.95383835e-01
3.85358095e-01 -2.62585700e-01 3.43994468e-01 -1.80350852e+00
-3.42444450e-01 -2.46292189e-01 -7.00792670e-01 1.26840103e+00
5.85056171e-02 -4.21354294e-01 1.24793708e+00 3.40388834e-01
1.14723317e-01 -1.03972173e+00 -7.46231973e-01 -9.97579515e-01
2.00405747e-01 3.23237300e-01 6.14738882e-01 1.72729266e+00
-3.66281047e-02 6.74989223e-01 -2.45112792e-01 -1.53596595e-01
6.69736683e-01 3.35060418e-01 7.64429152e-01 -1.59606051e+00
-1.03285886e-01 -3.11063856e-01 -6.68001771e-01 -8.94566000e-01
3.64469767e-01 -1.31237054e+00 -4.17484134e-01 -1.31038523e+00
3.85917753e-01 -4.67005461e-01 -5.97122490e-01 7.91545749e-01
-3.80279571e-01 -1.07738078e-01 -1.16110697e-01 -7.89135322e-02
-5.40285230e-01 8.25077295e-01 2.01930785e+00 -2.10795403e-01
-3.67471457e-01 5.16485795e-03 -8.97647619e-01 8.29948068e-01
8.07337880e-01 -6.60386980e-01 -1.06327569e+00 -8.57146084e-02
-7.67997876e-02 -3.30012500e-01 -1.00068040e-01 -9.97948825e-01
3.63308340e-01 -5.81669211e-01 7.23587275e-01 -6.07204378e-01
2.84029573e-01 -9.74996448e-01 -3.09168071e-01 6.08173132e-01
-5.44492602e-01 -3.27045619e-01 1.42678469e-01 6.98762655e-01
-3.19120288e-01 -5.37961066e-01 1.12039077e+00 -1.30089402e-01
-7.18969047e-01 2.23990053e-01 1.66401029e-01 3.33076537e-01
1.59627390e+00 -2.37848893e-01 -6.04986310e-01 -1.48201421e-01
-8.12150717e-01 3.35422844e-01 -5.95994107e-02 3.17425728e-01
9.23172891e-01 -1.75967515e+00 -5.58322728e-01 2.48412088e-01
-8.35063532e-02 3.89871716e-01 3.09223980e-01 5.05916059e-01
-3.41382086e-01 3.42555344e-01 -5.13345242e-01 -2.86258757e-01
-1.60018015e+00 6.14895821e-01 1.34750932e-01 -6.19670153e-01
-3.34283113e-01 1.20910394e+00 1.63844619e-02 -9.73182142e-01
4.10229713e-01 -1.33413985e-01 -6.21242106e-01 1.46218359e-01
5.18668711e-01 6.45144224e-01 -2.03352764e-01 -4.59849685e-01
1.11733064e-01 4.83410090e-01 -7.23285615e-01 4.05177295e-01
1.25890791e+00 2.80577987e-01 2.78810382e-01 2.31840357e-01
1.19054425e+00 -1.00932114e-01 -1.06021821e+00 -6.88676834e-01
-3.29627633e-01 -5.34032524e-01 1.56672284e-01 -9.24352705e-01
-1.43479562e+00 8.08883727e-01 7.00895190e-01 -1.37858167e-01
8.91628861e-01 -2.84097046e-01 6.47280216e-01 7.53899693e-01
-1.45327896e-01 -1.18311298e+00 6.24583423e-01 5.96744835e-01
5.12430072e-01 -1.26273572e+00 4.31191176e-01 -3.19225132e-01
-6.50801182e-01 9.01484132e-01 1.37375474e+00 -2.59166956e-01
7.47697771e-01 -2.54291356e-01 -1.47608995e-01 -4.23116144e-03
-6.91333771e-01 3.07331476e-02 5.57976127e-01 7.66758442e-01
2.60995805e-01 1.52720168e-01 -3.44912373e-02 6.00528955e-01
-2.29064047e-01 1.84992373e-01 5.28997600e-01 1.11127806e+00
-7.44615316e-01 -1.10585618e+00 -3.32320094e-01 5.06246507e-01
3.65806073e-02 7.76884258e-02 -6.57650471e-01 6.90607011e-01
4.19054836e-01 4.10359889e-01 -3.24647278e-01 -5.41674852e-01
3.18000942e-01 3.83994043e-01 6.06185794e-01 -9.62062180e-01
-5.89651465e-01 -5.94778001e-01 -5.13016701e-01 -2.20220730e-01
-5.99859774e-01 -2.29685962e-01 -1.21215582e+00 -3.01571190e-01
-4.96471703e-01 5.17752647e-01 3.60157818e-01 8.08196247e-01
4.02689070e-01 8.55210066e-01 6.91978157e-01 -2.50331819e-01
-7.98797905e-01 -8.16362798e-01 -8.04446161e-01 4.60364252e-01
-6.48729445e-04 -9.87864733e-01 -3.41437846e-01 7.80375153e-02] | [9.504691123962402, 3.3622353076934814] |
16e3954a-0526-4e91-8b93-0912f844ee3d | progressive-continual-learning-for-spoken | 2201.12546 | null | https://arxiv.org/abs/2201.12546v2 | https://arxiv.org/pdf/2201.12546v2.pdf | Progressive Continual Learning for Spoken Keyword Spotting | Catastrophic forgetting is a thorny challenge when updating keyword spotting (KWS) models after deployment. To tackle such challenges, we propose a progressive continual learning strategy for small-footprint spoken keyword spotting (PCL-KWS). Specifically, the proposed PCL-KWS framework introduces a network instantiator to generate the task-specific sub-networks for remembering previously learned keywords. As a result, the PCL-KWS approach incrementally learns new keywords without forgetting prior knowledge. Besides, the keyword-aware network scaling mechanism of PCL-KWS constrains the growth of model parameters while achieving high performance. Experimental results show that after learning five new tasks sequentially, our proposed PCL-KWS approach archives the new state-of-the-art performance of 92.8% average accuracy for all the tasks on Google Speech Command dataset compared with other baselines. | ['Nancy F. Chen', 'Nana Hou', 'Yizheng Huang'] | 2022-01-29 | null | null | null | null | ['keyword-spotting'] | ['speech'] | [-1.89744290e-02 8.96888226e-03 -3.45917642e-01 -3.50972682e-01
-7.29742408e-01 -2.70752370e-01 3.69141221e-01 -1.90997079e-01
-7.76736200e-01 7.55809426e-01 1.82984337e-01 -5.38513601e-01
-1.45753413e-01 -1.88074842e-01 -7.81716228e-01 -3.41552466e-01
-5.36926836e-02 3.79216611e-01 8.10396075e-01 -2.50687242e-01
1.53170437e-01 2.62995422e-01 -1.23733115e+00 1.29800811e-01
9.32903349e-01 8.88793945e-01 7.75914133e-01 6.58931911e-01
-2.83433080e-01 6.08566940e-01 -6.04429126e-01 -2.43826076e-01
-1.82688758e-02 1.13247216e-01 -7.42446840e-01 -3.83546889e-01
3.15666288e-01 -5.73590100e-01 -9.52332199e-01 5.79954207e-01
5.19329965e-01 3.90532136e-01 8.29814151e-02 -1.30904353e+00
-6.13605618e-01 1.32930970e+00 -3.03652644e-01 6.30394101e-01
8.30693990e-02 4.25988317e-01 7.89378583e-01 -1.40174735e+00
2.04217792e-01 1.06104720e+00 8.45580041e-01 6.54713869e-01
-9.64668155e-01 -1.11203909e+00 8.63750696e-01 5.11824727e-01
-1.81625533e+00 -8.58684719e-01 5.31824350e-01 2.02275872e-01
1.28326631e+00 3.45895022e-01 6.28266931e-01 1.29844177e+00
-1.13926165e-01 1.23254490e+00 7.05382466e-01 -3.01412195e-01
2.26662785e-01 1.61016002e-01 3.30145389e-01 5.65764964e-01
5.74303046e-02 -4.49898332e-01 -1.38200629e+00 -2.46007144e-01
2.91538596e-01 -4.05743904e-02 -5.18461406e-01 1.71448104e-02
-1.11684382e+00 1.86002761e-01 -1.23601668e-01 1.63248390e-01
-4.83838171e-01 2.57529140e-01 2.84973145e-01 3.77910554e-01
6.38165891e-01 3.84958535e-01 -9.45189297e-01 -5.38665235e-01
-1.26209378e+00 1.25072449e-01 7.94939876e-01 1.19311178e+00
7.29550183e-01 4.50807810e-01 -4.58044976e-01 1.08654571e+00
-1.89190209e-02 3.74458283e-01 9.18986201e-01 -6.65172875e-01
4.87826288e-01 2.41499305e-01 -1.44356079e-02 -6.82262599e-01
-3.11832488e-01 -7.27141082e-01 -5.50352812e-01 -7.79819429e-01
-2.75835812e-01 -1.32057279e-01 -1.10343981e+00 1.86575794e+00
1.95710525e-01 8.24899614e-01 -9.95998681e-02 4.34538305e-01
6.67228401e-01 7.57547557e-01 2.54085213e-02 -4.99798447e-01
9.42464471e-01 -1.42473841e+00 -8.53671432e-01 -5.83802938e-01
2.41315976e-01 -4.41979825e-01 1.86881125e+00 4.35064554e-01
-8.83331835e-01 -2.13520586e-01 -1.02046692e+00 3.38789999e-01
-2.92695284e-01 1.31059423e-01 5.28460801e-01 4.15595710e-01
-1.34245884e+00 2.92508096e-01 -7.22555399e-01 -2.73048192e-01
1.28884509e-01 2.96988249e-01 -2.46543400e-02 -5.02875187e-02
-1.44807446e+00 6.66931987e-01 7.13086069e-01 1.21963926e-01
-1.40815234e+00 -1.21269262e+00 -3.75821829e-01 3.80902857e-01
1.13297617e+00 -4.31754291e-01 1.82783091e+00 -4.09480274e-01
-1.94156516e+00 1.26583397e-01 -4.42881197e-01 -5.69838643e-01
3.31614077e-01 -5.60314536e-01 -6.83161199e-01 3.22451070e-02
-2.55275518e-01 5.79934120e-01 1.24696004e+00 -1.20362306e+00
-6.89691782e-01 9.33032110e-02 -2.24272180e-02 4.22495335e-01
-1.06180477e+00 -2.98939735e-01 -9.02403533e-01 -9.67362285e-01
-4.83215898e-02 -6.36180699e-01 7.78066143e-02 -3.92785013e-01
-4.67953056e-01 -6.28976643e-01 9.90295827e-01 -7.89484143e-01
2.01567101e+00 -2.15801764e+00 5.03170677e-03 -4.53022821e-03
4.24410850e-01 6.80962324e-01 -4.75704432e-01 4.49415028e-01
2.61183918e-01 1.52609438e-01 9.64477751e-03 -8.66986394e-01
-6.47482872e-02 2.30210185e-01 -3.80391598e-01 -1.17463373e-01
-2.08885923e-01 1.02633286e+00 -8.37213933e-01 -1.71170980e-01
-2.48745799e-01 3.17883313e-01 -2.41221249e-01 3.21341664e-01
-4.54381585e-01 -1.15935728e-01 4.87822741e-02 6.23442471e-01
4.81641233e-01 -1.12687819e-01 9.59629193e-02 1.09772924e-02
8.46647993e-02 4.80616659e-01 -1.06697130e+00 1.99360764e+00
-5.65875351e-01 1.55858904e-01 1.94819272e-01 -5.69496036e-01
6.70859993e-01 2.91251838e-01 -1.40789032e-01 -6.72847867e-01
-3.39667052e-01 1.91609025e-01 -3.50183398e-01 -2.90856749e-01
8.48549008e-01 3.34761977e-01 1.35704204e-01 7.38796890e-01
4.08210635e-01 1.27268493e-01 -1.66123986e-01 7.91187048e-01
1.07039702e+00 -4.33411509e-01 -1.17652744e-01 -1.02107726e-01
3.73802215e-01 -3.66188079e-01 6.83874309e-01 1.24216998e+00
-7.30330646e-02 1.46394685e-01 6.46549985e-02 -2.83279747e-01
-6.14496052e-01 -8.40935111e-01 7.29156077e-01 1.59554255e+00
-1.57505050e-01 -7.37172961e-01 -7.78783619e-01 -7.05263078e-01
8.78060386e-02 1.07453489e+00 -3.48723888e-01 -7.29546487e-01
-5.33624589e-01 -2.93022275e-01 9.47975516e-01 4.05027598e-01
5.58322310e-01 -8.98146212e-01 -1.91207245e-01 3.48420471e-01
-2.47397780e-01 -1.13719559e+00 -1.24756205e+00 2.54127830e-01
-6.45946622e-01 -7.12472141e-01 -8.06748629e-01 -7.16817617e-01
3.48584443e-01 7.96716094e-01 7.52756953e-01 2.12727636e-01
-1.69328079e-02 5.51012337e-01 -4.09888625e-01 -2.76420116e-01
-1.02282539e-01 8.82827699e-01 5.33538043e-01 1.38904020e-01
1.39997989e-01 -7.88565099e-01 -5.75203896e-01 1.38559535e-01
-8.75321865e-01 9.41643938e-02 6.32118165e-01 8.39704394e-01
3.91705930e-01 1.82887521e-02 9.39547896e-01 -6.84523880e-01
1.36695492e+00 -4.43097919e-01 -2.97169179e-01 6.96525514e-01
-1.31367004e+00 1.80721730e-02 5.48631728e-01 -1.12587380e+00
-1.11381769e+00 -1.92284063e-01 5.59720732e-02 -6.31023586e-01
2.20232189e-01 7.10699320e-01 4.13858593e-02 -1.70091525e-01
4.05308813e-01 1.03242064e+00 -2.02197716e-01 -7.98688173e-01
4.77239162e-01 8.88318300e-01 7.21173346e-01 -4.47761327e-01
6.70606792e-01 -5.96841499e-02 -1.02501023e+00 -8.31100523e-01
-7.69769251e-01 -4.84237850e-01 -3.01832497e-01 -2.33794659e-01
7.90588185e-02 -1.03219008e+00 -5.75450361e-01 9.55406547e-01
-1.16518390e+00 -7.52659917e-01 -1.85198903e-01 2.76636899e-01
-1.37387499e-01 3.20655048e-01 -6.53220952e-01 -9.09241378e-01
-8.17318022e-01 -9.85968113e-01 6.06979430e-01 1.68820426e-01
-2.69381851e-01 -6.67027473e-01 -1.13578975e-01 1.74398944e-01
1.08288658e+00 -8.82404685e-01 9.37239468e-01 -8.49333346e-01
-4.00816351e-01 1.44689351e-01 -9.10503268e-02 4.46430534e-01
1.93237737e-02 -5.53300500e-01 -9.48998153e-01 -4.96232063e-01
-2.06157923e-01 -3.67133409e-01 1.02848744e+00 -2.03773290e-01
1.46208954e+00 -7.75892437e-01 -3.42451334e-01 4.75672662e-01
8.52561474e-01 3.04698676e-01 1.88175038e-01 2.02702016e-01
6.93875730e-01 -3.83828133e-02 4.84160483e-01 5.81045747e-01
8.52991223e-01 5.86186230e-01 8.22253227e-02 3.62656564e-01
-4.04407024e-01 -7.05776036e-01 4.75246072e-01 1.41847336e+00
6.12118781e-01 -2.13946506e-01 -1.10320055e+00 7.49177396e-01
-1.90514767e+00 -3.78486753e-01 6.56759262e-01 2.06297851e+00
1.48390365e+00 2.86805689e-01 -4.06959429e-02 1.03692584e-01
5.70028841e-01 4.16810989e-01 -1.12877870e+00 -4.01591510e-02
5.09346416e-03 1.69244558e-01 4.26886111e-01 8.33269179e-01
-5.93369365e-01 1.63846827e+00 6.42207766e+00 1.21605146e+00
-1.21254456e+00 5.48547626e-01 3.72821093e-01 -3.67840081e-01
-3.28761876e-01 -7.23827779e-02 -1.14693105e+00 5.97966135e-01
1.19626462e+00 -5.26282430e-01 8.50041211e-01 8.24319065e-01
2.40158468e-01 -2.27719739e-01 -6.69434547e-01 1.07059348e+00
3.59750748e-01 -1.36318684e+00 3.99040014e-01 -6.65515304e-01
3.08716327e-01 2.33276367e-01 1.50493309e-01 5.69455206e-01
3.07585478e-01 -6.53374374e-01 9.65108693e-01 6.60453439e-01
9.84537661e-01 -7.42903292e-01 2.38096148e-01 7.56252170e-01
-9.37579215e-01 -6.83900192e-02 9.79676843e-02 1.43725291e-01
2.70351529e-01 6.31600916e-01 -1.37141705e+00 1.51858151e-01
9.13552880e-01 1.62352473e-01 -7.51284361e-01 9.96951640e-01
-2.61110187e-01 1.16429162e+00 -4.88768011e-01 -1.19652286e-01
2.03423083e-01 6.62955165e-01 5.57332158e-01 1.34869778e+00
8.77370536e-02 1.16619028e-01 1.44877151e-01 4.94922578e-01
-4.49836254e-01 -6.64541125e-02 -2.26372018e-01 -1.60075858e-01
1.33329451e+00 1.10564208e+00 -4.73735094e-01 -4.60226774e-01
-1.71673428e-02 1.20235133e+00 5.86456239e-01 6.62191868e-01
-6.80322170e-01 -5.66146195e-01 6.08893394e-01 2.15553030e-01
3.16175073e-01 -5.27650893e-01 2.26853400e-01 -1.09262538e+00
2.40412027e-01 -9.14210320e-01 1.29695997e-01 -9.34273720e-01
-8.81815672e-01 7.44739056e-01 1.85488254e-01 -2.99593091e-01
8.26693848e-02 1.06154874e-01 -5.06517231e-01 5.70477545e-01
-1.77979743e+00 -1.01284623e+00 -3.67716193e-01 7.50406086e-01
1.08950090e+00 -3.84056658e-01 6.78625882e-01 3.84316653e-01
-6.37996912e-01 1.16374123e+00 -1.38475433e-01 -4.51644808e-01
9.17912841e-01 -8.38994563e-01 7.88720071e-01 8.97407591e-01
1.94462851e-01 8.31367970e-01 4.63996589e-01 -9.39918220e-01
-1.53644550e+00 -1.08431578e+00 9.45005476e-01 2.57249177e-02
7.23746896e-01 -8.30093622e-01 -1.04649246e+00 6.90217257e-01
3.32749933e-01 -1.93747282e-01 3.35829288e-01 7.33426884e-02
-6.88211381e-01 -4.56165075e-01 -6.98241532e-01 5.96014977e-01
1.22719002e+00 -7.92060971e-01 -4.68737513e-01 4.39174026e-01
1.69616544e+00 -4.24971253e-01 -3.66423696e-01 2.88620710e-01
4.35961306e-01 -5.25649965e-01 8.04013371e-01 -6.45257771e-01
-5.25222421e-01 -9.93015915e-02 1.00236796e-01 -1.62191379e+00
-2.00383723e-01 -1.23648620e+00 -7.37919390e-01 1.21054399e+00
7.28915095e-01 -7.46410191e-01 6.63581729e-01 4.56820905e-01
-3.04442585e-01 -9.85122144e-01 -1.13579249e+00 -9.01923835e-01
-4.57155287e-01 -6.89923644e-01 7.65469491e-01 9.45941925e-01
-2.02627212e-01 3.28979909e-01 -7.70853758e-01 1.58517227e-01
3.27954292e-01 -6.70518219e-01 6.41074061e-01 -8.95537972e-01
-3.46121907e-01 -1.13566160e-01 3.66934508e-01 -1.33564949e+00
2.74772674e-01 -6.98198199e-01 2.06771508e-01 -1.22189462e+00
7.34333247e-02 -5.94633281e-01 -5.35563827e-01 9.37972248e-01
-3.69648159e-01 -3.14100564e-01 2.44905010e-01 3.04358751e-01
-1.09624588e+00 8.56597066e-01 7.36643970e-01 -1.62336215e-01
-3.99707049e-01 -9.69873592e-02 -7.83361316e-01 3.88486952e-01
9.15067136e-01 -6.10253811e-01 -8.55135977e-01 -7.51236737e-01
2.57723391e-01 -3.40897501e-01 -1.90925878e-02 -9.28574741e-01
1.08606815e+00 -1.59887940e-01 -3.02011847e-01 -7.50347376e-01
4.54941988e-01 -5.33114016e-01 -2.60933675e-03 2.16235623e-01
-4.27973688e-01 3.97216320e-01 3.78166109e-01 8.82378936e-01
-1.00985589e-03 -4.52894494e-02 3.17447394e-01 -1.53516412e-01
-9.46363032e-01 4.37311053e-01 -5.23703039e-01 1.62489250e-01
7.94956565e-01 4.14089225e-02 -3.35607558e-01 -5.09659350e-01
-7.13942111e-01 6.42493248e-01 -1.74238399e-01 9.26111460e-01
1.07322478e+00 -1.02854514e+00 -2.46304512e-01 2.20302299e-01
1.56278431e-01 1.04615875e-01 3.08675021e-01 8.26137543e-01
-2.48716474e-02 5.08893132e-01 4.74217772e-01 -2.66165227e-01
-1.20070195e+00 1.93460971e-01 8.35967734e-02 -1.39077678e-01
-6.24032378e-01 1.33156466e+00 -3.70927095e-01 -5.29489696e-01
1.05455160e+00 -3.45226735e-01 1.75912783e-01 -1.04103731e-02
7.26239085e-01 3.66638362e-01 3.55060756e-01 1.00640088e-01
-2.78852493e-01 -1.48820162e-01 -8.34385395e-01 -2.90146172e-01
1.49699926e+00 -4.88561869e-01 -4.52783518e-02 8.21089864e-01
5.73764503e-01 -2.15175822e-01 -1.33364880e+00 -9.34470057e-01
2.53304064e-01 -1.79640964e-01 2.69112349e-01 -1.31669223e+00
-1.00683224e+00 4.49377149e-01 3.02242488e-01 -3.85162532e-01
1.05621135e+00 -5.02509438e-02 1.21422005e+00 8.98871243e-01
5.15010476e-01 -1.46280885e+00 4.07196134e-01 9.54204738e-01
9.13068712e-01 -7.08601236e-01 -2.65039802e-01 -1.95391372e-01
-7.46610343e-01 7.50955224e-01 9.19092298e-01 4.35415775e-01
7.41926134e-01 4.08222944e-01 -1.38378302e-02 5.39316004e-03
-1.24349964e+00 2.82956421e-01 3.30168642e-02 2.04564184e-01
-2.64957994e-01 -4.39302735e-02 -4.40988280e-02 1.06263101e+00
-6.91833049e-02 1.24299429e-01 3.89261097e-01 9.75974798e-01
-5.47983408e-01 -9.92808759e-01 1.82604060e-01 6.04123771e-01
-3.68348733e-02 -7.00309813e-01 -3.83279771e-01 3.82312626e-01
-3.08544010e-01 7.22634435e-01 -3.51380855e-01 -8.42047513e-01
5.11510611e-01 6.81788564e-01 1.11091435e-01 -6.96685135e-01
-6.86336279e-01 -1.60003081e-01 3.01486906e-02 -5.82355678e-01
1.17306925e-01 -2.82740474e-01 -9.64684010e-01 -2.49592766e-01
-6.32301271e-01 4.18154657e-01 5.55808902e-01 7.99947143e-01
9.88656163e-01 5.90344906e-01 4.42868710e-01 -4.82920825e-01
-8.53461146e-01 -1.36628652e+00 -4.71575528e-01 -4.37656671e-01
2.79811114e-01 -7.20110655e-01 -4.38948810e-01 -3.05946767e-01] | [9.968682289123535, 3.614481210708618] |
f1334180-5437-4b75-8fa6-3512a285e3c4 | text-to-image-diffusion-models-can-be-easily | 2305.04175 | null | https://arxiv.org/abs/2305.04175v1 | https://arxiv.org/pdf/2305.04175v1.pdf | Text-to-Image Diffusion Models can be Easily Backdoored through Multimodal Data Poisoning | With the help of conditioning mechanisms, the state-of-the-art diffusion models have achieved tremendous success in guided image generation, particularly in text-to-image synthesis. To gain a better understanding of the training process and potential risks of text-to-image synthesis, we perform a systematic investigation of backdoor attack on text-to-image diffusion models and propose BadT2I, a general multimodal backdoor attack framework that tampers with image synthesis in diverse semantic levels. Specifically, we perform backdoor attacks on three levels of the vision semantics: Pixel-Backdoor, Object-Backdoor and Style-Backdoor. By utilizing a regularization loss, our methods efficiently inject backdoors into a large-scale text-to-image diffusion model while preserving its utility with benign inputs. We conduct empirical experiments on Stable Diffusion, the widely-used text-to-image diffusion model, demonstrating that the large-scale diffusion model can be easily backdoored within a few fine-tuning steps. We conduct additional experiments to explore the impact of different types of textual triggers. Besides, we discuss the backdoor persistence during further training, the findings of which provide insights for the development of backdoor defense methods. | ['Hang Su', 'Yuejian Fang', 'Shi Pu', 'Qingni Shen', 'Yinpeng Dong', 'Shengfang Zhai'] | 2023-05-07 | null | null | null | null | ['data-poisoning', 'backdoor-attack'] | ['adversarial', 'adversarial'] | [ 6.00248992e-01 5.12766279e-03 -3.32950085e-01 9.77860913e-02
-8.78488004e-01 -1.10044348e+00 1.06406665e+00 -3.02220792e-01
-1.01181261e-01 2.48199534e-02 1.28028661e-01 -8.69339406e-01
1.82753801e-01 -7.17615247e-01 -1.07755232e+00 -5.94207227e-01
7.63376132e-02 -1.39951736e-01 3.70353103e-01 -1.49240538e-01
3.79348397e-01 4.09384876e-01 -7.79058695e-01 6.52212620e-01
6.09219134e-01 7.92696834e-01 4.54730652e-02 9.09958780e-01
1.79539993e-01 9.85886633e-01 -8.41797411e-01 -9.87937450e-01
4.96369898e-01 -4.39817756e-01 -4.82554317e-01 2.57138163e-01
7.60923684e-01 -9.16158497e-01 -8.07077944e-01 1.27951467e+00
3.31955522e-01 -3.54448736e-01 5.65173745e-01 -1.29432964e+00
-1.22630954e+00 8.15580308e-01 -7.22267210e-01 8.24474543e-02
1.14947051e-01 7.46425152e-01 8.43044519e-01 -8.29371691e-01
7.92641342e-01 1.60004807e+00 4.01769131e-01 7.37923622e-01
-1.32115555e+00 -9.07142699e-01 9.23682656e-03 -1.61023930e-01
-1.01762331e+00 -4.22221601e-01 6.53881013e-01 -4.73064095e-01
8.28646421e-01 3.56440723e-01 1.67770788e-01 1.92353117e+00
4.08523351e-01 9.26129460e-01 1.44247937e+00 -3.34519267e-01
-3.45380791e-02 4.09532905e-01 -8.49244818e-02 9.50053811e-01
1.87664106e-01 6.97903872e-01 -6.89750612e-01 -3.55874777e-01
7.98132658e-01 -4.28300768e-01 -1.19981505e-01 -1.57500952e-01
-1.00989413e+00 1.10935140e+00 4.13486570e-01 -1.32463574e-01
1.31278083e-01 5.72519422e-01 3.65352869e-01 2.51354694e-01
2.81566024e-01 2.74779618e-01 5.89292943e-02 3.33687812e-01
-9.43396568e-01 3.62618774e-01 5.14633417e-01 8.79519820e-01
4.41891611e-01 1.84719577e-01 -6.65888369e-01 3.91871810e-01
2.85598218e-01 8.40565264e-01 2.53868699e-01 -9.98700678e-01
7.74497867e-01 -1.24378484e-02 -2.42646888e-01 -1.07903111e+00
3.39685529e-01 -2.61748701e-01 -6.10384166e-01 2.71861762e-01
2.43143812e-01 -2.86760151e-01 -1.10041022e+00 1.71134281e+00
1.60564750e-01 1.21812910e-01 7.02145249e-02 7.38488138e-01
2.47878626e-01 6.65138781e-01 -5.25853075e-02 1.92936093e-01
1.46447706e+00 -1.18084824e+00 -5.06541967e-01 -2.66505510e-01
6.48145974e-01 -8.98467422e-01 1.29744983e+00 3.17206889e-01
-9.17248011e-01 -2.81307131e-01 -1.22117484e+00 6.27201051e-03
-2.56639063e-01 4.53648642e-02 6.44737363e-01 1.33887208e+00
-7.78670073e-01 3.50800276e-01 -6.77605867e-01 -1.59712881e-01
6.71323061e-01 3.70494016e-02 -1.75216272e-01 -2.04671472e-01
-1.14911163e+00 4.44244325e-01 4.24075164e-02 -2.88661718e-01
-1.70878112e+00 -6.47898257e-01 -5.36989689e-01 -1.41582370e-01
6.52225256e-01 -8.89597058e-01 1.05666924e+00 -7.06310749e-01
-1.45932209e+00 8.67109001e-01 1.29161268e-01 -7.69514620e-01
8.95614386e-01 -3.06633174e-01 -2.20371842e-01 4.20351446e-01
2.28980958e-01 8.30918610e-01 1.78379631e+00 -1.41906011e+00
-3.19782853e-01 -2.77627230e-01 1.29540995e-01 -1.93026066e-01
-7.56394267e-01 1.58921957e-01 -5.71363866e-01 -1.18489420e+00
-3.84023726e-01 -1.17252016e+00 -1.47177681e-01 3.83576192e-02
-1.15935433e+00 2.50207722e-01 9.52344477e-01 -4.82600093e-01
1.02822649e+00 -2.29044676e+00 -7.35865012e-02 2.40417913e-01
2.28156552e-01 2.60863572e-01 -3.46447587e-01 5.41431606e-01
1.31238937e-01 4.69331622e-01 -9.88188609e-02 -5.48490644e-01
1.53596960e-02 8.94230977e-02 -1.13310349e+00 2.65345305e-01
7.04140142e-02 1.08907807e+00 -5.12199938e-01 -2.66519278e-01
-6.97767064e-02 3.58616024e-01 -8.33791852e-01 1.95414200e-02
-4.64130670e-01 1.59254894e-01 -4.60003525e-01 6.11666024e-01
6.39156699e-01 -2.56426871e-01 -8.15624744e-02 -1.46522671e-01
3.23238224e-01 -1.81105331e-01 -3.69543880e-01 1.37655437e+00
-1.73863575e-01 7.52526402e-01 5.17260395e-02 -3.00586969e-01
3.41251045e-01 -1.05729867e-02 -1.00313716e-01 -5.05998731e-01
1.08590007e-01 2.32577354e-01 -2.04304442e-01 -2.04995096e-01
8.32304299e-01 -1.99235789e-02 -2.77644657e-02 5.57566941e-01
-1.05708987e-01 -2.29239821e-01 -7.81560019e-02 8.19001377e-01
9.97275651e-01 -1.05662838e-01 -4.98614073e-01 -4.09960002e-02
1.85574546e-01 -3.47202271e-02 -4.01467085e-02 1.41807222e+00
-4.97907773e-02 5.86314023e-01 6.67909920e-01 1.51843339e-01
-1.02127075e+00 -1.17474854e+00 2.02890635e-01 9.59978938e-01
2.53487289e-01 -6.06139302e-01 -1.34736860e+00 -1.03649306e+00
1.51717708e-01 8.00072372e-01 -6.34479403e-01 -4.21364129e-01
-2.08248422e-01 -5.65274239e-01 1.48274052e+00 2.58261174e-01
6.11427248e-01 -6.32358193e-01 -2.46039346e-01 -2.40462065e-01
9.51339304e-02 -1.44649279e+00 -1.02079558e+00 -1.66553468e-01
-8.44202280e-01 -7.96112597e-01 -9.21106637e-01 -3.11831355e-01
7.20921218e-01 5.53527236e-01 5.51340997e-01 -1.90955447e-03
-3.67854923e-01 6.84328139e-01 -1.31014764e-01 -1.69635072e-01
-1.05936003e+00 -1.13525659e-01 -4.27561738e-02 2.18106315e-01
-3.00302953e-01 -1.73212633e-01 -5.99276185e-01 5.36869705e-01
-1.34623432e+00 -1.42411307e-01 4.50218201e-01 8.00093114e-01
1.86487138e-01 3.52548748e-01 -8.68660584e-02 -9.05783355e-01
9.82703209e-01 -2.28120252e-01 -8.17158759e-01 2.84159273e-01
-6.35088682e-01 5.17216250e-02 4.47971225e-01 -8.21935415e-01
-1.03007507e+00 -3.50801408e-01 -1.62257835e-01 -1.02233708e+00
3.33184332e-01 2.01803416e-01 -9.48850811e-02 -4.94774640e-01
9.20269907e-01 4.74680871e-01 -1.23598173e-01 -1.22386247e-01
1.10787094e+00 3.20529610e-01 6.17863297e-01 -9.18023705e-01
1.22230792e+00 9.32398796e-01 -1.65653691e-01 -9.57510412e-01
-6.44721985e-01 1.77969903e-01 2.56738901e-01 1.72008444e-02
8.74241352e-01 -8.78149748e-01 -5.35229325e-01 9.86037195e-01
-1.15054798e+00 -5.77947199e-01 -1.40427396e-01 2.78912857e-03
-4.81007218e-01 8.42982411e-01 -1.05684829e+00 -6.80898190e-01
-1.98456988e-01 -1.51865506e+00 1.15258002e+00 -1.33343667e-01
-5.03291115e-02 -9.86372232e-01 -3.31163794e-01 7.46179521e-01
2.72440583e-01 -2.10668027e-01 1.17477345e+00 -4.04663712e-01
-1.16225255e+00 -7.14699849e-02 -2.13174835e-01 6.38672233e-01
-3.18721801e-01 3.00764218e-02 -9.58916187e-01 -5.72616100e-01
3.11738104e-01 -5.99180400e-01 1.28322327e+00 1.92299247e-01
1.12857378e+00 -5.42803824e-01 -3.91593784e-01 8.51494431e-01
1.11066091e+00 -1.16190195e-01 7.70063937e-01 3.26024085e-01
8.00447166e-01 4.98709887e-01 2.79040337e-01 2.86926031e-01
1.76342018e-02 4.00823265e-01 4.49591666e-01 2.64778850e-03
-3.76945585e-01 -8.80073488e-01 7.43694782e-01 2.21496224e-01
5.85541487e-01 -6.95292771e-01 -4.55794662e-01 3.79896909e-01
-1.34773755e+00 -8.52346897e-01 1.50863826e-01 1.89306295e+00
6.76666021e-01 5.74000895e-01 1.23182433e-02 -8.10152367e-02
6.13435209e-01 5.48518538e-01 -4.71405029e-01 -2.89980531e-01
-3.17306608e-01 -1.21958241e-01 1.02678919e+00 5.58897018e-01
-1.02604485e+00 1.47476101e+00 7.03167200e+00 1.36858785e+00
-1.23229110e+00 2.13934392e-01 9.52714980e-01 -4.71297540e-02
-7.62017727e-01 2.12354392e-01 -9.77854908e-01 3.97424400e-01
4.70297158e-01 8.41391981e-02 6.74958289e-01 7.82269478e-01
-6.29489198e-02 1.38947800e-01 -9.84004259e-01 7.07227230e-01
2.52159103e-03 -1.67236102e+00 6.03891373e-01 5.72205305e-01
9.21932340e-01 -6.64118538e-03 9.66612041e-01 -7.66137540e-02
5.34632266e-01 -7.66931951e-01 1.14362860e+00 -2.54756678e-02
8.99102330e-01 -5.00346780e-01 -2.31408909e-01 2.64702767e-01
-5.56073487e-01 -3.79860669e-01 -2.09577248e-01 5.77415168e-01
9.44389999e-02 5.60467362e-01 -6.17279887e-01 3.28679442e-01
4.31990683e-01 1.59798354e-01 -4.90157574e-01 1.63777247e-01
-3.50774735e-01 8.88912439e-01 -3.00011128e-01 3.34641010e-01
5.20057857e-01 2.81817038e-02 9.10491109e-01 1.11189973e+00
4.12968516e-01 -2.56813884e-01 -4.72478159e-02 1.40056527e+00
-2.94830143e-01 -3.22689325e-01 -9.16806877e-01 -6.42119765e-01
3.50679040e-01 8.80533159e-01 -7.48276174e-01 -2.33407363e-01
-1.13109313e-01 1.58862841e+00 -4.49322872e-02 6.39272928e-01
-1.04730666e+00 -6.82182014e-02 7.57075608e-01 1.83732569e-01
3.90240431e-01 -4.19131160e-01 -2.60770351e-01 -1.17325330e+00
-1.05217464e-01 -1.29427660e+00 8.34780559e-02 -7.74531901e-01
-1.43048334e+00 4.14397389e-01 -1.60357095e-02 -5.13412893e-01
1.70699946e-04 -6.74681425e-01 -5.69416523e-01 6.37371004e-01
-1.24285567e+00 -1.40766680e+00 1.85940936e-01 8.91913712e-01
4.72415030e-01 -5.65367997e-01 3.45317632e-01 8.05438031e-04
-5.26102066e-01 1.14301777e+00 1.35141969e-01 2.83136755e-01
6.17827654e-01 -6.83257639e-01 1.00805593e+00 1.25753653e+00
4.28193510e-01 9.81205106e-01 5.40757954e-01 -1.04995906e+00
-1.71160054e+00 -1.08287358e+00 1.55482471e-01 -6.45577669e-01
9.52409446e-01 -5.58969378e-01 -4.62336600e-01 6.63800597e-01
3.45153749e-01 -1.98160499e-01 2.68994778e-01 -4.87574905e-01
-8.72479737e-01 1.75688341e-01 -9.55335319e-01 1.09290397e+00
9.72227514e-01 -1.19426918e+00 -1.53157283e-02 3.79171550e-01
1.14811790e+00 -5.11012733e-01 -3.62414300e-01 -5.60321659e-02
4.87357497e-01 -1.15403759e+00 1.24798059e+00 -4.10750240e-01
7.82789230e-01 1.22084036e-01 -4.24159169e-01 -9.09357190e-01
-6.53860494e-02 -1.41049147e+00 -2.09045470e-01 1.23293626e+00
3.25588763e-01 -6.81915879e-01 8.89325261e-01 1.98520884e-01
2.69353777e-01 -3.98276836e-01 -6.93904102e-01 -1.06134951e+00
3.01611125e-01 -7.26900160e-01 3.34453046e-01 7.09371328e-01
-4.70484674e-01 1.18398279e-01 -9.18604314e-01 3.68328929e-01
1.05523562e+00 -3.60065550e-01 8.46700668e-01 -1.94384724e-01
-7.31510699e-01 -4.43703949e-01 1.31554514e-01 -1.51754415e+00
9.29353684e-02 -9.46253240e-01 -3.41239929e-01 -6.89371347e-01
1.07015193e-01 -2.49737471e-01 3.64081562e-02 1.21960215e-01
-8.20162147e-02 4.26484317e-01 6.29477024e-01 4.07716542e-01
4.97753993e-02 4.32364911e-01 1.52130318e+00 -5.68181515e-01
1.73105940e-01 -2.08726376e-01 -8.97540212e-01 4.82368976e-01
4.20701772e-01 -7.83966780e-01 -9.14589465e-01 -5.95348299e-01
2.08572268e-01 4.27247100e-02 6.43474817e-01 -5.94396174e-01
1.49172157e-01 -1.11881390e-01 -9.09452662e-02 -2.76725233e-01
3.58751923e-01 -5.94782412e-01 -2.22226620e-01 5.46977818e-01
-6.48991883e-01 -2.46404037e-01 1.72565326e-01 9.56004202e-01
1.99858218e-01 -1.40179053e-01 7.46779203e-01 -8.21783915e-02
-3.72102350e-01 3.52881342e-01 -4.95215803e-01 1.24371191e-02
1.05042851e+00 -2.53067702e-01 -6.37656510e-01 -4.79109854e-01
-4.78314131e-01 -9.91930440e-02 6.26718760e-01 6.31153405e-01
8.33840668e-01 -1.07334101e+00 -4.50816035e-01 5.72047234e-01
-5.45771839e-03 -6.53444707e-01 1.86868921e-01 5.33942223e-01
-3.98554265e-01 3.41176659e-01 1.30472034e-01 -6.45761490e-01
-1.28757834e+00 9.08187747e-01 3.25729191e-01 -2.54381984e-01
-6.40453100e-01 1.07803786e+00 8.46577644e-01 5.61088696e-02
2.52822608e-01 -7.41342604e-02 6.30385101e-01 -3.05187762e-01
4.40320879e-01 2.20180258e-01 -3.19451481e-01 -2.90580750e-01
-6.02782406e-02 4.06022787e-01 -4.87707168e-01 -7.47390389e-01
5.04889369e-01 -1.37448713e-01 -7.05762953e-02 -2.43948951e-01
1.14509511e+00 3.55482012e-01 -1.40370834e+00 -1.27920538e-01
-4.18830186e-01 -8.25080991e-01 1.04497522e-01 -8.15973520e-01
-9.97690320e-01 1.08064032e+00 5.76519370e-01 2.16225967e-01
8.04086208e-01 -1.23182386e-01 1.26643205e+00 3.15315485e-01
2.04813987e-01 -9.09632862e-01 7.36758947e-01 2.65779942e-01
8.68894696e-01 -8.68854702e-01 -1.56477958e-01 -4.89932537e-01
-8.48773122e-01 6.76488757e-01 2.62840122e-01 -4.27869707e-02
5.27330995e-01 4.47476089e-01 1.85136631e-01 -1.41878739e-01
-6.04678571e-01 3.14632148e-01 1.39735535e-01 6.87384248e-01
-2.93256283e-01 -8.52226913e-02 3.74703616e-01 2.50784963e-01
-4.06493433e-02 -1.70615152e-01 5.39707482e-01 7.23090708e-01
-7.03327879e-02 -1.25064933e+00 -6.09439611e-01 6.86679687e-03
-7.60995924e-01 -3.66758198e-01 -7.44783342e-01 5.32707274e-01
-1.92944750e-01 1.16739261e+00 -3.81291151e-01 -6.44565105e-01
-5.21291494e-02 -2.23280981e-01 6.84043288e-01 -2.74969518e-01
-6.96351826e-01 2.84143537e-01 2.04238072e-02 -6.74793661e-01
1.57062426e-01 -2.30375215e-01 -8.56562138e-01 -6.68252349e-01
-3.72014910e-01 -4.15087372e-01 6.61242723e-01 7.15510190e-01
4.14670438e-01 2.33793527e-01 7.49863267e-01 -5.83161831e-01
-1.18917024e+00 -3.43417227e-01 -6.07834280e-01 5.44450998e-01
4.03977752e-01 -2.51449019e-01 -3.65462035e-01 5.14166243e-02] | [5.739577293395996, 7.845658779144287] |
9470da38-0ef6-43de-ac17-34ef18757070 | fusing-event-based-and-rgb-camera-for-robust | null | null | https://hal.archives-ouvertes.fr/hal-03591717/ | https://hal.archives-ouvertes.fr/hal-03591717/document | Fusing Event-based and RGB camera for Robust Object Detection in Adverse Conditions | The ability to detect objects, under image corruptions and different weather conditions is vital for deep learning models especially when applied to real-world applications such as autonomous driving. Traditional RGB-based detection fails under these conditions and it is thus important to design a sensor suite that is redundant to failures of the primary frame-based detection. Event-based cameras can complement frame-based cameras in low-light conditions and high dynamic range scenarios that an autonomous vehicle can encounter during navigation. Accordingly, we propose a redundant sensor fusion model of event-based and frame-based cameras that is robust to common image corruptions. The method utilizes a voxel grid representation for events as input and proposes a two-parallel feature extractor network for frames and events. Our sensor fusion approach is more robust to corruptions by over 30% compared to only frame-based detections and outperforms the only event-based detection. The model is trained and evaluated on the publicly released DSEC dataset. | ['Christian Laugier', 'Alessandro Renzaglia', 'Khushdeep Singh Mann', 'Anshul Paigwar', 'Abhishek Tomy'] | 2022-03-30 | null | null | null | icra-2022-3 | ['robust-object-detection', 'infrared-and-visible-image-fusion', 'stereo-lidar-fusion', 'event-based-vision'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 2.83009857e-01 -4.20490801e-01 4.14180458e-01 -4.72455382e-01
-6.21698976e-01 -4.40306753e-01 8.95837188e-01 8.80410969e-02
-9.19188201e-01 6.17198884e-01 -4.15745527e-01 -1.03535250e-01
1.78338364e-01 -8.19085658e-01 -1.06896377e+00 -8.13044190e-01
-1.57900639e-02 7.75368838e-03 1.05143356e+00 -1.66274175e-01
3.32041867e-02 9.45942938e-01 -2.28796840e+00 1.83408871e-01
1.78180397e-01 1.27458370e+00 4.37240124e-01 8.94409478e-01
2.21472740e-01 9.61024761e-01 -6.82947993e-01 1.95109487e-01
3.62622112e-01 -1.30861560e-02 4.02755626e-02 1.32394016e-01
8.37962568e-01 -6.39716506e-01 -5.99955440e-01 9.44304883e-01
2.30645239e-01 3.10134366e-02 2.90807217e-01 -1.45638835e+00
3.36388081e-01 -3.59428138e-01 -4.36432451e-01 8.07565629e-01
5.29795110e-01 4.07105595e-01 2.82638282e-01 -8.63196015e-01
4.79860306e-01 1.12571633e+00 7.29823530e-01 3.02545607e-01
-9.56342340e-01 -6.20213389e-01 -1.00057572e-01 7.56993115e-01
-1.04954600e+00 -6.28701150e-01 4.85971719e-01 -2.78517544e-01
1.43578243e+00 2.64712498e-02 7.27726102e-01 1.29313457e+00
6.82263553e-01 3.81286830e-01 1.10667169e+00 -1.69510081e-01
4.61075068e-01 -2.13314652e-01 -5.23466878e-02 5.41878521e-01
7.00195134e-01 6.19773686e-01 -9.98613298e-01 2.26616815e-01
5.53578615e-01 1.48454577e-01 -1.58164144e-01 -2.49056995e-01
-1.37745869e+00 6.24488533e-01 3.65732700e-01 -1.75429136e-01
-4.50280309e-01 6.02301776e-01 2.46734515e-01 2.93666512e-01
2.12066650e-01 -1.02545232e-01 -5.21930516e-01 -1.45845160e-01
-1.01704776e+00 3.15333307e-01 5.39054573e-01 9.22483146e-01
7.96852410e-01 3.61935705e-01 2.53752202e-01 1.34709477e-01
6.01932168e-01 9.37061071e-01 1.45830005e-01 -1.01232362e+00
1.71498820e-01 4.93688315e-01 7.05706775e-02 -6.75619185e-01
-7.95628309e-01 -1.33145332e-01 -6.40569270e-01 8.42571437e-01
2.65257835e-01 3.39158066e-02 -1.04591215e+00 1.24715018e+00
5.11510849e-01 6.34402514e-01 3.84805381e-01 8.64020467e-01
1.10412896e+00 5.55927277e-01 -1.82228759e-01 -2.36375004e-01
1.41928530e+00 -3.12118590e-01 -6.70677960e-01 -4.85545665e-01
2.96011388e-01 -7.74602652e-01 3.64484131e-01 6.51714206e-01
-6.61312521e-01 -5.65474808e-01 -1.54071140e+00 1.08433403e-01
-5.23322761e-01 -1.58357501e-01 4.77488518e-01 7.59782851e-01
-1.10246444e+00 2.15026200e-01 -1.33728182e+00 -5.83359778e-01
2.02317670e-01 4.13146853e-01 -6.95935190e-01 -1.84709340e-01
-9.82603252e-01 1.24096334e+00 4.76218492e-01 1.94164023e-01
-1.16717875e+00 -4.64111179e-01 -1.12367702e+00 -3.70720744e-01
1.90826416e-01 -3.92781556e-01 1.18890715e+00 -6.65368438e-01
-1.14968061e+00 6.88761950e-01 -3.87423038e-01 -8.96723688e-01
5.85348248e-01 -1.61227986e-01 -4.44347084e-01 4.14661378e-01
2.71860883e-02 5.85866094e-01 9.85687256e-01 -1.21335745e+00
-1.19630766e+00 -1.85210675e-01 3.96016091e-02 5.77911027e-02
1.60734281e-01 1.65916197e-02 -2.96405643e-01 1.85400937e-02
3.72680068e-01 -8.41392875e-01 -1.10332817e-01 1.75515190e-01
1.81448851e-02 9.24328342e-02 1.46834564e+00 -9.61524919e-02
2.79275835e-01 -2.01099420e+00 -3.13378006e-01 -1.28454819e-01
7.63132349e-02 1.53872102e-01 9.03419107e-02 1.97979137e-01
-1.51741207e-02 -6.02239668e-01 6.44506365e-02 -5.93221188e-01
-3.82550150e-01 6.40544891e-01 -2.63107926e-01 9.95894253e-01
3.65213394e-01 2.88203567e-01 -8.26098979e-01 -4.14043456e-01
1.12877822e+00 6.77685678e-01 -6.52769431e-02 1.67993218e-01
-2.99474783e-02 5.19195378e-01 -3.83581733e-03 9.10556316e-01
7.66144753e-01 3.05613190e-01 -3.60792130e-01 -4.18450624e-01
-5.04277766e-01 6.29729405e-02 -1.42445433e+00 1.34368348e+00
-3.04067135e-01 1.22448659e+00 6.28935099e-02 -5.00804722e-01
6.90438569e-01 3.22491288e-01 4.37549412e-01 -1.08095181e+00
1.24175474e-01 2.12191924e-01 -2.56937683e-01 -5.02698958e-01
7.19830275e-01 -7.02101290e-02 -5.71350008e-03 -1.64251730e-01
2.05171973e-01 -1.52665198e-01 2.66448647e-01 3.06361645e-01
1.66873896e+00 2.86050141e-01 1.43122658e-01 1.48645431e-01
5.25810063e-01 1.98381022e-01 6.60229921e-01 8.07516277e-01
-4.54963207e-01 7.15498984e-01 7.00379387e-02 -6.99338853e-01
-7.94842422e-01 -1.00905561e+00 -3.50524276e-01 6.83744907e-01
5.65654933e-01 -2.25103423e-01 -2.32780233e-01 -2.93792218e-01
8.49363394e-03 5.97093940e-01 -4.91939813e-01 -2.78814256e-01
-4.73731875e-01 -9.12079096e-01 4.80007976e-01 4.70554024e-01
6.44155145e-01 -7.56893396e-01 -1.57729828e+00 4.30275232e-01
-4.92027961e-02 -1.70515084e+00 4.77129996e-01 7.28362560e-01
-5.73315024e-01 -1.38994622e+00 1.98535360e-02 -1.27098486e-01
2.30492711e-01 7.11979389e-01 1.13936377e+00 -3.70102003e-02
-4.75031316e-01 8.02737832e-01 -5.74357092e-01 -8.06526482e-01
-2.12624520e-01 -7.02999592e-01 2.47692525e-01 1.47227794e-02
4.77325827e-01 -2.54734725e-01 -6.57036602e-01 2.69417793e-01
-1.16752446e+00 -1.81245789e-01 3.18210304e-01 4.78627741e-01
4.62621987e-01 3.34201790e-02 6.61889017e-02 -5.69702350e-02
-2.45879501e-01 -4.30475116e-01 -1.22889006e+00 -2.68367738e-01
-1.78444386e-01 -2.94425517e-01 3.12269390e-01 -1.31787390e-01
-1.01411712e+00 7.25009263e-01 9.35402513e-02 -5.15545487e-01
-5.29221535e-01 8.94061774e-02 -3.26206982e-02 -5.14981747e-01
7.16034114e-01 1.06749617e-01 -2.58606791e-01 4.61262390e-02
-2.45891884e-02 3.20064873e-01 7.70741343e-01 7.21221929e-03
8.44269276e-01 1.14367712e+00 4.67844427e-01 -1.26637101e+00
-3.59611779e-01 -8.53345633e-01 -6.34226680e-01 -7.82064557e-01
8.58415544e-01 -1.35938752e+00 -5.16361713e-01 8.56068611e-01
-1.53502452e+00 -2.90331021e-02 -4.90769595e-02 7.01805472e-01
-5.33739507e-01 2.28227600e-01 -1.22641310e-01 -1.03388703e+00
2.63422608e-01 -1.27313948e+00 1.41933775e+00 5.44511437e-01
3.65008801e-01 -6.80870891e-01 3.17781046e-02 5.99929616e-02
1.71326026e-01 6.66521907e-01 -2.93902904e-01 -1.23879589e-01
-1.14724326e+00 -3.29937249e-01 -2.58455932e-01 1.44601062e-01
-5.36023714e-02 1.82411984e-01 -1.34796965e+00 -1.72196597e-01
1.69270709e-01 -2.39116415e-01 1.29388106e+00 4.22881842e-01
5.57648838e-01 4.68062490e-01 -3.70263129e-01 9.51437056e-01
1.70505536e+00 3.07367861e-01 8.54504228e-01 8.86971116e-01
6.28072619e-01 2.81695873e-01 6.69213891e-01 5.01158953e-01
4.48974460e-01 5.39947033e-01 1.15603411e+00 -1.24064401e-01
-2.30927780e-01 5.10968804e-01 6.69833958e-01 -1.16079338e-01
-4.86912206e-02 -3.37558210e-01 -8.47897708e-01 5.23005843e-01
-1.87717688e+00 -1.07886791e+00 -6.62644565e-01 2.14519811e+00
2.59285122e-01 4.95289534e-01 -2.59118676e-01 3.43397915e-01
3.38239193e-01 1.76280305e-01 -5.09948552e-01 -6.00359924e-02
-4.66345280e-01 -1.06056244e-03 1.30395103e+00 8.15319717e-02
-1.38740218e+00 6.82084501e-01 6.40443182e+00 1.60188694e-02
-1.14169025e+00 2.80945927e-01 3.76874618e-02 -3.18311900e-01
3.15252244e-01 -1.28408998e-01 -1.23046839e+00 4.54038620e-01
1.22384477e+00 3.80242735e-01 -1.60061494e-01 5.50377071e-01
6.22948229e-01 -9.61874366e-01 -1.06439948e+00 1.18365157e+00
2.38651529e-01 -1.29025233e+00 -5.81107855e-01 -5.64474519e-03
3.41947019e-01 7.30245471e-01 -3.35100323e-01 -8.54110494e-02
2.94557333e-01 -7.32533455e-01 1.07952154e+00 4.92647797e-01
4.70932633e-01 -5.04359722e-01 8.31365347e-01 2.19614223e-01
-1.43969500e+00 7.79500008e-02 -2.31863037e-01 -2.34812602e-01
4.78051364e-01 7.05507398e-01 -8.46256733e-01 4.06798214e-01
1.19217110e+00 8.26817751e-01 -7.32298791e-01 1.28784275e+00
-5.26685454e-02 5.33848286e-01 -8.82041037e-01 2.92131275e-01
2.12009445e-01 2.72795677e-01 7.15269864e-01 1.19302702e+00
5.22664547e-01 -1.24470487e-01 3.91644947e-02 2.49957502e-01
4.04924154e-01 -9.05829370e-01 -8.74987841e-01 6.17573977e-01
4.31721032e-01 1.30286705e+00 -9.95241284e-01 -3.10945153e-01
-8.11932266e-01 9.77317750e-01 -9.07146484e-02 3.42409045e-01
-1.02979505e+00 -4.89653833e-02 7.34609008e-01 3.84933501e-03
6.06394768e-01 -7.33175933e-01 -7.07670599e-02 -1.03910267e+00
7.47155175e-02 -5.11289775e-01 2.31894940e-01 -1.06252003e+00
-7.52458036e-01 4.90948230e-01 2.22647786e-01 -1.41999650e+00
-1.49429217e-01 -9.66062546e-01 -4.24967945e-01 5.25750041e-01
-2.04622650e+00 -1.24441421e+00 -9.10276592e-01 7.78073192e-01
4.68449920e-01 -6.52643517e-02 4.96945977e-01 3.52916688e-01
-4.53431636e-01 -6.41963258e-02 6.64001107e-02 -1.14586569e-01
6.88845754e-01 -1.18486774e+00 1.50997639e-01 1.53017914e+00
1.51763737e-01 -6.27666339e-02 1.20611894e+00 -6.13803566e-01
-1.80963922e+00 -1.38054109e+00 5.42830884e-01 -5.54071605e-01
4.68803763e-01 -1.75287962e-01 -7.09745288e-01 6.74083710e-01
1.28172919e-01 6.01412177e-01 2.15613216e-01 -5.39800286e-01
-2.65097380e-01 -5.48035443e-01 -1.15064263e+00 4.15455699e-02
7.13947773e-01 -3.18671733e-01 -5.91914296e-01 5.55638492e-01
4.23762619e-01 -8.01717758e-01 -3.64207149e-01 5.55167735e-01
4.63491499e-01 -1.43963516e+00 1.09205806e+00 1.60976827e-01
-1.32997498e-01 -8.91451836e-01 -4.75393832e-01 -8.54631782e-01
7.94670880e-02 -2.21495986e-01 -2.43356973e-01 7.32067704e-01
-4.61894237e-02 -7.55244017e-01 6.85687780e-01 2.12714911e-01
-3.73139918e-01 9.34958160e-02 -1.57702816e+00 -9.21073496e-01
-8.08868229e-01 -9.58440125e-01 1.77823424e-01 3.53779107e-01
-8.55914176e-01 -1.98711291e-01 -2.93047756e-01 7.85432756e-01
9.40840065e-01 -4.10041451e-01 7.66005933e-01 -1.24204218e+00
7.93388709e-02 1.59956459e-02 -1.09184396e+00 -2.62483329e-01
-2.68503726e-01 -5.36857359e-02 4.41766053e-01 -1.51477396e+00
-2.57309645e-01 -1.12221077e-01 -3.24620783e-01 3.80575240e-01
1.30387396e-01 7.67957628e-01 6.81022853e-02 2.47654095e-01
-8.01996231e-01 2.76127458e-01 3.03902060e-01 9.96463299e-02
2.00194389e-01 -2.69024223e-01 1.93096712e-01 8.87700856e-01
5.16684473e-01 -5.90921342e-01 -2.01224089e-01 -4.10130799e-01
5.56088269e-01 -1.04647450e-01 1.06459379e+00 -1.77243471e+00
6.59695625e-01 3.04770470e-03 6.86559975e-01 -1.11711955e+00
7.19208717e-01 -1.17851317e+00 3.32144022e-01 6.33177459e-01
3.47681642e-01 2.95329362e-01 4.30835634e-01 8.71841490e-01
-2.20573887e-01 -9.46712419e-02 8.81422758e-01 -3.21011543e-01
-1.47496331e+00 -2.03726999e-03 -8.36171925e-01 -6.12674475e-01
1.39684856e+00 -6.34354413e-01 -4.79768515e-01 -1.64516091e-01
-1.34419173e-01 9.33848042e-03 4.25264418e-01 4.68472451e-01
8.84157300e-01 -1.04962754e+00 -5.46088755e-01 4.44515288e-01
3.83630574e-01 2.01519087e-01 6.27094507e-02 8.83763313e-01
-9.19129074e-01 4.19489563e-01 -2.33764529e-01 -1.20583284e+00
-1.57439709e+00 1.26530766e-01 4.17522967e-01 3.38809013e-01
-6.09242976e-01 9.34246004e-01 -1.21753082e-01 1.45035893e-01
3.52843732e-01 -6.34790182e-01 -1.38798475e-01 1.19066440e-01
7.36986518e-01 3.63939762e-01 7.01184690e-01 -1.00868261e+00
-8.05265129e-01 4.00036305e-01 2.41802618e-01 -2.75869578e-01
1.12486672e+00 -2.89433718e-01 2.37097263e-01 5.72105229e-01
7.73892105e-01 -5.68438411e-01 -1.63973260e+00 -8.79617184e-02
-2.31772602e-01 -4.14966434e-01 6.12526953e-01 -4.43198949e-01
-9.55302358e-01 7.22652316e-01 1.12873638e+00 1.67201146e-01
1.21652353e+00 -1.43152162e-01 5.39618492e-01 4.77754027e-01
5.38808048e-01 -9.88891602e-01 -4.83605117e-02 7.23972082e-01
4.86167520e-01 -1.56238949e+00 2.97672898e-01 -9.35265943e-02
3.06702405e-03 1.27285469e+00 6.63256943e-01 -1.46217152e-01
5.39767504e-01 6.91489577e-01 1.33102134e-01 -2.50185281e-01
-9.00722861e-01 -6.38757050e-01 -4.57166769e-02 8.15444231e-01
-7.40037635e-02 -2.60356337e-01 1.92749202e-01 -2.41882741e-01
1.30345970e-01 -1.85271762e-02 7.29211211e-01 1.39144313e+00
-6.97290421e-01 -7.00062633e-01 -1.03652728e+00 2.98008233e-01
-1.44521251e-01 1.40468225e-01 -6.02630563e-02 8.81377995e-01
7.03131139e-01 1.30597985e+00 4.55238312e-01 -3.83068949e-01
4.02507097e-01 -2.15745673e-01 5.23685575e-01 -3.43255311e-01
-3.47627163e-01 -3.50015126e-02 1.32527769e-01 -1.18693376e+00
-1.09559131e+00 -1.05019546e+00 -1.23596644e+00 -1.01936065e-01
-3.49055558e-01 -6.72979951e-01 1.11350274e+00 1.11578178e+00
1.78524643e-01 7.26016998e-01 3.11117947e-01 -1.38348675e+00
2.27563292e-01 -5.82166433e-01 -3.37484658e-01 1.00595661e-01
1.00525224e+00 -1.12664390e+00 -6.16095662e-01 2.89213896e-01] | [8.309062004089355, -1.2304203510284424] |
8e30496a-f2ed-4153-aa9a-110bccabe372 | pytouch-a-machine-learning-library-for-touch | 2105.12791 | null | https://arxiv.org/abs/2105.12791v1 | https://arxiv.org/pdf/2105.12791v1.pdf | PyTouch: A Machine Learning Library for Touch Processing | With the increased availability of rich tactile sensors, there is an equally proportional need for open-source and integrated software capable of efficiently and effectively processing raw touch measurements into high-level signals that can be used for control and decision-making. In this paper, we present PyTouch -- the first machine learning library dedicated to the processing of touch sensing signals. PyTouch, is designed to be modular, easy-to-use and provides state-of-the-art touch processing capabilities as a service with the goal of unifying the tactile sensing community by providing a library for building scalable, proven, and performance-validated modules over which applications and research can be built upon. We evaluate PyTouch on real-world data from several tactile sensors on touch processing tasks such as touch detection, slip and object pose estimations. PyTouch is open-sourced at https://github.com/facebookresearch/pytouch . | ['Roberto Calandra', 'Trevor Darrell', 'Shaoxiong Wang', 'Po-Wei Chou', 'Jingwei Xu', 'Huazhe Xu', 'Mike Lambeta'] | 2021-05-26 | null | null | null | null | ['touch-detection'] | ['robots'] | [ 2.86726534e-01 -4.95484263e-01 -2.30936497e-01 -2.67811716e-01
-6.98959768e-01 -5.55915058e-01 3.71465422e-02 2.15959370e-01
-2.65029967e-01 1.94786012e-01 -1.80880681e-01 -4.38738614e-02
-6.27900288e-02 -7.94337809e-01 -5.31049371e-01 -3.85164440e-01
2.29273960e-02 2.88160950e-01 4.85248834e-01 -5.44122458e-02
4.97941315e-01 4.04391408e-01 -2.01933193e+00 4.31957066e-01
4.11093295e-01 1.32214594e+00 3.53217423e-01 1.07154906e+00
-7.55983754e-04 -9.24198627e-02 -6.99675130e-03 -6.11339062e-02
2.66236633e-01 4.91979532e-02 -5.48625588e-01 -7.10475266e-01
4.29706603e-01 -4.55085516e-01 2.16487989e-01 4.08654511e-01
8.93183053e-01 -3.90315175e-01 4.18273181e-01 -9.38350737e-01
-1.79916441e-01 4.35155153e-01 -2.85129219e-01 -3.08869034e-01
6.87678277e-01 1.51056513e-01 8.40986073e-01 -8.35377395e-01
4.87941593e-01 1.05781662e+00 9.53535438e-01 6.22300506e-01
-1.15674615e+00 -5.22945523e-01 -3.78328443e-01 -2.74926007e-01
-7.51849294e-01 -3.16590995e-01 7.15716660e-01 -4.06217903e-01
1.10991561e+00 7.49548674e-01 5.60044110e-01 1.64823234e+00
4.41307724e-01 8.95641327e-01 1.39866579e+00 -4.35846955e-01
3.71143758e-01 -2.46055424e-01 2.47862160e-01 4.05257463e-01
3.74127895e-01 5.09608090e-02 -7.86070466e-01 -1.88920990e-01
1.19693947e+00 1.32580949e-02 2.93460578e-01 -4.53322351e-01
-1.23133790e+00 1.36642903e-01 3.76359671e-01 1.68278933e-01
-5.43706477e-01 4.52830315e-01 5.12573898e-01 2.01658964e-01
2.44830713e-01 4.64234889e-01 -7.47626066e-01 -9.54951346e-01
-3.51174980e-01 3.06028992e-01 1.15596831e+00 5.50926089e-01
4.41674620e-01 -4.88799691e-01 5.71581088e-02 9.73370671e-01
3.57805759e-01 7.36171365e-01 1.11694016e-01 -9.89489198e-01
4.30309713e-01 4.16514516e-01 3.12426299e-01 -7.93808699e-01
-4.47112888e-01 5.07711709e-01 -5.06847978e-01 4.89164382e-01
4.86952245e-01 -4.09034133e-01 -7.08545864e-01 1.08097482e+00
3.10537368e-01 -2.28593096e-01 -4.40788805e-01 1.06048071e+00
6.42743886e-01 8.50443095e-02 1.80872992e-01 5.23542583e-01
1.08659589e+00 -4.39272910e-01 -5.02052367e-01 -2.97201276e-01
7.49754310e-02 -1.11297917e+00 1.45815384e+00 7.62460589e-01
-8.73280466e-01 -6.66175723e-01 -9.66065884e-01 -1.56581819e-01
-6.86399043e-01 9.74607319e-02 1.13266754e+00 6.72545612e-01
-6.31153703e-01 1.07132709e+00 -1.53120279e+00 -7.28116989e-01
3.77133459e-01 2.64672846e-01 -3.95341963e-01 1.40419304e-01
-6.72450364e-01 9.09854650e-01 1.23173505e-01 1.08912401e-01
-1.33921817e-01 -7.07205296e-01 -5.37853241e-01 -3.84564281e-01
2.60233551e-01 -3.22612971e-01 1.43956888e+00 -3.19760591e-01
-1.98916626e+00 6.84606850e-01 8.43725577e-02 3.58850569e-01
4.67082053e-01 -8.61428857e-01 -1.21090841e-02 -4.93871942e-02
-4.38222066e-02 3.34075391e-01 6.50760353e-01 -1.16582561e+00
-3.88548017e-01 -5.94504714e-01 -5.74530244e-01 4.58926521e-03
-3.32495689e-01 -1.23765886e-01 -4.41526681e-01 -2.12628409e-01
2.36302689e-02 -7.79792964e-01 -5.09134680e-03 5.02213836e-01
-6.16567314e-01 -7.96570778e-02 9.93305981e-01 -5.84195793e-01
1.12996113e+00 -1.99924302e+00 -6.69944957e-02 4.01035279e-01
-7.40742385e-02 5.18803418e-01 -3.03571206e-02 1.05714130e+00
3.23208719e-01 -7.43123963e-02 4.94491868e-02 -2.55828649e-01
3.72069478e-01 2.09791020e-01 -1.13030516e-01 3.26545611e-02
2.64981031e-01 9.02305782e-01 -8.90103698e-01 -2.44415730e-01
1.01033652e+00 3.86499137e-01 -5.46948351e-02 2.93217987e-01
-1.92420483e-01 3.13907832e-01 -7.47741938e-01 1.21066952e+00
6.20637417e-01 5.80667611e-03 1.30932614e-01 -1.72710061e-01
-3.37733120e-01 1.05567910e-01 -1.31290078e+00 1.69715214e+00
-7.36067772e-01 3.86374384e-01 5.04222751e-01 -2.67237395e-01
1.06247759e+00 -1.40456975e-01 6.55808210e-01 -4.87346619e-01
6.40011132e-01 5.05421817e-01 -4.40393418e-01 -8.51957679e-01
2.57238179e-01 4.27656323e-01 -3.44851404e-01 6.04032218e-01
-1.27491578e-01 -3.79694849e-01 -1.41332671e-01 -3.89950871e-01
1.25403273e+00 5.48371553e-01 1.16997257e-01 -1.25208646e-01
-2.66263723e-01 -2.01769665e-01 -2.00907037e-01 5.91754019e-01
-1.17836975e-01 4.15925443e-01 1.98744759e-02 -6.48898855e-02
-1.12617183e+00 -1.33129418e+00 -4.99143004e-01 1.51265192e+00
1.06491536e-01 -2.12684140e-01 -6.25839293e-01 2.19321504e-01
8.82642448e-01 -2.58689001e-02 -6.99489594e-01 3.84207070e-01
-3.64628881e-02 -4.40851934e-02 6.06370687e-01 1.05380011e+00
4.65336025e-01 -1.43509924e+00 -1.21916163e+00 3.56866837e-01
1.88588873e-01 -8.23254228e-01 7.31428936e-02 5.27557433e-01
-7.34631181e-01 -9.76331055e-01 -5.57943285e-01 -5.27146041e-01
2.13719919e-01 -7.16525083e-03 7.26386487e-01 -1.93079785e-01
-9.84743118e-01 7.42177010e-01 -5.07188141e-01 -9.54787970e-01
1.46747887e-01 2.21834645e-01 -1.28277004e-01 -4.28107738e-01
5.95639884e-01 -4.59087402e-01 -4.89803135e-01 2.80629039e-01
-6.53887928e-01 -1.06591016e-01 6.22095048e-01 4.90240633e-01
6.48769498e-01 -7.10166574e-01 4.66672599e-01 -4.40565526e-01
9.43023682e-01 -2.88604468e-01 -3.06047261e-01 6.00973479e-02
-3.46178651e-01 -2.92725265e-01 2.00145021e-01 -5.12519121e-01
-8.11415315e-01 2.31375545e-01 -4.45115775e-01 -1.62221468e-03
-5.16969323e-01 4.60332692e-01 3.43948156e-02 -3.24234158e-01
6.29180431e-01 -3.77459913e-01 2.70119697e-01 -7.50444949e-01
4.71935332e-01 1.22956908e+00 5.17799675e-01 -7.50344872e-01
2.33209014e-01 2.94265896e-01 4.09875214e-02 -1.01772058e+00
-3.81444097e-01 -6.91532075e-01 -9.28714931e-01 -3.48720849e-01
6.35738671e-01 -1.64411083e-01 -1.23700714e+00 1.17490435e+00
-6.24557912e-01 -8.92546952e-01 1.39970258e-01 1.44998729e-01
-6.54336452e-01 2.48272866e-01 -8.27243924e-01 -1.14312029e+00
-7.09038198e-01 -5.63625038e-01 1.56067026e+00 4.70103800e-01
-7.13086069e-01 -8.86308014e-01 3.79986495e-01 1.76859930e-01
6.69383645e-01 5.17545700e-01 2.18561754e-01 -1.26784757e-01
-2.15410277e-01 -7.88369358e-01 -1.85340255e-01 3.51801097e-01
4.03548419e-01 4.50551122e-01 -1.26275718e+00 -1.59511387e-01
-8.13953280e-01 -8.93845379e-01 6.51966095e-01 3.07000488e-01
1.19779718e+00 2.56079853e-01 -5.70869148e-01 2.62610853e-01
1.45560884e+00 -1.43287614e-01 6.17052495e-01 1.00117885e-01
5.77591836e-01 5.62027693e-01 8.39307010e-01 6.32499158e-01
1.09760024e-01 5.44287384e-01 4.29382235e-01 -2.01743260e-01
2.53048241e-01 -2.79381067e-01 -1.73201635e-01 5.35277486e-01
-3.68591785e-01 1.68677479e-01 -8.81588876e-01 3.36420566e-01
-1.94853079e+00 -8.42259884e-01 -2.23070323e-01 2.37470412e+00
9.18829024e-01 -3.23503115e-03 1.49574444e-01 3.10427219e-01
2.85596848e-01 -2.59736627e-01 -6.95453167e-01 -8.71508956e-01
5.39214253e-01 7.82646239e-01 4.73072380e-01 3.52549732e-01
-1.40278351e+00 9.59355116e-01 6.76931334e+00 4.49320823e-01
-1.43059671e+00 -2.48925671e-01 -2.61448324e-02 -1.84648465e-02
3.26215625e-01 -5.95889449e-01 -3.56683284e-01 3.93726140e-01
6.07473373e-01 9.85727757e-02 6.36045575e-01 9.31251824e-01
2.80809142e-02 -5.97247422e-01 -1.17280066e+00 8.26009870e-01
-5.46026349e-01 -9.69098270e-01 -6.98822021e-01 -1.53466910e-01
2.16032684e-01 2.99242198e-01 8.68491307e-02 6.14258312e-02
3.58183756e-02 -7.99588084e-01 3.21511060e-01 7.99736738e-01
1.09931922e+00 -3.34122360e-01 8.54236245e-01 1.00014515e-01
-1.44797099e+00 -9.69062466e-03 -1.81050852e-01 -5.55089951e-01
-8.55716839e-02 5.37018597e-01 -5.66898346e-01 4.47227806e-01
1.17131543e+00 5.91618598e-01 -3.12972754e-01 9.19253469e-01
6.78537413e-02 3.89781386e-01 -6.84619427e-01 -7.33315349e-01
-3.48257363e-01 1.81245103e-01 6.35451600e-02 1.49474180e+00
7.68500566e-02 -8.23384821e-02 2.07954049e-01 6.53810680e-01
4.43366736e-01 -2.11293884e-02 -8.13437581e-01 -2.52505332e-01
8.41821611e-01 1.54418886e+00 -7.69939482e-01 9.11598802e-02
-1.02720603e-01 8.91293049e-01 1.33027554e-01 -6.78818766e-03
-3.62114996e-01 -1.05109048e+00 9.51260388e-01 7.29546398e-02
2.06853420e-01 -4.85590547e-01 -1.01158142e+00 -6.47790611e-01
4.16163594e-01 -5.07885396e-01 -1.25903502e-01 -7.49269009e-01
-1.88482451e+00 -2.82041989e-02 -6.29419461e-02 -1.06050050e+00
-2.25373894e-01 -1.26551962e+00 -7.27793694e-01 9.11371768e-01
-9.98745799e-01 -1.14813530e+00 -8.62126589e-01 3.90184760e-01
9.61020440e-02 4.86885041e-01 1.30681074e+00 1.52641252e-01
-3.70368123e-01 4.04414922e-01 1.09388325e-02 2.17227638e-01
9.05702889e-01 -1.29702842e+00 6.10008657e-01 1.93263710e-01
-4.51588392e-01 7.94336200e-01 5.03333151e-01 -8.06316018e-01
-2.05158591e+00 -5.77475667e-01 3.29198837e-01 -8.88744175e-01
9.81136680e-01 -6.60304308e-01 -8.95836949e-01 4.43985760e-01
-2.52416223e-01 -1.53257266e-01 7.70208478e-01 4.25874680e-01
-2.56165326e-01 -1.70569807e-01 -1.09529555e+00 4.19296890e-01
9.51628566e-01 -7.04965591e-01 -5.66342652e-01 -1.58761278e-01
-1.26018643e-01 -7.15266824e-01 -1.27140188e+00 5.21874130e-01
1.48399568e+00 -1.02060151e+00 9.04618979e-01 -3.73189375e-02
4.54508334e-01 2.33843878e-01 -2.38370877e-02 -9.95176911e-01
-2.04324454e-01 -4.79541302e-01 -1.12498514e-01 1.13941765e+00
2.33992204e-01 -9.12858546e-01 8.53377461e-01 7.53473043e-01
-6.25679865e-02 -1.22961462e+00 -8.21594000e-01 -5.62398016e-01
-5.47579490e-02 -8.52647066e-01 1.93953261e-01 5.35989463e-01
7.99118757e-01 -7.88175613e-02 -1.02172732e-01 -2.63511300e-01
5.86801529e-01 9.55565125e-02 1.06388521e+00 -1.52778637e+00
-1.94371074e-01 -4.77420747e-01 -3.36494803e-01 -8.83989096e-01
-5.97021937e-01 -2.91997761e-01 4.82933939e-01 -1.79083157e+00
-2.18903169e-01 -5.33748984e-01 -3.16553026e-01 1.15860796e+00
-6.35938942e-02 3.38277817e-01 8.37644339e-02 7.24827722e-02
-3.66931945e-01 -8.47497489e-03 1.39461768e+00 3.98210496e-01
-2.94249654e-01 1.67603478e-01 -4.53129917e-01 3.27167958e-01
1.09205377e+00 1.40504241e-01 4.21810150e-02 -2.57938623e-01
-1.56310499e-02 -2.43106201e-01 5.93617201e-01 -1.29232955e+00
-1.04578398e-01 -2.19610602e-01 6.24553680e-01 -5.35285771e-01
6.04926407e-01 -9.10961270e-01 -1.20480642e-01 4.86254930e-01
-2.18633100e-01 -3.50203991e-01 5.22919357e-01 1.52286515e-01
3.61603647e-01 3.29199374e-01 4.45634812e-01 1.30618840e-01
-7.62204468e-01 7.64969289e-02 -3.21953148e-01 -4.45602804e-01
1.12393081e+00 -4.17436481e-01 -5.98673761e-01 -2.74278503e-02
-4.92815852e-01 1.18972830e-01 5.30670583e-01 8.34151447e-01
6.36164606e-01 -1.09960532e+00 -1.93834379e-01 5.77152491e-01
3.18976253e-01 -3.66148412e-01 1.58121094e-01 4.99505430e-01
-5.91733932e-01 4.27622646e-01 -8.20007861e-01 -6.62398458e-01
-1.33468175e+00 1.44604057e-01 -3.52617353e-02 1.25603631e-01
-5.24340212e-01 7.50054240e-01 -9.46221828e-01 -7.41871774e-01
4.14983511e-01 -9.92626667e-01 3.51871669e-01 -2.88648307e-01
4.16894227e-01 6.37428463e-01 2.36401364e-01 1.89427599e-01
-4.52502429e-01 9.32002842e-01 4.70944107e-01 -8.39446858e-02
1.10815370e+00 2.34102234e-01 4.43850793e-02 1.13339472e+00
