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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
684060a2-2aa6-4337-8135-864b36e32469 | benchmarking-self-supervised-video | 2306.06010 | null | https://arxiv.org/abs/2306.06010v1 | https://arxiv.org/pdf/2306.06010v1.pdf | Benchmarking self-supervised video representation learning | Self-supervised learning is an effective way for label-free model pre-training, especially in the video domain where labeling is expensive. Existing self-supervised works in the video domain use varying experimental setups to demonstrate their effectiveness and comparison across approaches becomes challenging with no standard benchmark. In this work, we first provide a benchmark that enables a comparison of existing approaches on the same ground. Next, we study five different aspects of self-supervised learning important for videos; 1) dataset size, 2) complexity, 3) data distribution, 4) data noise, and, 5)feature analysis. To facilitate this study, we focus on seven different methods along with seven different network architectures and perform an extensive set of experiments on 5 different datasets with an evaluation of two different downstream tasks. We present several interesting insights from this study which span across different properties of pretraining and target datasets, pretext-tasks, and model architectures among others. We further put some of these insights to the real test and propose an approach that requires a limited amount of training data and outperforms existing state-of-the-art approaches which use 10x pretraining data. We believe this work will pave the way for researchers to a better understanding of self-supervised pretext tasks in video representation learning. | ['Yogesh Singh Rawat', 'Vibhav Vineet', 'Ashlesha Kumar', 'Akash Kumar'] | 2023-06-09 | null | null | null | null | ['self-supervised-action-recognition'] | ['computer-vision'] | [ 5.87349892e-01 -1.13662459e-01 -7.24859893e-01 -7.21291661e-01
-6.16676688e-01 -6.44861639e-01 6.66272819e-01 -1.76342409e-02
-5.12274444e-01 7.38723755e-01 1.87555090e-01 -1.71632811e-01
-8.04387927e-02 -4.26214457e-01 -9.23978448e-01 -5.53853869e-01
-4.36249167e-01 4.42892492e-01 2.15329751e-01 -1.13886446e-01
1.44861452e-02 5.31413704e-02 -1.76784170e+00 7.07513213e-01
4.09809560e-01 9.95615900e-01 3.13780010e-02 4.66783226e-01
2.38465384e-01 1.23596787e+00 -5.44294834e-01 -6.65638223e-02
3.57461393e-01 -4.50464100e-01 -1.17942584e+00 4.49186385e-01
9.05800283e-01 -2.75533617e-01 -3.56914431e-01 8.67155790e-01
4.26055133e-01 4.09238875e-01 7.66944468e-01 -1.41223264e+00
-4.26154286e-01 7.24113226e-01 -2.77352691e-01 3.04143906e-01
1.52494535e-01 2.19739512e-01 8.15279782e-01 -6.65158749e-01
7.77839482e-01 1.05218112e+00 7.00492680e-01 7.40069151e-01
-1.05506110e+00 -6.71800375e-01 1.55801788e-01 3.73738348e-01
-1.23632526e+00 -6.46050274e-01 6.74367487e-01 -5.25553405e-01
7.93533385e-01 3.76149788e-02 3.53077948e-01 1.66172659e+00
-2.46520981e-01 1.05136108e+00 1.43991721e+00 -6.16213262e-01
1.36825353e-01 4.30976421e-01 4.69206929e-01 8.35292637e-01
3.86977941e-02 3.02352369e-01 -6.49562418e-01 2.09920213e-01
6.67886019e-01 7.38892704e-02 -3.20284873e-01 -6.20236516e-01
-1.19120932e+00 8.83463919e-01 2.07138985e-01 4.44879860e-01
2.05937505e-01 1.86736092e-01 8.25755298e-01 7.47080982e-01
5.27334988e-01 3.40336472e-01 -7.71279514e-01 -9.10523757e-02
-1.07736468e+00 -1.02567174e-01 7.86678672e-01 1.23355126e+00
7.77864337e-01 2.67069727e-01 -1.79430753e-01 7.80730963e-01
1.40102476e-01 4.37998697e-02 9.16608810e-01 -9.38739181e-01
5.98410428e-01 2.15073660e-01 -3.23049515e-01 -5.33713877e-01
-3.13194215e-01 -5.52051067e-01 -8.62270057e-01 -1.86104681e-02
3.65492165e-01 -3.44652891e-01 -1.07945287e+00 1.75717235e+00
-1.82375565e-01 5.66216409e-01 2.55959034e-01 7.08068728e-01
1.16337955e+00 5.36714017e-01 2.26841033e-01 -3.03058356e-01
1.15463328e+00 -1.62137806e+00 -4.87886697e-01 -3.11184734e-01
1.09771585e+00 -6.05905592e-01 1.08052623e+00 3.15963477e-01
-8.08567345e-01 -9.41023350e-01 -9.36790287e-01 1.76383227e-01
-5.63849986e-01 5.06904542e-01 8.31067085e-01 5.81067801e-01
-1.21662843e+00 9.54040885e-01 -6.33455575e-01 -8.97412419e-01
6.46364450e-01 3.54449242e-01 -6.18118286e-01 -1.65536806e-01
-1.15918446e+00 8.45228076e-01 6.36746526e-01 -4.13436919e-01
-1.39339125e+00 -6.52023733e-01 -9.53363895e-01 -6.13633543e-02
7.25110054e-01 -5.51001549e-01 1.40148711e+00 -1.44469273e+00
-1.35166025e+00 1.21757448e+00 8.87635276e-02 -6.44188583e-01
3.51324409e-01 -3.61762911e-01 -2.86817491e-01 2.07119584e-01
-2.29211599e-02 9.08071101e-01 1.01243913e+00 -1.28139150e+00
-6.44456148e-01 4.33045663e-02 3.12226772e-01 2.85564452e-01
-5.08107483e-01 -1.26551660e-02 -4.46253419e-01 -8.53091776e-01
-3.78049910e-01 -1.06156862e+00 4.99358121e-03 -2.83313900e-01
-2.77188390e-01 -2.84989119e-01 8.72428596e-01 -4.67558131e-02
1.16927135e+00 -2.28061271e+00 -1.63323004e-02 -2.36817703e-01
1.63256571e-01 5.82214952e-01 -3.53631914e-01 4.64459687e-01
-5.73807240e-01 1.79807946e-01 -6.21548668e-02 -6.17695391e-01
-2.61760980e-01 2.24459693e-01 -2.07509562e-01 3.25900167e-01
1.00730419e-01 7.82252431e-01 -9.40392494e-01 -6.70436859e-01
1.86901867e-01 2.53249347e-01 -3.81725758e-01 4.15074289e-01
-1.73151284e-01 2.95172036e-01 -1.67984068e-01 6.73127532e-01
2.41763145e-01 -4.19967115e-01 1.51716366e-01 -4.22074795e-01
1.36868149e-01 1.39376968e-01 -9.91825640e-01 1.95169401e+00
-2.21873596e-01 8.89194369e-01 -2.11950138e-01 -1.54345560e+00
6.61054134e-01 4.04544979e-01 5.74275970e-01 -3.42312694e-01
3.81971270e-01 -1.90096274e-01 -1.30305514e-01 -7.54874647e-01
3.37128609e-01 -1.45905599e-01 2.67073035e-01 6.90624177e-01
8.22848856e-01 3.40559065e-01 6.81319296e-01 1.51831225e-01
1.15267730e+00 4.39393044e-01 2.36422777e-01 -4.18656528e-01
2.96713829e-01 1.04813889e-01 2.74131477e-01 8.69713366e-01
-3.40814680e-01 5.36042094e-01 2.87434250e-01 -5.79349995e-01
-7.12889552e-01 -5.36063612e-01 6.07674606e-02 1.47535956e+00
-1.57022066e-02 -7.35342979e-01 -7.48422921e-01 -1.26236773e+00
-1.85194418e-01 4.40480262e-01 -9.21529710e-01 -1.30689248e-01
-4.42228973e-01 -7.61714280e-01 5.50256610e-01 7.01898813e-01
6.42055809e-01 -1.10111272e+00 -5.32940865e-01 -1.87776819e-01
-1.38593465e-01 -1.33958936e+00 -1.31064683e-01 5.82988024e-01
-1.24150515e+00 -1.28434289e+00 -5.93609035e-01 -1.09452665e+00
7.24530280e-01 5.63594818e-01 1.36739075e+00 2.96171188e-01
1.18447132e-01 6.62629664e-01 -7.48789310e-01 -4.41564262e-01
-3.52157086e-01 4.11353618e-01 8.82053599e-02 -8.05019140e-02
4.76397306e-01 -4.31424737e-01 -3.56205732e-01 4.30234045e-01
-1.03960776e+00 9.36921164e-02 6.27223790e-01 8.21641028e-01
4.94385749e-01 2.66548753e-01 3.40555221e-01 -1.45809019e+00
3.34514290e-01 -7.51769543e-01 -2.11001918e-01 3.69256586e-01
-5.27003348e-01 -1.07162103e-01 6.16592169e-01 -5.94676018e-01
-9.22249258e-01 2.80335903e-01 9.55720544e-02 -6.56068861e-01
-5.21056175e-01 5.48414826e-01 1.15284428e-01 -5.30277193e-02
9.77500379e-01 8.57664719e-02 3.35021503e-02 -4.60806131e-01
3.03641051e-01 5.21541238e-01 2.36256555e-01 -4.79275137e-01
7.30265975e-01 5.89213550e-01 -7.28992447e-02 -7.93533027e-01
-1.34028256e+00 -5.76002717e-01 -9.35578704e-01 -2.23154679e-01
5.48410594e-01 -1.17476285e+00 -1.27116740e-01 4.04575437e-01
-7.27440059e-01 -8.53669822e-01 -5.12204468e-01 5.59646070e-01
-7.53481865e-01 4.07400578e-01 -6.45774901e-01 -4.29523259e-01
-6.47460222e-02 -1.11859834e+00 8.65097642e-01 -1.07608354e-02
-2.27993041e-01 -1.32663786e+00 -9.17423982e-03 4.68822628e-01
3.75064999e-01 6.94147870e-02 5.09723842e-01 -9.61887658e-01
-3.01660866e-01 8.14284682e-02 -1.15849189e-01 6.98621869e-01
1.52787939e-01 -1.18050314e-01 -1.16017687e+00 -5.78732133e-01
2.33818181e-02 -1.03690922e+00 1.24082553e+00 3.26109022e-01
1.14792633e+00 -1.83239937e-01 -3.92895520e-01 7.02093422e-01
1.43881583e+00 -1.22540437e-01 5.80084264e-01 4.79213744e-01
8.15647602e-01 7.74430394e-01 9.41184223e-01 -8.77259672e-03
2.59139419e-01 4.72562730e-01 4.25063133e-01 -3.20912927e-01
-3.79083514e-01 -1.10624045e-01 5.35287440e-01 8.33904624e-01
-1.90061584e-01 -4.71646249e-01 -6.01507604e-01 4.59190726e-01
-1.86854506e+00 -1.03310812e+00 1.54073671e-01 2.04160953e+00
9.25921500e-01 1.80934086e-01 3.19663972e-01 2.39594385e-01
6.16323113e-01 2.64239222e-01 -2.68834710e-01 3.18160132e-02
-7.80045763e-02 1.45173058e-01 4.66212124e-01 1.30199984e-01
-1.78137302e+00 1.20202971e+00 6.97223186e+00 8.36241841e-01
-1.27302825e+00 2.29857445e-01 7.14220047e-01 -6.00819429e-03
3.46093625e-01 7.44264647e-02 -8.81279409e-01 3.85961086e-01
1.01726043e+00 2.48549998e-01 1.91339076e-01 1.14613140e+00
-6.03310810e-03 -9.98218283e-02 -1.65122724e+00 1.17312789e+00
6.19265199e-01 -1.28864968e+00 1.57143235e-01 -3.72328043e-01
8.51168752e-01 3.32510144e-01 -6.94308281e-02 5.94435632e-01
3.42218399e-01 -1.02530611e+00 4.83292073e-01 3.36638361e-01
1.01176929e+00 -2.47315645e-01 7.66276777e-01 2.16042519e-01
-8.68036926e-01 -2.03508660e-01 -2.87805587e-01 -1.94800377e-01
2.35873478e-04 2.04110190e-01 -6.65774763e-01 5.05202293e-01
7.53190637e-01 1.21004140e+00 -8.33815694e-01 1.05975533e+00
-2.61978149e-01 1.07135069e+00 -8.70243832e-02 2.13514313e-01
3.95473331e-01 1.79124251e-01 1.23065263e-01 1.69216835e+00
1.30925374e-02 -1.13946445e-01 3.33087087e-01 1.85262188e-01
-2.40557805e-01 1.12361103e-01 -7.30072618e-01 -6.39506504e-02
2.26999909e-01 1.18851554e+00 -8.67653906e-01 -7.01904058e-01
-5.28984606e-01 7.67683506e-01 4.10678118e-01 5.15184462e-01
-7.33028173e-01 -4.81090602e-03 3.16028655e-01 2.29719713e-01
1.77576542e-01 -1.62940204e-01 1.74283534e-01 -1.35789096e+00
-3.01928550e-01 -1.15887403e+00 6.51572049e-01 -7.02304065e-01
-1.36818147e+00 6.44972265e-01 3.98329973e-01 -1.52515590e+00
-4.54406321e-01 -7.50378728e-01 -4.33047146e-01 4.15331312e-02
-1.80386734e+00 -1.17272747e+00 -4.45164680e-01 6.97594464e-01
9.91256595e-01 -7.11244941e-01 8.75773966e-01 4.35063601e-01
-7.01379478e-01 6.71231508e-01 8.63429718e-03 3.04994851e-01
1.15245032e+00 -9.98766422e-01 1.06821030e-01 6.19318187e-01
5.78597665e-01 5.04191101e-01 4.13254321e-01 -3.71409267e-01
-1.22757185e+00 -1.38180113e+00 4.93448704e-01 -6.05233014e-01
6.53516829e-01 -2.49055490e-01 -7.03276277e-01 1.17797434e+00
4.35321599e-01 9.34032202e-02 1.02320838e+00 1.08383931e-01
-2.81631768e-01 -1.49673194e-01 -7.99005210e-01 2.82685399e-01
1.34737122e+00 -3.33576620e-01 -4.82563555e-01 7.15580821e-01
5.48986912e-01 -3.36859226e-01 -6.66142702e-01 6.34820640e-01
3.75023246e-01 -1.03527641e+00 7.28437781e-01 -1.00526333e+00
4.39941406e-01 -6.81905225e-02 -9.05073136e-02 -1.30650532e+00
-3.89083385e-01 -5.98807871e-01 -3.24048966e-01 1.24877608e+00
3.57535064e-01 -2.02827036e-01 9.86269236e-01 1.39529884e-01
-1.88105583e-01 -9.18239295e-01 -4.10169780e-01 -9.63768184e-01
-1.20027088e-01 -5.48601925e-01 -2.12415936e-03 1.39254999e+00
-1.19783074e-01 7.92382777e-01 -6.04631245e-01 -3.23380381e-01
5.87560177e-01 -1.44988865e-01 9.60214436e-01 -1.16067982e+00
-2.07513049e-01 -8.50981027e-02 -3.88987035e-01 -1.21271348e+00
3.43892664e-01 -8.97537351e-01 -2.42363393e-01 -1.28990495e+00
3.77441168e-01 -5.50460696e-01 -3.93842101e-01 7.66823411e-01
1.23236366e-01 5.27675569e-01 1.83114186e-01 4.84735966e-01
-1.15443480e+00 2.36087352e-01 9.13073838e-01 -2.29871020e-01
-9.20747407e-03 -2.87133968e-03 -6.79384589e-01 9.10863936e-01
1.06104207e+00 -6.05776429e-01 -9.35831845e-01 -8.12772632e-01
-1.80370912e-01 -4.37461376e-01 1.04301378e-01 -1.18667328e+00
1.89849228e-01 -5.53678237e-02 3.04574639e-01 -3.06322098e-01
2.43604347e-01 -8.51345539e-01 -3.53968114e-01 2.70148098e-01
-5.49871624e-01 -1.87042356e-01 9.69656557e-02 4.32661802e-01
-4.12153184e-01 -6.68309212e-01 8.16506684e-01 -4.13994819e-01
-1.17532408e+00 3.00571501e-01 -2.65146047e-01 1.95334837e-01
1.25481093e+00 -2.43492812e-01 -3.01475048e-01 -5.51942348e-01
-9.43129301e-01 2.25744739e-01 3.59679639e-01 5.71700335e-01
4.11496967e-01 -1.16860163e+00 -7.34403729e-01 4.32707444e-02
1.95628449e-01 -2.99523592e-01 1.37922838e-01 6.58051610e-01
-3.27140212e-01 5.10425568e-01 -3.01043481e-01 -7.44035482e-01
-1.55462790e+00 7.50755131e-01 1.62431493e-01 -4.20691371e-01
-2.41443336e-01 8.79672050e-01 1.42377481e-01 -2.63943046e-01
7.22474098e-01 -1.32242337e-01 -6.34534538e-01 3.26915920e-01
4.27621663e-01 2.73636669e-01 -4.83348966e-02 -7.81172991e-01
-8.98295864e-02 5.51515520e-01 -2.66058922e-01 3.04135054e-01
1.34986198e+00 -5.96031472e-02 3.24712932e-01 6.01401329e-01
1.30726147e+00 -5.40671170e-01 -1.35925543e+00 -4.28936332e-01
1.69495475e-02 -2.07690507e-01 -5.83810620e-02 -5.87335825e-01
-1.32193542e+00 7.84491420e-01 7.19912887e-01 2.28652567e-01
1.15424252e+00 7.29411766e-02 3.99816364e-01 6.00807250e-01
2.54701167e-01 -1.29605007e+00 5.42213559e-01 5.43721735e-01
7.06651211e-01 -1.54832351e+00 3.00734997e-01 -5.53270578e-01
-7.93420672e-01 1.10703611e+00 8.12732875e-01 -1.14028975e-01
7.69479573e-01 1.94374248e-01 2.11013645e-01 -2.91062474e-01
-8.70151460e-01 -4.16332752e-01 2.77543366e-01 7.69750237e-01
6.66138649e-01 -4.66856629e-01 -4.91575226e-02 3.93958151e-01
-4.48163413e-02 3.19570154e-01 3.95119488e-01 1.31286836e+00
-1.73218548e-01 -1.24431551e+00 -3.01983748e-02 5.95844328e-01
-5.48413634e-01 -1.14277154e-01 -3.51669163e-01 9.58384752e-01
3.33020806e-01 9.52739418e-01 -7.01960083e-03 -5.42480707e-01
2.97535151e-01 8.03437606e-02 5.44824064e-01 -1.01430917e+00
-4.65550363e-01 -5.30781001e-02 3.45880479e-01 -5.58702707e-01
-1.21972656e+00 -5.66717207e-01 -8.32854271e-01 -2.78057158e-02
-4.91555631e-01 9.96303782e-02 4.52349335e-01 9.40474987e-01
3.42522502e-01 4.25749779e-01 5.76037228e-01 -9.56105411e-01
-4.09106642e-01 -1.27208853e+00 -4.07734901e-01 6.04051530e-01
1.14553779e-01 -8.82875323e-01 -3.24707627e-01 7.20051706e-01] | [9.227495193481445, 1.6222912073135376] |
ccef70e5-7132-4a95-9553-8ca660f5d526 | topological-interpretability-for-deep | 2305.08642 | null | https://arxiv.org/abs/2305.08642v1 | https://arxiv.org/pdf/2305.08642v1.pdf | Topological Interpretability for Deep-Learning | With the increasing adoption of AI-based systems across everyday life, the need to understand their decision-making mechanisms is correspondingly accelerating. The level at which we can trust the statistical inferences made from AI-based decision systems is an increasing concern, especially in high-risk systems such as criminal justice or medical diagnosis, where incorrect inferences may have tragic consequences. Despite their successes in providing solutions to problems involving real-world data, deep learning (DL) models cannot quantify the certainty of their predictions. And are frequently quite confident, even when their solutions are incorrect. This work presents a method to infer prominent features in two DL classification models trained on clinical and non-clinical text by employing techniques from topological and geometric data analysis. We create a graph of a model's prediction space and cluster the inputs into the graph's vertices by the similarity of features and prediction statistics. We then extract subgraphs demonstrating high-predictive accuracy for a given label. These subgraphs contain a wealth of information about features that the DL model has recognized as relevant to its decisions. We infer these features for a given label using a distance metric between probability measures, and demonstrate the stability of our method compared to the LIME interpretability method. This work demonstrates that we may gain insights into the decision mechanism of a DL model, which allows us to ascertain if the model is making its decisions based on information germane to the problem or identifies extraneous patterns within the data. | ['Georgia Tourassi', 'Lynne Penberthy', 'Heidi A. Hanson', 'Adam Spannaus'] | 2023-05-15 | null | null | null | null | ['medical-diagnosis'] | ['medical'] | [ 4.13253844e-01 7.61701703e-01 -1.39806300e-01 -5.77072859e-01
-4.79939669e-01 -4.04718548e-01 5.69543958e-01 9.15944755e-01
-3.70289870e-02 5.92785120e-01 2.21742928e-01 -6.33977115e-01
-6.82896972e-01 -8.60299587e-01 -3.15758020e-01 -6.09941602e-01
-3.03398669e-01 7.96754718e-01 -1.60170987e-01 -2.00846270e-02
3.88552755e-01 6.70004249e-01 -1.15514612e+00 6.71631932e-01
7.92942464e-01 1.07219994e+00 -3.15787315e-01 6.55054569e-01
4.74119931e-02 1.08352005e+00 -4.52073783e-01 -6.39309943e-01
1.71067119e-02 -3.63415897e-01 -9.95033205e-01 -2.18644291e-01
3.05776656e-01 -8.60565975e-02 -1.83055103e-01 9.97736037e-01
6.89201653e-02 -3.42868507e-01 1.23722649e+00 -1.31352973e+00
-4.67868239e-01 4.75652009e-01 -2.82602936e-01 2.17267275e-01
4.33524102e-01 1.09427229e-01 1.31204653e+00 -5.62470376e-01
7.21981108e-01 1.22559917e+00 1.07833683e+00 2.38796189e-01
-1.43756056e+00 -3.13293785e-01 -1.16913952e-02 4.06095773e-01
-1.39412045e+00 -3.30572844e-01 5.77378452e-01 -8.90971482e-01
7.14739442e-01 4.03836042e-01 6.63707495e-01 8.11244667e-01
7.65459836e-01 3.72889251e-01 1.15366912e+00 -3.93126398e-01
3.91191512e-01 1.95023626e-01 9.35551301e-02 9.90371048e-01
5.48477471e-01 3.24803814e-02 -3.02087367e-01 -4.28828150e-01
3.47881526e-01 9.76984873e-02 -1.12502970e-01 -2.06899986e-01
-1.22605240e+00 8.47035646e-01 6.57131732e-01 3.87647390e-01
-4.95030969e-01 -4.10055667e-02 1.70734361e-01 2.34256133e-01
6.48493826e-01 7.86164820e-01 -3.74631971e-01 1.01941429e-01
-8.99367511e-01 2.01102659e-01 8.76761198e-01 4.48609978e-01
4.95177895e-01 -5.89918196e-01 -1.89532340e-02 3.62187564e-01
3.36438626e-01 1.59881469e-02 3.43029857e-01 -7.15885937e-01
4.39578779e-02 1.03036952e+00 -2.21100330e-01 -1.76735079e+00
-7.00382888e-01 -4.89570111e-01 -9.06156659e-01 2.74988711e-01
4.76629823e-01 1.37990355e-01 -7.03546703e-01 1.51808274e+00
2.12498456e-01 -3.06890588e-02 -6.71750307e-02 6.02461994e-01
4.99130130e-01 7.40908161e-02 1.59415394e-01 4.46451008e-02
1.27103889e+00 6.49003126e-03 -4.99611765e-01 -1.50866345e-01
1.12002325e+00 -3.62335414e-01 6.25407815e-01 4.68807250e-01
-5.17968476e-01 -1.96918562e-01 -1.11858630e+00 2.31588259e-03
-4.67217892e-01 -9.45882350e-02 6.03595793e-01 3.44381213e-01
-9.33316052e-01 1.05000257e+00 -8.07078481e-01 -5.19141436e-01
7.58204043e-01 4.16064143e-01 -3.93806279e-01 -4.58139722e-04
-1.07361352e+00 1.33954215e+00 2.45329022e-01 1.77323669e-01
-3.23914170e-01 -5.57031810e-01 -6.22966886e-01 5.32177323e-03
8.06733221e-02 -8.38726103e-01 8.01524699e-01 -8.29632282e-01
-5.61911523e-01 1.07327318e+00 -3.56593914e-02 -5.05571663e-01
7.97822893e-01 3.43441367e-01 -4.95689929e-01 1.35663196e-01
2.05638170e-01 4.99537408e-01 6.82844400e-01 -1.09464586e+00
-7.46123254e-01 -6.25658333e-01 -8.70065019e-02 -6.59952685e-02
-2.26125903e-02 -2.21063256e-01 3.29581827e-01 -4.57245916e-01
4.49834228e-01 -9.69871700e-01 -3.93608063e-01 4.14364487e-01
-8.65787804e-01 -2.87713557e-01 5.56878150e-01 -4.82294619e-01
1.34352899e+00 -1.90945804e+00 -9.44259167e-02 6.03429496e-01
1.03577220e+00 -1.51043892e-01 3.21817875e-01 3.86410177e-01
-9.21440646e-02 4.96152729e-01 -5.21351039e-01 3.22245322e-02
-6.54762387e-02 1.57048911e-01 -1.47855043e-01 7.20678568e-01
4.82443154e-01 6.86739385e-01 -9.93805766e-01 -7.21832812e-01
9.87903178e-02 2.49998063e-01 -5.69532156e-01 8.78043398e-02
-3.15307453e-02 3.40594172e-01 -6.47314787e-01 6.06077969e-01
2.24117875e-01 -5.70496500e-01 4.42180097e-01 -2.69171596e-01
3.79455388e-01 3.35281640e-01 -7.84619093e-01 1.02269137e+00
-2.44684383e-01 9.33508396e-01 -3.72690141e-01 -1.27857661e+00
8.36478293e-01 8.43308866e-02 5.52822649e-01 -3.25955778e-01
1.06090233e-01 2.31532022e-01 3.84304494e-01 -5.32320499e-01
1.11103006e-01 -5.61388791e-01 -4.99496497e-02 5.55600286e-01
-3.08561623e-01 1.29687414e-02 -2.13309243e-01 2.74039924e-01
1.50362146e+00 -2.52683252e-01 6.44794524e-01 -5.04163921e-01
2.54013270e-01 3.36121887e-01 5.06775022e-01 6.31134152e-01
-2.17199743e-01 5.74688494e-01 9.16325390e-01 -9.03794527e-01
-1.00273657e+00 -1.01680517e+00 -6.59637690e-01 4.67132151e-01
-1.98216885e-01 -3.89667988e-01 -5.28389633e-01 -9.08379734e-01
3.96203101e-01 9.83330905e-01 -1.22962940e+00 -5.32921731e-01
-8.76811594e-02 -7.14708805e-01 5.42520821e-01 3.18304092e-01
-7.62975216e-02 -8.01415086e-01 -6.65994108e-01 1.21028155e-01
-5.66090979e-02 -9.77867424e-01 8.73534903e-02 1.00985199e-01
-9.62505102e-01 -1.44921827e+00 -1.04840733e-01 -4.88257557e-01
1.12161100e+00 -4.04417068e-01 1.27291274e+00 3.13774228e-01
-5.35591960e-01 2.39216760e-01 -1.16873175e-01 -5.86937785e-01
-8.89597356e-01 -1.59459561e-01 2.45312765e-01 3.11063584e-02
8.45271587e-01 -4.35520560e-01 -8.29947352e-01 2.71548599e-01
-7.39886463e-01 1.38989970e-01 5.46592653e-01 7.27571905e-01
5.01430094e-01 3.36530916e-02 5.40516198e-01 -1.18240571e+00
8.75309765e-01 -8.90123427e-01 3.55750881e-02 3.59617412e-01
-9.55534518e-01 2.64911711e-01 6.18797600e-01 -6.54219911e-02
-4.53378171e-01 -5.75120673e-02 8.56034905e-02 -1.90205321e-01
-3.18285525e-01 7.50518382e-01 3.26714098e-01 2.39641741e-01
9.35767293e-01 -2.42258966e-01 7.72564784e-02 -3.63290370e-01
1.12755336e-01 8.61540616e-01 2.99169928e-01 -4.81168240e-01
3.82806927e-01 5.12909651e-01 3.51054698e-01 -6.18718147e-01
-1.10452342e+00 -1.24024719e-01 -9.66220081e-01 -2.21473306e-01
8.78489792e-01 -2.99052179e-01 -9.59263146e-01 -1.32494524e-01
-1.04265440e+00 1.63626492e-01 -3.44918728e-01 4.03238684e-01
-5.47230661e-01 1.36383265e-01 -2.71054894e-01 -6.79248631e-01
-8.74100626e-02 -9.46610510e-01 9.15220797e-01 -2.32041761e-01
-1.09010553e+00 -1.45846367e+00 5.39607964e-02 2.57557809e-01
7.80604631e-02 6.57480001e-01 1.59921658e+00 -1.15233612e+00
-2.17764705e-01 -4.82468128e-01 -3.11129451e-01 1.18920840e-01
2.62570262e-01 1.14584498e-01 -9.72796142e-01 -7.02286735e-02
-5.45476377e-02 -4.30937558e-02 5.72737098e-01 3.75504017e-01
1.22823513e+00 -5.54557920e-01 -7.28432536e-01 2.27135301e-01
1.18178570e+00 7.61238933e-02 3.11222732e-01 2.38092527e-01
6.34201765e-01 9.75402117e-01 4.24258322e-01 2.75205165e-01
3.49975973e-01 5.26397109e-01 3.30871195e-01 -1.30398214e-01
1.15092836e-01 -4.32472825e-01 1.66367721e-02 5.18642068e-01
2.20041350e-02 2.82500852e-02 -1.37769473e+00 4.88524228e-01
-1.91807818e+00 -7.34913111e-01 -1.29064217e-01 2.23305321e+00
7.94749379e-01 4.50444788e-01 -1.25910103e-01 3.17287296e-01
5.01555502e-01 -3.80069703e-01 -7.74995625e-01 -7.28576243e-01
1.48191392e-01 -8.21535662e-02 1.89843759e-01 3.29124331e-01
-8.40873420e-01 4.59941059e-01 6.87765837e+00 5.74080408e-01
-9.89115000e-01 -3.23086858e-01 1.33797133e+00 2.41709545e-01
-3.75349760e-01 -8.50422233e-02 -3.44252229e-01 3.42791587e-01
1.06625569e+00 -3.61835003e-01 2.38711536e-02 6.84421003e-01
3.05605233e-01 -1.84560850e-01 -1.92923784e+00 5.96795976e-01
-6.07246459e-02 -1.55533230e+00 1.63580820e-01 3.62499475e-01
2.03140527e-01 -1.86694264e-01 1.25819724e-02 -1.47473305e-01
5.24806917e-01 -1.51811898e+00 5.94324052e-01 9.20202196e-01
9.04358447e-01 -6.43018663e-01 8.70602965e-01 4.58447009e-01
-5.98551333e-01 -5.06819598e-02 -2.37005755e-01 -2.45883510e-01
-3.22822541e-01 8.12702000e-01 -1.57655549e+00 1.27008021e-01
4.33100969e-01 8.34827542e-01 -7.36662805e-01 6.48731291e-01
-1.29775867e-01 6.23633265e-01 -2.43247703e-01 -1.46492139e-01
3.56489383e-02 -1.51493371e-01 4.62569416e-01 1.27735507e+00
2.77759224e-01 1.82139575e-01 -6.04779869e-02 1.07651007e+00
7.80706704e-02 1.37906075e-01 -1.10847878e+00 -1.99377462e-01
1.15235925e-01 1.13341653e+00 -9.46079314e-01 -2.86472976e-01
-6.37658164e-02 5.45007467e-01 4.02823418e-01 -7.75742158e-02
-4.02696759e-01 -2.87432134e-01 7.97890067e-01 3.90163422e-01
-3.45467687e-01 -2.98207924e-02 -8.90510082e-01 -8.71160746e-01
-1.42850295e-01 -8.43873739e-01 5.72575331e-01 -7.62548625e-01
-1.52828825e+00 6.86111093e-01 -2.26658776e-01 -1.27068126e+00
-4.53474998e-01 -8.41016233e-01 -5.11144996e-01 7.14492440e-01
-9.80691791e-01 -9.22515154e-01 9.24789906e-02 2.35895634e-01
-6.61629140e-02 -4.93910536e-02 1.04252267e+00 -2.07030594e-01
-3.62221152e-01 4.20650989e-01 1.02782205e-01 2.04892352e-01
4.76554543e-01 -1.28419256e+00 2.32827514e-01 3.13055128e-01
2.86655873e-01 6.03485942e-01 9.44704473e-01 -6.14178479e-01
-9.93316352e-01 -7.85081089e-01 1.21408153e+00 -1.01184535e+00
9.74943936e-01 -1.86711103e-01 -9.67181861e-01 6.95497155e-01
-4.73877996e-01 -9.59682744e-03 1.06412804e+00 3.69179338e-01
-3.34237903e-01 2.29143891e-02 -1.55690634e+00 5.72819352e-01
9.32456553e-01 -5.25933266e-01 -8.47120643e-01 4.90617156e-01
3.10007721e-01 1.17773630e-01 -1.04644966e+00 3.98637474e-01
6.77281857e-01 -1.02808416e+00 7.28741229e-01 -1.21048963e+00
8.06835830e-01 -6.40461817e-02 -6.19643442e-02 -1.49801993e+00
-4.59762514e-01 -1.74352005e-01 1.94108635e-01 7.04263806e-01
7.33688116e-01 -6.64876640e-01 7.46199787e-01 1.02929544e+00
2.39058346e-01 -1.22561932e+00 -8.81336987e-01 -2.00351745e-01
2.42368981e-01 -6.59252644e-01 4.85182554e-01 1.28912830e+00
3.75073045e-01 2.15527356e-01 1.11155756e-01 1.90710127e-01
6.41030252e-01 2.46647149e-01 2.93680340e-01 -1.71670878e+00
-4.68615703e-02 -5.87918222e-01 -1.00169826e+00 -1.47011176e-01
-9.17965267e-03 -1.21772456e+00 -2.27211341e-01 -1.71131420e+00
2.14961588e-01 -7.42166758e-01 -3.98164093e-01 6.10207617e-01
-1.39385713e-02 1.44117802e-01 -4.72918525e-02 4.08908844e-01
-2.87266880e-01 3.21548022e-02 1.02537274e+00 -1.78674310e-01
1.78539142e-01 7.35025033e-02 -9.16459322e-01 1.17965865e+00
6.47453964e-01 -7.43664682e-01 -1.30809695e-01 3.47037874e-02
5.92245758e-01 -2.85125356e-02 4.37227637e-01 -8.15320432e-01
1.68229192e-01 -2.36861184e-01 6.84701800e-01 1.73801146e-02
1.88553110e-02 -9.13384497e-01 1.04502060e-01 7.91183293e-01
-8.15169275e-01 3.17581110e-02 5.39673530e-02 6.76252604e-01
-7.84371421e-03 1.61941275e-02 6.99751258e-01 1.33795636e-02
-4.13481355e-01 9.62050036e-02 -4.02854413e-01 1.18624710e-01
1.01061213e+00 -2.62988895e-01 -2.27112085e-01 -4.89430904e-01
-1.11263943e+00 7.61456341e-02 3.69427353e-01 2.01212615e-01
7.58101761e-01 -1.12778115e+00 -9.57613289e-01 1.22129790e-01
3.91263396e-01 -2.60341048e-01 -8.11736211e-02 1.00039601e+00
-6.57174349e-01 4.68405604e-01 -1.32515669e-01 -7.18271911e-01
-1.16946173e+00 4.28949982e-01 4.42323655e-01 -2.89872408e-01
-7.43958294e-01 5.11321247e-01 2.87638515e-01 -2.70014793e-01
-6.74588084e-02 -5.07941723e-01 -1.96172267e-01 2.64025658e-01
3.33059609e-01 3.26962143e-01 1.28512770e-01 -5.86125433e-01
-5.56031823e-01 3.73882592e-01 -1.52904645e-01 2.61908770e-01
1.37898779e+00 6.52873665e-02 -1.67897001e-01 5.95951259e-01
1.07215738e+00 -1.62547976e-01 -8.19127083e-01 -1.06080562e-01
3.80602121e-01 -3.88578802e-01 1.61934182e-01 -1.04119051e+00
-7.19928443e-01 1.08367026e+00 4.34175462e-01 5.65154791e-01
8.15376222e-01 1.86355859e-01 3.08218181e-01 4.58678663e-01
2.39468127e-01 -8.95927072e-01 -2.47819036e-01 2.87658386e-02
9.34620023e-01 -1.39095414e+00 2.74339586e-01 -2.94688225e-01
-6.18133426e-01 1.36392391e+00 6.66966587e-02 -1.04783744e-01
8.91574323e-01 1.03407122e-01 7.95631558e-02 -5.54590762e-01
-9.17059720e-01 1.38104394e-01 6.50705695e-01 5.80030084e-01
5.09914577e-01 5.04861653e-01 -1.51515797e-01 4.86743063e-01
-4.55592930e-01 -5.46955429e-02 5.18895507e-01 5.77182531e-01
-3.79478216e-01 -6.84109509e-01 -1.99550688e-01 1.06443930e+00
-3.99181724e-01 -1.33961111e-01 -8.12074721e-01 6.35926187e-01
1.22511923e-01 8.37733746e-01 2.52086431e-01 -7.30315685e-01
6.97107166e-02 1.54592961e-01 1.75101295e-01 -6.25492334e-01
-2.61544287e-01 -5.41515231e-01 3.02484632e-01 -5.55908203e-01
-2.39464045e-02 -7.80103981e-01 -1.39123321e+00 -5.02045572e-01
-4.58888244e-03 5.83835021e-02 5.91576040e-01 1.18059909e+00
5.62708259e-01 4.82779920e-01 4.51432228e-01 -2.35798627e-01
-4.79993999e-01 -6.34644568e-01 -5.76116264e-01 5.86241722e-01
3.32984149e-01 -5.95619023e-01 -4.66004997e-01 -9.64284018e-02] | [8.613893508911133, 5.636249542236328] |
4c233cdb-0708-4c3b-89bd-9814d9cc5d64 | a-communication-efficient-adaptive-algorithm | 2301.08869 | null | https://arxiv.org/abs/2301.08869v2 | https://arxiv.org/pdf/2301.08869v2.pdf | A Communication-Efficient Adaptive Algorithm for Federated Learning under Cumulative Regret | We consider the problem of online stochastic optimization in a distributed setting with $M$ clients connected through a central server. We develop a distributed online learning algorithm that achieves order-optimal cumulative regret with low communication cost measured in the total number of bits transmitted over the entire learning horizon. This is in contrast to existing studies which focus on the offline measure of simple regret for learning efficiency. The holistic measure for communication cost also departs from the prevailing approach that \emph{separately} tackles the communication frequency and the number of bits in each communication round. | ['Qing Zhao', 'Kobi Cohen', 'Tamir Gabay', 'Sudeep Salgia'] | 2023-01-21 | null | null | null | null | ['stochastic-optimization'] | ['methodology'] | [-3.30981284e-01 4.83532310e-01 -2.48589098e-01 -5.94093502e-01
-1.29042172e+00 -8.59225571e-01 3.60129513e-02 2.14304939e-01
-8.93168628e-01 1.08485866e+00 -6.61787316e-02 -3.24978471e-01
-8.28521669e-01 -8.33229482e-01 -9.58370745e-01 -7.49895334e-01
-3.56724083e-01 3.63199592e-01 -4.30655062e-01 2.81504035e-01
1.12560377e-01 2.28598759e-01 -1.03059924e+00 -1.02775581e-01
5.81150472e-01 1.69312549e+00 -8.37007761e-02 7.26086438e-01
-2.88096458e-01 8.43190908e-01 -7.60020494e-01 -1.19094634e+00
6.74884677e-01 -6.89694583e-01 -8.75892341e-01 2.25525483e-01
2.82956157e-02 -7.45039880e-01 -5.11548042e-01 1.03688037e+00
7.89133549e-01 -7.93326870e-02 2.98700958e-01 -9.63748157e-01
2.02208888e-02 1.15511668e+00 -1.53852597e-01 2.28082508e-01
2.79263794e-01 -1.81062043e-01 1.21039057e+00 -1.90391615e-01
7.34336555e-01 1.06093919e+00 3.49840343e-01 4.34955388e-01
-1.01913393e+00 -4.44298834e-01 2.63362914e-01 4.78169888e-01
-1.14712250e+00 -3.35304499e-01 4.24213618e-01 -6.24685511e-02
7.42648005e-01 3.84844810e-01 7.13987052e-01 4.45074737e-01
3.31993341e-01 6.98298991e-01 9.81051743e-01 -5.41364193e-01
7.91478395e-01 3.16423506e-01 -1.30881935e-01 4.88008201e-01
4.24612403e-01 6.30930662e-02 -9.83395338e-01 -2.90610660e-02
5.17449915e-01 3.60268801e-02 -1.16500445e-01 -3.75580311e-01
-3.56290102e-01 8.88776720e-01 2.62503047e-02 7.79936314e-02
-4.18497711e-01 4.32168186e-01 3.32010180e-01 1.29952276e+00
6.66757643e-01 -1.59515709e-01 -5.82933009e-01 -7.51062632e-01
-7.48682499e-01 1.29928350e-01 1.52884030e+00 1.18280220e+00
6.59651518e-01 -1.40641659e-01 -1.81113899e-01 4.29293931e-01
1.54100403e-01 5.46590447e-01 -1.63965076e-01 -1.47654879e+00
7.97783256e-01 5.33418119e-01 4.83486623e-01 -2.00749427e-01
-7.92858824e-02 -8.76794815e-01 -5.39956987e-01 2.89231032e-01
6.41340256e-01 -9.56921160e-01 3.16513717e-01 1.52051401e+00
4.46322411e-01 -7.08351851e-01 1.22737594e-01 8.44700158e-01
1.15072802e-01 3.65008593e-01 -7.31375158e-01 -7.72030652e-01
5.09841800e-01 -1.00388849e+00 -8.86549652e-01 2.09842578e-01
6.39371634e-01 -5.42187870e-01 5.07916868e-01 7.20717967e-01
-1.58998644e+00 4.72518444e-01 -7.41119266e-01 3.05948466e-01
-1.88557252e-01 -2.91737437e-01 6.95696592e-01 1.32131267e+00
-1.16646910e+00 8.83683205e-01 -3.89003158e-01 -2.43572928e-02
5.66475451e-01 3.95317972e-01 3.89155716e-01 -3.21404904e-01
-4.58279938e-01 8.29282343e-01 7.92277977e-02 7.28902733e-03
-1.12302864e+00 -7.20711768e-01 -5.13206683e-02 2.42699832e-01
6.60447657e-01 -7.38029838e-01 1.67588902e+00 -1.15909159e+00
-2.10930872e+00 4.12806958e-01 1.77181557e-01 -6.87471867e-01
1.16841745e+00 -2.97727764e-01 4.47398484e-01 -5.95160089e-02
-7.35158443e-01 -3.22325230e-01 4.78752553e-01 -9.12628174e-01
-8.90555441e-01 -5.04192472e-01 5.41098297e-01 5.45082152e-01
-7.80672967e-01 -1.66492224e-01 -1.88333362e-01 -9.29733142e-02
-3.26415569e-01 -4.75348383e-01 -3.72083932e-01 1.82865456e-01
5.50605208e-02 -2.27424383e-01 3.80790442e-01 -3.09623361e-01
1.10497582e+00 -1.89599669e+00 3.12611073e-01 1.67519674e-01
3.00523303e-02 -4.77950454e-01 1.22371055e-01 9.35822487e-01
6.32825136e-01 -6.16988279e-02 2.82105148e-01 -5.08971334e-01
5.56687117e-01 1.80946618e-01 -1.25060335e-01 7.12735891e-01
-9.26400840e-01 6.01086617e-01 -6.90127790e-01 -3.68316263e-01
-9.89474729e-02 -5.88373467e-03 -5.98090947e-01 3.84747416e-01
-2.13280693e-01 1.70671139e-02 -4.38798547e-01 5.11525869e-01
4.24993306e-01 -4.02353317e-01 5.36271155e-01 7.34591246e-01
-1.36232346e-01 2.15177178e-01 -1.57786369e+00 1.66504908e+00
-7.81919658e-01 4.29369479e-01 8.17705750e-01 -9.84715521e-01
2.85304368e-01 5.16617715e-01 6.73531830e-01 -6.23701930e-01
2.23096475e-01 3.51484120e-01 -5.79880059e-01 -1.80238187e-01
-1.29076242e-01 -8.06998760e-02 6.70936182e-02 1.01604748e+00
2.50316679e-01 2.57879287e-01 -1.42062679e-02 2.51668185e-01
1.20272255e+00 -3.63929898e-01 8.67956281e-02 -2.37406865e-01
5.72973378e-02 -3.93183827e-01 4.14560109e-01 1.24533045e+00
-1.44863918e-01 -1.75947890e-01 6.63851917e-01 -3.51709187e-01
-1.06140113e+00 -1.03981650e+00 1.85016513e-01 1.54231799e+00
1.45568386e-01 -2.82218397e-01 -9.49707985e-01 -7.38446593e-01
3.47326756e-01 5.35685420e-01 -3.26000422e-01 1.74876541e-01
2.73410738e-01 -3.50360453e-01 1.29006281e-01 -3.71452384e-02
3.31448078e-01 -5.71148694e-01 -7.10217774e-01 3.23809057e-01
9.23869535e-02 -7.29179859e-01 -6.90572262e-01 3.02699536e-01
-9.80017662e-01 -9.91291225e-01 -6.89446926e-01 -1.19290300e-01
5.29231608e-01 -6.93947896e-02 1.06355214e+00 -3.63564968e-01
-3.78656417e-01 1.31746840e+00 -3.08172286e-01 -8.63166213e-01
2.56767333e-01 -7.06922710e-02 -1.64462730e-01 9.65178311e-02
3.26305889e-02 -4.99197870e-01 -1.08292186e+00 -1.28517419e-01
-6.97651565e-01 -5.49838305e-01 5.35302222e-01 5.11173189e-01
1.96612313e-01 1.83055505e-01 7.21086979e-01 -7.71006942e-01
7.58817077e-01 -4.84104544e-01 -9.00974035e-01 4.78598803e-01
-1.09901321e+00 -8.66981074e-02 7.74808288e-01 -1.03489332e-01
-1.14739335e+00 1.70488041e-02 2.83390224e-01 1.44097120e-01
4.12614346e-01 1.67574704e-01 1.50442615e-01 -5.59306085e-01
4.26833957e-01 6.17123067e-01 3.23338844e-02 -3.21165204e-01
4.28985655e-01 6.75748229e-01 -4.08186495e-01 -5.12807548e-01
4.46839482e-01 4.32707489e-01 1.34617999e-01 -5.12619734e-01
-9.01793063e-01 -2.24764168e-01 -1.20531596e-01 -5.88867188e-01
1.29502630e-04 -6.65835261e-01 -1.66492224e+00 1.88877419e-01
-9.59677041e-01 -2.96273023e-01 -8.63985121e-01 5.80933511e-01
-1.13456571e+00 2.80266970e-01 -6.87426805e-01 -1.72904015e+00
-7.16024399e-01 -3.03870767e-01 3.99515361e-01 2.17198625e-01
3.95624816e-01 -1.13620508e+00 2.64943708e-02 3.92616540e-01
8.77147377e-01 -6.00211807e-02 3.53113979e-01 -2.10622460e-01
-1.05241418e+00 -3.69242460e-01 -2.18994245e-02 4.13275033e-01
-2.87644833e-01 -8.11276495e-01 -5.78381181e-01 -8.43365669e-01
1.91031933e-01 -7.29901135e-01 4.62541431e-01 4.86793339e-01
1.08932376e+00 -1.15952647e+00 2.45907947e-01 4.51066852e-01
2.08203650e+00 1.85573682e-01 6.28473982e-02 1.60047859e-01
-2.31771603e-01 5.89745939e-01 3.66031259e-01 1.49413323e+00
2.04747453e-01 2.56767184e-01 5.49979210e-01 4.32564288e-01
5.71204782e-01 1.34976193e-01 4.71976757e-01 5.51943958e-01
-1.02703422e-01 -3.45484555e-01 -5.33911407e-01 4.41758394e-01
-2.07148123e+00 -8.76646280e-01 7.51667261e-01 2.59640527e+00
1.01365745e+00 2.69682817e-02 3.39744180e-01 9.29747596e-02
3.22555065e-01 -1.30763883e-02 -7.36903846e-01 -6.05960608e-01
1.56687140e-01 4.39737365e-02 1.27407455e+00 6.84221983e-01
-2.62586653e-01 3.20730746e-01 7.62297201e+00 1.04439986e+00
-5.82280636e-01 5.05403996e-01 8.07804286e-01 -1.17225218e+00
-3.34774107e-01 -1.53970182e-01 -8.91025186e-01 3.65697354e-01
1.31440580e+00 -6.96990609e-01 1.32688963e+00 1.12438524e+00
1.36446133e-01 -2.79216737e-01 -1.34106886e+00 1.12912166e+00
-3.57403874e-01 -1.28644133e+00 -1.51637375e-01 3.49381447e-01
8.46756935e-01 7.69901350e-02 -2.55726539e-02 1.42115057e-02
6.21086538e-01 -7.80791938e-01 5.85720420e-01 8.84834230e-01
5.96237779e-01 -1.25271499e+00 6.76342368e-01 7.82948136e-01
-7.18959749e-01 -6.18175745e-01 -3.14658165e-01 -3.77414674e-01
1.03017025e-01 8.91882122e-01 -3.88263881e-01 7.64178574e-01
9.66341853e-01 -1.08636089e-01 1.08975030e-01 1.21618509e+00
3.06448996e-01 6.44251585e-01 -6.05723619e-01 -5.64191341e-01
2.20161811e-01 -5.61524570e-01 4.31831747e-01 1.17570591e+00
1.90778285e-01 2.66449630e-01 -3.31219807e-02 4.78002816e-01
-3.64456445e-01 3.99642855e-01 -4.27857280e-01 -2.00773031e-01
5.71306527e-01 9.95685458e-01 -1.81793079e-01 -2.15941161e-01
-4.49206352e-01 1.03302372e+00 5.36746383e-01 4.05109107e-01
-3.18199098e-01 -6.70050979e-01 3.68256152e-01 -1.51892141e-01
4.98283446e-01 -2.38373429e-01 -3.41839641e-01 -6.98750257e-01
5.74276507e-01 -2.38089830e-01 5.21722019e-01 2.83785492e-01
-1.28038228e+00 -1.04075663e-01 -3.78546029e-01 -7.80297756e-01
-1.38179570e-01 -2.15913281e-01 -5.65849245e-01 4.33515251e-01
-1.48926580e+00 -3.87194902e-01 9.93387923e-02 6.94422841e-01
4.40759927e-01 -5.01136947e-03 8.24882507e-01 2.81184554e-01
-4.82918918e-01 1.08651984e+00 1.15633166e+00 -4.47147876e-01
2.12480247e-01 -1.47958040e+00 -7.43559837e-01 3.62426579e-01
-3.26897770e-01 1.83978617e-01 6.96683705e-01 -2.75622476e-02
-2.07051229e+00 -7.57469952e-01 1.02251983e+00 2.08308045e-02
4.25041050e-01 -3.92873138e-01 1.43086031e-01 5.14958262e-01
4.45785046e-01 -6.20419942e-02 1.01147020e+00 3.03876162e-01
-3.71380709e-02 -8.77102077e-01 -1.47636759e+00 1.66176081e-01
1.24754894e+00 -3.33047688e-01 3.60820323e-01 8.85684609e-01
5.75589240e-01 -3.08832139e-01 -1.31106400e+00 -2.62241006e-01
6.66542351e-01 -1.00485218e+00 4.49120730e-01 -4.45634991e-01
5.50866313e-02 7.57135630e-01 -3.02963257e-01 -1.06432974e+00
2.21111387e-01 -1.28782690e+00 -4.86927837e-01 1.12017179e+00
4.04719621e-01 -8.14851582e-01 1.28460526e+00 6.92649901e-01
3.57604027e-01 -1.18465555e+00 -1.77933264e+00 -1.19856989e+00
1.57746524e-01 -1.60834059e-01 7.30246902e-02 2.98163831e-01
4.23824072e-01 -2.26235643e-01 -2.87448674e-01 -3.26110214e-01
1.27325964e+00 2.14090347e-01 4.00680691e-01 -7.96646655e-01
-7.98771441e-01 -4.86085773e-01 -3.75112593e-02 -1.06589139e+00
-3.24374080e-01 -5.24839938e-01 -1.73269406e-01 -1.48742509e+00
4.70953166e-01 -1.50282502e-01 -3.74253571e-01 -8.28165039e-02
6.26718462e-01 -5.24025798e-01 2.65439928e-01 -1.09586917e-01
-1.57135689e+00 7.91744649e-01 9.64473665e-01 1.96001753e-01
3.86835076e-02 4.99244750e-01 -6.73456490e-01 2.65942007e-01
6.10297561e-01 -5.45885921e-01 -4.05580610e-01 -4.62958246e-01
5.46556652e-01 7.37063706e-01 -1.58067793e-01 -6.19761825e-01
6.27041221e-01 -4.83948588e-01 -1.98985450e-02 -3.27370316e-01
7.71601945e-02 -1.27682769e+00 -3.55342813e-02 7.77980864e-01
-5.24571061e-01 -2.54403055e-01 -4.68829930e-01 1.03415918e+00
1.55248106e-01 -4.56591398e-01 5.73593199e-01 -3.75692904e-01
2.73723632e-01 3.55880678e-01 -4.35774356e-01 5.29294461e-02
1.46866095e+00 3.21726613e-02 -1.52291507e-01 -1.02573740e+00
-7.31361985e-01 4.39198464e-01 4.04323786e-02 -2.51330525e-01
2.76045263e-01 -1.00322175e+00 -5.98844707e-01 -6.23440742e-01
-5.11943340e-01 -1.80769235e-01 3.86336923e-01 6.56611681e-01
-2.02762872e-01 6.91001475e-01 2.06963450e-01 -1.54755088e-02
-1.01198030e+00 2.85910636e-01 5.61082006e-01 -3.33292216e-01
-4.20870364e-01 8.99882138e-01 -7.03831017e-01 -8.53900462e-02
1.21041322e+00 3.07146877e-01 5.06169617e-01 6.46615475e-02
3.91684592e-01 1.07381177e+00 -9.79171973e-03 5.69151640e-01
-3.11958957e-02 7.76068717e-02 2.73136869e-02 -4.28102076e-01
1.58999300e+00 -6.80502236e-01 -1.10654958e-01 6.03538454e-01
1.24907291e+00 -2.41437450e-01 -1.54718792e+00 -6.77144408e-01
1.47344619e-01 -4.95825678e-01 1.69957995e-01 -1.11347246e+00
-1.09893274e+00 2.92653739e-01 9.21488702e-01 2.53294289e-01
1.08878112e+00 -3.17259878e-02 5.53800881e-01 8.60375464e-01
8.16800117e-01 -1.78275764e+00 8.69601313e-03 2.57152855e-01
7.11016953e-01 -9.31893289e-01 1.54263988e-01 -2.55698949e-01
-1.99505746e-01 1.12454665e+00 1.22741006e-01 -9.16243717e-02
7.72785485e-01 3.04522783e-01 -2.86866575e-01 -2.33470593e-02
-1.06362116e+00 -9.07445028e-02 -2.85901874e-01 8.43744949e-02
3.27094495e-01 1.27811611e-01 -1.07656848e+00 8.78341079e-01
-1.41096339e-01 1.72306299e-01 4.08640057e-01 1.39857960e+00
-8.88908446e-01 -1.20078707e+00 -1.54798269e-01 6.29777193e-01
-8.62152815e-01 3.11588049e-01 -5.60057163e-01 1.05573878e-01
-3.04132253e-01 1.15757322e+00 -6.62013292e-02 -1.63377691e-02
2.20551908e-01 6.72205538e-02 6.98294759e-01 -3.58632952e-01
-7.34238744e-01 -2.18745336e-01 4.80296500e-02 -8.21886718e-01
-2.46132106e-01 -5.23069322e-01 -8.47577214e-01 -6.23358965e-01
-3.25677931e-01 4.72496331e-01 9.91259396e-01 8.84511471e-01
3.08387756e-01 2.20140636e-01 1.36329997e+00 -3.06557596e-01
-1.44968033e+00 -7.11577296e-01 -1.10274661e+00 -3.90102185e-04
3.14524770e-01 1.03155598e-01 -7.17154205e-01 -4.51276690e-01] | [4.656428813934326, 3.4431564807891846] |
dd794aba-528a-4561-9c7b-2d6c55f3143d | meta-album-multi-domain-meta-dataset-for-few-1 | 2302.08909 | null | https://arxiv.org/abs/2302.08909v1 | https://arxiv.org/pdf/2302.08909v1.pdf | Meta-Album: Multi-domain Meta-Dataset for Few-Shot Image Classification | We introduce Meta-Album, an image classification meta-dataset designed to facilitate few-shot learning, transfer learning, meta-learning, among other tasks. It includes 40 open datasets, each having at least 20 classes with 40 examples per class, with verified licences. They stem from diverse domains, such as ecology (fauna and flora), manufacturing (textures, vehicles), human actions, and optical character recognition, featuring various image scales (microscopic, human scales, remote sensing). All datasets are preprocessed, annotated, and formatted uniformly, and come in 3 versions (Micro $\subset$ Mini $\subset$ Extended) to match users' computational resources. We showcase the utility of the first 30 datasets on few-shot learning problems. The other 10 will be released shortly after. Meta-Album is already more diverse and larger (in number of datasets) than similar efforts, and we are committed to keep enlarging it via a series of competitions. As competitions terminate, their test data are released, thus creating a rolling benchmark, available through OpenML.org. Our website https://meta-album.github.io/ contains the source code of challenge winning methods, baseline methods, data loaders, and instructions for contributing either new datasets or algorithms to our expandable meta-dataset. | ['Phan Anh Vu', 'Joaquin Vanschoren', 'Haozhe Sun', 'Jan N van Rijn', 'Felix Mohr', 'Mike Huisman', 'Isabelle Guyon', 'Sergio Escalera', 'Dustin Carrión-Ojeda', 'Ihsan Ullah'] | 2023-02-16 | meta-album-multi-domain-meta-dataset-for-few | https://meta-album.github.io/paper/Meta-Album.pdf | https://meta-album.github.io/paper/Meta-Album.pdf | neurips-2022-9 | ['optical-character-recognition', 'few-shot-image-classification'] | ['computer-vision', 'computer-vision'] | [ 3.58799756e-01 -1.57896698e-01 -3.82109523e-01 -5.65792397e-02
-1.03454828e+00 -5.41610241e-01 6.91017866e-01 -1.93540740e-03
-2.27960840e-01 8.07448685e-01 2.33617891e-02 -2.51207083e-01
-1.00110814e-01 -7.30660200e-01 -7.75106966e-01 -5.45399547e-01
-4.20335889e-01 3.64987046e-01 3.54339004e-01 -1.89326540e-01
3.63449961e-01 -5.66686206e-02 -2.09957504e+00 7.24861622e-01
4.63446647e-01 9.59708333e-01 2.89528489e-01 8.65862668e-01
-9.49147195e-02 9.05054688e-01 -5.07397115e-01 -3.72967541e-01
2.36734837e-01 -3.01007479e-01 -8.61912251e-01 -1.52453393e-01
5.25004447e-01 -1.57139618e-02 -1.57839954e-01 8.22152972e-01
7.27466583e-01 1.92738608e-01 7.50856042e-01 -1.56907046e+00
-9.29643452e-01 4.07928973e-01 -4.61975545e-01 2.93265134e-01
1.20303765e-01 4.46989357e-01 1.08651996e+00 -1.01053464e+00
9.72775698e-01 9.84798491e-01 7.65800774e-01 7.35263228e-01
-1.07966316e+00 -5.68361342e-01 -1.41653776e-01 4.99668747e-01
-1.19436049e+00 -5.67180991e-01 2.42769256e-01 -6.41633868e-01
1.17416275e+00 5.25475800e-01 5.01765311e-01 1.64878964e+00
1.41461909e-01 8.30302596e-01 1.15980506e+00 -3.94535363e-01
6.03272498e-01 4.10491303e-02 1.37367010e-01 6.66266441e-01
-8.59105662e-02 7.40295202e-02 -3.48223835e-01 -2.86725074e-01
3.69923770e-01 1.28485695e-01 -1.08441785e-01 -3.26909721e-01
-1.40638983e+00 6.25504732e-01 2.36941159e-01 1.26212940e-01
3.31721343e-02 -1.62740722e-02 6.51749730e-01 4.50965017e-01
6.75658107e-01 4.64602202e-01 -7.81886458e-01 -3.86662006e-01
-4.40972924e-01 2.96790183e-01 8.50009859e-01 1.14184785e+00
1.07215083e+00 -1.88829675e-01 7.87376836e-02 1.32932198e+00
-2.90130079e-01 1.96853817e-01 8.27783585e-01 -1.11393905e+00
2.81654984e-01 4.53693420e-01 6.39455020e-02 -5.00052154e-01
-3.68871421e-01 2.06050258e-02 -8.03698421e-01 2.26135001e-01
6.72093108e-02 -1.04624040e-01 -1.11804771e+00 1.38066685e+00
3.43743294e-01 7.50259101e-01 7.18354667e-03 5.24414301e-01
1.25411022e+00 8.11019301e-01 1.33596346e-01 -6.47094799e-03
1.32460797e+00 -1.07933331e+00 -6.10003434e-02 -1.20503090e-01
6.29826725e-01 -6.58102751e-01 1.25211644e+00 4.99491245e-01
-7.23474205e-01 -6.18724227e-01 -9.69612062e-01 1.19031675e-01
-1.09352016e+00 -4.54360962e-01 7.20644176e-01 3.49343807e-01
-7.41940081e-01 9.14214134e-01 -5.78028917e-01 -6.40229046e-01
6.06661141e-01 -1.31019711e-01 -4.01084036e-01 -8.57465193e-02
-1.18652964e+00 8.56388450e-01 2.68674076e-01 -4.62880969e-01
-1.15465045e+00 -1.05181909e+00 -9.94449556e-01 -3.60789657e-01
2.17468873e-01 -4.67648417e-01 1.20027518e+00 -6.93893313e-01
-1.16428661e+00 1.28422451e+00 3.09054226e-01 -1.84612468e-01
4.66849774e-01 7.49781877e-02 -5.56611001e-01 -2.29971424e-01
1.96170896e-01 8.59900713e-01 7.16842353e-01 -1.17127168e+00
-8.30809355e-01 -2.68920302e-01 2.00207546e-01 -2.53273398e-01
-1.71767160e-01 2.11961139e-02 -4.33037966e-01 -6.71530962e-01
-5.18360734e-01 -8.92348111e-01 -2.97729671e-01 -6.14163093e-03
-2.50733286e-01 -2.15853527e-01 8.01894426e-01 -1.91523656e-01
8.81313741e-01 -2.23749661e+00 1.31574199e-01 -5.51304221e-01
1.65068731e-01 3.19500327e-01 -7.22802699e-01 7.06738830e-01
-2.51279682e-01 3.57866973e-01 -4.45451200e-01 -4.23856735e-01
-4.13409509e-02 3.18793923e-01 -1.90517619e-01 4.39150304e-01
1.44075781e-01 1.04400086e+00 -1.01420248e+00 -4.78408635e-01
4.83683407e-01 1.18733056e-01 -2.47988701e-01 4.57133306e-03
-3.51567656e-01 5.28000332e-02 -1.77027836e-01 1.21350121e+00
6.29934192e-01 -4.40500379e-01 -2.30373353e-01 -5.06812446e-02
-1.95529938e-01 -2.60093212e-01 -1.17472172e+00 1.83021319e+00
-6.31774843e-01 5.75942278e-01 -6.56826124e-02 -9.95805800e-01
6.79418206e-01 -7.73720769e-03 7.30152369e-01 -6.55990422e-01
-4.44261581e-02 1.04558738e-02 -3.33611697e-01 -6.94843531e-01
3.55569601e-01 1.46576285e-01 -2.54523367e-01 5.40651858e-01
3.73457044e-01 -1.62948728e-01 4.99350339e-01 4.93784845e-02
1.20104301e+00 3.45624089e-01 4.79649663e-01 -1.44385219e-01
7.39421844e-02 2.48707041e-01 4.99320328e-01 7.81512499e-01
-4.71397996e-01 7.52757728e-01 3.46109509e-01 -8.24134529e-01
-1.34389651e+00 -1.17988431e+00 -5.09249568e-01 1.62685871e+00
1.61653720e-02 -4.69224781e-01 -5.65990508e-01 -4.69447196e-01
1.56958759e-01 4.19556081e-01 -1.07427847e+00 3.26134004e-02
-7.10686594e-02 -9.63569820e-01 5.09806573e-01 4.54726309e-01
3.34990144e-01 -1.34420204e+00 -6.85187459e-01 1.98816806e-02
-2.97866147e-02 -8.15556586e-01 -5.71461879e-02 3.73421133e-01
-6.08912647e-01 -1.48116469e+00 -6.49380803e-01 -7.42557347e-01
8.33258927e-02 2.95921773e-01 1.13010120e+00 4.18394059e-02
-9.21855211e-01 3.78687292e-01 -8.09004962e-01 -7.24723458e-01
-3.07970285e-01 2.16070786e-01 7.01760277e-02 -3.31610203e-01
2.49185964e-01 -5.30477524e-01 -3.76703620e-01 4.86173779e-01
-9.32397425e-01 4.61121686e-02 3.34883839e-01 9.95829165e-01
6.65160358e-01 -2.99161375e-01 5.53254366e-01 -1.14277256e+00
2.52620429e-01 -9.60176289e-01 -3.88535827e-01 4.14271444e-01
-3.63508880e-01 -5.32796741e-01 4.69983816e-01 -4.29737419e-01
-6.56035602e-01 -2.27172956e-01 -5.86602949e-02 -5.27922869e-01
-5.27500570e-01 4.14131343e-01 -3.90728423e-03 4.12415490e-02
1.07830119e+00 1.53021455e-01 6.24448014e-03 -5.83208859e-01
4.79876846e-01 8.79002035e-01 5.09045482e-01 -5.22790492e-01
4.91727322e-01 4.66816455e-01 -2.61514902e-01 -1.03684783e+00
-6.98537648e-01 -6.25899732e-01 -5.78973770e-01 -1.48209661e-01
5.72628796e-01 -9.38842356e-01 -2.90771991e-01 7.42139578e-01
-7.58333027e-01 -7.58037448e-01 -3.53493661e-01 2.02526316e-01
-7.05133140e-01 1.27009854e-01 -7.08667874e-01 -4.16056961e-01
-3.55140358e-01 -1.03099513e+00 1.14540660e+00 2.09667943e-02
-3.68230641e-02 -9.86506283e-01 4.31310862e-01 2.96315372e-01
4.32791740e-01 7.19631732e-01 8.12286437e-01 -6.05574787e-01
-1.87203556e-01 -1.39957786e-01 -1.81678280e-01 2.98230499e-01
2.00473890e-01 4.37911272e-01 -1.30328166e+00 -2.55238801e-01
-5.77590585e-01 -9.85215664e-01 1.14706767e+00 2.42785275e-01
1.45947242e+00 -1.03317469e-01 -4.51271862e-01 7.15298533e-01
1.47489631e+00 -2.34390926e-02 7.36954808e-01 5.93949378e-01
4.99095887e-01 6.49104714e-01 5.16233742e-01 8.51021826e-01
2.44435623e-01 3.85222375e-01 7.15621829e-01 9.73553434e-02
-1.65421322e-01 1.24389745e-01 2.66322225e-01 6.74186230e-01
-2.56679237e-01 -1.14874348e-01 -1.05026853e+00 5.65094411e-01
-1.82411313e+00 -1.32056975e+00 -1.06433153e-01 2.12626839e+00
7.82945335e-01 1.54278695e-03 2.25800306e-01 -1.96401998e-01
6.60374224e-01 2.83719569e-01 -8.96128178e-01 -2.56215096e-01
-1.79527611e-01 4.25444067e-01 2.41421759e-01 2.29266182e-01
-1.39903724e+00 9.15178537e-01 6.92723799e+00 1.03976977e+00
-1.04560578e+00 3.19727451e-01 8.07910204e-01 -2.53053963e-01
-1.81096360e-01 -1.43350124e-01 -6.20528758e-01 6.06533110e-01
1.08368063e+00 -2.06709757e-01 4.01398897e-01 1.14530730e+00
-2.07484812e-01 -8.93489048e-02 -1.09917843e+00 8.76676440e-01
8.79523158e-02 -1.78112590e+00 -2.16932476e-01 -3.07107382e-02
8.75205696e-01 9.35765684e-01 3.03090084e-02 8.86045218e-01
4.19844031e-01 -9.77313340e-01 6.80438578e-01 4.35316682e-01
1.14234531e+00 -5.85599422e-01 5.49563527e-01 4.27691340e-01
-1.21457958e+00 -2.10335732e-01 -7.13838816e-01 -2.14665323e-01
-1.94719106e-01 2.78524995e-01 -2.80229419e-01 5.15028179e-01
1.08563280e+00 1.06838810e+00 -8.67914438e-01 1.11427057e+00
2.29021296e-01 3.51600140e-01 7.20265359e-02 -3.99353802e-02
8.93088132e-02 -2.50057341e-03 2.64321923e-01 1.28216994e+00
8.36795792e-02 3.44007462e-02 4.10655677e-01 6.08924747e-01
-9.37093645e-02 1.33690700e-01 -8.35399508e-01 9.05731544e-02
4.13138896e-01 1.56939626e+00 -5.66845775e-01 -5.90419412e-01
-6.36603355e-01 7.81113565e-01 2.76725560e-01 1.37562409e-01
-1.00182652e+00 -6.95222795e-01 9.48297083e-01 -6.31093383e-02
1.74970716e-01 2.58453161e-01 -8.33428502e-02 -1.34394598e+00
-3.98421645e-01 -9.71535861e-01 6.73350394e-01 -8.46218646e-01
-1.69408619e+00 5.51460147e-01 1.74242109e-01 -1.52371323e+00
-5.25500216e-02 -1.03661036e+00 -6.43749535e-01 4.22113776e-01
-1.35917652e+00 -1.28401625e+00 -5.64639330e-01 4.71923769e-01
9.31751609e-01 -3.60342830e-01 1.11567163e+00 3.65802050e-01
-8.56456816e-01 3.74072403e-01 3.24054211e-01 3.15551870e-02
8.79464507e-01 -1.16028190e+00 6.65748537e-01 2.34856173e-01
9.91688147e-02 2.91656971e-01 4.64046329e-01 -4.54933941e-01
-1.40229523e+00 -1.38135421e+00 3.22914422e-01 -8.45283151e-01
1.05918503e+00 -4.79807794e-01 -8.21623921e-01 7.57829070e-01
2.67284691e-01 4.82986003e-01 9.76043940e-01 5.82080670e-02
-5.42040825e-01 -3.54121961e-02 -1.06630075e+00 3.21366966e-01
1.19806111e+00 -3.43381375e-01 -4.63403165e-01 5.64024806e-01
6.98562264e-01 -4.21883792e-01 -1.09578753e+00 4.62137401e-01
7.18687117e-01 -1.00383210e+00 8.55989754e-01 -9.24181700e-01
8.40512991e-01 7.14125903e-03 -4.50819284e-01 -1.28337264e+00
-4.80312645e-01 -2.91099757e-01 -1.71485469e-01 1.01685846e+00
5.78662753e-01 -5.31116009e-01 7.98576891e-01 1.17264099e-01
-4.34674203e-01 -1.00452960e+00 -8.50064397e-01 -1.06685627e+00
2.99052894e-01 -6.49776876e-01 5.39474845e-01 1.26278257e+00
1.29546508e-01 1.26420200e-01 -4.18060124e-01 -3.70916933e-01
7.26113856e-01 2.40299478e-01 1.04542756e+00 -1.28726399e+00
-4.41812754e-01 -6.78330004e-01 -3.67400795e-01 -3.22343081e-01
-3.70471254e-02 -1.01410174e+00 -5.79869375e-02 -1.54287851e+00
4.84415680e-01 -4.81523246e-01 -4.61392075e-01 8.86755943e-01
-2.10696757e-01 6.32361770e-01 6.47444576e-02 4.50109631e-01
-7.99228549e-01 3.01580757e-01 9.74567115e-01 -4.73360240e-01
8.08347017e-02 -8.91067982e-02 -4.68609929e-01 6.00465417e-01
9.09236789e-01 -1.47073910e-01 -1.95146203e-01 -2.58508027e-01
-5.67613915e-02 -3.43671650e-01 5.18596470e-01 -1.17343736e+00
-1.42553732e-01 -5.76272547e-01 4.06242877e-01 -2.94809729e-01
4.46579009e-01 -3.02231342e-01 1.51871935e-01 4.84700054e-01
-2.15445817e-01 4.08695340e-02 2.88971096e-01 4.05683368e-01
-1.19631756e-02 -2.46126413e-01 9.47837353e-01 -6.76861405e-01
-1.49299634e+00 6.84534192e-01 -1.54344872e-01 3.58135074e-01
1.43912792e+00 -2.68295676e-01 -7.25536644e-01 8.72095600e-02
-7.83565998e-01 9.95044634e-02 6.94838285e-01 7.87797928e-01
4.81074154e-01 -1.15913212e+00 -8.49458337e-01 1.02925682e-02
8.27471972e-01 -2.75712073e-01 5.49172401e-01 6.65561199e-01
-4.69906926e-01 8.74760374e-02 -5.26698232e-01 -4.60815877e-01
-9.89836991e-01 7.98895597e-01 2.79607922e-01 5.18190265e-02
-5.63658834e-01 8.92612755e-01 -1.85315907e-01 -7.98916698e-01
1.94594726e-01 -1.31522968e-01 4.81072953e-03 2.41489902e-01
9.40268099e-01 8.29793334e-01 8.33024308e-02 -2.92382717e-01
-3.42352569e-01 4.08620745e-01 4.04641256e-02 2.52125651e-01
1.70943034e+00 2.03182399e-01 -1.87602386e-01 8.00717950e-01
1.32742679e+00 -4.65769917e-01 -1.23588181e+00 -1.75994605e-01
-5.19736931e-02 -2.99851507e-01 -4.17434156e-01 -8.00175428e-01
-8.08543622e-01 7.95051277e-01 6.91888809e-01 2.43486091e-01
9.00772750e-01 2.33991832e-01 5.39912820e-01 3.90892684e-01
6.72051489e-01 -1.09606481e+00 2.23925352e-01 6.82652175e-01
7.78390944e-01 -1.59504664e+00 5.75855449e-02 9.12214443e-02
-5.99813700e-01 9.30993855e-01 9.75292206e-01 2.06752936e-03
7.46048331e-01 2.83954054e-01 1.15463100e-02 -8.64252299e-02
-1.23049641e+00 -2.86451727e-01 -6.33517429e-02 9.16354716e-01
4.08410668e-01 1.54480413e-01 1.61518812e-01 5.02599239e-01
2.82721743e-02 4.58681248e-02 5.91042817e-01 1.17712200e+00
-7.07363605e-01 -1.07019269e+00 -1.97977185e-01 7.02766955e-01
-1.19665831e-01 -2.24694431e-01 -2.99396843e-01 7.97590733e-01
6.49215281e-01 6.52297616e-01 2.56244630e-01 -5.89885473e-01
4.15732533e-01 1.99557934e-03 3.80269766e-01 -8.12300444e-01
-2.63635218e-01 -3.19577903e-01 1.14906020e-01 -6.94474280e-01
-3.00172865e-01 -6.38483703e-01 -8.13868105e-01 -5.18921793e-01
8.31385143e-03 -7.26651251e-02 5.55250704e-01 6.17103040e-01
5.31055093e-01 2.16523394e-01 5.98469555e-01 -1.12037587e+00
-3.52624416e-01 -1.28949964e+00 -5.95552623e-01 3.58186185e-01
5.56106083e-02 -7.21826911e-01 -3.56307358e-01 1.71578467e-01] | [9.855874061584473, 2.6736977100372314] |
8154d8ad-5798-43e1-81c4-364d02084908 | discourse-aware-emotion-cause-extraction-in | 2210.14419 | null | https://arxiv.org/abs/2210.14419v1 | https://arxiv.org/pdf/2210.14419v1.pdf | Discourse-Aware Emotion Cause Extraction in Conversations | Emotion Cause Extraction in Conversations (ECEC) aims to extract the utterances which contain the emotional cause in conversations. Most prior research focuses on modelling conversational contexts with sequential encoding, ignoring the informative interactions between utterances and conversational-specific features for ECEC. In this paper, we investigate the importance of discourse structures in handling utterance interactions and conversationspecific features for ECEC. To this end, we propose a discourse-aware model (DAM) for this task. Concretely, we jointly model ECEC with discourse parsing using a multi-task learning (MTL) framework and explicitly encode discourse structures via gated graph neural network (gated GNN), integrating rich utterance interaction information to our model. In addition, we use gated GNN to further enhance our ECEC model with conversation-specific features. Results on the benchmark corpus show that DAM outperform the state-of-theart (SOTA) systems in the literature. This suggests that the discourse structure may contain a potential link between emotional utterances and their corresponding cause expressions. It also verifies the effectiveness of conversationalspecific features. The codes of this paper will be available on GitHub. | ['Chen Gong', 'Guohong Fu', 'Yun Yuan', 'Nan Yu', 'Dexin Kong'] | 2022-10-26 | null | null | null | null | ['causal-emotion-entailment', 'discourse-parsing', 'emotion-cause-extraction'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 2.71142840e-01 7.19162941e-01 -1.77740436e-02 -6.61275506e-01
-5.61362088e-01 -4.60288882e-01 8.90430868e-01 8.01640451e-02
4.03284580e-02 7.36610711e-01 1.03217924e+00 -6.18050732e-02
2.47747581e-02 -6.04880571e-01 -3.22889656e-01 -5.05614042e-01
-2.47184277e-01 2.53847778e-01 -3.17457795e-01 -6.46616280e-01
1.09517612e-01 -6.81500137e-02 -1.37121594e+00 9.28746879e-01
7.03292787e-01 6.94881737e-01 -4.52766381e-03 9.77018237e-01
-5.20007372e-01 1.63765287e+00 -8.29251230e-01 -4.61608618e-01
-5.92441976e-01 -8.25431108e-01 -1.39101601e+00 2.76135325e-01
-2.27905482e-01 -1.52014524e-01 -6.65074065e-02 5.17782032e-01
4.39960957e-01 3.55467349e-01 5.57613432e-01 -1.49401534e+00
-2.36437514e-01 1.03072846e+00 -4.74369712e-02 2.87571680e-02
6.73290312e-01 -1.91585362e-01 1.35219502e+00 -6.06508195e-01
7.36551344e-01 1.59873438e+00 3.16389143e-01 8.53062809e-01
-6.89897418e-01 -2.38019273e-01 5.61959267e-01 4.75103676e-01
-7.04091787e-01 -6.11671507e-01 1.31464458e+00 -3.60976636e-01
1.52957344e+00 5.81677616e-01 4.12747443e-01 1.69056237e+00
6.64299577e-02 1.24690545e+00 8.85503352e-01 -6.13356948e-01
9.59065929e-02 -6.92696050e-02 5.46153784e-01 4.49686348e-01
-6.31895542e-01 -4.73240390e-02 -6.61161423e-01 -2.32186288e-01
1.40255496e-01 -4.88848805e-01 -3.81411165e-01 2.18570083e-01
-7.22858727e-01 1.17399955e+00 -1.98256113e-02 6.84665084e-01
-5.00629842e-01 3.74966450e-02 1.06175113e+00 3.88512224e-01
7.56093562e-01 4.11026478e-01 -4.59771544e-01 -5.45721352e-01
-1.66318312e-01 2.81288058e-01 1.38012826e+00 8.61355364e-01
4.70272511e-01 -1.00201927e-01 -5.16279638e-01 1.10625041e+00
1.52127355e-01 -1.58451453e-01 4.48476017e-01 -9.43269134e-01
4.31243569e-01 8.14958155e-01 -3.58369410e-01 -9.77036476e-01
-7.27758527e-01 -6.75462559e-02 -5.13760149e-01 -5.35866439e-01
1.40835193e-03 -7.06609845e-01 -1.34821132e-01 1.89728642e+00
3.62957805e-01 2.90806264e-01 6.14299178e-01 7.61732936e-01
1.49242926e+00 8.60533059e-01 2.47625709e-01 -5.09859264e-01
1.60054970e+00 -1.26778638e+00 -1.35652339e+00 -4.27357256e-01
9.95432794e-01 -5.97679198e-01 9.34113085e-01 2.64148235e-01
-7.88281620e-01 -1.99649498e-01 -7.10781872e-01 -1.26059547e-01
-1.90682918e-01 -8.45699571e-04 8.52672338e-01 4.99965847e-01
-8.02464068e-01 8.46281275e-02 -3.98880988e-01 -3.44834358e-01
-2.17580453e-01 1.42087564e-01 8.55162069e-02 1.52588412e-01
-1.72105753e+00 1.05100656e+00 4.67694491e-01 1.86133638e-01
-4.63185638e-01 -2.00594440e-01 -1.34546459e+00 8.22292343e-02
7.56973624e-01 -3.89842033e-01 1.42109644e+00 -1.22200751e+00
-1.93538380e+00 4.82522219e-01 -4.77803111e-01 -3.45954210e-01
-1.64968632e-02 -2.40335673e-01 -4.58270043e-01 2.71330774e-01
-4.11542684e-01 4.54271585e-01 4.66956079e-01 -1.34202623e+00
-3.37989271e-01 -6.13739640e-02 5.50225556e-01 4.74470735e-01
-4.29603487e-01 5.17906010e-01 -3.02795470e-01 -3.20507705e-01
-4.76464272e-01 -8.28197420e-01 -5.35642020e-02 -9.73951459e-01
-6.62352562e-01 -9.50893700e-01 9.90659356e-01 -6.80189550e-01
1.52792883e+00 -1.93277979e+00 4.89546865e-01 -9.02153105e-02
1.67076051e-01 -7.03661814e-02 -2.58715212e-01 8.37847054e-01
6.40312061e-02 8.24121013e-02 6.58164592e-03 -7.01022744e-01
3.00727785e-01 4.95906621e-01 -4.88869026e-02 -1.00470409e-01
4.34424311e-01 1.02651381e+00 -7.87029803e-01 -6.38900518e-01
2.88440138e-01 6.05214119e-01 -5.86130440e-01 6.35404408e-01
-4.97104228e-01 4.77006197e-01 -6.14146292e-01 2.16645867e-01
2.24657491e-01 -1.69919863e-01 6.86997831e-01 -1.26879826e-01
5.43538183e-02 6.38711691e-01 -5.74656546e-01 1.18567157e+00
-7.61997461e-01 7.24885464e-01 3.31979632e-01 -9.27152872e-01
9.32121754e-01 8.18009496e-01 2.63564289e-01 -4.88202125e-01
3.56189758e-01 -2.87871391e-01 1.26231581e-01 -1.06721008e+00
6.46339715e-01 -1.98649257e-01 -5.27853489e-01 4.52559829e-01
1.44810781e-01 1.98873840e-02 1.58226609e-01 3.21414471e-01
1.11265874e+00 6.54615611e-02 5.60493410e-01 1.95786893e-01
8.69260848e-01 -2.47065723e-01 6.38572812e-01 2.45677441e-01
-3.67826641e-01 2.08212420e-01 1.05135548e+00 -3.62907127e-02
-3.86972666e-01 -9.20039117e-02 3.25607836e-01 1.50389171e+00
-1.69694513e-01 -6.83608770e-01 -9.56820667e-01 -6.47234917e-01
-5.44418335e-01 1.18857324e+00 -7.20199525e-01 8.19059461e-03
-9.10149634e-01 -6.35615468e-01 7.14399219e-01 3.57600510e-01
4.06379402e-01 -1.47371864e+00 -5.17697334e-01 3.70686769e-01
-1.01201928e+00 -1.53816199e+00 -1.82499841e-01 2.36971304e-01
-2.08499998e-01 -1.16118479e+00 -9.49340016e-02 -5.79098105e-01
1.35608107e-01 -2.22202584e-01 1.14622271e+00 2.12888509e-01
1.41596392e-01 5.63073993e-01 -8.98917615e-01 -3.36964458e-01
-8.17541599e-01 1.13774270e-01 -5.30134320e-01 3.59303951e-02
6.33040130e-01 -3.39736998e-01 -1.90319777e-01 3.70037146e-02
-6.75041378e-01 3.38361859e-01 9.16137621e-02 8.05531979e-01
-1.97269410e-01 -2.10531488e-01 9.32848096e-01 -1.09837878e+00
1.16744745e+00 -7.26770282e-01 1.80782497e-01 2.57414371e-01
4.26422385e-03 -9.93440300e-02 5.50244927e-01 -1.45433560e-01
-1.64407313e+00 -2.62341291e-01 -4.20227468e-01 -1.43155068e-01
-4.09512371e-01 7.71132588e-01 -4.25299764e-01 5.27290583e-01
9.92888138e-02 -1.76461250e-01 -1.93596780e-01 -1.12612240e-01
3.50869775e-01 7.97206342e-01 1.50822744e-01 -8.69237721e-01
-1.88857600e-01 -5.62690049e-02 -3.36724341e-01 -1.00726950e+00
-1.09189844e+00 -4.83222693e-01 -4.16804701e-01 -7.14725137e-01
1.24800789e+00 -7.18869269e-01 -9.75878537e-01 1.79290116e-01
-1.59229517e+00 -4.92665797e-01 2.30037674e-01 4.30010259e-01
-6.26820147e-01 4.40986007e-01 -1.06045663e+00 -1.32895231e+00
-4.01228607e-01 -8.82744789e-01 9.89819467e-01 6.50149211e-02
-8.15200746e-01 -1.42595661e+00 1.55850612e-02 6.62874758e-01
1.34214878e-01 6.40935779e-01 1.07671034e+00 -1.06020379e+00
1.33337751e-02 2.50943869e-01 2.44341921e-02 1.36117041e-01
7.22598508e-02 5.57030365e-02 -1.32007217e+00 1.46385863e-01
3.16673696e-01 -6.53254330e-01 6.78886116e-01 4.27481413e-01
8.53915870e-01 -5.82162261e-01 -1.66033655e-01 -5.89819662e-02
9.50449109e-01 3.84972245e-01 4.66168702e-01 1.67000607e-01
5.81841826e-01 1.38129199e+00 5.73906839e-01 6.68801904e-01
9.99441743e-01 6.19410515e-01 4.14801240e-01 -2.42111441e-02
3.23731713e-02 5.03140204e-02 5.50151110e-01 1.36543071e+00
-1.11369193e-01 -7.29949951e-01 -8.61886859e-01 5.34192502e-01
-2.13878894e+00 -9.60705936e-01 -4.84684259e-01 1.14930832e+00
8.83567095e-01 -9.35513377e-02 -9.87238809e-03 -7.85546973e-02
7.59976208e-01 6.20239854e-01 -1.21878490e-01 -1.06960440e+00
-2.38019302e-01 -9.38886479e-02 -3.05507988e-01 8.09395909e-01
-1.13797772e+00 8.70166063e-01 5.33010340e+00 6.00743890e-01
-7.64079809e-01 2.85716534e-01 7.05513537e-01 1.42421108e-02
-3.76351297e-01 -1.63211718e-01 -6.29603505e-01 4.56416719e-02
1.26525354e+00 -1.95183828e-01 2.50425190e-01 7.10345387e-01
4.94594991e-01 9.54413638e-02 -1.27152658e+00 4.92952228e-01
4.12038475e-01 -1.09328628e+00 -2.51214445e-01 -9.35923457e-02
3.45989645e-01 -3.24416816e-01 -5.26030064e-01 8.51163805e-01
2.39677787e-01 -8.25045645e-01 3.53872925e-01 3.18537474e-01
8.17267373e-02 -8.82873535e-01 1.11559677e+00 3.18776160e-01
-1.07536960e+00 -7.32225627e-02 6.75462633e-02 -4.53062177e-01
4.56599712e-01 2.76202410e-01 -9.31230426e-01 8.59403610e-01
2.89317578e-01 5.67446828e-01 -2.23670900e-01 1.74398705e-01
-4.55369830e-01 8.65286410e-01 1.10920928e-01 -4.81556594e-01
4.79900718e-01 -7.96090961e-02 6.15059912e-01 1.75703263e+00
-2.75073070e-02 5.24532497e-01 3.31248105e-01 6.73369586e-01
-8.57830793e-02 3.03755730e-01 -5.42691529e-01 -1.87380075e-01
3.64185750e-01 1.33757222e+00 -4.84516174e-01 -3.54183584e-01
-5.25045276e-01 8.86942148e-01 4.26068425e-01 2.82341719e-01
-7.48881042e-01 -1.69896811e-01 4.76787895e-01 -7.51483142e-01
2.21571848e-01 1.36265576e-01 -6.17297105e-02 -9.44048464e-01
-1.14968009e-01 -9.55409586e-01 3.61880720e-01 -6.27873003e-01
-1.42839944e+00 9.74428654e-01 1.62763402e-01 -7.21317112e-01
-8.29921067e-01 -3.76920879e-01 -1.10074878e+00 4.14568782e-01
-1.56790733e+00 -1.21676183e+00 -1.76497042e-01 3.43796790e-01
1.02593553e+00 1.76703930e-01 1.14624655e+00 2.01192368e-02
-8.73587847e-01 3.08116853e-01 -5.95144570e-01 3.79977435e-01
4.74710137e-01 -1.26154733e+00 7.17829987e-02 4.73001659e-01
-3.56059939e-01 3.64339381e-01 8.89637887e-01 -6.14364266e-01
-1.16653347e+00 -9.84770536e-01 1.45773065e+00 -2.48013213e-01
6.29843414e-01 -4.92021173e-01 -1.03684342e+00 8.07138026e-01
1.00704420e+00 -5.80248833e-01 1.12746549e+00 6.16239369e-01
-7.46438801e-02 6.99039936e-01 -9.48996961e-01 4.96127605e-01
9.24366057e-01 -5.75809062e-01 -9.96947229e-01 2.84355521e-01
1.09968376e+00 -4.36968327e-01 -8.93528402e-01 3.12139928e-01
1.91217631e-01 -9.04922664e-01 5.33307731e-01 -6.11682475e-01
9.89531994e-01 3.97140443e-01 -1.38771147e-01 -1.45375443e+00
1.01013951e-01 -7.82649755e-01 -3.73366028e-01 1.72532880e+00
4.64246303e-01 -3.63400877e-01 3.21701586e-01 8.92839372e-01
-4.18676078e-01 -8.24252069e-01 -8.14676940e-01 -3.08358669e-01
2.57055238e-02 -5.82052767e-01 5.57177842e-01 1.23120034e+00
8.08272958e-01 9.61897969e-01 -5.12721658e-01 -1.29007563e-01
2.21472085e-02 1.25960946e-01 5.59279978e-01 -1.00142121e+00
-3.43868524e-01 -4.67882574e-01 2.23216489e-01 -7.48362422e-01
9.28847253e-01 -7.44012475e-01 2.35897571e-01 -1.68992686e+00
2.05594376e-01 4.34076712e-02 5.34290224e-02 5.39462209e-01
-4.35769707e-01 -5.46187103e-01 2.83186466e-01 -3.20494294e-01
-8.22535634e-01 9.52962995e-01 1.15237963e+00 6.28909543e-02
-3.19628954e-01 -2.01676175e-01 -6.33247435e-01 8.49156380e-01
8.84106874e-01 -1.60598055e-01 -4.16756690e-01 2.36598589e-02
3.09917033e-02 4.15186107e-01 1.82288215e-01 -1.77002117e-01
1.33281231e-01 -2.64382005e-01 -2.85459191e-01 -4.96056706e-01
6.37026906e-01 -4.61068898e-01 -2.75422424e-01 -2.15952545e-02
-7.91357279e-01 -2.84726590e-01 2.75529355e-01 3.05227488e-01
-5.39694309e-01 -3.85581553e-01 6.25416413e-02 -1.40110580e-02
-7.08243608e-01 -3.51155818e-01 -9.08987582e-01 1.36814460e-01
9.28257942e-01 1.16559364e-01 -5.23797989e-01 -9.53913450e-01
-7.55432546e-01 5.59175551e-01 -2.02310011e-01 7.01688170e-01
5.86238623e-01 -1.11940587e+00 -8.45537484e-01 -2.38479331e-01
8.73463154e-02 -2.47310802e-01 4.04194415e-01 7.98332393e-01
2.48860717e-01 5.53002179e-01 1.74722180e-01 -2.51215130e-01
-1.75667119e+00 2.47021332e-01 2.95043230e-01 -6.00124121e-01
-4.84406054e-01 8.20252001e-01 2.24722192e-01 -6.83152437e-01
2.53981113e-01 -4.14068043e-01 -8.43352735e-01 4.03214782e-01
2.92777419e-01 1.84944019e-01 -1.01374790e-01 -7.25442469e-01
-3.89153689e-01 -1.28119532e-02 -9.98327322e-03 -4.46493700e-02
1.43331325e+00 -3.99865150e-01 -3.45456481e-01 6.75251245e-01
1.36758208e+00 -1.13886960e-01 -1.04370713e+00 -2.45949656e-01
4.77523416e-01 1.11657575e-01 1.07027618e-02 -7.81584263e-01
-7.78085172e-01 7.86549211e-01 -1.52566850e-01 5.20833969e-01
1.03422499e+00 3.28770757e-01 7.83016562e-01 3.09210390e-01
-8.76395851e-02 -1.31603897e+00 2.09479466e-01 1.01355398e+00
1.30120194e+00 -1.12126088e+00 -2.87969977e-01 -8.39907706e-01
-1.16887367e+00 1.33179605e+00 9.29493308e-01 2.04914212e-01
3.60973477e-01 4.57295477e-01 2.29774669e-01 -6.29319072e-01
-1.46980238e+00 -3.01374495e-01 1.61896706e-01 3.24218869e-01
1.04325223e+00 1.40896216e-01 -5.41135013e-01 9.76563990e-01
-2.89316118e-01 -2.57342249e-01 5.91622233e-01 9.34055746e-01
-1.77994326e-01 -1.18542480e+00 -1.07184224e-01 1.91378415e-01
-4.03252006e-01 -1.36351496e-01 -9.95286405e-01 7.42416441e-01
-1.12027898e-01 1.59384596e+00 8.91811326e-02 -3.90698493e-01
2.84623444e-01 4.77394104e-01 1.75718591e-01 -6.87784553e-01
-1.17988729e+00 3.37327003e-01 1.25940311e+00 -5.52988410e-01
-9.90014672e-01 -6.26997769e-01 -1.52170408e+00 -1.55292079e-01
-3.14841807e-01 3.10277224e-01 4.43663716e-01 1.35418677e+00
2.43713871e-01 1.18573630e+00 6.92216992e-01 -7.11370051e-01
-6.11259192e-02 -1.24937010e+00 -3.82229015e-02 4.38698977e-01
1.84436545e-01 -5.30792952e-01 -5.09396136e-01 -1.41629189e-01] | [12.867389678955078, 6.378283977508545] |
61a84730-c3fa-4103-9b2a-0e7220dd3740 | proactive-robot-assistance-via-spatio | 2211.15501 | null | https://arxiv.org/abs/2211.15501v1 | https://arxiv.org/pdf/2211.15501v1.pdf | Proactive Robot Assistance via Spatio-Temporal Object Modeling | Proactive robot assistance enables a robot to anticipate and provide for a user's needs without being explicitly asked. We formulate proactive assistance as the problem of the robot anticipating temporal patterns of object movements associated with everyday user routines, and proactively assisting the user by placing objects to adapt the environment to their needs. We introduce a generative graph neural network to learn a unified spatio-temporal predictive model of object dynamics from temporal sequences of object arrangements. We additionally contribute the Household Object Movements from Everyday Routines (HOMER) dataset, which tracks household objects associated with human activities of daily living across 50+ days for five simulated households. Our model outperforms the leading baseline in predicting object movement, correctly predicting locations for 11.1% more objects and wrongly predicting locations for 11.5% fewer objects used by the human user. | ['Sonia Chernova', 'Maithili Patel'] | 2022-11-28 | null | null | null | null | ['temporal-sequences'] | ['reasoning'] | [ 2.31384503e-04 7.10589111e-01 -1.92685217e-01 -4.70655590e-01
-1.53373048e-01 -2.64195025e-01 3.59262049e-01 7.78455287e-03
-3.55736673e-01 7.15776980e-01 7.17612505e-01 1.35580469e-02
-1.82645157e-01 -6.65665030e-01 -9.27641690e-01 -4.52920288e-01
-6.40004575e-01 9.78252828e-01 -6.97743073e-02 -1.80649653e-01
-2.85211861e-01 6.63595498e-01 -1.40144157e+00 -7.60962293e-02
6.06010020e-01 5.75664580e-01 9.02922928e-01 9.58876193e-01
5.18486440e-01 1.22831345e+00 -7.66706839e-02 1.79458648e-01
1.39557600e-01 -1.52688608e-01 -8.97261560e-01 6.04098022e-01
-5.59662700e-01 -7.67464578e-01 -8.85980427e-01 5.39714098e-01
2.86308825e-01 6.86204195e-01 1.11204469e+00 -1.58929074e+00
-1.32407904e+00 6.69489384e-01 -5.74642979e-02 1.29444703e-01
1.03034890e+00 6.90504968e-01 7.39307225e-01 -4.60730731e-01
4.87014711e-01 1.44464254e+00 6.95109785e-01 4.04988974e-01
-1.25923300e+00 -4.33496907e-02 4.06984866e-01 4.77471977e-01
-1.30319273e+00 -6.39057994e-01 6.76904321e-01 -4.90535706e-01
1.73809743e+00 2.15197653e-02 7.37165570e-01 1.42302501e+00
2.45431036e-01 8.86883199e-01 1.14855774e-01 -1.14321131e-02
3.54263484e-01 -2.90267676e-01 2.77748585e-01 6.11889899e-01
2.19012171e-01 2.17234958e-02 -1.46899208e-01 -2.12381229e-01
7.68964887e-01 6.05722964e-01 -6.46016225e-02 -6.99300289e-01
-1.52758801e+00 5.43959260e-01 9.41276968e-01 1.71280652e-01
-1.32274950e+00 5.05142033e-01 -1.28086224e-01 -9.02186111e-02
8.15418884e-02 2.64296591e-01 -6.92727149e-01 -2.73136169e-01
2.98296750e-01 4.34329331e-01 8.47893536e-01 1.85883200e+00
8.09382856e-01 -1.18757049e-02 -1.82524443e-01 4.96981919e-01
5.44982255e-01 8.26767623e-01 4.34543878e-01 -1.25614619e+00
4.80091393e-01 8.12787592e-01 8.74785006e-01 -1.01038361e+00
-1.09877193e+00 5.45952380e-01 -8.65217805e-01 -5.50928771e-01
1.23134263e-01 -2.71694809e-01 -9.03168559e-01 1.62921035e+00
3.08760941e-01 -1.47058681e-01 2.67302603e-01 8.08500886e-01
3.80356073e-01 7.81610966e-01 6.45957112e-01 -1.06100507e-01
1.26808393e+00 -9.24524307e-01 -7.56576180e-01 -7.21791506e-01
8.69648814e-01 8.21697898e-03 7.36130357e-01 -8.12162012e-02
-7.98491716e-01 -6.04194760e-01 -3.65610987e-01 -1.04941674e-01
-2.40115121e-01 3.21224928e-01 8.09842169e-01 -2.39413500e-01
-1.34480643e+00 5.47039807e-01 -1.34546745e+00 -1.15632880e+00
2.13761672e-01 6.77407146e-01 -4.63918507e-01 -4.72187325e-02
-4.96882945e-01 9.04758990e-01 5.40088654e-01 4.08620350e-02
-6.91207767e-01 -3.00177664e-01 -1.39992881e+00 8.40632096e-02
1.91066712e-01 -8.03439975e-01 1.68557060e+00 -4.70250368e-01
-1.35425067e+00 5.52092969e-01 -1.85629576e-01 -7.32757092e-01
2.36804247e-01 -4.80907053e-01 -2.13854149e-01 -3.02147061e-01
5.60535252e-01 1.02939332e+00 3.94973487e-01 -1.04955280e+00
-8.04619789e-01 -3.41788650e-01 -2.84249157e-01 5.68491757e-01
1.99885577e-01 -4.03891236e-01 -2.39529535e-01 -3.13652873e-01
2.31135935e-01 -1.35294616e+00 -5.76624751e-01 -1.37973934e-01
-5.12069702e-01 -6.33800447e-01 5.78212321e-01 -7.00873375e-01
5.76649010e-01 -1.95632803e+00 4.72041041e-01 -1.87634438e-01
-5.69849834e-02 -6.73765361e-01 -2.86466599e-01 4.72444624e-01
1.42523810e-01 -6.56959653e-01 -8.31065848e-02 -7.35583305e-01
4.97971863e-01 5.67797482e-01 -1.88686445e-01 4.95945036e-01
7.15521351e-02 1.29953265e+00 -1.35636640e+00 -1.27397150e-01
5.83009779e-01 4.45282638e-01 -5.12477279e-01 6.63322747e-01
-2.73507446e-01 3.76809835e-01 -4.29940104e-01 7.30373502e-01
2.16531064e-02 -5.37636757e-01 6.40476286e-01 6.05630055e-02
3.50310028e-01 4.37464029e-01 -7.94106603e-01 1.39941180e+00
-2.55370498e-01 2.51555830e-01 -8.98917019e-02 -5.10607898e-01
6.54496789e-01 3.17326128e-01 8.59111786e-01 -6.58690572e-01
-3.04715298e-02 -3.05648029e-01 -4.00127470e-01 -8.49385738e-01
7.15320766e-01 5.88577092e-01 -5.55902600e-01 4.40556377e-01
-5.24685420e-02 1.11900479e-01 -2.18359262e-01 -1.12882815e-03
1.75539625e+00 2.22402856e-01 9.27536368e-01 -9.10285860e-02
-2.78153807e-01 2.50143051e-01 3.66267353e-01 8.04160774e-01
-5.92060626e-01 2.05102861e-01 -2.72136509e-01 -9.75549519e-01
-9.68000829e-01 -1.25514591e+00 7.19290614e-01 1.64931285e+00
2.07981318e-01 2.57936835e-01 -1.78802371e-01 -5.47719002e-01
3.40291113e-01 1.27269495e+00 -6.02595270e-01 -2.77840465e-01
-9.53004479e-01 -6.60181344e-01 -2.71515071e-01 8.93206954e-01
3.70549291e-01 -2.00995755e+00 -1.12876129e+00 5.16583502e-01
-3.70884895e-01 -1.06500494e+00 -8.35262060e-01 5.84326208e-01
-5.86356521e-01 -1.07078946e+00 -6.50745392e-01 -1.15183520e+00
8.72704327e-01 5.08419693e-01 1.21074259e+00 -2.41750076e-01
-9.57196578e-02 1.00235474e+00 -4.76356924e-01 -6.78086700e-03
-3.36040586e-01 3.78563702e-01 6.78072512e-01 -3.90819103e-01
5.27821302e-01 -8.61284852e-01 -5.62492371e-01 3.38185340e-01
-1.61395386e-01 3.17348093e-01 5.45727789e-01 2.56868929e-01
3.23204666e-01 -8.56754556e-03 4.90564793e-01 -2.64146715e-01
3.37681681e-01 -1.25885940e+00 -2.91371286e-01 1.17354505e-02
-2.82605559e-01 6.76753223e-02 2.46805966e-01 -8.57997000e-01
-9.26890492e-01 9.89502668e-01 2.49450594e-01 -1.23446621e-01
-7.68622637e-01 -2.07440466e-01 -2.96349168e-01 5.75828195e-01
5.51456332e-01 3.01799148e-01 -6.31775737e-01 -3.84138286e-01
6.68314517e-01 4.19337839e-01 1.01464188e+00 -4.36970115e-01
5.44548154e-01 1.32143334e-01 -4.05944139e-01 -6.97226286e-01
-1.00451790e-01 -5.35530269e-01 -9.83004391e-01 -1.46728173e-01
1.15341890e+00 -9.92452443e-01 -1.39516032e+00 4.73032117e-01
-1.38193393e+00 -1.39835763e+00 -3.20555538e-01 3.60411078e-01
-1.16277182e+00 -1.24614969e-01 -5.94726980e-01 -1.11613393e+00
-9.12961885e-02 -8.30895007e-01 1.50547099e+00 5.56768663e-02
-8.58261466e-01 -8.33827436e-01 -2.96022408e-02 -1.55661106e-01
2.16817081e-01 1.60055473e-01 5.63538373e-01 -5.48467100e-01
-7.82805681e-01 6.34701084e-03 -1.06210813e-01 -6.50746882e-01
8.26001823e-01 -7.87192404e-01 -4.19441819e-01 -1.21039025e-01
-2.75601417e-01 -2.78868880e-02 1.92168474e-01 5.86897612e-01
5.28604090e-01 -1.15528762e+00 -1.03410614e+00 2.39743754e-01
8.07437718e-01 8.17025125e-01 3.88548970e-01 3.25554818e-01
8.69616866e-01 7.10954428e-01 5.88095307e-01 6.09153807e-01
1.44841421e+00 8.34210038e-01 4.93926913e-01 4.68068540e-01
2.10660756e-01 -5.48539877e-01 4.77639943e-01 1.44233525e-01
1.79340199e-01 -5.89454353e-01 -1.15028417e+00 9.21154201e-01
-2.48005414e+00 -8.30491841e-01 1.61111653e-02 1.40859675e+00
3.19034845e-01 -2.76683748e-01 5.07442415e-01 -2.65952945e-01
5.07017076e-01 -1.42527848e-01 -1.00753224e+00 2.40267321e-01
4.62739885e-01 -9.10071909e-01 7.23506153e-01 5.68145454e-01
-1.28010380e+00 1.01054072e+00 6.71032476e+00 -3.85732353e-01
-2.13042364e-01 -9.13080871e-02 5.02849638e-01 8.44225194e-03
3.42375308e-01 -6.26213968e-01 -5.92691243e-01 5.45244455e-01
9.25815105e-01 1.22054338e-01 9.31846559e-01 1.43256426e+00
3.72147679e-01 -1.79578871e-01 -1.64749956e+00 1.09112453e+00
-1.10019475e-01 -7.94874907e-01 -3.74573857e-01 4.49363403e-02
5.36814094e-01 3.00546795e-01 -1.39694110e-01 5.75669527e-01
1.46521509e+00 -7.35147178e-01 9.67519283e-01 9.73230362e-01
1.26592413e-01 -2.57337540e-01 3.28961402e-01 7.61751831e-01
-1.42342997e+00 -5.38098752e-01 4.90473881e-02 -5.38704991e-01
6.31056428e-01 -1.47272572e-01 -1.50638533e+00 -1.03156097e-01
1.10285056e+00 9.14794564e-01 -4.24948990e-01 5.59740305e-01
-1.57947049e-01 3.58901508e-02 -5.23091614e-01 -2.30272189e-01
-3.97669002e-02 -2.50825360e-02 5.33755660e-01 9.50094223e-01
5.02467453e-01 7.29312420e-01 3.92437607e-01 7.00472116e-01
8.86945128e-02 -5.02773941e-01 -1.30773604e+00 2.30129912e-01
6.31068945e-01 8.26237082e-01 -6.65003359e-01 -3.15259844e-01
1.74760111e-02 1.73427367e+00 5.83124995e-01 5.69065511e-01
-5.97904921e-01 1.49171308e-01 7.84235001e-01 -7.06175938e-02
2.09706202e-01 -5.35931051e-01 -5.43178096e-02 -6.63702071e-01
-1.17267445e-01 -2.99378783e-01 1.97891816e-01 -1.44546962e+00
-1.28838038e+00 3.47788990e-01 1.33025199e-02 -6.93501413e-01
-9.06580865e-01 -3.36830258e-01 -3.87078732e-01 3.09217453e-01
-6.61320150e-01 -1.13964832e+00 -4.20202971e-01 4.53073442e-01
9.11670506e-01 5.57086244e-02 8.95708859e-01 -5.81133999e-02
-4.61079717e-01 5.52814119e-02 -2.10414037e-01 3.03783063e-02
2.15858176e-01 -1.39435458e+00 1.02644837e+00 5.25382876e-01
-1.64243728e-01 4.20755297e-01 9.03003275e-01 -1.11830485e+00
-1.68808961e+00 -1.65563357e+00 1.18146241e+00 -1.16355753e+00
4.88302469e-01 -3.23025405e-01 -7.56495357e-01 1.72689879e+00
-8.08292478e-02 -1.70184553e-01 1.40702501e-01 -2.35291168e-01
2.95337051e-01 3.91432852e-01 -1.48155236e+00 8.17506075e-01
1.85825229e+00 -2.96143293e-01 -7.20957339e-01 6.67668104e-01
1.25109005e+00 -4.07404304e-01 -5.50202787e-01 1.30658150e-01
2.33199582e-01 -4.34302419e-01 1.10802364e+00 -3.64622742e-01
-1.29999399e-01 -1.65094554e-01 -6.04286253e-01 -1.36959898e+00
-1.05462134e+00 -7.05005288e-01 -6.25330329e-01 9.01589036e-01
6.82885051e-02 -5.63664377e-01 7.83697903e-01 1.34381449e+00
-3.60703290e-01 -1.87679261e-01 -6.99361384e-01 -5.37248254e-01
-9.25747573e-01 -4.80743885e-01 1.08356714e+00 6.66217804e-01
2.92889953e-01 3.54865879e-01 -5.75477302e-01 6.24892116e-01
4.05003041e-01 -3.35930228e-01 9.70162868e-01 -9.44937229e-01
-1.15794465e-01 -1.99671730e-01 -3.98862332e-01 -1.40321708e+00
3.03617686e-01 -6.11068428e-01 8.93492341e-01 -2.09122992e+00
1.86646506e-01 -4.47109163e-01 2.85390228e-01 8.69504988e-01
-8.25794414e-02 -2.74054974e-01 1.46164075e-01 3.95843685e-01
-1.18723679e+00 7.52559662e-01 4.55154896e-01 -2.99012303e-01
-8.15172255e-01 2.05920205e-01 -5.19033372e-01 8.11766624e-01
7.13993609e-01 -3.11591923e-01 -3.63546222e-01 -6.58106506e-01
-5.23749627e-02 3.20051610e-01 5.40825188e-01 -1.18042719e+00
3.66140157e-01 -3.65968287e-01 5.91635466e-01 -7.49712884e-01
5.14375448e-01 -1.39497876e+00 5.56516826e-01 6.72017694e-01
-2.70560771e-01 5.96993387e-01 1.06636822e-01 9.50318873e-01
9.33009684e-01 5.07103622e-01 1.57035947e-01 8.92015770e-02
-1.17223179e+00 2.94441372e-01 -1.07248807e+00 -6.74450696e-01
1.31715953e+00 2.09226161e-02 5.63728437e-03 -5.47121465e-01
-1.10477412e+00 6.12390697e-01 6.10027492e-01 8.87813687e-01
4.06196356e-01 -1.61986542e+00 -4.06552153e-03 2.82479376e-02
1.74222022e-01 3.19911212e-01 9.94070545e-02 2.03882247e-01
-2.89523095e-01 6.90585911e-01 -1.04639851e-01 -5.71323216e-01
-7.30704367e-01 1.06767702e+00 2.68719524e-01 -3.62896547e-02
-9.30213690e-01 5.68409979e-01 1.62006810e-01 -9.13588643e-01
5.18672168e-01 -9.66613054e-01 -1.73922554e-01 -4.50345457e-01
2.12414414e-01 8.43009889e-01 -6.34880602e-01 -1.10130727e+00
-4.12324935e-01 8.25178772e-02 4.42086935e-01 1.21126674e-01
1.55956066e+00 -7.38959849e-01 2.24949598e-01 6.73307955e-01
8.77674162e-01 -9.68442619e-01 -1.74420142e+00 -3.35518658e-01
3.67590189e-01 1.09622307e-01 -6.67104661e-01 -7.39909828e-01
-4.29887056e-01 2.83245873e-02 6.19440973e-01 3.11250120e-01
8.10608268e-01 4.96809632e-01 8.57409477e-01 1.12190413e+00
1.17818248e+00 -5.93078136e-01 3.69163603e-01 4.71117437e-01
1.07843661e+00 -1.52714646e+00 -3.47302973e-01 -4.33323812e-03
-9.42786634e-01 4.74545598e-01 7.05561340e-01 -2.31884032e-01
5.97459137e-01 7.95498043e-02 -3.92848581e-01 -2.52958506e-01
-7.08431840e-01 -2.52202481e-01 1.76189914e-01 1.29485083e+00
-1.01792723e-01 6.76887691e-01 8.04366231e-01 7.34783769e-01
-3.53129923e-01 -2.66598649e-02 -3.47314030e-03 9.33701277e-01
-6.13572478e-01 -3.03721160e-01 -2.96360880e-01 5.74838340e-01
4.28601176e-01 4.35738683e-01 -2.73436844e-01 6.07163429e-01
-7.12541416e-02 1.10965836e+00 4.61662352e-01 -4.46448505e-01
6.89506769e-01 1.65130422e-01 2.15802580e-01 -7.50955224e-01
-5.08142412e-02 -4.33452815e-01 -1.41492516e-01 -9.63450193e-01
-1.20467782e-01 -1.03553665e+00 -1.59320045e+00 -2.74381250e-01
5.23560941e-01 -3.67821395e-01 7.34903142e-02 7.50350475e-01
5.99315941e-01 5.51774442e-01 3.51149529e-01 -1.93585467e+00
-1.52157396e-01 -1.19014847e+00 -4.33751613e-01 3.93111706e-01
6.53201461e-01 -6.52296007e-01 -1.16371013e-01 5.14369249e-01] | [4.823270797729492, 0.5902055501937866] |
a078b8be-9829-4fbb-b323-4a03d40c4c6c | transfer4d-a-framework-for-frugal-motion | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Maheshwari_Transfer4D_A_Framework_for_Frugal_Motion_Capture_and_Deformation_Transfer_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Maheshwari_Transfer4D_A_Framework_for_Frugal_Motion_Capture_and_Deformation_Transfer_CVPR_2023_paper.pdf | Transfer4D: A Framework for Frugal Motion Capture and Deformation Transfer | Animating a virtual character based on a real performance of an actor is a challenging task that currently requires expensive motion capture setups and additional effort by expert animators, rendering it accessible only to large production houses. The goal of our work is to democratize this task by developing a frugal alternative termed "Transfer4D" that uses only commodity depth sensors and further reduces animators' effort by automating the rigging and animation transfer process. To handle sparse, incomplete videos from depth video inputs and large variations between source and target objects, we propose to use skeletons as an intermediary representation between motion capture and transfer. We propose a novel skeleton extraction pipeline from single-view depth sequence that incorporates additional geometric information, resulting in superior performance in motion reconstruction and transfer in comparison to the contemporary methods. We use non-rigid reconstruction to track motion from the depth sequence, and then we rig the source object using skinning decomposition. Finally, the rig is embedded into the target object for motion retargeting. | ['Ramya Hebbalaguppe', 'Rahul Narain', 'Shubh Maheshwari'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['motion-retargeting'] | ['computer-vision'] | [ 3.95838886e-01 -8.15844834e-02 2.45585740e-01 1.05808571e-01
-2.85950869e-01 -7.43362963e-01 5.88681042e-01 -4.52257782e-01
-3.73208314e-01 3.40334833e-01 1.11917406e-03 6.23427629e-02
4.67482179e-01 -6.35772109e-01 -5.38207769e-01 -4.25202310e-01
2.37152323e-01 5.27283907e-01 9.20885921e-01 -1.34860337e-01
1.54406726e-01 7.36301780e-01 -1.48080468e+00 7.10537583e-02
4.34619844e-01 6.36064529e-01 4.32860553e-01 9.05070543e-01
3.04175355e-03 6.01623118e-01 -4.61768448e-01 -3.46632957e-01
5.44133782e-01 -6.24112785e-01 -6.57976270e-01 4.42877620e-01
5.04325807e-01 -7.22880363e-01 -4.06719595e-01 5.97189844e-01
5.71569145e-01 2.00404391e-01 4.02210891e-01 -8.50324690e-01
1.48007885e-01 1.47211835e-01 -9.01878178e-01 -9.12084877e-02
7.96232820e-01 1.64884016e-01 5.47100544e-01 -7.28954196e-01
1.19385171e+00 1.15922892e+00 6.54061735e-01 9.74384606e-01
-1.19225490e+00 -5.06630778e-01 1.33988351e-01 1.87794015e-01
-1.10092247e+00 -5.17542422e-01 1.17574453e+00 -5.75727046e-01
5.34988761e-01 3.13630849e-01 1.17247140e+00 1.18216586e+00
-1.28895625e-01 7.69728780e-01 5.43519139e-01 -4.93031710e-01
1.16611496e-01 -2.35520706e-01 -5.70050955e-01 8.18351328e-01
-5.78379221e-02 -5.44446073e-02 -5.09853542e-01 -5.83793921e-03
1.53123510e+00 -6.11912981e-02 -5.64674795e-01 -8.10038269e-01
-1.39845979e+00 5.30072868e-01 3.14873643e-02 2.99488269e-02
-5.11569798e-01 4.53343093e-01 2.24385723e-01 1.21300982e-03
2.80872405e-01 7.31731951e-02 -2.45361581e-01 -5.39499283e-01
-1.24149382e+00 4.23970252e-01 6.41817212e-01 8.66744339e-01
5.57543337e-01 1.62584275e-01 2.55988210e-01 5.28200030e-01
3.88642251e-01 3.52465570e-01 3.26510102e-01 -1.44595444e+00
4.34240848e-01 5.38596392e-01 7.23213330e-02 -1.05878603e+00
-2.86266744e-01 1.09910339e-01 -5.65540612e-01 6.02371097e-01
4.46798772e-01 -8.94678831e-02 -8.45043600e-01 1.33295834e+00
9.63321805e-01 3.11489522e-01 -2.83061475e-01 1.18194556e+00
7.13361204e-01 4.61483479e-01 -1.37467057e-01 -1.14098459e-01
1.27651560e+00 -9.60059583e-01 -5.06584227e-01 -6.21018820e-02
4.82975245e-01 -9.02557909e-01 9.86570537e-01 4.85749781e-01
-1.42661226e+00 -3.92586172e-01 -1.01818931e+00 -2.87578374e-01
3.63195062e-01 -1.93050370e-01 4.98724282e-01 4.86338466e-01
-7.77352810e-01 9.86660719e-01 -1.09746909e+00 -2.38788962e-01
2.53848493e-01 4.19397861e-01 -6.64427161e-01 6.29205182e-02
-8.07392776e-01 6.46351457e-01 2.18965858e-01 1.22934677e-01
-8.60495508e-01 -6.68589532e-01 -1.03862512e+00 -4.14266407e-01
3.25120807e-01 -1.20330739e+00 1.06692529e+00 -1.04727042e+00
-2.38566327e+00 8.62813234e-01 1.89059183e-01 4.64815386e-02
1.12807047e+00 -4.06397194e-01 2.22343475e-01 5.98048508e-01
-2.52887696e-01 6.65715635e-01 1.10422385e+00 -1.30792212e+00
-3.89507622e-01 -3.07905644e-01 -6.10880330e-02 2.60750771e-01
-6.26179799e-02 3.84396873e-02 -8.86431754e-01 -9.03070211e-01
3.38068098e-01 -1.08603489e+00 -1.35957316e-01 4.97771949e-01
-2.91058779e-01 3.21991146e-01 1.29365051e+00 -9.87105787e-01
1.04558170e+00 -1.94254506e+00 8.41600895e-01 3.59995402e-02
2.28716552e-01 4.10556406e-01 -1.15297578e-01 1.44014984e-01
2.70178672e-02 -3.82674873e-01 -3.48046958e-01 -6.33015573e-01
-3.31682295e-01 1.88253447e-01 6.88048080e-02 7.21957564e-01
-5.50041273e-02 7.23792076e-01 -7.97955096e-01 -7.23672867e-01
5.36252797e-01 6.43948197e-01 -7.65800834e-01 4.36606139e-01
-1.58152565e-01 9.53981996e-01 -4.48648959e-01 7.82017112e-01
5.60047448e-01 7.69647360e-02 1.01666659e-01 -1.34711012e-01
-2.60082453e-01 5.00228293e-02 -1.41643870e+00 2.41741657e+00
-2.65749067e-01 4.49589431e-01 4.03324842e-01 -5.30134499e-01
8.28367233e-01 3.37843299e-01 7.04936206e-01 -2.28271514e-01
2.13280544e-01 1.30338654e-01 -2.78758198e-01 -6.30020082e-01
4.47287649e-01 -2.87982166e-01 1.37901723e-01 3.72219920e-01
-2.41922483e-01 -4.98980343e-01 -1.13180503e-01 -5.75451367e-02
1.09257281e+00 1.10043836e+00 1.91708282e-01 2.30234697e-01
4.69272345e-01 1.32431939e-01 6.69034064e-01 -1.42910391e-01
1.04250580e-01 1.22162747e+00 3.12786430e-01 -4.43810374e-01
-1.31082296e+00 -9.61399853e-01 2.58022785e-01 6.17299557e-01
2.73894817e-01 -3.83279562e-01 -8.62982571e-01 -5.10466099e-01
-1.86882243e-01 1.49015784e-01 -5.18436790e-01 2.21380964e-02
-1.18340600e+00 -2.35874787e-01 4.12941098e-01 5.83929300e-01
4.63674158e-01 -8.26492727e-01 -1.25339901e+00 3.29124123e-01
-2.94079602e-01 -1.14717031e+00 -5.32305598e-01 -3.73992652e-01
-1.23564076e+00 -8.88796449e-01 -1.22800469e+00 -3.86705846e-01
4.49296266e-01 1.28354266e-01 9.31970239e-01 2.81053968e-02
-4.17260528e-01 2.69989580e-01 -4.12710488e-01 7.86480904e-02
-3.88596445e-01 -4.86270860e-02 -8.62772018e-03 1.39690787e-01
-4.60113943e-01 -8.22698236e-01 -9.77358043e-01 3.66329134e-01
-8.34176421e-01 4.66877639e-01 3.78174782e-01 4.69790757e-01
4.53867137e-01 -4.94298339e-01 -1.70229733e-01 -5.89914858e-01
-7.35968277e-02 -1.34477526e-01 -5.59213698e-01 -2.35715881e-01
1.14892371e-01 -2.11965948e-01 4.34279531e-01 -8.42426121e-01
-1.10242999e+00 4.91631746e-01 -1.35336429e-01 -9.43456113e-01
9.64950845e-02 -1.62286118e-01 -4.33220416e-01 -6.43771440e-02
4.16780263e-01 -1.44817485e-02 1.86185926e-01 -7.27224052e-01
3.57827902e-01 2.78272301e-01 7.50116110e-01 -5.30523956e-01
9.65307474e-01 8.82868826e-01 1.88979790e-01 -8.51576328e-01
-1.03608347e-01 -2.80160218e-01 -1.28831220e+00 -3.86830479e-01
9.85242426e-01 -7.69253492e-01 -7.01157153e-01 5.32541573e-01
-1.45011258e+00 -4.95331615e-01 -3.71481985e-01 6.37168348e-01
-7.31263161e-01 8.53530943e-01 -6.46057427e-01 -4.44394380e-01
-3.04163754e-01 -1.06888759e+00 1.38610303e+00 -1.59170236e-02
-6.06809139e-01 -9.85985756e-01 4.36536908e-01 4.93781865e-01
1.65873971e-02 9.90562081e-01 5.46213746e-01 3.23748916e-01
-7.48448193e-01 -2.69878000e-01 2.32683763e-01 1.65872186e-01
4.96293306e-02 4.02907670e-01 -8.36400926e-01 -2.58213699e-01
-1.50565773e-01 1.02417774e-01 4.57580507e-01 2.90235281e-01
7.39786983e-01 -3.25433686e-02 -2.79515445e-01 9.96564329e-01
1.17878234e+00 -4.45283111e-03 7.76659548e-01 6.03132367e-01
1.21859491e+00 9.61959124e-01 6.97190464e-01 6.15745723e-01
1.56759307e-01 1.10747063e+00 4.67518568e-01 -8.65566730e-02
-4.11583722e-01 -2.81265229e-01 5.49168110e-01 1.05959713e+00
-8.02541077e-01 7.07389265e-02 -8.39638591e-01 5.13368368e-01
-1.83285773e+00 -9.50305820e-01 -3.87161702e-01 2.26184821e+00
7.97135890e-01 -1.78534567e-01 3.39288473e-01 2.21295461e-01
5.24316311e-01 7.94492662e-03 -4.24440831e-01 -1.17368504e-01
2.04927504e-01 2.29975328e-01 1.55638352e-01 4.38148469e-01
-7.65347242e-01 9.61804152e-01 5.82233238e+00 5.54783285e-01
-1.17256594e+00 1.76671743e-01 -6.52488172e-02 -3.41064513e-01
-2.56504685e-01 1.25048757e-01 -4.18232650e-01 2.97680110e-01
6.07046723e-01 1.64834887e-01 4.18182574e-02 7.63514519e-01
4.20871943e-01 -1.41612262e-01 -1.22978389e+00 1.02936852e+00
1.46891281e-01 -1.08454025e+00 6.47872612e-02 8.18364844e-02
5.98050952e-01 -4.22256052e-01 -2.11124778e-01 -3.09814513e-01
2.57917028e-02 -8.04052293e-01 9.68794644e-01 5.35942614e-01
8.65379393e-01 -5.88709891e-01 3.30684632e-01 3.50942820e-01
-1.36824536e+00 3.75665069e-01 -2.02387363e-01 1.66721642e-02
6.46089554e-01 2.45077699e-01 -6.34269774e-01 5.72415590e-01
4.98922288e-01 6.28802478e-01 -2.44761810e-01 8.14031839e-01
-1.10747866e-01 1.83248129e-02 -2.15561688e-01 4.34686720e-01
-9.84825119e-02 -3.53857636e-01 8.17706823e-01 1.10536861e+00
3.70676994e-01 1.62055224e-01 7.87591860e-02 6.81615472e-01
1.68744549e-01 -5.52966772e-03 -5.96483171e-01 3.79910052e-01
2.04139482e-02 1.29825795e+00 -8.71959925e-01 -2.73933828e-01
-1.63916245e-01 1.66552675e+00 1.77465808e-02 1.12192705e-01
-1.01011801e+00 -2.14166299e-01 5.95520377e-01 6.72644734e-01
3.49292278e-01 -4.38866287e-01 -8.84779170e-03 -1.45000279e+00
1.22720279e-01 -7.51447737e-01 5.45973890e-02 -6.89590991e-01
-5.90003908e-01 4.95754033e-01 1.12657128e-02 -1.63106775e+00
-5.14832973e-01 -2.97648460e-01 -4.65702862e-01 6.82796419e-01
-1.19618928e+00 -1.28747618e+00 -7.15249836e-01 7.76877046e-01
8.13497603e-01 2.53483325e-01 6.20873451e-01 3.14936936e-01
-4.06073213e-01 2.83374012e-01 -3.09384674e-01 8.98321345e-03
6.58688903e-01 -1.02745199e+00 4.69872862e-01 8.21324468e-01
-9.93814319e-02 3.05273116e-01 7.23730922e-01 -6.93588257e-01
-1.52167130e+00 -8.29537570e-01 2.90374279e-01 -8.14884603e-01
3.29854965e-01 -3.72777283e-01 -8.42292249e-01 5.91426015e-01
-1.08621322e-01 1.41610444e-01 4.11513031e-01 -7.14245319e-01
-3.76881585e-02 2.37159654e-01 -1.14612651e+00 7.73698509e-01
1.17269647e+00 -2.16719136e-01 -5.61903298e-01 -1.26620054e-01
4.33735251e-01 -7.92153955e-01 -8.65817189e-01 2.05038875e-01
8.13498199e-01 -9.93982077e-01 1.17047012e+00 -1.68644547e-01
6.98367953e-01 -6.30088985e-01 2.64762081e-02 -9.27395880e-01
-4.40662764e-02 -1.02581930e+00 -4.64941084e-01 1.13180363e+00
-2.68648386e-01 6.31269440e-02 1.08127570e+00 7.29676902e-01
8.47672485e-03 -5.00233650e-01 -9.75869954e-01 -5.24408281e-01
-9.94196609e-02 -3.62920702e-01 2.50214219e-01 1.00247884e+00
-1.57973647e-01 7.97731802e-02 -4.92354095e-01 2.80411076e-02
8.86202753e-01 1.95127383e-01 1.25047636e+00 -1.21918011e+00
-6.16416752e-01 -1.21274084e-01 -6.90216482e-01 -1.27017140e+00
7.63015123e-03 -5.00267446e-01 -4.63078208e-02 -1.28408551e+00
-8.63271877e-02 -2.48335004e-01 6.37629986e-01 2.42857024e-01
9.37389955e-03 5.56145251e-01 3.27464193e-01 4.87564713e-01
-2.53418297e-01 6.34661138e-01 1.59611726e+00 2.28946611e-01
-4.36624885e-01 -7.70680532e-02 3.93849723e-02 1.26552224e+00
3.53027701e-01 -4.20791119e-01 -2.26531982e-01 -5.63005984e-01
-6.64211735e-02 4.93752450e-01 5.23925483e-01 -1.07161808e+00
5.29058501e-02 -1.48657799e-01 5.32084942e-01 -5.53900421e-01
7.36473143e-01 -9.49177086e-01 7.14900196e-01 6.54397726e-01
1.28031328e-01 3.39509845e-01 8.13784525e-02 5.57930529e-01
4.28292453e-02 -3.86901408e-01 9.09034252e-01 -4.04697835e-01
-6.60844147e-01 3.75592411e-01 -1.84911713e-01 -1.64993599e-01
1.35578847e+00 -6.63057208e-01 3.41743976e-01 -1.10103801e-01
-8.81617308e-01 -2.23985478e-01 1.04844499e+00 2.53832102e-01
8.56556833e-01 -1.39620245e+00 -6.40999079e-01 1.17090717e-01
-3.37485045e-01 5.32176852e-01 2.85836160e-01 7.81088173e-01
-1.39417517e+00 -1.59831196e-01 -3.61937165e-01 -8.02236438e-01
-1.69651711e+00 3.35983396e-01 2.58529007e-01 -2.25851592e-03
-1.29976928e+00 7.49687433e-01 2.50498891e-01 -1.44482955e-01
-5.00404388e-02 -1.83976948e-01 -1.03875957e-01 7.72953108e-02
5.50812602e-01 7.53501117e-01 -1.28338739e-01 -7.30250239e-01
-1.97721392e-01 1.11030447e+00 1.22453690e-01 -5.12235999e-01
1.29497194e+00 -2.43270889e-01 1.55891143e-02 4.03090239e-01
1.05916727e+00 2.30083257e-01 -1.85969055e+00 9.44051817e-02
-2.62104750e-01 -9.94986653e-01 -2.02527121e-01 -1.41562313e-01
-1.13084102e+00 9.52597439e-01 3.66580635e-01 -3.78705651e-01
9.29720998e-01 -1.42020747e-01 1.15586388e+00 -9.08728540e-02
6.18795753e-01 -9.09900963e-01 3.25912654e-01 1.28302947e-01
8.48757327e-01 -7.55936563e-01 2.52859741e-01 -6.26847804e-01
-4.60695952e-01 1.37276971e+00 5.72622776e-01 -1.49102181e-01
1.22631893e-01 2.94765472e-01 1.02623031e-01 -1.44716471e-01
-4.26000744e-01 2.19086275e-01 1.18375137e-01 7.55090415e-01
1.92881733e-01 -2.39246503e-01 -3.31777148e-02 1.13933221e-01
-1.40531734e-01 -7.18515925e-03 7.05627382e-01 1.13418591e+00
-2.28119418e-01 -1.34804714e+00 -6.76157653e-01 -2.88677424e-01
-4.90136981e-01 1.55069903e-01 -3.74984950e-01 8.17754388e-01
7.19929636e-02 4.38673019e-01 7.49146044e-02 -2.94720113e-01
6.15181327e-01 -1.32345960e-01 8.29883397e-01 -7.27511942e-01
-6.84623063e-01 4.20012712e-01 -9.50728431e-02 -8.33651662e-01
-6.59647405e-01 -8.71402562e-01 -1.29201472e+00 -4.83202308e-01
-2.39355996e-01 -3.67059916e-01 6.65210545e-01 5.71761668e-01
2.61922270e-01 5.37181079e-01 4.98586953e-01 -1.73111761e+00
-2.01821938e-01 -6.05319440e-01 -4.18203741e-01 5.05793750e-01
2.70575851e-01 -7.93779612e-01 -2.18320936e-01 4.87748533e-01] | [7.350462436676025, -0.9038503170013428] |
477e5c92-49c8-489c-b1e6-68633bcd6d49 | community-time-activity-trajectory-modelling | 2212.06366 | null | https://arxiv.org/abs/2212.06366v1 | https://arxiv.org/pdf/2212.06366v1.pdf | Community Time-Activity Trajectory Modelling based on Markov Chain Simulation and Dirichlet Regression | Accurate modeling of human time-activity trajectory is essential to support community resilience and emergency response strategies such as daily energy planning and urban seismic vulnerability assessment. However, existing modeling of time-activity trajectory is only driven by socio-demographic information with identical activity trajectories shared among the same group of people and neglects the influence of the environment. To further improve human time-activity trajectory modeling, this paper constructs community time-activity trajectory and analyzes how social-demographic and built environment influence people s activity trajectory based on Markov Chains and Dirichlet Regression. We use the New York area as a case study and gather data from American Time Use Survey, Policy Map, and the New York City Energy & Water Performance Map to evaluate the proposed method. To validate the regression model, Box s M Test and T-test are performed with 80% data training the model and the left 20% as the test sample. The modeling results align well with the actual human behavior trajectories, demonstrating the effectiveness of the proposed method. It also shows that both social-demographic and built environment factors will significantly impact a community's time-activity trajectory. Specifically, 1) Diversity and median age both have a significant influence on the proportion of time people assign to education activity. 2) Transportation condition affects people s activity trajectory in the way that longer commute time decreases the proportion of biological activity (eg. sleeping and eating) and increases people s working time. 3) Residential density affects almost all activities with a significant p-value for all biological needs, household management, working, education, and personal preference. | ['Jianli Chen', 'Yuqing Hu', 'Chen Xia'] | 2022-12-13 | null | null | null | null | ['trajectory-modeling'] | ['time-series'] | [-3.02471340e-01 -1.34224206e-01 -4.40191090e-01 3.34087983e-02
1.12763457e-01 -3.62568758e-02 4.59528834e-01 5.54937005e-01
-5.06667018e-01 7.56959260e-01 1.08615196e+00 -5.04117489e-01
-2.88442135e-01 -1.44882762e+00 -2.80214548e-01 -8.24813068e-01
-3.19977015e-01 1.62730515e-01 2.23738700e-01 -1.96049735e-01
1.00596689e-01 4.00078505e-01 -1.32232344e+00 -4.13508654e-01
1.02671182e+00 -1.26182092e-02 2.41541386e-01 3.76254231e-01
1.49910912e-01 4.10574645e-01 -1.48020893e-01 1.76560655e-01
-4.43552844e-02 -3.58788401e-01 -4.27834094e-01 -2.21418053e-01
-5.96511543e-01 -6.28333926e-01 -4.34685200e-01 3.90292794e-01
6.23297334e-01 4.92373705e-01 1.00238490e+00 -1.47086489e+00
-3.94888282e-01 7.42606997e-01 -3.32851291e-01 1.81287900e-01
4.66642201e-01 2.27824658e-01 2.63933301e-01 -5.34092367e-01
2.10796431e-01 1.16066504e+00 9.85527277e-01 5.95715223e-03
-1.14977694e+00 -7.76343405e-01 2.26076797e-01 1.69131041e-01
-1.86116207e+00 -3.23361814e-01 4.06546324e-01 -6.83934331e-01
1.16941011e+00 1.37591228e-01 1.22552514e+00 7.74449408e-01
2.40511417e-01 2.85426348e-01 8.93829465e-01 -1.81130216e-01
2.76076913e-01 -2.03173682e-01 1.79518715e-01 1.05825946e-01
3.45606625e-01 -3.12959775e-02 -1.55323848e-01 -4.57553744e-01
4.81233448e-01 6.68001592e-01 2.38202475e-02 2.35763982e-01
-1.10477388e+00 5.84805071e-01 3.20974469e-01 4.27096605e-01
-5.30480862e-01 3.77645135e-01 -2.17722822e-02 -2.23398775e-01
5.06763995e-01 -4.81883049e-01 -1.55295402e-01 -2.14613020e-01
-9.90151107e-01 5.05450666e-01 7.34856367e-01 5.74494064e-01
8.29849899e-01 -3.79556068e-03 -2.26335689e-01 5.78924477e-01
7.71467924e-01 1.28227901e+00 1.24577647e-02 -7.18068600e-01
3.44447523e-01 7.55949676e-01 2.67981768e-01 -1.32298422e+00
-6.03359342e-01 -8.83471742e-02 -9.78067994e-01 -4.71822083e-01
6.24976575e-01 -3.76261652e-01 -6.32212758e-01 1.68412602e+00
4.05733287e-01 2.78032839e-01 -1.71383098e-01 3.67311865e-01
2.28929654e-01 8.77077699e-01 7.52952158e-01 -4.74661171e-01
1.02028954e+00 -2.10006371e-01 -6.64618254e-01 9.96112004e-02
6.27708316e-01 -6.66919723e-02 8.34458351e-01 -4.05223370e-01
-8.72531116e-01 -3.52012426e-01 -4.87638384e-01 4.69415367e-01
-5.49772739e-01 -5.32386720e-01 3.96198481e-01 1.06816912e+00
-7.63248026e-01 4.25961137e-01 -1.33765423e+00 -9.84015405e-01
3.80339652e-01 9.35551226e-02 3.42025757e-02 -1.84010305e-02
-1.35303736e+00 8.81069303e-01 -1.11190699e-01 -1.48211464e-01
-7.13671505e-01 -8.85126173e-01 -6.40348375e-01 2.46139839e-01
-2.79111773e-01 -6.25619411e-01 4.22200739e-01 -3.13844234e-01
-8.85919273e-01 1.86761886e-01 -3.22479308e-01 -7.83600137e-02
4.74425673e-01 1.24754921e-01 -6.23197317e-01 -2.83937842e-01
5.09582281e-01 3.96553427e-01 -9.63784754e-02 -8.68539512e-01
-6.50483668e-01 -6.53998613e-01 -4.31776345e-01 1.79868728e-01
-4.37155902e-01 5.97164743e-02 1.68912336e-01 -4.38662738e-01
1.11370184e-01 -9.38817799e-01 -5.30287027e-01 -3.63544524e-01
-5.81550188e-02 -2.55662531e-01 3.73838097e-01 -8.57689559e-01
1.98095393e+00 -1.99404168e+00 -3.98664057e-01 6.37571275e-01
-1.52337730e-01 -3.76491487e-01 3.71400476e-01 1.15147638e+00
4.08533663e-01 4.69249606e-01 -4.25751507e-01 -3.30931321e-02
1.91354930e-01 1.73504800e-01 1.33262008e-01 7.58152068e-01
-2.66358376e-01 4.64576393e-01 -9.55421507e-01 -5.11002004e-01
4.10777867e-01 5.06345212e-01 -7.40082026e-01 -1.95465088e-01
3.92027289e-01 5.06842911e-01 -4.85553145e-01 4.19293404e-01
6.97895229e-01 1.71127543e-01 2.51264185e-01 5.01480281e-01
-8.32294047e-01 7.92845041e-02 -1.06341302e+00 9.13640141e-01
-1.01364531e-01 3.46786767e-01 -2.52393693e-01 -5.07417023e-01
8.03705394e-01 2.72775590e-01 8.68401408e-01 -8.51157725e-01
-3.00469369e-01 1.82595523e-03 1.48269292e-02 -7.77053416e-01
4.13234830e-01 -9.81614664e-02 -2.13134527e-01 8.66782129e-01
-7.52778172e-01 3.12445849e-01 -4.11882102e-02 1.30789906e-01
1.11074984e+00 5.09959310e-02 3.95828664e-01 -6.66067004e-01
1.53848082e-01 1.74980648e-02 7.57767260e-01 4.32238162e-01
-4.07448560e-01 -5.40606491e-02 2.83053458e-01 -1.77729949e-01
-1.18008173e+00 -1.23661411e+00 -3.70208293e-01 1.26117527e+00
1.43905491e-01 3.13508674e-04 -6.03967249e-01 1.41531751e-01
1.95924670e-01 1.12467170e+00 -5.50803661e-01 -4.72783074e-02
-7.12153554e-01 -1.24006414e+00 6.51291072e-01 5.14355481e-01
8.05280447e-01 -7.69032061e-01 -7.76338398e-01 2.77711630e-01
-4.94819582e-01 -2.98191249e-01 -2.65275687e-01 -3.99594933e-01
-8.65006924e-01 -7.76922703e-01 -7.41820514e-01 -3.90225321e-01
8.18456948e-01 3.79224628e-01 7.06862628e-01 2.58039236e-01
1.81702182e-01 5.90956867e-01 -4.05159086e-01 -3.51604253e-01
-1.25626624e-01 2.56828815e-01 2.42992699e-01 -2.29584277e-01
5.47467470e-01 -1.10393214e+00 -1.05184567e+00 6.57802939e-01
-5.45948088e-01 -1.03654936e-02 -8.28672275e-02 -2.46730335e-02
4.24230210e-02 3.09201866e-01 9.23926413e-01 -1.72563732e-01
2.95682251e-01 -1.32990277e+00 5.53304479e-02 1.64668933e-01
-8.90778959e-01 -5.56592226e-01 -1.20718196e-01 -3.92788410e-01
-1.07698977e+00 -1.40129849e-01 -4.72025946e-02 7.19276845e-01
-2.81686753e-01 5.55247605e-01 -4.47801232e-01 5.98979056e-01
4.30564553e-01 9.26811546e-02 -1.68548346e-01 -1.37943730e-01
-1.81516021e-01 6.85636699e-01 -3.80134583e-02 -5.26748478e-01
7.19652951e-01 6.60666108e-01 -4.89107408e-02 -1.19570911e+00
1.84301168e-01 -5.07903934e-01 -6.81375682e-01 -7.14736700e-01
9.78627086e-01 -9.50584292e-01 -1.03352487e+00 6.62135720e-01
-5.91848075e-01 -8.31054091e-01 2.24009246e-01 8.64094913e-01
-2.22909898e-01 1.19014889e-01 -1.26806945e-01 -1.35850811e+00
2.52396092e-02 -6.01930559e-01 3.96600604e-01 1.85621962e-01
-5.18645644e-01 -1.25005090e+00 4.76596326e-01 1.76259443e-01
5.28348088e-01 6.89938009e-01 1.00126624e+00 -2.46674806e-01
-4.13128883e-01 1.87514246e-01 -6.26372695e-02 -4.81757790e-01
9.31691974e-02 8.11691135e-02 -5.42096257e-01 -1.04660057e-01
-3.35528493e-01 4.61983949e-01 4.39803302e-01 6.90420449e-01
3.46239239e-01 -4.07686949e-01 -7.76517332e-01 2.47915238e-01
1.38256550e+00 3.16032648e-01 1.01650476e+00 3.53736103e-01
6.69174075e-01 9.70484853e-01 3.07483852e-01 8.07350814e-01
1.06864643e+00 3.92065942e-01 2.16711953e-01 -3.74148190e-02
2.36433506e-01 -4.85278189e-01 6.05412960e-01 4.73133743e-01
-5.55544674e-01 -2.90236682e-01 -1.56371987e+00 1.06609023e+00
-1.92763364e+00 -1.59805334e+00 -8.03929746e-01 2.27068591e+00
4.14335400e-01 -4.18860525e-01 5.70073247e-01 2.34074727e-01
5.30909956e-01 -5.22678234e-02 -4.12159532e-01 3.95737123e-03
1.01004787e-01 -2.96471775e-01 8.43559325e-01 6.62876666e-01
-3.90119523e-01 5.33928812e-01 6.57631493e+00 4.01945859e-01
-5.64793408e-01 2.34661952e-01 6.00129187e-01 -1.35950938e-01
-7.67724872e-01 1.15782283e-01 -7.89954722e-01 6.82185590e-01
1.30257130e+00 -3.47073734e-01 3.29477519e-01 3.11030984e-01
1.41412175e+00 -6.78873897e-01 -6.41456246e-01 3.11326325e-01
-3.80423367e-01 -7.73999155e-01 -5.55923343e-01 5.95165551e-01
7.37147629e-01 -1.47502214e-01 -3.78083885e-01 3.67714524e-01
5.93785584e-01 -9.31869388e-01 7.47298419e-01 1.05342722e+00
4.66602892e-01 -8.11947942e-01 3.05644631e-01 6.29020154e-01
-1.52811563e+00 -1.48740515e-01 1.16750218e-01 -5.57695508e-01
7.04009235e-01 5.70971906e-01 -7.43546128e-01 3.89209762e-02
7.66824484e-01 5.16812921e-01 -3.62914354e-01 7.83534169e-01
-3.30396704e-02 1.18667686e+00 -6.55082822e-01 -1.06220186e-01
1.17078170e-01 -3.60433340e-01 2.95596123e-01 1.20548356e+00
7.11549401e-01 6.18252993e-01 1.73704326e-01 7.04123557e-01
5.54180384e-01 9.28430259e-02 -7.59616315e-01 1.17518090e-01
9.33049500e-01 5.40941000e-01 -8.73951316e-01 -1.67697683e-01
-4.39228654e-01 2.49765381e-01 -2.16589838e-01 6.84890926e-01
-9.11422610e-01 -1.05423570e-01 8.14073622e-01 8.97931635e-01
-1.81882322e-01 -6.67223275e-01 -3.75070810e-01 -5.54668486e-01
-4.70094025e-01 -1.02886274e-01 1.04702845e-01 -4.58688289e-01
-8.41450691e-01 -3.45010221e-01 6.82018876e-01 -7.14497507e-01
-9.78648961e-02 1.81883872e-01 -1.00529253e+00 9.79961932e-01
-1.09737945e+00 -1.40378237e+00 -3.25453252e-01 6.61105096e-01
8.74886215e-02 3.10268849e-01 5.83240807e-01 5.98028123e-01
-7.36307323e-01 1.72898799e-01 3.63803238e-01 -1.03250798e-02
2.01401323e-01 -7.79335678e-01 3.58438671e-01 7.41665423e-01
-8.60663354e-01 7.84099579e-01 5.57385564e-01 -1.44439507e+00
-9.48122442e-01 -1.16052377e+00 1.43711889e+00 -3.35787088e-01
6.29902065e-01 -2.20859751e-01 -6.51759148e-01 5.26072502e-01
-3.91901433e-02 -7.90398300e-01 1.16142499e+00 2.50410587e-01
4.22762454e-01 9.46886465e-02 -1.37250161e+00 9.07174051e-01
1.25983465e+00 -3.10630351e-01 -2.24167719e-01 1.33955581e-02
5.43534100e-01 5.94755352e-01 -1.23552573e+00 1.79152340e-01
9.95250165e-01 -8.26725066e-01 8.45348597e-01 -1.31461427e-01
4.12025228e-02 -2.45797276e-01 -3.95787627e-01 -8.44287097e-01
-7.13411272e-01 -2.36895040e-01 4.18454409e-01 1.49739683e+00
2.56766886e-01 -7.59043396e-01 5.42749345e-01 1.22567844e+00
-1.34387851e-01 -3.45661372e-01 -7.81541884e-01 -5.52276611e-01
2.54010350e-01 -4.65960562e-01 1.16773164e+00 1.05510914e+00
1.39519781e-01 -2.74657756e-01 -1.84668675e-01 2.81913698e-01
6.37006760e-01 -5.18631399e-01 9.42020655e-01 -1.04129040e+00
4.07508254e-01 -3.51131171e-01 -3.80192548e-02 -4.57236320e-01
-1.45024031e-01 -4.95825082e-01 -2.58890867e-01 -1.86711407e+00
5.37357569e-01 -7.08654642e-01 3.21497381e-01 4.67337340e-01
1.88354310e-02 -2.75080621e-01 -1.53616816e-01 2.76922047e-01
1.83432385e-01 6.88712299e-01 7.46885121e-01 1.37072429e-01
-7.89734304e-01 2.73406059e-01 -3.04411203e-01 5.52028179e-01
8.35844219e-01 -5.62986612e-01 -4.91563827e-01 -1.20748788e-01
5.00515580e-01 2.90221244e-01 4.91627485e-01 -1.08554542e+00
1.74011260e-01 -8.17695200e-01 2.61532813e-01 -8.22870433e-01
-1.50344968e-01 -1.08888280e+00 8.62104654e-01 9.97806728e-01
5.36574572e-02 1.40057623e-01 -4.55831029e-02 5.98696470e-01
5.07117689e-01 8.41267928e-02 3.33049238e-01 9.82906520e-02
-8.18188936e-02 2.83093482e-01 -1.24013102e+00 -2.93450534e-01
1.10778379e+00 -8.91931057e-01 -1.81831524e-01 -2.73585379e-01
-8.62802207e-01 5.39467633e-01 8.63127887e-01 1.40441756e-03
1.38852403e-01 -1.54248703e+00 -7.31121480e-01 -1.33883115e-02
-1.74928367e-01 -3.76597941e-01 4.55792069e-01 1.14906144e+00
-4.84439135e-01 2.08306909e-01 -3.36390108e-01 -2.32884839e-01
-7.59239078e-01 1.78448811e-01 2.19925463e-01 -8.48448370e-03
-3.58009636e-01 4.37169448e-02 -7.65183046e-02 -3.08821142e-01
-2.96723582e-02 -4.92930502e-01 -2.87947088e-01 4.01933014e-01
4.28215653e-01 1.64890397e+00 -5.20180464e-01 -9.61775362e-01
-4.99823451e-01 5.15530646e-01 8.37138891e-01 -5.18688142e-01
1.35116255e+00 -7.52927542e-01 -1.46567613e-01 7.65960574e-01
7.44452715e-01 -5.64966798e-02 -1.14861679e+00 2.01429546e-01
-7.41351545e-02 -3.40871394e-01 -1.97038427e-01 -3.48233491e-01
-4.67198461e-01 6.51218414e-01 7.67485321e-01 4.12802249e-01
8.86413872e-01 -2.80339271e-01 7.91787326e-01 1.65888265e-01
4.20452148e-01 -1.23863018e+00 -2.47734696e-01 3.74966472e-01
5.71800351e-01 -7.38923013e-01 9.23359394e-02 -7.35574439e-02
-4.33637053e-01 3.99649024e-01 4.27518725e-01 1.37566641e-01
1.24224973e+00 5.72754852e-02 -5.90037227e-01 6.37942478e-02
-3.61467212e-01 -3.59616995e-01 -3.85317266e-01 8.58898580e-01
5.57993352e-01 5.90948701e-01 -5.06747186e-01 5.15480042e-01
-1.85157046e-01 -6.08682819e-02 3.14941168e-01 6.28831506e-01
-8.26987565e-01 -7.61547327e-01 -7.26652861e-01 3.63086492e-01
-1.58321530e-01 6.31298348e-02 2.33168174e-02 7.93033481e-01
3.47256511e-01 1.27811801e+00 3.33654106e-01 -1.29649103e-01
5.45406342e-02 1.91257045e-01 -3.65885794e-02 -3.02328736e-01
-7.61889338e-01 -5.04419804e-02 1.88354611e-01 -2.69452572e-01
-5.98685205e-01 -1.26506329e+00 -1.59164047e+00 -1.18330705e+00
2.62468792e-02 7.46092498e-02 5.79817295e-01 9.93591011e-01
1.65072978e-01 1.88826948e-01 6.38655961e-01 -8.48979414e-01
2.31527418e-01 -1.19909263e+00 -7.07589686e-01 1.18655875e-01
-3.42422053e-02 -5.36732376e-01 -2.94549733e-01 1.57426983e-01] | [6.409538269042969, 1.8763974905014038] |
9bc6268b-cee8-4573-908c-3b4f402d8ac4 | inkorrect-online-handwriting-spelling | 2202.13794 | null | https://arxiv.org/abs/2202.13794v1 | https://arxiv.org/pdf/2202.13794v1.pdf | Inkorrect: Online Handwriting Spelling Correction | We introduce Inkorrect, a data- and label-efficient approach for online handwriting (Digital Ink) spelling correction - DISC. Unlike previous work, the proposed method does not require multiple samples from the same writer, or access to character level segmentation. We show that existing automatic evaluation metrics do not fully capture and are not correlated with the human perception of the quality of the spelling correction, and propose new ones that correlate with human perception. We additionally surface an interesting phenomenon: a trade-off between the similarity and recognizability of the spell-corrected inks. We further create a family of models corresponding to different points on the Pareto frontier between those two axes. We show that Inkorrect's Pareto frontier dominates the points that correspond to prior work. | ['Claudiu Musat', 'Jesse Berent', 'Henry Rowley', 'Andrii Maksai'] | 2022-02-28 | null | null | null | null | ['spelling-correction'] | ['natural-language-processing'] | [ 2.03088894e-01 -4.01770502e-01 -3.65380079e-01 -2.72356808e-01
-4.53326285e-01 -1.12659800e+00 6.12406552e-01 5.37701666e-01
-4.43462074e-01 5.76139331e-01 -3.38748656e-02 -1.03519224e-01
-5.06968677e-01 -4.40109581e-01 -4.19807762e-01 -3.74850452e-01
4.53427404e-01 7.89856076e-01 3.32090467e-01 1.15388207e-01
9.29363489e-01 7.40935802e-01 -1.11543107e+00 2.68253177e-01
1.21293521e+00 6.39764965e-01 3.63263376e-02 1.22207379e+00
-3.09824914e-01 5.27609408e-01 -1.05844855e+00 -7.35891104e-01
2.98531592e-01 -5.97666085e-01 -6.04582250e-01 1.94144741e-01
9.36184287e-01 -8.38131532e-02 2.74059344e-02 1.28663123e+00
3.40941370e-01 -1.61574364e-01 1.06120610e+00 -1.13386488e+00
-1.05312443e+00 3.64937007e-01 -6.11130238e-01 -3.46146971e-02
3.82543147e-01 -2.03022342e-02 8.74541461e-01 -7.16099381e-01
8.91868234e-01 1.02472019e+00 8.31324339e-01 4.80507851e-01
-9.46080804e-01 -3.12087387e-01 -1.74742639e-01 4.44033928e-02
-1.14004576e+00 -1.09100170e-01 4.71688986e-01 -7.65731335e-01
1.13855466e-01 5.63693643e-01 6.31844938e-01 9.11538839e-01
9.76091698e-02 9.63960111e-01 1.73343027e+00 -6.74478829e-01
3.25708091e-01 2.62778431e-01 5.86911976e-01 5.36769867e-01
3.81239831e-01 -1.81505755e-01 -6.74196422e-01 4.83427346e-02
7.63656974e-01 -1.76362410e-01 -2.96516508e-01 -3.29807222e-01
-9.44935262e-01 1.29720747e-01 -1.87663838e-01 5.59826076e-01
-7.72813484e-02 5.72273694e-02 2.09194258e-01 3.59185845e-01
3.77097696e-01 9.14901853e-01 -2.30515990e-02 -5.06698489e-01
-1.44210482e+00 2.61107922e-01 9.26869154e-01 1.02173424e+00
5.62215090e-01 -2.67885506e-01 -5.95271707e-01 7.05978751e-01
-4.70122471e-02 5.40883243e-01 3.05267900e-01 -7.87971616e-01
3.46614182e-01 6.56470537e-01 5.89366436e-01 -8.22192311e-01
-1.91059053e-01 -3.88861597e-01 -3.94280881e-01 3.68412942e-01
1.09296036e+00 1.34023562e-01 -8.48603845e-01 1.02786970e+00
-3.06003779e-01 -1.50319055e-01 -2.25933537e-01 1.12784564e+00
3.43906060e-02 1.64923519e-01 -1.35075390e-01 -1.59573063e-01
1.00338531e+00 -1.13564110e+00 -8.39277923e-01 1.71026364e-01
2.42659703e-01 -9.19466436e-01 1.48901367e+00 7.72807837e-01
-1.15511620e+00 -5.27840436e-01 -1.18805683e+00 3.21083553e-02
-4.07522053e-01 3.18169296e-01 2.95932710e-01 1.11787689e+00
-1.00927675e+00 1.21496594e+00 -5.11636078e-01 -1.85727850e-01
2.82748669e-01 3.81809883e-02 -1.85632128e-02 7.58620501e-02
-5.00046909e-01 8.59356165e-01 1.50303155e-01 -1.21128283e-01
-7.25420654e-01 -4.70273584e-01 8.59045982e-02 -2.71227390e-01
1.78646654e-01 -3.02109029e-03 1.02522326e+00 -1.46995914e+00
-1.60911930e+00 1.06323755e+00 -7.55771399e-02 -1.67561620e-01
1.37547159e+00 -4.44501191e-01 -4.71934199e-01 -7.44910818e-03
-1.69179663e-01 4.13178951e-02 9.45775092e-01 -1.64714229e+00
-6.51616633e-01 -3.32738638e-01 -2.93988407e-01 2.66090315e-02
-3.97222400e-01 6.08398505e-02 -7.07416475e-01 -7.55848229e-01
-1.03392370e-01 -7.65890777e-01 2.81246156e-01 9.12205651e-02
-5.04040837e-01 -2.20001072e-01 6.27920628e-01 -7.15378165e-01
1.30601263e+00 -1.88995993e+00 1.39188752e-01 6.31647289e-01
2.53633857e-01 7.17198730e-01 -1.39844328e-01 3.73865187e-01
3.92991662e-01 4.23332930e-01 -2.06638455e-01 -4.52737719e-01
2.76087344e-01 2.33627539e-02 -1.90000743e-01 6.12094700e-01
-1.19948201e-01 5.96508086e-01 -1.21498871e+00 -5.10559559e-01
-1.00977302e-01 -3.96775231e-02 2.06182320e-02 9.88901258e-02
-3.48639637e-01 1.31311461e-01 -7.09006609e-03 5.90772688e-01
8.76655936e-01 5.42490035e-02 1.26239255e-01 1.99823469e-01
-4.38509494e-01 -2.59342134e-01 -1.21262980e+00 1.30822098e+00
5.14866435e-04 8.98895860e-01 -1.56828091e-01 5.22674173e-02
1.14348745e+00 -7.43662342e-02 -4.66515422e-02 -5.77410460e-01
6.40458688e-02 6.95407152e-01 -9.93121043e-02 -3.48854065e-01
1.03830409e+00 2.71427594e-02 3.06615710e-01 7.55541503e-01
-1.61143467e-01 -1.03898413e-01 4.79265422e-01 -1.01551659e-01
9.02025878e-01 1.11482441e-01 5.23094507e-03 -5.77936351e-01
3.53217870e-01 -9.71113592e-02 9.92494971e-02 1.01715946e+00
-3.53606492e-01 7.77760565e-01 8.88216138e-01 -7.17629343e-02
-1.18339586e+00 -1.06190825e+00 7.45088384e-02 1.04243147e+00
4.48395789e-01 -1.70350268e-01 -1.32002151e+00 -7.94542789e-01
1.04160815e-01 6.01585150e-01 -7.45263159e-01 3.23618293e-01
-4.35027599e-01 -3.49859715e-01 9.59157109e-01 5.22983193e-01
2.41547897e-01 -8.13288748e-01 -4.89036143e-01 -2.08341733e-01
2.10714519e-01 -7.98887789e-01 -7.07455039e-01 -4.03311327e-02
-7.83048213e-01 -1.07777619e+00 -1.17952585e+00 -3.97883832e-01
8.42022002e-01 -3.44766118e-02 1.03930950e+00 3.33182514e-01
1.15158081e-01 4.07375991e-01 -4.86820161e-01 -4.66812462e-01
-7.36875832e-01 1.87332198e-01 -3.79273272e-03 1.26102492e-01
3.06245238e-01 -2.19785925e-02 -4.78304684e-01 5.67655385e-01
-7.04592824e-01 -2.09232122e-01 5.87808311e-01 4.35579538e-01
6.29195333e-01 -1.99737608e-01 2.09677041e-01 -9.04715359e-01
1.08485651e+00 2.77736466e-02 -5.83005488e-01 7.46140361e-01
-8.68895352e-01 1.49161711e-01 5.82184136e-01 -5.05434513e-01
-8.67657661e-01 -2.30559781e-01 2.60959864e-01 -5.89543462e-01
-1.49527624e-01 -2.31789127e-02 -5.59027866e-02 -2.26588130e-01
7.16230094e-01 1.48845151e-01 -3.71648043e-01 -6.61805272e-01
2.93918252e-01 9.27109838e-01 8.87851954e-01 -7.45841920e-01
5.15047371e-01 3.79312277e-01 -7.13438392e-02 -6.83065355e-01
-5.19201458e-01 -4.38661397e-01 -8.84714365e-01 -5.59557259e-01
5.65557063e-01 -2.48797126e-02 -8.71244967e-01 8.95124435e-01
-1.27967536e+00 -4.32487816e-01 -5.20263791e-01 2.16091007e-01
-7.10025370e-01 7.50826061e-01 -7.01593161e-01 -1.12726426e+00
-1.73923001e-01 -7.34691322e-01 8.47176611e-01 3.91269594e-01
-4.54121709e-01 -9.06775832e-01 3.13086987e-01 1.57141924e-01
6.10545725e-02 1.89122275e-01 9.55659032e-01 -6.08119369e-01
-3.98323536e-01 -3.87899935e-01 -5.58763742e-01 3.86275828e-01
1.03638083e-01 3.69061321e-01 -9.52672422e-01 -1.93155438e-01
-3.03417683e-01 3.38885598e-02 8.16725850e-01 1.47252798e-01
1.02448988e+00 -1.48743257e-01 3.65816690e-02 3.55231553e-01
1.44540071e+00 3.04218024e-01 8.53296041e-01 2.46507347e-01
8.47661793e-01 7.90941834e-01 6.64381862e-01 3.34683925e-01
-1.63763344e-01 5.60578644e-01 2.08893269e-02 9.74846482e-02
-4.47364569e-01 -6.35812223e-01 3.25928539e-01 8.09732139e-01
-5.58926165e-01 -6.95912540e-01 -1.10548949e+00 5.80253363e-01
-1.77073658e+00 -5.62449038e-01 -6.28826857e-01 2.25373149e+00
7.60130286e-01 2.48898149e-01 3.54981661e-01 4.61661577e-01
9.13298130e-01 -7.60268271e-02 -4.54058290e-01 -7.09494352e-01
-3.07620198e-01 1.29792869e-01 9.48172510e-01 8.75037432e-01
-6.87048495e-01 1.10414362e+00 7.28058767e+00 1.12373078e+00
-9.05360579e-01 3.86522003e-02 3.19352776e-01 1.33905813e-01
-3.77417833e-01 2.57644542e-02 -5.26692033e-01 6.70837104e-01
6.31610513e-01 -8.18870962e-02 4.60934430e-01 4.53968644e-01
3.69433835e-02 -2.38028809e-01 -1.30032516e+00 8.85244668e-01
3.19933295e-01 -1.06662488e+00 1.79063946e-01 1.44431755e-01
1.07315159e+00 -5.98829627e-01 2.15497717e-01 -4.81924564e-01
2.50223160e-01 -1.01949406e+00 1.26399219e+00 1.15865970e+00
9.49867129e-01 -6.71880245e-01 6.61891460e-01 2.03176782e-01
-7.16994166e-01 1.90320358e-01 -1.21354975e-01 1.04245611e-01
-1.58880875e-01 4.00198221e-01 -7.08863258e-01 2.78989524e-01
1.07516065e-01 4.80602771e-01 -9.29738760e-01 1.16257274e+00
-2.97852367e-01 3.70292753e-01 2.93275923e-01 -4.44245428e-01
6.85630441e-02 -2.41438285e-01 6.00783169e-01 1.59161925e+00
2.86460370e-01 -2.93287218e-01 -2.37225085e-01 8.74626696e-01
-1.23101242e-01 1.98703483e-01 -1.44433066e-01 -4.49386120e-01
3.66348833e-01 7.99770713e-01 -1.11846411e+00 -3.96162748e-01
1.60627469e-01 1.49640858e+00 9.17579085e-02 1.97221115e-01
-4.54033524e-01 -5.94652832e-01 2.50350058e-01 1.86219886e-01
-1.11479193e-01 -4.47868973e-01 -9.67146218e-01 -6.70170307e-01
1.09379724e-01 -9.37639713e-01 -7.36462250e-02 -6.32046759e-01
-1.34289360e+00 3.75613570e-01 -4.75066036e-01 -1.08818173e+00
1.97937980e-01 -9.91433203e-01 -3.94875437e-01 8.71505618e-01
-1.25541830e+00 -7.05650389e-01 -4.49161470e-01 3.68455768e-01
3.95327717e-01 -2.21791729e-01 5.33698797e-01 1.88245937e-01
-3.18610311e-01 8.84751797e-01 4.86198664e-01 1.46959443e-02
8.94159257e-01 -1.68442678e+00 3.90027136e-01 8.57888103e-01
1.40563697e-01 4.89002824e-01 9.03882205e-01 -9.52991247e-01
-1.10088766e+00 -6.69497430e-01 8.02892625e-01 -9.58341420e-01
6.25172317e-01 -1.01890802e-01 -8.48294139e-01 2.09034562e-01
1.08373836e-01 -4.77052063e-01 5.55305064e-01 9.44834948e-03
-4.75602627e-01 1.40485331e-01 -1.11666167e+00 4.83565360e-01
1.04042315e+00 -6.80303335e-01 -4.15564328e-01 1.51866645e-01
-1.42602339e-01 -3.02529335e-01 -7.15309560e-01 -2.13558465e-01
8.99955332e-01 -1.18348968e+00 3.03885162e-01 -5.77950001e-01
6.68606043e-01 -1.64522037e-01 -5.21487892e-02 -1.25693119e+00
-2.19916791e-01 -7.15230525e-01 -1.28069026e-02 1.33816099e+00
2.71452457e-01 -2.34048560e-01 9.93786514e-01 4.80387241e-01
-4.68883589e-02 -4.94823515e-01 -5.34102559e-01 -1.20710528e+00
1.46009654e-01 -2.71088600e-01 6.14858210e-01 8.95782471e-01
5.09557547e-03 -2.99745262e-01 -4.69363987e-01 -1.64049909e-01
7.30598211e-01 1.52887672e-01 7.47749269e-01 -1.44109142e+00
-3.22415203e-01 -9.66603816e-01 -3.34244192e-01 -1.03924656e+00
-1.47682652e-01 -7.82668889e-01 1.41286209e-01 -1.45463824e+00
3.76960784e-01 -5.03875554e-01 -3.03456157e-01 1.23130389e-01
-6.93716109e-02 2.91423142e-01 6.28669500e-01 6.78947031e-01
-7.21744061e-01 -1.38419807e-01 1.44225752e+00 2.32253782e-02
-2.42969722e-01 -9.20874327e-02 -4.22977388e-01 9.92958307e-01
6.51050806e-01 -5.03910720e-01 7.72019774e-02 -5.14430165e-01
4.12165791e-01 -9.38472822e-02 3.04388195e-01 -1.00639498e+00
2.77504742e-01 -2.60508686e-01 3.28189403e-01 -4.51715112e-01
-1.17528014e-01 -5.82338989e-01 -1.31392255e-01 3.82283539e-01
-5.85741282e-01 8.49635750e-02 -1.65048227e-01 6.25093579e-01
1.10342994e-01 -8.05819511e-01 8.64856064e-01 4.49427366e-02
-4.58176613e-01 -1.30515054e-01 -4.16708231e-01 -1.74236074e-02
8.58733475e-01 -5.75417876e-01 -5.12496352e-01 -8.18797648e-02
-3.74857962e-01 -9.76039320e-02 9.72920656e-01 4.73573208e-01
3.35956484e-01 -1.08827698e+00 -6.08451068e-01 -8.76905099e-02
2.97497511e-02 -7.72440612e-01 -2.06762943e-02 5.65717578e-01
-1.09164333e+00 1.05462179e-01 -6.08077705e-01 -3.54063839e-01
-1.27512074e+00 1.77847207e-01 3.19335401e-01 -2.23971412e-01
-1.63584292e-01 6.43435419e-01 -5.14714479e-01 -1.19030848e-01
2.92674810e-01 -3.23157728e-01 1.08203217e-02 3.09461117e-01
5.35916805e-01 1.17396867e+00 8.68367106e-02 -4.16046828e-01
-1.65395066e-01 8.13942552e-01 -2.32350938e-02 -3.62572700e-01
7.90498912e-01 1.07389390e-01 -1.70876503e-01 7.08250999e-01
6.25758886e-01 5.54026961e-01 -1.38686526e+00 8.59240144e-02
4.97352183e-01 -9.17634428e-01 -3.58642012e-01 -1.18320596e+00
-6.80631280e-01 9.52971280e-01 9.14949596e-01 2.79032499e-01
7.89934337e-01 -2.99798936e-01 5.97651422e-01 2.91518211e-01
3.08753908e-01 -1.74417615e+00 2.51599759e-01 5.28409064e-01
8.77261877e-01 -7.91585207e-01 4.40014601e-02 -3.34845304e-01
-6.10698521e-01 1.30062962e+00 4.17965919e-01 -3.16962093e-01
3.51762064e-02 4.95086670e-01 2.13978842e-01 -8.08639079e-02
-1.03088163e-01 -7.34843779e-03 4.00244266e-01 6.29594028e-01
3.87266278e-01 4.46451634e-01 -7.22790718e-01 6.73171580e-01
-2.35610723e-01 1.42001256e-01 6.38198674e-01 7.03599334e-01
-4.28331763e-01 -1.27727425e+00 -5.40739357e-01 3.48595917e-01
-3.89231950e-01 1.49479136e-01 -1.25650251e+00 5.21236241e-01
4.00228262e-01 6.68433011e-01 6.23130202e-02 -3.40727389e-01
4.70112681e-01 6.13726117e-02 8.37024152e-01 -3.20365489e-01
-7.34437644e-01 -1.66525438e-01 -1.65404215e-01 -1.65871307e-01
-1.97231263e-01 -5.82801402e-01 -9.33390200e-01 -5.20156503e-01
-3.97100389e-01 1.19059853e-01 6.63053393e-01 7.17140794e-01
1.12474402e-02 2.00281546e-01 3.11435431e-01 -7.26703763e-01
-4.10472929e-01 -7.67121017e-01 -9.89893317e-01 4.91498739e-01
2.21129268e-01 -2.89096177e-01 -1.40745476e-01 -4.91203293e-02] | [11.861337661743164, 2.4577767848968506] |
54e6064c-e0c3-4ad3-b713-8efc66591903 | sugar-spherical-ultrafast-graph-attention | 2307.00511 | null | https://arxiv.org/abs/2307.00511v1 | https://arxiv.org/pdf/2307.00511v1.pdf | SUGAR: Spherical Ultrafast Graph Attention Framework for Cortical Surface Registration | Cortical surface registration plays a crucial role in aligning cortical functional and anatomical features across individuals. However, conventional registration algorithms are computationally inefficient. Recently, learning-based registration algorithms have emerged as a promising solution, significantly improving processing efficiency. Nonetheless, there remains a gap in the development of a learning-based method that exceeds the state-of-the-art conventional methods simultaneously in computational efficiency, registration accuracy, and distortion control, despite the theoretically greater representational capabilities of deep learning approaches. To address the challenge, we present SUGAR, a unified unsupervised deep-learning framework for both rigid and non-rigid registration. SUGAR incorporates a U-Net-based spherical graph attention network and leverages the Euler angle representation for deformation. In addition to the similarity loss, we introduce fold and multiple distortion losses, to preserve topology and minimize various types of distortions. Furthermore, we propose a data augmentation strategy specifically tailored for spherical surface registration, enhancing the registration performance. Through extensive evaluation involving over 10,000 scans from 7 diverse datasets, we showed that our framework exhibits comparable or superior registration performance in accuracy, distortion, and test-retest reliability compared to conventional and learning-based methods. Additionally, SUGAR achieves remarkable sub-second processing times, offering a notable speed-up of approximately 12,000 times in registering 9,000 subjects from the UK Biobank dataset in just 32 minutes. This combination of high registration performance and accelerated processing time may greatly benefit large-scale neuroimaging studies. | ['Hesheng Liu', 'Danhong Wang', 'Dan Hu', 'Ping Zhang', 'Qingyu Hu', 'Wei zhang', 'Ying Zhou', 'Weiwei Wang', 'Weigang Cui', 'Cong Lin', 'Zhenyu Sun', 'Danyang Wang', 'Youjia Zhang', 'Ning An', 'Jianxun Ren'] | 2023-07-02 | null | null | null | null | ['graph-attention'] | ['graphs'] | [ 1.42487377e-01 1.90942839e-01 -4.50188387e-03 -5.98581135e-01
-1.09794962e+00 -4.11777735e-01 5.38874030e-01 2.47495502e-01
-5.19379377e-01 4.03366655e-01 5.38918853e-01 1.60114720e-01
-2.64346451e-01 -4.79781479e-01 -5.68207741e-01 -4.66011763e-01
-4.30044323e-01 4.67235565e-01 -2.83837188e-02 8.69496241e-02
1.97191000e-01 5.68665028e-01 -8.52616787e-01 -1.41475111e-01
9.13950861e-01 7.38001049e-01 -1.20177217e-01 -1.88003592e-02
3.59726876e-01 1.21476669e-02 4.71583195e-02 -5.74345231e-01
3.95976067e-01 -1.01517677e-01 -9.33776438e-01 -2.79577047e-01
9.31812406e-01 -2.78209358e-01 -5.08698225e-01 8.30210507e-01
9.27189827e-01 3.30080688e-01 5.53214431e-01 -6.34229422e-01
-7.67632961e-01 4.62332278e-01 -6.21233821e-01 4.06694561e-01
2.60565549e-01 1.11714400e-01 9.08191979e-01 -6.98344409e-01
7.44307280e-01 7.98908770e-01 1.05333972e+00 5.89830339e-01
-1.61477208e+00 -5.78024387e-01 -1.36194378e-01 -7.74512738e-02
-1.42126322e+00 -8.09758663e-01 6.28240466e-01 -4.94796455e-01
1.17016542e+00 4.53860313e-02 7.23216891e-01 7.92258501e-01
4.14662600e-01 8.94892141e-02 1.12730014e+00 8.96457303e-03
-3.14667523e-02 -6.71933532e-01 -1.15974955e-01 6.15186155e-01
2.77250379e-01 -4.11865711e-02 -3.10517609e-01 -1.75805077e-01
1.11865807e+00 -1.23516366e-01 -1.08863369e-01 -3.23449463e-01
-1.42765927e+00 5.92877448e-01 8.51044059e-01 3.28121245e-01
-3.86732608e-01 2.51434654e-01 4.50028867e-01 -3.74613963e-02
7.19434500e-01 4.74121749e-01 -2.18008459e-01 8.45655892e-03
-1.26471794e+00 1.98766455e-01 2.77638763e-01 5.49091935e-01
2.80402154e-01 1.21839561e-01 -1.29441665e-02 8.23822081e-01
3.41339409e-01 3.08650106e-01 5.71685195e-01 -8.92353177e-01
5.26684761e-01 5.16562462e-01 -5.01215875e-01 -1.19932783e+00
-9.81281459e-01 -5.46390653e-01 -1.15209770e+00 4.55910675e-02
3.52270693e-01 4.00325209e-02 -7.26420105e-01 2.05020809e+00
3.36165547e-01 1.55894384e-01 -5.69193661e-01 9.38767612e-01
6.26110494e-01 -2.26611882e-01 7.13295043e-02 -1.12501323e-01
1.40168726e+00 -4.61141497e-01 -4.53912944e-01 -1.51146248e-01
5.37420869e-01 -7.43370533e-01 8.49017322e-01 -5.56301442e-04
-1.53058064e+00 -2.64170676e-01 -8.56940985e-01 -3.79675567e-01
-2.25168038e-02 -1.36542886e-01 8.59458506e-01 7.25916803e-01
-1.37996483e+00 6.13646388e-01 -1.25905752e+00 -3.64596725e-01
9.29383576e-01 8.48224521e-01 -7.72341728e-01 1.04496270e-01
-7.32212186e-01 9.91481662e-01 -2.67491728e-01 2.22677544e-01
-3.64244580e-01 -1.27566433e+00 -8.84730279e-01 -7.32085481e-02
-1.55262634e-01 -7.23985076e-01 7.48720229e-01 -4.50962335e-01
-1.34068763e+00 1.13368142e+00 -1.34735882e-01 -4.61596012e-01
5.11738539e-01 -6.72747195e-02 -2.96283871e-01 2.18079180e-01
1.94529846e-01 6.53169632e-01 4.56227988e-01 -5.29963374e-01
3.75611454e-01 -8.51600409e-01 -4.57555622e-01 3.14585030e-01
-2.34602392e-01 2.06812531e-01 -3.03031534e-01 -8.74914825e-01
5.59105337e-01 -8.36807251e-01 -3.44489247e-01 2.02803418e-01
-1.06481835e-01 4.66330945e-02 1.22795455e-01 -9.64380622e-01
6.31352425e-01 -2.00302958e+00 6.96047321e-02 4.84305203e-01
7.06669033e-01 8.02556649e-02 -1.45358145e-01 -4.58651744e-02
-3.00521374e-01 2.34088469e-02 -4.29037601e-01 -3.28634173e-01
-1.95804209e-01 -3.09504509e-01 1.91461980e-01 1.11199868e+00
1.50625840e-01 1.28301716e+00 -7.04515636e-01 -2.30078205e-01
-2.22355388e-02 8.52840424e-01 -9.25806642e-01 -1.08483367e-01
5.04009545e-01 7.59231865e-01 -1.65648505e-01 5.33831239e-01
6.55322194e-01 -1.60081238e-01 3.00269216e-01 -5.03009737e-01
7.94023946e-02 3.99904609e-01 -9.35504735e-01 2.02137327e+00
-2.52282381e-01 5.19164860e-01 8.30260068e-02 -1.17938328e+00
9.16916728e-01 3.61291468e-01 1.01094937e+00 -9.16655362e-01
2.06753537e-01 2.52970219e-01 3.32541704e-01 -2.33313024e-01
1.97503269e-01 -1.69154927e-01 1.76662236e-01 5.49985707e-01
5.90542704e-02 -1.82895139e-02 -1.67062610e-01 -9.59379971e-03
1.15344810e+00 1.59769759e-01 1.71711460e-01 -5.66446781e-01
1.37633741e-01 -4.59936649e-01 5.64629138e-01 2.78438866e-01
-2.93125451e-01 9.37225521e-01 2.51386166e-01 -4.76244062e-01
-1.15307128e+00 -1.20851111e+00 -4.12042767e-01 7.09971130e-01
-2.73145348e-01 -2.76925623e-01 -9.67601180e-01 -1.71689779e-01
2.53424030e-02 -6.63451180e-02 -6.37067139e-01 -1.80943087e-01
-8.41820478e-01 -1.14088321e+00 7.99010634e-01 6.50923550e-01
5.41536272e-01 -7.95797527e-01 -3.29239547e-01 2.72475570e-01
-1.41858459e-01 -1.21433854e+00 -8.92848670e-01 -3.22326571e-01
-1.31218553e+00 -1.03414011e+00 -7.11357117e-01 -8.24297726e-01
9.72926557e-01 5.24404421e-02 9.15330231e-01 5.41663282e-02
-5.02699971e-01 2.92059243e-01 1.19308680e-01 1.00609951e-01
2.67294079e-01 2.93708444e-01 4.22727495e-01 -1.93191633e-01
4.19308245e-02 -1.05899525e+00 -1.05006397e+00 5.05768508e-02
-6.15610480e-01 -2.46451914e-01 4.51813042e-01 5.90700388e-01
7.11925387e-01 -4.85269129e-01 7.12379456e-01 -7.02987790e-01
7.00014472e-01 -3.00167739e-01 -4.30634320e-01 -3.14566419e-02
-8.29628110e-01 1.17676854e-02 3.18093270e-01 -3.65014404e-01
-7.73088694e-01 3.66996378e-02 -2.62128472e-01 8.09052587e-02
6.87759295e-02 5.53230464e-01 1.70448259e-01 -6.90095186e-01
7.56079972e-01 7.17096729e-03 4.31823641e-01 -4.92947459e-01
4.25731629e-01 1.19413659e-01 7.38191605e-01 -5.55544734e-01
7.15838969e-01 6.22021258e-01 1.81060895e-01 -6.80210650e-01
-2.91219741e-01 -2.87916183e-01 -1.02597010e+00 7.44977072e-02
9.05433536e-01 -8.46584141e-01 -8.25124025e-01 4.74094242e-01
-8.48891258e-01 -2.88071334e-01 -7.56897181e-02 6.56793714e-01
-5.37829161e-01 4.13612574e-01 -5.22259712e-01 -1.96614668e-01
-9.78578806e-01 -1.37123370e+00 9.33134556e-01 5.94593249e-02
-3.82086724e-01 -1.11119878e+00 2.10383624e-01 4.68914121e-01
7.16430485e-01 4.55150872e-01 7.94528902e-01 -6.96546137e-01
-3.47866833e-01 -3.53753537e-01 -3.67401004e-01 -9.87548605e-02
1.78740367e-01 -5.82144618e-01 -6.93743944e-01 -4.85150218e-01
-5.62692210e-02 -7.81058595e-02 4.90932196e-01 6.01883233e-01
1.09756279e+00 -1.15025498e-01 -1.18610226e-01 1.07601440e+00
1.22067344e+00 -2.03449816e-01 8.06348383e-01 2.16582865e-01
8.84899557e-01 3.83009166e-01 -2.13279292e-01 -9.61187761e-03
7.17409790e-01 9.17151153e-01 1.05540507e-01 -3.84629160e-01
-4.48979765e-01 1.45706013e-01 9.45764482e-02 1.13623953e+00
-5.26291370e-01 6.22395873e-01 -1.06808555e+00 4.95552570e-01
-1.36696994e+00 -9.17199016e-01 -1.48445055e-01 2.44268346e+00
9.89558160e-01 -1.08062655e-01 2.45538577e-01 -2.38832831e-01
5.30217767e-01 4.06146906e-02 -6.35211229e-01 -1.45576820e-01
-1.18313298e-01 6.73529565e-01 4.64076996e-01 4.00860429e-01
-1.07866251e+00 8.16158175e-01 7.12548828e+00 2.55280256e-01
-1.24420476e+00 4.45072263e-01 5.86837173e-01 -3.47137302e-01
-3.01405996e-01 -2.89510548e-01 -3.95720303e-01 1.74563751e-01
8.72039378e-01 -3.96609455e-02 7.80469298e-01 3.63138497e-01
3.71799707e-01 2.53796667e-01 -9.55385804e-01 9.88231778e-01
3.04108381e-01 -1.50311422e+00 -3.05395573e-01 3.03597629e-01
6.71530187e-01 6.07589304e-01 5.90484124e-03 -3.01862985e-01
-9.23795477e-02 -1.22496343e+00 3.42846155e-01 6.45943999e-01
1.13978148e+00 -4.92743820e-01 5.15847981e-01 -3.29804748e-01
-1.28928840e+00 4.42757279e-01 -2.93006599e-01 3.05282362e-02
1.75011650e-01 6.28571987e-01 -4.16334987e-01 5.86771727e-01
6.05109632e-01 7.05794990e-01 -6.33228898e-01 1.25107098e+00
5.17958216e-02 3.94279540e-01 -3.70213687e-01 5.42706907e-01
-7.69295692e-02 -3.72552752e-01 4.02076572e-01 9.50337291e-01
2.16404453e-01 2.61331141e-01 -6.03228584e-02 9.88298655e-01
-2.99428225e-01 2.17940047e-01 -4.33456212e-01 2.64919758e-01
4.15749073e-01 1.42960322e+00 -8.98613513e-01 1.06911361e-01
-5.08494258e-01 7.82078803e-01 6.22901499e-01 1.96415007e-01
-6.58338010e-01 -3.58409315e-01 6.38203204e-01 3.29863936e-01
-9.26140621e-02 -6.54739916e-01 -7.50378072e-01 -1.17717409e+00
3.09203237e-01 -7.84390450e-01 1.61095053e-01 -3.82302165e-01
-1.39542699e+00 6.54974341e-01 -1.35376707e-01 -8.09457958e-01
5.71821444e-02 -3.33898425e-01 -6.37107491e-01 1.02744579e+00
-1.27880752e+00 -1.30886495e+00 -1.96048215e-01 6.17751420e-01
1.24868704e-02 -1.23856932e-01 7.60396242e-01 6.44261420e-01
-5.45510650e-01 9.62949395e-01 4.76494208e-02 2.91945428e-01
9.44187045e-01 -1.01749611e+00 7.31156349e-01 8.26898754e-01
1.13732673e-01 1.25807869e+00 7.62949213e-02 -7.10506201e-01
-1.50493050e+00 -9.81300533e-01 8.77166569e-01 -3.79930794e-01
7.32176065e-01 -3.80293727e-01 -1.02309752e+00 7.89739847e-01
-1.16830222e-01 3.00042957e-01 7.98088610e-01 2.84068584e-01
-4.46103394e-01 -1.40663013e-01 -1.29564476e+00 6.16996408e-01
1.32559013e+00 -6.49625123e-01 -6.59064114e-01 4.47818965e-01
4.01013374e-01 -5.78046203e-01 -1.47382474e+00 4.15381312e-01
8.52303863e-01 -5.11743069e-01 1.23224199e+00 -4.49794263e-01
1.61992982e-01 7.86879808e-02 1.57061994e-01 -1.14036000e+00
-6.30314708e-01 -7.30976820e-01 2.03422979e-01 1.19821513e+00
1.75073296e-01 -9.15229380e-01 7.68966556e-01 1.10353553e+00
-3.77051800e-01 -8.25018883e-01 -1.26154041e+00 -5.62472701e-01
4.68292236e-01 -3.71523201e-01 5.45800567e-01 1.06949043e+00
1.47343308e-01 -2.93075647e-02 -4.14183065e-02 -4.25570868e-02
8.17042649e-01 -1.98949948e-01 3.32337499e-01 -1.27164006e+00
1.19157083e-01 -6.21435523e-01 -7.53512323e-01 -5.82877755e-01
2.22752646e-01 -1.64889777e+00 -2.50078410e-01 -1.49245608e+00
3.51760775e-01 -5.79911292e-01 -1.80354759e-01 7.18629777e-01
-5.72292693e-02 7.71886766e-01 -9.93061289e-02 3.64665002e-01
-1.83858901e-01 4.75567281e-01 1.18678904e+00 1.62732512e-01
-2.67692924e-01 -3.75728101e-01 -8.98541093e-01 6.62723422e-01
8.04443717e-01 -2.63433188e-01 -2.04074055e-01 -8.34287941e-01
6.64056465e-02 -3.30797642e-01 4.35700178e-01 -9.60033715e-01
3.68446767e-01 3.32109541e-01 5.70736349e-01 -1.69784240e-02
1.04745790e-01 -5.21485686e-01 1.46010071e-01 3.88470829e-01
-2.75518924e-01 2.74075001e-01 2.14648873e-01 1.28869876e-01
2.46883422e-01 4.03695196e-01 8.97630930e-01 8.44145194e-02
-9.43909492e-03 7.45134592e-01 -5.39208800e-02 2.04923436e-01
5.66543937e-01 -2.25782782e-01 -2.29464144e-01 8.36094283e-03
-7.90206909e-01 -1.51194960e-01 3.29728812e-01 2.82178402e-01
5.25339127e-01 -1.52070808e+00 -8.05399597e-01 3.43048811e-01
-2.73834229e-01 -6.76089451e-02 4.02309716e-01 1.47593284e+00
-4.90350991e-01 3.36859822e-01 -5.02152979e-01 -4.58781630e-01
-9.49088633e-01 -6.87534586e-02 4.19257462e-01 -2.27852613e-01
-1.06754196e+00 6.44250512e-01 -3.40410136e-02 -6.45725787e-01
-1.67114228e-01 -1.81674346e-01 -5.02073318e-02 -7.43169039e-02
4.04012829e-01 4.55660850e-01 4.79037941e-01 -8.70083570e-01
-6.82070732e-01 8.23721707e-01 -5.32283150e-02 -1.47312149e-01
1.72917521e+00 -8.06701705e-02 -4.17013168e-01 -1.24797225e-01
1.15659952e+00 3.14018242e-02 -1.06159282e+00 -2.88524628e-01
1.55966058e-02 -1.67650044e-01 3.08327794e-01 -6.89553499e-01
-1.47855461e+00 5.92827857e-01 8.23565006e-01 -3.41783047e-01
1.02909052e+00 -1.64441410e-02 9.11430359e-01 2.12483876e-03
4.05397296e-01 -7.44046211e-01 -1.67694807e-01 3.51027459e-01
9.29491520e-01 -1.03247035e+00 3.95929694e-01 -3.09069544e-01
-2.86773264e-01 9.00042653e-01 3.94537121e-01 -5.31463504e-01
6.59484923e-01 3.90410274e-01 -1.61290616e-01 -5.54378867e-01
-1.42498761e-01 2.20294908e-01 8.36238861e-01 8.41012299e-01
9.10526276e-01 3.02047785e-02 -5.96283197e-01 7.08866596e-01
-4.70218569e-01 -1.21806882e-01 1.16956837e-01 5.09041965e-01
1.20233931e-03 -1.11179948e+00 -1.60137177e-01 7.69371212e-01
-6.76811576e-01 -3.96240711e-01 -1.08977191e-01 6.75183415e-01
-1.80065095e-01 5.22019923e-01 2.04136387e-01 -1.34775206e-01
2.94061095e-01 -1.25630740e-02 8.54582429e-01 -5.53760409e-01
-8.04019332e-01 1.69212356e-01 -1.52600393e-01 -8.91461372e-01
-5.35686493e-01 -9.50543463e-01 -1.38187206e+00 -3.02057207e-01
-1.63204521e-01 -3.56039554e-01 6.47634625e-01 1.10852420e+00
7.14470804e-01 4.30142105e-01 2.35891446e-01 -1.00567567e+00
-3.42693985e-01 -8.49328220e-01 -3.84527147e-01 5.22239685e-01
-3.50512122e-03 -7.18490422e-01 9.98497531e-02 4.27037850e-02] | [13.985238075256348, -2.5164852142333984] |
5714c76d-1436-4966-87a5-6102347185b6 | is-preprocessing-of-text-really-worth-your | 1806.02908 | null | http://arxiv.org/abs/1806.02908v2 | http://arxiv.org/pdf/1806.02908v2.pdf | Is preprocessing of text really worth your time for online comment classification? | A large proportion of online comments present on public domains are
constructive, however a significant proportion are toxic in nature. The
comments contain lot of typos which increases the number of features manifold,
making the ML model difficult to train. Considering the fact that the data
scientists spend approximately 80% of their time in collecting, cleaning and
organizing their data [1], we explored how much effort should we invest in the
preprocessing (transformation) of raw comments before feeding it to the
state-of-the-art classification models. With the help of four models on Jigsaw
toxic comment classification data, we demonstrated that the training of model
without any transformation produce relatively decent model. Applying even basic
transformations, in some cases, lead to worse performance and should be applied
with caution. | ['Fahim Mohammad'] | 2018-06-07 | null | null | null | null | ['toxic-comment-classification'] | ['natural-language-processing'] | [-1.45101503e-01 1.36659727e-01 1.75110735e-02 -3.44871432e-01
-4.33707207e-01 -7.60963321e-01 5.00090003e-01 7.31200516e-01
-4.42634672e-01 5.59965491e-01 4.07052398e-01 -6.51494026e-01
2.77577341e-01 -6.20762169e-01 -3.81533504e-01 -3.54458272e-01
3.65681857e-01 1.19353354e-01 1.01755597e-01 -3.26051354e-01
7.06174731e-01 -5.25155924e-02 -1.40229011e+00 5.80834985e-01
1.12914455e+00 6.26707256e-01 2.34154016e-02 5.51408529e-01
-6.27917826e-01 1.04933739e+00 -9.12123859e-01 -8.75552654e-01
1.95675060e-01 -2.08154187e-01 -9.16379869e-01 1.57266885e-01
2.07200319e-01 -6.19144365e-02 5.30843139e-02 1.03961837e+00
2.61476457e-01 1.98065005e-02 1.02183223e+00 -1.30036497e+00
-1.04451811e+00 6.98091328e-01 -3.07779372e-01 2.31818706e-01
3.76105636e-01 -4.01473977e-02 1.13714266e+00 -7.50957489e-01
7.27905452e-01 1.15980744e+00 8.06592345e-01 1.16751447e-01
-9.49048579e-01 -5.79746783e-01 6.77798688e-02 7.86714703e-02
-8.68622780e-01 -6.00731745e-02 5.54436386e-01 -9.91116464e-01
8.23544025e-01 3.21488678e-01 6.97662771e-01 1.12511086e+00
1.03786014e-01 6.00760937e-01 1.15842175e+00 -4.29205686e-01
-4.88924142e-03 8.86232615e-01 5.63248575e-01 4.66699839e-01
3.57421398e-01 -7.44803250e-01 -3.69678408e-01 -4.39385980e-01
1.65423434e-02 5.61262593e-02 -1.28557896e-02 2.16604799e-01
-7.90888250e-01 1.19484854e+00 1.24782927e-01 2.52987772e-01
-1.05076499e-01 -4.76365894e-01 5.45746624e-01 6.79126084e-01
9.04296577e-01 7.16438293e-01 -4.71917272e-01 -3.89889419e-01
-8.86036992e-01 1.80660725e-01 9.56114173e-01 8.58653128e-01
6.75715566e-01 -3.79977733e-01 2.11026385e-01 1.07394338e+00
2.31902495e-01 2.15526432e-01 5.48079431e-01 -4.33896691e-01
7.40765929e-01 1.32220376e+00 1.72295317e-01 -1.16420782e+00
-4.20848995e-01 -5.28735638e-01 -5.58894813e-01 2.46460643e-02
7.37466693e-01 -3.57156992e-01 -5.53434432e-01 1.02626431e+00
-7.07226172e-02 -7.45298803e-01 -1.94529220e-01 5.41255832e-01
5.54892361e-01 7.01115906e-01 1.92951962e-01 -8.13295841e-02
1.16257107e+00 -8.14460099e-01 -7.36617446e-01 -2.56460130e-01
1.18031847e+00 -1.17342162e+00 1.51288795e+00 6.61736608e-01
-7.68928349e-01 -3.19050699e-01 -1.01634610e+00 -2.13441387e-01
-7.77622998e-01 -4.21956647e-03 5.71369350e-01 7.83072472e-01
-4.30684984e-01 7.91605175e-01 -2.83291936e-01 -6.18704379e-01
4.78419423e-01 8.00872669e-02 -4.80121225e-01 9.47214290e-02
-9.93197143e-01 1.00697935e+00 5.98669536e-02 -1.22348443e-01
-2.71241635e-01 -7.66117454e-01 -3.24828863e-01 -7.66735226e-02
2.03199849e-01 -6.39588758e-02 1.11582160e+00 -1.14457834e+00
-9.41692829e-01 7.82613814e-01 -5.49530052e-02 -6.27148300e-02
6.00475490e-01 -3.60766590e-01 -3.04259032e-01 -3.21330726e-01
3.33593431e-04 2.14959189e-01 6.84194028e-01 -1.02853477e+00
-7.10647702e-01 -1.55596554e-01 1.54501006e-01 -2.14638248e-01
-8.29037189e-01 4.94798332e-01 2.73733377e-01 -6.45587206e-01
-8.05347785e-02 -8.84057641e-01 -5.16413674e-02 -1.33746564e-01
-4.33755785e-01 -4.44154412e-01 8.71994972e-01 -7.61714935e-01
1.43856287e+00 -2.22157454e+00 -1.91251755e-01 -7.21845729e-03
2.65347630e-01 2.18871385e-01 1.34003490e-01 9.37870681e-01
-8.61321390e-03 9.03076589e-01 -1.69677138e-02 -3.73393565e-01
1.13342799e-01 -8.21424201e-02 -3.46076846e-01 5.36299944e-01
7.09321648e-02 4.38300699e-01 -7.54048049e-01 -2.50066310e-01
-1.82403013e-01 4.69341166e-02 -7.74195135e-01 3.13231409e-01
-1.88037336e-01 4.73429449e-02 -4.17775422e-01 4.21565533e-01
6.64210737e-01 -3.79038543e-01 -1.32292986e-01 5.48831485e-02
-3.09579879e-01 7.05446661e-01 -8.66180241e-01 1.00304902e+00
-2.74174303e-01 9.45539117e-01 -2.19942212e-01 -7.90759861e-01
8.06259274e-01 1.34616092e-01 2.51201481e-01 -3.81422102e-01
3.60347867e-01 -1.70974084e-03 2.80381143e-01 -9.22117114e-01
5.49497187e-01 -1.04530558e-01 -1.18867435e-01 9.31599855e-01
-1.68908045e-01 5.23256660e-02 3.64785194e-01 4.74110633e-01
9.76048887e-01 -3.34915072e-01 2.38840565e-01 -3.11503440e-01
4.16097790e-01 2.36380950e-01 4.25498366e-01 5.46008646e-01
-8.77685100e-03 7.45440900e-01 9.66984212e-01 -5.05860746e-01
-1.24617970e+00 -5.66120327e-01 -5.30158542e-02 1.34114587e+00
-4.87018913e-01 -7.46797323e-01 -7.66974866e-01 -8.96314561e-01
4.26814258e-02 7.81797528e-01 -6.17327154e-01 3.40385847e-02
-2.39608735e-01 -1.00378203e+00 4.51450497e-01 1.35142490e-01
3.62724543e-01 -8.57341111e-01 -3.61875534e-01 1.83837205e-01
-2.55640894e-01 -7.87558496e-01 -2.03583464e-01 4.44024652e-01
-7.02879846e-01 -1.08571076e+00 -2.75169164e-01 -4.46538836e-01
8.80637050e-01 2.87651062e-01 9.22713459e-01 4.96024311e-01
-4.41509560e-02 -1.75096780e-01 -8.72226715e-01 -9.40838158e-01
-7.72707403e-01 3.73677075e-01 -2.75474548e-01 -2.98598498e-01
5.82590818e-01 -3.74255151e-01 -1.01052232e-01 1.74629670e-02
-8.89692366e-01 2.85225157e-02 3.70775253e-01 6.99120402e-01
-3.53103817e-01 1.08884811e-01 5.54656208e-01 -1.54285121e+00
1.10153008e+00 -7.90152848e-01 -8.11393484e-02 8.00722167e-02
-6.51949346e-01 -1.19017348e-01 9.12127972e-01 -5.26670337e-01
-7.57966638e-01 -3.86536121e-01 -2.12018013e-01 2.85127401e-01
-1.84400361e-02 8.24041486e-01 1.74984068e-01 2.25023761e-01
9.47918355e-01 -3.39168340e-01 5.11545427e-02 -7.97254145e-01
-9.02681053e-02 1.25040162e+00 -2.70419329e-01 -1.57743722e-01
9.32800710e-01 2.05483362e-01 -5.18987238e-01 -8.59642327e-01
-1.02543616e+00 -6.36123836e-01 -8.06586921e-01 -1.65887043e-01
6.24314964e-01 -4.91883487e-01 -4.41458285e-01 5.65158904e-01
-9.86515462e-01 -1.78565145e-01 9.82353538e-02 1.70883924e-01
4.79966365e-02 3.52971286e-01 -5.07657707e-01 -8.36261928e-01
-1.77619442e-01 -8.31554711e-01 2.27285802e-01 -9.87928286e-02
-7.07881927e-01 -1.07515466e+00 -1.19052887e-01 7.14904308e-01
5.20024598e-01 1.29646892e-02 1.43928981e+00 -1.05753708e+00
-1.79899588e-01 -3.45874995e-01 -2.10676640e-01 5.81540823e-01
2.65881181e-01 4.87995446e-01 -1.00468624e+00 -2.48770379e-02
2.04504564e-01 -4.22928840e-01 5.14557362e-01 -2.21386746e-01
1.02274144e+00 -6.78296208e-01 -2.06849817e-03 8.34544227e-02
1.25228095e+00 -1.23364367e-01 4.49260026e-01 4.96462315e-01
6.43872380e-01 7.60573387e-01 5.84593952e-01 6.04955614e-01
3.22115719e-01 1.55074418e-01 3.27579558e-01 6.41173348e-02
2.04927146e-01 -2.88692653e-01 4.77892399e-01 1.18347037e+00
1.10959105e-01 -3.95260096e-01 -9.97347116e-01 4.38593686e-01
-1.77356040e+00 -9.27954078e-01 -5.15747845e-01 1.83172798e+00
8.42714429e-01 3.30277175e-01 2.10400850e-01 4.71235216e-01
2.82093495e-01 3.59099135e-02 -5.79998493e-02 -7.80363202e-01
1.57041494e-02 -3.76266420e-01 5.55337191e-01 2.55358905e-01
-9.00348544e-01 6.19947493e-01 6.84892225e+00 5.60888112e-01
-1.21106744e+00 1.84147388e-01 6.27757251e-01 2.24737767e-02
-5.59187710e-01 1.36914626e-01 -7.47943103e-01 7.84746528e-01
1.14691293e+00 -2.52218276e-01 1.92506850e-01 9.70559716e-01
4.68162835e-01 -3.89825493e-01 -8.87692571e-01 6.18748009e-01
2.23363027e-01 -1.17251885e+00 1.39110506e-01 1.83716533e-03
6.58073723e-01 -2.46795416e-02 -1.16806127e-01 3.02784681e-01
4.37511355e-01 -1.06130278e+00 8.55409503e-01 2.69703597e-01
-2.68030893e-02 -6.56885326e-01 8.95091355e-01 8.00164521e-01
-2.08965793e-01 -3.23044419e-01 -7.81404257e-01 -6.69917464e-01
-8.03880617e-02 8.95437300e-01 -9.10144925e-01 1.37849480e-01
7.71294892e-01 7.86474884e-01 -1.06780529e+00 1.01966643e+00
-1.68520689e-01 9.60058868e-01 -1.74950629e-01 -5.08574545e-01
2.35753715e-01 -3.83935958e-01 1.67797536e-01 1.44106722e+00
1.95444763e-01 9.30190086e-03 -4.22946736e-03 5.78348875e-01
-8.66494179e-02 5.82757711e-01 -4.57677394e-01 -3.85564625e-01
3.01388919e-01 1.40306878e+00 -6.92166984e-01 -1.10793822e-01
-7.41533935e-01 4.98557955e-01 5.53925455e-01 1.66655377e-01
-6.54277205e-01 -2.10648701e-01 4.63366449e-01 5.40508091e-01
-1.46069319e-03 -1.75162748e-01 -7.66547084e-01 -9.92408156e-01
7.53322467e-02 -1.08442163e+00 1.95116147e-01 -6.35667264e-01
-1.64180446e+00 4.99992847e-01 -8.31033587e-02 -1.26606762e+00
1.20008606e-02 -5.94655037e-01 -6.92500770e-01 7.34103262e-01
-1.26544917e+00 -9.48582351e-01 -3.68814707e-01 1.86368182e-01
4.05569196e-01 -1.05863646e-01 6.36725426e-01 5.18599093e-01
-5.80137789e-01 6.02520645e-01 8.32792521e-02 2.48142436e-01
1.04350603e+00 -1.11107886e+00 1.89913586e-01 5.71674526e-01
-1.64844006e-01 9.30752695e-01 9.99538541e-01 -6.36924207e-01
-1.17070365e+00 -8.64019156e-01 1.21360958e+00 -7.07165837e-01
1.00718319e+00 -6.32665396e-01 -1.09473169e+00 6.39971852e-01
4.73473042e-01 -6.61666572e-01 1.13968253e+00 3.89639527e-01
-4.11376178e-01 6.59805164e-02 -1.02212703e+00 7.11507559e-01
6.17699206e-01 -4.28741932e-01 -8.69567096e-01 6.49864793e-01
3.98142487e-01 1.33243412e-01 -7.31210053e-01 -3.16221476e-01
4.29999828e-01 -8.72693896e-01 2.86270231e-01 -8.95923436e-01
6.52857065e-01 1.20023414e-02 1.66264117e-01 -1.51328659e+00
-3.23400676e-01 -4.52062845e-01 6.42027497e-01 1.65435433e+00
8.34037900e-01 -5.34554183e-01 6.53389931e-01 9.06957746e-01
-5.83573245e-02 -4.57200080e-01 -4.21417236e-01 -4.57840174e-01
4.25361007e-01 -3.43805760e-01 2.38195762e-01 1.17935705e+00
4.87810522e-01 3.48933429e-01 -3.12271446e-01 -3.61961454e-01
1.49640739e-01 -2.56351709e-01 9.24008667e-01 -1.68980598e+00
1.57299235e-01 -2.32421979e-01 -1.59130365e-01 -5.31160116e-01
4.58521284e-02 -8.63377690e-01 -2.14021578e-01 -1.59505308e+00
5.03647327e-01 -5.32099187e-01 1.06664710e-01 6.37221754e-01
-2.40748793e-01 1.47219181e-01 2.05982164e-01 4.05757695e-01
-2.14164004e-01 1.10360354e-01 9.82729614e-01 -7.25327507e-02
-3.17254551e-02 5.21890521e-02 -1.22367942e+00 1.04323399e+00
8.62014353e-01 -9.05308843e-01 -4.54900533e-01 -2.54213810e-01
9.95361924e-01 -6.94764972e-01 -5.66320233e-02 -8.43188286e-01
1.96679085e-01 -3.00170451e-01 2.41066813e-01 -5.40047646e-01
1.64806210e-02 -9.03979361e-01 3.89186181e-02 1.89096153e-01
-4.92365956e-01 3.69254738e-01 -1.11197397e-01 1.31967589e-01
-2.31371567e-01 -7.78076828e-01 7.43131757e-01 -3.78253996e-01
-1.27804473e-01 -2.61140227e-01 -8.40292931e-01 1.83040395e-01
7.40902364e-01 -5.88703938e-02 -8.59640181e-01 -5.00846267e-01
-3.69516939e-01 -1.94781229e-01 7.11672962e-01 4.96454477e-01
1.92648470e-02 -7.98603654e-01 -6.24877512e-01 -4.69247587e-02
1.98483597e-02 -2.49596864e-01 -5.12864664e-02 8.71181488e-01
-6.72034323e-01 2.47502625e-01 -2.29838744e-01 -9.99038070e-02
-1.30227506e+00 5.86455762e-01 -4.33588438e-02 -1.25382796e-01
-5.93322337e-01 6.01161540e-01 -1.33643389e-01 -3.76791388e-01
1.38936460e-01 -3.76991361e-01 -3.95931363e-01 6.68482840e-01
4.26446855e-01 5.78620791e-01 2.86662310e-01 -3.27802777e-01
-1.05349794e-01 2.10890114e-01 -4.02123183e-01 1.07684746e-01
1.51696372e+00 -3.22445370e-02 -2.74736732e-01 7.17567563e-01
1.30282164e+00 4.20940548e-01 -8.20093274e-01 2.22335324e-01
1.63661391e-01 -5.86584210e-01 -8.24130103e-02 -6.84944093e-01
-7.44201005e-01 8.60541880e-01 -4.08220179e-02 7.69318044e-01
6.47860289e-01 -2.08071023e-01 4.93970424e-01 5.28247476e-01
1.73454676e-02 -1.49132240e+00 8.64368007e-02 7.77526200e-01
8.06831777e-01 -1.35119593e+00 4.14303720e-01 -6.71585262e-01
-8.97739828e-01 1.13347816e+00 4.10273880e-01 -3.18334371e-01
8.55805099e-01 4.26966697e-01 4.07180011e-01 -1.49900213e-01
-1.03560424e+00 2.96286345e-01 4.50120978e-02 4.33206230e-01
8.44751000e-01 -2.96981752e-01 -7.08815932e-01 7.12355614e-01
-6.20184064e-01 -1.03774361e-01 1.18845510e+00 8.79662931e-01
-6.52944922e-01 -1.18124676e+00 -2.28162378e-01 8.70797098e-01
-8.64775836e-01 -2.22072765e-01 -8.32413912e-01 9.08205092e-01
7.73839355e-02 1.31876361e+00 -6.89477548e-02 -4.47545260e-01
2.39995509e-01 4.33862686e-01 -2.68581599e-01 -7.96996176e-01
-1.33544040e+00 -1.71623468e-01 4.49279100e-01 -1.41199799e-02
-1.91817120e-01 -8.69871676e-01 -1.04195487e+00 -7.77457416e-01
-3.58432442e-01 3.90269488e-01 7.82473505e-01 1.09495950e+00
2.09111676e-01 2.21608698e-01 4.76857513e-01 -1.89551309e-01
-6.46002173e-01 -1.53773081e+00 -5.19972980e-01 8.35283637e-01
1.04544885e-01 -4.56204474e-01 -9.03635085e-01 2.95152515e-01] | [8.881903648376465, 10.287038803100586] |
cfc4b977-b000-47ce-90e4-e507245b69b0 | unsupervised-semantic-aggregation-and | 2010.05517 | null | https://arxiv.org/abs/2010.05517v1 | https://arxiv.org/pdf/2010.05517v1.pdf | Unsupervised Semantic Aggregation and Deformable Template Matching for Semi-Supervised Learning | Unlabeled data learning has attracted considerable attention recently. However, it is still elusive to extract the expected high-level semantic feature with mere unsupervised learning. In the meantime, semi-supervised learning (SSL) demonstrates a promising future in leveraging few samples. In this paper, we combine both to propose an Unsupervised Semantic Aggregation and Deformable Template Matching (USADTM) framework for SSL, which strives to improve the classification performance with few labeled data and then reduce the cost in data annotating. Specifically, unsupervised semantic aggregation based on Triplet Mutual Information (T-MI) loss is explored to generate semantic labels for unlabeled data. Then the semantic labels are aligned to the actual class by the supervision of labeled data. Furthermore, a feature pool that stores the labeled samples is dynamically updated to assign proxy labels for unlabeled data, which are used as targets for cross-entropy minimization. Extensive experiments and analysis across four standard semi-supervised learning benchmarks validate that USADTM achieves top performance (e.g., 90.46$\%$ accuracy on CIFAR-10 with 40 labels and 95.20$\%$ accuracy with 250 labels). The code is released at https://github.com/taohan10200/USADTM. | ['Qi Wang', 'Yuan Yuan', 'Junyu Gao', 'Tao Han'] | 2020-10-12 | null | http://proceedings.neurips.cc/paper/2020/hash/71a58e8cb75904f24cde464161c3e766-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/71a58e8cb75904f24cde464161c3e766-Paper.pdf | neurips-2020-12 | ['template-matching'] | ['computer-vision'] | [ 2.13310376e-01 1.62954301e-01 -4.24279898e-01 -9.37885523e-01
-1.06665683e+00 -3.80972892e-01 3.14887911e-01 -3.62577289e-02
-2.94251263e-01 6.74711227e-01 -2.01167837e-02 9.49658453e-02
-3.72625962e-02 -4.73870307e-01 -3.78042966e-01 -9.37954068e-01
2.78265536e-01 3.67080271e-01 -1.43045932e-01 3.15465957e-01
4.59129997e-02 -3.59919071e-02 -1.42677999e+00 2.35890374e-01
1.10452533e+00 1.49756181e+00 1.00864850e-01 -2.80585945e-01
-3.18988383e-01 6.36319995e-01 -2.47075573e-01 -3.98048818e-01
4.07542825e-01 -5.00759304e-01 -8.16744089e-01 3.79846692e-01
3.66605312e-01 7.52656534e-02 4.25773161e-03 1.18310404e+00
4.60108161e-01 1.84199065e-01 7.52890289e-01 -1.30652404e+00
-6.62135899e-01 6.84603989e-01 -5.29822707e-01 -6.20919168e-02
-1.23567499e-01 3.21886353e-02 1.18696320e+00 -1.39822936e+00
3.46004367e-01 1.11719584e+00 5.22029340e-01 6.68926358e-01
-1.10037076e+00 -9.75776553e-01 1.50648400e-01 9.88348275e-02
-1.60421216e+00 -4.65816706e-01 9.76367533e-01 -3.09457332e-01
3.51200014e-01 1.33564368e-01 1.96451634e-01 8.15091312e-01
-4.04934824e-01 1.15317404e+00 1.25485408e+00 -4.16863531e-01
3.26325744e-01 3.57736081e-01 5.30032396e-01 8.43008161e-01
8.42414573e-02 1.23825520e-02 -5.93885958e-01 5.47986031e-02
2.47588545e-01 8.57729390e-02 -5.33054732e-02 -4.55305189e-01
-1.04466760e+00 8.81008267e-01 5.21295607e-01 5.18117622e-02
-4.12337780e-02 -2.91335583e-01 3.84776294e-01 2.33985573e-01
7.49118090e-01 1.52274206e-01 -6.06343448e-01 2.11579621e-01
-8.93294513e-01 -1.83050171e-01 3.57820183e-01 1.08611500e+00
1.12830031e+00 4.92196046e-02 -1.20236680e-01 1.12896192e+00
5.36706984e-01 5.64549804e-01 6.25397623e-01 -9.11863744e-01
6.86334252e-01 9.87809300e-01 -1.35297954e-01 -8.17763269e-01
-2.25227520e-01 -5.67044735e-01 -8.32479298e-01 -1.96740583e-01
2.20508114e-01 -1.24586537e-01 -8.59330595e-01 1.65454721e+00
4.31061566e-01 3.37155163e-01 2.09390506e-01 9.15860236e-01
9.21186626e-01 5.14821887e-01 2.57808447e-01 -2.97685623e-01
1.20182729e+00 -1.11564267e+00 -6.45587683e-01 -3.15181255e-01
9.18200016e-01 -5.96242487e-01 1.28162968e+00 2.40931526e-01
-6.21980846e-01 -6.28583372e-01 -9.37665224e-01 1.27926350e-01
-3.24532986e-01 4.60069120e-01 4.79286581e-01 8.07685971e-01
-6.02383912e-01 3.80137563e-01 -7.76270449e-01 -2.46839747e-02
7.85100341e-01 4.10231262e-01 -1.30688488e-01 -9.74071920e-02
-1.28270507e+00 3.49922091e-01 7.28172541e-01 7.80800208e-02
-5.54915845e-01 -5.76468110e-01 -9.34566438e-01 -9.11738724e-02
3.96402776e-01 -1.85218439e-01 1.05555296e+00 -1.08483517e+00
-1.38110518e+00 1.10170746e+00 -5.76476287e-03 -4.63507503e-01
3.43344271e-01 -2.10458394e-02 -4.28561062e-01 1.31090321e-02
2.41389528e-01 9.09535110e-01 6.93831980e-01 -1.27123690e+00
-6.67977571e-01 -6.10453129e-01 -5.11786103e-01 4.04096186e-01
-8.52384806e-01 -1.02807060e-01 -3.06969315e-01 -8.40025544e-01
4.97198522e-01 -1.03813004e+00 -1.26198485e-01 -1.64900512e-01
-5.14262915e-01 -4.79748905e-01 8.83226931e-01 -3.52778733e-01
1.24043798e+00 -2.23739839e+00 -2.53986359e-01 3.45613241e-01
1.10045798e-01 4.27260518e-01 -5.14381826e-02 -6.95926398e-02
-1.33425385e-01 1.38379587e-02 -6.06485009e-01 -5.55431783e-01
9.83937457e-02 -4.85299602e-02 -1.79770440e-01 3.40299398e-01
2.13479385e-01 1.08400989e+00 -7.46694207e-01 -7.89147079e-01
1.43355086e-01 2.34391555e-01 -3.33133698e-01 1.51495114e-01
-1.02179520e-01 5.16290724e-01 -7.77951658e-01 9.02198613e-01
7.49396861e-01 -4.87180918e-01 5.95312938e-02 -1.87476233e-01
2.23215759e-01 5.62501401e-02 -1.25243783e+00 1.77765512e+00
-2.65969694e-01 1.39114827e-01 -3.58986706e-01 -1.30660880e+00
1.24528122e+00 2.41052687e-01 6.01152837e-01 -7.10052669e-01
2.40940318e-01 4.41409975e-01 -4.34906602e-01 -3.23616534e-01
-2.56273337e-02 -7.16471747e-02 -3.96062613e-01 3.66537839e-01
1.13999590e-01 1.06609076e-01 4.48238254e-02 8.28354955e-02
6.11735225e-01 1.06736995e-01 1.88339710e-01 -4.24149156e-01
6.22877300e-01 -5.40972203e-02 9.61924672e-01 2.04742551e-01
-3.86570334e-01 5.37691236e-01 -1.02483854e-01 -3.23867321e-01
-5.74660897e-01 -1.20614946e+00 -3.34937125e-01 8.42565536e-01
3.52079839e-01 -1.59998685e-01 -8.88507605e-01 -1.02876687e+00
-9.40224752e-02 7.22398937e-01 -3.84187460e-01 -4.54246104e-01
-2.60752529e-01 -9.24520254e-01 3.04842979e-01 4.96179461e-01
7.84680843e-01 -1.09082484e+00 -1.64784059e-01 8.02790150e-02
-3.58799696e-01 -1.04908669e+00 -6.22503102e-01 1.42245188e-01
-1.02298331e+00 -8.51901233e-01 -4.93678898e-01 -1.16366005e+00
1.02491736e+00 2.93343872e-01 7.84333825e-01 -2.23747760e-01
-2.12569743e-01 6.16479516e-02 -4.97041136e-01 -3.01532388e-01
-5.45700565e-02 8.69215280e-02 1.28888100e-01 5.32181203e-01
7.49252319e-01 -3.49754930e-01 -6.36293292e-01 5.60513020e-01
-6.92635238e-01 6.56622350e-02 4.67131376e-01 9.85387623e-01
1.01929438e+00 1.86564843e-03 8.83547366e-01 -1.12307656e+00
1.32418811e-01 -5.77187359e-01 -5.36058486e-01 3.66509289e-01
-8.74041975e-01 5.92374913e-02 8.20353270e-01 -3.41872513e-01
-1.22483087e+00 3.72809201e-01 1.56717226e-01 -4.74383056e-01
-9.78081748e-02 4.20959055e-01 -4.92130548e-01 2.14258179e-01
4.95424777e-01 3.22912395e-01 -1.03479691e-01 -7.05741107e-01
2.20366672e-01 1.10508358e+00 2.50848085e-01 -5.68858981e-01
8.10350120e-01 5.24500430e-01 -3.20870310e-01 -3.70602816e-01
-1.50955045e+00 -5.16260922e-01 -6.48916900e-01 -1.01154439e-01
7.01274872e-01 -1.04712546e+00 -2.90165484e-01 5.41325271e-01
-4.79210645e-01 -2.61576921e-01 -4.24442142e-01 6.90425396e-01
-4.33018446e-01 3.36239547e-01 -2.73252010e-01 -6.81900442e-01
-5.40554225e-01 -1.18077016e+00 1.02480376e+00 4.17043507e-01
-1.88097749e-02 -9.22057390e-01 -2.82505929e-01 9.24447000e-01
1.29074574e-01 -1.00791166e-02 6.29239142e-01 -9.83315706e-01
-4.45230544e-01 -2.36644968e-01 -3.47015679e-01 7.17737079e-01
3.29512954e-01 -4.09833640e-01 -1.11887443e+00 -3.00592303e-01
1.14347912e-01 -7.31073022e-01 8.73769343e-01 2.21837699e-01
1.51476276e+00 -2.75515527e-01 -2.83764064e-01 6.13206267e-01
1.24724388e+00 2.87847489e-01 2.53625870e-01 1.85546651e-02
8.53718758e-01 7.09983349e-01 1.12265050e+00 4.75133777e-01
3.88022423e-01 5.35166740e-01 1.30854562e-01 4.98095080e-02
-1.19020931e-01 -3.98041666e-01 2.30406255e-01 1.19621313e+00
3.69436234e-01 3.22598666e-02 -9.79474545e-01 3.00385177e-01
-1.69256639e+00 -6.00990117e-01 -2.91816909e-02 2.30229425e+00
1.03761482e+00 1.43305033e-01 -6.97958767e-02 2.24062040e-01
1.00372207e+00 -3.91793773e-02 -8.50519598e-01 1.28786922e-01
-1.63748279e-01 1.80308312e-01 3.84713590e-01 2.40061879e-01
-1.47353482e+00 9.94583905e-01 4.36783266e+00 1.33043659e+00
-9.16421533e-01 2.70384401e-01 1.12233329e+00 8.60633031e-02
-1.66945323e-01 4.16083857e-02 -9.88372803e-01 7.88914621e-01
7.55537152e-01 -1.43179208e-01 1.02800600e-01 9.02517378e-01
2.88272381e-01 1.57240972e-01 -8.29556763e-01 1.10198784e+00
8.25867280e-02 -9.59202111e-01 -8.45945906e-03 -1.10623255e-01
9.76196408e-01 -6.66764304e-02 2.42229223e-01 3.02173048e-01
2.31465489e-01 -7.54495740e-01 5.48584580e-01 3.09099287e-01
9.75562334e-01 -7.57618070e-01 6.16885960e-01 4.78086412e-01
-1.41590464e+00 -2.12632582e-01 -3.24650288e-01 2.98510283e-01
-1.22476250e-01 8.17979932e-01 -5.49685240e-01 4.98039931e-01
5.96393406e-01 1.00326014e+00 -6.41811907e-01 8.47796798e-01
-1.36688977e-01 8.98048103e-01 -2.84090191e-01 5.26320562e-02
2.35764802e-01 -4.61548120e-01 9.06304643e-02 7.82294989e-01
1.46327749e-01 9.90599915e-02 5.75221598e-01 5.23473918e-01
-3.44017774e-01 5.50596416e-01 -2.36418784e-01 1.03709854e-01
7.36653209e-01 1.17379737e+00 -8.43780637e-01 -3.98297369e-01
-3.52942228e-01 7.69970059e-01 4.04616982e-01 2.55121350e-01
-9.15246546e-01 -2.58128107e-01 2.49250323e-01 -2.64920503e-01
-2.59417035e-02 2.06133693e-01 -4.97627050e-01 -1.37658238e+00
1.33760229e-01 -5.95083117e-01 7.09836364e-01 -4.72051203e-01
-1.50269127e+00 4.34469640e-01 -2.57193238e-01 -1.67633998e+00
1.70598418e-01 -2.89666474e-01 -3.08555067e-01 6.04898632e-01
-1.43309295e+00 -1.08032393e+00 -3.88610750e-01 4.57231849e-01
7.65267134e-01 -3.20155501e-01 8.00622165e-01 4.24814373e-01
-8.93793643e-01 9.99193728e-01 3.27992022e-01 3.32523465e-01
7.42991924e-01 -1.15140831e+00 4.38567288e-02 5.77691495e-01
2.81396031e-01 4.23594832e-01 1.59181640e-01 -4.83337224e-01
-1.08018339e+00 -1.45252526e+00 8.79881263e-01 -2.28882104e-01
4.42967355e-01 -3.77187192e-01 -9.39473927e-01 5.41340947e-01
-2.60415345e-01 3.70705664e-01 9.72784042e-01 -3.69770020e-01
-5.56111097e-01 -4.50517386e-01 -1.29387140e+00 4.02197689e-01
1.10622811e+00 -4.64933485e-01 -4.85856980e-01 5.42560339e-01
7.35194087e-01 -7.78334066e-02 -9.09224331e-01 5.81015587e-01
2.19246209e-01 -6.99077368e-01 8.37265074e-01 -3.60437721e-01
1.97614223e-01 -3.33479643e-01 -2.75247514e-01 -1.08288479e+00
4.68843766e-02 -2.08378479e-01 1.01133563e-01 1.48846030e+00
5.47118545e-01 -8.34144175e-01 1.07034373e+00 7.72344768e-01
-1.87866807e-01 -9.50663567e-01 -8.64633858e-01 -9.59706306e-01
-1.02531919e-02 -3.92801821e-01 4.56690907e-01 1.28374696e+00
-8.53805244e-03 3.51091892e-01 -1.74540803e-01 -5.27725741e-02
1.03222501e+00 4.34986442e-01 3.49856406e-01 -1.28958714e+00
1.63970403e-02 -1.91937566e-01 -2.94561625e-01 -9.90923226e-01
5.34254313e-01 -1.40108299e+00 -3.96044701e-02 -1.11651683e+00
4.03163731e-01 -9.16125834e-01 -7.15807319e-01 7.24975765e-01
-3.62004071e-01 4.44228709e-01 -2.29648352e-02 4.83872443e-01
-8.45717490e-01 8.52580667e-01 1.09650648e+00 -1.90340623e-01
-7.44756013e-02 1.52108908e-01 -7.23480344e-01 6.82709754e-01
1.28004181e+00 -5.98182619e-01 -6.51112378e-01 -2.57640719e-01
-2.99140841e-01 -2.25747347e-01 1.00162715e-01 -1.00389981e+00
3.34969312e-02 -1.74696475e-01 1.73944190e-01 -4.62752193e-01
1.94840789e-01 -8.05301428e-01 -5.87903112e-02 3.48059803e-01
-7.00622678e-01 -5.01137733e-01 -3.49390924e-01 5.68894148e-01
-4.54790801e-01 -2.90763080e-01 9.86659706e-01 1.18946016e-01
-8.83000553e-01 6.38780117e-01 2.91219950e-01 4.24190342e-01
1.18750215e+00 -2.81292409e-01 1.07328063e-02 6.51143640e-02
-8.20567250e-01 5.43442667e-01 2.46364608e-01 3.89242291e-01
5.10094702e-01 -1.56465924e+00 -6.35584831e-01 1.08186200e-01
5.19850791e-01 1.55500352e-01 3.19185287e-01 6.07099831e-01
-7.52634555e-02 4.05366451e-01 1.42517090e-01 -5.51156282e-01
-1.12001956e+00 4.58868265e-01 1.36959761e-01 -1.62906915e-01
-2.98716486e-01 9.83063579e-01 1.28958896e-01 -7.00234413e-01
2.73623496e-01 1.53748125e-01 -2.00743973e-01 2.94019997e-01
2.77398318e-01 2.57886678e-01 1.11351967e-01 -7.36931205e-01
-3.78344744e-01 5.71220458e-01 -1.28071919e-01 2.69689023e-01
1.28641558e+00 -3.19619596e-01 1.20402865e-01 4.81839359e-01
1.67738533e+00 -3.12310547e-01 -1.44875348e+00 -7.85837591e-01
3.73353928e-01 -4.37032163e-01 -7.40236267e-02 -6.87965989e-01
-1.39876533e+00 9.11937535e-01 7.38511682e-01 -1.11682415e-01
1.26540768e+00 5.69840148e-02 9.56536949e-01 3.41251701e-01
5.27436614e-01 -1.35934353e+00 2.19895050e-01 2.00795949e-01
3.18741947e-01 -1.62755156e+00 -2.06267998e-01 -6.13365352e-01
-8.43296230e-01 6.58533275e-01 7.69546330e-01 4.07351851e-02
8.31803024e-01 -1.49698341e-02 3.00415963e-01 -2.21323241e-02
-4.85269308e-01 -1.79856122e-01 2.73589551e-01 2.87601590e-01
2.85686791e-01 3.07445496e-01 -3.15477163e-01 8.34147930e-01
-1.68711059e-02 -8.50841850e-02 -7.46936724e-02 8.82019937e-01
-5.01442790e-01 -1.14693832e+00 -1.23970397e-01 7.35778093e-01
-3.52071375e-01 -1.98470861e-01 -1.97175607e-01 2.20159352e-01
2.17717379e-01 9.72767591e-01 -1.19139940e-01 -4.14469779e-01
5.52741811e-02 3.98853153e-01 -2.82634445e-03 -8.33468020e-01
-2.16970831e-01 1.57054737e-02 -1.58379152e-01 -3.79840523e-01
-6.10823572e-01 -6.59248054e-01 -1.50712788e+00 4.05615658e-01
-4.69312757e-01 3.93511385e-01 5.36809802e-01 9.69046235e-01
3.95918101e-01 -1.87097173e-02 1.21251595e+00 -3.31921011e-01
-7.68554747e-01 -7.76158154e-01 -6.16743505e-01 6.87756419e-01
-1.97634608e-01 -7.62213767e-01 -4.61814195e-01 3.22502524e-01] | [9.462248802185059, 3.7115354537963867] |
45df5d3e-38b6-4815-b186-ffb1f3dea1bb | controllable-data-augmentation-for-context | 2304.13902 | null | https://arxiv.org/abs/2304.13902v2 | https://arxiv.org/pdf/2304.13902v2.pdf | Controllable Data Augmentation for Context-Dependent Text-to-SQL | The limited scale of annotated data constraints existing context-dependent text-to-SQL models because of the complexity of labeling. The data augmentation method is a commonly used method to solve this problem. However, the data generated by current augmentation methods often lack diversity. In this paper, we introduce ConDA, which generates interactive questions and corresponding SQL results. We designed the SQL dialogue state to enhance the data diversity through the state transition. Meanwhile, we also present a filter method to ensure the data quality by a grounding model. Additionally, we utilize a grounding model to identify and filter low-quality questions that mismatch the state information. Experimental results on the SParC and CoSQL datasets show that ConDA boosts the baseline model to achieve an average improvement of $3.3\%$ on complex questions. Moreover, we analyze the augmented data, which reveals that the data generated by ConDA are of high quality in both SQL template hardness and types, turns, and question consistency. | ['Wanxiang Che', 'Longxu Dou', 'Dingzirui Wang'] | 2023-04-27 | null | null | null | null | ['text-to-sql'] | ['computer-code'] | [-3.18654180e-02 2.73328155e-01 -1.94766209e-01 -7.43423283e-01
-1.13022995e+00 -7.65369654e-01 3.83436471e-01 3.85486394e-01
-1.93807185e-01 5.27194083e-01 4.73393649e-01 -4.24634695e-01
-3.03768795e-02 -1.07093418e+00 -8.33245516e-01 1.94486365e-01
6.49690151e-01 4.77098018e-01 5.60651898e-01 -5.61113000e-01
2.67779678e-01 -1.50210202e-01 -1.61267698e+00 8.51405859e-01
1.68581474e+00 9.78044391e-01 7.36609623e-02 2.41416588e-01
-9.61691141e-01 7.99621582e-01 -6.80801630e-01 -7.65012980e-01
8.66794661e-02 -4.80884582e-01 -9.88653958e-01 1.20269299e-01
5.97944796e-01 -3.72173131e-01 -2.41415396e-01 7.36397445e-01
3.39999586e-01 6.90802094e-03 -8.22226629e-02 -1.27677476e+00
-6.76839232e-01 7.87235260e-01 1.37848422e-01 -2.35321268e-01
8.68346274e-01 1.00111753e-01 1.13838518e+00 -1.21867144e+00
6.21902466e-01 1.04296541e+00 4.90723491e-01 5.36588609e-01
-1.04542339e+00 -4.66651201e-01 2.92308003e-01 1.92569941e-01
-1.05875766e+00 -4.22713310e-01 4.99745041e-01 -2.19992131e-01
9.24837053e-01 7.65146375e-01 4.06133920e-01 7.81362414e-01
-4.99531358e-01 8.04626524e-01 1.19359136e+00 -4.19794887e-01
6.57158196e-02 4.00255263e-01 5.29584825e-01 8.41055989e-01
1.28462717e-01 -3.56208444e-01 -7.22489834e-01 -9.55770388e-02
2.54681438e-01 -9.36930105e-02 -1.58544183e-01 -7.93203786e-02
-8.93231094e-01 6.48018897e-01 8.13126713e-02 -6.82206526e-02
-1.34646416e-01 -5.57752132e-01 3.25192481e-01 3.66830021e-01
1.00181237e-01 8.87910187e-01 -6.03153527e-01 -2.32066914e-01
-4.95241970e-01 5.54503798e-01 9.89196539e-01 1.69490290e+00
9.21354890e-01 -3.11066657e-01 -7.83897460e-01 7.95915008e-01
1.43271580e-01 4.37323332e-01 2.80347705e-01 -1.03710365e+00
9.80099320e-01 1.57917976e+00 1.69343501e-01 -7.96963632e-01
-2.67576873e-01 -2.45079577e-01 -5.80264926e-01 -4.69202250e-01
3.20711523e-01 7.26027191e-02 -7.95691788e-01 1.56230676e+00
4.32642102e-01 -5.58239400e-01 5.56999668e-02 6.31935358e-01
1.24124694e+00 6.16832733e-01 -9.99408513e-02 -2.23013714e-01
1.43319631e+00 -1.03992414e+00 -1.28741121e+00 -1.64732620e-01
8.38552892e-01 -8.20805967e-01 2.04491568e+00 2.15573877e-01
-1.14748538e+00 -7.90095329e-01 -7.69722104e-01 -4.06628639e-01
-3.82032305e-01 1.28664866e-01 4.87086773e-01 8.88088763e-01
-4.67154890e-01 1.65690318e-01 -5.83071709e-01 -3.59859914e-01
1.77371483e-02 -1.89077929e-01 -6.24674968e-02 -1.52763382e-01
-1.24465263e+00 5.99378586e-01 4.20064867e-01 -1.88163429e-01
-3.11604410e-01 -1.02528369e+00 -8.53842556e-01 1.08330138e-01
9.55875397e-01 -3.57686937e-01 1.24695265e+00 1.05700277e-01
-1.33710456e+00 4.75705236e-01 -3.51698071e-01 -4.85996082e-02
3.59446347e-01 -4.66579616e-01 -5.33855498e-01 -2.24194705e-01
1.74099654e-01 4.14081097e-01 -4.83138263e-02 -1.13605332e+00
-6.57373488e-01 -2.75576413e-01 2.37639159e-01 1.01018175e-01
-6.14480495e-01 -1.20015942e-01 -1.06195664e+00 -5.62003493e-01
2.21109003e-01 -6.70071065e-01 -9.28752720e-02 -2.39924029e-01
-7.49494135e-01 -4.62506026e-01 4.28745359e-01 -8.27176630e-01
1.93601549e+00 -1.89617670e+00 -2.03638420e-01 2.19761088e-01
6.89900964e-02 1.46689609e-01 -3.22368056e-01 8.09755325e-01
2.75954783e-01 5.35351396e-01 -2.47070000e-01 -6.10705130e-02
2.44976267e-01 3.92626941e-01 -5.58214664e-01 -4.94828731e-01
4.33784992e-01 9.73460734e-01 -5.00982642e-01 -7.88197577e-01
-2.65772253e-01 -3.07345301e-01 -1.01633692e+00 8.32246780e-01
-9.05187607e-01 3.53363715e-02 -4.31095153e-01 6.69272423e-01
7.20022023e-01 -3.34232956e-01 1.61579728e-01 -4.32246149e-01
-1.29522726e-01 8.18870842e-01 -1.37414670e+00 1.88586307e+00
-3.03370059e-01 -1.56307146e-01 -2.69244283e-01 -2.13910714e-01
1.30937922e+00 5.90503439e-02 1.84417978e-01 -1.04297292e+00
-3.42629999e-01 2.24058911e-01 -2.38111869e-01 -1.05232370e+00
9.62178469e-01 3.71069103e-01 -2.90221661e-01 3.75180930e-01
-2.11962253e-01 -3.11077625e-01 6.79533660e-01 5.18277705e-01
1.11160481e+00 1.31053001e-01 -3.03201288e-01 6.42089173e-02
5.55879354e-01 4.15738285e-01 8.45729530e-01 8.25911343e-01
2.83447951e-01 5.65851748e-01 5.99344850e-01 -6.62014959e-03
-9.38320220e-01 -7.91479230e-01 -6.96174102e-03 1.04954135e+00
6.80761114e-02 -9.26929712e-01 -8.01873624e-01 -9.51419234e-01
2.61061341e-02 1.02293885e+00 -3.60811353e-01 -2.67297864e-01
-6.76510155e-01 -4.12498355e-01 5.28581500e-01 6.15024209e-01
6.95929527e-01 -9.08340812e-01 -1.53451636e-01 2.29993969e-01
-8.18523169e-01 -1.24847734e+00 -7.14084089e-01 -1.18789949e-01
-7.05642343e-01 -1.19545579e+00 8.61964002e-02 -6.31101251e-01
5.30553758e-01 1.38136894e-01 1.34662902e+00 4.12241131e-01
1.80031896e-01 8.91022533e-02 -8.12340975e-01 -3.30197126e-01
-5.32387555e-01 4.74334866e-01 -4.39624429e-01 -2.67462373e-01
5.43202758e-01 -6.93633780e-02 -4.25397784e-01 5.89602470e-01
-1.33451390e+00 2.00384215e-01 4.60011393e-01 7.21281409e-01
7.67064095e-01 -1.77345321e-01 7.77139783e-01 -1.30658758e+00
9.04357612e-01 -2.89233297e-01 -6.11208081e-01 8.11749637e-01
-1.25306058e+00 2.69268095e-01 6.41214669e-01 -1.62187200e-02
-1.30896330e+00 -9.92063507e-02 -3.38339120e-01 2.41150096e-01
-8.25450644e-02 8.79402637e-01 -7.10969388e-01 4.68958646e-01
7.31683791e-01 9.98373404e-02 -1.92163549e-02 -8.43141615e-01
5.03029048e-01 7.49854684e-01 5.81403673e-01 -7.84329653e-01
8.15002978e-01 1.57128237e-02 -4.55060810e-01 -1.74118966e-01
-1.00676417e+00 -3.23640436e-01 -2.09557801e-01 -8.75364840e-02
6.34428620e-01 -6.91976249e-01 -6.41748428e-01 1.92429513e-01
-1.15785789e+00 -3.55313234e-02 -5.38112700e-01 3.33774686e-02
-1.89657122e-01 3.70257825e-01 -5.60326517e-01 -6.59853041e-01
-3.61636281e-01 -1.03352952e+00 7.35761464e-01 2.88879007e-01
-1.88831523e-01 -3.26873064e-01 -7.01603219e-02 9.10159528e-01
3.92616242e-01 -7.07442388e-02 1.45027292e+00 -8.93648326e-01
-8.04289758e-01 -1.70405492e-01 -2.17033476e-01 1.66913331e-01
2.14400128e-01 4.97348905e-02 -5.54636002e-01 5.18683679e-02
-2.94625144e-02 -6.03314757e-01 3.83936733e-01 -5.72371066e-01
1.49301791e+00 -6.30553365e-01 2.65821397e-01 4.25125808e-01
1.21893835e+00 1.97055012e-01 7.77381599e-01 2.84609735e-01
6.57802939e-01 9.61967111e-01 9.74097192e-01 4.76493865e-01
8.35494936e-01 7.25740373e-01 5.05522847e-01 3.39278020e-02
-1.70674309e-01 -7.78495729e-01 5.14718406e-02 1.26657856e+00
5.97592652e-01 -2.26337567e-01 -9.06457782e-01 4.72830385e-01
-1.79146409e+00 -6.07793272e-01 -5.70414841e-01 2.06611156e+00
1.44266343e+00 1.50083631e-01 6.09315038e-02 2.01432824e-01
4.38484162e-01 -1.32395118e-01 -5.33030152e-01 -1.19753987e-01
-3.29773396e-01 1.81268543e-01 7.71719888e-02 4.94232208e-01
-4.89844233e-01 8.97729933e-01 6.04555559e+00 6.26190007e-01
-4.98239636e-01 -1.40954122e-01 2.39585161e-01 1.72512676e-03
-1.12961459e+00 9.57636982e-02 -1.12996542e+00 5.73253989e-01
6.80072188e-01 -3.00786257e-01 2.10007861e-01 7.56662488e-01
6.03895187e-02 5.51097542e-02 -1.03041649e+00 5.33390343e-01
-4.55459245e-02 -1.45978498e+00 3.17865372e-01 -1.93749949e-01
6.03998840e-01 -6.45680010e-01 -1.27485260e-01 8.20119262e-01
3.18158031e-01 -7.24891543e-01 6.23269498e-01 6.08373582e-01
7.60195374e-01 -6.66858375e-01 6.43289447e-01 4.70595241e-01
-8.96652222e-01 4.45605442e-02 -2.76923269e-01 2.11572737e-01
2.02679500e-01 5.53693533e-01 -7.88264930e-01 7.39044964e-01
7.52162814e-01 1.28754348e-01 -1.13206899e+00 8.14036131e-01
-2.31530383e-01 5.87177694e-01 -1.53573841e-01 -2.47616559e-01
-1.36932522e-01 -2.24151149e-01 3.13214734e-02 1.05549741e+00
2.68474400e-01 1.78157628e-01 3.75449866e-01 1.03843248e+00
-3.32895666e-01 3.69691819e-01 -1.95514351e-01 -1.18856229e-01
9.41615880e-01 1.13629186e+00 2.87483055e-02 -4.28449124e-01
-5.58298826e-01 5.78194082e-01 3.60172361e-01 2.11784557e-01
-7.46175110e-01 -7.25587189e-01 3.07885110e-01 3.38347703e-01
1.27579607e-02 -1.73708163e-02 -6.22690916e-01 -1.26333833e+00
8.42414737e-01 -1.47949922e+00 5.74196517e-01 -7.28313506e-01
-1.14015710e+00 6.92209244e-01 -1.89252570e-01 -1.08125031e+00
-1.61353469e-01 -1.13361113e-01 -2.45995298e-01 9.02593195e-01
-1.42509174e+00 -8.52219939e-01 -7.91754842e-01 5.11829615e-01
4.41641659e-01 1.59723789e-03 7.79241800e-01 5.81949532e-01
-7.84778297e-01 9.12701428e-01 -3.12021464e-01 1.71110988e-01
8.45407128e-01 -1.41478837e+00 5.47000647e-01 1.01469064e+00
-7.74564454e-04 9.19817388e-01 5.34283698e-01 -9.21373963e-01
-1.68771946e+00 -1.27329981e+00 1.19564784e+00 -6.01085484e-01
3.79088730e-01 -4.74312395e-01 -1.62531793e+00 4.46420789e-01
3.03049803e-01 -2.85363704e-01 8.19599032e-01 2.08939135e-01
-6.28557384e-01 -3.48969281e-01 -1.17320418e+00 5.92292607e-01
1.01758134e+00 -7.00389028e-01 -6.71594799e-01 2.21034065e-01
1.34890759e+00 -7.01275229e-01 -1.08330846e+00 6.76594019e-01
2.68824160e-01 -9.44607079e-01 4.78184998e-01 -9.93402064e-01
4.90450650e-01 -4.23248053e-01 -3.53254497e-01 -8.38481665e-01
-7.69847780e-02 -5.60873330e-01 -2.25303069e-01 1.95061600e+00
7.71721065e-01 -2.86867768e-01 1.16045547e+00 1.14779568e+00
-3.93329412e-01 -8.92067075e-01 -5.57377815e-01 -8.64459336e-01
-1.37869507e-01 -3.69127959e-01 1.23542285e+00 8.88416231e-01
6.97350800e-02 3.94326985e-01 -1.16467252e-01 3.47616225e-02
2.34249800e-01 3.31764728e-01 1.04559326e+00 -8.35358799e-01
-1.44238800e-01 -4.62468266e-02 5.14311969e-01 -1.37169278e+00
-3.17702889e-01 -8.61662745e-01 8.98649767e-02 -1.74382305e+00
1.13606431e-01 -7.17031181e-01 1.15405940e-01 4.27229583e-01
-7.69700527e-01 -6.09072030e-01 1.68580785e-01 9.77169499e-02
-6.67847276e-01 6.56297684e-01 1.21809053e+00 1.15820631e-01
-3.50207239e-01 -6.71157762e-02 -1.05065739e+00 2.22148836e-01
7.87979841e-01 -4.12656546e-01 -5.15441716e-01 -7.97369242e-01
6.58408225e-01 1.57988951e-01 -7.58231878e-02 -8.35827470e-01
3.45099509e-01 -4.28130656e-01 -1.33682385e-01 -1.02220786e+00
1.01621591e-01 -7.24429429e-01 -1.37977660e-01 2.60786176e-01
-8.51018369e-01 5.78691363e-01 3.84880230e-03 3.39353025e-01
-4.14885312e-01 -2.35962868e-01 4.45953667e-01 2.72867903e-02
-3.82118106e-01 3.29783410e-01 -2.16196058e-03 7.55779922e-01
4.83043879e-01 3.17253023e-01 -8.28914464e-01 -4.30426031e-01
-1.41717076e-01 7.35670388e-01 3.02317441e-01 6.06363893e-01
6.19117677e-01 -1.55782235e+00 -6.73012733e-01 2.82152623e-01
6.93208039e-01 3.08025330e-01 6.07221946e-02 3.74929726e-01
-3.63886923e-01 4.74777520e-01 4.79598939e-02 -3.32872450e-01
-1.06038582e+00 5.31698465e-01 -4.16091178e-03 -4.25718904e-01
-9.31226909e-02 6.76908255e-01 -4.28992271e-01 -1.02467418e+00
5.14374137e-01 -5.46695411e-01 -2.26002663e-01 5.63662648e-02
4.28509742e-01 3.98913294e-01 4.98214900e-01 1.39195576e-01
-1.78185508e-01 7.20586330e-02 -2.74102688e-01 -1.17938958e-01
1.08028030e+00 -1.93087310e-01 -2.17857987e-01 1.08317018e-01
8.90989602e-01 3.80789906e-01 -8.34124565e-01 -6.94106221e-01
4.47785765e-01 -5.38506985e-01 -5.31088412e-01 -1.21628273e+00
-8.10878754e-01 7.01487005e-01 3.03110868e-01 5.15137255e-01
1.12202001e+00 -9.18565467e-02 1.00230181e+00 7.22634017e-01
-3.64223234e-02 -1.15014684e+00 2.74751425e-01 6.64027870e-01
1.20210040e+00 -1.27854836e+00 -2.08956778e-01 -8.35090697e-01
-7.17526078e-01 7.84662664e-01 1.43572748e+00 6.40592635e-01
1.60454661e-01 2.90901423e-01 3.08057159e-01 -1.27202973e-01
-1.09223628e+00 -3.06887090e-01 2.80876786e-01 3.47605348e-01
5.08918524e-01 -3.26305836e-01 -5.04898012e-01 1.07367265e+00
-4.39538658e-01 -9.50825959e-03 5.51630139e-01 1.16464806e+00
-5.30879915e-01 -1.50891054e+00 -1.60870656e-01 6.55934572e-01
-3.03918570e-01 -2.55174726e-01 -5.69613516e-01 6.54802263e-01
-3.21748704e-02 1.32573676e+00 -1.03369974e-01 -6.95968866e-01
1.21110916e+00 4.22317505e-01 1.08423010e-01 -8.85367393e-01
-1.03478909e+00 -2.32884198e-01 4.45119649e-01 -6.36156678e-01
1.23673677e-01 -3.38170350e-01 -1.36777711e+00 -2.21980453e-01
-4.90568966e-01 5.63473344e-01 5.89988112e-01 5.94532847e-01
7.64240563e-01 4.99856025e-01 7.47169852e-01 7.55938947e-01
-8.07650328e-01 -9.35289860e-01 -1.39073459e-02 7.38570452e-01
-7.33887479e-02 -1.00229807e-01 -2.01934166e-02 -6.16329201e-02] | [9.904613494873047, 7.8536505699157715] |
0b7daaf2-2fb0-40ff-9576-14ce54cfb1b5 | cloud-detection-machine-learning-algorithms | 2012.10396 | null | https://arxiv.org/abs/2012.10396v1 | https://arxiv.org/pdf/2012.10396v1.pdf | Cloud detection machine learning algorithms for PROBA-V | This paper presents the development and implementation of a cloud detection algorithm for Proba-V. Accurate and automatic detection of clouds in satellite scenes is a key issue for a wide range of remote sensing applications. With no accurate cloud masking, undetected clouds are one of the most significant sources of error in both sea and land cover biophysical parameter retrieval. The objective of the algorithms presented in this paper is to detect clouds accurately providing a cloud flag per pixel. For this purpose, the method exploits the information of Proba-V using statistical machine learning techniques to identify the clouds present in Proba-V products. The effectiveness of the proposed method is successfully illustrated using a large number of real Proba-V images. | ['Gustau Camps-Valls', 'Jordi Muñoz-Marí', 'Gonzalo Mateo-García', 'Luis Gómez-Chova'] | 2020-12-09 | null | null | null | null | ['cloud-detection'] | ['computer-vision'] | [ 1.78938553e-01 -8.91280651e-01 1.73288897e-01 -2.18539283e-01
-6.39017701e-01 -9.46700752e-01 5.56572735e-01 2.14809254e-01
-1.31298378e-01 8.94872606e-01 -7.07248211e-01 -5.89837790e-01
-6.03548512e-02 -1.02823484e+00 -1.61938086e-01 -1.08617222e+00
-3.44071835e-01 1.31353572e-01 2.83443153e-01 -9.46713835e-02
1.98678270e-01 1.16263878e+00 -1.74672782e+00 7.88101405e-02
1.11797941e+00 1.09902871e+00 7.27433741e-01 7.27569461e-01
-2.14050099e-01 3.05476010e-01 -3.39944452e-01 3.87194276e-01
4.05080795e-01 -3.54377121e-01 -1.92323565e-01 2.31369451e-01
5.10394514e-01 -2.04030469e-01 3.81903112e-01 1.36528885e+00
5.28712235e-02 -2.12591901e-01 8.40589285e-01 -9.92424726e-01
2.30490297e-01 -7.30797499e-02 -6.38918698e-01 4.21768814e-01
-4.16626513e-01 2.05289964e-02 7.69530952e-01 -1.13097310e+00
8.06374475e-02 8.13652635e-01 5.38199484e-01 -2.81529248e-01
-1.03907609e+00 -8.72298658e-01 -2.51656204e-01 6.97733089e-02
-1.92648423e+00 -2.33064294e-01 3.94288659e-01 -9.30751145e-01
6.66733205e-01 8.33462179e-01 1.14408553e+00 -4.50952053e-01
2.77752697e-01 3.79936129e-01 1.76967549e+00 -6.05961502e-01
3.60344201e-01 4.28520083e-01 3.37252647e-01 3.26639235e-01
9.28075314e-01 2.19024211e-01 1.14230059e-01 -3.90026659e-01
5.26985765e-01 1.96728230e-01 -2.68774599e-01 -3.97287756e-02
-3.25395226e-01 9.30227399e-01 4.94379014e-01 2.76802897e-01
-6.77614868e-01 -1.17559277e-01 -1.11388437e-01 2.79502757e-02
6.41789258e-01 3.29512119e-01 -2.47826055e-01 4.02347744e-01
-1.42571294e+00 5.93356967e-01 2.39871398e-01 5.77041209e-01
8.38659108e-01 4.97008681e-01 1.54677361e-01 3.37292701e-01
5.80837548e-01 1.36406291e+00 -1.67736992e-01 -5.97778201e-01
2.51075197e-02 5.82225561e-01 6.83952451e-01 -1.09976256e+00
1.45881891e-01 -7.01284885e-01 -6.67462945e-01 6.66040361e-01
-7.03459606e-02 1.20274521e-01 -1.16684008e+00 6.46934748e-01
3.07890564e-01 1.68170840e-01 2.88397670e-01 1.01252699e+00
6.26699269e-01 1.03685308e+00 1.88276976e-01 -3.91011924e-01
1.44232857e+00 -1.52672753e-01 -7.97479153e-01 -3.47570628e-01
8.73016044e-02 -8.28680694e-01 3.92259657e-01 1.63526580e-01
-2.91472644e-01 -3.43704611e-01 -9.21720088e-01 7.82681704e-01
-6.23372853e-01 4.29741770e-01 7.68296599e-01 4.58331615e-01
-7.65837133e-01 2.84141749e-01 -6.83577240e-01 -2.82474399e-01
1.94860756e-01 1.03626601e-01 3.90985757e-02 5.10497130e-02
-9.29918408e-01 6.90773904e-01 3.95216972e-01 7.42190778e-01
-6.68678820e-01 -8.87754485e-02 -5.98128736e-01 1.87541723e-01
-7.38465711e-02 1.32472739e-01 8.31231952e-01 -1.20443261e+00
-3.89002979e-01 1.01476872e+00 -4.51997340e-01 -5.02452612e-01
2.18202531e-01 -1.22651733e-01 -5.66101313e-01 1.46184355e-01
1.25172734e-01 -7.22589493e-02 8.55348051e-01 -1.75287461e+00
-9.61556077e-01 -4.75963235e-01 -5.19304395e-01 1.60640255e-01
2.02107608e-01 2.39749104e-01 1.38698041e-01 -4.39052910e-01
5.17401636e-01 -9.15181220e-01 -2.18863875e-01 -6.29852787e-02
-1.04645351e-02 2.85459399e-01 1.26259053e+00 -6.47991538e-01
8.75613749e-01 -2.10664034e+00 -6.50737405e-01 6.31870031e-01
8.15450475e-02 7.83884108e-01 9.76620391e-02 4.19483393e-01
5.26676252e-02 3.02045017e-01 -6.16246581e-01 3.01580310e-01
-7.32674360e-01 4.40241128e-01 -6.48339331e-01 6.73066676e-01
4.47812915e-01 2.97636807e-01 -5.80317378e-01 -5.16852498e-01
4.12922055e-01 3.35819483e-01 2.58219212e-01 3.72564614e-01
-2.46654376e-01 1.96128607e-01 -6.06349707e-01 8.91790986e-01
1.40098906e+00 8.35169926e-02 5.17158285e-02 2.13772446e-01
-8.73010159e-01 -1.91909112e-02 -1.21799803e+00 2.26459801e-01
1.26507610e-01 8.30944657e-01 4.31465268e-01 -6.82868063e-01
1.00452971e+00 3.39624912e-01 4.67293188e-02 -3.11449051e-01
-4.89132181e-02 5.85714459e-01 8.80129859e-02 -6.66485369e-01
6.88374817e-01 -3.45185965e-01 6.44154131e-01 -6.76960647e-02
-6.34529173e-01 -5.94624698e-01 -1.19010724e-01 -1.45834172e-02
3.02511752e-01 -2.35317424e-01 5.11806488e-01 -4.78181332e-01
6.17851317e-01 7.00701177e-01 6.28375828e-01 7.82775104e-01
-2.51480788e-01 4.10456300e-01 -1.20664693e-01 -2.85250038e-01
-8.00781071e-01 -7.20822513e-01 -4.87796426e-01 3.79024684e-01
-1.39659513e-02 3.43783617e-01 -3.85019362e-01 1.24515601e-01
4.70228672e-01 5.42101145e-01 -4.03406978e-01 5.33012271e-01
-9.75475013e-02 -1.13817656e+00 -4.29124646e-02 1.54359907e-01
6.13436520e-01 -8.54069412e-01 -5.38090289e-01 1.34552822e-01
-7.07928985e-02 -1.08537090e+00 4.21319664e-01 1.95190743e-01
-1.25593948e+00 -1.03784227e+00 -4.27124918e-01 -5.56395113e-01
6.71806276e-01 1.08378494e+00 1.01224756e+00 5.39925277e-01
-5.04654348e-01 -2.13957846e-01 -4.98823464e-01 -7.52622485e-01
-4.47253466e-01 -3.44954073e-01 -1.98782086e-01 1.94489136e-01
4.60854650e-01 -1.75709441e-01 -4.89643663e-01 1.35275885e-01
-6.72826588e-01 -2.08499014e-01 6.37195408e-01 1.91354468e-01
8.52068186e-01 4.65766311e-01 -6.62557408e-02 -7.21117437e-01
1.13520600e-01 -4.26187217e-01 -1.69468164e+00 3.51345539e-02
-6.63850665e-01 -5.48152387e-01 -1.20688096e-01 1.62077606e-01
-7.01711714e-01 4.17813629e-01 3.30789894e-01 -4.56245154e-01
-2.49830663e-01 8.81016433e-01 -8.77399743e-03 -5.11521995e-01
3.73615026e-01 7.86332786e-01 -2.13182256e-01 -5.06057441e-01
-2.69936770e-01 9.75785792e-01 5.02460122e-01 4.23663482e-02
1.23152685e+00 7.09671795e-01 2.59563684e-01 -1.59760094e+00
-5.20349324e-01 -1.30186820e+00 -5.44049025e-01 -4.99419272e-01
7.14447260e-01 -1.31161022e+00 -1.80002481e-01 8.18391681e-01
-9.67377663e-01 4.22635339e-02 2.96839207e-01 6.34348214e-01
3.53577614e-01 2.06640244e-01 2.04226688e-01 -1.82525802e+00
-5.45067430e-01 -9.15082395e-01 7.82814205e-01 2.46155694e-01
2.95682341e-01 -6.92152441e-01 1.00397855e-01 2.36616448e-01
5.56180537e-01 4.34845060e-01 5.44950247e-01 -1.16757721e-01
-1.16249931e+00 -5.09383380e-01 -6.60551071e-01 4.87462282e-01
6.45867407e-01 7.21311331e-01 -1.22307003e+00 -3.42520595e-01
-7.87656307e-02 2.07596436e-01 9.22223806e-01 5.78226447e-01
7.75701880e-01 -2.38671497e-01 -1.96636364e-01 6.05324268e-01
2.20302439e+00 3.97802442e-01 6.97642863e-01 3.44772816e-01
4.19669211e-01 3.39179575e-01 1.10026014e+00 4.32833254e-01
-2.74268210e-01 2.53231287e-01 1.01596320e+00 -3.16301852e-01
1.45053491e-01 4.75464255e-01 -9.95729193e-02 5.57507575e-01
-4.67344582e-01 -1.19028546e-01 -1.16451037e+00 9.91867602e-01
-1.46780765e+00 -1.22268474e+00 -9.50312674e-01 2.24846935e+00
4.24648941e-01 -2.39903450e-01 -3.59355360e-01 2.18503609e-01
6.97631061e-01 2.59808034e-01 -4.24208380e-02 -1.46520734e-01
-2.97364503e-01 1.99180573e-01 1.07357943e+00 6.53727949e-01
-1.36537945e+00 8.60283256e-01 6.46805000e+00 3.55414540e-01
-1.51243854e+00 -1.81841608e-02 1.02700368e-01 5.00916302e-01
-6.20114803e-02 2.00893000e-01 -7.70455658e-01 3.17274779e-01
4.69939977e-01 1.89016178e-01 6.57920986e-02 7.52430081e-01
1.03245020e+00 -7.42911220e-01 1.87543314e-02 7.57074296e-01
-3.23953629e-01 -1.30599439e+00 1.06103227e-01 1.11918256e-01
8.30510616e-01 3.42047542e-01 -1.97472185e-01 -2.36995623e-01
1.73646688e-01 -8.95969927e-01 5.92619836e-01 4.28747952e-01
8.13256443e-01 -5.54181099e-01 9.94568288e-01 1.81156129e-01
-1.30655062e+00 1.77691966e-01 -4.22257423e-01 -5.04040837e-01
-3.52203518e-01 8.67858350e-01 -9.61911500e-01 3.37457359e-01
7.27308810e-01 3.72235984e-01 -4.13909972e-01 1.59441006e+00
-2.53834486e-01 1.17510819e+00 -4.38688487e-01 -5.61643019e-02
4.46584582e-01 -6.21035397e-01 5.55651367e-01 1.25699472e+00
4.62632716e-01 7.01812804e-01 -1.69081036e-02 5.71209490e-01
1.82983667e-01 2.23367915e-01 -6.33119285e-01 -3.29322010e-01
6.95521832e-01 1.18793488e+00 -6.06527448e-01 -3.10759097e-01
-2.63047606e-01 4.27713633e-01 -5.21958351e-01 3.63633066e-01
-3.14954489e-01 -2.03071386e-01 1.04232252e+00 5.55213392e-01
4.45635766e-01 -5.21407604e-01 -2.45038763e-01 -7.26921976e-01
-1.47497401e-01 -6.97276533e-01 -3.89961489e-02 -1.03589523e+00
-8.23340356e-01 4.46384460e-01 -4.81951535e-02 -1.58469176e+00
1.52929261e-01 -6.20152295e-01 -9.03230786e-01 1.57654786e+00
-2.03183126e+00 -1.09518647e+00 -1.06104517e+00 2.92238325e-01
1.94683552e-01 -1.17124438e-01 9.15287793e-01 -1.74245104e-01
-1.55881792e-01 -4.90449160e-01 8.43729317e-01 1.63916081e-01
1.57169297e-01 -1.21863449e+00 2.41177917e-01 1.36438823e+00
-2.32184112e-01 1.59382984e-01 1.08618557e+00 -9.67544794e-01
-1.18082249e+00 -1.45047176e+00 9.24606442e-01 3.72535139e-01
5.39164722e-01 -7.75593966e-02 -1.13078475e+00 3.57147843e-01
2.56861676e-03 2.58621812e-01 3.55059534e-01 -3.08577597e-01
1.46121010e-01 -5.69231987e-01 -1.17013705e+00 -1.90441068e-02
-1.98847607e-01 -4.94152427e-01 -2.24961340e-01 4.87672508e-01
7.57762939e-02 -1.21551514e-01 -4.27054465e-01 3.74474466e-01
5.03116310e-01 -7.92810857e-01 5.66060424e-01 -1.67380214e-01
-9.05542299e-02 -7.16865301e-01 -3.77096951e-01 -9.32085454e-01
-2.74969429e-01 -9.53672528e-02 5.05089283e-01 1.00186992e+00
1.68906793e-01 -6.70280159e-01 6.43213034e-01 1.46433771e-01
2.84006089e-01 1.26984745e-01 -8.78684402e-01 -8.35073709e-01
-1.60132080e-01 -2.31269032e-01 3.46524209e-01 1.03637075e+00
-9.50231791e-01 -4.71652627e-01 -1.42627552e-01 1.17596889e+00
1.00096512e+00 7.47989893e-01 7.83222139e-01 -1.65864360e+00
1.78117052e-01 6.91062212e-02 -4.18250024e-01 -4.85814475e-02
-4.44355220e-01 -3.73789668e-01 2.35260710e-01 -1.53645551e+00
4.20435280e-01 -8.44289064e-01 -6.28895611e-02 4.77923214e-01
-3.44820082e-01 2.91443974e-01 2.53382444e-01 8.06307375e-01
4.31747824e-01 3.24640840e-01 8.10779929e-01 -2.23935261e-01
-1.90895602e-01 4.31116611e-01 -8.04428533e-02 5.62667608e-01
9.60341215e-01 -8.64932656e-01 5.94261149e-03 -3.57363969e-01
-2.16044541e-02 -2.43537780e-02 7.80057847e-01 -1.10333276e+00
-3.23277235e-01 -6.46427512e-01 2.98823297e-01 -1.23084724e+00
3.06816041e-01 -1.07316685e+00 2.87753284e-01 7.09605157e-01
5.09139895e-01 -1.23164825e-01 4.79469568e-01 4.02294844e-01
-5.22075057e-01 -3.85869175e-01 1.03266680e+00 -3.22111577e-01
-9.72521186e-01 1.80062607e-01 -7.64689147e-01 -6.83282435e-01
9.93754804e-01 -2.21947744e-01 -1.52094260e-01 -2.86161631e-01
-3.98250222e-01 3.17114502e-01 5.24292827e-01 1.32105441e-03
7.66658306e-01 -7.55698562e-01 -1.02542961e+00 2.34428912e-01
3.22355092e-01 -1.15523383e-01 -1.22259296e-01 3.76495987e-01
-1.22103119e+00 5.38237989e-01 -4.20656018e-02 -8.85390699e-01
-1.92583930e+00 6.44231513e-02 6.93952978e-01 1.10915713e-01
-1.16214670e-01 4.43216681e-01 -1.05935872e-01 1.82132065e-01
-3.00125539e-01 -3.29675525e-01 -3.97033304e-01 3.61409783e-02
5.14888465e-01 -9.94935334e-02 1.38606399e-01 -9.32926536e-01
-5.64204633e-01 5.81989288e-01 2.58923918e-01 -6.21205308e-02
1.17596686e+00 -1.15718029e-01 -5.69508135e-01 5.21245062e-01
6.40126646e-01 2.85894901e-01 -8.62623572e-01 -2.77502209e-01
-1.47006169e-01 -1.02165341e+00 6.98174775e-01 -8.54707360e-01
-1.06025958e+00 1.03457308e+00 1.17442095e+00 1.14341348e-01
1.11015737e+00 -3.09981823e-01 -1.09937573e-02 3.43378156e-01
1.46306932e-01 -6.69621527e-01 -7.62893975e-01 1.46689862e-01
8.91657531e-01 -1.57057786e+00 6.27712131e-01 -7.83622324e-01
-3.22403967e-01 8.56810331e-01 6.64366260e-02 -3.59525867e-02
6.72234178e-01 9.17307660e-02 4.31141466e-01 -2.85576761e-01
-1.61706164e-01 -7.43136227e-01 -9.04011726e-02 4.97884959e-01
1.57535315e-01 6.81418478e-01 -3.86218667e-01 -1.87074482e-01
2.82642454e-01 3.81711237e-02 6.91040635e-01 1.16543472e+00
-1.12014580e+00 -5.56217313e-01 -1.27047396e+00 5.43699443e-01
-4.87631053e-01 -3.65913808e-01 -3.20792884e-01 9.42346752e-01
-3.65305622e-03 1.04403782e+00 1.10896602e-01 1.06829420e-01
-1.63767233e-01 -1.66683033e-01 -3.73650193e-02 -4.82082456e-01
-2.99664378e-01 4.25096393e-01 -1.31872138e-02 -6.78572804e-03
-7.91201711e-01 -1.00288141e+00 -1.01656854e+00 -1.38382778e-01
-7.08207548e-01 5.00260413e-01 1.14014995e+00 7.84560323e-01
-1.21992737e-01 2.81374399e-02 9.48403299e-01 -6.87036276e-01
-3.12929362e-01 -9.73630667e-01 -1.25159168e+00 -1.69956461e-01
7.73645580e-01 -6.66004181e-01 -5.93117535e-01 5.50524853e-02] | [9.68233871459961, -1.7948395013809204] |
cca88d25-3553-4829-a11c-ba2439929e55 | vis2mus-exploring-multimodal-representation | 2211.05543 | null | https://arxiv.org/abs/2211.05543v1 | https://arxiv.org/pdf/2211.05543v1.pdf | Vis2Mus: Exploring Multimodal Representation Mapping for Controllable Music Generation | In this study, we explore the representation mapping from the domain of visual arts to the domain of music, with which we can use visual arts as an effective handle to control music generation. Unlike most studies in multimodal representation learning that are purely data-driven, we adopt an analysis-by-synthesis approach that combines deep music representation learning with user studies. Such an approach enables us to discover \textit{interpretable} representation mapping without a huge amount of paired data. In particular, we discover that visual-to-music mapping has a nice property similar to equivariant. In other words, we can use various image transformations, say, changing brightness, changing contrast, style transfer, to control the corresponding transformations in the music domain. In addition, we released the Vis2Mus system as a controllable interface for symbolic music generation. | ['Gus Xia', 'Ying Shan', 'Kai Shao', 'Yixiao Zhang', 'Runbang Zhang'] | 2022-11-10 | null | null | null | null | ['music-generation', 'music-generation'] | ['audio', 'music'] | [ 2.84226060e-01 -2.24578172e-01 -1.79846466e-01 9.10136290e-03
-4.99047041e-01 -1.09341407e+00 8.14767659e-01 -3.78598809e-01
9.89553034e-02 3.46313596e-01 3.96323383e-01 -3.99930924e-02
-1.14156157e-01 -7.47123361e-01 -6.68804526e-01 -5.29831111e-01
6.04060471e-01 3.09763312e-01 -3.43176067e-01 -6.66136026e-01
3.16979200e-01 3.67229640e-01 -1.65862823e+00 4.78085190e-01
7.35357046e-01 5.73654056e-01 9.71039832e-02 4.82291371e-01
-3.98989737e-01 7.03959942e-01 -5.12084603e-01 -2.69110292e-01
2.12130412e-01 -8.65646005e-01 -4.64930028e-01 -1.38980895e-01
4.39881414e-01 1.30644571e-02 -2.96689391e-01 1.07595599e+00
7.94651747e-01 3.03152382e-01 8.61400604e-01 -1.46411610e+00
-1.08886015e+00 8.11276555e-01 -5.39139807e-01 -5.52845955e-01
8.41648936e-01 2.93385446e-01 9.58761871e-01 -7.03076243e-01
8.64566445e-01 1.52906251e+00 1.89560279e-01 7.41593897e-01
-1.43261731e+00 -9.22302246e-01 -1.41717002e-01 4.41967957e-02
-1.16008508e+00 -5.72831929e-01 1.15999854e+00 -7.31952012e-01
1.42485306e-01 6.82582676e-01 1.02425492e+00 1.47558081e+00
-3.29849780e-01 7.77053118e-01 1.07566202e+00 -6.51684165e-01
6.10316694e-02 7.09635392e-02 -5.91810107e-01 4.69548821e-01
-2.13555068e-01 2.55302548e-01 -8.50212455e-01 9.55921113e-02
1.26121926e+00 -1.55531451e-01 -2.67594665e-01 -4.98135984e-01
-1.77132690e+00 6.63627744e-01 3.30311000e-01 3.08122396e-01
2.93464754e-02 6.03936613e-01 3.38378310e-01 6.19445443e-01
-1.64619014e-01 1.06588042e+00 -3.23158018e-02 -9.83542725e-02
-7.45540023e-01 3.33481044e-01 4.64494526e-01 1.05096817e+00
4.34367597e-01 6.04247510e-01 -7.04062045e-01 8.06396127e-01
3.13732326e-01 8.34596813e-01 6.77946568e-01 -1.19683421e+00
1.80680215e-01 5.58306038e-01 1.45049943e-02 -9.29419816e-01
-9.56420079e-02 -1.52758569e-01 -9.67533588e-01 6.08338475e-01
3.79610747e-01 6.25461191e-02 -7.01491058e-01 1.92400038e+00
2.58690342e-02 2.05353647e-01 -2.15964820e-02 1.00508022e+00
1.05478692e+00 4.33730990e-01 -5.22894971e-02 -5.88756688e-02
1.29989183e+00 -7.02993691e-01 -1.00721991e+00 4.14569318e-01
2.27422759e-01 -9.73611951e-01 1.80459559e+00 4.61848199e-01
-1.12014353e+00 -8.05837512e-01 -9.60657597e-01 -2.15141699e-01
-3.81246120e-01 2.39460766e-01 6.64474308e-01 2.29681358e-01
-8.44458818e-01 6.97449446e-01 -4.12277848e-01 -2.58996129e-01
1.29524663e-01 1.93584129e-01 -4.55103695e-01 3.77687693e-01
-1.18576181e+00 4.48755354e-01 3.53710681e-01 -2.32819557e-01
-9.82719421e-01 -6.55642569e-01 -8.65249455e-01 -8.80897567e-02
1.71930537e-01 -1.10024607e+00 1.20293844e+00 -1.38874376e+00
-2.04663801e+00 7.83301055e-01 1.38531089e-01 2.93943375e-01
5.75006425e-01 -5.68837039e-02 -2.83414125e-01 -1.17987648e-01
-8.84173289e-02 7.39563763e-01 1.17978084e+00 -1.49996924e+00
-1.42572867e-02 -9.34398770e-02 1.52500555e-01 3.56287330e-01
-5.25409043e-01 -9.48523805e-02 -5.43552637e-01 -1.10168946e+00
-4.78075221e-02 -1.02606714e+00 -8.86775739e-03 2.82064378e-01
-6.76771522e-01 1.09441936e-01 6.13669455e-01 -3.78953964e-01
1.20761621e+00 -2.46502733e+00 9.14005816e-01 3.88907492e-01
1.12017736e-01 -5.78293577e-02 -4.52052742e-01 5.19844234e-01
-2.82026112e-01 2.82702118e-01 7.55730569e-02 -2.61849135e-01
3.27618778e-01 -1.83169141e-01 -6.98830485e-01 3.63726951e-02
-8.07612091e-02 1.06449151e+00 -9.72362518e-01 -4.94401127e-01
2.78626621e-01 5.08187592e-01 -7.01814890e-01 4.96246785e-01
-4.38047409e-01 1.00409698e+00 -5.43921590e-01 6.07627571e-01
2.73010433e-01 1.28533412e-02 1.51108488e-01 -3.44856113e-01
-1.51385516e-01 -1.86948299e-01 -1.15285385e+00 2.34314084e+00
-5.45020044e-01 6.75198436e-01 -1.78465530e-01 -5.60509980e-01
1.12054968e+00 4.30311263e-01 3.94323677e-01 -6.58789754e-01
7.57766962e-02 -7.46646523e-02 -3.56234983e-02 -5.46893060e-01
3.61300230e-01 -2.47574359e-01 -1.20609321e-01 4.93492812e-01
-6.54322952e-02 -5.39904296e-01 -9.91203636e-02 -4.94633387e-05
6.62397265e-01 6.59859896e-01 2.75591314e-01 3.56775969e-02
2.44366124e-01 -2.52372026e-01 1.24002255e-01 4.90841687e-01
3.58437985e-01 7.70419538e-01 5.69912612e-01 -7.48659223e-02
-1.03019273e+00 -1.28780091e+00 2.72330731e-01 1.39147711e+00
1.46201834e-01 -4.35479015e-01 -5.94701886e-01 -5.90140186e-02
-8.15451294e-02 7.76551366e-01 -6.19195223e-01 -3.09061199e-01
-3.99020791e-01 3.56498659e-02 6.56799614e-01 3.37479085e-01
4.20715690e-01 -1.53210521e+00 -1.93740115e-01 -1.58879682e-01
-1.32357210e-01 -5.81165433e-01 -6.67072415e-01 -2.33934432e-01
-6.60887599e-01 -7.81656623e-01 -8.04604471e-01 -7.69053519e-01
3.92055064e-01 2.30951756e-02 1.12360024e+00 -1.26149058e-01
-1.87392890e-01 4.90412563e-01 -3.55976611e-01 -4.15703624e-01
-4.13427621e-01 5.16616963e-02 -3.38497423e-02 -6.33055530e-03
-2.90231973e-01 -9.47576821e-01 -5.25352001e-01 1.90183729e-01
-1.01380062e+00 5.66682339e-01 4.42134053e-01 6.19580507e-01
7.14575708e-01 -4.90230620e-01 6.43602371e-01 -6.89734697e-01
9.25337374e-01 -2.46462315e-01 -3.38976860e-01 2.80999184e-01
-9.76170897e-02 9.97757465e-02 6.09691262e-01 -9.15657759e-01
-8.70905936e-01 2.50516444e-01 2.70163745e-01 -9.92086291e-01
-1.22729003e-01 5.12501955e-01 -4.00529325e-01 1.23020872e-01
8.43914866e-01 2.62838215e-01 1.28966793e-01 -5.29014230e-01
1.07097423e+00 4.97057498e-01 7.03573048e-01 -7.29095995e-01
1.11024797e+00 3.44426870e-01 1.11440532e-01 -5.43717802e-01
-3.09254557e-01 1.22920856e-01 -5.62679768e-01 -3.53575855e-01
1.00322628e+00 -8.05067182e-01 -1.15951443e+00 8.63246098e-02
-1.00257897e+00 -4.86691624e-01 -6.19986057e-01 3.36798579e-01
-1.12724721e+00 6.26204386e-02 -3.92461509e-01 -6.44963086e-01
-2.99792051e-01 -9.71733809e-01 1.20944142e+00 2.11391494e-01
-5.98567903e-01 -8.31015170e-01 2.31760949e-01 2.04660058e-01
3.43466669e-01 6.56379819e-01 1.01776636e+00 -2.61069924e-01
-5.87208509e-01 6.66829869e-02 -3.78191061e-02 -8.00692439e-02
3.45121473e-01 2.15142295e-01 -9.78974462e-01 -1.74164981e-01
-5.02173901e-01 -3.77817512e-01 5.24075031e-01 1.75312430e-01
1.56167626e+00 -7.89368972e-02 3.87575515e-02 9.10358608e-01
9.93165493e-01 2.66449869e-01 8.59506965e-01 3.11165363e-01
1.03049684e+00 3.66134226e-01 5.08764386e-01 5.20501196e-01
4.42693830e-02 1.13658035e+00 3.80096018e-01 -3.30509543e-01
-3.94563347e-01 -7.07653522e-01 4.72734988e-01 7.38239110e-01
-4.98725563e-01 -1.51881352e-01 -5.16609192e-01 1.04718029e-01
-1.99858201e+00 -1.08937621e+00 2.72175878e-01 2.17989230e+00
9.62640882e-01 -4.20889616e-01 1.58912733e-01 6.71254024e-02
7.11556375e-01 2.50278674e-02 -4.00034040e-01 -2.85012305e-01
-6.28775209e-02 3.61616820e-01 -2.11493850e-01 1.87812313e-01
-8.98618400e-01 1.26422632e+00 6.51761723e+00 9.34884489e-01
-1.28561342e+00 -2.33359843e-01 -1.09856904e-01 -1.78852379e-01
-7.40135312e-01 -1.48931488e-01 7.51453489e-02 2.99771190e-01
3.18584919e-01 -2.48953685e-01 1.10935271e+00 6.33278012e-01
4.43944037e-01 6.31520152e-01 -1.35317051e+00 1.49890113e+00
-1.32574579e-02 -1.36445761e+00 6.94252729e-01 -5.99039122e-02
6.45103633e-01 -8.43705833e-01 5.35702050e-01 2.59287417e-01
1.94374964e-01 -1.40735567e+00 8.91617715e-01 1.10285664e+00
1.48212707e+00 -6.41929328e-01 2.26398911e-02 -8.54984112e-03
-1.18283331e+00 9.50478688e-02 1.91346910e-02 3.13354582e-02
-3.19529474e-02 -2.09410846e-01 -3.61698776e-01 5.08627057e-01
4.02079433e-01 8.81699800e-01 -3.63770723e-01 8.62026691e-01
-3.87446761e-01 3.02539915e-01 2.97602803e-01 1.37148708e-01
-4.24397409e-01 -5.06617606e-01 6.11938596e-01 8.11148942e-01
8.14998448e-01 -7.97231495e-02 1.10943250e-01 1.18053436e+00
-2.75298953e-01 3.37549448e-01 -1.10406625e+00 -3.98440540e-01
3.95775527e-01 1.16463149e+00 -3.48587334e-01 -2.40761712e-01
-1.06147332e-02 1.12084889e+00 1.23619080e-01 5.86290061e-01
-7.03839064e-01 -3.18313688e-01 7.22415924e-01 1.29438385e-01
7.19840080e-02 -2.61024356e-01 -2.82541513e-01 -1.33488798e+00
-5.23103893e-01 -1.22413743e+00 2.33427039e-04 -1.44398975e+00
-1.21385241e+00 3.11557263e-01 -2.73818914e-02 -1.64786208e+00
-4.15534377e-01 -4.42169100e-01 -8.23411584e-01 6.16158187e-01
-9.04065728e-01 -1.43590653e+00 -3.96467358e-01 1.01321387e+00
2.37382144e-01 -6.96998656e-01 1.04694724e+00 3.06261867e-01
-1.88232571e-01 7.37649441e-01 -1.02558590e-01 1.39900774e-01
9.90617633e-01 -1.29515052e+00 2.94773071e-03 9.15773064e-02
5.88252366e-01 6.53544307e-01 6.40842199e-01 -2.94688553e-01
-1.71866715e+00 -7.86022007e-01 9.68037993e-02 -6.16938472e-01
6.24555290e-01 -4.42574143e-01 -6.10545278e-01 6.07194841e-01
4.98671710e-01 -5.26746631e-01 9.31325197e-01 1.25434831e-01
-5.44500768e-01 -2.88444217e-02 -6.82515383e-01 1.27586091e+00
1.28560770e+00 -7.40898609e-01 -4.38771665e-01 2.18165994e-01
8.57252240e-01 -4.99409258e-01 -8.43147576e-01 1.64462313e-01
7.70119727e-01 -5.51700473e-01 1.07387066e+00 -9.22903180e-01
6.26172602e-01 -6.94157600e-01 -2.32812852e-01 -1.54655170e+00
-4.56407875e-01 -9.70421910e-01 1.03333198e-01 1.51268613e+00
3.59053880e-01 -8.34682360e-02 2.89555848e-01 4.19460870e-02
-1.53248817e-01 -7.60757923e-03 -5.50595582e-01 -5.15471339e-01
-7.92533904e-03 -3.52114618e-01 8.95943403e-01 1.38654041e+00
6.15410134e-02 4.47434098e-01 -7.51971483e-01 -3.23131710e-01
2.77271956e-01 4.55838114e-01 1.37186289e+00 -1.19008684e+00
-5.77730656e-01 -7.27368832e-01 -2.66770989e-01 -5.35357118e-01
4.09160525e-01 -1.29820049e+00 -2.12704390e-01 -1.36325192e+00
2.64586270e-01 -1.82295024e-01 -4.00322050e-01 5.51518440e-01
1.46709666e-01 4.15563405e-01 7.60175228e-01 2.77687609e-01
-3.45092088e-01 7.03833163e-01 2.00258493e+00 -2.96636164e-01
-4.75548238e-01 -2.43843779e-01 -9.56062496e-01 6.54115319e-01
7.02820480e-01 -4.58401963e-02 -6.24474227e-01 -3.97747755e-01
7.24341273e-01 2.72114038e-01 3.68582517e-01 -8.14159751e-01
-1.45241991e-01 -5.21781445e-01 3.50261509e-01 4.58993055e-02
4.12686497e-01 -7.87078857e-01 5.42774439e-01 9.15006548e-02
-7.67502367e-01 -1.05542228e-01 2.74294168e-01 3.49377453e-01
-2.85607189e-01 -1.06476722e-02 4.44484085e-01 -8.80595297e-02
-4.90907490e-01 1.58277884e-01 -1.37902901e-01 2.76357727e-03
7.07742035e-01 -6.15313910e-02 -2.61125892e-01 -9.23870206e-01
-9.80282187e-01 -9.05314609e-02 5.57104945e-01 6.40731692e-01
4.92773741e-01 -2.06843877e+00 -6.55816376e-01 2.02398524e-01
3.70271385e-01 -4.05696154e-01 1.51458472e-01 4.36218500e-01
-4.21942294e-01 -1.73723862e-01 -5.50269783e-01 -5.81747532e-01
-1.17651927e+00 7.13799596e-01 9.34304073e-02 2.67132849e-01
-4.77868706e-01 2.50499636e-01 4.57241505e-01 -5.51455081e-01
1.40802979e-01 -1.69383645e-01 -4.35439885e-01 2.21672013e-01
3.82809222e-01 1.23161666e-01 -6.35679424e-01 -3.58842075e-01
1.30140126e-01 8.75589669e-01 6.52931750e-01 -5.37323356e-01
1.05504632e+00 2.38336384e-01 -1.37905464e-01 7.77792990e-01
7.59840310e-01 2.55390406e-01 -1.01718831e+00 1.74266100e-01
-3.89084816e-01 -5.54537654e-01 -2.30133057e-01 -7.52613962e-01
-1.11684263e+00 9.18095529e-01 7.34427214e-01 2.72562176e-01
1.16258419e+00 6.42363876e-02 2.01683089e-01 4.88110691e-01
7.60484533e-03 -9.31630433e-01 5.66174090e-01 2.83160537e-01
1.68149018e+00 -1.13329530e+00 -2.54421473e-01 -1.45523742e-01
-9.67782497e-01 1.17980468e+00 5.18282235e-01 -2.34816223e-02
3.81924480e-01 6.87479377e-02 1.88919306e-01 -1.03412457e-01
-6.24266565e-01 -2.82893717e-01 6.70955002e-01 7.75743008e-01
7.05353260e-01 3.33494514e-01 -5.85274696e-02 5.09618461e-01
-7.13699877e-01 1.60560608e-01 2.72920936e-01 4.05436844e-01
-1.69188663e-01 -1.23698258e+00 -4.98372167e-01 -3.91900986e-02
1.46916732e-01 -1.06455408e-01 -7.06203461e-01 9.34479594e-01
1.98364347e-01 6.23754799e-01 3.27087231e-02 -6.36448979e-01
5.56251228e-01 7.59208128e-02 7.07956135e-01 -4.86591160e-01
-5.28122306e-01 3.20629597e-01 -1.75709084e-01 -5.76624990e-01
-4.77527708e-01 -4.83172327e-01 -1.13157487e+00 -2.92666107e-01
1.61740839e-01 -3.14238787e-01 5.43875158e-01 5.73000491e-01
2.67314851e-01 8.95043850e-01 5.49892008e-01 -1.02597010e+00
-1.33112213e-02 -9.41747367e-01 -7.62148798e-01 8.14108729e-01
1.23503603e-01 -6.72980189e-01 -1.16952673e-01 3.46324980e-01] | [15.865254402160645, 5.385765552520752] |
20a2af42-bbff-4ef3-88c6-4ce91a8243ce | slaps-self-supervision-improves-structure-1 | 2102.05034 | null | https://arxiv.org/abs/2102.05034v2 | https://arxiv.org/pdf/2102.05034v2.pdf | SLAPS: Self-Supervision Improves Structure Learning for Graph Neural Networks | Graph neural networks (GNNs) work well when the graph structure is provided. However, this structure may not always be available in real-world applications. One solution to this problem is to infer a task-specific latent structure and then apply a GNN to the inferred graph. Unfortunately, the space of possible graph structures grows super-exponentially with the number of nodes and so the task-specific supervision may be insufficient for learning both the structure and the GNN parameters. In this work, we propose the Simultaneous Learning of Adjacency and GNN Parameters with Self-supervision, or SLAPS, a method that provides more supervision for inferring a graph structure through self-supervision. A comprehensive experimental study demonstrates that SLAPS scales to large graphs with hundreds of thousands of nodes and outperforms several models that have been proposed to learn a task-specific graph structure on established benchmarks. | ['Seyed Mehran Kazemi', 'Layla El Asri', 'Bahare Fatemi'] | 2021-02-09 | slaps-self-supervision-improves-structure | http://proceedings.neurips.cc/paper/2021/hash/bf499a12e998d178afd964adf64a60cb-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/bf499a12e998d178afd964adf64a60cb-Paper.pdf | neurips-2021-12 | ['graph-structure-learning'] | ['graphs'] | [ 3.15622836e-01 5.58804333e-01 -3.59754801e-01 -3.86230171e-01
-5.41603751e-02 -5.00637829e-01 3.98548603e-01 2.34934911e-01
-3.99364531e-03 7.72768736e-01 6.24786830e-04 -4.81880724e-01
-2.83583492e-01 -9.17269588e-01 -9.10887361e-01 -4.96903241e-01
-3.20883304e-01 8.49286199e-01 2.94731945e-01 -8.17947164e-02
-1.54135644e-01 3.16820771e-01 -1.14836323e+00 -1.25847757e-01
8.77021194e-01 5.90552866e-01 3.78525138e-01 6.66367650e-01
-1.39048815e-01 8.28822196e-01 -4.32472587e-01 -3.43510956e-01
3.28161865e-01 -2.99382716e-01 -9.26289022e-01 2.64946193e-01
7.39884675e-01 -1.76452726e-01 -5.57103276e-01 1.26354611e+00
4.61894982e-02 3.14925201e-02 4.77769226e-01 -1.48491240e+00
-4.78265047e-01 1.18234241e+00 -5.46395481e-01 1.80134311e-01
1.61632389e-01 -1.25155464e-01 1.29450524e+00 -2.11165115e-01
6.66589558e-01 1.19252932e+00 7.72225380e-01 2.51911700e-01
-1.46206713e+00 -6.22557580e-01 3.79796773e-01 7.58125037e-02
-1.28531456e+00 -1.76238745e-01 1.03205752e+00 -2.92213976e-01
1.10417330e+00 -1.88247368e-01 7.02208579e-01 1.28759658e+00
1.03654914e-01 4.44478214e-01 6.66237235e-01 -1.96321502e-01
7.71175399e-02 -2.41280213e-01 3.70143235e-01 1.15540481e+00
7.47600436e-01 -7.32350424e-02 -4.99411702e-01 -1.71908870e-01
1.06079578e+00 -7.18284249e-02 -7.99174607e-02 -6.51790917e-01
-1.07581675e+00 8.18849862e-01 8.80935669e-01 1.97691694e-01
-2.89769620e-01 3.86870205e-01 4.52668011e-01 5.38656592e-01
4.83608425e-01 5.57161868e-01 -5.01381695e-01 2.15177596e-01
-7.68879414e-01 -2.11846411e-01 1.15785968e+00 1.12661469e+00
1.20402598e+00 2.89991021e-01 2.58540481e-01 4.61035669e-01
1.99372962e-01 1.94286078e-01 8.63257200e-02 -8.14079046e-01
7.14399159e-01 1.00548708e+00 -4.45443332e-01 -1.36474693e+00
-5.74556291e-01 -6.83933616e-01 -1.27457118e+00 -2.19969824e-01
5.66312790e-01 -2.12385312e-01 -1.06866169e+00 1.87086630e+00
2.10657850e-01 4.41629946e-01 -3.50469053e-02 6.11891210e-01
8.77889335e-01 2.88107365e-01 -2.01858625e-01 9.52110216e-02
9.56410527e-01 -1.26848650e+00 -5.99016190e-01 -8.90894055e-01
7.15465784e-01 -9.47183445e-02 9.81153727e-01 1.07162401e-01
-6.49168849e-01 -4.86575603e-01 -1.01660275e+00 1.19753927e-01
-1.95570275e-01 -1.96092054e-01 1.00379193e+00 4.06227291e-01
-1.34296036e+00 7.27767885e-01 -8.96925747e-01 -4.63475108e-01
2.02773690e-01 5.51428795e-01 -7.20462739e-01 -2.43511930e-01
-1.14875889e+00 5.46015620e-01 9.79671061e-01 3.00150037e-01
-9.68240678e-01 -2.07358629e-01 -1.22755826e+00 3.98260623e-01
8.47020447e-01 -9.75054502e-01 7.22210586e-01 -1.01493943e+00
-1.13652778e+00 4.68943179e-01 1.13325259e-02 -6.54297650e-01
1.44188300e-01 7.23024979e-02 -1.44556478e-01 1.26829132e-01
4.88098757e-03 4.56101865e-01 8.53594899e-01 -1.11328077e+00
-9.24368426e-02 -4.26667482e-01 4.36514527e-01 1.57185987e-01
-4.03958559e-01 -4.20120537e-01 -5.78508615e-01 -2.69604355e-01
4.22839999e-01 -9.83882308e-01 -4.80301946e-01 -2.98081905e-01
-9.59577084e-01 -2.40366995e-01 7.86013901e-01 -5.47384858e-01
1.03919697e+00 -1.84713531e+00 3.30571085e-01 4.84001875e-01
6.79686606e-01 2.95696128e-02 -3.41834635e-01 5.18750608e-01
-9.49887484e-02 6.68654367e-02 -2.68364757e-01 -4.79464114e-01
-1.08065158e-01 7.27980971e-01 1.42399240e-02 3.34002316e-01
-9.01163742e-02 1.09005272e+00 -1.10998023e+00 -5.02944708e-01
-1.10290617e-01 3.22906226e-01 -4.59470630e-01 2.27487326e-01
-3.82430583e-01 2.62656182e-01 -3.83824348e-01 4.15760130e-01
5.38188398e-01 -1.18404174e+00 8.85125637e-01 -8.47368091e-02
7.17521846e-01 3.43043417e-01 -1.23903716e+00 1.67445290e+00
-2.94166028e-01 5.41229486e-01 2.08099499e-01 -1.20429146e+00
1.05503333e+00 6.69656768e-02 2.64689833e-01 -1.94227740e-01
-6.84729815e-02 -8.55566487e-02 2.18126714e-01 -2.37593025e-01
1.56494036e-01 9.18244421e-02 1.17397331e-01 5.47903955e-01
4.02928144e-01 2.33083576e-01 4.89867806e-01 6.44387007e-01
1.64451683e+00 -1.58807769e-01 4.09950614e-01 -1.99238673e-01
2.93321520e-01 -2.97819883e-01 5.71071148e-01 8.31832528e-01
8.71323422e-02 2.07554519e-01 9.73010600e-01 -6.11376584e-01
-8.96330535e-01 -8.83057714e-01 4.59387332e-01 1.01810277e+00
-6.38385564e-02 -8.67074907e-01 -7.99630165e-01 -1.06345189e+00
8.97644237e-02 4.63004440e-01 -6.18497133e-01 -2.40194917e-01
-5.71733952e-01 -3.94888431e-01 3.91234487e-01 5.83970368e-01
3.25459421e-01 -1.21890152e+00 -1.85379125e-02 2.38042831e-01
-1.54858872e-01 -1.64816809e+00 -3.85852367e-01 2.82049149e-01
-1.16982448e+00 -1.34897625e+00 -5.02215512e-02 -7.88906395e-01
1.24854290e+00 1.69262081e-01 1.47300160e+00 5.73014438e-01
1.26047269e-01 7.76453316e-02 -5.82968183e-02 2.78446376e-02
-5.64816117e-01 5.75698078e-01 -6.30364120e-02 -2.18169361e-01
4.84591769e-03 -1.14418769e+00 -1.33866623e-01 9.21858773e-02
-7.27288425e-01 2.99528003e-01 6.58535600e-01 7.99618602e-01
4.00509983e-01 4.68493015e-01 4.37475085e-01 -1.61641049e+00
6.12178385e-01 -3.57131541e-01 -8.19579184e-01 2.57876605e-01
-8.73445094e-01 4.27676231e-01 8.60194564e-01 -2.71340936e-01
-6.33469045e-01 2.31149480e-01 1.90535486e-01 -6.03174210e-01
-1.65769801e-01 1.08164668e+00 -1.55680344e-01 -9.65024717e-03
6.73152506e-01 8.58245865e-02 1.68742657e-01 -4.52395380e-01
2.37790048e-01 9.69732367e-03 4.30163145e-01 -5.08208454e-01
1.01217532e+00 2.28943527e-01 5.18156230e-01 -4.87748891e-01
-1.31140542e+00 -3.62427384e-01 -9.79337573e-01 1.37722328e-01
4.43921357e-01 -8.06105971e-01 -5.24018764e-01 2.74847716e-01
-1.13465190e+00 -7.04796433e-01 1.49892578e-02 1.71543166e-01
-4.19017255e-01 6.66135669e-01 -7.71541536e-01 -4.09385145e-01
-3.47013980e-01 -9.65997636e-01 8.34091723e-01 -9.76686403e-02
-2.47410908e-02 -1.38448834e+00 -4.29707691e-02 2.68000811e-01
3.19563866e-01 3.32906216e-01 9.56929862e-01 -6.98150575e-01
-9.16559339e-01 -1.00591280e-01 -4.62011665e-01 1.49256274e-01
4.54579920e-01 -1.92010462e-01 -5.12635589e-01 -5.92227876e-01
-3.20793808e-01 -3.21450323e-01 8.41832936e-01 3.89280856e-01
1.09485340e+00 -5.96805155e-01 -4.12843466e-01 8.23493719e-01
1.39772499e+00 -5.32205164e-01 4.21958387e-01 6.47530407e-02
1.32899249e+00 4.33957756e-01 1.33968458e-01 1.66276135e-02
5.72191536e-01 3.35508376e-01 8.21656704e-01 -1.21638693e-01
-7.82868266e-02 -6.74965620e-01 1.78187802e-01 1.02632260e+00
-1.09270930e-01 -5.12821138e-01 -1.24192333e+00 3.55625123e-01
-2.03725958e+00 -4.99120027e-01 -1.57578394e-01 1.83545887e+00
6.13836348e-01 4.13794607e-01 9.48743429e-03 -3.75792421e-02
8.55149209e-01 3.76501411e-01 -8.06735516e-01 7.46090785e-02
-3.18051353e-02 -2.48817176e-01 5.90700090e-01 5.93887269e-01
-8.41588676e-01 1.11230183e+00 6.89753532e+00 4.16707277e-01
-9.11461294e-01 -2.24832073e-01 3.49481821e-01 4.71606851e-01
-3.82524878e-01 4.09324557e-01 -7.35862315e-01 1.73715010e-01
1.10487902e+00 -2.27079064e-01 7.13022411e-01 8.56014371e-01
-2.79905856e-01 2.19709754e-01 -1.29285800e+00 8.37314129e-01
5.07170185e-02 -1.32026565e+00 1.69099614e-01 2.63288915e-01
6.36083245e-01 3.26730192e-01 -4.34329629e-01 3.00690234e-01
8.56540740e-01 -1.07046616e+00 1.06449507e-01 2.87784368e-01
7.43621469e-01 -5.22113681e-01 7.20799088e-01 7.62730718e-01
-1.25796092e+00 3.22844461e-02 -5.02556503e-01 -2.62485325e-01
3.30112837e-02 7.49090433e-01 -1.28822029e+00 7.38993526e-01
5.03251255e-01 1.07435095e+00 -9.06008124e-01 6.83279395e-01
-7.66001821e-01 8.31428111e-01 -4.85798508e-01 1.42373681e-01
2.09628627e-01 -3.58126551e-01 5.40202260e-01 8.60289752e-01
1.33686826e-01 -3.60303462e-01 5.30848980e-01 9.62524951e-01
-3.40806216e-01 -2.03233451e-01 -8.05979609e-01 -4.68986720e-01
4.22628641e-01 1.27053821e+00 -1.02602339e+00 -2.87928432e-01
-2.78405875e-01 9.13300872e-01 1.02259505e+00 5.67424715e-01
-5.68449616e-01 -1.45502612e-01 1.92862630e-01 2.53515452e-01
2.60695040e-01 -5.03327191e-01 9.22947377e-02 -1.15418494e+00
7.09949657e-02 -9.19393718e-01 8.04048657e-01 -7.28258908e-01
-1.46959305e+00 7.35221446e-01 -1.74813122e-02 -7.37739146e-01
-3.84794474e-01 -5.62227666e-01 -5.33316612e-01 6.28537476e-01
-1.33519793e+00 -1.27411926e+00 -4.61096853e-01 5.22170067e-01
4.23765145e-02 -1.45461470e-01 8.26243758e-01 -6.27339557e-02
-6.77326024e-01 3.94529223e-01 -2.38327950e-01 4.39440638e-01
4.63714600e-01 -1.58418286e+00 8.35341156e-01 8.52134764e-01
5.80432773e-01 6.71302736e-01 7.55420566e-01 -8.82943690e-01
-1.46388912e+00 -1.05024374e+00 7.90417314e-01 -2.19170675e-01
9.68283832e-01 -7.00383484e-01 -1.15024483e+00 1.29550421e+00
5.23299202e-02 5.13475776e-01 3.11634153e-01 7.11277425e-01
-5.74766636e-01 -9.36788470e-02 -8.13391328e-01 3.94596934e-01
1.57980978e+00 -6.71985865e-01 -1.98447570e-01 4.88623351e-01
9.82531846e-01 -6.18578494e-01 -8.56225729e-01 3.98309290e-01
5.90767246e-03 -7.85101712e-01 9.01068091e-01 -6.77986026e-01
1.88403398e-01 -2.95081556e-01 1.76448494e-01 -1.46639049e+00
-4.62201476e-01 -6.19003057e-01 -4.67479050e-01 9.73078609e-01
4.76758420e-01 -9.06201601e-01 1.27526343e+00 4.05460715e-01
2.79188645e-03 -5.80493331e-01 -6.45622969e-01 -8.85534942e-01
-3.82298350e-01 -1.00092843e-01 7.89018273e-01 1.25761724e+00
-2.64587939e-01 9.47522461e-01 -3.82846355e-01 4.70664531e-01
8.82074356e-01 3.72051269e-01 9.83552694e-01 -1.56013179e+00
-5.17349362e-01 -1.13601223e-01 -6.00512385e-01 -1.02097607e+00
6.98354602e-01 -1.24462903e+00 -1.56115979e-01 -1.99553812e+00
3.27485621e-01 -3.55266839e-01 -4.06759858e-01 8.75952721e-01
-2.72773087e-01 -3.20873499e-01 -6.52453005e-02 1.42254587e-02
-8.68455708e-01 4.33310330e-01 1.30229402e+00 -2.60802180e-01
-4.95236404e-02 7.25544095e-02 -7.27020323e-01 7.31968522e-01
7.52370298e-01 -6.24643981e-01 -8.55219603e-01 -4.74440783e-01
6.25270367e-01 1.80177540e-01 4.09317225e-01 -8.54287684e-01
5.28708100e-01 -1.23292156e-01 7.66476244e-02 -5.47303796e-01
1.19699895e-01 -8.85880768e-01 3.66110295e-01 3.44673544e-01
-2.23121375e-01 8.26514140e-02 -3.17683839e-03 9.85726476e-01
-3.04758519e-01 -1.25842780e-01 3.49495173e-01 -2.09816024e-01
-6.34158790e-01 6.51646376e-01 2.12271020e-01 1.35821313e-01
5.47124028e-01 -2.37105265e-01 -4.92490351e-01 -6.86942399e-01
-8.20816755e-01 3.84229273e-01 5.13095498e-01 3.49085182e-01
4.90167469e-01 -1.24374497e+00 -5.81318378e-01 1.80563197e-01
1.22214872e-02 5.11880398e-01 4.59045991e-02 5.82311809e-01
-3.62834394e-01 1.50383204e-01 -1.55713424e-01 -6.49692297e-01
-1.38288283e+00 7.35542357e-01 3.86638612e-01 -7.88540304e-01
-9.06084776e-01 8.25118780e-01 1.70287400e-01 -7.60843635e-01
2.36066177e-01 -1.69352427e-01 -7.07614869e-02 -3.64198148e-01
9.45406184e-02 3.35282907e-02 -6.85444996e-02 -5.21905541e-01
-1.40823588e-01 1.56696543e-01 -4.63539697e-02 5.45840740e-01
1.56459868e+00 -8.61018822e-02 -5.04384100e-01 4.50472951e-01
9.52878416e-01 -3.33243579e-01 -1.33739221e+00 -6.99491501e-01
3.30958962e-01 -2.00879589e-01 -1.16641186e-01 -3.64929289e-01
-1.36273611e+00 7.27117419e-01 -2.68067360e-01 4.69526321e-01
9.46181476e-01 2.26555511e-01 6.27172291e-01 9.77530479e-01
5.27005792e-01 -7.95092523e-01 2.99711287e-01 7.21555948e-01
7.28155375e-01 -1.23177385e+00 1.38735980e-01 -9.18414652e-01
-2.60264575e-01 1.07522666e+00 9.20478582e-01 -2.06713885e-01
8.40721905e-01 2.65745699e-01 -2.79317230e-01 -6.44083798e-01
-9.38662291e-01 -4.20910306e-02 4.41932678e-01 6.42226756e-01
2.31658891e-01 2.00832143e-01 4.98681903e-01 2.06120417e-01
-3.62502664e-01 -3.33487511e-01 4.81350601e-01 5.36002040e-01
-2.52675116e-01 -1.22847748e+00 1.33552894e-01 7.65380859e-01
-3.47942719e-03 -9.98326018e-02 -6.63347363e-01 7.12542057e-01
-3.63405198e-01 6.98511958e-01 -1.09962396e-01 -3.22950572e-01
6.21988922e-02 -1.22238852e-01 5.26822329e-01 -1.00299025e+00
-2.93529063e-01 -2.25992709e-01 4.49287951e-01 -6.62535548e-01
-2.99636394e-01 -1.53207511e-01 -1.25147402e+00 -4.01370078e-01
-4.14265692e-01 1.06389806e-01 4.38210160e-01 8.87485981e-01
4.08457786e-01 6.00820005e-01 1.45222887e-01 -5.78561306e-01
-3.68531227e-01 -9.48220789e-01 -8.28586698e-01 5.03620088e-01
2.92236120e-01 -5.40947974e-01 -5.35389364e-01 -2.31226072e-01] | [7.090664386749268, 6.20127534866333] |
2f1517b0-ae16-47a9-bf7a-8f6a19c5145f | clearumor-at-semeval-2019-task-7-convolving | 1904.03084 | null | http://arxiv.org/abs/1904.03084v1 | http://arxiv.org/pdf/1904.03084v1.pdf | CLEARumor at SemEval-2019 Task 7: ConvoLving ELMo Against Rumors | This paper describes our submission to SemEval-2019 Task 7: RumourEval:
Determining Rumor Veracity and Support for Rumors. We participated in both
subtasks. The goal of subtask A is to classify the type of interaction between
a rumorous social media post and a reply post as support, query, deny, or
comment. The goal of subtask B is to predict the veracity of a given rumor. For
subtask A, we implement a CNN-based neural architecture using ELMo embeddings
of post text combined with auxiliary features and achieve a F1-score of 44.6%.