8.43415022e-01 -1.25647381e-01 -1.45980167e+00 3.67143333e-01
-3.72316279e-02 -4.51593786e-01 -8.12384412e-02 -1.26652420e+00
-3.85325968e-01 9.19819474e-01 8.11847925e-01 3.47879738e-01
8.06707084e-01 -7.42786378e-03 1.02944791e+00 5.92862010e-01
1.02951634e+00 -1.47017121e+00 2.71549493e-01 4.42844540e-01
9.86791074e-01 -1.08985114e+00 -1.41619951e-01 -6.37455165e-01
-5.85772216e-01 1.14138985e+00 7.29431331e-01 -4.53781009e-01
8.40937912e-01 1.00645888e+00 5.93532205e-01 -2.22826339e-02
-5.56437016e-01 -1.67277366e-01 2.84400344e-01 9.29543734e-01
6.64718091e-01 4.78051186e-01 -1.27330363e-01 5.64896047e-01
-7.88224414e-02 5.83422840e-01 -2.79169410e-01 1.68015611e+00
-5.70822835e-01 -1.30733001e+00 -3.95803720e-01 8.62641335e-01
-2.04178840e-01 3.28938186e-01 -6.91030979e-01 6.47457778e-01
-8.15019235e-02 9.12407815e-01 -4.19031493e-02 -9.37036812e-01
5.16386867e-01 1.24999210e-01 8.10317278e-01 -5.93555868e-01
-6.10505104e-01 9.51512530e-03 2.76328534e-01 -1.07048094e+00
5.87504939e-04 -5.54161191e-01 -1.23026180e+00 -3.36746544e-01
-1.32525459e-01 -3.73151809e-01 1.14944267e+00 4.97222036e-01
6.33061945e-01 3.76403779e-01 3.02960247e-01 -1.78454196e+00
-4.75545496e-01 -1.21696305e+00 -7.13898480e-01 2.28533953e-01
2.75275018e-02 -7.97223508e-01 1.02211863e-01 -2.71642506e-01] | [5.922492027282715, -0.7362760305404663] |
975bfb71-4aea-4702-b062-e4fbebfd4482 | videoinr-learning-video-implicit-neural-1 | 2206.04647 | null | https://arxiv.org/abs/2206.04647v1 | https://arxiv.org/pdf/2206.04647v1.pdf | VideoINR: Learning Video Implicit Neural Representation for Continuous Space-Time Super-Resolution | Videos typically record the streaming and continuous visual data as discrete consecutive frames. Since the storage cost is expensive for videos of high fidelity, most of them are stored in a relatively low resolution and frame rate. Recent works of Space-Time Video Super-Resolution (STVSR) are developed to incorporate temporal interpolation and spatial super-resolution in a unified framework. However, most of them only support a fixed up-sampling scale, which limits their flexibility and applications. In this work, instead of following the discrete representations, we propose Video Implicit Neural Representation (VideoINR), and we show its applications for STVSR. The learned implicit neural representation can be decoded to videos of arbitrary spatial resolution and frame rate. We show that VideoINR achieves competitive performances with state-of-the-art STVSR methods on common up-sampling scales and significantly outperforms prior works on continuous and out-of-training-distribution scales. Our project page is at http://zeyuan-chen.com/VideoINR/ . | ['Xiaolong Wang', 'Humphrey Shi', 'Zhangyang Wang', 'Vidit Goel', 'Xingqian Xu', 'Jingwen Liu', 'Yinbo Chen', 'Zeyuan Chen'] | 2022-06-09 | videoinr-learning-video-implicit-neural | http://openaccess.thecvf.com//content/CVPR2022/html/Chen_VideoINR_Learning_Video_Implicit_Neural_Representation_for_Continuous_Space-Time_Super-Resolution_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Chen_VideoINR_Learning_Video_Implicit_Neural_Representation_for_Continuous_Space-Time_Super-Resolution_CVPR_2022_paper.pdf | cvpr-2022-1 | ['space-time-video-super-resolution', 'video-super-resolution'] | ['computer-vision', 'computer-vision'] | [ 4.27156627e-01 -4.18484002e-01 -6.16299868e-01 -2.13132247e-01
-8.68973911e-01 -1.00974545e-01 3.24758470e-01 -6.21640980e-01
-1.58020005e-01 8.83858621e-01 3.92203122e-01 1.18623652e-01
4.44742590e-02 -8.18671942e-01 -8.74890089e-01 -3.73690486e-01
-9.08722878e-02 -1.75970688e-01 5.75974464e-01 -1.36206567e-01
-4.35282178e-02 4.54382360e-01 -1.76955986e+00 7.96249747e-01
5.62120795e-01 8.45051408e-01 5.74576139e-01 8.89562786e-01
1.61126424e-02 1.19496167e+00 -3.80151987e-01 9.15045813e-02
3.10302526e-01 -4.98425007e-01 -7.60281384e-01 -3.63168828e-02
7.47127712e-01 -9.63869929e-01 -1.14359510e+00 1.05634928e+00
3.17418307e-01 3.94112319e-01 2.65453458e-01 -8.44020367e-01
-1.33558631e+00 5.10929823e-01 -6.62360609e-01 6.83984995e-01
4.59975034e-01 -2.89047807e-01 6.96372092e-01 -1.15649724e+00
8.00263584e-01 1.22999537e+00 6.33655250e-01 8.45370710e-01
-1.15685987e+00 -7.34245837e-01 3.64418000e-01 4.60132897e-01
-1.65328372e+00 -6.62711203e-01 4.66279417e-01 -2.48025090e-01
9.40121889e-01 1.07145064e-01 5.10390043e-01 1.03485608e+00
-2.60405988e-01 7.22451866e-01 8.40998769e-01 -1.59882337e-01
9.04168189e-02 -5.53900957e-01 -1.28401875e-01 4.00244951e-01
-5.52730188e-02 2.60316581e-01 -9.43563581e-01 2.73786634e-01
1.97554314e+00 4.18152690e-01 -6.01925671e-01 1.48349062e-01
-1.10489726e+00 7.53735900e-01 4.07715172e-01 4.29568887e-01
-3.49800825e-01 4.63694572e-01 3.94238949e-01 3.63374531e-01
5.75668454e-01 -1.93604812e-01 -3.02076906e-01 -2.10191593e-01
-1.47253847e+00 1.91038311e-01 4.20575701e-02 1.25406206e+00
6.23221040e-01 5.96481264e-01 -2.34446034e-01 9.93942916e-01
-7.56046250e-02 2.39300370e-01 6.31616116e-01 -1.68648124e+00
3.04566652e-01 -1.26063973e-01 2.60402530e-01 -7.67329812e-01
1.93665415e-01 -1.40998913e-02 -1.10334897e+00 1.95287034e-01
6.70577213e-02 2.36903727e-01 -9.70798016e-01 1.48642123e+00
5.64611033e-02 9.26033318e-01 2.99222544e-02 1.29449117e+00
1.06453657e+00 1.19634855e+00 -1.32047251e-01 -6.08204365e-01
1.22746360e+00 -1.10882878e+00 -1.01477635e+00 1.82331175e-01
-4.89100292e-02 -4.72945958e-01 9.29774106e-01 3.73763740e-01
-1.66798985e+00 -9.30606365e-01 -1.05223513e+00 -6.44379318e-01
8.95211250e-02 2.37835031e-02 3.61520588e-01 -1.71827115e-02
-1.42659688e+00 9.43951070e-01 -1.01352823e+00 1.36548537e-03
3.51531178e-01 1.87505931e-01 -3.35653007e-01 -3.88413459e-01
-1.35526514e+00 3.86652142e-01 2.25923076e-01 -7.06465170e-02
-1.03168905e+00 -8.43795657e-01 -8.43195379e-01 -3.39161009e-02
3.07690233e-01 -5.48066795e-01 1.38737440e+00 -1.16174722e+00
-1.51592791e+00 4.69891310e-01 -5.82075357e-01 -6.17187262e-01
5.08583128e-01 -4.31515157e-01 -6.36235476e-01 6.05482996e-01
-5.52093908e-02 6.54046237e-01 1.24901712e+00 -1.03042281e+00
-7.27799833e-01 -9.69107896e-02 1.81927264e-01 2.01602280e-01
-2.68183172e-01 3.16654742e-01 -8.08032990e-01 -1.10047412e+00
-8.77771154e-02 -3.84780675e-01 -1.36584148e-01 2.80574530e-01
4.00796533e-01 9.54053272e-03 9.87453103e-01 -1.02583981e+00
1.40840328e+00 -2.23261905e+00 2.63173759e-01 -5.89707375e-01
3.12859148e-01 3.66335750e-01 -1.88936204e-01 2.82491297e-02
-1.38696373e-01 1.08827323e-01 -3.59221473e-02 -3.14536572e-01
-4.55365956e-01 3.05321962e-01 -6.27123713e-01 4.06570941e-01
1.01738639e-01 6.46408796e-01 -9.03271019e-01 -6.07189655e-01
3.26576471e-01 1.28465259e+00 -5.60366035e-01 2.85388649e-01
1.00645116e-02 3.94474626e-01 -2.69430190e-01 5.48804641e-01
7.19239295e-01 -4.98162657e-01 1.37409568e-01 -4.61899549e-01
-3.16248298e-01 2.39987105e-01 -1.05012548e+00 1.87395024e+00
-3.62958223e-01 8.72685611e-01 1.25979304e-01 -7.61408806e-01
7.05608904e-01 5.29879689e-01 5.24763107e-01 -8.76439631e-01
-3.31183195e-01 1.21240012e-01 -6.67549074e-01 -2.95035839e-01
1.06673944e+00 3.44997235e-02 5.60311913e-01 -8.05441663e-02
6.59010708e-02 3.91029596e-01 3.05469155e-01 1.72016457e-01
8.23911130e-01 5.45375347e-01 3.53971988e-01 1.21878564e-01
3.14544708e-01 -3.97515237e-01 6.93758845e-01 5.93288124e-01
-1.51715696e-01 1.21813822e+00 3.68892476e-02 -5.74793935e-01
-1.55488205e+00 -1.20798659e+00 -2.30465978e-01 1.25196302e+00
2.99345493e-01 -4.26419973e-01 -5.99368036e-01 -1.13283359e-01
-2.45584890e-01 3.41975212e-01 -6.55353725e-01 2.95441329e-01
-1.03268039e+00 -8.10304731e-02 2.59980738e-01 8.85716140e-01
4.85817343e-01 -9.53475356e-01 -5.26737571e-01 3.28383237e-01
-3.74550849e-01 -1.39084339e+00 -6.64072037e-01 -4.28160101e-01
-1.16298807e+00 -6.11483812e-01 -1.32077241e+00 -8.05051565e-01
3.13663095e-01 8.85678053e-01 1.05327904e+00 1.08490951e-01
-1.55913442e-01 1.87602371e-01 -6.21856809e-01 4.40788388e-01
-1.78196073e-01 -2.67176151e-01 1.34482682e-01 -9.15962085e-02
-1.57974800e-03 -8.50273669e-01 -9.10249770e-01 3.59254122e-01
-1.11185944e+00 3.80593389e-01 2.35941991e-01 8.07258785e-01
1.08920288e+00 -2.06517037e-02 5.23271322e-01 -6.17226124e-01
1.23057947e-01 -3.76047641e-01 -5.62930405e-01 1.18228244e-02
-2.70575643e-01 -2.91817844e-01 7.54018366e-01 -7.35631526e-01
-9.63206649e-01 -2.55225420e-01 -1.10008061e-01 -1.24131608e+00
-4.73267166e-03 3.35529119e-01 2.18788221e-01 1.06322676e-01
4.78704453e-01 5.82536936e-01 -1.95220247e-01 -6.08401716e-01
3.88150573e-01 4.56708729e-01 7.72733569e-01 -3.59384537e-01
7.28403389e-01 5.62206388e-01 -3.71793270e-01 -9.67783868e-01
-7.57763326e-01 -1.91761494e-01 -5.42112648e-01 -2.27399930e-01
7.94059575e-01 -1.50907862e+00 -2.03721076e-01 1.97606668e-01
-9.96109486e-01 -6.10098720e-01 -3.82659405e-01 5.27540326e-01
-7.61505246e-01 5.17654359e-01 -1.26270199e+00 -6.33574426e-01
-1.40373871e-01 -1.10559273e+00 1.08853555e+00 2.66369343e-01
1.23468533e-01 -6.06227040e-01 -1.91270903e-01 2.37074271e-01
6.74540997e-01 8.33458006e-02 1.72793478e-01 2.69489199e-01
-9.71397102e-01 2.75030285e-01 -6.17024124e-01 2.76427299e-01
1.40378445e-01 8.11671019e-02 -7.19030142e-01 -6.65372849e-01
-1.09182313e-01 -2.36935109e-01 9.43969011e-01 7.85737038e-01
1.61775434e+00 -5.68741500e-01 3.77953984e-02 1.06805182e+00
1.68114889e+00 1.30840898e-01 1.04078627e+00 1.69609860e-01
7.06789196e-01 5.96362092e-02 7.12321341e-01 6.00721657e-01
1.70309424e-01 8.63631785e-01 2.32467830e-01 5.74473105e-02
-4.96980578e-01 -2.73249745e-01 6.71186388e-01 9.66146290e-01
-9.04337108e-01 3.35549489e-02 -2.85811394e-01 6.14132583e-01
-1.98066401e+00 -1.49933660e+00 1.94786400e-01 2.02570081e+00
9.06023920e-01 -2.11779863e-01 2.83016592e-01 -4.72683832e-02
1.12689698e+00 5.98747730e-01 -6.26917839e-01 -1.66264147e-01
-2.07881048e-01 1.16471075e-01 4.85563725e-01 5.70829570e-01
-9.96186912e-01 9.15545642e-01 6.70359468e+00 1.07052696e+00
-1.04012334e+00 4.21821803e-01 6.10799253e-01 -4.87308919e-01
-1.92567796e-01 -3.41721803e-01 -7.93984234e-01 4.42892283e-01
1.14342165e+00 -3.24036360e-01 9.94862318e-01 8.38653684e-01
3.69990200e-01 4.87615854e-01 -9.78114903e-01 1.49661124e+00
-8.49535465e-02 -1.81847334e+00 3.80755275e-01 -9.96177644e-02
9.05779958e-01 9.10301730e-02 2.79224664e-01 2.79968292e-01
1.01984084e-01 -1.32730126e+00 8.56234789e-01 5.65499127e-01
1.71862102e+00 -7.43160009e-01 2.67345339e-01 -4.23081443e-02
-1.82696712e+00 -4.86691445e-02 -7.64050066e-01 3.50545719e-02
4.05862749e-01 1.29768938e-01 -2.45905425e-02 4.72069949e-01
1.10181904e+00 1.25065923e+00 -1.42315194e-01 6.30386829e-01
1.07698500e-01 4.92733419e-01 5.59536880e-03 5.88356376e-01
-4.97450903e-02 1.64283288e-03 3.51060510e-01 1.32363737e+00
7.24417925e-01 5.12123406e-01 1.62531674e-01 6.93770587e-01
-2.37149313e-01 -2.90000796e-01 -4.62739378e-01 -8.33454654e-02
7.33268619e-01 8.51516843e-01 -3.16714883e-01 -5.52016854e-01
-8.17704916e-01 1.21509862e+00 3.14146191e-01 6.88984811e-01
-1.12424505e+00 -7.35082254e-02 7.46149421e-01 4.13196504e-01
7.52371013e-01 -4.68380123e-01 1.50367886e-01 -1.47304237e+00
-5.78465089e-02 -8.11036766e-01 3.23997736e-01 -1.05698550e+00
-1.14347720e+00 7.20265806e-01 7.80410692e-02 -1.61388934e+00
-3.36143970e-01 -3.56853038e-01 9.96555164e-02 8.22065115e-01
-1.75030577e+00 -1.00606024e+00 -4.93493706e-01 9.89970982e-01
1.17594588e+00 -9.21728685e-02 5.63298345e-01 6.13961041e-01
-2.56290287e-01 6.36008441e-01 3.23264837e-01 2.54967690e-01
5.43026268e-01 -7.67744184e-01 5.04815221e-01 9.62129056e-01
-6.79755509e-02 4.88564610e-01 5.84964991e-01 -5.99834621e-01
-1.39026296e+00 -1.36670411e+00 2.09393784e-01 2.99270749e-02
5.61886132e-01 -7.71953240e-02 -1.26740909e+00 8.46387327e-01
1.93115011e-01 7.01656163e-01 2.89636970e-01 -4.36816692e-01
-5.29252350e-01 2.90277079e-02 -1.00244546e+00 6.48647070e-01
1.36198032e+00 -8.89385104e-01 -4.16387469e-01 -5.63568203e-03
1.27616775e+00 -5.88985920e-01 -1.24862540e+00 3.87884527e-01
6.07380450e-01 -1.07095921e+00 1.34774876e+00 -4.15943444e-01
8.59536648e-01 -6.18420660e-01 -6.16507709e-01 -6.42240763e-01
-8.64221811e-01 -5.45058429e-01 -8.97625506e-01 1.00980997e+00
-1.49538219e-01 -2.45296761e-01 6.60803795e-01 2.68983573e-01
-7.50199258e-02 -6.53124928e-01 -9.50840116e-01 -7.77578652e-01
-1.80304110e-01 -2.06521362e-01 4.57434297e-01 9.66143727e-01
-4.83809382e-01 1.24115525e-02 -7.85592556e-01 3.12376976e-01
7.86518574e-01 1.27751395e-01 3.05956811e-01 -8.12522173e-01
-4.28410828e-01 -8.89561623e-02 -3.36132258e-01 -1.55324626e+00
-1.26324594e-01 -5.20472169e-01 -1.74911782e-01 -1.53898609e+00
1.99438840e-01 1.39606949e-02 -4.61062759e-01 1.65928379e-01
3.60152833e-02 5.82420886e-01 2.74167359e-01 5.87704003e-01
-6.84622943e-01 5.13893366e-01 1.38787997e+00 6.09720424e-02
-1.03226885e-01 -6.03672683e-01 -3.48811507e-01 6.78838491e-01
6.49554193e-01 -1.65943746e-02 -4.63303417e-01 -7.61557400e-01
-1.73777252e-01 6.49737477e-01 3.78531873e-01 -9.60162759e-01
2.89751813e-02 -4.27642882e-01 5.88730097e-01 -7.13156164e-01
6.85712218e-01 -5.89743614e-01 4.85893726e-01 3.04775909e-02
-5.91006756e-01 1.96669355e-01 -4.45038490e-02 6.78847492e-01
-3.09194088e-01 1.57554820e-02 1.11246252e+00 -2.84639865e-01
-1.12488341e+00 8.10779452e-01 -3.42938244e-01 4.07104865e-02
6.88302577e-01 -4.27480966e-01 -3.93227607e-01 -4.66535419e-01
-7.85571933e-01 -1.97755143e-01 7.91122794e-01 4.79511887e-01
1.30570126e+00 -1.59482622e+00 -8.98944736e-01 -5.92807755e-02
-3.41751158e-01 1.67936698e-01 8.28648984e-01 4.12783444e-01
-7.05136776e-01 8.01219642e-02 -3.75110716e-01 -5.45356333e-01
-1.33071840e+00 9.09980059e-01 1.70600519e-01 4.08995375e-02
-1.23816383e+00 6.09417260e-01 4.33830947e-01 4.72904295e-01
1.77427143e-01 -1.74404711e-01 -3.66572797e-01 -2.80795991e-01
1.34208870e+00 4.73567218e-01 -5.03219545e-01 -8.92477632e-01
-3.37728392e-03 6.88889325e-01 -1.74759924e-01 -6.84456751e-02
1.52595735e+00 -3.86916101e-01 1.28004387e-01 4.70533639e-01
1.25870144e+00 -2.31038675e-01 -1.84531724e+00 -4.35766757e-01
-5.74106038e-01 -1.01533413e+00 1.21853814e-01 -1.76496238e-01
-1.26796579e+00 8.25053990e-01 6.66575789e-01 6.78652376e-02
1.34922147e+00 -9.11004543e-02 9.27032351e-01 2.14285720e-02
6.50146604e-01 -9.93949711e-01 6.68756291e-02 4.98229295e-01
1.08151460e+00 -9.14235055e-01 1.58685908e-01 -4.49511707e-01
-5.20054400e-01 1.16232121e+00 4.80260402e-01 -5.35624444e-01
3.40608269e-01 5.11186719e-01 -3.02072942e-01 4.21519279e-01
-9.42591369e-01 1.44450990e-02 1.30238861e-01 8.29546571e-01
6.88430190e-01 -3.95948328e-02 -9.86163542e-02 3.79064709e-01
1.88773319e-01 5.47034979e-01 7.65095234e-01 6.48676097e-01
-4.28967953e-01 -6.56682611e-01 -3.78730208e-01 1.58065945e-01
-6.88247323e-01 -2.15574756e-01 4.23602164e-01 5.79995334e-01
-1.45630896e-01 6.09937310e-01 1.91282555e-01 -2.95793205e-01
-1.10543156e-02 -4.10757780e-01 6.66663826e-01 -3.39602888e-01
4.52686138e-02 2.82125533e-01 -2.17438452e-02 -9.89123285e-01
-7.84070611e-01 -4.96070236e-01 -1.32605064e+00 -4.97765332e-01
1.08685389e-01 -1.05113246e-01 1.76248580e-01 4.24226433e-01
4.13266659e-01 7.71853924e-01 4.24992770e-01 -1.34917402e+00
-1.83481112e-01 -7.14361131e-01 -6.74087763e-01 3.74603748e-01
7.52582371e-01 -5.93411922e-01 -7.39331543e-02 5.63192666e-01] | [11.084028244018555, -1.8229509592056274] |
9eccd9c5-1b99-446f-a9a0-96325211b023 | iterative-self-knowledge-distillation-from | 2202.02265 | null | https://arxiv.org/abs/2202.02265v1 | https://arxiv.org/pdf/2202.02265v1.pdf | Iterative Self Knowledge Distillation -- From Pothole Classification to Fine-Grained and COVID Recognition | Pothole classification has become an important task for road inspection vehicles to save drivers from potential car accidents and repair bills. Given the limited computational power and fixed number of training epochs, we propose iterative self knowledge distillation (ISKD) to train lightweight pothole classifiers. Designed to improve both the teacher and student models over time in knowledge distillation, ISKD outperforms the state-of-the-art self knowledge distillation method on three pothole classification datasets across four lightweight network architectures, which supports that self knowledge distillation should be done iteratively instead of just once. The accuracy relation between the teacher and student models shows that the student model can still benefit from a moderately trained teacher model. Implying that better teacher models generally produce better student models, our results justify the design of ISKD. In addition to pothole classification, we also demonstrate the efficacy of ISKD on six additional datasets associated with generic classification, fine-grained classification, and medical imaging application, which supports that ISKD can serve as a general-purpose performance booster without the need of a given teacher model and extra trainable parameters. | ['Kuan-Chuan Peng'] | 2022-02-04 | null | null | null | null | ['self-knowledge-distillation'] | ['computer-vision'] | [ 2.41697058e-02 6.49366200e-01 -3.20787340e-01 -5.19285560e-01
-6.16944671e-01 -2.31281012e-01 2.14757144e-01 3.57088536e-01
-7.86220491e-01 7.00013638e-01 -2.01748356e-01 -7.57725418e-01
-3.78011197e-01 -1.04412818e+00 -7.05390513e-01 -4.84196424e-01
3.15949112e-01 3.23874772e-01 6.95187569e-01 -1.80757549e-02
-1.97044879e-01 5.54720223e-01 -1.45732069e+00 1.57501817e-01
1.13188612e+00 1.05592895e+00 2.03591380e-02 6.71251535e-01
9.36578959e-02 9.97867644e-01 -4.37026739e-01 -4.61013049e-01
9.51682776e-02 -3.75286415e-02 -9.40049231e-01 -3.72086585e-01
6.18662059e-01 -3.76812249e-01 -3.66211444e-01 6.58854008e-01
7.37847567e-01 3.53691131e-01 4.26191419e-01 -1.22881567e+00
-6.53168380e-01 6.60213947e-01 -2.75279939e-01 4.83987927e-01
-4.58200067e-01 3.11564863e-01 8.24795127e-01 -7.82220960e-01
2.26751268e-02 9.99322712e-01 1.15964389e+00 5.49724340e-01
-1.11637866e+00 -8.74410152e-01 -1.20021835e-01 4.91232842e-01
-1.27097654e+00 -3.04660887e-01 6.22461915e-01 -2.77709574e-01
1.16178489e+00 9.20644253e-02 9.34659362e-01 7.41302729e-01
9.53841731e-02 8.43917191e-01 1.06784344e+00 -3.08682501e-01
2.12432146e-01 6.01232231e-01 5.50386190e-01 9.92087960e-01
4.60721761e-01 5.16616590e-02 -5.56819201e-01 -2.96097342e-02
6.30837679e-01 -1.67075604e-01 4.24284860e-02 -3.17974985e-01
-6.53211355e-01 9.27005589e-01 6.51613653e-01 1.48609020e-02
-2.34879479e-01 4.07094598e-01 3.48307967e-01 2.73807913e-01
6.43119395e-01 5.74104130e-01 -8.37503076e-01 -3.41287911e-01
-8.83407831e-01 -2.58427747e-02 9.05722618e-01 7.34725833e-01
1.17332721e+00 6.84981197e-02 -1.89087585e-01 8.11066568e-01
1.16179928e-01 3.93264502e-01 4.42529321e-01 -5.97412884e-01
6.17099889e-02 9.51272249e-01 -4.42699730e-01 -5.44616640e-01
-7.20385790e-01 -8.90393496e-01 -6.75419569e-01 1.76470742e-01
2.16406271e-01 -2.60493755e-01 -1.23324776e+00 1.33141768e+00
5.70064604e-01 5.85511327e-01 3.71649086e-01 3.49057525e-01
1.29820788e+00 2.06620663e-01 2.44582132e-01 5.80493696e-02
1.44479918e+00 -1.12063122e+00 -2.76041925e-01 -4.36264843e-01
1.15923452e+00 -2.96663642e-01 1.07506859e+00 2.84008414e-01
-8.37612212e-01 -5.63055456e-01 -1.06983495e+00 -1.31509349e-01
-5.08545995e-01 6.44850209e-02 9.36044812e-01 9.22326207e-01
-9.75419044e-01 6.86272860e-01 -7.99545884e-01 -2.40168557e-01
1.03120172e+00 4.30966973e-01 -1.89711437e-01 -3.71179953e-02
-1.21154320e+00 1.30873632e+00 5.27747571e-01 1.89607605e-01
-7.84154832e-01 -1.43391573e+00 -7.50605822e-01 1.97851822e-01
2.54968524e-01 -7.56676376e-01 1.47722697e+00 -4.55459535e-01
-1.51854050e+00 6.79529130e-01 3.41821045e-01 -7.58806169e-01
4.75594670e-01 -3.84009928e-01 -1.81943476e-01 2.98294574e-01
-2.27346137e-01 8.74530792e-01 8.44281256e-01 -7.10736752e-01
-7.36967981e-01 -1.63625926e-02 6.75776005e-02 2.08429039e-01
-5.27727544e-01 -4.93465155e-01 -1.34561136e-01 -1.60835460e-01
-1.86349675e-01 -7.69149661e-01 -2.01760590e-01 7.05334693e-02
-8.59089568e-02 -8.14751863e-01 1.10407138e+00 -2.94072509e-01
1.00159907e+00 -2.11790538e+00 -6.92565501e-01 2.26758778e-01
5.64121425e-01 6.91223502e-01 -1.78051725e-01 -1.30492061e-01
-1.95710406e-01 -2.37507205e-02 -1.58082489e-02 1.78302396e-02
-2.32892543e-01 6.05123758e-01 1.31277859e-01 3.26108724e-01
5.32773256e-01 1.25929022e+00 -1.11876464e+00 -4.96765584e-01
2.75357425e-01 5.71813166e-01 -6.36557817e-01 5.56516089e-02
1.85902789e-02 5.61052235e-03 -4.28036541e-01 3.07469547e-01
7.00792432e-01 -4.38134015e-01 -3.97123843e-01 -3.87505174e-01
2.45658204e-01 3.42824489e-01 -7.64452875e-01 1.33640480e+00
-5.38713217e-01 6.94004953e-01 -4.41945940e-01 -1.16044962e+00
9.81043994e-01 1.55326888e-01 1.74674720e-01 -9.42264974e-01
7.17036426e-02 2.46082604e-01 1.72714695e-01 -6.43975854e-01
4.45491850e-01 -1.53282121e-01 2.56207526e-01 3.11932892e-01
2.56639808e-01 -3.03427815e-01 -3.94282676e-02 2.36998111e-01
1.05367327e+00 -2.17734426e-01 -8.06952491e-02 -2.53837347e-01
8.33524019e-02 1.41368404e-01 3.67043048e-01 9.91726100e-01
-3.01241249e-01 1.14820667e-01 3.46952200e-01 -6.29618168e-01
-8.53919625e-01 -1.01718283e+00 -2.68194079e-01 1.28585947e+00
4.00689803e-02 -1.03735574e-01 -7.74859726e-01 -1.05043221e+00
2.37470806e-01 6.40717447e-01 -7.59102404e-01 -6.73301935e-01
-3.02645892e-01 -8.47980261e-01 9.77381766e-01 8.93745661e-01
9.41160262e-01 -6.64069533e-01 -8.90664876e-01 1.80369526e-01
1.68163836e-01 -1.13204324e+00 -4.09680009e-02 4.87928301e-01
-9.66204345e-01 -1.31936538e+00 -4.96422023e-01 -7.01238394e-01
6.09598935e-01 2.50593483e-01 1.02766895e+00 2.47696191e-01
-5.00683188e-01 6.18037701e-01 -2.85855353e-01 -7.70205855e-01
-4.22169238e-01 5.30641258e-01 4.94966917e-02 -3.45336646e-01
4.85415220e-01 -5.76704383e-01 -8.03720295e-01 1.17249943e-01
-5.78136027e-01 3.99987102e-01 8.76878858e-01 9.52416778e-01
3.37814152e-01 3.67250472e-01 9.35429394e-01 -9.34794962e-01
6.12911820e-01 -3.57639045e-01 -2.43300050e-01 2.60627717e-01
-9.96100962e-01 1.93899289e-01 3.26282799e-01 -6.17673218e-01
-1.01392794e+00 -2.46481404e-01 -2.13828936e-01 -4.24135089e-01
-1.37470029e-02 5.94834030e-01 4.08371180e-01 -3.84931803e-01
1.01651335e+00 -9.85700637e-03 1.99856296e-01 -4.60976928e-01
4.17166114e-01 5.49308896e-01 4.83142525e-01 -5.36201596e-01
7.74867296e-01 2.43319497e-01 -5.62184975e-02 -7.97782779e-01
-1.33706248e+00 -4.18050855e-01 -5.16949415e-01 -9.18203685e-03
7.33746111e-01 -9.31711674e-01 -9.57398534e-01 5.43322325e-01
-7.95534372e-01 -8.99948895e-01 -7.67831743e-01 4.44507986e-01
-1.98184252e-01 -1.09534584e-01 -2.29353383e-01 -5.60346067e-01
-6.17932677e-01 -8.46872032e-01 5.18624783e-01 5.97892046e-01
3.05748004e-02 -1.25652027e+00 -2.84368336e-01 4.98187095e-01
6.94339395e-01 -7.75692686e-02 1.00324547e+00 -8.99661779e-01
-3.10421795e-01 -2.39731386e-01 -6.28333807e-01 4.92724299e-01
1.33470967e-01 -3.78045231e-01 -1.35826838e+00 -2.10779756e-01
-2.75832355e-01 -8.04099619e-01 1.04446423e+00 -9.78943333e-03
1.24342394e+00 -3.54674041e-01 -2.33528405e-01 5.10602832e-01
1.09711635e+00 -1.67962968e-01 4.44179147e-01 3.60904366e-01
8.54502082e-01 3.39637637e-01 2.43051529e-01 -1.01186875e-02
6.65246665e-01 1.92902654e-01 2.49775991e-01 -5.33656240e-01
-4.55220014e-01 -2.39551783e-01 -4.02204366e-03 6.76946044e-01
1.11826852e-01 2.82067597e-01 -1.13345814e+00 7.84614444e-01
-1.72227478e+00 -5.76588154e-01 -1.01837508e-01 1.81159604e+00
1.09705234e+00 2.07518294e-01 3.85370851e-02 1.39005214e-01
9.14164931e-02 -1.81029573e-01 -7.47525513e-01 -5.76947212e-01
3.42301250e-01 5.76798439e-01 7.82652318e-01 2.75351852e-01
-9.35889423e-01 9.96947646e-01 6.46216631e+00 1.25732791e+00
-1.22553718e+00 3.05863082e-01 6.45206928e-01 8.49183723e-02
-1.04984276e-01 -1.41925722e-01 -9.39460516e-01 1.49928942e-01
1.23669100e+00 -9.05870181e-03 1.45820647e-01 1.01605785e+00
6.54260069e-02 -5.03320575e-01 -1.16204727e+00 8.11455250e-01
-2.73202449e-01 -1.54759300e+00 3.29414994e-04 -3.25030565e-01
7.48944998e-01 3.23420584e-01 5.57993464e-02 7.22646415e-01
5.16594708e-01 -1.19387448e+00 2.20653027e-01 2.26909995e-01
7.24694788e-01 -7.02563167e-01 7.29192615e-01 4.63976681e-01
-1.01880598e+00 -1.43798023e-01 -3.96234751e-01 -6.89265877e-02
-2.61617601e-01 7.27028072e-01 -1.41009474e+00 2.90847421e-01
9.04904127e-01 4.64638025e-01 -9.06937838e-01 1.14919293e+00
-3.87799859e-01 1.11198866e+00 -4.45011944e-01 -4.64195982e-02
3.98044020e-01 4.83412296e-01 -2.79189199e-02 1.35540628e+00
-5.26798889e-02 2.91809082e-01 2.00135082e-01 6.50725365e-01
-6.40405491e-02 -1.06420733e-01 -3.37312639e-01 6.41765893e-02
5.40276170e-01 1.30799294e+00 -7.98602104e-01 -4.59826380e-01
-3.15898567e-01 5.51364362e-01 4.91540343e-01 2.40993500e-01
-8.33131313e-01 -6.16347075e-01 4.64625388e-01 3.07087936e-02
3.37669134e-01 1.09967049e-02 -3.66196811e-01 -5.77625453e-01
-2.79686898e-01 -3.79814804e-01 3.11225593e-01 -7.02176213e-01
-1.02250934e+00 2.75207728e-01 9.24808756e-02 -8.11644733e-01
2.73125172e-01 -5.52784204e-01 -7.17710793e-01 6.68971479e-01
-2.20527697e+00 -1.26298964e+00 -7.24912703e-01 6.90286219e-01
2.92152375e-01 1.43563179e-02 7.72730887e-01 1.81450725e-01
-5.96585989e-01 1.12419927e+00 -2.11225346e-01 2.70879120e-01
5.42497694e-01 -1.11815763e+00 2.68658727e-01 3.76815349e-01
-1.33523494e-01 3.07085127e-01 4.18182641e-01 -4.45362210e-01
-1.29527676e+00 -1.43373990e+00 6.71240270e-01 -5.87121606e-01
7.32261002e-01 5.22842966e-02 -1.03291750e+00 4.24380392e-01
-1.74073741e-01 4.42594796e-01 7.31101334e-01 2.21001044e-01
-4.14568454e-01 -3.87839794e-01 -1.28397393e+00 2.10863084e-01
7.69646645e-01 -6.11188233e-01 -6.13016963e-01 3.48875374e-01
9.38120306e-01 -6.46727264e-01 -1.03381467e+00 5.71308553e-01
5.68763912e-01 -6.11436009e-01 7.80559003e-01 -5.98200560e-01
3.04943055e-01 4.58569303e-02 5.76131165e-01 -1.33178520e+00
-2.05553100e-01 -3.32242250e-01 -4.07014668e-01 9.28743720e-01
6.22359991e-01 -8.70729327e-01 1.19303226e+00 8.51140797e-01
-3.40058625e-01 -1.21415913e+00 -9.55504894e-01 -8.38989317e-01
1.09759748e-01 -6.37697160e-01 4.08790320e-01 1.01586890e+00
-1.96958080e-01 5.41485250e-01 -6.79771975e-02 1.38028771e-01
5.65447092e-01 -4.90997583e-01 7.01829612e-01 -1.47368097e+00
-2.12297201e-01 -2.28004262e-01 -4.94271129e-01 -9.02829468e-01
9.24554393e-02 -9.67140436e-01 -6.49263635e-02 -1.67012858e+00
2.97473073e-01 -1.10804200e+00 -4.09692734e-01 1.32233346e+00
-3.90848488e-01 4.66816336e-01 -1.22109480e-01 -2.12708473e-01
-6.20854557e-01 4.79195982e-01 1.42308569e+00 -2.90333271e-01
-4.42885190e-01 1.35520414e-01 -9.65695679e-01 7.06796467e-01
8.57356131e-01 -5.44350207e-01 -9.35930908e-01 -5.13586700e-01
3.06569129e-01 -4.63582814e-01 5.93907118e-01 -1.03738582e+00
5.66721201e-01 -1.95961352e-02 3.46436143e-01 -3.78098130e-01
1.68594927e-01 -7.92725682e-01 -2.83621818e-01 6.52898431e-01
-5.42308688e-02 -3.96305740e-01 7.65155971e-01 4.66175646e-01
2.64673203e-01 -2.16830075e-01 9.79644954e-01 7.39188269e-02
-9.07674432e-01 3.34902078e-01 -2.97827035e-01 1.11244857e-01
1.10818386e+00 -7.53270447e-01 -7.43769586e-01 6.47719055e-02
-7.15988696e-01 6.15098119e-01 -1.47227615e-01 4.05795187e-01
7.88332760e-01 -1.06502247e+00 -6.57543898e-01 3.15724969e-01
5.02850786e-02 5.89054346e-01 3.74380916e-01 1.02373147e+00
-4.81670111e-01 2.73464948e-01 4.13809307e-02 -6.87899768e-01
-1.19172883e+00 1.50308475e-01 6.02120340e-01 -3.22859883e-01
-9.14038539e-01 1.23560572e+00 -4.98191919e-03 -6.18304551e-01
3.10535908e-01 -5.61326325e-01 -1.41651988e-01 1.43702896e-02
5.10268033e-01 6.15190327e-01 4.73717719e-01 1.25730410e-01
-1.99921742e-01 1.80561095e-01 -4.48070079e-01 5.36479473e-01
1.38878715e+00 3.52340579e-01 2.27507979e-01 2.03992978e-01
1.00149393e+00 -5.81891119e-01 -1.39659142e+00 -2.79628873e-01
-1.58601299e-01 -3.09877060e-02 3.76957864e-01 -9.98960555e-01
-1.09601831e+00 9.04250681e-01 7.74728179e-01 -6.52994215e-02
1.02736235e+00 1.17584996e-01 7.67762661e-01 8.73739660e-01
1.17597014e-01 -1.05911684e+00 3.14064503e-01 6.04561806e-01
3.71341646e-01 -1.29072380e+00 4.73795943e-02 -3.75699729e-01
-3.93542439e-01 9.42799628e-01 9.34859753e-01 3.10812909e-02
7.53250003e-01 2.01633856e-01 1.22174628e-01 -3.96072865e-01
-7.60094166e-01 -1.79139659e-01 3.16880822e-01 6.94852471e-01
2.48288706e-01 -1.07943892e-01 1.17085375e-01 5.57862341e-01
-3.90427619e-01 2.24408731e-01 4.24873084e-01 8.48130524e-01
-7.10326672e-01 -7.25838900e-01 5.25937192e-02 9.21599805e-01
-1.45367101e-01 -2.09763944e-01 4.39988188e-02 8.05095732e-01
4.25959945e-01 7.10355580e-01 1.76556632e-01 -3.20124716e-01
4.10735071e-01 1.06710762e-01 5.32801270e-01 -8.29364061e-01
-7.43728042e-01 -7.20775425e-01 1.95287734e-01 -4.05330092e-01
-2.61911899e-01 1.37820944e-01 -1.32542467e+00 -3.57155651e-01
-8.56499732e-01 4.38012153e-01 5.44798255e-01 1.13950002e+00
3.32373410e-01 8.42338562e-01 2.79965937e-01 -5.71671426e-01
-6.86745763e-01 -9.43646014e-01 -4.00869280e-01 -2.94110216e-02
4.28406626e-01 -7.02763557e-01 -1.53605029e-01 -1.64973646e-01] | [9.478107452392578, 3.248389959335327] |
9ed72fe9-e48e-4639-8ba0-54b012b68e9c | big-little-adaptive-neural-networks-on-low | 2304.09695 | null | https://arxiv.org/abs/2304.09695v1 | https://arxiv.org/pdf/2304.09695v1.pdf | Big-Little Adaptive Neural Networks on Low-Power Near-Subthreshold Processors | This paper investigates the energy savings that near-subthreshold processors can obtain in edge AI applications and proposes strategies to improve them while maintaining the accuracy of the application. The selected processors deploy adaptive voltage scaling techniques in which the frequency and voltage levels of the processor core are determined at the run-time. In these systems, embedded RAM and flash memory size is typically limited to less than 1 megabyte to save power. This limited memory imposes restrictions on the complexity of the neural networks model that can be mapped to these devices and the required trade-offs between accuracy and battery life. To address these issues, we propose and evaluate alternative 'big-little' neural network strategies to improve battery life while maintaining prediction accuracy. The strategies are applied to a human activity recognition application selected as a demonstrator that shows that compared to the original network, the best configurations obtain an energy reduction measured at 80% while maintaining the original level of inference accuracy. | ['Jose Nunez-Yanez', 'Neil Howard', 'Zichao Shen'] | 2023-04-19 | null | null | null | null | ['human-activity-recognition', 'human-activity-recognition'] | ['computer-vision', 'time-series'] | [ 4.41426009e-01 -1.02293550e-03 -4.15093213e-01 -2.49252602e-01
-3.18852030e-02 -2.09283501e-01 1.54314175e-01 2.58875489e-01
-5.77783585e-01 8.76653552e-01 -5.22718608e-01 -4.11162615e-01
-3.18953484e-01 -8.03578019e-01 -5.98591805e-01 -7.32447982e-01
6.25648275e-02 2.19294026e-01 3.46825391e-01 1.04102947e-01
2.09477067e-01 7.12817848e-01 -2.06224298e+00 9.71840471e-02
6.38617218e-01 1.31607604e+00 2.17925422e-02 7.88012207e-01
2.85861582e-01 6.66522264e-01 -6.70065939e-01 4.14547533e-01
8.11901465e-02 -2.08511159e-01 -2.02600315e-01 -7.06485271e-01
-3.25727537e-02 -2.89442420e-01 -3.10504943e-01 9.56213057e-01
4.58770126e-01 5.34625836e-02 6.19331360e-01 -1.39431620e+00
3.60697694e-02 6.23601079e-01 -2.34909415e-01 6.83621645e-01
-3.39997411e-02 1.40274271e-01 4.07315880e-01 -1.56377897e-01
2.39754319e-01 4.45129484e-01 6.17380500e-01 5.55325985e-01
-1.31652069e+00 -7.30366111e-01 -2.50361621e-01 5.15809596e-01
-1.81443715e+00 -8.33514869e-01 7.01751888e-01 1.41308755e-01
1.93978143e+00 4.61932868e-01 8.47500682e-01 7.36144841e-01
9.59730864e-01 -7.11504668e-02 9.11803722e-01 -5.47187090e-01
1.02936351e+00 2.70801723e-01 4.19860989e-01 2.67726570e-01
8.12266588e-01 1.26460522e-01 -8.22955847e-01 -8.42455924e-02
3.30077678e-01 -4.46162760e-01 -1.58141345e-01 -8.01998898e-02
-4.79460984e-01 2.27192700e-01 2.06779316e-01 6.64111555e-01
-5.32821357e-01 7.36671865e-01 6.21120989e-01 -4.63280939e-02
7.60670230e-02 3.88194531e-01 -4.95875090e-01 -3.72550160e-01
-1.24276090e+00 -4.78704013e-02 9.57073510e-01 7.73216307e-01
4.19990569e-01 6.06333971e-01 8.92798826e-02 7.49883875e-02
2.15088248e-01 5.54503560e-01 5.75199187e-01 -8.42919528e-01
-2.67433394e-02 7.32590616e-01 -2.72388104e-02 -6.75831735e-01
-7.01040208e-01 -4.98179644e-01 -8.24160278e-01 5.66886246e-01
1.11825533e-01 -1.60464391e-01 -1.01344407e+00 1.55069423e+00
1.41206771e-01 -3.54511067e-02 5.94857000e-02 4.48409170e-01
2.61145294e-01 8.75284076e-01 3.39403629e-01 -4.42666739e-01
1.32888055e+00 -3.00944000e-01 -8.51283550e-01 -4.26657349e-01
4.25125629e-01 -2.88200304e-02 8.00927937e-01 3.66047442e-01
-1.18706942e+00 -4.96132374e-01 -1.93997133e+00 2.38118142e-01
-5.51048517e-01 4.85754572e-02 2.34128818e-01 1.04387152e+00
-1.24670112e+00 9.47565377e-01 -1.27320087e+00 -1.67029634e-01
5.76018309e-03 9.94562626e-01 3.50774884e-01 6.04241729e-01
-1.15093088e+00 1.20419466e+00 5.94673276e-01 7.19839782e-02
-6.36209548e-01 -7.75589526e-01 -4.33275104e-01 7.07043171e-01
6.95454031e-02 -6.55508816e-01 9.04796839e-01 -9.88719285e-01
-1.47994339e+00 3.82593870e-01 2.69781183e-02 -8.16083908e-01
-7.68640414e-02 2.31662124e-01 -6.62058413e-01 1.31526425e-01
-6.44281089e-01 3.78576070e-01 6.60393119e-01 -6.90306962e-01
-4.58750427e-01 -4.32468921e-01 -3.09658766e-01 -8.26992653e-03
-8.71912777e-01 -4.63290840e-01 1.21570468e-01 -2.98685171e-02
-3.23162973e-01 -9.32787061e-01 -1.78712040e-01 -2.49765649e-01
1.22968294e-01 -1.04305230e-01 9.78465199e-01 -2.53105432e-01
1.46579444e+00 -1.84595942e+00 -1.37721077e-01 5.78914464e-01
-1.32427096e-01 1.90228671e-01 4.10023600e-01 -2.48141754e-02
2.25605667e-01 -2.69627303e-01 1.17123704e-02 2.02974722e-01
-1.07521765e-01 3.33449125e-01 8.46955031e-02 4.09023136e-01
-2.58136988e-01 5.78770876e-01 -1.19646579e-01 -2.30504334e-01
2.13216811e-01 7.68730283e-01 -7.15086460e-02 -1.92333795e-02
-2.01657005e-02 -3.17254961e-01 -4.05283183e-01 4.63807166e-01
2.16926888e-01 -1.28131032e-01 6.26624167e-01 -5.39117634e-01
-3.20370644e-02 1.59552887e-01 -1.10608470e+00 1.23226380e+00
-5.13110697e-01 7.89572537e-01 1.00593865e-01 -8.74771655e-01
1.10576510e+00 2.34911859e-01 3.34581673e-01 -1.02699089e+00
6.71994984e-01 3.03468645e-01 2.49219969e-01 -1.51602358e-01
6.86823249e-01 1.08644798e-01 -5.49752489e-02 1.80337876e-01
-4.03871745e-01 9.79537740e-02 2.62227608e-03 -1.87377349e-01
1.25303566e+00 1.23144202e-01 2.69228399e-01 -1.02064216e+00
4.17262375e-01 2.07384124e-01 5.73924601e-01 3.24530303e-01
-2.19179645e-01 -2.95203000e-01 3.49280149e-01 -2.61220723e-01
-1.34166384e+00 -7.88156092e-01 -2.32012719e-01 9.15171862e-01
2.28661656e-01 -2.97675639e-01 -1.18030190e+00 -3.08662206e-02
-1.95195422e-01 1.22364187e+00 -5.19200683e-01 -6.56125128e-01
-5.81301928e-01 -8.78398657e-01 7.89442539e-01 1.07628274e+00
5.07761955e-01 -6.63186729e-01 -1.91140294e+00 2.24787042e-01
5.65572143e-01 -1.00728488e+00 3.78860347e-02 1.02672493e+00
-1.24503720e+00 -6.23675466e-01 1.80618167e-01 -3.43943805e-01
6.87531888e-01 -4.14045006e-01 1.15350640e+00 1.57235205e-01
-3.46506327e-01 7.39585906e-02 7.80236945e-02 -3.83221328e-01
-4.15721208e-01 3.25731009e-01 3.40516806e-01 -5.96762359e-01
6.94226980e-01 -9.51952457e-01 -6.58377349e-01 -1.61023736e-01
-6.31083488e-01 -9.23390985e-02 5.53495049e-01 4.86217320e-01
6.05798781e-01 4.52822626e-01 7.33357370e-01 -4.25584733e-01
4.23376799e-01 -1.90224215e-01 -9.21065509e-01 2.41747215e-01
-1.66336012e+00 3.32797080e-01 9.75037038e-01 -5.38637936e-01
-8.73685718e-01 3.64590704e-01 1.48309693e-01 -3.00930232e-01
1.24189034e-01 -6.22195331e-03 -3.15120220e-01 -3.04266125e-01
7.18376100e-01 1.70693383e-01 -1.63317546e-01 9.60698873e-02
-1.93135887e-01 6.75031006e-01 5.65105438e-01 -1.90164059e-01
1.10289529e-01 9.17307064e-02 5.82702994e-01 -6.30264342e-01
1.76742032e-01 4.08611447e-01 -4.74675030e-01 -4.41178173e-01
4.98545229e-01 -6.86539233e-01 -1.18545568e+00 1.26539066e-01
-5.42129755e-01 -3.01015168e-01 -3.88419479e-01 1.62925914e-01
-4.77665037e-01 -2.75768816e-01 -4.68147397e-01 -1.05600107e+00
-1.09590185e+00 -1.00613749e+00 4.84165758e-01 8.02128315e-01
-6.29285753e-01 -8.12739670e-01 -2.81622887e-01 -2.90230274e-01
8.25758874e-01 2.29671016e-01 1.05408823e+00 -4.15923208e-01
-3.80868495e-01 -1.30124271e-01 1.21679217e-01 1.16064534e-01
-2.49307007e-01 -5.77903949e-02 -1.12809551e+00 -4.02778655e-01
1.75030857e-01 1.73030794e-01 4.63394880e-01 5.46551943e-01
1.07820618e+00 3.43829207e-02 -8.33056211e-01 3.66970748e-01
1.73424387e+00 9.97758865e-01 6.61291480e-01 2.54396677e-01
1.94451451e-01 1.46331906e-01 3.04092854e-01 3.31778973e-01
-5.61496690e-02 6.24263346e-01 3.38141382e-01 3.98094863e-01
1.69709876e-01 6.14540130e-02 3.55205864e-01 5.34547985e-01
6.70597702e-02 -3.80013436e-01 -8.63600791e-01 5.24838030e-01
-1.56676221e+00 -5.67116797e-01 1.38312057e-01 2.34269261e+00
5.30237019e-01 7.07936525e-01 -7.23306760e-02 7.51483560e-01
5.00706673e-01 2.50981674e-02 -1.04593515e+00 -1.09407771e+00
2.32157961e-01 3.50540608e-01 1.01299989e+00 2.79015958e-01
-3.92571449e-01 3.46113741e-01 6.83530474e+00 8.06817949e-01
-1.24332929e+00 -1.05638020e-02 6.41381443e-01 -7.11593866e-01
-1.22539058e-01 -2.20864564e-01 -8.92936349e-01 6.83284342e-01
2.09914565e+00 -5.25599003e-01 4.34156060e-01 1.03471124e+00
1.18746251e-01 -6.57092571e-01 -1.34221458e+00 9.51205850e-01
4.03576083e-02 -1.19895709e+00 -5.25848091e-01 1.04070462e-01
1.26201436e-01 -2.91360140e-01 -1.02244921e-01 7.48227537e-02
-3.42818767e-01 -1.04096532e+00 6.12384915e-01 3.90407711e-01
1.03277683e+00 -1.39291751e+00 6.27112508e-01 3.27511162e-01
-1.23622108e+00 -3.27705413e-01 -3.15549523e-01 -2.96394587e-01
6.48146272e-02 6.84283376e-01 -6.07102394e-01 -1.10274062e-01
9.95745897e-01 -6.62852675e-02 -4.22780156e-01 4.08173203e-01
2.24203557e-01 5.79896748e-01 -8.32401931e-01 -8.61028552e-01
-2.49662742e-01 7.46010542e-02 2.23105609e-01 1.00430644e+00
5.41026175e-01 2.01800302e-01 -4.06215340e-01 7.29201615e-01
4.83207516e-02 -3.87315869e-01 -5.14405489e-01 -1.38345186e-03
1.03492379e+00 1.34489918e+00 -1.08428144e+00 -3.89875650e-01
1.63092643e-01 7.01320648e-01 2.41091684e-03 -8.59852955e-02
-8.97395134e-01 -6.45398676e-01 4.67709333e-01 4.75717664e-01
4.34081480e-02 -1.23327851e-01 -9.10678566e-01 -3.07010859e-01
-2.24644821e-02 -4.59125936e-01 1.80294603e-01 -6.84928060e-01
-4.07593578e-01 7.72605240e-01 2.15394482e-01 -6.60066962e-01
-5.05995333e-01 -6.05108619e-01 -5.33428073e-01 5.55045068e-01
-9.52915192e-01 -6.95331216e-01 -4.05125320e-01 1.60760611e-01
1.55744404e-01 -1.16137497e-01 8.79371822e-01 3.43680471e-01
-6.50340617e-01 7.99708664e-01 1.68029130e-01 -6.69756532e-01
-5.82736656e-02 -7.98860908e-01 3.00859809e-01 8.42090726e-01
-4.73719478e-01 2.26229250e-01 1.06717849e+00 -7.76372313e-01
-2.02972198e+00 -6.48174584e-01 5.28733611e-01 -1.38645545e-02
1.75590590e-01 -3.78292382e-01 -1.08438122e+00 4.49766219e-01
3.14380258e-01 -1.89300731e-01 3.83771002e-01 -1.69546932e-01
2.65177190e-01 -6.16554379e-01 -1.44976997e+00 4.67131168e-01
7.29480207e-01 -2.87786901e-01 -1.06791034e-01 -1.96683660e-01
1.64909586e-01 -3.15357774e-01 -9.99636412e-01 5.41078389e-01
7.77264178e-01 -7.66102076e-01 6.48783267e-01 -2.64487788e-02
-1.27969623e-01 -1.80167645e-01 -1.85676768e-01 -9.57250893e-01
-2.09804550e-01 -3.10492605e-01 -7.19530523e-01 1.05793703e+00
3.49530190e-01 -6.05616570e-01 1.22220516e+00 1.16310835e+00
1.13415666e-01 -1.04784632e+00 -1.41320348e+00 -7.39399552e-01
-1.37356266e-01 -2.51676261e-01 5.82166314e-01 2.85895228e-01
2.49793321e-01 4.50805753e-01 1.36784930e-02 8.07488337e-02
6.51298463e-01 -2.88534045e-01 -5.88475130e-02 -1.05485189e+00
-1.84353188e-01 -2.61773288e-01 -5.16316056e-01 -2.93367416e-01
8.55936185e-02 -4.69173104e-01 8.19711313e-02 -1.02034497e+00
1.11855090e-01 -4.18276310e-01 -5.38527131e-01 4.00184959e-01
2.29077280e-01 2.10410655e-01 -3.77598591e-02 -1.90634921e-01
-1.10394344e-01 9.74515080e-02 1.68141529e-01 6.45696297e-02
-4.09605980e-01 -3.82515341e-01 -2.57582933e-01 6.80876076e-01
7.86332607e-01 -5.57899654e-01 -8.30712259e-01 -2.00478122e-01
3.97007555e-01 1.61179990e-01 3.40586342e-02 -1.58865058e+00
7.43106782e-01 -1.60277113e-02 8.90369058e-01 -5.43605208e-01
5.92412174e-01 -1.31549740e+00 8.46999228e-01 1.04107940e+00
-1.36021152e-01 3.88021529e-01 5.69964051e-01 3.48384172e-01
5.11187971e-01 -3.56648147e-01 9.69837606e-01 4.96418804e-01
-7.97115326e-01 -3.30276191e-01 -8.50989223e-01 -6.15091920e-01
1.50923455e+00 -5.78663290e-01 -4.07568634e-01 -4.22223285e-02
-4.49107975e-01 5.48204035e-02 5.84768713e-01 1.08362168e-01
5.08670330e-01 -1.02171659e+00 -1.16282262e-01 2.86172211e-01
-3.14056277e-01 -4.67250526e-01 3.81285340e-01 6.04295671e-01
-6.35855138e-01 3.98846209e-01 -6.69889867e-01 -4.24516976e-01
-1.55268264e+00 7.63210416e-01 7.87682950e-01 -1.75258875e-01
-5.59098005e-01 2.79010922e-01 -9.24503684e-01 6.06805265e-01
1.52416810e-01 -2.01288417e-01 -1.58861414e-01 -1.23270914e-01
5.33864021e-01 8.37658763e-01 4.68232483e-01 -7.36003220e-02
-7.58770347e-01 4.88608152e-01 2.13348702e-01 -4.61221002e-02
1.12109315e+00 -5.56637682e-02 3.33226658e-02 4.74361807e-01
8.01054776e-01 -4.11054194e-01 -9.23572302e-01 3.40135932e-01
1.93480045e-01 -3.73310670e-02 7.01926649e-01 -6.71385705e-01
-1.03208840e+00 6.39581621e-01 1.15630972e+00 1.96608633e-01
1.79598713e+00 -6.17531478e-01 8.05294991e-01 5.43507278e-01
4.73927349e-01 -1.73232734e+00 -4.36244071e-01 -3.98993045e-02
1.54447183e-01 -3.62580776e-01 4.31343526e-01 -7.58112594e-02
-7.91003034e-02 1.18069935e+00 8.93643141e-01 -2.56024629e-01
6.17535472e-01 1.41534054e+00 -3.84838521e-01 8.73699710e-02
-1.07276380e+00 3.87099862e-01 -1.85597055e-02 4.09369111e-01
6.90847486e-02 4.00487818e-02 -4.33115691e-01 5.99023461e-01
-2.69105304e-02 4.21689987e-01 4.99272645e-01 1.09910989e+00
-6.66098893e-01 -7.32964337e-01 -3.19752604e-01 7.58237123e-01
-4.67187166e-01 2.61879236e-01 3.95267941e-02 7.10807920e-01
4.98982891e-02 8.68815958e-01 5.35796463e-01 -6.73649013e-01
3.59213948e-01 1.89217612e-01 5.88363528e-01 7.79847801e-02
-8.68970871e-01 -1.61450297e-01 3.33259255e-01 -4.97678667e-01
-4.77165915e-02 -1.78601414e-01 -1.76774526e+00 -5.36770463e-01
-3.25540572e-01 1.32904112e-01 1.11208832e+00 7.83512950e-01
5.73813796e-01 1.26329684e+00 1.98344126e-01 -7.33082771e-01
-4.11291152e-01 -7.27128088e-01 -7.29692042e-01 -3.71201009e-01
-1.02857567e-01 -5.51925361e-01 -2.97992557e-01 -1.59366146e-01] | [8.22741985321045, 2.5273778438568115] |
e1032f49-d2c8-49ec-be28-79d182972b45 | meta-regression-analysis-of-errors-in-short | 2305.18550 | null | https://arxiv.org/abs/2305.18550v1 | https://arxiv.org/pdf/2305.18550v1.pdf | Meta-Regression Analysis of Errors in Short-Term Electricity Load Forecasting | Forecasting electricity demand plays a critical role in ensuring reliable and cost-efficient operation of the electricity supply. With the global transition to distributed renewable energy sources and the electrification of heating and transportation, accurate load forecasts become even more important. While numerous empirical studies and a handful of review articles exist, there is surprisingly little quantitative analysis of the literature, most notably none that identifies the impact of factors on forecasting performance across the entirety of empirical studies. In this article, we therefore present a Meta-Regression Analysis (MRA) that examines factors that influence the accuracy of short-term electricity load forecasts. We use data from 421 forecast models published in 59 studies. While the grid level (esp. individual vs. aggregated vs. system), the forecast granularity, and the algorithms used seem to have a significant impact on the MAPE, bibliometric data, dataset sizes, and prediction horizon show no significant effect. We found the LSTM approach and a combination of neural networks with other approaches to be the best forecasting methods. The results help practitioners and researchers to make meaningful model choices. Yet, this paper calls for further MRA in the field of load forecasting to close the blind spots in research and practice of load forecasting. | ['Thorsten Staake', 'Hannah Hartstang', 'Konstantin Hopf'] | 2023-05-29 | null | null | null | null | ['load-forecasting'] | ['miscellaneous'] | [-3.35675955e-01 -3.38950723e-01 -7.43773162e-01 -1.99958026e-01
-1.79438785e-01 -5.92096627e-01 8.59992623e-01 5.77419363e-02
-9.93136093e-02 8.19291890e-01 3.95225346e-01 -9.90863502e-01
-4.91666555e-01 -9.81717229e-01 -1.82540402e-01 -7.80804574e-01
1.38360247e-01 4.05704767e-01 -5.17805398e-01 -2.34633684e-01
7.62880385e-01 4.42090452e-01 -1.37299526e+00 -2.04254210e-01
9.96428370e-01 9.66459274e-01 4.39060390e-01 -2.45984346e-02
-1.73097730e-01 7.46742964e-01 -6.59597814e-01 1.26854405e-02
8.33399221e-02 -4.77029473e-01 -5.26037812e-01 -5.46250582e-01
-3.95465791e-01 -3.61900926e-01 -4.73574502e-03 5.90791166e-01
9.17553067e-01 5.51382340e-02 6.91366315e-01 -1.26707816e+00
-7.34993041e-01 1.07081437e+00 -5.85040867e-01 6.65843844e-01
1.47891745e-01 1.32636741e-01 6.23257875e-01 -3.97583127e-01
2.38779888e-01 7.38234818e-01 6.64885879e-01 -2.52498657e-01
-1.14188838e+00 -9.08913910e-01 -9.60797945e-04 4.33089554e-01
-1.16709983e+00 -4.79925424e-01 6.55653298e-01 -6.55731678e-01
1.50672722e+00 2.18356386e-01 7.90327489e-01 6.41566396e-01
6.20879412e-01 -1.30484700e-01 1.37645531e+00 -5.85312307e-01
2.11091697e-01 3.48517269e-01 -1.12326354e-01 -2.35206351e-01
4.02789533e-01 2.49939084e-01 -1.06091283e-01 -9.68032554e-02
2.55647212e-01 -8.17905962e-02 -2.16865286e-01 4.67009395e-01
-1.03373110e+00 7.94787228e-01 1.77473709e-01 1.03647423e+00
-8.01323235e-01 -7.99522996e-02 6.28295481e-01 3.11448514e-01
7.29930282e-01 5.03006995e-01 -4.25535202e-01 -2.70934463e-01
-1.35284460e+00 -1.28567833e-02 9.56724107e-01 1.60964876e-01
2.84088612e-01 8.21706951e-01 -4.56586480e-02 5.57425737e-01
-4.18431051e-02 6.26376748e-01 8.19450438e-01 -7.13334441e-01
3.21938008e-01 3.23557377e-01 1.82699323e-01 -1.35815930e+00
-8.53092372e-01 -6.66212916e-01 -9.33197081e-01 -7.03497380e-02
1.88232183e-01 -5.12510121e-01 -3.16200376e-01 1.25919139e+00
-4.21612829e-01 -1.37825981e-01 -2.13559017e-01 6.46171689e-01
4.40429509e-01 9.95487809e-01 1.36415452e-01 -5.34158766e-01
1.03705180e+00 -5.52928984e-01 -8.28346789e-01 -2.37406995e-02
6.99114978e-01 -7.70934820e-01 4.52233762e-01 3.50386947e-01
-1.23556626e+00 -4.81126964e-01 -7.33739138e-01 3.91152829e-01
-8.46523464e-01 1.20012186e-01 4.58369404e-01 6.10445917e-01
-1.05687022e+00 6.31302953e-01 -5.34677207e-01 -4.71265703e-01
4.06906642e-02 2.05748498e-01 3.08405697e-01 3.54918331e-01
-1.35818744e+00 1.81136405e+00 5.55812895e-01 3.64588737e-01
-4.76037189e-02 -7.12659955e-01 -4.64591771e-01 3.21691543e-01
-1.14857189e-01 -5.47157168e-01 9.00007665e-01 -9.33982730e-01
-1.19548237e+00 3.47913033e-03 -1.64307073e-01 -7.98227191e-01
1.17304452e-01 5.11844575e-01 -7.41513610e-01 -4.48721707e-01
-6.64031431e-02 1.30227685e-01 1.93749920e-01 -8.14193070e-01
-9.09040809e-01 -3.17858011e-01 -3.46070766e-01 9.28690359e-02
-2.22149596e-01 1.29036456e-01 7.22652912e-01 -5.77880979e-01
-2.78854370e-01 -5.83875179e-01 -9.53164846e-02 -1.13527668e+00
-1.99452564e-02 -7.78075159e-01 5.89810073e-01 -1.12603879e+00
1.71126759e+00 -1.72242200e+00 -2.38136828e-01 3.52180034e-01
-2.47706592e-01 -7.31439814e-02 3.25383991e-01 1.00147939e+00
-3.82526189e-01 4.45673257e-01 2.00740188e-01 1.36124015e-01
4.12836760e-01 1.37238890e-01 -7.14378953e-01 5.60723066e-01
-2.63967186e-01 1.10558450e+00 -5.81120014e-01 6.16762862e-02
7.33289361e-01 5.06438434e-01 1.32552058e-01 -3.64105165e-01
2.11611196e-01 1.71247020e-01 -2.15136632e-01 2.34266669e-01
4.80734944e-01 -3.45687896e-01 1.50986671e-01 -1.09872147e-01
-8.09908628e-01 6.68891191e-01 -7.43613064e-01 1.01162410e+00
-7.63975024e-01 1.05142915e+00 -4.43700999e-01 -1.26146841e+00
9.17210639e-01 6.45977914e-01 6.04883611e-01 -1.27520800e+00
3.85673307e-02 5.68322599e-01 5.75347364e-01 -2.83686161e-01
4.94076431e-01 -1.56704918e-01 3.20922017e-01 6.12313807e-01
-4.08301353e-01 6.02350980e-02 4.29374307e-01 -3.06845665e-01
3.49341631e-01 -6.98523149e-02 1.81286961e-01 -7.46675551e-01
1.89616397e-01 -1.47497490e-01 4.67830718e-01 3.60613614e-01
2.86304295e-01 -9.10711065e-02 3.97558540e-01 -5.19451439e-01
-9.42904592e-01 -2.57655919e-01 -5.68257034e-01 8.88473451e-01
-5.68455100e-01 8.19812249e-03 -4.35825974e-01 1.41947612e-01
1.48365870e-01 1.57514369e+00 -4.76997465e-01 1.70041174e-01
-3.04642290e-01 -9.92522061e-01 1.58315033e-01 4.74441588e-01
3.70035529e-01 -1.01502740e+00 -9.26193416e-01 5.12054086e-01
-1.74737155e-01 -8.23893011e-01 6.19946644e-02 1.70684293e-01
-8.78513038e-01 -5.76873362e-01 -7.35715330e-01 -5.09787977e-01
4.08676475e-01 3.84877995e-02 1.29438460e+00 -7.41626024e-02
2.55464226e-01 7.52426907e-02 -3.75618115e-02 -6.99363291e-01
-1.79016843e-01 5.64223588e-01 2.54753605e-02 -6.41838431e-01
5.22846162e-01 -8.31400275e-01 -5.12103796e-01 1.32392853e-01
-5.45486629e-01 -1.13287549e-02 3.87833446e-01 4.53447163e-01
-1.31458119e-02 6.44359887e-01 1.43831420e+00 -4.50852126e-01
9.74751353e-01 -9.23798740e-01 -7.49299407e-01 1.55713066e-01
-1.49359202e+00 -3.84372234e-01 6.77643836e-01 5.53589724e-02
-8.26906383e-01 -6.58164918e-01 -4.07626890e-02 1.44994214e-01
-8.62442032e-02 1.15171528e+00 4.91020530e-01 7.70910606e-02
3.25907528e-01 3.97986293e-01 -1.44874170e-01 -4.47258204e-01
-2.35238299e-01 5.74030459e-01 4.95688543e-02 -1.01745173e-01
6.07557058e-01 -8.55186433e-02 -7.58247524e-02 -6.79045916e-01
-2.60072768e-01 1.98051594e-02 -2.39700884e-01 -3.59023154e-01
6.08220756e-01 -8.80411565e-01 -7.33742654e-01 1.68607473e-01
-7.81174362e-01 -4.92712289e-01 -1.29705384e-01 6.56900883e-01
-8.95676911e-02 -1.58914879e-01 -2.92283267e-01 -1.01877356e+00
-5.20601809e-01 -1.25008011e+00 2.06476718e-01 2.42796615e-01
-5.40577412e-01 -1.23656344e+00 8.28783736e-02 1.49706108e-02
1.26334178e+00 3.66134435e-01 1.11155736e+00 -5.52831948e-01
9.35741758e-04 -1.64247543e-01 -9.85898450e-02 1.09121263e-01
3.83511513e-01 1.63894281e-01 -8.64349127e-01 -2.36460268e-01
-7.06945583e-02 2.07698688e-01 4.40326124e-01 9.53419089e-01
1.18025613e+00 -6.52489305e-01 -2.96525598e-01 1.47910759e-01
1.59415746e+00 7.04398870e-01 6.56375527e-01 5.81579149e-01
1.65878072e-01 7.23087072e-01 1.11516126e-01 5.78293800e-01
8.00651371e-01 3.38491201e-01 6.01825817e-03 -4.31989378e-04
4.07846928e-01 8.64401609e-02 1.40764803e-01 1.11910248e+00
-4.17935312e-01 -2.04291716e-01 -1.21112633e+00 5.89285374e-01
-1.41481292e+00 -1.17245197e+00 -1.70220703e-01 2.11776781e+00
4.05479878e-01 2.50118345e-01 2.95857966e-01 3.64384979e-01
4.61979002e-01 2.22803652e-01 -3.65348220e-01 -8.84058237e-01
-2.50368178e-01 -2.06499249e-02 7.05457509e-01 2.89395392e-01
-3.23164850e-01 8.27827454e-02 6.53000450e+00 6.46215856e-01
-1.36742258e+00 -2.54759472e-02 1.12957335e+00 -1.08973488e-01
-6.35924399e-01 -6.26749471e-02 -4.63042319e-01 1.09873319e+00
1.60973406e+00 -7.55806744e-01 5.99276423e-01 3.84277850e-01
9.17089105e-01 -6.08322680e-01 -6.98122442e-01 6.10742629e-01
-2.69857258e-01 -1.24477160e+00 -3.84822220e-01 4.02818084e-01
9.15920138e-01 3.59696925e-01 1.16250468e-02 3.61335605e-01
-5.36779240e-02 -1.19390976e+00 7.02418506e-01 7.60443211e-01
4.12795037e-01 -8.92768502e-01 1.11084557e+00 4.72513735e-01
-9.25500274e-01 -3.05127144e-01 -2.07854494e-01 -6.70720994e-01
3.72190654e-01 8.51495385e-01 -2.73103625e-01 6.70556664e-01
7.85122693e-01 5.83728731e-01 -1.80069208e-01 6.42598510e-01
4.13002342e-01 1.09299541e+00 -3.27609032e-01 4.32000961e-05
2.26111546e-01 -4.78677213e-01 1.88356936e-02 9.70174193e-01
5.13851643e-01 7.12103993e-02 -3.04783136e-01 6.12772465e-01
1.47540137e-01 1.11740328e-01 -7.24743366e-01 -2.84097344e-01
7.86097884e-01 1.07544839e+00 -8.82692039e-01 -2.56982893e-01
-7.07700491e-01 1.98066637e-01 -3.92724812e-01 6.36850297e-01
-5.18114269e-01 -3.41646463e-01 3.16613436e-01 2.72460252e-01
4.13319394e-02 -1.21476971e-01 -8.46512258e-01 -4.77983177e-01
-1.58494309e-01 -7.19703555e-01 2.46884525e-01 -7.46732235e-01
-1.14348722e+00 1.95739836e-01 6.29330277e-02 -7.61146307e-01
-7.22843468e-01 -3.30714285e-02 -8.36352825e-01 1.53172052e+00
-1.75852644e+00 -5.36505699e-01 8.72508287e-02 -1.08465590e-01
4.33993101e-01 -2.19650969e-01 9.15679216e-01 2.52266616e-01
-6.55168891e-01 -4.27214801e-02 6.32637978e-01 -2.26571575e-01
3.72285098e-01 -1.09149992e+00 4.26328897e-01 4.27740455e-01
-4.62685615e-01 2.32795566e-01 6.56109035e-01 -6.59595072e-01
-1.13735843e+00 -4.57663268e-01 1.24525547e+00 -1.52614176e-01
5.39817512e-01 1.69587150e-01 -7.67475784e-01 6.60614491e-01
1.04015028e+00 -9.15511131e-01 6.18341863e-01 1.43706322e-01
2.88850844e-01 -3.67027491e-01 -1.10541630e+00 2.26457193e-01
2.18486741e-01 -2.36757234e-01 -2.64266521e-01 -3.29325497e-02
3.26631442e-02 4.96503301e-02 -1.38032138e+00 5.21086574e-01
6.77740753e-01 -8.99201810e-01 6.01717174e-01 9.46525410e-02
3.14479858e-01 1.47801667e-01 5.00331819e-02 -1.68730855e+00
-7.13577807e-01 -3.19962054e-01 -3.63271981e-02 1.31803823e+00
6.66693091e-01 -1.52246487e+00 2.87642032e-01 9.24196482e-01
-5.98673113e-02 -9.85642672e-01 -8.83513868e-01 -4.98812139e-01
5.00999868e-01 -2.77890116e-01 1.17367268e+00 1.35882044e+00
2.49957100e-01 -2.06265561e-02 -4.99400459e-02 -3.91648501e-01
3.17904532e-01 3.64733726e-01 5.23777306e-02 -1.26037240e+00
3.23771179e-01 -1.12813115e+00 -4.48412038e-02 -1.65043458e-01
1.54517248e-01 -7.96600938e-01 -5.75179815e-01 -2.22383046e+00
-7.14841187e-02 -3.79259586e-01 -3.65937591e-01 5.26293933e-01
3.05231452e-01 -8.44854489e-02 2.12207198e-01 2.98355907e-01
3.73924732e-01 4.05595750e-01 7.06372082e-01 6.06393069e-02
-1.80153117e-01 1.02470564e-02 -6.88441575e-01 5.70991933e-01
1.49556029e+00 -8.98011848e-02 -4.91523385e-01 -5.17905116e-01
6.17592871e-01 1.91965237e-01 -1.00974673e-02 -7.66791105e-01
2.91547745e-01 -3.29088509e-01 7.52237320e-01 -7.82081127e-01
-2.99152970e-01 -9.73457873e-01 6.88712776e-01 6.02834284e-01
-1.17938749e-01 8.77365887e-01 5.63469172e-01 -1.73248380e-01
-4.81428839e-02 -1.97758421e-01 1.72337025e-01 1.41757175e-01
-3.46466631e-01 -1.87161267e-01 -7.91511416e-01 -1.73545122e-01
7.83889651e-01 -2.83281863e-01 -3.93954188e-01 -6.28018796e-01
-2.46554449e-01 4.49987859e-01 3.24366629e-01 5.47523260e-01
-8.96483958e-02 -1.16710782e+00 -8.15019310e-01 -1.76781699e-01
-5.51399708e-01 -4.68500674e-01 2.70992875e-01 1.04798996e+00
-1.66590780e-01 1.20283318e+00 3.21628596e-03 -3.87557968e-02
-2.83061832e-01 1.94685444e-01 4.70722347e-01 -2.24427968e-01
-3.33948046e-01 -6.80467114e-02 -4.02457118e-01 -2.23618910e-01
1.93644285e-01 -4.72584158e-01 -4.01472270e-01 7.53281295e-01
3.35493773e-01 1.17171264e+00 3.88561368e-01 -6.06496453e-01
-1.68000385e-01 3.51466328e-01 4.97431785e-01 -2.51216744e-03
1.51509154e+00 -2.95833647e-01 -4.60970193e-01 1.03869152e+00
7.97569513e-01 -4.87981647e-01 -4.73360777e-01 2.48706520e-01
1.04709931e-01 -3.58395278e-02 5.44078767e-01 -1.28610802e+00
-1.21802914e+00 5.61966479e-01 6.47636533e-01 9.30227995e-01
1.34479082e+00 -5.63887000e-01 5.96833825e-01 -3.87326926e-02
2.96773255e-01 -1.40409398e+00 -9.55646873e-01 2.56731093e-01
9.27887857e-01 -1.01610970e+00 5.16121946e-02 4.95534867e-01
-3.59500825e-01 1.02596724e+00 6.01500925e-03 2.40110829e-01
8.91180217e-01 3.46608669e-01 -2.00782418e-01 1.02499910e-01
-7.78012156e-01 1.73873678e-01 2.33226880e-01 1.84333548e-01
7.72985339e-01 2.61900961e-01 -8.04895401e-01 7.08290815e-01
-5.63359082e-01 2.64032364e-01 4.00842577e-01 5.29190838e-01
-1.75052315e-01 -7.36331820e-01 -3.26792777e-01 1.18241572e+00
-6.87034726e-01 -1.48150444e-01 1.04319386e-01 8.50167513e-01
-7.29574338e-02 1.11503494e+00 5.68192303e-01 -1.20331228e-01
2.51697719e-01 3.70648146e-01 -4.36371304e-02 3.33888792e-02
-7.77074754e-01 -4.35961634e-02 8.73382464e-02 -9.42080989e-02
-4.83241409e-01 -9.03876781e-01 -9.54858124e-01 -1.00005698e+00
-5.50761998e-01 3.94255638e-01 1.04281306e+00 1.11307645e+00
4.01075274e-01 5.42314470e-01 9.18744028e-01 -9.09770131e-01
-4.63192165e-01 -1.43217909e+00 -7.51914084e-01 -4.41465288e-01
9.58896056e-02 -3.84500742e-01 -5.19648790e-01 -4.20868486e-01] | [6.119110107421875, 2.8624751567840576] |
4e21cda5-08a7-43ed-b003-a74b6e310eff | distanced-lstm-time-distanced-gates-in-long | 1909.05321 | null | https://arxiv.org/abs/1909.05321v1 | https://arxiv.org/pdf/1909.05321v1.pdf | Distanced LSTM: Time-Distanced Gates in Long Short-Term Memory Models for Lung Cancer Detection | The field of lung nodule detection and cancer prediction has been rapidly developing with the support of large public data archives. Previous studies have largely focused on cross-sectional (single) CT data. Herein, we consider longitudinal data. The Long Short-Term Memory (LSTM) model addresses learning with regularly spaced time points (i.e., equal temporal intervals). However, clinical imaging follows patient needs with often heterogeneous, irregular acquisitions. To model both regular and irregular longitudinal samples, we generalize the LSTM model with the Distanced LSTM (DLSTM) for temporally varied acquisitions. The DLSTM includes a Temporal Emphasis Model (TEM) that enables learning across regularly and irregularly sampled intervals. Briefly, (1) the time intervals between longitudinal scans are modeled explicitly, (2) temporally adjustable forget and input gates are introduced for irregular temporal sampling; and (3) the latest longitudinal scan has an additional emphasis term. We evaluate the DLSTM framework in three datasets including simulated data, 1794 National Lung Screening Trial (NLST) scans, and 1420 clinically acquired data with heterogeneous and irregular temporal accession. The experiments on the first two datasets demonstrate that our method achieves competitive performance on both simulated and regularly sampled datasets (e.g. improve LSTM from 0.6785 to 0.7085 on F1 score in NLST). In external validation of clinically and irregularly acquired data, the benchmarks achieved 0.8350 (CNN feature) and 0.8380 (LSTM) on the area under the ROC curve (AUC) score, while the proposed DLSTM achieves 0.8905. | ['Kim L. Sandler', 'Yuankai Huo', 'Sanja L. Antic', 'Steve Deppen', 'Riqiang Gao', 'Alexis B. Paulson', 'Yucheng Tang', 'Pierre P. Massion', 'Emily S. Epstein', 'Shunxing Bao', 'Bennett A. Landman', 'Aneri B. Balar'] | 2019-09-11 | null | null | null | null | ['lung-nodule-detection'] | ['medical'] | [ 2.87851214e-01 -2.59180933e-01 -7.01807678e-01 -2.49278575e-01
-9.57495034e-01 -1.89090565e-01 5.52760184e-01 8.07446465e-02
-6.92414284e-01 7.37459660e-01 2.37531021e-01 -4.96239305e-01
-4.02000368e-01 -7.36927688e-01 -7.19061255e-01 -8.01395714e-01
-4.24923331e-01 3.95259470e-01 6.17023826e-01 2.84903318e-01
-3.29568505e-01 2.80385643e-01 -8.27044129e-01 5.40216088e-01
7.28991032e-01 1.12896645e+00 3.96207094e-01 7.11663067e-01
3.43217887e-03 9.76516843e-01 -1.35880128e-01 1.13337785e-02
3.37910466e-02 -2.37588242e-01 -7.10779965e-01 2.09680703e-02
1.91513464e-01 -3.66594017e-01 -6.95966840e-01 6.28517687e-01
5.75982571e-01 -4.47891973e-04 3.70884806e-01 -6.80770934e-01
-5.30628502e-01 5.65867126e-01 -7.01341808e-01 6.26548171e-01
-2.95854449e-01 3.92605186e-01 5.30129015e-01 -7.01863647e-01
3.85418087e-01 7.13046253e-01 1.05575657e+00 4.42516476e-01
-9.23277020e-01 -4.18274969e-01 8.39907825e-02 -1.75108705e-02
-1.08153582e+00 1.79308727e-01 7.18794018e-02 -4.78462934e-01
8.94370437e-01 1.50627539e-01 7.70667732e-01 1.39092696e+00
9.79734063e-01 6.69240415e-01 9.14852202e-01 -6.31021038e-02
-1.45870581e-01 -2.33728051e-01 2.38257617e-01 8.43270123e-01
9.08028856e-02 5.59044063e-01 -1.67368695e-01 -2.27668673e-01
1.06927955e+00 6.59054279e-01 -8.35165158e-02 9.32307988e-02
-1.65895927e+00 7.00498879e-01 8.96887958e-01 6.59680486e-01
-6.46888673e-01 2.98447579e-01 5.73227882e-01 2.97102720e-01
3.01565289e-01 -4.52016555e-02 -3.68939728e-01 1.30132049e-01
-9.98621583e-01 -3.07816207e-01 3.26642931e-01 7.98504889e-01
5.77851050e-02 7.07588866e-02 -8.77579033e-01 8.22391808e-01
-3.47351544e-02 4.87495661e-01 1.18812144e+00 -4.15497094e-01
3.14212859e-01 2.66767710e-01 -5.02437539e-02 -4.50371504e-01
-6.58086777e-01 -7.51421750e-01 -1.38162470e+00 -5.32432079e-01
2.53747612e-01 -1.12276688e-01 -1.40031326e+00 1.69043422e+00
1.93476025e-02 4.96067435e-01 -8.75040442e-02 5.49254715e-01
7.74085522e-01 4.17706341e-01 4.54278529e-01 -4.60727900e-01
1.48214626e+00 -9.06600118e-01 -8.20426166e-01 -9.63541716e-02
8.26930523e-01 -3.40031564e-01 1.01656294e+00 -2.98770051e-02
-9.72783208e-01 -5.24718285e-01 -7.93905914e-01 1.33531809e-01
-1.30757377e-01 2.99566001e-01 3.60209137e-01 1.93761751e-01
-1.00810051e+00 6.83697939e-01 -1.49274981e+00 -3.88601989e-01
3.80252451e-01 4.01566029e-01 -5.89752682e-02 -3.09152991e-01
-1.38941085e+00 6.48774624e-01 2.28129148e-01 2.62308359e-01
-1.15120959e+00 -1.11198997e+00 -5.20151436e-01 -1.93027720e-01
3.75032634e-01 -9.11853552e-01 1.66293657e+00 -6.73837602e-01
-9.80175734e-01 6.29092991e-01 -1.87950537e-01 -9.36539531e-01
8.05271149e-01 3.99060063e-02 -5.30335665e-01 -2.59297192e-02
1.80427462e-01 6.03407383e-01 6.24238729e-01 -7.47078001e-01
-5.99240243e-01 -3.01061481e-01 -3.76847118e-01 -3.51450930e-04
-4.24144447e-01 -3.25711191e-01 -4.83505666e-01 -7.68555224e-01
1.98155060e-01 -1.10902286e+00 -4.27813441e-01 1.68064117e-01
-3.22437376e-01 -2.89478414e-02 7.54237890e-01 -7.84681976e-01
1.42167377e+00 -2.07369757e+00 -3.54700089e-01 -1.01056822e-01
4.42884862e-01 1.50065929e-01 9.43716690e-02 -1.45519730e-02
-1.89224780e-01 1.56529546e-01 -3.20818931e-01 -4.44677919e-02
-5.44789910e-01 3.72839183e-01 1.41716376e-01 2.98096120e-01
1.07013084e-01 1.39457047e+00 -9.85934973e-01 -6.18800819e-01
2.23920960e-02 4.69251096e-01 -1.08503871e-01 -1.02641419e-01
-7.11051598e-02 5.06674469e-01 -6.99816346e-01 6.40323877e-01
1.95428103e-01 -7.96339154e-01 -6.77876407e-03 -1.84065595e-01
-2.44096860e-01 1.22790389e-01 -3.61852676e-01 1.72088885e+00
-5.60223579e-01 4.67461616e-01 -5.04938245e-01 -4.06291991e-01
4.90138799e-01 7.86873460e-01 8.90794992e-01 -1.02140641e+00
-3.40387784e-03 4.27379340e-01 3.93990070e-01 -5.92884064e-01
1.19433716e-01 -4.25198466e-01 1.60413608e-01 3.56939465e-01
-2.81290919e-01 3.99328977e-01 -7.69099221e-02 -2.25625280e-02
1.48394179e+00 -4.04730201e-01 3.19886893e-01 -1.34511441e-01
2.48802140e-01 5.86225465e-02 5.21661222e-01 8.32832396e-01
-3.97800177e-01 4.54673767e-01 3.26006174e-01 -6.31561339e-01
-8.28764439e-01 -1.18908083e+00 -3.29895169e-01 7.37917960e-01
-3.71156484e-01 2.81621695e-01 -1.37225106e-01 -7.66867936e-01
4.30090241e-02 4.98598069e-01 -9.57604170e-01 -3.44962299e-01
-8.01901817e-01 -9.35424805e-01 6.34748101e-01 1.01003528e+00
6.36412561e-01 -1.18406475e+00 -7.61225343e-01 4.18228328e-01
-2.20626906e-01 -9.41757083e-01 -8.03473055e-01 3.61889362e-01
-1.36778581e+00 -9.01311874e-01 -9.48236942e-01 -8.31483722e-01
4.58966017e-01 2.55405635e-01 1.18611217e+00 -1.05583683e-01
-3.58755946e-01 3.38744849e-01 1.68502405e-02 -2.16411218e-01
-1.92754656e-01 1.48928091e-01 -2.84496367e-01 -2.93223530e-01
1.13592353e-02 -3.56617928e-01 -6.08940899e-01 3.48886162e-01
-9.84860301e-01 1.21095948e-01 1.05500507e+00 1.30725718e+00
1.11696541e+00 -1.62079886e-01 6.69526577e-01 -1.04033637e+00
2.51065850e-01 -7.87182093e-01 -3.56533408e-01 3.60418171e-01
-6.69439614e-01 4.74069873e-03 4.07870233e-01 -7.94330418e-01
-9.15300608e-01 -1.60704806e-01 1.52779981e-01 -6.71000838e-01
2.36800209e-01 8.56463015e-01 6.00031734e-01 1.04526237e-01
5.55721283e-01 5.63795030e-01 -4.82678153e-02 -1.45767197e-01
-2.94731051e-01 2.32283801e-01 3.56307417e-01 -3.89966220e-01
3.95455956e-01 5.00273824e-01 3.65045741e-02 -5.15831947e-01
-9.99421835e-01 -1.60688609e-01 -5.37496626e-01 -1.41827151e-01
8.64363968e-01 -8.12189281e-01 -4.24436927e-01 3.88054669e-01
-6.43002450e-01 -6.55039430e-01 -6.50722086e-01 8.53237808e-01
-4.29639757e-01 -1.02698833e-01 -9.78898108e-01 -4.31281626e-01
-5.27280509e-01 -1.14436960e+00 9.05807674e-01 3.73187400e-02
-4.52789105e-02 -1.20469952e+00 -8.09130147e-02 -1.84025764e-01
6.63457274e-01 5.37142813e-01 1.10660446e+00 -5.32376170e-01
-5.20161331e-01 -4.34864074e-01 -3.02536666e-01 7.39319324e-02
4.33420837e-01 -2.65659451e-01 -6.67926371e-01 -4.01947767e-01
1.97065979e-01 -1.96597189e-01 1.10061884e+00 1.03655851e+00
1.57859087e+00 -3.38515006e-02 -5.93968749e-01 5.41876137e-01
1.27896690e+00 6.36217177e-01 5.06605446e-01 1.51926875e-01
6.06659591e-01 1.58689424e-01 5.58296084e-01 2.54113942e-01
2.19983056e-01 1.64839983e-01 2.84766227e-01 -3.24284792e-01
-1.73587948e-01 -2.77179062e-01 1.84947446e-01 7.87614107e-01
3.59062441e-02 -1.50725797e-01 -1.17327690e+00 7.12794006e-01
-1.73449421e+00 -8.58083546e-01 -3.03499341e-01 2.39273190e+00
8.34232330e-01 3.64996135e-01 -1.59952849e-01 -2.40182564e-01
6.71019197e-01 3.02548140e-01 -9.50555503e-01 1.74000576e-01
7.85403326e-02 2.23843575e-01 8.10181856e-01 2.18665615e-01
-1.11909306e+00 4.46464866e-01 5.74717903e+00 8.20405245e-01
-1.34222996e+00 4.61189538e-01 9.71100390e-01 -3.64264548e-01
3.51303928e-02 -4.10130233e-01 -6.27207577e-01 4.90777463e-01
1.22909653e+00 -2.69340396e-01 -1.29979357e-01 3.20434600e-01
5.08886337e-01 -1.65531337e-02 -1.13957107e+00 5.37015736e-01
-4.09522951e-01 -1.17302871e+00 6.79163113e-02 1.20518647e-01
7.87423909e-01 5.50741136e-01 4.99676019e-01 5.86776733e-01
1.98896095e-01 -1.28518963e+00 1.77812591e-01 8.04693937e-01
1.21274698e+00 -3.93610656e-01 8.90537560e-01 4.85800564e-01
-1.35919487e+00 -5.83206415e-02 -7.34809935e-02 5.36865115e-01
-8.54968280e-03 6.94877148e-01 -9.96478856e-01 7.20803022e-01
7.16409683e-01 1.01233339e+00 -5.96828222e-01 1.00911152e+00
2.81459659e-01 7.46469975e-01 -2.74506927e-01 7.98496678e-02
6.10939980e-01 1.96000740e-01 2.40637422e-01 1.03721213e+00
6.38007939e-01 1.10046573e-01 2.76752621e-01 5.77922404e-01
-4.14963365e-02 -2.84298092e-01 -4.75327134e-01 -8.03921670e-02
3.44748318e-01 9.44531143e-01 -5.22071004e-01 -3.72038692e-01
-5.79493523e-01 5.00408351e-01 -1.10450290e-01 3.29214007e-01
-1.21802020e+00 2.66418248e-01 -1.25404820e-01 4.20250922e-01
2.28450492e-01 6.09518997e-02 -2.05399185e-01 -7.31806517e-01
-2.74722129e-02 -5.45569777e-01 9.01041627e-01 -4.71103817e-01
-1.42085898e+00 7.91784108e-01 -1.29647896e-01 -1.50031579e+00
-2.16183230e-01 -2.81553090e-01 -6.19074345e-01 7.48441577e-01
-1.60073423e+00 -1.12167549e+00 -4.71595347e-01 6.03084207e-01
6.31767690e-01 8.05501267e-02 6.45469844e-01 4.25670266e-01
-5.26580930e-01 5.13340890e-01 2.35330313e-01 2.22758099e-01
6.56674087e-01 -1.06575465e+00 5.09951375e-02 4.24544394e-01
-3.47486198e-01 4.78806913e-01 9.94728878e-02 -8.68112564e-01
-1.18848073e+00 -1.77283084e+00 7.20780611e-01 -2.15240389e-01
7.10687041e-01 3.75066221e-01 -1.10814810e+00 1.04058146e+00
-1.34646788e-01 5.78724861e-01 5.11700690e-01 -3.75563741e-01
3.41848074e-03 -6.41452894e-02 -1.21623731e+00 2.21273780e-01
8.18569243e-01 -3.18531483e-01 -1.45510614e-01 4.86584663e-01
9.01805460e-01 -6.69632494e-01 -1.26591766e+00 9.34549332e-01
6.92509353e-01 -5.45036852e-01 8.82244945e-01 -5.20562768e-01
3.33606333e-01 1.34465899e-02 1.59096718e-02 -8.70720625e-01
-5.16733170e-01 -3.68733704e-02 -2.32582793e-01 3.24141383e-01
5.48493207e-01 -7.24088073e-01 7.27086842e-01 3.14391613e-01
-3.72996897e-01 -1.38504672e+00 -1.06444073e+00 -8.74370754e-01
1.62112445e-01 -3.07940662e-01 1.27273262e-01 8.31764936e-01
-6.93000734e-01 -2.16734767e-01 -1.52607247e-01 8.75583887e-02
3.26259404e-01 1.94593798e-02 -1.89239785e-01 -9.41320062e-01
-3.56163681e-01 -3.86732042e-01 3.34063806e-02 -8.74411941e-01
-3.09875011e-01 -9.77660954e-01 -6.05281815e-02 -1.64700615e+00
6.70582771e-01 -3.38289440e-01 -8.50917399e-01 5.39814413e-01
-2.80102640e-01 -1.50283992e-01 -3.27427387e-01 2.84994572e-01
-3.84122521e-01 5.27923882e-01 1.63189018e+00 -2.26603508e-01
-1.98551536e-01 4.61100399e-01 -2.10367799e-01 4.48348284e-01
7.89004266e-01 -5.97357929e-01 -3.65328640e-01 -5.98476052e-01
-3.94576281e-01 7.03575075e-01 6.12647474e-01 -1.10304201e+00
3.47654104e-01 -4.02087606e-02 6.67988241e-01 -9.14932370e-01
1.06903881e-01 -7.40480363e-01 2.91334033e-01 1.29186261e+00
-5.08609295e-01 3.03498566e-01 3.98486465e-01 9.38859046e-01
-2.63934612e-01 1.95094869e-01 9.31419134e-01 -2.68231899e-01
-3.68753165e-01 9.68016624e-01 -2.71994740e-01 7.46859051e-03
1.00972033e+00 -1.37317762e-01 -1.37410015e-01 -1.17608450e-01
-1.04520023e+00 5.19336045e-01 -1.51594788e-01 3.12356263e-01
6.69804871e-01 -1.50486434e+00 -6.57266557e-01 5.52066565e-02
-1.51351979e-02 2.10832521e-01 5.72081387e-01 1.48731720e+00
-1.56411618e-01 6.34142935e-01 2.26957396e-01 -1.06804156e+00
-1.06421709e+00 4.40030485e-01 7.29572713e-01 -9.86186385e-01
-7.11405456e-01 7.66239762e-01 3.49054486e-01 -2.28021219e-01
2.20732048e-01 -8.44990015e-01 6.63640276e-02 -9.44821164e-02
2.08295688e-01 1.68677256e-01 2.80025095e-01 -1.28608495e-01
-1.79311171e-01 2.46218204e-01 -4.09996092e-01 6.58131465e-02
1.26844180e+00 2.65437543e-01 1.04124829e-01 9.34741318e-01
1.19208312e+00 -5.72678268e-01 -1.16436648e+00 -7.36108720e-01
2.43944526e-01 -5.25261685e-02 2.12546438e-01 -8.67485940e-01
-1.09791374e+00 8.25579286e-01 1.02035105e+00 9.39328969e-02
1.18531382e+00 -1.36547282e-01 1.16229773e+00 1.23427033e-01
1.92780539e-01 -4.94578511e-01 2.76511014e-01 7.07977057e-01
6.74987614e-01 -1.22392428e+00 -2.18738019e-01 4.80176089e-03
-6.78905010e-01 1.03118706e+00 6.33161306e-01 2.85831019e-02
7.43016362e-01 3.56613427e-01 -2.33410418e-01 -2.58905858e-01
-1.23702216e+00 2.46416882e-01 3.32263976e-01 6.00795634e-02
6.99603379e-01 2.22596765e-01 -1.57646298e-01 6.33261859e-01
2.56824821e-01 3.29545885e-01 1.84591323e-01 1.07827401e+00
-3.50532264e-01 -6.49886310e-01 -1.47140861e-01 1.10690224e+00
-5.17251313e-01 -1.94207296e-01 1.41015917e-01 1.14979970e+00
-1.15310261e-02 3.39943916e-01 1.23068824e-01 -1.80592969e-01
3.51998091e-01 5.69531247e-02 2.76806235e-01 -6.59145355e-01
-7.19136715e-01 3.02181989e-01 -3.45102966e-01 -5.20546675e-01
-2.79209822e-01 -6.94310188e-01 -1.51600349e+00 -9.61503312e-02
-2.26838753e-01 -1.57741129e-01 1.80337772e-01 8.13331008e-01
1.91422805e-01 1.30607891e+00 5.74714661e-01 -3.38697761e-01
-8.08279872e-01 -1.17299652e+00 -4.98961806e-01 -6.30901456e-02
6.97668791e-01 -4.78987157e-01 -1.89562172e-01 -2.14630570e-02] | [15.21770191192627, -2.0536999702453613] |
5c489e2e-3c58-4ffd-8525-261ca8ba55e7 | on-the-importance-of-attention-in-meta | 1806.00852 | null | http://arxiv.org/abs/1806.00852v1 | http://arxiv.org/pdf/1806.00852v1.pdf | On the Importance of Attention in Meta-Learning for Few-Shot Text Classification | Current deep learning based text classification methods are limited by their
ability to achieve fast learning and generalization when the data is scarce. We
address this problem by integrating a meta-learning procedure that uses the
knowledge learned across many tasks as an inductive bias towards better natural
language understanding. Based on the Model-Agnostic Meta-Learning framework
(MAML), we introduce the Attentive Task-Agnostic Meta-Learning (ATAML)
algorithm for text classification. The essential difference between MAML and
ATAML is in the separation of task-agnostic representation learning and
task-specific attentive adaptation. The proposed ATAML is designed to encourage
task-agnostic representation learning by way of task-agnostic parameterization
and facilitate task-specific adaptation via attention mechanisms. We provide
evidence to show that the attention mechanism in ATAML has a synergistic effect
on learning performance. In comparisons with models trained from random
initialization, pretrained models and meta trained MAML, our proposed ATAML
method generalizes better on single-label and multi-label classification tasks
in miniRCV1 and miniReuters-21578 datasets. | ['Hassan Chouaib', 'Mohammad Havaei', 'Stan Matwin', 'Nicolas Chapados', 'Xiang Jiang', 'Andrew Jesson', 'Thomas Vincent', 'Gabriel Chartrand'] | 2018-06-03 | null | null | null | null | ['few-shot-text-classification'] | ['natural-language-processing'] | [ 2.90623724e-01 1.92190677e-01 -4.19979870e-01 -7.02441573e-01
-5.83397031e-01 -2.96619266e-01 8.83212924e-01 4.88540858e-01
-6.62809253e-01 7.03056753e-01 4.19771999e-01 -2.18643427e-01
-3.04343611e-01 -5.49255908e-01 -4.59002763e-01 -5.67175806e-01
3.85576934e-01 6.62998319e-01 -2.31271058e-01 -3.16529691e-01
4.40488994e-01 1.02705799e-01 -1.33535731e+00 9.21509743e-01
1.10909534e+00 8.00444186e-01 4.68349010e-01 7.56352603e-01
-5.73388457e-01 1.04068637e+00 -5.13355136e-01 -3.97017926e-01
-2.29672879e-01 -3.50420952e-01 -1.48776376e+00 -6.58659190e-02
4.26647007e-01 1.50002047e-01 1.47975730e-02 7.23596036e-01
4.53936636e-01 4.58610684e-01 1.08614790e+00 -1.22124422e+00
-9.46698546e-01 1.04492271e+00 -6.00255430e-01 3.26723725e-01
-2.03809962e-01 -1.03715405e-01 8.40097308e-01 -9.63876843e-01
1.92906007e-01 1.25209892e+00 6.53728366e-01 1.09274411e+00
-1.04544413e+00 -5.15922844e-01 6.12562180e-01 1.83758289e-01
-1.01486480e+00 -3.19302022e-01 8.02532196e-01 -3.93506467e-01
1.10763013e+00 3.63039114e-02 3.42408493e-02 1.35037100e+00
4.38774168e-01 1.05803704e+00 1.28741968e+00 -8.13269317e-01
2.59016752e-01 4.67006296e-01 6.14612877e-01 4.41488951e-01
1.02569446e-01 -2.20974788e-01 -4.76118416e-01 -1.56888083e-01
1.26702681e-01 2.59873569e-01 -4.82585989e-02 -1.11751184e-01
-1.09346318e+00 8.04939687e-01 6.26226723e-01 5.16993761e-01
-3.93673599e-01 1.66867420e-01 8.58572066e-01 4.73759860e-01
8.73365402e-01 6.63827062e-01 -9.22038376e-01 1.06678151e-01
-7.98071861e-01 -2.46175691e-01 4.86759931e-01 8.20214748e-01
7.46261239e-01 3.27852041e-01 -3.65146577e-01 1.31810665e+00
4.49521691e-01 1.94381043e-01 1.41075337e+00 -4.92787629e-01
6.48611665e-01 9.78109896e-01 -3.64226133e-01 -4.12389338e-01
-7.39436269e-01 -8.30145955e-01 -9.63105321e-01 5.42603210e-02
1.03334980e-02 -2.72228807e-01 -1.06270444e+00 1.90224254e+00
-1.00061074e-02 -1.63370982e-01 5.13125479e-01 3.16790789e-01
1.00243068e+00 7.53010213e-01 8.83771062e-01 -2.87965149e-01
1.34907472e+00 -1.31496298e+00 -6.54101729e-01 -6.32507980e-01
1.33594584e+00 -3.76701027e-01 1.23216534e+00 1.18927017e-01
-8.33216786e-01 -7.85789192e-01 -1.00434721e+00 -7.25677684e-02
-7.86797941e-01 4.69395518e-02 5.91272235e-01 6.35929763e-01
-8.93598974e-01 2.72369891e-01 -3.52164358e-01 -6.94659233e-01
6.20558202e-01 3.33673298e-01 -1.78789377e-01 7.41483718e-02
-1.02061713e+00 1.14135253e+00 9.64933813e-01 -1.89226612e-01
-8.62034380e-01 -7.44226456e-01 -6.55557930e-01 2.68031418e-01
8.57992694e-02 -8.25921595e-01 1.52096736e+00 -1.64888155e+00
-1.52197838e+00 1.08232713e+00 4.33786446e-03 -5.44859052e-01
2.21296549e-01 -1.62618652e-01 -3.35627288e-01 -1.19391374e-01
-1.71249911e-01 8.52670014e-01 8.70143592e-01 -1.50680125e+00
-7.45600700e-01 -4.35215771e-01 -9.58120972e-02 5.76558590e-01
-8.73364866e-01 -3.28456730e-01 3.19021910e-01 -8.91693771e-01
-1.18585959e-01 -7.41472125e-01 -1.19948864e-01 -6.10618472e-01
-1.10127524e-01 -5.68267763e-01 1.02524054e+00 -1.49895221e-01
1.13406980e+00 -1.69148421e+00 1.97444990e-01 -2.93448418e-01
2.27018192e-01 5.65770090e-01 -4.43478256e-01 5.44470906e-01
-2.13123351e-01 8.97316113e-02 -5.22397421e-02 -5.13688684e-01
-1.69426709e-01 1.66993544e-01 -3.18286717e-01 2.46501658e-02
2.47582030e-02 1.11295474e+00 -1.00531411e+00 -4.67530459e-01
7.84846693e-02 8.50824639e-02 -3.83374631e-01 2.64121741e-01
-6.21041715e-01 3.28620046e-01 -5.66612840e-01 4.30300236e-01
3.69678169e-01 -5.16341031e-01 2.29555577e-01 -2.29998618e-01
9.80062410e-02 2.72548258e-01 -5.02070248e-01 1.91213381e+00
-1.06942248e+00 5.80683410e-01 -2.25841969e-01 -1.39919007e+00
8.13516319e-01 6.23576164e-01 1.83051661e-01 -7.22780168e-01
4.79507238e-01 5.10894395e-02 1.77660689e-01 -4.04194146e-01
5.46328187e-01 -2.35992506e-01 2.26022936e-02 1.09789920e+00
5.80566406e-01 2.51331717e-01 -1.09615244e-01 3.00309241e-01
6.77210450e-01 5.00698760e-02 5.32357872e-01 -7.09412277e-01
6.96404576e-01 4.87507023e-02 8.07508081e-02 9.51619446e-01
-7.36979395e-02 8.85837674e-02 -2.56437212e-01 -6.42569959e-01
-8.06185126e-01 -4.88304764e-01 -2.10184708e-01 2.13089371e+00
-2.52678245e-01 -3.09598595e-01 -5.35086811e-01 -1.19911098e+00
-1.57603398e-01 1.09271312e+00 -1.15302718e+00 -7.24127650e-01
-3.09483051e-01 -1.16899562e+00 1.90025002e-01 7.82336354e-01
8.00433934e-01 -1.31606400e+00 -2.90460497e-01 3.58342856e-01
4.61420901e-02 -6.93596482e-01 -3.53216439e-01 6.81280971e-01
-1.32968688e+00 -7.07290947e-01 -6.50852621e-01 -9.70777154e-01
8.68045270e-01 4.05984074e-01 1.27713311e+00 1.35285184e-01
7.55674690e-02 7.24972069e-01 -6.85721278e-01 -6.97299004e-01
-8.20427775e-01 6.36138618e-01 3.28661911e-02 1.46544144e-01
4.93795931e-01 -4.35963094e-01 -3.32175881e-01 1.99444398e-01
-8.45753372e-01 3.85993809e-01 6.73516333e-01 1.09476459e+00
-1.44518539e-02 -1.47944182e-01 1.21740830e+00 -1.39847946e+00
9.20733869e-01 -8.31677318e-01 5.39334826e-02 4.73359674e-01
-1.13095164e+00 3.78780842e-01 8.16465855e-01 -7.62941480e-01
-1.49188447e+00 -1.06270723e-01 -4.12591211e-02 1.96602140e-02
-1.89836457e-01 6.45346224e-01 4.34615370e-03 4.50762967e-03
1.05968392e+00 4.59807664e-01 -3.21938097e-01 -5.11907518e-01
4.84545708e-01 9.73527491e-01 1.79646492e-01 -9.37709153e-01
2.99860626e-01 2.41093010e-01 -3.60594451e-01 -5.55622399e-01
-1.47238088e+00 -3.91531438e-01 -8.81790340e-01 -8.02502632e-02
8.08208883e-01 -7.79990315e-01 -5.29766500e-01 4.64653909e-01
-1.00389183e+00 -7.38240421e-01 -3.76015216e-01 3.91475648e-01
-5.99521220e-01 -1.42100424e-01 -4.10836130e-01 -8.02535355e-01
-8.16914499e-01 -7.61275232e-01 6.87988281e-01 2.13816002e-01
-2.99272925e-01 -1.72953391e+00 1.70950323e-01 2.95527995e-01
6.88183606e-01 -3.61594260e-01 1.38175070e+00 -1.34511268e+00
1.99714452e-01 -8.07522908e-02 -1.44835070e-01 2.95071602e-01
2.67051697e-01 -4.70874846e-01 -1.47590768e+00 -4.68383014e-01
2.43079122e-02 -8.54141176e-01 1.10183334e+00 4.26164687e-01
1.27451324e+00 -5.86530626e-01 -3.76017243e-01 4.43365961e-01
1.34677887e+00 2.01473400e-01 1.84444055e-01 7.16757894e-01
7.60475338e-01 5.62806726e-01 3.03729802e-01 1.78472057e-01
5.34151852e-01 5.22516072e-01 2.65778840e-01 -1.19515896e-01
-1.82088032e-01 -7.83789679e-02 1.64131239e-01 8.49768817e-01
8.03049654e-02 -4.92668390e-01 -1.06335628e+00 4.57768738e-01
-2.01154089e+00 -7.90715098e-01 1.26377404e-01 1.97950256e+00
1.01337028e+00 -3.26622394e-03 7.50358030e-02 9.34349671e-02
5.44291914e-01 7.39598647e-02 -6.38990939e-01 -8.68026137e-01
-3.35543789e-02 3.95125989e-03 1.43465877e-01 4.31185097e-01
-1.17122972e+00 1.05068433e+00 6.15047741e+00 6.86016142e-01
-1.18521631e+00 5.94866216e-01 7.84944117e-01 1.14894390e-01
-2.44379044e-02 -3.57001573e-01 -8.37639511e-01 1.82526812e-01
1.34186363e+00 -3.14546913e-01 -9.69797000e-02 9.12665486e-01
-2.38652125e-01 2.91453540e-01 -1.24004006e+00 7.30880320e-01
3.16050708e-01 -1.29753280e+00 6.36743963e-01 -1.99890926e-01
9.43262160e-01 -1.20341242e-03 2.62907922e-01 9.14717376e-01
5.65969765e-01 -9.53956604e-01 4.43061411e-01 4.07108754e-01
2.65892297e-01 -6.76517487e-01 6.92604899e-01 5.43130577e-01
-7.58580804e-01 -5.58455527e-01 -5.12665212e-01 -5.45322634e-02
-4.40595299e-01 7.23378509e-02 -9.39230502e-01 4.36381817e-01
4.50311869e-01 7.02644050e-01 -8.26311171e-01 6.85600579e-01
-4.98037226e-02 7.59193957e-01 4.70364630e-01 -9.76585671e-02
3.50544244e-01 4.10481542e-01 1.75056502e-01 1.54478180e+00
-1.86940908e-01 6.69730082e-02 2.44560674e-01 5.14619648e-01
-4.00191575e-01 4.72232163e-01 -3.84645045e-01 -2.08685342e-02
3.24383408e-01 1.28641212e+00 -4.88828093e-01 -7.13009953e-01
-4.04628217e-01 9.28349674e-01 6.94024622e-01 3.79935592e-01
-2.66563743e-01 -5.35912849e-02 8.89165029e-02 -1.86708122e-01
-7.03843981e-02 1.59627989e-01 -5.99338889e-01 -1.09659827e+00
-6.74208343e-01 -9.35970306e-01 8.18849444e-01 -8.37170660e-01
-1.57470405e+00 8.12347293e-01 9.66374427e-02 -9.48220670e-01
-3.84786159e-01 -6.99425817e-01 -7.46152580e-01 8.18681717e-01
-1.50347006e+00 -1.59183419e+00 -4.42163080e-01 4.76815611e-01
1.19456351e+00 -6.70524538e-01 1.13029885e+00 -6.74130023e-03
-6.42699420e-01 8.87420952e-01 3.50457400e-01 1.87051538e-02
8.08917880e-01 -1.27625954e+00 8.99660774e-03 3.76581848e-01
4.28906158e-02 6.45476699e-01 4.63398457e-01 -4.29964751e-01
-1.08733857e+00 -1.44433653e+00 6.13390267e-01 -6.83227479e-01
3.96568239e-01 -3.62732261e-01 -1.21252656e+00 9.67327237e-01
4.15408313e-01 -2.48649016e-01 9.42278862e-01 4.85524803e-01
-6.23130262e-01 -1.06521122e-01 -1.16353738e+00 4.42043304e-01
7.05456316e-01 -3.17624152e-01 -8.91999841e-01 6.46237493e-01
5.92409790e-01 1.02519326e-01 -7.41969109e-01 3.52626324e-01
3.55770260e-01 -4.36872691e-01 7.27677643e-01 -1.23629820e+00
5.03108084e-01 3.57150525e-01 -1.29528075e-01 -1.61315608e+00
-6.57218933e-01 -8.75500292e-02 -2.40146443e-01 1.40293527e+00
5.42119384e-01 -7.48643816e-01 5.88343263e-01 2.66858339e-01
-4.58308905e-01 -5.95952988e-01 -6.79703593e-01 -6.72657549e-01
7.74726450e-01 -2.63083458e-01 4.32713091e-01 1.51091635e+00
2.69071668e-01 8.01970661e-01 -2.77242124e-01 -1.63129225e-01
4.72708285e-01 4.46540192e-02 5.97244442e-01 -1.51063848e+00
-1.18659258e-01 -6.29952967e-01 -2.22223960e-02 -7.27022648e-01
6.05227530e-01 -1.31875968e+00 -1.12642884e-01 -1.47746551e+00
6.80079043e-01 -5.81797481e-01 -7.27620780e-01 7.92822719e-01
-4.47842747e-01 7.53031299e-02 7.38387788e-03 4.16099042e-01
-8.48876834e-01 6.26846671e-01 1.10684228e+00 -4.06241089e-01
-1.99591354e-01 2.39728056e-02 -9.61215675e-01 7.84636199e-01
9.37755287e-01 -7.30774403e-01 -6.22511744e-01 -6.26502335e-01
7.31876716e-02 -3.63572627e-01 8.72302148e-03 -8.35439861e-01
2.30691463e-01 -2.12647781e-01 5.79403341e-01 -1.92731753e-01
1.36117905e-01 -7.36338913e-01 -4.55418169e-01 5.73135257e-01
-1.08377802e+00 -5.84923625e-02 4.65071142e-01 5.25736153e-01
1.09153695e-01 -6.92497075e-01 9.49021995e-01 -3.68288845e-01
-7.84860909e-01 2.35440582e-01 -5.93725562e-01 1.64134458e-01
6.93349719e-01 -1.24293543e-01 -7.76552618e-01 -1.79879814e-01
-7.73855805e-01 1.64280683e-01 1.79651231e-01 8.33709717e-01
3.45596164e-01 -1.29717362e+00 -9.40621972e-01 -6.52369410e-02
5.55990517e-01 -4.52353895e-01 4.22884464e-01 5.08340061e-01
1.17843427e-01 6.54404759e-01 -2.79179811e-01 -5.42733371e-01
-9.89936411e-01 6.96368337e-01 6.20160758e-01 -4.52068388e-01
-3.91575903e-01 9.40838754e-01 5.59970737e-01 -7.14196503e-01
3.08148891e-01 -1.14507623e-01 -5.06527603e-01 3.16605158e-02
6.65193260e-01 2.12004453e-01 2.21999958e-01 -4.77571696e-01
-5.21542318e-02 3.52158964e-01 -6.82005048e-01 1.00849614e-01
1.23378479e+00 -2.02782586e-01 2.27137342e-01 8.52287829e-01
1.12177348e+00 -4.81572568e-01 -9.40687358e-01 -5.67431331e-01
2.52222061e-01 1.11173071e-01 3.32540303e-01 -1.21209061e+00
-8.64185631e-01 1.10331082e+00 8.41590345e-01 -1.07253399e-02
9.52685356e-01 -3.32305506e-02 3.01246643e-01 7.62890220e-01
-3.14190127e-02 -1.13359058e+00 5.92027545e-01 7.08423018e-01
8.36543560e-01 -1.66019166e+00 3.92981432e-02 2.36100346e-01
-6.69141948e-01 1.14944077e+00 9.92416859e-01 2.29550049e-01
5.34724295e-01 -2.32609123e-01 2.13774651e-01 -2.07200393e-01
-1.19757974e+00 8.52517337e-02 5.67294836e-01 6.92807138e-01
8.91220391e-01 -2.11979225e-01 -1.38458535e-01 8.19071293e-01
1.41668066e-01 -2.11680248e-01 3.00687373e-01 1.16859138e+00
-7.78875351e-01 -1.11731589e+00 -1.60448745e-01 5.65720677e-01
-3.47191244e-01 -2.50111014e-01 -3.99612159e-01 6.90657139e-01
-1.06176816e-01 9.00052369e-01 8.68827775e-02 -2.24957541e-01
-1.27114460e-03 7.19276547e-01 5.19731700e-01 -1.06727123e+00
-1.04444659e+00 -4.94184256e-01 -1.17646977e-01 1.97528720e-01
-5.22073686e-01 -1.35413751e-01 -1.08961380e+00 2.86716819e-02
-4.42873538e-01 3.64176810e-01 8.02031398e-01 1.32957816e+00
3.98765355e-01 7.62472928e-01 5.87195098e-01 -6.72067106e-01
-5.41281700e-01 -1.45647788e+00 -2.42049128e-01 4.41248238e-01
1.73673049e-01 -5.90847671e-01 -2.46870264e-01 1.14819884e-01] | [10.761021614074707, 7.6909942626953125] |
63936196-bc55-4159-9fab-8f2b424c5280 | x-recosa-multi-scale-context-aggregation-for | 2303.07833 | null | https://arxiv.org/abs/2303.07833v1 | https://arxiv.org/pdf/2303.07833v1.pdf | X-ReCoSa: Multi-Scale Context Aggregation For Multi-Turn Dialogue Generation | In multi-turn dialogue generation, responses are not only related to the topic and background of the context but also related to words and phrases in the sentences of the context. However, currently widely used hierarchical dialog models solely rely on context representations from the utterance-level encoder, ignoring the sentence representations output by the word-level encoder. This inevitably results in a loss of information while decoding and generating. In this paper, we propose a new dialog model X-ReCoSa to tackle this problem which aggregates multi-scale context information for hierarchical dialog models. Specifically, we divide the generation decoder into upper and lower parts, namely the intention part and the generation part. Firstly, the intention part takes context representations as input to generate the intention of the response. Then the generation part generates words depending on sentence representations. Therefore, the hierarchical information has been fused into response generation. we conduct experiments on the English dataset DailyDialog. Experimental results exhibit that our method outperforms baseline models on both automatic metric-based and human-based evaluations. | ['Danqin Wu'] | 2023-03-14 | null | null | null | null | ['dialogue-generation', 'response-generation', 'dialogue-generation'] | ['natural-language-processing', 'natural-language-processing', 'speech'] | [ 1.51390135e-01 4.57448244e-01 9.26604569e-02 -6.55618668e-01
-8.83501649e-01 -2.66117036e-01 7.75524259e-01 1.23093374e-01
-2.88507372e-01 1.04703259e+00 9.67576563e-01 -9.20358002e-02
6.85185432e-01 -8.24265540e-01 -1.02378495e-01 -5.50427318e-01
6.67592406e-01 4.70058262e-01 2.46558815e-01 -6.30283177e-01
1.47646651e-01 -4.55025226e-01 -8.92688513e-01 6.90710843e-01
8.91797841e-01 7.68947482e-01 5.79919934e-01 6.29926383e-01
-5.07259965e-01 1.19033003e+00 -8.93481791e-01 -2.74696112e-01
-3.36888641e-01 -1.04579651e+00 -9.44006681e-01 2.61828721e-01
-3.02363068e-01 -7.10897624e-01 -1.80904180e-01 8.36960733e-01
5.87651908e-01 2.24158183e-01 5.37440121e-01 -7.58029401e-01
-5.22337317e-01 1.12822676e+00 -1.72534734e-01 -1.73121929e-01
6.46279633e-01 1.86660022e-01 1.02887094e+00 -9.11249101e-01
5.62045813e-01 1.69470716e+00 1.26984924e-01 9.22267735e-01
-9.52907205e-01 -3.23262215e-01 4.50111121e-01 5.00559993e-02
-1.05808580e+00 -5.55174887e-01 7.90323853e-01 -3.53695482e-01
9.90438819e-01 -6.60405541e-03 2.59142339e-01 1.23866034e+00
3.26067150e-01 8.76445591e-01 9.23861086e-01 -3.39128733e-01
1.68488443e-01 2.32316911e-01 3.79684657e-01 3.73046815e-01
-5.50353944e-01 -1.35949373e-01 -3.70566964e-01 -7.52678588e-02
5.38050592e-01 -3.22311282e-01 -2.98500896e-01 4.09554154e-01
-1.19053257e+00 1.16540778e+00 3.31661731e-01 2.60794997e-01
-5.94299853e-01 -1.42753094e-01 3.99147093e-01 2.10727662e-01
4.50923115e-01 6.07358404e-02 -1.26307815e-01 -1.64503679e-01
-5.45508981e-01 4.42633629e-01 1.10092950e+00 1.11288249e+00
6.87370837e-01 -5.43922298e-02 -7.47130930e-01 1.09702981e+00
6.11535430e-01 2.57329822e-01 6.99590266e-01 -6.04915202e-01
9.62536216e-01 7.04512358e-01 2.98951536e-01 -9.14860725e-01
-3.45768601e-01 -2.13860441e-02 -7.75915623e-01 -5.32263994e-01
2.05968112e-01 -8.36285949e-01 -5.83038568e-01 2.01281118e+00
4.59285140e-01 -2.53526211e-01 4.68548834e-01 1.11066687e+00
1.25117016e+00 1.11667216e+00 3.06458235e-01 -4.33257192e-01
1.63092017e+00 -1.15630198e+00 -1.10794270e+00 -4.41923201e-01
5.52341461e-01 -9.01067317e-01 9.39326227e-01 -1.06073067e-01
-1.22078204e+00 -6.46777868e-01 -9.99836981e-01 -3.50945562e-01
-1.63989905e-02 3.26139659e-01 9.28579941e-02 1.32465959e-01
-9.02635813e-01 -3.47288549e-02 -3.07709217e-01 -1.56755388e-01
-3.64797413e-01 1.29945660e-02 6.50522336e-02 1.88226074e-01
-1.85428655e+00 8.80559027e-01 5.51943958e-01 2.70765483e-01
-7.68421352e-01 4.05854918e-02 -9.02021110e-01 1.86156258e-01
3.12271595e-01 -6.50188208e-01 1.75554705e+00 -6.96494102e-01
-1.90097034e+00 3.64657730e-01 -3.70611012e-01 -3.84167582e-01
3.95264715e-01 -1.90472022e-01 -1.57203168e-01 7.54859895e-02
1.61352679e-01 8.25523913e-01 5.58872461e-01 -1.11406493e+00
-8.02711546e-01 -1.65303677e-01 3.53971064e-01 7.37271369e-01
-4.37066704e-02 -4.75517288e-02 -3.78709972e-01 -5.46103895e-01
-1.59707740e-01 -8.99512589e-01 -3.09460819e-01 -8.66196096e-01
-6.16403282e-01 -5.78980088e-01 5.94197154e-01 -9.61337030e-01
1.40321696e+00 -1.98603463e+00 4.05298501e-01 -3.80103707e-01
-2.75689811e-02 1.16720432e-02 -1.65525638e-02 8.02162826e-01
3.17220122e-01 -1.24017671e-01 -1.18420072e-01 -5.21993518e-01
8.64018872e-02 5.92048429e-02 -6.25883460e-01 -2.45096758e-01
2.21517175e-01 7.22367644e-01 -8.45264614e-01 -6.33047104e-01
1.76231205e-01 2.53863186e-01 -4.88144040e-01 8.75167251e-01
-6.77029610e-01 7.15604663e-01 -6.94997251e-01 9.08396393e-02
2.80265152e-01 -8.56432989e-02 3.77973288e-01 1.02672152e-01
-1.23278528e-01 9.37363267e-01 -6.94447637e-01 1.76246333e+00
-7.48393536e-01 1.07796580e-01 2.04910100e-01 -3.81818295e-01
1.08253372e+00 6.96374416e-01 -1.60671264e-01 -4.81955677e-01
2.87945896e-01 2.12185308e-02 1.92663297e-01 -4.75266457e-01
8.39065611e-01 -3.80470544e-01 -6.04001045e-01 6.75339401e-01
6.12297095e-02 -2.74485260e-01 1.06406480e-01 5.12211263e-01
6.35137379e-01 4.26310934e-02 3.78323972e-01 -5.71189150e-02
9.09541607e-01 -2.30135784e-01 4.30035412e-01 3.83547217e-01
2.65625212e-02 5.94536662e-01 7.85787761e-01 9.05928761e-02
-6.44926786e-01 -7.04554498e-01 2.94369847e-01 1.18377674e+00
3.45341146e-01 -5.29674590e-01 -1.13583958e+00 -6.80430591e-01
-4.50916737e-01 1.13091481e+00 -4.74877208e-01 -2.30198964e-01
-6.92897022e-01 -5.74303925e-01 2.58976132e-01 5.17723143e-01
7.04563081e-01 -1.44705057e+00 -1.79331332e-01 7.07429469e-01
-7.57629812e-01 -1.22243881e+00 -7.86102712e-01 -2.12544873e-01
-6.10971570e-01 -6.13322079e-01 -5.73019803e-01 -7.60669053e-01
3.42895836e-01 5.54125570e-02 9.09805000e-01 7.33657181e-02
4.19202298e-01 -1.46142736e-01 -5.44118762e-01 -1.26522645e-01
-7.80990720e-01 2.98196226e-01 -3.61078948e-01 1.50720507e-01
3.94835860e-01 -2.83958465e-01 -4.44208413e-01 3.00463676e-01
-6.56749368e-01 4.59548354e-01 6.25517607e-01 9.20224249e-01
6.52115718e-02 -5.75536728e-01 1.10053670e+00 -8.77250075e-01
1.27788591e+00 -6.94739938e-01 -3.15093696e-01 2.69915462e-01
-9.96438488e-02 1.06357038e-01 8.04820359e-01 -1.17357776e-01
-1.58418679e+00 -1.62025299e-02 -4.33296323e-01 2.82917291e-01
-1.81749508e-01 5.51988184e-01 -5.69617510e-01 8.55669022e-01
3.98492843e-01 4.84711111e-01 -1.38931409e-01 -3.99628758e-01
4.46650535e-01 1.20436847e+00 2.57683724e-01 -6.64517939e-01
1.52236789e-01 -2.09544733e-01 -6.11140192e-01 -8.71401608e-01
-8.52928877e-01 -1.69896722e-01 -3.23557734e-01 -3.43075663e-01
1.29837811e+00 -9.62353885e-01 -3.62103939e-01 3.73587608e-01
-1.65842032e+00 -3.07944536e-01 1.11642465e-01 3.76575768e-01
-4.92681146e-01 3.27473015e-01 -1.03886878e+00 -9.50489104e-01
-4.82583165e-01 -1.45888543e+00 1.26499009e+00 4.49938446e-01
-4.11575675e-01 -9.55072820e-01 8.25939700e-02 4.57581818e-01
2.95343190e-01 -1.44555390e-01 1.08877635e+00 -8.20062459e-01
-4.70961094e-01 1.76606831e-04 -2.02651501e-01 1.95536211e-01
2.58307904e-01 -4.67600763e-01 -9.35783625e-01 1.60249889e-01
3.91798884e-01 -6.09673679e-01 5.49539387e-01 -5.57278991e-02
6.25190198e-01 -5.69058239e-01 -4.14872132e-02 -1.20852537e-01
9.28513467e-01 3.70034814e-01 5.90175629e-01 -3.20816547e-01
5.16412020e-01 1.03239250e+00 7.90700257e-01 3.35446119e-01
9.17005956e-01 7.92560160e-01 1.12824656e-01 2.43633866e-01
-8.94602612e-02 -5.60748279e-01 7.10269511e-01 1.27783287e+00
3.11999559e-01 -5.70472956e-01 -6.34645641e-01 5.03691554e-01
-2.12003660e+00 -7.32460320e-01 -3.04679990e-01 1.82725120e+00
1.38074768e+00 3.53656597e-02 -1.84072573e-02 -4.39842999e-01
9.29001510e-01 3.63306731e-01 -2.77673542e-01 -5.49465120e-01
1.19532250e-01 -2.72983998e-01 -4.23108667e-01 1.03978312e+00
-7.50386357e-01 1.18282509e+00 5.11156893e+00 6.73088372e-01
-8.23280215e-01 1.41631037e-01 6.65487051e-01 1.32945821e-01
-3.67286175e-01 1.74948350e-01 -1.15554464e+00 6.14659131e-01
1.03457344e+00 -3.30738813e-01 1.42041147e-01 6.71620190e-01
4.23133850e-01 -1.24138862e-01 -1.25301385e+00 5.11830807e-01
1.15941186e-02 -9.41066980e-01 1.56639859e-01 -1.24353938e-01
4.54088092e-01 -5.19270360e-01 -3.12618613e-01 7.30255008e-01
3.54390621e-01 -7.36613870e-01 6.38562083e-01 4.33017224e-01
3.94859523e-01 -6.37531161e-01 8.86311531e-01 7.07382202e-01
-1.05330896e+00 1.65376693e-01 -4.05741215e-01 -2.04130232e-01
6.83767140e-01 3.91539335e-01 -1.21312809e+00 6.32561982e-01
-1.94325641e-01 3.71003091e-01 -1.46119133e-01 2.96983391e-01
-7.25118160e-01 5.14916778e-01 1.25197530e-01 -3.89618665e-01
4.54004258e-01 -3.25409740e-01 4.20548081e-01 1.20050836e+00
7.86862597e-02 4.57513332e-01 5.07107437e-01 7.80535400e-01
-1.96783751e-01 4.89716172e-01 -4.51989919e-01 8.77383724e-02
4.79424894e-01 1.24246562e+00 -3.24454665e-01 -7.01405048e-01
-4.81323451e-01 8.93957198e-01 3.83408695e-01 3.74428451e-01
-7.79661000e-01 -4.15318608e-01 2.67691404e-01 -1.71896666e-01
-1.74219534e-02 -1.27469242e-01 -1.54735640e-01 -1.12701428e+00
-6.30186424e-02 -8.69749784e-01 1.96633533e-01 -6.79837286e-01
-1.13645172e+00 9.17412162e-01 6.91596270e-02 -9.21239197e-01
-9.69192624e-01 -1.33168399e-01 -7.84673572e-01 1.29506648e+00
-1.25830424e+00 -9.24636424e-01 -3.16354841e-01 3.03570122e-01
1.20388234e+00 -5.41955158e-02 9.44894016e-01 7.56748840e-02
-6.74406528e-01 3.44483823e-01 -4.22689795e-01 2.94011354e-01
8.66306186e-01 -1.21850240e+00 3.32147121e-01 5.50329685e-01
-4.38956320e-01 7.49371350e-01 6.51518166e-01 -8.60632598e-01
-7.92104065e-01 -1.01112771e+00 1.35011661e+00 -1.29753321e-01
5.86435556e-01 -6.54371083e-01 -9.11609113e-01 6.10978425e-01
8.29774261e-01 -9.29484844e-01 7.81904042e-01 -1.05243698e-01
1.11465998e-01 1.91688508e-01 -9.51920450e-01 8.81065309e-01
6.15820110e-01 -5.15187800e-01 -1.15231216e+00 3.26505333e-01
1.19254267e+00 -5.54729223e-01 -6.77096069e-01 2.86693107e-02
3.14760178e-01 -8.32710862e-01 4.25663918e-01 -4.71748173e-01
9.22107756e-01 -5.71304932e-02 -1.89726517e-01 -1.37941301e+00
-1.98876509e-03 -5.33912718e-01 -8.73570666e-02 1.55234480e+00
6.04732633e-01 -3.49497318e-01 3.31884116e-01 7.35596240e-01
-3.95155072e-01 -6.24651372e-01 -7.44881332e-01 -1.18284151e-01
1.96818784e-01 5.07885292e-02 7.58075595e-01 5.50143838e-01
5.33716857e-01 1.39158821e+00 -5.15071750e-01 -2.87373394e-01
1.80176690e-01 4.45920050e-01 8.26981723e-01 -7.71348238e-01
-2.85910547e-01 -1.72349095e-01 3.22748989e-01 -1.76103222e+00
2.95762986e-01 -6.10774934e-01 6.87101662e-01 -1.65573788e+00
7.89636225e-02 -2.05496415e-01 3.43485266e-01 1.69114247e-01
-7.28803933e-01 -4.21390355e-01 2.60921150e-01 1.71759963e-01
-4.90421087e-01 1.01261652e+00 1.26053727e+00 5.19640781e-02
-3.88262182e-01 9.77828950e-02 -7.62825429e-01 5.98249495e-01
8.88241231e-01 -2.49733493e-01 -7.05451310e-01 -5.63919306e-01
-2.83443660e-01 8.99823070e-01 -1.58231482e-02 -7.12077022e-01
2.72397429e-01 -3.10312986e-01 1.23267565e-02 -6.50384784e-01
5.61214805e-01 -4.27036881e-01 -3.75585169e-01 2.88629591e-01
-9.37273264e-01 4.84122224e-02 -9.47575420e-02 4.76395875e-01
-3.73805344e-01 -4.57968175e-01 5.97176611e-01 -3.53912413e-01
-2.87363917e-01 2.67845038e-02 -6.41977966e-01 1.07273899e-01
7.64968455e-01 3.26340824e-01 -5.10454774e-01 -9.54137623e-01
-7.85730600e-01 6.18364811e-01 1.70152679e-01 6.28889501e-01
6.16404235e-01 -1.33377326e+00 -9.32535112e-01 -2.15394665e-02
-1.92320365e-02 1.75808206e-01 4.03489113e-01 6.76025510e-01
-1.05444245e-01 6.42893791e-01 1.67054683e-01 -3.99334997e-01
-1.03161418e+00 2.75378764e-01 1.49328917e-01 -6.07322156e-01
-3.18244159e-01 7.14160204e-01 6.88964546e-01 -4.03771609e-01
1.62530005e-01 -2.86398530e-01 -5.88554204e-01 2.86573470e-01
7.57101774e-01 -1.82980284e-01 -1.51422054e-01 -8.46639097e-01
2.63633998e-03 -1.37027232e-02 -3.62920344e-01 -6.86955988e-01
7.72569180e-01 -4.18447852e-01 -1.76584929e-01 6.75069809e-01
9.27201033e-01 -4.43509407e-02 -1.11052907e+00 -2.54813969e-01
-2.77678072e-02 -6.89775795e-02 -3.44204426e-01 -8.71265292e-01
-5.59801400e-01 1.14423847e+00 -7.78679028e-02 4.49268520e-01
8.26147139e-01 -1.13546558e-01 1.18944108e+00 3.47337604e-01
3.49808335e-01 -1.19389141e+00 2.16201887e-01 9.29354131e-01
1.25270367e+00 -9.78401423e-01 -3.70979041e-01 -4.17350560e-01
-1.13094735e+00 9.44116831e-01 1.12213671e+00 5.34848459e-02
2.67262518e-01 -5.68239950e-02 2.82982290e-01 8.78276303e-02
-1.22031164e+00 -2.21069176e-02 -1.12487450e-01 3.34684432e-01
8.51625323e-01 -2.50712130e-02 -7.50799656e-01 1.12029397e+00
-4.64763522e-01 -2.26213813e-01 5.67497492e-01 6.74645007e-01
-7.89497972e-01 -1.36134970e+00 -1.86321288e-01 9.36362073e-02
-3.40405226e-01 -3.01660657e-01 -8.45704377e-01 3.69898468e-01
-2.60401815e-01 1.48787427e+00 -1.49781287e-01 -5.21554053e-01
3.56676221e-01 5.01353204e-01 1.24178696e-02 -1.13660574e+00
-9.29085970e-01 3.33035946e-01 6.56675041e-01 -3.24384153e-01
-1.63419187e-01 -5.73702395e-01 -1.55383062e+00 6.34065196e-02
-3.85857522e-01 6.05939507e-01 4.37851578e-01 1.13835895e+00
2.36240238e-01 6.20226860e-01 8.82589698e-01 -6.24385893e-01
-8.56678665e-01 -1.63147926e+00 -2.00104237e-01 2.53497541e-01
1.61928639e-01 -3.47634852e-01 -1.38100609e-01 -4.79078665e-02] | [12.58167839050293, 8.189515113830566] |
fd358a98-4271-40f5-ac37-c0ef42674316 | incrementality-bidding-and-attribution | 2208.12809 | null | https://arxiv.org/abs/2208.12809v1 | https://arxiv.org/pdf/2208.12809v1.pdf | Incrementality Bidding and Attribution | The causal effect of showing an ad to a potential customer versus not, commonly referred to as "incrementality", is the fundamental question of advertising effectiveness. In digital advertising three major puzzle pieces are central to rigorously quantifying advertising incrementality: ad buying/bidding/pricing, attribution, and experimentation. Building on the foundations of machine learning and causal econometrics, we propose a methodology that unifies these three concepts into a computationally viable model of both bidding and attribution which spans the randomization, training, cross validation, scoring, and conversion attribution of advertising's causal effects. Implementation of this approach is likely to secure a significant improvement in the return on investment of advertising. | ['Jeffrey Wong', 'Randall Lewis'] | 2022-08-25 | null | null | null | null | ['econometrics'] | ['miscellaneous'] | [ 8.71185511e-02 2.74796665e-01 -1.02851129e+00 -6.97345316e-01
-7.81360924e-01 -6.22679055e-01 8.64625633e-01 1.69098407e-01
-2.85701960e-01 4.62256730e-01 6.93582475e-01 -1.13673878e+00
-4.41242546e-01 -6.37389481e-01 -8.94364953e-01 3.32825892e-02
-2.09404349e-01 3.45691949e-01 2.82761622e-02 4.35870588e-02
5.14224112e-01 2.85138398e-01 -1.33703721e+00 1.77170023e-01
3.11749160e-01 1.21122241e+00 -4.17277813e-01 2.64370620e-01
-9.44925696e-02 1.19095230e+00 -3.93914551e-01 -1.09354544e+00
4.10991400e-01 -2.97694523e-02 -2.77024746e-01 1.18452892e-01
5.08140743e-01 -8.16217363e-01 -2.22825259e-01 8.20252955e-01
1.70692787e-01 -6.31148934e-01 4.09828633e-01 -1.18701553e+00
-1.06893945e+00 1.04036188e+00 -5.69047928e-01 4.99433845e-01
2.90189803e-01 6.24675192e-02 1.62342644e+00 -2.97585279e-01
4.66614723e-01 1.33652997e+00 4.36823994e-01 -1.19230978e-01
-1.77342713e+00 -9.88950372e-01 2.25919396e-01 -1.80202931e-01
-8.46855879e-01 -4.40420151e-01 7.10108876e-01 -7.30800033e-01
5.35230003e-02 4.18875635e-01 7.41638660e-01 1.00349450e+00
3.30565929e-01 6.39493108e-01 1.74371660e+00 -4.98935401e-01
3.14895988e-01 5.01613677e-01 4.93740052e-01 6.48551285e-02
4.94622678e-01 9.74312365e-01 -3.19905698e-01 -5.97678185e-01
6.62460566e-01 -1.89605743e-01 3.41672271e-01 -2.35657722e-01
-3.90109986e-01 1.55421376e+00 3.63517731e-01 -4.36846048e-01
-6.33377790e-01 4.62050736e-01 -5.77373579e-02 4.46515471e-01
3.31151694e-01 4.64541674e-01 -5.19536316e-01 4.54158448e-02
-6.85609937e-01 6.10938907e-01 6.82110667e-01 5.64897239e-01
3.63848895e-01 -8.87067169e-02 5.11065573e-02 4.08360422e-01
5.80897391e-01 3.17819744e-01 2.21256852e-01 -1.03653884e+00
7.90165272e-03 4.85804141e-01 5.12459576e-01 -9.46215510e-01
1.32277794e-02 -5.85784793e-01 2.50020504e-01 1.66744635e-01
1.61890268e-01 -1.23513907e-01 -7.16799378e-01 1.72594333e+00
3.14073026e-01 -1.58172905e-01 -4.82360452e-01 7.47233868e-01
1.80929258e-01 1.71269000e-01 6.83953941e-01 -4.06897485e-01
1.43382251e+00 -4.45361435e-01 -5.86513400e-01 -4.65911925e-01
1.31470576e-01 -9.93219197e-01 1.05641234e+00 2.18260258e-01
-1.31215239e+00 -1.92632198e-01 -9.15344894e-01 2.56688833e-01
-2.79364157e-02 -5.47418118e-01 1.47657764e+00 7.77241826e-01
-6.61199510e-01 1.85955077e-01 -2.03406796e-01 2.61151522e-01
4.37225938e-01 2.45009139e-01 3.60536039e-01 1.62611890e-03
-1.12228978e+00 9.06597018e-01 -4.47982818e-01 -3.35824937e-01
-9.28949058e-01 -9.55428183e-01 -2.85635084e-01 -6.77702874e-02
3.19292486e-01 -4.33164954e-01 1.86405253e+00 -1.13625145e+00
-8.88564587e-01 5.02580345e-01 1.89726487e-01 -6.75718963e-01
4.95100796e-01 3.78564722e-03 -6.58346653e-01 -6.22665405e-01
4.43819284e-01 4.83551979e-01 9.25706923e-01 -1.17894387e+00
-8.26232374e-01 -8.23274851e-01 1.82943851e-01 -1.99242666e-01
-5.76574504e-02 4.34355766e-01 1.59963459e-01 -5.86349130e-01
6.58099875e-02 -9.65329528e-01 -3.33111763e-01 -4.50019389e-01
9.69089121e-02 -2.66625255e-01 2.84892052e-01 -6.41990960e-01
1.23710883e+00 -2.19144249e+00 -3.76714170e-01 8.63146037e-02
3.26375157e-01 -5.29317021e-01 1.50455102e-01 2.80803829e-01
-3.39727551e-01 3.35646152e-01 5.81905544e-01 4.06394780e-01
2.72400945e-01 7.84035921e-02 -5.69496751e-01 3.25475335e-01
2.79132426e-02 1.00985444e+00 -3.57003272e-01 -2.75980234e-01
-2.12200865e-01 1.10792905e-01 -5.80427051e-01 -2.11536437e-01
-3.48477960e-01 -2.59948492e-01 -4.25756931e-01 9.31220293e-01
5.38587689e-01 -1.26427427e-01 4.13286865e-01 -2.51066429e-03
-4.40433204e-01 9.17490661e-01 -9.46642578e-01 4.95132565e-01
-9.19515714e-02 4.41656262e-01 1.86140329e-01 -5.06332934e-01
4.87201482e-01 -2.87546571e-02 1.39674768e-01 -1.38664317e+00
1.98501065e-01 1.25846323e-02 2.03614786e-01 -2.11463407e-01
4.80643481e-01 -4.63734150e-01 -2.73260653e-01 7.26899326e-01
-5.13792813e-01 2.76984125e-01 -9.33521912e-02 3.10431421e-01
9.27730501e-01 -1.78167358e-01 1.96383253e-01 -3.23972702e-01
-3.77077043e-01 5.42483389e-01 4.84479606e-01 5.22889853e-01
-3.19904894e-01 -4.72251058e-01 7.21017778e-01 -3.59001219e-01
-1.17610765e+00 -1.08427179e+00 -4.19981986e-01 1.58637893e+00
-2.07518801e-01 2.23992374e-02 -3.63994330e-01 -4.89565313e-01
8.61522436e-01 1.34779048e+00 -9.72545922e-01 -5.27403392e-02
-2.43960455e-01 -8.69614363e-01 6.18117712e-02 4.03505892e-01
-1.43405423e-01 -3.59215945e-01 -4.94881302e-01 1.51408061e-01
2.46018976e-01 -8.10036123e-01 -2.88003266e-01 2.42463112e-01
-1.11207235e+00 -9.88540113e-01 5.74330389e-02 -1.57440871e-01
1.69004500e-01 5.77840209e-01 1.27832663e+00 -2.92110533e-01
-2.37406358e-01 3.64404351e-01 -1.32274628e-01 -7.72305727e-01
-6.03795588e-01 -6.38953626e-01 -1.37242928e-01 -2.25080222e-01
8.40642273e-01 -5.14183104e-01 -7.14929879e-01 6.38898134e-01
-6.08766675e-01 -6.64455056e-01 1.19747722e+00 3.46728116e-01
2.69664735e-01 1.89003214e-01 5.44381082e-01 -8.46914351e-01
9.57267702e-01 -7.33417392e-01 -9.37446892e-01 -8.87213796e-02
-1.42390478e+00 -1.05913192e-01 -1.33459985e-01 -6.64252818e-01
-7.57014394e-01 -1.69944987e-01 2.73123622e-01 2.55222470e-01
2.33086988e-01 6.64482832e-01 2.45928481e-01 -1.80783868e-01
8.14166903e-01 -2.94417828e-01 4.37289208e-01 -5.08984029e-01
8.47222328e-01 6.84169948e-01 1.21811070e-01 -2.00770676e-01
5.75933933e-01 4.50648755e-01 -3.68650198e-01 -2.40898162e-01
-9.44286406e-01 -1.44836843e-01 1.00894265e-01 -1.42528415e-02
4.01015043e-01 -1.09313619e+00 -9.16895092e-01 -5.92683375e-01
-4.73323107e-01 -7.57695362e-02 -3.92352045e-01 8.03522527e-01
-2.61584818e-01 -2.00980946e-01 -3.91018391e-01 -8.44408393e-01
3.18683475e-01 -1.18229508e+00 4.06110197e-01 -1.33948863e-01
-3.86800617e-01 -6.14587367e-01 1.19175911e-02 8.63717973e-01
6.13682508e-01 -9.08011496e-02 1.16072369e+00 -6.59531713e-01
-8.30926359e-01 -7.56067574e-01 -2.99362570e-01 9.06751454e-02
-2.20728546e-01 -2.64223595e-03 -4.05405372e-01 7.77932853e-02
3.01741242e-01 -2.04821005e-01 5.03366947e-01 1.05063498e+00
4.59733099e-01 -6.45803213e-01 -3.11498284e-01 -1.83438361e-01
1.40872371e+00 5.30325294e-01 5.74985802e-01 4.93396461e-01
-2.73027271e-01 8.33677649e-01 7.54383028e-01 4.30106491e-01
5.97155750e-01 7.51576900e-01 4.74482894e-01 -7.63769671e-02
-1.13481425e-01 -5.95201135e-01 4.40797925e-01 5.68710044e-02
1.98389411e-01 4.38286334e-01 -2.06721857e-01 3.54076475e-01
-1.32436430e+00 -1.07005203e+00 -3.76753956e-01 2.20950747e+00
7.12163687e-01 5.99531233e-01 6.34599149e-01 -1.86753377e-01
6.27598882e-01 6.16305834e-03 -5.24518907e-01 -8.09529901e-01
1.89301834e-01 -7.84559473e-02 1.22092605e+00 5.66428304e-01
-6.69119775e-01 6.24512970e-01 8.64007664e+00 2.94802308e-01
-6.55259550e-01 3.96872759e-01 1.05083954e+00 -1.44244611e-01
-7.30013907e-01 3.75721633e-01 -1.00864542e+00 5.39679348e-01
1.45360351e+00 -3.94760042e-01 5.58830857e-01 1.33191800e+00
5.88048756e-01 2.45348681e-02 -8.72624516e-01 1.06358014e-01
-1.52376845e-01 -1.22656322e+00 9.10962522e-02 7.98429132e-01
4.58868116e-01 -2.29381457e-01 4.88174200e-01 4.14417744e-01
1.31656182e+00 -7.40145624e-01 1.15896082e+00 -1.78628005e-02
5.27691782e-01 -3.09683263e-01 3.48019481e-01 -1.55854777e-01
-2.97588617e-01 -8.24749529e-01 -2.84909934e-01 -5.73169708e-01
2.75288999e-01 6.18700624e-01 -9.07543898e-01 -2.64896452e-01
4.35630172e-01 -1.16889231e-01 -4.88386750e-01 6.45336509e-01
-3.35273355e-01 8.73494446e-01 1.77642822e-01 -3.46162729e-02
2.94409305e-01 -1.10809214e-01 1.39181793e-01 6.40486538e-01
-2.20570058e-01 1.40423894e-01 -1.85686693e-01 9.25514281e-01
-4.29503508e-02 2.54614688e-02 -3.12470168e-01 -4.33947086e-01
5.54104388e-01 9.55050111e-01 -2.76920021e-01 -1.13058895e-01
-6.23722434e-01 2.20357120e-01 -1.46660181e-02 2.44758595e-02
-9.70257998e-01 4.83768553e-01 6.49592102e-01 7.75904655e-01
3.33841234e-01 -7.42253587e-02 -4.78007674e-01 -2.86915630e-01
-1.58325627e-01 -1.05304170e+00 4.54074711e-01 -6.17799342e-01
-1.23306656e+00 -3.48153920e-03 6.11727126e-02 -4.04246420e-01
-4.80030067e-02 -5.96635282e-01 -2.59240568e-01 6.15096271e-01
-1.41996217e+00 -1.02614796e+00 -5.44492900e-02 1.49577141e-01
3.67884606e-01 6.76170969e-03 4.52822030e-01 3.79313380e-01
-2.17377067e-01 4.49051708e-01 -2.76172996e-01 -5.63996077e-01
6.20942652e-01 -1.14224899e+00 2.79998153e-01 3.31260890e-01
-7.29729682e-02 7.72706985e-01 8.71125400e-01 -9.67052460e-01
-1.44747162e+00 -6.66455626e-01 8.60047638e-01 -5.65790415e-01
1.39608419e+00 -2.17921779e-01 -8.73974487e-02 1.18893445e+00
9.93129015e-02 -8.27142537e-01 9.09510970e-01 7.40253031e-01
-6.71994150e-01 -3.81920934e-01 -8.36462200e-01 5.93561053e-01
6.73688293e-01 -2.29144484e-01 -6.94749713e-01 6.53986633e-01
1.14450145e+00 4.62250769e-01 -9.80802536e-01 -7.86711276e-02
9.39867377e-01 -9.26171422e-01 1.05949843e+00 -1.00683463e+00
6.36590838e-01 3.97565871e-01 -7.07534790e-01 -8.48455012e-01
-8.69458914e-01 -6.59340501e-01 3.63408417e-01 9.43348110e-01
1.00306666e+00 -6.47882104e-01 7.66781926e-01 9.45494294e-01
1.57196134e-01 -4.80861962e-01 -6.42719984e-01 -6.40874922e-01
7.69396499e-02 -6.35550022e-01 8.25449765e-01 9.86786127e-01
-2.16793180e-01 5.66344440e-01 -1.45807058e-01 7.05484971e-02
1.08175623e+00 1.91883385e-01 7.02520728e-01 -1.60357368e+00
-4.77184832e-01 -5.77671528e-01 -4.23460305e-01 -9.36070561e-01
-2.32831359e-01 -5.94601870e-01 -5.71979761e-01 -1.11703813e+00
5.07866085e-01 -6.94393814e-01 -2.02353969e-01 -1.98679678e-02
1.68444902e-01 -4.25692089e-02 -1.51543647e-01 5.96506238e-01
-8.41405988e-02 -7.96169713e-02 7.81578660e-01 -4.41967919e-02
-3.18482935e-01 3.38695228e-01 -1.64281571e+00 5.76721549e-01
5.13622046e-01 -4.70929742e-01 -1.48338720e-01 -2.02795103e-01
4.03528571e-01 3.73944551e-01 7.47543991e-01 8.76302570e-02
1.01038329e-01 -6.23756468e-01 9.62847173e-02 -3.06361675e-01
1.39805555e-01 -8.55230272e-01 5.50042033e-01 5.80607891e-01
-9.09847260e-01 4.61265922e-01 2.15739254e-02 8.28077734e-01
8.25232863e-02 -1.33238867e-01 5.57024479e-01 -3.69480431e-01
-4.65709805e-01 -1.18509531e-01 5.09183519e-02 -1.33488536e-01
9.88388836e-01 -1.12844571e-01 -4.98788327e-01 -5.35516441e-01
-7.11684883e-01 -3.00025582e-01 2.46318817e-01 3.49710792e-01
-7.43917003e-02 -1.24662161e+00 -8.19791794e-01 -1.86760560e-01
-1.15624676e-02 -1.42674339e+00 -2.54001766e-01 7.70794511e-01
3.93946171e-02 9.40802693e-01 9.02816951e-02 5.83135989e-03
-1.00898576e+00 8.63455951e-01 -2.98973005e-02 -4.62193638e-02
-1.38221204e-01 8.47235143e-01 2.29722708e-01 3.15896899e-01
1.17325395e-01 2.22121373e-01 1.73758924e-01 1.87092453e-01
4.59896952e-01 3.37239861e-01 -9.45018902e-02 -3.61205995e-01
-1.86864957e-01 -1.88396931e-01 -5.60205996e-01 -5.37186980e-01
1.23691869e+00 -1.55446112e-01 -1.59660533e-01 3.22100192e-01
7.30304658e-01 4.48113173e-01 -1.02587199e+00 -2.71287002e-02
2.21530095e-01 -8.61109495e-01 7.26409316e-01 -1.42024088e+00
-7.64784634e-01 3.04640085e-01 7.78887868e-01 7.77480483e-01
6.40709341e-01 2.56318837e-01 3.35748196e-01 -4.60812479e-01
1.96035847e-01 -1.04691184e+00 1.12788873e-02 -8.26208830e-01
7.84284174e-01 -1.12211072e+00 4.98165905e-01 -3.93971622e-01
-6.58734143e-01 4.12510276e-01 1.06508538e-01 3.33252847e-02
7.88434982e-01 1.35081902e-01 -1.38951153e-01 -6.50535643e-01
-9.58198607e-01 1.59652289e-02 1.01527281e-03 1.23643525e-01
6.01185083e-01 7.55997002e-01 -1.02831542e+00 7.51775742e-01
-3.20086926e-01 4.03346047e-02 4.80634123e-01 5.97610414e-01
-5.10321736e-01 -1.08694053e+00 -3.24186414e-01 6.39620185e-01
-7.84663796e-01 -2.85721153e-01 -7.70783007e-01 1.23634839e+00
-2.08350681e-02 1.14832926e+00 2.14642450e-01 -4.83062744e-01
5.80568552e-01 -9.91627499e-02 1.23234093e-01 -1.28612056e-01
-3.30590487e-01 5.19901872e-01 4.14749742e-01 -4.73681599e-01
-3.24753135e-01 -1.00965381e+00 -3.28945458e-01 -6.87209964e-01
-7.13929236e-01 2.96851933e-01 1.24772859e+00 5.42233765e-01
5.05612671e-01 3.12924147e-01 1.03912902e+00 1.24389336e-01
-1.36543810e+00 -8.30618560e-01 -6.86425984e-01 1.74126521e-01
-9.11660418e-02 -1.08980060e+00 -9.45129156e-01 -2.96713740e-01] | [9.130826950073242, 5.775877952575684] |
f5dcecef-9f63-4b7a-a5a1-f3f9bc179b56 | unsupervised-multilingual-word-embeddings | 1808.08933 | null | http://arxiv.org/abs/1808.08933v2 | http://arxiv.org/pdf/1808.08933v2.pdf | Unsupervised Multilingual Word Embeddings | Multilingual Word Embeddings (MWEs) represent words from multiple languages
in a single distributional vector space. Unsupervised MWE (UMWE) methods
acquire multilingual embeddings without cross-lingual supervision, which is a
significant advantage over traditional supervised approaches and opens many new
possibilities for low-resource languages. Prior art for learning UMWEs,
however, merely relies on a number of independently trained Unsupervised
Bilingual Word Embeddings (UBWEs) to obtain multilingual embeddings. These
methods fail to leverage the interdependencies that exist among many languages.
To address this shortcoming, we propose a fully unsupervised framework for
learning MWEs that directly exploits the relations between all language pairs.