For subtask B, we employ a MLP neural network leveraging our estimates for
subtask A and achieve a F1-score of 30.1% (second place in the competition). We
provide results and analysis of our system performance and present ablation
experiments. | ['Lukas Schmelzeisen', 'Steffen Staab', 'Ipek Baris'] | 2019-04-05 | clearumor-at-semeval-2019-task-7-convolving-1 | https://aclanthology.org/S19-2193 | https://aclanthology.org/S19-2193.pdf | semeval-2019-6 | ['rumour-detection'] | ['natural-language-processing'] | [-3.85425836e-01 5.07472873e-01 -2.14994624e-01 -4.08591002e-01
-3.72321218e-01 -1.22972809e-01 9.86147344e-01 3.72778177e-01
-5.46868205e-01 6.52027965e-01 7.96158016e-01 -3.65015090e-01
5.92766881e-01 -3.87398601e-01 -5.85014403e-01 -1.65811703e-02
1.58342347e-02 4.96121109e-01 5.32254949e-02 -4.78150755e-01
7.85704076e-01 1.76921755e-01 -9.72996354e-01 9.43194091e-01
1.92315072e-01 1.10206366e+00 -5.29907107e-01 8.18458438e-01
7.79273659e-02 1.78580546e+00 -8.68038595e-01 -4.81072754e-01
-2.93490857e-01 -2.98915774e-01 -1.26059866e+00 -1.20586239e-01
4.40017045e-01 -7.11772799e-01 -7.13945210e-01 8.03211629e-01
3.96132678e-01 1.35845080e-01 7.75968432e-01 -1.18155479e+00
-1.13625169e+00 1.14786422e+00 -3.77201259e-01 6.08467221e-01
4.30737942e-01 8.60364810e-02 1.19176781e+00 -1.29274774e+00
4.78167295e-01 1.07193244e+00 7.17493117e-01 5.27970552e-01
-1.09241462e+00 -5.86326599e-01 -1.89091414e-01 5.71297169e-01
-1.01187372e+00 -9.52642798e-01 6.05483532e-01 -5.34910917e-01
1.24320006e+00 1.35544255e-01 2.75617242e-01 1.64903986e+00
2.91829884e-01 1.06891537e+00 1.03290451e+00 1.93994448e-01
1.58819258e-01 4.15550947e-01 5.92558801e-01 6.47364497e-01
-4.38873246e-02 -5.45344725e-02 -7.37022698e-01 -4.83220577e-01
1.76317498e-01 -6.76328614e-02 -3.87630284e-01 4.78569239e-01
-9.24476683e-01 1.26161289e+00 8.75687540e-01 2.26716772e-01
-8.02552581e-01 -6.24524094e-02 5.62834263e-01 7.73289263e-01
7.64100611e-01 8.38915288e-01 5.04255332e-02 -8.50325227e-02
-1.16263425e+00 6.56404495e-01 1.54236495e+00 4.93240178e-01
1.96696252e-01 1.17568471e-01 -3.90891999e-01 1.11639917e+00
-1.82551183e-02 -6.60221875e-02 6.39237285e-01 -7.03177512e-01
2.60139257e-01 1.77040845e-01 4.95081097e-01 -1.10571289e+00
-8.61957550e-01 -6.08578444e-01 -6.83966279e-01 1.35892093e-01
1.86238736e-01 -2.61046857e-01 -5.49730241e-01 1.20808315e+00
-3.06578994e-01 5.87389171e-02 1.41313016e-01 1.12093520e+00
1.18914342e+00 5.25304854e-01 -2.33102947e-01 -1.76929682e-01
1.16678190e+00 -1.50923133e+00 -5.63896537e-01 -3.13423395e-01
5.43694556e-01 -9.08946693e-01 6.73379660e-01 3.20730746e-01
-1.18492842e+00 -6.14090264e-02 -1.18624830e+00 -3.29468191e-01
-3.26001421e-02 -1.17744006e-01 1.56264201e-01 3.43565978e-02
-1.02847362e+00 6.91059530e-01 -5.32780111e-01 -2.12124079e-01
4.00222272e-01 -2.14208618e-01 -1.07291758e-01 1.07029065e-01
-1.41345143e+00 1.10919118e+00 4.58044671e-02 -8.93063396e-02
-1.26779926e+00 -5.14847279e-01 -5.98628819e-01 4.53920782e-01
7.33805299e-02 -3.62872630e-01 1.66832507e+00 -8.12811017e-01
-1.37286448e+00 9.94122207e-01 4.28526215e-02 -9.53633964e-01
8.65369439e-01 -2.61553377e-01 -6.08012617e-01 -1.27840817e-01
1.08958207e-01 7.09439889e-02 9.72283304e-01 -8.76768172e-01
-4.64023024e-01 -1.92118034e-01 -6.14306442e-02 5.24222627e-02
-1.68804631e-01 3.21731776e-01 1.92591250e-01 -3.67885113e-01
-5.74499555e-02 -7.16098487e-01 2.63108999e-01 -5.10955155e-01
-7.64983952e-01 -5.12591481e-01 6.17172062e-01 -8.14388394e-01
1.10217547e+00 -2.02628541e+00 -2.01866567e-01 -2.55201995e-01
8.67634952e-01 1.73388124e-01 -7.94438720e-02 5.39539754e-01
-1.01339472e-02 3.01745921e-01 1.46912262e-01 -7.15511978e-01
1.13626070e-01 -4.51602042e-01 -7.32608974e-01 5.83333313e-01
1.11210868e-01 7.46708453e-01 -7.10698724e-01 1.06314324e-01
-4.04270262e-01 2.63511449e-01 -6.79358125e-01 6.02926433e-01
-1.66296810e-01 -4.63546312e-04 -2.93590367e-01 3.96500230e-01
5.48426151e-01 -5.57069540e-01 -2.54295737e-01 1.76857226e-02
-5.57641089e-02 1.05419683e+00 -2.76597798e-01 7.09158838e-01
-4.72463876e-01 9.12693560e-01 2.64430851e-01 -5.39601505e-01
9.61993098e-01 4.72331226e-01 -5.32803917e-03 -4.60087806e-01
3.78994465e-01 2.91153967e-01 -2.15402067e-01 -5.82763493e-01
9.80665088e-01 -3.63397390e-01 -1.40232906e-01 1.07100677e+00
-1.99160159e-01 8.96565914e-02 -4.40805584e-01 6.29661500e-01
1.51138687e+00 -6.78422689e-01 3.27917993e-01 -4.63047177e-02
3.36661249e-01 -1.35907367e-01 3.60254422e-02 1.12236261e+00
-5.22051454e-01 6.57248020e-01 9.71249342e-01 -7.43864894e-01
-9.47688878e-01 -6.11230731e-01 1.45186737e-01 1.28411913e+00
-2.50368327e-01 -1.36635035e-01 -2.34729081e-01 -7.42026389e-01
3.14562768e-01 1.33541465e+00 -1.02015150e+00 -2.39551105e-02
-3.34785223e-01 -5.63295186e-01 7.07811534e-01 2.25868315e-01
4.39774305e-01 -1.38526571e+00 -2.93774515e-01 3.18285882e-01
-4.02598023e-01 -1.24745142e+00 -5.23468077e-01 1.32312790e-01
-4.92930710e-01 -1.00478721e+00 -5.14469147e-01 -4.73667502e-01
-7.62220323e-02 1.40183449e-01 1.25518346e+00 5.95497727e-01
3.92305940e-01 -2.77086258e-01 -3.14130574e-01 -8.95257741e-02
-6.21945620e-01 1.30863965e-01 2.00926155e-01 4.14703712e-02
4.30213809e-01 -6.13291025e-01 -4.66484308e-01 2.98949242e-01
-5.73701262e-01 -8.27385560e-02 1.75838426e-01 9.72470164e-01
-2.37807378e-01 -5.21438956e-01 8.74923110e-01 -1.13316417e+00
1.28815663e+00 -1.17970073e+00 -2.08196323e-02 -3.54656696e-01
-7.59763181e-01 -2.35657990e-01 6.94997549e-01 -1.35231122e-01
-6.93045974e-01 -7.44646251e-01 -3.51751894e-01 -3.56708974e-01
7.56921098e-02 7.09367931e-01 6.85430169e-01 4.60095584e-01
1.23866272e+00 1.85891151e-01 3.90750729e-02 -5.90873718e-01
1.59463465e-01 1.33280635e+00 7.18540728e-01 -9.50498506e-03
2.34769851e-01 1.98041484e-01 -7.27554262e-01 -7.61100590e-01
-1.52108717e+00 -6.22244596e-01 7.57763982e-02 1.00333206e-01
6.26036704e-01 -8.76642585e-01 -9.87484276e-01 4.86262411e-01
-1.48990285e+00 -3.73806566e-01 2.49165729e-01 3.86195332e-01
-4.19776410e-01 1.71947703e-02 -1.21994650e+00 -6.96891248e-01
-6.38881087e-01 -6.68806791e-01 2.15293035e-01 -1.93545192e-01
-5.53019881e-01 -7.27290928e-01 6.28701746e-02 9.04344261e-01
8.01695764e-01 -4.86010425e-02 5.10221303e-01 -1.60170722e+00
-8.34631771e-02 -5.13733387e-01 -5.52688181e-01 3.44308704e-01
-2.98082709e-01 -4.27277833e-01 -1.20834506e+00 -4.32556808e-01
1.94568634e-01 -1.07442307e+00 1.43625927e+00 -3.54916714e-02
1.02605283e+00 -9.85442936e-01 1.23384446e-01 2.78105527e-01
8.18034291e-01 -7.14095056e-01 4.13080096e-01 3.77302408e-01
4.15463477e-01 4.62001324e-01 1.29008358e-02 7.12238729e-01
5.62857568e-01 5.78104198e-01 7.00482130e-01 5.70033491e-01
-7.21541569e-02 -4.44090366e-01 6.20528162e-01 6.22085452e-01
2.96731591e-01 -5.13215899e-01 -7.73919046e-01 6.97658837e-01
-1.82047355e+00 -1.19011354e+00 -3.24679554e-01 1.73373330e+00
7.36575246e-01 3.82311046e-01 2.21371785e-01 -2.31006950e-01
5.39571524e-01 6.77915931e-01 -3.86692554e-01 -9.59342778e-01
1.47210747e-01 -4.97297943e-01 2.43718714e-01 1.01575458e+00
-9.49921489e-01 9.02642608e-01 6.48432016e+00 1.37468904e-01
-1.21334291e+00 3.81171137e-01 4.14530367e-01 -5.06885648e-01
-2.36769497e-01 -3.53660285e-01 -6.15309358e-01 4.93033648e-01
1.32761288e+00 -2.01603621e-01 6.82562351e-01 7.68252432e-01
3.50600958e-01 7.39749447e-02 -1.20685220e+00 6.20686352e-01
5.83689272e-01 -1.51435959e+00 -2.95282513e-01 -1.20907925e-01
6.60172403e-01 1.05484331e+00 6.87091053e-02 8.58929753e-01
5.92719436e-01 -1.42061388e+00 8.86041462e-01 4.34608877e-01
1.68993309e-01 -5.47407210e-01 1.07452071e+00 8.35061014e-01
1.61784768e-01 -9.57411751e-02 -4.11454737e-01 -4.66837555e-01
3.29766691e-01 7.31807828e-01 -1.54626226e+00 -1.71289399e-01
5.31944335e-01 7.49760866e-01 -3.87201905e-01 8.27567160e-01
-7.43543863e-01 1.04711866e+00 -9.85416304e-03 -3.34754020e-01
2.95277148e-01 6.86230659e-01 9.69002366e-01 1.37396359e+00
-2.22129077e-01 -6.50559589e-02 1.28637582e-01 1.12120759e+00
-9.48778689e-01 -1.11940287e-01 -2.08466575e-01 -4.41030651e-01
5.08970916e-01 1.04059219e+00 1.96246698e-01 -5.38066745e-01
-7.94102699e-02 1.39605343e+00 8.40665817e-01 2.45893627e-01
-6.99378669e-01 -2.69638509e-01 4.31357414e-01 3.03764343e-01
2.04690129e-01 1.51093155e-01 -4.51928020e-01 -1.40828168e+00
-3.55394870e-01 -8.51686418e-01 4.19474214e-01 -9.60490346e-01
-1.56034589e+00 8.57299626e-01 -6.04559541e-01 -4.12208319e-01
-3.94236594e-01 -2.61375576e-01 -8.61385822e-01 1.13114667e+00
-1.69849133e+00 -9.07521069e-01 -3.43566179e-01 2.94597358e-01
4.93623585e-01 -4.05597955e-01 8.30382466e-01 1.09041415e-01
-4.61865187e-01 5.58161557e-01 -8.10203031e-02 2.76556313e-01
8.78302157e-01 -1.18650603e+00 6.38483346e-01 3.00329626e-01
-2.35744283e-01 5.01007140e-01 1.23797810e+00 -4.68489647e-01
-7.75438845e-01 -1.17750037e+00 1.63591325e+00 -6.81325257e-01
1.11010039e+00 3.04370956e-03 -1.26509106e+00 1.02241504e+00
2.23805413e-01 4.54528406e-02 4.21395153e-01 6.98582232e-01
-8.71039033e-01 5.95044434e-01 -1.24885535e+00 4.02108788e-01
3.68622780e-01 -8.89679432e-01 -9.72873807e-01 4.67263132e-01
6.13963485e-01 -4.12843972e-01 -6.88015521e-01 -2.10708112e-01
5.52402020e-01 -1.22278321e+00 4.17423815e-01 -1.08304381e+00
1.34869504e+00 2.46848375e-01 -7.66024590e-02 -1.50745559e+00
-5.67115247e-01 -5.38347065e-01 -4.02606279e-01 4.73868102e-01
5.78812540e-01 -5.71685374e-01 6.75073028e-01 2.97431558e-01
-1.15321010e-01 -6.10869527e-01 -6.22392952e-01 -3.98529053e-01
3.58511925e-01 -3.00236851e-01 2.78599590e-01 1.28954172e+00
4.87579316e-01 8.85745347e-01 -8.24805856e-01 5.14281839e-02
4.73777562e-01 2.13512719e-01 7.08732069e-01 -1.03534317e+00
-2.93264657e-01 -6.92599237e-01 1.05963722e-02 -1.19467819e+00
4.18627471e-01 -1.19858384e+00 8.67128968e-02 -1.47124040e+00
4.68282402e-01 4.76868898e-02 -5.22044182e-01 6.74536824e-01
9.60471779e-02 5.09421289e-01 7.56327808e-02 6.92335308e-01
-6.32808149e-01 6.58083975e-01 7.99842358e-01 -2.99833626e-01
-1.72411785e-01 2.23019540e-01 -9.58072186e-01 5.78355849e-01
8.53979230e-01 -6.88932180e-01 3.06940675e-01 -2.31185704e-01
5.12291491e-01 4.81177419e-01 7.29709327e-01 -4.70487654e-01
2.99605370e-01 2.64261998e-02 4.05900627e-02 -4.73364562e-01
3.63607764e-01 -1.11667000e-01 -6.38291597e-01 3.07024270e-01
-9.92568195e-01 -1.20172031e-01 -1.42687663e-01 6.59189939e-01
-1.34828597e-01 -2.08555713e-01 9.15918887e-01 7.47029856e-02
-2.71223992e-01 -6.19609617e-02 -6.46435618e-01 2.47188315e-01
2.21870691e-01 4.63754445e-01 -6.93525493e-01 -1.06909430e+00
-9.95108843e-01 2.88275063e-01 3.20192605e-01 6.36038721e-01
7.85581827e-01 -9.57069099e-01 -1.49240875e+00 1.61639880e-02
2.66515791e-01 -7.23212600e-01 6.84365779e-02 1.15537930e+00
-2.62665212e-01 3.32863927e-01 -2.22887322e-01 -5.06224763e-03
-1.09742558e+00 2.34057143e-01 4.84101266e-01 -1.95680663e-01
-6.83873415e-01 1.05860281e+00 -3.09121042e-01 -3.41684610e-01
8.47670138e-02 2.97190875e-01 -4.12558466e-01 -6.29475415e-02
1.01786625e+00 5.06518483e-01 3.20545673e-01 -6.83310688e-01
-2.11094841e-01 -8.71219993e-01 -7.53050566e-01 -1.37748703e-01
1.48559690e+00 -4.71903943e-02 -2.48654723e-01 5.51122367e-01
1.46634483e+00 1.43552020e-01 -6.92156792e-01 -6.04958892e-01
-4.60876860e-02 -2.88816661e-01 6.39456391e-01 -1.26879084e+00
-7.87065625e-01 7.15026200e-01 -7.98994824e-02 6.04273200e-01
1.51909962e-01 6.62806779e-02 8.91739666e-01 5.73154986e-01
-1.09465659e-01 -8.29280019e-01 2.94176251e-01 1.19709933e+00
1.61518002e+00 -1.55875218e+00 1.28970221e-01 2.09670931e-01
-9.13639247e-01 9.80509877e-01 4.01506364e-01 -3.84757817e-01
5.31949461e-01 -2.58790821e-01 1.82641953e-01 -7.20019996e-01
-1.25689876e+00 5.33315003e-01 2.17209324e-01 -9.47836787e-02
6.95426941e-01 2.64208823e-01 -3.79889697e-01 9.06901121e-01
-5.47332287e-01 1.49174377e-01 1.22967124e+00 4.25140560e-01
-6.76530004e-01 -5.56930639e-02 -5.23587018e-02 6.04075134e-01
-8.04918826e-01 -2.33744532e-01 -7.38816559e-01 1.95835859e-01
-6.17748678e-01 1.29658055e+00 3.59814763e-02 -7.76346326e-01
1.01621568e-01 1.33282512e-01 -7.85882473e-02 -6.77345395e-01
-9.98501062e-01 -5.30682504e-01 7.48555183e-01 -5.31663418e-01
2.88721055e-01 -4.82027680e-01 -9.57007647e-01 -1.09190059e+00
-1.62125617e-01 2.76020646e-01 6.39918506e-01 8.94934475e-01
1.77290738e-01 2.88228273e-01 8.00570726e-01 -4.58621413e-01
-1.34916973e+00 -1.59061491e+00 -6.15777671e-01 3.96772444e-01
1.00387979e+00 -2.53787279e-01 -8.06097329e-01 -4.76894110e-01] | [8.216269493103027, 10.152532577514648] |
780ff761-615c-40cc-9113-ec1f32b4e0da | melody-transcription-via-generative-pre | 2212.01884 | null | https://arxiv.org/abs/2212.01884v1 | https://arxiv.org/pdf/2212.01884v1.pdf | Melody transcription via generative pre-training | Despite the central role that melody plays in music perception, it remains an open challenge in music information retrieval to reliably detect the notes of the melody present in an arbitrary music recording. A key challenge in melody transcription is building methods which can handle broad audio containing any number of instrument ensembles and musical styles - existing strategies work well for some melody instruments or styles but not all. To confront this challenge, we leverage representations from Jukebox (Dhariwal et al. 2020), a generative model of broad music audio, thereby improving performance on melody transcription by $20$% relative to conventional spectrogram features. Another obstacle in melody transcription is a lack of training data - we derive a new dataset containing $50$ hours of melody transcriptions from crowdsourced annotations of broad music. The combination of generative pre-training and a new dataset for this task results in $77$% stronger performance on melody transcription relative to the strongest available baseline. By pairing our new melody transcription approach with solutions for beat detection, key estimation, and chord recognition, we build Sheet Sage, a system capable of transcribing human-readable lead sheets directly from music audio. Audio examples can be found at https://chrisdonahue.com/sheetsage and code at https://github.com/chrisdonahue/sheetsage . | ['Percy Liang', 'John Thickstun', 'Chris Donahue'] | 2022-12-04 | null | null | null | null | ['chord-recognition', 'music-information-retrieval'] | ['audio', 'music'] | [ 2.36427501e-01 -2.97133297e-01 -9.53321997e-03 1.66209936e-01
-1.60161531e+00 -1.24850523e+00 2.32089475e-01 -6.91817626e-02
-1.46649957e-01 4.98638153e-01 6.89200521e-01 8.78905356e-02
-2.89746404e-01 -5.14043450e-01 -4.30073887e-01 -2.70174176e-01
4.37433943e-02 4.42985445e-01 -5.24862930e-02 -5.76504827e-01
2.81393409e-01 8.05601627e-02 -1.49649024e+00 6.66968584e-01
7.64093399e-02 1.06478679e+00 1.12187807e-02 1.21742070e+00
2.68505096e-01 7.36186326e-01 -8.51383805e-01 -2.65234917e-01
4.41605419e-01 -7.48942316e-01 -9.35057521e-01 -1.16598554e-01
5.01675248e-01 -2.09512144e-01 -4.53163609e-02 7.42560327e-01
9.89088297e-01 7.90586770e-02 4.61642683e-01 -8.74016583e-01
-2.78840512e-01 1.17615509e+00 -1.64288685e-01 2.05657214e-01
8.41895282e-01 1.48522019e-01 1.64506423e+00 -1.03977334e+00
4.29546118e-01 8.19735467e-01 1.02439117e+00 3.28411371e-01
-1.30831707e+00 -1.00819767e+00 -5.24492204e-01 -4.49404866e-02
-1.62839127e+00 -8.22555780e-01 7.77719080e-01 -5.53587854e-01
7.69514978e-01 6.40307844e-01 8.93403113e-01 1.04174590e+00
-4.28379804e-01 8.29651356e-01 7.98874617e-01 -4.94268209e-01
-1.43655604e-02 -4.31105822e-01 -1.67399928e-01 1.92773551e-01
-4.70210433e-01 -9.06624645e-03 -1.02167821e+00 -3.75543356e-01
9.71431255e-01 -4.45231169e-01 -3.91286612e-01 6.46689162e-02
-1.43914139e+00 5.11433661e-01 3.05715129e-02 2.40237072e-01
-1.88705072e-01 3.02515417e-01 7.25680530e-01 5.33506632e-01
-2.48969179e-02 1.03822327e+00 -3.04928511e-01 -8.54437411e-01
-1.38932669e+00 8.05554271e-01 9.27173734e-01 7.04863310e-01
2.62888551e-01 2.32299313e-01 -6.66250214e-02 1.10586751e+00
-2.00674772e-01 3.09960932e-01 4.77904618e-01 -1.13241494e+00
4.55242902e-01 2.16795340e-01 9.95759889e-02 -6.76552832e-01
-1.88668221e-01 -6.69830739e-01 -5.13083458e-01 1.27639279e-01
6.52404070e-01 6.01795129e-02 -3.18110734e-01 1.65022624e+00
-5.78017859e-03 7.76413307e-02 -3.09825689e-01 9.68817532e-01
7.32469440e-01 5.68955481e-01 -4.98742998e-01 -9.81346071e-02
1.42798841e+00 -5.92530251e-01 -3.56217653e-01 1.98124751e-01
1.28909305e-01 -1.36371958e+00 1.41738605e+00 8.47152770e-01
-1.43090594e+00 -7.42328465e-01 -1.04743958e+00 -3.46227437e-01
2.24419683e-01 1.60142481e-01 4.03272837e-01 4.39704180e-01
-7.18287289e-01 6.90375328e-01 -5.40676177e-01 1.04982086e-01
2.86345422e-01 3.08218867e-01 -1.03415839e-01 3.88789624e-01
-9.46038008e-01 2.87499964e-01 2.83284903e-01 -1.49105310e-01
-1.09344709e+00 -9.53301430e-01 -5.88401914e-01 -3.72098386e-02
3.29621106e-01 -4.03857887e-01 1.95181930e+00 -8.70676160e-01
-1.54646564e+00 7.60475755e-01 1.62397265e-01 -4.18080270e-01
4.77256954e-01 -4.51239944e-01 -4.72250134e-01 8.76548141e-02
1.16420977e-01 4.53292698e-01 8.68985713e-01 -7.89061368e-01
-5.95214605e-01 8.10675994e-02 -7.03252153e-03 2.69524902e-01
-2.01405734e-01 1.80039689e-01 -4.39066499e-01 -1.26132035e+00
-8.66877511e-02 -1.17154324e+00 1.60133168e-01 -5.42553306e-01
-6.67380810e-01 -4.30362560e-02 5.01323581e-01 -8.28260005e-01
1.58210242e+00 -2.24633908e+00 1.46399081e-01 1.72575146e-01
4.47334535e-02 7.96715356e-03 -2.51118600e-01 7.80840397e-01
-5.20865656e-02 -9.24754366e-02 -1.54466689e-01 -2.17473388e-01
3.02718043e-01 -2.36400425e-01 -8.92690241e-01 4.67272364e-02
-2.31386304e-01 9.96083498e-01 -7.46010661e-01 -1.18922517e-01
-7.75567144e-02 3.54205757e-01 -7.45572865e-01 1.40153006e-01
-3.37604791e-01 6.32662714e-01 3.30099352e-02 9.21225369e-01
7.24561047e-03 2.91660223e-02 1.87647596e-01 -1.25392452e-01
-1.22878812e-01 9.42852020e-01 -1.53961098e+00 2.25938296e+00
-2.83095628e-01 5.66170752e-01 9.14631113e-02 -3.61550093e-01
8.65551651e-01 7.74911344e-01 5.94622850e-01 -2.78928190e-01
2.70078313e-02 5.65854967e-01 2.29840070e-01 -3.40108387e-02
8.93789351e-01 -3.20554227e-01 -5.85474312e-01 7.69417584e-01
9.75760221e-02 -7.12859035e-01 2.66024739e-01 -1.05975755e-01
1.33149350e+00 1.76506847e-01 3.82112801e-01 -3.21483910e-02
6.52612597e-02 1.48837239e-01 3.44196647e-01 6.97559714e-01
4.46232110e-02 1.03906667e+00 1.41873777e-01 -3.44599396e-01
-1.16294515e+00 -1.29774535e+00 1.97196137e-02 1.59019852e+00
-4.74109083e-01 -1.07136869e+00 -6.72558546e-01 -3.41809131e-02
-1.85923010e-01 1.68962240e-01 -2.15445772e-01 2.01050222e-01
-6.95990860e-01 -3.38424981e-01 1.29233015e+00 5.29856324e-01
8.73998106e-02 -1.39794600e+00 -4.93750304e-01 4.36776459e-01
-4.86539751e-01 -7.15375483e-01 -9.40157831e-01 2.52213389e-01
-6.34507120e-01 -1.02228105e+00 -7.36711860e-01 -7.26509988e-01
-3.30433995e-01 -1.78791165e-01 1.51581430e+00 -1.49248213e-01
-6.00355744e-01 2.09407672e-01 -4.42721903e-01 -5.61423719e-01
-5.89600146e-01 4.71025437e-01 1.74839795e-01 -3.50972950e-01
8.45224932e-02 -9.55213785e-01 -6.76928520e-01 3.70502979e-01
-7.18619108e-01 5.20393737e-02 2.69781619e-01 6.36329234e-01
6.28771544e-01 -1.85564071e-01 8.29471648e-01 -5.89773774e-01
8.39733243e-01 -2.81390473e-02 -3.00641507e-01 -3.65152389e-01
-2.56983876e-01 -4.51041281e-01 4.89958316e-01 -5.01499951e-01
-2.34337583e-01 1.40776098e-01 -5.37925303e-01 -3.36392194e-01
-1.32818103e-01 4.51944381e-01 1.01790711e-01 4.19894248e-01
1.00259781e+00 6.41139895e-02 -2.53331095e-01 -8.36041868e-01
4.43714976e-01 8.56619596e-01 1.16303957e+00 -7.52117872e-01
7.28320658e-01 1.23997755e-01 -3.39637905e-01 -6.98302627e-01
-8.69290948e-01 -6.80611372e-01 -4.98468399e-01 -5.68647161e-02
4.82734829e-01 -1.08741653e+00 -8.44479859e-01 2.35371992e-01
-6.46606326e-01 -3.41180742e-01 -7.77890146e-01 4.07707214e-01
-9.70371664e-01 1.81017250e-01 -9.61706042e-01 -7.23910391e-01
-6.09578729e-01 -7.84604073e-01 1.12045109e+00 -2.94399530e-01
-1.21512520e+00 -4.55512583e-01 4.02935177e-01 4.95662570e-01
2.73064643e-01 1.47413701e-01 6.97619736e-01 -4.70964640e-01
-4.72345024e-01 -2.88577497e-01 3.54294837e-01 3.32172126e-01
2.06669316e-01 -1.99880391e-01 -1.24441588e+00 -2.33523101e-01
-1.64858997e-01 -7.43957698e-01 7.58149743e-01 1.80318072e-01
8.42022240e-01 -2.43941233e-01 3.49013895e-01 5.99351883e-01
9.02348340e-01 1.71567518e-02 5.29215753e-01 2.55144745e-01
5.96577942e-01 1.66456521e-01 5.24002790e-01 6.88412607e-01
6.26648515e-02 9.80473518e-01 1.29946964e-02 2.81952024e-01
-3.55187237e-01 -6.26968622e-01 6.08285487e-01 1.36966324e+00
-3.75239491e-01 -1.62106263e-03 -8.79905045e-01 7.81997502e-01
-1.73401284e+00 -1.29838836e+00 1.48926094e-01 2.18130255e+00
1.36097217e+00 1.40447691e-01 7.08436549e-01 8.49336684e-01
3.35167974e-01 4.48221825e-02 -3.05281937e-01 -2.55849063e-01
-1.59574926e-01 8.29138100e-01 3.03706545e-02 3.60920340e-01
-1.19296277e+00 8.08553934e-01 6.66589403e+00 1.16905916e+00
-1.09314299e+00 1.37398034e-01 1.30689144e-03 -8.73857558e-01
-1.92854226e-01 -3.63367759e-02 -6.07994199e-01 2.89778262e-01
9.37944293e-01 -2.41187941e-02 8.97183239e-01 5.52459896e-01
2.78373778e-01 3.50531161e-01 -1.35561049e+00 1.43586862e+00
5.12192473e-02 -1.52390718e+00 -8.49155113e-02 7.43712038e-02
5.72971165e-01 1.03750855e-01 4.82731126e-02 4.45069075e-01
2.03794315e-01 -1.29714775e+00 1.17205584e+00 3.50405753e-01
1.21813226e+00 -7.20852673e-01 2.97003657e-01 3.72252017e-01
-1.47713006e+00 -3.08302995e-02 9.92991552e-02 -4.65757132e-01
2.04685926e-01 2.00829700e-01 -1.14162719e+00 4.05357927e-01
5.64941049e-01 7.16025651e-01 -4.64530706e-01 1.05536258e+00
-1.69137180e-01 1.07110238e+00 -3.83383363e-01 3.86550784e-01
-2.09603757e-01 1.14182033e-01 7.85791576e-01 1.29331124e+00
6.45059884e-01 -1.11580119e-01 3.94208342e-01 6.64353549e-01
-2.86100894e-01 2.41044819e-01 -4.11229402e-01 -3.49020213e-01
6.13001406e-01 1.30729973e+00 -5.20137250e-01 -1.39100686e-01
1.07091732e-01 9.02568996e-01 2.02461760e-02 7.99347833e-02
-6.16351664e-01 -3.55263472e-01 6.44546270e-01 3.03326875e-01
2.60274708e-01 -2.68713534e-01 -3.23389977e-01 -9.53035712e-01
-3.43185365e-02 -1.58190525e+00 5.97121596e-01 -8.92365634e-01
-1.19059896e+00 6.75813794e-01 -4.28872257e-01 -1.60333097e+00
-8.43120158e-01 -2.79343069e-01 -3.54823261e-01 9.09127474e-01
-6.49625719e-01 -1.18431222e+00 -7.41399527e-02 6.63682878e-01
8.29783559e-01 -2.81143636e-01 1.31889093e+00 4.63254839e-01
1.54752582e-01 5.67186832e-01 -1.00841627e-01 4.12010014e-01
9.80353773e-01 -1.40430748e+00 6.11114204e-01 3.03715020e-01
1.21109688e+00 5.70837677e-01 7.15123177e-01 -3.69359374e-01
-1.27269304e+00 -8.26850355e-01 6.07247174e-01 -1.02265739e+00
6.41012549e-01 -6.15667343e-01 -5.91156662e-01 6.72230542e-01
1.77836493e-01 -4.39115018e-01 1.12472677e+00 4.70482647e-01
-5.68957627e-01 -6.84651062e-02 -3.50584388e-01 6.33011937e-01
1.07955432e+00 -1.24885595e+00 -7.80059576e-01 -1.94737408e-02
4.05061483e-01 -6.77348316e-01 -1.09471583e+00 3.11886996e-01
9.94454324e-01 -7.49566197e-01 9.48880136e-01 -2.66817033e-01
4.23475564e-01 -5.91637194e-01 -4.02221292e-01 -1.19909060e+00
-1.91536948e-01 -1.24458122e+00 -1.35027424e-01 1.27117109e+00
4.11819339e-01 1.93920299e-01 6.12462342e-01 -9.07069519e-02
-1.89294428e-01 -3.72844100e-01 -8.36948156e-01 -7.34303296e-01
-4.15089652e-02 -1.00853264e+00 5.49915731e-01 9.57065105e-01
3.22141081e-01 6.34124458e-01 -5.44814050e-01 -2.44534761e-01
2.52665102e-01 5.85402489e-01 1.20230138e+00 -1.08389747e+00
-1.01481235e+00 -5.50839841e-01 -3.50428164e-01 -9.21809316e-01
-2.95752168e-01 -1.20087504e+00 -2.38283556e-02 -1.14786017e+00
1.07341446e-01 -3.57648492e-01 -5.04122436e-01 5.79251826e-01
2.47958794e-01 1.07414019e+00 5.01412272e-01 6.97427690e-01
-5.18956602e-01 1.58655927e-01 9.66710389e-01 -1.01747878e-01
-4.24671769e-01 1.24018408e-01 -7.39564061e-01 8.85130644e-01
7.40829527e-01 -4.38889354e-01 -2.90405750e-01 -1.99836969e-01
4.91716355e-01 2.37630412e-01 4.80041355e-01 -1.34656310e+00
9.82825384e-02 2.75880516e-01 2.82569200e-01 -6.35026217e-01
7.36565888e-01 -1.10042445e-01 4.84065175e-01 4.57179844e-02
-6.24300718e-01 1.51946157e-01 3.20704997e-01 2.18298033e-01
-4.90958393e-01 -2.29913667e-01 3.08046848e-01 -2.45832115e-01
-3.56657982e-01 4.18155789e-02 -1.64845601e-01 4.43644434e-01
2.26288378e-01 -1.34306595e-01 -8.87939632e-02 -6.07419014e-01
-1.04310036e+00 -4.55045551e-01 5.13826787e-01 5.33847749e-01
2.34476045e-01 -1.43376160e+00 -7.63396919e-01 1.02380842e-01
1.15790755e-01 -1.88707143e-01 9.64482948e-02 7.21922338e-01
-4.99773443e-01 1.85834765e-01 -1.54240012e-01 -4.09083009e-01
-1.46247756e+00 1.06037296e-01 4.65764888e-02 -1.15653090e-01
-7.20266700e-01 9.79487121e-01 -3.58596481e-02 -3.32467198e-01
2.18500763e-01 -5.24032354e-01 9.81679708e-02 1.77720249e-01
6.97859824e-01 2.26189017e-01 1.56757802e-01 -5.86631656e-01
-2.20829993e-01 5.16251743e-01 3.29527199e-01 -7.88995802e-01
1.25820351e+00 1.04053319e-01 3.27322483e-02 1.10765660e+00
6.95061445e-01 9.53857005e-01 -9.75907624e-01 9.64786950e-03
6.39023855e-02 -3.51944238e-01 -3.42771143e-01 -1.04887187e+00
-2.62238443e-01 7.74762630e-01 3.45068306e-01 4.16238129e-01
1.10375476e+00 4.16081734e-02 1.03514493e+00 2.95710176e-01
5.03963053e-01 -1.04925191e+00 1.88970581e-01 7.15377867e-01
1.21037674e+00 -7.30161011e-01 -7.08463192e-02 -9.11999121e-02
-6.40986562e-01 1.07953548e+00 -4.23915312e-02 -2.97553122e-01
3.53978992e-01 5.58807075e-01 3.89116526e-01 -1.33451715e-01
-7.00177312e-01 -2.30995551e-01 6.29747450e-01 2.58359969e-01
9.03835595e-01 2.51001358e-01 9.14860889e-02 1.14967895e+00
-1.25249672e+00 1.48730293e-01 2.94717103e-01 6.44747555e-01
-2.99176663e-01 -1.31613314e+00 -5.98395824e-01 1.79065183e-01
-7.93219268e-01 -4.27250296e-01 -6.89682901e-01 4.53039974e-01
2.27393404e-01 9.83871162e-01 -1.64534897e-01 -4.82001960e-01
3.35026711e-01 3.95490438e-01 6.18015170e-01 -8.35915625e-01
-1.07074237e+00 7.68572390e-01 3.93576294e-01 -4.44401115e-01
-2.02805534e-01 -6.12831414e-01 -1.15150440e+00 -1.04044907e-01
2.14743968e-02 3.97562146e-01 3.42316002e-01 5.18221915e-01
2.89822608e-01 5.86150944e-01 2.41579086e-01 -1.12185156e+00
-6.80438280e-01 -1.17910624e+00 -8.53107214e-01 5.81363976e-01
3.54336351e-01 -1.47629857e-01 -1.01248294e-01 3.54170680e-01] | [15.904976844787598, 5.386116027832031] |
0ccdc824-5880-4176-a7b3-faf2e5b73688 | barlow-graph-auto-encoder-for-unsupervised | 2110.15742 | null | https://arxiv.org/abs/2110.15742v3 | https://arxiv.org/pdf/2110.15742v3.pdf | Barlow Graph Auto-Encoder for Unsupervised Network Embedding | Network embedding has emerged as a promising research field for network analysis. Recently, an approach, named Barlow Twins, has been proposed for self-supervised learning in computer vision by applying the redundancy-reduction principle to the embedding vectors corresponding to two distorted versions of the image samples. Motivated by this, we propose Barlow Graph Auto-Encoder, a simple yet effective architecture for learning network embedding. It aims to maximize the similarity between the embedding vectors of immediate and larger neighborhoods of a node, while minimizing the redundancy between the components of these projections. In addition, we also present the variation counterpart named as Barlow Variational Graph Auto-Encoder. Our approach yields promising results for inductive link prediction and is also on par with state of the art for clustering and downstream node classification, as demonstrated by extensive comparisons with several well-known techniques on three benchmark citation datasets. | ['Martin Kleinsteuber', 'Rayyan Ahmad Khan'] | 2021-10-29 | null | null | null | null | ['inductive-link-prediction'] | ['graphs'] | [ 1.52760014e-01 4.98062491e-01 -3.66096079e-01 -5.65978885e-02
-4.08011377e-02 -2.31341913e-01 8.74937892e-01 4.00646031e-01
-4.40087989e-02 3.90574872e-01 2.97567934e-01 -1.60675555e-01
-5.51667094e-01 -8.11589181e-01 -5.64655364e-01 -7.21965253e-01
-2.14998603e-01 5.06121278e-01 -4.98279184e-02 -7.10572004e-02
-2.15921924e-02 4.78500634e-01 -1.09024668e+00 -4.47116718e-02
4.70948428e-01 6.68667078e-01 5.85819967e-02 4.53702569e-01
6.34980500e-02 8.31217408e-01 4.99691479e-02 -7.89830029e-01
-1.48912845e-02 -4.14548963e-01 -7.96527445e-01 1.27042413e-01
3.22422564e-01 9.39007923e-02 -1.15951109e+00 9.88558233e-01
1.97942019e-01 -5.66940196e-02 1.00888705e+00 -1.37875211e+00
-9.80900705e-01 8.51949573e-01 -5.39487064e-01 1.69595808e-01
5.55703193e-02 -4.29599583e-01 1.72273111e+00 -8.05996716e-01
9.90426064e-01 1.20412433e+00 5.39516985e-01 3.38701755e-01
-1.83492279e+00 -1.41423970e-01 -2.36436218e-01 6.87081993e-01
-1.42092407e+00 -2.96806812e-01 1.17781436e+00 -5.61826229e-01
6.96860790e-01 3.56254913e-02 5.40064633e-01 1.06416368e+00
2.35594422e-01 5.89350045e-01 6.30458474e-01 -5.91185749e-01
6.74465746e-02 2.41851807e-01 8.88457224e-02 8.93223882e-01
1.19357832e-01 -7.12299645e-02 -4.58170503e-01 -1.67334229e-01
5.67009687e-01 -5.74686937e-03 -3.12480211e-01 -1.04510653e+00
-1.25007761e+00 1.05029786e+00 9.36032832e-01 5.04858553e-01
-2.71376401e-01 2.61627674e-01 3.69776428e-01 3.88288409e-01
4.83672112e-01 1.59841076e-01 3.98539081e-02 3.87637556e-01
-7.74438381e-01 -1.68514356e-01 8.61669600e-01 6.68787062e-01
7.47065008e-01 1.17409676e-02 5.68185095e-03 6.63757741e-01
6.38652682e-01 2.61375178e-02 3.60469937e-01 -1.11598289e+00
2.02441633e-01 7.56160975e-01 -7.59185970e-01 -1.41251624e+00
-2.56031156e-01 -4.90144730e-01 -1.25702405e+00 3.06639969e-02
1.13481522e-01 2.33886898e-01 -2.17026040e-01 1.61163664e+00
1.06807522e-01 4.75288033e-01 -1.39452014e-02 3.24018478e-01
8.17565322e-01 5.96607983e-01 -2.82917917e-01 -3.29956383e-01
9.25967395e-01 -8.58553886e-01 -6.49425685e-01 2.41757900e-01
6.08295798e-01 -2.57519245e-01 5.13160765e-01 -7.67013878e-02
-8.44518244e-01 -3.89455795e-01 -9.56063449e-01 -1.32947499e-02
-3.78998160e-01 5.45711257e-02 5.72370589e-01 3.87715757e-01
-1.28686297e+00 1.00344181e+00 -7.44880736e-01 -5.26319802e-01
4.94699210e-01 2.17695594e-01 -8.29856455e-01 -9.40870047e-02
-8.66769195e-01 6.40124142e-01 1.35673672e-01 -1.76084459e-01
-6.33497179e-01 -5.52763999e-01 -1.01338995e+00 3.97581071e-01
1.59013018e-01 -6.19163513e-01 3.29302490e-01 -5.32423913e-01
-1.29004693e+00 8.69891346e-01 -9.24217850e-02 -7.18642652e-01
1.60129949e-01 2.40089104e-01 -2.94151515e-01 2.80964077e-01
6.76740706e-02 7.92223752e-01 1.01060164e+00 -1.13193297e+00
1.54018402e-01 -5.89353085e-01 -1.38363421e-01 -4.03703634e-05
-8.35372686e-01 -3.46803963e-01 -2.63138652e-01 -6.18087769e-01
7.87073448e-02 -9.13353324e-01 -2.83668339e-01 3.27984124e-01
-5.85189104e-01 -3.30633938e-01 9.12317157e-01 -4.92207140e-01
1.14302802e+00 -2.10166049e+00 7.94620275e-01 4.93007034e-01
7.49130607e-01 2.63275262e-02 -3.56115133e-01 6.91446483e-01
-4.82714266e-01 6.88482970e-02 -2.10762933e-01 -4.24079180e-01
-1.68679029e-01 3.52112114e-01 -8.76540169e-02 6.82394624e-01
2.34541252e-01 9.96606171e-01 -9.35869277e-01 -6.30347490e-01
3.85574877e-01 7.31336534e-01 -5.00083148e-01 1.42960101e-02
1.38007849e-01 1.68713093e-01 -1.34497315e-01 2.01488093e-01
2.05972269e-01 -6.39925063e-01 6.00558341e-01 -4.73303139e-01
3.34686011e-01 -7.55815161e-03 -7.77994812e-01 1.53685808e+00
-1.12483978e-01 1.11657298e+00 -4.03310992e-02 -1.73278999e+00
9.14862692e-01 3.86485606e-01 7.63618231e-01 -2.71281809e-01
6.98872805e-02 -2.87903577e-01 1.73364773e-01 -3.01754743e-01
7.13551193e-02 2.49376968e-01 3.97483170e-01 3.68094236e-01
4.87141639e-01 2.03401014e-01 3.16479295e-01 7.67771184e-01
1.31260145e+00 -1.42917573e-01 3.50959331e-01 -1.79303631e-01
7.33150721e-01 -4.65793967e-01 1.70259520e-01 2.95012087e-01
-1.67200625e-01 3.92603308e-01 8.94572794e-01 -1.01705708e-01
-1.21526217e+00 -1.23023152e+00 -9.78900343e-02 5.17990053e-01
-8.69884416e-02 -7.18331873e-01 -7.08036721e-01 -7.21156597e-01
1.06658094e-01 4.13118750e-01 -6.69284999e-01 -3.03493887e-01
-3.17865521e-01 -3.81989956e-01 3.61028373e-01 3.48714381e-01
1.64042756e-01 -7.63134718e-01 3.39698613e-01 1.08316168e-01
7.24017248e-02 -1.13471973e+00 -2.66506523e-01 -5.24643138e-02
-9.96193707e-01 -1.22060406e+00 -5.77390790e-01 -9.73966181e-01
6.74705327e-01 3.11646432e-01 9.73084211e-01 3.48271988e-02
-5.55430770e-01 4.44009662e-01 -2.46702150e-01 3.19572568e-01
-5.02641916e-01 1.14820123e-01 4.83334102e-02 3.79633695e-01
8.08835588e-03 -8.82437229e-01 -3.82606685e-01 1.46479672e-02
-6.82582617e-01 -2.80482948e-01 4.99278486e-01 9.99628305e-01
5.17915308e-01 1.24135651e-01 5.80448210e-01 -8.85983109e-01
6.58141375e-01 -6.36670470e-01 -3.68961900e-01 3.26589555e-01
-1.03226590e+00 3.34224612e-01 6.79341257e-01 -6.12238795e-03
-4.38248873e-01 1.14546278e-02 2.07383573e-01 -7.48271406e-01
2.18966544e-01 5.95368564e-01 4.47652340e-02 -3.42254192e-01
4.68559176e-01 2.15357076e-02 5.26756287e-01 -3.56062382e-01
8.32919657e-01 4.90783691e-01 3.64789516e-01 -1.87624935e-02
9.81729567e-01 6.55462265e-01 5.58519185e-01 -1.03202939e+00
-4.26623583e-01 -5.34536481e-01 -9.73147988e-01 -3.04333031e-01
8.02735090e-01 -7.51608193e-01 -7.14303136e-01 -1.40698761e-01
-9.81493771e-01 1.76842779e-01 -3.22411150e-01 3.36579621e-01
-5.73273480e-01 7.62492001e-01 -7.44631052e-01 -4.63885069e-01
-2.38726437e-01 -8.85546863e-01 6.29396379e-01 -8.70670378e-02
-1.61624089e-01 -1.54356062e+00 3.54940861e-01 3.29388648e-01
8.38764310e-02 2.72997558e-01 1.17941403e+00 -4.70160723e-01
-4.25303042e-01 -3.15269679e-01 -3.60056937e-01 4.87747759e-01
3.42930928e-02 2.61530280e-01 -6.53407514e-01 -3.85896653e-01
-4.94491190e-01 3.82207818e-02 1.07112730e+00 4.37105924e-01
9.58525360e-01 -2.43378490e-01 -8.29182386e-01 5.02170980e-01
1.43561780e+00 -3.57442111e-01 4.67910320e-01 -7.71319196e-02
9.64699864e-01 7.34963953e-01 -1.11329548e-01 1.34546697e-01
3.91730189e-01 7.24891722e-01 5.03949165e-01 9.34838802e-02
-2.38454595e-01 -3.33157986e-01 1.87463850e-01 1.04005420e+00
-8.47916603e-02 -2.33441502e-01 -7.13444054e-01 6.72032654e-01
-2.00810909e+00 -1.10415435e+00 -3.02687258e-01 1.89833283e+00
2.40896538e-01 7.16401562e-02 8.01102072e-02 2.98666984e-01
8.16724181e-01 5.23835123e-01 -3.15111846e-01 -1.27775952e-01
-2.15477318e-01 1.49600029e-01 3.49551678e-01 5.59727311e-01
-9.80155706e-01 7.01889694e-01 6.94474459e+00 6.49907231e-01
-5.80138981e-01 7.62904882e-02 4.98691291e-01 1.40257791e-01
-4.35479403e-01 2.07788453e-01 -2.57823467e-01 2.87942767e-01
9.87553239e-01 -3.25538844e-01 5.17926514e-01 7.11509645e-01
4.17922586e-02 4.44595695e-01 -1.31521499e+00 9.55486894e-01
1.99503988e-01 -1.76248634e+00 1.14914075e-01 3.87134373e-01
7.68070996e-01 3.99532467e-02 1.83586940e-01 -9.12340283e-02
1.47815391e-01 -9.67061758e-01 5.51326834e-02 5.88449419e-01
5.40015221e-01 -7.03764558e-01 6.15043461e-01 9.97354612e-02
-1.23279452e+00 -9.79745481e-03 -5.22151530e-01 1.82819981e-02
6.50527626e-02 7.33179331e-01 -8.73900473e-01 6.04265869e-01
3.96924496e-01 1.39325607e+00 -5.84631264e-01 6.80232942e-01
-2.26319119e-01 5.66735983e-01 8.68543535e-02 8.73942077e-02
5.71271665e-02 -5.76650858e-01 7.27917314e-01 8.08406651e-01
-4.28155391e-03 -2.61934727e-01 -1.76679552e-01 1.01593244e+00
-4.62811142e-01 8.05349275e-02 -1.10367644e+00 -3.28763425e-01
4.74709094e-01 1.45916522e+00 -4.62998778e-01 -1.37326464e-01
-4.57801372e-01 1.05518329e+00 6.78100586e-01 2.65894502e-01
-4.02322561e-01 -2.60911852e-01 6.13948464e-01 1.05966814e-01
4.59482759e-01 -2.46004596e-01 -5.06204665e-02 -1.03974509e+00
-1.71407104e-01 -2.52820522e-01 3.27829301e-01 -6.51581407e-01
-1.45179939e+00 4.53619391e-01 -1.87957898e-01 -1.14695966e+00
-3.82191181e-01 -7.67434478e-01 -7.73651659e-01 4.79742080e-01
-1.30335963e+00 -1.12403500e+00 -2.80651748e-02 6.50301933e-01
1.40720725e-01 -5.71791768e-01 9.27904189e-01 2.57765472e-01
-7.25201547e-01 5.55260479e-01 6.90282762e-01 3.38959932e-01
3.75960410e-01 -1.32597196e+00 2.84368873e-01 4.84253466e-01
7.82019854e-01 2.99710274e-01 3.72996390e-01 -4.24492776e-01
-1.32621944e+00 -1.02734685e+00 1.10072625e+00 -2.81745821e-01
1.00562143e+00 -2.56077111e-01 -8.32702458e-01 8.51295829e-01
4.23623651e-01 1.54765189e-01 7.10390687e-01 1.49858639e-01
-4.56923813e-01 -2.66598225e-01 -8.78669202e-01 6.50976300e-01
9.92958009e-01 -8.23686481e-01 -3.10891807e-01 4.57514912e-01
7.47600973e-01 3.75201970e-01 -1.18646729e+00 2.07479984e-01
3.27679098e-01 -8.69721234e-01 1.26615214e+00 -5.95072746e-01
7.18549132e-01 -6.01675501e-03 -4.58883084e-02 -1.36703551e+00
-7.62397230e-01 -5.35663605e-01 -4.94108588e-01 1.19906664e+00
2.67029643e-01 -3.35582256e-01 1.09622967e+00 3.84881906e-02
1.60275772e-01 -8.38075399e-01 -9.28528547e-01 -6.00403368e-01
-1.18782409e-01 -4.38637584e-02 1.26071237e-02 1.11973011e+00
3.78206223e-01 7.95394719e-01 -3.35541606e-01 -4.87368703e-02
1.11968613e+00 -1.28319636e-01 3.96487027e-01 -1.50222707e+00
-2.22404882e-01 -6.24782264e-01 -1.05751336e+00 -7.04471588e-01
6.70337439e-01 -1.54155779e+00 -5.60577154e-01 -1.53557503e+00
2.53363818e-01 -6.33928254e-02 -2.68417567e-01 3.33397351e-02
1.72489315e-01 2.81729341e-01 1.16219819e-01 3.41730773e-01
-3.62243712e-01 7.69538641e-01 1.01306844e+00 -3.33303481e-01
1.14840217e-01 -3.41196470e-02 -4.42276001e-01 5.59110045e-01
5.17955542e-01 -2.34506473e-01 -3.59453738e-01 -1.19640060e-01
2.11729184e-01 1.56470478e-01 4.25916553e-01 -8.39944899e-01
3.69489104e-01 4.43390459e-01 1.26803517e-01 -5.22360265e-01
3.97144943e-01 -8.91941011e-01 1.09457247e-01 5.01088798e-01
-6.15928650e-01 5.76378312e-03 -4.80864227e-01 1.08100116e+00
-3.37260187e-01 -2.63081074e-01 6.66773140e-01 6.63364977e-02
-5.43220103e-01 4.59310919e-01 -1.49647281e-01 -2.36975163e-01
1.09951496e+00 3.62707581e-03 -1.64960772e-01 -5.18299401e-01
-9.54003453e-01 1.94055691e-01 2.16701403e-01 2.45995477e-01
7.94636548e-01 -1.60348332e+00 -8.09657276e-01 1.37380481e-01
6.20386116e-02 -4.82076854e-01 1.19683221e-01 8.55379879e-01
-3.52090210e-01 5.16189396e-01 -1.63595125e-01 -6.11468911e-01
-1.36837268e+00 7.68487453e-01 1.82698816e-01 -4.61621225e-01
-5.91062844e-01 6.48206651e-01 -1.12934403e-01 -3.26385289e-01
1.48895025e-01 1.88569605e-01 -5.01563251e-01 1.81157127e-01
1.97456867e-01 6.29725933e-01 -1.98391184e-01 -7.93489337e-01
-2.76565731e-01 6.08941257e-01 -6.47562137e-03 4.50623594e-02
1.36508715e+00 -1.22427866e-01 -5.44955313e-01 4.95259136e-01
1.69784224e+00 -2.90015519e-01 -8.66822541e-01 -6.30568385e-01
1.54875740e-01 -1.99924290e-01 1.29018515e-01 1.42101973e-01
-1.20320058e+00 9.73871648e-01 4.35891122e-01 6.04816079e-01
5.47420561e-01 4.11296219e-01 2.79623330e-01 4.81346130e-01
3.85456299e-03 -7.92544544e-01 3.57990652e-01 3.15914243e-01
6.67330921e-01 -1.32672262e+00 8.80783796e-02 -5.34207940e-01
-3.20480466e-01 1.14467120e+00 1.47067159e-01 -4.81360406e-01
1.11706507e+00 -2.29168475e-01 -4.92485136e-01 -3.96821231e-01
-9.04399812e-01 -1.30460411e-01 4.93560702e-01 7.44782209e-01
5.27993381e-01 2.54229486e-01 -2.01513439e-01 -2.00180903e-01
6.42057927e-03 -5.02228498e-01 4.57656294e-01 2.30762750e-01
-2.12701976e-01 -1.20613348e+00 1.22176960e-01 6.33877873e-01
-1.05061531e-01 -1.24933586e-01 -5.65466702e-01 6.71254873e-01
-3.18283528e-01 7.47715056e-01 2.51568109e-01 -5.89986086e-01
-1.07006244e-02 -3.31303701e-02 4.80831146e-01 -5.50925851e-01
-7.87926391e-02 -8.05942267e-02 -7.76974857e-02 -4.81606901e-01
-5.84484875e-01 -6.62000835e-01 -9.17497039e-01 -3.91418248e-01
-3.38709325e-01 1.08211011e-01 5.00348032e-01 7.93936789e-01
4.66666460e-01 5.24481416e-01 9.69699919e-01 -7.73644567e-01
-5.24691403e-01 -8.58172655e-01 -7.99520612e-01 5.28721869e-01
1.35713249e-01 -5.11289179e-01 -4.51522827e-01 1.66708622e-02] | [7.1565375328063965, 6.128909587860107] |
6ae0e98a-1dac-4e00-9eec-928eefcda99d | a-bilingual-openworld-video-text-dataset-and | 2112.04888 | null | https://arxiv.org/abs/2112.04888v1 | https://arxiv.org/pdf/2112.04888v1.pdf | A Bilingual, OpenWorld Video Text Dataset and End-to-end Video Text Spotter with Transformer | Most existing video text spotting benchmarks focus on evaluating a single language and scenario with limited data. In this work, we introduce a large-scale, Bilingual, Open World Video text benchmark dataset(BOVText). There are four features for BOVText. Firstly, we provide 2,000+ videos with more than 1,750,000+ frames, 25 times larger than the existing largest dataset with incidental text in videos. Secondly, our dataset covers 30+ open categories with a wide selection of various scenarios, e.g., Life Vlog, Driving, Movie, etc. Thirdly, abundant text types annotation (i.e., title, caption or scene text) are provided for the different representational meanings in video. Fourthly, the BOVText provides bilingual text annotation to promote multiple cultures live and communication. Besides, we propose an end-to-end video text spotting framework with Transformer, termed TransVTSpotter, which solves the multi-orient text spotting in video with a simple, but efficient attention-based query-key mechanism. It applies object features from the previous frame as a tracking query for the current frame and introduces a rotation angle prediction to fit the multiorient text instance. On ICDAR2015(video), TransVTSpotter achieves the state-of-the-art performance with 44.1% MOTA, 9 fps. The dataset and code of TransVTSpotter can be found at github:com=weijiawu=BOVText and github:com=weijiawu=TransVTSpotter, respectively. | ['Hong Zhou', 'Yejun Tang', 'Jiahong Li', 'Zhuang Li', 'Sibo Wang', 'Debing Zhang', 'Yuanqiang Cai', 'Weijia Wu'] | 2021-12-09 | null | null | null | null | ['text-spotting', 'text-annotation'] | ['computer-vision', 'natural-language-processing'] | [ 6.20701090e-02 -5.70590258e-01 -4.12322193e-01 -1.27327055e-01
-7.21069396e-01 -2.98314989e-01 5.21991849e-01 -2.52003938e-01
-5.31149209e-01 4.82572585e-01 5.75217545e-01 -1.28231004e-01
1.91922873e-01 -3.89823318e-01 -9.34675157e-01 -4.87134546e-01
4.46458876e-01 3.62582058e-01 3.74426007e-01 -2.91544676e-01
3.39639425e-01 -2.92208374e-01 -1.41371560e+00 6.98414087e-01
6.64792418e-01 1.20694923e+00 6.39035702e-01 5.94845653e-01
-2.51735628e-01 9.07210112e-01 -5.27265847e-01 -5.51353574e-01
9.73684639e-02 -2.52023369e-01 -7.25395918e-01 -4.88197058e-02
8.38075221e-01 -6.01484895e-01 -9.80009556e-01 8.60290647e-01
6.55427158e-01 7.92168453e-02 3.20039719e-01 -1.43767679e+00
-6.36425972e-01 6.20950818e-01 -7.92884052e-01 4.99655128e-01
7.07861245e-01 2.26826206e-01 9.63778079e-01 -1.20070386e+00
8.50739300e-01 1.23960817e+00 2.41769522e-01 2.69163221e-01
-5.81789553e-01 -1.02790511e+00 2.59362757e-01 6.33316219e-01
-1.60321152e+00 -5.25994956e-01 4.58537370e-01 -4.62667465e-01
7.73621261e-01 4.34182763e-01 7.84009397e-01 1.39890599e+00
2.57118434e-01 1.12805700e+00 5.64452589e-01 -9.01768133e-02
-2.66494155e-01 -3.04514378e-01 -3.14893603e-01 6.46814883e-01
-5.99573478e-02 -4.04143393e-01 -1.01767075e+00 2.90988892e-01
7.71458924e-01 1.91093698e-01 -6.25461817e-01 -3.77266966e-02
-1.97645843e+00 5.43687761e-01 -1.39951529e-02 2.28613079e-01
2.25367192e-02 3.30951959e-01 9.51903462e-01 5.21223545e-01
3.32558751e-01 -5.56160919e-02 -4.20838535e-01 -6.19856238e-01
-8.55544090e-01 6.40905201e-02 3.73066813e-01 1.56538332e+00
7.03488052e-01 6.87005445e-02 -6.82743192e-01 1.00314140e+00
6.74044490e-02 9.42778230e-01 7.54855931e-01 -6.91615939e-01
1.23646390e+00 4.56067801e-01 -2.21580014e-01 -1.19318306e+00
-2.00417101e-01 3.54642063e-01 -6.87249720e-01 -7.20102370e-01
-6.91220313e-02 -1.80118009e-01 -8.35939765e-01 1.16008830e+00
1.75143868e-01 2.97711700e-01 -1.60889134e-01 1.17027843e+00
1.38080919e+00 1.11033952e+00 -2.71959692e-01 -3.41405571e-01
1.52421749e+00 -1.41542733e+00 -8.31367731e-01 -1.56282052e-01
7.29286492e-01 -1.15705371e+00 1.30309904e+00 5.02416432e-01
-8.52337062e-01 -4.77563828e-01 -6.35389388e-01 -4.78142381e-01
-3.03726107e-01 4.35814321e-01 1.59212872e-01 2.38842055e-01
-8.42446566e-01 -2.44396571e-02 -4.49914753e-01 -7.73512363e-01
2.92478561e-01 2.19911069e-01 -3.52102816e-01 -3.85362506e-01
-1.13862848e+00 2.96908855e-01 5.18273771e-01 -1.82900101e-01
-8.83835137e-01 -5.22651851e-01 -7.89279878e-01 -1.73110664e-01
1.02483153e+00 -4.20173734e-01 9.31156635e-01 -1.02932930e+00
-1.34632456e+00 7.52460241e-01 -3.08247685e-01 -1.93382114e-01
6.88004375e-01 -4.27832752e-01 -7.31386065e-01 5.21181107e-01
3.34427923e-01 7.81387687e-01 1.08324313e+00 -5.96366763e-01
-8.22539151e-01 -3.50581668e-02 1.87744703e-02 4.28975999e-01
-7.56828785e-01 3.44590932e-01 -1.38158035e+00 -1.23218536e+00
-3.01868170e-01 -8.42628062e-01 5.40004730e-01 1.11184279e-02
-5.48240542e-01 -1.45244956e-01 1.37505376e+00 -7.87785053e-01
1.61189806e+00 -2.33816743e+00 1.76056966e-01 -3.04029673e-01
2.81288862e-01 3.86852436e-02 -8.46273173e-03 5.60968339e-01
-3.18952203e-02 1.13031484e-01 2.94431359e-01 -1.12132989e-01
7.69466683e-02 -1.47838503e-01 -2.89216816e-01 3.60649765e-01
-3.71539384e-01 1.06581867e+00 -6.15416348e-01 -9.21858132e-01
5.39106011e-01 1.52716354e-01 -4.09747601e-01 -9.42025613e-03
-2.70867527e-01 1.75740689e-01 -6.30205929e-01 7.74747252e-01
4.32397187e-01 -2.29718477e-01 -1.74689472e-01 -5.47170341e-01
-2.05859512e-01 -1.97629526e-01 -1.00737226e+00 2.12808871e+00
-2.16225937e-01 1.14253843e+00 -2.32476339e-01 -7.68307149e-01
5.49681187e-01 3.62645894e-01 8.88081074e-01 -8.12333643e-01
4.10078377e-01 7.10159168e-02 -5.70400119e-01 -8.82356882e-01
8.53541553e-01 5.65710783e-01 -2.37273142e-01 3.29567164e-01
1.33399963e-01 1.26178503e-01 6.31943285e-01 4.66322929e-01
9.12373722e-01 3.03557128e-01 7.85374083e-03 -1.25061497e-01
5.06925523e-01 -4.46895063e-02 4.39413518e-01 4.05821294e-01
-2.10513055e-01 7.48767257e-01 3.59871268e-01 -5.43246329e-01
-8.99042964e-01 -5.22096395e-01 1.97483134e-03 1.30949807e+00
5.41366220e-01 -9.04257774e-01 -5.81791759e-01 -4.85380262e-01
-2.07872942e-01 3.23725373e-01 -4.60572690e-01 -1.87146384e-02
-6.08223617e-01 -3.48539829e-01 6.61508739e-01 2.92702138e-01
8.38808596e-01 -1.00333095e+00 -5.21524131e-01 -6.20637201e-02
-1.01114261e+00 -1.63673341e+00 -1.32162941e+00 -4.09653097e-01
-4.88296270e-01 -9.63098645e-01 -1.07101274e+00 -9.23960209e-01
2.51784325e-01 5.61221063e-01 9.79111552e-01 1.49930105e-01
-2.64325947e-01 3.97513449e-01 -7.90354431e-01 8.90104398e-02
1.71741232e-01 -2.76151821e-02 -6.58897534e-02 3.01763862e-01
3.93703073e-01 9.25346538e-02 -6.99966729e-01 7.93771923e-01
-8.75340760e-01 6.83061779e-01 3.45269650e-01 8.75737190e-01
4.92242455e-01 -1.65967673e-01 7.46808648e-02 -3.13977420e-01
1.40499368e-01 -5.24623573e-01 -3.22210968e-01 3.30090672e-01
-7.17585534e-02 -5.15809298e-01 6.76193058e-01 -6.27121925e-01
-8.07350576e-01 -1.20090485e-01 -8.43647346e-02 -8.54633808e-01
-7.90675431e-02 4.05657262e-01 -9.63072255e-02 4.77125794e-02
1.35184586e-01 5.90119064e-01 -4.97669369e-01 -1.74792007e-01
6.90310746e-02 8.45357895e-01 4.47193533e-01 -6.65461898e-01
4.70451355e-01 3.96674216e-01 -5.28498948e-01 -9.38016295e-01
-4.03492570e-01 -5.69561005e-01 -2.99817175e-01 -5.39665937e-01
1.02521324e+00 -1.24663103e+00 -8.33490551e-01 6.43466532e-01
-8.42506528e-01 -4.86366212e-01 1.23054817e-01 6.35067761e-01
-6.82063818e-01 6.90344214e-01 -6.17378771e-01 -2.40893304e-01
-4.58524376e-01 -1.38482964e+00 1.56198406e+00 -5.69370724e-02
1.61923081e-01 -6.05697513e-01 -5.25115848e-01 6.87724173e-01
4.02450655e-03 -7.69606754e-02 5.30471563e-01 -3.84228230e-01
-7.22110033e-01 1.21074446e-01 -3.88931751e-01 -2.56414592e-01
-9.26499665e-02 9.62464288e-02 -5.57084560e-01 -4.88776118e-01
-4.76302534e-01 -4.00259435e-01 9.23458993e-01 2.68185258e-01
1.44613695e+00 -3.07755649e-01 -3.19641620e-01 8.65938723e-01
1.15162849e+00 4.21148747e-01 7.57079780e-01 6.04700446e-01
1.17516136e+00 1.90708935e-01 1.14214933e+00 6.51175261e-01
6.11419559e-01 1.01181710e+00 3.03733826e-01 3.83410417e-02
-1.31213233e-01 -2.71107256e-01 6.64735556e-01 1.07300162e+00
5.40010706e-02 -7.39199519e-01 -8.30575228e-01 3.65882933e-01
-1.98754168e+00 -1.12485373e+00 -2.65999824e-01 1.99888802e+00
5.68634987e-01 1.05296716e-01 1.42516896e-01 -1.91932861e-02
9.28222477e-01 3.49933177e-01 -6.01027668e-01 1.27143517e-01
-3.39633405e-01 -4.07884389e-01 5.91147482e-01 3.96649688e-02
-1.22056139e+00 1.27159393e+00 4.44438887e+00 1.47240579e+00
-1.32539070e+00 2.17985868e-01 7.05267549e-01 -6.40073419e-01
4.63862047e-02 -2.09360227e-01 -6.71317518e-01 1.00913858e+00
6.14641726e-01 -3.17420959e-01 5.77821612e-01 3.51984650e-01
4.30708438e-01 -1.59587011e-01 -8.50425184e-01 1.59433222e+00
6.18955553e-01 -1.49662495e+00 3.24928522e-01 -2.28451177e-01
6.13559425e-01 1.30859956e-01 -2.46168468e-02 3.49260539e-01
-2.70505786e-01 -6.43898666e-01 1.19295299e+00 3.16310287e-01
1.40995288e+00 -6.19985223e-01 4.25627619e-01 7.89332204e-03
-1.61002731e+00 -1.73911467e-01 -1.70386806e-01 4.47007716e-01
2.94708759e-01 2.34888092e-01 -2.40978360e-01 8.87241304e-01
1.16881192e+00 1.32513630e+00 -5.71505010e-01 9.28525865e-01
2.13514596e-01 4.83057797e-01 -3.10380727e-01 -1.61749765e-01
1.88486174e-01 -1.14814408e-01 5.73838890e-01 1.25715554e+00
6.16182923e-01 1.21548157e-02 4.16305453e-01 2.41019905e-01
-2.57694393e-01 5.88130534e-01 -4.16163921e-01 -1.13348253e-01
4.65476900e-01 1.10399711e+00 -8.56440842e-01 -5.61065972e-01
-6.67310536e-01 1.30806684e+00 -2.29285538e-01 5.26717126e-01
-1.17367995e+00 -4.40492511e-01 2.22873777e-01 -8.26188624e-02
6.03489518e-01 -3.19372900e-02 2.96046734e-01 -1.69118345e+00
3.31374645e-01 -1.17074394e+00 5.07965088e-01 -1.40749943e+00
-6.74567759e-01 5.73139429e-01 5.96333705e-02 -1.57783663e+00
3.21891487e-01 -5.12199879e-01 -2.42584601e-01 3.27145785e-01
-1.06690669e+00 -1.31213784e+00 -6.79370344e-01 1.12680113e+00
1.34012628e+00 -3.98495495e-01 1.89470097e-01 8.55246782e-01
-8.54655266e-01 8.06482315e-01 3.35830212e-01 3.18847656e-01
1.21772265e+00 -6.75661683e-01 5.25884092e-01 7.23156214e-01
2.30108619e-01 1.34513170e-01 4.63915318e-01 -8.28500211e-01
-1.89227724e+00 -1.32578242e+00 5.59841037e-01 -4.16000605e-01
8.16692770e-01 -5.51208794e-01 -4.62361187e-01 8.43154430e-01
3.54895651e-01 -3.94211560e-02 1.30427867e-01 -5.36383808e-01
-6.01542294e-02 -1.86724141e-01 -5.89659512e-01 8.29689980e-01
1.16178572e+00 -4.25977170e-01 -1.77843064e-01 5.88993669e-01
1.10494232e+00 -8.68581593e-01 -9.16217089e-01 6.17271960e-02
7.06180274e-01 -7.53835857e-01 8.32944930e-01 -1.39039040e-01
7.03572869e-01 -2.73443431e-01 -2.82562971e-01 -7.13465452e-01
2.14758646e-02 -8.58954608e-01 -7.21845403e-02 1.23416376e+00
9.43907648e-02 -5.14721572e-01 5.52348852e-01 7.13066980e-02
-4.73490477e-01 -7.13937104e-01 -1.12672746e+00 -5.67869425e-01
-2.64019728e-01 -5.95018864e-01 5.52252889e-01 1.14489973e+00
9.96076465e-02 3.54074538e-01 -9.75056350e-01 -4.81032670e-01
1.52710587e-01 1.10415123e-01 9.23472464e-01 -6.52654648e-01
-4.83625345e-02 -3.68928254e-01 -6.12692058e-01 -1.63242912e+00
-8.95565078e-02 -7.44839430e-01 -2.17057630e-01 -1.38298678e+00
3.10579956e-01 -5.82621172e-02 -2.17239354e-02 4.36402947e-01
-1.25238970e-02 4.14069295e-01 4.99106973e-01 4.24439311e-01
-1.13919902e+00 7.11995721e-01 1.46227074e+00 -3.83700401e-01
2.82213151e-01 -5.66340268e-01 -3.98645222e-01 5.70050955e-01
4.05073345e-01 -2.35508129e-01 -2.41789788e-01 -8.64078999e-01
3.03790241e-01 3.70701551e-01 3.76389921e-01 -8.71074438e-01
3.32631230e-01 -2.92928427e-01 2.26531535e-01 -1.04772484e+00
4.70215678e-01 -9.34106827e-01 9.28936899e-02 8.86302292e-02
-2.87331581e-01 4.00970846e-01 1.74115509e-01 5.56258559e-01
-2.28464425e-01 1.59668066e-02 2.48288214e-01 1.96537271e-01
-1.21864212e+00 6.81559682e-01 -3.39502245e-01 4.68522191e-01
1.13531470e+00 -3.65966380e-01 -7.06273675e-01 -4.46302265e-01
-3.90944600e-01 5.66099644e-01 5.62484860e-01 1.02207625e+00
6.90770268e-01 -1.48520315e+00 -7.12324739e-01 2.36296058e-02
5.15949070e-01 -1.78501591e-01 6.29863739e-01 1.17157829e+00
-5.79088628e-01 5.66437781e-01 -1.53543070e-01 -7.41490185e-01
-1.37850320e+00 6.98398709e-01 9.79787186e-02 2.13323355e-01
-9.28545892e-01 5.15652120e-01 4.90989625e-01 7.75461346e-02
2.55422026e-01 -4.01640445e-01 -1.95317268e-01 7.23258331e-02
6.61378145e-01 3.64698291e-01 5.43685295e-02 -1.10192156e+00
-3.01278740e-01 8.23999882e-01 -1.09193228e-01 1.24647647e-01
7.62990117e-01 -5.72871089e-01 6.00519404e-02 4.35645819e-01
1.23067021e+00 -1.00354180e-01 -1.06758630e+00 -2.37684384e-01
-3.72783273e-01 -8.88962984e-01 -2.30303574e-02 -6.42523587e-01
-1.38921559e+00 6.99901760e-01 6.45820916e-01 -2.15082884e-01
1.23663294e+00 -6.57540038e-02 1.19123197e+00 5.12230694e-01
4.25728917e-01 -1.38760149e+00 3.03158492e-01 6.05176032e-01
9.82853711e-01 -1.37822771e+00 1.45450190e-01 -2.24404439e-01
-8.83087456e-01 1.12492132e+00 7.48999596e-01 3.95169973e-01
2.60713577e-01 6.89832270e-02 2.53202617e-02 -1.37194365e-01
-9.61164951e-01 5.58476802e-03 2.03571528e-01 1.61414906e-01
3.81347507e-01 -1.29730925e-01 -2.17887029e-01 2.78579950e-01
-2.25370213e-01 2.89263017e-02 5.09557784e-01 8.88658941e-01
-2.44014576e-01 -4.97980744e-01 -4.26277220e-01 7.80510128e-01
-5.17960548e-01 -3.89442116e-01 -1.46952510e-01 6.74711704e-01
1.17371948e-02 1.04115617e+00 1.16839938e-01 -6.14429057e-01
2.45507687e-01 -3.07724833e-01 1.06148131e-01 -2.39862457e-01
-3.74605328e-01 5.26711702e-01 1.56448960e-01 -6.58024907e-01
-4.31797951e-01 -6.84864521e-01 -1.12170577e+00 -5.61789870e-01
-2.55285591e-01 1.31287137e-02 3.27099651e-01 9.01962817e-01
4.10371006e-01 5.11369228e-01 4.41846609e-01 -8.35987031e-01
3.97218585e-01 -9.77012932e-01 -3.49563390e-01 3.22033852e-01
2.43797675e-01 -8.01952600e-01 -1.93184949e-02 6.06240869e-01] | [11.89548110961914, 2.120465040206909] |
b24502b3-02eb-43f3-88d2-15d1337a026f | improved-batching-strategy-for-irregular-time | 2207.05708 | null | https://arxiv.org/abs/2207.05708v1 | https://arxiv.org/pdf/2207.05708v1.pdf | Improved Batching Strategy For Irregular Time-Series ODE | Irregular time series data are prevalent in the real world and are challenging to model with a simple recurrent neural network (RNN). Hence, a model that combines the use of ordinary differential equations (ODE) and RNN was proposed (ODE-RNN) to model irregular time series with higher accuracy, but it suffers from high computational costs. In this paper, we propose an improvement in the runtime on ODE-RNNs by using a different efficient batching strategy. Our experiments show that the new models reduce the runtime of ODE-RNN significantly ranging from 2 times up to 49 times depending on the irregularity of the data while maintaining comparable accuracy. Hence, our model can scale favorably for modeling larger irregular data sets. | ['Nathan Perlmutter', 'Ahmed Khorshid', 'Yony Bresler', 'Ting Fung Lam'] | 2022-07-12 | null | null | null | null | ['irregular-time-series'] | ['time-series'] | [-2.64397949e-01 -2.44849831e-01 3.26319695e-01 1.43906251e-01
-4.14285570e-01 -1.93215996e-01 3.81878823e-01 -7.74229243e-02
-2.33519062e-01 7.06041932e-01 4.45136651e-02 -6.46147430e-01
-6.69354871e-02 -7.31831849e-01 -5.82612216e-01 -5.23232341e-01
-3.04065496e-01 1.86741844e-01 9.38793942e-02 -5.19912064e-01
2.73934659e-02 7.80358076e-01 -1.55142021e+00 -4.93844673e-02
6.88531935e-01 1.02699625e+00 -3.71073233e-03 7.39676595e-01
-1.71249866e-01 1.37148094e+00 -6.23724043e-01 3.14005971e-01
4.32741582e-01 -3.56022239e-01 -5.17303109e-01 -4.34401333e-01
-2.30307654e-01 -2.36851364e-01 -7.86282063e-01 5.34047186e-01
6.38866007e-01 4.63938892e-01 3.40955079e-01 -8.86500716e-01
-5.94316959e-01 6.20285213e-01 -2.26811439e-01 2.02149466e-01
1.06788032e-01 -1.65281892e-01 5.59392631e-01 -6.47825420e-01
3.72542262e-01 9.70683098e-01 1.35890734e+00 4.50706899e-01
-1.31928277e+00 -4.67543364e-01 -2.99499612e-02 4.10086252e-02
-1.48992288e+00 -3.58078063e-01 9.87204492e-01 -4.85683158e-02
1.23763096e+00 4.38924521e-01 7.59796739e-01 1.15540314e+00
1.97945550e-01 5.89710355e-01 8.16185415e-01 -1.65488541e-01
3.10409039e-01 -6.46423519e-01 1.78402200e-01 2.91305125e-01
-1.14666549e-02 1.30202532e-01 1.80131327e-02 -2.19520003e-01
1.15034926e+00 4.65945899e-01 3.87983806e-02 2.31192827e-01
-1.02122331e+00 8.29401970e-01 3.74434441e-01 6.94171190e-01
-6.03546560e-01 2.50770479e-01 8.59744430e-01 6.23132646e-01
6.77258909e-01 5.88232100e-01 -5.29840648e-01 -6.08332098e-01
-9.83494759e-01 3.12361300e-01 1.02627397e+00 6.20936394e-01
3.91492128e-01 7.68652856e-01 6.18781969e-02 9.72882032e-01
-1.34823188e-01 3.83717418e-01 9.24553633e-01 -9.74692345e-01
1.94338366e-01 6.51240528e-01 -1.80606946e-01 -1.27709341e+00
-8.77803266e-01 -4.74586129e-01 -1.50610006e+00 -1.57636538e-01
2.51840502e-01 -4.49428916e-01 -9.79065955e-01 1.37649393e+00
-4.94909026e-02 5.33713758e-01 1.53684467e-01 4.74004269e-01
6.21702790e-01 1.17169690e+00 -3.71640444e-01 -3.70920807e-01
6.75178230e-01 -1.00801480e+00 -8.89653683e-01 2.44842038e-01
8.99697185e-01 -6.94565654e-01 7.48638511e-01 2.70833492e-01
-1.02792418e+00 -5.58173835e-01 -8.50572526e-01 -3.14145610e-02
-3.95811945e-01 3.08251698e-02 5.80422521e-01 1.50048301e-01
-1.27789295e+00 1.07011080e+00 -1.18307376e+00 -1.10842265e-01
-1.02878727e-01 4.98017341e-01 1.69104531e-01 4.78702456e-01
-1.33825529e+00 7.23969460e-01 2.25582436e-01 4.68832880e-01
-4.36995417e-01 -8.49514425e-01 -8.93068910e-01 -2.41826326e-02
7.22649395e-02 -2.75697559e-01 1.32257295e+00 -5.76350868e-01
-1.99952662e+00 4.34751920e-02 -1.76163137e-01 -1.04873109e+00
4.72209096e-01 -5.62654585e-02 -7.51241922e-01 -2.94562340e-01
-3.79896313e-01 -3.01020686e-03 6.49562836e-01 -6.64275944e-01
-8.13613459e-02 9.04847961e-03 -1.82779983e-01 -2.78189182e-01
-6.85326308e-02 -1.61570646e-02 -2.04630494e-01 -1.11217070e+00
8.35332870e-02 -1.32029784e+00 -7.17766762e-01 -4.14905339e-01
-8.87300000e-02 -3.78153622e-01 1.11990821e+00 -7.31083512e-01
1.75238407e+00 -2.24354410e+00 -1.67889431e-01 2.49653742e-01
2.38757491e-01 6.45717800e-01 -1.19473249e-01 7.62735784e-01
-2.68102288e-01 -7.72272572e-02 -3.04681569e-01 -4.44563568e-01
-2.13699117e-01 6.25045240e-01 -6.31541610e-01 2.50339478e-01
9.52845290e-02 9.59978998e-01 -5.75507700e-01 8.93801302e-02
1.94431171e-01 8.05582821e-01 -3.47784728e-01 2.91061044e-01
-1.90544277e-01 3.43060732e-01 -1.84884399e-01 3.76004547e-01
4.68226105e-01 -3.80028814e-01 -6.65371194e-02 -1.36356866e-02
-4.22746003e-01 2.47838348e-01 -1.24585044e+00 1.38696182e+00
-7.92508483e-01 9.15165663e-01 -3.91612977e-01 -1.13412619e+00
1.26646507e+00 4.77534622e-01 6.88734651e-01 -8.00602138e-01
7.39897415e-02 4.13778007e-01 -2.78716965e-04 -3.83370310e-01
5.52255869e-01 1.40503868e-01 2.72472911e-02 5.26302099e-01
-3.17524225e-01 9.00611356e-02 3.66852790e-01 -3.60312134e-01
1.23066282e+00 5.39441174e-03 1.36235774e-01 -1.67750120e-01
5.08133054e-01 -2.05656767e-01 6.04256094e-01 7.27987885e-01
2.13508576e-01 5.69876254e-01 4.81164634e-01 -1.13655663e+00
-1.22528315e+00 -5.17390788e-01 -7.88922682e-02 9.27705109e-01
-3.00642997e-01 -5.21162271e-01 -4.32330996e-01 2.14791065e-03
-1.13900341e-01 4.81370449e-01 -5.73528826e-01 1.65481586e-03
-1.16941988e+00 -8.04163039e-01 9.21620548e-01 7.14872181e-01
4.30403054e-01 -1.32930779e+00 -5.06380141e-01 6.25665307e-01
-6.25726581e-02 -1.09709585e+00 -4.17579353e-01 3.89199585e-01
-1.25021935e+00 -5.86372554e-01 -8.13678026e-01 -6.67098284e-01
3.84151757e-01 1.06626496e-01 1.02905047e+00 1.47239650e-02
-1.45848259e-01 -1.85618505e-01 -3.17780167e-01 -2.78230578e-01
-5.93271852e-01 2.69865394e-01 2.49536827e-01 -9.96378288e-02
-7.56080300e-02 -1.05077696e+00 -3.12241852e-01 1.72449991e-01
-1.15411258e+00 -1.28093278e-02 3.40699941e-01 6.55168116e-01
6.06543779e-01 8.76461491e-02 6.34623289e-01 -7.25588024e-01
9.52427089e-01 -3.60798806e-01 -7.46798277e-01 1.18648075e-02
-6.20445848e-01 4.31342870e-01 1.44713509e+00 -8.87162089e-01
-6.69423223e-01 -1.11789882e-01 -3.98114175e-01 -7.72501051e-01
1.78638786e-01 7.83499420e-01 7.53854454e-01 -1.92443416e-01
5.56929290e-01 4.16209161e-01 1.87646523e-01 -6.75775468e-01
-5.22846654e-02 4.98701870e-01 4.91804391e-01 -4.06942219e-01
5.11017025e-01 3.33383888e-01 1.86100230e-01 -8.89466047e-01
-5.27752042e-01 -1.84612036e-01 -2.50365943e-01 5.11478297e-02
2.73523271e-01 -9.65300977e-01 -7.89475799e-01 8.32617879e-01
-1.21224260e+00 -6.06675208e-01 -4.98249352e-01 5.53666890e-01
-4.26021010e-01 3.22226882e-01 -1.14876974e+00 -8.15675914e-01
-6.72799170e-01 -8.71775270e-01 7.95536339e-01 -1.01275090e-02
-2.54085451e-01 -1.15543759e+00 4.70873088e-01 -5.15806496e-01
1.11942828e+00 5.69668412e-01 6.51006222e-01 -5.45274138e-01
-2.37947211e-01 -4.36463594e-01 -1.95942461e-01 2.89427727e-01
1.53780028e-01 1.76718533e-01 -6.50065780e-01 -1.07289001e-01
3.42710763e-01 3.83084193e-02 4.86908674e-01 4.25275385e-01
1.43705630e+00 -4.00939941e-01 1.83480397e-01 7.30100870e-01
1.45535457e+00 3.63249213e-01 8.87128055e-01 2.03944921e-01
7.30208993e-01 5.79456389e-02 -5.27531728e-02 6.86662078e-01
2.05427423e-01 4.73131478e-01 1.31194010e-01 -2.28016078e-01
7.79393911e-02 7.14813452e-03 5.40970385e-01 1.69309294e+00
-4.87753779e-01 3.89065631e-02 -9.45044756e-01 5.65395892e-01
-2.01708937e+00 -9.64884877e-01 -3.31589937e-01 1.84893382e+00
9.08161044e-01 4.40509208e-02 2.83965439e-01 3.90725583e-01
5.29432178e-01 3.82326633e-01 -6.31654322e-01 -7.66396582e-01
-8.44543576e-02 2.22083867e-01 3.25115532e-01 2.84281850e-01
-8.20936799e-01 6.31278217e-01 7.52774715e+00 7.58051693e-01
-1.54026902e+00 -3.08157593e-01 2.86665738e-01 -5.34109324e-02
1.08048931e-01 -3.86194795e-01 -6.18424296e-01 7.11296916e-01
1.65669370e+00 -3.75671118e-01 7.45556533e-01 7.73167610e-01
4.91244078e-01 4.98626709e-01 -7.81410038e-01 1.23593056e+00
-9.83526781e-02 -1.32468832e+00 9.01900232e-03 -3.36063430e-02
8.06544423e-01 3.08020055e-01 -1.29564643e-01 6.15710437e-01
2.82393545e-01 -9.82161641e-01 9.10645872e-02 8.49103570e-01
4.38372016e-01 -8.04753661e-01 8.14865768e-01 3.54664981e-01
-1.42527962e+00 1.61957890e-02 -3.85199726e-01 -5.47232509e-01
2.06557035e-01 8.65878940e-01 -3.74188662e-01 4.77937698e-01
6.20415390e-01 1.05159187e+00 -3.49104166e-01 8.95153582e-01
3.79136384e-01 6.02383256e-01 -6.80112422e-01 -9.71003845e-02
4.10336435e-01 -1.56012118e-01 3.47439528e-01 1.06112278e+00
7.03072548e-01 -7.06071705e-02 2.84696370e-02 6.84807360e-01
7.19652250e-02 2.71663256e-02 -6.44796610e-01 2.66907999e-04
1.23027958e-01 9.01523471e-01 -4.26509023e-01 -2.76256651e-01
-3.90873373e-01 7.29880035e-01 3.34232032e-01 3.98590207e-01
-1.00322521e+00 -4.40479189e-01 4.96254802e-01 -1.28395230e-01
3.67178530e-01 -5.43503702e-01 -1.89292803e-01 -1.25561118e+00
2.31481507e-01 -9.35859144e-01 1.62415549e-01 -6.48199022e-01
-1.31675303e+00 1.10445261e+00 -4.87666577e-01 -1.48984885e+00
-6.88313782e-01 -5.32750010e-01 -5.16403854e-01 7.70023882e-01
-1.25964510e+00 -5.58444083e-01 -1.55602813e-01 6.68778479e-01
5.03101051e-01 -4.26422730e-02 1.06513870e+00 4.30678904e-01
-8.18104744e-01 3.73048723e-01 5.64147353e-01 1.96498722e-01
1.83648288e-01 -1.03939176e+00 8.47078979e-01 7.35335648e-01
-1.06363103e-01 5.77011704e-01 7.86512733e-01 -3.97354037e-01
-1.43309534e+00 -1.34219146e+00 1.04948783e+00 -5.02261408e-02
9.80206847e-01 -1.77956477e-01 -1.31769264e+00 5.91287732e-01
1.35544106e-01 1.99612483e-01 4.49299723e-01 -1.43414447e-02
-2.06070229e-01 -1.07519716e-01 -6.55184567e-01 7.57915497e-01
8.16975057e-01 -7.47141004e-01 -4.46360201e-01 1.04735874e-01
9.24738526e-01 -5.92880905e-01 -1.23658383e+00 5.80941260e-01
4.36612725e-01 -8.36518645e-01 7.23331690e-01 -3.64301890e-01
3.45547706e-01 -2.34734148e-01 1.97558459e-02 -1.31596410e+00
-2.32171863e-01 -1.06030726e+00 -8.66708934e-01 7.97816694e-01
2.81712055e-01 -8.45595598e-01 4.14557576e-01 5.35174191e-01
-5.09356745e-02 -1.00658965e+00 -8.35972846e-01 -1.10507846e+00
-1.55074194e-01 -6.54335856e-01 7.98920751e-01 1.00419068e+00
-2.33131692e-01 9.36943814e-02 -6.80148721e-01 -1.25637159e-01
6.32848069e-02 1.35426834e-01 5.60370684e-01 -1.53660667e+00
-9.82408673e-02 -4.98885602e-01 -3.33070397e-01 -1.17670012e+00
6.20305091e-02 -5.70225537e-01 -2.36335039e-01 -1.18301165e+00
-5.47376096e-01 -3.54238570e-01 -3.54631454e-01 3.67315561e-01
1.78163052e-01 1.95348278e-01 4.77290787e-02 4.71345752e-01
-5.47909498e-01 6.14484727e-01 9.36277449e-01 1.61534950e-01
-7.01737404e-01 1.90693393e-01 -2.75246322e-01 7.35094309e-01
9.59298313e-01 -3.92285049e-01 -1.83617070e-01 -4.11802560e-01
3.75461787e-01 3.83637518e-01 4.68985476e-02 -1.26563609e+00
3.29642594e-01 2.37164348e-01 1.25801295e-01 -8.36592317e-01
4.33863848e-02 -9.32868958e-01 4.83527988e-01 5.41340470e-01
-3.51189077e-01 7.66349792e-01 5.51744699e-01 4.31157917e-01
-3.45494598e-01 5.78216873e-02 5.00493705e-01 1.18688479e-01
-3.79622161e-01 3.16991270e-01 -6.31039679e-01 -1.17886476e-01
6.61155820e-01 -2.98656747e-02 -1.20840676e-01 -5.24184585e-01
-8.22103858e-01 -9.86818075e-02 1.64961293e-01 3.66067767e-01
3.86932433e-01 -1.59751308e+00 -5.53974748e-01 5.18504262e-01
-3.99896353e-01 1.32734746e-01 2.88347274e-01 9.45371807e-01
-7.92147934e-01 7.54563510e-01 -2.78786514e-02 -3.83491993e-01
-8.00255001e-01 5.68179429e-01 5.47569454e-01 -8.41809034e-01
-9.57152367e-01 6.14733621e-02 -4.18722868e-01 -6.63811147e-01
3.54866773e-01 -7.42018700e-01 -1.84914172e-01 -1.11976959e-01
5.57533205e-01 8.45477283e-01 2.74172485e-01 -4.39305604e-01
1.28167905e-02 6.08714283e-01 -6.44445792e-02 3.68644863e-01
1.68874753e+00 6.45638853e-02 -4.44387138e-01 8.11812520e-01
1.49226940e+00 -2.20054626e-01 -1.01292193e+00 -3.05232227e-01
-2.04726332e-03 2.30302945e-01 -5.64549603e-02 -1.65267423e-01
-1.20645988e+00 4.67855513e-01 4.77707535e-01 8.64785790e-01
1.25559044e+00 -5.42256236e-01 1.24381387e+00 8.30285966e-01
1.72333434e-01 -1.01604676e+00 -3.05838764e-01 1.18145502e+00
9.09033597e-01 -7.56307185e-01 -3.20750207e-01 8.78625140e-02
-2.49796063e-01 1.34576344e+00 2.61070162e-01 -7.27970958e-01
9.57032919e-01 5.78082383e-01 2.12010562e-01 2.71279719e-02
-9.57372725e-01 -5.07488772e-02 1.78341642e-01 1.76754564e-01
5.84347367e-01 -1.22886039e-01 -3.01952064e-01 4.28518116e-01
-1.35254547e-01 2.54929751e-01 5.27402222e-01 7.25054979e-01
1.68195412e-01 -9.34581161e-01 -3.33086938e-01 5.19172490e-01
-5.52138388e-01 -1.42222434e-01 6.33925349e-02 7.27300584e-01
-4.37426448e-01 7.03440070e-01 2.79448748e-01 -4.35874432e-01
4.11000788e-01 1.79638769e-02 6.16996130e-03 -3.76900174e-02
-7.92824507e-01 5.44307828e-02 -1.71788782e-01 -6.80331111e-01
-4.45656657e-01 -4.05746400e-01 -1.33427823e+00 -6.29248440e-01
-1.21879898e-01 -8.94092675e-03 6.14956319e-01 1.01895106e+00
7.81663299e-01 7.72201419e-01 9.07222748e-01 -9.98572707e-01
-5.18204689e-01 -1.03926897e+00 -6.40191674e-01 1.24722764e-01
6.06981575e-01 -1.61917359e-01 -5.66118658e-01 -1.97405398e-01] | [7.082246780395508, 3.2000296115875244] |
5593fd21-1088-4f42-907b-6e911e5499b9 | palm-predicting-actions-through-language | 2306.16545 | null | https://arxiv.org/abs/2306.16545v1 | https://arxiv.org/pdf/2306.16545v1.pdf | Palm: Predicting Actions through Language Models @ Ego4D Long-Term Action Anticipation Challenge 2023 | We present Palm, a solution to the Long-Term Action Anticipation (LTA) task utilizing vision-language and large language models. Given an input video with annotated action periods, the LTA task aims to predict possible future actions. We hypothesize that an optimal solution should capture the interdependency between past and future actions, and be able to infer future actions based on the structure and dependency encoded in the past actions. Large language models have demonstrated remarkable commonsense-based reasoning ability. Inspired by that, Palm chains an image captioning model and a large language model. It predicts future actions based on frame descriptions and action labels extracted from the input videos. Our method outperforms other participants in the EGO4D LTA challenge and achieves the best performance in terms of action prediction. Our code is available at https://github.com/DanDoge/Palm | ['Xi Wang', 'Luc van Gool', 'Otmar Hilliges', 'Daoji Huang'] | 2023-06-28 | null | null | null | null | ['action-anticipation', 'image-captioning'] | ['computer-vision', 'computer-vision'] | [ 7.58775026e-02 3.98115605e-01 -3.76920640e-01 -6.36610210e-01
-4.05625671e-01 -2.75831550e-01 8.95638227e-01 -2.67884701e-01
-2.45458841e-01 4.59919184e-01 1.15407252e+00 5.69353849e-02
2.28048131e-01 -4.24018592e-01 -8.56084406e-01 -5.74542545e-02
-2.63726383e-01 3.72565180e-01 -3.84506285e-02 1.58375964e-01
2.01926693e-01 2.37884626e-01 -1.45127940e+00 9.95284796e-01
1.80289984e-01 9.07094359e-01 5.99357307e-01 8.46111178e-01
1.45374253e-01 1.85596919e+00 1.98427781e-01 -4.08081949e-01
1.96956038e-01 -4.25148398e-01 -1.26375639e+00 3.15420210e-01
2.48922750e-01 -6.92113698e-01 -9.97117043e-01 6.60512567e-01
1.40819311e-01 1.29163995e-01 2.91705340e-01 -1.55401337e+00
-7.89292336e-01 7.97058821e-01 -1.28613606e-01 1.85404524e-01
8.72585654e-01 9.53560650e-01 1.11662209e+00 -6.39114916e-01
9.75707293e-01 1.64554465e+00 4.56883132e-01 8.86618137e-01
-9.97463942e-01 -1.10049278e-01 5.09492338e-01 8.36823225e-01
-9.50809598e-01 -7.72775710e-01 6.53240621e-01 -4.91854131e-01
1.51103127e+00 -1.82578832e-01 1.02984774e+00 1.52142370e+00
3.20432305e-01 1.42401087e+00 8.58698368e-01 -2.09522590e-01
3.22014652e-02 -4.72373486e-01 -1.03098452e-02 9.40383375e-01
-4.97905016e-01 8.54617208e-02 -9.15494442e-01 3.98612730e-02
6.28579199e-01 7.95709435e-03 -1.75080411e-02 -1.89282924e-01
-1.86188960e+00 3.63977194e-01 3.25547814e-01 7.92883039e-02
-9.74163592e-01 8.30098271e-01 4.20424312e-01 7.29419217e-02
2.60041654e-01 3.53619188e-01 -4.34737176e-01 -5.49557745e-01
-4.92025495e-01 3.11809093e-01 5.61425984e-01 8.76096070e-01
5.28917730e-01 -3.68654877e-01 -6.07095420e-01 2.95005798e-01
4.99091238e-01 5.70260584e-01 4.25647706e-01 -1.80798507e+00
3.88289720e-01 3.53832453e-01 4.30083543e-01 -8.89695048e-01
-3.09326649e-01 2.90975988e-01 -1.99318528e-01 -4.20949683e-02
4.14367795e-01 8.20633024e-02 -8.44136894e-01 1.87985313e+00
3.30478437e-02 5.25716484e-01 3.33216816e-01 7.51583815e-01
4.44238991e-01 7.33680665e-01 5.68064392e-01 -1.82068124e-01
1.14995706e+00 -1.16585445e+00 -7.96031296e-01 -7.87921369e-01
8.38198304e-01 -5.25068581e-01 1.02621925e+00 2.65915275e-01
-1.23426592e+00 -6.41013920e-01 -5.21309257e-01 -3.93957317e-01
-3.86706591e-02 3.39588344e-01 7.77423978e-01 -3.78778905e-01
-1.25536823e+00 5.39082885e-01 -1.13380146e+00 -6.52769148e-01
6.49812102e-01 6.18490996e-03 -3.13204050e-01 -2.45436102e-01
-9.93423045e-01 9.63615477e-01 5.08040249e-01 2.16010109e-01
-1.26265585e+00 -2.68711567e-01 -9.47191060e-01 -2.03063980e-01
4.17684823e-01 -8.69817257e-01 1.65643549e+00 -1.11897147e+00
-1.33055520e+00 9.95414019e-01 -5.53508461e-01 -1.14598918e+00
5.65690279e-01 -5.01614869e-01 -5.09096533e-02 2.91751057e-01
1.01792656e-01 1.21102595e+00 6.73480690e-01 -5.83791256e-01
-5.44733942e-01 -3.21134925e-01 4.49297965e-01 3.12456459e-01
7.71500096e-02 -2.54722815e-02 -3.23104799e-01 -2.67537206e-01
-1.31892338e-01 -1.24886191e+00 -2.41429746e-01 2.53303409e-01
-1.55493587e-01 -5.36436915e-01 6.52621627e-01 -7.93282688e-01
8.18757057e-01 -2.09585261e+00 1.81301862e-01 -4.95380968e-01
1.86089706e-02 2.05330253e-02 -4.80467796e-01 4.76936609e-01
-6.79309592e-02 -7.26109296e-02 3.45264852e-01 -3.24853718e-01
2.07785636e-01 3.82127643e-01 -6.23846173e-01 3.27153534e-01
1.13348939e-01 1.32604051e+00 -1.32421184e+00 -6.66539967e-01
6.07179940e-01 2.74499625e-01 -3.36042821e-01 5.65392435e-01
-9.60715473e-01 4.48608279e-01 -6.21986032e-01 5.22397339e-01
-3.40277590e-02 -5.75133324e-01 2.35480905e-01 -3.07416499e-01
-8.31740629e-03 3.63449901e-01 -5.19360721e-01 2.09687495e+00
-5.56117058e-01 9.73655462e-01 -4.83548164e-01 -6.98212683e-01
4.64460611e-01 5.31169355e-01 7.73921072e-01 -5.33359468e-01
-1.47501051e-01 -1.80009007e-01 -1.73093989e-01 -1.27746701e+00
5.40049709e-02 -9.23175812e-02 4.45678681e-02 5.73160172e-01
5.40319085e-02 -2.77746022e-02 2.15890795e-01 5.25824070e-01
1.24415815e+00 9.15693998e-01 4.26222712e-01 3.71427029e-01
3.88663173e-01 1.40773088e-01 4.77641284e-01 7.31199205e-01
-6.15271986e-01 3.19551021e-01 5.82167983e-01 -9.00199115e-01
-8.56848061e-01 -8.11637640e-01 6.90014720e-01 1.04604173e+00
-4.02070470e-02 -8.07457566e-01 -3.24036002e-01 -5.72695255e-01
-1.98992476e-01 1.23594737e+00 -6.89367175e-01 -1.24619760e-01
-5.42004228e-01 -6.26138747e-02 3.36638987e-01 7.81172752e-01
6.10592008e-01 -1.71989405e+00 -8.96416306e-01 -7.14461785e-03
-8.42227817e-01 -1.64465451e+00 -3.69871497e-01 -5.45610905e-01
-6.99471951e-01 -1.12697339e+00 -2.90398121e-01 -5.18204451e-01
3.14764440e-01 4.34996448e-02 1.22995079e+00 -1.61110654e-01
-1.55143753e-01 1.05037034e+00 -3.87539715e-01 -2.62605309e-01
-6.00851715e-01 -7.10955262e-01 1.24907956e-01 6.39306307e-02
5.69868386e-01 -5.22005856e-01 -6.72301650e-01 -8.07771385e-02
-4.00581360e-01 8.50691319e-01 6.81587934e-01 4.12814230e-01
5.65258443e-01 -4.35929239e-01 1.60628051e-01 -2.31282696e-01
1.36043370e-01 -4.27476764e-01 -2.72432834e-01 5.19634485e-01
-3.10766190e-01 1.29727498e-01 3.00117075e-01 -3.29275966e-01
-1.39735186e+00 5.75347126e-01 3.45927626e-02 -6.72963738e-01
-4.98523116e-01 3.31243485e-01 -3.83755122e-03 4.63447362e-01
3.46459508e-01 3.04569364e-01 -9.25679654e-02 -1.81721807e-01
7.35545099e-01 2.50178993e-01 5.88715136e-01 -4.73954290e-01
7.68240243e-02 6.91921711e-01 -8.27966481e-02 -5.22972643e-01
-1.34546089e+00 -3.51748973e-01 -8.68582964e-01 -7.22054839e-01
1.25939786e+00 -1.13463366e+00 -9.29323018e-01 5.31906903e-01
-1.55528915e+00 -8.85337055e-01 -4.23214346e-01 6.29720807e-01
-1.41202819e+00 3.53829086e-01 -6.46707237e-01 -7.50940323e-01
-1.08465709e-01 -9.41138804e-01 9.25110459e-01 5.59010766e-02
-5.60300887e-01 -9.52216864e-01 -6.16267882e-02 6.85395598e-01
-5.72240911e-02 2.21192643e-01 5.34333825e-01 -2.89175659e-01
-8.99891555e-01 -1.09930359e-01 -1.23869888e-01 2.19526142e-01
-4.34060721e-03 -5.69436513e-02 -6.71845257e-01 1.39874175e-01
-1.62141711e-01 -9.51523840e-01 1.01135337e+00 6.24031901e-01
1.27811754e+00 -4.78843182e-01 -3.67272019e-01 2.86329299e-01
1.06765747e+00 -4.62193750e-02 8.41130197e-01 1.85087532e-01
6.44222617e-01 5.99988461e-01 9.81817603e-01 6.35095298e-01
6.79907382e-01 5.02068162e-01 5.67989111e-01 6.65881753e-01
-5.27723312e-01 -7.67620444e-01 8.75484109e-01 1.65698752e-01
-2.30541706e-01 -4.44445491e-01 -1.17069590e+00 7.62406111e-01
-2.31094432e+00 -1.53340268e+00 1.18420972e-02 1.62996686e+00
5.18384397e-01 3.85345109e-02 -4.54088449e-02 -7.27672279e-01
3.66757691e-01 5.99696457e-01 -8.42469990e-01 -2.46612117e-01
3.62484269e-02 -4.75318432e-01 8.36717337e-02 5.21823406e-01
-1.05818594e+00 1.25537097e+00 6.47819042e+00 4.09022033e-01
-7.27672458e-01 8.89516547e-02 6.09578490e-01 -2.76171416e-01
-6.75317347e-02 3.35651964e-01 -5.26884079e-01 1.46724716e-01
1.18212998e+00 -3.13511789e-01 3.94048810e-01 6.39276147e-01
6.96518958e-01 -2.85406858e-01 -1.48061991e+00 1.11040759e+00
1.54214680e-01 -1.53960359e+00 2.18251556e-01 -1.15231603e-01
4.59428668e-01 4.99297619e-01 -2.22407311e-01 3.41215670e-01
4.74937797e-01 -8.43671501e-01 9.16137338e-01 1.10337961e+00
2.64265805e-01 -5.52915893e-02 2.98803955e-01 6.21207178e-01
-1.00655377e+00 -4.92038429e-01 -5.46056181e-02 -5.80472231e-01
6.50462091e-01 1.01666845e-01 -8.76645386e-01 4.71655391e-02
4.19670016e-01 1.55594528e+00 -5.14442861e-01 5.67242503e-01
-5.91296852e-01 4.80594367e-01 5.52799851e-02 1.09935045e-01
2.95856297e-01 -9.79017615e-02 5.81088305e-01 9.85585093e-01
2.10241035e-01 4.98923570e-01 3.26773643e-01 7.91902006e-01
-8.68876651e-03 -3.62552226e-01 -8.69093418e-01 -6.10994935e-01
-2.42894050e-02 7.70823956e-01 -3.53479892e-01 -5.85927069e-01
-5.24683774e-01 1.22427595e+00 3.86302054e-01 3.88134867e-01
-1.02730215e+00 5.47223985e-01 7.49535561e-01 -1.43424999e-02
1.67182416e-01 -3.90367001e-01 8.03628638e-02 -1.41604221e+00
5.17751127e-02 -6.12371385e-01 5.22666812e-01 -1.62925279e+00
-8.87979686e-01 3.42651635e-01 3.09552774e-02 -1.12954164e+00
-6.76570356e-01 -4.93980676e-01 -4.92174298e-01 3.12345266e-01
-1.17591310e+00 -1.67447054e+00 -2.69075722e-01 4.67164248e-01
1.17547476e+00 1.84249625e-01 7.28197694e-01 -3.98621112e-01
-2.03163490e-01 -3.53967488e-01 -6.06345832e-01 1.62828848e-01
4.80801523e-01 -9.01597261e-01 5.88484585e-01 8.10517371e-01
4.94117558e-01 1.39415190e-01 9.19189870e-01 -9.04743135e-01
-1.52830029e+00 -1.03589749e+00 1.27909648e+00 -8.90796423e-01
9.19016540e-01 2.29548011e-02 -4.81012404e-01 1.46658397e+00
3.66206318e-01 1.44291818e-01 3.52719218e-01 -8.12410861e-02
-2.77724087e-01 2.46689543e-01 -7.34285057e-01 7.47588277e-01
1.49452722e+00 -7.65263259e-01 -8.93908739e-01 9.39990938e-01
7.51264751e-01 -2.72399604e-01 -5.97832441e-01 1.22648098e-01
7.96818435e-01 -9.83519793e-01 1.08055305e+00 -1.00909352e+00
8.73367488e-01 -8.47072676e-02 -1.80665717e-01 -8.19236219e-01
-4.25668836e-01 -6.11222446e-01 -4.87853140e-01 8.23478281e-01
2.11246967e-01 -6.75611794e-02 6.11640751e-01 1.12266314e+00
3.48299332e-02 -5.64564884e-01 -6.74461603e-01 -8.09339881e-01
-4.73230213e-01 -7.68687069e-01 2.37606570e-01 5.33822477e-01
1.53113768e-01 2.10507065e-01 -6.90873325e-01 1.03295743e-01
5.43233991e-01 3.32946509e-01 7.15355158e-01 -6.49948478e-01
-2.83776283e-01 -3.98436673e-02 -4.39511150e-01 -1.24314570e+00
7.88907707e-01 -7.32499897e-01 2.30478346e-01 -1.91729081e+00
3.74934644e-01 1.85309663e-01 -2.08990380e-01 9.86952364e-01
2.04579785e-01 -1.82548482e-02 6.26853764e-01 2.44186535e-01
-1.21278763e+00 7.09177196e-01 1.11332786e+00 -1.54110745e-01
-8.87807906e-02 -1.71252880e-02 -1.72097117e-01 1.19398189e+00
8.21920097e-01 -2.82481730e-01 -4.72914070e-01 -8.39258075e-01
4.14446026e-01 5.23810506e-01 8.12665462e-01 -1.03587365e+00
2.32332245e-01 -6.73503637e-01 2.72033185e-01 -4.89666164e-01
7.27383673e-01 -5.37650883e-01 8.11809897e-02 6.70736015e-01
-9.82287705e-01 -8.68990198e-02 1.61628556e-02 8.28320801e-01
-1.22253783e-02 2.06532687e-01 3.44265491e-01 -6.03720486e-01
-1.63947320e+00 3.92711252e-01 -5.33589482e-01 -1.45865843e-01
1.16897559e+00 3.34890522e-02 -2.92275339e-01 -7.58970261e-01
-1.20256138e+00 5.34504831e-01 4.35349613e-01 7.33097911e-01
9.22823787e-01 -1.30401957e+00 -7.04331577e-01 -2.41227135e-01
2.97083318e-01 -5.52495182e-01 4.19841170e-01 9.02796209e-01
-2.57174134e-01 5.17874300e-01 -4.25983638e-01 -3.45198035e-01
-1.29396451e+00 7.55847692e-01 2.92309642e-01 -1.23089761e-01
-8.60793889e-01 9.28960145e-01 1.66499838e-01 1.36925712e-01
1.38863504e-01 -3.22226256e-01 -2.00219169e-01 -3.02361906e-01
8.34936678e-01 2.60661036e-01 -8.06085885e-01 -8.89528275e-01
-4.58068758e-01 1.98901534e-01 1.86371371e-01 -2.53252864e-01
1.36736000e+00 -5.70121408e-01 -1.03828132e-01 6.79818153e-01
9.71743524e-01 -6.11950815e-01 -1.79721618e+00 -3.72797817e-01
1.10105604e-01 -5.93837500e-01 -9.82047431e-03 -9.91995990e-01
-6.94540024e-01 8.38578284e-01 3.31559718e-01 -3.76834661e-01
1.04811406e+00 4.86859083e-01 9.68100727e-01 6.73142374e-01
5.04415691e-01 -1.00430572e+00 5.46630442e-01 6.06232047e-01
1.16050434e+00 -1.47117543e+00 5.87747134e-02 8.60570557e-03
-1.10475755e+00 1.19042540e+00 7.54361689e-01 -1.47864213e-02
4.40711558e-01 -1.74297050e-01 -4.78332117e-02 -3.04757595e-01
-1.52139020e+00 -3.74808073e-01 1.80514216e-01 5.72737932e-01
2.57091135e-01 1.63852900e-01 7.56671578e-02 1.89428657e-01
1.73439488e-01 7.13947237e-01 5.34922838e-01 6.87458754e-01
-3.33527267e-01 -7.21365750e-01 5.99027686e-02 2.30280414e-01
-8.37563798e-02 1.41390607e-01 -5.15083849e-01 2.86494136e-01
1.75870344e-01 9.65735555e-01 1.42030314e-01 -2.68751860e-01
-2.71327496e-02 4.10003483e-01 7.38598406e-01 -4.59519207e-01
1.04756616e-01 -2.88837343e-01 4.13474381e-01 -1.43974304e+00
-9.67041373e-01 -9.52347219e-01 -1.44779181e+00 -7.93194473e-02
5.00208795e-01 -3.42751026e-01 3.40420038e-01 1.20897448e+00
5.90278864e-01 1.64342746e-01 1.65421575e-01 -7.63157248e-01
-2.87708193e-01 -8.29062700e-01 -7.79204667e-02 6.21317387e-01
3.87812227e-01 -2.91196704e-01 -8.85977894e-02 7.51137733e-01] | [8.119029998779297, 0.5043160915374756] |
221be931-42d5-4a1e-b707-434d02a33223 | discovering-new-intents-via-constrained-deep | 1911.08891 | null | https://arxiv.org/abs/1911.08891v1 | https://arxiv.org/pdf/1911.08891v1.pdf | Discovering New Intents via Constrained Deep Adaptive Clustering with Cluster Refinement | Identifying new user intents is an essential task in the dialogue system. However, it is hard to get satisfying clustering results since the definition of intents is strongly guided by prior knowledge. Existing methods incorporate prior knowledge by intensive feature engineering, which not only leads to overfitting but also makes it sensitive to the number of clusters. In this paper, we propose constrained deep adaptive clustering with cluster refinement (CDAC+), an end-to-end clustering method that can naturally incorporate pairwise constraints as prior knowledge to guide the clustering process. Moreover, we refine the clusters by forcing the model to learn from the high confidence assignments. After eliminating low confidence assignments, our approach is surprisingly insensitive to the number of clusters. Experimental results on the three benchmark datasets show that our method can yield significant improvements over strong baselines. | ['Hanlei Zhang', 'Ting-En Lin', 'Hua Xu'] | 2019-11-20 | null | null | null | null | ['text-clustering', 'open-intent-discovery', 'short-text-clustering'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [-2.42393032e-01 5.71920127e-02 8.75244513e-02 -6.78868532e-01
-7.54394352e-01 -8.84419620e-01 6.88163459e-01 1.56690761e-01
-4.53435302e-01 3.97558302e-01 5.56385040e-01 -4.19315584e-02
5.92301637e-02 -2.74032027e-01 -2.62080729e-01 -5.04064858e-01
1.60907924e-01 8.42439950e-01 1.63172185e-01 -2.72733837e-01
3.23728055e-01 -1.58786982e-01 -1.20908070e+00 4.85337526e-01
1.14317596e+00 5.37508786e-01 1.48016423e-01 5.02334595e-01
6.20625094e-02 6.16298378e-01 -5.54214299e-01 -4.31035310e-01
2.11370140e-01 -5.46018422e-01 -1.05037487e+00 2.22191066e-01
-4.35291193e-02 -3.19623470e-01 1.46504894e-01 1.00671577e+00
2.97884464e-01 4.25609112e-01 5.46386123e-01 -1.22180617e+00
-3.19769025e-01 7.72201836e-01 -6.08522296e-01 -3.73342186e-01
2.86221355e-01 1.84362933e-01 1.40248895e+00 -8.86886239e-01
3.61003280e-01 1.25453913e+00 8.33938241e-01 5.23045778e-01
-1.41114604e+00 -5.04678905e-01 3.86406958e-01 6.64744750e-02
-1.59122300e+00 -5.83828509e-01 8.63806307e-01 -5.69886208e-01
6.51743352e-01 3.57363671e-01 3.03816289e-01 7.99046040e-01
-6.82227790e-01 8.24103177e-01 7.74007261e-01 -3.94262254e-01
4.66310740e-01 3.15294832e-01 5.19246757e-01 5.69157064e-01
-1.70774490e-01 -4.64176536e-01 -2.47384995e-01 -3.72725517e-01
3.26231271e-01 1.64721105e-02 -3.02231193e-01 -5.43243945e-01
-1.02007055e+00 1.08834207e+00 3.62251341e-01 1.59620851e-01
-1.39446497e-01 -9.27563757e-02 3.05177957e-01 1.19537758e-02
3.36007893e-01 8.15364182e-01 -6.89938486e-01 -4.51656163e-01
-9.20487404e-01 2.99371183e-01 9.60653722e-01 9.19979751e-01
9.19670820e-01 -6.21667087e-01 -3.13628733e-01 9.86375391e-01
3.03589046e-01 -1.70954451e-01 3.67004871e-01 -1.35552716e+00
3.41687560e-01 9.04433846e-01 3.49383712e-01 -9.94434476e-01
-5.75902700e-01 -2.13227451e-01 -7.94471383e-01 -1.28253296e-01
4.47841227e-01 -6.29196346e-01 -6.95107639e-01 1.94803941e+00
3.61830354e-01 3.92472465e-03 -1.50412604e-01 1.04289913e+00
4.91167665e-01 4.28604960e-01 -2.21578389e-01 -3.28475952e-01
1.03259027e+00 -1.10081613e+00 -7.99612820e-01 -1.49909094e-01
7.92440474e-01 -7.06636846e-01 1.36533165e+00 3.84302467e-01
-7.93311119e-01 -4.43927586e-01 -7.44791389e-01 -5.24773933e-02
5.41147254e-02 5.29069304e-02 6.51048362e-01 6.67061090e-01
-1.07823610e+00 3.42564046e-01 -9.71445799e-01 -2.91627288e-01
9.37480181e-02 6.20579123e-01 -1.28322348e-01 1.41917644e-02
-9.04968262e-01 5.70491672e-01 5.50164282e-01 -1.47271290e-01
-2.30002657e-01 -4.97064412e-01 -6.95919931e-01 1.96237564e-01
6.19955480e-01 -5.66082954e-01 1.46096599e+00 -8.17698956e-01
-1.69949007e+00 2.65839756e-01 -3.91898513e-01 -2.48254865e-01
5.28929889e-01 -2.78733969e-01 2.45392621e-01 -7.05721453e-02
2.61505339e-02 8.44784319e-01 4.97801065e-01 -1.38225389e+00
-6.90198421e-01 -1.03639044e-01 6.34276494e-02 6.29421234e-01
-5.59792280e-01 -6.00741319e-02 -1.04602373e+00 -1.74825743e-01
-2.26693023e-02 -1.17059982e+00 -5.81275642e-01 -5.12347460e-01
-6.13853812e-01 -4.71164435e-01 7.96267092e-01 -4.09364551e-01
1.29658425e+00 -2.30396652e+00 3.77693504e-01 3.94423872e-01
4.82833624e-01 1.00835755e-01 8.52574408e-02 3.25621724e-01
2.53224343e-01 2.32321247e-01 -2.74463058e-01 -6.47445858e-01
3.53438586e-01 7.72124380e-02 1.31326661e-01 2.35927880e-01
-9.41256806e-03 6.95174813e-01 -9.27153707e-01 -4.44146127e-01
2.13341594e-01 3.36240172e-01 -1.26316941e+00 4.54781592e-01
-4.16234910e-01 5.39554060e-01 -2.31658325e-01 4.52766307e-02
6.48713112e-01 -6.47197008e-01 5.33470869e-01 -1.99551880e-01
-4.87581976e-02 4.23112541e-01 -1.20661223e+00 1.81545019e+00
-3.12643170e-01 3.06670427e-01 2.09195554e-01 -7.26276815e-01
6.62256360e-01 2.71136999e-01 5.66320598e-01 2.34268699e-02
6.20584488e-02 -2.85483569e-01 2.08977953e-01 -2.01001123e-01
7.04594135e-01 2.58227475e-02 -2.97688782e-01 6.22506678e-01
-5.25463223e-02 -1.47177093e-02 1.39949635e-01 5.57219565e-01
9.35832798e-01 -2.93924242e-01 1.51815996e-01 -1.85355112e-01
1.66719273e-01 4.17753719e-02 9.15901065e-01 6.98372066e-01
-8.72710124e-02 8.27037692e-01 6.28503501e-01 -1.56652376e-01
-8.06143999e-01 -6.33457363e-01 3.06631103e-02 1.25196481e+00
6.40497357e-02 -1.01216483e+00 -9.06686366e-01 -9.92185712e-01
-3.53869081e-01 6.63652837e-01 -5.65112114e-01 -1.67802930e-01
-2.67538100e-01 -9.64028776e-01 3.19821954e-01 4.27597731e-01
3.27916831e-01 -5.95835626e-01 -2.26360276e-01 2.64555246e-01
-6.24552310e-01 -1.20434463e+00 -9.10218239e-01 2.29140103e-01
-3.36203784e-01 -1.06483853e+00 -2.82779336e-01 -7.51136065e-01
7.46968150e-01 2.22748503e-01 9.91527855e-01 4.95784581e-01
1.03413440e-01 1.62792608e-01 -5.12290955e-01 -5.15434444e-02
-4.35036033e-01 3.56216282e-01 1.23655155e-01 1.79615721e-01
6.88440025e-01 -4.10151035e-01 -6.00486934e-01 5.05194485e-01
-5.06372929e-01 1.76542148e-01 2.38797218e-01 8.83811772e-01
9.41377357e-02 3.16518426e-01 3.81196290e-01 -1.13784778e+00
8.10610116e-01 -3.07523787e-01 -3.95293176e-01 9.23359916e-02
-5.89941263e-01 1.69502392e-01 7.16800869e-01 -2.14374244e-01
-1.06244087e+00 5.68616629e-01 -1.02044225e-01 -3.11078578e-01
-3.06867272e-01 6.24718428e-01 -2.99452841e-01 2.90480137e-01
6.04657412e-01 -1.44095749e-01 -7.35344961e-02 -6.08128786e-01
6.60620332e-01 8.24650288e-01 5.41439772e-01 -6.18682921e-01
6.69531822e-01 2.53302544e-01 -6.67294204e-01 -8.25129807e-01
-9.12471294e-01 -9.17892814e-01 -1.19865537e+00 9.15009528e-02
8.47110629e-01 -9.65829074e-01 -8.59842837e-01 1.31256193e-01
-1.17158842e+00 -5.52829027e-01 2.04900414e-01 4.02752906e-01
-2.58111537e-01 7.28946924e-01 -6.38761342e-01 -7.32786477e-01
-2.02471614e-01 -1.24190962e+00 7.30642200e-01 1.41220689e-01
-8.11507642e-01 -1.06402457e+00 7.16489032e-02 5.44080973e-01
1.60435870e-01 9.74166580e-03 7.25680053e-01 -9.99336243e-01
-2.95426637e-01 -1.15339339e-01 -9.86234397e-02 2.32110739e-01
2.77744561e-01 1.57655373e-01 -7.96971083e-01 -3.17278475e-01
-9.39269960e-02 -4.12129015e-01 7.72926450e-01 1.04736783e-01
1.02662337e+00 -5.96567154e-01 -4.41098481e-01 4.71117824e-01
8.01838934e-01 -8.56068358e-03 3.57448637e-01 -8.65249485e-02
1.01695442e+00 6.60505414e-01 6.11477017e-01 7.78340518e-01
8.58500481e-01 8.65677357e-01 9.82807651e-02 -2.55247205e-01
2.64694065e-01 -5.10979518e-02 2.06996247e-01 1.06004012e+00
2.15807840e-01 -7.96326064e-03 -1.08240068e+00 4.68186915e-01
-2.14060378e+00 -8.16428602e-01 -2.54282355e-01 2.06075025e+00
1.39785624e+00 1.17957115e-01 4.78547812e-01 3.62497792e-02
7.96007872e-01 -2.76512206e-01 -3.82143110e-01 -8.64953846e-02
3.62053245e-01 -8.23225975e-02 8.23878963e-03 9.10730422e-01
-1.13765574e+00 1.27883124e+00 5.85904789e+00 4.24961656e-01
-7.05497026e-01 5.32050319e-02 5.66515505e-01 -1.45657569e-01
-1.40918717e-01 2.21124023e-01 -7.47994423e-01 5.23550570e-01
5.59953392e-01 7.67127499e-02 6.04707003e-01 7.59304702e-01
3.45404506e-01 1.66910589e-02 -1.26907420e+00 8.39406550e-01
4.45705578e-02 -9.96242642e-01 -2.57883430e-01 1.72913112e-02
7.94263303e-01 9.56142098e-02 -2.07451850e-01 4.37986612e-01
1.15446138e+00 -8.86811256e-01 1.74236000e-01 2.09733680e-01
3.84390205e-01 -1.12869298e+00 6.28104031e-01 7.07359135e-01
-8.72046232e-01 -1.47622880e-02 -3.16497445e-01 -1.83283105e-01
9.72047355e-03 6.03338122e-01 -1.32118452e+00 7.72334561e-02
6.68244600e-01 5.90982199e-01 -6.68436110e-01 8.81324053e-01
-1.62678048e-01 7.75907636e-01 -5.72827816e-01 -7.59008452e-02
2.65258729e-01 -2.38261625e-01 2.73318142e-01 1.41360950e+00
-2.03211531e-01 3.27789217e-01 8.09641480e-01 8.83696198e-01
-1.72045574e-01 4.32917066e-02 -3.38833958e-01 1.91194564e-01
7.84726977e-01 1.50975192e+00 -8.08735073e-01 -2.89918453e-01
-3.11560184e-01 1.12203002e+00 6.28035545e-01 2.46876225e-01
-5.57926297e-01 -3.68211657e-01 6.18207395e-01 -9.98501778e-02
3.53386819e-01 -3.34706903e-01 -4.11735445e-01 -1.23250139e+00
-2.44583368e-01 -1.01548266e+00 4.03396100e-01 -2.77442634e-01
-1.29727638e+00 3.97395521e-01 -1.85162991e-01 -8.33898485e-01
-4.19414341e-01 -6.05634600e-02 -7.08889425e-01 6.61182523e-01
-1.02306545e+00 -8.64028931e-01 -3.46808821e-01 6.09995782e-01
5.05436361e-01 8.44875127e-02 8.55123758e-01 1.62034377e-01
-6.97708786e-01 9.53137457e-01 6.87111095e-02 4.43866372e-01
1.09156609e+00 -1.57300770e+00 3.15189838e-01 7.23701358e-01
5.52831627e-02 1.02772439e+00 8.41275573e-01 -5.44655859e-01
-8.87458801e-01 -1.14669096e+00 7.80894578e-01 -8.87158096e-01
4.03658807e-01 -8.40839505e-01 -1.19252801e+00 7.40170896e-01
4.13161218e-01 -4.11644787e-01 1.17293191e+00 8.58452976e-01
-3.42940629e-01 2.40380466e-01 -8.87869477e-01 7.38586307e-01
7.77531385e-01 -2.97804058e-01 -4.35128182e-01 5.29452145e-01
9.41907763e-01 -3.36252034e-01 -8.87295067e-01 2.41586775e-01
1.44904107e-01 -9.61134911e-01 7.01678693e-01 -4.20044422e-01
1.21031903e-01 -4.93407726e-01 1.14996647e-02 -1.44607306e+00
-5.85425019e-01 -9.51324999e-01 -9.82453022e-03 1.51038384e+00
6.74405456e-01 -2.29276679e-02 7.68447816e-01 1.27621663e+00
-2.24556252e-01 -3.45422328e-01 -5.36318719e-01 -4.00823534e-01
9.18367580e-02 -2.94718564e-01 6.57769024e-01 1.53823268e+00
6.88846767e-01 9.32660758e-01 -5.33950865e-01 2.48822004e-01
4.98693615e-01 8.64447504e-02 1.20570576e+00 -1.48029459e+00
-5.85403740e-01 -5.40754378e-01 6.62333593e-02 -1.17846775e+00
1.79841489e-01 -7.61087596e-01 4.39742565e-01 -1.43273652e+00
3.84772986e-01 -5.78653336e-01 -8.99911076e-02 7.83326805e-01
-6.98617399e-01 4.05952409e-02 1.96327537e-01 2.94890612e-01
-1.35483158e+00 6.09140694e-01 6.41422987e-01 9.28355902e-02
-7.48873055e-01 1.81944266e-01 -1.07112777e+00 9.16185915e-01
9.56475258e-01 -4.56116647e-01 -5.01620770e-01 -1.35738492e-01
1.65906653e-01 -2.87872732e-01 -7.02975020e-02 -6.95095360e-01
5.52322090e-01 -1.71968907e-01 2.27249429e-01 -6.70450032e-01
3.51703614e-01 -6.63144827e-01 -3.04665357e-01 1.46877542e-01
-6.20008588e-01 -5.90836182e-02 -8.64575431e-02 5.38284957e-01
-8.88232738e-02 -1.05792858e-01 8.22563887e-01 -1.31643385e-01
-4.49683964e-01 6.93227723e-02 -4.14235801e-01 2.75535494e-01
7.39575267e-01 8.53502750e-02 1.03788354e-01 -6.36159599e-01
-6.80391550e-01 7.97125459e-01 7.14817941e-01 4.59317595e-01
2.59927571e-01 -1.26146734e+00 -6.40842259e-01 -6.99123740e-02
1.17077783e-01 4.24461365e-01 -2.06232622e-01 7.64063656e-01
-5.41184060e-02 2.25704163e-01 4.34908420e-01 -7.59013951e-01
-1.44110072e+00 6.60905421e-01 1.70442507e-01 -2.51558781e-01
-4.34316069e-01 7.98278987e-01 2.64824688e-01 -9.01768565e-01
6.33270681e-01 -3.32840234e-02 -3.19396406e-01 -6.85768202e-02
3.66069794e-01 4.67451736e-02 -1.13364637e-01 -5.05118132e-01
-3.65472376e-01 2.00401634e-01 -6.87140465e-01 -1.74489185e-01
1.31414664e+00 -3.93097490e-01 -3.94205116e-02 4.05401319e-01
1.24773204e+00 4.85742232e-03 -1.44051933e+00 -3.20734560e-01
3.57876807e-01 -3.14394265e-01 -7.49823302e-02 -8.83843005e-01
-7.91931450e-01 7.76256859e-01 1.94362924e-02 2.03549802e-01
7.58695185e-01 -1.63348153e-01 6.69718981e-01 7.36517668e-01
2.15347651e-02 -1.39979720e+00 3.89393538e-01 7.85348356e-01
4.81674373e-01 -1.54333007e+00 -1.33659691e-01 -3.56868833e-01
-1.10343695e+00 6.93912625e-01 9.10192370e-01 1.65557355e-01
6.20442271e-01 2.62231827e-01 2.57703722e-01 -1.62413061e-01
-9.85177457e-01 -1.27070248e-01 1.98929161e-01 5.10312974e-01
6.10702813e-01 -4.99407109e-03 -1.51748821e-01 9.23526466e-01
-3.57741445e-01 -3.40700567e-01 6.00588381e-01 7.37167060e-01
-4.38399523e-01 -1.14876401e+00 -2.34215662e-01 3.96454126e-01
-3.91681761e-01 -1.96668088e-01 -1.00037181e+00 4.77333128e-01
-4.61002216e-02 1.33335018e+00 -1.24735748e-02 -6.81427181e-01
1.80351183e-01 9.84536409e-02 -6.39966130e-02 -1.00525391e+00
-7.56713212e-01 3.86631161e-01 7.75831640e-02 -3.02020550e-01
-1.69388562e-01 -8.20798755e-01 -1.57046878e+00 -3.68553400e-01
-7.62353241e-01 7.19132662e-01 2.88337022e-01 9.52969909e-01
6.40183032e-01 2.84870476e-01 9.03047323e-01 -6.11793578e-01
-4.99440163e-01 -1.10668051e+00 -1.64566487e-01 7.62915075e-01
2.37158164e-01 -4.26076561e-01 -3.07945400e-01 7.81693235e-02] | [12.271068572998047, 7.408139705657959] |
4246c238-d065-4c21-9843-f454d9c4bf18 | solar-active-region-magnetogram-image-dataset | 2305.09492 | null | https://arxiv.org/abs/2305.09492v2 | https://arxiv.org/pdf/2305.09492v2.pdf | Solar Active Region Magnetogram Image Dataset for Studies of Space Weather | In this dataset we provide a comprehensive collection of magnetograms (images quantifying the strength of the magnetic field) from the National Aeronautics and Space Administration's (NASA's) Solar Dynamics Observatory (SDO). The dataset incorporates data from three sources and provides SDO Helioseismic and Magnetic Imager (HMI) magnetograms of solar active regions (regions of large magnetic flux, generally the source of eruptive events) as well as labels of corresponding flaring activity. This dataset will be useful for image analysis or solar physics research related to magnetic structure, its evolution over time, and its relation to solar flares. The dataset will be of interest to those researchers investigating automated solar flare prediction methods, including supervised and unsupervised machine learning (classical and deep), binary and multi-class classification, and regression. This dataset is a minimally processed, user configurable dataset of consistently sized images of solar active regions that can serve as a benchmark dataset for solar flare prediction research. | ['Ellery Wuest', 'Jeremy A. Grajeda', 'Ty Vincent', 'Laura E. Boucheron'] | 2023-05-16 | null | null | null | null | ['solar-flare-prediction'] | ['time-series'] | [ 3.37645859e-01 -4.31143820e-01 -3.70087802e-01 -4.45100069e-01
-1.76505372e-01 -6.14878058e-01 8.99617970e-01 4.37446125e-02
3.16919506e-01 6.99247539e-01 2.12366104e-01 -3.53502095e-01
-3.10784072e-01 -8.30200791e-01 -3.22744757e-01 -9.56620038e-01
1.95842180e-02 6.01860523e-01 -8.44943374e-02 -4.99737382e-01
6.62133992e-01 9.20156419e-01 -1.84966028e+00 -1.50374221e-02
9.20903623e-01 1.61094189e+00 2.02642635e-01 7.69848108e-01
4.12524641e-01 1.15859163e+00 -6.82770371e-01 7.94807136e-01
1.04441851e-01 -6.45930409e-01 -4.26954150e-01 3.88134681e-02
1.03524435e+00 3.32233280e-01 -2.28564665e-02 8.70495379e-01
5.29218495e-01 -8.46178755e-02 8.31683218e-01 -9.07313704e-01
-5.99759877e-01 4.46648477e-03 -7.31320977e-01 1.40776479e+00
1.91205740e-01 8.69592652e-02 6.69281602e-01 -4.19950068e-01
8.07626486e-01 7.05973327e-01 5.63161552e-01 -2.22446203e-01
-8.47445309e-01 -4.70370471e-01 -7.69355476e-01 4.44166005e-01
-8.58528078e-01 -5.87343574e-01 8.41150463e-01 -9.94973481e-01
1.29093897e+00 4.09656376e-01 8.78693759e-01 7.66719759e-01
4.39431429e-01 4.50737536e-01 1.62550449e+00 -5.25985658e-01
2.93073922e-01 1.46760985e-01 4.34289157e-01 5.72368622e-01
1.66519836e-01 5.93471169e-01 -6.81081116e-01 -1.98061809e-01
2.54933000e-01 -2.98426449e-01 -5.47160149e-01 -1.75994992e-01
-8.74868035e-01 1.01787519e+00 5.76358795e-01 5.35736084e-01
-3.90282661e-01 -3.43620211e-01 -5.22311330e-02 3.97441238e-01
1.53339311e-01 3.87130886e-01 -5.02369761e-01 3.94927524e-02
-1.04197800e+00 3.82008880e-01 5.76791227e-01 2.05905274e-01
1.00248587e+00 5.00988066e-01 2.23098546e-01 6.20758414e-01
1.35957196e-01 1.40220046e+00 8.91831338e-01 -6.83765948e-01
-8.37105587e-02 6.54544473e-01 9.43707302e-02 -1.04179490e+00
-7.74032652e-01 -1.43435493e-01 -8.74810338e-01 3.83419722e-01
-1.79551780e-01 -2.01094911e-01 -1.09295654e+00 1.10055876e+00
3.56489092e-01 -2.53472388e-01 2.62061536e-01 9.17693794e-01
1.08275652e+00 8.59171867e-01 -4.71067756e-01 -3.20459753e-01
1.28198016e+00 -6.55112147e-01 -5.09106636e-01 -4.35417920e-01
4.94394332e-01 -4.65357661e-01 5.49229801e-01 2.54232943e-01
-6.76371992e-01 -4.29515511e-01 -1.22595465e+00 2.07882226e-01
-8.46158803e-01 -3.50767956e-03 1.09046364e+00 3.82560462e-01
-8.10533881e-01 3.54973406e-01 -7.69522429e-01 -5.76357543e-01
3.87664229e-01 -2.01930091e-01 -4.90859635e-02 5.51679254e-01
-9.76449430e-01 1.16378915e+00 3.35399956e-01 -4.73930240e-02
-8.61090422e-01 -6.17814064e-01 -8.85562599e-01 -4.58249561e-02
-2.85672843e-01 -5.03663123e-01 1.10292578e+00 -1.11399603e+00
-8.61891508e-01 1.10181689e+00 -2.28742301e-01 -9.31058109e-01
-3.36976737e-01 5.43739319e-01 -8.72578323e-01 2.30865568e-01
8.60866830e-02 5.23684800e-01 1.10005903e+00 -7.47363567e-01
-8.79697025e-01 -8.62173021e-01 -5.93796551e-01 -4.54054438e-02
2.92892128e-01 1.68919489e-01 6.55564129e-01 -3.83114904e-01
-3.86183918e-03 -9.75546002e-01 1.62007138e-01 -6.89836025e-01
-9.85629112e-02 -3.38687807e-01 1.37001753e+00 -8.01180422e-01
8.94575179e-01 -1.92763364e+00 -2.66214982e-02 -2.53773592e-02
-1.56405240e-01 3.87969077e-01 5.21182656e-01 2.11885036e-03
-2.49517739e-01 -3.45158547e-01 -5.47148287e-01 6.14714682e-01
-4.54577237e-01 5.30541353e-02 -6.13072634e-01 5.30208111e-01
-9.59703699e-02 9.89278972e-01 -4.10439581e-01 -7.80685097e-02
5.76272428e-01 1.70646667e-01 2.03980878e-01 -2.39374027e-01
-4.22154993e-01 8.50291848e-01 -4.23880070e-01 9.97992277e-01
7.26277530e-01 -3.81297231e-01 -1.99607208e-01 5.72675057e-02
-5.65845907e-01 3.49371821e-01 -5.14191747e-01 1.00641382e+00
-1.29928052e-01 1.13594890e+00 -1.20782048e-01 -1.03155386e+00
1.23672771e+00 -1.90932631e-01 4.17099684e-01 -1.61688244e+00
1.27709508e-01 7.85807222e-02 -1.64917305e-01 -6.29228950e-01
4.50343490e-01 -3.16397160e-01 1.35506824e-01 4.82311219e-01
-8.08333978e-02 -2.90360928e-01 5.21009088e-01 1.96347237e-02
1.17022955e+00 -4.16083366e-01 4.21710074e-01 -7.90086329e-01
4.24385220e-01 4.46012586e-01 5.00601172e-01 6.37872875e-01
-4.16586548e-02 3.27706128e-01 8.39068294e-02 -8.50304544e-01
-1.00131774e+00 -8.68191004e-01 -1.06508422e+00 6.04300022e-01
2.27056071e-02 -9.75594968e-02 -2.08168268e-01 -1.23218678e-01
2.73894399e-01 5.26112080e-01 -5.59889972e-01 2.29150646e-05
-3.02383900e-01 -1.39869845e+00 1.43814966e-01 5.73582590e-01
5.20041943e-01 -1.74829006e+00 -1.12876904e+00 -3.64684522e-01
-9.33832154e-02 -7.17074394e-01 3.45610112e-01 5.48385441e-01
-8.02591741e-01 -1.42014372e+00 -1.35819823e-01 -7.37828553e-01
3.51972282e-01 9.88930911e-02 1.39122319e+00 -2.21772805e-01
-9.20817971e-01 3.71382743e-01 -3.95250797e-01 -7.02610314e-01
-1.64247677e-01 -2.85426944e-01 1.13964733e-02 -1.08663991e-01
5.52459896e-01 -4.43800867e-01 -5.22307992e-01 1.82352796e-01
-8.44834745e-01 -3.72344524e-01 1.56924531e-01 4.26940918e-01
6.80894434e-01 1.73198968e-01 5.43311298e-01 -6.79966927e-01
4.13586289e-01 -5.45175493e-01 -1.36214435e+00 2.33815134e-01
-7.32940972e-01 -1.65642083e-01 3.35662752e-01 1.67082444e-01
-1.24088800e+00 -2.27477908e-01 4.36658531e-01 1.47845000e-02
-3.93742919e-01 3.98715079e-01 3.38147283e-01 -4.66421574e-01
1.04315221e+00 1.74375162e-01 -4.27669197e-01 -3.39034051e-01
-3.20499912e-02 1.02047348e+00 9.52707350e-01 2.11790234e-01
4.85939264e-01 7.83331871e-01 3.07459563e-01 -1.36892426e+00
-1.32817471e+00 -6.69405818e-01 -5.14239311e-01 -3.68551552e-01
7.62917638e-01 -8.94018650e-01 -4.85062957e-01 6.88317776e-01
-3.82535607e-01 -2.62466550e-01 -1.14757143e-01 3.18788707e-01
-3.97193074e-01 1.65159255e-02 -2.71032244e-01 -7.43968487e-01
-6.62338078e-01 -5.88087022e-01 1.15878546e+00 9.72530305e-01
1.44116422e-02 -1.14121890e+00 8.14383388e-01 6.84401691e-01
5.35152495e-01 7.50174582e-01 8.16970348e-01 -6.61479607e-02
-5.53959489e-01 1.69587880e-01 -8.25564787e-02 2.77122825e-01
-8.10769200e-02 1.66079849e-02 -9.78636861e-01 -3.79565746e-01
5.21832407e-01 -6.53037071e-01 1.45572233e+00 8.18192422e-01
6.09536707e-01 -4.96455878e-02 -2.97166556e-01 7.45649576e-01
1.43986762e+00 4.75490153e-01 6.35393023e-01 6.83554351e-01
2.24578336e-01 1.42145142e-01 4.66085881e-01 5.22170246e-01
9.28172469e-02 4.11588401e-01 2.38961980e-01 3.01021077e-02
-5.41681461e-02 1.95031002e-01 3.69455516e-01 4.00939405e-01
1.18954636e-01 -4.82507050e-02 -9.84235644e-01 6.54318810e-01
-1.58223379e+00 -1.28794682e+00 -3.85415614e-01 2.19081807e+00
2.15137213e-01 -3.19528461e-01 -2.29659438e-01 2.20734790e-01
5.40372312e-01 6.88655496e-01 -7.72823870e-01 -4.11618322e-01
-8.76329601e-01 3.34638685e-01 5.97898304e-01 3.95551503e-01
-1.09789062e+00 5.03365338e-01 7.05229330e+00 2.09936857e-01
-1.64988530e+00 1.34839127e-02 2.66421825e-01 -1.34775162e-01
-2.09608063e-01 2.12470889e-01 -1.03194690e+00 6.81712151e-01
1.25413585e+00 -1.82370394e-01 7.30673492e-01 6.79587424e-01
-1.70021541e-02 -8.44788849e-01 -4.21768874e-01 9.78343308e-01
6.17500782e-01 -1.77539635e+00 -5.83280027e-01 -2.26421636e-02
1.12868595e+00 8.70029449e-01 -2.70467192e-01 7.57424980e-02
-1.63481474e-01 -9.05935705e-01 3.07252079e-01 9.86610055e-01
4.47471440e-01 -4.52988803e-01 3.54262114e-01 2.43331388e-01
-8.39989305e-01 -2.44158223e-01 -4.11005259e-01 -1.57449424e-01
4.54906344e-01 1.19785130e+00 -6.57803118e-01 1.82832554e-01
1.44790316e+00 1.07909274e+00 -1.13020945e+00 6.78038299e-01
-3.36657614e-02 6.99283659e-01 -8.15806568e-01 1.64088830e-01
3.33423391e-02 -6.29104376e-01 5.67442715e-01 6.57634735e-01
3.14609766e-01 -2.15140775e-01 7.46695697e-02 5.81926167e-01
8.08598027e-02 -2.71431267e-01 -1.08414268e+00 -3.12783778e-01
1.05721593e-01 1.76043284e+00 -7.33931005e-01 -4.33972985e-01
-1.67909980e-01 2.97607034e-01 9.08731762e-03 -1.21039487e-01
-4.51047570e-01 -3.93053025e-01 2.17431724e-01 3.93427223e-01
6.40763581e-01 9.73115265e-02 -3.97247076e-01 -9.84919965e-01
-2.11702865e-02 -4.42035824e-01 8.19853544e-01 -1.58223939e+00
-1.06678605e+00 3.23899388e-01 -1.72216609e-01 -5.20642817e-01
-3.52951884e-01 -5.18125534e-01 -9.62660015e-01 4.91440654e-01
-1.77548873e+00 -8.65542114e-01 -7.59550810e-01 3.32405835e-01
-1.41539276e-01 -5.94055831e-01 4.11094755e-01 6.29932210e-02
-1.54609412e-01 -4.10226405e-01 4.60795492e-01 -3.89256418e-01
4.07578856e-01 -1.31489265e+00 7.62616768e-02 5.32570302e-01
2.93168038e-01 -2.63740778e-01 6.77358449e-01 -7.78769851e-01
-1.29046822e+00 -1.07302535e+00 1.00264692e+00 -8.02658439e-01
7.05925226e-01 3.15740585e-01 -8.54448378e-01 9.03901279e-01
4.12650347e-01 -7.26055279e-02 3.86175156e-01 -6.26500994e-02
1.76839903e-01 -3.10688704e-01 -1.23736084e+00 -2.23068371e-01
4.39152062e-01 -4.49026287e-01 -7.52304554e-01 9.95450020e-01
-1.38778985e-01 -2.02593267e-01 -8.94432843e-01 1.19982481e+00
5.32849357e-02 -1.56914890e+00 5.43870389e-01 -3.40481222e-01
2.08297148e-01 -4.22599256e-01 3.16757597e-02 -1.08232903e+00
-1.94765583e-01 -4.36558753e-01 -2.85556048e-01 8.10471177e-01
3.41101468e-01 -4.97664720e-01 4.78968948e-01 -3.36721510e-01
-2.10565224e-01 -3.55330795e-01 -7.27416098e-01 -4.52124178e-01
-2.45287761e-01 -2.97688171e-02 2.78092623e-01 1.07992077e+00
-2.41826147e-01 7.26645529e-01 -1.45782441e-01 3.27232122e-01
4.14723665e-01 1.07132912e+00 5.97463787e-01 -1.61637533e+00
1.46048829e-01 -3.14401895e-01 -5.03039241e-01 -5.22604346e-01
1.47101387e-01 -1.17438662e+00 -9.80538353e-02 -1.27277744e+00
7.36928061e-02 -3.02045852e-01 -1.96602941e-01 4.78488326e-01
4.35055494e-01 3.71518165e-01 -2.11196512e-01 4.32468265e-01
-3.27084750e-01 4.58719105e-01 8.48060906e-01 -6.27211928e-02
-1.57475531e-01 -2.23633781e-01 -2.49464035e-01 6.66118979e-01
1.03462434e+00 -2.77113050e-01 1.03910342e-01 -1.24378137e-01
3.90058786e-01 3.41335595e-01 4.92597997e-01 -1.41623437e+00
1.11126840e-01 -1.33475915e-01 6.20791793e-01 -1.11925638e+00
8.10710117e-02 -2.85051256e-01 2.41087675e-01 2.75579542e-01
1.40043646e-01 -9.41443741e-02 -4.85470286e-03 1.20002113e-01
-3.82586688e-01 -3.82692367e-01 1.46989787e+00 -2.73804903e-01
-7.49803901e-01 1.79940730e-01 -4.51509207e-01 -4.45709005e-02
1.16210330e+00 2.50936985e-01 -8.24534118e-01 -2.50470847e-01
-4.44778055e-01 3.80799025e-01 6.46088243e-01 6.00149632e-01
7.11392164e-02 -1.12451839e+00 -3.45677167e-01 7.64548063e-01
3.87174189e-01 -2.63703436e-01 2.83939123e-01 1.04607749e+00
-7.07299352e-01 8.83369148e-01 -4.40783232e-01 -8.47602665e-01
-1.05661356e+00 4.72272485e-01 6.24122620e-01 4.98014353e-02
-8.56080472e-01 2.38427997e-01 -1.38898179e-01 -4.26448375e-01
-1.67987525e-01 -1.18686542e-01 -5.48362672e-01 3.32359612e-01
8.44841003e-01 3.51216555e-01 3.76252502e-01 -9.50423717e-01
-1.30573869e-01 5.66651881e-01 2.59199113e-01 4.24964756e-01
1.40536189e+00 -1.50284348e-02 -4.16391224e-01 6.79387093e-01
7.58851886e-01 -3.71401876e-01 -7.23922372e-01 9.51849390e-03
1.31651670e-01 -1.25407070e-01 6.34741783e-01 -1.06398356e+00
-1.20878565e+00 7.64608860e-01 1.04932439e+00 4.47847247e-01
1.26248443e+00 5.10145545e-01 7.63965130e-01 6.76691115e-01
7.07948431e-02 -1.11339164e+00 -3.86314094e-01 5.88310063e-01
8.91039312e-01 -1.42676830e+00 -7.36676082e-02 4.92295325e-01
-7.08078146e-01 9.71050203e-01 3.93084645e-01 -5.76609746e-03
9.53895271e-01 4.67530429e-01 4.62331623e-01 -8.37651730e-01
-7.39058912e-01 -3.36854786e-01 5.49281240e-02 5.75332403e-01
-8.15005973e-02 1.04920380e-01 -1.53673857e-01 -1.92506492e-01
-6.51473224e-01 -8.86246469e-03 4.79903400e-01 1.13787770e+00
-1.16283953e+00 -5.55981338e-01 -6.67280674e-01 9.94504154e-01
-6.26209974e-02 3.39604504e-02 -5.93713999e-01 1.48343593e-01
2.66556274e-02 7.67998815e-01 4.83610243e-01 2.75138140e-01
-6.60964847e-02 9.50434431e-02 5.42688310e-01 9.67548229e-03
-1.92415908e-01 -2.45359465e-01 1.19546812e-03 -1.79225877e-01
-8.17663968e-01 -7.38012195e-01 -1.08083260e+00 -1.14471175e-01
-1.89428136e-01 2.47195676e-01 1.08120346e+00 9.73788977e-01
3.46559793e-01 1.72448695e-01 1.06844139e+00 -1.04796171e+00
1.46950781e-01 -1.38914883e+00 -1.13502288e+00 5.92894137e-01
4.48487103e-01 -7.45195568e-01 -7.31857538e-01 1.49405196e-01] | [6.704005718231201, 2.859804630279541] |
155554cd-826a-418b-9671-9b2c1543f4bc | iris-r-cnn-accurate-iris-segmentation-in-non | 1903.10140 | null | http://arxiv.org/abs/1903.10140v1 | http://arxiv.org/pdf/1903.10140v1.pdf | Iris R-CNN: Accurate Iris Segmentation in Non-cooperative Environment | Despite the significant advances in iris segmentation, accomplishing accurate
iris segmentation in non-cooperative environment remains a grand challenge. In
this paper, we present a deep learning framework, referred to as Iris R-CNN, to
offer superior accuracy for iris segmentation. The proposed framework is
derived from Mask R-CNN, and several novel techniques are proposed to carefully
explore the unique characteristics of iris. First, we propose two novel
networks: (i) Double-Circle Region Proposal Network (DC-RPN), and (ii)
Double-Circle Classification and Regression Network (DC-CRN) to take into
account the iris and pupil circles to maximize the accuracy for iris
segmentation. Second, we propose a novel normalization scheme for Regions of
Interest (RoIs) to facilitate a radically new pooling operation over a
double-circle region. Experimental results on two challenging iris databases,
UBIRIS.v2 and MICHE, demonstrate the superior accuracy of the proposed approach
over other state-of-the-art methods. | ['Chunyang Feng', 'Yufeng Sun', 'Xin Li'] | 2019-03-25 | null | null | null | null | ['iris-segmentation'] | ['medical'] | [ 1.75019220e-01 6.51522831e-04 -3.21492940e-01 -4.25975323e-01
-3.21488202e-01 -1.58229902e-01 3.23538959e-01 -3.65564942e-01
-2.64475673e-01 5.50376654e-01 2.08584890e-01 -4.00519311e-01
-1.99277401e-01 -3.74270558e-01 -3.34690571e-01 -9.41147387e-01
1.79152414e-01 4.94305161e-04 -2.31712073e-01 3.92896160e-02
3.82036269e-01 9.05514479e-01 -1.59129465e+00 -1.10635430e-01
1.40791810e+00 9.15253818e-01 -5.03689170e-01 5.47834218e-01
2.48582080e-01 5.64222813e-01 -6.18814409e-01 -2.60480940e-01
6.46822751e-01 -5.68995833e-01 -7.80195713e-01 2.85335809e-01
5.63875794e-01 -3.44159156e-01 -1.83486506e-01 1.09221339e+00
8.59674752e-01 2.41482660e-01 4.32872325e-01 -5.66992521e-01
-8.23623300e-01 5.00363231e-01 -1.27677047e+00 3.65327805e-01
-1.61810264e-01 4.07437623e-01 5.69692731e-01 -4.93014604e-01
3.51348698e-01 1.00645900e+00 5.85243702e-01 5.97278714e-01
-1.04650378e+00 -6.97711051e-01 -2.21423447e-01 -2.58502573e-01
-1.77230453e+00 -4.93843853e-01 4.90684986e-01 -2.99372733e-01
5.62395513e-01 4.96763498e-01 6.99993432e-01 3.31307292e-01
-1.91910163e-01 8.82658124e-01 1.60018444e+00 -5.10080695e-01
-2.77913988e-01 7.73530528e-02 -7.20560551e-02 6.40334249e-01
1.94150984e-01 5.38641989e-01 -3.57634202e-02 2.14959770e-01
1.28712451e+00 -5.86003549e-02 -2.90423244e-01 -8.70472640e-02
-1.17924476e+00 4.60435539e-01 8.04338098e-01 2.93401569e-01
-2.54642844e-01 -3.14102262e-01 8.00192431e-02 -1.55457988e-01
4.51504439e-01 6.28608763e-01 -8.57150257e-02 2.01419085e-01
-1.21926093e+00 2.16342304e-02 2.79904753e-01 7.89416373e-01
4.33464944e-01 -1.51851535e-01 -6.88805163e-01 9.14159000e-01
4.59929585e-01 2.33728904e-02 4.32943434e-01 -4.63716030e-01
1.68896168e-01 9.29159820e-01 1.01679582e-02 -7.73889899e-01
-6.61308706e-01 -1.09807098e+00 -1.20160496e+00 3.21978569e-01
5.86047292e-01 -3.79449815e-01 -1.14436364e+00 1.04911804e+00
6.67158604e-01 5.95491469e-01 8.85195881e-02 1.15320873e+00
1.20249760e+00 6.81192055e-02 -1.45727277e-01 -9.64319985e-03
1.19860649e+00 -1.18276060e+00 -5.16431451e-01 3.23921174e-01
3.84168059e-01 -1.04930019e+00 4.95251447e-01 2.76897043e-01
-1.06223810e+00 -7.25560009e-01 -7.71299362e-01 -1.86733738e-01
-2.27929577e-01 8.57821286e-01 6.50171876e-01 9.65340734e-01
-1.07869029e+00 3.69510472e-01 -4.69929576e-01 -3.58946055e-01
8.01880658e-01 9.35698509e-01 -1.75183445e-01 3.69196922e-01
-6.36519253e-01 5.08582830e-01 3.70529950e-01 5.79792082e-01
-1.85144588e-01 -4.58659500e-01 -6.32370353e-01 -2.00056523e-01
2.07212642e-01 -5.62906265e-01 9.41095352e-01 -9.58192110e-01
-1.96627402e+00 1.31246185e+00 -3.72394741e-01 -4.40614492e-01
2.77593911e-01 8.97923931e-02 -2.75162905e-01 1.42526984e-01
-4.17638928e-01 7.18861401e-01 7.83848226e-01 -1.15706933e+00
-8.60601544e-01 -5.73340595e-01 1.16686985e-01 3.61398846e-01
3.41204405e-02 5.64367354e-01 -5.74357271e-01 -7.79878616e-01
1.54333308e-01 -8.66426885e-01 -4.69489872e-01 -4.24437039e-02
-9.35480118e-01 -5.08200943e-01 3.36940497e-01 -6.31975174e-01
1.48457205e+00 -1.95835781e+00 -8.27572569e-02 3.98228467e-01
5.88047981e-01 9.26551700e-01 -1.00414954e-01 -3.88045371e-01
-3.35247308e-01 1.68185458e-01 -8.98983926e-02 -5.29050469e-01
-3.79832447e-01 -1.21329732e-01 2.45260000e-01 8.56627882e-01
2.95776933e-01 1.04180384e+00 -4.62390244e-01 -6.61396861e-01
4.14093107e-01 5.15537500e-01 -3.33491653e-01 1.01321012e-01
2.85841048e-01 8.89530122e-01 -4.37034428e-01 1.30127418e+00
1.06754100e+00 -3.86042744e-01 -3.37316811e-01 -1.87769458e-01
-4.64241862e-01 -3.15254509e-01 -1.23934424e+00 1.46339500e+00
4.50194441e-02 4.03255254e-01 1.20117158e-01 -8.21264565e-01
1.06213295e+00 2.56327510e-01 3.77321631e-01 -3.08488399e-01
6.01090968e-01 1.14324994e-01 3.63805234e-01 -4.28010941e-01
5.48098505e-01 1.23399600e-01 6.81643426e-01 3.12973797e-01
-7.85144418e-02 2.74561733e-01 5.57846017e-02 -3.44702810e-01
2.95660943e-01 1.62493095e-01 4.53170151e-01 -1.52241468e-01
9.32576358e-01 -2.01124147e-01 7.90871084e-01 6.17877305e-01
-8.02677751e-01 8.26383770e-01 4.12909061e-01 -6.53977871e-01
-6.84074283e-01 -3.82081270e-01 -6.66764557e-01 4.59268242e-01
3.91393423e-01 1.19174302e-01 -9.16049302e-01 -5.86385667e-01
4.76311371e-02 -1.25141516e-01 -7.72632420e-01 4.57214117e-01
-4.44768786e-01 -1.30595481e+00 5.86967051e-01 3.29560816e-01
7.86776960e-01 -9.14277375e-01 -2.50721693e-01 -1.43153369e-01
1.47625819e-01 -9.29794073e-01 -6.72635317e-01 -4.52203900e-01
-7.80491114e-01 -1.34193003e+00 -1.01318228e+00 -8.78406107e-01
1.10727322e+00 2.49281526e-01 6.92817032e-01 4.03521031e-01
-7.16337204e-01 -1.80887327e-01 -2.14050293e-01 -4.61964279e-01
1.92591492e-02 7.94441700e-02 -2.61380952e-02 6.76521659e-01
7.48051882e-01 -1.11106269e-01 -9.94737744e-01 4.35051650e-01
-5.25043905e-01 -2.32131276e-02 8.76680195e-01 8.87871563e-01
8.99168730e-01 9.09607634e-02 3.36933821e-01 -8.83202970e-01
5.19678295e-01 -1.07396401e-01 -8.74527931e-01 3.43505710e-01
-6.11554503e-01 -4.68353689e-01 2.77914822e-01 -3.22313726e-01
-1.09257901e+00 2.58433163e-01 4.54172045e-02 -3.29225391e-01
-4.34463114e-01 2.64861971e-01 3.54024500e-01 -7.98702121e-01
6.61638439e-01 1.91811398e-01 2.84508467e-01 -5.38654089e-01
5.86996198e-01 1.23996067e+00 7.75799334e-01 -3.81617039e-01
8.52771223e-01 5.20923913e-01 1.44250631e-01 -5.82312167e-01
-7.95950294e-01 -8.51826727e-01 -1.00146496e+00 -6.20144494e-02
8.96050692e-01 -9.92171884e-01 -1.19681787e+00 8.37255716e-01
-8.98982108e-01 -2.39113858e-03 -3.00175458e-01 6.86083019e-01
-2.58450389e-01 3.77082855e-01 -4.69265729e-01 -8.01931739e-01
-6.83987916e-01 -1.42619920e+00 7.78106153e-01 1.46715629e+00
2.84329414e-01 -6.88779175e-01 -6.14181943e-02 8.10998380e-01
4.17238802e-01 4.71335739e-01 3.36689234e-01 -5.80028713e-01
-7.50597060e-01 -2.21214518e-01 -9.44871724e-01 3.41062874e-01
2.40430653e-01 5.15956700e-01 -1.08231246e+00 -4.17276591e-01
-4.82410461e-01 -8.80768150e-02 6.45222187e-01 9.63509083e-01
1.46891928e+00 -5.83121777e-02 -4.36990172e-01 1.34920990e+00
1.37069511e+00 2.95274645e-01 9.11495805e-01 1.60435274e-01
5.68526268e-01 4.80654776e-01 5.02847850e-01 3.10270667e-01
2.89180398e-01 5.52974284e-01 2.75392830e-01 -1.09146345e+00
-3.52062106e-01 5.17283008e-02 -5.48230350e-01 5.05919814e-01
-7.47935951e-01 7.85087794e-02 -6.64161742e-01 6.88566625e-01
-1.69449317e+00 -6.18786573e-01 -1.86437935e-01 2.19408298e+00
1.07428026e+00 -3.90964687e-01 2.48068482e-01 -3.91528964e-01
9.19046164e-01 -3.22449207e-02 -6.82778180e-01 -3.55664074e-01
-2.66821325e-01 6.04228675e-01 5.50050795e-01 2.02193692e-01
-1.46413326e+00 1.01923549e+00 6.47252989e+00 9.57583725e-01
-1.20990324e+00 -1.74196184e-01 9.23618555e-01 6.99884668e-02
4.09544080e-01 -1.53867170e-01 -1.07948160e+00 2.62805164e-01
3.52196842e-01 2.12879688e-01 6.73919857e-01 3.61132234e-01
1.80347934e-01 -4.83713150e-02 -4.42074716e-01 9.94092166e-01
1.23803563e-01 -1.34670711e+00 -4.28694099e-01 2.19699308e-01
1.00166965e+00 4.60378602e-02 2.91793555e-01 -6.72139972e-02
1.29173145e-01 -1.32586479e+00 -2.77890354e-01 8.68639410e-01
1.18532062e+00 -7.58373976e-01 8.63381803e-01 -3.52485776e-02
-1.07199836e+00 -3.23217362e-02 -4.50801343e-01 2.37628862e-01
-4.54239190e-01 5.10132313e-01 -7.14261532e-01 7.91825294e-01
6.48271561e-01 9.50604975e-01 -8.04949522e-01 1.68684936e+00
-2.02960387e-01 4.87528443e-01 -1.95971876e-01 1.28483593e-01
9.63861421e-02 -5.76531351e-01 5.81692100e-01 9.18267012e-01
-6.73626885e-02 3.42760652e-01 -2.15483546e-01 1.24592340e+00
-2.55988985e-01 4.82326388e-01 -1.23046190e-01 1.86331104e-02
2.60534555e-01 1.75394237e+00 -6.63871348e-01 -1.01391137e-01
-5.79263866e-01 4.73268330e-01 8.23923871e-02 4.00446326e-01
-4.40895319e-01 -6.71293795e-01 7.42356479e-01 -9.84834507e-02
1.89878047e-01 2.20645338e-01 -4.54524457e-01 -1.11934769e+00
-2.04460889e-01 -9.22565103e-01 4.77168113e-02 -5.34783721e-01
-1.15744293e+00 6.44378304e-01 -5.58734059e-01 -1.44622266e+00
4.06608850e-01 -4.91128176e-01 -8.06926429e-01 1.42976284e+00
-2.09940529e+00 -1.63462341e+00 -3.98748040e-01 5.99385023e-01
5.46891205e-02 -6.17775142e-01 5.29948950e-01 3.07798833e-01
-1.28951716e+00 9.77466226e-01 2.00427100e-01 5.23796856e-01
8.78046811e-01 -1.39823472e+00 2.84335107e-01 1.06115246e+00
-1.18654944e-01 9.64558721e-01 -6.27362281e-02 -4.50543106e-01
-8.71680021e-01 -1.10896957e+00 6.32769942e-01 -2.04884946e-01
1.46966666e-01 2.24478930e-01 -6.02863133e-01 5.25468409e-01
3.68626744e-01 4.33347285e-01 9.32215631e-01 2.22331226e-01
8.79520103e-02 -1.98690802e-01 -1.30495512e+00 6.02132320e-01
5.82303822e-01 -2.08761513e-01 -2.11631894e-01 3.81228417e-01
4.95143056e-01 -1.07387257e+00 -1.15502012e+00 7.19613373e-01
5.97272038e-01 -1.07960331e+00 1.02811694e+00 -5.81898272e-01
2.76364386e-01 -7.67631471e-01 5.25685430e-01 -8.17989707e-01
-1.86144039e-01 -1.13994908e+00 -1.26117747e-02 1.09221864e+00
2.14015737e-01 -7.70065784e-01 9.56123412e-01 4.57597941e-01
-4.88538891e-02 -1.25103486e+00 -8.99821162e-01 -2.31904134e-01
-2.29539610e-02 3.84432942e-01 1.03097463e+00 1.01348400e+00
-2.47618914e-01 -2.63841629e-01 -3.51839632e-01 3.77012283e-01
7.55812764e-01 3.16628814e-01 9.95087862e-01 -1.27898717e+00
-4.66759838e-02 -6.97819769e-01 -4.38097060e-01 -1.29503119e+00
-2.53200501e-01 -6.31627738e-01 -2.52241284e-01 -1.19520998e+00
1.36313409e-01 -6.87831342e-01 -4.18345332e-01 5.37572324e-01
-3.85149449e-01 5.67996740e-01 -2.16269135e-01 3.93782794e-01
-2.75910378e-01 1.85224757e-01 1.82737052e+00 -3.38162039e-03
-5.94395578e-01 6.46796286e-01 -1.01582992e+00 4.71313477e-01
7.59483695e-01 9.37241241e-02 3.91386962e-03 6.12564199e-03
-2.63482958e-01 -6.95328182e-03 2.25074068e-01 -7.71228194e-01
6.44544423e-01 -4.63247159e-03 6.00345314e-01 -8.10598731e-01
-1.28886402e-01 -4.75542128e-01 -3.22438508e-01 9.66644660e-02
-1.65659130e-01 -5.68527043e-01 3.10820490e-01 2.14356720e-01
-4.31758940e-01 3.28247435e-02 1.20749629e+00 1.26753777e-01
-2.95842618e-01 5.00005543e-01 2.87963837e-01 -2.10873589e-01
1.11446893e+00 -5.60060382e-01 -5.60226202e-01 9.04464126e-02
-8.03583026e-01 4.39348936e-01 3.35241169e-01 2.81492025e-01
5.02444386e-01 -1.02370000e+00 -8.17321301e-01 7.83424318e-01
1.21422879e-01 2.13891581e-01 3.79647404e-01 1.34771848e+00
-6.77747548e-01 6.18445337e-01 -1.28265172e-01 -6.39074147e-01
-1.61572742e+00 2.83956140e-01 9.47113574e-01 -2.56832898e-01
-5.84004402e-01 1.20401013e+00 -2.09252629e-02 -6.54851675e-01
3.08424801e-01 -4.48481530e-01 -7.66947687e-01 -2.07849517e-01
6.49208248e-01 3.07318181e-01 -1.95566863e-01 -9.02921617e-01
6.15439638e-02 1.07021987e+00 -3.21587205e-01 7.38244474e-01
1.00779533e+00 -2.54024029e-01 -7.03536451e-01 -5.06474733e-01
5.98471701e-01 -2.99059767e-02 -1.12021911e+00 -6.65389478e-01
-4.31517869e-01 -7.27425039e-01 4.00937796e-01 -1.08569670e+00
-1.38603508e+00 7.73352206e-01 1.01992834e+00 -1.69682682e-01
1.49549162e+00 -4.65809047e-01 7.06013620e-01 -3.56264055e-01
-8.60650763e-02 -9.53231990e-01 -6.39084518e-01 -4.95023206e-02
5.29066801e-01 -1.55339766e+00 3.07291120e-01 -5.29592812e-01
-4.93735433e-01 1.28243804e+00 7.61092484e-01 3.02054211e-02
7.16079354e-01 -2.62727380e-01 3.77889544e-01 -1.77884236e-01
2.70306189e-02 -6.79431379e-01 1.04395688e+00 6.26831532e-01
5.04393637e-01 2.24379122e-01 -3.29782873e-01 4.44706917e-01
7.37442747e-02 3.06766063e-01 5.49201906e-01 3.86438161e-01
-1.42050818e-01 -1.11727166e+00 -5.13530850e-01 7.97610104e-01
-6.18782222e-01 -8.74802321e-02 -3.21305484e-01 7.85792470e-01
5.37117183e-01 9.96761501e-01 1.17287949e-01 -3.66030693e-01
9.09026563e-02 -5.80141068e-01 3.50466996e-01 -6.32040560e-01
-9.92875040e-01 4.32531863e-01 -5.81257761e-01 -4.18266028e-01
-9.10635233e-01 -4.26809877e-01 -9.29133236e-01 -2.04949766e-01
-7.58087397e-01 -1.93869114e-01 6.59952760e-01 8.39961767e-01
4.55677152e-01 3.94675076e-01 7.89734125e-01 -6.23042524e-01
-3.13228220e-01 -1.02043855e+00 -9.67131197e-01 -9.44432616e-02
6.59953475e-01 -3.91149819e-01 -2.29573026e-01 -1.83858112e-01] | [3.763547897338867, -3.6194405555725098] |
1f1dbe07-3eca-427d-9a84-bbc4ce972859 | spring-a-fast-stochastic-proximal-alternating | 2002.12266 | null | https://arxiv.org/abs/2002.12266v3 | https://arxiv.org/pdf/2002.12266v3.pdf | SPRING: A fast stochastic proximal alternating method for non-smooth non-convex optimization | We introduce SPRING, a novel stochastic proximal alternating linearized minimization algorithm for solving a class of non-smooth and non-convex optimization problems. Large-scale imaging problems are becoming increasingly prevalent due to advances in data acquisition and computational capabilities. Motivated by the success of stochastic optimization methods, we propose a stochastic variant of proximal alternating linearized minimization (PALM) algorithm \cite{bolte2014proximal}. We provide global convergence guarantees, demonstrating that our proposed method with variance-reduced stochastic gradient estimators, such as SAGA \cite{SAGA} and SARAH \cite{sarah}, achieves state-of-the-art oracle complexities. We also demonstrate the efficacy of our algorithm via several numerical examples including sparse non-negative matrix factorization, sparse principal component analysis, and blind image deconvolution. | ['Carola-Bibiane Schönlieb', 'Mike Davies', 'Jingwei Liang', 'Junqi Tang', 'Derek Driggs'] | 2020-02-27 | null | null | null | null | ['image-deconvolution'] | ['computer-vision'] | [ 1.31752700e-01 -2.80463517e-01 2.12524876e-01 -3.15320194e-01
-1.66989112e+00 -6.56732202e-01 6.80352747e-02 -4.31210518e-01
-3.30238402e-01 7.09269881e-01 4.65221375e-01 -3.85955185e-01
-6.07803822e-01 2.50408556e-02 -1.02827954e+00 -8.37894738e-01
-4.91029948e-01 3.52016926e-01 -5.23342848e-01 1.40128523e-01
6.92295954e-02 2.29233369e-01 -5.35863400e-01 -1.33903474e-01
9.95727181e-01 1.00084996e+00 2.76704669e-01 6.46387696e-01
4.40939456e-01 5.94364464e-01 1.75084114e-01 -4.11944389e-01
6.59411907e-01 -4.73080695e-01 -6.16025090e-01 4.58900511e-01
4.81671095e-01 -3.28335673e-01 -3.30044508e-01 1.39894927e+00
7.41423070e-01 3.09605241e-01 4.69859391e-01 -7.62133837e-01
-5.76177001e-01 6.62091076e-01 -9.79237914e-01 2.86427408e-01
2.38322154e-01 1.32313192e-01 9.77222085e-01 -1.51592803e+00
5.04408836e-01 1.02032995e+00 1.00676823e+00 5.98175675e-02
-1.25930464e+00 -4.05996412e-01 1.64509967e-01 -1.59544691e-01
-1.35545504e+00 -8.38862419e-01 6.00021064e-01 -4.73191053e-01
4.89415586e-01 6.02115214e-01 2.62731552e-01 7.12197721e-01
-3.35960865e-01 1.23316109e+00 1.35638881e+00 -3.33304852e-01
2.66318440e-01 -9.37939957e-02 1.34292006e-01 1.02695549e+00
2.62678325e-01 -1.08661041e-01 -6.82099223e-01 -6.81592703e-01
8.87994707e-01 -4.80815135e-02 -6.20508492e-01 -4.58487391e-01
-1.49127364e+00 8.92708242e-01 1.95218146e-01 -4.28187139e-02
-8.54820371e-01 3.20493430e-01 1.88495204e-01 1.13016725e-01
5.90692043e-01 3.35766762e-01 -2.12196261e-01 -1.90476924e-02
-1.19769204e+00 2.40059033e-01 7.26393163e-01 8.08464468e-01
5.93467295e-01 4.70838666e-01 -2.22199664e-01 9.50056195e-01
5.48659325e-01 1.19673860e+00 2.99773186e-01 -1.35143983e+00
4.68120098e-01 -2.84974277e-01 2.72925407e-01 -1.04644275e+00
-3.99872571e-01 -8.95391524e-01 -1.07256556e+00 -6.48927093e-02
4.36292112e-01 -4.05128628e-01 -7.11364806e-01 1.68653154e+00
4.57456529e-01 4.44655567e-01 -2.15661600e-02 1.29477584e+00
3.93635273e-01 5.21766961e-01 -5.00289440e-01 -7.97102451e-01
1.01845765e+00 -1.03495872e+00 -6.73913658e-01 -4.83554572e-01
2.96649456e-01 -9.03543830e-01 9.42945242e-01 4.67627078e-01
-1.37795162e+00 2.19690517e-01 -6.53466344e-01 1.12566009e-01
6.43419862e-01 3.17519903e-01 1.02933443e+00 5.71249366e-01
-1.02160585e+00 4.13976878e-01 -1.06291354e+00 1.19210601e-01
5.38840950e-01 2.11719871e-01 -2.67276496e-01 -2.55799830e-01
-4.14556175e-01 6.33289069e-02 -3.67878884e-01 4.30847585e-01
-1.21242952e+00 -9.71159875e-01 -6.67542338e-01 -8.60329792e-02
5.61464071e-01 -9.06936109e-01 1.07424462e+00 -8.01948667e-01
-1.65910029e+00 7.72646189e-01 -5.43421388e-01 -6.25532389e-01
7.84157336e-01 -6.94090962e-01 1.68035865e-01 2.01103166e-01
3.62234324e-01 -2.99174786e-01 1.25575125e+00 -9.87781644e-01
5.26549742e-02 -5.04904866e-01 -3.06726605e-01 2.94570208e-01
-1.92399040e-01 2.35000819e-01 -4.40756023e-01 -9.85296607e-01
5.65753162e-01 -1.14153802e+00 -7.67782569e-01 -3.88199948e-02
-6.12261772e-01 3.05374891e-01 1.38501808e-01 -1.04416251e+00
9.43540156e-01 -2.15959120e+00 4.82360005e-01 4.07333225e-01
3.44111472e-01 -1.18407451e-01 -5.31028435e-02 2.67108828e-01
-1.85667381e-01 -1.81373298e-01 -7.72135854e-01 -6.35505676e-01
-6.84255734e-02 -1.61129117e-01 -4.07980710e-01 1.10198510e+00
-4.98636246e-01 7.33828485e-01 -1.08047771e+00 -8.60758200e-02
-3.03672366e-02 2.37372130e-01 -6.89738452e-01 1.85016960e-01
1.03409074e-01 4.22383696e-01 -6.44235134e-01 7.91460276e-01
7.90396750e-01 -8.24284494e-01 5.22241276e-03 -3.63721222e-01
-4.11221236e-02 -2.57177949e-01 -1.60763907e+00 1.96392584e+00
-4.25285190e-01 4.20134008e-01 9.47762907e-01 -1.20473325e+00
8.45354721e-02 3.39603156e-01 7.53879368e-01 -1.65720776e-01
-5.56833409e-02 5.38687766e-01 -5.42848468e-01 -4.12056118e-01
8.33854079e-02 -3.11577618e-01 3.10332209e-01 5.37525833e-01
-2.05228804e-03 -1.95061639e-01 2.07429752e-01 6.59369349e-01
1.02581859e+00 -1.81499392e-01 3.70313913e-01 -6.29856944e-01
3.59455645e-01 -2.87102252e-01 4.32206422e-01 1.28481591e+00
-1.43960655e-01 9.56508517e-01 -8.34154896e-03 -1.11138739e-01
-6.71292245e-01 -1.23432684e+00 -1.82237044e-01 1.02736354e+00
-2.21543536e-01 -2.59752303e-01 -5.99035501e-01 -4.48486656e-01
-4.07341216e-03 2.42179617e-01 -3.20573360e-01 4.59442139e-01
-4.44506556e-01 -1.37814653e+00 3.06283925e-02 8.53543878e-02
3.83090913e-01 -2.98654109e-01 -1.04529560e-01 1.50452584e-01
-3.20656002e-01 -1.14911199e+00 -1.01948178e+00 -1.02126434e-01
-1.05895936e+00 -9.40045297e-01 -1.23651624e+00 -4.89597172e-01
8.38465989e-01 7.27396131e-01 1.00080550e+00 -2.83553988e-01
-2.62797892e-01 9.14479434e-01 3.95638645e-02 -6.86097750e-03
1.39417708e-01 -4.43170011e-01 5.30094624e-01 4.53919530e-01
-6.31457210e-01 -8.15393925e-01 -8.76033664e-01 1.85675502e-01
-7.63555884e-01 -1.12097070e-01 6.63636804e-01 1.12106717e+00
1.09973621e+00 -4.36596453e-01 2.24920541e-01 -1.21388125e+00
6.58127487e-01 -6.38351500e-01 -7.69088387e-01 1.80646509e-01
-7.07114339e-01 1.61467403e-01 4.08967942e-01 -3.62834781e-01
-1.01673186e+00 2.80542523e-01 -1.72044653e-02 -6.64826214e-01
7.16047883e-01 1.00236285e+00 2.94580966e-01 -4.91093874e-01
7.58812964e-01 5.15029192e-01 7.02691525e-02 -5.20320773e-01
5.73533952e-01 3.47309828e-01 7.63878763e-01 -6.20961964e-01
9.61972475e-01 1.01732671e+00 -3.45173776e-02 -9.89176810e-01
-1.16775227e+00 -6.45975053e-01 7.91405663e-02 9.44096595e-04
4.37914371e-01 -1.27029216e+00 -7.83169687e-01 6.00459635e-01
-6.66633904e-01 -4.36388731e-01 -3.84849936e-01 7.64562666e-01
-8.46922517e-01 7.93357074e-01 -6.04370296e-01 -9.76667702e-01
-6.49311066e-01 -1.06154895e+00 1.02306890e+00 -2.44922414e-02
2.28943169e-01 -9.73700702e-01 -2.26215702e-02 5.22459149e-01
5.45774877e-01 2.65602339e-02 1.51552662e-01 -1.81476802e-01
-6.89620018e-01 -7.15603158e-02 -2.35719115e-01 6.12212181e-01
-1.22911803e-01 -5.36054850e-01 -5.49423635e-01 -7.26499140e-01
6.22044265e-01 -3.73041153e-01 7.22178102e-01 9.42316055e-01
1.14483488e+00 -6.96129560e-01 -8.23860392e-02 1.30364966e+00
1.46521223e+00 -4.81927127e-01 2.54970938e-01 1.09838946e-02
6.80832803e-01 -1.76066998e-03 2.88347453e-01 8.60148787e-01
9.19732302e-02 3.36265922e-01 1.69992507e-01 8.30992162e-02
-3.79671007e-02 1.33734643e-01 4.66611326e-01 1.29667771e+00
-1.33125596e-02 4.60873321e-02 -7.42574573e-01 7.84564972e-01
-1.96103001e+00 -9.24684286e-01 -3.03655088e-01 2.33744931e+00
8.14170599e-01 -4.64877725e-01 -1.29289150e-01 -4.56401944e-01
4.11922932e-01 3.36326331e-01 -7.89392829e-01 3.47942173e-01
-5.47460854e-01 7.59969652e-02 9.26882327e-01 7.60003746e-01
-1.03100860e+00 6.00884795e-01 6.77303219e+00 9.77774799e-01
-8.02649915e-01 4.36615646e-01 7.35049129e-01 -2.86109418e-01
-4.06317919e-01 4.39801589e-02 -3.01035464e-01 3.52790266e-01
6.30090535e-01 -3.21359962e-01 1.22207451e+00 7.84134448e-01
2.74987131e-01 -6.55490011e-02 -7.11590886e-01 1.60932028e+00
1.62524492e-01 -1.48590851e+00 -4.05139804e-01 -4.94764298e-02
1.27736259e+00 5.79798520e-01 4.57282037e-01 -1.74292818e-01
5.50861180e-01 -7.73637056e-01 6.40532494e-01 5.06164491e-01
6.85306966e-01 -2.74032056e-01 3.48161191e-01 1.95970267e-01
-7.18866348e-01 -3.98336388e-02 -2.21975893e-01 1.99030474e-01
6.47950649e-01 1.39643407e+00 -3.17498922e-01 4.76595402e-01
7.11324096e-01 6.49053633e-01 -2.60497741e-02 1.15739942e+00
-2.68766850e-01 1.14838409e+00 -8.49544406e-01 3.44554842e-01
1.36205330e-01 -7.46954978e-01 1.28731620e+00 1.10215414e+00
4.45797771e-01 4.85166401e-01 3.16696554e-01 7.43885279e-01
-2.99782366e-01 2.60963917e-01 -2.33805984e-01 -2.04772756e-01
2.45872632e-01 1.26372802e+00 -5.44961452e-01 -2.48908296e-01
-4.40854222e-01 1.13875782e+00 1.66158065e-01 7.95075953e-01
-5.63889980e-01 6.33092225e-02 6.06425822e-01 -1.11379214e-01
3.16171110e-01 -5.33533275e-01 -4.15471464e-01 -1.60909605e+00
2.69348621e-01 -1.19027388e+00 4.51735049e-01 -6.63331866e-01
-1.46762276e+00 2.37002507e-01 -2.51512855e-01 -7.69649208e-01
-7.58313611e-02 -3.86152476e-01 -2.93777645e-01 9.30755079e-01
-1.32600582e+00 -8.00220728e-01 -8.67603123e-02 7.68718898e-01
4.88162726e-01 -2.02229604e-01 5.60040712e-01 3.22643429e-01
-4.71834570e-01 5.32589614e-01 9.76620078e-01 -1.52021006e-01
4.85571563e-01 -1.28206301e+00 8.43206272e-02 1.21970892e+00
3.45406979e-01 8.74720216e-01 9.15999532e-01 -4.44686800e-01
-2.18904161e+00 -6.95182264e-01 1.19854674e-01 -1.86735429e-02
9.82161820e-01 -1.08304031e-01 -5.37912190e-01 8.90076995e-01
1.83069967e-02 4.08683717e-01 6.75214529e-01 7.98604265e-02
-2.13699594e-01 -9.12819952e-02 -1.08186793e+00 4.37912881e-01
1.21782994e+00 -7.26407409e-01 -6.62088916e-02 1.27090478e+00
3.74901533e-01 -6.22873068e-01 -6.41582251e-01 3.63615870e-01
3.65826815e-01 -7.80898929e-01 1.35379183e+00 -4.07354444e-01
2.31342107e-01 -4.79744561e-02 -4.53163654e-01 -1.28989661e+00
-3.54078561e-01 -1.46410549e+00 -4.83827472e-01 6.47743285e-01
2.67583072e-01 -8.16843033e-01 8.50520313e-01 5.05664766e-01
-1.18239164e-01 -5.94103456e-01 -9.84141409e-01 -8.07336330e-01
-1.69831440e-01 -5.01872778e-01 -2.57492453e-01 8.66859496e-01
-2.50448257e-01 1.17631786e-01 -9.19436872e-01 5.61176956e-01
1.34112549e+00 3.92983854e-01 7.06263840e-01 -4.18225139e-01
-1.12717879e+00 -1.47986516e-01 -6.47859424e-02 -1.50148332e+00
-9.92865190e-02 -1.00639153e+00 1.80104911e-01 -1.28905296e+00
5.05727470e-01 -3.70436817e-01 -2.54051626e-01 1.23999931e-01
-5.22676229e-01 3.55021387e-01 1.03284590e-01 4.70334351e-01
-7.31945217e-01 6.38305962e-01 9.62232530e-01 -1.18329488e-01
-5.05605899e-02 1.70240358e-01 -8.29407215e-01 7.09429383e-01
8.73066336e-02 -5.30381501e-01 -2.65777677e-01 -8.03557992e-01
5.42858362e-01 2.68377572e-01 2.43977547e-01 -7.84422219e-01
2.57352948e-01 -3.32794972e-02 -6.04615547e-02 -1.14058509e-01
2.82575816e-01 -4.30861115e-01 1.48326024e-01 3.77210438e-01
-1.54482618e-01 -6.50926158e-02 -1.85200304e-01 7.66706109e-01
-1.08359568e-01 -2.04297021e-01 7.23378479e-01 -1.62828907e-01
-1.33701667e-01 5.11922956e-01 -1.68038517e-01 5.12746572e-01
4.69694674e-01 4.66245830e-01 -1.42290106e-03 -8.26549947e-01
-1.06620240e+00 3.36044699e-01 2.51416191e-02 -3.18771034e-01
7.64815569e-01 -1.07988513e+00 -1.23020101e+00 2.25218721e-02
-3.46057177e-01 -1.06710449e-01 5.59918284e-01 1.55841064e+00
-4.39735085e-01 1.40690058e-01 6.57545149e-01 -5.89600503e-01
-1.07834232e+00 5.24802566e-01 2.23181739e-01 -2.70081043e-01
-5.80188930e-01 1.40838361e+00 4.10840183e-01 -3.84970367e-01
6.52460102e-03 4.90175858e-02 6.16148829e-01 -3.31180900e-01
6.61445677e-01 6.99123383e-01 -5.09083681e-02 -3.63483429e-01
-2.93919533e-01 2.90854007e-01 1.09246962e-01 -4.84197378e-01
1.65599430e+00 -5.17241418e-01 -4.56291616e-01 3.38189960e-01
1.39788616e+00 6.61217749e-01 -1.34209263e+00 -4.89657730e-01
-2.13092893e-01 -6.77266479e-01 4.27786082e-01 -6.54958487e-01
-1.26434743e+00 4.99924093e-01 6.90627277e-01 -2.28424966e-01
1.06879246e+00 -1.01537317e-01 8.55541587e-01 6.03724837e-01
5.44349134e-01 -7.33579397e-01 -1.17045164e-01 3.15677613e-01
1.06142306e+00 -1.46062934e+00 4.67862427e-01 -5.37078083e-01
-3.80693108e-01 6.97355151e-01 -4.08255070e-01 -4.38291579e-01
8.25352371e-01 1.04180098e-01 4.21312824e-03 -2.07335576e-01
-2.54801333e-01 4.18323576e-02 3.73194724e-01 2.46348277e-01
5.79955317e-02 1.76124871e-01 -5.14208257e-01 5.30448854e-01
1.05126716e-01 1.00689372e-02 4.68015611e-01 8.24034154e-01
-4.15472016e-02 -5.99921286e-01 -5.48830867e-01 6.60348594e-01
-7.58155107e-01 -5.84656775e-01 1.73149973e-01 7.05694631e-02
-7.64184177e-01 7.29932427e-01 -7.55820215e-01 2.11398825e-01
-1.78925619e-01 -3.17010283e-01 5.36748230e-01 -3.58704746e-01
-2.90266246e-01 4.83894259e-01 -1.30653575e-01 -8.54826391e-01
-4.24137741e-01 -1.08463633e+00 -9.57753181e-01 6.47640089e-03
-4.55118895e-01 4.37705755e-01 7.31565297e-01 9.22687769e-01
4.17708755e-01 2.02915266e-01 8.24581385e-01 -7.42764413e-01
-1.26496387e+00 -7.23459780e-01 -7.34640121e-01 3.76746774e-01
5.45228124e-01 -2.12174907e-01 -6.12947643e-01 2.07010135e-02] | [7.05544376373291, 4.500284194946289] |
b1ad88dd-d3b2-42a2-9405-4ab7b700bd0e | inforex-a-collaborative-systemfor-text | null | null | https://aclanthology.org/R19-1083 | https://aclanthology.org/R19-1083.pdf | Inforex --- a Collaborative Systemfor Text Corpora Annotation and Analysis Goes Open | In the paper we present the latest changes introduce to Inforex {---} a web-based system for qualitative and collaborative text corpora annotation and analysis. One of the most important news is the release of source codes. Now the system is available on the GitHub repository (https://github.com/CLARIN-PL/Inforex) as an open source project. The system can be easily setup and run in a Docker container what simplifies the installation process. The major improvements include: semi-automatic text annotation, multilingual text preprocessing using CLARIN-PL web services, morphological tagging of XML documents, improved editor for annotation attribute, batch annotation attribute editor, morphological disambiguation, extended word sense annotation. This paper contains a brief description of the mentioned improvements. We also present two use cases in which various Inforex features were used and tested in real-life projects. | ["Micha{\\l} Marci{\\'n}czuk", 'Marcin Oleksy'] | 2019-09-01 | null | null | null | ranlp-2019-9 | ['morphological-disambiguation', 'text-annotation', 'morphological-tagging'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [-2.85879314e-01 2.39062726e-01 1.75301090e-01 -5.23145199e-01
-8.99329305e-01 -1.12688839e+00 4.94626820e-01 8.49125385e-01
-5.85108578e-01 8.42422962e-01 5.76799989e-01 -3.06042284e-01
-2.21132651e-01 -2.95151234e-01 1.02784149e-01 -1.81051031e-01
3.12410116e-01 7.62856364e-01 1.47540197e-01 -6.32651031e-01
5.11858225e-01 2.73071796e-01 -1.46643186e+00 5.38200736e-01
6.74562752e-01 3.59428436e-01 5.73400438e-01 6.00446880e-01
-7.48439193e-01 4.58931059e-01 -6.39094710e-01 -6.42455161e-01
2.56561905e-01 4.72035781e-02 -1.39622295e+00 -2.57614583e-01
-2.36048102e-01 2.63507307e-01 4.67515290e-01 1.05997944e+00
8.28319907e-01 2.00167876e-02 1.42237857e-01 -1.19983268e+00
-5.55997729e-01 1.14148211e+00 -3.81931126e-01 -4.06072894e-03
1.08473194e+00 -3.02432090e-01 8.49282324e-01 -7.42032230e-01
1.15288734e+00 1.09462142e+00 7.53598571e-01 2.54647762e-01
-6.64198697e-01 -6.20526969e-01 -2.02627942e-01 1.02353327e-01
-1.44669640e+00 -4.75671589e-01 1.40014321e-01 -4.89395767e-01
1.49278057e+00 6.47046089e-01 3.54819745e-01 7.21457779e-01
1.55533165e-01 4.51458305e-01 1.25456846e+00 -9.47778940e-01
1.52369197e-02 4.30063754e-01 3.66773903e-01 5.03941298e-01
1.25222048e-02 -7.79455245e-01 -1.86980739e-01 -4.73661751e-01
2.55412310e-01 -1.70596614e-01 8.63744766e-02 2.74391145e-01
-1.22492826e+00 4.29291010e-01 -5.99353135e-01 1.10873675e+00
-1.41392633e-01 -5.37633419e-01 1.00035024e+00 3.27912271e-01
8.20539951e-01 3.62585306e-01 -1.10963488e+00 -7.69858241e-01
-6.05810940e-01 2.15956852e-01 1.02651119e+00 1.37876761e+00
5.44315934e-01 -4.89675075e-01 1.32488713e-01 1.33737433e+00
3.20897251e-01 2.93454289e-01 9.33913946e-01 -7.19463825e-01
4.02200580e-01 9.21175778e-01 3.98255676e-01 -4.19090599e-01
-6.87528789e-01 5.00798896e-02 -9.07675773e-02 -7.52832144e-02
4.07727122e-01 -4.94982809e-01 -3.68765652e-01 9.52055454e-01
5.76927185e-01 -7.00745523e-01 2.52187937e-01 4.10219759e-01
1.15117264e+00 4.00650203e-01 3.29095066e-01 -4.96008664e-01
1.88694263e+00 -5.79175770e-01 -1.44948947e+00 2.45153382e-01
9.31772470e-01 -1.85623372e+00 1.11568022e+00 2.78118730e-01
-1.18991017e+00 -1.18711710e-01 -6.49562240e-01 -3.52512568e-01
-1.27993488e+00 5.54523349e-01 6.20082200e-01 6.53186321e-01
-1.05258548e+00 3.77438009e-01 -7.90164471e-01 -1.21113384e+00
-3.07523817e-01 2.54237592e-01 -6.14588261e-01 3.93975496e-01
-1.00706911e+00 8.83380413e-01 6.75306380e-01 -3.31298023e-01
1.05482042e-01 -3.81081969e-01 -7.63163030e-01 -3.90464902e-01
5.41411400e-01 1.15505964e-01 1.69918191e+00 -8.51740658e-01
-1.45982540e+00 1.22966564e+00 1.91102852e-04 2.02987581e-01
4.11273718e-01 -2.46104568e-01 -7.65895844e-01 -1.24193266e-01
6.22767568e-01 1.35278434e-01 -2.53611535e-01 -8.49619329e-01
-8.83868158e-01 -4.95436281e-01 -1.44104108e-01 1.40605390e-01
-2.87153453e-01 1.24652493e+00 -4.67214465e-01 -8.88440669e-01
-4.07130778e-01 -6.81922317e-01 -7.37561733e-02 -6.04186654e-01
-6.45117313e-02 -6.82900310e-01 5.08617043e-01 -1.02518749e+00
1.61607587e+00 -1.94648242e+00 -4.38017756e-01 -1.62409082e-01
-2.23368704e-01 6.18671440e-02 2.81922549e-01 1.53358030e+00
-2.64009058e-01 4.39608693e-01 -7.12468028e-02 -1.30955786e-01
4.32610065e-01 3.19059253e-01 2.83130378e-01 3.25053126e-01
-5.02268672e-01 4.93152112e-01 -1.04483342e+00 -8.11096668e-01
1.37786165e-01 1.87327549e-01 9.01188627e-02 -1.60893068e-01
1.53667126e-02 1.47981271e-01 -4.35892850e-01 6.61142945e-01
6.95897818e-01 3.04731935e-01 5.94573379e-01 9.45737287e-02
-1.05099487e+00 2.38742620e-01 -1.58662832e+00 1.99900568e+00
-5.85098028e-01 5.57327449e-01 2.81663388e-01 -6.10990524e-01
8.15400481e-01 9.53188181e-01 7.77598739e-01 -2.23621413e-01
1.60142258e-01 4.79800105e-01 -4.92768168e-01 -1.12569606e+00
8.62038076e-01 3.18376929e-01 -3.64926726e-01 6.60961390e-01
2.61966467e-01 9.63623002e-02 8.15851629e-01 1.76516995e-01
8.36280644e-01 5.75851798e-01 1.04807055e+00 -6.34875119e-01
6.52500629e-01 3.70762795e-01 5.75050950e-01 8.44358876e-02
9.11799166e-03 2.06730053e-01 3.88260603e-01 -3.54992002e-01
-1.11613941e+00 -3.59621614e-01 -3.79300505e-01 1.48586535e+00
-5.18052042e-01 -1.12889063e+00 -9.32989120e-01 -6.93784118e-01
-6.24833167e-01 7.93777525e-01 -4.61477518e-01 8.08233738e-01
-3.33897322e-01 -5.42453587e-01 7.46949911e-01 2.62232602e-01
2.35347182e-01 -1.08773124e+00 -4.81262863e-01 2.86347747e-01
-4.11331683e-01 -1.03264773e+00 -3.05500686e-01 1.55885026e-01
-3.17671061e-01 -1.02272725e+00 -3.09187174e-01 -9.27810848e-01
5.04047751e-01 -3.04364592e-01 8.57436955e-01 1.43914118e-01
-3.75146419e-01 4.46557701e-01 -1.03896582e+00 -6.46707356e-01
-4.76332456e-01 2.01795697e-01 -2.86415815e-01 -5.80983698e-01
6.94514871e-01 -3.12556028e-01 -1.62655190e-01 1.85325786e-01
-1.15466070e+00 -1.19248338e-01 -7.97060225e-03 4.37746912e-01
4.47274655e-01 -9.84675586e-02 4.89289731e-01 -1.18697786e+00
9.84289229e-01 -4.62631226e-01 -5.31912684e-01 4.08092529e-01
-6.26866162e-01 -1.57240286e-01 3.61438990e-01 1.21206120e-01
-1.32993758e+00 2.90472001e-01 -8.69416595e-01 6.07079983e-01
-5.75143456e-01 7.51945972e-01 -2.25468621e-01 3.72414023e-01
4.42367107e-01 -2.34063774e-01 -1.43284604e-01 -1.18732738e+00
3.76445860e-01 1.44407392e+00 1.60804808e-01 -6.00458741e-01
2.11055428e-01 1.86760753e-01 -6.85817540e-01 -6.94938838e-01
-3.63702267e-01 -1.04784310e+00 -1.04492009e+00 -2.24018842e-01
8.06323528e-01 -7.07613051e-01 -5.39157093e-01 1.20253295e-01
-1.05488193e+00 -3.80157381e-01 -3.68967831e-01 3.23570848e-01
-2.14959577e-01 3.98568362e-01 -4.03627068e-01 -7.01126456e-01
-5.52246809e-01 -8.73498023e-01 9.15064931e-01 1.68850705e-01
-5.87128282e-01 -1.32417214e+00 2.39945576e-01 4.77609098e-01
1.07102498e-01 3.96468878e-01 7.80291378e-01 -1.37026989e+00
7.44652390e-01 -2.61449188e-01 1.66949153e-01 3.10268551e-02
1.54838473e-01 5.45628428e-01 -5.53003669e-01 8.05031434e-02
-4.23663765e-01 -1.73236988e-02 -3.08228225e-01 -1.85225770e-01
6.88984394e-01 -5.28195202e-01 -5.46137869e-01 -2.47757956e-01
1.50754476e+00 3.87728125e-01 5.25517404e-01 8.08179080e-01
6.60819337e-02 8.70137990e-01 1.15626180e+00 9.03573573e-01
5.04323363e-01 7.18160868e-01 -1.00176647e-01 3.09725404e-01
1.33349687e-01 1.90540671e-01 2.29477689e-01 1.31164372e+00
-2.54905343e-01 -2.00543076e-01 -1.60031796e+00 5.85869431e-01
-2.19050121e+00 -8.72428536e-01 -8.00614834e-01 2.00299668e+00
1.10647869e+00 -4.02815908e-01 3.23497593e-01 7.45027140e-02
7.27384210e-01 -4.04480726e-01 5.82065761e-01 -7.38824606e-01
3.05663627e-02 5.12630880e-01 3.26875478e-01 8.35648894e-01
-8.91781330e-01 9.85317886e-01 5.56792927e+00 8.37010205e-01
-8.76664639e-01 9.61948574e-01 6.60571977e-02 6.45658225e-02
-1.24695569e-01 1.87202454e-01 -1.02055132e+00 3.39993417e-01
1.00551355e+00 -5.72316110e-01 9.18487832e-03 8.53492498e-01
5.92636883e-01 -2.69756883e-01 -6.19191766e-01 4.34338987e-01
1.33353937e-02 -1.29415524e+00 -5.22546470e-01 -1.29034827e-02
3.72125298e-01 4.14329201e-01 -6.86189055e-01 3.15664053e-01
4.28610593e-01 -4.27168220e-01 9.42094564e-01 3.07424396e-01
8.62695634e-01 -6.83286846e-01 1.16906726e+00 1.37920901e-01
-1.11735356e+00 2.67908871e-01 -3.26627165e-01 8.07610527e-03
1.18855096e-01 4.66301143e-01 -7.12054789e-01 9.76466775e-01
1.01378548e+00 2.88450897e-01 -8.64073813e-01 1.07116580e+00
5.98096997e-02 3.08164716e-01 -2.68032372e-01 -3.81319702e-01
-1.74552985e-02 -3.28818023e-01 6.08398080e-01 1.94809651e+00
3.27636898e-01 2.78487921e-01 3.30301881e-01 5.86504266e-02
3.50299180e-01 1.12572610e+00 -2.18919322e-01 -2.37145703e-02
5.76440394e-01 1.75114143e+00 -1.18618655e+00 -4.52978700e-01
-4.23719913e-01 6.45347238e-01 -3.56105044e-02 -3.65847237e-02
-4.48261052e-01 -1.01068890e+00 2.60274559e-01 2.75581717e-01
1.64234415e-02 -2.27931410e-01 -1.75073359e-03 -8.97990584e-01
1.57414749e-02 -1.13925076e+00 6.29248500e-01 -9.85524237e-01
-8.83212268e-01 8.22811484e-01 4.52721775e-01 -1.03629339e+00
-1.54283375e-01 -6.82788789e-01 -2.24951580e-01 6.24448717e-01
-7.86791027e-01 -1.23987889e+00 -1.41219839e-01 5.65086186e-01
7.47001708e-01 -1.29553795e-01 1.32635784e+00 7.44962335e-01
-5.39059281e-01 2.26089433e-01 4.19476956e-01 2.13212535e-01
1.03365469e+00 -1.63400686e+00 -9.11410619e-03 4.80325997e-01
-2.61935149e-03 8.85092556e-01 7.94501126e-01 -8.70569468e-01
-9.55102563e-01 -6.55047059e-01 1.65734732e+00 -3.81029636e-01
1.18502593e+00 -6.95292413e-01 -5.01703322e-01 9.17140782e-01
1.29702854e+00 -4.04931694e-01 1.34916306e+00 2.70751631e-03
1.42270058e-01 -9.06601399e-02 -1.25606692e+00 4.51306731e-01
6.17603540e-01 -2.30372012e-01 -7.52920210e-01 1.02610958e+00
2.91659057e-01 -4.50858444e-01 -1.49671590e+00 -1.53415620e-01
4.45905030e-01 -5.51791370e-01 2.17022523e-01 -2.14769930e-01
-7.06687197e-03 -4.05049324e-01 1.07131712e-01 -1.00888669e+00
1.39282033e-01 -1.26240993e+00 9.34101760e-01 2.14643502e+00
7.68412530e-01 -7.98617721e-01 -1.22361392e-01 5.20228922e-01
-4.15742338e-01 -2.32944846e-01 -8.44597101e-01 -3.61960113e-01
-2.44362559e-02 -6.05224967e-01 7.41669059e-01 1.24574232e+00
1.09065998e+00 2.74736404e-01 1.81935221e-01 -1.06615044e-01
-1.19507697e-03 -2.85117477e-01 4.00009274e-01 -1.28789020e+00
-1.01433121e-01 -2.07378477e-01 -9.31199118e-02 -7.33753890e-02
7.35371327e-03 -1.07788134e+00 -2.06871793e-01 -1.92860925e+00
-3.77276056e-02 -3.95867229e-01 2.23727539e-01 1.08723056e+00
2.00734347e-01 1.84324071e-01 -2.93714534e-02 2.42896512e-01
-6.88693106e-01 -2.04572096e-01 5.83876967e-01 3.90771329e-01
-3.45481262e-02 -3.46961580e-02 -5.79844296e-01 6.01245701e-01
1.00278258e+00 -9.14566875e-01 5.66904321e-02 -2.96741992e-01
8.70186508e-01 -2.33821824e-01 -2.28976920e-01 -5.62262416e-01
1.80516765e-01 -2.71946996e-01 -2.51008500e-03 -6.54637933e-01
-2.51769632e-01 -9.50952709e-01 4.46319133e-01 1.62633374e-01
-1.94584146e-01 9.18501318e-01 4.58952367e-01 -2.29378611e-01
-3.78771797e-02 -1.05528939e+00 2.19384387e-01 -5.40111125e-01
-6.12155080e-01 -2.59221882e-01 -1.00313449e+00 -1.93206482e-02
1.23091853e+00 -1.25153482e-01 -4.49477583e-01 8.66688937e-02
-9.67255294e-01 1.83383569e-01 5.36113024e-01 6.05436325e-01
-1.81988761e-01 -9.65156376e-01 -6.41085982e-01 -2.77234524e-01
3.53290230e-01 -5.60524106e-01 -9.24737975e-02 9.80603516e-01
-1.10369313e+00 4.89911288e-01 -4.53639567e-01 1.49596231e-02
-1.72715199e+00 2.88760781e-01 -6.94131060e-03 -2.69629717e-01
-4.35311943e-01 2.78868109e-01 -8.15013111e-01 -8.26060176e-01
3.78875993e-02 -1.58194736e-01 -8.20984542e-01 3.17091435e-01
4.65957195e-01 5.33703029e-01 5.53373694e-01 -8.38022590e-01
-5.73349059e-01 2.80428827e-01 7.96620324e-02 -6.42623723e-01
1.67729723e+00 -5.41410625e-01 -7.54147172e-01 7.16578662e-01
8.44637096e-01 6.67542458e-01 -2.29096785e-01 2.37563550e-01
6.34506822e-01 -1.40441880e-01 -1.11003153e-01 -1.21246266e+00
-6.34438634e-01 2.65815347e-01 5.73061585e-01 6.11379445e-01
8.95693243e-01 5.73750772e-02 1.67043567e-01 3.56288254e-01
7.89567232e-02 -1.86506879e+00 -5.87758124e-01 4.95896727e-01
1.07107687e+00 -7.84895718e-01 1.80912882e-01 -4.39740390e-01
-7.36245096e-01 1.47264469e+00 2.65012950e-01 4.30069357e-01
8.65076303e-01 8.42165887e-01 5.35236418e-01 -4.60832834e-01
-7.10345566e-01 -2.06797063e-01 -1.65562958e-01 7.59748816e-01
1.36971176e+00 -8.76949131e-02 -1.50166321e+00 9.02541339e-01
-4.70789701e-01 -3.63401882e-02 7.57229686e-01 1.39002728e+00
-1.14053242e-01 -1.81787014e+00 -3.95287186e-01 1.50323510e-01
-1.31865883e+00 -1.93025112e-01 -7.30800807e-01 1.17024171e+00
4.09310430e-01 1.16611028e+00 -6.71143755e-02 1.30000964e-01
4.05528367e-01 4.32270229e-01 1.69935942e-01 -1.05976880e+00
-1.60803449e+00 4.37145203e-01 7.66970515e-01 -2.68103361e-01
-6.62729919e-01 -1.17412567e+00 -1.48941290e+00 -2.38026723e-01
-4.70988840e-01 8.46078813e-01 1.11744869e+00 8.59493196e-01
3.68995667e-01 9.04449970e-02 -2.70612855e-02 -3.25487792e-01
-5.28062554e-03 -1.40178978e+00 -4.32206988e-01 2.55861789e-01
-6.16321623e-01 -3.17184329e-01 8.42340067e-02 6.07183635e-01] | [9.66866397857666, 9.313729286193848] |
336e9550-2656-4a17-bf63-035185863458 | avface-towards-detailed-audio-visual-4d-face | 2304.13115 | null | https://arxiv.org/abs/2304.13115v2 | https://arxiv.org/pdf/2304.13115v2.pdf | AVFace: Towards Detailed Audio-Visual 4D Face Reconstruction | In this work, we present a multimodal solution to the problem of 4D face reconstruction from monocular videos. 3D face reconstruction from 2D images is an under-constrained problem due to the ambiguity of depth. State-of-the-art methods try to solve this problem by leveraging visual information from a single image or video, whereas 3D mesh animation approaches rely more on audio. However, in most cases (e.g. AR/VR applications), videos include both visual and speech information. We propose AVFace that incorporates both modalities and accurately reconstructs the 4D facial and lip motion of any speaker, without requiring any 3D ground truth for training. A coarse stage estimates the per-frame parameters of a 3D morphable model, followed by a lip refinement, and then a fine stage recovers facial geometric details. Due to the temporal audio and video information captured by transformer-based modules, our method is robust in cases when either modality is insufficient (e.g. face occlusions). Extensive qualitative and quantitative evaluation demonstrates the superiority of our method over the current state-of-the-art. | ['Dimitris Samaras', 'Aggelina Chatziagapi'] | 2023-04-25 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Chatziagapi_AVFace_Towards_Detailed_Audio-Visual_4D_Face_Reconstruction_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Chatziagapi_AVFace_Towards_Detailed_Audio-Visual_4D_Face_Reconstruction_CVPR_2023_paper.pdf | cvpr-2023-1 | ['3d-face-reconstruction', 'face-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 6.95716962e-02 1.07739188e-01 -2.21710876e-02 -1.14244044e-01
-6.09067976e-01 -4.78502721e-01 5.27795136e-01 -3.71188223e-01
-4.41094898e-02 4.05616671e-01 -8.38654116e-03 2.30264999e-02
3.39856893e-01 -4.53619629e-01 -6.26153827e-01 -6.36878788e-01
2.65476733e-01 5.02219081e-01 1.36789560e-01 -4.32576127e-02
8.25729370e-02 9.24852073e-01 -2.01893353e+00 9.97944921e-02
4.18054968e-01 1.25147903e+00 2.44032983e-02 6.41214907e-01
-4.03765678e-01 2.82789677e-01 -1.73342600e-02 -3.41786921e-01
3.04239959e-01 -3.62553060e-01 -4.68726128e-01 5.91777384e-01
7.25567758e-01 -5.76416552e-01 -2.62533516e-01 9.42434669e-01
5.10781646e-01 -1.57195106e-01 6.61222637e-01 -1.25014985e+00
1.98046565e-02 -3.07185829e-01 -7.02488184e-01 -4.16634202e-01
9.35631633e-01 -1.86461466e-03 4.58581120e-01 -1.30046690e+00
9.69606459e-01 1.47163033e+00 6.24674380e-01 7.81651616e-01
-1.50152791e+00 -5.92891753e-01 7.85072222e-02 5.72335087e-02
-1.46629643e+00 -9.91522551e-01 1.20981658e+00 -6.32482171e-01
5.36322117e-01 1.37726337e-01 8.90173256e-01 9.27582026e-01
-1.55051887e-01 5.15323281e-01 9.56083119e-01 -5.29355407e-01
1.67245016e-01 2.40717947e-01 -4.99658674e-01 8.62352431e-01
-2.29759783e-01 2.11358309e-01 -7.74878800e-01 -2.35777512e-01
9.96525109e-01 -1.30588770e-01 -4.25426483e-01 -7.64763355e-01
-7.49045074e-01 5.80071032e-01 -1.36713177e-01 1.50808528e-01
-3.86937827e-01 -9.69251171e-02 5.41846156e-02 2.02851266e-01
6.75942719e-01 -2.08150506e-01 -2.09301427e-01 -9.70995352e-02
-1.34853649e+00 2.40237907e-01 6.98517323e-01 9.00484025e-01
9.65972781e-01 1.16536610e-01 5.06173372e-01 6.86089337e-01
8.08434904e-01 6.18861973e-01 6.03595041e-02 -1.19078815e+00
7.73088038e-02 4.46807444e-01 4.16693613e-02 -9.53703821e-01
-1.54315710e-01 2.20493183e-01 -6.11480653e-01 5.40011108e-01
4.61806327e-01 1.83614448e-01 -8.98177981e-01 1.70577395e+00
9.25486982e-01 3.82253885e-01 -1.40983820e-01 9.93219614e-01
9.65566158e-01 3.36022347e-01 -4.30545241e-01 -5.67586482e-01
1.20586479e+00 -3.08400959e-01 -8.43731284e-01 -1.27551556e-01
4.33816761e-02 -1.04747033e+00 6.03933156e-01 3.35304320e-01
-1.49050236e+00 -4.70228434e-01 -7.08548248e-01 -1.75260663e-01
7.06326142e-02 -8.83959532e-02 1.58551440e-01 6.70281470e-01
-1.19721544e+00 3.57513249e-01 -9.32369888e-01 -3.57113570e-01
3.10320675e-01 4.55127358e-01 -9.17352736e-01 -2.69491464e-01
-8.02928984e-01 6.91863954e-01 -2.76835799e-01 2.08974734e-01
-7.73224294e-01 -7.61895835e-01 -1.11721098e+00 -2.69518524e-01
3.45643669e-01 -7.85559952e-01 1.14896166e+00 -8.10811996e-01
-1.95381439e+00 1.25410354e+00 -5.48508406e-01 1.16447881e-01
8.73147845e-01 6.43471405e-02 7.91071951e-02 6.56549752e-01
-3.29587162e-01 8.17812741e-01 1.21999586e+00 -1.55651593e+00
-1.48922995e-01 -6.58950388e-01 1.85058981e-01 2.07419083e-01
-9.66769829e-02 -1.62795614e-02 -8.51878405e-01 -3.94695610e-01
4.27370727e-01 -9.43217576e-01 1.10102117e-01 7.57763684e-01
-2.30011418e-01 1.71544537e-01 1.14339292e+00 -8.16462636e-01
6.37487531e-01 -2.15128827e+00 5.86294353e-01 3.48258540e-02
6.49035200e-02 1.41456917e-01 -6.48923516e-02 2.64572054e-01
6.57222494e-02 -2.92464439e-03 -1.28951445e-01 -1.08579528e+00
-8.56977552e-02 9.99951437e-02 4.45896434e-03 8.35685670e-01
9.03823450e-02 5.15917718e-01 -4.58141416e-01 -7.91599572e-01
6.29327536e-01 1.20064974e+00 -6.66615605e-01 4.47444201e-01
-2.59280324e-01 9.21126366e-01 -2.37646192e-01 9.92550552e-01
9.46716905e-01 2.09401939e-02 2.36919895e-01 -4.43444073e-01
-2.06058025e-01 -1.00119505e-02 -1.43439114e+00 2.19635844e+00
-5.41465759e-01 5.80714107e-01 8.52039158e-01 -7.02702582e-01
7.17319727e-01 6.78552866e-01 6.76490426e-01 -4.24056590e-01
2.46000633e-01 3.48616600e-01 -6.38132513e-01 -6.32556200e-01
1.65238842e-01 -3.61712575e-01 4.34114695e-01 2.91627705e-01
7.28630424e-02 -6.35836840e-01 -3.63461643e-01 -7.13011846e-02
6.39554501e-01 5.32483041e-01 1.35715887e-01 9.91846099e-02
7.25759506e-01 -4.10095632e-01 4.83383268e-01 -2.71703098e-02
4.82344292e-02 8.85352135e-01 3.78189951e-01 -1.78811818e-01
-9.22867894e-01 -8.58842075e-01 -2.99046785e-01 2.92638302e-01
1.52559236e-01 -4.90584344e-01 -9.35689926e-01 -3.80422920e-01
1.51146548e-02 7.32163489e-02 -6.03799701e-01 3.26360434e-01
-5.35682738e-01 -6.41118586e-02 1.39733717e-01 7.78230354e-02
3.90692532e-01 -5.82043767e-01 -7.64393032e-01 3.94974425e-02
-2.19157204e-01 -1.51878154e+00 -3.98712486e-01 -4.88212198e-01
-1.04546618e+00 -1.14560664e+00 -8.61039937e-01 -5.49287796e-01
6.75530612e-01 3.00033957e-01 8.61730695e-01 1.47241279e-01
-4.11153972e-01 8.23954105e-01 -5.06840348e-02 -3.43480259e-02
-5.19647598e-01 -3.67640913e-01 1.69106975e-01 3.81115735e-01
-2.18083113e-01 -8.67435873e-01 -5.60641587e-01 4.88642991e-01
-7.94435740e-01 2.45455176e-01 8.09947774e-02 5.54022610e-01
7.83203244e-01 -1.60291791e-01 -5.40706217e-02 -4.22681361e-01
-1.74388394e-01 -1.40722618e-01 -8.04191053e-01 -6.63881823e-02
-1.43986762e-01 -1.65750846e-01 1.98249936e-01 -6.43840551e-01
-1.01742387e+00 5.35269558e-01 -2.67832458e-01 -1.09760654e+00
-3.28815788e-01 1.26095399e-01 -5.98961890e-01 -3.24041516e-01
1.79185003e-01 2.23258659e-02 5.00453532e-01 -6.54756188e-01
9.38179865e-02 5.51411211e-01 4.00638580e-01 -3.16595733e-01
8.94379199e-01 8.42031598e-01 4.11961555e-01 -1.30857015e+00
-3.07869494e-01 -2.40479782e-01 -9.33983922e-01 -6.11092329e-01
8.90475750e-01 -9.39881146e-01 -9.49771404e-01 5.69289625e-01
-1.40281570e+00 -2.04696134e-01 -9.12718549e-02 5.78722596e-01
-7.94333458e-01 4.83534366e-01 -4.44434285e-01 -1.06394136e+00
5.27823865e-02 -1.40514755e+00 1.51991343e+00 -2.36639362e-02
-7.16302693e-02 -9.56907570e-01 -1.20736413e-01 4.44159865e-01
1.29946530e-01 5.27527094e-01 6.81216538e-01 2.92043656e-01
-8.27821195e-01 -9.59033966e-02 9.48910415e-02 1.16480760e-01
1.11679971e-01 1.46627292e-01 -1.36135948e+00 -1.43396974e-01
1.88635945e-01 -1.43063590e-01 4.47197229e-01 5.29805720e-01
7.44891644e-01 -2.52648056e-01 -1.90711811e-01 7.92447150e-01
1.28859389e+00 -1.35573328e-01 4.75742042e-01 -1.77504942e-01
7.47227609e-01 1.07272053e+00 6.10982239e-01 4.25488472e-01
4.49333191e-01 1.20630097e+00 9.09009576e-01 -5.30903377e-02
-4.64031637e-01 -2.57706940e-01 3.27322274e-01 5.83684266e-01
-2.03902960e-01 6.03566021e-02 -6.89299226e-01 3.06672543e-01
-1.55993199e+00 -8.41570079e-01 1.05454519e-01 2.26812649e+00
6.51955664e-01 -2.64025837e-01 6.38845861e-02 5.58195114e-01
6.05535328e-01 1.30324308e-02 -4.48825151e-01 -2.74674706e-02
-5.21820821e-02 2.32237466e-02 -2.73530893e-02 8.59506488e-01
-7.64301956e-01 8.09627116e-01 5.86809254e+00 6.37171030e-01
-1.30447924e+00 2.98406091e-03 2.10680485e-01 -1.77000761e-01
-5.84118366e-01 -6.29118755e-02 -7.03449428e-01 1.93195462e-01
5.14564157e-01 2.12074459e-01 3.94144624e-01 4.82558668e-01
3.34111631e-01 -2.07286164e-01 -1.22521496e+00 1.38982940e+00
3.43950331e-01 -1.30767190e+00 -8.93591866e-02 5.78035891e-01
3.68055999e-01 -4.07469153e-01 5.19102663e-02 -3.60120893e-01
-4.42802519e-01 -1.00724518e+00 1.05403209e+00 6.32252455e-01
1.24884462e+00 -5.73738039e-01 3.71621430e-01 4.92552906e-01
-1.23670352e+00 2.81685352e-01 4.09308635e-02 2.85453320e-01
4.32027191e-01 5.23104668e-01 -5.68021953e-01 4.64611471e-01
6.71027422e-01 7.26354659e-01 -7.82308728e-02 7.33830154e-01
-1.17717594e-01 1.62803829e-01 -4.96895105e-01 6.40176535e-01
-3.14559340e-01 -1.80944562e-01 7.72617877e-01 8.17330182e-01
5.19073904e-01 3.98132712e-01 7.45972991e-02 7.73899615e-01
-5.23191355e-02 1.60161033e-01 -8.05651128e-01 1.80278569e-01
2.59101242e-01 1.15183067e+00 -5.88130891e-01 1.62641317e-01
-4.60003793e-01 9.83856082e-01 -5.72949871e-02 2.94035822e-01
-4.86734957e-01 2.82471091e-01 8.66215050e-01 6.31052554e-01
3.08714747e-01 -4.89441782e-01 3.01104393e-02 -1.24379230e+00
5.42473383e-02 -6.63507700e-01 -9.61827561e-02 -8.92453372e-01
-7.09757388e-01 5.81499994e-01 1.18296975e-02 -1.30996954e+00
-6.11296535e-01 -5.43926775e-01 -2.46572152e-01 7.48199463e-01
-1.87559497e+00 -1.36162162e+00 -3.82019907e-01 8.16188157e-01
5.43763459e-01 3.45876515e-02 9.72642064e-01 2.93037176e-01
-1.97828770e-01 4.79470998e-01 -5.07798493e-01 -2.37627715e-01
7.19839752e-01 -7.23202646e-01 1.97094996e-02 4.88739103e-01
-3.89346927e-02 2.15570599e-01 8.04963887e-01 -4.63819087e-01
-2.01389003e+00 -5.26218295e-01 8.92524540e-01 -2.78018624e-01
1.50251225e-01 -5.01793027e-01 -8.47224891e-01 3.62009734e-01
-1.60695791e-01 3.86263847e-01 6.42418921e-01 -3.43751132e-01
-4.04037863e-01 -5.08039910e-03 -1.55937660e+00 3.88104558e-01
1.09993839e+00 -7.57030964e-01 -1.78178340e-01 -4.09992784e-02
3.25443238e-01 -7.51133800e-01 -8.84014845e-01 2.91243464e-01
8.93363416e-01 -1.25227714e+00 1.03197670e+00 -2.00959314e-02
1.84534088e-01 -4.04038966e-01 -4.21555221e-01 -1.00761259e+00
4.02371138e-01 -8.85312974e-01 -3.37450594e-01 1.19515908e+00
-9.99970287e-02 -2.77171284e-01 8.44274521e-01 5.80611467e-01
6.61239699e-02 -3.91302288e-01 -1.26646328e+00 -4.33045387e-01
-2.02980772e-01 -5.75416803e-01 5.13758540e-01 8.02564144e-01
-2.50789881e-01 2.61569694e-02 -4.25077498e-01 3.12314361e-01
8.66035521e-01 2.18109965e-01 1.00528765e+00 -1.53778267e+00
-6.97225854e-02 -1.89641103e-01 -6.09648347e-01 -9.66696322e-01
5.89993834e-01 -6.55735731e-01 -1.35107756e-01 -1.22807693e+00
-1.00145690e-01 -2.62474835e-01 6.58208668e-01 2.62116045e-01
4.13797796e-01 6.12891436e-01 2.01464534e-01 1.63737580e-01
5.77047430e-02 6.67969525e-01 1.25598681e+00 7.85981119e-02
-2.42230535e-01 -7.22260624e-02 -1.74589247e-01 9.97318089e-01
2.91622430e-01 -2.19446734e-01 -3.13571244e-01 -4.62926716e-01
2.21924521e-02 6.85612559e-01 5.85155308e-01 -6.38635099e-01
2.66553015e-01 -1.96048602e-01 3.83031875e-01 -5.75199306e-01
1.14398777e+00 -1.16493642e+00 4.43698913e-01 1.37528017e-01
8.18784162e-02 -9.97681320e-02 3.61559659e-01 4.63516563e-01
-2.81908184e-01 -3.87672186e-02 9.92661178e-01 -3.61517891e-02
-2.02056155e-01 5.96056759e-01 -2.48550549e-01 -1.79407820e-01
8.61890733e-01 -5.20797372e-01 2.60290295e-01 -6.17537498e-01
-9.16209340e-01 -2.24767953e-01 9.46601689e-01 1.76714078e-01
9.31065857e-01 -1.39343929e+00 -7.23024428e-01 7.51750410e-01
-1.54038191e-01 1.88433394e-01 3.71596813e-01 8.91462684e-01
-4.37152147e-01 1.28653035e-01 -7.62759522e-02 -1.06650937e+00
-1.84632039e+00 2.42024824e-01 5.64888597e-01 2.79344052e-01
-7.24834025e-01 5.49770594e-01 2.45499790e-01 -3.05202544e-01
3.63142669e-01 -1.20503409e-02 -1.11821041e-01 3.15537959e-01
5.10082960e-01 2.41156206e-01 1.83948383e-01 -1.20588577e+00
-3.95511657e-01 1.36423743e+00 2.81147361e-01 -5.52892506e-01
1.26877856e+00 -3.08913559e-01 4.65094708e-02 4.21604782e-01
1.27565038e+00 2.11142361e-01 -1.41595209e+00 -1.18252732e-01
-6.09917283e-01 -8.56386125e-01 2.41827860e-01 -1.80219010e-01
-1.24816787e+00 1.27480280e+00 4.71598148e-01 -6.50759190e-02
1.22416174e+00 3.16326730e-02 6.51637197e-01 -2.40636706e-01
4.01092917e-01 -7.47398019e-01 1.24698490e-01 2.62563199e-01
9.97615099e-01 -1.11723447e+00 1.41374161e-03 -7.98418283e-01
-2.29291454e-01 1.24561739e+00 1.89767942e-01 2.18472362e-01
1.03660381e+00 3.91498387e-01 2.36102864e-02 -6.98303711e-03
-5.30903220e-01 -1.01812378e-01 4.05780554e-01 6.76159978e-01
2.39605144e-01 -3.93442869e-01 2.09509715e-01 6.72940239e-02
-5.75440563e-02 -5.95264835e-04 4.25848246e-01 8.16185057e-01
-4.92975004e-02 -1.13832688e+00 -6.18357360e-01 -3.43098938e-01
-4.72749054e-01 2.73439348e-01 -3.06320131e-01 8.52810025e-01
8.85824040e-02 1.00164652e+00 1.22314647e-01 -1.98759422e-01
4.21227574e-01 1.18679389e-01 1.04409397e+00 -4.35181737e-01
-1.57743633e-01 7.04609036e-01 -1.73030838e-01 -7.77463675e-01
-7.31537223e-01 -9.83451426e-01 -1.18890667e+00 -4.07097191e-01
-2.71286964e-01 -1.89332917e-01 1.16380978e+00 6.57205045e-01
3.68710518e-01 1.08150661e-03 6.81722999e-01 -1.64420772e+00
-7.02931210e-02 -5.77012241e-01 -6.17994666e-01 1.79312095e-01
7.74644017e-01 -1.00009573e+00 -6.98736310e-01 1.89330816e-01] | [13.141488075256348, -0.09941704571247101] |
09f97267-f3d8-47f6-99d4-ec539e9232f5 | spectrogram-channels-u-net-a-source | 1810.11520 | null | http://arxiv.org/abs/1810.11520v2 | http://arxiv.org/pdf/1810.11520v2.pdf | Spectrogram-channels u-net: a source separation model viewing each channel as the spectrogram of each source | Sound source separation has attracted attention from Music Information
Retrieval(MIR) researchers, since it is related to many MIR tasks such as
automatic lyric transcription, singer identification, and voice conversion. In
this paper, we propose an intuitive spectrogram-based model for source
separation by adapting U-Net. We call it Spectrogram-Channels U-Net, which
means each channel of the output corresponds to the spectrogram of separated
source itself. The proposed model can be used for not only singing voice
separation but also multi-instrument separation by changing only the number of
output channels. In addition, we propose a loss function that balances volumes
between different sources. Finally, we yield performance that is
state-of-the-art on both separation tasks. | ['Se-Young Yun', 'Jaehoon Oh', 'Duyeon Kim'] | 2018-10-26 | null | null | null | null | ['singer-identification'] | ['music'] | [ 2.08330527e-01 -6.35272801e-01 -1.51324317e-01 1.93758860e-01
-1.02546275e+00 -8.11536789e-01 1.97853148e-01 -1.54466294e-02
-1.27986044e-01 4.19522554e-01 4.16141927e-01 -1.78291678e-01
-1.45644709e-01 -2.15351790e-01 -2.87150413e-01 -7.25928664e-01
3.89214516e-01 -1.72669724e-01 3.53280939e-02 -1.69271886e-01
1.55878840e-02 2.78708994e-01 -1.40990090e+00 1.27719596e-01
8.35968673e-01 1.00501287e+00 1.03021570e-01 8.25835049e-01
-1.18971087e-01 4.75287437e-01 -6.42544806e-01 -1.65314469e-02
2.47001752e-01 -1.10873067e+00 -4.62353230e-01 -1.83525801e-01
3.92449170e-01 8.29761252e-02 -2.21505195e-01 1.04952681e+00
8.33094895e-01 4.16292727e-01 6.88692570e-01 -8.07555795e-01
-3.50545228e-01 1.05787337e+00 -5.82539916e-01 2.00771570e-01
1.46318749e-01 -3.06791335e-01 1.42403877e+00 -8.56711507e-01
2.19279844e-02 1.13310230e+00 5.55147111e-01 5.31742990e-01
-1.11933160e+00 -7.85331547e-01 -2.66116500e-01 3.43903393e-01
-1.08303928e+00 -6.80347979e-01 1.15033722e+00 -3.90452564e-01
3.94536346e-01 6.72826827e-01 3.62484694e-01 5.82095206e-01
-4.01364416e-01 1.04318643e+00 8.33702922e-01 -7.65824676e-01
6.82810917e-02 -2.14493707e-01 2.02689543e-01 2.27035061e-01
-3.03898484e-01 -2.00589493e-01 -8.19508374e-01 -9.67382565e-02
7.94029415e-01 -4.17308271e-01 -6.13648295e-01 1.30984142e-01
-1.15108085e+00 6.05016887e-01 1.54111445e-01 5.34722149e-01
-7.26516694e-02 2.27552146e-01 5.29439747e-01 3.56604546e-01
3.33802015e-01 6.87650323e-01 -1.81663260e-01 -2.74893045e-01
-1.10356629e+00 1.79603532e-01 5.70334196e-01 4.32943523e-01
2.37365112e-01 5.62689364e-01 -4.52766269e-01 1.43419111e+00
1.54031003e-02 4.66362238e-01 8.41832876e-01 -1.03059947e+00
3.36528629e-01 -3.07656378e-02 -3.49201225e-02 -6.61841333e-01
-2.87729263e-01 -6.47929847e-01 -8.99429202e-01 -2.36766890e-01
3.77633393e-01 -2.17323527e-01 -4.70291436e-01 1.79948032e+00
1.52229026e-01 6.19575858e-01 -1.95058919e-02 9.21429336e-01
9.39216077e-01 8.10670376e-01 -6.03049994e-01 -5.69778025e-01
1.20975685e+00 -1.24752057e+00 -8.94421518e-01 7.68369362e-02
-9.91384909e-02 -1.05271482e+00 1.18681610e+00 5.72796345e-01
-1.20031238e+00 -8.27779114e-01 -9.48770106e-01 -5.62962554e-02
2.90333182e-01 4.19794083e-01 2.35191524e-01 6.36415780e-01
-5.91167808e-01 7.63306081e-01 -4.87163812e-01 2.18086600e-01
-6.07069097e-02 1.50696024e-01 -7.02065369e-03 5.30798197e-01
-1.05037498e+00 2.23630950e-01 7.06195012e-02 1.40632745e-02
-6.39201581e-01 -4.64592934e-01 -6.06328607e-01 3.67083430e-01
3.19320291e-01 -4.73997802e-01 1.51527858e+00 -8.22128296e-01
-1.89855409e+00 2.83617109e-01 -3.96163791e-01 -3.40943635e-01
1.27792493e-01 -4.88525301e-01 -7.18522549e-01 6.81478903e-02
-2.43463982e-02 -5.85269183e-03 1.23036826e+00 -1.08683360e+00
-5.28471053e-01 -3.14683616e-01 -4.84128892e-01 3.70261431e-01
-5.58318079e-01 2.86893725e-01 -4.78706926e-01 -1.08501422e+00
2.57073194e-01 -8.75931680e-01 1.57846257e-01 -6.10508204e-01
-6.65583253e-01 -9.72436741e-02 5.84446609e-01 -7.83075213e-01
1.70102024e+00 -2.55208468e+00 4.67220932e-01 4.92498837e-02
-3.30153517e-02 3.99274945e-01 -2.67047197e-01 1.89523518e-01
-6.14623465e-02 -1.89501017e-01 -2.48383731e-01 -5.19664288e-01
-8.13253745e-02 -3.75069052e-01 -6.62284017e-01 2.11071432e-01
-2.21342534e-01 5.78616321e-01 -7.26561606e-01 -1.87570840e-01
1.25524729e-01 4.23337728e-01 -4.72961217e-01 3.23909134e-01
5.59214354e-02 6.96933270e-01 9.15389881e-02 3.04290920e-01
4.42212731e-01 3.46602827e-01 6.37595505e-02 -1.63778350e-01
-2.19484136e-01 6.31149113e-01 -1.44351065e+00 1.78139591e+00
-6.74550414e-01 7.14843154e-01 4.01005685e-01 -5.30674994e-01
1.03834009e+00 6.33702278e-01 4.31113601e-01 -1.70347959e-01
1.80461019e-01 5.49280345e-01 3.24496746e-01 -1.27721921e-01
6.85607672e-01 -3.60228926e-01 7.43897771e-03 3.59781682e-01
1.50905505e-01 -3.03248435e-01 9.46386755e-02 -3.05462807e-01
7.09090590e-01 -1.78038254e-01 2.20660120e-01 -1.93892773e-02
7.87559092e-01 -6.33696437e-01 5.51903009e-01 5.15193522e-01
-1.38188293e-02 8.34679127e-01 1.91966444e-01 4.53395158e-01
-5.74926734e-01 -1.07562757e+00 -1.13843173e-01 1.28207767e+00
9.42567140e-02 -5.65401912e-01 -8.16003859e-01 -2.33488366e-01
-1.17473677e-01 5.52842796e-01 1.40110860e-02 -1.39180228e-01
-8.01508784e-01 -3.77369851e-01 9.68861163e-01 4.58893389e-01
5.03325686e-02 -1.10659695e+00 -1.34112850e-01 3.60526741e-01
-4.63701665e-01 -7.95718610e-01 -1.15856445e+00 2.75278330e-01
-7.78263748e-01 -7.11535454e-01 -9.67600942e-01 -7.86141574e-01
-1.84383184e-01 4.36697334e-01 6.79303408e-01 -4.84298199e-01
1.17836349e-01 2.16412321e-02 -2.42155284e-01 -3.89181107e-01
-4.93253350e-01 2.74020880e-01 4.26393867e-01 4.67583716e-01
-1.00176834e-01 -8.48102391e-01 -3.29994291e-01 2.78655142e-01
-8.49369049e-01 -1.92086354e-01 1.61777765e-01 6.84884965e-01
6.27921999e-01 1.46415964e-01 1.03943896e+00 -5.31384110e-01
1.18588650e+00 -1.07965231e-01 -2.48302534e-01 -4.56335470e-02
-3.08635920e-01 -7.06293583e-02 1.18864071e+00 -6.32420719e-01
-8.99083018e-01 -8.73184055e-02 -3.60870749e-01 -6.18211150e-01
-4.46830839e-02 2.29321152e-01 -4.10058558e-01 2.12278247e-01
4.65929091e-01 2.70466030e-01 -2.33574167e-01 -1.11083806e+00
5.49531877e-01 1.18477798e+00 9.94091630e-01 -3.28771383e-01
6.25708282e-01 1.06855333e-01 -1.61696687e-01 -1.01825476e+00
-8.64227295e-01 -9.96111155e-01 -3.81110191e-01 -8.25038478e-02
6.95821404e-01 -7.21733510e-01 -7.62315750e-01 6.55544043e-01
-1.18298697e+00 -7.72743747e-02 -3.66412401e-01 7.71054983e-01
-5.35572052e-01 2.34045073e-01 -8.19174051e-01 -9.75197077e-01
-5.79138935e-01 -1.02909398e+00 7.63420761e-01 3.27689618e-01
-2.59905368e-01 -6.84860229e-01 4.86103296e-01 2.34339505e-01
1.98300108e-01 -2.14392245e-01 6.81438446e-01 -5.61094165e-01
-9.76204872e-02 1.09512679e-01 1.23964131e-01 8.10221612e-01
4.89279896e-01 -3.30256909e-01 -1.29622185e+00 -5.14412895e-02
1.91350073e-01 -1.00290120e-01 1.14918315e+00 5.11543632e-01
1.15224981e+00 -3.52083623e-01 5.89207038e-02 7.29664207e-01
1.04636073e+00 2.80102402e-01 4.44879442e-01 -1.80914044e-01
1.01759815e+00 4.05129850e-01 2.98941225e-01 3.23139220e-01
-1.49885580e-01 1.01759934e+00 1.52807042e-03 -2.83076242e-02
-4.49856848e-01 -3.00663650e-01 4.72235888e-01 1.54647827e+00
3.42497928e-03 -1.96296811e-01 -4.34816778e-01 6.96480751e-01
-1.89226544e+00 -9.76706982e-01 -3.38281572e-01 2.42307663e+00
1.04183531e+00 -5.22021875e-02 4.64463800e-01 5.83221734e-01
7.65012205e-01 3.98516446e-01 -5.05021751e-01 -4.44831103e-01
-1.41631633e-01 5.44440806e-01 1.88331679e-01 6.40224457e-01
-1.00459790e+00 7.46267021e-01 6.07661343e+00 1.41278315e+00
-1.22334731e+00 2.82740355e-01 1.06589340e-01 -3.17384958e-01
-2.72454053e-01 -2.76846498e-01 -5.74390292e-01 4.95728999e-01
7.19851017e-01 -7.42630064e-02 7.33052254e-01 5.33246100e-01
3.21759611e-01 1.26729324e-01 -1.12083566e+00 1.34609520e+00
2.31814206e-01 -9.85346317e-01 -8.47952291e-02 -2.94004887e-01
6.71181560e-01 -3.44329894e-01 3.86750139e-02 4.53897193e-02
-3.75539154e-01 -9.09025609e-01 8.07775736e-01 2.09176317e-01
8.79845023e-01 -9.01660442e-01 2.35590324e-01 3.34639221e-01
-1.51965499e+00 -7.15660006e-02 1.24733858e-02 3.21329683e-02
1.75734341e-01 5.63095927e-01 -6.70296073e-01 6.96554005e-01
2.66973644e-01 5.64498961e-01 -1.11113012e-01 1.28089190e+00
-2.82194525e-01 1.12558782e+00 -1.41951472e-01 1.43855318e-01
-1.60506919e-01 -2.28852227e-01 9.74429190e-01 1.29013205e+00
4.21177864e-01 -1.66993901e-01 7.75862858e-02 5.95575094e-01
-2.73558617e-01 3.71708453e-01 -2.41925672e-01 1.04405895e-01
5.97030044e-01 1.17334712e+00 -4.49687570e-01 -4.39987779e-02
1.78204570e-02 1.03102446e+00 -6.99355304e-02 2.43382692e-01
-5.10010242e-01 -7.41525829e-01 8.32495511e-01 -5.31885140e-02
2.99475163e-01 -1.40762180e-01 -3.56539637e-01 -1.01902914e+00
-1.23677664e-01 -8.76831055e-01 7.87220150e-02 -4.70060736e-01
-1.08441162e+00 7.19603002e-01 -3.82273555e-01 -1.49303257e+00
-3.86487305e-01 -4.12100673e-01 -8.24414074e-01 9.66790080e-01
-1.41344380e+00 -5.97182810e-01 1.18665278e-01 5.43596089e-01
6.91684723e-01 -3.41984451e-01 8.16805661e-01 5.83850086e-01
-6.03847146e-01 6.67083263e-01 4.65377003e-01 1.41570047e-01
1.01765716e+00 -1.31206381e+00 2.62997240e-01 7.87249744e-01
8.14292431e-01 4.75611061e-01 5.90848327e-01 -1.71150506e-01
-1.00856185e+00 -9.66282547e-01 7.51485169e-01 1.18692976e-03
5.00791073e-01 -2.40210176e-01 -7.98472643e-01 1.20261945e-01
2.04206407e-01 -3.38399351e-01 9.19795156e-01 1.66729555e-01
-3.46856803e-01 -4.70250875e-01 -4.95034218e-01 5.13094962e-01
6.06013238e-01 -7.08297908e-01 -5.31284928e-01 -7.87755772e-02
7.53157437e-01 -3.45039517e-01 -3.54928404e-01 1.60064921e-01
6.40795350e-01 -9.42204177e-01 9.83025849e-01 -3.56162637e-01
3.33137631e-01 -5.36804914e-01 -1.75538227e-01 -1.66243339e+00
-5.01299798e-01 -9.48331892e-01 -3.64414901e-01 1.49324536e+00
2.57233709e-01 -2.24842072e-01 1.67415708e-01 -3.93675685e-01
-2.65189886e-01 -4.35166717e-01 -8.42413068e-01 -1.00576949e+00
-1.45173579e-01 -6.41936064e-01 5.25137901e-01 8.24083090e-01
1.33131087e-01 9.02573526e-01 -7.38309205e-01 -2.42078871e-01
4.57305610e-01 4.02500480e-01 6.27625108e-01 -1.16281736e+00
-1.00682890e+00 -8.54520500e-01 2.25728169e-01 -1.21836603e+00
1.55052885e-01 -9.22372401e-01 3.68928611e-01 -1.32413983e+00
1.76250059e-02 -3.27739388e-01 -6.73474729e-01 2.36666873e-01
-4.02119666e-01 4.46478069e-01 6.74094617e-01 4.47799385e-01
-3.05933625e-01 6.56597376e-01 1.13513672e+00 -6.82013184e-02
-7.86699712e-01 6.05307221e-01 -7.17113256e-01 8.99093330e-01
8.63774002e-01 -3.55413020e-01 -2.50563979e-01 -2.61874944e-01
-1.55072242e-01 4.67035860e-01 1.71797797e-01 -1.17089748e+00
5.69201410e-02 3.84318940e-02 -2.04695061e-01 -4.71240312e-01
6.32229745e-01 -4.68500614e-01 7.58023560e-02 3.38034295e-02
-6.07266545e-01 -4.88775581e-01 1.80862308e-01 3.57259482e-01
-6.17206514e-01 -4.90340769e-01 7.85556793e-01 1.33433387e-01
-6.62626773e-02 -1.63309630e-02 -1.04497269e-01 8.90532583e-02
3.25032920e-01 2.43464336e-01 -8.87092948e-02 -6.18559778e-01
-5.05270898e-01 -2.36517489e-01 -1.05647221e-01 3.94496173e-01
4.58080769e-01 -1.58332014e+00 -7.91669309e-01 1.14844009e-01
-1.58984616e-01 -2.36909300e-01 1.81850612e-01 7.92922497e-01
1.07646640e-02 3.35056335e-01 1.00616671e-01 -3.51415128e-01
-1.52082551e+00 2.84500629e-01 1.81672126e-01 -1.08206302e-01
-2.82811403e-01 9.06531155e-01 3.67129892e-01 -3.13838303e-01
5.11115730e-01 -2.00921685e-01 -3.28227133e-01 2.07091123e-01
7.33675241e-01 8.57571840e-01 -3.23455296e-02 -6.35502636e-01
-2.04504371e-01 6.35391533e-01 4.17972744e-01 -5.33908784e-01
9.91944194e-01 -1.38358906e-01 -2.66228676e-01 1.11324644e+00
1.10784149e+00 9.29368794e-01 -8.41527343e-01 -1.86158180e-01
-2.72760063e-01 -3.55132580e-01 2.37384006e-01 -7.15758145e-01
-8.82310987e-01 1.02320421e+00 3.42908055e-01 3.83574128e-01
1.34374404e+00 -1.06763311e-01 1.26548135e+00 1.65186197e-01
-7.28647709e-02 -1.04302108e+00 1.31249934e-01 6.50963068e-01
7.71541178e-01 -7.74662673e-01 -4.68095690e-01 -2.50893295e-01
-5.23416221e-01 9.90242839e-01 2.89725035e-01 -1.63529478e-02
5.86965382e-01 1.68604106e-01 1.88450754e-01 3.51774871e-01
-2.01902926e-01 -3.87529850e-01 7.44606674e-01 1.31176457e-01
7.22845495e-01 3.79643887e-01 -3.90173435e-01 1.15897703e+00
-3.87712747e-01 -2.05937624e-01 3.07118863e-01 1.48048520e-01
-6.17133677e-01 -1.39608264e+00 -8.12888324e-01 2.84852982e-01
-6.59395456e-01 -4.90166575e-01 -5.96874416e-01 -8.38565379e-02
1.20356567e-01 1.22274017e+00 -1.46058902e-01 -5.48430920e-01
3.38205338e-01 2.84069240e-01 4.85561460e-01 -7.44691074e-01
-7.26365685e-01 9.36067164e-01 -1.48092330e-01 3.09189893e-02
-3.06110293e-01 -5.62235117e-01 -1.28419065e+00 6.52021989e-02
-6.24651492e-01 4.69763100e-01 4.78036553e-01 6.62001312e-01
1.49346307e-01 9.44830298e-01 1.02375126e+00 -8.05837810e-01
-5.78583896e-01 -1.24665225e+00 -1.03174460e+00 2.64629275e-01
6.66629434e-01 -2.54428834e-01 -3.24925184e-01 6.52903467e-02] | [15.642047882080078, 5.577869415283203] |
75757269-13c9-4321-932b-c795a584a74a | curriculum-meta-learning-for-order-robust | 2101.01926 | null | https://arxiv.org/abs/2101.01926v3 | https://arxiv.org/pdf/2101.01926v3.pdf | Curriculum-Meta Learning for Order-Robust Continual Relation Extraction | Continual relation extraction is an important task that focuses on extracting new facts incrementally from unstructured text. Given the sequential arrival order of the relations, this task is prone to two serious challenges, namely catastrophic forgetting and order-sensitivity. We propose a novel curriculum-meta learning method to tackle the above two challenges in continual relation extraction. We combine meta learning and curriculum learning to quickly adapt model parameters to a new task and to reduce interference of previously seen tasks on the current task. We design a novel relation representation learning method through the distribution of domain and range types of relations. Such representations are utilized to quantify the difficulty of tasks for the construction of curricula. Moreover, we also present novel difficulty-based metrics to quantitatively measure the extent of order-sensitivity of a given model, suggesting new ways to evaluate model robustness. Our comprehensive experiments on three benchmark datasets show that our proposed method outperforms the state-of-the-art techniques. The code is available at the anonymous GitHub repository: https://github.com/wutong8023/AAAI_CML. | ['Guoqiang Xu', 'Yujin Zhu', 'Guilin Qi', 'Reza Haffari', 'Yuan-Fang Li', 'Xuekai Li', 'Tongtong Wu'] | 2021-01-06 | null | null | null | null | ['continual-relation-extraction'] | ['natural-language-processing'] | [ 2.61464659e-02 1.79767415e-01 -3.74513090e-01 -1.38779134e-01
-7.48827577e-01 -4.28919435e-01 3.78557175e-01 6.22318685e-01
-3.40451151e-01 9.89126503e-01 1.31740332e-01 -5.14806569e-01
-5.52038789e-01 -8.35468948e-01 -7.76790440e-01 -4.02196646e-01
-5.83862537e-04 5.47308445e-01 4.00821567e-01 -5.31336784e-01
3.23306322e-01 2.76042819e-01 -1.53700101e+00 3.13906163e-01
1.13733864e+00 6.45657122e-01 2.37471297e-01 4.74200189e-01
-8.70324522e-02 9.44281399e-01 -6.23177230e-01 -5.64841330e-01
-4.21450734e-02 -9.68210176e-02 -1.23561811e+00 -2.75258332e-01
1.60033524e-01 1.55677542e-01 -5.24405360e-01 1.06940830e+00
4.21824187e-01 2.58630097e-01 6.66622460e-01 -1.21687555e+00
-5.77407897e-01 9.65364575e-01 -5.53679109e-01 7.34452486e-01
4.12097961e-01 -2.72562623e-01 1.00067043e+00 -7.70540535e-01
6.27161264e-01 9.68362451e-01 3.92290294e-01 3.38828385e-01
-9.12681639e-01 -8.91528487e-01 6.53224111e-01 7.06488967e-01
-1.38120723e+00 -3.44647527e-01 8.96847725e-01 -5.19603133e-01
9.44227874e-01 3.05092096e-01 3.57194781e-01 1.05826473e+00
1.42136246e-01 8.64751995e-01 9.09850478e-01 -5.67454815e-01
7.42914677e-02 1.82334393e-01 5.45728862e-01 7.23067284e-01
6.53594911e-01 -1.66806921e-01 -8.42540026e-01 -6.63195401e-02
2.56363153e-01 -3.20901312e-02 -2.37269208e-01 -2.72038251e-01
-9.01146173e-01 4.20240253e-01 1.58479288e-01 3.69515181e-01
-5.84837981e-02 -2.38217860e-01 3.29789340e-01 5.34991086e-01
5.01406431e-01 6.74492717e-01 -9.74795163e-01 -1.71390831e-01
-3.54419678e-01 4.03639317e-01 7.79153824e-01 1.27070487e+00
7.16388643e-01 -4.73411173e-01 -1.22500956e-01 8.10755014e-01
-2.51486301e-01 -1.45596355e-01 5.93699038e-01 -4.33475316e-01
8.84411752e-01 9.94878054e-01 -1.06577650e-01 -9.51090515e-01
-4.47937965e-01 -5.16891479e-01 -7.98014462e-01 -3.18799496e-01
4.13351089e-01 -2.04385832e-01 -6.12275004e-01 1.59338188e+00
6.92381442e-01 2.05050230e-01 -3.75790112e-02 3.55272651e-01
9.88607049e-01 4.41803128e-01 2.16703624e-01 -4.93823141e-01
1.51681697e+00 -9.64194953e-01 -8.74328732e-01 -3.03294867e-01
7.88218141e-01 -7.47537434e-01 1.01276112e+00 5.03285825e-01
-9.85733092e-01 -5.12864888e-01 -1.12582672e+00 -1.73238471e-01
-6.41887426e-01 -1.71102583e-01 8.93117368e-01 3.44048321e-01
-3.83695126e-01 7.02205896e-01 -6.83183908e-01 -1.16537958e-01
3.04159731e-01 3.26442361e-01 -6.69631958e-02 -1.17476217e-01
-1.49815667e+00 9.79997575e-01 6.94429815e-01 -1.00689568e-01
-3.91052753e-01 -1.01633811e+00 -7.64016807e-01 1.28208801e-01
1.09262335e+00 -6.29840136e-01 1.21056676e+00 -1.58256397e-01
-1.12057793e+00 4.48923886e-01 -1.16397440e-01 -2.88043648e-01
3.92697334e-01 -5.56270242e-01 -4.43427414e-01 -2.20329493e-01
-6.33870661e-02 -1.54364295e-02 5.37653267e-01 -1.20625353e+00
-8.36864114e-01 -3.25812697e-01 2.30178148e-01 4.21732336e-01
-5.75610936e-01 -2.22339541e-01 -4.34447020e-01 -7.36397088e-01
2.73248702e-01 -6.83702767e-01 -1.25201404e-01 -7.64470935e-01
-4.33030546e-01 -6.55782640e-01 5.71710169e-01 -3.74467105e-01
1.82181084e+00 -1.79373240e+00 3.36001903e-01 -9.94821042e-02
4.18846160e-01 3.35383803e-01 3.60901318e-02 4.31011528e-01
-2.60158718e-01 2.21368477e-01 -5.46814827e-03 -2.30271816e-01
-3.74209911e-01 6.42829924e-04 -2.05996901e-01 -4.19849604e-02
1.96832374e-01 7.42259622e-01 -1.22325766e+00 -5.73281646e-01
-1.45789266e-01 1.21889189e-01 -2.12318331e-01 3.57957453e-01
-2.00304002e-01 3.00841451e-01 -4.96941954e-01 7.49515414e-01
3.77779663e-01 -2.82042712e-01 3.89153719e-01 -5.05189151e-02
1.39792353e-01 6.55013263e-01 -1.28234041e+00 1.51491177e+00
-4.69134927e-01 8.67159739e-02 -6.21742368e-01 -1.07911670e+00
8.17514300e-01 2.07471669e-01 2.28915021e-01 -5.88746428e-01
1.06649272e-01 1.20100237e-01 1.10384107e-01 -6.80981994e-01
6.59061134e-01 1.08668804e-01 -1.75207302e-01 3.15291256e-01
3.29467386e-01 -7.41376728e-02 5.72790325e-01 3.57551336e-01
1.21073472e+00 1.99319776e-02 6.21897817e-01 -2.37015918e-01
5.51055789e-01 -8.49762782e-02 8.07249904e-01 6.65143371e-01
-4.76940311e-02 6.75270930e-02 6.42683089e-01 -8.07276428e-01
-5.55323184e-01 -5.44583440e-01 8.38823989e-02 1.23958433e+00
9.62832645e-02 -7.82420337e-01 -2.69174933e-01 -1.01085305e+00
6.18752576e-02 7.20935106e-01 -6.90425396e-01 -4.88238305e-01
-5.96584141e-01 -8.38866651e-01 1.30288750e-01 4.06328887e-01
1.49347931e-01 -8.44335318e-01 -2.97517955e-01 1.36897996e-01
-4.06969458e-01 -1.09093165e+00 -2.73519099e-01 4.44685519e-01
-9.32852983e-01 -1.29450130e+00 -3.39784026e-01 -8.96018445e-01
6.77220941e-01 2.79303789e-01 1.22282207e+00 3.05699915e-01
-1.93133131e-01 -1.80098265e-02 -5.22551596e-01 -5.07001460e-01
-3.32872868e-01 7.47958481e-01 2.01753765e-01 -3.94536018e-01
4.67900902e-01 -7.03526080e-01 -4.07711774e-01 1.96787030e-01
-7.78573096e-01 1.59753472e-01 5.23633897e-01 8.82848740e-01
5.11557281e-01 6.22858584e-01 7.28745282e-01 -1.42202103e+00
9.02789354e-01 -6.55360997e-01 -4.89225209e-01 6.27366543e-01
-1.08954585e+00 2.15233460e-01 5.72955370e-01 -7.83119619e-01
-1.17193580e+00 2.19832920e-02 2.73790866e-01 -1.28486112e-01
9.36350226e-03 7.07003772e-01 -2.50005633e-01 1.94459379e-01
6.51799619e-01 2.71576941e-02 -5.70026219e-01 -4.40103680e-01
3.61916780e-01 2.68510967e-01 4.46270436e-01 -1.00462365e+00
9.17049527e-01 -1.55308377e-02 -3.35126407e-02 -4.60111678e-01
-1.35412169e+00 -5.52577257e-01 -7.76814401e-01 -3.07036191e-02
1.63833618e-01 -8.50045443e-01 -6.51046693e-01 3.98378849e-01
-1.03882265e+00 -3.47694695e-01 -2.84231782e-01 1.15315057e-01
-2.38701299e-01 3.48815441e-01 -6.19556904e-01 -7.58165121e-01
-3.12158495e-01 -9.38437581e-01 5.41259170e-01 4.36332256e-01
-2.95402437e-01 -9.82442200e-01 1.48433566e-01 4.88643467e-01
1.76945073e-03 1.13586940e-01 1.28344047e+00 -7.76797295e-01
-4.43970233e-01 -1.27004325e-01 9.78437886e-02 -6.48761094e-02
2.95985073e-01 -1.67825669e-01 -7.88976312e-01 -1.68100089e-01
-3.50132771e-02 -3.56092453e-01 1.02574599e+00 -6.31083921e-02
1.28587401e+00 -5.66296339e-01 -4.30184275e-01 4.33183789e-01
1.16173661e+00 2.40697131e-01 4.09237236e-01 5.93917072e-01
8.24202597e-01 6.03482068e-01 9.63303566e-01 3.77375931e-01
7.84826994e-01 4.56057519e-01 2.11901274e-02 2.48460203e-01
-7.77491555e-02 -4.29920763e-01 -7.29860216e-02 1.13968098e+00
-2.61659801e-01 -1.35466367e-01 -1.12850201e+00 6.93935812e-01
-2.03632641e+00 -6.93790853e-01 9.79254693e-02 2.21869445e+00
1.41298032e+00 6.22292459e-01 3.42594311e-02 4.90438491e-01
5.26492655e-01 -6.85143769e-02 -3.46211374e-01 -1.92502022e-01
1.21058904e-01 1.99405715e-01 2.96568185e-01 5.36976337e-01
-1.17995167e+00 1.24172473e+00 5.25852013e+00 9.36184883e-01
-6.15401447e-01 4.43902388e-02 7.57786095e-01 9.68931429e-03
-2.44121805e-01 6.44908547e-02 -1.14217222e+00 2.11280957e-01
8.35022807e-01 -5.85268438e-01 2.44382679e-01 7.21019268e-01
-3.97711992e-01 -1.38652503e-01 -1.19085670e+00 6.46499038e-01
-6.67781755e-02 -1.24780917e+00 8.40625539e-02 -2.28127345e-01
7.99110591e-01 -5.73515594e-01 7.64602721e-02 7.02918768e-01
3.79926264e-01 -6.36939287e-01 3.66314679e-01 4.91460890e-01
6.36825740e-01 -1.04490030e+00 7.05411077e-01 3.90784740e-01
-1.31723011e+00 -1.57673210e-01 -2.31304392e-01 -3.10267299e-01
-3.12806755e-01 8.02181244e-01 -9.51788127e-01 7.30194032e-01
7.12339342e-01 5.06664932e-01 -1.00965536e+00 8.36635172e-01
-5.21767020e-01 5.26611269e-01 -2.16661934e-02 -1.32964522e-01
-4.62655097e-01 3.07434708e-01 5.14538467e-01 1.04763722e+00
-2.77804537e-03 4.11015868e-01 1.01375870e-01 5.16153455e-01
-1.65459290e-01 2.20099375e-01 -5.33570170e-01 7.52918124e-02
9.10810351e-01 1.25766516e+00 -6.89332426e-01 -3.16317439e-01
-1.69588491e-01 6.17492616e-01 9.07645404e-01 1.89130500e-01
-4.97038782e-01 -4.46089536e-01 3.92031282e-01 -3.01131587e-02
6.85727000e-02 -2.80160904e-01 -4.16279942e-01 -1.27126884e+00
1.30408168e-01 -1.02876294e+00 7.93378115e-01 -3.13070595e-01
-1.19467473e+00 6.25675380e-01 2.85225183e-01 -1.05489278e+00
2.51289029e-02 -2.61201531e-01 -4.57566142e-01 4.32709515e-01
-1.51164651e+00 -8.36251438e-01 -2.51616806e-01 3.85274231e-01
7.45392501e-01 -2.37827629e-01 7.41895854e-01 3.86818796e-01
-9.76127744e-01 9.70428348e-01 -2.59760618e-01 -2.17567850e-02
7.64998853e-01 -1.30897582e+00 4.20531273e-01 9.16482210e-01
8.19439739e-02 8.37319970e-01 6.99661314e-01 -9.11525249e-01
-1.28525889e+00 -9.40889657e-01 1.22560334e+00 -5.92385054e-01
7.42097139e-01 -3.81975859e-01 -1.10574162e+00 7.02285409e-01
-1.59876049e-01 -6.37409016e-02 7.23976791e-01 7.67217934e-01
-5.30517638e-01 -2.64633477e-01 -8.23480308e-01 6.17105305e-01
1.23473012e+00 -1.63191602e-01 -7.56572664e-01 4.51599896e-01
1.17590988e+00 -7.00442314e-01 -8.69098067e-01 6.52203918e-01
3.57068211e-01 -5.83572686e-01 9.26037788e-01 -8.20937335e-01
6.05707169e-01 -1.00102998e-01 1.79191485e-01 -1.30131030e+00
-6.00894809e-01 -7.86222339e-01 -9.02512074e-01 1.46579814e+00
6.54441178e-01 -5.08585095e-01 6.46813631e-01 5.27875185e-01
1.75352991e-01 -1.23446679e+00 -7.54070699e-01 -8.49497020e-01
1.94806397e-01 -2.03653082e-01 8.31509709e-01 1.16212106e+00
2.98170894e-01 6.97384119e-01 -2.26144701e-01 2.62477726e-01
5.33694685e-01 1.38778552e-01 5.66598892e-01 -1.28236306e+00
-1.89332560e-01 -3.29100877e-01 -1.46010578e-01 -7.06035078e-01
3.97650339e-02 -7.98098087e-01 -2.06993073e-01 -1.28447640e+00
3.18467140e-01 -6.76272810e-01 -7.52191663e-01 5.91854393e-01
-8.74742985e-01 -4.60049689e-01 -4.34322022e-02 1.90589011e-01
-8.39113951e-01 5.12055576e-01 1.24329090e+00 -1.05081633e-01
-4.28336650e-01 1.98348328e-01 -1.07138741e+00 6.25801444e-01
1.07055640e+00 -7.65072584e-01 -6.46735966e-01 -3.03743452e-01
5.71322620e-01 -6.52758107e-02 -1.43211246e-01 -9.90205884e-01
5.32163620e-01 -4.48904932e-01 1.60190463e-01 -6.13122761e-01
1.70966480e-02 -6.10297263e-01 -1.50361985e-01 3.38046074e-01
-4.68488127e-01 2.06308722e-01 3.49865496e-01 6.35641575e-01
6.07692748e-02 -1.84305727e-01 4.29412663e-01 -1.78938180e-01
-6.15639865e-01 2.31715113e-01 2.11845543e-02 4.11731124e-01
9.26350534e-01 3.66039157e-01 -5.51353157e-01 -1.09290548e-01
-5.88823140e-01 3.88260841e-01 -1.15524709e-01 7.01245785e-01
6.21766448e-01 -1.13461912e+00 -6.42231047e-01 -3.12398169e-02
2.95388103e-01 4.71978903e-01 2.25558922e-01 5.77795625e-01
-4.36520837e-02 2.26069048e-01 4.35863808e-02 -8.95219296e-02
-1.28193259e+00 8.02745759e-01 9.15790573e-02 -8.19879770e-01
-4.10121322e-01 1.19186831e+00 -1.27681168e-02 -4.19583917e-01
4.97796178e-01 -4.63966042e-01 -4.58254635e-01 2.33775675e-01
4.65713084e-01 3.40338528e-01 3.94719601e-01 3.53189483e-02
-2.62160838e-01 1.43684998e-01 -7.65764058e-01 2.64287382e-01
1.37536550e+00 -9.91203338e-02 9.92647372e-03 6.39127851e-01
6.54338062e-01 -1.04363032e-01 -8.53552282e-01 -6.41638398e-01
5.46669245e-01 -3.13222468e-01 -2.13776931e-01 -8.83448243e-01
-8.44049513e-01 4.24972534e-01 2.03114256e-01 1.57981738e-01
1.13492596e+00 5.81444539e-02 7.60109782e-01 5.95316172e-01
3.41831356e-01 -1.12997437e+00 2.60206193e-01 7.37789273e-01
8.45104694e-01 -1.21309257e+00 6.05985582e-01 -8.40247869e-01
-4.65987504e-01 9.70669985e-01 1.04872453e+00 3.05119812e-01
7.48486280e-01 3.30013633e-01 -1.48639604e-01 -2.26752073e-01
-1.13998592e+00 -2.39747152e-01 3.81472528e-01 4.72870827e-01
5.42274117e-01 6.76707402e-02 -4.62935805e-01 1.00371146e+00
-4.20996696e-01 -7.96170160e-02 4.67312187e-01 1.15991735e+00
-2.25107670e-01 -1.41218734e+00 -6.35299832e-02 4.17148948e-01
-3.15656900e-01 -2.20467374e-01 -3.88234168e-01 6.83846235e-01
1.72410712e-01 8.27385306e-01 -3.99451584e-01 -7.02337742e-01
5.64308286e-01 3.38137537e-01 5.12944162e-01 -8.17724109e-01
-4.72995222e-01 -1.92444906e-01 2.08885238e-01 -8.87821168e-02
-1.18216902e-01 -6.19756997e-01 -1.08084691e+00 -1.36333197e-01
-4.93952155e-01 1.84558719e-01 2.60842413e-01 8.77394855e-01
3.31114829e-01 9.01558340e-01 4.56299007e-01 -2.63156384e-01
-5.68062484e-01 -9.32897508e-01 -2.31941268e-01 3.70484620e-01
1.99246272e-01 -1.04803431e+00 -2.56312042e-01 -4.86680865e-02] | [9.182561874389648, 8.528240203857422] |
107e25b1-0a26-45f6-a87e-ff3372b81535 | ts-moco-time-series-momentum-contrast-for | 2306.06522 | null | https://arxiv.org/abs/2306.06522v1 | https://arxiv.org/pdf/2306.06522v1.pdf | TS-MoCo: Time-Series Momentum Contrast for Self-Supervised Physiological Representation Learning | Limited availability of labeled physiological data often prohibits the use of powerful supervised deep learning models in the biomedical machine intelligence domain. We approach this problem and propose a novel encoding framework that relies on self-supervised learning with momentum contrast to learn representations from multivariate time-series of various physiological domains without needing labels. Our model uses a transformer architecture that can be easily adapted to classification problems by optimizing a linear output classification layer. We experimentally evaluate our framework using two publicly available physiological datasets from different domains, i.e., human activity recognition from embedded inertial sensory and emotion recognition from electroencephalography. We show that our self-supervised learning approach can indeed learn discriminative features which can be exploited in downstream classification tasks. Our work enables the development of domain-agnostic intelligent systems that can effectively analyze multivariate time-series data from physiological domains. | ['Enkelejda Kasneci', 'Tobias Grosse-Puppendahl', 'Ozan Özdenizci', 'David Bethge', 'Philipp Hallgarten'] | 2023-06-10 | null | null | null | null | ['activity-recognition', 'human-activity-recognition', 'human-activity-recognition'] | ['computer-vision', 'computer-vision', 'time-series'] | [ 5.70478439e-01 1.49477839e-01 -5.12816608e-02 -8.04923058e-01
-5.88867664e-01 -4.77394044e-01 4.23386931e-01 1.64546072e-01
-5.52083611e-01 1.04173255e+00 8.47916082e-02 -1.25681773e-01
-3.02501917e-01 -4.28540200e-01 -6.46423221e-01 -8.24993014e-01
-2.60460347e-01 2.68209130e-01 -3.55072260e-01 6.74791783e-02
-1.46545935e-02 4.91691738e-01 -1.27651525e+00 3.04330617e-01
7.96010017e-01 1.44191265e+00 1.22506030e-01 4.16214317e-01
2.88110375e-01 7.85558820e-01 -4.61839974e-01 3.46344084e-01
4.71522324e-02 -5.41099906e-01 -6.14778340e-01 -2.16329738e-01
-1.83566600e-01 5.05696274e-02 -2.59346038e-01 7.33954728e-01
6.99081540e-01 4.29526903e-02 8.32100868e-01 -1.04446208e+00
-3.93732965e-01 2.81675696e-01 9.37659591e-02 5.86085677e-01
1.73034161e-01 1.81768313e-01 7.59523451e-01 -4.54877168e-01
4.03729916e-01 5.24362981e-01 7.81216204e-01 7.06970751e-01
-1.60363364e+00 -6.08610690e-01 -1.38403997e-01 3.05487722e-01
-9.42512035e-01 -6.01820827e-01 1.07359290e+00 -4.71892715e-01
1.27379918e+00 6.78316578e-02 7.16762424e-01 1.87152755e+00
5.02222419e-01 5.48990905e-01 1.49223459e+00 -6.30248636e-02
7.74613619e-01 1.16449721e-01 1.47337869e-01 4.85197604e-01
-4.98151332e-02 3.23381603e-01 -8.65724504e-01 -2.35564128e-01
5.44805944e-01 1.79782599e-01 -2.15324193e-01 -3.92607540e-01
-1.34562719e+00 5.41201770e-01 3.72597098e-01 5.14211118e-01
-7.86624193e-01 1.91411659e-01 6.68564856e-01 6.58772886e-01
6.18060827e-01 8.70347977e-01 -9.99727547e-01 -3.52024466e-01
-9.83659208e-01 -1.98147044e-01 7.09321499e-01 4.83939767e-01
3.76287937e-01 2.78283179e-01 2.81058133e-01 5.80109000e-01
2.61215985e-01 3.93587947e-01 1.06586313e+00 -7.81197846e-01
-1.22551262e-01 4.68438834e-01 -1.37466297e-01 -6.15223587e-01
-9.82028008e-01 -5.28424740e-01 -9.16679859e-01 9.18549672e-02
1.76926896e-01 -4.52482849e-01 -5.16214907e-01 1.83758986e+00
-7.66475797e-02 4.43570256e-01 3.94503713e-01 9.33252394e-01
6.01065040e-01 3.07902843e-01 1.73358187e-01 -3.79242539e-01
1.12263834e+00 -3.69954169e-01 -6.83156371e-01 -3.23861301e-01
4.88338917e-01 1.94106326e-01 9.47942495e-01 6.09764934e-01
-7.84993768e-01 -3.30952436e-01 -1.13091230e+00 1.12783089e-01
-5.34006476e-01 1.59901142e-01 8.84713709e-01 3.94794405e-01
-8.77278686e-01 9.87699032e-01 -1.45407271e+00 -4.04386133e-01
5.80481589e-01 7.23673224e-01 -5.78351259e-01 5.93139708e-01
-1.08725452e+00 9.81126428e-01 3.89717281e-01 7.56364316e-02
-1.07971704e+00 -5.82122505e-01 -7.98822820e-01 1.42979354e-01
-2.14210644e-01 -5.83942831e-01 8.62364233e-01 -1.27927947e+00
-1.81764281e+00 9.00463879e-01 3.07003576e-02 -7.83184469e-01
-4.35729846e-02 -8.96010399e-02 -4.67993826e-01 2.76126653e-01
-7.15130121e-02 3.18954527e-01 1.20191669e+00 -4.38990533e-01
7.24952072e-02 -6.05987012e-01 -4.42711473e-01 -1.58590719e-01
-8.00234675e-01 -2.44069815e-01 6.12241030e-01 -5.71895599e-01
1.57718331e-01 -7.59161711e-01 -1.88895106e-01 -1.39760301e-01
-1.51155710e-01 1.09213434e-01 4.81690049e-01 -3.22148979e-01
6.63221002e-01 -2.27831888e+00 5.00601649e-01 1.53202310e-01
2.86822617e-01 -7.24967718e-02 -1.05484411e-01 1.92206651e-01
-5.44294596e-01 -3.56263310e-01 -3.54686588e-01 -1.15071326e-01
8.22390802e-03 2.77706683e-01 -5.30301869e-01 6.65538192e-01
6.56387985e-01 1.04231036e+00 -9.55775619e-01 -5.60629144e-02
1.24234065e-01 5.51660657e-01 -3.49918574e-01 5.52749455e-01
-1.61088988e-01 9.90852237e-01 -5.43000877e-01 5.69902241e-01
7.23802373e-02 -4.67000574e-01 2.60606587e-01 -3.07346165e-01
5.59744239e-02 7.06361115e-01 -3.93867671e-01 2.14561152e+00
-6.03981912e-01 6.67020679e-01 -2.14121521e-01 -1.86810851e+00
9.98494208e-01 4.93900448e-01 9.10074770e-01 -8.83515656e-01
3.81805927e-01 1.10750474e-01 -1.28293652e-02 -7.85023034e-01
-3.62447262e-01 -4.01919514e-01 -3.06970805e-01 5.56248665e-01
8.27579319e-01 9.15487409e-02 -3.73160183e-01 -3.42976838e-01
1.54297364e+00 3.18809062e-01 2.10218787e-01 -5.21059573e-01
4.09850985e-01 -1.41225934e-01 5.47603965e-01 6.06873989e-01
-2.88686752e-01 1.64608687e-01 4.39273208e-01 -6.22486234e-01
-7.72060990e-01 -1.17866850e+00 -4.43396777e-01 1.22689474e+00
-3.03700477e-01 -1.97329521e-01 -3.75740737e-01 -5.41865349e-01
-1.20408267e-01 3.16579282e-01 -7.27181852e-01 -6.12156689e-01
-1.77148521e-01 -9.94827807e-01 6.22329473e-01 8.01406503e-01
-1.20772272e-02 -1.21693194e+00 -1.10743380e+00 4.71242398e-01
2.22449839e-01 -1.10730636e+00 4.65705603e-01 1.19389629e+00
-1.29991710e+00 -8.47054660e-01 -4.31145310e-01 -6.20018303e-01
5.12909472e-01 -6.42508984e-01 8.90200198e-01 -6.05939209e-01
-5.11874855e-01 7.43723869e-01 -1.65540233e-01 -5.88508725e-01
-4.53652181e-02 2.74708867e-01 6.10072792e-01 1.32542759e-01
4.63138700e-01 -1.30185974e+00 -6.33770943e-01 -4.03949618e-02
-7.26412594e-01 -2.12730139e-01 4.32755142e-01 1.14349544e+00
6.20629132e-01 -3.48794937e-01 1.24796665e+00 -5.02294123e-01
6.28944695e-01 -8.36054623e-01 -4.12592143e-01 -1.61434188e-02
-5.08402884e-01 4.51629281e-01 1.10728538e+00 -5.23295999e-01
-6.67252839e-01 3.19190979e-01 -1.30160272e-01 -2.89022177e-01
-4.21958327e-01 6.79293394e-01 5.93466032e-03 -1.24243729e-01
8.32600474e-01 4.05560613e-01 7.83388540e-02 -3.97485346e-01
4.82501425e-02 6.83260381e-01 7.00398743e-01 -5.34913182e-01
1.88130036e-01 5.08877754e-01 4.99185100e-02 -7.16440797e-01
-6.71945930e-01 -2.95090348e-01 -6.24702513e-01 -4.89442237e-02
8.59801710e-01 -1.02474844e+00 -7.69761503e-01 2.86596775e-01
-6.93990827e-01 -4.35469478e-01 -4.41024780e-01 8.12580407e-01
-1.09982443e+00 4.24146019e-02 -7.62675643e-01 -7.02558935e-01
-5.38997889e-01 -6.51792169e-01 1.06185889e+00 2.45994758e-02
-4.34113473e-01 -1.10727537e+00 3.81071210e-01 -1.78777128e-01
6.56919122e-01 3.60159010e-01 7.32304156e-01 -1.01095641e+00
5.40365800e-02 -1.09822214e-01 3.17444980e-01 4.53080356e-01
2.27284417e-01 -6.67457700e-01 -1.45736086e+00 -2.46670842e-01
5.23557663e-01 -9.42249954e-01 8.48166823e-01 2.68348157e-01
1.36417937e+00 -1.79036021e-01 -2.86801100e-01 8.90824378e-01
1.02202010e+00 4.64679226e-02 4.70260084e-01 2.10563704e-01
3.47378433e-01 4.72117543e-01 1.92539379e-01 5.90950251e-01
1.72095492e-01 3.08625996e-01 1.79484412e-01 -2.79813334e-02
6.12333953e-01 8.04349557e-02 4.62143391e-01 9.98295605e-01
-7.91433677e-02 2.84141660e-01 -8.72699320e-01 3.31447899e-01
-1.80677843e+00 -9.12168026e-01 5.70539296e-01 1.87455738e+00
9.63221908e-01 -6.03438402e-03 -1.07964613e-01 3.92546862e-01
4.16733585e-02 -8.25099945e-02 -9.67550576e-01 -3.63439649e-01
8.08549076e-02 8.01683962e-01 4.38834168e-02 -1.40584543e-01
-1.36157596e+00 5.33743560e-01 6.92003822e+00 -4.67108265e-02
-1.64463544e+00 2.77372032e-01 5.34115434e-01 -3.35283339e-01
2.06906304e-01 -4.06015307e-01 -1.81762442e-01 4.38288063e-01
1.71453416e+00 -2.52773494e-01 5.66977143e-01 8.04292977e-01
1.27448276e-01 2.24913642e-01 -1.61009586e+00 1.31633878e+00
-9.29548293e-02 -1.03510213e+00 -7.02546537e-01 -3.11739951e-01
2.98610568e-01 4.83206838e-01 1.28851011e-01 3.99432808e-01
-1.97227951e-02 -1.12131429e+00 3.12667996e-01 8.63233387e-01
7.35406458e-01 -3.10307473e-01 4.31291401e-01 4.24163193e-01
-5.37238598e-01 -2.74305433e-01 -2.20413417e-01 -4.30385470e-01
-1.53327644e-01 6.92787230e-01 -6.12684309e-01 1.58931911e-01
7.49760747e-01 1.32788146e+00 -3.58120590e-01 6.42742693e-01
-5.86797744e-02 9.02092755e-01 -4.09435153e-01 5.57961725e-02
-4.22348566e-02 -4.86000180e-02 2.18888193e-01 1.04675734e+00
3.63892436e-01 1.65837154e-01 -9.50785205e-02 9.26613450e-01
1.08901203e-01 -1.13832116e-01 -7.96456516e-01 -4.57536578e-01
-7.33970944e-03 1.25685251e+00 -5.50526798e-01 -2.95247108e-01
-1.34802952e-01 1.15594089e+00 5.14634132e-01 3.26269358e-01
-7.78437436e-01 -2.85414875e-01 5.97127020e-01 -2.91535705e-01
2.33495563e-01 -3.08134347e-01 -3.06426585e-01 -1.72034931e+00
-1.21514425e-02 -6.32184267e-01 2.05259100e-01 -7.41966128e-01
-1.61201012e+00 8.58510017e-01 -2.30975717e-01 -1.24675941e+00
-7.30005860e-01 -9.05408502e-01 -4.27406996e-01 5.45964479e-01
-1.55005980e+00 -6.11534417e-01 -9.90099311e-02 9.46489275e-01
-1.42716020e-01 -3.18762720e-01 1.47596431e+00 2.44985610e-01
-5.50451458e-01 2.24929810e-01 2.79914021e-01 -4.78043780e-02
5.90809464e-01 -1.27167201e+00 -1.41031472e-02 3.53774220e-01
2.22878739e-01 6.02795243e-01 4.63643134e-01 -6.95589855e-02
-1.59708774e+00 -1.00748098e+00 5.84233105e-01 -4.19315368e-01
9.21654522e-01 -8.08361650e-01 -8.74190092e-01 6.77749395e-01
1.20015088e-02 5.10170877e-01 1.28009641e+00 1.16860867e-01
-2.87421972e-01 -3.51788968e-01 -1.11730516e+00 3.71890999e-02
9.57304001e-01 -1.07846296e+00 -9.33893919e-01 4.28113192e-01
1.09610528e-01 -6.32868484e-02 -1.21003377e+00 4.15161133e-01
6.13478899e-01 -6.53461874e-01 8.02551746e-01 -9.99168992e-01
2.69017756e-01 8.04321170e-02 1.54941408e-02 -1.60392368e+00
-1.11326367e-01 -5.83420396e-01 -4.50031430e-01 6.12599909e-01
3.20523888e-01 -9.09594774e-01 5.20263255e-01 5.67302465e-01
-9.44756866e-02 -6.89779937e-01 -1.16933596e+00 -7.07075000e-01
-6.35853410e-02 -5.81975341e-01 3.15467894e-01 1.06751013e+00
8.04633141e-01 5.06253421e-01 -1.87740192e-01 -4.34851721e-02
3.91380072e-01 2.12405547e-01 4.76387218e-02 -1.34961689e+00
-5.22206306e-01 -4.26337086e-02 -8.70405912e-01 -5.42715251e-01
6.72678292e-01 -1.17926288e+00 9.70950127e-02 -9.33791637e-01
-1.75399572e-01 -3.37928027e-01 -1.10308456e+00 9.56535757e-01
3.70237231e-01 3.42363268e-01 -3.02828580e-01 3.62966657e-02
-5.82175612e-01 7.27135658e-01 4.80249643e-01 -2.60279298e-01
-2.12451935e-01 -3.11214566e-01 -5.23509860e-01 7.62116313e-01
9.72727537e-01 -6.41863167e-01 -4.43172693e-01 -2.94153899e-01
1.87586382e-01 1.66251570e-01 6.07873678e-01 -1.40848541e+00
8.07492509e-02 1.94261834e-01 8.38861883e-01 2.87454814e-01
3.27628225e-01 -9.31265295e-01 -1.27908126e-01 4.56364810e-01
-6.89095795e-01 -5.61182499e-02 3.68018001e-01 5.05883217e-01
-1.87188208e-01 2.45013714e-01 3.62049907e-01 -7.85203651e-03
-5.92929482e-01 1.04838483e-01 -6.92083180e-01 -9.08203498e-02
9.37796474e-01 1.68802142e-01 -7.41504133e-02 -1.26350775e-01
-1.32830429e+00 -8.67810175e-02 1.16786867e-01 2.99723595e-01
6.40513778e-01 -1.29872930e+00 -4.83409733e-01 7.33193099e-01
2.80586571e-01 -6.15535080e-01 -1.39436737e-01 1.08550191e+00
8.16912502e-02 3.92163128e-01 -7.94491768e-01 -8.30916345e-01
-5.39615452e-01 5.14683485e-01 4.40454572e-01 -1.18249908e-01
-6.67835832e-01 4.66132343e-01 -1.09143883e-01 -5.00794351e-01
1.45120427e-01 -5.42680383e-01 -2.40993932e-01 -7.34728128e-02
5.13217151e-01 -1.83333963e-01 3.51767361e-01 -2.57014602e-01
-6.19445682e-01 -3.49499658e-02 1.22164860e-01 -1.29186464e-02
1.80559111e+00 2.26935923e-01 -2.03842167e-02 9.41335142e-01
1.36809659e+00 -7.04171896e-01 -1.23853123e+00 7.12600499e-02
2.59210199e-01 1.71257764e-01 1.81463614e-01 -9.13801014e-01
-9.60278630e-01 1.02258611e+00 9.47528481e-01 4.54741344e-03
1.55834389e+00 -1.15003325e-01 5.85558891e-01 9.66832340e-01
6.79933310e-01 -1.01125109e+00 9.07275379e-02 3.41612458e-01
6.86899364e-01 -1.29626250e+00 -2.93375105e-01 3.21663529e-01
-4.79959995e-01 1.21317220e+00 2.45023847e-01 -5.60386539e-01
7.77767897e-01 6.95584714e-01 1.68823041e-02 -3.12416315e-01
-1.08495021e+00 -8.64567459e-02 2.01349750e-01 6.93265557e-01
5.25308967e-01 -5.00218607e-02 -3.00201565e-01 1.27459109e+00
8.00288320e-02 4.31503206e-01 1.28506377e-01 1.07948828e+00
-6.34883344e-02 -9.65234458e-01 1.57037610e-03 7.74341583e-01
-4.68114644e-01 -1.50925100e-01 -2.50502020e-01 1.06286421e-01
-1.27881587e-01 5.49840569e-01 1.05404131e-01 -5.01257122e-01
2.46702526e-02 5.83884239e-01 6.57256424e-01 -5.24648786e-01
-5.83646059e-01 -7.20434710e-02 -2.31016830e-01 -8.66208196e-01
-7.74771035e-01 -5.34159660e-01 -1.30415916e+00 4.83924925e-01
3.76794599e-02 1.04171708e-01 6.65338993e-01 1.05206728e+00
7.65564024e-01 6.70151651e-01 6.70912981e-01 -1.00944233e+00
-5.31556368e-01 -1.01718044e+00 -6.41763210e-01 4.15816158e-01
6.04469240e-01 -7.38707840e-01 -2.22502977e-01 2.51814008e-01] | [13.1885347366333, 3.4859421253204346] |
e7c53082-6d93-4090-a5dd-fb53a692cb8e | siamixformer-a-siamese-transformer-network | 2208.00657 | null | https://arxiv.org/abs/2208.00657v1 | https://arxiv.org/pdf/2208.00657v1.pdf | SiamixFormer: A Siamese Transformer Network For Building Detection And Change Detection From Bi-Temporal Remote Sensing Images | Building detection and change detection using remote sensing images can help urban and rescue planning. Moreover, they can be used for building damage assessment after natural disasters. Currently, most of the existing models for building detection use only one image (pre-disaster image) to detect buildings. This is based on the idea that post-disaster images reduce the model's performance because of presence of destroyed buildings. In this paper, we propose a siamese model, called SiamixFormer, which uses pre- and post-disaster images as input. Our model has two encoders and has a hierarchical transformer architecture. The output of each stage in both encoders is given to a temporal transformer for feature fusion in a way that query is generated from pre-disaster images and (key, value) is generated from post-disaster images. To this end, temporal features are also considered in feature fusion. Another advantage of using temporal transformers in feature fusion is that they can better maintain large receptive fields generated by transformer encoders compared with CNNs. Finally, the output of the temporal transformer is given to a simple MLP decoder at each stage. The SiamixFormer model is evaluated on xBD, and WHU datasets, for building detection and on LEVIR-CD and CDD datasets for change detection and could outperform the state-of-the-art. | ['Foad Ghaderi', 'Amir mohammadian'] | 2022-08-01 | null | null | null | null | ['2d-semantic-segmentation', 'change-detection-for-remote-sensing-images', 'building-change-detection-for-remote-sensing', 'extracting-buildings-in-remote-sensing-images'] | ['computer-vision', 'miscellaneous', 'miscellaneous', 'miscellaneous'] | [ 2.18949094e-01 -3.73239458e-01 4.14522260e-01 -2.64012039e-01
-6.90891206e-01 -5.08801006e-02 6.75413847e-01 4.81004894e-01
-6.21490180e-01 4.03925627e-01 3.72136414e-01 -2.19514985e-02
-4.89157476e-02 -1.68363953e+00 -5.95916629e-01 -8.87916446e-01
-2.26487398e-01 7.69420043e-02 5.61715364e-01 -5.46604633e-01
4.45002578e-02 5.13439298e-01 -1.79337215e+00 5.47770858e-01
6.39426529e-01 1.29373038e+00 8.13539267e-01 5.15857518e-01
8.77591446e-02 7.74611354e-01 -4.74624455e-01 1.82413206e-01
1.68009818e-01 -9.78789181e-02 -6.85994685e-01 -6.84142020e-03
1.55749246e-01 -7.31778145e-01 -5.50558686e-01 7.44672894e-01
7.87911355e-01 2.02818006e-01 5.30665755e-01 -9.33977962e-01
-8.81200969e-01 7.99582839e-01 -4.58819121e-01 5.34480333e-01
3.51239532e-01 1.03251345e-01 9.94970620e-01 -1.21442759e+00
2.92164564e-01 1.30440450e+00 5.70730448e-01 -7.55011812e-02
-9.34732676e-01 -5.87477624e-01 2.72191256e-01 4.23885971e-01
-1.71763933e+00 -2.29954943e-01 7.76102066e-01 -5.07794261e-01
1.15522778e+00 3.11802983e-01 9.32599545e-01 4.09374416e-01
1.36391371e-01 9.37915087e-01 7.04296827e-01 -2.46277019e-01
1.54181004e-01 -3.33536386e-01 -1.28799215e-01 8.31639469e-01
-1.71532854e-01 2.15206325e-01 -4.49185908e-01 2.25532830e-01
6.63724780e-01 6.08321428e-01 -5.11186182e-01 1.76335022e-01
-1.24714136e+00 8.60020518e-01 1.32065320e+00 6.18979812e-01
-8.68061304e-01 1.36787981e-01 7.62130842e-02 2.33878896e-01
4.32971328e-01 -6.03705905e-02 -1.01400837e-01 5.29230654e-01
-1.11531281e+00 1.53125674e-01 -1.94314849e-02 4.27819699e-01
9.56669152e-01 -7.79736862e-02 -4.71871763e-01 7.32898474e-01
5.15709519e-01 6.30129516e-01 5.69868207e-01 -3.07071298e-01
6.46021307e-01 8.42169642e-01 -1.91551507e-01 -1.22063828e+00
-5.34926414e-01 -5.38950384e-01 -1.21403050e+00 1.47310510e-01
-1.84922397e-01 3.42158258e-01 -1.19669569e+00 1.38556540e+00
1.38722971e-01 9.85532925e-02 6.03672154e-02 7.71405399e-01
1.02082050e+00 1.00962090e+00 2.65946295e-02 1.07578948e-01
1.42976940e+00 -7.17433631e-01 -3.86075556e-01 -3.82252604e-01
2.77398616e-01 -5.45370698e-01 1.01892054e+00 1.16769141e-02
-7.75295794e-01 -7.92973399e-01 -1.09895539e+00 -2.55319625e-01
-6.59015179e-01 3.28550249e-01 -6.42259941e-02 4.48639318e-03
-1.21642244e+00 4.34571445e-01 -7.92843699e-01 -6.28749907e-01
3.66162717e-01 2.75311381e-01 -2.90286779e-01 -1.33308396e-01
-1.34179592e+00 8.38454783e-01 5.10680258e-01 4.26160067e-01
-1.05324495e+00 -4.23493385e-01 -9.33798254e-01 3.86566341e-01
-5.92149757e-02 -5.93056738e-01 9.48133588e-01 -2.37869948e-01
-7.51257002e-01 6.80946469e-01 1.09942779e-01 -4.92926836e-01
2.23301947e-01 -1.05899028e-01 -2.90278822e-01 1.80478647e-01
3.68952036e-01 8.63360941e-01 7.53261983e-01 -1.19302475e+00
-1.07896292e+00 -4.93478537e-01 1.44476563e-01 3.47968161e-01
-3.26176822e-01 -2.52714992e-01 -3.54229510e-01 -7.30081856e-01
6.13457739e-01 -5.85856140e-01 -2.44323477e-01 -3.07789985e-02
-4.18173760e-01 -1.33679807e-01 1.17718482e+00 -9.07299519e-01
1.52537417e+00 -2.27058172e+00 -6.20755889e-02 2.77820140e-01
1.21916346e-01 2.04256728e-01 -1.47584736e-01 6.65110648e-01
-1.13527708e-01 1.51114538e-01 -5.86083710e-01 -1.61839142e-01
-2.38008916e-01 1.64879709e-01 -3.50131005e-01 2.67760754e-01
1.74164057e-01 8.09531629e-01 -6.77603364e-01 -5.27389467e-01
6.09991372e-01 7.01756120e-01 -2.45927855e-01 1.88812446e-02
2.36780450e-01 1.78864881e-01 -4.97188747e-01 6.54522300e-01
7.46545076e-01 1.51682878e-02 -3.84796858e-01 -2.46180058e-01
-5.69478452e-01 2.84050107e-01 -1.06111205e+00 1.33096540e+00
-4.81740117e-01 4.68654573e-01 -2.82439172e-01 -8.65177453e-01
1.02913129e+00 2.84964055e-01 4.61672574e-01 -9.82601821e-01
-2.73759905e-02 7.38995448e-02 -5.16875446e-01 -4.19602513e-01
4.58052635e-01 -1.64587051e-01 -1.20535411e-01 2.49011755e-01
-3.14049542e-01 -2.59788424e-01 2.17833191e-01 -4.67468798e-03
1.11716819e+00 -3.14669997e-01 1.38550088e-01 1.14217155e-01
7.99352705e-01 -2.08808050e-01 3.80942374e-01 6.50371969e-01
7.84001574e-02 6.92876816e-01 -1.74682617e-01 -6.76814556e-01
-7.62856543e-01 -1.14624381e+00 6.46943003e-02 1.06291258e+00
1.97623149e-02 -2.40152061e-01 -4.63743091e-01 -4.72656101e-01
-2.06807368e-02 7.09376931e-01 -5.92200279e-01 -2.74719298e-01
-5.90363145e-01 -6.86215222e-01 6.18764460e-01 6.86817050e-01
1.26408303e+00 -1.23186624e+00 -1.13034415e+00 5.30340374e-01
-6.33060813e-01 -5.83901227e-01 -2.52617776e-01 2.87527770e-01
-8.12965453e-01 -8.84722650e-01 -6.92773759e-01 -1.00457501e+00
5.64747930e-01 5.60620248e-01 8.57915998e-01 2.39730895e-01
-8.49601105e-02 7.68322498e-02 -5.15164196e-01 -3.00388932e-01
9.62048955e-03 4.10483256e-02 -3.34918737e-01 4.85380292e-02
1.12797007e-01 -6.37483239e-01 -7.53992260e-01 2.42289320e-01
-1.21094096e+00 8.59585181e-02 7.41946995e-01 7.09694207e-01
6.20799899e-01 5.91487646e-01 1.25114873e-01 -1.67932212e-01
4.79496151e-01 -3.19862068e-01 -4.35952008e-01 2.89715439e-01
-5.12668312e-01 6.18761070e-02 4.75410372e-01 8.80930722e-02
-1.03154457e+00 2.88750201e-01 -3.90948743e-01 -1.13377854e-01
-4.79314439e-02 8.24463665e-01 -1.31679207e-01 3.04738164e-01
6.25949979e-01 5.49137414e-01 -4.63115394e-01 -6.09359384e-01
-4.39384878e-02 7.76530683e-01 6.47817969e-01 -1.55623415e-02
9.43570137e-01 6.22555733e-01 -3.14869553e-01 -9.75058198e-01
-5.74137986e-01 -5.61855495e-01 -7.45880127e-01 -4.09001976e-01
9.94183302e-01 -1.16780889e+00 -3.73291522e-01 9.70075309e-01
-1.16612780e+00 -5.60086742e-02 -3.53232294e-01 4.34831828e-01
-1.73271805e-01 5.86870089e-02 -4.95856762e-01 -8.44205022e-01
-5.54379284e-01 -9.93461549e-01 1.29633057e+00 2.28954792e-01
3.92203450e-01 -5.46802223e-01 -1.83256175e-02 2.74803154e-02
5.03488541e-01 3.70936126e-01 8.26353073e-01 -5.77125214e-02
-5.79211414e-01 -1.20227858e-01 -2.34807730e-01 3.07580590e-01
3.05284888e-01 -1.75087407e-01 -1.00193954e+00 -2.83991784e-01
1.12970341e-02 -6.05884525e-05 1.47682977e+00 5.03938496e-01
8.32928717e-01 -3.94659907e-01 -3.81620646e-01 3.02979827e-01
1.65456474e+00 2.71738052e-01 9.17958200e-01 5.89479208e-01
5.19798279e-01 4.05609190e-01 6.41162634e-01 5.58810890e-01
8.20218086e-01 5.26243627e-01 6.69824600e-01 -4.20815229e-01
-1.69745579e-01 -2.24184752e-01 5.38825989e-01 6.29540682e-01
-9.83586311e-02 -3.20555061e-01 -1.26669824e+00 8.26257944e-01
-1.85664272e+00 -1.32171202e+00 -2.45695665e-01 2.14308357e+00
4.63772357e-01 -2.58913897e-02 -2.52864063e-01 7.00966954e-01
6.74620330e-01 4.02395606e-01 -2.76551306e-01 1.71214700e-01
-4.16966587e-01 1.69785812e-01 3.70260477e-01 3.80469233e-01
-1.35454917e+00 7.56363273e-01 5.62023067e+00 5.06984532e-01
-1.28873968e+00 1.60572752e-01 5.02181590e-01 1.18066773e-01
-1.14540927e-01 -1.42439559e-01 -5.29892027e-01 2.92269289e-01
4.17965561e-01 2.55912811e-01 1.43460885e-01 6.57479048e-01
4.65364367e-01 -5.32438934e-01 -7.17437148e-01 9.45061624e-01
-9.98430103e-02 -1.13782299e+00 9.67147853e-03 -2.63318829e-02
4.95245278e-01 3.76521856e-01 5.52056320e-02 2.77653515e-01
1.96272016e-01 -7.85687685e-01 9.94002879e-01 6.25599027e-01
5.77993453e-01 -7.30967224e-01 8.18798184e-01 4.76798803e-01
-1.90064502e+00 -4.48245585e-01 -3.26716542e-01 -1.01127200e-01
2.52936870e-01 6.60771251e-01 -9.28196669e-01 7.20902085e-01
1.06231105e+00 6.93026721e-01 -8.47707450e-01 1.06768227e+00
-4.74627465e-01 5.27672410e-01 -6.40477061e-01 3.28215033e-01
4.45454985e-01 3.63872275e-02 1.41328007e-01 1.14839232e+00
7.57141590e-01 2.16336727e-01 4.07424122e-01 5.20886242e-01
1.43214628e-01 -1.21121556e-01 -7.11305857e-01 5.47159553e-01
4.62719023e-01 8.60537589e-01 -6.85537934e-01 -3.60052437e-01
-2.41635829e-01 9.38599646e-01 3.71210389e-02 2.53287762e-01
-7.74908781e-01 -4.72762316e-01 2.62881070e-01 3.96795034e-01
6.70668781e-01 -1.43911138e-01 -1.27286494e-01 -8.92155170e-01
4.58792523e-02 -4.19445157e-01 6.82328045e-01 -9.44594324e-01
-8.18182945e-01 6.93340003e-01 8.78614560e-02 -1.16620708e+00
-1.20072095e-02 -7.11917952e-02 -7.00285375e-01 7.30654895e-01
-1.76214457e+00 -1.45361161e+00 -6.69443309e-01 9.32249188e-01
5.95295370e-01 3.14261109e-01 6.16631687e-01 6.47042632e-01
-5.19492090e-01 8.95258337e-02 -6.13381341e-02 3.55607033e-01
2.23206297e-01 -1.06558597e+00 4.03618187e-01 1.19355392e+00
8.95289704e-02 6.68906420e-02 4.65655804e-01 -7.56686807e-01
-7.46705830e-01 -1.48217022e+00 1.25318861e+00 1.52704030e-01
9.93846878e-02 1.95065960e-02 -8.29111934e-01 5.11933506e-01
1.80871878e-02 -2.06410661e-01 3.28752965e-01 -2.73418635e-01
-5.02061666e-05 -4.84943986e-01 -1.30475104e+00 1.72852010e-01
8.65025997e-01 -6.49123609e-01 -6.76084518e-01 1.40317827e-01
4.67573076e-01 -5.42030632e-02 -6.39701664e-01 3.52546722e-01
3.83759618e-01 -9.70760047e-01 1.09232497e+00 2.67438978e-01
3.33672225e-01 -8.11812520e-01 -4.82064337e-01 -1.24174893e+00
-7.97025323e-01 1.90930501e-01 1.64405107e-01 1.19508529e+00
2.05742419e-01 -3.87893230e-01 3.10587525e-01 4.07814085e-02
-2.70313263e-01 -5.60081303e-01 -1.09775662e+00 -5.29567719e-01
-3.67277801e-01 -3.84896159e-01 9.24659014e-01 5.66277385e-01
-5.31462729e-01 3.96026164e-01 -1.68448359e-01 5.81225753e-01
2.85156131e-01 -6.66936040e-02 3.52282226e-01 -1.29093695e+00
2.88761675e-01 -3.76375347e-01 -4.51032043e-01 -9.99953806e-01
-5.76500356e-01 -7.91268289e-01 3.31779271e-01 -2.13975501e+00
7.38051161e-02 -3.65826845e-01 -4.93235528e-01 1.01738548e+00
-1.07360214e-01 4.44233179e-01 3.45990032e-01 3.97555947e-01
-1.31176710e-01 9.81190622e-01 9.88686264e-01 -5.69656074e-01
-2.79957771e-01 2.07887404e-02 -3.67229909e-01 5.38485229e-01
8.42972398e-01 -5.65730572e-01 -2.39631802e-01 -8.22168231e-01
4.81399037e-02 -1.49798885e-01 7.59529948e-01 -1.67346835e+00
4.71881956e-01 1.93663016e-01 5.57189584e-01 -1.36027312e+00
3.58402014e-01 -8.53293717e-01 2.86966056e-01 9.18202102e-01
9.51572657e-02 4.01087224e-01 -7.18843490e-02 3.73983830e-01
-5.51817596e-01 -1.26773184e-02 7.66666591e-01 -2.64820725e-01
-9.24126863e-01 3.73353094e-01 -6.10268295e-01 -6.08988464e-01
9.96924400e-01 -3.52865249e-01 -8.68745148e-02 -2.78053492e-01
-6.29668236e-01 4.06195134e-01 1.21841744e-01 3.16526592e-01
1.13249350e+00 -1.51472628e+00 -9.62254047e-01 3.61138433e-01
1.00533992e-01 3.60392123e-01 3.93744975e-01 7.35827744e-01
-5.47258794e-01 2.00052962e-01 -2.22204775e-01 -7.23953962e-01
-1.22343409e+00 4.98037845e-01 2.66991913e-01 -4.95747507e-01
-5.66107929e-01 8.36918056e-01 2.52165735e-01 -3.81265104e-01
7.45492056e-02 -7.37392128e-01 -5.12442112e-01 4.88280684e-01
5.59318781e-01 2.82742709e-01 2.57234186e-01 -9.62064922e-01
-5.55739760e-01 6.68201208e-01 2.05297902e-01 -2.80878723e-01
1.71865213e+00 -2.71943390e-01 -1.60190195e-01 3.20387721e-01
1.04902351e+00 -3.12654227e-01 -9.49633300e-01 -5.08334339e-01
-1.64765537e-01 -3.68170738e-01 4.44386750e-01 -6.93155468e-01
-1.32058084e+00 8.92586946e-01 1.12951159e+00 2.52873480e-01
1.74584138e+00 7.58115649e-02 9.09084618e-01 5.57080090e-01
2.31761366e-01 -1.04889214e+00 8.43211710e-02 5.60795903e-01
1.27578521e+00 -1.05656350e+00 -1.04874074e-01 -7.50571564e-02
-4.17071611e-01 1.03714144e+00 4.34656471e-01 9.61518958e-02
8.25673461e-01 1.90140568e-02 -1.29133388e-01 -4.42732334e-01
-4.28908706e-01 -6.79558277e-01 9.81144235e-02 4.50555801e-01
4.85925786e-02 9.80471373e-02 -5.33785149e-02 2.23550677e-01
-2.50050634e-01 -9.65398550e-02 9.02289823e-02 1.28555965e+00
-9.59329486e-01 -7.67871022e-01 -7.47027874e-01 4.46167678e-01
5.02594784e-02 -2.58791387e-01 -1.39053419e-01 5.43122351e-01
6.20278597e-01 1.25310540e+00 1.24566033e-01 -7.25661516e-01
6.10633850e-01 -2.27822199e-01 4.38301545e-03 -5.67177713e-01
-8.91770959e-01 3.45570408e-02 -2.91264743e-01 -3.46489429e-01
-4.48857874e-01 -6.90186679e-01 -1.26814449e+00 -2.62923986e-01
-4.69311506e-01 1.26643926e-01 4.47611243e-01 7.91384995e-01
2.43772939e-01 6.67162418e-01 9.13216531e-01 -8.39281380e-01
-1.60862103e-01 -1.20946908e+00 -5.61779916e-01 1.67813897e-01
5.75982034e-01 -6.22351110e-01 7.15078041e-02 5.70883974e-02] | [9.652117729187012, -1.2960597276687622] |
68f86d80-4eca-46c7-95c4-730241a8504c | analysis-of-augmentations-for-contrastive-ecg | 2206.07656 | null | https://arxiv.org/abs/2206.07656v1 | https://arxiv.org/pdf/2206.07656v1.pdf | Analysis of Augmentations for Contrastive ECG Representation Learning | This paper systematically investigates the effectiveness of various augmentations for contrastive self-supervised learning of electrocardiogram (ECG) signals and identifies the best parameters. The baseline of our proposed self-supervised framework consists of two main parts: the contrastive learning and the downstream task. In the first stage, we train an encoder using a number of augmentations to extract generalizable ECG signal representations. We then freeze the encoder and finetune a few linear layers with different amounts of labelled data for downstream arrhythmia detection. We then experiment with various augmentations techniques and explore a range of parameters. Our experiments are done on PTB-XL, a large and publicly available 12-lead ECG dataset. The results show that applying augmentations in a specific range of complexities works better for self-supervised contrastive learning. For instance, when adding Gaussian noise, a sigma in the range of 0.1 to 0.2 achieves better results, while poor training occurs when the added noise is too small or too large (outside of the specified range). A similar trend is observed with other augmentations, demonstrating the importance of selecting the optimum level of difficulty for the added augmentations, as augmentations that are too simple will not result in effective training, while augmentations that are too difficult will also prevent the model from effective learning of generalized representations. Our work can influence future research on self-supervised contrastive learning on bio-signals and aid in selecting optimum parameters for different augmentations. | ['Javad Hashemi', 'Ali Etemad', 'Sahar Soltanieh'] | 2022-05-30 | null | null | null | null | ['arrhythmia-detection'] | ['medical'] | [ 5.82664609e-01 1.29761100e-01 -1.64657742e-01 -4.64815974e-01
-7.38974392e-01 -4.51369107e-01 2.19295964e-01 3.36737126e-01
-5.36146760e-01 7.44232774e-01 1.37517005e-01 -5.03056169e-01
-2.89079726e-01 -5.76824903e-01 -4.98286784e-01 -6.67366505e-01
-4.69202816e-01 3.33509207e-01 1.29629791e-01 -2.99041629e-01
3.91705409e-02 5.15692949e-01 -1.22967112e+00 2.30263740e-01
7.71672606e-01 8.94439399e-01 -5.63269742e-02 7.93913841e-01
2.59778678e-01 4.77198482e-01 -9.00776446e-01 -7.18065873e-02
4.18243885e-01 -9.00221050e-01 -6.50016785e-01 8.73837098e-02
8.38286132e-02 -1.33815780e-01 5.76435179e-02 7.10346103e-01
1.03166842e+00 -1.68942407e-01 6.61523998e-01 -8.80889475e-01
2.12661969e-03 7.15467632e-01 -3.34717005e-01 7.17343330e-01
9.31539387e-02 3.20445150e-01 6.28886282e-01 -4.10287440e-01
2.36982018e-01 7.31513262e-01 1.04621494e+00 4.67301816e-01
-1.41367280e+00 -7.16385484e-01 -1.04368635e-01 -2.21780986e-01
-1.33073723e+00 -4.39490199e-01 9.33911324e-01 -2.91535735e-01
8.33892882e-01 2.69779414e-01 5.25783658e-01 1.09426641e+00
1.76304817e-01 1.74878940e-01 1.26635373e+00 -5.59374809e-01
2.91484267e-01 3.71123552e-01 5.81414513e-02 4.23567623e-01
3.28540415e-01 1.43257469e-01 -2.43934676e-01 -3.74037683e-01
8.65653157e-01 -2.36655369e-01 -2.15961993e-01 -1.73550293e-01
-1.12591136e+00 8.20894599e-01 2.01402336e-01 6.52019560e-01
-5.01568794e-01 5.18525355e-02 7.52404690e-01 6.24802411e-01
4.36623335e-01 1.09622860e+00 -6.80550933e-01 -1.61551237e-01
-1.08408928e+00 1.92005783e-02 6.08509362e-01 4.21574533e-01
4.65316921e-01 2.78945565e-01 -2.52324432e-01 9.67442095e-01
-8.70637521e-02 -6.88610971e-02 6.88412786e-01 -6.84895515e-01
4.53792512e-01 4.33537155e-01 -4.18618359e-02 -6.17899597e-01
-7.60033667e-01 -1.05691493e+00 -7.67184258e-01 1.46947473e-01
5.69536865e-01 -7.70671964e-01 -9.78638947e-01 1.67044032e+00
-3.24450098e-02 3.33659947e-01 1.70494884e-01 8.34275365e-01
7.32897282e-01 3.93532664e-01 2.10876033e-01 -4.28075850e-01
1.40242898e+00 -4.94979650e-01 -7.88090765e-01 -4.18755770e-01
9.53227282e-01 -5.35293400e-01 1.22717488e+00 3.92928064e-01
-1.19607258e+00 -7.15920746e-01 -1.33804429e+00 5.48697948e-01
-1.72012135e-01 2.21757084e-01 5.56490242e-01 7.82592237e-01
-9.49936628e-01 1.01481485e+00 -8.32594275e-01 -1.89982742e-01
5.54960012e-01 5.62598884e-01 -2.03722626e-01 2.88789123e-01
-1.46075726e+00 8.49320412e-01 4.73380536e-01 1.19318835e-01
-6.31749034e-01 -6.65327311e-01 -6.84465051e-01 7.23235309e-02
1.85550392e-01 -6.98935866e-01 8.02289903e-01 -1.25093424e+00
-1.16741753e+00 7.56556332e-01 2.72629112e-01 -8.81488204e-01
5.23264885e-01 -9.07870904e-02 -4.55116063e-01 4.51800711e-02
-2.01636463e-01 5.22587240e-01 1.03146398e+00 -1.23195171e+00
-1.59122422e-01 -2.64187545e-01 -2.99877338e-02 1.78490862e-01
-5.07366598e-01 2.32412759e-02 -1.77305371e-01 -8.84323955e-01
1.01417284e-02 -9.20011580e-01 -4.34840649e-01 -5.59855044e-01
-2.56324917e-01 2.62007833e-01 4.77604955e-01 -4.79349196e-01
1.37844265e+00 -2.18016648e+00 -1.61184654e-01 4.53501403e-01
1.95724770e-01 3.56217057e-01 -1.32973716e-01 3.51318270e-01
-5.69749773e-01 2.14801833e-01 -3.68344754e-01 -1.66805163e-01
-4.95308399e-01 1.78056836e-01 1.59450457e-01 2.72861063e-01
4.16662544e-01 7.68649995e-01 -8.41609120e-01 -4.27075565e-01
3.22009549e-02 4.75381702e-01 -4.92345512e-01 1.89874053e-01
1.91904217e-01 6.28617644e-01 -2.90280253e-01 1.85017973e-01
3.91096085e-01 -3.63840818e-01 1.95550874e-01 -2.50663400e-01
6.83119372e-02 4.40761507e-01 -1.24817777e+00 1.38981211e+00
-3.71943504e-01 5.02001405e-01 -3.51648241e-01 -1.11905468e+00
1.04735529e+00 4.73983705e-01 6.55890226e-01 -3.88931096e-01
2.13629201e-01 1.53510049e-01 6.13145053e-01 -4.72256958e-01
5.23661003e-02 -4.29791152e-01 2.26709887e-01 5.13280272e-01
8.59598368e-02 -4.47002016e-02 1.70570165e-01 -5.85573539e-02
1.35244894e+00 -6.29089773e-02 4.36598390e-01 -3.35151941e-01
4.43347365e-01 -2.11781323e-01 4.65840787e-01 9.24531877e-01
-1.92043051e-01 7.26702571e-01 6.86123371e-01 -4.26419139e-01
-9.09204006e-01 -7.15989113e-01 -3.27846378e-01 8.94030988e-01
-2.71741390e-01 -6.00204051e-01 -6.97679043e-01 -7.89088905e-01
-2.94328392e-01 4.45519060e-01 -7.14783549e-01 -4.37487483e-01
-6.53209090e-01 -1.26248431e+00 8.34140599e-01 8.26720059e-01
4.54297036e-01 -1.29390407e+00 -9.74074721e-01 2.23811075e-01
-3.06771677e-02 -9.48164344e-01 -1.51054814e-01 8.64918351e-01
-1.40193510e+00 -8.60551059e-01 -7.45872319e-01 -6.38765812e-01
7.76469052e-01 -3.49687427e-01 1.15143740e+00 2.92160332e-01
-1.29595533e-01 1.46171808e-01 -4.25286591e-01 -6.53960466e-01
-6.83224022e-01 2.47887373e-01 -1.46447539e-01 -1.38603732e-01
1.75992146e-01 -6.98047161e-01 -7.27986097e-01 8.23994875e-02
-7.35829651e-01 -2.80509025e-01 7.33745396e-01 1.20816362e+00
2.80545712e-01 1.77999854e-01 1.02838838e+00 -1.40834188e+00
1.03031659e+00 -4.39764678e-01 -1.71050936e-01 -1.01632096e-01
-7.70923913e-01 1.86273694e-01 7.73918211e-01 -6.62259638e-01
-6.75893128e-01 1.67954430e-01 -3.69403094e-01 -3.71841371e-01
-3.17014813e-01 4.78318334e-01 1.64325461e-01 -1.00652846e-02
1.13530135e+00 4.10927460e-03 1.23442188e-01 -1.83373779e-01
-3.00034583e-02 5.62172115e-01 2.89352685e-01 -3.67730677e-01
6.53414130e-01 1.13795526e-01 -1.29252806e-01 -6.53023899e-01
-4.99075919e-01 -2.64390171e-01 -6.96277261e-01 4.16511409e-02
5.22193909e-01 -6.72751069e-01 -1.88056305e-01 1.94801718e-01
-5.63050747e-01 -5.33938348e-01 -7.29769826e-01 5.25480747e-01
-5.26433766e-01 3.88763130e-01 -6.54493511e-01 -7.13620365e-01
-4.34089005e-01 -9.39910293e-01 8.28496337e-01 -4.69825454e-02
-6.54544652e-01 -1.00907707e+00 1.60012301e-02 1.15936972e-01
5.71955502e-01 3.40208620e-01 8.54697526e-01 -9.67061400e-01
8.81241858e-02 -1.30826622e-01 2.29592562e-01 7.60816574e-01
4.50151652e-01 -3.71981323e-01 -1.05849290e+00 -3.61950219e-01
1.65510640e-01 -3.22889477e-01 7.21651077e-01 4.45837647e-01
1.21275675e+00 -2.26351783e-01 -1.51750803e-01 6.34751320e-01
1.15854931e+00 4.74498808e-01 9.07476902e-01 4.84206885e-01
3.72015536e-01 3.16505224e-01 4.46761161e-01 3.64925891e-01
1.22254686e-02 3.88139635e-01 1.06098868e-01 -7.02087700e-01
5.14205694e-02 1.65032059e-01 -6.00056387e-02 4.59984958e-01
-3.32162052e-01 2.16724813e-01 -9.42007720e-01 4.59405243e-01
-1.42771828e+00 -6.98036373e-01 1.32922918e-01 2.44364858e+00
1.17702365e+00 5.67183197e-01 5.13674736e-01 9.37592566e-01
4.32481200e-01 -2.75944062e-02 -4.51991826e-01 -5.58177412e-01
3.46723646e-02 6.52916789e-01 4.09578264e-01 1.98113263e-01
-1.23804295e+00 7.34917223e-01 6.78305197e+00 3.11164767e-01
-1.33255327e+00 7.98396673e-03 8.48813534e-01 1.30474344e-01
6.61196485e-02 -7.22299367e-02 -4.00909841e-01 4.92433608e-01
1.28483105e+00 2.05693021e-01 1.01079725e-01 4.78764355e-01
3.71878088e-01 4.84488718e-02 -1.05316913e+00 9.16791022e-01
-8.04888979e-02 -1.15111339e+00 -3.30130279e-01 -1.60337761e-01
5.17843962e-01 -9.77386162e-02 -1.08658530e-01 2.04252899e-01
-3.01023602e-01 -1.09231758e+00 2.15050951e-01 3.70557666e-01
6.39159143e-01 -7.56534755e-01 1.11573291e+00 2.00520471e-01
-6.77778542e-01 -2.70068288e-01 -7.41292397e-03 -1.32895499e-01
-2.25088987e-02 5.56237578e-01 -1.03039670e+00 2.84969240e-01
5.36958039e-01 4.57924545e-01 -7.96162009e-01 1.02604532e+00
-2.06086244e-02 1.12171936e+00 -2.53883719e-01 2.66666412e-01
-4.19909135e-02 8.06780308e-02 4.00814652e-01 1.53687263e+00
-5.71664535e-02 6.71262592e-02 4.96135512e-03 5.78290761e-01
1.84578255e-01 2.70648330e-01 -6.26072347e-01 5.62654324e-02
4.80296880e-01 1.10676491e+00 -8.62378299e-01 -5.30809641e-01
-2.77598202e-01 8.29323828e-01 3.58042009e-02 4.18923825e-01
-7.36848891e-01 -4.75240231e-01 7.82772750e-02 4.53355849e-01
1.18138805e-01 1.43527344e-01 -7.34693825e-01 -7.08550453e-01
-1.14247710e-01 -1.23740721e+00 5.95538616e-01 -4.52589244e-01
-1.14840579e+00 9.13230300e-01 1.36086464e-01 -1.27614367e+00
-5.26159227e-01 -2.24528521e-01 -7.33750105e-01 8.05819869e-01
-1.27371562e+00 -5.90615153e-01 -3.33251685e-01 4.43969637e-01
2.83967316e-01 -9.22857299e-02 1.01993608e+00 3.61712575e-01
-3.27352434e-01 7.78619647e-01 -6.02876842e-01 1.61949590e-01
7.03896999e-01 -1.26071596e+00 2.70839065e-01 6.43112004e-01
1.68481112e-01 8.42855334e-01 8.76567960e-01 -4.16217417e-01
-7.69649088e-01 -9.72144365e-01 6.58304572e-01 -2.65324265e-01
2.73586482e-01 -2.67463207e-01 -1.10516334e+00 4.13683534e-01
1.03506774e-01 2.06681192e-01 7.90056050e-01 2.26273671e-01
1.06342703e-01 -2.71905661e-01 -1.09441590e+00 5.27395725e-01
8.40021789e-01 -3.43776941e-01 -6.70575380e-01 3.20937902e-01
2.67206609e-01 -3.77193540e-01 -1.21175051e+00 7.51135945e-01
3.91568989e-01 -8.70921373e-01 8.97126079e-01 -6.92974687e-01
1.16420358e-01 -9.45970267e-02 3.91500413e-01 -1.52092969e+00
-2.43513703e-01 -6.78746581e-01 -1.92860328e-03 1.04145741e+00
6.98114574e-01 -8.31313074e-01 8.09449673e-01 1.49186552e-01
-9.63149965e-02 -1.15202236e+00 -4.79236871e-01 -6.85136139e-01
8.82218629e-02 -4.58569765e-01 2.86109656e-01 1.03284335e+00
3.89535949e-02 6.43688023e-01 -3.55351567e-01 3.67673598e-02
3.79950583e-01 -2.99294561e-01 5.76579750e-01 -1.21682858e+00
-5.62698185e-01 -2.24198028e-01 -5.22960663e-01 -5.34030735e-01
-1.69023529e-01 -6.85673356e-01 -5.65139167e-02 -1.32098198e+00
-1.51251573e-02 -8.43334079e-01 -6.24087453e-01 6.73046291e-01
-4.77932841e-01 4.20272470e-01 9.10026282e-02 3.96430008e-02
-6.96160197e-02 -2.15431657e-02 9.30889249e-01 4.02402878e-02
-7.47045159e-01 3.05181146e-01 -8.88838053e-01 6.95720196e-01
1.08385098e+00 -4.85751241e-01 -7.75668383e-01 1.17983662e-01
1.82460584e-02 2.77292319e-02 1.90631479e-01 -1.11763287e+00
-1.69148728e-01 3.64532381e-01 6.37016833e-01 -2.64182687e-01
2.17780143e-01 -6.90348268e-01 -1.34721603e-02 5.92280924e-01
-5.93334138e-01 2.52450287e-01 4.82859492e-01 1.65743440e-01
-7.67928809e-02 -3.14050376e-01 8.54646623e-01 -1.52413547e-01
-1.54209465e-01 -1.20976292e-01 -4.05786902e-01 2.21309245e-01
7.44127333e-01 -4.71411318e-01 3.08661580e-01 -4.62598652e-01
-1.03768957e+00 -6.28665835e-03 5.11298366e-02 2.48674959e-01
4.61792052e-01 -9.60836232e-01 -7.50039399e-01 5.43112278e-01
8.03899840e-02 -2.94663548e-01 4.18046601e-02 9.64953661e-01
-3.41452271e-01 1.03561722e-01 -2.36075118e-01 -7.65088379e-01
-1.25314498e+00 3.27814043e-01 5.09912372e-01 -4.02989388e-01
-8.34259570e-01 5.90125918e-01 -6.43567741e-02 1.97882508e-03
4.10146713e-01 -5.19627273e-01 -5.22650361e-01 -5.68152890e-02
3.61091703e-01 1.88770443e-01 4.02623713e-01 -3.92103702e-01
-3.69030714e-01 3.70308965e-01 -1.16457112e-01 6.26269653e-02
1.38470149e+00 1.99563548e-01 4.29009140e-01 4.99952972e-01
9.25429046e-01 -1.97068736e-01 -1.08191168e+00 1.80950969e-01
1.51521429e-01 -1.35477841e-01 1.95880488e-01 -8.53155077e-01
-1.11471832e+00 7.75909841e-01 9.47642982e-01 2.06721410e-01
1.53629243e+00 -1.65722191e-01 4.12144095e-01 1.64878786e-01
1.70084596e-01 -1.02587688e+00 1.99888840e-01 1.94022328e-01
6.74199939e-01 -9.87016857e-01 1.23527683e-01 -1.59368843e-01
-9.21343505e-01 1.01304793e+00 4.56953615e-01 -3.09027851e-01
6.00068450e-01 6.52209282e-01 3.44559371e-01 -3.88397425e-01
-6.66719437e-01 -1.85438424e-01 3.01651657e-01 5.26612580e-01
7.52772570e-01 -1.89302817e-01 -6.08001769e-01 5.78315198e-01
-1.80179343e-01 1.45089567e-01 4.81046319e-01 1.00394857e+00
3.45347053e-03 -1.16202319e+00 -3.35746795e-01 8.51879120e-01
-7.97422588e-01 -1.99145272e-01 -2.79648811e-01 8.42256725e-01
2.92514086e-01 8.80468607e-01 2.80399770e-02 -2.81903774e-01
4.52353656e-01 3.36860746e-01 5.25295436e-01 -8.57969463e-01
-1.05348241e+00 3.35686654e-01 3.48857522e-01 -2.26798236e-01
-4.89338547e-01 -6.47931159e-01 -1.34818232e+00 4.42807585e-01
-4.91118670e-01 1.39389664e-01 4.65026766e-01 8.62057030e-01
2.94822216e-01 9.01124001e-01 5.79029620e-01 -4.91206825e-01
-6.72526598e-01 -1.24765682e+00 -4.53490287e-01 4.12113696e-01
4.41099942e-01 -6.02762699e-01 -5.07895470e-01 1.38353333e-01] | [14.25288200378418, 3.32623028755188] |
538f9bcd-25a2-4a9b-926b-a25dc615f4f6 | pose-guided-fashion-image-synthesis-using | 1906.07251 | null | https://arxiv.org/abs/1906.07251v2 | https://arxiv.org/pdf/1906.07251v2.pdf | Pose Guided Fashion Image Synthesis Using Deep Generative Model | Generating a photorealistic image with intended human pose is a promising yet challenging research topic for many applications such as smart photo editing, movie making, virtual try-on, and fashion display. In this paper, we present a novel deep generative model to transfer an image of a person from a given pose to a new pose while keeping fashion item consistent. In order to formulate the framework, we employ one generator and two discriminators for image synthesis. The generator includes an image encoder, a pose encoder and a decoder. The two encoders provide good representation of visual and geometrical context which will be utilized by the decoder in order to generate a photorealistic image. Unlike existing pose-guided image generation models, we exploit two discriminators to guide the synthesis process where one discriminator differentiates between generated image and real images (training samples), and another discriminator verifies the consistency of appearance between a target pose and a generated image. We perform end-to-end training of the network to learn the parameters through back-propagation given ground-truth images. The proposed generative model is capable of synthesizing a photorealistic image of a person given a target pose. We have demonstrated our results by conducting rigorous experiments on two data sets, both quantitatively and qualitatively. | ['Shanglin Yang', 'Wei Sun', 'Jawadul H. Bappy', 'Hui Zhou', 'Yi Xu', 'Tianfu Wu'] | 2019-06-17 | null | null | null | null | ['pose-guided-image-generation'] | ['computer-vision'] | [ 6.90656185e-01 3.06574792e-01 4.08280432e-01 -4.26622719e-01
-2.37304583e-01 -5.67513406e-01 8.60168338e-01 -3.70139718e-01
-2.86504682e-02 5.13180196e-01 6.42398149e-02 1.26696482e-01
4.03812855e-01 -9.69891548e-01 -1.05802238e+00 -5.45358717e-01
5.74197352e-01 2.94562638e-01 1.51068075e-02 -1.43280119e-01
9.00134444e-02 4.61602509e-01 -1.51575840e+00 2.09374607e-01
5.68136394e-01 1.01915681e+00 3.36075485e-01 6.70402050e-01
2.89310485e-01 4.21897471e-01 -7.13193119e-01 -7.48625040e-01
4.90233958e-01 -7.49867618e-01 -3.83014292e-01 6.38302684e-01
5.69212198e-01 -3.90469700e-01 -3.83925259e-01 8.81293952e-01
5.30081093e-01 7.55960345e-02 8.84076953e-01 -1.13926017e+00
-1.13254213e+00 6.65397346e-02 -5.32829762e-01 -3.71895105e-01
6.32276356e-01 3.97193372e-01 5.09735048e-01 -5.97029626e-01
8.92489910e-01 1.28117216e+00 3.46116781e-01 7.33682215e-01
-1.09129798e+00 -4.73853886e-01 -2.60922424e-02 -3.62035066e-01
-1.27256131e+00 -4.01762366e-01 1.16316426e+00 -5.45702219e-01
1.58901557e-01 2.75805861e-01 1.00525475e+00 1.13562918e+00
4.76745009e-01 6.74855649e-01 1.09973931e+00 -5.09227872e-01
2.66298115e-01 2.97506511e-01 -6.82892919e-01 7.21219122e-01
-9.35591906e-02 3.02953959e-01 -3.34772646e-01 9.88425463e-02
1.52302516e+00 -8.41547828e-03 -1.86959833e-01 -5.13949335e-01
-9.70101893e-01 6.05422020e-01 6.19657636e-01 -2.63709668e-02
-5.54244936e-01 1.57592490e-01 -1.33965135e-01 1.42975241e-01
2.86459059e-01 3.46669436e-01 2.07353756e-01 3.59188557e-01
-7.89537430e-01 3.62509280e-01 4.83257771e-01 1.04515421e+00
5.04878700e-01 6.75837845e-02 -4.61539865e-01 8.70613873e-01
3.46390158e-01 6.09995782e-01 3.02866012e-01 -8.08089554e-01
2.79616833e-01 5.93773186e-01 1.81599051e-01 -1.21562946e+00
7.30134398e-02 -5.22124231e-01 -9.35623765e-01 3.28218549e-01
1.72291473e-01 -2.02975988e-01 -1.02176404e+00 1.68403757e+00
5.74222147e-01 -9.29564685e-02 -1.11219473e-01 1.18237054e+00
8.16960931e-01 8.14618528e-01 -4.99791419e-03 7.91299157e-03
1.34340787e+00 -8.17843676e-01 -6.00821018e-01 -4.82910216e-01
-5.36539182e-02 -9.75729048e-01 1.03711295e+00 2.24253520e-01
-1.42019725e+00 -8.77352834e-01 -1.08633947e+00 -2.09197029e-01
-9.32550803e-03 5.92558026e-01 3.25786263e-01 4.05119985e-01
-8.47829640e-01 3.57409060e-01 -3.20924461e-01 -4.28123146e-01
1.38562605e-01 8.13593268e-02 -4.40513432e-01 1.60048887e-01
-1.02103484e+00 6.02806687e-01 1.54391766e-01 2.85657793e-01
-1.03456819e+00 -3.74219149e-01 -1.00521350e+00 -2.44260252e-01
-1.69928521e-02 -1.13002372e+00 1.05230594e+00 -1.21916223e+00
-1.76033390e+00 1.37347770e+00 1.89688861e-01 -1.76979661e-01
7.55747139e-01 1.12494551e-01 -3.28580528e-01 3.83227654e-02
1.67111009e-01 1.04052305e+00 1.13858175e+00 -1.71849978e+00
-3.61671090e-01 -3.08598131e-01 1.95953920e-01 4.58092213e-01
6.02144971e-02 -2.21321359e-01 -7.14804471e-01 -8.76343369e-01
7.52577931e-02 -9.69738722e-01 6.05069809e-02 2.22166240e-01
-8.23506176e-01 2.78918505e-01 7.41633594e-01 -8.20895970e-01
6.68882489e-01 -2.13908124e+00 3.83776069e-01 2.13118032e-01
-5.38327359e-02 2.27054894e-01 -9.76612940e-02 4.69916701e-01
-1.08241588e-01 -3.23612034e-01 -9.12233070e-02 -5.57949424e-01
-1.67210206e-01 -3.51501442e-02 -3.03711116e-01 4.63909507e-01
3.22439522e-01 1.06768560e+00 -7.71679521e-01 -2.83875942e-01
4.79749382e-01 7.48771667e-01 -5.19697428e-01 7.54021704e-01
-2.26104945e-01 8.79715919e-01 -2.93947786e-01 2.99461365e-01
6.96977258e-01 -1.08061396e-01 1.32749304e-01 -6.03456616e-01
8.92675743e-02 -1.28400907e-01 -1.21311784e+00 1.55482709e+00
-5.88436961e-01 3.84480655e-01 -5.37772998e-02 -4.81877267e-01
1.30138099e+00 1.11420259e-01 3.00933588e-02 -7.14754701e-01
4.94268388e-01 -1.64673910e-01 -2.25192562e-01 -5.47604620e-01
5.61900079e-01 -3.74269009e-01 -8.77764598e-02 4.28628474e-01
-1.11773677e-01 -5.20955443e-01 3.05731539e-02 3.75240436e-03
5.00075042e-01 3.37520629e-01 7.12639168e-02 7.04447031e-02
7.68042386e-01 -3.03319812e-01 2.56144077e-01 2.59442568e-01
3.39165717e-01 1.16815341e+00 2.29915023e-01 -5.29006124e-01
-1.46403384e+00 -1.18549716e+00 2.79780716e-01 5.71672618e-01
4.73964214e-01 -6.52084947e-02 -8.53888750e-01 -4.08420891e-01
-1.99357435e-01 7.25084186e-01 -7.64830053e-01 -3.55292141e-01
-5.35358191e-01 -7.88195580e-02 1.99630171e-01 4.39833760e-01
8.47488165e-01 -1.23942602e+00 -5.82504511e-01 -8.52075741e-02
-2.53126156e-02 -1.06363225e+00 -9.14846063e-01 -4.23716843e-01
-5.12728810e-01 -8.76249969e-01 -1.03273857e+00 -1.14437830e+00
1.26612890e+00 2.27723882e-01 8.98211956e-01 -5.81641980e-02
-4.02962029e-01 2.62707382e-01 -2.35286150e-02 -2.85003722e-01
-7.88563848e-01 -2.32139170e-01 -1.53744251e-01 4.17198837e-01
-2.42283136e-01 -5.16843438e-01 -1.11758208e+00 4.84703630e-01
-1.10989285e+00 7.65855789e-01 6.83257341e-01 6.10831201e-01
6.94137871e-01 -6.46500289e-02 1.36189297e-01 -7.13701904e-01
8.22228014e-01 -4.83129658e-02 -5.27759612e-01 2.00456724e-01
6.77358210e-02 2.24800948e-02 7.65196145e-01 -4.57913548e-01
-1.16019833e+00 3.24205369e-01 -1.73729777e-01 -5.11619270e-01
-1.65304109e-01 2.75185448e-03 -3.29765975e-01 3.90874147e-02
6.16106987e-01 4.32420611e-01 2.25443557e-01 -2.97963440e-01
5.75377941e-01 7.37653136e-01 8.66163909e-01 -5.07674575e-01
1.02451456e+00 5.61442614e-01 -1.53493717e-01 -6.15575016e-01
-5.08627295e-01 7.89769515e-02 -7.60753453e-01 -4.86021727e-01
9.84080374e-01 -8.20811093e-01 -7.97557533e-01 6.02155805e-01
-1.22624946e+00 -2.22033888e-01 -4.08888400e-01 9.98915285e-02
-6.47483170e-01 -4.06329371e-02 -3.21175456e-01 -6.22341990e-01
-4.16637748e-01 -1.17853260e+00 1.62710381e+00 4.93501067e-01
-1.41529158e-01 -9.41092551e-01 1.20789461e-01 4.96693820e-01
1.64350793e-01 5.58457971e-01 6.71095073e-01 1.41291991e-01
-5.31219840e-01 -3.74498814e-01 -1.94701970e-01 4.70080584e-01
3.69495362e-01 2.38605365e-02 -7.84803092e-01 -3.44211936e-01
-1.35829240e-01 -2.50858873e-01 2.69292146e-01 2.52890259e-01
9.98554111e-01 -3.03051978e-01 -2.34168082e-01 6.56306386e-01
1.24141014e+00 4.67341244e-01 9.14214134e-01 5.97562902e-02
8.56278300e-01 6.06144190e-01 3.67591143e-01 1.61290616e-01
2.97829479e-01 8.70201528e-01 2.78176159e-01 -3.38393569e-01
-5.42390287e-01 -9.79024470e-01 1.95182726e-01 4.01844233e-01
-2.02722937e-01 -4.00235265e-01 -4.18035179e-01 2.74677366e-01
-1.49558806e+00 -8.78582120e-01 1.70988172e-01 2.41481924e+00
7.17914224e-01 -3.12478635e-02 1.47083342e-01 -6.27444014e-02
9.17623341e-01 1.58623476e-02 -5.18560350e-01 -4.29798782e-01
9.88955274e-02 -9.37976688e-03 2.57455349e-01 3.77897054e-01
-8.50305438e-01 8.57274652e-01 6.11774588e+00 5.50886631e-01
-1.41027284e+00 -2.43903190e-01 8.09226632e-01 2.31958047e-01
-4.94188696e-01 -1.37335792e-01 -4.96015668e-01 4.92398292e-01
2.88479984e-01 -3.74967279e-03 2.93644845e-01 7.12513268e-01
3.37026894e-01 -3.26985717e-02 -1.16857576e+00 1.09461248e+00
3.46460819e-01 -1.32024860e+00 4.54849601e-01 1.05620071e-01
9.20152664e-01 -9.48062241e-01 3.31564665e-01 -1.71290815e-01
1.48105770e-01 -1.05529499e+00 1.14338672e+00 6.71093345e-01
1.09446251e+00 -5.93791187e-01 4.04862821e-01 3.21731955e-01
-9.66093779e-01 1.55096442e-01 -2.67740399e-01 -3.13844122e-02
4.63265717e-01 3.21793497e-01 -9.06090319e-01 6.21436536e-01
1.10281259e-01 3.96598756e-01 -4.90180075e-01 8.85533810e-01
-5.23602247e-01 -7.94746056e-02 -4.67101811e-03 1.48746997e-01
-2.12148726e-02 -4.98850852e-01 4.95872736e-01 7.37284005e-01
3.74292761e-01 -8.40949491e-02 5.81098907e-02 1.19324362e+00
-1.58805609e-01 2.01605242e-02 -7.87773132e-01 2.23260392e-02
2.84653306e-01 1.30042577e+00 -7.07079887e-01 -2.08660305e-01
2.15330601e-01 1.55098617e+00 8.04307535e-02 2.96491742e-01
-9.45772767e-01 -1.85812622e-01 2.31154338e-01 5.20591557e-01
1.14423461e-01 -2.99781263e-01 -2.25408509e-01 -8.98029745e-01
9.23508629e-02 -7.30044484e-01 -1.38933524e-01 -1.31560409e+00
-1.08971548e+00 8.56701195e-01 8.42412487e-02 -1.39310145e+00
-4.25097317e-01 -4.22034264e-01 -7.96882629e-01 1.07501745e+00
-9.33351994e-01 -1.55196440e+00 -6.34279847e-01 3.67120504e-01
5.38557112e-01 -1.84680820e-01 4.11779940e-01 2.90505260e-01
-3.80003333e-01 5.72690964e-01 -3.66309255e-01 2.04856902e-01
5.37409842e-01 -9.86866653e-01 5.31743705e-01 7.68128335e-01
1.92244519e-02 6.11982822e-01 8.61749947e-01 -7.39586651e-01
-1.30358458e+00 -1.18254185e+00 5.77619731e-01 -1.94722310e-01
-7.46111199e-02 -5.41424155e-01 -3.47982705e-01 6.24750674e-01
3.17811310e-01 -1.85814202e-01 2.00794816e-01 -4.90551412e-01
-5.72601473e-03 -1.62296981e-01 -1.29411638e+00 7.65611529e-01
1.14673865e+00 -3.36853325e-01 -3.97300214e-01 3.14104527e-01
3.62212360e-01 -7.04314947e-01 -5.65548658e-01 2.78312683e-01
7.47674048e-01 -1.05053449e+00 9.80648756e-01 -1.89978778e-01
8.50344300e-01 -5.48685133e-01 1.36271328e-01 -1.61077547e+00
-3.95418674e-01 -6.18762136e-01 2.76194364e-01 1.32618296e+00
4.38155271e-02 -3.19399625e-01 7.91336477e-01 5.32205999e-01
-5.34175001e-02 -7.51349270e-01 -4.16652650e-01 -6.22356474e-01
-1.51250541e-01 -1.14040561e-02 7.67765284e-01 4.95505095e-01
-4.54551101e-01 5.07237732e-01 -7.40501106e-01 8.90572891e-02
5.85158229e-01 2.58507311e-01 1.25480604e+00 -7.20218897e-01
-4.04696614e-01 -9.75684077e-02 -6.01297081e-01 -1.34982669e+00
-1.46244630e-01 -6.70403838e-01 5.19088022e-02 -1.72935379e+00
1.93861842e-01 -3.22768688e-01 6.14401937e-01 9.80488509e-02
-1.69766471e-02 6.25754058e-01 3.23828876e-01 1.30546376e-01
-9.87150967e-02 7.34112859e-01 1.82995701e+00 -1.61057234e-01
-3.18310887e-01 8.93007293e-02 -8.17828655e-01 5.84777832e-01
5.11032760e-01 3.11901164e-03 -7.34833598e-01 -4.68614727e-01
1.54250592e-01 8.66232738e-02 7.48171031e-01 -8.85187328e-01
3.22182886e-02 -1.06140085e-01 8.14017594e-01 -2.78805405e-01
5.74815989e-01 -6.69381559e-01 7.58089781e-01 3.97472769e-01
-4.57224786e-01 7.61042684e-02 -1.12858050e-01 3.53574812e-01
2.74420269e-02 -5.45033664e-02 8.85272861e-01 -9.82926488e-02
-4.64171648e-01 2.96207339e-01 1.50642335e-01 -2.88445055e-01
1.21491444e+00 -4.48577464e-01 -1.33476913e-01 -7.19756901e-01
-7.21745610e-01 -3.99059209e-04 7.36855686e-01 5.79665840e-01
9.25074995e-01 -1.73341572e+00 -6.35876596e-01 6.11636639e-01
8.45999792e-02 2.75446791e-02 4.14600700e-01 3.33597988e-01
-8.42075408e-01 1.38338149e-01 -4.74131793e-01 -5.50864041e-01
-1.23276269e+00 5.31443775e-01 4.73819017e-01 5.58915250e-02
-5.83262503e-01 7.54523575e-01 7.28732228e-01 -4.82418567e-01
7.04044402e-02 -1.14964835e-01 -4.59790081e-02 -4.84110057e-01
3.85886192e-01 5.96658094e-03 -1.51511133e-01 -9.58690464e-01
4.88866419e-02 7.90154755e-01 2.41957605e-02 -2.76542515e-01
1.09684193e+00 -1.22189604e-01 2.63026923e-01 2.78588444e-01
1.12830484e+00 -4.06972989e-02 -1.61772251e+00 -3.74785066e-02
-9.35348153e-01 -8.14965546e-01 -2.50764757e-01 -7.50053942e-01
-1.21803677e+00 7.59599805e-01 5.43273389e-01 4.51225713e-02
1.05974030e+00 -6.39740154e-02 8.87416184e-01 -3.66802663e-01
5.09433031e-01 -9.50241625e-01 4.35993016e-01 -9.82904956e-02
1.43367374e+00 -9.02404428e-01 -1.39234290e-01 -6.65486157e-01
-9.24770534e-01 9.16060209e-01 7.27839649e-01 -4.45385247e-01
2.70590663e-01 -4.78937477e-02 8.73382613e-02 -4.26829368e-01
-4.16112572e-01 3.44239995e-02 7.12980509e-01 7.43741632e-01
3.41618627e-01 2.56766025e-02 -2.52083480e-01 1.82368904e-01
-6.19136631e-01 1.72002271e-01 2.90824145e-01 6.72253430e-01
-1.05879851e-01 -1.07140338e+00 -4.58904624e-01 3.11714578e-02
1.29722385e-02 3.82375240e-01 -5.58524549e-01 6.26438320e-01
3.27003062e-01 7.04755664e-01 2.26265952e-01 -4.53065187e-01
5.43864429e-01 -3.10465246e-01 7.82590508e-01 -6.41073048e-01
-3.92883271e-01 8.46705120e-03 -6.50683567e-02 -1.88321516e-01
-2.14569837e-01 -3.80371451e-01 -8.38982761e-01 -2.43521810e-01
9.35746282e-02 -7.46019706e-02 7.05754101e-01 7.56706119e-01
3.50913972e-01 5.41129410e-01 9.02891576e-01 -1.09570682e+00
-2.28536412e-01 -7.29165733e-01 -5.19726753e-01 8.27124596e-01
-4.18836288e-02 -6.48744643e-01 -2.46650036e-02 4.99804348e-01] | [11.88856315612793, -0.7727894186973572] |
ec1914ed-c48f-41f6-9e0e-c98a26d15bb3 | white-box-multi-objective-adversarial-attack | 2305.03655 | null | https://arxiv.org/abs/2305.03655v2 | https://arxiv.org/pdf/2305.03655v2.pdf | White-Box Multi-Objective Adversarial Attack on Dialogue Generation | Pre-trained transformers are popular in state-of-the-art dialogue generation (DG) systems. Such language models are, however, vulnerable to various adversarial samples as studied in traditional tasks such as text classification, which inspires our curiosity about their robustness in DG systems. One main challenge of attacking DG models is that perturbations on the current sentence can hardly degrade the response accuracy because the unchanged chat histories are also considered for decision-making. Instead of merely pursuing pitfalls of performance metrics such as BLEU, ROUGE, we observe that crafting adversarial samples to force longer generation outputs benefits attack effectiveness -- the generated responses are typically irrelevant, lengthy, and repetitive. To this end, we propose a white-box multi-objective attack method called DGSlow. Specifically, DGSlow balances two objectives -- generation accuracy and length, via a gradient-based multi-objective optimizer and applies an adaptive searching mechanism to iteratively craft adversarial samples with only a few modifications. Comprehensive experiments on four benchmark datasets demonstrate that DGSlow could significantly degrade state-of-the-art DG models with a higher success rate than traditional accuracy-based methods. Besides, our crafted sentences also exhibit strong transferability in attacking other models. | ['Cong Liu', 'Yingfan Gao', 'Zexin Li', 'Yufei Li'] | 2023-05-05 | null | null | null | null | ['dialogue-generation', 'dialogue-generation'] | ['natural-language-processing', 'speech'] | [ 1.98685989e-01 3.98110420e-01 2.38083135e-02 -1.75771296e-01
-9.98387992e-01 -9.53750074e-01 7.44960070e-01 -1.63202256e-01
-2.99381018e-01 1.03562665e+00 2.12197274e-01 -4.89928544e-01
4.24763888e-01 -9.57739234e-01 -5.43569446e-01 -5.79324126e-01
2.90828586e-01 4.45675999e-01 8.78400877e-02 -9.11460161e-01
1.84776008e-01 1.44340068e-01 -6.92701936e-01 3.97226840e-01
1.17208087e+00 6.01795733e-01 -3.17603528e-01 8.23451817e-01
1.22014344e-01 9.50161159e-01 -1.39373922e+00 -1.25178313e+00
1.46501496e-01 -6.16554976e-01 -9.51234102e-01 -3.30252022e-01
1.50790304e-01 -6.31258786e-01 -5.01652718e-01 1.04411876e+00
7.95047343e-01 1.07416436e-01 4.39621240e-01 -1.31507313e+00
-8.28979075e-01 1.24074864e+00 -2.02091098e-01 -4.00573947e-02
5.19292772e-01 6.66633308e-01 1.22307038e+00 -5.54957390e-01
3.30892354e-01 1.63569546e+00 5.82717240e-01 1.11676168e+00
-1.19233608e+00 -6.40813708e-01 2.94654846e-01 -1.19713090e-01
-8.10432315e-01 -3.96566063e-01 9.18567121e-01 -7.01039359e-02
7.69279480e-01 7.07683444e-01 2.48472735e-01 1.90536058e+00
5.07925034e-01 9.38728690e-01 9.50918436e-01 -1.22719660e-01
1.26329422e-01 2.74046659e-01 -2.34218836e-01 5.80409586e-01
7.23026693e-02 1.51026636e-01 -4.04928237e-01 -4.23024118e-01
2.00143531e-01 -2.99348742e-01 -3.55233341e-01 3.17120254e-01
-1.11680222e+00 1.22312808e+00 8.30638111e-02 1.22364849e-01
-7.88758546e-02 1.36421680e-01 7.13058889e-01 4.69404519e-01
6.52829051e-01 9.95202899e-01 -3.27801049e-01 -3.11800867e-01
-5.39021194e-01 5.76209009e-01 9.35746372e-01 6.39161527e-01
2.66762435e-01 4.58141327e-01 -7.85535574e-01 7.36201406e-01
-1.73763447e-02 5.34795702e-01 6.42029285e-01 -6.17572248e-01
7.55252004e-01 4.81470019e-01 -7.00060325e-03 -1.03591335e+00
-9.07052532e-02 -4.24409419e-01 -9.31395054e-01 5.24340756e-03
5.23779988e-01 -6.76250517e-01 -4.84613538e-01 1.99342227e+00
2.49646112e-01 -3.06712776e-01 2.37686515e-01 8.61919343e-01
6.15659833e-01 5.76682329e-01 -8.45354348e-02 -1.61038324e-01
1.12625003e+00 -1.17890298e+00 -7.57445514e-01 -3.82739693e-01
7.67250896e-01 -7.54410267e-01 1.56328964e+00 4.62533683e-01
-1.18652904e+00 -2.72993237e-01 -1.06816852e+00 1.94254845e-01
-1.56127453e-01 -2.26277828e-01 6.28962159e-01 9.86336470e-01
-7.38651514e-01 7.81113207e-01 -4.81766999e-01 2.37589613e-01
1.48081496e-01 1.48229688e-01 -1.04778290e-01 2.69922525e-01
-1.74229157e+00 1.02217388e+00 7.72884190e-02 -6.20063841e-02
-1.19337606e+00 -6.91286027e-01 -6.55602098e-01 -1.36137577e-02
5.18777966e-01 -6.65688038e-01 1.60986257e+00 -7.49747157e-01
-2.12108111e+00 4.77382123e-01 1.96292847e-01 -6.26081169e-01
1.01382637e+00 -4.16334242e-01 -3.96148860e-01 -8.81621763e-02
-6.37242794e-02 1.40445441e-01 1.10411775e+00 -1.13218248e+00
-1.42009377e-01 3.99436951e-02 5.19935489e-01 7.37975538e-02
-6.20119154e-01 -3.69659364e-02 2.26339743e-01 -9.25941467e-01
-3.82397443e-01 -9.06141102e-01 -3.78033996e-01 -4.62512016e-01
-9.72090185e-01 -2.56212234e-01 7.36629009e-01 -6.61591053e-01
1.42411339e+00 -1.62456489e+00 2.34895259e-01 -1.41113669e-01
2.70390183e-01 4.87108946e-01 -2.48204038e-01 7.19151318e-01
1.34238899e-01 5.28112471e-01 -1.60484165e-01 -4.57966298e-01
2.49266058e-01 3.22149657e-02 -8.90989065e-01 3.23067605e-02
2.49634638e-01 1.12376308e+00 -1.15696573e+00 -2.11376235e-01
-8.74049440e-02 4.28110845e-02 -7.29923069e-01 5.42375028e-01
-5.90533674e-01 4.19874430e-01 -4.46369529e-01 4.66031194e-01
4.22821939e-01 4.51327413e-02 6.66513965e-02 1.10157661e-01
3.16912025e-01 6.16549671e-01 -5.70384264e-01 1.41140532e+00
-6.31441236e-01 4.32062894e-01 -2.33794257e-01 -8.16700935e-01
8.22515905e-01 4.16935742e-01 -9.58274007e-02 -4.98397529e-01
2.09314436e-01 9.52304378e-02 1.28250480e-01 -2.62432009e-01
7.53800750e-01 -8.80059600e-02 -4.20952797e-01 8.30189288e-01
-3.82991806e-02 -4.92571265e-01 1.42372400e-01 5.16717553e-01
1.21256733e+00 -2.93343991e-01 -1.47436075e-02 2.61004288e-02
6.16591454e-01 -2.58374125e-01 5.07853389e-01 1.05831182e+00
-1.10347584e-01 4.76345122e-01 9.06708241e-01 -2.54588306e-01
-9.76251781e-01 -8.06624949e-01 3.50604713e-01 1.25252259e+00
-1.32739037e-01 -6.48456573e-01 -1.20664442e+00 -1.37444162e+00
-9.34178755e-02 1.09390724e+00 -6.09453976e-01 -7.81921983e-01
-6.80020154e-01 -8.91503453e-01 1.35663080e+00 3.10240984e-01
4.59307075e-01 -1.05367029e+00 -2.17899173e-01 3.63847464e-01
-4.56642061e-01 -8.46745670e-01 -8.57439876e-01 -2.24105775e-01
-5.40060759e-01 -8.06642711e-01 -5.57085931e-01 -2.14537963e-01
5.53363681e-01 -1.84640512e-02 1.27521169e+00 2.59621471e-01
1.03643976e-01 5.22211939e-02 -4.82318878e-01 -4.94104177e-01
-1.16842914e+00 3.92574489e-01 1.14150196e-01 -1.61377024e-02
-5.14227562e-02 -6.17270231e-01 -5.79661846e-01 4.35138106e-01
-8.98103893e-01 -1.50463656e-01 4.32475716e-01 1.30200160e+00
-8.46845359e-02 -3.05122823e-01 1.09095085e+00 -1.10578918e+00
1.38066292e+00 -3.37703049e-01 -3.37955952e-01 3.75569671e-01
-6.82587504e-01 1.77999496e-01 1.35773742e+00 -8.99049520e-01
-8.96632493e-01 -6.96289361e-01 -4.62064117e-01 -2.44335398e-01
2.11104333e-01 2.53102005e-01 -4.14093286e-01 4.58137728e-02
1.05477715e+00 4.90255743e-01 4.87495363e-02 -1.26610726e-01
5.32052994e-01 6.89050674e-01 3.62215847e-01 -7.92130888e-01
1.10353911e+00 -8.03335197e-03 -4.18324828e-01 -2.96333969e-01
-9.31957304e-01 3.48099858e-01 3.66226211e-02 -1.39439151e-01
4.02090579e-01 -6.04641020e-01 -1.00539219e+00 6.86646461e-01
-1.39617872e+00 -4.82066274e-01 -1.82831109e-01 5.99274859e-02
-4.22049671e-01 5.51619768e-01 -9.20486331e-01 -8.80340099e-01
-8.27745199e-01 -1.28833079e+00 7.74745762e-01 -1.20138435e-03
-4.28965002e-01 -1.13441372e+00 6.02016188e-02 6.35137379e-01
5.60784698e-01 2.37311840e-01 9.81502473e-01 -1.08967340e+00
-2.19374746e-01 -4.17799056e-01 2.93042690e-01 5.98197222e-01
1.25599414e-01 -7.33247539e-03 -1.11797225e+00 -4.49289739e-01
1.42705530e-01 -7.11383879e-01 5.45053601e-01 -3.54288816e-01
1.24189973e+00 -1.02214432e+00 9.20164064e-02 3.34977448e-01
7.66822934e-01 1.33898541e-01 5.29874623e-01 1.72779605e-01
6.92967832e-01 3.72883826e-01 6.51562393e-01 5.69104731e-01
2.46652275e-01 6.79200351e-01 6.46881044e-01 8.51720199e-02
1.98989719e-01 -5.92616856e-01 8.21329653e-01 8.58539402e-01
1.66166723e-01 -6.53188527e-01 -5.68419099e-01 3.32975864e-01
-1.71054196e+00 -1.07042587e+00 2.01830134e-01 2.09071732e+00
1.44712365e+00 4.53789920e-01 2.36909002e-01 1.88856691e-01
5.48476875e-01 4.57999766e-01 -5.92417419e-01 -7.79050529e-01
-1.23420097e-01 1.32670477e-01 2.27471441e-01 5.46727598e-01
-8.58158112e-01 1.16981816e+00 6.01449251e+00 1.11707687e+00
-1.17919409e+00 1.66616514e-01 8.45068336e-01 -2.73778617e-01
-6.66818082e-01 -1.14158235e-01 -6.90712333e-01 6.85501635e-01
9.10304427e-01 -5.35792887e-01 6.73529208e-01 9.29260015e-01
2.41849437e-01 4.71823752e-01 -1.04706204e+00 3.66736412e-01
5.09561449e-02 -1.24024701e+00 3.90254080e-01 -1.50769591e-01
7.00819969e-01 -4.83751684e-01 2.63101012e-01 6.78795040e-01
8.29674780e-01 -1.18724453e+00 7.91602731e-01 8.14261660e-02
5.90804100e-01 -8.68525803e-01 6.67690873e-01 5.28638065e-01
-4.86334890e-01 -4.10510339e-02 -4.11291152e-01 5.97517844e-03
2.68115222e-01 5.58615804e-01 -1.04115438e+00 6.67817891e-01
1.88500479e-01 8.97763297e-02 -3.84352475e-01 3.09430301e-01
-6.33656383e-01 8.90640914e-01 8.23007151e-02 -3.57439131e-01
4.06000108e-01 -1.27535418e-03 8.41320753e-01 1.13724208e+00
4.71158288e-02 -7.61173153e-03 1.14204111e-02 8.90381694e-01
-4.38058794e-01 -1.15712006e-02 -6.55584455e-01 -2.40722910e-01
5.60393870e-01 1.28045201e+00 -2.15908978e-03 -2.29336888e-01
8.90500192e-03 1.05488813e+00 4.32106733e-01 1.55179530e-01
-1.13375103e+00 -5.29844284e-01 7.56999016e-01 -1.13581419e-01
-2.24924520e-01 -1.68264396e-02 -4.55320239e-01 -1.19656622e+00
6.82880580e-02 -1.68050802e+00 1.20562479e-01 -2.92725563e-01
-1.36112344e+00 9.37482893e-01 -4.15071279e-01 -1.07704091e+00
-5.51001787e-01 -1.87199309e-01 -9.96918321e-01 9.25325513e-01
-1.09310615e+00 -1.06623232e+00 -3.45517020e-03 5.59866071e-01
7.87791908e-01 -3.40932488e-01 1.00220394e+00 2.52583716e-02
-8.37872803e-01 1.55768263e+00 -6.61886334e-02 3.49849552e-01
8.88098359e-01 -1.35398793e+00 8.84140491e-01 9.20569122e-01
-9.89713357e-04 6.83827221e-01 8.89671683e-01 -5.87609231e-01
-1.40443850e+00 -1.13205123e+00 6.91039085e-01 -7.12497175e-01
7.69692540e-01 -5.80939889e-01 -8.53954971e-01 4.34087545e-01
2.49440819e-01 -3.63219887e-01 4.93659556e-01 -6.54907376e-02
-4.94957328e-01 9.48656797e-02 -1.20956886e+00 1.09138322e+00
9.96104062e-01 -5.78022540e-01 -4.05774146e-01 7.02888489e-01
1.16835225e+00 -6.43395662e-01 -8.08876395e-01 2.93525219e-01
3.71724755e-01 -7.99492717e-01 8.70917082e-01 -1.10757196e+00
7.65775621e-01 2.08705232e-01 5.60284108e-02 -1.85871756e+00
1.43836709e-02 -1.33317506e+00 -3.40787977e-01 1.37835228e+00
6.82235360e-01 -8.11709285e-01 6.56896830e-01 6.64726853e-01
-1.82407409e-01 -8.90593827e-01 -7.95979798e-01 -8.59884202e-01
4.97504234e-01 -6.62657842e-02 8.05486500e-01 9.08975780e-01
2.11368263e-01 5.23358703e-01 -8.56358945e-01 3.11051589e-02
4.33822215e-01 -2.17013255e-01 1.17911041e+00 -7.21058428e-01
-5.94103038e-01 -4.99734193e-01 1.26348883e-01 -9.45992470e-01
3.77359986e-01 -7.04381406e-01 -4.71283682e-02 -9.81750607e-01
-2.10649490e-01 -3.60877991e-01 5.41844293e-02 5.25328875e-01
-7.62934268e-01 6.01142235e-02 3.42743248e-01 2.60309000e-02
-3.87186348e-01 8.53764474e-01 1.50572014e+00 -4.12471950e-01
-1.03166245e-01 2.75765806e-01 -1.06507683e+00 3.97989303e-01
9.70092177e-01 -6.46375120e-01 -5.71711719e-01 -3.33156675e-01
1.67476237e-01 1.77232265e-01 1.27137110e-01 -6.33270800e-01
-3.90928276e-02 -3.53175163e-01 -1.80699140e-01 -4.90364432e-02
3.79143089e-01 -1.25011608e-01 -2.18514219e-01 5.65998137e-01
-7.01727092e-01 1.77508980e-01 2.06355788e-02 4.06858951e-01
-1.36278912e-01 -3.70396495e-01 6.74481332e-01 -2.35277653e-01
1.33965284e-01 2.15530708e-01 -3.77022654e-01 5.26554167e-01
7.79697359e-01 2.41567224e-01 -7.48941481e-01 -7.06011117e-01
-2.16866344e-01 1.49172679e-01 2.33718619e-01 5.26897132e-01
4.24856722e-01 -1.10024035e+00 -1.07471013e+00 5.84362037e-02
-1.16286270e-01 -7.25858212e-02 2.17443347e-01 5.30802906e-01
-2.65650958e-01 2.93354511e-01 1.12498835e-01 -1.76792651e-01
-1.16439509e+00 4.66263592e-01 4.68581557e-01 -9.32093561e-01
-2.97441155e-01 9.99165475e-01 1.92573354e-01 -7.10722864e-01
1.52006269e-01 4.57407720e-02 1.25583664e-01 -6.37372360e-02
4.64890510e-01 2.64232516e-01 1.61407709e-01 -2.78049726e-02
-2.21522562e-02 -1.44327775e-01 -5.32935858e-01 -2.45507583e-01
9.04571474e-01 2.74754614e-01 9.14479792e-02 1.60225496e-01
8.38816583e-01 2.95658857e-01 -1.18153763e+00 -1.86708748e-01
-3.16320211e-01 -4.01754409e-01 -3.48658115e-01 -1.13942337e+00
-9.34967637e-01 9.32651937e-01 -1.62086070e-01 6.17144167e-01
8.01247060e-01 -5.89754403e-01 1.13650119e+00 6.23788834e-01
3.20295334e-01 -9.50661838e-01 5.80005586e-01 6.91223443e-01
1.18905008e+00 -1.08834481e+00 -2.30499208e-01 -1.99492812e-01
-1.04319966e+00 8.45885634e-01 9.14762318e-01 3.23252641e-02
-5.88352941e-02 1.85723677e-01 1.92118168e-01 3.28534991e-01
-1.17011046e+00 3.96616966e-01 3.44328396e-02 4.33624208e-01
3.00696790e-01 2.30464432e-02 -3.57320040e-01 9.27015245e-01
-6.95985556e-01 -6.12676442e-01 7.70330787e-01 5.82577050e-01
-9.26454365e-02 -1.39241123e+00 -2.90051043e-01 2.91277528e-01
-8.73043895e-01 -2.45627269e-01 -7.56267726e-01 6.00122511e-01
-4.91925925e-01 1.25999236e+00 -5.67770243e-01 -9.22920763e-01
3.45122069e-01 7.66896904e-02 8.99610147e-02 -5.16810596e-01
-1.31360948e+00 -4.87142682e-01 4.62789804e-01 -4.17913705e-01
3.15428883e-01 -2.47561142e-01 -7.80948043e-01 -8.13673496e-01
-6.44640505e-01 3.75542074e-01 3.73192728e-01 9.41351116e-01
2.56465703e-01 5.73300362e-01 1.24326432e+00 -5.54855466e-01
-1.56751728e+00 -1.22521651e+00 -5.70868850e-02 4.61289674e-01
1.55919939e-01 -3.23857933e-01 -6.14248395e-01 -3.49972397e-01] | [6.168035507202148, 8.166693687438965] |
a719ac44-8cf0-4aff-99dc-d885066e58a9 | graspness-discovery-in-clutters-for-fast-and | null | null | http://openaccess.thecvf.com//content/ICCV2021/html/Wang_Graspness_Discovery_in_Clutters_for_Fast_and_Accurate_Grasp_Detection_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Wang_Graspness_Discovery_in_Clutters_for_Fast_and_Accurate_Grasp_Detection_ICCV_2021_paper.pdf | Graspness Discovery in Clutters for Fast and Accurate Grasp Detection | Efficient and robust grasp pose detection is vital for robotic manipulation. For general 6 DoF grasping, conventional methods treat all points in a scene equally and usually adopt uniform sampling to select grasp candidates. However, we discover that ignoring where to grasp greatly harms the speed and accuracy of current grasp pose detection methods. In this paper, we propose "graspness", a quality based on geometry cues that distinguishes graspable area in cluttered scenes. A look-ahead searching method is proposed for measuring the graspness and statistical results justify the rationality of our method. To quickly detect graspness in practice, we develop a neural network named graspness model to approximate the searching process. Extensive experiments verify the stability, generality and effectiveness of our graspness model, allowing it to be used as a plug-and-play module for different methods. A large improvement in accuracy is witnessed for various previous methods after equipping our graspness model. Moreover, we develop GSNet, an end-to-end network that incorporates our graspness model for early filtering of low-quality predictions. Experiments on a large-scale benchmark, GraspNet-1Billion, show that our method outperforms previous arts by a large margin (30+ AP) and achieves a high inference speed. Our code and model will be made publicly available. | ['Cewu Lu', 'Jin Gao', 'Hongjie Fang', 'Minghao Gou', 'Hao-Shu Fang', 'Chenxi Wang'] | 2021-01-01 | null | null | null | iccv-2021-1 | ['robotic-grasping'] | ['robots'] | [-1.39410764e-01 -2.88626403e-01 -2.74313658e-01 -3.15450788e-01
-6.48312092e-01 -5.74378908e-01 -5.24863154e-02 -1.09351547e-02
2.13489309e-03 1.66254997e-01 -1.59179151e-01 -9.83435437e-02
-3.93521518e-01 -8.11182082e-01 -1.06732297e+00 -7.25689173e-01
-3.99976730e-01 4.24919814e-01 5.99793077e-01 -1.50784865e-01
4.05424237e-01 7.64227271e-01 -1.44286048e+00 2.75274843e-01
8.17613244e-01 1.36072516e+00 1.01441634e+00 3.52950573e-01
2.16671973e-01 1.18296877e-01 -2.77880132e-01 -1.33148804e-01
7.74865568e-01 4.25281197e-01 -6.71687901e-01 -2.52386719e-01
3.39857280e-01 -9.27657127e-01 -4.30761129e-01 9.96711612e-01
6.03796780e-01 -3.05247842e-03 4.45630282e-01 -1.06011295e+00
-4.37106162e-01 8.20408106e-01 -5.00438094e-01 -2.36186683e-01
4.35839117e-01 3.89784962e-01 1.00285614e+00 -1.12505317e+00
2.69933581e-01 1.48210442e+00 6.87735617e-01 3.29903990e-01
-7.30655551e-01 -6.20108306e-01 4.67258185e-01 -7.52375945e-02
-1.00877512e+00 -5.63295484e-02 8.14002752e-01 -8.35933983e-02
6.51151001e-01 3.32670152e-01 5.62562883e-01 1.13580549e+00
2.87729472e-01 1.49724734e+00 7.75525630e-01 -1.97867334e-01
2.32223064e-01 -6.78080618e-01 1.17632328e-02 8.90666485e-01
3.60214710e-01 1.62614390e-01 -3.49627256e-01 -1.93543643e-01
1.21679628e+00 3.79049003e-01 -1.90445587e-01 -8.16528201e-01
-1.42018914e+00 4.56673145e-01 8.92360628e-01 -1.31273136e-01
-6.45333469e-01 4.69103932e-01 2.99706995e-01 2.00394928e-01
2.13264257e-01 4.38578099e-01 -5.04396617e-01 -4.21394547e-03
-5.18527746e-01 6.05518937e-01 8.95224273e-01 1.48518229e+00
3.44974339e-01 -3.96948159e-01 -4.86106247e-01 7.66139090e-01
1.89423427e-01 7.44071484e-01 -2.26337209e-01 -1.28453648e+00
4.51110959e-01 6.06560290e-01 4.38665479e-01 -1.20738196e+00
-4.25774783e-01 -1.63408458e-01 -6.30494654e-01 1.39883310e-01
3.48924786e-01 2.22865820e-01 -1.07807291e+00 1.31775105e+00
3.81503314e-01 -2.91537076e-01 -4.33214873e-01 1.30370748e+00
5.82926333e-01 2.47601420e-01 -1.18214726e-01 2.87028790e-01
1.08374596e+00 -1.07788336e+00 -2.80811310e-01 -1.09427795e-01
4.48497720e-02 -7.45149672e-01 1.14297998e+00 6.53278947e-01
-1.10870385e+00 -4.55138445e-01 -8.19771051e-01 1.83454230e-01
-1.19285740e-01 5.22551417e-01 1.18404222e+00 2.37449273e-01
-6.29769087e-01 1.17397606e+00 -1.25231266e+00 -3.06525528e-01
5.14295459e-01 4.86271322e-01 -1.75215807e-02 -3.64581347e-01
-5.46050310e-01 6.24012589e-01 4.70354259e-01 5.32689393e-01
-9.98228133e-01 -3.88907403e-01 -4.88393605e-01 8.07850510e-02
8.31173182e-01 -4.67326343e-01 1.32850766e+00 -8.48069638e-02
-1.53702009e+00 3.07982653e-01 4.69710603e-02 -9.28800181e-02
7.53576934e-01 -6.38000906e-01 3.99078071e-01 2.87534148e-01
6.48906827e-02 5.81713796e-01 8.70877326e-01 -1.59103143e+00
-6.24454081e-01 -4.67336774e-01 4.91204590e-01 -5.82805872e-02
-3.41084749e-02 -2.16681689e-01 -8.10760856e-01 -6.54019296e-01
8.35594356e-01 -9.08235729e-01 -3.72418016e-01 4.29180503e-01
-5.56671321e-01 -4.89249825e-01 6.70028806e-01 -6.16382241e-01
6.78640604e-01 -1.90055895e+00 1.83331117e-01 4.40937608e-01
1.87840536e-01 1.66691110e-01 -3.02881479e-01 7.22387612e-01
5.02278388e-01 -3.14438701e-01 -5.96049577e-02 -8.59640017e-02
3.87985915e-01 2.40531877e-01 -4.01324898e-01 4.82346058e-01
1.91413850e-01 1.06326520e+00 -1.25639439e+00 -1.97465211e-01
3.89974117e-01 7.46456236e-02 -6.97421372e-01 4.23118740e-01
-3.80705684e-01 6.28951713e-02 -9.84929621e-01 1.45919681e+00
9.33382809e-01 -1.94907993e-01 2.32609361e-02 -6.59258306e-01
-2.65338575e-03 6.64030686e-02 -1.19177830e+00 2.00945950e+00
-3.51686001e-01 -1.81551844e-01 4.58185554e-01 -8.02212358e-01
1.18350041e+00 -1.21787302e-01 5.71606517e-01 -2.29807228e-01
2.64855593e-01 4.52031493e-01 -1.54605120e-01 -5.21804094e-01
5.18536210e-01 6.26271307e-01 3.27191167e-02 2.75348052e-02
-9.93218571e-02 -2.87156463e-01 1.31919473e-01 -4.74803187e-02
1.24199700e+00 5.88843226e-01 -1.94666341e-01 -4.33519065e-01
-1.84769362e-01 -8.48551616e-02 3.00954580e-01 1.19192874e+00
-2.46867388e-01 6.44968867e-01 1.87445030e-01 -4.59926426e-01
-7.72477627e-01 -1.23337555e+00 -4.39484529e-02 1.10489035e+00
7.25740552e-01 -1.67180404e-01 -4.54887867e-01 -5.34554362e-01
5.74234486e-01 2.22402692e-01 -4.26530361e-01 -1.04678407e-01
-6.85896337e-01 -5.84472895e-01 5.57801314e-02 1.02290094e+00
3.77456844e-01 -1.11980319e+00 -9.03354883e-01 3.42784345e-01
-2.39746980e-02 -1.02922487e+00 -2.04915762e-01 3.32867414e-01
-9.53012466e-01 -1.17695308e+00 -8.93122613e-01 -8.26105058e-01
6.28469646e-01 6.62380934e-01 7.98718393e-01 2.37826034e-01
-4.35417414e-01 4.27508801e-01 -8.03299665e-01 -3.90076309e-01
-5.73691502e-02 1.74039185e-01 1.80535778e-01 -7.31265008e-01
2.22356647e-01 -5.31930506e-01 -9.17088449e-01 4.57059354e-01
-5.08264720e-01 -1.19374759e-01 1.06716239e+00 8.02477956e-01
4.32728857e-01 -3.03513974e-01 5.54642200e-01 -9.74023268e-02
6.20048165e-01 -1.75176412e-01 -6.58719361e-01 3.37717444e-01
-3.59982580e-01 -5.03623970e-02 1.58930779e-01 -4.78288591e-01
-5.59281409e-01 2.85284340e-01 -6.68366700e-02 -9.43024516e-01
3.54825594e-02 4.47323143e-01 -1.56204104e-01 -3.68951976e-01
1.79026529e-01 -3.04326974e-02 1.40209273e-01 -7.52778351e-01
3.32836717e-01 5.85293829e-01 6.92045748e-01 -1.18458283e+00
7.02893198e-01 3.37918609e-01 -5.87764941e-02 -3.75771999e-01
-6.39463365e-01 -5.73821008e-01 -6.10751748e-01 -2.29842156e-01
3.20805609e-01 -7.80442774e-01 -1.46429467e+00 4.82661277e-01
-1.33079004e+00 -2.58356154e-01 1.12832315e-01 2.77824134e-01
-6.47564769e-01 4.80308115e-01 -7.87687182e-01 -1.09102404e+00
-5.97346425e-01 -1.32083058e+00 1.54356885e+00 -4.74190079e-02
2.34710410e-01 -2.10905105e-01 -6.13960087e-01 -1.51652819e-03
4.48679835e-01 2.69919485e-01 4.37983543e-01 -3.83983731e-01
-1.09390450e+00 -4.22750235e-01 -6.18834078e-01 2.29629144e-01
2.95736164e-01 -8.27784687e-02 -6.88329756e-01 -6.80978537e-01
-3.35693240e-01 -4.14298624e-01 1.03495693e+00 4.21478510e-01
1.84101546e+00 -1.70887247e-01 -5.85966051e-01 3.97234946e-01
1.30206072e+00 -1.86951719e-02 4.00230646e-01 1.35416716e-01
6.98788643e-01 4.73587275e-01 1.07201660e+00 6.10508323e-01
1.32876441e-01 5.21355152e-01 1.03656995e+00 6.48769364e-02
1.81189105e-01 -2.38780916e-01 1.74744606e-01 7.76715696e-01
-4.99351084e-01 -2.21169874e-01 -8.68041337e-01 5.21713316e-01
-2.03256083e+00 -5.87368965e-01 1.66242540e-01 2.20373964e+00
6.69634283e-01 3.19039643e-01 -1.04532931e-02 -1.85794875e-01
4.60925251e-01 -8.82682055e-02 -7.80103743e-01 1.66630372e-01
3.59801769e-01 1.27685219e-01 8.15468073e-01 1.29845992e-01
-1.19491255e+00 1.11632001e+00 5.82022095e+00 1.04176760e+00
-8.91070724e-01 -4.06260878e-01 1.45861655e-01 1.15262970e-01
7.65697584e-02 -2.62074023e-01 -6.33192956e-01 2.73801178e-01
-6.07685111e-02 2.80855983e-01 6.25328958e-01 1.32361293e+00
1.48159461e-02 -5.74065410e-02 -1.39055371e+00 7.62205422e-01
-2.12169930e-01 -1.05389214e+00 2.68774200e-02 -3.78148675e-01
2.01279327e-01 1.58932075e-01 -2.01063111e-01 2.82227904e-01
5.04787385e-01 -6.92374229e-01 8.61886561e-01 4.64288771e-01
5.37575901e-01 -5.41345179e-01 6.75151885e-01 2.67297029e-01
-1.29973328e+00 -3.28642756e-01 -7.19156742e-01 1.94649259e-03
5.23402393e-02 3.76726598e-01 -7.40552366e-01 9.02676463e-01
1.10309768e+00 4.61556226e-01 -6.96594268e-02 1.34193444e+00
-9.10040289e-02 1.93765745e-01 -6.03863835e-01 -4.15099233e-01
2.94193417e-01 6.90892860e-02 7.24881530e-01 9.67545927e-01
3.39250743e-01 2.57312149e-01 7.71498919e-01 1.11426127e+00
8.63296092e-02 -2.97015905e-01 -2.32501462e-01 -5.45603298e-02
7.56129384e-01 9.75374758e-01 -9.81203496e-01 -1.08611491e-03
3.93293379e-03 9.29554939e-01 2.32023850e-01 1.79090828e-01
-6.04311526e-01 -5.40080547e-01 4.24996406e-01 -1.27371937e-01
6.52011871e-01 -4.34117466e-01 -2.45882154e-01 -9.69118357e-01
5.84914923e-01 -7.50392199e-01 -2.95745164e-01 -6.34681880e-01
-1.63866925e+00 3.34176898e-01 2.06786305e-01 -1.18066585e+00
1.93517894e-01 -1.18054402e+00 -3.98511291e-01 5.23412764e-01
-1.22981989e+00 -1.25409889e+00 -6.26827776e-01 1.26208335e-01
7.62327731e-01 2.32478842e-01 6.27344072e-01 -2.96772429e-04
-1.93864971e-01 3.94306600e-01 -5.60324751e-02 2.12466091e-01
5.18487215e-01 -1.11799860e+00 4.97698635e-01 4.71451938e-01
-3.36337119e-01 8.63297224e-01 8.02409351e-01 -8.24399054e-01
-2.21124172e+00 -8.63228738e-01 6.86787665e-02 -4.73368108e-01
6.49670601e-01 -4.75471914e-01 -8.34986866e-01 4.52622682e-01
-3.21287364e-01 -4.50141281e-02 -2.70625204e-01 1.70923516e-01
-9.22476873e-02 -9.94548947e-02 -1.12391019e+00 5.02580225e-01
1.38923049e+00 2.88501065e-02 -6.07944012e-01 5.13967097e-01
9.35749114e-01 -9.61470544e-01 -9.30375278e-01 1.11059415e+00
1.08044207e+00 -7.64733970e-01 1.28044724e+00 -3.44853193e-01
5.48382998e-01 -2.73716915e-02 -4.49680567e-01 -1.01339352e+00
-3.37805301e-01 -5.40178478e-01 -3.93115938e-01 6.75851226e-01
-5.39674871e-02 -3.23075920e-01 9.36870217e-01 4.95424569e-01
-4.28857476e-01 -1.29893923e+00 -7.91139126e-01 -1.17049730e+00
6.48330385e-03 -4.07837749e-01 7.76831985e-01 4.17440474e-01
-8.05043802e-03 -4.53812331e-01 -1.37940377e-01 5.56572855e-01
7.60966897e-01 6.44885302e-01 7.42981613e-01 -1.28011250e+00
-2.01833189e-01 -4.64440167e-01 -1.51788726e-01 -1.82551658e+00
-1.91523835e-01 -5.88901281e-01 6.97809994e-01 -1.67839813e+00
3.06598186e-01 -9.49375153e-01 -3.34710151e-01 7.41915166e-01
-1.91822633e-01 -1.94502786e-01 3.90197694e-01 2.64842123e-01
-5.95717132e-01 7.67316759e-01 1.56890845e+00 -2.48907700e-01
-1.56887636e-01 2.15679452e-01 -2.13751838e-01 7.13245273e-01
7.45854437e-01 -1.89157084e-01 -5.43581694e-03 -6.15094841e-01
-1.21530160e-01 4.29935530e-02 7.13459611e-01 -1.05887794e+00
1.89964965e-01 -3.31282496e-01 2.89676189e-01 -9.37532008e-01
4.26797777e-01 -1.02944207e+00 -4.90526140e-01 5.40955305e-01
-1.34525552e-01 -1.23910546e-01 1.59173727e-01 7.84050167e-01
2.65222281e-01 -2.70625204e-01 1.73541933e-01 -2.02160493e-01
-6.32454693e-01 6.94896460e-01 1.58432290e-01 -5.06043077e-01
7.82296777e-01 -7.20889196e-02 -3.52234431e-02 -5.24472445e-03
-4.98664737e-01 4.83095765e-01 3.96047533e-01 6.93792164e-01
8.59235346e-01 -1.19066823e+00 -5.34412622e-01 5.25111109e-02
1.23968124e-01 3.82641166e-01 -1.06028505e-01 7.32196510e-01
-5.33591032e-01 6.67570457e-02 -8.03063735e-02 -9.07258809e-01
-8.49798620e-01 5.70778251e-01 -1.33682892e-01 1.97660625e-01
-7.71701992e-01 9.87624884e-01 -1.90105647e-01 -4.56694752e-01
8.68180931e-01 -9.29137945e-01 4.01678413e-01 -6.43762708e-01
2.18421459e-01 3.98254871e-01 -3.97283398e-02 1.71720594e-01
-3.12287658e-01 4.21687096e-01 -4.61154841e-02 6.11618280e-01
1.61543119e+00 2.02219501e-01 -2.64663517e-01 2.75029093e-01
6.86538100e-01 -2.22510427e-01 -1.68200254e+00 -3.49012298e-05
-1.42334729e-01 -8.17316890e-01 -1.27239823e-01 -8.87276649e-01
-1.10568643e+00 6.11380816e-01 5.94362319e-01 1.02940038e-01
8.51044655e-01 1.69324100e-01 8.74239922e-01 1.12211883e+00
1.16823530e+00 -9.09148037e-01 9.46658850e-02 4.85033840e-01
1.41539538e+00 -1.44260156e+00 6.60099387e-02 -8.63631248e-01
-9.96157899e-02 1.34020245e+00 8.51682246e-01 -5.32877624e-01
4.44965720e-01 4.01695907e-01 -3.50476176e-01 -3.75369519e-01
-3.26986790e-01 1.07873388e-01 2.21383154e-01 4.02312815e-01
-1.61147844e-02 2.11460441e-01 -1.33163348e-01 6.82351410e-01
-7.52803981e-02 1.30789191e-01 -9.25094932e-02 1.36593509e+00
-7.78425276e-01 -8.99386644e-01 -1.33089900e-01 5.46260297e-01
-1.18130311e-01 1.16030335e-01 -1.20188363e-01 7.88811088e-01
-2.17153028e-01 7.33831704e-01 -2.20640361e-01 -5.32643080e-01
4.65313047e-01 -5.18456876e-01 8.32400024e-01 -3.31292897e-01
-2.94303030e-01 -4.84741814e-02 -1.81533054e-01 -1.16324472e+00
-3.69219393e-01 -3.80421907e-01 -1.19411576e+00 -2.90429562e-01
-7.56240487e-01 -2.05121592e-01 6.43867910e-01 7.54762411e-01
4.52622443e-01 4.42208380e-01 4.89612073e-01 -1.55208397e+00
-1.32621777e+00 -9.73478854e-01 -4.80640590e-01 1.36765376e-01
1.52518302e-01 -1.11821699e+00 -8.18023831e-02 -3.44883293e-01] | [5.790431022644043, -0.8974326848983765] |
d5dd4a54-c3b5-43ed-97a5-6c0db6ec1437 | faq-feature-aggregated-queries-for | 2303.08319 | null | https://arxiv.org/abs/2303.08319v2 | https://arxiv.org/pdf/2303.08319v2.pdf | FAQ: Feature Aggregated Queries for Transformer-based Video Object Detectors | Video object detection needs to solve feature degradation situations that rarely happen in the image domain. One solution is to use the temporal information and fuse the features from the neighboring frames. With Transformerbased object detectors getting a better performance on the image domain tasks, recent works began to extend those methods to video object detection. However, those existing Transformer-based video object detectors still follow the same pipeline as those used for classical object detectors, like enhancing the object feature representations by aggregation. In this work, we take a different perspective on video object detection. In detail, we improve the qualities of queries for the Transformer-based models by aggregation. To achieve this goal, we first propose a vanilla query aggregation module that weighted averages the queries according to the features of the neighboring frames. Then, we extend the vanilla module to a more practical version, which generates and aggregates queries according to the features of the input frames. Extensive experimental results validate the effectiveness of our proposed methods: On the challenging ImageNet VID benchmark, when integrated with our proposed modules, the current state-of-the-art Transformer-based object detectors can be improved by more than 2.4% on mAP and 4.2% on AP50. | ['Linjie Yang', 'Yiming Cui'] | 2023-03-15 | null | null | null | null | ['video-object-detection'] | ['computer-vision'] | [ 3.76398899e-02 -3.96673054e-01 -2.39023846e-02 -2.84423321e-01
-6.56104028e-01 -3.52590024e-01 5.10044515e-01 3.60083058e-02
-5.63845873e-01 2.91434109e-01 -1.82479277e-01 7.60845169e-02
1.43060654e-01 -7.65078187e-01 -7.05964744e-01 -6.38512373e-01
2.36956887e-02 1.60736650e-01 1.22368860e+00 -2.54241019e-01
9.54498723e-02 2.66972452e-01 -1.72638631e+00 5.48409343e-01
6.01140559e-01 1.38495600e+00 3.53076458e-01 5.62212825e-01
-9.02949646e-02 9.22395051e-01 -4.78756636e-01 -3.30183506e-01
4.13831472e-01 -2.75502920e-01 -5.11290908e-01 1.20087467e-01
5.48457742e-01 -7.91577935e-01 -6.36949182e-01 1.19586110e+00
3.88865858e-01 1.11403629e-01 1.74764290e-01 -1.35768652e+00
-5.40050745e-01 4.88080293e-01 -4.88059133e-01 6.48756444e-01
2.77203172e-01 1.75967768e-01 8.82037818e-01 -1.22385931e+00
3.74087095e-01 1.47849941e+00 4.10890430e-01 3.48515362e-01
-7.49451816e-01 -7.06142366e-01 5.83736181e-01 9.33185577e-01
-1.54919350e+00 -3.32890481e-01 3.30998719e-01 -3.62257034e-01
9.41746056e-01 2.47094810e-01 7.57970870e-01 7.13762462e-01
-9.54437628e-02 1.23549640e+00 7.29071677e-01 -5.51481247e-02
-1.26020266e-02 2.67921001e-01 2.23321542e-01 6.35762334e-01
3.07807863e-01 -4.68964726e-02 -5.30518532e-01 1.28959090e-01
4.63519186e-01 4.07820642e-01 -2.52019227e-01 -4.29384887e-01
-1.15695989e+00 7.63833821e-01 7.18554735e-01 4.20727789e-01
-4.09274668e-01 1.52494147e-01 4.11822826e-01 1.80382594e-01
3.49512875e-01 4.94434536e-02 -1.86167911e-01 1.88802943e-01
-1.16740561e+00 4.04684961e-01 4.56816852e-01 1.05544305e+00
7.09539831e-01 -6.09296672e-02 -7.97062695e-01 5.08214176e-01
4.04941082e-01 3.79878193e-01 3.19819748e-01 -6.44506097e-01
4.08716112e-01 6.99991405e-01 1.13470294e-01 -9.08901095e-01
-1.03177987e-01 -5.22181928e-01 -4.75991100e-01 1.42993838e-01
2.57250965e-01 1.91706836e-01 -9.67389345e-01 1.38251603e+00
3.81368846e-01 4.80033308e-01 -1.29011557e-01 1.15643418e+00
1.00096560e+00 6.61555886e-01 6.65788800e-02 -1.71580613e-01
1.62780833e+00 -1.12318194e+00 -6.50709748e-01 -3.92998233e-02
4.93547380e-01 -8.06427956e-01 5.90780497e-01 2.72086829e-01
-1.05613685e+00 -8.00767839e-01 -1.17152905e+00 4.16156501e-02
-4.10175294e-01 2.60481387e-01 2.72084475e-01 6.60154998e-01
-1.10549974e+00 3.78423095e-01 -9.09814239e-01 -3.83032560e-01
5.67086577e-01 3.53101373e-01 -1.44105941e-01 -1.88938186e-01
-1.03538287e+00 7.84634590e-01 6.11151874e-01 1.45108595e-01
-1.16259646e+00 -6.69034243e-01 -5.83887875e-01 2.61582375e-01
7.97018468e-01 -5.81815481e-01 1.23910022e+00 -7.98770547e-01
-1.05516028e+00 4.51809049e-01 -2.21785113e-01 -7.42810845e-01
6.52817130e-01 -4.23335344e-01 -1.46773145e-01 2.63110399e-01
1.71992794e-01 7.72451818e-01 7.96588123e-01 -8.20042610e-01
-1.28048015e+00 -2.35346064e-01 3.76680076e-01 6.90521225e-02
-5.57945609e-01 2.75057554e-01 -1.05697465e+00 -5.52327275e-01
-2.11699754e-02 -7.92356908e-01 -1.77253589e-01 3.45193774e-01
-1.01088889e-01 -5.37011564e-01 1.35126376e+00 -5.18097162e-01
1.19016349e+00 -2.27113771e+00 9.28449929e-02 -1.49644762e-01
4.35474008e-01 5.62364280e-01 -1.41755775e-01 2.76943389e-02
2.77621716e-01 -1.43417165e-01 -7.92493224e-02 -4.94405627e-01
-5.67354225e-02 -4.57334146e-02 -1.78634077e-01 2.93764293e-01
4.11711961e-01 8.64128590e-01 -9.08286035e-01 -7.41834104e-01
4.18857664e-01 4.03079957e-01 -7.60192394e-01 6.23820648e-02
-1.42754555e-01 1.14067152e-01 -5.83854079e-01 7.74070382e-01
7.87969708e-01 -9.65474099e-02 -8.65955800e-02 -5.17985344e-01
-2.01583862e-01 6.05780408e-02 -1.35590398e+00 1.46102643e+00
1.31033316e-01 7.72271574e-01 3.45613956e-02 -1.08516455e+00
6.41595244e-01 3.49251926e-01 8.34235370e-01 -7.81424046e-01
5.39456308e-02 2.92251885e-01 4.04628813e-02 -5.04529595e-01
4.93004143e-01 2.90148765e-01 1.81416482e-01 4.37839478e-02
2.97518641e-01 1.83541164e-01 5.86801767e-01 3.15975815e-01
1.00686789e+00 6.01823069e-02 1.97180361e-01 1.96944065e-02
7.59820879e-01 -1.79028437e-02 6.23965263e-01 8.95486891e-01
-3.27055693e-01 5.07492125e-01 1.83792382e-01 -5.61295569e-01
-8.36717308e-01 -1.07393003e+00 -4.62401658e-02 1.17451274e+00
4.40611243e-01 -6.43369019e-01 -7.02961504e-01 -8.37858558e-01
-5.47280954e-03 2.37558365e-01 -5.19831777e-01 -2.63056248e-01
-6.56489372e-01 -7.95321703e-01 5.22496641e-01 6.77290142e-01
9.50711846e-01 -8.10573161e-01 -8.24358106e-01 3.02576274e-01
-3.07588607e-01 -1.50611067e+00 -3.88425767e-01 -1.58297643e-01
-6.50845885e-01 -1.13060069e+00 -8.20526898e-01 -7.18051612e-01
2.12550551e-01 7.20391214e-01 8.06884825e-01 1.25512376e-01
-3.74242246e-01 4.09887582e-01 -5.76804757e-01 -5.55061877e-01
-4.45896760e-03 -2.12050956e-02 -2.87599652e-03 4.33097154e-01
3.78907174e-01 -2.24039540e-01 -7.25687623e-01 4.87294734e-01
-1.05784130e+00 -1.87477604e-01 7.43804097e-01 5.77325642e-01
4.17694867e-01 -8.30107257e-02 2.13146016e-01 -2.93856651e-01
1.66040048e-01 -4.42310691e-01 -6.71200752e-01 3.24446201e-01
-3.10206741e-01 -2.54092850e-02 2.37073138e-01 -5.35843432e-01
-8.97203684e-01 3.22747469e-01 -1.84319075e-02 -7.71697760e-01
1.82914302e-01 1.05683580e-01 -6.02561459e-02 -1.22762352e-01
3.80744278e-01 4.47784930e-01 -2.49360904e-01 -5.19997716e-01
2.65036106e-01 4.49111134e-01 4.13776845e-01 -2.72550434e-01
7.84415603e-01 6.23741150e-01 -1.80511594e-01 -6.64432466e-01
-6.50005043e-01 -7.62083948e-01 -3.18018675e-01 -3.32970679e-01
9.99495685e-01 -1.05179453e+00 -8.28105569e-01 4.26504284e-01
-1.22221303e+00 1.55643150e-01 -1.99752748e-01 6.67637050e-01
-3.99769694e-01 5.38961828e-01 -5.00863016e-01 -6.90770030e-01
-2.55305409e-01 -1.59821820e+00 1.26994574e+00 2.41348267e-01
3.44864964e-01 -3.67527783e-01 -2.48065606e-01 9.34567377e-02
5.36716580e-01 -1.82469800e-01 3.97034764e-01 -6.04451776e-01
-1.14322007e+00 -6.36930391e-02 -6.07661247e-01 4.22081232e-01
-1.18518975e-02 -1.97457075e-01 -1.00235915e+00 -3.27805936e-01
1.05323419e-01 -2.16315612e-02 1.39491594e+00 2.65293121e-01
9.69251752e-01 -7.50213042e-02 -5.37141442e-01 5.30258954e-01
1.38676727e+00 2.14976668e-01 5.41118681e-01 3.24731201e-01
6.38560295e-01 3.70038271e-01 8.54308128e-01 5.27918458e-01
3.19888383e-01 1.18445516e+00 7.24822879e-01 1.43949101e-02
-2.29245797e-01 -1.08875381e-02 6.37570143e-01 3.25807363e-01
-1.45643175e-01 -2.44051009e-01 -6.63137197e-01 6.25869930e-01
-2.17609215e+00 -1.04760039e+00 -1.35931194e-01 1.89735866e+00
3.10471386e-01 2.51929343e-01 3.19509149e-01 1.29831210e-01
7.15458453e-01 8.00640658e-02 -4.00762975e-01 2.33501866e-01
-1.52953476e-01 1.44957695e-02 5.24040341e-01 3.35784256e-02
-1.40516257e+00 1.10815859e+00 5.84453058e+00 8.78564179e-01
-1.11931252e+00 4.24537957e-01 1.70314237e-01 -3.75884622e-01
2.80718505e-01 -7.07743969e-03 -9.62183118e-01 3.53043675e-01
4.54625040e-01 -1.87374517e-01 7.98393562e-02 9.49146152e-01
1.88043252e-01 -6.12563491e-02 -1.25426126e+00 1.17582536e+00
1.39030024e-01 -1.32905817e+00 2.07176521e-01 -1.57194853e-01
4.42496151e-01 7.47722238e-02 7.85967261e-02 5.02393603e-01
-5.12367748e-02 -5.42075336e-01 1.12299585e+00 4.30596948e-01
2.64418811e-01 -4.14902240e-01 9.70378458e-01 3.01602811e-01
-1.62573266e+00 -3.75490725e-01 -4.58920926e-01 -4.70435992e-02
2.54560530e-01 4.60805982e-01 -8.56493235e-01 6.50094926e-01
1.27422786e+00 8.76812458e-01 -8.45839560e-01 1.52607524e+00
1.44544527e-01 4.35453773e-01 -4.42622155e-01 -5.51914833e-02
3.31423610e-01 1.20168164e-01 6.79035008e-01 1.31697857e+00
4.41861331e-01 1.03187449e-01 4.57767427e-01 7.07395077e-01
2.52264179e-02 1.97829738e-01 -5.12515247e-01 2.39391252e-01
2.12131307e-01 1.29509091e+00 -7.13467717e-01 -7.78009713e-01
-6.66450977e-01 9.38273013e-01 4.05034572e-02 3.08345050e-01
-1.22521317e+00 -1.84750065e-01 7.70897508e-01 1.45865887e-01
7.55396843e-01 -1.63673803e-01 3.35001200e-01 -1.15351939e+00
4.38447684e-01 -8.23966622e-01 4.99838352e-01 -6.11604273e-01
-9.48639274e-01 6.26391292e-01 1.61108166e-01 -1.45111012e+00
-1.14496268e-01 -7.55927861e-01 -4.20631498e-01 3.86106282e-01
-1.45309305e+00 -1.01343346e+00 -5.37339747e-01 7.55951226e-01
8.33858013e-01 -4.80266288e-02 3.16666454e-01 7.74483919e-01
-5.54083526e-01 5.25534570e-01 -1.50590494e-01 3.93895000e-01
6.13043129e-01 -8.30842197e-01 3.67181808e-01 1.14673507e+00
4.02170539e-01 2.30079696e-01 5.22416115e-01 -3.37845206e-01
-1.35024667e+00 -1.22945249e+00 5.47183692e-01 -3.08944076e-01
4.30435807e-01 -3.14494610e-01 -8.90652239e-01 6.41092539e-01
1.38018370e-01 4.65831935e-01 3.49444337e-03 -3.94720942e-01
-4.05185878e-01 -4.11962748e-01 -9.71587062e-01 4.23165202e-01
1.26557052e+00 -3.47035080e-01 -5.34747601e-01 3.32431316e-01
7.76006222e-01 -2.18534917e-01 -5.44801533e-01 6.27363324e-01
5.02110720e-01 -1.07216251e+00 1.18958890e+00 -5.37630439e-01
-1.31572157e-01 -8.72514427e-01 -4.28244859e-01 -8.74468803e-01
-4.39173907e-01 -2.84222513e-01 -1.73406601e-01 1.06114924e+00
-6.81738928e-02 -4.60795343e-01 6.02593839e-01 3.55088785e-02
-2.09821135e-01 -4.25620109e-01 -1.12308753e+00 -9.81457531e-01
-5.55038095e-01 -6.08187675e-01 4.62128550e-01 4.24137801e-01
-1.98227003e-01 1.88673139e-01 -3.32960099e-01 2.09376842e-01
6.72489703e-01 -1.50936050e-02 5.95529258e-01 -1.11218226e+00
-2.76657939e-01 -5.90897739e-01 -9.30956662e-01 -1.36493850e+00
-2.98369318e-01 -6.59681380e-01 2.34191164e-01 -1.39550781e+00
4.23942178e-01 -1.42984048e-01 -5.31420052e-01 3.35622191e-01
-4.29222226e-01 5.78520000e-01 8.82489920e-01 1.48487091e-01
-1.17765617e+00 5.29565990e-01 9.20201838e-01 -5.23705423e-01
-3.35018113e-02 -6.53194338e-02 -3.63905609e-01 6.45141780e-01
4.58587408e-01 -5.15957296e-01 -2.74801850e-01 -5.88628232e-01
-2.12054491e-01 -2.45477304e-01 6.60747349e-01 -1.45640910e+00
6.04977906e-01 1.41519725e-01 3.75096679e-01 -8.30418229e-01
4.59629178e-01 -9.44960952e-01 -2.37747598e-02 6.95969045e-01
4.33502719e-02 4.84424345e-02 2.81481594e-01 5.68351388e-01
-5.15899777e-01 2.42965613e-02 7.58849561e-01 -8.24655965e-02
-1.15171039e+00 4.78783786e-01 -4.07959074e-01 -1.46117359e-01
1.34793115e+00 -2.51607120e-01 -2.43543118e-01 -1.47959217e-01
-6.52796626e-01 4.15769756e-01 6.38829470e-02 7.43075132e-01
9.08038378e-01 -1.33412898e+00 -8.91008794e-01 5.97376935e-02
2.93830961e-01 -1.90035582e-01 2.80998111e-01 1.16780245e+00
-1.99554771e-01 6.74519122e-01 -3.37967038e-01 -1.04368782e+00
-1.45522451e+00 8.97200406e-01 4.81303543e-01 -1.42604396e-01
-4.50065047e-01 7.76326716e-01 6.25193238e-01 2.87695199e-01
2.81637967e-01 -6.34350598e-01 -3.29575300e-01 2.95522034e-01
7.94437408e-01 4.50902104e-01 1.91348642e-01 -6.23068750e-01
-6.58246756e-01 5.55360258e-01 -2.24243075e-01 2.70512179e-02
1.18011868e+00 4.10637297e-02 1.84629634e-01 1.27225026e-01
1.08349752e+00 -2.89116710e-01 -1.20668936e+00 -5.44401705e-01
-4.13959324e-02 -4.96235102e-01 2.41974555e-02 -2.01717436e-01
-1.27905154e+00 8.19838464e-01 9.89126265e-01 2.81786710e-01
1.35139477e+00 2.39476591e-01 5.52791059e-01 4.19754952e-01
4.08493787e-01 -8.33760262e-01 2.62954414e-01 3.93353283e-01
7.20868170e-01 -1.29895103e+00 -7.63270706e-02 -6.43548191e-01
-2.82427698e-01 1.02351141e+00 7.18092024e-01 -1.47004068e-01
6.23929024e-01 1.18309803e-01 -2.09543735e-01 -6.57679439e-02
-7.02978492e-01 -7.02664256e-01 4.51648146e-01 3.63576263e-01
1.02279566e-01 -1.86231181e-01 -2.87770927e-01 4.20008749e-01
4.00954515e-01 1.81216851e-01 2.52798736e-01 8.10645461e-01
-7.20910609e-01 -1.02681708e+00 -5.44787049e-01 2.96919346e-01
-3.66167128e-01 -1.15263656e-01 -1.71735793e-01 6.99718893e-01
4.17481959e-01 1.03817308e+00 1.51102811e-01 -5.08758366e-01
6.10673010e-01 -1.20656908e-01 3.47090781e-01 -5.21992147e-01
-4.80766058e-01 4.10864875e-02 -2.37022027e-01 -9.21879530e-01
-6.47632957e-01 -6.65667057e-01 -1.11396337e+00 -7.31436014e-02
-3.41353238e-01 -3.95376347e-02 4.58213985e-01 9.13710654e-01
2.93160468e-01 7.07857847e-01 2.59735525e-01 -9.35416400e-01
-6.08420730e-01 -9.55615044e-01 -1.96264237e-01 3.67189318e-01
3.83709133e-01 -9.39519644e-01 -5.48720136e-02 -3.13491002e-02] | [8.77437973022461, -0.18057717382907867] |
852346cf-7ce1-4aaf-8e0b-ca050d160150 | kracl-contrastive-learning-with-graph-context | 2208.07622 | null | https://arxiv.org/abs/2208.07622v2 | https://arxiv.org/pdf/2208.07622v2.pdf | KRACL: Contrastive Learning with Graph Context Modeling for Sparse Knowledge Graph Completion | Knowledge Graph Embeddings (KGE) aim to map entities and relations to low dimensional spaces and have become the \textit{de-facto} standard for knowledge graph completion. Most existing KGE methods suffer from the sparsity challenge, where it is harder to predict entities that appear less frequently in knowledge graphs. In this work, we propose a novel framework KRACL to alleviate the widespread sparsity in KGs with graph context and contrastive learning. Firstly, we propose the Knowledge Relational Attention Network (KRAT) to leverage the graph context by simultaneously projecting neighboring triples to different latent spaces and jointly aggregating messages with the attention mechanism. KRAT is capable of capturing the subtle semantic information and importance of different context triples as well as leveraging multi-hop information in knowledge graphs. Secondly, we propose the knowledge contrastive loss by combining the contrastive loss with cross entropy loss, which introduces more negative samples and thus enriches the feedback to sparse entities. Our experiments demonstrate that KRACL achieves superior results across various standard knowledge graph benchmarks, especially on WN18RR and NELL-995 which have large numbers of low in-degree entities. Extensive experiments also bear out KRACL's effectiveness in handling sparse knowledge graphs and robustness against noisy triples. | ['Minnan Luo', 'Jundong Li', 'Qinghua Zheng', 'Qingyue Zhang', 'Shangbin Feng', 'Zilong Chen', 'Zhaoxuan Tan'] | 2022-08-16 | null | null | null | null | ['knowledge-graph-embeddings', 'knowledge-graph-embeddings'] | ['graphs', 'methodology'] | [-3.05922419e-01 2.86367625e-01 -6.06864631e-01 -2.24915326e-01
-2.70918816e-01 -3.53844523e-01 3.37222487e-01 3.34492058e-01
-2.72546470e-01 9.01741564e-01 6.11881673e-01 5.04425131e-02
-5.20027339e-01 -1.22027099e+00 -8.49788725e-01 -4.86075789e-01
-2.27264181e-01 3.71948898e-01 -1.19616218e-01 -1.75880849e-01
-4.49600160e-01 1.08927958e-01 -9.54217792e-01 5.64983338e-02
1.03437006e+00 7.79401720e-01 -1.25628471e-01 1.10092677e-01
-2.88537979e-01 1.16833174e+00 -1.78373367e-01 -9.73416626e-01
6.08608052e-02 1.35111168e-01 -8.93730938e-01 -1.93787143e-01
3.74373674e-01 7.99261183e-02 -1.02646983e+00 1.08466756e+00
4.61351812e-01 2.82018244e-01 4.76585597e-01 -1.33362937e+00
-1.47960842e+00 1.13565528e+00 -6.99245811e-01 1.99110538e-01
4.45467532e-01 -2.20825836e-01 1.55179131e+00 -1.15152836e+00
6.31864607e-01 1.36100399e+00 8.22855294e-01 -7.31819123e-03
-1.10587347e+00 -7.70222962e-01 4.39733058e-01 5.75110734e-01
-1.83899212e+00 -9.83840376e-02 8.54078114e-01 -3.26530486e-01
1.07056999e+00 1.57738924e-01 6.34884655e-01 9.67437506e-01
-4.46180344e-01 8.31506371e-01 5.37237287e-01 -1.31853566e-01
-2.24043295e-01 2.36959327e-02 2.67177969e-01 9.11490679e-01
7.05911338e-01 -2.68016309e-01 -7.39945531e-01 -2.22291812e-01
7.11866379e-01 1.67386904e-01 -6.18064284e-01 -5.76554120e-01
-1.26805520e+00 7.95866966e-01 9.89458978e-01 8.96373615e-02
-4.71987903e-01 2.90806353e-01 3.58343095e-01 1.99177533e-01
4.33123469e-01 1.94061384e-01 -5.36420763e-01 2.66554803e-01
-2.37761050e-01 -1.44391721e-02 8.45646918e-01 1.33800781e+00
1.01925302e+00 2.32320819e-02 -3.68477941e-01 7.64145434e-01
5.06408513e-01 4.49567765e-01 2.24438518e-01 -4.45106149e-01
9.44468558e-01 1.05618620e+00 -1.05838761e-01 -1.58141363e+00
-2.14865118e-01 -5.71656764e-01 -9.65646625e-01 -7.79931784e-01
-1.99721605e-01 -1.98146105e-01 -8.20282578e-01 1.80109966e+00
4.20011342e-01 6.66736484e-01 1.74688697e-01 8.08839619e-01
1.18681228e+00 5.44420600e-01 2.54914582e-01 3.71037833e-02
1.20281839e+00 -8.57689500e-01 -8.90726030e-01 -1.40701726e-01
6.96871579e-01 -2.62281805e-01 1.03261495e+00 -1.96593419e-01
-5.74913383e-01 -5.82703985e-02 -8.60190690e-01 -2.80755758e-01
-6.31059408e-01 -8.20331797e-02 1.20314229e+00 4.17269588e-01
-8.78590226e-01 3.29182595e-01 -7.10953295e-01 -1.50390506e-01
7.35333323e-01 3.09162289e-01 -6.03507698e-01 -4.03246045e-01
-1.82070327e+00 5.38918316e-01 8.18639398e-01 3.19458187e-01
-3.23406696e-01 -9.54695761e-01 -1.34521759e+00 4.89334941e-01
8.12145472e-01 -9.21914995e-01 3.25631082e-01 -3.16341072e-01
-1.06239307e+00 4.66038555e-01 4.18868214e-02 -3.13134968e-01
1.04422487e-01 -1.68120489e-01 -7.55945861e-01 -1.13900237e-01
2.14228496e-01 2.67977953e-01 3.65642101e-01 -1.17861402e+00
-2.20307067e-01 -3.76329392e-01 2.52296537e-01 3.34453702e-01
-7.16651320e-01 -6.58346593e-01 -8.82221580e-01 -8.06301415e-01
8.69915262e-03 -6.94841862e-01 1.17127076e-01 -5.10815740e-01
-7.18545020e-01 -3.46025527e-01 7.54553616e-01 -6.17715836e-01
1.55806005e+00 -1.94227195e+00 3.74742031e-01 5.80226958e-01
6.88337624e-01 3.36630523e-01 -1.27204746e-01 5.49356163e-01
-1.18587606e-01 1.51008517e-01 2.12377347e-02 -2.20364198e-01
3.00429940e-01 5.18675447e-01 -3.23828846e-01 2.24482492e-01
1.87225528e-02 1.61053073e+00 -1.29383254e+00 -4.20037389e-01
-9.22443867e-02 7.25890398e-01 -5.12702465e-01 -2.06073016e-01
-1.46372423e-01 -7.59092197e-02 -7.35417068e-01 8.04755986e-01
5.44483900e-01 -9.02546406e-01 5.23499370e-01 -5.96002460e-01
5.59946120e-01 1.14009827e-01 -1.24033046e+00 1.65450382e+00
-2.93542624e-01 2.69693255e-01 -7.07236975e-02 -9.11296129e-01
6.86956882e-01 1.88708693e-01 3.32999825e-01 -4.91148889e-01
-2.15623286e-02 8.43602605e-03 -4.76687610e-01 -4.24152017e-01
6.58245444e-01 6.02967292e-02 -6.77300319e-02 4.96403761e-02
3.45698565e-01 4.70837325e-01 1.58939064e-01 8.63351643e-01
1.18944776e+00 -1.74915478e-01 1.67773098e-01 1.26127684e-02
3.27886641e-01 -4.27961498e-01 8.74872923e-01 5.08148134e-01
1.34079278e-01 4.02461551e-02 7.05220819e-01 -1.35992661e-01
-5.48365712e-01 -1.14727783e+00 2.45956063e-01 8.01465511e-01
3.37737173e-01 -8.07826400e-01 -1.00937016e-01 -1.05796754e+00
6.26678348e-01 4.45050925e-01 -7.71818101e-01 -4.55943257e-01
-2.10957080e-01 -8.59048545e-01 4.07546997e-01 6.31086469e-01
3.84859234e-01 -8.35850596e-01 8.11752617e-01 5.01859523e-02
-2.91139573e-01 -1.29668212e+00 -7.20444798e-01 -6.26988113e-02
-4.95275527e-01 -1.21627533e+00 -3.93362343e-01 -8.82635474e-01
7.80616939e-01 3.47611457e-01 1.18093693e+00 2.69558895e-02
-1.79752409e-01 7.21922517e-01 -6.01759851e-01 -1.30886417e-02
4.84855860e-01 2.12267295e-01 1.37157515e-01 2.14075863e-01
6.21731222e-01 -8.61039102e-01 -5.19999206e-01 -3.93735953e-02
-8.54269326e-01 -1.66030735e-01 6.44138634e-01 9.97054875e-01
8.53605151e-01 3.32092404e-01 8.52379203e-01 -1.38660932e+00
7.40061581e-01 -1.03270102e+00 -3.93676519e-01 5.05980611e-01
-8.08127761e-01 1.32893130e-01 4.16493416e-01 -4.55796979e-02
-8.77477646e-01 -1.70165405e-01 2.73499489e-01 -8.08152616e-01
6.67953014e-01 1.14538383e+00 -4.25779611e-01 -3.00741374e-01
3.40546906e-01 9.37949270e-02 -5.63370645e-01 -3.96320611e-01
7.91689098e-01 1.97962180e-01 4.27166939e-01 -5.86165369e-01
1.05858445e+00 3.51282924e-01 -5.74311279e-02 -5.60271382e-01
-1.16125643e+00 -6.23480797e-01 -2.20584109e-01 3.10768992e-01
7.00353503e-01 -1.28464782e+00 -9.17380750e-01 1.66664198e-01
-7.87913024e-01 -7.43078161e-03 -4.13701832e-01 5.65802813e-01
4.29486521e-02 6.33465588e-01 -6.19743526e-01 -4.00500566e-01
-4.02962744e-01 -6.13129020e-01 7.78597295e-01 2.27272883e-01
3.22930753e-01 -1.29372489e+00 1.70547254e-02 5.43608129e-01
1.36893347e-01 3.22988003e-01 9.35300112e-01 -4.87534374e-01
-9.68180120e-01 -1.74602255e-01 -6.10903680e-01 2.15807319e-01
3.56480330e-01 -4.88256097e-01 -6.54715538e-01 -2.08668590e-01
-7.61501014e-01 -4.48193252e-01 1.33839691e+00 1.64166857e-02
1.18452609e+00 -6.67416990e-01 -5.47443509e-01 7.70138085e-01
1.47017598e+00 -5.29910266e-01 4.76842582e-01 -5.25790490e-02
1.60192525e+00 2.60747313e-01 4.05470639e-01 3.43333870e-01
1.21140003e+00 5.20064414e-01 3.23519319e-01 3.22465487e-02
-1.25108555e-01 -8.46281290e-01 1.13871157e-01 1.34406638e+00
-5.79945073e-02 -3.02824110e-01 -8.49742532e-01 8.31235230e-01
-2.07041121e+00 -9.57710028e-01 -1.27020866e-01 1.98527300e+00
1.06647336e+00 -9.97570232e-02 -2.53956705e-01 -3.31686229e-01
7.88088202e-01 2.90642619e-01 -4.16522264e-01 2.37099916e-01
-4.54355240e-01 4.00966555e-02 7.63193190e-01 6.49232686e-01
-1.07145774e+00 1.01957643e+00 4.90836859e+00 9.79745328e-01
-5.94965100e-01 1.39929637e-01 1.66932970e-01 -3.17899063e-02
-9.90875244e-01 1.03941758e-03 -7.52438307e-01 4.85590726e-01
4.53282416e-01 -4.88308907e-01 4.93530929e-01 6.60184681e-01
-4.23774868e-01 3.05785835e-01 -8.64532709e-01 1.15549469e+00
4.30093370e-02 -1.43394244e+00 2.29498446e-01 -1.93286873e-02
1.14248312e+00 2.08288252e-01 -7.08495677e-02 7.42721379e-01
6.89393282e-01 -1.12814832e+00 1.13029875e-01 8.15791368e-01
6.52483642e-01 -7.94596136e-01 8.36630404e-01 -8.83852914e-02
-1.71500421e+00 2.89020650e-02 -4.58868802e-01 2.77259380e-01
1.68935955e-01 9.10017312e-01 -9.31312263e-01 1.21390271e+00
5.87691545e-01 1.14081967e+00 -4.36555415e-01 7.56455600e-01
-5.11717021e-01 4.79362935e-01 -3.19082648e-01 7.65736327e-02
1.20435782e-01 -2.31299132e-01 4.35808331e-01 1.15397668e+00
1.21517450e-01 1.63784832e-01 3.56339544e-01 9.00577128e-01
-8.46322358e-01 2.12234795e-01 -5.20761549e-01 -5.52466452e-01
9.15946007e-01 1.23888433e+00 -2.46823013e-01 -2.37371624e-01
-6.77866518e-01 9.62482512e-01 9.37239230e-01 7.96419382e-01
-6.66402519e-01 -6.56670928e-01 6.57338917e-01 -1.12321481e-01
6.13846004e-01 -1.64088592e-01 1.40891850e-01 -1.62822664e+00
3.59263152e-01 -2.48104617e-01 7.95935154e-01 -6.41174912e-01
-1.86103868e+00 2.05475658e-01 -1.82196900e-01 -6.80245280e-01
4.86325562e-01 -2.61609793e-01 -2.72653431e-01 8.70013952e-01
-2.05253530e+00 -1.46514785e+00 -4.14276123e-01 8.99648607e-01
-4.06165123e-01 -1.34594858e-01 6.26212060e-01 7.80231059e-01
-7.38972306e-01 9.05462444e-01 1.05272986e-01 4.48294997e-01
5.51927388e-01 -1.57294750e+00 1.73525751e-01 4.96079206e-01
2.71606803e-01 8.44222486e-01 1.38548538e-01 -1.00769246e+00
-1.71105921e+00 -1.45145059e+00 9.67178106e-01 -5.23291588e-01
9.57821786e-01 -2.97888547e-01 -1.15136385e+00 1.14935219e+00
-2.53612995e-01 5.36227584e-01 8.40178251e-01 6.87376440e-01
-7.94332743e-01 -2.35097721e-01 -8.30988526e-01 4.43863541e-01
1.27070832e+00 -9.10613596e-01 -5.88996470e-01 5.11329770e-01
1.25069904e+00 -3.97412300e-01 -1.50496161e+00 5.68153620e-01
1.48954406e-01 -2.78661847e-01 1.15235269e+00 -7.09162831e-01
1.12226501e-01 -4.14409250e-01 -1.14382192e-01 -1.31812835e+00
-5.09648383e-01 -5.85972130e-01 -8.70170534e-01 1.51836407e+00
5.21001697e-01 -6.68322563e-01 1.00966740e+00 5.19634664e-01
5.55845909e-02 -9.36681151e-01 -7.41682529e-01 -7.09598124e-01
-3.25409979e-01 -9.91051644e-02 7.29840875e-01 1.69321048e+00
3.27826172e-01 7.82962620e-01 -5.67197740e-01 4.58776832e-01
6.98975384e-01 1.82734951e-01 6.18145287e-01 -1.18297207e+00
-1.78502187e-01 -9.46293101e-02 -8.20531726e-01 -9.20929968e-01
4.11438823e-01 -1.47101605e+00 -5.79154491e-01 -1.80211151e+00
3.67285430e-01 -6.21499062e-01 -5.58268011e-01 8.30118477e-01
-6.86533153e-01 -1.62616864e-01 -7.81190023e-02 -3.17346081e-02
-1.06604707e+00 1.16855156e+00 1.25192440e+00 -3.19536358e-01
-1.53928995e-01 -5.27622759e-01 -9.72075701e-01 3.82607877e-01
3.35796207e-01 -2.87155539e-01 -8.08298886e-01 -5.75401843e-01
9.04768407e-01 -1.86806947e-01 4.72298414e-01 -4.53425586e-01
4.96032327e-01 -1.81721393e-02 2.28241846e-01 -2.60822117e-01
2.05583513e-01 -8.17579687e-01 2.38537624e-01 -1.67354450e-01
-7.72785172e-02 -3.40429455e-01 -2.14952640e-02 1.40056908e+00
-4.48403269e-01 3.44219178e-01 1.34875596e-01 3.82512212e-02
-1.04369581e+00 9.04553294e-01 6.84644222e-01 5.68421304e-01
8.65343690e-01 1.40765727e-01 -5.35466790e-01 -2.45753899e-01
-9.47981179e-01 9.42331612e-01 1.63175881e-01 4.85157758e-01
6.67194486e-01 -1.90116453e+00 -6.63225293e-01 2.08137818e-02
4.98108327e-01 2.32719719e-01 7.05527544e-01 8.62949491e-01
-1.44741684e-01 4.45440888e-01 3.61863524e-01 -7.21153840e-02
-8.07412088e-01 7.30092287e-01 1.23216949e-01 -4.76327658e-01
-7.48513818e-01 1.30093300e+00 7.63630494e-02 -6.72099113e-01
3.15016478e-01 -7.09812194e-02 -2.62407511e-01 1.31848887e-01
1.73077852e-01 4.35720146e-01 3.22052091e-02 -6.16962612e-01
-4.76676702e-01 2.35723615e-01 -3.38464886e-01 6.21213973e-01
1.40742183e+00 -2.44434550e-01 -2.87754983e-01 1.90506771e-01
1.25796509e+00 3.01517576e-01 -9.20821130e-01 -9.63488877e-01
3.74160968e-02 -5.59614062e-01 1.35423958e-01 -6.23534501e-01
-1.46048737e+00 3.18415493e-01 -1.03920393e-01 1.70550376e-01
8.82215619e-01 1.83083862e-01 1.03678596e+00 6.01783633e-01
3.88354003e-01 -9.82745349e-01 -1.10141240e-01 5.20451665e-01
7.54745543e-01 -1.25018847e+00 2.72251308e-01 -7.47410953e-01
-7.09270597e-01 5.14329493e-01 6.69771969e-01 9.15851966e-02
9.94684219e-01 -2.52586514e-01 -5.20824790e-01 -6.54712141e-01
-6.37134075e-01 -3.97558779e-01 5.03343999e-01 6.32709980e-01
1.68555006e-01 3.78862530e-01 -1.28365740e-01 1.07813287e+00
6.97681261e-03 -1.87078270e-03 1.64796546e-01 5.14253020e-01
-8.43781084e-02 -9.84847605e-01 1.52916312e-01 7.38970160e-01
-3.03606570e-01 -4.73328143e-01 -4.75294501e-01 6.75340891e-01
7.67270848e-02 5.84430933e-01 -4.86461908e-01 -6.34622514e-01
4.35120940e-01 -1.58651561e-01 2.19448239e-01 -7.02443421e-01
-1.85837671e-01 -4.98032898e-01 1.91984564e-01 -4.82277036e-01
-2.06338614e-01 -2.98246264e-01 -1.29411209e+00 -4.72784251e-01
-3.95057112e-01 3.41414720e-01 -7.15927593e-03 6.69144571e-01
8.61748576e-01 6.79661870e-01 3.68582100e-01 -1.86038688e-01
-2.64621139e-01 -6.47834122e-01 -1.06494820e+00 5.97915828e-01
1.98150918e-01 -8.94512475e-01 -2.13522956e-01 -2.72707880e-01] | [8.800956726074219, 7.889023303985596] |
a4104120-15f0-4b57-b8ae-93ad1479b2a6 | extraction-of-heart-rate-from-ppg-signal-a | null | null | https://ieeexplore.ieee.org/abstract/document/9068845 | https://www.researchgate.net/publication/340687904_Extraction_of_Heart_Rate_from_PPG_Signal_A_Machine_Learning_Approach_using_Decision_Tree_Regression_Algorithm | Extraction of Heart Rate from PPG Signal: A Machine Learning Approach using Decision Tree Regression Algorithm | This research article shows a novel technique which used to measure the heart rate (HR) from wearable devices such as fingertip device, wrist type device. For HR monitoring Photoplethysmography (PPG) signal is severely used. HR measurement precision is affected by noise and motion artifacts (MA) at the moment of physical movement. There are many conventional algorithms to decrease the effect of MA and measure the HR variations. In this work, a novel method called multi-model machine learning approach (MMMLA) is used to predict HR. The technique is shown in this work, which primarily trains and tests the algorithm for
various features and various data sets. K-means cluster is employed for splitting noisy and non noisy data. This process helps the machine to learn from noisy and non noisy data. After separation, Decision Tree Regression method is employed to fit data and measure HR from test data. In this research feature engineering is completed in different words, a special set of the feature is chosen and check their behavior with the projected model and therefore the error rate for each set of the feature was computed. In this work, model feature is also reduced and check the behavior in estimating HR. For each case root mean square (RMS) error and average absolute error of HR was computed. The minimum average absolute error was traced in his work was 1.18 beats per minute (BPM). The outcome of this work demonstrates that the algorithm has a significant possibility to be used for PPG-based HR monitoring. | ['Md. Abdullah Al Mahmud', 'A.H.M. Zadidul Karim', 'Md. Sazal Miah', 'Shikder Shafiul Bashar'] | 2019-12-22 | null | null | null | null | ['photoplethysmography-ppg'] | ['medical'] | [ 7.66374692e-02 -4.94839884e-02 7.96187967e-02 -9.83458832e-02
-1.38269916e-01 -1.89655855e-01 -6.22034743e-02 -4.67714891e-02
-2.71799058e-01 8.97207201e-01 -9.10369027e-03 6.46560937e-02
-2.46153221e-01 -5.69403827e-01 6.47261888e-02 -7.51119256e-01
-1.86579525e-02 -1.22846663e-01 -2.64421970e-01 -2.76995183e-04
4.48881745e-01 5.01352608e-01 -1.41502512e+00 -2.03443065e-01
6.94145203e-01 8.17723930e-01 -1.25655070e-01 9.53954935e-01
1.75565347e-01 3.96293402e-01 -7.10910797e-01 4.33350027e-01
4.60937589e-01 -1.09102118e+00 -4.38098073e-01 -6.14204518e-02
6.74834922e-02 -1.89661905e-01 1.52955011e-01 3.98001283e-01
1.07726169e+00 2.90222228e-01 6.39170527e-01 -1.10656679e+00
1.62089318e-01 2.02281520e-01 -5.39774776e-01 3.93357247e-01
5.42754829e-01 8.27526823e-02 -1.55974403e-01 -6.10139132e-01
1.11833729e-01 7.24856734e-01 9.75484014e-01 2.48082787e-01
-1.24221468e+00 -5.23387134e-01 -8.01910996e-01 7.85567388e-02
-1.50314093e+00 -2.27361947e-01 1.10106003e+00 -4.21486288e-01
7.92593896e-01 7.66822398e-01 8.88764441e-01 4.19326752e-01
6.33640707e-01 -2.43479893e-01 1.96660292e+00 -7.75873363e-01
6.02718145e-02 7.04371095e-01 4.83957738e-01 5.02230883e-01
2.38831118e-01 2.13345096e-01 -3.66215318e-01 -1.32531645e-02
6.52904987e-01 -1.93220749e-02 -3.03912550e-01 3.98421079e-01
-6.38263226e-01 4.81043398e-01 -2.58300006e-01 8.03621709e-01
-3.16461205e-01 -2.63232917e-01 3.85158092e-01 3.78320843e-01
-6.61839396e-02 3.44297528e-01 -5.58498681e-01 -5.94084680e-01
-9.58598435e-01 1.24448203e-02 1.19965053e+00 4.18448538e-01
4.17285174e-01 1.05984665e-01 -1.52765960e-01 7.04567373e-01
2.25822926e-01 5.22926390e-01 8.05651903e-01 -7.09745884e-01
-1.35368466e-01 7.10457385e-01 1.07760504e-01 -1.14862478e+00
-7.29478657e-01 -3.11692923e-01 -8.40587258e-01 4.45772469e-01
4.05137300e-01 -2.92616278e-01 -8.81522894e-01 1.20136762e+00
5.66250741e-01 2.80412018e-01 2.31067669e-02 9.15508389e-01
9.70708549e-01 5.62550187e-01 1.38707951e-01 -8.22818160e-01
1.27734184e+00 -2.95438200e-01 -8.73875260e-01 4.59359139e-01
1.98066413e-01 -9.00088966e-01 8.59053910e-01 6.40101850e-01
-7.66705334e-01 -1.18836260e+00 -1.09288919e+00 3.39432508e-01
-3.33086610e-01 2.01716721e-01 1.13672942e-01 1.01590645e+00
-5.76661587e-01 1.16119623e+00 -6.30721092e-01 -5.87299764e-01
-2.68602133e-01 2.56703734e-01 -1.52092338e-01 4.84010816e-01
-1.18255651e+00 1.28512120e+00 1.35796130e-01 6.08928859e-01
-1.54451340e-01 -5.88529706e-01 -4.36004788e-01 -2.19843477e-01
-3.55675519e-01 -4.84571666e-01 4.73173231e-01 -7.16335595e-01
-2.00095558e+00 6.35866821e-01 -2.35773727e-01 -1.42475501e-01
6.17483318e-01 -1.11820787e-01 -6.99344397e-01 1.31531447e-01
-4.77908194e-01 -2.28225753e-01 9.19227004e-01 -8.04014444e-01
-6.77011162e-02 -4.63318288e-01 -8.18688095e-01 2.57060200e-01
1.30647823e-01 -1.00399025e-01 2.33865261e-01 -6.33412749e-02
4.10658926e-01 -9.17142570e-01 8.43982771e-02 -5.82600236e-01
-1.86427951e-01 -4.83543314e-02 7.40220904e-01 -1.03814375e+00
1.48127484e+00 -1.83410728e+00 -1.83755353e-01 7.92893291e-01
2.00093742e-02 2.85491914e-01 7.39635527e-01 4.93519098e-01
-1.26883551e-01 1.70628667e-01 -1.45577505e-01 1.39890105e-01
-4.11152542e-01 5.25025465e-02 2.24188223e-01 6.19046986e-01
-3.32339168e-01 2.46021062e-01 -9.50196311e-02 -6.49921656e-01
7.76184022e-01 6.61630690e-01 3.10403198e-01 4.14963931e-01
7.51808822e-01 9.04814243e-01 -5.97117022e-02 3.80312175e-01
7.57145762e-01 4.18554246e-01 3.68306972e-03 -7.26943493e-01
-3.02491099e-01 -3.47391576e-01 -1.60212409e+00 1.11541581e+00
-4.79563594e-01 3.98996472e-01 -5.10962009e-01 -8.72325957e-01
1.78397548e+00 6.44170463e-01 7.94947684e-01 -6.94518447e-01
3.07294071e-01 1.66384831e-01 1.08110525e-01 -1.25745893e+00
-1.39556587e-01 -3.32987130e-01 4.69664782e-01 2.72014230e-01
-2.48990804e-01 -4.47894186e-02 -1.93447232e-01 -4.87064838e-01
6.36712193e-01 3.16108227e-01 7.22658038e-01 -3.84428263e-01
7.92177379e-01 -1.40406445e-01 5.18438280e-01 6.45290136e-01
-5.78010499e-01 4.42911983e-01 1.39644340e-01 -5.93568563e-01
-8.65680456e-01 -6.56550586e-01 -5.92927933e-01 1.80111304e-01
-1.81629267e-02 5.85061014e-02 -4.65740800e-01 2.35771649e-02
-5.22682890e-02 3.84830028e-01 -4.80305344e-01 -1.92452282e-01
-5.98950684e-01 -7.59323776e-01 4.98734683e-01 1.28774688e-01
5.86523890e-01 -1.17792237e+00 -1.07652998e+00 2.41928712e-01
1.65799052e-01 -5.84517300e-01 3.00172448e-01 1.88570973e-02
-1.28297949e+00 -1.09918475e+00 -4.75539297e-01 -3.53631556e-01
2.18529224e-01 -2.94544935e-01 7.69297600e-01 7.45247453e-02
-9.12249565e-01 3.89629781e-01 -4.03429806e-01 -7.48336017e-01
-2.80949265e-01 -3.92826587e-01 7.37483352e-02 -1.02803230e-01
6.32808983e-01 -7.37606764e-01 -9.03606534e-01 2.89303780e-01
-1.58199340e-01 -3.88318717e-01 4.20246243e-01 3.62122476e-01
5.24215698e-01 -5.50166219e-02 7.51711965e-01 -5.88496923e-01
8.29344213e-01 -1.88197955e-01 -4.81815964e-01 9.22098309e-02
-1.19967341e+00 -2.77929217e-01 6.80001438e-01 -3.19714665e-01
-7.83780158e-01 9.41156521e-02 2.16403827e-01 -1.85692385e-01
-4.52933520e-01 3.44173819e-01 2.36076057e-01 -1.35240644e-01
9.97708857e-01 2.80672461e-01 3.96691740e-01 -3.80706429e-01
-1.98400691e-01 8.14410269e-01 3.82001102e-01 -2.19833851e-01
5.12898684e-01 -1.21361107e-01 7.22622454e-01 -1.07530344e+00
-1.57972783e-01 -6.41425788e-01 -6.22209489e-01 -8.52088869e-01
9.09232259e-01 -5.87426662e-01 -1.01189637e+00 4.94867295e-01
-6.92032933e-01 -3.36734578e-02 7.61603490e-02 9.83997524e-01
-4.99781638e-01 3.58822882e-01 -4.75122541e-01 -1.58601105e+00
-9.91849303e-01 -5.31878710e-01 1.84022173e-01 8.98799896e-01
-4.11921471e-01 -9.18056369e-01 1.97671518e-01 2.66850322e-01
5.51024258e-01 8.93324733e-01 2.83350825e-01 -2.48133197e-01
2.57564783e-01 -5.96489251e-01 3.06686670e-01 5.05281687e-01
4.41879332e-01 2.05420226e-01 -1.13945079e+00 -6.70709834e-02
6.30888760e-01 3.16513106e-02 3.54768068e-01 5.61942935e-01
8.96986127e-01 4.56662513e-02 -6.12335019e-02 3.85773361e-01
1.98896980e+00 6.29240632e-01 1.07176888e+00 1.74212292e-01
4.89064574e-01 5.00918627e-01 8.32564116e-01 5.06883800e-01
-2.56298870e-01 4.55797046e-01 -1.28503948e-01 -3.81589115e-01
1.17645122e-01 1.02912843e-01 2.30859131e-01 5.18786430e-01
-6.78236306e-01 3.94742399e-01 -5.27018309e-01 -1.80865407e-01
-1.37426770e+00 -1.09215295e+00 -7.23039269e-01 2.62049174e+00
7.60672212e-01 -6.28393069e-02 4.33531344e-01 8.72129023e-01
8.07497561e-01 -3.52397591e-01 -2.38716647e-01 -1.03330016e+00
1.79458991e-01 7.28284597e-01 4.69149232e-01 6.38249338e-01
-7.67619491e-01 1.41079962e-01 5.68148994e+00 1.22919060e-01
-1.44656301e+00 -1.38428226e-01 4.64406133e-01 1.05683856e-01
4.05366600e-01 4.14637513e-02 -6.48223937e-01 7.33434439e-01
1.17819726e+00 -2.22741887e-01 3.23163867e-01 7.30592489e-01
9.86298323e-01 -7.66318262e-01 -6.91966236e-01 1.40822971e+00
2.01071739e-01 -5.07583797e-01 -7.67427325e-01 -2.01898113e-01
2.72305220e-01 -6.73322141e-01 -4.52734172e-01 1.17677808e-01
-8.55355561e-01 -1.02619219e+00 -1.93502843e-01 1.21521032e+00
6.40033305e-01 -6.40987635e-01 1.04774845e+00 2.43393369e-02
-1.11647320e+00 4.00390998e-02 -2.93823868e-01 -2.97751099e-01
2.78653868e-04 8.45356226e-01 -8.23667824e-01 5.64288378e-01
4.09506708e-01 2.67321795e-01 -4.13828731e-01 1.27657986e+00
1.12697490e-01 7.46586382e-01 -6.05884194e-01 -1.81248114e-01
-4.69330996e-01 -5.08930981e-01 4.74654734e-01 1.05444300e+00
4.77699518e-01 3.32687914e-01 -1.60134432e-03 9.67437267e-01
7.52621233e-01 5.66504836e-01 -6.61830187e-01 5.20554721e-01
5.12851596e-01 1.39177740e+00 -6.21692598e-01 -2.76698768e-01
-7.28174448e-02 8.09664369e-01 -4.74314570e-01 1.16858959e-01
-9.63358343e-01 -8.34833801e-01 -1.53812602e-01 3.01456928e-01
-4.99771684e-01 -2.62748450e-02 -7.55741835e-01 -4.72825795e-01
6.92105591e-02 -4.67338949e-01 4.72476482e-01 -6.68413341e-01
-8.05377722e-01 3.76943290e-01 1.38328657e-01 -1.39486206e+00
-1.33291870e-01 -2.36368433e-01 -8.80882740e-01 1.56527805e+00
-1.01046097e+00 -4.55265760e-01 -8.43192816e-01 6.46825790e-01
3.69578391e-01 8.34685564e-02 9.31461453e-01 3.50990325e-01
-7.94455647e-01 3.72147650e-01 -2.47862354e-01 -1.40311673e-01
7.88679183e-01 -1.24440920e+00 -6.75393045e-01 6.97318912e-01
-5.03207386e-01 6.50664091e-01 1.07156789e+00 -6.64602876e-01
-1.28543329e+00 -5.70376277e-01 8.77010107e-01 -1.20384507e-01
-6.54085055e-02 2.38585517e-01 -6.97166741e-01 2.46140733e-02
2.55839005e-02 1.47083297e-01 8.09472084e-01 -8.64712819e-02
3.70106071e-01 -6.67064130e-01 -1.62982929e+00 8.82090405e-02
1.38651788e-01 -2.92037219e-01 -5.20073950e-01 -6.94771251e-03
-1.30364150e-01 -3.98137152e-01 -1.58350408e+00 3.56887668e-01
1.09823966e+00 -1.13968968e+00 4.46329355e-01 -2.48813748e-01
1.30002610e-02 -4.90754128e-01 3.30560058e-01 -9.77730513e-01
-1.73216417e-01 -7.14630604e-01 -1.13675281e-01 1.31834269e+00
3.74247551e-01 -7.47268319e-01 6.47316813e-01 1.13395786e+00
6.96764067e-02 -7.05277681e-01 -5.62082410e-01 -6.44956946e-01
-4.70460802e-01 4.12111618e-02 -1.86928645e-01 9.89434659e-01
3.62188697e-01 2.68709183e-01 -6.12973630e-01 -6.73970506e-02
7.04105258e-01 -3.54281589e-02 7.88722873e-01 -1.22886264e+00
-3.83906603e-01 2.79728621e-01 -5.50237596e-01 2.24751644e-02
-6.56747162e-01 -3.31939459e-01 -2.68353075e-01 -1.41037965e+00
-1.22000083e-01 -4.09811229e-01 -5.28179109e-01 1.55285627e-01
-2.66265452e-01 1.48069218e-01 1.26781598e-01 9.00122523e-02
4.03516918e-01 -1.88354701e-01 9.94607627e-01 4.97128278e-01
-1.22015154e+00 4.05780107e-01 -6.03339821e-02 1.43147781e-01
1.29397547e+00 -4.52584952e-01 -5.13063729e-01 6.35328531e-01
-1.62151709e-01 4.41210061e-01 2.39083767e-01 -1.43445468e+00
-1.28277674e-01 -1.08523883e-01 1.00233185e+00 -4.42338258e-01
7.66250044e-02 -1.01935053e+00 8.17736268e-01 8.07697654e-01
-1.31153777e-01 1.00507215e-01 -3.69544029e-02 1.09413125e-01
1.68419033e-02 -2.68243074e-01 1.01880932e+00 -3.81668270e-01
-2.75543928e-01 -1.96887374e-01 -1.68355271e-01 -4.40948993e-01
1.08291590e+00 -9.35959041e-01 -1.70581654e-01 -2.53682166e-01
-1.14826703e+00 -2.52567738e-01 -1.34120896e-01 -1.00156009e-01
5.41010141e-01 -1.05112636e+00 -5.64954519e-01 1.19140424e-01
-4.41082828e-02 -6.12241030e-01 6.38259530e-01 1.48907113e+00
-6.59912407e-01 9.82032716e-02 -5.52452981e-01 -4.85898554e-01
-1.59666967e+00 4.19473469e-01 7.10134566e-01 1.16254173e-01
-6.26274228e-01 2.14077279e-01 -1.10152340e+00 4.07563806e-01
-2.82278168e-04 -2.50578851e-01 -7.67032802e-01 -5.97219691e-02
4.62076128e-01 9.17562127e-01 -2.94101071e-02 -2.54539162e-01
-1.24494620e-01 1.04562068e+00 5.35280049e-01 -1.18610926e-01
8.00044358e-01 -3.95477533e-01 -1.74987540e-01 8.27305198e-01
1.18115461e+00 1.95263162e-01 -7.04480886e-01 4.43427950e-01
-1.21204637e-01 -3.72041404e-01 -1.58202097e-01 -1.12228882e+00
-6.84349358e-01 6.03656054e-01 1.52224433e+00 2.11844295e-01
1.39527762e+00 -6.96892083e-01 3.31804842e-01 5.66095077e-02
3.76861691e-02 -1.57469773e+00 -4.22155946e-01 -2.60005414e-01
7.05444515e-01 -9.94551063e-01 2.65571147e-01 -3.73237193e-01
-7.00441778e-01 1.67905879e+00 6.07878208e-01 -2.47502878e-01
9.72035050e-01 1.74279496e-01 2.67183959e-01 6.07839040e-02
-1.88330427e-01 -4.92808782e-02 5.08787222e-02 7.15939581e-01
8.19515646e-01 1.15061037e-01 -1.44522262e+00 4.31626409e-01
-1.91235185e-01 7.19042778e-01 6.56247497e-01 8.63463938e-01
-7.62636483e-01 -7.23897874e-01 -5.92994988e-01 6.18768752e-01
-6.21351063e-01 1.87099040e-01 -5.40211499e-02 8.75440121e-01
3.53240877e-01 1.38065112e+00 -2.59079605e-01 -6.56772196e-01
4.90984529e-01 4.46360886e-01 5.28210759e-01 -5.69392182e-02
-8.86369288e-01 7.42749274e-02 6.03667013e-02 -4.59830105e-01
-5.62772095e-01 -3.64765137e-01 -1.29189277e+00 -1.54884011e-01
-9.02132615e-02 3.78973544e-01 1.22197187e+00 6.85063243e-01
9.08898413e-02 2.36880720e-01 1.01318395e+00 -3.56322140e-01
-3.56425107e-01 -1.31471074e+00 -7.95520723e-01 5.14305830e-01
1.69525012e-01 -4.52363461e-01 -7.68950939e-01 1.24386601e-01] | [13.967494010925293, 3.0176289081573486] |
de2075a6-785c-4799-bbc0-de02d63b7296 | dl-droid-deep-learning-based-android-malware | 1911.10113 | null | https://arxiv.org/abs/1911.10113v1 | https://arxiv.org/pdf/1911.10113v1.pdf | DL-Droid: Deep learning based android malware detection using real devices | The Android operating system has been the most popular for smartphones and tablets since 2012. This popularity has led to a rapid raise of Android malware in recent years. The sophistication of Android malware obfuscation and detection avoidance methods have significantly improved, making many traditional malware detection methods obsolete. In this paper, we propose DL-Droid, a deep learning system to detect malicious Android applications through dynamic analysis using stateful input generation. Experiments performed with over 30,000 applications (benign and malware) on real devices are presented. Furthermore, experiments were also conducted to compare the detection performance and code coverage of the stateful input generation method with the commonly used stateless approach using the deep learning system. Our study reveals that DL-Droid can achieve up to 97.8% detection rate (with dynamic features only) and 99.6% detection rate (with dynamic + static features) respectively which outperforms traditional machine learning techniques. Furthermore, the results highlight the significance of enhanced input generation for dynamic analysis as DL-Droid with the state-based input generation is shown to outperform the existing state-of-the-art approaches. | ['Suleiman Y. Yerima', 'Sakir Sezer', 'Mohammed K. Alzaylaee'] | 2019-11-22 | null | null | null | null | ['android-malware-detection'] | ['miscellaneous'] | [ 9.41664279e-02 -4.27359223e-01 -7.22123563e-01 2.47587025e-01
-4.28765833e-01 -7.65471756e-01 6.94629490e-01 -2.71613508e-01
-1.13469772e-01 5.01981020e-01 -3.70595574e-01 -1.04863811e+00
2.58821011e-01 -5.71141303e-01 -5.89249313e-01 -5.35456538e-01
-3.50897163e-01 -1.69808462e-01 2.84916371e-01 -2.36697480e-01
2.86372334e-01 4.49221730e-01 -1.73113012e+00 3.31663907e-01
7.74240315e-01 1.14644742e+00 -1.88638702e-01 1.07131445e+00
1.35197282e-01 6.31611764e-01 -1.08878803e+00 -4.90504950e-01
2.77690858e-01 9.96804535e-02 -5.48130989e-01 -5.33771157e-01
3.78407657e-01 -7.93141663e-01 -3.72109801e-01 1.21697736e+00
3.99558067e-01 -4.33879048e-01 2.51881808e-01 -1.62446845e+00
-7.31119633e-01 4.06266868e-01 -8.41020763e-01 5.62713683e-01
5.35344601e-01 2.55492806e-01 4.83663529e-01 -2.59904504e-01
1.61030695e-01 1.16082132e+00 7.11445749e-01 7.63086081e-01
-8.40750813e-01 -9.75731969e-01 -2.71184891e-01 7.10831881e-02
-1.24460173e+00 -2.91675121e-01 6.98269606e-01 -5.44993758e-01
1.35320675e+00 3.45836878e-01 5.21981716e-01 1.72138906e+00
7.92648971e-01 5.45946658e-01 9.30414081e-01 -1.78340867e-01
2.19313756e-01 1.80870548e-01 4.01573002e-01 8.83142054e-01
7.88360834e-01 4.10516381e-01 1.40347883e-01 -7.29102612e-01
6.11774214e-02 5.90713322e-01 1.26350850e-01 2.30523143e-02
-6.14636600e-01 9.81570423e-01 5.69319166e-02 1.04613714e-01
1.78222992e-02 2.00264186e-01 1.14577103e+00 7.80407637e-02
3.29636902e-01 3.29481512e-01 -6.08543336e-01 -6.91154063e-01
-8.68725717e-01 -1.21576134e-02 8.30940127e-01 5.21839201e-01
3.63872290e-01 5.64012706e-01 1.84366331e-01 3.52811396e-01
3.90255630e-01 8.02619457e-01 9.97446835e-01 -8.02266419e-01
2.16166705e-01 8.51932228e-01 -2.65055716e-01 -1.25842321e+00
5.49971722e-02 -2.71503180e-01 -6.80335939e-01 3.35847259e-01
1.20100193e-02 -4.26914066e-01 -7.08271682e-01 1.41850114e+00
1.18630245e-01 4.69744802e-01 1.03794836e-01 -1.29613683e-01
5.00447869e-01 8.12654316e-01 -2.56021768e-01 -2.24661723e-01
1.28483748e+00 -6.90114021e-01 -5.73218465e-01 -1.82853803e-01
8.37133229e-01 -5.51794052e-01 1.39866650e+00 6.53343737e-01
-3.66992831e-01 -5.31284451e-01 -1.31914198e+00 3.58157665e-01
-7.72134125e-01 2.89532989e-01 6.09207273e-01 1.52567708e+00
-1.04804862e+00 7.82010704e-02 -1.24148655e+00 -5.98301627e-02
7.42935359e-01 7.67653406e-01 -1.20634608e-01 4.84273285e-01
-7.52200484e-01 2.76937187e-01 3.73109162e-01 -5.45872986e-01
-1.39900112e+00 -2.76962548e-01 -7.18492091e-01 -9.46673080e-02
5.66337824e-01 2.10594937e-01 1.48643684e+00 -7.07675159e-01
-1.57672942e+00 5.44683337e-01 1.70688573e-02 -7.70565331e-01
1.24853529e-01 -5.76931953e-01 -7.94803023e-01 -4.81259078e-02
-4.72761951e-02 3.25982086e-03 1.21312380e+00 -9.40589428e-01
-4.96850759e-01 -2.81719148e-01 3.48464429e-01 -6.35167181e-01
-8.00351501e-01 4.08058204e-02 -1.33244604e-01 -3.96332771e-01
-9.24035013e-01 -1.42043579e+00 3.05499494e-01 -5.82124293e-01
-6.32295966e-01 -1.29265174e-01 2.00110793e+00 -7.79470623e-01
1.63661993e+00 -2.25623679e+00 -4.25811827e-01 -1.99233696e-01
5.13159931e-01 1.28556788e+00 1.64281830e-01 1.72163352e-01
-7.74778277e-02 5.07070899e-01 -1.10580675e-01 -2.00389728e-01
-3.76188129e-01 1.66200742e-01 -5.66922128e-01 3.54027748e-01
2.92137917e-02 9.19961035e-01 -9.00640845e-01 -1.74338862e-01
2.57637054e-01 6.69612408e-01 -6.49402916e-01 4.71818112e-02
-2.59647340e-01 -6.04181737e-02 -3.62145901e-01 9.39568579e-01
6.68100059e-01 -3.07364136e-01 1.18141435e-01 2.44778186e-01
1.16642609e-01 4.03515875e-01 -3.30650896e-01 7.87297130e-01
-6.63409591e-01 8.90661895e-01 -1.42375141e-01 -6.15413547e-01
7.15101361e-01 4.18051798e-03 2.28117675e-01 -3.55371445e-01
4.43559855e-01 2.08275706e-01 4.29305702e-01 -5.72318971e-01
4.63390112e-01 7.41789758e-01 -2.63394177e-01 7.03832984e-01
-2.51484632e-01 4.75429654e-01 -2.19068885e-01 3.74891162e-01
1.35966802e+00 -2.29047641e-01 5.99630475e-01 -9.23311338e-03
8.30660939e-01 -3.19796652e-01 1.25534892e-01 6.93691790e-01
-6.67610526e-01 -3.40875149e-01 7.64596939e-01 -2.95846194e-01
-5.53506970e-01 -8.73494864e-01 1.78850546e-01 1.21476948e+00
-8.83576348e-02 -7.14053988e-01 -1.32574821e+00 -1.44250607e+00
-1.13412172e-01 4.29070741e-01 -8.52607071e-01 -6.39994979e-01
-6.23363614e-01 -6.77207410e-01 1.16585457e+00 2.98297644e-01
9.57530260e-01 -9.69157279e-01 -7.97677040e-01 -2.01423436e-01
3.23389620e-01 -1.02529705e+00 -2.58071512e-01 -2.11993549e-02
-6.73562050e-01 -1.29692268e+00 -1.68815657e-01 -5.84939361e-01
2.03991726e-01 4.73693758e-01 6.55319571e-01 2.58802056e-01
-3.60040307e-01 2.69244552e-01 -3.27926517e-01 -3.45437646e-01
-1.04331028e+00 5.32455325e-01 4.23888355e-01 -1.66158155e-01
5.91158390e-01 -1.61670208e-01 -3.02296221e-01 1.58929795e-01
-9.35219646e-01 -7.62111068e-01 5.04129410e-01 5.63530862e-01
1.28492475e-01 3.17813337e-01 5.69110453e-01 -1.08271968e+00
8.93708885e-01 -8.47450793e-01 -7.77781188e-01 -1.35585010e-01
-8.05745304e-01 -1.32747829e-01 8.71713102e-01 -9.52651262e-01
-7.92308152e-01 -1.20962203e-01 -5.82873166e-01 -5.15837073e-01
-1.28183812e-01 -3.80087234e-02 -2.35972896e-01 -4.23060745e-01
8.83962512e-01 4.54705507e-01 4.28823292e-01 -1.48716524e-01
-3.15366052e-02 1.11298871e+00 3.26921463e-01 -6.80455267e-02
5.96048057e-01 3.67203951e-01 -1.74106315e-01 -1.11089826e+00
-3.43436807e-01 -6.14185221e-02 -2.14135155e-01 1.50084034e-01
7.28285551e-01 -6.50152564e-01 -1.13609302e+00 1.01446366e+00
-1.00812149e+00 -2.95243591e-01 2.18359679e-01 -1.60793260e-01
-7.14227557e-02 6.96409106e-01 -7.39456892e-01 -8.58087540e-01
-6.40772879e-01 -1.78303921e+00 1.16058433e+00 6.08754046e-02
-1.30252093e-01 -1.01607621e+00 9.59880203e-02 6.62140772e-02
5.50627172e-01 4.23321307e-01 7.02108920e-01 -1.08457005e+00
-2.36226335e-01 -5.37168443e-01 -7.25083277e-02 6.68825865e-01
7.29681373e-01 5.60419977e-01 -1.12735105e+00 -5.76729059e-01
2.28077099e-01 1.37386575e-01 4.81691778e-01 1.55167371e-01
1.58570945e+00 -9.28460300e-01 -7.38706648e-01 6.76038086e-01
1.15699780e+00 7.56403565e-01 5.99600911e-01 2.06814319e-01
1.01503372e+00 -5.60701303e-02 3.33234191e-01 3.24304432e-01
8.79186690e-02 5.75244188e-01 9.84257817e-01 3.06381941e-01
1.58558469e-02 -1.51426733e-01 1.02858770e+00 5.51072896e-01
2.87208974e-01 -2.90416926e-01 -8.89186740e-01 6.99626729e-02
-1.18480361e+00 -9.51349437e-01 -1.95131656e-02 2.34046245e+00
5.67907453e-01 5.64660847e-01 6.24361575e-01 4.71519291e-01
6.86060011e-01 3.23048651e-01 -6.55622423e-01 -1.08555520e+00
4.39480931e-01 2.30787560e-01 4.60294396e-01 4.25003946e-01
-1.23801517e+00 9.77062166e-01 6.69843769e+00 9.74796712e-01
-1.71652997e+00 4.39027607e-01 6.46677136e-01 3.03698573e-02
2.39546508e-01 -4.35289741e-01 -1.22044837e+00 1.10442054e+00
1.57190382e+00 -1.56075314e-01 4.83486205e-01 1.59462225e+00
-8.18733051e-02 1.16666839e-01 -6.74315572e-01 1.11564279e+00
2.25034937e-01 -1.33699882e+00 -6.37322143e-02 8.64819288e-01
6.76255643e-01 2.50129968e-01 6.00644946e-01 6.27965748e-01
1.64605737e-01 -7.98736572e-01 1.41786173e-01 -2.49631315e-01
1.13584864e+00 -9.70432699e-01 8.05864990e-01 3.75663340e-01
-1.09821486e+00 -6.75234020e-01 3.81743498e-02 -1.67416230e-01
-2.94519931e-01 2.21765772e-01 -1.25208414e+00 -1.07121401e-01
8.23425531e-01 7.82499969e-01 -1.06086648e+00 4.42681462e-01
3.81705537e-02 1.18807888e+00 -1.11648493e-01 -4.37021583e-01
2.05224574e-01 3.73745143e-01 4.92024571e-01 1.25718629e+00
3.57304662e-01 -6.49463236e-01 -3.54418792e-02 4.37817812e-01
-1.28907442e-01 -4.08637941e-01 -1.21286285e+00 -4.74822074e-01
5.11555672e-01 1.20824921e+00 -5.30865550e-01 -5.05580187e-01
-3.14536214e-01 8.78341496e-01 -1.10740505e-01 -4.65211757e-02
-1.16612637e+00 -5.65609634e-01 1.13800740e+00 4.08930033e-01
1.60155967e-01 -4.45053756e-01 4.39438112e-02 -8.71317506e-01
-3.22317511e-01 -1.35105324e+00 1.86615229e-01 -1.66897938e-01
-8.24994922e-01 7.46782422e-01 1.89768881e-01 -1.14713597e+00
-7.67101169e-01 -1.05076087e+00 -8.32339585e-01 2.79205501e-01
-8.85661125e-01 -8.50636959e-01 -1.59641072e-01 5.08632660e-01
6.44706964e-01 -6.62088811e-01 6.38934195e-01 2.94922769e-01
-9.51263011e-01 1.13768268e+00 2.09207252e-01 1.28954068e-01
1.98739350e-01 -9.55262959e-01 7.66795635e-01 1.04041910e+00
-2.87421763e-01 1.11860335e+00 2.88954735e-01 -1.00700879e+00
-1.76551723e+00 -1.20435524e+00 1.57014709e-02 -5.93177438e-01
8.33533585e-01 -5.55238426e-01 -9.49888408e-01 7.83157527e-01
1.94154546e-01 -5.70358746e-02 7.75361955e-01 -4.66171384e-01
-6.14629567e-01 -1.71055049e-01 -1.32551539e+00 6.35340571e-01
5.63123763e-01 -6.99662626e-01 6.73459796e-03 1.34766355e-01
1.09262121e+00 -2.47231156e-01 -2.57142723e-01 3.17564368e-01
6.25927031e-01 -1.11881673e+00 9.44451332e-01 -3.55469137e-01
1.02447808e-01 -5.76412678e-02 -1.33027896e-01 -7.29868233e-01
6.56385645e-02 -9.95622218e-01 -1.24963355e+00 9.60332572e-01
8.47802535e-02 -1.01837301e+00 9.22267318e-01 -2.45484620e-01
5.41453026e-02 -9.23437119e-01 -8.19680512e-01 -1.04021966e+00
-9.13982466e-02 -4.64751661e-01 7.72353947e-01 6.58441782e-01
-4.30789322e-01 1.15513146e-01 -3.78195435e-01 1.69920996e-02
2.83209920e-01 -4.81046468e-01 9.53276634e-01 -1.12798750e+00
-4.89527792e-01 -4.56327766e-01 -5.34470320e-01 -7.16490746e-01
2.97898859e-01 -5.83572328e-01 -5.66218376e-01 -5.94632328e-01
4.72908765e-02 2.20270157e-02 -4.35395427e-02 7.06536353e-01
-4.53493483e-02 4.99253452e-01 -1.34259433e-01 2.35528916e-01
-4.00703222e-01 1.60828773e-02 5.64912200e-01 -1.91755280e-01
-4.94943053e-01 4.95584279e-01 -8.34861100e-01 8.22428524e-01
1.20472360e+00 -3.15057993e-01 -6.62451923e-01 1.46636754e-01
2.74587274e-01 -6.28292680e-01 3.22720379e-01 -1.15418851e+00
-4.77921456e-01 2.22867832e-01 1.13793723e-01 -4.17869002e-01
3.47536087e-01 -5.61091959e-01 -2.14685842e-01 1.22814941e+00
1.45468175e-01 4.74406123e-01 6.60493791e-01 6.50305033e-01
3.26711804e-01 7.53961802e-02 7.83943653e-01 1.56275108e-01
-6.99401915e-01 1.75212696e-01 -6.85542762e-01 -1.40338436e-01
1.48331952e+00 -4.48561102e-01 -7.96450078e-01 -2.43643329e-01
-1.58569902e-01 -6.51664555e-01 5.36089003e-01 6.65542245e-01
6.63697720e-01 -9.89811420e-01 3.42514254e-02 6.08339489e-01
5.83665147e-02 -7.32031345e-01 -6.03580214e-02 5.75680435e-01
-6.74969614e-01 5.36105394e-01 -2.04447493e-01 -6.97145224e-01
-1.75268161e+00 1.01313102e+00 2.21666142e-01 -3.78503084e-01
-2.41199225e-01 5.53899467e-01 -1.61362559e-01 -3.50224823e-01
1.65281519e-01 -1.88127860e-01 -9.65197533e-02 -3.52156579e-01
8.42369318e-01 8.81431699e-01 1.43850043e-01 -8.04176092e-01
-6.61674201e-01 1.59845144e-01 -5.22394836e-01 4.05287385e-01
7.52217591e-01 2.56410003e-01 -2.18866289e-01 1.18820585e-01
1.56891429e+00 4.84323621e-01 -9.91007864e-01 3.34626377e-01
-1.44930661e-01 -4.94222671e-01 -7.88835958e-02 -7.58351803e-01
-1.04961431e+00 1.04613447e+00 1.19479573e+00 9.03227448e-01
1.01220965e+00 -2.49918789e-01 1.20099008e+00 3.25552613e-01
3.57708782e-01 -2.89505661e-01 4.63791609e-01 5.29516697e-01
2.83941656e-01 -1.29623628e+00 -1.90706894e-01 -2.31151313e-01
-2.24823594e-01 7.92127550e-01 8.67092431e-01 -1.61936224e-01
8.28582287e-01 6.50217831e-01 -2.89256662e-01 -7.79302195e-02
-4.12952900e-01 4.84552503e-01 1.54816598e-01 8.97552252e-01
1.94665924e-01 1.94424093e-01 -1.85106210e-02 2.98110992e-01
-2.17760220e-01 -2.27707312e-01 7.01931596e-01 1.00856161e+00
-3.31620574e-01 -1.20033813e+00 -3.34083527e-01 7.93976784e-01
-1.05501795e+00 -5.66965342e-02 -6.95715308e-01 8.42826903e-01
2.32210383e-01 1.12961054e+00 -1.40436083e-01 -1.14876747e+00
-4.46025699e-01 -2.15939909e-01 -3.74873951e-02 -7.72530913e-01
-6.78366899e-01 -3.57055008e-01 -2.51784414e-01 -6.47537351e-01
3.71196061e-01 -2.60417402e-01 -9.76750135e-01 -5.92852294e-01
-3.56956899e-01 -1.76040143e-01 9.48317289e-01 6.44268692e-01
9.97601926e-01 5.00228107e-01 6.61483526e-01 -8.95546257e-01
-5.97203135e-01 -7.27426231e-01 -1.66809503e-02 -2.07003266e-01
8.42925370e-01 -7.77380645e-01 -5.77672839e-01 -7.56229609e-02] | [14.420245170593262, 9.678712844848633] |
a833d59b-8b36-49cc-8193-655c3ec74492 | whittle-index-policy-for-crawling-ephemeral | 1503.08558 | null | http://arxiv.org/abs/1503.08558v1 | http://arxiv.org/pdf/1503.08558v1.pdf | Whittle Index Policy for Crawling Ephemeral Content | We consider a task of scheduling a crawler to retrieve content from several
sites with ephemeral content. A user typically loses interest in ephemeral
content, like news or posts at social network groups, after several days or
hours. Thus, development of timely crawling policy for such ephemeral
information sources is very important. We first formulate this problem as an
optimal control problem with average reward. The reward can be measured in the
number of clicks or relevant search requests. The problem in its initial
formulation suffers from the curse of dimensionality and quickly becomes
intractable even with moderate number of information sources. Fortunately, this
problem admits a Whittle index, which leads to problem decomposition and to a
very simple and efficient crawling policy. We derive the Whittle index and
provide its theoretical justification. | ['Borkar Vivek EE-IIT', 'Avrachenkov Konstantin INRIA Sophia Antipolis'] | 2015-03-30 | null | null | null | null | ['problem-decomposition'] | ['miscellaneous'] | [-2.26821527e-01 -1.22526288e-01 -4.05756027e-01 2.79700339e-01