Our model substantially outperforms previous approaches in the experiments on
multilingual word translation and cross-lingual word similarity. In addition,
our model even beats supervised approaches trained with cross-lingual
resources. | ['Claire Cardie', 'Xilun Chen'] | 2018-08-27 | unsupervised-multilingual-word-embeddings-1 | https://aclanthology.org/D18-1024 | https://aclanthology.org/D18-1024.pdf | emnlp-2018-10 | ['multilingual-word-embeddings'] | ['methodology'] | [-6.07499778e-01 -5.41575968e-01 -8.69745672e-01 -2.20215082e-01
-1.05895090e+00 -7.39301980e-01 7.40052581e-01 8.51108804e-02
-8.19481611e-01 7.87160456e-01 5.75906873e-01 -7.54323542e-01
2.26019442e-01 -6.55989707e-01 -5.49455404e-01 -3.99823129e-01
2.15215191e-01 5.03058493e-01 -2.74450004e-01 -4.87970531e-01
-1.56948388e-01 1.14856042e-01 -8.55342090e-01 -1.22505121e-01
1.15499735e+00 2.74882540e-02 3.32286000e-01 1.68797120e-01
-6.51579857e-01 1.86837912e-01 -1.01409266e-02 -6.51810646e-01
2.01330185e-01 -3.33873868e-01 -5.65175772e-01 -3.33119422e-01
4.22000796e-01 -7.54413605e-02 -3.53194565e-01 1.22347796e+00
6.19698644e-01 -2.15408862e-01 7.18015254e-01 -1.02435327e+00
-1.41197026e+00 9.70758140e-01 -5.63211024e-01 3.39017063e-01
2.40594998e-01 -1.50579184e-01 1.66160083e+00 -1.52964783e+00
7.50656366e-01 1.13393950e+00 8.29750299e-01 1.78580835e-01
-1.21690762e+00 -6.45141363e-01 1.22845441e-01 2.34852850e-01
-1.52056205e+00 -2.14070603e-01 6.12348437e-01 -5.41649222e-01
1.18995559e+00 -2.29338065e-01 4.71149832e-01 1.21704507e+00
2.25818858e-01 6.59072161e-01 1.24117959e+00 -8.98229659e-01
-4.10865873e-01 4.72845048e-01 3.04623991e-01 6.28276825e-01
3.76199901e-01 8.22784305e-02 -4.99460250e-01 -6.21442571e-02
5.71258783e-01 1.40136061e-02 -1.37748733e-01 -5.30618191e-01
-1.66052103e+00 1.07179880e+00 1.46471217e-01 7.83568799e-01
-1.00854494e-01 -1.93156734e-01 5.46904624e-01 6.29426181e-01
5.70321798e-01 5.63923001e-01 -7.28575528e-01 -2.74399389e-02
-5.58227301e-01 -3.70315641e-01 4.89826918e-01 9.53093708e-01
1.18013740e+00 1.25434160e-01 2.71559685e-01 1.19679499e+00
3.15428376e-01 8.37853611e-01 9.61014509e-01 -9.28993970e-02
4.56769645e-01 4.72254187e-01 -1.76751330e-01 -7.21607924e-01
8.00287575e-02 -2.33797565e-01 -4.45121586e-01 -3.63354087e-01
1.63747892e-01 -1.71772227e-01 -5.58195531e-01 1.75112879e+00
1.30736351e-01 -8.45891684e-02 3.75588268e-01 7.41702080e-01
5.49905658e-01 7.01695442e-01 3.21289487e-02 -9.09052417e-02
1.32729518e+00 -1.27480829e+00 -9.23941314e-01 -2.42724359e-01
8.90227914e-01 -1.29107404e+00 1.51507890e+00 -1.90558299e-01
-5.57244599e-01 -5.15121102e-01 -1.07057142e+00 -3.26630175e-01
-9.28726256e-01 2.74694353e-01 8.63845229e-01 5.87319255e-01
-9.34499800e-01 1.00736111e-01 -6.53879285e-01 -5.66604197e-01
-6.95954338e-02 1.73644394e-01 -7.14196682e-01 -4.13872302e-01
-1.62933707e+00 1.26018000e+00 4.36418235e-01 -1.91747308e-01
-5.74492097e-01 -5.93236506e-01 -1.31640434e+00 -3.22389692e-01
-2.31790021e-02 -2.03326181e-01 7.86850333e-01 -1.05089104e+00
-1.29982388e+00 9.25463557e-01 -2.65582204e-01 -1.08133800e-01
6.04635850e-02 -3.40145320e-01 -7.84784019e-01 -3.37444782e-01
3.94457221e-01 4.74105924e-01 4.37107533e-01 -1.05822313e+00
-4.02432561e-01 -1.09640412e-01 -9.56535414e-02 2.32461885e-01
-1.02542698e+00 3.36196184e-01 -4.84886616e-01 -9.71320748e-01
-2.46606603e-01 -7.75268734e-01 -1.70465589e-01 -4.07203406e-01
-2.94808708e-02 -4.18991268e-01 4.72566634e-01 -7.27632225e-01
1.33919859e+00 -2.01326275e+00 3.22081089e-01 -1.37393981e-01
-1.52239174e-01 3.44700634e-01 -5.95449865e-01 8.83535862e-01
-1.12380318e-01 1.11259162e-01 -2.60552093e-02 -3.03483397e-01
2.51218498e-01 7.45107353e-01 -4.07011777e-01 4.45435375e-01
2.78777272e-01 1.16132665e+00 -1.30880237e+00 -6.57835305e-01
1.81088626e-01 4.97323453e-01 -4.32824075e-01 1.61841810e-01
2.40684643e-01 2.17250869e-01 -2.17293069e-01 6.24631107e-01
5.44464171e-01 2.70162582e-01 7.24808991e-01 -1.59742445e-01
-3.94364923e-01 5.52796662e-01 -7.93333709e-01 2.04715633e+00
-9.54153180e-01 6.47317231e-01 -4.34320897e-01 -1.02904487e+00
9.11928594e-01 5.21034002e-01 2.75457561e-01 -5.86123526e-01
1.03384063e-01 8.29405129e-01 2.74052694e-02 -5.08438289e-01
5.43778002e-01 -4.34642345e-01 -3.85600775e-01 7.02274382e-01
7.37518668e-01 1.76820293e-01 3.30466777e-01 5.29146641e-02
4.70598429e-01 2.54002690e-01 6.66713357e-01 -6.31012142e-01
4.79088217e-01 -7.66500533e-02 8.57237697e-01 1.78393558e-01
-1.30639300e-01 2.00666383e-01 1.72057971e-01 -5.69152892e-01
-1.18840337e+00 -1.38747954e+00 -4.04576778e-01 1.29597735e+00
1.30950913e-01 -7.60287642e-01 -2.44861871e-01 -9.15789127e-01
3.52054462e-02 5.57087004e-01 -4.65589732e-01 1.55134276e-01
-5.91988921e-01 -8.80689323e-01 5.72952032e-01 6.71696544e-01
-2.65792638e-01 -7.95130551e-01 3.11633110e-01 1.50866106e-01
-1.94391161e-01 -1.35617507e+00 -9.43604946e-01 2.15990111e-01
-7.37580419e-01 -8.42709363e-01 -6.65493131e-01 -1.29481328e+00
7.43930578e-01 3.89159620e-01 1.28005230e+00 -3.38484824e-01
-1.47996545e-01 2.28156194e-01 -5.00735462e-01 -2.17769131e-01
-2.94261009e-01 3.63682389e-01 7.76610255e-01 -8.64697248e-02
1.17234576e+00 -6.68209672e-01 6.71974793e-02 1.03679925e-01
-8.65449846e-01 -3.70829284e-01 6.32251203e-01 1.11217546e+00
6.94306552e-01 -3.86934727e-01 7.48295963e-01 -8.21491301e-01
8.18463862e-01 -5.71015894e-01 -4.87000585e-01 6.08426392e-01
-6.96416914e-01 3.97910357e-01 7.08814859e-01 -6.79220080e-01
-6.89226627e-01 -3.96185577e-01 -1.54317439e-01 -1.91791043e-01
6.77764267e-02 7.36907601e-01 -1.45664617e-01 9.78704244e-02
3.46601695e-01 2.63399929e-01 -3.71340215e-01 -6.78278625e-01
8.73132408e-01 9.71962154e-01 2.22811028e-01 -8.64587486e-01
1.01130140e+00 -7.31514469e-02 -7.48126924e-01 -7.62486100e-01
-6.34876251e-01 -5.94336748e-01 -1.23766685e+00 2.56560385e-01
9.46616054e-01 -1.26323748e+00 2.91335970e-01 -1.02910638e-01
-1.26233804e+00 1.02821104e-01 -1.39873594e-01 1.16791832e+00
6.39845384e-03 3.93821627e-01 -6.88036323e-01 -3.59425575e-01
-9.23005342e-02 -1.15337169e+00 7.68620670e-01 -2.33928576e-01
-3.13912541e-01 -1.67065382e+00 6.55147135e-01 1.04614168e-01
3.50684434e-01 -2.52336204e-01 1.13710523e+00 -6.92253113e-01
-1.37713581e-01 -9.90628973e-02 -1.14276178e-01 6.41540825e-01
6.50777042e-01 -3.82143147e-02 -5.76797307e-01 -4.08477873e-01
-4.82429236e-01 -5.57816148e-01 5.99746943e-01 -9.57255438e-02
2.85121709e-01 -3.66297294e-03 -1.66263297e-01 5.82420766e-01
1.70368946e+00 -2.22722769e-01 6.49110004e-02 4.41986531e-01
9.47613716e-01 3.75444412e-01 3.31603885e-01 -4.89489101e-02
6.71913385e-01 4.69400704e-01 -2.07216039e-01 -2.76527435e-01
-1.30956545e-01 -5.66639602e-01 9.41401124e-01 2.21190810e+00
1.56089175e-03 2.38810524e-01 -1.07193792e+00 1.12496412e+00
-1.54640377e+00 -5.33541679e-01 -5.71742728e-02 2.11278296e+00
1.22630501e+00 -2.50883102e-01 -1.50913581e-01 -3.79588515e-01
5.46715915e-01 3.57668966e-01 -1.38632759e-01 -8.27831209e-01
-3.40831220e-01 5.52386343e-01 6.37069166e-01 8.38825762e-01
-9.19055939e-01 1.49246430e+00 6.61902428e+00 7.42550671e-01
-1.00947940e+00 7.79545009e-01 -1.38434827e-01 3.13154340e-01
-9.10167754e-01 8.43715668e-02 -8.25839579e-01 2.36694068e-01
7.38864601e-01 -3.44195426e-01 3.72585028e-01 7.22793341e-01
-2.15388969e-01 4.60031033e-01 -1.15849054e+00 9.01797116e-01
4.07695770e-01 -9.46277201e-01 2.92188466e-01 5.54216020e-02
1.22262502e+00 5.73625267e-01 8.33386406e-02 5.48922658e-01
8.04695666e-01 -8.94135237e-01 1.77032784e-01 -1.61011908e-02
1.16313469e+00 -9.12311971e-01 9.40674186e-01 2.28292957e-01
-1.32977045e+00 4.24783796e-01 -6.96540594e-01 3.12631577e-02
2.10849002e-01 5.97301543e-01 -5.17482519e-01 8.62763345e-01
3.20165813e-01 9.69417632e-01 -4.83397186e-01 2.73171335e-01
-6.11543000e-01 5.04606545e-01 -2.95928847e-02 1.67876735e-01
5.67831993e-01 -5.76518595e-01 3.46530527e-01 1.55070937e+00
4.42026407e-01 -6.96377039e-01 3.86846721e-01 4.17054385e-01
-2.98639417e-01 7.88271010e-01 -9.94613111e-01 -6.46015584e-01
4.20259476e-01 1.09389865e+00 -1.22133501e-01 -3.78288805e-01
-1.03192091e+00 1.15174127e+00 7.94646800e-01 4.11097139e-01
-6.71192765e-01 -4.79404896e-01 1.09595537e+00 -2.69386560e-01
2.32822746e-01 -8.08755219e-01 1.32699117e-01 -1.79742432e+00
3.39740589e-02 -1.03071201e+00 1.74770236e-01 -1.99632362e-01
-1.81970835e+00 8.60875070e-01 -3.23279440e-01 -1.38047445e+00
-2.35835403e-01 -9.66647327e-01 -4.19751018e-01 1.03241372e+00
-2.16888595e+00 -1.49631715e+00 5.35455287e-01 6.13174796e-01
6.17573738e-01 -5.89161277e-01 1.40605676e+00 6.90665841e-01
-5.54088652e-01 9.30066407e-01 5.97518027e-01 4.59933668e-01
1.27849174e+00 -1.31985521e+00 4.00636882e-01 9.04289961e-01
8.88581395e-01 1.12890542e+00 2.50374317e-01 -3.95938426e-01
-1.52660370e+00 -9.77106512e-01 1.91403961e+00 -6.39682055e-01
1.39409316e+00 -4.90009308e-01 -7.41185009e-01 1.01953375e+00
8.93609941e-01 1.52214631e-01 1.51942515e+00 7.09407210e-01
-9.40181375e-01 -1.79390237e-02 -3.28130990e-01 8.08005393e-01
9.04608011e-01 -1.11604464e+00 -1.13453209e+00 3.73553216e-01
7.98569858e-01 2.84010768e-01 -1.06266868e+00 1.65100187e-01
5.67729950e-01 -4.69077408e-01 9.33879256e-01 -1.02145123e+00
4.25277621e-01 -2.52310455e-01 -4.94702041e-01 -1.85097349e+00
-1.25441447e-01 -3.85163963e-01 2.77752757e-01 1.30118823e+00
6.31227553e-01 -9.25382674e-01 -1.13454439e-01 -2.82197356e-01
3.90280690e-03 -5.25034189e-01 -7.08920836e-01 -1.18433225e+00
7.20292628e-01 -4.48504061e-01 6.50493205e-01 1.79333496e+00
3.16131800e-01 5.56568861e-01 -5.26517630e-01 2.24139199e-01
6.06389165e-01 2.71103382e-01 5.96960664e-01 -8.77048433e-01
-2.41160989e-01 -4.02507514e-01 -4.05919552e-01 -1.01336682e+00
7.98326373e-01 -1.55827212e+00 -7.91010559e-02 -1.36829448e+00
2.28397012e-01 -4.27107930e-01 -7.05440998e-01 3.76679748e-01
-5.26478589e-01 4.61662263e-01 -1.37971014e-01 2.42477149e-01
-4.41621661e-01 6.76935792e-01 1.00103748e+00 -1.21390253e-01
7.45594427e-02 -8.67990255e-01 -5.20430267e-01 7.62318432e-01
7.53445625e-01 -5.96984684e-01 -2.54310906e-01 -1.19040596e+00
2.52600908e-01 -5.10769069e-01 -3.91361058e-01 -3.90707403e-01
3.18127945e-02 -3.03830475e-01 -4.57663182e-03 -2.48255670e-01
-1.28751472e-01 -7.72578418e-01 -3.77147317e-01 4.34813164e-02
-1.22260801e-01 6.19734406e-01 -1.03319101e-02 3.71977508e-01
-8.09862196e-01 -2.01273605e-01 4.28723842e-01 -5.26906066e-02
-8.54706883e-01 3.63618851e-01 -4.15678233e-01 2.91715145e-01
7.58145690e-01 2.51945198e-01 8.95941556e-02 1.99037373e-01
-4.93171573e-01 2.69156426e-01 5.32894909e-01 9.55713689e-01
2.98588514e-01 -2.12543368e+00 -7.66077638e-01 4.77310836e-01
7.31058598e-01 -6.81551039e-01 -3.24954420e-01 7.89112508e-01
-3.16591144e-01 4.40362841e-01 -3.18294525e-01 -1.92222044e-01
-9.74765182e-01 6.75413966e-01 -7.04041729e-03 -4.55569059e-01
-2.99607247e-01 6.29860938e-01 2.04666667e-02 -1.31534874e+00
-7.58913308e-02 -3.89703996e-02 -1.23011768e-01 3.12506288e-01
3.52625370e-01 -8.86610299e-02 -1.31413326e-01 -1.13057828e+00
-3.92103672e-01 8.87229204e-01 -2.02297360e-01 -2.20115885e-01
1.35648084e+00 -1.35852918e-01 -4.85774636e-01 9.67170894e-01
1.53771853e+00 6.71798944e-01 -5.09605885e-01 -7.95169592e-01
2.89074928e-01 -5.89393854e-01 -9.56557542e-02 -2.55376071e-01
-7.76559591e-01 1.17829561e+00 2.83041120e-01 -4.44699675e-01
6.05888605e-01 1.41082471e-02 1.13120842e+00 4.08552200e-01
4.85773057e-01 -1.33379221e+00 -1.20765641e-01 7.87018538e-01
3.78297836e-01 -1.54767132e+00 -2.30229750e-01 -1.02292195e-01
-6.45604610e-01 1.29901111e+00 4.43336010e-01 -1.14544556e-01
7.38234162e-01 3.20174664e-01 6.09165847e-01 1.24950185e-01
-4.63579267e-01 -4.50618654e-01 3.73150587e-01 5.38233697e-01
9.78240788e-01 4.55043346e-01 -8.71023953e-01 4.95349795e-01
-1.20151013e-01 -3.17568600e-01 8.13288689e-02 7.11541355e-01
-1.98343143e-01 -2.07210708e+00 -1.86667159e-01 -1.14756882e-01
-4.11531925e-01 -7.87476480e-01 -2.11598441e-01 8.35982919e-01
3.70946765e-01 6.16371930e-01 -8.50594342e-02 -3.19874048e-01
1.12228990e-01 4.71623629e-01 4.88724113e-01 -8.27395439e-01
-1.97495222e-01 3.39557379e-02 -2.82193441e-02 -2.25279912e-01
-5.28435826e-01 -5.15271306e-01 -8.28665435e-01 -1.70752764e-01
-3.23212117e-01 3.72647405e-01 4.95315999e-01 9.42508280e-01
1.00280561e-01 2.52389342e-01 7.39916503e-01 -4.84904051e-01
-4.50354367e-01 -1.00428355e+00 -5.61884880e-01 5.05316257e-01
1.49850458e-01 -5.28467238e-01 -2.92004168e-01 5.92192672e-02] | [11.067204475402832, 9.933663368225098] |
adf05f50-1db7-4672-8d7d-0a631b3c2f46 | bayesian-neural-ordinary-differential | 2012.07244 | null | https://arxiv.org/abs/2012.07244v4 | https://arxiv.org/pdf/2012.07244v4.pdf | Bayesian Neural Ordinary Differential Equations | Recently, Neural Ordinary Differential Equations has emerged as a powerful framework for modeling physical simulations without explicitly defining the ODEs governing the system, but instead learning them via machine learning. However, the question: "Can Bayesian learning frameworks be integrated with Neural ODE's to robustly quantify the uncertainty in the weights of a Neural ODE?" remains unanswered. In an effort to address this question, we primarily evaluate the following categories of inference methods: (a) The No-U-Turn MCMC sampler (NUTS), (b) Stochastic Gradient Hamiltonian Monte Carlo (SGHMC) and (c) Stochastic Langevin Gradient Descent (SGLD). We demonstrate the successful integration of Neural ODEs with the above Bayesian inference frameworks on classical physical systems, as well as on standard machine learning datasets like MNIST, using GPU acceleration. On the MNIST dataset, we achieve a posterior sample accuracy of 98.5% on the test ensemble of 10,000 images. Subsequently, for the first time, we demonstrate the successful integration of variational inference with normalizing flows and Neural ODEs, leading to a powerful Bayesian Neural ODE object. Finally, considering a predator-prey model and an epidemiological system, we demonstrate the probabilistic identification of model specification in partially-described dynamical systems using universal ordinary differential equations. Together, this gives a scientific machine learning tool for probabilistic estimation of epistemic uncertainties. | ['Krishna Vishal Vemula', 'Aslan Garcia-Valadez', 'Mohamed Tarek', 'Vaibhav Dixit', 'Karen Chung', 'Chris Rackauckas', 'Raj Dandekar'] | 2020-12-14 | null | null | null | null | ['physical-simulations'] | ['miscellaneous'] | [-1.26249090e-01 -2.65306771e-01 1.90399885e-01 -4.98371348e-02
-7.25184023e-01 -4.63095397e-01 9.87152874e-01 -1.99290588e-01
-7.00156629e-01 1.02650046e+00 -3.10532570e-01 -5.66478252e-01
-4.30838495e-01 -7.92356372e-01 -7.11917460e-01 -1.19958103e+00
2.33791526e-02 8.02195668e-01 2.51357406e-01 2.25165635e-01
1.32422894e-01 6.65055454e-01 -1.25209081e+00 -5.01193225e-01
8.26447070e-01 8.43309224e-01 -1.67529806e-01 7.80238330e-01
2.32349962e-01 5.48744857e-01 -2.28777945e-01 -6.38614357e-01
-8.87368694e-02 -2.86129415e-01 -3.75883311e-01 -4.86208141e-01
-2.38505751e-02 -4.43379164e-01 -3.82270306e-01 1.17963111e+00
4.72299010e-01 2.08702713e-01 1.44075274e+00 -8.78872633e-01
-4.26591784e-01 5.71012914e-01 -3.43178481e-01 3.03665668e-01
-5.62910475e-02 6.13092661e-01 6.92939103e-01 -6.05151117e-01
4.68112618e-01 1.41156006e+00 8.59142780e-01 5.28231323e-01
-1.69613814e+00 -6.43078685e-01 -3.49751592e-01 -2.12753797e-03
-1.65620577e+00 -1.22510590e-01 4.26404953e-01 -9.41119909e-01
8.48421812e-01 6.30308837e-02 6.67969763e-01 1.67895687e+00
6.05867565e-01 5.05422771e-01 1.38194966e+00 -1.67208493e-01
9.19570327e-01 1.23860471e-01 5.58812618e-01 6.69342458e-01
5.57749510e-01 6.87329590e-01 -3.02768886e-01 -7.70670712e-01
6.60953164e-01 -1.37772352e-01 1.82199106e-01 4.86020595e-02
-7.03458607e-01 1.21314597e+00 -1.09512150e-01 -2.62764424e-01
-4.27252442e-01 7.83384144e-01 3.09696972e-01 -5.00584126e-01
5.71767211e-01 8.84696394e-02 -2.36012504e-01 -3.05275340e-03
-9.27841246e-01 5.98648131e-01 1.21367967e+00 3.66885930e-01
5.62658370e-01 3.27656120e-01 -1.53046064e-02 5.23937583e-01
9.19643402e-01 1.23754406e+00 -1.74334452e-01 -1.25789547e+00
-3.48170578e-01 -4.11970504e-02 1.62384093e-01 -4.15341228e-01
-2.89334089e-01 -3.09716821e-01 -1.13694966e+00 4.93627667e-01
5.37861049e-01 -6.26845181e-01 -7.02053368e-01 1.87639832e+00
4.68658239e-01 3.28402728e-01 -3.70564088e-02 4.29079890e-01
4.37615544e-01 7.54872322e-01 1.63833395e-01 -3.12883705e-01
1.50080025e+00 -1.56330131e-02 -6.07234478e-01 2.77378410e-01
2.27100655e-01 -3.41421068e-01 4.58951354e-01 5.33723533e-01
-9.45857763e-01 -5.32019548e-02 -8.22865784e-01 3.46813411e-01
-3.83839190e-01 -7.57576674e-02 6.62465274e-01 9.96225536e-01
-8.24460566e-01 7.20100105e-01 -1.40414512e+00 -2.55654544e-01
4.31328952e-01 1.29152238e-01 2.81478345e-01 2.69125730e-01
-1.26426673e+00 9.91288185e-01 8.52957442e-02 1.40650898e-01
-1.43015873e+00 -8.08404982e-01 -5.99784255e-01 -1.51421770e-01
2.38147259e-01 -9.98671353e-01 9.48428571e-01 -4.37961444e-02
-1.80988050e+00 5.14081717e-01 2.57127248e-02 -7.44305134e-01
6.84795618e-01 3.83838639e-02 -9.96187106e-02 3.37855592e-02
-3.71000499e-01 4.82815325e-01 7.17658818e-01 -1.10033321e+00
-4.43517696e-03 -2.55120009e-01 -1.87234819e-01 -3.35244447e-01
1.80988163e-01 -2.84467768e-02 -5.94284162e-02 -4.04241353e-01
-2.35887885e-01 -1.14275444e+00 -2.61781603e-01 8.37184265e-02
-5.47232687e-01 -1.84360400e-01 2.74252623e-01 -6.60073102e-01
9.28521693e-01 -1.83809757e+00 2.05633342e-01 3.02883416e-01
7.09842443e-02 6.73989952e-02 4.35932636e-01 3.87164205e-01
4.07192111e-01 6.55840263e-02 -7.61014879e-01 -3.60603511e-01
3.92340034e-01 4.68277574e-01 -5.22599041e-01 6.88920796e-01
1.53019011e-01 8.07855368e-01 -6.96753621e-01 -4.69053596e-01
2.55741656e-01 9.16880131e-01 -7.05214262e-01 -5.22077270e-02
-2.70409197e-01 5.60898840e-01 -4.26582158e-01 2.02218413e-01
6.62939787e-01 -2.88482010e-01 -1.65829211e-02 -1.12120487e-01
-1.22846402e-01 -8.93840492e-02 -1.32451403e+00 1.02875280e+00
-1.34704322e-01 3.28411847e-01 -1.05971493e-01 -8.34848046e-01
5.32555699e-01 1.18174642e-01 1.43477246e-01 -6.11013472e-02
1.81833178e-01 1.96033701e-01 1.09887756e-01 -3.92894089e-01
6.21631183e-02 -3.47262144e-01 -1.13150433e-01 5.60123444e-01
3.68773878e-01 -4.33668613e-01 2.03268841e-01 1.99401915e-01
9.09054279e-01 3.73627871e-01 1.95000291e-01 -6.33909225e-01
2.33126789e-01 2.70524435e-02 2.88059026e-01 1.11493492e+00
-1.03681944e-01 2.92804629e-01 7.69845545e-01 -4.69461381e-02
-1.02056324e+00 -1.67085123e+00 -8.83306324e-01 3.98301870e-01
-2.89906204e-01 4.44786847e-02 -1.00612020e+00 4.06293236e-02
2.21525788e-01 8.04286957e-01 -7.06175923e-01 -2.44459748e-01
-6.01995066e-02 -1.80401945e+00 1.02378595e+00 1.56727150e-01
3.13919157e-01 -6.22870088e-01 -6.55583262e-01 1.40702546e-01
3.30125540e-01 -7.72653341e-01 2.60014474e-01 2.35842064e-01
-6.95537031e-01 -1.01769614e+00 -6.98796988e-01 2.66303420e-01
3.30261469e-01 -6.95809603e-01 8.01158071e-01 -4.81678724e-01
-5.42107582e-01 4.62023228e-01 2.82868922e-01 -2.16115236e-01
-7.38682985e-01 -2.72606373e-01 4.84442502e-01 -2.11730242e-01
7.57166967e-02 -6.82691336e-01 -5.94741464e-01 1.03303671e-01
-9.47440863e-01 -1.14414811e-01 3.73960257e-01 7.10949481e-01
5.00445426e-01 -1.56116784e-01 3.41198891e-01 -7.47951150e-01
4.87322181e-01 -7.91750193e-01 -1.14571416e+00 1.89996868e-01
-6.29403472e-01 3.66940886e-01 2.79786319e-01 -5.20857930e-01
-1.39234173e+00 -9.92552117e-02 -4.89230692e-01 -1.68026194e-01
-9.72252786e-02 7.24133015e-01 3.07440042e-01 1.15239896e-01
7.19813764e-01 1.75793588e-01 5.76972291e-02 -5.09057939e-01
3.21080297e-01 4.42227334e-01 5.87994635e-01 -1.01715219e+00
4.29575592e-01 9.61789131e-01 6.44772530e-01 -9.13912237e-01
-6.15830302e-01 7.42031112e-02 -3.54821503e-01 -2.84816176e-01
1.16257489e+00 -8.62160265e-01 -1.20631242e+00 9.66456056e-01
-1.20597720e+00 -6.56165361e-01 -2.65154272e-01 8.64630878e-01
-5.65933406e-01 3.43433440e-01 -9.13523018e-01 -1.66970479e+00
-6.28234297e-02 -1.31643856e+00 8.31201553e-01 2.04540268e-01
-2.16412321e-01 -1.31873953e+00 6.19529068e-01 -4.58686240e-02
3.33308905e-01 4.36005592e-01 8.55310678e-01 -4.25234467e-01
-4.08060938e-01 -1.37316167e-01 -2.13065803e-01 4.33621258e-01
-3.62041861e-01 7.71164834e-01 -1.21466100e+00 -2.77907997e-02
2.21053109e-01 -4.28403318e-02 1.04286754e+00 1.03384066e+00
6.52258456e-01 -1.66668013e-01 -2.15259597e-01 7.14301229e-01
1.40287471e+00 6.04061149e-02 4.87538218e-01 -4.46215004e-01
4.33018744e-01 6.15193546e-01 -2.21112594e-01 6.82100236e-01
1.65774360e-01 4.20939058e-01 2.96297014e-01 5.15597761e-01
2.09095955e-01 -8.68302882e-02 5.04154742e-01 7.61729062e-01
-8.83667618e-02 -1.04103610e-01 -1.05964684e+00 1.19277284e-01
-1.74536109e+00 -1.08715582e+00 -6.24174953e-01 2.10102034e+00
9.11219776e-01 1.17011644e-01 3.02148629e-02 -4.10286456e-01
6.10619545e-01 -9.57863107e-02 -9.88299310e-01 -1.35664687e-01
-2.44566336e-01 1.96459591e-01 5.69357574e-01 7.56264925e-01
-9.28753376e-01 5.26798785e-01 7.12433863e+00 1.11715579e+00
-7.24616230e-01 5.42664349e-01 6.08672857e-01 -4.71257195e-02
-1.47890568e-01 2.70499378e-01 -1.22862351e+00 8.05954337e-01
1.40888703e+00 -4.17860895e-02 6.64702415e-01 3.78426403e-01
3.51289690e-01 -7.16472328e-01 -7.56897330e-01 7.44941831e-01
-3.57536525e-01 -1.27863491e+00 -5.21374457e-02 5.03125012e-01
8.50453258e-01 4.74061489e-01 9.38488841e-02 2.45911032e-01
1.11065340e+00 -8.56101215e-01 6.75471783e-01 1.18965864e+00
5.45193732e-01 -6.40675306e-01 6.29136622e-01 5.15048325e-01
-5.50143540e-01 4.21084553e-01 -4.99586523e-01 -1.36866137e-01
4.31501001e-01 1.22335517e+00 -2.01695655e-02 2.27658734e-01
6.24887228e-01 3.12571853e-01 -1.59153357e-01 8.19414556e-01
-1.40272826e-01 9.90192235e-01 -9.94911134e-01 -1.70870036e-01
1.34447396e-01 -5.27549028e-01 7.38008320e-01 1.11532712e+00
3.06969434e-01 4.85606352e-03 -6.04191005e-01 1.65234554e+00
3.29759449e-01 -6.03396833e-01 -4.26956892e-01 3.94820757e-02
4.00813967e-01 1.11167872e+00 -7.76028395e-01 -3.49038780e-01
1.39742553e-01 4.88657534e-01 3.86110581e-02 6.59904718e-01
-1.01630056e+00 -8.38959664e-02 6.85752511e-01 -3.84519339e-01
2.27359265e-01 -3.03326756e-01 -9.73201990e-02 -1.20817697e+00
-5.44099867e-01 -3.01293671e-01 3.37828487e-01 -8.36729646e-01
-1.72165406e+00 -3.30107356e-03 6.65066838e-01 -4.24815685e-01
-5.72947145e-01 -1.12556660e+00 -6.24873340e-01 1.00488937e+00
-1.07727492e+00 -5.27060688e-01 2.97528863e-01 2.46545941e-01
-2.08888158e-01 9.70863830e-03 6.25944853e-01 1.47107795e-01
-9.63835895e-01 -4.57301438e-02 8.34071755e-01 -1.50813937e-01
1.75764978e-01 -1.27494967e+00 3.39361578e-01 7.40372837e-01
-1.92233026e-01 8.21886063e-01 1.04940701e+00 -8.57617438e-01
-1.51480484e+00 -8.16274464e-01 2.22779512e-01 -6.76676452e-01
9.40260470e-01 -4.61316884e-01 -8.32415104e-01 3.61314714e-01
-1.58894494e-01 -3.94127937e-03 5.36552131e-01 -2.94363111e-01
-3.32236975e-01 2.34988526e-01 -1.22703373e+00 5.83698213e-01
6.52490020e-01 -6.83480799e-01 -4.21617836e-01 3.84267062e-01
3.20920944e-01 -1.33528203e-01 -1.00312662e+00 2.76046246e-01
7.84678340e-01 -9.09355998e-01 1.17556179e+00 -5.69024324e-01
4.00325537e-01 -2.76985914e-01 -2.48501360e-01 -1.07029343e+00
1.88888207e-01 -5.87297678e-01 -4.72929150e-01 1.15726900e+00
1.32967144e-01 -9.30548310e-01 3.17909747e-01 7.01202929e-01
2.61401862e-01 -6.30532742e-01 -1.25069201e+00 -9.10218239e-01
6.33053660e-01 -8.91107619e-01 2.40872502e-01 5.18182993e-01
-4.37978983e-01 -1.19794473e-01 -3.16471964e-01 2.74820954e-01
1.35862350e+00 -5.64276040e-01 4.73919392e-01 -1.45669758e+00
-5.76154172e-01 -6.27020597e-01 -1.43920541e-01 -6.90694630e-01
4.35201019e-01 -8.23931992e-01 1.03957042e-01 -1.16690958e+00
6.79952443e-01 7.38486797e-02 2.24528555e-02 -1.55578196e-01
-5.95291704e-02 2.07469836e-01 -3.86570096e-01 2.34296843e-01
-1.98800504e-01 5.42852998e-01 5.30486405e-01 5.51157817e-02
4.54742283e-01 -1.06763043e-01 -1.77472621e-01 9.75838304e-01
4.89452720e-01 -8.00477147e-01 -6.45677745e-02 1.13947153e-01
5.99267364e-01 1.09321155e-01 1.11430943e+00 -7.68575251e-01
3.75864089e-01 -2.41952822e-01 1.73856899e-01 -7.64864087e-01
3.69018465e-01 -2.60690778e-01 5.89032769e-01 9.69244957e-01
-1.21221453e-01 -3.53428930e-01 2.47019947e-01 6.93042994e-01
3.36200953e-01 -5.98245740e-01 1.04399085e+00 5.99820949e-02
-1.01951420e-01 6.89449534e-02 -9.80305314e-01 1.96122870e-01
6.02494299e-01 3.69887441e-01 -4.70629305e-01 -2.79860407e-01
-7.63665080e-01 1.18290782e-01 2.69665629e-01 -6.31654024e-01
2.79765218e-01 -9.25136566e-01 -7.49799967e-01 -2.13470876e-01
-3.12674969e-01 -1.87096655e-01 6.20962858e-01 1.02375758e+00
-5.17276585e-01 2.59254366e-01 1.12124175e-01 -9.28887904e-01
-4.03715342e-01 2.61847168e-01 6.02744818e-01 -3.78122389e-01
-1.95050731e-01 6.37174726e-01 1.55188039e-01 -7.15188086e-01
-3.75661291e-02 -3.12806278e-01 2.04772517e-01 1.29920870e-01
3.09177518e-01 8.54832888e-01 -4.19691622e-01 -4.22661394e-01
-3.13841611e-01 4.72764641e-01 4.54636663e-01 -7.85510004e-01
1.15276206e+00 -1.60778865e-01 -1.79526702e-01 9.14379179e-01
9.68720913e-01 -4.31824505e-01 -1.73287690e+00 -2.77417731e-02
-9.97161865e-02 2.75785208e-01 2.29750469e-01 -6.80631042e-01
-6.53500557e-01 1.20921302e+00 6.94433630e-01 3.54376674e-01
2.26853907e-01 -4.00277041e-02 3.36709857e-01 6.78095758e-01
2.73997009e-01 -8.92833114e-01 -4.12000418e-01 4.90283430e-01
5.36972523e-01 -9.88714814e-01 3.19166929e-01 1.28527358e-01
-2.25133225e-01 9.29972470e-01 -9.58639309e-02 -4.38269168e-01
1.17338359e+00 5.69604456e-01 -5.05985618e-01 -3.47015500e-01
-7.71571815e-01 -7.60936514e-02 3.18753481e-01 3.46680194e-01
2.16906834e-02 1.15232542e-01 -2.02516779e-01 5.83286941e-01
5.91424108e-02 1.73108771e-01 5.10955036e-01 4.01243746e-01
-2.17624187e-01 -6.13542974e-01 -2.85698682e-01 6.34969056e-01
-5.47949672e-01 -1.31471187e-01 4.06852085e-03 7.59794354e-01
-1.68475002e-01 6.31875098e-01 2.02589035e-01 2.13614300e-01
-6.58081323e-02 3.24599445e-01 5.09691358e-01 -2.75361270e-01
-3.37244332e-01 5.24484366e-02 -2.10185826e-01 -2.00360835e-01
-3.54792506e-01 -1.13013852e+00 -9.85918403e-01 -6.10825539e-01
-2.85408318e-01 2.61237826e-02 9.49679494e-01 1.19652271e+00
2.21609958e-02 4.21101540e-01 1.21361883e-02 -1.06998312e+00
-1.13091564e+00 -8.60300481e-01 -8.71168554e-01 -6.09096400e-02
1.38533160e-01 -8.98037195e-01 -1.05160999e+00 -1.26574054e-01] | [6.969564914703369, 3.8650455474853516] |
92f22f3d-70c5-47b3-ae39-88657ebc15a9 | a-neural-network-multi-task-learning-approach | null | null | https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-017-1776-8 | https://bmcbioinformatics.biomedcentral.com/counter/pdf/10.1186/s12859-017-1776-8.pdf | A neural network multi-task learning approach to biomedical named entity recognition | Background
Named Entity Recognition (NER) is a key task in biomedical text mining. Accurate NER systems require task-specific, manually-annotated datasets, which are expensive to develop and thus limited in size. Since such datasets contain related but different information, an interesting question is whether it might be possible to use them together to improve NER performance. To investigate this, we develop supervised, multi-task, convolutional neural network models and apply them to a large number of varied existing biomedical named entity datasets. Additionally, we investigated the effect of dataset size on performance in both single- and multi-task settings.
Results
We present a single-task model for NER, a Multi-output multi-task model and a Dependent multi-task model. We apply the three models to 15 biomedical datasets containing multiple named entities including Anatomy, Chemical, Disease, Gene/Protein and Species. Each dataset represent a task. The results from the single-task model and the multi-task models are then compared for evidence of benefits from Multi-task Learning.
With the Multi-output multi-task model we observed an average F-score improvement of 0.8% when compared to the single-task model from an average baseline of 78.4%. Although there was a significant drop in performance on one dataset, performance improves significantly for five datasets by up to 6.3%. For the Dependent multi-task model we observed an average improvement of 0.4% when compared to the single-task model. There were no significant drops in performance on any dataset, and performance improves significantly for six datasets by up to 1.1%.
The dataset size experiments found that as dataset size decreased, the multi-output model’s performance increased compared to the single-task model’s. Using 50, 25 and 10% of the training data resulted in an average drop of approximately 3.4, 8 and 16.7% respectively for the single-task model but approximately 0.2, 3.0 and 9.8% for the multi-task model.
Conclusions
Our results show that, on average, the multi-task models produced better NER results than the single-task models trained on a single NER dataset. We also found that Multi-task Learning is beneficial for small datasets. Across the various settings the improvements are significant, demonstrating the benefit of Multi-task Learning for this task. | ['Anna Korhonen', 'Billy Chiu', 'Sampo Pyysalo', 'Gamal Crichton'] | 2017-08-15 | null | null | null | bmc-bioinformatics-2017-8 | ['multi-task-learning', 'anatomy', 'named-entity-recognition-ner'] | ['methodology', 'miscellaneous', 'natural-language-processing'] | [ 1.81606546e-01 1.14413776e-01 2.80589163e-01 -2.76219279e-01
-1.14491999e+00 -5.73142231e-01 3.43457758e-01 5.93967736e-01
-1.01240253e+00 8.65863144e-01 1.82239264e-01 -1.76937804e-01
-2.11081222e-01 -6.19644463e-01 -6.89360738e-01 -4.47172731e-01
8.47344846e-02 3.28483611e-01 9.45702791e-02 5.99825382e-02
4.79810387e-02 3.31478208e-01 -1.02786458e+00 5.17944276e-01
6.58413768e-01 4.46348369e-01 2.22310051e-01 7.62379766e-01
-3.31198305e-01 5.44900358e-01 -9.32286978e-01 -3.04260343e-01
7.62373731e-02 -6.25447035e-02 -9.60465193e-01 -4.59277272e-01
4.76254761e-01 2.21419677e-01 2.02614903e-01 6.39150620e-01
9.94495034e-01 1.07009411e-01 5.65336943e-01 -1.03243625e+00
-4.14747834e-01 5.75296640e-01 -3.24298054e-01 1.73064515e-01
2.68501103e-01 -2.01051310e-01 7.30230987e-01 -5.31544268e-01
7.91195989e-01 9.10849571e-01 1.22088301e+00 7.17405140e-01
-1.44128931e+00 -8.76890302e-01 -3.68460149e-01 -4.83850420e-01
-1.23403597e+00 -4.63309646e-01 -1.50689315e-02 -4.80552614e-01
1.33155572e+00 9.86396447e-02 5.22598810e-02 8.67368579e-01
5.75611889e-01 3.15415174e-01 1.27102172e+00 -3.95450354e-01
3.62229086e-02 3.68778020e-01 2.78396845e-01 5.19969761e-01
5.25991261e-01 -1.71759218e-01 -1.64064690e-01 -4.36815619e-01
4.42407817e-01 -8.77494216e-02 2.16679759e-02 2.49996200e-01
-1.33559513e+00 5.45913637e-01 2.02337429e-01 8.26525807e-01
-4.93412048e-01 -1.59446783e-02 6.84824049e-01 3.09564620e-01
6.66050196e-01 9.76442754e-01 -1.17279589e+00 -1.04458772e-01
-1.17229056e+00 -1.56191867e-02 1.12428248e+00 6.98271573e-01
5.74887455e-01 -1.12261727e-01 -3.75401795e-01 1.10139740e+00
-2.24204361e-01 2.06202254e-01 5.20318449e-01 -7.52978683e-01
3.01877767e-01 5.76595306e-01 1.72534678e-02 -8.22055280e-01
-1.20439029e+00 -3.55143368e-01 -6.81538522e-01 -4.91439477e-02
7.44398713e-01 -7.04453945e-01 -1.27770817e+00 1.89396036e+00
1.43562689e-01 -1.84907690e-02 2.63586462e-01 3.39501828e-01
1.10406435e+00 4.89257723e-01 7.50342131e-01 6.21453263e-02
1.66092992e+00 -6.72267973e-01 -7.91271925e-01 -1.07816428e-01
1.35579455e+00 -1.00512898e+00 7.11676836e-01 2.08163470e-01
-7.41707027e-01 -5.64160347e-01 -7.20674217e-01 -1.09824665e-01
-9.83182967e-01 3.27225864e-01 3.99095118e-01 9.16254818e-01
-1.11732841e+00 6.43567860e-01 -6.19245946e-01 -8.26583564e-01
4.40656811e-01 6.79493189e-01 -8.01670790e-01 -8.65818635e-02
-1.11931920e+00 1.26552796e+00 5.72502494e-01 -3.13594013e-01
-6.67933881e-01 -1.04278159e+00 -7.26620972e-01 1.35763079e-01
-5.99944368e-02 -6.31861985e-01 1.01646399e+00 -5.19620955e-01
-7.54150569e-01 9.34321105e-01 -2.04434127e-01 -3.16390008e-01
2.17066288e-01 -1.20198205e-01 -5.04989743e-01 -1.50332585e-01
5.34339070e-01 7.61451304e-01 -5.83555661e-02 -9.85483170e-01
-7.58514225e-01 -1.83916360e-01 -5.94134443e-02 -9.08663124e-02
-3.13991070e-01 3.61462057e-01 -1.03191857e-03 -5.02611041e-01
-4.91297394e-01 -9.82527971e-01 -3.72282386e-01 -3.50184590e-01
-3.34240675e-01 -2.65953094e-01 6.50943696e-01 -6.96860790e-01
1.10077631e+00 -1.93810511e+00 -4.12785023e-01 -3.15726489e-01
2.28286088e-01 3.83104235e-01 -3.50108862e-01 3.83541316e-01
-4.60071564e-01 6.56487763e-01 -2.10449055e-01 -2.29607761e-01
-4.88462031e-01 7.88465515e-02 5.74114323e-01 1.55008972e-01
4.50880110e-01 6.35133743e-01 -8.10493052e-01 -3.84072155e-01
1.46403667e-02 5.97162843e-01 -2.28581041e-01 -1.94363333e-02
2.40708798e-01 3.80146891e-01 -2.93807864e-01 5.08206129e-01
5.10341406e-01 -2.79821903e-01 6.58812821e-02 -3.93354565e-01
-2.30688885e-01 2.32235759e-01 -1.08973145e+00 1.57652926e+00
-6.32759929e-01 6.55319691e-01 -9.22264308e-02 -9.01619852e-01
9.00966704e-01 9.33506072e-01 7.10483789e-01 -6.39111221e-01
-7.96443820e-02 3.96723032e-01 3.78270954e-01 -4.90256190e-01
3.09479028e-01 -4.59229022e-01 -3.14008743e-01 4.69688565e-01
2.57236868e-01 3.80213886e-01 2.80887276e-01 1.58665314e-01
1.58026230e+00 -2.43847385e-01 4.36651856e-01 -5.08450091e-01
3.11969817e-01 2.13418886e-01 7.20993102e-01 7.10528970e-01
-1.71207041e-01 5.09261608e-01 5.57591319e-01 -4.83184248e-01
-9.31515753e-01 -4.28700179e-01 -4.54794317e-01 1.30651724e+00
-4.73135889e-01 -3.81387621e-01 -7.90388763e-01 -9.89221573e-01
-5.91127351e-02 6.66864395e-01 -6.90394938e-01 3.20004411e-02
-4.80480433e-01 -1.27976406e+00 1.16963625e+00 4.79167819e-01
5.11899948e-01 -1.22228873e+00 -6.85345471e-01 3.46898109e-01
-2.23771006e-01 -1.27556133e+00 -3.66321415e-01 6.32585645e-01
-9.58419919e-01 -9.78660941e-01 -9.29982841e-01 -8.16035390e-01
6.32751644e-01 -1.35349035e-01 1.05811727e+00 -9.00406670e-03
-3.93762290e-01 2.58263916e-01 -2.40659878e-01 -7.74580359e-01
-4.30476874e-01 4.82031941e-01 -1.01032570e-01 -4.73770499e-01
5.18988431e-01 -6.47834763e-02 -2.47724339e-01 2.44703323e-01
-9.85414624e-01 -2.54760593e-01 8.99246156e-01 1.18017066e+00
4.16788995e-01 -6.84509575e-02 1.06717682e+00 -1.32997882e+00
5.17939150e-01 -6.64987862e-01 2.82176156e-02 3.23304951e-01
-8.54476690e-01 2.07419902e-01 5.89195549e-01 -4.82760906e-01
-1.03156900e+00 2.46718660e-01 -1.09435797e-01 2.63938624e-02
-5.19950807e-01 5.92193305e-01 2.34819250e-03 1.41822333e-02
8.94560993e-01 -3.20800662e-01 -2.62610883e-01 -5.49668849e-01
-8.79875049e-02 9.37659800e-01 6.55657202e-02 -4.59764600e-01
2.56714910e-01 1.54172093e-01 -2.51314766e-03 -7.85763264e-01
-9.55778539e-01 -6.28443003e-01 -6.55736029e-01 2.05391929e-01
1.19384027e+00 -9.56310630e-01 -3.11244279e-01 3.51565838e-01
-1.19922721e+00 -5.16731441e-01 4.16615326e-03 5.57030737e-01
-4.65256535e-02 1.17591880e-01 -6.70056224e-01 -5.44859231e-01
-6.39802039e-01 -1.10448730e+00 1.03503823e+00 2.63541222e-01
-5.00435412e-01 -1.44994795e+00 -1.20668774e-02 3.68167698e-01
4.58409727e-01 4.69275266e-01 9.94487882e-01 -1.52044952e+00
2.41709620e-01 -1.21066988e-01 -3.01195860e-01 1.19341515e-01
5.37092924e-01 -4.56113428e-01 -9.93585944e-01 -3.30677718e-01
-2.08854094e-01 -2.68842369e-01 6.97561145e-01 3.21170002e-01
7.53363013e-01 1.97838396e-01 -7.10566938e-01 2.74691641e-01
1.42310667e+00 3.08285475e-01 5.18184483e-01 5.28832614e-01
5.87543964e-01 6.21884644e-01 5.08174121e-01 -5.25751971e-02
3.13462287e-01 3.65740567e-01 -1.43372938e-01 -5.66583753e-01
-2.61925966e-01 2.19059378e-01 1.81821734e-01 4.95517254e-01
1.07721098e-01 -1.72843948e-01 -1.29761970e+00 7.67194450e-01
-1.47357893e+00 -7.37707555e-01 -4.76047695e-01 2.17662382e+00
9.86920357e-01 -9.12217945e-02 9.11657959e-02 -1.89190045e-01
8.27101588e-01 -3.61097544e-01 -6.22430086e-01 -7.24025905e-01
5.87649904e-02 4.86732215e-01 7.66379535e-01 -1.04182763e-02
-1.27623391e+00 7.40635574e-01 6.63319063e+00 4.83657420e-01
-1.17232132e+00 1.74592063e-01 6.69941187e-01 -4.35485281e-02
2.49331504e-01 -1.56422138e-01 -1.08416760e+00 4.23722297e-01
1.61966169e+00 -3.65878761e-01 -9.94708017e-02 6.10727072e-01
1.41055718e-01 -1.48666248e-01 -1.08700657e+00 7.63887167e-01
2.82212324e-03 -1.06283057e+00 -2.14955360e-01 3.78614068e-01
7.99190938e-01 3.65918756e-01 -6.08156860e-01 4.91848350e-01
3.85413885e-01 -1.19301629e+00 7.55940750e-02 3.93731594e-01
8.69798720e-01 -5.71774840e-01 1.27441764e+00 2.97621399e-01
-1.10685372e+00 1.01482272e-01 -3.09693784e-01 2.02920943e-01
3.87102999e-02 6.54450476e-01 -8.69421601e-01 7.28465021e-01
9.67631578e-01 4.10388976e-01 -5.10901928e-01 1.25910532e+00
2.47690961e-01 5.66804230e-01 -3.41804743e-01 1.10452570e-01
8.78679752e-02 2.89083570e-01 2.67784953e-01 1.77067566e+00
3.05512518e-01 3.85461305e-03 1.57997310e-01 5.24303019e-01
-4.10416692e-01 2.17897132e-01 -5.62069654e-01 -2.23569438e-01
4.26474452e-01 1.66007149e+00 -8.67251396e-01 -3.73702049e-01
-5.55724263e-01 8.24805319e-01 2.87045836e-01 2.21852720e-01
-7.25144684e-01 -1.07612908e+00 4.46038812e-01 -1.89295739e-01
1.99052811e-01 1.13855951e-01 -4.73826736e-01 -7.61336684e-01
-3.42705816e-01 -7.46964514e-01 7.71627367e-01 -5.16504288e-01
-1.47465539e+00 6.51014805e-01 -1.07829526e-01 -6.74709082e-01
5.67401908e-02 -6.73963010e-01 -5.40751994e-01 1.17324722e+00
-1.47897875e+00 -9.66876805e-01 -2.45742381e-01 2.23625571e-01
3.81314903e-01 6.59734989e-03 1.37947547e+00 6.64203942e-01
-6.17394686e-01 8.22948694e-01 2.03186095e-01 5.47663808e-01
1.32565212e+00 -1.34696543e+00 2.48548433e-01 3.72352660e-01
-3.50569636e-02 9.80940938e-01 2.64715821e-01 -7.12840617e-01
-9.72839832e-01 -1.21312082e+00 1.32688284e+00 -6.73794985e-01
2.76050270e-01 -1.74749911e-01 -9.89504218e-01 8.52711141e-01
2.33388886e-01 -1.09073952e-01 1.29268396e+00 4.12767440e-01
-2.67274350e-01 1.07227549e-01 -1.71068573e+00 2.60232836e-01
7.08012164e-01 -2.05130860e-01 -6.94260359e-01 4.03294146e-01
5.90026200e-01 -2.69962460e-01 -1.68446445e+00 3.40856969e-01
6.55052066e-01 -4.62432534e-01 7.26261199e-01 -7.63541460e-01
2.80540079e-01 -1.48629040e-01 3.14267538e-02 -1.46325850e+00
-4.39934850e-01 -1.35930955e-01 5.44962168e-01 1.35890090e+00
9.98148084e-01 -7.85731673e-01 5.35263300e-01 9.36520755e-01
-2.68136948e-01 -5.51523030e-01 -7.40100086e-01 -7.85891235e-01
5.97891152e-01 -8.72695297e-02 2.59216189e-01 1.30112624e+00
-4.87516029e-03 6.34776354e-01 -7.16146305e-02 4.69694026e-02
2.95262039e-01 -4.55327123e-01 5.05046129e-01 -1.44625664e+00
1.10923015e-01 -1.61832035e-01 -2.93850720e-01 -3.22794348e-01
-8.05454478e-02 -1.11744261e+00 -5.87406987e-03 -2.11075616e+00
4.69716072e-01 -5.96784294e-01 -6.52807474e-01 1.21285379e+00
-3.91179621e-01 1.93273798e-01 1.68990389e-01 1.04368925e-01
-3.04111093e-01 -3.46865863e-01 6.98957264e-01 1.33165866e-01
-2.64908820e-01 -3.22330892e-01 -1.06926060e+00 4.19530571e-01
1.05291367e+00 -1.02612066e+00 1.39159663e-02 -5.76830089e-01
4.48552743e-02 -1.82229295e-01 3.74225043e-02 -1.09133625e+00
4.33190614e-01 3.14942181e-01 8.45194757e-01 -2.13627562e-01
1.12995729e-01 -5.38981557e-01 2.28038549e-01 5.81320345e-01
-4.71631885e-01 -7.92664010e-03 8.45750093e-01 3.68005693e-01
1.40426442e-01 -2.77587265e-01 7.27299929e-01 -3.18823665e-01
-5.59704423e-01 -8.56340453e-02 -4.54049259e-01 2.12514013e-01
9.76276338e-01 -2.27576748e-01 -3.87673169e-01 1.61036074e-01
-9.45055664e-01 3.09896499e-01 8.82081017e-02 3.50363970e-01
1.53174192e-01 -9.64420676e-01 -8.21790814e-01 -2.16367483e-01
8.84161890e-02 -1.51691183e-01 1.26445353e-01 9.19931948e-01
-1.65972054e-01 5.95051467e-01 -2.55337894e-01 -4.36173826e-01
-1.60326838e+00 1.98074818e-01 4.68996733e-01 -6.90740347e-01
-1.30125612e-01 4.69155163e-01 3.31485458e-02 -9.87292230e-01
-1.04383752e-01 -2.94641405e-01 -3.17577809e-01 2.23466590e-01
3.71832579e-01 6.46327674e-01 2.39439383e-01 -5.03057003e-01
-5.51100671e-01 6.06296062e-01 -3.87137115e-01 7.78602958e-02
1.47830033e+00 3.24742526e-01 -6.18151128e-02 5.45480967e-01
1.35114956e+00 -1.79122120e-01 -4.40427572e-01 1.55036062e-01
6.17349982e-01 2.46375516e-01 3.35554555e-02 -1.31250715e+00
-9.02671754e-01 6.18479609e-01 6.19317055e-01 7.86838606e-02
1.09856987e+00 -3.22286645e-03 7.07121313e-01 4.96306479e-01
3.46109688e-01 -9.24217820e-01 -2.81187683e-01 5.45123577e-01
3.95543218e-01 -1.26839316e+00 1.15232274e-01 -1.99512929e-01
-5.78579783e-01 1.10442793e+00 6.04620874e-01 1.35756448e-01
6.37021184e-01 4.29927230e-01 1.59157172e-01 -3.91941488e-01
-5.85824370e-01 -1.49711445e-01 1.48748279e-01 6.14160299e-01
1.08281255e+00 -8.14493001e-02 -4.98954177e-01 7.17899501e-01
-5.53083280e-03 3.52581739e-01 4.69677180e-01 9.85976994e-01
-1.97999194e-01 -1.32785523e+00 -2.60098457e-01 8.47125351e-01
-1.14473367e+00 -2.44453937e-01 -3.46257299e-01 1.02471602e+00
2.53118545e-01 1.22313511e+00 -3.79308388e-02 -3.15222025e-01
6.66066110e-01 5.42779565e-01 1.59741845e-02 -9.82680082e-01
-1.26884091e+00 -3.40722390e-02 5.44099629e-01 -3.65424842e-01
-5.16543627e-01 -6.24184251e-01 -1.63397539e+00 -5.06277103e-03
-5.55066049e-01 3.23405921e-01 8.55820537e-01 8.26775014e-01
8.12509835e-01 7.10316300e-01 9.12058949e-02 -2.87454545e-01
-2.62597054e-01 -1.18805087e+00 -3.46510828e-01 2.88667768e-01
-2.75456416e-03 -3.73898864e-01 -3.38373661e-01 1.83180004e-01] | [8.489190101623535, 8.78288459777832] |
c7abd68c-c102-44a7-ad44-dd9d014a52ea | reinforcement-learning-with-partial | 2304.13223 | null | https://arxiv.org/abs/2304.13223v1 | https://arxiv.org/pdf/2304.13223v1.pdf | Reinforcement Learning with Partial Parametric Model Knowledge | We adapt reinforcement learning (RL) methods for continuous control to bridge the gap between complete ignorance and perfect knowledge of the environment. Our method, Partial Knowledge Least Squares Policy Iteration (PLSPI), takes inspiration from both model-free RL and model-based control. It uses incomplete information from a partial model and retains RL's data-driven adaption towards optimal performance. The linear quadratic regulator provides a case study; numerical experiments demonstrate the effectiveness and resulting benefits of the proposed method. | ['R. Bhushan Gopaluni', 'Michael G. Forbes', 'Nathan P. Lawrence', 'Philip D. Loewen', 'Shuyuan Wang'] | 2023-04-26 | null | null | null | null | ['continuous-control'] | ['playing-games'] | [-1.56605393e-01 5.03955364e-01 -7.94991970e-01 2.56079108e-01
-7.45551288e-01 -3.91567588e-01 4.70853031e-01 -3.62291753e-01
-5.06444454e-01 1.71088016e+00 -1.84112057e-01 -3.54296029e-01
-4.37430948e-01 -3.80418807e-01 -8.35381031e-01 -8.89833391e-01
-2.15088174e-01 8.00409093e-02 -3.17329690e-02 -5.90449274e-01
2.38449708e-01 1.88634485e-01 -7.19946444e-01 -3.10609937e-01
8.62518609e-01 7.83651114e-01 5.22263288e-01 6.12047255e-01
3.85017365e-01 1.35018337e+00 -2.66248286e-01 5.14003217e-01
6.86950982e-01 -2.18246475e-01 -4.83267277e-01 8.12262204e-03
-4.69277352e-01 -5.59718251e-01 -6.07190073e-01 9.28866744e-01
6.06791735e-01 2.17079461e-01 7.80181140e-02 -1.26508975e+00
-3.68245155e-01 4.33569402e-01 -4.44129080e-01 1.19894966e-02
2.56627113e-01 7.60864735e-01 5.43624401e-01 -3.89555514e-01
5.08918226e-01 1.51932216e+00 5.90737581e-01 6.54373884e-01
-1.11268616e+00 -5.47498107e-01 3.59300375e-01 7.96580464e-02
-1.28189850e+00 -3.63594145e-01 6.25779390e-01 -2.71578878e-01
1.06864727e+00 -1.01703949e-01 8.57885122e-01 8.88481319e-01
4.64020997e-01 8.96816432e-01 1.64639771e+00 -4.30488288e-01
5.74921846e-01 2.51680195e-01 -4.60505843e-01 7.41089463e-01
4.30850714e-01 1.19638097e+00 -3.34837943e-01 -1.76927865e-01
1.09549487e+00 -2.98276395e-01 -3.90051425e-01 -8.43828380e-01
-9.83120739e-01 7.17691779e-01 3.36875916e-01 -1.80535838e-01
-8.12835634e-01 5.42881906e-01 -1.86001435e-02 9.19805527e-01
-9.89102945e-02 9.55862522e-01 -7.53462553e-01 -2.74589568e-01
-4.12151873e-01 4.45618272e-01 9.67354715e-01 9.99380887e-01
7.34854639e-01 9.15782094e-01 7.05389725e-03 2.55682856e-01
4.22060788e-01 9.33911920e-01 6.58677459e-01 -1.66242540e+00
2.64419794e-01 3.94955099e-01 1.12392497e+00 -6.05102360e-01
-4.01479721e-01 -4.78767395e-01 -2.58465379e-01 7.78627455e-01
5.70576936e-02 -9.42083120e-01 -6.75792396e-01 1.90779674e+00
2.58452177e-01 1.93681836e-01 6.46278322e-01 8.68236184e-01
-1.11001059e-01 5.84527433e-01 -2.01210499e-01 -8.22563112e-01
3.74392658e-01 -8.02547395e-01 -9.18802738e-01 -3.82126212e-01
3.20714504e-01 -1.07818738e-01 7.45341182e-01 6.31176233e-01
-1.05151713e+00 -4.42188501e-01 -1.34766221e+00 7.86986113e-01
-1.90778345e-01 -1.58340380e-01 2.68645942e-01 4.24799800e-01
-1.18520188e+00 9.53801274e-01 -9.57877517e-01 3.68879326e-02
8.28928500e-02 8.37484896e-01 5.30306948e-03 2.77754515e-01
-1.44114125e+00 1.30384302e+00 5.65353155e-01 7.08294213e-02
-1.27816331e+00 -5.86474895e-01 -6.44817829e-01 -3.81577730e-01
1.30328798e+00 -5.77201843e-01 1.80420315e+00 -1.14616358e+00
-2.34494162e+00 -1.09705143e-01 8.38643089e-02 -8.10029268e-01
5.95693886e-01 -5.15607119e-01 -1.02293149e-01 7.97312707e-02
-2.12049678e-01 3.17077667e-01 1.08519268e+00 -1.67712855e+00
-5.89885890e-01 -6.67973980e-02 1.14632696e-01 3.83753419e-01
2.43307650e-01 -5.73283136e-01 2.41768375e-01 -2.59766012e-01
-3.84902418e-01 -1.01958251e+00 -7.48674572e-01 -2.10433736e-01
-1.89457208e-01 2.12053120e-01 8.10951173e-01 -4.93863136e-01
1.23013198e+00 -1.72645986e+00 9.95339751e-02 3.64742041e-01
-9.75531936e-02 4.50174510e-01 -2.02061504e-01 7.95144320e-01
1.82168484e-01 -2.02741906e-01 6.07204065e-02 1.10094130e-01
1.78267777e-01 7.46753693e-01 -4.18912977e-01 3.34141612e-01
1.32144123e-01 1.00395453e+00 -1.26246130e+00 -2.91889578e-01
2.89686352e-01 1.35860443e-01 -2.46154130e-01 3.34763497e-01
-6.02698267e-01 6.31587505e-01 -1.20136118e+00 3.89068902e-01
8.16221312e-02 8.56473297e-03 5.63307822e-01 1.71255112e-01
-5.97572803e-01 -3.49596411e-01 -1.33011401e+00 1.33024752e+00
-6.24819279e-01 -2.15368606e-02 6.08979166e-01 -1.06089246e+00
7.84238040e-01 5.10322571e-01 6.99736178e-01 -9.56645131e-01
1.31297916e-01 2.42456481e-01 -1.93160325e-02 -7.12419271e-01
1.70269057e-01 -1.79896176e-01 -8.97269044e-03 1.51997373e-01
-1.42578438e-01 -5.58640957e-01 -1.12045966e-01 -1.86661184e-01
1.07694769e+00 7.69642055e-01 1.01710117e+00 -3.79212499e-01
4.68290299e-01 3.10250103e-01 9.50890660e-01 1.14362454e+00
-5.74326158e-01 -3.03020597e-01 3.71256143e-01 -2.88908988e-01
-9.13858891e-01 -7.67047524e-01 4.22026604e-01 9.08965766e-01
1.66784853e-01 1.64019763e-01 -2.94598758e-01 -4.23400909e-01
3.90508413e-01 8.28965902e-01 -5.68407774e-01 -3.89579475e-01
-6.36819243e-01 -2.61900544e-01 5.96774463e-03 3.31284553e-01
4.95811880e-01 -9.55834746e-01 -7.59610593e-01 5.44777811e-01
3.77547622e-01 -7.13612735e-01 -1.40661120e-01 3.32791060e-01
-6.99930906e-01 -1.26279557e+00 -3.85685742e-01 -4.47385699e-01
4.29112613e-01 4.66718338e-02 6.34006977e-01 -2.21406043e-01
1.09977104e-01 1.05766666e+00 5.93835935e-02 -6.91596687e-01
-6.56875908e-01 -4.81840938e-01 6.59998775e-01 -1.51666671e-01
-3.13282281e-01 -3.38435888e-01 -4.01708603e-01 2.15442866e-01
-3.61224055e-01 -2.02584967e-01 8.62967610e-01 1.12239695e+00
7.31267214e-01 2.34152064e-01 9.53805923e-01 -6.59688056e-01
1.05685198e+00 -3.77758592e-01 -1.29092288e+00 2.98296005e-01
-1.12134576e+00 3.68845701e-01 8.35056782e-01 -4.42819983e-01
-1.46546662e+00 3.93342853e-01 5.66036463e-01 -7.01250076e-01
1.66975230e-01 3.85721654e-01 -3.06321383e-02 -5.95108747e-01
5.75665057e-01 4.40496564e-01 5.82737744e-01 -3.28670591e-01
5.06948054e-01 3.90477359e-01 4.88425285e-01 -7.62362361e-01
7.52146304e-01 1.88050628e-01 2.63679057e-01 -6.25218511e-01
-6.63154244e-01 -2.36028999e-01 -4.61136937e-01 -2.09160879e-01
3.83579224e-01 -1.06048274e+00 -1.06428957e+00 2.69717515e-01
-5.92208624e-01 -1.10062349e+00 -7.27499843e-01 6.18099391e-01
-1.42037439e+00 9.42863077e-02 -2.20165521e-01 -1.43960297e+00
-1.25860557e-01 -8.28006923e-01 3.76530081e-01 3.29501808e-01
1.67068198e-01 -1.08052754e+00 5.65877855e-01 -2.81307608e-01
5.68860948e-01 4.71920609e-01 5.22063434e-01 -3.48563612e-01
-5.60227931e-01 1.09756580e-02 3.63210231e-01 4.41987038e-01
4.85137813e-02 -4.57380563e-01 -5.31896055e-01 -9.13745582e-01
3.08177561e-01 -8.19314122e-01 4.86574829e-01 3.88079673e-01
3.97983551e-01 -8.85830402e-01 -2.73116887e-01 3.89989503e-02
1.88640904e+00 6.23534262e-01 1.65565699e-01 6.28982544e-01
3.85674655e-01 2.20739543e-01 1.04193211e+00 9.20667350e-01
3.67190808e-01 5.10400951e-01 7.08423555e-01 4.36709775e-03
2.82111913e-01 -5.78906238e-01 7.03115523e-01 5.25309265e-01
-5.10546267e-02 4.00425106e-01 -5.69095016e-01 4.29973215e-01
-2.32451868e+00 -9.70899224e-01 5.17995238e-01 2.39289999e+00
1.13843429e+00 8.18311125e-02 1.14629842e-01 -2.46168807e-01
5.14248610e-01 -2.00860798e-01 -1.35110366e+00 -5.03796160e-01
5.13938628e-02 -1.29982352e-01 9.73845601e-01 8.04296732e-01
-8.25938284e-01 9.73939836e-01 7.43796635e+00 7.86100686e-01
-8.29440773e-01 -1.70340076e-01 1.02446854e-01 -4.21788804e-02
2.14321941e-01 1.92005038e-01 -6.51721776e-01 1.94323555e-01
1.01381695e+00 -5.74433148e-01 1.08235013e+00 9.42676246e-01
9.95680630e-01 -3.80688161e-01 -7.05602288e-01 4.81291652e-01
-4.99334425e-01 -9.30769324e-01 -5.68201005e-01 5.76715693e-02
1.17142582e+00 3.40378401e-03 6.45512789e-02 8.76348853e-01
1.17027438e+00 -7.16832280e-01 7.42151856e-01 9.46082175e-01
4.09967124e-01 -7.93934166e-01 6.05767846e-01 7.48929143e-01
-9.18427944e-01 -9.49811697e-01 -2.90603369e-01 -5.32770932e-01
-4.05010022e-02 -5.90759367e-02 -9.13647175e-01 3.84275675e-01
5.14248423e-02 4.78857219e-01 -1.13775827e-01 9.33830440e-01
-5.89244545e-01 5.45833349e-01 -2.93162167e-01 -2.48047784e-01
5.70771456e-01 -2.04029754e-01 8.12338591e-01 5.92949629e-01
-6.18380457e-02 1.38285890e-01 7.98631966e-01 7.45128810e-01
6.10836387e-01 -2.76053607e-01 -8.79302621e-01 -1.15112064e-03
4.87718195e-01 6.93169236e-01 1.90147057e-01 -2.26487041e-01
-2.14288786e-01 2.32776225e-01 2.92120904e-01 6.85391366e-01
-4.91881013e-01 -2.06465155e-01 5.88266969e-01 -4.55951244e-01
4.96283352e-01 -4.90089804e-01 1.53346896e-01 -7.10918248e-01
-3.39278668e-01 -1.08524930e+00 -2.68356632e-02 -5.77100277e-01
-9.59418356e-01 -1.05814017e-01 1.26464799e-01 -1.26764441e+00
-8.58226180e-01 -4.34011221e-01 -2.53651500e-01 5.05939841e-01
-1.91207075e+00 -8.84641588e-01 3.90783042e-01 6.16173089e-01
4.55173641e-01 -2.51295596e-01 6.31909192e-01 -5.00582933e-01
-5.39156139e-01 1.26122758e-01 8.35133791e-01 -3.97806168e-01
4.19325322e-01 -1.49830568e+00 -1.71469465e-01 4.04797375e-01
-6.60229325e-01 3.22309911e-01 9.04183984e-01 -6.90284252e-01
-2.14402533e+00 -1.06734419e+00 3.20769213e-02 -1.38847575e-01
9.75701928e-01 3.27292144e-01 -5.10991752e-01 5.94570160e-01
9.13728550e-02 -2.34758034e-01 -1.65687710e-01 -5.40943742e-01
1.32085413e-01 -3.04499447e-01 -1.16933858e+00 6.87035978e-01
6.05130017e-01 -1.82800218e-01 -6.24280214e-01 5.73291220e-02
9.08666372e-01 -4.92461324e-01 -9.41394091e-01 3.51433516e-01
4.35315788e-01 -4.11404103e-01 8.54934692e-01 -9.30126369e-01
-3.35083008e-01 -2.92539090e-01 -1.64498948e-02 -1.88186800e+00
-3.45276475e-01 -1.52622676e+00 -5.15226901e-01 5.74152768e-01
2.43873313e-01 -7.94443905e-01 6.11123204e-01 5.63876867e-01
-4.93542589e-02 -8.13002169e-01 -1.02079129e+00 -1.32086289e+00
2.49910310e-01 -2.24717915e-01 3.08915228e-01 7.20452249e-01
4.45881873e-01 -9.81065538e-03 -8.45188916e-01 3.29029232e-01
7.94697821e-01 7.85971209e-02 7.58399367e-01 -4.71756279e-01
-6.94586515e-01 -1.01663783e-01 2.07564011e-01 -8.79251659e-01
3.03764373e-01 -8.37717429e-02 3.45425755e-01 -1.38006532e+00
-2.29322448e-01 -3.00842583e-01 -6.28903985e-01 6.78312540e-01
-5.19864261e-02 -5.17518878e-01 3.74907523e-01 8.38056877e-02
-1.07547855e+00 9.13094461e-01 1.67299092e+00 -1.01397015e-01
-8.10875535e-01 3.70351017e-01 -6.94783807e-01 8.47074866e-01
1.18079638e+00 -3.19925785e-01 -8.31780255e-01 -1.20059386e-01
1.43875867e-01 7.22134829e-01 1.18769020e-01 -1.04163527e+00
2.05520540e-01 -9.68643010e-01 -3.70527394e-02 4.04366404e-02
2.86140442e-01 -9.56103683e-01 1.47508634e-02 1.24713397e+00
-5.48774183e-01 -1.07672460e-01 2.64675736e-01 1.08291662e+00
3.29419002e-02 -8.40257332e-02 8.59550774e-01 -5.04363656e-01
-1.04482818e+00 7.59009868e-02 -3.48955333e-01 2.29229450e-01
1.02347767e+00 1.31118700e-01 -2.77896106e-01 -7.58646846e-01
-9.26949322e-01 8.65066230e-01 1.78906515e-01 1.62339404e-01
7.15041816e-01 -1.03719127e+00 -3.44733477e-01 1.62186120e-02
-4.31780398e-01 -5.68723977e-01 -1.93512872e-01 7.43386090e-01
4.66127060e-02 8.26726079e-01 -1.51299775e-01 -1.13264702e-01
-4.78897691e-01 8.55986476e-01 6.90583825e-01 -4.40968424e-01
-2.26588860e-01 1.50467172e-01 -1.03683971e-01 -4.37559128e-01
2.88136035e-01 -1.99691355e-01 -1.24436989e-01 -6.81703210e-01
4.30908293e-01 7.42303610e-01 -4.33350116e-01 -8.62400308e-02
1.74133152e-01 4.60604548e-01 2.44160503e-01 -4.72121596e-01
1.14138293e+00 -5.44859111e-01 3.45037997e-01 7.37774670e-01
5.44877112e-01 -4.36432838e-01 -2.20427847e+00 -4.27382022e-01
1.12409942e-01 -3.09532762e-01 2.53361970e-01 -1.18875015e+00
-6.05545878e-01 3.21518630e-01 6.29706860e-01 -4.68536653e-02
7.60570884e-01 -7.50119448e-01 4.81698453e-01 1.10270166e+00
9.21909630e-01 -1.56801260e+00 2.36289173e-01 6.87332392e-01
9.56223726e-01 -1.36077654e+00 2.74824500e-01 2.27983519e-01
-9.76162016e-01 7.67591536e-01 1.02625728e+00 -4.39453274e-01
6.33670270e-01 2.65312314e-01 1.08498000e-01 2.57733315e-01
-1.33306122e+00 -3.86653453e-01 -3.02131742e-01 9.74531233e-01
-5.54138720e-01 8.41493532e-02 -2.35338092e-01 5.00921965e-01
3.10997993e-01 5.20582438e-01 5.29985189e-01 1.40950346e+00
-9.13779318e-01 -1.05470479e+00 -5.13136268e-01 1.48506030e-01
-1.13752805e-01 4.34348971e-01 -2.46283799e-01 1.23794448e+00
-1.65567040e-01 1.09495187e+00 -5.22002399e-01 -7.14116497e-03
4.79921430e-01 -1.74367145e-01 6.08495414e-01 -3.65468591e-01
-2.94030279e-01 1.99354723e-01 1.18930839e-01 -1.11974299e+00
-2.82246292e-01 -2.34096333e-01 -1.45648086e+00 6.56175017e-02
-3.54025453e-01 1.84703097e-01 6.07546985e-01 9.72149074e-01
4.13139522e-01 3.66013259e-01 1.07881749e+00 -7.58125842e-01
-1.68670762e+00 -5.70215523e-01 -6.89673603e-01 -1.88879490e-01
9.21472490e-01 -8.97988439e-01 -3.76775563e-01 -3.05836678e-01] | [4.262612342834473, 2.1694116592407227] |
cbd1ee25-bd85-4342-81d8-c2fa4d73a1ee | tempoqr-temporal-question-reasoning-over | 2112.05785 | null | https://arxiv.org/abs/2112.05785v1 | https://arxiv.org/pdf/2112.05785v1.pdf | TempoQR: Temporal Question Reasoning over Knowledge Graphs | Knowledge Graph Question Answering (KGQA) involves retrieving facts from a Knowledge Graph (KG) using natural language queries. A KG is a curated set of facts consisting of entities linked by relations. Certain facts include also temporal information forming a Temporal KG (TKG). Although many natural questions involve explicit or implicit time constraints, question answering (QA) over TKGs has been a relatively unexplored area. Existing solutions are mainly designed for simple temporal questions that can be answered directly by a single TKG fact. This paper puts forth a comprehensive embedding-based framework for answering complex questions over TKGs. Our method termed temporal question reasoning (TempoQR) exploits TKG embeddings to ground the question to the specific entities and time scope it refers to. It does so by augmenting the question embeddings with context, entity and time-aware information by employing three specialized modules. The first computes a textual representation of a given question, the second combines it with the entity embeddings for entities involved in the question, and the third generates question-specific time embeddings. Finally, a transformer-based encoder learns to fuse the generated temporal information with the question representation, which is used for answer predictions. Extensive experiments show that TempoQR improves accuracy by 25--45 percentage points on complex temporal questions over state-of-the-art approaches and it generalizes better to unseen question types. | ['George Karypis', 'Nagib Hakim', 'Tetiana Grinberg', 'Phillip R. Howard', 'Soji Adeshina', 'Vassilis N. Ioannidis', 'Prasanna Lakkur Subramanyam', 'Costas Mavromatis'] | 2021-12-10 | null | null | null | null | ['graph-question-answering'] | ['graphs'] | [-2.33815864e-01 5.48039496e-01 -3.17428917e-01 -3.50605667e-01
-7.50940323e-01 -7.50835061e-01 5.83501399e-01 6.93207979e-01
-1.12329811e-01 6.38214409e-01 4.96019006e-01 -5.25462031e-01
-5.20566344e-01 -1.43091178e+00 -6.08948112e-01 -1.68652266e-01
-2.40228564e-01 6.29650235e-01 8.76628518e-01 -6.81526780e-01
-2.94044297e-02 7.43321776e-02 -1.47008789e+00 5.05212426e-01
9.57969129e-01 1.35110652e+00 -2.28602007e-01 7.19610989e-01
-7.93947339e-01 1.63794541e+00 -5.54659963e-01 -8.31859887e-01
-2.31269553e-01 -3.20325524e-01 -1.66444468e+00 -2.96888649e-01
3.28923166e-01 -2.33170941e-01 -7.94261575e-01 4.54594791e-01
6.57319576e-02 5.34204364e-01 2.15373412e-01 -1.43779981e+00
-1.00559294e+00 4.42862749e-01 2.53932148e-01 6.69336617e-01
1.01598394e+00 -2.43658513e-01 1.56871855e+00 -5.28697312e-01
7.76642263e-01 1.11943626e+00 4.73582655e-01 3.31067234e-01
-7.08575487e-01 -4.59530577e-03 3.90686691e-01 1.03029728e+00
-1.02746403e+00 2.39517599e-01 8.14009011e-01 -1.39055640e-01
1.53894722e+00 2.88566768e-01 7.75353491e-01 6.69818103e-01
3.46978873e-01 6.62551105e-01 7.17896700e-01 -1.86830744e-01
2.40943909e-01 -1.00701638e-01 6.12315655e-01 9.45540369e-01
-2.17820510e-01 -1.75511986e-01 -7.22885728e-01 -1.60453588e-01
3.07484388e-01 -1.77703276e-01 -2.21982047e-01 -1.57120928e-01
-1.19394672e+00 1.09168255e+00 6.47245407e-01 6.11330032e-01
-4.39073145e-01 3.65782738e-01 5.12758434e-01 6.17433131e-01
3.78983915e-01 4.20768082e-01 -8.20238888e-01 -3.85523587e-02
-4.97556001e-01 5.97689092e-01 1.22531760e+00 8.96776080e-01
9.14342701e-01 -3.27136010e-01 -5.31819224e-01 3.37616920e-01
4.18341644e-02 7.00969398e-02 4.98640746e-01 -9.14011240e-01
6.02162778e-01 1.30748725e+00 1.64553002e-01 -1.25821376e+00
-3.67195308e-01 -1.11168243e-01 -3.92759554e-02 -7.38744974e-01
3.18612903e-01 6.46195859e-02 -7.12914407e-01 1.76364088e+00
8.65376055e-01 3.10643464e-01 2.05752105e-01 6.86371744e-01
1.31555510e+00 1.03626788e+00 7.27173612e-02 -1.67787522e-01
1.85554683e+00 -8.71940255e-01 -1.06524456e+00 -1.10244475e-01
6.92589939e-01 -1.29176691e-01 8.38460922e-01 -3.65397781e-02
-7.92923510e-01 -2.87316352e-01 -7.48481095e-01 -7.79532254e-01
-1.12209666e+00 -2.95123816e-01 7.56727457e-01 1.58058360e-01
-1.06594825e+00 4.26250666e-01 -3.78908485e-01 -6.17137551e-01
-1.15272757e-02 3.62555087e-01 -2.65962243e-01 -2.04526424e-01
-2.06164980e+00 1.15957773e+00 7.66947985e-01 -4.54673618e-02
-6.20961607e-01 -9.13112342e-01 -1.14038634e+00 8.60515237e-02
9.32399392e-01 -1.02825212e+00 1.28935826e+00 -1.33055039e-02
-1.22166610e+00 7.20172405e-01 -4.44833279e-01 -7.50462174e-01
-4.70716096e-02 -1.43997326e-01 -1.08747196e+00 6.83238566e-01
3.72052372e-01 4.11762625e-01 7.30483353e-01 -8.24753165e-01
-6.10117733e-01 -3.94481003e-01 9.98061240e-01 1.20883226e-01
-1.73538998e-01 -1.55221939e-01 -6.54094100e-01 -4.39644217e-01
1.02288298e-01 -5.29438078e-01 9.83096883e-02 -2.29308218e-01
-1.11188598e-01 -9.63486373e-01 1.09982395e+00 -1.08861160e+00
1.73029995e+00 -1.58267438e+00 1.92875266e-01 -1.04416445e-01
2.01698959e-01 -2.60923672e-02 1.74006056e-02 1.09013259e+00
-2.05671668e-01 -1.78709641e-01 -6.41778335e-02 2.59791702e-01
2.97514826e-01 7.13916421e-01 -9.11179006e-01 -3.92394550e-02
3.53117496e-01 1.44353449e+00 -1.31319976e+00 -7.24289060e-01
7.80152828e-02 1.75492726e-02 -5.80559134e-01 2.00176448e-01
-1.10173917e+00 -9.93485749e-02 -7.13767648e-01 6.40799224e-01
7.28537664e-02 -4.11803424e-01 1.79375365e-01 -4.75586832e-01
2.01068074e-01 6.08811200e-01 -8.76115680e-01 1.70474041e+00
-5.37163258e-01 3.91268909e-01 -7.02390373e-01 -9.87943351e-01
8.54892910e-01 7.24544585e-01 4.62493509e-01 -7.03774810e-01
-5.58480956e-02 3.79313268e-02 -5.96739948e-01 -1.08872128e+00
8.08638513e-01 -2.03855947e-01 -3.60486805e-01 2.94414878e-01
4.39853340e-01 -4.29253459e-01 4.18405563e-01 7.17459202e-01
1.40619159e+00 2.16959849e-01 3.96637470e-01 1.17338896e-01
6.88905239e-01 3.32336873e-01 3.97657990e-01 1.16703644e-01
-2.46054530e-01 -7.51601085e-02 7.63625979e-01 -6.70196176e-01
-4.23232138e-01 -1.17395782e+00 3.52624804e-01 9.85511541e-01
1.76606134e-01 -8.96760046e-01 -3.06297421e-01 -1.16889799e+00
3.62020046e-01 1.03203404e+00 -1.08540928e+00 -3.60533148e-01
-6.62871301e-01 8.17745030e-02 4.10778672e-01 6.24894261e-01
5.11538625e-01 -1.07829773e+00 -5.09032845e-01 4.00306046e-01
-7.48094201e-01 -1.32384896e+00 -4.21411961e-01 -4.64472063e-02
-7.95702636e-01 -1.47041667e+00 -9.30205584e-02 -8.44639480e-01
2.45966211e-01 2.56351195e-02 1.38686907e+00 9.25198477e-03
-1.02203384e-01 1.03731203e+00 -6.89800143e-01 -2.96756923e-01
1.56882048e-01 -1.29339695e-01 -1.54090494e-01 2.06720352e-01
5.85901916e-01 -6.22855961e-01 -5.68341434e-01 6.69545680e-02
-1.14791071e+00 -5.20119846e-01 -1.89437494e-02 5.03921270e-01
5.56419373e-01 2.54653662e-01 8.95107985e-01 -7.66402066e-01
7.73240805e-01 -9.08896327e-01 -4.72675264e-01 7.82352149e-01
-4.18359786e-01 4.35715050e-01 6.58314526e-01 -4.12117720e-01
-1.24978864e+00 -5.39603770e-01 5.12604639e-02 -3.65816742e-01
3.55935484e-01 9.67186332e-01 7.71618336e-02 2.28410825e-01
6.38697088e-01 2.66412824e-01 -5.22921383e-01 -4.25358489e-02
1.00969231e+00 2.98108403e-02 6.49069130e-01 -9.70112026e-01
9.82518852e-01 4.18625265e-01 9.73920617e-03 -5.32917678e-01
-1.26834595e+00 -6.84906363e-01 -3.40820372e-01 -3.23628306e-01
1.07922173e+00 -5.67547977e-01 -1.00043786e+00 -3.04301769e-01
-1.15938914e+00 -1.16967879e-01 -7.87582874e-01 2.26083606e-01
-6.50376976e-01 4.31702107e-01 -5.36035597e-01 -7.05705702e-01
-3.07880968e-01 -3.31767052e-01 1.06674993e+00 1.41216889e-01
-3.25838983e-01 -1.37056100e+00 2.91498154e-01 7.62521386e-01
2.17755049e-01 4.85339075e-01 1.36015511e+00 -6.43844068e-01
-9.26334023e-01 -1.92526072e-01 -1.03474729e-01 -9.43196192e-02
2.25924388e-01 -3.37580681e-01 -7.31476247e-01 8.62020850e-02
8.35740045e-02 -4.25619870e-01 6.61121607e-01 -5.39129935e-02
8.80576968e-01 -5.63901126e-01 -2.56207466e-01 3.42799984e-02
1.52754569e+00 2.65140861e-01 6.47509575e-01 2.18881458e-01
3.79651070e-01 7.15886772e-01 8.41280758e-01 2.57174551e-01
1.21742666e+00 5.22612989e-01 5.12641490e-01 7.20170856e-01
-3.43164690e-02 -6.41811490e-01 2.18365535e-01 9.72289920e-01
1.51198804e-01 -2.31504083e-01 -8.27486038e-01 1.22858047e+00
-1.89348304e+00 -1.23526978e+00 -2.37052694e-01 1.69629800e+00
1.10244668e+00 -2.58012235e-01 -4.86337692e-02 3.27891618e-01
2.03789145e-01 4.41693038e-01 -6.39295280e-01 -3.99868131e-01
1.24006188e-02 4.36412245e-01 6.19258247e-02 5.71141362e-01
-7.67129183e-01 1.10018289e+00 5.51002836e+00 5.68680942e-01
-5.14828086e-01 2.11423397e-01 -2.76353121e-01 2.26423711e-01
-7.31522679e-01 4.00582910e-01 -5.04362106e-01 -2.99426224e-02
1.02387607e+00 -8.08444679e-01 3.99322599e-01 4.56649810e-01
-2.04092547e-01 -1.62343040e-01 -1.16871464e+00 7.94663250e-01
1.13048986e-01 -1.49616492e+00 2.27476045e-01 -4.62319016e-01
6.11442208e-01 -5.14597714e-01 -2.24311739e-01 8.59757423e-01
2.20800593e-01 -6.51001632e-01 5.82199335e-01 8.60195458e-01
4.68819499e-01 -5.62090039e-01 5.96974313e-01 1.70199275e-01
-1.81472492e+00 -3.48324209e-01 -1.76050410e-01 8.78041014e-02
2.40986675e-01 4.85356361e-01 -9.53419030e-01 1.51078856e+00
8.17747593e-01 4.91735280e-01 -6.99131489e-01 6.39358878e-01
-5.68253577e-01 5.14765441e-01 -1.62538439e-01 -1.53958574e-01
3.41223598e-01 -2.00982336e-02 3.60845327e-01 7.56419182e-01
2.84567028e-01 6.66302860e-01 -1.30015656e-01 6.95306838e-01
-1.08739980e-01 5.56404777e-02 -6.74851894e-01 -3.19953978e-01
3.16128939e-01 1.09230816e+00 -4.03147310e-01 -4.98118967e-01
-4.72279549e-01 1.07578969e+00 4.31631684e-01 5.70283115e-01
-1.02080882e+00 -5.49977958e-01 4.33929533e-01 -9.39204022e-02
5.41841269e-01 -1.44958004e-01 2.98707962e-01 -1.34285641e+00
3.83715183e-01 -5.59603930e-01 1.29368925e+00 -1.10768509e+00
-1.14158487e+00 5.10767937e-01 2.54494786e-01 -9.41738844e-01
-4.21050757e-01 -3.78904551e-01 -5.71573615e-01 7.22553432e-01
-1.74692249e+00 -1.19949567e+00 -3.55559707e-01 1.06769943e+00
2.48093650e-01 2.90159076e-01 8.90190959e-01 2.17108056e-01
-1.67107597e-01 1.26997724e-01 -6.98115051e-01 -1.82938036e-02
2.78695762e-01 -1.53625810e+00 2.03267410e-01 6.27763629e-01
2.72111118e-01 6.37483537e-01 5.83825588e-01 -7.20264196e-01
-1.77379775e+00 -1.38154495e+00 1.77434373e+00 -8.16237867e-01
1.13488245e+00 1.13176614e-01 -1.09070814e+00 1.08551908e+00