-8.33855271e-01 -8.26903701e-01 5.62235296e-01 1.77607134e-01
-4.82574075e-01 9.66851354e-01 -1.27209201e-01 -3.30504179e-01
-5.64580023e-01 -8.18915725e-01 -5.91039121e-01 -7.11205125e-01
-3.34407032e-01 4.88296807e-01 4.87009317e-01 -1.52444869e-01
2.76522011e-01 1.90751791e-01 -1.23081768e+00 -5.01827776e-01
7.15205133e-01 1.16607976e+00 7.87028134e-01 4.67712969e-01
-2.80694723e-01 5.78026652e-01 -4.42294091e-01 -1.66741818e-01
6.15127742e-01 -2.14786857e-01 -8.24788451e-01 -1.40887573e-01
-4.75513577e-01 -5.43924630e-01 -2.10354209e-01 9.95772362e-01
4.52884704e-01 3.97816710e-02 2.89106160e-01 -1.54684126e+00
-2.19331682e-01 4.38814253e-01 -9.25614059e-01 5.64933181e-01
2.56317973e-01 -3.30563575e-01 1.20437872e+00 -4.06055808e-01
8.26483190e-01 5.93338788e-01 2.23013029e-01 6.00634143e-03
-7.13061929e-01 -2.01683372e-01 2.04911008e-01 1.11381449e-01
-9.48663473e-01 -3.87618989e-02 7.72732913e-01 -4.85329963e-02
3.94295484e-01 2.24093914e-01 7.64415741e-01 9.41514552e-01
2.37788379e-01 9.47069228e-01 1.00443685e+00 -3.07692233e-02
3.72361004e-01 1.83003590e-01 1.36615619e-01 2.11496860e-01
8.34989071e-01 -2.87004977e-01 -6.41612530e-01 -4.34696466e-01
5.58030665e-01 1.18880942e-01 -2.36815736e-01 -3.44386548e-01
-9.11179841e-01 8.47378850e-01 9.42369848e-02 1.83401659e-01
-7.33901560e-01 6.82118610e-02 1.08805634e-02 8.48824143e-01
1.16171882e-01 4.66138572e-01 -4.00054187e-01 -6.28691375e-01
-6.82297707e-01 3.42035860e-01 1.29899991e+00 1.01257443e+00
5.56459069e-01 -2.67159194e-01 3.39085162e-01 6.52747691e-01
-3.04783117e-02 7.08876729e-01 4.10582453e-01 -8.35140169e-01
5.30128121e-01 1.65833607e-01 9.08853114e-01 -1.05914974e+00
-2.30535194e-01 -5.34192383e-01 -4.69810158e-01 -3.70139033e-01
6.13404274e-01 -5.90332389e-01 1.55010121e-02 1.39705515e+00
6.11644864e-01 -7.78500503e-03 -4.35949534e-01 8.48154306e-01
-1.44511685e-01 7.11948574e-01 -4.05575067e-01 -8.64014149e-01
1.05498409e+00 -8.20381701e-01 -8.45242918e-01 -1.73528737e-03
2.66363442e-01 -7.02809513e-01 9.24880803e-01 5.04365265e-01
-1.20904160e+00 4.09676224e-01 -5.80866933e-01 3.21352392e-01
-3.90019834e-01 -5.15144587e-01 4.03552741e-01 1.91154867e-01
-7.75464833e-01 5.77178776e-01 -5.22801936e-01 -5.61050534e-01
-1.56454235e-01 1.63158983e-01 2.66031504e-01 5.65713458e-02
-1.08051360e+00 6.87452316e-01 -5.11799045e-02 -2.87955403e-01
-9.38211560e-01 -4.37310636e-01 -4.69675623e-02 2.56426543e-01
9.49070692e-01 -8.04398358e-02 1.27548289e+00 -3.42913240e-01
-1.31228197e+00 2.82128990e-01 1.50033385e-01 -2.85889715e-01
6.65823698e-01 -3.14791083e-01 -2.61895835e-01 2.00822815e-01
5.40198445e-01 -2.81587869e-01 9.84236658e-01 -1.17848635e+00
-9.80956078e-01 -1.56321481e-01 2.14127913e-01 2.35900164e-01
-8.36506128e-01 8.22619349e-02 -4.87605304e-01 -3.68642122e-01
-2.24176913e-01 -8.32674205e-01 -3.10846001e-01 -3.92001241e-01
-1.36381865e-01 -3.08909267e-01 1.02266109e+00 -5.52081645e-01
1.33329272e+00 -1.78036058e+00 -1.18619777e-01 1.51719809e-01
1.53158441e-01 -5.32985181e-02 -1.03716902e-01 8.27407718e-01
5.13770521e-01 2.47674629e-01 2.89066821e-01 3.22206974e-01
-4.78824042e-02 2.03400850e-01 -3.35263133e-01 2.13188201e-01
-3.19832146e-01 5.59414864e-01 -1.31854308e+00 -3.97530794e-01
-3.98237199e-01 -6.83230627e-03 -2.29403287e-01 3.96709740e-01
-2.60607183e-01 8.98416340e-03 -1.39057672e+00 5.63595951e-01
5.11639237e-01 -9.30128574e-01 1.67355940e-01 4.51950073e-01
-3.71449620e-01 3.09544533e-01 -1.15906823e+00 1.16805375e+00
-6.45187080e-01 3.42406332e-01 7.73908973e-01 -1.07384586e+00
5.50047159e-01 4.15417969e-01 1.12855363e+00 -4.01279718e-01
2.54627138e-01 9.22456831e-02 -6.08075798e-01 -6.70763075e-01
4.78955984e-01 4.31687593e-01 -5.61450347e-02 1.30770755e+00
-5.58956742e-01 3.25270057e-01 5.61693370e-01 4.76774335e-01
1.62305486e+00 -3.09229314e-01 2.51703292e-01 -4.35100049e-01
9.72359627e-03 1.39041394e-01 3.82184744e-01 1.13251531e+00
-1.35938331e-01 -4.50166762e-02 8.52048695e-01 -1.91029832e-01
-9.28601861e-01 -9.26925302e-01 5.63869579e-03 1.16877174e+00
5.21174908e-01 -3.93147945e-01 -6.73049569e-01 -6.73053622e-01
-2.07720567e-02 3.87939334e-01 -4.01954263e-01 2.49294713e-01
-4.37057674e-01 -8.67543280e-01 -2.59616345e-01 -1.14947371e-01
6.16430461e-01 -9.45085049e-01 -7.72709608e-01 5.84330320e-01
-4.09206331e-01 -1.11141706e+00 -6.65638745e-01 1.63168252e-01
-6.76750779e-01 -1.29354775e+00 -1.07898176e+00 -6.14475965e-01
6.54482007e-01 1.00686085e+00 1.00071526e+00 1.17264554e-01
-3.65335941e-01 8.03616047e-01 -7.18282640e-01 -3.05123478e-01
1.99888483e-01 2.88388163e-01 9.50256512e-02 -3.38900904e-03
2.00654715e-01 -7.14120269e-01 -9.00149167e-01 5.24800241e-01
-9.96877730e-01 -3.68637025e-01 4.33247834e-01 7.86656797e-01
4.83236760e-01 5.19437015e-01 1.11679363e+00 -7.49860048e-01
1.38287520e+00 -1.22710192e+00 -1.01294589e+00 3.25627148e-01
-8.70258391e-01 1.36904623e-02 7.71135807e-01 -5.57986140e-01
-8.91249299e-01 -6.02028817e-02 6.19890749e-01 2.97956288e-01
4.02162015e-01 6.85428023e-01 2.79406458e-01 4.43632789e-02
4.65532184e-01 3.66214097e-01 -1.19339794e-01 -6.69755876e-01
1.58022299e-01 9.20587957e-01 8.16444382e-02 -8.32460284e-01
9.48302269e-01 4.00137514e-01 -1.81571707e-01 -9.38520133e-01
-7.10778296e-01 -6.88481092e-01 5.04148453e-02 -2.11220667e-01
1.87845692e-01 -4.73112583e-01 -8.83168280e-01 1.31013036e-01
-9.16980624e-01 -2.28306696e-01 -1.53673157e-01 1.51577279e-01
-4.39664185e-01 4.73298460e-01 -5.41934252e-01 -8.48915219e-01
-1.92491904e-01 -6.20252311e-01 5.14462888e-01 2.76581317e-01
1.69839203e-01 -5.80729544e-01 1.72757640e-01 1.40038311e-01
8.39454234e-01 4.92778569e-02 6.21693254e-01 -4.70282316e-01
-9.40282047e-01 -8.12097371e-01 -3.67437422e-01 -3.77402037e-01
2.47015491e-01 -4.69900042e-01 -9.21348035e-02 -5.19694567e-01
1.96187466e-01 -2.32423484e-01 1.25524655e-01 1.19711384e-01
9.39985871e-01 -7.50836551e-01 -3.51916432e-01 4.34874883e-03
1.46265292e+00 1.35005996e-01 -7.72190616e-02 5.17340600e-01
2.66489014e-02 4.56661850e-01 8.56263638e-01 1.22798920e+00
1.30713418e-01 4.57622945e-01 6.55231655e-01 5.84288955e-01
5.78637123e-01 -2.39237815e-01 2.07438782e-01 8.03694248e-01
-1.09929860e-01 -4.24666464e-01 -4.47751641e-01 8.24429810e-01
-1.95087767e+00 -6.84335530e-01 2.41614297e-01 2.49081254e+00
8.51640582e-01 -2.52478905e-02 4.41621780e-01 -9.38712731e-02
7.67181337e-01 7.03411847e-02 -8.87821913e-01 9.05630961e-02
1.98862448e-01 -3.11630100e-01 7.68882990e-01 2.95019597e-01
-4.52739686e-01 7.64507174e-01 6.96650267e+00 7.31027305e-01
-7.71082222e-01 1.97537556e-01 3.89233530e-01 -4.51479316e-01
-1.81764022e-01 3.71956266e-02 -9.16984677e-01 7.30641782e-01
7.63821363e-01 -9.80413198e-01 1.10450637e+00 1.24288201e+00
6.29182637e-01 -3.65742892e-01 -6.01858497e-01 8.17073464e-01
-3.72195423e-01 -9.84111786e-01 -2.28837311e-01 2.97047079e-01
8.47916484e-01 1.53628528e-01 -1.30354881e-01 -1.30037397e-01
9.26232576e-01 -2.99108237e-01 -3.28652449e-02 -1.10954121e-01
4.15922076e-01 -5.94256163e-01 2.51272470e-01 6.81116700e-01
-1.04066396e+00 -2.20556512e-01 -4.84334916e-01 -1.15109757e-01
6.56522989e-01 7.65942574e-01 -4.26866472e-01 1.75509512e-01
9.12917435e-01 5.14060140e-01 -1.22892587e-02 1.38592589e+00
-1.32946432e-01 5.49162984e-01 -8.39442968e-01 -7.47951031e-01
3.54978055e-01 -4.12109673e-01 6.46934867e-01 5.18461704e-01
6.26860619e-01 3.96605015e-01 -1.86027642e-02 3.62095147e-01
-2.32540965e-01 2.67467529e-01 -6.42965972e-01 -2.96890616e-01
6.99536264e-01 1.54948866e+00 -9.27701950e-01 -1.77862078e-01
-4.33152944e-01 7.76184261e-01 3.50398511e-01 5.76707125e-01
-6.04944348e-01 -5.70418060e-01 3.75078857e-01 6.00686073e-01
1.46910429e-01 -3.40423435e-01 1.76991478e-01 -1.14017522e+00
3.03216398e-01 -7.44455040e-01 2.95671612e-01 -4.97533947e-01
-1.16714132e+00 4.34425831e-01 1.60727240e-02 -9.55938876e-01
-3.63016784e-01 -3.00830960e-01 -7.84998655e-01 2.30397642e-01
-1.48539019e+00 -2.52196938e-01 -2.86510408e-01 8.80298793e-01
4.67572600e-01 -8.43370780e-02 3.02813053e-01 1.29854873e-01
-5.76203346e-01 2.68495500e-01 8.84985507e-01 -3.01794767e-01
4.26851749e-01 -1.22921169e+00 -5.24271727e-02 5.71217895e-01
-2.40996793e-01 4.68289495e-01 9.23147261e-01 -6.43275559e-01
-1.99260473e+00 -6.76329732e-01 9.08217549e-01 -1.49351612e-01
1.25445461e+00 -3.47117260e-02 -4.39044446e-01 2.89636940e-01
2.93647528e-01 -1.04561277e-01 3.58105540e-01 -2.76187100e-02
1.79705128e-01 -1.18324846e-01 -1.18717194e+00 7.57169366e-01
8.51954937e-01 -7.69958869e-02 -4.35272977e-02 1.05362070e+00
5.36180854e-01 1.93945751e-01 -7.42313445e-01 -2.30679914e-01
4.46438015e-01 -4.47351635e-01 5.99217534e-01 -5.43895006e-01
1.74669847e-01 1.67005405e-01 1.42214939e-01 -1.32339883e+00
-1.77413747e-01 -1.39036584e+00 -3.06782693e-01 8.70168269e-01
3.77511978e-01 -7.41557240e-01 9.02867675e-01 7.71984994e-01
6.36383653e-01 -6.98209763e-01 -8.60825360e-01 -9.25522208e-01
-1.85839772e-01 2.56408513e-01 4.04014081e-01 7.73205936e-01
5.90883613e-01 2.72958428e-01 -9.49307621e-01 -1.35367781e-01
1.01135004e+00 2.12703213e-01 6.85640872e-01 -1.19706714e+00
-3.10248673e-01 -9.01995823e-02 6.23148382e-01 -1.48026967e+00
-4.03333962e-01 -6.23201072e-01 -5.00216819e-02 -1.34412968e+00
1.86022818e-01 -6.41203105e-01 -3.63068670e-01 2.54484236e-01
2.42936134e-01 -3.27406168e-01 1.01345003e-01 5.27947307e-01
-1.20609331e+00 2.53215760e-01 1.29279745e+00 8.65514278e-02
-2.92826623e-01 4.06443566e-01 -8.33291769e-01 4.46000516e-01
1.05874014e+00 -8.20278406e-01 -5.23334682e-01 -1.79188743e-01
8.69958222e-01 7.39616156e-01 -1.49203420e-01 -3.12824428e-01
3.38305593e-01 -5.35396338e-01 -4.81757969e-01 -6.16365910e-01
8.02907273e-02 -1.05440915e+00 1.44719049e-01 3.71879071e-01
-4.26586270e-01 1.00854166e-01 -5.80877006e-01 1.06991506e+00
-1.49197550e-02 -6.35962844e-01 4.02594537e-01 -3.40977132e-01
-9.18097571e-02 5.03045142e-01 -7.20539749e-01 3.90577555e-01
1.10499275e+00 1.38427541e-01 -3.06380898e-01 -9.00156558e-01
-6.07724428e-01 7.05887079e-01 2.36715138e-01 5.77428937e-01
2.34425545e-01 -1.15565979e+00 -3.72148752e-01 -1.52917922e-01
-4.10642356e-01 -2.67483860e-01 -2.71555573e-01 6.95946276e-01
-3.07807714e-01 4.81878430e-01 -3.67633998e-02 -2.22801249e-02
-8.01674247e-01 5.64305186e-01 -3.35026860e-01 -6.56240284e-01
-5.26649833e-01 2.03861400e-01 -3.97161156e-01 4.37901706e-01
3.96534860e-01 1.97188690e-01 -2.37742886e-01 2.45577753e-01
6.08124197e-01 8.07264268e-01 -2.34700292e-01 2.14210358e-02
1.43763870e-01 4.40140337e-01 -4.55145359e-01 -2.26040147e-02
1.49816012e+00 -5.43604016e-01 -2.74040371e-01 -7.55156502e-02
1.19815171e+00 -2.84272373e-01 -1.11506212e+00 -3.45084965e-01
7.83853307e-02 -6.96393967e-01 1.14829779e-01 -4.61178035e-01
-1.10944259e+00 4.07490790e-01 1.79116681e-01 1.04348040e+00
8.26996207e-01 -3.80719341e-02 1.17818832e+00 1.04969990e+00
8.71151924e-01 -1.48219168e+00 2.55029231e-01 3.78136903e-01
6.79438710e-01 -1.12736487e+00 -6.30242452e-02 -3.17020833e-01
-2.54554540e-01 8.04630637e-01 2.67019510e-01 -2.17117384e-01
1.19826722e+00 4.59913820e-01 -3.22208226e-01 -1.28459081e-01
-8.24362993e-01 1.37375686e-02 -5.86274028e-01 2.82057524e-01
-3.26206237e-02 -1.00594290e-01 -1.07974195e+00 4.71159190e-01
-4.75531630e-03 3.42651345e-02 7.10201263e-01 1.10072577e+00
-9.30051029e-01 -1.14089906e+00 -1.02782585e-01 9.45541263e-01
-9.12597001e-01 -4.45208065e-02 -1.13584101e-01 4.03464973e-01
-7.63562799e-01 1.09278572e+00 -3.79328579e-01 1.71222966e-02
-1.38445199e-01 -3.73125196e-01 -3.23148578e-01 -6.58700913e-02
-8.87501836e-02 1.87103912e-01 -3.47403251e-02 -5.79085469e-01
-2.58813258e-02 -6.88381910e-01 -8.28761101e-01 -3.18368375e-01
-6.73675060e-01 7.98546672e-01 6.89858854e-01 6.59581184e-01
3.91661257e-01 -7.68055916e-02 1.20747149e+00 -4.21275556e-01
-9.90044236e-01 -6.80281758e-01 -1.05292189e+00 3.02546769e-01
1.14127569e-01 -6.24182105e-01 -7.14090645e-01 -2.90547729e-01] | [4.642611026763916, 3.4131360054016113] |
bb48a2a8-c831-473d-9deb-add9eede3d56 | earthnets-empowering-ai-in-earth-observation | 2210.04936 | null | https://arxiv.org/abs/2210.04936v2 | https://arxiv.org/pdf/2210.04936v2.pdf | EarthNets: Empowering AI in Earth Observation | Earth observation, aiming at monitoring the state of planet Earth using remote sensing data, is critical for improving our daily lives and living environment. With a growing number of satellites in orbit, an increasing number of datasets with diverse sensors and research domains are being published to facilitate the research of the remote sensing community. In this paper, we present a comprehensive review of more than 400 publicly published datasets, including applications like land use/cover, change/disaster monitoring, scene understanding, agriculture, climate change, and weather forecasting. We systematically analyze these Earth observation datasets with respect to five aspects volume, bibliometric analysis, resolution distributions, research domains, and the correlation between datasets. Based on the dataset attributes, we propose to measure, rank, and select datasets to build a new benchmark for model evaluation. Furthermore, a new platform for Earth observation, termed EarthNets, is released as a means of achieving a fair and consistent evaluation of deep learning methods on remote sensing data. EarthNets supports standard dataset libraries and cutting-edge deep learning models to bridge the gap between the remote sensing and machine learning communities. Based on this platform, extensive deep learning methods are evaluated on the new benchmark. The insightful results are beneficial to future research. The platform and dataset collections are publicly available at https://earthnets.github.io/. | ['Xiao Xiang Zhu', 'Yilei Shi', 'Yi Wang', 'Fahong Zhang', 'Zhitong Xiong'] | 2022-10-10 | null | null | null | null | ['weather-forecasting'] | ['miscellaneous'] | [-2.42200315e-01 -6.34766281e-01 -6.12753808e-01 -4.23842251e-01
-2.29883984e-01 -4.49459344e-01 6.49719238e-01 1.19052701e-01
-1.30581513e-01 6.67901337e-01 3.15806448e-01 -5.13412714e-01
-2.64241725e-01 -1.45050621e+00 -3.81468743e-01 -7.43066728e-01
-3.06179166e-01 1.08830355e-01 -1.93826109e-01 -3.36571127e-01
1.05856270e-01 6.24333501e-01 -1.72601700e+00 1.91071518e-02
8.79680693e-01 1.18483222e+00 4.73080903e-01 2.95526862e-01
-1.42283782e-01 3.83442879e-01 -1.64657205e-01 2.32734144e-01
3.74810576e-01 -6.06075116e-02 -6.11215115e-01 -3.48650247e-01
3.12426656e-01 -1.88306004e-01 -3.73059511e-01 1.24765861e+00
7.37432122e-01 -3.71952578e-02 3.63861382e-01 -1.02600420e+00
-1.03016555e+00 6.28405094e-01 -5.76860428e-01 4.40818310e-01
-2.74213534e-02 1.06070377e-03 1.19361722e+00 -7.80145586e-01
3.82957101e-01 9.13636029e-01 6.29189074e-01 3.80105264e-02
-4.73079652e-01 -1.01538134e+00 -2.83425255e-03 3.86445671e-01
-1.48305273e+00 -4.15450275e-01 6.08251989e-01 -6.81190491e-01
6.95569754e-01 4.39931244e-01 7.05440223e-01 7.36157119e-01
4.36890572e-01 3.64145815e-01 1.02755499e+00 -1.18512668e-01
6.64340854e-02 -1.43544659e-01 1.21725023e-01 4.79755580e-01
5.41586936e-01 2.43256807e-01 -2.89018035e-01 1.37736872e-01
6.44116104e-01 5.99847913e-01 -2.92025357e-01 -3.10388254e-03
-1.30719757e+00 7.19762206e-01 9.62200820e-01 4.70923990e-01
-7.59333253e-01 -5.33295190e-03 2.17605457e-01 4.93158072e-01
9.60695267e-01 4.30460066e-01 -6.47928059e-01 2.42432356e-01
-8.86980057e-01 2.12699577e-01 6.97325408e-01 6.87069356e-01
9.04290617e-01 1.55444562e-01 1.81525499e-01 8.83040905e-01
4.52454865e-01 8.91806185e-01 3.46669823e-01 -5.98217189e-01
1.64290592e-01 9.80212510e-01 9.88239422e-02 -1.34880495e+00
-6.32099867e-01 -6.26161695e-01 -1.32503498e+00 -5.69089465e-02
-3.25457036e-01 -2.80888557e-01 -8.50494742e-01 1.20078456e+00
4.00692463e-01 2.01631054e-01 -7.62423053e-02 9.35922086e-01
1.53983164e+00 9.25500333e-01 -2.82080844e-02 1.18076883e-01
1.33616030e+00 -7.20188975e-01 -7.25154757e-01 -1.20376647e-01
6.06476486e-01 -5.75743139e-01 1.05416739e+00 8.26400220e-02
-3.50143105e-01 -4.07993495e-01 -9.03289676e-01 1.05388515e-01
-9.94182527e-01 4.34208781e-01 8.77862871e-01 3.26384366e-01
-9.36962605e-01 5.26670337e-01 -7.36445665e-01 -6.65465653e-01
5.61528027e-01 -1.92483112e-01 -1.47109866e-01 -2.97333170e-02
-1.58762836e+00 7.47725189e-01 4.34297919e-01 2.94229358e-01
-1.00874162e+00 -7.87602782e-01 -6.17650270e-01 1.48830324e-01
-6.47441447e-02 -4.82502788e-01 6.66140318e-01 -3.49906534e-01
-1.11625540e+00 1.13001537e+00 3.84718031e-01 -2.10601777e-01
2.46718392e-01 -2.14216322e-01 -9.53864753e-01 -4.19552147e-01
3.30230445e-01 3.22573960e-01 8.58845338e-02 -7.47856736e-01
-7.87537277e-01 -5.53388596e-01 -1.82496205e-01 1.27226785e-01
-6.12613559e-01 2.10703418e-01 -3.04085165e-01 -4.95574445e-01
2.10102677e-01 -8.73577774e-01 -3.27984869e-01 7.18045048e-03
-3.06752205e-01 -3.09050262e-01 7.13954389e-01 -5.88025928e-01
1.25374162e+00 -2.17988920e+00 -1.42953366e-01 4.34340024e-03
2.93673545e-01 2.14083940e-01 -2.53202915e-01 4.85484749e-01
-8.99950564e-02 3.44266653e-01 -2.07668707e-01 2.12133944e-01
-2.82153964e-01 4.73708287e-02 -5.62531233e-01 7.68200874e-01
-1.97322905e-01 8.25185299e-01 -8.55678618e-01 -2.04481304e-01
3.01616669e-01 4.23140705e-01 3.59469801e-02 6.75652772e-02
-2.94691324e-01 6.34970546e-01 -7.89047003e-01 1.08967543e+00
8.15112889e-01 -4.98518765e-01 -3.65237612e-03 7.35494718e-02
-5.75157404e-01 4.02964175e-01 -1.14102638e+00 1.44745135e+00
-5.05288422e-01 7.91676641e-01 -7.16632092e-03 -8.34164858e-01
1.44784045e+00 5.03450707e-02 5.93655527e-01 -8.54216874e-01
1.97694432e-02 3.68870467e-01 -2.48276383e-01 -5.22647798e-01
4.88720566e-01 2.49186873e-01 6.46046996e-02 3.01120639e-01
-3.78169000e-01 1.10847682e-01 1.72428600e-02 -2.82499582e-01
6.57023787e-01 -1.69834107e-01 4.35586721e-01 -6.11140072e-01
2.29499996e-01 2.52847344e-01 5.70511460e-01 4.37478125e-01
-1.91504434e-01 3.15461695e-01 1.74920373e-02 -1.22274232e+00
-8.62234652e-01 -6.07856631e-01 -7.93953180e-01 1.34754288e+00
1.54714271e-01 -7.78709278e-02 -2.18826681e-02 3.04281358e-02
2.96215564e-01 3.57345194e-01 -6.07074976e-01 9.19564664e-02
-9.20245126e-02 -1.28413761e+00 6.98621333e-01 4.24458861e-01
1.09463298e+00 -1.19941139e+00 -4.06017482e-01 -4.22096662e-02
-4.32363719e-01 -7.62297928e-01 2.72519588e-01 -1.74165130e-01
-9.61957812e-01 -9.35234427e-01 -7.03645885e-01 -4.66911227e-01
-4.06880453e-02 6.28872335e-01 1.31416392e+00 2.14287024e-02
-1.68114483e-01 -1.88476443e-01 -3.60131860e-01 -7.83278525e-01
6.18637502e-02 5.00625074e-01 1.83695629e-01 -3.03191572e-01
6.04398727e-01 -6.09560728e-01 -6.98904037e-01 3.22939247e-01
-9.43701923e-01 -3.28446656e-01 6.56761467e-01 2.57832438e-01
6.34133339e-01 2.29278430e-01 5.42098641e-01 -4.82472956e-01
3.69677484e-01 -1.24979782e+00 -8.64859164e-01 1.72736973e-01
-8.32825541e-01 -3.63502264e-01 1.45107687e-01 5.26241474e-02
-7.89391875e-01 -1.94787353e-01 -4.82472368e-02 -8.25633258e-02
-3.28162998e-01 1.26275158e+00 -6.29850850e-02 -3.26191657e-03
9.10503507e-01 2.44689435e-01 -4.11146522e-01 -9.06749904e-01
2.17993110e-01 1.19070685e+00 3.55149895e-01 -1.22802280e-01
4.70728070e-01 5.80529869e-01 -1.58333540e-01 -1.07801688e+00
-1.04988599e+00 -8.01762342e-01 -4.29263562e-01 -1.49562106e-01
6.33897007e-01 -1.54429746e+00 -2.70883739e-01 6.72545731e-01
-9.56277549e-01 -2.49269605e-01 1.64173782e-01 7.03890145e-01
1.57649845e-01 -6.36553913e-02 -1.58714786e-01 -4.77736294e-01
-7.01040566e-01 -7.44990170e-01 8.98642957e-01 2.91381031e-01
2.85494894e-01 -1.10054946e+00 5.84256709e-01 6.11892827e-02
8.41257095e-01 4.47148472e-01 4.04779106e-01 -4.47079718e-01
-6.03259325e-01 -2.40227222e-01 -5.39535403e-01 2.71052957e-01
4.69912678e-01 2.75665849e-01 -9.55867469e-01 -2.55600214e-01
-2.72018820e-01 -1.36116102e-01 1.44079220e+00 6.62139893e-01
1.32310557e+00 -1.83484837e-01 -6.40919328e-01 1.15845072e+00
1.50168800e+00 -9.86242145e-02 7.50092626e-01 7.99896836e-01
8.58014226e-01 4.24092740e-01 5.41584492e-01 6.40426397e-01
4.96127903e-01 2.40795135e-01 8.40745270e-01 -3.37395012e-01
1.99083865e-01 1.87761620e-01 9.22261551e-02 7.44090438e-01
-4.63700145e-01 -1.50069222e-01 -1.61832869e+00 8.54806900e-01
-1.46937001e+00 -1.01962399e+00 -3.88771683e-01 2.02925014e+00
6.72443807e-01 -5.34659922e-01 -4.55864631e-02 -2.10022241e-01
7.55509973e-01 8.28697324e-01 -7.33337760e-01 2.92714238e-01
-3.05473894e-01 -1.35199711e-01 1.03821588e+00 9.42383613e-03
-1.47707129e+00 1.06732464e+00 6.35208416e+00 4.32053238e-01
-1.72309494e+00 9.59154144e-02 3.29264164e-01 1.36066005e-01
-4.26760852e-01 -2.15703428e-01 -1.04948080e+00 4.92851436e-01
9.15166795e-01 -3.57768834e-01 2.00554013e-01 1.05209243e+00
3.48304629e-01 2.73697436e-01 -5.25413692e-01 8.02738905e-01
-3.62236023e-01 -1.76784730e+00 9.87823531e-02 1.86379567e-01
9.24884379e-01 1.31413448e+00 1.48540705e-01 1.51422247e-01
5.89696825e-01 -1.07714629e+00 1.82419300e-01 6.48346603e-01
1.05595350e+00 -4.51129258e-01 8.17075253e-01 2.05089256e-01
-1.41242778e+00 -2.10861549e-01 -8.11556101e-01 -3.41187656e-01
-4.10229951e-01 1.24453485e+00 -2.14521274e-01 8.63948047e-01
1.44414330e+00 1.51060092e+00 -3.51611823e-01 1.18128276e+00
-1.67781875e-01 7.61204720e-01 -2.98038423e-01 5.86576462e-02
1.88496426e-01 -3.38927329e-01 5.29732168e-01 9.98090148e-01
5.14673293e-01 9.01427940e-02 2.21623868e-01 8.29074323e-01
-3.78778875e-01 2.47812927e-01 -9.25843894e-01 -2.63354927e-01
7.83127367e-01 1.44015050e+00 -2.99330920e-01 -1.82696640e-01
-4.24933940e-01 1.45726845e-01 -7.48623610e-02 2.90532261e-01
-6.41090453e-01 -4.85734403e-01 1.11471653e+00 -2.37980811e-03
-1.02239959e-01 -4.36792850e-01 -4.93969291e-01 -1.28138435e+00
-1.07319750e-01 -6.35182500e-01 4.71500278e-01 -7.27928162e-01
-1.32472277e+00 5.05366087e-01 -1.35642797e-01 -1.54847312e+00
4.03058022e-01 -4.91824716e-01 -8.59425247e-01 9.73739862e-01
-2.35130739e+00 -9.54152763e-01 -9.79893148e-01 3.42749625e-01
1.14207976e-01 -3.61741453e-01 8.48329246e-01 6.09693229e-01
-7.15699375e-01 -6.14846088e-02 7.55620837e-01 1.98497832e-01
7.37604856e-01 -9.44259346e-01 7.61658370e-01 5.91042578e-01
-5.37173934e-02 3.75808388e-01 3.32148343e-01 -5.81009209e-01
-1.26993847e+00 -1.75862134e+00 8.94572377e-01 -1.85078919e-01
9.42174733e-01 1.76438943e-01 -1.03185880e+00 8.25329125e-01
-5.11888117e-02 1.73894033e-01 7.46924698e-01 3.35378736e-01
-2.08137602e-01 -5.44083655e-01 -8.78084540e-01 1.29812613e-01
8.34295273e-01 -3.68128866e-01 -2.69827843e-01 7.04902470e-01
7.04126477e-01 -1.28728241e-01 -1.13666499e+00 6.56182587e-01
5.51946640e-01 -5.91643453e-01 7.92211711e-01 -7.22771704e-01
8.05610418e-01 -2.34543115e-01 -3.66126448e-01 -1.38220632e+00
-7.69720793e-01 8.47733095e-02 2.39576817e-01 1.00918210e+00
2.94388175e-01 -8.31740677e-01 4.42793608e-01 -1.00107349e-01
-2.50332981e-01 -4.97254997e-01 -5.60466111e-01 -7.68941641e-01
3.92526180e-01 -1.64579019e-01 1.18494105e+00 1.60241878e+00
-4.98264641e-01 1.98891476e-01 -4.20282662e-01 5.26432693e-01
3.85136127e-01 5.75900078e-01 7.89351821e-01 -1.92575395e+00
4.28569168e-01 -5.35072565e-01 -2.90712744e-01 -8.43608201e-01
-1.79650992e-01 -9.49557185e-01 -2.26126179e-01 -1.94282186e+00
8.17518383e-02 -6.82577550e-01 -6.07943773e-01 8.27808321e-01
-1.74367800e-01 2.22416297e-01 -4.54964668e-01 8.18295419e-01
-3.23564738e-01 7.10601151e-01 1.10986078e+00 -2.93895811e-01
-2.74699211e-01 3.07974108e-02 -6.30495310e-01 7.02441335e-01
1.15506089e+00 -4.58720833e-01 1.68105513e-01 -1.01798439e+00
7.55170882e-01 -3.07164431e-01 3.67124707e-01 -1.07789159e+00
6.39569834e-02 -6.13788188e-01 2.37796009e-01 -8.31336260e-01
-3.59643102e-01 -6.23647511e-01 2.64958888e-01 3.49246353e-01
-1.96890816e-01 -9.24078003e-03 1.99027225e-01 3.08547974e-01
-4.51347739e-01 3.83766711e-01 7.88926482e-01 -2.07545236e-01
-1.16954720e+00 8.09261799e-01 -8.78961310e-02 -2.66446084e-01
7.95412898e-01 2.30912447e-01 -5.82145870e-01 -4.49436978e-02
-4.40684557e-01 5.81464410e-01 4.69909012e-01 6.21336520e-01
2.87330955e-01 -1.40742660e+00 -1.30860949e+00 2.12338373e-01
5.05349100e-01 9.74668413e-02 3.40057343e-01 6.75119102e-01
-7.36911237e-01 5.45845807e-01 -4.93433416e-01 -5.51716387e-01
-8.34851265e-01 1.06668301e-01 5.81167758e-01 -1.14937581e-01
-5.79956591e-01 5.64685464e-01 -5.91865182e-02 -1.06763625e+00
-1.90345123e-02 -3.80277544e-01 -6.62129164e-01 2.33176708e-01
6.70826912e-01 5.50007224e-01 2.21545145e-01 -4.65565681e-01
-5.17032564e-01 5.33074677e-01 3.65299523e-01 4.16391522e-01
1.92904639e+00 -2.27011517e-01 -4.79976267e-01 7.31389582e-01
9.06317294e-01 -2.74392933e-01 -6.99261010e-01 -5.18607497e-01
-7.48357624e-02 -4.45492536e-01 6.35194123e-01 -7.81179190e-01
-1.50859547e+00 9.06119525e-01 8.20845604e-01 4.96571451e-01
1.07671702e+00 4.81842272e-02 6.82596684e-01 6.84303820e-01
4.43873033e-02 -8.42894912e-01 -3.75236720e-01 6.96090817e-01
1.04369521e+00 -1.60366559e+00 3.24113935e-01 -3.14271972e-02
-1.77987203e-01 8.56474221e-01 3.57082307e-01 -3.08676790e-02
1.17462790e+00 -2.52998881e-02 3.23268980e-01 -5.29306352e-01
-4.86527443e-01 -3.84147644e-01 2.22659945e-01 3.67739677e-01
7.16387749e-01 5.69224894e-01 -1.76938727e-01 3.07558954e-01
-1.60454452e-01 2.29091898e-01 3.69047999e-01 6.06810331e-01
-8.53465140e-01 -5.69582820e-01 -4.04316247e-01 7.24251807e-01
-4.86971557e-01 -4.28260952e-01 -3.37031245e-01 5.22881329e-01
-9.30869505e-02 7.81846702e-01 3.00140887e-01 -3.63267809e-01
1.71388298e-01 -3.56493890e-01 -3.76201063e-01 -5.83733499e-01
-7.65570626e-02 -4.08204883e-01 -1.29190952e-01 -4.81169909e-01
-7.10029125e-01 -3.58501047e-01 -8.99619579e-01 -7.18841672e-01
-2.52706647e-01 2.37703528e-02 1.17574584e+00 8.10710549e-01
8.02813530e-01 4.47762460e-01 8.60578179e-01 -8.14081550e-01
-3.84039968e-01 -1.32405519e+00 -9.56823409e-01 -1.38571382e-01
2.09798500e-01 -6.52104914e-01 -3.55505824e-01 -2.48725146e-01] | [9.673970222473145, -1.4104920625686646] |
99808287-2eb1-40e9-843d-76a6bb26d885 | acl-spc-adaptive-closed-loop-system-for-self | 2303.01979 | null | https://arxiv.org/abs/2303.01979v3 | https://arxiv.org/pdf/2303.01979v3.pdf | ACL-SPC: Adaptive Closed-Loop system for Self-Supervised Point Cloud Completion | Point cloud completion addresses filling in the missing parts of a partial point cloud obtained from depth sensors and generating a complete point cloud. Although there has been steep progress in the supervised methods on the synthetic point cloud completion task, it is hardly applicable in real-world scenarios due to the domain gap between the synthetic and real-world datasets or the requirement of prior information. To overcome these limitations, we propose a novel self-supervised framework ACL-SPC for point cloud completion to train and test on the same data. ACL-SPC takes a single partial input and attempts to output the complete point cloud using an adaptive closed-loop (ACL) system that enforces the output same for the variation of an input. We evaluate our proposed ACL-SPC on various datasets to prove that it can successfully learn to complete a partial point cloud as the first self-supervised scheme. Results show that our method is comparable with unsupervised methods and achieves superior performance on the real-world dataset compared to the supervised methods trained on the synthetic dataset. Extensive experiments justify the necessity of self-supervised learning and the effectiveness of our proposed method for the real-world point cloud completion task. The code is publicly available from https://github.com/Sangminhong/ACL-SPC_PyTorch | ['Kyoung Mu Lee', 'Reyhaneh Neshatavar', 'Mohsen Yavartanoo', 'Sangmin Hong'] | 2023-03-02 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Hong_ACL-SPC_Adaptive_Closed-Loop_System_for_Self-Supervised_Point_Cloud_Completion_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Hong_ACL-SPC_Adaptive_Closed-Loop_System_for_Self-Supervised_Point_Cloud_Completion_CVPR_2023_paper.pdf | cvpr-2023-1 | ['point-cloud-completion'] | ['computer-vision'] | [ 1.86214164e-01 5.46442270e-02 7.35184327e-02 -2.78905362e-01
-1.03827834e+00 -5.68678260e-01 4.34513658e-01 7.08009079e-02
-2.56405205e-01 6.55265987e-01 -5.35929978e-01 -2.00471625e-01
-4.10322957e-02 -7.09596813e-01 -1.13387036e+00 -5.65555573e-01
6.57754317e-02 9.66750205e-01 3.96529227e-01 1.62268523e-02
4.35991555e-01 7.92546809e-01 -1.75023067e+00 -1.24625620e-02
1.04400933e+00 6.24769568e-01 7.32556045e-01 5.89538217e-01
-8.67320821e-02 1.50357023e-01 -3.21449339e-01 2.00547855e-02
6.00952625e-01 2.30058189e-02 -6.30369484e-01 4.24711406e-01
4.57004398e-01 -1.29627079e-01 2.39976663e-02 9.36836362e-01
3.66034865e-01 -2.34667212e-01 6.05940521e-01 -1.62325978e+00
-8.61414708e-03 -1.00892998e-01 -5.15395701e-01 -5.32606483e-01
4.90032971e-01 1.28072038e-01 5.75551569e-01 -1.23859978e+00
6.80216670e-01 8.30986738e-01 7.50714839e-01 4.81761903e-01
-1.22325563e+00 -7.24151254e-01 -2.50872940e-01 -2.01367393e-01
-1.52070260e+00 -2.63727993e-01 9.32025015e-01 -5.41004896e-01
5.61727464e-01 -5.57598285e-03 6.68330193e-01 7.04010248e-01
-3.41805406e-02 7.23577857e-01 1.05938363e+00 -4.55043375e-01
6.31569326e-01 1.19405515e-01 -2.26691693e-01 6.62902534e-01
2.60186136e-01 2.51555353e-01 -3.36144924e-01 -3.20760339e-01
9.68279183e-01 1.33455351e-01 -1.14050262e-01 -8.74059439e-01
-1.39586127e+00 5.98495960e-01 4.34408307e-01 -6.07578233e-02
-3.02090883e-01 -9.60011221e-03 1.80845737e-01 3.11932176e-01
3.80383492e-01 2.78694957e-01 -5.31548798e-01 -2.24068481e-02
-1.11327279e+00 5.17601788e-01 7.10223794e-01 1.56003761e+00
1.16028392e+00 -1.46790788e-01 6.06779873e-01 5.71450531e-01
2.81989872e-01 7.82331586e-01 1.35057226e-01 -1.08547485e+00
6.88622952e-01 7.91355550e-01 3.00417900e-01 -6.64670646e-01
-1.95873782e-01 -1.98667705e-01 -7.11530924e-01 7.97601819e-01
2.93821007e-01 -1.40483052e-01 -9.27276969e-01 1.25975633e+00
5.10133564e-01 4.92609859e-01 1.92109972e-01 7.48765171e-01
6.09062910e-01 6.58353031e-01 -3.27696532e-01 -8.36845711e-02
7.05956340e-01 -6.75523341e-01 -2.76634872e-01 -2.56435066e-01
5.39356112e-01 -8.10061872e-01 1.19024706e+00 5.23056507e-01
-8.89270604e-01 -7.77183890e-01 -1.16472721e+00 1.66210815e-01
-2.29395777e-01 3.52209181e-01 4.59975958e-01 1.73834234e-01
-9.01560664e-01 6.45913839e-01 -1.03288126e+00 -3.62288147e-01
4.38460201e-01 4.71169859e-01 -5.72754025e-01 -2.50043213e-01
-5.57620585e-01 4.47092772e-01 4.54696178e-01 7.42866052e-03
-8.69473636e-01 -8.08060944e-01 -8.41665626e-01 -3.82979929e-01
2.63045400e-01 -6.48194313e-01 1.18153632e+00 -6.98129654e-01
-1.36864674e+00 8.25840831e-01 -1.56733245e-01 -1.21581331e-01
7.20184565e-01 -2.49411434e-01 6.43297806e-02 1.35157496e-01
5.15201509e-01 1.07849693e+00 8.69873405e-01 -1.71961725e+00
-7.46892452e-01 -3.82573009e-01 -2.30765253e-01 2.79293180e-01
2.40023240e-01 -5.58514655e-01 -7.36290753e-01 -3.85868102e-01
6.23637915e-01 -1.25774360e+00 -3.39812845e-01 3.42950314e-01
-2.74223030e-01 -1.39731586e-01 1.16044855e+00 -5.56181632e-02
4.34775442e-01 -2.09374881e+00 -5.68461232e-02 2.03627422e-01
-1.33853853e-01 2.08986312e-01 -1.34009123e-01 8.63075435e-01
-2.03294337e-01 -2.41150439e-01 -8.12362194e-01 -8.43411982e-01
-3.19694847e-01 4.82301384e-01 -5.02754509e-01 6.03925765e-01
1.99806169e-01 5.00737965e-01 -1.03739142e+00 -6.25621796e-01
5.51912367e-01 3.36152166e-01 -4.66341406e-01 2.45668277e-01
-4.66998965e-01 8.50121319e-01 -4.13050830e-01 7.73700953e-01
1.11763501e+00 -1.50046691e-01 -2.65500784e-01 1.17795244e-01
-1.45495683e-01 -2.92946100e-01 -1.46308565e+00 2.28772116e+00
-3.36667091e-01 3.66415739e-01 -2.96977330e-02 -7.72451997e-01
1.17033803e+00 3.54591906e-01 5.78160524e-01 -2.86969692e-01
-2.01970428e-01 5.23673296e-01 -3.49535048e-01 -4.88345593e-01
3.42692643e-01 -1.75370201e-01 5.05842604e-02 3.47464085e-01
2.28950717e-02 -9.73991275e-01 -2.04929039e-02 3.04781586e-01
8.92459691e-01 6.58213019e-01 1.01682618e-01 -2.29910463e-02
7.28864491e-01 4.60454881e-01 5.28364837e-01 5.07754147e-01
-4.38080356e-02 1.20151412e+00 1.93135828e-01 -3.65388989e-01
-1.38667870e+00 -1.12235010e+00 -2.71319628e-01 2.50137269e-01
4.83482867e-01 -3.22056562e-01 -6.21821582e-01 -6.42270565e-01
1.96247131e-01 5.64049900e-01 -2.00455442e-01 2.55695999e-01
-3.71094555e-01 -1.34407744e-01 3.02237958e-01 3.57640207e-01
4.96904552e-01 -1.11238670e+00 -5.81023693e-01 -6.73291981e-02
-7.55019188e-02 -1.44678712e+00 -1.30064739e-02 -2.31603328e-02
-1.29954123e+00 -1.24348116e+00 -5.89318871e-01 -9.41309690e-01
1.06414700e+00 2.90128559e-01 9.25975800e-01 2.42170349e-01
-1.18418392e-02 4.83498365e-01 -3.31062019e-01 -5.25627375e-01
-3.61056387e-01 1.17335275e-01 1.58983096e-01 -8.30026716e-02
1.72896504e-01 -7.50908077e-01 -4.74128962e-01 4.34532732e-01
-9.78130758e-01 2.17398718e-01 5.08746207e-01 5.28695941e-01
9.48214173e-01 1.17109992e-01 4.13298637e-01 -8.70256126e-01
2.71621644e-01 -3.83904457e-01 -1.00265205e+00 -1.24591097e-01
-4.77158397e-01 -3.27804834e-02 6.41721666e-01 -1.46829084e-01
-7.41202116e-01 8.63562882e-01 -2.05067307e-01 -8.44944119e-01
-3.46194029e-01 2.12033123e-01 -6.86277896e-02 -2.48681456e-01
6.74544632e-01 3.22429746e-01 2.37482265e-01 -4.44150448e-01
7.83427060e-02 6.85331881e-01 7.46081829e-01 -7.61672139e-01
1.29957521e+00 8.57562900e-01 2.21665993e-01 -8.49502325e-01
-5.60991466e-01 -8.11326146e-01 -1.10058987e+00 -2.45720848e-01
5.53953409e-01 -1.06765175e+00 -6.61090016e-01 4.42410171e-01
-1.35850811e+00 -2.35203400e-01 -3.83501172e-01 4.21985388e-01
-9.91450667e-01 5.61043978e-01 -7.43863061e-02 -7.45167136e-01
-2.14806736e-01 -1.15231538e+00 1.54129708e+00 -3.63843441e-02
1.54025525e-01 -8.58191133e-01 3.38218778e-01 1.97748974e-01
-1.39163762e-01 5.90285897e-01 3.55477601e-01 -2.70824581e-01
-8.94896924e-01 -4.54164326e-01 3.53295319e-02 3.53389233e-01
2.41015226e-01 1.06335767e-01 -8.36011827e-01 -4.99529719e-01
1.63026899e-01 -4.79071438e-01 3.75352830e-01 -5.79309240e-02
1.07751894e+00 1.27007052e-01 -3.69395971e-01 6.64068341e-01
1.88038182e+00 -5.43083102e-02 6.34108841e-01 2.45211646e-01
6.23359621e-01 4.35884714e-01 1.02567315e+00 4.53699797e-01
3.22548598e-01 4.71171558e-01 7.44084001e-01 -1.55765936e-01
1.35613382e-01 -6.02400482e-01 1.13214642e-01 7.37344086e-01
5.11892103e-02 1.12026237e-01 -1.35568428e+00 7.04843283e-01
-1.94977045e+00 -5.26424706e-01 -3.18439215e-01 2.44340730e+00
5.62487125e-01 2.00808480e-01 -1.77627504e-01 4.08042341e-01
6.02068067e-01 -1.50134757e-01 -6.26982391e-01 1.12450190e-01
9.02486220e-02 2.24707976e-01 5.68162382e-01 7.04228699e-01
-9.76467192e-01 1.06475711e+00 5.74400759e+00 6.39122307e-01
-1.16458929e+00 -6.08807690e-02 -2.76186708e-02 3.00808400e-01
-8.87835547e-02 2.79639751e-01 -6.45522594e-01 3.99771303e-01
4.56796438e-01 1.19213695e-02 2.11768031e-01 1.04029202e+00
3.02416444e-01 -3.27846080e-01 -1.35706317e+00 1.31420064e+00
-1.45820621e-03 -1.40801656e+00 -4.46279831e-02 -6.36881217e-02
8.08920383e-01 3.15943152e-01 -2.91945815e-01 7.33633414e-02
1.39075711e-01 -6.29873574e-01 5.95956206e-01 5.80948353e-01
1.01542878e+00 -5.92960060e-01 6.56356096e-01 1.04325771e+00
-1.14118826e+00 2.33102337e-01 -5.74497342e-01 -1.34179756e-01
3.82317118e-02 5.39536774e-01 -1.24946094e+00 8.40253770e-01
5.07479250e-01 8.29530180e-01 -5.60272634e-01 1.19692969e+00
-3.32552671e-01 4.38991994e-01 -5.80889165e-01 3.33606631e-01
9.87266898e-02 -4.70728397e-01 6.80498958e-01 8.39604974e-01
2.93389529e-01 2.29156353e-02 2.83515751e-01 8.53756011e-01
3.34313273e-01 1.27631472e-02 -9.54297185e-01 2.91334897e-01
4.46943849e-01 1.04939592e+00 -6.84450328e-01 -3.06124121e-01
-2.82485366e-01 9.16524172e-01 3.26835990e-01 3.25139701e-01
-5.72719991e-01 -3.70084107e-01 3.11569452e-01 2.82124102e-01
1.99072033e-01 -5.69880843e-01 -6.19032562e-01 -1.08609354e+00
3.31537038e-01 -6.37564600e-01 1.54932998e-02 -1.04373729e+00
-1.14893222e+00 6.04431272e-01 1.95586845e-01 -2.08501053e+00
-2.01147959e-01 -4.24042851e-01 -6.63630366e-01 7.37684906e-01
-1.51229203e+00 -1.16101778e+00 -7.65979528e-01 8.11200440e-01
7.84648776e-01 -2.60423392e-01 7.93273330e-01 1.11465968e-01
-4.86800335e-02 1.30693942e-01 6.87166154e-02 -5.00972196e-02
7.29906499e-01 -1.14735651e+00 5.76043785e-01 7.88366437e-01
-4.20291759e-02 3.58703762e-01 6.30812347e-01 -9.41928029e-01
-1.51353741e+00 -1.17461789e+00 5.54546177e-01 -5.40006459e-01
2.32529476e-01 -5.84985256e-01 -9.53616440e-01 6.84020162e-01
-1.02340207e-01 3.25393856e-01 1.71116740e-01 -4.68094766e-01
-3.25270966e-02 -2.07582086e-01 -1.37217867e+00 3.53338599e-01
8.82291734e-01 -3.32547665e-01 -6.59362972e-01 6.39536262e-01
7.91243076e-01 -6.90891087e-01 -6.07085705e-01 6.35733306e-01
1.40721947e-01 -9.34739769e-01 8.05897474e-01 -1.33777363e-02
6.05302393e-01 -7.94539750e-01 -1.50375679e-01 -1.21287763e+00
2.16135308e-01 -5.18779159e-01 2.20167592e-01 8.64696145e-01
2.34292537e-01 -7.12020695e-01 1.28065622e+00 2.88142592e-01
-2.06039250e-01 -6.25713766e-01 -8.28614652e-01 -8.45889449e-01
9.27682668e-02 -7.10983515e-01 6.59186542e-01 7.98863888e-01
-2.23700657e-01 5.47466092e-02 -1.33285642e-01 5.92651427e-01
1.10277700e+00 2.40293175e-01 1.44643581e+00 -1.24969232e+00
-5.11797927e-02 3.92990977e-01 -5.37764132e-01 -1.11575627e+00
1.68664157e-01 -7.80246198e-01 1.29401341e-01 -1.75123370e+00
-2.88608283e-01 -1.08426857e+00 4.63331014e-01 4.25289124e-01
1.02062397e-01 1.85852572e-01 2.49603361e-01 6.39331818e-01
-3.89229625e-01 5.97853422e-01 1.28509748e+00 4.12951745e-02
-3.58589500e-01 3.89702797e-01 -8.14144537e-02 7.64193773e-01
7.86222041e-01 -6.57066703e-01 -5.99821389e-01 -4.27949876e-01
1.40640020e-01 3.68679196e-01 3.02718073e-01 -1.45542932e+00
5.68049431e-01 -2.22122282e-01 2.08963215e-01 -1.20874393e+00
4.46782410e-01 -1.18838060e+00 2.62996018e-01 4.16019022e-01
1.86768979e-01 3.32042836e-02 1.74856588e-01 5.18437266e-01
-3.10193360e-01 -3.42629820e-01 6.30236804e-01 -2.28647426e-01
-5.33719420e-01 5.39199710e-01 2.33528376e-01 -2.01397583e-01
1.17289567e+00 -6.00232482e-01 1.05722502e-01 -2.78059006e-01
-6.24531031e-01 4.30640429e-01 9.60476339e-01 3.50201279e-01
9.37040508e-01 -1.22208631e+00 -8.50862682e-01 5.86969495e-01
4.35682297e-01 9.26870346e-01 -2.67579928e-02 3.52254212e-01
-1.07907546e+00 2.06070483e-01 -1.71392336e-01 -1.22618771e+00
-9.37699080e-01 6.18426323e-01 2.11164221e-01 1.00537039e-01
-8.01695406e-01 2.28772804e-01 -1.16135431e-02 -9.59871650e-01
2.92947084e-01 -2.55948603e-01 2.60694057e-01 -5.68071425e-01
1.89522859e-02 1.85100868e-01 2.44862184e-01 -4.97564644e-01
-1.73817441e-01 9.60914075e-01 2.49682799e-01 -2.69312561e-01
1.40910697e+00 1.02468371e-01 -3.32637466e-02 4.73548084e-01
1.14345717e+00 1.70893166e-02 -1.41485739e+00 -2.92959660e-01
-1.45775661e-01 -6.74039841e-01 -3.95897835e-01 -3.71745229e-01
-7.94282734e-01 7.03760862e-01 4.79443729e-01 -2.29632989e-01
9.23861861e-01 -5.49991913e-02 5.90124249e-01 4.77788568e-01
7.90555716e-01 -7.46455252e-01 6.20106421e-02 4.35961783e-01
1.21120405e+00 -1.42658234e+00 1.97738349e-01 -7.55821884e-01
-5.92980921e-01 1.04327667e+00 6.97894514e-01 -5.87504029e-01
7.52773821e-01 3.13392818e-01 1.77544549e-01 -2.83374548e-01
-5.34719348e-01 3.98266409e-03 -1.41531065e-01 6.55505478e-01
-1.62514582e-01 -4.77358773e-02 -1.54598206e-01 -1.50639132e-01
-4.92520571e-01 3.55997860e-01 6.01586878e-01 1.20495188e+00
-2.42550746e-01 -1.24584365e+00 -6.65727794e-01 7.67610967e-02
1.88730374e-01 3.61350030e-01 -2.98449934e-01 9.01131451e-01
2.17350423e-01 6.77662015e-01 2.17236996e-01 -2.35020667e-01
4.37035471e-01 -8.44750330e-02 3.43410879e-01 -9.46870387e-01
-1.70903325e-01 5.26186004e-02 -4.47331548e-01 -4.27912116e-01
-5.10035634e-01 -8.11644852e-01 -1.58079636e+00 9.69761163e-02
-1.27247050e-01 2.36524001e-01 9.64374363e-01 7.24799097e-01
4.34590399e-01 -4.53244932e-02 9.81935680e-01 -1.25311935e+00
-4.21482623e-01 -7.97158599e-01 -2.94532597e-01 5.98475456e-01
5.06317735e-01 -6.68549955e-01 -5.02485752e-01 1.87574074e-01] | [8.184493064880371, -3.201247215270996] |
a735781f-e201-4983-8df3-4b7bb8a2a289 | codetalker-speech-driven-3d-facial-animation | 2301.02379 | null | https://arxiv.org/abs/2301.02379v2 | https://arxiv.org/pdf/2301.02379v2.pdf | CodeTalker: Speech-Driven 3D Facial Animation with Discrete Motion Prior | Speech-driven 3D facial animation has been widely studied, yet there is still a gap to achieving realism and vividness due to the highly ill-posed nature and scarcity of audio-visual data. Existing works typically formulate the cross-modal mapping into a regression task, which suffers from the regression-to-mean problem leading to over-smoothed facial motions. In this paper, we propose to cast speech-driven facial animation as a code query task in a finite proxy space of the learned codebook, which effectively promotes the vividness of the generated motions by reducing the cross-modal mapping uncertainty. The codebook is learned by self-reconstruction over real facial motions and thus embedded with realistic facial motion priors. Over the discrete motion space, a temporal autoregressive model is employed to sequentially synthesize facial motions from the input speech signal, which guarantees lip-sync as well as plausible facial expressions. We demonstrate that our approach outperforms current state-of-the-art methods both qualitatively and quantitatively. Also, a user study further justifies our superiority in perceptual quality. | ['Tien-Tsin Wong', 'Jue Wang', 'Xiaodong Cun', 'Yuechen Zhang', 'Menghan Xia', 'Jinbo Xing'] | 2023-01-06 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Xing_CodeTalker_Speech-Driven_3D_Facial_Animation_With_Discrete_Motion_Prior_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Xing_CodeTalker_Speech-Driven_3D_Facial_Animation_With_Discrete_Motion_Prior_CVPR_2023_paper.pdf | cvpr-2023-1 | ['3d-face-animation'] | ['computer-vision'] | [ 1.19553581e-01 2.62791276e-01 -8.23957995e-02 -2.90849805e-01
-1.14181077e+00 -1.95603281e-01 7.47560859e-01 -7.51780689e-01
4.55688909e-02 4.23511654e-01 5.66964865e-01 2.53895432e-01
1.97766200e-01 -2.22596064e-01 -5.41376948e-01 -8.88258874e-01
-4.58837301e-02 5.32767661e-02 -2.70902157e-01 -2.72488147e-01
-1.29673272e-01 4.85905766e-01 -1.89608300e+00 8.11131671e-02
7.35290587e-01 8.92895222e-01 5.07022114e-03 7.28718400e-01
2.46598502e-03 7.05890298e-01 -3.31385493e-01 -4.34482604e-01
7.50081539e-02 -7.36808896e-01 -4.80449796e-01 4.77531582e-01
3.08038980e-01 -3.67676765e-01 -2.66761750e-01 1.17531502e+00
5.06578684e-01 2.00088605e-01 7.10801661e-01 -1.64527237e+00
-5.67241251e-01 -1.05315121e-02 -7.34219313e-01 -4.95856911e-01
5.06908059e-01 2.60719419e-01 9.22809422e-01 -1.07545197e+00
7.95466006e-01 1.74692786e+00 2.46184617e-01 1.07788706e+00
-1.49543512e+00 -7.98610032e-01 1.10490561e-01 -2.01338325e-02
-1.62063098e+00 -1.02799010e+00 1.04435515e+00 -5.26020825e-01
2.95911252e-01 3.63106012e-01 7.35728979e-01 1.27092862e+00
-1.60135493e-01 6.71292126e-01 8.10546756e-01 -4.11760598e-01
3.38805765e-01 2.78819632e-02 -9.32106555e-01 6.08743668e-01
-5.28093696e-01 3.30617577e-01 -7.04518437e-01 -1.98470429e-01
9.99294817e-01 -3.81297261e-01 -2.66210675e-01 -6.27947569e-01
-9.07122254e-01 8.44071925e-01 -9.41356085e-03 8.00145864e-02
-4.77478832e-01 2.91960806e-01 2.61338353e-01 -1.18446685e-01
6.56044602e-01 -1.40157059e-01 1.84617758e-01 -3.62441957e-01
-1.26756942e+00 4.47488248e-01 4.21799481e-01 8.82297397e-01
5.26243806e-01 5.72907746e-01 -9.42308363e-03 9.38361943e-01
8.21671903e-01 7.06851125e-01 8.00459757e-02 -1.55929661e+00
-3.65307368e-02 -2.81277448e-02 1.39196992e-01 -1.20594907e+00
1.09448776e-01 8.78507644e-02 -8.13133657e-01 6.23290360e-01
1.84523150e-01 -1.24022458e-02 -6.34871602e-01 2.26248002e+00
6.61077619e-01 4.98645157e-01 9.18420125e-03 1.22736168e+00
6.23493373e-01 7.95662224e-01 3.52164328e-01 -6.41483068e-01
1.24859142e+00 -6.99808657e-01 -1.20093000e+00 5.46122976e-02
1.38229996e-01 -8.32182705e-01 1.19263446e+00 3.99504900e-01
-1.48875093e+00 -5.29709995e-01 -6.87506199e-01 -2.00176369e-02
4.33688730e-01 9.31119360e-03 2.17984259e-01 5.78690171e-01
-1.17113960e+00 2.06808478e-01 -8.29518855e-01 -1.32438734e-01
4.25695151e-01 9.26965922e-02 -4.81814861e-01 2.30939835e-01
-1.10005748e+00 6.59563124e-01 -2.28221059e-01 2.50491500e-02
-1.07711792e+00 -8.07302952e-01 -1.12960649e+00 -1.10278994e-01
2.16810539e-01 -4.88940299e-01 1.30119979e+00 -1.32487226e+00
-2.13001537e+00 9.38084960e-01 -4.36061203e-01 3.84605979e-03
6.25958979e-01 -8.79108161e-03 -5.19361436e-01 4.18649584e-01
-2.03674007e-02 1.15549123e+00 1.41480100e+00 -1.65120423e+00
-3.38738561e-01 -1.28747731e-01 -1.63523570e-01 3.56081307e-01
-3.41253638e-01 1.56735137e-01 -8.33073318e-01 -9.78187144e-01
-9.82127860e-02 -1.00075829e+00 -2.28745446e-01 6.28443539e-01
-1.58424571e-01 -2.50414293e-02 9.50729311e-01 -5.95866501e-01
1.26877904e+00 -2.44820952e+00 3.40346754e-01 -3.29943411e-02
7.82231912e-02 6.45482726e-03 -3.02105457e-01 2.17762679e-01
-2.84636974e-01 -5.66763729e-02 -1.94576532e-01 -8.45871449e-01
5.76967001e-02 2.45639399e-01 -4.96217132e-01 7.26582706e-01
3.97814989e-01 6.85128808e-01 -8.09169471e-01 -7.49022722e-01
3.66834551e-01 1.10534704e+00 -7.84250855e-01 4.39339072e-01
-3.30270052e-01 8.60243857e-01 -2.71479845e-01 5.95888317e-01
7.35694230e-01 7.85123780e-02 -1.31457016e-01 -1.54134393e-01
-5.17094089e-03 -1.53865531e-01 -1.19665325e+00 2.09169555e+00
-5.19102931e-01 6.16180062e-01 5.68041861e-01 -6.72235847e-01
1.02971542e+00 5.93860567e-01 7.64408171e-01 -5.10440588e-01
1.07728370e-01 -4.35580686e-03 -3.59683841e-01 -6.92883432e-01
2.96792626e-01 -5.11874795e-01 1.41224638e-01 1.77906379e-01
-9.65125263e-02 -4.38707083e-01 -3.95770818e-01 3.14900875e-02
2.75904208e-01 5.03676236e-01 2.90211085e-02 -2.25698531e-01
6.99959099e-01 -4.25755143e-01 5.17361641e-01 -1.17737249e-01
-2.77084917e-01 7.77953327e-01 5.84649563e-01 1.05282748e-02
-1.10864532e+00 -9.81450796e-01 4.42432277e-02 9.62859094e-01
7.76490867e-02 -3.51161808e-01 -1.16999400e+00 -1.64517447e-01
-2.95212507e-01 6.98589444e-01 -7.19166934e-01 -1.53629184e-01
-4.54798490e-01 -2.97210701e-02 4.76005435e-01 1.32057741e-01
1.34700179e-01 -9.35994864e-01 -5.02683342e-01 6.65339306e-02
-3.38130116e-01 -1.19964004e+00 -7.66292155e-01 -6.38292611e-01
-6.11997545e-01 -6.01579905e-01 -1.12598193e+00 -7.12830603e-01
5.51289201e-01 1.45912796e-01 8.13628733e-01 -2.85906702e-01
-2.72023410e-01 4.24513131e-01 -5.82304746e-02 -1.05495453e-01
-7.58609772e-01 -6.41104221e-01 2.23276913e-01 5.39139926e-01
-9.94426906e-02 -7.02395439e-01 -6.43618286e-01 3.07031333e-01
-9.09142494e-01 3.22183341e-01 1.67667583e-01 8.13275337e-01
5.40537119e-01 -2.68243372e-01 4.68371272e-01 -2.01515660e-01
4.15064722e-01 -3.32743734e-01 -5.32201409e-01 -1.78080112e-01
-2.40845084e-01 -3.39039788e-02 3.71795475e-01 -7.14302361e-01
-1.24513376e+00 2.71956056e-01 -3.20149481e-01 -1.07193434e+00
-1.68327317e-01 5.53105213e-02 -5.38202465e-01 -1.88159272e-02
4.87702161e-01 1.21484995e-01 6.38743103e-01 -1.80708364e-01
7.26657748e-01 5.26000142e-01 7.49863803e-01 -6.49618626e-01
8.35302711e-01 7.89704204e-01 1.39762908e-01 -1.12448931e+00
-3.30336392e-01 2.37416644e-02 -3.27335954e-01 -6.34705245e-01
8.84220123e-01 -1.08686376e+00 -1.13698423e+00 3.03508103e-01
-1.19461298e+00 -2.74027765e-01 -2.27131516e-01 4.20568854e-01
-1.11472654e+00 5.03583372e-01 -3.97496074e-01 -1.27024233e+00
-1.47794411e-01 -1.44091594e+00 1.38496959e+00 -6.79551959e-02
-4.04097617e-01 -7.63436675e-01 8.23570192e-02 1.56558305e-01
2.22404733e-01 6.03299856e-01 6.46806002e-01 2.63367951e-01
-4.17465210e-01 8.95131677e-02 1.25409737e-01 2.45290607e-01
1.72827005e-01 4.51290041e-01 -1.23952532e+00 -2.10702553e-01
-6.22562766e-02 -3.87863457e-01 3.08862418e-01 6.79228902e-01
1.06317711e+00 -4.88692552e-01 7.38458708e-02 7.07653224e-01
9.38098550e-01 1.69629045e-02 6.09754562e-01 -2.18265653e-01
4.40359145e-01 1.21770096e+00 8.57658803e-01 7.00457692e-01
2.57710695e-01 9.97040927e-01 6.74442351e-01 -1.86645359e-01
-2.16154158e-01 -5.18864274e-01 4.50365156e-01 6.29959583e-01
-8.06560814e-02 8.27388391e-02 -4.72121626e-01 5.26919782e-01
-1.68645298e+00 -1.03659701e+00 2.22816661e-01 2.03312087e+00
9.72000599e-01 -3.45153868e-01 4.17170107e-01 9.13646817e-02
7.14435399e-01 3.24092031e-01 -4.02763247e-01 -1.84569210e-01
1.49151096e-02 -1.03545412e-02 -2.55819172e-01 7.95078874e-01
-9.16060209e-01 9.87797678e-01 5.94021511e+00 1.19907200e+00
-1.36224484e+00 -5.41063119e-03 7.55557060e-01 -1.83939919e-01
-5.12420774e-01 -1.97047710e-01 -3.48052472e-01 3.83386523e-01
8.12932074e-01 -1.99797735e-01 4.27490294e-01 8.28910053e-01
8.36025298e-01 1.74801618e-01 -9.54602420e-01 1.35580671e+00
5.55533469e-02 -1.28995216e+00 1.52081117e-01 1.85504079e-01
6.95361197e-01 -7.36935616e-01 5.32454610e-01 -3.78137231e-02
8.71557221e-02 -1.19892383e+00 1.10910153e+00 5.65151930e-01
1.47760832e+00 -1.02217674e+00 -2.13171914e-01 2.39523649e-01
-1.31893909e+00 2.50266969e-01 -1.70802325e-03 3.40906501e-01
4.63649809e-01 1.14650033e-01 -3.94048870e-01 9.49229747e-02
4.84357715e-01 6.23609245e-01 2.31715202e-01 3.90838742e-01
-1.06767759e-01 3.69275033e-01 -3.04448698e-02 1.88741714e-01
1.39262840e-01 -3.35374117e-01 7.47005701e-01 1.08179438e+00
4.24698651e-01 2.67842352e-01 -2.80753393e-02 1.13195217e+00
1.13781527e-01 1.80557072e-01 -5.62880218e-01 2.80446690e-02
4.51257825e-01 1.20811141e+00 -2.27085099e-01 -5.85433133e-02
-2.87062556e-01 1.03758562e+00 -1.16544493e-01 5.74976683e-01
-1.01239336e+00 9.90597755e-02 1.26018250e+00 8.69885981e-02
1.39836930e-02 -2.74158418e-01 -9.97082889e-02 -9.19242978e-01
-8.79901648e-02 -1.04086244e+00 -8.55354145e-02 -9.07037973e-01
-9.00881469e-01 7.97052085e-01 -7.01163546e-04 -1.52354300e+00
-7.10420966e-01 -1.52153760e-01 -3.93344492e-01 9.68508005e-01
-1.37589455e+00 -1.23525131e+00 -3.56681108e-01 8.92034292e-01
7.19586492e-01 -5.80519773e-02 8.70028198e-01 2.87077665e-01
-2.55327910e-01 7.41372406e-01 -2.28063747e-01 -1.26482427e-01
7.84452081e-01 -5.91710389e-01 2.11956933e-01 5.89015782e-01
-6.03940785e-02 3.42049330e-01 1.07010472e+00 -3.64488363e-01
-1.50138366e+00 -1.01923859e+00 6.50473773e-01 -1.17002495e-01
6.17547512e-01 -4.52759296e-01 -9.63779628e-01 1.19806670e-01
1.25789538e-01 2.76752621e-01 5.87980568e-01 -5.55734634e-01
-3.31583411e-01 -7.53492862e-02 -1.23333943e+00 1.00405490e+00
8.91557574e-01 -6.44726574e-01 -2.92975623e-02 -1.48604214e-01
6.86466455e-01 -3.09367269e-01 -7.81118631e-01 1.85120657e-01
6.97363973e-01 -9.66305614e-01 1.01789165e+00 -5.08440495e-01
5.11505425e-01 -2.70486116e-01 -2.50344247e-01 -1.13164151e+00
-3.13321799e-02 -1.28124726e+00 -1.63332015e-01 1.51344490e+00
-2.00509280e-02 2.81476323e-02 8.78430665e-01 6.17847741e-01
1.69571742e-01 -6.31724954e-01 -1.14606738e+00 -5.65017223e-01
9.89368856e-02 -6.47343695e-01 3.97477716e-01 8.50949168e-01
4.62997071e-02 -4.34746332e-02 -7.10239589e-01 2.25736365e-01
8.00562501e-01 -7.71413073e-02 9.01449680e-01 -8.76442969e-01
-1.23532534e-01 -5.46640813e-01 -3.30865443e-01 -1.13430929e+00
6.16503894e-01 -3.91352028e-01 3.44583154e-01 -8.90969098e-01
-2.48606149e-02 -2.18297154e-01 4.67413247e-01 4.00203429e-02
4.57454138e-02 3.11440587e-01 2.30128229e-01 2.40931243e-01
-1.80100724e-01 1.09908700e+00 1.57208025e+00 1.08757474e-01
-2.40892738e-01 -1.41538739e-01 -3.89348686e-01 8.41349185e-01
3.13001662e-01 -2.03244120e-01 -6.22196078e-01 -1.60351247e-01
-2.12782800e-01 7.43692577e-01 4.92314279e-01 -6.87927246e-01
8.61406177e-02 -4.35438186e-01 -4.11826074e-02 -3.07781905e-01
1.10703385e+00 -8.42350900e-01 3.52901518e-01 3.15869421e-01
-5.73827624e-01 -2.93899149e-01 1.63781360e-01 6.22769356e-01
-3.38495731e-01 2.91001737e-01 1.31406081e+00 3.31456602e-01
-3.38606596e-01 5.86793959e-01 -3.70907515e-01 1.35866389e-01
9.72481728e-01 -2.31401235e-01 3.63249183e-01 -1.04110646e+00
-8.37495089e-01 -7.98246190e-02 5.46276152e-01 4.86192793e-01
8.23721170e-01 -1.78277075e+00 -9.15786266e-01 4.31815863e-01
7.90492073e-02 -1.61214098e-01 5.24016619e-01 7.61138022e-01
-2.23541379e-01 9.80419666e-02 -2.07127795e-01 -9.27699625e-01
-1.50062907e+00 5.69542944e-01 3.38524252e-01 4.48815197e-01
-5.57599664e-01 7.34662414e-01 3.48747581e-01 1.24242030e-01
5.15459299e-01 7.54778460e-02 -1.51614174e-01 4.66045141e-02
5.43547213e-01 2.40050033e-01 -4.36360836e-01 -1.26157975e+00
-2.61874139e-01 9.02455509e-01 5.07069409e-01 -7.14022875e-01
1.01295733e+00 -3.55514467e-01 1.40686631e-01 2.57451862e-01
1.44106877e+00 1.97659865e-01 -1.93959963e+00 -1.52486488e-01
-2.95720577e-01 -7.23942101e-01 -1.38066104e-02 -1.56889230e-01
-1.17828333e+00 1.07772923e+00 4.96871382e-01 -1.70143723e-01
1.24955916e+00 -4.41834331e-03 5.54500163e-01 -3.79820794e-01
1.60369277e-01 -7.93725073e-01 3.58214617e-01 9.95016396e-02
1.16433406e+00 -1.11942375e+00 -3.92905474e-01 -6.16766930e-01
-9.81097996e-01 1.01482320e+00 3.29022437e-01 -1.36917859e-01
6.99669063e-01 4.67207789e-01 2.23440468e-01 1.54123992e-01
-7.61729896e-01 -4.96258177e-02 4.50474173e-01 8.53368521e-01
3.76430899e-01 -1.94556713e-01 1.42075296e-04 4.75339532e-01
-1.70942381e-01 -3.74973216e-03 2.25870028e-01 3.94377112e-01
-2.40011483e-01 -8.16286564e-01 -5.04638374e-01 -4.83828038e-01
-3.64050210e-01 4.66002412e-02 -2.82063875e-02 7.39819765e-01
-1.56140432e-01 1.06228471e+00 5.63762672e-02 -1.77007928e-01
1.69760942e-01 -1.13654770e-01 4.19925183e-01 -2.27587640e-01
3.13316211e-02 8.05240214e-01 -1.11461997e-01 -9.12895620e-01
-5.40344536e-01 -7.84149170e-01 -1.30354273e+00 -3.80624384e-01
-2.72340849e-02 9.98311788e-02 6.31264925e-01 5.02368987e-01
4.29352939e-01 2.66030699e-01 9.73079741e-01 -1.18603826e+00
-6.18772328e-01 -6.83714747e-01 -5.71420729e-01 5.37308633e-01
7.37769186e-01 -8.11137795e-01 -4.51330721e-01 4.54631299e-01] | [13.233363151550293, -0.4067801535129547] |
fc00b3ea-2b0f-4db4-bda0-9f0d6dbc3b8e | inter-frame-accelerate-attack-against-video | 2305.06540 | null | https://arxiv.org/abs/2305.06540v1 | https://arxiv.org/pdf/2305.06540v1.pdf | Inter-frame Accelerate Attack against Video Interpolation Models | Deep learning based video frame interpolation (VIF) method, aiming to synthesis the intermediate frames to enhance video quality, have been highly developed in the past few years. This paper investigates the adversarial robustness of VIF models. We apply adversarial attacks to VIF models and find that the VIF models are very vulnerable to adversarial examples. To improve attack efficiency, we suggest to make full use of the property of video frame interpolation task. The intuition is that the gap between adjacent frames would be small, leading to the corresponding adversarial perturbations being similar as well. Then we propose a novel attack method named Inter-frame Accelerate Attack (IAA) that initializes the perturbation as the perturbation for the previous adjacent frame and reduces the number of attack iterations. It is shown that our method can improve attack efficiency greatly while achieving comparable attack performance with traditional methods. Besides, we also extend our method to video recognition models which are higher level vision tasks and achieves great attack efficiency. | ['Xiaochun Cao', 'Baoyuan Wu', 'Wenyuan Yang', 'Liang Yi', 'Zhikai Chen', 'Junpei Liao'] | 2023-05-11 | null | null | null | null | ['video-recognition', 'video-frame-interpolation'] | ['computer-vision', 'computer-vision'] | [ 2.43704587e-01 -1.46201193e-01 1.12726642e-02 1.46586010e-02
-5.47340989e-01 -6.39851451e-01 5.26163816e-01 -4.56099212e-01
-2.83701539e-01 5.11866987e-01 1.61231339e-01 -4.91630822e-01
2.77913243e-01 -7.82796919e-01 -1.03750682e+00 -7.23154604e-01
-2.25074098e-01 -4.70930040e-01 4.56061810e-01 -3.14593315e-01
2.82870919e-01 5.83345950e-01 -9.28838789e-01 2.72310287e-01
5.90557635e-01 9.44808602e-01 -3.37359577e-01 9.96493399e-01
2.35422626e-01 1.18065619e+00 -8.80058169e-01 -7.93999195e-01
8.17026675e-01 -4.13840741e-01 -7.66701818e-01 -7.27992058e-02
4.54116911e-01 -9.57422197e-01 -1.04697132e+00 1.34692657e+00
6.09516978e-01 9.00287405e-02 3.97782445e-01 -1.71574068e+00
-5.71113646e-01 5.10301948e-01 -6.24189019e-01 2.92078733e-01
2.83858985e-01 3.08839113e-01 4.27478611e-01 -5.91152549e-01
2.86572576e-01 1.43698525e+00 7.93540657e-01 9.05703187e-01
-7.40110934e-01 -8.41756344e-01 1.57103866e-01 4.78970438e-01
-1.26317394e+00 -3.29932690e-01 8.30593050e-01 -1.77627131e-01
5.53777337e-01 4.52173829e-01 2.86383808e-01 1.18657470e+00
3.63736123e-01 8.22421491e-01 7.87069499e-01 -2.91754514e-01
1.96227208e-01 -2.64533907e-01 -3.75729084e-01 5.69215000e-01
-2.48545539e-02 4.21983600e-01 -1.66832760e-01 -3.23894531e-01
9.47340608e-01 -3.71513963e-02 -6.75054967e-01 1.63262248e-01
-1.08551478e+00 9.30762112e-01 4.52244043e-01 -1.37139618e-01
-1.89007118e-01 5.20819366e-01 6.75071299e-01 3.91022593e-01
-2.55074576e-02 6.26143366e-02 -5.12815192e-02 1.50520682e-01
-6.70126677e-01 5.46035051e-01 5.54223716e-01 7.92048275e-01
3.42908919e-01 6.07947946e-01 -2.99688458e-01 3.92486274e-01
4.03686523e-01 4.56782758e-01 1.84371293e-01 -1.52001679e+00
5.14256179e-01 -1.88734040e-01 1.34639874e-01 -1.29364467e+00
2.01785907e-01 1.26566499e-01 -9.68716443e-01 7.33485937e-01
5.36026478e-01 -3.70571852e-01 -5.82725406e-01 1.87267137e+00
2.35502869e-01 8.03180218e-01 1.87014937e-01 8.54659438e-01
5.84542751e-01 8.87063742e-01 -6.59076199e-02 -1.02198102e-01
1.03501475e+00 -7.79340863e-01 -6.26122296e-01 1.60691425e-01
1.73032284e-01 -1.07745004e+00 8.95173490e-01 4.06893671e-01
-1.18812597e+00 -7.26620078e-01 -1.20354974e+00 2.56272942e-01
7.24022985e-02 -5.13709188e-01 4.36063439e-01 9.36318815e-01
-9.00623620e-01 6.66681588e-01 -6.41310632e-01 1.64172426e-01
5.60812652e-01 4.97342110e-01 -2.91331828e-01 1.11283600e-01
-1.59035742e+00 5.06875277e-01 3.07223588e-01 1.06311902e-01
-1.26265502e+00 -7.57834733e-01 -7.85925567e-01 -8.36704448e-02
2.54262596e-01 -6.54310048e-01 1.04246247e+00 -1.31427944e+00
-1.71163905e+00 3.15765023e-01 -3.01170405e-02 -9.28159595e-01
8.32511485e-01 -3.02928120e-01 -4.45284545e-01 2.87750125e-01
-3.50009739e-01 5.81659794e-01 1.45902526e+00 -1.26193488e+00
-4.79710370e-01 -1.07077532e-03 6.43243253e-01 -2.11538717e-01
-4.04144585e-01 2.39609465e-01 -2.25593254e-01 -1.12689006e+00
-5.20763278e-01 -1.00241649e+00 -4.23267692e-01 3.31613719e-01
-3.19134623e-01 1.30863786e-01 1.20445979e+00 -6.16932809e-01
1.28397858e+00 -2.23694277e+00 -1.07226428e-02 3.46656680e-01
2.42223069e-01 7.25237548e-01 -2.23205850e-01 2.12971583e-01
-3.29288393e-01 2.11834162e-01 -1.67784020e-01 1.84555620e-01
-9.24207717e-02 2.19636887e-01 -8.20250869e-01 6.37391984e-01
2.22231373e-01 7.65510917e-01 -7.51376987e-01 -2.69950807e-01
2.33091757e-01 6.06897235e-01 -8.78934741e-01 3.41585368e-01
1.12067379e-01 3.20603967e-01 -3.08142871e-01 3.80127132e-01
1.07634485e+00 1.88391626e-01 -1.43578008e-01 -4.88383800e-01
3.03647488e-01 -3.88899356e-01 -1.28138828e+00 1.13057625e+00
-2.81507105e-01 7.25147545e-01 9.66939107e-02 -9.39976871e-01
7.37671494e-01 4.59026366e-01 3.55822951e-01 -1.76514104e-01
3.26304168e-01 -6.51088730e-02 1.84364051e-01 -4.12916631e-01
2.04321608e-01 -1.56680103e-02 1.42384022e-01 1.25023844e-02
-4.13208932e-01 1.17056053e-02 -1.52174249e-01 4.04522955e-01
9.46096838e-01 2.26404574e-02 3.12606573e-01 -2.82881521e-02
1.11753047e+00 -5.64945996e-01 7.12574601e-01 8.94198656e-01
-5.22045851e-01 4.79897320e-01 5.22112906e-01 -5.46776831e-01
-1.12236369e+00 -1.26854646e+00 2.55697101e-01 4.41089094e-01
3.96378487e-01 -5.54994285e-01 -1.05979288e+00 -8.55454504e-01
-6.52495101e-02 2.57665724e-01 -5.18057108e-01 -4.32893783e-01
-9.02363896e-01 -4.44910318e-01 1.03149712e+00 6.72322214e-01
9.44386899e-01 -7.65123248e-01 -2.37118423e-01 6.69648349e-02
-1.39750615e-01 -1.12148488e+00 -7.44479477e-01 -5.03167033e-01
-5.64025104e-01 -1.03050756e+00 -8.03034961e-01 -7.75031626e-01
5.20381331e-01 2.70279616e-01 8.48570824e-01 2.86207438e-01
1.42991081e-01 2.27260008e-01 -2.61377335e-01 -3.20582777e-01
-8.72039974e-01 -6.24375105e-01 2.65727788e-01 1.89868167e-01
1.20639063e-01 -5.07585406e-01 -6.37413204e-01 4.06462103e-01
-1.18784034e+00 -3.22338939e-01 7.70524051e-03 1.02568030e+00
4.12303895e-01 1.95147753e-01 5.43878496e-01 -4.88694400e-01
4.27992702e-01 -3.12308639e-01 -7.39875138e-01 1.28706917e-01
-1.52145177e-01 -1.00753568e-01 1.17649353e+00 -7.00332999e-01
-7.79037714e-01 -1.81853160e-01 -6.26639724e-01 -9.76266325e-01
4.94454578e-02 -5.04270196e-03 -4.16700631e-01 -8.14472616e-01
6.13342762e-01 1.66427568e-02 1.60248116e-01 7.56277591e-02
2.17312232e-01 5.15890837e-01 8.38084519e-01 -5.19828022e-01
1.27763319e+00 5.71859539e-01 2.23204538e-01 -7.25573719e-01
-2.74460316e-01 2.64670581e-01 2.82495599e-02 -4.17513549e-01
6.82487905e-01 -9.22165453e-01 -1.20487761e+00 6.49372876e-01
-1.34990728e+00 -2.88815916e-01 4.22909409e-02 4.62456644e-01
-6.24975264e-01 1.05006886e+00 -9.69334781e-01 -5.34657001e-01
-3.18312347e-01 -1.51697493e+00 6.35016799e-01 1.06433131e-01
3.78950872e-02 -9.41988409e-01 -1.92276314e-01 1.24646965e-02
3.34215611e-01 6.94494188e-01 5.41844726e-01 -2.43303254e-01
-7.38230288e-01 -2.21511573e-02 -6.55906051e-02 6.60053134e-01
1.25138924e-01 3.09926212e-01 -7.40358531e-01 -5.47169387e-01
3.74793321e-01 -5.39415218e-02 6.90241456e-01 3.30977082e-01
1.62542093e+00 -7.65434623e-01 1.79431245e-01 1.01239944e+00
1.49209666e+00 3.91732782e-01 1.31618345e+00 5.48613012e-01
8.43623579e-01 2.17011511e-01 5.06607890e-01 4.88533586e-01
-1.88694373e-01 7.51796186e-01 7.10272014e-01 7.67421583e-03
6.73910137e-03 -6.59714043e-02 8.11417937e-01 3.40833873e-01
-2.18315825e-01 -3.28557551e-01 -4.10748065e-01 2.38774955e-01
-1.67039990e+00 -1.37434602e+00 1.56540535e-02 2.12938690e+00
6.83986247e-01 3.17012817e-01 1.59469128e-01 4.82411236e-01
8.66672873e-01 3.48818630e-01 -4.69489247e-01 -6.58970892e-01
-3.06211747e-02 2.78041691e-01 7.30839849e-01 7.07176864e-01
-1.31232309e+00 7.98712850e-01 6.54799700e+00 1.25555933e+00
-1.02674985e+00 -1.00313053e-01 7.17865050e-01 1.73772916e-01
-2.39506528e-01 -6.60646558e-02 -4.99395639e-01 7.07955003e-01
8.47861886e-01 -2.62266606e-01 6.72356009e-01 7.00890243e-01
1.82332501e-01 5.71928680e-01 -8.86122644e-01 1.02034366e+00
1.90723091e-02 -1.50236678e+00 5.81042349e-01 -1.45011947e-01
6.83675647e-01 -5.26295483e-01 2.61760861e-01 -1.84747472e-03
2.87724495e-01 -9.77854669e-01 6.05641305e-01 1.22276023e-01
7.10403144e-01 -1.15066218e+00 6.61871612e-01 -2.95367725e-02
-1.25280654e+00 -2.45647341e-01 -5.39394617e-01 -7.31306989e-03
2.69244015e-01 4.46284525e-02 -3.09387863e-01 5.62294304e-01
7.41525590e-01 5.68460405e-01 -3.20700586e-01 9.32686567e-01
-2.34674990e-01 7.18913496e-01 -1.72135487e-01 4.28461522e-01
4.28702265e-01 4.61229496e-02 7.48409033e-01 9.46404874e-01
2.72497535e-01 7.96884969e-02 1.42251462e-01 4.69865918e-01
-1.35556191e-01 -1.97721124e-01 -6.76282644e-01 3.43269765e-01
4.78868872e-01 7.79102743e-01 -3.14929783e-01 -3.11807692e-01
-6.67649150e-01 1.13906145e+00 -2.87164688e-01 4.06489998e-01
-1.50961781e+00 -6.05767429e-01 1.09640181e+00 -2.45324463e-01
3.26917231e-01 -1.34451479e-01 -7.18111023e-02 -1.19808090e+00
-1.59836486e-02 -1.49353373e+00 2.12578669e-01 -4.55843776e-01
-1.17538881e+00 6.18968725e-01 -2.66533136e-01 -1.73704541e+00
-3.13631624e-01 -4.53721136e-01 -8.12627733e-01 7.50941873e-01
-1.44992113e+00 -9.77770627e-01 -2.57096022e-01 1.27254045e+00
5.70121109e-01 -4.24391657e-01 4.88381684e-01 5.04063904e-01
-4.76713628e-01 1.23921192e+00 9.80348140e-02 5.83284318e-01
6.74571455e-01 -7.26245344e-01 8.27152550e-01 1.49071705e+00
-1.19073965e-01 6.31413877e-01 7.10277140e-01 -4.42801088e-01
-1.33755696e+00 -1.30378008e+00 2.58425981e-01 -1.18474312e-01
6.61255419e-01 -4.93699946e-02 -9.93759215e-01 5.29749393e-01
2.88593471e-01 1.90104455e-01 3.69974643e-01 -9.27371442e-01
-6.30908608e-01 -2.25275889e-01 -1.36965895e+00 9.86560822e-01
8.25463712e-01 -6.25811517e-01 -3.86855692e-01 2.09238857e-01
9.31823194e-01 -5.42999327e-01 -9.09720480e-01 4.34103400e-01
4.97636557e-01 -1.09254110e+00 1.48305643e+00 -6.13172114e-01
5.70552170e-01 -7.00226605e-01 -3.97028267e-01 -8.55086267e-01
-2.96781093e-01 -1.24222612e+00 -4.09469903e-01 1.15257108e+00
-1.64205283e-01 -6.29514575e-01 7.08265126e-01 4.66430277e-01
6.70499131e-02 -4.88494009e-01 -8.41813982e-01 -1.06481397e+00
1.44733325e-01 -2.38316938e-01 7.95796096e-01 7.19599903e-01
-3.99787426e-01 -2.64675349e-01 -8.90593469e-01 6.54761374e-01
6.90582275e-01 -4.09710556e-01 9.94622409e-01 -5.14165521e-01
-5.67978978e-01 -1.84862092e-01 -7.60922134e-01 -1.09552681e+00
3.56940210e-01 -3.99566650e-01 -2.04360962e-01 -6.87618136e-01
-2.49304280e-01 -1.11003786e-01 -3.05251777e-01 1.95340931e-01
-5.23988426e-01 5.76785505e-01 6.87950790e-01 1.26650274e-01
-4.73806709e-02 3.34757417e-01 1.13139403e+00 -3.02398384e-01
1.64254218e-01 9.61729214e-02 -5.02435207e-01 8.99983346e-01
6.88024044e-01 -1.90257043e-01 -4.25229639e-01 -4.99070525e-01
-2.04091847e-01 2.23450720e-01 5.77467501e-01 -1.20075703e+00
-2.90440721e-03 -1.36106029e-01 4.69206691e-01 -1.67589039e-01
1.73298091e-01 -1.08695292e+00 3.06369096e-01 8.41449499e-01
-4.24100846e-01 1.99053958e-01 2.06980258e-01 5.45335829e-01
-3.31040710e-01 -2.02815920e-01 1.21382821e+00 -3.95939536e-02
-7.73533940e-01 6.14719152e-01 -3.46570164e-01 -1.13312028e-01
1.14279473e+00 -2.09078133e-01 -2.30734885e-01 -6.49365425e-01
-4.96287078e-01 -8.95616189e-02 5.22003651e-01 3.11758161e-01
7.99282968e-01 -1.66441536e+00 -8.36346030e-01 2.54914075e-01
-5.00786901e-01 -3.64873379e-01 3.12159270e-01 3.37884665e-01
-9.64371264e-01 -1.30902633e-01 -4.29493308e-01 -3.20649981e-01
-1.56045461e+00 1.16520226e+00 4.26406384e-01 -4.74543124e-02
-4.37287778e-01 8.05233896e-01 3.94864738e-01 2.48402372e-01
3.65425467e-01 5.46047427e-02 -5.45631722e-02 -5.21537960e-01
1.08803260e+00 6.45265996e-01 -3.79937887e-01 -6.30686522e-01
-3.17614049e-01 7.22138226e-01 -1.45251185e-01 1.96894005e-01
9.06757951e-01 -1.47916839e-01 -2.13152147e-03 -4.45050120e-01
1.44697070e+00 1.84777737e-01 -1.63411224e+00 4.90834303e-02
-5.06039679e-01 -1.01055837e+00 -1.46823063e-01 -6.97670504e-02
-1.45488298e+00 8.20671141e-01 7.70860493e-01 1.79374069e-01
1.43879664e+00 -6.87054038e-01 1.29800475e+00 2.47124568e-01
3.91480267e-01 -5.97179830e-01 6.91044629e-02 4.29585338e-01
7.46162713e-01 -9.83829796e-01 -1.62723288e-02 -5.65179408e-01
-4.89251137e-01 1.18907750e+00 5.72914660e-01 -6.98910177e-01
4.91449803e-01 5.77766478e-01 3.43397725e-03 6.35041654e-01
-3.45574468e-01 2.33613804e-01 -1.88722759e-02 8.71889114e-01
-9.68257803e-03 -4.12225723e-01 -2.91036218e-01 1.88645631e-01
4.90484200e-02 6.93169162e-02 7.37968385e-01 6.11042321e-01
-1.23472705e-01 -1.33863735e+00 -8.06275547e-01 -6.60668761e-02
-1.05715287e+00 -1.01514749e-01 -3.06247044e-02 6.22235000e-01
-3.36434245e-02 9.83243346e-01 -1.54240906e-01 -6.99840486e-01
1.41594723e-01 -3.90226424e-01 4.91330683e-01 2.95509100e-01
-6.55933499e-01 -3.98415048e-03 -2.16090903e-01 -7.84746051e-01
-5.00941098e-01 -3.09599936e-01 -1.06585860e+00 -9.05115426e-01
-1.10832870e-01 8.37590545e-02 3.72722857e-02 6.90053463e-01
1.63457051e-01 4.38943774e-01 1.21800935e+00 -9.41303849e-01
-6.67146683e-01 -2.39832669e-01 -2.82018811e-01 5.31475782e-01
5.80850065e-01 -3.61674696e-01 -5.40005147e-01 3.04467797e-01] | [5.382009983062744, 7.935990810394287] |
6fc3aa8c-990a-4e18-afde-fb71f5836050 | point-cloud-quality-assessment-using-3d | 2209.15475 | null | https://arxiv.org/abs/2209.15475v1 | https://arxiv.org/pdf/2209.15475v1.pdf | Point Cloud Quality Assessment using 3D Saliency Maps | Point cloud quality assessment (PCQA) has become an appealing research field in recent days. Considering the importance of saliency detection in quality assessment, we propose an effective full-reference PCQA metric which makes the first attempt to utilize the saliency information to facilitate quality prediction, called point cloud quality assessment using 3D saliency maps (PQSM). Specifically, we first propose a projection-based point cloud saliency map generation method, in which depth information is introduced to better reflect the geometric characteristics of point clouds. Then, we construct point cloud local neighborhoods to derive three structural descriptors to indicate the geometry, color and saliency discrepancies. Finally, a saliency-based pooling strategy is proposed to generate the final quality score. Extensive experiments are performed on four independent PCQA databases. The results demonstrate that the proposed PQSM shows competitive performances compared to multiple state-of-the-art PCQA metrics. | ['Shan Liu', 'Jun Sun', 'Yiling Xu', 'Qi Yang', 'Yujie Zhang', 'Zhengyu Wang'] | 2022-09-30 | null | null | null | null | ['saliency-detection'] | ['computer-vision'] | [-2.55654231e-02 -5.59108973e-01 1.14016846e-01 -2.06646994e-01
-9.91995156e-01 -1.60112590e-01 4.62442547e-01 3.38718414e-01
9.49768648e-02 2.99755871e-01 2.06269890e-01 1.48590580e-01
-1.89377040e-01 -7.94230759e-01 -4.56459045e-01 -5.22552550e-01
1.86726555e-01 7.66100883e-02 7.10478187e-01 -3.06040555e-01
9.26875770e-01 8.17985773e-01 -1.67459345e+00 1.11580782e-01
1.51553178e+00 1.27780688e+00 4.83478069e-01 9.09614861e-02
-2.23467514e-01 1.92310184e-01 -5.91227949e-01 -3.68402302e-01
2.82378376e-01 -5.43169491e-02 -5.94270587e-01 1.95559874e-01
1.84186354e-01 -1.68005198e-01 -1.00823358e-01 1.34608507e+00
4.65584844e-01 1.04892485e-01 4.44435000e-01 -1.36542964e+00
-7.63402343e-01 -2.19250500e-01 -6.28773093e-01 5.19397497e-01
5.53625107e-01 3.22885275e-01 1.18028724e+00 -1.44237864e+00
2.41119772e-01 1.30318522e+00 3.18864197e-01 1.82452556e-02
-8.16912115e-01 -5.98972201e-01 -1.51742306e-02 6.84993088e-01
-1.43141162e+00 -8.81692097e-02 1.36100221e+00 -1.80900782e-01
5.21780908e-01 4.15059596e-01 7.74479210e-01 2.15400979e-01
1.73200712e-01 1.03140211e+00 1.12033117e+00 -2.11980924e-01
4.20797706e-01 -1.09450392e-01 -2.61019498e-01 6.09947443e-01
8.77839178e-02 2.95011327e-02 -6.11609697e-01 -2.24220768e-01
9.54504132e-01 1.78699479e-01 -2.30763644e-01 -7.93003976e-01
-1.31164527e+00 5.41387677e-01 1.00708544e+00 1.35553613e-01
-5.65486789e-01 -1.74311712e-01 4.39971872e-02 -1.24178439e-01
5.62074065e-01 6.01936758e-01 -3.87865864e-02 1.12768291e-02
-9.31953847e-01 4.76396382e-01 -1.59815811e-02 9.28111553e-01
8.71929884e-01 -1.61346003e-01 -3.25531721e-01 8.15690815e-01
4.03059274e-01 7.73834765e-01 2.45165318e-01 -1.08503354e+00
6.04442120e-01 1.05491054e+00 3.56937438e-01 -1.52126145e+00
-1.88373506e-01 -3.73223543e-01 -4.75545257e-01 3.32280666e-01
-2.61839926e-01 5.30181527e-01 -6.94534242e-01 9.27002728e-01
3.77546668e-01 3.57161075e-01 -1.44459605e-01 1.44528127e+00
8.48669171e-01 6.67537570e-01 -7.09278360e-02 8.76034945e-02
1.11188149e+00 -9.25508201e-01 -2.94444799e-01 1.35900706e-01
-5.17115965e-02 -7.46365488e-01 1.29181194e+00 2.46080503e-01
-1.22993124e+00 -6.78626001e-01 -1.24983716e+00 -1.85625069e-02
-1.69412211e-01 3.66396122e-02 4.98827130e-01 4.24421966e-01
-1.07441533e+00 5.53254783e-01 -8.28645349e-01 -5.65413684e-02
5.78553021e-01 2.32913882e-01 -8.77791084e-03 -1.70809105e-01
-1.12234879e+00 8.82722139e-01 1.53803915e-01 -5.22085056e-02
-7.79346526e-01 -8.22318316e-01 -7.30628192e-01 2.28907183e-01
1.30200744e-01 -5.62130213e-01 9.81787026e-01 -5.49529672e-01
-1.48597026e+00 6.95593059e-01 -4.26897466e-01 -9.51375347e-03
1.15417674e-01 -1.62330016e-01 -5.24547040e-01 6.99773610e-01
3.87045175e-01 8.24063361e-01 8.70182157e-01 -1.39295888e+00
-9.11026537e-01 -3.46111000e-01 5.11339754e-02 7.12658644e-01
-8.72849207e-03 2.27559894e-01 -7.37295747e-01 -4.87941712e-01
6.06154621e-01 -5.72387278e-01 -1.47721887e-01 9.36093926e-02
-2.70797849e-01 -3.75459105e-01 6.48858428e-01 -4.57209140e-01
1.12450421e+00 -2.03984189e+00 1.99881345e-01 4.01512921e-01
3.25030416e-01 2.05222249e-01 -1.37972176e-01 1.05713837e-01
6.97836056e-02 -8.70183930e-02 -4.34268832e-01 -1.58398122e-01
-6.34606853e-02 -2.65571386e-01 -3.26959491e-01 2.75898695e-01
8.29621792e-01 1.01651740e+00 -1.10424197e+00 -7.18951166e-01
5.46454310e-01 2.07870379e-01 -5.01301885e-01 6.22502714e-02
-2.28206837e-03 3.65399063e-01 -9.14325356e-01 1.07842147e+00
1.12112772e+00 -4.27418917e-01 -7.20533192e-01 -1.79462686e-01
-3.49150956e-01 2.54780591e-01 -8.68000865e-01 1.91823554e+00
-2.09059983e-01 3.70628655e-01 -3.30345482e-01 -4.56149548e-01
1.29611480e+00 -1.71169657e-02 5.90068281e-01 -8.93616557e-01
6.08480386e-02 3.10586184e-01 -2.45279223e-01 -5.71628101e-03
7.88209915e-01 -6.47671446e-02 1.02270216e-01 -9.89454519e-03
-1.83298022e-01 -6.91561043e-01 -1.55249447e-01 8.35409835e-02
8.62869620e-01 -3.40702981e-02 -4.00151052e-02 -1.99601725e-01
8.70548785e-01 2.54554451e-01 6.28604054e-01 7.38474131e-02
-6.67390466e-01 1.29031968e+00 1.32118061e-01 -2.94068843e-01
-1.01375699e+00 -1.47033918e+00 7.11112330e-03 5.29156923e-01
1.02642703e+00 -2.64768779e-01 -5.59323311e-01 -4.82452273e-01
-5.72918281e-02 5.02517462e-01 -2.49947235e-01 -1.71148151e-01
-4.64513272e-01 -3.89777064e-01 -1.38724416e-01 3.23602945e-01
8.68208110e-01 -1.26747179e+00 -5.38248301e-01 1.21296786e-01
-3.92965823e-01 -8.27459693e-01 -3.90822262e-01 -6.53943419e-01
-1.10987747e+00 -8.85510802e-01 -1.01429164e+00 -7.65529692e-01
5.31372428e-01 9.91668642e-01 1.09229100e+00 2.42385775e-01
1.20046742e-01 1.41137972e-01 -5.36848307e-01 -4.78925973e-01
3.00201923e-01 -4.57103811e-02 -6.48886040e-02 3.69805954e-02
5.88943005e-01 -4.70767200e-01 -1.03205407e+00 4.18604910e-01
-7.46623158e-01 9.61106047e-02 9.29511189e-01 2.73053676e-01
9.43129659e-01 5.76937534e-02 3.99573892e-01 -1.30711392e-01
7.06836879e-01 -3.16495866e-01 -6.97650075e-01 2.78188705e-01
-5.06575167e-01 -1.40764713e-01 3.07927251e-01 1.44302234e-01
-9.20249701e-01 -2.12729514e-01 -7.64259845e-02 -7.98429012e-01
9.70184896e-03 4.30051863e-01 -4.19061482e-01 -3.91743273e-01
2.36884743e-01 4.99645710e-01 -2.80232310e-01 -5.11466622e-01
2.85453767e-01 4.83212501e-01 6.28617644e-01 -3.82324129e-01
1.08763909e+00 6.42953873e-01 2.32604630e-02 -4.96829778e-01
-7.24577904e-01 -7.65875340e-01 -6.82761431e-01 -3.04547340e-01
7.28585720e-01 -8.54869008e-01 -5.20164073e-01 3.93400818e-01
-1.31413543e+00 5.20945668e-01 -2.16163080e-02 2.77765065e-01
-6.06369793e-01 5.79081178e-01 -3.11150581e-01 -5.93679011e-01
-5.51441908e-01 -1.40524292e+00 1.37164402e+00 5.33781350e-01
3.63175303e-01 -5.41879892e-01 -2.81707570e-02 2.76268959e-01
3.47087204e-01 1.66433036e-01 5.83610833e-01 -2.23924424e-02
-1.14667404e+00 1.94990411e-02 -7.96002150e-01 1.80757329e-01
2.28201315e-01 -1.41089901e-01 -6.02851927e-01 4.10982035e-02
1.07322693e-01 1.74160823e-01 5.16464233e-01 3.61327916e-01
1.21494818e+00 1.56295687e-01 -2.35610440e-01 4.96365309e-01
1.40109956e+00 1.79750994e-01 8.04400861e-01 5.24568200e-01
8.14571261e-01 3.46841305e-01 1.22606051e+00 3.78495246e-01
5.53523183e-01 8.68527591e-01 5.89559138e-01 -1.36199892e-01
-4.65152338e-02 -4.23706859e-01 5.21117449e-02 1.03448331e+00
6.10635094e-02 4.15157586e-01 -9.91638362e-01 7.51342058e-01
-1.59243786e+00 -7.02780962e-01 -1.58646211e-01 2.06238818e+00
5.96987724e-01 3.04784387e-01 -1.08285611e-02 2.01754689e-01
8.65628004e-01 1.31784990e-01 -5.18915474e-01 -9.34599563e-02
-1.54563963e-01 2.90993840e-01 2.78199136e-01 2.26386815e-01
-1.06751478e+00 1.00601399e+00 6.13669538e+00 9.17495191e-01
-1.19559157e+00 1.09413445e-01 3.45824480e-01 6.10661656e-02
-5.11772692e-01 -2.49874517e-02 -4.72152293e-01 7.66189754e-01
3.74303460e-01 -3.32893938e-01 1.56217486e-01 9.24370766e-01
2.71427006e-01 -8.83035362e-02 -6.15349948e-01 1.25272775e+00
5.84769845e-02 -1.31272471e+00 3.33276421e-01 -1.46977842e-01
8.70871365e-01 8.08692724e-02 4.57318693e-01 -1.16381332e-01
8.25352892e-02 -6.77438021e-01 7.12383628e-01 8.18880558e-01
6.66196823e-01 -1.03807580e+00 6.86684012e-01 -3.13108414e-02
-1.34274030e+00 9.09554362e-02 -7.10328877e-01 9.27696154e-02
2.26439893e-01 6.26005709e-01 -6.31138563e-01 8.22301805e-01
9.53658640e-01 8.10425341e-01 -9.23522353e-01 1.88555253e+00
-2.58420825e-01 3.20703357e-01 -1.31222904e-01 -1.04971938e-01
2.96794504e-01 -3.04552734e-01 8.65712345e-01 5.59369266e-01
5.29299498e-01 3.22738737e-01 -1.00239605e-01 1.10977161e+00
2.53200173e-01 2.36415997e-01 -2.26140738e-01 3.28279614e-01
5.93130887e-01 1.13999081e+00 -6.15643024e-01 -2.04421267e-01
-4.13296312e-01 9.63880897e-01 1.53873190e-01 1.29955143e-01
-5.70761740e-01 -5.90720594e-01 8.19405556e-01 -9.27791819e-02
4.17684495e-01 -2.23337278e-01 -6.27858877e-01 -1.22125924e+00
2.18651891e-01 -4.99161810e-01 -3.19056064e-02 -1.39766622e+00
-1.24662852e+00 4.65259045e-01 -2.43166275e-02 -1.97450757e+00
2.40372092e-01 -2.45358527e-01 -8.76464128e-01 1.16881263e+00
-1.98644423e+00 -1.04833424e+00 -6.39330685e-01 4.07568127e-01
5.61864257e-01 1.16678486e-02 3.19893301e-01 1.30078629e-01
-1.95384428e-01 2.17638642e-01 -9.11515877e-02 -2.18410671e-01
3.34431320e-01 -1.13599110e+00 6.31518066e-01 9.62656081e-01
-5.61193191e-02 5.70143580e-01 6.48090243e-01 -6.19388700e-01
-1.14383245e+00 -9.75418031e-01 7.26665020e-01 -5.94185293e-01
4.18208092e-01 5.30875735e-02 -1.19608223e+00 -2.30696172e-01
-1.83261916e-01 3.11840177e-02 2.79974699e-01 -3.32392752e-01
-2.43387595e-02 -2.96551228e-01 -1.19071078e+00 6.38096690e-01
8.87875259e-01 -5.01478612e-01 -9.71853852e-01 9.55908746e-03
1.14947748e+00 -4.06562537e-01 -6.74357295e-01 5.96376002e-01
2.35264167e-01 -1.02695298e+00 1.12496746e+00 8.89520422e-02
5.33142269e-01 -7.93662190e-01 -4.15957086e-02 -1.39859557e+00
-6.15285814e-01 -1.21351264e-01 6.11352026e-02 9.80595112e-01
1.83386788e-01 -3.18424195e-01 9.70259786e-01 4.60694522e-01
-5.01293123e-01 -7.10193098e-01 -8.70281339e-01 -5.65791786e-01
-1.85825169e-01 -3.20118666e-01 1.20740187e+00 5.97508490e-01
1.59404520e-03 5.40095456e-02 3.89759168e-02 4.65651780e-01
5.85390568e-01 4.89881098e-01 4.98882860e-01 -1.46279585e+00
2.67748147e-01 -6.39508426e-01 -7.91679025e-01 -1.15448606e+00
-1.58729374e-01 -7.06624925e-01 8.34843218e-02 -1.73635828e+00
1.04920387e-01 -6.07846916e-01 -8.18281472e-01 -8.98268223e-02
-5.68302095e-01 3.82874995e-01 3.61117244e-01 5.84992766e-01
-9.49846268e-01 1.18145537e+00 1.55404937e+00 -2.19776973e-01
-2.91561544e-01 9.26581174e-02 -7.38652527e-01 4.36984032e-01
7.17306077e-01 -1.52049467e-01 -2.82975197e-01 -4.61012334e-01
-6.07458269e-03 3.75184864e-02 4.19640005e-01 -1.55529344e+00
1.62124529e-01 -3.71687740e-01 3.79019111e-01 -1.24821985e+00
5.02877414e-01 -7.61432409e-01 -2.81844139e-01 2.22324386e-01
6.57842085e-02 2.21182153e-01 1.82928387e-02 5.45641124e-01
-6.03534520e-01 2.46386416e-02 6.26958847e-01 3.95930707e-02
-1.02712047e+00 5.57176888e-01 3.72539490e-01 -2.48822153e-01
9.25432682e-01 -3.51862639e-01 -1.83618098e-01 -1.91872716e-01
-1.77958369e-01 4.09356445e-01 8.37292254e-01 6.11237109e-01
1.41520894e+00 -1.75240433e+00 -6.16228878e-01 9.12681967e-02
5.66394150e-01 4.59537506e-02 3.04326892e-01 5.75359106e-01
-8.29870164e-01 6.14979923e-01 -5.36268830e-01 -8.00522685e-01
-1.00606763e+00 6.35610402e-01 -3.86912189e-02 4.72442061e-02
-4.77594197e-01 7.27493465e-01 2.14153379e-01 -1.76636562e-01
-2.71759033e-01 -4.79175091e-01 -4.59076613e-01 -5.19758821e-01
5.49294233e-01 3.83798480e-01 3.32389653e-01 -8.92260373e-01
-5.34938693e-01 8.16574872e-01 1.79706365e-01 -1.51382759e-01
1.12232769e+00 -3.09707612e-01 -7.44550824e-02 1.81734756e-01
9.68235373e-01 -1.01460107e-01 -1.38529503e+00 -4.31122094e-01
1.62626997e-01 -1.10303676e+00 1.58370107e-01 -6.47430360e-01
-1.04646218e+00 1.16263676e+00 8.46798003e-01 3.82787287e-02
1.40414858e+00 1.57293223e-03 1.03927267e+00 -1.06710255e-01
7.11094916e-01 -8.41689229e-01 3.05275649e-01 3.00356537e-01
1.25805199e+00 -1.36599660e+00 1.07134357e-01 -7.49937057e-01
-8.83813739e-01 6.25192702e-01 6.86772823e-01 -4.52356309e-01
5.20594716e-01 -7.20378757e-01 -1.20252728e-01 -3.40534240e-01
-4.77425486e-01 -3.57250005e-01 8.29600990e-01 7.25005150e-01
-1.78052768e-01 1.86814040e-01 -2.88019359e-01 4.56976175e-01
-3.93698871e-01 -1.00006275e-01 2.41197646e-01 8.35907221e-01
-7.86190093e-01 -8.35791230e-01 -4.84224051e-01 4.67139423e-01
-1.47790208e-01 -1.11530624e-01 -3.71944338e-01 2.54546911e-01
-6.52453825e-02 1.05413616e+00 1.79779500e-01 -5.98357618e-01
4.82576042e-01 -3.82292986e-01 2.27241173e-01 -7.69237399e-01
-3.04442316e-01 -1.07172184e-01 -6.10331059e-01 -7.62460768e-01
-4.89966154e-01 -6.31030917e-01 -1.32451725e+00 -5.33896685e-02
-3.24385345e-01 3.35740179e-01 6.49010062e-01 8.02072346e-01
6.61677420e-01 9.57140550e-02 8.99473727e-01 -1.18997633e+00
-7.03115389e-02 -7.83343911e-01 -4.94002610e-01 4.54988748e-01
3.42206001e-01 -1.14642143e+00 -2.16153756e-01 -4.09580708e-01] | [9.754021644592285, -0.6240609288215637] |
9a3cb7c1-76a1-4ac4-8ff6-0b31bcd6edf6 | vrconvmf-visual-recurrent-convolutional | 2202.10241 | null | https://arxiv.org/abs/2202.10241v1 | https://arxiv.org/pdf/2202.10241v1.pdf | VRConvMF: Visual Recurrent Convolutional Matrix Factorization for Movie Recommendation | Sparsity of user-to-item rating data becomes one of challenging issues in the recommender systems, which severely deteriorates the recommendation performance. Fortunately, context-aware recommender systems can alleviate the sparsity problem by making use of some auxiliary information, such as the information of both the users and items. In particular, the visual information of items, such as the movie poster, can be considered as the supplement for item description documents, which helps to obtain more item features. In this paper, we focus on movie recommender system and propose a probabilistic matrix factorization based recommendation scheme called visual recurrent convolutional matrix factorization (VRConvMF), which utilizes the textual and multi-level visual features extracted from the descriptive texts and posters respectively. We implement the proposed VRConvMF and conduct extensive experiments on three commonly used real world datasets to validate its effectiveness. The experimental results illustrate that the proposed VRConvMF outperforms the existing schemes. | ['Feng Xia', 'Nan Jiang', 'Kai Lin', 'Zhe Li', 'Honglong Chen', 'Zhu Wang'] | 2022-02-16 | null | null | null | null | ['movie-recommendation'] | ['miscellaneous'] | [-2.69687861e-01 -8.19503427e-01 -2.24689603e-01 -2.30278596e-01
-2.06482023e-01 -6.46260798e-01 3.21224779e-01 -2.42914129e-02
-1.01063676e-01 3.21799010e-01 8.77529204e-01 -1.88960716e-01
-4.01429385e-01 -5.32177091e-01 -1.94754362e-01 -7.27408469e-01
3.51813048e-01 -2.21337691e-01 -6.83968561e-03 -3.45705003e-01
4.89331245e-01 1.21898077e-01 -1.55133390e+00 8.90156925e-01
5.54311156e-01 1.00394213e+00 6.15772367e-01 1.72869816e-01
-3.48339081e-01 5.72345376e-01 -3.08688462e-01 -1.92322105e-01
1.11737773e-01 -7.51811638e-02 -6.87507465e-02 4.07618940e-01
2.36780971e-01 -5.68582416e-01 -5.49438417e-01 8.30636621e-01
3.19849521e-01 7.26616919e-01 7.59124935e-01 -9.16483045e-01
-1.42018640e+00 7.73108959e-01 -8.08738947e-01 4.76965874e-01
3.21595550e-01 -3.16287249e-01 1.17110205e+00 -1.30164850e+00
4.25472796e-01 1.28640449e+00 3.71645778e-01 3.32431108e-01
-4.93073821e-01 -5.92207730e-01 8.09665740e-01 4.24156934e-01
-1.46739900e+00 -2.02196434e-01 7.12339997e-01 -4.69010890e-01
6.82366490e-01 2.78613687e-01 6.14783704e-01 8.40142667e-01
-1.68715626e-01 1.00280881e+00 9.46094453e-01 -7.22960681e-02
-1.04352042e-01 3.74816269e-01 3.92417282e-01 4.34737980e-01
1.98506102e-01 -1.68811724e-01 -2.35554934e-01 -1.35034621e-01
1.07694280e+00 9.29541409e-01 -5.61172009e-01 -1.88966691e-01
-1.03452313e+00 8.67238045e-01 5.65992713e-01 3.37115586e-01
-3.23292732e-01 -3.01319212e-01 4.16793197e-01 1.91417158e-01
2.65891999e-01 -1.28117442e-01 -2.38039166e-01 2.34251738e-01
-6.75090969e-01 1.47185745e-02 2.12981358e-01 8.76621962e-01
4.20300692e-01 3.08433354e-01 -3.26272994e-01 1.03567731e+00
8.50235045e-01 4.18298781e-01 7.69468486e-01 -4.41226274e-01
5.49393237e-01 6.02704048e-01 3.10052246e-01 -1.37838519e+00
-2.68344045e-01 -7.19524920e-01 -8.27641249e-01 -4.40973312e-01
-1.89683795e-01 -6.81468099e-02 -7.86689699e-01 1.25181639e+00
2.99520463e-01 3.20713967e-01 1.10717425e-02 1.48314261e+00
1.68443906e+00 1.10334802e+00 -1.15510397e-01 -5.94261706e-01
1.34611285e+00 -1.09625494e+00 -7.51271248e-01 1.37522221e-01
4.64784920e-01 -1.02545595e+00 1.10970449e+00 6.76154494e-01
-6.02239370e-01 -9.55664992e-01 -1.01685858e+00 9.46709514e-02
-6.43118992e-02 8.32646251e-01 8.68513107e-01 3.85756761e-01
-5.67879140e-01 3.10555041e-01 -4.13901538e-01 -1.28037080e-01
2.64634907e-01 2.40743890e-01 -3.99968147e-01 -4.76406634e-01
-1.06622958e+00 2.94266641e-01 1.15474656e-01 4.13970947e-01
-8.34829986e-01 -2.97905594e-01 -6.00782275e-01 2.86402702e-01
3.07022274e-01 -3.67554039e-01 8.87979925e-01 -8.44771087e-01
-1.17807388e+00 2.34244000e-02 -7.97477812e-02 2.78306723e-01
-1.16071656e-01 -3.01732481e-01 -7.43039548e-01 -1.21833123e-01
-2.56503701e-01 -3.04708555e-02 8.24128211e-01 -1.19448602e+00
-6.45572484e-01 -3.95590425e-01 3.94429594e-01 5.15620172e-01
-7.87757814e-01 2.16227956e-03 -6.71338499e-01 -9.20991957e-01
8.11435562e-03 -7.86275387e-01 -2.86028832e-01 -3.04502964e-01
-1.73775956e-01 -3.83433759e-01 9.09970164e-01 -5.43603718e-01
1.36576784e+00 -2.36396408e+00 1.54350400e-01 2.10536808e-01
2.75098562e-01 3.42008621e-01 -2.82759935e-01 4.60255086e-01
-7.47778360e-03 -3.05186719e-01 4.59491700e-01 -2.53459692e-01
-4.09246534e-01 2.62434125e-01 -5.64369321e-01 3.96355331e-01
-5.16214848e-01 6.82519495e-01 -8.62858295e-01 -3.37398440e-01
3.48549247e-01 7.49268651e-01 -5.89058161e-01 2.07272232e-01
-2.01716926e-03 3.63394946e-01 -8.22620034e-01 3.14371586e-01
8.69851232e-01 -5.19811094e-01 6.96262345e-02 -5.27787805e-01
-1.14493541e-01 1.52663887e-01 -1.50429308e+00 1.73028314e+00
-4.20293510e-01 9.76537392e-02 -9.54332426e-02 -7.95353055e-01
9.27698255e-01 2.29161248e-01 2.72710860e-01 -5.86378813e-01
2.64898121e-01 -3.71501952e-01 -1.02540910e-01 -5.46828985e-01
8.94317150e-01 -6.79071173e-02 1.68447867e-01 6.33679152e-01
7.18851164e-02 7.29686916e-01 1.85176153e-02 8.19315195e-01
6.03407383e-01 1.62065551e-01 1.55036867e-01 1.87535060e-03
7.13164628e-01 -3.02914351e-01 6.35267913e-01 3.75780910e-01
3.42753500e-01 5.11081100e-01 -9.30622593e-02 -2.08206221e-01
-6.85600638e-01 -8.50551486e-01 8.25143233e-02 1.14988697e+00
3.70083392e-01 -8.74493837e-01 -1.41949385e-01 -9.05330956e-01
-8.24817941e-02 2.79872149e-01 -6.68744624e-01 -5.43258898e-02
-1.56448692e-01 -6.29725277e-01 -4.10906225e-01 7.57420301e-01
-5.37761562e-02 -9.00701880e-01 1.72577277e-01 1.44288242e-01
-3.85600813e-02 -7.34434664e-01 -7.87234485e-01 -2.70415157e-01
-8.92462790e-01 -7.69016743e-01 -8.23062956e-01 -7.36505985e-01
7.42316723e-01 1.51417196e+00 6.60301745e-01 5.19360542e-01
1.90272376e-01 2.21523285e-01 -9.87971663e-01 1.69235915e-01
3.44583333e-01 -2.83300966e-01 3.24943632e-01 3.77463192e-01
2.44308263e-01 -5.65543652e-01 -9.29146707e-01 5.57291806e-01
-7.94077337e-01 3.31285857e-02 5.96038163e-01 7.69988000e-01
6.99821532e-01 1.66374311e-01 6.10007524e-01 -1.12109709e+00
8.98217380e-01 -8.01155329e-01 -2.34435767e-01 3.22367400e-01
-5.59227347e-01 -2.55577326e-01 1.05288661e+00 -7.32131958e-01
-1.26338363e+00 -1.14160433e-01 -4.95889001e-02 -6.76624417e-01
3.94398980e-02 9.39172983e-01 -2.44340479e-01 1.31669745e-01
4.41989243e-01 3.13503683e-01 -6.24556839e-01 -1.14744163e+00
6.42959356e-01 1.02940440e+00 3.10296685e-01 -1.95269763e-01
4.94922310e-01 3.57641786e-01 -4.03022408e-01 -5.67494690e-01
-1.02494371e+00 -9.60211337e-01 -2.35452875e-01 -3.67261499e-01
3.57665837e-01 -1.15086746e+00 -5.80241084e-01 -1.82842627e-01
-8.50644767e-01 5.18273771e-01 -5.80575019e-02 7.78446138e-01
1.10236257e-01 7.10294843e-01 -7.89898455e-01 -7.67615914e-01
-4.38520700e-01 -1.04538834e+00 7.86710978e-01 6.15389347e-01
3.01032513e-01 -7.25024164e-01 8.71438533e-02 4.16017592e-01
2.33191133e-01 -4.37794894e-01 6.94199324e-01 -6.18228137e-01
-2.95392156e-01 1.10087525e-02 -5.84190190e-01 2.79385448e-01
3.43120933e-01 -6.97075054e-02 -6.77010298e-01 -4.62854564e-01
-7.63595626e-02 -1.17992729e-01 1.08673608e+00 1.88009679e-01
1.11956906e+00 -3.92831087e-01 -3.18254739e-01 4.44112629e-01
1.36793685e+00 1.58418328e-01 4.28717226e-01 -1.54327676e-01
1.28268874e+00 2.42267564e-01 9.90394175e-01 8.90538096e-01
3.89616430e-01 6.09342694e-01 3.60383302e-01 1.34843364e-01
-7.35156536e-02 -4.98766690e-01 3.94066840e-01 1.58174312e+00
-2.79888362e-01 -3.67038697e-01 -5.06497510e-02 3.53192061e-01
-2.25163746e+00 -9.20401633e-01 -4.87629473e-01 1.97725558e+00
2.87364721e-01 -1.50082350e-01 1.04653493e-01 1.09568805e-01
7.07171917e-01 1.29886106e-01 -3.56581658e-01 -2.50917286e-01
-7.77549390e-03 -2.14165270e-01 -6.82169013e-03 3.99751626e-02
-1.02494442e+00 6.97521389e-01 5.54339027e+00 1.06807852e+00
-1.05722976e+00 2.60882020e-01 1.25097409e-01 -3.71596277e-01
-6.92048669e-01 -1.16722167e-01 -7.98203826e-01 5.81405818e-01
5.71349382e-01 1.05158351e-02 5.27292073e-01 1.10665011e+00
4.60818797e-01 3.00885260e-01 -6.09426856e-01 1.18444514e+00
4.35105979e-01 -1.42392755e+00 4.90215749e-01 4.56164889e-02
8.40106606e-01 -2.13931426e-01 4.31721240e-01 3.59111965e-01
3.33429068e-01 -8.08413744e-01 3.66660088e-01 6.82441175e-01
6.54949129e-01 -8.95175755e-01 7.98460245e-01 3.08545589e-01
-1.60057604e+00 -1.77142128e-01 -1.07046664e+00 -2.23171845e-01
1.87570125e-01 5.41403472e-01 -2.40751043e-01 7.25373983e-01
7.41020918e-01 1.57109189e+00 -4.29062963e-01 1.24591911e+00
-2.86205590e-01 6.11791611e-01 5.88683970e-03 -1.00112051e-01
2.64933974e-01 -3.90403569e-01 2.45392010e-01 9.51224446e-01
5.02915263e-01 5.05264223e-01 5.17330289e-01 2.07766294e-01
-1.08650833e-01 8.80858123e-01 -2.65933543e-01 -1.29479453e-01
2.45290056e-01 1.80578446e+00 -6.99609637e-01 -2.09166616e-01
-7.73540974e-01 7.67722845e-01 3.03608894e-01 2.59271383e-01
-6.30191743e-01 -1.26341179e-01 3.76873970e-01 -8.95998776e-02
8.26588809e-01 -9.69552919e-02 1.96374059e-01 -1.57484090e+00
-1.35933846e-01 -7.41110384e-01 5.71452379e-01 -1.06471181e+00
-1.45802867e+00 4.96900618e-01 -4.38435584e-01 -1.72593427e+00
3.04715276e-01 -4.60500866e-01 -6.12622082e-01 7.54042923e-01
-1.25622547e+00 -1.44099700e+00 -4.35789138e-01 1.02933133e+00
6.23133481e-01 -4.41009939e-01 6.24024808e-01 5.82300305e-01
-5.80178678e-01 7.03960776e-01 5.59837580e-01 -1.63591638e-01
7.01753259e-01 -8.20880413e-01 -1.15351871e-01 6.49494469e-01
5.25165141e-01 1.03993297e+00 3.91933858e-01 -6.41751468e-01
-1.61963153e+00 -1.11088157e+00 3.11045647e-01 -1.35360599e-01
4.94613826e-01 -5.14201149e-02 -8.08219075e-01 5.90172470e-01
-5.27253114e-02 8.11536536e-02 1.15456307e+00 3.71793181e-01
-6.04333460e-01 -7.33660460e-02 -8.33250880e-01 4.30216581e-01
1.01435184e+00 -3.62532377e-01 -7.66499400e-01 3.97000581e-01
8.30752611e-01 -6.87045082e-02 -7.48106599e-01 6.55826405e-02
8.17552328e-01 -7.30175734e-01 1.05004048e+00 -8.68321717e-01
5.36979556e-01 -7.14961529e-01 -3.88549060e-01 -1.35587502e+00
-9.92350101e-01 4.78730649e-02 -3.53532195e-01 1.42593610e+00
8.28278959e-02 -3.62105928e-02 6.75219536e-01 1.22489378e-01
-2.17313722e-01 -7.73856878e-01 -4.39688146e-01 -3.46945643e-01
-3.46499950e-01 -2.37638578e-01 6.43397212e-01 1.00723803e+00
2.22331911e-01 7.88967669e-01 -1.10326850e+00 6.74726740e-02
1.55499995e-01 7.28097260e-01 5.86134791e-01 -1.23103070e+00
-6.04911029e-01 -3.38089168e-02 -2.55299032e-01 -1.54660726e+00
-1.97433084e-01 -9.20726299e-01 -1.99822992e-01 -1.75546134e+00
5.67211270e-01 -4.72839385e-01 -9.81816828e-01 1.91117093e-01
-2.28989810e-01 4.54271466e-01 1.48808643e-01 5.44893980e-01
-1.04285586e+00 6.25505209e-01 1.50656557e+00 -1.48724630e-01
-1.21138901e-01 -5.46428300e-02 -1.14819038e+00 6.58702493e-01
3.89751554e-01 -3.25900346e-01 -7.88706601e-01 -5.26937485e-01
3.96494746e-01 2.39367828e-01 -3.05685848e-01 -4.43685681e-01
1.72258109e-01 -3.40173602e-01 7.30797052e-01 -7.98948467e-01
3.74032140e-01 -7.39200532e-01 1.08463362e-01 2.50817444e-02
-2.74027258e-01 2.05228463e-01 -1.27992019e-01 9.62765694e-01
-9.17733684e-02 -2.91189909e-01 3.18840444e-01 -2.00208887e-01
-9.02317584e-01 5.37266731e-01 -3.63153100e-01 -4.23257768e-01
6.32908821e-01 5.79952113e-02 -4.51463193e-01 -4.12500530e-01
-8.19873989e-01 2.54825503e-01 3.41035098e-01 6.50920749e-01
1.16145861e+00 -1.63032937e+00 -5.94809890e-01 1.54600668e-04
4.51125473e-01 -6.16585791e-01 1.14215076e+00 6.42952025e-01
-4.09676917e-02 4.58550841e-01 -1.90735072e-01 -1.89108208e-01
-1.46962488e+00 1.24087763e+00 -3.64586174e-01 -1.71335965e-01
-6.60476863e-01 9.29031372e-01 5.23210227e-01 -8.09579939e-02
1.49201110e-01 -2.32071858e-02 -1.28313720e+00 9.23485681e-02
1.03733408e+00 1.27906010e-01 -1.29443452e-01 -1.03195572e+00
-3.31201226e-01 6.03173137e-01 -6.83643401e-01 2.31566444e-01
1.41806030e+00 -5.32501340e-01 1.34493515e-01 2.78591216e-01
9.80722487e-01 4.09986019e-01 -8.63980055e-01 -6.76487446e-01
-7.26968229e-01 -8.15677762e-01 4.74284500e-01 -5.41347623e-01
-1.67833817e+00 9.60108459e-01 4.92655516e-01 6.90413043e-02
9.71008062e-01 -2.41936281e-01 1.03157842e+00 3.79453719e-01
4.24828082e-01 -7.50083029e-01 2.79536098e-02 1.91765592e-01
8.18824112e-01 -1.10857379e+00 4.99501705e-01 -3.62566233e-01
-9.09764469e-01 1.01744759e+00 6.53741300e-01 -5.11182129e-01
1.02581573e+00 -3.08087766e-01 2.32425462e-02 -6.43713325e-02
-8.03879440e-01 -1.63431555e-01 5.94519377e-01 4.37965214e-01
4.40420330e-01 1.57635435e-01 -5.03569484e-01 1.52888370e+00
1.72146261e-01 4.71905731e-02 4.79216367e-01 7.18700469e-01
-5.37785411e-01 -1.17434287e+00 -2.57712096e-01 9.49962020e-01
-4.14942056e-01 -2.12692425e-01 -1.28575966e-01 6.24320693e-02
2.77936578e-01 1.17250574e+00 -1.43108144e-01 -9.38289583e-01
2.18485966e-01 -5.97013474e-01 2.16403514e-01 -8.26832175e-01
-6.54040933e-01 3.79496008e-01 -2.52018999e-02 -3.44964683e-01
-4.59601045e-01 -6.45220101e-01 -1.09444010e+00 -2.47893304e-01
-6.44358099e-01 5.93934357e-01 3.70199889e-01 1.05803168e+00
3.29691648e-01 4.48142499e-01 8.86006773e-01 -7.05567420e-01
-1.62110388e-01 -9.81889129e-01 -9.45191681e-01 4.82077718e-01
7.37273023e-02 -1.06479764e+00 -1.83372974e-01 -1.68646932e-01] | [10.193473815917969, 5.572422981262207] |
db2812b7-8954-403b-a022-f6ec323d6b03 | unsupervised-statistical-feature-guided | 2306.05285 | null | https://arxiv.org/abs/2306.05285v1 | https://arxiv.org/pdf/2306.05285v1.pdf | Unsupervised Statistical Feature-Guided Diffusion Model for Sensor-based Human Activity Recognition | Recognizing human activities from sensor data is a vital task in various domains, but obtaining diverse and labeled sensor data remains challenging and costly. In this paper, we propose an unsupervised statistical feature-guided diffusion model for sensor-based human activity recognition. The proposed method aims to generate synthetic time-series sensor data without relying on labeled data, addressing the scarcity and annotation difficulties associated with real-world sensor data. By conditioning the diffusion model on statistical information such as mean, standard deviation, Z-score, and skewness, we generate diverse and representative synthetic sensor data. We conducted experiments on public human activity recognition datasets and compared the proposed method to conventional oversampling methods and state-of-the-art generative adversarial network methods. The experimental results demonstrate that the proposed method can improve the performance of human activity recognition and outperform existing techniques. | ['Paul Lukowicz', 'Stephan Sigg', 'Sungho Suh', 'Vitor Fortes Rey', 'Si Zuo'] | 2023-05-30 | null | null | null | null | ['human-activity-recognition', 'human-activity-recognition'] | ['computer-vision', 'time-series'] | [ 8.46662402e-01 -1.03726119e-01 -8.00113305e-02 -2.93682456e-01
-8.34070265e-01 -4.45369065e-01 7.44733751e-01 -1.01616040e-01
-3.19862276e-01 1.09075320e+00 4.65205342e-01 2.03716993e-01
-9.19018239e-02 -8.75916421e-01 -6.05908394e-01 -9.58045483e-01
-8.44189227e-02 1.74927220e-01 -1.11268096e-01 2.31230706e-01
-7.64386728e-02 4.07481313e-01 -1.38397181e+00 -5.02092689e-02
9.08566773e-01 1.26410592e+00 -4.96418923e-01 5.73458612e-01
2.57565737e-01 9.28212047e-01 -1.15844023e+00 -3.43115181e-02
2.57367194e-01 -8.58336389e-01 -9.92392227e-02 4.66331720e-01
-6.78153560e-02 -1.46145597e-01 -5.37376165e-01 8.85621011e-01
6.81475043e-01 2.29539707e-01 8.76329064e-01 -1.70239806e+00
-6.35946035e-01 3.73278558e-01 -4.46674496e-01 2.16048419e-01
7.82178402e-01 3.19407344e-01 3.12125534e-01 -5.07921755e-01
3.57680440e-01 7.13751078e-01 7.91710794e-01 6.72552943e-01
-1.09273386e+00 -8.98751378e-01 4.60216925e-02 -1.41204804e-01
-1.50041151e+00 -3.52129221e-01 1.17664909e+00 -4.85674381e-01
5.18158078e-01 -1.67390071e-02 9.06794369e-01 2.04789925e+00
4.29043174e-03 9.37679291e-01 1.16789401e+00 1.01540104e-01
7.66421080e-01 -4.54689890e-01 -2.35620648e-01 4.34653133e-01
6.90486968e-01 1.05613045e-01 -6.91787481e-01 -4.74871606e-01
8.19978535e-01 5.37799656e-01 -1.87382162e-01 -4.38964397e-01
-1.57154202e+00 6.01785243e-01 -2.65979674e-02 1.47739545e-01
-8.98823440e-01 1.90537259e-01 3.11996877e-01 -1.36932969e-01
4.56259817e-01 -8.49790052e-02 -1.24211229e-01 -5.50820649e-01
-1.11903322e+00 3.03279310e-01 7.92533278e-01 9.04833376e-01
1.81167781e-01 5.53897262e-01 -3.94843698e-01 5.68741918e-01
1.76440388e-01 1.11589038e+00 7.83528507e-01 -8.80356908e-01
6.11706197e-01 6.01615310e-01 3.44202548e-01 -1.01707923e+00
-2.30263621e-01 -2.24224657e-01 -1.32475126e+00 -1.76676422e-01
6.72385693e-01 -6.62659347e-01 -9.91895139e-01 1.72467518e+00
2.61685073e-01 5.81252694e-01 2.80449778e-01 6.98899508e-01
4.02051270e-01 5.92489600e-01 4.03989792e-01 -3.62072378e-01
8.24724317e-01 -5.60336173e-01 -8.16040576e-01 -2.95711368e-01
9.38174427e-02 -1.55910909e-01 8.38863254e-01 4.97064739e-01
-6.82397604e-01 -5.49417615e-01 -1.18591285e+00 7.18966901e-01
-1.27919450e-01 -6.30504116e-02 7.17516899e-01 9.30206478e-01
-3.30142766e-01 5.57338178e-01 -1.29708338e+00 -3.00364882e-01
7.54582047e-01 5.15311584e-02 -3.39409053e-01 -5.42791001e-02
-1.11834300e+00 1.76502168e-01 2.81640887e-01 -7.16793388e-02
-1.27444041e+00 -3.75400931e-01 -1.09304404e+00 -2.47431308e-01
3.22453111e-01 -5.78460395e-01 1.04083562e+00 -7.31003106e-01
-1.42281449e+00 3.30235660e-01 -9.15724784e-02 -6.53184533e-01
6.62533581e-01 -1.39849827e-01 -7.95343757e-01 -3.42425257e-02
-4.05435497e-03 1.59043655e-01 9.78588700e-01 -1.01434374e+00
-3.86367440e-01 -5.30175745e-01 -5.93693733e-01 -1.49718046e-01
-3.38242173e-01 -4.95023519e-01 2.35471898e-03 -1.13633716e+00
4.63904403e-02 -7.89138138e-01 -2.64494658e-01 -1.25245273e-01
-3.59947950e-01 1.27715066e-01 7.62355208e-01 -7.07335770e-01
1.02800786e+00 -1.99105775e+00 -2.60748982e-01 4.41727310e-01
-7.67089799e-02 1.89468533e-01 1.46236578e-02 4.03949946e-01
2.34093085e-01 -2.97640264e-01 -6.66945815e-01 -8.01577941e-02
1.43799543e-01 3.04049999e-01 -1.45746142e-01 5.49792409e-01
1.22204281e-01 9.94278252e-01 -1.13952363e+00 -4.23346162e-01
2.00408757e-01 7.33473241e-01 -1.00202672e-01 2.74752438e-01
-1.81531399e-01 8.39692354e-01 -6.04819834e-01 1.08962834e+00
3.16127658e-01 -1.89791515e-01 4.96396273e-02 3.74779589e-02
6.03195488e-01 -9.58811790e-02 -1.47625339e+00 1.79645216e+00
-8.11852068e-02 2.48270735e-01 -5.39404511e-01 -1.14110112e+00
1.13546038e+00 4.14212763e-01 1.07147825e+00 -7.04815567e-01
1.48577645e-01 -2.86678504e-02 -3.38958502e-01 -4.47531611e-01
6.56029508e-02 1.42936349e-01 -6.58529460e-01 6.32404268e-01
-8.59371871e-02 2.24800687e-02 2.21661210e-01 -1.92610845e-01
1.54415560e+00 2.89241463e-01 3.52547258e-01 2.81729877e-01
3.88919950e-01 -1.65269479e-01 8.17690432e-01 7.40614474e-01
-5.35816371e-01 4.75534916e-01 2.30106622e-01 -1.90982521e-01
-9.55361009e-01 -1.39077282e+00 3.34222555e-01 5.43086946e-01
-1.02033183e-01 -1.44924998e-01 -1.00209534e+00 -8.81861508e-01
4.89888005e-02 5.81666529e-01 -8.08363855e-01 -2.98904419e-01
-3.89019400e-01 -1.00093031e+00 1.14595485e+00 8.90622616e-01
9.07297194e-01 -1.21707046e+00 -6.19406998e-01 4.22572196e-01
-2.51009852e-01 -1.13337219e+00 -4.90176082e-01 -8.37875232e-02
-8.37696075e-01 -1.17112029e+00 -9.29030001e-01 -3.45103323e-01
7.41790712e-01 -2.81914752e-02 8.82644832e-01 -7.98498034e-01
-1.92048192e-01 5.57495832e-01 -3.08462113e-01 -5.93029559e-01
-2.94890523e-01 -4.95937988e-02 2.31644541e-01 5.36224425e-01
6.97604895e-01 -8.65479231e-01 -8.65583479e-01 4.89429593e-01
-1.21780181e+00 -4.53953803e-01 5.31462610e-01 6.99836493e-01
8.21149349e-01 2.17312723e-01 8.57275963e-01 -7.38672614e-01
9.56531167e-01 -6.51522875e-01 -3.10296714e-01 3.20168324e-02
-7.17118204e-01 6.40461966e-02 6.00152671e-01 -1.03969097e+00
-1.00903118e+00 4.38211679e-01 1.50847450e-01 -3.35706025e-01
-4.46069181e-01 2.59764850e-01 -4.57173526e-01 2.38317341e-01
8.17854464e-01 6.89457715e-01 -6.59320652e-02 -2.32208297e-01
1.85475573e-01 6.95951819e-01 8.57376635e-01 -5.59121668e-01
7.27935910e-01 7.14141071e-01 -1.22469530e-01 -6.30844891e-01
-7.24014163e-01 -2.40899459e-01 -3.47216755e-01 -2.58211195e-01
7.19101667e-01 -1.02659857e+00 -3.80174458e-01 9.74101365e-01
-7.29306519e-01 -3.70088488e-01 -7.67434239e-01 4.86223131e-01
-7.18485236e-01 2.87785023e-01 -2.53102005e-01 -1.05730760e+00
-4.72113103e-01 -5.32978177e-01 1.11668229e+00 2.80099869e-01
-5.60472131e-01 -9.55733418e-01 3.62313241e-01 5.37540317e-01
4.84601915e-01 1.10477507e+00 -6.66523874e-02 -8.81448269e-01
-3.23895305e-01 -6.62983298e-01 3.31882805e-01 5.20115554e-01
5.53784668e-01 -4.15272325e-01 -7.97859251e-01 -1.42225102e-01
3.59709188e-02 -4.08631742e-01 3.57172310e-01 1.72263443e-01
1.28135622e+00 -3.21520090e-01 -2.78485894e-01 4.50882047e-01
9.76258397e-01 3.72892082e-01 7.21938610e-01 -3.92330140e-02
6.56737447e-01 8.29995871e-02 6.98878944e-01 1.03744948e+00
1.48338020e-01 1.74341768e-01 1.54992238e-01 1.72292620e-01
2.13382885e-01 -6.80113375e-01 4.77344304e-01 4.50763464e-01
-1.30289853e-01 -2.96669632e-01 -7.81257510e-01 7.51236498e-01
-1.85089600e+00 -1.37366891e+00 2.87556827e-01 2.12877035e+00
7.60633767e-01 2.67732352e-01 3.25489491e-01 7.19771981e-01
5.70115387e-01 4.25469905e-01 -1.01323152e+00 2.97008842e-01
-1.43951729e-01 4.31629211e-01 6.21501148e-01 -1.17535897e-01
-1.15968835e+00 4.00786340e-01 6.71136284e+00 6.57714963e-01
-8.14846098e-01 1.21527269e-01 3.73678535e-01 -1.73845097e-01
6.79258704e-02 -6.57732844e-01 -5.56072116e-01 8.65950346e-01
1.04628718e+00 -2.32511014e-01 4.17734116e-01 9.60886896e-01
3.32496971e-01 -7.25994408e-02 -1.07288456e+00 1.27712917e+00
3.84933203e-01 -7.54107177e-01 -9.37580615e-02 6.05113730e-02
1.02859843e+00 -2.26111770e-01 -2.28629932e-01 2.51104951e-01
5.90301633e-01 -1.10386395e+00 4.51116681e-01 9.72464561e-01
5.86869538e-01 -6.87325299e-01 7.08888113e-01 6.03342593e-01
-1.18736446e+00 2.53167637e-02 2.82395244e-01 -3.05986926e-02
3.12233835e-01 7.86295652e-01 -4.95882273e-01 2.12946787e-01
4.74074215e-01 7.57770479e-01 -4.44491208e-01 7.46234715e-01
-3.17658126e-01 8.95780802e-01 -4.53895420e-01 -2.68980145e-01
-1.91606626e-01 -1.65960476e-01 2.17807546e-01 1.04864168e+00
6.19728327e-01 -9.20923334e-03 1.38989940e-01 7.17264593e-01
-1.95786625e-01 -3.25229049e-01 -8.73087168e-01 -5.83734453e-01
4.73348647e-01 7.50045300e-01 -5.00052452e-01 -5.22174239e-01
-2.60100037e-01 1.16200507e+00 -3.05795580e-01 4.25426602e-01
-1.12691140e+00 -4.08739924e-01 4.95354146e-01 1.72128603e-01
2.37771228e-01 -3.43395084e-01 -3.30797106e-01 -1.25696445e+00
3.57010245e-01 -1.11920273e+00 4.91345584e-01 -6.44420803e-01
-1.45363307e+00 2.74928719e-01 9.03531313e-02 -1.46092486e+00
-6.70510650e-01 -1.66578457e-01 -3.63195330e-01 5.38266122e-01
-9.56455231e-01 -1.02193666e+00 -8.51457775e-01 1.08786166e+00
4.31949139e-01 -3.52893114e-01 8.75020266e-01 4.49725032e-01
-3.82905066e-01 5.07037997e-01 -6.49895659e-03 6.07436061e-01
4.22021180e-01 -9.06807065e-01 5.83521962e-01 9.81415331e-01
2.07403198e-01 3.74550551e-01 6.17222011e-01 -8.56463850e-01
-1.31116509e+00 -1.36721158e+00 5.27134538e-01 -4.32942092e-01
5.41969478e-01 -3.68356586e-01 -6.24147892e-01 5.57399988e-01
-8.36054087e-02 3.76701683e-01 9.99024272e-01 -6.19418144e-01
-1.22506216e-01 -3.04623097e-01 -1.49613011e+00 3.74342531e-01
1.34683990e+00 -4.07962352e-01 -6.73433065e-01 1.53169662e-01
1.72061279e-01 -1.45740464e-01 -1.17588830e+00 6.41983747e-01
5.79551816e-01 -5.87964237e-01 1.03208363e+00 -4.79501456e-01
-4.61793737e-03 -3.89279932e-01 -2.71223336e-01 -1.45690584e+00
8.99016410e-02 -7.47671604e-01 -7.36579120e-01 1.16192174e+00
1.35091856e-01 -6.36290967e-01 1.23735857e+00 6.88117504e-01
3.27013701e-01 -2.39286870e-01 -7.53744066e-01 -8.91445696e-01
-4.73127365e-01 -5.69816947e-01 8.37517858e-01 8.63192081e-01
-2.42081001e-01 -2.43824478e-02 -5.73045373e-01 4.53077555e-02
1.08887720e+00 -2.33915225e-01 1.13918221e+00 -1.25458562e+00
-4.29929346e-01 1.12631552e-01 -6.01287603e-01 -8.87712777e-01
-4.76058386e-02 -2.94317722e-01 -6.05691411e-03 -1.41885471e+00
-1.11536793e-01 -5.73745696e-03 -5.91961801e-01 3.73734146e-01
-3.33528258e-02 4.37067032e-01 -2.70110846e-01 -4.53118309e-02
-6.44153535e-01 7.94134796e-01 1.09418845e+00 -5.26445866e-01
-1.74628526e-01 2.26752967e-01 -5.62477648e-01 7.20425069e-01
1.01969385e+00 -6.59829259e-01 -7.93377340e-01 1.50479421e-01
-1.55526266e-01 2.78231770e-01 4.09098923e-01 -1.59255564e+00
1.30857542e-01 -4.08000231e-01 8.43555212e-01 -3.94600660e-01
3.86056215e-01 -9.58238721e-01 5.00923991e-01 5.88510633e-01
-2.57526338e-01 1.32792536e-02 -2.43845388e-01 1.03671253e+00
-2.54904985e-01 3.59961361e-01 4.72275019e-01 4.56429087e-02
-4.31384146e-01 4.67878282e-01 -4.16422486e-01 3.46185565e-01
1.23363435e+00 -2.73778439e-01 -6.26035109e-02 -6.13890529e-01
-8.66896868e-01 -1.66473314e-01 3.56038451e-01 5.39415002e-01
7.85485268e-01 -1.72168660e+00 -5.47951519e-01 5.11813819e-01
4.37097549e-01 2.75321044e-02 -8.04777741e-02 5.38038731e-01
-2.29329273e-01 -7.08853304e-02 -2.87005305e-01 -3.99145663e-01
-7.83833563e-01 4.67754811e-01 2.11506501e-01 -3.29316318e-01
-3.60241324e-01 1.76561296e-01 -5.96336722e-01 -2.81659096e-01
4.47137892e-01 -4.35133696e-01 1.04222737e-01 -1.60480157e-01
4.95925754e-01 5.59194803e-01 -1.47950366e-01 -5.45146108e-01
-3.02102089e-01 3.31782430e-01 6.29149377e-01 -2.95598805e-01
1.12296915e+00 1.74471647e-01 5.66918850e-01 4.95352566e-01
7.85286188e-01 -3.95321660e-02 -1.65507531e+00 -2.58620352e-01
-3.41844447e-02 -3.35452050e-01 -4.29270953e-01 -6.42499566e-01
-1.09171963e+00 5.49486697e-01 7.73506403e-01 2.39213958e-01
1.12402940e+00 -3.59549552e-01 1.15588903e+00 4.36325341e-01
6.21890903e-01 -1.23904181e+00 3.51786733e-01 -3.29357721e-02
4.52782810e-01 -1.33529639e+00 -7.36515895e-02 -1.01710454e-01
-7.60800004e-01 3.46135765e-01 4.60015208e-01 -2.90610224e-01
6.39651477e-01 4.14735019e-01 6.72488362e-02 6.75999224e-02
-3.20303082e-01 -1.29138425e-01 3.75063270e-02 1.14017844e+00
7.70334154e-02 1.41217694e-01 -2.50354469e-01 8.46173227e-01
-9.95439570e-03 5.36112249e-01 1.46149933e-01 1.35495210e+00
-1.54850617e-01 -1.02605319e+00 -6.62525952e-01 5.18290997e-01
-6.00839972e-01 4.04628515e-01 -3.88248920e-01 5.49843788e-01
1.63443565e-01 1.11718452e+00 -5.06452806e-02 -5.11357725e-01
6.26306236e-01 4.52159405e-01 3.92506212e-01 -1.03617743e-01
-1.49375871e-01 -2.01156244e-01 6.49453700e-02 -5.74088991e-01
-7.92657614e-01 -9.33204889e-01 -1.12616432e+00 -1.15203157e-01
2.76524216e-01 5.21619916e-02 7.41177976e-01 8.23548198e-01
4.47284400e-01 6.16587877e-01 6.47290170e-01 -8.42165470e-01
-7.17388988e-01 -1.15774238e+00 -6.60208702e-01 1.16160250e+00
9.79829803e-02 -6.86666429e-01 -3.51525962e-01 5.94218791e-01] | [7.818946361541748, 1.0461063385009766] |
6acbebc5-95df-43d2-bf74-ebba0c34a38f | dense-rgb-slam-with-neural-implicit-maps | 2301.08930 | null | https://arxiv.org/abs/2301.08930v2 | https://arxiv.org/pdf/2301.08930v2.pdf | Dense RGB SLAM with Neural Implicit Maps | There is an emerging trend of using neural implicit functions for map representation in Simultaneous Localization and Mapping (SLAM). Some pioneer works have achieved encouraging results on RGB-D SLAM. In this paper, we present a dense RGB SLAM method with neural implicit map representation. To reach this challenging goal without depth input, we introduce a hierarchical feature volume to facilitate the implicit map decoder. This design effectively fuses shape cues across different scales to facilitate map reconstruction. Our method simultaneously solves the camera motion and the neural implicit map by matching the rendered and input video frames. To facilitate optimization, we further propose a photometric warping loss in the spirit of multi-view stereo to better constrain the camera pose and scene geometry. We evaluate our method on commonly used benchmarks and compare it with modern RGB and RGB-D SLAM systems. Our method achieves favorable results than previous methods and even surpasses some recent RGB-D SLAM methods.The code is at poptree.github.io/DIM-SLAM/. | ['Ping Tan', 'Zilong Dong', 'Luwei Yang', 'Weihao Yuan', 'Xiaodong Gu', 'Heng Li'] | 2023-01-21 | null | null | null | null | ['simultaneous-localization-and-mapping'] | ['computer-vision'] | [ 4.22025360e-02 -2.41376668e-01 -4.85754684e-02 -5.69558322e-01
-7.24143445e-01 -7.28112519e-01 5.85160792e-01 -2.81379938e-01
-4.12042201e-01 5.10319054e-01 2.59364545e-01 -1.85904875e-01
1.09984711e-01 -7.35787868e-01 -9.35173392e-01 -4.03864771e-01
2.32431680e-01 6.16863489e-01 2.19990999e-01 -3.20687979e-01
4.96962637e-01 6.33074045e-01 -1.42956817e+00 4.12681028e-02
6.76330149e-01 8.83681595e-01 4.45810854e-01 5.62536061e-01
-1.46860123e-01 7.82801449e-01 2.72236448e-02 -7.94235319e-02
7.03933418e-01 -1.46182910e-01 -8.64579320e-01 -3.00538577e-02
9.14193213e-01 -3.37138295e-01 -6.74405932e-01 1.01761293e+00
6.84600770e-01 2.67956704e-01 4.42815535e-02 -1.26433945e+00
-3.83023381e-01 -2.20976979e-01 -6.66900754e-01 -1.14903584e-01
7.09696114e-01 2.82406136e-02 8.03626478e-01 -1.10907674e+00
7.22777903e-01 1.22808778e+00 1.05235326e+00 2.16207758e-01
-1.21937358e+00 -5.64840913e-01 1.24692321e-02 8.52035880e-02
-1.86469686e+00 -6.26724541e-01 7.46784210e-01 -2.96893209e-01
1.08703029e+00 1.91419065e-01 7.15267777e-01 6.90558016e-01
2.70172268e-01 5.11536896e-01 1.10326898e+00 -2.15898305e-01
1.20770432e-01 -2.41983920e-01 -2.74156153e-01 9.57540870e-01
-1.09827742e-01 1.57562807e-01 -1.15527594e+00 -4.40255478e-02
1.20629048e+00 2.01620549e-01 -5.15187204e-01 -1.02523637e+00
-1.41405690e+00 8.29286218e-01 9.67685878e-01 -1.60240203e-01
-2.12530151e-01 7.50607550e-01 4.69496436e-02 2.48596698e-01
3.12905461e-01 3.46775770e-01 -1.88231438e-01 -1.81253940e-01
-9.59475160e-01 5.93208596e-02 7.06009448e-01 1.20858431e+00
1.46926892e+00 -2.12335438e-01 5.41130185e-01 4.29649681e-01
4.41306442e-01 7.85779476e-01 1.55822486e-01 -1.55792713e+00
5.97128093e-01 5.76545179e-01 1.69977650e-01 -1.23635328e+00
-5.71504593e-01 -2.07023118e-02 -6.57218575e-01 2.44653106e-01
7.60368630e-02 4.21206057e-01 -9.24680293e-01 1.29518640e+00
4.00382727e-01 4.01399791e-01 -1.60357863e-01 1.16534543e+00
7.23938942e-01 5.27196288e-01 -6.45502687e-01 3.88197929e-01
8.22117567e-01 -1.09062183e+00 -5.67730248e-01 -5.49497128e-01
5.36205292e-01 -6.78674221e-01 1.02012742e+00 1.63230672e-01
-7.40032434e-01 -2.67546982e-01 -1.10380936e+00 -5.88609040e-01
-2.97985554e-01 -1.11234881e-01 8.89292061e-01 3.76251876e-01
-1.54839635e+00 4.90480840e-01 -1.05403209e+00 -5.40558219e-01
1.00264534e-01 6.59184933e-01 -7.67741442e-01 -2.77602851e-01
-8.27575564e-01 9.17348385e-01 3.25281024e-01 2.12995976e-01
-6.59638643e-01 -4.25031811e-01 -1.22446942e+00 -5.78542531e-01
2.05397829e-01 -9.31901872e-01 9.64478970e-01 -5.81039846e-01
-1.56654322e+00 1.13145101e+00 -3.89155656e-01 -2.78814137e-01
4.79387432e-01 -5.05802453e-01 3.53368700e-01 3.19373095e-03
1.79408878e-01 1.13825285e+00 4.41931069e-01 -1.34571326e+00
-4.73427147e-01 -4.88794506e-01 2.91916728e-01 5.65607607e-01
2.16230586e-01 -5.41255176e-01 -8.80505264e-01 -8.92435387e-02
9.25542355e-01 -1.18763351e+00 -4.20899093e-01 2.77749121e-01
-2.95914680e-01 3.92692238e-01 5.86930335e-01 -5.57706714e-01
6.35248661e-01 -2.02797389e+00 4.22619402e-01 1.34776518e-01
3.10167044e-01 -4.56442952e-01 -8.56815577e-02 2.87851006e-01
5.02700746e-01 -2.92146504e-01 -2.28785872e-01 -9.77440655e-01
7.23595396e-02 6.57488227e-01 -3.64073545e-01 1.07175648e+00
-2.47203097e-01 1.01397502e+00 -7.77102470e-01 -3.44751835e-01
5.34759462e-01 7.34783471e-01 -6.27422869e-01 3.11736390e-02
6.05042912e-02 7.60165811e-01 -2.23776445e-01 8.22794914e-01
9.40232873e-01 -3.23780328e-01 -1.05667792e-01 -1.36192188e-01
-4.78274852e-01 4.88369405e-01 -1.22254002e+00 2.95806575e+00
-6.04081631e-01 8.55609059e-01 2.40358129e-01 -4.54786301e-01
9.77991402e-01 -1.68496132e-01 4.89877790e-01 -9.33259904e-01
5.58300056e-02 3.98125231e-01 -6.18782640e-01 1.64907143e-01
8.62833202e-01 1.90470681e-01 2.41051447e-02 3.03120129e-02
-1.14057853e-03 -6.95226848e-01 -4.53513980e-01 1.26334012e-01
1.11250210e+00 8.07747424e-01 3.82758766e-01 -3.00626040e-01
4.25254375e-01 3.54241192e-01 6.19754910e-01 6.85859084e-01
-6.59340173e-02 7.86916792e-01 -3.05073373e-02 -8.99670839e-01
-1.08440840e+00 -1.22297645e+00 -5.52228466e-03 6.45675361e-01
8.88320684e-01 -5.15072346e-01 -2.38894865e-01 -2.19560742e-01
3.56429577e-01 2.04650372e-01 -5.47215581e-01 2.03246742e-01
-7.44300365e-01 -3.29368591e-01 4.16880190e-01 4.24323529e-01
7.18092203e-01 -6.59802973e-01 -7.36368775e-01 9.88764614e-02
-3.20355415e-01 -1.21244013e+00 -3.01252991e-01 4.09730852e-01
-9.33629215e-01 -8.56578767e-01 -3.85043144e-01 -6.48140550e-01
6.86723948e-01 6.61950648e-01 1.12339413e+00 -1.42145958e-02
-5.41103221e-02 3.93882781e-01 -2.52382219e-01 1.72730591e-02
-1.67592727e-02 2.00550735e-01 3.13859612e-01 -3.63277406e-01
2.23287672e-01 -8.46276999e-01 -6.58005536e-01 5.79386890e-01
-5.16942799e-01 5.41934907e-01 3.64067137e-01 5.49380243e-01
9.97043014e-01 -6.73381686e-01 -2.22672015e-01 -2.98472524e-01
-1.38171330e-01 -1.15193173e-01 -1.08937812e+00 -1.64314270e-01
-5.76315820e-01 1.06853306e-01 -6.32103533e-04 5.09423353e-02
-4.44733620e-01 8.11229289e-01 -1.87815994e-01 -6.07615411e-01
9.41063166e-02 2.58428663e-01 -6.38412833e-02 -8.24598551e-01
4.39908445e-01 3.27918559e-01 6.46759346e-02 -5.31235397e-01
4.44189340e-01 3.53767931e-01 7.40544677e-01 -4.03578788e-01
9.52018201e-01 9.94293630e-01 2.28223816e-01 -5.13509929e-01
-7.05793619e-01 -7.76803493e-01 -1.04683781e+00 -1.48355946e-01
6.68169379e-01 -1.25304866e+00 -7.02791333e-01 2.33804584e-01
-1.44537568e+00 -6.52624488e-01 2.29557045e-02 5.09797156e-01
-1.04032660e+00 3.33361238e-01 -4.88905758e-01 -4.55058843e-01
-1.40515178e-01 -1.39311814e+00 1.61888242e+00 1.31333172e-02
2.60136891e-02 -8.90437305e-01 5.10857344e-01 1.40205592e-01
3.62088084e-01 4.27218348e-01 2.14376636e-02 1.88095272e-01
-1.23518360e+00 4.54838760e-02 -2.69030809e-01 -1.96935803e-01
1.29847407e-01 -6.29858732e-01 -9.48703110e-01 -4.44112837e-01
-2.82211173e-02 -2.03830242e-01 8.45268846e-01 2.33164385e-01
6.59997225e-01 6.06602523e-03 -3.79811585e-01 1.65358210e+00
1.80890751e+00 -1.76830351e-01 6.93652511e-01 8.99289489e-01
1.10425901e+00 3.60327095e-01 5.57006121e-01 3.83081615e-01
7.39491284e-01 9.25846517e-01 9.03480411e-01 -2.18751684e-01
1.25926390e-01 -3.83603394e-01 3.66750330e-01 9.51349258e-01
-5.87148108e-02 -2.44305171e-02 -1.20419371e+00 2.98006028e-01
-2.01495862e+00 -4.43587154e-01 -1.28134921e-01 2.19439149e+00
5.40817976e-01 -2.19380617e-01 -3.44451100e-01 -3.10597509e-01
4.66727197e-01 4.36906010e-01 -5.46553016e-01 -4.80682924e-02
-3.69460732e-01 -3.05758920e-02 1.06277406e+00 9.88201797e-01
-1.14040303e+00 1.30169153e+00 5.64090872e+00 2.94719517e-01
-1.08318663e+00 2.32200876e-01 1.63910985e-02 -3.19468588e-01
-4.33108091e-01 3.74824077e-01 -8.58209074e-01 -2.28198767e-02
5.96041858e-01 7.47295246e-02 7.48943985e-01 7.53520191e-01
9.02187601e-02 -3.54669154e-01 -1.15631211e+00 1.53516030e+00
1.44373879e-01 -1.68488872e+00 -2.26296902e-01 4.19068545e-01
7.91457891e-01 7.96198070e-01 -1.59192398e-01 -1.07033722e-01
3.08582425e-01 -1.00281906e+00 8.98462951e-01 4.17575449e-01
9.00416493e-01 -5.29245019e-01 6.74751222e-01 3.02659005e-01
-1.47021890e+00 3.20727378e-01 -5.16782165e-01 -2.85898924e-01
3.81736577e-01 3.66799027e-01 -8.55498314e-01 7.03906655e-01
7.25685239e-01 1.08406639e+00 -6.20290935e-01 9.94321227e-01
-2.10911423e-01 -2.62531579e-01 -7.02722728e-01 3.03062141e-01
3.64524156e-01 -3.49158078e-01 5.11947989e-01 9.54029858e-01
4.86863136e-01 -1.83284894e-01 2.93390095e-01 9.08429146e-01
1.13070616e-02 -4.03429978e-02 -1.02585816e+00 4.05791581e-01
4.67109352e-01 9.76126790e-01 -6.39054537e-01 -1.05368076e-02
-4.44648862e-01 1.49446845e+00 4.52863485e-01 1.72630668e-01
-8.45039904e-01 -8.94861147e-02 8.62690985e-01 1.47263240e-02
-7.94758182e-03 -8.67171288e-01 -3.73063534e-01 -1.26967716e+00
5.54259457e-02 -3.71333808e-01 -1.13254555e-01 -1.01861131e+00
-6.84414506e-01 5.28666139e-01 -3.53199184e-01 -1.27751172e+00
-1.76427826e-01 -6.33846760e-01 1.89823061e-02 9.70764756e-01
-1.71856642e+00 -1.11581647e+00 -9.44370687e-01 7.30202436e-01
3.56826127e-01 2.32234135e-01 7.00196207e-01 3.42342168e-01
5.31339161e-02 2.42291540e-02 2.23847643e-01 8.42720345e-02
8.40603769e-01 -1.29130089e+00 9.50245976e-01 7.79368639e-01
3.54570717e-01 6.28783643e-01 4.15797234e-01 -6.84543133e-01
-2.01055932e+00 -9.23089743e-01 4.97558475e-01 -8.09444249e-01
4.61098999e-01 -7.42298543e-01 -5.33265412e-01 1.03210437e+00
-4.20948751e-02 2.19549790e-01 2.54800797e-01 -1.11972295e-01
-3.72467697e-01 -1.52289629e-01 -8.66354048e-01 3.57590824e-01
1.43551743e+00 -8.53931963e-01 -2.63589740e-01 2.56739974e-01
9.69230235e-01 -1.06285918e+00 -6.74632728e-01 4.38926876e-01
6.38217926e-01 -1.34650064e+00 1.11815810e+00 1.00852482e-01
7.34969676e-02 -7.67494917e-01 -8.64247143e-01 -1.13525116e+00
-2.39906400e-01 -7.30569601e-01 -9.80098471e-02 8.09524179e-01
1.03015900e-01 -7.85709083e-01 1.07185817e+00 3.51702631e-01
-3.88854384e-01 -4.99784976e-01 -1.27177763e+00 -7.16198206e-01
-3.81694525e-01 -3.96197975e-01 6.35716200e-01 8.48250568e-01
-4.33798105e-01 -6.78324774e-02 -5.33212364e-01 5.63695312e-01
7.62096941e-01 2.70503163e-01 1.24641037e+00 -9.79051828e-01
2.68288446e-03 -1.81914523e-01 -8.17213893e-01 -1.60973418e+00
1.35741830e-01 -9.85415936e-01 3.82679909e-01 -1.77069819e+00
-1.08913831e-01 -5.91479003e-01 -9.16818529e-02 4.87858087e-01
3.71296108e-01 6.71318173e-01 2.50130177e-01 5.08158624e-01
-7.79649794e-01 5.93659997e-01 1.06572521e+00 7.10966364e-02
-3.33647519e-01 -4.99800205e-01 -1.71048731e-01 7.45221078e-01
6.64036155e-01 -4.71949816e-01 -1.23191737e-01 -1.03440583e+00
5.75958490e-01 1.54218167e-01 7.27979660e-01 -1.22559547e+00
5.33723831e-01 -1.48564592e-01 2.50211209e-01 -9.71559525e-01
8.63130689e-01 -9.17092443e-01 4.93017018e-01 4.82217938e-01
1.20431706e-01 4.48999971e-01 8.96569192e-02 5.21938622e-01
-2.94538826e-01 2.14586973e-01 5.31507432e-01 -3.19661617e-01
-1.31462061e+00 6.44817173e-01 2.79907256e-01 -2.14011580e-01
5.66984236e-01 -4.29727942e-01 -2.54407942e-01 -4.62571591e-01
-4.34237272e-01 2.70645857e-01 1.34273398e+00 4.96605873e-01
8.64317954e-01 -1.54314291e+00 -3.83262962e-01 3.25962603e-01
2.22997412e-01 4.80783939e-01 -8.11392590e-02 1.23630273e+00
-1.32425499e+00 4.84396279e-01 -2.81536043e-01 -1.11616862e+00
-9.68115270e-01 2.27932319e-01 5.61804950e-01 2.46554703e-01
-7.92183399e-01 8.36756647e-01 3.61611426e-01 -9.01513815e-01
2.63003945e-01 -2.69681424e-01 5.13969779e-01 -4.40577626e-01
3.98337990e-01 3.87238204e-01 5.42132631e-02 -9.40225720e-01
-7.98978508e-01 9.56058025e-01 5.39013863e-01 -2.40624741e-01
1.42630720e+00 -5.99290550e-01 -4.57132578e-01 5.76654971e-01
1.41861331e+00 1.09538540e-01 -1.54859102e+00 -2.21566454e-01
-4.38198820e-02 -7.52649784e-01 1.51815251e-01 -2.39138111e-01
-8.11249495e-01 9.19309258e-01 6.22259676e-01 -3.15022916e-01
9.40386593e-01 -9.29827918e-04 6.86131060e-01 6.35489166e-01
1.13103950e+00 -7.98997879e-01 -1.14874169e-01 9.45601106e-01
8.97413909e-01 -1.53014410e+00 2.79058516e-01 -3.58454943e-01
-2.90643275e-01 1.17472506e+00 5.30230999e-01 -2.77096957e-01
2.45833546e-01 3.28548044e-01 1.84788942e-01 -2.71832556e-01
-1.26492903e-01 -1.98199168e-01 1.94584951e-01 4.99013066e-01
1.55856952e-01 -1.19767196e-01 3.50073487e-01 -2.55264908e-01
-3.57450873e-01 -4.48697098e-02 2.99173832e-01 1.03350544e+00
-5.40697396e-01 -1.00850999e+00 -4.96675342e-01 -1.41599879e-01
2.25180745e-01 -2.08879426e-01 -4.69287127e-01 8.96308661e-01
2.56132502e-02 4.57178444e-01 1.13109596e-01 -6.29466474e-01
1.76716194e-01 -1.43256038e-01 7.12182641e-01 -4.23689842e-01
-1.83608457e-01 -6.58229878e-03 -9.94678289e-02 -1.44275868e+00
-6.15469038e-01 -5.67275584e-01 -1.40069699e+00 -5.04495859e-01
-1.02556594e-01 -1.46243647e-01 1.10523951e+00 6.77064717e-01
5.24173141e-01 3.24866213e-02 4.88582253e-01 -1.50494075e+00
-2.75274962e-02 -4.82886195e-01 -5.11667371e-01 6.19762465e-02
6.72590435e-01 -8.03549170e-01 -3.68537843e-01 -1.52952671e-01] | [7.571420669555664, -2.269552230834961] |
43a2ad52-4db0-4da5-b802-4ad2ea1fcbf2 | symmetric-dilated-convolution-for-surgical | 2007.06373 | null | https://arxiv.org/abs/2007.06373v2 | https://arxiv.org/pdf/2007.06373v2.pdf | Symmetric Dilated Convolution for Surgical Gesture Recognition | Automatic surgical gesture recognition is a prerequisite of intra-operative computer assistance and objective surgical skill assessment. Prior works either require additional sensors to collect kinematics data or have limitations on capturing temporal information from long and untrimmed surgical videos. To tackle these challenges, we propose a novel temporal convolutional architecture to automatically detect and segment surgical gestures with corresponding boundaries only using RGB videos. We devise our method with a symmetric dilation structure bridged by a self-attention module to encode and decode the long-term temporal patterns and establish the frame-to-frame relationship accordingly. We validate the effectiveness of our approach on a fundamental robotic suturing task from the JIGSAWS dataset. The experiment results demonstrate the ability of our method on capturing long-term frame dependencies, which largely outperform the state-of-the-art methods on the frame-wise accuracy up to ~6 points and the F1@50 score ~6 points. | ['Jian Chang', 'Jian Jun Zhang', 'Xiaosong Yang', 'Yao Lyu', 'Yinyu Nie', 'Jinglu Zhang', 'Hailin Li'] | 2020-07-13 | null | null | null | null | ['surgical-gesture-recognition'] | ['medical'] | [ 2.78525084e-01 3.74764167e-02 -4.09316897e-01 -2.93906212e-01
-8.12725902e-01 -3.90855193e-01 3.04044306e-01 -3.56781185e-01
-8.19948077e-01 2.04853103e-01 3.45903963e-01 -3.24150503e-01
-3.25769842e-01 -1.00373197e-02 -7.15048432e-01 -6.65576577e-01
-2.50289172e-01 -5.81367053e-02 3.09683651e-01 -1.49684235e-01
3.12390059e-01 3.16971272e-01 -1.23197424e+00 3.81591290e-01
5.08881867e-01 1.11914802e+00 2.98367560e-01 9.47278559e-01
2.27591112e-01 9.28885520e-01 -1.51939556e-01 9.61545706e-02
4.27298099e-01 -3.95059526e-01 -8.55168760e-01 -1.13497376e-01
2.12045521e-01 -6.01797462e-01 -8.30553293e-01 7.57385254e-01
6.68680608e-01 -1.33353904e-01 4.35306653e-02 -6.71815515e-01
-7.71568939e-02 4.31538731e-01 -2.49321520e-01 4.48212564e-01
3.05241227e-01 5.11739254e-01 4.58110452e-01 -5.92747986e-01
9.15350854e-01 7.27057695e-01 6.93397284e-01 7.98391759e-01
-7.16583908e-01 -7.30981827e-01 1.46485910e-01 1.14194520e-01
-9.84971166e-01 -3.92304629e-01 7.00581074e-01 -4.29266810e-01
7.36721575e-01 1.39087811e-01 8.43388617e-01 1.30297685e+00
7.70571232e-01 8.91172528e-01 9.58070815e-01 -2.02872738e-01
-1.83331296e-01 -8.62263978e-01 -3.65289510e-03 1.10392511e+00
-1.29623249e-01 6.56179368e-01 -7.28225231e-01 5.77329695e-01
1.36949432e+00 3.11876476e-01 -5.41254997e-01 -4.17688757e-01
-1.93836367e+00 3.68803829e-01 8.35430086e-01 4.89279866e-01
-4.25991356e-01 7.08157301e-01 4.96633470e-01 2.74611205e-01
-1.07176296e-01 4.09368426e-01 -3.90061736e-01 -6.87400699e-01
-6.73788488e-01 -4.41823065e-01 4.78937060e-01 1.09783280e+00
-4.61362153e-02 -3.48881394e-01 -3.98604900e-01 2.13475868e-01
8.48422945e-02 8.69731903e-02 8.71681154e-01 -9.28195775e-01
3.05026174e-01 3.73121500e-01 8.20612535e-02 -4.65638846e-01
-9.08820748e-01 -4.32714403e-01 -6.31143808e-01 2.27925748e-01
5.06664753e-01 -6.55864775e-02 -1.36558735e+00 1.47439313e+00
2.45914072e-01 5.34062564e-01 -1.22031726e-01 1.19619787e+00
9.93002474e-01 -8.30435604e-02 3.67447883e-02 -3.35652888e-01
1.22800660e+00 -1.17487085e+00 -8.21342885e-01 -9.10783485e-02
8.55584860e-01 -6.53413892e-01 1.08476651e+00 2.90332109e-01
-8.79377127e-01 -5.58493733e-01 -1.07819903e+00 -1.92720592e-01
2.80833542e-01 4.66638863e-01 9.22210038e-01 2.78846860e-01
-7.06069589e-01 6.78928077e-01 -1.87514436e+00 -1.59610704e-01
4.52377230e-01 9.03548062e-01 -7.00058818e-01 -8.43910277e-02
-6.17044508e-01 7.79327631e-01 1.69785202e-01 6.66589200e-01
-9.44789350e-01 -8.32814515e-01 -9.62534606e-01 -2.80899972e-01
2.86894590e-01 -6.40500486e-01 1.41621959e+00 -7.65074372e-01
-1.92312062e+00 8.36634457e-01 -9.08956304e-03 -3.09196293e-01
6.99256659e-01 -4.75998998e-01 -1.41379759e-02 4.10644710e-01
-4.15159911e-01 5.79985559e-01 5.97320557e-01 -7.95098424e-01
-5.85740805e-01 -3.60107422e-01 2.98827171e-01 2.32866764e-01
-3.07369143e-01 -1.16629742e-01 -7.25910127e-01 -9.49181020e-01
5.01438320e-01 -1.36374998e+00 -4.28623855e-01 3.69201571e-01
-9.60549787e-02 2.71305412e-01 6.40093088e-01 -5.89394093e-01
1.21535361e+00 -2.16234660e+00 4.39415962e-01 -1.33047625e-01
2.35869661e-01 1.65754318e-01 -2.54891608e-02 -9.00764763e-02
-1.52107440e-02 -4.14413154e-01 1.32309407e-01 -1.55810177e-01
-3.75067443e-01 4.86184031e-01 -3.74512263e-02 7.72334576e-01
-3.00168861e-02 1.20727682e+00 -1.20111978e+00 -4.68857318e-01
4.64024544e-01 4.51381892e-01 -5.88421345e-01 3.99268746e-01
2.03070655e-01 1.13362944e+00 -4.90396440e-01 7.90338993e-01
8.38832036e-02 -1.98978648e-01 2.42141515e-01 -4.03849810e-01
-2.82799806e-02 2.65236974e-01 -5.27099192e-01 2.90311360e+00
-5.71267307e-01 5.49809456e-01 1.52103409e-01 -7.84370661e-01
4.36616868e-01 3.72606248e-01 1.07775331e+00 -7.38459706e-01
6.42182767e-01 3.49445552e-01 3.26178670e-01 -8.41949224e-01
3.33825685e-02 -2.24703804e-01 -2.70099640e-01 1.00907810e-01
2.68848062e-01 3.61682773e-02 -7.15561882e-02 -2.06969082e-01
1.44896901e+00 4.26498383e-01 9.94390845e-02 -2.28047892e-01
2.55005538e-01 1.36485994e-01 5.50172567e-01 5.84018350e-01
-6.23988509e-01 7.74376333e-01 4.53928351e-01 -7.29028702e-01
-6.40776575e-01 -8.38179648e-01 1.49309579e-02 6.37258232e-01
5.58225334e-01 -1.73953682e-01 -4.94352102e-01 -1.06034207e+00
-1.73849821e-01 -7.50006288e-02 -9.77103949e-01 -3.50597918e-01
-1.04090214e+00 -3.35110217e-01 3.35476756e-01 9.46677446e-01
1.99683651e-01 -9.17607963e-01 -1.15594220e+00 2.49256045e-01
-2.04074472e-01 -1.47789180e+00 -6.40968502e-01 1.86303541e-01
-1.07263362e+00 -1.30298221e+00 -7.11328626e-01 -1.02916753e+00
9.01787162e-01 1.16980344e-01 5.22689641e-01 4.77052294e-02
-6.62483156e-01 2.95290798e-01 -5.64937532e-01 -2.67408073e-01
-1.73316777e-01 1.54357746e-01 -1.42382026e-01 -2.51916975e-01
6.54374016e-03 -4.16915804e-01 -1.10853779e+00 5.10470986e-01
-8.53762627e-01 3.85422260e-01 9.96344686e-01 1.19923651e+00
4.92877930e-01 -7.02616513e-01 6.23250008e-02 -6.23145163e-01
-1.70383435e-02 4.98471148e-02 -5.53355455e-01 9.74678546e-02
-3.25006068e-01 2.24911332e-01 3.03091735e-01 -6.29217803e-01
-5.95259130e-01 6.91215158e-01 -8.61861706e-02 -8.97243559e-01
5.67887835e-02 5.32866836e-01 4.88015443e-01 -4.91453111e-01
4.75788802e-01 -4.77969758e-02 3.97429347e-01 -1.00618437e-01
1.10140562e-01 2.58689225e-01 1.13175011e+00 -3.91607702e-01
5.22138834e-01 7.38184631e-01 1.22191206e-01 -1.78830996e-01
-7.77310193e-01 -8.05805504e-01 -1.00693607e+00 -4.80960965e-01
9.86730456e-01 -9.32754934e-01 -1.03264499e+00 4.33198422e-01
-9.99701381e-01 -5.27544558e-01 -9.52265784e-02 1.16623855e+00
-8.26953709e-01 2.80076265e-01 -9.31629777e-01 -1.35556430e-01
-4.72646803e-01 -1.60902286e+00 1.39307702e+00 5.67992851e-02
-8.68532956e-02 -6.34039104e-01 -1.64620262e-02 1.96792275e-01
2.80100554e-01 7.31820405e-01 4.27174896e-01 -1.12752743e-01
-7.42872477e-01 -5.21922767e-01 -6.05586320e-02 -6.52259737e-02
2.14527830e-01 -3.96956652e-01 -5.67576230e-01 -4.94781375e-01
-1.17669918e-01 -2.62648523e-01 8.21997225e-01 5.92373013e-01
1.36934090e+00 7.88652524e-02 -3.79955173e-01 1.13801467e+00
1.25689828e+00 3.53012197e-02 6.93791211e-01 2.00448081e-01
8.44900906e-01 2.96666056e-01 8.51953506e-01 2.09980816e-01
1.99797064e-01 7.32765138e-01 6.52544081e-01 -9.87052694e-02
-3.10087115e-01 6.23511989e-03 2.76658535e-01 1.01847804e+00
-4.19582397e-01 2.44285911e-01 -9.72503841e-01 5.40637791e-01
-1.99356914e+00 -6.96041644e-01 1.42329350e-01 2.09127831e+00
8.26794803e-01 1.18876003e-01 -4.28686887e-01 -7.27359280e-02
1.32140338e-01 -1.24662668e-02 -5.84649920e-01 2.27144092e-01
5.13110101e-01 5.25537312e-01 8.30793917e-01 4.47469205e-01
-1.31721318e+00 8.96702409e-01 6.17801905e+00 4.64903057e-01
-1.71734965e+00 -3.05081182e-03 1.82322145e-01 -6.44257009e-01
2.68768132e-01 -2.34535635e-01 -2.36589760e-01 3.31272125e-01
7.03530908e-01 2.97218859e-01 1.35571897e-01 5.65674901e-01
2.39731446e-01 -4.62136827e-02 -1.33588111e+00 1.10102308e+00
1.02835156e-01 -1.52771604e+00 -5.40741026e-01 -1.00453645e-02
6.76198721e-01 1.37742326e-01 -3.34594958e-02 1.60099417e-01
6.87783211e-02 -1.18114257e+00 5.64397037e-01 6.94131315e-01
1.28154182e+00 -3.03905517e-01 8.34002256e-01 1.41471684e-01
-1.17849267e+00 -2.98072129e-01 1.19781226e-01 5.01845777e-02
9.15329233e-02 -1.27902389e-01 -7.76868939e-01 6.20328903e-01
5.25306642e-01 1.02215922e+00 -1.60759345e-01 1.03931177e+00
-2.88274586e-01 2.44488016e-01 -3.09546590e-01 1.63326144e-01
3.99646401e-01 4.14219201e-01 2.88844228e-01 9.91520882e-01
3.69348019e-01 4.77092206e-01 2.07767814e-01 -6.23722980e-03
6.26647323e-02 -3.42706412e-01 -2.62105942e-01 -9.98120569e-03
-1.66616485e-01 1.14806283e+00 -8.47973645e-01 1.30176079e-02
-4.07762140e-01 1.18672347e+00 3.34635973e-02 1.86523423e-01
-1.02164698e+00 -2.72838354e-01 6.92733109e-01 -1.15970038e-01
3.45318347e-01 -6.34355545e-01 -4.47520018e-01 -1.29998481e+00
3.08824062e-01 -6.61481678e-01 3.52226496e-01 -6.31668270e-01
-4.97572690e-01 6.23194098e-01 -3.74466151e-01 -1.92253721e+00
-4.04927611e-01 -9.88180995e-01 -3.17876458e-01 1.99662030e-01
-1.61813104e+00 -1.44328523e+00 -1.01181495e+00 7.68730640e-01
5.80887496e-01 2.80318886e-01 8.65397453e-01 1.30121455e-01
-4.57162827e-01 5.96316695e-01 -9.30097401e-02 5.38962841e-01
9.25284445e-01 -8.39297235e-01 1.85734387e-02 8.97850692e-01
-8.19038972e-02 6.64920628e-01 4.21893239e-01 -3.31442297e-01
-1.95969713e+00 -7.73681819e-01 3.67049009e-01 -7.29692996e-01
5.46058893e-01 -2.09579095e-01 -2.36356795e-01 8.39653611e-01
-9.10180584e-02 7.41881490e-01 5.36106944e-01 -2.64228135e-01
-8.27883109e-02 -5.83024472e-02 -7.46695757e-01 4.56454992e-01
1.55582631e+00 -3.54511678e-01 -5.23736775e-01 2.40663424e-01
7.21400678e-01 -1.46564209e+00 -1.06780279e+00 9.18873429e-01
1.19360399e+00 -7.71254182e-01 9.44230258e-01 -6.93123281e-01
8.99854183e-01 -2.57332891e-01 1.65736884e-01 -8.70611787e-01
-8.01250935e-02 -8.08938146e-01 -2.28586718e-02 9.49510261e-02
7.63132349e-02 -1.84097454e-01 9.36060429e-01 4.87562865e-01
-7.28866220e-01 -1.07849526e+00 -1.34160113e+00 -5.91624618e-01
-2.04450786e-01 -6.05185390e-01 1.31631643e-02 7.80373037e-01
4.81718957e-01 -3.84671271e-01 -2.33405486e-01 2.34595224e-01
1.75947785e-01 2.80019194e-01 6.72392845e-01 -7.15824842e-01
-2.37206250e-01 -5.18555105e-01 -8.17431688e-01 -1.21523631e+00
-1.85102634e-02 -7.59270370e-01 5.64863384e-01 -1.34420860e+00
-1.03407100e-01 -2.53203928e-01 -6.45254791e-01 5.90651989e-01
-8.24060738e-02 4.11313117e-01 5.83332032e-02 3.41224432e-01
-6.12724423e-01 3.49301070e-01 1.62734032e+00 3.35318968e-02
-1.94980249e-01 -2.61646092e-01 -1.42446354e-01 5.84475279e-01
2.99443543e-01 -3.58328789e-01 -2.86155194e-01 -7.97794878e-01
-2.62196362e-01 3.77458721e-01 4.38092619e-01 -1.12577569e+00
5.94924152e-01 -1.04930691e-01 3.49566132e-01 -3.53237927e-01
2.18626231e-01 -1.02951586e+00 -1.11970797e-01 1.01972568e+00
-3.61172020e-01 1.35843521e-02 2.77903825e-01 5.61616659e-01
-3.73801351e-01 4.42653865e-01 6.61342978e-01 3.15341651e-02
-9.42619741e-01 6.08238280e-01 -2.46706214e-02 -1.85526282e-01
1.20468020e+00 -1.89639837e-01 -1.47734061e-01 3.20112407e-02
-7.92330980e-01 1.38603270e-01 3.80950600e-01 6.27386510e-01
8.08595061e-01 -1.21068931e+00 -4.07199323e-01 5.09613395e-01
3.69657367e-01 2.07859248e-01 3.82755101e-01 1.64373183e+00
-1.03572154e+00 5.77583075e-01 -4.45576340e-01 -9.60953116e-01
-1.14239669e+00 3.85675699e-01 4.26877022e-01 -4.35833514e-01
-1.12523592e+00 1.02703977e+00 2.69775301e-01 -1.09937198e-01
4.63595003e-01 -9.58163679e-01 -2.67775264e-02 -4.77596819e-01
4.05556291e-01 -3.68706495e-01 1.87860385e-01 -3.92898560e-01
-4.64355797e-01 7.91086853e-01 1.01267891e-02 -2.26914044e-02
1.36118484e+00 1.45785391e-01 2.84378558e-01 1.82476297e-01
1.22164893e+00 -2.90558279e-01 -1.82843292e+00 -1.59034014e-01
-2.18953371e-01 -6.70540333e-01 2.48509988e-01 -8.18806231e-01
-1.19474435e+00 7.48853445e-01 7.43617237e-01 -6.54026866e-01
1.31338716e+00 -3.68503928e-02 1.06116891e+00 9.75286141e-02
6.28063083e-01 -7.52371371e-01 1.59163624e-01 5.06755352e-01
9.38661039e-01 -1.37817872e+00 -1.55085802e-01 -4.35049027e-01
-5.33261538e-01 1.39543164e+00 5.62556088e-01 -2.95623124e-01
7.06871450e-01 5.80766261e-01 3.91217440e-01 -2.68745512e-01
-5.30365348e-01 -1.99565277e-01 6.24109030e-01 2.33268827e-01
5.52537680e-01 -1.28571346e-01 -5.40390372e-01 5.30079067e-01
-8.98679122e-02 6.14607632e-01 2.49984652e-01 1.29786479e+00
1.18580431e-01 -7.93281138e-01 2.63704449e-01 1.93999097e-01
-5.69242120e-01 3.43528613e-02 2.48388797e-01 9.65230525e-01
-1.95578277e-01 5.20921350e-01 -4.44275439e-02 -5.73664248e-01
5.35004377e-01 -1.88211828e-01 8.29521775e-01 -4.10193890e-01
-6.87831879e-01 2.36035168e-01 -9.57561806e-02 -1.36438322e+00
-6.85989320e-01 -5.15833437e-01 -1.62297499e+00 8.36392492e-02
-1.73245549e-01 -3.81242007e-01 7.54200757e-01 9.66539443e-01
3.86283547e-01 1.05269063e+00 4.30355757e-01 -1.09684896e+00
-3.17745358e-01 -8.67541552e-01 6.02234155e-02 4.70723391e-01
7.46437311e-01 -6.58531606e-01 6.70771254e-03 2.47786492e-01] | [14.068485260009766, -3.3385677337646484] |
0bc2bf60-9e8e-44d9-b7fa-c8129536a80e | click-is-not-equal-to-like-counterfactual | 2009.09945 | null | https://arxiv.org/abs/2009.09945v4 | https://arxiv.org/pdf/2009.09945v4.pdf | Clicks can be Cheating: Counterfactual Recommendation for Mitigating Clickbait Issue | Recommendation is a prevalent and critical service in information systems. To provide personalized suggestions to users, industry players embrace machine learning, more specifically, building predictive models based on the click behavior data. This is known as the Click-Through Rate (CTR) prediction, which has become the gold standard for building personalized recommendation service. However, we argue that there is a significant gap between clicks and user satisfaction -- it is common that a user is "cheated" to click an item by the attractive title/cover of the item. This will severely hurt user's trust on the system if the user finds the actual content of the clicked item disappointing. What's even worse, optimizing CTR models on such flawed data will result in the Matthew Effect, making the seemingly attractive but actually low-quality items be more frequently recommended. In this paper, we formulate the recommendation models as a causal graph that reflects the cause-effect factors in recommendation, and address the clickbait issue by performing counterfactual inference on the causal graph. We imagine a counterfactual world where each item has only exposure features (i.e., the features that the user can see before making a click decision). By estimating the click likelihood of a user in the counterfactual world, we are able to reduce the direct effect of exposure features and eliminate the clickbait issue. Experiments on real-world datasets demonstrate that our method significantly improves the post-click satisfaction of CTR models. | ['Tat-Seng Chua', 'Xiangnan He', 'Wenjie Wang', 'Hanwang Zhang', 'Fuli Feng'] | 2020-09-21 | null | null | null | null | ['counterfactual-inference'] | ['miscellaneous'] | [-5.16663156e-02 1.58398133e-02 -7.17846155e-01 -4.33739185e-01
-2.78072834e-01 -4.25164342e-01 3.45726669e-01 -8.08987394e-02
-2.96410233e-01 5.19625902e-01 4.18654114e-01 -7.99600005e-01
-2.27513880e-01 -1.02282143e+00 -8.10525417e-01 -3.00187379e-01
7.21752197e-02 6.25987574e-02 9.45233107e-02 -2.65159011e-01
6.30944848e-01 -1.28859386e-01 -1.50304890e+00 4.35365319e-01
8.46002340e-01 1.13714361e+00 3.11586022e-01 3.06016088e-01
-1.13240734e-01 4.40014422e-01 -2.75318444e-01 -1.02306867e+00
3.77324522e-01 -8.60952884e-02 -3.68916631e-01 -1.76152021e-01
1.13402970e-01 -5.30717254e-01 -3.54926884e-01 9.64453042e-01
-1.02778999e-02 -9.13565829e-02 3.89953554e-01 -1.29987311e+00
-8.86725605e-01 7.76035607e-01 -6.06584370e-01 1.20475590e-01
4.23844129e-01 7.16590285e-02 1.74538505e+00 -6.51546359e-01
3.86374325e-01 1.15330005e+00 4.34164464e-01 2.53851920e-01
-1.29247689e+00 -7.02804029e-01 5.30438960e-01 2.17740327e-01
-9.47611094e-01 1.66085109e-01 7.17290521e-01 -5.77350497e-01
2.81767905e-01 7.55535185e-01 6.23218179e-01 1.07373679e+00
6.00272834e-01 9.68431354e-01 9.95810449e-01 -1.40049890e-01
2.63185352e-01 3.97989064e-01 2.84218818e-01 1.09453797e-01
6.94436908e-01 4.00929093e-01 -4.48541611e-01 -3.44678909e-01
4.29039270e-01 6.90989017e-01 -2.46488988e-01 -1.46851510e-01
-7.01535940e-01 1.14987290e+00 5.46134531e-01 -1.04627227e-02
-6.74382210e-01 -1.64878443e-01 -6.66124448e-02 4.43008870e-01
4.31977421e-01 5.49712837e-01 -7.63499379e-01 -3.77036594e-02
-5.51638186e-01 4.98874575e-01 1.02294469e+00 5.63242972e-01
7.27200806e-01 -4.14038867e-01 -3.45102400e-01 4.53493834e-01
5.38626015e-01 3.89562309e-01 4.70486790e-01 -5.57961822e-01
3.28522950e-01 5.46616137e-01 6.07409894e-01 -1.12217128e+00
1.14376374e-01 -6.71975017e-01 -3.33722472e-01 -1.09528132e-01
3.83242428e-01 -1.28312752e-01 -6.12431169e-01 1.57643330e+00
-9.06949192e-02 9.27462205e-02 -6.18377268e-01 9.66336608e-01
2.07553487e-02 3.10358584e-01 2.23269776e-01 -6.11337602e-01
1.14167309e+00 -5.60955465e-01 -9.98335958e-01 -1.80033296e-01
6.29414201e-01 -6.84595287e-01 1.53834200e+00 5.38033783e-01
-6.83714747e-01 -3.20963532e-01 -8.84714663e-01 5.33870041e-01
-1.92972988e-01 -3.45709115e-01 1.00801337e+00 6.79199815e-01
-2.14966029e-01 9.14596498e-01 -3.79271805e-01 1.26740644e-02
3.22365373e-01 1.63543925e-01 3.95867616e-01 -2.19436929e-01
-1.40072215e+00 5.19981682e-01 -3.07818353e-01 -2.00987816e-01
-5.01590729e-01 -1.05436683e+00 2.15488337e-02 3.09620470e-01
7.77427971e-01 -5.37573278e-01 1.45374966e+00 -9.28033710e-01
-9.65695441e-01 -3.68504375e-02 -1.38837710e-01 -4.21466321e-01
6.43549562e-01 -4.90584880e-01 -7.39384115e-01 -8.04088175e-01
2.46070132e-01 -3.58026117e-01 8.82390261e-01 -1.36025512e+00
-9.91817772e-01 -6.42167091e-01 1.23675831e-01 -5.28917909e-02
-3.06232125e-01 -3.67402017e-01 -2.88561553e-01 -4.33624208e-01
5.04937246e-02 -9.61685717e-01 -2.94710100e-01 -5.05468488e-01
-6.56883001e-01 -3.76035720e-01 6.29712403e-01 -5.22385776e-01
1.83902955e+00 -2.07333803e+00 -8.81137967e-01 4.31942791e-01
2.63553798e-01 -9.77583695e-03 1.18077219e-01 4.23383594e-01
-1.84243510e-03 6.27508163e-01 6.39386296e-01 2.79982358e-01
1.54599473e-01 -3.32097299e-02 -7.14008868e-01 2.04615042e-01
-3.97307336e-01 7.89058387e-01 -7.19129920e-01 1.01280577e-01
-1.47401720e-01 4.78066802e-02 -9.88892078e-01 3.06179494e-01
-3.54032069e-01 3.04604396e-02 -8.24429572e-01 2.30048805e-01
6.73228502e-01 -6.08784139e-01 2.20854327e-01 -2.33753875e-01
-2.85514534e-01 7.27447927e-01 -1.10319960e+00 8.55123281e-01
-5.39121926e-01 1.62474588e-01 -4.01083469e-01 -5.60602188e-01
5.74028075e-01 3.54945548e-02 3.65663290e-01 -1.04967368e+00
-4.12553586e-02 3.79485898e-02 2.73912638e-01 -6.54472530e-01
4.16214138e-01 -1.59794703e-01 8.56004879e-02 6.61002278e-01
-6.31529927e-01 4.42071885e-01 -8.07126239e-02 4.38167572e-01
1.16053653e+00 -2.70110846e-01 1.28351763e-01 5.73144294e-02
-3.97808244e-03 -1.66945890e-01 6.32074952e-01 1.10872507e+00
1.38020322e-01 8.82178992e-02 6.35319650e-01 -2.14556605e-01
-9.15389359e-01 -1.01575351e+00 1.32268801e-01 1.33341980e+00
7.03633279e-02 -5.44135034e-01 -3.32050323e-01 -8.71724069e-01
3.72957289e-01 1.46307731e+00 -6.43223107e-01 -3.20670724e-01
-3.21405157e-02 -6.09612346e-01 -3.71567011e-01 1.55401245e-01
1.98808551e-01 -4.65331316e-01 3.00876033e-02 1.40497386e-01
-3.18482816e-01 -5.26544929e-01 -7.06071198e-01 -3.79622936e-01
-7.15581179e-01 -9.65047002e-01 -2.04153284e-01 1.74057439e-01
4.15206909e-01 6.26092255e-01 1.00002694e+00 2.14016348e-01
3.66195440e-02 1.21053249e-01 -4.37856346e-01 -5.56414366e-01
-6.85787108e-03 -3.31648082e-01 1.02868892e-01 3.71989161e-01
5.16115069e-01 -3.97310197e-01 -1.20629978e+00 7.24754035e-01
-5.94683528e-01 -1.49528816e-01 5.81079662e-01 7.20146954e-01
3.71527225e-01 2.31515944e-01 8.36703360e-01 -1.32605100e+00
9.32570040e-01 -8.32258165e-01 -5.42948484e-01 1.51777789e-01
-1.59003830e+00 -1.54846281e-01 5.12423158e-01 -5.90823174e-01
-1.28362870e+00 -3.04682553e-01 4.12258282e-02 6.11792020e-02
1.81453466e-01 7.41964400e-01 -8.58739242e-02 3.30473304e-01
6.36258900e-01 -1.55923918e-01 -3.17871690e-01 -8.64444971e-01
4.82301027e-01 7.39364386e-01 -1.24196090e-01 -1.02060340e-01
7.60398567e-01 4.13279831e-01 -2.73583919e-01 -2.86487252e-01
-1.24644709e+00 -4.75918651e-01 8.54367167e-02 -3.00599754e-01
4.52128440e-01 -5.74987710e-01 -1.17973506e+00 -4.43849057e-01
-7.50802994e-01 2.25900963e-01 -1.52515918e-01 7.67607272e-01
-3.44387233e-01 3.51385884e-02 -2.84991980e-01 -1.03365755e+00
1.55391499e-01 -9.23305571e-01 3.30917418e-01 2.43245915e-01
-1.72542050e-01 -7.53488898e-01 2.64342278e-02 5.89345276e-01
5.60502291e-01 -4.09257591e-01 1.04010129e+00 -1.09953094e+00
-8.49412799e-01 -9.39713061e-01 -2.35098049e-01 1.90910742e-01
1.38854325e-01 -1.09270886e-01 -7.77909398e-01 -1.53461501e-01
1.06448069e-01 2.89778471e-01 5.28596103e-01 5.41329026e-01
1.39417124e+00 -8.52147460e-01 -5.51970541e-01 -1.70579284e-01
1.48962021e+00 3.11437488e-01 6.65149629e-01 1.71144679e-01
3.94375086e-01 6.10518932e-01 1.14209139e+00 7.60257304e-01
3.31312388e-01 7.50532687e-01 6.11286819e-01 1.85234547e-01
2.78486848e-01 -9.80113029e-01 2.30929941e-01 4.18476224e-01
-1.06293514e-01 -3.64534587e-01 -2.31703401e-01 1.63066089e-01
-2.00128651e+00 -9.08750653e-01 -4.93775964e-01 2.83055639e+00
4.70950991e-01 3.14612716e-01 1.64216578e-01 -5.85037097e-02
7.13917077e-01 -2.20670134e-01 -5.46768129e-01 -1.77831978e-01
2.99519747e-01 -2.59099245e-01 8.38553369e-01 4.58321899e-01
-3.75390917e-01 5.31012356e-01 5.73207140e+00 7.36285031e-01
-9.48663890e-01 4.25948501e-02 7.51957178e-01 -2.06972405e-01
-7.82007992e-01 2.57737279e-01 -9.82513428e-01 8.06087375e-01
1.02257204e+00 -6.88630521e-01 5.17759025e-01 1.00261509e+00
9.92729068e-01 -1.09523460e-01 -1.17107892e+00 5.51423550e-01
-2.27193326e-01 -1.29217410e+00 6.63181245e-02 6.30738258e-01
7.09348798e-01 -1.41722217e-01 3.61406028e-01 4.07774866e-01
5.42271256e-01 -6.06098235e-01 4.60386276e-01 7.73930192e-01
1.82221204e-01 -5.15163898e-01 7.50011086e-01 6.15341187e-01
-5.18050730e-01 -5.70343614e-01 -2.53167242e-01 -3.86790037e-01
2.74828255e-01 9.60951686e-01 -1.04310715e+00 2.38513157e-01
5.51580131e-01 4.06303048e-01 -4.16243136e-01 1.32234561e+00
-1.19163170e-01 9.32058930e-01 1.59752607e-01 -3.33647013e-01
-2.54278392e-01 -3.45307767e-01 3.45596075e-01 5.79803765e-01
4.80888754e-01 6.05548173e-02 -5.62069528e-02 9.66750920e-01
-1.61334038e-01 2.95971274e-01 -6.20647192e-01 -2.31210694e-01
2.94434249e-01 1.14046061e+00 -2.84358919e-01 -5.16596138e-02
-8.13505352e-01 6.46857560e-01 -1.08210407e-01 5.40229023e-01
-7.33509064e-01 3.31588537e-02 5.85901320e-01 7.84759402e-01
2.55509287e-01 2.79782176e-01 -5.83934426e-01 -1.06725907e+00
-5.03381677e-02 -6.51113033e-01 3.38670850e-01 -7.39248574e-01
-1.65370142e+00 -8.70723054e-02 -4.35405105e-01 -1.15301621e+00
7.22339898e-02 -3.92435044e-01 -7.20630109e-01 8.44902098e-01
-1.32308507e+00 -6.48597360e-01 8.24933201e-02 4.27577913e-01
4.58473742e-01 2.57538855e-01 3.53721231e-01 5.09927273e-01
-3.24236989e-01 5.43178439e-01 7.98841640e-02 -3.73330057e-01
7.26847947e-01 -1.10753548e+00 7.72600099e-02 3.81086051e-01
1.71490893e-01 1.07932460e+00 1.09261429e+00 -1.09726346e+00
-1.40462232e+00 -8.18492532e-01 1.15718484e+00 -5.83935797e-01
1.00339401e+00 -1.56819895e-01 -7.49515712e-01 7.35951960e-01
-2.12487444e-01 -5.47182500e-01 9.45789695e-01 7.64907420e-01
-3.12415600e-01 -2.20193520e-01 -9.08874393e-01 9.24949527e-01
7.92237103e-01 -2.59090275e-01 -4.81821448e-01 5.38320780e-01
9.06854689e-01 5.86702645e-01 -6.87671363e-01 2.22934783e-01
7.95353532e-01 -1.04721403e+00 7.27747798e-01 -1.06114841e+00
4.79769260e-01 -5.34073496e-03 -3.43919367e-01 -1.34165847e+00
-6.25409126e-01 -5.20049453e-01 1.53249688e-02 1.03458977e+00
9.16509628e-01 -5.76265633e-01 9.71483707e-01 1.24363387e+00
1.72632217e-01 -7.09502280e-01 -4.69787896e-01 -7.01762915e-01
-3.52185547e-01 -7.05831826e-01 7.10913420e-01 7.53442526e-01
2.58546919e-01 5.63711584e-01 -6.82152987e-01 9.10690520e-04
5.97638905e-01 2.88972259e-01 8.27897251e-01 -1.44942617e+00
-4.72488612e-01 -2.42770106e-01 3.16608638e-01 -1.39376152e+00
-3.57334107e-01 -5.72516024e-01 -2.34001592e-01 -1.19969678e+00
4.75892425e-01 -5.15965998e-01 -5.61790586e-01 1.38979778e-02
-2.08740905e-01 -1.71449482e-01 6.33466393e-02 4.47278112e-01
-6.41894877e-01 2.38795817e-01 1.48982155e+00 1.98403150e-01
-2.92727262e-01 9.65888202e-01 -1.23504829e+00 7.60443747e-01
5.79485714e-01 -4.24240559e-01 -3.14361930e-01 2.23961651e-01
9.23252642e-01 2.49663398e-01 2.50198394e-01 9.70670767e-03
9.11200121e-02 -5.95633805e-01 1.21489763e-01 -5.90149343e-01
1.73633024e-01 -1.02288973e+00 1.68222204e-01 4.72978950e-01
-8.06401372e-01 -8.82620960e-02 -5.63335001e-01 1.24429166e+00
1.84066415e-01 -7.70718530e-02 1.64881408e-01 -9.34569016e-02
-1.60463437e-01 4.33124393e-01 -2.23043665e-01 -4.30491418e-01
6.02517486e-01 5.23273684e-02 -3.74394596e-01 -8.52210224e-01
-7.86681831e-01 1.99779868e-02 1.44245267e-01 6.40828788e-01
4.66405839e-01 -1.29604149e+00 -3.92686039e-01 3.84295695e-02
2.36053690e-01 -9.99755561e-01 3.56118470e-01 7.72881866e-01
4.81510371e-01 6.72471642e-01 3.69944036e-01 -1.89758334e-02
-1.00766766e+00 7.18956828e-01 -9.48213339e-02 -3.23994458e-01
-5.67261279e-02 5.60090065e-01 4.60669249e-01 1.44563004e-01
6.22892864e-02 8.42520446e-02 -2.59957999e-01 -1.07424207e-01
5.36952257e-01 6.05743647e-01 -5.73242782e-03 1.39795907e-03
2.21884158e-02 -2.08032385e-01 -5.84938169e-01 -5.04005998e-02
1.18769729e+00 -4.65249330e-01 1.35886848e-01 5.25778055e-01
9.68240142e-01 3.99162382e-01 -1.10798621e+00 -2.40315154e-01
7.85565525e-02 -1.26335406e+00 2.08777517e-01 -1.22719455e+00
-1.01196718e+00 4.77646440e-01 5.76138020e-01 9.87909794e-01
6.04153037e-01 4.19241488e-02 7.36925364e-01 2.02275664e-01
3.08355391e-01 -1.20768094e+00 1.96417451e-01 -1.65481865e-01
7.57636428e-01 -1.39440358e+00 6.47046268e-02 -5.70967793e-01
-7.80676961e-01 5.50926864e-01 5.94645262e-01 5.44651598e-03
1.07282722e+00 -3.18509430e-01 -3.14561814e-01 -1.82523414e-01
-1.10740876e+00 1.15249760e-01 3.33774805e-01 1.46611005e-01
6.51135147e-01 5.81504285e-01 -9.45659161e-01 1.24853754e+00
-1.39832526e-01 1.07097261e-01 5.98149896e-01 4.22294050e-01
-5.67361653e-01 -1.09923828e+00 2.04257842e-04 1.30753636e+00
-7.92413712e-01 -2.71824896e-01 -1.84471324e-01 5.55434585e-01
-8.06513503e-02 1.21572185e+00 -5.16646616e-02 -8.52965891e-01
6.55693710e-01 -6.31889477e-02 -1.64857462e-01 -5.91104746e-01
-1.86808363e-01 3.19828838e-01 1.95174485e-01 -7.14208543e-01
1.50201648e-01 -6.72514915e-01 -8.51405799e-01 -6.43988967e-01
-9.08436000e-01 3.78024668e-01 9.22279000e-01 9.91637588e-01
4.25841033e-01 3.81987721e-01 1.23718667e+00 -1.14690445e-01
-1.03174531e+00 -8.21637154e-01 -9.05427754e-01 7.94988573e-01
1.34611815e-01 -7.21877873e-01 -8.38905275e-01 -2.34763950e-01] | [9.83612060546875, 5.578089237213135] |
db7c9ae4-c168-4435-98d4-911c22f4ee7e | suggestion-mining-from-opinionated-text | null | null | https://aclanthology.org/P16-3018 | https://aclanthology.org/P16-3018.pdf | Suggestion Mining from Opinionated Text | null | ['Sapna Negi'] | 2016-08-01 | null | null | null | acl-2016-8 | ['suggestion-mining'] | ['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] | [-7.208926677703857, 3.8354179859161377] |
ab072d0d-f89c-4320-9029-0e794b1a5be9 | cuet-nlp-dravidianlangtech-acl2022 | null | null | https://aclanthology.org/2022.dravidianlangtech-1.27 | https://aclanthology.org/2022.dravidianlangtech-1.27.pdf | CUET-NLP@DravidianLangTech-ACL2022: Investigating Deep Learning Techniques to Detect Multimodal Troll Memes | With the substantial rise of internet usage, social media has become a powerful communication medium to convey information, opinions, and feelings on various issues. Recently, memes have become a popular way of sharing information on social media. Usually, memes are visuals with text incorporated into them and quickly disseminate hatred and offensive content. Detecting or classifying memes is challenging due to their region-specific interpretation and multimodal nature. This work presents a meme classification technique in Tamil developed by the CUET NLP team under the shared task (DravidianLangTech-ACL2022). Several computational models have been investigated to perform the classification task. This work also explored visual and textual features using VGG16, ResNet50, VGG19, CNN and CNN+LSTM models. Multimodal features are extracted by combining image (VGG16) and text (CNN, LSTM+CNN) characteristics. Results demonstrate that the textual strategy with CNN+LSTM achieved the highest weighted f_1-score (0.52) and recall (0.57). Moreover, the CNN-Text+VGG16 outperformed the other models concerning the multimodal memes detection by achieving the highest f_1-score of 0.49, but the LSTM+CNN model allowed the team to achieve 4^{th} place in the shared task. | ['Mohammed Moshiul Hoque', 'Omar Sharif', 'Eftekhar Hossain', 'Nusratul Jannat', 'Md Hasan'] | null | null | null | null | dravidianlangtech-acl-2022-5 | ['meme-classification'] | ['natural-language-processing'] | [-2.94065356e-01 -1.45389974e-01 5.60130998e-02 4.03789520e-01
-5.07082701e-01 -5.07775187e-01 1.02683949e+00 5.17825007e-01
-6.44970596e-01 6.21518910e-01 2.50917792e-01 -3.31617221e-02
2.83788264e-01 -6.31506562e-01 -3.16949248e-01 -4.68616396e-01
5.88057674e-02 -1.29930809e-01 5.81314228e-03 -5.01690805e-01
7.52274096e-01 1.18507691e-01 -1.43878353e+00 7.94223905e-01
7.67414093e-01 1.17344499e+00 1.84419885e-01 7.45676696e-01
-5.66062629e-01 1.21732640e+00 -8.78894627e-01 -5.84092855e-01
-2.70043850e-01 -2.50669628e-01 -6.75059676e-01 -3.13356332e-02
5.99867880e-01 -9.06101242e-02 -3.86331469e-01 9.24200177e-01
5.95113397e-01 2.30062813e-01 6.29193723e-01 -1.20620048e+00
-1.04358041e+00 3.56369972e-01 -7.57973135e-01 3.73082727e-01
4.05555964e-01 -1.72894657e-01 5.67177832e-01 -1.13376629e+00
8.53784323e-01 1.21810150e+00 5.43202400e-01 2.40499750e-01
-8.46462786e-01 -6.28744781e-01 -1.75448880e-01 2.22324729e-01
-1.33214355e+00 1.51642874e-01 6.35807037e-01 -7.20780730e-01
1.05712688e+00 1.03814773e-01 6.43937826e-01 1.46026695e+00
5.84158361e-01 9.26145792e-01 1.41283238e+00 -2.30432659e-01
-1.01237424e-01 6.70935214e-01 1.28086060e-01 9.05574679e-01
-2.54537284e-01 -5.38967133e-01 -9.39494193e-01 -2.80472040e-01
3.24519902e-01 1.78089365e-01 1.46422451e-02 4.90057856e-01
-1.19385183e+00 1.12957013e+00 6.86735570e-01 7.84084857e-01
-3.30228180e-01 2.07642280e-02 9.22522545e-01 4.72568333e-01
1.23801446e+00 6.02689028e-01 1.62326664e-01 -4.23273928e-02
-9.53432620e-01 7.79988095e-02 5.20134449e-01 5.13238490e-01
6.82940483e-01 1.40914442e-02 -4.11017388e-01 1.20762932e+00
4.53186594e-02 8.30131888e-01 5.46828210e-01 -2.97782540e-01
6.01108670e-01 8.78460288e-01 -1.19905360e-01 -2.09839487e+00
-5.63274801e-01 -4.17700946e-01 -9.42391813e-01 -1.74736798e-01
7.36926720e-02 -2.51757056e-01 -8.36148798e-01 1.24410856e+00
-9.90510136e-02 -4.09515083e-01 -3.56958121e-01 6.34314120e-01
1.25362110e+00 1.16223300e+00 3.60662133e-01 9.34150144e-02
1.21206176e+00 -7.50813186e-01 -9.99646842e-01 -6.48427457e-02
8.27424347e-01 -1.06081617e+00 7.63372719e-01 3.35155785e-01
-8.00119758e-01 -5.03063619e-01 -7.78208852e-01 -3.68340872e-02
-1.14039421e+00 2.29627669e-01 2.16007426e-01 5.29510856e-01
-1.13737237e+00 3.93947542e-01 -1.02392785e-01 -9.18984115e-01
4.88799781e-01 1.02907427e-01 -5.03444552e-01 3.03117394e-01
-1.23347700e+00 1.13596070e+00 4.05813158e-01 1.42734066e-01
-6.83047414e-01 -3.06608200e-01 -6.71356738e-01 -4.71644789e-01
-2.06422247e-03 -1.50639042e-01 4.50476706e-01 -1.09825408e+00
-9.80680645e-01 1.28155363e+00 1.86486572e-01 -2.79123664e-01
5.28288484e-01 -3.45953166e-01 -6.29791677e-01 3.15495551e-01
1.56192198e-01 9.19167638e-01 9.07840014e-01 -1.18193591e+00
-4.33584541e-01 -3.28288823e-01 -9.14319828e-02 -4.20145243e-02
-1.11867535e+00 4.57198977e-01 -2.52229851e-02 -7.58296967e-01
-8.54377970e-02 -9.66700613e-01 3.04774880e-01 -3.12680781e-01
-7.49220312e-01 -3.05943489e-01 1.28511560e+00 -1.18034768e+00
1.27689922e+00 -2.12517881e+00 9.43152308e-02 1.65858835e-01
6.87270582e-01 4.68667984e-01 -5.94168790e-02 9.46223855e-01
3.96328062e-01 4.53791410e-01 6.02953993e-02 -5.22166789e-01
-1.20756738e-01 -2.45267138e-01 -2.70546615e-01 5.57758033e-01
2.12659035e-02 8.85472119e-01 -7.21709549e-01 -6.94375813e-01
1.93579316e-01 6.82343662e-01 -3.69283184e-02 4.56632636e-02
3.00263404e-03 1.55361503e-01 -4.23095554e-01 5.30885696e-01
5.41770339e-01 -4.25140619e-01 -2.62876749e-02 -2.85371915e-02
-4.40953195e-01 -4.68304902e-01 -4.54955846e-01 1.16108644e+00
-4.39332932e-01 1.49598551e+00 -3.50812115e-02 -4.36546594e-01
1.31838679e+00 3.21757317e-01 2.04630852e-01 -9.03452158e-01
6.48858726e-01 1.56370863e-01 -6.15292609e-01 -8.90454888e-01
9.38082814e-01 5.27436510e-02 -2.36626580e-01 3.81561220e-01
-1.75736528e-02 2.78942287e-01 1.07954331e-02 4.84903783e-01
7.60382235e-01 -1.95382074e-01 9.53856558e-02 -2.36033246e-01
4.42027807e-01 1.32643819e-01 -1.89131916e-01 7.14853168e-01
-2.65360236e-01 5.16345799e-01 5.41852415e-01 -4.71098930e-01
-1.08421361e+00 -4.33459103e-01 1.72234744e-01 1.35547626e+00
-2.53716558e-02 -1.20136812e-01 -7.67343044e-01 -5.07229567e-01
-1.02913946e-01 4.01866674e-01 -1.09816909e+00 3.22661214e-02
-2.63611168e-01 -6.79738224e-01 7.76311338e-01 1.23946376e-01
1.05625927e+00 -1.37549686e+00 -3.87581736e-01 4.64144628e-03
-5.52797079e-01 -1.04136312e+00 -1.53559402e-01 -1.67797193e-01
-4.75128502e-01 -8.72256398e-01 -1.18081033e+00 -7.77137041e-01
4.68811423e-01 3.56111526e-01 5.62674761e-01 2.79018164e-01
-9.09260064e-02 3.00233513e-01 -6.19687080e-01 -3.07902515e-01
-2.84983665e-01 3.59019935e-01 -3.14124286e-01 3.43329161e-01
2.62341559e-01 -5.03630796e-03 -3.08857471e-01 1.31909326e-01
-1.08672059e+00 2.45874614e-01 2.21972644e-01 6.97564304e-01
-1.58523783e-01 -3.43160540e-01 5.69443941e-01 -8.31009984e-01
8.89200687e-01 -9.58968580e-01 1.40499711e-01 3.28421108e-02
-1.25669271e-01 -6.68525100e-01 4.11068648e-01 -6.54503763e-01
-9.41857338e-01 -5.82477212e-01 2.03565061e-01 -4.52112347e-01
-2.62119826e-02 7.58425355e-01 4.39755350e-01 -1.56623125e-01
6.68065786e-01 1.07107289e-01 -3.78952250e-02 -3.79694998e-01
-6.14290386e-02 9.13407505e-01 2.80146487e-02 9.29986164e-02
5.13205171e-01 4.09539372e-01 -3.13346893e-01 -1.29275596e+00
-9.32956934e-01 -4.49423432e-01 -3.82127017e-01 -8.61295938e-01
1.25209498e+00 -7.45439589e-01 -8.61865461e-01 9.76472497e-01
-1.39675879e+00 1.44849233e-02 6.23938024e-01 2.90726572e-01
5.99929225e-03 2.67523497e-01 -1.00300860e+00 -1.08907950e+00
-4.21596825e-01 -6.69641018e-01 7.32889473e-01 9.42751467e-02
-3.64183605e-01 -1.30151868e+00 -5.63890524e-02 6.03445172e-01
7.25264490e-01 8.96469593e-01 7.73180366e-01 -8.78611088e-01
8.03232118e-02 -2.45346919e-01 -6.71770513e-01 3.13015640e-01
-2.70853996e-01 -1.26038631e-02 -1.17580938e+00 -2.59529710e-01
-2.65614748e-01 -5.31850636e-01 1.11261320e+00 1.21586286e-01
8.71419251e-01 -4.55993384e-01 -1.61215529e-01 -7.02026337e-02
1.36989284e+00 1.42459810e-01 7.46680617e-01 8.10406029e-01
1.03999007e+00 8.87954295e-01 3.47924680e-01 5.81119895e-01
1.53043881e-01 3.99042457e-01 5.95131695e-01 -2.25863591e-01
9.82961208e-02 -2.61351228e-01 5.33027709e-01 9.43243206e-01
-1.66321665e-01 -6.06561601e-01 -1.18988955e+00 4.61572111e-01
-1.88994086e+00 -1.11535239e+00 -6.13377333e-01 1.66925418e+00
1.97812498e-01 -1.31293669e-01 3.05009872e-01 1.90383479e-01
1.21537626e+00 4.89225417e-01 9.27178189e-02 -6.64260089e-01
-5.72044969e-01 -3.41220438e-01 4.55267817e-01 2.64140666e-01
-1.20747530e+00 1.06036448e+00 5.66157198e+00 1.05566204e+00
-1.40531528e+00 4.86794502e-01 7.71284938e-01 2.35299747e-02
2.62633879e-02 -6.12349629e-01 -5.59522092e-01 7.22301364e-01
9.87695932e-01 2.44664833e-01 1.55833572e-01 4.01616603e-01
2.46358916e-01 -3.51598978e-01 -2.99358457e-01 9.72352803e-01
6.56175911e-01 -1.55559075e+00 2.40615070e-01 2.15814024e-01
8.86087179e-01 6.39902726e-02 5.38178980e-01 2.62882769e-01
-3.75878811e-01 -1.09295380e+00 8.28055322e-01 6.50620401e-01
5.80114126e-01 -8.79545987e-01 1.04651475e+00 1.91268593e-01
-7.52947628e-01 -2.14551970e-01 -2.94935226e-01 -1.07256502e-01
-3.47899017e-03 4.66439009e-01 -7.52447307e-01 2.82592803e-01
8.64430189e-01 9.58218157e-01 -9.57049906e-01 7.32686222e-01
7.91810229e-02 6.01052284e-01 1.86728388e-01 -8.16192508e-01
6.80747271e-01 8.97708163e-02 7.10232913e-01 1.83513570e+00
2.04563230e-01 -2.66761392e-01 1.70862511e-01 7.20961094e-01
-1.06772840e-01 5.16583323e-01 -7.45863974e-01 -5.95530570e-01
1.56030208e-01 1.47272539e+00 -1.00315166e+00 -2.43560180e-01
-7.73903206e-02 8.98728311e-01 2.57670850e-01 3.13429713e-01
-9.25726473e-01 -5.61953962e-01 -1.83895722e-01 1.47108153e-01
-1.34857103e-01 -2.00039759e-01 -2.23724112e-01 -7.21086442e-01
-3.13172370e-01 -6.43755555e-01 3.41188610e-01 -1.07862151e+00
-1.42302144e+00 9.46541965e-01 -1.72418356e-01 -1.06407702e+00
1.85624242e-01 -7.66294420e-01 -4.86718804e-01 6.55887604e-01
-9.34410155e-01 -1.55100322e+00 -4.43285167e-01 5.60549915e-01
4.28200603e-01 -5.06186783e-01 5.70585132e-01 3.72248918e-01
-6.80685937e-01 3.19002926e-01 5.19791879e-02 3.11866254e-01
6.99040294e-01 -8.23558569e-01 -2.35253856e-01 3.35442483e-01
-4.10235941e-01 4.77732092e-01 6.94140196e-01 -9.79493856e-01
-1.04961586e+00 -1.07276952e+00 1.10830975e+00 -3.66896182e-01
1.01771188e+00 -2.98813522e-01 -7.70533621e-01 3.61003846e-01
7.60154128e-01 -6.01152778e-01 6.90820515e-01 -5.89531884e-02
-4.35290247e-01 4.64938164e-01 -1.31636393e+00 4.73440498e-01
4.94077891e-01 -8.76313031e-01 -3.12105894e-01 4.59768891e-01
3.86296540e-01 6.11595176e-02 -9.30511236e-01 -2.33161822e-01
4.01290983e-01 -8.42868805e-01 4.19786245e-01 -4.52523828e-01
1.10766029e+00 -3.44034322e-02 -2.08194599e-01 -1.22853124e+00
-1.07046954e-01 -3.75505835e-01 1.35011271e-01 1.37241983e+00
5.63817322e-01 -7.18324184e-01 4.69763070e-01 7.63125643e-02
1.40276097e-04 -4.27911669e-01 -6.76766217e-01 -4.08736259e-01
1.04150595e-02 -2.52666712e-01 -2.05822363e-01 1.47088695e+00
1.16429172e-01 2.55539834e-01 -8.61177623e-01 -3.71763647e-01
2.67659664e-01 -4.43057001e-01 5.26284873e-01 -1.12805927e+00
4.49180335e-01 -5.01828969e-01 -5.53882897e-01 -3.37251753e-01
1.77142486e-01 -9.02804255e-01 -4.30929482e-01 -1.68332946e+00
6.69006348e-01 1.08581536e-01 -3.00151676e-01 5.51828980e-01
1.93666726e-01 8.73820484e-01 4.82395083e-01 2.70825297e-01
-8.02962720e-01 1.46346435e-01 1.37558734e+00 -4.44201440e-01
-1.01308323e-01 -5.83085954e-01 -3.55027735e-01 5.75446069e-01
9.55892205e-01 -3.54754120e-01 1.61251977e-01 -1.91207111e-01
9.14688230e-01 -1.56606600e-01 6.52215660e-01 -8.33305538e-01
2.71604478e-01 9.99743268e-02 6.86927795e-01 -9.17015254e-01
5.31835496e-01 -2.58092731e-01 -1.11523636e-01 5.01794696e-01
-5.99981189e-01 3.69429320e-01 4.86516267e-01 3.23962599e-01
-3.87360364e-01 3.81243117e-02 7.02166617e-01 -1.29640236e-01
-6.57550633e-01 -5.51610477e-02 -1.07862878e+00 -1.15811311e-01
8.08986306e-01 -3.82678568e-01 -9.04480636e-01 -6.87565982e-01
-8.33651543e-01 -1.83656644e-02 3.17992419e-01 8.03608060e-01
7.67127395e-01 -1.22017825e+00 -8.46139431e-01 -4.16187584e-01
2.42280200e-01 -9.32141542e-01 5.83646297e-01 1.08971131e+00
-5.23444533e-01 3.48332465e-01 -5.92190087e-01 -4.58662540e-01
-1.32150996e+00 2.91486204e-01 1.67438313e-01 -2.31685471e-02
-3.09961259e-01 6.71458960e-01 2.87069958e-02 -2.98201889e-01
3.57517190e-02 5.63886523e-01 -6.88228667e-01 7.27212131e-01
6.65795743e-01 9.17611837e-01 1.88025013e-02 -1.12357950e+00
-1.38643056e-01 4.00371730e-01 -1.20718949e-01 -1.41270071e-01
1.36577976e+00 -1.94290593e-01 -4.96896088e-01 7.92630434e-01
1.56800437e+00 -4.03958783e-02 -3.67300183e-01 1.34890527e-01
8.62723142e-02 -4.16544527e-01 3.03535223e-01 -8.58530343e-01
-1.00926995e+00 9.84654069e-01 5.84578395e-01 7.09863007e-01
5.72576821e-01 -2.09196717e-01 9.55293417e-01 2.61227190e-01
3.79852727e-02 -1.53388369e+00 6.47584260e-01 8.29124570e-01
1.08145714e+00 -1.20459974e+00 -1.48907661e-01 -1.35984764e-01
-9.00608301e-01 1.14068842e+00 6.29832566e-01 -3.36811095e-02
4.88783687e-01 -2.55795538e-01 1.75375625e-01 -6.10977709e-01
-4.00001258e-01 3.36466841e-02 6.55512214e-01 3.16002786e-01
7.01256752e-01 -2.73252465e-02 -3.60925972e-01 2.56045330e-02
4.16737981e-02 -4.61842984e-01 5.66927791e-01 8.93087924e-01
-7.38361955e-01 -3.26328963e-01 -5.42500079e-01 4.14207101e-01
-8.28945458e-01 4.19675037e-02 -9.21544492e-01 9.56371725e-01
2.31292263e-01 1.22627211e+00 1.60682604e-01 -8.62577915e-01
-9.37839970e-02 -7.46342242e-02 -8.18318035e-03 -4.03540462e-01
-1.33247852e+00 -1.20090730e-01 2.88626194e-01 -3.29417527e-01
-5.32146335e-01 -3.39993030e-01 -8.85980248e-01 -8.33131969e-01
-2.73018718e-01 9.32239667e-02 1.10874486e+00 8.68202031e-01
2.91683048e-01 3.48082125e-01 6.37079179e-01 -9.55302417e-01
2.62248397e-01 -1.21120584e+00 -6.00954831e-01 4.22571093e-01
1.55745029e-01 -5.62093854e-01 -4.38838333e-01 -2.01523647e-01] | [8.515472412109375, 10.733499526977539] |