4.06413734e-01 1.56698361e-01 6.95899427e-01 6.00837767e-02
-7.47894645e-01 -2.99630582e-01 -1.04975474e+00 4.88138616e-01
9.21502888e-01 -1.12872803e+00 -1.59394085e+00 5.73964238e-01
1.52273607e+00 -3.74196351e-01 -1.33186412e+00 1.45383462e-01
-4.72546322e-03 -5.63400328e-01 8.72927248e-01 -1.07160115e+00
2.11006269e-01 -6.29931748e-01 -3.16706240e-01 -1.05295193e+00
-5.78203760e-02 -5.55696785e-01 -9.24503028e-01 1.13416505e+00
2.51898855e-01 -6.22968137e-01 6.61266685e-01 5.07520616e-01
-5.39029762e-02 -9.23079014e-01 -1.01909971e+00 -8.34337115e-01
-1.26917288e-01 -5.60897470e-01 8.53468299e-01 1.18684494e+00
4.40941125e-01 4.87481534e-01 4.84799035e-02 4.97587025e-01
1.72113642e-01 6.02339208e-01 2.50825077e-01 -1.22381902e+00
5.38501777e-02 1.12411372e-01 -3.73846292e-01 -8.45307171e-01
2.31251851e-01 -1.00868714e+00 -3.38046432e-01 -2.33716273e+00
-5.36331475e-01 1.51352569e-01 3.79754673e-03 5.88709950e-01
-2.62825727e-01 -4.40939963e-01 -4.97415774e-02 -4.56104219e-01
-8.64599168e-01 7.78583407e-01 1.24695480e+00 -3.12597752e-01
-3.08485162e-02 -3.44494134e-01 -5.24442255e-01 2.34994024e-01
6.09777749e-01 -2.79786348e-01 -9.80474949e-01 -3.22181135e-01
7.70328999e-01 6.73449397e-01 4.38363910e-01 -7.91965783e-01
7.28448927e-01 -2.57616460e-01 -2.86743999e-01 -8.07442784e-01
5.41460216e-01 -8.98943961e-01 3.59336548e-02 2.22818539e-01
-1.54718950e-01 2.69686520e-01 3.60905141e-01 8.14655483e-01
-7.45860219e-01 -9.49253212e-04 5.02129793e-02 -3.46468464e-02
-1.10449076e+00 5.75618684e-01 1.43457493e-02 5.63971817e-01
1.15811551e+00 -1.32661313e-02 -4.37441975e-01 -6.90448701e-01
-8.99087667e-01 7.63946950e-01 -3.87997597e-01 6.29694879e-01
8.37171793e-01 -1.50390804e+00 -2.30429128e-01 -5.09974360e-01
5.49602091e-01 -8.33782032e-02 4.64584023e-01 6.91558599e-01
-3.52421731e-01 8.35402310e-01 3.60645652e-01 2.21746620e-02
-8.17473829e-01 1.00645244e+00 4.67946053e-01 -6.48047745e-01
-4.64937061e-01 8.57438147e-01 -2.70209610e-01 -5.30527771e-01
8.12363774e-02 -9.26918209e-01 -5.74606478e-01 5.55209458e-01
2.96611309e-01 2.39105195e-01 1.41177103e-01 -4.25424308e-01
-5.20731509e-01 6.55876517e-01 2.09184721e-01 -5.11273779e-02
1.14294565e+00 1.01299316e-01 -3.31071287e-01 5.42981744e-01
1.20781767e+00 -8.26334506e-02 -4.86099035e-01 -6.68562651e-01
3.61943752e-01 -1.15962707e-01 -2.62456417e-01 -7.64930665e-01
-8.13016117e-01 5.39636910e-01 -1.56358704e-01 7.17377722e-01
1.15812504e+00 4.11715567e-01 1.18719721e+00 6.95638001e-01
5.30968547e-01 -9.34174359e-01 4.42094982e-01 8.32935035e-01
1.16670120e+00 -8.21245611e-01 -2.49950022e-01 -4.56312716e-01
-5.41394174e-01 1.18362093e+00 5.95590353e-01 1.06985770e-01
5.72424054e-01 -5.34863174e-01 -2.05388054e-01 -8.61463726e-01
-1.23915899e+00 -5.72404087e-01 4.72254217e-01 3.44312847e-01
-7.66028166e-02 -1.42230406e-01 -3.08573931e-01 6.83039486e-01
-4.04561073e-01 3.38118613e-01 1.85677692e-01 1.03313887e+00
-3.32691789e-01 -1.07026362e+00 -1.70074522e-01 4.65794563e-01
-2.25485966e-01 7.53190517e-02 -3.39592010e-01 8.31468463e-01
1.74795836e-01 1.28875089e+00 -2.06389248e-01 -5.59236526e-01
6.97919965e-01 6.00011826e-01 4.72923398e-01 -6.97679043e-01
-7.33732462e-01 -9.40821111e-01 4.75749135e-01 -7.92561054e-01
-3.67089093e-01 -4.15758669e-01 -1.58571172e+00 -1.16092458e-01
-1.63265795e-01 6.17931843e-01 1.15211174e-01 1.14736605e+00
2.85856396e-01 7.53467023e-01 3.96818787e-01 3.55531663e-01
-2.72466809e-01 -4.96300787e-01 -4.07566458e-01 3.91795665e-01
4.29562211e-01 -6.70045078e-01 -3.18537652e-01 8.22001025e-02] | [10.392597198486328, 7.945568084716797] |
1c479a2e-9e8a-4a0b-b1b1-6243e72b8722 | jl-dcf-joint-learning-and-densely-cooperative | 2004.08515 | null | https://arxiv.org/abs/2004.08515v1 | https://arxiv.org/pdf/2004.08515v1.pdf | JL-DCF: Joint Learning and Densely-Cooperative Fusion Framework for RGB-D Salient Object Detection | This paper proposes a novel joint learning and densely-cooperative fusion (JL-DCF) architecture for RGB-D salient object detection. Existing models usually treat RGB and depth as independent information and design separate networks for feature extraction from each. Such schemes can easily be constrained by a limited amount of training data or over-reliance on an elaborately-designed training process. In contrast, our JL-DCF learns from both RGB and depth inputs through a Siamese network. To this end, we propose two effective components: joint learning (JL), and densely-cooperative fusion (DCF). The JL module provides robust saliency feature learning, while the latter is introduced for complementary feature discovery. Comprehensive experiments on four popular metrics show that the designed framework yields a robust RGB-D saliency detector with good generalization. As a result, JL-DCF significantly advances the top-1 D3Net model by an average of ~1.9% (S-measure) across six challenging datasets, showing that the proposed framework offers a potential solution for real-world applications and could provide more insight into the cross-modality complementarity task. The code will be available at https://github.com/kerenfu/JLDCF/. | ['Ge-Peng Ji', 'Deng-Ping Fan', 'Qijun Zhao', 'Keren Fu'] | 2020-04-18 | jl-dcf-joint-learning-and-densely-cooperative-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Fu_JL-DCF_Joint_Learning_and_Densely-Cooperative_Fusion_Framework_for_RGB-D_Salient_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Fu_JL-DCF_Joint_Learning_and_Densely-Cooperative_Fusion_Framework_for_RGB-D_Salient_CVPR_2020_paper.pdf | cvpr-2020-6 | ['rgb-d-salient-object-detection'] | ['computer-vision'] | [ 2.61472519e-02 -6.40953556e-02 -3.84100199e-01 -3.69155049e-01
-9.13682163e-01 -4.15550649e-01 6.54596269e-01 1.59325242e-01
-4.82207537e-01 3.99052322e-01 5.37028238e-02 -4.45357822e-02
9.61273387e-02 -4.41623122e-01 -7.77907193e-01 -7.28209138e-01
1.68986246e-01 -6.41811341e-02 6.62632883e-01 -2.69347250e-01
2.43334800e-01 5.17937064e-01 -1.69637907e+00 3.87190096e-02
9.82094884e-01 1.48486328e+00 6.82153225e-01 1.85109496e-01
3.96692082e-02 5.32886684e-01 -3.08398157e-01 -1.75836891e-01
4.31509078e-01 -2.15201065e-01 -4.71479326e-01 -9.65495184e-02
3.20996046e-01 -4.00939077e-01 -2.99309015e-01 1.06845927e+00
6.03108883e-01 5.86819462e-02 4.10630524e-01 -1.30854833e+00
-6.33887947e-01 3.30922961e-01 -8.21189404e-01 3.27847451e-01
3.23074728e-01 2.11585954e-01 9.80729163e-01 -1.16587579e+00
3.58217925e-01 1.08374453e+00 4.50400054e-01 5.23999333e-01
-1.04375231e+00 -6.05703115e-01 2.38309205e-01 1.98630005e-01
-1.36520493e+00 -3.85573179e-01 1.19442904e+00 -1.33347675e-01
7.27330506e-01 2.51244813e-01 8.06871891e-01 9.20360446e-01
-2.19541546e-02 1.44814324e+00 1.11087382e+00 -1.38046935e-01
3.30419511e-01 7.82837123e-02 -6.04639985e-02 7.85983503e-01
3.05924535e-01 7.02072456e-02 -8.10284376e-01 5.66173308e-02
9.92760241e-01 3.82143527e-01 -4.04126287e-01 -7.83734560e-01
-1.31253350e+00 6.51600480e-01 1.13363361e+00 1.57388166e-01
-2.41201818e-01 1.32933781e-01 1.74125418e-01 -8.80533084e-02
3.68338138e-01 3.03301960e-01 -2.92466164e-01 2.28662431e-01
-9.92474437e-01 1.84510484e-01 2.66885817e-01 9.73610580e-01
8.74310911e-01 -7.68788680e-02 -7.03765487e-04 6.73650444e-01
5.14292419e-01 7.56480932e-01 4.64390695e-01 -7.51712263e-01
2.44225532e-01 7.29182422e-01 1.96279868e-01 -9.51279938e-01
-5.30294299e-01 -6.27885163e-01 -6.21589065e-01 3.48484457e-01
3.66100639e-01 1.49858028e-01 -9.71026182e-01 1.69068885e+00
4.00130391e-01 1.67343199e-01 -4.63388525e-02 1.44860411e+00
1.15754366e+00 2.99414665e-01 -1.84256230e-02 2.32007191e-01
1.20042431e+00 -1.11329496e+00 -4.60183442e-01 -3.59978884e-01
3.14336568e-01 -6.90716565e-01 1.10114396e+00 1.22541264e-01
-1.25141108e+00 -5.49670100e-01 -1.23645663e+00 -3.40450197e-01
-2.45187208e-01 4.13802028e-01 8.39833617e-01 2.90177166e-01
-1.09247839e+00 2.07714334e-01 -1.02282548e+00 -2.01977253e-01
7.06334352e-01 5.06313980e-01 -3.12291205e-01 -1.65535763e-01
-9.77992475e-01 8.30747008e-01 9.53883901e-02 2.32244536e-01
-9.68198776e-01 -5.13578117e-01 -8.93438518e-01 -1.41016543e-01
3.82111937e-01 -6.88341081e-01 1.26305282e+00 -7.83418953e-01
-1.39443445e+00 8.64956617e-01 -1.54657245e-01 -2.68611938e-01
2.98473388e-01 -4.42070991e-01 -1.59832790e-01 4.20456648e-01
1.99458703e-01 8.01571012e-01 9.92135584e-01 -1.37461782e+00
-6.61170185e-01 -4.43913072e-01 1.08294263e-01 2.66409278e-01
-2.34802261e-01 -1.16126254e-01 -6.31687224e-01 -6.30144060e-01
4.84157503e-01 -7.44206369e-01 -1.88313231e-01 4.22097415e-01
-3.31457824e-01 -1.31097838e-01 7.44859278e-01 -3.54481548e-01
9.11840737e-01 -2.15386033e+00 2.63151556e-01 -8.35791081e-02
5.27272224e-01 4.20758337e-01 -6.27080873e-02 4.76433262e-02
2.70560473e-01 -2.48734728e-01 -3.39141726e-01 -5.89865983e-01
-7.63233304e-02 -9.49896276e-02 -7.27892667e-02 6.76329613e-01
4.16310996e-01 1.27214921e+00 -1.18114650e+00 -4.01034564e-01
4.04336393e-01 5.35356045e-01 -4.36521053e-01 2.03327984e-01
-5.78068867e-02 3.11470330e-01 -6.19147956e-01 1.14378679e+00
7.70594358e-01 -4.56173390e-01 -1.46222755e-01 -4.65185255e-01
-1.60625383e-01 1.71404883e-01 -9.93312418e-01 2.25877428e+00
-2.17302278e-01 4.19097334e-01 1.47052631e-01 -8.83179069e-01
1.05075598e+00 -1.05289057e-01 3.56012106e-01 -9.31527495e-01
4.20756131e-01 6.00250781e-01 -3.20540905e-01 -2.69677430e-01
3.54942948e-01 1.70843399e-04 -1.51849970e-01 2.87702173e-01
3.10414106e-01 -1.73641115e-01 -8.34525973e-02 3.08210790e-01
9.29969907e-01 1.96865231e-01 2.97958046e-01 -2.36214429e-01
5.34118474e-01 -2.23173946e-01 7.60445178e-01 6.07771933e-01
-6.79736614e-01 6.76634967e-01 1.26034513e-01 -1.46928206e-01
-6.61241353e-01 -1.35018253e+00 -6.01500599e-03 9.26439762e-01
9.50847626e-01 -1.02267273e-01 -3.45760494e-01 -9.91854668e-01
2.00779125e-01 3.16264898e-01 -5.53445458e-01 -3.99341702e-01
-3.10908735e-01 -4.20750856e-01 2.71263510e-01 7.17136502e-01
7.34099090e-01 -7.49395370e-01 -8.31443429e-01 -4.03599702e-02
-5.30709140e-02 -1.11874998e+00 -4.27376896e-01 4.29637432e-01
-9.01988149e-01 -9.11059618e-01 -9.85064566e-01 -9.45705175e-01
5.81740201e-01 8.98148358e-01 8.09246540e-01 -1.58225391e-02
-1.08346567e-01 3.29622507e-01 -4.41852748e-01 -2.90376872e-01
2.67833143e-01 3.68260555e-02 1.30073307e-02 -4.24020551e-02
4.91354406e-01 -5.88943303e-01 -1.04731166e+00 2.70761639e-01
-7.80329525e-01 1.74996451e-01 9.48704660e-01 7.46865988e-01
6.64056361e-01 -5.99190712e-01 6.38325095e-01 -1.09511770e-01
1.81383923e-01 -5.01789451e-01 -5.81244171e-01 9.52329934e-02
-4.02929634e-01 -3.20982412e-02 2.55950004e-01 -1.40218675e-01
-9.95747685e-01 3.59130949e-01 5.88835217e-03 -6.89666271e-01
-1.79569855e-01 1.59388885e-01 -2.85207510e-01 -3.30278873e-01
4.27734911e-01 2.78484941e-01 2.79121213e-02 -5.42948604e-01
3.53442222e-01 5.46351254e-01 6.47024274e-01 -3.43020856e-01
8.09314609e-01 7.03069866e-01 -5.40195927e-02 -4.86911833e-01
-1.07205749e+00 -7.09909797e-01 -6.40002847e-01 -3.41722071e-01
7.46063948e-01 -1.33805680e+00 -4.10985112e-01 6.33768380e-01
-9.06090319e-01 -3.28169197e-01 -1.30908757e-01 5.79903245e-01
-4.04539287e-01 2.32568100e-01 -4.04754817e-01 -5.96931934e-01
-2.56450593e-01 -1.18369210e+00 1.49382722e+00 5.00721335e-01
1.98531196e-01 -9.06546116e-01 -1.87876761e-01 4.27882791e-01
4.86716330e-01 2.49690667e-01 2.84686506e-01 -4.12347585e-01
-7.76408434e-01 -1.85441166e-01 -6.22059345e-01 2.92064339e-01
2.75388092e-01 -3.64569038e-01 -1.08075118e+00 -3.21575522e-01
-1.11585362e-02 -6.08221710e-01 1.17761290e+00 3.31710160e-01
9.78237510e-01 2.56734848e-01 -2.94274300e-01 7.08692610e-01
1.50439990e+00 -1.35995552e-01 2.66856372e-01 3.92357051e-01
8.11868727e-01 1.32398471e-01 7.30145097e-01 5.09632826e-01
6.42963171e-01 6.81849897e-01 8.05164933e-01 -3.95496786e-01
-4.26526874e-01 -1.86541498e-01 3.15044969e-01 6.48660362e-01
-5.89760095e-02 6.68298155e-02 -9.34559524e-01 6.74019754e-01
-1.96454561e+00 -6.70042217e-01 1.95894693e-03 1.96145999e+00
7.68498361e-01 1.92450270e-01 2.72702962e-01 1.60284355e-01
5.97207427e-01 2.49293476e-01 -8.45167339e-01 2.08227396e-01
-4.02883649e-01 1.20264247e-01 4.67792183e-01 2.93800563e-01
-1.26447964e+00 9.11143661e-01 5.24004078e+00 8.49378645e-01
-1.44195342e+00 2.75791287e-01 3.26130807e-01 -4.89297122e-01
-3.55561376e-01 -1.50765345e-01 -7.76711047e-01 4.37919050e-01
2.76685447e-01 -1.94082744e-02 1.49077624e-01 8.13803554e-01
3.08270510e-02 -4.30660874e-01 -7.84818470e-01 1.12203431e+00
1.68404043e-01 -1.24215674e+00 -9.69159380e-02 -1.96338490e-01
7.92267442e-01 3.64217758e-01 2.24646673e-01 1.11244254e-01
-2.58175991e-02 -5.19900680e-01 1.00248086e+00 4.98246431e-01
5.55576801e-01 -6.39980853e-01 6.73239112e-01 2.03890592e-01
-1.23508716e+00 -2.28413299e-01 -2.90747762e-01 7.95247685e-03
1.32819459e-01 9.49184597e-01 -3.64531457e-01 6.35665834e-01
8.57297063e-01 1.07440817e+00 -7.45869875e-01 1.25654233e+00
-3.87992948e-01 2.72886485e-01 -4.34100479e-01 -5.83751425e-02
3.65920603e-01 1.94054469e-01 5.85780561e-01 1.04143465e+00
1.90646574e-01 -1.33864000e-01 2.13356361e-01 8.14943790e-01
3.32402922e-02 -9.70507562e-02 -2.95304984e-01 2.66916037e-01
4.80546713e-01 1.44225156e+00 -9.53664243e-01 -1.51618570e-01
-5.64769864e-01 1.04181123e+00 4.44809020e-01 3.47826302e-01
-1.01488650e+00 -3.71488988e-01 6.55588388e-01 -1.24578722e-01
6.86676025e-01 -2.42325217e-01 -4.45034742e-01 -1.28874648e+00
2.10966304e-01 -4.42823023e-01 9.81637239e-02 -8.21262777e-01
-1.03987062e+00 5.43183029e-01 -2.71652311e-01 -1.41450572e+00
2.55786330e-01 -5.90415239e-01 -4.32515055e-01 6.51497245e-01
-2.06249022e+00 -1.26864386e+00 -6.80771768e-01 6.73084676e-01
2.53633350e-01 -9.86190438e-02 3.89826357e-01 2.78407365e-01
-6.38942659e-01 5.80641091e-01 -3.62338200e-02 -7.53253698e-02
6.37054443e-01 -1.20490229e+00 9.72718224e-02 9.12909329e-01
1.49129955e-02 4.28715050e-01 4.16582137e-01 -3.98028523e-01
-1.54411876e+00 -1.00701046e+00 5.54398954e-01 -2.96786964e-01
5.44112027e-01 -4.37711388e-01 -7.38470018e-01 2.10023031e-01
-2.48317188e-03 4.51047748e-01 5.29682100e-01 -3.08860838e-01
-4.28042561e-01 -2.95581937e-01 -1.16674733e+00 3.64768952e-01
1.08877289e+00 -6.55516267e-01 -5.69832444e-01 9.62739438e-02
8.10530186e-01 -3.06166202e-01 -7.29320049e-01 5.38229942e-01
5.32346904e-01 -1.28948832e+00 9.94611204e-01 1.18717030e-02
5.03462076e-01 -7.48931110e-01 -3.93641233e-01 -1.01242459e+00
-3.47032011e-01 -3.42318267e-01 -5.41808665e-01 9.45660353e-01
1.99585557e-01 -4.27485168e-01 6.83788180e-01 2.78544307e-01
-3.46832305e-01 -1.18579829e+00 -8.62572849e-01 -8.71882975e-01
-2.83079743e-01 -4.08059776e-01 5.55514097e-01 7.03100443e-01
-2.94189733e-02 1.27815455e-01 -1.55276597e-01 2.74885148e-01
8.51697087e-01 4.26282197e-01 5.16637206e-01 -1.08308089e+00
-1.34011328e-01 -7.23530889e-01 -5.60263872e-01 -1.47618401e+00
1.26977097e-02 -1.02509487e+00 9.69703719e-02 -1.55935562e+00
1.91777885e-01 -4.80082572e-01 -5.13373435e-01 6.07730567e-01
-3.64530653e-01 5.38227677e-01 1.95947826e-01 3.63903731e-01
-9.43858147e-01 1.01798058e+00 1.41218960e+00 -3.41099501e-02
-1.91938475e-01 -7.65784010e-02 -8.82546782e-01 6.66988075e-01
1.02757776e+00 -1.75781608e-01 -3.56562823e-01 -4.55589205e-01
-7.59205744e-02 -2.28937685e-01 7.63867319e-01 -1.12508321e+00
4.32433575e-01 -4.08523679e-02 5.18929482e-01 -5.73398113e-01
5.07398725e-01 -6.96509302e-01 -5.20601749e-01 4.00325507e-01
-2.84571555e-02 -3.16712677e-01 1.96815521e-01 6.68906093e-01
-3.42134684e-01 1.97372124e-01 8.03422749e-01 -4.04094942e-02
-1.05698490e+00 2.68017083e-01 7.47646317e-02 -5.36138378e-02
1.06184983e+00 -2.89438367e-01 -3.58932942e-01 -1.40868932e-01
-4.69817132e-01 2.94988394e-01 6.60208106e-01 4.40888345e-01
8.58795643e-01 -1.44513977e+00 -3.03846955e-01 3.84737194e-01
2.91230768e-01 2.39017636e-01 2.01089501e-01 1.13894475e+00
-1.26293540e-01 3.81000370e-01 -3.46493781e-01 -9.05067563e-01
-8.89190435e-01 5.20315588e-01 9.29392651e-02 9.59590003e-02
-4.02035326e-01 1.23108029e+00 2.13179022e-01 -3.24294388e-01
3.35673332e-01 -4.24124122e-01 3.03293280e-02 7.63684735e-02
4.20257181e-01 2.25265533e-01 9.67612024e-03 -6.93183959e-01
-6.74293697e-01 5.29974163e-01 -8.01147223e-02 1.15589850e-01
1.47378504e+00 -2.98725665e-01 9.87810194e-02 4.52946335e-01
1.22056043e+00 -2.45222524e-01 -1.66433156e+00 -5.19803226e-01
-1.72880128e-01 -4.45423901e-01 4.47692335e-01 -7.30423629e-01
-1.30932736e+00 7.09131300e-01 6.82904720e-01 -1.70034140e-01
1.49973202e+00 3.25392634e-01 8.69618475e-01 2.78089374e-01
5.02967238e-01 -9.10438120e-01 3.50136459e-01 2.91473746e-01
9.75730777e-01 -1.59322453e+00 1.16311774e-01 -3.19994748e-01
-6.68268681e-01 7.69856691e-01 7.35521555e-01 -3.55931461e-01
7.94313788e-01 5.47421724e-02 7.28815794e-02 -3.27767432e-01
-4.77588415e-01 -5.87437034e-01 5.24506688e-01 4.49130863e-01
1.92750826e-01 1.49268871e-02 2.79382279e-04 6.92748427e-01
8.55425820e-02 -5.76082580e-02 1.40766948e-01 1.23336744e+00
-5.81103384e-01 -8.12557876e-01 -1.94487870e-01 3.68750036e-01
-1.03663251e-01 7.37924427e-02 -4.22194511e-01 6.97645068e-01
1.50859401e-01 7.52456486e-01 -1.06414795e-01 -5.58309913e-01
2.37462848e-01 -3.03676367e-01 5.59078276e-01 -4.34733748e-01
-4.07952487e-01 3.09052706e-01 -3.87149751e-01 -9.24472809e-01
-8.36868346e-01 -7.68205166e-01 -1.29266429e+00 3.91844809e-02
-5.29826581e-01 -2.28582203e-01 6.35223508e-01 8.64842176e-01
4.89698201e-01 4.00775135e-01 6.63505733e-01 -1.42725766e+00
-2.92321056e-01 -7.30973959e-01 -6.11213148e-01 1.48424968e-01
6.78985178e-01 -1.08587229e+00 -4.52819854e-01 -2.66331106e-01] | [9.638601303100586, -0.7910177707672119] |
11013305-bb07-46f9-a2f8-c4219f04827b | starnet-style-aware-3d-point-cloud-generation | 2303.15805 | null | https://arxiv.org/abs/2303.15805v1 | https://arxiv.org/pdf/2303.15805v1.pdf | StarNet: Style-Aware 3D Point Cloud Generation | This paper investigates an open research task of reconstructing and generating 3D point clouds. Most existing works of 3D generative models directly take the Gaussian prior as input for the decoder to generate 3D point clouds, which fail to learn disentangled latent codes, leading noisy interpolated results. Most of the GAN-based models fail to discriminate the local geometries, resulting in the point clouds generated not evenly distributed at the object surface, hence degrading the point cloud generation quality. Moreover, prevailing methods adopt computation-intensive frameworks, such as flow-based models and Markov chains, which take plenty of time and resources in the training phase. To resolve these limitations, this paper proposes a unified style-aware network architecture combining both point-wise distance loss and adversarial loss, StarNet which is able to reconstruct and generate high-fidelity and even 3D point clouds using a mapping network that can effectively disentangle the Gaussian prior from input's high-level attributes in the mapped latent space to generate realistic interpolated objects. Experimental results demonstrate that our framework achieves comparable state-of-the-art performance on various metrics in the point cloud reconstruction and generation tasks, but is more lightweight in model size, requires much fewer parameters and less time for model training. | ['Chunyan Miao', 'Zhiqi Shen', 'Vun Chan Hua Nicholas', 'Guosheng Lin', 'Hao Wang', 'Yunfan Zhang'] | 2023-03-28 | null | null | null | null | ['point-cloud-reconstruction', 'point-cloud-generation', 'generating-3d-point-clouds'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 1.64694060e-02 1.57921955e-01 6.75982162e-02 -2.97077429e-02
-8.97663116e-01 -6.10893548e-01 7.36574411e-01 -4.71996814e-01
2.77191848e-01 7.98777640e-01 -8.76352787e-02 -1.84962943e-01
3.29587311e-01 -1.32375872e+00 -1.10309219e+00 -7.60354757e-01
3.08051646e-01 8.73864293e-01 -1.64658323e-01 -3.25270891e-02
1.37421608e-01 7.66590118e-01 -1.46793091e+00 2.66202632e-03
1.22597051e+00 7.64368653e-01 2.81691432e-01 5.51979303e-01
-3.78118992e-01 4.20563102e-01 -5.13960540e-01 -6.46804869e-01
4.99857932e-01 -4.30417746e-01 -3.62591326e-01 -2.37851199e-02
4.65044886e-01 -5.18204272e-01 -5.85480511e-01 9.56045389e-01
5.44266522e-01 -2.15612322e-01 1.02393055e+00 -1.41322899e+00
-1.03427529e+00 1.49609566e-01 -3.64749819e-01 -5.25851607e-01
1.08884938e-01 3.94606650e-01 5.24753153e-01 -1.07545149e+00
4.75333065e-01 1.43755817e+00 6.54995024e-01 7.73744404e-01
-1.37256539e+00 -1.01763678e+00 -1.44276559e-01 -2.94212312e-01
-1.60663271e+00 -2.08021730e-01 9.58747447e-01 -6.06790960e-01
7.15472817e-01 3.28393131e-01 8.93810272e-01 1.46392345e+00
2.41670564e-01 5.04820883e-01 9.63601768e-01 -6.20545372e-02
2.80979007e-01 2.86676530e-02 -8.52191627e-01 4.92603064e-01
2.47774422e-01 4.95600164e-01 -4.89555687e-01 -3.10117304e-01
1.57318330e+00 8.28679278e-02 -1.82560936e-01 -5.39586246e-01
-1.31911504e+00 8.81816506e-01 5.65745473e-01 -1.99410886e-01
-3.53656024e-01 3.68759990e-01 -1.46576256e-01 4.52287942e-02
7.14081168e-01 2.35174194e-01 3.78058176e-03 -8.14938694e-02
-1.05465198e+00 6.27780855e-01 5.05759478e-01 1.55259931e+00
8.56887579e-01 3.71417493e-01 -2.86766976e-01 3.89965981e-01
7.37220943e-01 8.96266043e-01 -1.64691195e-01 -9.35397327e-01
7.62578189e-01 4.77324098e-01 2.92564243e-01 -1.04958236e+00
2.99211413e-01 -4.08720523e-01 -1.24681342e+00 6.13204956e-01
-9.43004936e-02 -2.02783681e-02 -1.19045532e+00 1.60488474e+00
3.00634921e-01 6.79185688e-01 3.96402329e-02 1.03807652e+00
6.30472183e-01 8.40086520e-01 -6.41693994e-02 1.93694204e-01
9.39990878e-01 -6.72227681e-01 -4.85238880e-01 2.16867346e-02
-2.99108122e-02 -9.67661023e-01 8.96057725e-01 6.20225780e-02
-1.49248910e+00 -7.37049282e-01 -1.14508522e+00 -2.80247569e-01
-7.88165703e-02 -1.87863588e-01 6.78970814e-01 5.02676010e-01
-1.01136744e+00 6.75448537e-01 -1.05430710e+00 3.04964930e-01
7.74773180e-01 1.40841633e-01 -1.65559873e-01 -2.63204515e-01
-9.62959886e-01 7.31474400e-01 9.20850113e-02 1.97617307e-01
-1.18552923e+00 -1.09241652e+00 -7.95685887e-01 -3.41033675e-02
-1.93906248e-01 -1.56216061e+00 8.73674929e-01 -4.90762889e-01
-1.66346967e+00 6.29799604e-01 -9.56344903e-02 -5.51553667e-02
9.54715788e-01 1.89069891e-03 -1.22181349e-01 -1.11593977e-01
1.94414645e-01 1.02146173e+00 8.96300673e-01 -1.65755773e+00
-2.59784490e-01 -3.10559809e-01 -1.36911005e-01 2.84797162e-01
3.69509071e-01 -6.43352032e-01 -4.04043794e-01 -8.76663089e-01
3.55586499e-01 -9.88118529e-01 -3.47627550e-01 2.93111771e-01
-4.65112567e-01 4.11351994e-02 9.06370699e-01 -5.33550858e-01
4.96789128e-01 -2.05527687e+00 3.70009035e-01 -6.99123815e-02
2.76381820e-01 -5.84745640e-03 -3.26556177e-03 3.86238068e-01
7.56576732e-02 4.35902596e-01 -3.88708860e-01 -8.52396965e-01
1.86308533e-01 4.58052039e-01 -6.72628164e-01 3.03895801e-01
4.12422568e-01 1.15490389e+00 -9.60533321e-01 -3.64602923e-01
4.96596843e-01 9.96719778e-01 -6.98371053e-01 3.93541694e-01
-4.21293288e-01 9.95355308e-01 -5.39175391e-01 5.01651525e-01
1.19146347e+00 -2.44211048e-01 -5.31325758e-01 -4.51618768e-02
7.79113267e-03 2.20494211e-01 -1.00065219e+00 2.37253356e+00
-6.58992589e-01 3.36193353e-01 -1.89652458e-01 -3.88526291e-01
1.12171090e+00 5.46441495e-01 3.03176641e-01 -2.51044065e-01
-1.57595694e-01 3.52613956e-01 -5.63677669e-01 -1.21189781e-01
5.23585975e-01 -3.03302437e-01 -7.13079236e-03 1.24803022e-01
-1.01008207e-01 -7.99088776e-01 -5.56723714e-01 1.85631067e-01
7.38703549e-01 6.40615404e-01 -5.12339115e-01 1.47932246e-01
1.91000238e-01 -6.70789182e-02 5.34741521e-01 4.54828829e-01
6.04121149e-01 1.29137886e+00 2.39687294e-01 -2.43729755e-01
-1.62680304e+00 -1.64993632e+00 -5.78993298e-02 -7.44105950e-02
4.08363372e-01 -1.00611962e-01 -5.43262720e-01 -4.29212928e-01
1.28712907e-01 1.05910480e+00 -3.80865395e-01 -3.03031892e-01
-5.77802241e-01 -4.92935389e-01 6.18963063e-01 4.77038532e-01
5.51485300e-01 -7.36141086e-01 -2.29615971e-01 1.44669339e-01
-3.07423264e-01 -9.56411839e-01 -2.78981805e-01 -3.73516649e-01
-1.27332032e+00 -6.00888610e-01 -7.71133304e-01 -5.85096061e-01
1.02755511e+00 7.89899305e-02 1.40649509e+00 -1.58006936e-01
3.58935446e-02 -3.60511132e-02 -1.42729089e-01 -4.15268958e-01
-5.68927169e-01 -8.09502602e-02 -2.74607360e-01 -5.61914369e-02
2.24112440e-02 -1.02189314e+00 -7.03548014e-01 3.66870373e-01
-9.61565971e-01 6.00464761e-01 5.63035846e-01 7.86552131e-01
9.00545359e-01 -1.79170296e-01 3.57710093e-01 -4.59260792e-01
2.52806574e-01 -6.23459160e-01 -6.97824299e-01 -9.01452303e-02
-4.53662634e-01 9.14641693e-02 4.99551535e-01 -2.86102563e-01
-1.00797641e+00 6.93182498e-02 -2.79783756e-01 -1.17538023e+00
-1.65820524e-01 -2.67731510e-02 -5.13462484e-01 -1.09842576e-01
5.21948874e-01 5.11759996e-01 -9.50334296e-02 -5.23621440e-01
5.41749120e-01 3.19112927e-01 5.41879296e-01 -6.78795815e-01
1.35816586e+00 6.42734468e-01 2.90825099e-01 -4.16986704e-01
-4.12938237e-01 1.12667479e-01 -5.22329807e-01 -1.21011801e-01
9.73414242e-01 -1.29930031e+00 -3.91140640e-01 4.38986868e-01
-1.67027748e+00 -4.83588502e-03 -5.12017787e-01 4.07915801e-01
-9.06601369e-01 1.34311318e-01 -4.30879682e-01 -7.12710500e-01
-3.17872286e-01 -1.25947940e+00 1.53258276e+00 2.60518081e-02
1.20147645e-01 -7.85572946e-01 -2.02999096e-02 2.88235396e-01
4.14653361e-01 7.40994036e-01 1.07742906e+00 3.21477294e-01
-1.60457718e+00 -2.47765779e-01 -1.31704912e-01 2.66739994e-01
1.35712892e-01 -6.67381361e-02 -1.01955831e+00 -2.10408479e-01
-2.73365881e-02 -3.80839407e-02 3.80058795e-01 2.48912826e-01
1.29017925e+00 -3.38890344e-01 -3.51978272e-01 1.18921888e+00
1.59417319e+00 -6.41352162e-02 1.09106040e+00 -3.22321117e-01
1.21964467e+00 1.36820063e-01 2.59226471e-01 3.38634312e-01
4.33577359e-01 6.81191802e-01 8.76110196e-01 -8.29821676e-02
-3.48201334e-01 -9.66932178e-01 7.48942941e-02 9.86370444e-01
-3.85841221e-01 -4.46079493e-01 -7.07752824e-01 4.50960726e-01
-1.72372758e+00 -8.07092488e-01 -1.71622157e-01 2.13845873e+00
6.59534097e-01 -1.31084993e-01 -4.55072612e-01 -1.07125677e-01
6.21915340e-01 8.58813673e-02 -6.64071977e-01 5.68209514e-02
-1.11548401e-01 3.69823515e-01 4.65393901e-01 3.68286729e-01
-5.79289496e-01 8.83993447e-01 5.86755800e+00 9.89632368e-01
-1.15144587e+00 1.73513114e-01 5.09307027e-01 -2.30946839e-01
-1.00777996e+00 1.33696750e-01 -5.76667488e-01 7.31372952e-01
5.87983906e-01 -8.83929282e-02 4.11054224e-01 7.93482840e-01
1.97134987e-01 4.49398637e-01 -1.19603527e+00 1.31203735e+00
-1.11093223e-01 -1.71035421e+00 6.01209760e-01 2.94395119e-01
9.43717480e-01 7.51066729e-02 2.43171930e-01 7.43410289e-02
4.32763457e-01 -1.29882741e+00 1.09231532e+00 8.98721516e-01
1.16735768e+00 -8.82427692e-01 3.23583692e-01 6.48544371e-01
-8.65245938e-01 6.27786636e-01 -5.54823518e-01 1.19909398e-01
4.70489442e-01 6.70208275e-01 -7.44438887e-01 9.16042268e-01
5.77876091e-01 5.45126796e-01 -9.73588675e-02 9.68170047e-01
-3.74426156e-01 3.81036758e-01 -3.83922637e-01 3.05982441e-01
6.20952435e-02 -4.02158588e-01 8.81729007e-01 5.12900889e-01
9.12455261e-01 1.40757766e-02 -7.31479330e-03 1.88857305e+00
-9.83698964e-02 -3.33247811e-01 -7.91086972e-01 1.17264777e-01
5.65519333e-01 8.42914224e-01 -3.97859991e-01 -1.27633914e-01
-6.97484761e-02 1.08766699e+00 6.19883835e-02 4.14640784e-01
-1.06646228e+00 -1.77918747e-01 8.56788337e-01 3.51526797e-01
2.91762710e-01 -6.43189073e-01 -6.63661540e-01 -1.18829024e+00
1.73714593e-01 -3.51048023e-01 -3.57008159e-01 -1.16199982e+00
-1.46341121e+00 7.19707489e-01 -1.18051739e-02 -1.69805503e+00
-4.14055735e-01 -2.00071901e-01 -4.48271304e-01 1.48520470e+00
-1.51371944e+00 -1.44247162e+00 -5.03650010e-01 5.27020991e-01
3.56941909e-01 -1.16116725e-01 8.27520013e-01 4.75351930e-01
-2.45841108e-02 4.41513479e-01 -3.16004977e-02 -1.24319941e-01
2.78762072e-01 -1.10550261e+00 1.01648331e+00 7.62699187e-01
9.05151069e-02 4.42842424e-01 5.28152525e-01 -8.15441251e-01
-1.43219066e+00 -1.34014046e+00 6.31125271e-01 -8.04000556e-01
-1.55540466e-01 -6.29785836e-01 -8.01120937e-01 4.51949120e-01
1.93242319e-02 1.02388419e-01 4.00764436e-01 -5.22943676e-01
-2.09407732e-01 2.30298176e-01 -1.34205043e+00 6.87570453e-01
1.37602437e+00 -4.77046400e-01 -2.32436582e-01 2.78267354e-01
1.06646240e+00 -8.09054792e-01 -8.56977522e-01 4.16937441e-01
1.56404704e-01 -9.89149034e-01 1.13968456e+00 -3.09650570e-01
8.91010106e-01 -6.46983445e-01 -1.01729900e-01 -1.34853601e+00
-3.91137570e-01 -6.53629839e-01 -2.69239038e-01 1.24397969e+00
3.45099717e-01 -5.25687814e-01 1.00957406e+00 4.99140561e-01
-4.64888483e-01 -7.28164136e-01 -1.05335045e+00 -6.50335610e-01
3.85178000e-01 -3.67971212e-01 1.40051508e+00 8.72968853e-01
-9.47783113e-01 1.76342323e-01 -4.30373639e-01 5.30146182e-01
9.02493000e-01 3.12092185e-01 9.79407907e-01 -1.18171823e+00
-3.62764411e-02 -2.42037281e-01 -4.00688738e-01 -1.33976781e+00
-9.08639189e-03 -1.00866556e+00 -2.72652339e-02 -1.66638350e+00
-4.15795386e-01 -1.13717639e+00 2.64202803e-01 4.93106954e-02
1.15639038e-01 2.97188729e-01 1.67061552e-01 4.04393196e-01
1.23178020e-01 1.12933958e+00 1.77084136e+00 -2.32190177e-01
-8.56977981e-03 8.85341913e-02 -5.97840309e-01 4.60189223e-01
4.07514781e-01 -6.08652830e-01 -6.18753493e-01 -1.05218506e+00
3.91211838e-01 3.25656027e-01 8.65666091e-01 -1.12793911e+00
-5.12652565e-03 -3.21484774e-01 6.83311522e-01 -9.43161547e-01
7.10619986e-01 -1.06314683e+00 9.47858989e-01 4.12101336e-02
2.00748250e-01 1.69558618e-02 1.07675850e-01 6.48783445e-01
-2.31888264e-01 9.63651016e-02 6.53565884e-01 -2.70281166e-01
-1.76192727e-02 9.63178813e-01 3.28313023e-01 -1.59113094e-01
9.66966927e-01 -3.80240709e-01 -4.13072184e-02 -1.93556726e-01
-4.27175343e-01 -5.84363975e-02 9.52398837e-01 5.98195970e-01
7.73984432e-01 -1.91650724e+00 -9.40530181e-01 5.67478895e-01
-5.28452471e-02 1.06353068e+00 5.18880963e-01 1.44005090e-01
-9.39083576e-01 1.75536379e-01 -2.20359504e-01 -9.63962615e-01
-4.27233458e-01 4.62775469e-01 3.00605327e-01 4.40918729e-02
-8.56736183e-01 8.29510570e-01 2.56792456e-01 -6.58308208e-01
-1.83033392e-01 -2.45820567e-01 4.93317723e-01 -4.01204169e-01
-1.80265978e-02 2.19812497e-01 -8.02617893e-02 -6.36241317e-01
6.07239604e-02 6.72041178e-01 4.06776220e-01 -1.53141990e-01
1.18545949e+00 8.13278258e-02 1.98885165e-02 2.47050837e-01
1.10270596e+00 -4.38982956e-02 -1.63274908e+00 1.01809718e-01
-8.00984800e-01 -9.60421741e-01 4.12301756e-02 -5.67853868e-01
-1.17031050e+00 1.05467999e+00 5.15277207e-01 -8.50552768e-02
7.38209605e-01 -1.22953199e-01 9.97450173e-01 -2.82047182e-01
8.63819718e-01 -4.85598773e-01 -2.37800255e-01 2.30818525e-01
1.13943398e+00 -8.89298737e-01 -1.23411596e-01 -7.43079245e-01
-4.24551904e-01 6.78697228e-01 6.23099089e-01 -4.57084060e-01
6.02425456e-01 1.97141200e-01 -1.20104916e-01 -7.21658021e-02
-6.07643425e-01 4.55019206e-01 2.51776338e-01 8.37920427e-01
-3.28737497e-02 1.74826607e-01 2.95916021e-01 3.30313981e-01
-6.35717809e-01 -2.31746659e-02 2.18452960e-01 5.98860085e-01
2.53251553e-01 -1.17796171e+00 -3.65635812e-01 1.90131053e-01
4.92556486e-03 -1.11375391e-01 6.51573390e-02 4.64287519e-01
4.47733998e-01 4.86940593e-01 4.05630916e-01 -3.41304183e-01
3.84630889e-01 -1.19570270e-01 6.48321807e-01 -4.59310651e-01
-4.31321450e-02 -1.33241683e-01 -4.95120525e-01 -4.25760627e-01
-2.53336966e-01 -5.15353382e-01 -1.01551950e+00 -6.12736642e-01
-3.17354172e-01 1.05893314e-01 7.24813163e-01 4.29208875e-01
8.34435701e-01 5.09857297e-01 7.48315632e-01 -1.27971327e+00
-4.04408336e-01 -7.51162827e-01 -4.59340692e-01 3.60482693e-01
3.41243774e-01 -8.12284946e-01 -2.84569889e-01 -3.17021422e-02] | [8.86362361907959, -3.6505799293518066] |
b9815043-9f97-49c8-81af-3c9d1285b9a3 | code-prompting-a-neural-symbolic-method-for | 2305.18507 | null | https://arxiv.org/abs/2305.18507v1 | https://arxiv.org/pdf/2305.18507v1.pdf | Code Prompting: a Neural Symbolic Method for Complex Reasoning in Large Language Models | Large language models (LLMs) have scaled up to unlock a wide range of complex reasoning tasks with the aid of various prompting methods. However, current prompting methods generate natural language intermediate steps to help reasoning, which can cause imperfect task reduction and confusion. To mitigate such limitations, we explore code prompting, a neural symbolic prompting method with both zero-shot and few-shot versions which triggers code as intermediate steps. We conduct experiments on 7 widely-used benchmarks involving symbolic reasoning and arithmetic reasoning. Code prompting generally outperforms chain-of-thought (CoT) prompting. To further understand the performance and limitations of code prompting, we perform extensive ablation studies and error analyses, and identify several exclusive advantages of using symbolic promptings compared to natural language. We also consider the ensemble of code prompting and CoT prompting to combine the strengths of both. Finally, we show through experiments how code annotations and their locations affect code prompting. | ['Muhan Zhang', 'Zhouchen Lin', 'Haotong Yang', 'Yi Hu'] | 2023-05-29 | null | null | null | null | ['arithmetic-reasoning'] | ['reasoning'] | [ 3.29291910e-01 3.42877418e-01 -2.28823125e-01 -4.49612796e-01
-7.24484444e-01 -6.52733088e-01 8.77431393e-01 3.93826187e-01
-1.83656618e-01 3.51840734e-01 6.47653639e-01 -9.11064386e-01
-2.31355359e-03 -4.91001010e-01 -5.86413682e-01 7.88159482e-03
2.40563601e-02 -4.30416800e-02 1.30813152e-01 -3.86767894e-01
7.16386437e-01 9.24887508e-02 -1.32527864e+00 7.41377831e-01
1.15344620e+00 2.32434064e-01 1.49007872e-01 6.64945126e-01
-6.90626860e-01 1.78583515e+00 -7.51474619e-01 -2.87283957e-01
2.11342469e-01 -2.56671697e-01 -1.05061817e+00 -5.56173086e-01
3.47593397e-01 -4.51976180e-01 -1.39253378e-01 9.89092588e-01
2.77087510e-01 2.19673425e-01 3.46560925e-01 -1.50537419e+00
-9.37730432e-01 1.20052171e+00 -5.24582684e-01 3.60384524e-01
8.62062752e-01 6.04880273e-01 8.23424220e-01 -8.85940790e-01
4.00114596e-01 1.33483100e+00 8.13876748e-01 6.12972558e-01
-1.42284715e+00 -5.54064810e-01 3.50013644e-01 -5.74446693e-02
-1.23536134e+00 -5.36470711e-01 5.69335043e-01 -7.17853487e-01
1.58460295e+00 -4.14626859e-02 2.40513518e-01 1.10251284e+00
2.68738419e-01 7.65437722e-01 9.65405822e-01 -6.18615150e-01
3.01854819e-01 -1.95005551e-01 5.82578063e-01 1.01638651e+00
2.32359737e-01 1.78267062e-01 -6.09747767e-01 -6.33334935e-01
5.04298270e-01 1.66411683e-01 -1.00156978e-01 1.12260357e-01
-1.38638461e+00 7.59788096e-01 5.05118728e-01 1.86582953e-01
-1.45365641e-01 5.93900084e-01 5.59163213e-01 3.86927724e-01
4.43046764e-02 1.05629063e+00 -3.45548302e-01 -5.26510715e-01
-9.72040415e-01 3.72558743e-01 1.08187485e+00 1.13328683e+00
6.21748328e-01 1.56069845e-01 -8.34985971e-01 6.26853824e-01
1.40022263e-01 2.70545334e-01 3.94910902e-01 -1.05911314e+00
6.64723992e-01 1.03403223e+00 2.10917994e-01 -9.23503518e-01
-5.30883849e-01 5.73682897e-02 -1.59665704e-01 2.08565488e-01
5.46368480e-01 -2.96396226e-01 -7.44472086e-01 1.74743080e+00
-2.32026368e-01 -7.26747513e-02 -4.72287927e-03 6.96882248e-01
6.94641590e-01 3.65956783e-01 5.07013142e-01 2.36786023e-01
1.32869351e+00 -1.27837133e+00 -4.59808737e-01 -5.07219672e-01
1.14893210e+00 -6.38941109e-01 1.64898109e+00 2.38229379e-01
-8.49674046e-01 -3.92187953e-01 -7.52381146e-01 -3.35428715e-01
-1.99088112e-01 7.29117468e-02 1.06471741e+00 3.28922987e-01
-1.05461907e+00 3.06727260e-01 -9.11644518e-01 -1.84408918e-01
2.08375007e-01 3.47779952e-02 1.05717994e-01 -1.31791160e-01
-1.02186346e+00 9.85134006e-01 1.62390918e-01 -3.98724973e-01
-8.70151639e-01 -9.57854748e-01 -1.04833651e+00 2.41310909e-01
7.10276008e-01 -5.33758879e-01 1.76908875e+00 -9.41652894e-01
-1.18536282e+00 6.32559717e-01 -2.81662017e-01 -4.29564714e-01
2.60800362e-01 -6.98429525e-01 -9.59313363e-02 -8.15656483e-02
2.44545370e-01 7.29105532e-01 4.39676762e-01 -8.24507117e-01
-1.52626842e-01 3.08468759e-01 6.78923905e-01 -1.00479156e-01
-3.79910246e-02 3.17170292e-01 7.21850917e-02 -6.19152486e-01
-2.72927463e-01 -8.01272988e-01 -2.92834938e-01 -3.99480052e-02
-3.79947990e-01 -3.03085655e-01 3.02540541e-01 -5.23098350e-01
1.33684945e+00 -2.34381032e+00 -3.93150374e-02 -1.66685715e-01
5.12582898e-01 1.57483965e-01 -4.56623167e-01 6.07893467e-01
-2.31679857e-01 4.20483887e-01 -1.77492216e-01 4.32503745e-02
1.88000783e-01 4.56829555e-02 -7.30292559e-01 -9.25507322e-02
3.86771142e-01 1.23815215e+00 -1.14130807e+00 -4.81049597e-01
-3.78002003e-02 9.76778120e-02 -9.86242235e-01 1.90341577e-01
-7.55105615e-01 -5.57046868e-02 -3.83592755e-01 8.56499493e-01
1.37022346e-01 -5.39996982e-01 -1.38154969e-01 3.74945015e-01
-1.14211075e-01 6.39605820e-01 -6.73716843e-01 1.79427791e+00
-7.06304669e-01 6.43735468e-01 -2.09768370e-01 -4.71041888e-01
8.07496190e-01 3.16764206e-01 -1.92636967e-01 -7.43254840e-01
-1.64932072e-01 1.68802485e-01 2.06530407e-01 -7.96093643e-01
5.35939097e-01 -3.37919942e-03 -2.72338212e-01 8.62143993e-01
-1.45256907e-01 -1.61133155e-01 1.84404314e-01 7.17643023e-01
1.72481799e+00 2.75514662e-01 6.67940617e-01 -3.66346776e-01
2.21840933e-01 3.32308769e-01 1.92279145e-01 1.26508987e+00
-2.14586571e-01 2.24965870e-01 8.26908410e-01 -5.08723676e-01
-7.24728465e-01 -7.69160628e-01 4.25774574e-01 1.55555964e+00
-3.45797464e-02 -7.81126678e-01 -6.76198244e-01 -7.29454458e-01
1.01124570e-01 1.25123060e+00 -4.96369511e-01 -4.82376873e-01
-4.71404552e-01 -3.41668606e-01 1.08143759e+00 8.31738055e-01
3.35545719e-01 -1.25016606e+00 -1.15866709e+00 4.99297567e-02
-2.34747902e-01 -8.97657335e-01 -4.76127237e-01 3.94738615e-01
-6.45546138e-01 -1.10791540e+00 -2.80895829e-01 -3.80383521e-01
6.98731422e-01 3.79705399e-01 1.60358262e+00 8.03609133e-01
-2.30579853e-01 5.84238529e-01 -5.89980304e-01 -2.54182398e-01
-6.59698904e-01 1.01385392e-01 -2.34046191e-01 -8.75974655e-01
5.82062006e-01 -4.41645712e-01 -8.23551044e-02 -7.92862549e-02
-7.51925468e-01 3.80844206e-01 6.51741445e-01 7.17198312e-01
-2.64974147e-01 -5.32272041e-01 5.07316351e-01 -9.82281506e-01
1.15307999e+00 -5.89969158e-01 -5.31873822e-01 4.31879699e-01
-5.90131402e-01 5.39934278e-01 7.63229191e-01 -4.68253076e-01
-1.09783626e+00 -1.14640720e-01 1.04393594e-01 -3.30538452e-02
-3.04809928e-01 7.78510511e-01 4.84794617e-01 -1.12428196e-01
1.21391666e+00 1.14606671e-01 -2.18871623e-01 -2.26734400e-01
4.71314102e-01 4.26571727e-01 4.88321275e-01 -1.48917353e+00
6.68964326e-01 -6.96483180e-02 -4.57804561e-01 -2.70422637e-01
-8.38994026e-01 -1.14467070e-01 -2.60244817e-01 8.15581009e-02
6.44700646e-01 -9.34332907e-01 -7.34912813e-01 8.64650309e-02
-1.54011202e+00 -8.82542312e-01 -2.13397309e-01 2.27831565e-02
-4.87550795e-01 1.89243689e-01 -8.63464296e-01 -9.43738878e-01
-2.13731751e-01 -1.33259451e+00 1.08705187e+00 1.75055429e-01
-9.37091768e-01 -7.96880960e-01 -4.32081893e-02 -2.65512336e-02
6.57096922e-01 8.24545696e-02 1.30274177e+00 -9.84704971e-01
-7.04813242e-01 2.02156782e-01 -3.68738472e-01 -2.14347214e-01
-1.78474814e-01 -1.11966464e-03 -8.13542366e-01 6.63329437e-02
-1.77272424e-01 -6.75365329e-01 5.68617105e-01 -1.42609745e-01
9.04420435e-01 -4.98856038e-01 -2.45647624e-01 3.98322731e-01
1.26778257e+00 2.68201113e-01 3.97701025e-01 3.99033546e-01
6.00597680e-01 4.28389549e-01 5.13814688e-01 3.87468785e-01
4.11759913e-01 3.74484658e-01 1.61790386e-01 1.40694261e-01
9.95173864e-03 -3.74078184e-01 5.67781091e-01 4.78552788e-01
1.72825411e-01 9.99458432e-02 -1.63275361e+00 4.26666051e-01
-1.75112879e+00 -8.68932903e-01 -1.02493048e-01 1.81859314e+00
1.14931345e+00 1.84800074e-01 -6.25371560e-02 -4.90420274e-02
3.43997836e-01 6.87293475e-03 -4.50061351e-01 -6.42916322e-01
4.97881085e-01 1.81223214e-01 1.31149471e-01 5.20788372e-01
-5.19443631e-01 1.05060613e+00 7.03759241e+00 4.27084446e-01
-1.13952851e+00 1.98734235e-02 2.98242599e-01 -9.31252167e-02
-5.28005481e-01 2.18113482e-01 -5.73684990e-01 2.84623295e-01
9.97812331e-01 -3.27856749e-01 9.37640488e-01 1.21578228e+00
1.76439136e-02 -4.70546335e-01 -1.69023836e+00 7.03942835e-01
-5.18574528e-02 -1.28043795e+00 -5.83643280e-02 -6.27457261e-01
6.81717992e-01 2.05063954e-01 -3.58842075e-01 8.72136056e-01
9.36217010e-01 -1.09925354e+00 1.04582381e+00 4.35074300e-01
5.87948322e-01 -2.52524436e-01 4.09445941e-01 5.79936564e-01
-9.65443969e-01 -3.22545469e-01 7.19026551e-02 -8.85586321e-01
-4.82747629e-02 4.48505193e-01 -8.95054996e-01 1.42848521e-01
3.92653793e-01 2.41747811e-01 -1.07560599e+00 7.08568871e-01
-6.14796638e-01 6.62777483e-01 -1.50698870e-01 -2.86463588e-01
2.44841531e-01 5.27037144e-01 2.32978880e-01 1.50847065e+00
8.87761265e-02 3.23282570e-01 4.25780535e-01 1.51283371e+00
1.76559284e-01 -3.35917979e-01 -6.54722273e-01 -5.05270600e-01
9.00600255e-01 8.99090946e-01 -9.81128037e-01 -6.58462465e-01
-6.42436028e-01 6.39589310e-01 4.57649142e-01 6.04748905e-01
-8.85530055e-01 -5.95684528e-01 4.55744237e-01 -6.83148876e-02
-3.77823353e-01 -4.55038339e-01 -7.40375459e-01 -1.21182036e+00
2.26858128e-02 -1.18103015e+00 1.30929962e-01 -1.17761970e+00
-9.56366777e-01 3.71872485e-01 3.00003916e-01 -6.19768918e-01
-3.20744902e-01 -4.97376978e-01 -7.86785543e-01 9.01849627e-01
-1.26237345e+00 -5.13315439e-01 -5.16308486e-01 2.53834367e-01
6.35498285e-01 -4.97176573e-02 8.86497498e-01 1.74262404e-01
-2.09228292e-01 4.85644937e-01 -7.04915643e-01 2.04180449e-01
6.67065561e-01 -1.26181185e+00 8.88380527e-01 1.01207328e+00
-7.79197142e-02 1.45615602e+00 7.47239590e-01 -7.76431620e-01
-1.29774463e+00 -8.90306950e-01 9.93056297e-01 -8.16870928e-01
6.21047795e-01 -3.20105910e-01 -9.99764681e-01 1.01838601e+00
1.43198490e-01 -1.70702726e-01 4.88162875e-01 3.10314208e-01
-7.80426979e-01 5.81120968e-01 -9.09013033e-01 9.57135379e-01
1.04631424e+00 -9.71069038e-01 -1.11996305e+00 1.41552106e-01
1.21634686e+00 -5.43283224e-01 -3.08120877e-01 1.12052262e-01
3.12770009e-01 -1.12001121e+00 8.38095725e-01 -5.78847289e-01
1.04526842e+00 -2.90016890e-01 -1.02430865e-01 -1.11624110e+00
-2.71980435e-01 -6.57136679e-01 -7.86875263e-02 1.29040396e+00
2.95668125e-01 -5.11608779e-01 2.04409122e-01 1.13075113e+00
-2.54193813e-01 -5.01499593e-01 -2.39938319e-01 -6.29687726e-01
2.88058147e-02 -5.21417975e-01 7.49767661e-01 1.04925406e+00
7.62768924e-01 3.68108153e-01 -6.32494688e-02 -2.45670617e-01
-3.67281027e-02 1.43257141e-01 7.10506797e-01 -9.59894538e-01
-5.86310863e-01 -4.45424497e-01 2.06364930e-01 -9.89674807e-01
3.29544455e-01 -1.20220268e+00 3.82949710e-01 -1.38635015e+00
3.69536251e-01 -3.05614591e-01 4.26309966e-02 1.10244358e+00
-5.68090618e-01 -3.72002035e-01 4.54166681e-01 1.65105656e-01
-7.92334795e-01 1.94150507e-01 8.07127893e-01 -1.04844579e-02
-1.86324388e-01 -4.33297515e-01 -9.93706763e-01 7.56436765e-01
5.53077042e-01 -5.63357055e-01 -5.44296086e-01 -7.55903840e-01
7.80798376e-01 3.81013602e-01 6.46229923e-01 -9.67294931e-01
4.01286006e-01 -4.05016422e-01 -1.29980624e-01 6.84607551e-02
-2.91922390e-01 -5.27513504e-01 -2.05445930e-01 6.38966203e-01
-1.04348254e+00 4.16942328e-01 4.41076934e-01 2.67997712e-01
-6.45906292e-03 -4.37873244e-01 5.30849993e-01 -3.87724489e-01
-9.33034539e-01 -3.79022419e-01 -7.18261003e-01 4.49126482e-01
8.35536063e-01 6.25202507e-02 -7.49183893e-01 -2.24010646e-01
-2.13921100e-01 2.57376134e-01 6.38238668e-01 5.77118635e-01
3.89150530e-01 -1.05064774e+00 -2.80397028e-01 1.52457565e-01
4.26026970e-01 -9.75472778e-02 -3.22635740e-01 9.29403484e-01
-5.83018541e-01 5.63055634e-01 -1.31818965e-01 -3.98871601e-01
-8.41510713e-01 8.74298155e-01 1.02228791e-01 -4.18286286e-02
-7.01907456e-01 9.25373673e-01 1.18432239e-01 -6.45230472e-01
3.04710597e-01 -9.21101153e-01 2.35628635e-01 -4.40302551e-01
6.64904892e-01 2.28836909e-01 -6.29492998e-02 2.01995745e-01
-4.23261970e-01 -5.76804355e-02 -1.81290507e-02 -7.55445063e-02
9.16229546e-01 2.83192694e-01 -2.77614892e-01 5.47591209e-01
5.48512876e-01 1.68785416e-02 -1.16388988e+00 -1.91044599e-01
6.47858083e-01 -3.87930870e-01 -3.67755502e-01 -1.24632466e+00
-3.49162519e-01 8.94774437e-01 -1.07583836e-01 1.79808781e-01
8.19815636e-01 -1.44807875e-01 3.71863842e-01 9.14086878e-01
6.78639770e-01 -6.06536567e-01 2.96132058e-01 9.33134735e-01
8.04805934e-01 -1.16229331e+00 -2.49841064e-01 -2.33370975e-01
-4.64428663e-01 1.20460474e+00 1.04640245e+00 -4.42273803e-02
5.41828014e-02 9.16246176e-01 1.63989663e-01 -3.47759396e-01
-1.25271297e+00 1.87823504e-01 -2.89725006e-01 4.48912948e-01
1.12026143e+00 -1.12576045e-01 -2.85449296e-01 7.84106255e-01
-1.91550210e-01 4.94213492e-01 6.25669718e-01 1.30352747e+00
-3.52103800e-01 -7.98730135e-01 -5.77480257e-01 5.30107260e-01
-3.42924923e-01 -7.16251314e-01 -6.03001475e-01 6.72680795e-01
-2.25886270e-01 9.21794295e-01 -2.42033571e-01 -3.87930095e-01
1.03495605e-01 4.89579350e-01 2.94364184e-01 -1.12886965e+00
-9.86880124e-01 -6.97406590e-01 1.03679635e-01 -8.40021014e-01
1.18915997e-02 -3.19521397e-01 -1.63777566e+00 -4.86591667e-01
-1.89973459e-01 8.27691182e-02 2.32098967e-01 1.24230158e+00
4.65384901e-01 8.68226647e-01 -2.42477924e-01 -8.76772642e-01
-9.40786600e-01 -8.19879889e-01 1.17851771e-01 3.80166352e-01
5.08178115e-01 -6.35480523e-01 -2.66754001e-01 2.81013191e-01] | [8.070457458496094, 7.681636333465576] |
ea1aefa3-d77b-4168-8389-ce652f31fd03 | near-field-beam-management-for-extremely | 2306.16206 | null | https://arxiv.org/abs/2306.16206v1 | https://arxiv.org/pdf/2306.16206v1.pdf | Near-Field Beam Management for Extremely Large-Scale Array Communications | Extremely large-scale arrays (XL-arrays) have emerged as a promising technology to achieve super-high spectral efficiency and spatial resolution in future wireless systems. The large aperture of XL-arrays means that spherical rather than planar wavefronts must be considered, and a paradigm shift from far-field to near-field communications is necessary. Unlike existing works that have mainly considered far-field beam management, we study the new near-field beam management for XL-arrays. We first provide an overview of near-field communications and introduce various applications of XL-arrays in both outdoor and indoor scenarios. Then, three typical near-field beam management methods for XL-arrays are discussed: near-field beam training, beam tracking, and beam scheduling. We point out their main design issues and propose promising solutions to address them. Moreover, other important directions in near-field communications are also highlighted to motivate future research. | ['A. Lee Swindlehurst', 'Linglong Dai', 'Li Chen', 'Beixiong Zheng', 'Yong Zeng', 'Chenyu Wu', 'Yunpu Zhang', 'Changsheng You'] | 2023-06-28 | null | null | null | null | ['management'] | ['miscellaneous'] | [ 2.84832984e-01 -2.56030381e-01 2.41596922e-01 -5.05997181e-01
-4.76798773e-01 -5.47973812e-01 1.67417333e-01 -4.61288124e-01
2.61017233e-02 9.17169333e-01 4.54248458e-01 -5.14091849e-01
-8.44618440e-01 -9.54960227e-01 -2.09546342e-01 -9.12257016e-01
-3.70104611e-01 -7.40880892e-02 7.51560777e-02 -3.15427572e-01
-1.48994759e-01 7.95985043e-01 -1.20370996e+00 -2.86456831e-02
6.33482873e-01 1.22606647e+00 5.34200013e-01 8.43521297e-01
1.64797530e-01 4.98940438e-01 -6.58796132e-01 5.90497963e-02
1.48495629e-01 -4.81884390e-01 6.66709691e-02 2.15077117e-01
5.19112527e-01 -1.04748696e-01 -3.13522041e-01 9.47878122e-01
1.10133994e+00 -2.10476205e-01 3.85796219e-01 -8.92591536e-01
-1.67912856e-01 4.37833965e-01 -5.98391593e-01 3.83321881e-01
5.11069000e-01 -8.84842873e-02 2.36960366e-01 -1.04192460e+00
4.97785583e-02 8.71417642e-01 8.69741023e-01 1.17793754e-01
-2.72199124e-01 -9.48382020e-01 -3.29093963e-01 1.36686608e-01
-1.52394140e+00 -6.64843917e-01 4.42703426e-01 -3.78828228e-01
3.53548378e-01 7.94643879e-01 6.63068533e-01 4.26717520e-01
4.41738546e-01 3.22666205e-02 9.71001327e-01 -7.05131054e-01
4.99993831e-01 -2.76856929e-01 -1.69613481e-01 3.79350424e-01
7.11029112e-01 4.07426953e-01 -6.25792325e-01 -1.44219160e-01
9.68947351e-01 -3.46178085e-01 -7.90651262e-01 -4.30138171e-01
-1.31173170e+00 4.13299888e-01 6.77247643e-01 8.40701222e-01
-2.50885606e-01 8.23399574e-02 -6.40009105e-01 -3.48783880e-02
3.80603820e-01 8.50310385e-01 3.30530033e-02 3.89749944e-01
-8.45448017e-01 1.55604511e-01 5.54627299e-01 1.34146059e+00
4.11413848e-01 3.49723212e-02 -6.00111365e-01 6.93338573e-01
4.61499989e-01 1.09068346e+00 -4.44785684e-01 -6.38800144e-01
5.02777874e-01 -2.76039988e-01 4.61403430e-01 -7.08304286e-01
-8.39901567e-01 -1.71476793e+00 -1.07100964e+00 8.39100629e-02
2.42355183e-01 -8.13554525e-01 -5.80133796e-01 1.25370646e+00
2.74742037e-01 3.39852631e-01 1.01872012e-01 7.49651730e-01
8.50375652e-01 6.72715962e-01 -6.61798358e-01 -9.58259404e-01
1.13602662e+00 -8.23699117e-01 -1.23071110e+00 -6.28418982e-01
3.72874260e-01 -1.04681933e+00 1.30096644e-01 3.45748335e-01
-1.31921172e+00 -2.36107826e-01 -1.10024011e+00 8.38735998e-01
3.35536599e-01 1.12322494e-02 5.92382431e-01 1.07173860e+00
-1.01700139e+00 -6.51768893e-02 -4.31035340e-01 -2.69504935e-01
5.35140097e-01 -9.25543532e-02 4.42853689e-01 -4.12173986e-01
-6.71505451e-01 2.75244325e-01 -2.94012994e-01 1.34321958e-01
1.60355404e-01 -1.12009656e+00 -5.13646305e-01 8.10931474e-02
2.69398868e-01 -8.48717749e-01 1.42212605e+00 2.35064626e-01
-1.31098425e+00 2.10767204e-04 -5.27607143e-01 -2.81060040e-01
-3.26323301e-01 -1.77949011e-01 -9.83808458e-01 -8.02264735e-02
2.75123894e-01 -3.20146114e-01 3.73483568e-01 -1.24490857e+00
-1.12442708e+00 -7.92149529e-02 -2.85830945e-01 3.27650636e-01
-3.72529805e-01 3.22930850e-02 -8.70564282e-02 -6.30408466e-01
6.01225078e-01 -7.33786523e-01 -5.87187946e-01 -5.40593565e-02
-4.81675148e-01 3.76091227e-02 7.00847268e-01 3.34638298e-01
1.83368587e+00 -1.98928201e+00 -4.17024493e-01 3.47419649e-01
3.35523725e-01 1.01289347e-01 -2.54432801e-02 7.16946304e-01
-2.02287629e-01 -5.17736793e-01 8.26751068e-02 4.73656096e-02
-6.42733932e-01 -2.85198539e-01 -2.35972077e-01 6.44332051e-01
-7.91098714e-01 6.86719000e-01 -8.31789196e-01 1.69059545e-01
2.06240565e-01 8.05776045e-02 -6.46620393e-01 8.16216320e-02
2.15570569e-01 8.95008743e-01 -8.98911297e-01 8.36736023e-01
1.33193493e+00 -5.26040733e-01 -2.15227127e-01 -3.69295299e-01
-8.65218341e-01 -4.64994997e-01 -1.24689877e+00 1.37468278e+00
-5.80857754e-01 5.30141652e-01 6.56902254e-01 -7.23357797e-01
5.53758740e-01 3.55103910e-01 6.33587420e-01 -1.06296039e+00
5.82338013e-02 4.42600161e-01 5.15175387e-02 -4.22610521e-01
-5.45311272e-02 -3.51146996e-01 1.30301833e-01 2.61768788e-01
-2.64652222e-01 -3.86569411e-01 9.67127532e-02 -4.77806246e-03
1.36431766e+00 -6.11744285e-01 5.14310777e-01 -7.57315457e-01
6.36522293e-01 -6.78133368e-02 2.61674225e-01 9.45339859e-01
4.43529218e-01 4.16253209e-01 -7.67684758e-01 -1.56123161e-01
-3.98966402e-01 -1.06561315e+00 -5.41731954e-01 6.44961953e-01
7.23172545e-01 -5.81512213e-01 -3.05427104e-01 1.22713692e-01
-8.71605054e-02 8.02122533e-01 1.75369546e-01 4.30293620e-01
-5.74876845e-01 -1.10520625e+00 -7.55063444e-02 4.07242000e-01
5.25230229e-01 -2.18533397e-01 -6.47264540e-01 2.48567641e-01
-2.15849996e-01 -1.25837886e+00 -4.25514355e-02 -4.73673362e-03
-3.47743124e-01 -1.00095201e+00 -8.79625082e-01 -5.68896234e-01
5.08280396e-01 8.57579112e-01 1.05254149e+00 -3.72304171e-01
-4.16351974e-01 7.34049737e-01 -5.76032579e-01 -7.61586130e-01
2.33586848e-01 -6.28524482e-01 4.12555367e-01 1.45110413e-01
-1.79315567e-01 -8.04332256e-01 -1.04388964e+00 6.96558774e-01
-4.52241264e-02 -1.96424574e-02 8.47256839e-01 5.86424351e-01
4.34309840e-01 6.82427049e-01 5.47150254e-01 -4.54468429e-01
3.40601295e-01 -4.10971105e-01 -9.09632504e-01 -3.47483121e-02
-3.24193239e-01 -5.16990781e-01 3.90397102e-01 2.52498090e-01
-1.41096830e+00 -2.37255901e-01 -6.65483952e-01 3.37949425e-01
-9.52831283e-02 2.74640799e-01 -1.60139799e-01 -7.61926234e-01
1.02374816e+00 -5.02362661e-02 -6.61866426e-01 -4.18256670e-01
8.07326138e-02 7.33770609e-01 4.34222788e-01 1.37499660e-01
1.55059850e+00 7.92781651e-01 5.93261242e-01 -1.36872542e+00
-1.42782080e+00 -9.28707063e-01 -1.62405461e-01 -2.70293534e-01
7.26616502e-01 -1.09736538e+00 -5.15631199e-01 1.83889881e-01
-1.05212295e+00 -2.76503265e-01 -3.06129634e-01 9.58414137e-01
-2.54907310e-01 -4.69814390e-02 5.71190789e-02 -1.02192080e+00
-1.04616635e-01 -6.16415918e-01 1.14371383e+00 3.76513094e-01
-1.26665467e-02 -1.03926492e+00 4.01551202e-02 1.31402969e-01
1.07805300e+00 -5.73044084e-02 5.33469141e-01 3.75896275e-01
-8.00521255e-01 -1.08965337e-01 -4.95558186e-03 -4.82572526e-01
4.00957853e-01 -1.17686284e+00 -8.36352170e-01 -7.76272595e-01
4.32111740e-01 6.56207576e-02 2.77808279e-01 1.34558237e+00
1.21377432e+00 -1.15248092e-01 -1.39139938e+00 1.20680225e+00
1.58718073e+00 4.16379422e-01 5.99635422e-01 -1.01758949e-01
2.62402087e-01 1.92437291e-01 9.30184424e-01 5.80640197e-01
-1.14455521e-01 7.58747458e-01 4.65273827e-01 -4.10033464e-01
-4.63681728e-01 3.53576422e-01 -4.76788074e-01 3.86205912e-01
1.67910412e-01 -1.20255244e+00 -4.95117098e-01 8.92010853e-02
-1.39932644e+00 -1.08805370e+00 -7.07378745e-01 2.10284686e+00
1.51853606e-01 -1.86177954e-01 -3.96256000e-01 6.60334975e-02
5.61230719e-01 1.78242490e-01 -1.56279266e-01 6.14654899e-01
-2.56700069e-01 3.96645099e-01 7.01823533e-01 8.46893311e-01
-1.02890253e+00 1.51660874e-01 7.49574804e+00 9.24637198e-01
-9.50649619e-01 2.59190261e-01 -2.25236211e-02 -2.91152269e-01
-5.12115657e-01 -2.92169303e-01 -1.23024201e+00 2.42305875e-01
2.66975164e-01 6.85254335e-02 -1.47706628e-01 3.37283939e-01
2.03549773e-01 -4.70530689e-01 -8.34560633e-01 1.60493565e+00
-2.32813671e-01 -1.70239711e+00 -4.73098963e-01 2.12613538e-01
9.46108997e-01 -1.65467680e-01 -1.58289466e-02 -2.80014455e-01
-1.51424125e-01 -9.89784062e-01 2.89372414e-01 5.05122185e-01
1.12968016e+00 -5.37636578e-01 3.78698528e-01 8.21750641e-01
-1.64341807e+00 -5.32986641e-01 -3.18317056e-01 -3.87174100e-01
8.17064226e-01 1.63538647e+00 -4.49032784e-01 9.68675375e-01
9.04115081e-01 5.49007654e-01 -1.61346927e-01 1.55758405e+00
4.85567283e-03 7.43973613e-01 -5.39440930e-01 -1.04424745e-01
-1.13840654e-01 -1.82617068e-01 9.11580384e-01 9.07440662e-01
1.26036799e+00 5.24055660e-01 2.42953137e-01 4.90993947e-01
4.78412241e-01 -1.99231729e-01 -6.48861706e-01 4.43960279e-01
8.38253498e-01 1.29394925e+00 -4.05615360e-01 1.68808833e-01
-8.66731584e-01 2.94347525e-01 -5.93844533e-01 6.29564643e-01
-4.16221797e-01 -4.54128593e-01 6.65696084e-01 7.22237706e-01
2.31381163e-01 -5.08247972e-01 -3.62286538e-01 -5.29140711e-01
-1.47573307e-01 -7.90733024e-02 1.04656247e-02 -9.19177830e-01
-1.12277842e+00 4.88262683e-01 1.55209392e-01 -1.60187495e+00
6.65562286e-04 -3.50996554e-01 -5.99445820e-01 9.79169786e-01
-1.79512691e+00 -8.28014135e-01 -7.06055045e-01 3.78784090e-01
3.88260931e-01 -3.13406616e-01 6.76039636e-01 4.29082543e-01
2.27450207e-01 2.67617881e-01 4.38907444e-01 -3.19266409e-01
4.81718063e-01 -8.00677121e-01 1.72114789e-01 1.06916559e+00
-2.36788288e-01 2.83751458e-01 9.10928130e-01 -7.17807114e-01
-1.64691091e+00 -1.36337781e+00 5.02320647e-01 -9.62186083e-02
2.86245167e-01 -6.22352660e-01 8.03184211e-02 4.05263484e-01
4.45700437e-01 4.70586240e-01 9.06325459e-01 1.81328990e-02
4.68256176e-01 -2.44071111e-01 -8.05303335e-01 4.02082860e-01
1.97837186e+00 4.88362700e-01 2.57628441e-01 1.02170479e+00
1.84038684e-01 -6.66792572e-01 -4.80526954e-01 1.17758107e+00
3.85345638e-01 -1.18374836e+00 1.19187415e+00 5.09524822e-01
-3.35463494e-01 -5.00888824e-01 -4.29125696e-01 -1.59084094e+00
-1.31787395e+00 -9.18013394e-01 -6.37458116e-02 8.43783498e-01
1.72671434e-02 -7.73682237e-01 1.09958875e+00 -7.63298035e-01
-5.74056268e-01 -7.86749125e-01 -8.14196527e-01 -9.31399822e-01
-2.96532094e-01 -4.53538507e-01 4.33381855e-01 4.53581363e-01
-1.81053579e-01 7.46578038e-01 -1.92346618e-01 1.10344017e+00
1.11093295e+00 4.04509842e-01 6.62934780e-01 -1.48632717e+00
-5.73039949e-01 -2.43836150e-01 -1.60845876e-01 -1.72414196e+00
-4.61487234e-01 -5.58223248e-01 2.47451365e-01 -2.06530476e+00
-2.42024049e-01 -9.42534566e-01 4.38375890e-01 -1.92356065e-01
1.59828767e-01 8.64866018e-01 -4.38740194e-01 -1.67043641e-01
-3.54611009e-01 3.88273895e-01 1.51529217e+00 2.08729163e-01
6.00144975e-02 1.09098697e+00 -9.25894022e-01 6.71104133e-01
4.83337402e-01 6.86256140e-02 -7.69624472e-01 -5.92086732e-01
4.99271780e-01 4.36190218e-01 8.29811841e-02 -1.85072792e+00
8.43848825e-01 -3.29028010e-01 7.49110818e-01 -9.07723546e-01
5.15125930e-01 -1.00299549e+00 1.82475761e-01 4.44500834e-01
5.11602759e-01 -6.10446870e-01 9.28820763e-03 7.20244706e-01
-1.12435676e-01 -9.26505253e-02 1.04340398e+00 -3.22493017e-02
-5.50093353e-01 3.10716659e-01 -6.17689371e-01 -2.73909062e-01
1.51615047e+00 -3.15998852e-01 -4.70056921e-01 -7.14021504e-01
-5.25125682e-01 5.06961763e-01 -2.48044476e-01 -8.76245424e-02
5.64098835e-01 -1.47805607e+00 -7.87740111e-01 4.67000961e-01
3.02997380e-01 3.43051814e-02 5.56760311e-01 8.15719664e-01
-1.86058417e-01 9.98885930e-01 3.47127944e-01 -9.06350315e-01
-1.32780230e+00 7.40978718e-02 7.09845006e-01 1.92943528e-01
-4.49279368e-01 1.49708784e+00 7.55434811e-01 -8.19284841e-02
9.53503251e-02 7.12600071e-03 -3.57198894e-01 -6.47709072e-01
8.54018986e-01 3.00233305e-01 3.66198838e-01 -3.46608728e-01
-4.55633610e-01 1.35778725e+00 8.21759284e-01 6.86701760e-02
1.26248705e+00 -5.45681477e-01 -2.10184790e-03 -1.35182738e-01
4.34002757e-01 1.03127289e+00 -7.67440319e-01 -3.48149836e-01
-5.93238294e-01 -7.38646448e-01 3.46294791e-01 -8.58975351e-01
-5.95780909e-01 5.35150528e-01 7.53447056e-01 1.90564945e-01
1.53034413e+00 4.24130678e-01 3.72531563e-01 5.36637843e-01
8.86298954e-01 -4.38263237e-01 2.27484420e-01 5.87847590e-01
8.56558979e-01 -5.97882569e-01 3.82861346e-01 -1.33247435e+00
4.94375259e-01 9.27202284e-01 5.27528763e-01 3.24737459e-01
1.32480311e+00 1.08150387e+00 -1.32782549e-01 -4.45642143e-01
-2.66229540e-01 -1.61106318e-01 2.57577181e-01 9.75554287e-01
5.24022460e-01 -2.14315996e-01 -1.98203370e-01 4.10397619e-01
-4.90131795e-01 -1.60773054e-01 3.29150766e-01 1.15728414e+00
-9.93168890e-01 -1.21066213e+00 -1.08416462e+00 6.57332242e-01
-6.33103400e-02 4.20275144e-02 4.56129044e-01 2.38836005e-01
3.33855391e-01 1.52602756e+00 1.39595225e-01 -1.53923988e-01
5.76236665e-01 -5.51347911e-01 6.83418751e-01 -1.09246790e+00
2.94459313e-01 4.72335100e-01 3.26780640e-02 -4.63328928e-01
-4.61540937e-01 -1.99997947e-01 -6.89299524e-01 2.27337793e-01
-5.23997784e-01 5.39555490e-01 5.08885205e-01 9.48465288e-01
3.93185914e-01 9.22747970e-01 7.70091832e-01 -9.45345819e-01
-1.64271712e-01 -8.53739381e-01 -8.98955286e-01 -6.17711365e-01
5.10714054e-01 -7.54394710e-01 -1.94733560e-01 -4.81204480e-01] | [6.385447025299072, 1.2423003911972046] |
01fa99f8-1d0c-4bac-8518-3806db71da10 | rethinking-surgical-instrument-segmentation-a | 2206.11804 | null | https://arxiv.org/abs/2206.11804v4 | https://arxiv.org/pdf/2206.11804v4.pdf | Rethinking Surgical Instrument Segmentation: A Background Image Can Be All You Need | Data diversity and volume are crucial to the success of training deep learning models, while in the medical imaging field, the difficulty and cost of data collection and annotation are especially huge. Specifically in robotic surgery, data scarcity and imbalance have heavily affected the model accuracy and limited the design and deployment of deep learning-based surgical applications such as surgical instrument segmentation. Considering this, we rethink the surgical instrument segmentation task and propose a one-to-many data generation solution that gets rid of the complicated and expensive process of data collection and annotation from robotic surgery. In our method, we only utilize a single surgical background tissue image and a few open-source instrument images as the seed images and apply multiple augmentations and blending techniques to synthesize amounts of image variations. In addition, we also introduce the chained augmentation mixing during training to further enhance the data diversities. The proposed approach is evaluated on the real datasets of the EndoVis-2018 and EndoVis-2017 surgical scene segmentation. Our empirical analysis suggests that without the high cost of data collection and annotation, we can achieve decent surgical instrument segmentation performance. Moreover, we also observe that our method can deal with novel instrument prediction in the deployment domain. We hope our inspiring results will encourage researchers to emphasize data-centric methods to overcome demanding deep learning limitations besides data shortage, such as class imbalance, domain adaptation, and incremental learning. Our code is available at https://github.com/lofrienger/Single_SurgicalScene_For_Segmentation. | ['Hongliang Ren', 'Mengya Xu', 'Mobarakol Islam', 'An Wang'] | 2022-06-23 | null | null | null | null | ['scene-segmentation'] | ['computer-vision'] | [ 1.66820604e-02 2.22601235e-01 -4.85022277e-01 -2.93146431e-01
-6.92196131e-01 -6.47938788e-01 5.80620430e-02 -2.93926708e-02
-4.46077913e-01 5.59292853e-01 -6.56601116e-02 -6.30831420e-01
-2.09954217e-01 -4.81366009e-01 -6.70489132e-01 -8.38510275e-01
3.34006608e-01 4.38680500e-01 -1.40631884e-01 -1.12055756e-01
1.25098694e-02 4.43367779e-01 -1.20493269e+00 1.61530580e-02
1.13822174e+00 1.15578890e+00 3.82964194e-01 3.80066216e-01
-2.76168138e-01 6.33859277e-01 -5.12253702e-01 -3.45467687e-01
7.77684152e-01 -1.27706021e-01 -5.80952287e-01 2.61268348e-01
2.58369505e-01 -4.70722288e-01 -4.14654672e-01 9.26518738e-01
9.46154952e-01 -2.56635338e-01 3.54200006e-01 -1.11001658e+00
-2.64157057e-01 6.27853811e-01 -7.29154885e-01 3.63620594e-02
-4.67679650e-01 4.62784708e-01 5.14341652e-01 -8.41748178e-01
7.18009353e-01 4.67290401e-01 8.21019709e-01 6.67781532e-01
-7.80450225e-01 -9.04219329e-01 4.65438701e-02 -1.74811915e-01
-1.09643984e+00 -2.79260039e-01 9.57029879e-01 -7.37172842e-01
1.12656243e-01 2.10560575e-01 8.30835164e-01 1.14712369e+00
2.42278859e-01 1.04653525e+00 7.95093298e-01 -2.09680706e-01
-4.75387834e-02 7.24267215e-02 -1.06170855e-01 9.25614715e-01
4.95767087e-01 3.54985483e-02 2.24984623e-02 1.07291400e-01
1.12162268e+00 3.75818700e-01 -2.70191342e-01 -6.47016287e-01
-1.36215127e+00 6.56681001e-01 5.80837011e-01 2.32087016e-01
-3.99616420e-01 -9.13465396e-02 7.52736270e-01 4.19505835e-02
1.98290512e-01 7.65876472e-01 -6.31700218e-01 -8.45016688e-02
-8.31797063e-01 -7.15781674e-02 6.80247962e-01 1.30727053e+00
5.03412306e-01 -1.20919190e-01 -1.70060322e-01 9.55335915e-01
-2.02929854e-01 6.58066347e-02 8.30276370e-01 -8.04206908e-01
4.35418785e-01 7.12694407e-01 -1.22846365e-01 -8.27000916e-01
-8.49058986e-01 -8.56861293e-01 -1.08604050e+00 -1.62216783e-01
4.22709852e-01 -4.83865380e-01 -1.26698196e+00 1.35160244e+00
4.46050853e-01 3.48293036e-02 -2.19453931e-01 1.04522336e+00
1.10808599e+00 1.56082101e-02 -5.45512699e-02 -2.30961174e-01
1.27194536e+00 -1.13472402e+00 -7.62721360e-01 -2.24383235e-01
1.01779521e+00 -8.11118305e-01 1.33889079e+00 5.41624665e-01
-8.94121408e-01 -5.16797662e-01 -1.06489122e+00 -1.96226284e-01
1.22499736e-02 6.63766921e-01 1.14018214e+00 4.76560622e-01
-4.04938042e-01 2.73000240e-01 -1.10266471e+00 -7.43263587e-02
1.04155207e+00 5.83266795e-01 -2.23060340e-01 -1.71837926e-01
-6.72492981e-01 5.69000959e-01 4.26030636e-01 3.51258993e-01
-9.04713869e-01 -9.93212640e-01 -5.33722103e-01 -3.48962933e-01
7.42765188e-01 -6.97920740e-01 1.24517488e+00 -9.64835823e-01
-1.37323582e+00 1.03040671e+00 4.50248986e-01 -1.77723825e-01
6.74041331e-01 -1.47976056e-01 -6.20029904e-02 -1.01615742e-01
-1.66169688e-01 5.44501722e-01 5.38724065e-01 -1.46206856e+00
-4.68792558e-01 -3.85402143e-01 9.20943096e-02 2.25428686e-01
-4.80964035e-01 -5.05046129e-01 -7.85833776e-01 -7.15868652e-01
2.78449923e-01 -1.25478756e+00 -5.66763699e-01 2.73216873e-01
-4.99991000e-01 3.50301743e-01 4.79514897e-01 -7.58794010e-01
1.15709746e+00 -2.38720465e+00 1.10502969e-02 1.75101031e-02
4.60859329e-01 3.27107459e-01 -4.56164405e-02 -1.64403349e-01
6.59364415e-03 1.44795284e-01 -1.92787349e-01 -2.72872239e-01
-3.82274628e-01 4.18436825e-01 7.74330050e-02 4.45218295e-01
-5.40593155e-02 8.79296482e-01 -7.61910737e-01 -8.64068210e-01
1.86173290e-01 1.12859212e-01 -9.77150738e-01 2.11250082e-01
-9.07560214e-02 9.45080400e-01 -5.62663615e-01 9.94258165e-01
7.72749305e-01 -2.89696902e-01 1.46386236e-01 -6.25060439e-01
1.55981734e-01 -2.42594676e-03 -9.12988782e-01 2.31299567e+00
-6.94676101e-01 2.14346334e-01 2.24329680e-01 -1.00421691e+00
8.78552675e-01 1.65434241e-01 9.97785151e-01 -6.15348339e-01
5.28430820e-01 4.25734580e-01 4.47703242e-01 -8.24446917e-01
4.19378787e-01 -1.49618402e-01 -2.77445372e-02 -1.32748798e-01
1.17882118e-01 -4.16278720e-01 8.08700621e-02 1.08372432e-03
8.29957724e-01 -7.51969144e-02 1.68437824e-01 -1.91869453e-01
1.22600175e-01 4.09039617e-01 8.70180786e-01 6.14514470e-01
-4.39431757e-01 7.73273945e-01 6.58548534e-01 -5.22508442e-01
-1.07759547e+00 -8.98530126e-01 -2.89090633e-01 8.49765420e-01
3.28773022e-01 2.11236365e-02 -6.02466881e-01 -8.34303260e-01