d7b3e00c-bf06-4186-a0e0-858dea3ed781 | debiased-contrastive-learning-for-sequential | 2303.11780 | null | https://arxiv.org/abs/2303.11780v1 | https://arxiv.org/pdf/2303.11780v1.pdf | Debiased Contrastive Learning for Sequential Recommendation | Current sequential recommender systems are proposed to tackle the dynamic user preference learning with various neural techniques, such as Transformer and Graph Neural Networks (GNNs). However, inference from the highly sparse user behavior data may hinder the representation ability of sequential pattern encoding. To address the label shortage issue, contrastive learning (CL) methods are proposed recently to perform data augmentation in two fashions: (i) randomly corrupting the sequence data (e.g. stochastic masking, reordering); (ii) aligning representations across pre-defined contrastive views. Although effective, we argue that current CL-based methods have limitations in addressing popularity bias and disentangling of user conformity and real interest. In this paper, we propose a new Debiased Contrastive learning paradigm for Recommendation (DCRec) that unifies sequential pattern encoding with global collaborative relation modeling through adaptive conformity-aware augmentation. This solution is designed to tackle the popularity bias issue in recommendation systems. Our debiased contrastive learning framework effectively captures both the patterns of item transitions within sequences and the dependencies between users across sequences. Our experiments on various real-world datasets have demonstrated that DCRec significantly outperforms state-of-the-art baselines, indicating its efficacy for recommendation. To facilitate reproducibility of our results, we make our implementation of DCRec publicly available at: https://github.com/HKUDS/DCRec. | ['Kangyi Lin', 'Da Luo', 'Chunzhen Huang', 'Lianghao Xia', 'Chao Huang', 'Yuhao Yang'] | 2023-03-21 | null | null | null | null | ['sequential-recommendation'] | ['miscellaneous'] | [ 3.77004802e-01 -5.35466611e-01 -6.95351303e-01 -3.60633790e-01
-2.09514990e-01 -6.37774229e-01 5.06520092e-01 2.33856648e-01
-3.01900446e-01 3.67693305e-01 6.29866302e-01 -5.25716305e-01
-2.95174956e-01 -6.53840423e-01 -7.00687170e-01 -5.89304030e-01
-1.62017360e-01 3.77443522e-01 -2.57369429e-02 -3.41484219e-01
3.48322928e-01 2.28321627e-01 -1.72698390e+00 6.95818067e-01
9.69025195e-01 9.91344929e-01 2.10355908e-01 4.39903975e-01
-1.52090907e-01 9.02254939e-01 -1.85033366e-01 -3.26690555e-01
3.01148951e-01 -3.37892085e-01 -6.42193735e-01 -8.14520270e-02
6.67133391e-01 -3.95794511e-01 -5.45620322e-01 9.19170737e-01
4.92988020e-01 5.65488935e-01 5.49753547e-01 -1.34607399e+00
-1.32584524e+00 1.09317124e+00 -7.19868958e-01 5.46935201e-01
3.52444500e-01 2.42744256e-02 1.52057552e+00 -9.42721128e-01
3.58581543e-01 9.92874444e-01 6.15705431e-01 5.10312498e-01
-1.26081419e+00 -8.35423946e-01 8.32774639e-01 5.24970353e-01
-1.19225073e+00 -2.54375666e-01 9.56539154e-01 -3.85246813e-01
7.54205883e-01 5.63382685e-01 6.01071060e-01 1.67550719e+00
-1.41824931e-01 1.06395495e+00 9.53960538e-01 -1.26275063e-01
1.36966899e-01 -1.78858265e-01 5.76795995e-01 5.26455283e-01
2.72624791e-01 1.37576059e-01 -6.75858378e-01 -3.89016449e-01
7.89796710e-01 5.41830719e-01 -4.34069186e-01 -5.26047826e-01
-1.05923963e+00 7.39423215e-01 3.95802230e-01 1.96017191e-01
-3.89587522e-01 -1.79171145e-01 4.38216805e-01 5.78578889e-01
5.05674601e-01 5.68273365e-01 -6.49681270e-01 6.05830885e-02
-7.26149976e-01 2.32085288e-01 6.39582157e-01 1.02820551e+00
3.30857754e-01 4.23434824e-02 -4.76117313e-01 1.08362508e+00
3.89066696e-01 3.25654298e-01 8.19907188e-01 -5.95401764e-01
4.17502731e-01 4.43074733e-01 -3.42102610e-02 -1.20931816e+00
-3.82152349e-01 -8.36403131e-01 -9.82285261e-01 -4.54832613e-01
3.85813087e-01 1.26785666e-01 -8.59306097e-01 1.89102066e+00
1.25032693e-01 4.58344400e-01 -2.57144451e-01 8.92153502e-01
1.02817655e+00 6.47520602e-01 7.62071013e-02 -3.44044060e-01
1.20149541e+00 -1.23969102e+00 -5.82058609e-01 3.73200290e-02
5.96489549e-01 -4.91886735e-01 1.44247842e+00 5.56329548e-01
-9.18781340e-01 -6.54366553e-01 -9.10751581e-01 1.71412215e-01
-2.34526753e-01 -8.45361724e-02 6.30219758e-01 5.09732783e-01
-8.53845894e-01 4.94974285e-01 -5.51888168e-01 -1.59700707e-01
2.91508585e-01 4.35522825e-01 -1.14843301e-01 -1.61810130e-01
-1.34799469e+00 4.43226784e-01 1.13206245e-01 -1.54474028e-03
-5.40515721e-01 -8.14333677e-01 -7.72533119e-01 2.97447085e-01
7.14879453e-01 -6.10833824e-01 1.13836884e+00 -1.18774831e+00
-1.32261431e+00 3.65295678e-01 6.16107807e-02 -2.65161723e-01
1.06372938e-01 -4.01984572e-01 -8.40739071e-01 -4.47739065e-01
-3.22193176e-01 1.52423203e-01 7.73839474e-01 -1.34983659e+00
-6.37863100e-01 -3.13132912e-01 2.20630035e-01 4.01668876e-01
-7.14330077e-01 -2.70373225e-01 -5.19816875e-01 -1.14303505e+00
4.19211537e-02 -9.39833879e-01 -2.45078385e-01 -3.45292479e-01
-1.88080564e-01 -3.38686675e-01 5.47374308e-01 -4.47067678e-01
1.77870548e+00 -2.07309437e+00 2.41357744e-01 3.00867498e-01
3.16070616e-01 3.30853760e-01 -6.82613194e-01 5.01889467e-01
-6.49517626e-02 -2.21263133e-02 1.48588672e-01 -2.73287922e-01
-7.66028464e-02 1.77603602e-01 -3.80962729e-01 3.99438322e-01
-3.55248719e-01 8.51298451e-01 -1.11807251e+00 -1.68684889e-02
-3.40928584e-02 3.02193701e-01 -1.01913869e+00 4.93795782e-01
-2.43913159e-01 4.20011580e-01 -1.24257341e-01 5.64954042e-01
6.94067776e-01 -6.21543944e-01 7.31657863e-01 -5.11687696e-01
1.11208513e-01 6.85091674e-01 -1.32309294e+00 1.54493034e+00
-2.60248542e-01 -8.17428455e-02 -2.64653265e-01 -8.95831704e-01
7.04967082e-01 6.16959557e-02 5.79594135e-01 -9.65827703e-01
8.22446309e-03 1.39313191e-02 2.21661985e-01 -3.58940095e-01
9.48946536e-01 2.40199149e-01 5.59263676e-02 5.73813260e-01
1.41392738e-01 7.03941584e-01 1.25081599e-01 3.61824870e-01
9.09444988e-01 1.42217904e-01 2.47058690e-01 -2.16341883e-01
4.12049383e-01 -5.21730721e-01 6.28347754e-01 1.01009655e+00
2.96092983e-02 5.14817357e-01 1.41484544e-01 -4.62034404e-01
-8.21321607e-01 -9.22872007e-01 2.52560556e-01 1.78050411e+00
3.90568137e-01 -6.85222983e-01 -1.84156504e-02 -1.02044785e+00
7.16945529e-02 7.17319667e-01 -6.70214295e-01 -4.07970101e-01
-6.60629272e-01 -8.09413850e-01 1.43852204e-01 6.25481784e-01
2.98297126e-02 -1.06276608e+00 1.28348574e-01 2.46489316e-01
-2.55110204e-01 -7.58050680e-01 -1.06550944e+00 1.35029390e-01
-7.91112483e-01 -9.24774468e-01 -5.02226293e-01 -6.61459684e-01
5.58125377e-01 9.41774547e-01 1.31523645e+00 2.25681335e-01
4.06515986e-01 4.31791186e-01 -8.60424101e-01 2.04648256e-01
-1.20637566e-01 2.73884296e-01 3.10366094e-01 2.04611868e-01
3.44914854e-01 -8.94661844e-01 -9.39595819e-01 5.36515772e-01
-1.02772117e+00 9.81859937e-02 4.58669931e-01 9.48368967e-01
5.21446049e-01 -2.26659626e-01 5.48945904e-01 -1.44125056e+00
8.61493349e-01 -9.89380538e-01 -2.12749004e-01 3.13186526e-01
-1.12698364e+00 1.91573948e-02 8.23675513e-01 -8.77715409e-01
-8.97558093e-01 -4.68213350e-01 -1.86044127e-01 -5.28478563e-01
-2.00603884e-02 8.23899984e-01 -2.17897631e-02 3.42820168e-01
6.04549587e-01 4.08504665e-01 -1.61848083e-01 -6.76585317e-01
6.02317154e-01 6.89258099e-01 2.83220112e-01 -7.11278021e-01
4.38899279e-01 8.74218643e-02 -4.84727889e-01 -3.79711956e-01
-9.42938566e-01 -6.94567323e-01 -3.21703166e-01 -8.96289945e-02
1.12739325e-01 -8.61692786e-01 -5.80998778e-01 3.76081437e-01
-4.51438487e-01 -4.34038460e-01 -2.63923317e-01 3.17168474e-01
-2.38196820e-01 6.95340395e-01 -8.24868560e-01 -6.64242446e-01
-4.95825619e-01 -9.35327768e-01 5.61153948e-01 8.24892372e-02
-3.08527410e-01 -9.05421376e-01 1.68366611e-01 5.10120273e-01
5.78252137e-01 -3.93694371e-01 9.31992114e-01 -1.08735180e+00
-2.73132890e-01 8.43222141e-02 -6.58712983e-02 1.70422941e-01
3.39462399e-01 -2.34691754e-01 -5.91002762e-01 -6.86672032e-01
-3.16560477e-01 -3.75774831e-01 7.91268051e-01 2.44988114e-01
1.67965066e+00 -6.99558735e-01 -1.78573057e-01 5.54606736e-01
1.08227229e+00 2.45200351e-01 4.62491214e-01 2.96154141e-01
1.06137276e+00 4.08746868e-01 5.64626873e-01 5.55982709e-01
6.95796072e-01 7.72082746e-01 4.92500097e-01 6.53107390e-02
-1.95772097e-01 -5.36136448e-01 3.23095083e-01 1.25212264e+00
-1.97301075e-01 -8.04531455e-01 -4.73610133e-01 3.49656224e-01
-2.18154049e+00 -1.01891041e+00 -1.04909308e-01 2.23567367e+00
8.01980793e-01 3.24702375e-02 4.26141322e-01 -3.38484049e-02
6.85949326e-01 4.15206701e-01 -5.73015273e-01 -1.98303193e-01
-6.22948743e-02 -1.52880207e-01 3.49573761e-01 2.44864255e-01
-1.03911924e+00 6.60725236e-01 5.46262598e+00 8.13253999e-01
-1.21396458e+00 8.08611512e-02 6.00105345e-01 -2.10467905e-01
-6.46021903e-01 -4.09957677e-01 -6.65028095e-01 7.79550076e-01
7.47796416e-01 2.21288696e-01 7.54183292e-01 6.18340015e-01
-1.67440064e-02 5.26866317e-01 -1.06856263e+00 1.11596024e+00
2.70248353e-01 -1.35601878e+00 4.14139450e-01 4.30379882e-02
8.68809402e-01 4.94752936e-02 4.78809744e-01 5.51933348e-01
5.20255804e-01 -7.00634837e-01 6.96890473e-01 4.31823134e-01
6.31072104e-01 -5.94715416e-01 6.15696251e-01 2.92310625e-01
-1.15549910e+00 -3.47959638e-01 -1.62370980e-01 -1.11461654e-01
3.83082144e-02 4.61963534e-01 -4.53194916e-01 7.13720918e-01
6.83332801e-01 1.11386955e+00 -5.28557718e-01 9.26986039e-01
6.11728653e-02 9.45198953e-01 4.33027521e-02 4.02424335e-02
4.79911864e-02 -3.06291997e-01 4.65807885e-01 1.21250653e+00
2.83828855e-01 3.83570492e-02 3.31990540e-01 4.93914694e-01
-2.77149379e-01 3.47474307e-01 -2.54127502e-01 -1.84925459e-02
6.42512977e-01 1.18152797e+00 -5.08759499e-01 -2.75922537e-01
-7.21856356e-01 8.82656395e-01 5.76629996e-01 3.80261868e-01
-8.53756547e-01 2.69211173e-01 6.34978414e-01 1.74364313e-01
6.16446018e-01 -1.42862305e-01 -1.13180414e-01 -1.49163330e+00
-2.07350373e-01 -1.30065882e+00 8.92990351e-01 -3.13380182e-01
-1.96645844e+00 5.95546663e-01 -1.14637371e-02 -1.42866385e+00
9.95704606e-02 -3.75550359e-01 -3.78053755e-01 4.40781444e-01
-1.38563848e+00 -1.13348377e+00 -1.10020511e-01 6.41527355e-01
8.05857122e-01 -2.37336367e-01 5.93907595e-01 5.87961972e-01
-6.49458349e-01 1.15671766e+00 1.89731225e-01 -2.63013214e-01
7.92711377e-01 -1.21312726e+00 3.80087793e-01 6.63566351e-01
3.95190299e-01 9.64216053e-01 6.12974167e-01 -5.54376841e-01
-1.59098613e+00 -1.11154675e+00 4.78671521e-01 -3.49269420e-01
6.10915661e-01 -3.42767715e-01 -1.16218114e+00 6.10781074e-01
7.61635378e-02 3.88592407e-02 1.13857210e+00 5.79718232e-01
-8.18551302e-01 -1.01020925e-01 -7.32106388e-01 8.21912408e-01
1.51482356e+00 -3.19189399e-01 -4.76227641e-01 4.32808101e-02
6.48998737e-01 -2.53573865e-01 -8.36386025e-01 6.19093895e-01
8.31539512e-01 -8.09274077e-01 1.08358359e+00 -8.66220176e-01
3.91878426e-01 -2.78548867e-01 -3.29211861e-01 -1.32545125e+00
-9.35226142e-01 -5.15593827e-01 -8.47522199e-01 1.23792207e+00
4.53797907e-01 -5.68298459e-01 6.84982300e-01 3.43274713e-01
-8.12301338e-02 -9.33294058e-01 -3.18365782e-01 -6.79192126e-01
-2.16033190e-01 -2.21398473e-01 7.68451333e-01 1.46381414e+00
1.05772100e-01 5.57082534e-01 -1.05466855e+00 5.53529002e-02
3.13265562e-01 4.54430193e-01 5.25339901e-01 -1.11311829e+00
-6.37776971e-01 -5.22099972e-01 1.78792372e-01 -1.53780794e+00
-4.81507033e-02 -1.10868430e+00 -2.09692135e-01 -1.31994379e+00
4.55360681e-01 -6.24140382e-01 -1.01543772e+00 2.97083884e-01
-4.70391750e-01 1.86303362e-01 -3.15959491e-02 4.57043022e-01
-9.27956104e-01 6.22001648e-01 1.38024223e+00 -1.66458756e-01
-4.09525752e-01 1.14927150e-01 -1.09750199e+00 3.40833157e-01
8.74216616e-01 -3.26544613e-01 -8.36076140e-01 -3.09900522e-01
4.49030817e-01 -2.37006724e-01 -1.42800301e-01 -5.13490856e-01
3.10422748e-01 -2.81520426e-01 2.32603505e-01 -5.28049529e-01
2.20532399e-02 -7.34818757e-01 1.71910957e-01 2.62568802e-01
-8.33166122e-01 2.83517241e-01 -4.76320833e-02 9.15799558e-01
1.34906590e-01 1.31447732e-01 3.72164458e-01 2.82334685e-02
-7.90751278e-01 6.82385504e-01 -4.28737342e-01 2.04385370e-02
3.77818227e-01 -1.65885519e-02 -4.80481952e-01 -5.00808954e-01
-6.41630828e-01 2.28577971e-01 3.53431702e-01 8.28095257e-01
5.73667765e-01 -1.50377655e+00 -4.98337746e-01 2.77939767e-01
3.41405690e-01 -5.04058540e-01 5.08613646e-01 8.65643382e-01
8.13708305e-02 1.01661138e-01 -1.63617358e-01 -1.84059709e-01
-1.41281295e+00 9.39122677e-01 2.01860722e-02 -4.88001585e-01
-5.62969685e-01 8.98034692e-01 3.03509414e-01 -6.60960734e-01
5.01158416e-01 -3.29586357e-01 -6.84486687e-01 1.34101138e-01
4.68420833e-01 4.02419925e-01 7.96518475e-02 -4.62868959e-01
-1.47075668e-01 3.18746343e-02 -6.51608706e-01 3.34718645e-01
1.40373600e+00 -2.82321364e-01 -2.92887427e-02 4.00138885e-01
8.88368785e-01 8.05197060e-02 -1.11431026e+00 -7.07868934e-01
-7.69088641e-02 -5.70151389e-01 4.09698300e-02 -9.00880992e-01
-1.10808992e+00 3.29259843e-01 5.50589263e-01 5.06975651e-01
1.16165030e+00 -1.42243132e-01 8.94520640e-01 2.43735194e-01
2.94322938e-01 -8.28275859e-01 2.59312242e-01 5.42651117e-01
8.55488956e-01 -1.15698814e+00 2.18846817e-02 -3.87510657e-01
-7.36052334e-01 7.98839271e-01 9.01619554e-01 -2.67132334e-02
8.08340013e-01 -2.24031694e-02 -8.38323534e-02 -1.85250729e-01
-1.00989926e+00 -8.14137459e-02 7.54131556e-01 2.51805007e-01
7.64488995e-01 7.03873858e-02 -5.04464567e-01 9.67527092e-01
7.60434940e-02 -1.88958898e-01 2.89830923e-01 8.42509508e-01
-1.12868026e-01 -1.35887980e+00 2.76577454e-02 7.99945056e-01
-3.34046930e-01 -4.59908009e-01 -1.36492610e-01 3.20354283e-01
8.96672383e-02 8.21259677e-01 9.31167305e-02 -8.14613402e-01
3.97770315e-01 -1.33782595e-01 3.65846783e-01 -6.41459823e-01
-6.49977624e-01 1.32856876e-01 1.69799075e-01 -4.83628273e-01
-4.13957238e-01 -6.61528051e-01 -6.73525691e-01 -2.92948216e-01
-3.94084603e-01 1.89662963e-01 2.26383433e-02 7.22424090e-01
5.76611221e-01 6.53106391e-01 7.05192626e-01 -7.36365199e-01
-7.41545081e-01 -9.78898585e-01 -6.63668692e-01 7.65310585e-01
2.74457574e-01 -7.75525630e-01 -2.41531849e-01 -1.56270117e-01] | [10.142488479614258, 5.595671653747559] |
45abafad-cbf3-4caa-b7a2-4c06e5d64264 | single-image-3d-face-reconstruction-under | 2205.04126 | null | https://arxiv.org/abs/2205.04126v2 | https://arxiv.org/pdf/2205.04126v2.pdf | Towards 3D Face Reconstruction in Perspective Projection: Estimating 6DoF Face Pose from Monocular Image | In 3D face reconstruction, orthogonal projection has been widely employed to substitute perspective projection to simplify the fitting process. This approximation performs well when the distance between camera and face is far enough. However, in some scenarios that the face is very close to camera or moving along the camera axis, the methods suffer from the inaccurate reconstruction and unstable temporal fitting due to the distortion under the perspective projection. In this paper, we aim to address the problem of single-image 3D face reconstruction under perspective projection. Specifically, a deep neural network, Perspective Network (PerspNet), is proposed to simultaneously reconstruct 3D face shape in canonical space and learn the correspondence between 2D pixels and 3D points, by which the 6DoF (6 Degrees of Freedom) face pose can be estimated to represent perspective projection. Besides, we contribute a large ARKitFace dataset to enable the training and evaluation of 3D face reconstruction solutions under the scenarios of perspective projection, which has 902,724 2D facial images with ground-truth 3D face mesh and annotated 6DoF pose parameters. Experimental results show that our approach outperforms current state-of-the-art methods by a significant margin. The code and data are available at https://github.com/cbsropenproject/6dof_face. | ['Zhen Lei', 'Xiaobo Li', 'Yuanzhang Chang', 'Xiangyu Zhu', 'Jiangjing Lyu', 'Miao Xu', 'Bowen Pan', 'Yueying Kao'] | 2022-05-09 | null | null | null | null | ['3d-face-reconstruction', 'face-reconstruction'] | ['computer-vision', 'computer-vision'] | [-2.50490934e-01 -8.97345096e-02 2.21036952e-02 -5.42487323e-01
-3.69972169e-01 -1.90554962e-01 2.97260940e-01 -9.98220682e-01
1.11069111e-02 1.32641211e-01 6.48645237e-02 1.64500520e-01
2.00743690e-01 -4.97536361e-01 -8.23499382e-01 -6.78296030e-01
3.14274609e-01 7.86837220e-01 -4.42617774e-01 6.28916547e-02
-9.98749211e-03 1.03913462e+00 -1.57851875e+00 -5.59009844e-03
1.69284239e-01 1.02347040e+00 -2.06738055e-01 2.22595841e-01
-2.79082377e-02 -8.66359696e-02 -2.11056963e-01 -7.90602386e-01
5.76116502e-01 -1.65322885e-01 -1.81203738e-01 3.88307065e-01
9.23785985e-01 -8.38677645e-01 -5.73011875e-01 9.42549169e-01
8.65673661e-01 -1.58694446e-01 4.38550979e-01 -1.06231618e+00
-4.66495246e-01 -1.94702193e-01 -1.03001821e+00 -2.93508589e-01
6.58945084e-01 1.70777477e-02 1.11318253e-01 -1.68773746e+00
8.02162945e-01 1.61291218e+00 8.73014033e-01 1.01286447e+00
-9.89106238e-01 -9.22365069e-01 1.58147328e-02 -1.09550446e-01
-1.69276631e+00 -1.05375934e+00 1.16610861e+00 -3.10402900e-01
6.19793296e-01 2.76177675e-02 9.48932946e-01 1.27620971e+00
6.94576884e-03 4.97565299e-01 9.23433959e-01 -2.20356137e-01
-2.62228519e-01 -1.28409475e-01 -3.35924238e-01 8.50722075e-01
-1.09835654e-01 3.52572769e-01 -5.60937166e-01 -2.80437827e-01
1.22418439e+00 3.20542336e-01 -4.51562107e-01 -4.73105371e-01
-6.39537156e-01 4.95570183e-01 2.34212905e-01 -7.30090737e-02
-3.27187479e-01 -1.56835049e-01 7.02536628e-02 -6.70607807e-03
8.17588389e-01 -3.26608598e-01 -4.34223980e-01 1.30452171e-01
-8.24922025e-01 3.31119299e-01 6.25087678e-01 1.07468319e+00
4.20834839e-01 2.90489793e-01 3.77663225e-01 9.41376328e-01
6.93509936e-01 7.67093003e-01 6.75725341e-02 -1.34661663e+00
3.21579367e-01 5.10182977e-01 -1.83664076e-02 -1.18605816e+00
-4.41775531e-01 -2.63728738e-01 -7.80142546e-01 2.19719470e-01
4.65468675e-01 1.74577553e-02 -7.63774455e-01 1.51376867e+00
8.64501178e-01 3.35607082e-01 -2.48123929e-01 1.18846118e+00
1.13033700e+00 4.99430597e-01 -7.17777967e-01 -5.45529604e-01
1.35468102e+00 -6.40545070e-01 -8.05990219e-01 -1.24219589e-01
1.10010728e-02 -9.86818433e-01 8.11958611e-01 3.85114968e-01
-1.40272677e+00 -4.49186772e-01 -6.27090693e-01 -1.43868223e-01
2.45912522e-01 1.93764642e-01 2.31258124e-01 4.19507205e-01
-1.04743886e+00 2.91683435e-01 -9.05359983e-01 -2.53281146e-01
6.47207260e-01 3.69809180e-01 -8.74782264e-01 -3.83836716e-01
-7.44773865e-01 6.58755779e-01 -2.64675081e-01 6.11619234e-01
-8.33675027e-01 -7.89553404e-01 -7.20497906e-01 -1.37924567e-01
3.44545543e-01 -7.98259735e-01 1.04764414e+00 -7.17566192e-01
-1.76803124e+00 1.37250960e+00 -3.60823303e-01 3.39428872e-01
7.50062287e-01 -1.77297547e-01 -2.98323423e-01 9.75604877e-02
-1.51656538e-01 6.05302215e-01 1.03533351e+00 -1.33439088e+00
2.37305865e-01 -1.21714604e+00 -2.06083819e-01 4.17094916e-01
-1.94307253e-01 2.32169554e-01 -1.08875406e+00 -3.31483096e-01
5.31298339e-01 -1.03056180e+00 1.62039503e-01 7.33857453e-01
-2.81116903e-01 -3.93467247e-02 1.02492166e+00 -8.41866851e-01
5.75701654e-01 -2.24696827e+00 1.28330797e-01 -1.68049008e-01
1.18750848e-01 3.11591506e-01 -5.38217789e-03 1.29219756e-01
-4.00245637e-01 -2.10039005e-01 -8.63283351e-02 -8.46894920e-01
-2.78115779e-01 -1.07837114e-02 -1.67405233e-02 9.44588482e-01
-2.46536985e-01 6.96693301e-01 -3.81169736e-01 -3.90352994e-01
1.04983047e-01 1.14983726e+00 -5.84828198e-01 3.48287404e-01
1.01760395e-01 7.12277234e-01 -2.33084485e-01 8.97767246e-01
1.27819216e+00 5.25823832e-02 6.54629171e-02 -2.99237043e-01
7.52231851e-02 -2.21648827e-01 -1.10996246e+00 1.91497421e+00
-3.07831436e-01 3.99697900e-01 4.69190806e-01 -5.48042834e-01
1.06611443e+00 5.82493544e-01 7.41273999e-01 -4.12945300e-01
3.85174662e-01 5.50464019e-02 -3.89267862e-01 -4.48740453e-01
-1.83683664e-01 -3.23954970e-01 5.27963877e-01 3.07726383e-01
-5.25842346e-02 -2.49873310e-01 -4.55844074e-01 -3.99720490e-01
2.99571395e-01 5.20254850e-01 7.75872990e-02 1.03788882e-01
5.46708107e-01 -6.06789768e-01 7.38381982e-01 -4.09820855e-01
-1.72384739e-01 9.61357236e-01 4.44659203e-01 -8.52297783e-01
-8.71412635e-01 -9.76741970e-01 -5.11561871e-01 3.08732539e-01
-5.52000962e-02 -2.43693918e-01 -9.67036903e-01 -5.54185271e-01
6.30765185e-02 2.17846468e-01 -4.78131384e-01 3.02215777e-02
-9.35270429e-01 -4.99976844e-01 4.68865961e-01 3.64704490e-01
5.28416038e-01 -7.52147019e-01 -2.29130745e-01 -2.31731549e-01
-7.97949582e-02 -1.17181146e+00 -8.32468688e-01 -5.96476853e-01
-1.12657285e+00 -1.21242464e+00 -8.28397095e-01 -5.44939816e-01
1.13948655e+00 4.43032444e-01 8.54969203e-01 1.03324071e-01
-9.79463905e-02 2.95892775e-01 2.55604923e-01 -2.02434912e-01
6.65770518e-03 -5.77682257e-01 4.39823389e-01 1.85894027e-01
5.07577121e-01 -7.20843434e-01 -7.36538708e-01 6.71288610e-01
-4.70413119e-01 1.00995257e-01 6.03752844e-02 7.66020417e-01
8.10576022e-01 -1.46760002e-01 -4.19492535e-02 -5.52676082e-01
-7.03237776e-04 -1.91600844e-01 -9.06169653e-01 -8.77359509e-02
-3.53610992e-01 -5.35258412e-01 4.92305815e-01 -4.57626641e-01
-1.12375963e+00 3.34695518e-01 -4.68813092e-01 -1.31442845e+00
-2.26573765e-01 5.39962724e-02 -6.73761904e-01 -2.06859723e-01
3.75558555e-01 2.02469394e-01 5.51807821e-01 -6.97654963e-01
9.68137532e-02 5.15000522e-01 3.42724144e-01 -3.20021898e-01
7.47003257e-01 8.38381410e-01 2.08370239e-01 -8.95888925e-01
-7.36434460e-01 -3.87767404e-02 -9.05647635e-01 -5.44642925e-01
5.43781579e-01 -1.06341147e+00 -1.04593813e+00 7.10527360e-01
-1.36245894e+00 4.80594411e-02 1.11281939e-01 4.87230599e-01
-5.78909039e-01 4.24833894e-01 -5.22745550e-01 -7.81206191e-01
-4.83931243e-01 -1.37787259e+00 1.32305062e+00 1.93582952e-01
1.57851562e-01 -6.78281903e-01 -2.08919510e-01 6.01087391e-01
-1.73305664e-02 3.04993123e-01 4.28783596e-01 -6.60097413e-03
-4.52529937e-01 -3.41897547e-01 2.54704151e-02 2.53095388e-01
4.53188177e-03 2.42841870e-01 -1.18796110e+00 -4.08258557e-01
5.42488635e-01 -6.87491670e-02 2.82516956e-01 6.87106669e-01
1.17898560e+00 -3.20931196e-01 -2.17389643e-01 1.33249092e+00
1.11646616e+00 1.70890287e-01 5.67490876e-01 -2.42259532e-01
8.33516836e-01 8.43848884e-01 4.82666761e-01 5.16484022e-01
3.59558582e-01 1.08701432e+00 6.22224033e-01 1.97134793e-01
-2.42654502e-01 -5.27673066e-01 2.93742001e-01 8.99041235e-01
-3.00073504e-01 4.98010553e-02 -7.72961438e-01 2.88005248e-02
-1.32478702e+00 -7.15038776e-01 -8.89099613e-02 2.33449364e+00
4.66938227e-01 -2.49645188e-01 4.55328338e-02 -1.39794707e-01
7.07419276e-01 6.65498078e-02 -8.17990780e-01 -4.00097594e-02
9.30816606e-02 -2.51270890e-01 7.88700357e-02 5.45017958e-01
-6.59259200e-01 9.44422722e-01 5.70221901e+00 5.80092132e-01
-1.40702391e+00 1.67012528e-01 6.15140200e-01 -4.05567944e-01
-1.46888122e-01 -1.90503970e-01 -1.15697050e+00 4.53182817e-01
5.71313798e-01 2.78648645e-01 6.83008492e-01 7.80768633e-01
2.72070318e-01 3.49515319e-01 -9.72227216e-01 1.50328434e+00
5.40515661e-01 -1.04334927e+00 -4.39830609e-02 4.01941866e-01
6.46095395e-01 -1.67831779e-01 2.63591677e-01 -1.90442860e-01
-3.64410281e-01 -1.05056238e+00 5.16040802e-01 7.78916478e-01
1.33624184e+00 -7.32431829e-01 3.47508550e-01 5.90904832e-01
-9.28123891e-01 1.86332762e-01 -4.98536974e-01 1.36920065e-01
2.87889451e-01 3.77615154e-01 -6.72599912e-01 3.63159418e-01
8.24681640e-01 8.58730495e-01 -9.17258263e-02 5.46958268e-01
-1.23240516e-01 9.34279338e-02 -4.76337641e-01 6.35602176e-01
-3.46430898e-01 -6.21250689e-01 7.01858938e-01 3.09439301e-01
7.02064574e-01 3.32828760e-01 -1.51683658e-01 9.05481279e-01
-2.46864125e-01 -5.52556254e-02 -1.00608277e+00 1.77095339e-01
4.54711616e-01 1.23518801e+00 -5.28953552e-01 2.12954402e-01
-4.15979505e-01 8.92121017e-01 2.28112429e-01 4.06307757e-01
-8.01060081e-01 3.78294706e-01 8.25112760e-01 5.51875234e-01
-9.26350616e-03 -2.86625803e-01 -3.00337858e-02 -1.24642217e+00
2.73074239e-01 -6.80950880e-01 3.47201712e-02 -9.89555359e-01
-1.04137206e+00 7.73984432e-01 2.36341842e-02 -1.30108297e+00
-7.78152049e-02 -7.41714299e-01 -4.61317748e-01 8.35015416e-01
-1.20250881e+00 -1.31593490e+00 -4.20449018e-01 6.86010540e-01
6.03640676e-01 -2.97779948e-01 8.11094522e-01 4.33261573e-01
-7.37014353e-01 8.26678216e-01 -1.54749140e-01 3.71310189e-02
8.01902771e-01 -5.21621108e-01 3.65187138e-01 4.01409984e-01
6.28441274e-02 7.59436309e-01 4.75564927e-01 -5.05140722e-01
-1.98603272e+00 -8.23185503e-01 6.51833653e-01 -6.51797533e-01
-2.57428773e-02 -5.14182746e-01 -7.62937844e-01 8.38450015e-01
-2.56623983e-01 6.70930743e-01 5.33217490e-01 -2.78817892e-01
-2.91405708e-01 -3.29161018e-01 -1.45603609e+00 5.28479397e-01
1.43254936e+00 -4.58762884e-01 -1.24540605e-01 4.22076494e-01
3.59039545e-01 -1.03310716e+00 -9.60573554e-01 4.25202310e-01
9.29088950e-01 -1.16928828e+00 1.16165674e+00 -2.09130958e-01
2.35067859e-01 -2.24444851e-01 -2.04729483e-01 -1.05705082e+00
1.44875064e-01 -6.28004372e-01 -4.01214629e-01 1.04122257e+00
-1.76428840e-01 -6.49342835e-01 1.03672075e+00 5.22528887e-01
-1.02160290e-01 -1.14024639e+00 -1.32514668e+00 -4.73638713e-01
-8.33970979e-02 -2.61310637e-01 8.11656117e-01 8.31794560e-01
-5.84236979e-01 2.16355577e-01 -5.10156333e-01 3.38351995e-01
6.17298543e-01 1.36969909e-01 9.51291203e-01 -1.24273908e+00
1.41680554e-01 -1.69724837e-01 -2.59848207e-01 -1.20184839e+00
6.51294589e-01 -7.82519698e-01 -3.79374474e-01 -1.03901780e+00
8.20062757e-02 -1.99221089e-01 5.00747323e-01 1.15339458e-01
2.99276203e-01 4.70009506e-01 6.78075571e-03 3.12489957e-01
1.59217417e-01 8.18968236e-01 1.79033899e+00 2.16408789e-01
-4.85072285e-02 3.15493912e-01 -4.54283059e-01 1.06138527e+00
3.63175124e-01 -2.87907958e-01 -3.33029032e-01 -7.66872108e-01
-1.81818157e-02 5.65702319e-01 3.56930673e-01 -5.01590431e-01
2.05271631e-01 4.87847300e-03 8.15546036e-01 -8.32613230e-01
1.17590857e+00 -1.22828829e+00 6.97661757e-01 2.13915512e-01
1.72324941e-01 2.83688098e-01 1.29786298e-01 3.11703026e-01
-3.15967537e-02 -7.06101209e-02 9.70029473e-01 -2.23209485e-01
-1.29839694e-02 1.12540126e+00 3.59949052e-01 -3.09129059e-01
9.82397020e-01 -5.34747958e-01 3.50188278e-02 -3.72696102e-01
-6.95954680e-01 -8.12117830e-02 8.75579715e-01 2.69842148e-01
1.15066934e+00 -1.62530398e+00 -7.92521417e-01 8.90741527e-01
-1.43888026e-01 2.98287541e-01 5.06631136e-01 9.18550849e-01
-6.87249959e-01 2.12289721e-01 -1.31214753e-01 -8.30662549e-01
-1.56029248e+00 5.82838416e-01 7.08066463e-01 4.61777300e-01
-8.81565213e-01 9.14766192e-01 2.73487180e-01 -9.60248172e-01
2.32141361e-01 2.78779507e-01 -1.00787945e-01 -1.89175922e-02
5.31420350e-01 3.18270117e-01 3.39770243e-02 -1.26222563e+00
-2.88377732e-01 1.29308832e+00 7.97472745e-02 1.17159404e-01
1.38696456e+00 -1.01448298e-01 -3.00110251e-01 1.32473454e-01
1.40298498e+00 5.80749027e-02 -1.55602407e+00 -1.03451833e-01
-7.46838272e-01 -1.07784748e+00 -3.95612279e-03 -2.47892022e-01
-1.64171207e+00 1.21714997e+00 7.51299798e-01 -5.57040095e-01
1.10816014e+00 -1.21949345e-01 7.93294013e-01 2.59403791e-02
4.37442780e-01 -5.32208502e-01 -4.63608392e-02 3.90036076e-01
1.38460135e+00 -1.11505723e+00 1.16419084e-01 -6.29024684e-01
-3.68889064e-01 1.39036870e+00 1.00269580e+00 -9.14440230e-02
9.39693153e-01 2.31528088e-01 1.47177905e-01 -4.55487549e-01
-4.53898638e-01 5.43687940e-01 2.68639088e-01 5.30480325e-01
3.51455241e-01 -1.56928927e-01 1.99244723e-01 2.98316061e-01
-2.96404868e-01 -6.20162860e-02 1.84050977e-01 5.39720714e-01
2.00619638e-01 -8.79318714e-01 -6.46752238e-01 9.58981961e-02
-4.66487437e-01 2.93256760e-01 -2.46687964e-01 7.89354861e-01
7.93447644e-02 3.35434973e-01 1.37766704e-01 -1.25311196e-01
4.80960101e-01 6.92043379e-02 8.25433016e-01 -4.40420479e-01
3.36569212e-02 5.03258228e-01 -2.95623451e-01 -7.08769262e-01
-2.48187497e-01 -7.84226418e-01 -1.00417733e+00 -5.94779074e-01
-3.12515587e-01 -2.93155760e-01 7.75601864e-01 5.64697623e-01
5.30269325e-01 -9.68528986e-02 9.04864371e-01 -1.27672172e+00
-4.58915800e-01 -8.74897420e-01 -6.85567975e-01 7.62388706e-02
3.37904125e-01 -9.21386838e-01 -3.93069506e-01 -4.82470654e-02] | [13.204896926879883, 0.08960448205471039] |
9877a4ff-96d0-4c3e-9920-92710790b751 | reliable-image-dehazing-by-nerf | 2303.09153 | null | https://arxiv.org/abs/2303.09153v1 | https://arxiv.org/pdf/2303.09153v1.pdf | Reliable Image Dehazing by NeRF | We present an image dehazing algorithm with high quality, wide application, and no data training or prior needed. We analyze the defects of the original dehazing model, and propose a new and reliable dehazing reconstruction and dehazing model based on the combination of optical scattering model and computer graphics lighting rendering model. Based on the new haze model and the images obtained by the cameras, we can reconstruct the three-dimensional space, accurately calculate the objects and haze in the space, and use the transparency relationship of haze to perform accurate haze removal. To obtain a 3D simulation dataset we used the Unreal 5 computer graphics rendering engine. In order to obtain real shot data in different scenes, we used fog generators, array cameras, mobile phones, underwater cameras and drones to obtain haze data. We use formula derivation, simulation data set and real shot data set result experimental results to prove the feasibility of the new method. Compared with various other methods, we are far ahead in terms of calculation indicators (4 dB higher quality average scene), color remains more natural, and the algorithm is more robust in different scenarios and best in the subjective perception. | ['Yueting Chen', 'Qi Li', 'Zhihai Xu', 'Huajun Feng', 'Shiqi Chen', 'Zheyan Jin'] | 2023-03-16 | null | null | null | null | ['image-dehazing'] | ['computer-vision'] | [-2.20092535e-02 -6.62215352e-01 1.04270768e+00 -7.44071007e-02
1.32701710e-01 -1.80973798e-01 2.27887258e-01 -5.24624109e-01
-4.26441669e-01 5.82665861e-01 8.08568075e-02 -4.11985852e-02
1.96640231e-02 -1.09273458e+00 -3.52078587e-01 -1.08844888e+00
-5.40804677e-02 -4.93955091e-02 7.97437847e-01 -5.25804400e-01
3.00548255e-01 2.91583538e-01 -1.88305879e+00 5.32068163e-02
1.13036573e+00 8.38704944e-01 4.63487983e-01 7.92251468e-01
3.68995517e-02 7.09472001e-01 -9.24637794e-01 -1.14722066e-01
5.99333704e-01 -5.94125211e-01 1.77503437e-01 3.59098673e-01
2.04409927e-01 -9.13657129e-01 -3.91669989e-01 1.34470057e+00
5.03932655e-01 2.52210110e-01 7.25743413e-01 -1.01480901e+00
-7.99814880e-01 -5.22158742e-01 -5.93912363e-01 2.91579932e-01
1.98104069e-01 1.87690988e-01 -1.10121310e-01 -8.54879916e-01
1.10016242e-01 1.16922486e+00 5.47496915e-01 5.27402043e-01
-2.51634717e-01 -7.84627557e-01 -1.97855532e-01 3.51887345e-01
-1.55727804e+00 -2.39677250e-01 6.14833832e-01 -2.69077361e-01
4.23796088e-01 4.14364129e-01 1.03976488e+00 2.98236430e-01
5.51443696e-01 1.76695809e-01 1.43446529e+00 -5.25262237e-01
1.95563599e-01 3.27981830e-01 1.52526617e-01 1.00613296e+00
6.38208687e-01 2.98765093e-01 -2.63400435e-01 7.35655203e-02
7.85478652e-01 3.55089277e-01 -7.03742325e-01 1.50918096e-01
-6.49755478e-01 4.78178680e-01 1.70764461e-01 1.69784669e-02
-1.44401789e-01 -4.19191197e-02 -4.35919613e-01 3.77381265e-01
5.81047177e-01 -8.30313042e-02 2.28836104e-01 2.46356368e-01
-6.65772021e-01 1.77660808e-01 6.32117569e-01 9.62814629e-01
8.53210986e-01 5.81969082e-01 3.06797475e-01 5.94425857e-01
7.17412233e-01 1.09412003e+00 3.90697241e-01 -9.57397401e-01
1.17972612e-01 2.57627189e-01 3.20424408e-01 -1.07118285e+00
1.63902253e-01 1.12146951e-01 -8.95354748e-01 1.00941455e+00
-1.44848794e-01 -1.47653475e-01 -1.31230450e+00 8.11295390e-01
2.95973182e-01 5.82835495e-01 3.98510754e-01 1.20364678e+00
9.38830912e-01 1.49073040e+00 -6.59587383e-01 -5.13778269e-01
1.08764720e+00 -6.86497927e-01 -1.16707897e+00 2.12472647e-01
1.82901248e-01 -8.53180885e-01 9.32636142e-01 1.11895239e+00
-1.03049529e+00 -4.87112761e-01 -1.55979514e+00 5.45250401e-02
-4.21299726e-01 -3.02312106e-01 3.00555587e-01 1.00514495e+00
-1.04049528e+00 2.17675135e-01 -5.18447936e-01 -5.31526133e-02
2.64785588e-02 -5.81337400e-02 -7.46776769e-03 -6.03787959e-01
-1.11970627e+00 1.00117791e+00 2.34828457e-01 3.84166211e-01
-1.19709587e+00 -5.90561986e-01 -6.14437580e-01 -1.50195226e-01
1.01045839e-01 -5.90315044e-01 7.28417814e-01 -6.90981448e-01
-1.61907959e+00 3.92872125e-01 1.33100837e-01 -1.23395488e-01
3.03836882e-01 -3.10209960e-01 -8.83831322e-01 2.91893154e-01
-4.04879808e-01 -1.33956984e-01 9.20261323e-01 -1.81752229e+00
-6.72958434e-01 -2.78295964e-01 6.40994590e-03 3.94352317e-01
-7.62271695e-03 7.37703145e-02 -3.80940229e-01 -2.55981773e-01
3.38293947e-02 -4.69260424e-01 -1.48222983e-01 3.22085500e-01
8.43406692e-02 6.17502630e-01 1.12984395e+00 -9.23991144e-01
1.09455991e+00 -2.54272866e+00 -1.08414844e-01 2.71861166e-01
3.68379444e-01 5.70421398e-01 2.85126835e-01 2.97086596e-01
3.42691094e-01 -9.14055109e-02 -4.08418328e-01 -1.69034272e-01
-3.64355147e-01 3.81397605e-01 -1.24518864e-01 6.50003016e-01
-4.69668418e-01 1.17982673e-02 -6.23600364e-01 -4.93754715e-01
5.22921860e-01 5.83751023e-01 -4.30882514e-01 5.83389044e-01
9.22631621e-02 1.29458725e-01 -1.52292430e-01 5.85648298e-01
1.38669574e+00 5.63987613e-01 -7.22655118e-01 -1.83997631e-01
-4.37498629e-01 -6.69010103e-01 -1.51278698e+00 1.04727757e+00
-4.84284610e-01 6.25487268e-01 4.47108626e-01 -2.79085159e-01
1.10377622e+00 2.75971383e-01 -1.18393846e-01 -6.99790001e-01
1.53823242e-01 1.32169172e-01 -3.26612890e-01 -1.05761743e+00
5.32197118e-01 -5.35326004e-01 6.42286539e-01 9.47460532e-02
-3.17799538e-01 -7.07782507e-01 -2.01250017e-01 1.29877806e-01
6.09516263e-01 -1.99608505e-01 -5.84627427e-02 -5.19694805e-01
3.55319947e-01 -6.12592697e-02 3.66233021e-01 2.98448414e-01
1.40511423e-01 9.13746595e-01 -1.96780503e-01 -5.16317606e-01
-9.60503042e-01 -1.08766997e+00 1.52101349e-02 3.16281676e-01
1.00681460e+00 -1.06917001e-01 -8.07214618e-01 3.22735041e-01
-5.37453711e-01 8.74214649e-01 -3.13120365e-01 -3.40910435e-01
-3.47252041e-01 -1.03798485e+00 6.49145767e-02 -4.53428894e-01
1.23353648e+00 -4.86224532e-01 -4.27982986e-01 -3.99262644e-02
1.00917690e-01 -8.98152053e-01 -1.11697301e-01 -5.65338075e-01
-7.02361166e-01 -1.15048695e+00 -6.79434359e-01 -7.68360496e-01
5.35607457e-01 1.10624182e+00 7.56492019e-01 6.39114559e-01
-3.76542807e-01 5.64484417e-01 -5.87670386e-01 -8.72876883e-01
-4.61727917e-01 -1.14553630e+00 3.94459963e-02 2.29458615e-01
1.23318039e-01 -6.88326418e-01 -8.22535753e-01 3.87257129e-01
-1.34083295e+00 1.69721350e-01 2.03518242e-01 2.36281291e-01
3.13225724e-02 8.59090209e-01 -2.83083200e-01 -6.02306306e-01
4.43030000e-01 -4.61910844e-01 -1.08750892e+00 -2.24806946e-02
-6.61311209e-01 -5.43853223e-01 4.22537833e-01 -4.98443544e-02
-1.47746468e+00 -6.10271096e-01 7.78749064e-02 -6.24134958e-01
-1.07293285e-01 9.93383974e-02 -3.92975628e-01 -3.69132757e-01
6.23588085e-01 5.61886489e-01 1.09742180e-01 -5.65814853e-01
-5.47689721e-02 8.53058875e-01 5.90258777e-01 -6.03134558e-02
1.20070171e+00 9.03553367e-01 -4.83447164e-02 -1.36415923e+00
-2.28296757e-01 -2.94976532e-01 -2.39240304e-01 -6.44697845e-01
1.35336924e+00 -1.10372901e+00 -5.78520954e-01 9.44259346e-01
-1.34993076e+00 -1.30038172e-01 2.62993574e-01 8.23595524e-01
6.04399107e-02 6.64650857e-01 -6.50385976e-01 -1.19416189e+00
-2.44846553e-01 -9.79123294e-01 7.30678618e-01 2.65691906e-01
9.79338646e-01 -9.22390282e-01 4.30626348e-02 1.60481274e-01
4.90965843e-01 4.17104334e-01 7.20668733e-01 5.55710793e-01
-1.19608152e+00 -8.25551376e-02 -1.85032234e-01 7.56647527e-01
2.29340225e-01 5.10367692e-01 -9.80207086e-01 -4.07513604e-02
7.23861396e-01 4.48731452e-01 9.57536936e-01 1.83897898e-01
9.20835316e-01 -1.99258342e-01 1.31301165e-01 1.16173792e+00
2.03945804e+00 6.32514417e-01 1.32160366e+00 2.92533040e-01
6.81105196e-01 4.56106514e-01 7.37450242e-01 4.45471555e-01
8.33141282e-02 3.86226326e-01 9.18005168e-01 -3.30462873e-01
-2.39262968e-01 2.43369997e-01 4.90906894e-01 1.21719480e+00
-7.47710943e-01 -7.55321205e-01 -4.55384135e-01 2.57689267e-01
-1.27878392e+00 -8.64232421e-01 -6.01574421e-01 2.28326416e+00
3.14102948e-01 -3.59555446e-02 -2.56637454e-01 1.65800527e-01
7.20961213e-01 -8.55528414e-02 1.45851955e-01 -2.99366355e-01
-1.30097777e-01 -3.21592614e-02 7.11500704e-01 9.15983677e-01
-4.99368787e-01 6.92532063e-01 6.58289242e+00 6.69050872e-01
-8.43359590e-01 2.54929394e-01 -2.62761046e-03 1.43150985e-02
-5.37305415e-01 -1.52277693e-01 -6.05298162e-01 8.87433290e-01
5.78216910e-01 -1.78587571e-01 7.40980566e-01 2.52748460e-01
4.03351873e-01 -4.36557591e-01 -4.09975290e-01 1.34414816e+00
4.56870168e-01 -1.02133250e+00 3.24779451e-01 5.35190776e-02
7.30340779e-01 -2.55420178e-01 1.55906277e-02 -1.64764374e-01
3.24088812e-01 -8.07805717e-01 5.94630003e-01 9.05964196e-01
5.61993182e-01 -5.36991775e-01 1.07391953e+00 4.43902135e-01
-7.75358796e-01 2.39375412e-01 -1.03882229e+00 -3.42075616e-01
2.08635285e-01 7.58436263e-01 -3.56444925e-01 8.00423026e-01
8.80844355e-01 4.08330351e-01 -3.39207441e-01 1.33183146e+00
-1.81983918e-01 5.72476923e-01 -3.88664663e-01 1.85905341e-02
1.95685521e-01 -8.94463062e-01 4.34914023e-01 9.87437189e-01
7.93843150e-01 8.40870082e-01 -2.68340975e-01 6.70681238e-01
5.46821773e-01 -1.94861993e-01 -6.64393663e-01 5.46239793e-01
1.08249851e-01 8.98019671e-01 -3.23630750e-01 -4.24653530e-01
-5.25359631e-01 9.49601710e-01 -6.18701041e-01 6.53246880e-01
-9.59578395e-01 -7.28376925e-01 4.79440153e-01 1.18741997e-01
-7.65414312e-02 -5.85264385e-01 9.03573185e-02 -1.14604723e+00
-1.55549482e-01 -6.13536775e-01 -6.35673404e-02 -1.53097057e+00
-1.11944735e+00 4.81963962e-01 2.36776620e-01 -1.51005685e+00
4.58589524e-01 -9.41386521e-01 -8.67047310e-01 1.00096524e+00
-1.74116969e+00 -7.14064300e-01 -1.14226794e+00 9.90037084e-01
4.96018797e-01 -9.23984721e-02 6.55543208e-01 5.23643017e-01
-3.63639683e-01 -3.67586762e-02 6.05354249e-01 -3.27724278e-01
4.96303618e-01 -9.06834602e-01 -1.66073442e-01 1.22133124e+00
-3.90395671e-01 1.31979600e-01 1.02485406e+00 -5.70308685e-01
-1.36859322e+00 -8.64149272e-01 -1.59902066e-01 -3.61683756e-01
2.14616567e-01 -4.28820252e-01 -1.07829416e+00 3.05609167e-01
4.84315634e-01 -1.69685960e-01 4.60441738e-01 -7.57211328e-01
1.05874382e-01 -3.41804057e-01 -1.33203793e+00 6.23781681e-01
7.78997481e-01 1.47432284e-02 -7.33022153e-01 3.83229256e-01
1.08190024e+00 -4.60295856e-01 -4.56567824e-01 1.83520436e-01
3.26721251e-01 -1.48229206e+00 7.88843215e-01 1.01096302e-01
2.10064948e-01 -8.07322085e-01 -3.61968011e-01 -1.29179525e+00
-3.39857548e-01 -6.10268891e-01 3.59669626e-01 9.79370475e-01
6.26837537e-02 -8.60085368e-01 4.48939621e-01 3.64505053e-01
-5.61803281e-01 -9.68049094e-02 -5.76995730e-01 -8.01093042e-01
-4.03863579e-01 -1.56712756e-01 8.49541545e-01 7.01951325e-01
-5.36153555e-01 -4.03693058e-02 -1.00305212e+00 9.66989517e-01
1.01414716e+00 -6.17048480e-02 8.88279259e-01 -1.02241838e+00
-2.69722879e-01 2.73864448e-01 -6.24204814e-01 -8.03183377e-01
-3.90888453e-01 -9.88001376e-03 1.66514665e-01 -1.89446342e+00
-1.34245127e-01 -1.46858186e-01 -3.15158330e-02 -2.47248545e-01
-2.67232835e-01 4.40396458e-01 1.26487657e-01 3.81913334e-01
-1.89782828e-01 6.50771022e-01 1.46444893e+00 5.83717076e-04
-1.04304656e-01 7.20616952e-02 -2.68333256e-01 8.48313451e-01
6.01967990e-01 -3.49516511e-01 -5.67360997e-01 -9.67587769e-01
1.06073646e-02 -6.67935088e-02 4.28502977e-01 -1.24535203e+00
3.16711366e-01 -2.43929416e-01 3.97574067e-01 -4.35669005e-01
7.90775597e-01 -1.29003751e+00 2.74185061e-01 4.63777244e-01
5.44291139e-01 -1.08311906e-01 8.91430229e-02 6.40503705e-01
-2.64990538e-01 -5.75222850e-01 9.12549138e-01 -4.34558094e-01
-9.76970732e-01 4.21687961e-01 -6.10892057e-01 -2.39112049e-01
9.79002655e-01 -7.31452644e-01 -5.93270004e-01 -6.51232898e-01
-4.10546094e-01 -1.24222837e-01 7.96123028e-01 -2.14548200e-01
1.07933462e+00 -9.82524872e-01 -7.42447197e-01 5.54771662e-01
-1.14152864e-01 2.71562189e-01 3.24025184e-01 2.49712899e-01
-1.69558525e+00 -3.93459857e-01 -2.82100379e-01 -2.72599071e-01
-1.35422671e+00 7.05334783e-01 4.31362301e-01 7.37218976e-01
-7.83163249e-01 7.00373590e-01 6.13793433e-01 1.33395731e-01
-9.32139456e-02 -3.32517207e-01 -2.02449232e-01 -6.93152726e-01
1.00107217e+00 7.47267783e-01 -3.41244340e-02 -2.98604935e-01
1.11012414e-01 9.53713715e-01 3.74135882e-01 -1.50100619e-01
1.30718517e+00 -4.24579173e-01 -4.67706859e-01 3.58065784e-01
8.97580624e-01 4.29160923e-01 -1.15375423e+00 1.12350903e-01
-8.72171044e-01 -1.21907747e+00 2.27844074e-01 -4.01629031e-01
-1.04428804e+00 1.05339491e+00 1.00330627e+00 5.40576935e-01
1.47707474e+00 -4.62800026e-01 7.39400923e-01 3.13258469e-01
5.39063036e-01 -6.85012996e-01 -3.40194777e-02 1.61763132e-01
6.12882078e-01 -8.62252831e-01 4.02762830e-01 -9.10229027e-01
-4.20108199e-01 9.95687306e-01 7.12682009e-01 -3.75720203e-01
8.42389703e-01 3.42154771e-01 3.51404250e-01 -3.23385388e-01
-4.48530942e-01 -1.41585454e-01 -2.02865914e-01 9.23897684e-01
-3.31640124e-01 -3.91279459e-01 7.39501566e-02 1.63135946e-01
-1.39643222e-01 -1.05748244e-01 1.21126795e+00 7.16958582e-01
-1.26323211e+00 -4.82313842e-01 -8.57046902e-01 1.01827495e-01
-3.07807058e-01 -1.86285257e-01 1.16081536e-01 8.82795990e-01
4.57509726e-01 1.20369565e+00 7.98635650e-03 -4.24977690e-01
4.02718216e-01 -4.03555334e-01 5.85140347e-01 -5.07070601e-01
1.89317256e-01 2.81107575e-02 -1.49944767e-01 -1.68670595e-01
-5.31041682e-01 1.59067102e-02 -1.12713885e+00 -6.27497256e-01
-5.50026953e-01 4.56771821e-01 7.23030925e-01 6.11943245e-01
-1.08827502e-01 5.14744163e-01 8.34831595e-01 -8.06125820e-01
1.67111263e-01 -8.38357449e-01 -1.37656140e+00 2.33470157e-01
4.24837768e-01 -6.82786107e-01 -9.63696718e-01 2.40795538e-01] | [10.836413383483887, -3.161076307296753] |
c8064d9f-c190-4c6e-a13a-af6a67cae562 | joint-falsification-and-fidelity-settings | 2305.06111 | null | https://arxiv.org/abs/2305.06111v1 | https://arxiv.org/pdf/2305.06111v1.pdf | Joint Falsification and Fidelity Settings Optimization for Validation of Safety-Critical Systems: A Theoretical Analysis | Safety validation is a crucial component in the development and deployment of autonomous systems, such as self-driving vehicles and robotic systems. Ensuring safe operation necessitates extensive testing and verification of control policies, typically conducted in simulation environments. High-fidelity simulators accurately model real-world dynamics but entail high computational costs, limiting their scalability for exhaustive testing. Conversely, low-fidelity simulators offer efficiency but may not capture the intricacies of high-fidelity simulators, potentially yielding false conclusions. We propose a joint falsification and fidelity optimization framework for safety validation of autonomous systems. Our mathematical formulation combines counterexample searches with simulator fidelity improvement, facilitating more efficient exploration of the critical environmental configurations challenging the control system. Our contributions encompass a set of theorems addressing counterexample sensitivity analysis, sample complexity, convergence, the interplay between the outer and inner optimization loops, and regret bound analysis. The proposed joint optimization approach enables a more targeted and efficient testing process, optimizes the use of available computational resources, and enhances confidence in autonomous system safety validation. | ['Mykel J. Kochenderfer', 'Ali Baheri'] | 2023-05-10 | null | null | null | null | ['efficient-exploration'] | ['methodology'] | [ 2.29986638e-01 2.38508359e-01 -3.43313456e-01 -2.28277221e-02
-5.95253527e-01 -7.18847871e-01 4.50436711e-01 1.28788829e-01
-3.07379603e-01 9.94779885e-01 -6.67252779e-01 -9.71983194e-01
-4.27336544e-01 -7.20923960e-01 -9.68774319e-01 -5.03634453e-01
-5.20297348e-01 2.05923036e-01 1.99630186e-01 -4.50190865e-02
-1.17583089e-01 5.46064615e-01 -1.74946129e+00 -6.99050009e-01
1.00184929e+00 1.03389859e+00 -2.85473228e-01 5.80381811e-01
8.48513842e-01 3.03959250e-01 -3.68022859e-01 -1.19452104e-01
3.08305830e-01 -3.27064842e-01 -2.87648410e-01 -6.70537278e-02
-2.79911309e-01 -5.67105412e-01 -1.64611816e-01 1.21943343e+00
2.12715045e-01 -2.15987004e-02 2.18558505e-01 -2.15180659e+00
4.89328653e-01 3.89518768e-01 -1.51302189e-01 -2.90571064e-01
1.36036023e-01 6.99918807e-01 5.92077255e-01 -1.73859894e-01
1.79158643e-01 8.82863462e-01 5.54408073e-01 2.91328520e-01
-1.14471030e+00 -9.49814200e-01 1.30535543e-01 -1.61778912e-01
-1.58548295e+00 -4.65252280e-01 3.32379818e-01 -5.06929457e-01
9.43190575e-01 3.05208117e-01 7.48610556e-01 9.69494820e-01
6.65890872e-01 3.02395195e-01 9.52346444e-01 3.04105058e-02
5.88591039e-01 2.86001295e-01 -3.22808564e-01 5.98186553e-01
1.02920151e+00 9.71722364e-01 7.61540011e-02 -3.51974219e-01
3.67973387e-01 -2.04704836e-01 -9.79162827e-02 -7.18538880e-01
-9.47816193e-01 6.36259139e-01 -1.21091492e-01 -4.31623966e-01
-2.15890408e-01 4.63389128e-01 5.86777031e-01 3.59569967e-01
-2.60243982e-01 6.20162249e-01 -4.12358671e-01 -4.06485140e-01
-6.13163054e-01 5.44073224e-01 8.10595810e-01 1.10251975e+00
3.89079839e-01 3.47676814e-01 5.78929186e-02 -1.37161553e-01
3.55073422e-01 6.87942863e-01 -6.35295287e-02 -1.10233176e+00
3.51369530e-01 5.81438363e-01 6.55089617e-01 -6.89886332e-01
-4.52264309e-01 -6.68801546e-01 -5.24752319e-01 5.32618642e-01
2.95865275e-02 -4.61829454e-01 -4.22362715e-01 1.81980002e+00
5.80439925e-01 1.72636405e-01 1.70052126e-01 7.95099020e-01
-4.73696202e-01 3.62181216e-01 6.99561760e-02 -5.63696742e-01
9.78709936e-01 -2.40312412e-01 -5.25528371e-01 -3.61336410e-01
8.06353867e-01 -1.69562429e-01 9.12477553e-01 3.13353658e-01
-1.08969069e+00 -1.31810024e-01 -1.68296719e+00 8.21803451e-01
-1.12996101e-02 -5.28330579e-02 5.38035154e-01 9.09503162e-01
-6.02350593e-01 5.22360265e-01 -8.62906516e-01 5.68634868e-02
3.55347127e-01 7.56147325e-01 -1.18183762e-01 1.65693998e-01
-1.19979596e+00 1.18542814e+00 3.30241770e-01 2.49540105e-01
-1.44471717e+00 -8.45667422e-01 -1.03219843e+00 -2.21717954e-02
8.01180482e-01 -3.94590944e-01 1.41407633e+00 -3.36224735e-01
-1.52036309e+00 9.52561423e-02 2.83506155e-01 -8.78865719e-01
7.53195822e-01 1.15558850e-02 -4.16344017e-01 -2.16047063e-01
-4.21703830e-02 1.86954483e-01 7.02866137e-01 -9.48932528e-01
-5.93459070e-01 -1.40613215e-02 2.05635056e-01 -1.86816409e-01
1.11456424e-01 -4.40622598e-01 1.08438089e-01 -5.32257520e-02
-4.86651421e-01 -1.25330067e+00 -4.46517467e-01 -2.13717178e-01
-1.03750020e-01 4.58375573e-01 6.62608564e-01 -2.04959616e-01
1.19354737e+00 -2.12755895e+00 -1.70895889e-01 6.11636162e-01
-1.61187172e-01 1.50899947e-01 1.15718730e-01 5.41955352e-01
1.40035406e-01 9.52185169e-02 -9.27231982e-02 2.88359344e-01
3.53569597e-01 1.68716982e-01 -3.55245292e-01 8.81806433e-01
6.11105502e-01 5.90810955e-01 -7.38964677e-01 -3.19060028e-01
5.59946537e-01 -5.67614809e-02 -5.77916145e-01 6.69943616e-02
-2.97107309e-01 3.63706052e-01 -6.18763804e-01 2.86010385e-01
3.25388551e-01 -3.39469984e-02 3.10562462e-01 4.85775992e-02
-3.54629517e-01 2.42188096e-01 -1.21640468e+00 7.63155818e-01
-6.08067095e-01 3.94633353e-01 3.19900960e-01 -7.57245183e-01
6.59378052e-01 1.78939849e-01 3.31610411e-01 -8.31448019e-01
6.91752851e-01 1.89759180e-01 1.71091825e-01 -2.93426692e-01
4.30644572e-01 -3.28992844e-01 -5.48066080e-01 2.65031993e-01
-3.80155504e-01 -6.09131753e-01 1.92867443e-01 -9.67231393e-02
1.03175282e+00 -4.37404215e-03 6.08630478e-01 -2.72102386e-01
5.57473898e-01 2.80339658e-01 6.58777773e-01 5.30318975e-01
-4.83027220e-01 -3.84010524e-01 8.57427478e-01 1.57178536e-01
-1.22072935e+00 -6.58484340e-01 -1.90065473e-01 1.40950769e-01
4.94098902e-01 -1.53644770e-01 -5.92694879e-01 -3.77526104e-01
4.25120473e-01 1.04443014e+00 -5.49874723e-01 -9.31273460e-01
-2.66603261e-01 -3.56488407e-01 6.65408492e-01 4.21134323e-01
5.63372448e-02 -2.59781569e-01 -1.31383741e+00 1.70679882e-01
4.55654860e-01 -1.07593811e+00 -1.20764725e-01 1.76379815e-01
-6.79467857e-01 -1.24742866e+00 2.70707279e-01 -1.30028158e-01
5.97610831e-01 -1.30478755e-01 5.86336493e-01 -1.64727326e-02
-3.22287858e-01 1.76754788e-01 1.20407894e-01 -5.24638355e-01
-8.38604450e-01 -4.22733098e-01 4.08699334e-01 -4.42312598e-01
-2.07935348e-01 -2.92687297e-01 -2.88018495e-01 7.32845128e-01
-6.80852473e-01 -1.39263943e-01 6.16028965e-01 9.74026024e-01
5.97744107e-01 5.46873033e-01 6.53302968e-01 -3.82332176e-01
6.91852391e-01 -3.71668965e-01 -1.75535250e+00 2.42257312e-01
-9.56852555e-01 2.82636315e-01 8.98161352e-01 -3.86264384e-01
-5.56341767e-01 5.83994016e-02 1.62118882e-01 -3.14600408e-01
2.16083065e-01 4.10898179e-01 -2.99144059e-01 -4.00006860e-01
4.61759955e-01 -1.22831002e-01 4.38488603e-01 2.57268250e-01
-4.23735753e-02 2.52112776e-01 4.11617488e-01 -7.86575198e-01
1.04365289e+00 1.49414077e-01 5.16903281e-01 -4.88333851e-01
-1.85593173e-01 -2.94289980e-02 1.35222927e-01 -4.71731752e-01
4.14972268e-02 -8.42876315e-01 -1.48420954e+00 -8.19606856e-02
-5.02413511e-01 -3.32231671e-01 -4.39193606e-01 5.12554824e-01
-6.18882358e-01 1.90387800e-01 8.92495140e-02 -1.42067933e+00
2.07303807e-01 -1.53707087e+00 9.67299879e-01 1.46834791e-01
-4.29003924e-01 -4.50526953e-01 8.31150934e-02 -6.32554963e-02
1.90691158e-01 5.79922855e-01 7.62271702e-01 -2.08178312e-01
-6.16619110e-01 -6.71848536e-01 1.71244353e-01 1.63981378e-01
-3.77913624e-01 1.49380997e-01 -6.49681866e-01 -7.92357266e-01
-1.08304590e-01 -3.30756307e-01 -5.36350124e-02 2.93673158e-01
8.40215445e-01 -4.59319323e-01 -3.47576082e-01 2.13933825e-01
1.38920116e+00 2.57411361e-01 3.74042302e-01 3.59886438e-01
-9.13043022e-02 6.73369646e-01 1.30447054e+00 8.30416977e-01
1.32267103e-01 4.84375805e-01 8.57684970e-01 2.93379217e-01
7.26849258e-01 -2.41761237e-01 5.01179874e-01 1.64259635e-02
4.53290790e-01 -7.10827485e-02 -8.71964157e-01 5.87659359e-01
-1.63352454e+00 -5.78507364e-01 1.86330348e-01 2.80717206e+00
5.70823312e-01 5.94401121e-01 3.93684119e-01 4.72949564e-01
6.12651825e-01 -6.01877868e-01 -8.75707269e-01 -4.45967495e-01
3.38081777e-01 -2.01062169e-02 1.05812228e+00 4.80214506e-01
-7.93123543e-01 3.65869045e-01 6.38321447e+00 4.60500628e-01
-9.64719176e-01 -2.38735363e-01 2.57197917e-01 -2.27922723e-01
-1.29225343e-01 3.58343989e-01 -6.50792360e-01 2.25772068e-01
1.51843941e+00 -9.20411944e-01 3.22322279e-01 1.12058902e+00
7.05576718e-01 -4.17951316e-01 -1.17452276e+00 3.99805844e-01
-6.21109843e-01 -1.03945482e+00 -6.24374270e-01 2.15975001e-01
6.47000015e-01 -1.67066246e-01 1.88346639e-01 4.59739208e-01
4.19444144e-01 -9.85121489e-01 1.07108748e+00 5.60756922e-02
8.46983314e-01 -1.21229327e+00 9.04895425e-01 3.01874995e-01
-8.53658617e-01 -4.10439968e-01 1.92489564e-01 -2.55644202e-01
3.65027905e-01 3.44103098e-01 -9.52173710e-01 4.33696806e-01
2.98770368e-01 3.98579799e-02 -3.14976782e-01 1.01712692e+00
-2.64755934e-02 4.31852549e-01 -5.37311435e-01 -3.78843009e-01
2.16416404e-01 -1.09400094e-01 6.33083761e-01 6.67263985e-01
2.49014005e-01 -1.07472330e-01 2.30730474e-01 8.82191956e-01
4.80436891e-01 -4.89586353e-01 -7.01373696e-01 -5.13358057e-01
7.16546595e-01 9.11701083e-01 -5.04095435e-01 -2.15498246e-02
-8.80969167e-02 -2.53959075e-02 -2.63378292e-01 2.21595511e-01
-1.42406607e+00 -3.67996365e-01 1.06569266e+00 1.21216243e-02
7.98186883e-02 -4.39376175e-01 -5.20287633e-01 -4.98018116e-01
1.30250648e-01 -1.08382070e+00 7.42122754e-02 -2.01175407e-01
-5.10811687e-01 1.19581766e-01 2.87496924e-01 -1.55866122e+00
-7.27460921e-01 -4.77921575e-01 -3.05965483e-01 6.43796980e-01
-1.22044587e+00 -4.86820549e-01 -1.72388833e-02 1.17717668e-01
-2.34002750e-02 1.24967374e-01 4.89053905e-01 2.21025333e-01
-9.94928122e-01 6.48256481e-01 4.19483846e-03 -6.90760195e-01
1.79211423e-01 -5.48984528e-01 2.20986292e-01 1.18450654e+00
-9.59777951e-01 5.71244776e-01 1.34969056e+00 -7.83911467e-01
-1.92416263e+00 -1.29449880e+00 8.81517753e-02 -5.84336743e-02
1.03287172e+00 -3.16625983e-01 -5.18521428e-01 2.42115080e-01
-3.71379524e-01 -2.05674872e-01 1.02852292e-01 -1.89992711e-01
1.33604303e-01 -2.39374965e-01 -1.26772523e+00 9.28189933e-01
7.44172990e-01 -3.94978166e-01 -2.64196582e-02 4.73170588e-03
7.36688972e-01 -2.56136239e-01 -9.32476103e-01 7.43586361e-01
6.36738360e-01 -5.55602491e-01 5.53188145e-01 -4.39220846e-01
-1.27557039e-01 -5.76646745e-01 2.72682961e-02 -8.83270264e-01
1.91651389e-01 -9.85596299e-01 -2.31053486e-01 7.67570615e-01
5.82884431e-01 -8.07574868e-01 4.99102563e-01 7.89656401e-01
-9.38914940e-02 -7.17274964e-01 -1.02637911e+00 -1.26022410e+00
-1.28963023e-01 -7.59332180e-01 7.19653428e-01 4.40327764e-01
3.13449919e-01 -2.06477586e-02 -2.48884633e-01 7.49807835e-01
7.38043904e-01 -6.33622780e-02 6.95291758e-01 -7.73350596e-01
-4.60241616e-01 -4.51684892e-01 -5.99380910e-01 -2.09405035e-01
3.14038396e-01 -2.19276577e-01 4.10350710e-01 -8.09870720e-01
-2.39233419e-01 -4.62203324e-01 9.33545157e-02 2.13559628e-01
1.60106331e-01 -1.30101487e-01 -2.66113400e-01 -2.86390185e-01
-5.63444197e-01 7.78874218e-01 8.82521212e-01 -1.13290943e-01
-8.53877217e-02 2.11108670e-01 -2.77366012e-01 4.12467659e-01
7.22354174e-01 -2.85164475e-01 -8.32533002e-01 1.92027986e-01
5.07849634e-01 4.06871259e-01 6.61038160e-01 -1.13179302e+00
1.36688715e-02 -6.48937106e-01 -2.09298357e-01 -4.88411307e-01
1.33563980e-01 -1.04694557e+00 6.76193476e-01 9.82798934e-01
-2.21578270e-01 1.62417322e-01 6.99311197e-01 7.53211141e-01
-2.07312316e-01 -5.89347146e-02 8.91218185e-01 4.81854618e-01
-5.03262877e-01 1.90876544e-01 -5.21376193e-01 -1.61819682e-01
1.64445972e+00 -2.12517783e-01 -1.54155657e-01 -3.49577159e-01
-2.85221308e-01 8.23832333e-01 7.00525701e-01 3.64769489e-01
4.01848137e-01 -9.44598675e-01 -2.94242591e-01 3.62979084e-01
2.07101077e-01 -2.26054922e-01 2.42304504e-01 1.01813507e+00
-3.53367567e-01 5.64366817e-01 -1.40976876e-01 -5.09487212e-01
-9.67917621e-01 5.28322995e-01 4.67307866e-01 3.56945172e-02
-2.66898125e-01 3.14888597e-01 2.17944071e-01 -1.98530167e-01
1.24949172e-01 -4.14419115e-01 4.16488171e-01 -5.94833732e-01
2.08306223e-01 5.00737727e-01 2.06155196e-01 -1.95438847e-01
-6.55548990e-01 -5.88276982e-03 2.33331010e-01 -3.77432704e-01
8.11094284e-01 -7.14014098e-02 2.26471826e-01 1.89751402e-01
7.99848914e-01 -4.57863629e-01 -1.45054340e+00 6.01333857e-01
-2.28292663e-02 -3.77056926e-01 2.47740045e-01 -5.73550284e-01
-3.87412727e-01 3.99373859e-01 4.77034777e-01 1.95515320e-01
1.08890522e+00 -5.73259532e-01 4.28552121e-01 5.24279296e-01
7.60376871e-01 -1.24999416e+00 -4.10852611e-01 2.77175099e-01
6.09076083e-01 -8.75135064e-01 1.62426904e-01 -4.01376903e-01
-5.87713003e-01 5.89730442e-01 7.89712310e-01 -1.75307810e-01
3.23084801e-01 8.31853330e-01 -5.45170426e-01 1.87439293e-01
-1.08893013e+00 1.69688448e-01 -1.00406474e-02 2.81098068e-01
-2.20930189e-01 2.26304650e-01 -3.97008121e-01 6.88701510e-01
-5.71703278e-02 -2.77379658e-02 6.24969900e-01 1.18762910e+00
-2.77071834e-01 -8.66057754e-01 -3.50446105e-01 3.13599229e-01
5.21247089e-02 4.91406441e-01 -1.33595616e-01 1.21053076e+00
-1.59282118e-01 9.47873890e-01 -1.62581474e-01 -5.77066243e-01
5.10201573e-01 -1.42692581e-01 4.55836713e-01 -1.71476647e-01
-2.29813173e-01 -1.91516966e-01 6.41145885e-01 -8.57201457e-01
9.37447473e-02 -6.87036991e-01 -1.16318512e+00 -4.89536822e-01
-7.22098768e-01 3.92486364e-01 9.49246466e-01 9.86176193e-01
5.68137467e-01 6.53303742e-01 8.73367250e-01 -5.14393747e-01
-1.27465904e+00 -4.03243959e-01 -4.39216793e-01 -3.17621559e-01
5.26723921e-01 -1.00978959e+00 -4.34853017e-01 -4.95270878e-01] | [4.922143936157227, 2.255983591079712] |
626ccc71-7fec-4b8e-8d81-2ac380229d04 | multi-modal-gait-recognition-via-effective | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Cui_Multi-Modal_Gait_Recognition_via_Effective_Spatial-Temporal_Feature_Fusion_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Cui_Multi-Modal_Gait_Recognition_via_Effective_Spatial-Temporal_Feature_Fusion_CVPR_2023_paper.pdf | Multi-Modal Gait Recognition via Effective Spatial-Temporal Feature Fusion | Gait recognition is a biometric technology that identifies people by their walking patterns. The silhouettes-based method and the skeletons-based method are the two most popular approaches. However, the silhouette data are easily affected by clothing occlusion, and the skeleton data lack body shape information. To obtain a more robust and comprehensive gait representation for recognition, we propose a transformer-based gait recognition framework called MMGaitFormer, which effectively fuses and aggregates the spatial-temporal information from the skeletons and silhouettes. Specifically, a Spatial Fusion Module (SFM) and a Temporal Fusion Module (TFM) are proposed for effective spatial-level and temporal-level feature fusion, respectively. The SFM performs fine-grained body parts spatial fusion and guides the alignment of each part of the silhouette and each joint of the skeleton through the attention mechanism. The TFM performs temporal modeling through Cycle Position Embedding (CPE) and fuses temporal information of two modalities. Experiments demonstrate that our MMGaitFormer achieves state-of-the-art performance on popular gait datasets. For the most challenging "CL" (i.e., walking in different clothes) condition in CASIA-B, our method achieves a rank-1 accuracy of 94.8%, which outperforms the state-of-the-art single-modal methods by a large margin. | ['Yimei Kang', 'Yufeng Cui'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['gait-recognition'] | ['computer-vision'] | [-1.60196781e-01 -7.11403906e-01 -3.06660861e-01 -1.20222807e-01
-5.35735548e-01 2.56287046e-02 3.97980452e-01 -2.46491238e-01
-2.88188934e-01 4.35641438e-01 4.51716214e-01 5.09284198e-01
1.42047346e-01 -7.30422795e-01 -2.77515352e-01 -1.09833109e+00
-3.43502052e-02 1.71182647e-01 2.43683353e-01 -1.88609123e-01
-7.66530037e-02 2.38892555e-01 -1.43022692e+00 -2.35917456e-02
6.70586467e-01 1.13534212e+00 -3.06011826e-01 3.70769799e-01
1.59714058e-01 1.98252201e-01 -3.52911830e-01 -7.36286342e-01
3.14382985e-02 -2.14142382e-01 -2.76998222e-01 1.34582072e-01
5.70819020e-01 -1.70853853e-01 -8.13993752e-01 8.76966596e-01
8.22923541e-01 2.35448452e-03 6.01140141e-01 -1.28900588e+00
-5.85070729e-01 7.65784178e-03 -1.18907714e+00 1.19894825e-01
7.66714692e-01 4.46851879e-01 8.00456047e-01 -9.63607430e-01
3.34227115e-01 1.48218131e+00 8.87081027e-01 5.52924335e-01
-1.12793887e+00 -6.50159717e-01 8.49843249e-02 5.12391746e-01
-1.65351236e+00 -3.98027956e-01 9.83282983e-01 -3.82404566e-01
7.26541221e-01 1.71910584e-01 9.91042376e-01 1.19413710e+00
5.30786276e-01 1.01250613e+00 9.16754901e-01 -1.42055631e-01
-2.69345969e-01 -7.13665187e-01 2.00231060e-01 1.04356647e+00
5.40708721e-01 1.69179514e-01 -9.68443871e-01 -1.48267254e-01
9.53564942e-01 3.07854265e-01 -2.20839798e-01 -2.82739431e-01
-1.41445005e+00 3.08988452e-01 2.70927876e-01 2.28031069e-01
-4.57032442e-01 3.00465196e-01 5.31543016e-01 -5.28839156e-02
2.00165659e-01 -4.60952103e-01 -2.21318696e-02 -2.84554899e-01
-1.15068734e+00 3.32041174e-01 4.41940755e-01 5.54917276e-01
3.73651147e-01 1.09887972e-01 -4.52225655e-01 9.77344036e-01
7.97395110e-01 9.82811332e-01 6.51546061e-01 -5.77918351e-01
6.87328100e-01 6.32958531e-01 -2.06828844e-02 -1.45141029e+00
-2.91303247e-01 -2.37801194e-01 -1.12897098e+00 -3.19008321e-01
5.84948182e-01 1.86607108e-01 -9.49941516e-01 1.76107907e+00
3.69130403e-01 4.18930829e-01 -4.03432578e-01 1.09350443e+00
7.47499406e-01 3.47490042e-01 2.41905138e-01 -1.42271653e-01
1.79582632e+00 -7.79201627e-01 -7.46315360e-01 -1.80048376e-01
1.22820601e-01 -6.31659269e-01 8.24105442e-01 1.36564806e-01
-1.02831924e+00 -8.42335999e-01 -1.14701605e+00 1.10140346e-01
-1.91547513e-01 4.21446681e-01 2.69978493e-01 7.82638907e-01
-6.61750495e-01 4.95110065e-01 -1.19891775e+00 -4.62686330e-01
1.55034438e-01 3.79228055e-01 -6.77384853e-01 -1.99206084e-01
-1.18806040e+00 7.37020195e-01 7.90539570e-03 3.06356579e-01
-4.06315327e-01 -2.80373693e-01 -1.15611148e+00 -2.15369314e-01
-4.42374274e-02 -9.43240225e-01 4.92365181e-01 -1.16097771e-01
-1.28400195e+00 9.26055014e-01 -5.14503479e-01 -2.58347690e-01
5.19293368e-01 -2.09126398e-01 -6.73809767e-01 3.27835351e-01
2.75789410e-01 3.04450303e-01 1.26228416e+00 -7.17833459e-01
-4.38028634e-01 -7.99810648e-01 -6.47955835e-01 1.18596084e-01
-3.58095765e-01 -1.98141247e-01 -8.89912605e-01 -1.12812972e+00
4.04857606e-01 -8.40288222e-01 2.06116289e-01 1.06980160e-01
-2.42576987e-01 -1.67629942e-01 8.87254357e-01 -1.24243927e+00
1.61684752e+00 -2.18370104e+00 3.69765043e-01 2.08412275e-01
1.97565988e-01 2.08439350e-01 1.18285209e-01 2.45016515e-01
5.75067475e-03 -2.26882175e-01 -1.31250486e-01 -7.24756956e-01
2.55023897e-01 2.33574316e-01 1.46427959e-01 7.78953910e-01
1.19760279e-02 1.00548160e+00 -7.70777583e-01 -8.60562801e-01
4.43157703e-01 6.32314205e-01 -2.95775592e-01 3.87846716e-02
5.95658779e-01 4.85090792e-01 -3.92374843e-01 1.23496091e+00
6.63879156e-01 -8.23394880e-02 7.42732286e-02 -8.16635966e-01
2.16939881e-01 -1.34657070e-01 -1.24691701e+00 1.76333463e+00
-6.97794482e-02 3.47925089e-02 -1.75713897e-01 -8.54557872e-01
1.06840611e+00 3.84588063e-01 9.61675465e-01 -7.06017375e-01
6.36334196e-02 1.91502690e-01 -2.49683335e-01 -5.39212465e-01
2.50629783e-01 -9.16657448e-02 -3.34901273e-01 1.81101188e-01
2.17757791e-01 6.53426349e-01 3.06657523e-01 -2.14830354e-01
9.08648551e-01 3.93315673e-01 3.37771863e-01 -8.28662366e-02
7.68483162e-01 -6.61413789e-01 9.83425915e-01 1.74611092e-01
-7.25202560e-01 7.23459303e-01 -4.15853299e-02 -5.13087034e-01
-8.27182591e-01 -1.32392991e+00 1.23320423e-01 6.67589605e-01
3.46790671e-01 -6.28031790e-01 -6.34209573e-01 -5.46286941e-01
3.67842972e-01 -1.98973909e-01 -7.17778683e-01 -4.57046986e-01
-7.98658252e-01 -8.99859905e-01 8.36407959e-01 9.20059919e-01
1.19346952e+00 -5.22173464e-01 -5.21381974e-01 2.90933996e-01
-6.95361018e-01 -9.85809863e-01 -1.16807067e+00 -6.46073759e-01
-8.10448885e-01 -1.09412444e+00 -1.20507216e+00 -5.56043386e-01
4.87520695e-01 1.57283708e-01 7.06944823e-01 7.22630844e-02
-4.08505291e-01 4.84278709e-01 -1.93040833e-01 3.86226952e-01
5.01678765e-01 -2.78796017e-01 4.69723076e-01 6.99544847e-01
5.13878465e-01 -7.58233368e-01 -9.49998736e-01 5.69375694e-01
-3.30335528e-01 -2.52526309e-02 5.18181443e-01 1.11002660e+00
4.99246776e-01 -4.92042042e-02 2.20603317e-01 6.54212758e-02
3.01367760e-01 -1.84113383e-01 1.50613142e-02 3.51629674e-01
-3.43907803e-01 -1.06832586e-01 1.56716585e-01 -5.14184117e-01
-7.57487118e-01 4.54365723e-02 -1.11102961e-01 -6.64072812e-01
-3.26771922e-02 4.39574867e-01 -4.77928400e-01 -2.26906642e-01
1.76405117e-01 6.78402960e-01 4.53582965e-02 -6.67704821e-01
4.65743840e-02 3.85116696e-01 1.03043079e+00 -9.52607155e-01
7.52466500e-01 5.09192407e-01 1.41724765e-01 -8.28814328e-01
-1.27549976e-01 -6.44643128e-01 -8.62408757e-01 -5.51584065e-01
8.71621788e-01 -9.11419094e-01 -9.64515090e-01 9.95175719e-01
-9.13955629e-01 3.39062184e-01 -5.03410921e-02 4.77652490e-01
-5.38829863e-01 8.66680622e-01 -9.22931910e-01 -7.77879357e-01
-5.35466194e-01 -9.58013594e-01 1.42780650e+00 4.23316270e-01
-2.65439600e-01 -7.78510749e-01 1.35000665e-02 5.84239423e-01
1.10067045e-02 5.88157296e-01 4.72545266e-01 -8.40400755e-02
-1.50577769e-01 -4.22903866e-01 -1.65312096e-01 5.53649524e-03
3.50258440e-01 -1.58798061e-02 -6.14432693e-01 -4.84349757e-01
-2.75908500e-01 2.62597017e-02 1.00139463e+00 4.31953281e-01
5.02134025e-01 -2.38140717e-01 -3.79249692e-01 5.94155550e-01
1.14848292e+00 1.55006856e-01 7.75050640e-01 1.41556054e-01
8.99780571e-01 4.10382271e-01 6.72699630e-01 5.53723335e-01
7.02684641e-01 1.05693793e+00 -8.43086746e-03 1.16753206e-01
-3.97234410e-01 -3.71112436e-01 6.78568661e-01 1.00035954e+00
-6.15492880e-01 2.75137603e-01 -8.87559772e-01 6.42361641e-01
-2.08540177e+00 -1.07935846e+00 1.36284962e-01 2.23690724e+00
5.86882889e-01 -2.05417480e-02 6.98157668e-01 4.20799494e-01
8.76371861e-01 3.29736799e-01 -5.26396275e-01 3.15415412e-01
-1.65879652e-01 1.24724172e-01 3.73294622e-01 8.76584053e-02
-1.37658215e+00 6.33206666e-01 5.50518560e+00 8.86717558e-01
-9.00400102e-01 1.21607050e-01 1.69205472e-01 1.23068012e-01
4.00853872e-01 -4.75085825e-01 -7.90519297e-01 9.91603017e-01
5.92441022e-01 3.17055658e-02 8.99232998e-02 4.92610186e-01
2.25267887e-01 5.83558418e-02 -7.56385684e-01 1.51275885e+00
2.01595515e-01 -9.41578507e-01 7.21190311e-03 1.75754726e-01
1.83424518e-01 -4.75706547e-01 1.19411252e-01 1.17505461e-01
-2.34170467e-01 -7.65818000e-01 6.70000672e-01 8.84178579e-01
9.12637293e-01 -6.72357619e-01 7.75175929e-01 -1.00453831e-01
-2.22198033e+00 4.20984030e-02 1.62143223e-02 2.14779064e-01
3.67484242e-01 5.04965007e-01 8.38173702e-02 9.78043199e-01
8.49203169e-01 1.26545465e+00 -5.43478727e-01 1.21500885e+00
-9.81024727e-02 5.42264044e-01 -3.92065883e-01 3.48915160e-01
-1.67996690e-01 -2.32983738e-01 5.77270031e-01 1.34100056e+00
3.54954571e-01 1.21475302e-01 3.75455022e-01 5.36640644e-01
2.35172361e-01 -5.63634858e-02 -2.54221797e-01 -4.35429923e-02
3.23617786e-01 1.00493300e+00 -4.86739665e-01 -2.52836615e-01
-3.98390412e-01 1.15518022e+00 -1.56375557e-01 2.35421643e-01
-8.00096095e-01 -2.40418896e-01 9.32715237e-01 5.90256676e-02
2.45283842e-01 -4.18961048e-01 -2.71618664e-01 -1.66991568e+00
3.07041943e-01 -7.83142209e-01 7.50744879e-01 -4.67889220e-01
-1.57679391e+00 7.93240592e-02 -7.18967542e-02 -1.49511361e+00
-7.23033771e-02 -5.43459237e-01 -6.32049918e-01 8.99092793e-01
-9.44026649e-01 -1.74108458e+00 -4.05131698e-01 9.08526003e-01
3.07398736e-01 -2.32983246e-01 7.07292914e-01 7.75369108e-01
-8.61213684e-01 9.34493423e-01 -3.76549512e-01 5.76798737e-01
7.28342235e-01 -9.12274003e-01 5.74366570e-01 1.01729143e+00
-1.10112563e-01 9.39303041e-01 5.77516079e-01 -9.32639122e-01
-1.67291522e+00 -9.01630044e-01 8.33238363e-01 -1.87595904e-01
4.90881592e-01 -1.39694931e-02 -7.89745152e-01 2.94541061e-01
-3.41011763e-01 9.17972848e-02 6.96605206e-01 -6.44426644e-02
-5.37303448e-01 -3.02945405e-01 -1.12552118e+00 5.90985715e-01
1.12175310e+00 -5.71534932e-01 -8.20450366e-01 -2.07433954e-01
1.63593963e-01 -3.29487443e-01 -1.27529728e+00 5.85667729e-01
1.26181197e+00 -6.65292144e-01 1.31455207e+00 -2.86463290e-01
-9.26870387e-03 -7.48415232e-01 -4.35398519e-01 -8.83328676e-01
-4.45338339e-01 -3.91554594e-01 -7.43548810e-01 1.23357666e+00
-3.86797041e-01 -5.81695795e-01 8.88918638e-01 4.35687542e-01
2.06715912e-01 -7.48150408e-01 -1.23018634e+00 -9.39760029e-01
-4.26287711e-01 -2.54805148e-01 5.55331767e-01 8.23503375e-01
6.70772269e-02 8.77359733e-02 -8.92589092e-01 -1.21778948e-02
1.05725753e+00 8.86399075e-02 6.74998462e-01 -1.14816833e+00
-2.18590900e-01 -4.59656894e-01 -1.02746129e+00 -1.06679142e+00
-1.32708892e-01 -5.82695603e-01 -2.61434823e-01 -1.33986115e+00
2.35079885e-01 1.40001729e-01 -4.35845256e-01 6.00721955e-01
-2.16539234e-01 5.44289827e-01 3.58241916e-01 3.52647692e-01
-4.32561785e-01 8.08974326e-01 1.20339000e+00 -4.65006649e-01
8.71484503e-02 -1.17748044e-01 -2.54094660e-01 6.68763280e-01
4.18883145e-01 4.98465076e-02 2.38236468e-02 -7.21323043e-02
-4.58857208e-01 9.01630074e-02 6.28733099e-01 -1.26823056e+00
3.87444496e-01 4.35345387e-03 8.19298387e-01 -8.45012724e-01
7.52017856e-01 -5.74682951e-01 3.42422456e-01 8.84930730e-01
3.54154646e-01 1.66611716e-01 -8.26890171e-02 6.70389116e-01
-2.58176416e-01 5.78988373e-01 8.02189052e-01 5.94250970e-02
-6.76446855e-01 7.34933317e-01 -1.57385767e-01 -2.69470572e-01
6.24979794e-01 -6.49202883e-01 -2.36746162e-01 -1.92329735e-01
-1.03915989e+00 2.17968971e-01 3.59324396e-01 5.16083658e-01
9.52754200e-01 -1.95436823e+00 -6.23363078e-01 4.87850100e-01
1.94157094e-01 -6.88855469e-01 5.03131449e-01 1.28995323e+00
-2.46248424e-01 3.01160604e-01 -4.92017359e-01 -7.52761126e-01
-1.43391109e+00 1.97610408e-01 4.13027197e-01 -5.05370677e-01
-6.68454647e-01 5.80932856e-01 -3.79394442e-02 -7.70815536e-02
9.90492702e-02 2.07903590e-02 -1.68228880e-01 2.05248177e-01
6.62385285e-01 7.06825018e-01 -6.12516031e-02 -1.31618202e+00
-8.13462853e-01 1.25672042e+00 2.40095288e-01 -1.49402827e-01
8.78058016e-01 -2.48035744e-01 -4.39233221e-02 3.31532270e-01
1.00923979e+00 -3.13333035e-01 -1.18249893e+00 -3.72869372e-01
-1.54822916e-01 -7.26451516e-01 -3.59665751e-01 -5.54913700e-01
-1.26449621e+00 1.10936892e+00 7.59958267e-01 -3.62295002e-01
1.25262547e+00 -4.59932745e-01 1.32522631e+00 -1.48778141e-01
7.41395891e-01 -1.01001227e+00 3.01290423e-01 3.50548983e-01
7.52473891e-01 -8.58770490e-01 6.64063841e-02 -3.59387368e-01
-4.86104727e-01 9.71648932e-01 5.62450767e-01 -1.58174694e-01
6.69595301e-01 -3.55380662e-02 -3.29066604e-01 -3.79080400e-02
-3.74754637e-01 -2.88728088e-01 8.04892302e-01 6.65090978e-01
3.48327488e-01 2.83572853e-01 -5.97081482e-01 1.00567305e+00
3.41682024e-02 -1.07078157e-01 -4.36116815e-01 9.75927055e-01
-2.66089350e-01 -1.13908052e+00 -6.56813443e-01 3.68257761e-01
-3.44490290e-01 2.80318648e-01 -2.09969327e-01 5.84348023e-01
3.56410980e-01 7.93299198e-01 -2.21567079e-01 -1.05231249e+00
4.21866864e-01 1.81114390e-01 7.43067741e-01 -1.05020501e-01
-3.16064239e-01 3.90224904e-01 -8.19262769e-03 -9.13285196e-01
-5.49531758e-01 -1.03961897e+00 -9.56321418e-01 -6.51305854e-01
7.69651830e-02 -1.11540049e-01 9.30871218e-02 1.01177251e+00
2.29063526e-01 2.64243156e-01 3.96638840e-01 -1.10889828e+00
-2.73077995e-01 -8.08374345e-01 -8.34355056e-01 6.79194748e-01
2.12891102e-01 -1.24850190e+00 6.66573867e-02 3.42188865e-01] | [14.285103797912598, 1.4211947917938232] |
1e28e6cb-d4e4-4a1d-94dc-93dc32a4b0dd | chexpert-a-large-chest-radiograph-dataset | 1901.07031 | null | http://arxiv.org/abs/1901.07031v1 | http://arxiv.org/pdf/1901.07031v1.pdf | CheXpert: A Large Chest Radiograph Dataset with Uncertainty Labels and Expert Comparison | Large, labeled datasets have driven deep learning methods to achieve
expert-level performance on a variety of medical imaging tasks. We present
CheXpert, a large dataset that contains 224,316 chest radiographs of 65,240
patients. We design a labeler to automatically detect the presence of 14
observations in radiology reports, capturing uncertainties inherent in
radiograph interpretation. We investigate different approaches to using the
uncertainty labels for training convolutional neural networks that output the
probability of these observations given the available frontal and lateral
radiographs. On a validation set of 200 chest radiographic studies which were
manually annotated by 3 board-certified radiologists, we find that different
uncertainty approaches are useful for different pathologies. We then evaluate
our best model on a test set composed of 500 chest radiographic studies
annotated by a consensus of 5 board-certified radiologists, and compare the
performance of our model to that of 3 additional radiologists in the detection
of 5 selected pathologies. On Cardiomegaly, Edema, and Pleural Effusion, the
model ROC and PR curves lie above all 3 radiologist operating points. We
release the dataset to the public as a standard benchmark to evaluate
performance of chest radiograph interpretation models.
The dataset is freely available at
https://stanfordmlgroup.github.io/competitions/chexpert . | ['Jesse K. Sandberg', 'Safwan S. Halabi', 'Katie Shpanskaya', 'Behzad Haghgoo', 'Silviana Ciurea-Ilcus', 'Pranav Rajpurkar', 'Henrik Marklund', 'Andrew Y. Ng', 'Robyn Ball', 'Ricky Jones', 'David B. Larson', 'David A. Mong', 'Yifan Yu', 'Jeremy Irvin', 'Curtis P. Langlotz', 'Chris Chute', 'Michael Ko', 'Matthew P. Lungren', 'Jayne Seekins', 'Bhavik N. Patel'] | 2019-01-21 | null | null | null | null | ['lung-disease-classification'] | ['medical'] | [-2.04269644e-02 3.82437557e-01 -2.17524797e-01 -8.69126558e-01
-1.73792922e+00 -5.61547101e-01 1.67286266e-02 4.34167206e-01
-5.52196860e-01 6.74258947e-01 2.44740456e-01 -7.77685463e-01
-2.65188903e-01 -3.19705755e-01 -9.46086943e-01 -4.45928037e-01
-1.00176908e-01 9.56310689e-01 3.00941437e-01 6.65887535e-01
-3.29497367e-01 1.46592572e-01 -5.48245430e-01 7.76538193e-01
3.86243433e-01 1.27815855e+00 -8.44493061e-02 1.04606915e+00
4.58674073e-01 1.13509977e+00 -4.72048700e-01 -6.02395236e-01
1.07520640e-01 -2.73449153e-01 -8.26578557e-01 1.10195674e-01
3.88607889e-01 -7.37446964e-01 -5.66241741e-01 7.58738339e-01
7.88734853e-01 -2.47383058e-01 1.02766526e+00 -7.30869591e-01
-7.68789113e-01 1.20917165e+00 -6.47285998e-01 7.63388872e-01
3.44621614e-02 2.25717440e-01 8.79320502e-01 -8.75008762e-01
4.59031254e-01 6.37857020e-01 8.99758399e-01 3.22849184e-01
-8.55822384e-01 -4.97736007e-01 -3.04188281e-01 5.92862144e-02
-1.28545892e+00 1.26621068e-01 1.01558037e-01 -5.41292250e-01
4.29491848e-01 2.19455510e-01 1.27549812e-01 1.37290299e+00
7.24715710e-01 5.58248341e-01 8.72943640e-01 -2.01265126e-01
3.74435782e-01 3.57263759e-02 5.80763280e-01 8.76765132e-01
4.06547189e-01 3.39350253e-02 -9.75317657e-02 -6.93197072e-01
6.14492595e-01 1.12953596e-01 -4.47965741e-01 -1.64037868e-01
-1.31421638e+00 7.87016869e-01 6.63283646e-01 -1.43014062e-02
-4.01064873e-01 3.95376742e-01 3.28218073e-01 -1.27807945e-01
5.84228598e-02 4.42927808e-01 -4.03105736e-01 2.76427537e-01
-8.25631440e-01 -1.75429024e-02 7.89825201e-01 7.25916445e-01
-1.49566248e-01 -4.21122193e-01 -5.22963166e-01 7.15765178e-01
3.39317381e-01 5.36660254e-01 6.83593035e-01 -1.12599707e+00
3.13611656e-01 1.44306809e-01 -5.41959982e-03 -3.01430285e-01
-9.13100123e-01 -5.20878375e-01 -8.06386530e-01 -1.60283372e-02
5.08646548e-01 -4.66672212e-01 -1.27992260e+00 1.29652178e+00
1.90469790e-02 4.50419728e-03 -2.17057183e-01 9.55259204e-01
1.18427896e+00 9.27788913e-02 2.05469161e-01 8.52001980e-02
1.77945375e+00 -7.11171210e-01 -4.13262904e-01 -4.77598235e-03
7.75013626e-01 -5.74495971e-01 8.06816816e-01 4.48441327e-01
-1.05325472e+00 -3.16985548e-01 -9.80775416e-01 2.44509295e-01
2.24674731e-01 4.92267698e-01 9.41151604e-02 4.56982791e-01
-7.90035903e-01 4.61773664e-01 -1.32370734e+00 2.21973956e-01
7.56157041e-01 3.55075151e-01 -2.52611309e-01 -5.20746373e-02
-1.21881366e+00 9.43906903e-01 3.18426311e-01 -1.39167935e-01
-1.10297585e+00 -7.72284448e-01 -8.21628869e-01 7.26127923e-02
4.99411374e-01 -7.40805686e-01 1.95219839e+00 -2.03378439e-01
-6.52887762e-01 1.09056723e+00 3.60890895e-01 -8.06566894e-01
9.42500889e-01 -8.94310474e-02 -2.78845131e-01 4.87906635e-01
2.06645191e-01 5.32362163e-01 5.38114309e-01 -1.06948209e+00
-5.25974095e-01 -1.09777510e-01 -2.75378019e-01 -1.62537590e-01
2.57974446e-01 -6.65315762e-02 -4.00547594e-01 -7.05812693e-01
1.49174243e-01 -1.14031613e+00 -4.52739626e-01 7.52398968e-02
-5.93420088e-01 -7.53146410e-02 2.17812672e-01 -5.63739955e-01
9.07515943e-01 -2.08616805e+00 -5.73431551e-01 3.25795233e-01
7.20450580e-01 -2.30989367e-01 4.29093152e-01 -2.20439792e-01
-4.28305179e-01 3.71573806e-01 -6.04386747e-01 -1.78397775e-01
-1.83390006e-01 1.84454843e-01 -5.07015362e-02 4.63619918e-01
3.30499142e-01 1.00704789e+00 -7.92654037e-01 -8.85203242e-01
1.43179312e-01 2.99944520e-01 -3.25826198e-01 1.50675580e-01
-7.03572258e-02 6.27316654e-01 -5.48523784e-01 7.10964024e-01
3.54981840e-01 -1.14858985e+00 1.07314102e-01 -1.24046676e-01
7.51777649e-01 1.97560892e-01 -8.80613089e-01 1.51293802e+00
-3.29800665e-01 4.09361243e-01 -2.56365448e-01 -1.72503635e-01
5.18901527e-01 5.46294332e-01 4.90008980e-01 -5.48110642e-02
4.56314921e-01 3.77366781e-01 3.69546294e-01 -5.36782324e-01
-2.19690964e-01 -1.63221419e-01 -1.64444789e-01 8.69689226e-01
3.48938294e-02 -3.46404910e-01 -6.32255301e-02 3.98059547e-01
1.67142248e+00 -4.62000191e-01 4.14134711e-01 -8.21679905e-02
2.24166542e-01 3.58127803e-03 2.12901101e-01 1.26289809e+00
-4.84636873e-01 1.32160330e+00 5.83006740e-01 -5.53735793e-01
-8.26676965e-01 -1.57471931e+00 -8.24061096e-01 7.19458282e-01
-2.38970205e-01 -1.13031618e-01 -3.87186170e-01 -1.12117243e+00
5.05456068e-02 8.40126753e-01 -9.68882859e-01 -7.25121722e-02
-4.36728477e-01 -9.28896904e-01 7.02533841e-01 1.00087273e+00
1.22589216e-01 -9.42778885e-01 -9.83464956e-01 -7.23802820e-02
-2.06467837e-01 -1.07343900e+00 -4.97351050e-01 6.01110458e-01
-6.65883899e-01 -1.55229390e+00 -8.51925671e-01 -4.37858701e-01
5.73375285e-01 -3.95187855e-01 1.47862780e+00 1.45979419e-01
-6.45525694e-01 7.08391249e-01 -1.56635359e-01 -5.89694619e-01
-6.92759693e-01 7.22350478e-02 -1.84164777e-01 -6.08190477e-01
2.08134472e-01 -4.33702320e-02 -8.26696217e-01 3.09462100e-01
-9.92352724e-01 -3.55829865e-01 6.40643179e-01 1.05855072e+00
8.53072286e-01 -2.98997968e-01 3.84595782e-01 -1.25328767e+00
6.34777904e-01 -8.22180808e-01 -3.17008197e-01 4.19594288e-01
-4.66157943e-01 1.21776462e-01 2.42933854e-01 -2.17085611e-02
-1.07297122e+00 1.15122363e-01 -1.80914745e-01 -6.13708258e-01
-2.58241326e-01 4.21815395e-01 7.88919747e-01 3.72277707e-01
1.12014818e+00 -3.30485135e-01 -4.13894653e-01 -2.10462749e-01
8.28025714e-02 6.34355664e-01 9.01655734e-01 -6.00068808e-01
2.15897188e-01 3.40082258e-01 -4.25085463e-02 -6.78107664e-02
-1.39089274e+00 -2.98365504e-01 -4.65419084e-01 -1.03114098e-02
1.02604425e+00 -7.85830319e-01 -3.17771494e-01 -7.89550394e-02
-8.98874521e-01 -9.98147950e-02 -6.22468412e-01 1.03223598e+00
-4.09086466e-01 3.10898066e-01 -9.54083562e-01 -3.44707161e-01
-5.17417192e-01 -1.71319103e+00 1.31340313e+00 9.40497965e-02
-4.73630816e-01 -7.75790155e-01 1.22343771e-01 2.68283397e-01
2.51298863e-02 3.16210568e-01 9.68935966e-01 -1.21769989e+00
-3.12862992e-01 -3.62628102e-01 -3.17162603e-01 3.60660046e-01
5.71120046e-02 1.41243368e-01 -1.04747903e+00 2.19868254e-02
2.56751895e-01 -7.80956030e-01 1.10316050e+00 8.58886719e-01
1.82522225e+00 3.97048831e-01 -3.05463463e-01 6.10654950e-01
1.05810034e+00 1.14623874e-01 1.34461612e-01 1.74438000e-01
3.92750829e-01 -1.44256642e-02 1.50885448e-01 5.62129259e-01
3.25147450e-01 -1.26343846e-01 5.25671363e-01 2.33227722e-02
-9.69442632e-03 6.29458576e-02 -4.83474433e-01 4.69224215e-01
1.89409897e-01 -4.05442029e-01 -1.61738861e+00 3.12575638e-01
-1.48193121e+00 -1.24143034e-01 -6.06414601e-02 1.78543711e+00
8.51446331e-01 4.80606884e-01 -3.50283325e-01 -3.38365465e-01
7.60232866e-01 -3.18723656e-02 -7.69622386e-01 6.38094023e-02
1.90161347e-01 4.31205630e-01 8.31009507e-01 1.26686990e-01
-1.38605666e+00 1.37096941e-02 6.80843449e+00 4.49486405e-01
-8.66581142e-01 2.13171080e-01 1.34465063e+00 -2.79489368e-01
-1.57712623e-01 -5.40193200e-01 -3.50706905e-01 4.39429998e-01
1.19924963e+00 -8.56161863e-03 -4.05723631e-01 1.06811094e+00
-8.57595727e-02 -3.14994991e-01 -1.54346859e+00 8.49863887e-01
-4.30580638e-02 -1.49574280e+00 -3.01434517e-01 -2.11681187e-01
8.79501104e-01 4.87280190e-01 4.08349752e-01 2.81223148e-01
5.87352216e-01 -1.24301684e+00 3.10063034e-01 3.71697128e-01
1.07397473e+00 -3.56470257e-01 1.34941375e+00 5.65439165e-02
-4.45535183e-01 1.98557913e-01 -1.28900304e-01 6.57320023e-01
1.31550223e-01 5.31362414e-01 -1.41137946e+00 4.05967116e-01
9.29507375e-01 3.09030056e-01 -8.20255399e-01 1.09252489e+00
-4.14058596e-01 9.50389683e-01 -5.34037709e-01 4.55526710e-01
3.70275408e-01 7.26185977e-01 1.34136185e-01 1.17049599e+00
1.21443614e-01 4.58447903e-01 1.53055624e-03 9.27669346e-01
-6.00230277e-01 -1.09763600e-01 -3.17485571e-01 3.46625686e-01
3.74400765e-01 1.17276073e+00 -9.59374666e-01 -6.43088758e-01
-2.59687543e-01 3.46773982e-01 9.74006802e-02 -2.27599079e-03
-1.28048456e+00 1.60074934e-01 -4.63812947e-01 -1.42009007e-02
9.86836553e-02 2.87963420e-01 -4.96826708e-01 -9.24002409e-01
-5.50114177e-02 -8.09750199e-01 1.03771591e+00 -1.05006611e+00
-1.68408680e+00 8.31904948e-01 1.95904568e-01 -9.99178886e-01
-5.40888011e-01 -8.21942210e-01 -5.14334738e-01 6.10861838e-01
-1.14089799e+00 -5.33129275e-01 -4.39213723e-01 2.67782062e-01
5.38045943e-01 9.24238563e-03 7.28169620e-01 1.43323243e-01
-3.78342032e-01 6.25304401e-01 1.40010387e-01 5.28603852e-01
1.13279116e+00 -1.56125081e+00 3.43858361e-01 4.54597712e-01
5.52060455e-02 2.95934290e-01 2.70832360e-01 -6.20469749e-01
-5.94148457e-01 -1.09687245e+00 1.96206555e-01 -1.05754030e+00
6.13979578e-01 3.96933854e-01 -1.08764195e+00 1.12415206e+00
-9.73105431e-04 5.12442231e-01 1.00750279e+00 -3.54901642e-01
-2.98993617e-01 4.77094889e-01 -1.23915040e+00 1.29593998e-01
5.47278225e-01 -2.83127427e-01 -1.05661559e+00 6.33280098e-01
6.31097138e-01 -1.13613057e+00 -1.26221514e+00 7.30439425e-01
5.77484727e-01 -8.07647228e-01 7.88344383e-01 -7.53879249e-01
8.59330297e-01 1.05770759e-01 -7.09990785e-02 -1.13944507e+00
-1.09146476e-01 2.18227401e-01 4.56169620e-02 3.18471253e-01
6.78071856e-01 -5.41258335e-01 6.70111597e-01 9.56134200e-01
-1.09222971e-01 -1.09143698e+00 -8.17225873e-01 -2.84804255e-01
2.56993979e-01 -7.48493373e-01 2.64028490e-01 6.39564097e-01
-3.38115066e-01 -7.18699023e-02 1.27108738e-01 3.05371463e-01
6.61256075e-01 -7.48507902e-02 4.33421917e-02 -8.13407481e-01
-6.88504279e-01 -2.79172450e-01 -8.39603022e-02 -3.78626853e-01
-1.37458056e-01 -1.07153487e+00 5.86637631e-02 -1.60022116e+00
6.04227781e-01 -4.61815119e-01 -8.01921427e-01 4.71425325e-01
-3.98783088e-01 2.16704488e-01 -3.90828289e-02 4.37180310e-01
-7.29361236e-01 3.28780822e-02 1.26938725e+00 -2.65359968e-01
3.07235241e-01 3.66413295e-01 -6.36410594e-01 1.09674227e+00
7.74470508e-01 -8.79887879e-01 -3.20053399e-01 -5.64249456e-01
-1.65518816e-03 4.13745016e-01 4.26816612e-01 -1.09450221e+00
2.55644973e-02 2.81059742e-01 9.92402673e-01 -9.30305660e-01
8.07986502e-03 -7.75763273e-01 8.88443086e-03 6.95763290e-01
-9.11532164e-01 2.33897984e-01 2.24612042e-01 7.64281034e-01
-5.83468527e-02 -4.31082755e-01 8.95262122e-01 -6.63356483e-01
-2.24769935e-01 2.91908115e-01 -3.56666982e-01 5.71975350e-01
1.00534534e+00 4.77779061e-01 -3.41141284e-01 -3.77992988e-01
-1.27614236e+00 4.86246616e-01 -1.37636522e-02 1.45970076e-01
6.56461000e-01 -9.44051325e-01 -1.05211747e+00 -2.15159595e-01
1.89856201e-01 4.25148368e-01 4.37156975e-01 8.76523912e-01
-8.06186199e-01 4.98660237e-01 1.63491189e-01 -1.28881788e+00
-9.11466241e-01 3.27636451e-01 7.61817813e-01 -6.30553365e-01
-6.25906587e-01 1.06630290e+00 3.87851030e-01 -3.44591349e-01
2.61895359e-01 -8.37316453e-01 2.71759331e-01 -3.75246346e-01
4.75662798e-01 2.44852528e-01 3.42574239e-01 -1.25667034e-02
-5.77990532e-01 9.70483199e-02 -5.02234757e-01 -5.53148016e-02
1.00802433e+00 2.02326491e-01 3.68106157e-01 5.03028393e-01
9.59254801e-01 -1.88311026e-01 -1.03201485e+00 -3.23014677e-01
9.13539082e-02 -1.14404634e-01 8.93819612e-03 -1.08907628e+00
-1.05005944e+00 7.30580151e-01 8.77329469e-01 -1.44687712e-01
6.32858038e-01 4.29185271e-01 4.86153066e-01 4.95093077e-01
4.51124758e-02 -6.30849361e-01 2.14309335e-01 6.56423643e-02
7.68120944e-01 -1.86897624e+00 1.48047671e-01 -7.54913092e-02
-1.10998940e+00 1.14431930e+00 6.79530323e-01 -1.56653374e-01
9.13212478e-01 4.53209043e-01 3.32313120e-01 -5.05035758e-01
-9.46029305e-01 3.43737155e-01 4.10735011e-01 2.10378006e-01
7.21735895e-01 3.74082476e-01 -1.69107363e-01 9.59577918e-01
-8.32131878e-02 2.20563933e-01 7.60188699e-01 9.97780204e-01
-2.20366597e-01 -5.98717511e-01 -6.20596945e-01 1.13035882e+00
-1.04131556e+00 -6.50075376e-02 -9.25180614e-02 9.95769322e-01
1.21897431e-02 5.20877123e-01 -6.50132000e-02 -8.74627307e-02
2.78312206e-01 1.80086493e-01 2.87123054e-01 -8.72881770e-01
-5.45515239e-01 5.32148406e-03 7.41884112e-02 -2.83761472e-01
6.22479096e-02 -5.37089169e-01 -1.63835633e+00 4.39728975e-01
-1.98737115e-01 4.67889085e-02 4.03804988e-01 7.01144576e-01
-5.88772222e-02 1.08749938e+00 3.10259372e-01 -1.82853013e-01
-1.14311397e+00 -9.03387785e-01 -4.49139804e-01 4.17372167e-01
3.99652809e-01 -6.43580973e-01 -4.00249869e-01 1.42991096e-01] | [15.190557479858398, -1.9663363695144653] |
ea98db92-a034-4e7a-9b48-99f33931e8c0 | tag-enhanced-tree-structured-neural-networks | 1803.01165 | null | http://arxiv.org/abs/1803.01165v1 | http://arxiv.org/pdf/1803.01165v1.pdf | Tag-Enhanced Tree-Structured Neural Networks for Implicit Discourse Relation Classification | Identifying implicit discourse relations between text spans is a challenging
task because it requires understanding the meaning of the text. To tackle this
task, recent studies have tried several deep learning methods but few of them
exploited the syntactic information. In this work, we explore the idea of
incorporating syntactic parse tree into neural networks. Specifically, we
employ the Tree-LSTM model and Tree-GRU model, which are based on the tree
structure, to encode the arguments in a relation. Moreover, we further leverage
the constituent tags to control the semantic composition process in these
tree-structured neural networks. Experimental results show that our method
achieves state-of-the-art performance on PDTB corpus. | ['Houfeng Wang', 'Xu sun', 'Jingfeng Yang', 'Yizhong Wang', 'Sujian Li'] | 2018-03-03 | tag-enhanced-tree-structured-neural-networks-2 | https://aclanthology.org/I17-1050 | https://aclanthology.org/I17-1050.pdf | ijcnlp-2017-11 | ['implicit-discourse-relation-classification'] | ['natural-language-processing'] | [ 2.48369783e-01 6.16661906e-01 -3.83502752e-01 -5.10719299e-01
-2.56674826e-01 -2.73600906e-01 5.84181905e-01 1.00927033e-01
-3.46628189e-01 6.75876796e-01 8.02972496e-01 -5.40899396e-01
1.03715777e-01 -9.38656569e-01 -5.45927525e-01 -4.84103352e-01
4.52590846e-02 1.09130479e-01 2.61292279e-01 -2.88151711e-01
1.69148356e-01 -2.50820309e-01 -9.76148367e-01 8.26373219e-01
8.17580044e-01 1.12758863e+00 3.18036199e-01 2.37998039e-01
-7.72576332e-01 1.72882831e+00 -6.04596019e-01 -4.46674049e-01
-3.84330541e-01 -3.75239700e-01 -1.43207085e+00 -3.95088851e-01
-7.15411231e-02 -2.90044516e-01 -5.12302160e-01 1.12238526e+00
1.22332804e-01 1.92970946e-01 4.59114522e-01 -7.31242716e-01
-6.63557053e-01 1.34862745e+00 -3.75297129e-01 3.81220579e-01
1.71430543e-01 -3.48918378e-01 1.46751893e+00 -4.08014506e-01
4.17598218e-01 1.63453269e+00 5.12625873e-01 4.63896960e-01
-8.37368548e-01 -4.83434737e-01 5.95297575e-01 6.46349609e-01
-7.08691776e-01 -4.03602988e-01 9.83412266e-01 -3.84210795e-01
1.29584384e+00 -8.20964295e-03 3.99204791e-01 1.12711883e+00
-1.45749301e-02 1.10766006e+00 9.80931103e-01 -7.14779496e-01
-1.34929210e-01 -5.84821165e-01 8.25466931e-01 7.81475186e-01
-9.04771313e-02 -2.27664322e-01 -3.97821009e-01 1.45567566e-01
4.45304990e-01 -3.64744872e-01 -3.54660213e-01 1.11948840e-01
-7.75125206e-01 1.07257414e+00 5.95251024e-01 6.21139526e-01
-1.94855973e-01 2.36397624e-01 8.16726863e-01 7.79287815e-02
6.33421600e-01 3.41404706e-01 -6.10151052e-01 -9.77969021e-02
-2.09564626e-01 -3.02754492e-02 1.06609643e+00 6.14759803e-01
5.08371949e-01 -1.21038981e-01 -4.23382998e-01 1.02237964e+00
4.30159509e-01 -6.81399330e-02 6.95997238e-01 -9.45922971e-01
8.72648001e-01 9.46625888e-01 -4.31858361e-01 -1.05911064e+00
-4.53524053e-01 9.55032483e-02 -8.44745398e-01 -3.88467699e-01
2.97943115e-01 -3.40406090e-01 -7.56258011e-01 1.76166952e+00
2.83529133e-01 6.67826682e-02 3.26433390e-01 7.09982634e-01
1.31745875e+00 6.41636431e-01 4.52085078e-01 -1.73484579e-01
1.69510472e+00 -1.24411690e+00 -1.19399917e+00 -2.97287822e-01
9.17825043e-01 -3.97367656e-01 9.27172720e-01 -1.65688261e-01
-1.03230798e+00 -3.16776276e-01 -9.98024821e-01 -5.34966469e-01
-3.06376666e-01 8.42054635e-02 6.85359120e-01 1.42412186e-01
-5.48714459e-01 5.89411736e-01 -8.83250713e-01 -2.27335826e-01
5.23851037e-01 1.37291417e-01 -6.07501529e-02 1.97559640e-01
-1.70627391e+00 9.11809266e-01 1.13472855e+00 3.78380090e-01
-3.87714684e-01 -1.03692479e-01 -1.23745692e+00 4.16526973e-01
6.13213837e-01 -6.11675799e-01 1.58306658e+00 -8.81318867e-01
-1.78286874e+00 8.82539988e-01 -3.37319762e-01 -7.21106291e-01
-1.58788368e-01 -5.50187826e-01 1.40121318e-02 1.34441126e-02
-1.20784469e-01 3.36494148e-01 3.33733022e-01 -9.40026402e-01
-7.05136299e-01 -4.18739319e-01 4.78736818e-01 3.07192266e-01
-4.52968448e-01 2.94622838e-01 -9.26923826e-02 -5.67566633e-01
1.66851714e-01 -5.19198239e-01 -8.04133713e-02 -6.55375600e-01
-6.13093495e-01 -1.03178310e+00 9.58883941e-01 -9.34615731e-01
1.54373705e+00 -1.96639550e+00 2.60770977e-01 -3.94091368e-01
2.46782869e-01 5.25064647e-01 -2.61480678e-02 4.13590908e-01
1.26628861e-01 2.65780330e-01 -1.19543746e-01 -1.58301026e-01
1.31427661e-01 6.24976814e-01 -6.12106323e-01 2.83899065e-02
3.21729183e-01 1.05524969e+00 -8.05231810e-01 -7.41346061e-01
-1.14571519e-01 1.59135342e-01 -2.12581441e-01 5.79955280e-01
-5.69533169e-01 1.62020043e-01 -8.38588834e-01 1.51125357e-01
2.57830173e-01 -1.66289389e-01 7.62994230e-01 -3.38985473e-01
1.04536843e-02 1.43122876e+00 -4.27022159e-01 1.66602325e+00
-3.89799535e-01 6.31595969e-01 6.19368441e-02 -1.50147521e+00
8.80751848e-01 5.50933123e-01 2.91153155e-02 -6.20924115e-01
5.44976175e-01 3.25265676e-02 4.70919758e-01 -8.84388268e-01
2.88157046e-01 -2.91861385e-01 -1.18166566e-01 4.10874248e-01
5.97074977e-04 3.50539446e-01 1.94538206e-01 -1.34420767e-01
1.13527346e+00 4.21124488e-01 5.37314177e-01 -1.22098267e-01
7.49101460e-01 -2.12260067e-01 7.87891686e-01 3.95424217e-01
-1.40334755e-01 1.30696639e-01 1.07752454e+00 -6.38518393e-01
-5.01444101e-01 -6.00682080e-01 1.71813611e-02 1.23897862e+00
-3.76388170e-02 -5.42192698e-01 -9.71501946e-01 -1.07539880e+00
-3.56287271e-01 8.01880002e-01 -6.81226611e-01 -5.38246334e-02
-1.36971796e+00 -5.53680837e-01 6.69016123e-01 7.57181048e-01
9.51160312e-01 -1.42482567e+00 -6.29677534e-01 3.57878298e-01
-6.23956323e-01 -1.54546201e+00 -9.50596407e-02 4.02566612e-01
-8.47126305e-01 -1.34423137e+00 -1.43279776e-01 -1.20444238e+00
2.21358612e-01 -7.27514699e-02 1.14349306e+00 2.04112276e-01
3.33463520e-01 -2.84044474e-01 -6.71820700e-01 -2.20598325e-01
-2.47025654e-01 6.37565076e-01 -4.96079713e-01 -1.34445354e-01
5.67498982e-01 -6.21542335e-01 -3.10933530e-01 -9.09347534e-02
-5.00341952e-01 2.78882116e-01 3.05448949e-01 9.52399492e-01
2.40493581e-01 -3.16727608e-02 5.69074988e-01 -1.20523179e+00
8.55184674e-01 -4.84246790e-01 -1.30557954e-01 3.40517759e-01
-7.48122334e-02 3.78169328e-01 5.49355805e-01 -2.77304947e-01
-1.53697109e+00 -3.12404841e-01 -3.05836260e-01 -3.95890512e-03
-1.22972779e-01 8.89704525e-01 -4.55984950e-01 5.83415687e-01
1.33420259e-01 -1.15149014e-01 -2.99668312e-01 -7.10389853e-01
5.09386301e-01 8.66777837e-01 3.90195787e-01 -1.00636327e+00
1.80694118e-01 1.47985101e-01 -2.02362597e-01 -5.30681610e-01
-1.76194775e+00 -1.76239371e-01 -7.16910779e-01 2.47774139e-01
1.33160281e+00 -6.23831987e-01 -8.26013803e-01 4.50515687e-01
-1.79287899e+00 -5.91823578e-01 2.95049399e-02 2.05365360e-01
-4.47993428e-01 4.76898462e-01 -1.00815797e+00 -7.30584741e-01
-5.76939940e-01 -9.19217110e-01 7.69665480e-01 2.69149154e-01
-3.97694767e-01 -1.29630280e+00 -1.85098890e-02 3.76601100e-01
1.66489944e-01 2.47261554e-01 1.43308783e+00 -9.70051765e-01
-3.00055504e-01 3.39060098e-01 -5.46244621e-01 2.59599298e-01
3.93721849e-01 -4.77276981e-01 -1.07262516e+00 1.71108350e-01
2.34333679e-01 -5.12698531e-01 1.08887208e+00 2.89033890e-01
1.33672857e+00 -6.90286398e-01 -4.47653443e-01 5.15261114e-01
7.72228777e-01 2.06133649e-01 6.51031673e-01 5.53516924e-01
9.17575657e-01 1.02256167e+00 4.54311699e-01 4.60960418e-02
7.94171393e-01 5.26179731e-01 2.78074503e-01 1.84976265e-01
-2.55800486e-01 -2.51536131e-01 3.02486181e-01 1.14090621e+00
-5.70712350e-02 -1.86494038e-01 -9.82677460e-01 3.71954083e-01
-2.28130245e+00 -7.14199066e-01 -3.21688473e-01 1.28724992e+00
1.26207817e+00 3.30317974e-01 -3.70663971e-01 -3.01172715e-02
7.49809682e-01 6.58232331e-01 -1.70275614e-01 -5.44794559e-01
-1.38633372e-02 1.95162028e-01 -3.30778174e-02 3.95389140e-01
-1.38609219e+00 1.39616036e+00 5.89548206e+00 7.19035923e-01
-9.83789682e-01 2.03786805e-01 6.76943541e-01 2.75724411e-01
-2.69470140e-02 6.85381070e-02 -8.04553092e-01 2.93473631e-01
9.12033319e-01 -9.78798792e-03 2.73914598e-02 6.40836596e-01
7.94388074e-03 -7.99931437e-02 -1.07025218e+00 4.30089682e-01
-1.11646809e-01 -1.19136012e+00 9.57420934e-03 -1.87019765e-01
3.12199384e-01 -3.77012864e-02 -4.61094677e-01 6.29482090e-01
6.87642932e-01 -1.03821468e+00 4.35486466e-01 2.33741984e-01
4.13985431e-01 -4.73420739e-01 8.82474780e-01 3.72595519e-01
-1.30052543e+00 -8.08862969e-02 -4.14006561e-01 -6.51584804e-01
2.79326826e-01 4.12160039e-01 -5.88737667e-01 5.85408270e-01
7.35997558e-01 7.93194830e-01 -1.48419112e-01 4.64136660e-01
-1.10271466e+00 1.04641461e+00 -1.11257635e-01 -1.47052154e-01
4.74662840e-01 4.10569506e-03 3.16067010e-01 1.25105500e+00
-4.92589101e-02 4.08440053e-01 3.70628297e-01 8.12831938e-01
-3.21585536e-01 -2.76242457e-02 -4.29217935e-01 -1.01762936e-01
4.33891356e-01 1.16908145e+00 -4.37774718e-01 -3.17723691e-01
-3.36300939e-01 5.63726068e-01 1.06789362e+00 3.00403446e-01
-6.22908294e-01 -4.02154028e-01 5.79691410e-01 -2.17380852e-01
4.83819544e-01 -3.57024819e-01 -3.96884531e-01 -1.03428948e+00
9.04659331e-02 -6.09342158e-01 6.50126100e-01 -6.92994237e-01
-1.34115756e+00 7.11011648e-01 7.84252509e-02 -4.82280880e-01
-2.04805389e-01 -7.49043167e-01 -9.64921832e-01 8.20480287e-01
-1.79577708e+00 -1.40536630e+00 -9.08636972e-02 1.90094933e-01
7.12616503e-01 -3.85037363e-02 9.49834883e-01 -1.00931317e-01
-7.67595291e-01 2.96292484e-01 -3.52651596e-01 7.70572126e-01
1.99958429e-01 -1.14073098e+00 5.02117276e-01 5.74094772e-01
6.13081083e-02 6.97805941e-01 1.96918160e-01 -5.71107209e-01
-9.04103994e-01 -9.14506197e-01 1.19057822e+00 -1.63914591e-01
9.88181114e-01 -2.22108006e-01 -1.34426212e+00 9.68395829e-01
8.87787521e-01 -2.05462307e-01 7.68928826e-01 6.58245385e-01
-4.23708975e-01 1.46071047e-01 -4.61081982e-01 4.54201728e-01
1.12532949e+00 -5.10636926e-01 -1.47267437e+00 -6.22267760e-02
1.27245831e+00 -5.93308389e-01 -7.69734263e-01 4.99821186e-01
1.79267794e-01 -6.95482492e-01 6.14068210e-01 -9.93638098e-01
8.94919455e-01 1.33791091e-02 -1.34972066e-01 -1.21434700e+00
-2.34694541e-01 -3.06929231e-01 -5.63641429e-01 1.67754614e+00
2.75203764e-01 -5.05941749e-01 5.22742033e-01 4.97478843e-01
-3.70808274e-01 -9.98775065e-01 -1.02379012e+00 -3.99755329e-01
3.17119271e-01 -2.00370282e-01 5.75117886e-01 1.09127486e+00
4.32791799e-01 1.07513797e+00 -9.92226321e-03 -2.23508000e-01
1.24727041e-01 2.22579524e-01 3.08245808e-01 -1.22360241e+00
-1.44059658e-01 -5.99004507e-01 5.87371290e-02 -1.29284513e+00
9.56299245e-01 -9.74456012e-01 9.78953019e-02 -1.88655305e+00
1.16861984e-01 -3.59969705e-01 -3.08669060e-01 8.69707406e-01
-5.74612916e-01 -4.63059872e-01 8.97531882e-02 1.56037882e-01
-6.81554735e-01 9.68428671e-01 1.14046669e+00 -2.93463647e-01
-5.99634349e-02 -3.06482583e-01 -5.54887474e-01 1.07105052e+00
1.04604208e+00 -4.47792619e-01 -2.20894024e-01 -8.58674467e-01
1.29778817e-01 -3.65893990e-02 1.26917526e-01 -5.70938528e-01
1.63047194e-01 -2.29424298e-01 -5.08116558e-02 -6.13553345e-01
2.36794740e-01 -5.67547739e-01 -7.90740490e-01 3.03435445e-01
-6.46510243e-01 -1.68164685e-01 1.31329611e-01 3.07349652e-01
-5.53807974e-01 -5.24047136e-01 3.57679665e-01 -3.14563543e-01
-6.58507526e-01 -1.18648194e-01 -5.26143432e-01 3.72430384e-01
5.71751356e-01 4.08338964e-01 -5.83095610e-01 -1.31330481e-02
-4.84433264e-01 5.16061485e-01 -2.64083713e-01 5.09517670e-01
4.15287226e-01 -1.28631103e+00 -5.42445123e-01 -2.30176762e-01
-2.11168036e-01 6.31159663e-01 -9.84655619e-02 6.23858392e-01
-2.07972184e-01 4.86502886e-01 1.54532194e-01 -3.15144330e-01
-1.32909107e+00 4.20170635e-01 4.76494223e-01 -7.93390930e-01
-7.21722305e-01 8.11439097e-01 5.61125696e-01 -4.14086461e-01
3.58826965e-01 -4.90345985e-01 -8.13283443e-01 2.24455401e-01
3.76521051e-01 -9.90965590e-02 -3.05742800e-01 -5.00074506e-01
-1.50202125e-01 3.48775327e-01 -2.88945615e-01 1.93308443e-01
1.37447166e+00 -1.09272733e-01 -6.49933279e-01 4.88838017e-01
1.07968390e+00 -4.83041912e-01 -1.01549113e+00 -5.09310305e-01
6.92596018e-01 1.34855151e-01 1.08931169e-01 -5.86220980e-01
-9.36828256e-01 1.23486853e+00 -2.30873093e-01 4.83347028e-01
9.22534525e-01 1.23241983e-01 1.16168845e+00 7.06894040e-01
-4.38804626e-02 -1.04526246e+00 1.66092254e-02 1.36897886e+00
7.19123662e-01 -9.96445477e-01 -2.19384938e-01 -6.87703550e-01
-3.57949495e-01 1.36072409e+00 1.01979542e+00 -4.70591784e-02
5.28352976e-01 2.32483149e-01 1.82820961e-01 -3.54089528e-01
-8.98222208e-01 -3.44916821e-01 1.76627725e-01 3.81243229e-01
9.66143548e-01 -8.76351073e-02 -6.28333747e-01 1.05187345e+00
-2.48174518e-01 -8.16553459e-02 1.23092167e-01 8.91132474e-01
-5.73525965e-01 -1.35500276e+00 -5.31531423e-02 1.63857803e-01
-5.85119188e-01 -4.04876500e-01 -5.64891934e-01 6.24466479e-01
-1.91496909e-02 8.96700799e-01 1.42287433e-01 -2.71797508e-01
3.71561438e-01 3.91861737e-01 4.47512627e-01 -7.69391358e-01
-7.77156293e-01 -7.26346001e-02 8.10105383e-01 -3.21312726e-01
-8.08844805e-01 -4.35013622e-01 -1.69429302e+00 7.26106577e-03
-3.17093819e-01 2.47922853e-01 2.23954469e-01 1.39736891e+00
1.53140485e-01 9.68798757e-01 2.98484862e-01 -5.14104784e-01
-4.85131413e-01 -1.45213711e+00 -1.56846657e-01 3.99050891e-01
9.32471305e-02 -6.40063107e-01 1.20823517e-01 9.12597179e-02] | [10.56445026397705, 9.240799903869629] |
9d25f77e-a6d4-4def-81bb-ba736d53eda2 | de-confounded-variational-encoder-decoder-for | null | null | https://aclanthology.org/2021.acl-long.430 | https://aclanthology.org/2021.acl-long.430.pdf | De-Confounded Variational Encoder-Decoder for Logical Table-to-Text Generation | Logical table-to-text generation aims to automatically generate fluent and logically faithful text from tables. The task remains challenging where deep learning models often generated linguistically fluent but logically inconsistent text. The underlying reason may be that deep learning models often capture surface-level spurious correlations rather than the causal relationships between the table $\boldsymbol{x}$ and the sentence $\boldsymbol{y}$. Specifically, in the training stage, a model can get a low empirical loss without understanding $\boldsymbol{x}$ and use spurious statistical cues instead. In this paper, we propose a de-confounded variational encoder-decoder (DCVED) based on causal intervention, learning the objective $p(\boldsymbol{y}|\textrm{do}(\boldsymbol{x}))$. Firstly, we propose to use variational inference to estimate the confounders in the latent space and cooperate with the causal intervention based on Pearl{'}s do-calculus to alleviate the spurious correlations. Secondly, to make the latent confounder meaningful, we propose a back-prediction process to predict the not-used entities but linguistically similar to the exactly selected ones. Finally, since our variational model can generate multiple candidates, we train a table-text selector to find out the best candidate sentence for the given table. An extensive set of experiments show that our model outperforms the baselines and achieves new state-of-the-art performance on two logical table-to-text datasets in terms of logical fidelity. | ['Yaohui Jin', 'Hao He', 'Yitian Li', 'Jidong Tian', 'Wenqing Chen'] | 2021-08-01 | null | null | null | acl-2021-5 | ['table-to-text-generation'] | ['natural-language-processing'] | [ 1.86372548e-01 6.40067935e-01 -3.69193256e-01 -5.33584952e-01
-1.06021845e+00 -3.70505631e-01 7.55444109e-01 6.01829477e-02
-3.88287492e-02 1.27240181e+00 5.68192303e-01 -4.81830776e-01
-2.91467495e-02 -1.15745342e+00 -1.30150437e+00 -2.35445783e-01
2.16483116e-01 7.75002897e-01 -2.62123525e-01 -2.27731854e-01
8.14695805e-02 -3.89432348e-02 -1.44301355e+00 6.81613803e-01
1.21200001e+00 5.81719100e-01 1.73287526e-01 2.13826984e-01
-4.50988680e-01 1.31107974e+00 -7.36562967e-01 -1.10386443e+00
4.31753024e-02 -9.06571090e-01 -7.88568139e-01 -2.99428701e-01
3.95050049e-01 -3.93167764e-01 -2.00217441e-01 1.07878089e+00
1.59791872e-01 5.08149108e-03 1.02109027e+00 -1.05893254e+00
-9.78625298e-01 1.59107053e+00 -5.14168859e-01 3.85817625e-02
4.04375076e-01 1.94972798e-01 1.52872789e+00 -1.15641713e+00
8.18904877e-01 1.69247174e+00 3.97742271e-01 6.58734441e-01
-1.40042806e+00 -8.14381123e-01 3.45754564e-01 5.22681475e-02
-1.43040979e+00 -4.79016453e-01 6.85537338e-01 -3.54755640e-01
8.22782874e-01 1.87534854e-01 2.63702184e-01 1.48267519e+00
3.03751677e-01 8.22417140e-01 7.13331103e-01 -4.88875628e-01
6.83576912e-02 1.95173666e-01 -1.31140068e-01 1.03658521e+00
4.13136303e-01 -6.48339465e-02 -9.34958458e-01 3.69895436e-02
5.65579414e-01 -3.51867735e-01 -3.52298878e-02 2.95727938e-01
-1.32643819e+00 9.84760284e-01 1.33643523e-01 2.18789488e-01
-4.15640980e-01 6.07412517e-01 -2.89709717e-02 -1.97173674e-02
1.39630198e-01 5.12494862e-01 -5.93748450e-01 7.98399001e-03
-1.00685489e+00 6.51277483e-01 7.04420030e-01 1.12832427e+00
7.18909025e-01 4.63414490e-02 -5.95225215e-01 5.53515673e-01
6.34128630e-01 5.69457412e-01 -1.52404025e-01 -9.70263362e-01
9.47407067e-01 5.10888457e-01 1.92549706e-01 -8.09998691e-01
-6.50497675e-02 4.04209271e-02 -9.86056983e-01 -1.85640723e-01
5.19972205e-01 -3.58860910e-01 -7.62639880e-01 2.04744673e+00
-6.56408295e-02 -1.62945062e-01 1.19425140e-01 6.54619277e-01
8.70624065e-01 8.39541256e-01 3.50418746e-01 -3.80035728e-01
1.27908742e+00 -4.91616338e-01 -1.11535859e+00 -3.60421360e-01
5.87796092e-01 -5.42282283e-01 1.61308229e+00 1.84525818e-01
-1.27553439e+00 -3.49203736e-01 -9.80443358e-01 -5.35174251e-01
-2.65895694e-01 2.29935884e-01 6.69091582e-01 3.69470596e-01
-6.94575608e-01 7.12846756e-01 -6.16389871e-01 2.43962169e-01
4.36376095e-01 1.25587702e-01 -3.12833078e-02 9.67515036e-02
-1.82036173e+00 7.84061074e-01 4.67465311e-01 1.37746453e-01
-8.64776611e-01 -9.68770385e-01 -9.99895453e-01 1.81552798e-01
6.81863129e-01 -9.96378899e-01 1.03171217e+00 -5.98618746e-01
-1.45312953e+00 8.19809318e-01 -5.74792027e-01 -3.79192770e-01
7.28504658e-01 -2.06272170e-01 -2.30078489e-01 -3.64034742e-01
5.91224074e-01 6.41385198e-01 5.58092475e-01 -1.35538328e+00
-4.10064310e-01 -1.41626313e-01 9.26259626e-03 -6.89163199e-03
1.55270904e-01 -1.03942037e-01 -2.44212031e-01 -7.53109097e-01
7.08763003e-02 -4.95775163e-01 5.59295304e-02 -3.90158474e-01
-1.09332430e+00 -6.03319466e-01 4.07375693e-02 -5.68489671e-01
1.48177338e+00 -1.68580365e+00 3.06216061e-01 8.85650236e-03
2.36745000e-01 -2.79964983e-01 2.91197866e-01 4.34188843e-01
-2.33008564e-02 7.69492984e-01 -4.44962800e-01 -3.81059945e-01
4.18401241e-01 1.26600593e-01 -6.39711261e-01 -5.87040782e-02
4.98601437e-01 1.03827381e+00 -9.03537929e-01 -7.32099891e-01
-9.92491990e-02 1.82982609e-01 -7.40431190e-01 1.07181251e-01
-9.90799546e-01 1.86191723e-01 -3.18768859e-01 4.32162374e-01
4.37630415e-01 -3.21916282e-01 3.86960626e-01 -1.16677329e-01
-1.73255838e-02 9.29809093e-01 -1.19530118e+00 1.59363794e+00
-3.94179255e-01 3.74245167e-01 -3.81479263e-01 -5.67973733e-01
7.59934843e-01 1.50022864e-01 7.72479847e-02 -3.97859871e-01
7.97300190e-02 1.90970048e-01 4.51421505e-03 -4.10467923e-01
4.89931107e-01 -5.84571958e-01 -4.03195858e-01 4.64630067e-01
-3.49577777e-02 -3.17538023e-01 3.15993965e-01 5.06511807e-01
8.85414481e-01 3.79669040e-01 1.33272931e-01 -2.60784060e-01
5.48864245e-01 -6.10907190e-02 6.86386645e-01 9.60343361e-01
2.56294936e-01 3.50021064e-01 1.16886771e+00 -2.03134820e-01
-8.54995847e-01 -1.37092876e+00 -7.66760483e-02 7.75956810e-01
1.25575617e-01 -6.94722354e-01 -7.10217595e-01 -9.13437247e-01
4.62553017e-02 1.67092550e+00 -8.00705314e-01 -9.97507349e-02
-7.05651462e-01 -7.20883727e-01 7.61523247e-01 4.55237389e-01
3.13103110e-01 -1.38452303e+00 -1.72780946e-01 2.88484901e-01
-6.48253083e-01 -8.30583751e-01 -4.16154206e-01 2.12531894e-01
-5.33030331e-01 -8.14465523e-01 -2.54479237e-02 -4.41306114e-01
6.50558591e-01 -6.99427605e-01 1.44096947e+00 1.96365261e-04
2.11698622e-01 -4.75585848e-01 1.25053406e-01 -6.01549625e-01
-7.82606840e-01 -1.43460184e-01 -9.08872336e-02 -1.30837351e-01
4.96668518e-01 -2.87009090e-01 -1.76374152e-01 -2.06215173e-01
-7.11448729e-01 4.29921240e-01 4.02716786e-01 1.05887377e+00
8.53361428e-01 1.42322019e-01 5.64012408e-01 -1.24056745e+00
7.49297798e-01 -3.89033854e-01 -5.84290922e-01 3.83392483e-01
-6.78464353e-01 6.61585689e-01 9.56502080e-01 -8.88342261e-02
-1.33099425e+00 -3.31272662e-01 -1.01721168e-01 -2.23662615e-01
1.30866200e-01 8.50404859e-01 -7.15901077e-01 1.29217792e+00
7.60248303e-01 4.60687801e-02 -5.34868836e-01 -1.88448414e-01
7.45682120e-01 2.04905987e-01 3.65985334e-01 -8.39918256e-01
7.14332283e-01 2.95646548e-01 5.46256788e-02 -1.49313733e-01
-1.18450773e+00 5.97401202e-01 -5.18801272e-01 -8.90034717e-03
1.13150501e+00 -8.76200497e-01 -8.99382591e-01 -4.05570380e-02
-1.50829518e+00 -5.24743497e-01 -3.53889048e-01 2.35780865e-01
-4.16401953e-01 -5.28952107e-02 -5.35661876e-01 -9.26744461e-01
-1.60548598e-01 -1.01459098e+00 1.15643322e+00 -9.07085650e-03
-5.87512672e-01 -9.36176121e-01 -3.73266399e-01 2.77650684e-01
-2.26343840e-01 2.67119408e-01 1.59465241e+00 -3.56251895e-01
-1.04252648e+00 9.86853242e-02 -2.11337015e-01 3.46009880e-02
8.77164751e-02 2.92914271e-01 -7.21977532e-01 4.80395317e-01
-2.26826280e-01 -3.22763145e-01 9.30143535e-01 3.97654951e-01
1.08995986e+00 -7.49716640e-01 -2.92882383e-01 2.77687401e-01
1.20497751e+00 7.97418226e-03 8.12733829e-01 -1.77979857e-01
8.20536494e-01 6.51159883e-01 3.07525694e-01 4.45082814e-01
8.00293565e-01 4.01275247e-01 1.18642211e-01 3.31916511e-01
-1.44714028e-01 -1.06162441e+00 6.59509063e-01 4.52798754e-01
3.17904353e-01 -5.28579712e-01 -8.07556331e-01 5.17306030e-01
-1.65500832e+00 -1.25216842e+00 -4.36807156e-01 2.04507017e+00
1.42183781e+00 4.87200201e-01 -4.63551909e-01 -3.51931565e-02
5.31594634e-01 2.38774389e-01 -3.51389050e-01 -3.85588050e-01
-4.21928018e-01 3.60402733e-01 1.88506454e-01 9.62767899e-01
-7.05007195e-01 1.37072349e+00 5.06862497e+00 8.14451039e-01
-7.00143158e-01 6.62001818e-02 8.50435853e-01 -1.66878581e-01
-1.15933681e+00 1.83510810e-01 -1.13801587e+00 5.78200400e-01
9.13266778e-01 -3.72567832e-01 4.64441925e-01 4.26083565e-01
4.29636300e-01 -2.02515841e-01 -1.53327751e+00 5.65450549e-01
-1.18753966e-02 -1.64630890e+00 4.90659386e-01 -1.11316606e-01
7.98979044e-01 -6.08953476e-01 -1.74270831e-02 4.77045238e-01
7.84848809e-01 -1.31924975e+00 1.21115839e+00 6.94429517e-01
1.02355051e+00 -6.18414760e-01 3.84233326e-01 4.57928181e-01
-9.53421235e-01 1.99868530e-01 -1.75488904e-01 -1.06704831e-01
1.92688897e-01 8.36043477e-01 -7.70076454e-01 5.08130908e-01
4.67907727e-01 5.07392645e-01 -2.02515841e-01 -1.61079034e-01
-9.18381572e-01 5.46281934e-01 -1.02343485e-01 -4.46570456e-01
-2.27089431e-02 -1.09364711e-01 4.75559145e-01 1.01785195e+00
2.21417189e-01 2.96933353e-01 -1.31511003e-01 2.01807070e+00
-6.72442079e-01 8.35331306e-02 -6.15306973e-01 -1.83069274e-01
6.22423649e-01 5.95931232e-01 -3.91532362e-01 -4.65403646e-01
-6.36641979e-02 8.22140396e-01 4.91632074e-01 3.55222046e-01
-1.05487025e+00 -2.81303614e-01 4.65843499e-01 1.23084329e-01
1.43851489e-01 1.19989499e-01 -9.91485596e-01 -1.27861857e+00
2.13271230e-01 -7.56814361e-01 3.06560993e-01 -8.82697582e-01
-1.30870557e+00 4.03060049e-01 1.53817326e-01 -6.24257267e-01
-4.92613405e-01 -2.85467207e-01 -4.38712507e-01 1.39328229e+00
-1.16828060e+00 -9.68833447e-01 2.04233602e-01 4.13653553e-01
4.89138067e-01 -6.36545103e-03 7.81466961e-01 -1.37193431e-03
-5.94922125e-01 7.88118303e-01 -3.68448883e-01 2.88769692e-01
6.16513014e-01 -1.50662065e+00 3.66455197e-01 1.18652689e+00
1.34457141e-01 9.35833931e-01 7.72892535e-01 -1.18032134e+00
-1.33794570e+00 -1.15282035e+00 1.77989709e+00 -8.99433732e-01
5.73238552e-01 -6.81913614e-01 -7.81002820e-01 9.40430820e-01
3.35514605e-01 -5.12483656e-01 4.81100738e-01 2.12177113e-01
-4.45872933e-01 -4.15445343e-02 -9.93443966e-01 1.17679191e+00
1.11506879e+00 -5.92366397e-01 -7.44544506e-01 4.33569521e-01
9.66192067e-01 -5.07474601e-01 -5.65734029e-01 2.46539161e-01
2.29044199e-01 -8.21122050e-01 5.59329391e-01 -6.89835668e-01
1.29542661e+00 -3.30883741e-01 -1.25091910e-01 -1.13201475e+00
-1.47118509e-01 -9.04404819e-01 -2.24474102e-01 1.39226818e+00
1.14034617e+00 -1.63964227e-01 7.86915421e-01 8.66218209e-01
-8.08288306e-02 -6.13421261e-01 -8.99597824e-01 -3.16071361e-01
5.03685772e-01 -5.74051738e-01 7.36051500e-01 9.16959882e-01
-1.40013103e-03 7.07257569e-01 -4.55921888e-01 8.31255913e-02
6.47202969e-01 1.50269493e-01 5.87960899e-01 -1.09013343e+00
-3.15047085e-01 -4.97911513e-01 4.75120008e-01 -9.39621568e-01
5.34090161e-01 -1.20837903e+00 2.77568996e-01 -1.82083166e+00
1.31447852e-01 -4.47062880e-01 4.04236130e-02 6.09022498e-01
-7.27917492e-01 -5.10126472e-01 1.65280178e-01 -7.53189996e-02
-1.99908286e-01 6.92728639e-01 1.36601770e+00 -2.38878012e-01
-1.81747079e-01 -3.32329214e-01 -1.07589746e+00 5.73558331e-01
4.41013634e-01 -7.00343251e-01 -4.85993534e-01 -5.74685574e-01
7.07797229e-01 4.43443954e-01 3.42607856e-01 -3.04624736e-01
-5.22486633e-03 -5.16137302e-01 3.80846918e-01 -7.14373112e-01
7.40285814e-02 -3.60691398e-01 1.09459914e-01 2.74053067e-01
-8.63882244e-01 5.76245077e-02 8.18232745e-02 3.81258339e-01
-6.72234735e-03 -1.66569561e-01 4.88918602e-01 -2.77484655e-01
-5.09063713e-02 2.04272829e-02 -4.22975123e-01 5.49300134e-01
5.61010242e-01 3.04339349e-01 -3.16399425e-01 -3.27088684e-01
-3.59207302e-01 2.62310326e-01 1.02098107e-01 3.23524654e-01
5.91248512e-01 -1.25486684e+00 -1.00553930e+00 1.16024271e-01
-1.03068635e-01 3.36172879e-01 1.11165702e-01 4.68417078e-01
-3.31957161e-01 4.86237824e-01 3.24346930e-01 -1.14288665e-01
-5.09853899e-01 4.72077787e-01 3.86019737e-01 -5.84445298e-01
-3.09233457e-01 1.25559020e+00 2.76788175e-01 -3.14151585e-01
2.14382008e-01 -6.62244439e-01 2.07316071e-01 6.14685304e-02
2.82374650e-01 4.90169227e-02 -2.01569587e-01 -2.07692534e-01
-3.75908643e-01 -1.87943697e-01 -3.89036648e-02 -5.24861813e-01
9.85534072e-01 9.40873008e-03 -3.49749029e-01 7.06617951e-01
9.16393757e-01 3.72124046e-01 -1.13065493e+00 -5.46701327e-02
2.18935534e-01 -1.60224706e-01 -2.09633902e-01 -1.18796241e+00
-6.70761764e-01 9.83253002e-01 -1.80934787e-01 -9.81247891e-03
6.57552123e-01 8.95460397e-02 6.06171310e-01 2.82405674e-01
1.14538059e-01 -9.77073014e-01 4.75269333e-02 5.85807025e-01
1.16338217e+00 -1.15560305e+00 -8.06659460e-02 -5.77090144e-01
-9.04733479e-01 9.12844181e-01 7.24673390e-01 1.48985580e-01
5.54036021e-01 3.52358043e-01 -1.91407409e-02 -2.91943550e-01
-1.11801600e+00 5.36396876e-02 1.57517523e-01 1.75667793e-01
8.31200778e-01 2.08034351e-01 -5.08352280e-01 1.07192755e+00
-6.08576477e-01 -4.61736843e-02 7.21166730e-01 4.42233741e-01
7.24640023e-03 -1.25492656e+00 -1.86843991e-01 5.31727910e-01
-5.39426148e-01 -7.21032143e-01 -5.47539234e-01 7.04964995e-01
4.23583657e-01 9.93583679e-01 -9.07775834e-02 -2.60529608e-01
3.66598606e-01 4.57846314e-01 5.18595934e-01 -7.12583423e-01
-6.01265728e-01 9.76353139e-02 2.23042801e-01 -2.55638450e-01
-5.88587821e-02 -8.92583251e-01 -1.82625365e+00 -4.78453457e-01
4.37142849e-02 1.15398347e-01 2.62845665e-01 1.07187426e+00
2.60839641e-01 8.06357563e-01 3.17923397e-01 -1.26749277e-01
-6.08624458e-01 -9.20648575e-01 -5.26745677e-01 4.79741871e-01
9.56216604e-02 -7.28394330e-01 -2.76651770e-01 3.16554934e-01] | [11.457183837890625, 8.743942260742188] |
d525e643-1a79-44e6-b73f-799cf11c77c2 | manet-multimodal-attention-network-based | 2002.12573 | null | https://arxiv.org/abs/2002.12573v1 | https://arxiv.org/pdf/2002.12573v1.pdf | MANet: Multimodal Attention Network based Point- View fusion for 3D Shape Recognition | 3D shape recognition has attracted more and more attention as a task of 3D vision research. The proliferation of 3D data encourages various deep learning methods based on 3D data. Now there have been many deep learning models based on point-cloud data or multi-view data alone. However, in the era of big data, integrating data of two different modals to obtain a unified 3D shape descriptor is bound to improve the recognition accuracy. Therefore, this paper proposes a fusion network based on multimodal attention mechanism for 3D shape recognition. Considering the limitations of multi-view data, we introduce a soft attention scheme, which can use the global point-cloud features to filter the multi-view features, and then realize the effective fusion of the two features. More specifically, we obtain the enhanced multi-view features by mining the contribution of each multi-view image to the overall shape recognition, and then fuse the point-cloud features and the enhanced multi-view features to obtain a more discriminative 3D shape descriptor. We have performed relevant experiments on the ModelNet40 dataset, and experimental results verify the effectiveness of our method. | ['Yaxin Zhao', 'Tangkun Zhang', 'Jichao Jiao'] | 2020-02-28 | null | null | null | null | ['3d-shape-recognition'] | ['computer-vision'] | [-3.99959296e-01 -6.30506575e-01 6.62842169e-02 -4.95042831e-01
-5.72919607e-01 -3.67760837e-01 5.52925229e-01 -1.28678858e-01
-7.49805868e-02 -1.82800516e-01 2.98458904e-01 1.68756947e-01
-3.42597991e-01 -7.85347044e-01 -4.24366474e-01 -9.19028342e-01
4.81764466e-01 3.98653060e-01 1.38233125e-01 -8.20935890e-02
2.03090131e-01 9.15964186e-01 -1.44851840e+00 3.39169413e-01
5.40670156e-01 1.35490072e+00 2.07112968e-01 1.73359606e-02
-4.94931668e-01 5.19524887e-02 -1.40570551e-01 -2.41941214e-01
4.86691922e-01 1.56513721e-01 -1.61481738e-01 3.75797212e-01
4.08766240e-01 -5.05859852e-01 -4.87254024e-01 1.04151988e+00
9.50815260e-01 -5.03724776e-02 4.63193595e-01 -1.14754128e+00
-9.82558966e-01 -3.03466823e-02 -8.25696051e-01 1.53869763e-01
1.26478896e-01 2.49373708e-02 8.07078183e-01 -1.42570901e+00
3.12559962e-01 1.59674406e+00 3.57005358e-01 2.69030631e-01
-7.61725903e-01 -5.35640121e-01 2.38260493e-01 2.44180962e-01
-1.31667113e+00 -3.45592707e-01 1.31546640e+00 -2.84395993e-01
1.07348883e+00 4.80124466e-02 7.29185343e-01 7.25006521e-01
-4.71475609e-02 1.16351402e+00 8.26153517e-01 -1.95198730e-01
-3.21529835e-01 -7.65354410e-02 -9.94765833e-02 6.16463125e-01
2.46223077e-01 1.24462143e-01 -1.84526384e-01 -1.04423210e-01
8.77833545e-01 7.29360282e-01 -9.41450894e-02 -6.76313818e-01
-1.04689050e+00 6.23995483e-01 7.10734606e-01 4.85298693e-01
-5.47029614e-01 -2.30158672e-01 2.66833812e-01 8.15440789e-02
7.65502930e-01 -2.05577463e-01 -4.27300870e-01 -6.22512214e-03
-5.17920494e-01 -6.16120640e-04 2.57094115e-01 7.57978559e-01
7.19802082e-01 8.82051736e-02 1.62133574e-01 1.16010642e+00
7.35081077e-01 9.78771269e-01 2.75877923e-01 -4.53636289e-01
5.69526732e-01 1.43296444e+00 -2.90029049e-01 -1.31473017e+00
-2.99573809e-01 -4.24296468e-01 -9.59381282e-01 -3.65807414e-02
1.55888628e-02 1.71541363e-01 -9.01142061e-01 1.27037334e+00
3.91428620e-01 -9.50095206e-02 -1.68675572e-01 1.34580922e+00
1.19713986e+00 4.86759573e-01 -3.40437025e-01 1.18646078e-01
1.05165839e+00 -5.41809738e-01 -3.60600293e-01 9.39616337e-02
4.40816164e-01 -8.27798963e-01 6.16839230e-01 3.08936611e-02
-9.80408132e-01 -6.86724603e-01 -1.06261635e+00 -9.07974392e-02
-3.96113008e-01 1.99506536e-01 4.55774069e-01 4.17938828e-01
-5.92589140e-01 6.39462620e-02 -8.30744267e-01 -1.83350146e-01
7.74383485e-01 3.36808860e-01 -5.53434730e-01 -4.12894070e-01
-7.94689775e-01 6.77652240e-01 1.85153663e-01 2.51511365e-01
-4.77132887e-01 -3.53023559e-01 -7.65147984e-01 1.54171497e-01
2.50265151e-01 -6.52028561e-01 6.71356142e-01 -5.34717798e-01
-9.59652007e-01 9.08200026e-01 -1.35054842e-01 3.38243127e-01
7.50943124e-02 2.76491940e-02 -3.97140414e-01 1.57321990e-01
1.57175977e-02 5.20108044e-01 6.33377612e-01 -1.49055457e+00
-5.25509000e-01 -1.09285212e+00 6.37608580e-03 4.72362250e-01
-3.89270723e-01 -7.71886930e-02 -5.98658800e-01 -8.89380127e-02
7.99592733e-01 -7.41602063e-01 1.54662445e-01 -1.53862044e-01
-1.37630776e-01 -5.90492487e-01 1.11332893e+00 -3.24762583e-01
6.79314137e-01 -2.42386985e+00 2.93110877e-01 1.84875160e-01
3.47062886e-01 3.33208799e-01 -2.71432936e-01 2.13958174e-01
-3.31026576e-02 -5.70869744e-02 -1.32796844e-03 -2.66824126e-01
-3.04344837e-02 1.86643928e-01 -1.69122458e-01 3.67313385e-01
3.27426165e-01 1.10528266e+00 -6.02121592e-01 -4.05964166e-01
4.08682197e-01 5.73733032e-01 -3.87740523e-01 1.55176297e-01
2.12423220e-01 2.41786808e-01 -9.08441901e-01 9.84018147e-01
9.45330143e-01 -3.04558396e-01 -4.70338225e-01 -5.59511423e-01
-7.20542371e-02 -2.48531014e-01 -1.10278010e+00 1.66049647e+00
-3.64128590e-01 5.36041744e-02 4.63538505e-02 -1.10299611e+00
1.49402952e+00 1.61478341e-01 9.32180047e-01 -7.28031874e-01
3.28529030e-01 4.12345916e-01 -1.03370212e-02 -5.96223116e-01
3.91306169e-02 -1.95730448e-01 1.56434119e-01 3.45498562e-01
1.22475855e-01 -1.83991715e-01 -3.42599690e-01 -1.02556296e-01
4.60898429e-01 6.26867041e-02 -1.11331288e-02 1.79486185e-01
8.13277781e-01 -3.05734992e-01 3.73100668e-01 1.73904717e-01
-1.88119635e-01 6.11491978e-01 1.29926264e-01 -8.87970924e-01
-9.48649287e-01 -8.09689105e-01 -2.29224246e-02 5.40589094e-01
4.47904289e-01 -1.31908402e-01 -9.96861607e-02 -9.00568485e-01
3.28623921e-01 2.01328605e-01 -3.50619644e-01 -2.94174403e-01
-4.49464351e-01 -5.36111593e-01 3.55037212e-01 5.58384418e-01
7.08493173e-01 -8.14926565e-01 -4.60533530e-01 7.85443857e-02
8.83302391e-02 -1.00419593e+00 -5.53306222e-01 -2.22724050e-01
-9.39236224e-01 -1.00598168e+00 -8.33316982e-01 -7.38936484e-01
7.33253360e-01 9.38604057e-01 5.40842116e-01 2.72693396e-01
-6.35423558e-03 8.63571048e-01 -3.96316051e-01 -5.60517013e-01
1.77493349e-01 -1.41400367e-01 8.52208436e-02 4.97529209e-01
7.19825387e-01 -6.80062532e-01 -3.54112178e-01 1.95585713e-01
-7.89162934e-01 -1.02737367e-01 8.13075066e-01 8.91001642e-01
8.01708937e-01 -1.92366198e-01 5.47779560e-01 -1.15434855e-01
4.77598578e-01 -2.82281399e-01 -5.10219276e-01 2.16079652e-01
-3.73538882e-01 -1.52748749e-01 3.80934954e-01 -2.43325412e-01
-9.08526659e-01 3.12268417e-02 -2.38887817e-01 -1.13831198e+00
-3.86483401e-01 6.79288924e-01 -7.34193385e-01 -2.47140452e-01
2.47718636e-02 7.85767674e-01 3.42278808e-01 -7.85990715e-01
2.50408202e-01 8.52806509e-01 -4.09155227e-02 -2.96378344e-01
6.51380181e-01 6.23189986e-01 6.21606335e-02 -8.61187577e-01
-7.73315430e-01 -4.85504955e-01 -7.34606028e-01 -3.33193272e-01
8.45388591e-01 -9.63050067e-01 -9.18998480e-01 6.59782350e-01
-1.29367542e+00 6.47455871e-01 5.26404008e-02 5.81311941e-01
-3.45742077e-01 6.41921461e-01 -2.37829477e-01 -8.39869738e-01
-4.04736757e-01 -1.39231861e+00 1.45691156e+00 4.35262650e-01
5.28155625e-01 -8.40455890e-01 -9.37949866e-02 3.89311224e-01
2.36283198e-01 2.30646431e-02 8.82830024e-01 -8.36257398e-01
-7.93909311e-01 -5.62296212e-01 -6.31054521e-01 3.17558199e-01
3.20043921e-01 -3.23975146e-01 -8.40292215e-01 -2.55821854e-01
3.04742932e-01 -2.87539661e-01 8.48216534e-01 3.52278531e-01
1.13790786e+00 1.42260745e-01 -2.50286698e-01 6.76307261e-01
1.37435865e+00 4.05893087e-01 2.51509994e-01 1.04521073e-01
1.09551525e+00 3.15322816e-01 3.44243020e-01 4.66314346e-01
6.17840290e-01 6.72006547e-01 6.81206524e-01 -4.99672536e-03
1.24773711e-01 -3.20874274e-01 -1.94018185e-02 1.27596831e+00
-2.67679960e-01 2.47299835e-01 -9.44323063e-01 5.16708553e-01
-1.67257845e+00 -1.08483064e+00 -1.62282698e-02 1.90592980e+00
8.70825499e-02 3.84135954e-02 -9.92888138e-02 8.16545449e-03
6.32501304e-01 2.57548004e-01 -7.07581758e-01 -5.08063883e-02
-3.78456652e-01 -2.98277557e-01 2.91904956e-02 4.65719812e-02
-1.12732410e+00 4.86796468e-01 4.84378195e+00 9.45837140e-01
-1.43814266e+00 -1.49944678e-01 3.40102494e-01 1.01289395e-02
-5.72907150e-01 -1.83228731e-01 -7.84220099e-01 3.22361708e-01
-2.73669306e-02 -7.12297484e-02 2.84417003e-01 8.57536316e-01
6.26828149e-02 3.93685699e-01 -9.20992017e-01 1.58358181e+00
4.32539076e-01 -1.07625794e+00 4.13829505e-01 4.46839184e-01
5.51787138e-01 2.24238813e-01 2.46945601e-02 2.39030659e-01
-2.12448001e-01 -7.22494960e-01 3.90348613e-01 7.49213398e-01
4.89307612e-01 -7.65519619e-01 7.84478188e-01 7.42280006e-01
-1.31811309e+00 -5.77799678e-02 -6.11077011e-01 1.16791591e-01
2.43846402e-02 6.56855285e-01 -3.35288674e-01 1.11756265e+00
5.72675824e-01 1.22829461e+00 -4.73284215e-01 1.19431865e+00
2.70715922e-01 -2.65027899e-02 -5.32784581e-01 -1.63336471e-01
3.94821167e-01 -4.85584289e-01 7.00892508e-01 5.66379011e-01
5.80197930e-01 4.02035952e-01 4.50462133e-01 1.00964665e+00
1.79148182e-01 1.75182506e-01 -9.70678985e-01 -7.90481362e-03
3.06963265e-01 1.35264480e+00 -5.01645207e-01 -3.12192529e-01
-8.47792387e-01 5.48612833e-01 2.30253592e-01 1.30927175e-01
-4.37753379e-01 -2.23430693e-01 4.64915484e-01 -1.95056379e-01
5.33107698e-01 -5.18426836e-01 -3.99970889e-01 -1.32136858e+00
1.69091403e-01 -6.02447569e-01 3.14295799e-01 -8.78460944e-01
-1.60303915e+00 5.09460270e-01 -6.38610944e-02 -1.50505686e+00
3.14255655e-01 -6.40445590e-01 -5.74293494e-01 1.04111087e+00
-1.50091970e+00 -1.37105966e+00 -3.34967107e-01 6.69610023e-01
3.66113067e-01 -5.55385053e-01 6.61994934e-01 4.39541042e-01
-3.17069829e-01 3.73643577e-01 -6.89372420e-02 2.36673132e-01
4.08491075e-01 -9.03317750e-01 2.09797576e-01 3.40478688e-01
3.40500534e-01 4.26592767e-01 -1.72939107e-01 -6.31979287e-01
-1.84765530e+00 -8.42573345e-01 7.25507975e-01 -5.26260555e-01
7.88354650e-02 1.03575513e-02 -1.03349543e+00 2.53000259e-01
-1.32939965e-01 1.63082540e-01 6.74732804e-01 -4.52731596e-03
-4.43390667e-01 -3.18741232e-01 -8.53754163e-01 2.14958295e-01
1.13827729e+00 -5.37266970e-01 -8.18605602e-01 -7.63107277e-03
5.38080275e-01 -4.15672928e-01 -9.42696154e-01 7.74762094e-01
4.72685158e-01 -9.96737540e-01 1.05176401e+00 -5.61946750e-01
3.10453713e-01 -4.08766568e-01 -5.21721363e-01 -1.42027140e+00
-4.97818589e-01 3.22750151e-01 9.59538892e-02 1.04245341e+00
3.42603363e-02 -5.46525717e-01 6.75572157e-01 3.39478850e-01
-3.52804899e-01 -1.06290770e+00 -1.10461199e+00 -5.49121082e-01
-4.70594540e-02 -4.20109510e-01 9.88390625e-01 8.42611492e-01
-2.25463644e-01 5.34550011e-01 -1.74044937e-01 2.15218186e-01
6.17434263e-01 9.21458006e-01 7.36481845e-01 -1.40473402e+00
7.32899904e-02 -6.90656543e-01 -7.07635283e-01 -1.37521732e+00
1.21955477e-01 -1.31044126e+00 -4.52603310e-01 -1.61045945e+00
4.42583591e-01 -3.42736363e-01 -3.57478231e-01 5.14618397e-01
-7.18267858e-02 8.60367119e-02 4.82890487e-01 3.22827667e-01
-6.25023007e-01 1.13403857e+00 1.81262398e+00 -4.63806212e-01
-7.73850977e-02 -1.84667613e-02 -6.06921554e-01 6.17870808e-01
4.49097067e-01 -3.80795412e-02 -2.30208352e-01 -8.06005538e-01
-1.16995834e-01 9.76732299e-02 5.60264826e-01 -6.24387026e-01
4.75869566e-01 3.06990314e-02 8.59165311e-01 -1.19780111e+00
6.77579761e-01 -1.22171819e+00 -2.55742371e-01 2.44447157e-01
2.79860139e-01 -9.07033980e-02 2.29005873e-01 5.55074215e-01
-5.65970421e-01 2.00823098e-01 5.46077073e-01 -3.37012380e-01
-6.93328977e-01 8.28467131e-01 3.53557348e-01 -2.50778615e-01
8.11861932e-01 -5.61428726e-01 -2.40596905e-02 -2.62703031e-01
-7.02992499e-01 3.29398632e-01 1.68852761e-01 8.57295215e-01
1.11464190e+00 -2.02688694e+00 -7.71044970e-01 8.01114976e-01
2.23042846e-01 7.84650594e-02 5.85243404e-01 8.49370122e-01
2.51384340e-02 5.72543979e-01 -3.88440490e-01 -1.04952872e+00
-1.18496573e+00 6.10871375e-01 4.61168468e-01 1.92332223e-01
-4.97423053e-01 6.14242494e-01 3.38282675e-01 -6.86382353e-01
5.50855473e-02 -1.38071030e-01 -4.01142359e-01 1.89528018e-01
3.20728362e-01 1.03885725e-01 1.95273921e-01 -9.94314075e-01
-6.35336280e-01 1.22449672e+00 5.79550117e-02 2.67551124e-01
1.73104119e+00 -1.05331026e-01 -2.29144260e-01 3.69847983e-01
1.47394335e+00 -1.79704830e-01 -1.03727138e+00 -5.41603804e-01
-2.76646197e-01 -7.76569188e-01 1.61435455e-01 -5.03036559e-01
-1.41933131e+00 1.23070753e+00 5.33371031e-01 3.15485656e-01
1.14403105e+00 2.16587167e-02 8.76212955e-01 3.86061043e-01
3.28452855e-01 -6.74911380e-01 6.19822368e-02 7.60675013e-01
1.08108222e+00 -1.43087530e+00 -1.27060652e-01 -1.47392914e-01
-4.44366783e-01 1.17408490e+00 7.53106236e-01 -1.88299686e-01
1.01777601e+00 -1.86872125e-01 2.89806295e-02 -4.86772180e-01
-5.93368530e-01 -3.09486121e-01 6.79326594e-01 4.95691508e-01
3.24203260e-02 -1.49349216e-02 3.91472094e-02 1.04957795e+00
3.31443429e-01 -2.13507369e-01 -3.34916003e-02 6.25663579e-01
-5.62666237e-01 -1.10342717e+00 -4.32701707e-01 5.70749760e-01
-3.87254134e-02 2.63332337e-01 -5.94869137e-01 4.43299085e-01
2.11343646e-01 6.36155665e-01 -6.43213689e-02 -6.31420672e-01
7.39242911e-01 2.31227905e-01 5.87212384e-01 -2.96382725e-01
-1.37626544e-01 4.40216333e-01 -5.38118303e-01 -3.02327931e-01
-6.11542404e-01 -6.12770498e-01 -9.37118769e-01 -1.98946163e-01
-3.79843265e-01 -2.16568291e-01 6.83263958e-01 1.04436803e+00
7.73538530e-01 1.33064494e-01 1.03645968e+00 -1.08620656e+00
-7.31326580e-01 -7.02948630e-01 -5.12744427e-01 5.43704689e-01
1.46246642e-01 -9.72726524e-01 -3.11475307e-01 -4.54601556e-01] | [8.131691932678223, -3.855595588684082] |
56207355-7b54-4acd-88c4-1d18e7513309 | better-fine-tuning-by-reducing | 2008.03156 | null | https://arxiv.org/abs/2008.03156v1 | https://arxiv.org/pdf/2008.03156v1.pdf | Better Fine-Tuning by Reducing Representational Collapse | Although widely adopted, existing approaches for fine-tuning pre-trained language models have been shown to be unstable across hyper-parameter settings, motivating recent work on trust region methods. In this paper, we present a simplified and efficient method rooted in trust region theory that replaces previously used adversarial objectives with parametric noise (sampling from either a normal or uniform distribution), thereby discouraging representation change during fine-tuning when possible without hurting performance. We also introduce a new analysis to motivate the use of trust region methods more generally, by studying representational collapse; the degradation of generalizable representations from pre-trained models as they are fine-tuned for a specific end task. Extensive experiments show that our fine-tuning method matches or exceeds the performance of previous trust region methods on a range of understanding and generation tasks (including DailyMail/CNN, Gigaword, Reddit TIFU, and the GLUE benchmark), while also being much faster. We also show that it is less prone to representation collapse; the pre-trained models maintain more generalizable representations every time they are fine-tuned. | ['Sonal Gupta', 'Luke Zettlemoyer', 'Armen Aghajanyan', 'Akshat Shrivastava', 'Naman Goyal', 'Anchit Gupta'] | 2020-08-06 | null | https://openreview.net/forum?id=OQ08SN70M1V | https://openreview.net/pdf?id=OQ08SN70M1V | iclr-2021-1 | ['cross-lingual-natural-language-inference'] | ['natural-language-processing'] | [-5.72046973e-02 3.42394978e-01 -3.69744718e-01 -4.24239069e-01
-9.30153310e-01 -1.02500832e+00 9.35064077e-01 -2.79035531e-02
-3.67334396e-01 9.39019740e-01 3.67476434e-01 -3.29275727e-01
1.44364256e-02 -6.18183553e-01 -9.59485352e-01 -3.90574872e-01
2.09389955e-01 6.54753685e-01 1.10285319e-01 -5.45998037e-01
3.45495731e-01 4.05654222e-01 -1.27380919e+00 2.96121269e-01
7.90207326e-01 7.12714851e-01 -4.28833216e-01 6.68191075e-01
3.12596709e-01 7.74374187e-01 -8.32935035e-01 -8.26513350e-01
3.46871525e-01 -2.14672163e-01 -1.12011051e+00 -2.36949146e-01
5.51188588e-01 -2.76802450e-01 -2.81721264e-01 9.09066200e-01
3.34641814e-01 5.53800166e-01 6.83488905e-01 -1.39115655e+00
-1.17412329e+00 1.02455962e+00 -2.73317248e-01 2.39943042e-01
1.85716271e-01 3.76927331e-02 8.81152511e-01 -7.85823047e-01
5.06408155e-01 1.48530924e+00 8.78252625e-01 8.99171293e-01
-1.54853833e+00 -7.75688648e-01 4.85962272e-01 -2.70780772e-01
-1.34941256e+00 -7.05306053e-01 3.04992825e-01 -3.13719958e-01
1.22094178e+00 2.30249599e-01 2.00535566e-01 1.56661141e+00
3.75101447e-01 5.78987598e-01 1.11280882e+00 -2.92405993e-01
1.76394820e-01 4.77809191e-01 1.11557223e-01 7.47371852e-01
2.60500520e-01 3.91532719e-01 -3.48544389e-01 -6.52322650e-01
6.90344334e-01 -3.21215510e-01 -3.86124551e-01 -6.34910643e-01
-1.26240885e+00 1.17130256e+00 4.71353203e-01 3.65485460e-01
-2.18139395e-01 5.30347764e-01 6.98686600e-01 7.02916503e-01
7.78721571e-01 9.90007579e-01 -6.08336329e-01 -4.71210256e-02
-1.05294943e+00 4.07641530e-01 8.54560673e-01 8.44236434e-01
5.81732810e-01 3.41318071e-01 -6.34882629e-01 7.29014575e-01
3.01280946e-01 1.18491566e-03 8.99457872e-01 -1.04079652e+00
3.54884505e-01 2.54789531e-01 2.36271873e-01 -6.53780460e-01
1.86782591e-02 -5.01351416e-01 -8.58722985e-01 3.79497051e-01
4.92585778e-01 -1.51888371e-01 -1.03984916e+00 1.99467075e+00
2.59816460e-02 1.16834633e-01 2.44763285e-01 5.78269005e-01
6.93883538e-01 5.30790627e-01 7.26650655e-02 6.14456134e-03
9.50711250e-01 -1.28060043e+00 -4.19535190e-01 -1.34852082e-01
6.36738539e-01 -6.71007991e-01 1.19110954e+00 5.02397418e-01
-1.21563900e+00 -6.68337762e-01 -1.09533525e+00 -9.40644890e-02
-5.21347523e-01 -4.12800133e-01 6.97846711e-01 8.33734334e-01
-1.35835814e+00 1.00530481e+00 -5.15931547e-01 -2.51154572e-01
4.06619787e-01 3.55616212e-01 -5.44512570e-01 5.00238985e-02
-1.47417498e+00 1.24121118e+00 4.30336297e-01 -3.13135445e-01
-1.31792080e+00 -1.10985637e+00 -7.90428758e-01 6.19344898e-02
1.25821248e-01 -1.02491856e+00 1.59299958e+00 -1.04404759e+00
-1.67698765e+00 8.14507723e-01 2.12272912e-01 -8.25452626e-01
7.55557299e-01 -4.93498921e-01 -2.95006484e-01 -3.14822555e-01
-1.50233284e-01 8.78173113e-01 1.42212331e+00 -1.35386860e+00
6.33470388e-03 8.28327686e-02 3.31900418e-01 1.45083606e-01
-4.02862787e-01 4.60131578e-02 -4.93735448e-02 -9.69881058e-01
-4.38726097e-01 -9.38740611e-01 -4.24016327e-01 -2.72104621e-01
-5.94908595e-01 -2.82794923e-01 4.91049081e-01 -1.06395207e-01
1.22032499e+00 -1.86614335e+00 3.41764510e-01 2.18184933e-01
3.17982703e-01 3.87407601e-01 -3.77378941e-01 4.27552104e-01
-5.37211120e-01 6.91211760e-01 -1.82535425e-01 -3.94312203e-01
1.14647470e-01 3.15193653e-01 -7.84625053e-01 4.39566195e-01
-3.23385112e-02 9.31659639e-01 -9.52409208e-01 -2.14013115e-01
-1.68513954e-02 4.05008733e-01 -7.63107002e-01 1.69207215e-01
-2.59079933e-01 1.98513389e-01 -2.20163703e-01 2.19782099e-01
4.52478141e-01 -1.47048056e-01 -3.21494848e-01 1.14312887e-01
2.43277267e-01 1.76003814e-01 -9.21561182e-01 1.73442864e+00
-6.20264947e-01 3.42188299e-01 -1.31527245e-01 -7.51498818e-01
1.02033043e+00 4.27449554e-01 8.11184347e-02 -3.30967188e-01
9.12858769e-02 6.75582439e-02 -1.47588193e-01 1.62434652e-01
8.03889215e-01 -3.74179393e-01 -2.79721826e-01 9.39231157e-01
2.07778409e-01 -5.43316483e-01 1.74594633e-02 4.83578563e-01
1.11475050e+00 1.47016114e-02 3.19193989e-01 -4.76675689e-01
3.10132474e-01 -1.04591109e-01 2.68782079e-01 1.15790272e+00
-2.14860737e-01 6.89498782e-01 4.42274511e-01 -4.18147773e-01
-1.09202480e+00 -1.13708460e+00 5.36296610e-03 1.62272131e+00
-4.26885754e-01 -5.25803506e-01 -8.31896544e-01 -1.12805152e+00
2.67740548e-01 1.03325129e+00 -1.06553113e+00 -7.52799273e-01
-3.66771430e-01 -4.91562426e-01 1.01797998e+00 5.63282073e-01
3.24196696e-01 -1.05974519e+00 -9.66378301e-02 2.19004527e-01
7.12134764e-02 -6.09635115e-01 -5.67232728e-01 3.46654534e-01
-9.30122495e-01 -7.05757916e-01 -8.34390163e-01 -3.38836938e-01
6.96054220e-01 1.10846281e-01 1.64043450e+00 1.64937750e-01
2.53842920e-01 3.11530560e-01 -3.19344044e-01 -2.70218015e-01
-8.67159486e-01 5.04675388e-01 2.19620049e-01 -3.45839918e-01
-1.37329116e-01 -3.45045060e-01 -2.81212032e-01 4.18666542e-01
-9.37194467e-01 -3.44891727e-01 2.86240339e-01 1.02465475e+00
3.27518493e-01 -2.27989167e-01 8.09065163e-01 -1.34760535e+00
1.21445572e+00 -6.62741184e-01 -3.42614174e-01 3.57257545e-01
-9.60354924e-01 4.08303022e-01 8.46681058e-01 -5.02926946e-01
-1.04950643e+00 -2.18255758e-01 6.34289533e-02 -7.42959499e-01
2.16891110e-01 1.64249077e-01 2.36614257e-01 -1.05094865e-01
1.36208272e+00 -2.03764122e-02 9.18275788e-02 -2.78195590e-01
8.95105422e-01 2.07875162e-01 5.81793308e-01 -8.50172877e-01
1.24759102e+00 2.89153337e-01 -5.77223778e-01 -1.36037558e-01
-1.11656928e+00 -1.30775928e-01 -4.62011427e-01 2.22772509e-01
2.06010252e-01 -9.47017789e-01 -2.33198971e-01 4.92592789e-02
-9.40162480e-01 -5.25879025e-01 -7.91091383e-01 2.41585270e-01
-7.04185307e-01 2.86445081e-01 -5.32068431e-01 -3.12032402e-01
-5.74503183e-01 -9.94153202e-01 8.09787631e-01 3.06513626e-02
-6.48628592e-01 -1.30179310e+00 5.57933927e-01 2.66619205e-01
7.20903277e-01 1.40439302e-01 9.58690882e-01 -9.40454543e-01
-5.65616926e-03 -2.21899509e-01 -5.40339686e-02 7.03329742e-01
3.65825812e-03 2.27754861e-01 -1.07947755e+00 -6.57007515e-01
-1.58456340e-01 -7.17968643e-01 1.18853414e+00 1.05191469e-01
1.19586194e+00 -7.15425968e-01 -2.58026868e-01 6.45637989e-01
1.17461228e+00 -3.56020927e-01 6.64698601e-01 4.21240330e-01
5.42649627e-01 3.42938274e-01 5.44628918e-01 4.79519427e-01
4.47235964e-02 4.39911693e-01 5.13057768e-01 -9.23035219e-02
7.85093531e-02 -3.69708210e-01 5.92984378e-01 3.72270882e-01
-1.73231065e-02 -2.86399484e-01 -5.43356121e-01 5.06856024e-01
-1.96603251e+00 -1.00166428e+00 3.71678948e-01 2.27451420e+00
1.21667039e+00 2.29665995e-01 1.79011196e-01 -1.89797096e-02
6.68553412e-01 2.53912687e-01 -7.04027414e-01 -9.63516951e-01
4.04533222e-02 6.45394981e-01 4.69266117e-01 7.06451416e-01
-1.01394892e+00 1.43307853e+00 7.81134605e+00 8.18966568e-01
-8.42083633e-01 1.29006401e-01 6.55996144e-01 -2.45618254e-01
-7.14401722e-01 -1.66371129e-02 -6.45581245e-01 -5.82776144e-02
1.08956337e+00 -4.41793144e-01 5.98097444e-01 1.03697979e+00
-1.24559440e-01 4.21364367e-01 -1.39154553e+00 4.79164988e-01
2.59415954e-02 -1.44185305e+00 2.58555084e-01 -2.10102707e-01
1.06529737e+00 1.29346922e-01 3.45199585e-01 8.10642421e-01
1.15874517e+00 -1.43693697e+00 6.85936153e-01 5.47196031e-01
7.74975598e-01 -9.33490634e-01 4.97696102e-01 2.89863169e-01
-6.53997004e-01 -2.47154664e-02 -5.56737661e-01 3.02752666e-02
-2.19014078e-01 4.90344465e-01 -1.14871359e+00 2.42927760e-01
4.87730891e-01 4.54454452e-01 -8.14217508e-01 6.59612834e-01
-2.88002789e-01 5.80389917e-01 -9.59657580e-02 2.31754526e-01
3.84631127e-01 3.48231673e-01 5.00657320e-01 1.26309013e+00
1.45452514e-01 -4.39706206e-01 5.27837239e-02 9.66641605e-01
-4.84387785e-01 -6.92455396e-02 -1.01575971e+00 -8.10273290e-02
5.01232386e-01 1.21798539e+00 -3.58049870e-01 -4.88871247e-01
-1.42344320e-02 1.03550696e+00 6.15935564e-01 4.21580315e-01
-1.08139384e+00 -2.92043060e-01 8.16229999e-01 -8.08263347e-02
2.44033605e-01 2.45968159e-02 -1.57050610e-01 -1.15558827e+00
-3.09336752e-01 -1.16764677e+00 4.67476606e-01 -7.23324001e-01
-1.42551041e+00 8.07339787e-01 1.41186863e-01 -9.31812704e-01
-7.85697937e-01 -1.63572937e-01 -5.67379832e-01 7.81896591e-01
-1.36676443e+00 -1.03219426e+00 1.75609253e-02 8.45358133e-01
3.90782237e-01 -3.20958376e-01 1.13258111e+00 -2.29087755e-01
-6.40398741e-01 1.13188434e+00 1.97110459e-01 -2.24640090e-02
9.81289148e-01 -1.32723892e+00 8.22509587e-01 6.95734441e-01
2.74581432e-01 9.99841690e-01 8.65320861e-01 -5.15725255e-01
-9.00320351e-01 -1.20262468e+00 7.60325193e-01 -7.53132582e-01
6.87982619e-01 -3.63639385e-01 -1.18321347e+00 9.47665930e-01
4.34079111e-01 1.87474504e-01 4.55028445e-01 2.91310549e-01
-1.07165313e+00 1.04135193e-01 -1.43667352e+00 6.35388732e-01
9.26338613e-01 -6.36535525e-01 -7.80700684e-01 4.95415986e-01
9.51822937e-01 -2.97257304e-01 -9.82237101e-01 2.61362165e-01
4.20732379e-01 -1.03275919e+00 1.16312838e+00 -1.15188456e+00
2.52631307e-01 -5.44110276e-02 -8.47695321e-02 -2.00446653e+00
-7.77568221e-01 -1.12649035e+00 -1.80442259e-01 1.35078955e+00
5.05091727e-01 -7.03672528e-01 6.89882040e-01 5.26228189e-01
-1.39839053e-01 -5.83540440e-01 -9.57980037e-01 -8.90591741e-01
6.82606876e-01 -2.07005456e-01 6.54706597e-01 1.06619406e+00
-2.68390030e-02 4.40056741e-01 -3.54964346e-01 -1.82361335e-01
3.82301182e-01 -1.37561202e-01 7.88904369e-01 -1.07616973e+00
-3.93027425e-01 -6.04044259e-01 -1.57889146e-02 -6.92149282e-01
4.39116269e-01 -1.04528153e+00 2.07352079e-02 -1.10887003e+00
8.02652165e-02 -5.00917256e-01 -5.05691171e-01 7.34124124e-01
-1.74676493e-01 3.02815765e-01 1.16086498e-01 2.37576544e-01
-5.22014916e-01 5.92934966e-01 1.30016482e+00 -1.82513729e-01
-6.86212480e-02 8.50331932e-02 -1.33709574e+00 6.02875710e-01
9.22009706e-01 -5.95072150e-01 -6.97016656e-01 -3.24557871e-01
3.98652226e-01 -3.29157442e-01 2.21848071e-01 -8.16233158e-01
-2.17562802e-02 -4.11971211e-02 3.76770914e-01 4.74067181e-02
2.54981488e-01 -4.24487174e-01 -4.60124910e-02 3.91997755e-01
-8.17098796e-01 4.16829765e-01 3.64058256e-01 5.58752120e-01
3.33817340e-02 -4.56868619e-01 1.09611773e+00 -2.99022317e-01
-4.37789053e-01 2.09125280e-01 -3.52264255e-01 3.76974732e-01
9.85342085e-01 -9.53134336e-03 -5.43646336e-01 -4.97912675e-01
-6.87910914e-01 -2.97921766e-02 5.20988345e-01 7.05628395e-01
5.33587754e-01 -1.35946631e+00 -9.76620436e-01 2.36709833e-01
2.49667197e-01 -1.61074594e-01 -5.15715592e-03 2.39168942e-01
-2.05178782e-01 2.73673415e-01 -2.10217893e-01 -2.00474307e-01
-1.01355326e+00 7.35243559e-01 4.20946300e-01 -6.06736124e-01
-3.36709112e-01 1.31236947e+00 2.94678152e-01 -6.48268163e-01
2.67634481e-01 -4.81827140e-01 -1.01191029e-01 4.37508300e-02
3.86938483e-01 1.59369498e-01 1.51987761e-01 -3.19069326e-01
-1.72682405e-01 2.11901113e-01 -6.62576497e-01 -2.01933712e-01
1.06839991e+00 1.71628907e-01 1.44545749e-01 2.93503076e-01
1.15076196e+00 -1.17039807e-01 -1.27795410e+00 -2.80424774e-01
-2.94093370e-01 -3.24580520e-01 -3.74004021e-02 -8.64644349e-01
-1.06343699e+00 6.56306446e-01 3.54241997e-01 3.77106518e-01
7.11624682e-01 -3.93582322e-02 5.58497071e-01 6.25872791e-01
2.46301547e-01 -9.80911791e-01 3.09482843e-01 6.72705173e-01
1.31989515e+00 -1.02136517e+00 7.58968070e-02 7.76504725e-02
-8.52685750e-01 8.04202974e-01 6.39452636e-01 -4.51851577e-01
4.75930214e-01 1.31310821e-01 -1.79642625e-03 4.74378429e-02
-1.14485598e+00 1.74583226e-01 2.47284353e-01 8.80324602e-01
4.47774857e-01 1.29751535e-02 2.71465033e-01 4.44003135e-01
-6.37889445e-01 -1.54298365e-01 5.79363465e-01 6.10843718e-01
-3.52086723e-01 -1.16944325e+00 -3.96541595e-01 3.71337861e-01
-6.43997371e-01 -3.47240478e-01 -6.53833389e-01 9.76187766e-01
-1.69310614e-01 8.99304628e-01 -1.81622222e-01 -4.23195928e-01
3.17589551e-01 2.29114443e-01 5.33229232e-01 -8.89473438e-01
-1.28959751e+00 -4.70239758e-01 1.88209429e-01 -6.43075705e-01
1.59904901e-02 -5.13260484e-01 -9.89875793e-01 -5.32416999e-01
-3.92941177e-01 4.38831031e-01 3.72448951e-01 7.32468724e-01
3.21372122e-01 5.04073083e-01 6.92103624e-01 -6.71853364e-01
-1.16401958e+00 -1.11279082e+00 -2.48804063e-01 4.33029532e-01
3.09546947e-01 -6.79788411e-01 -5.17230630e-01 -1.15835004e-01] | [10.572759628295898, 8.180859565734863] |
c4f335fd-2fc0-4196-82eb-b487c9c6fd2f | baa-ngp-bundle-adjusting-accelerated-neural | 2306.04166 | null | https://arxiv.org/abs/2306.04166v2 | https://arxiv.org/pdf/2306.04166v2.pdf | BAA-NGP: Bundle-Adjusting Accelerated Neural Graphics Primitives | Implicit neural representation has emerged as a powerful method for reconstructing 3D scenes from 2D images. Given a set of camera poses and associated images, the models can be trained to synthesize novel, unseen views. In order to expand the use cases for implicit neural representations, we need to incorporate camera pose estimation capabilities as part of the representation learning, as this is necessary for reconstructing scenes from real-world video sequences where cameras are generally not being tracked. Existing approaches like COLMAP and, most recently, bundle-adjusting neural radiance field methods often suffer from lengthy processing times. These delays ranging from hours to days, arise from laborious feature matching, hardware limitations, dense point sampling, and long training times required by a multi-layer perceptron structure with a large number of parameters. To address these challenges, we propose a framework called bundle-adjusting accelerated neural graphics primitives (BAA-NGP). Our approach leverages accelerated sampling and hash encoding to expedite both pose refinement/estimation and 3D scene reconstruction. Experimental results demonstrate that our method achieves a more than 10 to 20 $\times$ speed improvement in novel view synthesis compared to other bundle-adjusting neural radiance field methods without sacrificing the quality of pose estimation. | ['Michael Yip', 'Alexey Supikov', 'Shreya Saha', 'Jingpei Lu', 'Shan Lin', 'Sainan Liu'] | 2023-06-07 | null | null | null | null | ['pose-estimation', '3d-scene-reconstruction', 'novel-view-synthesis'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 4.42565978e-01 -1.48171768e-01 2.18410417e-01 -3.80253792e-01
-8.38450789e-01 -6.10328913e-01 5.07496297e-01 -1.69104502e-01
-2.62168169e-01 3.16786647e-01 1.69330686e-01 -1.41152680e-01
2.57133633e-01 -7.36572206e-01 -1.09068930e+00 -3.39948893e-01
2.57948011e-01 4.89092976e-01 1.02839202e-01 3.24079059e-02
3.84152889e-01 9.72937346e-01 -1.79397571e+00 -3.12235992e-04
3.68065983e-01 9.83662784e-01 2.73469388e-01 7.70845175e-01
-1.85287464e-02 5.84330678e-01 -3.53937209e-01 -2.38760963e-01
6.24840617e-01 1.41829714e-01 -3.63350004e-01 4.04808551e-01
9.79348063e-01 -1.15103042e+00 -4.24778938e-01 7.17616618e-01
4.68967587e-01 2.76296377e-01 4.64489251e-01 -9.74176943e-01
3.33504342e-02 6.97801728e-03 -7.48529613e-01 -1.30175516e-01
5.10522425e-01 2.09277362e-01 6.86707258e-01 -1.18374872e+00
5.47439516e-01 1.20310986e+00 9.92239714e-01 3.29862505e-01
-1.28255177e+00 -6.67973936e-01 1.91321850e-01 -6.83721900e-02
-1.40826166e+00 -7.78620958e-01 9.40206230e-01 -4.71565157e-01
1.28132439e+00 1.08019643e-01 7.70032346e-01 9.56151187e-01
-8.55948403e-02 6.14131570e-01 5.96045732e-01 -2.41324261e-01
9.77946669e-02 -1.28516659e-01 -1.61490500e-01 8.95920932e-01
1.10785551e-01 1.13450475e-01 -7.31189132e-01 -2.61866659e-01
1.47191155e+00 2.33154580e-01 -4.58396077e-01 -7.62745380e-01
-1.34791267e+00 6.81425869e-01 4.71821845e-01 -6.55630350e-01
-5.21581411e-01 5.43699145e-01 1.42319605e-01 2.72290353e-02
4.02255028e-01 4.68999982e-01 -4.58452970e-01 -2.21645117e-01
-9.59801614e-01 4.06125784e-01 6.26087308e-01 1.27040946e+00
1.10521710e+00 2.53102988e-01 4.66215789e-01 5.41370273e-01
2.86200672e-01 4.83788401e-01 1.34493643e-02 -1.40762591e+00
4.11508828e-01 4.07931983e-01 6.58473596e-02 -1.16274798e+00
-3.18942994e-01 -3.24979573e-01 -7.76195526e-01 3.21530759e-01
1.22234359e-01 1.31758332e-01 -9.50048745e-01 1.40894270e+00
6.16687596e-01 3.52622271e-01 -2.90589213e-01 9.13557410e-01
5.75217366e-01 8.33436728e-01 -3.93113613e-01 5.42962514e-02
9.76887465e-01 -9.53970790e-01 -1.09780259e-01 -2.56783277e-01
3.85277480e-01 -9.23859119e-01 8.04399192e-01 4.86958712e-01
-1.31669915e+00 -5.37988126e-01 -1.18062472e+00 -5.09567857e-01
9.10065398e-02 -2.72601843e-01 8.13676119e-01 4.93536383e-01
-1.07484019e+00 5.89461088e-01 -1.00953555e+00 -7.52260759e-02
3.85675550e-01 6.02197349e-01 -3.88678342e-01 -3.50561261e-01
-3.68819773e-01 4.68156338e-01 1.21372767e-01 5.51529154e-02
-7.97348320e-01 -9.35768306e-01 -1.03137565e+00 7.68042952e-02
4.30249602e-01 -1.12075877e+00 1.27222574e+00 -6.16764367e-01
-1.44578898e+00 5.72790504e-01 -1.09392367e-01 -1.70275480e-01
2.83501089e-01 -5.31685472e-01 1.46785900e-01 2.17226788e-01
-1.11363471e-01 9.86403048e-01 1.09313190e+00 -1.35185969e+00
-4.33617741e-01 -4.06272739e-01 1.74280554e-01 6.48788214e-01
-2.15787031e-02 -3.16562951e-01 -7.98915327e-01 -6.29961312e-01
6.06057763e-01 -9.89991903e-01 -4.72895920e-01 6.52126849e-01
-1.30029067e-01 2.48170048e-01 9.83863413e-01 -6.55825496e-01
4.61807340e-01 -2.09162116e+00 1.95404753e-01 2.21572578e-01
2.83585370e-01 -1.02556974e-01 3.32847685e-02 3.41069065e-02
1.26449898e-01 -3.42233181e-01 -6.75806627e-02 -7.72750437e-01
-1.38605773e-01 2.07174569e-01 -4.73530650e-01 4.75314111e-01
6.77427873e-02 4.03655559e-01 -5.97072482e-01 -2.02004328e-01
5.41357100e-01 9.09928024e-01 -9.97597218e-01 2.60371387e-01
-3.47365737e-01 4.75388914e-01 -2.69808084e-01 7.56447315e-01
7.17564464e-01 -4.46870506e-01 -5.54641411e-02 -5.25828421e-01
-1.14091821e-01 3.36932868e-01 -1.32581699e+00 2.38322687e+00
-4.60351735e-01 5.82895875e-01 1.06496140e-01 -5.16204357e-01
7.73318112e-01 1.53567165e-01 5.29398441e-01 -4.52037930e-01
2.25595891e-01 4.09993418e-02 -7.02748716e-01 -1.45831153e-01
8.92020345e-01 1.96648017e-03 1.18742689e-01 4.34965700e-01
-1.65145248e-01 -5.79776227e-01 -3.29964995e-01 3.64659242e-02
9.74688888e-01 7.34044969e-01 1.45980358e-01 2.52353817e-01
8.44372213e-02 1.08140424e-01 7.18199134e-01 3.96477550e-01
2.45024741e-01 1.06346345e+00 1.93806291e-02 -8.05172384e-01
-1.31381524e+00 -9.86917377e-01 -1.43313315e-02 7.95753777e-01
1.51043221e-01 -3.31183523e-01 -6.08601809e-01 -2.74401277e-01
-3.45877521e-02 5.91475487e-01 -1.32442221e-01 6.43099174e-02
-9.73102927e-01 -3.20024848e-01 2.25384474e-01 6.55008197e-01
3.97098213e-01 -5.28755784e-01 -1.02363360e+00 3.07347655e-01
2.79147048e-02 -1.39386463e+00 -5.51940799e-01 1.31423220e-01
-1.19451666e+00 -8.68027985e-01 -6.22379243e-01 -6.18257284e-01
8.97579193e-01 7.65877485e-01 1.18137681e+00 1.79567747e-02
-2.66093314e-01 5.95331609e-01 1.70149598e-02 -2.11222038e-01
-1.93511546e-01 -8.30954388e-02 1.40646577e-01 -2.13712037e-01
1.09996796e-01 -9.67386127e-01 -8.29988897e-01 2.73388326e-01
-8.60205948e-01 6.48602605e-01 6.94793463e-01 6.92941189e-01
8.86573911e-01 -3.42548281e-01 -1.81170985e-01 -6.98761523e-01
-7.76930675e-02 -1.90875322e-01 -9.98440444e-01 -1.67365119e-01
-4.16737258e-01 1.40544936e-01 5.12375653e-01 -4.00268286e-01
-9.29993153e-01 5.56671619e-01 -9.92612541e-02 -1.00887418e+00
-8.57179835e-02 3.02377969e-01 -1.17353693e-01 -3.05662721e-01
6.36563957e-01 1.88320905e-01 9.94594470e-02 -5.76229632e-01
3.78369242e-01 2.29656428e-01 7.48354733e-01 -5.47422409e-01
9.47997153e-01 5.23148000e-01 1.30716249e-01 -8.45713139e-01
-6.51385069e-01 -4.22860235e-01 -8.29048932e-01 -1.84540898e-01
7.55779624e-01 -1.32621610e+00 -5.39330304e-01 4.09738570e-01
-1.42539358e+00 -2.13085487e-01 -8.91851783e-02 5.47633231e-01
-7.78614998e-01 4.06237543e-01 -5.41998148e-01 -4.88344580e-01
-3.84430885e-01 -1.45266593e+00 1.40998423e+00 4.58384342e-02
-2.86887169e-01 -6.63681149e-01 -7.98858106e-02 5.57499945e-01
2.60624975e-01 5.36648870e-01 8.29911828e-01 2.69614726e-01
-1.43800247e+00 -1.69786915e-01 -2.68398196e-01 8.47894549e-02
-1.32037431e-01 -2.36571178e-01 -1.11290181e+00 -5.29044211e-01
9.87092853e-02 -3.81223172e-01 3.72674465e-01 3.14585567e-01
1.27683926e+00 -3.05720985e-01 -1.51114464e-01 1.32854724e+00
1.46315360e+00 1.08276263e-01 4.90007937e-01 3.39152038e-01
1.12026525e+00 4.04179931e-01 3.71785283e-01 6.56027675e-01
4.82798040e-01 7.66365767e-01 6.16613328e-01 -7.60279149e-02
-1.48466855e-01 -4.27052319e-01 1.46392643e-01 1.02662194e+00
-8.32770765e-02 6.38430864e-02 -7.67971992e-01 2.81124473e-01
-1.49243164e+00 -7.32171297e-01 1.73993289e-01 2.43975949e+00
5.94852269e-01 4.96167727e-02 -1.83783159e-01 9.13934503e-03
3.04018974e-01 2.20880628e-01 -8.89697373e-01 -1.86072215e-01
1.16120651e-01 1.23225063e-01 7.04360306e-01 4.76062596e-01
-7.98546910e-01 8.03722858e-01 5.86822701e+00 2.87525684e-01
-1.18381095e+00 -1.74474314e-01 4.67225462e-01 -4.39326555e-01
-4.57596958e-01 1.15221888e-01 -7.63528824e-01 -1.19497038e-01
6.49043918e-01 2.26170897e-01 5.93182981e-01 9.48010564e-01
7.74566922e-03 -2.12067813e-02 -1.32948864e+00 1.56269157e+00
4.04463530e-01 -1.51070905e+00 2.27077827e-01 2.19350696e-01
5.29437065e-01 2.76686043e-01 8.22305009e-02 -2.98164543e-02
1.61693797e-01 -8.42774332e-01 8.49267066e-01 3.96978855e-01
1.05812216e+00 -8.15764129e-01 4.18224968e-02 2.95923114e-01
-1.05741048e+00 1.48517177e-01 -4.96012330e-01 -3.30580249e-02
4.46568191e-01 4.78800178e-01 -1.00525439e+00 2.28066280e-01
7.99080551e-01 7.50676513e-01 -2.66972423e-01 1.01518071e+00
1.30966559e-01 1.44858703e-01 -6.22744381e-01 2.78431892e-01
1.53273270e-01 -3.13805342e-02 5.99755943e-01 6.86491847e-01
6.16393328e-01 1.60428360e-01 8.32150225e-03 7.49191582e-01
-7.56749660e-02 -2.98882753e-01 -8.64990413e-01 2.55289942e-01
4.24599737e-01 1.21334386e+00 -7.18629360e-01 -1.22090019e-01
-5.09017289e-01 1.20956302e+00 2.64920115e-01 2.99044937e-01
-7.83982217e-01 -1.81833357e-02 7.20698059e-01 4.07619417e-01
4.38727438e-01 -8.38718712e-01 -2.84500480e-01 -1.13537014e+00
3.49191017e-02 -8.46878767e-01 -1.28703848e-01 -1.12424314e+00
-8.55810940e-01 4.97209519e-01 9.33103263e-02 -1.40663254e+00
-5.20568252e-01 -6.20835841e-01 -2.87310220e-02 6.95235550e-01
-1.44885921e+00 -1.08415091e+00 -5.57478964e-01 5.73146701e-01
8.78430247e-01 -1.14645675e-01 8.02703023e-01 3.00481826e-01
-2.56502837e-01 3.75315070e-01 -1.33386299e-01 -1.19621582e-01
3.77097756e-01 -8.70763719e-01 7.96497583e-01 7.20452845e-01
2.51627147e-01 5.79128683e-01 5.40054023e-01 -4.09032822e-01
-2.00654125e+00 -9.97502267e-01 2.84981698e-01 -5.03431797e-01
3.21639851e-02 -5.95625579e-01 -7.30391383e-01 8.74726474e-01
-8.04172307e-02 -4.78050336e-02 6.95419848e-01 3.44612598e-02
-5.41246176e-01 -1.07248239e-01 -9.90913451e-01 8.54002833e-01
1.18842447e+00 -5.75389683e-01 -2.03871816e-01 2.12045208e-01
9.02440310e-01 -8.25897217e-01 -8.47967446e-01 4.14300561e-01
6.76344573e-01 -1.18868494e+00 1.32293749e+00 1.68320502e-03
2.67212033e-01 -4.34301734e-01 -5.21298468e-01 -9.46264803e-01
-4.99173015e-01 -7.37872601e-01 -2.99461037e-01 6.77928805e-01
9.40974057e-02 -3.06647986e-01 1.05945909e+00 9.15353060e-01
-3.59065950e-01 -6.70318782e-01 -6.61380172e-01 -2.62399495e-01
-4.98722494e-01 -6.41266227e-01 8.18596542e-01 8.70013595e-01
-8.16826642e-01 4.91051078e-01 -3.82692069e-01 5.65633059e-01
9.33656514e-01 6.89643621e-02 1.32801557e+00 -1.00734961e+00
-6.83300972e-01 -2.80603647e-01 -4.93588924e-01 -1.70691085e+00
-1.35714591e-01 -4.50856209e-01 9.34961811e-02 -1.32971752e+00
-7.95642734e-02 -6.86150253e-01 1.84414193e-01 3.17539990e-01
1.40529737e-01 3.75785410e-01 1.77734986e-01 4.02902961e-01
-2.90230960e-01 6.20868802e-01 1.06920326e+00 7.57768005e-02
-2.02217594e-01 -1.56341597e-01 -4.78937387e-01 9.69264925e-01
3.43367994e-01 -3.04751188e-01 -7.34671831e-01 -1.31388259e+00
4.71208006e-01 3.52058142e-01 4.68636155e-01 -1.31714892e+00
3.98342580e-01 3.95142520e-03 7.21907079e-01 -1.10010087e+00
1.02974558e+00 -1.03436136e+00 5.64208865e-01 1.04333840e-01
-3.23532633e-02 5.32618999e-01 2.48513326e-01 6.41526401e-01
8.27883855e-02 1.27043445e-02 4.97702062e-01 -2.53502876e-01
-7.79608130e-01 6.48679554e-01 -1.16431043e-01 -4.15730476e-01
7.26486146e-01 -7.59346366e-01 1.00387275e-01 -3.86701494e-01
-2.06207186e-01 -1.32977083e-01 8.87163103e-01 3.38913411e-01
1.08721220e+00 -1.21432161e+00 -4.22187537e-01 5.60583174e-01
2.57810466e-02 9.65040267e-01 2.91206062e-01 4.67203736e-01
-1.12707412e+00 1.30533397e-01 -1.03377521e-01 -9.58329618e-01
-1.22246087e+00 3.37087780e-01 1.69659823e-01 1.49342775e-01
-1.01329100e+00 9.27164197e-01 4.16215420e-01 -4.77248996e-01
3.67308527e-01 -3.40978712e-01 3.23395967e-01 -3.82405728e-01
3.86906266e-01 3.50706339e-01 3.74881215e-02 -6.79548562e-01
7.91029632e-02 8.78830969e-01 -2.57632613e-01 -1.85973868e-01
1.48425710e+00 -8.58285353e-02 1.07783794e-01 3.51678669e-01
1.35305738e+00 -8.14020634e-02 -1.87125003e+00 -3.15509886e-01
-4.72924054e-01 -7.25899518e-01 4.16786015e-01 -2.45052472e-01
-1.03391862e+00 8.03035676e-01 5.02274513e-01 -4.17689860e-01
1.12308693e+00 -3.17518324e-01 1.19879305e+00 6.76918745e-01
7.71484733e-01 -8.09047520e-01 -5.99079840e-02 4.17880267e-01
8.09686303e-01 -1.07337725e+00 3.99730653e-01 -4.87625986e-01
-1.71688035e-01 1.17789221e+00 5.49588919e-01 -1.28046572e-01
5.34935832e-01 3.39469343e-01 7.61191025e-02 -2.62564451e-01
-5.42885005e-01 4.61222112e-01 2.51248568e-01 5.38676977e-01
1.51320025e-01 -2.22322419e-01 6.86532021e-01 -2.09504247e-01
-4.98929918e-01 -2.62847692e-01 4.49441731e-01 1.12956583e+00
-2.88282990e-01 -8.89591336e-01 -5.04719675e-01 3.57613832e-01
8.12423900e-02 -1.54358167e-02 1.46364197e-01 6.01297557e-01
-1.46661818e-01 3.73703450e-01 4.12865520e-01 -3.66180182e-01
3.60920399e-01 -2.44691417e-01 7.61909366e-01 -6.47338986e-01
-3.38353336e-01 3.03106010e-01 1.12601025e-02 -9.53708708e-01
-4.06672716e-01 -6.96169376e-01 -1.07751846e+00 -4.09751594e-01
-1.76843256e-01 -5.20284653e-01 9.17594194e-01 6.77579701e-01
5.91918409e-01 2.20725775e-01 6.96610332e-01 -1.55519128e+00
-4.36421335e-01 -4.40837145e-01 -2.22431615e-01 2.19055504e-01
4.15339231e-01 -7.11634099e-01 -8.95411372e-02 1.86100975e-01] | [8.744787216186523, -2.8266215324401855] |
44ac45e0-d291-42e0-80ec-4888bc5ce17e | change-point-models-for-real-time-cyber | 2003.04185 | null | https://arxiv.org/abs/2003.04185v1 | https://arxiv.org/pdf/2003.04185v1.pdf | Change Point Models for Real-time Cyber Attack Detection in Connected Vehicle Environment | Connected vehicle (CV) systems are cognizant of potential cyber attacks because of increasing connectivity between its different components such as vehicles, roadside infrastructure, and traffic management centers. However, it is a challenge to detect security threats in real-time and develop appropriate or effective countermeasures for a CV system because of the dynamic behavior of such attacks, high computational power requirement, and a historical data requirement for training detection models. To address these challenges, statistical models, especially change point models, have potentials for real-time anomaly detections. Thus, the objective of this study is to investigate the efficacy of two change point models, Expectation Maximization (EM) and two forms of Cumulative Summation (CUSUM) algorithms (i.e., typical and adaptive), for real-time V2I cyber attack detection in a CV Environment. To prove the efficacy of these models, we evaluated these two models for three different type of cyber attack, denial of service (DOS), impersonation, and false information, using basic safety messages (BSMs) generated from CVs through simulation. Results from numerical analysis revealed that EM, CUSUM, and adaptive CUSUM could detect these cyber attacks, DOS, impersonation, and false information, with an accuracy of (99%, 100%, 100%), (98%, 10%, 100%), and (100%, 98%, 100%) respectively. | ['Gurcan Comert', 'Mhafuzul Islam', 'Mizanur Rahman', 'Mashrur Chowdhury'] | 2020-03-05 | null | null | null | null | ['cyber-attack-detection'] | ['miscellaneous'] | [-2.27689669e-01 -4.45624650e-01 1.59462795e-01 2.64172405e-01
5.54201454e-02 -6.72316730e-01 5.35254717e-01 5.28451502e-01
-1.54338211e-01 6.97888136e-01 -6.80599809e-01 -1.05490530e+00
-2.15447992e-01 -8.59335959e-01 -4.64400023e-01 -4.73416537e-01
-5.43666422e-01 -8.49763975e-02 7.38470674e-01 -1.34332761e-01
3.98740351e-01 9.25051570e-01 -1.48090148e+00 -2.13372841e-01
7.74298787e-01 1.18510020e+00 -4.53733981e-01 7.49720573e-01
-6.25454262e-02 4.31080818e-01 -1.07291055e+00 -3.53232831e-01
2.40047738e-01 1.80592481e-02 -1.61288108e-03 -3.64028871e-01
-3.46057415e-01 -2.96989083e-01 -3.72826427e-01 1.12385452e+00
7.31221437e-02 8.70906748e-04 7.56841123e-01 -2.30621886e+00
3.75613458e-02 -1.14029059e-02 -8.10788453e-01 6.93143189e-01
8.53682160e-02 3.52132350e-01 8.49549323e-02 -5.30688882e-01
1.64182976e-01 1.02361941e+00 3.82144809e-01 3.30207467e-01
-6.85177028e-01 -1.17053306e+00 8.01725015e-02 5.00860929e-01
-1.39716578e+00 -7.80228451e-02 5.65643549e-01 -3.70844156e-01
8.73561740e-01 3.38863820e-01 4.21729356e-01 8.88546824e-01
8.56384456e-01 3.03560555e-01 7.82317877e-01 -2.40777940e-01
4.16956931e-01 4.85592008e-01 4.23512489e-01 2.64165308e-02
9.30847824e-01 5.86471736e-01 2.40035370e-01 -6.15023315e-01
6.96732402e-02 1.85373545e-01 2.14665756e-02 2.16522038e-01
-2.98000187e-01 6.85654461e-01 -2.16352388e-01 1.48661703e-01
-6.02701247e-01 -1.43595517e-01 6.32981658e-01 3.23678672e-01
1.60756528e-01 -6.31354526e-02 -4.37815636e-01 -2.03715786e-01
-3.00007284e-01 -1.33516297e-01 7.37042487e-01 9.94616210e-01
2.28128374e-01 6.92024827e-01 4.73489761e-02 8.23945850e-02
5.12691557e-01 1.08621144e+00 2.43574262e-01 -2.16721028e-01
1.45206988e-01 4.23317015e-01 -9.79333930e-03 -1.31545055e+00
-2.42541149e-01 -3.82689685e-01 -8.30060065e-01 2.62278318e-01
-2.38315731e-01 -4.89054173e-01 -7.94604003e-01 1.54110920e+00
4.71180052e-01 8.48052740e-01 3.18643272e-01 2.00241491e-01
1.29232749e-01 8.73909354e-01 4.59717959e-01 -6.28049850e-01
8.81455362e-01 1.43996686e-01 -8.08861196e-01 8.97726044e-02
5.33405423e-01 -7.69283533e-01 3.46026450e-01 2.09603742e-01
-5.96830249e-01 -1.81499958e-01 -1.27546692e+00 1.39644086e+00
-3.61809492e-01 -4.60537136e-01 2.90171653e-01 1.18102038e+00
-5.77958345e-01 2.11653467e-02 -6.33216560e-01 -2.78930306e-01
6.52303547e-02 8.46377611e-02 2.65127905e-02 4.55408879e-02
-1.34224260e+00 8.08548868e-01 7.42498711e-02 4.93847020e-02
-1.05111849e+00 -6.33764565e-01 -5.63541472e-01 -5.09810485e-02
3.58223766e-01 1.58320069e-01 8.19244027e-01 -4.39466953e-01
-8.42354774e-01 7.19526634e-02 1.02447301e-01 -6.20995045e-01
1.26073450e-01 2.14350820e-01 -1.10551012e+00 7.49812089e-03
-1.93341851e-01 -3.01700950e-01 3.53082627e-01 -1.29145980e+00
-7.41104841e-01 -4.94414181e-01 -9.41363573e-02 -5.12450457e-01
-8.21700320e-02 4.27088588e-01 5.02377272e-01 -6.58874214e-02
-1.70934230e-01 -8.63765359e-01 -5.67471907e-02 -6.42638087e-01
-4.97533679e-01 -2.31644660e-01 1.45441675e+00 -5.46138167e-01
1.23555553e+00 -2.22932482e+00 -1.07969177e+00 8.74533534e-01
-1.18963614e-01 8.41525197e-01 1.16736874e-01 7.08876252e-01
-9.63258594e-02 1.80203184e-01 -9.41893272e-03 4.60243881e-01
-2.81462133e-01 1.32491171e-01 -4.04434264e-01 5.63821554e-01
5.38870953e-02 1.93905577e-01 -6.65802300e-01 -2.30785329e-02
5.92988670e-01 2.09245041e-01 -1.70534253e-01 8.46100003e-02
1.44298196e-01 1.22188358e-02 -5.74808836e-01 7.25993514e-01
9.73531008e-01 4.25141305e-01 7.16250166e-02 6.57321960e-02
-3.24058771e-01 -2.91402757e-01 -1.43481958e+00 -2.29981422e-01
-1.80938020e-01 7.67539024e-01 -1.05401427e-01 -1.09265220e+00
9.81902540e-01 6.32337928e-01 5.55331230e-01 -1.00096262e+00
6.97043240e-01 1.67896554e-01 1.67325854e-01 -4.38366383e-01
1.32932216e-01 3.23252082e-01 5.82337976e-02 2.86143273e-01
-3.95871133e-01 5.39648354e-01 1.37383193e-02 5.19417644e-01
1.24698651e+00 -1.08246732e+00 2.08591372e-01 -9.49739069e-02
7.14632213e-01 2.29177065e-02 5.50108492e-01 4.56384778e-01
-6.21981084e-01 -4.09767061e-01 6.84506953e-01 -6.73390552e-02
-8.64789307e-01 -1.10028613e+00 -1.18973754e-01 -2.17748675e-02
5.84760785e-01 3.58884856e-02 -4.93776888e-01 -4.52704817e-01
1.14547044e-01 1.30836141e+00 -1.10095277e-01 -9.43855703e-01
-3.77243906e-01 -7.71306098e-01 6.91729367e-01 2.87494659e-01
4.97472227e-01 -5.56432962e-01 -6.32176340e-01 1.22845478e-01
2.68751144e-01 -1.36492836e+00 -2.06744559e-02 -3.16047698e-01
-3.41997266e-01 -1.59813654e+00 3.19591999e-01 -1.34016499e-01
6.96679354e-01 7.74590850e-01 4.72826093e-01 2.84996927e-01
-3.96833628e-01 6.93614066e-01 -2.80277014e-01 -8.75135362e-01
-8.90120685e-01 -8.75066936e-01 4.58712786e-01 2.18722463e-01
5.96714020e-01 -2.44450241e-01 -3.88639838e-01 5.52810550e-01
-9.99254405e-01 -8.41258407e-01 4.30909485e-01 2.29905128e-01
1.71832293e-01 5.25035262e-01 9.84557569e-01 -3.63191277e-01
8.31533015e-01 -1.11892700e+00 -1.00033307e+00 1.79963708e-02
-9.60416675e-01 -4.81519759e-01 5.96554935e-01 -7.56011665e-01
-7.06859350e-01 -3.79527062e-01 -4.47231382e-02 -5.35160244e-01
-3.36284429e-01 4.14215297e-01 -1.93104655e-01 -9.71150547e-02
5.92874467e-01 1.82878956e-01 3.34451050e-01 1.47339031e-01
-2.27754056e-01 9.31571007e-01 1.58204421e-01 2.02671606e-02
1.12738931e+00 1.70034513e-01 2.54699439e-01 -9.82197165e-01
1.35581195e-01 -4.91868287e-01 1.93059072e-01 -6.25280440e-01
4.06691700e-01 -6.53952181e-01 -1.13368642e+00 7.06790864e-01
-1.02670467e+00 2.03585833e-01 3.09237927e-01 6.83478057e-01
2.68003613e-01 6.92784190e-01 -3.74127358e-01 -1.47476149e+00
-3.54120255e-01 -8.78140450e-01 8.55073407e-02 4.50632334e-01
1.03435822e-01 -7.58255899e-01 -2.39265889e-01 -1.22481413e-01
7.17267334e-01 6.83550894e-01 8.80390108e-01 -1.13508415e+00
-4.97914672e-01 -1.03471446e+00 -2.45895565e-01 5.51707208e-01
5.29558165e-04 6.18035674e-01 -4.53531444e-01 -4.58376139e-01
1.92295745e-01 3.26864064e-01 -1.81091666e-01 1.97673842e-01
6.85185134e-01 -4.81853485e-01 -6.02552176e-01 2.33200919e-02
1.39303923e+00 1.13040721e+00 7.62707114e-01 3.34402025e-02
-3.58043564e-03 3.18456739e-01 8.67233872e-01 6.43573582e-01
2.19541624e-01 3.75299215e-01 9.50144231e-01 3.69147003e-01
4.31046575e-01 7.13463053e-02 4.80896324e-01 2.28558183e-01
4.19027001e-01 -6.12162828e-01 -8.91904831e-01 3.68748695e-01
-1.27733779e+00 -1.21118605e+00 -5.83117604e-01 2.62519145e+00
2.29860097e-02 6.20757341e-01 1.53652430e-01 5.88830769e-01
1.01882493e+00 -3.75909626e-01 -4.72016007e-01 -7.42876709e-01
7.79049993e-02 -2.40436599e-01 8.10228348e-01 6.13550982e-03
-7.15388775e-01 4.17033613e-01 5.52751541e+00 8.43608618e-01
-1.29311931e+00 -1.09798908e-02 6.87347353e-01 3.10636371e-01
1.35751903e-01 1.69866234e-01 -7.10796237e-01 9.09448326e-01
1.64318240e+00 -5.37469268e-01 2.77394950e-02 9.08506930e-01
5.19147217e-01 -2.63446748e-01 -3.97358924e-01 5.36058307e-01
9.46902782e-02 -8.64921570e-01 -9.18930471e-02 8.38529617e-02
4.32713419e-01 -6.94850013e-02 -1.53407529e-01 3.84160489e-01
2.94951439e-01 -5.51937640e-01 3.57477367e-01 2.08466515e-01
2.49527484e-01 -1.07099545e+00 1.15761864e+00 4.79427874e-01
-1.14321744e+00 -3.28405619e-01 -5.01489602e-02 1.48104489e-01
3.72691035e-01 4.82320726e-01 -6.18163228e-01 4.52250123e-01
3.54822665e-01 -1.87448025e-01 -2.12665424e-01 1.24760473e+00
1.17330924e-01 8.83106470e-01 -6.00568831e-01 -4.87917989e-01
1.01311207e-01 2.38509864e-01 9.33822691e-01 6.92899227e-01
2.62106031e-01 2.81612068e-01 2.32389241e-01 5.11310101e-01
6.01449072e-01 -2.11737696e-02 -5.26400268e-01 8.58273432e-02
1.08275437e+00 1.02888227e+00 -6.49405837e-01 -6.75627068e-02
-3.32591534e-01 9.73726902e-03 -5.56030929e-01 5.22720337e-01
-1.20492458e+00 -7.50620663e-01 8.36991787e-01 4.16164935e-01
-1.61351264e-01 -5.03338039e-01 -2.80600041e-01 -4.74455208e-01
-5.14191091e-02 -7.29006231e-01 3.89490664e-01 -3.49079639e-01
-9.20114279e-01 6.01722479e-01 2.97691464e-01 -1.43255949e+00
-2.13816479e-01 -3.59430432e-01 -1.21140933e+00 6.26226306e-01
-1.16763186e+00 -5.56595087e-01 -3.28121573e-01 7.53590107e-01
1.30647659e-01 -4.25569922e-01 1.86313897e-01 5.06176174e-01
-7.47682750e-01 8.22278023e-01 8.65296572e-02 8.52108151e-02
-8.20126906e-02 -2.44904757e-01 1.40106916e-01 1.19895279e+00
-3.92327219e-01 2.95277357e-01 8.81951392e-01 -9.08226490e-01
-1.56536102e+00 -1.00553060e+00 4.51277673e-01 -1.00388549e-01
5.87027073e-01 -1.00530073e-01 -8.14259052e-01 2.79251903e-01
-9.60738361e-02 -1.59017831e-01 4.17492867e-01 -3.84119570e-01
-8.40502530e-02 -9.30548236e-02 -1.58457136e+00 6.57465398e-01
2.77482808e-01 8.80888328e-02 8.21965337e-02 1.28803551e-01
5.80008447e-01 7.50412990e-04 -4.83839571e-01 6.49604857e-01
2.20619485e-01 -7.52967060e-01 7.72713184e-01 -4.86212969e-01
-5.60423672e-01 -1.94115058e-01 -1.86277151e-01 -1.03612471e+00
8.79189968e-02 -4.45746124e-01 -8.97869617e-02 1.18982911e+00
5.01837194e-01 -1.39972317e+00 2.57227898e-01 6.69302523e-01
-2.56121188e-01 -6.06344879e-01 -1.22058558e+00 -1.10192215e+00
-3.24514538e-01 -8.09928834e-01 6.77510560e-01 8.01340342e-01
1.19586907e-01 -6.19849376e-02 -1.57195836e-01 8.71678770e-01
6.95066452e-01 -4.58707035e-01 8.66679192e-01 -1.13179839e+00
3.15884680e-01 -3.12576234e-01 -1.06452847e+00 -9.56824888e-03
-2.45487958e-01 -8.19015503e-02 -3.21080863e-01 -1.02163839e+00
-1.34577319e-01 -4.53675479e-01 -4.59985316e-01 2.06482708e-01
5.15494235e-02 -2.64316142e-01 -9.63331666e-03 1.42630637e-01
-3.98259668e-04 4.32805836e-01 1.99885309e-01 -1.57596301e-02
-1.81295604e-01 4.57778811e-01 -1.09785087e-01 5.10318339e-01
9.70281184e-01 -6.49109662e-01 -6.96897805e-01 2.49427587e-01
-6.56413883e-02 4.09450352e-01 5.31746566e-01 -1.12608147e+00
3.40416133e-01 -4.77994025e-01 -8.96730423e-02 -8.19127321e-01
-6.36899918e-02 -8.55732322e-01 2.77398318e-01 8.94967139e-01
3.94473642e-01 4.00539577e-01 6.08596444e-01 7.53546715e-01
-2.60962725e-01 -4.16349731e-02 7.69584358e-01 4.72550660e-01
-5.95594347e-01 5.39635837e-01 -9.27775860e-01 -2.41603017e-01
1.88837218e+00 -5.21527946e-01 -5.61729014e-01 -5.04972279e-01
-2.19726875e-01 3.61589313e-01 -1.33183941e-01 5.95146060e-01
8.62007916e-01 -9.97693837e-01 -4.42091882e-01 3.93227756e-01
-1.06911466e-01 -6.64504528e-01 5.16343772e-01 9.49808836e-01
-3.17776799e-01 4.65073049e-01 -1.80543989e-01 -3.47834349e-01
-1.35266960e+00 7.68162429e-01 1.54756591e-01 1.93900466e-01
-1.47274718e-01 9.35359448e-02 -2.12906525e-01 1.58057466e-01
1.83049008e-01 1.60180435e-01 -1.84721917e-01 -3.58816504e-01
4.80324239e-01 8.15933704e-01 -1.25617776e-02 -7.11755276e-01
-5.76257288e-01 1.38477579e-01 -3.46997716e-02 2.26458460e-01
3.84109974e-01 -1.45144761e-01 1.64730713e-01 6.27259240e-02
7.77752280e-01 -9.76130962e-02 -6.69989049e-01 1.03229970e-01
8.06634799e-02 -4.83613789e-01 -8.16867501e-02 -6.30398035e-01
-1.03728592e+00 5.18787503e-01 8.04158866e-01 8.42113793e-01
9.64033723e-01 -3.39659929e-01 9.55029547e-01 -1.18590504e-01
5.71502030e-01 -7.73490906e-01 -1.39757782e-01 3.38768095e-01
2.21347794e-01 -9.23175097e-01 -4.49960470e-01 -4.85733390e-01
-4.45496768e-01 7.53647625e-01 9.55302596e-01 -2.20280975e-01
1.18242586e+00 4.14279699e-01 -1.18352629e-01 -1.22826666e-01
-1.03383064e+00 1.99794576e-01 -1.47361681e-01 6.67208672e-01
-3.04850370e-01 1.30934089e-01 -5.61777115e-01 5.72447240e-01
2.28311718e-01 -3.61640215e-01 9.97158587e-01 1.05168426e+00
-6.32187605e-01 -4.71365333e-01 -4.02556479e-01 5.10680437e-01
-3.73916060e-01 2.45305553e-01 -9.69177708e-02 9.04532790e-01
-2.66516685e-01 1.68478990e+00 1.78936958e-01 -8.31953108e-01
4.62483972e-01 -1.60821825e-01 -4.19239730e-01 1.21854052e-01
-2.37054238e-03 -4.84856308e-01 5.88288680e-02 -3.27944905e-01
3.77035558e-01 -4.95492309e-01 -1.37219357e+00 -8.36446524e-01
-6.86443686e-01 5.01449525e-01 1.18455005e+00 1.16036606e+00
6.05478585e-01 3.78957748e-01 1.17836618e+00 1.00740336e-01
-9.54149485e-01 -9.05245662e-01 -4.93684977e-01 2.03336895e-01
-3.27432640e-02 -9.19604242e-01 -9.05248940e-01 -7.91302800e-01] | [5.306484699249268, 7.2557148933410645] |
7f07692f-11dc-410c-a579-5ad26a8942d2 | a-hybrid-supervisedunsupervised-machine | 1706.07103 | null | http://arxiv.org/abs/1706.07103v1 | http://arxiv.org/pdf/1706.07103v1.pdf | A hybrid supervised/unsupervised machine learning approach to solar flare prediction | We introduce a hybrid approach to solar flare prediction, whereby a
supervised regularization method is used to realize feature importance and an
unsupervised clustering method is used to realize the binary flare/no-flare
decision. The approach is validated against NOAA SWPC data. | ['Federico Benvenuto', 'Cristina Campi', 'Michele Piana', 'Anna Maria Massone'] | 2017-06-21 | null | null | null | null | ['solar-flare-prediction'] | ['time-series'] | [ 3.25198293e-01 -4.18940246e-01 -1.53733164e-01 -7.08411634e-01
-4.34648067e-01 -5.21727264e-01 3.84574234e-01 -8.12726468e-02
1.90091133e-01 9.88905549e-01 4.95628417e-02 -2.67674532e-02
-6.73401773e-01 -6.44916236e-01 5.61242960e-02 -1.01388514e+00
3.51192743e-01 2.62240261e-01 1.01166293e-01 -3.26677233e-01
4.21982855e-01 8.07170510e-01 -1.93029380e+00 -8.01615715e-02
1.33575928e+00 1.16967630e+00 -3.38436030e-02 8.00522804e-01
-5.16545027e-02 9.58954155e-01 -5.60445189e-01 3.87427092e-01
3.62400234e-01 -8.39176714e-01 -7.74382129e-02 4.18715864e-01
5.11968136e-01 4.42750871e-01 2.24131882e-01 1.03270662e+00
4.59909946e-01 3.15319955e-01 1.09031224e+00 -9.03243840e-01
6.14976361e-02 -7.53270909e-02 -4.40724552e-01 5.10952532e-01
3.60343039e-01 1.41649753e-01 9.26744640e-01 -4.97464538e-01
4.10611510e-01 6.29793882e-01 8.05013478e-01 -1.69698626e-01
-1.37298524e+00 -3.92674416e-01 -3.73161346e-01 5.46323098e-02
-1.36791348e+00 -6.25248671e-01 8.90920937e-01 -6.47815585e-01
1.32848549e+00 3.36428761e-01 8.59652877e-01 -9.29066017e-02
5.87334812e-01 2.34091610e-01 1.36421001e+00 -5.41170418e-01
5.21346033e-01 -6.57669976e-02 6.37523115e-01 4.91184860e-01
5.93704760e-01 8.15252125e-01 -3.20942730e-01 -3.69417429e-01
2.59307414e-01 -2.54076362e-01 -3.82205278e-01 -7.30654597e-02
7.95200467e-02 1.24668920e+00 3.05052191e-01 4.89464581e-01
-2.90464759e-01 -2.75824100e-01 -1.93077505e-01 6.88044667e-01
2.64429957e-01 5.63140452e-01 -4.36157078e-01 3.90990913e-01
-1.52258301e+00 3.12973797e-01 6.90200508e-01 1.94993168e-01
1.16520071e+00 5.58929861e-01 1.18097566e-01 7.30445325e-01
5.74000597e-01 1.17978930e+00 3.14913422e-01 -1.03647304e+00
-3.09313834e-01 7.73681819e-01 1.69638216e-01 -1.11022067e+00
-4.92823035e-01 -1.35983691e-01 -6.92357957e-01 8.27328622e-01
-2.03090519e-01 -4.53843266e-01 -1.37965643e+00 9.03013885e-01
-1.11301772e-01 -8.60456675e-02 5.15767753e-01 7.66727805e-01
7.99830794e-01 5.32138348e-01 2.08039936e-02 -9.14546967e-01
5.10861337e-01 -5.77652693e-01 -1.05927396e+00 -3.54843169e-01
6.08687460e-01 -5.77546716e-01 4.06322747e-01 5.53945065e-01
-8.33237827e-01 -5.44684231e-01 -1.11094022e+00 2.46687025e-01
-6.05160177e-01 8.61838013e-02 7.93582201e-01 6.97535098e-01
-8.25630724e-01 3.91862810e-01 -6.95744932e-01 -4.62721586e-01
2.30607912e-01 3.12990844e-01 1.83789916e-02 5.58201551e-01
-7.13760376e-01 8.09656918e-01 5.11438370e-01 2.16356024e-01
-1.41695440e-01 -3.80076766e-01 -6.37027144e-01 1.23201367e-02
-3.42121184e-01 -5.02493501e-01 6.42939210e-01 -1.30431724e+00
-1.44948697e+00 6.43634677e-01 -2.77043700e-01 -6.98367298e-01
-2.00866669e-01 2.91496903e-01 -9.83576834e-01 2.74579108e-01
-3.01834822e-01 1.33365914e-01 1.08856702e+00 -1.43029678e+00
-9.39464927e-01 -2.22792968e-01 -6.66579902e-01 3.24587971e-01
4.63663071e-01 -9.54873487e-02 4.38433558e-01 -2.67225116e-01
3.86842072e-01 -6.50535524e-01 -2.23062649e-01 -4.08986956e-01
8.53633974e-03 -1.68541417e-01 1.00606000e+00 -5.15922964e-01
1.61804223e+00 -1.83856285e+00 1.02808416e-01 8.09927642e-01
-1.45624697e-01 3.59051436e-01 5.54346025e-01 1.79642573e-01
-2.19327182e-01 -2.78100908e-01 -1.00341487e+00 1.86976030e-01
-1.86668709e-01 6.85500801e-01 -2.07002684e-01 3.08663040e-01
3.53110969e-01 5.68307638e-01 -4.34568882e-01 -1.65676624e-01
4.34685647e-01 1.30047396e-01 -6.10775590e-01 2.12175965e-01
-3.05974483e-01 2.18629941e-01 -4.07702416e-01 7.62310207e-01
9.28607047e-01 8.85333344e-02 1.97608918e-01 3.47136483e-02
-6.32335305e-01 -1.39623746e-01 -1.08540797e+00 1.12301195e+00
1.60992235e-01 4.53518748e-01 1.89316288e-01 -8.57222140e-01
1.37804997e+00 -6.03385307e-02 2.27822080e-01 -6.60837352e-01
2.92112410e-01 2.79806852e-01 -7.70021230e-02 -7.59399533e-01
2.44846061e-01 -6.74143970e-01 5.00974357e-02 -2.82906312e-02
-3.02669848e-03 -9.19117808e-01 3.90603185e-01 -2.81312436e-01
1.12085319e+00 1.73974633e-01 4.11959738e-01 -9.20498371e-01
6.59601927e-01 4.67618644e-01 7.84879029e-01 4.37127978e-01
-8.06683078e-02 5.31474590e-01 3.53699505e-01 -3.90415758e-01
-1.56383127e-01 -8.39249909e-01 -5.16877234e-01 5.84108591e-01
3.96673195e-02 -8.77384096e-02 -3.32292467e-01 -4.75126237e-01
2.98489094e-01 9.92900491e-01 -4.01171297e-01 -3.75616760e-03
-8.22948199e-03 -1.25387239e+00 -2.25200094e-02 3.94232303e-01
4.37570363e-01 -1.30392385e+00 -6.27135694e-01 4.41019423e-02
1.54752940e-01 -1.23951003e-01 1.95062965e-01 8.44298959e-01
-8.68333161e-01 -1.05954719e+00 2.48730987e-01 -7.67287612e-01
5.00981092e-01 2.12454405e-02 1.06315625e+00 2.59689897e-01
-6.51849627e-01 9.67562348e-02 -3.57257545e-01 -4.28161412e-01
-6.12813383e-02 -4.15082961e-01 -1.17414579e-01 -1.23367712e-01
8.92740309e-01 -7.69129157e-01 -3.64012331e-01 -1.68003291e-01
-9.26597953e-01 -9.76791322e-01 2.03776032e-01 5.19780815e-01
7.75126219e-01 4.50125813e-01 6.11098409e-01 -9.48026359e-01
4.11203474e-01 -3.70679200e-01 -8.70683491e-01 1.50644451e-01
-8.81792545e-01 -4.84006293e-02 1.29986957e-01 -7.16782361e-02
-1.44107211e+00 4.87436533e-01 -2.29175702e-01 -2.19750911e-01
-3.95193547e-01 8.42314303e-01 -2.45646518e-02 -6.40878737e-01
9.42281425e-01 -7.77870491e-02 -3.70654881e-01 -4.53371286e-01
4.46723253e-02 6.75991118e-01 5.67277849e-01 1.50016874e-01
9.96542275e-01 3.26643854e-01 6.87063485e-02 -1.24765265e+00
-1.05960429e+00 -6.83689237e-01 -7.80318379e-01 -3.59399438e-01
9.11030710e-01 -7.42336571e-01 -1.53454930e-01 5.44991255e-01
-4.77002740e-01 -1.76022485e-01 -3.84565741e-01 5.96931159e-01
-6.30358040e-01 1.66420549e-01 9.53595564e-02 -1.13812435e+00
-3.64286393e-01 -4.12704676e-01 6.67202890e-01 8.60969901e-01
9.02131796e-02 -1.12377679e+00 7.80964315e-01 2.22058803e-01
2.65440166e-01 4.86055762e-01 7.35208690e-01 -3.54552716e-01
-1.14299476e-01 -9.44526270e-02 -1.55580953e-01 4.50099945e-01
3.48813355e-01 4.33143109e-01 -8.43368471e-01 -1.19234718e-01
1.99145615e-01 -5.69191098e-01 1.14575338e+00 6.81847036e-01
5.26996195e-01 2.12786168e-01 -1.58539787e-01 7.83986807e-01
2.06694746e+00 6.22693717e-01 8.72423768e-01 -1.35824442e-01
2.65648931e-01 7.62359872e-02 7.38467813e-01 2.60924429e-01
-1.07097872e-01 4.71738018e-02 1.40436724e-01 -2.10178778e-01
-6.02241382e-02 1.03778370e-01 7.92359933e-02 7.48244643e-01
-1.07498050e-01 -2.05811784e-01 -6.11703575e-01 3.76300544e-01
-1.77682853e+00 -1.13897800e+00 -5.55333793e-01 2.07754970e+00
3.17679912e-01 1.23245031e-01 -2.24304408e-01 4.56197441e-01
4.05884981e-01 4.20328349e-01 -2.63571590e-01 -6.00130677e-01
-6.23564720e-01 4.72494721e-01 4.13242072e-01 8.94410968e-01
-1.05061352e+00 8.79615247e-01 8.33013821e+00 5.26250541e-01
-1.06829405e+00 -1.97440103e-01 2.34687217e-02 1.91001236e-01
-3.01234573e-01 1.71034217e-01 -6.88299060e-01 2.13097185e-01
6.89927280e-01 -1.95042219e-03 4.53569591e-01 3.69800419e-01
2.85369903e-01 -8.21984231e-01 -1.72103271e-01 7.32533932e-01
5.31876087e-01 -8.95702243e-01 -2.58273423e-01 -2.76798040e-01
9.21164930e-01 -1.19891977e-02 -6.76085889e-01 2.95058548e-01
2.98753738e-01 -7.30403304e-01 -7.37026334e-02 1.54553890e+00
4.00242537e-01 -9.51911926e-01 5.29188514e-01 1.36655450e-01
-1.29364634e+00 -2.17715159e-01 -6.42070711e-01 -3.03509533e-01
3.11850220e-01 9.04058695e-01 -2.38100186e-01 6.36888564e-01
7.68683970e-01 8.28246891e-01 -7.48379588e-01 1.29556775e+00
-3.02099288e-01 5.99472344e-01 -8.93458128e-01 3.46417397e-01
4.53439094e-02 -8.44621718e-01 7.29779184e-01 1.01364923e+00
4.21007782e-01 3.91319335e-01 1.40430480e-01 5.15187025e-01
5.99801302e-01 1.02178834e-01 -7.02872097e-01 -1.22766405e-01
-5.71279526e-02 1.39393580e+00 -7.78037548e-01 -3.65485810e-02
-2.30205223e-01 5.89743435e-01 -5.38710393e-02 3.20743978e-01
-4.27074492e-01 -7.04254568e-01 4.72837836e-01 1.13727883e-01
7.20426381e-01 1.24453403e-01 -2.89847940e-01 -1.12383330e+00
-4.72987503e-01 -3.58168155e-01 6.25324428e-01 -1.36300564e+00
-1.34977865e+00 4.62580860e-01 -2.20137239e-01 -1.08803868e+00
-2.47537524e-01 -5.34550667e-01 -1.24500513e+00 7.72946000e-01
-1.79818475e+00 -6.78431034e-01 -3.09670061e-01 4.56842929e-01
8.81223232e-02 -4.50496167e-01 8.48165214e-01 1.55753389e-01
-2.99386501e-01 -1.21326305e-01 2.44240522e-01 -5.53443730e-01
5.12080610e-01 -1.25397837e+00 -7.56304383e-01 8.54422390e-01
-2.78477043e-01 -2.01295301e-01 9.60704625e-01 -1.13549948e+00
-1.01900196e+00 -1.10403669e+00 1.31886113e+00 -1.77662089e-01
7.60290682e-01 5.00749290e-01 -1.05780864e+00 5.65981269e-01
2.86530018e-01 -1.14105158e-01 6.98750734e-01 7.48236626e-02
7.74540240e-03 -2.95957625e-01 -1.48625374e+00 -6.53354600e-02
5.68328500e-01 -8.44642520e-02 -1.24316740e+00 3.73803318e-01
-2.04375875e-03 8.76507089e-02 -1.09582853e+00 1.17818511e+00
2.53395855e-01 -1.33154798e+00 5.27886450e-01 -2.63323963e-01
-3.71451750e-02 -7.80741513e-01 -2.79203385e-01 -1.20435321e+00
-4.78546649e-01 -5.79664469e-01 8.80702361e-02 1.09132624e+00
5.78733385e-01 -4.93101627e-01 6.44488275e-01 -9.36913565e-02
-1.18013382e-01 -5.76366633e-02 -7.95280397e-01 -3.45359653e-01
-1.74846724e-01 3.35176140e-02 -2.65288875e-02 9.45243776e-01
-2.17257708e-01 4.07460481e-01 -1.69949621e-01 3.92208248e-01
5.73505938e-01 5.56108594e-01 2.14458585e-01 -1.78490043e+00
-1.80856556e-01 -4.34940428e-01 -4.01421219e-01 -2.82162070e-01
9.78296101e-02 -8.16182733e-01 4.86960679e-01 -1.38760686e+00
-2.10090548e-01 -1.25017926e-01 -3.04711431e-01 3.83922070e-01
3.83818359e-03 5.98969936e-01 -1.74894467e-01 9.22524184e-02
-2.49765500e-01 4.12557006e-01 4.39586073e-01 2.25089267e-01
-8.33961010e-01 2.11483642e-01 -3.98418844e-01 9.48092103e-01
9.35343266e-01 -3.93527716e-01 -2.90916741e-01 3.02938633e-02
2.28945121e-01 3.21333438e-01 1.04746334e-01 -1.49451184e+00
1.94491416e-01 -3.90682131e-01 4.17446405e-01 -1.01062071e+00
1.89561963e-01 -8.67924631e-01 4.23108965e-01 3.63488019e-01
1.95591390e-01 -5.47982454e-01 1.08470418e-01 4.36158478e-01
-3.91959220e-01 -5.79375386e-01 1.15360796e+00 2.70010173e-01
-4.17520106e-01 -1.72474265e-01 -7.72562504e-01 -4.86169249e-01
7.67461419e-01 -1.62832335e-01 -3.20099354e-01 -1.45456284e-01
-1.18724501e+00 5.11069357e-01 4.63159263e-01 -8.91800970e-02
2.92614371e-01 -1.26943302e+00 -5.75737357e-01 7.42905438e-01
2.27749899e-01 -6.27524614e-01 8.67381841e-02 8.68722618e-01
-6.28896058e-01 2.69287825e-01 -2.86888152e-01 -5.22046268e-01
-1.36373782e+00 2.68790811e-01 5.51371753e-01 -8.71592760e-02
-3.51304412e-01 4.40306723e-01 -3.18275303e-01 -3.16071570e-01
-1.73285976e-02 1.24292197e-02 -7.34055102e-01 4.52165633e-01
5.51172376e-01 2.77558297e-01 2.55252272e-01 -8.58460605e-01
-1.79317966e-01 1.13087642e+00 6.80303931e-01 3.65459234e-01
1.19240487e+00 -2.78521121e-01 -2.56187350e-01 7.24285603e-01
6.45394325e-01 1.17086589e-01 -1.08990479e+00 1.26458243e-01
1.38879001e-01 -3.73431891e-02 7.21609652e-01 -1.23726070e+00
-9.83135760e-01 4.57560271e-01 9.08733666e-01 5.41584194e-01
1.79836237e+00 -3.24255347e-01 1.87282920e-01 8.40085328e-01
-6.99125081e-02 -1.27521074e+00 -7.29724526e-01 6.34887576e-01
8.26024711e-01 -1.04422271e+00 1.92881122e-01 -4.61410522e-01
-8.56090426e-01 1.02047312e+00 1.13232359e-01 -6.14035189e-01
1.24810970e+00 7.12772310e-01 3.30277771e-01 -3.50014776e-01
-7.64091492e-01 -1.00104809e+00 3.54603499e-01 3.93972665e-01
2.95739472e-02 -2.95388121e-02 -9.24822807e-01 5.84246814e-01
-7.53643289e-02 2.53677636e-01 1.44369319e-01 1.16498530e+00
-1.28821599e+00 -8.17305505e-01 -4.90034372e-01 1.11142468e+00
-2.00847927e-02 1.32031262e-01 -7.15524554e-01 5.76461315e-01
4.09962267e-01 1.19004536e+00 3.77681226e-01 -1.61176294e-01
2.56029516e-01 5.31552970e-01 4.05090302e-01 -4.69052970e-01
-8.74099314e-01 7.27434099e-01 2.30690390e-01 -3.05366963e-01
-1.16137195e+00 -6.27041221e-01 -1.24511373e+00 9.10053104e-02
-4.11595911e-01 6.34274244e-01 4.02872920e-01 1.03650820e+00
-4.00277935e-02 2.39289150e-01 1.06980669e+00 -9.47105408e-01
2.43378848e-01 -9.03541923e-01 -1.14961183e+00 2.82423710e-03
2.06230670e-01 -7.55458117e-01 -1.32521212e+00 2.81996369e-01] | [6.614551544189453, 2.784780979156494] |
8e77d7ca-dc64-4fbb-84cc-156651cb1c5e | uncovering-the-hidden-dynamics-of-video-self | 2306.02014 | null | https://arxiv.org/abs/2306.02014v1 | https://arxiv.org/pdf/2306.02014v1.pdf | Uncovering the Hidden Dynamics of Video Self-supervised Learning under Distribution Shifts | Video self-supervised learning (VSSL) has made significant progress in recent years. However, the exact behavior and dynamics of these models under different forms of distribution shift are not yet known. In this paper, we comprehensively study the behavior of six popular self-supervised methods (v-SimCLR, v-MOCO, v-BYOL, v-SimSiam, v-DINO, v-MAE) in response to various forms of natural distribution shift, i.e., (i) context shift, (ii) viewpoint shift, (iii) actor shift, (iv) source shift, (v) generalizability to unknown classes (zero-shot), and (vi) open-set recognition. To perform this extensive study, we carefully craft a test bed consisting of $17$ in-distribution and out-of-distribution benchmark pairs using available public datasets and a series of evaluation protocols to stress-test the different methods under the intended shifts. Our study uncovers a series of intriguing findings and interesting behaviors of VSSL methods. For instance, we observe that while video models generally struggle with context shifts, v-MAE and supervised learning exhibit more robustness. Moreover, our study shows that v-MAE is a strong temporal learner, whereas contrastive methods, v-SimCLR and v-MOCO, exhibit strong performances against viewpoint shifts. When studying the notion of open-set recognition, we notice a trade-off between closed-set and open-set recognition performance, particularly if the pretrained VSSL encoders are used without finetuning. We hope that our work will contribute to the development of robust video representation learning frameworks for various real-world scenarios. | ['Ali Etemad', 'Ahmad Beirami', 'Pritam Sarkar'] | 2023-06-03 | null | null | null | null | ['open-set-learning'] | ['miscellaneous'] | [ 2.54809856e-01 -4.57364410e-01 -5.00024259e-01 -2.84737140e-01
-5.93859136e-01 -8.50798607e-01 6.22104168e-01 -6.21245280e-02
-3.07367891e-01 6.72334909e-01 4.34188172e-02 -3.14882487e-01
-5.29963821e-02 -4.92131680e-01 -1.03448951e+00 -8.25950682e-01
-2.95040369e-01 3.29830498e-01 4.95958537e-01 -3.34629208e-01
1.89023003e-01 2.60678113e-01 -1.69222772e+00 3.54473054e-01
7.10069954e-01 1.08518887e+00 -1.41513765e-01 5.48591614e-01
1.28756300e-01 9.29583788e-01 -4.53587770e-01 -5.65118194e-01
6.23434424e-01 -4.44440275e-01 -5.82106292e-01 4.35852319e-01
8.50185812e-01 -2.11494446e-01 -7.42473364e-01 1.10016322e+00
3.60070437e-01 2.13131070e-01 7.52288222e-01 -1.50845027e+00
-6.44636393e-01 6.38754725e-01 -8.21130276e-01 6.33947194e-01
4.16761696e-01 6.10864699e-01 8.79998267e-01 -7.80880809e-01
8.41022730e-01 1.19841230e+00 6.85807168e-01 6.54058576e-01
-1.25627232e+00 -8.47871900e-01 2.98853129e-01 2.72026569e-01
-1.43756413e+00 -5.29847562e-01 6.33992195e-01 -3.96812677e-01
6.88420236e-01 1.70501217e-01 3.16865861e-01 1.68319154e+00
1.13261836e-02 9.67519224e-01 1.37099051e+00 -3.59838933e-01
4.56185699e-01 3.39400113e-01 3.28418970e-01 6.44713640e-01
3.60571355e-01 1.82925314e-01 -8.15522194e-01 -2.31441990e-01
6.25006437e-01 9.93242860e-02 -4.09577012e-01 -6.38873339e-01
-1.10562980e+00 6.21917307e-01 1.22156583e-01 2.81342745e-01
1.18220948e-01 9.79967490e-02 6.40414655e-01 7.43804514e-01
3.91838491e-01 2.08062306e-01 -5.04507899e-01 -3.95796150e-01
-8.37921202e-01 2.18016312e-01 7.60002315e-01 1.28940833e+00
7.72170722e-01 2.18218893e-01 -2.12283432e-01 7.28841543e-01
-1.03998862e-01 4.60106403e-01 8.48186791e-01 -8.48132908e-01
6.11593723e-01 2.99464166e-01 -1.32105380e-01 -8.62674057e-01
-2.04005256e-01 -4.18023169e-01 -9.40451264e-01 -7.73197366e-03
6.76253974e-01 -5.95061742e-02 -7.55607486e-01 1.90173793e+00
-6.08969256e-02 4.38256860e-01 2.31815115e-01 6.67882085e-01
7.91684568e-01 4.16005701e-01 -4.02989835e-02 -3.57083082e-01
1.15224457e+00 -1.04167449e+00 -4.11476195e-01 -5.60832918e-02
7.86535800e-01 -3.94353092e-01 1.24371552e+00 3.61159205e-01
-8.91669750e-01 -6.83633864e-01 -9.97691095e-01 4.08082217e-01
-2.45995075e-01 -2.16778532e-01 5.45562565e-01 6.24522746e-01
-1.03021896e+00 5.49142718e-01 -6.74002171e-01 -7.38365233e-01
5.93415201e-01 2.60266364e-01 -6.05174601e-01 -2.65810579e-01
-9.27756727e-01 3.93285692e-01 2.59191692e-01 -4.05650944e-01
-1.19815373e+00 -6.68735683e-01 -6.28441095e-01 3.27274809e-03
7.26477206e-01 -4.91729051e-01 1.11675417e+00 -1.55989075e+00
-1.27811670e+00 1.01837564e+00 -4.57240902e-02 -5.24250448e-01
7.63560832e-01 -1.71836033e-01 -5.48705637e-01 2.03149915e-01
1.02292355e-02 6.76514626e-01 1.02149093e+00 -1.36155736e+00
-4.96021599e-01 -3.61015528e-01 9.39984992e-02 1.52845621e-01
-5.13146698e-01 -8.24006796e-02 -3.06082994e-01 -8.52919638e-01
-1.43398076e-01 -1.13006282e+00 2.22938731e-01 -1.71185255e-01
-5.00527859e-01 -2.92841196e-01 9.82307911e-01 -1.06411949e-01
1.45603395e+00 -2.52174902e+00 6.16030954e-02 3.91539969e-02
1.37639120e-01 2.63930112e-01 -3.72346014e-01 5.71215510e-01
-3.61005187e-01 1.53508708e-01 -1.71825871e-01 -2.12197810e-01
-7.45231956e-02 3.12389344e-01 -5.10596395e-01 5.05629599e-01
-7.51077905e-02 7.11271167e-01 -8.98243010e-01 -4.06817108e-01
-1.87661409e-01 4.03391831e-02 -7.61508346e-01 2.31583685e-01
-1.98164195e-01 2.98992485e-01 6.20127022e-02 9.30959165e-01
5.20407736e-01 -4.62318629e-01 1.71035364e-01 -2.22486660e-01
2.03624800e-01 -2.44249806e-01 -1.14749694e+00 1.38297236e+00
3.61890942e-02 6.67670071e-01 -3.30586165e-01 -1.05670846e+00
6.01397753e-01 6.25970513e-02 4.47438657e-01 -7.03852713e-01
5.50830662e-02 2.12862059e-01 5.90267032e-02 -5.00915766e-01
5.18469036e-01 -3.23894769e-02 1.34837344e-01 4.29975808e-01
4.59883481e-01 4.04750168e-01 3.37992847e-01 4.87825453e-01
1.34193826e+00 2.92466599e-02 3.36135000e-01 -2.88977087e-01
2.99527198e-01 -2.88275987e-01 5.86079776e-01 1.06692815e+00
-5.34513175e-01 6.60699844e-01 7.43345082e-01 -4.47416872e-01
-8.68017673e-01 -1.08273363e+00 -8.16820264e-02 1.41183412e+00
3.29111308e-01 -4.08810437e-01 -7.93362081e-01 -9.77143288e-01
1.33892864e-01 6.63080871e-01 -5.76077044e-01 -4.23151463e-01
-4.65210497e-01 -6.46386206e-01 6.82151318e-01 5.95881522e-01
6.97419524e-01 -8.31957161e-01 -5.11619270e-01 -2.77953029e-01
2.11595604e-03 -1.24633861e+00 -8.54115546e-01 2.02539250e-01
-7.49234080e-01 -1.09375525e+00 -6.03595793e-01 -6.75132573e-01
6.13595188e-01 6.31353796e-01 1.15262914e+00 -6.60806745e-02
8.77717957e-02 7.58175731e-01 -5.94476700e-01 -1.72428682e-01
-3.70701104e-01 1.60237834e-01 5.03370523e-01 2.89019674e-01
3.75981331e-01 -5.91213286e-01 -4.65725362e-01 7.88082302e-01
-1.19595695e+00 -3.06425095e-01 3.93443674e-01 9.81632650e-01
4.94928420e-01 4.30433936e-02 5.08304536e-01 -1.11423934e+00
3.74444991e-01 -9.05277550e-01 -5.09526253e-01 3.41656387e-01
-7.32661307e-01 -1.63715240e-02 5.84311783e-01 -8.63282919e-01
-9.14763629e-01 -6.96845278e-02 1.84087381e-01 -9.97088194e-01
-1.72334194e-01 2.92603701e-01 -2.09226683e-01 -1.68478061e-02
9.36854243e-01 5.63427687e-01 -8.39309022e-02 -2.26569191e-01
2.64299452e-01 7.37480402e-01 5.37791252e-01 -6.87922120e-01
9.97290134e-01 5.95785141e-01 -3.47249836e-01 -9.58011270e-01
-8.56396377e-01 -4.65909421e-01 -5.35948992e-01 -2.01346293e-01
4.98381704e-01 -1.18531072e+00 -3.74881744e-01 6.01412952e-01
-6.53591275e-01 -6.51328325e-01 -3.55761737e-01 1.02918684e-01
-5.53912818e-01 6.77799165e-01 -3.58221978e-01 -5.44003069e-01
-1.83459502e-02 -1.18932509e+00 8.91375184e-01 2.59434551e-01
-1.79204136e-01 -1.10086834e+00 1.41085684e-02 3.78671587e-01
3.12789947e-01 -1.34373149e-02 7.88628221e-01 -1.07501376e+00
-5.28179646e-01 9.48148146e-02 -1.28507614e-01 4.92796361e-01
8.58112201e-02 7.60230497e-02 -8.81218374e-01 -8.00817311e-01
-3.05366069e-01 -7.46234357e-01 9.59305227e-01 1.33960456e-01
1.34577239e+00 -4.74125952e-01 -2.26922691e-01 9.45026040e-01
1.36204338e+00 2.26146251e-01 7.54933655e-01 3.79724801e-01
5.34560442e-01 5.76381832e-02 3.88517648e-01 5.05438149e-01
4.23882723e-01 6.94855452e-01 3.75809848e-01 1.46598607e-01
-1.08179599e-01 -4.21818525e-01 8.36253285e-01 6.54459476e-01
-8.16477165e-02 -5.57939827e-01 -6.91919565e-01 3.16007227e-01
-1.80211604e+00 -1.18008554e+00 2.76300639e-01 2.32324553e+00
7.75775254e-01 2.55629003e-01 5.39597273e-01 8.79150704e-02
6.30438626e-01 4.33295846e-01 -7.33231008e-01 -6.01464175e-02
-5.17135918e-01 -1.27201136e-02 6.42252624e-01 -2.16265470e-01
-1.28398073e+00 1.02455473e+00 6.40338421e+00 1.10576880e+00
-1.22546339e+00 1.36404291e-01 7.47617185e-01 -1.34940714e-01
-1.26206353e-01 3.52088623e-02 -8.82002234e-01 7.25662649e-01
7.90716887e-01 -1.99038219e-02 5.39836466e-01 1.05152082e+00
-3.95072669e-01 3.16420756e-02 -1.47457218e+00 1.41374934e+00
4.57550645e-01 -1.31501532e+00 2.52379596e-01 -1.15493841e-01
9.04944241e-01 2.75521964e-01 1.14110053e-01 7.40560353e-01
4.03268933e-01 -7.82327235e-01 7.24454045e-01 2.62767226e-01
1.10295296e+00 -3.98263961e-01 5.22912085e-01 2.75173664e-01
-8.65092635e-01 -4.10884887e-01 -3.25244814e-01 1.75248057e-01
-2.21201316e-01 2.98705280e-01 -3.65646243e-01 4.48800445e-01
6.86088264e-01 8.66545737e-01 -7.07167864e-01 7.72818208e-01
2.17412159e-01 1.09814489e+00 -1.30866587e-01 3.64665985e-01
1.60752475e-01 4.53625880e-02 7.27067411e-01 1.35349393e+00
2.01401979e-01 -6.08754493e-02 2.82606244e-01 5.04240870e-01
-3.50888938e-01 -1.92447409e-01 -6.85087085e-01 -2.49828156e-02
4.34676170e-01 8.75762045e-01 -7.00896502e-01 -4.32567209e-01
-6.01267874e-01 9.89663124e-01 3.51215690e-01 5.01331210e-01
-9.40061629e-01 4.66813631e-02 6.18272781e-01 5.40651500e-01
4.31583732e-01 -1.16952315e-01 2.37486660e-01 -1.67747915e+00
-1.21190473e-01 -1.34699953e+00 6.77097559e-01 -6.05110824e-01
-1.61197841e+00 5.40507674e-01 2.20889583e-01 -1.45826030e+00
9.77004915e-02 -5.30782580e-01 -5.05400479e-01 -3.09689082e-02
-1.40903890e+00 -6.87210023e-01 -3.90947431e-01 9.08791244e-01
7.79388666e-01 -7.84273267e-01 3.50726128e-01 3.56932253e-01
-7.68779457e-01 1.18696821e+00 2.52314925e-01 3.15917492e-01
9.59717870e-01 -9.77063656e-01 2.02367106e-03 6.32155538e-01
4.44472998e-01 4.54578191e-01 6.21507883e-01 -3.44557315e-01
-1.56140828e+00 -1.33938479e+00 1.61341876e-01 -5.11770904e-01
8.35958004e-01 -3.15912694e-01 -1.02492368e+00 9.08304095e-01
1.05366088e-01 5.49260259e-01 7.37216771e-01 -4.11287509e-02
-8.98957849e-01 -3.23394030e-01 -9.35354829e-01 4.48002309e-01
1.49611318e+00 -5.50347686e-01 -4.60473984e-01 4.67356026e-01
6.52676344e-01 -4.17287409e-01 -6.52897179e-01 4.34395820e-01
5.05286157e-01 -1.28248894e+00 7.41939366e-01 -8.17939281e-01
6.91318095e-01 3.39584835e-02 -5.80290854e-01 -1.27931702e+00
-3.09583604e-01 -6.61401391e-01 -3.09207261e-01 1.29142547e+00
1.51797548e-01 -7.14230120e-01 7.26849318e-01 1.50080249e-01
-1.95872299e-02 -7.68660486e-01 -1.02449262e+00 -1.28811955e+00
1.20841553e-02 -2.53132164e-01 2.87277222e-01 1.15724671e+00
-3.40379849e-02 1.59073010e-01 -4.96226132e-01 -8.23885277e-02
5.61710477e-01 -1.60552517e-01 1.09733319e+00 -8.17125797e-01
-5.85233271e-01 -4.42290306e-01 -5.85002482e-01 -1.22945964e+00
2.95308590e-01 -9.50613320e-01 -4.14854646e-01 -7.31559992e-01
5.72507203e-01 -4.20281082e-01 -6.09672368e-01 4.18838859e-01
-1.78146437e-01 1.44073203e-01 4.01232153e-01 5.72160780e-01
-1.21539080e+00 3.99255723e-01 8.53420079e-01 -1.53665140e-01
-1.54355198e-01 -3.13901603e-02 -7.01451838e-01 6.60890996e-01
5.92967272e-01 -3.56383473e-01 -4.98461068e-01 -3.67414415e-01
1.62889048e-01 -1.53088430e-02 4.06314075e-01 -1.14730895e+00
9.19473469e-02 -1.57773376e-01 1.20481789e-01 -1.51894003e-01
-7.01328889e-02 -6.35526419e-01 -1.42222211e-01 2.81157047e-01
-3.99880052e-01 2.13914230e-01 5.34558520e-02 7.61066556e-01
-2.82227963e-01 -1.36759523e-02 8.43995035e-01 -3.87149230e-02
-9.73316491e-01 5.45990705e-01 -2.83912688e-01 5.30982196e-01
1.18956769e+00 -4.29274082e-01 -5.15125215e-01 -5.76583028e-01
-4.58930820e-01 1.48758441e-01 5.00470698e-01 4.91446555e-01
4.95379239e-01 -1.26234984e+00 -6.25073731e-01 2.47232035e-01
4.81618673e-01 -1.92987487e-01 3.91308904e-01 8.39164019e-01
-2.78272122e-01 9.89549235e-02 -9.71194133e-02 -9.04990017e-01
-1.27924657e+00 6.74406588e-01 2.37347662e-01 -3.28540117e-01
-5.03898680e-01 8.60135257e-01 4.58255768e-01 -2.48482943e-01
4.79898006e-01 -1.56194329e-01 8.98046494e-02 4.68053184e-02
3.90613705e-01 4.59517509e-01 -2.82715112e-01 -7.20488966e-01
-2.50171989e-01 4.24920052e-01 -1.72526449e-01 3.69223744e-01
1.18308914e+00 -4.83391583e-02 2.72682995e-01 5.54673851e-01
1.19872129e+00 -4.09933515e-02 -1.50463235e+00 -5.59931397e-01
-9.46017206e-02 -4.31885719e-01 -3.12630147e-01 -5.96984386e-01
-1.22852707e+00 5.46696901e-01 6.68275118e-01 8.29822198e-02
1.12830317e+00 6.85370490e-02 6.25900507e-01 4.23255980e-01
5.62849402e-01 -1.09818256e+00 5.37873745e-01 4.46237773e-01
7.76460767e-01 -1.51014400e+00 -5.56411631e-02 -1.21122330e-01
-8.83753419e-01 8.23572576e-01 9.14876580e-01 -7.69697651e-02
7.35592902e-01 2.17762485e-01 -9.84126776e-02 -6.04504049e-02
-1.08856344e+00 -1.14577301e-01 -4.49333638e-02 5.27587116e-01
1.23192593e-01 -8.90937224e-02 2.12052256e-01 5.32438099e-01
-7.51589015e-02 4.73835040e-03 6.11550987e-01 8.22319865e-01
-1.69863537e-01 -8.90264034e-01 -3.79184484e-01 4.78263229e-01
-2.56037742e-01 1.25882685e-01 -3.41432542e-01 1.01645541e+00
1.74016014e-01 6.94819450e-01 1.38831094e-01 -6.45559430e-01
3.55574071e-01 6.28915802e-02 4.85444248e-01 -5.34582615e-01
-5.03295720e-01 -1.33041099e-01 -1.60299316e-01 -4.31991369e-01
-4.30607647e-01 -7.92500794e-01 -8.41962874e-01 -3.15824360e-01
-2.65721023e-01 -1.42636567e-01 -3.29150930e-02 8.58765364e-01
4.10362750e-01 2.79752642e-01 6.63882375e-01 -6.36105597e-01
-9.60516453e-01 -8.31032574e-01 -6.41204178e-01 9.93886590e-01
3.18845779e-01 -8.68737996e-01 -6.69134736e-01 1.08159438e-01] | [9.188011169433594, 1.2706488370895386] |
97be0c09-9d9e-4602-a0ea-f6bc419caf21 | active-learning-with-effective-scoring | 2208.14856 | null | https://arxiv.org/abs/2208.14856v2 | https://arxiv.org/pdf/2208.14856v2.pdf | Active Learning with Effective Scoring Functions for Semi-Supervised Temporal Action Localization | Temporal Action Localization (TAL) aims to predict both action category and temporal boundary of action instances in untrimmed videos, i.e., start and end time. Fully-supervised solutions are usually adopted in most existing works, and proven to be effective. One of the practical bottlenecks in these solutions is the large amount of labeled training data required. To reduce expensive human label cost, this paper focuses on a rarely investigated yet practical task named semi-supervised TAL and proposes an effective active learning method, named AL-STAL. We leverage four steps for actively selecting video samples with high informativeness and training the localization model, named \emph{Train, Query, Annotate, Append}. Two scoring functions that consider the uncertainty of localization model are equipped in AL-STAL, thus facilitating the video sample rank and selection. One takes entropy of predicted label distribution as measure of uncertainty, named Temporal Proposal Entropy (TPE). And the other introduces a new metric based on mutual information between adjacent action proposals and evaluates the informativeness of video samples, named Temporal Context Inconsistency (TCI). To validate the effectiveness of proposed method, we conduct extensive experiments on two benchmark datasets THUMOS'14 and ActivityNet 1.3. Experiment results show that AL-STAL outperforms the existing competitors and achieves satisfying performance compared with fully-supervised learning. | ['Wensheng Zhang', 'Chenyang Zhang', 'Yongqiang Tang', 'Xuebing Yang', 'Ding Li'] | 2022-08-31 | null | null | null | null | ['action-localization'] | ['computer-vision'] | [ 3.57907951e-01 -1.77899301e-02 -7.42803097e-01 -2.93644249e-01
-1.10338950e+00 -3.39386106e-01 3.50948811e-01 6.56354707e-03
-5.87044477e-01 7.45328903e-01 3.44421744e-01 2.50157028e-01
-1.72372863e-01 -1.79418877e-01 -5.07235527e-01 -9.16737020e-01
-2.12385744e-01 1.42370239e-01 6.78199351e-01 6.43151462e-01
3.36064726e-01 -1.52480863e-02 -1.37865329e+00 3.80022019e-01
8.69254172e-01 1.45359719e+00 1.55556202e-01 1.24846019e-01
-1.03217112e-02 1.36480451e+00 -3.80489141e-01 -1.45804673e-01
1.70727223e-01 -5.89581311e-01 -7.54525125e-01 3.59149128e-01
2.30915584e-02 -3.26700062e-01 -2.63608154e-02 9.14791703e-01
4.18119043e-01 3.30004096e-01 3.91077965e-01 -1.43940747e+00
5.62532656e-02 6.87203526e-01 -5.98636389e-01 5.24056196e-01
5.35266757e-01 2.42822528e-01 9.50224698e-01 -1.06727540e+00
7.01488853e-01 9.61628497e-01 6.08887613e-01 2.67195642e-01
-8.89064610e-01 -5.30937135e-01 3.58121008e-01 7.62737513e-01
-1.53951931e+00 -5.87356627e-01 8.33855450e-01 -4.64612961e-01
4.51747328e-01 1.26670107e-01 5.65294564e-01 1.09095418e+00
-8.82125925e-03 1.33941519e+00 1.20800066e+00 -2.01975644e-01
5.88451684e-01 7.86587000e-02 -1.25121236e-01 7.77114213e-01
-1.80757865e-01 -1.65848315e-01 -7.79520690e-01 -7.96097666e-02
4.84388024e-01 8.59022737e-02 -3.70364934e-01 -5.44888556e-01
-1.36595917e+00 3.91728312e-01 1.01849414e-01 1.78759471e-01
-4.78833944e-01 -7.73488507e-02 6.57184005e-01 -2.39457022e-02
4.58505392e-01 2.25049704e-02 -3.88225198e-01 -5.52423000e-01
-8.40275764e-01 -1.45319894e-01 2.20246390e-01 8.36917937e-01
4.92954344e-01 -3.12909901e-01 -4.95590657e-01 5.64567029e-01
3.20398033e-01 7.37764016e-02 6.51257217e-01 -1.07370329e+00
5.66356003e-01 8.56468379e-01 1.92681029e-01 -8.71643424e-01
-1.07696325e-01 -1.98652327e-01 -3.48819613e-01 -1.25832453e-01
2.76875913e-01 -6.21636659e-02 -5.56915522e-01 1.47843266e+00
4.29810196e-01 6.50844574e-01 -1.13406815e-01 9.75289643e-01
7.26646721e-01 5.67373931e-01 4.03702885e-01 -8.11173975e-01
8.08305383e-01 -1.24626529e+00 -8.44593346e-01 -1.89413220e-01
7.09115803e-01 -6.81697130e-01 8.80594969e-01 4.58968520e-01
-1.01964688e+00 -6.08904243e-01 -8.50252151e-01 4.04012084e-01
-2.35132966e-02 5.79880774e-01 5.37769735e-01 1.01937979e-01
-5.08691132e-01 6.40012622e-01 -1.04937685e+00 -2.32949004e-01
7.55150259e-01 1.59592435e-01 -4.84631896e-01 -4.32482585e-02
-1.10936856e+00 5.23743093e-01 8.00019503e-01 1.87875092e-01
-1.09984386e+00 -2.41560727e-01 -7.62213230e-01 -9.75127593e-02
9.59459722e-01 1.33575976e-01 1.21270084e+00 -1.14981306e+00
-1.29173100e+00 5.90730369e-01 -4.74555716e-02 -5.18293321e-01
6.47342861e-01 -3.05876523e-01 -4.43967640e-01 3.28052759e-01
3.17206442e-01 6.81187630e-01 8.11145067e-01 -9.75197017e-01
-9.74602461e-01 -1.84819862e-01 -3.01213507e-02 5.56529224e-01
-3.62669885e-01 -5.45723364e-03 -8.11585367e-01 -5.72246075e-01
2.84153432e-01 -8.89975429e-01 -4.49135080e-02 -4.64191437e-02
-2.35078216e-01 -5.99331915e-01 9.38050449e-01 -6.83032513e-01
1.60241854e+00 -2.15141940e+00 4.08331864e-02 5.34111597e-02
-5.46559654e-02 3.64292443e-01 1.33522570e-01 2.53875881e-01
1.02908760e-02 -2.14063600e-01 -1.85019180e-01 -2.29526773e-01
-1.23934962e-01 7.66369998e-02 -1.34969102e-02 4.38052267e-01
1.00137040e-01 7.00565755e-01 -1.16139925e+00 -1.11528552e+00
3.54075640e-01 2.24523082e-01 -2.52455652e-01 2.22843483e-01
-2.15994641e-01 7.82206833e-01 -6.98269308e-01 9.01886463e-01
2.77547538e-01 -3.67164850e-01 1.63831830e-01 -3.59911323e-01
-1.44409657e-01 1.52390614e-01 -1.31895065e+00 1.78353119e+00
-8.56218636e-02 3.75249773e-01 -4.52043176e-01 -1.00693130e+00
6.09456182e-01 4.47881132e-01 8.91972959e-01 -6.33012235e-01
1.23103984e-01 1.28387555e-01 -2.31344342e-01 -8.05269659e-01
6.12088591e-02 4.15479571e-01 -2.94768903e-02 2.56196439e-01
1.63384512e-01 7.10781693e-01 4.52488154e-01 2.60963947e-01
1.21905804e+00 8.86757255e-01 5.87939203e-01 1.13092363e-01
8.40750575e-01 -2.09802926e-01 1.11373472e+00 5.15007317e-01
-8.27840865e-01 2.73862273e-01 6.87473536e-01 -1.77016735e-01
-4.85888690e-01 -8.17853749e-01 1.16926461e-01 1.15139341e+00
5.58874309e-01 -5.20240068e-01 -6.98751092e-01 -1.35573936e+00
-3.77145082e-01 5.67302942e-01 -5.69101810e-01 -7.71603286e-02
-5.34338951e-01 -4.55213726e-01 2.45595500e-01 7.07065463e-01
8.37073922e-01 -1.25893271e+00 -6.53793514e-01 1.56440698e-02
-5.49853146e-01 -1.22706008e+00 -5.26715577e-01 -5.75328842e-02
-8.68534923e-01 -1.13766503e+00 -4.12335515e-01 -5.13155103e-01
6.75840795e-01 1.69933766e-01 8.32472205e-01 -1.38211325e-01
-8.58979151e-02 2.84839600e-01 -6.65054321e-01 1.99911073e-02
4.49581221e-02 -2.03802630e-01 -1.41106565e-02 5.09900391e-01
3.98079187e-01 -3.81918758e-01 -9.15119946e-01 7.68847585e-01
-6.69779539e-01 -6.72555938e-02 8.79588425e-01 4.96146262e-01
9.94366050e-01 3.24643612e-01 5.75344026e-01 -6.92874610e-01
1.98521301e-01 -4.67037320e-01 -3.92675370e-01 4.51013178e-01
-5.93370557e-01 -1.89510077e-01 4.02302414e-01 -4.00417089e-01
-1.23448551e+00 4.01166350e-01 1.71940401e-01 -5.62613606e-01
-1.05610698e-01 5.87251723e-01 -2.50555128e-01 1.34555981e-01
3.43940645e-01 3.70211899e-01 -3.23102802e-01 -2.80377567e-01
1.74380876e-02 5.44695377e-01 4.33644146e-01 -1.05977625e-01
4.25797969e-01 4.53340024e-01 -1.74099326e-01 -4.11905855e-01
-1.05206919e+00 -8.32524359e-01 -7.38945484e-01 -7.62396991e-01
8.40282023e-01 -9.39048588e-01 -5.54941654e-01 3.34504902e-01
-7.51994550e-01 -1.10551253e-01 -1.90379560e-01 7.88807094e-01
-6.42717659e-01 5.65581858e-01 -1.95473865e-01 -1.02516878e+00
-1.19663224e-01 -1.13625026e+00 1.03994954e+00 2.23287776e-01
-2.11799398e-01 -8.05596173e-01 -2.04879493e-02 6.84658468e-01
-6.98981658e-02 3.25387836e-01 3.26806158e-01 -7.79878855e-01
-8.38679254e-01 -4.08920556e-01 -1.42993880e-02 3.74422908e-01
4.09755222e-02 -2.02572316e-01 -8.54930341e-01 -1.75243810e-01
-1.05382249e-01 -6.78129971e-01 8.86431158e-01 3.09265435e-01
1.32809603e+00 -2.49829978e-01 -4.46980953e-01 8.54077190e-02
1.16767049e+00 5.29291868e-01 7.30069458e-01 6.77217841e-02
4.49972510e-01 3.58727276e-01 1.60215509e+00 5.86579621e-01
9.63986292e-02 8.46980214e-01 4.03256595e-01 2.84639359e-01
9.82169583e-02 -3.91302735e-01 7.31825352e-01 8.99282753e-01
-3.05316061e-01 -1.93873346e-01 -5.78662097e-01 3.20702285e-01
-2.29390717e+00 -1.21299016e+00 2.78163433e-01 2.38314605e+00
6.74883664e-01 5.96291006e-01 2.23959371e-01 3.90047073e-01
7.97664583e-01 2.29430377e-01 -7.21112788e-01 4.28957105e-01
8.60798135e-02 -3.03565562e-01 3.31933290e-01 1.08832702e-01
-1.62003624e+00 8.28257740e-01 4.63268042e+00 1.24681795e+00
-8.36433351e-01 3.47762883e-01 7.04487681e-01 -1.98318064e-01
4.51201409e-01 1.08751573e-01 -6.21234596e-01 8.79568219e-01
5.14690876e-01 1.38443723e-01 -5.86574851e-03 1.12224400e+00
6.29091561e-01 -4.86247331e-01 -1.17966139e+00 1.10415578e+00
2.67776340e-01 -9.47880328e-01 -3.41621697e-01 -4.00357485e-01
6.16746128e-01 -3.25242549e-01 -1.33379802e-01 5.39533556e-01
-2.90460080e-01 -4.34189558e-01 6.90168977e-01 6.96876645e-01
5.56178749e-01 -5.93365073e-01 8.93720567e-01 4.86960739e-01
-1.46696222e+00 -4.24279988e-01 -1.36033457e-03 2.84046620e-01
2.45105043e-01 2.66881734e-01 -6.78269744e-01 4.57488656e-01
6.91986263e-01 1.11496305e+00 -6.68741643e-01 1.33906364e+00
-4.07961369e-01 8.21411788e-01 -3.68974395e-02 2.24624611e-02
3.25931430e-01 -1.31866142e-01 5.80161810e-01 8.15260530e-01
1.30576506e-01 1.23437673e-01 5.56067944e-01 3.93463343e-01
7.70118833e-03 2.37044588e-01 -1.80473998e-01 -2.68994123e-01
6.55104756e-01 1.26718962e+00 -1.04874730e+00 -3.26407552e-01
-2.50789285e-01 1.03869987e+00 9.25962776e-02 1.04067326e-01
-1.25689793e+00 -1.24490388e-01 -1.01071961e-01 4.80706804e-02
2.93970376e-01 1.24596730e-01 2.07990527e-01 -1.01996076e+00
3.32448959e-01 -7.07813084e-01 6.66383922e-01 -7.38411665e-01
-9.14072454e-01 2.65928715e-01 1.52595460e-01 -1.90654385e+00
-1.15707800e-01 -1.31441221e-01 -6.25370562e-01 1.71721786e-01
-1.07260382e+00 -9.87926304e-01 -4.65943515e-01 5.77755094e-01
1.05237353e+00 -1.84357062e-01 3.68983030e-01 5.29959381e-01
-9.60860193e-01 5.11055410e-01 -1.68848947e-01 9.52887312e-02
7.28534997e-01 -9.63733077e-01 -2.36678496e-01 8.14223588e-01
7.55336285e-02 3.48882645e-01 4.42281961e-01 -7.52407789e-01
-1.18494928e+00 -1.12154555e+00 7.71839321e-01 -3.22746068e-01
5.31588614e-01 6.17935583e-02 -5.64333200e-01 5.85527003e-01
-1.48949310e-01 1.67919293e-01 7.01752126e-01 -3.78759712e-01
9.22219530e-02 -2.11670637e-01 -9.72358644e-01 3.93342614e-01
1.19443393e+00 -2.91000783e-01 -4.90584701e-01 5.78376055e-01
6.02697492e-01 -2.47895211e-01 -8.04326057e-01 6.48197711e-01
5.55546582e-01 -1.22907817e+00 6.94684684e-01 -3.96154702e-01
3.61100405e-01 -4.57012385e-01 2.67751366e-02 -7.70738542e-01
-2.55315661e-01 -4.91477251e-01 -4.05698061e-01 1.33820891e+00
3.02221835e-01 -1.89655602e-01 8.50028276e-01 4.00135279e-01
-1.14043951e-01 -1.23760116e+00 -9.35824335e-01 -8.27677906e-01
-8.52157354e-01 -3.32688153e-01 1.94586232e-01 9.40146565e-01
1.72657743e-01 2.04483077e-01 -5.57054818e-01 -8.64484757e-02
6.33150756e-01 -7.36667588e-02 4.30598080e-01 -9.81753051e-01
-1.73169076e-01 -2.28872523e-01 -6.41932189e-01 -1.09634483e+00
2.37953309e-02 -5.15443146e-01 1.57345638e-01 -1.35741937e+00
2.52850205e-01 -2.89514512e-01 -7.09563613e-01 5.45047641e-01
-2.11641714e-01 3.16829443e-01 -7.58897588e-02 5.77203810e-01
-1.63121569e+00 6.47914052e-01 9.33343947e-01 5.13292663e-02
-2.91563779e-01 2.40081117e-01 -4.95983027e-02 1.14703166e+00
5.38996220e-01 -3.80230874e-01 -7.55402029e-01 -2.84873825e-02
-1.69979874e-02 2.98235655e-01 2.60369718e-01 -1.34646845e+00
2.65909612e-01 -2.29841799e-01 3.13958585e-01 -9.08986688e-01
2.72902876e-01 -9.15661335e-01 -3.76535952e-02 1.86976850e-01
-7.58171439e-01 -3.18167925e-01 -4.16281641e-01 9.46156740e-01
-3.97804528e-01 -2.82054543e-01 7.17594504e-01 -1.65189892e-01
-1.27745640e+00 5.55956185e-01 -1.11252517e-01 5.07153757e-02
1.38771427e+00 -3.60486358e-01 -7.61300465e-03 -1.96372256e-01
-8.60523701e-01 2.90562183e-01 1.52138710e-01 4.14525419e-01
6.54268086e-01 -1.56122243e+00 -2.60704398e-01 -1.96252748e-01
4.14111078e-01 -2.84102231e-01 4.37774390e-01 1.43730307e+00
-3.07370484e-01 3.62994403e-01 8.19118023e-02 -7.19734430e-01
-1.42319083e+00 5.12765646e-01 7.37888962e-02 -3.50240320e-01
-3.98690432e-01 7.48402178e-01 -2.27173436e-02 1.12924732e-01
6.97522998e-01 -3.51958722e-02 -5.71661890e-01 2.51327306e-01
4.40982044e-01 6.07911050e-01 -7.98795223e-02 -8.87785733e-01
-4.87091780e-01 3.45771611e-01 -4.25633043e-03 9.53191891e-02
9.69026327e-01 -2.15480715e-01 7.85241574e-02 5.91558516e-01
1.05265439e+00 -3.02536517e-01 -1.58631980e+00 -4.54869270e-01
3.84251714e-01 -6.55679464e-01 4.34100628e-02 -7.68384516e-01
-9.78729904e-01 6.25662029e-01 9.21426713e-01 2.84990817e-02
1.18903136e+00 -1.18633479e-01 7.93758273e-01 4.10706222e-01
5.16743898e-01 -1.45575511e+00 4.70038652e-01 9.04182717e-02
5.91014028e-01 -1.45219338e+00 2.14647070e-01 -4.30113673e-01
-1.08544660e+00 6.71074867e-01 8.83692861e-01 2.76347905e-01
4.61957574e-01 -2.07140744e-01 -1.57671541e-01 -1.15051605e-02
-7.46752858e-01 -1.54821545e-01 3.76619816e-01 1.44449577e-01
2.36580148e-01 -2.51184404e-01 -6.16450667e-01 5.44817448e-01
6.57275081e-01 4.43316996e-01 -7.51859844e-02 1.23671412e+00
-5.76769352e-01 -7.82661617e-01 -1.37972578e-01 4.35997397e-01
-4.22708929e-01 3.55548561e-01 -3.72838140e-01 5.16321242e-01
6.40596509e-01 8.70942056e-01 -4.17627133e-02 -5.05619466e-01
1.07496427e-02 1.92617550e-01 2.70803243e-01 -3.72521996e-01
-1.65946573e-01 3.01086217e-01 1.55261442e-01 -8.31609309e-01
-9.98461664e-01 -8.66416216e-01 -1.36027443e+00 3.90049160e-01
-5.15713871e-01 2.90894955e-01 1.56139791e-01 1.13665402e+00
2.88416535e-01 2.61378169e-01 8.71061206e-01 -6.42406285e-01
-4.50786620e-01 -9.41644728e-01 -4.10870463e-01 5.54866076e-01
-1.69951499e-01 -1.05413771e+00 -4.67022538e-01 2.77518094e-01] | [8.486466407775879, 0.6189032793045044] |
18c2fc1f-8003-4209-b6a6-ebad06c2519d | combining-learned-skills-and-reinforcement | 1908.00722 | null | https://arxiv.org/abs/1908.00722v3 | https://arxiv.org/pdf/1908.00722v3.pdf | Learning to combine primitive skills: A step towards versatile robotic manipulation | Manipulation tasks such as preparing a meal or assembling furniture remain highly challenging for robotics and vision. Traditional task and motion planning (TAMP) methods can solve complex tasks but require full state observability and are not adapted to dynamic scene changes. Recent learning methods can operate directly on visual inputs but typically require many demonstrations and/or task-specific reward engineering. In this work we aim to overcome previous limitations and propose a reinforcement learning (RL) approach to task planning that learns to combine primitive skills. First, compared to previous learning methods, our approach requires neither intermediate rewards nor complete task demonstrations during training. Second, we demonstrate the versatility of our vision-based task planning in challenging settings with temporary occlusions and dynamic scene changes. Third, we propose an efficient training of basic skills from few synthetic demonstrations by exploring recent CNN architectures and data augmentation. Notably, while all of our policies are learned on visual inputs in simulated environments, we demonstrate the successful transfer and high success rates when applying such policies to manipulation tasks on a real UR5 robotic arm. | ['Ivan Laptev', 'Cordelia Schmid', 'Josef Sivic', 'Igor Kalevatykh', 'Alexander Pashevich', 'Robin Strudel'] | 2019-08-02 | null | null | null | null | ['industrial-robots'] | ['robots'] | [ 3.25618356e-01 2.37113953e-01 -1.15975142e-02 -3.81372757e-02
-3.84881407e-01 -6.94109023e-01 7.69433677e-01 -1.71592399e-01
-5.56329668e-01 9.63849187e-01 -1.04751930e-01 -3.77060473e-01
-1.53250232e-01 -3.23329419e-01 -1.12022197e+00 -2.87515491e-01
-1.45778164e-01 5.22676468e-01 3.19316298e-01 -4.75039303e-01
2.83263564e-01 6.30442142e-01 -1.64367402e+00 -4.22526933e-02
7.45016456e-01 5.69831848e-01 8.28286886e-01 8.95304501e-01
3.20543647e-01 9.05564547e-01 -3.92583042e-01 4.61111277e-01
5.91629684e-01 -1.82186157e-01 -9.40073907e-01 2.92812169e-01
3.50864470e-01 -5.79638004e-01 -7.00043142e-01 6.37835026e-01
3.83817524e-01 3.63241911e-01 4.99216646e-01 -1.58336508e+00
-6.34546816e-01 2.67971486e-01 -7.74428993e-02 -6.35506585e-02
4.90227252e-01 9.47003543e-01 6.26223266e-01 -6.23256326e-01
8.04574251e-01 1.17125726e+00 3.73233289e-01 9.36433673e-01
-1.29133892e+00 -2.76047856e-01 4.05543655e-01 2.36943528e-01
-6.20604336e-01 -2.94436544e-01 4.49090898e-01 -5.02253771e-01
1.29207051e+00 -2.43673667e-01 9.61404681e-01 1.47733104e+00
-9.49228257e-02 1.00579786e+00 1.26244330e+00 -2.81256706e-01
2.46826157e-01 -3.82355660e-01 -3.22407335e-01 9.19363022e-01
5.90562187e-02 5.93120277e-01 -3.46907437e-01 1.91499636e-01
1.27257514e+00 1.61659062e-01 -2.67197847e-01 -1.00301528e+00
-1.97704113e+00 5.37569463e-01 6.45108104e-01 -3.75459231e-02
-4.73581821e-01 8.83188725e-01 1.96098626e-01 5.95814288e-01
-4.57389385e-01 9.98881459e-01 -7.20179379e-01 -2.59652644e-01
-3.09268713e-01 4.69166577e-01 8.29975128e-01 1.36992717e+00
6.34702265e-01 3.37937474e-01 -3.22720021e-01 3.78652781e-01
-2.24037073e-03 6.27507269e-01 3.13863426e-01 -1.49009264e+00
4.22359675e-01 2.31350154e-01 6.42241716e-01 -4.27812070e-01
-6.64433718e-01 6.76950738e-02 -2.68149257e-01 9.54991162e-01
6.62465811e-01 -2.68561661e-01 -1.30915010e+00 1.67966378e+00
2.52593309e-01 1.15177773e-01 2.27460608e-01 8.87177050e-01
4.95304644e-01 4.24416155e-01 1.21554390e-01 1.08859479e-01
9.19669688e-01 -1.44149351e+00 -3.84839594e-01 -3.93308580e-01
4.07172650e-01 -4.43628937e-01 1.35611188e+00 4.46925968e-01
-1.18942595e+00 -6.08380497e-01 -9.48233843e-01 -5.78338057e-02
-1.90562546e-01 2.09631458e-01 8.09697866e-01 -1.03256265e-02
-1.10064328e+00 9.17396247e-01 -1.25116932e+00 -5.72584867e-01
3.57073337e-01 6.38300061e-01 -5.06416738e-01 -1.05709754e-01
-6.11027777e-01 1.25783598e+00 3.90475303e-01 2.60947011e-02
-1.82946014e+00 -4.76620853e-01 -1.08358431e+00 8.50789994e-02
7.05662966e-01 -9.29118395e-01 1.71911335e+00 -5.44408858e-01
-2.05707216e+00 4.20727164e-01 2.50826150e-01 -4.97197032e-01
6.08441472e-01 -4.23976958e-01 3.43605936e-01 2.20771983e-01
-4.07570153e-02 1.16547561e+00 9.32566822e-01 -1.28452277e+00
-4.93836224e-01 2.58288532e-01 5.71328819e-01 3.23922396e-01
1.90344781e-01 -4.20971423e-01 -7.97405243e-02 -2.68765181e-01
1.43845044e-02 -1.21582246e+00 -5.92111170e-01 4.48418289e-01
-5.41789085e-02 -2.46101767e-01 7.94960797e-01 -1.63820371e-01
7.88062066e-02 -1.89315140e+00 6.59780264e-01 -2.40043908e-01
-6.65164590e-02 2.54036874e-01 -4.92306113e-01 6.38598025e-01
2.84117669e-01 -3.03737074e-01 -7.20071495e-02 -1.38806224e-01
3.07561725e-01 5.66195548e-01 -3.43565971e-01 2.21581519e-01
4.56734836e-01 1.37037098e+00 -1.36601460e+00 -2.24124908e-01
4.87098515e-01 2.05027655e-01 -6.88605726e-01 3.26892734e-01
-7.82185435e-01 1.17144978e+00 -4.72811848e-01 7.02004611e-01
-1.05839036e-02 -2.11727813e-01 2.22014636e-01 1.96915880e-01
-1.55159369e-01 3.52080017e-01 -8.96745801e-01 2.25306296e+00
-7.01430023e-01 6.00906014e-01 2.68137366e-01 -1.09014285e+00
6.90686285e-01 2.51871943e-01 3.55669051e-01 -5.79108715e-01
-3.80924419e-02 2.59651452e-01 1.77338883e-01 -8.50574791e-01
4.64565724e-01 4.92701679e-02 -2.27588952e-01 3.49375278e-01
3.15143168e-01 -1.09164381e+00 1.53956831e-01 -1.10667665e-02
1.32675874e+00 1.02726698e+00 2.53199220e-01 3.54437307e-02
3.12788635e-02 5.35282433e-01 3.09425861e-01 1.00032783e+00
-3.99193585e-01 4.32720572e-01 3.85189831e-01 -5.38000941e-01
-1.33879709e+00 -1.11468136e+00 4.28761870e-01 1.07926548e+00
2.93966621e-01 7.93954432e-02 -2.48819053e-01 -6.85862780e-01
3.96209657e-02 5.91947079e-01 -5.08668005e-01 -1.46644292e-02
-1.01212513e+00 -4.53246981e-02 2.23892316e-01 6.46015048e-01
3.08765858e-01 -1.89644480e+00 -1.27689767e+00 2.73283243e-01
1.52208894e-01 -1.26964247e+00 -1.86732858e-01 3.33394051e-01
-7.61923373e-01 -1.22501743e+00 -7.93708801e-01 -1.17468631e+00
7.50802159e-01 4.11751866e-01 9.58046675e-01 1.53692653e-02
-4.28804129e-01 9.95011628e-01 -2.49983653e-01 -3.39556754e-01
-4.64993596e-01 2.69702002e-02 1.89690173e-01 -7.97487319e-01
-3.91821116e-01 -8.69551420e-01 -7.01069593e-01 3.22049886e-01
-5.27909636e-01 2.63642192e-01 9.06885207e-01 1.24711680e+00
3.85635734e-01 -5.52424014e-01 5.58713973e-01 -4.05806750e-01
6.32027805e-01 -1.56992242e-01 -8.32385719e-01 6.65663108e-02
-3.38178366e-01 1.90458700e-01 5.66675603e-01 -8.61189127e-01
-7.81509936e-01 4.58677918e-01 1.87414482e-01 -5.11323094e-01
-3.72681856e-01 1.30164638e-01 4.29898977e-01 -1.68683931e-01
7.85388947e-01 2.87757456e-01 2.59913336e-02 -5.39653450e-02
5.50406814e-01 1.28940027e-02 6.47256374e-01 -9.34071720e-01
9.05184627e-01 5.01800239e-01 1.36462718e-01 -5.11063993e-01
-4.16352302e-01 -1.29859895e-01 -8.28429937e-01 -3.08320522e-02
8.32659423e-01 -7.87042260e-01 -1.36222172e+00 2.46339664e-01
-1.09463334e+00 -1.27928138e+00 -5.88332593e-01 6.69161320e-01
-1.31587303e+00 2.28376105e-01 -6.01681054e-01 -5.52175820e-01
7.30180889e-02 -1.34989822e+00 1.14785647e+00 1.00663416e-01
-1.90012649e-01 -4.46413428e-01 -1.07608840e-01 -1.47794206e-02
5.62047839e-01 4.57790434e-01 6.82034910e-01 -8.02929327e-03
-1.07743466e+00 1.48917645e-01 -1.34864837e-01 -1.72614120e-02
1.17467560e-01 -3.33743125e-01 -4.32497531e-01 -6.49585426e-01
-4.07507509e-01 -1.03900778e+00 6.79354072e-01 1.20852768e-01
1.01970315e+00 -5.85551299e-02 -3.70190978e-01 4.84768063e-01
1.02730262e+00 2.55245775e-01 4.27268356e-01 4.78346884e-01
6.46766603e-01 6.35716915e-01 9.73131418e-01 2.38223374e-01
3.27399135e-01 6.65807962e-01 9.18273807e-01 1.02197036e-01
-1.82603285e-01 -3.91206175e-01 5.03729343e-01 3.82653207e-01
-1.83534622e-01 1.78881824e-01 -8.10767651e-01 7.73329139e-01
-2.10555768e+00 -8.04113865e-01 1.66035950e-01 1.84571469e+00
6.73091173e-01 1.60759389e-01 2.03579217e-02 -1.91642448e-01
8.51271823e-02 -1.03306353e-01 -1.05567753e+00 -2.43081957e-01
3.60097110e-01 2.88274348e-01 3.63787919e-01 3.66513819e-01
-9.44987416e-01 1.24866772e+00 6.86075354e+00 4.40690182e-02
-9.60902035e-01 -9.95926335e-02 -2.05956280e-01 -4.04465169e-01
5.90585694e-02 1.37719139e-01 -3.23743045e-01 -1.93183254e-02
3.16263974e-01 2.04720706e-01 8.08039486e-01 1.04031086e+00
8.20828825e-02 -2.53696114e-01 -1.46177983e+00 8.23625445e-01
-3.43587816e-01 -1.06075990e+00 -4.17727828e-01 -1.12787955e-01
7.55847394e-01 2.66266376e-01 1.21088736e-01 8.62134695e-01
1.08101678e+00 -1.32304144e+00 6.31996214e-01 3.13703537e-01
6.80354655e-01 -2.90595323e-01 3.22594345e-02 6.71275675e-01
-7.18125939e-01 -4.17617530e-01 -3.69598538e-01 -4.11964357e-01
3.05067569e-01 -2.84428775e-01 -1.17905414e+00 1.35943696e-01
6.01915240e-01 6.29713416e-01 -2.39229053e-02 8.25388432e-01
-6.82151079e-01 9.42274034e-02 -3.03800851e-01 -3.56342733e-01
4.16303247e-01 8.79175738e-02 5.53959906e-01 7.72384644e-01
2.65263379e-01 -5.00772744e-02 6.41757131e-01 8.80102098e-01
1.82946697e-01 -4.72429276e-01 -9.62485433e-01 -6.92743659e-02
2.26859078e-01 1.15898287e+00 -5.02568662e-01 -2.38507181e-01
-2.15612486e-01 1.09356403e+00 7.77202487e-01 5.47958434e-01
-7.66412437e-01 -1.43007636e-01 7.80628204e-01 -1.87599316e-01
6.59968913e-01 -1.06794596e+00 7.14270994e-02 -1.16127503e+00
5.02169058e-02 -8.80091906e-01 -2.33436272e-01 -9.70328033e-01
-7.54416585e-01 3.37632895e-01 -3.85737121e-02 -1.32725585e+00
-4.24315423e-01 -8.72328043e-01 -4.69440997e-01 5.33099830e-01
-1.79821420e+00 -1.29829514e+00 -5.78191280e-01 7.23906338e-01
8.91989231e-01 -2.00637817e-01 9.11125600e-01 -2.58792251e-01
-9.58489254e-02 4.91579138e-02 -2.96521217e-01 -1.79324567e-01
5.31431675e-01 -1.40084553e+00 6.09891653e-01 5.65113962e-01
6.14387030e-03 4.96835440e-01 7.01267362e-01 -4.72123384e-01
-1.69498384e+00 -6.94406986e-01 5.13051189e-02 -6.32077456e-01
6.65757477e-01 -4.70587462e-01 -5.45703351e-01 1.03522921e+00
3.39139163e-01 1.90572202e-01 -2.07844481e-01 -1.34136528e-01
-1.70182690e-01 3.90411496e-01 -1.04078841e+00 9.71376896e-01
1.52552176e+00 -3.05292159e-01 -8.48981380e-01 6.15515471e-01
9.17637169e-01 -8.54142249e-01 -6.41618967e-01 5.34236968e-01
6.21977746e-01 -5.70840061e-01 1.03580141e+00 -1.00467467e+00
5.49437582e-01 -6.22036099e-01 4.16789912e-02 -1.48672116e+00
-2.85963893e-01 -9.02220309e-01 -2.32166588e-01 3.96663874e-01
2.68242866e-01 -4.46891963e-01 7.16215909e-01 3.74118596e-01
-3.50717038e-01 -5.51627278e-01 -7.24572003e-01 -9.07531798e-01
4.40414026e-02 -2.22101942e-01 3.48556459e-01 8.13152015e-01
2.77718753e-01 1.08816385e-01 -6.12916887e-01 1.11147016e-01
2.87235469e-01 2.61719435e-01 1.26820290e+00 -8.44820082e-01
-6.49158061e-01 -2.04238072e-01 -1.66460216e-01 -1.32361162e+00
2.94552982e-01 -7.37132967e-01 5.72849631e-01 -1.80813980e+00
-1.83249846e-01 -5.68784356e-01 1.82061959e-02 8.08944285e-01
1.49695113e-01 -1.29737273e-01 5.08790791e-01 3.49308997e-01
-7.44652510e-01 7.56879747e-01 1.93980837e+00 -1.68254465e-01
-3.44184697e-01 -9.67823341e-03 -2.28089973e-01 5.33610582e-01
9.17270243e-01 -3.03223282e-01 -5.71614683e-01 -6.69032753e-01
1.95282683e-01 3.05032164e-01 7.85408854e-01 -1.09686494e+00
2.56165713e-01 -4.85142380e-01 2.85730213e-01 -2.65175849e-01
5.54785132e-01 -8.99292529e-01 -2.09198251e-01 1.03007078e+00
-3.34579319e-01 1.51834741e-01 2.98844695e-01 8.44500303e-01
1.86314970e-01 -9.43164304e-02 4.97939795e-01 -5.33941388e-01
-1.06332314e+00 1.75587371e-01 -6.31884098e-01 -2.62364715e-01
1.31955802e+00 -9.09743682e-02 -3.97449762e-01 -3.43174905e-01
-1.10421515e+00 6.81363642e-01 6.46878898e-01 5.51758707e-01
6.88546062e-01 -1.13750088e+00 -5.15624940e-01 2.17122454e-02
7.20259026e-02 4.05483931e-01 -1.78792514e-02 7.45398343e-01
-7.16762543e-01 3.57299358e-01 -7.44413435e-01 -6.62033617e-01
-9.00357842e-01 8.44612420e-01 2.23777264e-01 -2.54480958e-01
-9.14046407e-01 5.13073504e-01 2.57375985e-01 -8.62094998e-01
2.42160425e-01 -6.50799632e-01 -4.50510979e-02 -6.39781237e-01
-1.28696896e-02 5.48401512e-02 -4.78331119e-01 6.98205158e-02
-2.81989407e-02 3.64133418e-01 7.99564943e-02 -3.35049748e-01
1.42639947e+00 1.58930466e-01 3.27999920e-01 2.06436470e-01
5.83144963e-01 -5.15672207e-01 -2.17831922e+00 -1.74374267e-01
2.75514796e-02 -3.09057862e-01 -5.86931348e-01 -8.20547879e-01
-5.34219325e-01 9.72709537e-01 2.74879485e-01 -9.83655453e-02
7.11848676e-01 -7.29456320e-02 5.80808580e-01 1.21738338e+00
8.40924442e-01 -1.17356086e+00 8.19261849e-01 8.67740691e-01
1.22079527e+00 -1.50265837e+00 -1.20646477e-01 -2.02662349e-01
-8.00043166e-01 1.08638597e+00 9.06305075e-01 -4.85320330e-01
1.63745090e-01 2.56474644e-01 2.08407976e-02 -1.49308905e-01
-9.21309352e-01 -5.41462600e-01 -1.58126857e-02 1.07710338e+00
-1.29803672e-01 -1.52907476e-01 4.11874708e-03 -2.97812790e-01
-1.51147977e-01 1.21841006e-01 6.99153066e-01 1.46202481e+00
-4.62870181e-01 -9.66999054e-01 4.09243852e-02 9.57034975e-02
-1.20067615e-02 1.56819984e-01 -1.95499778e-01 1.10163641e+00
-2.05849931e-01 5.80519438e-01 -2.32273221e-01 -8.00570026e-02
5.00161052e-01 -1.10152029e-02 1.11641335e+00 -9.59355652e-01
-5.59688568e-01 -2.54019201e-01 -2.06569117e-02 -8.08815539e-01
-4.80879009e-01 -7.84447849e-01 -1.42370152e+00 1.47571340e-01
7.11414963e-02 -4.02640939e-01 8.98124814e-01 8.97206724e-01
3.90576601e-01 8.57689738e-01 2.52360851e-01 -1.54207325e+00
-8.04318488e-01 -1.02369773e+00 -1.83389172e-01 4.10724252e-01
5.97853601e-01 -9.24260676e-01 8.17575082e-02 1.84188504e-02] | [4.591673851013184, 0.7569230794906616] |
7d7779b6-fc7c-4d96-9036-576e321f2e3a | exploring-faithful-rationale-for-multi-hop | 2212.01060 | null | https://arxiv.org/abs/2212.01060v1 | https://arxiv.org/pdf/2212.01060v1.pdf | Exploring Faithful Rationale for Multi-hop Fact Verification via Salience-Aware Graph Learning | The opaqueness of the multi-hop fact verification model imposes imperative requirements for explainability. One feasible way is to extract rationales, a subset of inputs, where the performance of prediction drops dramatically when being removed. Though being explainable, most rationale extraction methods for multi-hop fact verification explore the semantic information within each piece of evidence individually, while ignoring the topological information interaction among different pieces of evidence. Intuitively, a faithful rationale bears complementary information being able to extract other rationales through the multi-hop reasoning process. To tackle such disadvantages, we cast explainable multi-hop fact verification as subgraph extraction, which can be solved based on graph convolutional network (GCN) with salience-aware graph learning. In specific, GCN is utilized to incorporate the topological interaction information among multiple pieces of evidence for learning evidence representation. Meanwhile, to alleviate the influence of noisy evidence, the salience-aware graph perturbation is induced into the message passing of GCN. Moreover, the multi-task model with three diagnostic properties of rationale is elaborately designed to improve the quality of an explanation without any explicit annotations. Experimental results on the FEVEROUS benchmark show significant gains over previous state-of-the-art methods for both rationale extraction and fact verification. | ['Deyu Zhou', 'Yingjie Zhu', 'Jiasheng Si'] | 2022-12-02 | null | null | null | null | ['fact-verification'] | ['natural-language-processing'] | [ 2.52532870e-01 8.51002753e-01 -6.07584596e-01 -1.36634022e-01
-7.37179816e-01 -5.78833103e-01 4.91555154e-01 6.27772808e-01
4.70706433e-01 6.37062728e-01 6.74123466e-01 -5.97751677e-01
-4.37757939e-01 -9.06201720e-01 -9.42120671e-01 -4.24638838e-01
-1.77881315e-01 1.48543790e-01 2.87798613e-01 -2.45551407e-01
2.65119702e-01 7.14765638e-02 -1.16884923e+00 7.65948772e-01
8.39816868e-01 8.92764568e-01 -3.23188275e-01 3.22655469e-01
-2.10004285e-01 1.12180698e+00 -3.00364584e-01 -4.91215110e-01
4.69568633e-02 -2.42052227e-01 -9.11621153e-01 -1.21295061e-02
5.43424070e-01 -2.29660943e-01 -4.12173599e-01 1.02362716e+00
-2.16369182e-02 -1.56175882e-01 4.90869284e-01 -1.53630364e+00
-7.89535880e-01 1.26145053e+00 -7.09429622e-01 2.48808876e-01
4.24567550e-01 2.19664678e-01 1.69339073e+00 -8.59835148e-01
6.09459579e-01 1.01036394e+00 3.92741054e-01 1.79146931e-01
-8.85961890e-01 -5.92755675e-01 7.93533087e-01 3.91531527e-01
-1.10401881e+00 -4.66192188e-03 1.30184555e+00 -3.05280894e-01
7.69072473e-01 4.47517425e-01 7.92800248e-01 1.04970384e+00
8.57568458e-02 9.42157388e-01 7.13743806e-01 -6.18619844e-02
1.41227871e-01 -1.99818820e-01 4.98216897e-01 1.38492560e+00
7.24572122e-01 -9.49670561e-03 -8.66303682e-01 -1.91339657e-01
4.13752913e-01 9.44365636e-02 -4.27948058e-01 -4.92299855e-01
-1.29105818e+00 6.98039711e-01 9.40556586e-01 1.02053672e-01
-2.14460358e-01 3.13698381e-01 3.02233994e-01 9.86074805e-02
2.39474490e-01 4.70268369e-01 -6.92766726e-01 3.94031852e-01
-7.51827896e-01 7.11116865e-02 4.87292230e-01 9.80897188e-01
7.31277168e-01 8.47719982e-02 -6.18594587e-01 1.36204762e-02
4.87613887e-01 2.23588776e-02 4.79489192e-02 -3.65767211e-01
1.02022922e+00 1.50319600e+00 -2.19822660e-01 -1.53402984e+00
-6.49460852e-01 -7.95486510e-01 -9.89378572e-01 -9.81344506e-02
2.50980765e-01 1.72808275e-01 -9.22203898e-01 1.54037523e+00
5.19171357e-01 3.01802188e-01 -9.88409817e-02 1.10654879e+00
1.23959923e+00 2.35123396e-01 -1.56196892e-01 5.86203113e-02
1.58463538e+00 -1.02717006e+00 -5.18275976e-01 -2.89596349e-01
8.11058879e-01 -3.25492233e-01 1.14043581e+00 2.53963381e-01
-7.37751245e-01 -2.83226103e-01 -1.24555147e+00 -2.09056184e-01
-4.22673255e-01 -8.74446034e-02 1.10442722e+00 3.54930341e-01
-6.24486566e-01 3.91332954e-01 -3.46962005e-01 4.02833164e-01
6.30390525e-01 2.95186073e-01 -3.58543009e-01 -3.08371305e-01
-1.45566154e+00 4.95884031e-01 4.68328029e-01 1.81702241e-01
-7.98913300e-01 -8.99685323e-01 -1.06356072e+00 5.17103374e-01
1.03624928e+00 -9.55415130e-01 6.22677624e-01 -3.65762889e-01
-8.33074570e-01 2.97057360e-01 -8.13731030e-02 -2.26107165e-01
3.34514350e-01 2.83208527e-02 -5.12103677e-01 1.98253378e-01
2.23199293e-01 3.16394210e-01 8.86982918e-01 -1.38928878e+00
-2.64251113e-01 -2.26477772e-01 6.13160491e-01 -5.25657320e-03
-1.62043989e-01 -6.19717181e-01 -5.50425828e-01 -6.43383503e-01
4.67094034e-01 -7.92528570e-01 -1.16312280e-01 -2.49806523e-01
-1.20604479e+00 -1.03737079e-01 7.21869588e-01 -5.22667468e-01
1.40120423e+00 -1.91251862e+00 2.84801304e-01 5.36876202e-01
9.16643500e-01 -1.07542746e-01 -2.50521511e-01 3.07445467e-01
-2.65555471e-01 6.17327213e-01 -2.84076273e-01 9.14755315e-02
9.18452218e-02 5.35292514e-02 -4.56842482e-01 2.42256641e-01
4.93597567e-01 1.14266133e+00 -1.12763155e+00 -5.46796501e-01
-9.34004411e-03 2.68746763e-01 -7.96563506e-01 -2.10715026e-01
-4.60094124e-01 3.36778551e-01 -8.92834663e-01 8.98737073e-01
6.51016116e-01 -1.00451767e+00 2.41710886e-01 -4.68570054e-01
3.56471866e-01 7.30258882e-01 -1.20172942e+00 1.53822458e+00
-2.41240442e-01 1.94987297e-01 -2.28262782e-01 -7.84621179e-01
5.41194618e-01 3.01087767e-01 3.57683867e-01 -3.00380945e-01
-3.66523713e-02 -1.15460518e-03 6.13325313e-02 -3.60057175e-01
4.82286185e-01 1.80062920e-01 -1.95532143e-01 4.56395805e-01
-2.77662873e-01 1.53209731e-01 9.47853401e-02 7.61280477e-01
1.41587496e+00 -1.43751204e-01 4.15663719e-01 9.03825313e-02
4.85190541e-01 2.84084886e-01 6.01609468e-01 6.68752968e-01
6.39416650e-02 5.55765450e-01 8.44152927e-01 -6.50692821e-01
-4.52703029e-01 -9.10963476e-01 3.95397902e-01 6.59708500e-01
4.57871586e-01 -8.92985940e-01 -2.47124299e-01 -1.43566918e+00
2.61818141e-01 6.23237014e-01 -8.28546584e-01 -3.77334535e-01
-4.83418107e-01 -5.76806605e-01 5.19755006e-01 4.94759530e-01
4.33585614e-01 -8.26418221e-01 -2.14753926e-01 3.42356525e-02
-4.69445527e-01 -1.12062466e+00 -3.80236506e-01 1.19950458e-01
-7.62597203e-01 -1.69746423e+00 1.63324162e-01 -2.24385083e-01
9.40877736e-01 5.62321186e-01 1.18006170e+00 9.40662384e-01
-5.49345799e-02 1.61035240e-01 -2.05327749e-01 -1.06100552e-01
-3.14759128e-02 1.06943771e-01 -4.20797825e-01 -1.24124624e-01
5.03709354e-02 -5.19052505e-01 -5.25477231e-01 1.20479114e-01
-9.64699864e-01 3.91589344e-01 7.30078399e-01 9.01380420e-01
6.93853021e-01 4.03855532e-01 3.81446719e-01 -1.11213529e+00
3.61993492e-01 -5.92736721e-01 -3.08002561e-01 3.12411726e-01
-6.80518925e-01 4.49355662e-01 7.96369433e-01 -1.38623700e-01
-8.93558323e-01 -7.35120624e-02 3.29840869e-01 -2.73322612e-01
1.73956633e-01 8.26196015e-01 -4.47195888e-01 4.60289344e-02
3.32554609e-01 -7.65049979e-02 -6.62506223e-01 -1.08582817e-01
7.34764397e-01 -9.78627279e-02 2.50565082e-01 -5.46355128e-01
1.19421983e+00 5.70498049e-01 4.19852406e-01 -2.13545442e-01
-1.26969695e+00 -2.98303485e-01 -5.49365640e-01 -1.76296294e-01
5.67525864e-01 -7.96974719e-01 -7.91323960e-01 -3.84840041e-01
-1.23800325e+00 9.29326341e-02 -1.70242310e-01 2.70305991e-01
-8.35054070e-02 3.13288569e-01 -4.75386411e-01 -4.54428166e-01
-1.32725820e-01 -1.06029224e+00 1.21226251e+00 -1.16021022e-01
-3.10289353e-01 -9.23316240e-01 -2.31611177e-01 6.95587158e-01
-2.11308956e-01 3.91167372e-01 1.43123353e+00 -7.37184703e-01
-1.34480441e+00 -1.80258468e-01 -4.79044408e-01 -4.40097302e-01
2.07993522e-01 -2.38684881e-02 -8.89363945e-01 1.12134151e-01
-3.32470328e-01 -1.23072252e-01 1.23415756e+00 1.54208466e-01
1.28073323e+00 -8.70338678e-01 -5.92899859e-01 4.38370883e-01
1.22727406e+00 -5.76857865e-01 2.49506339e-01 2.76472449e-01
1.30074120e+00 4.78037387e-01 5.28590620e-01 3.47046554e-01
8.40984523e-01 3.62437993e-01 1.04656720e+00 -2.44312748e-01
-4.03242052e-01 -5.29311836e-01 1.03736512e-01 6.41600847e-01
-8.31370358e-04 -3.34352106e-01 -8.59588861e-01 4.50327903e-01
-2.03535271e+00 -9.34189737e-01 -7.61503518e-01 1.66948712e+00
6.06638730e-01 4.49237466e-01 -1.53540388e-01 4.81489867e-01
5.81030130e-01 5.07613540e-01 -3.44618589e-01 -2.96222903e-02
-1.75556228e-01 -4.73757565e-01 2.21020713e-01 4.76026177e-01
-8.51632774e-01 7.21831739e-01 5.20924520e+00 6.10808730e-01
-7.07754791e-01 -5.60341403e-02 4.33773965e-01 -2.00666804e-02
-1.19903135e+00 4.04683769e-01 -5.48401475e-01 2.39354521e-01
1.83919445e-01 1.96470693e-01 2.88504958e-01 7.05275893e-01
-2.82159261e-02 3.80005725e-02 -1.15236998e+00 3.16966653e-01
-1.39421567e-01 -1.72321069e+00 4.41098303e-01 -3.33209634e-02
8.46211910e-01 -3.45670670e-01 4.92424592e-02 2.20569655e-01
3.59035790e-01 -8.49663973e-01 6.93811119e-01 3.72768015e-01
4.42876518e-01 -5.68311572e-01 6.96403682e-01 2.57514477e-01
-1.63216889e+00 -1.18760630e-01 -2.14523509e-01 -1.29453942e-01
-3.23641673e-02 1.17836833e+00 -9.92159426e-01 1.17275822e+00
2.86795020e-01 7.25256920e-01 -7.50738740e-01 7.00053990e-01
-9.00037825e-01 5.95781446e-01 -1.18795753e-01 5.33126220e-02
3.02732438e-01 2.18031585e-01 7.38569677e-01 1.03511393e+00
7.60857686e-02 2.46479020e-01 2.15192094e-01 1.01585817e+00
-4.71015543e-01 -2.54398167e-01 -7.93691874e-01 -4.22056355e-02
2.80280143e-01 1.21831250e+00 -9.55991924e-01 -3.87968630e-01
-4.49226767e-01 6.62328243e-01 6.48811936e-01 3.95900637e-01
-9.55918550e-01 1.11773275e-01 3.22694421e-01 1.63777322e-01
3.42476159e-01 2.07619146e-02 -8.64462912e-01 -1.27842343e+00
3.66245449e-01 -6.42681658e-01 8.03050280e-01 -7.14254916e-01
-1.15855455e+00 3.93569678e-01 -1.08830385e-01 -1.32466662e+00
1.91031918e-01 -2.81835109e-01 -7.29824364e-01 5.42337716e-01
-1.99780726e+00 -1.39840913e+00 -3.70919645e-01 5.24053216e-01
3.23096395e-01 1.82130694e-01 4.50318664e-01 -5.16261067e-03
-5.60028613e-01 4.31479156e-01 -7.69596756e-01 9.99543536e-03
2.91024476e-01 -1.41186965e+00 2.43619695e-01 1.05602682e+00
4.25956756e-01 8.27237368e-01 4.99557793e-01 -1.09976959e+00
-1.66300476e+00 -1.22808814e+00 1.03419578e+00 -5.66679597e-01
8.23772430e-01 -2.12884828e-01 -1.00687778e+00 7.69385576e-01
-1.50533080e-01 3.45369995e-01 5.45303047e-01 5.91981411e-01
-9.72618163e-01 1.15575872e-01 -8.05681586e-01 9.18778241e-01
1.29887831e+00 -6.16841614e-01 -7.08255529e-01 3.77309173e-01
1.10585856e+00 -4.00810450e-01 -5.93741119e-01 5.80513656e-01
3.28086048e-01 -1.08337045e+00 1.07161510e+00 -7.91970313e-01
7.76923299e-01 -6.56259716e-01 1.49043053e-01 -1.09466791e+00
-4.67872769e-01 -5.69744468e-01 -7.65045047e-01 1.16869211e+00
7.46554554e-01 -2.19500363e-01 8.31341505e-01 4.81222928e-01
-3.14370215e-01 -9.19243336e-01 -7.69509733e-01 -4.54820305e-01
-6.10767722e-01 -5.37254512e-01 1.02515495e+00 1.15885949e+00
5.03306687e-01 5.64751148e-01 -4.93941307e-01 7.58920133e-01
5.99057794e-01 7.29287386e-01 5.83938599e-01 -1.29599679e+00
-3.09715152e-01 -3.93437535e-01 -1.64519832e-01 -8.30598891e-01
1.61083177e-01 -1.32527781e+00 -1.35981530e-01 -1.97399485e+00
2.73389041e-01 -2.97508091e-01 -4.84934807e-01 9.05635595e-01
-6.33570313e-01 -9.32197422e-02 2.05190301e-01 7.02573583e-02
-7.75125623e-01 5.40172160e-01 1.63724852e+00 -6.12315834e-01
-1.58921704e-02 -2.04218492e-01 -1.11245835e+00 9.12571430e-01
5.83552361e-01 -7.80032933e-01 -7.18148947e-01 -3.75414848e-01
8.43206406e-01 1.92297965e-01 7.37998128e-01 -5.45567036e-01
3.60684156e-01 -2.71975696e-01 3.62817705e-01 -6.50521278e-01
1.45342216e-01 -9.44149077e-01 -5.44637023e-03 3.39904994e-01
-4.32227015e-01 -1.16114870e-01 1.32911503e-01 9.90982175e-01
-1.80750698e-01 2.35085025e-01 9.41294506e-02 -3.76347601e-02
-4.66729939e-01 4.54158753e-01 3.56379263e-02 1.23398729e-01
5.62337339e-01 -9.92077366e-02 -9.16414201e-01 -1.63628727e-01
-5.66294014e-01 3.71337950e-01 1.89691246e-01 3.52634668e-01
8.80542099e-01 -1.35196459e+00 -4.46556717e-01 9.55342576e-02
4.23616827e-01 1.74275488e-01 3.84084821e-01 9.38128114e-01
2.30428964e-01 4.40559655e-01 1.72102168e-01 -3.24576467e-01
-1.04312217e+00 8.43518317e-01 -1.06539473e-01 -8.14132035e-01
-8.29008639e-01 7.78496742e-01 3.51789743e-01 -2.37796709e-01
5.77867664e-02 -7.64121115e-01 -2.37007171e-01 -8.12202469e-02
2.24268004e-01 2.23594844e-01 1.76209494e-01 -2.52952784e-01
-6.10033810e-01 3.68237555e-01 -2.91539379e-03 4.00088191e-01
1.19977295e+00 -2.16190089e-02 -5.30710109e-02 2.05786929e-01
7.21226692e-01 5.06874144e-01 -1.09595561e+00 -5.21716848e-03
2.16720313e-01 -4.46538776e-01 6.15486167e-02 -9.32355464e-01
-1.21214318e+00 7.12171853e-01 -5.11464417e-01 4.61252242e-01
9.13282275e-01 1.25324562e-01 5.85448325e-01 3.58854979e-01
1.54361233e-01 -6.57259643e-01 3.82676333e-01 3.28539670e-01
1.00240397e+00 -1.31643474e+00 2.87714690e-01 -1.05462062e+00
-6.83955193e-01 1.11799729e+00 6.52548015e-01 1.66573361e-01
5.57455361e-01 1.61382943e-01 -4.09805357e-01 -9.57417667e-01
-8.30747366e-01 -1.12046279e-01 8.77772450e-01 3.18363845e-01
1.76126093e-01 4.38583530e-02 1.96919385e-02 1.08758032e+00
-9.35579613e-02 -4.00701880e-01 5.45723081e-01 5.40624857e-01
-3.32815409e-01 -7.64756501e-01 -1.33807853e-01 4.98759300e-01
-2.11672604e-01 -5.69804072e-01 -7.99423277e-01 9.59685981e-01
1.51276916e-01 1.04618859e+00 -5.58917880e-01 -5.50021291e-01
2.39725962e-01 -3.71499173e-02 5.45198433e-02 -6.93051994e-01
-7.21402943e-01 -2.01125160e-01 4.59647685e-01 -7.94883549e-01
-2.69991010e-01 -1.83171272e-01 -1.78672945e+00 -2.64833629e-01
-4.74977255e-01 1.28760010e-01 2.76901752e-01 1.28361440e+00
4.82160985e-01 1.10811913e+00 3.80233377e-01 -1.06913656e-01
-5.60267270e-02 -4.32129025e-01 -4.84349221e-01 2.00367048e-01
7.11560428e-01 -7.89376974e-01 -5.35500884e-01 -2.47022837e-01] | [9.041824340820312, 7.718967437744141] |
82f48131-d8e9-4081-a69d-e77181a11439 | denoising-diffusion-post-processing-for-low | 2303.09627 | null | https://arxiv.org/abs/2303.09627v2 | https://arxiv.org/pdf/2303.09627v2.pdf | Denoising Diffusion Post-Processing for Low-Light Image Enhancement | Low-light image enhancement (LLIE) techniques attempt to increase the visibility of images captured in low-light scenarios. However, as a result of enhancement, a variety of image degradations such as noise and color bias are revealed. Furthermore, each particular LLIE approach may introduce a different form of flaw within its enhanced results. To combat these image degradations, post-processing denoisers have widely been used, which often yield oversmoothed results lacking detail. We propose using a diffusion model as a post-processing approach, and we introduce Low-light Post-processing Diffusion Model (LPDM) in order to model the conditional distribution between under-exposed and normally-exposed images. We apply LPDM in a manner which avoids the computationally expensive generative reverse process of typical diffusion models, and post-process images in one pass through LPDM. Extensive experiments demonstrate that our approach outperforms competing post-processing denoisers by increasing the perceptual quality of enhanced low-light images on a variety of challenging low-light datasets. Source code is available at https://github.com/savvaki/LPDM. | ['Anna S. Bosman', 'Savvas Panagiotou'] | 2023-03-16 | null | null | null | null | ['image-enhancement', 'low-light-image-enhancement'] | ['computer-vision', 'computer-vision'] | [ 5.72945654e-01 -3.90371650e-01 4.96266484e-01 -2.53492892e-01
-5.12718022e-01 -3.39374900e-01 6.99652076e-01 -8.83392990e-02
-2.41437823e-01 5.94161630e-01 1.78971797e-01 -4.74786796e-02
1.35244653e-01 -9.47625756e-01 -6.10753715e-01 -1.06349492e+00
4.02660608e-01 -4.97226626e-01 3.34661126e-01 -9.76217911e-02
2.48187691e-01 4.43954200e-01 -1.62612224e+00 2.87521392e-01
1.01814020e+00 7.15595901e-01 4.10052359e-01 8.18809330e-01
7.19551817e-02 7.48857200e-01 -4.29181933e-01 -3.97671431e-01
5.30808747e-01 -5.79241335e-01 -2.06628039e-01 3.89767379e-01
4.93754476e-01 -3.90182555e-01 -2.83475935e-01 1.53062427e+00
5.77249944e-01 1.63393661e-01 4.87403274e-01 -8.74103129e-01
-8.41448665e-01 -1.20563559e-01 -9.64703321e-01 4.50724870e-01
3.44924241e-01 7.14807510e-01 4.53025877e-01 -1.10770214e+00
4.89703923e-01 1.27200711e+00 5.95901251e-01 2.99731255e-01
-1.57493281e+00 -5.25568962e-01 -2.77047485e-01 1.41116306e-01
-1.05496669e+00 -7.19469130e-01 9.19281542e-01 -1.35157123e-01
4.72732663e-01 1.99007854e-01 4.92129326e-01 9.61510956e-01
7.14514017e-01 4.10835594e-01 1.82667565e+00 -5.54049730e-01
2.09751859e-01 -5.45827262e-02 -1.52497113e-01 4.44502324e-01
4.22398239e-01 5.10200143e-01 -5.32542050e-01 3.77992280e-02
4.99782056e-01 3.60684693e-02 -5.08923054e-01 -1.34467140e-01
-8.50420475e-01 1.92514658e-01 4.36803967e-01 7.00282604e-02
-6.28929734e-01 4.45386656e-02 -9.20350477e-02 1.55402869e-01
8.88520896e-01 4.51044738e-02 4.96021770e-02 1.98478699e-01
-8.62745941e-01 9.73638296e-02 1.98965058e-01 4.45298642e-01
9.60661769e-01 1.02680428e-02 -3.14266860e-01 1.06532037e+00
4.08803254e-01 6.55108273e-01 7.42483810e-02 -1.14997113e+00
2.10386366e-01 1.47062600e-01 2.46474221e-01 -9.68126178e-01
-6.09364621e-02 -3.98883313e-01 -1.04400957e+00 8.68261695e-01
3.58630478e-01 1.36386186e-01 -1.10298049e+00 1.39421201e+00
4.51248437e-01 6.93138614e-02 1.05335303e-01 1.01007390e+00
5.44260800e-01 9.66563225e-01 1.17957711e-01 -4.56229359e-01
1.19271350e+00 -8.35384190e-01 -9.67819571e-01 -3.38497341e-01
-1.25331819e-01 -1.21471715e+00 1.13328540e+00 6.83451295e-01
-1.55329692e+00 -6.71141505e-01 -9.51188803e-01 -2.77740836e-01
-1.07130043e-01 -1.14497446e-01 2.16957822e-01 7.60719240e-01
-1.11297500e+00 5.20178437e-01 -6.26347542e-01 -8.00749883e-02
4.30379361e-01 -1.96711734e-01 -9.60417464e-02 -7.00679481e-01
-8.75497460e-01 8.56046498e-01 -7.81374401e-04 3.62396717e-01
-1.00610566e+00 -8.32850754e-01 -7.38702595e-01 -2.86508471e-01
3.93743575e-01 -6.64740920e-01 8.81153524e-01 -7.60669291e-01
-1.42887759e+00 8.57239842e-01 -4.10487026e-01 -7.23307729e-02
6.13098025e-01 -1.59388334e-01 -5.48610866e-01 2.12762281e-01
-9.29654203e-03 3.48082721e-01 1.45294189e+00 -1.89883292e+00
-3.92132193e-01 -1.40384734e-01 -7.16748387e-02 2.79118747e-01
4.91505712e-02 3.21278498e-02 -5.63933909e-01 -7.74706125e-01
8.86981487e-02 -7.12126911e-01 -1.76620483e-01 1.63358659e-01
-3.12022507e-01 3.38458449e-01 8.41280341e-01 -6.88839853e-01
1.09714007e+00 -2.38066411e+00 -3.64572674e-01 -3.15014757e-02
4.90802467e-01 4.41725701e-01 -1.91465974e-01 2.68986583e-01
2.10764091e-02 -9.73530263e-02 -6.51557446e-01 -5.70733249e-01
-3.19468528e-01 2.50733346e-01 -9.37498584e-02 7.46398509e-01
1.83643594e-01 6.67873085e-01 -1.00286794e+00 -3.21273148e-01
5.36850512e-01 8.84461403e-01 -2.95318812e-01 2.27449328e-01
-3.74635980e-02 4.81415600e-01 1.42982647e-01 7.00701118e-01
1.21579957e+00 3.72001342e-02 -2.05955163e-01 -5.47340035e-01
-3.75354290e-01 -1.43703461e-01 -1.05306280e+00 1.37102056e+00
-6.16071880e-01 7.83861339e-01 2.14325652e-01 -2.93575227e-01
7.43842244e-01 -1.74583700e-02 2.26552874e-01 -9.49388862e-01
1.40617520e-01 1.62716478e-01 -1.92609161e-01 -6.04030252e-01
5.70354104e-01 -4.96802807e-01 6.14326954e-01 5.00672340e-01
-4.24974591e-01 -3.37895334e-01 2.94730663e-01 1.00928433e-01
8.99484634e-01 1.19004294e-01 7.76442662e-02 1.98573731e-02
5.17780662e-01 -4.25298661e-01 5.12374103e-01 7.29633868e-01
-3.56835812e-01 1.03845310e+00 -5.06224856e-02 -1.40360922e-01
-1.10787189e+00 -1.19108450e+00 -2.23479390e-01 6.43073499e-01
3.79030347e-01 -1.09685600e-01 -6.76092923e-01 -2.00068638e-01
-3.71269822e-01 9.14267421e-01 -3.97110552e-01 -2.11763769e-01
-2.97680855e-01 -1.17445600e+00 1.88625172e-01 -1.08183451e-01
8.17174852e-01 -9.31428194e-01 -6.43401384e-01 8.15232024e-02
-3.00704688e-01 -1.14020491e+00 -3.50975633e-01 1.67143151e-01
-6.76293790e-01 -1.04159641e+00 -9.36197579e-01 -4.09357697e-01
9.03160930e-01 8.05723011e-01 1.08489263e+00 1.21029548e-01
-5.14201820e-01 3.72895151e-01 -2.71764815e-01 -2.31865019e-01
-7.30773509e-01 -7.18510926e-01 -1.29471362e-01 5.21845579e-01
1.94265053e-01 -4.92999345e-01 -1.14648485e+00 2.12176949e-01
-1.35528851e+00 2.05439523e-01 7.34660804e-01 7.48478651e-01
6.64703488e-01 7.29097962e-01 2.09340360e-02 -6.57925010e-01
9.56178606e-01 -2.09304139e-01 -6.16709173e-01 4.56062146e-02
-1.02794242e+00 -1.70758486e-01 4.23927635e-01 -3.73513460e-01
-1.76891637e+00 -3.16943288e-01 5.08633815e-02 -3.58037591e-01
-3.94075900e-01 4.13658023e-02 -1.11532301e-01 -2.70462513e-01
6.69727445e-01 3.59170467e-01 6.38400912e-02 -5.20456910e-01
3.89854163e-01 5.44425607e-01 5.99452198e-01 -2.10866570e-01
1.06548989e+00 8.56811523e-01 2.34289587e-01 -9.68096197e-01
-8.20707977e-01 -3.19887042e-01 -2.97662377e-01 -6.26749337e-01
7.82257915e-01 -8.47396374e-01 -3.78842294e-01 8.56061995e-01
-1.00737369e+00 -5.80331326e-01 -3.48988801e-01 2.18393117e-01
-2.36427978e-01 6.92035556e-01 -5.90213358e-01 -9.50473070e-01
-1.44199044e-01 -1.10039997e+00 8.40536118e-01 4.59553838e-01
2.63093531e-01 -9.32961047e-01 2.38450561e-02 3.19003165e-01
6.12270057e-01 1.45826608e-01 7.54900038e-01 6.42249584e-01
-6.57441318e-01 1.45226372e-02 -5.70328176e-01 8.99631023e-01
3.10258687e-01 3.82084586e-02 -1.14373899e+00 -3.97005737e-01
3.58113021e-01 -6.46871747e-03 8.40144753e-01 7.01846242e-01
7.69307673e-01 5.76502550e-03 6.91668987e-02 8.01010013e-01
1.73486090e+00 1.09152138e-01 1.06175506e+00 4.57848489e-01
5.15591800e-01 6.49726033e-01 6.14526510e-01 2.44110599e-01
1.92904860e-01 5.56023657e-01 5.83613396e-01 -6.39944196e-01
-8.93806040e-01 -7.06330920e-03 5.09326100e-01 4.15090799e-01
-2.15882778e-01 -4.42570090e-01 -6.20241702e-01 4.72105414e-01
-1.27375889e+00 -9.54054773e-01 -6.41337812e-01 2.14252949e+00
8.66664588e-01 -5.03963791e-02 -1.97263449e-01 3.24963599e-01
6.60519540e-01 3.34057689e-01 -5.14759481e-01 -8.58333409e-02
-5.26119173e-01 1.22772515e-01 6.30980790e-01 7.58698702e-01
-6.83645248e-01 6.01449251e-01 6.05278921e+00 6.66234970e-01
-9.66727674e-01 2.96926290e-01 7.14813650e-01 -8.27563778e-02
-3.35008055e-01 5.73271103e-02 -4.60454434e-01 7.21357942e-01
5.34683168e-01 2.48391889e-02 5.34296811e-01 1.14962123e-01
7.87613571e-01 -7.05274105e-01 -4.67060626e-01 1.16024530e+00
1.87950179e-01 -7.97646403e-01 -1.13012806e-01 1.55950069e-01
8.43302608e-01 -7.37857595e-02 3.44778240e-01 -3.61701578e-01
1.55662090e-01 -5.33802211e-01 6.58305049e-01 7.13211000e-01
7.80781746e-01 -4.98620331e-01 4.33693767e-01 8.51933360e-02
-8.88319850e-01 -2.39928123e-02 -4.24506038e-01 1.51346818e-01
6.08715177e-01 1.18305969e+00 -3.21198851e-01 5.41129231e-01
9.77818489e-01 6.78854942e-01 -7.27988839e-01 1.18764222e+00
-5.86390972e-01 4.74061906e-01 -1.40471280e-01 5.94461918e-01
-6.61380365e-02 -6.00556850e-01 8.01464498e-01 1.17058551e+00
3.24964613e-01 1.53482243e-01 -2.16308355e-01 1.01123714e+00
1.24761097e-01 -3.08968872e-01 -4.11535710e-01 2.75458306e-01
-7.91860595e-02 1.26620245e+00 -7.77054489e-01 -1.51215687e-01
-4.61492270e-01 1.33779085e+00 -4.40210611e-01 7.39653170e-01
-6.60578072e-01 -3.15126568e-01 5.64242184e-01 4.92373794e-01
-1.11750528e-01 -1.81066230e-01 -3.23170811e-01 -1.07448030e+00
1.88385472e-01 -9.10959542e-01 9.21724588e-02 -1.41354609e+00
-1.24272656e+00 5.40030897e-01 -3.64722535e-02 -1.29440355e+00
3.57095540e-01 -2.93426871e-01 -5.10854840e-01 1.02824485e+00
-2.17829061e+00 -1.00759149e+00 -6.72956705e-01 7.41355836e-01
7.36024022e-01 2.97979593e-01 9.29451138e-02 5.43209255e-01
-4.02650803e-01 -9.44062993e-02 1.77833721e-01 -4.79706734e-01
9.03808177e-01 -1.05285943e+00 1.41377062e-01 1.55418372e+00
-1.44460589e-01 4.11957771e-01 9.88851905e-01 -7.14749157e-01
-1.18151379e+00 -9.81275260e-01 5.91806173e-01 -1.60546124e-01
3.07823062e-01 -6.87492564e-02 -1.06253719e+00 1.77453935e-01
5.26029825e-01 1.00024134e-01 3.53607565e-01 -4.85074580e-01
-7.80636296e-02 -2.24195525e-01 -1.21051824e+00 7.06553876e-01
8.30058515e-01 -4.96139288e-01 -2.63859421e-01 2.47093663e-02
2.49278635e-01 -2.46466249e-01 -5.44623017e-01 2.53996730e-01
3.53747636e-01 -1.49816382e+00 1.21293879e+00 3.57889444e-01
6.17878497e-01 -7.59009600e-01 -3.11017949e-02 -1.41085112e+00
-3.73473048e-01 -7.52864540e-01 -1.43697038e-01 1.32697737e+00
5.44856489e-02 -7.07026720e-01 3.07548046e-01 4.84881312e-01
-1.02875298e-02 -3.68502051e-01 -3.91875625e-01 -6.75371408e-01
-4.81622040e-01 -5.07471383e-01 1.94143966e-01 6.55135453e-01
-7.45016098e-01 9.26556811e-03 -5.55790484e-01 3.84986311e-01
1.36688519e+00 -1.05798490e-01 7.65019476e-01 -7.97711790e-01
-2.31848389e-01 -2.60690331e-01 5.45785092e-02 -8.75546992e-01
-3.16469431e-01 -5.94046533e-01 3.03613693e-01 -1.70835376e+00
2.24794403e-01 -3.70449185e-01 -3.24599683e-01 2.11635098e-01
-3.92905116e-01 8.11749041e-01 1.54292464e-01 3.84750307e-01
-1.91329762e-01 5.61320722e-01 1.49254715e+00 -1.54603213e-01
-1.89085320e-01 -1.06814943e-01 -5.94018936e-01 7.45729446e-01
6.67457402e-01 -6.24127269e-01 -4.81287211e-01 -5.54632127e-01
3.98241252e-01 -3.27977061e-01 6.95323110e-01 -9.65561271e-01
1.59686416e-01 -1.40322417e-01 5.60831010e-01 -4.27741051e-01
4.35499519e-01 -9.04711366e-01 3.03735882e-01 4.27745193e-01
3.80230844e-02 -1.97807446e-01 1.33827791e-01 6.93759978e-01
-2.47257397e-01 -1.75553545e-01 1.30047894e+00 -1.48266301e-01
-5.46873748e-01 1.12420224e-01 -5.30418634e-01 -2.95507282e-01
8.73934627e-01 -4.34654921e-01 -5.05051136e-01 -3.96112382e-01
-4.41399038e-01 -2.44512513e-01 7.78356254e-01 1.93788618e-01
7.83883274e-01 -9.67510045e-01 -6.41714990e-01 3.41032684e-01
-4.57602851e-02 -2.00430036e-01 7.23381102e-01 1.11033344e+00
-5.69214642e-01 -2.90809393e-01 -1.75823346e-01 -5.21773458e-01
-1.38871348e+00 5.03929198e-01 2.39923641e-01 -8.90851244e-02
-9.84475553e-01 6.45293593e-01 4.37852412e-01 1.56304419e-01
-1.97687998e-01 -2.13518187e-01 1.31692991e-01 -1.84519932e-01
7.13577986e-01 6.35759592e-01 1.56014219e-01 -6.04520559e-01
2.77972985e-02 6.35304749e-01 -2.57636067e-02 -4.65924591e-02
1.47679758e+00 -9.28506076e-01 -2.25893572e-01 3.02355736e-01
8.56076419e-01 2.86684424e-01 -1.65088093e+00 -2.21229881e-01
-4.32033509e-01 -1.04018116e+00 6.43485844e-01 -8.35881114e-01
-1.10837984e+00 7.77632236e-01 9.26974893e-01 3.47423613e-01
1.76131725e+00 -3.51648867e-01 7.77926266e-01 -2.96766073e-01
6.57874569e-02 -9.72860694e-01 1.83080658e-01 -9.61642787e-02
8.07044923e-01 -1.29118514e+00 2.70313650e-01 -5.18278301e-01
-3.97397935e-01 9.10770595e-01 2.80756950e-01 1.09449051e-01
5.41737974e-01 4.19057310e-01 4.40658748e-01 -3.01757246e-01
-5.85677922e-01 -3.79857779e-01 3.15389559e-02 8.09945703e-01
2.39798501e-01 -3.75866860e-01 -2.55803019e-01 -1.00511752e-01
3.04394037e-01 8.04916695e-02 8.66183758e-01 8.76271784e-01
-3.03048432e-01 -1.04159391e+00 -8.93421888e-01 6.01725914e-02
-6.77803993e-01 -2.64058679e-01 -3.17624509e-02 3.12892377e-01
2.46171698e-01 1.32304764e+00 -1.86931476e-01 -4.83889617e-02
2.81527758e-01 -2.91650265e-01 4.65123862e-01 -2.55894631e-01
-2.90597290e-01 4.32194918e-01 -3.04531753e-01 -6.68980062e-01
-7.69936204e-01 -5.09654641e-01 -6.05742335e-01 -3.43737930e-01
-3.19181979e-01 -4.02504385e-01 6.98915422e-01 5.01482785e-01
2.37789482e-01 7.10103214e-01 7.01936126e-01 -1.18191397e+00
-3.15322503e-02 -7.77664185e-01 -7.68639445e-01 6.49601936e-01
6.44874334e-01 -6.71542645e-01 -7.05739796e-01 4.09433722e-01] | [10.834832191467285, -2.5727505683898926] |
d2ca233e-8850-4ae8-86a4-187e7e175685 | opinion-based-relational-pivoting-for-cross | null | null | https://openreview.net/forum?id=38nXXqT4hC6 | https://openreview.net/pdf?id=38nXXqT4hC6 | Opinion-based Relational Pivoting for Cross-domain Aspect Term Extraction | Domain adaptation methods often exploit domain-transferable input features, a.k.a. pivots.
The task of Aspect and Opinion Term Extraction presents a special challenge for domain transfer: while opinion terms largely transfer across domains, aspects change drastically from one domain to another (e.g. from \textit{restaurants} to \textit{laptops}).
In this paper, we investigate and establish empirically a prior conjecture, which suggests that the linguistic relations connecting opinion terms to their aspects transfer well across domains and therefore can be leveraged for cross-domain aspect term extraction.
We present several analyses supporting this conjecture, via experiments with four linguistic dependency formalisms to represent relation patterns.
Following, we present an aspect term extraction method that drives models to consider opinion--aspect relations via explicit multitask objectives.
This method provides significant performance gains, even on top of a prior state-of-the-art linguistically-informed model, which are shown in analysis to stem from the relational pivoting signal.
| ['Anonymous'] | 2021-05-16 | null | null | null | acl-arr-may-2021-5 | ['term-extraction'] | ['natural-language-processing'] | [ 2.49940649e-01 3.28200817e-01 -5.75336754e-01 -7.03522921e-01
-1.07721317e+00 -1.04205215e+00 1.09150326e+00 4.46853489e-01
-3.32785070e-01 9.36245382e-01 4.50639725e-01 -5.17985940e-01
-3.25664729e-01 -6.78602934e-01 -7.21743047e-01 -3.62332702e-01
-5.30424193e-02 8.87508273e-01 1.17785484e-02 -7.17840612e-01
1.31967291e-01 8.14922601e-02 -1.13310790e+00 6.02922440e-01
7.54926264e-01 8.75679314e-01 -3.56681079e-01 2.22755104e-01
-6.29198909e-01 6.96192086e-01 -7.99435854e-01 -6.69163108e-01
4.98006605e-02 -2.24079236e-01 -1.14354801e+00 1.00986294e-01
3.27048510e-01 1.85865298e-01 2.85219401e-01 7.73486376e-01
3.42296094e-01 2.37567768e-01 1.26201844e+00 -1.25981462e+00
-1.23581815e+00 7.79863238e-01 -4.86420155e-01 2.32838005e-01
4.83204395e-01 -2.31878683e-01 1.55695915e+00 -8.56271863e-01
9.54333007e-01 1.24415946e+00 5.61168790e-01 2.70675391e-01
-1.59165001e+00 -4.92873013e-01 7.72803545e-01 5.38986921e-03
-1.04218209e+00 -5.13238966e-01 7.42628634e-01 -6.75254703e-01
1.43441129e+00 -4.49097604e-02 4.29461628e-01 1.28654480e+00
8.28391239e-02 8.84052932e-01 1.47021317e+00 -4.55715597e-01
1.76423803e-01 6.82468832e-01 4.20155972e-01 2.99308360e-01
4.94136930e-01 -3.80109213e-02 -1.05663669e+00 -1.45463482e-01
2.84356683e-01 -5.92429101e-01 -8.23869184e-02 -5.35300553e-01
-1.04751539e+00 8.12517762e-01 1.33817732e-01 5.33712208e-01
-1.91322967e-01 -3.28185081e-01 4.23311234e-01 7.38918602e-01
1.01726425e+00 6.71694875e-01 -1.25211799e+00 2.17308328e-02
-6.41640425e-01 4.22690272e-01 1.24499369e+00 1.32651985e+00
1.09678674e+00 -3.15186620e-01 -9.26172808e-02 9.44230735e-01
2.21026972e-01 6.18442118e-01 2.84268051e-01 -6.63456261e-01
7.52632260e-01 6.00665510e-01 8.65901187e-02 -7.16098070e-01
-3.35851401e-01 -2.53321439e-01 -3.10862273e-01 -5.37496433e-02
3.25833827e-01 -3.88173908e-01 -8.24189723e-01 1.75393343e+00
3.30298632e-01 -4.39124882e-01 1.63952351e-01 5.82905650e-01
7.85577834e-01 2.18380749e-01 4.82579291e-01 -4.04278412e-02
1.99077165e+00 -7.20182717e-01 -6.55235887e-01 -7.46628463e-01
5.90178251e-01 -7.46974349e-01 1.09298146e+00 2.16023445e-01
-8.39491010e-01 -4.21361655e-01 -8.36710095e-01 -2.36623496e-01
-8.74794900e-01 -1.36394262e-01 9.41719055e-01 4.97123927e-01
-9.61685181e-01 3.26044887e-01 -6.10702515e-01 -6.09416366e-01
2.53477544e-01 3.48791718e-01 -5.66952348e-01 2.01076549e-02
-1.54730535e+00 1.14264488e+00 3.32851887e-01 -4.95200187e-01
-3.36089730e-01 -8.48254442e-01 -1.09468114e+00 -2.07784578e-01
3.99000615e-01 -1.10424793e+00 1.38439965e+00 -1.27716899e+00
-1.39151716e+00 1.27873433e+00 -6.02846980e-01 -2.21806049e-01
-9.38577652e-02 -5.22376418e-01 -6.10760689e-01 -5.32382689e-02
5.37732244e-01 5.10727823e-01 9.37334776e-01 -1.34148657e+00
-8.97178113e-01 -4.62296009e-01 5.55992186e-01 2.84586996e-01
-5.02135336e-01 3.07358325e-01 -1.51045009e-01 -7.71491885e-01
-2.71963090e-01 -9.46576476e-01 4.48931232e-02 -2.29566723e-01
-2.76867896e-01 -4.70199883e-01 4.12998945e-01 -3.74625981e-01
1.05541742e+00 -1.73566926e+00 3.35964411e-01 1.42136812e-01
1.65791940e-02 6.82574585e-02 -1.05460405e-01 4.36035335e-01
-3.29054862e-01 2.40228355e-01 -3.48864019e-01 -3.55647892e-01
3.06960762e-01 2.66819596e-01 -4.14590180e-01 1.56746238e-01
7.55810082e-01 9.70093846e-01 -6.88024163e-01 -4.19127584e-01
-2.16198057e-01 3.57848614e-01 -5.91353774e-01 5.82739785e-02
-4.32967544e-01 3.48359466e-01 -6.48082674e-01 6.26230776e-01
5.47662854e-01 -3.76520425e-01 1.51160389e-01 -2.91565835e-01
-5.05812131e-02 9.30781305e-01 -6.81934357e-01 1.92675972e+00
-8.42432320e-01 5.86815000e-01 9.12105665e-02 -1.16833961e+00
1.01176262e+00 3.40872794e-01 4.71099406e-01 -5.52696228e-01
-3.23211215e-02 1.55207008e-01 -1.98350668e-01 -4.54986274e-01
6.97763562e-01 -6.21241271e-01 -4.85324562e-01 6.21656299e-01
5.02744317e-01 -4.88726974e-01 3.58391970e-01 1.23007454e-01
8.61732543e-01 4.26066339e-01 6.13808572e-01 -6.49774313e-01
6.51260376e-01 4.12337452e-01 3.17799062e-01 3.51147652e-01
-8.96518379e-02 2.52322882e-01 7.05999911e-01 -2.87849963e-01
-6.95950389e-01 -1.04587865e+00 -3.07592243e-01 1.76922524e+00
-7.04223663e-02 -4.19000328e-01 -3.50267708e-01 -9.99706686e-01
2.79781222e-01 9.60058928e-01 -6.97282672e-01 -6.48406073e-02
-4.81800824e-01 -6.10081911e-01 2.59064138e-01 4.66378003e-01
1.81230068e-01 -9.07080352e-01 -8.36027935e-02 2.27478191e-01
-2.86208123e-01 -1.34588635e+00 -1.63007155e-01 5.64608991e-01
-6.33540392e-01 -6.97564483e-01 -6.73855424e-01 -7.76443124e-01
2.81677306e-01 -2.68542860e-02 1.95485795e+00 -6.17599189e-01
2.83095270e-01 5.80585241e-01 -5.23764849e-01 -7.09659219e-01
-2.04690352e-01 6.72091186e-01 -1.52787762e-02 -3.67139071e-01
1.05799532e+00 -8.59403253e-01 -4.28696632e-01 2.94087470e-01
-7.78589785e-01 -4.27146256e-01 7.09353566e-01 5.46367943e-01
5.09305000e-01 -3.16645801e-01 6.09670341e-01 -1.39255822e+00
9.39250112e-01 -8.16739142e-01 -3.30159992e-01 2.46070519e-01
-6.23328269e-01 3.33116859e-01 3.87731701e-01 -3.75112563e-01
-1.31665623e+00 -3.07267010e-01 7.12184161e-02 2.87414193e-01
-4.21122283e-01 7.63739049e-01 -2.21325427e-01 3.05408210e-01
8.97184491e-01 -1.98046006e-02 -3.20908964e-01 -3.56537968e-01
7.97139108e-01 7.12163329e-01 1.26520559e-01 -1.10923588e+00
8.43318701e-01 6.20656729e-01 -2.40323931e-01 -9.08612192e-01
-1.45152080e+00 -5.98852813e-01 -7.56404996e-01 3.16637993e-01
8.15521121e-01 -1.16747499e+00 -1.12389907e-01 7.24220723e-02
-1.24745548e+00 -1.49351969e-01 -4.58342582e-01 2.31675848e-01
-4.93487507e-01 -2.20686197e-02 -4.44901973e-01 -4.93305653e-01
-1.76427901e-01 -5.96418917e-01 1.43650675e+00 -2.70228386e-02
-7.65583336e-01 -1.53359878e+00 2.78286189e-01 4.15292144e-01
4.17619765e-01 9.99611318e-02 1.12880659e+00 -1.03195786e+00
-3.18666577e-01 2.95664817e-01 -2.38955706e-01 3.81141603e-01
5.71049094e-01 -3.50223899e-01 -1.04732966e+00 -3.51503007e-02
6.29661977e-02 -2.94910252e-01 7.86787570e-01 2.27380276e-01
2.96199650e-01 -2.22549111e-01 -4.70987439e-01 3.66846144e-01
1.13407791e+00 -1.58577383e-01 1.30908862e-01 6.24364853e-01
5.97400784e-01 9.29223299e-01 7.23212898e-01 1.49797976e-01
8.89316916e-01 5.99778235e-01 -3.61419320e-01 9.28360298e-02
-1.65612489e-01 -8.59869644e-02 6.33229375e-01 8.58907700e-01
2.97937803e-02 -1.66347817e-01 -1.04834342e+00 9.87649500e-01
-1.52308738e+00 -5.19216120e-01 3.44703309e-02 1.73771548e+00
1.29060829e+00 3.21595162e-01 7.11821318e-02 -3.27712417e-01
3.44642997e-01 3.49683553e-01 -5.34543872e-01 -5.17791390e-01
-2.68446594e-01 6.81708574e-01 2.41872460e-01 5.92417419e-01
-1.24944377e+00 1.22177184e+00 6.13198853e+00 6.28780365e-01
-9.48034704e-01 1.69863209e-01 3.03897887e-01 1.66062459e-01
-7.46880770e-01 1.68689653e-01 -9.86209750e-01 -2.15077028e-02
9.91296768e-01 -4.19375896e-01 1.52159199e-01 9.07693505e-01
-1.96207300e-01 1.32398261e-02 -1.55566919e+00 4.21670616e-01
9.58216190e-02 -8.02572548e-01 2.82246768e-01 -2.36985758e-02
9.67022061e-01 1.04141347e-01 1.82854652e-01 5.98634422e-01
6.09021246e-01 -8.90804589e-01 5.31807780e-01 2.21004635e-01
8.01739693e-01 -6.49460554e-01 5.23706496e-01 7.67181907e-03
-1.28747892e+00 2.73531884e-01 -2.03945652e-01 -7.42762983e-02
3.18082273e-01 8.64382505e-01 -6.65000021e-01 8.17076802e-01
6.14654303e-01 8.87548983e-01 -4.72879916e-01 1.33761555e-01
-7.18073010e-01 5.51882207e-01 -1.94136590e-01 1.05081722e-01
1.24767579e-01 -1.88436806e-01 7.09965289e-01 1.43491352e+00
1.54225528e-01 -6.95855021e-02 7.31101036e-02 8.51016283e-01
3.22162732e-02 2.60248661e-01 -9.96301949e-01 -1.22456737e-01
2.52352744e-01 1.06542969e+00 -4.15971398e-01 -4.75317448e-01
-7.70624161e-01 9.27959085e-01 4.77240026e-01 6.38371527e-01
-4.45905000e-01 -4.48924750e-01 1.14902210e+00 1.37574419e-01
6.93006635e-01 -2.54429758e-01 -4.51877266e-01 -1.36037087e+00
1.95915371e-01 -1.02881885e+00 3.70036602e-01 -4.41611946e-01
-1.78707945e+00 7.81702638e-01 2.94592410e-01 -1.09416687e+00
-6.24418139e-01 -8.68695617e-01 -2.11367190e-01 1.08678257e+00
-1.98544991e+00 -1.41739011e+00 1.64426580e-01 7.57184088e-01
5.15936613e-01 -2.40908280e-01 1.03845358e+00 1.40150800e-01
2.38226238e-03 5.43805003e-01 -3.26015264e-01 8.28433782e-02
1.12356234e+00 -1.52675831e+00 7.53084600e-01 4.74906802e-01
3.53107452e-01 1.14596105e+00 7.85905302e-01 -5.13311625e-01
-9.99555826e-01 -1.01634657e+00 1.43760359e+00 -1.13076901e+00
1.12874413e+00 -3.87048364e-01 -8.69278014e-01 9.35315192e-01
6.38639569e-01 -1.80455312e-01 1.16022670e+00 9.87974763e-01
-9.05224144e-01 -6.37729317e-02 -8.83266211e-01 5.28301775e-01
1.13755059e+00 -9.71357286e-01 -1.29224765e+00 2.67875403e-01
8.84482920e-01 -2.70780530e-02 -1.04935312e+00 3.24478447e-01
4.15981710e-01 -9.25913572e-01 9.30926979e-01 -1.03216040e+00
5.62627375e-01 -2.78644234e-01 -1.42796248e-01 -1.73440623e+00
-2.04919040e-01 -6.42834902e-01 -2.77434886e-01 1.64890409e+00
1.00749302e+00 -8.16472352e-01 4.71805871e-01 7.34109998e-01
2.23200932e-01 -2.88923651e-01 -9.00412679e-01 -8.36266577e-01
6.48316920e-01 -5.18142402e-01 5.63148141e-01 1.14551461e+00
2.30654493e-01 1.23017216e+00 2.29039252e-01 6.62039667e-02
3.69496346e-01 3.43909055e-01 4.87676322e-01 -1.24826229e+00
-5.44875562e-01 -3.78908604e-01 -2.38295808e-01 -1.27075076e+00
6.38156593e-01 -1.03998268e+00 -1.26872510e-01 -1.56771231e+00
-1.00455694e-01 -4.78961349e-01 -4.35812503e-01 4.10320222e-01
-3.91294062e-01 -1.26064196e-01 -5.85894659e-02 -7.52876922e-02
-6.62207901e-01 5.19503534e-01 1.07176816e+00 -3.11569899e-01
-1.67107925e-01 1.94025740e-01 -1.36273217e+00 8.27611804e-01
7.37634301e-01 -6.13894701e-01 -5.87390900e-01 -7.12453127e-01
5.26848376e-01 -3.34915429e-01 -1.75645351e-01 -4.74160850e-01
-2.33204127e-03 -1.95410978e-02 -4.81329411e-02 -1.34576514e-01
3.84851456e-01 -9.12689090e-01 -4.23807204e-01 -2.34401941e-01
-4.06530321e-01 -4.12986707e-03 4.08846676e-01 5.49939275e-01
-4.93784010e-01 7.15485141e-02 3.59981745e-01 -1.33438364e-01
-5.74843884e-01 2.71671219e-03 -4.47190613e-01 7.59256244e-01
6.27134919e-01 1.71839267e-01 -3.06576580e-01 -4.06486928e-01
-8.15629363e-01 6.55440837e-02 2.27641448e-01 7.05917478e-01
2.87440736e-02 -1.31944335e+00 -8.21590126e-01 3.73090953e-02
5.79483390e-01 6.76865727e-02 -1.55058831e-01 6.00509703e-01
2.44908884e-01 6.42681241e-01 2.38434784e-02 -4.82005239e-01
-1.00775039e+00 5.16817808e-01 1.01879379e-02 -6.22414231e-01
-8.18377733e-02 9.68553126e-01 3.97080779e-01 -7.64919043e-01
-2.39792347e-01 -6.58563733e-01 -2.11673468e-01 5.99314809e-01
3.27361524e-01 -1.17974959e-01 2.26858616e-01 -5.87822378e-01
-5.58235884e-01 8.11141312e-01 -3.14284384e-01 -4.37454581e-01
1.39938140e+00 -2.73033172e-01 -3.21057200e-01 6.20256364e-01
1.18384707e+00 1.23139352e-01 -8.29662919e-01 -6.58826172e-01
5.54304957e-01 -8.41589347e-02 -3.45249772e-01 -1.03287649e+00
-7.00366855e-01 6.43520653e-01 -2.36788113e-03 3.73637199e-01
9.61206138e-01 4.73275393e-01 5.31354189e-01 5.99418283e-01
3.75862062e-01 -1.12758207e+00 -2.33965576e-01 9.73905504e-01
8.99385691e-01 -1.32029927e+00 2.16816038e-01 -4.69699323e-01
-1.02489638e+00 7.10422635e-01 3.80884916e-01 -1.38444886e-01
9.74982679e-01 3.13640565e-01 2.74062514e-01 -3.53615016e-01
-9.65173006e-01 -6.06469691e-01 5.12956560e-01 1.07979381e+00
8.65554690e-01 1.08762667e-01 -2.19186783e-01 8.22446525e-01
-4.14235383e-01 -1.10666312e-01 -8.07286128e-02 9.84778821e-01
-2.48690620e-01 -1.56992483e+00 -1.84078917e-01 2.91998863e-01
-5.99168062e-01 -4.76398528e-01 -8.06914628e-01 8.94127429e-01
1.04331784e-01 1.02617884e+00 -1.76526770e-01 3.81246209e-04
6.93357348e-01 4.04758006e-01 2.97620475e-01 -9.43264127e-01
-7.01156497e-01 -1.78096756e-01 6.63995147e-01 -3.11933041e-01
-8.87120008e-01 -7.81622171e-01 -8.04730475e-01 -3.76660451e-02
-7.13176429e-02 3.40549588e-01 5.41007578e-01 1.17254972e+00
5.50092101e-01 6.03544652e-01 1.31110325e-01 -6.02584481e-01
-3.19614828e-01 -1.06897140e+00 -5.11000812e-01 5.39090633e-01
3.86419624e-01 -6.93641245e-01 -3.41361403e-01 3.57937783e-01] | [11.337977409362793, 6.864887237548828] |
8e027908-720a-4a18-bc20-769794e96914 | realistic-saliency-guided-image-enhancement-1 | 2306.06092 | null | https://arxiv.org/abs/2306.06092v1 | https://arxiv.org/pdf/2306.06092v1.pdf | Realistic Saliency Guided Image Enhancement | Common editing operations performed by professional photographers include the cleanup operations: de-emphasizing distracting elements and enhancing subjects. These edits are challenging, requiring a delicate balance between manipulating the viewer's attention while maintaining photo realism. While recent approaches can boast successful examples of attention attenuation or amplification, most of them also suffer from frequent unrealistic edits. We propose a realism loss for saliency-guided image enhancement to maintain high realism across varying image types, while attenuating distractors and amplifying objects of interest. Evaluations with professional photographers confirm that we achieve the dual objective of realism and effectiveness, and outperform the recent approaches on their own datasets, while requiring a smaller memory footprint and runtime. We thus offer a viable solution for automating image enhancement and photo cleanup operations. | ['Yağız Aksoy', 'Eli Shechtman', 'Eric Kee', 'Zoya Bylinskii', 'S. Mahdi H. Miangoleh'] | 2023-06-09 | realistic-saliency-guided-image-enhancement | http://openaccess.thecvf.com//content/CVPR2023/html/Miangoleh_Realistic_Saliency_Guided_Image_Enhancement_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Miangoleh_Realistic_Saliency_Guided_Image_Enhancement_CVPR_2023_paper.pdf | cvpr-2023-1 | ['image-enhancement'] | ['computer-vision'] | [ 5.35818517e-01 -2.37102471e-02 2.96457767e-01 -3.25877070e-01
-4.74906385e-01 -4.04441029e-01 4.78404492e-01 1.81726158e-01
-4.84222114e-01 4.48186129e-01 3.12740028e-01 -3.38075534e-02
1.51146874e-01 -4.56097782e-01 -6.28630757e-01 -4.06390220e-01
2.03686550e-01 -2.63621688e-01 3.40653330e-01 -3.07578862e-01
6.63510382e-01 5.00740767e-01 -1.84486723e+00 -1.36028584e-02
1.20106208e+00 6.85041010e-01 3.10462028e-01 8.80180717e-01
1.51342571e-01 1.01615226e+00 -8.00049484e-01 -7.09682465e-01
4.76585388e-01 -3.89976770e-01 -5.36381543e-01 3.94283950e-01
1.16008353e+00 -6.46554351e-01 -2.44785041e-01 1.32229209e+00
7.86444902e-01 3.81757766e-01 1.12583280e-01 -1.20221865e+00
-1.27141142e+00 2.62618721e-01 -1.05971706e+00 4.41659987e-01
4.26438928e-01 5.83690882e-01 8.46680462e-01 -8.48088562e-01
4.55307722e-01 1.22340858e+00 7.12840140e-01 5.41730106e-01
-1.39161170e+00 -5.58090448e-01 2.09489793e-01 3.16951334e-01
-1.20242083e+00 -8.92441869e-01 6.11318231e-01 -2.76920438e-01
7.67253757e-01 6.94621444e-01 6.10234797e-01 7.85002530e-01
3.87383737e-02 4.90373671e-01 1.21898103e+00 -3.70550513e-01
2.58545429e-01 4.07887876e-01 -4.79205027e-02 4.20497537e-01
2.29431063e-01 -4.56901491e-02 -5.99240303e-01 2.49886792e-02
9.24695075e-01 -1.20563798e-01 -6.27456427e-01 -7.84699097e-02
-9.96123016e-01 4.44759667e-01 3.79519522e-01 -2.19457880e-01
-5.64592957e-01 3.10391515e-01 2.53774136e-01 1.45550137e-02
4.51311797e-01 8.63737464e-01 4.45146598e-02 -1.51105106e-01
-9.74673271e-01 3.33723813e-01 2.11771309e-01 1.11876476e+00
7.58377910e-01 1.23132385e-01 -6.06947839e-01 8.84705067e-01
-1.42085671e-01 5.20399094e-01 1.94421232e-01 -1.25552177e+00
1.42382935e-01 3.55981469e-01 4.12124544e-01 -1.34876108e+00
-7.22911432e-02 -3.39151382e-01 -6.08341932e-01 8.76083076e-01
2.42003098e-01 -4.63034473e-02 -9.21593726e-01 1.76319516e+00
4.20650750e-01 -2.92057656e-02 -4.51571494e-01 1.23545563e+00
5.60412109e-01 5.55062890e-01 4.80175287e-01 -1.16932996e-01
1.51751459e+00 -1.25498724e+00 -1.10610342e+00 -5.97252727e-01
-9.97831207e-03 -1.01481366e+00 1.70134115e+00 4.48907375e-01
-1.76696241e+00 -5.96883118e-01 -1.11521304e+00 -6.40888691e-01
-1.56673603e-02 1.01887628e-01 6.17415130e-01 6.57921255e-01
-1.22266412e+00 6.20006442e-01 -4.65760142e-01 -6.18509725e-02
4.49915558e-01 1.42237604e-01 -1.86926812e-01 6.71841726e-02
-8.29025745e-01 1.20673323e+00 -2.44597688e-01 7.34438747e-02
-8.03038895e-01 -1.27559769e+00 -6.97261393e-01 3.65592390e-01
3.64321500e-01 -6.57652855e-01 1.28376305e+00 -1.37981808e+00
-1.46350408e+00 9.88223791e-01 5.30932238e-03 -3.84281009e-01
6.67304397e-01 -5.37695944e-01 -3.84676993e-01 2.31694922e-01
-6.85419068e-02 8.31660450e-01 1.22012937e+00 -1.34608066e+00
-4.81581718e-01 -1.23459317e-01 4.48310912e-01 5.03118813e-01
-7.40356326e-01 2.03586936e-01 -3.90918553e-01 -8.58980954e-01
-2.70756185e-01 -7.30835557e-01 -2.71549374e-01 5.18407941e-01
-3.90163362e-01 4.34595495e-01 7.75568366e-01 -7.46099532e-01
1.28447509e+00 -2.13124895e+00 -4.39683571e-02 -2.38966554e-01
6.05457425e-01 5.08544922e-01 -1.86079979e-01 2.26676047e-01
-1.11966833e-01 7.45074525e-02 -5.88199198e-02 -6.47698462e-01
-1.40932828e-01 -1.67518929e-01 -4.06074524e-01 5.29198408e-01
2.46853620e-01 9.02473569e-01 -9.44408178e-01 -4.75145876e-01
5.43947458e-01 6.46783233e-01 -7.23022759e-01 1.62263840e-01
-2.27020197e-02 2.77498141e-02 2.48344213e-01 4.67860878e-01
9.68410015e-01 -1.85715824e-01 -7.39000812e-02 -4.43402737e-01
-2.18132228e-01 1.60782814e-01 -1.17998648e+00 1.41661978e+00
-6.27498567e-01 1.02873528e+00 3.16774011e-01 -3.23992848e-01
6.60890102e-01 -2.77156413e-01 1.25167921e-01 -9.46636856e-01
1.51468515e-01 -2.53051996e-01 -2.43879139e-01 -6.62350655e-01
1.14243829e+00 -2.56349862e-01 4.96438563e-01 3.75975847e-01
-3.62020999e-01 -2.25231200e-01 4.02743928e-02 3.66878092e-01
9.11026299e-01 -5.80286346e-02 3.27873826e-01 -3.31706017e-01
1.89132482e-01 -1.44301191e-01 2.94996917e-01 6.18261158e-01
-4.08698499e-01 8.16398382e-01 1.77881613e-01 -1.80582941e-01
-1.13408256e+00 -1.15281999e+00 2.55046755e-01 1.28990936e+00
6.09800160e-01 -5.33258855e-01 -8.47435117e-01 -1.81064606e-01
-2.53606915e-01 7.75680363e-01 -5.98777592e-01 -3.35919470e-01
-3.76748681e-01 -3.77387166e-01 3.06035787e-01 4.46436048e-01
5.36267877e-01 -8.34794939e-01 -1.03436983e+00 -1.57477945e-01
-3.69160652e-01 -1.12748837e+00 -1.01900518e+00 -3.94804418e-01
-4.93592352e-01 -8.08534265e-01 -8.97053063e-01 -5.96728921e-01
1.08432508e+00 1.06784809e+00 1.20198858e+00 2.66081572e-01
-5.91559410e-01 3.57020169e-01 -1.66616842e-01 -5.43728530e-01
-2.34613091e-01 -4.22665864e-01 -1.16728105e-01 -2.49237474e-02
-1.70526683e-01 -7.82067657e-01 -1.09683418e+00 4.12519932e-01
-1.08155894e+00 3.27648520e-01 6.24824226e-01 5.28991103e-01
3.73300195e-01 -8.92664716e-02 9.75130573e-02 -7.12335289e-01
6.97516382e-01 7.72514492e-02 -5.83832860e-01 2.12698400e-01
-6.35526001e-01 -2.34554365e-01 5.80184937e-01 -7.86713779e-01
-1.41643095e+00 -1.03794746e-01 1.98428184e-01 -6.60425127e-01
1.26757100e-01 -1.98049337e-01 -4.02628407e-02 -4.75585312e-01
9.92684305e-01 8.64486955e-03 6.53735036e-03 -2.44512841e-01
6.85654759e-01 2.26976797e-01 8.88059199e-01 -2.04970643e-01
8.30188096e-01 7.31125951e-01 -2.54731238e-01 -8.37952793e-01
-7.86400020e-01 -2.50462234e-01 -1.47110179e-01 -5.19641697e-01
6.44802094e-01 -8.35728109e-01 -8.15280557e-01 3.81280363e-01
-1.04300559e+00 -2.45044708e-01 -5.14142871e-01 1.01153150e-01
-2.81723648e-01 6.66178703e-01 -4.01422709e-01 -7.26851761e-01
-3.50381136e-01 -1.10823214e+00 1.03764832e+00 4.49491292e-01
-3.55687886e-01 -6.76841497e-01 -1.51936635e-01 2.99573213e-01
8.17837894e-01 1.78167000e-01 4.86342758e-01 3.42116147e-01
-6.00038111e-01 -1.68908481e-02 -5.37488520e-01 3.34501535e-01
1.85844719e-01 1.60933152e-01 -1.15221894e+00 -3.00441772e-01
-1.58093125e-01 -1.68227375e-01 5.74973166e-01 4.60044175e-01
1.19471419e+00 -4.91282612e-01 6.13906719e-02 7.13845968e-01
1.40856874e+00 -1.84858814e-02 1.08958101e+00 1.67402908e-01
5.82924247e-01 5.34211934e-01 7.21693337e-01 5.44666886e-01
2.19115391e-01 7.81678498e-01 4.54883128e-01 -7.03646719e-01
-5.37969232e-01 -2.25412771e-01 2.60956198e-01 2.74965435e-01
-9.32334214e-02 -1.46319941e-01 -5.57312310e-01 6.22127056e-01
-1.49873662e+00 -1.12803829e+00 -2.56857038e-01 2.25509501e+00
1.00789893e+00 -2.52217855e-02 1.35956049e-01 8.70731771e-02
8.43842506e-01 2.65313894e-01 -4.16235328e-01 -4.04005229e-01
-6.20055757e-02 2.16676101e-01 6.02899730e-01 7.14497805e-01
-8.02031159e-01 8.74452770e-01 7.44005680e+00 6.72882915e-01
-1.15751505e+00 4.44511548e-02 6.54763460e-01 -7.26807892e-01
-4.46491271e-01 5.32651022e-02 -5.66507041e-01 5.42212129e-01
3.86561036e-01 -3.23171228e-01 6.25311494e-01 9.54299629e-01
5.94036162e-01 -5.09211063e-01 -8.86640251e-01 1.22706938e+00
3.55433255e-01 -1.30925107e+00 -2.83121616e-02 -2.10274130e-01
7.20247865e-01 -4.66850430e-01 4.44981903e-01 -1.54368877e-01
1.95649847e-01 -9.97178972e-01 9.55362380e-01 4.62327600e-01
8.35158229e-01 -5.56371272e-01 8.85566548e-02 -2.16702297e-01
-7.46165514e-01 -8.60063210e-02 -3.20826024e-01 -2.91387737e-01
4.75762039e-01 7.82067895e-01 -3.47167820e-01 -5.76406680e-02
7.75171936e-01 3.73897344e-01 -8.34098935e-01 1.12663722e+00
-1.92421988e-01 1.16816744e-01 3.90421897e-02 9.16363373e-02
-5.98288588e-02 -1.87324882e-01 6.63522661e-01 1.25000417e+00
9.96068567e-02 2.39941120e-01 -3.61126393e-01 1.06025672e+00
-9.95253562e-04 3.70057784e-02 -3.53244513e-01 1.90082788e-01
5.34237325e-01 1.45448351e+00 -6.27794445e-01 -4.33255285e-01
-6.23492971e-02 1.35675418e+00 2.28738889e-01 3.28518629e-01
-1.38070357e+00 -6.48593187e-01 1.04468584e+00 3.61639738e-01
1.64397657e-01 -1.41931131e-01 -5.62871397e-01 -9.60647881e-01
2.49447852e-01 -1.03401470e+00 -1.00531541e-01 -1.38465321e+00
-9.10487533e-01 5.20249665e-01 -1.81651831e-01 -1.22279978e+00
4.40458238e-01 -2.66525865e-01 -6.33527637e-01 7.55195260e-01
-1.60728717e+00 -9.42687035e-01 -9.65451241e-01 5.47579587e-01
6.65357947e-01 5.21843672e-01 2.94984579e-01 6.19079947e-01
-3.88617933e-01 7.36668527e-01 -4.56539601e-01 -5.88310361e-01
9.87865031e-01 -1.17320931e+00 3.54597956e-01 1.13256967e+00
-6.91079050e-02 7.38646388e-01 1.25663197e+00 -4.02986258e-01
-1.40137088e+00 -8.90661299e-01 5.84315658e-01 -2.63418287e-01
3.01283121e-01 -4.01186734e-01 -8.72130752e-01 5.01033008e-01
5.51877499e-01 -1.73318952e-01 3.30449700e-01 -1.01994932e-01
-3.59562576e-01 -3.32169011e-02 -1.13529432e+00 1.14767039e+00
1.23022437e+00 -4.73430187e-01 -3.16596895e-01 3.48040760e-01
6.58941567e-01 -4.23421353e-01 -5.00153720e-01 -3.86802368e-02
4.76430029e-01 -1.32305396e+00 1.33816528e+00 -2.87649751e-01
8.05071175e-01 -4.23568726e-01 2.36841068e-01 -1.46034312e+00
-5.93012094e-01 -1.10438108e+00 1.23998962e-01 1.44543195e+00
-2.15490051e-02 -2.69991547e-01 4.18243259e-01 8.79732251e-01
-9.12220776e-02 -3.76635343e-01 -3.47951442e-01 -6.12556159e-01
-6.00574136e-01 -3.14214796e-01 4.19676960e-01 9.58919525e-01
1.85149625e-01 1.43023029e-01 -9.08387363e-01 1.31449446e-01
7.36877799e-01 -3.20447564e-01 1.01448524e+00 -5.52159727e-01
-2.38879889e-01 -5.96253514e-01 -3.00085425e-01 -9.64024186e-01
-3.38260233e-01 -2.11817369e-01 1.86380818e-01 -1.24962842e+00
3.11607599e-01 -2.12078735e-01 1.63708821e-01 2.81361550e-01
-6.21164680e-01 6.21994793e-01 2.12870613e-01 2.58324649e-02
-5.71420431e-01 4.72895592e-01 1.41670549e+00 -4.49914066e-03
-1.98286474e-01 -4.02702034e-01 -1.23108423e+00 7.33670890e-01
4.88694578e-01 -1.14265271e-01 -8.14618111e-01 -7.62957394e-01
2.98411876e-01 -3.96194786e-01 7.46702492e-01 -9.66341496e-01
3.01914632e-01 -4.34099495e-01 2.52716005e-01 -2.99961686e-01
5.07840812e-01 -6.16308093e-01 2.04279348e-01 2.95541734e-01
-5.10093808e-01 2.89537460e-01 4.10907865e-01 5.10513902e-01
3.16488068e-03 9.05761868e-02 1.36017501e+00 -1.66515913e-02
-8.00026774e-01 9.21906978e-02 -2.84307182e-01 1.42740816e-01
1.15943885e+00 -4.04241174e-01 -6.35845184e-01 -6.88539743e-01
-3.55454773e-01 2.46606451e-02 8.35319638e-01 3.37235421e-01
7.88344204e-01 -1.12020886e+00 -5.90272665e-01 8.96508992e-02
-7.03281723e-03 -4.23401445e-01 7.46594071e-01 8.71332347e-01
-6.66409671e-01 -2.44393021e-01 -6.08503044e-01 -3.02479386e-01
-1.58738101e+00 8.77109945e-01 3.46936911e-01 2.93447763e-01
-8.11931312e-01 9.71863747e-01 4.85872895e-01 4.75501478e-01
3.69828194e-01 -1.91592067e-01 3.78360376e-02 -2.51227189e-02
9.95071113e-01 7.64699578e-01 2.53293496e-02 -2.00815052e-01
-1.82695180e-01 4.19650972e-01 -1.58267617e-01 -6.17142692e-02
1.12308395e+00 -6.68327868e-01 2.54165120e-02 -1.10572062e-01
8.16527367e-01 4.55442935e-01 -1.47053361e+00 9.20629781e-03
-5.00294507e-01 -1.24973166e+00 2.59326637e-01 -8.21937680e-01
-1.12030292e+00 7.82841146e-01 5.36415219e-01 1.60142317e-01
1.51931119e+00 -3.89939278e-01 7.21894085e-01 -1.18238293e-01
2.33408040e-03 -1.18076301e+00 4.11864936e-01 -1.60430446e-01
1.18301630e+00 -1.19946957e+00 3.22610736e-01 -7.17494488e-01
-9.81063545e-01 5.30073166e-01 8.77577901e-01 -2.17873394e-01
1.67239666e-01 5.14490664e-01 2.13765606e-01 -2.53110528e-01
-7.13674486e-01 -1.23287082e-01 2.14271381e-01 7.62987912e-01
2.60922015e-01 -2.78818280e-01 -2.89196402e-01 1.42380223e-01
-1.58926353e-01 -1.36111185e-01 7.57466495e-01 6.56178117e-01
-3.40855926e-01 -4.64240879e-01 -4.90187258e-01 1.11680448e-01
-3.48418683e-01 -3.58475059e-01 -4.34223592e-01 8.64933431e-01
-4.27602381e-02 9.95955825e-01 4.64740321e-02 -1.71658576e-01
5.27947962e-01 -5.46310723e-01 5.05656183e-01 -3.29082280e-01
-7.57196963e-01 -1.12430237e-01 6.60241842e-02 -8.24809313e-01
-3.70141745e-01 -3.62362951e-01 -6.72766328e-01 -6.28912389e-01
-4.29001808e-01 -2.72771418e-01 5.22402406e-01 3.81459683e-01
6.22713566e-01 6.55363381e-01 5.55708706e-01 -1.14711630e+00
-5.47647357e-01 -8.17385375e-01 -5.43783069e-01 8.27349365e-01
3.88454527e-01 -7.57615149e-01 -3.94014895e-01 4.05940801e-01] | [11.161893844604492, -1.146140217781067] |
6827aa7a-ec26-4b97-9516-ad69a8c6a2cb | learning-to-adapt-for-stereo | 1904.02957 | null | http://arxiv.org/abs/1904.02957v1 | http://arxiv.org/pdf/1904.02957v1.pdf | Learning to Adapt for Stereo | Real world applications of stereo depth estimation require models that are
robust to dynamic variations in the environment. Even though deep learning
based stereo methods are successful, they often fail to generalize to unseen
variations in the environment, making them less suitable for practical
applications such as autonomous driving. In this work, we introduce a
"learning-to-adapt" framework that enables deep stereo methods to continuously
adapt to new target domains in an unsupervised manner. Specifically, our
approach incorporates the adaptation procedure into the learning objective to
obtain a base set of parameters that are better suited for unsupervised online
adaptation. To further improve the quality of the adaptation, we learn a
confidence measure that effectively masks the errors introduced during the
unsupervised adaptation. We evaluate our method on synthetic and real-world
stereo datasets and our experiments evidence that learning-to-adapt is, indeed
beneficial for online adaptation on vastly different domains. | ['Philip H. S. Torr', 'Thomas Joy', 'Thalaiyasingam Ajanthan', 'Oscar Rahnama', 'Luigi Di Stefano', 'Alessio Tonioni'] | 2019-04-05 | learning-to-adapt-for-stereo-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Tonioni_Learning_to_Adapt_for_Stereo_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Tonioni_Learning_to_Adapt_for_Stereo_CVPR_2019_paper.pdf | cvpr-2019-6 | ['stereo-depth-estimation'] | ['computer-vision'] | [ 2.25666523e-01 -1.17474273e-01 5.94525300e-02 -7.23646939e-01
-5.00970781e-01 -5.73654115e-01 5.88126659e-01 -1.57599188e-02
-5.73387325e-01 8.87145042e-01 1.36620291e-02 1.42879248e-01
1.02684595e-01 -7.28359938e-01 -1.06209528e+00 -6.48296118e-01
2.67314970e-01 6.33440971e-01 7.36150801e-01 -2.49506906e-01
4.08170998e-01 4.93121833e-01 -1.93471384e+00 5.09485882e-03
9.75935757e-01 7.26767123e-01 4.11397815e-01 6.11656010e-01
8.93621668e-02 4.83809978e-01 -2.53163129e-01 -1.17271334e-01
5.76555490e-01 -2.20933661e-01 -5.73417664e-01 3.67808849e-01
4.78502870e-01 -3.84420753e-01 -3.45452338e-01 1.09276783e+00
5.24072289e-01 3.53280574e-01 5.55916905e-01 -6.86844766e-01
6.12307340e-02 -4.49495018e-02 -1.15096048e-01 3.33365023e-01
3.10188860e-01 4.67705995e-01 5.78213990e-01 -7.04594254e-01
8.66937280e-01 1.06589913e+00 6.43489838e-01 6.27987325e-01
-1.27371991e+00 -4.07327563e-01 3.10752213e-01 3.49946469e-01
-1.01229286e+00 -7.61420012e-01 7.68537879e-01 -4.29645032e-01
6.66452348e-01 -1.85282320e-01 7.46211767e-01 1.23379695e+00
3.25140119e-01 6.30100965e-01 1.17618740e+00 -3.01248372e-01
7.62053072e-01 3.43940079e-01 -4.46378827e-01 4.61355507e-01
2.33193859e-01 5.53479016e-01 -9.19325292e-01 2.50254214e-01
6.77636027e-01 -2.85087883e-01 -6.22018017e-02 -1.04305756e+00
-1.09411669e+00 7.50443280e-01 4.67479885e-01 -2.79387180e-02
-3.56807053e-01 -7.67326206e-02 3.35959703e-01 3.99775147e-01
4.98702824e-01 5.10058761e-01 -5.08949637e-01 -2.42795229e-01
-6.69841230e-01 2.30940238e-01 6.43346250e-01 8.60043347e-01
1.18694890e+00 9.63147134e-02 1.85168758e-01 9.94146168e-01
6.66746274e-02 4.47491646e-01 6.33756399e-01 -1.15067887e+00
3.39706004e-01 4.20998693e-01 1.91088960e-01 -5.59722662e-01
-3.04422259e-01 -4.76244658e-01 -3.91736984e-01 5.42500079e-01
4.64924306e-01 -1.49951831e-01 -1.02495527e+00 1.74194133e+00
6.74327314e-01 1.67726651e-01 1.21630579e-01 7.90099740e-01
1.37772486e-01 2.20376357e-01 -1.39027253e-01 9.33345687e-03
4.78492379e-01 -8.92011464e-01 -3.17851245e-01 -7.65163779e-01
4.95998830e-01 -5.98544061e-01 1.09404445e+00 4.78457361e-01
-8.05384219e-01 -6.15688682e-01 -1.10804641e+00 1.83006629e-01
-3.91927987e-01 -4.25216079e-01 4.87829387e-01 4.83119696e-01
-1.05941844e+00 6.64288878e-01 -8.20591688e-01 -6.95975900e-01
3.77566606e-01 4.54706848e-01 -5.03876150e-01 -5.42148128e-02
-9.03672576e-01 8.51866782e-01 7.75158823e-01 -6.34921715e-02
-1.06756723e+00 -4.22888070e-01 -8.89256716e-01 -3.57298464e-01
4.10946339e-01 -8.10627401e-01 1.43291426e+00 -1.28426564e+00
-1.91186190e+00 8.96735609e-01 -2.96540946e-01 -7.20454454e-01
7.45374501e-01 -3.53034168e-01 -1.61042109e-01 1.33427195e-02
2.16780659e-02 8.61524522e-01 1.13436460e+00 -1.47891283e+00
-9.82650816e-01 -4.26414222e-01 2.32601628e-01 5.02040148e-01
-2.86629319e-01 -6.86166823e-01 -5.57068110e-01 -3.57928008e-01
2.61947423e-01 -1.09351420e+00 -4.80191052e-01 7.49523798e-03
2.30780020e-02 4.09238636e-01 6.63193166e-01 -2.23998636e-01
6.98365152e-01 -2.06504083e+00 2.19820246e-01 2.57010102e-01
-7.32841566e-02 1.46101251e-01 4.28328775e-02 1.67721778e-01
3.92045885e-01 -5.63879907e-01 -3.25182885e-01 -3.17430109e-01
-2.16520354e-01 5.24832785e-01 1.51057485e-02 3.40622276e-01
1.30532011e-01 4.40566570e-01 -1.00855577e+00 -3.17367882e-01
4.96711463e-01 1.95437774e-01 -9.30657625e-01 4.30141628e-01
-3.55101466e-01 1.00910413e+00 -3.84222806e-01 4.16610539e-01
4.24575716e-01 9.13687646e-02 -1.00315452e-01 1.07060671e-01
-1.75214931e-02 1.08896784e-01 -1.17644155e+00 1.71551692e+00
-6.84376419e-01 7.33251512e-01 -6.96054921e-02 -1.04022014e+00
1.03627992e+00 -1.07576773e-01 3.76114756e-01 -9.09226179e-01
-6.87835366e-03 3.88101071e-01 3.35060433e-02 -5.36034405e-01
6.34470522e-01 -2.74837792e-01 2.04092726e-01 -1.14165023e-01
1.49221778e-01 -5.33921063e-01 1.48831114e-01 -2.21073538e-01
8.53958845e-01 3.00301045e-01 3.42376232e-01 -2.41818175e-01
7.20299661e-01 -1.41075738e-02 7.77200341e-01 8.11426401e-01
-3.36531132e-01 7.03047395e-01 -1.46238789e-01 -6.72855735e-01
-1.36692035e+00 -1.07658350e+00 -2.01724350e-01 8.42646539e-01
3.05441976e-01 1.58962011e-01 -7.79087961e-01 -6.86272860e-01
5.08014411e-02 6.79051638e-01 -5.82843006e-01 -4.37831998e-01
-4.96840745e-01 -4.38099205e-01 -3.20252739e-02 5.02831638e-01
5.03205001e-01 -1.16555238e+00 -8.88785899e-01 5.12150943e-01
-2.97357817e-03 -1.26814723e+00 -3.06981653e-01 2.94274777e-01
-1.20968139e+00 -9.07821655e-01 -4.18652713e-01 -4.85299796e-01
7.20242143e-01 6.63041323e-02 1.04676712e+00 -2.87539601e-01
-3.39935757e-02 6.75571501e-01 -2.35195145e-01 -5.57943106e-01
-6.26647890e-01 2.75764585e-01 2.34832764e-01 1.33969530e-01
3.36860329e-01 -8.00393164e-01 -5.15277922e-01 3.83823216e-01
-9.10098493e-01 -1.26233995e-01 5.74489594e-01 9.76668298e-01
6.76926255e-01 -1.29369602e-01 4.04374719e-01 -1.03078187e+00
2.79246211e-01 -2.17019796e-01 -9.81877744e-01 -1.21400855e-01
-8.69310200e-01 3.76550853e-01 8.04114521e-01 -4.91856962e-01
-1.21124542e+00 4.56159383e-01 -2.08962142e-01 -2.72737801e-01
-3.63839269e-01 3.33524197e-01 -8.13624784e-02 -4.41504478e-01
8.33595812e-01 2.94672996e-01 -1.49175778e-01 -3.26958835e-01
3.63267697e-02 3.29814583e-01 7.08362818e-01 -6.02636337e-01
1.08315456e+00 7.33601034e-01 6.73838928e-02 -8.06974411e-01
-9.11232948e-01 -5.27304411e-01 -9.54906821e-01 -4.87725616e-01
5.85440218e-01 -8.96862447e-01 -2.28862807e-01 6.03993475e-01
-6.93090856e-01 -7.05154359e-01 -4.41938579e-01 5.14953434e-01
-1.06575704e+00 3.01441163e-01 -3.41642909e-02 -6.33072972e-01
1.13472946e-01 -1.11236918e+00 1.00159812e+00 5.68932116e-01
-6.18082529e-04 -1.28719485e+00 4.09259528e-01 2.31908455e-01
3.58006090e-01 1.47641227e-01 4.07990932e-01 -4.46588606e-01
-6.93740308e-01 -9.83958691e-02 1.69338524e-01 4.58511710e-01
1.49480775e-01 -7.15147853e-02 -1.11779106e+00 -5.26570857e-01
-5.96004501e-02 -3.95787537e-01 8.85620296e-01 2.86275655e-01
1.14883161e+00 3.60597111e-02 -2.06254303e-01 9.98017371e-01
1.49551785e+00 3.20879906e-01 7.17542708e-01 9.38759029e-01
4.49250817e-01 5.69952071e-01 8.44727576e-01 4.97137100e-01
3.45131427e-01 8.14022481e-01 5.52872360e-01 1.37930334e-01
9.35364664e-02 -4.48226333e-01 3.90585721e-01 6.16311073e-01
-5.25453733e-03 9.12844911e-02 -9.25854623e-01 7.70331144e-01
-1.81212342e+00 -7.33603477e-01 5.20855963e-01 2.61685491e+00
7.37901330e-01 6.53545380e-01 4.97344360e-02 -6.16335534e-02
4.44143891e-01 -2.34379042e-02 -1.01131439e+00 -3.28952700e-01
-1.40694246e-01 1.83884904e-01 6.57841444e-01 6.46511018e-01
-1.07241452e+00 1.17825854e+00 6.16760445e+00 2.34173164e-01
-1.28973496e+00 -1.55352369e-01 3.04038525e-01 2.23845094e-01
-2.84841329e-01 1.20327801e-01 -7.05172241e-01 3.51350904e-01
8.32046926e-01 -1.40355572e-01 5.49249649e-01 9.35062230e-01
3.20503891e-01 -2.98388779e-01 -1.21572256e+00 8.98904264e-01
-1.63753048e-01 -1.22777569e+00 2.29144096e-02 1.92263462e-02
1.01858664e+00 2.01541886e-01 2.94767432e-02 1.93387285e-01
3.25696528e-01 -6.13055825e-01 7.68627405e-01 6.77157462e-01
3.75122041e-01 -6.96696639e-01 8.21632326e-01 4.93938267e-01
-7.81775534e-01 -1.77589372e-01 -3.93752456e-01 -1.45697836e-02
1.00248501e-01 3.75826150e-01 -1.08099711e+00 3.84199411e-01
7.69125760e-01 7.93417573e-01 -7.27450132e-01 1.32227898e+00
-1.56487286e-01 3.41242284e-01 -3.27554911e-01 9.79062244e-02
2.56101102e-01 -1.77291572e-01 8.22848737e-01 9.74528730e-01
4.18345273e-01 -2.34207585e-01 1.31107062e-01 4.05638307e-01
1.82741120e-01 4.01126593e-02 -9.32425082e-01 2.50063598e-01
4.95750934e-01 7.30228961e-01 -5.04126787e-01 -2.30093107e-01
-1.18983597e-01 1.07909846e+00 5.51509142e-01 4.14371461e-01
-4.92641658e-01 2.96814609e-02 6.08597577e-01 7.35209733e-02
5.72174072e-01 -4.71422464e-01 -1.48833930e-01 -1.24950254e+00
1.72628254e-01 -9.72134054e-01 2.47772634e-01 -5.87573528e-01
-1.18171823e+00 4.31788623e-01 -4.30209041e-02 -1.51929092e+00
-6.52653337e-01 -5.29264450e-01 -3.28221351e-01 4.36112791e-01
-1.72771251e+00 -7.36133397e-01 -5.54659903e-01 6.55221522e-01
7.63805211e-01 -2.73351729e-01 6.67399049e-01 6.22712933e-02
-1.79955512e-01 4.91450906e-01 4.46588248e-01 -2.91222751e-01
8.79132271e-01 -1.30787194e+00 2.99200028e-01 9.19098258e-01
1.87050685e-01 2.58400559e-01 1.17809725e+00 -3.37970912e-01
-1.03904152e+00 -9.90255415e-01 2.68013120e-01 -3.03846180e-01
4.54420567e-01 -2.63761997e-01 -9.97500062e-01 5.66992342e-01
-8.89452621e-02 1.28729679e-02 8.71855319e-02 5.23218885e-02
-2.51381904e-01 -5.79493105e-01 -1.16165495e+00 7.48353958e-01
1.02170706e+00 -5.40960491e-01 -4.26627755e-01 2.50096768e-01
3.63515496e-01 -6.84183121e-01 -6.29197121e-01 5.75576186e-01
6.30168676e-01 -1.47352803e+00 8.45272005e-01 -4.09151852e-01
1.73335493e-01 -2.79853642e-01 -1.47300616e-01 -1.61091793e+00
-7.92484954e-02 -5.25651515e-01 -4.78531234e-03 7.72297144e-01
3.22521001e-01 -9.76583838e-01 9.81569231e-01 2.95893967e-01
-1.75750628e-01 -3.36878031e-01 -1.07356191e+00 -9.60022867e-01
-7.35178292e-02 -3.52246702e-01 4.31402653e-01 6.29272521e-01
-3.86537910e-01 1.08505845e-01 -2.91422635e-01 3.15647572e-01
8.41601431e-01 -1.75764665e-01 1.18855858e+00 -1.22532344e+00
-5.04943192e-01 -2.02651680e-01 -7.80483246e-01 -1.09474766e+00
2.39424407e-01 -2.61013001e-01 5.36783636e-01 -1.08087409e+00
-3.55814323e-02 -4.78342801e-01 -2.12813854e-01 3.42993997e-02
-1.95142224e-01 8.31573457e-02 -1.02152035e-01 1.87601134e-01
-5.13995886e-01 7.27806449e-01 1.07244885e+00 -1.40164182e-01
-4.13569897e-01 2.85481095e-01 -1.46780923e-01 8.49478483e-01
1.08430552e+00 -4.14093375e-01 -5.54810464e-01 -5.22557497e-01
1.91938058e-02 -1.72021955e-01 1.55759558e-01 -1.48365867e+00
1.52816117e-01 -3.19799334e-01 2.90611923e-01 -2.83210874e-01
4.66341496e-01 -8.55307817e-01 -4.08767909e-02 3.28372449e-01
-2.22578183e-01 -1.51425511e-01 3.87430280e-01 7.48525798e-01
-3.07631493e-01 -3.16063136e-01 1.09822834e+00 -1.83624700e-01
-1.22426438e+00 3.22221875e-01 -3.27256471e-01 1.08248144e-01
8.72583389e-01 -6.27520502e-01 1.37061685e-01 -7.01322615e-01
-8.28005195e-01 1.17929332e-01 9.59570289e-01 4.15150613e-01
5.14772236e-01 -1.02462566e+00 -3.98207843e-01 3.55548650e-01
5.17124295e-01 2.46545658e-01 9.30552706e-02 4.87669200e-01
-6.15310729e-01 2.56431639e-01 -4.72420782e-01 -8.79445076e-01
-9.33597207e-01 5.06284654e-01 6.12555861e-01 -7.97999576e-02
-6.18551135e-01 6.64422989e-01 2.47396991e-01 -5.93984962e-01
3.16307724e-01 -2.25600377e-01 -6.96947500e-02 -4.43975449e-01
1.61282614e-01 5.56200035e-02 1.71943188e-01 -4.54361230e-01
-1.44035146e-01 5.99228084e-01 -1.52834684e-01 -3.20026487e-01
1.31194293e+00 -5.81201971e-01 4.00475115e-01 7.69413352e-01
8.44089150e-01 -6.63397759e-02 -1.96491933e+00 -4.67431337e-01
3.19351226e-01 -6.87963784e-01 -9.83914435e-02 -4.84707057e-01
-7.93972790e-01 7.89015353e-01 7.86151886e-01 -1.31637976e-01
1.13304472e+00 -2.17112198e-01 6.99465990e-01 7.07011819e-01
7.21783876e-01 -1.60639644e+00 1.56003430e-01 6.93088651e-01
5.94370306e-01 -1.52714205e+00 -1.88991502e-01 -1.38865411e-01
-5.15518725e-01 1.28215683e+00 7.41034329e-01 1.62140336e-02
4.54955012e-01 1.27741233e-01 2.94257641e-01 1.00650981e-01
-7.58472025e-01 -3.63018155e-01 1.87534317e-01 8.60064447e-01
-5.69876097e-02 -2.59464383e-01 1.26001239e-01 -3.17243397e-01
-2.19533682e-01 2.51780171e-02 5.25180399e-01 9.34525669e-01
-6.41223907e-01 -1.08028972e+00 -2.05278590e-01 7.84294829e-02
1.31701022e-01 2.73112416e-01 -2.35404715e-01 7.86770880e-01
1.65479839e-01 6.42131984e-01 -8.76162946e-02 -3.56747985e-01
5.51620662e-01 5.06563894e-02 5.38114429e-01 -6.25256300e-01
-1.39211252e-01 -1.06337309e-01 1.24726892e-02 -7.49766946e-01
-5.09437025e-01 -9.79790092e-01 -1.01705945e+00 9.43458639e-03
-6.36955872e-02 -1.58224612e-01 5.75493574e-01 1.20725322e+00
1.70199350e-01 2.79044151e-01 7.83773780e-01 -9.23721313e-01
-5.60895264e-01 -6.03989422e-01 -2.92004079e-01 6.49536848e-01
5.50634086e-01 -9.56927240e-01 -4.34774041e-01 3.20043296e-01] | [8.691291809082031, -2.3132879734039307] |
3d889169-213a-4b9b-bb4f-6acca7e8b9f8 | when-worlds-collide-integrating-different | null | null | http://papers.nips.cc/paper/7220-when-worlds-collide-integrating-different-counterfactual-assumptions-in-fairness | http://papers.nips.cc/paper/7220-when-worlds-collide-integrating-different-counterfactual-assumptions-in-fairness.pdf | When Worlds Collide: Integrating Different Counterfactual Assumptions in Fairness | Machine learning is now being used to make crucial decisions about people's lives. For nearly all of these decisions there is a risk that individuals of a certain race, gender, sexual orientation, or any other subpopulation are unfairly discriminated against. Our recent method has demonstrated how to use techniques from counterfactual inference to make predictions fair across different subpopulations. This method requires that one provides the causal model that generated the data at hand. In general, validating all causal implications of the model is not possible without further assumptions. Hence, it is desirable to integrate competing causal models to provide counterfactually fair decisions, regardless of which causal "world" is the correct one. In this paper, we show how it is possible to make predictions that are approximately fair with respect to multiple possible causal models at once, thus mitigating the problem of exact causal specification. We frame the goal of learning a fair classifier as an optimization problem with fairness constraints entailed by competing causal explanations. We show how this optimization problem can be efficiently solved using gradient-based methods. We demonstrate the flexibility of our model on two real-world fair classification problems. We show that our model can seamlessly balance fairness in multiple worlds with prediction accuracy. | ['Matt J. Kusner', 'Chris Russell', 'Ricardo Silva', 'Joshua Loftus'] | 2017-12-01 | null | null | null | neurips-2017-12 | ['counterfactual-inference'] | ['miscellaneous'] | [ 3.21642309e-01 4.03840303e-01 -8.50112438e-01 -8.58727872e-01
-4.77175385e-01 -2.51204103e-01 6.09796584e-01 3.53112936e-01
-6.82290435e-01 1.35065448e+00 4.39495325e-01 -8.24192822e-01
-2.43393004e-01 -9.49132442e-01 -6.14364147e-01 -4.42777663e-01
1.33774638e-01 4.74549949e-01 -3.83747369e-01 -2.77325436e-02
3.68054599e-01 3.38233680e-01 -1.42052877e+00 2.47941568e-01
1.10936308e+00 6.49901450e-01 -6.10366583e-01 5.18527865e-01
2.34680977e-02 9.26615179e-01 -3.85772228e-01 -1.23171115e+00
3.00485492e-01 -4.52333272e-01 -8.22444081e-01 -4.00587618e-01
6.64759874e-01 -5.64477205e-01 1.09664142e-01 1.09707785e+00
2.81427890e-01 1.05364077e-01 9.95067835e-01 -1.63480246e+00
-4.42230374e-01 7.99646139e-01 -6.53619826e-01 -1.78830236e-01
2.15546697e-01 9.86243971e-03 1.30153203e+00 -8.53463933e-02
5.16570210e-01 1.64748597e+00 5.11206627e-01 7.43558407e-01
-1.60169959e+00 -1.04450405e+00 2.40049258e-01 -3.23496275e-02
-8.53853106e-01 -5.74216127e-01 5.00049770e-01 -5.33178687e-01
2.57408231e-01 7.41001666e-01 5.96538663e-01 1.00839472e+00
4.22673196e-01 3.59615117e-01 1.27750683e+00 -3.70576203e-01
3.75819564e-01 -4.78605442e-02 1.97039515e-01 4.59168583e-01
7.61462331e-01 4.90524918e-01 -4.88044918e-01 -6.66392207e-01
2.03362316e-01 -8.83940607e-02 -8.52773637e-02 -3.11371624e-01
-8.10148716e-01 1.28945184e+00 4.13062006e-01 -2.36008421e-01
-1.78782612e-01 3.03170919e-01 1.90365270e-01 1.55190185e-01
5.11563003e-01 6.09460294e-01 -3.73815626e-01 2.26847857e-01
-9.42977667e-01 9.12625849e-01 9.12348449e-01 2.89514959e-01
3.29369992e-01 -3.21582764e-01 -2.33350396e-01 4.59988594e-01
3.78914535e-01 4.24866557e-01 4.07167524e-02 -1.32206595e+00
4.54700589e-01 3.59023929e-01 6.28231764e-01 -9.87624109e-01
-2.69018412e-01 -8.25268105e-02 -7.66585767e-01 4.74484205e-01
1.04508936e+00 -4.25113648e-01 -4.91066545e-01 2.24968076e+00
4.24855232e-01 8.56015086e-02 -6.24806657e-02 9.07798231e-01
1.80299908e-01 2.84343690e-01 6.79946065e-01 -4.97686297e-01
1.13824403e+00 1.45505080e-02 -6.60631239e-01 -1.67451546e-01
4.95781392e-01 -6.09459221e-01 7.95474410e-01 9.78712365e-02
-8.90104353e-01 1.02328897e-01 -8.91180873e-01 -9.71608832e-02
-5.36826951e-03 -5.00365257e-01 1.11717057e+00 1.24056864e+00
-4.65303957e-01 9.18574750e-01 -2.66773015e-01 -9.35180187e-02
6.52272284e-01 4.56478059e-01 -2.17916563e-01 8.65018144e-02
-1.41518152e+00 8.71388495e-01 4.75927368e-02 -9.02459696e-02
-5.45873582e-01 -9.83081520e-01 -5.66816628e-01 2.55295008e-01
4.22406584e-01 -9.46654022e-01 1.22596014e+00 -1.27056968e+00
-9.72707152e-01 8.69359076e-01 -2.10701242e-01 -5.82454503e-01
1.13195002e+00 -1.26929104e-01 -1.99211270e-01 -5.11233330e-01
4.32679266e-01 5.31730831e-01 5.25314510e-01 -1.09731126e+00
-8.01579177e-01 -5.58859468e-01 3.39704305e-01 1.99895442e-01
5.72970174e-02 2.61648417e-01 4.88241613e-01 -4.08572435e-01
-3.82691890e-01 -9.84755874e-01 -3.32151145e-01 2.26132944e-01
-7.39069164e-01 -8.02893937e-03 1.91687316e-01 -3.30465287e-01
1.00704300e+00 -1.75955367e+00 -3.71868849e-01 3.07394296e-01
2.00976491e-01 -8.37241039e-02 2.40913346e-01 -1.38502732e-01
-1.27720281e-01 6.86620057e-01 -2.04115316e-01 -1.55189604e-01
2.06280023e-01 5.65195829e-03 -5.98749161e-01 6.76581860e-01
-6.14870042e-02 6.51744366e-01 -8.12553585e-01 -6.26820266e-01
-4.24656235e-02 6.56009540e-02 -9.07349229e-01 3.96026410e-02
-7.51418024e-02 9.27719846e-02 -3.10941011e-01 1.44358009e-01
7.66267598e-01 1.57785341e-01 5.48935175e-01 1.32372022e-01
-7.75459483e-02 4.60791677e-01 -1.19748724e+00 7.79648781e-01
-2.48979896e-01 3.12135220e-01 -7.61236027e-02 -9.26304162e-01
4.63244617e-01 -1.45581772e-03 4.14023027e-02 -3.38269562e-01
1.86811149e-01 2.29705527e-01 2.11448044e-01 -1.95203885e-01
3.14796209e-01 -1.12949038e+00 -2.54979342e-01 6.47836685e-01
-5.23439944e-01 2.92257387e-02 -1.75148785e-01 9.03874263e-02
5.92331171e-01 -2.33783752e-01 5.66319883e-01 -4.71331537e-01
9.52224880e-02 2.44932212e-02 1.02120245e+00 1.03148770e+00
-4.55273539e-01 3.13657284e-01 1.01383555e+00 -5.50867379e-01
-9.24818695e-01 -1.20819068e+00 -2.54298598e-01 9.24283981e-01
1.09298207e-01 2.43697301e-01 -5.25362730e-01 -7.58470237e-01
4.65563655e-01 1.37497640e+00 -9.61002648e-01 -2.28174001e-01
-2.32084543e-01 -1.18444502e+00 5.20279884e-01 9.14673209e-02
2.85705119e-01 -4.90462691e-01 -8.17468941e-01 -2.01199278e-01
-2.65763879e-01 -4.61554021e-01 -4.44723547e-01 -2.08291799e-01
-6.94247246e-01 -1.21965599e+00 -4.34143007e-01 1.25110582e-01
4.78706181e-01 -2.63860393e-02 1.10092533e+00 1.56549945e-01
-1.40878940e-02 -2.09009185e-01 1.53524846e-01 -5.98564565e-01
-5.68125308e-01 -3.74346733e-01 1.75212085e-01 1.36453927e-01
3.30184698e-01 -4.03989404e-01 -5.20445406e-01 1.07712477e-01
-5.02144933e-01 2.17860177e-01 3.60424444e-02 8.92559528e-01
1.84066162e-01 6.36815233e-03 6.82357907e-01 -1.43296134e+00
7.12486267e-01 -6.94890022e-01 -6.56172574e-01 4.70006615e-01
-6.59797132e-01 2.94536620e-01 5.14229119e-01 -2.64364839e-01
-1.58433354e+00 -1.89189419e-01 1.58104286e-01 2.82680541e-01
-1.02329627e-01 2.81447172e-01 -5.04345059e-01 3.65603119e-01
7.14035511e-01 -5.53373992e-01 -6.07116818e-02 -2.13557765e-01
4.70906764e-01 6.93340540e-01 1.62346557e-01 -9.05564725e-01
4.49936390e-01 5.02929986e-01 3.60126972e-01 -3.46037328e-01
-1.16625834e+00 3.27702194e-01 -2.64681876e-01 -1.83251172e-01
8.47540736e-01 -5.58384120e-01 -1.20164251e+00 8.69975388e-02
-1.04841268e+00 -3.34654778e-01 -2.57573396e-01 5.85943937e-01
-4.77822423e-01 -3.73820737e-02 -1.62064865e-01 -1.21010160e+00
6.18798174e-02 -9.02007043e-01 5.09271324e-01 2.03959584e-01
-7.09825337e-01 -8.97355318e-01 3.00843623e-02 6.07088566e-01
1.37597874e-01 4.09326196e-01 1.10639954e+00 -5.43170571e-01
-2.56923974e-01 -1.85689017e-01 -3.14262658e-01 -2.74957597e-01
9.60543901e-02 2.30355561e-01 -9.00315344e-01 4.69606966e-02
-1.33570969e-01 -3.07837099e-01 8.88917148e-01 7.49417961e-01
1.07396173e+00 -8.99508357e-01 -3.62610608e-01 3.87516558e-01
1.27987695e+00 1.29061088e-01 4.50400054e-01 -1.16698095e-03
4.31931436e-01 1.17824876e+00 5.59051037e-01 3.38675469e-01
6.76024377e-01 8.32490206e-01 3.22136909e-01 6.79145232e-02
3.56241435e-01 -4.78437483e-01 6.89985529e-02 -5.85560620e-01
-1.60335094e-01 -7.84264728e-02 -7.65898228e-01 4.45119828e-01
-1.94100094e+00 -1.44613826e+00 -2.56330401e-01 2.70159912e+00
8.97232413e-01 1.21799573e-01 3.28034431e-01 -1.08359307e-01
8.28914762e-01 -1.56005800e-01 -6.74934268e-01 -9.70875084e-01
6.93735927e-02 -1.62733749e-01 7.24888265e-01 9.48806167e-01
-1.10762131e+00 4.86084878e-01 6.39494276e+00 7.77295589e-01
-7.54098892e-01 -1.92281343e-02 1.47969782e+00 -5.40532887e-01
-8.70083690e-01 2.59308040e-01 -5.92619598e-01 5.77339292e-01
7.48672068e-01 -7.79388607e-01 3.11841279e-01 5.12428463e-01
4.74766195e-01 -4.01082635e-01 -1.22232103e+00 4.06735986e-01
-4.09605443e-01 -1.29644418e+00 1.63113445e-01 2.09994137e-01
8.16452384e-01 -5.57603359e-01 6.43886253e-02 -4.71754223e-02
1.07365739e+00 -1.39651835e+00 9.85284209e-01 4.07289207e-01
7.53835261e-01 -1.04704916e+00 6.34562790e-01 4.16807443e-01
-4.19084817e-01 -2.62345940e-01 -2.90929645e-01 -6.32758677e-01
1.26704359e-02 1.01328802e+00 -5.88849843e-01 3.76239568e-01
3.27381879e-01 1.21118195e-01 6.09020144e-02 8.44539821e-01
-3.29858601e-01 5.66897273e-01 -2.31813759e-01 -1.49474600e-02
-1.42743185e-01 1.40807375e-01 3.04331005e-01 9.52960253e-01
2.88325429e-01 3.91224176e-01 -2.50699013e-01 1.10006940e+00
-3.19632381e-01 1.79580227e-01 -6.25571370e-01 2.37428933e-01
5.32779396e-01 7.93659031e-01 -5.30471802e-01 -2.21204549e-01
-2.44831264e-01 4.65921938e-01 2.12110534e-01 1.59662515e-01
-7.13406563e-01 1.09357662e-01 1.14056230e+00 1.46336362e-01
-5.81762910e-01 4.99446243e-01 -8.08093309e-01 -1.32732332e+00
-3.48235726e-01 -7.89192796e-01 8.03889275e-01 -4.59770441e-01
-1.55939865e+00 -2.52686858e-01 1.53658614e-01 -6.42896593e-01
-2.82199770e-01 -3.56818646e-01 -6.83803022e-01 1.19032109e+00
-1.28134704e+00 -7.51854956e-01 4.38740373e-01 2.51207888e-01
-5.82157932e-02 1.68910950e-01 6.76471949e-01 -7.52144679e-03
-5.20517647e-01 5.86963832e-01 -2.69934207e-01 -1.06971659e-01
7.75890291e-01 -1.39125705e+00 1.16376683e-01 8.90966475e-01
-2.19082028e-01 6.84843898e-01 1.11415064e+00 -8.00679028e-01
-7.07138777e-01 -8.56381893e-01 1.55458152e+00 -2.77760953e-01
5.22079349e-01 -1.73768014e-01 -5.70723653e-01 6.44789398e-01
-1.98141798e-01 -2.64989406e-01 9.44642305e-01 6.76547587e-01
-5.44534862e-01 -2.31388226e-01 -1.56281209e+00 9.24374223e-01
1.03839195e+00 -1.53201208e-01 -5.76003015e-01 6.69307113e-02
5.26133597e-01 1.99557096e-02 -3.32444608e-01 2.02222958e-01
9.94974554e-01 -1.26425147e+00 9.00717497e-01 -1.18508804e+00
8.21500361e-01 -3.53509234e-03 -2.61696070e-01 -1.39575493e+00
-6.99174404e-02 -4.26762342e-01 5.57483315e-01 1.26462269e+00
6.25942469e-01 -8.59170377e-01 7.65329301e-01 1.64371610e+00
5.91372132e-01 -5.33542275e-01 -1.25641823e+00 -3.76466960e-01
6.93334162e-01 -6.28871977e-01 1.05719090e+00 1.28523958e+00
1.85328364e-01 2.10683599e-01 -7.21371889e-01 3.89245376e-02
1.05789483e+00 5.24245620e-01 6.68242931e-01 -1.39158225e+00
-2.48278439e-01 -5.80668032e-01 -1.00216985e-01 -2.49062613e-01
4.02965188e-01 -7.64713228e-01 -1.64799392e-01 -1.03494453e+00
7.68718660e-01 -8.04593742e-01 -9.22899880e-03 4.74805355e-01
-5.94494820e-01 -1.35968253e-01 4.93394673e-01 -1.66958719e-02
-5.00312038e-02 2.50024199e-01 1.04948533e+00 -1.39092728e-01
5.79207437e-03 2.35783666e-01 -1.40105009e+00 8.54549825e-01
9.74630117e-01 -7.12860107e-01 -3.86374325e-01 -2.66186707e-02
5.15599191e-01 4.26407903e-01 7.00068295e-01 -3.05349827e-01
3.13458499e-03 -9.59611118e-01 4.13101077e-01 2.40103588e-01
1.38405571e-02 -5.72998941e-01 4.08367753e-01 7.25403547e-01
-9.67722833e-01 -4.59722281e-01 -5.51398657e-02 4.29898828e-01
2.27525279e-01 -2.33584389e-01 8.82265270e-01 -7.80474395e-02
-2.97319368e-02 7.46561810e-02 -1.26942039e-01 2.44103417e-01
9.62663949e-01 2.88392603e-01 -5.00835061e-01 -5.53667128e-01
-5.78560174e-01 3.23185593e-01 6.21979117e-01 7.65148401e-02
2.09765330e-01 -1.24360251e+00 -9.97474790e-01 -2.33454973e-01
-1.38676420e-01 -7.11797595e-01 2.00575575e-01 4.75201815e-01
-5.56855462e-02 3.40633422e-01 -2.85279989e-01 2.36475229e-01
-1.29040313e+00 5.29985905e-01 6.35975659e-01 -4.07212764e-01
6.76873624e-02 7.33023763e-01 5.45900822e-01 -3.81973803e-01
-3.19814593e-01 1.08691506e-01 -5.38460799e-02 1.77738532e-01
5.42443514e-01 5.87328434e-01 -4.10841048e-01 -5.39067447e-01
-3.62538069e-01 1.00323364e-01 1.02524705e-01 -3.74826074e-01
1.14086914e+00 -1.82303891e-01 -1.71633005e-01 2.58823901e-01
7.99398541e-01 3.00639123e-01 -1.31897295e+00 3.89798015e-01
-1.63954139e-01 -1.00920284e+00 -1.34105235e-01 -1.06424654e+00
-8.60072315e-01 8.26307952e-01 2.89094508e-01 2.58119822e-01
8.77506912e-01 -4.15439010e-01 2.02568904e-01 -4.40723598e-02
4.12860453e-01 -8.70645165e-01 -8.08865011e-01 -3.21718812e-01
7.62516797e-01 -1.45340538e+00 2.44975790e-01 -3.76425594e-01
-6.53249025e-01 8.27938318e-01 2.84469485e-01 2.59648450e-02
5.46393037e-01 9.47798789e-02 -9.12320837e-02 1.80164024e-01
-8.26535106e-01 6.23422451e-02 2.89872020e-01 3.41000795e-01
5.81935525e-01 9.29908812e-01 -8.31235588e-01 6.33729637e-01
-5.39159238e-01 2.02723235e-01 6.54901743e-01 3.11624169e-01
-2.28050053e-01 -1.25704408e+00 -5.43828428e-01 9.96748507e-01
-8.69261503e-01 -2.30050702e-02 -5.08389652e-01 6.28090322e-01
3.03761125e-01 1.15038264e+00 7.40045682e-03 -5.04141636e-02
1.31312817e-01 -4.82196250e-04 3.90791416e-01 -4.27974910e-01
-2.03628570e-01 -4.36918139e-01 5.27319014e-01 -7.04528272e-01
-4.30685967e-01 -1.05973220e+00 -9.01087046e-01 -9.99215543e-01
-3.41744274e-02 2.00934112e-01 2.36645088e-01 1.05447364e+00
-1.62053555e-02 -1.61803905e-02 4.28016424e-01 -4.30127293e-01
-7.39223301e-01 -2.24446803e-01 -6.31878614e-01 4.16394591e-01
4.44785327e-01 -7.25538194e-01 -5.64528942e-01 -1.81112215e-01] | [8.766182899475098, 5.404683589935303] |
e0c03d7c-d4b5-4aec-a094-b56a1475d549 | leafmask-towards-greater-accuracy-on-leaf | 2108.03568 | null | https://arxiv.org/abs/2108.03568v1 | https://arxiv.org/pdf/2108.03568v1.pdf | LeafMask: Towards Greater Accuracy on Leaf Segmentation | Leaf segmentation is the most direct and effective way for high-throughput plant phenotype data analysis and quantitative researches of complex traits. Currently, the primary goal of plant phenotyping is to raise the accuracy of the autonomous phenotypic measurement. In this work, we present the LeafMask neural network, a new end-to-end model to delineate each leaf region and count the number of leaves, with two main components: 1) the mask assembly module merging position-sensitive bases of each predicted box after non-maximum suppression (NMS) and corresponding coefficients to generate original masks; 2) the mask refining module elaborating leaf boundaries from the mask assembly module by the point selection strategy and predictor. In addition, we also design a novel and flexible multi-scale attention module for the dual attention-guided mask (DAG-Mask) branch to effectively enhance information expression and produce more accurate bases. Our main contribution is to generate the final improved masks by combining the mask assembly module with the mask refining module under the anchor-free instance segmentation paradigm. We validate our LeafMask through extensive experiments on Leaf Segmentation Challenge (LSC) dataset. Our proposed model achieves the 90.09% BestDice score outperforming other state-of-the-art approaches. | ['Jun Yue', 'Zhenbo Li', 'Dantong Niu', 'Liao Qu', 'Ruohao Guo'] | 2021-08-08 | null | null | null | null | ['plant-phenotyping'] | ['computer-vision'] | [ 6.36958063e-01 -8.90594814e-03 -2.14266554e-01 -3.66691709e-01
-5.87011814e-01 -9.38201487e-01 -1.17302522e-01 1.42562717e-01
1.32759482e-01 4.13706899e-01 -3.04408699e-01 -4.36283797e-01
1.87893547e-02 -8.26262116e-01 -5.31226993e-01 -9.06966090e-01
2.55456716e-01 3.35983753e-01 4.74224031e-01 -1.17456272e-01
1.05897315e-01 6.72345340e-01 -1.43843806e+00 5.12431085e-01
1.41677248e+00 1.36196625e+00 8.81014943e-01 7.25833356e-01
-4.25942838e-01 -3.38598117e-02 -5.39915860e-01 1.39269233e-01
2.97950208e-01 -5.12526333e-01 -5.00477254e-01 3.90817821e-01
2.82419950e-01 -1.41788870e-01 4.74188298e-01 9.17881846e-01
7.86446035e-01 -4.71291870e-01 3.09819043e-01 -1.10471117e+00
-6.94206953e-01 8.93833101e-01 -1.24202764e+00 -1.47606164e-01
-4.82096910e-01 2.00852737e-01 1.06540692e+00 -7.69582570e-01
4.59650367e-01 9.54511583e-01 4.63866711e-01 1.99415818e-01
-1.54196990e+00 -2.44796395e-01 5.22144258e-01 9.50968461e-05
-1.30548286e+00 -2.10575342e-01 5.89867353e-01 -2.05658272e-01
3.49555075e-01 5.77934384e-01 5.48099339e-01 1.68553650e-01
-1.53373387e-02 9.70674992e-01 6.06164694e-01 -2.31811002e-01
1.24105796e-01 -2.36425579e-01 6.49021566e-02 8.45143974e-01
9.32409465e-02 -1.95825532e-01 1.24130502e-01 1.10028043e-01
6.18267715e-01 -2.63478160e-01 -4.44302887e-01 -4.31740493e-01
-8.24106455e-01 6.09653831e-01 7.70179749e-01 3.47992152e-01
-6.26751721e-01 -1.41670287e-01 1.41931042e-01 -5.40664077e-01
1.42340258e-01 4.76088375e-01 -1.04609728e+00 3.94803375e-01
-1.17158234e+00 7.68946409e-02 7.11816072e-01 1.06528926e+00
4.85872418e-01 2.94742510e-02 -1.04755223e+00 9.48503852e-01
2.80858397e-01 3.17794681e-01 2.18976587e-01 -1.07236111e+00
-3.68362926e-02 1.01476955e+00 -1.03113903e-02 -7.44834125e-01
-5.17851651e-01 -7.27867365e-01 -9.78700697e-01 3.27894121e-01
3.43736768e-01 -1.72508910e-01 -1.09990799e+00 1.86420619e+00
4.37840134e-01 7.23036081e-02 -1.83344856e-01 8.88549089e-01
1.19375288e+00 7.43039966e-01 3.57437253e-01 -1.45030707e-01
1.67657137e+00 -8.16909313e-01 -6.42983258e-01 -2.36206487e-01
6.78172290e-01 -6.70709133e-01 8.77945423e-01 1.39694408e-01
-1.04535949e+00 -7.78468549e-01 -1.13209462e+00 -5.91929369e-02
-3.62342060e-01 1.17365479e+00 7.04170346e-01 3.26167822e-01
-7.91673362e-01 7.22974241e-01 -6.02429450e-01 -3.20843220e-01
1.03074086e+00 4.82423484e-01 -1.30194321e-01 2.57705748e-01
-3.14766824e-01 3.95473927e-01 6.58662438e-01 6.88952148e-01
-8.26017976e-01 -8.71406615e-01 -5.09061038e-01 5.63638031e-01
6.77341521e-01 -3.06238145e-01 9.20612872e-01 -8.25500131e-01
-1.58735824e+00 1.11010373e+00 -3.20701003e-01 -2.38717943e-01
-1.20106891e-01 9.22713056e-02 5.79304919e-02 -2.02121735e-01
1.92784041e-01 1.35771203e+00 6.13076746e-01 -1.50724161e+00
-8.69877815e-01 -7.48360753e-01 -5.36071479e-01 9.60180461e-02
2.93445066e-02 -2.14971691e-01 -5.61389029e-01 -4.08353478e-01
5.97770810e-01 -9.01055813e-01 -3.09512675e-01 3.83680344e-01
-7.88062334e-01 1.04575232e-01 1.33615351e+00 -1.07051837e+00
1.21415699e+00 -1.98475063e+00 2.30741516e-01 -1.68103337e-01
2.20813140e-01 8.85186076e-01 -5.00910521e-01 -1.31508812e-01
-3.13616693e-01 3.60343277e-01 -7.12186515e-01 -4.56049703e-02
-3.49726737e-01 3.38704605e-03 -2.49618009e-01 8.92582089e-02
6.83986604e-01 1.25906682e+00 -4.35478091e-01 -5.11922538e-01
2.53766924e-01 3.72206396e-03 -2.81117082e-01 3.57307374e-01
-6.09556437e-01 5.45106947e-01 -3.65921289e-01 1.27314985e+00
1.22191823e+00 -1.14177912e-01 1.43234521e-01 -4.29960668e-01
-6.51979029e-01 -3.38076085e-01 -9.71145272e-01 1.58669698e+00
8.46205801e-02 2.65502781e-01 6.70418143e-01 -9.07918513e-01
1.28309369e+00 2.61740871e-02 6.47645295e-01 -1.36493042e-01
1.84552178e-01 -1.83555428e-02 3.38458151e-01 -1.27573818e-01
7.64045268e-02 7.13993251e-01 2.14473069e-01 -1.67177722e-01
3.31963003e-02 -1.78573489e-01 4.36983794e-01 1.34796882e-02
9.53258872e-01 5.26164532e-01 2.10923135e-01 -4.89780188e-01
6.85036659e-01 6.39366880e-02 1.21047199e+00 4.32590038e-01
-6.07908964e-01 5.84417164e-01 7.30858207e-01 -1.07565895e-01
-6.87936187e-01 -5.20965338e-01 -5.10805286e-02 1.13320923e+00
5.53009734e-02 -3.71936381e-01 -1.01389444e+00 -8.75324667e-01
1.42828196e-01 8.88196588e-01 -5.04901826e-01 -2.13780001e-01
-2.91568100e-01 -1.17524922e+00 5.58646441e-01 6.18551016e-01
9.11936462e-01 -1.39091241e+00 -7.17374623e-01 3.48691642e-01
-1.17059328e-01 -1.06213963e+00 -4.26259965e-01 7.15180635e-01
-7.11415112e-01 -8.30408633e-01 -4.58261490e-01 -9.01048601e-01
5.56713581e-01 3.55708778e-01 8.37269306e-01 1.81758940e-01
-7.27595747e-01 -5.95213532e-01 -3.42669219e-01 -5.35998523e-01
-1.98702008e-01 4.46135372e-01 -8.30686033e-01 4.20555398e-02
1.55253885e-02 -5.77255487e-01 -5.76956391e-01 2.83442646e-01
-7.87816823e-01 3.30441833e-01 7.74788260e-01 9.11995232e-01
1.04446661e+00 -4.76781614e-02 5.40712535e-01 -7.80183554e-01
1.50829390e-01 -7.22484961e-02 -1.32700491e+00 6.62279785e-01
-3.36671025e-01 -1.87086686e-01 6.91390634e-01 -1.34001762e-01
-1.01524282e+00 6.71871006e-01 -2.28200644e-01 1.40341744e-01
-4.52592820e-01 4.34740603e-01 -1.08458006e+00 -1.57505155e-01
3.32933277e-01 1.59218803e-01 -8.64640027e-02 -6.41962111e-01
5.92185855e-01 5.28051972e-01 7.96111524e-01 -4.12284046e-01
6.49538457e-01 2.83101231e-01 3.87437344e-01 -8.58109355e-01
-1.01681376e+00 -2.73061007e-01 -1.17479789e+00 -5.38668782e-02
8.67679358e-01 -4.76836622e-01 -1.09488606e+00 7.12045431e-01
-1.17976165e+00 -4.39858854e-01 -3.14351231e-01 -2.52398163e-01
-3.31687748e-01 2.72365153e-01 -3.73947620e-01 -6.42646074e-01
-6.45449817e-01 -1.25445485e+00 1.43294942e+00 9.28125322e-01
2.28493229e-01 -4.11892176e-01 -4.13668752e-01 4.34848852e-02
2.67597109e-01 4.08991843e-01 9.43713486e-01 -3.79168868e-01
-7.21653938e-01 -1.79115087e-01 -8.80067229e-01 2.81606585e-01
2.31723145e-01 6.30572736e-01 -1.04588675e+00 -5.66603765e-02
-4.33305353e-01 -1.48676544e-01 1.03009439e+00 9.64622319e-01
1.80411160e+00 2.48066917e-01 -4.56876189e-01 1.13541162e+00
1.62287116e+00 3.91156882e-01 5.65512300e-01 -6.45882189e-02
8.71303976e-01 6.67662740e-01 6.69805646e-01 4.17947918e-01
-7.46789994e-03 5.95627606e-01 7.68362880e-01 -5.81336617e-01
-1.61178142e-01 -5.94522059e-02 -1.79199383e-01 2.43253455e-01
3.96973908e-01 -7.80369341e-01 -6.11470580e-01 4.48391587e-01
-1.93736029e+00 -6.30905747e-01 -4.04501677e-01 2.05036426e+00
7.78610110e-01 5.34052029e-02 -1.47286311e-01 3.34302969e-02
8.92571151e-01 1.19277380e-01 -1.02418184e+00 -3.03937942e-01
-5.92349231e-01 1.35166422e-01 8.10817301e-01 3.65757227e-01
-1.48739481e+00 1.60133278e+00 5.22832203e+00 8.94612372e-01
-8.66378903e-01 -2.39044979e-01 1.03843927e+00 5.08965552e-01
9.35082957e-02 3.98720771e-01 -1.24850881e+00 3.72212827e-01
5.04970849e-01 5.06103814e-01 3.56261492e-01 7.24087536e-01
3.67952734e-01 -2.62792647e-01 -5.95744789e-01 2.46318102e-01
-3.33219141e-01 -1.22744536e+00 -2.38506626e-02 -4.54039685e-02
4.78176832e-01 -3.62896591e-01 -2.62474865e-01 1.75289456e-02
4.47007805e-01 -8.14835727e-01 4.10269976e-01 2.47132659e-01
9.05274868e-01 -2.28740275e-01 4.07056481e-01 3.37976813e-01
-1.45941305e+00 -3.09072495e-01 -6.48591876e-01 5.23607016e-01
2.66854484e-02 8.11516047e-01 -7.91518927e-01 6.25530779e-01
4.06272888e-01 4.61710066e-01 -9.64783549e-01 1.32186306e+00
-2.36550674e-01 1.02163637e+00 -4.97539431e-01 2.12639123e-01
2.04443276e-01 -2.90574282e-01 5.01364589e-01 9.79983270e-01
2.90511310e-01 1.35002611e-02 5.36997080e-01 1.46781158e+00
6.01768009e-02 1.17298089e-01 -9.67981219e-02 -1.99945509e-01
6.21473908e-01 1.96715486e+00 -1.34612358e+00 -1.39838085e-01
-5.84041141e-03 1.19937027e+00 2.11845636e-01 -5.19850152e-03
-7.08166361e-01 -4.34871286e-01 3.54602784e-01 -2.84151107e-01
7.49899805e-01 -7.01745375e-05 -7.08883107e-01 -4.84640867e-01
-2.19643682e-01 -8.13726008e-01 2.35492349e-01 -8.22935998e-01
-8.36423576e-01 3.69461477e-01 -4.48767513e-01 -4.79589760e-01
3.96335363e-01 -6.02851391e-01 -8.52041841e-01 1.31700635e+00
-1.24790800e+00 -1.49333167e+00 -8.76758099e-01 -2.51682341e-01
5.62514544e-01 7.61003271e-02 1.13065183e+00 1.71275139e-01
-1.10714626e+00 4.40370709e-01 2.56940514e-01 -4.99356352e-02
5.54016709e-01 -1.23678195e+00 5.91854274e-01 1.02223742e+00
-6.87585250e-02 -1.05101820e-02 2.32042894e-01 -8.90657723e-01
-1.09652150e+00 -1.34658766e+00 5.40787756e-01 -3.28326002e-02
1.75194617e-03 -2.94929445e-01 -8.77805650e-01 4.87334907e-01
-7.71412328e-02 -5.92153035e-02 2.13562265e-01 -2.89271772e-01
2.71336019e-01 -4.53779548e-01 -1.18449402e+00 6.89692557e-01
7.51712620e-01 3.25141907e-01 3.29046875e-01 1.58519164e-01
1.04048419e+00 -2.84291983e-01 -5.91919482e-01 9.13657427e-01
4.42152709e-01 -5.69114685e-01 8.42369556e-01 -3.78038704e-01
3.16875279e-01 -8.03425372e-01 -2.99895376e-01 -1.30886734e+00
-1.14687598e+00 -6.51977539e-01 1.45176530e-01 1.90869164e+00
4.82892424e-01 -7.28282779e-02 9.18926239e-01 1.89480618e-01
-5.07482886e-01 -8.85015190e-01 -3.34069282e-01 -3.03252876e-01
-2.24220216e-01 3.97278257e-02 9.61626112e-01 3.87246937e-01
-8.66284847e-01 1.77275971e-01 -9.32434946e-02 3.80061567e-01
2.96200514e-01 5.41829050e-01 4.68342245e-01 -1.40738010e+00
-8.59684870e-02 -7.32461214e-01 1.78612527e-02 -1.14789927e+00
-3.87421101e-02 -1.01859879e+00 5.15469730e-01 -1.41161358e+00
1.90890253e-01 -3.64535809e-01 -2.78871320e-02 6.61116123e-01
-6.46532536e-01 -8.50689635e-02 2.59090483e-01 -2.88212448e-01
-1.66490033e-01 4.97957587e-01 1.41731989e+00 -2.12771177e-01
-4.81385976e-01 3.19594800e-01 -1.16310275e+00 6.64779365e-01
1.09700406e+00 -2.48005316e-01 -2.96463042e-01 -3.90632808e-01
-5.57182312e-01 -2.01417506e-01 3.68117392e-01 -8.18111360e-01
-2.18825907e-01 -2.26614356e-01 7.42111444e-01 -1.22873819e+00
2.32441187e-01 -6.75334275e-01 -1.24440782e-01 4.85881567e-01
-4.76169765e-01 -2.67011747e-02 4.93712604e-01 1.87124297e-01
3.88724357e-01 -6.89368367e-01 9.99859393e-01 -1.07082613e-01
-8.50360572e-01 4.03758019e-01 -2.60464609e-01 -2.08017573e-01
1.29241407e+00 -6.23951927e-02 -3.77273053e-01 1.04259469e-01
-5.22373974e-01 5.72188377e-01 -6.34564599e-03 2.91526109e-01
2.40750611e-01 -1.01793897e+00 -9.11685646e-01 4.28874254e-01
9.04388577e-02 1.08052276e-01 1.01518750e-01 5.75416207e-01
-7.56709635e-01 5.65262079e-01 -3.97171855e-01 -7.28686750e-01
-1.29594624e+00 3.80658835e-01 2.70871311e-01 -3.47649336e-01
-3.67268831e-01 1.18589151e+00 5.18506825e-01 -5.99803925e-01
3.93941849e-01 -5.28912961e-01 -4.19373244e-01 -1.37878507e-01
1.88847050e-01 1.89068288e-01 -1.54474517e-02 -5.44971704e-01
-2.16253862e-01 3.61745447e-01 2.27458343e-01 2.29883045e-01
1.15477741e+00 7.09143132e-02 -4.04936403e-01 2.94375211e-01
4.45085198e-01 -3.65659863e-01 -1.33159781e+00 2.49986470e-01
2.22893119e-01 -3.21935475e-01 2.82024145e-01 -1.48117757e+00
-1.22787333e+00 7.15296805e-01 9.78917837e-01 2.72347093e-01
1.52356017e+00 -3.58482540e-01 8.08071196e-01 5.32284565e-02
-2.07551643e-01 -9.24148321e-01 -7.64375746e-01 2.20832407e-01
9.26869214e-01 -1.25829399e+00 -2.83237338e-01 -1.00916934e+00
-4.10167515e-01 7.63535380e-01 1.07645643e+00 1.86185017e-01
7.28539824e-01 5.75272739e-01 -7.98165277e-02 3.18547338e-02
-8.83869588e-01 -5.94457328e-01 1.37014955e-01 7.74241090e-01
6.56292677e-01 2.41081238e-01 -5.40170133e-01 1.08480036e+00
2.86504924e-01 -8.52737427e-02 2.55179852e-02 7.06037939e-01
-7.24317610e-01 -1.26815331e+00 -3.74143511e-01 5.71221828e-01
-3.10998499e-01 -1.46231696e-01 -8.06068659e-01 3.17786634e-01
5.14732122e-01 7.98936963e-01 -1.98517486e-01 -2.26435125e-01
2.00126410e-01 1.69296607e-01 2.05765888e-01 -6.87141061e-01
-6.35727406e-01 2.81297356e-01 -2.40470126e-01 -4.73247737e-01
1.83371142e-01 -6.05686367e-01 -1.03813195e+00 -1.44608440e-02
-7.50722468e-01 -2.26213381e-01 7.26519048e-01 5.88935137e-01
9.47967768e-01 1.08160985e+00 5.21483243e-01 -6.67549491e-01
-3.37322772e-01 -1.14257753e+00 -6.03044152e-01 6.06993735e-02
-1.73434675e-01 -3.45371962e-01 3.43020737e-01 2.13048086e-01] | [9.05756664276123, -1.6268718242645264] |
399204a1-02e2-445d-9545-0e07213bef41 | afrinames-most-asr-models-butcher-african | 2306.00253 | null | https://arxiv.org/abs/2306.00253v2 | https://arxiv.org/pdf/2306.00253v2.pdf | AfriNames: Most ASR models "butcher" African Names | Useful conversational agents must accurately capture named entities to minimize error for downstream tasks, for example, asking a voice assistant to play a track from a certain artist, initiating navigation to a specific location, or documenting a laboratory result for a patient. However, where named entities such as ``Ukachukwu`` (Igbo), ``Lakicia`` (Swahili), or ``Ingabire`` (Rwandan) are spoken, automatic speech recognition (ASR) models' performance degrades significantly, propagating errors to downstream systems. We model this problem as a distribution shift and demonstrate that such model bias can be mitigated through multilingual pre-training, intelligent data augmentation strategies to increase the representation of African-named entities, and fine-tuning multilingual ASR models on multiple African accents. The resulting fine-tuned models show an 81.5\% relative WER improvement compared with the baseline on samples with African-named entities. | ['Sahib Singh', 'Amina Mardiyyah Rufai', 'Chris Chinenye Emezue', 'Atnafu Lambebo Tonja', 'Bonaventure F. P. Dossou', 'Tejumade Afonja', 'Tobi Olatunji'] | 2023-06-01 | null | null | null | null | ['automatic-speech-recognition'] | ['speech'] | [ 2.90937304e-01 4.04986233e-01 5.32882437e-02 -5.89426994e-01
-1.18650746e+00 -7.01122463e-01 5.42370617e-01 5.29679433e-02
-7.68766403e-01 1.14640307e+00 7.25809157e-01 -6.65783048e-01
1.57779574e-01 -5.23188233e-01 -4.08682644e-01 -4.67409730e-01
4.07884978e-02 8.24588120e-01 -1.25950933e-01 -4.25110191e-01
3.24702151e-02 3.30523044e-01 -4.66778129e-01 3.59570265e-01
9.37131286e-01 2.11494073e-01 2.73126066e-01 8.83897305e-01
-2.16168493e-01 1.01098740e+00 -1.28560543e+00 -5.38652778e-01
-1.71403185e-01 -3.85163397e-01 -1.16417849e+00 -2.37305939e-01
-1.74440481e-02 -2.20034003e-01 -4.01061267e-01 8.57796371e-01
5.88352263e-01 3.40609789e-01 4.89756495e-01 -7.57776082e-01
-6.84339106e-01 1.18252563e+00 -1.79631785e-01 2.59382099e-01
2.94107795e-01 7.40619674e-02 8.85489643e-01 -7.84113824e-01
8.40631783e-01 1.08442366e+00 7.46164382e-01 7.83016384e-01
-9.16457057e-01 -7.23067939e-01 2.32129768e-01 -1.18695833e-01
-1.44604266e+00 -9.51504111e-01 2.89583176e-01 -2.01933384e-01
1.09241331e+00 4.06055868e-01 7.63478056e-02 1.29116130e+00
-3.30453545e-01 6.78094387e-01 7.77841330e-01 -2.83128709e-01
-1.37014268e-02 2.43037060e-01 -2.40485501e-02 4.35676575e-01
-2.88937867e-01 -9.36477557e-02 -9.09340560e-01 -2.88970739e-01
3.74274373e-01 -4.26253289e-01 -4.43742365e-01 4.55285698e-01
-1.40921247e+00 7.39045560e-01 3.06171805e-01 4.15457129e-01
-3.86498392e-01 -1.33709148e-01 2.81520158e-01 3.27859938e-01
4.11225259e-01 8.36606324e-01 -9.11554158e-01 -6.43714249e-01
-6.80971503e-01 -6.54327422e-02 8.95347834e-01 1.17296469e+00
3.42183203e-01 3.45961124e-01 -4.11204621e-02 1.35957277e+00
2.20009819e-01 6.48610473e-01 4.17396247e-01 -9.34727788e-01
7.61643946e-01 3.44076484e-01 3.52318704e-01 -2.09315538e-01
-4.27520186e-01 -2.83297271e-01 -5.36718369e-01 -2.94562757e-01
9.33933854e-01 -8.99508476e-01 -9.74138021e-01 1.92489326e+00
1.81295723e-01 -1.50214314e-01 6.70928657e-01 8.34830582e-01
7.05413640e-01 8.85244668e-01 3.84198546e-01 -3.30820978e-01
1.46618140e+00 -9.84727025e-01 -9.24607813e-01 -4.86369640e-01
7.27064967e-01 -9.92453873e-01 7.01375902e-01 -9.67362002e-02
-9.79016304e-01 -3.03425550e-01 -4.91568148e-01 9.92257446e-02
-4.33848858e-01 4.76450399e-02 4.20773119e-01 9.05986369e-01
-9.56110001e-01 1.28260374e-01 -7.07169712e-01 -6.56740844e-01
-1.13342717e-01 2.71655381e-01 -4.27328855e-01 -1.26360971e-02
-1.57509208e+00 9.43156481e-01 1.60567939e-01 3.05772722e-01
-6.89185619e-01 -7.41684318e-01 -1.07701397e+00 -2.64824569e-01
4.93507944e-02 -2.69402146e-01 1.46228349e+00 -7.68262506e-01
-1.58956909e+00 7.39343405e-01 -5.48399508e-01 -3.56012851e-01
3.45205635e-01 -3.35463434e-01 -9.25936282e-01 -1.18508123e-01
1.11613601e-01 5.07694304e-01 4.32077497e-01 -8.49976659e-01
-9.25496459e-01 -4.48793322e-01 -4.12961096e-02 4.24462199e-01
6.53772429e-02 6.07709885e-01 -1.49789870e-01 -8.57183576e-01
-1.27523348e-01 -1.03153026e+00 -2.31238589e-01 -9.28246617e-01
-4.99996603e-01 -1.49486095e-01 5.48741698e-01 -1.29244351e+00
1.27625418e+00 -2.08264589e+00 4.77595218e-02 1.51527405e-01
-2.78967917e-01 4.29363728e-01 -3.72543484e-02 4.73437458e-01
-5.51084848e-03 1.51532426e-01 -3.87630492e-01 -9.07948911e-02
-1.98664859e-01 1.74318969e-01 -9.72931460e-02 1.33801520e-01
3.95071089e-01 6.47991538e-01 -9.45505142e-01 -1.35338558e-02
-5.88359311e-02 5.68130016e-01 -3.31550717e-01 2.44287580e-01
1.32340595e-01 7.16543198e-01 -8.09622258e-02 4.03462708e-01
2.56354630e-01 1.11702539e-01 5.39773762e-01 2.16963664e-01
-2.39706039e-01 9.88532424e-01 -9.21605051e-01 1.47375453e+00
-6.78284883e-01 6.57439530e-01 4.82669115e-01 -4.60609972e-01
8.09235096e-01 6.42845988e-01 7.61824250e-02 -4.04767126e-01
-7.62661826e-03 4.46193397e-01 4.16131198e-01 -2.70924181e-01
7.88089693e-01 -1.74340203e-01 -2.67726421e-01 3.84586662e-01
-3.18057120e-01 -1.64609961e-02 -2.17018470e-01 1.31382391e-01
1.22340798e+00 -2.69691378e-01 1.89145893e-01 -6.78270012e-02
5.37818134e-01 1.77073821e-01 6.11238599e-01 7.27145255e-01
-4.81579691e-01 5.11952698e-01 1.02821670e-01 1.93554293e-02
-9.75228667e-01 -9.77382421e-01 -4.25015017e-02 1.55215001e+00
-3.36872578e-01 -1.97896823e-01 -7.67930984e-01 -7.16403902e-01
-3.63945156e-01 1.18627715e+00 -7.54781514e-02 5.98797053e-02
-1.13662589e+00 -5.88759840e-01 1.03849471e+00 6.19781435e-01
3.67962599e-01 -1.20972645e+00 3.12750600e-02 3.93923700e-01
-5.14859736e-01 -1.21125185e+00 -8.39086354e-01 2.88507104e-01
-2.74492443e-01 -5.84670782e-01 -9.77886856e-01 -9.24432456e-01
4.43586677e-01 -1.40958562e-01 9.78379190e-01 -2.95583695e-01
1.71351895e-01 1.34771004e-01 -4.06972229e-01 -1.53367057e-01
-8.23408246e-01 4.08218682e-01 2.96002239e-01 -1.29404932e-01
5.33794463e-01 -1.47966474e-01 -2.81140894e-01 3.29312056e-01
-2.64636666e-01 -2.20228404e-01 4.41238254e-01 8.62039387e-01
-8.27002749e-02 -3.48849446e-01 1.06075358e+00 -1.17869616e+00
6.80830956e-01 -4.07919526e-01 -1.12083644e-01 2.93127716e-01
-1.61517382e-01 9.21052247e-02 5.72576344e-01 -3.83189142e-01
-1.45305705e+00 -9.18958336e-02 -6.73620641e-01 2.61931479e-01
-5.28535187e-01 2.73923516e-01 -3.81403029e-01 4.93646830e-01
5.90740860e-01 2.48540804e-01 -3.25054437e-01 -2.09157616e-01
3.61513436e-01 1.30181193e+00 6.21357918e-01 -4.08625901e-01
3.67107064e-01 2.64973547e-02 -9.18822289e-01 -1.27651250e+00
-4.22954530e-01 -5.77637494e-01 -3.80258679e-01 3.81337032e-02
1.04447305e+00 -9.88512635e-01 -7.90089846e-01 5.93349934e-01
-1.39197958e+00 -4.84295279e-01 1.93953682e-02 7.75095820e-01
-2.34700024e-01 1.99366566e-02 -9.03659225e-01 -1.12072992e+00
-2.91338265e-01 -1.07032371e+00 7.96863198e-01 1.82279363e-01
-8.39949191e-01 -1.01541948e+00 2.85702795e-02 8.50734591e-01
5.83418429e-01 -4.28755194e-01 9.47751284e-01 -1.31721509e+00
-1.50806025e-01 2.01615497e-01 -2.11584225e-01 2.24096939e-01
3.39929193e-01 -3.15356612e-01 -8.53329062e-01 -7.05052614e-02
-4.25671488e-01 2.03296933e-02 2.23574370e-01 -4.89109531e-02
1.87357277e-01 -5.09182930e-01 -2.09950477e-01 1.85631484e-01
5.45930147e-01 8.33639860e-01 1.41364202e-01 9.28730369e-02
6.02782965e-01 7.25904644e-01 2.83781916e-01 2.70433933e-01
7.21840262e-01 6.21655285e-01 -2.36247241e-01 1.79700986e-01
-1.90876439e-01 -1.69331387e-01 6.58693731e-01 1.09080887e+00
4.35313806e-02 -3.47692728e-01 -1.09218001e+00 1.02082658e+00
-1.51810026e+00 -7.59745181e-01 -1.63570181e-01 2.07385802e+00
1.07733524e+00 -1.29220158e-01 3.58344689e-02 -3.18818510e-01
1.10098195e+00 1.08125836e-01 -3.54634225e-01 -5.14034152e-01
2.49220263e-02 2.51333535e-01 6.31136477e-01 8.74172151e-01
-8.31061661e-01 1.38164711e+00 6.10450315e+00 4.21055824e-01
-9.63447630e-01 2.59790719e-01 5.22283375e-01 6.36152029e-02
-2.29554132e-01 -1.10782482e-01 -9.51117337e-01 3.51611227e-01
1.46570182e+00 1.98559821e-01 4.76677626e-01 7.11488307e-01
3.35033000e-01 1.70266479e-01 -9.62483168e-01 8.72068226e-01
-4.44210041e-03 -9.78074312e-01 -2.63074100e-01 -1.56009272e-01
5.65194666e-01 2.67083108e-01 -1.77146047e-01 4.70584035e-01
1.02951562e+00 -9.75869536e-01 4.98437405e-01 2.07313195e-01
5.20425200e-01 -1.18516362e+00 7.57274389e-01 3.35663885e-01
-8.15037191e-01 2.33651385e-01 -1.49373198e-02 2.31554538e-01
4.47536767e-01 1.43536061e-01 -1.55038905e+00 4.21861023e-01
4.41261530e-01 -1.46127284e-01 2.93410365e-02 5.82426012e-01
-5.02901971e-01 1.00623894e+00 -2.89700538e-01 -1.14712425e-01
2.63997972e-01 -8.68091360e-03 7.84934759e-01 1.51445520e+00
2.41262302e-01 3.79840732e-01 -1.48201391e-01 2.35479921e-01
-3.40670228e-01 2.22059578e-01 -4.00226444e-01 -3.06187421e-01
7.42532253e-01 9.34130073e-01 -3.42476189e-01 -2.53944814e-01
-2.73356557e-01 1.29722953e+00 4.59325135e-01 5.80745101e-01
-5.49532592e-01 -7.56106853e-01 1.20319152e+00 5.24774119e-02
4.12214436e-02 -4.43815112e-01 -1.44532576e-01 -8.72594774e-01
-3.17787409e-01 -1.17247427e+00 4.02804583e-01 -7.20384777e-01
-1.02822959e+00 9.98173773e-01 -5.17675400e-01 -5.54747403e-01
-7.57271767e-01 -4.63898540e-01 -2.62004107e-01 1.03681171e+00
-1.33168542e+00 -1.12595749e+00 3.13771218e-01 4.98809814e-01
7.09295273e-01 -3.95653814e-01 9.59020615e-01 4.46154803e-01
-5.07592559e-01 8.05678844e-01 -1.03632085e-01 7.56888688e-01
1.01382935e+00 -1.23270679e+00 6.76573932e-01 9.87568319e-01
2.28495121e-01 9.07482922e-01 5.91326773e-01 -5.78380764e-01
-1.00566280e+00 -1.05922771e+00 1.61490810e+00 -6.84581339e-01
7.77814031e-01 -3.65658760e-01 -9.56531942e-01 9.36752081e-01
4.37471211e-01 -4.85419482e-01 9.90915358e-01 4.44348037e-01
-3.22882384e-01 1.75504535e-01 -1.07246578e+00 7.70205081e-01
1.14075565e+00 -8.11453938e-01 -6.72857463e-01 2.86394209e-01
9.51182902e-01 -4.43370789e-01 -9.59183335e-01 9.63524431e-02
4.43717122e-01 -5.72366118e-01 5.55245399e-01 -1.05553937e+00
-1.78089380e-01 -2.34253123e-01 -1.28009170e-01 -1.69270098e+00
-8.30540583e-02 -1.09512687e+00 3.62601846e-01 1.53305793e+00
9.88147855e-01 -6.00911438e-01 4.97680187e-01 7.85846531e-01
-4.05190796e-01 1.32564247e-01 -9.92502332e-01 -3.84987831e-01
-9.09106955e-02 -6.66691184e-01 6.08982682e-01 1.24469090e+00
2.41534129e-01 4.90298569e-01 -4.36890692e-01 6.46251082e-01
1.15291066e-01 -4.85695273e-01 5.24561048e-01 -6.01060748e-01
-1.47282124e-01 -1.92295149e-01 -4.29465398e-02 -1.14976335e+00
3.63040745e-01 -6.66690946e-01 4.29162920e-01 -1.41904473e+00
-4.32524979e-01 -9.16814625e-01 4.42007743e-03 5.81043065e-01
-3.57276827e-01 1.39646903e-01 1.00569487e-01 -1.57427624e-01
-2.90855736e-01 1.86698988e-01 7.19618678e-01 -4.58393944e-03
-4.44929302e-01 3.26959580e-01 -7.41211236e-01 5.90084970e-01
6.39205396e-01 -4.62471873e-01 -6.65361360e-02 -7.17909098e-01
1.63446903e-01 5.25479257e-01 -3.21326762e-01 -6.42412066e-01
4.43689018e-01 -2.33860984e-01 1.19857229e-01 -8.52968544e-02
5.71392834e-01 -6.32544696e-01 9.56202671e-02 3.17341715e-01
-4.88425076e-01 6.10521100e-02 1.18240722e-01 2.26131067e-01
-3.14377785e-01 -1.81640014e-01 7.68196642e-01 -1.12769485e-01
-7.65506506e-01 -2.29807511e-01 -9.09768760e-01 3.26536536e-01
6.96706057e-01 1.42074570e-01 -5.96193671e-01 -7.62723207e-01
-9.25509453e-01 2.99265295e-01 6.80252761e-02 5.85682809e-01
3.53737712e-01 -9.77308750e-01 -9.56368923e-01 2.36498579e-01
1.52284324e-01 -1.39662594e-01 2.56381035e-01 6.21123374e-01
-4.73307997e-01 4.73608702e-01 2.24971727e-01 -8.93158987e-02
-1.20317328e+00 -4.49013300e-02 2.71053553e-01 -1.73309624e-01
-1.27542421e-01 1.35485148e+00 3.07996292e-02 -1.17804778e+00
1.75352007e-01 -8.38506222e-02 -7.91827366e-02 1.25137269e-01
4.54610586e-01 5.07015884e-01 2.26724789e-01 -1.04997718e+00
-7.93761730e-01 -4.27999310e-02 -2.39333376e-01 -4.53986138e-01
1.03121889e+00 -3.30933988e-01 -1.79841351e-02 4.15578395e-01
8.31465483e-01 7.40509391e-01 -7.96612859e-01 -1.29689112e-01
1.71128675e-01 9.70012173e-02 -4.53366339e-01 -1.11009204e+00
-6.31572902e-01 5.99158287e-01 1.47837132e-01 -1.30070979e-02
6.07685149e-01 1.41147673e-01 8.89666736e-01 5.13220251e-01
4.51634139e-01 -1.13003302e+00 -4.97014314e-01 1.08278978e+00
5.80918014e-01 -1.15235877e+00 -6.84951425e-01 -1.53229952e-01
-1.08269382e+00 8.31216514e-01 4.04122412e-01 4.06303406e-01
4.86537129e-01 2.69249111e-01 7.28518069e-01 2.06805080e-01
-6.61966980e-01 -2.96254307e-01 -3.02857179e-02 6.98286235e-01
6.96142495e-01 3.43269110e-01 -1.06201626e-01 3.91802937e-01