7.68458890e-03 2.29763433e-01 -7.44163334e-01 -1.90557033e-01
-6.17040157e-01 -9.28799808e-01 3.72071981e-01 5.19367456e-01
2.30798960e-01 -8.19098651e-01 -3.75441849e-01 1.56463459e-01
-1.84706956e-01 -1.14722979e+00 -5.05991042e-01 2.50978917e-01
-1.09868073e+00 -1.16655540e+00 -7.15698957e-01 -8.82148743e-01
9.18551862e-01 1.22866981e-01 8.98026824e-01 1.89522907e-01
-7.67166018e-01 1.92381963e-02 -3.04176509e-01 -8.78243327e-01
-3.83931130e-01 5.82100570e-01 -1.76595122e-01 -2.94183671e-01
-3.63966115e-02 -3.36736143e-01 -1.10582840e+00 3.56178612e-01
-1.00359452e+00 3.73209000e-01 9.14220989e-01 1.10259581e+00
6.76930666e-01 -2.92232484e-01 4.28477615e-01 -1.11004937e+00
4.15104449e-01 -5.19200742e-01 -5.27294040e-01 4.04478535e-02
-5.73355913e-01 -3.26176167e-01 5.49813032e-01 -6.80831254e-01
-9.32496965e-01 2.42566988e-01 -1.00348100e-01 -7.03378141e-01
4.04403023e-02 6.33033633e-01 2.14313958e-02 -1.27036408e-01
8.15648377e-01 -1.43403709e-01 4.82775062e-01 -2.91448534e-01
2.10748523e-01 6.12608969e-01 4.90269989e-01 -5.86990833e-01
4.35493946e-01 4.41960901e-01 -1.50451586e-01 -3.71536255e-01
-8.33980203e-01 -4.32507366e-01 -6.08754992e-01 -2.47016579e-01
6.40986264e-01 -9.91824746e-01 -5.67895412e-01 4.51954722e-01
-8.09755087e-01 -4.54493344e-01 -4.10918295e-01 5.67653000e-01
-3.57668012e-01 6.99967071e-02 -6.32501841e-01 -2.52761662e-01
-5.43166518e-01 -1.69719923e+00 8.77488315e-01 4.26430523e-01
2.01582104e-01 -8.61459672e-01 -2.93549895e-01 5.93138516e-01
3.79695266e-01 4.25465345e-01 7.70413280e-01 -6.69909954e-01
-7.01415122e-01 -4.23233777e-01 -1.73281297e-01 4.35873270e-01
3.83054256e-01 -5.80133051e-02 -6.75499678e-01 -3.66511256e-01
-3.14157177e-03 -3.04507554e-01 5.12103200e-01 5.40419877e-01
1.97166038e+00 -5.81645332e-02 -3.55497956e-01 1.04915702e+00
1.34489882e+00 1.32507041e-01 4.06752706e-01 2.95322865e-01
1.07410645e+00 5.06894708e-01 8.68980527e-01 6.05394304e-01
4.42644686e-01 4.45852131e-01 6.50093257e-01 -6.46682262e-01
-1.76176757e-01 -5.63410707e-02 -4.10142392e-01 1.03412950e+00
-1.52325749e-01 2.51097232e-02 -1.17337775e+00 7.09659815e-01
-1.59132981e+00 -3.59451205e-01 -4.81547341e-02 2.05288363e+00
1.16189599e+00 -7.80235231e-02 -1.32058889e-01 -2.71233231e-01
4.02955025e-01 -1.03659555e-01 -8.40967178e-01 7.06835836e-02
2.26914644e-01 1.35943681e-01 9.20027256e-01 1.16470736e-02
-1.06316066e+00 8.90624583e-01 5.09161425e+00 8.88093412e-01
-1.62000883e+00 2.56904274e-01 8.56061935e-01 -4.44939673e-01
-9.48751643e-02 -5.07170022e-01 -5.24221897e-01 5.03832519e-01
3.83417428e-01 -5.34819216e-02 2.02820688e-01 1.14252460e+00
1.03192322e-01 3.98894772e-02 -9.74790692e-01 1.18295062e+00
-1.07355408e-01 -1.63834155e+00 -2.34542921e-01 1.20578803e-01
8.68580639e-01 1.59433156e-01 1.53843716e-01 3.92535150e-01
-7.20126107e-02 -9.14151192e-01 2.94881672e-01 3.34261805e-01
8.53485584e-01 -4.30462778e-01 9.34967756e-01 1.97436318e-01
-6.76588833e-01 -2.88583308e-01 -2.46112764e-01 3.69579136e-01
-1.47145614e-01 7.51787603e-01 -1.14994407e+00 7.15793788e-01
5.84597230e-01 4.52463835e-01 -4.75054950e-01 1.19910073e+00
1.20996110e-01 3.40236276e-01 -1.82067990e-01 3.19973081e-01
5.41858748e-03 -6.46035299e-02 1.09902985e-01 8.84640813e-01
3.74856055e-01 8.03554878e-02 2.68666327e-01 5.65259755e-01
-3.73606175e-01 2.72486955e-01 -4.10988659e-01 9.57129896e-02
4.61666316e-01 1.48328125e+00 -6.36023998e-01 -1.37425363e-01
-4.67631668e-01 5.70421159e-01 6.19580708e-02 1.18301369e-01
-9.60620463e-01 -3.39310989e-03 5.83681643e-01 1.62568226e-01
-8.93549901e-03 -1.97714433e-01 -9.95070994e-01 -1.19885540e+00
6.88156113e-02 -1.00732005e+00 3.27313960e-01 -3.54200721e-01
-1.08948982e+00 5.26923001e-01 -2.35699013e-01 -1.58098781e+00
7.88035244e-02 -6.96254373e-01 -2.92095602e-01 5.08579671e-01
-1.43951869e+00 -1.19931912e+00 -8.27041209e-01 5.44619143e-01
6.06550515e-01 -1.85748994e-01 4.35761094e-01 6.57337308e-01
-7.56452024e-01 8.19890261e-01 1.29826635e-01 3.65763903e-01
9.61948752e-01 -9.13360894e-01 -5.55870570e-02 4.33562338e-01
-2.92999625e-01 6.12189054e-01 4.30018693e-01 -4.67458576e-01
-1.50356221e+00 -1.00976980e+00 -9.64767188e-02 -3.44013482e-01
5.74494421e-01 -3.63768846e-01 -7.68122435e-01 7.97093987e-01
-1.83977440e-01 2.86105812e-01 8.92770112e-01 1.13877155e-01
2.66926587e-01 -3.71927649e-01 -1.22323692e+00 7.51723170e-01
1.13777661e+00 3.80104259e-02 -1.77979231e-01 6.37321174e-01
7.44753122e-01 -1.27915585e+00 -1.18408298e+00 8.91037047e-01
6.21047318e-01 -6.98290348e-01 9.14189696e-01 -4.45942491e-01
7.08842814e-01 1.10602295e-02 1.59822702e-01 -1.08610630e+00
7.74967447e-02 -4.19935554e-01 1.86595038e-01 9.19942021e-01
5.79504490e-01 -5.74304760e-01 1.08729219e+00 7.98391044e-01
-6.91511750e-01 -1.23912740e+00 -7.73656189e-01 -4.77039039e-01
3.55314314e-01 -3.07426780e-01 5.87141514e-01 1.23998129e+00
-2.74029136e-01 -2.25283086e-01 -3.60156476e-01 9.72329080e-02
3.60573411e-01 3.73893864e-02 1.10854089e+00 -1.08872771e+00
-2.47259051e-01 -4.07277077e-01 -1.32574812e-01 -7.88753092e-01
-3.02316815e-01 -8.70265305e-01 1.43114164e-01 -1.48146474e+00
2.49655053e-01 -9.77376223e-01 -3.61064553e-01 5.25600016e-01
-2.73402154e-01 8.26284066e-02 2.85361767e-01 5.57401717e-01
-1.90766931e-01 4.26915050e-01 1.96432424e+00 -4.89956588e-02
-3.65913242e-01 1.69299036e-01 -8.62600803e-01 7.47865975e-01
8.95914137e-01 -3.69094104e-01 -5.13738751e-01 -6.71029508e-01
-8.29483718e-02 1.21853292e-01 1.53892592e-01 -9.12442029e-01
2.93831348e-01 -8.88233334e-02 3.09262574e-01 -4.04090673e-01
-2.87342351e-02 -1.06730938e+00 -5.77465668e-02 5.62933624e-01
-1.04586035e-01 -1.82977602e-01 4.68941689e-01 1.54204145e-01
-1.71175674e-01 -8.79578888e-02 8.34861994e-01 -2.48033464e-01
-5.24463832e-01 7.73927867e-01 2.48833269e-01 1.04318500e-01
1.09656596e+00 -1.71711102e-01 -3.15832764e-01 1.30432978e-01
-7.55819380e-01 3.56527001e-01 4.68612224e-01 6.47405028e-01
4.31641906e-01 -1.10426390e+00 -5.57095230e-01 2.22323760e-01
2.33160421e-01 7.87928522e-01 7.14679897e-01 1.33515584e+00
-8.29259098e-01 3.40810120e-02 -3.14199954e-01 -8.31072450e-01
-8.95358682e-01 5.93400836e-01 4.75078017e-01 -3.90893877e-01
-6.01830184e-01 1.02588999e+00 4.78544474e-01 -8.05786729e-01
4.33675826e-01 -6.15887403e-01 2.22305119e-01 9.83910635e-04
1.93057656e-02 -9.72722098e-02 1.53574705e-01 9.13236439e-02
-2.79522896e-01 3.02832931e-01 -4.16738629e-01 4.01481062e-01
1.24652493e+00 6.63048178e-02 9.66650769e-02 1.87970772e-01
1.08691704e+00 -1.28522605e-01 -1.39862764e+00 -1.10468000e-01
-3.63757432e-01 -5.21870017e-01 1.56723484e-01 -1.00715804e+00
-1.56777453e+00 7.76104748e-01 7.56503522e-01 -2.48472124e-01
1.29885983e+00 -4.75208126e-02 8.71848583e-01 1.28320381e-02
3.57887208e-01 -1.10598052e+00 1.03103764e-01 1.55279666e-01
9.98878658e-01 -1.70133066e+00 1.15322940e-01 -5.87084234e-01
-9.02540267e-01 8.85912657e-01 9.33437288e-01 1.19146004e-01
5.53138733e-01 5.99976599e-01 3.78606886e-01 -1.95494220e-01
-3.35439205e-01 1.40379474e-01 1.17746174e-01 4.07728642e-01
5.15646338e-01 1.31473616e-01 -4.98277336e-01 8.09628189e-01
-2.57580549e-01 3.64859581e-01 5.15026510e-01 9.04174387e-01
2.47832667e-02 -9.87713575e-01 -8.97946581e-02 8.08386445e-01
-4.68649149e-01 -1.91525653e-01 1.53719589e-01 1.03414273e+00
3.04033756e-01 4.99195278e-01 -1.81376487e-02 -3.01827788e-01
5.16809464e-01 -3.49011570e-01 4.26661819e-01 -6.11451447e-01
-5.96719205e-01 1.78875670e-01 -1.07479401e-01 -5.24050593e-01
-1.59380674e-01 -4.03824359e-01 -1.35067368e+00 -1.26948431e-01
-3.63877326e-01 -1.51193812e-01 1.06361663e+00 7.08536029e-01
5.17336130e-01 1.16282654e+00 6.19511604e-01 -8.77166271e-01
-6.01524889e-01 -9.54635084e-01 -4.15083498e-01 3.65134180e-01
3.23604792e-01 -6.62388682e-01 -1.33890212e-01 9.68644097e-02] | [14.553197860717773, -2.2471048831939697] |
f3e7ed18-21b5-421d-8c07-17c01c2ac0c5 | linking-convolutional-neural-networks-with | null | null | https://www.researchgate.net/publication/335620542_Linking_convolutional_neural_networks_with_graph_convolutional_networks_application_in_pulmonary_artery-vein_separation | https://bit.ly/2kNpbdv | Linking convolutional neural networks with graph convolutional networks: application in pulmonary artery-vein separation | Graph Convolutional Networks (GCNs) are a novel and powerful method for dealing with non-Euclidean data, while Convolutional Neural Networks (CNNs) can learn features from Euclidean data such as images. In this work, we propose a novel method to combine CNNs with GCNs (CNN-GCN), that can consider both Euclidean and non-Euclidean features and can be trained end-to-end. We applied this method to separate the pulmonary vascular trees into arteries and veins (A/V). Chest CT scans were pre-processed by vessel segmentation and skeletonization, from which a graph was constructed: voxels on the skeletons resulting in a vertex set and their connections in an adjacency matrix. 3D patches centered around each vertex were extracted from the CT scans, oriented perpendicularly to the vessel. The proposed CNN-GCN classifier was trained and applied on the constructed vessel graphs, where each node is then labeled as artery or vein. The proposed method was trained and validated on data from one hospital (11 patient, 22 lungs), and tested on independent data from a different hospital (10 patients, 10 lungs). A baseline CNN method and human observer performance were used for comparison. The CNN-GCN method obtained a median accuracy of 0.773 (0.738) in the validation (test) set, compared to a median accuracy of 0.817 by the observers, and 0.727 (0.693) by the CNN. In conclusion, the proposed CNN-GCN method combines local image information with graph connectivity information, improving pulmonary A/V separation over a baseline CNN method, approaching the performance of human observers. | ['Boudewijn P. F. Lelieveldt', 'Berend C. Stoel', 'M. Els Bakker', 'Marius Staring', 'Lucia J. Kroft', 'Gudula J.A.M. Boon', 'Xiaojuan Xiao', 'Qiuxia Xie', 'Zhiwei Zhai', 'Xuhui Zhou', 'Frederikus A. Klok'] | 2019-09-01 | null | null | null | preprint-2019-9 | ['pulmonary-arteryvein-classification', 'pulmorary-vessel-segmentation', '3d-medical-imaging-segmentation'] | ['computer-vision', 'computer-vision', 'medical'] | [-5.74324057e-02 4.68316019e-01 4.18536291e-02 -1.23370783e-02
-1.72771081e-01 -3.83736461e-01 2.95730054e-01 3.80552262e-01
-4.95289534e-01 3.08960229e-01 -5.43833971e-02 -4.50371742e-01
-3.83121103e-01 -1.15410721e+00 -2.59910196e-01 -6.89828932e-01
-3.14460337e-01 5.51972270e-01 4.93990511e-01 2.13359788e-01
-9.25592780e-02 1.28641319e+00 -9.32439864e-01 6.18499704e-02
5.81562459e-01 1.16025770e+00 -2.66123146e-01 1.06095910e+00
-1.95468724e-01 8.09201777e-01 -4.42366391e-01 -2.85842896e-01
4.59649265e-01 -5.85560083e-01 -7.40939975e-01 2.39740431e-01
5.83907604e-01 -7.48140141e-02 -5.64649582e-01 6.06868207e-01
5.26885509e-01 5.32658920e-02 9.17698801e-01 -8.77393365e-01
-3.35191995e-01 2.43959084e-01 -3.93652648e-01 2.39908755e-01
-2.97003746e-01 1.03177898e-01 8.67839217e-01 -6.50891006e-01
5.64883769e-01 7.17590153e-01 9.59130287e-01 4.74106103e-01
-1.00322866e+00 -3.38606358e-01 -4.28377151e-01 -3.89795639e-02
-1.43204045e+00 1.97328553e-01 5.28701127e-01 -6.85866773e-01
7.81063795e-01 1.76817775e-01 1.17365766e+00 4.43059623e-01
4.31983083e-01 1.17908850e-01 8.46557081e-01 -3.35001349e-01
2.65484191e-02 -1.12309679e-01 -3.84669490e-02 1.26638865e+00
3.93143952e-01 1.24948323e-01 7.28241354e-02 -2.03495145e-01
1.33028901e+00 1.53184608e-01 -3.53006572e-01 -5.28468132e-01
-1.31848991e+00 9.36920285e-01 1.24685431e+00 5.87389588e-01
-6.03857875e-01 8.24906528e-02 2.91197568e-01 -5.41227162e-02
5.03130257e-02 3.51111144e-01 -5.08863218e-02 2.68784791e-01
-8.48208070e-01 -3.23346555e-01 1.05109048e+00 6.98978484e-01
4.57029998e-01 -3.39357257e-02 -3.24022859e-01 6.58132315e-01
2.35413030e-01 2.56014824e-01 5.47132790e-01 -5.61631143e-01
2.56331533e-01 8.53353798e-01 -7.03684092e-01 -1.12541854e+00
-1.00523508e+00 -7.86077559e-01 -1.44387019e+00 2.38356590e-01
6.96681201e-01 -1.73944011e-01 -9.62211549e-01 1.17142415e+00
2.74755985e-01 1.23835117e-01 -1.87426861e-02 9.31011677e-01
1.31261015e+00 2.19551623e-01 2.86868177e-02 1.37863740e-01
1.28871882e+00 -1.21232212e+00 -2.40910590e-01 3.42166156e-01
8.97382677e-01 -5.31118453e-01 5.91722488e-01 2.00769454e-01
-8.56977940e-01 -5.76951146e-01 -8.82258654e-01 1.99983925e-01
-2.71610320e-01 3.10013145e-01 2.45928541e-01 5.33936560e-01
-1.32051885e+00 9.09388959e-01 -8.80861104e-01 -6.02762818e-01
7.75595486e-01 4.53935623e-01 -6.75605714e-01 1.83026880e-01
-7.62220740e-01 7.40363419e-01 1.50747880e-01 1.92633808e-01
-8.62190366e-01 -4.03657913e-01 -6.01901174e-01 1.59570888e-01
1.30085155e-01 -1.02576184e+00 8.89531791e-01 -7.54721165e-01
-1.38282311e+00 9.28933799e-01 4.19096082e-01 -4.91923273e-01
7.34246492e-01 3.88895333e-01 -2.77016461e-01 6.68008149e-01
-1.03889331e-01 4.42572147e-01 6.19028926e-01 -9.19422388e-01
-4.43990320e-01 -2.88677514e-01 -3.57100993e-01 -8.22588056e-02
-1.73170730e-01 -3.08824807e-01 -3.01886290e-01 -6.07025266e-01
4.03383166e-01 -9.52707887e-01 -4.41776335e-01 5.41421533e-01
-6.95181787e-01 -2.08053857e-01 6.32520854e-01 -7.27116406e-01
1.00192118e+00 -1.88959265e+00 1.05424963e-01 8.14895511e-01
1.03955138e+00 4.54354644e-01 -4.31471504e-02 2.72353679e-01
-2.82419652e-01 2.07606077e-01 -5.57806313e-01 -4.80222888e-02
-6.79671168e-01 2.72027314e-01 7.69062817e-01 6.49184704e-01
5.87343425e-02 1.00058043e+00 -8.29084158e-01 -8.68147969e-01
3.86124790e-01 6.30008221e-01 -2.20269874e-01 2.89508581e-01
3.63970160e-01 5.80792725e-01 -3.83381546e-01 4.51400578e-01
5.38984299e-01 -4.97419536e-01 1.78163916e-01 -2.42211848e-01
1.39526516e-01 -4.04846728e-01 -8.70284021e-01 1.21930826e+00
-6.42600954e-01 7.52703249e-01 3.07583380e-02 -9.13300812e-01
1.27032936e+00 5.35938501e-01 7.91108727e-01 -3.29472482e-01
5.49304307e-01 2.74650246e-01 4.96114522e-01 -6.64652586e-01
-3.90904278e-01 -1.98908880e-01 5.43760896e-01 1.77506328e-01
1.95013598e-01 -2.68249840e-01 1.57564759e-01 1.18807875e-01
1.29174137e+00 -4.36087519e-01 4.77743566e-01 -2.72091329e-01
9.16714728e-01 2.18649749e-02 1.63108081e-01 5.42905033e-01
-3.36943895e-01 8.82454574e-01 8.90638709e-01 -6.63267970e-01
-8.87134910e-01 -1.13531208e+00 -2.33282030e-01 1.90209836e-01
-2.19242930e-01 -4.24285650e-01 -9.36298788e-01 -1.08198702e+00
-3.60059030e-02 9.44085345e-02 -7.35611141e-01 2.59221382e-02
-7.49368787e-01 -1.64340273e-01 5.72350383e-01 6.65462852e-01
5.68004906e-01 -1.09094393e+00 -5.96422434e-01 3.88375014e-01
8.78454000e-02 -1.11509657e+00 -2.47996494e-01 1.56143680e-01
-1.24298346e+00 -1.62431419e+00 -8.46583426e-01 -9.02894497e-01
8.71192336e-01 1.56297404e-02 1.09060204e+00 5.68645835e-01
-6.38951302e-01 3.24492007e-01 -2.83091992e-01 -4.71388847e-02
-4.78752643e-01 1.38753071e-01 -4.23557699e-01 2.02823162e-01
-1.84227809e-01 -6.19487584e-01 -7.52217770e-01 3.49120706e-01
-6.79335356e-01 -1.90746829e-01 9.24703836e-01 8.65790844e-01
6.96665227e-01 -1.76634341e-02 2.12343752e-01 -1.08647668e+00
4.42369223e-01 -3.93323094e-01 -4.85582262e-01 5.44699281e-02
-5.78803420e-01 -2.90779740e-01 1.03493273e+00 -1.16321802e-01
-5.17460227e-01 2.76285470e-01 -1.21275626e-01 -5.91511726e-01
-2.87851363e-01 5.92788160e-01 2.48933986e-01 -4.67887700e-01
9.33340788e-01 -1.98727891e-01 4.43092197e-01 -2.81363904e-01
2.75852621e-01 5.17431259e-01 5.41470706e-01 -1.76759318e-01
7.60982871e-01 4.57159013e-01 7.99016297e-01 -9.77833211e-01
-5.44925034e-01 -6.16478801e-01 -1.30618322e+00 -4.88917917e-01
1.21517873e+00 -4.37936634e-01 -3.58684689e-01 3.19788814e-01
-9.70990419e-01 -3.40352267e-01 -5.33757210e-01 8.52257907e-01
-3.88001829e-01 5.29700041e-01 -6.19389117e-01 -1.57163367e-01
-6.12955034e-01 -9.82239485e-01 6.17850065e-01 2.70492226e-01
-1.47351772e-01 -1.47786963e+00 6.93854913e-02 2.92687625e-01
6.65793419e-01 7.43169904e-01 1.00408173e+00 -9.73880470e-01
-2.81884909e-01 -5.96524239e-01 -5.06849468e-01 6.28378212e-01
1.68350860e-01 3.38741273e-01 -7.12578595e-01 -1.20635696e-01
-1.50991112e-01 1.65188625e-01 6.43023074e-01 4.70512986e-01
1.33523810e+00 -2.21005499e-01 -3.75013322e-01 7.56727934e-01
1.51618481e+00 7.27128908e-02 5.64417303e-01 7.46896192e-02
1.02944410e+00 4.24451083e-01 -4.23160270e-02 2.73670793e-01
-7.61821494e-02 2.44354293e-01 7.11703062e-01 -7.66249716e-01
-6.07591927e-01 8.31405222e-02 -3.26723099e-01 8.08997691e-01
-6.15237355e-01 -1.80620119e-01 -1.07661760e+00 4.34299618e-01
-1.46532965e+00 -4.33700025e-01 -8.37523878e-01 2.03741241e+00
2.14719474e-01 1.02058299e-01 3.02736044e-01 6.61812872e-02
8.35297525e-01 -5.11097386e-02 -1.35965660e-01 -5.47397196e-01
1.79708183e-01 8.48280489e-01 7.08990753e-01 1.97177589e-01
-1.12265968e+00 4.45053369e-01 5.62500525e+00 4.65350360e-01
-1.45708656e+00 -7.65787289e-02 5.35025358e-01 2.85239160e-01
3.38588268e-01 -2.16237739e-01 -1.68013111e-01 -9.17222425e-02
9.65327799e-01 -7.72179663e-03 8.65520611e-02 9.18562293e-01
-5.73771726e-03 1.42401934e-01 -8.85401011e-01 8.70295942e-01
3.09737753e-02 -1.22303009e+00 -3.18656638e-02 2.26613104e-01
6.28641546e-01 1.22570679e-01 -2.94033170e-01 -8.72384682e-02
-1.05598114e-01 -1.34157538e+00 1.22433096e-01 5.29896677e-01
8.49762797e-01 -4.44796264e-01 1.08160222e+00 2.38732308e-01
-1.45405817e+00 1.35316938e-01 -3.22237790e-01 2.37077206e-01
-4.89560068e-02 5.34093559e-01 -1.34079945e+00 8.39477360e-01
6.28483891e-01 8.78295183e-01 -8.15170169e-01 1.37292349e+00
-3.41744989e-01 5.81305146e-01 -2.23322406e-01 -2.33098030e-01
4.59939539e-01 -3.93017709e-01 4.52607751e-01 1.24059713e+00
4.33855057e-01 7.09804986e-03 -8.02875459e-02 8.10760677e-01
-1.59079969e-01 6.58618748e-01 -5.75568140e-01 1.40601709e-01
-2.75800023e-02 1.84985316e+00 -1.25114214e+00 -4.65199977e-01
-6.00845158e-01 5.09735823e-01 3.06516141e-01 1.56106263e-01
-8.07510376e-01 -6.24927938e-01 -3.46237123e-02 4.45413500e-01
2.53400981e-01 1.62754990e-02 -2.02853009e-01 -7.22434342e-01
-1.18229248e-01 -3.74816120e-01 7.52663076e-01 -6.64338112e-01
-1.47666728e+00 1.03459036e+00 -1.97344422e-01 -1.41294920e+00
1.95129707e-01 -9.90190327e-01 -9.22041118e-01 8.57376218e-01
-1.23765242e+00 -1.10046780e+00 -1.05105269e+00 6.02664292e-01
-1.18630878e-05 -1.41295433e-01 7.50652432e-01 8.36018696e-02
-4.20283020e-01 2.43008539e-01 -9.57612693e-02 7.34071016e-01
3.77926379e-01 -1.35029244e+00 -3.64669748e-02 5.00396907e-01
1.30985789e-02 4.39022809e-01 -2.93768644e-01 -5.32627165e-01
-7.52152920e-01 -1.50213599e+00 8.20579410e-01 -6.51000664e-02
4.86298233e-01 1.85485795e-01 -8.42200935e-01 5.73216975e-01
1.07273415e-01 7.35816717e-01 7.84832478e-01 -3.62125993e-01
-2.47091621e-01 1.91572737e-02 -1.29753804e+00 3.03900748e-01
8.40301335e-01 -2.32096434e-01 -2.86235601e-01 5.57660282e-01
2.92717338e-01 -5.48215985e-01 -1.43543041e+00 4.56956357e-01
4.86662388e-01 -1.16751444e+00 8.85007858e-01 -4.09519881e-01
4.16388154e-01 -3.28118324e-01 2.68841028e-01 -1.17266238e+00
-5.22016168e-01 -1.33375525e-01 2.18823254e-01 5.07914484e-01
5.44741094e-01 -8.38771820e-01 8.59007895e-01 1.19323999e-01
-3.71678263e-01 -1.20418704e+00 -1.08082712e+00 -7.79573143e-01
3.47729355e-01 -8.47400427e-02 4.04938549e-01 7.68785894e-01
-4.31697965e-01 2.37468377e-01 2.69491464e-01 -4.15298529e-02
4.48917896e-01 -1.10030472e-01 6.99613929e-01 -1.48939157e+00
-9.37653631e-02 -8.40246618e-01 -9.38094914e-01 -5.29819846e-01
-1.64382398e-01 -1.42064571e+00 -3.68254721e-01 -2.07513881e+00
-5.82515597e-02 -4.65402275e-01 -2.92389154e-01 3.97024900e-01
2.30494663e-01 2.41068125e-01 6.02583066e-02 3.06807250e-01
-6.76623955e-02 1.12625778e-01 1.89163184e+00 -3.96187641e-02
-1.26079068e-01 3.08011830e-01 -3.15696269e-01 6.37915611e-01
9.91864085e-01 -3.75437409e-01 -1.15410358e-01 1.50600925e-01
-3.80811483e-01 3.62117797e-01 5.78177452e-01 -1.39252245e+00
3.77463907e-01 3.91863495e-01 5.86812913e-01 -4.83425349e-01
-3.84649113e-02 -1.08763289e+00 2.95882970e-01 9.16770816e-01
-1.61849186e-01 -9.43521708e-02 -6.34717345e-02 4.38002646e-01
-2.50575632e-01 -2.90342063e-01 1.06733227e+00 -3.64958465e-01
-1.24747567e-01 7.28253245e-01 -4.43690330e-01 -1.10820882e-01
1.31527483e+00 -3.39487016e-01 -1.39244705e-01 -2.18398660e-01
-1.20929933e+00 -1.72092795e-01 1.64156035e-01 -1.50503010e-01
7.86870182e-01 -1.29187322e+00 -7.30401397e-01 2.88998395e-01
1.36215463e-01 4.56211060e-01 1.33336157e-01 1.38511896e+00
-1.37627077e+00 3.61919343e-01 -2.67363280e-01 -9.77957726e-01
-1.38206530e+00 4.79393333e-01 1.02579129e+00 -4.57939833e-01
-9.91116107e-01 7.30428398e-01 1.13192543e-01 -5.32102883e-01
1.59542397e-01 -7.20696151e-01 -4.76373702e-01 -1.43006608e-01
-1.32096425e-01 5.36665082e-01 2.98227370e-01 -8.04280281e-01
-2.98296809e-01 1.04988956e+00 4.15110201e-01 2.28885397e-01
1.12533700e+00 4.76889968e-01 -3.56600076e-01 -5.83111010e-02
1.56328952e+00 -8.49144533e-02 -6.61724567e-01 -2.43623719e-01
-2.43154362e-01 -3.28038573e-01 -2.21047457e-02 -5.11256218e-01
-1.72892404e+00 1.20726240e+00 6.82815552e-01 4.59026009e-01
1.02564216e+00 6.92520067e-02 7.72116840e-01 1.86246544e-01
-1.45349637e-01 -5.97135663e-01 3.35707575e-01 3.25121552e-01
8.64534080e-01 -9.28565204e-01 5.34484275e-02 -7.86607563e-01
-5.15357137e-01 1.84502578e+00 6.47135913e-01 -2.83143908e-01
1.11201453e+00 -1.18602335e-01 2.71103501e-01 -6.88626349e-01
-1.39270917e-01 -2.97780633e-01 5.13110280e-01 7.36568093e-01
5.52316010e-01 3.07915956e-01 -1.44707382e-01 2.80443639e-01
-1.74332142e-01 -1.16454199e-01 4.60730821e-01 8.52796912e-01
-3.03618878e-01 -7.34441042e-01 -2.46489346e-01 8.11732829e-01
-3.21186692e-01 1.74239621e-01 -5.29635847e-01 1.21033561e+00
2.82355636e-01 7.33677626e-01 2.46139303e-01 -3.11863929e-01
6.85548604e-01 -2.87163675e-01 3.30550611e-01 -6.14815772e-01
-9.65277672e-01 1.72538236e-01 9.97858196e-02 -4.95612383e-01
-2.71599382e-01 -3.53050470e-01 -1.48092186e+00 -1.41241491e-01
-4.36938077e-01 1.49076760e-01 5.12485325e-01 4.65251774e-01
1.24964407e-02 1.02420545e+00 8.12482655e-01 -3.90064389e-01
-3.16131264e-01 -8.62599730e-01 -7.70439625e-01 4.37910348e-01
2.05493957e-01 -5.18390357e-01 -7.24967599e-01 -1.09609887e-01] | [15.10130500793457, -2.2451276779174805] |
6dd0ac3b-0229-4aa6-89f5-8f6f9d32e7bc | a-multi-scale-multiple-instance-video | 1505.05914 | null | http://arxiv.org/abs/1505.05914v3 | http://arxiv.org/pdf/1505.05914v3.pdf | A Multi-scale Multiple Instance Video Description Network | Generating natural language descriptions for in-the-wild videos is a
challenging task. Most state-of-the-art methods for solving this problem borrow
existing deep convolutional neural network (CNN) architectures (AlexNet,
GoogLeNet) to extract a visual representation of the input video. However,
these deep CNN architectures are designed for single-label centered-positioned
object classification. While they generate strong semantic features, they have
no inherent structure allowing them to detect multiple objects of different
sizes and locations in the frame. Our paper tries to solve this problem by
integrating the base CNN into several fully convolutional neural networks
(FCNs) to form a multi-scale network that handles multiple receptive field
sizes in the original image. FCNs, previously applied to image segmentation,
can generate class heat-maps efficiently compared to sliding window mechanisms,
and can easily handle multiple scales. To further handle the ambiguity over
multiple objects and locations, we incorporate the Multiple Instance Learning
mechanism (MIL) to consider objects in different positions and at different
scales simultaneously. We integrate our multi-scale multi-instance architecture
with a sequence-to-sequence recurrent neural network to generate sentence
descriptions based on the visual representation. Ours is the first end-to-end
trainable architecture that is capable of multi-scale region processing.
Evaluation on a Youtube video dataset shows the advantage of our approach
compared to the original single-scale whole frame CNN model. Our flexible and
efficient architecture can potentially be extended to support other video
processing tasks. | ['Kate Saenko', 'Vasili Ramanishka', 'Subhashini Venugopalan', 'Marcus Rohrbach', 'Huijuan Xu'] | 2015-05-21 | null | null | null | null | ['video-description'] | ['computer-vision'] | [ 3.95594597e-01 -1.00097749e-02 -2.10211016e-02 -3.95060956e-01
-8.61155391e-01 -6.20874107e-01 3.24407369e-01 -2.43176728e-01
-4.35932487e-01 4.65183377e-01 -1.46267593e-01 -2.04386637e-01
3.86090696e-01 -8.22218120e-01 -1.11906338e+00 -4.99472290e-01
2.54868418e-01 3.48398358e-01 9.14715886e-01 -3.58364344e-01
6.26190975e-02 6.71900690e-01 -1.61712241e+00 9.89430487e-01
3.60845923e-01 1.18787074e+00 5.88302433e-01 8.24098587e-01
-2.96994358e-01 9.17771518e-01 -8.79987061e-01 -3.04289628e-02
3.91761214e-01 -4.17036682e-01 -6.74308836e-01 3.44707042e-01
8.43238950e-01 -4.31199342e-01 -2.51261026e-01 8.30211997e-01
3.06173623e-01 1.96586862e-01 4.58162457e-01 -1.22766793e+00
-5.65878749e-01 2.96222001e-01 -4.74641323e-01 3.00810426e-01
2.95564085e-01 2.23785043e-01 7.01211631e-01 -8.55266929e-01
7.69561231e-01 1.21732843e+00 6.49647176e-01 7.79110253e-01
-1.04365838e+00 -5.21860540e-01 3.33586395e-01 1.36556491e-01
-1.15774643e+00 -2.45264813e-01 5.96057057e-01 -3.46428245e-01
1.25843990e+00 1.83410063e-01 7.22520530e-01 1.17459953e+00
6.16640039e-02 1.01162148e+00 7.98816144e-01 -2.52865463e-01
1.80193573e-01 -1.17600560e-01 -3.41982782e-01 7.30791330e-01
2.36977413e-02 -2.68757403e-01 -2.36561820e-01 1.73381120e-01
1.19636643e+00 4.80735719e-01 -1.78332061e-01 -3.83529603e-01
-1.23054349e+00 8.71586323e-01 7.61366129e-01 3.54832947e-01
-2.42698625e-01 5.06020069e-01 5.59756875e-01 -5.33479899e-02
4.74557251e-01 3.39403331e-01 -5.88037789e-01 1.21708751e-01
-1.40141535e+00 4.23155755e-01 5.39029837e-01 1.10716033e+00
8.29519868e-01 2.63381690e-01 -4.05137539e-01 5.72107732e-01
-1.26824598e-03 2.67370284e-01 6.58208430e-01 -1.06570494e+00
5.26392579e-01 7.36586392e-01 -1.68570906e-01 -7.82797217e-01
-5.50997496e-01 -2.11679146e-01 -7.43512392e-01 3.71334404e-01
4.05509859e-01 -1.19655788e-01 -1.28335094e+00 1.42004645e+00
2.90765047e-01 4.37585622e-01 1.99537557e-02 1.10664546e+00
1.16009820e+00 1.00056326e+00 1.84739009e-01 2.18278363e-01
1.59603858e+00 -1.37396491e+00 -3.00835043e-01 -4.66035485e-01
3.65070373e-01 -6.06695712e-01 8.42024326e-01 1.17459647e-01
-1.29087448e+00 -8.19653988e-01 -8.77350032e-01 -4.50925469e-01
-5.67624271e-01 3.63862634e-01 4.47560161e-01 2.50905275e-01
-1.26155221e+00 4.02514309e-01 -7.04308212e-01 -3.98006588e-01
5.30677974e-01 2.39479244e-01 -2.86498934e-01 1.63151294e-01
-1.00619471e+00 5.19605398e-01 5.95878780e-01 1.46731332e-01
-9.50354874e-01 -5.72211564e-01 -1.09248316e+00 2.98844248e-01
5.65002978e-01 -8.31002891e-01 1.19251943e+00 -1.74546683e+00
-1.30363667e+00 7.42169857e-01 -1.43882141e-01 -5.85449219e-01
4.59651262e-01 6.62189350e-02 3.35017033e-02 6.34243011e-01
2.34140515e-01 1.27880609e+00 1.08394516e+00 -1.07193005e+00
-6.58140659e-01 -3.17241997e-02 3.72210771e-01 1.89118702e-02
-4.42270301e-02 2.50555158e-01 -6.65653348e-01 -1.10118663e+00
-1.91029489e-01 -9.32719409e-01 -4.39100444e-01 2.77750164e-01
-3.42667639e-01 -1.74507156e-01 9.57039833e-01 -4.72796202e-01
8.71343613e-01 -1.98493659e+00 1.59309894e-01 -2.04224691e-01
2.05918662e-02 6.23436987e-01 -4.62960660e-01 1.31891251e-01
-1.32714450e-01 9.58509147e-02 -3.24519664e-01 -5.07398367e-01
-3.94596875e-01 2.10456282e-01 -2.79337168e-01 9.29489508e-02
6.58170581e-01 1.18570924e+00 -7.83340573e-01 -5.12555301e-01
3.22590917e-01 6.15694880e-01 -7.45272517e-01 3.27069998e-01
-5.12936592e-01 1.33096293e-01 -2.45840415e-01 6.35920167e-01
5.76396942e-01 -3.89669180e-01 9.56332311e-02 -3.12962472e-01
-1.38485609e-02 -1.59354568e-01 -1.14290571e+00 1.81470847e+00
-5.48041880e-01 8.31567407e-01 -2.53332383e-03 -1.06026161e+00
8.45538557e-01 2.02604845e-01 3.72897506e-01 -5.22854745e-01
-1.06046554e-02 1.19654238e-01 -3.20522517e-01 -6.22095942e-01
6.61878228e-01 1.65029336e-02 -1.82704926e-01 2.51635581e-01
2.64526546e-01 2.38754209e-02 4.34561610e-01 1.93019688e-01
9.63060021e-01 5.96235573e-01 2.25529507e-01 3.73979397e-02
4.77389038e-01 5.48025034e-02 5.79035282e-01 6.64395750e-01
-6.73445761e-02 1.43621373e+00 4.91363376e-01 -9.27653670e-01
-1.29876232e+00 -8.51923764e-01 2.31744185e-01 1.32359660e+00
8.74765143e-02 -2.86221951e-01 -8.65058959e-01 -8.03437173e-01
-2.26586565e-01 2.35105425e-01 -5.98622859e-01 2.51947999e-01
-7.47209728e-01 -2.95159996e-01 4.54533219e-01 9.12937045e-01
5.25608242e-01 -1.56195092e+00 -1.26925516e+00 4.80578274e-01
-2.94356167e-01 -1.52358294e+00 -5.61587095e-01 1.03927635e-01
-5.77407241e-01 -1.10668004e+00 -1.04120529e+00 -9.61961985e-01
6.68861449e-01 1.89104214e-01 1.13331163e+00 1.02032840e-01
-6.81955338e-01 2.10618928e-01 -4.52920228e-01 -2.46434465e-01
-4.20081168e-01 1.13312691e-01 -3.31405997e-01 1.74195975e-01
7.69469440e-02 -1.70828104e-01 -8.26916397e-01 2.78037846e-01
-1.42759299e+00 2.58949995e-01 4.93478656e-01 6.77102089e-01
6.43605947e-01 -3.17953646e-01 5.94302177e-01 -7.52119124e-01
2.42551297e-01 -3.19460928e-01 -5.69266796e-01 3.21535766e-01
9.30113345e-02 2.87412703e-02 7.60496080e-01 -5.39883494e-01
-8.99283588e-01 4.93265331e-01 -2.31292620e-01 -7.95656025e-01
-4.78190571e-01 7.29774758e-02 1.69551894e-01 5.63738830e-02
4.51904237e-01 3.00054610e-01 4.51254360e-02 -2.49643058e-01
4.69779819e-01 4.78238553e-01 4.45368528e-01 -2.71480948e-01
3.49955052e-01 5.91351390e-01 -1.21466130e-01 -6.38837457e-01
-9.58664954e-01 -5.97637951e-01 -7.97760904e-01 -1.80637762e-01
1.36155486e+00 -1.14763904e+00 -4.04829592e-01 3.45670670e-01
-1.53056180e+00 -5.99533379e-01 -2.58164763e-01 -5.28683253e-02
-7.41753280e-01 2.10816383e-01 -6.60628736e-01 -5.05739331e-01
-4.02484775e-01 -1.38801622e+00 1.61769497e+00 3.48549306e-01
-6.99830800e-02 -1.02118349e+00 -3.42502177e-01 3.16990525e-01
5.53514123e-01 4.72141743e-01 4.15104210e-01 -5.42868614e-01
-7.50801265e-01 1.10662235e-02 -4.30776209e-01 3.45080525e-01
-1.71965808e-01 2.24041015e-01 -7.88217545e-01 -2.47310594e-01
-3.14118564e-01 -4.46582973e-01 1.17175233e+00 4.49591577e-01
1.51284707e+00 -3.86562645e-01 -1.62587568e-01 7.74770617e-01
1.53520906e+00 -1.19277267e-02 8.06218624e-01 5.62519968e-01
8.51817846e-01 4.34519261e-01 4.74404961e-01 2.96071857e-01
3.53200555e-01 9.10733223e-01 6.28666937e-01 -5.97628474e-01
-3.56546521e-01 -5.07639982e-02 4.92806047e-01 9.12267640e-02
-3.45988125e-02 -1.85394794e-01 -6.49815619e-01 5.64674258e-01
-2.04636359e+00 -1.26962304e+00 9.18557495e-02 1.63608086e+00
5.30206978e-01 -4.93753664e-02 3.64831299e-01 -3.20507854e-01
7.61018574e-01 2.21893698e-01 -4.62362558e-01 -6.89777672e-01
-1.76062763e-01 2.86506653e-01 5.16472578e-01 7.50210881e-02
-1.22689533e+00 1.04741561e+00 5.99358416e+00 9.21749830e-01
-1.46739888e+00 1.28475696e-01 9.02021587e-01 -3.06593001e-01
3.36419186e-03 -2.18029052e-01 -9.31844413e-01 3.50336969e-01
7.93319345e-01 4.09700751e-01 1.82007879e-01 1.04471707e+00
1.98820844e-01 5.29296137e-02 -9.57355499e-01 9.35401022e-01
3.27404290e-01 -1.74843597e+00 3.28964144e-01 -2.01122880e-01
8.49180937e-01 1.65092498e-01 -1.41856328e-01 3.26301694e-01
-1.18453316e-01 -1.03371584e+00 1.09128368e+00 3.72387558e-01
9.61406827e-01 -6.38445616e-01 6.63361013e-01 2.62130797e-01
-1.42112231e+00 -3.00183177e-01 -5.14831066e-01 6.60398602e-02
2.75234729e-01 1.77008152e-01 -8.67290735e-01 2.91812509e-01
9.02557552e-01 6.01821363e-01 -8.08677971e-01 9.90415931e-01
-3.06871980e-02 1.87163457e-01 -2.00175703e-01 -8.18293169e-02
6.48758769e-01 1.34398669e-01 2.95253068e-01 1.56456423e+00
4.81066406e-01 -1.54446587e-01 3.47175241e-01 1.10414183e+00
-6.54658377e-02 -1.20411381e-01 -5.56841671e-01 2.36031622e-01
1.38199672e-01 1.45932996e+00 -1.16934741e+00 -6.70804203e-01
-5.13837695e-01 1.22218156e+00 3.93564641e-01 5.33690333e-01
-1.09421897e+00 -4.81326967e-01 5.83764613e-01 4.04431283e-01
8.91788960e-01 -1.90043509e-01 6.30720779e-02 -1.25945628e+00
3.24215554e-02 -8.40703070e-01 3.86575341e-01 -1.00693488e+00
-1.06182015e+00 9.67177808e-01 2.86883023e-02 -1.38566899e+00
-5.22842586e-01 -8.22297633e-01 -6.75677538e-01 5.01206696e-01
-1.66307759e+00 -1.40608978e+00 -3.76046687e-01 7.73251653e-01
1.23553288e+00 -7.58464783e-02 6.14503980e-01 8.22249278e-02
-3.74467194e-01 3.99992973e-01 -2.16127694e-01 4.23430115e-01
4.45295095e-01 -1.17235565e+00 5.54883003e-01 8.42463970e-01
2.38425076e-01 4.02110100e-01 2.95030743e-01 -3.64088386e-01
-1.03434050e+00 -1.59081650e+00 4.77485925e-01 -3.70313585e-01
4.27788347e-01 -6.39455795e-01 -7.80050933e-01 6.52378798e-01
2.88683623e-01 5.27085781e-01 3.70114446e-01 -5.37599266e-01
-3.44032466e-01 -5.31795993e-02 -1.09916556e+00 4.47850287e-01
8.98479104e-01 -3.81985128e-01 -4.51989204e-01 3.33785027e-01
7.97473669e-01 -6.60822511e-01 -4.83038664e-01 3.68369341e-01
3.11223626e-01 -1.16919339e+00 1.14328861e+00 -6.03581190e-01
8.15795004e-01 -4.92708206e-01 1.52190477e-01 -1.00876427e+00
-1.42453089e-01 -4.73109037e-01 3.60813774e-02 9.99417305e-01
4.13912565e-01 -3.93244684e-01 7.73432076e-01 5.08730650e-01
-2.18807757e-01 -9.17379379e-01 -1.00275147e+00 -5.13871491e-01
-1.55265361e-01 -3.69965971e-01 4.13474411e-01 4.31699663e-01
-4.76189882e-01 1.75356433e-01 -3.72349173e-01 -1.17607145e-02
3.31641734e-01 3.76148492e-01 7.55373657e-01 -8.32010269e-01
-3.42741787e-01 -4.11273271e-01 -6.34226859e-01 -1.09831476e+00
2.99214095e-01 -7.55269587e-01 1.16836570e-01 -1.72165835e+00
1.77949995e-01 -2.66203046e-01 -1.60297100e-02 7.71075010e-01
-1.00029580e-01 9.03182089e-01 6.54958248e-01 -6.71462342e-03
-9.35932159e-01 2.38706395e-01 1.21995699e+00 -1.48376092e-01
-1.69730455e-01 -1.65364981e-01 -3.85529488e-01 6.89807475e-01
6.37634635e-01 -3.06360245e-01 -2.36415327e-01 -4.58929986e-01
4.98955175e-02 -3.49877737e-02 6.94217205e-01 -1.12245023e+00
1.03294417e-01 -7.17902184e-02 6.76436841e-01 -4.26469624e-01
3.98093253e-01 -8.44875991e-01 9.33531299e-02 2.50371337e-01
-3.52634400e-01 1.90330550e-01 3.80144715e-01 4.31533724e-01
-3.41388553e-01 -2.36453235e-01 8.37036133e-01 -5.71381688e-01
-1.06863511e+00 4.10517037e-01 -3.97021323e-01 -6.48509711e-02
1.34299171e+00 -3.86605978e-01 -2.97905684e-01 -2.13091403e-01
-7.51534760e-01 6.53292462e-02 5.50780892e-01 7.54782140e-01
9.68562365e-01 -1.19946337e+00 -6.43439710e-01 2.85970837e-01
7.06819147e-02 3.90650034e-01 3.33589822e-01 4.75785494e-01
-9.56444740e-01 3.39382291e-01 -3.50601643e-01 -8.59791160e-01
-1.21314752e+00 6.75100327e-01 4.87122089e-01 -2.04770640e-01
-7.66779840e-01 7.53018796e-01 5.02382278e-01 -2.11055771e-01
2.17368943e-03 -6.09134853e-01 -1.71845153e-01 1.09774508e-01
7.38723159e-01 -4.61242199e-02 5.94050661e-02 -7.78123140e-01
-2.24707201e-01 7.75112152e-01 3.42085212e-02 1.67681694e-01
1.47066665e+00 -3.69542316e-02 3.45032588e-02 2.33418658e-01
1.41750467e+00 -3.48611414e-01 -1.59875524e+00 4.62981872e-02
-3.00761074e-01 -1.93884030e-01 -3.01262528e-01 -4.92743880e-01
-1.33334672e+00 9.14643466e-01 4.23638135e-01 1.91503480e-01
1.08564353e+00 2.70163938e-02 8.77614677e-01 1.09009176e-01
1.73437834e-01 -1.06444001e+00 4.37083960e-01 5.13247192e-01
8.91314209e-01 -1.28215361e+00 -1.96095556e-01 -2.32871652e-01
-8.72797430e-01 1.70291340e+00 7.94098914e-01 -3.60311985e-01
1.95515946e-01 4.01266396e-01 2.09558889e-01 -1.21789999e-01
-6.74206257e-01 -3.70810270e-01 3.70635271e-01 5.13778806e-01
3.55567783e-01 -1.93064541e-01 8.73070136e-02 2.68082052e-01
2.90431648e-01 6.74142838e-02 7.36206353e-01 8.69925678e-01
-3.72099757e-01 -8.85919750e-01 -3.39580208e-01 2.57618755e-01
-5.40323019e-01 -9.47515592e-02 -6.74439669e-02 7.30973244e-01
2.62353003e-01 6.16317868e-01 4.66290265e-01 1.55714806e-02
2.54153460e-01 -3.55070718e-02 2.73399353e-01 -8.41154099e-01
-7.75595129e-01 1.61800966e-01 -2.28160188e-01 -9.31029856e-01
-7.46836662e-01 -4.35552597e-01 -1.55105186e+00 3.50902796e-01
1.09510915e-02 -3.02062720e-01 7.67126083e-01 8.54732454e-01
5.15416205e-01 7.39474118e-01 3.39566827e-01 -1.48394418e+00
-2.07704857e-01 -7.47988880e-01 -2.58759737e-01 6.66567087e-01
5.87245464e-01 -3.39712203e-01 -1.47286147e-01 2.97779202e-01] | [9.281471252441406, 0.044471219182014465] |
40fa3400-1002-414d-a16a-613d0e74119c | visil-fine-grained-spatio-temporal-video | 1908.07410 | null | https://arxiv.org/abs/1908.07410v1 | https://arxiv.org/pdf/1908.07410v1.pdf | ViSiL: Fine-grained Spatio-Temporal Video Similarity Learning | In this paper we introduce ViSiL, a Video Similarity Learning architecture that considers fine-grained Spatio-Temporal relations between pairs of videos -- such relations are typically lost in previous video retrieval approaches that embed the whole frame or even the whole video into a vector descriptor before the similarity estimation. By contrast, our Convolutional Neural Network (CNN)-based approach is trained to calculate video-to-video similarity from refined frame-to-frame similarity matrices, so as to consider both intra- and inter-frame relations. In the proposed method, pairwise frame similarity is estimated by applying Tensor Dot (TD) followed by Chamfer Similarity (CS) on regional CNN frame features - this avoids feature aggregation before the similarity calculation between frames. Subsequently, the similarity matrix between all video frames is fed to a four-layer CNN, and then summarized using Chamfer Similarity (CS) into a video-to-video similarity score -- this avoids feature aggregation before the similarity calculation between videos and captures the temporal similarity patterns between matching frame sequences. We train the proposed network using a triplet loss scheme and evaluate it on five public benchmark datasets on four different video retrieval problems where we demonstrate large improvements in comparison to the state of the art. The implementation of ViSiL is publicly available. | ['Ioannis Kompatsiaris', 'Giorgos Kordopatis-Zilos', 'Symeon Papadopoulos', 'Ioannis Patras'] | 2019-08-20 | visil-fine-grained-spatio-temporal-video-1 | http://openaccess.thecvf.com/content_ICCV_2019/html/Kordopatis-Zilos_ViSiL_Fine-Grained_Spatio-Temporal_Video_Similarity_Learning_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Kordopatis-Zilos_ViSiL_Fine-Grained_Spatio-Temporal_Video_Similarity_Learning_ICCV_2019_paper.pdf | iccv-2019-10 | ['video-similarity'] | ['computer-vision'] | [ 1.13846980e-01 -6.99571550e-01 -1.87817931e-01 -4.43414301e-01
-8.50288153e-01 -4.98461723e-01 8.15713644e-01 4.80824679e-01
-6.91030085e-01 3.26380223e-01 1.89975038e-01 2.22885758e-01
-4.05757517e-01 -6.34660840e-01 -7.98235655e-01 -5.62813699e-01
-3.26088011e-01 9.47519690e-02 4.24861878e-01 -2.03223944e-01
4.26397830e-01 6.29185259e-01 -1.70880210e+00 5.69661558e-01
1.57206044e-01 1.66115081e+00 5.92167415e-02 7.29043424e-01
1.74293399e-01 1.27163959e+00 -2.32489794e-01 -2.48617560e-01
5.21139681e-01 -3.64985734e-01 -8.87355089e-01 1.72855556e-02
1.07437479e+00 -4.89927769e-01 -7.31129706e-01 9.45124626e-01
3.62168849e-01 5.55390835e-01 4.36826974e-01 -1.31461799e+00
-6.65818274e-01 1.34317175e-01 -3.68647546e-01 6.16504371e-01
8.72141480e-01 -5.45982607e-02 1.37145925e+00 -1.02226448e+00
8.85223866e-01 1.23957098e+00 7.49361455e-01 -9.41834301e-02
-8.14713180e-01 -3.38439375e-01 -2.03240678e-01 9.43335712e-01
-1.57187080e+00 -3.19791973e-01 7.24294841e-01 -4.64910090e-01
1.21584427e+00 3.28722060e-01 9.94978011e-01 5.45268893e-01
3.41643840e-01 8.04882824e-01 3.35372895e-01 -7.75104202e-03
-5.07788584e-02 -6.37732565e-01 -4.32871953e-02 7.08578050e-01
-2.84940869e-01 1.27828240e-01 -5.92042565e-01 1.95714105e-02
6.27173901e-01 4.94783044e-01 -2.86385119e-01 -5.11553228e-01
-1.66111267e+00 6.60444140e-01 8.25908482e-01 6.19291127e-01
-4.56899256e-01 2.52304018e-01 8.76340628e-01 6.66889727e-01
4.19905424e-01 2.62144864e-01 4.46824916e-02 -1.79795727e-01
-1.48024881e+00 5.33172131e-01 5.26107132e-01 8.20105493e-01
1.00392354e+00 -3.03485811e-01 -4.92360443e-01 6.78424776e-01
1.51954576e-01 2.51115590e-01 4.95179772e-01 -1.21129942e+00
4.07617837e-01 4.29765850e-01 -1.59917608e-01 -1.66538358e+00
-2.03505963e-01 6.59245700e-02 -8.84670734e-01 -2.45378241e-01
3.06221098e-01 4.58054721e-01 -4.74745095e-01 1.38068187e+00
1.41977042e-01 5.90760946e-01 -2.64771670e-01 1.22729588e+00
1.10391319e+00 6.27594829e-01 -1.29363149e-01 -2.15724707e-01
1.08023226e+00 -1.04161739e+00 -5.69938481e-01 3.94235194e-01
5.60108483e-01 -1.04960775e+00 5.20718396e-01 -7.94472620e-02
-1.49949610e+00 -9.36743021e-01 -1.04757881e+00 -3.73723984e-01
-3.99244785e-01 -4.48220335e-02 1.05843499e-01 -3.61265279e-02
-1.52585077e+00 8.76911402e-01 -7.41971910e-01 -5.49713910e-01
2.75556505e-01 4.16961879e-01 -7.78524160e-01 -2.18948379e-01
-1.23452699e+00 7.87294328e-01 3.98579568e-01 9.60461274e-02
-7.07008779e-01 -7.58153975e-01 -1.16504908e+00 1.36191845e-01
1.59480914e-01 -7.47141242e-01 9.07123029e-01 -1.15237737e+00
-1.16214681e+00 7.70148993e-01 -9.30731222e-02 -4.85662252e-01
3.53045911e-01 -8.37701708e-02 -2.93752193e-01 6.03902161e-01
1.49879813e-01 8.40412736e-01 8.55235517e-01 -7.29123175e-01
-8.94980609e-01 5.85736101e-03 4.73675370e-01 1.69009924e-01
-2.51641329e-02 3.50770354e-01 -7.20366716e-01 -8.64121974e-01
2.73331344e-01 -9.90252614e-01 1.37776107e-01 4.68567550e-01
-5.88842146e-02 -5.03065705e-01 9.23626125e-01 -4.82780308e-01
1.22549272e+00 -2.05809784e+00 5.32289386e-01 2.19162360e-01
2.52755672e-01 3.79567206e-01 -5.24258494e-01 4.64501351e-01
-3.34058106e-01 -3.26337039e-01 -4.86830398e-02 -3.97101641e-01
-8.75834152e-02 1.11413077e-02 4.77844439e-02 7.62545705e-01
3.15091997e-01 9.13764358e-01 -1.15348732e+00 -7.13841558e-01
5.81383467e-01 6.77985847e-01 -6.89063549e-01 3.87301087e-01
1.69518232e-01 8.51671547e-02 -2.87610590e-01 6.02622509e-01
6.17980599e-01 -4.11661416e-02 -9.46490541e-02 -8.03794384e-01
-7.22246990e-02 1.93268191e-02 -8.94443810e-01 2.07613015e+00
-3.84332150e-01 1.01979077e+00 -4.23624605e-01 -1.27246547e+00
6.97258234e-01 2.53233224e-01 1.09666264e+00 -6.43217862e-01
2.15347230e-01 1.29516199e-01 -2.78984100e-01 -5.52796721e-01
5.44408321e-01 3.55683833e-01 1.61982104e-01 3.62551451e-01
2.93664485e-01 -1.18407849e-02 5.92628896e-01 3.41822833e-01
1.00729036e+00 1.91810697e-01 3.59988689e-01 -1.18123315e-01
1.09881747e+00 -2.49432653e-01 2.29629770e-01 5.33548772e-01
-2.60564536e-01 8.72025669e-01 2.87692904e-01 -9.70272183e-01
-1.18147039e+00 -1.12141275e+00 7.14571774e-02 9.50334787e-01
3.94480735e-01 -7.18986869e-01 -6.15302682e-01 -5.43104947e-01
3.07594594e-02 -2.39026040e-01 -6.73370779e-01 -1.74705416e-01
-6.68979645e-01 1.64643042e-02 3.18259656e-01 4.53111887e-01
7.30414569e-01 -9.58366752e-01 -4.20963049e-01 2.25093737e-01
-3.60349983e-01 -1.15551484e+00 -1.02481437e+00 -4.36168581e-01
-4.66101676e-01 -1.28075671e+00 -8.48006785e-01 -1.03227830e+00
2.88899571e-01 4.92353052e-01 1.16140461e+00 4.20746565e-01
-3.14954609e-01 5.54585814e-01 -6.17290676e-01 4.71466005e-01
-3.72462943e-02 -2.44742289e-01 8.88275076e-03 2.47131079e-01
4.87002343e-01 -3.91953409e-01 -8.05445969e-01 3.81488830e-01
-1.02882874e+00 -3.30342084e-01 1.86594471e-01 9.00256693e-01
7.81907618e-01 -4.47466969e-01 -7.87137225e-02 -1.95916574e-02
2.01118186e-01 -4.24359262e-01 -5.05371928e-01 5.01501024e-01
9.08676386e-02 -2.03204229e-01 4.09839451e-01 -3.00829828e-01
-3.19564730e-01 -9.75725800e-02 -2.17098314e-02 -1.01801252e+00
8.15235451e-02 5.13776422e-01 2.01790452e-01 -5.30063212e-01
2.18208164e-01 2.35655367e-01 -1.27428189e-01 -1.27203152e-01
1.55115739e-01 4.17415231e-01 6.32437885e-01 -4.03303862e-01
7.68560112e-01 4.86395389e-01 2.00637400e-01 -7.44179547e-01
-7.38092661e-01 -8.58783364e-01 -9.73126590e-01 -3.03086936e-01
1.13625610e+00 -9.96901095e-01 -8.00880313e-01 3.58257174e-01
-1.34715724e+00 1.81822047e-01 -1.78661719e-01 8.04329097e-01
-7.59752691e-01 8.50877464e-01 -7.88481712e-01 -4.98166271e-02
-3.47504795e-01 -1.46575272e+00 1.29021668e+00 -4.80300672e-02
-1.89784750e-01 -1.09510589e+00 2.01158419e-01 2.52983183e-01
4.67696697e-01 2.07349986e-01 5.19824088e-01 -6.59761429e-01
-7.43416488e-01 -3.63298386e-01 -5.37127435e-01 3.83254796e-01
2.09444731e-01 2.95675546e-01 -4.12785888e-01 -5.40701151e-01
-1.82388276e-01 -1.98933020e-01 8.83611917e-01 4.42461729e-01
1.17005181e+00 -1.41877860e-01 -5.73374741e-02 9.02644336e-01
1.38029957e+00 5.68631850e-02 5.83883643e-01 5.82267165e-01
8.86838555e-01 2.93247670e-01 8.33165169e-01 5.86891055e-01
5.53235412e-01 1.11449695e+00 2.39625484e-01 8.50815251e-02
-2.58466989e-01 1.46200359e-01 4.14489150e-01 1.01959574e+00
-3.56362134e-01 -1.93579942e-01 -5.30601919e-01 5.05221367e-01
-1.94687974e+00 -1.45077825e+00 8.28891546e-02 2.12493896e+00
4.82928246e-01 -3.10937732e-01 1.63711771e-01 8.28453079e-02
7.49365568e-01 5.02954960e-01 -8.52597579e-02 -2.07913563e-01
-2.23105878e-01 1.40334755e-01 2.53082603e-01 4.37094003e-01
-1.46819353e+00 7.15929925e-01 5.36184883e+00 8.74000072e-01
-1.28129041e+00 4.47753519e-02 3.40402305e-01 -3.36414218e-01
8.58423263e-02 -1.26842797e-01 -2.12734088e-01 4.12987381e-01
5.43436825e-01 -1.57655597e-01 3.70146126e-01 5.06015480e-01
1.35791143e-02 -7.81082585e-02 -1.59748340e+00 1.41797435e+00
4.53415036e-01 -1.69340849e+00 3.63504440e-01 -4.59130555e-01
6.73954725e-01 1.29163310e-01 -2.79421546e-02 2.00203550e-03
-2.98757046e-01 -7.16364205e-01 9.52051878e-01 7.42975712e-01
7.17834234e-01 -7.54745662e-01 9.44876850e-01 -4.75017339e-01
-1.82967556e+00 6.96916655e-02 -5.70294738e-01 2.28240877e-01
1.76301867e-01 2.80514419e-01 -3.48947704e-01 8.58226120e-01
1.13536811e+00 1.53795934e+00 -6.40573859e-01 1.27936959e+00
3.85864049e-01 -9.80209261e-02 -1.60331309e-01 3.27379644e-01
6.47392511e-01 -1.56485915e-01 5.35830855e-01 1.36424053e+00
5.70349634e-01 -7.01799840e-02 3.17479312e-01 4.31004941e-01
3.15163843e-02 1.34397328e-01 -5.32838821e-01 8.30001831e-02
2.18208373e-01 1.18644381e+00 -4.59054828e-01 -5.12928486e-01
-6.25880718e-01 1.02010548e+00 1.12225175e-01 2.57723272e-01
-8.07732761e-01 -5.20314574e-01 9.72468257e-01 -9.78403538e-02
4.89506692e-01 -1.51364580e-01 5.67720056e-01 -1.26689899e+00
2.36056745e-01 -8.30412507e-01 5.99517524e-01 -8.73993158e-01
-1.44389963e+00 8.07312012e-01 2.01989673e-02 -1.69589841e+00
-5.28052926e-01 -3.51952910e-01 -5.43000758e-01 6.18564010e-01
-1.57387948e+00 -1.19316411e+00 -4.98942316e-01 1.07887113e+00
6.31341219e-01 -3.73356819e-01 4.30813044e-01 6.28143668e-01
-1.64131671e-01 7.35312939e-01 1.24119624e-01 3.25818330e-01
8.18324208e-01 -8.85270536e-01 4.73704129e-01 5.66788495e-01
1.81165382e-01 3.98337871e-01 3.01739991e-01 -3.58759046e-01
-1.40990496e+00 -1.19759166e+00 1.15682185e+00 -2.81007886e-01
9.60962415e-01 5.79541773e-02 -9.46949065e-01 4.67267334e-01
2.92391449e-01 7.08441615e-01 5.76949894e-01 -6.08871639e-01
-5.12521267e-01 -3.57666969e-01 -8.83666158e-01 3.52612048e-01
1.06364620e+00 -1.10238397e+00 -4.88245368e-01 3.62608194e-01
5.37497818e-01 -2.57593930e-01 -1.27124262e+00 3.46405119e-01
8.57243419e-01 -1.29165030e+00 1.31459105e+00 -6.02164567e-01
6.17040157e-01 -6.22299016e-01 -6.09130740e-01 -1.08292091e+00
-4.16898906e-01 -5.17872036e-01 2.27129921e-01 7.89800167e-01
-1.27513155e-01 -9.22325701e-02 3.67103785e-01 1.00070558e-01
-1.03643127e-01 -7.10953176e-01 -1.05756593e+00 -7.17307270e-01
-2.63414085e-01 -3.12451780e-01 5.79959154e-01 1.08088422e+00
5.13478555e-02 -2.71425515e-01 -3.64310801e-01 -3.28103304e-02
5.15982747e-01 1.27540499e-01 5.21406472e-01 -8.80581856e-01
-4.97327819e-02 -5.70032954e-01 -1.25438941e+00 -8.55470896e-01
4.52766478e-01 -1.00089467e+00 -2.55308282e-02 -1.30471945e+00
3.27877194e-01 -6.40450269e-02 -6.22720897e-01 2.54804730e-01
2.00119745e-02 7.39378691e-01 6.62829936e-01 2.97401309e-01
-1.20718849e+00 5.27603745e-01 1.24298728e+00 -4.17699695e-01
2.63443828e-01 -3.92622739e-01 2.56626576e-01 5.20932317e-01
1.81283355e-01 -2.69314021e-01 -1.78386256e-01 -5.79238713e-01
9.93271265e-03 3.09256703e-01 6.59256995e-01 -1.45517695e+00
5.33702910e-01 2.33082026e-02 3.90233189e-01 -6.72513366e-01
4.72616881e-01 -9.99419808e-01 2.52320826e-01 3.97116274e-01
-5.07328570e-01 5.84794104e-01 -2.22529426e-01 4.13203090e-01
-8.15828800e-01 1.02579109e-01 6.27310455e-01 5.24802022e-02
-8.45454574e-01 7.99800813e-01 -8.83371085e-02 -1.03399932e-01
1.06058228e+00 -3.35966378e-01 -9.51522812e-02 -4.34783638e-01
-6.99839294e-01 7.12105632e-03 5.96845090e-01 7.07206130e-01
9.09267783e-01 -1.85083270e+00 -6.76711261e-01 2.52547830e-01
1.96941882e-01 -3.62873822e-01 4.52228189e-01 1.17774081e+00
-8.10332537e-01 5.10476291e-01 -5.62731445e-01 -9.04519498e-01
-1.49394965e+00 6.42200589e-01 4.75221664e-01 -6.33936673e-02
-3.91515523e-01 8.85975420e-01 1.52367815e-01 5.95353059e-02
2.52851695e-01 -3.84048104e-01 -2.83508897e-01 3.07550251e-01
6.35792911e-01 3.39882165e-01 1.49713203e-01 -1.35216928e+00
-5.82279921e-01 1.17792106e+00 -6.14469759e-02 3.61142635e-01
1.22190082e+00 -1.42037749e-01 -4.45618689e-01 1.97109938e-01
2.09695649e+00 -6.66741431e-01 -1.11689472e+00 -6.13052189e-01
3.11579350e-02 -8.58659029e-01 9.47695747e-02 -3.19107808e-02
-1.48983002e+00 8.72694790e-01 7.45692611e-01 -1.59262307e-02
1.16912436e+00 4.70543653e-02 1.09083128e+00 6.31534278e-01
1.71637163e-01 -9.61415350e-01 3.36094171e-01 6.70692325e-01
8.82304609e-01 -1.32821655e+00 2.21249238e-01 -7.11151138e-02
-4.41749603e-01 1.36519396e+00 1.73481181e-01 -4.82575357e-01
8.29815388e-01 -2.37166226e-01 -1.57588825e-01 -1.61479175e-01
-7.08441377e-01 -3.01431626e-01 9.28331792e-01 3.16120744e-01
5.59537411e-01 -3.52621704e-01 -1.05977528e-01 -1.01908617e-01
9.27430987e-02 9.80219245e-02 9.40655023e-02 9.10524547e-01
-1.99037939e-01 -1.03732955e+00 -1.33144528e-01 3.79951745e-01
-3.67616713e-01 -1.67660818e-01 -1.76881835e-01 6.12485707e-01
2.43347198e-01 8.67952347e-01 6.57162428e-01 -6.62457347e-01
2.27244243e-01 -4.60559160e-01 5.22305131e-01 -2.60677516e-01
-9.33175266e-01 6.04362190e-02 -2.61745989e-01 -1.27186620e+00
-1.00830543e+00 -8.54668975e-01 -8.67626965e-01 -5.27607739e-01
2.33812444e-02 9.01357271e-03 3.36605549e-01 1.05186260e+00
1.69318795e-01 1.25150725e-01 8.74743283e-01 -1.49784994e+00
-1.88180357e-01 -6.17622733e-01 -3.67010504e-01 8.83956492e-01
6.82415366e-01 -7.43755519e-01 -2.67166853e-01 1.45650879e-01] | [9.91629409790039, 0.6412017345428467] |
74359535-c9aa-4a5d-bbaa-62b986d2397e | hierarchical-deep-learning-model-for | 2303.03386 | null | https://arxiv.org/abs/2303.03386v1 | https://arxiv.org/pdf/2303.03386v1.pdf | Hierarchical Deep Learning Model for Degradation Prediction per Look-Ahead Scheduled Battery Usage Profile | Batteries can effectively improve the security of energy systems and mitigate climate change by facilitating wind and solar power. The installed capacity of battery energy storage system (BESS), mainly the lithium ion batteries are increasing significantly in recent years. However, the battery degradation cannot be accurately quantified and integrated into energy management system with existing heuristic battery degradation models. This paper proposed a hierarchical deep learning based battery degradation quantification (HDL-BDQ) model to quantify the battery degradation given scheduled BESS daily operations. Particularly, two sequential and cohesive deep neural networks are proposed to accurately estimate the degree of degradation using inputs of battery operational profiles and it can significantly outperform existing fixed or linear rate based degradation models as well as single-stage deep neural models. Training results show the high accuracy of the proposed system. Moreover, a learning and optimization decoupled algorithm is implemented to strategically take advantage of the proposed HDL-BDQ model in optimization-based look-ahead scheduling (LAS) problems. Case studies demonstrate the effectiveness of the proposed HDL-BDQ model in LAS of a microgrid testbed. | ['Xingpeng Li', 'Cunzhi Zhao'] | 2023-03-06 | null | null | null | null | ['energy-management'] | ['time-series'] | [-8.37544203e-01 -6.33650541e-01 -2.11247981e-01 -2.43240073e-01
-2.18385935e-01 -3.94290149e-01 2.58598179e-01 1.11306958e-01
-1.06049636e-02 1.37866187e+00 1.49660986e-02 -5.43853998e-01
-4.22241330e-01 -9.27610576e-01 -6.58030450e-01 -1.17966104e+00
-2.09654436e-01 5.82976878e-01 -2.36653298e-01 -3.05800885e-01
-1.42048627e-01 6.42698050e-01 -1.49343538e+00 -4.07053024e-01
1.20545709e+00 1.57942712e+00 4.92924869e-01 5.94696343e-01
3.00459802e-01 4.43325520e-01 -6.78679109e-01 3.88330013e-01
1.55937016e-01 3.92552055e-02 1.05131352e-02 -4.21928942e-01
-6.41201556e-01 -9.13616776e-01 -3.77674848e-01 1.01092339e+00
1.01172543e+00 2.32183710e-01 4.68819857e-01 -1.71112800e+00
-5.47374249e-01 5.53425193e-01 -1.01755103e-02 6.33944571e-01
-5.12576699e-01 2.71843672e-01 5.27387440e-01 -7.32376754e-01
-3.50388616e-01 9.69824493e-01 4.40721333e-01 2.93587834e-01
-4.42019194e-01 -6.69600725e-01 1.71509936e-01 7.58820176e-01
-1.25021720e+00 -2.63129890e-01 7.62737453e-01 -2.10941672e-01
1.54734635e+00 2.79159695e-01 1.20398843e+00 3.29607397e-01
6.15524650e-01 8.55435073e-01 1.17087984e+00 -1.38815120e-01
5.41977346e-01 8.30517802e-03 1.86935350e-01 1.73385188e-01
6.55331910e-01 2.86315709e-01 -3.35081518e-01 2.56513059e-01
1.28671870e-01 9.38822143e-03 -2.27255359e-01 -2.31199518e-01
-5.18267989e-01 3.63079518e-01 5.44342756e-01 -3.99336517e-02
-4.22336996e-01 3.84700119e-01 5.65599084e-01 4.23455797e-02
5.18320560e-01 9.76779535e-02 -6.48902297e-01 -1.04918577e-01
-9.05878603e-01 1.21609949e-01 6.75154448e-01 9.78302002e-01
2.33687997e-01 1.10383546e+00 -3.20210010e-01 3.24811667e-01
3.15096140e-01 6.75816238e-01 8.23851526e-01 -7.14922130e-01
-5.56794144e-02 4.56486583e-01 7.13736653e-01 -2.39200160e-01
-7.97104359e-01 -9.52588320e-01 -1.15809906e+00 2.14407697e-01
-4.39923197e-01 -2.89887667e-01 -8.29155266e-01 1.41842508e+00
1.45832926e-01 -2.50469655e-01 2.40228608e-01 7.95252264e-01
3.46403837e-01 9.55162823e-01 -1.85728475e-01 -7.08406806e-01
9.09775257e-01 -9.71660078e-01 -1.31379724e+00 8.12994391e-02
3.19789857e-01 7.91816562e-02 7.11364090e-01 5.10505438e-01
-1.50390625e+00 -3.86750758e-01 -1.66597950e+00 1.84745789e-01
-8.65006864e-01 4.05892611e-01 3.45716149e-01 6.89913630e-01
-1.36122024e+00 5.80689967e-01 -1.14833391e+00 -2.11408380e-02
3.75826836e-01 5.71849704e-01 6.11119807e-01 2.93028653e-01
-1.51115942e+00 1.34277368e+00 8.41410577e-01 9.15422320e-01
-1.52627945e+00 -6.84380710e-01 -5.03935635e-01 4.54841942e-01
1.69124484e-01 -6.70398712e-01 1.45266879e+00 -4.44040477e-01
-1.62576210e+00 -3.24517667e-01 -2.10542362e-02 -9.78924096e-01
4.01381701e-01 -3.13077897e-01 -3.53176028e-01 -3.13092291e-01
-5.98924756e-01 2.39595518e-01 3.41976494e-01 -1.24456131e+00
-8.05530071e-01 -2.52623051e-01 -1.63280994e-01 3.90644401e-01
-9.14184630e-01 -4.83016819e-01 1.82586551e-01 -3.34641218e-01
-6.93085730e-01 -4.70841169e-01 -9.19190869e-02 -3.54863554e-01
5.36463074e-02 -5.71197927e-01 1.06020808e+00 -1.19472814e+00
1.61350536e+00 -1.53708529e+00 9.68512073e-02 -2.08714962e-01
-2.56223410e-01 3.14734638e-01 3.89869124e-01 3.16695094e-01
1.78341612e-01 2.58914456e-02 -6.61594095e-03 -4.69205827e-01
4.48453665e-01 6.85321212e-01 -2.21304014e-01 2.94648260e-01
-9.54887122e-02 9.78063464e-01 -9.22934890e-01 2.28857607e-01
3.09352100e-01 1.21290192e-01 2.27973551e-01 3.29040200e-01
-5.18834412e-01 -6.57367110e-02 -3.80850025e-02 9.72279668e-01
7.46404350e-01 2.54518148e-02 1.44840136e-01 -4.23199743e-01
-3.11454445e-01 1.48472190e-01 -6.61862671e-01 1.04555976e+00
-7.31260478e-01 6.04766071e-01 3.22273552e-01 -1.32166469e+00
7.25075245e-01 5.20359725e-02 3.62618536e-01 -1.14511514e+00
2.56819338e-01 4.41637486e-01 -6.60490151e-03 -2.33426705e-01
4.71833229e-01 -1.60937816e-01 1.97735742e-01 1.02586627e-01
-2.06371859e-01 8.93370733e-02 2.94885784e-01 -2.21626386e-01
7.01797128e-01 -8.80875066e-03 2.97462255e-01 -6.80046856e-01
6.68968678e-01 -1.69302285e-01 8.19343686e-01 2.58278996e-01
-2.57887691e-01 -3.84705514e-01 1.86524570e-01 -3.44819605e-01
-1.08328593e+00 -9.84071612e-01 -6.49616197e-02 6.92378461e-01
3.01271111e-01 1.68719441e-01 -5.77936232e-01 -2.34036639e-01
3.61885786e-01 1.32283759e+00 -3.44299823e-01 -5.22164941e-01
-1.74775392e-01 -1.36114371e+00 9.35048535e-02 9.86680865e-01
7.63649106e-01 -6.19581223e-01 -7.31073916e-01 3.16087574e-01
3.09375376e-01 -5.73897004e-01 -6.15921244e-02 7.92309999e-01
-7.84164548e-01 -5.77083707e-01 -3.73705953e-01 -4.48881328e-01
5.17019272e-01 1.14466213e-01 1.00439715e+00 -8.31039697e-02
6.26489893e-02 2.79905498e-01 -1.84013112e-03 -7.46834934e-01
-3.14558655e-01 -8.55222866e-02 7.68223584e-01 -4.78433341e-01
8.26595500e-02 -4.43211734e-01 -8.68625104e-01 2.60995328e-01
-6.38462245e-01 -2.59760506e-02 5.02787530e-01 1.13954031e+00
5.95968068e-01 1.13120782e+00 1.48783970e+00 2.02785105e-01
1.00481701e+00 -6.84453785e-01 -1.12221432e+00 3.70859563e-01
-1.69662559e+00 -1.00711606e-01 8.07910621e-01 1.73669271e-02
-9.21212494e-01 -3.29233676e-01 2.70072557e-02 -6.64618134e-01
5.36017716e-01 6.73088908e-01 -6.75925553e-01 1.67095482e-01
-4.70144093e-01 6.59238875e-01 1.38137877e-01 -4.49625164e-01
-3.39035764e-02 8.70894730e-01 6.14531100e-01 -3.14133108e-01
4.80664223e-01 -2.11894214e-02 1.87060833e-01 -2.54265994e-01
-4.06804264e-01 -1.52952254e-01 -1.78008914e-01 -4.98017937e-01
4.41748530e-01 -1.30989683e+00 -1.14312756e+00 1.02187371e+00
-8.23357880e-01 -6.55857444e-01 -1.10287167e-01 4.06469889e-02
-3.18887383e-01 7.38340542e-02 -4.57892090e-01 -1.30691528e+00
-8.98032606e-01 -1.16493189e+00 4.69285697e-01 5.49207032e-01
5.39729595e-01 -1.23501062e+00 -3.98961157e-01 1.29353628e-02
7.96938479e-01 1.88756704e-01 1.08992684e+00 -5.14735878e-01
-5.78919649e-01 -2.95028031e-01 -2.52219997e-02 1.16015673e+00
3.05769742e-01 -4.24408734e-01 -7.19236732e-01 -1.05239141e+00
1.72186926e-01 -7.44939074e-02 4.22270834e-01 5.66289008e-01
1.30153704e+00 -6.20569587e-01 -2.06399307e-01 3.73165876e-01
1.79680502e+00 7.70004869e-01 3.99023712e-01 6.27804339e-01
4.71852630e-01 -7.44003132e-02 5.85363209e-01 6.50164843e-01
7.75110364e-01 3.50961685e-01 1.10662043e+00 2.94230729e-01
9.71461013e-02 1.30467832e-01 7.91393578e-01 1.17238712e+00
4.19963986e-01 -6.72946393e-01 -6.96976364e-01 8.31248820e-01
-1.90015590e+00 -4.50909704e-01 1.01215445e-01 2.24328947e+00
6.95318818e-01 4.95814681e-01 5.08012017e-03 3.70131433e-01
3.50199163e-01 -2.58078516e-01 -1.29534495e+00 -6.57679260e-01
-3.32336217e-01 -2.65236169e-01 1.02293026e+00 3.29582542e-01
-6.12337828e-01 2.33441114e-01 5.96692657e+00 8.54590893e-01
-1.05598748e+00 4.41592425e-01 6.84771061e-01 -4.44956750e-01
-2.65597075e-01 -4.31654125e-01 -1.05625653e+00 1.11102664e+00
1.49993038e+00 -8.93972933e-01 8.03609669e-01 7.76648939e-01
9.34359431e-01 -3.22244853e-01 -1.04716516e+00 6.32238090e-01
-2.42607724e-02 -1.34330726e+00 7.36564547e-02 1.01474173e-01
9.27033901e-01 1.16868265e-01 -1.34563819e-01 4.79975671e-01
2.54313558e-01 -8.88467252e-01 9.42600310e-01 9.87420321e-01
4.92476195e-01 -1.12018156e+00 1.25646412e+00 4.47065622e-01
-1.09485185e+00 -9.53860581e-01 -3.32204729e-01 -3.94034028e-01
2.99207479e-01 7.58610427e-01 -8.54408026e-01 8.04267287e-01
1.14048862e+00 5.50334871e-01 -4.72836584e-01 1.11135077e+00
-3.53066474e-02 7.12188005e-01 -5.93088388e-01 -4.59106147e-01
1.71450347e-01 -1.89760432e-01 3.56229186e-01 8.11670601e-01
6.17549419e-01 -2.01659113e-01 9.49169472e-02 6.26701474e-01
-1.91886216e-01 -6.96107507e-01 -1.63184568e-01 -2.00194895e-01
7.96142995e-01 1.40300810e+00 -3.92630845e-01 -4.38244373e-01
-1.08225413e-01 3.76443148e-01 -2.74234444e-01 3.46966654e-01
-8.40666890e-01 -1.49135172e-01 7.44464755e-01 -5.36522567e-02
2.74360210e-01 -4.30003434e-01 -6.84899032e-01 -5.29148281e-01
1.55608421e-02 -3.67295325e-01 2.55839340e-02 -1.29959548e+00
-1.27209842e+00 9.81954634e-02 1.01552688e-01 -9.93093729e-01
-2.27425158e-01 -8.00530255e-01 -9.20681834e-01 1.05314577e+00
-2.17791486e+00 -9.80660737e-01 -5.46335101e-01 7.28540123e-02
8.32610369e-01 -3.00901562e-01 4.28334922e-01 4.55798507e-01
-1.21460378e+00 4.09594178e-01 1.19831216e+00 -6.08939290e-01
1.19716749e-01 -1.70685875e+00 -1.23899586e-01 1.11209106e+00
-1.10791802e+00 9.26103741e-02 8.85420620e-01 -7.45819390e-01
-1.94216144e+00 -1.31066489e+00 1.71205192e-03 1.15016848e-01
9.16447103e-01 -2.02499777e-01 -7.95267642e-01 3.17020386e-01
7.84461439e-01 -2.12947682e-01 2.89158314e-01 -7.42041528e-01
6.73052311e-01 -9.00413334e-01 -1.03709757e+00 3.75218317e-02
5.52657306e-01 -1.50554240e-01 -1.19720563e-01 4.55045432e-01
6.96862042e-01 -4.54377443e-01 -9.94679511e-01 6.86813235e-01
4.48462605e-01 -4.09708589e-01 6.56532884e-01 -4.03458595e-01
-1.81314513e-01 -5.31315863e-01 -2.65380085e-01 -1.83127213e+00
-1.75480679e-01 -5.07552505e-01 -1.19683933e+00 1.35705197e+00
-9.99756753e-02 -8.02543402e-01 3.71405661e-01 3.88174385e-01
-8.29509318e-01 -1.29440939e+00 -1.31044424e+00 -1.15195227e+00
2.33648017e-01 7.38791078e-02 1.01809275e+00 4.03578997e-01
-2.57776439e-01 -1.07990168e-01 -1.93302855e-01 6.87358141e-01
6.17569745e-01 -1.46525502e-01 1.68645605e-01 -9.68783975e-01
2.23433033e-01 -6.73582435e-01 -7.96937197e-03 -6.43159091e-01
1.37576342e-01 -6.10067546e-01 1.88245103e-01 -2.17375422e+00
-7.31403902e-02 -3.34998965e-01 -1.09395206e+00 5.30847549e-01
-2.64927056e-02 -3.53092402e-01 -1.70124844e-01 2.88809445e-02
-3.39308321e-01 1.25571454e+00 6.56618059e-01 -4.91857111e-01
1.70898512e-01 3.50787640e-02 -3.44990283e-01 4.00016665e-01
1.06355476e+00 4.33650501e-02 -7.21638203e-01 -3.76001358e-01
4.39094931e-01 5.32352179e-02 3.67250770e-01 -1.00401211e+00
5.06649971e-01 -8.14210251e-02 6.56094432e-01 -1.13704884e+00
5.22106774e-02 -9.66165006e-01 5.38343847e-01 9.06367958e-01
3.20832223e-01 4.85193580e-01 5.53914964e-01 7.17157066e-01
1.82244461e-02 -3.89848165e-02 5.51069438e-01 2.32427359e-01
-1.03932607e+00 1.29892930e-01 -5.05046844e-01 -8.35990608e-01
1.21025002e+00 -8.64486769e-02 -6.38645053e-01 -2.82654315e-01
-5.90230048e-01 1.27970409e+00 1.76689610e-01 3.91950190e-01
5.70198536e-01 -1.40435481e+00 -1.87293530e-01 -1.26039177e-01
-4.82871473e-01 1.58878773e-01 2.01800391e-01 8.39773953e-01
-3.51420850e-01 7.00245380e-01 -2.30738282e-01 -4.06298846e-01
-5.41847527e-01 6.56691730e-01 9.66951907e-01 -3.50130022e-01
-2.04970362e-03 3.74457210e-01 -4.97749150e-01 -7.21039902e-03
4.73791569e-01 -5.31475961e-01 4.16952148e-02 3.17041755e-01
2.19495848e-01 1.04052091e+00 5.89929700e-01 3.22525501e-02
-3.02511811e-01 -1.87838241e-01 3.76888156e-01 4.01085585e-01
1.42676616e+00 -6.30923152e-01 -3.63574564e-01 7.06366897e-01
7.62200713e-01 -8.75385165e-01 -1.40589237e+00 3.68150800e-01
-1.38212770e-01 1.61249176e-01 8.58993888e-01 -1.47258472e+00
-1.07212985e+00 7.70540714e-01 1.05562222e+00 4.61308122e-01
1.76064134e+00 -8.63056719e-01 9.84250546e-01 3.21503282e-01
4.01232302e-01 -1.40350711e+00 -3.02229702e-01 2.03904197e-01
8.62598538e-01 -8.99944305e-01 1.30260170e-01 7.05646038e-01
-7.85896108e-02 1.24233568e+00 8.51344347e-01 2.45817512e-01
7.37428069e-01 5.14185727e-01 -5.00403702e-01 2.33253837e-01
-9.77343619e-01 1.08866408e-01 -2.17414349e-01 3.23857397e-01
-3.55077893e-01 2.89683253e-01 -4.69396740e-01 1.04438877e+00
1.11645050e-01 1.77913874e-01 7.22508550e-01 9.94696558e-01
-5.01255214e-01 -5.95755696e-01 -1.85551912e-01 6.51248991e-01
-1.70004833e-02 1.67237937e-01 3.51398736e-01 5.12235343e-01
1.62180722e-01 7.98331559e-01 2.32879713e-01 -2.87243754e-01
2.63911694e-01 4.23295088e-02 1.68195918e-01 4.14571241e-02
-1.86362356e-01 -5.35811856e-02 -1.60561632e-02 -8.86839181e-02
-1.64691687e-01 -4.28895831e-01 -1.34978843e+00 -2.27781609e-01
-5.54643154e-01 3.10497731e-01 1.16517758e+00 1.10261023e+00
3.55437368e-01 1.16727412e+00 1.01602602e+00 -1.08856416e+00
-1.06371856e+00 -1.06848145e+00 -7.78486788e-01 -6.99976444e-01
5.56556821e-01 -9.16379392e-01 -6.00115061e-01 -2.64237851e-01] | [5.669229030609131, 2.5439224243164062] |
76520016-dcb9-47fe-b5ba-de23ebdaf5c2 | occlusion-coherence-detecting-and-localizing | 1506.08347 | null | http://arxiv.org/abs/1506.08347v2 | http://arxiv.org/pdf/1506.08347v2.pdf | Occlusion Coherence: Detecting and Localizing Occluded Faces | The presence of occluders significantly impacts object recognition accuracy.
However, occlusion is typically treated as an unstructured source of noise and
explicit models for occluders have lagged behind those for object appearance
and shape. In this paper we describe a hierarchical deformable part model for
face detection and landmark localization that explicitly models part occlusion.
The proposed model structure makes it possible to augment positive training
data with large numbers of synthetically occluded instances. This allows us to
easily incorporate the statistics of occlusion patterns in a discriminatively
trained model. We test the model on several benchmarks for landmark
localization and detection including challenging new data sets featuring
significant occlusion. We find that the addition of an explicit occlusion model
yields a detection system that outperforms existing approaches for occluded
instances while maintaining competitive accuracy in detection and landmark
localization for unoccluded instances. | ['Charless C. Fowlkes', 'Golnaz Ghiasi'] | 2015-06-28 | null | null | null | null | ['occluded-face-detection'] | ['computer-vision'] | [ 6.63563162e-02 -6.28804490e-02 -3.00825208e-01 -5.54869294e-01
-9.06713545e-01 -5.88535726e-01 5.41797400e-01 -5.39361276e-02
-4.66873646e-02 4.02280152e-01 -1.55191831e-02 4.84404825e-02
3.52279752e-01 -4.62332845e-01 -6.27451658e-01 -6.95995033e-01
-2.30141595e-01 6.95728004e-01 3.02491903e-01 1.75236985e-01
-9.94564965e-02 1.03358209e+00 -1.73859060e+00 2.35514686e-01
3.43991607e-01 8.71607602e-01 -2.04467520e-01 4.34906751e-01
-6.43346906e-02 1.79340065e-01 -7.61915922e-01 -2.68176019e-01
5.27474821e-01 2.81215042e-01 -3.56252134e-01 5.78464627e-01
1.43448472e+00 -3.68030280e-01 -2.17887744e-01 5.81584275e-01
6.17983282e-01 6.50037313e-04 8.48953068e-01 -1.11367965e+00
-4.65524137e-01 2.87710056e-02 -1.17439687e+00 3.24989378e-01
4.76907521e-01 2.22274065e-02 8.43916237e-01 -1.24927616e+00
7.82369852e-01 1.62224519e+00 7.62573719e-01 4.95506436e-01
-1.78142333e+00 -5.17589808e-01 4.93830383e-01 -1.66267067e-01
-1.72613895e+00 -9.12607014e-01 8.41863930e-01 -5.25271297e-01
9.02109146e-01 3.55253398e-01 3.11553121e-01 1.00631428e+00
-1.45394146e-01 9.87326980e-01 7.44411707e-01 -4.73186433e-01
8.03922862e-02 2.92951316e-01 3.02433163e-01 8.14698279e-01
4.83661205e-01 8.34150538e-02 -2.49634534e-01 -4.61111277e-01
6.51323199e-01 -4.19256240e-02 -5.07947467e-02 -7.71016479e-01
-3.77387673e-01 6.55946493e-01 3.76531035e-01 7.37970397e-02
-8.62148255e-02 1.87096149e-01 1.67644143e-01 4.85463291e-02
7.11035490e-01 1.33858100e-01 -3.91215891e-01 4.82842863e-01
-1.17951155e+00 2.05991030e-01 7.92333663e-01 9.48016226e-01
8.05216432e-01 4.26718861e-01 -5.32576501e-01 1.18543625e+00
3.82348478e-01 5.65324068e-01 5.22577502e-02 -8.26359451e-01
6.29385337e-02 7.33841062e-01 1.42951354e-01 -1.03188431e+00
-3.16050231e-01 -5.44107735e-01 -2.07538158e-01 3.67504179e-01
4.79544014e-01 1.89144671e-01 -1.46523225e+00 1.75060701e+00
8.27955365e-01 3.03018242e-01 -4.01190400e-01 5.71033239e-01
1.07851994e+00 1.40595570e-01 1.51850641e-01 -1.43414393e-01
1.49015307e+00 -8.44369352e-01 -7.84825146e-01 -3.27018321e-01
3.51640105e-01 -1.00130403e+00 8.55529964e-01 1.74512729e-01
-9.18445826e-01 -5.77741206e-01 -9.20423746e-01 -2.31956154e-01
-3.06038946e-01 5.59572875e-01 8.38361323e-01 9.47312057e-01
-1.14217854e+00 1.77769110e-01 -8.49794149e-01 -3.68624777e-01
8.26345444e-01 8.00444603e-01 -5.98975420e-01 -2.65491456e-01
-2.88118482e-01 7.68445194e-01 -1.48734823e-01 -3.05290464e-02
-9.94865119e-01 -5.43305933e-01 -1.11311150e+00 -1.25367075e-01
2.56949306e-01 -3.36230755e-01 1.00564826e+00 -6.70158327e-01
-9.24411118e-01 1.02542436e+00 -6.63649321e-01 -1.62713990e-01
5.98535955e-01 -1.63679883e-01 -2.63897091e-01 5.99630103e-02
2.36257724e-02 8.66920173e-01 1.29769099e+00 -1.57070637e+00
-2.93644547e-01 -5.22034049e-01 -8.75539407e-02 -4.39902805e-02
-3.18656117e-01 1.36150837e-01 -8.86848152e-01 -7.07076728e-01
3.08742166e-01 -8.33174527e-01 -2.13419255e-02 4.59034353e-01
-3.17715436e-01 -2.36368656e-01 1.35874724e+00 -5.29722273e-01
8.02947640e-01 -2.13930702e+00 -2.64065981e-01 1.71194836e-01
2.13249326e-01 1.77170888e-01 -4.39775050e-01 8.43764618e-02
-1.05136506e-01 1.28055200e-01 1.32121518e-01 -1.01491225e+00
-5.98761886e-02 4.51816618e-01 -2.03736454e-01 6.54368460e-01
4.10370678e-01 8.86410117e-01 -3.44282866e-01 -5.82753122e-01
2.52090693e-01 8.33479822e-01 -8.08395267e-01 1.75055023e-02
-4.60577272e-02 2.03535467e-01 -2.03628406e-01 1.29869187e+00
9.64656830e-01 -4.11941968e-02 -1.33882001e-01 -1.97847933e-01
2.91527957e-01 1.09638609e-01 -1.50870335e+00 1.29333472e+00
-1.64106920e-01 7.39220738e-01 3.67528349e-01 -4.47016418e-01
8.99355114e-01 3.09654504e-01 4.94021088e-01 -2.54777461e-01
2.00469449e-01 -7.39397407e-02 -1.30894572e-01 -2.06042886e-01
2.87120938e-01 4.06893156e-02 4.29537803e-01 1.04832470e-01
-3.87161225e-02 -3.93333333e-03 1.69461921e-01 -1.28034540e-02
1.15437496e+00 5.95188327e-02 1.43587261e-01 -1.89621657e-01
3.43431920e-01 -2.03163564e-01 6.38950467e-01 8.41918766e-01
-3.11759740e-01 8.32316875e-01 7.46381581e-02 -7.27179587e-01
-7.76027203e-01 -1.13723814e+00 -8.36277485e-01 1.10313654e+00
1.37448981e-02 -3.52988541e-01 -7.35470414e-01 -7.59635389e-01
4.36397702e-01 3.00903261e-01 -8.37075830e-01 1.19847819e-01
-5.94442368e-01 -1.02180934e+00 4.22620565e-01 6.04977310e-01
2.60141399e-02 -9.79370236e-01 -3.59588908e-03 7.63958693e-02
1.50121614e-01 -1.46905231e+00 -5.29892981e-01 1.05569862e-01
-7.95283377e-01 -1.13704920e+00 -3.76312733e-01 -8.80736768e-01
1.15711689e+00 4.08917814e-01 1.07171392e+00 3.66705418e-01
-1.14636695e+00 6.73167706e-01 8.08463395e-02 -4.66761351e-01
-1.30179346e-01 -1.45453066e-01 1.87856376e-01 1.91403553e-01
4.69722062e-01 -4.33087081e-01 -5.67083538e-01 5.25095999e-01
-6.63687170e-01 -4.04747158e-01 2.87124366e-01 8.13366115e-01
8.10896039e-01 -3.13493550e-01 2.69098699e-01 -8.88110936e-01
6.37283176e-02 -1.64912149e-01 -6.78659737e-01 2.16207907e-01
-1.55548573e-01 -1.35246471e-01 -4.46001701e-02 -9.01702225e-01
-9.25378501e-01 4.97393638e-01 -1.30934626e-01 -5.29889643e-01
-3.67846161e-01 -2.90314645e-01 -3.03852022e-01 -7.23040164e-01
6.48620546e-01 -8.29570740e-02 -1.83465555e-01 -9.12523508e-01
1.61080673e-01 3.30384105e-01 3.44273746e-01 -7.88149774e-01
9.70166743e-01 7.96431661e-01 -5.91639876e-02 -1.07383132e+00
-5.43390095e-01 -5.17716408e-01 -7.64179170e-01 1.46775812e-01
4.65974361e-01 -1.03566742e+00 -5.17114282e-01 1.44247442e-01
-1.17020941e+00 -5.74065819e-02 -3.86302590e-01 1.66481599e-01
-2.12914035e-01 3.33289176e-01 -5.74860752e-01 -1.00640869e+00
-4.02669720e-02 -1.29189467e+00 1.66733086e+00 3.03903706e-02
-1.16081603e-01 -8.52009118e-01 -8.80272873e-03 3.41970056e-01
1.72070920e-01 4.15294528e-01 4.68956709e-01 -5.75677037e-01
-7.73439527e-01 -9.03300107e-01 -1.82151720e-01 2.29089931e-01
3.52871716e-01 3.19684386e-01 -1.53590703e+00 -6.07599616e-01
-6.61601350e-02 -1.70523435e-01 9.52261925e-01 4.61813152e-01
1.06433463e+00 -2.30396688e-01 -7.91777730e-01 6.67103112e-01
1.12738752e+00 6.33047298e-02 4.90750462e-01 -1.54101744e-01
7.59868205e-01 5.78034222e-01 2.88618565e-01 2.23781794e-01
4.33172844e-02 1.08785510e+00 3.13516527e-01 -4.05370414e-01
-4.78513837e-01 -1.57125175e-01 1.15759328e-01 8.43264312e-02
1.17008246e-01 -1.79295257e-01 -8.10576260e-01 5.13041556e-01
-1.40162897e+00 -7.65706837e-01 -1.14538066e-01 2.11980152e+00
6.55200303e-01 -5.73727712e-02 2.23534927e-02 1.23765185e-01
7.28709102e-01 8.05839598e-02 -3.46358448e-01 -7.05993250e-02
-1.05633438e-01 3.13922405e-01 2.86532015e-01 9.12481785e-01
-1.50403595e+00 1.13703632e+00 7.41166449e+00 8.90722454e-01
-8.81105363e-01 2.43076533e-01 6.09357119e-01 -2.58965194e-01
2.32561282e-03 -1.61537364e-01 -1.50428557e+00 1.77532524e-01
2.50561774e-01 4.18302655e-01 7.59333149e-02 1.20053792e+00
-4.04103436e-02 -9.31496546e-02 -1.29980457e+00 9.09261525e-01
2.38633186e-01 -8.80659997e-01 1.20099202e-01 2.79264003e-01
8.26427221e-01 -1.06879182e-01 3.82463098e-01 1.90087318e-01
1.81576237e-01 -1.14685512e+00 4.11916912e-01 -2.97292694e-02
6.46376193e-01 -2.58839846e-01 3.70908260e-01 6.68139607e-02
-1.37955999e+00 9.04159024e-02 -5.50645769e-01 1.13853477e-01
-8.45685601e-02 2.67772794e-01 -1.16200066e+00 -3.43753733e-02
4.57759082e-01 2.96085626e-01 -9.61624026e-01 1.35794866e+00
-1.34257555e-01 4.91273016e-01 -8.26417148e-01 4.70234364e-01
-1.83000222e-01 1.25062153e-01 6.24100208e-01 1.28527987e+00
-1.14386410e-01 -9.31866542e-02 5.74726701e-01 6.28292978e-01
-1.25840962e-01 2.90765077e-01 -8.54916811e-01 4.05398846e-01
4.53165621e-01 1.39644825e+00 -8.80943596e-01 -2.25807473e-01
-5.58557987e-01 6.96623743e-01 5.24709582e-01 5.73918521e-01
-7.13083446e-01 3.16870511e-01 8.16179037e-01 3.96561474e-01
5.07126808e-01 -1.65421069e-01 -3.16226155e-01 -1.04173040e+00
2.33133614e-01 -8.43192637e-01 3.74731064e-01 -2.59035915e-01
-1.17247772e+00 6.78159118e-01 9.97376814e-03 -8.88331115e-01
4.11905870e-02 -6.08828843e-01 -5.01080692e-01 7.29962707e-01
-1.42515159e+00 -1.43802547e+00 -3.60476315e-01 2.56129950e-01
6.26630187e-01 1.93057097e-02 8.87849092e-01 5.74268639e-01
-8.32499325e-01 9.20681834e-01 -3.42457294e-02 2.56473392e-01
7.35395253e-01 -1.16339040e+00 4.17968661e-01 6.90841496e-01
7.12398708e-01 8.52538288e-01 7.36756146e-01 -6.48065746e-01
-1.15650928e+00 -1.03006136e+00 5.17807126e-01 -9.57656980e-01
1.86443031e-01 -8.28321040e-01 -1.11113155e+00 9.53921080e-01
-3.40122551e-01 6.82229519e-01 6.34092271e-01 3.08264494e-01
-7.14811981e-01 -2.00546876e-01 -1.45592737e+00 4.88649994e-01
1.17013586e+00 -4.93277997e-01 -3.47603440e-01 6.15883470e-01
2.57335514e-01 -4.76884246e-01 -4.33650196e-01 7.30239749e-01
7.02940941e-01 -6.67619467e-01 1.41754866e+00 -5.66824138e-01
-3.22368771e-01 -3.47397417e-01 -2.81651825e-01 -5.39672673e-01
-2.50678301e-01 -4.40823972e-01 -3.71607333e-01 1.31231225e+00
3.02295715e-01 -4.40535843e-01 1.12868178e+00 1.01279986e+00
1.33209229e-01 -6.73655391e-01 -1.05841756e+00 -9.04630125e-01
-3.46869737e-01 -2.78724998e-01 2.87055820e-01 6.80666327e-01
-6.07815444e-01 6.30910322e-02 -2.84030616e-01 4.03919429e-01
7.04989016e-01 4.45829034e-02 9.69635129e-01 -1.40149415e+00
-2.10246101e-01 -2.23106623e-01 -7.33672857e-01 -1.06358922e+00
3.59406501e-01 -8.48993421e-01 -1.44729748e-01 -1.30250013e+00
2.81528234e-01 -6.20635986e-01 9.00097750e-03 8.67211998e-01
-2.79850274e-01 1.03131545e+00 -1.26528740e-01 1.33577853e-01
-4.36977923e-01 4.22920942e-01 1.07022476e+00 -2.99427807e-01
-3.43153179e-01 5.54522537e-02 -4.77118462e-01 1.00850856e+00
4.23394173e-01 -6.05508447e-01 -1.95853725e-01 -5.13735175e-01
-6.85233653e-01 -3.69605631e-01 4.21717644e-01 -9.56193984e-01
-6.50659576e-02 1.24730773e-01 1.00190854e+00 -5.67000151e-01
9.91027653e-01 -9.48052526e-01 1.61973640e-01 2.65490532e-01
-4.89401966e-02 -1.45395964e-01 6.77293718e-01 5.45149803e-01
-1.79866832e-02 -1.43212065e-01 9.49022532e-01 1.90894634e-01
-4.20382172e-01 5.34401119e-01 7.03666825e-04 -2.70299967e-02
1.07274246e+00 -4.18677390e-01 -1.82809323e-01 -8.55144113e-02
-1.02547979e+00 -8.25432241e-02 6.88767314e-01 7.04143047e-01
5.19258618e-01 -1.24182689e+00 -5.82594752e-01 5.50009251e-01
7.18273595e-02 -2.67386496e-01 -3.14483531e-02 5.59300363e-01
-4.08885241e-01 2.51431525e-01 2.01217026e-01 -7.97945380e-01
-1.87115109e+00 6.79343879e-01 3.31608355e-01 3.99687234e-03
-5.12973547e-01 9.35889304e-01 5.70433259e-01 -2.98357099e-01
7.01094329e-01 -1.68130293e-01 -1.11745603e-01 7.92590529e-02
6.26325190e-01 1.56366125e-01 1.02864213e-01 -9.02556539e-01
-4.62650448e-01 8.04749489e-01 -5.38306415e-01 2.37983614e-01
1.10311091e+00 2.07352906e-01 5.30426055e-02 1.13619238e-01
1.15207481e+00 3.87947559e-01 -1.25009918e+00 -2.88444579e-01
6.62652925e-02 -7.80763030e-01 -8.93639773e-02 -6.19971216e-01
-1.12751067e+00 9.16985631e-01 8.54953110e-01 -1.02541991e-01
6.05140984e-01 9.41629484e-02 1.93687037e-01 4.08558369e-01
4.96599734e-01 -5.52353919e-01 1.18824869e-01 1.86717078e-01
9.46398795e-01 -1.39977098e+00 3.19043010e-01 -9.54967916e-01
-1.46124348e-01 7.20149517e-01 8.93439710e-01 -6.34913519e-02
7.33805656e-01 7.08012998e-01 1.77348077e-01 -2.24528164e-01
-5.30794561e-01 -3.67166013e-01 7.45973110e-01 7.43265808e-01
3.65765303e-01 -9.24004018e-02 -1.02856241e-01 1.81167409e-01
1.30798072e-01 -5.36793888e-01 -1.02619402e-01 9.39459682e-01
-3.59570801e-01 -1.38698673e+00 -5.88615537e-01 4.05180961e-01
-5.96090317e-01 4.61026095e-02 -6.44982159e-01 1.04805374e+00
1.98581487e-01 7.64368236e-01 1.12325117e-01 2.50057094e-02
1.57284766e-01 2.68892616e-01 6.67641580e-01 -1.01857042e+00
-4.93992001e-01 4.47916567e-01 -4.62293699e-02 -3.99111599e-01
1.47577440e-02 -7.98857331e-01 -1.00235748e+00 3.64335366e-02
-7.36647725e-01 2.51533240e-02 6.07933760e-01 6.26299262e-01
4.23598021e-01 4.55747694e-01 3.31870437e-01 -1.09634554e+00
-3.61320496e-01 -9.39843476e-01 -5.79568565e-01 4.83136684e-01
5.25428653e-01 -1.20811808e+00 -5.05021572e-01 4.16794419e-02] | [13.35744571685791, 0.3366940915584564] |
acee5a47-8fb9-4882-a154-b5fa86bd9d28 | dnnshield-dynamic-randomized-model | 2208.00498 | null | https://arxiv.org/abs/2208.00498v1 | https://arxiv.org/pdf/2208.00498v1.pdf | DNNShield: Dynamic Randomized Model Sparsification, A Defense Against Adversarial Machine Learning | DNNs are known to be vulnerable to so-called adversarial attacks that manipulate inputs to cause incorrect results that can be beneficial to an attacker or damaging to the victim. Recent works have proposed approximate computation as a defense mechanism against machine learning attacks. We show that these approaches, while successful for a range of inputs, are insufficient to address stronger, high-confidence adversarial attacks. To address this, we propose DNNSHIELD, a hardware-accelerated defense that adapts the strength of the response to the confidence of the adversarial input. Our approach relies on dynamic and random sparsification of the DNN model to achieve inference approximation efficiently and with fine-grain control over the approximation error. DNNSHIELD uses the output distribution characteristics of sparsified inference compared to a dense reference to detect adversarial inputs. We show an adversarial detection rate of 86% when applied to VGG16 and 88% when applied to ResNet50, which exceeds the detection rate of the state of the art approaches, with a much lower overhead. We demonstrate a software/hardware-accelerated FPGA prototype, which reduces the performance impact of DNNSHIELD relative to software-only CPU and GPU implementations. | ['Radu Teodorescu', 'Kristin Barber', 'Saikat Majumdar', 'Mohammad Hossein Samavatian'] | 2022-07-31 | null | null | null | null | ['machine-learning', 'machine-learning'] | ['methodology', 'miscellaneous'] | [ 4.14685309e-02 -5.10832034e-02 -1.74561776e-02 -3.43924999e-01
-4.61761832e-01 -1.11510098e+00 5.18682241e-01 8.67190510e-02
-5.50431609e-01 5.78686476e-01 -7.78321028e-02 -6.77330673e-01
3.91047210e-01 -1.17288470e+00 -1.05176592e+00 -3.70958984e-01
-8.96521807e-02 3.22493494e-01 5.57488024e-01 -2.76326537e-01
2.17405766e-01 9.40931797e-01 -1.41565418e+00 5.73793769e-01
3.86832148e-01 1.05041564e+00 -7.43398607e-01 1.00040531e+00
2.79730558e-01 6.51104093e-01 -1.25317335e+00 -7.49619722e-01
7.57341146e-01 -6.10312484e-02 -3.73317957e-01 -8.37635875e-01
1.11623013e+00 -9.83026266e-01 -8.40150774e-01 1.45863640e+00
5.56502998e-01 -3.12331766e-01 2.60826945e-01 -1.39211059e+00
-1.11878857e-01 9.93856370e-01 -1.93380281e-01 4.36744183e-01
8.36453773e-03 6.18318021e-01 4.97843951e-01 -3.47333610e-01
3.97078633e-01 1.37903333e+00 8.42982888e-01 7.68870115e-01
-1.29507768e+00 -1.19871664e+00 -1.86510667e-01 -3.03637590e-02
-1.42299819e+00 -3.12473625e-01 2.73541540e-01 8.32502265e-03
1.01495647e+00 4.67431277e-01 4.03581619e-01 1.39888561e+00
5.92437267e-01 3.23962063e-01 1.10864329e+00 -1.06192119e-01
7.61015475e-01 4.78282534e-02 2.72051811e-01 4.68086272e-01
8.10918570e-01 7.31449127e-01 -3.12522531e-01 -9.22103763e-01
6.33515656e-01 -1.93713695e-01 -2.30608985e-01 2.57889628e-01
-6.92420185e-01 1.00845039e+00 7.16143489e-01 -8.38307887e-02
-1.90079913e-01 8.34363520e-01 1.00459862e+00 4.11707401e-01
-4.80865315e-02 5.90931296e-01 -5.41609108e-01 -7.40449205e-02
-1.06611347e+00 5.57226896e-01 1.16945338e+00 7.50646949e-01
4.99971330e-01 6.69579089e-01 -1.61810860e-01 -2.18205407e-01
4.74626720e-02 7.70373344e-01 1.84334423e-02 -7.71622121e-01
3.34675878e-01 1.19938388e-01 -2.68580943e-01 -1.04917312e+00
-9.22846645e-02 -5.82477212e-01 -7.47652829e-01 9.40270066e-01
5.55368006e-01 -4.43968207e-01 -8.74247968e-01 1.70857811e+00
2.42786109e-01 4.85399425e-01 1.24646783e-01 7.75027812e-01
2.36144483e-01 6.22751951e-01 8.50918368e-02 4.09994811e-01
1.30538702e+00 -3.83033246e-01 -2.31820032e-01 -1.79997966e-01
5.17727077e-01 -5.53102195e-01 8.87786031e-01 6.65291667e-01
-8.35623562e-01 -3.43023151e-01 -1.70832443e+00 2.92661816e-01
-4.12260264e-01 -3.47314298e-01 5.91530442e-01 1.53824902e+00
-9.10435200e-01 7.60812342e-01 -1.01532197e+00 2.70588666e-01
7.15711713e-01 5.92034698e-01 -1.11025693e-02 1.06167190e-01
-1.43313861e+00 7.71275997e-01 4.68415767e-01 -2.91647226e-01
-1.19592559e+00 -1.15252483e+00 -5.54925561e-01 2.30882570e-01
9.46710780e-02 -5.12362838e-01 9.82966244e-01 -8.98285151e-01
-1.30406582e+00 4.53580350e-01 6.23138309e-01 -1.37190545e+00
7.46158481e-01 -1.48041397e-01 -6.09857976e-01 2.43173257e-01
-4.05756146e-01 3.65912735e-01 1.07159436e+00 -8.01464617e-01
-3.12725276e-01 -2.68826693e-01 3.10745627e-01 -5.20489514e-01
-4.91457283e-01 -7.96668530e-02 1.74935937e-01 -9.54739511e-01
-4.88190502e-01 -1.15585399e+00 -4.56944853e-01 3.82527113e-01
-4.43542540e-01 5.68544984e-01 1.13615286e+00 -2.39894390e-01
1.04037607e+00 -2.12660408e+00 -5.62182426e-01 7.69181073e-01
3.13755423e-01 1.03307712e+00 6.00225329e-02 1.84478998e-01
8.18363354e-02 1.05498299e-01 -1.27537340e-01 6.35896474e-02
6.97290376e-02 3.92218083e-01 -1.23307097e+00 7.64369309e-01
2.12502386e-03 6.12226069e-01 -8.09375763e-01 1.11857191e-01
3.30858864e-02 5.30206978e-01 -9.03288960e-01 7.08338320e-02
-1.90282598e-01 -3.74285728e-01 -3.88032824e-01 4.84080285e-01
1.04526997e+00 3.77217144e-01 8.25213641e-02 -4.48960036e-01
3.75295639e-01 2.36296818e-01 -1.26452637e+00 9.81120825e-01
-3.56284052e-01 6.57001376e-01 2.22603843e-01 -3.84805977e-01
9.38624442e-01 -1.26021132e-02 -4.67480659e-01 -3.39996397e-01
3.21952671e-01 4.18999881e-01 1.50378823e-01 4.59283471e-01
4.45298761e-01 1.73360005e-01 -3.90647709e-01 5.72945118e-01
-8.13408792e-02 -3.37273240e-01 -4.20568377e-01 3.65058780e-01
1.59410477e+00 -3.85057062e-01 1.33921504e-01 -5.56515038e-01
4.79756594e-01 -3.75515148e-02 3.73504817e-01 1.32771695e+00
-1.67216048e-01 1.74890175e-01 5.71254492e-01 -8.10875833e-01
-1.08054972e+00 -1.39281034e+00 -2.21212223e-01 7.76509881e-01
1.83608271e-02 -5.20247102e-01 -1.22874713e+00 -9.39213753e-01
3.97572547e-01 1.08744013e+00 -5.61686039e-01 -6.45963907e-01
-5.82933247e-01 -5.46211421e-01 1.85889971e+00 8.38431835e-01
7.17715561e-01 -5.94585717e-01 -8.17033291e-01 1.23349234e-01
7.59283900e-01 -1.22491002e+00 -4.37130749e-01 3.09744984e-01
-7.30719745e-01 -9.43413496e-01 1.71883807e-01 -8.08158964e-02
5.96793771e-01 -3.28767955e-01 1.15250707e+00 2.26091027e-01
-3.61824334e-01 1.27802253e-01 1.52648855e-02 -5.22658229e-01
-1.02496696e+00 -9.81223360e-02 4.70333427e-01 -3.78678411e-01
2.01911986e-01 -8.56849492e-01 -3.98530811e-01 2.46188626e-01
-1.21719074e+00 -7.08699465e-01 3.91103417e-01 8.05357516e-01
5.23253977e-01 8.13792199e-02 3.36123228e-01 -1.09994411e+00
5.70822835e-01 -1.95070416e-01 -1.07999384e+00 -1.49097487e-01
-5.24850607e-01 2.39119813e-01 1.39776516e+00 -9.09420013e-01
-4.08888519e-01 6.58517238e-03 -5.73001325e-01 -7.99301386e-01
-4.13686596e-03 -1.34658054e-01 -3.28829169e-01 -7.28147626e-01
1.37115753e+00 -1.11348252e-03 -1.78394899e-01 6.67177588e-02
3.15986097e-01 2.95755595e-01 9.21884418e-01 -8.43472958e-01
1.07296705e+00 6.09078288e-01 4.83126968e-01 -5.20759404e-01
-5.02193868e-01 5.60181081e-01 -1.97091214e-02 1.71279788e-01
2.66308188e-01 -8.02308202e-01 -9.05917406e-01 6.12861335e-01
-1.07150030e+00 -3.91223252e-01 -4.25883114e-01 1.41351923e-01
-2.37750933e-01 2.18622997e-01 -8.85704935e-01 -4.15033996e-01
-7.66437352e-01 -1.25322223e+00 6.17766559e-01 -1.16442055e-01
-2.23045930e-01 -7.10875392e-01 -2.81053364e-01 -2.98067391e-01
8.16893816e-01 6.52855217e-01 7.24223256e-01 -1.20423555e+00
-5.93257487e-01 -5.53435028e-01 -1.46017060e-01 7.00858712e-01
-5.25140166e-01 1.02898441e-01 -1.11361325e+00 -5.22477448e-01
1.46459460e-01 -3.94801676e-01 7.03021526e-01 -5.68986014e-02
1.25205517e+00 -8.60417128e-01 -1.13261342e-01 1.02846992e+00
1.56775784e+00 -7.27384835e-02 9.58412111e-01 2.76687413e-01
6.67654455e-01 -2.31223106e-01 3.51107389e-01 5.73089182e-01
-3.26365680e-01 5.61590910e-01 8.53622913e-01 1.06870532e-01
1.46782945e-03 -2.60205090e-01 5.49824834e-01 -1.64320797e-01
4.16054130e-01 -2.79988617e-01 -7.76546359e-01 2.37554330e-02
-1.24628162e+00 -1.04084039e+00 -6.26816377e-02 2.17634296e+00
8.63907039e-01 8.92295361e-01 -4.87388745e-02 2.76286155e-01
5.02458811e-01 1.16282262e-01 -6.70200527e-01 -1.12482846e+00
-2.37748902e-02 8.58936369e-01 1.22609341e+00 3.86082768e-01
-9.24028933e-01 9.34292912e-01 6.64865017e+00 1.14116991e+00
-1.15299261e+00 -1.30090732e-02 5.99804878e-01 -1.25173077e-01
-2.21709982e-01 -5.98779805e-02 -1.00561297e+00 6.59100175e-01
1.41961920e+00 -2.58477330e-01 5.66548467e-01 1.24466062e+00
-3.46906811e-01 3.29790920e-01 -1.23830140e+00 6.47255182e-01
-2.94888113e-02 -1.54516554e+00 3.18477929e-01 1.08825557e-01
5.84690809e-01 5.10239899e-02 3.51274610e-01 2.84147978e-01
7.40739107e-01 -1.07985353e+00 7.28446841e-01 6.60188347e-02
7.71799803e-01 -1.42807090e+00 9.64691043e-01 1.95265338e-01
-8.69044662e-01 -2.35856473e-02 -3.97837728e-01 -1.22809976e-01
-1.71709150e-01 3.92106235e-01 -9.69132841e-01 -1.38517335e-01
6.01858377e-01 -2.14015588e-01 -5.19415319e-01 4.28131551e-01
-4.31803256e-01 9.47558165e-01 -6.85029030e-01 -1.54329106e-01
4.23621535e-01 5.51332593e-01 7.16640592e-01 1.40327883e+00
1.06666707e-01 4.80266148e-03 8.75639990e-02 8.90590549e-01
-1.42565653e-01 -5.07018268e-01 -5.34805357e-01 1.28261000e-01
8.86097550e-01 8.19365799e-01 -5.51813781e-01 -4.45170999e-01
1.24308810e-01 1.01468730e+00 9.55299288e-03 -7.60934725e-02
-1.19495285e+00 -6.90056741e-01 1.38918853e+00 1.36631191e-01
3.82143766e-01 -1.65841475e-01 -4.76492524e-01 -8.81289840e-01
-2.90778250e-01 -1.37224531e+00 3.75732958e-01 -1.72919780e-01
-1.23902559e+00 7.98188925e-01 -8.37550834e-02 -9.81782913e-01
-3.43051314e-01 -6.98385775e-01 -5.84206223e-01 8.85587811e-01
-7.68948972e-01 -9.28528666e-01 -8.39959309e-02 7.10048676e-01
6.60910085e-02 -3.64057690e-01 9.67985332e-01 2.14525953e-01
-1.63556531e-03 1.47584188e+00 -9.68354344e-02 3.63849610e-01
5.08923709e-01 -9.38263178e-01 1.22036147e+00 1.29842114e+00
3.57502699e-02 6.72443807e-01 1.02672958e+00 -6.78947866e-01
-1.67176068e+00 -1.42993939e+00 2.81400960e-02 -3.09689283e-01
8.57079327e-01 -5.74886441e-01 -9.54555273e-01 6.28136039e-01
-1.95484534e-01 6.44145787e-01 4.98577774e-01 -4.53307271e-01
-1.06286037e+00 -2.31722116e-01 -1.89026713e+00 8.83478820e-01
7.46548533e-01 -7.47428954e-01 -2.86195070e-01 1.01381294e-01
9.07776177e-01 -1.00088298e+00 -8.12640786e-01 1.73702121e-01
5.36145031e-01 -1.01723552e+00 1.24533892e+00 -6.06734514e-01
1.30368039e-01 -3.81508678e-01 -5.15879095e-01 -1.00807488e+00
-5.44973202e-02 -6.95072055e-01 -5.18845141e-01 9.23764408e-01
9.58700553e-02 -9.90864277e-01 9.22984123e-01 4.90122199e-01
-2.98937913e-02 -5.15182674e-01 -9.98964071e-01 -9.19000089e-01
1.27461135e-01 -6.23788714e-01 1.04199183e+00 7.54002869e-01
-4.80503529e-01 -3.29366982e-01 -2.83895582e-01 8.80640686e-01
8.82056832e-01 -3.59466314e-01 9.12883282e-01 -6.08485222e-01
-6.58602715e-01 -3.07886630e-01 -1.22162557e+00 -6.36104643e-01
2.02997103e-01 -6.44454777e-01 -2.56253779e-01 -4.52355802e-01
-5.34081578e-01 -3.62228185e-01 -1.10056207e-01 6.04325652e-01
8.80295858e-02 6.59566522e-01 1.63214952e-01 -2.09785372e-01
-6.21759892e-02 -4.31899242e-02 4.14084166e-01 -2.39812791e-01
6.55992329e-02 1.11711986e-01 -5.67914128e-01 7.96121061e-01
7.14126647e-01 -7.96130478e-01 -3.85418028e-01 -3.13414574e-01
9.31087360e-02 -3.38286102e-01 6.12209380e-01 -1.41502321e+00
3.86442006e-01 2.05406070e-01 5.16875327e-01 -3.89320374e-01
1.61436230e-01 -8.94638121e-01 2.37185597e-01 1.07371902e+00
-3.17680776e-01 3.42770994e-01 8.15486014e-01 6.15783572e-01
1.93454534e-01 -1.84165224e-01 1.03526878e+00 2.27539927e-01
-4.14924830e-01 4.05986995e-01 -5.28073311e-01 1.65388882e-01
1.04173255e+00 3.00298035e-02 -6.70865119e-01 -1.72145829e-01
-2.96130806e-01 -1.88984632e-01 6.85216606e-01 -5.58935627e-02
6.67803824e-01 -1.25185692e+00 -6.29029155e-01 3.26386690e-01
-3.07560951e-01 -4.58602548e-01 4.64056581e-02 -5.47966138e-02
-1.03009379e+00 -2.88514998e-02 -5.66411793e-01 -3.09876919e-01
-1.35637641e+00 7.99274862e-01 5.33985019e-01 -2.20682725e-01
-5.95753372e-01 8.83614898e-01 -1.81480557e-01 -1.46732584e-01
2.92959511e-01 -3.82997960e-01 6.84906423e-01 -5.35031557e-01
9.52758014e-01 4.80138183e-01 3.59895200e-01 -2.07090899e-01
-5.75099111e-01 -1.32603332e-01 -2.69725442e-01 1.78433415e-02
8.14661741e-01 6.92518115e-01 1.29226595e-01 -2.99448818e-01
1.14572215e+00 2.19267786e-01 -1.07967591e+00 -4.56433520e-02
-4.88521338e-01 -7.36686885e-01 2.75280803e-01 -8.51745605e-01
-1.16571522e+00 7.23454893e-01 7.35452235e-01 6.22440241e-02
1.10209262e+00 -5.94358087e-01 1.06392241e+00 5.85362375e-01
5.97511411e-01 -6.35285199e-01 -1.64969698e-01 5.15105546e-01
3.31719339e-01 -6.31518602e-01 2.84369558e-01 -3.72235686e-01
-2.68639028e-01 1.26079237e+00 7.46647239e-01 -7.29206622e-01
5.68824351e-01 1.25105309e+00 -1.87645838e-01 1.84142739e-01
-5.99705398e-01 6.02115571e-01 -1.01199694e-01 7.67877162e-01
-3.97638500e-01 4.59359027e-02 1.12026259e-01 5.34660757e-01
-4.22867209e-01 -3.28367621e-01 6.31515265e-01 8.54479551e-01
-1.62936151e-01 -9.92062569e-01 -6.56359911e-01 3.77909184e-01
-7.67950177e-01 -2.58573085e-01 -2.65035868e-01 7.29001820e-01
6.74467906e-02 6.22626007e-01 2.05116019e-01 -7.61593401e-01
2.41953462e-01 -5.10359466e-01 4.09678251e-01 -2.74058193e-01
-1.33334124e+00 -6.25080407e-01 1.04982957e-01 -1.00308907e+00
5.60856223e-01 -3.19314778e-01 -1.19842935e+00 -9.64968562e-01
-1.42493583e-02 -1.80290446e-01 6.22308969e-01 6.06833637e-01
4.38284844e-01 4.27495539e-01 5.83379984e-01 -8.31816971e-01
-1.24093258e+00 -7.45383129e-02 -4.51112032e-01 6.77400008e-02
3.30074966e-01 -2.16228560e-01 -6.92391694e-01 -5.66122055e-01] | [5.670760154724121, 7.71614933013916] |
abd8f278-fb58-43b4-a7a5-540b80409ae5 | latent-factor-guided-convolutional-neural | null | null | http://openaccess.thecvf.com/content_cvpr_2016/html/Wen_Latent_Factor_Guided_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Wen_Latent_Factor_Guided_CVPR_2016_paper.pdf | Latent Factor Guided Convolutional Neural Networks for Age-Invariant Face Recognition | While considerable progresses have been made on face recognition, age-invariant face recognition (AIFR) still remains a major challenge in real world applications of face recognition systems. The major difficulty of AIFR arises from the fact that the facial appearance is subject to significant intra-personal changes caused by the aging process over time. In order to address this problem, we propose a novel deep face recognition framework to learn the age-invariant deep face features through a carefully designed CNN model. To the best of our knowledge, this is the first attempt to show the effectiveness of deep CNNs in advancing the state-of-the-art of AIFR. Extensive experiments are conducted on several public domain face aging datasets (MORPH Album2, FGNET, and CACD-VS) to demonstrate the effectiveness of the proposed model over the state-of-the-art. We also verify the excellent generalization of our new model on the famous LFW dataset. | ['Yandong Wen', 'Yu Qiao', 'Zhifeng Li'] | 2016-06-01 | null | null | null | cvpr-2016-6 | ['age-invariant-face-recognition'] | ['computer-vision'] | [-1.21132255e-01 -1.80462018e-01 -4.74780723e-02 -7.09229767e-01
-6.53891712e-02 2.75056571e-01 3.26946288e-01 -4.80066419e-01
-2.04155445e-01 6.60695314e-01 -4.77639809e-02 1.11535914e-01
-5.55288941e-02 -7.29853749e-01 -5.35719097e-01 -6.45973802e-01
-3.88166428e-01 3.31959836e-02 -2.94318467e-01 -3.35280120e-01
-1.21964879e-01 8.09818208e-01 -1.98776257e+00 -6.00915179e-02
8.83378208e-01 1.58840191e+00 -4.69163269e-01 5.83864786e-02
1.20249003e-01 4.72599357e-01 -4.78246361e-01 -7.91880608e-01
2.78439492e-01 -1.17992379e-01 -4.66440737e-01 1.33604050e-01
1.09892714e+00 -6.16107047e-01 -8.24081838e-01 9.79381621e-01
8.41479540e-01 -1.20322645e-01 4.90540355e-01 -1.49579275e+00
-1.19405282e+00 1.41552627e-01 -6.31040394e-01 1.51236087e-01
1.36644736e-01 -6.56400546e-02 5.09896278e-01 -1.15047240e+00
3.57616514e-01 1.66026676e+00 8.80373120e-01 1.05796325e+00
-6.57842398e-01 -1.13899970e+00 2.84060001e-01 4.76077497e-01
-1.33010185e+00 -7.13843465e-01 8.95332038e-01 -1.66482851e-01
5.02094328e-01 -1.39103472e-01 4.69917327e-01 1.30507815e+00
2.58146793e-01 6.01591825e-01 1.05174243e+00 -3.36360991e-01
4.71349172e-02 -4.76069599e-01 2.29160488e-01 1.15241659e+00
2.99938262e-01 5.67197539e-02 -7.38514185e-01 -2.43827458e-02
6.44522667e-01 1.91067487e-01 -3.24955672e-01 -1.47873491e-01
-4.30544555e-01 4.82898146e-01 3.59456480e-01 4.18524027e-01
-2.11153314e-01 1.37696087e-01 3.40192795e-01 4.99155819e-01
8.43186796e-01 -1.91049606e-01 -5.77789307e-01 1.30140468e-01
-8.30371559e-01 2.52901614e-01 4.42537308e-01 5.31706154e-01
4.89051819e-01 4.03609037e-01 -1.43604651e-01 1.08255839e+00
2.95759529e-01 6.19361460e-01 4.96370167e-01 -6.12222135e-01
-8.26447085e-02 6.35975838e-01 -2.99630463e-01 -9.88310337e-01
-3.07518035e-01 -4.45609212e-01 -9.70237374e-01 3.65934849e-01
5.43725431e-01 1.78553477e-01 -1.14554441e+00 1.94729936e+00
2.27408186e-01 2.89188921e-01 -3.04993927e-01 3.38259399e-01
9.47197735e-01 9.23442468e-02 3.12780738e-01 -3.51807505e-01
1.45515072e+00 -6.51165962e-01 -8.46399248e-01 -6.90169483e-02
1.57741264e-01 -5.31217337e-01 7.46014833e-01 3.57721925e-01
-7.94076622e-01 -9.40435350e-01 -1.05173254e+00 3.02491896e-03
-3.20927799e-01 4.70791876e-01 8.76573026e-01 1.02832329e+00
-1.22352910e+00 5.76517642e-01 -6.43641710e-01 -5.27317405e-01
1.11608016e+00 6.62206113e-01 -8.40701103e-01 -4.82606471e-01
-1.15623438e+00 6.45965278e-01 -1.30215868e-01 4.07930076e-01
-9.64176297e-01 -9.84055758e-01 -5.80864251e-01 -1.40224621e-01
2.39855900e-01 -4.57425267e-01 1.09499383e+00 -1.17680371e+00
-1.33916473e+00 1.15571249e+00 -3.09709072e-01 -1.86955199e-01
4.91448015e-01 -5.72484195e-01 -9.54094350e-01 -4.09539528e-02
-2.87652791e-01 5.39971530e-01 1.18645763e+00 -1.00660050e+00
-2.75351226e-01 -1.08753860e+00 -1.01107173e-01 -4.70370561e-01
-1.02016473e+00 2.81975657e-01 -3.68648797e-01 -6.44941986e-01
-2.16647565e-01 -9.03283477e-01 3.82308811e-01 7.30012894e-01
7.28609040e-02 -7.77276337e-01 1.21214283e+00 -9.15523410e-01
1.20369840e+00 -2.24932933e+00 -2.66678296e-02 -3.55785191e-01
2.80138701e-01 5.77700853e-01 -4.08542216e-01 -3.55449840e-02
-5.36913753e-01 -9.59637109e-03 -3.40718478e-02 -5.19377172e-01
-2.55871892e-01 8.51048306e-02 -1.85732201e-01 5.97091556e-01
3.20665181e-01 7.99425542e-01 -5.64864933e-01 -2.79370546e-01
-2.70436823e-01 8.39317918e-01 -3.40017200e-01 3.13007891e-01
4.10191491e-02 2.43064359e-01 -3.82792681e-01 1.14588666e+00
1.04227233e+00 3.22916120e-01 -8.73144269e-02 -4.62986171e-01
9.04914662e-02 -4.92550045e-01 -6.31353438e-01 1.59201229e+00
-1.18994549e-01 3.03091556e-01 -6.12719283e-02 -1.01387572e+00
9.69982028e-01 3.25863689e-01 5.67148268e-01 -7.51611650e-01
4.39481199e-01 2.48822033e-01 7.35109225e-02 -4.24888283e-01
2.06022605e-01 -4.61904611e-03 4.81979132e-01 1.34542435e-01
4.05920208e-01 5.39587379e-01 1.27292439e-01 -2.15508193e-01
8.87360096e-01 3.00622154e-02 -3.13781248e-03 -3.82434189e-01
7.45326519e-01 -1.02042043e+00 7.53539979e-01 2.55290180e-01
-8.50372016e-01 3.89794797e-01 3.40272427e-01 -9.60269392e-01
-7.36267030e-01 -1.10273755e+00 -4.13181573e-01 9.70171928e-01
-5.05239844e-01 -1.98797837e-01 -9.05232549e-01 -1.13571739e+00
2.39575297e-01 6.42622188e-02 -1.16490614e+00 -3.91838044e-01
-6.21730208e-01 -9.32012260e-01 6.12785280e-01 6.52866364e-01
9.71519113e-01 -9.15729463e-01 8.24208036e-02 -6.44345805e-02
2.01261982e-01 -1.20540857e+00 -5.70205748e-01 -7.12760091e-01
-8.78469348e-01 -1.34927607e+00 -8.60824168e-01 -8.22429180e-01
8.58117044e-01 1.92941815e-01 1.03688037e+00 4.49098438e-01
-5.92906058e-01 4.93712991e-01 -2.18349800e-01 -4.98815805e-01
-7.71938935e-02 8.61257464e-02 5.11731029e-01 4.33755279e-01
4.43310767e-01 -5.85359991e-01 -8.44762504e-01 3.28687131e-01
-7.79958665e-01 -5.84813535e-01 3.23066592e-01 6.92982674e-01
1.63973182e-01 1.49585381e-01 1.07396662e+00 -6.82571352e-01
4.31618184e-01 -1.93998888e-01 -4.09660339e-01 4.50823367e-01
-6.90723360e-01 -2.61897027e-01 3.50631386e-01 -4.07761812e-01
-1.04765391e+00 -1.11465201e-01 -3.74499977e-01 -3.03422213e-01
-1.19050153e-01 1.03612684e-01 -5.56404054e-01 -4.23023015e-01
4.43332136e-01 -2.32799947e-02 3.32221597e-01 -5.36660492e-01
1.47400215e-01 7.04585075e-01 6.93001449e-01 -6.95971251e-01
9.00108933e-01 5.49769223e-01 1.47248700e-01 -9.16538239e-01
-9.13708806e-01 1.29009321e-01 -5.72668970e-01 -4.27460194e-01
6.52787030e-01 -1.03257322e+00 -8.77583623e-01 1.35174906e+00
-8.43501091e-01 -1.89105570e-02 9.47545469e-02 -1.00659393e-01
-8.80100653e-02 6.02368951e-01 -6.55402958e-01 -8.28726590e-01
-7.46956885e-01 -1.01782167e+00 9.23745036e-01 4.03218389e-01
2.53513902e-01 -7.35963643e-01 -2.38023773e-01 4.58356202e-01
6.60818160e-01 3.17464918e-01 8.39065135e-01 -1.52959689e-01
-2.35196665e-01 -4.00508463e-01 -2.82763779e-01 6.94380820e-01
5.64153373e-01 4.26470608e-01 -1.26454318e+00 -6.26103759e-01
-8.84697810e-02 -4.46257353e-01 1.00512600e+00 1.43511280e-01
1.77682829e+00 6.00103848e-02 -7.40480050e-02 4.86316830e-01
1.18058789e+00 1.75521389e-01 1.03485513e+00 -5.57216220e-02
4.80188102e-01 4.86768782e-01 3.24549288e-01 5.00689745e-01
3.95995528e-01 6.22479022e-01 4.87539649e-01 -9.14226323e-02
-4.59937781e-01 -2.01587547e-02 3.55165690e-01 5.93012691e-01
-4.77424234e-01 -3.13991643e-02 -7.08772838e-01 2.77077436e-01
-1.44965971e+00 -9.08857226e-01 2.48910531e-01 2.00260139e+00
6.15904093e-01 -2.45111197e-01 -1.22372126e-02 2.64437646e-01
6.17588878e-01 3.02826375e-01 -4.90882039e-01 -6.60310686e-02
-1.90972790e-01 6.52213216e-01 -2.22530048e-02 -6.89039081e-02
-1.17213261e+00 7.97617316e-01 6.17744875e+00 7.09253252e-01
-1.39228690e+00 4.65291440e-02 1.06441271e+00 1.20929398e-01
3.15869838e-01 -6.86312079e-01 -9.89118755e-01 4.90635067e-01
7.53016710e-01 1.87579840e-02 4.42664236e-01 1.02940357e+00
-1.31195232e-01 2.70420611e-01 -1.06220734e+00 1.13865769e+00
5.48301756e-01 -8.26499403e-01 6.88623786e-02 1.10878423e-01
6.90080583e-01 -3.74407023e-01 6.17782176e-01 4.35576379e-01
-2.82102883e-01 -1.24035347e+00 3.51728678e-01 5.55131495e-01
1.19023967e+00 -8.15256357e-01 6.41402781e-01 -4.55999970e-01
-1.21460021e+00 -3.94160569e-01 -4.32796538e-01 5.11995740e-02
-4.23433483e-01 7.18134880e-01 -2.88360655e-01 4.94696409e-01
9.21663284e-01 7.77027249e-01 -9.42713320e-01 8.51001203e-01
-5.02906106e-02 5.16247153e-01 1.60334632e-01 5.83276868e-01
-4.00018662e-01 1.66555062e-01 -6.27673715e-02 5.90508521e-01
5.93727887e-01 -5.71780764e-02 -1.70391187e-01 4.07662749e-01
-7.99704015e-01 1.13544472e-01 -3.96938711e-01 -3.14319044e-01
2.27684885e-01 1.36441123e+00 -1.82565093e-01 -5.66758495e-03
-5.90322673e-01 8.47623169e-01 4.72424656e-01 2.45813996e-01
-6.11651897e-01 -4.28900197e-02 1.05212700e+00 2.46615842e-01
1.86683536e-01 -3.01832438e-01 1.69701979e-01 -1.18325090e+00
1.57491967e-01 -1.19246209e+00 3.43023956e-01 -3.97752136e-01
-1.63682783e+00 7.88867116e-01 -3.84601086e-01 -6.70830309e-01
2.86269069e-01 -9.28404808e-01 -5.52525222e-01 6.64094210e-01
-1.71072888e+00 -1.52851188e+00 -6.64325237e-01 7.42881894e-01
4.81850982e-01 -7.01860368e-01 9.99777913e-01 7.29620337e-01
-8.55745494e-01 1.20427668e+00 -1.88350692e-01 3.41584057e-01
9.78605032e-01 -6.86546981e-01 4.10086721e-01 8.39923859e-01
-4.03680980e-01 6.98197603e-01 3.72356713e-01 -5.50833523e-01
-1.63493276e+00 -1.14430869e+00 6.89248741e-01 -3.33285302e-01
5.14525890e-01 -2.86261261e-01 -1.04129601e+00 4.50962484e-01
1.23054683e-01 5.26281953e-01 7.38953829e-01 2.94092417e-01
-9.56064582e-01 -8.46607804e-01 -1.39296830e+00 4.29304987e-01
1.39007914e+00 -4.88527954e-01 -2.26665795e-01 1.08566411e-01
4.25414443e-01 -2.74445694e-02 -1.06347716e+00 9.50201213e-01
9.99482930e-01 -8.91628563e-01 1.00479817e+00 -6.13004267e-01
4.13865894e-01 -1.01884730e-01 -8.62449557e-02 -1.11923110e+00
-1.84810653e-01 -3.49959135e-01 -3.92208070e-01 1.51005340e+00
-2.61606663e-01 -6.61958933e-01 8.25374186e-01 5.23321927e-01
-5.32054529e-03 -9.37661529e-01 -1.18194330e+00 -7.71655917e-01
1.72723100e-01 -1.61230221e-01 8.39068890e-01 8.57851207e-01
-7.72518456e-01 -1.75534308e-01 -5.55631280e-01 1.51566297e-01
8.32892478e-01 -3.11418712e-01 6.61347330e-01 -1.71701014e+00
1.61557719e-01 -3.17181706e-01 -6.98084831e-01 -3.13663304e-01
6.55401587e-01 -4.76144791e-01 -2.80189544e-01 -9.82230842e-01
2.00202376e-01 -1.77473024e-01 -6.15850449e-01 6.65098548e-01
-2.91452497e-01 5.72167993e-01 2.13993881e-02 -2.65624374e-01
-3.80023509e-01 9.32809949e-01 1.27544188e+00 -1.97333768e-01
3.49851668e-01 -1.40090048e-01 -8.35639894e-01 4.92019206e-01
8.02044213e-01 -4.32183817e-02 -3.19545776e-01 -4.33198512e-01
-2.06605941e-01 -4.31398869e-01 1.77064642e-01 -1.21646261e+00
-9.31650996e-02 1.31713226e-01 7.92553008e-01 -1.44521639e-01
4.08931792e-01 -6.22997165e-01 -3.36586572e-02 5.35971105e-01
1.09008171e-01 1.77096575e-01 3.57267052e-01 3.39158982e-01
-1.26343966e-01 2.70495385e-01 1.06920600e+00 2.04082355e-01
-6.31093204e-01 1.23184347e+00 9.07617733e-02 -1.37583002e-01
8.76009941e-01 -1.00872211e-01 -5.29961169e-01 -1.44837201e-01
-3.63742054e-01 2.39870069e-03 3.87782753e-01 9.16418195e-01
7.17465162e-01 -1.51262081e+00 -7.82410443e-01 4.94018346e-01
2.67409414e-01 -6.09641850e-01 5.67852557e-01 5.22626400e-01
-1.76141933e-01 1.32848814e-01 -6.37801349e-01 -1.27489358e-01
-1.59861040e+00 6.47551715e-01 5.76373816e-01 -2.02260371e-02
-2.12757289e-01 9.49086905e-01 2.53377318e-01 -1.40063465e-01
5.04330516e-01 1.51984215e-01 -1.84571907e-01 6.59626275e-02
9.87591028e-01 4.27338392e-01 3.48736554e-01 -7.11894751e-01
-3.24068993e-01 6.35015011e-01 -4.74865526e-01 7.12267339e-01
1.48739004e+00 1.31200641e-01 -4.10817653e-01 -5.65748662e-02
1.10070956e+00 -3.08365136e-01 -1.21435654e+00 -1.61580667e-01
-2.81557411e-01 -6.10708296e-01 -2.89664604e-03 -7.03369796e-01
-1.78092504e+00 8.99508119e-01 1.31252015e+00 -3.05718094e-01
1.36925948e+00 -3.70209247e-01 9.19341266e-01 2.39973053e-01
5.80641448e-01 -1.00845587e+00 4.33240473e-01 2.83572346e-01
1.12511456e+00 -1.28566146e+00 4.98332083e-03 -4.77719218e-01
2.05563549e-02 1.28212035e+00 1.04285741e+00 5.82217388e-02
1.10524619e+00 -1.08080700e-01 1.11804113e-01 3.28512341e-02
-4.00439680e-01 1.99161731e-02 3.34524125e-01 6.18193686e-01
6.93078697e-01 -1.55001611e-01 -4.23101604e-01 5.17308354e-01
3.29560861e-02 5.15397787e-01 4.47048880e-02 7.46765196e-01
-2.57624179e-01 -1.44058979e+00 -1.20776534e-01 4.69936818e-01
-9.34434474e-01 2.36976877e-01 -2.60368198e-01 6.56438887e-01
3.26109231e-01 8.04460764e-01 -1.74916372e-01 -2.99877912e-01
3.57528418e-01 3.05662215e-01 8.63324881e-01 -1.43136144e-01
-3.57983768e-01 -5.39811850e-01 -1.83600262e-01 -5.57591081e-01
-4.63500023e-01 -5.65252304e-01 -7.63064146e-01 -5.64043641e-01
9.15788263e-02 -3.31378430e-01 5.73422611e-01 8.17829251e-01
4.42980140e-01 4.90405768e-01 7.41632700e-01 -6.67020977e-01
-6.33493245e-01 -9.86108840e-01 -7.14733005e-01 6.14238083e-01
2.47168452e-01 -1.05112970e+00 -1.37239367e-01 -1.11801364e-01] | [13.339977264404297, 0.6981818675994873] |
32842ff6-b38c-463f-ba9e-495a89e9c8ac | pl-unext-per-stage-edge-detail-and-line | 2303.04413 | null | https://arxiv.org/abs/2303.04413v1 | https://arxiv.org/pdf/2303.04413v1.pdf | PL-UNeXt: Per-stage Edge Detail and Line Feature Guided Segmentation for Power Line Detection | Power line detection is a critical inspection task for electricity companies and is also useful in avoiding drone obstacles. Accurately separating power lines from the surrounding area in the aerial image is still challenging due to the intricate background and low pixel ratio. In order to properly capture the guidance of the spatial edge detail prior and line features, we offer PL-UNeXt, a power line segmentation model with a booster training strategy. We design edge detail heads computing the loss in edge space to guide the lower-level detail learning and line feature heads generating auxiliary segmentation masks to supervise higher-level line feature learning. Benefited from this design, our model can reach 70.6 F1 score (+1.9%) on TTPLA and 68.41 mIoU (+5.2%) on VITL (without utilizing IR images), while preserving a real-time performance due to few inference parameters. | ['Daming Liu', 'Zhen Chen', 'Yang Cheng'] | 2023-03-08 | null | null | null | null | ['line-detection'] | ['computer-vision'] | [ 3.06988001e-01 -2.28232279e-01 -3.89371701e-02 -1.96972966e-01
-6.79378629e-01 -9.43645120e-01 1.69417337e-01 -7.54429847e-02
-1.23439118e-01 5.80107450e-01 -7.54158020e-01 -5.31485498e-01
-9.71651375e-02 -1.00683451e+00 -5.34603596e-01 -8.38882923e-01
-3.58645052e-01 -4.97124232e-02 3.52721661e-01 -1.35183483e-01
1.93195134e-01 8.54300320e-01 -1.01080775e+00 6.32635951e-02
1.47125638e+00 1.18237531e+00 1.99082360e-01 9.71822143e-01
2.58417904e-01 5.45623720e-01 -6.93663299e-01 -1.75387293e-01
4.27063733e-01 -4.00278658e-01 -3.86567086e-01 4.87442523e-01
6.02574527e-01 -6.82026088e-01 -4.77194399e-01 1.34169197e+00
6.20902538e-01 1.14607066e-01 5.96007407e-01 -1.37184024e+00
-2.64581740e-01 4.10452962e-01 -1.31011295e+00 4.05656934e-01
1.95844695e-01 3.41007680e-01 9.64138687e-01 -4.35405165e-01
-4.61774841e-02 4.36699569e-01 8.85544181e-01 -4.14090306e-02
-1.07530236e+00 -5.60147226e-01 1.26781791e-01 3.49323571e-01
-1.58484304e+00 -9.22841579e-02 7.96278894e-01 -1.87934488e-01
7.63693929e-01 3.04942787e-01 8.17288995e-01 3.95042390e-01
1.75100416e-01 1.11611044e+00 7.53926337e-01 -3.85461897e-01
-5.53238690e-02 1.32309034e-01 -7.33499154e-02 1.16978490e+00
8.70317370e-02 -1.92388780e-02 -3.09084862e-01 4.03398395e-01
9.05483484e-01 -3.97454798e-01 -6.63197637e-01 -9.09527689e-02
-5.04903615e-01 5.17377973e-01 7.60956466e-01 1.09807760e-01
-1.82130277e-01 -6.85021281e-02 -1.03058778e-02 1.07325176e-02
2.40000352e-01 2.85853535e-01 -5.24170578e-01 -1.05878688e-01
-1.40677822e+00 -1.81769252e-01 6.42757773e-01 1.30931604e+00
7.02510834e-01 4.02622849e-01 -3.87653112e-01 6.18904591e-01
1.77229717e-01 8.00624490e-01 -2.49334335e-01 -6.61443889e-01
6.27899945e-01 6.05189621e-01 -1.32686973e-01 -9.65866446e-01
-7.09533215e-01 -8.60043585e-01 -9.52387929e-01 3.71006012e-01
3.23646843e-01 -6.08192444e-01 -8.53168011e-01 9.13466990e-01
1.35571331e-01 1.91244811e-01 -1.80827901e-01 5.96546769e-01
5.29661000e-01 1.03082228e+00 -2.23866984e-01 -1.40363742e-02
1.60457277e+00 -1.09439564e+00 -5.84918976e-01 -5.76006830e-01
8.63822460e-01 -4.95954692e-01 8.22988749e-01 7.81146407e-01
-1.00514758e+00 -5.82584023e-01 -1.50518727e+00 -2.15401929e-02
-2.20617890e-01 8.74230921e-01 7.54031479e-01 8.47082913e-01
-9.23003495e-01 4.74358767e-01 -7.06706405e-01 -9.77103710e-02
8.51112366e-01 3.90733302e-01 -2.42193881e-02 -9.76500586e-02
-8.86168778e-01 5.93755782e-01 4.48259920e-01 5.21456480e-01
-6.57841206e-01 -9.29538727e-01 -9.58708882e-01 3.83334696e-01
6.83542430e-01 -1.56416148e-01 8.84174585e-01 -3.10889214e-01
-1.42106831e+00 4.36373264e-01 1.91871449e-01 -4.61560786e-01
7.84345031e-01 -3.22480083e-01 -2.95312434e-01 5.90884566e-01
-1.02541737e-01 7.69468904e-01 9.21397984e-01 -1.00203121e+00
-1.39112568e+00 -1.55660197e-01 -9.24807191e-02 3.32097143e-01
-2.71810502e-01 -4.52196836e-01 -6.81186736e-01 -6.03123724e-01
-9.08788517e-02 -6.93527997e-01 -3.60438041e-02 2.42411986e-01
-4.76902753e-01 -1.21012554e-01 1.33071935e+00 -9.70300436e-01
1.18124890e+00 -2.04320717e+00 -5.48333526e-01 2.76061058e-01
-3.27002220e-02 4.06693846e-01 -7.15312660e-02 -9.39130485e-02
1.06597051e-01 -3.83086987e-02 -4.66877908e-01 1.19094580e-01
-5.66168986e-02 -3.68079692e-01 -7.79186636e-02 5.92346191e-01
4.82869357e-01 9.81954813e-01 -8.21156025e-01 -3.39202911e-01
6.59934521e-01 4.87362891e-01 -3.90545636e-01 1.92755740e-02
1.39060840e-01 1.47122473e-01 -3.58221233e-01 7.20023096e-01
1.08122635e+00 3.44899669e-02 -1.34393498e-01 -6.03226483e-01
-2.40347579e-01 -3.35587800e-01 -1.08384562e+00 1.41279268e+00
-3.87243509e-01 1.02272892e+00 2.02878311e-01 -6.56233430e-01
9.78623092e-01 1.09125845e-01 4.83601689e-01 -7.07824349e-01
3.06427181e-01 -2.16226935e-01 -7.58350417e-02 -3.29038739e-01
4.19569343e-01 4.98938560e-01 -5.41267619e-02 -1.00095116e-03
-4.89731729e-02 -7.71308899e-01 3.82399529e-01 4.97904606e-02
9.04496133e-01 3.59541774e-01 3.68621312e-02 -4.25058931e-01
4.72307444e-01 1.67227220e-02 6.06479287e-01 6.86244369e-01
-1.66799754e-01 4.86518323e-01 4.90768015e-01 -2.64018923e-01
-6.45833910e-01 -8.83787572e-01 -4.36390370e-01 6.13664746e-01
3.57958615e-01 -1.24199823e-01 -9.95663285e-01 -8.49510372e-01
-2.68832237e-01 9.51542079e-01 1.63742546e-02 -2.89174020e-01
-2.43686408e-01 -1.00197339e+00 7.52569437e-01 7.21132040e-01
1.15526199e+00 -6.14535987e-01 -6.40461028e-01 1.20855279e-01
2.52652112e-02 -1.46031487e+00 -5.77882469e-01 4.64622825e-01
-5.20890594e-01 -9.40794051e-01 -5.80710471e-01 -9.18427169e-01
9.33881521e-01 2.99502939e-01 9.75693405e-01 2.40805253e-01
-8.00283015e-01 1.96930743e-03 -2.22357377e-01 -2.75113195e-01
2.16468930e-01 2.44655728e-01 -4.35978264e-01 -1.73873678e-01
2.36952558e-01 -7.53608346e-02 -4.29967880e-01 2.88203120e-01
-6.79061472e-01 1.59333855e-01 6.68414593e-01 8.59676182e-01
1.79819271e-01 1.03375149e+00 1.60012633e-01 -5.57248950e-01
1.79774910e-01 1.05125219e-01 -1.50137365e+00 2.97250301e-01
-4.20294225e-01 -4.89540935e-01 7.31154740e-01 2.93569956e-02
-1.04001343e+00 1.60752371e-01 -1.93845838e-01 1.00120895e-01
-2.75966942e-01 2.15842843e-01 -5.13188958e-01 -4.64140445e-01
2.64635265e-01 1.21301860e-01 -5.77554643e-01 7.66864195e-02
1.15072615e-01 6.95495665e-01 6.90389395e-01 -2.84132987e-01
1.28136778e+00 3.58858556e-01 2.01088265e-01 -1.38256168e+00
-1.04578686e+00 -2.91473240e-01 -9.20782328e-01 -2.81756461e-01
8.35905492e-01 -8.22670817e-01 -9.56477225e-01 6.61344528e-01
-9.13812995e-01 -6.93688512e-01 -1.63697720e-01 3.03599715e-01
-1.71132907e-01 6.71606481e-01 -6.17580831e-01 -8.36474538e-01
-4.02816325e-01 -1.16956151e+00 9.07347381e-01 7.60598123e-01
1.94654047e-01 -9.57429945e-01 -7.99089372e-01 3.94494087e-01
6.88781142e-02 1.60172775e-01 9.10011649e-01 -2.18329594e-01
-7.72016943e-01 -4.16462511e-01 -6.32105231e-01 6.47165656e-01
1.70256495e-01 3.60239148e-01 -1.16470635e+00 -4.46083575e-01
-2.25965023e-01 -1.26878858e-01 6.44504607e-01 5.45948446e-01
1.23283303e+00 8.53468254e-02 -3.66099566e-01 9.63670373e-01
1.48765957e+00 6.02548420e-01 6.96753681e-01 1.49684653e-01
7.97840953e-01 2.54600137e-01 7.04858363e-01 6.67826831e-01
1.37895420e-01 3.55411530e-01 2.64419496e-01 -8.23023081e-01
1.48171544e-01 -1.78275421e-01 2.34455273e-01 2.64468968e-01
4.33467984e-01 -7.39103258e-01 -6.86309159e-01 5.21304488e-01
-1.37273026e+00 -8.34918916e-01 -3.58632863e-01 1.97227943e+00
6.42291009e-01 5.42880893e-01 -1.64568931e-01 3.52905512e-01
6.22992039e-01 7.61056542e-02 -5.35966873e-01 -4.28904183e-02
-2.78486073e-01 -3.20972763e-02 9.87812400e-01 6.04730844e-01
-1.46501446e+00 9.76098120e-01 5.70155001e+00 1.20957851e+00
-9.20488358e-01 -7.63624787e-01 7.93944001e-01 4.67616707e-01
6.02838695e-02 -2.11296096e-01 -9.97147381e-01 3.90536845e-01
2.70036161e-01 1.82765603e-01 4.27564204e-01 5.79354405e-01
2.45658413e-01 -5.78341544e-01 -9.59300816e-01 1.04258835e+00
1.43248543e-01 -9.76114869e-01 -2.30542406e-01 -7.80173987e-02
9.45120931e-01 -2.09757596e-01 1.09560549e-01 1.25980735e-01
1.57143563e-01 -1.01317203e+00 4.29907590e-01 1.88919857e-01
8.02602828e-01 -1.06609905e+00 7.39917278e-01 3.82503271e-01
-1.45726943e+00 -1.23730500e-03 -3.06156248e-01 2.42168948e-01
2.35698596e-01 6.71451330e-01 -9.58201349e-01 8.71975422e-01
7.15221822e-01 7.51485348e-01 -5.00121057e-01 1.29284751e+00
-8.43607605e-01 8.11371982e-01 -6.53257191e-01 2.57222235e-01
5.21512926e-01 -6.31812036e-01 2.67618060e-01 1.25087452e+00
3.08896273e-01 -5.04484437e-02 4.52870369e-01 8.34684789e-01
-9.93369371e-02 -4.14012343e-01 -4.03807312e-01 1.78591236e-01
4.95205879e-01 1.77738178e+00 -1.08301330e+00 6.95381910e-02
-5.51665783e-01 1.27506971e+00 -2.37828836e-01 4.26037580e-01
-1.20248187e+00 -1.11706412e+00 2.96129167e-01 -1.61187500e-01
5.35366952e-01 -2.51229316e-01 -2.96591043e-01 -7.68022776e-01
-4.96843942e-02 -4.86117929e-01 2.58727551e-01 -9.31707144e-01
-1.06764722e+00 5.98361969e-01 -1.84508950e-01 -1.18256593e+00
2.70555794e-01 -8.60026002e-01 -7.53786564e-01 8.40149283e-01
-2.08275127e+00 -1.22498250e+00 -7.17450321e-01 1.67515129e-01
8.14442515e-01 -7.87204504e-03 2.68029958e-01 2.80853599e-01
-9.01594818e-01 6.77799165e-01 3.08976285e-02 6.10253394e-01
3.20651352e-01 -1.47409999e+00 3.79131436e-01 1.21395957e+00
1.59771964e-01 -1.41433045e-01 4.23716098e-01 -4.15385634e-01
-1.26822221e+00 -1.14584637e+00 3.15057933e-01 -2.20062673e-01
5.72836578e-01 -2.93028086e-01 -8.05263460e-01 5.27661383e-01
3.72871488e-01 2.08306182e-02 4.86425966e-01 -4.68194544e-01
3.87822449e-01 -3.12216729e-01 -1.35685217e+00 7.03970194e-01
7.59168446e-01 -3.37089658e-01 -1.82773232e-01 1.72593504e-01
1.14763819e-01 -4.42455709e-01 -9.01795089e-01 5.53791821e-01
1.34878635e-01 -8.54649603e-01 7.97330081e-01 2.77481467e-01
-1.03785060e-02 -7.76076376e-01 3.88691038e-01 -1.34771454e+00
-4.83119309e-01 -8.98975313e-01 2.50995249e-01 1.46584547e+00
4.61806536e-01 -3.96320373e-01 1.09812665e+00 3.86212051e-01
-2.74435073e-01 -5.39544463e-01 -5.79872072e-01 -5.54595709e-01
-4.35341924e-01 -4.34506118e-01 4.83871341e-01 7.21354961e-01
-3.07600677e-01 3.34884733e-01 -1.51485831e-01 9.21497107e-01
7.93171942e-01 4.44764160e-02 4.81575429e-01 -1.14273226e+00
4.15402800e-02 -3.83179903e-01 -3.54799092e-01 -1.66523969e+00
5.92826456e-02 -5.62370658e-01 2.54941076e-01 -1.77370584e+00
-3.16769809e-01 -3.95222545e-01 9.42615047e-02 6.16425693e-01
-2.32148737e-01 5.40464044e-01 1.94412246e-01 -6.14676178e-01
-3.96895021e-01 5.96515179e-01 1.27914178e+00 -5.36860943e-01
3.34150307e-02 1.12664863e-01 -4.18616503e-01 9.78223860e-01
9.36250269e-01 -2.33142704e-01 -6.34694159e-01 -6.23404980e-01
4.72675934e-02 -1.96956992e-01 3.31266701e-01 -1.20970786e+00
4.94790614e-01 8.16496834e-02 9.47902501e-01 -1.09835780e+00
1.48472250e-01 -1.20638049e+00 -5.50518751e-01 4.65307355e-01
2.07859457e-01 -1.50428832e-01 6.14436507e-01 1.80721924e-01
2.92321127e-02 -5.38253248e-01 8.46484005e-01 2.49039933e-01
-1.15843916e+00 3.53150576e-01 -6.02988958e-01 3.48418579e-02
1.26928258e+00 -7.07264781e-01 -2.99599975e-01 -2.15000942e-01
-2.36206174e-01 8.65636885e-01 2.78927058e-01 6.77825883e-02
5.33910871e-01 -8.40549111e-01 -6.97547078e-01 7.34291732e-01
-1.75880164e-01 3.93386424e-01 4.53425407e-01 7.83040822e-01
-1.08579624e+00 2.83114880e-01 -1.98386103e-01 -6.55035436e-01
-1.01887000e+00 4.89299409e-02 5.47701776e-01 -1.57150075e-01
-7.94222295e-01 9.80276704e-01 7.65242651e-02 -3.13800089e-02
2.94541895e-01 -4.99997526e-01 -1.73534781e-01 3.64443436e-02
5.08771956e-01 6.81350768e-01 2.88482487e-01 -7.94430375e-02
-2.08001390e-01 9.04380322e-01 9.40430723e-03 3.37773860e-01
1.11563039e+00 -1.28212467e-01 1.35525137e-01 -1.37448832e-01
8.76371384e-01 1.34655103e-01 -1.90203607e+00 2.56352216e-01
3.90647165e-02 -4.44303811e-01 7.15026915e-01 -1.02737224e+00
-1.43292797e+00 1.05591238e+00 6.34395719e-01 3.65238905e-01
1.47142231e+00 -5.75551689e-01 8.49079728e-01 4.64449495e-01
3.12255293e-01 -1.34152937e+00 -2.68692911e-01 3.07357460e-01
3.54199290e-01 -1.09862065e+00 2.56003797e-01 -9.20099258e-01
-6.52998567e-01 1.28053367e+00 7.47214198e-01 9.71237570e-03
1.76145494e-01 9.32167232e-01 6.19719997e-02 -3.08332853e-02
-2.93237895e-01 -2.11943507e-01 3.80460501e-01 8.59619141e-01
1.50281042e-01 -3.35558295e-01 5.54052413e-01 2.71701843e-01
7.44396374e-02 -2.59411275e-01 5.01331925e-01 8.65374863e-01
-4.96415108e-01 -5.53502381e-01 -2.52917945e-01 5.33307552e-01
-3.47840428e-01 -1.84832945e-01 7.09230155e-02 7.35768497e-01
-8.15808214e-03 1.09630406e+00 3.06721389e-01 -1.01393647e-01
5.51757991e-01 -6.49590850e-01 3.95687133e-01 -3.22412521e-01
-4.63781595e-01 1.74148977e-01 -7.21779168e-02 -1.51991561e-01
1.61715418e-01 -2.44902521e-01 -1.40672696e+00 -1.46920010e-01
-8.99425328e-01 2.67196149e-01 3.57841074e-01 7.04820037e-01
-4.06607948e-02 9.65537429e-01 8.69861424e-01 -8.87477398e-01
-3.80059272e-01 -6.03685200e-01 -1.01660383e+00 -3.59986275e-01
2.81820804e-01 -2.85643399e-01 -3.80248040e-01 1.48856595e-01] | [8.857515335083008, -0.9611779451370239] |
0c147ade-f3ca-4a3e-b0e0-47d2109e158b | asl-recognition-with-metric-learning-based-1 | 2004.05054 | null | https://arxiv.org/abs/2004.05054v1 | https://arxiv.org/pdf/2004.05054v1.pdf | ASL Recognition with Metric-Learning based Lightweight Network | In the past decades the set of human tasks that are solved by machines was extended dramatically. From simple image classification problems researchers now move towards solving more sophisticated and vital problems, like, autonomous driving and language translation. The case of language translation includes a challenging area of sign language translation that incorporates both image and language processing. We make a step in that direction by proposing a lightweight network for ASL gesture recognition with a performance sufficient for practical applications. The proposed solution demonstrates impressive robustness on MS-ASL dataset and in live mode for continuous sign gesture recognition scenario. Additionally, we describe how to combine action recognition model training with metric-learning to train the network on the database of limited size. The training code is available as part of Intel OpenVINO Training Extensions. | ['Evgeny Izutov'] | 2020-04-10 | asl-recognition-with-metric-learning-based | null | null | arxiv-org-2020-4 | ['sign-language-translation'] | ['computer-vision'] | [ 5.48333704e-01 -2.38974109e-01 -3.60884488e-01 -6.32857561e-01
-7.90459573e-01 -4.50571120e-01 7.62490392e-01 -7.35143304e-01
-9.20210183e-01 6.51577771e-01 1.37930840e-01 -5.31729162e-01
1.85660318e-01 -3.46286267e-01 -6.11025810e-01 -4.05628890e-01
2.63920844e-01 5.00640213e-01 4.05708164e-01 -3.78211379e-01
3.82718563e-01 8.15744758e-01 -1.75619888e+00 2.68021554e-01
4.29475754e-01 7.85831094e-01 3.74705996e-03 9.88282025e-01
-2.73433983e-01 8.83565724e-01 -3.88704807e-01 -2.09585607e-01
5.63218951e-01 -5.03622949e-01 -7.68114686e-01 -2.51908246e-02
8.64711225e-01 -5.86934388e-01 -5.00049353e-01 7.76430428e-01
8.96772981e-01 1.60210028e-01 2.82002628e-01 -1.66824698e+00
-5.34227550e-01 3.72402936e-01 -7.20512867e-02 -7.70205259e-02
3.80011767e-01 6.04204535e-01 5.99864244e-01 -9.02821004e-01
6.44166648e-01 1.11872482e+00 4.65438038e-01 9.89057243e-01
-5.64428031e-01 -8.11461389e-01 -8.26642439e-02 4.97926980e-01
-1.06540227e+00 -6.03379250e-01 5.25854528e-01 -2.04555482e-01
1.34128177e+00 1.91685557e-01 5.27902782e-01 1.19185698e+00
-3.70764852e-01 1.09080267e+00 1.44073987e+00 -6.04504347e-01
1.48921296e-01 -2.65452683e-01 1.94144249e-01 7.33984232e-01
6.84665442e-02 1.62650615e-01 -8.41662645e-01 2.72342771e-01
6.77164733e-01 -2.56595165e-01 5.54297194e-02 -2.79791266e-01
-1.43726039e+00 4.86120433e-01 1.20467894e-01 2.76216686e-01
-1.02941632e-01 4.70663279e-01 5.04343271e-01 7.45846748e-01
-1.96571857e-01 1.91794798e-01 -4.09563720e-01 -7.17741191e-01
-1.10812104e+00 3.14182222e-01 8.07473779e-01 1.08811629e+00
4.34656322e-01 1.50853440e-01 -1.39723569e-01 4.61468577e-01
2.68617123e-01 9.31674421e-01 6.39193773e-01 -1.01482737e+00
6.73651636e-01 5.33807755e-01 -3.19143245e-03 -2.86522388e-01
-4.26210225e-01 6.29211962e-02 -4.92182046e-01 6.82977736e-01
7.52047241e-01 -1.04372449e-01 -1.34755850e+00 1.53618598e+00
2.14077473e-01 7.78370872e-02 1.30294546e-01 1.14547575e+00
8.89009178e-01 2.51296341e-01 5.71413897e-02 3.83022606e-01
1.10087442e+00 -1.02103508e+00 -5.69506109e-01 -1.57645702e-01
8.47283125e-01 -7.74770439e-01 1.37823105e+00 3.43686789e-01
-9.37451959e-01 -4.68106955e-01 -1.11696541e+00 -2.92890251e-01
-4.71942455e-01 2.97874063e-01 8.09598625e-01 7.44387925e-01
-1.17933679e+00 2.57058531e-01 -1.08227098e+00 -1.05209672e+00
4.41020608e-01 6.48848355e-01 -4.76622641e-01 -2.62815896e-02
-8.85650337e-01 1.08826840e+00 2.63321787e-01 2.67616630e-01
-5.88701129e-01 -4.53516170e-02 -7.64588118e-01 -4.78394359e-01
2.70551622e-01 -5.46857059e-01 1.36137533e+00 -9.03998315e-01
-1.98706830e+00 1.05830801e+00 -8.98694992e-02 -6.04168236e-01
1.11398709e+00 -2.66861230e-01 -3.88103098e-01 -5.83721213e-02
-1.35120675e-01 9.25273180e-01 8.36647332e-01 -6.05121493e-01
-7.84874082e-01 -4.39832628e-01 -5.44446409e-02 2.37160102e-01
-8.64229798e-02 3.66133153e-01 -2.52806962e-01 -3.82078171e-01
2.09895596e-02 -1.14549065e+00 -1.55775547e-01 4.22501713e-01
-3.52204405e-02 -1.82758301e-01 1.03624082e+00 -5.58571875e-01
7.37068653e-01 -1.74150348e+00 9.40720458e-03 1.68280780e-01
-2.99605727e-01 6.81153238e-01 -4.49690580e-01 4.51063335e-01
2.64751554e-01 -2.70898938e-01 -3.48171830e-01 -2.54399776e-01
2.76726127e-01 4.44055825e-01 -4.50801492e-01 3.73730212e-01
2.54886687e-01 1.34640491e+00 -7.52002299e-01 -3.31103623e-01
4.17277634e-01 2.87958324e-01 -2.67029852e-01 1.74632475e-01
-1.34138435e-01 7.13174522e-01 -3.14651191e-01 8.74014199e-01
3.84924740e-01 2.66072787e-02 -2.01854885e-01 9.23636928e-02
-2.62982100e-01 1.67259097e-01 -1.19248700e+00 2.10054159e+00
-3.98348480e-01 9.27666008e-01 -5.22912154e-03 -9.39275086e-01
8.89265299e-01 1.92011327e-01 4.44146097e-01 -8.06479037e-01
2.03557491e-01 5.51041961e-01 1.27437413e-01 -8.67806554e-01
5.07410347e-01 1.15533568e-01 2.68949359e-03 6.83919191e-01
-7.29062036e-03 -2.83451289e-01 3.20828080e-01 -2.04131439e-01
1.22794986e+00 7.32543528e-01 9.16792825e-02 2.57305861e-01
6.17961228e-01 4.39166665e-01 3.13773230e-02 8.60344648e-01
-6.39890969e-01 4.05578852e-01 -2.27761164e-01 -5.17549515e-01
-1.02872181e+00 -1.06571054e+00 1.57196105e-01 1.18365133e+00
2.54201666e-02 8.85604173e-02 -6.41117811e-01 -6.44084513e-01
-4.30659316e-02 4.27339524e-01 -1.00601867e-01 1.15157336e-01
-1.09107339e+00 -3.26491207e-01 1.16142488e+00 8.73049736e-01
9.80904341e-01 -1.49933910e+00 -1.16081774e+00 3.70583534e-02
1.13119930e-01 -1.37126708e+00 -3.44281018e-01 3.24490070e-02
-8.10877621e-01 -9.51919556e-01 -9.04293478e-01 -1.04894233e+00
4.50723588e-01 4.59484607e-02 6.13138676e-01 4.47152369e-02
-6.37839794e-01 7.07378507e-01 -4.84310597e-01 -3.99975896e-01
-5.11909187e-01 2.01610446e-01 7.35525936e-02 -1.50870502e-01
8.70219469e-01 -5.16080558e-01 -3.84011984e-01 2.88482606e-01
-7.99411774e-01 1.76792175e-01 7.86376417e-01 8.68603468e-01
1.84073865e-01 -7.88547277e-01 1.97110623e-01 -2.34374795e-02
5.48894048e-01 2.31387571e-01 -6.78674519e-01 3.98976058e-01
-3.15440923e-01 3.42666954e-01 2.73744881e-01 -5.34247041e-01
-8.03646028e-01 5.76718211e-01 -3.51400316e-01 -8.19065198e-02
-5.73960662e-01 1.12173602e-01 -1.27670933e-02 -3.30426067e-01
5.80009878e-01 5.63529611e-01 4.55534995e-01 -3.42592239e-01
7.09922493e-01 1.10150027e+00 7.98839509e-01 -3.50600034e-01
9.13133681e-01 6.73241079e-01 1.62155792e-01 -9.14442778e-01
-1.54986739e-01 -4.61293757e-01 -1.04971278e+00 -3.89204860e-01
7.71162271e-01 -6.53441310e-01 -8.29217374e-01 7.04876065e-01
-1.14259434e+00 -7.34315574e-01 -3.66388768e-01 7.36485183e-01
-9.33329225e-01 2.84082741e-01 -2.14061484e-01 -6.06255591e-01
-3.05865020e-01 -1.07514942e+00 1.23421407e+00 2.26648360e-01
-3.59307855e-01 -6.44064605e-01 8.87665972e-02 5.22285581e-01
6.60985351e-01 1.51268020e-01 3.61882508e-01 -4.71207172e-01
-8.78781199e-01 -2.36255571e-01 -3.13583732e-01 3.96195114e-01
1.12193912e-01 -2.63741940e-01 -8.18232417e-01 -2.38661557e-01
-4.76078480e-01 -6.60667539e-01 8.89346540e-01 2.78322231e-02
5.73092461e-01 9.19667557e-02 -1.44149825e-01 7.47035503e-01
1.08691192e+00 1.76378578e-01 5.56390047e-01 4.93427634e-01
7.82196820e-01 5.05913258e-01 6.20191276e-01 8.44359398e-02
3.61832023e-01 7.97812641e-01 -3.96943092e-02 -1.80054724e-01
-6.65773988e-01 -3.76299024e-01 5.51676869e-01 7.64700830e-01
-4.25921500e-01 3.22391577e-02 -1.02706170e+00 4.19895828e-01
-2.08875823e+00 -1.01586807e+00 -5.73107600e-02 1.92254436e+00
5.11191607e-01 -1.31415203e-01 2.00161457e-01 1.56348467e-01