-7.11752236e-01 -6.59909070e-01 4.64312643e-01 7.09757268e-01
-2.10052580e-01 -1.00546813e+00 -3.70505333e-01 1.63935184e-01
-7.82134652e-01 -5.60231566e-01 -5.98521173e-01 6.72073007e-01
-6.84392378e-02 1.35300744e+00 3.44583005e-01 -2.02769369e-01
4.48110819e-01 7.05739379e-01 -4.02412079e-02 -8.51130068e-01
-9.87963974e-01 7.50261843e-02 8.47617805e-01 3.07421312e-02
-2.43869931e-01 -7.20915079e-01 -1.45144868e+00 -3.08867753e-01
-2.79245913e-01 5.45258164e-01 6.65431380e-01 9.86656547e-01
3.80484670e-01 4.95158434e-01 3.53202283e-01 -9.27279815e-02
-2.16102809e-01 -1.30195630e+00 -2.60558695e-01 1.92139506e-01
2.84084409e-01 3.09069199e-03 -3.37308973e-01 9.92765501e-02] | [14.059050559997559, 6.968752861022949] |
3256c2a0-6e6a-41b1-8b21-6ef9df75ef73 | meeting-summarization-with-pre-training-and | 2111.08210 | null | https://arxiv.org/abs/2111.08210v1 | https://arxiv.org/pdf/2111.08210v1.pdf | Meeting Summarization with Pre-training and Clustering Methods | Automatic meeting summarization is becoming increasingly popular these days. The ability to automatically summarize meetings and to extract key information could greatly increase the efficiency of our work and life. In this paper, we experiment with different approaches to improve the performance of query-based meeting summarization. We started with HMNet\cite{hmnet}, a hierarchical network that employs both a word-level transformer and a turn-level transformer, as the baseline. We explore the effectiveness of pre-training the model with a large news-summarization dataset. We investigate adding the embeddings of queries as a part of the input vectors for query-based summarization. Furthermore, we experiment with extending the locate-then-summarize approach of QMSum\cite{qmsum} with an intermediate clustering step. Lastly, we compare the performance of our baseline models with BART, a state-of-the-art language model that is effective for summarization. We achieved improved performance by adding query embeddings to the input of the model, by using BART as an alternative language model, and by using clustering methods to extract key information at utterance level before feeding the text into summarization models. | ['Xiang Xiao', 'Wei Ji', 'Andras Huebner'] | 2021-11-16 | null | null | null | null | ['meeting-summarization'] | ['natural-language-processing'] | [ 2.21776024e-01 6.61378920e-01 1.99858636e-01 -3.29846114e-01
-1.36035550e+00 -3.66961420e-01 7.87527204e-01 5.82043111e-01
-6.32474303e-01 6.04520023e-01 1.29967213e+00 -8.35520551e-02
2.16102764e-01 -5.34239292e-01 -5.64870834e-01 -1.89951330e-01
2.22918794e-01 5.69352567e-01 2.84084350e-01 -4.09963906e-01
3.73601884e-01 -4.78952266e-02 -1.22534311e+00 6.57977104e-01
1.03726709e+00 3.63711864e-01 1.36115372e-01 1.09560943e+00
-2.96339691e-01 8.07648897e-01 -1.13747215e+00 -2.77004212e-01
-2.79255629e-01 -7.50440717e-01 -1.09057486e+00 6.73149303e-02
5.37276804e-01 -2.78932899e-01 -3.85517567e-01 3.82599145e-01
9.44985986e-01 3.64011019e-01 6.50658786e-01 -6.83058441e-01
-4.36828554e-01 1.20667195e+00 -3.98316145e-01 2.03001782e-01
6.49159491e-01 -1.81660950e-01 1.18720102e+00 -8.05656612e-01
6.35461450e-01 1.24155211e+00 4.88594353e-01 5.23856103e-01
-1.16544938e+00 -3.13268632e-01 2.09627420e-01 2.00424492e-01
-1.17222583e+00 -9.68205512e-01 6.93319142e-01 -1.09578982e-01
1.48187864e+00 5.44779718e-01 5.78311145e-01 8.62800837e-01
1.14206620e-01 1.09844160e+00 1.05468176e-01 -5.54763913e-01
1.35600314e-01 9.98988077e-02 4.13083762e-01 4.32298332e-01
1.82670981e-01 -8.16465557e-01 -6.75897419e-01 -3.67153198e-01
5.02216257e-02 -2.04557687e-01 -3.34573388e-01 6.83838427e-02
-1.17647505e+00 7.40196526e-01 2.48060167e-01 4.45994258e-01
-5.84338784e-01 1.90907478e-01 7.29784191e-01 2.15597689e-01
7.64466822e-01 9.39404905e-01 -1.02365017e-01 -2.92371958e-02
-1.57079625e+00 4.40725207e-01 1.24805272e+00 8.26716542e-01
5.54462075e-01 -2.47532921e-03 -9.34517801e-01 9.42892849e-01
-5.16958535e-02 -1.83485784e-02 6.56692445e-01 -1.15283179e+00
6.97051406e-01 5.76268494e-01 1.46454901e-01 -6.23934925e-01
-4.26045954e-01 -5.36422074e-01 -7.18904674e-01 -6.55157506e-01
-1.87931269e-01 -3.31357986e-01 -7.31323361e-01 1.50959301e+00
2.84950305e-02 1.68738037e-01 3.72225314e-01 3.63189638e-01
1.24390090e+00 9.68244910e-01 -2.01427087e-01 -5.27466834e-01
1.16222131e+00 -1.40171206e+00 -8.88417780e-01 -1.64168954e-01
9.66358602e-01 -8.65450740e-01 8.75075638e-01 1.62902474e-01
-1.39591599e+00 -5.48089683e-01 -1.02349758e+00 -4.48078126e-01
-5.85528836e-02 3.00382793e-01 9.47871208e-02 9.40333009e-02
-1.57925141e+00 6.04949355e-01 -9.37031507e-01 -7.30046153e-01
-3.25332992e-02 3.13908398e-01 -1.44995853e-01 4.00584675e-02
-1.04290545e+00 8.64701450e-01 4.84708518e-01 -2.22026125e-01
-4.50993747e-01 -5.05118370e-01 -1.18175042e+00 4.16807503e-01
2.84293443e-01 -1.12154222e+00 1.70381522e+00 -3.58348250e-01
-1.51417172e+00 3.68490577e-01 -6.38695657e-01 -5.68445802e-01
1.02264233e-01 -4.48921174e-01 -8.25003535e-02 2.99361467e-01
2.78612286e-01 7.78020501e-01 4.05781418e-01 -1.16980982e+00
-4.82639611e-01 -1.78238884e-01 1.62965711e-02 6.20466828e-01
-4.08007860e-01 -7.07174093e-02 -5.38323998e-01 -4.45909441e-01
-1.99052632e-01 -6.24040723e-01 -3.49208325e-01 -9.90490675e-01
-7.60237277e-01 -5.66742003e-01 5.63952625e-01 -1.14175034e+00
1.91865551e+00 -2.09581232e+00 3.75887215e-01 -6.16962127e-02
1.08005643e-01 2.90328056e-01 -4.02857512e-01 1.17340374e+00
2.28313252e-01 2.56774157e-01 -2.56926745e-01 -1.03036749e+00
-1.68517437e-02 -7.84375817e-02 -3.02852899e-01 -1.29962966e-01
2.27914795e-01 9.90172446e-01 -8.19193065e-01 -7.05776334e-01
-1.32970586e-01 1.89073056e-01 -7.07805037e-01 3.30898225e-01
-2.79722035e-01 9.58586708e-02 -3.28167796e-01 4.96876985e-02
1.15958139e-01 -3.09145395e-02 7.65269175e-02 -1.10596776e-01
-1.41342565e-01 8.79954696e-01 -8.12897384e-01 1.92707741e+00
-5.48116744e-01 7.20784903e-01 3.57000045e-02 -9.20601308e-01
7.17515588e-01 5.49361587e-01 3.93130630e-01 -3.81426692e-01
-3.76285496e-03 -7.69513845e-02 -9.09832492e-02 -4.18732673e-01
1.33753693e+00 2.10983574e-01 -4.23328698e-01 7.03634739e-01
1.10866219e-01 -5.43523610e-01 5.02177894e-01 8.24391067e-01
1.38638914e+00 -1.92255899e-01 3.44088554e-01 -1.18278705e-01
4.28939670e-01 7.14507699e-02 1.17474549e-01 8.04300487e-01
3.39527041e-01 7.65336037e-01 4.79605198e-01 2.11580709e-01
-8.91494215e-01 -8.06972146e-01 4.60475475e-01 1.26170599e+00
-3.41496140e-01 -9.48589027e-01 -9.80568409e-01 -4.85389233e-01
-2.45119661e-01 1.35129082e+00 -2.64216185e-01 -3.39844137e-01
-8.70960534e-01 -4.61266339e-01 5.83762109e-01 5.86994827e-01
4.04627979e-01 -1.04347610e+00 -3.56846541e-01 5.32634437e-01
-6.75968409e-01 -9.67925727e-01 -8.80836070e-01 1.77173987e-01
-7.40091801e-01 -4.72216696e-01 -6.76147759e-01 -8.51148605e-01
3.38001758e-01 2.48869732e-01 1.17961001e+00 -8.29208791e-02
7.97956437e-02 5.43603182e-01 -3.84330750e-01 -6.38854980e-01
-5.70927382e-01 7.93716490e-01 -1.06262140e-01 -3.74834567e-01
1.92299351e-01 -4.93812144e-01 -5.18622994e-01 -3.49097580e-01
-1.11019123e+00 1.13176502e-01 5.80193520e-01 5.19261122e-01
4.63804863e-02 -3.53651941e-01 9.56678033e-01 -9.26338434e-01
1.38586295e+00 -1.81377575e-01 2.77512252e-01 2.89807528e-01
-7.67867789e-02 2.78789043e-01 5.63183844e-01 -1.90542132e-01
-9.25914824e-01 -1.96366310e-01 -5.44067562e-01 1.14661366e-01
3.01933289e-01 7.55686760e-01 3.62684317e-02 7.37612724e-01
6.94192410e-01 2.66239822e-01 -4.95966058e-03 -3.86919916e-01
6.06880784e-01 1.20622265e+00 4.32010174e-01 -1.86016813e-01
4.20167625e-01 1.63904324e-01 -6.89970732e-01 -1.39023101e+00
-1.10864365e+00 -1.05948496e+00 -4.95982170e-01 1.38726383e-01
9.30982232e-01 -9.82994437e-01 -2.05196172e-01 5.14145978e-02
-1.52346969e+00 -2.81972110e-01 -5.75720608e-01 4.05521899e-01
-4.67296064e-01 6.29211843e-01 -7.40834117e-01 -6.38385892e-01
-8.65872383e-01 -7.19489992e-01 1.38326252e+00 1.91192120e-01
-8.35801601e-01 -9.50174809e-01 2.47876257e-01 4.47323114e-01
4.34567362e-01 2.65575908e-02 8.10911119e-01 -1.22659886e+00
-1.83548242e-01 -5.82842678e-02 7.04933405e-02 4.56434250e-01
1.96054280e-01 -2.25300938e-01 -8.44748199e-01 -3.38511586e-01
6.70846738e-03 -2.47337982e-01 1.47431433e+00 5.03491998e-01
8.14421475e-01 -6.56043768e-01 -3.95345718e-01 1.35267064e-01
9.15639460e-01 -1.78912342e-01 5.20004570e-01 1.08522616e-01
5.87273717e-01 5.84172010e-01 3.40032637e-01 3.90017390e-01
7.79610038e-01 4.35295314e-01 -1.36456490e-01 -6.77025169e-02
-1.32316276e-01 -2.94124842e-01 6.14642978e-01 1.69759512e+00
1.72576472e-01 -5.94227135e-01 -6.45416141e-01 7.98256278e-01
-2.09951282e+00 -1.01491559e+00 5.31615354e-02 1.89486313e+00
1.11499524e+00 1.13341443e-01 1.31720811e-01 1.29864542e-02
3.91183555e-01 4.71993238e-01 -2.07395330e-01 -9.01370049e-01
2.19989419e-01 1.76527649e-01 2.47488886e-01 7.68086433e-01
-1.03092718e+00 1.04257107e+00 6.19935179e+00 5.89119017e-01
-8.85245085e-01 6.05905727e-02 1.65308505e-01 -2.24314779e-01
-4.25049514e-01 -1.97290167e-01 -9.08735156e-01 5.57245761e-02
1.23156047e+00 -4.96705800e-01 5.11030555e-02 4.18021530e-01
4.34559703e-01 -2.54692852e-01 -1.47631586e+00 5.44343233e-01
7.47893631e-01 -1.41565955e+00 3.41037303e-01 -1.24588408e-01
7.61030853e-01 -1.36143908e-01 -4.37739521e-01 7.64451683e-01
3.38784546e-01 -8.46571803e-01 2.95362145e-01 5.45901060e-01
4.49037224e-01 -5.88847697e-01 1.07764542e+00 6.05512500e-01
-1.03187740e+00 2.01732904e-01 -3.71239930e-01 -1.15907744e-01
4.32271153e-01 4.45613563e-01 -1.28942454e+00 9.20925736e-01
2.05782250e-01 6.01254165e-01 -6.65014386e-01 9.84867930e-01
-1.81370094e-01 7.04057336e-01 -3.62292886e-01 -1.91779003e-01
3.29130650e-01 7.15357438e-02 7.00330615e-01 1.64298117e+00
2.29366183e-01 -6.11846410e-02 5.13138115e-01 3.51383895e-01
-3.41868728e-01 2.24699825e-01 -6.42594099e-01 -7.52703846e-02
3.93432021e-01 9.74831760e-01 -3.31108689e-01 -7.63196349e-01
-2.09583074e-01 1.24376929e+00 2.78386533e-01 4.56859291e-01
-5.50948620e-01 -7.37293124e-01 1.37347013e-01 2.31321365e-01
3.93337458e-01 -3.07912290e-01 4.71758917e-02 -1.07185006e+00
6.60541654e-02 -8.21803689e-01 3.13722879e-01 -5.37050426e-01
-8.53287339e-01 4.68060136e-01 2.52094239e-01 -6.70528173e-01
-6.40379667e-01 2.69235939e-01 -1.02660072e+00 7.49993086e-01
-1.40079606e+00 -1.06739557e+00 -1.13156639e-01 1.48334369e-01
1.09486592e+00 1.41895369e-01 9.03509676e-01 -2.71103848e-02
-4.74697262e-01 4.81175631e-01 2.25481749e-01 4.66281585e-02
1.03317046e+00 -1.34394932e+00 5.48925698e-01 7.25784481e-01
1.06993355e-01 6.87308669e-01 8.58761072e-01 -6.32954717e-01
-9.28424597e-01 -1.39067900e+00 1.36625946e+00 -4.67213213e-01
4.84907418e-01 -4.45604593e-01 -9.23028052e-01 8.64589334e-01
9.37033892e-01 -1.09363449e+00 8.43391299e-01 3.29651564e-01
2.52136111e-01 -5.60006239e-02 -5.77834427e-01 7.93033957e-01
8.23902071e-01 -4.02765483e-01 -1.37724185e+00 3.32222462e-01
1.38717973e+00 -1.56703934e-01 -6.57624364e-01 1.72169030e-01
2.46739864e-01 -5.49557328e-01 6.69519782e-01 -4.19661045e-01
4.98156369e-01 9.47599765e-03 8.87276679e-02 -1.83380520e+00
-2.81051219e-01 -6.16338134e-01 -1.15589112e-01 1.56754613e+00
6.18744493e-01 -3.55269343e-01 5.91184497e-01 2.37238407e-01
-6.09363198e-01 -6.56726599e-01 -6.77561462e-01 -4.52765495e-01
5.61727502e-04 2.44329795e-02 3.13200861e-01 3.57844263e-01
3.74256223e-01 1.09326541e+00 -6.68486133e-02 -1.29917532e-01
2.02324182e-01 -8.49421695e-02 1.04349983e+00 -1.04521859e+00
-2.50722110e-01 -4.48880523e-01 1.14689833e-02 -1.31930721e+00
1.94842398e-01 -9.00795341e-01 2.53231794e-01 -2.67704535e+00
3.13400477e-01 3.68078411e-01 -6.20288439e-02 2.13619247e-01
-4.05749559e-01 -1.06555626e-01 2.63596386e-01 1.11349963e-01
-9.93849218e-01 7.67446399e-01 9.90346491e-01 -3.44277322e-01
-7.24096775e-01 8.13732371e-02 -1.08050978e+00 5.73305845e-01
8.70373905e-01 -3.14801723e-01 -4.25928235e-01 -5.47589242e-01
-1.70322433e-02 1.76981792e-01 -1.50007188e-01 -1.00827336e+00
6.77984655e-01 3.88416737e-01 9.42329019e-02 -9.69712257e-01
4.72024024e-01 -1.49540812e-01 -3.75353813e-01 1.96121946e-01
-7.53722727e-01 1.37047812e-01 2.23658457e-01 3.59964490e-01
-5.22673488e-01 -3.86602402e-01 2.90502727e-01 -2.10516706e-01
-1.93842024e-01 -3.12297970e-01 -6.35167360e-01 1.63887292e-01
6.58292294e-01 -1.01192854e-01 -2.00831831e-01 -7.91602135e-01
-7.14093029e-01 7.38101184e-01 3.34324121e-01 3.01153034e-01
5.68813205e-01 -8.44955921e-01 -1.13101709e+00 -1.27492875e-01
-3.59909907e-02 2.78091371e-01 1.46792699e-02 7.95964837e-01
-4.48190004e-01 5.72142363e-01 3.87657732e-01 -4.42009330e-01
-1.39862299e+00 2.49121025e-01 -4.70645837e-02 -8.46700132e-01
-4.69778061e-01 5.90631962e-01 7.31909275e-02 -4.84943599e-01
4.44048196e-01 -5.42643130e-01 -5.79051554e-01 4.28098470e-01
5.44061780e-01 4.13408220e-01 1.95221782e-01 -3.03035557e-01
-3.11291695e-01 1.98235720e-01 -3.82814288e-01 -5.02080381e-01
1.41800618e+00 -2.41493881e-01 -2.98698276e-01 7.48187542e-01
1.11540520e+00 3.43081504e-01 -4.63940501e-01 -2.34997556e-01
3.76824439e-01 4.00653690e-01 -8.41024965e-02 -5.80956221e-01
-2.19476938e-01 7.26947308e-01 -2.88007349e-01 3.31553727e-01
9.20716226e-01 1.25530511e-01 1.01045132e+00 8.16956878e-01
-1.62074566e-01 -1.26364267e+00 3.56507957e-01 7.38322139e-01
1.08897698e+00 -9.54504728e-01 3.19396198e-01 -1.55812323e-01
-7.44750857e-01 9.26510096e-01 3.81150514e-01 -1.51050135e-01
2.84556180e-01 1.11535035e-01 -3.78292217e-03 -1.48053080e-01
-1.04759717e+00 -1.97147191e-01 3.42819095e-01 1.45492524e-01
7.26693094e-01 -6.15261495e-03 -5.52484989e-01 5.99218845e-01
-6.25648320e-01 -1.66826338e-01 8.55657876e-01 9.87970769e-01
-8.88906479e-01 -1.01892030e+00 -8.61118808e-02 7.82366335e-01
-5.64486742e-01 -3.39297384e-01 -8.42708707e-01 6.83726370e-01
-3.75557512e-01 1.37241840e+00 1.40379697e-01 -2.71339267e-01
6.76833570e-01 4.39934671e-01 9.90947187e-02 -1.38100445e+00
-9.22604382e-01 1.47441387e-01 6.50960922e-01 -2.73559421e-01
-3.12274188e-01 -5.64633489e-01 -1.16806090e+00 -6.10723831e-02
-2.80752629e-01 6.86923325e-01 5.26179373e-01 8.58879626e-01
6.21287048e-01 1.01970673e+00 4.59779739e-01 -1.01136172e+00
-7.71283269e-01 -1.56528127e+00 -1.61973700e-01 8.84305760e-02
4.14318621e-01 1.59634531e-01 -1.76086918e-01 6.60397559e-02] | [12.523730278015137, 9.476539611816406] |
a672cfe2-2ca7-48c4-a4dd-6fed6889b716 | enhancing-modality-agnostic-representations | 2302.04308 | null | https://arxiv.org/abs/2302.04308v1 | https://arxiv.org/pdf/2302.04308v1.pdf | Enhancing Modality-Agnostic Representations via Meta-Learning for Brain Tumor Segmentation | In the medical vision domain, different imaging modalities provide complementary information. However, in practice, not all modalities may be available during inference. Previous approaches, e.g., knowledge distillation or image synthesis, often assume the availability of full modalities for all patients during training; this is unrealistic and impractical owing to the variability in data collection across sites. We propose a novel approach to learn enhanced modality-agnostic representations by employing a novel meta-learning strategy in training, even when only a fraction of full modality patients are available. Meta-learning enhances partial modality representations to full modality representations by meta-training on partial modality data and meta-testing on limited full modality samples. Additionally, we co-supervise this feature enrichment by introducing an auxiliary adversarial learning branch. More specifically, a missing modality detector is used as a discriminator to mimic the full modality setting. Our segmentation framework significantly outperforms state-of-the-art brain tumor segmentation techniques in missing modality scenarios, as demonstrated on two brain tumor MRI datasets. | ['Prateek Prasanna', 'Chao Chen', 'Joseph Bae', 'Xuan Xu', 'Xiaoling Hu', 'Aishik Konwer'] | 2023-02-08 | null | null | null | null | ['tumor-segmentation', 'brain-tumor-segmentation'] | ['computer-vision', 'medical'] | [ 8.61536562e-01 4.46320653e-01 -5.53765118e-01 -3.12919021e-01
-1.24937904e+00 -5.88656604e-01 7.78385758e-01 -6.67629614e-02
-5.76642931e-01 1.02668655e+00 2.44027808e-01 -4.34694827e-01
-3.20526212e-02 -6.67467654e-01 -1.02324247e+00 -8.08055758e-01
2.66747713e-01 5.49522161e-01 -2.64250517e-01 7.61676803e-02
-2.52234489e-01 3.88710976e-01 -1.08179152e+00 4.78109181e-01
1.12588823e+00 7.47024119e-01 2.50899822e-01 6.92921221e-01
-1.17630005e-01 6.85510397e-01 -4.09684330e-01 -3.11884254e-01
2.86544114e-01 -4.80943322e-01 -8.27864349e-01 3.89662117e-01
4.86863643e-01 -5.56279302e-01 -7.10507751e-01 9.45412815e-01
5.10711133e-01 3.93848643e-02 8.24945927e-01 -1.33399916e+00
-4.85664248e-01 7.80566156e-01 -6.87368035e-01 1.86473221e-01
2.53523737e-01 3.15473616e-01 6.82530820e-01 -5.78734696e-01
7.40162849e-01 6.17183089e-01 4.46576834e-01 8.99369240e-01
-1.38482666e+00 -5.99847496e-01 1.53508395e-01 5.15198633e-02
-1.02048910e+00 -3.55972081e-01 8.49653244e-01 -4.09460902e-01
5.64485192e-01 9.20340866e-02 5.48872709e-01 1.56632662e+00
8.58197957e-02 1.11371863e+00 1.42849743e+00 -4.20108080e-01
1.41751796e-01 3.21477726e-02 -1.25924572e-01 7.24530220e-01
7.16058388e-02 2.28762344e-01 -4.72792685e-01 -3.20668787e-01
8.83128285e-01 2.88920581e-01 -3.37863743e-01 -4.82099384e-01
-1.68536603e+00 7.09052920e-01 5.65255284e-01 1.63185835e-01
-6.95937276e-01 1.24837607e-02 4.54662114e-01 1.55008689e-01
2.44297668e-01 3.15243632e-01 -2.19559178e-01 3.57678592e-01
-1.26745784e+00 6.30339012e-02 6.12926781e-01 8.04869115e-01
5.20492315e-01 -3.89990471e-02 -3.05093706e-01 7.40161240e-01
-3.03911604e-02 5.09144187e-01 7.17069924e-01 -9.76925492e-01
5.20374119e-01 3.41260672e-01 -2.86647469e-01 -4.77263033e-02
-4.84853625e-01 -6.20518684e-01 -1.06994951e+00 1.08598538e-01
5.34673691e-01 -3.18422467e-01 -1.59260690e+00 2.06966329e+00
3.16685557e-01 7.21519113e-01 1.32936522e-01 7.89040804e-01
9.39538658e-01 1.35927483e-01 3.07666421e-01 -1.24874130e-01
1.38697886e+00 -1.10696757e+00 -5.99399686e-01 -2.47118965e-01
4.92935508e-01 -3.55050951e-01 8.41645956e-01 2.01615691e-01
-1.15198851e+00 -2.40924418e-01 -8.63153636e-01 2.62449354e-01
-4.05679822e-01 -1.38473660e-01 8.49979639e-01 6.65216804e-01
-8.17291677e-01 2.19489187e-01 -1.07079065e+00 1.13950698e-02
9.14694428e-01 3.86577249e-01 -8.22086811e-01 -5.38581073e-01
-1.07558775e+00 8.34967852e-01 4.74868298e-01 -7.37093538e-02
-1.34186780e+00 -1.27123547e+00 -9.57880855e-01 -2.37908557e-01
5.81700623e-01 -1.32727981e+00 1.25189030e+00 -1.08638787e+00
-1.26074636e+00 9.52750981e-01 -1.02315098e-02 -5.47159195e-01
8.61794949e-01 2.62114495e-01 -4.09896851e-01 5.90879858e-01
1.86734961e-03 9.34300244e-01 1.08296609e+00 -1.36181390e+00
-5.01896679e-01 -3.04296941e-01 2.72506773e-01 4.13559318e-01
5.51910959e-02 -5.26812375e-01 -4.11050737e-01 -7.77548373e-01
-2.68705897e-02 -9.74590480e-01 -4.24877107e-01 1.76143229e-01
-7.25286901e-01 5.38984597e-01 4.80660111e-01 -7.70064175e-01
4.61625218e-01 -1.93037450e+00 4.36759979e-01 4.16573554e-01
5.55704534e-01 -4.54899520e-02 -2.63861954e-01 -1.08729877e-01
-2.08152056e-01 -6.80865049e-02 -7.19075322e-01 -5.70436716e-01
-2.11955681e-01 6.19695306e-01 4.89824871e-03 4.60802317e-01
2.37409487e-01 1.06817257e+00 -1.03163719e+00 -7.11827397e-01
5.22506773e-01 6.19581521e-01 -6.23043954e-01 2.07046755e-02
-8.60456228e-02 1.25616264e+00 -3.96297276e-01 1.12588954e+00
7.13608623e-01 -4.64313179e-01 1.24918245e-01 -2.05254167e-01
5.41586101e-01 -2.24066973e-01 -7.59994864e-01 2.40079665e+00
-7.87106574e-01 2.00527295e-01 1.21380091e-01 -1.31892323e+00
1.03819236e-01 5.34797728e-01 8.43755603e-01 -6.21696055e-01
2.08447993e-01 2.97316283e-01 2.17942387e-01 -3.62517893e-01
2.22906515e-01 -5.90359151e-01 1.64203346e-02 3.55999798e-01
4.17606682e-01 -1.53805181e-01 -7.70953149e-02 1.81516781e-01
1.21868324e+00 2.34217141e-02 1.57152519e-01 3.47375333e-01
3.73828322e-01 4.91817780e-02 4.36330318e-01 1.06663418e+00
-3.64938915e-01 8.81510198e-01 3.80562425e-01 2.04484433e-01
-9.60505545e-01 -1.39044511e+00 -3.45151961e-01 7.51776636e-01
-4.21754904e-02 3.48050356e-01 -6.53819919e-01 -1.14774454e+00
3.55686210e-02 6.84888780e-01 -1.05288947e+00 -3.11833948e-01
-4.32775497e-01 -8.65459263e-01 6.45042360e-01 7.10966706e-01
3.37418646e-01 -8.62574756e-01 -5.42775273e-01 1.73386291e-01
-1.80482760e-01 -1.37720227e+00 -2.74554640e-01 2.55806953e-01
-1.14566433e+00 -1.14047897e+00 -1.33236313e+00 -5.29381633e-01
9.44618702e-01 -2.77807179e-04 1.00935352e+00 -1.97827816e-01
-4.21821415e-01 6.42447233e-01 -2.33115628e-01 -1.45145208e-01
-6.13814592e-01 1.98463812e-01 -2.67230242e-01 -7.71914143e-03
-1.67531043e-01 -5.51549017e-01 -7.80875266e-01 -2.12157175e-01
-1.11067665e+00 3.14009815e-01 1.01644599e+00 1.51509500e+00
9.83326316e-01 -2.55513310e-01 5.73207855e-01 -1.39801943e+00
2.38245517e-01 -7.30971754e-01 -9.16851908e-02 4.53346521e-01
-2.21130237e-01 6.38109297e-02 4.22612101e-01 -6.46261156e-01
-1.14767480e+00 1.75911576e-01 -6.24033995e-02 -8.01828623e-01
-3.70259225e-01 6.98502898e-01 -8.06894600e-02 -1.62261412e-01
5.43275535e-01 3.94421279e-01 2.70435721e-01 -1.58115670e-01
5.81827343e-01 3.06611776e-01 8.99395287e-01 -5.89575827e-01
6.99334204e-01 7.22898304e-01 2.73641407e-01 -4.64723021e-01
-7.16955364e-01 -2.89038658e-01 -6.39947712e-01 -4.37913612e-02
6.99130893e-01 -9.24307346e-01 -2.46448845e-01 4.90570366e-01
-6.21707201e-01 -4.40357655e-01 -5.78306437e-01 7.38317251e-01
-8.51598203e-01 1.40548483e-01 -3.14660519e-01 -2.26436004e-01
-3.30062717e-01 -1.57394457e+00 1.11292338e+00 2.25483701e-01
2.34203473e-01 -1.26568985e+00 -1.28053576e-01 5.37959874e-01
3.42826068e-01 6.86257720e-01 9.11523640e-01 -7.57673681e-01
-3.86457324e-01 -3.27474773e-01 -1.34696543e-01 2.84275562e-01
3.67780864e-01 -4.87522721e-01 -1.02655602e+00 -2.96581924e-01
-4.32923824e-01 -4.62247759e-01 1.12977779e+00 6.30410433e-01
1.41380572e+00 1.90011069e-01 -5.81253827e-01 8.03789675e-01
1.26732194e+00 -1.37931019e-01 3.51507545e-01 1.51163042e-01
7.60042965e-01 3.03885490e-01 2.35444993e-01 3.35376233e-01
3.33773255e-01 3.27818632e-01 5.37516952e-01 -3.54062587e-01
-4.63223845e-01 -2.20234782e-01 -1.76952496e-01 1.72352374e-01
8.78781304e-02 -1.02049679e-01 -9.04722929e-01 6.30208910e-01
-1.48093724e+00 -8.42895567e-01 6.54570043e-01 2.17531538e+00
9.39279616e-01 2.46489886e-02 1.05533786e-01 9.90219191e-02
5.78215718e-01 6.89876499e-03 -9.76627350e-01 1.62997007e-01
-1.51329428e-01 3.41343999e-01 7.09623039e-01 2.41864577e-01
-1.15031850e+00 5.09962440e-01 6.49767351e+00 5.10914207e-01
-1.17824769e+00 5.68626046e-01 5.18211126e-01 -2.45254889e-01
-4.81433809e-01 -2.51972407e-01 -1.58220127e-01 3.64904135e-01
8.49955916e-01 6.51196614e-02 3.66992265e-01 5.18318176e-01
-2.77127951e-01 -1.44460946e-01 -1.19052732e+00 9.34594035e-01
2.62733072e-01 -1.40310240e+00 7.14459270e-02 1.40646636e-01
8.80129755e-01 1.31261259e-01 2.94942051e-01 4.43516105e-01
1.23677269e-01 -1.07193506e+00 3.46541286e-01 7.59216249e-01
1.06699538e+00 -7.16267169e-01 6.36541963e-01 3.37677747e-01
-6.58423245e-01 -3.40667516e-02 1.46754727e-01 7.25171685e-01
3.36220920e-01 4.39165413e-01 -9.71744359e-01 1.00565922e+00
2.01824054e-01 5.41199625e-01 -4.60792214e-01 1.15135431e+00
-8.11783820e-02 5.54236233e-01 -3.91437858e-01 7.66144991e-01
1.46825001e-01 3.13534588e-01 5.20326078e-01 9.20632184e-01
1.71134859e-01 -6.62102327e-02 2.59292394e-01 7.28166521e-01
-3.27293128e-01 -3.05623204e-01 -5.64662755e-01 -1.31667554e-01
4.12825316e-01 1.30117774e+00 -5.13895631e-01 -4.31269020e-01
-8.45600307e-01 1.05314708e+00 4.32608724e-02 3.81943852e-01
-8.86332214e-01 5.30038821e-03 2.19738171e-01 -1.32451460e-01
2.39238907e-02 1.46118075e-01 -4.00443465e-01 -1.27942824e+00
-3.40469211e-01 -8.93589199e-01 7.38852680e-01 -6.52490735e-01
-1.48666632e+00 4.68468666e-01 2.13540077e-01 -1.30271447e+00
-5.69754481e-01 -5.64358532e-01 -4.26949590e-01 1.07423615e+00
-1.89474976e+00 -1.78474319e+00 -3.42175901e-01 1.00007892e+00
2.44141072e-01 -3.31792206e-01 8.63767922e-01 2.84899712e-01
-4.55062211e-01 1.09278047e+00 1.11265816e-02 1.48078993e-01
6.81011856e-01 -1.25971019e+00 -1.60327435e-01 6.05312109e-01
-9.28327218e-02 4.79401141e-01 4.65545505e-01 -5.15738904e-01
-1.47661531e+00 -8.73269200e-01 5.08635230e-02 -2.40951478e-01
7.20253348e-01 3.70277651e-03 -9.12885368e-01 8.95291090e-01
1.16563648e-01 6.14323139e-01 1.04541659e+00 -5.26595414e-02
-2.34011695e-01 2.06615731e-01 -1.54663384e+00 5.49999416e-01
8.22866678e-01 -6.92917526e-01 -7.88885891e-01 4.58256096e-01
6.03006780e-01 -8.45771968e-01 -1.00320792e+00 6.28182769e-01
5.71004868e-01 -3.36974919e-01 1.09355712e+00 -7.11800337e-01
5.53609848e-01 -4.93595749e-02 -2.71589309e-01 -1.42811120e+00
2.16557741e-01 -1.71402305e-01 -4.70334142e-01 7.73240328e-01
4.90304440e-01 -7.95603871e-01 1.01317418e+00 5.73806643e-01
-2.44525179e-01 -7.69288719e-01 -1.20418060e+00 -6.11561894e-01
3.80272985e-01 -3.65966946e-01 5.88394284e-01 1.12850237e+00
-2.51657993e-01 -2.73306936e-01 -2.83277899e-01 3.60295475e-01
7.95962393e-01 1.33792117e-01 3.82600486e-01 -7.86839306e-01
-8.15457106e-01 -4.16406393e-01 -2.37089664e-01 -4.82778251e-01
5.31528354e-01 -1.26799393e+00 -1.67359948e-01 -1.59201360e+00
5.00574648e-01 -4.39085960e-01 -7.85101116e-01 6.80172920e-01
-2.40040466e-01 3.47369611e-01 1.35602970e-02 1.45181507e-01
-3.88945900e-02 3.29531312e-01 1.86023533e+00 -5.07469416e-01
-3.11549231e-02 7.24912733e-02 -8.00508678e-01 5.24934411e-01
5.78303277e-01 -3.33290994e-01 -6.38391495e-01 -3.28095675e-01
-4.09354508e-01 5.39194345e-01 7.19416499e-01 -8.56836140e-01
2.22145706e-01 -1.62790969e-01 6.77163184e-01 -4.37545925e-01
5.70944905e-01 -9.16090548e-01 6.41249418e-02 3.94859970e-01
-4.71934855e-01 -4.69740868e-01 2.94749677e-01 6.60125017e-01
-3.37667853e-01 -2.71488428e-01 6.81146979e-01 -3.38091463e-01
-6.91066563e-01 6.21252596e-01 -1.12077892e-02 1.01622500e-01
1.00855696e+00 -2.23211125e-01 -3.44759077e-01 -1.65291056e-01
-1.20658243e+00 2.39353597e-01 2.81162918e-01 1.65538296e-01
6.36782825e-01 -1.19200528e+00 -7.47849524e-01 2.80384421e-01
2.39132971e-01 1.66992649e-01 8.22439730e-01 1.29350269e+00
-1.77043229e-01 9.42975748e-03 -3.49878609e-01 -7.63259053e-01
-9.18877065e-01 5.48404753e-01 5.72378814e-01 -4.14344430e-01
-7.55001187e-01 5.65491855e-01 3.04782122e-01 -6.21427357e-01
8.22532400e-02 -4.21622097e-02 1.52421114e-03 -5.07133678e-02
4.70786303e-01 -4.48223427e-02 1.01881735e-01 -5.54307282e-01
-3.70382845e-01 2.24609941e-01 -4.72258270e-01 -4.56062287e-01
1.05956757e+00 1.16885073e-01 3.56012315e-01 2.08275512e-01
9.52987611e-01 -1.95109859e-01 -1.37160456e+00 -5.95769823e-01
-6.35769904e-01 -3.30996007e-01 4.15443003e-01 -1.30864596e+00
-1.37927067e+00 8.68545890e-01 7.13941216e-01 -3.99044991e-01
1.44377983e+00 1.15422435e-01 6.69712186e-01 2.43500531e-01
3.18808287e-01 -7.99572170e-01 -1.36362672e-01 7.69670755e-02
6.96254134e-01 -1.61179721e+00 -5.17026000e-02 -3.22369814e-01
-7.66148984e-01 7.99714386e-01 4.48495239e-01 4.00198102e-02
6.66132689e-01 2.91169673e-01 -2.33805832e-02 3.52675579e-02
-4.98786658e-01 -2.01572403e-01 3.56991857e-01 7.87569225e-01
1.60030276e-01 3.78991544e-01 7.68320039e-02 5.28287828e-01
1.44381061e-01 1.72969118e-01 4.64814484e-01 1.24568653e+00
1.82892993e-01 -1.17560279e+00 -3.23905975e-01 7.53998935e-01
-5.45479000e-01 -1.32831573e-01 7.75555223e-02 1.01784074e+00
1.12536214e-01 4.94758785e-01 -2.81196218e-02 -1.52695835e-01
1.83435038e-01 1.67575940e-01 9.23636615e-01 -6.35180473e-01
-3.38217765e-01 -1.69608854e-02 -3.14496756e-01 -3.27763021e-01
-5.76944053e-01 -8.36048365e-01 -1.25974262e+00 -9.85930637e-02
-1.69655219e-01 -3.03021938e-01 7.10211694e-01 1.12810826e+00
1.12178765e-01 1.15542221e+00 5.17984927e-01 -8.87938440e-01
-5.86209297e-01 -7.99152613e-01 -4.75319594e-01 3.70094597e-01
6.19384050e-01 -9.76654172e-01 -1.32091254e-01 1.49087999e-02] | [14.547357559204102, -2.139209508895874] |
4a488659-75dc-4bfb-b5b9-d3d409d00826 | an-ensemble-teacher-student-learning-approach | 2210.06382 | null | https://arxiv.org/abs/2210.06382v1 | https://arxiv.org/pdf/2210.06382v1.pdf | An Ensemble Teacher-Student Learning Approach with Poisson Sub-sampling to Differential Privacy Preserving Speech Recognition | We propose an ensemble learning framework with Poisson sub-sampling to effectively train a collection of teacher models to issue some differential privacy (DP) guarantee for training data. Through boosting under DP, a student model derived from the training data suffers little model degradation from the models trained with no privacy protection. Our proposed solution leverages upon two mechanisms, namely: (i) a privacy budget amplification via Poisson sub-sampling to train a target prediction model that requires less noise to achieve a same level of privacy budget, and (ii) a combination of the sub-sampling technique and an ensemble teacher-student learning framework that introduces DP-preserving noise at the output of the teacher models and transfers DP-preserving properties via noisy labels. Privacy-preserving student models are then trained with the noisy labels to learn the knowledge with DP-protection from the teacher model ensemble. Experimental evidences on spoken command recognition and continuous speech recognition of Mandarin speech show that our proposed framework greatly outperforms existing DP-preserving algorithms in both speech processing tasks. | ['Chin-Hui Lee', 'Sabato Marco Siniscalchi', 'Jun Qi', 'Chao-Han Huck Yang'] | 2022-10-12 | null | null | null | null | ['spoken-command-recognition'] | ['speech'] | [ 5.12550831e-01 5.14043510e-01 -1.17070787e-01 -9.03683424e-01
-1.27135837e+00 -7.85743892e-01 5.00243425e-01 3.25549021e-02
-3.93643081e-01 7.20990717e-01 -2.64212079e-02 -4.77533191e-01
1.98845580e-01 -6.23336434e-01 -9.84853804e-01 -1.24607360e+00
1.20389961e-01 7.48958960e-02 -3.02014239e-02 3.98010373e-01
-1.12000421e-01 3.55155170e-01 -1.50392568e+00 4.41447824e-01
1.07540894e+00 1.15863991e+00 -4.40346509e-01 5.32772243e-01
2.14815825e-01 7.51585722e-01 -7.42065728e-01 -5.77012539e-01
7.53986895e-01 -3.25130910e-01 -5.18997669e-01 -5.22092767e-02
6.46610260e-01 -7.09501207e-01 -4.13957864e-01 1.27658951e+00
5.54033756e-01 2.24075183e-01 3.68013889e-01 -1.39477766e+00
-5.12256622e-01 5.91079533e-01 -5.49482465e-01 -2.60978162e-01
-1.31214023e-01 8.19820911e-02 6.13819718e-01 -5.82655013e-01
4.14912738e-02 1.13314235e+00 4.27657634e-01 9.25962090e-01
-1.40960038e+00 -1.25659180e+00 3.35970521e-02 -1.91903174e-01
-1.20526457e+00 -8.25028658e-01 5.13792574e-01 4.45085578e-02
4.84332532e-01 7.64232755e-01 2.15829328e-01 1.19130945e+00
7.97817782e-02 1.08798873e+00 1.77080703e+00 -3.24246973e-01
5.13271213e-01 9.65597987e-01 3.78456742e-01 4.18602169e-01
-1.55801266e-01 3.80175918e-01 -9.27951932e-01 -7.35010684e-01
1.32314339e-01 2.43366463e-03 -5.02642035e-01 -4.17622298e-01
-3.45625937e-01 5.89056253e-01 -6.97144866e-02 -3.18598568e-01
-4.91033457e-02 -1.82393119e-01 5.96730113e-01 7.57868648e-01
9.03667629e-01 -1.64792106e-01 -7.48994291e-01 3.49338919e-01
-8.81776869e-01 2.62118399e-01 1.07508659e+00 1.19213164e+00
7.96221137e-01 -1.47038139e-03 -2.45313153e-01 6.41290367e-01
3.82196277e-01 6.18367374e-01 3.59475672e-01 -7.47735739e-01
5.17641008e-01 1.66438282e-01 -4.28294167e-02 -5.63641667e-01
6.72211170e-01 -5.70637345e-01 -6.46579325e-01 3.81268144e-01
2.04284400e-01 -4.33330357e-01 -6.94006085e-01 2.05193710e+00
8.82408679e-01 4.96623486e-01 3.36863428e-01 5.92170119e-01
4.61967796e-01 6.31808698e-01 6.22275881e-02 -3.81257594e-01
1.37094164e+00 -7.82329321e-01 -6.31431401e-01 1.23899557e-01
6.72033310e-01 -3.38573515e-01 6.86527073e-01 4.40935791e-01
-1.08886147e+00 -1.05544418e-01 -1.07338738e+00 -1.21765900e-02
-6.50031567e-02 -1.36417300e-01 2.41335645e-01 1.31325316e+00
-1.14331543e+00 3.10412973e-01 -7.87187576e-01 2.33912513e-01
9.14399862e-01 7.26446271e-01 -5.23272455e-01 -7.54063763e-03
-7.97865689e-01 3.17156255e-01 -8.16780552e-02 -3.25287700e-01
-1.38283014e+00 -1.26007521e+00 -7.44024277e-01 1.61098003e-01
2.17485085e-01 -6.08306944e-01 1.31281602e+00 -1.26022792e+00
-1.96771753e+00 9.12438929e-01 -1.46010607e-01 -7.66941011e-01
1.06980860e+00 -1.74393147e-01 2.87961066e-02 -1.77793741e-01
-2.77558774e-01 3.21455449e-01 1.41737986e+00 -1.40787017e+00
-1.04973149e+00 -9.72759843e-01 -4.40799087e-01 2.81700492e-01
-6.45946324e-01 -6.30865525e-03 1.95654392e-01 -5.94393790e-01
-2.52613015e-02 -8.19604754e-01 -7.12286085e-02 1.06488399e-01
-4.16803211e-01 -1.81390464e-01 1.60196662e+00 -7.50466645e-01
8.01731408e-01 -2.32973933e+00 -4.00082529e-01 4.80935574e-01
3.25979739e-01 4.30517286e-01 2.02353019e-02 1.86609589e-02
6.84814602e-02 -1.33720897e-02 -3.56701791e-01 -1.01366782e+00
-4.39040400e-02 2.95555770e-01 -1.06510878e+00 6.86253011e-01
4.55901809e-02 4.22740996e-01 -6.84707582e-01 -1.68453917e-01
-8.26089382e-02 5.54814816e-01 -6.18109167e-01 7.15537608e-01
-1.23468593e-01 5.60520053e-01 -5.71709871e-01 3.67581755e-01
1.34359276e+00 5.02530932e-01 1.92088351e-01 3.73658270e-01
1.86558843e-01 5.05481303e-01 -1.01672459e+00 1.26748562e+00
-2.16883913e-01 1.35814875e-01 8.82023573e-01 -1.05493271e+00
9.79543746e-01 5.95374048e-01 3.24853435e-02 -1.89795390e-01
1.76261589e-01 -2.46743392e-03 -4.41246003e-01 -2.76953667e-01
1.45312279e-01 6.02369681e-02 9.14790034e-02 1.00440800e+00
2.20451280e-01 1.43206239e-01 -1.07417727e+00 4.69340160e-02
9.94758189e-01 -2.09328711e-01 -1.93787560e-01 -4.83758330e-01
6.53401434e-01 -7.36472607e-01 9.84799981e-01 1.01558793e+00
-5.83005428e-01 -5.41338418e-03 1.55716211e-01 -2.91075379e-01
-8.22740912e-01 -1.05442715e+00 -1.57972068e-01 1.38905704e+00
-3.59366119e-01 -8.62799361e-02 -1.16700780e+00 -1.25014699e+00
1.20035678e-01 7.85679281e-01 -5.35781503e-01 -1.46378174e-01
-2.07408041e-01 -2.44418085e-01 8.71846616e-01 4.36837152e-02
6.73480809e-01 -4.24014390e-01 -2.63245165e-01 -1.32546887e-01
4.23487991e-01 -7.79722452e-01 -8.08846235e-01 4.04646993e-01
-9.51732159e-01 -6.25550687e-01 -2.69632638e-01 -8.97970259e-01
8.83022308e-01 2.45610520e-01 4.12777156e-01 -1.45265058e-01
2.46368751e-01 4.77983743e-01 -1.90618879e-03 -8.60797763e-01
-8.18554044e-01 -1.77043706e-01 3.21738929e-01 5.32763183e-01
6.02204919e-01 -9.38903630e-01 -4.50951338e-01 8.12188536e-02
-8.92621636e-01 -1.72989383e-01 1.78839773e-01 1.00555432e+00
4.55393612e-01 1.73486490e-02 5.42291224e-01 -1.06192696e+00
4.61265832e-01 -3.96359742e-01 -7.94080853e-01 4.73217279e-01
-7.72860944e-01 2.97215418e-03 6.94094062e-01 -6.25285804e-01
-1.37652540e+00 3.22272241e-01 9.65814069e-02 -7.34595060e-01
-3.21039796e-01 -4.54480469e-01 -6.04725003e-01 -4.09421951e-01
4.01024580e-01 7.32186317e-01 4.03370976e-01 -4.60594177e-01
4.68180090e-01 1.20005083e+00 4.26820666e-01 -9.64384317e-01
7.22408295e-01 5.01152217e-01 -2.32921585e-01 -7.84552753e-01
-4.77709472e-01 -6.82624476e-03 -2.67160535e-01 2.61704493e-02
5.43670297e-01 -1.23304832e+00 -8.66284430e-01 8.08764994e-01
-9.58064616e-01 3.58785652e-02 -3.71016562e-01 3.86602014e-01
-4.00659531e-01 3.66541594e-01 -4.97537196e-01 -1.52756560e+00
-8.10299754e-01 -1.14269829e+00 8.65813434e-01 2.91413128e-01
2.83471465e-01 -5.94773293e-01 -2.07099691e-01 7.42989004e-01
2.33380005e-01 -5.35678677e-02 7.78736055e-01 -1.32992303e+00
-7.49392509e-01 -3.21868032e-01 3.25870067e-01 9.57473695e-01
1.78407133e-02 -2.97702223e-01 -1.88126075e+00 -7.37890363e-01
8.82379234e-01 -7.65555918e-01 6.71289325e-01 1.59317136e-01
1.50352085e+00 -1.09225774e+00 -1.98791951e-01 8.54167044e-01
9.30598974e-01 1.51474416e-01 1.92834154e-01 -2.22062260e-01
4.80506182e-01 9.43980575e-01 1.53385282e-01 5.41347504e-01
3.39854509e-01 1.24012537e-01 2.29158074e-01 1.63268685e-01
5.40046871e-01 -6.61394417e-01 5.58323741e-01 8.93987179e-01
7.28681564e-01 1.32872388e-01 -4.11164522e-01 4.12896663e-01
-1.70913506e+00 -7.46562183e-01 3.17983717e-01 2.55046797e+00
1.33686411e+00 -4.30076271e-01 5.55742718e-03 -2.87447032e-02
5.34130096e-01 6.20414875e-02 -7.78583348e-01 -6.20331407e-01
-2.33803187e-02 3.70053530e-01 6.06028497e-01 4.60791916e-01
-1.19901705e+00 7.39786685e-01 5.31091261e+00 1.12117767e+00
-1.05216587e+00 4.53286976e-01 1.23400414e+00 -2.19955638e-01
-4.62880373e-01 1.35444596e-01 -6.98154271e-01 4.55930203e-01
9.01606679e-01 -4.58979666e-01 3.63476694e-01 1.16160262e+00
2.99727209e-02 3.67982298e-01 -1.26834035e+00 6.02346778e-01
-1.06487781e-01 -9.34510350e-01 -1.55459732e-01 3.74414355e-01
7.84630537e-01 -8.33335817e-02 5.04606485e-01 2.57920533e-01
9.20337081e-01 -8.12448800e-01 5.62253356e-01 1.94943026e-01
4.73844707e-01 -9.81958985e-01 2.91189939e-01 8.95208478e-01
-5.07500768e-01 -1.77767470e-01 -4.59068865e-01 2.09380791e-01
-6.09748483e-01 6.07752562e-01 -1.10020494e+00 4.98290718e-01
8.53644431e-01 1.44997358e-01 4.58300486e-02 4.10041928e-01
-3.71261746e-01 1.16471303e+00 -5.77428222e-01 1.17152512e-01
-1.10863056e-02 -4.81743813e-01 7.02670097e-01 9.45229352e-01
7.01353624e-02 2.32364953e-01 5.81719913e-02 8.76879156e-01
-5.35573483e-01 1.31046683e-01 -8.06947172e-01 3.81572068e-01
1.21359587e+00 1.04813004e+00 2.40408167e-01 -1.50658250e-01
-3.31391126e-01 9.29394662e-01 4.30096716e-01 2.74697453e-01
-4.03896511e-01 5.57319541e-03 1.17443073e+00 -3.70188475e-01
1.52591974e-01 2.62874126e-01 -4.47483838e-01 -9.29194391e-01
1.58275217e-01 -1.32105017e+00 4.82760876e-01 -1.41515434e-01
-1.54660201e+00 2.90890932e-01 -4.39569175e-01 -8.38994086e-01
4.79717106e-02 4.72300053e-02 -6.51966333e-01 1.27740145e+00
-1.25735164e+00 -1.29638970e+00 4.39796895e-01 7.86595523e-01
2.10647017e-01 -3.96774709e-01 1.11617732e+00 -7.00642616e-02
-7.07491100e-01 1.51162767e+00 4.81223643e-01 1.96817145e-01
6.64373040e-01 -1.07458413e+00 1.18915796e-01 8.46683443e-01
1.06755190e-01 6.71970308e-01 4.19604003e-01 -4.53302622e-01
-1.47631896e+00 -1.39655018e+00 8.10474396e-01 -5.33425450e-01
1.88532487e-01 -8.27647984e-01 -1.22072899e+00 8.76805186e-01
3.07144046e-01 -3.41582170e-04 1.19329870e+00 -2.97053963e-01
-8.67183268e-01 -4.59370732e-01 -1.92866135e+00 2.79179573e-01
4.51003581e-01 -9.66248453e-01 -3.74051929e-01 2.33651146e-01
1.10624933e+00 -4.14638430e-01 -8.96202743e-01 1.17874578e-01
5.59518337e-01 -8.74494255e-01 5.79364121e-01 -1.03978610e+00
-9.61909518e-02 -5.62228858e-02 -2.50202268e-01 -1.14686954e+00
3.19593996e-01 -1.16996789e+00 -1.35688230e-01 1.75533497e+00
2.43084535e-01 -9.40603197e-01 1.12440622e+00 1.29887748e+00
2.21163452e-01 -6.47834659e-01 -1.51987636e+00 -5.99828243e-01
4.69679505e-01 -3.78253549e-01 1.03580999e+00 1.30299807e+00
-3.72854829e-01 -8.59424993e-02 -5.30211270e-01 6.99606359e-01
1.26300442e+00 -2.90257692e-01 1.03061068e+00 -7.18744874e-01
-2.57095039e-01 2.09392831e-01 -1.48915961e-01 -1.10098159e+00
7.60867178e-01 -8.75998378e-01 5.76660372e-02 -1.60201818e-01
1.06982470e-01 -6.58349991e-01 -3.67060930e-01 5.98729312e-01
-1.90082163e-01 -3.88248622e-01 2.40374990e-02 1.05376646e-01
-2.04071403e-01 7.98029542e-01 8.32592309e-01 1.32811097e-02
-1.80197671e-01 4.99294460e-01 -7.76205719e-01 3.72947156e-01
6.05446517e-01 -9.94526923e-01 -6.30264103e-01 -3.35511237e-01
-6.27699554e-01 3.76977250e-02 2.81843215e-01 -5.34712911e-01
4.96649683e-01 1.62133239e-02 2.35563397e-01 -2.71470726e-01
6.41997233e-02 -1.26843894e+00 -1.01596855e-01 4.77587610e-01
-8.00248563e-01 -5.50422728e-01 1.51468730e-02 8.39570343e-01
1.97322533e-01 9.46063697e-02 1.02248991e+00 3.21343005e-01
1.75078064e-01 4.68897939e-01 -2.43857563e-01 -6.77796081e-02
1.13952887e+00 -4.39225733e-02 -3.53462547e-01 -4.96674627e-01
-4.70150322e-01 5.33801377e-01 3.74605745e-01 2.73211390e-01
4.41085905e-01 -9.98508573e-01 -8.32603216e-01 9.17813182e-01
-2.52480149e-01 2.29879469e-01 2.14794651e-01 4.39858735e-01
3.64407837e-01 7.82236606e-02 2.46809557e-01 -7.16853559e-01
-1.78118563e+00 3.26558352e-01 3.04766029e-01 -1.18479572e-01
-3.07641864e-01 1.38070571e+00 3.38310063e-01 -1.02713144e+00
7.85999596e-01 -1.41134530e-01 6.08090401e-01 -3.22529912e-01
6.34873807e-01 3.26274484e-01 1.69146314e-01 -3.87261331e-01
-2.33905628e-01 -2.57000923e-01 -4.77359414e-01 -2.39832520e-01
1.30855191e+00 -1.85875386e-01 -2.28994742e-01 4.56048027e-02
1.33442104e+00 1.02311939e-01 -1.65442932e+00 -7.80995607e-01
-1.49576426e-01 -7.26937592e-01 2.49259457e-01 -8.22194338e-01
-1.03762484e+00 7.72951722e-01 8.51245165e-01 3.21896151e-02
1.42864990e+00 -3.54727864e-01 7.86171138e-01 1.63600251e-01
5.24853945e-01 -9.80746090e-01 -5.54573953e-01 2.67132640e-01
4.04689550e-01 -1.16021574e+00 -2.08866790e-01 -1.97275147e-01
-6.27396822e-01 6.05938077e-01 4.92461681e-01 6.57698810e-02
9.36664462e-01 7.03250468e-01 3.61401498e-01 5.35928607e-01
-1.15071607e+00 7.22503006e-01 1.04293332e-01 9.56761837e-01
-8.80811810e-02 2.72409588e-01 1.01291090e-01 1.18832290e+00
-3.24648112e-01 -9.66455787e-02 1.59657728e-02 9.90922809e-01
-3.53460312e-01 -1.30473971e+00 -4.41268355e-01 5.08631527e-01
-7.05880761e-01 1.81352571e-02 -6.56872392e-01 1.44840807e-01
-1.72779132e-02 1.06928289e+00 6.53743073e-02 -6.24493778e-01
1.87213451e-01 4.81206596e-01 1.45943850e-01 -4.16268080e-01
-1.29115152e+00 -1.80461079e-01 -2.26949424e-01 -4.86046553e-01
-9.14867315e-03 -8.17177594e-01 -7.57902861e-01 -5.60368240e-01
-4.85502958e-01 4.35085058e-01 7.55640924e-01 7.63641059e-01
6.84269369e-01 -9.35618505e-02 1.49109483e+00 -1.40059248e-01
-1.62182558e+00 -6.61987305e-01 -8.56786251e-01 9.15504247e-02
6.34246647e-01 1.68818519e-01 -7.71920323e-01 -5.43818362e-02] | [5.885386943817139, 6.760809898376465] |
376701f0-cf59-4172-be01-3c23383eed37 | generating-templated-caption-for-video | 2301.05997 | null | https://arxiv.org/abs/2301.05997v2 | https://arxiv.org/pdf/2301.05997v2.pdf | Exploiting Prompt Caption for Video Grounding | Video grounding aims to locate a moment of interest matching the given query sentence from an untrimmed video. Previous works ignore the \emph{sparsity dilemma} in video annotations, which fails to provide the context information between potential events and query sentences in the dataset. In this paper, we contend that exploiting easily available captions which describe general actions \ie, prompt captions (PC) defined in our paper, will significantly boost the performance. To this end, we propose a Prompt Caption Network (PCNet) for video grounding. Specifically, we first introduce dense video captioning to generate dense captions and then obtain prompt captions by Non-Prompt Caption Suppression (NPCS). To capture the potential information in prompt captions, we propose Caption Guided Attention (CGA) project the semantic relations between prompt captions and query sentences into temporal space and fuse them into visual representations. Considering the gap between prompt captions and ground truth, we propose Asymmetric Cross-modal Contrastive Learning (ACCL) for constructing more negative pairs to maximize cross-modal mutual information. Without bells and whistles, extensive experiments on three public datasets (\ie, ActivityNet Captions, TACoS and ActivityNet-CG) demonstrate that our method significantly outperforms state-of-the-art methods. | ['Yuexian Zou', 'Yaowei Li', 'Zhihong Zhu', 'Xuxin Cheng', 'Meng Cao', 'Hongxiang Li'] | 2023-01-15 | null | null | null | null | ['video-grounding', 'dense-video-captioning'] | ['computer-vision', 'computer-vision'] | [ 4.46633190e-01 4.56955209e-02 -4.52384591e-01 -3.99335355e-01
-1.10246849e+00 -5.66304088e-01 6.42483532e-01 -1.26981974e-01
-3.17773938e-01 7.03611791e-01 8.00396979e-01 1.20212115e-01
2.54436642e-01 -2.44602188e-01 -1.08089066e+00 -4.08507019e-01
-9.65509638e-02 1.98395993e-03 1.23836756e-01 -1.19456284e-01
4.06666212e-02 -1.67537913e-01 -1.34226322e+00 9.24912035e-01
6.11851394e-01 1.07426620e+00 4.93635893e-01 5.45036912e-01
-1.18545974e-02 1.30622447e+00 -4.36876327e-01 -4.83602941e-01
-3.56395580e-02 -8.91177177e-01 -8.09179842e-01 2.69562483e-01
6.81754172e-01 -4.97759730e-01 -9.24758732e-01 1.11238015e+00
2.69963861e-01 2.76727825e-01 2.77053386e-01 -1.75002861e+00
-8.17881584e-01 6.05863214e-01 -6.66844368e-01 6.42178655e-01
8.86868536e-01 3.73914748e-01 1.32980633e+00 -1.02057409e+00
9.55163121e-01 1.09023964e+00 2.64880598e-01 7.22786367e-01
-9.83604431e-01 -6.19143486e-01 6.21409953e-01 5.55510342e-01
-1.39741373e+00 -3.96515459e-01 1.01679695e+00 -3.58253896e-01
6.51519895e-01 5.22389650e-01 6.92265868e-01 1.72619355e+00
-2.14955091e-01 1.19921291e+00 6.15157425e-01 -1.50122121e-01
5.82153313e-02 -4.50358838e-02 -5.38399257e-02 5.96304297e-01
-2.24845290e-01 -9.28611308e-02 -1.07476544e+00 -1.38905749e-01
6.25801146e-01 1.03795595e-01 -7.24983454e-01 -1.23217158e-01
-1.91770172e+00 7.09818423e-01 6.04389191e-01 3.24790508e-01
-6.69461846e-01 3.05000365e-01 4.26239967e-01 -8.52059498e-02
3.84336859e-01 3.73980343e-01 -6.53864592e-02 -1.02563627e-01
-8.91732931e-01 2.40606904e-01 2.28472054e-01 1.19836020e+00
7.28822172e-01 -1.38585970e-01 -9.40571070e-01 5.15367568e-01
2.98789650e-01 6.62560105e-01 4.04778361e-01 -1.07369947e+00
9.98358727e-01 4.53925550e-01 2.56960183e-01 -1.42645061e+00
-5.97369485e-02 -1.11375123e-01 -6.11911774e-01 -7.88792491e-01
6.69992045e-02 -2.65943587e-01 -8.65780532e-01 1.99169517e+00
4.99090031e-02 9.65651751e-01 1.48454681e-01 1.38297963e+00
1.16735029e+00 1.01947844e+00 4.53311324e-01 -3.98292810e-01
1.42487097e+00 -1.06146491e+00 -9.54074085e-01 -6.23802423e-01
5.51537693e-01 -5.09860992e-01 1.14836013e+00 -1.95673347e-01
-8.04000258e-01 -4.90760505e-01 -8.00168753e-01 2.07466204e-02
-2.55609043e-02 6.36164322e-02 4.05211389e-01 -1.81910485e-01
-9.40698326e-01 1.95271242e-02 -6.21278763e-01 -3.87092680e-01
4.07161981e-01 -1.41567010e-02 -4.66461986e-01 -2.28996798e-01
-1.56498003e+00 4.18698788e-01 4.54059750e-01 2.09346071e-01
-1.20603800e+00 -5.40429771e-01 -1.23685002e+00 -1.68042794e-01
6.38159871e-01 -7.19600737e-01 1.23463762e+00 -1.47990823e+00
-8.48825276e-01 1.02450383e+00 -4.15410787e-01 -6.26325011e-01
2.02002153e-01 -4.44168091e-01 -5.28014183e-01 7.62035191e-01
4.36133623e-01 1.33374858e+00 8.00647438e-01 -1.23203707e+00
-4.00486946e-01 1.66691005e-01 3.59241009e-01 4.17991906e-01
-2.84101129e-01 2.67425835e-01 -8.75098467e-01 -7.55332947e-01
1.32312542e-02 -8.96690607e-01 -1.71005964e-01 -1.26396403e-01
-4.33046997e-01 -1.80484772e-01 8.31800699e-01 -8.36349308e-01
1.25735581e+00 -2.32889771e+00 2.26857245e-01 -1.44401893e-01
1.40064538e-01 7.14974552e-02 -4.84716654e-01 4.68708813e-01
-9.65730697e-02 -2.46792883e-02 -1.52193487e-01 -1.97650611e-01
-1.30322054e-01 4.34562236e-01 -6.26683772e-01 3.82567853e-01
4.44189429e-01 1.32092869e+00 -1.40376663e+00 -8.83209586e-01
1.88237831e-01 4.82923985e-01 -5.62904477e-01 5.89647591e-01
-5.81551790e-01 6.43321693e-01 -5.93230307e-01 6.66319907e-01
1.68152422e-01 -5.86716950e-01 8.12810287e-02 -5.41650355e-01
2.55263537e-01 1.09062288e-02 -6.40461624e-01 2.12382841e+00
-1.63140461e-01 8.68874788e-01 -1.12382896e-01 -9.36219811e-01
7.00472236e-01 6.44985259e-01 6.37994230e-01 -8.95995975e-01
1.02525530e-02 -2.42572054e-01 -5.63506842e-01 -1.08712530e+00
4.06717926e-01 2.09361419e-01 -1.89626679e-01 1.39702767e-01
1.78922579e-01 3.45501482e-01 1.67560637e-01 7.45509624e-01
9.10599351e-01 2.84497678e-01 4.36026119e-02 2.24558920e-01
5.04455090e-01 6.79432377e-02 5.93079865e-01 6.32611156e-01
-4.20010805e-01 1.03827024e+00 7.19693661e-01 -1.97607413e-01
-7.55085468e-01 -1.00084567e+00 5.65580189e-01 1.32373846e+00
5.81246257e-01 -5.92776179e-01 -6.91944361e-01 -8.07996988e-01
-4.92011100e-01 7.34243155e-01 -7.43219197e-01 -1.09882504e-01
-7.51811981e-01 -4.14823800e-01 2.92301059e-01 5.98117769e-01
5.28782129e-01 -1.30922949e+00 -4.21851099e-01 -6.80267736e-02
-1.19290650e+00 -1.68202078e+00 -9.85509396e-01 -3.61971349e-01
-4.60044503e-01 -1.08850050e+00 -9.48876023e-01 -9.15003955e-01
6.94752455e-01 4.27205771e-01 1.31226540e+00 -9.14520323e-02
1.21909879e-01 7.88840473e-01 -6.20997429e-01 6.47810400e-02
9.13647562e-02 -3.25763613e-01 -1.60137549e-01 5.36937416e-01
5.85859060e-01 -2.70873845e-01 -7.16073215e-01 2.32694775e-01
-8.58705342e-01 4.71239001e-01 4.36950207e-01 7.38756180e-01
7.95225918e-01 -7.63657451e-01 5.12352109e-01 -3.97089601e-01
2.98736155e-01 -9.70637321e-01 -1.24517195e-01 2.29388833e-01
-2.43229489e-03 -2.62668014e-01 2.68716812e-01 -4.92615640e-01
-8.63339722e-01 2.06785858e-01 1.78753629e-01 -1.13215446e+00
-1.13864794e-01 4.95904416e-01 -1.67770222e-01 4.09628451e-01
4.45050269e-01 5.29514849e-01 -2.56606191e-01 -1.27339929e-01
4.39377308e-01 4.59147513e-01 1.03135169e+00 -4.67843622e-01
5.36252677e-01 5.30292749e-01 -2.41814315e-01 -5.69833577e-01
-1.32607925e+00 -6.32484198e-01 -3.15564364e-01 -6.17730141e-01
1.32902896e+00 -1.29408658e+00 -5.74164987e-01 -2.20870435e-01
-1.46242118e+00 -3.10737863e-02 -8.99371058e-02 5.83836198e-01
-8.41234028e-01 3.77885044e-01 -3.78842950e-01 -6.31793082e-01
-1.95626244e-01 -1.01662445e+00 1.35385990e+00 1.92581117e-01
-2.66337037e-01 -7.46329606e-01 -7.25167543e-02 5.68742335e-01
4.83203419e-02 5.92339814e-01 1.22971520e-01 -7.51922846e-01
-8.47359121e-01 -7.81200901e-02 -5.15351355e-01 1.76092803e-01
-6.97065517e-02 -3.67076784e-01 -8.87635708e-01 -5.95806167e-02
-1.91123232e-01 -3.96372229e-01 7.12722600e-01 3.89321476e-01
1.12179482e+00 -7.17304885e-01 -3.71499717e-01 5.88856041e-01
1.12236071e+00 9.52720270e-03 6.32049561e-01 1.84392378e-01
8.62543166e-01 6.03883862e-01 9.87464964e-01 4.89810079e-01
4.56825763e-01 6.55723512e-01 7.87663162e-01 -5.02911247e-02
-8.88671651e-02 -7.35207498e-01 7.36465275e-01 5.29262900e-01
1.37682604e-02 -5.03675044e-01 -7.75763869e-01 8.55829656e-01
-2.23050237e+00 -1.26555336e+00 -5.74586876e-02 1.68437541e+00
5.97342849e-01 -2.03381479e-01 -2.50394791e-02 -4.25852388e-01
1.08584917e+00 5.53101063e-01 -3.10054183e-01 2.74736166e-01
-2.12645754e-01 -4.07273412e-01 2.18577355e-01 3.09290588e-01
-1.21998763e+00 8.47135603e-01 5.27317047e+00 6.46403611e-01
-1.00470793e+00 3.34505200e-01 7.29125977e-01 -3.01522791e-01
-4.59705472e-01 -1.67769909e-01 -4.85682666e-01 7.25502431e-01
9.60763693e-01 -2.66525149e-01 3.02626550e-01 7.21935749e-01
5.39262831e-01 9.47030932e-02 -1.32428658e+00 1.36493504e+00
5.20072043e-01 -1.48542416e+00 2.37050131e-01 -4.45149660e-01
7.81373978e-01 9.30136666e-02 -4.74410988e-02 3.47049057e-01
-1.12923019e-01 -7.78590560e-01 8.62641811e-01 5.69921911e-01
7.00228989e-01 -3.54660720e-01 6.79100394e-01 4.56176884e-02
-1.27945769e+00 8.32157657e-02 -8.87647048e-02 7.68478811e-02
5.91079116e-01 1.23468503e-01 -6.57451868e-01 6.47795141e-01
7.30539560e-01 1.06705916e+00 -5.85371554e-01 9.36911643e-01
-2.85881460e-01 6.51967227e-01 -9.08565819e-02 1.45722069e-02
7.51421392e-01 1.44064263e-01 7.16788471e-01 1.29618454e+00
2.41103500e-01 3.14737588e-01 3.59347612e-01 7.67201960e-01
-2.27709144e-01 1.21526241e-01 -7.09801614e-01 -2.94422865e-01
3.92098248e-01 9.85495389e-01 -6.97158694e-01 -4.77106303e-01
-6.91811621e-01 1.20416427e+00 1.02232933e-01 7.97862709e-01
-1.21696424e+00 -1.51990410e-02 4.95490104e-01 6.05698638e-02
2.74066389e-01 1.07349698e-02 2.10929871e-01 -1.46478617e+00
3.39303344e-01 -7.62195468e-01 6.46694720e-01 -1.32826889e+00
-1.21198940e+00 7.13132977e-01 3.17121238e-01 -1.54383969e+00
-2.93342382e-01 -1.07398860e-01 -5.48753440e-01 5.10398805e-01
-1.36620712e+00 -1.18186867e+00 -6.59177363e-01 9.20317352e-01
8.97125840e-01 2.24487484e-01 2.01911524e-01 4.97909725e-01
-4.63414878e-01 3.75735164e-01 -6.07300401e-01 3.53282928e-01
7.87387311e-01 -7.04831600e-01 8.58720914e-02 1.03756332e+00
4.83424127e-01 3.42408985e-01 8.54134619e-01 -7.91279972e-01
-1.27138364e+00 -1.44245863e+00 1.01518369e+00 -5.19500971e-01
7.51001716e-01 -2.56114960e-01 -8.23396564e-01 1.06665313e+00
5.65453827e-01 2.80163497e-01 6.11614585e-01 -4.29343075e-01
-4.10224408e-01 1.13174044e-01 -6.28421724e-01 6.75369322e-01
1.31055701e+00 -7.89413571e-01 -8.76760840e-01 7.36521661e-01
1.26848769e+00 -4.17599887e-01 -3.64735276e-01 3.42488468e-01
2.00411886e-01 -7.38239884e-01 1.05069149e+00 -8.83709848e-01
6.03832126e-01 -3.14981192e-01 -4.83859837e-01 -9.09598827e-01
-3.28368768e-02 -8.96417022e-01 -2.49464840e-01 1.42487097e+00
2.89996177e-01 1.84629947e-01 6.94236219e-01 5.93749762e-01
-3.23797166e-01 -4.89390671e-01 -8.82817328e-01 -6.22217894e-01
-6.54682875e-01 -6.23634875e-01 4.07626629e-01 1.13288677e+00
2.34789371e-01 4.86906707e-01 -8.65371287e-01 2.97702312e-01
3.22708786e-01 1.75772712e-01 4.70987648e-01 -5.54958582e-01
-9.66970101e-02 5.35567012e-03 -3.72408330e-01 -1.27781999e+00
4.42928106e-01 -7.46790528e-01 2.81496346e-01 -1.51297081e+00
4.35787827e-01 1.86243936e-01 -4.41356331e-01 5.57303786e-01
-4.51605350e-01 6.15258992e-01 3.84975404e-01 2.40311205e-01
-1.43896866e+00 7.10623026e-01 1.31080246e+00 -3.03206027e-01
-8.06969255e-02 -5.05875230e-01 -5.53146660e-01 6.22539520e-01
2.65381813e-01 -4.85425651e-01 -4.90457237e-01 -6.18977964e-01
2.28516027e-01 4.49867994e-01 7.93986022e-01 -8.77798975e-01
2.13860810e-01 -3.75847369e-01 5.99199794e-02 -6.01384103e-01
4.65155214e-01 -6.69290781e-01 2.54288930e-02 7.68350437e-02
-7.28723586e-01 1.93240359e-01 4.25032377e-02 9.69552994e-01
-5.73624909e-01 -6.76217154e-02 3.04511935e-01 -1.04967207e-01
-1.10618961e+00 4.76666600e-01 -2.16247156e-01 4.20532167e-01
1.18861246e+00 5.73267639e-02 -3.08600783e-01 -7.45456219e-01
-8.73144507e-01 6.07373655e-01 1.79100603e-01 7.24842846e-01
8.31819355e-01 -1.67833352e+00 -8.76387477e-01 -1.77330092e-01
4.64377820e-01 -1.83864892e-01 5.43853998e-01 9.63713586e-01
-2.06014976e-01 5.39903581e-01 -6.20612614e-02 -7.54311979e-01
-1.21918380e+00 9.35147524e-01 2.14214791e-02 1.37455136e-01
-6.17378116e-01 9.78714406e-01 6.29274964e-01 3.25062782e-01
3.66691262e-01 -2.56842434e-01 -2.69997686e-01 1.16922617e-01
6.67348206e-01 -1.45531952e-01 -5.65260291e-01 -1.09974980e+00
-4.77495909e-01 3.84271801e-01 4.54618745e-02 -9.32728648e-02
1.01733935e+00 -3.71268928e-01 1.43311396e-01 3.15703809e-01
1.34977877e+00 -2.84976333e-01 -1.45864403e+00 -1.42870992e-01
-7.97042400e-02 -5.91562569e-01 -1.00540243e-01 -4.68022406e-01
-1.06812751e+00 6.31202757e-01 2.52342373e-01 -3.61655839e-02
1.24798131e+00 5.20053566e-01 9.37274933e-01 3.93314391e-01
2.67880689e-02 -6.89138174e-01 6.22195125e-01 1.47223115e-01
1.17612004e+00 -1.52559912e+00 -3.50056589e-01 -4.56295937e-01
-1.13371885e+00 6.73303187e-01 9.68922079e-01 4.32392908e-03
6.88577667e-02 -3.17265093e-01 2.74121687e-02 -2.71477848e-01
-9.02665496e-01 -4.19142693e-01 4.11153406e-01 5.85049570e-01
1.67358622e-01 -2.95126706e-01 -1.91096276e-01 7.30618358e-01
2.45493352e-01 2.65646074e-02 4.32808131e-01 7.25090683e-01
-2.60120213e-01 -4.97719467e-01 -4.35497284e-01 2.18201339e-01
-4.38412338e-01 -3.32946718e-01 -5.16120315e-01 5.14820099e-01
4.48571108e-02 1.03037214e+00 2.63515025e-01 -4.93077844e-01
1.26212165e-01 1.63455345e-02 1.76778480e-01 -5.50959945e-01
-3.50489289e-01 1.06885443e-02 1.55648276e-01 -9.36588407e-01
-7.80808926e-01 -6.21277869e-01 -1.09151113e+00 2.61563510e-01
5.02027981e-02 3.38441581e-01 7.48599470e-02 9.03726339e-01
5.62678635e-01 4.30250168e-01 4.61635947e-01 -8.92941713e-01
1.13808341e-01 -8.43043327e-01 -2.94617321e-02 8.14026654e-01
4.40675497e-01 -6.64958596e-01 -4.63247806e-01 6.65268958e-01] | [10.304241180419922, 0.7369306683540344] |
e5ce2c6a-deb5-4a39-b487-d6cb819c71ff | benchmarking-automatic-machine-learning | 1808.06492 | null | http://arxiv.org/abs/1808.06492v1 | http://arxiv.org/pdf/1808.06492v1.pdf | Benchmarking Automatic Machine Learning Frameworks | AutoML serves as the bridge between varying levels of expertise when
designing machine learning systems and expedites the data science process. A
wide range of techniques is taken to address this, however there does not exist
an objective comparison of these techniques. We present a benchmark of current
open source AutoML solutions using open source datasets. We test auto-sklearn,
TPOT, auto_ml, and H2O's AutoML solution against a compiled set of regression
and classification datasets sourced from OpenML and find that auto-sklearn
performs the best across classification datasets and TPOT performs the best
across regression datasets. | ['Alexander Allen', 'Adithya Balaji'] | 2018-08-17 | null | null | null | null | ['automated-feature-engineering'] | ['methodology'] | [-6.21809304e-01 1.69128031e-01 -2.21052811e-01 -6.16774440e-01
-6.96488738e-01 -8.73189509e-01 5.56661546e-01 5.10138981e-02
-1.62683725e-01 9.58650708e-01 9.97895300e-02 -5.04134417e-01
-3.67338508e-01 -3.93229395e-01 -5.82713544e-01 -1.47695586e-01
3.10591310e-01 7.82697976e-01 5.04630841e-02 -9.80959460e-02
1.03101447e-01 2.75956780e-01 -1.66934943e+00 3.64221960e-01
7.29499102e-01 7.38610089e-01 -1.36488587e-01 7.42026389e-01
-5.27128994e-01 1.22337925e+00 -5.12113333e-01 -3.81221026e-01
3.98296744e-01 -1.71062544e-01 -9.21839595e-01 -5.66529453e-01
6.31079674e-01 6.80712283e-01 3.21011655e-02 3.74497354e-01
6.47489548e-01 -7.43677616e-02 3.70737404e-01 -1.56079531e+00
-5.18355370e-01 6.58991516e-01 2.13317931e-01 1.57905951e-01
4.27411258e-01 5.32458067e-01 1.03000140e+00 -6.80139363e-01
7.95168281e-01 1.14735401e+00 1.07994795e+00 5.46640694e-01
-1.38368666e+00 -6.31977379e-01 -4.59258318e-01 -5.08853644e-02
-1.28852272e+00 -5.64079762e-01 3.71968895e-01 -6.65647388e-01
1.56704450e+00 3.39269042e-01 6.40629232e-01 1.06315923e+00
3.61555815e-01 7.91858733e-01 1.49732482e+00 -4.43571001e-01
2.64073521e-01 6.21344984e-01 6.69573843e-01 8.11835706e-01
2.60979712e-01 8.18782747e-02 -7.84517765e-01 -1.76689833e-01
3.60175014e-01 -4.62804049e-01 1.20985419e-01 -1.63097054e-01
-9.04003203e-01 4.90533322e-01 -2.16471646e-02 3.74227762e-01
-2.24639684e-01 6.87578768e-02 2.64236778e-01 8.08138609e-01
5.05305767e-01 1.15948987e+00 -1.28183568e+00 -4.84137058e-01
-7.54926026e-01 3.75735790e-01 1.34791648e+00 1.22467816e+00
1.02397656e+00 -1.28551079e-02 2.35092431e-01 6.47974551e-01
3.46551508e-01 1.64497957e-01 6.48701131e-01 -9.58120763e-01
4.04768407e-01 1.05148470e+00 -2.19810873e-01 -5.00124931e-01
-4.33280289e-01 -2.12314025e-01 -1.82896331e-01 2.13403419e-01
3.41680974e-01 -3.56094718e-01 -5.91365755e-01 1.20643330e+00
2.91396886e-01 -1.78224389e-02 4.40633178e-01 2.75949597e-01
1.21436918e+00 5.15203059e-01 4.10031348e-01 -9.22463834e-03
6.53731942e-01 -1.07397902e+00 -6.17748678e-01 -4.91607606e-01
1.15332651e+00 -6.04581535e-01 9.99144435e-01 6.27500951e-01
-9.43360269e-01 -5.86688399e-01 -9.75373626e-01 -3.46745819e-01
-9.79421318e-01 -2.65036315e-01 9.78836894e-01 6.20493412e-01
-8.38072121e-01 5.59790015e-01 -6.19887948e-01 -4.78422940e-01
3.27142030e-01 2.01322660e-01 -7.29904890e-01 -3.57166044e-02
-9.27601814e-01 1.37126398e+00 7.29838610e-01 -2.22116366e-01
-7.19584346e-01 -1.03309703e+00 -7.52783597e-01 -3.73046905e-01
3.04592997e-01 -5.00319779e-01 1.33962822e+00 -7.26580143e-01
-1.12326360e+00 8.92457724e-01 2.53760815e-01 -4.64574307e-01
5.35900652e-01 -3.41537356e-01 -5.92813194e-01 -7.63138652e-01
-1.26779273e-01 4.59795684e-01 2.65153348e-01 -8.99849772e-01
-6.62572920e-01 -3.37756902e-01 -2.34423444e-01 3.14066969e-02
-5.93952872e-02 9.17936787e-02 -1.91626206e-01 1.65586863e-02
-5.81420958e-01 -7.94649720e-01 -1.86973810e-01 -3.67766172e-01
-1.32509440e-01 -5.79895020e-01 8.23258996e-01 -4.50490743e-01
1.50026011e+00 -1.86875069e+00 1.55896887e-01 -2.59347856e-02
2.76859522e-01 2.78258979e-01 -2.62066990e-01 6.50788844e-01
-1.37446299e-01 3.32810670e-01 1.16657265e-01 -8.62705633e-02
1.02634057e-02 5.88304818e-01 1.18319586e-01 2.02591449e-01
-8.56579170e-02 1.02515531e+00 -7.27604866e-01 -7.14767694e-01
2.64526099e-01 2.23322168e-01 -4.37019259e-01 4.55574244e-01
-7.12718427e-01 1.88212425e-01 -4.24589604e-01 7.27858722e-01
-5.10208942e-02 -3.02009284e-01 1.80775359e-01 -6.25853473e-03
-4.66476977e-01 2.34579459e-01 -9.94590402e-01 1.98085272e+00
-6.30197346e-01 7.96096683e-01 -2.61087269e-01 -4.53915179e-01
1.03009880e+00 5.01127899e-01 4.33318794e-01 -3.41894448e-01
-3.80230583e-02 2.63334125e-01 7.43913278e-02 -9.05137956e-01
-4.95077185e-02 -1.96329523e-02 -1.96305856e-01 6.97112262e-01
6.25914395e-01 -2.73572117e-01 3.27005833e-01 8.80945623e-02
1.62000787e+00 5.47286391e-01 7.25549877e-01 -3.85700375e-01
3.74946982e-01 3.25690269e-01 4.18746412e-01 6.59403265e-01
4.25867327e-02 1.70572177e-01 3.16859990e-01 -8.18711519e-01
-1.01159656e+00 -6.22725606e-01 -1.66681692e-01 1.29031277e+00
-7.12614834e-01 -9.86441672e-01 -6.26386762e-01 -9.65740681e-01
3.41085106e-01 1.09522688e+00 -6.40977204e-01 1.86310783e-01
-6.37142127e-03 -4.28732753e-01 7.23496795e-01 3.46108317e-01
2.76200294e-01 -1.23963821e+00 -5.47856748e-01 2.28577569e-01
1.29477203e-01 -1.00422072e+00 2.46070594e-01 4.13165331e-01
-7.44999468e-01 -1.36103833e+00 5.22688687e-01 -3.21333379e-01
2.65120417e-01 -3.29023004e-01 1.61058033e+00 8.83155018e-02
-5.51688135e-01 6.10422671e-01 -4.46787387e-01 -9.33829606e-01
-8.35765541e-01 5.67733288e-01 6.77083898e-03 -6.53225183e-01
8.96497369e-01 -4.87321943e-01 7.96757042e-02 3.01212162e-01
-6.69753313e-01 1.00551590e-01 5.03079951e-01 4.96155292e-01
3.24393451e-01 1.67015493e-01 6.07188046e-01 -1.20534408e+00
6.96550131e-01 -8.62638950e-01 -7.82946885e-01 5.26324451e-01
-1.27401149e+00 3.21944356e-01 6.56875491e-01 3.57340835e-03
-7.41662085e-01 1.36107400e-01 -1.99712232e-01 -2.82379270e-01
-3.79457563e-01 7.48947263e-01 -2.45669454e-01 1.29106380e-02
9.87323642e-01 -1.23411603e-01 5.98900430e-02 -9.41685081e-01
5.18391669e-01 8.79950285e-01 1.48116991e-01 -6.28024161e-01
4.62129921e-01 -2.07987085e-01 -2.86338687e-01 -7.36918271e-01
-1.09163690e+00 -3.51346165e-01 -7.77825296e-01 -1.98581636e-01
5.52315772e-01 -6.75561547e-01 -5.39910138e-01 5.67457005e-02
-8.56529415e-01 -6.64626718e-01 -5.30494273e-01 1.46005675e-01
-6.27970815e-01 -3.45768660e-01 -1.39948562e-01 -5.11682689e-01
-4.33014005e-01 -9.30507839e-01 5.29361248e-01 1.46646902e-03
-7.75920033e-01 -1.34249651e+00 5.62887728e-01 6.38849676e-01
5.17921388e-01 3.27783316e-01 9.58329618e-01 -1.25448585e+00
-2.93133765e-01 -5.12985826e-01 2.42366254e-01 5.30084908e-01
1.75639212e-01 3.70022058e-01 -1.12893605e+00 -1.52525716e-02
-4.69749570e-01 -7.65976667e-01 2.60872483e-01 -1.68960005e-01
9.83740687e-01 -3.13199669e-01 -3.26920271e-01 5.63676059e-01
1.62051463e+00 -7.42391199e-02 4.95984137e-01 5.09072602e-01
5.87474644e-01 5.75763881e-01 5.82497299e-01 2.51083523e-01
3.16153497e-01 2.11574882e-01 6.80886060e-02 2.25139171e-01
-1.26363322e-01 -2.83897877e-01 3.02178323e-01 8.52003336e-01
1.37645053e-02 -1.23573661e-01 -1.46365106e+00 2.16085061e-01
-2.02666903e+00 -8.34067762e-01 -3.71243298e-01 1.87108099e+00
1.19715989e+00 6.07522912e-02 1.47753945e-02 3.79093587e-02
-1.11971311e-01 -1.09200098e-01 -5.80634356e-01 -5.25175214e-01
-4.32979129e-02 4.28794086e-01 2.38488615e-01 4.93629187e-01
-7.33440161e-01 1.23600662e+00 8.22746277e+00 5.04058599e-01
-7.30181694e-01 1.40011311e-01 3.08215227e-02 -3.15235466e-01
-3.38305682e-02 3.05943727e-01 -1.10345006e+00 3.48231673e-01
1.68000066e+00 -5.21067441e-01 5.43841124e-01 1.14952755e+00
-3.97415757e-02 -1.00790031e-01 -1.64920712e+00 6.29234612e-01
-5.49063012e-02 -1.39970911e+00 -2.76596606e-01 -1.61759317e-01
6.08280361e-01 5.51678419e-01 -4.29316819e-01 7.20400989e-01
9.48435068e-01 -1.38113248e+00 5.05927622e-01 7.84762442e-01
4.27996427e-01 -5.65330684e-01 4.80963588e-01 6.20269835e-01
-6.24844491e-01 -1.30351156e-01 -1.91905290e-01 -2.49007985e-01
-4.35765117e-01 5.09211779e-01 -1.13656020e+00 5.62569857e-01
7.83747852e-01 9.65797126e-01 -1.05020666e+00 8.88388932e-01
-3.01943153e-01 8.23664129e-01 -1.87340721e-01 7.42785335e-02
9.84229520e-02 6.96743950e-02 1.96765810e-01 1.44983506e+00
-1.44279674e-01 -1.98427603e-01 1.45492330e-01 7.16707647e-01
2.23988950e-01 2.80762523e-01 -9.49035168e-01 -3.63147378e-01
5.29202700e-01 1.43567681e+00 1.80544388e-02 -1.44470245e-01
-7.62489676e-01 3.30021650e-01 4.74532336e-01 -2.37219539e-02
-4.26041275e-01 -1.26764029e-01 6.93607926e-01 8.84968340e-02
-1.36004508e-01 -5.86752035e-02 -2.23153502e-01 -8.64068210e-01
-5.83715856e-01 -1.54014099e+00 7.36754060e-01 -1.19074559e+00
-1.48192465e+00 4.70024586e-01 5.12045696e-02 -6.79937184e-01
-4.03107554e-01 -6.93753719e-01 -2.55369782e-01 7.85745978e-01
-1.15482914e+00 -1.23740149e+00 -4.55988437e-01 3.57905358e-01
5.88537753e-01 -5.61255515e-01 1.15890300e+00 8.18426460e-02
-8.02035689e-01 1.46450296e-01 2.76821293e-02 -1.09738357e-01
9.10728276e-01 -1.49443054e+00 4.79384303e-01 3.51702362e-01
-7.82906543e-04 7.51673281e-01 8.31858218e-01 -8.81000459e-01
-1.80531967e+00 -1.25873017e+00 6.34139717e-01 -1.38687921e+00
8.88850868e-01 -2.52778649e-01 -9.68003213e-01 1.31353343e+00
5.63220739e-01 6.10707551e-02 1.16189480e+00 4.51263428e-01
-5.17870009e-01 -1.80880159e-01 -1.08770251e+00 1.30966663e-01
9.41324532e-01 -3.32774222e-01 -8.86212587e-01 3.95970196e-01
4.78764623e-01 -3.00098926e-01 -1.52633989e+00 2.65678078e-01
7.47564256e-01 -9.11379874e-01 4.23541069e-01 -1.22096717e+00
4.49346870e-01 -1.15198582e-01 -1.43076852e-01 -1.45295084e+00
-1.63211733e-01 -6.78882122e-01 -3.06583792e-01 1.46390688e+00
9.11244810e-01 -7.98501015e-01 5.97674787e-01 9.52367246e-01
-1.29099175e-01 -8.54279220e-01 -5.32428622e-01 -8.74193192e-01
1.57592505e-01 -7.35780716e-01 6.42735124e-01 1.21328628e+00
-2.65491307e-02 5.86663544e-01 -2.86542606e-02 -2.59536028e-01
5.39161503e-01 -6.64517805e-02 1.45610368e+00 -1.68051267e+00
-2.54321963e-01 -1.68061152e-01 -3.03639561e-01 -1.68592528e-01
4.58689988e-01 -1.33794367e+00 -2.74484456e-01 -1.67037976e+00
1.65032238e-01 -7.50527918e-01 -1.81143627e-01 1.18289661e+00
2.86483228e-01 -2.52021641e-01 -1.98071152e-01 1.63322121e-01
-7.75336683e-01 -6.57704994e-02 8.42230141e-01 1.19619258e-01
-1.23807959e-01 -3.69118780e-01 -8.62567544e-01 7.39369094e-01
7.45253444e-01 -6.42857611e-01 -4.40361977e-01 -2.44846344e-01
7.02232301e-01 -1.96738228e-01 2.16950439e-02 -1.20163155e+00
2.48894766e-01 -3.25163662e-01 4.13180381e-01 -4.48500872e-01
-1.93678401e-02 -9.13750112e-01 5.79822838e-01 3.71942818e-02
-4.87693906e-01 2.39450410e-01 4.36876595e-01 1.45084321e-01
1.75607458e-01 -2.97311187e-01 6.21970713e-01 -4.08138156e-01
-7.56959319e-01 1.46685645e-01 -2.50353575e-01 2.27567345e-01
1.11236393e+00 -5.77940308e-02 -5.72370052e-01 7.94106871e-02
-6.96667552e-01 7.04619527e-01 5.85660934e-01 8.14875841e-01
2.30386838e-01 -8.83531988e-01 -6.11381292e-01 4.32082057e-01
2.34313279e-01 -5.55448085e-02 -3.05268466e-01 5.86867213e-01
-5.20730197e-01 6.96296334e-01 -3.32204401e-01 -3.32261145e-01
-1.38381922e+00 4.56387222e-01 6.35384500e-01 -2.96701938e-01
-3.23112547e-01 6.70896590e-01 -6.13711476e-01 -1.07486391e+00
2.35614091e-01 4.52877320e-02 -6.12087324e-02 -6.68132082e-02
2.81092376e-01 5.74589193e-01 2.77436972e-01 -1.33800238e-01
-1.12876937e-01 2.43911147e-01 -4.75107990e-02 1.01816170e-01
1.72621608e+00 2.48880967e-01 -2.81076759e-01 1.13851476e+00
8.57088566e-01 -2.33774289e-01 -7.58973360e-01 -5.91919534e-02
4.20558482e-01 -2.37036780e-01 1.59849122e-01 -1.29997241e+00
-6.70741260e-01 4.77088392e-01 3.49458545e-01 3.71973485e-01
6.66744888e-01 9.57351327e-02 4.33838628e-02 7.16431856e-01
4.64191884e-01 -1.20453215e+00 -1.87494174e-01 4.96651947e-01
9.05779839e-01 -1.12858999e+00 3.53641659e-01 -9.73102637e-04
-6.10579610e-01 9.52851892e-01 1.17886448e+00 8.98032784e-02
6.99670136e-01 5.88115931e-01 2.87159592e-01 -4.83552635e-01
-1.55746126e+00 -6.68409020e-02 7.93148428e-02 4.08837408e-01
8.60797644e-01 -2.15484917e-01 -2.27929965e-01 4.78994906e-01
-4.11804229e-01 3.33158463e-01 1.73807666e-01 1.23079848e+00
-3.06321114e-01 -1.34919727e+00 -2.35372528e-01 6.10049963e-01
-3.01465720e-01 8.35357606e-03 -8.91086757e-01 1.21526039e+00
2.99908996e-01 7.40144014e-01 -2.82593727e-01 -3.79201710e-01
3.87437522e-01 7.64013588e-01 6.36482596e-01 -8.55525553e-01
-1.05201471e+00 -4.10464317e-01 5.82016349e-01 -6.51382625e-01
-3.25194240e-01 -6.62367821e-01 -1.07386589e+00 -1.78431273e-01
-3.80140841e-01 3.80370677e-01 1.01272190e+00 9.20084655e-01
5.46436548e-01 3.95753920e-01 2.30497256e-01 -2.44835556e-01
-3.21454078e-01 -1.05943561e+00 -2.97801226e-01 1.84903517e-01
2.71329414e-02 -4.60847378e-01 -4.33946908e-01 2.59965267e-02] | [8.950610160827637, 7.333078861236572] |
568963e6-d295-4e41-a4b2-897b62124295 | armoured-adversarially-robust-models-using | null | null | https://openreview.net/forum?id=JoCR4h9O3Ew | https://openreview.net/pdf?id=JoCR4h9O3Ew | ARMOURED: Adversarially Robust MOdels using Unlabeled data by REgularizing Diversity | Adversarial attacks pose a major challenge for modern deep neural networks. Recent advancements show that adversarially robust generalization requires a huge amount of labeled data for training. If annotation becomes a burden, can unlabeled data help bridge the gap? In this paper, we propose ARMOURED, an adversarially robust training method based on semi-supervised learning that consists of two components. The first component applies multi-view learning to simultaneously optimize multiple independent networks and utilizes unlabeled data to enforce labeling consistency. The second component reduces adversarial transferability among the networks via diversity regularizers inspired by determinantal point processes and entropy maximization. Notably, ARMOURED does not rely on generating adversarial samples during training. We demonstrate the robustness of ARMOURED on CIFAR-10 and SVHN datasets against state-of-the-art benchmarks in both the adversarial robust training and the semi-supervised training domains. Experimental results show that under projected gradient descent attacks with bounded $\ell_{\infty}$ norm, ARMOURED achieves substantial gains in accuracy, while maintaining high accuracy on clean samples. | ['Chuan-Sheng Foo', 'Yu Jing Goh', 'Kiran Chari', 'Xun Xu', 'Cuong Manh Nguyen', 'Kangkang Lu'] | 2021-01-01 | null | null | null | iclr-2021-1 | ['multi-view-learning'] | ['computer-vision'] | [ 1.90272212e-01 1.44911617e-01 -7.19961524e-02 -5.22681713e-01
-1.14642286e+00 -1.05171263e+00 3.00607920e-01 -3.85293007e-01
-4.25327033e-01 9.59671080e-01 -1.33024141e-01 -3.97135735e-01
7.19487444e-02 -5.41795194e-01 -9.70427573e-01 -8.14541340e-01
6.51510507e-02 2.95865029e-01 -1.83550760e-01 -4.03150469e-01
-4.40754503e-01 5.67163289e-01 -6.52506530e-01 1.08982332e-01
9.17741358e-01 1.19194257e+00 -6.74245059e-01 4.63801146e-01
2.98038185e-01 1.09061301e+00 -6.60121977e-01 -1.07333815e+00
7.62163818e-01 -4.44181144e-01 -6.36062086e-01 -8.05183277e-02
8.33489954e-01 -3.71578783e-01 -6.17870390e-01 1.40579486e+00
7.16680229e-01 1.41166463e-01 6.54808164e-01 -1.37406135e+00
-1.02121639e+00 6.76173389e-01 -4.07250404e-01 1.43386930e-01
-4.95281100e-01 1.49818122e-01 9.25640166e-01 -8.97930384e-01
4.26694274e-01 1.07837713e+00 9.47052360e-01 9.60785151e-01
-1.28458107e+00 -9.69350994e-01 3.53549719e-01 -2.97177285e-02
-1.12147164e+00 -5.88971436e-01 1.02437687e+00 -4.46652740e-01
4.96947825e-01 3.59878689e-01 1.22540243e-01 1.49749565e+00
9.48276222e-02 7.66464710e-01 1.01486087e+00 -1.10599659e-01
4.26645219e-01 2.72839248e-01 -5.27846031e-02 6.12051725e-01
1.18857801e-01 2.79926777e-01 -2.45048627e-01 -2.61271268e-01
4.70300466e-01 -1.10793887e-02 -4.20292675e-01 -4.57691014e-01
-6.74162686e-01 1.07735038e+00 4.27488834e-01 -1.70064673e-01
-8.66970886e-03 2.87820920e-02 7.25574076e-01 5.35548389e-01
7.26177037e-01 4.51585025e-01 -4.51568872e-01 3.84653300e-01
-9.15666819e-01 -4.18814570e-02 9.47868049e-01 1.00253034e+00
3.66407454e-01 7.57678866e-01 1.87299978e-02 8.07186186e-01
2.45785475e-01 7.47690260e-01 2.75144547e-01 -9.24676180e-01
7.99123347e-01 2.34416842e-01 -3.46604019e-01 -6.46352828e-01
-1.13554383e-02 -7.29952276e-01 -1.29460239e+00 5.60453832e-01
4.62898493e-01 -6.34690940e-01 -1.03624916e+00 2.02074146e+00
2.18421564e-01 1.35085061e-02 3.00924987e-01 7.01368630e-01
6.28223360e-01 2.49863163e-01 -4.25434895e-02 -1.19787917e-01
6.66795790e-01 -1.20933580e+00 -5.85705876e-01 -3.77500623e-01
3.15467507e-01 -5.34061670e-01 6.95632875e-01 4.03751463e-01
-1.17541230e+00 -3.28760445e-01 -1.33227813e+00 1.32716402e-01
-2.73993254e-01 -2.33473837e-01 2.41438910e-01 9.83775496e-01
-7.39308000e-01 7.99415469e-01 -8.22368681e-01 5.33316672e-01
1.07710445e+00 3.53054732e-01 -6.67448997e-01 -1.58579677e-01
-1.40958548e+00 7.22050071e-01 3.13513219e-01 1.44734651e-01
-1.33575714e+00 -7.57327914e-01 -1.06740272e+00 -1.91946421e-02
2.70490944e-01 -4.61376935e-01 9.12253320e-01 -1.33262432e+00
-1.51679361e+00 7.60722756e-01 4.47218806e-01 -7.68009126e-01
1.04124773e+00 -3.93353105e-01 -5.06651938e-01 1.99162707e-01
-9.65966061e-02 2.89177060e-01 1.32088387e+00 -1.46875417e+00
-9.95785072e-02 -5.88780880e-01 -1.04575083e-01 3.89873832e-02
-6.83107615e-01 -1.37398079e-01 -1.56475574e-01 -1.07421958e+00
-1.74326062e-01 -1.27796161e+00 -3.27006906e-01 1.31287336e-01
-7.51019597e-01 2.14524478e-01 9.56905782e-01 -7.99202263e-01
7.36177027e-01 -2.13576484e+00 3.13115090e-01 2.85665601e-01
3.74552757e-01 7.20159471e-01 -6.83180392e-02 1.81024149e-01
-4.62334156e-01 3.68604958e-01 -5.99732161e-01 -6.37865365e-01
1.57996058e-01 1.47428751e-01 -7.54629970e-01 8.66972506e-01
1.42832756e-01 8.74083340e-01 -7.53561080e-01 -2.08518043e-01
-6.01610430e-02 5.93085766e-01 -4.21503663e-01 1.65493175e-01
-1.49307782e-02 6.29005313e-01 -2.37484679e-01 5.68073153e-01
8.01668644e-01 -6.55163899e-02 -6.71667792e-03 1.55520186e-01
5.22188485e-01 -3.21934938e-01 -1.01720345e+00 1.72669828e+00
-3.61062795e-01 4.87770617e-01 3.45852762e-01 -1.14884865e+00
6.86387181e-01 4.40156698e-01 2.48864532e-01 -2.10015967e-01
2.89692611e-01 1.55952051e-01 -2.37505451e-01 -1.25805706e-01
-4.95129675e-02 -1.70488000e-01 -1.92958012e-01 2.86034882e-01
4.39288169e-01 -1.39203772e-01 -1.33609608e-01 2.85678715e-01
9.94324684e-01 7.98518062e-02 8.64105821e-02 -1.23033606e-01
5.63652873e-01 -3.28131109e-01 8.57074916e-01 7.07445621e-01
-4.65352207e-01 7.11168408e-01 4.83573526e-01 -5.01508653e-01
-9.38252151e-01 -1.36303401e+00 -1.33668065e-01 1.15530264e+00
-1.49997577e-01 3.45967487e-02 -8.90368998e-01 -1.45569324e+00
2.84524523e-02 7.36767471e-01 -8.79921854e-01 -4.59657967e-01
-3.75576198e-01 -8.11602592e-01 1.00122273e+00 6.39442503e-01
6.62209868e-01 -7.17978120e-01 1.19661234e-01 -1.63858950e-01
1.79878712e-01 -1.03732145e+00 -6.34177446e-01 4.04053897e-01
-6.95837021e-01 -8.53318751e-01 -8.60543668e-01 -6.66060686e-01
9.32142615e-01 -3.47500667e-02 1.06847179e+00 -3.02978843e-01
-7.85353556e-02 3.03220183e-01 -1.78776175e-01 -6.81815088e-01
-5.94789147e-01 -4.20262627e-02 2.95135856e-01 1.84532925e-01
-1.44073620e-01 -9.70609486e-01 -4.18428749e-01 3.65953058e-01
-1.05862737e+00 -4.99879539e-01 3.88230324e-01 1.16699505e+00
9.09537375e-01 -8.95906389e-02 8.05464625e-01 -1.23873472e+00
3.73233318e-01 -5.71340859e-01 -6.61055028e-01 4.32291269e-01
-5.73340297e-01 1.08278058e-01 1.30532527e+00 -4.97698843e-01
-8.14141333e-01 1.66243955e-01 -2.82012701e-01 -1.15953350e+00
1.63102541e-02 1.58696115e-01 -5.01983523e-01 -4.30168688e-01
1.10928440e+00 1.57664984e-01 -7.56831095e-02 -2.42422059e-01
6.17089331e-01 2.85866946e-01 6.23468935e-01 -5.34125090e-01
1.36780930e+00 7.24370599e-01 5.50348647e-02 -4.13783312e-01
-1.19088221e+00 -2.19654534e-02 -6.64591312e-01 -1.19700618e-01
7.72183537e-01 -1.11389041e+00 -3.72105092e-01 6.45475924e-01
-7.43197381e-01 -4.05103058e-01 -6.31681979e-01 2.78040946e-01
-5.03775895e-01 4.15083319e-01 -5.76283693e-01 -5.23082256e-01
-6.87625945e-01 -1.18345499e+00 6.63921535e-01 3.56071144e-02
2.93073207e-01 -1.38485539e+00 8.45828801e-02 5.48806071e-01
3.56403232e-01 7.42007792e-01 5.84816575e-01 -1.35490072e+00
-2.02854946e-01 -6.42658830e-01 2.62541980e-01 1.18410516e+00
1.43569767e-01 -2.80365676e-01 -1.37029612e+00 -6.32373571e-01
4.17146653e-01 -9.50278878e-01 1.02410030e+00 1.72905289e-02
1.41406643e+00 -5.60755849e-01 2.10312337e-01 1.20504594e+00
1.25878274e+00 -1.07855782e-01 4.99488443e-01 8.60924125e-02
1.08444989e+00 4.80296522e-01 5.55867404e-02 1.05026245e-01
1.64325088e-02 1.71740562e-01 8.05107236e-01 -1.59662321e-01
-5.04068434e-02 -1.92702323e-01 4.28845555e-01 8.72155547e-01
1.61549672e-01 -3.55858296e-01 -8.34279537e-01 3.71839285e-01
-1.61755514e+00 -1.24067807e+00 3.37014616e-01 2.13412952e+00
9.48594272e-01 2.62079477e-01 -1.83968812e-01 1.51282743e-01
6.61593020e-01 3.72936517e-01 -1.07798314e+00 -3.09972823e-01
-3.40625912e-01 2.16822997e-01 7.57463276e-01 4.22766209e-01
-1.58221900e+00 7.30645180e-01 5.50401306e+00 9.30872440e-01
-1.14630628e+00 3.17578167e-01 1.13792157e+00 -3.88955534e-01
-2.70157814e-01 -4.31086183e-01 -6.00382328e-01 2.86544383e-01
8.49814177e-01 2.90236389e-03 4.94498044e-01 1.26668179e+00
-3.70250612e-01 7.03118980e-01 -1.05548942e+00 6.84228361e-01
3.52766275e-01 -1.25893700e+00 -4.16736454e-02 -7.04297423e-02
1.33535218e+00 3.02424818e-01 6.40745401e-01 3.30350101e-01
8.05568218e-01 -1.19712889e+00 7.76551008e-01 1.86765134e-01
9.03493524e-01 -1.07313657e+00 6.01924539e-01 3.85206968e-01
-7.70315647e-01 -6.21563531e-02 -3.73986155e-01 6.21368825e-01
1.08277807e-02 5.71743011e-01 -4.91090775e-01 6.74496651e-01
5.03368676e-01 6.34881616e-01 -4.21755940e-01 4.58212972e-01
-5.32584429e-01 7.70154715e-01 -2.97677368e-02 5.68126202e-01
1.45998240e-01 -1.35389179e-01 6.88407600e-01 9.22116816e-01
-9.90595892e-02 -2.24616259e-01 2.80900955e-01 7.23986804e-01
-7.38037825e-01 -9.25676897e-02 -6.88640237e-01 7.87056386e-02
3.93361270e-01 1.23475993e+00 -3.11262935e-01 -4.65341173e-02
-2.81850517e-01 1.24141455e+00 5.82598388e-01 5.09805083e-01
-1.02830791e+00 -4.82841969e-01 5.71828485e-01 -4.08383727e-01
3.30278784e-01 -1.85507648e-02 -3.97749454e-01 -1.32910848e+00
1.77063316e-01 -1.18091631e+00 3.78016144e-01 -1.73628896e-01
-1.77569413e+00 9.10848796e-01 -4.35713977e-01 -1.27877152e+00
-2.02755496e-01 -5.88405311e-01 -6.35787070e-01 1.12431312e+00
-1.52288401e+00 -1.41924655e+00 -4.11226451e-02 9.84576225e-01
4.67291355e-01 -6.50392711e-01 1.04519522e+00 2.77573645e-01
-1.01039171e+00 1.27889609e+00 5.95079541e-01 6.95973039e-01
8.75609815e-01 -1.34074807e+00 7.10300922e-01 1.13409185e+00
3.17314506e-01 3.66451859e-01 3.54244918e-01 -4.75471258e-01
-1.16251826e+00 -1.59786797e+00 1.30525306e-01 -5.59668362e-01
7.75708556e-01 -4.58065420e-01 -1.00872064e+00 9.26012874e-01
1.24567248e-01 8.08696449e-01 1.14651155e+00 -2.78763026e-01
-9.96003091e-01 -1.80160969e-01 -1.33066142e+00 4.96613353e-01
9.45851624e-01 -6.94190383e-01 -3.70747626e-01 6.47016823e-01
7.20414937e-01 -5.86580932e-01 -1.10755742e+00 3.41520965e-01
2.70041794e-01 -8.17193568e-01 1.14792216e+00 -1.13433433e+00
4.29135531e-01 1.90493524e-01 -3.30962181e-01 -1.67421567e+00
-9.96747240e-02 -1.06266522e+00 -3.64812493e-01 1.29892516e+00
6.24177516e-01 -6.26727641e-01 8.63874614e-01 7.65897453e-01
-2.95242935e-01 -8.92530799e-01 -9.76289451e-01 -1.06571925e+00
6.91500425e-01 -3.02968353e-01 3.02079707e-01 1.21432543e+00
-3.98450464e-01 5.77015802e-02 -6.10144556e-01 3.29734266e-01
1.07728660e+00 -3.17497402e-01 5.85453689e-01 -1.03433239e+00
-4.24240053e-01 -7.20663145e-02 -2.89569557e-01 -3.64439458e-01
8.45143497e-01 -1.15199924e+00 -6.72257245e-02 -9.23385322e-01
-2.61364371e-01 -5.03363669e-01 -3.95471990e-01 5.86080253e-01
-3.90914857e-01 4.38786268e-01 1.64618120e-01 2.28095725e-01
-6.13987207e-01 7.67893136e-01 1.00858021e+00 -3.15689474e-01
4.75632735e-02 2.28029892e-01 -8.08457136e-01 9.42674756e-01
9.48373258e-01 -5.91828704e-01 -6.60943091e-01 -7.06310451e-01
9.51226279e-02 -1.76062927e-01 3.70722204e-01 -1.10420239e+00
1.10313021e-01 -4.86729182e-02 5.16376138e-01 1.00824818e-01
4.78878140e-01 -9.18868542e-01 -1.30982175e-01 4.49675500e-01
-6.22183383e-01 -5.87468520e-02 2.79676437e-01 9.64166641e-01
-1.96831271e-01 -2.35445991e-01 1.27342534e+00 -3.46719585e-02
-2.26476073e-01 8.00726771e-01 1.71591505e-01 7.55382121e-01
1.20611417e+00 4.49556947e-01 -5.32602787e-01 -3.83856922e-01
-8.81638050e-01 2.10313827e-01 3.79962444e-01 2.41445944e-01
5.59236765e-01 -1.50312078e+00 -8.90958607e-01 2.26333618e-01
-8.29989016e-02 1.00839049e-01 4.05748487e-01 4.18011606e-01
-4.71325934e-01 -3.78839150e-02 -1.22727051e-01 -2.04608649e-01
-1.03062892e+00 7.03500450e-01 6.04230464e-01 -4.13286030e-01
-3.14126521e-01 1.44387579e+00 1.12252109e-01 -6.08505189e-01
5.31364202e-01 4.17157978e-01 2.93899830e-02 -7.21914768e-02
5.50803483e-01 2.87937343e-01 1.07941225e-01 -7.59454012e-01
-1.59272954e-01 9.75269154e-02 -3.36233199e-01 -1.56084418e-01
1.33612835e+00 2.42965326e-01 1.40237138e-01 1.31771356e-01
1.62361217e+00 1.07137524e-01 -1.78675401e+00 -4.10584450e-01
-4.35510665e-01 -1.98028252e-01 -1.48260733e-02 -7.16972470e-01
-1.54238987e+00 1.05858433e+00 5.35990536e-01 1.07776903e-01
1.06261921e+00 -4.24918264e-01 8.24526668e-01 5.40736616e-01
7.58510828e-02 -9.59347010e-01 1.03743918e-01 5.65541089e-01
9.84160542e-01 -1.47794569e+00 -2.85396408e-02 -2.42194086e-01
-8.94678116e-01 8.87831390e-01 6.51919365e-01 -3.60455543e-01
8.13576519e-01 2.87351310e-01 2.26041615e-01 1.32803202e-01
-3.87200773e-01 3.97917926e-01 3.90387803e-01 6.12238824e-01
-1.05837263e-01 -2.01602295e-01 3.89432669e-01 7.32281089e-01
-1.21059217e-01 -5.63187718e-01 9.94527340e-02 8.45949650e-01
-4.33681086e-02 -1.08805442e+00 -3.00072253e-01 3.41836333e-01
-1.08128464e+00 -2.70153731e-01 -4.29094344e-01 5.50550759e-01
6.55425116e-02 8.58683407e-01 -4.34757501e-01 -5.43329120e-01
1.96651191e-01 1.95676729e-01 3.28582555e-01 -4.31909651e-01
-8.07210624e-01 -4.97350365e-01 -1.23875089e-01 -4.52967882e-01
-1.38310626e-01 -5.55956304e-01 -8.57633829e-01 -1.43345222e-01
-3.39770675e-01 1.18893959e-01 6.55991077e-01 9.26189184e-01
2.57229924e-01 5.90594172e-01 1.05650544e+00 -8.07651043e-01
-1.30541134e+00 -6.75835133e-01 -4.60709184e-01 5.71057558e-01
5.19955456e-01 -1.97612435e-01 -7.23274291e-01 1.85184702e-01] | [5.614206314086914, 7.904454708099365] |
55de8a8a-6b22-492f-9e8c-4993648a025c | exploiting-the-complementarity-of-2d-and-3d | 2304.02991 | null | https://arxiv.org/abs/2304.02991v1 | https://arxiv.org/pdf/2304.02991v1.pdf | Exploiting the Complementarity of 2D and 3D Networks to Address Domain-Shift in 3D Semantic Segmentation | 3D semantic segmentation is a critical task in many real-world applications, such as autonomous driving, robotics, and mixed reality. However, the task is extremely challenging due to ambiguities coming from the unstructured, sparse, and uncolored nature of the 3D point clouds. A possible solution is to combine the 3D information with others coming from sensors featuring a different modality, such as RGB cameras. Recent multi-modal 3D semantic segmentation networks exploit these modalities relying on two branches that process the 2D and 3D information independently, striving to maintain the strength of each modality. In this work, we first explain why this design choice is effective and then show how it can be improved to make the multi-modal semantic segmentation more robust to domain shift. Our surprisingly simple contribution achieves state-of-the-art performances on four popular multi-modal unsupervised domain adaptation benchmarks, as well as better results in a domain generalization scenario. | ['Luigi Di Stefano', 'Samuele Salti', 'Pierluigi Zama Ramirez', 'Adriano Cardace'] | 2023-04-06 | null | null | null | null | ['mixed-reality'] | ['computer-vision'] | [ 2.96718508e-01 6.90914169e-02 -2.10361853e-01 -3.34780514e-01
-6.41715348e-01 -8.47235322e-01 5.84451497e-01 1.60870567e-01
-4.39236909e-01 3.92590493e-01 -7.17758834e-02 -1.18641600e-01
-3.28393020e-02 -7.37792790e-01 -6.18948400e-01 -7.79963315e-01
3.09981287e-01 8.76327574e-01 6.95093453e-01 -3.38500470e-01
1.05338968e-01 5.23859501e-01 -1.78323865e+00 2.79456563e-02
8.32928777e-01 1.03104591e+00 2.89462417e-01 1.94617808e-01
-4.67063695e-01 1.45550609e-01 -2.34161243e-01 -1.41227007e-01
5.02672493e-01 -2.88621604e-01 -7.50113785e-01 4.31888431e-01
2.99871832e-01 5.68729602e-02 5.49495891e-02 1.23180962e+00
2.70007193e-01 2.74058402e-01 6.47792757e-01 -1.26102436e+00
-2.78050900e-01 1.72927454e-01 -7.93375075e-01 -2.73596406e-01
2.70031363e-01 -9.59880184e-03 7.37049699e-01 -4.03481543e-01
8.53853047e-01 1.29873884e+00 4.87828732e-01 6.35104239e-01
-1.28437364e+00 -4.31615055e-01 3.40197444e-01 1.29856303e-01
-1.07985783e+00 -1.34195730e-01 1.20728219e+00 -4.16612267e-01
7.01555192e-01 -7.49592036e-02 5.13969898e-01 1.07170510e+00
-2.45713353e-01 8.38179171e-01 1.45935190e+00 -2.72754580e-01
5.49352169e-01 1.53922692e-01 1.00306943e-01 3.47347409e-01
1.72383249e-01 -1.28849208e-01 -5.01868308e-01 1.67762414e-01
4.75503981e-01 -5.76071581e-03 -1.09905843e-02 -1.13150346e+00
-1.20127761e+00 7.70179391e-01 6.81517422e-01 3.46187741e-01
-2.41368935e-01 -2.41373464e-01 2.67449737e-01 2.50030100e-01
4.54941034e-01 4.44858253e-01 -6.01027668e-01 -5.06735705e-02
-8.51101279e-01 3.92630026e-02 5.13088822e-01 9.20678020e-01
1.09226060e+00 -2.03614056e-01 4.17098999e-01 9.02309060e-01
4.15641099e-01 6.42051101e-01 2.95118928e-01 -1.04245245e+00
3.23982060e-01 7.54854858e-01 -1.00829817e-01 -6.44493639e-01
-6.90295100e-01 -2.34515518e-01 -6.95560873e-01 5.43904066e-01
6.96766615e-01 1.58342376e-01 -1.44384265e+00 1.62703073e+00
6.22236907e-01 -1.64230037e-02 3.39004174e-02 1.13499820e+00
9.17425036e-01 1.91593632e-01 9.25560370e-02 2.44965672e-01
1.31696975e+00 -7.97718227e-01 -2.48537809e-01 -6.81832433e-01
4.48492229e-01 -4.52403128e-01 8.73204112e-01 3.14751565e-01
-5.54858565e-01 -4.57423657e-01 -1.09205282e+00 -3.36375624e-01
-7.31677413e-01 -3.34876567e-01 5.08031964e-01 7.99745917e-01
-6.10117376e-01 2.42621496e-01 -9.08069551e-01 -6.80393398e-01
5.61269701e-01 2.05786213e-01 -6.67010307e-01 -3.62809271e-01
-9.88537371e-01 8.62513602e-01 6.90678120e-01 -3.38275462e-01
-4.14037734e-01 -4.48544413e-01 -1.00026536e+00 -3.59718233e-01
6.31085753e-01 -7.30041027e-01 8.14891577e-01 -7.47965813e-01
-1.29612911e+00 1.35923445e+00 -2.63021048e-02 -2.95266807e-01
4.86493170e-01 -4.06021737e-02 -5.05048856e-02 2.35239685e-01
2.12782562e-01 9.34923768e-01 8.47325742e-01 -1.66136575e+00
-6.80365622e-01 -1.01189244e+00 1.88903928e-01 2.98421890e-01
-8.08274001e-02 -5.91904581e-01 -6.17153287e-01 -2.75864094e-01
6.87106311e-01 -1.32654154e+00 -3.51819873e-01 -6.84131309e-02
-3.55038702e-01 -5.18709458e-02 1.05295682e+00 -2.79203385e-01
2.48691618e-01 -2.31899667e+00 5.69611728e-01 2.35221237e-01
7.70701393e-02 8.94039348e-02 -1.82503127e-02 -1.06683597e-02
1.90352827e-01 -3.61765884e-02 -5.92517853e-01 -4.55428690e-01
1.84831217e-01 4.83921111e-01 -2.35260446e-02 4.21570152e-01
2.00565338e-01 7.67619967e-01 -8.75963807e-01 -4.85321373e-01
5.55171311e-01 4.25551325e-01 -2.99454808e-01 -1.79654002e-01
-4.78247076e-01 9.88982379e-01 -5.81221104e-01 6.77388608e-01
8.44700336e-01 -2.10207298e-01 -5.05172983e-02 -1.04004748e-01
6.70336857e-02 1.59521893e-01 -1.31909215e+00 2.36520004e+00
-2.48136982e-01 3.00607562e-01 1.93162128e-01 -1.23487699e+00
8.76002789e-01 -3.07672266e-02 7.54925132e-01 -8.97507370e-01
2.04566315e-01 3.97820681e-01 -3.40959400e-01 -2.82826364e-01
5.91168284e-01 -3.83438498e-01 -3.60407650e-01 2.25973278e-01
1.44297317e-01 -7.18734205e-01 1.14536926e-01 7.52563700e-02
7.73127854e-01 4.33582634e-01 6.79091290e-02 8.37510815e-05
2.80925214e-01 4.28930312e-01 6.08732104e-01 5.84914863e-01
-3.42379838e-01 8.23038161e-01 4.04683083e-01 -9.59182829e-02
-9.00600255e-01 -1.17656708e+00 -1.71694934e-01 7.75784791e-01
6.71587884e-01 1.34990454e-01 -5.15967906e-01 -9.01657164e-01
1.62940934e-01 6.38395488e-01 -5.32549500e-01 -2.61068583e-01
-3.41879129e-01 -6.10082984e-01 1.91951767e-01 4.93783593e-01
6.69811249e-01 -6.37142360e-01 -7.89181352e-01 -2.96342690e-02
-3.04983258e-01 -1.58562613e+00 2.15708673e-01 5.56958914e-01
-1.00178325e+00 -9.38561857e-01 -7.26812422e-01 -4.51081157e-01
3.81540507e-01 6.56104147e-01 1.07165623e+00 -4.08813208e-01
1.17550679e-01 5.04746854e-01 -6.40639067e-01 -3.90344739e-01
-2.40024433e-01 3.34765464e-01 -7.14876130e-02 -1.95415374e-02
6.86945558e-01 -6.81405485e-01 -1.85556680e-01 2.43957594e-01
-1.11883414e+00 -8.59436020e-03 4.64094818e-01 5.55079699e-01
9.09395218e-01 8.23092982e-02 3.15951377e-01 -9.29529786e-01
-1.03081046e-02 -3.62783909e-01 -5.02256691e-01 -1.29427090e-02
-2.66611159e-01 1.04092091e-01 3.11721712e-01 -1.97082102e-01
-8.82909060e-01 6.10506952e-01 -1.09359242e-01 -5.67789674e-01
-7.30201781e-01 2.39495963e-01 -4.46087033e-01 -8.08892623e-02
6.83096707e-01 1.91798694e-02 1.28908530e-01 -5.95488250e-01
5.99730313e-01 5.26657820e-01 4.07779783e-01 -4.81493294e-01
9.61505115e-01 9.20485079e-01 2.52466142e-01 -9.87514496e-01
-8.33236158e-01 -7.67605603e-01 -9.45847273e-01 -1.70762733e-01
1.17236984e+00 -8.58906388e-01 -1.45100385e-01 5.60876191e-01
-9.18566167e-01 -3.43480706e-01 -2.58330703e-01 3.72451067e-01
-6.52583301e-01 3.30473542e-01 -5.82974367e-02 -5.50833404e-01
2.00399071e-01 -1.30310893e+00 1.40904927e+00 3.14954191e-01
1.04655869e-01 -8.62444103e-01 -7.33910725e-02 6.84955239e-01
1.28129885e-01 4.33555007e-01 9.61750865e-01 -6.56005681e-01
-5.10827005e-01 -4.03342955e-02 -3.14558566e-01 3.33584398e-01
1.54851541e-01 -3.49183798e-01 -1.13614857e+00 -9.76613611e-02
-1.25548631e-01 -3.24017495e-01 1.14032090e+00 2.45002791e-01
8.56303453e-01 5.96649587e-01 -3.77122730e-01 5.26448429e-01
1.33394015e+00 -8.26898590e-02 4.29367781e-01 4.31057125e-01
9.47690129e-01 8.93806398e-01 7.33289897e-01 8.56948942e-02
5.79377174e-01 7.44350553e-01 8.52151752e-01 -1.05750605e-01
-1.59663260e-01 -1.60181839e-02 -1.04625002e-01 4.29754674e-01
1.14839770e-01 -2.50323806e-02 -1.10787392e+00 6.61703169e-01
-1.97259188e+00 -6.01484954e-01 -2.12197587e-01 2.21148467e+00
4.39076841e-01 3.33153099e-01 3.15632850e-01 2.90811777e-01
5.01230717e-01 2.05899909e-01 -7.75524974e-01 -1.23577036e-01
-5.11988163e-01 1.01035632e-01 7.81367898e-01 8.74264389e-02
-1.30208969e+00 8.15596104e-01 5.14405775e+00 6.23879552e-01
-1.12364912e+00 9.28692669e-02 3.52000207e-01 -3.80071066e-02
-3.92452538e-01 -8.82390812e-02 -5.76570570e-01 3.87115002e-01
4.45681781e-01 4.66487467e-01 1.86194211e-01 7.83141792e-01
-1.04380026e-01 -4.65824395e-01 -9.53972936e-01 1.07372940e+00
2.23360267e-02 -1.04009581e+00 -2.11010233e-01 1.44034937e-01
8.61649692e-01 4.11292583e-01 -5.99484816e-02 1.50953621e-01
2.67322391e-01 -6.50667310e-01 6.78561211e-01 2.54626065e-01
5.17177999e-01 -5.97054362e-01 5.16325355e-01 5.44969022e-01
-1.05074215e+00 -5.69765549e-03 -6.24611750e-02 1.34221628e-01
3.14089358e-01 8.47979188e-01 -3.61645609e-01 7.17717528e-01
1.07313335e+00 9.41357195e-01 -6.15160227e-01 9.88079071e-01
-2.26081967e-01 1.06739245e-01 -6.62557006e-01 2.18714088e-01
3.02048951e-01 -3.25080574e-01 8.01117599e-01 8.54219615e-01
1.40799582e-01 -1.21153206e-01 2.24351451e-01 7.44095385e-01
1.28335981e-02 -2.60955483e-01 -8.38411689e-01 3.69888060e-02
3.39986145e-01 1.15460193e+00 -1.10381281e+00 -1.42403319e-01
-5.63949406e-01 9.28442061e-01 1.19141832e-01 3.96342039e-01
-6.20386243e-01 -1.63754746e-01 7.49498963e-01 3.76285352e-02
5.06785691e-01 -6.89693093e-01 -7.72184551e-01 -1.18117952e+00
1.87731519e-01 -6.37092412e-01 4.23435539e-01 -7.72143304e-01
-1.23211086e+00 2.79048651e-01 1.17769510e-01 -1.36092138e+00
-3.42965089e-02 -7.44117856e-01 1.03887811e-01 6.35299563e-01
-1.85764647e+00 -1.30187345e+00 -4.72222209e-01 6.51556134e-01
3.14534605e-01 1.39337286e-01 5.13538480e-01 3.24607581e-01
-2.78199494e-01 4.05350588e-02 1.07330373e-02 -1.37101024e-01
7.92698264e-01 -1.39973247e+00 1.64968625e-01 7.96675742e-01
1.76926553e-01 1.06704272e-01 6.18679941e-01 -4.21980023e-01
-1.43613410e+00 -9.63630140e-01 4.54048067e-01 -4.88184273e-01
4.09633309e-01 -3.52189392e-01 -9.12806332e-01 3.74390364e-01
-2.66156673e-01 1.17074706e-01 4.27795023e-01 1.44903749e-01
-6.42375350e-01 -7.44328797e-02 -1.31276345e+00 4.35598016e-01
1.16141486e+00 -5.30243576e-01 -6.13903999e-01 1.70298785e-01
6.45119965e-01 -6.45365596e-01 -7.97889173e-01 5.73861361e-01
2.89561450e-01 -1.16248655e+00 1.12802231e+00 -3.37063521e-01
3.05725098e-01 -5.23424685e-01 -4.24218714e-01 -1.32394898e+00
-6.19348744e-03 -4.03835848e-02 1.09181069e-01 9.78822470e-01
2.86307365e-01 -6.99723184e-01 9.25432384e-01 4.83438700e-01
-2.65665591e-01 -1.71497002e-01 -1.26004326e+00 -8.23109508e-01
2.02946022e-01 -5.97959757e-01 5.87917089e-01 1.01327562e+00
-3.94987881e-01 5.75780988e-01 -7.34248459e-02 3.27927560e-01
7.00349271e-01 5.34756899e-01 8.91505718e-01 -1.71728933e+00
-4.03993875e-02 -5.03473878e-01 -7.32435465e-01 -1.28954458e+00
1.46761477e-01 -8.15645456e-01 1.90529346e-01 -1.61026061e+00
-4.11303714e-02 -7.09926903e-01 -2.44914234e-01 5.34664989e-01
5.25067225e-02 6.49769545e-01 3.85147899e-01 5.61914891e-02
-7.79442608e-01 6.05077028e-01 1.34587777e+00 -3.74084800e-01
-4.11069274e-01 -1.83048964e-01 -6.09913230e-01 6.99562192e-01
7.55908430e-01 -3.87435049e-01 -4.17986929e-01 -7.35599697e-01
1.85544014e-01 -1.58909202e-01 6.26052856e-01 -9.95182991e-01
1.01065673e-01 -2.07893535e-01 1.87493786e-01 -6.86784208e-01
7.52700508e-01 -1.13149714e+00 -9.12846476e-02 -2.19761506e-02
1.38173431e-01 -5.70949078e-01 2.61562258e-01 6.89891458e-01
-2.92917639e-01 -6.16970770e-02 1.00423265e+00 -2.60022134e-01
-1.27267301e+00 1.74606949e-01 3.60722304e-04 8.79838020e-02
1.03496361e+00 -5.88699102e-01 -2.08786413e-01 -1.47443846e-01
-7.89205194e-01 2.88481236e-01 9.24329042e-01 7.80414760e-01
3.15408766e-01 -1.15089953e+00 -3.46413165e-01 2.18774229e-01
5.46206474e-01 5.60024142e-01 3.26655775e-01 9.83677268e-01
-9.54386145e-02 3.02064478e-01 -4.75053906e-01 -1.14394617e+00
-9.85719860e-01 4.21629667e-01 3.57922018e-01 8.48381668e-02
-6.61204815e-01 5.82056761e-01 1.04745097e-01 -7.74645567e-01
9.60900262e-02 -2.25007907e-01 -1.35684595e-01 2.36868978e-01
8.60739201e-02 1.60031587e-01 3.56230408e-01 -9.78503346e-01
-6.85154438e-01 1.00163114e+00 2.98554629e-01 -1.23857826e-01
1.15251970e+00 -4.24508393e-01 5.67827113e-02 6.29709840e-01
1.06447697e+00 -3.19143116e-01 -1.22237730e+00 -4.51708317e-01
-6.90601692e-02 -3.25035483e-01 1.03133574e-01 -6.52482927e-01
-1.14994323e+00 9.46596384e-01 5.43758988e-01 3.81081700e-01
1.21706736e+00 3.44734430e-01 8.28633904e-01 2.26848543e-01
7.08117604e-01 -1.30363381e+00 -1.39142662e-01 6.07022405e-01
3.47136497e-01 -1.73755527e+00 2.85214721e-03 -4.38350499e-01
-6.62253678e-01 8.19624603e-01 3.78619462e-01 4.36474383e-02
6.73684895e-01 -1.74955234e-01 2.68175185e-01 -1.35225400e-01
-1.31076112e-01 -6.96719050e-01 2.57049561e-01 9.43055153e-01
-9.57001075e-02 6.95096329e-02 1.23034850e-01 2.82398999e-01
5.01859225e-02 -2.65581191e-01 2.41210103e-01 1.03476405e+00
-4.59303409e-01 -1.19587862e+00 -5.87241590e-01 2.76359349e-01
4.96121049e-02 4.74163204e-01 -6.29141748e-01 8.06450427e-01
2.95458734e-01 1.06173766e+00 5.81888929e-02 -3.69140387e-01
4.61429775e-01 1.34362802e-01 5.56933403e-01 -5.49813092e-01
-1.06608205e-01 1.00184575e-01 -1.06930815e-01 -6.29852235e-01
-9.34956968e-01 -8.06110799e-01 -1.52577055e+00 -4.96577658e-02
-9.96493502e-04 -2.42661595e-01 1.05310535e+00 1.16119432e+00
4.08179492e-01 4.56414759e-01 4.12488699e-01 -1.13978326e+00
-8.13840926e-02 -6.53418243e-01 -5.94578505e-01 7.49256849e-01
4.54018950e-01 -1.08623147e+00 -2.57499546e-01 9.69903916e-03] | [8.262799263000488, -2.49794864654541] |
0c90cb85-09fd-439d-87eb-ac88fe6ac18a | using-user-s-local-context-to-support-local | 2205.12408 | null | https://arxiv.org/abs/2205.12408v2 | https://arxiv.org/pdf/2205.12408v2.pdf | Using user's local context to support local news | American local newspapers have been experiencing a large loss of reader retention and business within the past 15 years due to the proliferation of online news sources. Local media companies are starting to shift from an advertising-supported business model to one based on subscriptions to mitigate this problem. With this subscription model, there is a need to increase user engagement and personalization, and recommender systems are one way for these news companies to accomplish this goal. However, using standard modeling approaches that focus on users' global preferences is not appropriate in this context because the local preferences of users exhibit some specific characteristics which do not necessarily match their long-term or global preferences in the news. Our research explores a localized session-based recommendation approach, using recommendations based on local news articles and articles pertaining to the different local news categories. Experiments performed on a news dataset from a local newspaper show that these local models, particularly certain categories of items, do indeed provide more accuracy and effectiveness for personalization which, in turn, may lead to more user engagement with local news content. | ['Bamshad Mobasher', 'Payam Pourashraf'] | 2022-05-24 | null | null | null | null | ['session-based-recommendations'] | ['miscellaneous'] | [-2.67892808e-01 -1.71932057e-02 -8.05989444e-01 -6.43620610e-01
-5.41769981e-01 -2.96066642e-01 7.43221760e-01 5.81512809e-01
-5.48108876e-01 5.36138058e-01 1.02738917e+00 -1.39164358e-01
-4.28297907e-01 -9.66058552e-01 -4.27868396e-01 -2.82181859e-01
4.09569651e-01 4.58517820e-01 4.25536215e-01 -6.60810590e-01
6.75291955e-01 2.72022456e-01 -1.50344241e+00 5.19771278e-01
7.44347632e-01 4.88276422e-01 3.28530163e-01 1.32080063e-01
-5.95791280e-01 4.02509004e-01 -3.66230905e-01 -4.35098290e-01
1.82328567e-01 -1.89978749e-01 -6.06532514e-01 -9.84047130e-02
3.27113122e-01 -2.78655946e-01 -6.17059544e-02 8.18180501e-01
2.51322001e-01 6.08802915e-01 2.79270142e-01 -3.54246676e-01
-5.81111073e-01 1.01241779e+00 -4.41921294e-01 4.03104633e-01
3.74480724e-01 -6.59721613e-01 1.35418153e+00 -4.90001976e-01
7.79018879e-01 9.78070796e-01 5.95831394e-01 7.92714134e-02
-1.40286541e+00 -3.18949670e-01 7.56897211e-01 -2.09040210e-01
-9.42812800e-01 -2.41283089e-01 7.14412093e-01 -3.15614432e-01
8.13014925e-01 5.75529277e-01 7.96995640e-01 1.10474694e+00
4.25172001e-01 3.02512974e-01 1.12410033e+00 -4.90777612e-01
2.15089083e-01 8.77210975e-01 5.01685858e-01 -3.86274785e-01
3.45721424e-01 -3.16725463e-01 -4.68495518e-01 -2.64200300e-01
3.68010670e-01 4.71725464e-01 -1.52310133e-01 -2.92989284e-01
-8.96681547e-01 1.10360360e+00 1.75494269e-01 8.12922478e-01
-6.51494622e-01 -5.80044448e-01 2.47427180e-01 5.28542697e-01
1.00045848e+00 9.18487966e-01 -4.49061513e-01 -2.95264482e-01
-8.64922464e-01 4.55065697e-01 1.15892994e+00 5.63742280e-01
5.46903253e-01 -4.34860349e-01 -9.85559523e-02 1.19320965e+00
4.61243629e-01 1.12856694e-01 5.56925833e-01 -4.90924239e-01
1.13597780e-01 5.34480989e-01 3.44844043e-01 -1.38748789e+00
-4.40459549e-01 -1.03457963e+00 4.67570610e-02 -1.87299624e-01
3.93002093e-01 -4.13212702e-02 -6.27101362e-01 1.57941461e+00
2.04412013e-01 -4.30201262e-01 -1.20879658e-01 6.52617514e-01
4.01849508e-01 9.23753738e-01 3.12269092e-01 -5.13685226e-01
1.21730256e+00 -7.70552278e-01 -7.59166658e-01 -4.34489757e-01
4.48352039e-01 -9.92293417e-01 1.07689786e+00 3.79757047e-01
-8.51959229e-01 -3.45791876e-01 -7.60067940e-01 3.43723476e-01
-5.08381844e-01 -3.35476786e-01 5.76695085e-01 7.68370926e-01
-6.93531334e-01 6.04019523e-01 -5.14871120e-01 -1.07906914e+00
-1.17831141e-01 5.91114685e-02 8.82309601e-02 -1.65852889e-01
-1.14146113e+00 1.01052451e+00 -7.53095374e-02 -2.72596955e-01
-6.55807704e-02 -5.47382534e-01 -7.46636316e-02 1.62345633e-01
2.34639823e-01 -3.82439375e-01 1.11474478e+00 -1.05027795e+00
-1.27287555e+00 3.05749655e-01 1.53719187e-01 -2.51286715e-01
1.15652695e-01 -1.79033726e-01 -9.80859041e-01 -4.23347771e-01
2.12456152e-01 8.81693065e-02 5.61668694e-01 -1.19187164e+00
-1.32219553e+00 -3.21172625e-01 2.41254225e-01 3.40302348e-01
-7.33469963e-01 2.80897439e-01 -2.70005465e-01 -4.95205760e-01
2.37327233e-01 -8.22219551e-01 -2.89801538e-01 -8.87248039e-01
2.88660198e-01 -2.91205376e-01 3.68732065e-01 -6.69654608e-01
1.44457304e+00 -1.98450196e+00 -3.31691951e-01 6.11333787e-01
-1.02689959e-01 -1.43961981e-01 2.61559665e-01 7.53858447e-01
4.10737514e-01 1.31285354e-01 7.00528979e-01 7.94997811e-02
-1.31743744e-01 7.66155422e-02 -4.36028302e-01 9.33792740e-02
-7.02271879e-01 1.65391505e-01 -8.30973208e-01 -2.92225629e-02
-2.96657950e-01 2.89603651e-01 -6.63794756e-01 -4.32311505e-01
-1.88201651e-01 2.96358258e-01 -7.08848238e-01 4.74565059e-01
3.75940591e-01 -2.86827713e-01 5.41719258e-01 1.08144671e-01
-4.84906882e-01 7.30793178e-01 -8.56331408e-01 1.21145582e+00
-5.30822635e-01 2.96436459e-01 2.32370514e-02 -8.01323354e-01
1.07501686e+00 1.44740239e-01 7.61804044e-01 -1.03192329e+00
1.40887290e-01 2.40089834e-01 6.54090643e-02 -3.06566983e-01
1.02077079e+00 -1.33617282e-01 -1.16313659e-01 4.66578096e-01
-4.05324876e-01 5.75977623e-01 2.32634917e-01 2.13502586e-01
7.64027238e-01 6.53611347e-02 1.41607270e-01 -5.66999435e-01
-5.19380383e-02 3.47818285e-01 6.36403680e-01 8.20023596e-01
1.29964054e-01 1.47181645e-01 1.73443258e-02 -2.74180025e-01
-9.37336206e-01 -5.96201599e-01 -2.74061322e-01 1.56782126e+00
9.25568864e-02 -5.67259908e-01 -2.79897422e-01 -5.02494454e-01
1.17868921e-02 1.27276218e+00 -4.31568682e-01 1.14354379e-02
-3.13870907e-01 -7.93511093e-01 -4.48037207e-01 2.03723013e-01
1.57687813e-01 -7.65027702e-01 -1.76146984e-01 7.92469442e-01
-2.87511110e-01 -5.52301109e-01 -5.94457805e-01 5.92943728e-02
-8.94777119e-01 -2.31123537e-01 -7.31374919e-01 -5.13552368e-01
6.48306727e-01 6.21797860e-01 9.07982528e-01 -3.68466645e-01
5.56808710e-01 6.57950282e-01 -8.45247805e-01 -4.93093282e-01
-3.41044992e-01 4.22118336e-01 1.03456257e-02 2.42005378e-01
6.32729530e-01 -6.13547504e-01 -6.09404266e-01 5.34009457e-01
-7.08876610e-01 -7.48982057e-02 6.43736362e-01 5.50649643e-01
1.89426824e-01 1.53158754e-01 1.01010454e+00 -1.34877872e+00
7.80012131e-01 -9.81087089e-01 -3.17501456e-01 1.39954701e-01
-1.37863564e+00 -3.74113679e-01 3.37388635e-01 -6.70508087e-01
-1.52257895e+00 -4.48254019e-01 -1.09702773e-01 6.82658613e-01
-1.19309239e-02 1.18619168e+00 1.86111376e-01 1.32766977e-01
1.01629496e+00 -1.84827462e-01 8.37228671e-02 -7.54384041e-01
-2.52400134e-02 8.00751150e-01 -5.04273809e-02 -3.21135759e-01
4.12372977e-01 3.66779506e-01 -7.29645371e-01 -7.02937663e-01
-9.85209286e-01 -8.04091334e-01 1.75172966e-02 -4.43306178e-01
5.89403033e-01 -7.47158945e-01 -1.43460467e-01 -2.04394907e-01
-4.36805069e-01 3.97388823e-02 -3.37584406e-01 9.96164382e-01
-2.05686279e-02 8.54844674e-02 -3.78346473e-01 -6.31501317e-01
1.37274712e-02 -7.55583107e-01 1.06881514e-01 4.59621966e-01
-5.01374364e-01 -1.05467105e+00 1.28706157e-01 7.23392010e-01
9.94590461e-01 -2.32755721e-01 8.07213247e-01 -1.19436455e+00
-2.98371971e-01 -6.98210657e-01 1.61953419e-01 1.62097216e-01
3.93420905e-01 -1.70058802e-01 -4.26917166e-01 -4.31246400e-01
-1.12876343e-02 4.72683340e-01 4.79591668e-01 4.90379989e-01
2.86287278e-01 -4.18788731e-01 -6.87166631e-01 -1.04123019e-01
1.32366705e+00 5.09300292e-01 5.25390565e-01 9.08231914e-01
1.85571939e-01 7.89877772e-01 9.41233337e-01 5.65039098e-01
4.57031071e-01 8.25215280e-01 4.43530940e-02 2.40001548e-02
2.78925180e-01 -3.89786571e-01 4.38659668e-01 8.13348591e-01
-2.53971606e-01 -2.17166454e-01 -6.06889546e-01 5.60714006e-01
-1.96384978e+00 -1.20895779e+00 -1.01598747e-01 2.45282292e+00
6.25442505e-01 5.36431313e-01 4.48284686e-01 -4.03745770e-01
7.36022711e-01 -5.94064072e-02 -4.91027460e-02 -6.86703265e-01
6.14331439e-02 -2.84203947e-01 8.50664020e-01 3.19153428e-01
-5.68651080e-01 5.54723024e-01 6.46476316e+00 4.83034551e-01
-1.12592661e+00 -8.24103057e-02 6.51302099e-01 -9.16531608e-02
-8.63034546e-01 4.08240676e-01 -1.24432051e+00 5.54202080e-01
1.13634515e+00 -5.40332556e-01 3.46408457e-01 9.86902893e-01
5.14535367e-01 -1.97781011e-01 -5.70478559e-01 2.90750414e-01
1.93288121e-02 -1.41782153e+00 -4.57225665e-02 4.20256019e-01
8.26546609e-01 -8.48728716e-02 2.18451414e-02 5.45497060e-01
2.81547815e-01 -3.99383694e-01 5.31491697e-01 7.65333831e-01
-1.15457751e-01 -7.92626500e-01 7.06659794e-01 5.12664855e-01
-4.94822413e-01 -3.66844475e-01 -4.20970768e-01 -1.82774097e-01
2.25089222e-01 6.56544864e-01 -7.84043670e-01 1.91205204e-01
1.02811551e+00 4.00726587e-01 -3.13380510e-01 1.02529716e+00
4.82056826e-01 7.79183209e-01 -2.26800978e-01 -2.36312017e-01
1.65338814e-01 -4.07852352e-01 6.09273076e-01 9.46092129e-01
5.35739064e-01 -1.24769118e-02 3.30628343e-02 1.87793508e-01
3.07636052e-01 7.70758927e-01 -2.59882361e-01 -1.65587559e-01
4.21129256e-01 1.36369514e+00 -8.38661313e-01 -1.76419914e-01
-1.03831291e+00 2.79586315e-01 -1.59432217e-01 2.23225415e-01
-5.78875959e-01 -1.41160190e-01 4.11489546e-01 1.02924109e+00
2.60731012e-01 -4.28737625e-02 -2.16270342e-01 -8.94846916e-01
-4.10399437e-01 -1.08630753e+00 4.00815010e-01 -2.73417950e-01
-1.24681413e+00 3.27039301e-01 2.17527837e-01 -1.05236638e+00
-1.57950036e-02 5.00407349e-03 -3.53329092e-01 6.13666832e-01
-1.09452391e+00 -1.08874762e+00 -2.14655232e-02 2.98512220e-01
4.77699876e-01 -2.03739852e-01 6.50315344e-01 5.56676507e-01
2.51951050e-02 4.24851120e-01 5.64693987e-01 -6.76569462e-01
1.15135801e+00 -8.69951367e-01 -9.39884260e-02 5.75996757e-01
6.71635196e-02 1.02774966e+00 1.13347983e+00 -8.99898589e-01
-1.04606104e+00 -6.08393908e-01 1.37919867e+00 -3.25193197e-01
5.76113641e-01 -2.36039162e-02 -8.36957753e-01 8.23820412e-01
2.74018437e-01 -8.87086272e-01 9.74309266e-01 9.64119315e-01
3.11741866e-02 -5.47757149e-01 -1.07032657e+00 7.57542074e-01
7.56404757e-01 -9.21485648e-02 -5.52927554e-01 4.69822049e-01
3.95754725e-01 3.57120385e-04 -1.09675324e+00 1.60816163e-01
7.09406674e-01 -6.72185242e-01 7.28707016e-01 -3.47289830e-01
5.68670481e-02 -8.39042142e-02 -2.41690859e-01 -1.40120423e+00
-9.38504636e-01 -3.96668792e-01 4.90387619e-01 1.46751451e+00
9.08497989e-01 -9.74036276e-01 7.74428725e-01 1.03537679e+00
-2.55254298e-01 -5.76201200e-01 -4.27163333e-01 -4.42616701e-01
-1.34364754e-01 -1.51964873e-01 5.31723142e-01 1.05647337e+00
3.98961604e-01 4.64431047e-01 -5.88064730e-01 -5.51923225e-03
7.27733001e-02 2.69337237e-01 8.59572589e-01 -1.52044487e+00
-4.88440752e-01 -4.86903191e-01 -8.02411959e-02 -1.11498487e+00
-4.29519683e-01 -6.14762366e-01 -2.20238760e-01 -1.76124597e+00
2.35061988e-01 -5.47898531e-01 -5.36351860e-01 7.86932334e-02
2.57924139e-01 -7.85954297e-03 -2.74120867e-01 5.58083475e-01
-3.94610673e-01 -1.83841884e-01 1.13037217e+00 2.12057233e-02
-6.72967970e-01 5.42558014e-01 -1.48582923e+00 3.90308440e-01
7.61662066e-01 -4.55996484e-01 -4.83105719e-01 -1.09317765e-01
9.21913207e-01 -8.81970152e-02 -2.89504200e-01 -7.04415739e-01
3.23992670e-01 -6.34992480e-01 1.75056100e-01 -4.26055610e-01
1.83042914e-01 -9.22913909e-01 7.07566619e-01 4.43532765e-02
-6.07128620e-01 -1.57194003e-01 -1.27948627e-01 7.34844387e-01
-1.78633034e-01 -2.13272154e-01 5.91967165e-01 7.91435391e-02
-6.40069127e-01 5.05442433e-02 -7.07870066e-01 -5.06703556e-01
9.92757738e-01 -4.24784511e-01 -3.01354498e-01 -8.33888590e-01
-1.08594370e+00 -4.02363651e-02 5.18266678e-01 6.82336509e-01
4.44288999e-02 -1.02060187e+00 -5.58991790e-01 -4.62663267e-03
2.98455417e-01 -7.31854022e-01 3.12856317e-01 1.05343676e+00
8.64672102e-03 4.09478337e-01 -2.67232537e-01 -2.45535709e-02
-1.14035475e+00 3.86424243e-01 -6.53474480e-02 -2.74267316e-01
-6.80305839e-01 6.45895958e-01 -5.61892688e-02 -3.89032625e-02
-1.47455968e-02 -7.39643201e-02 -8.25771093e-01 6.11188591e-01
3.06465507e-01 3.78563225e-01 9.37907696e-02 -5.95046103e-01
-8.89654309e-02 -2.86919274e-03 -7.89234996e-01 -1.67166442e-01
1.62453616e+00 -5.78662395e-01 1.81986079e-01 5.97266853e-01
7.05276728e-01 8.22089553e-01 -7.70708382e-01 -5.77400148e-01
1.97198927e-01 -7.65610099e-01 4.08894360e-01 -8.74396741e-01
-8.67463768e-01 -2.09800646e-01 3.76757443e-01 7.94201791e-01
8.57274830e-01 1.16939984e-01 7.06550181e-01 1.32218421e-01
3.05789977e-01 -1.73685420e+00 -3.79561901e-01 2.45436713e-01
6.60191417e-01 -8.07309151e-01 1.56953499e-01 -8.51134434e-02
-7.81902134e-01 9.10243750e-01 2.43564025e-01 7.11639673e-02
1.03455865e+00 -3.12321901e-01 9.02482718e-02 -4.73989081e-03
-7.67382681e-01 3.35012041e-02 1.04200035e-01 1.49715453e-01
8.26640904e-01 1.58265144e-01 -1.19015586e+00 7.53032923e-01
-7.49316067e-02 -2.25299776e-01 5.92777491e-01 1.00954545e+00
-8.53495598e-01 -1.50413632e+00 -3.09472591e-01 1.11021340e+00
-8.00181091e-01 1.31895170e-01 -4.11218479e-02 7.24877954e-01
-1.00194126e-01 1.15297449e+00 6.33830503e-02 -3.05503339e-01
3.67473125e-01 1.31698802e-01 4.97614667e-02 -7.50576079e-01
-6.95520759e-01 6.93048835e-01 6.69155836e-01 -2.01737285e-01
-4.99731272e-01 -1.14638793e+00 -7.76991189e-01 -6.27216041e-01
-7.00017512e-01 5.29169142e-01 9.76579309e-01 7.13707685e-01
4.92719501e-01 3.74325514e-01 7.48738170e-01 -3.05885166e-01
-6.09991312e-01 -9.46484506e-01 -9.47607398e-01 4.12847549e-01
-2.06830949e-01 -5.58926702e-01 -3.20430875e-01 -2.93594629e-01] | [9.993576049804688, 5.866909503936768] |
e5a13451-d79f-45b3-ac6a-b7e44f48e64d | adaptive-elastic-training-for-sparse-deep | 2110.07029 | null | https://arxiv.org/abs/2110.07029v1 | https://arxiv.org/pdf/2110.07029v1.pdf | Adaptive Elastic Training for Sparse Deep Learning on Heterogeneous Multi-GPU Servers | Motivated by extreme multi-label classification applications, we consider training deep learning models over sparse data in multi-GPU servers. The variance in the number of non-zero features across training batches and the intrinsic GPU heterogeneity combine to limit accuracy and increase the time to convergence. We address these challenges with Adaptive SGD, an adaptive elastic model averaging stochastic gradient descent algorithm for heterogeneous multi-GPUs that is characterized by dynamic scheduling, adaptive batch size scaling, and normalized model merging. Instead of statically partitioning batches to GPUs, batches are routed based on the relative processing speed. Batch size scaling assigns larger batches to the faster GPUs and smaller batches to the slower ones, with the goal to arrive at a steady state in which all the GPUs perform the same number of model updates. Normalized model merging computes optimal weights for every GPU based on the assigned batches such that the combined model achieves better accuracy. We show experimentally that Adaptive SGD outperforms four state-of-the-art solutions in time-to-accuracy and is scalable with the number of GPUs. | ['Alexander Sim', 'Kesheng Wu', 'Florin Rusu', 'Yujing Ma'] | 2021-10-13 | null | null | null | null | ['extreme-multi-label-classification'] | ['methodology'] | [-3.02179605e-01 -5.41141927e-01 8.53400975e-02 -6.09840810e-01
-7.99049556e-01 -2.94136077e-01 1.92795694e-01 6.80548370e-01
-6.91606700e-01 2.65307128e-01 -2.24258855e-01 -1.14867665e-01
1.32262617e-01 -6.07141256e-01 -3.70536268e-01 -7.06405222e-01
5.51629923e-02 9.55335140e-01 4.25245285e-01 2.46175662e-01
1.91459760e-01 4.26658511e-01 -1.58423209e+00 5.37110209e-01
4.96028632e-01 1.22242653e+00 1.96563512e-01 1.01449716e+00
-2.62306750e-01 7.71855056e-01 -1.81204721e-01 -1.86344028e-01
4.35896784e-01 -4.54914309e-02 -5.36126614e-01 1.08123541e-01
6.11041546e-01 -5.08458912e-01 2.20095739e-01 7.90943503e-01
7.53473818e-01 2.39354119e-01 5.26955605e-01 -1.27122486e+00
9.01225954e-02 1.62403509e-01 -1.00022542e+00 2.53333002e-01
-1.47120520e-01 -8.16664472e-02 8.71631682e-01 -7.42126465e-01
4.64612335e-01 9.62637901e-01 9.33936238e-01 5.01172066e-01
-1.40096736e+00 -4.70564902e-01 3.78427833e-01 -7.86751062e-02
-1.20007086e+00 -4.27750796e-01 7.13011771e-02 -5.62914610e-01
1.18896687e+00 3.63805145e-01 4.01092470e-01 4.68605638e-01
1.38202071e-01 5.32965064e-01 6.62068486e-01 -5.04091620e-01
6.87365234e-01 1.08275339e-01 7.11035669e-01 8.28842402e-01
2.74057359e-01 -5.54270566e-01 -7.94752121e-01 -9.02559936e-01
3.84598702e-01 -8.12122785e-03 2.61001438e-01 -3.15532446e-01
-7.93647349e-01 9.38790560e-01 -4.36110087e-02 3.82320993e-02
-7.05915630e-01 2.51179248e-01 1.03554726e+00 3.30265284e-01
9.31526959e-01 1.62422564e-02 -9.44197714e-01 -2.95829475e-01
-1.41094005e+00 2.92744279e-01 9.94418263e-01 7.35508204e-01
7.47536540e-01 -2.30635494e-01 -1.47473678e-01 9.45418656e-01
4.59289588e-02 8.66524652e-02 7.13774443e-01 -9.79739428e-01
3.81118953e-01 4.31877255e-01 5.24649955e-02 -4.60386246e-01
-7.67745554e-01 -6.25082731e-01 -8.09124231e-01 9.45380777e-02
3.69928986e-01 -6.60018444e-01 -5.37416160e-01 1.62618244e+00
1.00296640e+00 4.56649393e-01 -3.63663852e-01 7.94529736e-01
3.59830916e-01 5.38437963e-01 3.61961335e-01 -1.75187960e-01
1.32308090e+00 -1.38020158e+00 -9.79549251e-03 -2.02951953e-01
1.42440104e+00 -9.13859308e-01 9.60038424e-01 2.78253436e-01
-8.20578992e-01 -5.36988854e-01 -8.03905249e-01 -8.70344937e-02
1.17047988e-01 2.65875906e-01 6.33564353e-01 5.77687204e-01
-1.25601208e+00 7.88145721e-01 -1.46382761e+00 -2.53641218e-01
2.76918501e-01 5.24042666e-01 -1.00943834e-01 -9.00975093e-02
-2.46762097e-01 2.59009123e-01 2.50069350e-01 -2.69964367e-01
-3.55908334e-01 -1.07530642e+00 -3.80647212e-01 7.03949258e-02
-2.09195986e-01 -1.04333770e+00 1.39030421e+00 -1.06042433e+00
-1.45252717e+00 9.01493728e-01 -2.92196810e-01 -4.18450385e-01
8.50216269e-01 -1.37298614e-01 2.97455996e-01 -3.18990678e-01
-4.38307831e-03 4.13032562e-01 7.37328231e-01 -7.55941987e-01
-1.02960348e+00 -6.90391004e-01 -4.56515580e-01 5.81212223e-01
-7.53453612e-01 2.62929022e-01 -6.07935429e-01 -1.23243975e-02
2.56067216e-01 -1.21322381e+00 -4.67797846e-01 -1.32849157e-01
5.47233485e-02 -3.08087111e-01 7.37519026e-01 -2.59056389e-01
1.00826633e+00 -2.18465114e+00 4.47909981e-02 1.62164390e-01
5.55234969e-01 -3.15386355e-02 -9.09205079e-02 4.34782878e-02
2.60821074e-01 -5.74557543e-01 1.58087179e-01 -1.06367290e+00
-1.25769898e-01 1.58192873e-01 1.31647095e-01 5.95603287e-01
-3.01135987e-01 2.43667528e-01 -8.27282012e-01 -5.53637087e-01
-1.51068270e-01 2.95391798e-01 -9.43339944e-01 2.21912295e-01
-1.41961575e-01 2.79766917e-01 -5.72351992e-01 1.74873799e-01
9.91378546e-01 -7.75135100e-01 3.25619370e-01 7.46638104e-02
-1.07835874e-01 9.33585968e-03 -1.50758648e+00 1.36512589e+00
-7.78685391e-01 2.73618549e-01 2.39764586e-01 -7.37017334e-01
5.94252884e-01 5.30555099e-02 6.57924473e-01 -2.07600698e-01
2.07169250e-01 4.87694681e-01 -2.43105575e-01 -1.37098834e-01
6.19612455e-01 1.35470659e-01 2.36940578e-01 9.09939647e-01
-8.62686038e-02 -6.78460076e-02 3.26447725e-01 2.08112597e-01
1.10944843e+00 -4.93097827e-02 -8.76535103e-02 -4.44045246e-01
1.38749644e-01 6.61985800e-02 4.06793147e-01 9.15351391e-01
-1.39488885e-02 4.92007673e-01 4.34748620e-01 -9.51193869e-01
-1.25410628e+00 -5.22723615e-01 -1.59885347e-01 2.06660676e+00
-2.84484982e-01 -4.12375897e-01 -1.01329398e+00 -4.11670953e-01
6.98769316e-02 4.98799771e-01 -5.54010570e-01 1.58902153e-01
-4.23504770e-01 -1.38464189e+00 -8.16790015e-02 3.54062587e-01
2.24346012e-01 -6.35758460e-01 -9.07406986e-01 6.44955635e-01
2.48761445e-01 -1.17026556e+00 -5.33411562e-01 5.00217259e-01
-1.19146216e+00 -6.69561744e-01 -4.97630060e-01 -7.24374413e-01
9.61401343e-01 2.63766825e-01 1.25988317e+00 2.95126319e-01
-2.18361348e-01 1.31115004e-01 -1.19291581e-01 -2.53298670e-01
-2.90648580e-01 5.38421452e-01 1.57814771e-01 1.25770688e-01
3.37909013e-01 -4.72517371e-01 -5.80573022e-01 -4.03139330e-02
-8.02343309e-01 5.22153914e-01 -1.52354807e-01 7.58060694e-01
9.27947402e-01 -2.44163975e-01 -3.51312235e-02 -1.11467397e+00
6.95219755e-01 -6.13604724e-01 -6.89827442e-01 1.69998989e-01
-4.86194819e-01 5.03016002e-02 9.92455065e-01 -7.11178958e-01
-6.67501032e-01 2.69541442e-01 -2.27835268e-01 -4.07345027e-01
8.08369368e-02 1.61214381e-01 5.65488577e-01 -2.29444608e-01
7.29957163e-01 1.79789379e-01 -1.66084409e-01 -5.40339708e-01
9.04309973e-02 6.52067244e-01 -7.34750405e-02 -8.29890430e-01
-2.22948968e-01 4.04290229e-01 1.50275290e-01 -6.96493745e-01
-9.26052034e-01 -6.72189415e-01 -5.23475826e-01 -1.14572421e-01
6.07212663e-01 -1.08895397e+00 -6.58916414e-01 8.90223026e-01
-9.81632888e-01 -8.90718281e-01 -1.93014920e-01 4.21622336e-01
-4.60803717e-01 9.75684226e-02 -1.03932238e+00 -5.54178059e-01
-9.30250645e-01 -1.24582779e+00 1.33259499e+00 1.50975928e-01
-1.86269000e-01 -9.17157352e-01 2.32424721e-01 1.32978424e-01
5.01681507e-01 -1.06908984e-01 7.50548720e-01 -8.64918768e-01
2.89313607e-02 -4.63743657e-01 -3.59373212e-01 1.55290812e-01
-3.22076619e-01 1.12932837e-02 -8.72219324e-01 -6.11013055e-01
2.99076498e-01 -3.21434200e-01 6.18475795e-01 5.57666004e-01
1.22812045e+00 -1.19844995e-01 -4.04225022e-01 7.90112078e-01
1.59853077e+00 -3.62658054e-01 -1.98626637e-01 3.07876527e-01
8.28072786e-01 1.85349226e-01 2.43645579e-01 1.01916158e+00
3.89789820e-01 7.31951952e-01 1.43329054e-01 -2.58134939e-02
1.33320943e-01 4.14862335e-01 2.08201334e-02 1.06343555e+00
1.98697463e-01 -9.35685355e-03 -1.08482850e+00 2.61299998e-01
-2.07765079e+00 -3.07319254e-01 -5.02249658e-01 2.49287033e+00
5.96645892e-01 1.58289060e-01 2.21718833e-01 -3.47759575e-01
6.19093001e-01 -1.16315551e-01 -6.63736999e-01 -7.19466686e-01
2.91998029e-01 4.46821563e-02 6.34764433e-01 4.92342174e-01
-9.41702485e-01 8.92373323e-01 5.56144524e+00 9.91742074e-01
-1.36610568e+00 6.72704816e-01 1.19577157e+00 -8.20235014e-01
3.31194699e-01 -2.37693712e-01 -1.17110384e+00 6.38665617e-01
1.22013462e+00 -6.57298341e-02 4.37754393e-01 1.26517987e+00
6.42656460e-02 -2.63208091e-01 -1.15593541e+00 9.89214301e-01
-2.14615434e-01 -1.25966072e+00 -2.03234896e-01 3.50358710e-02
1.21512580e+00 9.59608197e-01 -1.08140402e-01 1.69383779e-01
5.90421975e-01 -1.13526493e-01 7.94663131e-01 1.52122527e-01
3.21826071e-01 -6.99736714e-01 6.82925463e-01 5.82333446e-01
-1.03829682e+00 -1.16484247e-01 -5.57329714e-01 -2.16779560e-01
-1.55165475e-02 1.03273571e+00 -5.59046984e-01 9.51588005e-02
6.73145175e-01 -6.30301656e-03 -4.69258785e-01 8.25391889e-01
6.87524259e-01 5.35481751e-01 -9.30344999e-01 -2.12433904e-01
4.03372735e-01 -2.66604513e-01 -3.67909856e-02 1.32019818e+00
2.57187396e-01 -4.87601943e-03 5.22794902e-01 7.21009001e-02
-8.82760137e-02 6.80443347e-01 4.42124419e-02 7.08077669e-01
3.28839988e-01 1.36690986e+00 -1.16198552e+00 -7.39183426e-01
-6.43025696e-01 1.12144744e+00 9.24140811e-01 -4.59894873e-02
-7.63692796e-01 1.59769312e-01 8.01921606e-01 7.78718367e-02
2.21902683e-01 -4.80147362e-01 -6.32921517e-01 -1.17054594e+00
-1.07956737e-01 -5.81979871e-01 6.24756157e-01 -2.87276745e-01
-1.39479935e+00 8.09612870e-01 -6.05278790e-01 -8.14246356e-01
-8.69032815e-02 -3.29755783e-01 -3.87431413e-01 8.10690820e-01
-1.09237289e+00 -8.19969058e-01 -3.33982557e-01 3.17533135e-01
6.13859475e-01 2.38349847e-02 1.03479290e+00 3.57817829e-01
-7.16976166e-01 7.93211877e-01 8.56279552e-01 -3.89629483e-01
5.71947694e-01 -9.13288176e-01 7.48723090e-01 4.79564518e-01
2.59119812e-02 1.46093845e-01 6.89448416e-01 -4.60610121e-01
-1.16413796e+00 -1.18817329e+00 8.64514291e-01 -9.65045318e-02
7.50943363e-01 -4.97443646e-01 -8.82462561e-01 5.03729403e-01
-3.89144778e-01 5.91519713e-01 9.12544429e-01 3.68456751e-01
-1.60021409e-01 -1.23476803e-01 -1.24474025e+00 4.95357990e-01
7.46768177e-01 -2.13320851e-01 4.99014348e-01 9.93242681e-01
2.91631043e-01 -8.18115175e-01 -6.94359720e-01 -3.97161804e-02
5.17989695e-01 -8.43772113e-01 4.74579066e-01 -7.83229947e-01
2.40109354e-01 4.75994460e-02 -2.26555094e-01 -1.12558866e+00
-5.26980400e-01 -2.81489193e-01 -8.05108249e-02 7.75111854e-01
2.89093435e-01 -6.78288400e-01 1.00097609e+00 1.14840615e+00
5.50586879e-02 -9.92622197e-01 -1.05461645e+00 -5.73511481e-01
-7.93098882e-02 -4.26590472e-01 5.43789387e-01 8.49681199e-01
-4.27686870e-01 2.87679762e-01 -1.86415210e-01 1.78084940e-01
8.10998678e-01 3.48445654e-01 8.00045788e-01 -1.26440680e+00
-7.36233830e-01 -3.41422141e-01 -1.78854302e-01 -8.93983781e-01
1.07741699e-01 -7.30576634e-01 -2.39725366e-01 -1.15122223e+00
5.28710008e-01 -8.24500382e-01 -2.52643973e-01 5.64251006e-01
-3.04833353e-01 4.43750471e-01 1.15934417e-01 3.74010593e-01
-1.02810121e+00 6.18357360e-02 6.88165843e-01 3.01682681e-01
-5.70907235e-01 1.93547234e-01 -2.06192508e-01 6.67308569e-01
6.83116257e-01 -8.23823035e-01 -5.44819161e-02 -9.61167991e-01
1.75704286e-01 4.22376543e-02 -3.73621553e-01 -1.21022797e+00
5.26429415e-01 1.43855497e-01 3.50480288e-01 -3.10545191e-02
1.96403816e-01 -6.17445409e-01 2.18025386e-01 5.94719589e-01
-3.91386241e-01 2.59768456e-01 3.07255328e-01 3.20205390e-01
2.38407403e-01 -3.15288275e-01 1.09464920e+00 -5.01595438e-03
-2.35764980e-01 5.50280571e-01 -3.55505764e-01 5.97948059e-02
1.02003729e+00 7.99707100e-02 8.75548422e-02 1.00730263e-01
-8.11148286e-01 2.04432219e-01 4.81563538e-01 1.16584085e-01
-1.99392736e-01 -1.12397325e+00 -9.48999047e-01 2.08048284e-01
-1.43788338e-01 4.30569761e-02 7.89255321e-01 8.47208977e-01
-7.70492077e-01 -8.79634842e-02 3.46969105e-02 -8.20692897e-01
-1.65654922e+00 2.87811071e-01 3.81239146e-01 -7.15589583e-01
-5.15075326e-01 1.18049300e+00 1.33364528e-01 -2.80557245e-01
1.01396620e-01 -9.25589446e-03 3.21806222e-01 9.26992819e-02
5.78313768e-01 5.44701517e-01 8.02435756e-01 -3.98809522e-01
-2.76133448e-01 4.22344536e-01 -5.98145783e-01 1.95252135e-01
1.63225186e+00 -1.17500955e-02 -3.39066267e-01 3.16223294e-01
1.45638239e+00 -2.72180587e-01 -1.35852849e+00 -3.03737402e-01
-1.66918650e-01 -3.13873142e-01 4.64424282e-01 -2.13789061e-01
-1.20564258e+00 4.61798728e-01 8.90057504e-01 1.58188537e-01
9.79630768e-01 -2.54920602e-01 9.32882667e-01 2.28916287e-01
6.10084295e-01 -1.32338500e+00 -2.72988796e-01 7.85797834e-01
2.06177030e-02 -1.19597280e+00 1.34141907e-01 -1.06777951e-01
-4.64371264e-01 1.09544504e+00 7.23556876e-01 -2.08623037e-01
8.04710448e-01 7.21289754e-01 7.42856739e-03 1.21021485e-02
-1.04528081e+00 3.69325668e-01 -2.06859514e-01 1.62462667e-02
4.91466999e-01 2.60690063e-01 -4.34963346e-01 3.96896571e-01
-1.21533081e-01 1.57609731e-01 1.42131463e-01 9.07631397e-01
-2.60552883e-01 -1.09163237e+00 -2.21213371e-01 1.04062724e+00
-5.54433048e-01 -2.26274505e-01 5.38289011e-01 -3.11715156e-02
5.70612028e-02 6.91340268e-01 5.58923781e-01 -1.41793489e-01
1.42412022e-01 1.24798782e-01 4.90006238e-01 -7.26428509e-01
-9.90216017e-01 -4.42667492e-02 -1.77279279e-01 -3.18680108e-01
3.71645331e-01 -7.46154189e-01 -1.18856490e+00 -6.83948576e-01
-3.27852756e-01 1.76266000e-01 1.21110153e+00 7.74393916e-01
8.29840660e-01 3.37077647e-01 6.64262295e-01 -1.14188373e+00
-8.09027374e-01 -8.26555550e-01 -6.45442903e-01 3.56329739e-01
2.10895166e-02 -1.96849152e-01 -3.32379282e-01 -1.87808648e-01] | [8.503183364868164, 3.427165985107422] |
3c65668a-c012-42a9-87ef-5311cdb75c37 | evaluating-translation-quality-and-clir | null | null | https://aclanthology.org/L16-1064 | https://aclanthology.org/L16-1064.pdf | Evaluating Translation Quality and CLIR Performance of Query Sessions | This paper presents the evaluation of the translation quality and Cross-Lingual Information Retrieval (CLIR) performance when using session information as the context of queries. The hypothesis is that previous queries provide context that helps to solve ambiguous translations in the current query. We tested several strategies on the TREC 2010 Session track dataset, which includes query reformulations grouped by generalization, specification, and drifting types. We study the Basque to English direction, evaluating both the translation quality and CLIR performance, with positive results in both cases. The results show that the quality of translation improved, reducing error rate by 12{\%} (HTER) when using session information, which improved CLIR results 5{\%} (nDCG). We also provide an analysis of the improvements across the three kinds of sessions: generalization, specification, and drifting. Translation quality improved in all three types (generalization, specification, and drifting), and CLIR improved for generalization and specification sessions, preserving the performance in drifting sessions. | ['I{\\~n}aki Alegria', 'Xabier Saralegi', 'Eneko Agirre'] | 2016-05-01 | evaluating-translation-quality-and-clir-1 | https://aclanthology.org/L16-1064 | https://aclanthology.org/L16-1064.pdf | lrec-2016-5 | ['cross-lingual-information-retrieval'] | ['natural-language-processing'] | [-1.32738769e-01 -4.71907467e-01 -4.62881207e-01 -2.16753483e-01
-1.67092204e+00 -1.32934570e+00 9.43543971e-01 4.42411721e-01
-9.37552512e-01 8.18291068e-01 5.80471039e-01 -5.64841807e-01
-5.76612830e-01 -3.91947389e-01 -5.30324578e-01 -4.89562511e-01
2.31247678e-01 7.57910430e-01 5.94623566e-01 -8.45796704e-01
4.69009191e-01 2.09781617e-01 -1.40264845e+00 8.34251523e-01
1.12069631e+00 4.84876931e-01 8.18967670e-02 4.59836662e-01
-3.65239680e-01 -9.61664617e-02 -7.85268188e-01 -5.54870665e-01
8.29718858e-02 -3.43643814e-01 -9.24621344e-01 -5.50471544e-01
3.46149743e-01 2.21062109e-01 1.46739423e-01 7.66424835e-01
8.45379233e-01 6.05067611e-02 5.08266091e-01 -8.83162260e-01
-7.02769637e-01 5.83633423e-01 -2.49459848e-01 4.44779664e-01
9.31810081e-01 -4.00011271e-01 1.01032364e+00 -9.83266413e-01
1.03604043e+00 1.19832492e+00 3.13184887e-01 3.50702941e-01
-1.23254657e+00 -5.64305663e-01 -3.49673163e-03 2.85233259e-02
-1.68759453e+00 -4.34883267e-01 1.49463667e-02 -2.98250794e-01
1.23103642e+00 7.44452417e-01 1.85684279e-01 8.44236672e-01
1.84133559e-01 5.72914720e-01 1.12042916e+00 -7.48833597e-01
7.22968578e-02 6.85103357e-01 3.27271819e-01 -1.05515793e-01
3.52645278e-01 1.59941778e-01 -7.05201983e-01 -4.16313767e-01
2.38089580e-02 -3.22612315e-01 -2.79034257e-01 2.82250881e-01
-1.11822867e+00 4.81718957e-01 2.24076107e-01 5.48172235e-01
-4.38015044e-01 -3.77654433e-01 3.83881867e-01 8.31980944e-01
6.79306030e-01 7.02234983e-01 -6.83505893e-01 -2.98528552e-01
-7.82675683e-01 6.01803362e-01 7.52624214e-01 1.27662480e+00
7.88569212e-01 -8.99320006e-01 -7.46365190e-01 1.13920963e+00
3.32913958e-02 1.01633775e+00 5.52926362e-01 -6.28105819e-01
7.78227568e-01 6.08140886e-01 3.68885309e-01 -5.15787840e-01
-3.94398242e-01 -7.24500656e-01 -9.64207500e-02 -6.56239510e-01
2.29744583e-01 -7.22019300e-02 -8.93045843e-01 1.64452040e+00
-1.01305433e-02 -1.04704475e+00 7.95126706e-02 6.44436181e-01
8.52832556e-01 7.56745398e-01 7.37111121e-02 -6.65777504e-01
1.42625034e+00 -6.64120734e-01 -1.11078191e+00 -6.57019541e-02
1.40726519e+00 -1.35906994e+00 1.40627563e+00 4.18991223e-02
-9.25465941e-01 -3.36875170e-01 -6.38619959e-01 -4.66171876e-02
-7.42294669e-01 -5.61167262e-02 5.45576438e-02 8.69527400e-01
-1.61477363e+00 1.72168285e-01 -3.71370614e-01 -7.64360309e-01
-6.97027504e-01 1.80488840e-01 -1.76242441e-01 -3.84728253e-01
-1.49974775e+00 9.96861339e-01 1.78802118e-01 -4.03649032e-01
-8.94063413e-02 -7.43534207e-01 -3.51884663e-01 1.02498438e-02
2.13061780e-01 -4.05910194e-01 1.16482878e+00 -3.60729545e-01
-8.66508484e-01 8.82069051e-01 -7.19620049e-01 -3.74388695e-02
4.07790184e-01 -2.51254141e-01 -7.10752368e-01 -3.42133522e-01
5.59762537e-01 5.14345765e-01 3.08852196e-02 -1.09736419e+00
-9.66099679e-01 -6.88845873e-01 2.73873080e-02 6.04522407e-01
-1.59195870e-01 4.57117081e-01 -8.13892961e-01 -5.16712785e-01
3.83094102e-01 -1.23728156e+00 3.97297353e-01 -7.76011705e-01
-1.99200045e-02 -7.77286887e-01 5.91086090e-01 -8.43657553e-01
2.01447940e+00 -2.14391279e+00 -6.68350533e-02 2.38387078e-01
-4.65628177e-01 -8.01797069e-05 -2.26907328e-01 1.00765014e+00
7.39998966e-02 6.22491539e-01 2.42768228e-01 -3.96531895e-02
5.06360084e-02 1.83618814e-01 -1.81485400e-01 -1.47872224e-01
-3.07535380e-01 1.03154206e+00 -6.09637856e-01 -3.82315695e-01
-4.25737888e-01 2.75308192e-01 -4.05770302e-01 -3.53951603e-01
-1.31209433e-01 8.22057277e-02 -1.95445761e-01 6.47498012e-01
3.77804965e-01 6.30411357e-02 -2.71436162e-02 1.74808994e-01
-3.99182141e-01 7.96175539e-01 -7.14309156e-01 1.72423911e+00
-4.02670979e-01 5.80624580e-01 -3.54634494e-01 -1.08541615e-01
8.12647820e-01 6.31366313e-01 3.36432517e-01 -1.65874279e+00
-5.24419248e-01 6.68932974e-01 -1.58628091e-01 -4.88820344e-01
8.70109200e-01 3.14005256e-01 -2.74869442e-01 7.29268014e-01
-5.21459162e-01 2.81044841e-03 7.15372980e-01 2.99502283e-01
7.82146811e-01 -6.35439828e-02 -1.67892233e-01 -5.76238930e-01
4.15298402e-01 3.67732942e-01 2.68024743e-01 1.05428028e+00
5.30813076e-03 3.26425433e-01 2.43940726e-01 -7.96213597e-02
-7.34018862e-01 -8.93201947e-01 -1.74707860e-01 1.66754067e+00
2.01569483e-01 -7.12463081e-01 -6.90182924e-01 -7.22508788e-01
-1.82128429e-01 1.07000351e+00 -4.07962441e-01 -3.44969600e-01
-5.01742661e-01 -8.26080441e-01 4.81236875e-01 1.20712571e-01
4.01542604e-01 -7.45397925e-01 -4.93303947e-02 -5.55008240e-02
-9.78986621e-01 -8.89841139e-01 -8.06147635e-01 -2.47483328e-02
-8.90227854e-01 -4.10387397e-01 -9.45673883e-01 -6.41488910e-01
8.29654261e-02 5.23127556e-01 1.30010045e+00 1.27452001e-01
1.44501865e-01 5.29962659e-01 -7.05940366e-01 -4.29888666e-01
-2.85079718e-01 3.70983750e-01 -1.25491887e-01 -7.09042609e-01
7.17467308e-01 4.43700179e-02 -4.87231553e-01 7.05510557e-01
-1.05451775e+00 -6.74502134e-01 5.66363215e-01 5.38790703e-01
4.84849483e-01 -1.76027238e-01 6.02183223e-01 -7.26007402e-01
1.30552030e+00 -3.73848081e-01 -3.24422032e-01 8.27058673e-01
-1.47330773e+00 2.55849928e-01 -2.15256840e-01 -1.16041332e-01
-9.67529655e-01 -5.14816403e-01 -1.91759609e-03 3.38093132e-01
6.15022443e-02 9.54392970e-01 2.42408499e-01 1.53468713e-01
1.07952142e+00 3.32314044e-01 -4.69860613e-01 -7.64915109e-01
7.87885860e-03 9.27448511e-01 -1.92078184e-02 -6.56511307e-01
4.62337464e-01 -2.80245915e-02 -5.77261090e-01 -6.93058074e-01
-3.31498891e-01 -9.62645650e-01 -3.94466460e-01 1.66705754e-02
7.80840278e-01 -7.78840423e-01 -2.64068991e-01 8.90640076e-03
-1.05494678e+00 1.11765772e-01 2.42236666e-02 6.42710268e-01
-1.68103293e-01 1.69868022e-01 -3.64974558e-01 -6.30550623e-01
-3.02778691e-01 -1.32673144e+00 1.39354348e+00 -1.41035591e-03
-4.70420212e-01 -9.47935998e-01 2.36645520e-01 5.11159360e-01
5.79899490e-01 -4.22320873e-01 1.32746327e+00 -8.96018028e-01
-2.68157512e-01 -3.92210305e-01 -1.12601154e-01 -2.58392125e-01
3.45131725e-01 -3.21486026e-01 -6.90925717e-01 -4.69597638e-01
-3.58648837e-01 1.34336218e-01 4.63981360e-01 2.58754283e-01
3.12810481e-01 -3.31672788e-01 -4.05284137e-01 -1.26797870e-01
1.24408901e+00 6.89261198e-01 6.92843974e-01 6.59191728e-01
-7.51555637e-02 6.67959630e-01 7.95662642e-01 -3.11266650e-02
3.42479765e-01 1.39348865e+00 -3.20369929e-01 3.02423965e-02
-1.40253663e-01 -1.06348678e-01 4.08108115e-01 5.66976666e-01
-8.82924646e-02 -5.06137669e-01 -8.90836716e-01 5.17975867e-01
-1.57924771e+00 -7.08546102e-01 2.32438166e-02 2.60078311e+00
8.50912750e-01 -4.47735190e-03 2.50122994e-01 -1.32581949e-01
5.93040645e-01 -3.46981406e-01 -5.08257374e-02 -5.67325950e-01
-2.73545712e-01 1.30172506e-01 5.16030669e-01 8.23117375e-01
-5.26704490e-01 1.00793827e+00 6.93263149e+00 6.88071012e-01
-1.04900908e+00 1.64218526e-02 1.91639289e-01 -7.85420686e-02
-6.95967436e-01 5.95035627e-02 -1.07537282e+00 4.15239930e-01
1.11434555e+00 -7.87566185e-01 2.85500377e-01 4.20425624e-01
3.43921989e-01 -2.60989279e-01 -1.08411145e+00 7.37439036e-01
2.31041778e-02 -1.11484468e+00 2.92627841e-01 1.45778358e-01
6.72739625e-01 1.27972096e-01 1.27965420e-01 6.66865885e-01
9.33360402e-03 -4.58084553e-01 4.25528347e-01 2.46422067e-01
6.40851140e-01 -5.94653249e-01 9.54788327e-01 5.22758842e-01
-8.40267181e-01 2.64548451e-01 -8.32989737e-02 2.52761692e-01
-1.48756891e-01 2.00796023e-01 -9.78118777e-01 1.14420104e+00
1.04905343e+00 4.55293685e-01 -9.58637178e-01 9.75073278e-01
1.17674865e-01 3.97909194e-01 -3.80118519e-01 -2.53019542e-01
3.08464468e-01 -1.50958851e-01 6.27971053e-01 1.34933162e+00
4.08667356e-01 -1.55976266e-01 7.35067427e-02 3.05951715e-01
6.96374327e-02 4.40046519e-01 -5.42000055e-01 2.20021289e-02
7.84961045e-01 4.85684812e-01 -5.99664092e-01 -4.59702700e-01
-1.89346388e-01 9.57927465e-01 -3.39989185e-01 7.44873643e-01
-4.37548429e-01 -5.63348413e-01 3.96270543e-01 2.51956433e-01
-1.96420595e-01 5.97419553e-02 -3.53005350e-01 -8.72056663e-01
5.33109307e-01 -9.54724073e-01 8.97141099e-01 -7.19628036e-01
-7.20444024e-01 6.27904177e-01 4.49867606e-01 -1.08747995e+00
-6.13345861e-01 -7.53627345e-02 1.57971084e-01 1.32616508e+00
-1.34241867e+00 -5.14081597e-01 1.23961650e-01 5.24384260e-01
6.22282088e-01 1.60176754e-01 9.70932543e-01 8.24411809e-01
-1.26321465e-01 1.08744442e+00 4.19018000e-01 -2.02057451e-01
1.35902858e+00 -9.22554851e-01 -1.56549774e-02 7.75895357e-01
1.74194440e-01 1.03384125e+00 7.65878677e-01 -7.75955915e-01
-1.29241502e+00 -8.02883208e-01 1.93914711e+00 -6.01783097e-01
2.18414649e-01 -1.75792083e-01 -8.47831130e-01 5.55101693e-01
3.29159796e-01 -8.34178925e-01 4.82603282e-01 6.71540320e-01
-4.45281327e-01 -2.47019917e-01 -1.07741451e+00 7.09525108e-01
8.77609849e-01 -8.15658867e-01 -4.92261827e-01 7.45095015e-01
1.14885235e+00 -2.94093698e-01 -9.60367858e-01 4.42207277e-01
6.27657533e-01 -6.29761755e-01 9.83307779e-01 -5.20440340e-01
8.16220045e-03 -1.41600594e-01 -3.73588204e-01 -1.22123706e+00
-4.60120231e-01 -3.75275671e-01 7.47724950e-01 1.06549954e+00
1.26064622e+00 -7.75947571e-01 3.24417114e-01 7.28577971e-01
-3.50365430e-01 -3.62673640e-01 -9.67051566e-01 -1.00571680e+00
3.55015874e-01 -1.39949948e-01 6.17553294e-01 7.52488434e-01
1.42733350e-01 5.48379183e-01 1.38725668e-01 -4.92822789e-02
1.89907961e-02 -1.03964582e-01 4.24084067e-01 -8.63799334e-01
5.69763556e-02 -6.17267907e-01 -6.59773424e-02 -9.49836314e-01
-4.25205231e-01 -1.06776226e+00 -1.78974628e-01 -1.77240729e+00
7.50069246e-02 -3.32721561e-01 -2.02758998e-01 3.29420984e-01
-3.74168098e-01 7.81816691e-02 -1.15011614e-02 5.91919601e-01
-4.52188641e-01 -3.69877741e-02 9.47325706e-01 -8.20001215e-02
-6.07123315e-01 2.82481045e-01 -8.27899337e-01 -2.70708114e-01
4.02745843e-01 -5.87028861e-01 -5.43530881e-01 -9.17291760e-01
6.01705015e-01 5.76147288e-02 -1.63465351e-01 -4.64419574e-01
4.17261571e-01 8.74518454e-02 2.61665117e-02 -8.87970626e-01
7.30530769e-02 -6.35441363e-01 1.89080775e-01 5.47059596e-01
-8.76236260e-01 8.24198484e-01 4.21590477e-01 1.75000593e-01
-4.20763671e-01 1.00118034e-01 3.24601442e-01 -2.00686092e-03
-5.42492211e-01 -1.51646629e-01 -3.87620717e-01 2.55015016e-01
4.64151561e-01 -9.70623866e-02 -4.02931511e-01 -6.43516958e-01
-8.91391814e-01 4.10638839e-01 2.99541801e-01 9.04165804e-01
1.69901639e-01 -1.31747746e+00 -9.87974882e-01 1.14892095e-01
8.36617529e-01 -6.52073205e-01 -1.18949533e-01 1.02344596e+00
-8.38362649e-02 1.20886743e+00 2.59557277e-01 -8.22300375e-01
-1.40703261e+00 2.32041985e-01 1.54374510e-01 -4.17038143e-01
-1.00198332e-02 7.29937494e-01 -1.21647734e-02 -6.25348151e-01
4.95641917e-01 -1.55876383e-01 -2.21349388e-01 3.31900924e-01
5.94490469e-01 4.93388951e-01 9.19148028e-01 -2.86336303e-01
-5.10822952e-01 6.84626997e-01 -4.92785990e-01 -7.88064659e-01
5.50384998e-01 -7.02278972e-01 -3.60116452e-01 5.13250113e-01
1.28463101e+00 1.57670118e-02 1.20286196e-01 -6.51533425e-01
8.21326196e-01 -3.70245188e-01 -2.84661241e-02 -1.59280658e+00
-3.87698978e-01 4.56383646e-01 9.49849188e-01 3.01931739e-01
1.17691219e+00 6.30189702e-02 3.71466815e-01 6.65189862e-01
5.63596427e-01 -1.18132603e+00 -5.26972830e-01 7.22775578e-01
9.65797305e-01 -1.05738282e+00 -3.48102927e-01 -2.27846369e-01
-5.37194967e-01 6.65229261e-01 1.45884737e-01 6.75279737e-01
5.52761793e-01 -2.80090511e-01 4.22570050e-01 -2.04350680e-01
-7.58744299e-01 -1.25125080e-01 6.14152193e-01 2.78652698e-01
1.01724052e+00 -2.54921499e-03 -1.24622393e+00 1.15769438e-01
-3.15372616e-01 -2.24400148e-01 1.49092078e-02 1.08593869e+00
-3.78113747e-01 -1.29879820e+00 -3.93417209e-01 2.47170895e-01
-5.53789079e-01 -4.64442223e-01 -9.86352801e-01 8.92982244e-01
-2.72078007e-01 1.34734702e+00 4.25360203e-02 -4.52345669e-01
7.93219805e-01 5.31173468e-01 1.73755094e-01 -5.79130113e-01
-9.49606180e-01 5.88330507e-01 4.40180123e-01 -6.11575723e-01
-2.47424364e-01 -7.27446616e-01 -8.18248630e-01 -2.59562165e-01
-4.67333674e-01 1.06265509e+00 9.66700971e-01 8.91542733e-01
6.44284248e-01 1.01553269e-01 7.45753646e-01 1.42803922e-01
-4.98755157e-01 -1.19602847e+00 -1.88723668e-01 4.42022383e-01
2.24451438e-01 -3.49566013e-01 -4.66687083e-01 -1.61442786e-01] | [11.516316413879395, 9.840658187866211] |
e12fc9e9-57b0-4b87-866a-ff1755f63306 | neural-complexity-statistical-mechanical | 2303.03128 | null | https://arxiv.org/abs/2303.03128v1 | https://arxiv.org/pdf/2303.03128v1.pdf | Neural complexity -- Statistical-mechanical approach of human electroencephalograms | The brain is a complex system whose understanding enables potentially deeper approaches to mental phenomena. Dynamics of wide classes of complex systems have been satisfactorily described within $q$-statistics, a current generalization of Boltzmann-Gibbs (BG) statistics. Here, we study human electroencephalograms of typical human adults (EEG), very specifically their inter-occurrence times across an arbitrarily chosen threshold of the signal (observed, for instance, at the midparietal location in scalp). The distributions of these inter-occurrence times differ from those usually emerging within BG statistical mechanics. They are instead well approached within the $q$-statistical theory, based on non-additive entropies characterized by the index $q$. The present method points towards a suitable tool for quantitatively accessing brain complexity, thus potentially opening useful studies of the properties of both typical and altered brain physiology. | ['Henrique Santos Lima', 'Constantino Tsallis', 'Dimitri Marques Abramov'] | 2023-03-06 | null | null | null | null | ['eeg', 'eeg'] | ['methodology', 'time-series'] | [ 8.55584219e-02 -2.66654268e-02 3.95414084e-01 -2.41182759e-01
-2.93824971e-01 -3.25532228e-01 7.19567537e-01 4.81418908e-01
-7.94602573e-01 8.94370914e-01 -2.88399607e-01 1.23008139e-01
-4.78403449e-01 -5.53515255e-01 -4.62607950e-01 -1.09836197e+00
-7.96840310e-01 2.29882374e-01 9.73139107e-02 -8.81902650e-02
4.66749430e-01 2.79983878e-01 -1.64040458e+00 -2.40331382e-01
8.91326666e-01 1.02517176e+00 1.17810920e-01 5.56942761e-01
2.60058612e-01 -2.36577392e-02 -5.61102748e-01 -2.37390772e-01
-1.53032795e-01 -6.40145540e-01 -5.43149650e-01 -3.14764917e-01
-1.38295412e-01 3.80792379e-01 -7.37168640e-02 1.38732791e+00
3.68443161e-01 3.17196995e-01 1.03743446e+00 -9.46499050e-01
-2.72758514e-01 4.49964553e-01 -4.46975172e-01 8.75942528e-01
5.42591631e-01 9.95991975e-02 8.22588861e-01 -6.09055758e-01
2.72754520e-01 7.75205076e-01 3.92741531e-01 3.00646514e-01
-1.53270364e+00 -5.69038033e-01 -2.85347492e-01 5.65816283e-01
-1.79563475e+00 -3.57193917e-01 5.41540027e-01 -6.89345717e-01
1.01825261e+00 1.30489348e-02 1.03764749e+00 9.18040574e-01
1.00366974e+00 1.06741965e-01 1.61393940e+00 -2.70304531e-01
7.45544076e-01 4.26981971e-03 3.16227734e-01 3.83152872e-01
4.21900898e-01 1.14846237e-01 -8.28069568e-01 -1.88140720e-01
5.05203605e-01 -3.77939552e-01 -4.93848145e-01 -1.68655157e-01
-1.26592839e+00 3.60005260e-01 2.56377086e-02 8.56472433e-01
-5.28848827e-01 3.46082836e-01 1.60436198e-01 1.63749997e-02
2.00246707e-01 2.33127922e-01 -4.59623575e-01 -4.95702088e-01
-9.58557844e-01 2.52116561e-01 5.96137345e-01 7.12310195e-01
6.47081673e-01 -1.02691278e-01 9.09020975e-02 4.28352386e-01
3.07003319e-01 6.79248333e-01 5.83195210e-01 -3.66902888e-01
6.76880479e-02 3.35266069e-02 8.57275799e-02 -6.73027337e-01
-7.60313213e-01 -4.88198668e-01 -8.84369969e-01 4.42704894e-02
7.00367510e-01 7.36261606e-02 -5.46834946e-01 1.88916874e+00
-1.09952882e-01 1.33875906e-01 -3.18899304e-01 4.13405597e-01
1.90863162e-01 3.09700966e-01 3.51203442e-01 -5.75621367e-01
1.53702736e+00 3.30714494e-01 -7.22904146e-01 5.50537445e-02
1.04511231e-01 -1.12058714e-01 9.42893744e-01 8.02482665e-01
-1.22707427e+00 -2.91996449e-01 -9.02536213e-01 6.12381399e-01
-5.17484248e-01 -6.24594808e-01 3.39963913e-01 8.50527883e-01
-1.25828207e+00 1.00981677e+00 -8.87960136e-01 -3.44508111e-01
3.14969271e-01 3.59387785e-01 -3.42623562e-01 2.12372497e-01
-1.23115659e+00 1.15393209e+00 4.48468924e-01 2.10668683e-01
-7.00528562e-01 -4.50209618e-01 -4.98773932e-01 8.46984312e-02
-1.53211087e-01 -3.51246804e-01 8.82266581e-01 -4.43866313e-01
-1.45400536e+00 9.25236285e-01 -2.36087531e-01 -1.75173759e-01
5.76717742e-02 1.77893698e-01 -5.26589453e-01 2.51513541e-01
-2.54238635e-01 1.52763188e-01 9.23257172e-01 -8.57978344e-01
6.54070377e-02 -6.36992037e-01 -3.82951230e-01 -9.84677151e-02
1.28161758e-01 -9.97755229e-02 4.31708425e-01 -2.86242753e-01
4.15897489e-01 -7.15330124e-01 -9.08999294e-02 -7.51209736e-01
-2.83643156e-01 -4.84562635e-01 -2.76543498e-01 -4.68707174e-01
1.27338624e+00 -2.15839696e+00 2.47986972e-01 5.67487419e-01
3.13363284e-01 -2.91411877e-01 2.39627078e-01 5.51092863e-01
-3.49542052e-01 7.75540844e-02 -4.22293305e-01 2.62138188e-01
3.52634579e-01 -1.82831421e-01 1.14989340e-01 1.03716290e+00
1.14797950e-02 6.92460418e-01 -9.06769216e-01 -1.87280014e-01
4.56145667e-02 4.53901887e-01 -3.77960682e-01 -1.29684746e-01
1.50566176e-01 4.71362054e-01 -3.06310385e-01 2.59000421e-01
5.10905921e-01 3.35304402e-02 -7.73261413e-02 1.15737021e-01
-2.45870575e-01 1.45652190e-01 -9.37188923e-01 1.33343887e+00
-5.64979762e-02 6.95176899e-01 -1.69323996e-01 -1.26248133e+00
7.60352790e-01 4.81682748e-01 3.81640762e-01 -7.07095623e-01
3.93351257e-01 2.58048892e-01 6.92233145e-01 -4.62976515e-01
4.99805324e-02 -4.90071505e-01 -4.85250540e-02 4.26110536e-01
2.93180883e-01 -4.35061634e-01 2.50623018e-01 -2.04713702e-01
1.20787430e+00 -2.23238871e-01 6.94247305e-01 -1.07737207e+00
5.02538443e-01 -7.68742859e-01 4.89631779e-02 6.63353264e-01
-5.96416831e-01 2.27561519e-01 7.73156285e-01 7.60278106e-02
-7.39872336e-01 -1.44406366e+00 -8.59477460e-01 6.32148325e-01
1.23647131e-01 -3.78875911e-01 -1.16723657e+00 3.13122392e-01
-2.14773476e-01 6.77104294e-01 -7.53352404e-01 -5.40433109e-01
-2.93371022e-01 -1.27897120e+00 2.85937905e-01 5.42868488e-02
3.01222801e-01 -1.44652033e+00 -1.03013110e+00 3.41507167e-01
4.37739231e-02 -9.63266075e-01 1.09548710e-01 6.09276235e-01
-9.41073179e-01 -7.07741082e-01 -6.12723768e-01 -3.95367146e-02
3.66102219e-01 -3.70207131e-01 1.14566994e+00 -1.61507592e-01
-7.60464013e-01 6.57699704e-01 -1.10534035e-01 -4.59167659e-01
-9.63158440e-03 -1.70663804e-01 4.10567373e-01 -6.27299696e-02
6.78371251e-01 -1.01161277e+00 -8.78260911e-01 -4.09907289e-03
-7.84676790e-01 -7.71408081e-01 3.71864051e-01 3.79950523e-01
3.20412844e-01 3.44752610e-01 6.71044052e-01 -1.36316046e-01
7.94360161e-01 -5.84434390e-01 -4.01733041e-01 -1.47479907e-01
-5.62008500e-01 7.15161264e-02 6.16115808e-01 -2.28989780e-01
-5.81644297e-01 -5.85942149e-01 1.05473481e-01 4.64758396e-01
-5.75145543e-01 4.87236768e-01 -1.49010286e-01 -2.04163697e-02
5.23686588e-01 5.83653152e-01 -4.15214896e-01 3.39824781e-02
-1.30350108e-03 4.67756391e-01 3.41892213e-01 -6.04908526e-01
4.38396156e-01 3.46805215e-01 1.91909268e-01 -1.29684329e+00
-2.07619146e-01 -2.99071342e-01 -6.75923586e-01 -3.71400803e-01
1.01090050e+00 -2.60991722e-01 -1.25834846e+00 5.04736185e-01
-9.09090936e-01 -2.25614518e-01 -2.26847768e-01 8.51565301e-01
-9.94196951e-01 3.05995017e-01 -5.07067263e-01 -1.34537148e+00
2.76509114e-02 -8.43890727e-01 8.26771438e-01 2.55212694e-01
-5.75227678e-01 -1.06468213e+00 3.62309664e-01 -1.49683550e-01
2.70317614e-01 2.27000996e-01 1.25898159e+00 -4.68015999e-01
-2.39194006e-01 -2.88766652e-01 1.89224809e-01 1.00791998e-01
-2.85461217e-01 3.23974825e-02 -1.04416609e+00 -2.19617382e-01
2.96015799e-01 1.16791174e-01 6.24297500e-01 7.51439035e-01
1.12194586e+00 3.87097001e-01 -1.56462267e-01 1.21358477e-01
1.45856285e+00 5.27933955e-01 7.67362654e-01 9.37115923e-02
-1.67683989e-01 7.54268289e-01 2.80479225e-03 7.99374223e-01
-3.25568542e-02 2.81125695e-01 3.62816870e-01 6.66020751e-01
7.20280826e-01 3.45101714e-01 3.63590777e-01 8.98991704e-01
-6.07928991e-01 -3.99043970e-02 -9.88854110e-01 5.01853049e-01
-1.29845476e+00 -1.25927901e+00 -2.37400681e-01 2.61858702e+00
7.94791758e-01 4.40365523e-01 1.50124982e-01 3.20712537e-01
6.17612064e-01 1.24043055e-01 -4.10489947e-01 -4.58122820e-01
3.20077799e-02 9.15091753e-01 3.61353904e-01 3.60213697e-01
-4.98603046e-01 3.41780692e-01 7.63330030e+00 6.29576325e-01
-7.72444129e-01 1.45900384e-01 2.03082129e-01 -1.73506185e-01
-1.07063316e-01 -2.64913887e-01 -5.52019715e-01 7.62336850e-01
1.54633892e+00 -5.83954930e-01 6.90881312e-01 2.68818438e-01
4.16536033e-01 -8.09410691e-01 -1.17337513e+00 1.22386742e+00
-1.21458329e-01 -5.79233050e-01 -4.47067946e-01 4.27340686e-01
3.53847295e-01 -6.87059537e-02 2.14102283e-01 1.02384418e-01
-2.07791880e-01 -1.37653768e+00 5.96246958e-01 9.43612874e-01
7.73771286e-01 -6.89830065e-01 5.31690061e-01 4.56537664e-01
-9.73927855e-01 1.14569247e-01 -4.22902346e-01 -9.63966399e-02
6.39801249e-02 1.00403237e+00 -1.34286702e-01 1.47649541e-01
7.85521090e-01 4.36468601e-01 -3.73119146e-01 1.08948553e+00
1.09216742e-01 3.86130363e-01 -4.16637748e-01 -4.34226692e-01
-1.46316560e-02 -6.47829652e-01 5.22545576e-01 1.10202491e+00
5.94425499e-01 4.46215242e-01 -6.87872469e-01 1.07643247e+00
4.37827408e-01 1.45394504e-01 -5.47457099e-01 -3.24543953e-01
2.94698775e-01 1.16265726e+00 -1.09104300e+00 -1.67217851e-01
-8.80381018e-02 6.18356943e-01 8.32227990e-02 3.14163476e-01
-7.05466211e-01 -5.48011363e-01 5.70547283e-01 1.80561081e-01
5.44071682e-02 -4.65370536e-01 -1.86099738e-01 -1.18751538e+00
1.07921667e-01 -3.99157792e-01 -1.19939245e-01 -6.82328522e-01
-1.11645997e+00 8.36062729e-01 4.81216997e-01 -8.11528921e-01
-2.12593302e-01 -8.36588204e-01 -6.66736960e-01 9.84875500e-01
-1.13017929e+00 9.06077959e-03 -1.08715855e-01 8.60371470e-01
-1.83566675e-01 2.62837589e-01 9.80123401e-01 2.72511020e-02
-4.78034139e-01 1.17751986e-01 2.47714296e-01 -3.67961079e-01
3.08699965e-01 -1.61427283e+00 5.96652590e-02 5.81450045e-01
-3.63763818e-03 8.12894881e-01 1.18028402e+00 -4.09341604e-01
-1.03251755e+00 -2.99481571e-01 7.56341577e-01 -4.31144595e-01
7.45202959e-01 -6.34664893e-01 -8.41308594e-01 8.75878334e-02
3.50644171e-01 -1.48082718e-01 8.57318640e-01 5.98292649e-02
1.48989037e-02 -7.17734993e-02 -1.17895651e+00 5.94078720e-01
1.01302302e+00 -6.98499918e-01 -7.39134789e-01 2.42663264e-01
-1.61974311e-01 3.33290100e-01 -1.05544448e+00 2.93307245e-01
7.07863748e-01 -1.51967025e+00 6.47958577e-01 -2.56563514e-01
1.22505657e-01 6.85847774e-02 -7.77634531e-02 -1.57368946e+00
-3.70030314e-01 -7.82657325e-01 1.00412153e-01 5.74134529e-01
6.04901239e-02 -1.06716609e+00 3.57928425e-01 5.19169688e-01
2.22205937e-01 -6.85304463e-01 -1.42953575e+00 -8.28873575e-01
1.86783135e-01 -5.87638438e-01 2.83302397e-01 5.65522134e-01
8.57819557e-01 -4.36022021e-02 3.64326745e-01 -8.84862319e-02
8.30656230e-01 -4.65074897e-01 -5.26498118e-03 -1.45752895e+00
-1.92734510e-01 -8.62208486e-01 -8.19469929e-01 -4.85382527e-01
8.82992148e-02 -8.23438823e-01 3.90173644e-01 -9.82134283e-01
2.22745061e-01 -5.89450449e-02 -6.93629444e-01 -2.76458353e-01
-5.12585882e-03 1.67519987e-01 -2.16457948e-01 -1.30067810e-01
-1.88091323e-01 4.74332720e-01 9.41262722e-01 4.66972709e-01
-2.02045012e-02 -1.58698201e-01 -2.46556804e-01 8.78940761e-01
8.10961902e-01 -4.79376823e-01 -3.13914955e-01 4.72655892e-01
5.74484229e-01 1.72113240e-01 3.64021331e-01 -1.24900007e+00
-9.02841762e-02 -1.31755203e-01 3.12968493e-01 -3.71948063e-01
2.92885125e-01 -6.89265490e-01 1.64975628e-01 4.79116112e-01
-8.35864767e-02 1.63565725e-01 1.37775615e-02 6.01725876e-01
-6.99149668e-02 -4.71772909e-01 9.79858935e-01 -2.87300706e-01
-1.89302519e-01 3.56194109e-01 -9.92983222e-01 2.59169698e-01
1.24846804e+00 -4.27087039e-01 -2.84656525e-01 -3.27793002e-01
-1.08006001e+00 -3.70833308e-01 9.86927599e-02 -3.31057310e-01
4.59943891e-01 -9.57832575e-01 -2.64827341e-01 3.55175197e-01
-3.02708689e-02 -5.29430091e-01 3.99706125e-01 1.39310682e+00
-3.35821450e-01 4.22116518e-01 -4.64468420e-01 -6.34859204e-01
-6.31668448e-01 3.31685096e-01 4.98046339e-01 5.59357218e-02
-3.08446437e-01 8.04476142e-01 2.85754621e-01 2.78407246e-01
-8.44698548e-02 -6.54896438e-01 -1.73786640e-01 2.62579560e-01
5.08543074e-01 3.56356800e-01 1.87875312e-02 -5.11792660e-01
-6.11661792e-01 6.25641465e-01 2.72998601e-01 -5.05038202e-01
1.29827237e+00 -1.57296389e-01 -5.82162499e-01 1.09035850e+00
9.92500186e-01 -9.06434432e-02 -8.67585123e-01 2.53853410e-01
2.60087460e-01 -1.08463593e-01 -1.94410861e-01 -4.93296653e-01
-4.00345385e-01 1.17673695e+00 5.60322881e-01 9.18334961e-01
1.05260885e+00 1.51490197e-01 3.35381567e-01 2.08827779e-01
7.83183336e-01 -1.14598131e+00 -1.00220203e-01 2.60224789e-01
7.22396314e-01 -6.86079383e-01 -5.35834990e-02 -8.50694478e-02
-1.67033821e-01 1.24201131e+00 1.66024312e-01 -6.61009669e-01
1.31156898e+00 2.67357558e-01 -5.92129588e-01 -3.98652881e-01
-7.43374705e-01 -2.40593836e-01 2.33710945e-01 5.50049365e-01
7.37313747e-01 3.01005512e-01 -6.47318959e-01 5.43547511e-01
-4.28229749e-01 -2.67193258e-01 6.75531924e-01 6.70556486e-01
-5.92714608e-01 -6.80005014e-01 -3.50915253e-01 5.26072204e-01
-6.43124104e-01 -3.30788046e-01 5.94244823e-02 7.13387549e-01
1.69640541e-01 7.95610011e-01 1.68041885e-01 -1.62043661e-01
2.09854171e-01 5.95543921e-01 1.12127674e+00 -6.43227816e-01
-3.04578304e-01 1.03058899e-02 -4.67661262e-01 -6.63490057e-01
-3.97937357e-01 -1.06607628e+00 -1.23373711e+00 -1.86212316e-01
-2.25328237e-01 3.80091816e-01 7.91798294e-01 1.19243300e+00
-9.65797231e-02 3.44355255e-01 2.42941111e-01 -1.06644928e+00
-4.91175622e-01 -1.12855649e+00 -1.51117325e+00 1.32068977e-01
3.78547549e-01 -8.67763042e-01 -8.01473379e-01 1.85295288e-02] | [12.84582805633545, 3.4634621143341064] |
a99b2ca8-7175-41a5-aa45-7017f9bdf704 | action2vec-a-crossmodal-embedding-approach-to | 1901.00484 | null | http://arxiv.org/abs/1901.00484v1 | http://arxiv.org/pdf/1901.00484v1.pdf | Action2Vec: A Crossmodal Embedding Approach to Action Learning | We describe a novel cross-modal embedding space for actions, named
Action2Vec, which combines linguistic cues from class labels with
spatio-temporal features derived from video clips. Our approach uses a
hierarchical recurrent network to capture the temporal structure of video
features. We train our embedding using a joint loss that combines
classification accuracy with similarity to Word2Vec semantics. We evaluate
Action2Vec by performing zero shot action recognition and obtain state of the
art results on three standard datasets. In addition, we present two novel
analogy tests which quantify the extent to which our joint embedding captures
distributional semantics. This is the first joint embedding space to combine
verbs and action videos, and the first to be thoroughly evaluated with respect
to its distributional semantics. | ['Meera Hahn', 'Andrew Silva', 'James M. Rehg'] | 2019-01-02 | null | null | null | null | ['zero-shot-action-recognition'] | ['computer-vision'] | [ 1.53954148e-01 -1.75469309e-01 -6.42870784e-01 -2.30114713e-01
-7.61640668e-01 -4.83281195e-01 1.07366037e+00 1.14909850e-01
-7.41382003e-01 5.11528730e-01 1.13494575e+00 2.36531675e-01
-1.63379133e-01 -5.53084373e-01 -2.49598116e-01 -5.41393280e-01
-3.32568198e-01 8.93457211e-04 2.14097589e-01 -1.07319467e-01
4.47645746e-02 3.57592225e-01 -1.51731944e+00 5.46275318e-01
1.18472621e-01 1.00466263e+00 -2.94510424e-01 5.99701345e-01
-5.01467753e-03 1.12123787e+00 -4.03514981e-01 -2.74714142e-01
7.96664432e-02 -4.96331036e-01 -8.32649648e-01 1.11627564e-01
4.03272539e-01 -2.27617353e-01 -7.96531022e-01 6.51442647e-01
3.42501938e-01 5.67588627e-01 9.46341813e-01 -1.40287077e+00
-8.82815003e-01 3.41031879e-01 -4.33126390e-02 4.79409277e-01
6.62079453e-01 3.91735852e-01 1.64223433e+00 -7.41996408e-01
1.04480743e+00 1.23068845e+00 6.75859451e-01 4.57941204e-01
-1.29660738e+00 -1.48033753e-01 5.23464195e-02 6.46223545e-01
-1.15903616e+00 -2.27534279e-01 7.81220555e-01 -7.67840445e-01
1.37963367e+00 -8.79366547e-02 7.50299394e-01 1.62999320e+00
3.12978595e-01 1.15083277e+00 6.91224694e-01 -2.16992170e-01
1.20476872e-01 -1.46493763e-01 1.57893091e-01 5.77322900e-01
-2.69281060e-01 7.88537711e-02 -7.37093925e-01 -7.79390857e-02
6.18180037e-01 2.60825217e-01 -1.21836089e-01 -7.23950148e-01
-1.28786218e+00 1.15407860e+00 2.00941935e-01 6.96573377e-01
-4.75465328e-01 6.21360123e-01 9.54313517e-01 2.80109376e-01
4.92246211e-01 3.98894787e-01 -3.33090901e-01 -6.67867124e-01
-6.46209121e-01 1.15880482e-01 4.16388750e-01 3.95547211e-01
4.03109580e-01 1.75604388e-01 -6.45008087e-01 7.92539299e-01
2.38378987e-01 3.97213101e-01 8.19482207e-01 -1.12958062e+00
8.16995800e-02 5.48517227e-01 -1.02040485e-01 -9.64789152e-01
-1.33782044e-01 4.19341981e-01 -2.06551522e-01 8.71843323e-02
9.58950594e-02 1.57272056e-01 -5.47929883e-01 1.87782276e+00
-4.39305529e-02 4.90104049e-01 1.90440446e-01 7.31539786e-01
8.04038227e-01 5.06842971e-01 3.80869627e-01 -8.88544172e-02
1.18437457e+00 -9.01517868e-01 -7.96630800e-01 1.42317906e-01
9.37707722e-01 -2.34760836e-01 1.10201967e+00 1.34344250e-01
-8.20122540e-01 -4.98031735e-01 -9.41263974e-01 -1.62439808e-01
-5.73566496e-01 -3.92284468e-02 6.46547437e-01 2.47159451e-01
-1.00477552e+00 8.99040282e-01 -1.04115582e+00 -5.41740417e-01
3.32072377e-01 -4.34533134e-03 -8.15780640e-01 1.07314304e-01
-1.44770336e+00 1.00547171e+00 3.33133131e-01 -6.90818548e-01
-9.54563439e-01 -4.88083243e-01 -1.48591709e+00 -1.89469364e-02
1.44829974e-01 -2.95462310e-01 1.06996441e+00 -8.69392455e-01
-1.43768859e+00 8.25526357e-01 -8.74920562e-02 -6.60982311e-01
2.63411760e-01 -1.82119757e-01 -5.56096256e-01 5.01196504e-01
2.79198468e-01 6.00422800e-01 6.98581696e-01 -6.07360423e-01
-3.98088604e-01 -3.15734535e-01 2.48199776e-01 1.05903774e-01
-6.20736301e-01 2.30011158e-02 -2.03977570e-01 -9.29729462e-01
-4.65039074e-01 -7.46312320e-01 6.68168720e-03 2.20851392e-01
3.44488993e-02 -6.40314758e-01 7.94103801e-01 -5.01846969e-01
1.31813812e+00 -2.53423786e+00 5.25185764e-01 -1.90233901e-01
5.69367893e-02 1.46976933e-02 -5.40262938e-01 7.09111631e-01
-1.48258761e-01 1.22953929e-01 3.38588133e-02 -3.04170042e-01
4.01624262e-01 3.63422811e-01 -3.44982266e-01 6.54540122e-01
3.40006649e-01 1.10599673e+00 -1.06406295e+00 -5.35707772e-01
3.81282628e-01 7.61375189e-01 -6.71709120e-01 7.69546255e-02
-1.59786060e-01 -1.55735910e-01 -4.36076462e-01 4.77340072e-01
-2.62936354e-01 -1.00381866e-01 2.45259747e-01 -4.00212795e-01
6.29728809e-02 1.87614679e-01 -8.49824667e-01 1.90913713e+00
-3.24856728e-01 9.54308748e-01 -5.98405182e-01 -1.05943501e+00
5.31492889e-01 5.17111540e-01 9.98013318e-01 -7.07111955e-01
2.48266608e-01 -2.68445164e-01 -4.43987101e-01 -7.54080653e-01
4.40842986e-01 -2.88323492e-01 -4.86351967e-01 6.22144282e-01
7.67530859e-01 1.65968880e-01 3.03122759e-01 3.02657962e-01
1.47169042e+00 4.29738104e-01 5.97504377e-01 4.47910428e-02
2.85530210e-01 -1.33392051e-01 3.72012764e-01 4.74139392e-01
-5.91312587e-01 4.49248850e-01 6.85023427e-01 -4.12902564e-01
-7.53177285e-01 -1.28407264e+00 -1.09979929e-02 1.24038613e+00
-8.82384777e-02 -8.04390132e-01 -1.28077760e-01 -1.02313495e+00
1.08377233e-01 7.99791932e-01 -1.14323056e+00 -4.29620057e-01
-1.76714912e-01 8.11804235e-02 7.27564692e-01 1.05966139e+00
-1.40054701e-02 -1.33054996e+00 -7.68024206e-01 4.65187877e-02
-2.77143940e-02 -1.10588002e+00 -5.15857875e-01 9.74256322e-02
-5.15924692e-01 -1.27474284e+00 -5.04255116e-01 -4.66810912e-01
-1.91294566e-01 2.47286446e-03 1.03005254e+00 -4.22665447e-01
-4.73117918e-01 1.16005051e+00 -8.57079566e-01 1.17066614e-02
-2.38911659e-02 -4.29443121e-01 3.46896350e-01 1.30162621e-02
7.85469711e-01 -4.83808219e-01 -3.76343787e-01 -1.02701355e-02
-1.02509153e+00 -5.19823253e-01 2.96674103e-01 7.79411554e-01
5.74817121e-01 -2.68748105e-01 1.00619923e-02 -3.10878575e-01
5.13228714e-01 -5.59233904e-01 3.45811509e-02 2.61264831e-01
-3.42051685e-02 2.47231528e-01 3.30331713e-01 -5.20334065e-01
-5.52185893e-01 1.14838909e-02 1.03033535e-01 -9.59445298e-01
-2.47925967e-01 5.92414141e-01 1.10905983e-01 3.71681690e-01
4.36272413e-01 8.58137608e-02 1.05432853e-01 -3.47849101e-01
7.93307722e-01 2.88029432e-01 3.90488774e-01 -2.05429897e-01
2.71420836e-01 7.16607511e-01 -1.68453470e-01 -1.07905459e+00
-8.34448516e-01 -7.54227757e-01 -8.52577388e-01 -1.70850560e-01
1.56230402e+00 -7.05580175e-01 -5.95456660e-01 -1.90481246e-02
-1.02894390e+00 -2.92045295e-01 -7.50869989e-01 9.90350544e-01
-1.08985949e+00 5.02998948e-01 -7.74395466e-01 -5.76328039e-01
1.46788895e-01 -9.25014973e-01 1.23526728e+00 -4.84360695e-01
-6.45102322e-01 -1.35483170e+00 6.45671785e-01 2.61556227e-02
7.97820743e-03 5.37510872e-01 6.94396615e-01 -8.70063305e-01
-2.41435133e-02 -2.60011464e-01 -9.68700424e-02 4.21650857e-01
1.98668927e-01 -5.21231443e-03 -7.55983770e-01 -1.83061093e-01
-2.78156817e-01 -6.79120421e-01 1.33462024e+00 4.16139454e-01
8.25239718e-01 -9.93096456e-02 -1.92103386e-01 2.85332978e-01
1.50868094e+00 5.17377555e-02 5.58194518e-01 2.32919604e-01
6.09957457e-01 4.22176331e-01 5.48466861e-01 7.08996534e-01
1.00703485e-01 9.42519426e-01 2.74891227e-01 3.02501172e-01
-4.29675728e-02 -4.39676017e-01 8.88330996e-01 7.04450905e-01
-2.01702584e-02 -2.26899564e-01 -7.29647696e-01 7.25111961e-01
-1.95873606e+00 -1.52913654e+00 3.29874963e-01 1.75178814e+00
5.78053296e-01 1.77887958e-02 2.97838151e-01 7.60413706e-02
2.03391105e-01 8.13330948e-01 -3.03334355e-01 -4.91539896e-01
-1.18907019e-02 4.06669885e-01 2.69034654e-01 4.69096422e-01
-1.46540105e+00 1.08700430e+00 7.05442476e+00 8.20364714e-01
-8.42674613e-01 3.38953584e-01 -1.00662887e-01 -3.47365499e-01
-3.38195831e-01 -1.87103406e-01 -3.12145472e-01 2.58527219e-01
9.81490791e-01 -2.89084703e-01 3.18394713e-02 8.24364483e-01
2.05771565e-01 1.84072435e-01 -1.41443205e+00 9.85385239e-01
6.22372448e-01 -1.40917695e+00 2.68454254e-01 6.93027582e-03
4.43565100e-01 1.21313177e-01 1.62033865e-03 6.23408258e-01
5.91592193e-01 -1.03535008e+00 5.08844376e-01 7.82204688e-01
8.82509530e-01 -5.02096355e-01 5.68357944e-01 -2.31616050e-01
-1.40099406e+00 -1.75844133e-01 -1.40058756e-01 -1.47150785e-01
4.93811518e-01 -1.30457997e-01 -3.80513370e-01 2.17020407e-01
6.12114012e-01 1.52568364e+00 -5.09481847e-01 6.08407617e-01
-2.39897370e-01 4.49753553e-01 1.00621983e-01 -4.22685444e-02
7.31383562e-01 4.72076349e-02 6.03663981e-01 1.28160131e+00
1.30079553e-01 7.91331604e-02 3.71158421e-01 5.49044609e-01
3.91373746e-02 8.74404982e-02 -1.19744647e+00 -6.37444019e-01
7.75168315e-02 7.34874547e-01 -4.53412861e-01 -4.39145118e-01
-6.20960057e-01 1.19291258e+00 3.87194127e-01 2.55614698e-01
-9.82167423e-01 -3.85931462e-01 1.17894292e+00 -2.00881884e-01
5.11991858e-01 -2.91759610e-01 2.85749227e-01 -1.35661542e+00
-1.96443051e-01 -4.30022120e-01 6.35480404e-01 -7.20416009e-01
-1.34530556e+00 3.39527309e-01 1.58696517e-01 -1.45446527e+00
-5.04471064e-01 -6.94861472e-01 -5.08040726e-01 1.44146055e-01
-1.01033008e+00 -1.03763688e+00 1.02894582e-01 7.32533634e-01
8.10302317e-01 -3.95735860e-01 1.02583873e+00 1.79341435e-01
-2.96064138e-01 5.08679509e-01 4.83397506e-02 3.70612353e-01
6.37800932e-01 -1.26317418e+00 -8.88336673e-02 4.08157825e-01
6.35579944e-01 2.76580155e-01 5.55109620e-01 -4.62087214e-01
-1.13720512e+00 -1.03448355e+00 9.98562932e-01 -8.44416678e-01
1.09463048e+00 -1.62556931e-01 -6.13360822e-01 9.55966711e-01
2.66282260e-01 1.22223482e-01 1.12514400e+00 1.24116413e-01
-7.11317301e-01 2.71130085e-01 -7.91124642e-01 5.67807317e-01
1.08481061e+00 -1.07889235e+00 -1.10936809e+00 2.75832653e-01
1.00186861e+00 4.87741202e-01 -9.87391531e-01 3.52365375e-01
7.64842391e-01 -8.86154056e-01 1.02225518e+00 -1.22119987e+00
7.62896121e-01 3.90579551e-02 -7.96570480e-01 -1.32573390e+00
-5.66591620e-01 1.94720589e-02 -8.73625800e-02 8.88475835e-01
2.05291122e-01 -1.35333747e-01 5.33304751e-01 2.60982931e-01
1.17618874e-01 -6.61707759e-01 -8.36858392e-01 -1.15369141e+00
6.12784401e-02 -7.32238233e-01 1.06633089e-01 1.07338822e+00
4.21482146e-01 4.19157028e-01 -4.76918042e-01 -4.39948857e-01
2.95791894e-01 -3.57115269e-02 3.65741640e-01 -9.04813766e-01
-2.68135965e-01 -6.81761682e-01 -1.30772996e+00 -6.73302591e-01
7.42489100e-01 -1.04736650e+00 -2.03785151e-01 -1.56283653e+00
3.17115128e-01 5.12166739e-01 -7.61796117e-01 6.24068975e-01
2.48671800e-01 6.43905580e-01 2.84742057e-01 -2.70830914e-02
-1.18017733e+00 1.17701280e+00 7.76154816e-01 -3.67995590e-01
-5.49983010e-02 -6.23154521e-01 -9.01485607e-02 7.92079091e-01
4.53614920e-01 -2.13909313e-01 -4.62693632e-01 -2.66548991e-01
-1.56587377e-01 1.09810129e-01 4.32016611e-01 -9.77859914e-01
-2.23486274e-01 -3.29208493e-01 1.45472869e-01 -3.29077661e-01
7.14849353e-01 -6.68336093e-01 -2.98666120e-01 2.76974350e-01
-8.91963899e-01 -4.99257110e-02 8.23995024e-02 9.00867343e-01
-5.43102741e-01 1.57852113e-01 4.96061772e-01 -2.07519345e-02
-1.25956297e+00 3.58870029e-01 -6.66677058e-01 2.81136453e-01
1.25581515e+00 -2.87701815e-01 1.51629061e-01 -5.67685127e-01
-1.29696214e+00 2.65329033e-02 3.88161659e-01 7.69101799e-01
7.61969566e-01 -1.99916553e+00 -5.00204742e-01 -5.57330698e-02
5.01750469e-01 -1.21191120e+00 2.54661977e-01 8.78731310e-01
-7.99637362e-02 5.00176191e-01 -4.12207991e-01 -3.98658633e-01
-1.37777543e+00 8.38995397e-01 6.97536990e-02 -3.00666749e-01
-6.86861575e-01 6.56020463e-01 3.85017432e-02 -2.22324640e-01
2.79093921e-01 -2.09689364e-01 -5.53635597e-01 5.12548864e-01
6.10672653e-01 4.11042720e-01 -5.22076428e-01 -1.02475595e+00
-5.85956573e-01 3.65976542e-01 3.89817566e-01 -4.51401502e-01
1.34522247e+00 1.95441917e-01 1.43452555e-01 1.01244700e+00
1.84504354e+00 -3.91774654e-01 -1.23297167e+00 -3.77375692e-01
9.46205035e-02 -6.43668413e-01 -3.71749103e-02 -3.39023501e-01
-8.50951254e-01 9.11172926e-01 5.78551829e-01 1.12114400e-01
6.83568358e-01 2.73241550e-01 7.54988074e-01 3.12823951e-01
1.28340945e-01 -1.19626200e+00 6.96402550e-01 7.08556294e-01
8.63583267e-01 -1.17814565e+00 1.86066981e-02 2.69906402e-01
-1.10271013e+00 8.91133904e-01 4.41390932e-01 -5.21206677e-01
1.05975223e+00 -1.30854294e-01 -1.78515553e-01 -4.89514530e-01
-9.75239277e-01 -7.06750989e-01 4.74165767e-01 3.96577179e-01
4.74907696e-01 1.57870561e-01 -5.77665389e-01 5.91526091e-01
1.08927876e-01 7.84316845e-03 1.58928916e-01 9.07623053e-01
-2.72882283e-01 -9.71981049e-01 3.56961161e-01 4.37858254e-01
-4.36173588e-01 6.24255426e-02 -5.63624740e-01 7.76947021e-01
-3.16644013e-02 6.30822778e-01 3.10941696e-01 -7.17671871e-01
3.21898937e-01 3.90854210e-01 4.01719660e-01 -6.79667413e-01
-1.45110026e-01 -7.63614774e-02 1.55104071e-01 -1.27131557e+00
-7.24244952e-01 -7.31174111e-01 -1.11895132e+00 1.07191205e-01
9.58932787e-02 -2.13117269e-03 2.65361458e-01 9.78256762e-01
2.76272595e-01 4.66381103e-01 4.54179853e-01 -9.12699759e-01
-6.02175355e-01 -8.12579751e-01 -7.87690938e-01 1.05794883e+00
2.54982591e-01 -1.05193079e+00 -4.68299538e-01 3.18244286e-02] | [8.667016983032227, 0.9513075351715088] |
82d8475b-57ed-4fac-98c5-5557f5198220 | clevr-ref-diagnosing-visual-reasoning-with | 1901.00850 | null | http://arxiv.org/abs/1901.00850v2 | http://arxiv.org/pdf/1901.00850v2.pdf | CLEVR-Ref+: Diagnosing Visual Reasoning with Referring Expressions | Referring object detection and referring image segmentation are important
tasks that require joint understanding of visual information and natural
language. Yet there has been evidence that current benchmark datasets suffer
from bias, and current state-of-the-art models cannot be easily evaluated on
their intermediate reasoning process. To address these issues and complement
similar efforts in visual question answering, we build CLEVR-Ref+, a synthetic
diagnostic dataset for referring expression comprehension. The precise
locations and attributes of the objects are readily available, and the
referring expressions are automatically associated with functional programs.
The synthetic nature allows control over dataset bias (through sampling
strategy), and the modular programs enable intermediate reasoning ground truth
without human annotators.
In addition to evaluating several state-of-the-art models on CLEVR-Ref+, we
also propose IEP-Ref, a module network approach that significantly outperforms
other models on our dataset. In particular, we present two interesting and
important findings using IEP-Ref: (1) the module trained to transform feature
maps into segmentation masks can be attached to any intermediate module to
reveal the entire reasoning process step-by-step; (2) even if all training data
has at least one object referred, IEP-Ref can correctly predict no-foreground
when presented with false-premise referring expressions. To the best of our
knowledge, this is the first direct and quantitative proof that neural modules
behave in the way they are intended. | ['Yutong Bai', 'Runtao Liu', 'Alan Yuille', 'Chenxi Liu'] | 2019-01-03 | clevr-ref-diagnosing-visual-reasoning-with-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Liu_CLEVR-Ref_Diagnosing_Visual_Reasoning_With_Referring_Expressions_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Liu_CLEVR-Ref_Diagnosing_Visual_Reasoning_With_Referring_Expressions_CVPR_2019_paper.pdf | cvpr-2019-6 | ['referring-expression-segmentation'] | ['computer-vision'] | [ 3.90731692e-01 7.59999573e-01 -2.99032569e-01 -3.82870972e-01
-7.03461766e-01 -9.30654228e-01 5.98921061e-01 1.03687435e-01
-1.49588183e-01 5.12366414e-01 1.90023948e-02 -5.58921695e-01
7.51247704e-02 -8.04148197e-01 -1.21324408e+00 -1.41327277e-01
2.99088061e-01 7.29716241e-01 3.99096340e-01 -7.57616758e-02
1.92510650e-01 4.78285253e-01 -1.47821093e+00 9.66325104e-01
7.55672336e-01 7.36155868e-01 2.81931330e-02 6.19169950e-01
-2.57743955e-01 1.33930361e+00 -6.65869176e-01 -7.89287686e-01
-1.02105990e-01 -6.21436119e-01 -1.39449954e+00 7.37661272e-02
8.08656394e-01 -2.93175220e-01 -2.17253298e-01 1.05572832e+00
7.19993934e-02 -2.44540259e-01 6.63917184e-01 -1.52927721e+00
-1.29128027e+00 5.35336375e-01 -3.99578571e-01 3.79506081e-01
7.61438310e-01 6.90996230e-01 1.26907408e+00 -8.15290451e-01
1.23713827e+00 1.48029220e+00 5.32725513e-01 7.80611694e-01
-1.72113454e+00 -2.70349950e-01 4.01762009e-01 2.51111329e-01
-1.18325841e+00 -2.45714143e-01 6.57046556e-01 -5.97726703e-01
1.19415522e+00 3.10749620e-01 7.66597211e-01 1.22893989e+00
-1.09687641e-01 1.12761986e+00 1.14875031e+00 -2.95868039e-01
1.67138860e-01 2.56690085e-01 3.00164819e-01 1.00418365e+00
1.98526472e-01 3.08958925e-02 -3.94162208e-01 1.22411445e-01
7.17232764e-01 -3.22642982e-01 -4.71102268e-01 -5.97365558e-01
-1.29374933e+00 6.08574510e-01 8.49273682e-01 3.44273537e-01
-1.65047154e-01 4.71119761e-01 3.14093500e-01 1.27358258e-01
2.76428144e-02 7.50251710e-01 -3.52442116e-01 1.65883452e-01
-8.41258526e-01 3.10782254e-01 6.88152671e-01 1.04407167e+00
5.43649077e-01 -1.90801173e-01 -5.39612591e-01 2.88111657e-01
2.70910025e-01 2.08731741e-01 5.38239628e-02 -1.24345577e+00
6.34126246e-01 1.15656328e+00 -4.07123119e-02 -9.40401316e-01
-5.03944993e-01 -2.26038978e-01 -2.47712895e-01 2.59263277e-01
8.96363676e-01 2.84635007e-01 -7.49131858e-01 1.91204023e+00
3.77103761e-02 -4.35905546e-01 9.56601351e-02 9.53367651e-01
1.24143469e+00 3.33436847e-01 4.62154835e-01 3.32100928e-01
1.75797272e+00 -7.22010911e-01 -4.75388259e-01 -6.94853187e-01
7.38946974e-01 -3.36651981e-01 1.61023235e+00 1.02487110e-01
-1.37573862e+00 -4.93051052e-01 -9.62060452e-01 -4.38419580e-01
-5.91860116e-01 2.47862041e-01 8.26629937e-01 4.92117703e-01
-1.21918237e+00 1.43077791e-01 -5.78160346e-01 -5.21236718e-01
1.14291120e+00 1.13453552e-01 -4.23518300e-01 -2.46410400e-01
-8.47296894e-01 1.16828310e+00 3.28944296e-01 6.77350760e-02
-1.03317964e+00 -7.63635457e-01 -9.73675311e-01 5.29319644e-02
3.93211216e-01 -1.04157221e+00 1.40939939e+00 -1.41542208e+00
-5.66831231e-01 1.70795465e+00 -5.69097996e-01 -4.00343895e-01
6.34337664e-01 8.54389518e-02 -1.76261351e-01 6.47531569e-01
2.24032313e-01 1.19348800e+00 5.61755538e-01 -1.62364495e+00
-1.81007758e-01 -4.59725976e-01 5.77437103e-01 8.23719706e-03
3.73785496e-01 -4.01821807e-02 -4.65388358e-01 -2.92218924e-01
2.94551402e-01 -4.05870229e-01 2.30562523e-01 6.40956104e-01
-5.51569581e-01 -3.43617857e-01 6.53506458e-01 -5.39485991e-01
7.14235246e-01 -2.03544593e+00 1.75894454e-01 -3.13411593e-01
3.44025403e-01 -6.16194569e-02 -4.26232144e-02 5.96774966e-02
-4.07826483e-01 5.60819268e-01 -3.32744151e-01 4.47856635e-02
1.62490934e-01 3.23841482e-01 -5.59599340e-01 4.31296259e-01
8.52828562e-01 1.61148882e+00 -9.85056520e-01 -8.95009637e-01
7.37820789e-02 1.11520521e-01 -4.26567882e-01 8.75403509e-02
-7.13568389e-01 3.20822805e-01 -1.95104405e-01 9.38434541e-01
5.11239231e-01 -5.13235927e-01 4.71510105e-02 -5.22669256e-01
2.83379227e-01 3.17554802e-01 -7.17273057e-01 1.51092362e+00
-1.46907583e-01 1.02872241e+00 -4.25200090e-02 -1.02032626e+00
4.30973589e-01 2.00983867e-01 -2.51286924e-01 -9.55305278e-01
1.13054611e-01 -1.23676090e-02 5.01369052e-02 -8.86611283e-01
2.67265588e-01 -2.44895503e-01 -6.97716698e-02 4.18981165e-01
2.53187537e-01 -4.60820436e-01 4.73305136e-01 5.62589169e-01
9.96708989e-01 5.33487260e-01 2.18567804e-01 -1.17946021e-01
4.00625259e-01 7.00481534e-01 2.90209085e-01 7.97207296e-01
-4.06173468e-01 5.98667085e-01 1.25315487e+00 -3.15412104e-01
-9.22086060e-01 -1.37072718e+00 -2.24482238e-01 1.07940698e+00
2.48911291e-01 5.44394739e-02 -8.50605488e-01 -9.66226637e-01
-9.44922641e-02 1.29360962e+00 -1.01578009e+00 -5.76004498e-02
-4.18772608e-01 -5.15878618e-01 9.12388086e-01 8.64889205e-01
3.92780483e-01 -1.42304111e+00 -9.36751485e-01 -2.30326623e-01
-2.32528150e-01 -1.21394277e+00 2.22511999e-02 1.95290953e-01
-9.04127836e-01 -1.60885596e+00 -3.34299326e-01 -9.19117749e-01
1.14402604e+00 -3.85352187e-02 1.75141156e+00 3.22949231e-01
-3.19150299e-01 7.11621463e-01 -2.55508982e-02 -1.32353574e-01
-4.76381749e-01 -2.79433846e-01 -9.37852561e-01 -5.32921195e-01
4.87673461e-01 -1.80614784e-01 -4.58364218e-01 3.55351895e-01
-8.29149663e-01 3.38617504e-01 6.01460636e-01 7.48815656e-01
6.46857440e-01 -4.09605175e-01 3.74193609e-01 -1.23303211e+00
5.00052631e-01 -3.77224714e-01 -5.00424862e-01 6.17552757e-01
-1.76022425e-01 1.12760343e-01 3.69134158e-01 -2.59590387e-01
-1.20586252e+00 3.25706340e-02 1.14727810e-01 -1.58339664e-01
-4.85829026e-01 8.75776112e-02 -2.86800712e-01 2.20532432e-01
9.96931374e-01 1.45386219e-01 -3.17197561e-01 1.52172938e-01
8.64311814e-01 8.07426870e-02 8.81192088e-01 -8.31775784e-01
6.51722133e-01 7.42546797e-01 1.11133363e-02 -3.52931201e-01
-1.13758373e+00 -6.48898929e-02 -8.58865917e-01 -2.03524962e-01
1.26877177e+00 -8.49819779e-01 -9.78288352e-01 5.20006241e-03
-1.40563345e+00 -6.23833954e-01 -5.38343906e-01 -1.98678821e-01
-8.10726225e-01 8.86546522e-02 -6.22856081e-01 -7.67444432e-01
6.57544583e-02 -1.12265253e+00 1.10451186e+00 2.18427673e-01
-5.69878757e-01 -9.29445922e-01 -5.07723391e-01 6.45388186e-01
1.00972079e-01 3.95578980e-01 1.33781254e+00 -5.79287410e-01
-9.02286768e-01 -5.33554330e-02 -8.34462523e-01 -1.27515972e-01
-3.47888857e-01 1.46399230e-01 -1.22018194e+00 2.35792488e-01
-2.22229615e-01 -7.14719892e-01 7.87230790e-01 9.84083489e-02
1.32331133e+00 -1.07119091e-01 -5.03809333e-01 3.82575303e-01
1.25377584e+00 2.13364959e-02 7.23844290e-01 1.78665474e-01
6.57565534e-01 8.80827308e-01 5.36987722e-01 -3.53945434e-01
5.99201083e-01 3.10263067e-01 7.74229944e-01 -2.48992920e-01
-5.45256913e-01 -4.61585790e-01 9.51155648e-02 -2.75307953e-01
3.30422997e-01 -1.87053889e-01 -1.13326681e+00 8.96081865e-01
-1.70025945e+00 -1.13873911e+00 -4.79910910e-01 1.81354356e+00
8.04219067e-01 2.74068981e-01 -1.16854191e-01 -5.11592738e-02
4.59780067e-01 5.50666191e-02 -6.78252935e-01 -4.59534436e-01
-4.16708916e-01 1.67324468e-01 8.60186890e-02 2.34999791e-01
-9.55340743e-01 1.03454697e+00 6.71176958e+00 3.33051026e-01
-7.74578393e-01 1.76178128e-01 8.38639259e-01 5.65660000e-02
-5.66353261e-01 2.00162187e-01 -5.66868424e-01 -1.19865470e-01
5.48177660e-01 2.27134228e-01 3.90595883e-01 8.42689395e-01
-1.89530298e-01 -4.51187253e-01 -1.74193609e+00 7.86956966e-01
3.57855618e-01 -1.33361745e+00 1.34803861e-01 -3.78827631e-01
4.93080765e-01 -3.09340030e-01 1.13781549e-01 4.39698160e-01
3.85239869e-01 -1.39794052e+00 1.04358506e+00 5.94878912e-01
6.91932619e-01 -3.51683408e-01 6.11996233e-01 2.05193549e-01
-8.64346206e-01 -2.72996407e-02 -1.35288075e-01 1.35318846e-01
1.44171432e-01 1.66191593e-01 -8.43819678e-01 1.40843004e-01
6.19704187e-01 4.78881776e-01 -1.16830826e+00 6.96575701e-01
-9.14306223e-01 5.27914941e-01 6.04361892e-02 -5.91373965e-02
6.11813553e-02 3.00201833e-01 2.68254310e-01 1.12760806e+00
-2.07932726e-01 -5.01604229e-02 -3.02868158e-01 1.94460905e+00
-2.12922588e-01 -2.52667636e-01 -6.47768915e-01 -2.55634189e-01
2.37707630e-01 1.08952153e+00 -8.14339578e-01 -3.85148495e-01
-5.34952998e-01 9.89299417e-01 6.03770852e-01 6.49675488e-01
-8.61582518e-01 -8.50122645e-02 2.42853120e-01 1.91678882e-01
2.89273232e-01 1.81075409e-01 -6.20747507e-01 -9.17314649e-01
1.82366461e-01 -7.84373045e-01 5.46809614e-01 -1.64523804e+00
-1.28025067e+00 3.44134927e-01 -5.39493747e-02 -7.58008003e-01
-1.28145427e-01 -1.07177567e+00 -5.31151831e-01 8.26124787e-01
-1.34927905e+00 -1.41475153e+00 -4.51106668e-01 5.49043298e-01
3.30647200e-01 2.84728229e-01 8.81304562e-01 -2.33676508e-01
-5.53682208e-01 4.59751904e-01 -8.17104876e-01 3.46925706e-01
3.81358176e-01 -1.41214144e+00 3.51615429e-01 8.58622968e-01
3.21030378e-01 8.84323537e-01 7.42443383e-01 -4.28923219e-01
-1.34638417e+00 -9.07820165e-01 6.10091865e-01 -1.13045657e+00
6.01882041e-01 -3.42373073e-01 -1.02392602e+00 1.20812964e+00
2.47237578e-01 2.93163121e-01 4.63938832e-01 7.91396350e-02
-7.66856134e-01 2.66349912e-01 -1.41181517e+00 7.79958367e-01
1.19425535e+00 -7.96163499e-01 -1.24654353e+00 4.52946693e-01
7.09494293e-01 -4.74971414e-01 -5.99185884e-01 1.86438963e-01
1.86012164e-01 -1.14085448e+00 1.00900757e+00 -9.56612527e-01
7.53777325e-01 -5.51183939e-01 -2.80668050e-01 -8.17078590e-01
1.12335123e-01 5.26155569e-02 -1.73613757e-01 1.23036444e+00
6.85558558e-01 -3.56096864e-01 6.68414593e-01 9.11109447e-01
-9.98285636e-02 -7.23954499e-01 -7.87478685e-01 -4.28497165e-01
7.60425404e-02 -6.86764061e-01 2.14214846e-01 8.00085306e-01
1.97348669e-01 3.98747295e-01 3.96999568e-01 1.07368879e-01
3.97336096e-01 3.25108677e-01 7.03339696e-01 -9.43073809e-01
-4.58370119e-01 -5.84408998e-01 -3.56805176e-01 -9.99471605e-01
4.21955556e-01 -1.09125805e+00 -5.73360212e-02 -1.98091376e+00
4.32171851e-01 -1.94482468e-02 1.96159750e-01 7.39810407e-01
-1.54055580e-01 2.10079759e-01 1.81669027e-01 -6.31222054e-02
-7.71915913e-01 1.19204831e-03 1.53852129e+00 -3.38149130e-01
1.95001140e-01 -2.59542614e-01 -1.05810606e+00 9.80729520e-01
6.02573037e-01 -2.70055652e-01 -4.24274087e-01 -6.11520529e-01
6.51473761e-01 -2.10895576e-03 1.17517877e+00 -5.22655845e-01
6.36132360e-02 -8.67576972e-02 6.37255490e-01 -5.21755099e-01
3.43183041e-01 -7.34944701e-01 -1.56237110e-01 3.95702869e-01
-4.02685374e-01 4.90211062e-02 4.54643518e-01 3.63268316e-01
-1.15398899e-01 -4.38242972e-01 6.03184223e-01 -5.55251896e-01
-9.90676582e-01 -3.60031933e-01 -3.93741518e-01 5.54218352e-01
1.08574414e+00 -2.20399082e-01 -1.15879238e+00 -2.35483453e-01
-7.60030866e-01 2.96833158e-01 6.49628937e-01 1.31915331e-01
6.56632721e-01 -9.54452634e-01 -3.79125804e-01 -4.32914607e-02
6.38888836e-01 1.61543787e-01 2.02691913e-01 9.10803080e-01
-6.45614147e-01 4.32602376e-01 -1.26849771e-01 -8.06761086e-01
-1.02161944e+00 9.23192143e-01 5.66418409e-01 2.36369148e-02
-4.24474746e-01 1.11935580e+00 4.74961221e-01 -6.05727196e-01
1.13256171e-01 -5.15737414e-01 -1.21693471e-02 1.01236373e-01
3.62799853e-01 -9.56990868e-02 -2.37250462e-01 -6.87771916e-01
-4.97062624e-01 4.42302734e-01 1.09094463e-01 -8.19553509e-02
9.51714993e-01 1.13734491e-01 -3.26891392e-01 5.40631831e-01
7.53042459e-01 -2.14684919e-01 -1.10938156e+00 8.03698450e-02
5.50413989e-02 -2.97097683e-01 -3.52894545e-01 -1.23600328e+00
-9.30163622e-01 1.28341353e+00 3.22606444e-01 8.90224501e-02
1.03570402e+00 6.78507030e-01 9.08241794e-02 3.53518009e-01
2.18922585e-01 -7.49307990e-01 2.77594537e-01 1.85163647e-01
1.01936519e+00 -1.23863792e+00 -3.16748992e-02 -5.55842817e-01
-7.92888939e-01 8.28619599e-01 1.04790938e+00 7.22352043e-02
-1.39036998e-01 3.76378968e-02 7.72383958e-02 -6.65104687e-01
-8.29016924e-01 -3.48487049e-01 3.11455667e-01 8.48672748e-01
6.85702324e-01 -8.70723277e-02 2.95541175e-02 5.14892876e-01
-1.36556238e-01 -8.59770104e-02 3.60673368e-01 6.95948303e-01
-2.01440156e-01 -6.24599874e-01 -3.89860660e-01 4.81797218e-01
-4.09452170e-01 -1.31528035e-01 -7.83659041e-01 1.18874681e+00
1.37089282e-01 8.41768444e-01 2.18407631e-01 1.75395861e-01
5.28520763e-01 3.46640408e-01 8.16420674e-01 -6.12727463e-01
-6.37994766e-01 -5.51276803e-01 3.14358175e-01 -6.65177643e-01
-3.35614294e-01 -6.17219269e-01 -1.62237954e+00 1.09020352e-01
-2.94585396e-02 -4.36089367e-01 2.42116243e-01 9.80251193e-01
1.54033676e-01 9.70301688e-01 -4.21550810e-01 -5.47890484e-01
-2.28041008e-01 -6.72514319e-01 -1.07004493e-01 8.94531965e-01
1.76715657e-01 -6.40661597e-01 -3.33878100e-01 1.97526738e-01] | [10.764100074768066, 1.8905428647994995] |
6f2c5960-14c7-493b-8aa2-2be231f7b1ca | soccernet-caption-dense-video-captioning-for | 2304.04565 | null | https://arxiv.org/abs/2304.04565v1 | https://arxiv.org/pdf/2304.04565v1.pdf | SoccerNet-Caption: Dense Video Captioning for Soccer Broadcasts Commentaries | Soccer is more than just a game - it is a passion that transcends borders and unites people worldwide. From the roar of the crowds to the excitement of the commentators, every moment of a soccer match is a thrill. Yet, with so many games happening simultaneously, fans cannot watch them all live. Notifications for main actions can help, but lack the engagement of live commentary, leaving fans feeling disconnected. To fulfill this need, we propose in this paper a novel task of dense video captioning focusing on the generation of textual commentaries anchored with single timestamps. To support this task, we additionally present a challenging dataset consisting of almost 37k timestamped commentaries across 715.9 hours of soccer broadcast videos. Additionally, we propose a first benchmark and baseline for this task, highlighting the difficulty of temporally anchoring commentaries yet showing the capacity to generate meaningful commentaries. By providing broadcasters with a tool to summarize the content of their video with the same level of engagement as a live game, our method could help satisfy the needs of the numerous fans who follow their team but cannot necessarily watch the live game. We believe our method has the potential to enhance the accessibility and understanding of soccer content for a wider audience, bringing the excitement of the game to more people. | ['Marc Van Droogenbroeck', 'Bernard Ghanem', 'Silvio Giancola', 'Anthony Cioppa', 'Hassan Mkhallati'] | 2023-04-10 | null | null | null | null | ['video-captioning', 'dense-video-captioning'] | ['computer-vision', 'computer-vision'] | [ 1.50030136e-01 1.08011469e-01 -9.15799737e-02 -8.84615555e-02
-1.20697546e+00 -9.26471949e-01 5.64809501e-01 2.62352079e-01
-3.93988788e-01 7.62165964e-01 9.63218212e-01 -9.38671455e-03
4.56927389e-01 -4.37591732e-01 -5.45861483e-01 -3.18290323e-01
5.55225313e-02 1.46141142e-01 5.79702258e-01 -7.82231808e-01
2.30539590e-01 -2.12281510e-01 -1.68997264e+00 8.53102028e-01
2.64474154e-01 8.39652121e-01 1.03510909e-01 7.69228339e-01
1.79074347e-01 1.27773321e+00 -9.73719656e-01 -6.46999300e-01
7.81095698e-02 -6.10538185e-01 -4.69827563e-01 -5.00928424e-02
6.55865729e-01 -3.77000332e-01 -6.86593473e-01 5.45540333e-01
5.87512374e-01 4.00453061e-01 6.32751435e-02 -1.43646467e+00
-3.93183202e-01 9.21923161e-01 -2.93562591e-01 3.91494423e-01
1.05402434e+00 2.94266254e-01 1.18491518e+00 -5.62369525e-01
8.82692933e-01 7.68715382e-01 8.85627687e-01 3.86519879e-01
-6.72159731e-01 -7.25277841e-01 3.28822732e-01 1.89002767e-01
-1.15110314e+00 -6.51679873e-01 6.52128220e-01 -6.76328540e-01
5.41722357e-01 5.95906973e-01 9.88016129e-01 1.54977822e+00
-1.26286387e-01 6.97788894e-01 4.53474253e-01 1.50612906e-01
1.13599084e-01 -1.10355787e-01 -1.10638440e-01 1.80158705e-01
-1.18283540e-01 -3.98010731e-01 -1.34241283e+00 -1.27135560e-01
2.89907694e-01 1.59503549e-01 -3.89300555e-01 1.39313146e-01
-1.75372064e+00 7.86579370e-01 2.78252698e-02 6.57308847e-02
-2.04168394e-01 3.32276106e-01 8.86250973e-01 7.41854832e-02
5.36983550e-01 3.85461926e-01 1.28279686e-01 -9.83304203e-01
-1.27728438e+00 6.37718499e-01 9.96352375e-01 8.91658962e-01
2.07378641e-01 -1.73663557e-01 -5.66426754e-01 3.36800337e-01
-1.56951681e-01 4.49386358e-01 2.19000787e-01 -9.69397366e-01
7.78657019e-01 5.37700772e-01 5.51798105e-01 -1.58563125e+00
-3.16927791e-01 -2.60479689e-01 -2.55577803e-01 -1.20474912e-01
7.73196399e-01 -3.17639619e-01 -1.12276703e-01 1.56324863e+00
4.13399190e-02 2.18921214e-01 -5.52622318e-01 1.28141832e+00
9.16494906e-01 9.39490020e-01 -2.64912426e-01 -3.26910466e-02
1.62617266e+00 -8.77660751e-01 -7.38234997e-01 -3.40658039e-01
3.99353027e-01 -7.86962390e-01 1.34049237e+00 4.56351519e-01
-1.16398096e+00 -3.46512616e-01 -1.12864387e+00 -1.69622332e-01
1.24337338e-01 -3.30585808e-01 4.27721292e-01 3.25975180e-01
-7.52650380e-01 3.27373385e-01 -7.47734487e-01 -4.56948847e-01
2.00648531e-01 -1.51891544e-01 -5.05478203e-01 -1.65555961e-02
-1.23075974e+00 5.99353194e-01 1.27450645e-01 -1.37938485e-01
-7.71215022e-01 -8.15731764e-01 -8.01580310e-01 -1.37479052e-01
6.97315753e-01 -2.13499606e-01 1.49509394e+00 -1.12369061e+00
-1.15984857e+00 7.72285581e-01 -1.13206208e-01 -5.18220663e-01
8.22302997e-01 -4.54659283e-01 -4.18607146e-01 3.62347662e-01
4.27161276e-01 4.50086713e-01 4.09141928e-01 -8.12825978e-01
-8.62443447e-01 9.28669795e-02 3.90643626e-01 3.12233865e-01
-4.19725180e-01 3.36741596e-01 -5.04872859e-01 -6.59914017e-01
-2.45376199e-01 -8.43841910e-01 1.32264256e-01 -1.82555020e-01
-2.07826048e-01 1.34433389e-01 9.28054869e-01 -8.24168801e-01
1.72339332e+00 -2.52092576e+00 1.16881877e-01 -2.28161409e-01
4.08637315e-01 -2.45226666e-01 2.19291911e-01 1.06655657e+00
2.75975019e-01 4.04469557e-02 1.06345639e-01 -3.02358449e-01
2.03219607e-01 -5.47725223e-02 -5.45438409e-01 5.56075394e-01
-2.57590920e-01 9.87493634e-01 -1.27035534e+00 -3.61618578e-01
-2.62767285e-01 3.75534028e-01 -6.53235614e-01 1.05698369e-02
-3.03386807e-01 5.46900690e-01 -1.24688976e-01 4.63923633e-01
-6.27974048e-03 -4.21540618e-01 3.52813154e-02 -4.77515683e-02
-3.74420345e-01 3.97419661e-01 -1.10827541e+00 2.07882476e+00
-7.49593377e-02 1.23203790e+00 1.71883717e-01 -5.06597161e-01
6.56452537e-01 5.34456491e-01 6.56621397e-01 -7.40052283e-01
4.32894789e-02 1.35423639e-03 -2.88048685e-01 -4.96575415e-01
1.17450309e+00 -1.41199708e-01 -8.21875393e-01 7.81457365e-01
-3.51407051e-01 5.26784174e-02 4.68587697e-01 6.55270994e-01
1.35041559e+00 8.30110684e-02 -2.54804529e-02 1.59028977e-01
-1.15757607e-01 3.76560658e-01 4.33395088e-01 7.79266238e-01
-2.62875557e-01 1.08713782e+00 7.39402294e-01 -6.67138696e-01
-1.24991274e+00 -8.70936394e-01 5.61070919e-01 1.55170393e+00
3.14592987e-01 -9.63710904e-01 -5.16259670e-01 -3.54284257e-01
-1.71985790e-01 3.27298433e-01 -5.30735612e-01 -7.82166943e-02
-7.08345294e-01 -2.04119027e-01 6.75221980e-01 1.59197390e-01
4.30623144e-01 -8.06350827e-01 -9.12604034e-01 3.70330900e-01
-1.21935880e+00 -1.41996551e+00 -1.01031423e+00 -4.67471570e-01
-8.00863057e-02 -8.67311776e-01 -8.20648968e-01 -4.88217235e-01
7.14126378e-02 4.69329387e-01 1.44402409e+00 1.48600861e-01
-4.02109697e-02 4.58170414e-01 -7.94413269e-01 -2.52230257e-01
-3.15029591e-01 1.64851800e-01 6.53690286e-03 1.75350443e-01
2.37673625e-01 -8.09337139e-01 -6.71073020e-01 4.54008847e-01
-7.64941871e-01 4.87538069e-01 -1.11390002e-01 3.66052389e-01
1.95424572e-01 -3.87803078e-01 5.37853897e-01 -5.16413391e-01
4.86031026e-01 -9.88558829e-01 -4.18060906e-02 -3.43577951e-01
2.57108539e-01 -6.06099427e-01 4.21498984e-01 -5.42212009e-01
-6.02618337e-01 -1.86547562e-01 1.26319930e-01 2.87020840e-02
1.78378418e-01 4.52945501e-01 2.62815624e-01 5.57460368e-01
9.50962305e-01 5.03657851e-03 -1.26031429e-01 -3.06057990e-01
3.13007027e-01 6.24760330e-01 9.86473620e-01 -4.63219225e-01
6.88408911e-01 7.77071357e-01 -4.05393511e-01 -8.53792310e-01
-1.03307927e+00 -7.78126359e-01 -8.07099938e-02 -8.45427036e-01
8.58519077e-01 -1.30427718e+00 -1.05865002e+00 2.33761683e-01
-1.18819404e+00 -3.43589664e-01 -4.65229273e-01 2.54424363e-01
-4.00732607e-01 4.20977712e-01 -6.18139625e-01 -5.20384490e-01
4.40915488e-02 -8.17587197e-01 1.03461814e+00 2.29446113e-01
-6.33347094e-01 -5.42969584e-01 1.40948519e-01 6.32614017e-01
3.07145506e-01 7.22899914e-01 -1.71821475e-01 -7.13395894e-01
-5.11663437e-01 -3.50921690e-01 -7.63275400e-02 6.91722438e-04
1.74713954e-02 -1.02969758e-01 -8.03659976e-01 -2.06623435e-01
-4.13414836e-01 -4.39758658e-01 5.84396303e-01 7.42737129e-02
6.30896509e-01 -5.17657697e-01 -1.43825989e-02 2.92255968e-01
1.09835482e+00 -1.51265562e-01 6.08685017e-01 4.26580548e-01
6.06153488e-01 4.06224430e-01 7.57798493e-01 9.21044469e-01
7.28953063e-01 8.56436431e-01 5.68167984e-01 1.31710067e-01
-1.50024161e-01 -6.93417847e-01 7.47187436e-01 8.91028345e-01
-2.38918468e-01 -4.92016613e-01 -9.53482687e-01 8.88341904e-01
-2.07971001e+00 -1.40880096e+00 -2.62269795e-01 2.11830521e+00
6.62539303e-01 2.70653337e-01 6.44102633e-01 1.73623472e-01
7.13074267e-01 5.60398161e-01 -3.65915969e-02 -3.86808157e-01
-1.36648551e-01 -3.21127653e-01 4.09082413e-01 3.56155366e-01
-8.62405419e-01 4.90448177e-01 5.72458076e+00 6.96177185e-01
-1.18344653e+00 2.63343304e-01 3.00428510e-01 -7.73137569e-01
-3.22121620e-01 1.09253950e-01 -7.97782421e-01 8.75855327e-01
1.05881464e+00 -2.60385782e-01 4.40672725e-01 6.03563786e-01
6.64761424e-01 -3.94233942e-01 -1.20531821e+00 1.13946640e+00
5.08185327e-01 -1.70714474e+00 -3.17530841e-01 -5.56544624e-02
8.10468197e-01 5.07651269e-03 -2.05918122e-02 2.53444314e-01
1.84142336e-01 -1.13779044e+00 1.43895924e+00 4.87983137e-01
7.54161835e-01 -4.62795258e-01 5.88301599e-01 2.64007837e-01
-1.17555690e+00 -8.77465755e-02 1.60269320e-01 -8.49264801e-01
8.44532549e-01 4.82696295e-01 -6.65498853e-01 4.04250801e-01
7.15894043e-01 8.87574494e-01 -5.58802426e-01 1.01816273e+00
-4.72921580e-02 6.27578378e-01 -3.10086519e-01 -1.77963331e-01
3.47315073e-01 2.02492461e-01 8.35443377e-01 1.31757426e+00
5.42819738e-01 1.71909720e-01 2.81554401e-01 3.58555019e-01
-6.95484355e-02 4.71014678e-02 -6.57284737e-01 -3.38768959e-01
5.75521588e-01 1.09985423e+00 -8.47639263e-01 -2.46971130e-01
-4.19077486e-01 1.01635659e+00 -1.25229120e-01 1.42065406e-01
-1.22008467e+00 -1.93660617e-01 6.33943141e-01 6.13513470e-01
9.95777845e-02 -3.10938448e-01 2.01630276e-02 -1.23080289e+00
3.88564169e-01 -1.14232230e+00 3.75290424e-01 -1.01052988e+00
-9.80745196e-01 3.82910818e-01 -1.34674758e-01 -1.25232589e+00
-2.65210718e-01 7.02551976e-02 -6.97965503e-01 2.52592683e-01
-8.54968429e-01 -1.02389455e+00 -6.91476107e-01 4.85829920e-01
6.54126167e-01 3.15658808e-01 2.38192946e-01 6.34866595e-01
-8.64946321e-02 2.55993843e-01 -2.03324512e-01 1.32041603e-01
1.11192811e+00 -8.96230459e-01 4.52304691e-01 7.58029461e-01
4.30058748e-01 4.21497524e-02 1.27489614e+00 -5.25457025e-01
-1.31646776e+00 -6.26666248e-01 9.64408994e-01 -9.75710392e-01
9.09123778e-01 -6.58508360e-01 -6.13343894e-01 7.46563196e-01
1.81424707e-01 -4.06267613e-01 6.99863911e-01 4.85599339e-02
-1.66084707e-01 -2.93485876e-02 -3.61516714e-01 6.81430280e-01
1.00422919e+00 -7.07011938e-01 -6.72407329e-01 4.39545393e-01
9.34984684e-01 -6.58115566e-01 -5.17631233e-01 -2.33026475e-01
7.15156913e-01 -1.17177022e+00 7.69573689e-01 -3.17454457e-01
7.32480884e-01 -3.70130599e-01 -1.15995720e-01 -1.09923100e+00
1.09768212e-01 -1.18163025e+00 2.48312190e-01 1.48444295e+00
3.39695156e-01 3.73483114e-02 8.49694073e-01 6.83679104e-01
-3.57458502e-01 -2.11749554e-01 -9.32980835e-01 -4.57933098e-01
-3.55129629e-01 -8.58196259e-01 3.46434355e-01 8.26241016e-01
5.98744631e-01 2.41721123e-01 -8.05954337e-01 -1.16056599e-01
2.64308691e-01 -2.75538206e-01 1.03518558e+00 -1.00122142e+00
-3.98444563e-01 -1.33057788e-01 -5.56376100e-01 -1.05400288e+00
-2.85320789e-01 -7.23533034e-01 1.75225109e-01 -1.58143210e+00
3.09091300e-01 -8.49857833e-03 1.64266527e-01 3.34378690e-01
1.73896149e-01 8.63978028e-01 4.52592403e-01 1.93956330e-01
-1.26914835e+00 1.83298159e-02 1.13868809e+00 6.64411560e-02
-2.11687461e-01 -1.30637765e-01 -8.90672326e-01 5.77434182e-01
5.05623519e-01 -4.14461941e-01 -2.87266344e-01 -3.35258573e-01
1.14010012e+00 2.35796288e-01 6.80465937e-01 -1.29431915e+00
5.87816358e-01 -1.43388817e-02 -2.57910460e-01 -4.29624319e-01
5.51135778e-01 -4.36542451e-01 5.44588506e-01 1.23885296e-01
-3.26268822e-01 1.58537194e-01 1.84186190e-01 7.30160236e-01
-4.44797397e-01 2.84292758e-01 2.31649905e-01 -1.26614630e-01
-6.69918537e-01 1.05011187e-01 -6.47998035e-01 5.87310374e-01
9.01804328e-01 -4.20860171e-01 -5.34104943e-01 -1.04610801e+00
-6.93136573e-01 3.15864056e-01 7.98763216e-01 5.76553822e-01
2.67739594e-01 -1.34454775e+00 -1.00177193e+00 -2.20785782e-01
2.83761859e-01 -8.52707401e-02 4.85681146e-01 9.09933448e-01
-6.16392493e-01 1.67790696e-01 -2.11074680e-01 -6.12378597e-01
-1.08305585e+00 2.84894556e-01 -2.63438612e-01 -6.31492510e-02
-1.03171659e+00 7.87942350e-01 8.46779346e-02 2.89039224e-01
3.48038495e-01 -3.13895673e-01 -1.33893758e-01 5.81495881e-01
1.04612780e+00 4.25250262e-01 -8.00005570e-02 -7.64509559e-01
-3.31959754e-01 4.38075811e-02 2.55943745e-01 -5.08834720e-01
1.30271053e+00 -4.26825970e-01 1.95807248e-01 8.81346285e-01
8.20969284e-01 6.44512832e-01 -1.59636569e+00 1.30681261e-01
-2.67000884e-01 -5.02217889e-01 -4.45168972e-01 -7.32452035e-01
-7.23705471e-01 7.14860201e-01 -1.47022650e-01 4.80897009e-01
5.44990599e-01 1.26922220e-01 1.20408738e+00 -3.06575857e-02
3.91479224e-01 -1.11933243e+00 4.81976122e-01 3.68913025e-01
8.08335125e-01 -1.16471207e+00 -1.81381822e-01 -1.58632677e-02
-1.06360173e+00 8.12241614e-01 1.84559509e-01 -2.77651893e-03
6.40913472e-02 2.96893805e-01 2.66692396e-02 -2.39059001e-01
-1.01421201e+00 1.55327693e-01 6.82118163e-02 4.97832358e-01
3.87643576e-01 -8.52376148e-02 -2.21820951e-01 9.74688470e-01
-4.48444605e-01 -1.79210771e-02 1.05738175e+00 7.75878727e-01
-6.37937903e-01 -5.96106052e-01 -4.88560170e-01 4.63961549e-02
-6.90636337e-01 3.81980725e-02 -4.17857409e-01 6.48904979e-01
6.74529523e-02 1.23677886e+00 1.61585256e-01 -4.56604242e-01
3.86153042e-01 -3.78793525e-03 -7.98260570e-02 -6.45421565e-01
-9.01430130e-01 -2.59541333e-01 5.50483406e-01 -7.29932785e-01
-3.34245801e-01 -6.36550069e-01 -9.58933234e-01 -8.77384365e-01
2.04672903e-01 4.84133661e-01 6.50747955e-01 7.23658621e-01
3.02315086e-01 3.05258751e-01 3.31852168e-01 -1.15282309e+00
3.80269885e-02 -7.67741263e-01 -4.50387269e-01 5.77275753e-01
4.01674271e-01 -4.87045467e-01 -4.87385362e-01 3.53082448e-01] | [10.378576278686523, 0.7320815920829773] |
8cc96861-838f-46a2-9776-05389ccee8fd | twin-s-a-digital-twin-for-skull-base-surgery | 2211.11863 | null | https://arxiv.org/abs/2211.11863v2 | https://arxiv.org/pdf/2211.11863v2.pdf | Twin-S: A Digital Twin for Skull-base Surgery | Purpose: Digital twins are virtual interactive models of the real world, exhibiting identical behavior and properties. In surgical applications, computational analysis from digital twins can be used, for example, to enhance situational awareness. Methods: We present a digital twin framework for skull-base surgeries, named Twin-S, which can be integrated within various image-guided interventions seamlessly. Twin-S combines high-precision optical tracking and real-time simulation. We rely on rigorous calibration routines to ensure that the digital twin representation precisely mimics all real-world processes. Twin-S models and tracks the critical components of skull-base surgery, including the surgical tool, patient anatomy, and surgical camera. Significantly, Twin-S updates and reflects real-world drilling of the anatomical model in frame rate. Results: We extensively evaluate the accuracy of Twin-S, which achieves an average 1.39 mm error during the drilling process. We further illustrate how segmentation masks derived from the continuously updated digital twin can augment the surgical microscope view in a mixed reality setting, where bone requiring ablation is highlighted to provide surgeons additional situational awareness. Conclusion: We present Twin-S, a digital twin environment for skull-base surgery. Twin-S tracks and updates the virtual model in real-time given measurements from modern tracking technologies. Future research on complementing optical tracking with higher-precision vision-based approaches may further increase the accuracy of Twin-S. | ['Mathias Unberath', 'Adnan Munawar', 'Russell H. Taylor', 'Francis X. Creighton', 'Manish Sahu', 'Nimesh Nagururu', 'Hao Ding', 'Xiangyu Zhang', 'Anna Goodridge', 'Zhaoshuo Li', 'Ruixing Liang', 'Hongchao Shu'] | 2022-11-21 | null | null | null | null | ['mixed-reality'] | ['computer-vision'] | [-4.49947640e-02 1.77961275e-01 2.04223245e-01 3.88885736e-01
-5.55132270e-01 -6.14585757e-01 2.50564933e-01 9.10115615e-02
-3.07894260e-01 2.03222960e-01 -5.60893863e-02 -5.47996283e-01
-4.89143580e-02 -5.08261800e-01 -4.32146400e-01 -3.79521608e-01
-1.02830417e-01 6.63964272e-01 5.27474284e-01 -5.13849735e-01
1.61694616e-01 6.32127166e-01 -1.32924747e+00 2.63943467e-02
5.65448523e-01 6.86840355e-01 8.02979887e-01 8.76038790e-01
2.65826195e-01 1.95366114e-01 -4.01922166e-01 -2.43969664e-01
6.27695501e-01 1.38654843e-01 -1.50839716e-01 -1.68261006e-01
3.21839631e-01 -4.06858295e-01 -2.69775242e-01 7.14558482e-01
6.68873072e-01 -3.54898602e-01 -5.89043535e-02 -7.23872840e-01
2.68957824e-01 1.22837901e-01 -4.35873538e-01 1.88107103e-01
5.56904435e-01 3.73465896e-01 -8.05668533e-02 -8.15592170e-01
1.31722713e+00 6.11169279e-01 1.02049088e+00 4.52105522e-01
-1.25495517e+00 -8.52255464e-01 -1.80780739e-01 -4.01901782e-01
-1.17571831e+00 -3.85844857e-01 2.97045201e-01 -7.42756903e-01
5.39856911e-01 5.62629461e-01 1.52358496e+00 5.67620635e-01
1.04861820e+00 3.41809213e-01 1.06786895e+00 -6.22303843e-01
-8.87567922e-02 1.23309851e-01 -3.59660447e-01 1.07507622e+00
3.56905758e-01 7.91788220e-01 -6.29992485e-01 -9.53089595e-02
1.63317060e+00 1.13206260e-01 -8.68202031e-01 -9.31928396e-01
-1.63001263e+00 1.28896341e-01 1.66979000e-01 2.12434158e-01
-4.21558052e-01 3.60272259e-01 2.32230395e-01 -6.25755191e-02
2.50189126e-01 4.74570900e-01 3.92332822e-02 -4.97948110e-01
-7.95027316e-01 -1.88315555e-01 4.96724844e-01 1.20630825e+00
3.25696282e-02 -2.64517944e-02 9.46190506e-02 3.49079400e-01
3.02747101e-01 5.12573980e-02 6.32016063e-01 -7.55509079e-01
-1.75470620e-01 4.79499876e-01 1.83926057e-02 -4.86104548e-01
-8.16356957e-01 -6.15186274e-01 -2.01137662e-01 6.83793783e-01
1.44045949e-01 -5.02681248e-02 -1.18694770e+00 1.08653533e+00
6.63903177e-01 6.60891831e-01 -4.09666777e-01 7.59980738e-01
7.84107685e-01 -3.97244126e-01 -5.04441559e-01 -3.94621372e-01
1.48359728e+00 -5.01022041e-01 -8.05822909e-01 -2.47276574e-01
8.65550697e-01 -9.57449257e-01 7.98892677e-01 5.15691757e-01
-1.23164666e+00 4.33768705e-02 -1.11646926e+00 4.35917288e-01
1.21890105e-01 1.59695745e-01 6.43294513e-01 9.07278121e-01
-1.24507976e+00 5.02198458e-01 -1.43097866e+00 -2.58202076e-01
3.60450864e-01 8.13068330e-01 -6.51500642e-01 2.41825432e-01
-3.99992853e-01 1.45171082e+00 -1.50849327e-01 3.56964380e-01
-8.56979072e-01 -1.23904979e+00 -8.99223745e-01 -6.29660308e-01
4.25854236e-01 -9.71866310e-01 1.69730818e+00 -9.88366455e-02
-1.76217270e+00 1.16275978e+00 1.05684012e-01 -2.63928622e-01
8.44055057e-01 -1.80032492e-01 -4.46333617e-01 3.98424864e-01
-3.57585549e-02 -9.05592181e-03 5.80524147e-01 -1.33252490e+00
-3.65057886e-01 -6.59787774e-01 8.45838860e-02 2.29707554e-01
2.04644337e-01 -1.13026835e-01 -8.25844228e-01 -7.54373789e-01
6.52766109e-01 -1.32292843e+00 -7.06346512e-01 9.36091006e-01
4.87896940e-03 9.57326114e-01 7.72398651e-01 -4.62789625e-01
1.13591242e+00 -2.00599599e+00 -2.20457003e-01 9.64176133e-02
6.60433769e-01 1.78465068e-01 6.25197530e-01 1.36284828e-01
-1.12485990e-01 -4.18784410e-01 1.42595330e-02 -5.65078616e-01
-6.88115001e-01 1.82081580e-01 1.70744076e-01 8.63112748e-01
-9.79537129e-01 6.88629627e-01 -1.01498532e+00 -5.29513478e-01
6.68201447e-01 3.91775757e-01 -5.70500135e-01 -6.67027012e-02
2.11283207e-01 8.23567569e-01 -1.81617185e-01 7.36460447e-01
5.92830002e-01 8.93272310e-02 3.36418658e-01 -3.57180029e-01
-4.53896314e-01 8.70916899e-03 -1.04563808e+00 2.39583755e+00
-1.03152084e+00 5.08756936e-01 6.14252746e-01 9.16107595e-02
6.21704817e-01 6.37812674e-01 6.75995290e-01 -6.85050786e-01
5.32110989e-01 5.31190336e-01 1.39061520e-02 -4.45551634e-01
6.48475528e-01 -7.79241145e-01 3.03537369e-01 4.03751224e-01
-1.24587446e-01 -1.04443514e+00 -4.01814222e-01 3.21162909e-01
6.43616199e-01 1.53567195e-01 3.49973232e-01 -3.55095237e-01
-1.71178058e-01 1.99627966e-01 2.25675315e-01 5.04839540e-01
-3.97794306e-01 8.65203857e-01 2.99014039e-02 -3.37491304e-01
-5.72837055e-01 -1.20592320e+00 -5.10811090e-01 -7.71419927e-02
7.10798681e-01 -5.58215439e-01 -5.07245600e-01 -3.96733224e-01
-8.63754936e-03 5.78450143e-01 -8.65042031e-01 -2.59876311e-01
-5.02241611e-01 -2.15199754e-01 1.39093608e-01 4.56242085e-01
-4.94552493e-01 -3.07104319e-01 -1.52514768e+00 5.21516383e-01
1.17862940e-01 -1.23580503e+00 -4.01765198e-01 2.39315405e-01
-1.10351968e+00 -1.17806232e+00 -4.91420478e-01 -4.36507761e-01
6.99318588e-01 3.02676946e-01 8.47468078e-01 1.34461880e-01
-9.35946703e-01 8.16774726e-01 -2.13580713e-01 -6.47901297e-01
-4.76744533e-01 -7.99799502e-01 3.30623627e-01 -4.41887230e-01
-5.39325535e-01 -6.14526033e-01 -8.03556740e-01 5.09375572e-01
-5.32971799e-01 6.77238464e-01 2.58866876e-01 5.00078797e-01
4.44782466e-01 -7.26719737e-01 -3.92919958e-01 -6.66136920e-01
4.35772181e-01 1.73404559e-01 -8.40388954e-01 6.52394518e-02
-2.28835031e-01 -3.19868743e-01 2.06766382e-01 -3.30288440e-01
-1.00892365e+00 6.43732399e-02 2.52925545e-01 -9.18011963e-01
3.08608294e-01 4.77838099e-01 6.93466842e-01 -6.00658834e-01
7.23483682e-01 -2.47154132e-01 5.70753157e-01 -8.30247402e-02
4.43145968e-02 1.89403415e-01 1.04387224e+00 -3.99637491e-01
5.89148879e-01 1.14282537e+00 1.46601200e-02 -4.68005329e-01
-2.44897813e-01 -3.47578079e-01 -8.58215332e-01 -7.94373989e-01
5.46037138e-01 -5.43345630e-01 -7.33688116e-01 2.15526208e-01
-1.02873218e+00 -3.94870281e-01 -3.77646804e-01 1.18790638e+00
-6.02180779e-01 4.91980091e-02 -3.46359462e-01 -7.25583494e-01
-2.67680645e-01 -1.80997527e+00 1.38450849e+00 9.54401493e-02
-3.83618087e-01 -1.05302989e+00 2.68703341e-01 1.27987072e-01
1.83725342e-01 8.49470377e-01 2.24833071e-01 2.12533716e-02
-8.08342040e-01 -6.10314727e-01 5.76014876e-01 -4.72107142e-01
2.74407953e-01 2.35786811e-01 -8.58056843e-01 -2.70548433e-01
1.51654169e-01 4.99505848e-01 2.39274334e-02 8.45563293e-01
7.90101886e-01 4.57577229e-01 -8.02788615e-01 1.00879133e+00
1.36464548e+00 4.42184389e-01 3.23735505e-01 5.31933844e-01
3.28925133e-01 2.66111672e-01 9.77338970e-01 5.92771828e-01
1.91565275e-01 1.11018538e+00 6.51879907e-01 -1.36270449e-01
-2.25207895e-01 9.61524844e-02 -5.27574569e-02 8.55321705e-01
-2.24199831e-01 4.88329291e-01 -1.31258678e+00 3.19362789e-01
-1.30634606e+00 -4.76475775e-01 -2.91418403e-01 2.67087221e+00
4.61737514e-01 2.01163366e-01 -4.33209151e-01 -2.15260282e-01
6.04912341e-01 -2.21941665e-01 -2.90142894e-01 -2.70436943e-01
5.21991551e-01 4.61582124e-01 7.31102288e-01 4.68695194e-01
-5.65119982e-01 6.58333480e-01 6.71505690e+00 2.91189939e-01
-1.54723692e+00 2.79957145e-01 -2.51514405e-01 -6.38652205e-01
-2.31255382e-01 1.55198589e-01 -2.29762688e-01 7.54545629e-02
6.00979269e-01 -5.48010468e-01 -1.22894369e-01 5.18335342e-01
3.48074883e-01 -6.12555385e-01 -1.13140738e+00 1.17369068e+00
2.02089265e-01 -2.00694013e+00 -2.77747244e-01 3.65193039e-01
3.79377902e-01 -1.11405917e-01 1.00821480e-01 -2.24941865e-01
2.75623769e-01 -6.53278947e-01 9.47899640e-01 5.79700351e-01
1.30565572e+00 -2.33746842e-01 3.83856773e-01 3.32717896e-01
-1.20334029e+00 1.25733331e-01 3.89295608e-01 2.16303945e-01
7.77068555e-01 3.75319690e-01 -1.40241385e+00 7.94329226e-01
5.94273686e-01 4.73291159e-01 -1.41424686e-01 1.48749804e+00
3.51862222e-01 -2.89064318e-01 -4.89745885e-01 6.00601077e-01
-1.06207199e-01 -3.11791480e-01 9.47026014e-01 7.14390099e-01
3.24737906e-01 4.91982788e-01 -2.11577892e-01 2.96357065e-01
3.57828915e-01 -2.69544572e-01 -6.17832184e-01 5.68583369e-01
3.60455424e-01 1.21207929e+00 -1.02284086e+00 -1.74826622e-01
-8.41945708e-02 7.96971738e-01 -1.77104577e-01 -2.72699922e-01
-8.50500107e-01 1.86233982e-01 7.48551190e-01 5.19351959e-01
-1.92432478e-01 -5.21569073e-01 -6.10948026e-01 -9.66772377e-01
6.47131726e-02 -3.95135909e-01 6.72400594e-02 -1.33084333e+00
7.73116797e-02 7.75668621e-01 -1.25209481e-01 -2.03695798e+00
7.19588473e-02 -5.82141459e-01 -5.09135067e-01 4.88624156e-01
-9.97583747e-01 -1.05140555e+00 -7.52623320e-01 5.44947207e-01
4.91396308e-01 3.46417904e-01 9.34547305e-01 -2.45981757e-02
-4.24929559e-01 1.87517866e-01 1.20039158e-01 -2.85369068e-01
8.55991662e-01 -9.81622636e-01 2.53036648e-01 8.04725111e-01
-1.98808640e-01 8.59864533e-01 7.48089075e-01 -8.63178611e-01
-1.52147615e+00 -4.69379485e-01 -2.63936728e-01 -1.02267349e+00
6.02716923e-01 -3.29004943e-01 -2.64286488e-01 1.00946307e+00
-2.78718233e-01 4.13631439e-01 7.29639292e-01 -2.82783777e-01
2.72093654e-01 8.71131793e-02 -1.31783259e+00 9.34743881e-01
1.10555410e+00 -5.20145357e-01 -6.03707850e-01 4.10484433e-01
4.81622845e-01 -1.79723263e+00 -9.49899793e-01 4.22627717e-01
1.11779165e+00 -1.13834989e+00 9.36427176e-01 5.48929051e-02
7.47245699e-02 -2.33159348e-01 3.70461524e-01 -1.50180185e+00
2.33808950e-01 -1.04205596e+00 1.37030810e-01 5.60643487e-02
9.95097607e-02 -9.23687875e-01 8.30234230e-01 9.07772958e-01
-8.59328687e-01 -8.23284864e-01 -1.40057611e+00 -6.17926598e-01
-4.38783199e-01 -8.65952909e-01 2.42707431e-01 7.65913844e-01
4.45480198e-01 -8.30708444e-01 2.71377265e-01 5.70587695e-01
4.64410633e-01 6.51478171e-02 7.38741279e-01 -1.22425616e+00
-8.05746168e-02 -3.21331561e-01 -7.79277444e-01 -5.43811083e-01
-5.26206613e-01 -6.24415278e-01 -1.49726078e-01 -1.55466771e+00
-3.88082832e-01 -8.14912915e-01 2.32068822e-01 -3.06554884e-02
2.74064124e-01 2.67216206e-01 1.69771641e-01 2.93024331e-01
3.76722485e-01 3.66544515e-01 1.87176168e+00 5.75106084e-01
-3.34509015e-01 -1.97569677e-03 -2.76363015e-01 1.06781268e+00
2.47984499e-01 -3.44205081e-01 -2.44919628e-01 -6.39408648e-01
-1.91593036e-01 4.77502465e-01 3.97774696e-01 -1.06151509e+00
8.37427855e-01 1.58523262e-01 6.77406788e-02 -5.54605782e-01
8.45312715e-01 -1.21993053e+00 7.31564760e-01 9.60389495e-01
5.17180443e-01 1.37581408e-01 7.01107740e-01 4.84901547e-01
1.08948909e-01 7.37127513e-02 8.95398021e-01 -3.87475371e-01
-2.89023221e-01 3.60651493e-01 -2.55070597e-01 -3.95357579e-01
1.57537365e+00 -1.12787259e+00 -1.94526613e-01 1.12716411e-03
-1.26265919e+00 -1.56078100e-01 7.81128764e-01 2.58197278e-01
1.01342499e+00 -7.82215059e-01 -3.96961778e-01 5.17200232e-01
1.81453779e-01 3.75236988e-01 5.30297935e-01 1.71384299e+00
-1.41987252e+00 1.59347221e-01 -2.36833677e-01 -9.93140638e-01
-1.48141396e+00 2.72694975e-01 1.10851407e+00 1.36635751e-01
-1.20607972e+00 9.60821986e-01 4.99846399e-01 -4.13909554e-01
-1.94781318e-01 -4.69063669e-01 3.55807185e-01 -2.42882207e-01
4.16430503e-01 -8.43518749e-02 6.19379520e-01 -3.94605786e-01
-5.37589610e-01 8.49358499e-01 -2.64655232e-01 -3.77604753e-01
1.24310076e+00 -2.14389101e-01 1.70449853e-01 2.14403450e-01
3.67938787e-01 3.31863135e-01 -1.03657854e+00 3.29304748e-04
-5.70329368e-01 -9.76175010e-01 5.82820296e-01 -8.44681382e-01
-1.15299964e+00 3.84564638e-01 8.27950060e-01 -6.28464758e-01
9.09181416e-01 -1.41650423e-01 4.94363159e-01 -4.43563342e-01
1.20464063e+00 -7.72900879e-01 -3.09951842e-01 8.42352980e-04
1.06705940e+00 -6.60528719e-01 2.00497627e-01 -8.57315421e-01
-4.84466612e-01 1.22391987e+00 6.84906960e-01 1.52685866e-01
8.77657950e-01 1.14290428e+00 3.70597154e-01 -7.46843457e-01
-4.56681162e-01 2.46395975e-01 -1.00306377e-01 5.86697280e-01
2.59022385e-01 2.13542134e-01 3.94648202e-02 6.98777735e-02
-6.40008748e-01 1.85069978e-01 8.76246393e-01 1.44938922e+00
1.47310980e-02 -6.40855074e-01 -7.03916848e-01 4.78995204e-01
-9.37291905e-02 -1.19724594e-01 2.97531992e-01 9.59660590e-01
-3.84007357e-02 6.05043113e-01 4.89158444e-02 -1.21128619e-01
7.18575001e-01 -5.47582209e-01 8.93835247e-01 -8.67494285e-01
-1.03686357e+00 1.81684017e-01 3.40873711e-02 -1.09155548e+00
-3.86458039e-02 -6.84638321e-01 -1.22435188e+00 -7.70091824e-03
-7.88381398e-01 -1.87908143e-01 1.10011208e+00 6.41158164e-01
2.20224544e-01 9.01427805e-01 1.75839484e-01 -1.24113297e+00
1.71253994e-01 -3.38140190e-01 -5.96561611e-01 -2.30037585e-01
5.75325191e-01 -1.10285127e+00 -1.74565792e-01 -1.55821726e-01] | [13.743624687194824, -3.025155544281006] |
ec243af8-d6c9-4708-99a7-8f1cb4105445 | patch-based-medical-image-segmentation-using | 2109.07138 | null | https://arxiv.org/abs/2109.07138v2 | https://arxiv.org/pdf/2109.07138v2.pdf | Patch-based Medical Image Segmentation using Matrix Product State Tensor Networks | Tensor networks are efficient factorisations of high-dimensional tensors into a network of lower-order tensors. They have been most commonly used to model entanglement in quantum many-body systems and more recently are witnessing increased applications in supervised machine learning. In this work, we formulate image segmentation in a supervised setting with tensor networks. The key idea is to first lift the pixels in image patches to exponentially high-dimensional feature spaces and using a linear decision hyper-plane to classify the input pixels into foreground and background classes. The high-dimensional linear model itself is approximated using the matrix product state (MPS) tensor network. The MPS is weight-shared between the non-overlapping image patches resulting in our strided tensor network model. The performance of the proposed model is evaluated on three 2D- and one 3D- biomedical imaging datasets. The performance of the proposed tensor network segmentation model is compared with relevant baseline methods. In the 2D experiments, the tensor network model yields competitive performance compared to the baseline methods while being more resource efficient. | ['Jens Petersen', 'Søren Alexander Flensborg', 'Erik B Dam', 'Raghavendra Selvan'] | 2021-09-15 | null | null | null | null | ['tensor-networks'] | ['methodology'] | [ 3.56286347e-01 2.37890020e-01 -6.40855506e-02 -2.07125887e-01
-4.66807514e-01 -4.37736690e-01 4.55194294e-01 -2.23110378e-01
-6.48718417e-01 2.97662377e-01 9.64656174e-02 -3.14205766e-01
-2.14404404e-01 -3.78933012e-01 -3.62330377e-01 -1.21653438e+00
-4.19268459e-01 5.89921117e-01 2.71588027e-01 2.03829736e-01
1.34517133e-01 5.48463404e-01 -9.11016166e-01 1.65665776e-01
7.27000892e-01 1.02483678e+00 -3.87653798e-01 8.44641566e-01
2.18808979e-01 7.33684003e-01 -4.65401448e-02 -5.17733335e-01
5.24848402e-01 -3.30455184e-01 -1.25819063e+00 1.35217547e-01
5.11125565e-01 -1.49515003e-01 -7.54688621e-01 1.38884151e+00
2.82524824e-01 1.55138567e-01 6.97475135e-01 -1.05671394e+00
-3.33226323e-01 4.29137379e-01 -2.96939284e-01 4.71276760e-01
-2.07946047e-01 5.27662113e-02 1.15486062e+00 -6.25996292e-01
6.81332231e-01 1.17555177e+00 2.87744820e-01 5.70534587e-01
-1.69160545e+00 -1.96011245e-01 -5.78617811e-01 2.70866364e-01
-1.09914052e+00 -3.51480246e-02 8.78474116e-01 -6.57618940e-01
8.52705717e-01 3.78029972e-01 7.83933938e-01 6.20790184e-01
7.17691779e-01 7.76477396e-01 1.50433588e+00 -4.02266353e-01
2.16726556e-01 -2.30571091e-01 5.77376246e-01 1.11465228e+00
1.15848579e-01 -1.17083117e-01 -6.08873725e-01 -5.69087088e-01
8.68486822e-01 -2.39434224e-02 2.02587433e-02 -5.58969438e-01
-1.88275528e+00 6.81210160e-01 7.74090409e-01 4.39843655e-01
-4.77274835e-01 5.10235548e-01 4.97798383e-01 -1.41561493e-01
5.70365071e-01 4.26975846e-01 7.11841807e-02 8.44733194e-02
-9.34458315e-01 2.02554733e-01 7.27115571e-01 6.22075379e-01
7.24591255e-01 -2.75885552e-01 -2.81124830e-01 4.94079858e-01
4.40630496e-01 4.36111659e-01 1.19318597e-01 -1.02623975e+00
2.70145863e-01 6.01114511e-01 -8.71885866e-02 -9.97001052e-01
-6.40855670e-01 -3.26369017e-01 -1.35365033e+00 -1.91051304e-01
5.30980468e-01 7.29494765e-02 -1.00271106e+00 1.36204267e+00
6.66937530e-01 4.05378968e-01 1.62872940e-01 1.08891737e+00
3.85289520e-01 6.24429047e-01 -2.03339905e-01 -9.50247720e-02
1.74816501e+00 -9.67504561e-01 -8.09028268e-01 3.30689222e-01
8.33357275e-01 -7.61664927e-01 3.13979149e-01 4.96180385e-01
-7.12680101e-01 -7.76510984e-02 -9.08824980e-01 -2.64808834e-01
-1.34999409e-01 2.46472925e-01 1.16612136e+00 8.85335088e-01
-9.14939523e-01 9.39540744e-01 -1.29374123e+00 -1.93144709e-01
5.52370310e-01 6.78305686e-01 -4.86726791e-01 -4.94231246e-02
-1.00668347e+00 5.90662241e-01 4.17174160e-01 6.91823959e-01
-7.48467326e-01 -3.69217038e-01 -5.57661653e-01 -4.21333522e-01
3.78704928e-02 -8.74484599e-01 8.30901206e-01 -1.31058514e-01
-1.54413056e+00 9.37715411e-01 -6.32970035e-03 -4.24322128e-01
1.99218839e-01 2.50346303e-01 -4.40054014e-02 4.40588266e-01
8.03552195e-03 5.76211870e-01 8.93528879e-01 -7.36938179e-01
-1.69310421e-01 -5.91315687e-01 2.80006260e-01 9.07539204e-02
-6.19822815e-02 -8.33390877e-02 -1.98030621e-01 -2.02016756e-01
8.57121587e-01 -1.50465715e+00 -5.57726681e-01 -2.88650185e-01
-9.18530762e-01 -2.79811084e-01 6.13942742e-01 -4.07961100e-01
9.16037321e-01 -1.93223643e+00 7.14402854e-01 2.59590179e-01
7.76431441e-01 1.88123155e-02 3.64548340e-02 1.21384293e-01
-8.72652903e-02 1.12858824e-01 -1.40930966e-01 -3.47363830e-01
1.09044485e-01 5.65993428e-01 1.50624171e-01 9.85033631e-01
1.71459168e-01 9.30187941e-01 -8.98751497e-01 -7.03327298e-01
7.00729340e-02 4.76144582e-01 -6.40034914e-01 -1.15989774e-01
1.07384417e-02 7.43691146e-01 -6.38800085e-01 3.93688977e-01
8.54317605e-01 -4.50150996e-01 5.53111769e-02 -9.96437073e-01
-4.90849512e-03 9.99190286e-03 -9.85109091e-01 1.86587143e+00
8.78032893e-02 5.29009402e-01 -4.70596552e-02 -1.13038254e+00
3.64727974e-01 2.96155989e-01 7.43485987e-01 -3.33687186e-01
3.61753404e-01 2.94941634e-01 5.40761232e-01 -5.71167767e-01
5.14382422e-01 -4.02373135e-01 -2.02594355e-01 5.31361163e-01
4.31982219e-01 -3.69400859e-01 3.63866866e-01 4.61329579e-01
1.04509854e+00 -7.29424804e-02 -1.72544494e-01 -6.12017035e-01
6.00213110e-01 1.00002969e-02 5.19216835e-01 5.12174547e-01
-4.70375419e-01 2.20582753e-01 9.83733892e-01 -8.06333423e-01
-1.21781290e+00 -8.48664403e-01 -4.64057148e-01 4.45261478e-01
1.61689252e-01 -3.47364873e-01 -1.10767436e+00 -6.92655563e-01
-3.10725391e-01 1.18127473e-01 -7.15922058e-01 -8.30542110e-03
-4.81037498e-01 -1.21625268e+00 7.47362971e-01 -1.43123209e-01
6.79690778e-01 -4.76012021e-01 -3.80301923e-01 4.35311757e-02
-2.62728721e-01 -1.42922533e+00 -4.66125935e-01 2.72061825e-01
-1.06211710e+00 -9.72628236e-01 -7.40650058e-01 -4.26369965e-01
1.03497124e+00 8.88666362e-02 6.30884528e-01 -1.82720631e-01
-4.81642783e-01 3.00993025e-01 -1.66798204e-01 7.67441466e-02
-1.73975796e-01 8.57654437e-02 4.29563403e-01 5.56111991e-01
1.45350397e-01 -2.01214343e-01 -7.38225639e-01 1.87441990e-01
-1.26186347e+00 2.99330115e-01 5.96109807e-01 1.19590425e+00
6.62269652e-01 2.16516722e-02 -3.82061809e-01 -7.82529175e-01
4.24870491e-01 -1.28226519e-01 -5.62496126e-01 1.88917175e-01
-1.35066763e-01 6.95532143e-01 2.12837204e-01 -2.89510936e-01
-6.84613466e-01 1.90605015e-01 3.14109623e-01 -2.77802348e-01
1.48069680e-01 4.09968227e-01 3.49870831e-01 -6.77303612e-01
4.41081554e-01 -1.51637243e-02 -1.71598107e-01 -2.10261077e-01
3.09544504e-01 4.26915377e-01 2.25857675e-01 -9.56609845e-01
8.72803628e-01 7.72846103e-01 8.51533890e-01 -1.01863909e+00
-9.73030210e-01 -6.25762403e-01 -1.27642059e+00 -2.57240117e-01
1.45815456e+00 -3.19919616e-01 -8.84385943e-01 7.14150190e-01
-1.49056220e+00 -5.40305935e-02 -1.30658761e-01 8.84707451e-01
-4.74238813e-01 6.58029556e-01 -9.28952098e-01 -8.18441629e-01
-1.75267950e-01 -1.63823462e+00 1.05478477e+00 2.74983227e-01
5.27468503e-01 -1.07960153e+00 2.84809232e-01 7.69128203e-01
1.43145099e-01 3.66620719e-01 1.09836423e+00 -4.58897620e-01
-8.98431361e-01 -1.12846650e-01 -3.43955338e-01 6.08899534e-01
-3.52218926e-01 -3.10545027e-01 -8.78748357e-01 -8.67621228e-02
1.01566680e-01 -3.02703112e-01 7.46329606e-01 2.58003771e-01
7.73522139e-01 -2.59044301e-02 -2.59945691e-01 5.05719543e-01
1.20139050e+00 -2.76711673e-01 4.74245667e-01 -1.42117396e-01
1.05882275e+00 5.26721895e-01 1.95607752e-01 -7.98387229e-02
3.05691272e-01 6.63666666e-01 4.27321225e-01 -1.27774373e-01
6.77376091e-02 2.81116545e-01 5.68084158e-02 1.40868771e+00
-4.72768277e-01 2.58288115e-01 -1.08293259e+00 2.63501674e-01
-1.91889167e+00 -6.89956903e-01 -6.88847184e-01 1.98585069e+00
5.47360003e-01 -7.72312209e-02 -1.05415538e-01 -1.19596273e-01
4.09279138e-01 1.49181455e-01 -3.99720401e-01 -3.76787245e-01
-8.79042894e-02 1.90782726e-01 7.92640686e-01 3.64564002e-01
-1.28150117e+00 8.54693651e-01 6.04207563e+00 5.51685691e-01
-1.11327255e+00 4.36120450e-01 6.37875617e-01 1.15490049e-01
2.34820065e-03 2.48589411e-01 -4.48610663e-01 -4.17027026e-02
9.21257436e-01 2.44995415e-01 5.72951198e-01 2.70990044e-01
9.66383889e-02 -2.76444376e-01 -1.15037680e+00 1.31878269e+00
-1.25744373e-01 -1.38946104e+00 4.96134497e-02 5.20860434e-01
8.65742922e-01 3.41937423e-01 1.19061656e-01 -8.30006078e-02
-9.11031812e-02 -8.78640234e-01 5.05185544e-01 7.05205500e-01
5.55852592e-01 -4.81376499e-01 8.68395567e-01 2.36236840e-01
-8.31053734e-01 3.09171677e-01 -4.46151912e-01 1.32711232e-01
1.83474228e-01 7.86752403e-01 -4.68778253e-01 8.02791595e-01
4.93215889e-01 4.46992040e-01 -3.35275143e-01 1.01382434e+00
1.80377886e-01 6.52233183e-01 -4.96352136e-01 -8.88451040e-02
6.27496958e-01 -8.51791799e-01 6.24502838e-01 9.46441233e-01
-5.52015565e-02 2.75995761e-01 7.90487230e-02 9.20604467e-01
-1.77031487e-01 -1.15811145e-02 -3.77852798e-01 -5.42403936e-01
-3.96384507e-01 1.69676900e+00 -1.04037786e+00 -3.05763006e-01
-1.50645822e-01 1.16316390e+00 1.64273635e-01 3.21437657e-01
-6.45204306e-01 -3.77007782e-01 5.25171816e-01 -3.69222641e-01
-3.44556235e-02 -6.18734777e-01 2.05550399e-02 -1.60069454e+00
-9.75785702e-02 -4.66406375e-01 -1.60029214e-02 -3.72106403e-01
-1.37610507e+00 6.47224724e-01 7.00748712e-02 -9.27804947e-01
2.35907555e-01 -1.12868237e+00 -6.82931691e-02 7.95733392e-01
-1.27978206e+00 -1.32515597e+00 1.11428209e-01 5.07387757e-01
-2.56662160e-01 2.13603646e-01 9.10320222e-01 2.64723003e-01
-8.04679275e-01 3.86034660e-02 3.16705704e-01 3.53047907e-01
1.74641579e-01 -1.47258162e+00 1.72576129e-01 8.48682404e-01
2.15729848e-01 8.89755905e-01 4.83727306e-01 -4.77984637e-01
-1.75981450e+00 -7.53932595e-01 7.14300215e-01 -3.31275493e-01
9.89619613e-01 -6.88901842e-01 -5.80050945e-01 5.00507593e-01
2.06449583e-01 8.20551813e-01 7.78133392e-01 -1.44500598e-01
-3.61638516e-01 -7.67793553e-03 -1.33797204e+00 2.83785462e-01
6.23195648e-01 -7.89674163e-01 -1.84447140e-01 9.00722504e-01
3.06988657e-01 -5.21438241e-01 -1.31300533e+00 -1.32343024e-01
4.91006881e-01 -8.74618471e-01 7.01797783e-01 -7.12812781e-01
3.89086381e-02 -5.87727666e-01 -8.37668628e-02 -1.11784446e+00
-4.12938744e-01 -1.08292007e+00 8.06559399e-02 4.38125759e-01
1.87597290e-01 -5.73879182e-01 9.37062621e-01 7.55564749e-01
-8.17351267e-02 -8.41858864e-01 -1.50159323e+00 -5.02526641e-01
8.21220689e-03 -4.00944710e-01 1.47003621e-01 8.07204008e-01
2.32680693e-01 5.60507953e-01 -3.50278080e-01 2.76924968e-01
1.31646109e+00 -2.90526688e-01 4.28206623e-01 -1.16475534e+00
-2.27837503e-01 -1.17048293e-01 -1.01434350e+00 -1.12323165e+00
4.17469814e-02 -1.35739696e+00 -1.40064865e-01 -1.17889833e+00
6.24535680e-01 -4.48432654e-01 -2.79273659e-01 2.10130140e-01
2.74176717e-01 5.76353967e-01 8.23300034e-02 3.92823219e-01
-8.04238558e-01 3.96272540e-01 1.63729334e+00 -2.65695095e-01
1.36282206e-01 -1.86469808e-01 3.95737626e-02 4.51568544e-01
5.10931849e-01 -6.56847656e-01 -2.70149112e-02 -3.35147113e-01
3.82204950e-01 7.13816583e-02 5.43459594e-01 -9.50119734e-01
4.89790171e-01 7.84747526e-02 6.73954487e-02 -5.71030021e-01
4.14608508e-01 -6.26600146e-01 7.00973198e-02 6.10699415e-01
-1.43008590e-01 -2.81684279e-01 -1.23936966e-01 5.48889697e-01
-1.28459530e-02 -3.59944463e-01 9.05133188e-01 -8.41473714e-02
-1.41320214e-01 5.51090956e-01 -2.09547147e-01 -2.74517745e-01
1.02492642e+00 8.01886469e-02 -2.59483218e-01 2.58600980e-01
-9.86526608e-01 -2.03291789e-01 1.49831325e-01 -5.27515076e-02
4.55820411e-01 -1.40941858e+00 -2.88985908e-01 6.93948641e-02
6.97956979e-02 -4.82049659e-02 5.14032245e-01 1.59943509e+00
-8.81116450e-01 8.51628959e-01 -1.97155923e-01 -1.07289898e+00
-1.11722004e+00 1.57095537e-01 7.18512595e-01 -6.26258492e-01
-4.78956580e-01 7.69769073e-01 1.99159998e-02 -7.00835824e-01
-1.59361422e-01 -8.28136861e-01 1.01692662e-01 -3.89307976e-01
3.38066399e-01 3.95472974e-01 -1.07713901e-01 -1.12857497e+00
-3.06310922e-01 6.01213872e-01 -1.80028841e-01 -2.30564743e-01
1.29688060e+00 1.45651057e-01 -8.90831292e-01 5.02422452e-01
1.45260572e+00 -4.17484283e-01 -8.10964882e-01 -4.97019410e-01
3.26815285e-02 -1.23094849e-01 5.31240106e-01 -2.10145101e-01
-9.67947960e-01 1.10217130e+00 8.00739527e-01 4.34710443e-01
5.55598378e-01 2.08101183e-01 8.90796363e-01 5.19022584e-01
4.47728604e-01 -8.55204225e-01 1.52560502e-01 5.31725883e-01
3.32272649e-01 -1.01647747e+00 7.16891885e-02 -4.53123510e-01
-3.65736634e-01 1.36890519e+00 -1.06056452e-01 -3.63177955e-01
6.51007593e-01 -1.54008731e-01 -2.33373135e-01 -7.43343771e-01
-1.75408065e-01 6.32166415e-02 6.66996598e-01 3.27942818e-01
3.71700794e-01 3.86458725e-01 -1.72392517e-01 -2.32075620e-02
-6.29261211e-02 -9.04252976e-02 6.34051323e-01 6.27009630e-01
6.42874986e-02 -1.18810010e+00 -4.47779864e-01 5.49222708e-01
-6.53950274e-01 -2.97426488e-02 -2.18956489e-02 2.91116506e-01
2.58461982e-01 6.95770562e-01 -1.45979553e-01 -3.91436368e-01
1.16802149e-01 3.55100483e-02 9.43155706e-01 -5.30721128e-01
-4.42198753e-01 1.40122429e-01 -1.22088417e-01 -7.72484243e-01
-8.05258036e-01 -8.53568375e-01 -1.25688279e+00 -1.26852050e-01
-5.40835857e-01 2.76353002e-01 1.03770137e+00 1.04561567e+00
2.42027834e-01 4.63019043e-01 4.02702540e-01 -1.10115027e+00
-6.88123167e-01 -8.54214072e-01 -7.35078573e-01 1.80159941e-01
4.38456029e-01 -6.78193212e-01 -3.43651652e-01 -1.58151947e-02] | [5.835435390472412, 4.976510047912598] |
e7c9151f-80ca-409a-9987-cfa4c17db8ee | selective-text-augmentation-with-word-roles | 2209.01560 | null | https://arxiv.org/abs/2209.01560v1 | https://arxiv.org/pdf/2209.01560v1.pdf | Selective Text Augmentation with Word Roles for Low-Resource Text Classification | Data augmentation techniques are widely used in text classification tasks to improve the performance of classifiers, especially in low-resource scenarios. Most previous methods conduct text augmentation without considering the different functionalities of the words in the text, which may generate unsatisfactory samples. Different words may play different roles in text classification, which inspires us to strategically select the proper roles for text augmentation. In this work, we first identify the relationships between the words in a text and the text category from the perspectives of statistical correlation and semantic similarity and then utilize them to divide the words into four roles -- Gold, Venture, Bonus, and Trivial words, which have different functionalities for text classification. Based on these word roles, we present a new augmentation technique called STA (Selective Text Augmentation) where different text-editing operations are selectively applied to words with specific roles. STA can generate diverse and relatively clean samples, while preserving the original core semantics, and is also quite simple to implement. Extensive experiments on 5 benchmark low-resource text classification datasets illustrate that augmented samples produced by STA successfully boost the performance of classification models which significantly outperforms previous non-selective methods, including two large language model-based techniques. Cross-dataset experiments further indicate that STA can help the classifiers generalize better to other datasets than previous methods. | ['Hailiang Huang', 'Songqiao Han', 'Biyang Guo'] | 2022-09-04 | null | null | null | null | ['text-augmentation'] | ['natural-language-processing'] | [ 3.25280368e-01 -2.38609642e-01 -2.93465555e-01 -3.83323491e-01
-1.68659657e-01 -2.13986322e-01 7.33322024e-01 3.67868960e-01
-6.24907613e-01 7.81389713e-01 5.19432902e-01 -2.42392868e-01
1.90603331e-01 -8.46836567e-01 -2.51056608e-02 -8.07879806e-01
5.88718534e-01 4.74623710e-01 1.14954144e-01 -6.27074003e-01
2.89056420e-01 1.25695378e-01 -1.64731634e+00 5.31454623e-01
1.16916156e+00 6.16278410e-01 1.78661525e-01 2.59929597e-02
-8.91308069e-01 3.62995863e-01 -8.09076607e-01 -4.77588415e-01
7.40216300e-02 -3.94220054e-01 -7.20092058e-01 9.69899073e-02
3.43916402e-03 -4.40003313e-02 -1.51466891e-01 1.00224650e+00
3.93854916e-01 2.81300008e-01 6.76701844e-01 -1.17645860e+00
-7.45082736e-01 1.16830337e+00 -6.32584512e-01 5.42368852e-02
1.91301838e-01 -2.96444654e-01 1.08970022e+00 -1.12545252e+00
2.24510759e-01 1.35215509e+00 7.64706910e-01 5.75227857e-01
-1.11157942e+00 -1.01225746e+00 6.25540495e-01 7.08306059e-02
-1.35072958e+00 -2.60572881e-01 7.68338144e-01 -2.40562439e-01
6.33303940e-01 5.21775723e-01 4.14113998e-01 1.24006426e+00
-1.23488098e-01 1.00399470e+00 1.00431204e+00 -6.90255404e-01
-8.00627172e-02 4.62530792e-01 5.71755707e-01 2.43994057e-01
5.55062473e-01 -4.90054429e-01 -4.49170321e-01 -1.43475324e-01
1.83410212e-01 3.00038457e-01 -3.38914067e-01 -3.22145447e-02
-1.41669655e+00 8.79243374e-01 1.81320101e-01 6.00693166e-01
-1.04216211e-01 -4.40030009e-01 7.40888715e-01 1.73237875e-01
8.39814961e-01 6.23893440e-01 -7.61092782e-01 1.02486156e-01
-4.20898527e-01 1.46601722e-01 4.19587791e-01 9.59338427e-01
6.45718336e-01 5.32861948e-02 -5.78913391e-01 1.17653728e+00
1.17043592e-01 2.97452509e-01 1.24024105e+00 1.95082456e-01
6.25575483e-01 1.11820352e+00 -1.41184673e-01 -8.73188853e-01
-4.08111662e-01 -5.81879735e-01 -1.18304420e+00 -4.72105473e-01
5.27495220e-02 -6.96248785e-02 -1.04012108e+00 1.59043002e+00
3.03503513e-01 -1.10188410e-01 1.32651955e-01 4.39047426e-01
9.98984277e-01 5.84878504e-01 2.55299777e-01 -3.16251487e-01
1.59583855e+00 -9.32502806e-01 -1.14607084e+00 -4.23586518e-01
1.30712152e+00 -8.09095681e-01 1.66123879e+00 2.51856893e-01
-3.19555461e-01 -5.67600012e-01 -9.94307995e-01 1.94586650e-01
-8.32240224e-01 3.72939527e-01 7.08210528e-01 1.00469422e+00
-3.69852871e-01 3.28818381e-01 -1.48202136e-01 -2.17414349e-01
4.52939063e-01 1.88829958e-01 -2.97459692e-01 -2.17340537e-03
-1.59929967e+00 6.71832800e-01 8.31992447e-01 -4.86993194e-02
-7.64856637e-02 -6.12082899e-01 -8.31162095e-01 -6.27980158e-02
4.89494413e-01 -3.86283755e-01 9.36870873e-01 -9.66640711e-01
-1.03105903e+00 5.92489183e-01 -2.97775149e-01 -1.93564326e-01
3.57771754e-01 -1.37732983e-01 -6.81692123e-01 -4.61355031e-01
1.96084306e-01 4.06009227e-01 9.01154935e-01 -1.23776007e+00
-8.32019210e-01 -3.63064438e-01 -2.28441268e-01 3.90536159e-01
-1.34808707e+00 -8.62331390e-02 -4.55126524e-01 -1.24259186e+00
2.96012372e-01 -7.91951716e-01 -1.94458798e-01 -4.96104270e-01
-6.73059046e-01 -3.76275450e-01 1.00963032e+00 -3.66696358e-01
1.56456757e+00 -2.06684875e+00 -1.86169222e-01 6.99335039e-02
1.94195509e-01 4.61041063e-01 -1.40631601e-01 2.62424469e-01
-3.30997288e-01 4.83654410e-01 -1.49171665e-01 -5.00291467e-01
-3.06414336e-01 2.86442369e-01 -5.14216959e-01 -3.58055569e-02
-2.44786660e-03 8.25560927e-01 -7.55215287e-01 -5.80289364e-01
7.46591687e-02 1.00245580e-01 -3.92425597e-01 -8.39814171e-02
-2.38425583e-01 2.67765164e-01 -7.36890793e-01 4.38191175e-01
6.32694066e-01 -9.95130688e-02 4.76133153e-02 -3.32646787e-01
2.43131205e-01 3.01452756e-01 -1.17589200e+00 1.14288235e+00
-5.08776248e-01 3.47710967e-01 -3.86598587e-01 -9.77289498e-01
1.10312772e+00 6.14370666e-02 2.17847213e-01 -4.76937234e-01
3.19249600e-01 3.63793448e-02 3.28297429e-02 -3.57551873e-01
9.33185756e-01 -2.66841531e-01 -1.14849530e-01 6.18990362e-01
-2.79370487e-01 9.02860612e-02 4.30921346e-01 2.14087471e-01
5.91743886e-01 -2.17534438e-01 2.27495626e-01 -2.74039984e-01
7.07510352e-01 -1.29884128e-02 5.12092590e-01 7.30525792e-01
5.30693941e-02 3.64228249e-01 3.20582360e-01 -3.48063529e-01
-8.41338754e-01 -3.38119328e-01 -2.16410354e-01 1.54491246e+00
1.94624454e-01 -6.89098716e-01 -4.81210917e-01 -1.10111845e+00
-5.24332412e-02 1.01566947e+00 -9.79628384e-01 -5.94507277e-01
-3.61252636e-01 -1.31614530e+00 5.01071095e-01 6.46874368e-01
5.37402213e-01 -9.93039131e-01 2.31640339e-01 3.71901989e-02
-4.35088426e-01 -9.64286268e-01 -6.39102340e-01 2.10541576e-01
-8.73007298e-01 -8.06864202e-01 -5.17566681e-01 -8.01812947e-01
1.00841856e+00 7.32285619e-01 8.69375944e-01 2.73393005e-01
6.20002076e-02 -2.17547193e-01 -1.10058236e+00 -7.40962982e-01
-5.94221175e-01 4.55212533e-01 2.03839436e-01 2.73067325e-01
7.15122104e-01 -1.52887374e-01 -8.47663507e-02 5.25980651e-01
-1.15534627e+00 3.49649310e-01 5.30059338e-01 1.17815781e+00
4.00963575e-01 2.84326553e-01 7.44394898e-01 -1.34319329e+00
1.06743872e+00 -3.41813326e-01 1.46004250e-02 3.69849026e-01
-8.33133101e-01 1.81903362e-01 9.77817178e-01 -8.10731411e-01
-1.15783417e+00 -2.59557277e-01 -4.88203615e-02 -9.70150232e-02
4.84206490e-02 5.48907042e-01 -4.86642212e-01 3.10566276e-01
7.67469108e-01 5.14746189e-01 -6.91837743e-02 -4.80455846e-01
2.27341026e-01 9.81308937e-01 -1.16987355e-01 -4.87561613e-01
8.20996344e-01 4.02734101e-01 -2.86869854e-01 -8.19914699e-01
-1.15380573e+00 -5.23470938e-01 -7.54817545e-01 1.63829476e-01
3.99888575e-01 -7.55994081e-01 -1.99545369e-01 4.87198621e-01
-8.46300185e-01 1.02859549e-01 -3.59148890e-01 4.01892871e-01
1.62426710e-01 4.79160428e-01 -2.27734625e-01 -6.67294383e-01
-5.32389343e-01 -1.02091706e+00 9.22620177e-01 1.43845975e-01
-3.84348482e-01 -1.05615544e+00 -4.96177733e-01 4.13997114e-01
3.50281894e-01 -3.08106601e-01 1.24240792e+00 -1.41013873e+00
2.24677265e-01 -5.28223336e-01 -8.06168541e-02 5.68208814e-01
6.66327357e-01 -1.06955022e-01 -1.02671969e+00 -2.92639554e-01
-1.03630893e-01 -1.31286919e-01 9.39662099e-01 -3.74373165e-03
1.58815348e+00 -2.88740218e-01 -5.23459613e-01 4.22400683e-01
8.04442704e-01 2.89720684e-01 5.65458298e-01 4.51672047e-01
9.76108074e-01 6.97944939e-01 9.59971607e-01 4.52562660e-01
1.09475004e-02 5.43194175e-01 1.16865396e-01 -3.00066203e-01
1.73554152e-01 -2.82643169e-01 2.05187976e-01 9.99433279e-01
2.87586421e-01 -3.91370147e-01 -8.33256781e-01 4.05405968e-01
-1.63650417e+00 -6.54572785e-01 -4.83376145e-01 2.17933297e+00
1.18947673e+00 4.29612339e-01 -7.25967214e-02 5.56850255e-01
9.24285769e-01 3.98108922e-02 -4.47494179e-01 -1.55210301e-01
-3.38854581e-01 1.41745374e-01 3.04475784e-01 2.57415008e-02
-1.13222837e+00 1.11147833e+00 5.87097311e+00 1.33574450e+00
-9.80980217e-01 4.39970382e-02 7.59435117e-01 -2.24206503e-02
-4.88611966e-01 -1.77460849e-01 -1.19503212e+00 5.63219845e-01
4.60425586e-01 -4.13910747e-01 6.63847430e-03 8.56831670e-01
1.62819356e-01 2.14282453e-01 -9.57827926e-01 8.81342828e-01
2.88049221e-01 -1.18996406e+00 8.84086847e-01 -1.48319975e-01
6.61778927e-01 -3.89636338e-01 -5.16716838e-02 6.09122097e-01
2.68331945e-01 -9.11995053e-01 4.68160123e-01 1.23011179e-01
6.38961494e-01 -8.97295535e-01 1.03036845e+00 4.41643953e-01
-1.08085024e+00 -1.18270285e-01 -5.31173587e-01 6.00204170e-02
-2.60921806e-01 6.70732498e-01 -9.62967455e-01 6.09444976e-01
5.69712102e-01 6.54084563e-01 -9.55298841e-01 4.22087491e-01
-1.24122545e-01 6.14688158e-01 2.78226193e-02 -4.89680558e-01
6.50113150e-02 -1.75535932e-01 3.75926644e-01 1.11483729e+00
1.49455622e-01 -7.94728249e-02 1.86182842e-01 5.81341505e-01
-3.70187700e-01 7.59259462e-01 -4.82045114e-01 -1.10479757e-01
6.29619718e-01 1.26538575e+00 -8.58722210e-01 -6.59819782e-01
-3.89098018e-01 7.56473720e-01 -7.72272646e-02 1.23769663e-01
-6.13925219e-01 -7.58059442e-01 5.99520624e-01 7.11555185e-04
-7.38127679e-02 1.70944452e-01 -6.58827722e-01 -1.27881336e+00
1.30615070e-01 -1.07013059e+00 4.63747948e-01 -6.20182633e-01
-1.38629186e+00 7.17630029e-01 -2.81416699e-02 -1.49883640e+00
2.28602052e-01 -5.22452116e-01 -4.79222715e-01 1.03063786e+00
-1.22857559e+00 -1.07345343e+00 -5.48471689e-01 5.15233994e-01
8.59074116e-01 -4.80401367e-01 8.25085402e-01 2.14776680e-01
-8.20400536e-01 9.41174746e-01 3.26057196e-01 3.70717973e-01
8.71354520e-01 -1.06342053e+00 5.77981889e-01 7.73957133e-01
9.47013199e-02 9.46672440e-01 5.14358282e-01 -8.08955729e-01
-1.04509485e+00 -1.33654714e+00 8.45218539e-01 -3.59663606e-01
6.18242383e-01 -5.63725650e-01 -1.28007817e+00 5.56086183e-01
-1.36365190e-01 -3.45074266e-01 9.79805410e-01 3.72800827e-01
-3.15226316e-01 -1.42051116e-01 -9.60493684e-01 1.11191881e+00
8.95137608e-01 -2.10054308e-01 -8.05492103e-01 5.56054890e-01
9.78103280e-01 2.18739212e-02 -3.03792864e-01 4.20718014e-01
2.83514887e-01 -5.04714608e-01 7.89856553e-01 -9.87589300e-01
4.12529558e-01 -1.39559120e-01 2.05152873e-02 -1.54487097e+00
-1.63150787e-01 -2.65441597e-01 2.07267374e-01 1.63261604e+00
5.54233789e-01 -1.02505898e+00 7.98526406e-01 4.05589849e-01
-9.51427221e-02 -6.85845375e-01 -6.25197828e-01 -6.95256174e-01
2.14922860e-01 -4.63086545e-01 9.49527025e-01 1.52974772e+00
-2.95683392e-03 5.66525340e-01 -3.92725945e-01 -2.67362028e-01
1.08382098e-01 -5.89111224e-02 8.42224061e-01 -1.46239591e+00
1.91097990e-01 -7.08852410e-01 -1.13506816e-01 -9.59805965e-01
2.41204128e-01 -1.12071669e+00 -5.10086752e-02 -1.30573034e+00
3.14595699e-01 -7.99056888e-01 -2.46257082e-01 8.37129176e-01
-6.79555118e-01 1.58779636e-01 -1.53848961e-01 2.73973614e-01
-2.01082945e-01 9.86104786e-01 1.35282600e+00 -2.63545364e-01
-3.46236825e-01 1.50202349e-01 -1.24227262e+00 7.73314238e-01
9.04858887e-01 -4.61980790e-01 -5.33264816e-01 -2.51446217e-01
2.53701389e-01 -7.67850399e-01 -3.46494555e-01 -5.92953801e-01
-8.72562677e-02 -2.47087806e-01 4.18861091e-01 -5.64120889e-01
5.92183918e-02 -8.67556930e-01 -3.62304151e-01 4.38674480e-01
-6.55827224e-01 1.09667648e-02 3.68141681e-01 4.51462090e-01
-1.63281873e-01 -5.20003796e-01 6.62915051e-01 1.75766349e-02
-5.74420094e-01 2.11260557e-01 -3.77751559e-01 1.32984251e-01
1.02452040e+00 -3.74836504e-01 -2.44701117e-01 -1.05470464e-01
-5.21755457e-01 2.39747792e-01 1.70177549e-01 8.65531504e-01
5.35192966e-01 -1.50084269e+00 -9.08901155e-01 5.75687170e-01
5.02637029e-01 3.65466364e-02 1.26386583e-01 6.10878050e-01
-1.29664660e-01 3.53745282e-01 1.49863079e-01 -4.27880406e-01
-1.59698606e+00 4.83099371e-01 1.48651034e-01 -2.89120883e-01
-5.47500074e-01 7.38788247e-01 4.49562490e-01 -5.97515225e-01
1.96039721e-01 -3.79252136e-01 -6.73420131e-01 6.08317316e-01
7.35272050e-01 1.40063986e-01 4.15768355e-01 -6.09793782e-01
-1.06708989e-01 2.73971051e-01 -7.33500421e-01 2.70881355e-01
1.06718481e+00 -1.13633059e-01 -2.45523021e-01 6.25165105e-01
9.59992647e-01 2.21355662e-01 -4.19606268e-01 -6.11785293e-01
9.04934630e-02 -6.49062276e-01 4.80683260e-02 -5.90266764e-01
-1.05463481e+00 7.53311098e-01 2.97652245e-01 4.50148821e-01
1.13349080e+00 -2.30473816e-01 6.50824964e-01 4.86752391e-01
8.16447437e-02 -1.12559688e+00 1.75482810e-01 7.65409887e-01
8.51358593e-01 -1.16622138e+00 6.99657798e-02 -6.95671260e-01
-8.55023861e-01 9.47756171e-01 8.84932995e-01 5.48404574e-01
4.79647577e-01 3.30154933e-02 3.15660089e-02 2.15807453e-01
-6.02986634e-01 -2.00029954e-01 3.30741525e-01 3.87715489e-01
6.10675693e-01 -4.47653942e-02 -5.73175490e-01 8.57437491e-01
-2.56132126e-01 -5.74613452e-01 5.19699812e-01 9.12521839e-01
-4.53782260e-01 -1.36972296e+00 -4.79675293e-01 1.06536222e+00
-3.24626118e-01 -4.12386149e-01 -5.31246662e-01 8.34528387e-01
1.30926058e-01 9.27161753e-01 -3.15056592e-02 -7.22007573e-01
4.41110611e-01 4.69514787e-01 -1.24156743e-01 -9.18033600e-01
-6.51218414e-01 -1.66817799e-01 1.42748550e-01 1.62932888e-01
-1.20869860e-01 -5.56797564e-01 -1.23585904e+00 -2.89492458e-01
-8.91469181e-01 3.88465166e-01 5.35968423e-01 1.09020782e+00
2.23558798e-01 7.55730987e-01 8.07402492e-01 -4.84908253e-01
-5.45287251e-01 -1.45286238e+00 -5.03216982e-01 7.10797071e-01
-7.79464990e-02 -7.48892188e-01 -4.64412779e-01 -5.37666865e-02] | [10.497540473937988, 7.582983493804932] |
7d19660b-423f-423e-9026-4519329d366e | spectrum-shaping-for-multiple-link-discovery | 2108.04932 | null | https://arxiv.org/abs/2108.04932v1 | https://arxiv.org/pdf/2108.04932v1.pdf | Spectrum Shaping For Multiple Link Discovery in 6G THz Systems | This paper presents a novel antenna configuration to measure directions of multiple signal sources at the receiver in a THz mobile network via a single channel measurement. Directional communication is an intrinsic attribute of THz wireless networks and the knowledge of direction should be harvested continuously to maintain link quality. Direction discovery can potentially impose an immense burden on the network that limits its communication capacity exceedingly. To utterly mitigate direction discovery overhead, we propose a novel technique called spectrum shaping capable of measuring direction, power, and relative distance of propagation paths via a single measurement. We demonstrate that the proposed technique is also able to measure the transmitter antenna orientation. We evaluate the performance of the proposed design in several scenarios and show that the introduced technique performs similar to a large array of antennas while attaining a much simpler hardware architecture. Results show that the spectrum shaping with only two antennas placed 0.5 mm, 5 mm, and 1 cm apart performs direction of arrival estimation similar to a much more complex uniform linear array equipped with 7, 60, and 120 antennas, respectively. | ['Nazanin Rahnavard', 'Katarina Vuckovic', 'Farzam Hejazi'] | 2021-08-10 | null | null | null | null | ['direction-of-arrival-estimation'] | ['audio'] | [ 5.10721862e-01 5.36122441e-01 2.93823212e-01 -1.92471936e-01
-2.15317219e-01 -8.12165201e-01 4.86023068e-01 3.47985158e-04
-3.69179279e-01 7.25233078e-01 -8.70739296e-02 -7.75944293e-01
-5.63463449e-01 -9.79781985e-01 -2.56730169e-01 -1.23291600e+00
-2.73942709e-01 -2.37998385e-02 2.11194277e-01 1.08677313e-01
2.47495428e-01 6.87895179e-01 -9.90040779e-01 -4.62003410e-01
5.00453591e-01 1.27536368e+00 3.53537232e-01 6.22686267e-01
6.63912296e-02 1.69549257e-01 -8.10547709e-01 -4.52542752e-02
-1.19217493e-01 -1.53916031e-01 1.11740656e-01 -2.32347682e-01
-4.32353094e-02 -5.00097454e-01 -1.96716309e-01 7.16019630e-01
8.04064989e-01 -6.56569779e-01 7.24135160e-01 -1.04987192e+00
3.12512040e-01 8.96814466e-01 -4.39298689e-01 -3.49315256e-02
2.22815067e-01 -3.74123037e-01 1.04413547e-01 -1.29552320e-01
4.83791232e-01 3.86378109e-01 6.11957908e-01 -9.35082585e-02
-5.94686687e-01 -8.42869639e-01 -7.10140228e-01 -1.45529732e-02
-1.10765719e+00 -7.09452391e-01 1.12330234e+00 -5.83580434e-02
2.11385056e-01 4.85343128e-01 6.30466938e-01 9.30868685e-01
4.99571562e-01 -2.31449813e-01 1.00966573e+00 -7.62250543e-01
5.80754697e-01 3.33626419e-01 -2.47778282e-01 7.54384220e-01
7.75656819e-01 3.92345488e-02 -3.71589333e-01 -2.37779886e-01
6.35082126e-01 -5.25091052e-01 -4.67471063e-01 -5.18851757e-01
-1.52703679e+00 2.73658186e-01 4.97653991e-01 9.89670157e-01
-5.69639146e-01 4.52854395e-01 -4.64420050e-01 2.92596310e-01
3.07838656e-02 7.07514942e-01 1.61157280e-01 -9.08485427e-02
-7.89610445e-01 -3.48999828e-01 8.48070920e-01 1.22315168e+00
2.07356349e-01 -7.00757205e-02 -2.90204167e-01 2.98119009e-01
3.95738393e-01 1.04513884e+00 -7.72602670e-03 -6.76555812e-01
1.82979241e-01 -9.59898233e-02 2.22552538e-01 -1.11401761e+00
-1.04421878e+00 -1.43469262e+00 -9.76215541e-01 -1.52783170e-01
5.70856929e-01 -5.59574187e-01 -6.21833563e-01 1.65123093e+00
3.30672443e-01 -4.10032459e-02 1.67111620e-01 3.77768099e-01
6.94801807e-01 6.10966146e-01 -3.56588364e-01 -6.96811855e-01
1.13520098e+00 -2.92787701e-01 -8.12098086e-01 -9.52020288e-02
5.57062149e-01 -7.55803823e-01 5.87196425e-02 5.44276655e-01
-9.94124472e-01 5.68701141e-02 -1.50256991e+00 4.06391591e-01
-5.50652593e-02 4.82960284e-01 6.23823106e-01 1.46349299e+00
-9.64274228e-01 2.00895816e-01 -5.23056090e-01 -3.45459044e-01
1.82659879e-01 2.76082814e-01 6.13064617e-02 -1.00158326e-01
-6.16961539e-01 3.88236076e-01 -1.22351512e-01 6.59721792e-02
-1.80870458e-01 -8.38323355e-01 -1.08383030e-01 1.58916488e-01
-7.59489015e-02 -5.27816474e-01 1.00111294e+00 1.55126601e-01
-1.75546217e+00 3.78964841e-02 -2.84362167e-01 -3.08604330e-01
4.03282285e-01 5.85592687e-01 -6.76075101e-01 4.86506850e-01
6.77196309e-02 -9.62634161e-02 4.34174538e-01 -1.17015529e+00
-4.85791057e-01 -5.21087408e-01 -2.98923850e-01 -2.28442341e-01
-6.37850404e-01 -4.82708633e-01 2.48201981e-01 -1.11588381e-01
1.04594815e+00 -8.88266623e-01 -1.50919273e-01 5.97517975e-02
-6.98898315e-01 3.44075531e-01 9.39063489e-01 7.57241026e-02
8.58611822e-01 -1.93250430e+00 -3.36454481e-01 7.61818349e-01
3.91678363e-01 -2.07840458e-01 -7.95824230e-02 9.27765131e-01
2.34267935e-01 -1.08546272e-01 -1.12576611e-01 1.16247207e-01
-4.02438283e-01 -4.62708026e-01 2.51047850e-01 8.31797600e-01
-7.21508205e-01 3.75237077e-01 -7.85643220e-01 1.89343661e-01
-1.28538013e-01 4.02308345e-01 -2.70579368e-01 1.29690617e-02
2.76911020e-01 4.98794198e-01 -7.63073623e-01 7.43718684e-01
1.21605217e+00 -1.49426535e-01 5.28172493e-01 -4.59212333e-01
-7.38861501e-01 1.96525734e-02 -9.40926015e-01 1.35910261e+00
-4.28205848e-01 7.24428117e-01 3.60763252e-01 -9.89558280e-01
1.17923915e+00 5.40344119e-01 7.21445918e-01 -1.11107171e+00
1.73245341e-01 4.50867444e-01 3.42644215e-01 -7.25740492e-01
8.92750621e-02 -7.36988932e-02 -1.72172129e-01 5.67291915e-01
-6.72953799e-02 -1.38500169e-01 -1.61123082e-01 1.78314939e-01
1.46250498e+00 -5.66126108e-01 3.33524615e-01 -4.74581659e-01
2.93107539e-01 -2.88566083e-01 2.39616677e-01 7.88191080e-01
4.06644605e-02 -9.92092863e-02 4.52794462e-01 1.46047533e-01
-7.22754896e-01 -1.25081098e+00 -2.87701815e-01 1.18715428e-01
6.68626845e-01 -2.69993633e-01 -2.98117816e-01 -9.78136733e-02
2.60191429e-02 7.02245176e-01 -1.05617046e-01 -6.01313217e-03
-2.11517245e-01 -9.55707073e-01 5.87439835e-01 -1.50077045e-01
8.88999045e-01 1.40144033e-02 -5.92306018e-01 1.71819434e-01
-3.78781147e-02 -1.25603819e+00 3.55267942e-01 1.05087265e-01
-5.51499426e-01 -6.28781974e-01 -6.25651956e-01 -5.38782775e-01
7.82762587e-01 6.87705457e-01 4.80952770e-01 -4.44503635e-01
1.27539203e-01 6.44473255e-01 -2.97327101e-01 -6.72077119e-01
-1.06563523e-01 1.19729407e-01 5.58096208e-02 1.85238812e-02
-2.12950096e-01 -1.21214592e+00 -8.77050698e-01 5.55528283e-01
-2.44995892e-01 -1.52694881e-01 9.63374019e-01 2.89627258e-02
-6.71562403e-02 4.04759437e-01 1.03591096e+00 -3.79153699e-01
4.78706837e-01 -7.43847370e-01 -7.38568068e-01 9.88226831e-02
-4.93526876e-01 -3.00143920e-02 4.12999630e-01 7.71286562e-02
-8.90736699e-01 -1.95138529e-02 -1.30245104e-01 8.07774365e-01
-6.78615198e-02 4.25445199e-01 -3.23540628e-01 -8.98538172e-01
4.65062827e-01 2.24672839e-01 -1.80280820e-01 -1.30763948e-01
1.78034440e-01 7.84215510e-01 2.22035319e-01 8.90489072e-02
1.11613846e+00 6.03967130e-01 7.67184794e-01 -1.50586319e+00
6.62824810e-02 -5.18821895e-01 -1.54351696e-01 -5.22242904e-01
3.91883284e-01 -7.63535678e-01 -9.45601225e-01 3.34877789e-01
-1.22709024e+00 1.43387571e-01 3.95349324e-01 9.32308912e-01
-4.56605069e-02 1.71762362e-01 9.60741490e-02 -9.92686927e-01
-2.24816740e-01 -5.91130078e-01 7.85558581e-01 3.00999254e-01
-6.80254493e-03 -7.22494721e-01 -1.07339539e-01 -3.86615433e-02
1.06810665e+00 4.04300869e-01 9.71248507e-01 -7.22242445e-02
-7.28637815e-01 -6.09089196e-01 -5.39353117e-02 -7.54273415e-01
4.08269286e-01 -5.32983959e-01 -1.06700337e+00 -2.47022182e-01
3.06588203e-01 4.16085869e-01 4.27544415e-01 7.14730620e-01
1.00704503e+00 6.41232729e-02 -9.75852609e-01 5.30385613e-01
1.43702793e+00 5.21764398e-01 7.37752736e-01 7.29529485e-02
1.36840731e-01 3.23751986e-01 1.83945701e-01 3.39997590e-01
1.46480545e-01 4.25009102e-01 7.01085269e-01 5.00924475e-02
2.56463252e-02 2.44812742e-01 -1.36322245e-01 6.94291115e-01
-1.20811298e-01 -1.11648989e+00 -3.63400668e-01 1.29627258e-01
-1.14144599e+00 -9.01818752e-01 -3.86159122e-01 2.25322771e+00
-5.50770089e-02 -4.61941883e-02 -1.70254663e-01 4.05953825e-01
3.52105767e-01 5.85878715e-02 -1.83944464e-01 1.42466605e-01
8.48954320e-02 -6.68022037e-02 9.52867389e-01 4.17172849e-01
-5.76370299e-01 -9.55929756e-02 5.94988298e+00 4.64274973e-01
-1.46123827e+00 -1.12182356e-01 5.40527166e-04 -2.89282892e-02
-8.39832544e-01 -2.03535736e-01 -4.34872061e-01 5.16513109e-01
7.40674078e-01 -2.05467045e-01 -2.13570982e-01 3.33949655e-01
3.62458497e-01 -5.46988308e-01 -7.96423852e-01 1.06509626e+00
-3.27687800e-01 -1.32478964e+00 -3.19535524e-01 5.86670578e-01
3.33171338e-01 -5.28997004e-01 1.68185383e-01 -4.23998564e-01
-6.32665277e-01 -4.54731047e-01 3.95838380e-01 5.46201706e-01
5.61944783e-01 -6.55446231e-01 5.69098353e-01 7.30745077e-01
-1.04698431e+00 -2.24045426e-01 -1.36575669e-01 3.62256579e-02
2.89730400e-01 1.53919148e+00 -1.29113626e+00 7.95913935e-01
1.09973855e-01 1.24457262e-01 -9.83004346e-02 1.20000994e+00
3.43584195e-02 7.35975921e-01 -7.29891121e-01 -6.33754909e-01
1.30456656e-01 -2.62286127e-01 9.89099205e-01 7.82515585e-01
1.41645992e+00 8.59962329e-02 -7.62630463e-01 5.34847558e-01
-2.55451262e-01 -2.51555443e-01 -7.84890532e-01 7.04655796e-02
9.73254085e-01 1.41211832e+00 -8.71075928e-01 2.48936862e-01
-1.72084942e-01 6.63612902e-01 -3.17732930e-01 3.77186149e-01
-6.28130317e-01 -6.77428842e-01 2.55978674e-01 4.26571637e-01
2.60940224e-01 -8.47972512e-01 -4.12321895e-01 -4.62083191e-01
2.97702730e-01 8.19970295e-02 -3.94791424e-01 -4.28238869e-01
-4.89850521e-01 3.35332185e-01 -3.26831728e-01 -1.46026993e+00
-5.02152964e-02 -2.58079767e-01 -7.03965068e-01 6.51596606e-01
-1.51183546e+00 -6.96765184e-01 -7.74405539e-01 2.00637300e-02
-3.18392843e-01 -3.36599573e-02 8.52010667e-01 3.82451683e-01
-1.39670402e-01 4.65533704e-01 7.19805598e-01 -3.83325487e-01
4.85561877e-01 -7.93900728e-01 1.29023949e-02 6.43816829e-01
-2.46735170e-01 4.88571793e-01 7.60158062e-01 -5.19074976e-01
-2.02047753e+00 -7.83862591e-01 6.93355858e-01 3.42839926e-01
3.37946936e-02 -6.56572878e-01 7.49554262e-02 -7.64294863e-02
3.34000409e-01 -2.85971493e-01 8.55461836e-01 -9.83045623e-02
2.84581155e-01 -4.70594466e-01 -1.75180161e+00 3.99645329e-01
1.22804117e+00 7.32940361e-02 5.08356750e-01 4.87105876e-01
2.22533181e-01 -1.69751093e-01 -6.61872327e-01 4.46651965e-01
1.13473058e+00 -1.00611353e+00 8.83572102e-01 5.41720808e-01
-6.11913390e-02 -1.13553494e-01 -3.81804466e-01 -1.48418713e+00
-5.90973020e-01 -7.48439193e-01 3.31846148e-01 1.12502730e+00
5.89342952e-01 -1.14663112e+00 9.99477386e-01 -1.47553280e-01
4.23587561e-02 -6.34194493e-01 -1.07347059e+00 -9.88138437e-01
-2.44213015e-01 -2.18533710e-01 6.29875243e-01 5.73459148e-01
2.55138963e-01 3.22203130e-01 -5.94047196e-02 9.17434633e-01
1.20318627e+00 -2.38459498e-01 4.44677293e-01 -1.36236811e+00
-2.04256341e-01 -2.97671467e-01 -6.70802951e-01 -1.26611829e+00
-5.70370853e-01 -6.57554150e-01 -9.65588912e-02 -1.63619888e+00
-4.95633632e-01 -1.01422656e+00 5.93554191e-02 -3.77361715e-01
7.15308666e-01 2.92097181e-01 -4.79626864e-01 -4.19467419e-01
1.97199896e-01 5.35183609e-01 1.11537695e+00 -7.85631314e-02
-3.35426554e-02 5.78594327e-01 -5.46722651e-01 4.87321854e-01
7.92859077e-01 -3.31864536e-01 -7.87847102e-01 -4.59561348e-01
2.22556144e-01 4.33822304e-01 2.60438770e-01 -1.58896244e+00
7.19495535e-01 2.30603978e-01 7.51316369e-01 -4.90592360e-01
6.24245167e-01 -1.09562624e+00 3.41486275e-01 7.05143452e-01
-1.43956505e-02 -6.58894718e-01 -1.18820116e-01 7.70431757e-01
4.86051142e-01 -5.43313771e-02 5.33818185e-01 2.28546306e-01
-1.53129980e-01 -9.99646038e-02 -6.35395169e-01 -9.47795033e-01
1.32072234e+00 -2.32720762e-01 -7.19412863e-01 -6.79653883e-01
-2.35275909e-01 -1.92943048e-02 -1.07987940e-01 -2.17206851e-01
4.43752736e-01 -1.22268534e+00 -3.44003379e-01 9.78389606e-02
-8.11593980e-02 -6.90527678e-01 2.13630974e-01 8.64617705e-01
-2.79179215e-01 8.09180021e-01 -1.45967856e-01 -6.38432384e-01
-1.35881150e+00 -1.33066475e-01 4.28026825e-01 1.95291653e-01
-3.02124828e-01 7.72350132e-01 -2.60147423e-01 -4.71504740e-02
9.82139483e-02 -3.13207507e-01 -2.48554870e-01 -1.14181370e-01
2.78597891e-01 4.59713727e-01 3.54663044e-01 1.02262288e-01
-4.40929323e-01 1.03386509e+00 3.87695521e-01 -2.77887136e-01
1.17650747e+00 -3.97128791e-01 1.35657340e-02 -4.93941568e-02
1.10083139e+00 8.02034438e-01 -3.74047548e-01 5.99192791e-02
-3.39979053e-01 -3.44379634e-01 2.84134150e-01 -8.47356021e-01
-7.11810768e-01 3.45959663e-01 8.35949004e-01 7.30998814e-01
9.78012621e-01 4.80703972e-02 4.28284287e-01 7.55900443e-01
8.99207354e-01 -6.67136431e-01 -2.06888676e-01 -1.95915192e-01
6.11808419e-01 -7.26167381e-01 1.42627701e-01 -1.05178094e+00
5.99847138e-01 1.24883044e+00 5.70307188e-02 4.90563035e-01
9.28839505e-01 6.60205662e-01 -1.76584587e-01 -4.16915536e-01
-2.02674568e-01 2.11971119e-01 -2.17687428e-01 8.40059161e-01
5.89559674e-01 1.69694468e-01 -5.32964289e-01 2.28484645e-01
-3.78006428e-01 -1.37425005e-01 9.14403439e-01 1.13313699e+00
-9.40912127e-01 -1.07963610e+00 -3.68080646e-01 5.16936719e-01
-1.57819211e-01 3.10668528e-01 -1.67581365e-01 5.18841684e-01
6.01163954e-02 1.37682164e+00 -6.47028312e-02 -3.64622176e-01
5.27157843e-01 -3.23176205e-01 7.30449796e-01 1.32907210e-02
4.19753820e-01 -5.82578555e-02 4.13202703e-01 -4.05054569e-01
-3.04414183e-01 -3.34663868e-01 -8.70326400e-01 -2.91321706e-02
-2.10755900e-01 3.71717811e-01 1.39868426e+00 9.16711211e-01
4.14425731e-01 6.38131559e-01 1.10794461e+00 -4.46839482e-01
-3.38606507e-01 -7.90881455e-01 -9.71297503e-01 -8.39575827e-01
5.16954124e-01 -5.90452909e-01 -3.59027773e-01 -7.43882716e-01] | [6.310864448547363, 1.1832473278045654] |
fa7532ba-3b69-4715-837b-7fde2cb3135c | neural-volumetric-reconstruction-for-coherent | 2306.09909 | null | https://arxiv.org/abs/2306.09909v1 | https://arxiv.org/pdf/2306.09909v1.pdf | Neural Volumetric Reconstruction for Coherent Synthetic Aperture Sonar | Synthetic aperture sonar (SAS) measures a scene from multiple views in order to increase the resolution of reconstructed imagery. Image reconstruction methods for SAS coherently combine measurements to focus acoustic energy onto the scene. However, image formation is typically under-constrained due to a limited number of measurements and bandlimited hardware, which limits the capabilities of existing reconstruction methods. To help meet these challenges, we design an analysis-by-synthesis optimization that leverages recent advances in neural rendering to perform coherent SAS imaging. Our optimization enables us to incorporate physics-based constraints and scene priors into the image formation process. We validate our method on simulation and experimental results captured in both air and water. We demonstrate both quantitatively and qualitatively that our method typically produces superior reconstructions than existing approaches. We share code and data for reproducibility. | ['Suren Jayasuriya', 'Daniel C. Brown', 'Adithya Pediredla', 'Thomas Blanford', 'Juhyeon Kim', 'Albert W. Reed'] | 2023-06-16 | null | null | null | null | ['image-reconstruction', 'neural-rendering'] | ['computer-vision', 'computer-vision'] | [ 7.39259779e-01 -2.47180164e-01 6.81224108e-01 -3.55894774e-01
-1.00707686e+00 -6.37149215e-01 3.52130145e-01 -3.78193408e-01
-2.07202181e-01 4.89340186e-01 2.98730344e-01 -9.59787741e-02
3.05017736e-02 -5.28876066e-01 -8.21701527e-01 -7.12704182e-01
-8.73987302e-02 -1.69895142e-01 -1.48341730e-02 5.01380228e-02
1.60566613e-01 3.23493958e-01 -1.52302706e+00 -9.03749540e-02
8.23429346e-01 7.33584583e-01 6.96134269e-01 9.97747362e-01
3.92627209e-01 7.62912273e-01 -2.25481868e-01 3.34167778e-01
4.96108353e-01 -6.62653685e-01 1.25808897e-03 3.35783482e-01
6.98928952e-01 -7.55830467e-01 -4.31497008e-01 1.06607807e+00
5.63144088e-01 3.60177130e-01 4.01466668e-01 -5.47270119e-01
-1.70999184e-01 -1.35584059e-03 -4.60736722e-01 2.48588756e-01
4.18151706e-01 3.50489795e-01 7.26370156e-01 -9.86769617e-01
3.10432941e-01 8.79232466e-01 5.86424708e-01 3.99730235e-01
-1.25798202e+00 -5.28599262e-01 -1.88694537e-01 -4.06128794e-01
-1.15490592e+00 -1.04085243e+00 7.65527308e-01 -2.43572593e-01
7.95743585e-01 2.50990033e-01 8.96665990e-01 6.57108247e-01
1.63282424e-01 4.47763324e-01 1.22130668e+00 -3.01757157e-01
3.24055254e-01 -2.59388715e-01 -2.15736195e-01 8.19785178e-01
1.52461052e-01 4.61578459e-01 -7.99979806e-01 -4.84707654e-02
1.15944171e+00 -2.77885377e-01 -7.72745550e-01 -5.07453561e-01
-1.03775668e+00 5.75022697e-01 2.74742872e-01 -4.54897910e-01
-2.86763221e-01 5.35367668e-01 -3.67689580e-01 -1.34670615e-01
3.35517079e-01 8.71716261e-01 1.68426614e-02 -1.65631980e-01
-1.02004611e+00 1.59021810e-01 8.31597090e-01 7.31965363e-01
5.82200110e-01 7.10826516e-01 3.68344784e-01 9.12920177e-01
6.48875535e-01 1.13304305e+00 -4.42956528e-03 -1.61700320e+00
-3.96990310e-03 -2.48916820e-01 4.09549206e-01 -9.04526234e-01
-2.83363670e-01 -6.39512360e-01 -4.05695081e-01 5.62698603e-01
1.27483830e-01 -4.29157525e-01 -1.22025990e+00 1.61381614e+00
2.27716461e-01 6.40875638e-01 4.27352399e-01 1.40468431e+00
8.50365102e-01 9.62614238e-01 -5.51882923e-01 -3.79318535e-01
1.03194797e+00 -6.90221727e-01 -7.77265191e-01 -5.97255766e-01
-3.10174134e-02 -9.41066742e-01 9.28911567e-01 5.42727947e-01
-1.36087334e+00 -1.43524408e-01 -1.40720809e+00 5.55236936e-02
5.28186023e-01 -8.25129598e-02 4.53827143e-01 4.82483953e-01
-9.79603350e-01 2.15046927e-01 -1.29438865e+00 -5.73250316e-02
1.81525946e-01 2.83552222e-02 1.26110632e-02 -1.95926100e-01
-6.20756626e-01 6.99503958e-01 -3.10670614e-01 1.25868678e-01
-1.22130537e+00 -1.00289369e+00 -1.19640219e+00 -2.41146218e-02
2.62166947e-01 -8.08706164e-01 1.43780589e+00 -3.83037090e-01
-1.89278734e+00 3.24731827e-01 -1.80763096e-01 -5.22679985e-01
1.24464452e-01 -4.21438754e-01 -6.57024011e-02 5.28756261e-01
-1.47327185e-01 4.25279319e-01 8.69506478e-01 -1.76035571e+00
-4.79071826e-01 1.57595128e-01 9.20728594e-02 7.49682009e-01
2.05877334e-01 -3.57817024e-01 -5.01824498e-01 -3.42471153e-01
5.88790119e-01 -8.90489280e-01 -5.15225708e-01 1.48803622e-01
-1.98293015e-01 1.03331947e+00 5.80893874e-01 -4.37794834e-01
5.98233759e-01 -2.05833578e+00 5.83946705e-02 -4.27028537e-02
2.81793356e-01 -6.14063963e-02 -9.46968198e-02 4.79563504e-01
3.34439784e-01 -4.63759333e-01 -5.47379792e-01 -6.85883224e-01
-3.79989922e-01 3.37169379e-01 -5.93170166e-01 7.34999120e-01
-3.31320800e-02 4.26143855e-01 -8.68633747e-01 -2.57467151e-01
5.77298164e-01 6.71499908e-01 -8.63661706e-01 5.26424408e-01
-1.80999845e-01 8.61112714e-01 -4.05170500e-01 5.93573749e-01
7.86973357e-01 -3.12692821e-01 9.51770544e-02 -3.74860823e-01
-4.08044994e-01 2.86357015e-01 -1.14389801e+00 1.73485053e+00
-8.39655221e-01 7.58997202e-01 7.75714934e-01 -7.26636529e-01
5.51110864e-01 6.84911534e-02 5.43328524e-01 -6.31945193e-01
-8.13006982e-02 9.22222957e-02 1.61767397e-02 -5.93120158e-01
5.89493930e-01 -3.32594633e-01 2.69598812e-01 3.91934067e-01
-1.71390995e-01 -1.14864361e+00 -1.66996777e-01 2.35639080e-01
1.09026849e+00 2.20194697e-01 2.88812190e-01 -1.96028858e-01
-1.60340831e-01 2.09175393e-01 6.43633246e-01 9.62754607e-01
1.56863362e-01 1.10797369e+00 -1.99562877e-01 1.67567767e-02
-1.12279499e+00 -1.45224333e+00 -3.06112856e-01 4.86578524e-01
5.34324288e-01 -9.34690759e-02 -5.54568052e-01 1.34930238e-01
-1.48360372e-01 7.31757224e-01 -2.82765776e-01 2.22907588e-01
-5.80600142e-01 -8.06905031e-01 2.50753403e-01 1.07190602e-01
4.60144520e-01 -6.49591386e-01 -1.24536622e+00 3.97885472e-01
-2.14328125e-01 -1.45869243e+00 -2.30146721e-01 1.34515287e-02
-6.61566019e-01 -9.73097920e-01 -4.80892479e-01 -3.29408646e-01
5.61193466e-01 8.14391732e-01 1.01003027e+00 -1.19558536e-01
-4.63170886e-01 9.96479571e-01 -1.96416512e-01 -5.14045775e-01
-4.55422163e-01 -9.12220657e-01 1.88060150e-01 -1.17116988e-01
-5.98305583e-01 -1.01639569e+00 -9.29102659e-01 1.75011218e-01
-8.74634802e-01 4.44874913e-01 5.61421931e-01 8.42257500e-01
5.26322007e-01 -7.02541843e-02 -1.02001533e-01 -3.96154374e-01
3.58697087e-01 -2.38504544e-01 -1.12010944e+00 -1.90977767e-01
-3.50600123e-01 -8.12456310e-02 2.72332937e-01 -2.02861831e-01
-1.38692462e+00 1.81979880e-01 4.67385054e-02 -5.09376645e-01
7.01866299e-02 8.36981058e-01 2.68346936e-01 -5.30751944e-01
4.97037321e-01 4.48962718e-01 2.68553704e-01 -1.40044972e-01
6.98985606e-02 2.61694908e-01 8.61195624e-01 -3.86142433e-01
7.83715487e-01 8.42564106e-01 1.38887793e-01 -1.38906038e+00
-9.62911069e-01 -2.62680560e-01 2.49136309e-03 -3.96243066e-01
7.29243159e-01 -1.26242459e+00 -6.23153806e-01 2.92110980e-01
-8.65611076e-01 -6.44893646e-01 -1.84939176e-01 1.08953977e+00
-5.72519422e-01 4.26518321e-01 -5.27407706e-01 -1.14140785e+00
-1.57895029e-01 -1.25117207e+00 1.00071871e+00 3.54064614e-01
1.65650249e-01 -7.78129458e-01 2.36597925e-01 4.23164517e-01
4.42414045e-01 2.53235012e-01 1.60662994e-01 4.41928655e-01
-1.17132080e+00 2.51658976e-01 -1.80580780e-01 2.48769462e-01
5.18252216e-02 5.21293050e-03 -1.12384307e+00 -4.34065580e-01
4.00863945e-01 -3.94168556e-01 1.00672805e+00 1.04525256e+00
6.34692788e-01 -2.59771099e-04 -2.27199152e-01 1.12467611e+00
1.69467485e+00 1.50435328e-01 5.14546156e-01 -6.48084283e-02
5.27657807e-01 4.18844759e-01 5.28962672e-01 6.52750611e-01
2.72665113e-01 5.85024357e-01 5.67228675e-01 -1.52427629e-01
-2.79576868e-01 -3.19050625e-02 1.84090301e-01 8.78774703e-01
1.29156724e-01 -6.46972001e-01 -7.07499087e-01 4.82868373e-01
-1.43695545e+00 -7.11893857e-01 2.89640073e-02 2.08287954e+00
5.75902879e-01 -1.48036510e-01 -7.78439045e-01 -3.48559707e-01
7.92542398e-02 4.20650125e-01 -4.97408092e-01 -1.42115913e-02
-5.45297824e-02 2.05279812e-01 6.17978871e-01 9.69087124e-01
-8.43891382e-01 6.31990314e-01 7.26154232e+00 2.47988120e-01
-1.22851658e+00 -1.10953204e-01 6.61694026e-03 -4.05382961e-01
-5.83106637e-01 -8.19936115e-03 -4.63279963e-01 1.31962910e-01
6.23496592e-01 8.39976892e-02 7.62744904e-01 3.07018787e-01
5.61374724e-01 -6.77975953e-01 -8.46865058e-01 1.01776445e+00
2.19941795e-01 -1.50518739e+00 -2.40700573e-01 1.58022031e-01
7.40219712e-01 3.62974793e-01 2.83573836e-01 -4.37834352e-01
7.75119185e-01 -8.80084515e-01 6.03794098e-01 4.91132468e-01
8.15587938e-01 -4.14770365e-01 1.30927280e-01 2.89101601e-01
-9.28006470e-01 9.80942324e-02 -1.99657649e-01 -2.55277485e-01
6.95750833e-01 6.69470787e-01 -8.65233541e-01 3.85755271e-01
6.47591710e-01 6.40187144e-01 1.26363203e-01 1.09821415e+00
-1.52475700e-01 9.41231668e-01 -6.53608620e-01 2.35222667e-01
2.25028187e-01 -4.83515590e-01 1.08936679e+00 8.20227563e-01
6.22590601e-01 8.72868121e-01 3.82588893e-01 1.10774231e+00
3.17675292e-01 -5.20176411e-01 -7.62862146e-01 7.34462664e-02
5.10659397e-01 1.25166535e+00 -4.39900935e-01 -3.38683762e-02
-4.44865018e-01 5.80189884e-01 -9.44863930e-02 6.23228490e-01
-6.29461646e-01 3.30618583e-02 8.23743463e-01 -1.41985817e-02
2.89150417e-01 -8.98554444e-01 -2.63200790e-01 -1.21389556e+00
-1.75752431e-01 -6.18781209e-01 -2.05374151e-01 -1.29509556e+00
-8.15846503e-01 3.16745251e-01 9.13601518e-02 -1.34256375e+00
-1.23419613e-01 -2.65680194e-01 -4.21370894e-01 8.31780076e-01
-1.66893959e+00 -9.87494588e-01 -5.54849565e-01 1.66177228e-02
5.96285641e-01 6.91106990e-02 7.98545122e-01 -2.75737923e-02
-2.65940547e-01 -4.92622890e-02 3.17251086e-01 -1.79053709e-01
4.13853198e-01 -1.13064408e+00 3.46658796e-01 1.33176935e+00
8.12867507e-02 6.47380471e-01 1.31869435e+00 -5.00267684e-01
-1.86229980e+00 -5.41420102e-01 -3.05023283e-01 -1.40913829e-01
4.39938128e-01 -3.22812945e-01 -7.89848685e-01 5.30646741e-01
3.44951183e-01 2.49837503e-01 6.40235364e-01 -2.95685321e-01
3.20688970e-02 3.43369916e-02 -9.72891271e-01 5.73739827e-01
7.94670820e-01 -2.81980753e-01 -3.20822597e-01 9.24723893e-02
6.76546812e-01 -1.02752066e+00 -6.01632535e-01 5.65847218e-01
7.69437671e-01 -1.15626037e+00 1.19890225e+00 1.41042560e-01
5.77790141e-01 -5.44045508e-01 -5.27307153e-01 -1.65454483e+00
-1.26424566e-01 -7.07306683e-01 -8.82657710e-03 6.73609734e-01
1.52838737e-01 -8.25633526e-01 6.63986206e-01 3.94838125e-01
-5.64292490e-01 -3.53977710e-01 -7.47345150e-01 -4.60215122e-01
-3.83220494e-01 -6.30769491e-01 1.44069463e-01 6.44190311e-01
-3.28880996e-01 1.21460706e-01 -4.80910391e-01 1.06031311e+00
1.30391121e+00 4.41630810e-01 8.41447353e-01 -6.67463422e-01
-6.51685834e-01 8.91235769e-02 -4.71307859e-02 -1.52256680e+00
-2.13913634e-01 -2.49064475e-01 7.44679511e-01 -1.52773607e+00
4.87302467e-02 -6.43839657e-01 2.04439104e-01 -9.02724788e-02
5.56965098e-02 6.10986054e-01 3.74197587e-02 1.77872598e-01
-1.96161121e-01 6.49475455e-01 1.43490255e+00 2.58013636e-01
-2.28780434e-01 -1.51132971e-01 -4.19912070e-01 8.57447267e-01
4.19695735e-01 -2.43735418e-01 -7.05119669e-01 -9.89563763e-01
2.47042939e-01 6.10580623e-01 6.15067005e-01 -1.21731174e+00
3.70708197e-01 -3.52673709e-01 5.06184340e-01 -5.78891456e-01
1.05732453e+00 -6.59215868e-01 1.63847923e-01 2.67311364e-01
-3.25024098e-01 -5.04670441e-01 2.20518827e-01 7.86064744e-01
-1.74732909e-01 -8.58876109e-02 9.84591544e-01 -2.54652590e-01
-5.92235744e-01 8.50786269e-02 -5.37033677e-01 8.04512799e-02
4.37353671e-01 -8.56457353e-02 -2.97342151e-01 -9.94649947e-01
-3.99919748e-01 2.78806835e-01 6.08065605e-01 -1.40866220e-01
9.60495472e-01 -8.05308282e-01 -6.69407308e-01 4.52394754e-01
-1.86167657e-02 3.95043164e-01 3.83269370e-01 7.89272428e-01
-1.01352251e+00 5.71737178e-02 9.96831805e-02 -9.06952560e-01
-1.19059086e+00 1.63032636e-02 6.95774436e-01 3.00027281e-01
-9.15174961e-01 9.27612484e-01 5.65806985e-01 -4.61071789e-01
-1.59464300e-01 -4.10143077e-01 6.93246722e-02 -6.50092840e-01
5.30592442e-01 1.91938654e-01 -3.20621222e-01 -3.12716901e-01
-5.58235049e-02 7.51260281e-01 3.10522914e-01 -8.12611282e-01
1.47801304e+00 -4.90860790e-01 2.42161632e-01 2.83898264e-01
9.53069866e-01 4.25578624e-01 -1.99928081e+00 -2.15223610e-01
-9.24793124e-01 -6.90171897e-01 6.75452948e-01 -7.13829100e-01
-8.63766313e-01 8.07665527e-01 4.31375146e-01 1.11846261e-01
1.14245069e+00 -2.50975847e-01 9.08956409e-01 3.99442106e-01
3.34491730e-01 -7.22210765e-01 1.64615601e-01 4.12367254e-01
7.37246454e-01 -1.27186811e+00 4.94345337e-01 -6.67423725e-01
-4.39292848e-01 1.00870287e+00 2.92282522e-01 -3.06694686e-01
6.82252407e-01 1.03018570e+00 4.03227001e-01 -4.06192094e-01
-6.49570644e-01 3.40951197e-02 1.59256876e-01 4.47545409e-01
6.30278662e-02 -1.63356021e-01 2.09347337e-01 2.69532297e-02
-2.51953751e-01 -1.62789226e-01 1.00399292e+00 1.06585765e+00
-7.23753154e-01 -4.83940601e-01 -5.04320025e-01 2.66095757e-01
-2.96837956e-01 -2.24279106e-01 2.97396719e-01 5.41714668e-01
-4.58999485e-01 1.06336117e+00 1.34520859e-01 -7.91559070e-02
1.27647832e-01 -6.94608629e-01 8.27942252e-01 -7.09065437e-01
1.83402121e-01 6.66853726e-01 2.42426708e-01 -6.79548740e-01
-5.90116799e-01 -7.78097093e-01 -1.42791450e+00 1.17826045e-01
-3.19093794e-01 7.34401401e-03 9.30103481e-01 8.04251790e-01
3.55875254e-01 7.19083607e-01 6.76612973e-01 -1.19536459e+00
-4.05055374e-01 -6.33234978e-01 -6.51380718e-01 -1.90512836e-01
8.36675823e-01 -6.32766902e-01 -6.73509419e-01 1.76072925e-01] | [11.263164520263672, -2.618006467819214] |
fbca6167-1e29-4184-8916-37cea808b8a6 | finding-the-most-transferable-tasks-for-brain | 2301.00934 | null | https://arxiv.org/abs/2301.00934v1 | https://arxiv.org/pdf/2301.00934v1.pdf | Finding the Most Transferable Tasks for Brain Image Segmentation | Although many studies have successfully applied transfer learning to medical image segmentation, very few of them have investigated the selection strategy when multiple source tasks are available for transfer. In this paper, we propose a prior knowledge guided and transferability based framework to select the best source tasks among a collection of brain image segmentation tasks, to improve the transfer learning performance on the given target task. The framework consists of modality analysis, RoI (region of interest) analysis, and transferability estimation, such that the source task selection can be refined step by step. Specifically, we adapt the state-of-the-art analytical transferability estimation metrics to medical image segmentation tasks and further show that their performance can be significantly boosted by filtering candidate source tasks based on modality and RoI characteristics. Our experiments on brain matter, brain tumor, and white matter hyperintensities segmentation datasets reveal that transferring from different tasks under the same modality is often more successful than transferring from the same task under different modalities. Furthermore, within the same modality, transferring from the source task that has stronger RoI shape similarity with the target task can significantly improve the final transfer performance. And such similarity can be captured using the Structural Similarity index in the label space. | ['Xiao-Ping Zhang', 'Yang Li', 'Jingyun Yang', 'Yang Tan', 'Yicong Li'] | 2023-01-03 | null | null | null | null | ['brain-image-segmentation'] | ['medical'] | [ 6.00774288e-01 -5.54315280e-04 -2.42200956e-01 -4.68804926e-01
-1.03115869e+00 -3.85468185e-01 5.17211020e-01 2.43919566e-01
-6.90110207e-01 7.39096642e-01 8.69912580e-02 -9.42656025e-02
-5.08507252e-01 -6.81892633e-01 -6.57174706e-01 -9.21492219e-01
2.46453851e-01 4.90494519e-01 5.21489978e-01 1.45853944e-02
4.53299880e-01 2.59637505e-01 -1.13930547e+00 4.59869266e-01
1.39995027e+00 1.08618021e+00 6.03650093e-01 -3.13866436e-02
-2.38985494e-01 5.40652812e-01 -2.60027707e-01 -1.44453585e-01
1.39865503e-01 -7.45655894e-01 -1.25675011e+00 2.04958227e-02
1.90322921e-01 1.78961992e-01 1.83888480e-01 1.21806645e+00
5.98527610e-01 1.31546333e-01 1.04550660e+00 -1.10112238e+00
-5.28841317e-01 6.39401138e-01 -5.44896066e-01 5.24428248e-01
-6.13308437e-02 -1.45754263e-01 7.30184615e-01 -7.91164815e-01
6.89177811e-01 9.01036799e-01 3.95437151e-01 3.33208829e-01
-1.16969216e+00 -7.83318102e-01 1.32824212e-01 3.98341000e-01
-1.14531326e+00 -1.00962214e-01 8.76399279e-01 -6.80350959e-01
3.50734442e-01 -2.84907073e-02 3.02790910e-01 8.24125290e-01
1.61510482e-01 8.75857770e-01 1.57171059e+00 -3.84131551e-01
1.34257063e-01 2.59508610e-01 2.63929486e-01 5.85518837e-01
-4.78979088e-02 -1.56172529e-01 -3.60652357e-01 1.12215191e-01
6.14635646e-01 -7.07534999e-02 -3.80423307e-01 -4.87729788e-01
-1.73747385e+00 8.61952305e-01 9.03230786e-01 7.98624039e-01
-4.95503217e-01 -4.04030412e-01 4.38644379e-01 4.04998779e-01
6.64304614e-01 5.08057714e-01 -4.82765228e-01 3.07059914e-01
-9.70730722e-01 -2.75212944e-01 3.32085758e-01 7.22201109e-01
1.00121737e+00 -4.17875051e-01 -6.68566823e-01 1.02160144e+00
1.36650875e-01 2.78619379e-01 7.58725762e-01 -6.96435750e-01
4.30724651e-01 6.71443462e-01 -4.24481213e-01 -6.12239599e-01
-4.76167083e-01 -5.71473777e-01 -6.89061701e-01 7.33992532e-02
5.82641661e-01 -9.99133140e-02 -1.02709317e+00 1.67888319e+00
4.02138650e-01 2.01409683e-01 -1.10771686e-01 9.63531435e-01
8.45668077e-01 2.69441664e-01 2.59138137e-01 -2.19319835e-01
1.44818556e+00 -1.04612744e+00 -3.87022853e-01 -1.34932265e-01
7.86070347e-01 -6.64283216e-01 1.02766383e+00 5.15433699e-02
-7.81446040e-01 -5.72542012e-01 -7.93605566e-01 3.57554913e-01
-2.36235291e-01 -3.99611099e-03 3.72014463e-01 4.45114195e-01
-8.14288616e-01 5.46560228e-01 -6.14494383e-01 -3.80336076e-01
5.85460186e-01 2.83768773e-01 -2.79205769e-01 -1.44562632e-01
-1.21496451e+00 1.10252690e+00 7.08623588e-01 -3.78186733e-01
-8.40052545e-01 -1.02153456e+00 -4.49425071e-01 -7.02490732e-02
4.62571532e-01 -7.22311974e-01 7.54864693e-01 -1.33341789e+00
-1.30999458e+00 1.07035935e+00 8.82247239e-02 -3.08376133e-01
5.50498068e-01 2.59364486e-01 -2.55214930e-01 4.91812408e-01
3.24771672e-01 9.03618336e-01 1.10844195e+00 -1.20890558e+00
-7.94283807e-01 -5.52523792e-01 -9.01371092e-02 4.09469604e-01
-5.07132292e-01 9.51455832e-02 -2.58096367e-01 -6.03544116e-01
9.15763527e-02 -9.85706687e-01 6.53119683e-02 -1.84530199e-01
-2.10627005e-01 -3.62222821e-01 4.85298306e-01 -7.07942069e-01
7.72822261e-01 -1.96683025e+00 5.22358179e-01 4.34667885e-01
1.72764167e-01 5.73785082e-02 -2.69841939e-01 -1.96199790e-01
-1.61954790e-01 1.11422099e-01 -4.37742949e-01 1.54675499e-01
-1.97613731e-01 -1.45443574e-01 2.39989191e-01 3.46365213e-01
1.62570011e-02 8.35329950e-01 -8.39023471e-01 -9.22876954e-01
6.37575909e-02 2.26965621e-01 -2.82953978e-01 1.95839092e-01
2.46623248e-01 1.17235541e+00 -7.81100571e-01 5.61572254e-01
4.77035552e-01 -3.26214254e-01 -1.77607939e-01 -4.76519883e-01
2.83164561e-01 -1.41409844e-01 -6.06352508e-01 1.90460253e+00
-5.72042644e-01 3.65645707e-01 9.34093520e-02 -1.47430277e+00
9.77732420e-01 2.85299003e-01 1.01005781e+00 -9.06384885e-01
2.65590131e-01 3.60432595e-01 4.92979378e-01 -5.32027543e-01
-3.35174352e-02 -3.56320322e-01 1.73225120e-01 5.36199272e-01
2.08281368e-01 -3.08828596e-02 1.63769424e-01 -1.00587696e-01
8.58671248e-01 1.55272484e-01 1.45200759e-01 -4.42213506e-01
7.32200384e-01 2.66943052e-02 4.47067201e-01 5.76351166e-01
-5.33309519e-01 5.41369021e-01 2.41467044e-01 5.97011819e-02
-7.33838022e-01 -1.03571105e+00 -4.86867309e-01 1.49015057e+00
1.92739964e-01 3.66893649e-01 -9.54937160e-01 -1.10863638e+00
-1.50382295e-01 5.34297705e-01 -8.40852499e-01 -3.04242402e-01
-5.65075397e-01 -9.05955195e-01 4.28417593e-01 5.13908863e-01
7.28007853e-01 -1.42125726e+00 -5.38444698e-01 -4.38292176e-02
-4.47905660e-01 -9.39746141e-01 -6.88185453e-01 1.57287031e-01
-1.11600232e+00 -1.15757251e+00 -1.43958712e+00 -1.07351339e+00
8.72795224e-01 2.30855197e-01 9.31885600e-01 9.41769686e-04
2.99003664e-02 5.29832184e-01 -6.06579661e-01 -2.11571410e-01
-4.77772743e-01 4.29808319e-01 -3.48376781e-01 3.32198918e-01
1.20785981e-01 -3.40812325e-01 -6.47355616e-01 6.24276936e-01
-9.92367983e-01 1.77235365e-01 8.70311499e-01 9.02526438e-01
6.96337521e-01 -1.63386047e-01 8.72612178e-01 -8.63125145e-01
5.54999173e-01 -6.41425192e-01 -8.06436986e-02 5.80306947e-01
-8.07206810e-01 -7.15048378e-03 2.63691515e-01 -5.87372839e-01
-1.32969511e+00 -1.06294908e-01 2.82195151e-01 -3.27315211e-01
-2.56864816e-01 8.27816546e-01 -5.18873073e-02 -2.70229757e-01
6.94737911e-01 2.42215320e-01 1.56336352e-01 -3.21140856e-01
2.05378905e-01 4.36441243e-01 2.23025948e-01 -7.50004828e-01
4.51125383e-01 5.01767397e-01 -4.10692319e-02 -3.58863175e-01
-7.87218034e-01 -5.80482960e-01 -1.11565506e+00 -4.09271389e-01
1.19029069e+00 -5.21360219e-01 -2.26345181e-01 3.87184173e-01
-7.80008018e-01 -4.40292746e-01 -6.47876561e-02 7.36797869e-01
-6.27063811e-01 3.53279144e-01 -3.14599007e-01 -1.38718918e-01
-2.99762011e-01 -1.59459162e+00 8.73871326e-01 2.24129096e-01
7.69688785e-02 -1.24325275e+00 -8.95307213e-02 5.71759760e-01
5.12197733e-01 1.72571927e-01 1.27302945e+00 -9.74454880e-01
-2.63638347e-01 2.21271604e-01 -4.97218460e-01 3.68264318e-01
4.72941369e-01 -5.53868830e-01 -6.45072699e-01 -4.07576561e-01
1.03035972e-01 -2.22740859e-01 1.19108880e+00 7.29675651e-01
1.18116879e+00 2.30723396e-01 -5.64728916e-01 4.37546104e-01
1.14158547e+00 2.99609393e-01 4.61578429e-01 4.87292141e-01
7.68723667e-01 9.95272815e-01 7.79342473e-01 -1.73927933e-01
2.61094213e-01 6.24549448e-01 1.54702812e-01 -4.10006523e-01
-3.77196670e-01 2.87862509e-01 3.00751626e-01 8.21903527e-01
-2.23903835e-01 1.76178768e-01 -1.09037912e+00 5.86284220e-01
-1.84533191e+00 -6.64086044e-01 1.56642005e-01 2.18979526e+00
9.96160567e-01 1.19578037e-02 2.38277867e-01 -1.10572070e-01
1.05117702e+00 -1.47205859e-01 -6.59868360e-01 1.18388541e-01
1.25456288e-01 4.15323414e-02 4.09744978e-01 1.54397681e-01
-1.09949219e+00 7.60901093e-01 5.61566734e+00 1.08785057e+00
-1.14988184e+00 5.03569424e-01 8.07160437e-01 3.02754313e-01
-2.61146486e-01 -1.52174443e-01 -3.98279130e-01 4.43168402e-01
6.36795640e-01 -2.18009397e-01 3.02638650e-01 4.14454788e-01
-8.94470662e-02 -2.42566615e-01 -1.22281611e+00 7.51508534e-01
1.68473169e-01 -9.04753447e-01 5.79888821e-02 8.34439322e-02
7.48794794e-01 -2.84228027e-02 1.67388603e-01 3.94208640e-01
-1.17724471e-01 -9.21726048e-01 4.32516575e-01 7.13264585e-01
5.48883975e-01 -6.24497354e-01 7.81231761e-01 2.37918600e-01
-1.08373404e+00 -3.04000191e-02 -2.68104017e-01 6.03179574e-01
-4.73678820e-02 5.16341388e-01 -8.57818305e-01 7.94744730e-01
7.80668259e-01 6.99674487e-01 -7.68710494e-01 1.25068045e+00
6.41920865e-02 7.54764140e-01 1.13973543e-01 1.70557365e-01
1.08081348e-01 -4.59279031e-01 5.86298883e-01 1.08110535e+00
3.96220177e-01 -1.78501844e-01 4.78973389e-01 9.01943505e-01
1.79549363e-02 6.08534455e-01 -3.99376035e-01 4.35952842e-01
2.17714369e-01 1.32097375e+00 -1.29404008e+00 -3.69366854e-01
-3.69998097e-01 7.73956358e-01 2.04385430e-01 3.94588947e-01
-8.83015335e-01 -2.24977896e-01 -2.15751920e-02 2.35483591e-02
3.04198384e-01 3.22204560e-01 -5.26280105e-01 -1.00748682e+00
-2.62203842e-01 -6.26564205e-01 6.46454036e-01 -6.17521465e-01
-1.62853539e+00 7.23462641e-01 4.12695885e-01 -1.36958992e+00
1.17332079e-01 -4.91538674e-01 -5.94638407e-01 7.83328235e-01
-1.65917373e+00 -1.24770629e+00 -4.18904901e-01 9.09688175e-01
3.94471616e-01 -3.33296925e-01 5.26035845e-01 4.40259188e-01
-3.15338463e-01 4.90372598e-01 1.89764276e-02 1.14441939e-01
1.04197097e+00 -1.18871939e+00 -6.08620167e-01 5.26209831e-01
-1.29488081e-01 3.04714501e-01 1.97399721e-01 -6.57790184e-01
-7.53000557e-01 -1.16002178e+00 2.86301076e-01 -1.77636489e-01
5.86280942e-01 1.71752766e-01 -1.26294303e+00 4.28705305e-01
2.45536432e-01 -1.57708395e-02 7.46761799e-01 1.30799994e-01
-1.77006245e-01 -1.41180858e-01 -1.18716872e+00 3.55400294e-01
8.15531373e-01 -4.09681052e-01 -8.66682112e-01 4.50046957e-01
5.37006736e-01 -2.19255865e-01 -1.33030605e+00 5.53007901e-01
2.43653417e-01 -5.72592318e-01 9.68182564e-01 -5.86643755e-01
4.26096767e-01 -1.69520244e-01 9.37512368e-02 -1.61108732e+00
-4.45881307e-01 1.08404547e-01 4.50925738e-01 1.29434931e+00
5.91257811e-01 -6.74395323e-01 4.41997975e-01 4.90962863e-01
-3.77035856e-01 -4.97098655e-01 -1.08019459e+00 -8.46406162e-01
6.65562034e-01 -4.35889550e-02 4.05026197e-01 1.09028614e+00
-1.33676991e-01 3.29044461e-01 -7.17561617e-02 -1.82287857e-01
6.24680877e-01 3.39326173e-01 3.66859198e-01 -1.44705951e+00
-1.37169249e-02 -7.98701763e-01 -1.36829153e-01 -4.46332604e-01
4.56438154e-01 -1.66108465e+00 2.82545298e-01 -1.61918116e+00
5.48726797e-01 -5.44923782e-01 -1.00555885e+00 7.14815915e-01
-4.80094433e-01 3.05196792e-01 2.33395949e-01 5.73743463e-01
-5.88460207e-01 4.18829650e-01 1.78996062e+00 -3.70898306e-01
-2.63046533e-01 1.23041444e-01 -6.33013546e-01 5.88278949e-01
8.08297694e-01 -6.02318108e-01 -4.75170016e-01 -2.82225192e-01
-1.91716269e-01 1.68396279e-01 3.13002855e-01 -9.14118588e-01
1.26691803e-01 -1.23559833e-01 3.90458345e-01 -1.58997700e-01
-7.52476305e-02 -8.38729739e-01 -2.45622739e-01 5.41584551e-01
-5.84120393e-01 -2.82882482e-01 9.91141945e-02 4.77289289e-01
-2.20042646e-01 -4.78664368e-01 1.01965058e+00 -2.94935610e-02
-7.77243912e-01 4.20191079e-01 -2.96431094e-01 1.44913018e-01
1.14310360e+00 -3.15525085e-01 -1.27451897e-01 -8.51630419e-02
-8.88545394e-01 4.29184809e-02 3.76594514e-02 4.15958792e-01
4.83591855e-01 -1.35748255e+00 -1.04740405e+00 -2.85322100e-01
2.39947245e-01 -3.43247354e-01 2.54968822e-01 1.51130569e+00
4.74792942e-02 2.22486958e-01 -7.83695877e-01 -9.48144555e-01
-1.21106207e+00 4.38444018e-01 3.80498350e-01 -5.22595584e-01
-4.61955577e-01 5.35913527e-01 6.12906456e-01 -4.63809520e-01
-1.04009859e-01 -3.01212281e-01 -5.94600797e-01 3.65236849e-01
2.49053538e-01 5.37058175e-01 3.08202833e-01 -8.43455851e-01
-4.04687136e-01 8.17745507e-01 -1.02980323e-01 -1.83713958e-02
1.17808390e+00 -1.28902659e-01 -1.84376284e-01 3.19872916e-01
1.25223160e+00 -3.62672538e-01 -1.07945955e+00 -6.69290245e-01
1.17995270e-01 -1.90241784e-01 2.79358149e-01 -9.39953566e-01
-1.54185224e+00 8.10001075e-01 9.49335635e-01 1.04965262e-01
1.32575393e+00 2.67494738e-01 6.68476582e-01 6.79392740e-02
3.90322655e-01 -1.15020597e+00 2.92513460e-01 4.54686284e-01
8.18629026e-01 -1.43252397e+00 -1.26642808e-01 -5.25536954e-01
-9.99581218e-01 8.90548885e-01 6.30081356e-01 -3.73262689e-02
8.41838539e-01 -2.63382703e-01 -1.65754557e-02 -3.05432439e-01
-7.12837577e-02 -3.59096617e-01 8.86109650e-01 6.18785858e-01
3.71022344e-01 1.79600224e-01 -4.16026235e-01 5.03846586e-01
2.87459075e-01 -1.37478928e-03 -1.95684619e-02 6.70658946e-01
-4.77376580e-01 -9.76297438e-01 -4.41804111e-01 8.46598268e-01
-3.57777983e-01 -2.28412170e-02 -1.19021326e-01 5.80667078e-01
1.21775940e-01 7.91704476e-01 -4.92116436e-02 -1.47256956e-01
1.46857440e-01 2.05337077e-01 6.92922652e-01 -6.49850786e-01
-8.33763301e-01 1.57404035e-01 -3.11629057e-01 -2.94153214e-01
-9.93590951e-01 -5.94912589e-01 -1.38723695e+00 2.82856137e-01
-4.30635870e-01 1.00611605e-01 3.87452424e-01 1.27339363e+00
2.22548455e-01 7.80988395e-01 6.39121056e-01 -8.10611427e-01
-2.61294007e-01 -1.14486504e+00 -5.07568240e-01 7.58074641e-01
-1.23706840e-01 -1.07186472e+00 -6.32044971e-02 1.71583444e-01] | [14.561254501342773, -2.0802745819091797] |
cf60f5df-fb30-4d8f-bbe5-55149ce0e5e9 | task-arithmetic-in-the-tangent-space-improved | 2305.12827 | null | https://arxiv.org/abs/2305.12827v2 | https://arxiv.org/pdf/2305.12827v2.pdf | Task Arithmetic in the Tangent Space: Improved Editing of Pre-Trained Models | Task arithmetic has recently emerged as a cost-effective and scalable approach to edit pre-trained models directly in weight space: By adding the fine-tuned weights of different tasks, the model's performance can be improved on these tasks, while negating them leads to task forgetting. Yet, our understanding of the effectiveness of task arithmetic and its underlying principles remains limited. We present a comprehensive study of task arithmetic in vision-language models and show that weight disentanglement is the crucial factor that makes it effective. This property arises during pre-training and manifests when distinct directions in weight space govern separate, localized regions in function space associated with the tasks. Notably, we show that fine-tuning models in their tangent space by linearizing them amplifies weight disentanglement. This leads to substantial performance improvements across multiple task arithmetic benchmarks and diverse models. Building on these findings, we provide theoretical and empirical analyses of the neural tangent kernel (NTK) of these models and establish a compelling link between task arithmetic and the spatial localization of the NTK eigenfunctions. Overall, our work uncovers novel insights into the fundamental mechanisms of task arithmetic and offers a more reliable and effective approach to edit pre-trained models through the NTK linearization. | ['Pascal Frossard', 'Alessandro Favero', 'Guillermo Ortiz-Jimenez'] | 2023-05-22 | null | null | null | null | ['disentanglement'] | ['methodology'] | [ 4.16653514e-01 -2.91297376e-01 -9.86665674e-03 -3.48854899e-01
-3.08039963e-01 -4.63025868e-01 7.05950797e-01 1.53869167e-01
-6.17204249e-01 2.96397537e-01 5.12247980e-01 -2.25840017e-01
-3.54938567e-01 -3.83533001e-01 -7.74266422e-01 -6.96871579e-01
1.33561134e-01 2.14985982e-01 7.33624175e-02 -4.48044389e-01
7.67071903e-01 2.53406525e-01 -1.20414484e+00 3.67945850e-01
1.07580292e+00 6.25590384e-01 4.64313954e-01 7.16359437e-01
4.91415858e-02 6.63891971e-01 -5.91059089e-01 -5.43404520e-01
2.97898620e-01 -1.36944816e-01 -5.95857799e-01 -5.43996394e-01
9.53876019e-01 -3.79302800e-01 -5.07794440e-01 9.05844212e-01
3.23172390e-01 2.52080083e-01 1.01284611e+00 -9.98598516e-01
-1.27642357e+00 6.67209566e-01 -4.42900926e-01 4.29404795e-01
-1.92211151e-01 3.25647295e-01 1.08561432e+00 -9.76759672e-01
2.42045179e-01 1.04216099e+00 9.32773292e-01 4.79446590e-01
-1.78335381e+00 -4.61588055e-01 2.00340852e-01 1.95299596e-01
-1.28394532e+00 -6.99201524e-01 5.99881351e-01 -7.20344067e-01
1.27019918e+00 8.20830241e-02 8.95553589e-01 1.01284266e+00
8.68357062e-01 5.92197657e-01 1.09877348e+00 -5.23717165e-01
-1.16868988e-01 -2.80970931e-01 4.48466331e-01 7.33069241e-01
4.41684395e-01 1.37725130e-01 -1.05445135e+00 -4.52277288e-02
9.80419517e-01 -1.40610486e-01 -3.53278190e-01 -5.53545773e-01
-1.45540571e+00 4.99547601e-01 4.48472261e-01 9.17247832e-02
-1.64411113e-01 7.06709504e-01 5.83125532e-01 3.69665414e-01
2.81972319e-01 1.05606079e+00 -5.37310898e-01 -1.71859190e-01
-8.62872660e-01 2.50305504e-01 2.21293658e-01 7.83302069e-01
6.88981116e-01 4.15847749e-01 -5.85875869e-01 7.05538869e-01
1.18848920e-01 4.01100785e-01 5.16742229e-01 -1.22426546e+00
4.64143693e-01 3.55680972e-01 -1.37250289e-01 -9.91180301e-01
-5.12238622e-01 -6.62658989e-01 -4.43584800e-01 3.85019839e-01
7.71418750e-01 1.91105887e-01 -7.78081477e-01 2.23744178e+00
-2.40100354e-01 -4.61483002e-02 -3.11286151e-01 6.37808681e-01
2.14533761e-01 2.26283908e-01 4.21819061e-01 2.53335118e-01
1.34603202e+00 -8.69169712e-01 -6.95615232e-01 -4.22134697e-01
7.96963513e-01 -8.11306417e-01 1.56063795e+00 2.69210547e-01
-1.43142724e+00 -7.06726551e-01 -1.21209300e+00 -6.07169509e-01
-3.77129734e-01 1.42792150e-01 9.52762425e-01 6.13946855e-01
-1.22884977e+00 8.62485588e-01 -7.74462163e-01 -6.50462732e-02
6.40088618e-01 4.26492155e-01 -1.75693855e-01 -1.47933230e-01
-9.80477631e-01 1.38583326e+00 2.49787167e-01 5.15980944e-02
-7.05131292e-01 -1.36093688e+00 -7.54668593e-01 3.10970873e-01
4.99603413e-02 -1.11719286e+00 1.18968832e+00 -5.28251410e-01
-1.27602565e+00 7.18091905e-01 -2.08475351e-01 -2.98782945e-01
8.33648369e-02 -4.46967214e-01 1.00082248e-01 -1.07885525e-01
-1.51241213e-01 7.33342707e-01 1.21043754e+00 -9.77605283e-01
-9.63897705e-02 -3.85234475e-01 1.31205255e-02 3.17094415e-01
-6.30367219e-01 -2.23474011e-01 -5.21577075e-02 -8.88076603e-01
4.85844575e-02 -6.87164783e-01 2.95556337e-01 2.45298758e-01
-3.51049416e-02 -3.62772830e-02 2.03826249e-01 -7.47740746e-01
1.30214703e+00 -2.11046100e+00 4.41004634e-01 3.47093008e-02
7.84236014e-01 8.22198242e-02 -3.45782965e-01 3.29489380e-01
4.84970100e-02 2.94249109e-03 -1.66364461e-01 -3.65519613e-01
2.11777836e-01 4.79001254e-02 -5.45594752e-01 4.20091182e-01
2.26769656e-01 1.46150470e+00 -9.15559053e-01 -1.10265486e-01
2.76321054e-01 5.93192339e-01 -8.18034112e-01 -2.50420272e-01
8.42021778e-03 1.80341929e-01 -7.43677095e-02 1.89233720e-01
6.37724876e-01 -1.32561997e-01 -4.11948003e-02 -7.03661025e-01
-2.65790485e-02 4.86810178e-01 -5.47750294e-01 1.84717858e+00
-4.68142420e-01 1.03306746e+00 -1.62553057e-01 -1.04460025e+00
4.88029480e-01 -1.49295717e-01 3.23312916e-02 -8.52448285e-01
-6.73913956e-03 8.53611454e-02 5.48498750e-01 -3.02644759e-01
7.21802592e-01 -1.11117676e-01 1.94298044e-01 5.66158116e-01
3.60922605e-01 -4.75618571e-01 6.33366555e-02 2.15554014e-01
9.54782009e-01 2.32668325e-01 1.89479411e-01 -6.58751011e-01
7.44123906e-02 -1.14926867e-01 2.05941051e-02 9.77546096e-01
-4.70837355e-01 4.51894969e-01 3.61126631e-01 -3.42407405e-01
-1.23813105e+00 -1.38053656e+00 -2.72757620e-01 1.72129869e+00
-3.45625788e-01 -5.42530000e-01 -6.43444955e-01 -2.51465231e-01
2.37107113e-01 8.44896019e-01 -1.06546187e+00 -9.25284624e-01
-8.21221352e-01 -5.10290623e-01 8.47466409e-01 6.42687201e-01
4.52123791e-01 -6.50017083e-01 -6.57532454e-01 -2.87300032e-02
-1.62920818e-01 -8.21559310e-01 -8.73778045e-01 3.02716762e-01
-1.17343223e+00 -9.09990132e-01 -8.35921764e-01 -6.21857226e-01
7.24617958e-01 6.05304956e-01 1.36643159e+00 9.94700491e-02
-3.75281125e-01 5.24675667e-01 3.01539660e-01 -4.34005231e-01
-1.35581300e-01 3.12947720e-01 2.57847160e-02 -4.57683206e-01
2.59542465e-01 -5.03601432e-01 -6.70430422e-01 1.56853080e-01
-6.65769160e-01 1.97086662e-01 5.32084405e-01 8.31613839e-01
3.93164493e-02 -3.60839188e-01 5.42507648e-01 -7.81866312e-01
1.10313773e+00 -3.04898359e-02 -3.82736176e-01 3.71639460e-01
-6.99381113e-01 5.76260388e-01 4.49483901e-01 -5.29701889e-01
-1.24062455e+00 -2.31690541e-01 2.36791208e-01 -1.78634375e-01
4.01671112e-01 3.85174036e-01 4.29472238e-01 -3.74232143e-01
1.01869154e+00 3.06483626e-01 -1.23791449e-01 -1.17275380e-01
6.30031288e-01 -1.00799628e-01 4.65487957e-01 -9.39946949e-01
8.10311913e-01 5.11096299e-01 -4.70283777e-02 -8.12863410e-01
-9.36266124e-01 -1.49634361e-01 -6.41739011e-01 -5.45733571e-02
9.54879999e-01 -8.56706679e-01 -7.78601110e-01 6.11295640e-01
-1.13225532e+00 -8.49461317e-01 -3.94106060e-01 5.04432201e-01
-5.39700091e-01 1.79793343e-01 -7.19429255e-01 -4.49662954e-01
-8.52522254e-02 -1.07507908e+00 9.69129145e-01 3.57004330e-02
-5.31937718e-01 -1.29425609e+00 -1.90129019e-02 2.19860569e-01
9.51497018e-01 -3.63696188e-01 1.71271420e+00 -2.07708910e-01
-4.57387537e-01 1.21473819e-01 -4.99750763e-01 2.80720115e-01
-1.25534743e-01 -5.51084653e-02 -1.11252105e+00 -1.06418796e-01
2.14388013e-01 -4.80272442e-01 1.29365706e+00 8.17870617e-01
1.21630764e+00 1.83012448e-02 -7.19077215e-02 7.56510139e-01
1.03427029e+00 -2.73774713e-01 6.30435050e-01 1.52837485e-01
7.56901264e-01 5.53779006e-01 1.00113712e-01 1.06345743e-01
3.82115543e-01 6.87853158e-01 8.40821266e-02 1.44159272e-01
-3.28369439e-01 -2.08917648e-01 3.53875965e-01 8.94686818e-01
-3.84581238e-01 4.57185835e-01 -9.13162351e-01 2.99591243e-01
-1.59190178e+00 -9.92577493e-01 -9.95604023e-02 2.42695069e+00
1.25300860e+00 3.41879457e-01 -4.01654691e-01 -3.40231843e-02
2.32986689e-01 3.07353854e-01 -5.90971708e-01 -5.56608558e-01
-1.53453708e-01 5.30499697e-01 4.83964175e-01 7.05112398e-01
-7.81792283e-01 1.04743624e+00 7.96614981e+00 8.49335074e-01
-9.61842954e-01 1.63998380e-01 2.82669365e-01 -2.65124649e-01
-3.21455449e-01 -2.64286011e-01 -5.66281438e-01 1.12654669e-02
8.96034598e-01 -2.22027242e-01 8.39766026e-01 5.44791222e-01
5.77301607e-02 -1.41004205e-01 -1.40760100e+00 7.59624898e-01
2.08148897e-01 -1.47814536e+00 1.17237851e-01 -2.98724277e-03
7.93639243e-01 1.45066958e-02 7.29978740e-01 6.59834325e-01
3.36453766e-01 -1.27517247e+00 9.12707627e-01 6.59240484e-01
8.05954933e-01 -3.03669184e-01 1.57226369e-01 1.35889351e-01
-1.15913880e+00 -2.06511602e-01 -7.25274444e-01 -4.10126239e-01
-4.15017791e-02 4.28951681e-01 -5.26906908e-01 -6.30556867e-02
2.76080221e-01 4.98871356e-01 -7.95590281e-01 9.11951125e-01
-1.50078714e-01 3.49504650e-01 2.69522637e-01 3.19146633e-01
-1.82541564e-01 -1.41555816e-01 2.00523168e-01 1.12870526e+00
2.08687201e-01 -2.36458960e-03 -3.51812065e-01 1.18516636e+00
-2.55281595e-03 -2.71224380e-01 -6.16121471e-01 -1.03041604e-01
3.83319676e-01 8.89841497e-01 -4.07934040e-01 -2.99562305e-01
-3.13308358e-01 1.14609313e+00 8.60183060e-01 8.00148010e-01
-9.15368438e-01 -4.16001737e-01 9.58936155e-01 -2.22042631e-02
5.06228656e-02 -8.90502572e-01 -1.01372623e+00 -1.04225552e+00
-1.06818408e-01 -6.95864618e-01 -1.47303149e-01 -1.00693345e+00
-1.07695079e+00 1.29362345e-01 9.49676335e-02 -4.68976468e-01
1.99726805e-01 -1.08156157e+00 -4.69258130e-01 1.18113196e+00
-1.51817036e+00 -9.71702516e-01 -2.26913527e-01 5.27596176e-01
6.23278201e-01 6.23735366e-03 9.42432046e-01 1.45874649e-01
-3.54223639e-01 8.77984762e-01 2.09914476e-01 -1.11376822e-01
1.03365004e+00 -1.24353611e+00 6.89129770e-01 5.33267796e-01
1.07077248e-01 1.46594954e+00 6.21859729e-01 -6.21729493e-01
-1.43227565e+00 -7.90456414e-01 8.87413859e-01 -9.65529144e-01
6.56280756e-01 -6.43040955e-01 -1.06902814e+00 7.58349955e-01
1.68387845e-01 -2.72314757e-01 6.62040949e-01 3.77850384e-01
-1.01381874e+00 6.72010332e-02 -7.25904703e-01 9.71383452e-01
1.27153015e+00 -9.14704978e-01 -8.13296854e-01 2.99877644e-01
8.37807834e-01 -3.66230041e-01 -5.97203195e-01 1.05477445e-01
8.52834642e-01 -9.68253434e-01 1.35466242e+00 -8.81608605e-01
6.20329916e-01 3.25932689e-02 -1.77476183e-01 -1.62335694e+00
-7.76342750e-01 -2.98417896e-01 -2.80779332e-01 5.97480536e-01
4.43602085e-01 -6.69815183e-01 3.52628678e-01 9.31514680e-01
-4.43332911e-01 -7.21737444e-01 -8.09210658e-01 -6.95938349e-01
5.26569307e-01 -5.23950338e-01 2.41630808e-01 8.58541250e-01
-9.16489516e-04 2.27621466e-01 -1.11075051e-01 -3.32270533e-01
5.68339109e-01 -5.25458574e-01 3.97895575e-01 -1.21192038e+00
-2.48111591e-01 -8.57470632e-01 -1.90904409e-01 -1.26346111e+00
2.47845933e-01 -9.97726202e-01 -8.77568349e-02 -1.38465858e+00
2.52694190e-01 -5.40765226e-01 -3.47352743e-01 3.71573210e-01
-3.10598105e-01 2.11493999e-01 4.24585313e-01 3.91521573e-01
-2.65214086e-01 6.59250677e-01 1.46171296e+00 -9.21099409e-02
-6.67327344e-02 -3.52530271e-01 -9.25553441e-01 7.61724234e-01
8.44736636e-01 -2.35071927e-01 -6.40438139e-01 -1.22904289e+00
5.00483811e-01 -6.67486131e-01 4.15589005e-01 -9.93061006e-01
1.36998519e-01 -3.00009698e-01 6.79695249e-01 -4.31889892e-02
5.43392241e-01 -5.25170505e-01 -5.65754235e-01 5.97741485e-01
-7.26052344e-01 2.99948514e-01 4.86604005e-01 5.33360183e-01
3.85539711e-01 -2.21257672e-01 7.73392379e-01 -3.79462875e-02
-6.25558078e-01 -1.15656644e-01 -4.14064586e-01 4.83801067e-01
5.66997468e-01 -3.71882707e-01 -8.42249572e-01 -8.47568959e-02
-4.58702803e-01 -1.44407451e-01 3.47530484e-01 2.65835911e-01
5.48632324e-01 -1.38578725e+00 -6.86641693e-01 3.60660911e-01
1.93077311e-01 -5.04823864e-01 4.62294489e-01 1.05047309e+00
-3.30291808e-01 5.63401937e-01 -5.39812624e-01 -4.70749825e-01
-8.29395711e-01 2.49024346e-01 2.92784363e-01 -2.06662834e-01
-3.53765279e-01 1.06715107e+00 5.42254269e-01 -3.02449077e-01
2.78821737e-01 -7.84339607e-01 6.58362545e-03 -8.31370130e-02
5.35390615e-01 3.80559653e-01 7.62379691e-02 -3.41592044e-01
-6.53484315e-02 7.62511373e-01 -3.05758536e-01 -2.79503763e-01
1.00291264e+00 -6.00109398e-02 -6.75017983e-02 6.13378882e-01
9.04233515e-01 -7.56846666e-02 -1.52271938e+00 -2.41484404e-01
-2.42237940e-01 -6.24883831e-01 -7.16820806e-02 -7.85818458e-01
-7.65819430e-01 1.26381063e+00 3.39564502e-01 -1.95003182e-01
8.52354288e-01 -2.44454339e-01 4.08298314e-01 5.96162736e-01
1.73899338e-01 -1.35801554e+00 5.51800549e-01 1.05601668e+00
1.09164715e+00 -1.04124200e+00 -1.69269647e-02 -1.92147881e-01
-6.38978839e-01 1.10252845e+00 7.49767959e-01 -2.57860124e-01
4.94181424e-01 5.38636036e-02 -2.65744358e-01 -2.55691767e-01
-6.86854064e-01 1.21346869e-01 4.57852542e-01 9.77910101e-01
7.47321367e-01 7.39675481e-03 -1.08740449e-01 4.72615600e-01
-4.80050892e-01 -2.03889489e-01 3.50005746e-01 8.19646657e-01
-6.62262857e-01 -8.40893447e-01 -2.22775444e-01 6.50059760e-01
-1.57417193e-01 -7.49979615e-01 -1.49465173e-01 7.51182139e-01
-6.14358373e-02 4.52439159e-01 6.51139244e-02 -2.48692274e-01
2.36266702e-01 6.13448501e-01 1.07405758e+00 -6.41745687e-01
-6.72665834e-01 -3.79272044e-01 -1.57078654e-01 -4.76373881e-01
1.51361957e-01 -3.56758296e-01 -1.11509907e+00 -5.37765741e-01
-2.69978225e-01 -4.20475304e-01 5.77324867e-01 8.32989633e-01
4.93137598e-01 9.24294591e-01 -1.13005124e-01 -1.05467141e+00
-9.42625880e-01 -8.35704029e-01 -4.28770155e-01 4.43817914e-01
2.97549665e-01 -9.38956141e-01 -1.40205100e-01 2.74576724e-01] | [10.009806632995605, 2.453312635421753] |
a2d320a8-4225-4bbb-be3d-299db5d574a7 | relation-mention-extraction-from-noisy-data | 1811.01237 | null | http://arxiv.org/abs/1811.01237v1 | http://arxiv.org/pdf/1811.01237v1.pdf | Relation Mention Extraction from Noisy Data with Hierarchical Reinforcement Learning | In this paper we address a task of relation mention extraction from noisy
data: extracting representative phrases for a particular relation from noisy
sentences that are collected via distant supervision. Despite its significance
and value in many downstream applications, this task is less studied on noisy
data. The major challenges exists in 1) the lack of annotation on mention
phrases, and more severely, 2) handling noisy sentences which do not express a
relation at all. To address the two challenges, we formulate the task as a
semi-Markov decision process and propose a novel hierarchical reinforcement
learning model. Our model consists of a top-level sentence selector to remove
noisy sentences, a low-level mention extractor to extract relation mentions,
and a reward estimator to provide signals to guide data denoising and mention
extraction without explicit annotations. Experimental results show that our
model is effective to extract relation mentions from noisy data. | ['Minlie Huang', 'Yijie Zhang', 'Jun Feng', 'Xiaoyan Zhu', 'Yang Yang'] | 2018-11-03 | null | null | null | null | ['relation-mention-extraction'] | ['natural-language-processing'] | [ 3.55222672e-01 7.70870745e-01 -1.48093939e-01 -5.44996500e-01
-1.52618563e+00 -2.58628488e-01 1.55899301e-01 4.13283527e-01
-5.33427656e-01 1.11963224e+00 6.05570316e-01 -1.43935949e-01
7.42794499e-02 -7.29327142e-01 -6.39770508e-01 -6.69537604e-01
5.88706098e-02 3.30527455e-01 5.36679514e-02 -1.83211312e-01
-8.83798823e-02 9.57560539e-02 -8.99141312e-01 2.81041026e-01
1.02091527e+00 7.19901502e-01 2.44816586e-01 5.08402884e-01
-1.66714475e-01 9.81256545e-01 -8.78043771e-01 -7.16098964e-01
-4.85256910e-02 -5.66347837e-01 -1.17975235e+00 2.60972053e-01
-2.31363386e-01 -1.56267971e-01 -2.25309029e-01 1.37127030e+00
7.41353750e-01 1.70346513e-01 3.83110881e-01 -7.43889749e-01
-4.38390821e-01 1.38109267e+00 -4.72461253e-01 3.46128434e-01
3.02876592e-01 1.11887805e-01 1.52581632e+00 -5.59602737e-01
7.44467318e-01 1.35848057e+00 6.11018240e-01 6.24259055e-01
-1.52810442e+00 -3.57715935e-01 5.87753832e-01 -2.71285675e-03
-1.14938772e+00 -5.71945548e-01 6.78918779e-01 -2.33211756e-01
1.19207335e+00 7.74418041e-02 3.92451555e-01 1.32578635e+00
4.78501199e-03 1.07624161e+00 7.45717108e-01 -2.18362674e-01
4.74746734e-01 -1.14457607e-01 3.74338418e-01 2.13688716e-01
9.41049382e-02 -1.98893771e-01 -7.79070079e-01 -1.97631702e-01
7.19836280e-02 -4.39092338e-01 -3.02109033e-01 2.30785042e-01
-7.66924143e-01 8.12866926e-01 3.32955360e-01 4.06175703e-01
-7.53095627e-01 2.06341356e-01 4.55986977e-01 4.25949842e-01
8.21660042e-01 4.87509459e-01 -7.36796141e-01 -2.00855166e-01
-7.68579125e-01 2.18533263e-01 9.35494006e-01 1.18736601e+00
6.74708545e-01 -2.50810355e-01 -3.01410466e-01 8.15880835e-01
2.38449350e-01 2.73762017e-01 2.79209703e-01 -7.12816596e-01
1.01586223e+00 3.18816036e-01 6.64995685e-02 -7.61871159e-01
-4.68361408e-01 -3.01748633e-01 -8.19129050e-01 -4.25019473e-01
2.44742677e-01 -5.93299806e-01 -6.58554792e-01 1.81275320e+00
3.82593423e-01 1.26686811e-01 4.30013776e-01 9.20391142e-01
1.06542087e+00 6.15162134e-01 4.11886215e-01 -5.31250060e-01
1.36849535e+00 -6.63001120e-01 -1.13135123e+00 -4.81678277e-01
7.37676978e-01 -4.58248436e-01 6.06319606e-01 2.49203071e-01
-1.02673519e+00 -4.22640406e-02 -7.02312231e-01 -2.93172389e-01
-8.45086947e-02 1.83820091e-02 5.14101923e-01 2.28050098e-01
-6.66219532e-01 7.99783766e-01 -9.06689286e-01 -8.47982094e-02
3.96026522e-01 4.72848415e-01 -1.84867293e-01 1.77654460e-01
-1.66865742e+00 8.88361156e-01 4.20703202e-01 4.50480551e-01
-7.13036895e-01 -3.60148817e-01 -1.18314791e+00 3.86652499e-01
7.52855480e-01 -6.13034904e-01 1.57678521e+00 -7.05270410e-01
-1.49042666e+00 4.56352234e-01 -4.49163944e-01 -7.66429722e-01
3.21556419e-01 -4.74087149e-01 -1.84092104e-01 -3.40159126e-02
4.52938646e-01 9.95886326e-02 5.55733681e-01 -1.21061385e+00
-8.65144193e-01 -3.42947036e-01 -1.79858908e-01 3.26588422e-01
-4.19189222e-02 4.98386063e-02 -3.31887156e-01 -6.10553980e-01
3.71617198e-01 -5.52669525e-01 -7.01430559e-01 -8.57348025e-01
-1.08445430e+00 -5.19580245e-01 4.62943166e-01 -7.73468733e-01
1.21999896e+00 -1.95406103e+00 4.08285230e-01 1.76946312e-01
3.22032273e-01 1.37618318e-01 -1.72147289e-01 4.07279283e-01
-1.38690863e-02 2.88506657e-01 -3.80681515e-01 -6.62115812e-01
-1.79343790e-01 3.68296176e-01 -4.19797063e-01 1.99945062e-01
9.08993304e-01 9.31784987e-01 -1.45271456e+00 -5.26524007e-01
-2.52341330e-01 3.65807891e-01 -2.88942426e-01 4.81376380e-01
-3.24470937e-01 6.34820044e-01 -8.03124666e-01 5.03774107e-01
4.97487754e-01 -4.87108231e-02 4.37703341e-01 -1.63682237e-01
1.75626054e-01 9.07302320e-01 -1.36826146e+00 1.40903246e+00
-3.14962894e-01 1.48774475e-01 3.79613608e-01 -1.12212801e+00
9.90241468e-01 5.27636647e-01 2.67041594e-01 -2.71544099e-01
2.61555582e-01 1.04403406e-01 -1.34750292e-01 -8.87129366e-01
4.88190472e-01 -3.56917590e-01 -3.78906697e-01 2.37656459e-01
3.39782685e-01 -2.93740958e-01 4.14144516e-01 4.73350078e-01
1.67467320e+00 3.50990854e-02 3.35121840e-01 1.40804444e-02
3.39707494e-01 -3.07639033e-01 1.34916985e+00 8.99437129e-01
-3.18783224e-01 5.32469213e-01 9.62713063e-01 4.88452055e-02
-4.79154021e-01 -8.43392551e-01 1.35180533e-01 8.00509691e-01
7.32321218e-02 -6.39374197e-01 -7.11454332e-01 -9.80671287e-01
-1.70205936e-01 5.61205864e-01 -3.76874179e-01 -2.25654423e-01
-6.17474556e-01 -9.98584926e-01 4.78507549e-01 4.31489676e-01
2.08063453e-01 -1.20831978e+00 -1.46743193e-01 4.26633656e-01
-6.81634009e-01 -1.39753914e+00 -4.64024812e-01 1.01937366e+00
-7.83104658e-01 -8.59652460e-01 -2.18583286e-01 -8.41698885e-01
4.90115434e-01 -4.82790247e-02 1.33721554e+00 2.87652779e-02
2.21015364e-01 -2.16731250e-01 -5.77821910e-01 -4.29020703e-01
-4.86561418e-01 4.13156152e-01 -4.00958024e-03 -6.03533424e-02
6.82924986e-01 -5.46832860e-01 -9.49563086e-02 -1.99371576e-01
-7.14365959e-01 -3.80117148e-01 7.41678953e-01 1.10684907e+00
7.10313499e-01 1.61216229e-01 1.01495540e+00 -1.37949848e+00
9.01086748e-01 -4.86733258e-01 -5.17533183e-01 5.11273462e-03
-2.33902559e-01 3.39557290e-01 5.51919043e-01 -7.61698335e-02
-1.27708590e+00 5.41434407e-01 -4.72129226e-01 2.08150521e-01
-3.01426053e-01 6.85765207e-01 -5.94359636e-01 7.44803011e-01
5.21127284e-01 -2.68985648e-02 -3.77548397e-01 -5.51402330e-01
4.44664299e-01 9.21108842e-01 4.05546784e-01 -5.32798469e-01
8.01777065e-01 1.37117639e-01 -3.58041555e-01 -8.96103919e-01
-1.48945701e+00 -6.93874538e-01 -9.60383594e-01 3.64115328e-01
1.01432014e+00 -1.08185625e+00 -7.62878776e-01 1.24421977e-02
-1.46204603e+00 -2.31351629e-01 -4.46315765e-01 4.23498154e-01
-4.41971183e-01 4.44108278e-01 -1.30873632e+00 -1.15058148e+00
-5.44561923e-01 -1.20370746e+00 1.18727291e+00 1.54710144e-01
-4.17527795e-01 -6.10777497e-01 4.03851047e-02 4.38267499e-01
-1.36693329e-01 -3.68423313e-02 6.65955067e-01 -8.37759018e-01
-7.06942737e-01 3.45893875e-02 6.00404758e-03 2.99623549e-01
2.37284526e-01 -3.49722832e-01 -1.15145028e+00 1.50519952e-01
1.60279661e-01 -6.15544319e-01 1.27025175e+00 3.17190379e-01
7.00608790e-01 -3.91423970e-01 -3.53324324e-01 3.10789585e-01
1.07390583e+00 8.95998674e-04 2.04646438e-01 2.32873440e-01
5.38417399e-01 8.28537524e-01 7.51432002e-01 3.62245530e-01
2.65178740e-01 2.87731856e-01 3.79508525e-01 3.89540941e-02
9.45546106e-02 -3.19588929e-01 2.93616503e-01 9.32176232e-01
2.08450615e-01 -2.54753143e-01 -5.69283307e-01 8.09234262e-01
-2.16130686e+00 -1.00588024e+00 -3.05302918e-01 1.54346621e+00
1.53978622e+00 3.21623355e-01 -2.10189167e-02 3.22314322e-01
7.74441302e-01 -7.47854449e-03 -3.85019988e-01 -2.86386937e-01
-1.77810282e-01 6.61872998e-02 2.54527569e-01 8.11919630e-01
-1.42272198e+00 1.26363516e+00 5.67717552e+00 6.89341903e-01
-5.19085288e-01 -6.30446747e-02 5.85753560e-01 9.35476571e-02
-2.33012155e-01 9.38661173e-02 -1.14893186e+00 2.98632264e-01
9.08405304e-01 9.67881903e-02 1.13133796e-01 6.65764093e-01
5.86419404e-01 -2.56552905e-01 -1.20024490e+00 5.10836363e-01
-2.99922824e-01 -9.48783159e-01 -1.70880884e-01 -3.03331971e-01
5.83535790e-01 -4.82950471e-02 -3.84255767e-01 4.12008315e-01
7.24465609e-01 -8.69372725e-01 5.24008572e-01 3.68848592e-01
1.26607120e-01 -9.82053339e-01 1.17999625e+00 5.40185511e-01
-8.98573339e-01 6.11674897e-02 -4.46113288e-01 -1.46741778e-01
5.42226553e-01 1.41888797e+00 -8.57939482e-01 6.94128573e-01
5.96079886e-01 8.73297513e-01 -1.19849421e-01 7.56384909e-01
-9.03517723e-01 8.70199919e-01 -2.34297216e-01 -1.82928741e-01
9.23606232e-02 -1.59088463e-01 4.83847350e-01 1.56080902e+00
-2.00752303e-01 4.11887854e-01 2.96369225e-01 8.18394721e-01
-4.08472747e-01 7.95842484e-02 -4.77199465e-01 1.05984852e-01
4.63166505e-01 1.32892621e+00 -6.97192073e-01 -2.55850852e-01
-4.15266901e-01 1.04543507e+00 7.58407235e-01 4.47894543e-01
-5.02371311e-01 -5.67373812e-01 6.91865027e-01 -3.76102239e-01
3.64438504e-01 1.22710143e-03 -5.74318409e-01 -1.30309188e+00
1.80668399e-01 -9.50705290e-01 4.09256667e-01 -3.17465097e-01
-1.62768674e+00 6.81889236e-01 -4.45270628e-01 -7.78861105e-01
-2.18379751e-01 -1.10050485e-01 -4.34147924e-01 9.63081539e-01
-1.70046949e+00 -7.14492083e-01 2.71111190e-01 6.55779913e-02
5.11366427e-01 2.21722662e-01 8.33764255e-01 3.03293824e-01
-9.15716529e-01 3.75365168e-01 -5.72100431e-02 6.32880807e-01
5.43042839e-01 -1.64647365e+00 5.84740996e-01 9.87604499e-01
2.66598195e-01 6.14986658e-01 9.41775262e-01 -9.64594781e-01
-9.77242887e-01 -1.24781895e+00 1.54341054e+00 -4.50584143e-01
7.14857280e-01 -6.30589247e-01 -9.94405508e-01 7.88113594e-01
8.86409953e-02 4.65564243e-02 6.21988654e-01 4.73218918e-01
-6.96043596e-02 -3.83653305e-02 -1.04198730e+00 5.10077298e-01
9.57365394e-01 -4.84735310e-01 -8.17117333e-01 4.58527297e-01
1.01002049e+00 -4.21894908e-01 -8.34765851e-01 2.49996230e-01
-2.49720678e-01 -5.19844353e-01 6.23997092e-01 -7.39916563e-01
5.15854955e-01 -1.74646571e-01 2.73459335e-03 -1.69332266e+00
-1.66280121e-01 -8.85590971e-01 -3.28919053e-01 1.73467982e+00
9.18955386e-01 1.14863310e-02 9.33909833e-01 7.16195524e-01
-3.87312919e-02 -8.05421650e-01 -1.06304359e+00 -5.10667801e-01
-6.46656528e-02 -4.74952519e-01 2.46649504e-01 6.98202074e-01
5.38265288e-01 1.22689533e+00 -5.18982649e-01 4.43947405e-01
4.33463216e-01 2.01074347e-01 3.73349637e-01 -1.11861241e+00
-5.36464155e-01 4.37877439e-02 -6.78889900e-02 -1.25966668e+00
4.75740612e-01 -8.37670088e-01 8.92002642e-01 -1.59181035e+00
1.97434485e-01 -2.68323004e-01 3.53128947e-02 5.65834761e-01
-7.07157075e-01 -2.25969493e-01 -7.69441053e-02 -8.66371170e-02
-8.37536573e-01 8.82475257e-01 9.02422547e-01 -3.13898534e-01
-3.66486073e-01 3.38457674e-01 -1.00728536e+00 7.79812634e-01
7.14376450e-01 -8.84759307e-01 -2.07825914e-01 -3.04436326e-01
2.81155586e-01 3.02983284e-01 -8.83845538e-02 -3.94924074e-01
8.50588903e-02 -8.83006454e-02 6.88148066e-02 -4.76022333e-01
2.56911188e-01 -5.55739641e-01 -6.53533757e-01 3.18263173e-01
-8.21602046e-01 -2.96872646e-01 -2.18053430e-01 8.87234688e-01
-3.67415518e-01 -5.77378094e-01 5.93166411e-01 -2.19677165e-01
-2.90717125e-01 -6.24034833e-03 -6.83981478e-01 4.61468339e-01
6.51725352e-01 6.19218409e-01 -1.23520069e-01 -5.28695047e-01
-1.30012035e+00 5.68735838e-01 -3.22624594e-01 2.51507759e-01
4.79756713e-01 -9.71406400e-01 -8.97944033e-01 -1.74506918e-01
-7.66896680e-02 5.93566000e-01 -1.98409468e-01 5.55397451e-01
1.92906991e-01 2.31216818e-01 6.44266605e-01 -2.33216465e-01
-1.28684092e+00 5.43173194e-01 2.09337369e-01 -7.53499091e-01
-6.97767615e-01 1.16165721e+00 -1.54762223e-01 -4.27099913e-01
5.60541689e-01 -6.03455424e-01 -3.63846868e-01 2.41176069e-01
5.86464524e-01 7.83251077e-02 3.33933622e-01 -5.46892405e-01
-2.08571136e-01 5.10712005e-02 -1.18728183e-01 2.49129422e-02
1.67871547e+00 -3.97182554e-01 -1.34065375e-01 5.08208215e-01
9.03076708e-01 -5.72490580e-02 -1.22881401e+00 -5.16009867e-01
8.01610351e-01 -3.35637480e-02 -7.66986832e-02 -6.13013923e-01
-9.48792696e-01 6.65455520e-01 -1.37287840e-01 4.06852514e-01
8.31836283e-01 2.50523180e-01 1.04958451e+00 7.98622906e-01
1.01801187e-01 -1.35411596e+00 -1.00950219e-01 8.70721817e-01
6.81369126e-01 -1.40255988e+00 -1.83949515e-01 -9.00136769e-01
-7.05316126e-01 8.28617632e-01 6.11729443e-01 4.96071093e-02
5.65179229e-01 7.73904860e-01 2.46650174e-01 -6.39039725e-02
-9.61647213e-01 -6.92177653e-01 -1.60818592e-01 8.38138402e-01
6.97158575e-01 -2.30750948e-01 -5.59652925e-01 1.17118084e+00
-1.42564312e-01 -2.13198200e-01 6.96635604e-01 7.47744381e-01
-5.37306786e-01 -1.25448120e+00 -1.25688881e-01 5.08656144e-01
-8.56912434e-01 -3.66023749e-01 -4.92270917e-01 6.54942766e-02
5.99111952e-02 1.49863422e+00 -4.11510199e-01 -3.06974322e-01
6.51008010e-01 1.53359652e-01 -1.37617037e-01 -1.15696633e+00
-9.53464746e-01 2.01447532e-01 5.16675532e-01 -4.24534291e-01
-3.99913907e-01 -8.16526413e-01 -1.42916489e+00 2.25423679e-01
-7.57067144e-01 7.09640920e-01 2.56891847e-01 1.31579447e+00
2.31510729e-01 9.03491080e-01 4.63531762e-01 -7.35550940e-01
-9.57897604e-01 -1.10617185e+00 -7.46008098e-01 5.29913545e-01
6.19546473e-01 -3.27088922e-01 -3.31450760e-01 7.43730962e-02] | [9.355069160461426, 8.720480918884277] |
2735c24e-9f37-4ff0-a959-a3729a3954e7 | lightweight-modeling-of-user-context | 2306.16029 | null | https://arxiv.org/abs/2306.16029v1 | https://arxiv.org/pdf/2306.16029v1.pdf | Lightweight Modeling of User Context Combining Physical and Virtual Sensor Data | The multitude of data generated by sensors available on users' mobile devices, combined with advances in machine learning techniques, support context-aware services in recognizing the current situation of a user (i.e., physical context) and optimizing the system's personalization features. However, context-awareness performances mainly depend on the accuracy of the context inference process, which is strictly tied to the availability of large-scale and labeled datasets. In this work, we present a framework developed to collect datasets containing heterogeneous sensing data derived from personal mobile devices. The framework has been used by 3 voluntary users for two weeks, generating a dataset with more than 36K samples and 1331 features. We also propose a lightweight approach to model the user context able to efficiently perform the entire reasoning process on the user mobile device. To this aim, we used six dimensionality reduction techniques in order to optimize the context classification. Experimental results on the generated dataset show that we achieve a 10x speed up and a feature reduction of more than 90% while keeping the accuracy loss less than 3%. | ['Pan Hui', 'Franca Delmastro', 'Dimitris Chatzopoulos', 'Mattia Giovanni Campana'] | 2023-06-28 | null | null | null | null | ['dimensionality-reduction'] | ['methodology'] | [ 4.22535270e-01 -7.96099007e-02 -2.48049662e-01 -5.80824018e-01
-5.37152946e-01 -3.85610431e-01 4.64151442e-01 4.86882657e-01
-3.00030142e-01 6.23499811e-01 2.52124101e-01 -1.39806792e-01
-2.93537527e-01 -9.15786147e-01 -2.45889828e-01 -4.77001876e-01
-5.30998297e-02 2.49938428e-01 -7.32011208e-03 -1.14383432e-03
2.86071151e-01 3.96009326e-01 -2.16822839e+00 4.80775714e-01
8.00017297e-01 1.46390796e+00 3.96367460e-01 6.93739891e-01
-4.47435603e-02 5.32282054e-01 -3.52569610e-01 1.08753540e-01
-1.14285730e-01 -2.92297509e-02 -6.29041135e-01 -1.40635585e-02
7.67504424e-02 -1.17499940e-01 3.80903840e-01 6.12861991e-01
4.94454861e-01 5.31922095e-02 2.19081089e-01 -8.92892599e-01
-9.58082750e-02 3.00011724e-01 1.49838746e-01 3.91075648e-02
8.24585378e-01 -2.47016296e-01 8.87908936e-01 -7.45453715e-01
4.64997083e-01 7.52715230e-01 6.11016989e-01 4.66962218e-01
-9.84600961e-01 -2.64699906e-01 2.31343240e-01 4.43822771e-01
-1.68785048e+00 -5.74526429e-01 8.29890311e-01 -3.02366316e-01
1.02843022e+00 8.64411414e-01 9.32367325e-01 1.18870091e+00
-7.64108077e-02 4.29343253e-01 1.01978230e+00 -6.74357176e-01
8.52155447e-01 5.46342611e-01 2.42188141e-01 2.98879147e-01
3.50038409e-01 -3.79090488e-01 -5.31819940e-01 -5.66641927e-01
-4.44395803e-02 3.73170406e-01 1.74665034e-01 -3.29677075e-01
-8.83866787e-01 2.45866790e-01 -2.52277236e-02 8.73771906e-01
-6.23716354e-01 -2.37032264e-01 1.07424445e-01 -1.84472930e-02
2.47231945e-01 1.26164362e-01 -5.76273024e-01 -6.14053607e-01
-6.69444978e-01 1.23568401e-01 1.07288051e+00 9.32125270e-01
9.65144455e-01 -4.74886149e-01 -3.32124233e-02 5.99382341e-01
2.78269678e-01 7.07866967e-01 7.11372316e-01 -7.49155939e-01
5.02880573e-01 1.12200069e+00 3.50052029e-01 -1.20508778e+00
-6.74758434e-01 -1.20062411e-01 -7.58875310e-01 -4.15578961e-01
4.23495136e-02 -1.89125791e-01 -2.46084407e-01 1.52312720e+00
6.01019979e-01 3.03213179e-01 -1.07116826e-01 4.76928562e-01
5.66118993e-02 1.83127090e-01 1.71087876e-01 -3.85218829e-01
1.44679272e+00 -2.72155285e-01 -7.97556520e-01 -1.92795292e-01
6.91929042e-01 -4.25083578e-01 1.40497029e+00 5.52075446e-01
-6.60094976e-01 -6.36325181e-01 -1.13172734e+00 2.78890312e-01
-7.30009139e-01 1.38742447e-01 6.96623802e-01 1.19641316e+00
-6.86670482e-01 4.41528291e-01 -8.58049929e-01 -6.52672768e-01
3.96704644e-01 5.73183775e-01 -1.55385181e-01 8.86768922e-02
-1.00665855e+00 5.53200603e-01 3.58208239e-01 -3.40021215e-02
-4.09921587e-01 -5.04824281e-01 -4.23628509e-01 2.24856630e-01
3.70205671e-01 -5.19329250e-01 8.63853514e-01 -7.09859133e-01
-1.41779339e+00 6.54884100e-01 -4.27748144e-01 -4.78613585e-01
2.39918292e-01 -4.08239692e-01 -9.93342161e-01 -8.02413374e-02
-1.66153133e-01 -1.41960263e-01 6.49479508e-01 -1.04560828e+00
-1.08443904e+00 -8.69098008e-01 -2.91240327e-02 1.12286262e-01
-9.80099559e-01 -2.69243777e-01 -4.82442379e-01 -1.03104256e-01
7.97899161e-03 -1.19082403e+00 -1.69367671e-01 -5.94108045e-01
-1.46222323e-01 1.41797796e-01 1.09306681e+00 -5.70643723e-01
1.78984153e+00 -2.13122821e+00 -2.76255995e-01 6.17016971e-01
-1.74831495e-01 4.43381041e-01 3.37961793e-01 2.71503597e-01
3.37024629e-01 1.49015486e-02 -2.46298820e-01 -6.58334315e-01
6.62446255e-03 5.55203974e-01 -1.09593123e-01 9.08851251e-02
-3.04779798e-01 7.24599063e-01 -7.49092937e-01 -4.09682244e-01
4.96484607e-01 6.56679451e-01 -5.92856407e-01 2.64855832e-01
-1.52982801e-01 4.37688589e-01 -7.48857796e-01 7.96853185e-01
3.92339468e-01 -1.45036384e-01 8.36780787e-01 -9.90534388e-03
1.19526878e-01 2.19042927e-01 -1.49056923e+00 1.92567301e+00
-9.88406181e-01 3.18296738e-02 5.15662655e-02 -7.10125625e-01
8.34641933e-01 2.53396958e-01 8.38160276e-01 -7.82199085e-01
1.28917992e-01 1.16720416e-01 -6.82217777e-01 -7.56682754e-01
5.63428581e-01 3.54048103e-01 -3.08022499e-01 5.49688876e-01
-4.37106580e-01 4.53782439e-01 -2.63051456e-03 -3.13614964e-01
1.05048561e+00 8.32983330e-02 5.50755322e-01 -2.34968245e-01
9.77290809e-01 -3.54769975e-01 5.01680315e-01 6.22324407e-01
-2.24834025e-01 3.31411660e-02 1.81629043e-02 -6.20381594e-01
-4.56791461e-01 -6.52736187e-01 -4.47161086e-02 1.21198297e+00
1.01024084e-01 -6.89290822e-01 -9.62679267e-01 -6.61572158e-01
-6.52963109e-03 7.26718128e-01 -6.28545046e-01 -9.16153565e-02
-4.45190787e-01 -8.63504350e-01 2.77166843e-01 2.99342066e-01
5.15468299e-01 -8.49595129e-01 -1.05525362e+00 2.54232675e-01
-3.57760340e-01 -1.22971117e+00 7.15776309e-02 -1.92078520e-02
-1.02452791e+00 -1.07225060e+00 1.98742017e-01 -1.94800690e-01
4.26012546e-01 1.38951734e-01 9.55691755e-01 5.95691241e-02
-2.76640803e-01 6.20849311e-01 -3.57795924e-01 -2.47491077e-01
-8.83381367e-02 4.65534806e-01 3.08775187e-01 4.86837596e-01
7.97231078e-01 -9.41104949e-01 -5.67492604e-01 3.04753691e-01
-8.48596513e-01 -3.42837811e-01 3.91494960e-01 1.75130308e-01
8.69767010e-01 2.90359557e-01 6.14868462e-01 -9.71873045e-01
5.85333467e-01 -6.36582971e-01 -3.48351747e-01 4.37547624e-01
-1.24626243e+00 -3.47643942e-02 5.34299791e-01 -3.90202820e-01
-1.22073960e+00 4.22547311e-01 -2.56614745e-01 1.91423684e-01
-6.16687298e-01 1.37013927e-01 -7.80406475e-01 1.05291262e-01
7.21089184e-01 8.85306224e-02 -5.39921939e-01 -7.04911828e-01
4.20294225e-01 1.19137120e+00 2.90432423e-01 -5.27031958e-01
3.08601022e-01 6.70012116e-01 3.29978727e-02 -9.93168056e-01
-7.43481219e-01 -6.15238249e-01 -6.55539334e-01 -1.72437474e-01
6.50189698e-01 -7.82772362e-01 -8.69327307e-01 1.50359631e-01
-4.61438179e-01 -1.29187942e-01 -3.64121884e-01 2.40621328e-01
-5.03557980e-01 1.77654803e-01 1.03086613e-01 -1.17431784e+00
-5.98653913e-01 -7.00373948e-01 1.08023787e+00 1.05589606e-01
-3.76697421e-01 -8.68549168e-01 9.24704596e-02 5.53202629e-01
6.38446271e-01 2.10673466e-01 5.19181013e-01 -6.15448594e-01
-4.62359428e-01 -3.92741680e-01 1.27275199e-01 9.33690146e-02
5.90000987e-01 -5.00846446e-01 -1.50531662e+00 -2.10797004e-02
2.87422210e-01 2.96714813e-01 1.37220845e-01 -9.58106667e-02
1.41477585e+00 -3.62774760e-01 -4.95354474e-01 3.22467059e-01
1.43860579e+00 3.00851494e-01 6.14715457e-01 1.85645610e-01
7.60457098e-01 3.87237251e-01 9.16515291e-01 7.97457814e-01
6.49017215e-01 9.58354354e-01 4.82258022e-01 4.78554934e-01
2.05018833e-01 -1.47631124e-01 1.63235143e-01 5.59687376e-01
-1.67551130e-01 4.53695282e-02 -8.10313165e-01 4.58480984e-01
-1.97155881e+00 -8.76173079e-01 -1.30076529e-02 2.38172150e+00
4.94589984e-01 1.90948561e-01 1.27650961e-01 6.57810986e-01
5.17526031e-01 -3.99855971e-02 -5.52108824e-01 -3.74415576e-01
2.28614166e-01 1.67070106e-02 1.61494732e-01 6.28550470e-01
-9.68779385e-01 5.77368677e-01 5.74191380e+00 3.61175895e-01
-1.06580949e+00 3.28079343e-01 3.53465348e-01 -3.37243199e-01
-3.04814786e-01 -1.77134469e-01 -8.70404959e-01 8.31842959e-01
1.40977573e+00 1.83559492e-01 6.52693570e-01 1.23521781e+00
4.07666504e-01 -2.60978043e-01 -9.71889019e-01 1.26288128e+00
4.80114594e-02 -9.88404036e-01 -2.24137917e-01 3.73155206e-01
5.66664457e-01 -1.36853069e-01 -1.18598424e-01 1.97324663e-01
-5.41442037e-01 -4.05139744e-01 4.20968413e-01 8.21231425e-01
5.41195750e-01 -8.78233731e-01 4.82764065e-01 5.65734029e-01
-1.28791428e+00 -6.25808001e-01 1.12109855e-01 -3.60581219e-01
-6.71404600e-02 1.20292377e+00 -9.35995936e-01 4.75763112e-01
8.17771196e-01 3.52251649e-01 -7.77056694e-01 3.14090490e-01
1.50712579e-01 3.82782727e-01 -5.99629939e-01 -1.81205779e-01
-4.33666408e-01 -1.77683115e-01 2.27449760e-01 1.18829501e+00
4.85323608e-01 -2.33890694e-02 -5.10955974e-02 3.88974547e-01
1.87825173e-01 3.31095487e-01 -6.04291260e-01 1.55956551e-01
6.36039078e-01 1.39866400e+00 -4.10716027e-01 -4.32066232e-01
-1.62233979e-01 9.73140836e-01 -9.57314577e-03 -5.50493784e-02
-6.51892960e-01 -7.14904517e-02 7.87963450e-01 2.70703942e-01
2.79003561e-01 -2.55503744e-01 -5.24505973e-01 -1.17833149e+00
5.36158442e-01 -5.83644629e-01 2.91749686e-01 -3.89703363e-01
-7.25800395e-01 3.79624695e-01 -2.37101659e-01 -9.52374458e-01
-6.01022899e-01 -3.13556969e-01 -2.53633112e-01 7.33291864e-01
-1.32591736e+00 -1.12122893e+00 -6.91634715e-01 8.33945453e-01
8.94612968e-02 1.64245013e-02 1.38740325e+00 7.08360732e-01
-5.12008727e-01 3.65125149e-01 -8.23951289e-02 -5.71487069e-01
3.36990118e-01 -1.09146023e+00 2.26926073e-01 7.34483778e-01
1.48331612e-01 7.51333237e-01 4.54681963e-01 -6.00268662e-01
-1.73470128e+00 -1.03481996e+00 1.31698704e+00 -7.04377592e-01
3.72514695e-01 -5.44006467e-01 -6.51622474e-01 4.36734289e-01
-4.43144888e-01 1.59903333e-01 1.19163024e+00 4.57461894e-01
-4.19536568e-02 -6.66710258e-01 -1.63075817e+00 4.25456613e-01
1.33780026e+00 -8.10749114e-01 -3.46039742e-01 3.05536129e-02
3.09090823e-01 -1.30860105e-01 -1.19067132e+00 3.14914882e-01
8.18867207e-01 -1.22860491e+00 8.97912502e-01 -2.12875888e-01
-4.25043404e-01 -3.62718731e-01 -6.29596472e-01 -5.90627193e-01
-3.79445055e-03 -4.60823029e-01 -9.69454587e-01 1.36894393e+00
1.92091182e-01 -5.68315685e-01 8.84129047e-01 1.06281555e+00
4.08440679e-01 -7.00101614e-01 -8.40413868e-01 -3.41504812e-01
-8.77250314e-01 -1.05386996e+00 1.24675739e+00 8.06596816e-01
1.99708492e-02 1.58055767e-01 -2.70367324e-01 3.57606322e-01
4.40095395e-01 2.94655174e-01 6.69766784e-01 -1.46588993e+00
-3.04112941e-01 3.89461704e-02 -4.27962303e-01 -5.31616092e-01
-1.40863299e-01 -4.86312658e-01 -4.11243320e-01 -1.21515977e+00
-1.35967523e-01 -7.72165596e-01 -5.85730016e-01 3.92052293e-01
-1.07877189e-02 7.79504478e-02 -1.86110154e-01 2.25715041e-02
-8.20705533e-01 6.24558628e-02 3.82835329e-01 1.06682010e-01
-6.94121957e-01 3.48745406e-01 -8.88977349e-01 6.48210764e-01
1.08323717e+00 -1.72836944e-01 -6.53380930e-01 -1.75308436e-01
5.63171804e-01 -2.70941228e-01 2.09478572e-01 -1.40461934e+00
4.71973419e-02 -2.43883237e-01 3.86573285e-01 -3.85482788e-01
5.79770625e-01 -1.54863691e+00 5.25489688e-01 3.81441951e-01
-1.43110737e-01 -3.86910349e-01 -2.82665696e-02 6.00232065e-01
1.70977116e-01 6.80271685e-02 2.89016396e-01 1.11752950e-01
-8.60671401e-01 1.92143489e-02 -2.52952069e-01 -4.65247631e-01
1.13310552e+00 -9.72686335e-03 1.92481279e-01 -3.18369605e-02
-1.08223832e+00 -2.45071158e-01 5.16096890e-01 6.66353583e-01
2.90991455e-01 -1.35665464e+00 9.90561470e-02 4.58699673e-01
4.23881471e-01 -3.18806857e-01 4.74037588e-01 4.17583644e-01
5.98169416e-02 5.45609295e-01 5.40704802e-02 -6.12377584e-01
-1.37469018e+00 6.96873307e-01 1.92650720e-01 -2.90538609e-01
-3.69288057e-01 -1.07049495e-02 -8.10249567e-01 -3.22200030e-01
2.75030345e-01 -5.36268175e-01 -3.77350569e-01 2.84232944e-01
9.98216689e-01 7.18922198e-01 6.19802117e-01 -4.86724555e-01
-6.62160456e-01 5.63640058e-01 3.51872087e-01 6.57086521e-02
1.17140567e+00 -6.63079619e-01 1.99520081e-01 5.55365443e-01
1.06337857e+00 3.75828862e-01 -8.65338683e-01 -2.44263262e-01
3.50002438e-01 -4.77190852e-01 -9.75686163e-02 -7.98599243e-01
-6.24108136e-01 4.30607915e-01 1.34168768e+00 5.97531855e-01
1.46141148e+00 -2.96373397e-01 7.89759576e-01 7.05284894e-01
8.46966684e-01 -1.48935378e+00 -5.23123324e-01 2.08301380e-01
3.27173710e-01 -1.32792234e+00 1.41759845e-03 -2.02257454e-01
-4.72956866e-01 5.78983009e-01 1.22711197e-01 3.40998292e-01
8.80938351e-01 1.21757261e-01 -1.10993706e-01 2.17812415e-02
-5.97344577e-01 -3.28639477e-01 1.33147866e-01 7.25917637e-01
1.68526843e-01 5.14446259e-01 -1.39148921e-01 9.28268850e-01
-1.21450067e-01 4.20470208e-01 -3.89709976e-03 1.18585670e+00
-5.21100402e-01 -1.44175482e+00 -4.35680062e-01 3.55084360e-01
-4.11735177e-01 3.25799793e-01 -2.31157154e-01 1.54848382e-01
5.99759817e-01 1.46498787e+00 -1.97324455e-01 -6.77297592e-01
5.70831537e-01 3.10081422e-01 2.10526064e-01 -4.13383275e-01
-5.56065261e-01 -4.22069430e-01 1.50802344e-01 -1.07365727e+00
-5.69804430e-01 -8.25333774e-01 -1.18453360e+00 -9.48337838e-02
1.99484229e-01 1.14685874e-02 1.27351284e+00 1.12418497e+00
9.72585261e-01 4.25437361e-01 9.00443435e-01 -7.04002500e-01
-1.10360704e-01 -7.55307436e-01 -4.21087831e-01 5.08039117e-01
1.62621632e-01 -4.19059902e-01 -1.08371131e-01 1.84684113e-01] | [7.490678787231445, 1.390409231185913] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.