3.60309154e-01 -4.73564006e-02 -6.68286920e-01 -5.97558558e-01
-2.65139133e-01 3.72774601e-01 9.17163193e-01 5.84381878e-01
-1.12368751e+00 1.44534874e+00 6.53961754e+00 4.62334633e-01
-1.51819837e+00 3.03404331e-01 -7.00216144e-02 -5.48734292e-02
2.34937936e-01 -1.03807539e-01 -8.43564928e-01 1.03705853e-01
7.76087403e-01 1.19224153e-01 5.26862860e-01 6.69516802e-01
4.72144574e-01 -6.71115667e-02 -1.22341561e+00 1.26254976e+00
2.89463878e-01 -1.09633625e+00 8.95780697e-02 -3.28738540e-02
4.40438509e-01 5.81288099e-01 4.17878293e-02 3.32484215e-01
2.63351589e-01 -1.14389694e+00 6.85950577e-01 5.00939786e-01
1.10296440e+00 -2.29198307e-01 4.95528221e-01 4.40595061e-01
-1.05551088e+00 -7.31448233e-02 -2.88945669e-03 -5.27439892e-01
4.07803804e-01 -4.20162737e-01 -9.27333891e-01 -1.35723837e-02
4.30448025e-01 7.13286638e-01 -5.18465757e-01 1.01588356e+00
-4.22999591e-01 6.08118832e-01 -4.31850195e-01 -3.80614966e-01
4.12343264e-01 -1.56320795e-01 4.95446801e-01 1.29514706e+00
5.36049366e-01 -9.38056484e-02 8.89513493e-02 5.86608112e-01
2.04641744e-01 2.19375804e-01 -8.74240458e-01 -5.75594716e-02
4.36305068e-02 7.64391661e-01 -6.43117070e-01 -4.07321393e-01
-4.49364096e-01 1.33055711e+00 6.44776598e-02 3.57912838e-01
-7.66280353e-01 -3.34119260e-01 7.61534810e-01 -1.06753558e-01
2.65358329e-01 -7.59465992e-01 -5.04699469e-01 -1.13401067e+00
4.40846622e-01 -9.42467213e-01 4.23640572e-02 -6.82916939e-01
-8.34518850e-01 5.17617762e-01 -3.90037708e-02 -1.50294065e+00
-5.43686688e-01 -1.10224640e+00 -3.40200394e-01 6.05880380e-01
-1.56509316e+00 -1.67004693e+00 -6.18033350e-01 7.32395172e-01
6.71054721e-01 -5.32900751e-01 7.93194950e-01 4.52871799e-01
-1.30771911e-02 7.72563815e-01 -8.06467235e-02 4.17940915e-01
7.97714353e-01 -8.72327745e-01 5.86020231e-01 1.00960374e+00
4.33022738e-01 2.22262561e-01 4.14512694e-01 -5.13494492e-01
-1.53631842e+00 -9.27861810e-01 1.16180873e+00 -6.80494905e-01
7.23247290e-01 -3.77523005e-01 -3.57610643e-01 6.63935483e-01
1.34464413e-01 4.21978235e-02 1.99613601e-01 -6.75767899e-01
-3.11877370e-01 6.91878330e-03 -1.01835334e+00 7.85296381e-01
1.60650361e+00 -6.14420116e-01 -5.47904611e-01 4.86669481e-01
2.71435767e-01 -6.68864489e-01 -3.93545479e-01 3.24326724e-01
8.86471629e-01 -5.33689499e-01 6.62235677e-01 -8.83667409e-01
1.35171667e-01 -5.64107597e-01 -3.32509339e-01 -8.13252747e-01
3.76747936e-01 -8.84879172e-01 9.74916369e-02 7.23553836e-01
3.25951397e-01 -6.71386898e-01 1.16115224e+00 6.14337862e-01
4.42000013e-03 -3.67098123e-01 -1.24699640e+00 -9.27338064e-01
-7.49036148e-02 -8.25408816e-01 5.54755867e-01 3.94923240e-01
-1.54836103e-01 -1.12466663e-01 -5.33152521e-01 -1.57280296e-01
6.44252121e-01 7.78990537e-02 1.29598975e+00 -8.12311649e-01
-2.09148481e-01 -4.97693449e-01 -9.69939470e-01 -1.37989247e+00
1.84580624e-01 -1.08199239e+00 1.94864914e-01 -1.53347111e+00
-2.98072517e-01 -2.21792743e-01 -5.54182790e-02 7.58843482e-01
4.06500787e-01 6.38796031e-01 5.47963560e-01 2.42313236e-01
-5.24212003e-01 1.86045960e-01 1.23332632e+00 -4.16565448e-01
-1.31223783e-01 1.00945733e-01 -7.52119347e-02 5.07449090e-01
7.75129795e-01 -2.21344337e-01 -2.26831660e-01 -5.91511905e-01
-2.73948222e-01 -2.87416518e-01 5.60330868e-01 -1.32715046e+00
5.43938875e-01 -1.62298545e-01 -7.60944635e-02 -5.39681494e-01
4.06479061e-01 -8.24564874e-01 -2.00809211e-01 6.36150002e-01
-5.61214030e-01 1.40017048e-01 2.20151711e-02 -1.85601022e-02
-2.92229503e-01 1.39723763e-01 5.38508177e-01 -4.38816193e-03
-1.38092780e+00 1.59790963e-01 -3.33686292e-01 -1.02449603e-01
1.07744455e+00 -6.53122723e-01 -1.82466775e-01 -5.38605928e-01
-8.52913499e-01 1.82752281e-01 3.49333644e-01 8.39470506e-01
7.76526630e-01 -1.36985207e+00 -7.84544289e-01 4.38818693e-01
2.59031922e-01 -3.83519024e-01 -3.16731095e-01 9.66618538e-01
-9.96292889e-01 7.54278421e-01 -5.93740165e-01 -6.46478951e-01
-1.66549170e+00 -4.17365506e-03 3.27922285e-01 1.09456867e-01
-7.74906576e-01 5.90541065e-01 -5.84686041e-01 -7.68092930e-01
6.53578043e-01 -6.02988183e-01 1.42805696e-01 -4.67186958e-01
6.23867452e-01 3.54340881e-01 2.75499448e-02 -9.55002010e-01
-5.15668273e-01 7.63134360e-01 2.80771226e-01 -4.62952942e-01
1.20819247e+00 5.34926876e-02 1.06950834e-01 3.67375851e-01
1.07224941e+00 -5.95234931e-01 -1.26753616e+00 -1.73748538e-01
2.90772915e-01 -2.29816332e-01 -2.09765896e-01 -1.09306955e+00
-5.99618316e-01 1.08058918e+00 1.19386494e+00 -7.22156465e-01
1.05406106e+00 -1.24780193e-01 9.29433465e-01 1.04409289e+00
7.79433310e-01 -1.39933002e+00 -1.51514053e-01 8.68231356e-01
9.03926909e-01 -1.57420790e+00 -3.65497828e-01 1.66883141e-01
-5.74942648e-01 1.25388622e+00 4.73394334e-01 -2.54064739e-01
4.49289799e-01 6.03924215e-01 6.89440250e-01 1.18718408e-01
-2.29435906e-01 -5.10092795e-01 2.74087250e-01 8.15363348e-01
3.26894879e-01 1.68798506e-01 -5.58967352e-01 -1.53573286e-02
-3.42369080e-01 6.75987959e-01 3.03026170e-01 1.26597238e+00
-3.59170258e-01 -1.41840863e+00 -4.38256592e-01 2.41575584e-01
9.81361866e-02 5.51401116e-02 -6.14436984e-01 8.51475716e-01
1.72485426e-01 7.60543406e-01 -1.48633327e-02 -3.94846916e-01
4.05898184e-01 3.55104566e-01 8.13027978e-01 -2.74402440e-01
-4.78499383e-01 -5.09995818e-01 5.93190640e-02 -7.72663057e-01
-7.82806158e-01 -7.41868138e-01 -1.55391550e+00 -7.57718608e-02
8.61517265e-02 -5.33355474e-01 1.03375542e+00 1.25817418e+00
1.66825190e-01 1.18464828e-01 -1.73565112e-02 -9.04215932e-01
-7.96090424e-01 -9.45297420e-01 -1.45083517e-01 4.56345260e-01
2.03294829e-01 -4.17224795e-01 -4.52122055e-02 7.20525458e-02] | [9.165234565734863, -6.4789252281188965] |
2dcb72a2-f7f2-4550-ad85-a9d18e216c9b | evaluating-gesture-generation-in-a-large | 2303.08737 | null | https://arxiv.org/abs/2303.08737v1 | https://arxiv.org/pdf/2303.08737v1.pdf | Evaluating gesture-generation in a large-scale open challenge: The GENEA Challenge 2022 | This paper reports on the second GENEA Challenge to benchmark data-driven automatic co-speech gesture generation. Participating teams used the same speech and motion dataset to build gesture-generation systems. Motion generated by all these systems was rendered to video using a standardised visualisation pipeline and evaluated in several large, crowdsourced user studies. Unlike when comparing different research papers, differences in results are here only due to differences between methods, enabling direct comparison between systems. The dataset was based on 18 hours of full-body motion capture, including fingers, of different persons engaging in a dyadic conversation. Ten teams participated in the challenge across two tiers: full-body and upper-body gesticulation. For each tier, we evaluated both the human-likeness of the gesture motion and its appropriateness for the specific speech signal. Our evaluations decouple human-likeness from gesture appropriateness, which has been a difficult problem in the field. The evaluation results are a revolution, and a revelation. Some synthetic conditions are rated as significantly more human-like than human motion capture. To the best of our knowledge, this has never been shown before on a high-fidelity avatar. On the other hand, all synthetic motion is found to be vastly less appropriate for the speech than the original motion-capture recordings. We also find that conventional objective metrics do not correlate well with subjective human-likeness ratings in this large evaluation. The one exception is the Fr\'echet gesture distance (FGD), which achieves a Kendall's tau rank correlation of around -0.5. Based on the challenge results we formulate numerous recommendations for system building and evaluation. | ['Gustav Eje Henter', 'Mihail Tsakov', 'Teodor Nikolov', 'Carla Viegas', 'Youngwoo Yoon', 'Pieter Wolfert', 'Taras Kucherenko'] | 2023-03-15 | null | null | null | null | ['gesture-generation'] | ['robots'] | [ 3.63846570e-02 1.41064942e-01 1.73891291e-01 -1.98775977e-01
-1.01004016e+00 -8.49142611e-01 1.06894827e+00 -5.38908243e-01
-5.45233309e-01 5.10147333e-01 9.28156495e-01 5.14351465e-02
1.30313054e-01 3.36926244e-02 -2.26125896e-01 -5.06372690e-01
5.78364693e-02 4.14925545e-01 3.00265729e-01 -3.39322001e-01
7.11409822e-02 3.12092423e-01 -1.59487939e+00 5.20822227e-01
3.48634794e-02 5.32478869e-01 -2.87711667e-03 1.35510671e+00
2.54400760e-01 4.83068168e-01 -9.92938757e-01 -2.42730916e-01
3.00277859e-01 -8.47000897e-01 -8.61243844e-01 -2.85597565e-03
5.25709271e-01 -5.56905210e-01 -3.16329092e-01 4.15080249e-01
1.18609834e+00 3.56123865e-01 4.40265894e-01 -1.40770113e+00
-2.28012398e-01 2.04478115e-01 1.16395392e-02 1.60023302e-01
1.26915026e+00 8.10263038e-01 9.44920242e-01 -7.33020186e-01
1.12099659e+00 1.43966901e+00 7.95800209e-01 1.02145147e+00
-9.22743082e-01 -3.82078290e-01 -2.92497993e-01 -3.01887602e-01
-1.19214582e+00 -7.15435266e-01 2.87421256e-01 -7.06621230e-01
1.07097137e+00 7.64982879e-01 9.94008899e-01 1.78140378e+00
-3.39591354e-01 6.47619903e-01 9.84556556e-01 -3.62153113e-01
1.20701984e-01 -1.58593222e-01 -4.37322825e-01 3.15006673e-01
-1.26055002e-01 2.53715068e-01 -9.11832333e-01 -1.50067896e-01
9.22275364e-01 -5.76416194e-01 -5.45267403e-01 -1.48840416e-02
-1.80370069e+00 3.26576382e-01 -2.08159015e-01 6.32496834e-01
-1.49773151e-01 4.09657866e-01 3.32167655e-01 3.22010726e-01
1.29001558e-01 4.09487873e-01 -1.41656831e-01 -1.23286390e+00
-9.41082776e-01 7.63188601e-01 1.12765539e+00 9.08969820e-01
-2.22702324e-01 -8.47698152e-02 -4.38160092e-01 5.15009522e-01
3.69421393e-01 4.96704102e-01 6.71898007e-01 -1.22033012e+00
4.91722167e-01 2.30411544e-01 4.46240097e-01 -1.01055539e+00
-5.96096039e-01 2.65513122e-01 -2.46780366e-01 5.92643380e-01
1.09163618e+00 -3.90962243e-01 -7.42015123e-01 1.68512249e+00
2.64035016e-01 -5.63154183e-02 -8.97790045e-02 1.56186473e+00
1.12699211e+00 5.01799464e-01 9.13484842e-02 -6.47940785e-02
1.30387366e+00 -7.90746868e-01 -7.90934563e-01 5.97426742e-02
7.25070119e-01 -1.00894356e+00 1.53959501e+00 3.71877968e-01
-1.22658551e+00 -4.70086217e-01 -1.01343155e+00 3.34030092e-02
6.27995282e-02 -1.06726646e-01 4.05752778e-01 9.85227466e-01
-1.20428801e+00 6.33080959e-01 -7.22180068e-01 -7.88163722e-01
-1.30329445e-01 4.36781570e-02 -6.10470295e-01 3.48512143e-01
-9.77819800e-01 9.06423986e-01 -1.09731235e-01 -3.34134176e-02
-5.71525931e-01 -3.69703680e-01 -6.46084070e-01 -6.67455494e-01
3.49321477e-02 -6.52007937e-01 1.83296025e+00 -8.86555433e-01
-1.86602366e+00 9.63107824e-01 -8.67582262e-02 2.89378245e-03
1.22975373e+00 -3.15816015e-01 -5.28702199e-01 8.71704593e-02
7.91012645e-02 7.43400872e-01 5.45036674e-01 -1.24352014e+00
-5.64234316e-01 5.10796681e-02 -9.22548473e-02 6.42301023e-01
9.51736644e-02 3.63124371e-01 -5.68343699e-01 -9.34568286e-01
-1.89792141e-02 -1.14732027e+00 1.26105472e-01 -1.97328534e-02
-1.90289766e-01 -2.00223345e-02 8.04252684e-01 -8.68351340e-01
1.22957277e+00 -1.98169124e+00 1.35441914e-01 7.96293244e-02
1.81904599e-01 2.55643547e-01 -4.74056713e-02 5.83365262e-01
1.81632489e-01 4.12300855e-01 -2.57608928e-02 -4.17161733e-01
1.95797384e-01 5.44895139e-03 -9.41453651e-02 4.87613231e-01
-1.91282198e-01 1.04132652e+00 -1.07153380e+00 -4.29963112e-01
2.30200931e-01 5.39500296e-01 -2.92194873e-01 1.84601530e-01
5.47066517e-02 8.20740640e-01 -4.27576117e-02 5.26217759e-01
-7.19575584e-02 8.84296000e-02 4.86368500e-02 -2.80296039e-02
-1.44092530e-01 2.85536081e-01 -1.18665743e+00 1.97299170e+00
-1.52257040e-01 1.21690571e+00 -5.70176542e-02 -1.16418183e-01
6.80983365e-01 8.95065784e-01 5.02730072e-01 -5.58830678e-01
-1.76227260e-02 2.92744368e-01 2.60301411e-01 -9.58077073e-01
6.53627932e-01 -9.63302851e-02 -1.81946173e-01 6.38069212e-01
-2.04248905e-01 -4.36074555e-01 4.27296907e-02 -3.91351357e-02
1.26716340e+00 5.46055198e-01 2.05860555e-01 7.91105628e-02
-1.13540262e-01 3.39906067e-01 8.53831097e-02 7.41712689e-01
-4.93657500e-01 1.34295988e+00 3.32916528e-01 -3.28063041e-01
-1.17899084e+00 -1.03567839e+00 3.72968793e-01 1.07449293e+00
6.93198144e-02 -6.36577189e-01 -1.00604260e+00 -3.23366910e-01
-4.42090660e-01 4.09906566e-01 -4.44141984e-01 2.82522321e-01
-6.93093717e-01 -3.92253995e-01 1.28838801e+00 5.41215658e-01
2.28537127e-01 -1.34099340e+00 -1.21038234e+00 8.19465145e-02
-4.41209465e-01 -1.26787710e+00 -7.12477207e-01 -6.63535893e-01
-4.66187239e-01 -8.66942644e-01 -1.21112156e+00 -4.66597378e-01
1.17632508e-01 9.81423333e-02 1.11806536e+00 2.01629251e-01
-2.38250092e-01 7.14785993e-01 -7.13990688e-01 -1.79612577e-01
-5.71731031e-01 -1.53552756e-01 2.57574111e-01 -3.16954046e-01
1.68526009e-01 -3.24258983e-01 -6.75276458e-01 7.66577542e-01
-5.91279328e-01 7.41525069e-02 2.91170359e-01 6.32653773e-01
5.28700138e-03 -6.63734794e-01 1.01895690e-01 -1.13345265e-01
1.12900019e+00 -1.75067917e-01 1.62755638e-01 -9.33483541e-02
-1.77322388e-01 -3.30929965e-01 -2.49604750e-02 -6.92440093e-01
-8.45998466e-01 6.88158348e-02 -2.48284694e-02 -2.27184966e-01
-4.19692487e-01 2.91573186e-03 -2.35060174e-02 2.51392752e-01
9.45712388e-01 -2.03852803e-01 1.15289629e-01 -2.93898761e-01
3.90061110e-01 8.34036469e-01 9.45369899e-01 -4.90658373e-01
6.08404875e-01 5.19022942e-01 -4.02893901e-01 -1.06868494e+00
1.49508342e-01 -3.91258895e-01 -5.96501291e-01 -6.89457834e-01
1.07107317e+00 -7.47848988e-01 -9.48645413e-01 5.23666263e-01
-1.25887275e+00 -8.82886112e-01 -3.82651448e-01 7.90600300e-01
-7.94341207e-01 2.71945506e-01 -3.79775912e-01 -1.03229356e+00
-3.94770317e-02 -1.18952787e+00 1.29551542e+00 8.69678631e-02
-1.31591451e+00 -7.27044702e-01 3.65541816e-01 4.67630655e-01
1.70230359e-01 7.61679292e-01 -1.05419680e-01 -5.31630874e-01
-4.99316538e-03 -4.12590951e-01 8.00419673e-02 -2.67746031e-01
2.13724419e-01 2.24564731e-01 -1.07767081e+00 -1.43788129e-01
-3.45452368e-01 -3.26386362e-01 2.32217908e-01 1.63048297e-01
7.33422115e-02 -2.49420717e-01 -9.17400941e-02 1.48850739e-01
5.45349419e-01 3.30991805e-01 7.58354783e-01 3.24268401e-01
8.32183838e-01 8.81308556e-01 7.50539541e-01 4.48627800e-01
2.50457436e-01 1.31129229e+00 8.71921796e-03 1.29964456e-01
-5.18213034e-01 -2.65190512e-01 7.21406877e-01 5.53745568e-01
-7.47952580e-01 -4.56380039e-01 -1.20899189e+00 5.28421223e-01
-1.83029699e+00 -1.23022366e+00 -5.29780686e-01 2.07260776e+00
5.64166784e-01 -1.78546354e-01 8.02505910e-01 2.02871472e-01
6.10271156e-01 1.82670593e-01 -5.22599183e-02 -4.96434450e-01
-1.89726099e-01 -8.04747865e-02 7.75415301e-02 6.42996550e-01
-8.28616738e-01 7.98411965e-01 6.87191582e+00 4.48722035e-01
-1.08094442e+00 1.59134358e-01 2.75179446e-01 -7.75823712e-01
-1.84419677e-01 -1.36653915e-01 -3.29674661e-01 5.49716771e-01
9.51609313e-01 8.49411339e-02 3.73622328e-01 4.71989512e-01
7.44723856e-01 -1.75805464e-01 -1.19929326e+00 1.15274882e+00
1.21748760e-01 -1.08859324e+00 -2.78620601e-01 1.05471678e-01
5.36670089e-01 -1.90664098e-01 -1.64758414e-01 1.44230630e-02
2.10600376e-01 -1.37900543e+00 1.16408217e+00 6.38251126e-01
1.10368335e+00 -3.14164251e-01 4.95562792e-01 2.88635314e-01
-1.15605319e+00 4.06782001e-01 3.34218621e-01 -3.25162232e-01
7.11574793e-01 -1.65076941e-01 -9.34203386e-01 8.45239013e-02
9.30768430e-01 1.73833519e-01 -2.67679751e-01 7.18921840e-01
-1.34225950e-01 3.62624556e-01 -3.97162318e-01 -3.73328149e-01
2.43149027e-01 1.52146801e-01 1.00764811e+00 1.64869368e+00
3.64882141e-01 3.36713850e-01 -1.64630145e-01 3.84539306e-01
3.30271930e-01 1.04290266e-02 -8.46461535e-01 -3.00421178e-01
4.42199260e-01 9.79971647e-01 -7.17171729e-01 -1.83386132e-01
-2.04203948e-01 1.25087142e+00 -3.09257150e-01 3.09391677e-01
-6.99505448e-01 -8.75286907e-02 8.02772462e-01 1.19699597e-01
-2.41933316e-01 -4.18960333e-01 -2.71837592e-01 -7.87188888e-01
3.10403496e-01 -1.09457457e+00 5.56010306e-02 -9.98149157e-01
-8.41657877e-01 6.76559627e-01 2.12934062e-01 -1.55189049e+00
-1.00613940e+00 -4.31383193e-01 -5.29764116e-01 9.08967137e-01
-2.03098029e-01 -9.32493269e-01 -7.08524108e-01 3.92921507e-01
6.06336951e-01 -1.26209840e-01 9.04022694e-01 1.76759452e-01
-2.46991649e-01 6.24040127e-01 -4.68277216e-01 1.48573458e-01
8.43519509e-01 -1.10632956e+00 9.66275632e-01 6.67962074e-01
3.70194554e-01 5.38964272e-01 1.03433204e+00 -6.38186038e-01
-1.19699550e+00 -4.09697294e-01 7.33214676e-01 -1.18146479e+00
4.93858576e-01 -2.69280106e-01 -4.67688799e-01 4.44392473e-01
2.40636349e-01 -6.44592717e-02 4.87352341e-01 -3.85691524e-01
-1.07234269e-01 6.43684566e-01 -1.05687642e+00 9.18489397e-01
1.62714303e+00 -5.53425968e-01 -5.80290496e-01 6.14787191e-02
4.43258077e-01 -7.94956028e-01 -8.36436987e-01 1.80567324e-01
1.38335705e+00 -1.14084566e+00 8.12361598e-01 -6.69053018e-01
3.42655778e-01 -3.41169775e-01 -2.75653154e-01 -1.31798112e+00
4.89684828e-02 -1.25923729e+00 4.09646332e-02 1.21429026e+00
3.70669097e-01 -1.39529333e-01 8.02047849e-01 9.69656229e-01
-3.12098060e-02 -4.39755380e-01 -1.01988256e+00 -8.43579471e-01
-2.21098170e-01 -8.64686072e-01 4.55563724e-01 9.91472304e-01
3.83512408e-01 1.39391735e-01 -5.11148870e-01 -2.51326412e-01
1.30472183e-01 -5.33799887e-01 1.46166968e+00 -1.03630030e+00
-4.03435051e-01 -7.84656107e-01 -6.85813904e-01 -9.11009490e-01
-3.53063107e-01 -5.28464317e-01 2.51588792e-01 -1.57815087e+00
-1.93261176e-01 -7.53409602e-03 6.38906896e-01 3.08954597e-01
1.04392849e-01 4.84183431e-01 5.43403089e-01 5.98357439e-01
-2.41702870e-01 8.87614563e-02 1.35105324e+00 2.07530558e-01
-6.49647474e-01 -1.00815050e-01 -3.21826011e-01 7.90593684e-01
7.51985312e-01 -1.39320642e-01 -4.90251720e-01 -3.85992736e-01
6.14891388e-02 2.07788363e-01 6.91165268e-01 -1.14139199e+00
7.83710107e-02 -1.91309914e-01 1.55430645e-01 -1.60709158e-01
5.93484521e-01 -4.05935079e-01 5.86257577e-01 3.59162152e-01
-3.05514067e-01 2.21369222e-01 1.38009116e-01 3.07392091e-01
8.18094388e-02 2.33651415e-01 2.94464856e-01 -2.36426786e-01
-8.28334808e-01 -2.05695212e-01 -5.38489938e-01 3.37051749e-01
7.95251191e-01 -8.13910007e-01 -2.19173357e-01 -1.08162379e+00
-8.66841137e-01 -1.93729386e-01 5.57724357e-01 8.54231477e-01
4.78211999e-01 -1.51979673e+00 -8.01706254e-01 -1.84323698e-01
2.09005401e-01 -1.30026430e-01 -1.05271220e-01 8.59148622e-01
-7.66537607e-01 2.02896073e-01 -2.02363774e-01 -7.99976945e-01
-1.43069673e+00 -2.22865358e-01 2.79539704e-01 1.82516143e-01
-8.84245753e-01 7.22660959e-01 -2.27028370e-01 -1.84664041e-01
2.20699430e-01 -3.80576640e-01 1.07948847e-01 1.06522679e-01
7.38318741e-01 7.47797072e-01 -9.59473550e-02 -1.21653295e+00
-4.41321999e-01 5.31325519e-01 5.89650810e-01 -1.25343168e+00
8.51770103e-01 -1.07870884e-01 5.69554687e-01 7.73321807e-01
9.16162908e-01 1.45081505e-01 -1.48059356e+00 4.54991162e-01
-1.72216818e-02 -4.95577365e-01 -4.74586964e-01 -8.44146609e-01
-5.16089261e-01 5.99433661e-01 6.57111526e-01 3.43083411e-01
6.00502491e-01 3.38878413e-03 6.66229248e-01 3.05644751e-01
4.27944899e-01 -1.17965376e+00 2.33431071e-01 3.15448225e-01
1.39417136e+00 -1.00331652e+00 -2.42886528e-01 -8.61076042e-02
-1.11855042e+00 8.93335402e-01 5.55214942e-01 2.15384707e-01
1.00165024e-01 3.32075119e-01 5.36388874e-01 -1.07030034e-01
-4.46857363e-01 -1.97396994e-01 4.63425785e-01 8.41971159e-01
7.32000709e-01 2.18616769e-01 -4.74389374e-01 4.83596861e-01
-8.57477784e-01 1.00450441e-01 4.51097548e-01 8.92344534e-01
-1.88034493e-02 -8.44129443e-01 -6.67152405e-01 -4.97904643e-02
-2.15224177e-01 3.03573042e-01 -9.80562210e-01 1.16958225e+00
-1.50666475e-01 1.31736302e+00 -2.24546771e-02 -8.47161591e-01
6.41091406e-01 1.15864687e-01 6.34699047e-01 -4.96412039e-01
-7.87285566e-01 -5.88714480e-02 5.63665569e-01 -9.32688951e-01
-5.86252570e-01 -8.99712324e-01 -1.33868420e+00 -4.95997787e-01
2.33742371e-01 -2.47427076e-01 7.15776026e-01 9.19701517e-01
1.36426911e-01 2.86331624e-01 5.82408980e-02 -1.61485171e+00
-2.18837231e-01 -1.20898616e+00 -2.22216517e-01 9.85297918e-01
2.82256454e-01 -5.95659912e-01 -4.35874611e-01 2.14066938e-01] | [5.606679439544678, -0.0901964008808136] |
9c93ead3-eb06-42b3-bc85-9bd8212bc7b2 | neural-ode-and-dae-modules-for-power-system | 2110.12981 | null | https://arxiv.org/abs/2110.12981v5 | https://arxiv.org/pdf/2110.12981v5.pdf | Feasibility Study of Neural ODE and DAE Modules for Power System Dynamic Component Modeling | In the context of high penetration of renewables, the need to build dynamic models of power system components based on accessible measurement data has become urgent. To address this challenge, firstly, a neural ordinary differential equations (ODE) module and a neural differential-algebraic equations (DAE) module are proposed to form a data-driven modeling framework that accurately captures components' dynamic characteristics and flexibly adapts to various interface settings. Secondly, analytical models and data-driven models learned by the neural ODE and DAE modules are integrated together and simulated simultaneously using unified transient stability simulation methods. Finally, the neural ODE and DAE modules are implemented with Python and made public on GitHub. Using the portal measurements, three simple but representative cases of excitation controller modeling, photovoltaic power plant modeling, and equivalent load modeling of a regional power network are carried out in the IEEE-39 system and 2383wp system. Neural dynamic model-integrated simulations are compared with the original model-based ones to verify the feasibility and potentiality of the proposed neural ODE and DAE modules. | ['Shaowei Huang', 'Huizhe Guan', 'Tirui He', 'Ying Chen', 'Tannan Xiao'] | 2021-10-25 | null | null | null | null | ['numerical-integration'] | ['miscellaneous'] | [-5.90433657e-01 -3.19311470e-01 1.17689602e-01 3.72541398e-02
1.29663959e-01 -9.62886393e-01 2.55939484e-01 -1.52315587e-01
5.65371513e-01 8.58424664e-01 -4.75945473e-01 -4.59080130e-01
-6.10144794e-01 -7.45958924e-01 -2.94408262e-01 -8.36946905e-01
-3.07410210e-01 1.23151444e-01 -3.77036989e-01 -3.66942972e-01
-3.00770730e-01 7.91069806e-01 -1.40475464e+00 -6.11776531e-01
1.28806579e+00 9.49704707e-01 1.40998676e-01 4.67928499e-01
1.89049512e-01 3.04542333e-01 -5.83096325e-01 4.10855234e-01
3.06396008e-01 -8.40316787e-02 -1.42852068e-01 -2.79775113e-01
-5.86496830e-01 -4.17283744e-01 -3.28313917e-01 1.02667403e+00
1.01707232e+00 4.24490184e-01 8.05090368e-01 -1.58404374e+00
-6.93368733e-01 3.68529320e-01 -2.22754106e-01 4.98135164e-02
2.85819083e-01 3.72811079e-01 4.10641313e-01 -7.75335431e-01
1.57681882e-01 8.34985733e-01 9.78623152e-01 1.77391663e-01
-1.28238559e+00 -5.55012882e-01 -7.94042274e-03 1.15057915e-01
-1.50629759e+00 9.30485576e-02 8.94149542e-01 -4.56002504e-01
1.24821973e+00 3.90107125e-01 1.08464074e+00 7.43077099e-01
2.96248913e-01 2.96256900e-01 9.44872618e-01 -7.79517069e-02
4.94763166e-01 2.14294612e-01 4.48718548e-01 2.40602121e-01
2.70027280e-01 2.51740158e-01 1.30717665e-01 -2.11256146e-01
8.74986172e-01 -2.56517112e-01 -5.21791875e-01 -2.24318072e-01
-4.85739082e-01 2.92752177e-01 2.97255963e-01 4.72475171e-01
-5.65108776e-01 -2.23744497e-01 3.13669711e-01 3.50939810e-01
4.20367680e-02 3.52813154e-01 -6.62335277e-01 -8.15578550e-03
-7.86221504e-01 -3.40293236e-02 1.30951989e+00 1.00419009e+00
2.28354856e-01 1.10810173e+00 1.63959324e-01 8.51775467e-01
4.34616596e-01 6.68713391e-01 5.85761547e-01 -8.81232977e-01
-3.39207470e-01 8.89145195e-01 2.76865691e-01 -8.60069871e-01
-7.95196414e-01 -7.26408184e-01 -1.28949797e+00 2.51261741e-01
-1.10760517e-01 -7.55237162e-01 -5.72778881e-01 1.50465846e+00
4.40996557e-01 -8.38297382e-02 1.06254756e-01 6.21848404e-01
4.94868189e-01 1.38278651e+00 -2.82891184e-01 -7.79782176e-01
8.78377378e-01 -5.82989514e-01 -9.05151486e-01 5.23228347e-01
3.12452078e-01 -3.29106033e-01 8.72905612e-01 2.51669973e-01
-1.28067219e+00 -5.70827484e-01 -1.30360484e+00 2.29903430e-01
-9.18772638e-01 5.55064917e-01 1.56362150e-02 5.37807941e-01
-1.39514279e+00 7.59036720e-01 -8.25121462e-01 -5.34799337e-01
-1.93632752e-01 4.22906756e-01 3.21761221e-01 9.16087508e-01
-1.34768260e+00 1.17861795e+00 5.72951138e-01 6.71687424e-01
-7.41734505e-01 -1.19237781e+00 -3.61276835e-01 5.89478493e-01
1.49585083e-01 -9.57184792e-01 1.15618944e+00 -7.55330324e-01
-2.27003241e+00 -4.14197713e-01 2.46148616e-01 -3.26411813e-01
9.92813632e-02 2.66046554e-01 -7.57484913e-01 -1.84682667e-01
-6.52494013e-01 -2.13799864e-01 4.41814452e-01 -1.10472369e+00
-2.34056756e-01 7.34592751e-02 -1.34472534e-01 1.21910416e-01
-5.17495096e-01 -3.44744056e-01 2.51414716e-01 -4.10917610e-01
-2.92696416e-01 -4.60221559e-01 -2.97917873e-01 1.09945826e-01
-4.47836995e-01 -2.99140632e-01 1.11355019e+00 -1.18615091e+00
1.23881471e+00 -1.70850945e+00 2.49521554e-01 6.23355031e-01
-3.91033500e-01 6.14182889e-01 1.96597084e-01 8.36624086e-01
-3.13704133e-01 -9.26408172e-02 5.65252677e-02 3.22949439e-01
5.31446457e-01 2.34838963e-01 -1.17284060e-01 2.63573676e-01
6.05855212e-02 5.88253856e-01 -3.21542442e-01 2.24241704e-01
7.22581863e-01 7.92921007e-01 -7.45208636e-02 1.38535291e-01
-2.10913539e-01 4.01893616e-01 -6.24308944e-01 6.13258719e-01
7.19379485e-01 -9.90057811e-02 2.75724113e-01 -6.34325206e-01
-5.92529058e-01 -2.52439588e-01 -1.64725971e+00 1.26667225e+00
-8.95559728e-01 4.28363651e-01 7.70535469e-01 -1.16813016e+00
9.57274556e-01 5.05243719e-01 6.51737452e-01 -6.76276684e-01
4.68210012e-01 2.05109209e-01 -5.83590120e-02 -4.88702029e-01
2.05003321e-01 3.86086494e-01 3.45112801e-01 3.65245521e-01
5.00732839e-01 -3.52877587e-01 3.63877207e-01 -7.73222223e-02
5.57247818e-01 3.21002215e-01 4.57282722e-01 -8.64915252e-01
9.86589849e-01 -6.22439347e-02 8.14741194e-01 5.81507459e-02
1.54301047e-01 -3.90284844e-02 2.22790867e-01 -3.00322741e-01
-1.17471850e+00 -8.48030448e-01 -3.98755789e-01 3.22039664e-01
-1.23752452e-01 -1.82606637e-01 -7.38561213e-01 1.79071665e-01
-1.01214424e-01 1.09799016e+00 -1.95672497e-01 -3.39234680e-01
-1.64380565e-01 -8.40653598e-01 1.16294220e-01 6.57102168e-01
4.98336434e-01 -4.42253888e-01 -4.78947610e-01 4.34760541e-01
4.31893229e-01 -6.45278990e-01 2.24349409e-01 5.37972510e-01
-6.84750915e-01 -8.80468726e-01 -6.60326302e-01 -7.18365669e-01
6.14366770e-01 -4.60461557e-01 9.36096787e-01 -3.19842659e-02
-3.54811907e-01 7.31964529e-01 2.22670868e-01 -2.26494938e-01
-5.22041857e-01 9.49578211e-02 6.89123392e-01 -4.44559753e-01
-4.34486181e-01 -1.24792290e+00 -2.25426868e-01 2.59399384e-01
-5.43609381e-01 8.49656239e-02 3.04920942e-01 6.50912225e-01
3.56540799e-01 5.04465461e-01 1.08173645e+00 7.05158934e-02
9.48561192e-01 -7.14177907e-01 -1.62404537e+00 5.67036271e-01
-1.05192447e+00 -2.53671706e-01 1.32364285e+00 -4.32241142e-01
-1.16459608e+00 1.59260049e-01 -1.25641152e-01 -5.23756683e-01
-6.01993240e-02 6.82781160e-01 -5.07707477e-01 -2.48722807e-01
2.19865404e-02 4.62518632e-01 1.37996465e-01 -6.30266190e-01
1.39422506e-01 6.80704534e-01 6.89329207e-01 -7.23503172e-01
1.14545429e+00 -2.71430284e-01 2.27204993e-01 -8.29880059e-01
2.44873866e-01 2.05825195e-01 -4.74830687e-01 -4.81704503e-01
2.18010843e-01 -9.61193025e-01 -1.23231971e+00 9.68500018e-01
-1.08003521e+00 -4.17375147e-01 -6.04738176e-01 2.04316512e-01
-3.98171753e-01 7.10115358e-02 -5.23444712e-01 -1.05513453e+00
-8.07578743e-01 -1.10368574e+00 2.30280399e-01 9.35116768e-01
7.37368837e-02 -1.27704680e+00 1.03765443e-01 -5.46563148e-01
9.69717145e-01 5.11312664e-01 1.10455275e+00 -4.42004263e-01
-3.48377258e-01 -2.81087011e-01 1.33855805e-01 7.27084816e-01
-3.31001356e-02 8.17628384e-01 -6.33727551e-01 -5.89413345e-01
2.77261317e-01 1.39763713e-01 -6.60623014e-01 5.36485612e-01
8.28718543e-01 -4.51868296e-01 -2.29872450e-01 6.15231812e-01
2.08451915e+00 5.95795035e-01 2.83776402e-01 -8.02851543e-02
2.37907439e-01 1.08249083e-01 -1.31603897e-01 5.82216680e-01
3.91042411e-01 2.39570081e-01 4.66774881e-01 -1.16646118e-01
6.02870464e-01 2.11851001e-01 4.87425923e-01 1.33012891e+00
-8.60855952e-02 -1.01355582e-01 -6.85480356e-01 2.98487037e-01
-1.58584285e+00 -5.40804207e-01 -2.28948027e-01 1.90106356e+00
4.00204867e-01 -3.51725876e-01 8.53877813e-02 1.83580667e-01
7.95808613e-01 -3.06535810e-01 -8.72571409e-01 -6.51807964e-01
-1.99180633e-01 9.21018347e-02 2.36248538e-01 2.13036045e-01
-3.69164437e-01 -1.57692119e-01 6.48894787e+00 8.23084772e-01
-1.34333289e+00 2.50359561e-04 2.78791755e-01 -2.87013338e-03
1.65558428e-01 -1.28869250e-01 -5.52366316e-01 6.25573397e-01
1.47673905e+00 -1.17880261e+00 8.74756455e-01 9.99739587e-01
8.40031564e-01 -1.61088064e-01 -8.43872070e-01 5.61058700e-01
-4.86714989e-01 -1.18477237e+00 -2.39430398e-01 -7.42512792e-02
9.93315935e-01 -1.36433337e-02 -2.29200244e-01 6.65386379e-01
3.75863194e-01 -4.84802812e-01 4.97603327e-01 9.40094113e-01
2.19704837e-01 -7.03424692e-01 6.69424176e-01 5.53072095e-01
-1.37867069e+00 -6.43765867e-01 -2.53830492e-01 -2.36283615e-01
5.06157696e-01 7.27338970e-01 -3.41050059e-01 1.29096377e+00
7.90325403e-01 7.55239606e-01 -3.08269829e-01 1.21477425e+00
5.51615842e-02 6.05922222e-01 -7.67414868e-01 -1.73272386e-01
-2.22574860e-01 -5.67894220e-01 6.63488388e-01 6.03967428e-01
7.17316985e-01 1.04975462e-01 -1.13718927e-01 1.39655507e+00
3.71574789e-01 -4.71270122e-02 -3.05790722e-01 -9.35529731e-03
6.30624235e-01 1.88592434e+00 -3.32178026e-02 -2.36061811e-01
-2.32895792e-01 2.10051775e-01 -3.14088970e-01 8.29779804e-01
-1.17951989e+00 -5.01168847e-01 6.17534220e-01 -3.87013733e-01
-1.41130984e-01 -3.18979651e-01 -2.77429461e-01 -1.03539443e+00
-1.06468918e-02 -6.53535068e-01 3.08601502e-02 -1.42127991e+00
-1.41463971e+00 3.46666783e-01 8.23913440e-02 -1.48442745e+00
-8.58021200e-01 -7.36203492e-01 -1.30754733e+00 1.32460678e+00
-1.18137193e+00 -7.70482242e-01 -3.00962359e-01 4.46455210e-01
3.00132155e-01 -2.56650090e-01 9.21295583e-01 4.99620497e-01
-1.33323336e+00 -1.29823595e-01 9.21979845e-01 -4.49932933e-01
-2.86214590e-01 -1.24570012e+00 -1.00808404e-01 9.79806066e-01
-8.28277528e-01 5.12651980e-01 7.17982054e-01 -2.51774609e-01
-1.90376925e+00 -8.22972834e-01 7.79115558e-02 3.08430970e-01
8.31076920e-01 -1.53294712e-01 -1.22074068e+00 3.35520804e-01
9.51151550e-01 -1.13330573e-01 2.17310518e-01 -6.00093007e-01
5.30564785e-01 -4.07865465e-01 -1.35206616e+00 4.42378640e-01
2.95521706e-01 -3.55197161e-01 -2.21766666e-01 9.11892876e-02
2.32909828e-01 -3.96628946e-01 -1.54008996e+00 5.05251825e-01
4.52455819e-01 -3.43771368e-01 1.03907120e+00 -4.39125933e-02
-2.82664746e-01 -8.07814956e-01 4.03437801e-02 -1.89452612e+00
-2.15678528e-01 -1.18449175e+00 -7.85366714e-01 1.72525847e+00
2.62057483e-01 -1.04061723e+00 -8.15493613e-02 8.11495960e-01
-2.22219959e-01 -8.87100220e-01 -1.04810107e+00 -9.31269705e-01
3.26266080e-01 3.62761840e-02 9.11271572e-01 1.07576275e+00
1.90594211e-01 1.61264092e-01 2.26050615e-01 7.10931063e-01
2.59160757e-01 -1.41699627e-01 4.51754481e-01 -1.30530512e+00
-6.54535666e-02 -4.19227660e-01 1.90826073e-01 -3.69343549e-01
1.37690231e-01 -4.87719804e-01 -3.85260463e-01 -1.82732558e+00
-3.26009005e-01 -9.12564099e-02 -3.45932573e-01 4.13846821e-01
3.39776158e-01 -5.98261297e-01 5.22505306e-02 3.78324799e-02
3.61966223e-01 1.04314864e+00 7.92207599e-01 -3.08860671e-02
-3.47494751e-01 2.17342116e-02 4.24515605e-02 6.86726630e-01
9.36308086e-01 2.30563492e-01 -5.78587115e-01 -2.41168335e-01
-6.47335798e-02 4.53348398e-01 6.41667247e-01 -1.56971443e+00
6.12545967e-01 6.93336725e-02 4.75478888e-01 -9.23647881e-01
1.37127161e-01 -1.37031674e+00 8.66417527e-01 5.72961330e-01
2.98749566e-01 4.64370787e-01 6.75471485e-01 1.47414625e-01
1.05120383e-01 1.49557926e-02 5.55542707e-01 5.03953874e-01
-3.21338177e-01 1.12302229e-02 -8.06158185e-01 -4.68018711e-01
1.08787239e+00 -2.11595520e-01 -5.98827660e-01 -3.16574037e-01
-1.01155150e+00 7.32497573e-01 2.13077009e-01 2.89830238e-01
-1.08031005e-01 -1.28902388e+00 -2.74560958e-01 3.02820265e-01
-7.69735157e-01 -1.11236244e-01 5.27072430e-01 7.99439490e-01
-2.12143853e-01 4.01597053e-01 -5.57804763e-01 -5.77387929e-01
-5.25836825e-01 4.51903611e-01 1.15938175e+00 -5.73909581e-02
-2.59725958e-01 -6.78877681e-02 -5.22684991e-01 -8.25504243e-01
3.68619822e-02 -6.08391881e-01 -1.09162927e-01 4.12224047e-02
-4.42737713e-02 9.69173908e-01 4.88185361e-02 -2.75172591e-01
-1.04411915e-01 4.73421335e-01 7.91853428e-01 1.30305052e-01
1.60034192e+00 -3.35235178e-01 -1.50083706e-01 4.62358385e-01
8.57516408e-01 -5.88843584e-01 -1.23466337e+00 2.44436815e-01
-4.49214697e-01 4.33348835e-01 2.92706162e-01 -1.24263227e+00
-1.55093729e+00 6.35691106e-01 8.55048120e-01 7.88289070e-01
1.53401828e+00 -1.02311933e+00 3.10505807e-01 5.76499701e-01
3.54902476e-01 -1.56675160e+00 -6.75945878e-01 5.28187633e-01
9.65660334e-01 -2.20900759e-01 -5.50628640e-02 -1.97950706e-01
1.18383817e-01 1.45994186e+00 8.02610457e-01 -1.64781213e-01
9.64534998e-01 9.34767365e-01 -2.95002550e-01 3.06724399e-01
-8.21967304e-01 4.20796275e-01 7.93145373e-02 6.19967580e-01
1.31295919e-01 -3.29996347e-01 -4.56191599e-01 1.18265152e+00
1.72040954e-01 5.40107250e-01 7.35752881e-01 9.55361247e-01
1.49423301e-01 -6.87771678e-01 -3.78204286e-01 2.54142135e-01
-1.75561652e-01 4.81042892e-01 4.57505658e-02 1.02630877e+00
-1.44421563e-01 6.16608977e-01 5.49389841e-03 -8.23957250e-02
5.64823985e-01 2.10628867e-01 2.15054844e-02 1.48972571e-01
-9.31070864e-01 -9.99060124e-02 -1.33476049e-01 -3.25070411e-01
-1.89197958e-02 -3.73486102e-01 -1.28129387e+00 -3.32788020e-01
-6.89122736e-01 2.82135338e-01 1.03032458e+00 6.70758665e-01
5.91883421e-01 1.06211841e+00 9.00012553e-01 -1.15103447e+00
-8.73658240e-01 -8.51883531e-01 -7.35326767e-01 -5.23398399e-01
-3.81631702e-02 -5.57931781e-01 -5.32090008e-01 -1.56687945e-01] | [5.652680397033691, 2.5572242736816406] |
9b93cc70-901d-48b5-a40b-ccba457e018e | evopose2d-pushing-the-boundaries-of-2d-human | 2011.08446 | null | https://arxiv.org/abs/2011.08446v2 | https://arxiv.org/pdf/2011.08446v2.pdf | EvoPose2D: Pushing the Boundaries of 2D Human Pose Estimation using Accelerated Neuroevolution with Weight Transfer | Neural architecture search has proven to be highly effective in the design of efficient convolutional neural networks that are better suited for mobile deployment than hand-designed networks. Hypothesizing that neural architecture search holds great potential for human pose estimation, we explore the application of neuroevolution, a form of neural architecture search inspired by biological evolution, in the design of 2D human pose networks for the first time. Additionally, we propose a new weight transfer scheme that enables us to accelerate neuroevolution in a flexible manner. Our method produces network designs that are more efficient and more accurate than state-of-the-art hand-designed networks. In fact, the generated networks process images at higher resolutions using less computation than previous hand-designed networks at lower resolutions, allowing us to push the boundaries of 2D human pose estimation. Our base network designed via neuroevolution, which we refer to as EvoPose2D-S, achieves comparable accuracy to SimpleBaseline while being 50% faster and 12.7x smaller in terms of file size. Our largest network, EvoPose2D-L, achieves new state-of-the-art accuracy on the Microsoft COCO Keypoints benchmark, is 4.3x smaller than its nearest competitor, and has similar inference speed. The code is publicly available at https://github.com/wmcnally/evopose2d. | ['John McPhee', 'Alexander Wong', 'Kanav Vats', 'William McNally'] | 2020-11-17 | null | null | null | null | ['2d-human-pose-estimation'] | ['computer-vision'] | [-1.60837144e-01 6.96692020e-02 -9.61115435e-02 7.18812197e-02
-2.61027753e-01 -5.31264424e-01 2.13605404e-01 -3.78710091e-01
-7.66452789e-01 7.02899456e-01 -3.27106453e-02 -3.69940102e-01
-1.78490534e-01 -7.34193563e-01 -1.01760423e+00 -2.81336784e-01
-2.90079206e-01 7.20357835e-01 1.77522436e-01 -5.66512764e-01
-7.08798692e-02 6.06168568e-01 -1.66123605e+00 -1.69607088e-01
4.90200430e-01 9.80544031e-01 1.26165748e-01 8.75756979e-01
3.65998656e-01 8.67081527e-03 -7.36508727e-01 -5.64970613e-01
4.11228001e-01 -1.81609929e-01 -7.35547781e-01 -6.93124771e-01
7.11311400e-01 -1.58313110e-01 -3.88912499e-01 6.03716850e-01
1.08693516e+00 1.31423980e-01 4.26081806e-01 -1.16173530e+00
-5.80255449e-01 5.71812391e-01 -3.34784687e-01 2.99790770e-01
1.28839642e-01 2.47883886e-01 7.32255161e-01 -7.19976664e-01
6.77876413e-01 1.18199444e+00 1.45554864e+00 1.06749618e+00
-1.10579169e+00 -7.39628434e-01 -8.20618570e-02 -1.85802560e-02
-1.54856122e+00 -4.03461456e-01 4.74316299e-01 -4.50476184e-02
1.22806215e+00 4.00994718e-01 1.11500764e+00 1.42471373e+00
3.64175171e-01 8.43546391e-01 4.47070777e-01 -2.48474553e-01
2.10577548e-01 -3.64251256e-01 -4.40071464e-01 1.01492631e+00
5.01291573e-01 3.46026480e-01 -5.84956110e-01 -1.76998451e-01
9.83079374e-01 -2.99386948e-01 -1.70185015e-01 -4.80430573e-01
-1.30135787e+00 7.22966552e-01 7.40374446e-01 1.25289649e-01
-4.82766420e-01 7.48842299e-01 3.62995714e-01 1.43589433e-02
2.89003849e-01 1.05987239e+00 -7.81893015e-01 -4.62002128e-01
-8.22742701e-01 5.87773621e-01 7.60790944e-01 7.14114785e-01
5.06082624e-02 1.69586375e-01 8.02452862e-02 7.94613004e-01
4.68497314e-02 5.34247041e-01 6.74390376e-01 -1.14926422e+00
7.17120022e-02 2.98538238e-01 -1.61596954e-01 -1.02174902e+00
-7.49207199e-01 -8.39327753e-01 -8.74476850e-01 2.62258708e-01
3.28724384e-01 -4.08176929e-01 -9.94416058e-01 1.94418871e+00
4.92496252e-01 1.89165726e-01 -9.52528119e-02 6.57384336e-01
7.72875249e-01 4.04068768e-01 -2.88991839e-01 4.11453843e-01
1.20025945e+00 -1.08128202e+00 -6.12075441e-02 -1.99455008e-01
4.24432635e-01 -3.79710168e-01 8.55193615e-01 4.70379382e-01
-1.20460010e+00 -7.82117307e-01 -1.23705792e+00 1.82743594e-01
-3.17722827e-01 7.47134269e-04 9.07644808e-01 9.06654418e-01
-1.28561497e+00 8.06959569e-01 -1.01235080e+00 -3.66869986e-01
5.10646462e-01 5.92296541e-01 -2.12260067e-01 5.13506293e-01
-1.22935176e+00 9.28186297e-01 6.18832946e-01 1.45467162e-01
-6.52175605e-01 -7.45430171e-01 -6.90160811e-01 -1.37487814e-01
3.12126577e-01 -1.09012914e+00 1.28959131e+00 -6.16651535e-01
-1.63217938e+00 5.18417418e-01 1.94987908e-01 -8.19255352e-01
5.64733088e-01 -2.37406179e-01 -2.90821820e-01 1.62739921e-02
-2.65895188e-01 1.30778062e+00 6.75669014e-01 -8.23338926e-01
-5.61607957e-01 -3.03444296e-01 -1.52933285e-01 2.11104959e-01
-5.64711928e-01 -4.28208828e-01 -9.70082581e-01 -9.74922359e-01
-1.76218942e-01 -1.45277178e+00 -2.97276735e-01 1.89655542e-01
-2.90636212e-01 -2.53145248e-01 6.08973384e-01 -5.48503458e-01
1.26239562e+00 -1.67143774e+00 4.70639110e-01 1.64433718e-01
4.83284026e-01 7.05100358e-01 -2.98628986e-01 8.67386833e-02
9.03913751e-02 1.71484619e-01 -7.80166388e-02 -4.09538060e-01
-1.05961956e-01 1.10754400e-01 3.03949803e-01 2.05929920e-01
-5.27309552e-02 1.25546789e+00 -6.54633999e-01 -3.49180609e-01
-1.76260784e-01 6.43324256e-01 -8.48384321e-01 -1.80317253e-01
-2.29068667e-01 1.31218061e-01 -3.42401356e-01 6.97762072e-01
3.20713639e-01 -1.84128404e-01 -6.67345300e-02 -1.47504866e-01
1.19035197e-02 -1.94964021e-01 -8.42695117e-01 2.01337433e+00
-4.12275434e-01 8.14060152e-01 -3.33501965e-01 -5.46079278e-01
8.97073746e-01 -8.05540010e-04 4.52021956e-01 -5.33480048e-01
3.01164716e-01 1.26337275e-01 1.11347184e-01 -3.01822335e-01
5.20520747e-01 4.29857403e-01 -1.31987870e-01 2.88625479e-01
1.78980112e-01 -2.38395587e-01 9.96354148e-02 -1.12207860e-01
1.17158651e+00 2.74278760e-01 1.30681500e-01 -1.45591632e-01
-7.17627332e-02 -1.25429615e-01 4.62013215e-01 8.18926275e-01
-2.73956180e-01 5.57152927e-01 7.77351111e-02 -8.28607082e-01
-1.29931772e+00 -9.86158907e-01 -1.09223604e-01 1.06290710e+00
2.55451612e-02 -5.09189606e-01 -9.98632252e-01 -5.17970264e-01
2.82935500e-01 4.32317823e-01 -9.57506716e-01 -3.20138693e-01
-6.87350571e-01 -6.36805117e-01 1.44578433e+00 7.70538032e-01
6.49609447e-01 -1.02259302e+00 -1.11320591e+00 1.10680968e-01
-1.93194095e-02 -6.20835543e-01 -3.99259418e-01 1.36259466e-01
-8.66859734e-01 -8.70032668e-01 -1.04600418e+00 -6.51930571e-01
2.46562541e-01 -1.81972086e-01 1.21840560e+00 3.42001617e-01
-4.79702890e-01 1.99809477e-01 -2.14775190e-01 -7.51533806e-01
-2.06663132e-01 8.02470386e-01 4.01369095e-01 -6.68049037e-01
9.64510739e-02 -5.96291959e-01 -6.62203312e-01 3.59532863e-01
-6.80778325e-01 5.70457615e-03 6.66047096e-01 9.96964097e-01
5.54830968e-01 7.06201121e-02 3.90862137e-01 -5.59169233e-01
8.11102808e-01 -3.19104403e-01 -3.09419394e-01 2.01605663e-01
-7.22676754e-01 3.28127772e-01 4.72665578e-01 -6.42810225e-01
-6.92518890e-01 1.75122157e-01 -4.70595062e-01 -5.46548665e-01
-9.75869223e-02 3.84294868e-01 3.61093879e-01 -4.01126027e-01
1.11384213e+00 9.54927504e-03 3.10727477e-01 -4.41067845e-01
3.22401345e-01 3.50035161e-01 7.82316804e-01 -5.71083069e-01
6.78222775e-01 1.30058914e-01 1.39676511e-01 -8.58188629e-01
-3.14129204e-01 2.90126875e-02 -5.46131611e-01 -1.40115753e-01
8.08778167e-01 -6.82081103e-01 -9.38204825e-01 7.22990751e-01
-9.28572595e-01 -5.83381057e-01 -9.62871537e-02 2.70756245e-01
-5.93633950e-01 -1.21773742e-01 -5.71518302e-01 -4.27942127e-01
-7.43229449e-01 -1.00708628e+00 9.88619208e-01 2.90812820e-01
-5.60044229e-01 -8.63414824e-01 2.85421342e-01 2.16080338e-01
6.19014561e-01 4.02434230e-01 5.71006238e-01 -4.74584281e-01
-5.50785959e-02 -3.06587934e-01 1.80331886e-01 1.13487206e-01
-2.21144050e-01 1.04093283e-01 -5.22137105e-01 -4.48058635e-01
-5.54477036e-01 -4.52048510e-01 8.60429823e-01 4.86095339e-01
1.33540237e+00 -1.24757707e-01 -5.56969821e-01 1.04806292e+00
9.79215682e-01 1.90097213e-01 6.02133989e-01 7.17954993e-01
6.83436811e-01 2.39919052e-01 1.91380590e-01 2.77015716e-01
4.41476583e-01 9.58291590e-01 4.40293431e-01 -1.62603036e-01
-2.14082897e-01 -3.16571504e-01 -1.19870290e-01 6.31456137e-01
-5.47568262e-01 -3.84289980e-01 -1.11132324e+00 3.43139380e-01
-1.78223872e+00 -8.81964207e-01 3.94459456e-01 1.94326854e+00
7.91996539e-01 3.43143225e-01 5.26064873e-01 -3.09175402e-02
5.27253807e-01 -1.13686055e-01 -8.67086172e-01 -3.49778742e-01
-7.18914717e-02 5.41234851e-01 6.56194806e-01 8.18840265e-02
-1.13668621e+00 8.71693134e-01 6.57586861e+00 8.72174382e-01
-1.08126867e+00 -6.31891489e-02 5.77996433e-01 -5.70306540e-01
6.62924871e-02 -5.31971276e-01 -8.84293556e-01 5.36845803e-01
1.07116437e+00 1.55980289e-01 5.63062966e-01 1.01670635e+00
-1.87810123e-01 2.52901584e-01 -8.75254750e-01 9.48388994e-01
1.50844932e-01 -1.66587985e+00 -4.97920141e-02 9.86859053e-02
6.93953454e-01 2.28055090e-01 2.81496912e-01 2.73861051e-01
1.77628607e-01 -1.18831491e+00 9.94778991e-01 4.37458724e-01
8.36437643e-01 -1.21099102e+00 7.67534912e-01 2.55409181e-01
-1.21814287e+00 -2.29806244e-01 -3.17596376e-01 9.33884755e-02
8.93564299e-02 5.89531064e-02 -9.64660645e-01 1.83705345e-01
1.20735109e+00 3.23488444e-01 -9.73710239e-01 1.21874201e+00
-2.11575422e-02 3.09658676e-01 -5.15765429e-01 -6.63685203e-01
1.82743341e-01 5.49008489e-01 5.94361722e-01 1.05673766e+00
4.70160753e-01 -2.15432897e-01 -2.04176247e-01 4.97862726e-01
-2.28074417e-01 -3.02234739e-01 -5.57723463e-01 8.32616091e-02
9.02902365e-01 9.18044865e-01 -6.90547824e-01 -1.12770639e-01
3.83221179e-01 1.06624389e+00 4.45214093e-01 -2.09186487e-02
-1.00027931e+00 -6.36167705e-01 7.25859344e-01 -1.04870103e-01
5.22527337e-01 -3.29932690e-01 -1.50680870e-01 -8.35395813e-01
-1.80676401e-01 -1.19631577e+00 1.40740097e-01 -7.58073926e-01
-9.10703063e-01 1.01567447e+00 6.66675195e-02 -8.31982613e-01
-4.78981763e-01 -7.77323782e-01 -3.79774630e-01 5.45808792e-01
-5.78391731e-01 -1.11732721e+00 -3.77587140e-01 3.59715968e-01
4.91033792e-01 -4.28346783e-01 9.64777291e-01 2.40721762e-01
-6.95340455e-01 1.25127482e+00 2.72048842e-02 7.35572726e-02
3.61563742e-01 -1.01565385e+00 1.27722299e+00 6.51337862e-01
1.85942575e-01 8.61451745e-01 7.17292786e-01 -7.23788917e-01
-1.53268182e+00 -9.61246371e-01 4.93622810e-01 -6.79003119e-01
3.38382065e-01 -3.02435070e-01 -4.79362071e-01 4.60678548e-01
-2.97558773e-02 -2.16553256e-01 4.21973109e-01 2.78064668e-01
-2.37981945e-01 6.20014407e-02 -1.11728585e+00 8.96719515e-01
1.51473236e+00 -1.33550078e-01 -4.49266195e-01 2.82861292e-02
9.82943833e-01 -8.39232206e-01 -8.48238349e-01 6.04510248e-01
1.17057920e+00 -7.48598576e-01 1.37945080e+00 -5.71125805e-01
3.21117699e-01 -1.45100564e-01 2.77943797e-02 -1.43346214e+00
-5.45654297e-01 -5.66545784e-01 -4.53938961e-01 6.62897468e-01
7.75225699e-01 -7.37054229e-01 1.28194654e+00 3.15330476e-01
-1.70483544e-01 -1.39475095e+00 -9.93194878e-01 -1.03776014e+00
1.80869699e-01 -2.54151762e-01 8.59462202e-01 6.61703527e-01
-4.55258876e-01 2.11905286e-01 -5.41168451e-01 -1.69028550e-01
5.96719742e-01 -3.63059580e-01 9.88454342e-01 -1.23141778e+00
-7.22411931e-01 -7.84031749e-01 -6.17875934e-01 -1.03755248e+00
5.74280694e-02 -5.57883263e-01 6.81663156e-02 -1.06793463e+00
-1.09396107e-01 -5.13818800e-01 -8.63772929e-02 6.56121254e-01
-8.76550823e-02 6.26839936e-01 2.81951368e-01 9.00968164e-03
-2.67249376e-01 4.88568187e-01 1.09690642e+00 -2.57030968e-02
-3.81945729e-01 7.83158541e-02 -6.71432436e-01 6.83597684e-01
1.09884894e+00 -2.15120167e-01 -3.32720488e-01 -6.19851410e-01
2.96052992e-01 -2.37535387e-01 2.72596687e-01 -1.50570107e+00
3.52769583e-01 3.20215553e-01 9.10807371e-01 -5.46231985e-01
5.65080762e-01 -4.27011013e-01 5.32261312e-01 8.11961293e-01
-3.43959779e-01 6.51227236e-01 5.55500984e-01 4.09151286e-01
3.18327963e-01 -2.07685307e-01 4.76776004e-01 -2.85562277e-01
-8.19630563e-01 3.94504249e-01 -2.76078999e-01 -2.83023529e-02
7.53107965e-01 -4.54017937e-01 -2.64193118e-01 -2.29641110e-01
-4.43289578e-01 7.71141872e-02 5.15345037e-01 6.63538992e-01
6.70159578e-01 -1.39295053e+00 -4.99452084e-01 1.10277459e-02
-3.20667736e-02 -4.49413322e-02 1.62140459e-01 4.65440303e-01
-1.03760767e+00 4.63707745e-01 -5.00767231e-01 -5.01461267e-01
-1.37073898e+00 3.39989394e-01 5.90177178e-01 -7.54673779e-02
-4.19864446e-01 1.44909942e+00 -4.38364804e-01 -7.55465984e-01
4.19530153e-01 -5.65433428e-02 7.04701543e-02 -2.19771117e-01
4.30033177e-01 6.46864891e-01 1.12581901e-01 -4.95129734e-01
-4.81876671e-01 5.64566076e-01 1.18242584e-01 -9.25199762e-02
1.51680052e+00 3.31262022e-01 1.39913008e-01 -2.68235132e-02
1.06270564e+00 -2.56670684e-01 -1.31045783e+00 1.26795113e-01
-2.71552533e-01 -2.36474544e-01 6.50888532e-02 -9.97506559e-01
-1.12249374e+00 4.60084796e-01 8.72805119e-01 -3.23285535e-02
1.10023558e+00 -1.56602606e-01 1.07793272e+00 8.08329284e-01
6.36750817e-01 -1.02003264e+00 3.72579277e-01 5.77257276e-01
9.22479272e-01 -8.61523628e-01 7.91879445e-02 1.62219092e-01
-3.78542185e-01 1.08156860e+00 9.47713494e-01 -3.74304946e-03
4.15402025e-01 3.08421850e-01 -5.02985753e-02 -2.40627900e-01
-6.84739947e-01 1.05731741e-01 5.44467568e-01 7.06751168e-01
2.33782411e-01 1.12849101e-01 1.10847801e-01 3.72021407e-01
-7.64419615e-01 -1.94383860e-02 -1.65026352e-01 1.01228940e+00
-3.37133497e-01 -9.76557493e-01 -2.78216720e-01 3.46109152e-01
-3.49140137e-01 -1.15781002e-01 -5.21771729e-01 9.54919338e-01
3.81750762e-01 4.37883377e-01 5.71239665e-02 -9.67573404e-01
4.96695250e-01 -2.47892171e-01 7.08514333e-01 -1.69073299e-01
-7.15647519e-01 -3.79204035e-01 2.47892439e-01 -7.04530776e-01
-1.18051767e-01 -4.62163091e-01 -9.80910659e-01 -6.94933712e-01
-2.84904510e-01 -3.11440919e-02 7.95196414e-01 5.55234194e-01
7.83288836e-01 6.22586966e-01 7.16217160e-02 -1.44781184e+00
-4.39922035e-01 -7.64973581e-01 -2.66610887e-02 -8.71609990e-03
1.35271221e-01 -7.93660164e-01 2.19590850e-02 -3.80087197e-01] | [7.164283275604248, -0.8075579404830933] |
d15ea8f1-80b0-4818-9fb0-4c3b77501a35 | stable-table-generation-framework-for-encoder | 2206.04045 | null | https://arxiv.org/abs/2206.04045v2 | https://arxiv.org/pdf/2206.04045v2.pdf | STable: Table Generation Framework for Encoder-Decoder Models | The output structure of database-like tables, consisting of values structured in horizontal rows and vertical columns identifiable by name, can cover a wide range of NLP tasks. Following this constatation, we propose a framework for text-to-table neural models applicable to problems such as extraction of line items, joint entity and relation extraction, or knowledge base population. The permutation-based decoder of our proposal is a generalized sequential method that comprehends information from all cells in the table. The training maximizes the expected log-likelihood for a table's content across all random permutations of the factorization order. During the content inference, we exploit the model's ability to generate cells in any order by searching over possible orderings to maximize the model's confidence and avoid substantial error accumulation, which other sequential models are prone to. Experiments demonstrate a high practical value of the framework, which establishes state-of-the-art results on several challenging datasets, outperforming previous solutions by up to 15%. | ['Łukasz Garncarek', 'Dawid Jurkiewicz', 'Karolina Szyndler', 'Gabriela Pałka', 'Tomasz Dwojak', 'Łukasz Borchmann', 'Michał Turski', 'Michał Pietruszka'] | 2022-06-08 | null | null | null | null | ['knowledge-base-population', 'joint-entity-and-relation-extraction'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.79731333e-01 6.11025214e-01 -5.84495902e-01 -3.13676715e-01
-8.64384234e-01 -8.60937476e-01 3.62590343e-01 5.28421879e-01
-1.81953236e-01 1.31105042e+00 3.50262940e-01 -6.26739502e-01
-2.30489314e-01 -1.28838468e+00 -1.37238896e+00 -1.97850987e-01
-1.76600546e-01 9.16279733e-01 9.44167096e-03 -1.54036105e-01
1.29621580e-01 3.16947073e-01 -1.44239926e+00 8.91679466e-01
7.25259840e-01 1.13324606e+00 7.41269365e-02 3.49061996e-01
-5.26645303e-01 1.07615328e+00 -6.54290795e-01 -9.41422105e-01
1.77474067e-01 -2.08423227e-01 -9.12456989e-01 -2.58866876e-01
1.56780288e-01 -3.59544396e-01 -2.48119891e-01 1.05202770e+00
8.48417878e-02 -3.54539007e-02 7.43120372e-01 -9.39594865e-01
-7.43655384e-01 1.46552777e+00 -3.16636771e-01 1.27836496e-01
4.88529801e-01 -3.22778851e-01 1.61738741e+00 -8.64325643e-01
1.02493167e+00 1.27111781e+00 5.22426009e-01 1.77731812e-01
-1.40365875e+00 -3.05964053e-01 2.07947955e-01 1.36866957e-01
-1.50639987e+00 -4.00839359e-01 1.01707324e-01 -2.04716548e-01
1.31582487e+00 3.68100137e-01 3.74673724e-01 9.09414232e-01
3.08509707e-01 1.06551480e+00 6.54074907e-01 -5.04742682e-01
2.57206082e-01 3.70472103e-01 -1.70925315e-02 5.70339382e-01
8.80132437e-01 -4.48043674e-01 -1.14886868e+00 -1.95912108e-01
5.78859389e-01 -5.36070228e-01 8.00946653e-02 -3.89543772e-01
-1.39929414e+00 7.19407201e-01 1.90677792e-01 1.37078390e-01
-3.89618993e-01 -3.04248817e-02 2.17150763e-01 1.42129049e-01
7.89548233e-02 6.31131411e-01 -9.45445240e-01 7.03623286e-03
-8.23491037e-01 5.61190009e-01 1.23637390e+00 1.57347763e+00
4.94893938e-01 -3.70384067e-01 -5.38636148e-01 5.64080238e-01
2.04812273e-01 4.49008793e-01 1.23635523e-01 -5.24380445e-01
1.10483444e+00 7.22170651e-01 3.45061958e-01 -7.54458547e-01
-4.34154421e-01 -5.17636061e-01 -8.41256261e-01 -5.75827777e-01
7.33744204e-01 -2.31067777e-01 -7.12623119e-01 1.85251355e+00
2.98196495e-01 -6.41609490e-01 3.04039121e-01 2.01882586e-01
6.68459952e-01 7.12620914e-01 -9.37396884e-02 -4.07122225e-01
1.74337626e+00 -4.30137783e-01 -1.02726281e+00 -3.12846720e-01
7.20017552e-01 -3.37367922e-01 7.08959222e-01 4.17436242e-01
-1.21129715e+00 -3.36768955e-01 -1.02438879e+00 -4.00406182e-01
-6.46538258e-01 4.00375366e-01 9.02491629e-01 6.39499903e-01
-8.12234104e-01 6.38675869e-01 -6.34701073e-01 -1.86060965e-01
4.45136249e-01 5.10361195e-01 -4.80215311e-01 1.81650177e-01
-1.47463667e+00 7.53464520e-01 1.07775605e+00 4.29575652e-01
-2.60165721e-01 -6.68690503e-01 -8.14330041e-01 4.64129567e-01
9.49097753e-01 -1.07672286e+00 1.11413777e+00 -2.25076109e-01
-1.07832730e+00 5.59016645e-01 -5.55402815e-01 -8.98523271e-01
5.20856440e-01 -3.72516543e-01 -2.64345080e-01 -1.98078200e-01
1.96621776e-01 5.47516465e-01 3.09222609e-01 -1.10056365e+00
-6.84103310e-01 -2.45693624e-01 8.84024985e-03 1.84431151e-01
-2.33803079e-01 -2.62131274e-01 -6.92596614e-01 -5.35888493e-01
2.28984222e-01 -6.11120164e-01 -1.51013747e-01 -4.32216793e-01
-1.29548645e+00 -1.28338352e-01 -2.90678322e-01 -8.57673466e-01
1.79135454e+00 -1.60619307e+00 1.22903939e-02 4.59243745e-01
4.32394594e-02 -3.81259203e-01 2.53947705e-01 7.86557972e-01
2.33819976e-01 4.60833430e-01 -1.54283062e-01 -2.10907515e-02
3.31967920e-01 1.13229886e-01 -5.89948058e-01 -1.51759401e-01
4.71464396e-01 1.10358393e+00 -4.43961859e-01 -5.26533186e-01
-3.71604234e-01 -1.19836047e-01 -8.01479816e-01 1.00008145e-01
-8.41158450e-01 -1.35418013e-01 -3.62263113e-01 6.15702510e-01
5.93990266e-01 -5.68119943e-01 9.78281260e-01 -2.86943465e-01
9.56455022e-02 8.49554539e-01 -1.56548345e+00 1.39940917e+00
-1.75969929e-01 8.17628652e-02 -3.55598837e-01 -5.83928287e-01
6.73573911e-01 7.77800754e-02 7.72732571e-02 -4.85825270e-01
2.64208727e-02 1.69266015e-01 1.39380455e-01 -3.96942645e-01
8.77164125e-01 2.56117165e-01 -3.92865539e-01 3.41937333e-01
2.40540326e-01 4.25857306e-01 8.53128791e-01 4.80454534e-01
9.28494513e-01 1.84789732e-01 8.35244000e-01 -3.57186079e-01
4.17031974e-01 -6.67597651e-02 5.21323919e-01 1.25504696e+00
6.35876477e-01 1.84125900e-01 1.08768082e+00 -4.47424918e-01
-9.75537241e-01 -1.15672183e+00 -2.59983420e-01 9.08902884e-01
-2.48543113e-01 -7.25820184e-01 -7.08066106e-01 -6.62202775e-01
3.77114981e-01 1.00166035e+00 -9.78175581e-01 1.90015644e-01
-5.99138200e-01 -9.65987682e-01 3.46581489e-01 6.80514097e-01
8.17089677e-02 -1.18947327e+00 -2.18498975e-01 5.04844427e-01
-5.23988724e-01 -1.33310640e+00 -8.21181387e-02 8.08938444e-01
-8.60361874e-01 -8.62343132e-01 -2.98809037e-02 -6.46799564e-01
6.99727297e-01 -4.65279758e-01 1.58718288e+00 -2.71184564e-01
1.89629465e-01 -3.23551744e-01 -8.25229734e-02 -5.00624418e-01
-5.14708638e-01 6.46869302e-01 -1.91981569e-02 -5.52258529e-02
5.49226165e-01 -3.68550658e-01 -9.16991234e-02 8.27901140e-02
-8.81444991e-01 1.87214553e-01 8.25894594e-01 8.59626591e-01
8.23905408e-01 1.89619526e-01 6.02722645e-01 -1.56860566e+00
5.64190686e-01 -6.86472178e-01 -6.38745129e-01 6.61700606e-01
-5.61091304e-01 7.64180303e-01 8.37464690e-01 -7.93126300e-02
-9.75497365e-01 -2.60483697e-02 -7.87576213e-02 2.82788694e-01
1.88325886e-02 9.15872812e-01 -6.91621840e-01 8.07187319e-01
4.88911271e-01 4.72020179e-01 -3.75242323e-01 -5.52452743e-01
6.23234928e-01 2.79893249e-01 5.61671793e-01 -8.81589651e-01
7.56058216e-01 2.33437583e-01 1.29213780e-01 -4.62746203e-01
-1.03676486e+00 -7.85981491e-03 -8.43693316e-01 2.68279344e-01
6.43477201e-01 -1.02755201e+00 -1.05034411e+00 3.38372141e-02
-1.21358323e+00 -3.20595969e-03 -2.54654795e-01 2.19344348e-01
-5.19154131e-01 1.39255494e-01 -6.30271435e-01 -7.67581344e-01
-2.09307343e-01 -7.30613410e-01 8.19102347e-01 1.85149953e-01
-5.13764739e-01 -8.08052063e-01 -4.38794255e-01 6.10516183e-02
-8.17807466e-02 -7.44520873e-02 1.66977644e+00 -1.08515263e+00
-1.17250180e+00 3.30816470e-02 -1.25375122e-01 -1.54307231e-01
-9.40261632e-02 -1.63460314e-01 -6.40191019e-01 2.12961026e-02
-3.84966850e-01 -3.50526035e-01 8.07397008e-01 2.17120394e-01
1.33450663e+00 -7.73116589e-01 -4.82619017e-01 5.62658727e-01
1.50070894e+00 2.69937277e-01 7.48227835e-01 5.04599392e-01
5.27583778e-01 6.87697768e-01 5.70419073e-01 6.12508476e-01
5.11331499e-01 3.34761381e-01 2.95619667e-01 4.44129527e-01
4.34930801e-01 -9.34446692e-01 -1.58134550e-02 5.64933717e-01
2.60435224e-01 -8.42102528e-01 -5.42841554e-01 6.31670475e-01
-1.69539094e+00 -9.70866859e-01 -3.08398250e-02 2.29270554e+00
1.14408433e+00 6.41781569e-01 -1.59302237e-03 1.37280315e-01
5.26062369e-01 -5.28672636e-02 -4.83046025e-01 -4.01856303e-01
-2.97769189e-01 1.14249118e-01 6.62767470e-01 3.20273668e-01
-1.06089580e+00 9.59515691e-01 6.82756615e+00 9.87695336e-01
-3.05468589e-01 -4.90942985e-01 8.49115372e-01 -5.59830330e-02
-7.64662683e-01 -6.79340437e-02 -1.68868923e+00 3.53208452e-01
1.09744072e+00 -2.26433352e-01 6.10891044e-01 7.27878034e-01
-2.04543233e-01 -3.36131662e-01 -1.45230222e+00 6.17098987e-01
-1.32640660e-01 -1.68541181e+00 4.88104403e-01 1.07355729e-01
5.70158243e-01 -5.13349295e-01 -2.99635585e-02 3.64098281e-01
5.92846692e-01 -9.80909288e-01 1.11182296e+00 4.73138273e-01
9.75187123e-01 -7.84614503e-01 5.40728450e-01 4.99660224e-01
-1.09478557e+00 -1.56711981e-01 -4.01157826e-01 1.40092045e-01
2.07930014e-01 6.89619303e-01 -1.13080359e+00 7.30112255e-01
5.18965542e-01 2.80085891e-01 -5.21612525e-01 5.45594871e-01
-3.48122954e-01 3.93170953e-01 -4.66699213e-01 -5.21075010e-01
-1.31326079e-01 -5.36484532e-02 2.27352753e-01 1.27475727e+00
2.91231602e-01 -2.16731980e-01 -2.47958615e-01 9.67711210e-01
-5.95934153e-01 2.68678576e-01 -6.82714105e-01 -2.55134195e-01
8.61064196e-01 9.03086483e-01 -9.62626636e-01 -4.84292299e-01
-2.30373666e-01 6.45516038e-01 6.74989760e-01 1.69692382e-01
-6.27149880e-01 -4.35089707e-01 2.34225824e-01 5.31617403e-02
8.54109824e-01 1.06683008e-01 -8.35695863e-01 -1.02937138e+00
5.51528752e-01 -9.98109877e-01 6.05859697e-01 -6.07124925e-01
-1.10939395e+00 6.64146662e-01 3.02955210e-01 -9.75690305e-01
-7.24050879e-01 -7.51389444e-01 1.68942928e-01 9.13175821e-01
-1.11555266e+00 -9.24606144e-01 5.09932756e-01 2.29451597e-01
1.81347042e-01 -1.53289184e-01 8.47555995e-01 5.99841513e-02
-4.21625733e-01 8.58497024e-01 2.69995272e-01 4.58311081e-01
3.62022102e-01 -1.55494785e+00 9.03391600e-01 9.40944672e-01
4.51104283e-01 1.25129092e+00 6.60543501e-01 -8.67358387e-01
-1.67709792e+00 -8.90401125e-01 1.66051483e+00 -7.25233793e-01
7.31383085e-01 -8.79156888e-01 -7.49147892e-01 1.06750071e+00
5.23349904e-02 -3.56506318e-01 7.64606476e-01 5.95276594e-01
-5.60177088e-01 -2.01054573e-01 -9.76416647e-01 6.30794108e-01
1.13676643e+00 -3.73432964e-01 -5.37729800e-01 2.40741432e-01
7.73332834e-01 -6.92058265e-01 -1.00496542e+00 2.39806369e-01
9.07112062e-01 -8.17408442e-01 8.10188293e-01 -9.03292835e-01
7.92045057e-01 -3.56991202e-01 -3.27992737e-01 -9.91334319e-01
-3.78510892e-01 -5.99448144e-01 -7.63554990e-01 1.30390155e+00
1.13126540e+00 -4.32781368e-01 9.36193705e-01 3.93425524e-01
1.83647454e-01 -9.83110785e-01 -7.68489242e-01 -5.39604425e-01
-5.73401786e-02 -3.88146669e-01 9.52027559e-01 4.61416066e-01
-1.09985201e-02 6.42552257e-01 -4.16631699e-01 2.87249029e-01
4.34963882e-01 3.13290477e-01 7.81755090e-01 -1.12575102e+00
-5.69351315e-01 -1.33278251e-01 1.30558401e-01 -1.39015377e+00
-8.82691965e-02 -1.07348704e+00 1.99548360e-02 -1.74708164e+00
2.57213354e-01 -3.07400197e-01 -1.83138549e-01 4.68969166e-01
-3.23685378e-01 -1.98878095e-01 1.91669419e-01 -8.08328912e-02
-6.57028198e-01 5.14540561e-02 1.05808401e+00 -1.09051168e-01
-8.95517543e-02 -1.26608759e-01 -1.34809542e+00 3.50804985e-01
4.66288924e-01 -7.24464238e-01 -4.45966631e-01 -3.63314509e-01
1.02716005e+00 4.67478842e-01 -2.69120455e-01 -7.30162382e-01
2.06262454e-01 -1.54977679e-01 7.62589574e-01 -1.09938693e+00
1.02971390e-01 -7.13652492e-01 2.87882328e-01 3.50561827e-01
-6.75292671e-01 2.76922524e-01 2.95363426e-01 6.36635005e-01
-1.43632904e-01 -1.76092610e-01 2.02877194e-01 -2.69689262e-01
-4.22887951e-01 9.48621556e-02 -4.00404066e-01 3.07727456e-01
5.14737070e-01 2.64067724e-02 -3.36936146e-01 -1.47467211e-01
-5.81483483e-01 1.25543103e-01 -4.22336385e-02 2.84595042e-01
2.08148628e-01 -1.23603380e+00 -5.70368648e-01 4.19758618e-01
2.52936184e-01 1.61985129e-01 1.51846886e-01 4.60151702e-01
-4.87712294e-01 8.42038751e-01 -7.78199881e-02 -2.66874194e-01
-7.49646604e-01 7.59840071e-01 -2.49189511e-02 -1.06179631e+00
-3.13202590e-01 9.01933610e-01 2.32844371e-02 -3.74197692e-01
1.74762428e-01 -8.12682807e-01 -3.24283302e-01 3.14529002e-01
6.14637077e-01 -2.10524555e-02 4.73861247e-01 -9.99798775e-02
-2.60989785e-01 -1.03016444e-01 -5.33977747e-01 -1.43296629e-01
1.07516241e+00 -2.17882209e-02 -2.62925506e-01 6.12412095e-01
7.27852881e-01 4.02567416e-01 -8.61697137e-01 -3.29628378e-01
3.32237244e-01 -2.48355776e-01 -6.73054218e-01 -1.18174207e+00
-4.27947402e-01 4.65990573e-01 -2.20142439e-01 3.36834103e-01
7.76608407e-01 -1.09162614e-01 5.49893737e-01 9.47554648e-01
6.17042601e-01 -9.19429779e-01 -4.52057332e-01 7.88191199e-01
6.20265663e-01 -8.52642775e-01 8.32915306e-02 -7.26860762e-01
-6.24289334e-01 1.15739691e+00 4.46766287e-01 3.09968501e-01
4.32439148e-01 6.81154191e-01 -3.27399135e-01 1.87644348e-01
-1.29778779e+00 -6.55962154e-02 2.46039197e-01 4.99990225e-01
4.95305866e-01 7.22257644e-02 -3.95559341e-01 9.74692702e-01
-6.41144395e-01 2.81303772e-03 4.49829012e-01 8.75050485e-01
-3.64393234e-01 -1.20903337e+00 -3.21000487e-01 9.67297494e-01
-8.08694005e-01 -4.45806414e-01 -2.29459956e-01 8.64462733e-01
9.61641595e-02 7.41952658e-01 1.83125302e-01 -2.14444414e-01
1.62794232e-01 4.53926444e-01 5.26318312e-01 -5.38987517e-01
-5.02574682e-01 -2.85092704e-02 5.54971397e-01 -2.28241876e-01
1.02382831e-01 -7.25299895e-01 -1.07874334e+00 -3.63971829e-01
-8.14125240e-02 3.52699369e-01 4.23662215e-01 8.77156436e-01
3.94176245e-01 5.98279059e-01 1.64749026e-01 4.65960504e-04
-7.20223904e-01 -1.04415929e+00 -7.21274137e-01 2.27728084e-01
8.03920329e-02 -5.40119886e-01 1.81998864e-01 2.26181090e-01] | [9.626346588134766, 7.9447245597839355] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.